diff options
102 files changed, 4604 insertions, 2561 deletions
diff --git a/Documentation/device-mapper/dm-dust.txt b/Documentation/device-mapper/dm-dust.txt new file mode 100644 index 000000000000..954d402a1f6a --- /dev/null +++ b/Documentation/device-mapper/dm-dust.txt @@ -0,0 +1,272 @@ +dm-dust +======= + +This target emulates the behavior of bad sectors at arbitrary +locations, and the ability to enable the emulation of the failures +at an arbitrary time. + +This target behaves similarly to a linear target. At a given time, +the user can send a message to the target to start failing read +requests on specific blocks (to emulate the behavior of a hard disk +drive with bad sectors). + +When the failure behavior is enabled (i.e.: when the output of +"dmsetup status" displays "fail_read_on_bad_block"), reads of blocks +in the "bad block list" will fail with EIO ("Input/output error"). + +Writes of blocks in the "bad block list will result in the following: + +1. Remove the block from the "bad block list". +2. Successfully complete the write. + +This emulates the "remapped sector" behavior of a drive with bad +sectors. + +Normally, a drive that is encountering bad sectors will most likely +encounter more bad sectors, at an unknown time or location. +With dm-dust, the user can use the "addbadblock" and "removebadblock" +messages to add arbitrary bad blocks at new locations, and the +"enable" and "disable" messages to modulate the state of whether the +configured "bad blocks" will be treated as bad, or bypassed. +This allows the pre-writing of test data and metadata prior to +simulating a "failure" event where bad sectors start to appear. + +Table parameters: +----------------- +<device_path> <offset> <blksz> + +Mandatory parameters: + <device_path>: path to the block device. + <offset>: offset to data area from start of device_path + <blksz>: block size in bytes + (minimum 512, maximum 1073741824, must be a power of 2) + +Usage instructions: +------------------- + +First, find the size (in 512-byte sectors) of the device to be used: + +$ sudo blockdev --getsz /dev/vdb1 +33552384 + +Create the dm-dust device: +(For a device with a block size of 512 bytes) +$ sudo dmsetup create dust1 --table '0 33552384 dust /dev/vdb1 0 512' + +(For a device with a block size of 4096 bytes) +$ sudo dmsetup create dust1 --table '0 33552384 dust /dev/vdb1 0 4096' + +Check the status of the read behavior ("bypass" indicates that all I/O +will be passed through to the underlying device): +$ sudo dmsetup status dust1 +0 33552384 dust 252:17 bypass + +$ sudo dd if=/dev/mapper/dust1 of=/dev/null bs=512 count=128 iflag=direct +128+0 records in +128+0 records out + +$ sudo dd if=/dev/zero of=/dev/mapper/dust1 bs=512 count=128 oflag=direct +128+0 records in +128+0 records out + +Adding and removing bad blocks: +------------------------------- + +At any time (i.e.: whether the device has the "bad block" emulation +enabled or disabled), bad blocks may be added or removed from the +device via the "addbadblock" and "removebadblock" messages: + +$ sudo dmsetup message dust1 0 addbadblock 60 +kernel: device-mapper: dust: badblock added at block 60 + +$ sudo dmsetup message dust1 0 addbadblock 67 +kernel: device-mapper: dust: badblock added at block 67 + +$ sudo dmsetup message dust1 0 addbadblock 72 +kernel: device-mapper: dust: badblock added at block 72 + +These bad blocks will be stored in the "bad block list". +While the device is in "bypass" mode, reads and writes will succeed: + +$ sudo dmsetup status dust1 +0 33552384 dust 252:17 bypass + +Enabling block read failures: +----------------------------- + +To enable the "fail read on bad block" behavior, send the "enable" message: + +$ sudo dmsetup message dust1 0 enable +kernel: device-mapper: dust: enabling read failures on bad sectors + +$ sudo dmsetup status dust1 +0 33552384 dust 252:17 fail_read_on_bad_block + +With the device in "fail read on bad block" mode, attempting to read a +block will encounter an "Input/output error": + +$ sudo dd if=/dev/mapper/dust1 of=/dev/null bs=512 count=1 skip=67 iflag=direct +dd: error reading '/dev/mapper/dust1': Input/output error +0+0 records in +0+0 records out +0 bytes copied, 0.00040651 s, 0.0 kB/s + +...and writing to the bad blocks will remove the blocks from the list, +therefore emulating the "remap" behavior of hard disk drives: + +$ sudo dd if=/dev/zero of=/dev/mapper/dust1 bs=512 count=128 oflag=direct +128+0 records in +128+0 records out + +kernel: device-mapper: dust: block 60 removed from badblocklist by write +kernel: device-mapper: dust: block 67 removed from badblocklist by write +kernel: device-mapper: dust: block 72 removed from badblocklist by write +kernel: device-mapper: dust: block 87 removed from badblocklist by write + +Bad block add/remove error handling: +------------------------------------ + +Attempting to add a bad block that already exists in the list will +result in an "Invalid argument" error, as well as a helpful message: + +$ sudo dmsetup message dust1 0 addbadblock 88 +device-mapper: message ioctl on dust1 failed: Invalid argument +kernel: device-mapper: dust: block 88 already in badblocklist + +Attempting to remove a bad block that doesn't exist in the list will +result in an "Invalid argument" error, as well as a helpful message: + +$ sudo dmsetup message dust1 0 removebadblock 87 +device-mapper: message ioctl on dust1 failed: Invalid argument +kernel: device-mapper: dust: block 87 not found in badblocklist + +Counting the number of bad blocks in the bad block list: +-------------------------------------------------------- + +To count the number of bad blocks configured in the device, run the +following message command: + +$ sudo dmsetup message dust1 0 countbadblocks + +A message will print with the number of bad blocks currently +configured on the device: + +kernel: device-mapper: dust: countbadblocks: 895 badblock(s) found + +Querying for specific bad blocks: +--------------------------------- + +To find out if a specific block is in the bad block list, run the +following message command: + +$ sudo dmsetup message dust1 0 queryblock 72 + +The following message will print if the block is in the list: +device-mapper: dust: queryblock: block 72 found in badblocklist + +The following message will print if the block is in the list: +device-mapper: dust: queryblock: block 72 not found in badblocklist + +The "queryblock" message command will work in both the "enabled" +and "disabled" modes, allowing the verification of whether a block +will be treated as "bad" without having to issue I/O to the device, +or having to "enable" the bad block emulation. + +Clearing the bad block list: +---------------------------- + +To clear the bad block list (without needing to individually run +a "removebadblock" message command for every block), run the +following message command: + +$ sudo dmsetup message dust1 0 clearbadblocks + +After clearing the bad block list, the following message will appear: + +kernel: device-mapper: dust: clearbadblocks: badblocks cleared + +If there were no bad blocks to clear, the following message will +appear: + +kernel: device-mapper: dust: clearbadblocks: no badblocks found + +Message commands list: +---------------------- + +Below is a list of the messages that can be sent to a dust device: + +Operations on blocks (requires a <blknum> argument): + +addbadblock <blknum> +queryblock <blknum> +removebadblock <blknum> + +...where <blknum> is a block number within range of the device + (corresponding to the block size of the device.) + +Single argument message commands: + +countbadblocks +clearbadblocks +disable +enable +quiet + +Device removal: +--------------- + +When finished, remove the device via the "dmsetup remove" command: + +$ sudo dmsetup remove dust1 + +Quiet mode: +----------- + +On test runs with many bad blocks, it may be desirable to avoid +excessive logging (from bad blocks added, removed, or "remapped"). +This can be done by enabling "quiet mode" via the following message: + +$ sudo dmsetup message dust1 0 quiet + +This will suppress log messages from add / remove / removed by write +operations. Log messages from "countbadblocks" or "queryblock" +message commands will still print in quiet mode. + +The status of quiet mode can be seen by running "dmsetup status": + +$ sudo dmsetup status dust1 +0 33552384 dust 252:17 fail_read_on_bad_block quiet + +To disable quiet mode, send the "quiet" message again: + +$ sudo dmsetup message dust1 0 quiet + +$ sudo dmsetup status dust1 +0 33552384 dust 252:17 fail_read_on_bad_block verbose + +(The presence of "verbose" indicates normal logging.) + +"Why not...?" +------------- + +scsi_debug has a "medium error" mode that can fail reads on one +specified sector (sector 0x1234, hardcoded in the source code), but +it uses RAM for the persistent storage, which drastically decreases +the potential device size. + +dm-flakey fails all I/O from all block locations at a specified time +frequency, and not a given point in time. + +When a bad sector occurs on a hard disk drive, reads to that sector +are failed by the device, usually resulting in an error code of EIO +("I/O error") or ENODATA ("No data available"). However, a write to +the sector may succeed, and result in the sector becoming readable +after the device controller no longer experiences errors reading the +sector (or after a reallocation of the sector). However, there may +be bad sectors that occur on the device in the future, in a different, +unpredictable location. + +This target seeks to provide a device that can exhibit the behavior +of a bad sector at a known sector location, at a known time, based +on a large storage device (at least tens of gigabytes, not occupying +system memory). diff --git a/Documentation/device-mapper/dm-integrity.txt b/Documentation/device-mapper/dm-integrity.txt index 297251b0d2d5..d63d78ffeb73 100644 --- a/Documentation/device-mapper/dm-integrity.txt +++ b/Documentation/device-mapper/dm-integrity.txt @@ -21,6 +21,13 @@ mode it calculates and verifies the integrity tag internally. In this mode, the dm-integrity target can be used to detect silent data corruption on the disk or in the I/O path. +There's an alternate mode of operation where dm-integrity uses bitmap +instead of a journal. If a bit in the bitmap is 1, the corresponding +region's data and integrity tags are not synchronized - if the machine +crashes, the unsynchronized regions will be recalculated. The bitmap mode +is faster than the journal mode, because we don't have to write the data +twice, but it is also less reliable, because if data corruption happens +when the machine crashes, it may not be detected. When loading the target for the first time, the kernel driver will format the device. But it will only format the device if the superblock contains @@ -59,6 +66,10 @@ Target arguments: either both data and tag or none of them are written. The journaled mode degrades write throughput twice because the data have to be written twice. + B - bitmap mode - data and metadata are written without any + synchronization, the driver maintains a bitmap of dirty + regions where data and metadata don't match. This mode can + only be used with internal hash. R - recovery mode - in this mode, journal is not replayed, checksums are not checked and writes to the device are not allowed. This mode is useful for data recovery if the @@ -79,6 +90,10 @@ interleave_sectors:number a power of two. If the device is already formatted, the value from the superblock is used. +meta_device:device + Don't interleave the data and metadata on on device. Use a + separate device for metadata. + buffer_sectors:number The number of sectors in one buffer. The value is rounded down to a power of two. @@ -146,6 +161,15 @@ block_size:number Supported values are 512, 1024, 2048 and 4096 bytes. If not specified the default block size is 512 bytes. +sectors_per_bit:number + In the bitmap mode, this parameter specifies the number of + 512-byte sectors that corresponds to one bitmap bit. + +bitmap_flush_interval:number + The bitmap flush interval in milliseconds. The metadata buffers + are synchronized when this interval expires. + + The journal mode (D/J), buffer_sectors, journal_watermark, commit_time can be changed when reloading the target (load an inactive table and swap the tables with suspend and resume). The other arguments should not be changed @@ -167,7 +191,13 @@ The layout of the formatted block device: provides (i.e. the size of the device minus the size of all metadata and padding). The user of this target should not send bios that access data beyond the "provided data sectors" limit. - * flags - a flag is set if journal_mac is used + * flags + SB_FLAG_HAVE_JOURNAL_MAC - a flag is set if journal_mac is used + SB_FLAG_RECALCULATING - recalculating is in progress + SB_FLAG_DIRTY_BITMAP - journal area contains the bitmap of dirty + blocks + * log2(sectors per block) + * a position where recalculating finished * journal The journal is divided into sections, each section contains: * metadata area (4kiB), it contains journal entries diff --git a/Documentation/networking/rxrpc.txt b/Documentation/networking/rxrpc.txt index cd7303d7fa25..180e07d956a7 100644 --- a/Documentation/networking/rxrpc.txt +++ b/Documentation/networking/rxrpc.txt @@ -796,7 +796,9 @@ The kernel interface functions are as follows: s64 tx_total_len, gfp_t gfp, rxrpc_notify_rx_t notify_rx, - bool upgrade); + bool upgrade, + bool intr, + unsigned int debug_id); This allocates the infrastructure to make a new RxRPC call and assigns call and connection numbers. The call will be made on the UDP port that @@ -824,6 +826,13 @@ The kernel interface functions are as follows: the server upgrade the service to a better one. The resultant service ID is returned by rxrpc_kernel_recv_data(). + intr should be set to true if the call should be interruptible. If this + is not set, this function may not return until a channel has been + allocated; if it is set, the function may return -ERESTARTSYS. + + debug_id is the call debugging ID to be used for tracing. This can be + obtained by atomically incrementing rxrpc_debug_id. + If this function is successful, an opaque reference to the RxRPC call is returned. The caller now holds a reference on this and it must be properly ended. @@ -1056,6 +1065,16 @@ The kernel interface functions are as follows: This value can be used to determine if the remote client has been restarted as it shouldn't change otherwise. + (*) Set the maxmimum lifespan on a call. + + void rxrpc_kernel_set_max_life(struct socket *sock, + struct rxrpc_call *call, + unsigned long hard_timeout) + + This sets the maximum lifespan on a call to hard_timeout (which is in + jiffies). In the event of the timeout occurring, the call will be + aborted and -ETIME or -ETIMEDOUT will be returned. + ======================= CONFIGURABLE PARAMETERS diff --git a/arch/arm/configs/mini2440_defconfig b/arch/arm/configs/mini2440_defconfig index 8b0f7c4c3f09..7d26ca0b1302 100644 --- a/arch/arm/configs/mini2440_defconfig +++ b/arch/arm/configs/mini2440_defconfig @@ -152,7 +152,7 @@ CONFIG_SPI_S3C24XX=y CONFIG_SPI_SPIDEV=y CONFIG_GPIO_SYSFS=y CONFIG_SENSORS_LM75=y -CONFIG_THERMAL=m +CONFIG_THERMAL=y CONFIG_WATCHDOG=y CONFIG_S3C2410_WATCHDOG=y CONFIG_FB=y diff --git a/arch/arm/configs/pxa_defconfig b/arch/arm/configs/pxa_defconfig index f6d24d762a7f..07ebbdce3645 100644 --- a/arch/arm/configs/pxa_defconfig +++ b/arch/arm/configs/pxa_defconfig @@ -387,7 +387,7 @@ CONFIG_SENSORS_LM75=m CONFIG_SENSORS_LM90=m CONFIG_SENSORS_LM95245=m CONFIG_SENSORS_NTC_THERMISTOR=m -CONFIG_THERMAL=m +CONFIG_THERMAL=y CONFIG_WATCHDOG=y CONFIG_XILINX_WATCHDOG=m CONFIG_SA1100_WATCHDOG=m diff --git a/arch/mips/configs/ip22_defconfig b/arch/mips/configs/ip22_defconfig index ff40fbc2f439..21a1168ae301 100644 --- a/arch/mips/configs/ip22_defconfig +++ b/arch/mips/configs/ip22_defconfig @@ -228,7 +228,7 @@ CONFIG_SERIAL_IP22_ZILOG=m # CONFIG_HW_RANDOM is not set CONFIG_RAW_DRIVER=m # CONFIG_HWMON is not set -CONFIG_THERMAL=m +CONFIG_THERMAL=y CONFIG_WATCHDOG=y CONFIG_INDYDOG=m # CONFIG_VGA_CONSOLE is not set diff --git a/arch/mips/configs/ip27_defconfig b/arch/mips/configs/ip27_defconfig index 81c47e18131b..54db5dedf776 100644 --- a/arch/mips/configs/ip27_defconfig +++ b/arch/mips/configs/ip27_defconfig @@ -271,7 +271,7 @@ CONFIG_I2C_PARPORT_LIGHT=m CONFIG_I2C_TAOS_EVM=m CONFIG_I2C_STUB=m # CONFIG_HWMON is not set -CONFIG_THERMAL=m +CONFIG_THERMAL=y CONFIG_MFD_PCF50633=m CONFIG_PCF50633_ADC=m CONFIG_PCF50633_GPIO=m diff --git a/arch/unicore32/configs/unicore32_defconfig b/arch/unicore32/configs/unicore32_defconfig index aebd01fc28e5..360cc9abcdb0 100644 --- a/arch/unicore32/configs/unicore32_defconfig +++ b/arch/unicore32/configs/unicore32_defconfig @@ -119,7 +119,7 @@ CONFIG_I2C_PUV3=y # Hardware Monitoring support #CONFIG_SENSORS_LM75=m # Generic Thermal sysfs driver -#CONFIG_THERMAL=m +#CONFIG_THERMAL=y #CONFIG_THERMAL_HWMON=y # Multimedia support diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index 2210c1b9491b..e5009a34f9c2 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -934,7 +934,7 @@ static struct rbd_client *rbd_get_client(struct ceph_options *ceph_opts) struct rbd_client *rbdc; int ret; - mutex_lock_nested(&client_mutex, SINGLE_DEPTH_NESTING); + mutex_lock(&client_mutex); rbdc = rbd_client_find(ceph_opts); if (rbdc) { ceph_destroy_options(ceph_opts); @@ -1326,7 +1326,7 @@ static void rbd_obj_zero_range(struct rbd_obj_request *obj_req, u32 off, zero_bvecs(&obj_req->bvec_pos, off, bytes); break; default: - rbd_assert(0); + BUG(); } } @@ -1581,7 +1581,7 @@ static void rbd_obj_request_destroy(struct kref *kref) kfree(obj_request->bvec_pos.bvecs); break; default: - rbd_assert(0); + BUG(); } kfree(obj_request->img_extents); @@ -1781,7 +1781,7 @@ static void rbd_osd_req_setup_data(struct rbd_obj_request *obj_req, u32 which) &obj_req->bvec_pos); break; default: - rbd_assert(0); + BUG(); } } @@ -2036,7 +2036,7 @@ static int __rbd_img_fill_request(struct rbd_img_request *img_req) ret = rbd_obj_setup_zeroout(obj_req); break; default: - rbd_assert(0); + BUG(); } if (ret < 0) return ret; @@ -2383,7 +2383,7 @@ static int rbd_obj_read_from_parent(struct rbd_obj_request *obj_req) &obj_req->bvec_pos); break; default: - rbd_assert(0); + BUG(); } } else { ret = rbd_img_fill_from_bvecs(child_img_req, @@ -2515,7 +2515,7 @@ static int rbd_obj_issue_copyup_ops(struct rbd_obj_request *obj_req, u32 bytes) num_osd_ops += count_zeroout_ops(obj_req); break; default: - rbd_assert(0); + BUG(); } obj_req->osd_req = rbd_osd_req_create(obj_req, num_osd_ops); @@ -2542,7 +2542,7 @@ static int rbd_obj_issue_copyup_ops(struct rbd_obj_request *obj_req, u32 bytes) __rbd_obj_setup_zeroout(obj_req, which); break; default: - rbd_assert(0); + BUG(); } ret = ceph_osdc_alloc_messages(obj_req->osd_req, GFP_NOIO); @@ -3842,8 +3842,12 @@ static void rbd_queue_workfn(struct work_struct *work) goto err_rq; } - rbd_assert(op_type == OBJ_OP_READ || - rbd_dev->spec->snap_id == CEPH_NOSNAP); + if (op_type != OBJ_OP_READ && rbd_dev->spec->snap_id != CEPH_NOSNAP) { + rbd_warn(rbd_dev, "%s on read-only snapshot", + obj_op_name(op_type)); + result = -EIO; + goto err; + } /* * Quit early if the mapped snapshot no longer exists. It's diff --git a/drivers/hwmon/hwmon.c b/drivers/hwmon/hwmon.c index cd91510a5387..e694c46ff039 100644 --- a/drivers/hwmon/hwmon.c +++ b/drivers/hwmon/hwmon.c @@ -118,9 +118,7 @@ static DEFINE_IDA(hwmon_ida); * The complex conditional is necessary to avoid a cyclic dependency * between hwmon and thermal_sys modules. */ -#if IS_REACHABLE(CONFIG_THERMAL) && defined(CONFIG_THERMAL_OF) && \ - (!defined(CONFIG_THERMAL_HWMON) || \ - !(defined(MODULE) && IS_MODULE(CONFIG_THERMAL))) +#ifdef CONFIG_THERMAL_OF static int hwmon_thermal_get_temp(void *data, int *temp) { struct hwmon_thermal_data *tdata = data; diff --git a/drivers/md/Kconfig b/drivers/md/Kconfig index 2557f198e175..db269a348b20 100644 --- a/drivers/md/Kconfig +++ b/drivers/md/Kconfig @@ -436,6 +436,15 @@ config DM_DELAY If unsure, say N. +config DM_DUST + tristate "Bad sector simulation target" + depends on BLK_DEV_DM + ---help--- + A target that simulates bad sector behavior. + Useful for testing. + + If unsure, say N. + config DM_INIT bool "DM \"dm-mod.create=\" parameter support" depends on BLK_DEV_DM=y diff --git a/drivers/md/Makefile b/drivers/md/Makefile index a52b703e588e..be7a6eb92abc 100644 --- a/drivers/md/Makefile +++ b/drivers/md/Makefile @@ -48,6 +48,7 @@ obj-$(CONFIG_DM_BUFIO) += dm-bufio.o obj-$(CONFIG_DM_BIO_PRISON) += dm-bio-prison.o obj-$(CONFIG_DM_CRYPT) += dm-crypt.o obj-$(CONFIG_DM_DELAY) += dm-delay.o +obj-$(CONFIG_DM_DUST) += dm-dust.o obj-$(CONFIG_DM_FLAKEY) += dm-flakey.o obj-$(CONFIG_DM_MULTIPATH) += dm-multipath.o dm-round-robin.o obj-$(CONFIG_DM_MULTIPATH_QL) += dm-queue-length.o diff --git a/drivers/md/dm-cache-metadata.c b/drivers/md/dm-cache-metadata.c index 6fc93834da44..151aa95775be 100644 --- a/drivers/md/dm-cache-metadata.c +++ b/drivers/md/dm-cache-metadata.c @@ -1167,11 +1167,18 @@ static int __load_discards(struct dm_cache_metadata *cmd, if (r) return r; - for (b = 0; b < from_dblock(cmd->discard_nr_blocks); b++) { + for (b = 0; ; b++) { r = fn(context, cmd->discard_block_size, to_dblock(b), dm_bitset_cursor_get_value(&c)); if (r) break; + + if (b >= (from_dblock(cmd->discard_nr_blocks) - 1)) + break; + + r = dm_bitset_cursor_next(&c); + if (r) + break; } dm_bitset_cursor_end(&c); diff --git a/drivers/md/dm-crypt.c b/drivers/md/dm-crypt.c index 7f6462f74ac8..1b16d34bb785 100644 --- a/drivers/md/dm-crypt.c +++ b/drivers/md/dm-crypt.c @@ -946,6 +946,7 @@ static int crypt_integrity_ctr(struct crypt_config *cc, struct dm_target *ti) { #ifdef CONFIG_BLK_DEV_INTEGRITY struct blk_integrity *bi = blk_get_integrity(cc->dev->bdev->bd_disk); + struct mapped_device *md = dm_table_get_md(ti->table); /* From now we require underlying device with our integrity profile */ if (!bi || strcasecmp(bi->profile->name, "DM-DIF-EXT-TAG")) { @@ -965,7 +966,7 @@ static int crypt_integrity_ctr(struct crypt_config *cc, struct dm_target *ti) if (crypt_integrity_aead(cc)) { cc->integrity_tag_size = cc->on_disk_tag_size - cc->integrity_iv_size; - DMINFO("Integrity AEAD, tag size %u, IV size %u.", + DMDEBUG("%s: Integrity AEAD, tag size %u, IV size %u.", dm_device_name(md), cc->integrity_tag_size, cc->integrity_iv_size); if (crypto_aead_setauthsize(any_tfm_aead(cc), cc->integrity_tag_size)) { @@ -973,7 +974,7 @@ static int crypt_integrity_ctr(struct crypt_config *cc, struct dm_target *ti) return -EINVAL; } } else if (cc->integrity_iv_size) - DMINFO("Additional per-sector space %u bytes for IV.", + DMDEBUG("%s: Additional per-sector space %u bytes for IV.", dm_device_name(md), cc->integrity_iv_size); if ((cc->integrity_tag_size + cc->integrity_iv_size) != bi->tag_size) { @@ -1031,11 +1032,11 @@ static u8 *org_iv_of_dmreq(struct crypt_config *cc, return iv_of_dmreq(cc, dmreq) + cc->iv_size; } -static uint64_t *org_sector_of_dmreq(struct crypt_config *cc, +static __le64 *org_sector_of_dmreq(struct crypt_config *cc, struct dm_crypt_request *dmreq) { u8 *ptr = iv_of_dmreq(cc, dmreq) + cc->iv_size + cc->iv_size; - return (uint64_t*) ptr; + return (__le64 *) ptr; } static unsigned int *org_tag_of_dmreq(struct crypt_config *cc, @@ -1071,7 +1072,7 @@ static int crypt_convert_block_aead(struct crypt_config *cc, struct bio_vec bv_out = bio_iter_iovec(ctx->bio_out, ctx->iter_out); struct dm_crypt_request *dmreq; u8 *iv, *org_iv, *tag_iv, *tag; - uint64_t *sector; + __le64 *sector; int r = 0; BUG_ON(cc->integrity_iv_size && cc->integrity_iv_size != cc->iv_size); @@ -1143,9 +1144,11 @@ static int crypt_convert_block_aead(struct crypt_config *cc, r = crypto_aead_decrypt(req); } - if (r == -EBADMSG) - DMERR_LIMIT("INTEGRITY AEAD ERROR, sector %llu", + if (r == -EBADMSG) { + char b[BDEVNAME_SIZE]; + DMERR_LIMIT("%s: INTEGRITY AEAD ERROR, sector %llu", bio_devname(ctx->bio_in, b), (unsigned long long)le64_to_cpu(*sector)); + } if (!r && cc->iv_gen_ops && cc->iv_gen_ops->post) r = cc->iv_gen_ops->post(cc, org_iv, dmreq); @@ -1166,7 +1169,7 @@ static int crypt_convert_block_skcipher(struct crypt_config *cc, struct scatterlist *sg_in, *sg_out; struct dm_crypt_request *dmreq; u8 *iv, *org_iv, *tag_iv; - uint64_t *sector; + __le64 *sector; int r = 0; /* Reject unexpected unaligned bio. */ @@ -1788,7 +1791,8 @@ static void kcryptd_async_done(struct crypto_async_request *async_req, error = cc->iv_gen_ops->post(cc, org_iv_of_dmreq(cc, dmreq), dmreq); if (error == -EBADMSG) { - DMERR_LIMIT("INTEGRITY AEAD ERROR, sector %llu", + char b[BDEVNAME_SIZE]; + DMERR_LIMIT("%s: INTEGRITY AEAD ERROR, sector %llu", bio_devname(ctx->bio_in, b), (unsigned long long)le64_to_cpu(*org_sector_of_dmreq(cc, dmreq))); io->error = BLK_STS_PROTECTION; } else if (error < 0) @@ -1887,7 +1891,7 @@ static int crypt_alloc_tfms_skcipher(struct crypt_config *cc, char *ciphermode) * algorithm implementation is used. Help people debug performance * problems by logging the ->cra_driver_name. */ - DMINFO("%s using implementation \"%s\"", ciphermode, + DMDEBUG_LIMIT("%s using implementation \"%s\"", ciphermode, crypto_skcipher_alg(any_tfm(cc))->base.cra_driver_name); return 0; } @@ -1907,7 +1911,7 @@ static int crypt_alloc_tfms_aead(struct crypt_config *cc, char *ciphermode) return err; } - DMINFO("%s using implementation \"%s\"", ciphermode, + DMDEBUG_LIMIT("%s using implementation \"%s\"", ciphermode, crypto_aead_alg(any_tfm_aead(cc))->base.cra_driver_name); return 0; } diff --git a/drivers/md/dm-delay.c b/drivers/md/dm-delay.c index fddffe251bf6..f496213f8b67 100644 --- a/drivers/md/dm-delay.c +++ b/drivers/md/dm-delay.c @@ -121,7 +121,8 @@ static void delay_dtr(struct dm_target *ti) { struct delay_c *dc = ti->private; - destroy_workqueue(dc->kdelayd_wq); + if (dc->kdelayd_wq) + destroy_workqueue(dc->kdelayd_wq); if (dc->read.dev) dm_put_device(ti, dc->read.dev); diff --git a/drivers/md/dm-dust.c b/drivers/md/dm-dust.c new file mode 100644 index 000000000000..845f376a72d9 --- /dev/null +++ b/drivers/md/dm-dust.c @@ -0,0 +1,515 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Copyright (c) 2018 Red Hat, Inc. + * + * This is a test "dust" device, which fails reads on specified + * sectors, emulating the behavior of a hard disk drive sending + * a "Read Medium Error" sense. + * + */ + +#include <linux/device-mapper.h> +#include <linux/module.h> +#include <linux/rbtree.h> + +#define DM_MSG_PREFIX "dust" + +struct badblock { + struct rb_node node; + sector_t bb; +}; + +struct dust_device { + struct dm_dev *dev; + struct rb_root badblocklist; + unsigned long long badblock_count; + spinlock_t dust_lock; + unsigned int blksz; + unsigned int sect_per_block; + sector_t start; + bool fail_read_on_bb:1; + bool quiet_mode:1; +}; + +static struct badblock *dust_rb_search(struct rb_root *root, sector_t blk) +{ + struct rb_node *node = root->rb_node; + + while (node) { + struct badblock *bblk = rb_entry(node, struct badblock, node); + + if (bblk->bb > blk) + node = node->rb_left; + else if (bblk->bb < blk) + node = node->rb_right; + else + return bblk; + } + + return NULL; +} + +static bool dust_rb_insert(struct rb_root *root, struct badblock *new) +{ + struct badblock *bblk; + struct rb_node **link = &root->rb_node, *parent = NULL; + sector_t value = new->bb; + + while (*link) { + parent = *link; + bblk = rb_entry(parent, struct badblock, node); + + if (bblk->bb > value) + link = &(*link)->rb_left; + else if (bblk->bb < value) + link = &(*link)->rb_right; + else + return false; + } + + rb_link_node(&new->node, parent, link); + rb_insert_color(&new->node, root); + + return true; +} + +static int dust_remove_block(struct dust_device *dd, unsigned long long block) +{ + struct badblock *bblock; + unsigned long flags; + + spin_lock_irqsave(&dd->dust_lock, flags); + bblock = dust_rb_search(&dd->badblocklist, block * dd->sect_per_block); + + if (bblock == NULL) { + if (!dd->quiet_mode) { + DMERR("%s: block %llu not found in badblocklist", + __func__, block); + } + spin_unlock_irqrestore(&dd->dust_lock, flags); + return -EINVAL; + } + + rb_erase(&bblock->node, &dd->badblocklist); + dd->badblock_count--; + if (!dd->quiet_mode) + DMINFO("%s: badblock removed at block %llu", __func__, block); + kfree(bblock); + spin_unlock_irqrestore(&dd->dust_lock, flags); + + return 0; +} + +static int dust_add_block(struct dust_device *dd, unsigned long long block) +{ + struct badblock *bblock; + unsigned long flags; + + bblock = kmalloc(sizeof(*bblock), GFP_KERNEL); + if (bblock == NULL) { + if (!dd->quiet_mode) + DMERR("%s: badblock allocation failed", __func__); + return -ENOMEM; + } + + spin_lock_irqsave(&dd->dust_lock, flags); + bblock->bb = block * dd->sect_per_block; + if (!dust_rb_insert(&dd->badblocklist, bblock)) { + if (!dd->quiet_mode) { + DMERR("%s: block %llu already in badblocklist", + __func__, block); + } + spin_unlock_irqrestore(&dd->dust_lock, flags); + kfree(bblock); + return -EINVAL; + } + + dd->badblock_count++; + if (!dd->quiet_mode) + DMINFO("%s: badblock added at block %llu", __func__, block); + spin_unlock_irqrestore(&dd->dust_lock, flags); + + return 0; +} + +static int dust_query_block(struct dust_device *dd, unsigned long long block) +{ + struct badblock *bblock; + unsigned long flags; + + spin_lock_irqsave(&dd->dust_lock, flags); + bblock = dust_rb_search(&dd->badblocklist, block * dd->sect_per_block); + if (bblock != NULL) + DMINFO("%s: block %llu found in badblocklist", __func__, block); + else + DMINFO("%s: block %llu not found in badblocklist", __func__, block); + spin_unlock_irqrestore(&dd->dust_lock, flags); + + return 0; +} + +static int __dust_map_read(struct dust_device *dd, sector_t thisblock) +{ + struct badblock *bblk = dust_rb_search(&dd->badblocklist, thisblock); + + if (bblk) + return DM_MAPIO_KILL; + + return DM_MAPIO_REMAPPED; +} + +static int dust_map_read(struct dust_device *dd, sector_t thisblock, + bool fail_read_on_bb) +{ + unsigned long flags; + int ret = DM_MAPIO_REMAPPED; + + if (fail_read_on_bb) { + spin_lock_irqsave(&dd->dust_lock, flags); + ret = __dust_map_read(dd, thisblock); + spin_unlock_irqrestore(&dd->dust_lock, flags); + } + + return ret; +} + +static void __dust_map_write(struct dust_device *dd, sector_t thisblock) +{ + struct badblock *bblk = dust_rb_search(&dd->badblocklist, thisblock); + + if (bblk) { + rb_erase(&bblk->node, &dd->badblocklist); + dd->badblock_count--; + kfree(bblk); + if (!dd->quiet_mode) { + sector_div(thisblock, dd->sect_per_block); + DMINFO("block %llu removed from badblocklist by write", + (unsigned long long)thisblock); + } + } +} + +static int dust_map_write(struct dust_device *dd, sector_t thisblock, + bool fail_read_on_bb) +{ + unsigned long flags; + + if (fail_read_on_bb) { + spin_lock_irqsave(&dd->dust_lock, flags); + __dust_map_write(dd, thisblock); + spin_unlock_irqrestore(&dd->dust_lock, flags); + } + + return DM_MAPIO_REMAPPED; +} + +static int dust_map(struct dm_target *ti, struct bio *bio) +{ + struct dust_device *dd = ti->private; + int ret; + + bio_set_dev(bio, dd->dev->bdev); + bio->bi_iter.bi_sector = dd->start + dm_target_offset(ti, bio->bi_iter.bi_sector); + + if (bio_data_dir(bio) == READ) + ret = dust_map_read(dd, bio->bi_iter.bi_sector, dd->fail_read_on_bb); + else + ret = dust_map_write(dd, bio->bi_iter.bi_sector, dd->fail_read_on_bb); + + return ret; +} + +static bool __dust_clear_badblocks(struct rb_root *tree, + unsigned long long count) +{ + struct rb_node *node = NULL, *nnode = NULL; + + nnode = rb_first(tree); + if (nnode == NULL) { + BUG_ON(count != 0); + return false; + } + + while (nnode) { + node = nnode; + nnode = rb_next(node); + rb_erase(node, tree); + count--; + kfree(node); + } + BUG_ON(count != 0); + BUG_ON(tree->rb_node != NULL); + + return true; +} + +static int dust_clear_badblocks(struct dust_device *dd) +{ + unsigned long flags; + struct rb_root badblocklist; + unsigned long long badblock_count; + + spin_lock_irqsave(&dd->dust_lock, flags); + badblocklist = dd->badblocklist; + badblock_count = dd->badblock_count; + dd->badblocklist = RB_ROOT; + dd->badblock_count = 0; + spin_unlock_irqrestore(&dd->dust_lock, flags); + + if (!__dust_clear_badblocks(&badblocklist, badblock_count)) + DMINFO("%s: no badblocks found", __func__); + else + DMINFO("%s: badblocks cleared", __func__); + + return 0; +} + +/* + * Target parameters: + * + * <device_path> <offset> <blksz> + * + * device_path: path to the block device + * offset: offset to data area from start of device_path + * blksz: block size (minimum 512, maximum 1073741824, must be a power of 2) + */ +static int dust_ctr(struct dm_target *ti, unsigned int argc, char **argv) +{ + struct dust_device *dd; + unsigned long long tmp; + char dummy; + unsigned int blksz; + unsigned int sect_per_block; + sector_t DUST_MAX_BLKSZ_SECTORS = 2097152; + sector_t max_block_sectors = min(ti->len, DUST_MAX_BLKSZ_SECTORS); + + if (argc != 3) { + ti->error = "Invalid argument count"; + return -EINVAL; + } + + if (kstrtouint(argv[2], 10, &blksz) || !blksz) { + ti->error = "Invalid block size parameter"; + return -EINVAL; + } + + if (blksz < 512) { + ti->error = "Block size must be at least 512"; + return -EINVAL; + } + + if (!is_power_of_2(blksz)) { + ti->error = "Block size must be a power of 2"; + return -EINVAL; + } + + if (to_sector(blksz) > max_block_sectors) { + ti->error = "Block size is too large"; + return -EINVAL; + } + + sect_per_block = (blksz >> SECTOR_SHIFT); + + if (sscanf(argv[1], "%llu%c", &tmp, &dummy) != 1 || tmp != (sector_t)tmp) { + ti->error = "Invalid device offset sector"; + return -EINVAL; + } + + dd = kzalloc(sizeof(struct dust_device), GFP_KERNEL); + if (dd == NULL) { + ti->error = "Cannot allocate context"; + return -ENOMEM; + } + + if (dm_get_device(ti, argv[0], dm_table_get_mode(ti->table), &dd->dev)) { + ti->error = "Device lookup failed"; + kfree(dd); + return -EINVAL; + } + + dd->sect_per_block = sect_per_block; + dd->blksz = blksz; + dd->start = tmp; + + /* + * Whether to fail a read on a "bad" block. + * Defaults to false; enabled later by message. + */ + dd->fail_read_on_bb = false; + + /* + * Initialize bad block list rbtree. + */ + dd->badblocklist = RB_ROOT; + dd->badblock_count = 0; + spin_lock_init(&dd->dust_lock); + + dd->quiet_mode = false; + + BUG_ON(dm_set_target_max_io_len(ti, dd->sect_per_block) != 0); + + ti->num_discard_bios = 1; + ti->num_flush_bios = 1; + ti->private = dd; + + return 0; +} + +static void dust_dtr(struct dm_target *ti) +{ + struct dust_device *dd = ti->private; + + __dust_clear_badblocks(&dd->badblocklist, dd->badblock_count); + dm_put_device(ti, dd->dev); + kfree(dd); +} + +static int dust_message(struct dm_target *ti, unsigned int argc, char **argv, + char *result_buf, unsigned int maxlen) +{ + struct dust_device *dd = ti->private; + sector_t size = i_size_read(dd->dev->bdev->bd_inode) >> SECTOR_SHIFT; + bool invalid_msg = false; + int result = -EINVAL; + unsigned long long tmp, block; + unsigned long flags; + char dummy; + + if (argc == 1) { + if (!strcasecmp(argv[0], "addbadblock") || + !strcasecmp(argv[0], "removebadblock") || + !strcasecmp(argv[0], "queryblock")) { + DMERR("%s requires an additional argument", argv[0]); + } else if (!strcasecmp(argv[0], "disable")) { + DMINFO("disabling read failures on bad sectors"); + dd->fail_read_on_bb = false; + result = 0; + } else if (!strcasecmp(argv[0], "enable")) { + DMINFO("enabling read failures on bad sectors"); + dd->fail_read_on_bb = true; + result = 0; + } else if (!strcasecmp(argv[0], "countbadblocks")) { + spin_lock_irqsave(&dd->dust_lock, flags); + DMINFO("countbadblocks: %llu badblock(s) found", + dd->badblock_count); + spin_unlock_irqrestore(&dd->dust_lock, flags); + result = 0; + } else if (!strcasecmp(argv[0], "clearbadblocks")) { + result = dust_clear_badblocks(dd); + } else if (!strcasecmp(argv[0], "quiet")) { + if (!dd->quiet_mode) + dd->quiet_mode = true; + else + dd->quiet_mode = false; + result = 0; + } else { + invalid_msg = true; + } + } else if (argc == 2) { + if (sscanf(argv[1], "%llu%c", &tmp, &dummy) != 1) + return result; + + block = tmp; + sector_div(size, dd->sect_per_block); + if (block > size) { + DMERR("selected block value out of range"); + return result; + } + + if (!strcasecmp(argv[0], "addbadblock")) + result = dust_add_block(dd, block); + else if (!strcasecmp(argv[0], "removebadblock")) + result = dust_remove_block(dd, block); + else if (!strcasecmp(argv[0], "queryblock")) + result = dust_query_block(dd, block); + else + invalid_msg = true; + + } else + DMERR("invalid number of arguments '%d'", argc); + + if (invalid_msg) + DMERR("unrecognized message '%s' received", argv[0]); + + return result; +} + +static void dust_status(struct dm_target *ti, status_type_t type, + unsigned int status_flags, char *result, unsigned int maxlen) +{ + struct dust_device *dd = ti->private; + unsigned int sz = 0; + + switch (type) { + case STATUSTYPE_INFO: + DMEMIT("%s %s %s", dd->dev->name, + dd->fail_read_on_bb ? "fail_read_on_bad_block" : "bypass", + dd->quiet_mode ? "quiet" : "verbose"); + break; + + case STATUSTYPE_TABLE: + DMEMIT("%s %llu %u", dd->dev->name, + (unsigned long long)dd->start, dd->blksz); + break; + } +} + +static int dust_prepare_ioctl(struct dm_target *ti, struct block_device **bdev) +{ + struct dust_device *dd = ti->private; + struct dm_dev *dev = dd->dev; + + *bdev = dev->bdev; + + /* + * Only pass ioctls through if the device sizes match exactly. + */ + if (dd->start || + ti->len != i_size_read(dev->bdev->bd_inode) >> SECTOR_SHIFT) + return 1; + + return 0; +} + +static int dust_iterate_devices(struct dm_target *ti, iterate_devices_callout_fn fn, + void *data) +{ + struct dust_device *dd = ti->private; + + return fn(ti, dd->dev, dd->start, ti->len, data); +} + +static struct target_type dust_target = { + .name = "dust", + .version = {1, 0, 0}, + .module = THIS_MODULE, + .ctr = dust_ctr, + .dtr = dust_dtr, + .iterate_devices = dust_iterate_devices, + .map = dust_map, + .message = dust_message, + .status = dust_status, + .prepare_ioctl = dust_prepare_ioctl, +}; + +static int __init dm_dust_init(void) +{ + int result = dm_register_target(&dust_target); + + if (result < 0) + DMERR("dm_register_target failed %d", result); + + return result; +} + +static void __exit dm_dust_exit(void) +{ + dm_unregister_target(&dust_target); +} + +module_init(dm_dust_init); +module_exit(dm_dust_exit); + +MODULE_DESCRIPTION(DM_NAME " dust test target"); +MODULE_AUTHOR("Bryan Gurney <dm-devel@redhat.com>"); +MODULE_LICENSE("GPL"); diff --git a/drivers/md/dm-exception-store.h b/drivers/md/dm-exception-store.h index 721efc493942..3f4139ac1f60 100644 --- a/drivers/md/dm-exception-store.h +++ b/drivers/md/dm-exception-store.h @@ -11,6 +11,7 @@ #define _LINUX_DM_EXCEPTION_STORE #include <linux/blkdev.h> +#include <linux/list_bl.h> #include <linux/device-mapper.h> /* @@ -27,7 +28,7 @@ typedef sector_t chunk_t; * chunk within the device. */ struct dm_exception { - struct list_head hash_list; + struct hlist_bl_node hash_list; chunk_t old_chunk; chunk_t new_chunk; diff --git a/drivers/md/dm-init.c b/drivers/md/dm-init.c index 4b76f84424c3..352e803f566e 100644 --- a/drivers/md/dm-init.c +++ b/drivers/md/dm-init.c @@ -160,7 +160,7 @@ static int __init dm_parse_table(struct dm_device *dev, char *str) while (table_entry) { DMDEBUG("parsing table \"%s\"", str); - if (++dev->dmi.target_count >= DM_MAX_TARGETS) { + if (++dev->dmi.target_count > DM_MAX_TARGETS) { DMERR("too many targets %u > %d", dev->dmi.target_count, DM_MAX_TARGETS); return -EINVAL; @@ -242,9 +242,9 @@ static int __init dm_parse_devices(struct list_head *devices, char *str) return -ENOMEM; list_add_tail(&dev->list, devices); - if (++ndev >= DM_MAX_DEVICES) { - DMERR("too many targets %u > %d", - dev->dmi.target_count, DM_MAX_TARGETS); + if (++ndev > DM_MAX_DEVICES) { + DMERR("too many devices %lu > %d", + ndev, DM_MAX_DEVICES); return -EINVAL; } diff --git a/drivers/md/dm-integrity.c b/drivers/md/dm-integrity.c index c27c32cf4a30..44e76cda087a 100644 --- a/drivers/md/dm-integrity.c +++ b/drivers/md/dm-integrity.c @@ -15,6 +15,7 @@ #include <linux/rbtree.h> #include <linux/delay.h> #include <linux/random.h> +#include <linux/reboot.h> #include <crypto/hash.h> #include <crypto/skcipher.h> #include <linux/async_tx.h> @@ -24,6 +25,7 @@ #define DEFAULT_INTERLEAVE_SECTORS 32768 #define DEFAULT_JOURNAL_SIZE_FACTOR 7 +#define DEFAULT_SECTORS_PER_BITMAP_BIT 32768 #define DEFAULT_BUFFER_SECTORS 128 #define DEFAULT_JOURNAL_WATERMARK 50 #define DEFAULT_SYNC_MSEC 10000 @@ -33,6 +35,8 @@ #define METADATA_WORKQUEUE_MAX_ACTIVE 16 #define RECALC_SECTORS 8192 #define RECALC_WRITE_SUPER 16 +#define BITMAP_BLOCK_SIZE 4096 /* don't change it */ +#define BITMAP_FLUSH_INTERVAL (10 * HZ) /* * Warning - DEBUG_PRINT prints security-sensitive data to the log, @@ -48,6 +52,7 @@ #define SB_MAGIC "integrt" #define SB_VERSION_1 1 #define SB_VERSION_2 2 +#define SB_VERSION_3 3 #define SB_SECTORS 8 #define MAX_SECTORS_PER_BLOCK 8 @@ -60,12 +65,14 @@ struct superblock { __u64 provided_data_sectors; /* userspace uses this value */ __u32 flags; __u8 log2_sectors_per_block; - __u8 pad[3]; + __u8 log2_blocks_per_bitmap_bit; + __u8 pad[2]; __u64 recalc_sector; }; #define SB_FLAG_HAVE_JOURNAL_MAC 0x1 #define SB_FLAG_RECALCULATING 0x2 +#define SB_FLAG_DIRTY_BITMAP 0x4 #define JOURNAL_ENTRY_ROUNDUP 8 @@ -151,9 +158,18 @@ struct dm_integrity_c { struct workqueue_struct *metadata_wq; struct superblock *sb; unsigned journal_pages; + unsigned n_bitmap_blocks; + struct page_list *journal; struct page_list *journal_io; struct page_list *journal_xor; + struct page_list *recalc_bitmap; + struct page_list *may_write_bitmap; + struct bitmap_block_status *bbs; + unsigned bitmap_flush_interval; + int synchronous_mode; + struct bio_list synchronous_bios; + struct delayed_work bitmap_flush_work; struct crypto_skcipher *journal_crypt; struct scatterlist **journal_scatterlist; @@ -180,6 +196,7 @@ struct dm_integrity_c { __s8 log2_metadata_run; __u8 log2_buffer_sectors; __u8 sectors_per_block; + __u8 log2_blocks_per_bitmap_bit; unsigned char mode; int suspending; @@ -232,17 +249,20 @@ struct dm_integrity_c { bool journal_uptodate; bool just_formatted; + bool recalculate_flag; struct alg_spec internal_hash_alg; struct alg_spec journal_crypt_alg; struct alg_spec journal_mac_alg; atomic64_t number_of_mismatches; + + struct notifier_block reboot_notifier; }; struct dm_integrity_range { sector_t logical_sector; - unsigned n_sectors; + sector_t n_sectors; bool waiting; union { struct rb_node node; @@ -288,6 +308,16 @@ struct journal_io { struct journal_completion *comp; }; +struct bitmap_block_status { + struct work_struct work; + struct dm_integrity_c *ic; + unsigned idx; + unsigned long *bitmap; + struct bio_list bio_queue; + spinlock_t bio_queue_lock; + +}; + static struct kmem_cache *journal_io_cache; #define JOURNAL_IO_MEMPOOL 32 @@ -423,7 +453,9 @@ static void wraparound_section(struct dm_integrity_c *ic, unsigned *sec_ptr) static void sb_set_version(struct dm_integrity_c *ic) { - if (ic->meta_dev || ic->sb->flags & cpu_to_le32(SB_FLAG_RECALCULATING)) + if (ic->mode == 'B' || ic->sb->flags & cpu_to_le32(SB_FLAG_DIRTY_BITMAP)) + ic->sb->version = SB_VERSION_3; + else if (ic->meta_dev || ic->sb->flags & cpu_to_le32(SB_FLAG_RECALCULATING)) ic->sb->version = SB_VERSION_2; else ic->sb->version = SB_VERSION_1; @@ -447,6 +479,137 @@ static int sync_rw_sb(struct dm_integrity_c *ic, int op, int op_flags) return dm_io(&io_req, 1, &io_loc, NULL); } +#define BITMAP_OP_TEST_ALL_SET 0 +#define BITMAP_OP_TEST_ALL_CLEAR 1 +#define BITMAP_OP_SET 2 +#define BITMAP_OP_CLEAR 3 + +static bool block_bitmap_op(struct dm_integrity_c *ic, struct page_list *bitmap, + sector_t sector, sector_t n_sectors, int mode) +{ + unsigned long bit, end_bit, this_end_bit, page, end_page; + unsigned long *data; + + if (unlikely(((sector | n_sectors) & ((1 << ic->sb->log2_sectors_per_block) - 1)) != 0)) { + DMCRIT("invalid bitmap access (%llx,%llx,%d,%d,%d)", + (unsigned long long)sector, + (unsigned long long)n_sectors, + ic->sb->log2_sectors_per_block, + ic->log2_blocks_per_bitmap_bit, + mode); + BUG(); + } + + if (unlikely(!n_sectors)) + return true; + + bit = sector >> (ic->sb->log2_sectors_per_block + ic->log2_blocks_per_bitmap_bit); + end_bit = (sector + n_sectors - 1) >> + (ic->sb->log2_sectors_per_block + ic->log2_blocks_per_bitmap_bit); + + page = bit / (PAGE_SIZE * 8); + bit %= PAGE_SIZE * 8; + + end_page = end_bit / (PAGE_SIZE * 8); + end_bit %= PAGE_SIZE * 8; + +repeat: + if (page < end_page) { + this_end_bit = PAGE_SIZE * 8 - 1; + } else { + this_end_bit = end_bit; + } + + data = lowmem_page_address(bitmap[page].page); + + if (mode == BITMAP_OP_TEST_ALL_SET) { + while (bit <= this_end_bit) { + if (!(bit % BITS_PER_LONG) && this_end_bit >= bit + BITS_PER_LONG - 1) { + do { + if (data[bit / BITS_PER_LONG] != -1) + return false; + bit += BITS_PER_LONG; + } while (this_end_bit >= bit + BITS_PER_LONG - 1); + continue; + } + if (!test_bit(bit, data)) + return false; + bit++; + } + } else if (mode == BITMAP_OP_TEST_ALL_CLEAR) { + while (bit <= this_end_bit) { + if (!(bit % BITS_PER_LONG) && this_end_bit >= bit + BITS_PER_LONG - 1) { + do { + if (data[bit / BITS_PER_LONG] != 0) + return false; + bit += BITS_PER_LONG; + } while (this_end_bit >= bit + BITS_PER_LONG - 1); + continue; + } + if (test_bit(bit, data)) + return false; + bit++; + } + } else if (mode == BITMAP_OP_SET) { + while (bit <= this_end_bit) { + if (!(bit % BITS_PER_LONG) && this_end_bit >= bit + BITS_PER_LONG - 1) { + do { + data[bit / BITS_PER_LONG] = -1; + bit += BITS_PER_LONG; + } while (this_end_bit >= bit + BITS_PER_LONG - 1); + continue; + } + __set_bit(bit, data); + bit++; + } + } else if (mode == BITMAP_OP_CLEAR) { + if (!bit && this_end_bit == PAGE_SIZE * 8 - 1) + clear_page(data); + else while (bit <= this_end_bit) { + if (!(bit % BITS_PER_LONG) && this_end_bit >= bit + BITS_PER_LONG - 1) { + do { + data[bit / BITS_PER_LONG] = 0; + bit += BITS_PER_LONG; + } while (this_end_bit >= bit + BITS_PER_LONG - 1); + continue; + } + __clear_bit(bit, data); + bit++; + } + } else { + BUG(); + } + + if (unlikely(page < end_page)) { + bit = 0; + page++; + goto repeat; + } + + return true; +} + +static void block_bitmap_copy(struct dm_integrity_c *ic, struct page_list *dst, struct page_list *src) +{ + unsigned n_bitmap_pages = DIV_ROUND_UP(ic->n_bitmap_blocks, PAGE_SIZE / BITMAP_BLOCK_SIZE); + unsigned i; + + for (i = 0; i < n_bitmap_pages; i++) { + unsigned long *dst_data = lowmem_page_address(dst[i].page); + unsigned long *src_data = lowmem_page_address(src[i].page); + copy_page(dst_data, src_data); + } +} + +static struct bitmap_block_status *sector_to_bitmap_block(struct dm_integrity_c *ic, sector_t sector) +{ + unsigned bit = sector >> (ic->sb->log2_sectors_per_block + ic->log2_blocks_per_bitmap_bit); + unsigned bitmap_block = bit / (BITMAP_BLOCK_SIZE * 8); + + BUG_ON(bitmap_block >= ic->n_bitmap_blocks); + return &ic->bbs[bitmap_block]; +} + static void access_journal_check(struct dm_integrity_c *ic, unsigned section, unsigned offset, bool e, const char *function) { @@ -455,8 +618,8 @@ static void access_journal_check(struct dm_integrity_c *ic, unsigned section, un if (unlikely(section >= ic->journal_sections) || unlikely(offset >= limit)) { - printk(KERN_CRIT "%s: invalid access at (%u,%u), limit (%u,%u)\n", - function, section, offset, ic->journal_sections, limit); + DMCRIT("%s: invalid access at (%u,%u), limit (%u,%u)", + function, section, offset, ic->journal_sections, limit); BUG(); } #endif @@ -756,12 +919,12 @@ static void complete_journal_io(unsigned long error, void *context) complete_journal_op(comp); } -static void rw_journal(struct dm_integrity_c *ic, int op, int op_flags, unsigned section, - unsigned n_sections, struct journal_completion *comp) +static void rw_journal_sectors(struct dm_integrity_c *ic, int op, int op_flags, + unsigned sector, unsigned n_sectors, struct journal_completion *comp) { struct dm_io_request io_req; struct dm_io_region io_loc; - unsigned sector, n_sectors, pl_index, pl_offset; + unsigned pl_index, pl_offset; int r; if (unlikely(dm_integrity_failed(ic))) { @@ -770,9 +933,6 @@ static void rw_journal(struct dm_integrity_c *ic, int op, int op_flags, unsigned return; } - sector = section * ic->journal_section_sectors; - n_sectors = n_sections * ic->journal_section_sectors; - pl_index = sector >> (PAGE_SHIFT - SECTOR_SHIFT); pl_offset = (sector << SECTOR_SHIFT) & (PAGE_SIZE - 1); @@ -805,6 +965,17 @@ static void rw_journal(struct dm_integrity_c *ic, int op, int op_flags, unsigned } } +static void rw_journal(struct dm_integrity_c *ic, int op, int op_flags, unsigned section, + unsigned n_sections, struct journal_completion *comp) +{ + unsigned sector, n_sectors; + + sector = section * ic->journal_section_sectors; + n_sectors = n_sections * ic->journal_section_sectors; + + rw_journal_sectors(ic, op, op_flags, sector, n_sectors, comp); +} + static void write_journal(struct dm_integrity_c *ic, unsigned commit_start, unsigned commit_sections) { struct journal_completion io_comp; @@ -988,6 +1159,12 @@ static void wait_and_add_new_range(struct dm_integrity_c *ic, struct dm_integrit } while (unlikely(new_range->waiting)); } +static void add_new_range_and_wait(struct dm_integrity_c *ic, struct dm_integrity_range *new_range) +{ + if (unlikely(!add_new_range(ic, new_range, true))) + wait_and_add_new_range(ic, new_range); +} + static void init_journal_node(struct journal_node *node) { RB_CLEAR_NODE(&node->node); @@ -1204,6 +1381,14 @@ static void do_endio(struct dm_integrity_c *ic, struct bio *bio) int r = dm_integrity_failed(ic); if (unlikely(r) && !bio->bi_status) bio->bi_status = errno_to_blk_status(r); + if (unlikely(ic->synchronous_mode) && bio_op(bio) == REQ_OP_WRITE) { + unsigned long flags; + spin_lock_irqsave(&ic->endio_wait.lock, flags); + bio_list_add(&ic->synchronous_bios, bio); + queue_delayed_work(ic->commit_wq, &ic->bitmap_flush_work, 0); + spin_unlock_irqrestore(&ic->endio_wait.lock, flags); + return; + } bio_endio(bio); } @@ -1477,7 +1662,8 @@ static int dm_integrity_map(struct dm_target *ti, struct bio *bio) else wanted_tag_size *= ic->tag_size; if (unlikely(wanted_tag_size != bip->bip_iter.bi_size)) { - DMERR("Invalid integrity data size %u, expected %u", bip->bip_iter.bi_size, wanted_tag_size); + DMERR("Invalid integrity data size %u, expected %u", + bip->bip_iter.bi_size, wanted_tag_size); return DM_MAPIO_KILL; } } @@ -1681,7 +1867,7 @@ retry: unsigned ws, we, range_sectors; dio->range.n_sectors = min(dio->range.n_sectors, - ic->free_sectors << ic->sb->log2_sectors_per_block); + (sector_t)ic->free_sectors << ic->sb->log2_sectors_per_block); if (unlikely(!dio->range.n_sectors)) { if (from_map) goto offload_to_thread; @@ -1764,6 +1950,20 @@ offload_to_thread: goto journal_read_write; } + if (ic->mode == 'B' && dio->write) { + if (!block_bitmap_op(ic, ic->may_write_bitmap, dio->range.logical_sector, + dio->range.n_sectors, BITMAP_OP_TEST_ALL_SET)) { + struct bitmap_block_status *bbs; + + bbs = sector_to_bitmap_block(ic, dio->range.logical_sector); + spin_lock(&bbs->bio_queue_lock); + bio_list_add(&bbs->bio_queue, bio); + spin_unlock(&bbs->bio_queue_lock); + queue_work(ic->writer_wq, &bbs->work); + return; + } + } + dio->in_flight = (atomic_t)ATOMIC_INIT(2); if (need_sync_io) { @@ -1790,10 +1990,15 @@ offload_to_thread: if (need_sync_io) { wait_for_completion_io(&read_comp); - if (unlikely(ic->recalc_wq != NULL) && - ic->sb->flags & cpu_to_le32(SB_FLAG_RECALCULATING) && + if (ic->sb->flags & cpu_to_le32(SB_FLAG_RECALCULATING) && dio->range.logical_sector + dio->range.n_sectors > le64_to_cpu(ic->sb->recalc_sector)) goto skip_check; + if (ic->mode == 'B') { + if (!block_bitmap_op(ic, ic->recalc_bitmap, dio->range.logical_sector, + dio->range.n_sectors, BITMAP_OP_TEST_ALL_CLEAR)) + goto skip_check; + } + if (likely(!bio->bi_status)) integrity_metadata(&dio->work); else @@ -1831,8 +2036,16 @@ static void pad_uncommitted(struct dm_integrity_c *ic) wraparound_section(ic, &ic->free_section); ic->n_uncommitted_sections++; } - WARN_ON(ic->journal_sections * ic->journal_section_entries != - (ic->n_uncommitted_sections + ic->n_committed_sections) * ic->journal_section_entries + ic->free_sectors); + if (WARN_ON(ic->journal_sections * ic->journal_section_entries != + (ic->n_uncommitted_sections + ic->n_committed_sections) * + ic->journal_section_entries + ic->free_sectors)) { + DMCRIT("journal_sections %u, journal_section_entries %u, " + "n_uncommitted_sections %u, n_committed_sections %u, " + "journal_section_entries %u, free_sectors %u", + ic->journal_sections, ic->journal_section_entries, + ic->n_uncommitted_sections, ic->n_committed_sections, + ic->journal_section_entries, ic->free_sectors); + } } static void integrity_commit(struct work_struct *w) @@ -1981,8 +2194,7 @@ static void do_journal_write(struct dm_integrity_c *ic, unsigned write_start, io->range.n_sectors = (k - j) << ic->sb->log2_sectors_per_block; spin_lock_irq(&ic->endio_wait.lock); - if (unlikely(!add_new_range(ic, &io->range, true))) - wait_and_add_new_range(ic, &io->range); + add_new_range_and_wait(ic, &io->range); if (likely(!from_replay)) { struct journal_node *section_node = &ic->journal_tree[i * ic->journal_section_entries]; @@ -2120,11 +2332,14 @@ static void integrity_recalc(struct work_struct *w) sector_t area, offset; sector_t metadata_block; unsigned metadata_offset; + sector_t logical_sector, n_sectors; __u8 *t; unsigned i; int r; unsigned super_counter = 0; + DEBUG_print("start recalculation... (position %llx)\n", le64_to_cpu(ic->sb->recalc_sector)); + spin_lock_irq(&ic->endio_wait.lock); next_chunk: @@ -2133,21 +2348,49 @@ next_chunk: goto unlock_ret; range.logical_sector = le64_to_cpu(ic->sb->recalc_sector); - if (unlikely(range.logical_sector >= ic->provided_data_sectors)) + if (unlikely(range.logical_sector >= ic->provided_data_sectors)) { + if (ic->mode == 'B') { + DEBUG_print("queue_delayed_work: bitmap_flush_work\n"); + queue_delayed_work(ic->commit_wq, &ic->bitmap_flush_work, 0); + } goto unlock_ret; + } get_area_and_offset(ic, range.logical_sector, &area, &offset); range.n_sectors = min((sector_t)RECALC_SECTORS, ic->provided_data_sectors - range.logical_sector); if (!ic->meta_dev) - range.n_sectors = min(range.n_sectors, (1U << ic->sb->log2_interleave_sectors) - (unsigned)offset); - - if (unlikely(!add_new_range(ic, &range, true))) - wait_and_add_new_range(ic, &range); + range.n_sectors = min(range.n_sectors, ((sector_t)1U << ic->sb->log2_interleave_sectors) - (unsigned)offset); + add_new_range_and_wait(ic, &range); spin_unlock_irq(&ic->endio_wait.lock); + logical_sector = range.logical_sector; + n_sectors = range.n_sectors; + + if (ic->mode == 'B') { + if (block_bitmap_op(ic, ic->recalc_bitmap, logical_sector, n_sectors, BITMAP_OP_TEST_ALL_CLEAR)) { + goto advance_and_next; + } + while (block_bitmap_op(ic, ic->recalc_bitmap, logical_sector, + ic->sectors_per_block, BITMAP_OP_TEST_ALL_CLEAR)) { + logical_sector += ic->sectors_per_block; + n_sectors -= ic->sectors_per_block; + cond_resched(); + } + while (block_bitmap_op(ic, ic->recalc_bitmap, logical_sector + n_sectors - ic->sectors_per_block, + ic->sectors_per_block, BITMAP_OP_TEST_ALL_CLEAR)) { + n_sectors -= ic->sectors_per_block; + cond_resched(); + } + get_area_and_offset(ic, logical_sector, &area, &offset); + } + + DEBUG_print("recalculating: %lx, %lx\n", logical_sector, n_sectors); if (unlikely(++super_counter == RECALC_WRITE_SUPER)) { recalc_write_super(ic); + if (ic->mode == 'B') { + queue_delayed_work(ic->commit_wq, &ic->bitmap_flush_work, ic->bitmap_flush_interval); + } super_counter = 0; } @@ -2162,7 +2405,7 @@ next_chunk: io_req.client = ic->io; io_loc.bdev = ic->dev->bdev; io_loc.sector = get_data_sector(ic, area, offset); - io_loc.count = range.n_sectors; + io_loc.count = n_sectors; r = dm_io(&io_req, 1, &io_loc, NULL); if (unlikely(r)) { @@ -2171,8 +2414,8 @@ next_chunk: } t = ic->recalc_tags; - for (i = 0; i < range.n_sectors; i += ic->sectors_per_block) { - integrity_sector_checksum(ic, range.logical_sector + i, ic->recalc_buffer + (i << SECTOR_SHIFT), t); + for (i = 0; i < n_sectors; i += ic->sectors_per_block) { + integrity_sector_checksum(ic, logical_sector + i, ic->recalc_buffer + (i << SECTOR_SHIFT), t); t += ic->tag_size; } @@ -2184,6 +2427,9 @@ next_chunk: goto err; } +advance_and_next: + cond_resched(); + spin_lock_irq(&ic->endio_wait.lock); remove_range_unlocked(ic, &range); ic->sb->recalc_sector = cpu_to_le64(range.logical_sector + range.n_sectors); @@ -2199,6 +2445,103 @@ unlock_ret: recalc_write_super(ic); } +static void bitmap_block_work(struct work_struct *w) +{ + struct bitmap_block_status *bbs = container_of(w, struct bitmap_block_status, work); + struct dm_integrity_c *ic = bbs->ic; + struct bio *bio; + struct bio_list bio_queue; + struct bio_list waiting; + + bio_list_init(&waiting); + + spin_lock(&bbs->bio_queue_lock); + bio_queue = bbs->bio_queue; + bio_list_init(&bbs->bio_queue); + spin_unlock(&bbs->bio_queue_lock); + + while ((bio = bio_list_pop(&bio_queue))) { + struct dm_integrity_io *dio; + + dio = dm_per_bio_data(bio, sizeof(struct dm_integrity_io)); + + if (block_bitmap_op(ic, ic->may_write_bitmap, dio->range.logical_sector, + dio->range.n_sectors, BITMAP_OP_TEST_ALL_SET)) { + remove_range(ic, &dio->range); + INIT_WORK(&dio->work, integrity_bio_wait); + queue_work(ic->wait_wq, &dio->work); + } else { + block_bitmap_op(ic, ic->journal, dio->range.logical_sector, + dio->range.n_sectors, BITMAP_OP_SET); + bio_list_add(&waiting, bio); + } + } + + if (bio_list_empty(&waiting)) + return; + + rw_journal_sectors(ic, REQ_OP_WRITE, REQ_FUA | REQ_SYNC, + bbs->idx * (BITMAP_BLOCK_SIZE >> SECTOR_SHIFT), + BITMAP_BLOCK_SIZE >> SECTOR_SHIFT, NULL); + + while ((bio = bio_list_pop(&waiting))) { + struct dm_integrity_io *dio = dm_per_bio_data(bio, sizeof(struct dm_integrity_io)); + + block_bitmap_op(ic, ic->may_write_bitmap, dio->range.logical_sector, + dio->range.n_sectors, BITMAP_OP_SET); + + remove_range(ic, &dio->range); + INIT_WORK(&dio->work, integrity_bio_wait); + queue_work(ic->wait_wq, &dio->work); + } + + queue_delayed_work(ic->commit_wq, &ic->bitmap_flush_work, ic->bitmap_flush_interval); +} + +static void bitmap_flush_work(struct work_struct *work) +{ + struct dm_integrity_c *ic = container_of(work, struct dm_integrity_c, bitmap_flush_work.work); + struct dm_integrity_range range; + unsigned long limit; + struct bio *bio; + + dm_integrity_flush_buffers(ic); + + range.logical_sector = 0; + range.n_sectors = ic->provided_data_sectors; + + spin_lock_irq(&ic->endio_wait.lock); + add_new_range_and_wait(ic, &range); + spin_unlock_irq(&ic->endio_wait.lock); + + dm_integrity_flush_buffers(ic); + if (ic->meta_dev) + blkdev_issue_flush(ic->dev->bdev, GFP_NOIO, NULL); + + limit = ic->provided_data_sectors; + if (ic->sb->flags & cpu_to_le32(SB_FLAG_RECALCULATING)) { + limit = le64_to_cpu(ic->sb->recalc_sector) + >> (ic->sb->log2_sectors_per_block + ic->log2_blocks_per_bitmap_bit) + << (ic->sb->log2_sectors_per_block + ic->log2_blocks_per_bitmap_bit); + } + /*DEBUG_print("zeroing journal\n");*/ + block_bitmap_op(ic, ic->journal, 0, limit, BITMAP_OP_CLEAR); + block_bitmap_op(ic, ic->may_write_bitmap, 0, limit, BITMAP_OP_CLEAR); + + rw_journal_sectors(ic, REQ_OP_WRITE, REQ_FUA | REQ_SYNC, 0, + ic->n_bitmap_blocks * (BITMAP_BLOCK_SIZE >> SECTOR_SHIFT), NULL); + + spin_lock_irq(&ic->endio_wait.lock); + remove_range_unlocked(ic, &range); + while (unlikely((bio = bio_list_pop(&ic->synchronous_bios)) != NULL)) { + bio_endio(bio); + spin_unlock_irq(&ic->endio_wait.lock); + spin_lock_irq(&ic->endio_wait.lock); + } + spin_unlock_irq(&ic->endio_wait.lock); +} + + static void init_journal(struct dm_integrity_c *ic, unsigned start_section, unsigned n_sections, unsigned char commit_seq) { @@ -2395,9 +2738,37 @@ clear_journal: init_journal_node(&ic->journal_tree[i]); } +static void dm_integrity_enter_synchronous_mode(struct dm_integrity_c *ic) +{ + DEBUG_print("dm_integrity_enter_synchronous_mode\n"); + + if (ic->mode == 'B') { + ic->bitmap_flush_interval = msecs_to_jiffies(10) + 1; + ic->synchronous_mode = 1; + + cancel_delayed_work_sync(&ic->bitmap_flush_work); + queue_delayed_work(ic->commit_wq, &ic->bitmap_flush_work, 0); + flush_workqueue(ic->commit_wq); + } +} + +static int dm_integrity_reboot(struct notifier_block *n, unsigned long code, void *x) +{ + struct dm_integrity_c *ic = container_of(n, struct dm_integrity_c, reboot_notifier); + + DEBUG_print("dm_integrity_reboot\n"); + + dm_integrity_enter_synchronous_mode(ic); + + return NOTIFY_DONE; +} + static void dm_integrity_postsuspend(struct dm_target *ti) { struct dm_integrity_c *ic = (struct dm_integrity_c *)ti->private; + int r; + + WARN_ON(unregister_reboot_notifier(&ic->reboot_notifier)); del_timer_sync(&ic->autocommit_timer); @@ -2406,6 +2777,9 @@ static void dm_integrity_postsuspend(struct dm_target *ti) if (ic->recalc_wq) drain_workqueue(ic->recalc_wq); + if (ic->mode == 'B') + cancel_delayed_work_sync(&ic->bitmap_flush_work); + queue_work(ic->commit_wq, &ic->commit_work); drain_workqueue(ic->commit_wq); @@ -2416,6 +2790,18 @@ static void dm_integrity_postsuspend(struct dm_target *ti) dm_integrity_flush_buffers(ic); } + if (ic->mode == 'B') { + dm_integrity_flush_buffers(ic); +#if 1 + /* set to 0 to test bitmap replay code */ + init_journal(ic, 0, ic->journal_sections, 0); + ic->sb->flags &= ~cpu_to_le32(SB_FLAG_DIRTY_BITMAP); + r = sync_rw_sb(ic, REQ_OP_WRITE, REQ_FUA); + if (unlikely(r)) + dm_integrity_io_error(ic, "writing superblock", r); +#endif + } + WRITE_ONCE(ic->suspending, 0); BUG_ON(!RB_EMPTY_ROOT(&ic->in_progress)); @@ -2426,11 +2812,70 @@ static void dm_integrity_postsuspend(struct dm_target *ti) static void dm_integrity_resume(struct dm_target *ti) { struct dm_integrity_c *ic = (struct dm_integrity_c *)ti->private; + int r; + DEBUG_print("resume\n"); + + if (ic->sb->flags & cpu_to_le32(SB_FLAG_DIRTY_BITMAP)) { + DEBUG_print("resume dirty_bitmap\n"); + rw_journal_sectors(ic, REQ_OP_READ, 0, 0, + ic->n_bitmap_blocks * (BITMAP_BLOCK_SIZE >> SECTOR_SHIFT), NULL); + if (ic->mode == 'B') { + if (ic->sb->log2_blocks_per_bitmap_bit == ic->log2_blocks_per_bitmap_bit) { + block_bitmap_copy(ic, ic->recalc_bitmap, ic->journal); + block_bitmap_copy(ic, ic->may_write_bitmap, ic->journal); + if (!block_bitmap_op(ic, ic->journal, 0, ic->provided_data_sectors, + BITMAP_OP_TEST_ALL_CLEAR)) { + ic->sb->flags |= cpu_to_le32(SB_FLAG_RECALCULATING); + ic->sb->recalc_sector = cpu_to_le64(0); + } + } else { + DEBUG_print("non-matching blocks_per_bitmap_bit: %u, %u\n", + ic->sb->log2_blocks_per_bitmap_bit, ic->log2_blocks_per_bitmap_bit); + ic->sb->log2_blocks_per_bitmap_bit = ic->log2_blocks_per_bitmap_bit; + block_bitmap_op(ic, ic->recalc_bitmap, 0, ic->provided_data_sectors, BITMAP_OP_SET); + block_bitmap_op(ic, ic->may_write_bitmap, 0, ic->provided_data_sectors, BITMAP_OP_SET); + block_bitmap_op(ic, ic->journal, 0, ic->provided_data_sectors, BITMAP_OP_SET); + rw_journal_sectors(ic, REQ_OP_WRITE, REQ_FUA | REQ_SYNC, 0, + ic->n_bitmap_blocks * (BITMAP_BLOCK_SIZE >> SECTOR_SHIFT), NULL); + ic->sb->flags |= cpu_to_le32(SB_FLAG_RECALCULATING); + ic->sb->recalc_sector = cpu_to_le64(0); + } + } else { + if (!(ic->sb->log2_blocks_per_bitmap_bit == ic->log2_blocks_per_bitmap_bit && + block_bitmap_op(ic, ic->journal, 0, ic->provided_data_sectors, BITMAP_OP_TEST_ALL_CLEAR))) { + ic->sb->flags |= cpu_to_le32(SB_FLAG_RECALCULATING); + ic->sb->recalc_sector = cpu_to_le64(0); + } + init_journal(ic, 0, ic->journal_sections, 0); + replay_journal(ic); + ic->sb->flags &= ~cpu_to_le32(SB_FLAG_DIRTY_BITMAP); + } + r = sync_rw_sb(ic, REQ_OP_WRITE, REQ_FUA); + if (unlikely(r)) + dm_integrity_io_error(ic, "writing superblock", r); + } else { + replay_journal(ic); + if (ic->mode == 'B') { + int mode; + ic->sb->flags |= cpu_to_le32(SB_FLAG_DIRTY_BITMAP); + ic->sb->log2_blocks_per_bitmap_bit = ic->log2_blocks_per_bitmap_bit; + r = sync_rw_sb(ic, REQ_OP_WRITE, REQ_FUA); + if (unlikely(r)) + dm_integrity_io_error(ic, "writing superblock", r); + + mode = ic->recalculate_flag ? BITMAP_OP_SET : BITMAP_OP_CLEAR; + block_bitmap_op(ic, ic->journal, 0, ic->provided_data_sectors, mode); + block_bitmap_op(ic, ic->recalc_bitmap, 0, ic->provided_data_sectors, mode); + block_bitmap_op(ic, ic->may_write_bitmap, 0, ic->provided_data_sectors, mode); + rw_journal_sectors(ic, REQ_OP_WRITE, REQ_FUA | REQ_SYNC, 0, + ic->n_bitmap_blocks * (BITMAP_BLOCK_SIZE >> SECTOR_SHIFT), NULL); + } + } - replay_journal(ic); - - if (ic->recalc_wq && ic->sb->flags & cpu_to_le32(SB_FLAG_RECALCULATING)) { + DEBUG_print("testing recalc: %x\n", ic->sb->flags); + if (ic->sb->flags & cpu_to_le32(SB_FLAG_RECALCULATING)) { __u64 recalc_pos = le64_to_cpu(ic->sb->recalc_sector); + DEBUG_print("recalc pos: %lx / %lx\n", (long)recalc_pos, ic->provided_data_sectors); if (recalc_pos < ic->provided_data_sectors) { queue_work(ic->recalc_wq, &ic->recalc_work); } else if (recalc_pos > ic->provided_data_sectors) { @@ -2438,6 +2883,16 @@ static void dm_integrity_resume(struct dm_target *ti) recalc_write_super(ic); } } + + ic->reboot_notifier.notifier_call = dm_integrity_reboot; + ic->reboot_notifier.next = NULL; + ic->reboot_notifier.priority = INT_MAX - 1; /* be notified after md and before hardware drivers */ + WARN_ON(register_reboot_notifier(&ic->reboot_notifier)); + +#if 0 + /* set to 1 to stress test synchronous mode */ + dm_integrity_enter_synchronous_mode(ic); +#endif } static void dm_integrity_status(struct dm_target *ti, status_type_t type, @@ -2462,10 +2917,14 @@ static void dm_integrity_status(struct dm_target *ti, status_type_t type, __u64 watermark_percentage = (__u64)(ic->journal_entries - ic->free_sectors_threshold) * 100; watermark_percentage += ic->journal_entries / 2; do_div(watermark_percentage, ic->journal_entries); - arg_count = 5; + arg_count = 3; arg_count += !!ic->meta_dev; arg_count += ic->sectors_per_block != 1; arg_count += !!(ic->sb->flags & cpu_to_le32(SB_FLAG_RECALCULATING)); + arg_count += ic->mode == 'J'; + arg_count += ic->mode == 'J'; + arg_count += ic->mode == 'B'; + arg_count += ic->mode == 'B'; arg_count += !!ic->internal_hash_alg.alg_string; arg_count += !!ic->journal_crypt_alg.alg_string; arg_count += !!ic->journal_mac_alg.alg_string; @@ -2475,13 +2934,19 @@ static void dm_integrity_status(struct dm_target *ti, status_type_t type, DMEMIT(" meta_device:%s", ic->meta_dev->name); if (ic->sectors_per_block != 1) DMEMIT(" block_size:%u", ic->sectors_per_block << SECTOR_SHIFT); - if (ic->sb->flags & cpu_to_le32(SB_FLAG_RECALCULATING)) + if (ic->recalculate_flag) DMEMIT(" recalculate"); DMEMIT(" journal_sectors:%u", ic->initial_sectors - SB_SECTORS); DMEMIT(" interleave_sectors:%u", 1U << ic->sb->log2_interleave_sectors); DMEMIT(" buffer_sectors:%u", 1U << ic->log2_buffer_sectors); - DMEMIT(" journal_watermark:%u", (unsigned)watermark_percentage); - DMEMIT(" commit_time:%u", ic->autocommit_msec); + if (ic->mode == 'J') { + DMEMIT(" journal_watermark:%u", (unsigned)watermark_percentage); + DMEMIT(" commit_time:%u", ic->autocommit_msec); + } + if (ic->mode == 'B') { + DMEMIT(" sectors_per_bit:%llu", (unsigned long long)ic->sectors_per_block << ic->log2_blocks_per_bitmap_bit); + DMEMIT(" bitmap_flush_interval:%u", jiffies_to_msecs(ic->bitmap_flush_interval)); + } #define EMIT_ALG(a, n) \ do { \ @@ -2562,7 +3027,7 @@ static int calculate_device_limits(struct dm_integrity_c *ic) if (last_sector < ic->start || last_sector >= ic->meta_device_sectors) return -EINVAL; } else { - __u64 meta_size = ic->provided_data_sectors * ic->tag_size; + __u64 meta_size = (ic->provided_data_sectors >> ic->sb->log2_sectors_per_block) * ic->tag_size; meta_size = (meta_size + ((1U << (ic->log2_buffer_sectors + SECTOR_SHIFT)) - 1)) >> (ic->log2_buffer_sectors + SECTOR_SHIFT); meta_size <<= ic->log2_buffer_sectors; @@ -2659,37 +3124,37 @@ static void dm_integrity_set(struct dm_target *ti, struct dm_integrity_c *ic) blk_queue_max_integrity_segments(disk->queue, UINT_MAX); } -static void dm_integrity_free_page_list(struct dm_integrity_c *ic, struct page_list *pl) +static void dm_integrity_free_page_list(struct page_list *pl) { unsigned i; if (!pl) return; - for (i = 0; i < ic->journal_pages; i++) - if (pl[i].page) - __free_page(pl[i].page); + for (i = 0; pl[i].page; i++) + __free_page(pl[i].page); kvfree(pl); } -static struct page_list *dm_integrity_alloc_page_list(struct dm_integrity_c *ic) +static struct page_list *dm_integrity_alloc_page_list(unsigned n_pages) { - size_t page_list_desc_size = ic->journal_pages * sizeof(struct page_list); struct page_list *pl; unsigned i; - pl = kvmalloc(page_list_desc_size, GFP_KERNEL | __GFP_ZERO); + pl = kvmalloc_array(n_pages + 1, sizeof(struct page_list), GFP_KERNEL | __GFP_ZERO); if (!pl) return NULL; - for (i = 0; i < ic->journal_pages; i++) { + for (i = 0; i < n_pages; i++) { pl[i].page = alloc_page(GFP_KERNEL); if (!pl[i].page) { - dm_integrity_free_page_list(ic, pl); + dm_integrity_free_page_list(pl); return NULL; } if (i) pl[i - 1].next = &pl[i]; } + pl[i].page = NULL; + pl[i].next = NULL; return pl; } @@ -2702,7 +3167,8 @@ static void dm_integrity_free_journal_scatterlist(struct dm_integrity_c *ic, str kvfree(sl); } -static struct scatterlist **dm_integrity_alloc_journal_scatterlist(struct dm_integrity_c *ic, struct page_list *pl) +static struct scatterlist **dm_integrity_alloc_journal_scatterlist(struct dm_integrity_c *ic, + struct page_list *pl) { struct scatterlist **sl; unsigned i; @@ -2721,7 +3187,8 @@ static struct scatterlist **dm_integrity_alloc_journal_scatterlist(struct dm_int unsigned idx; page_list_location(ic, i, 0, &start_index, &start_offset); - page_list_location(ic, i, ic->journal_section_sectors - 1, &end_index, &end_offset); + page_list_location(ic, i, ic->journal_section_sectors - 1, + &end_index, &end_offset); n_pages = (end_index - start_index + 1); @@ -2842,7 +3309,7 @@ static int create_journal(struct dm_integrity_c *ic, char **error) } ic->journal_pages = journal_pages; - ic->journal = dm_integrity_alloc_page_list(ic); + ic->journal = dm_integrity_alloc_page_list(ic->journal_pages); if (!ic->journal) { *error = "Could not allocate memory for journal"; r = -ENOMEM; @@ -2874,7 +3341,7 @@ static int create_journal(struct dm_integrity_c *ic, char **error) DEBUG_print("cipher %s, block size %u iv size %u\n", ic->journal_crypt_alg.alg_string, blocksize, ivsize); - ic->journal_io = dm_integrity_alloc_page_list(ic); + ic->journal_io = dm_integrity_alloc_page_list(ic->journal_pages); if (!ic->journal_io) { *error = "Could not allocate memory for journal io"; r = -ENOMEM; @@ -2898,7 +3365,7 @@ static int create_journal(struct dm_integrity_c *ic, char **error) goto bad; } - ic->journal_xor = dm_integrity_alloc_page_list(ic); + ic->journal_xor = dm_integrity_alloc_page_list(ic->journal_pages); if (!ic->journal_xor) { *error = "Could not allocate memory for journal xor"; r = -ENOMEM; @@ -2922,7 +3389,8 @@ static int create_journal(struct dm_integrity_c *ic, char **error) sg_set_buf(&sg[i], &ic->commit_ids, sizeof ic->commit_ids); memset(crypt_iv, 0x00, ivsize); - skcipher_request_set_crypt(req, sg, sg, PAGE_SIZE * ic->journal_pages + sizeof ic->commit_ids, crypt_iv); + skcipher_request_set_crypt(req, sg, sg, + PAGE_SIZE * ic->journal_pages + sizeof ic->commit_ids, crypt_iv); init_completion(&comp.comp); comp.in_flight = (atomic_t)ATOMIC_INIT(1); if (do_crypt(true, req, &comp)) @@ -3063,7 +3531,7 @@ bad: * device * offset from the start of the device * tag size - * D - direct writes, J - journal writes, R - recovery mode + * D - direct writes, J - journal writes, B - bitmap mode, R - recovery mode * number of optional arguments * optional arguments: * journal_sectors @@ -3071,10 +3539,14 @@ bad: * buffer_sectors * journal_watermark * commit_time + * meta_device + * block_size + * sectors_per_bit + * bitmap_flush_interval * internal_hash * journal_crypt * journal_mac - * block_size + * recalculate */ static int dm_integrity_ctr(struct dm_target *ti, unsigned argc, char **argv) { @@ -3087,10 +3559,13 @@ static int dm_integrity_ctr(struct dm_target *ti, unsigned argc, char **argv) {0, 9, "Invalid number of feature args"}, }; unsigned journal_sectors, interleave_sectors, buffer_sectors, journal_watermark, sync_msec; - bool recalculate; bool should_write_sb; __u64 threshold; unsigned long long start; + __s8 log2_sectors_per_bitmap_bit = -1; + __s8 log2_blocks_per_bitmap_bit; + __u64 bits_in_journal; + __u64 n_bitmap_bits; #define DIRECT_ARGUMENTS 4 @@ -3114,6 +3589,7 @@ static int dm_integrity_ctr(struct dm_target *ti, unsigned argc, char **argv) init_waitqueue_head(&ic->copy_to_journal_wait); init_completion(&ic->crypto_backoff); atomic64_set(&ic->number_of_mismatches, 0); + ic->bitmap_flush_interval = BITMAP_FLUSH_INTERVAL; r = dm_get_device(ti, argv[0], dm_table_get_mode(ti->table), &ic->dev); if (r) { @@ -3136,10 +3612,11 @@ static int dm_integrity_ctr(struct dm_target *ti, unsigned argc, char **argv) } } - if (!strcmp(argv[3], "J") || !strcmp(argv[3], "D") || !strcmp(argv[3], "R")) + if (!strcmp(argv[3], "J") || !strcmp(argv[3], "B") || + !strcmp(argv[3], "D") || !strcmp(argv[3], "R")) { ic->mode = argv[3][0]; - else { - ti->error = "Invalid mode (expecting J, D, R)"; + } else { + ti->error = "Invalid mode (expecting J, B, D, R)"; r = -EINVAL; goto bad; } @@ -3149,7 +3626,6 @@ static int dm_integrity_ctr(struct dm_target *ti, unsigned argc, char **argv) buffer_sectors = DEFAULT_BUFFER_SECTORS; journal_watermark = DEFAULT_JOURNAL_WATERMARK; sync_msec = DEFAULT_SYNC_MSEC; - recalculate = false; ic->sectors_per_block = 1; as.argc = argc - DIRECT_ARGUMENTS; @@ -3161,6 +3637,7 @@ static int dm_integrity_ctr(struct dm_target *ti, unsigned argc, char **argv) while (extra_args--) { const char *opt_string; unsigned val; + unsigned long long llval; opt_string = dm_shift_arg(&as); if (!opt_string) { r = -EINVAL; @@ -3182,7 +3659,8 @@ static int dm_integrity_ctr(struct dm_target *ti, unsigned argc, char **argv) dm_put_device(ti, ic->meta_dev); ic->meta_dev = NULL; } - r = dm_get_device(ti, strchr(opt_string, ':') + 1, dm_table_get_mode(ti->table), &ic->meta_dev); + r = dm_get_device(ti, strchr(opt_string, ':') + 1, + dm_table_get_mode(ti->table), &ic->meta_dev); if (r) { ti->error = "Device lookup failed"; goto bad; @@ -3196,6 +3674,14 @@ static int dm_integrity_ctr(struct dm_target *ti, unsigned argc, char **argv) goto bad; } ic->sectors_per_block = val >> SECTOR_SHIFT; + } else if (sscanf(opt_string, "sectors_per_bit:%llu%c", &llval, &dummy) == 1) { + log2_sectors_per_bitmap_bit = !llval ? 0 : __ilog2_u64(llval); + } else if (sscanf(opt_string, "bitmap_flush_interval:%u%c", &val, &dummy) == 1) { + if (val >= (uint64_t)UINT_MAX * 1000 / HZ) { + r = -EINVAL; + ti->error = "Invalid bitmap_flush_interval argument"; + } + ic->bitmap_flush_interval = msecs_to_jiffies(val); } else if (!strncmp(opt_string, "internal_hash:", strlen("internal_hash:"))) { r = get_alg_and_key(opt_string, &ic->internal_hash_alg, &ti->error, "Invalid internal_hash argument"); @@ -3212,7 +3698,7 @@ static int dm_integrity_ctr(struct dm_target *ti, unsigned argc, char **argv) if (r) goto bad; } else if (!strcmp(opt_string, "recalculate")) { - recalculate = true; + ic->recalculate_flag = true; } else { r = -EINVAL; ti->error = "Invalid argument"; @@ -3228,7 +3714,7 @@ static int dm_integrity_ctr(struct dm_target *ti, unsigned argc, char **argv) if (!journal_sectors) { journal_sectors = min((sector_t)DEFAULT_MAX_JOURNAL_SECTORS, - ic->data_device_sectors >> DEFAULT_JOURNAL_SIZE_FACTOR); + ic->data_device_sectors >> DEFAULT_JOURNAL_SIZE_FACTOR); } if (!buffer_sectors) @@ -3263,6 +3749,12 @@ static int dm_integrity_ctr(struct dm_target *ti, unsigned argc, char **argv) else ic->log2_tag_size = -1; + if (ic->mode == 'B' && !ic->internal_hash) { + r = -EINVAL; + ti->error = "Bitmap mode can be only used with internal hash"; + goto bad; + } + ic->autocommit_jiffies = msecs_to_jiffies(sync_msec); ic->autocommit_msec = sync_msec; timer_setup(&ic->autocommit_timer, autocommit_fn, 0); @@ -3308,7 +3800,7 @@ static int dm_integrity_ctr(struct dm_target *ti, unsigned argc, char **argv) } INIT_WORK(&ic->commit_work, integrity_commit); - if (ic->mode == 'J') { + if (ic->mode == 'J' || ic->mode == 'B') { ic->writer_wq = alloc_workqueue("dm-integrity-writer", WQ_MEM_RECLAIM, 1); if (!ic->writer_wq) { ti->error = "Cannot allocate workqueue"; @@ -3349,7 +3841,7 @@ static int dm_integrity_ctr(struct dm_target *ti, unsigned argc, char **argv) should_write_sb = true; } - if (!ic->sb->version || ic->sb->version > SB_VERSION_2) { + if (!ic->sb->version || ic->sb->version > SB_VERSION_3) { r = -EINVAL; ti->error = "Unknown version"; goto bad; @@ -3409,6 +3901,27 @@ try_smaller_buffer: ti->error = "The device is too small"; goto bad; } + + if (log2_sectors_per_bitmap_bit < 0) + log2_sectors_per_bitmap_bit = __fls(DEFAULT_SECTORS_PER_BITMAP_BIT); + if (log2_sectors_per_bitmap_bit < ic->sb->log2_sectors_per_block) + log2_sectors_per_bitmap_bit = ic->sb->log2_sectors_per_block; + + bits_in_journal = ((__u64)ic->journal_section_sectors * ic->journal_sections) << (SECTOR_SHIFT + 3); + if (bits_in_journal > UINT_MAX) + bits_in_journal = UINT_MAX; + while (bits_in_journal < (ic->provided_data_sectors + ((sector_t)1 << log2_sectors_per_bitmap_bit) - 1) >> log2_sectors_per_bitmap_bit) + log2_sectors_per_bitmap_bit++; + + log2_blocks_per_bitmap_bit = log2_sectors_per_bitmap_bit - ic->sb->log2_sectors_per_block; + ic->log2_blocks_per_bitmap_bit = log2_blocks_per_bitmap_bit; + if (should_write_sb) { + ic->sb->log2_blocks_per_bitmap_bit = log2_blocks_per_bitmap_bit; + } + n_bitmap_bits = ((ic->provided_data_sectors >> ic->sb->log2_sectors_per_block) + + (((sector_t)1 << log2_blocks_per_bitmap_bit) - 1)) >> log2_blocks_per_bitmap_bit; + ic->n_bitmap_blocks = DIV_ROUND_UP(n_bitmap_bits, BITMAP_BLOCK_SIZE * 8); + if (!ic->meta_dev) ic->log2_buffer_sectors = min(ic->log2_buffer_sectors, (__u8)__ffs(ic->metadata_run)); @@ -3433,25 +3946,21 @@ try_smaller_buffer: DEBUG_print(" journal_sections %u\n", (unsigned)le32_to_cpu(ic->sb->journal_sections)); DEBUG_print(" journal_entries %u\n", ic->journal_entries); DEBUG_print(" log2_interleave_sectors %d\n", ic->sb->log2_interleave_sectors); - DEBUG_print(" device_sectors 0x%llx\n", (unsigned long long)ic->device_sectors); + DEBUG_print(" data_device_sectors 0x%llx\n", i_size_read(ic->dev->bdev->bd_inode) >> SECTOR_SHIFT); DEBUG_print(" initial_sectors 0x%x\n", ic->initial_sectors); DEBUG_print(" metadata_run 0x%x\n", ic->metadata_run); DEBUG_print(" log2_metadata_run %d\n", ic->log2_metadata_run); DEBUG_print(" provided_data_sectors 0x%llx (%llu)\n", (unsigned long long)ic->provided_data_sectors, (unsigned long long)ic->provided_data_sectors); DEBUG_print(" log2_buffer_sectors %u\n", ic->log2_buffer_sectors); + DEBUG_print(" bits_in_journal %llu\n", (unsigned long long)bits_in_journal); - if (recalculate && !(ic->sb->flags & cpu_to_le32(SB_FLAG_RECALCULATING))) { + if (ic->recalculate_flag && !(ic->sb->flags & cpu_to_le32(SB_FLAG_RECALCULATING))) { ic->sb->flags |= cpu_to_le32(SB_FLAG_RECALCULATING); ic->sb->recalc_sector = cpu_to_le64(0); } - if (ic->sb->flags & cpu_to_le32(SB_FLAG_RECALCULATING)) { - if (!ic->internal_hash) { - r = -EINVAL; - ti->error = "Recalculate is only valid with internal hash"; - goto bad; - } + if (ic->internal_hash) { ic->recalc_wq = alloc_workqueue("dm-integrity-recalc", WQ_MEM_RECLAIM, 1); if (!ic->recalc_wq ) { ti->error = "Cannot allocate workqueue"; @@ -3488,6 +3997,45 @@ try_smaller_buffer: r = create_journal(ic, &ti->error); if (r) goto bad; + + } + + if (ic->mode == 'B') { + unsigned i; + unsigned n_bitmap_pages = DIV_ROUND_UP(ic->n_bitmap_blocks, PAGE_SIZE / BITMAP_BLOCK_SIZE); + + ic->recalc_bitmap = dm_integrity_alloc_page_list(n_bitmap_pages); + if (!ic->recalc_bitmap) { + r = -ENOMEM; + goto bad; + } + ic->may_write_bitmap = dm_integrity_alloc_page_list(n_bitmap_pages); + if (!ic->may_write_bitmap) { + r = -ENOMEM; + goto bad; + } + ic->bbs = kvmalloc_array(ic->n_bitmap_blocks, sizeof(struct bitmap_block_status), GFP_KERNEL); + if (!ic->bbs) { + r = -ENOMEM; + goto bad; + } + INIT_DELAYED_WORK(&ic->bitmap_flush_work, bitmap_flush_work); + for (i = 0; i < ic->n_bitmap_blocks; i++) { + struct bitmap_block_status *bbs = &ic->bbs[i]; + unsigned sector, pl_index, pl_offset; + + INIT_WORK(&bbs->work, bitmap_block_work); + bbs->ic = ic; + bbs->idx = i; + bio_list_init(&bbs->bio_queue); + spin_lock_init(&bbs->bio_queue_lock); + + sector = i * (BITMAP_BLOCK_SIZE >> SECTOR_SHIFT); + pl_index = sector >> (PAGE_SHIFT - SECTOR_SHIFT); + pl_offset = (sector << SECTOR_SHIFT) & (PAGE_SIZE - 1); + + bbs->bitmap = lowmem_page_address(ic->journal[pl_index].page) + pl_offset; + } } if (should_write_sb) { @@ -3512,6 +4060,17 @@ try_smaller_buffer: if (r) goto bad; } + if (ic->mode == 'B') { + unsigned max_io_len = ((sector_t)ic->sectors_per_block << ic->log2_blocks_per_bitmap_bit) * (BITMAP_BLOCK_SIZE * 8); + if (!max_io_len) + max_io_len = 1U << 31; + DEBUG_print("max_io_len: old %u, new %u\n", ti->max_io_len, max_io_len); + if (!ti->max_io_len || ti->max_io_len > max_io_len) { + r = dm_set_target_max_io_len(ti, max_io_len); + if (r) + goto bad; + } + } if (!ic->internal_hash) dm_integrity_set(ti, ic); @@ -3520,6 +4079,7 @@ try_smaller_buffer: ti->flush_supported = true; return 0; + bad: dm_integrity_dtr(ti); return r; @@ -3542,10 +4102,9 @@ static void dm_integrity_dtr(struct dm_target *ti) destroy_workqueue(ic->writer_wq); if (ic->recalc_wq) destroy_workqueue(ic->recalc_wq); - if (ic->recalc_buffer) - vfree(ic->recalc_buffer); - if (ic->recalc_tags) - kvfree(ic->recalc_tags); + vfree(ic->recalc_buffer); + kvfree(ic->recalc_tags); + kvfree(ic->bbs); if (ic->bufio) dm_bufio_client_destroy(ic->bufio); mempool_exit(&ic->journal_io_mempool); @@ -3555,9 +4114,11 @@ static void dm_integrity_dtr(struct dm_target *ti) dm_put_device(ti, ic->dev); if (ic->meta_dev) dm_put_device(ti, ic->meta_dev); - dm_integrity_free_page_list(ic, ic->journal); - dm_integrity_free_page_list(ic, ic->journal_io); - dm_integrity_free_page_list(ic, ic->journal_xor); + dm_integrity_free_page_list(ic->journal); + dm_integrity_free_page_list(ic->journal_io); + dm_integrity_free_page_list(ic->journal_xor); + dm_integrity_free_page_list(ic->recalc_bitmap); + dm_integrity_free_page_list(ic->may_write_bitmap); if (ic->journal_scatterlist) dm_integrity_free_journal_scatterlist(ic, ic->journal_scatterlist); if (ic->journal_io_scatterlist) @@ -3595,7 +4156,7 @@ static void dm_integrity_dtr(struct dm_target *ti) static struct target_type integrity_target = { .name = "integrity", - .version = {1, 2, 0}, + .version = {1, 3, 0}, .module = THIS_MODULE, .features = DM_TARGET_SINGLETON | DM_TARGET_INTEGRITY, .ctr = dm_integrity_ctr, diff --git a/drivers/md/dm-ioctl.c b/drivers/md/dm-ioctl.c index c740153b4e52..1e03bc89e20f 100644 --- a/drivers/md/dm-ioctl.c +++ b/drivers/md/dm-ioctl.c @@ -2069,7 +2069,7 @@ int __init dm_early_create(struct dm_ioctl *dmi, /* alloc table */ r = dm_table_create(&t, get_mode(dmi), dmi->target_count, md); if (r) - goto err_destroy_dm; + goto err_hash_remove; /* add targets */ for (i = 0; i < dmi->target_count; i++) { @@ -2116,6 +2116,10 @@ int __init dm_early_create(struct dm_ioctl *dmi, err_destroy_table: dm_table_destroy(t); +err_hash_remove: + (void) __hash_remove(__get_name_cell(dmi->name)); + /* release reference from __get_name_cell */ + dm_put(md); err_destroy_dm: dm_put(md); dm_destroy(md); diff --git a/drivers/md/dm-mpath.c b/drivers/md/dm-mpath.c index 2ee5e357a0a7..dbcc1e41cd57 100644 --- a/drivers/md/dm-mpath.c +++ b/drivers/md/dm-mpath.c @@ -544,8 +544,23 @@ static int multipath_clone_and_map(struct dm_target *ti, struct request *rq, return DM_MAPIO_REMAPPED; } -static void multipath_release_clone(struct request *clone) +static void multipath_release_clone(struct request *clone, + union map_info *map_context) { + if (unlikely(map_context)) { + /* + * non-NULL map_context means caller is still map + * method; must undo multipath_clone_and_map() + */ + struct dm_mpath_io *mpio = get_mpio(map_context); + struct pgpath *pgpath = mpio->pgpath; + + if (pgpath && pgpath->pg->ps.type->end_io) + pgpath->pg->ps.type->end_io(&pgpath->pg->ps, + &pgpath->path, + mpio->nr_bytes); + } + blk_put_request(clone); } @@ -882,6 +897,7 @@ static struct pgpath *parse_path(struct dm_arg_set *as, struct path_selector *ps if (attached_handler_name || m->hw_handler_name) { INIT_DELAYED_WORK(&p->activate_path, activate_path_work); r = setup_scsi_dh(p->path.dev->bdev, m, &attached_handler_name, &ti->error); + kfree(attached_handler_name); if (r) { dm_put_device(ti, p->path.dev); goto bad; @@ -896,7 +912,6 @@ static struct pgpath *parse_path(struct dm_arg_set *as, struct path_selector *ps return p; bad: - kfree(attached_handler_name); free_pgpath(p); return ERR_PTR(r); } diff --git a/drivers/md/dm-rq.c b/drivers/md/dm-rq.c index b66745bd08bb..5f7063f05ae0 100644 --- a/drivers/md/dm-rq.c +++ b/drivers/md/dm-rq.c @@ -168,7 +168,7 @@ static void dm_end_request(struct request *clone, blk_status_t error) struct request *rq = tio->orig; blk_rq_unprep_clone(clone); - tio->ti->type->release_clone_rq(clone); + tio->ti->type->release_clone_rq(clone, NULL); rq_end_stats(md, rq); blk_mq_end_request(rq, error); @@ -201,7 +201,7 @@ static void dm_requeue_original_request(struct dm_rq_target_io *tio, bool delay_ rq_end_stats(md, rq); if (tio->clone) { blk_rq_unprep_clone(tio->clone); - tio->ti->type->release_clone_rq(tio->clone); + tio->ti->type->release_clone_rq(tio->clone, NULL); } dm_mq_delay_requeue_request(rq, delay_ms); @@ -398,7 +398,7 @@ static int map_request(struct dm_rq_target_io *tio) case DM_MAPIO_REMAPPED: if (setup_clone(clone, rq, tio, GFP_ATOMIC)) { /* -ENOMEM */ - ti->type->release_clone_rq(clone); + ti->type->release_clone_rq(clone, &tio->info); return DM_MAPIO_REQUEUE; } @@ -408,7 +408,7 @@ static int map_request(struct dm_rq_target_io *tio) ret = dm_dispatch_clone_request(clone, rq); if (ret == BLK_STS_RESOURCE || ret == BLK_STS_DEV_RESOURCE) { blk_rq_unprep_clone(clone); - tio->ti->type->release_clone_rq(clone); + tio->ti->type->release_clone_rq(clone, &tio->info); tio->clone = NULL; return DM_MAPIO_REQUEUE; } diff --git a/drivers/md/dm-snap.c b/drivers/md/dm-snap.c index a168963b757d..3107f2b1988b 100644 --- a/drivers/md/dm-snap.c +++ b/drivers/md/dm-snap.c @@ -13,6 +13,7 @@ #include <linux/init.h> #include <linux/kdev_t.h> #include <linux/list.h> +#include <linux/list_bl.h> #include <linux/mempool.h> #include <linux/module.h> #include <linux/slab.h> @@ -44,11 +45,11 @@ static const char dm_snapshot_merge_target_name[] = "snapshot-merge"; struct dm_exception_table { uint32_t hash_mask; unsigned hash_shift; - struct list_head *table; + struct hlist_bl_head *table; }; struct dm_snapshot { - struct mutex lock; + struct rw_semaphore lock; struct dm_dev *origin; struct dm_dev *cow; @@ -76,7 +77,9 @@ struct dm_snapshot { atomic_t pending_exceptions_count; - /* Protected by "lock" */ + spinlock_t pe_allocation_lock; + + /* Protected by "pe_allocation_lock" */ sector_t exception_start_sequence; /* Protected by kcopyd single-threaded callback */ @@ -457,9 +460,9 @@ static int __find_snapshots_sharing_cow(struct dm_snapshot *snap, if (!bdev_equal(s->cow->bdev, snap->cow->bdev)) continue; - mutex_lock(&s->lock); + down_read(&s->lock); active = s->active; - mutex_unlock(&s->lock); + up_read(&s->lock); if (active) { if (snap_src) @@ -618,6 +621,36 @@ static void unregister_snapshot(struct dm_snapshot *s) * The lowest hash_shift bits of the chunk number are ignored, allowing * some consecutive chunks to be grouped together. */ +static uint32_t exception_hash(struct dm_exception_table *et, chunk_t chunk); + +/* Lock to protect access to the completed and pending exception hash tables. */ +struct dm_exception_table_lock { + struct hlist_bl_head *complete_slot; + struct hlist_bl_head *pending_slot; +}; + +static void dm_exception_table_lock_init(struct dm_snapshot *s, chunk_t chunk, + struct dm_exception_table_lock *lock) +{ + struct dm_exception_table *complete = &s->complete; + struct dm_exception_table *pending = &s->pending; + + lock->complete_slot = &complete->table[exception_hash(complete, chunk)]; + lock->pending_slot = &pending->table[exception_hash(pending, chunk)]; +} + +static void dm_exception_table_lock(struct dm_exception_table_lock *lock) +{ + hlist_bl_lock(lock->complete_slot); + hlist_bl_lock(lock->pending_slot); +} + +static void dm_exception_table_unlock(struct dm_exception_table_lock *lock) +{ + hlist_bl_unlock(lock->pending_slot); + hlist_bl_unlock(lock->complete_slot); +} + static int dm_exception_table_init(struct dm_exception_table *et, uint32_t size, unsigned hash_shift) { @@ -625,12 +658,12 @@ static int dm_exception_table_init(struct dm_exception_table *et, et->hash_shift = hash_shift; et->hash_mask = size - 1; - et->table = dm_vcalloc(size, sizeof(struct list_head)); + et->table = dm_vcalloc(size, sizeof(struct hlist_bl_head)); if (!et->table) return -ENOMEM; for (i = 0; i < size; i++) - INIT_LIST_HEAD(et->table + i); + INIT_HLIST_BL_HEAD(et->table + i); return 0; } @@ -638,15 +671,16 @@ static int dm_exception_table_init(struct dm_exception_table *et, static void dm_exception_table_exit(struct dm_exception_table *et, struct kmem_cache *mem) { - struct list_head *slot; - struct dm_exception *ex, *next; + struct hlist_bl_head *slot; + struct dm_exception *ex; + struct hlist_bl_node *pos, *n; int i, size; size = et->hash_mask + 1; for (i = 0; i < size; i++) { slot = et->table + i; - list_for_each_entry_safe (ex, next, slot, hash_list) + hlist_bl_for_each_entry_safe(ex, pos, n, slot, hash_list) kmem_cache_free(mem, ex); } @@ -660,7 +694,7 @@ static uint32_t exception_hash(struct dm_exception_table *et, chunk_t chunk) static void dm_remove_exception(struct dm_exception *e) { - list_del(&e->hash_list); + hlist_bl_del(&e->hash_list); } /* @@ -670,11 +704,12 @@ static void dm_remove_exception(struct dm_exception *e) static struct dm_exception *dm_lookup_exception(struct dm_exception_table *et, chunk_t chunk) { - struct list_head *slot; + struct hlist_bl_head *slot; + struct hlist_bl_node *pos; struct dm_exception *e; slot = &et->table[exception_hash(et, chunk)]; - list_for_each_entry (e, slot, hash_list) + hlist_bl_for_each_entry(e, pos, slot, hash_list) if (chunk >= e->old_chunk && chunk <= e->old_chunk + dm_consecutive_chunk_count(e)) return e; @@ -721,7 +756,8 @@ static void free_pending_exception(struct dm_snap_pending_exception *pe) static void dm_insert_exception(struct dm_exception_table *eh, struct dm_exception *new_e) { - struct list_head *l; + struct hlist_bl_head *l; + struct hlist_bl_node *pos; struct dm_exception *e = NULL; l = &eh->table[exception_hash(eh, new_e->old_chunk)]; @@ -731,7 +767,7 @@ static void dm_insert_exception(struct dm_exception_table *eh, goto out; /* List is ordered by old_chunk */ - list_for_each_entry_reverse(e, l, hash_list) { + hlist_bl_for_each_entry(e, pos, l, hash_list) { /* Insert after an existing chunk? */ if (new_e->old_chunk == (e->old_chunk + dm_consecutive_chunk_count(e) + 1) && @@ -752,12 +788,24 @@ static void dm_insert_exception(struct dm_exception_table *eh, return; } - if (new_e->old_chunk > e->old_chunk) + if (new_e->old_chunk < e->old_chunk) break; } out: - list_add(&new_e->hash_list, e ? &e->hash_list : l); + if (!e) { + /* + * Either the table doesn't support consecutive chunks or slot + * l is empty. + */ + hlist_bl_add_head(&new_e->hash_list, l); + } else if (new_e->old_chunk < e->old_chunk) { + /* Add before an existing exception */ + hlist_bl_add_before(&new_e->hash_list, &e->hash_list); + } else { + /* Add to l's tail: e is the last exception in this slot */ + hlist_bl_add_behind(&new_e->hash_list, &e->hash_list); + } } /* @@ -766,6 +814,7 @@ out: */ static int dm_add_exception(void *context, chunk_t old, chunk_t new) { + struct dm_exception_table_lock lock; struct dm_snapshot *s = context; struct dm_exception *e; @@ -778,7 +827,17 @@ static int dm_add_exception(void *context, chunk_t old, chunk_t new) /* Consecutive_count is implicitly initialised to zero */ e->new_chunk = new; + /* + * Although there is no need to lock access to the exception tables + * here, if we don't then hlist_bl_add_head(), called by + * dm_insert_exception(), will complain about accessing the + * corresponding list without locking it first. + */ + dm_exception_table_lock_init(s, old, &lock); + + dm_exception_table_lock(&lock); dm_insert_exception(&s->complete, e); + dm_exception_table_unlock(&lock); return 0; } @@ -807,7 +866,7 @@ static int calc_max_buckets(void) { /* use a fixed size of 2MB */ unsigned long mem = 2 * 1024 * 1024; - mem /= sizeof(struct list_head); + mem /= sizeof(struct hlist_bl_head); return mem; } @@ -927,7 +986,7 @@ static int remove_single_exception_chunk(struct dm_snapshot *s) int r; chunk_t old_chunk = s->first_merging_chunk + s->num_merging_chunks - 1; - mutex_lock(&s->lock); + down_write(&s->lock); /* * Process chunks (and associated exceptions) in reverse order @@ -942,7 +1001,7 @@ static int remove_single_exception_chunk(struct dm_snapshot *s) b = __release_queued_bios_after_merge(s); out: - mutex_unlock(&s->lock); + up_write(&s->lock); if (b) flush_bios(b); @@ -1001,9 +1060,9 @@ static void snapshot_merge_next_chunks(struct dm_snapshot *s) if (linear_chunks < 0) { DMERR("Read error in exception store: " "shutting down merge"); - mutex_lock(&s->lock); + down_write(&s->lock); s->merge_failed = 1; - mutex_unlock(&s->lock); + up_write(&s->lock); } goto shut; } @@ -1044,10 +1103,10 @@ static void snapshot_merge_next_chunks(struct dm_snapshot *s) previous_count = read_pending_exceptions_done_count(); } - mutex_lock(&s->lock); + down_write(&s->lock); s->first_merging_chunk = old_chunk; s->num_merging_chunks = linear_chunks; - mutex_unlock(&s->lock); + up_write(&s->lock); /* Wait until writes to all 'linear_chunks' drain */ for (i = 0; i < linear_chunks; i++) @@ -1089,10 +1148,10 @@ static void merge_callback(int read_err, unsigned long write_err, void *context) return; shut: - mutex_lock(&s->lock); + down_write(&s->lock); s->merge_failed = 1; b = __release_queued_bios_after_merge(s); - mutex_unlock(&s->lock); + up_write(&s->lock); error_bios(b); merge_shutdown(s); @@ -1188,10 +1247,11 @@ static int snapshot_ctr(struct dm_target *ti, unsigned int argc, char **argv) s->snapshot_overflowed = 0; s->active = 0; atomic_set(&s->pending_exceptions_count, 0); + spin_lock_init(&s->pe_allocation_lock); s->exception_start_sequence = 0; s->exception_complete_sequence = 0; s->out_of_order_tree = RB_ROOT; - mutex_init(&s->lock); + init_rwsem(&s->lock); INIT_LIST_HEAD(&s->list); spin_lock_init(&s->pe_lock); s->state_bits = 0; @@ -1357,9 +1417,9 @@ static void snapshot_dtr(struct dm_target *ti) /* Check whether exception handover must be cancelled */ (void) __find_snapshots_sharing_cow(s, &snap_src, &snap_dest, NULL); if (snap_src && snap_dest && (s == snap_src)) { - mutex_lock(&snap_dest->lock); + down_write(&snap_dest->lock); snap_dest->valid = 0; - mutex_unlock(&snap_dest->lock); + up_write(&snap_dest->lock); DMERR("Cancelling snapshot handover."); } up_read(&_origins_lock); @@ -1390,8 +1450,6 @@ static void snapshot_dtr(struct dm_target *ti) dm_exception_store_destroy(s->store); - mutex_destroy(&s->lock); - dm_put_device(ti, s->cow); dm_put_device(ti, s->origin); @@ -1467,6 +1525,13 @@ static void __invalidate_snapshot(struct dm_snapshot *s, int err) dm_table_event(s->ti->table); } +static void invalidate_snapshot(struct dm_snapshot *s, int err) +{ + down_write(&s->lock); + __invalidate_snapshot(s, err); + up_write(&s->lock); +} + static void pending_complete(void *context, int success) { struct dm_snap_pending_exception *pe = context; @@ -1475,43 +1540,63 @@ static void pending_complete(void *context, int success) struct bio *origin_bios = NULL; struct bio *snapshot_bios = NULL; struct bio *full_bio = NULL; + struct dm_exception_table_lock lock; int error = 0; + dm_exception_table_lock_init(s, pe->e.old_chunk, &lock); + if (!success) { /* Read/write error - snapshot is unusable */ - mutex_lock(&s->lock); - __invalidate_snapshot(s, -EIO); + invalidate_snapshot(s, -EIO); error = 1; + + dm_exception_table_lock(&lock); goto out; } e = alloc_completed_exception(GFP_NOIO); if (!e) { - mutex_lock(&s->lock); - __invalidate_snapshot(s, -ENOMEM); + invalidate_snapshot(s, -ENOMEM); error = 1; + + dm_exception_table_lock(&lock); goto out; } *e = pe->e; - mutex_lock(&s->lock); + down_read(&s->lock); + dm_exception_table_lock(&lock); if (!s->valid) { + up_read(&s->lock); free_completed_exception(e); error = 1; + goto out; } - /* Check for conflicting reads */ - __check_for_conflicting_io(s, pe->e.old_chunk); - /* - * Add a proper exception, and remove the - * in-flight exception from the list. + * Add a proper exception. After inserting the completed exception all + * subsequent snapshot reads to this chunk will be redirected to the + * COW device. This ensures that we do not starve. Moreover, as long + * as the pending exception exists, neither origin writes nor snapshot + * merging can overwrite the chunk in origin. */ dm_insert_exception(&s->complete, e); + up_read(&s->lock); + + /* Wait for conflicting reads to drain */ + if (__chunk_is_tracked(s, pe->e.old_chunk)) { + dm_exception_table_unlock(&lock); + __check_for_conflicting_io(s, pe->e.old_chunk); + dm_exception_table_lock(&lock); + } out: + /* Remove the in-flight exception from the list */ dm_remove_exception(&pe->e); + + dm_exception_table_unlock(&lock); + snapshot_bios = bio_list_get(&pe->snapshot_bios); origin_bios = bio_list_get(&pe->origin_bios); full_bio = pe->full_bio; @@ -1519,8 +1604,6 @@ out: full_bio->bi_end_io = pe->full_bio_end_io; increment_pending_exceptions_done_count(); - mutex_unlock(&s->lock); - /* Submit any pending write bios */ if (error) { if (full_bio) @@ -1660,43 +1743,59 @@ __lookup_pending_exception(struct dm_snapshot *s, chunk_t chunk) } /* - * Looks to see if this snapshot already has a pending exception - * for this chunk, otherwise it allocates a new one and inserts - * it into the pending table. + * Inserts a pending exception into the pending table. * - * NOTE: a write lock must be held on snap->lock before calling - * this. + * NOTE: a write lock must be held on the chunk's pending exception table slot + * before calling this. */ static struct dm_snap_pending_exception * -__find_pending_exception(struct dm_snapshot *s, - struct dm_snap_pending_exception *pe, chunk_t chunk) +__insert_pending_exception(struct dm_snapshot *s, + struct dm_snap_pending_exception *pe, chunk_t chunk) { - struct dm_snap_pending_exception *pe2; - - pe2 = __lookup_pending_exception(s, chunk); - if (pe2) { - free_pending_exception(pe); - return pe2; - } - pe->e.old_chunk = chunk; bio_list_init(&pe->origin_bios); bio_list_init(&pe->snapshot_bios); pe->started = 0; pe->full_bio = NULL; + spin_lock(&s->pe_allocation_lock); if (s->store->type->prepare_exception(s->store, &pe->e)) { + spin_unlock(&s->pe_allocation_lock); free_pending_exception(pe); return NULL; } pe->exception_sequence = s->exception_start_sequence++; + spin_unlock(&s->pe_allocation_lock); dm_insert_exception(&s->pending, &pe->e); return pe; } +/* + * Looks to see if this snapshot already has a pending exception + * for this chunk, otherwise it allocates a new one and inserts + * it into the pending table. + * + * NOTE: a write lock must be held on the chunk's pending exception table slot + * before calling this. + */ +static struct dm_snap_pending_exception * +__find_pending_exception(struct dm_snapshot *s, + struct dm_snap_pending_exception *pe, chunk_t chunk) +{ + struct dm_snap_pending_exception *pe2; + + pe2 = __lookup_pending_exception(s, chunk); + if (pe2) { + free_pending_exception(pe); + return pe2; + } + + return __insert_pending_exception(s, pe, chunk); +} + static void remap_exception(struct dm_snapshot *s, struct dm_exception *e, struct bio *bio, chunk_t chunk) { @@ -1714,6 +1813,7 @@ static int snapshot_map(struct dm_target *ti, struct bio *bio) int r = DM_MAPIO_REMAPPED; chunk_t chunk; struct dm_snap_pending_exception *pe = NULL; + struct dm_exception_table_lock lock; init_tracked_chunk(bio); @@ -1723,13 +1823,15 @@ static int snapshot_map(struct dm_target *ti, struct bio *bio) } chunk = sector_to_chunk(s->store, bio->bi_iter.bi_sector); + dm_exception_table_lock_init(s, chunk, &lock); /* Full snapshots are not usable */ /* To get here the table must be live so s->active is always set. */ if (!s->valid) return DM_MAPIO_KILL; - mutex_lock(&s->lock); + down_read(&s->lock); + dm_exception_table_lock(&lock); if (!s->valid || (unlikely(s->snapshot_overflowed) && bio_data_dir(bio) == WRITE)) { @@ -1752,15 +1854,9 @@ static int snapshot_map(struct dm_target *ti, struct bio *bio) if (bio_data_dir(bio) == WRITE) { pe = __lookup_pending_exception(s, chunk); if (!pe) { - mutex_unlock(&s->lock); + dm_exception_table_unlock(&lock); pe = alloc_pending_exception(s); - mutex_lock(&s->lock); - - if (!s->valid || s->snapshot_overflowed) { - free_pending_exception(pe); - r = DM_MAPIO_KILL; - goto out_unlock; - } + dm_exception_table_lock(&lock); e = dm_lookup_exception(&s->complete, chunk); if (e) { @@ -1771,13 +1867,22 @@ static int snapshot_map(struct dm_target *ti, struct bio *bio) pe = __find_pending_exception(s, pe, chunk); if (!pe) { + dm_exception_table_unlock(&lock); + up_read(&s->lock); + + down_write(&s->lock); + if (s->store->userspace_supports_overflow) { - s->snapshot_overflowed = 1; - DMERR("Snapshot overflowed: Unable to allocate exception."); + if (s->valid && !s->snapshot_overflowed) { + s->snapshot_overflowed = 1; + DMERR("Snapshot overflowed: Unable to allocate exception."); + } } else __invalidate_snapshot(s, -ENOMEM); + up_write(&s->lock); + r = DM_MAPIO_KILL; - goto out_unlock; + goto out; } } @@ -1789,7 +1894,10 @@ static int snapshot_map(struct dm_target *ti, struct bio *bio) bio->bi_iter.bi_size == (s->store->chunk_size << SECTOR_SHIFT)) { pe->started = 1; - mutex_unlock(&s->lock); + + dm_exception_table_unlock(&lock); + up_read(&s->lock); + start_full_bio(pe, bio); goto out; } @@ -1797,9 +1905,12 @@ static int snapshot_map(struct dm_target *ti, struct bio *bio) bio_list_add(&pe->snapshot_bios, bio); if (!pe->started) { - /* this is protected by snap->lock */ + /* this is protected by the exception table lock */ pe->started = 1; - mutex_unlock(&s->lock); + + dm_exception_table_unlock(&lock); + up_read(&s->lock); + start_copy(pe); goto out; } @@ -1809,7 +1920,8 @@ static int snapshot_map(struct dm_target *ti, struct bio *bio) } out_unlock: - mutex_unlock(&s->lock); + dm_exception_table_unlock(&lock); + up_read(&s->lock); out: return r; } @@ -1845,7 +1957,7 @@ static int snapshot_merge_map(struct dm_target *ti, struct bio *bio) chunk = sector_to_chunk(s->store, bio->bi_iter.bi_sector); - mutex_lock(&s->lock); + down_write(&s->lock); /* Full merging snapshots are redirected to the origin */ if (!s->valid) @@ -1876,12 +1988,12 @@ redirect_to_origin: bio_set_dev(bio, s->origin->bdev); if (bio_data_dir(bio) == WRITE) { - mutex_unlock(&s->lock); + up_write(&s->lock); return do_origin(s->origin, bio); } out_unlock: - mutex_unlock(&s->lock); + up_write(&s->lock); return r; } @@ -1913,7 +2025,7 @@ static int snapshot_preresume(struct dm_target *ti) down_read(&_origins_lock); (void) __find_snapshots_sharing_cow(s, &snap_src, &snap_dest, NULL); if (snap_src && snap_dest) { - mutex_lock(&snap_src->lock); + down_read(&snap_src->lock); if (s == snap_src) { DMERR("Unable to resume snapshot source until " "handover completes."); @@ -1923,7 +2035,7 @@ static int snapshot_preresume(struct dm_target *ti) "source is suspended."); r = -EINVAL; } - mutex_unlock(&snap_src->lock); + up_read(&snap_src->lock); } up_read(&_origins_lock); @@ -1969,11 +2081,11 @@ static void snapshot_resume(struct dm_target *ti) (void) __find_snapshots_sharing_cow(s, &snap_src, &snap_dest, NULL); if (snap_src && snap_dest) { - mutex_lock(&snap_src->lock); - mutex_lock_nested(&snap_dest->lock, SINGLE_DEPTH_NESTING); + down_write(&snap_src->lock); + down_write_nested(&snap_dest->lock, SINGLE_DEPTH_NESTING); __handover_exceptions(snap_src, snap_dest); - mutex_unlock(&snap_dest->lock); - mutex_unlock(&snap_src->lock); + up_write(&snap_dest->lock); + up_write(&snap_src->lock); } up_read(&_origins_lock); @@ -1988,9 +2100,9 @@ static void snapshot_resume(struct dm_target *ti) /* Now we have correct chunk size, reregister */ reregister_snapshot(s); - mutex_lock(&s->lock); + down_write(&s->lock); s->active = 1; - mutex_unlock(&s->lock); + up_write(&s->lock); } static uint32_t get_origin_minimum_chunksize(struct block_device *bdev) @@ -2030,7 +2142,7 @@ static void snapshot_status(struct dm_target *ti, status_type_t type, switch (type) { case STATUSTYPE_INFO: - mutex_lock(&snap->lock); + down_write(&snap->lock); if (!snap->valid) DMEMIT("Invalid"); @@ -2055,7 +2167,7 @@ static void snapshot_status(struct dm_target *ti, status_type_t type, DMEMIT("Unknown"); } - mutex_unlock(&snap->lock); + up_write(&snap->lock); break; @@ -2107,9 +2219,10 @@ static int __origin_write(struct list_head *snapshots, sector_t sector, int r = DM_MAPIO_REMAPPED; struct dm_snapshot *snap; struct dm_exception *e; - struct dm_snap_pending_exception *pe; + struct dm_snap_pending_exception *pe, *pe2; struct dm_snap_pending_exception *pe_to_start_now = NULL; struct dm_snap_pending_exception *pe_to_start_last = NULL; + struct dm_exception_table_lock lock; chunk_t chunk; /* Do all the snapshots on this origin */ @@ -2121,52 +2234,59 @@ static int __origin_write(struct list_head *snapshots, sector_t sector, if (dm_target_is_snapshot_merge(snap->ti)) continue; - mutex_lock(&snap->lock); - - /* Only deal with valid and active snapshots */ - if (!snap->valid || !snap->active) - goto next_snapshot; - /* Nothing to do if writing beyond end of snapshot */ if (sector >= dm_table_get_size(snap->ti->table)) - goto next_snapshot; + continue; /* * Remember, different snapshots can have * different chunk sizes. */ chunk = sector_to_chunk(snap->store, sector); + dm_exception_table_lock_init(snap, chunk, &lock); - /* - * Check exception table to see if block - * is already remapped in this snapshot - * and trigger an exception if not. - */ - e = dm_lookup_exception(&snap->complete, chunk); - if (e) + down_read(&snap->lock); + dm_exception_table_lock(&lock); + + /* Only deal with valid and active snapshots */ + if (!snap->valid || !snap->active) goto next_snapshot; pe = __lookup_pending_exception(snap, chunk); if (!pe) { - mutex_unlock(&snap->lock); - pe = alloc_pending_exception(snap); - mutex_lock(&snap->lock); - - if (!snap->valid) { - free_pending_exception(pe); - goto next_snapshot; - } - + /* + * Check exception table to see if block is already + * remapped in this snapshot and trigger an exception + * if not. + */ e = dm_lookup_exception(&snap->complete, chunk); - if (e) { - free_pending_exception(pe); + if (e) goto next_snapshot; - } - pe = __find_pending_exception(snap, pe, chunk); - if (!pe) { - __invalidate_snapshot(snap, -ENOMEM); - goto next_snapshot; + dm_exception_table_unlock(&lock); + pe = alloc_pending_exception(snap); + dm_exception_table_lock(&lock); + + pe2 = __lookup_pending_exception(snap, chunk); + + if (!pe2) { + e = dm_lookup_exception(&snap->complete, chunk); + if (e) { + free_pending_exception(pe); + goto next_snapshot; + } + + pe = __insert_pending_exception(snap, pe, chunk); + if (!pe) { + dm_exception_table_unlock(&lock); + up_read(&snap->lock); + + invalidate_snapshot(snap, -ENOMEM); + continue; + } + } else { + free_pending_exception(pe); + pe = pe2; } } @@ -2193,7 +2313,8 @@ static int __origin_write(struct list_head *snapshots, sector_t sector, } next_snapshot: - mutex_unlock(&snap->lock); + dm_exception_table_unlock(&lock); + up_read(&snap->lock); if (pe_to_start_now) { start_copy(pe_to_start_now); diff --git a/drivers/md/dm-target.c b/drivers/md/dm-target.c index 314d17ca6466..64dd0b34fcf4 100644 --- a/drivers/md/dm-target.c +++ b/drivers/md/dm-target.c @@ -136,7 +136,8 @@ static int io_err_clone_and_map_rq(struct dm_target *ti, struct request *rq, return DM_MAPIO_KILL; } -static void io_err_release_clone_rq(struct request *clone) +static void io_err_release_clone_rq(struct request *clone, + union map_info *map_context) { } diff --git a/drivers/md/dm-thin-metadata.c b/drivers/md/dm-thin-metadata.c index ed3caceaed07..7f0840601737 100644 --- a/drivers/md/dm-thin-metadata.c +++ b/drivers/md/dm-thin-metadata.c @@ -202,6 +202,13 @@ struct dm_pool_metadata { bool fail_io:1; /* + * Set once a thin-pool has been accessed through one of the interfaces + * that imply the pool is in-service (e.g. thin devices created/deleted, + * thin-pool message, metadata snapshots, etc). + */ + bool in_service:1; + + /* * Reading the space map roots can fail, so we read it into these * buffers before the superblock is locked and updated. */ @@ -367,6 +374,32 @@ static int subtree_equal(void *context, const void *value1_le, const void *value /*----------------------------------------------------------------*/ +/* + * Variant that is used for in-core only changes or code that + * shouldn't put the pool in service on its own (e.g. commit). + */ +static inline void __pmd_write_lock(struct dm_pool_metadata *pmd) + __acquires(pmd->root_lock) +{ + down_write(&pmd->root_lock); +} +#define pmd_write_lock_in_core(pmd) __pmd_write_lock((pmd)) + +static inline void pmd_write_lock(struct dm_pool_metadata *pmd) +{ + __pmd_write_lock(pmd); + if (unlikely(!pmd->in_service)) + pmd->in_service = true; +} + +static inline void pmd_write_unlock(struct dm_pool_metadata *pmd) + __releases(pmd->root_lock) +{ + up_write(&pmd->root_lock); +} + +/*----------------------------------------------------------------*/ + static int superblock_lock_zero(struct dm_pool_metadata *pmd, struct dm_block **sblock) { @@ -790,6 +823,9 @@ static int __commit_transaction(struct dm_pool_metadata *pmd) */ BUILD_BUG_ON(sizeof(struct thin_disk_superblock) > 512); + if (unlikely(!pmd->in_service)) + return 0; + r = __write_changed_details(pmd); if (r < 0) return r; @@ -853,6 +889,7 @@ struct dm_pool_metadata *dm_pool_metadata_open(struct block_device *bdev, pmd->time = 0; INIT_LIST_HEAD(&pmd->thin_devices); pmd->fail_io = false; + pmd->in_service = false; pmd->bdev = bdev; pmd->data_block_size = data_block_size; @@ -903,7 +940,6 @@ int dm_pool_metadata_close(struct dm_pool_metadata *pmd) DMWARN("%s: __commit_transaction() failed, error = %d", __func__, r); } - if (!pmd->fail_io) __destroy_persistent_data_objects(pmd); @@ -1032,10 +1068,10 @@ int dm_pool_create_thin(struct dm_pool_metadata *pmd, dm_thin_id dev) { int r = -EINVAL; - down_write(&pmd->root_lock); + pmd_write_lock(pmd); if (!pmd->fail_io) r = __create_thin(pmd, dev); - up_write(&pmd->root_lock); + pmd_write_unlock(pmd); return r; } @@ -1123,10 +1159,10 @@ int dm_pool_create_snap(struct dm_pool_metadata *pmd, { int r = -EINVAL; - down_write(&pmd->root_lock); + pmd_write_lock(pmd); if (!pmd->fail_io) r = __create_snap(pmd, dev, origin); - up_write(&pmd->root_lock); + pmd_write_unlock(pmd); return r; } @@ -1166,10 +1202,10 @@ int dm_pool_delete_thin_device(struct dm_pool_metadata *pmd, { int r = -EINVAL; - down_write(&pmd->root_lock); + pmd_write_lock(pmd); if (!pmd->fail_io) r = __delete_device(pmd, dev); - up_write(&pmd->root_lock); + pmd_write_unlock(pmd); return r; } @@ -1180,7 +1216,7 @@ int dm_pool_set_metadata_transaction_id(struct dm_pool_metadata *pmd, { int r = -EINVAL; - down_write(&pmd->root_lock); + pmd_write_lock(pmd); if (pmd->fail_io) goto out; @@ -1194,7 +1230,7 @@ int dm_pool_set_metadata_transaction_id(struct dm_pool_metadata *pmd, r = 0; out: - up_write(&pmd->root_lock); + pmd_write_unlock(pmd); return r; } @@ -1225,7 +1261,12 @@ static int __reserve_metadata_snap(struct dm_pool_metadata *pmd) * We commit to ensure the btree roots which we increment in a * moment are up to date. */ - __commit_transaction(pmd); + r = __commit_transaction(pmd); + if (r < 0) { + DMWARN("%s: __commit_transaction() failed, error = %d", + __func__, r); + return r; + } /* * Copy the superblock. @@ -1283,10 +1324,10 @@ int dm_pool_reserve_metadata_snap(struct dm_pool_metadata *pmd) { int r = -EINVAL; - down_write(&pmd->root_lock); + pmd_write_lock(pmd); if (!pmd->fail_io) r = __reserve_metadata_snap(pmd); - up_write(&pmd->root_lock); + pmd_write_unlock(pmd); return r; } @@ -1331,10 +1372,10 @@ int dm_pool_release_metadata_snap(struct dm_pool_metadata *pmd) { int r = -EINVAL; - down_write(&pmd->root_lock); + pmd_write_lock(pmd); if (!pmd->fail_io) r = __release_metadata_snap(pmd); - up_write(&pmd->root_lock); + pmd_write_unlock(pmd); return r; } @@ -1377,19 +1418,19 @@ int dm_pool_open_thin_device(struct dm_pool_metadata *pmd, dm_thin_id dev, { int r = -EINVAL; - down_write(&pmd->root_lock); + pmd_write_lock_in_core(pmd); if (!pmd->fail_io) r = __open_device(pmd, dev, 0, td); - up_write(&pmd->root_lock); + pmd_write_unlock(pmd); return r; } int dm_pool_close_thin_device(struct dm_thin_device *td) { - down_write(&td->pmd->root_lock); + pmd_write_lock_in_core(td->pmd); __close_device(td); - up_write(&td->pmd->root_lock); + pmd_write_unlock(td->pmd); return 0; } @@ -1570,10 +1611,10 @@ int dm_thin_insert_block(struct dm_thin_device *td, dm_block_t block, { int r = -EINVAL; - down_write(&td->pmd->root_lock); + pmd_write_lock(td->pmd); if (!td->pmd->fail_io) r = __insert(td, block, data_block); - up_write(&td->pmd->root_lock); + pmd_write_unlock(td->pmd); return r; } @@ -1657,10 +1698,10 @@ int dm_thin_remove_block(struct dm_thin_device *td, dm_block_t block) { int r = -EINVAL; - down_write(&td->pmd->root_lock); + pmd_write_lock(td->pmd); if (!td->pmd->fail_io) r = __remove(td, block); - up_write(&td->pmd->root_lock); + pmd_write_unlock(td->pmd); return r; } @@ -1670,10 +1711,10 @@ int dm_thin_remove_range(struct dm_thin_device *td, { int r = -EINVAL; - down_write(&td->pmd->root_lock); + pmd_write_lock(td->pmd); if (!td->pmd->fail_io) r = __remove_range(td, begin, end); - up_write(&td->pmd->root_lock); + pmd_write_unlock(td->pmd); return r; } @@ -1696,13 +1737,13 @@ int dm_pool_inc_data_range(struct dm_pool_metadata *pmd, dm_block_t b, dm_block_ { int r = 0; - down_write(&pmd->root_lock); + pmd_write_lock(pmd); for (; b != e; b++) { r = dm_sm_inc_block(pmd->data_sm, b); if (r) break; } - up_write(&pmd->root_lock); + pmd_write_unlock(pmd); return r; } @@ -1711,13 +1752,13 @@ int dm_pool_dec_data_range(struct dm_pool_metadata *pmd, dm_block_t b, dm_block_ { int r = 0; - down_write(&pmd->root_lock); + pmd_write_lock(pmd); for (; b != e; b++) { r = dm_sm_dec_block(pmd->data_sm, b); if (r) break; } - up_write(&pmd->root_lock); + pmd_write_unlock(pmd); return r; } @@ -1765,10 +1806,10 @@ int dm_pool_alloc_data_block(struct dm_pool_metadata *pmd, dm_block_t *result) { int r = -EINVAL; - down_write(&pmd->root_lock); + pmd_write_lock(pmd); if (!pmd->fail_io) r = dm_sm_new_block(pmd->data_sm, result); - up_write(&pmd->root_lock); + pmd_write_unlock(pmd); return r; } @@ -1777,12 +1818,16 @@ int dm_pool_commit_metadata(struct dm_pool_metadata *pmd) { int r = -EINVAL; - down_write(&pmd->root_lock); + /* + * Care is taken to not have commit be what + * triggers putting the thin-pool in-service. + */ + __pmd_write_lock(pmd); if (pmd->fail_io) goto out; r = __commit_transaction(pmd); - if (r <= 0) + if (r < 0) goto out; /* @@ -1790,7 +1835,7 @@ int dm_pool_commit_metadata(struct dm_pool_metadata *pmd) */ r = __begin_transaction(pmd); out: - up_write(&pmd->root_lock); + pmd_write_unlock(pmd); return r; } @@ -1806,7 +1851,7 @@ int dm_pool_abort_metadata(struct dm_pool_metadata *pmd) { int r = -EINVAL; - down_write(&pmd->root_lock); + pmd_write_lock(pmd); if (pmd->fail_io) goto out; @@ -1817,7 +1862,7 @@ int dm_pool_abort_metadata(struct dm_pool_metadata *pmd) pmd->fail_io = true; out: - up_write(&pmd->root_lock); + pmd_write_unlock(pmd); return r; } @@ -1948,10 +1993,10 @@ int dm_pool_resize_data_dev(struct dm_pool_metadata *pmd, dm_block_t new_count) { int r = -EINVAL; - down_write(&pmd->root_lock); + pmd_write_lock(pmd); if (!pmd->fail_io) r = __resize_space_map(pmd->data_sm, new_count); - up_write(&pmd->root_lock); + pmd_write_unlock(pmd); return r; } @@ -1960,29 +2005,29 @@ int dm_pool_resize_metadata_dev(struct dm_pool_metadata *pmd, dm_block_t new_cou { int r = -EINVAL; - down_write(&pmd->root_lock); + pmd_write_lock(pmd); if (!pmd->fail_io) { r = __resize_space_map(pmd->metadata_sm, new_count); if (!r) __set_metadata_reserve(pmd); } - up_write(&pmd->root_lock); + pmd_write_unlock(pmd); return r; } void dm_pool_metadata_read_only(struct dm_pool_metadata *pmd) { - down_write(&pmd->root_lock); + pmd_write_lock_in_core(pmd); dm_bm_set_read_only(pmd->bm); - up_write(&pmd->root_lock); + pmd_write_unlock(pmd); } void dm_pool_metadata_read_write(struct dm_pool_metadata *pmd) { - down_write(&pmd->root_lock); + pmd_write_lock_in_core(pmd); dm_bm_set_read_write(pmd->bm); - up_write(&pmd->root_lock); + pmd_write_unlock(pmd); } int dm_pool_register_metadata_threshold(struct dm_pool_metadata *pmd, @@ -1992,9 +2037,9 @@ int dm_pool_register_metadata_threshold(struct dm_pool_metadata *pmd, { int r; - down_write(&pmd->root_lock); + pmd_write_lock_in_core(pmd); r = dm_sm_register_threshold_callback(pmd->metadata_sm, threshold, fn, context); - up_write(&pmd->root_lock); + pmd_write_unlock(pmd); return r; } @@ -2005,7 +2050,7 @@ int dm_pool_metadata_set_needs_check(struct dm_pool_metadata *pmd) struct dm_block *sblock; struct thin_disk_superblock *disk_super; - down_write(&pmd->root_lock); + pmd_write_lock(pmd); pmd->flags |= THIN_METADATA_NEEDS_CHECK_FLAG; r = superblock_lock(pmd, &sblock); @@ -2019,7 +2064,7 @@ int dm_pool_metadata_set_needs_check(struct dm_pool_metadata *pmd) dm_bm_unlock(sblock); out: - up_write(&pmd->root_lock); + pmd_write_unlock(pmd); return r; } diff --git a/drivers/md/dm-writecache.c b/drivers/md/dm-writecache.c index f7822875589e..1cb137f0ef9d 100644 --- a/drivers/md/dm-writecache.c +++ b/drivers/md/dm-writecache.c @@ -190,7 +190,6 @@ struct writeback_struct { struct dm_writecache *wc; struct wc_entry **wc_list; unsigned wc_list_n; - unsigned page_offset; struct page *page; struct wc_entry *wc_list_inline[WB_LIST_INLINE]; struct bio bio; @@ -546,21 +545,20 @@ static struct wc_entry *writecache_find_entry(struct dm_writecache *wc, e = container_of(node, struct wc_entry, rb_node); if (read_original_sector(wc, e) == block) break; + node = (read_original_sector(wc, e) >= block ? e->rb_node.rb_left : e->rb_node.rb_right); if (unlikely(!node)) { - if (!(flags & WFE_RETURN_FOLLOWING)) { + if (!(flags & WFE_RETURN_FOLLOWING)) return NULL; - } if (read_original_sector(wc, e) >= block) { - break; + return e; } else { node = rb_next(&e->rb_node); - if (unlikely(!node)) { + if (unlikely(!node)) return NULL; - } e = container_of(node, struct wc_entry, rb_node); - break; + return e; } } } @@ -571,7 +569,7 @@ static struct wc_entry *writecache_find_entry(struct dm_writecache *wc, node = rb_prev(&e->rb_node); else node = rb_next(&e->rb_node); - if (!node) + if (unlikely(!node)) return e; e2 = container_of(node, struct wc_entry, rb_node); if (read_original_sector(wc, e2) != block) @@ -804,7 +802,7 @@ static void writecache_discard(struct dm_writecache *wc, sector_t start, sector_ writecache_free_entry(wc, e); } - if (!node) + if (unlikely(!node)) break; e = container_of(node, struct wc_entry, rb_node); @@ -1478,10 +1476,9 @@ static void __writecache_writeback_pmem(struct dm_writecache *wc, struct writeba bio = bio_alloc_bioset(GFP_NOIO, max_pages, &wc->bio_set); wb = container_of(bio, struct writeback_struct, bio); wb->wc = wc; - wb->bio.bi_end_io = writecache_writeback_endio; - bio_set_dev(&wb->bio, wc->dev->bdev); - wb->bio.bi_iter.bi_sector = read_original_sector(wc, e); - wb->page_offset = PAGE_SIZE; + bio->bi_end_io = writecache_writeback_endio; + bio_set_dev(bio, wc->dev->bdev); + bio->bi_iter.bi_sector = read_original_sector(wc, e); if (max_pages <= WB_LIST_INLINE || unlikely(!(wb->wc_list = kmalloc_array(max_pages, sizeof(struct wc_entry *), GFP_NOIO | __GFP_NORETRY | @@ -1507,12 +1504,12 @@ static void __writecache_writeback_pmem(struct dm_writecache *wc, struct writeba wb->wc_list[wb->wc_list_n++] = f; e = f; } - bio_set_op_attrs(&wb->bio, REQ_OP_WRITE, WC_MODE_FUA(wc) * REQ_FUA); + bio_set_op_attrs(bio, REQ_OP_WRITE, WC_MODE_FUA(wc) * REQ_FUA); if (writecache_has_error(wc)) { bio->bi_status = BLK_STS_IOERR; - bio_endio(&wb->bio); + bio_endio(bio); } else { - submit_bio(&wb->bio); + submit_bio(bio); } __writeback_throttle(wc, wbl); diff --git a/drivers/md/dm-zoned-metadata.c b/drivers/md/dm-zoned-metadata.c index fa68336560c3..d8334cd45d7c 100644 --- a/drivers/md/dm-zoned-metadata.c +++ b/drivers/md/dm-zoned-metadata.c @@ -1169,6 +1169,9 @@ static int dmz_init_zones(struct dmz_metadata *zmd) goto out; } + if (!nr_blkz) + break; + /* Process report */ for (i = 0; i < nr_blkz; i++) { ret = dmz_init_zone(zmd, zone, &blkz[i]); @@ -1204,6 +1207,8 @@ static int dmz_update_zone(struct dmz_metadata *zmd, struct dm_zone *zone) /* Get zone information from disk */ ret = blkdev_report_zones(zmd->dev->bdev, dmz_start_sect(zmd, zone), &blkz, &nr_blkz, GFP_NOIO); + if (!nr_blkz) + ret = -EIO; if (ret) { dmz_dev_err(zmd->dev, "Get zone %u report failed", dmz_id(zmd, zone)); diff --git a/drivers/md/dm-zoned-target.c b/drivers/md/dm-zoned-target.c index 8865c1709e16..51d029bbb740 100644 --- a/drivers/md/dm-zoned-target.c +++ b/drivers/md/dm-zoned-target.c @@ -643,7 +643,8 @@ static int dmz_get_zoned_device(struct dm_target *ti, char *path) q = bdev_get_queue(dev->bdev); dev->capacity = i_size_read(dev->bdev->bd_inode) >> SECTOR_SHIFT; - aligned_capacity = dev->capacity & ~(blk_queue_zone_sectors(q) - 1); + aligned_capacity = dev->capacity & + ~((sector_t)blk_queue_zone_sectors(q) - 1); if (ti->begin || ((ti->len != dev->capacity) && (ti->len != aligned_capacity))) { ti->error = "Partial mapping not supported"; diff --git a/drivers/md/dm.c b/drivers/md/dm.c index 043f0761e4a0..1fb1333fefec 100644 --- a/drivers/md/dm.c +++ b/drivers/md/dm.c @@ -781,7 +781,8 @@ static void close_table_device(struct table_device *td, struct mapped_device *md } static struct table_device *find_table_device(struct list_head *l, dev_t dev, - fmode_t mode) { + fmode_t mode) +{ struct table_device *td; list_for_each_entry(td, l, list) @@ -792,7 +793,8 @@ static struct table_device *find_table_device(struct list_head *l, dev_t dev, } int dm_get_table_device(struct mapped_device *md, dev_t dev, fmode_t mode, - struct dm_dev **result) { + struct dm_dev **result) +{ int r; struct table_device *td; @@ -1906,7 +1908,6 @@ static void cleanup_mapped_device(struct mapped_device *md) static struct mapped_device *alloc_dev(int minor) { int r, numa_node_id = dm_get_numa_node(); - struct dax_device *dax_dev = NULL; struct mapped_device *md; void *old_md; @@ -1969,11 +1970,10 @@ static struct mapped_device *alloc_dev(int minor) sprintf(md->disk->disk_name, "dm-%d", minor); if (IS_ENABLED(CONFIG_DAX_DRIVER)) { - dax_dev = alloc_dax(md, md->disk->disk_name, &dm_dax_ops); - if (!dax_dev) + md->dax_dev = alloc_dax(md, md->disk->disk_name, &dm_dax_ops); + if (!md->dax_dev) goto bad; } - md->dax_dev = dax_dev; add_disk_no_queue_reg(md->disk); format_dev_t(md->name, MKDEV(_major, minor)); diff --git a/drivers/md/persistent-data/dm-space-map-common.c b/drivers/md/persistent-data/dm-space-map-common.c index 0a3b8ae4a29c..b8a62188f6be 100644 --- a/drivers/md/persistent-data/dm-space-map-common.c +++ b/drivers/md/persistent-data/dm-space-map-common.c @@ -190,6 +190,8 @@ static int sm_find_free(void *addr, unsigned begin, unsigned end, static int sm_ll_init(struct ll_disk *ll, struct dm_transaction_manager *tm) { + memset(ll, 0, sizeof(struct ll_disk)); + ll->tm = tm; ll->bitmap_info.tm = tm; diff --git a/drivers/net/ethernet/mellanox/mlxsw/Kconfig b/drivers/net/ethernet/mellanox/mlxsw/Kconfig index b6b3ff0fe17f..7ccb950aa7d4 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/Kconfig +++ b/drivers/net/ethernet/mellanox/mlxsw/Kconfig @@ -22,7 +22,6 @@ config MLXSW_CORE_HWMON config MLXSW_CORE_THERMAL bool "Thermal zone support for Mellanox Technologies Switch ASICs" depends on MLXSW_CORE && THERMAL - depends on !(MLXSW_CORE=y && THERMAL=m) default y ---help--- Say Y here if you want to automatically control fans speed according diff --git a/drivers/thermal/Kconfig b/drivers/thermal/Kconfig index 66a709d5d6b9..15bdd25780be 100644 --- a/drivers/thermal/Kconfig +++ b/drivers/thermal/Kconfig @@ -3,7 +3,7 @@ # menuconfig THERMAL - tristate "Generic Thermal sysfs driver" + bool "Generic Thermal sysfs driver" help Generic Thermal Sysfs driver offers a generic mechanism for thermal management. Usually it's made up of one or more thermal @@ -11,7 +11,7 @@ menuconfig THERMAL Each thermal zone contains its own temperature, trip points, cooling devices. All platforms with ACPI thermal support can use this driver. - If you want this support, you should say Y or M here. + If you want this support, you should say Y here. if THERMAL @@ -24,7 +24,6 @@ config THERMAL_STATISTICS config THERMAL_EMERGENCY_POWEROFF_DELAY_MS int "Emergency poweroff delay in milli-seconds" - depends on THERMAL default 0 help Thermal subsystem will issue a graceful shutdown when @@ -149,10 +148,9 @@ config THERMAL_GOV_POWER_ALLOCATOR allocating and limiting power to devices. config CPU_THERMAL - bool "generic cpu cooling support" + bool "Generic cpu cooling support" depends on CPU_FREQ depends on THERMAL_OF - depends on THERMAL=y help This implements the generic cpu cooling mechanism through frequency reduction. An ACPI version of this already exists diff --git a/drivers/thermal/intel/Kconfig b/drivers/thermal/intel/Kconfig index 2e013eeb4a1d..2c727a820759 100644 --- a/drivers/thermal/intel/Kconfig +++ b/drivers/thermal/intel/Kconfig @@ -1,6 +1,5 @@ config INTEL_POWERCLAMP tristate "Intel PowerClamp idle injection driver" - depends on THERMAL depends on X86 depends on CPU_SUP_INTEL help diff --git a/drivers/thermal/intel/int340x_thermal/int3403_thermal.c b/drivers/thermal/intel/int340x_thermal/int3403_thermal.c index 0c19fcd56a0d..79a7df2baa92 100644 --- a/drivers/thermal/intel/int340x_thermal/int3403_thermal.c +++ b/drivers/thermal/intel/int340x_thermal/int3403_thermal.c @@ -220,6 +220,7 @@ static int int3403_add(struct platform_device *pdev) { struct int3403_priv *priv; int result = 0; + unsigned long long tmp; acpi_status status; priv = devm_kzalloc(&pdev->dev, sizeof(struct int3403_priv), @@ -234,19 +235,18 @@ static int int3403_add(struct platform_device *pdev) goto err; } - status = acpi_evaluate_integer(priv->adev->handle, "PTYP", - NULL, &priv->type); - if (ACPI_FAILURE(status)) { - unsigned long long tmp; - status = acpi_evaluate_integer(priv->adev->handle, "_TMP", - NULL, &tmp); + status = acpi_evaluate_integer(priv->adev->handle, "_TMP", + NULL, &tmp); + if (ACPI_FAILURE(status)) { + status = acpi_evaluate_integer(priv->adev->handle, "PTYP", + NULL, &priv->type); if (ACPI_FAILURE(status)) { result = -EINVAL; goto err; - } else { - priv->type = INT3403_TYPE_SENSOR; } + } else { + priv->type = INT3403_TYPE_SENSOR; } platform_set_drvdata(pdev, priv); diff --git a/drivers/thermal/intel/int340x_thermal/processor_thermal_device.c b/drivers/thermal/intel/int340x_thermal/processor_thermal_device.c index 8e1cf4d789be..2e6071a82da2 100644 --- a/drivers/thermal/intel/int340x_thermal/processor_thermal_device.c +++ b/drivers/thermal/intel/int340x_thermal/processor_thermal_device.c @@ -81,22 +81,13 @@ static ssize_t power_limit_##index##_##suffix##_show(struct device *dev, \ struct device_attribute *attr, \ char *buf) \ { \ - struct pci_dev *pci_dev; \ - struct platform_device *pdev; \ - struct proc_thermal_device *proc_dev; \ + struct proc_thermal_device *proc_dev = dev_get_drvdata(dev); \ \ if (proc_thermal_emum_mode == PROC_THERMAL_NONE) { \ dev_warn(dev, "Attempted to get power limit before device was initialized!\n"); \ return 0; \ } \ \ - if (proc_thermal_emum_mode == PROC_THERMAL_PLATFORM_DEV) { \ - pdev = to_platform_device(dev); \ - proc_dev = platform_get_drvdata(pdev); \ - } else { \ - pci_dev = to_pci_dev(dev); \ - proc_dev = pci_get_drvdata(pci_dev); \ - } \ return sprintf(buf, "%lu\n",\ (unsigned long)proc_dev->power_limits[index].suffix * 1000); \ } @@ -274,7 +265,7 @@ static void proc_thermal_notify(acpi_handle handle, u32 event, void *data) THERMAL_DEVICE_POWER_CAPABILITY_CHANGED); break; default: - dev_err(proc_priv->dev, "Unsupported event [0x%x]\n", event); + dev_dbg(proc_priv->dev, "Unsupported event [0x%x]\n", event); break; } } diff --git a/drivers/thermal/qcom/Kconfig b/drivers/thermal/qcom/Kconfig index cdb455ffd575..3ce20fec86a2 100644 --- a/drivers/thermal/qcom/Kconfig +++ b/drivers/thermal/qcom/Kconfig @@ -1,6 +1,5 @@ config QCOM_TSENS tristate "Qualcomm TSENS Temperature Alarm" - depends on THERMAL depends on QCOM_QFPROM depends on ARCH_QCOM || COMPILE_TEST help diff --git a/drivers/thermal/thermal_core.c b/drivers/thermal/thermal_core.c index e0b530603db6..46cfb7de4eb2 100644 --- a/drivers/thermal/thermal_core.c +++ b/drivers/thermal/thermal_core.c @@ -266,7 +266,7 @@ static int __init thermal_register_governors(void) return thermal_gov_power_allocator_register(); } -static void thermal_unregister_governors(void) +static void __init thermal_unregister_governors(void) { thermal_gov_step_wise_unregister(); thermal_gov_fair_share_unregister(); @@ -941,7 +941,7 @@ static void bind_cdev(struct thermal_cooling_device *cdev) */ static struct thermal_cooling_device * __thermal_cooling_device_register(struct device_node *np, - char *type, void *devdata, + const char *type, void *devdata, const struct thermal_cooling_device_ops *ops) { struct thermal_cooling_device *cdev; @@ -1015,7 +1015,7 @@ __thermal_cooling_device_register(struct device_node *np, * ERR_PTR. Caller must check return value with IS_ERR*() helpers. */ struct thermal_cooling_device * -thermal_cooling_device_register(char *type, void *devdata, +thermal_cooling_device_register(const char *type, void *devdata, const struct thermal_cooling_device_ops *ops) { return __thermal_cooling_device_register(NULL, type, devdata, ops); @@ -1039,7 +1039,7 @@ EXPORT_SYMBOL_GPL(thermal_cooling_device_register); */ struct thermal_cooling_device * thermal_of_cooling_device_register(struct device_node *np, - char *type, void *devdata, + const char *type, void *devdata, const struct thermal_cooling_device_ops *ops) { return __thermal_cooling_device_register(np, type, devdata, ops); @@ -1543,6 +1543,7 @@ static int thermal_pm_notify(struct notifier_block *nb, unsigned long mode, void *_unused) { struct thermal_zone_device *tz; + enum thermal_device_mode tz_mode; switch (mode) { case PM_HIBERNATION_PREPARE: @@ -1555,6 +1556,13 @@ static int thermal_pm_notify(struct notifier_block *nb, case PM_POST_SUSPEND: atomic_set(&in_suspend, 0); list_for_each_entry(tz, &thermal_tz_list, node) { + tz_mode = THERMAL_DEVICE_ENABLED; + if (tz->ops->get_mode) + tz->ops->get_mode(tz, &tz_mode); + + if (tz_mode == THERMAL_DEVICE_DISABLED) + continue; + thermal_zone_device_init(tz); thermal_zone_device_update(tz, THERMAL_EVENT_UNSPECIFIED); @@ -1612,19 +1620,4 @@ error: mutex_destroy(&poweroff_lock); return result; } - -static void __exit thermal_exit(void) -{ - unregister_pm_notifier(&thermal_pm_nb); - of_thermal_destroy_zones(); - genetlink_exit(); - class_unregister(&thermal_class); - thermal_unregister_governors(); - ida_destroy(&thermal_tz_ida); - ida_destroy(&thermal_cdev_ida); - mutex_destroy(&thermal_list_lock); - mutex_destroy(&thermal_governor_lock); -} - fs_initcall(thermal_init); -module_exit(thermal_exit); diff --git a/fs/afs/addr_list.c b/fs/afs/addr_list.c index 967db336d11a..9eaff55df7b4 100644 --- a/fs/afs/addr_list.c +++ b/fs/afs/addr_list.c @@ -251,7 +251,7 @@ struct afs_vlserver_list *afs_dns_query(struct afs_cell *cell, time64_t *_expiry _enter("%s", cell->name); ret = dns_query("afsdb", cell->name, cell->name_len, "srv=1", - &result, _expiry); + &result, _expiry, true); if (ret < 0) { _leave(" = %d [dns]", ret); return ERR_PTR(ret); diff --git a/fs/afs/afs.h b/fs/afs/afs.h index d12ffb457e47..3f4e460c6655 100644 --- a/fs/afs/afs.h +++ b/fs/afs/afs.h @@ -23,6 +23,9 @@ #define AFSPATHMAX 1024 /* Maximum length of a pathname plus NUL */ #define AFSOPAQUEMAX 1024 /* Maximum length of an opaque field */ +#define AFS_VL_MAX_LIFESPAN (120 * HZ) +#define AFS_PROBE_MAX_LIFESPAN (30 * HZ) + typedef u64 afs_volid_t; typedef u64 afs_vnodeid_t; typedef u64 afs_dataversion_t; @@ -69,8 +72,8 @@ typedef enum { struct afs_callback { time64_t expires_at; /* Time at which expires */ - unsigned version; /* Callback version */ - afs_callback_type_t type; /* Type of callback */ + //unsigned version; /* Callback version */ + //afs_callback_type_t type; /* Type of callback */ }; struct afs_callback_break { @@ -144,6 +147,15 @@ struct afs_file_status { u32 abort_code; /* Abort if bulk-fetching this failed */ }; +struct afs_status_cb { + struct afs_file_status status; + struct afs_callback callback; + unsigned int cb_break; /* Pre-op callback break counter */ + bool have_status; /* True if status record was retrieved */ + bool have_cb; /* True if cb record was retrieved */ + bool have_error; /* True if status.abort_code indicates an error */ +}; + /* * AFS file status change request */ diff --git a/fs/afs/callback.c b/fs/afs/callback.c index 128f2dbe256a..d441bef72163 100644 --- a/fs/afs/callback.c +++ b/fs/afs/callback.c @@ -94,15 +94,15 @@ int afs_register_server_cb_interest(struct afs_vnode *vnode, struct afs_server *server = entry->server; again: - if (vnode->cb_interest && - likely(vnode->cb_interest == entry->cb_interest)) + vcbi = rcu_dereference_protected(vnode->cb_interest, + lockdep_is_held(&vnode->io_lock)); + if (vcbi && likely(vcbi == entry->cb_interest)) return 0; read_lock(&slist->lock); cbi = afs_get_cb_interest(entry->cb_interest); read_unlock(&slist->lock); - vcbi = vnode->cb_interest; if (vcbi) { if (vcbi == cbi) { afs_put_cb_interest(afs_v2net(vnode), cbi); @@ -114,8 +114,9 @@ again: */ if (cbi && vcbi->server == cbi->server) { write_seqlock(&vnode->cb_lock); - old = vnode->cb_interest; - vnode->cb_interest = cbi; + old = rcu_dereference_protected(vnode->cb_interest, + lockdep_is_held(&vnode->cb_lock.lock)); + rcu_assign_pointer(vnode->cb_interest, cbi); write_sequnlock(&vnode->cb_lock); afs_put_cb_interest(afs_v2net(vnode), old); return 0; @@ -160,8 +161,9 @@ again: */ write_seqlock(&vnode->cb_lock); - old = vnode->cb_interest; - vnode->cb_interest = cbi; + old = rcu_dereference_protected(vnode->cb_interest, + lockdep_is_held(&vnode->cb_lock.lock)); + rcu_assign_pointer(vnode->cb_interest, cbi); vnode->cb_s_break = cbi->server->cb_s_break; vnode->cb_v_break = vnode->volume->cb_v_break; clear_bit(AFS_VNODE_CB_PROMISED, &vnode->flags); @@ -191,10 +193,11 @@ void afs_put_cb_interest(struct afs_net *net, struct afs_cb_interest *cbi) vi = NULL; write_unlock(&cbi->server->cb_break_lock); - kfree(vi); + if (vi) + kfree_rcu(vi, rcu); afs_put_server(net, cbi->server); } - kfree(cbi); + kfree_rcu(cbi, rcu); } } @@ -218,14 +221,8 @@ void __afs_break_callback(struct afs_vnode *vnode) vnode->cb_break++; afs_clear_permits(vnode); - spin_lock(&vnode->lock); - - _debug("break callback"); - - if (list_empty(&vnode->granted_locks) && - !list_empty(&vnode->pending_locks)) + if (vnode->lock_state == AFS_VNODE_LOCK_WAITING_FOR_CB) afs_lock_may_be_available(vnode); - spin_unlock(&vnode->lock); } } diff --git a/fs/afs/cell.c b/fs/afs/cell.c index 9de46116c749..9c3b07ba2222 100644 --- a/fs/afs/cell.c +++ b/fs/afs/cell.c @@ -123,6 +123,7 @@ static struct afs_cell *afs_alloc_cell(struct afs_net *net, const char *name, unsigned int namelen, const char *addresses) { + struct afs_vlserver_list *vllist; struct afs_cell *cell; int i, ret; @@ -151,18 +152,14 @@ static struct afs_cell *afs_alloc_cell(struct afs_net *net, atomic_set(&cell->usage, 2); INIT_WORK(&cell->manager, afs_manage_cell); - cell->flags = ((1 << AFS_CELL_FL_NOT_READY) | - (1 << AFS_CELL_FL_NO_LOOKUP_YET)); INIT_LIST_HEAD(&cell->proc_volumes); rwlock_init(&cell->proc_lock); rwlock_init(&cell->vl_servers_lock); - /* Fill in the VL server list if we were given a list of addresses to - * use. + /* Provide a VL server list, filling it in if we were given a list of + * addresses to use. */ if (addresses) { - struct afs_vlserver_list *vllist; - vllist = afs_parse_text_addrs(net, addresses, strlen(addresses), ':', VL_SERVICE, AFS_VL_PORT); @@ -171,19 +168,32 @@ static struct afs_cell *afs_alloc_cell(struct afs_net *net, goto parse_failed; } - rcu_assign_pointer(cell->vl_servers, vllist); + vllist->source = DNS_RECORD_FROM_CONFIG; + vllist->status = DNS_LOOKUP_NOT_DONE; cell->dns_expiry = TIME64_MAX; - __clear_bit(AFS_CELL_FL_NO_LOOKUP_YET, &cell->flags); } else { + ret = -ENOMEM; + vllist = afs_alloc_vlserver_list(0); + if (!vllist) + goto error; + vllist->source = DNS_RECORD_UNAVAILABLE; + vllist->status = DNS_LOOKUP_NOT_DONE; cell->dns_expiry = ktime_get_real_seconds(); } + rcu_assign_pointer(cell->vl_servers, vllist); + + cell->dns_source = vllist->source; + cell->dns_status = vllist->status; + smp_store_release(&cell->dns_lookup_count, 1); /* vs source/status */ + _leave(" = %p", cell); return cell; parse_failed: if (ret == -EINVAL) printk(KERN_ERR "kAFS: bad VL server IP address\n"); +error: kfree(cell); _leave(" = %d", ret); return ERR_PTR(ret); @@ -208,6 +218,7 @@ struct afs_cell *afs_lookup_cell(struct afs_net *net, { struct afs_cell *cell, *candidate, *cursor; struct rb_node *parent, **pp; + enum afs_cell_state state; int ret, n; _enter("%s,%s", name, vllist); @@ -267,18 +278,16 @@ struct afs_cell *afs_lookup_cell(struct afs_net *net, wait_for_cell: _debug("wait_for_cell"); - ret = wait_on_bit(&cell->flags, AFS_CELL_FL_NOT_READY, TASK_INTERRUPTIBLE); - smp_rmb(); - - switch (READ_ONCE(cell->state)) { - case AFS_CELL_FAILED: + wait_var_event(&cell->state, + ({ + state = smp_load_acquire(&cell->state); /* vs error */ + state == AFS_CELL_ACTIVE || state == AFS_CELL_FAILED; + })); + + /* Check the state obtained from the wait check. */ + if (state == AFS_CELL_FAILED) { ret = cell->error; goto error; - default: - _debug("weird %u %d", cell->state, cell->error); - goto error; - case AFS_CELL_ACTIVE: - break; } _leave(" = %p [cell]", cell); @@ -360,16 +369,46 @@ int afs_cell_init(struct afs_net *net, const char *rootcell) /* * Update a cell's VL server address list from the DNS. */ -static void afs_update_cell(struct afs_cell *cell) +static int afs_update_cell(struct afs_cell *cell) { - struct afs_vlserver_list *vllist, *old; + struct afs_vlserver_list *vllist, *old = NULL, *p; unsigned int min_ttl = READ_ONCE(afs_cell_min_ttl); unsigned int max_ttl = READ_ONCE(afs_cell_max_ttl); time64_t now, expiry = 0; + int ret = 0; _enter("%s", cell->name); vllist = afs_dns_query(cell, &expiry); + if (IS_ERR(vllist)) { + ret = PTR_ERR(vllist); + + _debug("%s: fail %d", cell->name, ret); + if (ret == -ENOMEM) + goto out_wake; + + ret = -ENOMEM; + vllist = afs_alloc_vlserver_list(0); + if (!vllist) + goto out_wake; + + switch (ret) { + case -ENODATA: + case -EDESTADDRREQ: + vllist->status = DNS_LOOKUP_GOT_NOT_FOUND; + break; + case -EAGAIN: + case -ECONNREFUSED: + vllist->status = DNS_LOOKUP_GOT_TEMP_FAILURE; + break; + default: + vllist->status = DNS_LOOKUP_GOT_LOCAL_FAILURE; + break; + } + } + + _debug("%s: got list %d %d", cell->name, vllist->source, vllist->status); + cell->dns_status = vllist->status; now = ktime_get_real_seconds(); if (min_ttl > max_ttl) @@ -379,48 +418,47 @@ static void afs_update_cell(struct afs_cell *cell) else if (expiry > now + max_ttl) expiry = now + max_ttl; - if (IS_ERR(vllist)) { - switch (PTR_ERR(vllist)) { - case -ENODATA: - case -EDESTADDRREQ: + _debug("%s: status %d", cell->name, vllist->status); + if (vllist->source == DNS_RECORD_UNAVAILABLE) { + switch (vllist->status) { + case DNS_LOOKUP_GOT_NOT_FOUND: /* The DNS said that the cell does not exist or there * weren't any addresses to be had. */ - set_bit(AFS_CELL_FL_NOT_FOUND, &cell->flags); - clear_bit(AFS_CELL_FL_DNS_FAIL, &cell->flags); cell->dns_expiry = expiry; break; - case -EAGAIN: - case -ECONNREFUSED: + case DNS_LOOKUP_BAD: + case DNS_LOOKUP_GOT_LOCAL_FAILURE: + case DNS_LOOKUP_GOT_TEMP_FAILURE: + case DNS_LOOKUP_GOT_NS_FAILURE: default: - set_bit(AFS_CELL_FL_DNS_FAIL, &cell->flags); cell->dns_expiry = now + 10; break; } - - cell->error = -EDESTADDRREQ; } else { - clear_bit(AFS_CELL_FL_DNS_FAIL, &cell->flags); - clear_bit(AFS_CELL_FL_NOT_FOUND, &cell->flags); - - /* Exclusion on changing vl_addrs is achieved by a - * non-reentrant work item. - */ - old = rcu_dereference_protected(cell->vl_servers, true); - rcu_assign_pointer(cell->vl_servers, vllist); cell->dns_expiry = expiry; - - if (old) - afs_put_vlserverlist(cell->net, old); } - if (test_and_clear_bit(AFS_CELL_FL_NO_LOOKUP_YET, &cell->flags)) - wake_up_bit(&cell->flags, AFS_CELL_FL_NO_LOOKUP_YET); + /* Replace the VL server list if the new record has servers or the old + * record doesn't. + */ + write_lock(&cell->vl_servers_lock); + p = rcu_dereference_protected(cell->vl_servers, true); + if (vllist->nr_servers > 0 || p->nr_servers == 0) { + rcu_assign_pointer(cell->vl_servers, vllist); + cell->dns_source = vllist->source; + old = p; + } + write_unlock(&cell->vl_servers_lock); + afs_put_vlserverlist(cell->net, old); - now = ktime_get_real_seconds(); - afs_set_cell_timer(cell->net, cell->dns_expiry - now); - _leave(""); +out_wake: + smp_store_release(&cell->dns_lookup_count, + cell->dns_lookup_count + 1); /* vs source/status */ + wake_up_var(&cell->dns_lookup_count); + _leave(" = %d", ret); + return ret; } /* @@ -491,8 +529,7 @@ void afs_put_cell(struct afs_net *net, struct afs_cell *cell) now = ktime_get_real_seconds(); cell->last_inactive = now; expire_delay = 0; - if (!test_bit(AFS_CELL_FL_DNS_FAIL, &cell->flags) && - !test_bit(AFS_CELL_FL_NOT_FOUND, &cell->flags)) + if (cell->vl_servers->nr_servers) expire_delay = afs_cell_gc_delay; if (atomic_dec_return(&cell->usage) > 1) @@ -623,11 +660,13 @@ again: goto final_destruction; if (cell->state == AFS_CELL_FAILED) goto done; - cell->state = AFS_CELL_UNSET; + smp_store_release(&cell->state, AFS_CELL_UNSET); + wake_up_var(&cell->state); goto again; case AFS_CELL_UNSET: - cell->state = AFS_CELL_ACTIVATING; + smp_store_release(&cell->state, AFS_CELL_ACTIVATING); + wake_up_var(&cell->state); goto again; case AFS_CELL_ACTIVATING: @@ -635,28 +674,29 @@ again: if (ret < 0) goto activation_failed; - cell->state = AFS_CELL_ACTIVE; - smp_wmb(); - clear_bit(AFS_CELL_FL_NOT_READY, &cell->flags); - wake_up_bit(&cell->flags, AFS_CELL_FL_NOT_READY); + smp_store_release(&cell->state, AFS_CELL_ACTIVE); + wake_up_var(&cell->state); goto again; case AFS_CELL_ACTIVE: if (atomic_read(&cell->usage) > 1) { - time64_t now = ktime_get_real_seconds(); - if (cell->dns_expiry <= now && net->live) - afs_update_cell(cell); + if (test_and_clear_bit(AFS_CELL_FL_DO_LOOKUP, &cell->flags)) { + ret = afs_update_cell(cell); + if (ret < 0) + cell->error = ret; + } goto done; } - cell->state = AFS_CELL_DEACTIVATING; + smp_store_release(&cell->state, AFS_CELL_DEACTIVATING); + wake_up_var(&cell->state); goto again; case AFS_CELL_DEACTIVATING: - set_bit(AFS_CELL_FL_NOT_READY, &cell->flags); if (atomic_read(&cell->usage) > 1) goto reverse_deactivation; afs_deactivate_cell(net, cell); - cell->state = AFS_CELL_INACTIVE; + smp_store_release(&cell->state, AFS_CELL_INACTIVE); + wake_up_var(&cell->state); goto again; default: @@ -669,17 +709,13 @@ activation_failed: cell->error = ret; afs_deactivate_cell(net, cell); - cell->state = AFS_CELL_FAILED; - smp_wmb(); - if (test_and_clear_bit(AFS_CELL_FL_NOT_READY, &cell->flags)) - wake_up_bit(&cell->flags, AFS_CELL_FL_NOT_READY); + smp_store_release(&cell->state, AFS_CELL_FAILED); /* vs error */ + wake_up_var(&cell->state); goto again; reverse_deactivation: - cell->state = AFS_CELL_ACTIVE; - smp_wmb(); - clear_bit(AFS_CELL_FL_NOT_READY, &cell->flags); - wake_up_bit(&cell->flags, AFS_CELL_FL_NOT_READY); + smp_store_release(&cell->state, AFS_CELL_ACTIVE); + wake_up_var(&cell->state); _leave(" [deact->act]"); return; @@ -739,11 +775,16 @@ void afs_manage_cells(struct work_struct *work) } if (usage == 1) { + struct afs_vlserver_list *vllist; time64_t expire_at = cell->last_inactive; - if (!test_bit(AFS_CELL_FL_DNS_FAIL, &cell->flags) && - !test_bit(AFS_CELL_FL_NOT_FOUND, &cell->flags)) + read_lock(&cell->vl_servers_lock); + vllist = rcu_dereference_protected( + cell->vl_servers, + lockdep_is_held(&cell->vl_servers_lock)); + if (vllist->nr_servers > 0) expire_at += afs_cell_gc_delay; + read_unlock(&cell->vl_servers_lock); if (purging || expire_at <= now) sched_cell = true; else if (expire_at < next_manage) @@ -751,10 +792,8 @@ void afs_manage_cells(struct work_struct *work) } if (!purging) { - if (cell->dns_expiry <= now) + if (test_bit(AFS_CELL_FL_DO_LOOKUP, &cell->flags)) sched_cell = true; - else if (cell->dns_expiry <= next_manage) - next_manage = cell->dns_expiry; } if (sched_cell) diff --git a/fs/afs/cmservice.c b/fs/afs/cmservice.c index 748090014519..01437cfe5432 100644 --- a/fs/afs/cmservice.c +++ b/fs/afs/cmservice.c @@ -213,7 +213,7 @@ static int afs_find_cm_server_by_peer(struct afs_call *call) return 0; } - call->cm_server = server; + call->server = server; return afs_record_cm_probe(call, server); } @@ -234,7 +234,7 @@ static int afs_find_cm_server_by_uuid(struct afs_call *call, return 0; } - call->cm_server = server; + call->server = server; return afs_record_cm_probe(call, server); } @@ -260,8 +260,8 @@ static void SRXAFSCB_CallBack(struct work_struct *work) * server holds up change visibility till it receives our reply so as * to maintain cache coherency. */ - if (call->cm_server) - afs_break_callbacks(call->cm_server, call->count, call->request); + if (call->server) + afs_break_callbacks(call->server, call->count, call->request); afs_send_empty_reply(call); afs_put_call(call); @@ -376,10 +376,10 @@ static void SRXAFSCB_InitCallBackState(struct work_struct *work) { struct afs_call *call = container_of(work, struct afs_call, work); - _enter("{%p}", call->cm_server); + _enter("{%p}", call->server); - if (call->cm_server) - afs_init_callback_state(call->cm_server); + if (call->server) + afs_init_callback_state(call->server); afs_send_empty_reply(call); afs_put_call(call); _leave(""); diff --git a/fs/afs/dir.c b/fs/afs/dir.c index 9a466be583d2..79d93a26759a 100644 --- a/fs/afs/dir.c +++ b/fs/afs/dir.c @@ -18,6 +18,7 @@ #include <linux/sched.h> #include <linux/task_io_accounting_ops.h> #include "internal.h" +#include "afs_fs.h" #include "xdr_fs.h" static struct dentry *afs_lookup(struct inode *dir, struct dentry *dentry, @@ -102,8 +103,8 @@ struct afs_lookup_cookie { bool found; bool one_only; unsigned short nr_fids; - struct afs_file_status *statuses; - struct afs_callback *callbacks; + struct inode **inodes; + struct afs_status_cb *statuses; struct afs_fid fids[50]; }; @@ -638,12 +639,14 @@ static struct inode *afs_do_lookup(struct inode *dir, struct dentry *dentry, struct key *key) { struct afs_lookup_cookie *cookie; - struct afs_cb_interest *cbi = NULL; + struct afs_cb_interest *dcbi, *cbi = NULL; struct afs_super_info *as = dir->i_sb->s_fs_info; - struct afs_iget_data data; + struct afs_status_cb *scb; + struct afs_iget_data iget_data; struct afs_fs_cursor fc; - struct afs_vnode *dvnode = AFS_FS_I(dir); - struct inode *inode = NULL; + struct afs_server *server; + struct afs_vnode *dvnode = AFS_FS_I(dir), *vnode; + struct inode *inode = NULL, *ti; int ret, i; _enter("{%lu},%p{%pd},", dir->i_ino, dentry, dentry); @@ -657,10 +660,14 @@ static struct inode *afs_do_lookup(struct inode *dir, struct dentry *dentry, cookie->nr_fids = 1; /* slot 0 is saved for the fid we actually want */ read_seqlock_excl(&dvnode->cb_lock); - if (dvnode->cb_interest && - dvnode->cb_interest->server && - test_bit(AFS_SERVER_FL_NO_IBULK, &dvnode->cb_interest->server->flags)) - cookie->one_only = true; + dcbi = rcu_dereference_protected(dvnode->cb_interest, + lockdep_is_held(&dvnode->cb_lock.lock)); + if (dcbi) { + server = dcbi->server; + if (server && + test_bit(AFS_SERVER_FL_NO_IBULK, &server->flags)) + cookie->one_only = true; + } read_sequnlock_excl(&dvnode->cb_lock); for (i = 0; i < 50; i++) @@ -678,24 +685,43 @@ static struct inode *afs_do_lookup(struct inode *dir, struct dentry *dentry, goto out; /* Check to see if we already have an inode for the primary fid. */ - data.volume = dvnode->volume; - data.fid = cookie->fids[0]; - inode = ilookup5(dir->i_sb, cookie->fids[0].vnode, afs_iget5_test, &data); + iget_data.fid = cookie->fids[0]; + iget_data.volume = dvnode->volume; + iget_data.cb_v_break = dvnode->volume->cb_v_break; + iget_data.cb_s_break = 0; + inode = ilookup5(dir->i_sb, cookie->fids[0].vnode, + afs_iget5_test, &iget_data); if (inode) goto out; /* Need space for examining all the selected files */ inode = ERR_PTR(-ENOMEM); - cookie->statuses = kcalloc(cookie->nr_fids, sizeof(struct afs_file_status), - GFP_KERNEL); + cookie->statuses = kvcalloc(cookie->nr_fids, sizeof(struct afs_status_cb), + GFP_KERNEL); if (!cookie->statuses) goto out; - cookie->callbacks = kcalloc(cookie->nr_fids, sizeof(struct afs_callback), - GFP_KERNEL); - if (!cookie->callbacks) + cookie->inodes = kcalloc(cookie->nr_fids, sizeof(struct inode *), + GFP_KERNEL); + if (!cookie->inodes) goto out_s; + for (i = 1; i < cookie->nr_fids; i++) { + scb = &cookie->statuses[i]; + + /* Find any inodes that already exist and get their + * callback counters. + */ + iget_data.fid = cookie->fids[i]; + ti = ilookup5_nowait(dir->i_sb, iget_data.fid.vnode, + afs_iget5_test, &iget_data); + if (!IS_ERR_OR_NULL(ti)) { + vnode = AFS_FS_I(ti); + scb->cb_break = afs_calc_vnode_cb_break(vnode); + cookie->inodes[i] = ti; + } + } + /* Try FS.InlineBulkStatus first. Abort codes for the individual * lookups contained therein are stored in the reply without aborting * the whole operation. @@ -704,7 +730,7 @@ static struct inode *afs_do_lookup(struct inode *dir, struct dentry *dentry, goto no_inline_bulk_status; inode = ERR_PTR(-ERESTARTSYS); - if (afs_begin_vnode_operation(&fc, dvnode, key)) { + if (afs_begin_vnode_operation(&fc, dvnode, key, true)) { while (afs_select_fileserver(&fc)) { if (test_bit(AFS_SERVER_FL_NO_IBULK, &fc.cbi->server->flags)) { @@ -712,11 +738,12 @@ static struct inode *afs_do_lookup(struct inode *dir, struct dentry *dentry, fc.ac.error = -ECONNABORTED; break; } + iget_data.cb_v_break = dvnode->volume->cb_v_break; + iget_data.cb_s_break = fc.cbi->server->cb_s_break; afs_fs_inline_bulk_status(&fc, afs_v2net(dvnode), cookie->fids, cookie->statuses, - cookie->callbacks, cookie->nr_fids, NULL); } @@ -737,15 +764,16 @@ no_inline_bulk_status: * any of the lookups fails - so, for the moment, revert to * FS.FetchStatus for just the primary fid. */ - cookie->nr_fids = 1; inode = ERR_PTR(-ERESTARTSYS); - if (afs_begin_vnode_operation(&fc, dvnode, key)) { + if (afs_begin_vnode_operation(&fc, dvnode, key, true)) { while (afs_select_fileserver(&fc)) { + iget_data.cb_v_break = dvnode->volume->cb_v_break; + iget_data.cb_s_break = fc.cbi->server->cb_s_break; + scb = &cookie->statuses[0]; afs_fs_fetch_status(&fc, afs_v2net(dvnode), cookie->fids, - cookie->statuses, - cookie->callbacks, + scb, NULL); } @@ -757,26 +785,36 @@ no_inline_bulk_status: if (IS_ERR(inode)) goto out_c; - for (i = 0; i < cookie->nr_fids; i++) - cookie->statuses[i].abort_code = 0; - success: /* Turn all the files into inodes and save the first one - which is the * one we actually want. */ - if (cookie->statuses[0].abort_code != 0) - inode = ERR_PTR(afs_abort_to_error(cookie->statuses[0].abort_code)); + scb = &cookie->statuses[0]; + if (scb->status.abort_code != 0) + inode = ERR_PTR(afs_abort_to_error(scb->status.abort_code)); for (i = 0; i < cookie->nr_fids; i++) { - struct inode *ti; + struct afs_status_cb *scb = &cookie->statuses[i]; + + if (!scb->have_status && !scb->have_error) + continue; + + if (cookie->inodes[i]) { + afs_vnode_commit_status(&fc, AFS_FS_I(cookie->inodes[i]), + scb->cb_break, NULL, scb); + continue; + } - if (cookie->statuses[i].abort_code != 0) + if (scb->status.abort_code != 0) continue; - ti = afs_iget(dir->i_sb, key, &cookie->fids[i], - &cookie->statuses[i], - &cookie->callbacks[i], - cbi, dvnode); + iget_data.fid = cookie->fids[i]; + ti = afs_iget(dir->i_sb, key, &iget_data, scb, cbi, dvnode); + if (!IS_ERR(ti)) + afs_cache_permit(AFS_FS_I(ti), key, + 0 /* Assume vnode->cb_break is 0 */ + + iget_data.cb_v_break, + scb); if (i == 0) { inode = ti; } else { @@ -787,9 +825,13 @@ success: out_c: afs_put_cb_interest(afs_v2net(dvnode), cbi); - kfree(cookie->callbacks); + if (cookie->inodes) { + for (i = 0; i < cookie->nr_fids; i++) + iput(cookie->inodes[i]); + kfree(cookie->inodes); + } out_s: - kfree(cookie->statuses); + kvfree(cookie->statuses); out: kfree(cookie); return inode; @@ -1114,9 +1156,8 @@ void afs_d_release(struct dentry *dentry) */ static void afs_vnode_new_inode(struct afs_fs_cursor *fc, struct dentry *new_dentry, - struct afs_fid *newfid, - struct afs_file_status *newstatus, - struct afs_callback *newcb) + struct afs_iget_data *new_data, + struct afs_status_cb *new_scb) { struct afs_vnode *vnode; struct inode *inode; @@ -1125,7 +1166,7 @@ static void afs_vnode_new_inode(struct afs_fs_cursor *fc, return; inode = afs_iget(fc->vnode->vfs_inode.i_sb, fc->key, - newfid, newstatus, newcb, fc->cbi, fc->vnode); + new_data, new_scb, fc->cbi, fc->vnode); if (IS_ERR(inode)) { /* ENOMEM or EINTR at a really inconvenient time - just abandon * the new directory on the server. @@ -1136,22 +1177,29 @@ static void afs_vnode_new_inode(struct afs_fs_cursor *fc, vnode = AFS_FS_I(inode); set_bit(AFS_VNODE_NEW_CONTENT, &vnode->flags); - afs_vnode_commit_status(fc, vnode, 0); + if (fc->ac.error == 0) + afs_cache_permit(vnode, fc->key, vnode->cb_break, new_scb); d_instantiate(new_dentry, inode); } +static void afs_prep_for_new_inode(struct afs_fs_cursor *fc, + struct afs_iget_data *iget_data) +{ + iget_data->volume = fc->vnode->volume; + iget_data->cb_v_break = fc->vnode->volume->cb_v_break; + iget_data->cb_s_break = fc->cbi->server->cb_s_break; +} + /* * create a directory on an AFS filesystem */ static int afs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode) { - struct afs_file_status newstatus; + struct afs_iget_data iget_data; + struct afs_status_cb *scb; struct afs_fs_cursor fc; - struct afs_callback newcb; struct afs_vnode *dvnode = AFS_FS_I(dir); - struct afs_fid newfid; struct key *key; - u64 data_version = dvnode->status.data_version; int ret; mode |= S_IFDIR; @@ -1159,23 +1207,32 @@ static int afs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode) _enter("{%llx:%llu},{%pd},%ho", dvnode->fid.vid, dvnode->fid.vnode, dentry, mode); + ret = -ENOMEM; + scb = kcalloc(2, sizeof(struct afs_status_cb), GFP_KERNEL); + if (!scb) + goto error; + key = afs_request_key(dvnode->volume->cell); if (IS_ERR(key)) { ret = PTR_ERR(key); - goto error; + goto error_scb; } ret = -ERESTARTSYS; - if (afs_begin_vnode_operation(&fc, dvnode, key)) { + if (afs_begin_vnode_operation(&fc, dvnode, key, true)) { + afs_dataversion_t data_version = dvnode->status.data_version + 1; + while (afs_select_fileserver(&fc)) { fc.cb_break = afs_calc_vnode_cb_break(dvnode); - afs_fs_create(&fc, dentry->d_name.name, mode, data_version, - &newfid, &newstatus, &newcb); + afs_prep_for_new_inode(&fc, &iget_data); + afs_fs_create(&fc, dentry->d_name.name, mode, + &scb[0], &iget_data.fid, &scb[1]); } - afs_check_for_remote_deletion(&fc, fc.vnode); - afs_vnode_commit_status(&fc, dvnode, fc.cb_break); - afs_vnode_new_inode(&fc, dentry, &newfid, &newstatus, &newcb); + afs_check_for_remote_deletion(&fc, dvnode); + afs_vnode_commit_status(&fc, dvnode, fc.cb_break, + &data_version, &scb[0]); + afs_vnode_new_inode(&fc, dentry, &iget_data, &scb[1]); ret = afs_end_vnode_operation(&fc); if (ret < 0) goto error_key; @@ -1185,15 +1242,18 @@ static int afs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode) if (ret == 0 && test_bit(AFS_VNODE_DIR_VALID, &dvnode->flags)) - afs_edit_dir_add(dvnode, &dentry->d_name, &newfid, + afs_edit_dir_add(dvnode, &dentry->d_name, &iget_data.fid, afs_edit_dir_for_create); key_put(key); + kfree(scb); _leave(" = 0"); return 0; error_key: key_put(key); +error_scb: + kfree(scb); error: d_drop(dentry); _leave(" = %d", ret); @@ -1220,15 +1280,19 @@ static void afs_dir_remove_subdir(struct dentry *dentry) */ static int afs_rmdir(struct inode *dir, struct dentry *dentry) { + struct afs_status_cb *scb; struct afs_fs_cursor fc; struct afs_vnode *dvnode = AFS_FS_I(dir), *vnode = NULL; struct key *key; - u64 data_version = dvnode->status.data_version; int ret; _enter("{%llx:%llu},{%pd}", dvnode->fid.vid, dvnode->fid.vnode, dentry); + scb = kzalloc(sizeof(struct afs_status_cb), GFP_KERNEL); + if (!scb) + return -ENOMEM; + key = afs_request_key(dvnode->volume->cell); if (IS_ERR(key)) { ret = PTR_ERR(key); @@ -1250,14 +1314,16 @@ static int afs_rmdir(struct inode *dir, struct dentry *dentry) } ret = -ERESTARTSYS; - if (afs_begin_vnode_operation(&fc, dvnode, key)) { + if (afs_begin_vnode_operation(&fc, dvnode, key, true)) { + afs_dataversion_t data_version = dvnode->status.data_version + 1; + while (afs_select_fileserver(&fc)) { fc.cb_break = afs_calc_vnode_cb_break(dvnode); - afs_fs_remove(&fc, vnode, dentry->d_name.name, true, - data_version); + afs_fs_remove(&fc, vnode, dentry->d_name.name, true, scb); } - afs_vnode_commit_status(&fc, dvnode, fc.cb_break); + afs_vnode_commit_status(&fc, dvnode, fc.cb_break, + &data_version, scb); ret = afs_end_vnode_operation(&fc); if (ret == 0) { afs_dir_remove_subdir(dentry); @@ -1272,6 +1338,7 @@ static int afs_rmdir(struct inode *dir, struct dentry *dentry) error_key: key_put(key); error: + kfree(scb); return ret; } @@ -1285,32 +1352,27 @@ error: * However, if we didn't have a callback promise outstanding, or it was * outstanding on a different server, then it won't break it either... */ -int afs_dir_remove_link(struct dentry *dentry, struct key *key, - unsigned long d_version_before, - unsigned long d_version_after) +static int afs_dir_remove_link(struct afs_vnode *dvnode, struct dentry *dentry, + struct key *key) { - bool dir_valid; int ret = 0; - /* There were no intervening changes on the server if the version - * number we got back was incremented by exactly 1. - */ - dir_valid = (d_version_after == d_version_before + 1); - if (d_really_is_positive(dentry)) { struct afs_vnode *vnode = AFS_FS_I(d_inode(dentry)); if (test_bit(AFS_VNODE_DELETED, &vnode->flags)) { /* Already done */ - } else if (dir_valid) { + } else if (test_bit(AFS_VNODE_DIR_VALID, &dvnode->flags)) { + write_seqlock(&vnode->cb_lock); drop_nlink(&vnode->vfs_inode); if (vnode->vfs_inode.i_nlink == 0) { set_bit(AFS_VNODE_DELETED, &vnode->flags); - clear_bit(AFS_VNODE_CB_PROMISED, &vnode->flags); + __afs_break_callback(vnode); } + write_sequnlock(&vnode->cb_lock); ret = 0; } else { - clear_bit(AFS_VNODE_CB_PROMISED, &vnode->flags); + afs_break_callback(vnode); if (test_bit(AFS_VNODE_DELETED, &vnode->flags)) kdebug("AFS_VNODE_DELETED"); @@ -1331,11 +1393,10 @@ int afs_dir_remove_link(struct dentry *dentry, struct key *key, static int afs_unlink(struct inode *dir, struct dentry *dentry) { struct afs_fs_cursor fc; + struct afs_status_cb *scb; struct afs_vnode *dvnode = AFS_FS_I(dir), *vnode = NULL; struct key *key; - unsigned long d_version = (unsigned long)dentry->d_fsdata; bool need_rehash = false; - u64 data_version = dvnode->status.data_version; int ret; _enter("{%llx:%llu},{%pd}", @@ -1344,10 +1405,15 @@ static int afs_unlink(struct inode *dir, struct dentry *dentry) if (dentry->d_name.len >= AFSNAMEMAX) return -ENAMETOOLONG; + ret = -ENOMEM; + scb = kcalloc(2, sizeof(struct afs_status_cb), GFP_KERNEL); + if (!scb) + goto error; + key = afs_request_key(dvnode->volume->cell); if (IS_ERR(key)) { ret = PTR_ERR(key); - goto error; + goto error_scb; } /* Try to make sure we have a callback promise on the victim. */ @@ -1374,30 +1440,34 @@ static int afs_unlink(struct inode *dir, struct dentry *dentry) spin_unlock(&dentry->d_lock); ret = -ERESTARTSYS; - if (afs_begin_vnode_operation(&fc, dvnode, key)) { + if (afs_begin_vnode_operation(&fc, dvnode, key, true)) { + afs_dataversion_t data_version = dvnode->status.data_version + 1; + afs_dataversion_t data_version_2 = vnode->status.data_version; + while (afs_select_fileserver(&fc)) { fc.cb_break = afs_calc_vnode_cb_break(dvnode); + fc.cb_break_2 = afs_calc_vnode_cb_break(vnode); if (test_bit(AFS_SERVER_FL_IS_YFS, &fc.cbi->server->flags) && !test_bit(AFS_SERVER_FL_NO_RM2, &fc.cbi->server->flags)) { yfs_fs_remove_file2(&fc, vnode, dentry->d_name.name, - data_version); + &scb[0], &scb[1]); if (fc.ac.error != -ECONNABORTED || fc.ac.abort_code != RXGEN_OPCODE) continue; set_bit(AFS_SERVER_FL_NO_RM2, &fc.cbi->server->flags); } - afs_fs_remove(&fc, vnode, dentry->d_name.name, false, - data_version); + afs_fs_remove(&fc, vnode, dentry->d_name.name, false, &scb[0]); } - afs_vnode_commit_status(&fc, dvnode, fc.cb_break); + afs_vnode_commit_status(&fc, dvnode, fc.cb_break, + &data_version, &scb[0]); + afs_vnode_commit_status(&fc, vnode, fc.cb_break_2, + &data_version_2, &scb[1]); ret = afs_end_vnode_operation(&fc); - if (ret == 0) - ret = afs_dir_remove_link( - dentry, key, d_version, - (unsigned long)dvnode->status.data_version); + if (ret == 0 && !(scb[1].have_status || scb[1].have_error)) + ret = afs_dir_remove_link(dvnode, dentry, key); if (ret == 0 && test_bit(AFS_VNODE_DIR_VALID, &dvnode->flags)) afs_edit_dir_remove(dvnode, &dentry->d_name, @@ -1409,6 +1479,8 @@ static int afs_unlink(struct inode *dir, struct dentry *dentry) error_key: key_put(key); +error_scb: + kfree(scb); error: _leave(" = %d", ret); return ret; @@ -1420,13 +1492,11 @@ error: static int afs_create(struct inode *dir, struct dentry *dentry, umode_t mode, bool excl) { + struct afs_iget_data iget_data; struct afs_fs_cursor fc; - struct afs_file_status newstatus; - struct afs_callback newcb; + struct afs_status_cb *scb; struct afs_vnode *dvnode = AFS_FS_I(dir); - struct afs_fid newfid; struct key *key; - u64 data_version = dvnode->status.data_version; int ret; mode |= S_IFREG; @@ -1444,17 +1514,26 @@ static int afs_create(struct inode *dir, struct dentry *dentry, umode_t mode, goto error; } + ret = -ENOMEM; + scb = kcalloc(2, sizeof(struct afs_status_cb), GFP_KERNEL); + if (!scb) + goto error_scb; + ret = -ERESTARTSYS; - if (afs_begin_vnode_operation(&fc, dvnode, key)) { + if (afs_begin_vnode_operation(&fc, dvnode, key, true)) { + afs_dataversion_t data_version = dvnode->status.data_version + 1; + while (afs_select_fileserver(&fc)) { fc.cb_break = afs_calc_vnode_cb_break(dvnode); - afs_fs_create(&fc, dentry->d_name.name, mode, data_version, - &newfid, &newstatus, &newcb); + afs_prep_for_new_inode(&fc, &iget_data); + afs_fs_create(&fc, dentry->d_name.name, mode, + &scb[0], &iget_data.fid, &scb[1]); } - afs_check_for_remote_deletion(&fc, fc.vnode); - afs_vnode_commit_status(&fc, dvnode, fc.cb_break); - afs_vnode_new_inode(&fc, dentry, &newfid, &newstatus, &newcb); + afs_check_for_remote_deletion(&fc, dvnode); + afs_vnode_commit_status(&fc, dvnode, fc.cb_break, + &data_version, &scb[0]); + afs_vnode_new_inode(&fc, dentry, &iget_data, &scb[1]); ret = afs_end_vnode_operation(&fc); if (ret < 0) goto error_key; @@ -1463,13 +1542,16 @@ static int afs_create(struct inode *dir, struct dentry *dentry, umode_t mode, } if (test_bit(AFS_VNODE_DIR_VALID, &dvnode->flags)) - afs_edit_dir_add(dvnode, &dentry->d_name, &newfid, + afs_edit_dir_add(dvnode, &dentry->d_name, &iget_data.fid, afs_edit_dir_for_create); + kfree(scb); key_put(key); _leave(" = 0"); return 0; +error_scb: + kfree(scb); error_key: key_put(key); error: @@ -1485,15 +1567,12 @@ static int afs_link(struct dentry *from, struct inode *dir, struct dentry *dentry) { struct afs_fs_cursor fc; - struct afs_vnode *dvnode, *vnode; + struct afs_status_cb *scb; + struct afs_vnode *dvnode = AFS_FS_I(dir); + struct afs_vnode *vnode = AFS_FS_I(d_inode(from)); struct key *key; - u64 data_version; int ret; - vnode = AFS_FS_I(d_inode(from)); - dvnode = AFS_FS_I(dir); - data_version = dvnode->status.data_version; - _enter("{%llx:%llu},{%llx:%llu},{%pd}", vnode->fid.vid, vnode->fid.vnode, dvnode->fid.vid, dvnode->fid.vnode, @@ -1503,14 +1582,21 @@ static int afs_link(struct dentry *from, struct inode *dir, if (dentry->d_name.len >= AFSNAMEMAX) goto error; + ret = -ENOMEM; + scb = kcalloc(2, sizeof(struct afs_status_cb), GFP_KERNEL); + if (!scb) + goto error; + key = afs_request_key(dvnode->volume->cell); if (IS_ERR(key)) { ret = PTR_ERR(key); - goto error; + goto error_scb; } ret = -ERESTARTSYS; - if (afs_begin_vnode_operation(&fc, dvnode, key)) { + if (afs_begin_vnode_operation(&fc, dvnode, key, true)) { + afs_dataversion_t data_version = dvnode->status.data_version + 1; + if (mutex_lock_interruptible_nested(&vnode->io_lock, 1) < 0) { afs_end_vnode_operation(&fc); goto error_key; @@ -1519,11 +1605,14 @@ static int afs_link(struct dentry *from, struct inode *dir, while (afs_select_fileserver(&fc)) { fc.cb_break = afs_calc_vnode_cb_break(dvnode); fc.cb_break_2 = afs_calc_vnode_cb_break(vnode); - afs_fs_link(&fc, vnode, dentry->d_name.name, data_version); + afs_fs_link(&fc, vnode, dentry->d_name.name, + &scb[0], &scb[1]); } - afs_vnode_commit_status(&fc, dvnode, fc.cb_break); - afs_vnode_commit_status(&fc, vnode, fc.cb_break_2); + afs_vnode_commit_status(&fc, dvnode, fc.cb_break, + &data_version, &scb[0]); + afs_vnode_commit_status(&fc, vnode, fc.cb_break_2, + NULL, &scb[1]); ihold(&vnode->vfs_inode); d_instantiate(dentry, &vnode->vfs_inode); @@ -1540,11 +1629,14 @@ static int afs_link(struct dentry *from, struct inode *dir, afs_edit_dir_for_link); key_put(key); + kfree(scb); _leave(" = 0"); return 0; error_key: key_put(key); +error_scb: + kfree(scb); error: d_drop(dentry); _leave(" = %d", ret); @@ -1557,12 +1649,11 @@ error: static int afs_symlink(struct inode *dir, struct dentry *dentry, const char *content) { + struct afs_iget_data iget_data; struct afs_fs_cursor fc; - struct afs_file_status newstatus; + struct afs_status_cb *scb; struct afs_vnode *dvnode = AFS_FS_I(dir); - struct afs_fid newfid; struct key *key; - u64 data_version = dvnode->status.data_version; int ret; _enter("{%llx:%llu},{%pd},%s", @@ -1577,24 +1668,32 @@ static int afs_symlink(struct inode *dir, struct dentry *dentry, if (strlen(content) >= AFSPATHMAX) goto error; + ret = -ENOMEM; + scb = kcalloc(2, sizeof(struct afs_status_cb), GFP_KERNEL); + if (!scb) + goto error; + key = afs_request_key(dvnode->volume->cell); if (IS_ERR(key)) { ret = PTR_ERR(key); - goto error; + goto error_scb; } ret = -ERESTARTSYS; - if (afs_begin_vnode_operation(&fc, dvnode, key)) { + if (afs_begin_vnode_operation(&fc, dvnode, key, true)) { + afs_dataversion_t data_version = dvnode->status.data_version + 1; + while (afs_select_fileserver(&fc)) { fc.cb_break = afs_calc_vnode_cb_break(dvnode); - afs_fs_symlink(&fc, dentry->d_name.name, - content, data_version, - &newfid, &newstatus); + afs_prep_for_new_inode(&fc, &iget_data); + afs_fs_symlink(&fc, dentry->d_name.name, content, + &scb[0], &iget_data.fid, &scb[1]); } - afs_check_for_remote_deletion(&fc, fc.vnode); - afs_vnode_commit_status(&fc, dvnode, fc.cb_break); - afs_vnode_new_inode(&fc, dentry, &newfid, &newstatus, NULL); + afs_check_for_remote_deletion(&fc, dvnode); + afs_vnode_commit_status(&fc, dvnode, fc.cb_break, + &data_version, &scb[0]); + afs_vnode_new_inode(&fc, dentry, &iget_data, &scb[1]); ret = afs_end_vnode_operation(&fc); if (ret < 0) goto error_key; @@ -1603,15 +1702,18 @@ static int afs_symlink(struct inode *dir, struct dentry *dentry, } if (test_bit(AFS_VNODE_DIR_VALID, &dvnode->flags)) - afs_edit_dir_add(dvnode, &dentry->d_name, &newfid, + afs_edit_dir_add(dvnode, &dentry->d_name, &iget_data.fid, afs_edit_dir_for_symlink); key_put(key); + kfree(scb); _leave(" = 0"); return 0; error_key: key_put(key); +error_scb: + kfree(scb); error: d_drop(dentry); _leave(" = %d", ret); @@ -1626,11 +1728,11 @@ static int afs_rename(struct inode *old_dir, struct dentry *old_dentry, unsigned int flags) { struct afs_fs_cursor fc; + struct afs_status_cb *scb; struct afs_vnode *orig_dvnode, *new_dvnode, *vnode; struct dentry *tmp = NULL, *rehash = NULL; struct inode *new_inode; struct key *key; - u64 orig_data_version, new_data_version; bool new_negative = d_is_negative(new_dentry); int ret; @@ -1644,8 +1746,6 @@ static int afs_rename(struct inode *old_dir, struct dentry *old_dentry, vnode = AFS_FS_I(d_inode(old_dentry)); orig_dvnode = AFS_FS_I(old_dir); new_dvnode = AFS_FS_I(new_dir); - orig_data_version = orig_dvnode->status.data_version; - new_data_version = new_dvnode->status.data_version; _enter("{%llx:%llu},{%llx:%llu},{%llx:%llu},{%pd}", orig_dvnode->fid.vid, orig_dvnode->fid.vnode, @@ -1653,10 +1753,15 @@ static int afs_rename(struct inode *old_dir, struct dentry *old_dentry, new_dvnode->fid.vid, new_dvnode->fid.vnode, new_dentry); + ret = -ENOMEM; + scb = kcalloc(2, sizeof(struct afs_status_cb), GFP_KERNEL); + if (!scb) + goto error; + key = afs_request_key(orig_dvnode->volume->cell); if (IS_ERR(key)) { ret = PTR_ERR(key); - goto error; + goto error_scb; } /* For non-directories, check whether the target is busy and if so, @@ -1690,31 +1795,43 @@ static int afs_rename(struct inode *old_dir, struct dentry *old_dentry, new_dentry = tmp; rehash = NULL; new_negative = true; - orig_data_version = orig_dvnode->status.data_version; - new_data_version = new_dvnode->status.data_version; } } ret = -ERESTARTSYS; - if (afs_begin_vnode_operation(&fc, orig_dvnode, key)) { + if (afs_begin_vnode_operation(&fc, orig_dvnode, key, true)) { + afs_dataversion_t orig_data_version; + afs_dataversion_t new_data_version; + struct afs_status_cb *new_scb = &scb[1]; + + orig_data_version = orig_dvnode->status.data_version + 1; + if (orig_dvnode != new_dvnode) { if (mutex_lock_interruptible_nested(&new_dvnode->io_lock, 1) < 0) { afs_end_vnode_operation(&fc); goto error_rehash; } + new_data_version = new_dvnode->status.data_version; + } else { + new_data_version = orig_data_version; + new_scb = &scb[0]; } + while (afs_select_fileserver(&fc)) { fc.cb_break = afs_calc_vnode_cb_break(orig_dvnode); fc.cb_break_2 = afs_calc_vnode_cb_break(new_dvnode); afs_fs_rename(&fc, old_dentry->d_name.name, new_dvnode, new_dentry->d_name.name, - orig_data_version, new_data_version); + &scb[0], new_scb); } - afs_vnode_commit_status(&fc, orig_dvnode, fc.cb_break); - afs_vnode_commit_status(&fc, new_dvnode, fc.cb_break_2); - if (orig_dvnode != new_dvnode) + afs_vnode_commit_status(&fc, orig_dvnode, fc.cb_break, + &orig_data_version, &scb[0]); + if (new_dvnode != orig_dvnode) { + afs_vnode_commit_status(&fc, new_dvnode, fc.cb_break_2, + &new_data_version, &scb[1]); mutex_unlock(&new_dvnode->io_lock); + } ret = afs_end_vnode_operation(&fc); if (ret < 0) goto error_rehash; @@ -1754,6 +1871,8 @@ error_tmp: if (tmp) dput(tmp); key_put(key); +error_scb: + kfree(scb); error: _leave(" = %d", ret); return ret; diff --git a/fs/afs/dir_silly.c b/fs/afs/dir_silly.c index f6f89fdab6b2..28f4aa015229 100644 --- a/fs/afs/dir_silly.c +++ b/fs/afs/dir_silly.c @@ -24,21 +24,28 @@ static int afs_do_silly_rename(struct afs_vnode *dvnode, struct afs_vnode *vnode struct key *key) { struct afs_fs_cursor fc; - u64 dir_data_version = dvnode->status.data_version; + struct afs_status_cb *scb; int ret = -ERESTARTSYS; _enter("%pd,%pd", old, new); + scb = kzalloc(sizeof(struct afs_status_cb), GFP_KERNEL); + if (!scb) + return -ENOMEM; + trace_afs_silly_rename(vnode, false); - if (afs_begin_vnode_operation(&fc, dvnode, key)) { + if (afs_begin_vnode_operation(&fc, dvnode, key, true)) { + afs_dataversion_t dir_data_version = dvnode->status.data_version + 1; + while (afs_select_fileserver(&fc)) { fc.cb_break = afs_calc_vnode_cb_break(dvnode); afs_fs_rename(&fc, old->d_name.name, dvnode, new->d_name.name, - dir_data_version, dir_data_version); + scb, scb); } - afs_vnode_commit_status(&fc, dvnode, fc.cb_break); + afs_vnode_commit_status(&fc, dvnode, fc.cb_break, + &dir_data_version, scb); ret = afs_end_vnode_operation(&fc); } @@ -64,6 +71,7 @@ static int afs_do_silly_rename(struct afs_vnode *dvnode, struct afs_vnode *vnode fsnotify_nameremove(old, 0); } + kfree(scb); _leave(" = %d", ret); return ret; } @@ -143,31 +151,37 @@ static int afs_do_silly_unlink(struct afs_vnode *dvnode, struct afs_vnode *vnode struct dentry *dentry, struct key *key) { struct afs_fs_cursor fc; - u64 dir_data_version = dvnode->status.data_version; + struct afs_status_cb *scb; int ret = -ERESTARTSYS; _enter(""); + scb = kcalloc(2, sizeof(struct afs_status_cb), GFP_KERNEL); + if (!scb) + return -ENOMEM; + trace_afs_silly_rename(vnode, true); - if (afs_begin_vnode_operation(&fc, dvnode, key)) { + if (afs_begin_vnode_operation(&fc, dvnode, key, false)) { + afs_dataversion_t dir_data_version = dvnode->status.data_version + 1; + while (afs_select_fileserver(&fc)) { fc.cb_break = afs_calc_vnode_cb_break(dvnode); if (test_bit(AFS_SERVER_FL_IS_YFS, &fc.cbi->server->flags) && !test_bit(AFS_SERVER_FL_NO_RM2, &fc.cbi->server->flags)) { yfs_fs_remove_file2(&fc, vnode, dentry->d_name.name, - dir_data_version); + &scb[0], &scb[1]); if (fc.ac.error != -ECONNABORTED || fc.ac.abort_code != RXGEN_OPCODE) continue; set_bit(AFS_SERVER_FL_NO_RM2, &fc.cbi->server->flags); } - afs_fs_remove(&fc, vnode, dentry->d_name.name, false, - dir_data_version); + afs_fs_remove(&fc, vnode, dentry->d_name.name, false, &scb[0]); } - afs_vnode_commit_status(&fc, dvnode, fc.cb_break); + afs_vnode_commit_status(&fc, dvnode, fc.cb_break, + &dir_data_version, &scb[0]); ret = afs_end_vnode_operation(&fc); if (ret == 0) { drop_nlink(&vnode->vfs_inode); @@ -182,6 +196,7 @@ static int afs_do_silly_unlink(struct afs_vnode *dvnode, struct afs_vnode *vnode afs_edit_dir_for_unlink); } + kfree(scb); _leave(" = %d", ret); return ret; } diff --git a/fs/afs/dynroot.c b/fs/afs/dynroot.c index a9ba81ddf154..af1689d1f32e 100644 --- a/fs/afs/dynroot.c +++ b/fs/afs/dynroot.c @@ -46,7 +46,7 @@ static int afs_probe_cell_name(struct dentry *dentry) return 0; } - ret = dns_query("afsdb", name, len, "srv=1", NULL, NULL); + ret = dns_query("afsdb", name, len, "srv=1", NULL, NULL, false); if (ret == -ENODATA) ret = -EDESTADDRREQ; return ret; @@ -261,8 +261,7 @@ int afs_dynroot_populate(struct super_block *sb) struct afs_net *net = afs_sb2net(sb); int ret; - if (mutex_lock_interruptible(&net->proc_cells_lock) < 0) - return -ERESTARTSYS; + mutex_lock(&net->proc_cells_lock); net->dynroot_sb = sb; hlist_for_each_entry(cell, &net->proc_cells, proc_link) { diff --git a/fs/afs/file.c b/fs/afs/file.c index e8d6619890a9..11e69c5fb7ab 100644 --- a/fs/afs/file.c +++ b/fs/afs/file.c @@ -170,11 +170,12 @@ int afs_release(struct inode *inode, struct file *file) { struct afs_vnode *vnode = AFS_FS_I(inode); struct afs_file *af = file->private_data; + int ret = 0; _enter("{%llx:%llu},", vnode->fid.vid, vnode->fid.vnode); if ((file->f_mode & FMODE_WRITE)) - return vfs_fsync(file, 0); + ret = vfs_fsync(file, 0); file->private_data = NULL; if (af->wb) @@ -182,8 +183,8 @@ int afs_release(struct inode *inode, struct file *file) key_put(af->key); kfree(af); afs_prune_wb_keys(vnode); - _leave(" = 0"); - return 0; + _leave(" = %d", ret); + return ret; } /* @@ -227,6 +228,7 @@ static void afs_file_readpage_read_complete(struct page *page, int afs_fetch_data(struct afs_vnode *vnode, struct key *key, struct afs_read *desc) { struct afs_fs_cursor fc; + struct afs_status_cb *scb; int ret; _enter("%s{%llx:%llu.%u},%x,,,", @@ -236,15 +238,22 @@ int afs_fetch_data(struct afs_vnode *vnode, struct key *key, struct afs_read *de vnode->fid.unique, key_serial(key)); + scb = kzalloc(sizeof(struct afs_status_cb), GFP_KERNEL); + if (!scb) + return -ENOMEM; + ret = -ERESTARTSYS; - if (afs_begin_vnode_operation(&fc, vnode, key)) { + if (afs_begin_vnode_operation(&fc, vnode, key, true)) { + afs_dataversion_t data_version = vnode->status.data_version; + while (afs_select_fileserver(&fc)) { fc.cb_break = afs_calc_vnode_cb_break(vnode); - afs_fs_fetch_data(&fc, desc); + afs_fs_fetch_data(&fc, scb, desc); } - afs_check_for_remote_deletion(&fc, fc.vnode); - afs_vnode_commit_status(&fc, vnode, fc.cb_break); + afs_check_for_remote_deletion(&fc, vnode); + afs_vnode_commit_status(&fc, vnode, fc.cb_break, + &data_version, scb); ret = afs_end_vnode_operation(&fc); } @@ -254,6 +263,7 @@ int afs_fetch_data(struct afs_vnode *vnode, struct key *key, struct afs_read *de &afs_v2net(vnode)->n_fetch_bytes); } + kfree(scb); _leave(" = %d", ret); return ret; } @@ -404,10 +414,10 @@ static int afs_readpage(struct file *file, struct page *page) /* * Make pages available as they're filled. */ -static void afs_readpages_page_done(struct afs_call *call, struct afs_read *req) +static void afs_readpages_page_done(struct afs_read *req) { #ifdef CONFIG_AFS_FSCACHE - struct afs_vnode *vnode = call->reply[0]; + struct afs_vnode *vnode = req->vnode; #endif struct page *page = req->pages[req->index]; @@ -461,6 +471,7 @@ static int afs_readpages_one(struct file *file, struct address_space *mapping, return -ENOMEM; refcount_set(&req->usage, 1); + req->vnode = vnode; req->page_done = afs_readpages_page_done; req->pos = first->index; req->pos <<= PAGE_SHIFT; diff --git a/fs/afs/flock.c b/fs/afs/flock.c index adc88eff7849..ed3ac03682d7 100644 --- a/fs/afs/flock.c +++ b/fs/afs/flock.c @@ -41,9 +41,6 @@ void afs_lock_may_be_available(struct afs_vnode *vnode) { _enter("{%llx:%llu}", vnode->fid.vid, vnode->fid.vnode); - if (vnode->lock_state != AFS_VNODE_LOCK_WAITING_FOR_CB) - return; - spin_lock(&vnode->lock); if (vnode->lock_state == AFS_VNODE_LOCK_WAITING_FOR_CB) afs_next_locker(vnode, 0); @@ -77,7 +74,7 @@ static void afs_schedule_lock_extension(struct afs_vnode *vnode) */ void afs_lock_op_done(struct afs_call *call) { - struct afs_vnode *vnode = call->reply[0]; + struct afs_vnode *vnode = call->lvnode; if (call->error == 0) { spin_lock(&vnode->lock); @@ -185,6 +182,7 @@ static void afs_kill_lockers_enoent(struct afs_vnode *vnode) static int afs_set_lock(struct afs_vnode *vnode, struct key *key, afs_lock_type_t type) { + struct afs_status_cb *scb; struct afs_fs_cursor fc; int ret; @@ -195,18 +193,23 @@ static int afs_set_lock(struct afs_vnode *vnode, struct key *key, vnode->fid.unique, key_serial(key), type); + scb = kzalloc(sizeof(struct afs_status_cb), GFP_KERNEL); + if (!scb) + return -ENOMEM; + ret = -ERESTARTSYS; - if (afs_begin_vnode_operation(&fc, vnode, key)) { + if (afs_begin_vnode_operation(&fc, vnode, key, true)) { while (afs_select_fileserver(&fc)) { fc.cb_break = afs_calc_vnode_cb_break(vnode); - afs_fs_set_lock(&fc, type); + afs_fs_set_lock(&fc, type, scb); } - afs_check_for_remote_deletion(&fc, fc.vnode); - afs_vnode_commit_status(&fc, vnode, fc.cb_break); + afs_check_for_remote_deletion(&fc, vnode); + afs_vnode_commit_status(&fc, vnode, fc.cb_break, NULL, scb); ret = afs_end_vnode_operation(&fc); } + kfree(scb); _leave(" = %d", ret); return ret; } @@ -216,6 +219,7 @@ static int afs_set_lock(struct afs_vnode *vnode, struct key *key, */ static int afs_extend_lock(struct afs_vnode *vnode, struct key *key) { + struct afs_status_cb *scb; struct afs_fs_cursor fc; int ret; @@ -226,18 +230,23 @@ static int afs_extend_lock(struct afs_vnode *vnode, struct key *key) vnode->fid.unique, key_serial(key)); + scb = kzalloc(sizeof(struct afs_status_cb), GFP_KERNEL); + if (!scb) + return -ENOMEM; + ret = -ERESTARTSYS; - if (afs_begin_vnode_operation(&fc, vnode, key)) { + if (afs_begin_vnode_operation(&fc, vnode, key, false)) { while (afs_select_current_fileserver(&fc)) { fc.cb_break = afs_calc_vnode_cb_break(vnode); - afs_fs_extend_lock(&fc); + afs_fs_extend_lock(&fc, scb); } - afs_check_for_remote_deletion(&fc, fc.vnode); - afs_vnode_commit_status(&fc, vnode, fc.cb_break); + afs_check_for_remote_deletion(&fc, vnode); + afs_vnode_commit_status(&fc, vnode, fc.cb_break, NULL, scb); ret = afs_end_vnode_operation(&fc); } + kfree(scb); _leave(" = %d", ret); return ret; } @@ -247,6 +256,7 @@ static int afs_extend_lock(struct afs_vnode *vnode, struct key *key) */ static int afs_release_lock(struct afs_vnode *vnode, struct key *key) { + struct afs_status_cb *scb; struct afs_fs_cursor fc; int ret; @@ -257,18 +267,23 @@ static int afs_release_lock(struct afs_vnode *vnode, struct key *key) vnode->fid.unique, key_serial(key)); + scb = kzalloc(sizeof(struct afs_status_cb), GFP_KERNEL); + if (!scb) + return -ENOMEM; + ret = -ERESTARTSYS; - if (afs_begin_vnode_operation(&fc, vnode, key)) { + if (afs_begin_vnode_operation(&fc, vnode, key, false)) { while (afs_select_current_fileserver(&fc)) { fc.cb_break = afs_calc_vnode_cb_break(vnode); - afs_fs_release_lock(&fc); + afs_fs_release_lock(&fc, scb); } - afs_check_for_remote_deletion(&fc, fc.vnode); - afs_vnode_commit_status(&fc, vnode, fc.cb_break); + afs_check_for_remote_deletion(&fc, vnode); + afs_vnode_commit_status(&fc, vnode, fc.cb_break, NULL, scb); ret = afs_end_vnode_operation(&fc); } + kfree(scb); _leave(" = %d", ret); return ret; } @@ -736,7 +751,7 @@ static int afs_do_getlk(struct file *file, struct file_lock *fl) posix_test_lock(file, fl); if (fl->fl_type == F_UNLCK) { /* no local locks; consult the server */ - ret = afs_fetch_status(vnode, key, false); + ret = afs_fetch_status(vnode, key, false, NULL); if (ret < 0) goto error; diff --git a/fs/afs/fs_probe.c b/fs/afs/fs_probe.c index 5d3abde52a0f..9b7266209343 100644 --- a/fs/afs/fs_probe.c +++ b/fs/afs/fs_probe.c @@ -33,8 +33,8 @@ static bool afs_fs_probe_done(struct afs_server *server) void afs_fileserver_probe_result(struct afs_call *call) { struct afs_addr_list *alist = call->alist; - struct afs_server *server = call->reply[0]; - unsigned int server_index = (long)call->reply[1]; + struct afs_server *server = call->server; + unsigned int server_index = call->server_index; unsigned int index = call->addr_ix; unsigned int rtt = UINT_MAX; bool have_result = false; diff --git a/fs/afs/fsclient.c b/fs/afs/fsclient.c index 1296f5dc4c1e..48298408d6ac 100644 --- a/fs/afs/fsclient.c +++ b/fs/afs/fsclient.c @@ -60,78 +60,17 @@ static void xdr_dump_bad(const __be32 *bp) } /* - * Update the core inode struct from a returned status record. - */ -void afs_update_inode_from_status(struct afs_vnode *vnode, - struct afs_file_status *status, - const afs_dataversion_t *expected_version, - u8 flags) -{ - struct timespec64 t; - umode_t mode; - - t = status->mtime_client; - vnode->vfs_inode.i_ctime = t; - vnode->vfs_inode.i_mtime = t; - vnode->vfs_inode.i_atime = t; - - if (flags & (AFS_VNODE_META_CHANGED | AFS_VNODE_NOT_YET_SET)) { - vnode->vfs_inode.i_uid = make_kuid(&init_user_ns, status->owner); - vnode->vfs_inode.i_gid = make_kgid(&init_user_ns, status->group); - set_nlink(&vnode->vfs_inode, status->nlink); - - mode = vnode->vfs_inode.i_mode; - mode &= ~S_IALLUGO; - mode |= status->mode; - barrier(); - vnode->vfs_inode.i_mode = mode; - } - - if (!(flags & AFS_VNODE_NOT_YET_SET)) { - if (expected_version && - *expected_version != status->data_version) { - _debug("vnode modified %llx on {%llx:%llu} [exp %llx]", - (unsigned long long) status->data_version, - vnode->fid.vid, vnode->fid.vnode, - (unsigned long long) *expected_version); - vnode->invalid_before = status->data_version; - if (vnode->status.type == AFS_FTYPE_DIR) { - if (test_and_clear_bit(AFS_VNODE_DIR_VALID, &vnode->flags)) - afs_stat_v(vnode, n_inval); - } else { - set_bit(AFS_VNODE_ZAP_DATA, &vnode->flags); - } - } else if (vnode->status.type == AFS_FTYPE_DIR) { - /* Expected directory change is handled elsewhere so - * that we can locally edit the directory and save on a - * download. - */ - if (test_bit(AFS_VNODE_DIR_VALID, &vnode->flags)) - flags &= ~AFS_VNODE_DATA_CHANGED; - } - } - - if (flags & (AFS_VNODE_DATA_CHANGED | AFS_VNODE_NOT_YET_SET)) { - inode_set_iversion_raw(&vnode->vfs_inode, status->data_version); - i_size_write(&vnode->vfs_inode, status->size); - } -} - -/* * decode an AFSFetchStatus block */ -static int xdr_decode_AFSFetchStatus(struct afs_call *call, - const __be32 **_bp, - struct afs_file_status *status, - struct afs_vnode *vnode, - const afs_dataversion_t *expected_version, - struct afs_read *read_req) +static int xdr_decode_AFSFetchStatus(const __be32 **_bp, + struct afs_call *call, + struct afs_status_cb *scb) { const struct afs_xdr_AFSFetchStatus *xdr = (const void *)*_bp; + struct afs_file_status *status = &scb->status; bool inline_error = (call->operation_ID == afs_FS_InlineBulkStatus); u64 data_version, size; u32 type, abort_code; - u8 flags = 0; abort_code = ntohl(xdr->abort_code); @@ -144,6 +83,7 @@ static int xdr_decode_AFSFetchStatus(struct afs_call *call, * case. */ status->abort_code = abort_code; + scb->have_error = true; return 0; } @@ -161,44 +101,25 @@ static int xdr_decode_AFSFetchStatus(struct afs_call *call, case AFS_FTYPE_FILE: case AFS_FTYPE_DIR: case AFS_FTYPE_SYMLINK: - if (type != status->type && - vnode && - !test_bit(AFS_VNODE_UNSET, &vnode->flags)) { - pr_warning("Vnode %llx:%llx:%x changed type %u to %u\n", - vnode->fid.vid, - vnode->fid.vnode, - vnode->fid.unique, - status->type, type); - goto bad; - } status->type = type; break; default: goto bad; } -#define EXTRACT_M(FIELD) \ - do { \ - u32 x = ntohl(xdr->FIELD); \ - if (status->FIELD != x) { \ - flags |= AFS_VNODE_META_CHANGED; \ - status->FIELD = x; \ - } \ - } while (0) - - EXTRACT_M(nlink); - EXTRACT_M(author); - EXTRACT_M(owner); - EXTRACT_M(caller_access); /* call ticket dependent */ - EXTRACT_M(anon_access); - EXTRACT_M(mode); - EXTRACT_M(group); + status->nlink = ntohl(xdr->nlink); + status->author = ntohl(xdr->author); + status->owner = ntohl(xdr->owner); + status->caller_access = ntohl(xdr->caller_access); /* Ticket dependent */ + status->anon_access = ntohl(xdr->anon_access); + status->mode = ntohl(xdr->mode) & S_IALLUGO; + status->group = ntohl(xdr->group); + status->lock_count = ntohl(xdr->lock_count); status->mtime_client.tv_sec = ntohl(xdr->mtime_client); status->mtime_client.tv_nsec = 0; status->mtime_server.tv_sec = ntohl(xdr->mtime_server); status->mtime_server.tv_nsec = 0; - status->lock_count = ntohl(xdr->lock_count); size = (u64)ntohl(xdr->size_lo); size |= (u64)ntohl(xdr->size_hi) << 32; @@ -206,25 +127,10 @@ static int xdr_decode_AFSFetchStatus(struct afs_call *call, data_version = (u64)ntohl(xdr->data_version_lo); data_version |= (u64)ntohl(xdr->data_version_hi) << 32; - if (data_version != status->data_version) { - status->data_version = data_version; - flags |= AFS_VNODE_DATA_CHANGED; - } - - if (read_req) { - read_req->data_version = data_version; - read_req->file_size = size; - } + status->data_version = data_version; + scb->have_status = true; *_bp = (const void *)*_bp + sizeof(*xdr); - - if (vnode) { - if (test_bit(AFS_VNODE_UNSET, &vnode->flags)) - flags |= AFS_VNODE_NOT_YET_SET; - afs_update_inode_from_status(vnode, status, expected_version, - flags); - } - return 0; bad: @@ -232,77 +138,22 @@ bad: return afs_protocol_error(call, -EBADMSG, afs_eproto_bad_status); } -/* - * Decode the file status. We need to lock the target vnode if we're going to - * update its status so that stat() sees the attributes update atomically. - */ -static int afs_decode_status(struct afs_call *call, - const __be32 **_bp, - struct afs_file_status *status, - struct afs_vnode *vnode, - const afs_dataversion_t *expected_version, - struct afs_read *read_req) +static time64_t xdr_decode_expiry(struct afs_call *call, u32 expiry) { - int ret; - - if (!vnode) - return xdr_decode_AFSFetchStatus(call, _bp, status, vnode, - expected_version, read_req); - - write_seqlock(&vnode->cb_lock); - ret = xdr_decode_AFSFetchStatus(call, _bp, status, vnode, - expected_version, read_req); - write_sequnlock(&vnode->cb_lock); - return ret; + return ktime_divns(call->reply_time, NSEC_PER_SEC) + expiry; } -/* - * decode an AFSCallBack block - */ -static void xdr_decode_AFSCallBack(struct afs_call *call, - struct afs_vnode *vnode, - const __be32 **_bp) +static void xdr_decode_AFSCallBack(const __be32 **_bp, + struct afs_call *call, + struct afs_status_cb *scb) { - struct afs_cb_interest *old, *cbi = call->cbi; + struct afs_callback *cb = &scb->callback; const __be32 *bp = *_bp; - u32 cb_expiry; - - write_seqlock(&vnode->cb_lock); - - if (!afs_cb_is_broken(call->cb_break, vnode, cbi)) { - vnode->cb_version = ntohl(*bp++); - cb_expiry = ntohl(*bp++); - vnode->cb_type = ntohl(*bp++); - vnode->cb_expires_at = cb_expiry + ktime_get_real_seconds(); - old = vnode->cb_interest; - if (old != call->cbi) { - vnode->cb_interest = cbi; - cbi = old; - } - set_bit(AFS_VNODE_CB_PROMISED, &vnode->flags); - } else { - bp += 3; - } - write_sequnlock(&vnode->cb_lock); - call->cbi = cbi; - *_bp = bp; -} - -static ktime_t xdr_decode_expiry(struct afs_call *call, u32 expiry) -{ - return ktime_add_ns(call->reply_time, expiry * NSEC_PER_SEC); -} - -static void xdr_decode_AFSCallBack_raw(struct afs_call *call, - const __be32 **_bp, - struct afs_callback *cb) -{ - const __be32 *bp = *_bp; - - cb->version = ntohl(*bp++); + bp++; /* version */ cb->expires_at = xdr_decode_expiry(call, ntohl(*bp++)); - cb->type = ntohl(*bp++); + bp++; /* type */ + scb->have_cb = true; *_bp = bp; } @@ -395,7 +246,6 @@ static void xdr_decode_AFSFetchVolumeStatus(const __be32 **_bp, */ static int afs_deliver_fs_fetch_status_vnode(struct afs_call *call) { - struct afs_vnode *vnode = call->reply[0]; const __be32 *bp; int ret; @@ -403,16 +253,13 @@ static int afs_deliver_fs_fetch_status_vnode(struct afs_call *call) if (ret < 0) return ret; - _enter("{%llx:%llu}", vnode->fid.vid, vnode->fid.vnode); - /* unmarshall the reply once we've received all of it */ bp = call->buffer; - ret = afs_decode_status(call, &bp, &vnode->status, vnode, - &call->expected_version, NULL); + ret = xdr_decode_AFSFetchStatus(&bp, call, call->out_scb); if (ret < 0) return ret; - xdr_decode_AFSCallBack(call, vnode, &bp); - xdr_decode_AFSVolSync(&bp, call->reply[1]); + xdr_decode_AFSCallBack(&bp, call, call->out_scb); + xdr_decode_AFSVolSync(&bp, call->out_volsync); _leave(" = 0 [done]"); return 0; @@ -431,8 +278,8 @@ static const struct afs_call_type afs_RXFSFetchStatus_vnode = { /* * fetch the status information for a file */ -int afs_fs_fetch_file_status(struct afs_fs_cursor *fc, struct afs_volsync *volsync, - bool new_inode) +int afs_fs_fetch_file_status(struct afs_fs_cursor *fc, struct afs_status_cb *scb, + struct afs_volsync *volsync) { struct afs_vnode *vnode = fc->vnode; struct afs_call *call; @@ -440,7 +287,7 @@ int afs_fs_fetch_file_status(struct afs_fs_cursor *fc, struct afs_volsync *volsy __be32 *bp; if (test_bit(AFS_SERVER_FL_IS_YFS, &fc->cbi->server->flags)) - return yfs_fs_fetch_file_status(fc, volsync, new_inode); + return yfs_fs_fetch_file_status(fc, scb, volsync); _enter(",%x,{%llx:%llu},,", key_serial(fc->key), vnode->fid.vid, vnode->fid.vnode); @@ -453,10 +300,8 @@ int afs_fs_fetch_file_status(struct afs_fs_cursor *fc, struct afs_volsync *volsy } call->key = fc->key; - call->reply[0] = vnode; - call->reply[1] = volsync; - call->expected_version = new_inode ? 1 : vnode->status.data_version; - call->want_reply_time = true; + call->out_scb = scb; + call->out_volsync = volsync; /* marshall the parameters */ bp = call->request; @@ -465,10 +310,10 @@ int afs_fs_fetch_file_status(struct afs_fs_cursor *fc, struct afs_volsync *volsy bp[2] = htonl(vnode->fid.vnode); bp[3] = htonl(vnode->fid.unique); - call->cb_break = fc->cb_break; afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_call(call, &vnode->fid); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } @@ -478,8 +323,7 @@ int afs_fs_fetch_file_status(struct afs_fs_cursor *fc, struct afs_volsync *volsy */ static int afs_deliver_fs_fetch_data(struct afs_call *call) { - struct afs_vnode *vnode = call->reply[0]; - struct afs_read *req = call->reply[2]; + struct afs_read *req = call->read_request; const __be32 *bp; unsigned int size; int ret; @@ -541,7 +385,7 @@ static int afs_deliver_fs_fetch_data(struct afs_call *call) if (req->offset == PAGE_SIZE) { req->offset = 0; if (req->page_done) - req->page_done(call, req); + req->page_done(req); req->index++; if (req->remain > 0) goto begin_page; @@ -575,12 +419,14 @@ static int afs_deliver_fs_fetch_data(struct afs_call *call) return ret; bp = call->buffer; - ret = afs_decode_status(call, &bp, &vnode->status, vnode, - &vnode->status.data_version, req); + ret = xdr_decode_AFSFetchStatus(&bp, call, call->out_scb); if (ret < 0) return ret; - xdr_decode_AFSCallBack(call, vnode, &bp); - xdr_decode_AFSVolSync(&bp, call->reply[1]); + xdr_decode_AFSCallBack(&bp, call, call->out_scb); + xdr_decode_AFSVolSync(&bp, call->out_volsync); + + req->data_version = call->out_scb->status.data_version; + req->file_size = call->out_scb->status.size; call->unmarshall++; @@ -593,7 +439,7 @@ static int afs_deliver_fs_fetch_data(struct afs_call *call) zero_user_segment(req->pages[req->index], req->offset, PAGE_SIZE); if (req->page_done) - req->page_done(call, req); + req->page_done(req); req->offset = 0; } @@ -603,7 +449,7 @@ static int afs_deliver_fs_fetch_data(struct afs_call *call) static void afs_fetch_data_destructor(struct afs_call *call) { - struct afs_read *req = call->reply[2]; + struct afs_read *req = call->read_request; afs_put_read(req); afs_flat_call_destructor(call); @@ -629,7 +475,9 @@ static const struct afs_call_type afs_RXFSFetchData64 = { /* * fetch data from a very large file */ -static int afs_fs_fetch_data64(struct afs_fs_cursor *fc, struct afs_read *req) +static int afs_fs_fetch_data64(struct afs_fs_cursor *fc, + struct afs_status_cb *scb, + struct afs_read *req) { struct afs_vnode *vnode = fc->vnode; struct afs_call *call; @@ -643,11 +491,9 @@ static int afs_fs_fetch_data64(struct afs_fs_cursor *fc, struct afs_read *req) return -ENOMEM; call->key = fc->key; - call->reply[0] = vnode; - call->reply[1] = NULL; /* volsync */ - call->reply[2] = req; - call->expected_version = vnode->status.data_version; - call->want_reply_time = true; + call->out_scb = scb; + call->out_volsync = NULL; + call->read_request = req; /* marshall the parameters */ bp = call->request; @@ -661,9 +507,9 @@ static int afs_fs_fetch_data64(struct afs_fs_cursor *fc, struct afs_read *req) bp[7] = htonl(lower_32_bits(req->len)); refcount_inc(&req->usage); - call->cb_break = fc->cb_break; afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_call(call, &vnode->fid); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } @@ -671,7 +517,9 @@ static int afs_fs_fetch_data64(struct afs_fs_cursor *fc, struct afs_read *req) /* * fetch data from a file */ -int afs_fs_fetch_data(struct afs_fs_cursor *fc, struct afs_read *req) +int afs_fs_fetch_data(struct afs_fs_cursor *fc, + struct afs_status_cb *scb, + struct afs_read *req) { struct afs_vnode *vnode = fc->vnode; struct afs_call *call; @@ -679,12 +527,12 @@ int afs_fs_fetch_data(struct afs_fs_cursor *fc, struct afs_read *req) __be32 *bp; if (test_bit(AFS_SERVER_FL_IS_YFS, &fc->cbi->server->flags)) - return yfs_fs_fetch_data(fc, req); + return yfs_fs_fetch_data(fc, scb, req); if (upper_32_bits(req->pos) || upper_32_bits(req->len) || upper_32_bits(req->pos + req->len)) - return afs_fs_fetch_data64(fc, req); + return afs_fs_fetch_data64(fc, scb, req); _enter(""); @@ -693,11 +541,9 @@ int afs_fs_fetch_data(struct afs_fs_cursor *fc, struct afs_read *req) return -ENOMEM; call->key = fc->key; - call->reply[0] = vnode; - call->reply[1] = NULL; /* volsync */ - call->reply[2] = req; - call->expected_version = vnode->status.data_version; - call->want_reply_time = true; + call->out_scb = scb; + call->out_volsync = NULL; + call->read_request = req; /* marshall the parameters */ bp = call->request; @@ -709,9 +555,9 @@ int afs_fs_fetch_data(struct afs_fs_cursor *fc, struct afs_read *req) bp[5] = htonl(lower_32_bits(req->len)); refcount_inc(&req->usage); - call->cb_break = fc->cb_break; afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_call(call, &vnode->fid); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } @@ -721,28 +567,24 @@ int afs_fs_fetch_data(struct afs_fs_cursor *fc, struct afs_read *req) */ static int afs_deliver_fs_create_vnode(struct afs_call *call) { - struct afs_vnode *vnode = call->reply[0]; const __be32 *bp; int ret; - _enter("{%u}", call->unmarshall); - ret = afs_transfer_reply(call); if (ret < 0) return ret; /* unmarshall the reply once we've received all of it */ bp = call->buffer; - xdr_decode_AFSFid(&bp, call->reply[1]); - ret = afs_decode_status(call, &bp, call->reply[2], NULL, NULL, NULL); + xdr_decode_AFSFid(&bp, call->out_fid); + ret = xdr_decode_AFSFetchStatus(&bp, call, call->out_scb); if (ret < 0) return ret; - ret = afs_decode_status(call, &bp, &vnode->status, vnode, - &call->expected_version, NULL); + ret = xdr_decode_AFSFetchStatus(&bp, call, call->out_dir_scb); if (ret < 0) return ret; - xdr_decode_AFSCallBack_raw(call, &bp, call->reply[3]); - /* xdr_decode_AFSVolSync(&bp, call->reply[X]); */ + xdr_decode_AFSCallBack(&bp, call, call->out_scb); + xdr_decode_AFSVolSync(&bp, call->out_volsync); _leave(" = 0 [done]"); return 0; @@ -771,24 +613,23 @@ static const struct afs_call_type afs_RXFSMakeDir = { int afs_fs_create(struct afs_fs_cursor *fc, const char *name, umode_t mode, - u64 current_data_version, + struct afs_status_cb *dvnode_scb, struct afs_fid *newfid, - struct afs_file_status *newstatus, - struct afs_callback *newcb) + struct afs_status_cb *new_scb) { - struct afs_vnode *vnode = fc->vnode; + struct afs_vnode *dvnode = fc->vnode; struct afs_call *call; - struct afs_net *net = afs_v2net(vnode); + struct afs_net *net = afs_v2net(dvnode); size_t namesz, reqsz, padsz; __be32 *bp; if (test_bit(AFS_SERVER_FL_IS_YFS, &fc->cbi->server->flags)){ if (S_ISDIR(mode)) - return yfs_fs_make_dir(fc, name, mode, current_data_version, - newfid, newstatus, newcb); + return yfs_fs_make_dir(fc, name, mode, dvnode_scb, + newfid, new_scb); else - return yfs_fs_create_file(fc, name, mode, current_data_version, - newfid, newstatus, newcb); + return yfs_fs_create_file(fc, name, mode, dvnode_scb, + newfid, new_scb); } _enter(""); @@ -804,19 +645,16 @@ int afs_fs_create(struct afs_fs_cursor *fc, return -ENOMEM; call->key = fc->key; - call->reply[0] = vnode; - call->reply[1] = newfid; - call->reply[2] = newstatus; - call->reply[3] = newcb; - call->expected_version = current_data_version + 1; - call->want_reply_time = true; + call->out_dir_scb = dvnode_scb; + call->out_fid = newfid; + call->out_scb = new_scb; /* marshall the parameters */ bp = call->request; *bp++ = htonl(S_ISDIR(mode) ? FSMAKEDIR : FSCREATEFILE); - *bp++ = htonl(vnode->fid.vid); - *bp++ = htonl(vnode->fid.vnode); - *bp++ = htonl(vnode->fid.unique); + *bp++ = htonl(dvnode->fid.vid); + *bp++ = htonl(dvnode->fid.vnode); + *bp++ = htonl(dvnode->fid.unique); *bp++ = htonl(namesz); memcpy(bp, name, namesz); bp = (void *) bp + namesz; @@ -825,41 +663,38 @@ int afs_fs_create(struct afs_fs_cursor *fc, bp = (void *) bp + padsz; } *bp++ = htonl(AFS_SET_MODE | AFS_SET_MTIME); - *bp++ = htonl(vnode->vfs_inode.i_mtime.tv_sec); /* mtime */ + *bp++ = htonl(dvnode->vfs_inode.i_mtime.tv_sec); /* mtime */ *bp++ = 0; /* owner */ *bp++ = 0; /* group */ *bp++ = htonl(mode & S_IALLUGO); /* unix mode */ *bp++ = 0; /* segment size */ afs_use_fs_server(call, fc->cbi); - trace_afs_make_fs_call1(call, &vnode->fid, name); + trace_afs_make_fs_call1(call, &dvnode->fid, name); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } /* - * Deliver reply data to any operation that returns file status and volume + * Deliver reply data to any operation that returns directory status and volume * sync. */ -static int afs_deliver_fs_status_and_vol(struct afs_call *call) +static int afs_deliver_fs_dir_status_and_vol(struct afs_call *call) { - struct afs_vnode *vnode = call->reply[0]; const __be32 *bp; int ret; - _enter("{%u}", call->unmarshall); - ret = afs_transfer_reply(call); if (ret < 0) return ret; /* unmarshall the reply once we've received all of it */ bp = call->buffer; - ret = afs_decode_status(call, &bp, &vnode->status, vnode, - &call->expected_version, NULL); + ret = xdr_decode_AFSFetchStatus(&bp, call, call->out_dir_scb); if (ret < 0) return ret; - /* xdr_decode_AFSVolSync(&bp, call->reply[X]); */ + xdr_decode_AFSVolSync(&bp, call->out_volsync); _leave(" = 0 [done]"); return 0; @@ -871,14 +706,14 @@ static int afs_deliver_fs_status_and_vol(struct afs_call *call) static const struct afs_call_type afs_RXFSRemoveFile = { .name = "FS.RemoveFile", .op = afs_FS_RemoveFile, - .deliver = afs_deliver_fs_status_and_vol, + .deliver = afs_deliver_fs_dir_status_and_vol, .destructor = afs_flat_call_destructor, }; static const struct afs_call_type afs_RXFSRemoveDir = { .name = "FS.RemoveDir", .op = afs_FS_RemoveDir, - .deliver = afs_deliver_fs_status_and_vol, + .deliver = afs_deliver_fs_dir_status_and_vol, .destructor = afs_flat_call_destructor, }; @@ -886,7 +721,7 @@ static const struct afs_call_type afs_RXFSRemoveDir = { * remove a file or directory */ int afs_fs_remove(struct afs_fs_cursor *fc, struct afs_vnode *vnode, - const char *name, bool isdir, u64 current_data_version) + const char *name, bool isdir, struct afs_status_cb *dvnode_scb) { struct afs_vnode *dvnode = fc->vnode; struct afs_call *call; @@ -895,7 +730,7 @@ int afs_fs_remove(struct afs_fs_cursor *fc, struct afs_vnode *vnode, __be32 *bp; if (test_bit(AFS_SERVER_FL_IS_YFS, &fc->cbi->server->flags)) - return yfs_fs_remove(fc, vnode, name, isdir, current_data_version); + return yfs_fs_remove(fc, vnode, name, isdir, dvnode_scb); _enter(""); @@ -910,9 +745,7 @@ int afs_fs_remove(struct afs_fs_cursor *fc, struct afs_vnode *vnode, return -ENOMEM; call->key = fc->key; - call->reply[0] = dvnode; - call->reply[1] = vnode; - call->expected_version = current_data_version + 1; + call->out_dir_scb = dvnode_scb; /* marshall the parameters */ bp = call->request; @@ -930,6 +763,7 @@ int afs_fs_remove(struct afs_fs_cursor *fc, struct afs_vnode *vnode, afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_call1(call, &dvnode->fid, name); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } @@ -939,7 +773,6 @@ int afs_fs_remove(struct afs_fs_cursor *fc, struct afs_vnode *vnode, */ static int afs_deliver_fs_link(struct afs_call *call) { - struct afs_vnode *dvnode = call->reply[0], *vnode = call->reply[1]; const __be32 *bp; int ret; @@ -951,14 +784,13 @@ static int afs_deliver_fs_link(struct afs_call *call) /* unmarshall the reply once we've received all of it */ bp = call->buffer; - ret = afs_decode_status(call, &bp, &vnode->status, vnode, NULL, NULL); + ret = xdr_decode_AFSFetchStatus(&bp, call, call->out_scb); if (ret < 0) return ret; - ret = afs_decode_status(call, &bp, &dvnode->status, dvnode, - &call->expected_version, NULL); + ret = xdr_decode_AFSFetchStatus(&bp, call, call->out_dir_scb); if (ret < 0) return ret; - /* xdr_decode_AFSVolSync(&bp, call->reply[X]); */ + xdr_decode_AFSVolSync(&bp, call->out_volsync); _leave(" = 0 [done]"); return 0; @@ -978,7 +810,9 @@ static const struct afs_call_type afs_RXFSLink = { * make a hard link */ int afs_fs_link(struct afs_fs_cursor *fc, struct afs_vnode *vnode, - const char *name, u64 current_data_version) + const char *name, + struct afs_status_cb *dvnode_scb, + struct afs_status_cb *vnode_scb) { struct afs_vnode *dvnode = fc->vnode; struct afs_call *call; @@ -987,7 +821,7 @@ int afs_fs_link(struct afs_fs_cursor *fc, struct afs_vnode *vnode, __be32 *bp; if (test_bit(AFS_SERVER_FL_IS_YFS, &fc->cbi->server->flags)) - return yfs_fs_link(fc, vnode, name, current_data_version); + return yfs_fs_link(fc, vnode, name, dvnode_scb, vnode_scb); _enter(""); @@ -1000,9 +834,8 @@ int afs_fs_link(struct afs_fs_cursor *fc, struct afs_vnode *vnode, return -ENOMEM; call->key = fc->key; - call->reply[0] = dvnode; - call->reply[1] = vnode; - call->expected_version = current_data_version + 1; + call->out_dir_scb = dvnode_scb; + call->out_scb = vnode_scb; /* marshall the parameters */ bp = call->request; @@ -1023,6 +856,7 @@ int afs_fs_link(struct afs_fs_cursor *fc, struct afs_vnode *vnode, afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_call1(call, &vnode->fid, name); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } @@ -1032,7 +866,6 @@ int afs_fs_link(struct afs_fs_cursor *fc, struct afs_vnode *vnode, */ static int afs_deliver_fs_symlink(struct afs_call *call) { - struct afs_vnode *vnode = call->reply[0]; const __be32 *bp; int ret; @@ -1044,15 +877,14 @@ static int afs_deliver_fs_symlink(struct afs_call *call) /* unmarshall the reply once we've received all of it */ bp = call->buffer; - xdr_decode_AFSFid(&bp, call->reply[1]); - ret = afs_decode_status(call, &bp, call->reply[2], NULL, NULL, NULL); + xdr_decode_AFSFid(&bp, call->out_fid); + ret = xdr_decode_AFSFetchStatus(&bp, call, call->out_scb); if (ret < 0) return ret; - ret = afs_decode_status(call, &bp, &vnode->status, vnode, - &call->expected_version, NULL); + ret = xdr_decode_AFSFetchStatus(&bp, call, call->out_dir_scb); if (ret < 0) return ret; - /* xdr_decode_AFSVolSync(&bp, call->reply[X]); */ + xdr_decode_AFSVolSync(&bp, call->out_volsync); _leave(" = 0 [done]"); return 0; @@ -1074,19 +906,19 @@ static const struct afs_call_type afs_RXFSSymlink = { int afs_fs_symlink(struct afs_fs_cursor *fc, const char *name, const char *contents, - u64 current_data_version, + struct afs_status_cb *dvnode_scb, struct afs_fid *newfid, - struct afs_file_status *newstatus) + struct afs_status_cb *new_scb) { - struct afs_vnode *vnode = fc->vnode; + struct afs_vnode *dvnode = fc->vnode; struct afs_call *call; - struct afs_net *net = afs_v2net(vnode); + struct afs_net *net = afs_v2net(dvnode); size_t namesz, reqsz, padsz, c_namesz, c_padsz; __be32 *bp; if (test_bit(AFS_SERVER_FL_IS_YFS, &fc->cbi->server->flags)) - return yfs_fs_symlink(fc, name, contents, current_data_version, - newfid, newstatus); + return yfs_fs_symlink(fc, name, contents, dvnode_scb, + newfid, new_scb); _enter(""); @@ -1104,17 +936,16 @@ int afs_fs_symlink(struct afs_fs_cursor *fc, return -ENOMEM; call->key = fc->key; - call->reply[0] = vnode; - call->reply[1] = newfid; - call->reply[2] = newstatus; - call->expected_version = current_data_version + 1; + call->out_dir_scb = dvnode_scb; + call->out_fid = newfid; + call->out_scb = new_scb; /* marshall the parameters */ bp = call->request; *bp++ = htonl(FSSYMLINK); - *bp++ = htonl(vnode->fid.vid); - *bp++ = htonl(vnode->fid.vnode); - *bp++ = htonl(vnode->fid.unique); + *bp++ = htonl(dvnode->fid.vid); + *bp++ = htonl(dvnode->fid.vnode); + *bp++ = htonl(dvnode->fid.unique); *bp++ = htonl(namesz); memcpy(bp, name, namesz); bp = (void *) bp + namesz; @@ -1130,14 +961,15 @@ int afs_fs_symlink(struct afs_fs_cursor *fc, bp = (void *) bp + c_padsz; } *bp++ = htonl(AFS_SET_MODE | AFS_SET_MTIME); - *bp++ = htonl(vnode->vfs_inode.i_mtime.tv_sec); /* mtime */ + *bp++ = htonl(dvnode->vfs_inode.i_mtime.tv_sec); /* mtime */ *bp++ = 0; /* owner */ *bp++ = 0; /* group */ *bp++ = htonl(S_IRWXUGO); /* unix mode */ *bp++ = 0; /* segment size */ afs_use_fs_server(call, fc->cbi); - trace_afs_make_fs_call1(call, &vnode->fid, name); + trace_afs_make_fs_call1(call, &dvnode->fid, name); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } @@ -1147,29 +979,24 @@ int afs_fs_symlink(struct afs_fs_cursor *fc, */ static int afs_deliver_fs_rename(struct afs_call *call) { - struct afs_vnode *orig_dvnode = call->reply[0], *new_dvnode = call->reply[1]; const __be32 *bp; int ret; - _enter("{%u}", call->unmarshall); - ret = afs_transfer_reply(call); if (ret < 0) return ret; /* unmarshall the reply once we've received all of it */ bp = call->buffer; - ret = afs_decode_status(call, &bp, &orig_dvnode->status, orig_dvnode, - &call->expected_version, NULL); + ret = xdr_decode_AFSFetchStatus(&bp, call, call->out_dir_scb); if (ret < 0) return ret; - if (new_dvnode != orig_dvnode) { - ret = afs_decode_status(call, &bp, &new_dvnode->status, new_dvnode, - &call->expected_version_2, NULL); + if (call->out_dir_scb != call->out_scb) { + ret = xdr_decode_AFSFetchStatus(&bp, call, call->out_scb); if (ret < 0) return ret; } - /* xdr_decode_AFSVolSync(&bp, call->reply[X]); */ + xdr_decode_AFSVolSync(&bp, call->out_volsync); _leave(" = 0 [done]"); return 0; @@ -1186,14 +1013,14 @@ static const struct afs_call_type afs_RXFSRename = { }; /* - * create a symbolic link + * Rename/move a file or directory. */ int afs_fs_rename(struct afs_fs_cursor *fc, const char *orig_name, struct afs_vnode *new_dvnode, const char *new_name, - u64 current_orig_data_version, - u64 current_new_data_version) + struct afs_status_cb *orig_dvnode_scb, + struct afs_status_cb *new_dvnode_scb) { struct afs_vnode *orig_dvnode = fc->vnode; struct afs_call *call; @@ -1204,8 +1031,8 @@ int afs_fs_rename(struct afs_fs_cursor *fc, if (test_bit(AFS_SERVER_FL_IS_YFS, &fc->cbi->server->flags)) return yfs_fs_rename(fc, orig_name, new_dvnode, new_name, - current_orig_data_version, - current_new_data_version); + orig_dvnode_scb, + new_dvnode_scb); _enter(""); @@ -1225,10 +1052,8 @@ int afs_fs_rename(struct afs_fs_cursor *fc, return -ENOMEM; call->key = fc->key; - call->reply[0] = orig_dvnode; - call->reply[1] = new_dvnode; - call->expected_version = current_orig_data_version + 1; - call->expected_version_2 = current_new_data_version + 1; + call->out_dir_scb = orig_dvnode_scb; + call->out_scb = new_dvnode_scb; /* marshall the parameters */ bp = call->request; @@ -1257,6 +1082,7 @@ int afs_fs_rename(struct afs_fs_cursor *fc, afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_call2(call, &orig_dvnode->fid, orig_name, new_name); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } @@ -1266,7 +1092,6 @@ int afs_fs_rename(struct afs_fs_cursor *fc, */ static int afs_deliver_fs_store_data(struct afs_call *call) { - struct afs_vnode *vnode = call->reply[0]; const __be32 *bp; int ret; @@ -1278,13 +1103,10 @@ static int afs_deliver_fs_store_data(struct afs_call *call) /* unmarshall the reply once we've received all of it */ bp = call->buffer; - ret = afs_decode_status(call, &bp, &vnode->status, vnode, - &call->expected_version, NULL); + ret = xdr_decode_AFSFetchStatus(&bp, call, call->out_scb); if (ret < 0) return ret; - /* xdr_decode_AFSVolSync(&bp, call->reply[X]); */ - - afs_pages_written_back(vnode, call); + xdr_decode_AFSVolSync(&bp, call->out_volsync); _leave(" = 0 [done]"); return 0; @@ -1314,7 +1136,8 @@ static int afs_fs_store_data64(struct afs_fs_cursor *fc, struct address_space *mapping, pgoff_t first, pgoff_t last, unsigned offset, unsigned to, - loff_t size, loff_t pos, loff_t i_size) + loff_t size, loff_t pos, loff_t i_size, + struct afs_status_cb *scb) { struct afs_vnode *vnode = fc->vnode; struct afs_call *call; @@ -1332,13 +1155,12 @@ static int afs_fs_store_data64(struct afs_fs_cursor *fc, call->key = fc->key; call->mapping = mapping; - call->reply[0] = vnode; call->first = first; call->last = last; call->first_offset = offset; call->last_to = to; call->send_pages = true; - call->expected_version = vnode->status.data_version + 1; + call->out_scb = scb; /* marshall the parameters */ bp = call->request; @@ -1362,6 +1184,7 @@ static int afs_fs_store_data64(struct afs_fs_cursor *fc, *bp++ = htonl((u32) i_size); trace_afs_make_fs_call(call, &vnode->fid); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } @@ -1371,7 +1194,8 @@ static int afs_fs_store_data64(struct afs_fs_cursor *fc, */ int afs_fs_store_data(struct afs_fs_cursor *fc, struct address_space *mapping, pgoff_t first, pgoff_t last, - unsigned offset, unsigned to) + unsigned offset, unsigned to, + struct afs_status_cb *scb) { struct afs_vnode *vnode = fc->vnode; struct afs_call *call; @@ -1380,7 +1204,7 @@ int afs_fs_store_data(struct afs_fs_cursor *fc, struct address_space *mapping, __be32 *bp; if (test_bit(AFS_SERVER_FL_IS_YFS, &fc->cbi->server->flags)) - return yfs_fs_store_data(fc, mapping, first, last, offset, to); + return yfs_fs_store_data(fc, mapping, first, last, offset, to, scb); _enter(",%x,{%llx:%llu},,", key_serial(fc->key), vnode->fid.vid, vnode->fid.vnode); @@ -1401,7 +1225,7 @@ int afs_fs_store_data(struct afs_fs_cursor *fc, struct address_space *mapping, if (pos >> 32 || i_size >> 32 || size >> 32 || (pos + size) >> 32) return afs_fs_store_data64(fc, mapping, first, last, offset, to, - size, pos, i_size); + size, pos, i_size, scb); call = afs_alloc_flat_call(net, &afs_RXFSStoreData, (4 + 6 + 3) * 4, @@ -1411,13 +1235,12 @@ int afs_fs_store_data(struct afs_fs_cursor *fc, struct address_space *mapping, call->key = fc->key; call->mapping = mapping; - call->reply[0] = vnode; call->first = first; call->last = last; call->first_offset = offset; call->last_to = to; call->send_pages = true; - call->expected_version = vnode->status.data_version + 1; + call->out_scb = scb; /* marshall the parameters */ bp = call->request; @@ -1439,6 +1262,7 @@ int afs_fs_store_data(struct afs_fs_cursor *fc, struct address_space *mapping, afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_call(call, &vnode->fid); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } @@ -1448,7 +1272,6 @@ int afs_fs_store_data(struct afs_fs_cursor *fc, struct address_space *mapping, */ static int afs_deliver_fs_store_status(struct afs_call *call) { - struct afs_vnode *vnode = call->reply[0]; const __be32 *bp; int ret; @@ -1460,11 +1283,10 @@ static int afs_deliver_fs_store_status(struct afs_call *call) /* unmarshall the reply once we've received all of it */ bp = call->buffer; - ret = afs_decode_status(call, &bp, &vnode->status, vnode, - &call->expected_version, NULL); + ret = xdr_decode_AFSFetchStatus(&bp, call, call->out_scb); if (ret < 0) return ret; - /* xdr_decode_AFSVolSync(&bp, call->reply[X]); */ + xdr_decode_AFSVolSync(&bp, call->out_volsync); _leave(" = 0 [done]"); return 0; @@ -1498,7 +1320,8 @@ static const struct afs_call_type afs_RXFSStoreData64_as_Status = { * set the attributes on a very large file, using FS.StoreData rather than * FS.StoreStatus so as to alter the file size also */ -static int afs_fs_setattr_size64(struct afs_fs_cursor *fc, struct iattr *attr) +static int afs_fs_setattr_size64(struct afs_fs_cursor *fc, struct iattr *attr, + struct afs_status_cb *scb) { struct afs_vnode *vnode = fc->vnode; struct afs_call *call; @@ -1517,8 +1340,7 @@ static int afs_fs_setattr_size64(struct afs_fs_cursor *fc, struct iattr *attr) return -ENOMEM; call->key = fc->key; - call->reply[0] = vnode; - call->expected_version = vnode->status.data_version + 1; + call->out_scb = scb; /* marshall the parameters */ bp = call->request; @@ -1538,6 +1360,7 @@ static int afs_fs_setattr_size64(struct afs_fs_cursor *fc, struct iattr *attr) afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_call(call, &vnode->fid); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } @@ -1546,7 +1369,8 @@ static int afs_fs_setattr_size64(struct afs_fs_cursor *fc, struct iattr *attr) * set the attributes on a file, using FS.StoreData rather than FS.StoreStatus * so as to alter the file size also */ -static int afs_fs_setattr_size(struct afs_fs_cursor *fc, struct iattr *attr) +static int afs_fs_setattr_size(struct afs_fs_cursor *fc, struct iattr *attr, + struct afs_status_cb *scb) { struct afs_vnode *vnode = fc->vnode; struct afs_call *call; @@ -1558,7 +1382,7 @@ static int afs_fs_setattr_size(struct afs_fs_cursor *fc, struct iattr *attr) ASSERT(attr->ia_valid & ATTR_SIZE); if (attr->ia_size >> 32) - return afs_fs_setattr_size64(fc, attr); + return afs_fs_setattr_size64(fc, attr, scb); call = afs_alloc_flat_call(net, &afs_RXFSStoreData_as_Status, (4 + 6 + 3) * 4, @@ -1567,8 +1391,7 @@ static int afs_fs_setattr_size(struct afs_fs_cursor *fc, struct iattr *attr) return -ENOMEM; call->key = fc->key; - call->reply[0] = vnode; - call->expected_version = vnode->status.data_version + 1; + call->out_scb = scb; /* marshall the parameters */ bp = call->request; @@ -1585,6 +1408,7 @@ static int afs_fs_setattr_size(struct afs_fs_cursor *fc, struct iattr *attr) afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_call(call, &vnode->fid); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } @@ -1593,7 +1417,8 @@ static int afs_fs_setattr_size(struct afs_fs_cursor *fc, struct iattr *attr) * set the attributes on a file, using FS.StoreData if there's a change in file * size, and FS.StoreStatus otherwise */ -int afs_fs_setattr(struct afs_fs_cursor *fc, struct iattr *attr) +int afs_fs_setattr(struct afs_fs_cursor *fc, struct iattr *attr, + struct afs_status_cb *scb) { struct afs_vnode *vnode = fc->vnode; struct afs_call *call; @@ -1601,10 +1426,10 @@ int afs_fs_setattr(struct afs_fs_cursor *fc, struct iattr *attr) __be32 *bp; if (test_bit(AFS_SERVER_FL_IS_YFS, &fc->cbi->server->flags)) - return yfs_fs_setattr(fc, attr); + return yfs_fs_setattr(fc, attr, scb); if (attr->ia_valid & ATTR_SIZE) - return afs_fs_setattr_size(fc, attr); + return afs_fs_setattr_size(fc, attr, scb); _enter(",%x,{%llx:%llu},,", key_serial(fc->key), vnode->fid.vid, vnode->fid.vnode); @@ -1616,8 +1441,7 @@ int afs_fs_setattr(struct afs_fs_cursor *fc, struct iattr *attr) return -ENOMEM; call->key = fc->key; - call->reply[0] = vnode; - call->expected_version = vnode->status.data_version; + call->out_scb = scb; /* marshall the parameters */ bp = call->request; @@ -1630,6 +1454,7 @@ int afs_fs_setattr(struct afs_fs_cursor *fc, struct iattr *attr) afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_call(call, &vnode->fid); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } @@ -1659,7 +1484,7 @@ static int afs_deliver_fs_get_volume_status(struct afs_call *call) return ret; bp = call->buffer; - xdr_decode_AFSFetchVolumeStatus(&bp, call->reply[1]); + xdr_decode_AFSFetchVolumeStatus(&bp, call->out_volstatus); call->unmarshall++; afs_extract_to_tmp(call); @@ -1675,7 +1500,7 @@ static int afs_deliver_fs_get_volume_status(struct afs_call *call) return afs_protocol_error(call, -EBADMSG, afs_eproto_volname_len); size = (call->count + 3) & ~3; /* It's padded */ - afs_extract_begin(call, call->reply[2], size); + afs_extract_to_buf(call, size); call->unmarshall++; /* Fall through - and extract the volume name */ @@ -1685,7 +1510,7 @@ static int afs_deliver_fs_get_volume_status(struct afs_call *call) if (ret < 0) return ret; - p = call->reply[2]; + p = call->buffer; p[call->count] = 0; _debug("volname '%s'", p); afs_extract_to_tmp(call); @@ -1703,7 +1528,7 @@ static int afs_deliver_fs_get_volume_status(struct afs_call *call) return afs_protocol_error(call, -EBADMSG, afs_eproto_offline_msg_len); size = (call->count + 3) & ~3; /* It's padded */ - afs_extract_begin(call, call->reply[2], size); + afs_extract_to_buf(call, size); call->unmarshall++; /* Fall through - and extract the offline message */ @@ -1713,7 +1538,7 @@ static int afs_deliver_fs_get_volume_status(struct afs_call *call) if (ret < 0) return ret; - p = call->reply[2]; + p = call->buffer; p[call->count] = 0; _debug("offline '%s'", p); @@ -1732,7 +1557,7 @@ static int afs_deliver_fs_get_volume_status(struct afs_call *call) return afs_protocol_error(call, -EBADMSG, afs_eproto_motd_len); size = (call->count + 3) & ~3; /* It's padded */ - afs_extract_begin(call, call->reply[2], size); + afs_extract_to_buf(call, size); call->unmarshall++; /* Fall through - and extract the message of the day */ @@ -1742,7 +1567,7 @@ static int afs_deliver_fs_get_volume_status(struct afs_call *call) if (ret < 0) return ret; - p = call->reply[2]; + p = call->buffer; p[call->count] = 0; _debug("motd '%s'", p); @@ -1757,23 +1582,13 @@ static int afs_deliver_fs_get_volume_status(struct afs_call *call) } /* - * destroy an FS.GetVolumeStatus call - */ -static void afs_get_volume_status_call_destructor(struct afs_call *call) -{ - kfree(call->reply[2]); - call->reply[2] = NULL; - afs_flat_call_destructor(call); -} - -/* * FS.GetVolumeStatus operation type */ static const struct afs_call_type afs_RXFSGetVolumeStatus = { .name = "FS.GetVolumeStatus", .op = afs_FS_GetVolumeStatus, .deliver = afs_deliver_fs_get_volume_status, - .destructor = afs_get_volume_status_call_destructor, + .destructor = afs_flat_call_destructor, }; /* @@ -1786,27 +1601,19 @@ int afs_fs_get_volume_status(struct afs_fs_cursor *fc, struct afs_call *call; struct afs_net *net = afs_v2net(vnode); __be32 *bp; - void *tmpbuf; if (test_bit(AFS_SERVER_FL_IS_YFS, &fc->cbi->server->flags)) return yfs_fs_get_volume_status(fc, vs); _enter(""); - tmpbuf = kmalloc(AFSOPAQUEMAX, GFP_KERNEL); - if (!tmpbuf) - return -ENOMEM; - - call = afs_alloc_flat_call(net, &afs_RXFSGetVolumeStatus, 2 * 4, 12 * 4); - if (!call) { - kfree(tmpbuf); + call = afs_alloc_flat_call(net, &afs_RXFSGetVolumeStatus, 2 * 4, + max(12 * 4, AFSOPAQUEMAX + 1)); + if (!call) return -ENOMEM; - } call->key = fc->key; - call->reply[0] = vnode; - call->reply[1] = vs; - call->reply[2] = tmpbuf; + call->out_volstatus = vs; /* marshall the parameters */ bp = call->request; @@ -1815,6 +1622,7 @@ int afs_fs_get_volume_status(struct afs_fs_cursor *fc, afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_call(call, &vnode->fid); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } @@ -1835,7 +1643,7 @@ static int afs_deliver_fs_xxxx_lock(struct afs_call *call) /* unmarshall the reply once we've received all of it */ bp = call->buffer; - /* xdr_decode_AFSVolSync(&bp, call->reply[X]); */ + xdr_decode_AFSVolSync(&bp, call->out_volsync); _leave(" = 0 [done]"); return 0; @@ -1876,7 +1684,8 @@ static const struct afs_call_type afs_RXFSReleaseLock = { /* * Set a lock on a file */ -int afs_fs_set_lock(struct afs_fs_cursor *fc, afs_lock_type_t type) +int afs_fs_set_lock(struct afs_fs_cursor *fc, afs_lock_type_t type, + struct afs_status_cb *scb) { struct afs_vnode *vnode = fc->vnode; struct afs_call *call; @@ -1884,7 +1693,7 @@ int afs_fs_set_lock(struct afs_fs_cursor *fc, afs_lock_type_t type) __be32 *bp; if (test_bit(AFS_SERVER_FL_IS_YFS, &fc->cbi->server->flags)) - return yfs_fs_set_lock(fc, type); + return yfs_fs_set_lock(fc, type, scb); _enter(""); @@ -1893,8 +1702,8 @@ int afs_fs_set_lock(struct afs_fs_cursor *fc, afs_lock_type_t type) return -ENOMEM; call->key = fc->key; - call->reply[0] = vnode; - call->want_reply_time = true; + call->lvnode = vnode; + call->out_scb = scb; /* marshall the parameters */ bp = call->request; @@ -1906,6 +1715,7 @@ int afs_fs_set_lock(struct afs_fs_cursor *fc, afs_lock_type_t type) afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_calli(call, &vnode->fid, type); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } @@ -1913,7 +1723,7 @@ int afs_fs_set_lock(struct afs_fs_cursor *fc, afs_lock_type_t type) /* * extend a lock on a file */ -int afs_fs_extend_lock(struct afs_fs_cursor *fc) +int afs_fs_extend_lock(struct afs_fs_cursor *fc, struct afs_status_cb *scb) { struct afs_vnode *vnode = fc->vnode; struct afs_call *call; @@ -1921,7 +1731,7 @@ int afs_fs_extend_lock(struct afs_fs_cursor *fc) __be32 *bp; if (test_bit(AFS_SERVER_FL_IS_YFS, &fc->cbi->server->flags)) - return yfs_fs_extend_lock(fc); + return yfs_fs_extend_lock(fc, scb); _enter(""); @@ -1930,8 +1740,8 @@ int afs_fs_extend_lock(struct afs_fs_cursor *fc) return -ENOMEM; call->key = fc->key; - call->reply[0] = vnode; - call->want_reply_time = true; + call->lvnode = vnode; + call->out_scb = scb; /* marshall the parameters */ bp = call->request; @@ -1942,6 +1752,7 @@ int afs_fs_extend_lock(struct afs_fs_cursor *fc) afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_call(call, &vnode->fid); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } @@ -1949,7 +1760,7 @@ int afs_fs_extend_lock(struct afs_fs_cursor *fc) /* * release a lock on a file */ -int afs_fs_release_lock(struct afs_fs_cursor *fc) +int afs_fs_release_lock(struct afs_fs_cursor *fc, struct afs_status_cb *scb) { struct afs_vnode *vnode = fc->vnode; struct afs_call *call; @@ -1957,7 +1768,7 @@ int afs_fs_release_lock(struct afs_fs_cursor *fc) __be32 *bp; if (test_bit(AFS_SERVER_FL_IS_YFS, &fc->cbi->server->flags)) - return yfs_fs_release_lock(fc); + return yfs_fs_release_lock(fc, scb); _enter(""); @@ -1966,7 +1777,8 @@ int afs_fs_release_lock(struct afs_fs_cursor *fc) return -ENOMEM; call->key = fc->key; - call->reply[0] = vnode; + call->lvnode = vnode; + call->out_scb = scb; /* marshall the parameters */ bp = call->request; @@ -1977,6 +1789,7 @@ int afs_fs_release_lock(struct afs_fs_cursor *fc) afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_call(call, &vnode->fid); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } @@ -2071,14 +1884,6 @@ static int afs_deliver_fs_get_capabilities(struct afs_call *call) return 0; } -static void afs_destroy_fs_get_capabilities(struct afs_call *call) -{ - struct afs_server *server = call->reply[0]; - - afs_put_server(call->net, server); - afs_flat_call_destructor(call); -} - /* * FS.GetCapabilities operation type */ @@ -2087,7 +1892,7 @@ static const struct afs_call_type afs_RXFSGetCapabilities = { .op = afs_FS_GetCapabilities, .deliver = afs_deliver_fs_get_capabilities, .done = afs_fileserver_probe_result, - .destructor = afs_destroy_fs_get_capabilities, + .destructor = afs_flat_call_destructor, }; /* @@ -2110,11 +1915,11 @@ struct afs_call *afs_fs_get_capabilities(struct afs_net *net, return ERR_PTR(-ENOMEM); call->key = key; - call->reply[0] = afs_get_server(server); - call->reply[1] = (void *)(long)server_index; + call->server = afs_get_server(server); + call->server_index = server_index; call->upgrade = true; - call->want_reply_time = true; call->async = true; + call->max_lifespan = AFS_PROBE_MAX_LIFESPAN; /* marshall the parameters */ bp = call->request; @@ -2131,10 +1936,6 @@ struct afs_call *afs_fs_get_capabilities(struct afs_net *net, */ static int afs_deliver_fs_fetch_status(struct afs_call *call) { - struct afs_file_status *status = call->reply[1]; - struct afs_callback *callback = call->reply[2]; - struct afs_volsync *volsync = call->reply[3]; - struct afs_fid *fid = call->reply[0]; const __be32 *bp; int ret; @@ -2142,16 +1943,13 @@ static int afs_deliver_fs_fetch_status(struct afs_call *call) if (ret < 0) return ret; - _enter("{%llx:%llu}", fid->vid, fid->vnode); - /* unmarshall the reply once we've received all of it */ bp = call->buffer; - ret = afs_decode_status(call, &bp, status, NULL, - &call->expected_version, NULL); + ret = xdr_decode_AFSFetchStatus(&bp, call, call->out_scb); if (ret < 0) return ret; - xdr_decode_AFSCallBack_raw(call, &bp, callback); - xdr_decode_AFSVolSync(&bp, volsync); + xdr_decode_AFSCallBack(&bp, call, call->out_scb); + xdr_decode_AFSVolSync(&bp, call->out_volsync); _leave(" = 0 [done]"); return 0; @@ -2173,15 +1971,14 @@ static const struct afs_call_type afs_RXFSFetchStatus = { int afs_fs_fetch_status(struct afs_fs_cursor *fc, struct afs_net *net, struct afs_fid *fid, - struct afs_file_status *status, - struct afs_callback *callback, + struct afs_status_cb *scb, struct afs_volsync *volsync) { struct afs_call *call; __be32 *bp; if (test_bit(AFS_SERVER_FL_IS_YFS, &fc->cbi->server->flags)) - return yfs_fs_fetch_status(fc, net, fid, status, callback, volsync); + return yfs_fs_fetch_status(fc, net, fid, scb, volsync); _enter(",%x,{%llx:%llu},,", key_serial(fc->key), fid->vid, fid->vnode); @@ -2193,12 +1990,9 @@ int afs_fs_fetch_status(struct afs_fs_cursor *fc, } call->key = fc->key; - call->reply[0] = fid; - call->reply[1] = status; - call->reply[2] = callback; - call->reply[3] = volsync; - call->expected_version = 1; /* vnode->status.data_version */ - call->want_reply_time = true; + call->out_fid = fid; + call->out_scb = scb; + call->out_volsync = volsync; /* marshall the parameters */ bp = call->request; @@ -2207,9 +2001,9 @@ int afs_fs_fetch_status(struct afs_fs_cursor *fc, bp[2] = htonl(fid->vnode); bp[3] = htonl(fid->unique); - call->cb_break = fc->cb_break; afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_call(call, fid); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } @@ -2219,9 +2013,7 @@ int afs_fs_fetch_status(struct afs_fs_cursor *fc, */ static int afs_deliver_fs_inline_bulk_status(struct afs_call *call) { - struct afs_file_status *statuses; - struct afs_callback *callbacks; - struct afs_vnode *vnode = call->reply[0]; + struct afs_status_cb *scb; const __be32 *bp; u32 tmp; int ret; @@ -2260,10 +2052,8 @@ static int afs_deliver_fs_inline_bulk_status(struct afs_call *call) return ret; bp = call->buffer; - statuses = call->reply[1]; - ret = afs_decode_status(call, &bp, &statuses[call->count], - call->count == 0 ? vnode : NULL, - NULL, NULL); + scb = &call->out_scb[call->count]; + ret = xdr_decode_AFSFetchStatus(&bp, call, scb); if (ret < 0) return ret; @@ -2302,13 +2092,8 @@ static int afs_deliver_fs_inline_bulk_status(struct afs_call *call) _debug("unmarshall CB array"); bp = call->buffer; - callbacks = call->reply[2]; - callbacks[call->count].version = ntohl(bp[0]); - callbacks[call->count].expires_at = xdr_decode_expiry(call, ntohl(bp[1])); - callbacks[call->count].type = ntohl(bp[2]); - statuses = call->reply[1]; - if (call->count == 0 && vnode && statuses[0].abort_code == 0) - xdr_decode_AFSCallBack(call, vnode, &bp); + scb = &call->out_scb[call->count]; + xdr_decode_AFSCallBack(&bp, call, scb); call->count++; if (call->count < call->count2) goto more_cbs; @@ -2323,7 +2108,7 @@ static int afs_deliver_fs_inline_bulk_status(struct afs_call *call) return ret; bp = call->buffer; - xdr_decode_AFSVolSync(&bp, call->reply[3]); + xdr_decode_AFSVolSync(&bp, call->out_volsync); call->unmarshall++; @@ -2351,8 +2136,7 @@ static const struct afs_call_type afs_RXFSInlineBulkStatus = { int afs_fs_inline_bulk_status(struct afs_fs_cursor *fc, struct afs_net *net, struct afs_fid *fids, - struct afs_file_status *statuses, - struct afs_callback *callbacks, + struct afs_status_cb *statuses, unsigned int nr_fids, struct afs_volsync *volsync) { @@ -2361,7 +2145,7 @@ int afs_fs_inline_bulk_status(struct afs_fs_cursor *fc, int i; if (test_bit(AFS_SERVER_FL_IS_YFS, &fc->cbi->server->flags)) - return yfs_fs_inline_bulk_status(fc, net, fids, statuses, callbacks, + return yfs_fs_inline_bulk_status(fc, net, fids, statuses, nr_fids, volsync); _enter(",%x,{%llx:%llu},%u", @@ -2376,12 +2160,9 @@ int afs_fs_inline_bulk_status(struct afs_fs_cursor *fc, } call->key = fc->key; - call->reply[0] = NULL; /* vnode for fid[0] */ - call->reply[1] = statuses; - call->reply[2] = callbacks; - call->reply[3] = volsync; + call->out_scb = statuses; + call->out_volsync = volsync; call->count2 = nr_fids; - call->want_reply_time = true; /* marshall the parameters */ bp = call->request; @@ -2393,9 +2174,9 @@ int afs_fs_inline_bulk_status(struct afs_fs_cursor *fc, *bp++ = htonl(fids[i].unique); } - call->cb_break = fc->cb_break; afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_call(call, &fids[0]); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } @@ -2405,7 +2186,6 @@ int afs_fs_inline_bulk_status(struct afs_fs_cursor *fc, */ static int afs_deliver_fs_fetch_acl(struct afs_call *call) { - struct afs_vnode *vnode = call->reply[1]; struct afs_acl *acl; const __be32 *bp; unsigned int size; @@ -2430,7 +2210,7 @@ static int afs_deliver_fs_fetch_acl(struct afs_call *call) acl = kmalloc(struct_size(acl, data, size), GFP_KERNEL); if (!acl) return -ENOMEM; - call->reply[0] = acl; + call->ret_acl = acl; acl->size = call->count2; afs_extract_begin(call, acl->data, size); call->unmarshall++; @@ -2451,11 +2231,10 @@ static int afs_deliver_fs_fetch_acl(struct afs_call *call) return ret; bp = call->buffer; - ret = afs_decode_status(call, &bp, &vnode->status, vnode, - &vnode->status.data_version, NULL); + ret = xdr_decode_AFSFetchStatus(&bp, call, call->out_scb); if (ret < 0) return ret; - xdr_decode_AFSVolSync(&bp, call->reply[2]); + xdr_decode_AFSVolSync(&bp, call->out_volsync); call->unmarshall++; @@ -2469,7 +2248,7 @@ static int afs_deliver_fs_fetch_acl(struct afs_call *call) static void afs_destroy_fs_fetch_acl(struct afs_call *call) { - kfree(call->reply[0]); + kfree(call->ret_acl); afs_flat_call_destructor(call); } @@ -2486,7 +2265,8 @@ static const struct afs_call_type afs_RXFSFetchACL = { /* * Fetch the ACL for a file. */ -struct afs_acl *afs_fs_fetch_acl(struct afs_fs_cursor *fc) +struct afs_acl *afs_fs_fetch_acl(struct afs_fs_cursor *fc, + struct afs_status_cb *scb) { struct afs_vnode *vnode = fc->vnode; struct afs_call *call; @@ -2503,10 +2283,9 @@ struct afs_acl *afs_fs_fetch_acl(struct afs_fs_cursor *fc) } call->key = fc->key; - call->reply[0] = NULL; - call->reply[1] = vnode; - call->reply[2] = NULL; /* volsync */ - call->ret_reply0 = true; + call->ret_acl = NULL; + call->out_scb = scb; + call->out_volsync = NULL; /* marshall the parameters */ bp = call->request; @@ -2515,7 +2294,6 @@ struct afs_acl *afs_fs_fetch_acl(struct afs_fs_cursor *fc) bp[2] = htonl(vnode->fid.vnode); bp[3] = htonl(vnode->fid.unique); - call->cb_break = fc->cb_break; afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_call(call, &vnode->fid); afs_make_call(&fc->ac, call, GFP_KERNEL); @@ -2523,19 +2301,43 @@ struct afs_acl *afs_fs_fetch_acl(struct afs_fs_cursor *fc) } /* + * Deliver reply data to any operation that returns file status and volume + * sync. + */ +static int afs_deliver_fs_file_status_and_vol(struct afs_call *call) +{ + const __be32 *bp; + int ret; + + ret = afs_transfer_reply(call); + if (ret < 0) + return ret; + + bp = call->buffer; + ret = xdr_decode_AFSFetchStatus(&bp, call, call->out_scb); + if (ret < 0) + return ret; + xdr_decode_AFSVolSync(&bp, call->out_volsync); + + _leave(" = 0 [done]"); + return 0; +} + +/* * FS.StoreACL operation type */ static const struct afs_call_type afs_RXFSStoreACL = { .name = "FS.StoreACL", .op = afs_FS_StoreACL, - .deliver = afs_deliver_fs_status_and_vol, + .deliver = afs_deliver_fs_file_status_and_vol, .destructor = afs_flat_call_destructor, }; /* * Fetch the ACL for a file. */ -int afs_fs_store_acl(struct afs_fs_cursor *fc, const struct afs_acl *acl) +int afs_fs_store_acl(struct afs_fs_cursor *fc, const struct afs_acl *acl, + struct afs_status_cb *scb) { struct afs_vnode *vnode = fc->vnode; struct afs_call *call; @@ -2555,8 +2357,8 @@ int afs_fs_store_acl(struct afs_fs_cursor *fc, const struct afs_acl *acl) } call->key = fc->key; - call->reply[0] = vnode; - call->reply[2] = NULL; /* volsync */ + call->out_scb = scb; + call->out_volsync = NULL; /* marshall the parameters */ bp = call->request; diff --git a/fs/afs/inode.c b/fs/afs/inode.c index c4652b42d545..b42d9d09669c 100644 --- a/fs/afs/inode.c +++ b/fs/afs/inode.c @@ -23,6 +23,7 @@ #include <linux/namei.h> #include <linux/iversion.h> #include "internal.h" +#include "afs_fs.h" static const struct inode_operations afs_symlink_inode_operations = { .get_link = page_get_link, @@ -58,38 +59,50 @@ static noinline void dump_vnode(struct afs_vnode *vnode, struct afs_vnode *paren * Initialise an inode from the vnode status. */ static int afs_inode_init_from_status(struct afs_vnode *vnode, struct key *key, - struct afs_vnode *parent_vnode) + struct afs_cb_interest *cbi, + struct afs_vnode *parent_vnode, + struct afs_status_cb *scb) { + struct afs_cb_interest *old_cbi = NULL; + struct afs_file_status *status = &scb->status; struct inode *inode = AFS_VNODE_TO_I(vnode); + struct timespec64 t; _debug("FS: ft=%d lk=%d sz=%llu ver=%Lu mod=%hu", - vnode->status.type, - vnode->status.nlink, - (unsigned long long) vnode->status.size, - vnode->status.data_version, - vnode->status.mode); + status->type, + status->nlink, + (unsigned long long) status->size, + status->data_version, + status->mode); - read_seqlock_excl(&vnode->cb_lock); + write_seqlock(&vnode->cb_lock); - afs_update_inode_from_status(vnode, &vnode->status, NULL, - AFS_VNODE_NOT_YET_SET); + vnode->status = *status; - switch (vnode->status.type) { + t = status->mtime_client; + inode->i_ctime = t; + inode->i_mtime = t; + inode->i_atime = t; + inode->i_uid = make_kuid(&init_user_ns, status->owner); + inode->i_gid = make_kgid(&init_user_ns, status->group); + set_nlink(&vnode->vfs_inode, status->nlink); + + switch (status->type) { case AFS_FTYPE_FILE: - inode->i_mode = S_IFREG | vnode->status.mode; + inode->i_mode = S_IFREG | status->mode; inode->i_op = &afs_file_inode_operations; inode->i_fop = &afs_file_operations; inode->i_mapping->a_ops = &afs_fs_aops; break; case AFS_FTYPE_DIR: - inode->i_mode = S_IFDIR | vnode->status.mode; + inode->i_mode = S_IFDIR | status->mode; inode->i_op = &afs_dir_inode_operations; inode->i_fop = &afs_dir_file_operations; inode->i_mapping->a_ops = &afs_dir_aops; break; case AFS_FTYPE_SYMLINK: /* Symlinks with a mode of 0644 are actually mountpoints. */ - if ((vnode->status.mode & 0777) == 0644) { + if ((status->mode & 0777) == 0644) { inode->i_flags |= S_AUTOMOUNT; set_bit(AFS_VNODE_MOUNTPOINT, &vnode->flags); @@ -99,7 +112,7 @@ static int afs_inode_init_from_status(struct afs_vnode *vnode, struct key *key, inode->i_fop = &afs_mntpt_file_operations; inode->i_mapping->a_ops = &afs_fs_aops; } else { - inode->i_mode = S_IFLNK | vnode->status.mode; + inode->i_mode = S_IFLNK | status->mode; inode->i_op = &afs_symlink_inode_operations; inode->i_mapping->a_ops = &afs_fs_aops; } @@ -107,7 +120,7 @@ static int afs_inode_init_from_status(struct afs_vnode *vnode, struct key *key, break; default: dump_vnode(vnode, parent_vnode); - read_sequnlock_excl(&vnode->cb_lock); + write_sequnlock(&vnode->cb_lock); return afs_protocol_error(NULL, -EBADMSG, afs_eproto_file_type); } @@ -116,17 +129,175 @@ static int afs_inode_init_from_status(struct afs_vnode *vnode, struct key *key, * for consistency with other AFS clients. */ inode->i_blocks = ((i_size_read(inode) + 1023) >> 10) << 1; - vnode->invalid_before = vnode->status.data_version; + i_size_write(&vnode->vfs_inode, status->size); + + vnode->invalid_before = status->data_version; + inode_set_iversion_raw(&vnode->vfs_inode, status->data_version); + + if (!scb->have_cb) { + /* it's a symlink we just created (the fileserver + * didn't give us a callback) */ + vnode->cb_expires_at = ktime_get_real_seconds(); + } else { + vnode->cb_expires_at = scb->callback.expires_at; + old_cbi = rcu_dereference_protected(vnode->cb_interest, + lockdep_is_held(&vnode->cb_lock.lock)); + if (cbi != old_cbi) + rcu_assign_pointer(vnode->cb_interest, afs_get_cb_interest(cbi)); + else + old_cbi = NULL; + set_bit(AFS_VNODE_CB_PROMISED, &vnode->flags); + } - read_sequnlock_excl(&vnode->cb_lock); + write_sequnlock(&vnode->cb_lock); + afs_put_cb_interest(afs_v2net(vnode), old_cbi); return 0; } /* + * Update the core inode struct from a returned status record. + */ +static void afs_apply_status(struct afs_fs_cursor *fc, + struct afs_vnode *vnode, + struct afs_status_cb *scb, + const afs_dataversion_t *expected_version) +{ + struct afs_file_status *status = &scb->status; + struct timespec64 t; + umode_t mode; + bool data_changed = false; + + BUG_ON(test_bit(AFS_VNODE_UNSET, &vnode->flags)); + + if (status->type != vnode->status.type) { + pr_warning("Vnode %llx:%llx:%x changed type %u to %u\n", + vnode->fid.vid, + vnode->fid.vnode, + vnode->fid.unique, + status->type, vnode->status.type); + afs_protocol_error(NULL, -EBADMSG, afs_eproto_bad_status); + return; + } + + if (status->nlink != vnode->status.nlink) + set_nlink(&vnode->vfs_inode, status->nlink); + + if (status->owner != vnode->status.owner) + vnode->vfs_inode.i_uid = make_kuid(&init_user_ns, status->owner); + + if (status->group != vnode->status.group) + vnode->vfs_inode.i_gid = make_kgid(&init_user_ns, status->group); + + if (status->mode != vnode->status.mode) { + mode = vnode->vfs_inode.i_mode; + mode &= ~S_IALLUGO; + mode |= status->mode; + WRITE_ONCE(vnode->vfs_inode.i_mode, mode); + } + + t = status->mtime_client; + vnode->vfs_inode.i_ctime = t; + vnode->vfs_inode.i_mtime = t; + vnode->vfs_inode.i_atime = t; + + if (vnode->status.data_version != status->data_version) + data_changed = true; + + vnode->status = *status; + + if (expected_version && + *expected_version != status->data_version) { + kdebug("vnode modified %llx on {%llx:%llu} [exp %llx] %s", + (unsigned long long) status->data_version, + vnode->fid.vid, vnode->fid.vnode, + (unsigned long long) *expected_version, + fc->type ? fc->type->name : "???"); + vnode->invalid_before = status->data_version; + if (vnode->status.type == AFS_FTYPE_DIR) { + if (test_and_clear_bit(AFS_VNODE_DIR_VALID, &vnode->flags)) + afs_stat_v(vnode, n_inval); + } else { + set_bit(AFS_VNODE_ZAP_DATA, &vnode->flags); + } + } else if (vnode->status.type == AFS_FTYPE_DIR) { + /* Expected directory change is handled elsewhere so + * that we can locally edit the directory and save on a + * download. + */ + if (test_bit(AFS_VNODE_DIR_VALID, &vnode->flags)) + data_changed = false; + } + + if (data_changed) { + inode_set_iversion_raw(&vnode->vfs_inode, status->data_version); + i_size_write(&vnode->vfs_inode, status->size); + } +} + +/* + * Apply a callback to a vnode. + */ +static void afs_apply_callback(struct afs_fs_cursor *fc, + struct afs_vnode *vnode, + struct afs_status_cb *scb, + unsigned int cb_break) +{ + struct afs_cb_interest *old; + struct afs_callback *cb = &scb->callback; + + if (!afs_cb_is_broken(cb_break, vnode, fc->cbi)) { + vnode->cb_expires_at = cb->expires_at; + old = rcu_dereference_protected(vnode->cb_interest, + lockdep_is_held(&vnode->cb_lock.lock)); + if (old != fc->cbi) { + rcu_assign_pointer(vnode->cb_interest, afs_get_cb_interest(fc->cbi)); + afs_put_cb_interest(afs_v2net(vnode), old); + } + set_bit(AFS_VNODE_CB_PROMISED, &vnode->flags); + } +} + +/* + * Apply the received status and callback to an inode all in the same critical + * section to avoid races with afs_validate(). + */ +void afs_vnode_commit_status(struct afs_fs_cursor *fc, + struct afs_vnode *vnode, + unsigned int cb_break, + const afs_dataversion_t *expected_version, + struct afs_status_cb *scb) +{ + if (fc->ac.error != 0) + return; + + write_seqlock(&vnode->cb_lock); + + if (scb->have_error) { + if (scb->status.abort_code == VNOVNODE) { + set_bit(AFS_VNODE_DELETED, &vnode->flags); + clear_nlink(&vnode->vfs_inode); + __afs_break_callback(vnode); + } + } else { + if (scb->have_status) + afs_apply_status(fc, vnode, scb, expected_version); + if (scb->have_cb) + afs_apply_callback(fc, vnode, scb, cb_break); + } + + write_sequnlock(&vnode->cb_lock); + + if (fc->ac.error == 0 && scb->have_status) + afs_cache_permit(vnode, fc->key, cb_break, scb); +} + +/* * Fetch file status from the volume. */ -int afs_fetch_status(struct afs_vnode *vnode, struct key *key, bool new_inode) +int afs_fetch_status(struct afs_vnode *vnode, struct key *key, bool is_new, + afs_access_t *_caller_access) { + struct afs_status_cb *scb; struct afs_fs_cursor fc; int ret; @@ -135,18 +306,38 @@ int afs_fetch_status(struct afs_vnode *vnode, struct key *key, bool new_inode) vnode->fid.vid, vnode->fid.vnode, vnode->fid.unique, vnode->flags); + scb = kzalloc(sizeof(struct afs_status_cb), GFP_KERNEL); + if (!scb) + return -ENOMEM; + ret = -ERESTARTSYS; - if (afs_begin_vnode_operation(&fc, vnode, key)) { + if (afs_begin_vnode_operation(&fc, vnode, key, true)) { + afs_dataversion_t data_version = vnode->status.data_version; + while (afs_select_fileserver(&fc)) { fc.cb_break = afs_calc_vnode_cb_break(vnode); - afs_fs_fetch_file_status(&fc, NULL, new_inode); + afs_fs_fetch_file_status(&fc, scb, NULL); } - afs_check_for_remote_deletion(&fc, fc.vnode); - afs_vnode_commit_status(&fc, vnode, fc.cb_break); + if (fc.error) { + /* Do nothing. */ + } else if (is_new) { + ret = afs_inode_init_from_status(vnode, key, fc.cbi, + NULL, scb); + fc.error = ret; + if (ret == 0) + afs_cache_permit(vnode, key, fc.cb_break, scb); + } else { + afs_vnode_commit_status(&fc, vnode, fc.cb_break, + &data_version, scb); + } + afs_check_for_remote_deletion(&fc, vnode); ret = afs_end_vnode_operation(&fc); } + if (ret == 0 && _caller_access) + *_caller_access = scb->status.caller_access; + kfree(scb); _leave(" = %d", ret); return ret; } @@ -156,10 +347,10 @@ int afs_fetch_status(struct afs_vnode *vnode, struct key *key, bool new_inode) */ int afs_iget5_test(struct inode *inode, void *opaque) { - struct afs_iget_data *data = opaque; + struct afs_iget_data *iget_data = opaque; struct afs_vnode *vnode = AFS_FS_I(inode); - return memcmp(&vnode->fid, &data->fid, sizeof(data->fid)) == 0; + return memcmp(&vnode->fid, &iget_data->fid, sizeof(iget_data->fid)) == 0; } /* @@ -177,17 +368,19 @@ static int afs_iget5_pseudo_dir_test(struct inode *inode, void *opaque) */ static int afs_iget5_set(struct inode *inode, void *opaque) { - struct afs_iget_data *data = opaque; + struct afs_iget_data *iget_data = opaque; struct afs_vnode *vnode = AFS_FS_I(inode); - vnode->fid = data->fid; - vnode->volume = data->volume; + vnode->fid = iget_data->fid; + vnode->volume = iget_data->volume; + vnode->cb_v_break = iget_data->cb_v_break; + vnode->cb_s_break = iget_data->cb_s_break; /* YFS supports 96-bit vnode IDs, but Linux only supports * 64-bit inode numbers. */ - inode->i_ino = data->fid.vnode; - inode->i_generation = data->fid.unique; + inode->i_ino = iget_data->fid.vnode; + inode->i_generation = iget_data->fid.unique; return 0; } @@ -197,38 +390,42 @@ static int afs_iget5_set(struct inode *inode, void *opaque) */ struct inode *afs_iget_pseudo_dir(struct super_block *sb, bool root) { - struct afs_iget_data data; struct afs_super_info *as; struct afs_vnode *vnode; struct inode *inode; static atomic_t afs_autocell_ino; + struct afs_iget_data iget_data = { + .cb_v_break = 0, + .cb_s_break = 0, + }; + _enter(""); as = sb->s_fs_info; if (as->volume) { - data.volume = as->volume; - data.fid.vid = as->volume->vid; + iget_data.volume = as->volume; + iget_data.fid.vid = as->volume->vid; } if (root) { - data.fid.vnode = 1; - data.fid.unique = 1; + iget_data.fid.vnode = 1; + iget_data.fid.unique = 1; } else { - data.fid.vnode = atomic_inc_return(&afs_autocell_ino); - data.fid.unique = 0; + iget_data.fid.vnode = atomic_inc_return(&afs_autocell_ino); + iget_data.fid.unique = 0; } - inode = iget5_locked(sb, data.fid.vnode, + inode = iget5_locked(sb, iget_data.fid.vnode, afs_iget5_pseudo_dir_test, afs_iget5_set, - &data); + &iget_data); if (!inode) { _leave(" = -ENOMEM"); return ERR_PTR(-ENOMEM); } _debug("GOT INODE %p { ino=%lu, vl=%llx, vn=%llx, u=%x }", - inode, inode->i_ino, data.fid.vid, data.fid.vnode, - data.fid.unique); + inode, inode->i_ino, iget_data.fid.vid, iget_data.fid.vnode, + iget_data.fid.unique); vnode = AFS_FS_I(inode); @@ -299,23 +496,24 @@ static void afs_get_inode_cache(struct afs_vnode *vnode) * inode retrieval */ struct inode *afs_iget(struct super_block *sb, struct key *key, - struct afs_fid *fid, struct afs_file_status *status, - struct afs_callback *cb, struct afs_cb_interest *cbi, + struct afs_iget_data *iget_data, + struct afs_status_cb *scb, + struct afs_cb_interest *cbi, struct afs_vnode *parent_vnode) { - struct afs_iget_data data = { .fid = *fid }; struct afs_super_info *as; struct afs_vnode *vnode; + struct afs_fid *fid = &iget_data->fid; struct inode *inode; int ret; _enter(",{%llx:%llu.%u},,", fid->vid, fid->vnode, fid->unique); as = sb->s_fs_info; - data.volume = as->volume; + iget_data->volume = as->volume; inode = iget5_locked(sb, fid->vnode, afs_iget5_test, afs_iget5_set, - &data); + iget_data); if (!inode) { _leave(" = -ENOMEM"); return ERR_PTR(-ENOMEM); @@ -332,43 +530,25 @@ struct inode *afs_iget(struct super_block *sb, struct key *key, return inode; } - if (!status) { + if (!scb) { /* it's a remotely extant inode */ - ret = afs_fetch_status(vnode, key, true); + ret = afs_fetch_status(vnode, key, true, NULL); if (ret < 0) goto bad_inode; } else { - /* it's an inode we just created */ - memcpy(&vnode->status, status, sizeof(vnode->status)); - - if (!cb) { - /* it's a symlink we just created (the fileserver - * didn't give us a callback) */ - vnode->cb_version = 0; - vnode->cb_type = 0; - vnode->cb_expires_at = ktime_get(); - } else { - vnode->cb_version = cb->version; - vnode->cb_type = cb->type; - vnode->cb_expires_at = cb->expires_at; - vnode->cb_interest = afs_get_cb_interest(cbi); - set_bit(AFS_VNODE_CB_PROMISED, &vnode->flags); - } - - vnode->cb_expires_at += ktime_get_real_seconds(); + ret = afs_inode_init_from_status(vnode, key, cbi, parent_vnode, + scb); + if (ret < 0) + goto bad_inode; } - ret = afs_inode_init_from_status(vnode, key, parent_vnode); - if (ret < 0) - goto bad_inode; - afs_get_inode_cache(vnode); /* success */ clear_bit(AFS_VNODE_UNSET, &vnode->flags); inode->i_flags |= S_NOATIME; unlock_new_inode(inode); - _leave(" = %p [CB { v=%u t=%u }]", inode, vnode->cb_version, vnode->cb_type); + _leave(" = %p", inode); return inode; /* failure */ @@ -400,6 +580,66 @@ void afs_zap_data(struct afs_vnode *vnode) } /* + * Check the validity of a vnode/inode. + */ +bool afs_check_validity(struct afs_vnode *vnode) +{ + struct afs_cb_interest *cbi; + struct afs_server *server; + struct afs_volume *volume = vnode->volume; + time64_t now = ktime_get_real_seconds(); + bool valid, need_clear = false; + unsigned int cb_break, cb_s_break, cb_v_break; + int seq = 0; + + do { + read_seqbegin_or_lock(&vnode->cb_lock, &seq); + cb_v_break = READ_ONCE(volume->cb_v_break); + cb_break = vnode->cb_break; + + if (test_bit(AFS_VNODE_CB_PROMISED, &vnode->flags)) { + cbi = rcu_dereference(vnode->cb_interest); + server = rcu_dereference(cbi->server); + cb_s_break = READ_ONCE(server->cb_s_break); + + if (vnode->cb_s_break != cb_s_break || + vnode->cb_v_break != cb_v_break) { + vnode->cb_s_break = cb_s_break; + vnode->cb_v_break = cb_v_break; + need_clear = true; + valid = false; + } else if (test_bit(AFS_VNODE_ZAP_DATA, &vnode->flags)) { + need_clear = true; + valid = false; + } else if (vnode->cb_expires_at - 10 <= now) { + need_clear = true; + valid = false; + } else { + valid = true; + } + } else if (test_bit(AFS_VNODE_DELETED, &vnode->flags)) { + valid = true; + } else { + vnode->cb_v_break = cb_v_break; + valid = false; + } + + } while (need_seqretry(&vnode->cb_lock, seq)); + + done_seqretry(&vnode->cb_lock, seq); + + if (need_clear) { + write_seqlock(&vnode->cb_lock); + if (cb_break == vnode->cb_break) + __afs_break_callback(vnode); + write_sequnlock(&vnode->cb_lock); + valid = false; + } + + return valid; +} + +/* * validate a vnode/inode * - there are several things we need to check * - parent dir data changes (rm, rmdir, rename, mkdir, create, link, @@ -410,7 +650,6 @@ void afs_zap_data(struct afs_vnode *vnode) */ int afs_validate(struct afs_vnode *vnode, struct key *key) { - time64_t now = ktime_get_real_seconds(); bool valid; int ret; @@ -418,36 +657,9 @@ int afs_validate(struct afs_vnode *vnode, struct key *key) vnode->fid.vid, vnode->fid.vnode, vnode->flags, key_serial(key)); - /* Quickly check the callback state. Ideally, we'd use read_seqbegin - * here, but we have no way to pass the net namespace to the RCU - * cleanup for the server record. - */ - read_seqlock_excl(&vnode->cb_lock); - - if (test_bit(AFS_VNODE_CB_PROMISED, &vnode->flags)) { - if (vnode->cb_s_break != vnode->cb_interest->server->cb_s_break || - vnode->cb_v_break != vnode->volume->cb_v_break) { - vnode->cb_s_break = vnode->cb_interest->server->cb_s_break; - vnode->cb_v_break = vnode->volume->cb_v_break; - valid = false; - } else if (vnode->status.type == AFS_FTYPE_DIR && - (!test_bit(AFS_VNODE_DIR_VALID, &vnode->flags) || - vnode->cb_expires_at - 10 <= now)) { - valid = false; - } else if (test_bit(AFS_VNODE_ZAP_DATA, &vnode->flags) || - vnode->cb_expires_at - 10 <= now) { - valid = false; - } else { - valid = true; - } - } else if (test_bit(AFS_VNODE_DELETED, &vnode->flags)) { - valid = true; - } else { - vnode->cb_v_break = vnode->volume->cb_v_break; - valid = false; - } - - read_sequnlock_excl(&vnode->cb_lock); + rcu_read_lock(); + valid = afs_check_validity(vnode); + rcu_read_unlock(); if (test_bit(AFS_VNODE_DELETED, &vnode->flags)) clear_nlink(&vnode->vfs_inode); @@ -463,7 +675,7 @@ int afs_validate(struct afs_vnode *vnode, struct key *key) * access */ if (!test_bit(AFS_VNODE_CB_PROMISED, &vnode->flags)) { _debug("not promised"); - ret = afs_fetch_status(vnode, key, false); + ret = afs_fetch_status(vnode, key, false, NULL); if (ret < 0) { if (ret == -ENOENT) { set_bit(AFS_VNODE_DELETED, &vnode->flags); @@ -534,6 +746,7 @@ int afs_drop_inode(struct inode *inode) */ void afs_evict_inode(struct inode *inode) { + struct afs_cb_interest *cbi; struct afs_vnode *vnode; vnode = AFS_FS_I(inode); @@ -550,10 +763,14 @@ void afs_evict_inode(struct inode *inode) truncate_inode_pages_final(&inode->i_data); clear_inode(inode); - if (vnode->cb_interest) { - afs_put_cb_interest(afs_i2net(inode), vnode->cb_interest); - vnode->cb_interest = NULL; + write_seqlock(&vnode->cb_lock); + cbi = rcu_dereference_protected(vnode->cb_interest, + lockdep_is_held(&vnode->cb_lock.lock)); + if (cbi) { + afs_put_cb_interest(afs_i2net(inode), cbi); + rcu_assign_pointer(vnode->cb_interest, NULL); } + write_sequnlock(&vnode->cb_lock); while (!list_empty(&vnode->wb_keys)) { struct afs_wb_key *wbk = list_entry(vnode->wb_keys.next, @@ -573,6 +790,7 @@ void afs_evict_inode(struct inode *inode) } #endif + afs_prune_wb_keys(vnode); afs_put_permits(rcu_access_pointer(vnode->permit_cache)); key_put(vnode->silly_key); vnode->silly_key = NULL; @@ -587,9 +805,10 @@ void afs_evict_inode(struct inode *inode) int afs_setattr(struct dentry *dentry, struct iattr *attr) { struct afs_fs_cursor fc; + struct afs_status_cb *scb; struct afs_vnode *vnode = AFS_FS_I(d_inode(dentry)); struct key *key; - int ret; + int ret = -ENOMEM; _enter("{%llx:%llu},{n=%pd},%x", vnode->fid.vid, vnode->fid.vnode, dentry, @@ -601,6 +820,10 @@ int afs_setattr(struct dentry *dentry, struct iattr *attr) return 0; } + scb = kzalloc(sizeof(struct afs_status_cb), GFP_KERNEL); + if (!scb) + goto error; + /* flush any dirty data outstanding on a regular file */ if (S_ISREG(vnode->vfs_inode.i_mode)) filemap_write_and_wait(vnode->vfs_inode.i_mapping); @@ -611,25 +834,33 @@ int afs_setattr(struct dentry *dentry, struct iattr *attr) key = afs_request_key(vnode->volume->cell); if (IS_ERR(key)) { ret = PTR_ERR(key); - goto error; + goto error_scb; } } ret = -ERESTARTSYS; - if (afs_begin_vnode_operation(&fc, vnode, key)) { + if (afs_begin_vnode_operation(&fc, vnode, key, false)) { + afs_dataversion_t data_version = vnode->status.data_version; + + if (attr->ia_valid & ATTR_SIZE) + data_version++; + while (afs_select_fileserver(&fc)) { fc.cb_break = afs_calc_vnode_cb_break(vnode); - afs_fs_setattr(&fc, attr); + afs_fs_setattr(&fc, attr, scb); } - afs_check_for_remote_deletion(&fc, fc.vnode); - afs_vnode_commit_status(&fc, vnode, fc.cb_break); + afs_check_for_remote_deletion(&fc, vnode); + afs_vnode_commit_status(&fc, vnode, fc.cb_break, + &data_version, scb); ret = afs_end_vnode_operation(&fc); } if (!(attr->ia_valid & ATTR_FILE)) key_put(key); +error_scb: + kfree(scb); error: _leave(" = %d", ret); return ret; diff --git a/fs/afs/internal.h b/fs/afs/internal.h index b3cd6e8ad59d..2073c1a3ab4b 100644 --- a/fs/afs/internal.h +++ b/fs/afs/internal.h @@ -66,6 +66,8 @@ struct afs_fs_context { struct afs_iget_data { struct afs_fid fid; struct afs_volume *volume; /* volume on which resides */ + unsigned int cb_v_break; /* Pre-fetch volume break count */ + unsigned int cb_s_break; /* Pre-fetch server break count */ }; enum afs_call_state { @@ -111,8 +113,12 @@ struct afs_call { struct rxrpc_call *rxcall; /* RxRPC call handle */ struct key *key; /* security for this call */ struct afs_net *net; /* The network namespace */ - struct afs_server *cm_server; /* Server affected by incoming CM call */ + union { + struct afs_server *server; + struct afs_vlserver *vlserver; + }; struct afs_cb_interest *cbi; /* Callback interest for server used */ + struct afs_vnode *lvnode; /* vnode being locked */ void *request; /* request data (first part) */ struct address_space *mapping; /* Pages being written from */ struct iov_iter iter; /* Buffer iterator */ @@ -122,7 +128,20 @@ struct afs_call { struct bio_vec bvec[1]; }; void *buffer; /* reply receive buffer */ - void *reply[4]; /* Where to put the reply */ + union { + long ret0; /* Value to reply with instead of 0 */ + struct afs_addr_list *ret_alist; + struct afs_vldb_entry *ret_vldb; + struct afs_acl *ret_acl; + }; + struct afs_fid *out_fid; + struct afs_status_cb *out_dir_scb; + struct afs_status_cb *out_scb; + struct yfs_acl *out_yacl; + struct afs_volsync *out_volsync; + struct afs_volume_status *out_volstatus; + struct afs_read *read_request; + unsigned int server_index; pgoff_t first; /* first page in mapping to deal with */ pgoff_t last; /* last page in mapping to deal with */ atomic_t usage; @@ -131,10 +150,10 @@ struct afs_call { int error; /* error code */ u32 abort_code; /* Remote abort ID or 0 */ u32 epoch; + unsigned int max_lifespan; /* Maximum lifespan to set if not 0 */ unsigned request_size; /* size of request data */ unsigned reply_max; /* maximum size of reply */ unsigned first_offset; /* offset into mapping[first] */ - unsigned int cb_break; /* cb_break + cb_s_break before the call */ union { unsigned last_to; /* amount of mapping[last] */ unsigned count2; /* count used in unmarshalling */ @@ -145,9 +164,9 @@ struct afs_call { bool send_pages; /* T if data from mapping should be sent */ bool need_attention; /* T if RxRPC poked us */ bool async; /* T if asynchronous */ - bool ret_reply0; /* T if should return reply[0] on success */ bool upgrade; /* T to request service upgrade */ - bool want_reply_time; /* T if want reply_time */ + bool have_reply_time; /* T if have got reply_time */ + bool intr; /* T if interruptible */ u16 service_id; /* Actual service ID (after upgrade) */ unsigned int debug_id; /* Trace ID */ u32 operation_ID; /* operation ID for an incoming call */ @@ -159,8 +178,6 @@ struct afs_call { } __attribute__((packed)); __be64 tmp64; }; - afs_dataversion_t expected_version; /* Updated version expected from store */ - afs_dataversion_t expected_version_2; /* 2nd updated version expected from store */ ktime_t reply_time; /* Time of first reply packet */ }; @@ -221,7 +238,8 @@ struct afs_read { unsigned int index; /* Which page we're reading into */ unsigned int nr_pages; unsigned int offset; /* offset into current page */ - void (*page_done)(struct afs_call *, struct afs_read *); + struct afs_vnode *vnode; + void (*page_done)(struct afs_read *); struct page **pages; struct page *array[]; }; @@ -367,13 +385,13 @@ struct afs_cell { time64_t last_inactive; /* Time of last drop of usage count */ atomic_t usage; unsigned long flags; -#define AFS_CELL_FL_NOT_READY 0 /* The cell record is not ready for use */ -#define AFS_CELL_FL_NO_GC 1 /* The cell was added manually, don't auto-gc */ -#define AFS_CELL_FL_NOT_FOUND 2 /* Permanent DNS error */ -#define AFS_CELL_FL_DNS_FAIL 3 /* Failed to access DNS */ -#define AFS_CELL_FL_NO_LOOKUP_YET 4 /* Not completed first DNS lookup yet */ +#define AFS_CELL_FL_NO_GC 0 /* The cell was added manually, don't auto-gc */ +#define AFS_CELL_FL_DO_LOOKUP 1 /* DNS lookup requested */ enum afs_cell_state state; short error; + enum dns_record_source dns_source:8; /* Latest source of data from lookup */ + enum dns_lookup_status dns_status:8; /* Latest status of data from lookup */ + unsigned int dns_lookup_count; /* Counter of DNS lookups */ /* Active fileserver interaction state. */ struct list_head proc_volumes; /* procfs volume list */ @@ -538,7 +556,10 @@ struct afs_server { struct afs_vol_interest { struct hlist_node srv_link; /* Link in server->cb_volumes */ struct hlist_head cb_interests; /* List of callback interests on the server */ - afs_volid_t vid; /* Volume ID to match */ + union { + struct rcu_head rcu; + afs_volid_t vid; /* Volume ID to match */ + }; unsigned int usage; }; @@ -550,7 +571,10 @@ struct afs_cb_interest { struct afs_vol_interest *vol_interest; struct afs_server *server; /* Server on which this interest resides */ struct super_block *sb; /* Superblock on which inodes reside */ - afs_volid_t vid; /* Volume ID to match */ + union { + struct rcu_head rcu; + afs_volid_t vid; /* Volume ID to match */ + }; refcount_t usage; }; @@ -660,15 +684,13 @@ struct afs_vnode { afs_lock_type_t lock_type : 8; /* outstanding callback notification on this file */ - struct afs_cb_interest *cb_interest; /* Server on which this resides */ + struct afs_cb_interest __rcu *cb_interest; /* Server on which this resides */ unsigned int cb_s_break; /* Mass break counter on ->server */ unsigned int cb_v_break; /* Mass break counter on ->volume */ unsigned int cb_break; /* Break counter on vnode */ seqlock_t cb_lock; /* Lock for ->cb_interest, ->status, ->cb_*break */ time64_t cb_expires_at; /* time at which callback expires */ - unsigned cb_version; /* callback version */ - afs_callback_type_t cb_type; /* type of callback */ }; static inline struct fscache_cookie *afs_vnode_cache(struct afs_vnode *vnode) @@ -755,6 +777,7 @@ struct afs_vl_cursor { * Cursor for iterating over a set of fileservers. */ struct afs_fs_cursor { + const struct afs_call_type *type; /* Type of call done */ struct afs_addr_cursor ac; struct afs_vnode *vnode; struct afs_server_list *server_list; /* Current server list (pins ref) */ @@ -772,6 +795,7 @@ struct afs_fs_cursor { #define AFS_FS_CURSOR_VNOVOL 0x0008 /* Set if seen VNOVOL */ #define AFS_FS_CURSOR_CUR_ONLY 0x0010 /* Set if current server only (file lock held) */ #define AFS_FS_CURSOR_NO_VSLEEP 0x0020 /* Set to prevent sleep on VBUSY, VOFFLINE, ... */ +#define AFS_FS_CURSOR_INTR 0x0040 /* Set if op is interruptible */ unsigned short nr_iterations; /* Number of server iterations */ }; @@ -882,7 +906,6 @@ extern const struct address_space_operations afs_dir_aops; extern const struct dentry_operations afs_fs_dentry_operations; extern void afs_d_release(struct dentry *); -extern int afs_dir_remove_link(struct dentry *, struct key *, unsigned long, unsigned long); /* * dir_edit.c @@ -940,50 +963,48 @@ extern int afs_flock(struct file *, int, struct file_lock *); /* * fsclient.c */ -#define AFS_VNODE_NOT_YET_SET 0x01 -#define AFS_VNODE_META_CHANGED 0x02 -#define AFS_VNODE_DATA_CHANGED 0x04 -extern void afs_update_inode_from_status(struct afs_vnode *, struct afs_file_status *, - const afs_dataversion_t *, u8); - -extern int afs_fs_fetch_file_status(struct afs_fs_cursor *, struct afs_volsync *, bool); +extern int afs_fs_fetch_file_status(struct afs_fs_cursor *, struct afs_status_cb *, + struct afs_volsync *); extern int afs_fs_give_up_callbacks(struct afs_net *, struct afs_server *); -extern int afs_fs_fetch_data(struct afs_fs_cursor *, struct afs_read *); -extern int afs_fs_create(struct afs_fs_cursor *, const char *, umode_t, u64, - struct afs_fid *, struct afs_file_status *, struct afs_callback *); -extern int afs_fs_remove(struct afs_fs_cursor *, struct afs_vnode *, const char *, bool, u64); -extern int afs_fs_link(struct afs_fs_cursor *, struct afs_vnode *, const char *, u64); -extern int afs_fs_symlink(struct afs_fs_cursor *, const char *, const char *, u64, - struct afs_fid *, struct afs_file_status *); +extern int afs_fs_fetch_data(struct afs_fs_cursor *, struct afs_status_cb *, struct afs_read *); +extern int afs_fs_create(struct afs_fs_cursor *, const char *, umode_t, + struct afs_status_cb *, struct afs_fid *, struct afs_status_cb *); +extern int afs_fs_remove(struct afs_fs_cursor *, struct afs_vnode *, const char *, bool, + struct afs_status_cb *); +extern int afs_fs_link(struct afs_fs_cursor *, struct afs_vnode *, const char *, + struct afs_status_cb *, struct afs_status_cb *); +extern int afs_fs_symlink(struct afs_fs_cursor *, const char *, const char *, + struct afs_status_cb *, struct afs_fid *, struct afs_status_cb *); extern int afs_fs_rename(struct afs_fs_cursor *, const char *, - struct afs_vnode *, const char *, u64, u64); + struct afs_vnode *, const char *, + struct afs_status_cb *, struct afs_status_cb *); extern int afs_fs_store_data(struct afs_fs_cursor *, struct address_space *, - pgoff_t, pgoff_t, unsigned, unsigned); -extern int afs_fs_setattr(struct afs_fs_cursor *, struct iattr *); + pgoff_t, pgoff_t, unsigned, unsigned, struct afs_status_cb *); +extern int afs_fs_setattr(struct afs_fs_cursor *, struct iattr *, struct afs_status_cb *); extern int afs_fs_get_volume_status(struct afs_fs_cursor *, struct afs_volume_status *); -extern int afs_fs_set_lock(struct afs_fs_cursor *, afs_lock_type_t); -extern int afs_fs_extend_lock(struct afs_fs_cursor *); -extern int afs_fs_release_lock(struct afs_fs_cursor *); +extern int afs_fs_set_lock(struct afs_fs_cursor *, afs_lock_type_t, struct afs_status_cb *); +extern int afs_fs_extend_lock(struct afs_fs_cursor *, struct afs_status_cb *); +extern int afs_fs_release_lock(struct afs_fs_cursor *, struct afs_status_cb *); extern int afs_fs_give_up_all_callbacks(struct afs_net *, struct afs_server *, struct afs_addr_cursor *, struct key *); extern struct afs_call *afs_fs_get_capabilities(struct afs_net *, struct afs_server *, struct afs_addr_cursor *, struct key *, unsigned int); extern int afs_fs_inline_bulk_status(struct afs_fs_cursor *, struct afs_net *, - struct afs_fid *, struct afs_file_status *, - struct afs_callback *, unsigned int, - struct afs_volsync *); + struct afs_fid *, struct afs_status_cb *, + unsigned int, struct afs_volsync *); extern int afs_fs_fetch_status(struct afs_fs_cursor *, struct afs_net *, - struct afs_fid *, struct afs_file_status *, - struct afs_callback *, struct afs_volsync *); + struct afs_fid *, struct afs_status_cb *, + struct afs_volsync *); struct afs_acl { u32 size; u8 data[]; }; -extern struct afs_acl *afs_fs_fetch_acl(struct afs_fs_cursor *); -extern int afs_fs_store_acl(struct afs_fs_cursor *, const struct afs_acl *); +extern struct afs_acl *afs_fs_fetch_acl(struct afs_fs_cursor *, struct afs_status_cb *); +extern int afs_fs_store_acl(struct afs_fs_cursor *, const struct afs_acl *, + struct afs_status_cb *); /* * fs_probe.c @@ -995,15 +1016,20 @@ extern int afs_wait_for_fs_probes(struct afs_server_list *, unsigned long); /* * inode.c */ -extern int afs_fetch_status(struct afs_vnode *, struct key *, bool); +extern void afs_vnode_commit_status(struct afs_fs_cursor *, + struct afs_vnode *, + unsigned int, + const afs_dataversion_t *, + struct afs_status_cb *); +extern int afs_fetch_status(struct afs_vnode *, struct key *, bool, afs_access_t *); extern int afs_iget5_test(struct inode *, void *); extern struct inode *afs_iget_pseudo_dir(struct super_block *, bool); extern struct inode *afs_iget(struct super_block *, struct key *, - struct afs_fid *, struct afs_file_status *, - struct afs_callback *, + struct afs_iget_data *, struct afs_status_cb *, struct afs_cb_interest *, struct afs_vnode *); extern void afs_zap_data(struct afs_vnode *); +extern bool afs_check_validity(struct afs_vnode *); extern int afs_validate(struct afs_vnode *, struct key *); extern int afs_getattr(const struct path *, struct kstat *, u32, unsigned int); extern int afs_setattr(struct dentry *, struct iattr *); @@ -1096,7 +1122,7 @@ static inline void afs_put_sysnames(struct afs_sysnames *sysnames) {} * rotate.c */ extern bool afs_begin_vnode_operation(struct afs_fs_cursor *, struct afs_vnode *, - struct key *); + struct key *, bool); extern bool afs_select_fileserver(struct afs_fs_cursor *); extern bool afs_select_current_fileserver(struct afs_fs_cursor *); extern int afs_end_vnode_operation(struct afs_fs_cursor *); @@ -1121,6 +1147,12 @@ extern void afs_send_simple_reply(struct afs_call *, const void *, size_t); extern int afs_extract_data(struct afs_call *, bool); extern int afs_protocol_error(struct afs_call *, int, enum afs_eproto_cause); +static inline void afs_set_fc_call(struct afs_call *call, struct afs_fs_cursor *fc) +{ + call->intr = fc->flags & AFS_FS_CURSOR_INTR; + fc->type = call->type; +} + static inline void afs_extract_begin(struct afs_call *call, void *buf, size_t size) { call->kvec[0].iov_base = buf; @@ -1201,7 +1233,8 @@ static inline void afs_set_call_complete(struct afs_call *call, */ extern void afs_put_permits(struct afs_permits *); extern void afs_clear_permits(struct afs_vnode *); -extern void afs_cache_permit(struct afs_vnode *, struct key *, unsigned int); +extern void afs_cache_permit(struct afs_vnode *, struct key *, unsigned int, + struct afs_status_cb *); extern void afs_zap_permits(struct rcu_head *); extern struct key *afs_request_key(struct afs_cell *); extern int afs_check_permit(struct afs_vnode *, struct key *, afs_access_t *); @@ -1327,7 +1360,6 @@ extern int afs_write_end(struct file *file, struct address_space *mapping, struct page *page, void *fsdata); extern int afs_writepage(struct page *, struct writeback_control *); extern int afs_writepages(struct address_space *, struct writeback_control *); -extern void afs_pages_written_back(struct afs_vnode *, struct afs_call *); extern ssize_t afs_file_write(struct kiocb *, struct iov_iter *); extern int afs_fsync(struct file *, loff_t, loff_t, int); extern vm_fault_t afs_page_mkwrite(struct vm_fault *vmf); @@ -1343,33 +1375,36 @@ extern ssize_t afs_listxattr(struct dentry *, char *, size_t); /* * yfsclient.c */ -extern int yfs_fs_fetch_file_status(struct afs_fs_cursor *, struct afs_volsync *, bool); -extern int yfs_fs_fetch_data(struct afs_fs_cursor *, struct afs_read *); -extern int yfs_fs_create_file(struct afs_fs_cursor *, const char *, umode_t, u64, - struct afs_fid *, struct afs_file_status *, struct afs_callback *); -extern int yfs_fs_make_dir(struct afs_fs_cursor *, const char *, umode_t, u64, - struct afs_fid *, struct afs_file_status *, struct afs_callback *); -extern int yfs_fs_remove_file2(struct afs_fs_cursor *, struct afs_vnode *, const char *, u64); -extern int yfs_fs_remove(struct afs_fs_cursor *, struct afs_vnode *, const char *, bool, u64); -extern int yfs_fs_link(struct afs_fs_cursor *, struct afs_vnode *, const char *, u64); -extern int yfs_fs_symlink(struct afs_fs_cursor *, const char *, const char *, u64, - struct afs_fid *, struct afs_file_status *); -extern int yfs_fs_rename(struct afs_fs_cursor *, const char *, - struct afs_vnode *, const char *, u64, u64); +extern int yfs_fs_fetch_file_status(struct afs_fs_cursor *, struct afs_status_cb *, + struct afs_volsync *); +extern int yfs_fs_fetch_data(struct afs_fs_cursor *, struct afs_status_cb *, struct afs_read *); +extern int yfs_fs_create_file(struct afs_fs_cursor *, const char *, umode_t, struct afs_status_cb *, + struct afs_fid *, struct afs_status_cb *); +extern int yfs_fs_make_dir(struct afs_fs_cursor *, const char *, umode_t, struct afs_status_cb *, + struct afs_fid *, struct afs_status_cb *); +extern int yfs_fs_remove_file2(struct afs_fs_cursor *, struct afs_vnode *, const char *, + struct afs_status_cb *, struct afs_status_cb *); +extern int yfs_fs_remove(struct afs_fs_cursor *, struct afs_vnode *, const char *, bool, + struct afs_status_cb *); +extern int yfs_fs_link(struct afs_fs_cursor *, struct afs_vnode *, const char *, + struct afs_status_cb *, struct afs_status_cb *); +extern int yfs_fs_symlink(struct afs_fs_cursor *, const char *, const char *, + struct afs_status_cb *, struct afs_fid *, struct afs_status_cb *); +extern int yfs_fs_rename(struct afs_fs_cursor *, const char *, struct afs_vnode *, const char *, + struct afs_status_cb *, struct afs_status_cb *); extern int yfs_fs_store_data(struct afs_fs_cursor *, struct address_space *, - pgoff_t, pgoff_t, unsigned, unsigned); -extern int yfs_fs_setattr(struct afs_fs_cursor *, struct iattr *); + pgoff_t, pgoff_t, unsigned, unsigned, struct afs_status_cb *); +extern int yfs_fs_setattr(struct afs_fs_cursor *, struct iattr *, struct afs_status_cb *); extern int yfs_fs_get_volume_status(struct afs_fs_cursor *, struct afs_volume_status *); -extern int yfs_fs_set_lock(struct afs_fs_cursor *, afs_lock_type_t); -extern int yfs_fs_extend_lock(struct afs_fs_cursor *); -extern int yfs_fs_release_lock(struct afs_fs_cursor *); +extern int yfs_fs_set_lock(struct afs_fs_cursor *, afs_lock_type_t, struct afs_status_cb *); +extern int yfs_fs_extend_lock(struct afs_fs_cursor *, struct afs_status_cb *); +extern int yfs_fs_release_lock(struct afs_fs_cursor *, struct afs_status_cb *); extern int yfs_fs_fetch_status(struct afs_fs_cursor *, struct afs_net *, - struct afs_fid *, struct afs_file_status *, - struct afs_callback *, struct afs_volsync *); + struct afs_fid *, struct afs_status_cb *, + struct afs_volsync *); extern int yfs_fs_inline_bulk_status(struct afs_fs_cursor *, struct afs_net *, - struct afs_fid *, struct afs_file_status *, - struct afs_callback *, unsigned int, - struct afs_volsync *); + struct afs_fid *, struct afs_status_cb *, + unsigned int, struct afs_volsync *); struct yfs_acl { struct afs_acl *acl; /* Dir/file/symlink ACL */ @@ -1382,8 +1417,10 @@ struct yfs_acl { }; extern void yfs_free_opaque_acl(struct yfs_acl *); -extern struct yfs_acl *yfs_fs_fetch_opaque_acl(struct afs_fs_cursor *, unsigned int); -extern int yfs_fs_store_opaque_acl2(struct afs_fs_cursor *, const struct afs_acl *); +extern struct yfs_acl *yfs_fs_fetch_opaque_acl(struct afs_fs_cursor *, struct yfs_acl *, + struct afs_status_cb *); +extern int yfs_fs_store_opaque_acl2(struct afs_fs_cursor *, const struct afs_acl *, + struct afs_status_cb *); /* * Miscellaneous inline functions. @@ -1398,14 +1435,6 @@ static inline struct inode *AFS_VNODE_TO_I(struct afs_vnode *vnode) return &vnode->vfs_inode; } -static inline void afs_vnode_commit_status(struct afs_fs_cursor *fc, - struct afs_vnode *vnode, - unsigned int cb_break) -{ - if (fc->ac.error == 0) - afs_cache_permit(vnode, fc->key, cb_break); -} - static inline void afs_check_for_remote_deletion(struct afs_fs_cursor *fc, struct afs_vnode *vnode) { diff --git a/fs/afs/proc.c b/fs/afs/proc.c index be2ee3bbd0a9..371501d28e08 100644 --- a/fs/afs/proc.c +++ b/fs/afs/proc.c @@ -53,7 +53,7 @@ static int afs_proc_cells_show(struct seq_file *m, void *v) seq_printf(m, "%3u %6lld %2u %s\n", atomic_read(&cell->usage), cell->dns_expiry - ktime_get_real_seconds(), - vllist ? vllist->nr_servers : 0, + vllist->nr_servers, cell->name); return 0; } @@ -296,8 +296,8 @@ static int afs_proc_cell_vlservers_show(struct seq_file *m, void *v) if (v == SEQ_START_TOKEN) { seq_printf(m, "# source %s, status %s\n", - dns_record_sources[vllist->source], - dns_lookup_statuses[vllist->status]); + dns_record_sources[vllist ? vllist->source : 0], + dns_lookup_statuses[vllist ? vllist->status : 0]); return 0; } @@ -336,7 +336,7 @@ static void *afs_proc_cell_vlservers_start(struct seq_file *m, loff_t *_pos) if (pos == 0) return SEQ_START_TOKEN; - if (!vllist || pos - 1 >= vllist->nr_servers) + if (pos - 1 >= vllist->nr_servers) return NULL; return &vllist->servers[pos - 1]; diff --git a/fs/afs/rotate.c b/fs/afs/rotate.c index c3ae324781f8..b00c739e0e63 100644 --- a/fs/afs/rotate.c +++ b/fs/afs/rotate.c @@ -25,7 +25,7 @@ * them here also using the io_lock. */ bool afs_begin_vnode_operation(struct afs_fs_cursor *fc, struct afs_vnode *vnode, - struct key *key) + struct key *key, bool intr) { memset(fc, 0, sizeof(*fc)); fc->vnode = vnode; @@ -33,10 +33,15 @@ bool afs_begin_vnode_operation(struct afs_fs_cursor *fc, struct afs_vnode *vnode fc->ac.error = SHRT_MAX; fc->error = -EDESTADDRREQ; - if (mutex_lock_interruptible(&vnode->io_lock) < 0) { - fc->error = -EINTR; - fc->flags |= AFS_FS_CURSOR_STOP; - return false; + if (intr) { + fc->flags |= AFS_FS_CURSOR_INTR; + if (mutex_lock_interruptible(&vnode->io_lock) < 0) { + fc->error = -EINTR; + fc->flags |= AFS_FS_CURSOR_STOP; + return false; + } + } else { + mutex_lock(&vnode->io_lock); } if (vnode->lock_state != AFS_VNODE_LOCK_NONE) @@ -61,7 +66,8 @@ static bool afs_start_fs_iteration(struct afs_fs_cursor *fc, fc->untried = (1UL << fc->server_list->nr_servers) - 1; fc->index = READ_ONCE(fc->server_list->preferred); - cbi = vnode->cb_interest; + cbi = rcu_dereference_protected(vnode->cb_interest, + lockdep_is_held(&vnode->io_lock)); if (cbi) { /* See if the vnode's preferred record is still available */ for (i = 0; i < fc->server_list->nr_servers; i++) { @@ -82,8 +88,8 @@ static bool afs_start_fs_iteration(struct afs_fs_cursor *fc, /* Note that the callback promise is effectively broken */ write_seqlock(&vnode->cb_lock); - ASSERTCMP(cbi, ==, vnode->cb_interest); - vnode->cb_interest = NULL; + ASSERTCMP(cbi, ==, rcu_access_pointer(vnode->cb_interest)); + rcu_assign_pointer(vnode->cb_interest, NULL); if (test_and_clear_bit(AFS_VNODE_CB_PROMISED, &vnode->flags)) vnode->cb_break++; write_sequnlock(&vnode->cb_lock); @@ -118,10 +124,14 @@ static void afs_busy(struct afs_volume *volume, u32 abort_code) */ static bool afs_sleep_and_retry(struct afs_fs_cursor *fc) { - msleep_interruptible(1000); - if (signal_pending(current)) { - fc->error = -ERESTARTSYS; - return false; + if (fc->flags & AFS_FS_CURSOR_INTR) { + msleep_interruptible(1000); + if (signal_pending(current)) { + fc->error = -ERESTARTSYS; + return false; + } + } else { + msleep(1000); } return true; @@ -408,7 +418,9 @@ selected_server: if (error < 0) goto failed_set_error; - fc->cbi = afs_get_cb_interest(vnode->cb_interest); + fc->cbi = afs_get_cb_interest( + rcu_dereference_protected(vnode->cb_interest, + lockdep_is_held(&vnode->io_lock))); read_lock(&server->fs_lock); alist = rcu_dereference_protected(server->addresses, @@ -459,6 +471,8 @@ no_more_servers: s->probe.abort_code); } + error = e.error; + failed_set_error: fc->error = error; failed: @@ -476,12 +490,15 @@ failed: bool afs_select_current_fileserver(struct afs_fs_cursor *fc) { struct afs_vnode *vnode = fc->vnode; - struct afs_cb_interest *cbi = vnode->cb_interest; + struct afs_cb_interest *cbi; struct afs_addr_list *alist; int error = fc->ac.error; _enter(""); + cbi = rcu_dereference_protected(vnode->cb_interest, + lockdep_is_held(&vnode->io_lock)); + switch (error) { case SHRT_MAX: if (!cbi) { @@ -490,7 +507,7 @@ bool afs_select_current_fileserver(struct afs_fs_cursor *fc) return false; } - fc->cbi = afs_get_cb_interest(vnode->cb_interest); + fc->cbi = afs_get_cb_interest(cbi); read_lock(&cbi->server->fs_lock); alist = rcu_dereference_protected(cbi->server->addresses, diff --git a/fs/afs/rxrpc.c b/fs/afs/rxrpc.c index a34a89c75c6a..4fa5ce92b9b9 100644 --- a/fs/afs/rxrpc.c +++ b/fs/afs/rxrpc.c @@ -188,7 +188,7 @@ void afs_put_call(struct afs_call *call) if (call->type->destructor) call->type->destructor(call); - afs_put_server(call->net, call->cm_server); + afs_put_server(call->net, call->server); afs_put_cb_interest(call->net, call->cbi); afs_put_addrlist(call->alist); kfree(call->request); @@ -417,6 +417,7 @@ void afs_make_call(struct afs_addr_cursor *ac, struct afs_call *call, gfp_t gfp) afs_wake_up_async_call : afs_wake_up_call_waiter), call->upgrade, + call->intr, call->debug_id); if (IS_ERR(rxcall)) { ret = PTR_ERR(rxcall); @@ -426,6 +427,10 @@ void afs_make_call(struct afs_addr_cursor *ac, struct afs_call *call, gfp_t gfp) call->rxcall = rxcall; + if (call->max_lifespan) + rxrpc_kernel_set_max_life(call->net->socket, rxcall, + call->max_lifespan); + /* send the request */ iov[0].iov_base = call->request; iov[0].iov_len = call->request_size; @@ -529,11 +534,11 @@ static void afs_deliver_to_call(struct afs_call *call) return; } - if (call->want_reply_time && + if (!call->have_reply_time && rxrpc_kernel_get_reply_time(call->net->socket, call->rxcall, &call->reply_time)) - call->want_reply_time = false; + call->have_reply_time = true; ret = call->type->deliver(call); state = READ_ONCE(call->state); @@ -648,7 +653,7 @@ long afs_wait_for_call_to_complete(struct afs_call *call, break; } - if (timeout == 0 && + if (call->intr && timeout == 0 && life == last_life && signal_pending(current)) { if (stalled) break; @@ -691,10 +696,9 @@ long afs_wait_for_call_to_complete(struct afs_call *call, ret = ac->error; switch (ret) { case 0: - if (call->ret_reply0) { - ret = (long)call->reply[0]; - call->reply[0] = NULL; - } + ret = call->ret0; + call->ret0 = 0; + /* Fall through */ case -ECONNABORTED: ac->responded = true; diff --git a/fs/afs/security.c b/fs/afs/security.c index 5f58a9a17e69..5d8ece98561e 100644 --- a/fs/afs/security.c +++ b/fs/afs/security.c @@ -87,11 +87,9 @@ void afs_clear_permits(struct afs_vnode *vnode) permits = rcu_dereference_protected(vnode->permit_cache, lockdep_is_held(&vnode->lock)); RCU_INIT_POINTER(vnode->permit_cache, NULL); - vnode->cb_break++; spin_unlock(&vnode->lock); - if (permits) - afs_put_permits(permits); + afs_put_permits(permits); } /* @@ -118,10 +116,10 @@ static void afs_hash_permits(struct afs_permits *permits) * as the ACL *may* have changed. */ void afs_cache_permit(struct afs_vnode *vnode, struct key *key, - unsigned int cb_break) + unsigned int cb_break, struct afs_status_cb *scb) { struct afs_permits *permits, *xpermits, *replacement, *zap, *new = NULL; - afs_access_t caller_access = READ_ONCE(vnode->status.caller_access); + afs_access_t caller_access = scb->status.caller_access; size_t size = 0; bool changed = false; int i, j; @@ -148,7 +146,7 @@ void afs_cache_permit(struct afs_vnode *vnode, struct key *key, } if (afs_cb_is_broken(cb_break, vnode, - vnode->cb_interest)) { + rcu_dereference(vnode->cb_interest))) { changed = true; break; } @@ -178,7 +176,7 @@ void afs_cache_permit(struct afs_vnode *vnode, struct key *key, } } - if (afs_cb_is_broken(cb_break, vnode, vnode->cb_interest)) + if (afs_cb_is_broken(cb_break, vnode, rcu_dereference(vnode->cb_interest))) goto someone_else_changed_it; /* We need a ref on any permits list we want to copy as we'll have to @@ -255,14 +253,16 @@ found: kfree(new); + rcu_read_lock(); spin_lock(&vnode->lock); zap = rcu_access_pointer(vnode->permit_cache); - if (!afs_cb_is_broken(cb_break, vnode, vnode->cb_interest) && + if (!afs_cb_is_broken(cb_break, vnode, rcu_dereference(vnode->cb_interest)) && zap == permits) rcu_assign_pointer(vnode->permit_cache, replacement); else zap = replacement; spin_unlock(&vnode->lock); + rcu_read_unlock(); afs_put_permits(zap); out_put: afs_put_permits(permits); @@ -322,13 +322,12 @@ int afs_check_permit(struct afs_vnode *vnode, struct key *key, */ _debug("no valid permit"); - ret = afs_fetch_status(vnode, key, false); + ret = afs_fetch_status(vnode, key, false, _access); if (ret < 0) { *_access = 0; _leave(" = %d", ret); return ret; } - *_access = vnode->status.caller_access; } _leave(" = 0 [access %x]", *_access); diff --git a/fs/afs/server.c b/fs/afs/server.c index 65b33b6da48b..52c170b59cfd 100644 --- a/fs/afs/server.c +++ b/fs/afs/server.c @@ -521,8 +521,15 @@ static noinline bool afs_update_server_record(struct afs_fs_cursor *fc, struct a alist = afs_vl_lookup_addrs(fc->vnode->volume->cell, fc->key, &server->uuid); if (IS_ERR(alist)) { - fc->ac.error = PTR_ERR(alist); - _leave(" = f [%d]", fc->ac.error); + if ((PTR_ERR(alist) == -ERESTARTSYS || + PTR_ERR(alist) == -EINTR) && + !(fc->flags & AFS_FS_CURSOR_INTR) && + server->addresses) { + _leave(" = t [intr]"); + return true; + } + fc->error = PTR_ERR(alist); + _leave(" = f [%d]", fc->error); return false; } @@ -574,7 +581,11 @@ retry: ret = wait_on_bit(&server->flags, AFS_SERVER_FL_UPDATING, TASK_INTERRUPTIBLE); if (ret == -ERESTARTSYS) { - fc->ac.error = ret; + if (!(fc->flags & AFS_FS_CURSOR_INTR) && server->addresses) { + _leave(" = t [intr]"); + return true; + } + fc->error = ret; _leave(" = f [intr]"); return false; } diff --git a/fs/afs/super.c b/fs/afs/super.c index 783c68cd1a35..f18911e8d770 100644 --- a/fs/afs/super.c +++ b/fs/afs/super.c @@ -426,7 +426,7 @@ static int afs_set_super(struct super_block *sb, struct fs_context *fc) static int afs_fill_super(struct super_block *sb, struct afs_fs_context *ctx) { struct afs_super_info *as = AFS_FS_S(sb); - struct afs_fid fid; + struct afs_iget_data iget_data; struct inode *inode = NULL; int ret; @@ -451,11 +451,13 @@ static int afs_fill_super(struct super_block *sb, struct afs_fs_context *ctx) } else { sprintf(sb->s_id, "%llu", as->volume->vid); afs_activate_volume(as->volume); - fid.vid = as->volume->vid; - fid.vnode = 1; - fid.vnode_hi = 0; - fid.unique = 1; - inode = afs_iget(sb, ctx->key, &fid, NULL, NULL, NULL, NULL); + iget_data.fid.vid = as->volume->vid; + iget_data.fid.vnode = 1; + iget_data.fid.vnode_hi = 0; + iget_data.fid.unique = 1; + iget_data.cb_v_break = as->volume->cb_v_break; + iget_data.cb_s_break = 0; + inode = afs_iget(sb, ctx->key, &iget_data, NULL, NULL, NULL); } if (IS_ERR(inode)) @@ -677,13 +679,12 @@ static struct inode *afs_alloc_inode(struct super_block *sb) vnode->volume = NULL; vnode->lock_key = NULL; vnode->permit_cache = NULL; - vnode->cb_interest = NULL; + RCU_INIT_POINTER(vnode->cb_interest, NULL); #ifdef CONFIG_AFS_FSCACHE vnode->cache = NULL; #endif vnode->flags = 1 << AFS_VNODE_UNSET; - vnode->cb_type = 0; vnode->lock_state = AFS_VNODE_LOCK_NONE; init_rwsem(&vnode->rmdir_lock); @@ -708,7 +709,7 @@ static void afs_destroy_inode(struct inode *inode) _debug("DESTROY INODE %p", inode); - ASSERTCMP(vnode->cb_interest, ==, NULL); + ASSERTCMP(rcu_access_pointer(vnode->cb_interest), ==, NULL); atomic_dec(&afs_count_active_inodes); } @@ -741,7 +742,7 @@ static int afs_statfs(struct dentry *dentry, struct kstatfs *buf) return PTR_ERR(key); ret = -ERESTARTSYS; - if (afs_begin_vnode_operation(&fc, vnode, key)) { + if (afs_begin_vnode_operation(&fc, vnode, key, true)) { fc.flags |= AFS_FS_CURSOR_NO_VSLEEP; while (afs_select_fileserver(&fc)) { fc.cb_break = afs_calc_vnode_cb_break(vnode); @@ -749,7 +750,6 @@ static int afs_statfs(struct dentry *dentry, struct kstatfs *buf) } afs_check_for_remote_deletion(&fc, fc.vnode); - afs_vnode_commit_status(&fc, vnode, fc.cb_break); ret = afs_end_vnode_operation(&fc); } diff --git a/fs/afs/vl_list.c b/fs/afs/vl_list.c index b4f1a84519b9..61e25010ff33 100644 --- a/fs/afs/vl_list.c +++ b/fs/afs/vl_list.c @@ -232,18 +232,16 @@ struct afs_vlserver_list *afs_extract_vlserver_list(struct afs_cell *cell, if (bs.status > NR__dns_lookup_status) bs.status = NR__dns_lookup_status; + /* See if we can update an old server record */ server = NULL; - if (previous) { - /* See if we can update an old server record */ - for (i = 0; i < previous->nr_servers; i++) { - struct afs_vlserver *p = previous->servers[i].server; - - if (p->name_len == bs.name_len && - p->port == bs.port && - strncasecmp(b, p->name, bs.name_len) == 0) { - server = afs_get_vlserver(p); - break; - } + for (i = 0; i < previous->nr_servers; i++) { + struct afs_vlserver *p = previous->servers[i].server; + + if (p->name_len == bs.name_len && + p->port == bs.port && + strncasecmp(b, p->name, bs.name_len) == 0) { + server = afs_get_vlserver(p); + break; } } diff --git a/fs/afs/vl_probe.c b/fs/afs/vl_probe.c index b05e0de04f42..beb991563939 100644 --- a/fs/afs/vl_probe.c +++ b/fs/afs/vl_probe.c @@ -33,8 +33,8 @@ static bool afs_vl_probe_done(struct afs_vlserver *server) void afs_vlserver_probe_result(struct afs_call *call) { struct afs_addr_list *alist = call->alist; - struct afs_vlserver *server = call->reply[0]; - unsigned int server_index = (long)call->reply[1]; + struct afs_vlserver *server = call->vlserver; + unsigned int server_index = call->server_index; unsigned int index = call->addr_ix; unsigned int rtt = UINT_MAX; bool have_result = false; diff --git a/fs/afs/vl_rotate.c b/fs/afs/vl_rotate.c index 7adde83a0648..3f845489a9f0 100644 --- a/fs/afs/vl_rotate.c +++ b/fs/afs/vl_rotate.c @@ -43,11 +43,29 @@ bool afs_begin_vlserver_operation(struct afs_vl_cursor *vc, struct afs_cell *cel static bool afs_start_vl_iteration(struct afs_vl_cursor *vc) { struct afs_cell *cell = vc->cell; + unsigned int dns_lookup_count; + + if (cell->dns_source == DNS_RECORD_UNAVAILABLE || + cell->dns_expiry <= ktime_get_real_seconds()) { + dns_lookup_count = smp_load_acquire(&cell->dns_lookup_count); + set_bit(AFS_CELL_FL_DO_LOOKUP, &cell->flags); + queue_work(afs_wq, &cell->manager); + + if (cell->dns_source == DNS_RECORD_UNAVAILABLE) { + if (wait_var_event_interruptible( + &cell->dns_lookup_count, + smp_load_acquire(&cell->dns_lookup_count) + != dns_lookup_count) < 0) { + vc->error = -ERESTARTSYS; + return false; + } + } - if (wait_on_bit(&cell->flags, AFS_CELL_FL_NO_LOOKUP_YET, - TASK_INTERRUPTIBLE)) { - vc->error = -ERESTARTSYS; - return false; + /* Status load is ordered after lookup counter load */ + if (cell->dns_source == DNS_RECORD_UNAVAILABLE) { + vc->error = -EDESTADDRREQ; + return false; + } } read_lock(&cell->vl_servers_lock); @@ -55,7 +73,7 @@ static bool afs_start_vl_iteration(struct afs_vl_cursor *vc) rcu_dereference_protected(cell->vl_servers, lockdep_is_held(&cell->vl_servers_lock))); read_unlock(&cell->vl_servers_lock); - if (!vc->server_list || !vc->server_list->nr_servers) + if (!vc->server_list->nr_servers) return false; vc->untried = (1UL << vc->server_list->nr_servers) - 1; diff --git a/fs/afs/vlclient.c b/fs/afs/vlclient.c index dd9ba4e96fb3..3d4b9836a2e2 100644 --- a/fs/afs/vlclient.c +++ b/fs/afs/vlclient.c @@ -34,7 +34,7 @@ static int afs_deliver_vl_get_entry_by_name_u(struct afs_call *call) /* unmarshall the reply once we've received all of it */ uvldb = call->buffer; - entry = call->reply[0]; + entry = call->ret_vldb; nr_servers = ntohl(uvldb->nServers); if (nr_servers > AFS_NMAXNSERVERS) @@ -110,7 +110,7 @@ static int afs_deliver_vl_get_entry_by_name_u(struct afs_call *call) static void afs_destroy_vl_get_entry_by_name_u(struct afs_call *call) { - kfree(call->reply[0]); + kfree(call->ret_vldb); afs_flat_call_destructor(call); } @@ -155,8 +155,8 @@ struct afs_vldb_entry *afs_vl_get_entry_by_name_u(struct afs_vl_cursor *vc, } call->key = vc->key; - call->reply[0] = entry; - call->ret_reply0 = true; + call->ret_vldb = entry; + call->max_lifespan = AFS_VL_MAX_LIFESPAN; /* Marshall the parameters */ bp = call->request; @@ -214,7 +214,7 @@ static int afs_deliver_vl_get_addrs_u(struct afs_call *call) if (!alist) return -ENOMEM; alist->version = uniquifier; - call->reply[0] = alist; + call->ret_alist = alist; call->count = count; call->count2 = nentries; call->unmarshall++; @@ -229,7 +229,7 @@ static int afs_deliver_vl_get_addrs_u(struct afs_call *call) if (ret < 0) return ret; - alist = call->reply[0]; + alist = call->ret_alist; bp = call->buffer; count = min(call->count, 4U); for (i = 0; i < count; i++) @@ -249,8 +249,7 @@ static int afs_deliver_vl_get_addrs_u(struct afs_call *call) static void afs_vl_get_addrs_u_destructor(struct afs_call *call) { - afs_put_server(call->net, (struct afs_server *)call->reply[0]); - kfree(call->reply[1]); + afs_put_addrlist(call->ret_alist); return afs_flat_call_destructor(call); } @@ -287,8 +286,8 @@ struct afs_addr_list *afs_vl_get_addrs_u(struct afs_vl_cursor *vc, return ERR_PTR(-ENOMEM); call->key = vc->key; - call->reply[0] = NULL; - call->ret_reply0 = true; + call->ret_alist = NULL; + call->max_lifespan = AFS_VL_MAX_LIFESPAN; /* Marshall the parameters */ bp = call->request; @@ -358,9 +357,7 @@ static int afs_deliver_vl_get_capabilities(struct afs_call *call) static void afs_destroy_vl_get_capabilities(struct afs_call *call) { - struct afs_vlserver *server = call->reply[0]; - - afs_put_vlserver(call->net, server); + afs_put_vlserver(call->net, call->vlserver); afs_flat_call_destructor(call); } @@ -398,11 +395,11 @@ struct afs_call *afs_vl_get_capabilities(struct afs_net *net, return ERR_PTR(-ENOMEM); call->key = key; - call->reply[0] = afs_get_vlserver(server); - call->reply[1] = (void *)(long)server_index; + call->vlserver = afs_get_vlserver(server); + call->server_index = server_index; call->upgrade = true; - call->want_reply_time = true; call->async = true; + call->max_lifespan = AFS_PROBE_MAX_LIFESPAN; /* marshall the parameters */ bp = call->request; @@ -460,7 +457,7 @@ static int afs_deliver_yfsvl_get_endpoints(struct afs_call *call) if (!alist) return -ENOMEM; alist->version = uniquifier; - call->reply[0] = alist; + call->ret_alist = alist; if (call->count == 0) goto extract_volendpoints; @@ -488,7 +485,7 @@ static int afs_deliver_yfsvl_get_endpoints(struct afs_call *call) if (ret < 0) return ret; - alist = call->reply[0]; + alist = call->ret_alist; bp = call->buffer; switch (call->count2) { case YFS_ENDPOINT_IPV4: @@ -609,7 +606,6 @@ static int afs_deliver_yfsvl_get_endpoints(struct afs_call *call) break; } - alist = call->reply[0]; _leave(" = 0 [done]"); return 0; } @@ -644,8 +640,8 @@ struct afs_addr_list *afs_yfsvl_get_endpoints(struct afs_vl_cursor *vc, return ERR_PTR(-ENOMEM); call->key = vc->key; - call->reply[0] = NULL; - call->ret_reply0 = true; + call->ret_alist = NULL; + call->max_lifespan = AFS_VL_MAX_LIFESPAN; /* Marshall the parameters */ bp = call->request; diff --git a/fs/afs/write.c b/fs/afs/write.c index 0122d7445fba..8bcab95f1127 100644 --- a/fs/afs/write.c +++ b/fs/afs/write.c @@ -314,6 +314,46 @@ static void afs_redirty_pages(struct writeback_control *wbc, } /* + * completion of write to server + */ +static void afs_pages_written_back(struct afs_vnode *vnode, + pgoff_t first, pgoff_t last) +{ + struct pagevec pv; + unsigned long priv; + unsigned count, loop; + + _enter("{%llx:%llu},{%lx-%lx}", + vnode->fid.vid, vnode->fid.vnode, first, last); + + pagevec_init(&pv); + + do { + _debug("done %lx-%lx", first, last); + + count = last - first + 1; + if (count > PAGEVEC_SIZE) + count = PAGEVEC_SIZE; + pv.nr = find_get_pages_contig(vnode->vfs_inode.i_mapping, + first, count, pv.pages); + ASSERTCMP(pv.nr, ==, count); + + for (loop = 0; loop < count; loop++) { + priv = page_private(pv.pages[loop]); + trace_afs_page_dirty(vnode, tracepoint_string("clear"), + pv.pages[loop]->index, priv); + set_page_private(pv.pages[loop], 0); + end_page_writeback(pv.pages[loop]); + } + first += count; + __pagevec_release(&pv); + } while (first <= last); + + afs_prune_wb_keys(vnode); + _leave(""); +} + +/* * write to a file */ static int afs_store_data(struct address_space *mapping, @@ -322,6 +362,7 @@ static int afs_store_data(struct address_space *mapping, { struct afs_vnode *vnode = AFS_FS_I(mapping->host); struct afs_fs_cursor fc; + struct afs_status_cb *scb; struct afs_wb_key *wbk = NULL; struct list_head *p; int ret = -ENOKEY, ret2; @@ -333,6 +374,10 @@ static int afs_store_data(struct address_space *mapping, vnode->fid.unique, first, last, offset, to); + scb = kzalloc(sizeof(struct afs_status_cb), GFP_NOFS); + if (!scb) + return -ENOMEM; + spin_lock(&vnode->wb_lock); p = vnode->wb_keys.next; @@ -351,6 +396,7 @@ try_next_key: spin_unlock(&vnode->wb_lock); afs_put_wb_key(wbk); + kfree(scb); _leave(" = %d [no keys]", ret); return ret; @@ -361,14 +407,19 @@ found_key: _debug("USE WB KEY %u", key_serial(wbk->key)); ret = -ERESTARTSYS; - if (afs_begin_vnode_operation(&fc, vnode, wbk->key)) { + if (afs_begin_vnode_operation(&fc, vnode, wbk->key, false)) { + afs_dataversion_t data_version = vnode->status.data_version + 1; + while (afs_select_fileserver(&fc)) { fc.cb_break = afs_calc_vnode_cb_break(vnode); - afs_fs_store_data(&fc, mapping, first, last, offset, to); + afs_fs_store_data(&fc, mapping, first, last, offset, to, scb); } - afs_check_for_remote_deletion(&fc, fc.vnode); - afs_vnode_commit_status(&fc, vnode, fc.cb_break); + afs_check_for_remote_deletion(&fc, vnode); + afs_vnode_commit_status(&fc, vnode, fc.cb_break, + &data_version, scb); + if (fc.ac.error == 0) + afs_pages_written_back(vnode, first, last); ret = afs_end_vnode_operation(&fc); } @@ -393,6 +444,7 @@ found_key: } afs_put_wb_key(wbk); + kfree(scb); _leave(" = %d", ret); return ret; } @@ -679,46 +731,6 @@ int afs_writepages(struct address_space *mapping, } /* - * completion of write to server - */ -void afs_pages_written_back(struct afs_vnode *vnode, struct afs_call *call) -{ - struct pagevec pv; - unsigned long priv; - unsigned count, loop; - pgoff_t first = call->first, last = call->last; - - _enter("{%llx:%llu},{%lx-%lx}", - vnode->fid.vid, vnode->fid.vnode, first, last); - - pagevec_init(&pv); - - do { - _debug("done %lx-%lx", first, last); - - count = last - first + 1; - if (count > PAGEVEC_SIZE) - count = PAGEVEC_SIZE; - pv.nr = find_get_pages_contig(vnode->vfs_inode.i_mapping, - first, count, pv.pages); - ASSERTCMP(pv.nr, ==, count); - - for (loop = 0; loop < count; loop++) { - priv = page_private(pv.pages[loop]); - trace_afs_page_dirty(vnode, tracepoint_string("clear"), - pv.pages[loop]->index, priv); - set_page_private(pv.pages[loop], 0); - end_page_writeback(pv.pages[loop]); - } - first += count; - __pagevec_release(&pv); - } while (first <= last); - - afs_prune_wb_keys(vnode); - _leave(""); -} - -/* * write to an AFS file */ ssize_t afs_file_write(struct kiocb *iocb, struct iov_iter *from) diff --git a/fs/afs/xattr.c b/fs/afs/xattr.c index c81f85003fc7..17f58fea7ec1 100644 --- a/fs/afs/xattr.c +++ b/fs/afs/xattr.c @@ -47,40 +47,52 @@ static int afs_xattr_get_acl(const struct xattr_handler *handler, void *buffer, size_t size) { struct afs_fs_cursor fc; + struct afs_status_cb *scb; struct afs_vnode *vnode = AFS_FS_I(inode); struct afs_acl *acl = NULL; struct key *key; - int ret; + int ret = -ENOMEM; + + scb = kzalloc(sizeof(struct afs_status_cb), GFP_NOFS); + if (!scb) + goto error; key = afs_request_key(vnode->volume->cell); - if (IS_ERR(key)) - return PTR_ERR(key); + if (IS_ERR(key)) { + ret = PTR_ERR(key); + goto error_scb; + } ret = -ERESTARTSYS; - if (afs_begin_vnode_operation(&fc, vnode, key)) { + if (afs_begin_vnode_operation(&fc, vnode, key, true)) { + afs_dataversion_t data_version = vnode->status.data_version; + while (afs_select_fileserver(&fc)) { fc.cb_break = afs_calc_vnode_cb_break(vnode); - acl = afs_fs_fetch_acl(&fc); + acl = afs_fs_fetch_acl(&fc, scb); } afs_check_for_remote_deletion(&fc, fc.vnode); - afs_vnode_commit_status(&fc, vnode, fc.cb_break); + afs_vnode_commit_status(&fc, vnode, fc.cb_break, + &data_version, scb); ret = afs_end_vnode_operation(&fc); } if (ret == 0) { ret = acl->size; if (size > 0) { - ret = -ERANGE; - if (acl->size > size) - return -ERANGE; - memcpy(buffer, acl->data, acl->size); - ret = acl->size; + if (acl->size <= size) + memcpy(buffer, acl->data, acl->size); + else + ret = -ERANGE; } kfree(acl); } key_put(key); +error_scb: + kfree(scb); +error: return ret; } @@ -93,41 +105,53 @@ static int afs_xattr_set_acl(const struct xattr_handler *handler, const void *buffer, size_t size, int flags) { struct afs_fs_cursor fc; + struct afs_status_cb *scb; struct afs_vnode *vnode = AFS_FS_I(inode); struct afs_acl *acl = NULL; struct key *key; - int ret; + int ret = -ENOMEM; if (flags == XATTR_CREATE) return -EINVAL; - key = afs_request_key(vnode->volume->cell); - if (IS_ERR(key)) - return PTR_ERR(key); + scb = kzalloc(sizeof(struct afs_status_cb), GFP_NOFS); + if (!scb) + goto error; acl = kmalloc(sizeof(*acl) + size, GFP_KERNEL); - if (!acl) { - key_put(key); - return -ENOMEM; + if (!acl) + goto error_scb; + + key = afs_request_key(vnode->volume->cell); + if (IS_ERR(key)) { + ret = PTR_ERR(key); + goto error_acl; } acl->size = size; memcpy(acl->data, buffer, size); ret = -ERESTARTSYS; - if (afs_begin_vnode_operation(&fc, vnode, key)) { + if (afs_begin_vnode_operation(&fc, vnode, key, true)) { + afs_dataversion_t data_version = vnode->status.data_version; + while (afs_select_fileserver(&fc)) { fc.cb_break = afs_calc_vnode_cb_break(vnode); - afs_fs_store_acl(&fc, acl); + afs_fs_store_acl(&fc, acl, scb); } afs_check_for_remote_deletion(&fc, fc.vnode); - afs_vnode_commit_status(&fc, vnode, fc.cb_break); + afs_vnode_commit_status(&fc, vnode, fc.cb_break, + &data_version, scb); ret = afs_end_vnode_operation(&fc); } - kfree(acl); key_put(key); +error_acl: + kfree(acl); +error_scb: + kfree(scb); +error: return ret; } @@ -146,12 +170,12 @@ static int afs_xattr_get_yfs(const struct xattr_handler *handler, void *buffer, size_t size) { struct afs_fs_cursor fc; + struct afs_status_cb *scb; struct afs_vnode *vnode = AFS_FS_I(inode); struct yfs_acl *yacl = NULL; struct key *key; - unsigned int flags = 0; char buf[16], *data; - int which = 0, dsize, ret; + int which = 0, dsize, ret = -ENOMEM; if (strcmp(name, "acl") == 0) which = 0; @@ -164,65 +188,81 @@ static int afs_xattr_get_yfs(const struct xattr_handler *handler, else return -EOPNOTSUPP; + yacl = kzalloc(sizeof(struct yfs_acl), GFP_KERNEL); + if (!yacl) + goto error; + if (which == 0) - flags |= YFS_ACL_WANT_ACL; + yacl->flags |= YFS_ACL_WANT_ACL; else if (which == 3) - flags |= YFS_ACL_WANT_VOL_ACL; + yacl->flags |= YFS_ACL_WANT_VOL_ACL; + + scb = kzalloc(sizeof(struct afs_status_cb), GFP_NOFS); + if (!scb) + goto error_yacl; key = afs_request_key(vnode->volume->cell); - if (IS_ERR(key)) - return PTR_ERR(key); + if (IS_ERR(key)) { + ret = PTR_ERR(key); + goto error_scb; + } ret = -ERESTARTSYS; - if (afs_begin_vnode_operation(&fc, vnode, key)) { + if (afs_begin_vnode_operation(&fc, vnode, key, true)) { + afs_dataversion_t data_version = vnode->status.data_version; + while (afs_select_fileserver(&fc)) { fc.cb_break = afs_calc_vnode_cb_break(vnode); - yacl = yfs_fs_fetch_opaque_acl(&fc, flags); + yfs_fs_fetch_opaque_acl(&fc, yacl, scb); } afs_check_for_remote_deletion(&fc, fc.vnode); - afs_vnode_commit_status(&fc, vnode, fc.cb_break); + afs_vnode_commit_status(&fc, vnode, fc.cb_break, + &data_version, scb); ret = afs_end_vnode_operation(&fc); } - if (ret == 0) { - switch (which) { - case 0: - data = yacl->acl->data; - dsize = yacl->acl->size; - break; - case 1: - data = buf; - dsize = snprintf(buf, sizeof(buf), "%u", - yacl->inherit_flag); - break; - case 2: - data = buf; - dsize = snprintf(buf, sizeof(buf), "%u", - yacl->num_cleaned); - break; - case 3: - data = yacl->vol_acl->data; - dsize = yacl->vol_acl->size; - break; - default: - ret = -EOPNOTSUPP; - goto out; - } + if (ret < 0) + goto error_key; + + switch (which) { + case 0: + data = yacl->acl->data; + dsize = yacl->acl->size; + break; + case 1: + data = buf; + dsize = snprintf(buf, sizeof(buf), "%u", yacl->inherit_flag); + break; + case 2: + data = buf; + dsize = snprintf(buf, sizeof(buf), "%u", yacl->num_cleaned); + break; + case 3: + data = yacl->vol_acl->data; + dsize = yacl->vol_acl->size; + break; + default: + ret = -EOPNOTSUPP; + goto error_key; + } - ret = dsize; - if (size > 0) { - if (dsize > size) { - ret = -ERANGE; - goto out; - } - memcpy(buffer, data, dsize); + ret = dsize; + if (size > 0) { + if (dsize > size) { + ret = -ERANGE; + goto error_key; } + memcpy(buffer, data, dsize); } -out: - yfs_free_opaque_acl(yacl); +error_key: key_put(key); +error_scb: + kfree(scb); +error_yacl: + yfs_free_opaque_acl(yacl); +error: return ret; } @@ -235,42 +275,54 @@ static int afs_xattr_set_yfs(const struct xattr_handler *handler, const void *buffer, size_t size, int flags) { struct afs_fs_cursor fc; + struct afs_status_cb *scb; struct afs_vnode *vnode = AFS_FS_I(inode); struct afs_acl *acl = NULL; struct key *key; - int ret; + int ret = -ENOMEM; if (flags == XATTR_CREATE || strcmp(name, "acl") != 0) return -EINVAL; - key = afs_request_key(vnode->volume->cell); - if (IS_ERR(key)) - return PTR_ERR(key); + scb = kzalloc(sizeof(struct afs_status_cb), GFP_NOFS); + if (!scb) + goto error; acl = kmalloc(sizeof(*acl) + size, GFP_KERNEL); - if (!acl) { - key_put(key); - return -ENOMEM; - } + if (!acl) + goto error_scb; acl->size = size; memcpy(acl->data, buffer, size); + key = afs_request_key(vnode->volume->cell); + if (IS_ERR(key)) { + ret = PTR_ERR(key); + goto error_acl; + } + ret = -ERESTARTSYS; - if (afs_begin_vnode_operation(&fc, vnode, key)) { + if (afs_begin_vnode_operation(&fc, vnode, key, true)) { + afs_dataversion_t data_version = vnode->status.data_version; + while (afs_select_fileserver(&fc)) { fc.cb_break = afs_calc_vnode_cb_break(vnode); - yfs_fs_store_opaque_acl2(&fc, acl); + yfs_fs_store_opaque_acl2(&fc, acl, scb); } afs_check_for_remote_deletion(&fc, fc.vnode); - afs_vnode_commit_status(&fc, vnode, fc.cb_break); + afs_vnode_commit_status(&fc, vnode, fc.cb_break, + &data_version, scb); ret = afs_end_vnode_operation(&fc); } +error_acl: kfree(acl); key_put(key); +error_scb: + kfree(scb); +error: return ret; } diff --git a/fs/afs/yfsclient.c b/fs/afs/yfsclient.c index 6cf7d161baa1..10de675dc6fc 100644 --- a/fs/afs/yfsclient.c +++ b/fs/afs/yfsclient.c @@ -183,24 +183,19 @@ static void xdr_dump_bad(const __be32 *bp) /* * Decode a YFSFetchStatus block */ -static int xdr_decode_YFSFetchStatus(struct afs_call *call, - const __be32 **_bp, - struct afs_file_status *status, - struct afs_vnode *vnode, - const afs_dataversion_t *expected_version, - struct afs_read *read_req) +static int xdr_decode_YFSFetchStatus(const __be32 **_bp, + struct afs_call *call, + struct afs_status_cb *scb) { const struct yfs_xdr_YFSFetchStatus *xdr = (const void *)*_bp; + struct afs_file_status *status = &scb->status; u32 type; - u8 flags = 0; status->abort_code = ntohl(xdr->abort_code); if (status->abort_code != 0) { - if (vnode && status->abort_code == VNOVNODE) { - set_bit(AFS_VNODE_DELETED, &vnode->flags); + if (status->abort_code == VNOVNODE) status->nlink = 0; - __afs_break_callback(vnode); - } + scb->have_error = true; return 0; } @@ -209,77 +204,28 @@ static int xdr_decode_YFSFetchStatus(struct afs_call *call, case AFS_FTYPE_FILE: case AFS_FTYPE_DIR: case AFS_FTYPE_SYMLINK: - if (type != status->type && - vnode && - !test_bit(AFS_VNODE_UNSET, &vnode->flags)) { - pr_warning("Vnode %llx:%llx:%x changed type %u to %u\n", - vnode->fid.vid, - vnode->fid.vnode, - vnode->fid.unique, - status->type, type); - goto bad; - } status->type = type; break; default: goto bad; } -#define EXTRACT_M4(FIELD) \ - do { \ - u32 x = ntohl(xdr->FIELD); \ - if (status->FIELD != x) { \ - flags |= AFS_VNODE_META_CHANGED; \ - status->FIELD = x; \ - } \ - } while (0) - -#define EXTRACT_M8(FIELD) \ - do { \ - u64 x = xdr_to_u64(xdr->FIELD); \ - if (status->FIELD != x) { \ - flags |= AFS_VNODE_META_CHANGED; \ - status->FIELD = x; \ - } \ - } while (0) - -#define EXTRACT_D8(FIELD) \ - do { \ - u64 x = xdr_to_u64(xdr->FIELD); \ - if (status->FIELD != x) { \ - flags |= AFS_VNODE_DATA_CHANGED; \ - status->FIELD = x; \ - } \ - } while (0) - - EXTRACT_M4(nlink); - EXTRACT_D8(size); - EXTRACT_D8(data_version); - EXTRACT_M8(author); - EXTRACT_M8(owner); - EXTRACT_M8(group); - EXTRACT_M4(mode); - EXTRACT_M4(caller_access); /* call ticket dependent */ - EXTRACT_M4(anon_access); - - status->mtime_client = xdr_to_time(xdr->mtime_client); - status->mtime_server = xdr_to_time(xdr->mtime_server); - status->lock_count = ntohl(xdr->lock_count); - - if (read_req) { - read_req->data_version = status->data_version; - read_req->file_size = status->size; - } + status->nlink = ntohl(xdr->nlink); + status->author = xdr_to_u64(xdr->author); + status->owner = xdr_to_u64(xdr->owner); + status->caller_access = ntohl(xdr->caller_access); /* Ticket dependent */ + status->anon_access = ntohl(xdr->anon_access); + status->mode = ntohl(xdr->mode) & S_IALLUGO; + status->group = xdr_to_u64(xdr->group); + status->lock_count = ntohl(xdr->lock_count); + + status->mtime_client = xdr_to_time(xdr->mtime_client); + status->mtime_server = xdr_to_time(xdr->mtime_server); + status->size = xdr_to_u64(xdr->size); + status->data_version = xdr_to_u64(xdr->data_version); + scb->have_status = true; *_bp += xdr_size(xdr); - - if (vnode) { - if (test_bit(AFS_VNODE_UNSET, &vnode->flags)) - flags |= AFS_VNODE_NOT_YET_SET; - afs_update_inode_from_status(vnode, status, expected_version, - flags); - } - return 0; bad: @@ -288,73 +234,20 @@ bad: } /* - * Decode the file status. We need to lock the target vnode if we're going to - * update its status so that stat() sees the attributes update atomically. - */ -static int yfs_decode_status(struct afs_call *call, - const __be32 **_bp, - struct afs_file_status *status, - struct afs_vnode *vnode, - const afs_dataversion_t *expected_version, - struct afs_read *read_req) -{ - int ret; - - if (!vnode) - return xdr_decode_YFSFetchStatus(call, _bp, status, vnode, - expected_version, read_req); - - write_seqlock(&vnode->cb_lock); - ret = xdr_decode_YFSFetchStatus(call, _bp, status, vnode, - expected_version, read_req); - write_sequnlock(&vnode->cb_lock); - return ret; -} - -/* * Decode a YFSCallBack block */ -static void xdr_decode_YFSCallBack(struct afs_call *call, - struct afs_vnode *vnode, - const __be32 **_bp) -{ - struct yfs_xdr_YFSCallBack *xdr = (void *)*_bp; - struct afs_cb_interest *old, *cbi = call->cbi; - u64 cb_expiry; - - write_seqlock(&vnode->cb_lock); - - if (!afs_cb_is_broken(call->cb_break, vnode, cbi)) { - cb_expiry = xdr_to_u64(xdr->expiration_time); - do_div(cb_expiry, 10 * 1000 * 1000); - vnode->cb_version = ntohl(xdr->version); - vnode->cb_type = ntohl(xdr->type); - vnode->cb_expires_at = cb_expiry + ktime_get_real_seconds(); - old = vnode->cb_interest; - if (old != call->cbi) { - vnode->cb_interest = cbi; - cbi = old; - } - set_bit(AFS_VNODE_CB_PROMISED, &vnode->flags); - } - - write_sequnlock(&vnode->cb_lock); - call->cbi = cbi; - *_bp += xdr_size(xdr); -} - -static void xdr_decode_YFSCallBack_raw(const __be32 **_bp, - struct afs_callback *cb) +static void xdr_decode_YFSCallBack(const __be32 **_bp, + struct afs_call *call, + struct afs_status_cb *scb) { struct yfs_xdr_YFSCallBack *x = (void *)*_bp; - u64 cb_expiry; - - cb_expiry = xdr_to_u64(x->expiration_time); - do_div(cb_expiry, 10 * 1000 * 1000); - cb->version = ntohl(x->version); - cb->type = ntohl(x->type); - cb->expires_at = cb_expiry + ktime_get_real_seconds(); + struct afs_callback *cb = &scb->callback; + ktime_t cb_expiry; + cb_expiry = call->reply_time; + cb_expiry = ktime_add(cb_expiry, xdr_to_u64(x->expiration_time) * 100); + cb->expires_at = ktime_divns(cb_expiry, NSEC_PER_SEC); + scb->have_cb = true; *_bp += xdr_size(x); } @@ -442,11 +335,10 @@ static void xdr_decode_YFSFetchVolumeStatus(const __be32 **_bp, } /* - * deliver reply data to an FS.FetchStatus + * Deliver a reply that's a status, callback and volsync. */ -static int yfs_deliver_fs_fetch_status_vnode(struct afs_call *call) +static int yfs_deliver_fs_status_cb_and_volsync(struct afs_call *call) { - struct afs_vnode *vnode = call->reply[0]; const __be32 *bp; int ret; @@ -454,16 +346,36 @@ static int yfs_deliver_fs_fetch_status_vnode(struct afs_call *call) if (ret < 0) return ret; - _enter("{%llx:%llu}", vnode->fid.vid, vnode->fid.vnode); - /* unmarshall the reply once we've received all of it */ bp = call->buffer; - ret = yfs_decode_status(call, &bp, &vnode->status, vnode, - &call->expected_version, NULL); + ret = xdr_decode_YFSFetchStatus(&bp, call, call->out_scb); if (ret < 0) return ret; - xdr_decode_YFSCallBack(call, vnode, &bp); - xdr_decode_YFSVolSync(&bp, call->reply[1]); + xdr_decode_YFSCallBack(&bp, call, call->out_scb); + xdr_decode_YFSVolSync(&bp, call->out_volsync); + + _leave(" = 0 [done]"); + return 0; +} + +/* + * Deliver reply data to operations that just return a file status and a volume + * sync record. + */ +static int yfs_deliver_status_and_volsync(struct afs_call *call) +{ + const __be32 *bp; + int ret; + + ret = afs_transfer_reply(call); + if (ret < 0) + return ret; + + bp = call->buffer; + ret = xdr_decode_YFSFetchStatus(&bp, call, call->out_scb); + if (ret < 0) + return ret; + xdr_decode_YFSVolSync(&bp, call->out_volsync); _leave(" = 0 [done]"); return 0; @@ -475,15 +387,15 @@ static int yfs_deliver_fs_fetch_status_vnode(struct afs_call *call) static const struct afs_call_type yfs_RXYFSFetchStatus_vnode = { .name = "YFS.FetchStatus(vnode)", .op = yfs_FS_FetchStatus, - .deliver = yfs_deliver_fs_fetch_status_vnode, + .deliver = yfs_deliver_fs_status_cb_and_volsync, .destructor = afs_flat_call_destructor, }; /* * Fetch the status information for a file. */ -int yfs_fs_fetch_file_status(struct afs_fs_cursor *fc, struct afs_volsync *volsync, - bool new_inode) +int yfs_fs_fetch_file_status(struct afs_fs_cursor *fc, struct afs_status_cb *scb, + struct afs_volsync *volsync) { struct afs_vnode *vnode = fc->vnode; struct afs_call *call; @@ -505,9 +417,8 @@ int yfs_fs_fetch_file_status(struct afs_fs_cursor *fc, struct afs_volsync *volsy } call->key = fc->key; - call->reply[0] = vnode; - call->reply[1] = volsync; - call->expected_version = new_inode ? 1 : vnode->status.data_version; + call->out_scb = scb; + call->out_volsync = volsync; /* marshall the parameters */ bp = call->request; @@ -516,9 +427,9 @@ int yfs_fs_fetch_file_status(struct afs_fs_cursor *fc, struct afs_volsync *volsy bp = xdr_encode_YFSFid(bp, &vnode->fid); yfs_check_req(call, bp); - call->cb_break = fc->cb_break; afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_call(call, &vnode->fid); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } @@ -528,8 +439,7 @@ int yfs_fs_fetch_file_status(struct afs_fs_cursor *fc, struct afs_volsync *volsy */ static int yfs_deliver_fs_fetch_data64(struct afs_call *call) { - struct afs_vnode *vnode = call->reply[0]; - struct afs_read *req = call->reply[2]; + struct afs_read *req = call->read_request; const __be32 *bp; unsigned int size; int ret; @@ -586,7 +496,7 @@ static int yfs_deliver_fs_fetch_data64(struct afs_call *call) if (req->offset == PAGE_SIZE) { req->offset = 0; if (req->page_done) - req->page_done(call, req); + req->page_done(req); req->index++; if (req->remain > 0) goto begin_page; @@ -623,12 +533,14 @@ static int yfs_deliver_fs_fetch_data64(struct afs_call *call) return ret; bp = call->buffer; - ret = yfs_decode_status(call, &bp, &vnode->status, vnode, - &vnode->status.data_version, req); + ret = xdr_decode_YFSFetchStatus(&bp, call, call->out_scb); if (ret < 0) return ret; - xdr_decode_YFSCallBack(call, vnode, &bp); - xdr_decode_YFSVolSync(&bp, call->reply[1]); + xdr_decode_YFSCallBack(&bp, call, call->out_scb); + xdr_decode_YFSVolSync(&bp, call->out_volsync); + + req->data_version = call->out_scb->status.data_version; + req->file_size = call->out_scb->status.size; call->unmarshall++; @@ -642,7 +554,7 @@ static int yfs_deliver_fs_fetch_data64(struct afs_call *call) zero_user_segment(req->pages[req->index], req->offset, PAGE_SIZE); if (req->page_done) - req->page_done(call, req); + req->page_done(req); req->offset = 0; } @@ -652,9 +564,7 @@ static int yfs_deliver_fs_fetch_data64(struct afs_call *call) static void yfs_fetch_data_destructor(struct afs_call *call) { - struct afs_read *req = call->reply[2]; - - afs_put_read(req); + afs_put_read(call->read_request); afs_flat_call_destructor(call); } @@ -671,7 +581,8 @@ static const struct afs_call_type yfs_RXYFSFetchData64 = { /* * Fetch data from a file. */ -int yfs_fs_fetch_data(struct afs_fs_cursor *fc, struct afs_read *req) +int yfs_fs_fetch_data(struct afs_fs_cursor *fc, struct afs_status_cb *scb, + struct afs_read *req) { struct afs_vnode *vnode = fc->vnode; struct afs_call *call; @@ -693,11 +604,9 @@ int yfs_fs_fetch_data(struct afs_fs_cursor *fc, struct afs_read *req) return -ENOMEM; call->key = fc->key; - call->reply[0] = vnode; - call->reply[1] = NULL; /* volsync */ - call->reply[2] = req; - call->expected_version = vnode->status.data_version; - call->want_reply_time = true; + call->out_scb = scb; + call->out_volsync = NULL; + call->read_request = req; /* marshall the parameters */ bp = call->request; @@ -709,9 +618,9 @@ int yfs_fs_fetch_data(struct afs_fs_cursor *fc, struct afs_read *req) yfs_check_req(call, bp); refcount_inc(&req->usage); - call->cb_break = fc->cb_break; afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_call(call, &vnode->fid); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } @@ -721,7 +630,6 @@ int yfs_fs_fetch_data(struct afs_fs_cursor *fc, struct afs_read *req) */ static int yfs_deliver_fs_create_vnode(struct afs_call *call) { - struct afs_vnode *vnode = call->reply[0]; const __be32 *bp; int ret; @@ -733,16 +641,15 @@ static int yfs_deliver_fs_create_vnode(struct afs_call *call) /* unmarshall the reply once we've received all of it */ bp = call->buffer; - xdr_decode_YFSFid(&bp, call->reply[1]); - ret = yfs_decode_status(call, &bp, call->reply[2], NULL, NULL, NULL); + xdr_decode_YFSFid(&bp, call->out_fid); + ret = xdr_decode_YFSFetchStatus(&bp, call, call->out_scb); if (ret < 0) return ret; - ret = yfs_decode_status(call, &bp, &vnode->status, vnode, - &call->expected_version, NULL); + ret = xdr_decode_YFSFetchStatus(&bp, call, call->out_dir_scb); if (ret < 0) return ret; - xdr_decode_YFSCallBack_raw(&bp, call->reply[3]); - xdr_decode_YFSVolSync(&bp, NULL); + xdr_decode_YFSCallBack(&bp, call, call->out_scb); + xdr_decode_YFSVolSync(&bp, call->out_volsync); _leave(" = 0 [done]"); return 0; @@ -764,14 +671,13 @@ static const struct afs_call_type afs_RXFSCreateFile = { int yfs_fs_create_file(struct afs_fs_cursor *fc, const char *name, umode_t mode, - u64 current_data_version, + struct afs_status_cb *dvnode_scb, struct afs_fid *newfid, - struct afs_file_status *newstatus, - struct afs_callback *newcb) + struct afs_status_cb *new_scb) { - struct afs_vnode *vnode = fc->vnode; + struct afs_vnode *dvnode = fc->vnode; struct afs_call *call; - struct afs_net *net = afs_v2net(vnode); + struct afs_net *net = afs_v2net(dvnode); size_t namesz, reqsz, rplsz; __be32 *bp; @@ -795,24 +701,23 @@ int yfs_fs_create_file(struct afs_fs_cursor *fc, return -ENOMEM; call->key = fc->key; - call->reply[0] = vnode; - call->reply[1] = newfid; - call->reply[2] = newstatus; - call->reply[3] = newcb; - call->expected_version = current_data_version + 1; + call->out_dir_scb = dvnode_scb; + call->out_fid = newfid; + call->out_scb = new_scb; /* marshall the parameters */ bp = call->request; bp = xdr_encode_u32(bp, YFSCREATEFILE); bp = xdr_encode_u32(bp, 0); /* RPC flags */ - bp = xdr_encode_YFSFid(bp, &vnode->fid); + bp = xdr_encode_YFSFid(bp, &dvnode->fid); bp = xdr_encode_string(bp, name, namesz); bp = xdr_encode_YFSStoreStatus_mode(bp, mode); bp = xdr_encode_u32(bp, yfs_LockNone); /* ViceLockType */ yfs_check_req(call, bp); afs_use_fs_server(call, fc->cbi); - trace_afs_make_fs_call1(call, &vnode->fid, name); + trace_afs_make_fs_call1(call, &dvnode->fid, name); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } @@ -830,14 +735,13 @@ static const struct afs_call_type yfs_RXFSMakeDir = { int yfs_fs_make_dir(struct afs_fs_cursor *fc, const char *name, umode_t mode, - u64 current_data_version, + struct afs_status_cb *dvnode_scb, struct afs_fid *newfid, - struct afs_file_status *newstatus, - struct afs_callback *newcb) + struct afs_status_cb *new_scb) { - struct afs_vnode *vnode = fc->vnode; + struct afs_vnode *dvnode = fc->vnode; struct afs_call *call; - struct afs_net *net = afs_v2net(vnode); + struct afs_net *net = afs_v2net(dvnode); size_t namesz, reqsz, rplsz; __be32 *bp; @@ -860,23 +764,22 @@ int yfs_fs_make_dir(struct afs_fs_cursor *fc, return -ENOMEM; call->key = fc->key; - call->reply[0] = vnode; - call->reply[1] = newfid; - call->reply[2] = newstatus; - call->reply[3] = newcb; - call->expected_version = current_data_version + 1; + call->out_dir_scb = dvnode_scb; + call->out_fid = newfid; + call->out_scb = new_scb; /* marshall the parameters */ bp = call->request; bp = xdr_encode_u32(bp, YFSMAKEDIR); bp = xdr_encode_u32(bp, 0); /* RPC flags */ - bp = xdr_encode_YFSFid(bp, &vnode->fid); + bp = xdr_encode_YFSFid(bp, &dvnode->fid); bp = xdr_encode_string(bp, name, namesz); bp = xdr_encode_YFSStoreStatus_mode(bp, mode); yfs_check_req(call, bp); afs_use_fs_server(call, fc->cbi); - trace_afs_make_fs_call1(call, &vnode->fid, name); + trace_afs_make_fs_call1(call, &dvnode->fid, name); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } @@ -886,8 +789,6 @@ int yfs_fs_make_dir(struct afs_fs_cursor *fc, */ static int yfs_deliver_fs_remove_file2(struct afs_call *call) { - struct afs_vnode *dvnode = call->reply[0]; - struct afs_vnode *vnode = call->reply[1]; struct afs_fid fid; const __be32 *bp; int ret; @@ -898,20 +799,18 @@ static int yfs_deliver_fs_remove_file2(struct afs_call *call) if (ret < 0) return ret; - /* unmarshall the reply once we've received all of it */ bp = call->buffer; - ret = yfs_decode_status(call, &bp, &dvnode->status, dvnode, - &call->expected_version, NULL); + ret = xdr_decode_YFSFetchStatus(&bp, call, call->out_dir_scb); if (ret < 0) return ret; xdr_decode_YFSFid(&bp, &fid); - ret = yfs_decode_status(call, &bp, &vnode->status, vnode, NULL, NULL); + ret = xdr_decode_YFSFetchStatus(&bp, call, call->out_scb); if (ret < 0) return ret; /* Was deleted if vnode->status.abort_code == VNOVNODE. */ - xdr_decode_YFSVolSync(&bp, NULL); + xdr_decode_YFSVolSync(&bp, call->out_volsync); return 0; } @@ -929,7 +828,8 @@ static const struct afs_call_type yfs_RXYFSRemoveFile2 = { * Remove a file and retrieve new file status. */ int yfs_fs_remove_file2(struct afs_fs_cursor *fc, struct afs_vnode *vnode, - const char *name, u64 current_data_version) + const char *name, struct afs_status_cb *dvnode_scb, + struct afs_status_cb *vnode_scb) { struct afs_vnode *dvnode = fc->vnode; struct afs_call *call; @@ -954,9 +854,8 @@ int yfs_fs_remove_file2(struct afs_fs_cursor *fc, struct afs_vnode *vnode, return -ENOMEM; call->key = fc->key; - call->reply[0] = dvnode; - call->reply[1] = vnode; - call->expected_version = current_data_version + 1; + call->out_dir_scb = dvnode_scb; + call->out_scb = vnode_scb; /* marshall the parameters */ bp = call->request; @@ -968,6 +867,7 @@ int yfs_fs_remove_file2(struct afs_fs_cursor *fc, struct afs_vnode *vnode, afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_call1(call, &dvnode->fid, name); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } @@ -977,7 +877,6 @@ int yfs_fs_remove_file2(struct afs_fs_cursor *fc, struct afs_vnode *vnode, */ static int yfs_deliver_fs_remove(struct afs_call *call) { - struct afs_vnode *dvnode = call->reply[0]; const __be32 *bp; int ret; @@ -987,14 +886,12 @@ static int yfs_deliver_fs_remove(struct afs_call *call) if (ret < 0) return ret; - /* unmarshall the reply once we've received all of it */ bp = call->buffer; - ret = yfs_decode_status(call, &bp, &dvnode->status, dvnode, - &call->expected_version, NULL); + ret = xdr_decode_YFSFetchStatus(&bp, call, call->out_dir_scb); if (ret < 0) return ret; - xdr_decode_YFSVolSync(&bp, NULL); + xdr_decode_YFSVolSync(&bp, call->out_volsync); return 0; } @@ -1019,7 +916,8 @@ static const struct afs_call_type yfs_RXYFSRemoveDir = { * remove a file or directory */ int yfs_fs_remove(struct afs_fs_cursor *fc, struct afs_vnode *vnode, - const char *name, bool isdir, u64 current_data_version) + const char *name, bool isdir, + struct afs_status_cb *dvnode_scb) { struct afs_vnode *dvnode = fc->vnode; struct afs_call *call; @@ -1042,9 +940,7 @@ int yfs_fs_remove(struct afs_fs_cursor *fc, struct afs_vnode *vnode, return -ENOMEM; call->key = fc->key; - call->reply[0] = dvnode; - call->reply[1] = vnode; - call->expected_version = current_data_version + 1; + call->out_dir_scb = dvnode_scb; /* marshall the parameters */ bp = call->request; @@ -1056,6 +952,7 @@ int yfs_fs_remove(struct afs_fs_cursor *fc, struct afs_vnode *vnode, afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_call1(call, &dvnode->fid, name); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } @@ -1065,7 +962,6 @@ int yfs_fs_remove(struct afs_fs_cursor *fc, struct afs_vnode *vnode, */ static int yfs_deliver_fs_link(struct afs_call *call) { - struct afs_vnode *dvnode = call->reply[0], *vnode = call->reply[1]; const __be32 *bp; int ret; @@ -1075,16 +971,14 @@ static int yfs_deliver_fs_link(struct afs_call *call) if (ret < 0) return ret; - /* unmarshall the reply once we've received all of it */ bp = call->buffer; - ret = yfs_decode_status(call, &bp, &vnode->status, vnode, NULL, NULL); + ret = xdr_decode_YFSFetchStatus(&bp, call, call->out_scb); if (ret < 0) return ret; - ret = yfs_decode_status(call, &bp, &dvnode->status, dvnode, - &call->expected_version, NULL); + ret = xdr_decode_YFSFetchStatus(&bp, call, call->out_dir_scb); if (ret < 0) return ret; - xdr_decode_YFSVolSync(&bp, NULL); + xdr_decode_YFSVolSync(&bp, call->out_volsync); _leave(" = 0 [done]"); return 0; } @@ -1103,7 +997,9 @@ static const struct afs_call_type yfs_RXYFSLink = { * Make a hard link. */ int yfs_fs_link(struct afs_fs_cursor *fc, struct afs_vnode *vnode, - const char *name, u64 current_data_version) + const char *name, + struct afs_status_cb *dvnode_scb, + struct afs_status_cb *vnode_scb) { struct afs_vnode *dvnode = fc->vnode; struct afs_call *call; @@ -1127,9 +1023,8 @@ int yfs_fs_link(struct afs_fs_cursor *fc, struct afs_vnode *vnode, return -ENOMEM; call->key = fc->key; - call->reply[0] = dvnode; - call->reply[1] = vnode; - call->expected_version = current_data_version + 1; + call->out_dir_scb = dvnode_scb; + call->out_scb = vnode_scb; /* marshall the parameters */ bp = call->request; @@ -1142,6 +1037,7 @@ int yfs_fs_link(struct afs_fs_cursor *fc, struct afs_vnode *vnode, afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_call1(call, &vnode->fid, name); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } @@ -1151,7 +1047,6 @@ int yfs_fs_link(struct afs_fs_cursor *fc, struct afs_vnode *vnode, */ static int yfs_deliver_fs_symlink(struct afs_call *call) { - struct afs_vnode *vnode = call->reply[0]; const __be32 *bp; int ret; @@ -1163,15 +1058,14 @@ static int yfs_deliver_fs_symlink(struct afs_call *call) /* unmarshall the reply once we've received all of it */ bp = call->buffer; - xdr_decode_YFSFid(&bp, call->reply[1]); - ret = yfs_decode_status(call, &bp, call->reply[2], NULL, NULL, NULL); + xdr_decode_YFSFid(&bp, call->out_fid); + ret = xdr_decode_YFSFetchStatus(&bp, call, call->out_scb); if (ret < 0) return ret; - ret = yfs_decode_status(call, &bp, &vnode->status, vnode, - &call->expected_version, NULL); + ret = xdr_decode_YFSFetchStatus(&bp, call, call->out_dir_scb); if (ret < 0) return ret; - xdr_decode_YFSVolSync(&bp, NULL); + xdr_decode_YFSVolSync(&bp, call->out_volsync); _leave(" = 0 [done]"); return 0; @@ -1193,9 +1087,9 @@ static const struct afs_call_type yfs_RXYFSSymlink = { int yfs_fs_symlink(struct afs_fs_cursor *fc, const char *name, const char *contents, - u64 current_data_version, + struct afs_status_cb *dvnode_scb, struct afs_fid *newfid, - struct afs_file_status *newstatus) + struct afs_status_cb *vnode_scb) { struct afs_vnode *dvnode = fc->vnode; struct afs_call *call; @@ -1222,10 +1116,9 @@ int yfs_fs_symlink(struct afs_fs_cursor *fc, return -ENOMEM; call->key = fc->key; - call->reply[0] = dvnode; - call->reply[1] = newfid; - call->reply[2] = newstatus; - call->expected_version = current_data_version + 1; + call->out_dir_scb = dvnode_scb; + call->out_fid = newfid; + call->out_scb = vnode_scb; /* marshall the parameters */ bp = call->request; @@ -1239,6 +1132,7 @@ int yfs_fs_symlink(struct afs_fs_cursor *fc, afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_call1(call, &dvnode->fid, name); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } @@ -1248,8 +1142,6 @@ int yfs_fs_symlink(struct afs_fs_cursor *fc, */ static int yfs_deliver_fs_rename(struct afs_call *call) { - struct afs_vnode *orig_dvnode = call->reply[0]; - struct afs_vnode *new_dvnode = call->reply[1]; const __be32 *bp; int ret; @@ -1259,20 +1151,17 @@ static int yfs_deliver_fs_rename(struct afs_call *call) if (ret < 0) return ret; - /* unmarshall the reply once we've received all of it */ bp = call->buffer; - ret = yfs_decode_status(call, &bp, &orig_dvnode->status, orig_dvnode, - &call->expected_version, NULL); + ret = xdr_decode_YFSFetchStatus(&bp, call, call->out_dir_scb); if (ret < 0) return ret; - if (new_dvnode != orig_dvnode) { - ret = yfs_decode_status(call, &bp, &new_dvnode->status, new_dvnode, - &call->expected_version_2, NULL); + if (call->out_dir_scb != call->out_scb) { + ret = xdr_decode_YFSFetchStatus(&bp, call, call->out_scb); if (ret < 0) return ret; } - xdr_decode_YFSVolSync(&bp, NULL); + xdr_decode_YFSVolSync(&bp, call->out_volsync); _leave(" = 0 [done]"); return 0; } @@ -1294,8 +1183,8 @@ int yfs_fs_rename(struct afs_fs_cursor *fc, const char *orig_name, struct afs_vnode *new_dvnode, const char *new_name, - u64 current_orig_data_version, - u64 current_new_data_version) + struct afs_status_cb *orig_dvnode_scb, + struct afs_status_cb *new_dvnode_scb) { struct afs_vnode *orig_dvnode = fc->vnode; struct afs_call *call; @@ -1321,10 +1210,8 @@ int yfs_fs_rename(struct afs_fs_cursor *fc, return -ENOMEM; call->key = fc->key; - call->reply[0] = orig_dvnode; - call->reply[1] = new_dvnode; - call->expected_version = current_orig_data_version + 1; - call->expected_version_2 = current_new_data_version + 1; + call->out_dir_scb = orig_dvnode_scb; + call->out_scb = new_dvnode_scb; /* marshall the parameters */ bp = call->request; @@ -1338,46 +1225,18 @@ int yfs_fs_rename(struct afs_fs_cursor *fc, afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_call2(call, &orig_dvnode->fid, orig_name, new_name); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } /* - * Deliver reply data to a YFS.StoreData64 operation. - */ -static int yfs_deliver_fs_store_data(struct afs_call *call) -{ - struct afs_vnode *vnode = call->reply[0]; - const __be32 *bp; - int ret; - - _enter(""); - - ret = afs_transfer_reply(call); - if (ret < 0) - return ret; - - /* unmarshall the reply once we've received all of it */ - bp = call->buffer; - ret = yfs_decode_status(call, &bp, &vnode->status, vnode, - &call->expected_version, NULL); - if (ret < 0) - return ret; - xdr_decode_YFSVolSync(&bp, NULL); - - afs_pages_written_back(vnode, call); - - _leave(" = 0 [done]"); - return 0; -} - -/* * YFS.StoreData64 operation type. */ static const struct afs_call_type yfs_RXYFSStoreData64 = { .name = "YFS.StoreData64", .op = yfs_FS_StoreData64, - .deliver = yfs_deliver_fs_store_data, + .deliver = yfs_deliver_status_and_volsync, .destructor = afs_flat_call_destructor, }; @@ -1386,7 +1245,8 @@ static const struct afs_call_type yfs_RXYFSStoreData64 = { */ int yfs_fs_store_data(struct afs_fs_cursor *fc, struct address_space *mapping, pgoff_t first, pgoff_t last, - unsigned offset, unsigned to) + unsigned offset, unsigned to, + struct afs_status_cb *scb) { struct afs_vnode *vnode = fc->vnode; struct afs_call *call; @@ -1424,13 +1284,12 @@ int yfs_fs_store_data(struct afs_fs_cursor *fc, struct address_space *mapping, call->key = fc->key; call->mapping = mapping; - call->reply[0] = vnode; call->first = first; call->last = last; call->first_offset = offset; call->last_to = to; call->send_pages = true; - call->expected_version = vnode->status.data_version + 1; + call->out_scb = scb; /* marshall the parameters */ bp = call->request; @@ -1445,51 +1304,25 @@ int yfs_fs_store_data(struct afs_fs_cursor *fc, struct address_space *mapping, afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_call(call, &vnode->fid); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } /* - * deliver reply data to an FS.StoreStatus - */ -static int yfs_deliver_fs_store_status(struct afs_call *call) -{ - struct afs_vnode *vnode = call->reply[0]; - const __be32 *bp; - int ret; - - _enter(""); - - ret = afs_transfer_reply(call); - if (ret < 0) - return ret; - - /* unmarshall the reply once we've received all of it */ - bp = call->buffer; - ret = yfs_decode_status(call, &bp, &vnode->status, vnode, - &call->expected_version, NULL); - if (ret < 0) - return ret; - xdr_decode_YFSVolSync(&bp, NULL); - - _leave(" = 0 [done]"); - return 0; -} - -/* * YFS.StoreStatus operation type */ static const struct afs_call_type yfs_RXYFSStoreStatus = { .name = "YFS.StoreStatus", .op = yfs_FS_StoreStatus, - .deliver = yfs_deliver_fs_store_status, + .deliver = yfs_deliver_status_and_volsync, .destructor = afs_flat_call_destructor, }; static const struct afs_call_type yfs_RXYFSStoreData64_as_Status = { .name = "YFS.StoreData64", .op = yfs_FS_StoreData64, - .deliver = yfs_deliver_fs_store_status, + .deliver = yfs_deliver_status_and_volsync, .destructor = afs_flat_call_destructor, }; @@ -1497,7 +1330,8 @@ static const struct afs_call_type yfs_RXYFSStoreData64_as_Status = { * Set the attributes on a file, using YFS.StoreData64 rather than * YFS.StoreStatus so as to alter the file size also. */ -static int yfs_fs_setattr_size(struct afs_fs_cursor *fc, struct iattr *attr) +static int yfs_fs_setattr_size(struct afs_fs_cursor *fc, struct iattr *attr, + struct afs_status_cb *scb) { struct afs_vnode *vnode = fc->vnode; struct afs_call *call; @@ -1518,8 +1352,7 @@ static int yfs_fs_setattr_size(struct afs_fs_cursor *fc, struct iattr *attr) return -ENOMEM; call->key = fc->key; - call->reply[0] = vnode; - call->expected_version = vnode->status.data_version + 1; + call->out_scb = scb; /* marshall the parameters */ bp = call->request; @@ -1534,6 +1367,7 @@ static int yfs_fs_setattr_size(struct afs_fs_cursor *fc, struct iattr *attr) afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_call(call, &vnode->fid); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } @@ -1542,7 +1376,8 @@ static int yfs_fs_setattr_size(struct afs_fs_cursor *fc, struct iattr *attr) * Set the attributes on a file, using YFS.StoreData64 if there's a change in * file size, and YFS.StoreStatus otherwise. */ -int yfs_fs_setattr(struct afs_fs_cursor *fc, struct iattr *attr) +int yfs_fs_setattr(struct afs_fs_cursor *fc, struct iattr *attr, + struct afs_status_cb *scb) { struct afs_vnode *vnode = fc->vnode; struct afs_call *call; @@ -1550,7 +1385,7 @@ int yfs_fs_setattr(struct afs_fs_cursor *fc, struct iattr *attr) __be32 *bp; if (attr->ia_valid & ATTR_SIZE) - return yfs_fs_setattr_size(fc, attr); + return yfs_fs_setattr_size(fc, attr, scb); _enter(",%x,{%llx:%llu},,", key_serial(fc->key), vnode->fid.vid, vnode->fid.vnode); @@ -1565,8 +1400,7 @@ int yfs_fs_setattr(struct afs_fs_cursor *fc, struct iattr *attr) return -ENOMEM; call->key = fc->key; - call->reply[0] = vnode; - call->expected_version = vnode->status.data_version; + call->out_scb = scb; /* marshall the parameters */ bp = call->request; @@ -1578,6 +1412,7 @@ int yfs_fs_setattr(struct afs_fs_cursor *fc, struct iattr *attr) afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_call(call, &vnode->fid); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } @@ -1607,7 +1442,7 @@ static int yfs_deliver_fs_get_volume_status(struct afs_call *call) return ret; bp = call->buffer; - xdr_decode_YFSFetchVolumeStatus(&bp, call->reply[1]); + xdr_decode_YFSFetchVolumeStatus(&bp, call->out_volstatus); call->unmarshall++; afs_extract_to_tmp(call); @@ -1623,7 +1458,7 @@ static int yfs_deliver_fs_get_volume_status(struct afs_call *call) return afs_protocol_error(call, -EBADMSG, afs_eproto_volname_len); size = (call->count + 3) & ~3; /* It's padded */ - afs_extract_begin(call, call->reply[2], size); + afs_extract_to_buf(call, size); call->unmarshall++; /* Fall through - and extract the volume name */ @@ -1633,7 +1468,7 @@ static int yfs_deliver_fs_get_volume_status(struct afs_call *call) if (ret < 0) return ret; - p = call->reply[2]; + p = call->buffer; p[call->count] = 0; _debug("volname '%s'", p); afs_extract_to_tmp(call); @@ -1651,7 +1486,7 @@ static int yfs_deliver_fs_get_volume_status(struct afs_call *call) return afs_protocol_error(call, -EBADMSG, afs_eproto_offline_msg_len); size = (call->count + 3) & ~3; /* It's padded */ - afs_extract_begin(call, call->reply[2], size); + afs_extract_to_buf(call, size); call->unmarshall++; /* Fall through - and extract the offline message */ @@ -1661,7 +1496,7 @@ static int yfs_deliver_fs_get_volume_status(struct afs_call *call) if (ret < 0) return ret; - p = call->reply[2]; + p = call->buffer; p[call->count] = 0; _debug("offline '%s'", p); @@ -1680,7 +1515,7 @@ static int yfs_deliver_fs_get_volume_status(struct afs_call *call) return afs_protocol_error(call, -EBADMSG, afs_eproto_motd_len); size = (call->count + 3) & ~3; /* It's padded */ - afs_extract_begin(call, call->reply[2], size); + afs_extract_to_buf(call, size); call->unmarshall++; /* Fall through - and extract the message of the day */ @@ -1690,7 +1525,7 @@ static int yfs_deliver_fs_get_volume_status(struct afs_call *call) if (ret < 0) return ret; - p = call->reply[2]; + p = call->buffer; p[call->count] = 0; _debug("motd '%s'", p); @@ -1706,23 +1541,13 @@ static int yfs_deliver_fs_get_volume_status(struct afs_call *call) } /* - * Destroy a YFS.GetVolumeStatus call. - */ -static void yfs_get_volume_status_call_destructor(struct afs_call *call) -{ - kfree(call->reply[2]); - call->reply[2] = NULL; - afs_flat_call_destructor(call); -} - -/* * YFS.GetVolumeStatus operation type */ static const struct afs_call_type yfs_RXYFSGetVolumeStatus = { .name = "YFS.GetVolumeStatus", .op = yfs_FS_GetVolumeStatus, .deliver = yfs_deliver_fs_get_volume_status, - .destructor = yfs_get_volume_status_call_destructor, + .destructor = afs_flat_call_destructor, }; /* @@ -1735,28 +1560,21 @@ int yfs_fs_get_volume_status(struct afs_fs_cursor *fc, struct afs_call *call; struct afs_net *net = afs_v2net(vnode); __be32 *bp; - void *tmpbuf; _enter(""); - tmpbuf = kmalloc(AFSOPAQUEMAX, GFP_KERNEL); - if (!tmpbuf) - return -ENOMEM; - call = afs_alloc_flat_call(net, &yfs_RXYFSGetVolumeStatus, sizeof(__be32) * 2 + sizeof(struct yfs_xdr_u64), - sizeof(struct yfs_xdr_YFSFetchVolumeStatus) + - sizeof(__be32)); - if (!call) { - kfree(tmpbuf); + max_t(size_t, + sizeof(struct yfs_xdr_YFSFetchVolumeStatus) + + sizeof(__be32), + AFSOPAQUEMAX + 1)); + if (!call) return -ENOMEM; - } call->key = fc->key; - call->reply[0] = vnode; - call->reply[1] = vs; - call->reply[2] = tmpbuf; + call->out_volstatus = vs; /* marshall the parameters */ bp = call->request; @@ -1767,39 +1585,12 @@ int yfs_fs_get_volume_status(struct afs_fs_cursor *fc, afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_call(call, &vnode->fid); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } /* - * Deliver reply data to operations that just return a file status and a volume - * sync record. - */ -static int yfs_deliver_status_and_volsync(struct afs_call *call) -{ - struct afs_vnode *vnode = call->reply[0]; - const __be32 *bp; - int ret; - - _enter("{%u}", call->unmarshall); - - ret = afs_transfer_reply(call); - if (ret < 0) - return ret; - - /* unmarshall the reply once we've received all of it */ - bp = call->buffer; - ret = yfs_decode_status(call, &bp, &vnode->status, vnode, - &call->expected_version, NULL); - if (ret < 0) - return ret; - xdr_decode_YFSVolSync(&bp, NULL); - - _leave(" = 0 [done]"); - return 0; -} - -/* * YFS.SetLock operation type */ static const struct afs_call_type yfs_RXYFSSetLock = { @@ -1834,7 +1625,8 @@ static const struct afs_call_type yfs_RXYFSReleaseLock = { /* * Set a lock on a file */ -int yfs_fs_set_lock(struct afs_fs_cursor *fc, afs_lock_type_t type) +int yfs_fs_set_lock(struct afs_fs_cursor *fc, afs_lock_type_t type, + struct afs_status_cb *scb) { struct afs_vnode *vnode = fc->vnode; struct afs_call *call; @@ -1853,8 +1645,8 @@ int yfs_fs_set_lock(struct afs_fs_cursor *fc, afs_lock_type_t type) return -ENOMEM; call->key = fc->key; - call->reply[0] = vnode; - call->want_reply_time = true; + call->lvnode = vnode; + call->out_scb = scb; /* marshall the parameters */ bp = call->request; @@ -1866,6 +1658,7 @@ int yfs_fs_set_lock(struct afs_fs_cursor *fc, afs_lock_type_t type) afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_calli(call, &vnode->fid, type); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } @@ -1873,7 +1666,7 @@ int yfs_fs_set_lock(struct afs_fs_cursor *fc, afs_lock_type_t type) /* * extend a lock on a file */ -int yfs_fs_extend_lock(struct afs_fs_cursor *fc) +int yfs_fs_extend_lock(struct afs_fs_cursor *fc, struct afs_status_cb *scb) { struct afs_vnode *vnode = fc->vnode; struct afs_call *call; @@ -1891,8 +1684,8 @@ int yfs_fs_extend_lock(struct afs_fs_cursor *fc) return -ENOMEM; call->key = fc->key; - call->reply[0] = vnode; - call->want_reply_time = true; + call->lvnode = vnode; + call->out_scb = scb; /* marshall the parameters */ bp = call->request; @@ -1903,6 +1696,7 @@ int yfs_fs_extend_lock(struct afs_fs_cursor *fc) afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_call(call, &vnode->fid); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } @@ -1910,7 +1704,7 @@ int yfs_fs_extend_lock(struct afs_fs_cursor *fc) /* * release a lock on a file */ -int yfs_fs_release_lock(struct afs_fs_cursor *fc) +int yfs_fs_release_lock(struct afs_fs_cursor *fc, struct afs_status_cb *scb) { struct afs_vnode *vnode = fc->vnode; struct afs_call *call; @@ -1928,7 +1722,8 @@ int yfs_fs_release_lock(struct afs_fs_cursor *fc) return -ENOMEM; call->key = fc->key; - call->reply[0] = vnode; + call->lvnode = vnode; + call->out_scb = scb; /* marshall the parameters */ bp = call->request; @@ -1939,48 +1734,18 @@ int yfs_fs_release_lock(struct afs_fs_cursor *fc) afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_call(call, &vnode->fid); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } /* - * Deliver reply data to an FS.FetchStatus with no vnode. - */ -static int yfs_deliver_fs_fetch_status(struct afs_call *call) -{ - struct afs_file_status *status = call->reply[1]; - struct afs_callback *callback = call->reply[2]; - struct afs_volsync *volsync = call->reply[3]; - struct afs_vnode *vnode = call->reply[0]; - const __be32 *bp; - int ret; - - ret = afs_transfer_reply(call); - if (ret < 0) - return ret; - - _enter("{%llx:%llu}", vnode->fid.vid, vnode->fid.vnode); - - /* unmarshall the reply once we've received all of it */ - bp = call->buffer; - ret = yfs_decode_status(call, &bp, status, vnode, - &call->expected_version, NULL); - if (ret < 0) - return ret; - xdr_decode_YFSCallBack_raw(&bp, callback); - xdr_decode_YFSVolSync(&bp, volsync); - - _leave(" = 0 [done]"); - return 0; -} - -/* * YFS.FetchStatus operation type */ static const struct afs_call_type yfs_RXYFSFetchStatus = { .name = "YFS.FetchStatus", .op = yfs_FS_FetchStatus, - .deliver = yfs_deliver_fs_fetch_status, + .deliver = yfs_deliver_fs_status_cb_and_volsync, .destructor = afs_flat_call_destructor, }; @@ -1990,8 +1755,7 @@ static const struct afs_call_type yfs_RXYFSFetchStatus = { int yfs_fs_fetch_status(struct afs_fs_cursor *fc, struct afs_net *net, struct afs_fid *fid, - struct afs_file_status *status, - struct afs_callback *callback, + struct afs_status_cb *scb, struct afs_volsync *volsync) { struct afs_call *call; @@ -2012,11 +1776,8 @@ int yfs_fs_fetch_status(struct afs_fs_cursor *fc, } call->key = fc->key; - call->reply[0] = NULL; /* vnode for fid[0] */ - call->reply[1] = status; - call->reply[2] = callback; - call->reply[3] = volsync; - call->expected_version = 1; /* vnode->status.data_version */ + call->out_scb = scb; + call->out_volsync = volsync; /* marshall the parameters */ bp = call->request; @@ -2025,9 +1786,9 @@ int yfs_fs_fetch_status(struct afs_fs_cursor *fc, bp = xdr_encode_YFSFid(bp, fid); yfs_check_req(call, bp); - call->cb_break = fc->cb_break; afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_call(call, fid); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } @@ -2037,9 +1798,7 @@ int yfs_fs_fetch_status(struct afs_fs_cursor *fc, */ static int yfs_deliver_fs_inline_bulk_status(struct afs_call *call) { - struct afs_file_status *statuses; - struct afs_callback *callbacks; - struct afs_vnode *vnode = call->reply[0]; + struct afs_status_cb *scb; const __be32 *bp; u32 tmp; int ret; @@ -2078,10 +1837,8 @@ static int yfs_deliver_fs_inline_bulk_status(struct afs_call *call) return ret; bp = call->buffer; - statuses = call->reply[1]; - ret = yfs_decode_status(call, &bp, &statuses[call->count], - call->count == 0 ? vnode : NULL, - NULL, NULL); + scb = &call->out_scb[call->count]; + ret = xdr_decode_YFSFetchStatus(&bp, call, scb); if (ret < 0) return ret; @@ -2120,13 +1877,8 @@ static int yfs_deliver_fs_inline_bulk_status(struct afs_call *call) _debug("unmarshall CB array"); bp = call->buffer; - callbacks = call->reply[2]; - xdr_decode_YFSCallBack_raw(&bp, &callbacks[call->count]); - statuses = call->reply[1]; - if (call->count == 0 && vnode && statuses[0].abort_code == 0) { - bp = call->buffer; - xdr_decode_YFSCallBack(call, vnode, &bp); - } + scb = &call->out_scb[call->count]; + xdr_decode_YFSCallBack(&bp, call, scb); call->count++; if (call->count < call->count2) goto more_cbs; @@ -2141,7 +1893,7 @@ static int yfs_deliver_fs_inline_bulk_status(struct afs_call *call) return ret; bp = call->buffer; - xdr_decode_YFSVolSync(&bp, call->reply[3]); + xdr_decode_YFSVolSync(&bp, call->out_volsync); call->unmarshall++; @@ -2170,8 +1922,7 @@ static const struct afs_call_type yfs_RXYFSInlineBulkStatus = { int yfs_fs_inline_bulk_status(struct afs_fs_cursor *fc, struct afs_net *net, struct afs_fid *fids, - struct afs_file_status *statuses, - struct afs_callback *callbacks, + struct afs_status_cb *statuses, unsigned int nr_fids, struct afs_volsync *volsync) { @@ -2194,10 +1945,8 @@ int yfs_fs_inline_bulk_status(struct afs_fs_cursor *fc, } call->key = fc->key; - call->reply[0] = NULL; /* vnode for fid[0] */ - call->reply[1] = statuses; - call->reply[2] = callbacks; - call->reply[3] = volsync; + call->out_scb = statuses; + call->out_volsync = volsync; call->count2 = nr_fids; /* marshall the parameters */ @@ -2209,9 +1958,9 @@ int yfs_fs_inline_bulk_status(struct afs_fs_cursor *fc, bp = xdr_encode_YFSFid(bp, &fids[i]); yfs_check_req(call, bp); - call->cb_break = fc->cb_break; afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_call(call, &fids[0]); + afs_set_fc_call(call, fc); afs_make_call(&fc->ac, call, GFP_NOFS); return afs_wait_for_call_to_complete(call, &fc->ac); } @@ -2221,9 +1970,7 @@ int yfs_fs_inline_bulk_status(struct afs_fs_cursor *fc, */ static int yfs_deliver_fs_fetch_opaque_acl(struct afs_call *call) { - struct afs_volsync *volsync = call->reply[2]; - struct afs_vnode *vnode = call->reply[1]; - struct yfs_acl *yacl = call->reply[0]; + struct yfs_acl *yacl = call->out_yacl; struct afs_acl *acl; const __be32 *bp; unsigned int size; @@ -2308,11 +2055,10 @@ static int yfs_deliver_fs_fetch_opaque_acl(struct afs_call *call) bp = call->buffer; yacl->inherit_flag = ntohl(*bp++); yacl->num_cleaned = ntohl(*bp++); - ret = yfs_decode_status(call, &bp, &vnode->status, vnode, - &call->expected_version, NULL); + ret = xdr_decode_YFSFetchStatus(&bp, call, call->out_scb); if (ret < 0) return ret; - xdr_decode_YFSVolSync(&bp, volsync); + xdr_decode_YFSVolSync(&bp, call->out_volsync); call->unmarshall++; @@ -2333,12 +2079,6 @@ void yfs_free_opaque_acl(struct yfs_acl *yacl) } } -static void yfs_destroy_fs_fetch_opaque_acl(struct afs_call *call) -{ - yfs_free_opaque_acl(call->reply[0]); - afs_flat_call_destructor(call); -} - /* * YFS.FetchOpaqueACL operation type */ @@ -2346,18 +2086,18 @@ static const struct afs_call_type yfs_RXYFSFetchOpaqueACL = { .name = "YFS.FetchOpaqueACL", .op = yfs_FS_FetchOpaqueACL, .deliver = yfs_deliver_fs_fetch_opaque_acl, - .destructor = yfs_destroy_fs_fetch_opaque_acl, + .destructor = afs_flat_call_destructor, }; /* * Fetch the YFS advanced ACLs for a file. */ struct yfs_acl *yfs_fs_fetch_opaque_acl(struct afs_fs_cursor *fc, - unsigned int flags) + struct yfs_acl *yacl, + struct afs_status_cb *scb) { struct afs_vnode *vnode = fc->vnode; struct afs_call *call; - struct yfs_acl *yacl; struct afs_net *net = afs_v2net(vnode); __be32 *bp; @@ -2370,19 +2110,15 @@ struct yfs_acl *yfs_fs_fetch_opaque_acl(struct afs_fs_cursor *fc, sizeof(__be32) * 2 + sizeof(struct yfs_xdr_YFSFetchStatus) + sizeof(struct yfs_xdr_YFSVolSync)); - if (!call) - goto nomem; - - yacl = kzalloc(sizeof(struct yfs_acl), GFP_KERNEL); - if (!yacl) - goto nomem_call; + if (!call) { + fc->ac.error = -ENOMEM; + return ERR_PTR(-ENOMEM); + } - yacl->flags = flags; call->key = fc->key; - call->reply[0] = yacl; - call->reply[1] = vnode; - call->reply[2] = NULL; /* volsync */ - call->ret_reply0 = true; + call->out_yacl = yacl; + call->out_scb = scb; + call->out_volsync = NULL; /* marshall the parameters */ bp = call->request; @@ -2391,17 +2127,10 @@ struct yfs_acl *yfs_fs_fetch_opaque_acl(struct afs_fs_cursor *fc, bp = xdr_encode_YFSFid(bp, &vnode->fid); yfs_check_req(call, bp); - call->cb_break = fc->cb_break; afs_use_fs_server(call, fc->cbi); trace_afs_make_fs_call(call, &vnode->fid); afs_make_call(&fc->ac, call, GFP_KERNEL); return (struct yfs_acl *)afs_wait_for_call_to_complete(call, &fc->ac); - -nomem_call: - afs_put_call(call); -nomem: - fc->ac.error = -ENOMEM; - return ERR_PTR(-ENOMEM); } /* @@ -2417,7 +2146,8 @@ static const struct afs_call_type yfs_RXYFSStoreOpaqueACL2 = { /* * Fetch the YFS ACL for a file. */ -int yfs_fs_store_opaque_acl2(struct afs_fs_cursor *fc, const struct afs_acl *acl) +int yfs_fs_store_opaque_acl2(struct afs_fs_cursor *fc, const struct afs_acl *acl, + struct afs_status_cb *scb) { struct afs_vnode *vnode = fc->vnode; struct afs_call *call; @@ -2441,8 +2171,8 @@ int yfs_fs_store_opaque_acl2(struct afs_fs_cursor *fc, const struct afs_acl *acl } call->key = fc->key; - call->reply[0] = vnode; - call->reply[2] = NULL; /* volsync */ + call->out_scb = scb; + call->out_volsync = NULL; /* marshall the parameters */ bp = call->request; diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c index 36a8dc699448..72f8e1311392 100644 --- a/fs/ceph/caps.c +++ b/fs/ceph/caps.c @@ -892,8 +892,8 @@ int __ceph_caps_issued_mask(struct ceph_inode_info *ci, int mask, int touch) int have = ci->i_snap_caps; if ((have & mask) == mask) { - dout("__ceph_caps_issued_mask %p snap issued %s" - " (mask %s)\n", &ci->vfs_inode, + dout("__ceph_caps_issued_mask ino 0x%lx snap issued %s" + " (mask %s)\n", ci->vfs_inode.i_ino, ceph_cap_string(have), ceph_cap_string(mask)); return 1; @@ -904,8 +904,8 @@ int __ceph_caps_issued_mask(struct ceph_inode_info *ci, int mask, int touch) if (!__cap_is_valid(cap)) continue; if ((cap->issued & mask) == mask) { - dout("__ceph_caps_issued_mask %p cap %p issued %s" - " (mask %s)\n", &ci->vfs_inode, cap, + dout("__ceph_caps_issued_mask ino 0x%lx cap %p issued %s" + " (mask %s)\n", ci->vfs_inode.i_ino, cap, ceph_cap_string(cap->issued), ceph_cap_string(mask)); if (touch) @@ -916,8 +916,8 @@ int __ceph_caps_issued_mask(struct ceph_inode_info *ci, int mask, int touch) /* does a combination of caps satisfy mask? */ have |= cap->issued; if ((have & mask) == mask) { - dout("__ceph_caps_issued_mask %p combo issued %s" - " (mask %s)\n", &ci->vfs_inode, + dout("__ceph_caps_issued_mask ino 0x%lx combo issued %s" + " (mask %s)\n", ci->vfs_inode.i_ino, ceph_cap_string(cap->issued), ceph_cap_string(mask)); if (touch) { @@ -2257,8 +2257,6 @@ int ceph_fsync(struct file *file, loff_t start, loff_t end, int datasync) if (datasync) goto out; - inode_lock(inode); - dirty = try_flush_caps(inode, &flush_tid); dout("fsync dirty caps are %s\n", ceph_cap_string(dirty)); @@ -2273,7 +2271,6 @@ int ceph_fsync(struct file *file, loff_t start, loff_t end, int datasync) ret = wait_event_interruptible(ci->i_cap_wq, caps_are_flushed(inode, flush_tid)); } - inode_unlock(inode); out: dout("fsync %p%s result=%d\n", inode, datasync ? " datasync" : "", ret); return ret; @@ -2528,9 +2525,14 @@ static void __take_cap_refs(struct ceph_inode_info *ci, int got, * to (when applicable), and check against max_size here as well. * Note that caller is responsible for ensuring max_size increases are * requested from the MDS. + * + * Returns 0 if caps were not able to be acquired (yet), a 1 if they were, + * or a negative error code. + * + * FIXME: how does a 0 return differ from -EAGAIN? */ static int try_get_cap_refs(struct ceph_inode_info *ci, int need, int want, - loff_t endoff, bool nonblock, int *got, int *err) + loff_t endoff, bool nonblock, int *got) { struct inode *inode = &ci->vfs_inode; struct ceph_mds_client *mdsc = ceph_inode_to_client(inode)->mdsc; @@ -2550,8 +2552,7 @@ again: if ((file_wanted & need) != need) { dout("try_get_cap_refs need %s file_wanted %s, EBADF\n", ceph_cap_string(need), ceph_cap_string(file_wanted)); - *err = -EBADF; - ret = 1; + ret = -EBADF; goto out_unlock; } @@ -2572,10 +2573,8 @@ again: if (endoff >= 0 && endoff > (loff_t)ci->i_max_size) { dout("get_cap_refs %p endoff %llu > maxsize %llu\n", inode, endoff, ci->i_max_size); - if (endoff > ci->i_requested_max_size) { - *err = -EAGAIN; - ret = 1; - } + if (endoff > ci->i_requested_max_size) + ret = -EAGAIN; goto out_unlock; } /* @@ -2610,8 +2609,7 @@ again: * task isn't in TASK_RUNNING state */ if (nonblock) { - *err = -EAGAIN; - ret = 1; + ret = -EAGAIN; goto out_unlock; } @@ -2640,8 +2638,7 @@ again: if (session_readonly) { dout("get_cap_refs %p needed %s but mds%d readonly\n", inode, ceph_cap_string(need), ci->i_auth_cap->mds); - *err = -EROFS; - ret = 1; + ret = -EROFS; goto out_unlock; } @@ -2650,16 +2647,14 @@ again: if (READ_ONCE(mdsc->fsc->mount_state) == CEPH_MOUNT_SHUTDOWN) { dout("get_cap_refs %p forced umount\n", inode); - *err = -EIO; - ret = 1; + ret = -EIO; goto out_unlock; } mds_wanted = __ceph_caps_mds_wanted(ci, false); if (need & ~(mds_wanted & need)) { dout("get_cap_refs %p caps were dropped" " (session killed?)\n", inode); - *err = -ESTALE; - ret = 1; + ret = -ESTALE; goto out_unlock; } if (!(file_wanted & ~mds_wanted)) @@ -2710,7 +2705,7 @@ static void check_max_size(struct inode *inode, loff_t endoff) int ceph_try_get_caps(struct ceph_inode_info *ci, int need, int want, bool nonblock, int *got) { - int ret, err = 0; + int ret; BUG_ON(need & ~CEPH_CAP_FILE_RD); BUG_ON(want & ~(CEPH_CAP_FILE_CACHE|CEPH_CAP_FILE_LAZYIO|CEPH_CAP_FILE_SHARED)); @@ -2718,15 +2713,8 @@ int ceph_try_get_caps(struct ceph_inode_info *ci, int need, int want, if (ret < 0) return ret; - ret = try_get_cap_refs(ci, need, want, 0, nonblock, got, &err); - if (ret) { - if (err == -EAGAIN) { - ret = 0; - } else if (err < 0) { - ret = err; - } - } - return ret; + ret = try_get_cap_refs(ci, need, want, 0, nonblock, got); + return ret == -EAGAIN ? 0 : ret; } /* @@ -2737,7 +2725,7 @@ int ceph_try_get_caps(struct ceph_inode_info *ci, int need, int want, int ceph_get_caps(struct ceph_inode_info *ci, int need, int want, loff_t endoff, int *got, struct page **pinned_page) { - int _got, ret, err = 0; + int _got, ret; ret = ceph_pool_perm_check(ci, need); if (ret < 0) @@ -2747,21 +2735,19 @@ int ceph_get_caps(struct ceph_inode_info *ci, int need, int want, if (endoff > 0) check_max_size(&ci->vfs_inode, endoff); - err = 0; _got = 0; ret = try_get_cap_refs(ci, need, want, endoff, - false, &_got, &err); - if (ret) { - if (err == -EAGAIN) - continue; - if (err < 0) - ret = err; - } else { + false, &_got); + if (ret == -EAGAIN) { + continue; + } else if (!ret) { + int err; + DEFINE_WAIT_FUNC(wait, woken_wake_function); add_wait_queue(&ci->i_cap_wq, &wait); - while (!try_get_cap_refs(ci, need, want, endoff, - true, &_got, &err)) { + while (!(err = try_get_cap_refs(ci, need, want, endoff, + true, &_got))) { if (signal_pending(current)) { ret = -ERESTARTSYS; break; @@ -2770,19 +2756,14 @@ int ceph_get_caps(struct ceph_inode_info *ci, int need, int want, } remove_wait_queue(&ci->i_cap_wq, &wait); - if (err == -EAGAIN) continue; - if (err < 0) - ret = err; } - if (ret < 0) { - if (err == -ESTALE) { - /* session was killed, try renew caps */ - ret = ceph_renew_caps(&ci->vfs_inode); - if (ret == 0) - continue; - } + if (ret == -ESTALE) { + /* session was killed, try renew caps */ + ret = ceph_renew_caps(&ci->vfs_inode); + if (ret == 0) + continue; return ret; } @@ -4099,7 +4080,7 @@ void ceph_put_fmode(struct ceph_inode_info *ci, int fmode) } /* - * For a soon-to-be unlinked file, drop the AUTH_RDCACHE caps. If it + * For a soon-to-be unlinked file, drop the LINK caps. If it * looks like the link count will hit 0, drop any other caps (other * than PIN) we don't specifically want (due to the file still being * open). diff --git a/fs/ceph/debugfs.c b/fs/ceph/debugfs.c index 98365e74cb4a..b3fc5fe26a1a 100644 --- a/fs/ceph/debugfs.c +++ b/fs/ceph/debugfs.c @@ -37,7 +37,7 @@ static int mdsmap_show(struct seq_file *s, void *p) struct ceph_entity_addr *addr = &mdsmap->m_info[i].addr; int state = mdsmap->m_info[i].state; seq_printf(s, "\tmds%d\t%s\t(%s)\n", i, - ceph_pr_addr(&addr->in_addr), + ceph_pr_addr(addr), ceph_mds_state_name(state)); } return 0; @@ -88,7 +88,7 @@ static int mdsc_show(struct seq_file *s, void *p) req->r_dentry, path ? path : ""); spin_unlock(&req->r_dentry->d_lock); - kfree(path); + ceph_mdsc_free_path(path, pathlen); } else if (req->r_path1) { seq_printf(s, " #%llx/%s", req->r_ino1.ino, req->r_path1); @@ -108,7 +108,7 @@ static int mdsc_show(struct seq_file *s, void *p) req->r_old_dentry, path ? path : ""); spin_unlock(&req->r_old_dentry->d_lock); - kfree(path); + ceph_mdsc_free_path(path, pathlen); } else if (req->r_path2 && req->r_op != CEPH_MDS_OP_SYMLINK) { if (req->r_ino2.ino) seq_printf(s, " #%llx/%s", req->r_ino2.ino, @@ -124,18 +124,48 @@ static int mdsc_show(struct seq_file *s, void *p) return 0; } +static int caps_show_cb(struct inode *inode, struct ceph_cap *cap, void *p) +{ + struct seq_file *s = p; + + seq_printf(s, "0x%-17lx%-17s%-17s\n", inode->i_ino, + ceph_cap_string(cap->issued), + ceph_cap_string(cap->implemented)); + return 0; +} + static int caps_show(struct seq_file *s, void *p) { struct ceph_fs_client *fsc = s->private; - int total, avail, used, reserved, min; + struct ceph_mds_client *mdsc = fsc->mdsc; + int total, avail, used, reserved, min, i; ceph_reservation_status(fsc, &total, &avail, &used, &reserved, &min); seq_printf(s, "total\t\t%d\n" "avail\t\t%d\n" "used\t\t%d\n" "reserved\t%d\n" - "min\t%d\n", + "min\t\t%d\n\n", total, avail, used, reserved, min); + seq_printf(s, "ino issued implemented\n"); + seq_printf(s, "-----------------------------------------------\n"); + + mutex_lock(&mdsc->mutex); + for (i = 0; i < mdsc->max_sessions; i++) { + struct ceph_mds_session *session; + + session = __ceph_lookup_mds_session(mdsc, i); + if (!session) + continue; + mutex_unlock(&mdsc->mutex); + mutex_lock(&session->s_mutex); + ceph_iterate_session_caps(session, caps_show_cb, s); + mutex_unlock(&session->s_mutex); + ceph_put_mds_session(session); + mutex_lock(&mdsc->mutex); + } + mutex_unlock(&mdsc->mutex); + return 0; } diff --git a/fs/ceph/export.c b/fs/ceph/export.c index 3c59ad180ef0..d3ef7ee429ec 100644 --- a/fs/ceph/export.c +++ b/fs/ceph/export.c @@ -22,18 +22,77 @@ struct ceph_nfs_confh { u64 ino, parent_ino; } __attribute__ ((packed)); +/* + * fh for snapped inode + */ +struct ceph_nfs_snapfh { + u64 ino; + u64 snapid; + u64 parent_ino; + u32 hash; +} __attribute__ ((packed)); + +static int ceph_encode_snapfh(struct inode *inode, u32 *rawfh, int *max_len, + struct inode *parent_inode) +{ + const static int snap_handle_length = + sizeof(struct ceph_nfs_snapfh) >> 2; + struct ceph_nfs_snapfh *sfh = (void *)rawfh; + u64 snapid = ceph_snap(inode); + int ret; + bool no_parent = true; + + if (*max_len < snap_handle_length) { + *max_len = snap_handle_length; + ret = FILEID_INVALID; + goto out; + } + + ret = -EINVAL; + if (snapid != CEPH_SNAPDIR) { + struct inode *dir; + struct dentry *dentry = d_find_alias(inode); + if (!dentry) + goto out; + + rcu_read_lock(); + dir = d_inode_rcu(dentry->d_parent); + if (ceph_snap(dir) != CEPH_SNAPDIR) { + sfh->parent_ino = ceph_ino(dir); + sfh->hash = ceph_dentry_hash(dir, dentry); + no_parent = false; + } + rcu_read_unlock(); + dput(dentry); + } + + if (no_parent) { + if (!S_ISDIR(inode->i_mode)) + goto out; + sfh->parent_ino = sfh->ino; + sfh->hash = 0; + } + sfh->ino = ceph_ino(inode); + sfh->snapid = snapid; + + *max_len = snap_handle_length; + ret = FILEID_BTRFS_WITH_PARENT; +out: + dout("encode_snapfh %llx.%llx ret=%d\n", ceph_vinop(inode), ret); + return ret; +} + static int ceph_encode_fh(struct inode *inode, u32 *rawfh, int *max_len, struct inode *parent_inode) { + const static int handle_length = + sizeof(struct ceph_nfs_fh) >> 2; + const static int connected_handle_length = + sizeof(struct ceph_nfs_confh) >> 2; int type; - struct ceph_nfs_fh *fh = (void *)rawfh; - struct ceph_nfs_confh *cfh = (void *)rawfh; - int connected_handle_length = sizeof(*cfh)/4; - int handle_length = sizeof(*fh)/4; - /* don't re-export snaps */ if (ceph_snap(inode) != CEPH_NOSNAP) - return -EINVAL; + return ceph_encode_snapfh(inode, rawfh, max_len, parent_inode); if (parent_inode && (*max_len < connected_handle_length)) { *max_len = connected_handle_length; @@ -44,6 +103,7 @@ static int ceph_encode_fh(struct inode *inode, u32 *rawfh, int *max_len, } if (parent_inode) { + struct ceph_nfs_confh *cfh = (void *)rawfh; dout("encode_fh %llx with parent %llx\n", ceph_ino(inode), ceph_ino(parent_inode)); cfh->ino = ceph_ino(inode); @@ -51,6 +111,7 @@ static int ceph_encode_fh(struct inode *inode, u32 *rawfh, int *max_len, *max_len = connected_handle_length; type = FILEID_INO32_GEN_PARENT; } else { + struct ceph_nfs_fh *fh = (void *)rawfh; dout("encode_fh %llx\n", ceph_ino(inode)); fh->ino = ceph_ino(inode); *max_len = handle_length; @@ -59,7 +120,7 @@ static int ceph_encode_fh(struct inode *inode, u32 *rawfh, int *max_len, return type; } -static struct dentry *__fh_to_dentry(struct super_block *sb, u64 ino) +static struct inode *__lookup_inode(struct super_block *sb, u64 ino) { struct ceph_mds_client *mdsc = ceph_sb_to_client(sb)->mdsc; struct inode *inode; @@ -81,7 +142,7 @@ static struct dentry *__fh_to_dentry(struct super_block *sb, u64 ino) mask = CEPH_STAT_CAP_INODE; if (ceph_security_xattr_wanted(d_inode(sb->s_root))) mask |= CEPH_CAP_XATTR_SHARED; - req->r_args.getattr.mask = cpu_to_le32(mask); + req->r_args.lookupino.mask = cpu_to_le32(mask); req->r_ino1 = vino; req->r_num_caps = 1; @@ -91,16 +152,114 @@ static struct dentry *__fh_to_dentry(struct super_block *sb, u64 ino) ihold(inode); ceph_mdsc_put_request(req); if (!inode) - return ERR_PTR(-ESTALE); - if (inode->i_nlink == 0) { - iput(inode); - return ERR_PTR(-ESTALE); - } + return err < 0 ? ERR_PTR(err) : ERR_PTR(-ESTALE); } + return inode; +} + +struct inode *ceph_lookup_inode(struct super_block *sb, u64 ino) +{ + struct inode *inode = __lookup_inode(sb, ino); + if (IS_ERR(inode)) + return inode; + if (inode->i_nlink == 0) { + iput(inode); + return ERR_PTR(-ESTALE); + } + return inode; +} +static struct dentry *__fh_to_dentry(struct super_block *sb, u64 ino) +{ + struct inode *inode = __lookup_inode(sb, ino); + if (IS_ERR(inode)) + return ERR_CAST(inode); + if (inode->i_nlink == 0) { + iput(inode); + return ERR_PTR(-ESTALE); + } return d_obtain_alias(inode); } +static struct dentry *__snapfh_to_dentry(struct super_block *sb, + struct ceph_nfs_snapfh *sfh, + bool want_parent) +{ + struct ceph_mds_client *mdsc = ceph_sb_to_client(sb)->mdsc; + struct ceph_mds_request *req; + struct inode *inode; + struct ceph_vino vino; + int mask; + int err; + bool unlinked = false; + + if (want_parent) { + vino.ino = sfh->parent_ino; + if (sfh->snapid == CEPH_SNAPDIR) + vino.snap = CEPH_NOSNAP; + else if (sfh->ino == sfh->parent_ino) + vino.snap = CEPH_SNAPDIR; + else + vino.snap = sfh->snapid; + } else { + vino.ino = sfh->ino; + vino.snap = sfh->snapid; + } + inode = ceph_find_inode(sb, vino); + if (inode) + return d_obtain_alias(inode); + + req = ceph_mdsc_create_request(mdsc, CEPH_MDS_OP_LOOKUPINO, + USE_ANY_MDS); + if (IS_ERR(req)) + return ERR_CAST(req); + + mask = CEPH_STAT_CAP_INODE; + if (ceph_security_xattr_wanted(d_inode(sb->s_root))) + mask |= CEPH_CAP_XATTR_SHARED; + req->r_args.lookupino.mask = cpu_to_le32(mask); + if (vino.snap < CEPH_NOSNAP) { + req->r_args.lookupino.snapid = cpu_to_le64(vino.snap); + if (!want_parent && sfh->ino != sfh->parent_ino) { + req->r_args.lookupino.parent = + cpu_to_le64(sfh->parent_ino); + req->r_args.lookupino.hash = + cpu_to_le32(sfh->hash); + } + } + + req->r_ino1 = vino; + req->r_num_caps = 1; + err = ceph_mdsc_do_request(mdsc, NULL, req); + inode = req->r_target_inode; + if (inode) { + if (vino.snap == CEPH_SNAPDIR) { + if (inode->i_nlink == 0) + unlinked = true; + inode = ceph_get_snapdir(inode); + } else if (ceph_snap(inode) == vino.snap) { + ihold(inode); + } else { + /* mds does not support lookup snapped inode */ + err = -EOPNOTSUPP; + inode = NULL; + } + } + ceph_mdsc_put_request(req); + + if (want_parent) { + dout("snapfh_to_parent %llx.%llx\n err=%d\n", + vino.ino, vino.snap, err); + } else { + dout("snapfh_to_dentry %llx.%llx parent %llx hash %x err=%d", + vino.ino, vino.snap, sfh->parent_ino, sfh->hash, err); + } + if (!inode) + return ERR_PTR(-ESTALE); + /* see comments in ceph_get_parent() */ + return unlinked ? d_obtain_root(inode) : d_obtain_alias(inode); +} + /* * convert regular fh to dentry */ @@ -110,6 +269,11 @@ static struct dentry *ceph_fh_to_dentry(struct super_block *sb, { struct ceph_nfs_fh *fh = (void *)fid->raw; + if (fh_type == FILEID_BTRFS_WITH_PARENT) { + struct ceph_nfs_snapfh *sfh = (void *)fid->raw; + return __snapfh_to_dentry(sb, sfh, false); + } + if (fh_type != FILEID_INO32_GEN && fh_type != FILEID_INO32_GEN_PARENT) return NULL; @@ -163,13 +327,49 @@ static struct dentry *__get_parent(struct super_block *sb, static struct dentry *ceph_get_parent(struct dentry *child) { - /* don't re-export snaps */ - if (ceph_snap(d_inode(child)) != CEPH_NOSNAP) - return ERR_PTR(-EINVAL); - - dout("get_parent %p ino %llx.%llx\n", - child, ceph_vinop(d_inode(child))); - return __get_parent(child->d_sb, child, 0); + struct inode *inode = d_inode(child); + struct dentry *dn; + + if (ceph_snap(inode) != CEPH_NOSNAP) { + struct inode* dir; + bool unlinked = false; + /* do not support non-directory */ + if (!d_is_dir(child)) { + dn = ERR_PTR(-EINVAL); + goto out; + } + dir = __lookup_inode(inode->i_sb, ceph_ino(inode)); + if (IS_ERR(dir)) { + dn = ERR_CAST(dir); + goto out; + } + /* There can be multiple paths to access snapped inode. + * For simplicity, treat snapdir of head inode as parent */ + if (ceph_snap(inode) != CEPH_SNAPDIR) { + struct inode *snapdir = ceph_get_snapdir(dir); + if (dir->i_nlink == 0) + unlinked = true; + iput(dir); + if (IS_ERR(snapdir)) { + dn = ERR_CAST(snapdir); + goto out; + } + dir = snapdir; + } + /* If directory has already been deleted, futher get_parent + * will fail. Do not mark snapdir dentry as disconnected, + * this prevent exportfs from doing futher get_parent. */ + if (unlinked) + dn = d_obtain_root(dir); + else + dn = d_obtain_alias(dir); + } else { + dn = __get_parent(child->d_sb, child, 0); + } +out: + dout("get_parent %p ino %llx.%llx err=%ld\n", + child, ceph_vinop(inode), (IS_ERR(dn) ? PTR_ERR(dn) : 0)); + return dn; } /* @@ -182,6 +382,11 @@ static struct dentry *ceph_fh_to_parent(struct super_block *sb, struct ceph_nfs_confh *cfh = (void *)fid->raw; struct dentry *dentry; + if (fh_type == FILEID_BTRFS_WITH_PARENT) { + struct ceph_nfs_snapfh *sfh = (void *)fid->raw; + return __snapfh_to_dentry(sb, sfh, true); + } + if (fh_type != FILEID_INO32_GEN_PARENT) return NULL; if (fh_len < sizeof(*cfh) / 4) @@ -194,14 +399,115 @@ static struct dentry *ceph_fh_to_parent(struct super_block *sb, return dentry; } +static int __get_snap_name(struct dentry *parent, char *name, + struct dentry *child) +{ + struct inode *inode = d_inode(child); + struct inode *dir = d_inode(parent); + struct ceph_fs_client *fsc = ceph_inode_to_client(inode); + struct ceph_mds_request *req = NULL; + char *last_name = NULL; + unsigned next_offset = 2; + int err = -EINVAL; + + if (ceph_ino(inode) != ceph_ino(dir)) + goto out; + if (ceph_snap(inode) == CEPH_SNAPDIR) { + if (ceph_snap(dir) == CEPH_NOSNAP) { + strcpy(name, fsc->mount_options->snapdir_name); + err = 0; + } + goto out; + } + if (ceph_snap(dir) != CEPH_SNAPDIR) + goto out; + + while (1) { + struct ceph_mds_reply_info_parsed *rinfo; + struct ceph_mds_reply_dir_entry *rde; + int i; + + req = ceph_mdsc_create_request(fsc->mdsc, CEPH_MDS_OP_LSSNAP, + USE_AUTH_MDS); + if (IS_ERR(req)) { + err = PTR_ERR(req); + req = NULL; + goto out; + } + err = ceph_alloc_readdir_reply_buffer(req, inode); + if (err) + goto out; + + req->r_direct_mode = USE_AUTH_MDS; + req->r_readdir_offset = next_offset; + req->r_args.readdir.flags = + cpu_to_le16(CEPH_READDIR_REPLY_BITFLAGS); + if (last_name) { + req->r_path2 = last_name; + last_name = NULL; + } + + req->r_inode = dir; + ihold(dir); + req->r_dentry = dget(parent); + + inode_lock(dir); + err = ceph_mdsc_do_request(fsc->mdsc, NULL, req); + inode_unlock(dir); + + if (err < 0) + goto out; + + rinfo = &req->r_reply_info; + for (i = 0; i < rinfo->dir_nr; i++) { + rde = rinfo->dir_entries + i; + BUG_ON(!rde->inode.in); + if (ceph_snap(inode) == + le64_to_cpu(rde->inode.in->snapid)) { + memcpy(name, rde->name, rde->name_len); + name[rde->name_len] = '\0'; + err = 0; + goto out; + } + } + + if (rinfo->dir_end) + break; + + BUG_ON(rinfo->dir_nr <= 0); + rde = rinfo->dir_entries + (rinfo->dir_nr - 1); + next_offset += rinfo->dir_nr; + last_name = kstrndup(rde->name, rde->name_len, GFP_KERNEL); + if (!last_name) { + err = -ENOMEM; + goto out; + } + + ceph_mdsc_put_request(req); + req = NULL; + } + err = -ENOENT; +out: + if (req) + ceph_mdsc_put_request(req); + kfree(last_name); + dout("get_snap_name %p ino %llx.%llx err=%d\n", + child, ceph_vinop(inode), err); + return err; +} + static int ceph_get_name(struct dentry *parent, char *name, struct dentry *child) { struct ceph_mds_client *mdsc; struct ceph_mds_request *req; + struct inode *inode = d_inode(child); int err; - mdsc = ceph_inode_to_client(d_inode(child))->mdsc; + if (ceph_snap(inode) != CEPH_NOSNAP) + return __get_snap_name(parent, name, child); + + mdsc = ceph_inode_to_client(inode)->mdsc; req = ceph_mdsc_create_request(mdsc, CEPH_MDS_OP_LOOKUPNAME, USE_ANY_MDS); if (IS_ERR(req)) @@ -209,8 +515,8 @@ static int ceph_get_name(struct dentry *parent, char *name, inode_lock(d_inode(parent)); - req->r_inode = d_inode(child); - ihold(d_inode(child)); + req->r_inode = inode; + ihold(inode); req->r_ino2 = ceph_vino(d_inode(parent)); req->r_parent = d_inode(parent); set_bit(CEPH_MDS_R_PARENT_LOCKED, &req->r_req_flags); @@ -224,10 +530,10 @@ static int ceph_get_name(struct dentry *parent, char *name, memcpy(name, rinfo->dname, rinfo->dname_len); name[rinfo->dname_len] = 0; dout("get_name %p ino %llx.%llx name %s\n", - child, ceph_vinop(d_inode(child)), name); + child, ceph_vinop(inode), name); } else { dout("get_name %p ino %llx.%llx err %d\n", - child, ceph_vinop(d_inode(child)), err); + child, ceph_vinop(inode), err); } ceph_mdsc_put_request(req); diff --git a/fs/ceph/file.c b/fs/ceph/file.c index 84725b53ac21..305daf043eb0 100644 --- a/fs/ceph/file.c +++ b/fs/ceph/file.c @@ -929,7 +929,7 @@ ceph_direct_read_write(struct kiocb *iocb, struct iov_iter *iter, dout("sync_direct_%s on file %p %lld~%u snapc %p seq %lld\n", (write ? "write" : "read"), file, pos, (unsigned)count, - snapc, snapc->seq); + snapc, snapc ? snapc->seq : 0); ret = filemap_write_and_wait_range(inode->i_mapping, pos, pos + count - 1); diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c index 35dae6d5493a..f85355bf49c4 100644 --- a/fs/ceph/inode.c +++ b/fs/ceph/inode.c @@ -2266,43 +2266,72 @@ int ceph_permission(struct inode *inode, int mask) return err; } +/* Craft a mask of needed caps given a set of requested statx attrs. */ +static int statx_to_caps(u32 want) +{ + int mask = 0; + + if (want & (STATX_MODE|STATX_UID|STATX_GID|STATX_CTIME)) + mask |= CEPH_CAP_AUTH_SHARED; + + if (want & (STATX_NLINK|STATX_CTIME)) + mask |= CEPH_CAP_LINK_SHARED; + + if (want & (STATX_ATIME|STATX_MTIME|STATX_CTIME|STATX_SIZE| + STATX_BLOCKS)) + mask |= CEPH_CAP_FILE_SHARED; + + if (want & (STATX_CTIME)) + mask |= CEPH_CAP_XATTR_SHARED; + + return mask; +} + /* - * Get all attributes. Hopefully somedata we'll have a statlite() - * and can limit the fields we require to be accurate. + * Get all the attributes. If we have sufficient caps for the requested attrs, + * then we can avoid talking to the MDS at all. */ int ceph_getattr(const struct path *path, struct kstat *stat, u32 request_mask, unsigned int flags) { struct inode *inode = d_inode(path->dentry); struct ceph_inode_info *ci = ceph_inode(inode); - int err; + int err = 0; - err = ceph_do_getattr(inode, CEPH_STAT_CAP_INODE_ALL, false); - if (!err) { - generic_fillattr(inode, stat); - stat->ino = ceph_translate_ino(inode->i_sb, inode->i_ino); - if (ceph_snap(inode) == CEPH_NOSNAP) - stat->dev = inode->i_sb->s_dev; + /* Skip the getattr altogether if we're asked not to sync */ + if (!(flags & AT_STATX_DONT_SYNC)) { + err = ceph_do_getattr(inode, statx_to_caps(request_mask), + flags & AT_STATX_FORCE_SYNC); + if (err) + return err; + } + + generic_fillattr(inode, stat); + stat->ino = ceph_translate_ino(inode->i_sb, inode->i_ino); + if (ceph_snap(inode) == CEPH_NOSNAP) + stat->dev = inode->i_sb->s_dev; + else + stat->dev = ci->i_snapid_map ? ci->i_snapid_map->dev : 0; + + if (S_ISDIR(inode->i_mode)) { + if (ceph_test_mount_opt(ceph_sb_to_client(inode->i_sb), + RBYTES)) + stat->size = ci->i_rbytes; else - stat->dev = ci->i_snapid_map ? ci->i_snapid_map->dev : 0; - - if (S_ISDIR(inode->i_mode)) { - if (ceph_test_mount_opt(ceph_sb_to_client(inode->i_sb), - RBYTES)) - stat->size = ci->i_rbytes; - else - stat->size = ci->i_files + ci->i_subdirs; - stat->blocks = 0; - stat->blksize = 65536; - /* - * Some applications rely on the number of st_nlink - * value on directories to be either 0 (if unlinked) - * or 2 + number of subdirectories. - */ - if (stat->nlink == 1) - /* '.' + '..' + subdirs */ - stat->nlink = 1 + 1 + ci->i_subdirs; - } + stat->size = ci->i_files + ci->i_subdirs; + stat->blocks = 0; + stat->blksize = 65536; + /* + * Some applications rely on the number of st_nlink + * value on directories to be either 0 (if unlinked) + * or 2 + number of subdirectories. + */ + if (stat->nlink == 1) + /* '.' + '..' + subdirs */ + stat->nlink = 1 + 1 + ci->i_subdirs; } + + /* Mask off any higher bits (e.g. btime) until we have support */ + stat->result_mask = request_mask & STATX_BASIC_STATS; return err; } diff --git a/fs/ceph/locks.c b/fs/ceph/locks.c index 9dae2ec7e1fa..ac9b53b89365 100644 --- a/fs/ceph/locks.c +++ b/fs/ceph/locks.c @@ -237,15 +237,6 @@ int ceph_lock(struct file *file, int cmd, struct file_lock *fl) spin_lock(&ci->i_ceph_lock); if (ci->i_ceph_flags & CEPH_I_ERROR_FILELOCK) { err = -EIO; - } else if (op == CEPH_MDS_OP_SETFILELOCK) { - /* - * increasing i_filelock_ref closes race window between - * handling request reply and adding file_lock struct to - * inode. Otherwise, i_auth_cap may get trimmed in the - * window. Caller function will decrease the counter. - */ - fl->fl_ops = &ceph_fl_lock_ops; - atomic_inc(&ci->i_filelock_ref); } spin_unlock(&ci->i_ceph_lock); if (err < 0) { @@ -299,10 +290,6 @@ int ceph_flock(struct file *file, int cmd, struct file_lock *fl) spin_lock(&ci->i_ceph_lock); if (ci->i_ceph_flags & CEPH_I_ERROR_FILELOCK) { err = -EIO; - } else { - /* see comment in ceph_lock */ - fl->fl_ops = &ceph_fl_lock_ops; - atomic_inc(&ci->i_filelock_ref); } spin_unlock(&ci->i_ceph_lock); if (err < 0) { diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c index 9049c2a3e972..959b1bf7c327 100644 --- a/fs/ceph/mds_client.c +++ b/fs/ceph/mds_client.c @@ -550,15 +550,9 @@ void ceph_put_mds_session(struct ceph_mds_session *s) struct ceph_mds_session *__ceph_lookup_mds_session(struct ceph_mds_client *mdsc, int mds) { - struct ceph_mds_session *session; - if (mds >= mdsc->max_sessions || !mdsc->sessions[mds]) return NULL; - session = mdsc->sessions[mds]; - dout("lookup_mds_session %p %d\n", session, - refcount_read(&session->s_ref)); - get_session(session); - return session; + return get_session(mdsc->sessions[mds]); } static bool __have_session(struct ceph_mds_client *mdsc, int mds) @@ -1284,9 +1278,9 @@ static void cleanup_session_requests(struct ceph_mds_client *mdsc, * * Caller must hold session s_mutex. */ -static int iterate_session_caps(struct ceph_mds_session *session, - int (*cb)(struct inode *, struct ceph_cap *, - void *), void *arg) +int ceph_iterate_session_caps(struct ceph_mds_session *session, + int (*cb)(struct inode *, struct ceph_cap *, + void *), void *arg) { struct list_head *p; struct ceph_cap *cap; @@ -1451,7 +1445,7 @@ static void remove_session_caps(struct ceph_mds_session *session) LIST_HEAD(dispose); dout("remove_session_caps on %p\n", session); - iterate_session_caps(session, remove_session_caps_cb, fsc); + ceph_iterate_session_caps(session, remove_session_caps_cb, fsc); wake_up_all(&fsc->mdsc->cap_flushing_wq); @@ -1534,8 +1528,8 @@ static int wake_up_session_cb(struct inode *inode, struct ceph_cap *cap, static void wake_up_session_caps(struct ceph_mds_session *session, int ev) { dout("wake_up_session_caps %p mds%d\n", session, session->s_mds); - iterate_session_caps(session, wake_up_session_cb, - (void *)(unsigned long)ev); + ceph_iterate_session_caps(session, wake_up_session_cb, + (void *)(unsigned long)ev); } /* @@ -1768,7 +1762,7 @@ int ceph_trim_caps(struct ceph_mds_client *mdsc, session->s_mds, session->s_nr_caps, max_caps, trim_caps); if (trim_caps > 0) { session->s_trim_caps = trim_caps; - iterate_session_caps(session, trim_caps_cb, session); + ceph_iterate_session_caps(session, trim_caps_cb, session); dout("trim_caps mds%d done: %d / %d, trimmed %d\n", session->s_mds, session->s_nr_caps, max_caps, trim_caps - session->s_trim_caps); @@ -1861,7 +1855,8 @@ again: num_cap_releases--; head = msg->front.iov_base; - le32_add_cpu(&head->num, 1); + put_unaligned_le32(get_unaligned_le32(&head->num) + 1, + &head->num); item = msg->front.iov_base + msg->front.iov_len; item->ino = cpu_to_le64(cap->cap_ino); item->cap_id = cpu_to_le64(cap->cap_id); @@ -2089,43 +2084,29 @@ static inline u64 __get_oldest_tid(struct ceph_mds_client *mdsc) * Encode hidden .snap dirs as a double /, i.e. * foo/.snap/bar -> foo//bar */ -char *ceph_mdsc_build_path(struct dentry *dentry, int *plen, u64 *base, +char *ceph_mdsc_build_path(struct dentry *dentry, int *plen, u64 *pbase, int stop_on_nosnap) { struct dentry *temp; char *path; - int len, pos; + int pos; unsigned seq; + u64 base; if (!dentry) return ERR_PTR(-EINVAL); -retry: - len = 0; - seq = read_seqbegin(&rename_lock); - rcu_read_lock(); - for (temp = dentry; !IS_ROOT(temp);) { - struct inode *inode = d_inode(temp); - if (inode && ceph_snap(inode) == CEPH_SNAPDIR) - len++; /* slash only */ - else if (stop_on_nosnap && inode && - ceph_snap(inode) == CEPH_NOSNAP) - break; - else - len += 1 + temp->d_name.len; - temp = temp->d_parent; - } - rcu_read_unlock(); - if (len) - len--; /* no leading '/' */ - - path = kmalloc(len+1, GFP_NOFS); + path = __getname(); if (!path) return ERR_PTR(-ENOMEM); - pos = len; - path[pos] = 0; /* trailing null */ +retry: + pos = PATH_MAX - 1; + path[pos] = '\0'; + + seq = read_seqbegin(&rename_lock); rcu_read_lock(); - for (temp = dentry; !IS_ROOT(temp) && pos != 0; ) { + temp = dentry; + for (;;) { struct inode *inode; spin_lock(&temp->d_lock); @@ -2143,83 +2124,54 @@ retry: spin_unlock(&temp->d_lock); break; } - strncpy(path + pos, temp->d_name.name, - temp->d_name.len); + memcpy(path + pos, temp->d_name.name, temp->d_name.len); } spin_unlock(&temp->d_lock); - if (pos) - path[--pos] = '/'; temp = temp->d_parent; + + /* Are we at the root? */ + if (IS_ROOT(temp)) + break; + + /* Are we out of buffer? */ + if (--pos < 0) + break; + + path[pos] = '/'; } + base = ceph_ino(d_inode(temp)); rcu_read_unlock(); - if (pos != 0 || read_seqretry(&rename_lock, seq)) { + if (pos < 0 || read_seqretry(&rename_lock, seq)) { pr_err("build_path did not end path lookup where " - "expected, namelen is %d, pos is %d\n", len, pos); + "expected, pos is %d\n", pos); /* presumably this is only possible if racing with a rename of one of the parent directories (we can not lock the dentries above us to prevent this, but retrying should be harmless) */ - kfree(path); goto retry; } - *base = ceph_ino(d_inode(temp)); - *plen = len; + *pbase = base; + *plen = PATH_MAX - 1 - pos; dout("build_path on %p %d built %llx '%.*s'\n", - dentry, d_count(dentry), *base, len, path); - return path; -} - -/* Duplicate the dentry->d_name.name safely */ -static int clone_dentry_name(struct dentry *dentry, const char **ppath, - int *ppathlen) -{ - u32 len; - char *name; - -retry: - len = READ_ONCE(dentry->d_name.len); - name = kmalloc(len + 1, GFP_NOFS); - if (!name) - return -ENOMEM; - - spin_lock(&dentry->d_lock); - if (dentry->d_name.len != len) { - spin_unlock(&dentry->d_lock); - kfree(name); - goto retry; - } - memcpy(name, dentry->d_name.name, len); - spin_unlock(&dentry->d_lock); - - name[len] = '\0'; - *ppath = name; - *ppathlen = len; - return 0; + dentry, d_count(dentry), base, *plen, path + pos); + return path + pos; } static int build_dentry_path(struct dentry *dentry, struct inode *dir, const char **ppath, int *ppathlen, u64 *pino, bool *pfreepath, bool parent_locked) { - int ret; char *path; rcu_read_lock(); if (!dir) dir = d_inode_rcu(dentry->d_parent); - if (dir && ceph_snap(dir) == CEPH_NOSNAP) { + if (dir && parent_locked && ceph_snap(dir) == CEPH_NOSNAP) { *pino = ceph_ino(dir); rcu_read_unlock(); - if (parent_locked) { - *ppath = dentry->d_name.name; - *ppathlen = dentry->d_name.len; - } else { - ret = clone_dentry_name(dentry, ppath, ppathlen); - if (ret) - return ret; - *pfreepath = true; - } + *ppath = dentry->d_name.name; + *ppathlen = dentry->d_name.len; return 0; } rcu_read_unlock(); @@ -2331,9 +2283,9 @@ static struct ceph_msg *create_request_message(struct ceph_mds_client *mdsc, (!!req->r_inode_drop + !!req->r_dentry_drop + !!req->r_old_inode_drop + !!req->r_old_dentry_drop); if (req->r_dentry_drop) - len += req->r_dentry->d_name.len; + len += pathlen1; if (req->r_old_dentry_drop) - len += req->r_old_dentry->d_name.len; + len += pathlen2; msg = ceph_msg_new2(CEPH_MSG_CLIENT_REQUEST, len, 1, GFP_NOFS, false); if (!msg) { @@ -2410,10 +2362,10 @@ static struct ceph_msg *create_request_message(struct ceph_mds_client *mdsc, out_free2: if (freepath2) - kfree((char *)path2); + ceph_mdsc_free_path((char *)path2, pathlen2); out_free1: if (freepath1) - kfree((char *)path1); + ceph_mdsc_free_path((char *)path1, pathlen1); out: return msg; } @@ -2427,8 +2379,7 @@ static void complete_request(struct ceph_mds_client *mdsc, { if (req->r_callback) req->r_callback(mdsc, req); - else - complete_all(&req->r_completion); + complete_all(&req->r_completion); } /* @@ -2670,28 +2621,11 @@ static void kick_requests(struct ceph_mds_client *mdsc, int mds) } } -void ceph_mdsc_submit_request(struct ceph_mds_client *mdsc, +int ceph_mdsc_submit_request(struct ceph_mds_client *mdsc, struct inode *dir, struct ceph_mds_request *req) { - dout("submit_request on %p\n", req); - mutex_lock(&mdsc->mutex); - __register_request(mdsc, req, NULL); - __do_request(mdsc, req); - mutex_unlock(&mdsc->mutex); -} - -/* - * Synchrously perform an mds request. Take care of all of the - * session setup, forwarding, retry details. - */ -int ceph_mdsc_do_request(struct ceph_mds_client *mdsc, - struct inode *dir, - struct ceph_mds_request *req) -{ int err; - dout("do_request on %p\n", req); - /* take CAP_PIN refs for r_inode, r_parent, r_old_dentry */ if (req->r_inode) ceph_get_cap_refs(ceph_inode(req->r_inode), CEPH_CAP_PIN); @@ -2701,18 +2635,21 @@ int ceph_mdsc_do_request(struct ceph_mds_client *mdsc, ceph_get_cap_refs(ceph_inode(req->r_old_dentry_dir), CEPH_CAP_PIN); - /* issue */ + dout("submit_request on %p for inode %p\n", req, dir); mutex_lock(&mdsc->mutex); __register_request(mdsc, req, dir); __do_request(mdsc, req); + err = req->r_err; + mutex_unlock(&mdsc->mutex); + return err; +} - if (req->r_err) { - err = req->r_err; - goto out; - } +static int ceph_mdsc_wait_request(struct ceph_mds_client *mdsc, + struct ceph_mds_request *req) +{ + int err; /* wait */ - mutex_unlock(&mdsc->mutex); dout("do_request waiting\n"); if (!req->r_timeout && req->r_wait_for_completion) { err = req->r_wait_for_completion(mdsc, req); @@ -2753,8 +2690,26 @@ int ceph_mdsc_do_request(struct ceph_mds_client *mdsc, err = req->r_err; } -out: mutex_unlock(&mdsc->mutex); + return err; +} + +/* + * Synchrously perform an mds request. Take care of all of the + * session setup, forwarding, retry details. + */ +int ceph_mdsc_do_request(struct ceph_mds_client *mdsc, + struct inode *dir, + struct ceph_mds_request *req) +{ + int err; + + dout("do_request on %p\n", req); + + /* issue */ + err = ceph_mdsc_submit_request(mdsc, dir, req); + if (!err) + err = ceph_mdsc_wait_request(mdsc, req); dout("do_request %p done, result %d\n", req, err); return err; } @@ -3485,7 +3440,7 @@ out_freeflocks: ceph_pagelist_encode_string(pagelist, path, pathlen); ceph_pagelist_append(pagelist, &rec, sizeof(rec.v1)); out_freepath: - kfree(path); + ceph_mdsc_free_path(path, pathlen); } out_err: @@ -3642,7 +3597,7 @@ static void send_mds_reconnect(struct ceph_mds_client *mdsc, recon_state.msg_version = 2; } /* trsaverse this session's caps */ - err = iterate_session_caps(session, encode_caps_cb, &recon_state); + err = ceph_iterate_session_caps(session, encode_caps_cb, &recon_state); spin_lock(&session->s_cap_lock); session->s_cap_reconnect = 0; @@ -4125,6 +4080,8 @@ int ceph_mdsc_init(struct ceph_fs_client *fsc) mdsc->max_sessions = 0; mdsc->stopping = 0; atomic64_set(&mdsc->quotarealms_count, 0); + mdsc->quotarealms_inodes = RB_ROOT; + mutex_init(&mdsc->quotarealms_inodes_mutex); mdsc->last_snap_seq = 0; init_rwsem(&mdsc->snap_rwsem); mdsc->snap_realms = RB_ROOT; @@ -4216,6 +4173,8 @@ void ceph_mdsc_pre_umount(struct ceph_mds_client *mdsc) * their inode/dcache refs */ ceph_msgr_flush(); + + ceph_cleanup_quotarealms_inodes(mdsc); } /* diff --git a/fs/ceph/mds_client.h b/fs/ceph/mds_client.h index 50385a481fdb..a83f28bc2387 100644 --- a/fs/ceph/mds_client.h +++ b/fs/ceph/mds_client.h @@ -326,6 +326,18 @@ struct ceph_snapid_map { }; /* + * node for list of quotarealm inodes that are not visible from the filesystem + * mountpoint, but required to handle, e.g. quotas. + */ +struct ceph_quotarealm_inode { + struct rb_node node; + u64 ino; + unsigned long timeout; /* last time a lookup failed for this inode */ + struct mutex mutex; + struct inode *inode; +}; + +/* * mds client state */ struct ceph_mds_client { @@ -344,6 +356,12 @@ struct ceph_mds_client { int stopping; /* true if shutting down */ atomic64_t quotarealms_count; /* # realms with quota */ + /* + * We keep a list of inodes we don't see in the mountpoint but that we + * need to track quota realms. + */ + struct rb_root quotarealms_inodes; + struct mutex quotarealms_inodes_mutex; /* * snap_rwsem will cover cap linkage into snaprealms, and @@ -447,8 +465,9 @@ extern int ceph_alloc_readdir_reply_buffer(struct ceph_mds_request *req, struct inode *dir); extern struct ceph_mds_request * ceph_mdsc_create_request(struct ceph_mds_client *mdsc, int op, int mode); -extern void ceph_mdsc_submit_request(struct ceph_mds_client *mdsc, - struct ceph_mds_request *req); +extern int ceph_mdsc_submit_request(struct ceph_mds_client *mdsc, + struct inode *dir, + struct ceph_mds_request *req); extern int ceph_mdsc_do_request(struct ceph_mds_client *mdsc, struct inode *dir, struct ceph_mds_request *req); @@ -468,8 +487,18 @@ extern void ceph_flush_cap_releases(struct ceph_mds_client *mdsc, struct ceph_mds_session *session); extern void ceph_queue_cap_reclaim_work(struct ceph_mds_client *mdsc); extern void ceph_reclaim_caps_nr(struct ceph_mds_client *mdsc, int nr); +extern int ceph_iterate_session_caps(struct ceph_mds_session *session, + int (*cb)(struct inode *, + struct ceph_cap *, void *), + void *arg); extern void ceph_mdsc_pre_umount(struct ceph_mds_client *mdsc); +static inline void ceph_mdsc_free_path(char *path, int len) +{ + if (path) + __putname(path - (PATH_MAX - 1 - len)); +} + extern char *ceph_mdsc_build_path(struct dentry *dentry, int *plen, u64 *base, int stop_on_nosnap); diff --git a/fs/ceph/mdsmap.c b/fs/ceph/mdsmap.c index 1a2c5d390f7f..701b4fb0fb5a 100644 --- a/fs/ceph/mdsmap.c +++ b/fs/ceph/mdsmap.c @@ -205,7 +205,7 @@ struct ceph_mdsmap *ceph_mdsmap_decode(void **p, void *end) dout("mdsmap_decode %d/%d %lld mds%d.%d %s %s\n", i+1, n, global_id, mds, inc, - ceph_pr_addr(&addr.in_addr), + ceph_pr_addr(&addr), ceph_mds_state_name(state)); if (mds < 0 || state <= 0) diff --git a/fs/ceph/quota.c b/fs/ceph/quota.c index 9455d3aef0c3..c4522212872c 100644 --- a/fs/ceph/quota.c +++ b/fs/ceph/quota.c @@ -22,7 +22,16 @@ void ceph_adjust_quota_realms_count(struct inode *inode, bool inc) static inline bool ceph_has_realms_with_quotas(struct inode *inode) { struct ceph_mds_client *mdsc = ceph_inode_to_client(inode)->mdsc; - return atomic64_read(&mdsc->quotarealms_count) > 0; + struct super_block *sb = mdsc->fsc->sb; + + if (atomic64_read(&mdsc->quotarealms_count) > 0) + return true; + /* if root is the real CephFS root, we don't have quota realms */ + if (sb->s_root->d_inode && + (sb->s_root->d_inode->i_ino == CEPH_INO_ROOT)) + return false; + /* otherwise, we can't know for sure */ + return true; } void ceph_handle_quota(struct ceph_mds_client *mdsc, @@ -68,6 +77,108 @@ void ceph_handle_quota(struct ceph_mds_client *mdsc, iput(inode); } +static struct ceph_quotarealm_inode * +find_quotarealm_inode(struct ceph_mds_client *mdsc, u64 ino) +{ + struct ceph_quotarealm_inode *qri = NULL; + struct rb_node **node, *parent = NULL; + + mutex_lock(&mdsc->quotarealms_inodes_mutex); + node = &(mdsc->quotarealms_inodes.rb_node); + while (*node) { + parent = *node; + qri = container_of(*node, struct ceph_quotarealm_inode, node); + + if (ino < qri->ino) + node = &((*node)->rb_left); + else if (ino > qri->ino) + node = &((*node)->rb_right); + else + break; + } + if (!qri || (qri->ino != ino)) { + /* Not found, create a new one and insert it */ + qri = kmalloc(sizeof(*qri), GFP_KERNEL); + if (qri) { + qri->ino = ino; + qri->inode = NULL; + qri->timeout = 0; + mutex_init(&qri->mutex); + rb_link_node(&qri->node, parent, node); + rb_insert_color(&qri->node, &mdsc->quotarealms_inodes); + } else + pr_warn("Failed to alloc quotarealms_inode\n"); + } + mutex_unlock(&mdsc->quotarealms_inodes_mutex); + + return qri; +} + +/* + * This function will try to lookup a realm inode which isn't visible in the + * filesystem mountpoint. A list of these kind of inodes (not visible) is + * maintained in the mdsc and freed only when the filesystem is umounted. + * + * Note that these inodes are kept in this list even if the lookup fails, which + * allows to prevent useless lookup requests. + */ +static struct inode *lookup_quotarealm_inode(struct ceph_mds_client *mdsc, + struct super_block *sb, + struct ceph_snap_realm *realm) +{ + struct ceph_quotarealm_inode *qri; + struct inode *in; + + qri = find_quotarealm_inode(mdsc, realm->ino); + if (!qri) + return NULL; + + mutex_lock(&qri->mutex); + if (qri->inode) { + /* A request has already returned the inode */ + mutex_unlock(&qri->mutex); + return qri->inode; + } + /* Check if this inode lookup has failed recently */ + if (qri->timeout && + time_before_eq(jiffies, qri->timeout)) { + mutex_unlock(&qri->mutex); + return NULL; + } + in = ceph_lookup_inode(sb, realm->ino); + if (IS_ERR(in)) { + pr_warn("Can't lookup inode %llx (err: %ld)\n", + realm->ino, PTR_ERR(in)); + qri->timeout = jiffies + msecs_to_jiffies(60 * 1000); /* XXX */ + } else { + qri->timeout = 0; + qri->inode = in; + } + mutex_unlock(&qri->mutex); + + return in; +} + +void ceph_cleanup_quotarealms_inodes(struct ceph_mds_client *mdsc) +{ + struct ceph_quotarealm_inode *qri; + struct rb_node *node; + + /* + * It should now be safe to clean quotarealms_inode tree without holding + * mdsc->quotarealms_inodes_mutex... + */ + mutex_lock(&mdsc->quotarealms_inodes_mutex); + while (!RB_EMPTY_ROOT(&mdsc->quotarealms_inodes)) { + node = rb_first(&mdsc->quotarealms_inodes); + qri = rb_entry(node, struct ceph_quotarealm_inode, node); + rb_erase(node, &mdsc->quotarealms_inodes); + iput(qri->inode); + kfree(qri); + } + mutex_unlock(&mdsc->quotarealms_inodes_mutex); +} + /* * This function walks through the snaprealm for an inode and returns the * ceph_snap_realm for the first snaprealm that has quotas set (either max_files @@ -76,9 +187,15 @@ void ceph_handle_quota(struct ceph_mds_client *mdsc, * * Note that the caller is responsible for calling ceph_put_snap_realm() on the * returned realm. + * + * Callers of this function need to hold mdsc->snap_rwsem. However, if there's + * a need to do an inode lookup, this rwsem will be temporarily dropped. Hence + * the 'retry' argument: if rwsem needs to be dropped and 'retry' is 'false' + * this function will return -EAGAIN; otherwise, the snaprealms walk-through + * will be restarted. */ static struct ceph_snap_realm *get_quota_realm(struct ceph_mds_client *mdsc, - struct inode *inode) + struct inode *inode, bool retry) { struct ceph_inode_info *ci = NULL; struct ceph_snap_realm *realm, *next; @@ -88,6 +205,7 @@ static struct ceph_snap_realm *get_quota_realm(struct ceph_mds_client *mdsc, if (ceph_snap(inode) != CEPH_NOSNAP) return NULL; +restart: realm = ceph_inode(inode)->i_snap_realm; if (realm) ceph_get_snap_realm(mdsc, realm); @@ -95,11 +213,25 @@ static struct ceph_snap_realm *get_quota_realm(struct ceph_mds_client *mdsc, pr_err_ratelimited("get_quota_realm: ino (%llx.%llx) " "null i_snap_realm\n", ceph_vinop(inode)); while (realm) { + bool has_inode; + spin_lock(&realm->inodes_with_caps_lock); - in = realm->inode ? igrab(realm->inode) : NULL; + has_inode = realm->inode; + in = has_inode ? igrab(realm->inode) : NULL; spin_unlock(&realm->inodes_with_caps_lock); - if (!in) + if (has_inode && !in) break; + if (!in) { + up_read(&mdsc->snap_rwsem); + in = lookup_quotarealm_inode(mdsc, inode->i_sb, realm); + down_read(&mdsc->snap_rwsem); + if (IS_ERR_OR_NULL(in)) + break; + ceph_put_snap_realm(mdsc, realm); + if (!retry) + return ERR_PTR(-EAGAIN); + goto restart; + } ci = ceph_inode(in); has_quota = __ceph_has_any_quota(ci); @@ -125,9 +257,22 @@ bool ceph_quota_is_same_realm(struct inode *old, struct inode *new) struct ceph_snap_realm *old_realm, *new_realm; bool is_same; +restart: + /* + * We need to lookup 2 quota realms atomically, i.e. with snap_rwsem. + * However, get_quota_realm may drop it temporarily. By setting the + * 'retry' parameter to 'false', we'll get -EAGAIN if the rwsem was + * dropped and we can then restart the whole operation. + */ down_read(&mdsc->snap_rwsem); - old_realm = get_quota_realm(mdsc, old); - new_realm = get_quota_realm(mdsc, new); + old_realm = get_quota_realm(mdsc, old, true); + new_realm = get_quota_realm(mdsc, new, false); + if (PTR_ERR(new_realm) == -EAGAIN) { + up_read(&mdsc->snap_rwsem); + if (old_realm) + ceph_put_snap_realm(mdsc, old_realm); + goto restart; + } is_same = (old_realm == new_realm); up_read(&mdsc->snap_rwsem); @@ -166,6 +311,7 @@ static bool check_quota_exceeded(struct inode *inode, enum quota_check_op op, return false; down_read(&mdsc->snap_rwsem); +restart: realm = ceph_inode(inode)->i_snap_realm; if (realm) ceph_get_snap_realm(mdsc, realm); @@ -173,12 +319,23 @@ static bool check_quota_exceeded(struct inode *inode, enum quota_check_op op, pr_err_ratelimited("check_quota_exceeded: ino (%llx.%llx) " "null i_snap_realm\n", ceph_vinop(inode)); while (realm) { + bool has_inode; + spin_lock(&realm->inodes_with_caps_lock); - in = realm->inode ? igrab(realm->inode) : NULL; + has_inode = realm->inode; + in = has_inode ? igrab(realm->inode) : NULL; spin_unlock(&realm->inodes_with_caps_lock); - if (!in) + if (has_inode && !in) break; - + if (!in) { + up_read(&mdsc->snap_rwsem); + in = lookup_quotarealm_inode(mdsc, inode->i_sb, realm); + down_read(&mdsc->snap_rwsem); + if (IS_ERR_OR_NULL(in)) + break; + ceph_put_snap_realm(mdsc, realm); + goto restart; + } ci = ceph_inode(in); spin_lock(&ci->i_ceph_lock); if (op == QUOTA_CHECK_MAX_FILES_OP) { @@ -314,7 +471,7 @@ bool ceph_quota_update_statfs(struct ceph_fs_client *fsc, struct kstatfs *buf) bool is_updated = false; down_read(&mdsc->snap_rwsem); - realm = get_quota_realm(mdsc, d_inode(fsc->sb->s_root)); + realm = get_quota_realm(mdsc, d_inode(fsc->sb->s_root), true); up_read(&mdsc->snap_rwsem); if (!realm) return false; diff --git a/fs/ceph/super.c b/fs/ceph/super.c index 285edda4fc3b..c864b44c8341 100644 --- a/fs/ceph/super.c +++ b/fs/ceph/super.c @@ -845,6 +845,12 @@ static void ceph_umount_begin(struct super_block *sb) return; } +static int ceph_remount(struct super_block *sb, int *flags, char *data) +{ + sync_filesystem(sb); + return 0; +} + static const struct super_operations ceph_super_ops = { .alloc_inode = ceph_alloc_inode, .destroy_inode = ceph_destroy_inode, @@ -853,6 +859,7 @@ static const struct super_operations ceph_super_ops = { .drop_inode = ceph_drop_inode, .sync_fs = ceph_sync_fs, .put_super = ceph_put_super, + .remount_fs = ceph_remount, .show_options = ceph_show_options, .statfs = ceph_statfs, .umount_begin = ceph_umount_begin, diff --git a/fs/ceph/super.h b/fs/ceph/super.h index c5b4a05905c0..6edab9a750f8 100644 --- a/fs/ceph/super.h +++ b/fs/ceph/super.h @@ -1083,6 +1083,7 @@ extern long ceph_ioctl(struct file *file, unsigned int cmd, unsigned long arg); /* export.c */ extern const struct export_operations ceph_export_ops; +struct inode *ceph_lookup_inode(struct super_block *sb, u64 ino); /* locks.c */ extern __init void ceph_flock_init(void); @@ -1133,5 +1134,6 @@ extern bool ceph_quota_is_max_bytes_approaching(struct inode *inode, loff_t newlen); extern bool ceph_quota_update_statfs(struct ceph_fs_client *fsc, struct kstatfs *buf); +extern void ceph_cleanup_quotarealms_inodes(struct ceph_mds_client *mdsc); #endif /* _FS_CEPH_SUPER_H */ diff --git a/fs/cifs/dns_resolve.c b/fs/cifs/dns_resolve.c index 7ede7306599f..1e21b2528cfb 100644 --- a/fs/cifs/dns_resolve.c +++ b/fs/cifs/dns_resolve.c @@ -77,7 +77,7 @@ dns_resolve_server_name_to_ip(const char *unc, char **ip_addr) goto name_is_IP_address; /* Perform the upcall */ - rc = dns_query(NULL, hostname, len, NULL, ip_addr, NULL); + rc = dns_query(NULL, hostname, len, NULL, ip_addr, NULL, false); if (rc < 0) cifs_dbg(FYI, "%s: unable to resolve: %*.*s\n", __func__, len, len, hostname); diff --git a/fs/nfs/dns_resolve.c b/fs/nfs/dns_resolve.c index a7d3df85736d..e6a700f01452 100644 --- a/fs/nfs/dns_resolve.c +++ b/fs/nfs/dns_resolve.c @@ -22,7 +22,7 @@ ssize_t nfs_dns_resolve_name(struct net *net, char *name, size_t namelen, char *ip_addr = NULL; int ip_len; - ip_len = dns_query(NULL, name, namelen, NULL, &ip_addr, NULL); + ip_len = dns_query(NULL, name, namelen, NULL, &ip_addr, NULL, false); if (ip_len > 0) ret = rpc_pton(net, ip_addr, ip_len, sa, salen); else diff --git a/include/linux/ceph/ceph_fs.h b/include/linux/ceph/ceph_fs.h index 4903deb0777a..3ac0feaf2b5e 100644 --- a/include/linux/ceph/ceph_fs.h +++ b/include/linux/ceph/ceph_fs.h @@ -436,6 +436,12 @@ union ceph_mds_request_args { __le64 length; /* num bytes to lock from start */ __u8 wait; /* will caller wait for lock to become available? */ } __attribute__ ((packed)) filelock_change; + struct { + __le32 mask; /* CEPH_CAP_* */ + __le64 snapid; + __le64 parent; + __le32 hash; + } __attribute__ ((packed)) lookupino; } __attribute__ ((packed)); #define CEPH_MDS_FLAG_REPLAY 1 /* this is a replayed op */ diff --git a/include/linux/ceph/messenger.h b/include/linux/ceph/messenger.h index 800a2128d411..23895d178149 100644 --- a/include/linux/ceph/messenger.h +++ b/include/linux/ceph/messenger.h @@ -323,7 +323,8 @@ struct ceph_connection { }; -extern const char *ceph_pr_addr(const struct sockaddr_storage *ss); +extern const char *ceph_pr_addr(const struct ceph_entity_addr *addr); + extern int ceph_parse_ips(const char *c, const char *end, struct ceph_entity_addr *addr, int max_count, int *count); diff --git a/include/linux/ceph/osdmap.h b/include/linux/ceph/osdmap.h index 5675b1f09bc5..e081b56f1c1d 100644 --- a/include/linux/ceph/osdmap.h +++ b/include/linux/ceph/osdmap.h @@ -110,17 +110,16 @@ struct ceph_object_id { int name_len; }; +#define __CEPH_OID_INITIALIZER(oid) { .name = (oid).inline_name } + +#define CEPH_DEFINE_OID_ONSTACK(oid) \ + struct ceph_object_id oid = __CEPH_OID_INITIALIZER(oid) + static inline void ceph_oid_init(struct ceph_object_id *oid) { - oid->name = oid->inline_name; - oid->name_len = 0; + *oid = (struct ceph_object_id) __CEPH_OID_INITIALIZER(*oid); } -#define CEPH_OID_INIT_ONSTACK(oid) \ - ({ ceph_oid_init(&oid); oid; }) -#define CEPH_DEFINE_OID_ONSTACK(oid) \ - struct ceph_object_id oid = CEPH_OID_INIT_ONSTACK(oid) - static inline bool ceph_oid_empty(const struct ceph_object_id *oid) { return oid->name == oid->inline_name && !oid->name_len; diff --git a/include/linux/device-mapper.h b/include/linux/device-mapper.h index b0672756d056..e1f51d607cc5 100644 --- a/include/linux/device-mapper.h +++ b/include/linux/device-mapper.h @@ -62,7 +62,8 @@ typedef int (*dm_clone_and_map_request_fn) (struct dm_target *ti, struct request *rq, union map_info *map_context, struct request **clone); -typedef void (*dm_release_clone_request_fn) (struct request *clone); +typedef void (*dm_release_clone_request_fn) (struct request *clone, + union map_info *map_context); /* * Returns: diff --git a/include/linux/dns_resolver.h b/include/linux/dns_resolver.h index 34a744a1bafc..f2b3ae22e6b7 100644 --- a/include/linux/dns_resolver.h +++ b/include/linux/dns_resolver.h @@ -27,6 +27,7 @@ #include <uapi/linux/dns_resolver.h> extern int dns_query(const char *type, const char *name, size_t namelen, - const char *options, char **_result, time64_t *_expiry); + const char *options, char **_result, time64_t *_expiry, + bool invalidate); #endif /* _LINUX_DNS_RESOLVER_H */ diff --git a/include/linux/list.h b/include/linux/list.h index d3b4db895340..e951228db4b2 100644 --- a/include/linux/list.h +++ b/include/linux/list.h @@ -789,7 +789,7 @@ static inline void hlist_add_behind(struct hlist_node *n, struct hlist_node *prev) { n->next = prev->next; - WRITE_ONCE(prev->next, n); + prev->next = n; n->pprev = &prev->next; if (n->next) diff --git a/include/linux/list_bl.h b/include/linux/list_bl.h index 3fc2cc57ba1b..ae1b541446c9 100644 --- a/include/linux/list_bl.h +++ b/include/linux/list_bl.h @@ -86,6 +86,32 @@ static inline void hlist_bl_add_head(struct hlist_bl_node *n, hlist_bl_set_first(h, n); } +static inline void hlist_bl_add_before(struct hlist_bl_node *n, + struct hlist_bl_node *next) +{ + struct hlist_bl_node **pprev = next->pprev; + + n->pprev = pprev; + n->next = next; + next->pprev = &n->next; + + /* pprev may be `first`, so be careful not to lose the lock bit */ + WRITE_ONCE(*pprev, + (struct hlist_bl_node *) + ((uintptr_t)n | ((uintptr_t)*pprev & LIST_BL_LOCKMASK))); +} + +static inline void hlist_bl_add_behind(struct hlist_bl_node *n, + struct hlist_bl_node *prev) +{ + n->next = prev->next; + n->pprev = &prev->next; + prev->next = n; + + if (n->next) + n->next->pprev = &n->next; +} + static inline void __hlist_bl_del(struct hlist_bl_node *n) { struct hlist_bl_node *next = n->next; diff --git a/include/linux/slab_def.h b/include/linux/slab_def.h index 9a5eafb7145b..abc7de77b988 100644 --- a/include/linux/slab_def.h +++ b/include/linux/slab_def.h @@ -61,9 +61,6 @@ struct kmem_cache { atomic_t allocmiss; atomic_t freehit; atomic_t freemiss; -#ifdef CONFIG_DEBUG_SLAB_LEAK - atomic_t store_user_clean; -#endif /* * If debugging is enabled, then the allocator can add additional diff --git a/include/linux/thermal.h b/include/linux/thermal.h index 4a22099ed8c0..15a4ca5d7099 100644 --- a/include/linux/thermal.h +++ b/include/linux/thermal.h @@ -442,10 +442,10 @@ void thermal_zone_device_update(struct thermal_zone_device *, enum thermal_notify_event); void thermal_zone_set_trips(struct thermal_zone_device *); -struct thermal_cooling_device *thermal_cooling_device_register(char *, void *, - const struct thermal_cooling_device_ops *); +struct thermal_cooling_device *thermal_cooling_device_register(const char *, + void *, const struct thermal_cooling_device_ops *); struct thermal_cooling_device * -thermal_of_cooling_device_register(struct device_node *np, char *, void *, +thermal_of_cooling_device_register(struct device_node *np, const char *, void *, const struct thermal_cooling_device_ops *); struct thermal_cooling_device * devm_thermal_of_cooling_device_register(struct device *dev, diff --git a/include/linux/wait_bit.h b/include/linux/wait_bit.h index 2b0072fa5e92..7dec36aecbd9 100644 --- a/include/linux/wait_bit.h +++ b/include/linux/wait_bit.h @@ -305,6 +305,19 @@ do { \ __ret; \ }) +#define __wait_var_event_interruptible(var, condition) \ + ___wait_var_event(var, condition, TASK_INTERRUPTIBLE, 0, 0, \ + schedule()) + +#define wait_var_event_interruptible(var, condition) \ +({ \ + int __ret = 0; \ + might_sleep(); \ + if (!(condition)) \ + __ret = __wait_var_event_interruptible(var, condition); \ + __ret; \ +}) + /** * clear_and_wake_up_bit - clear a bit and wake up anyone waiting on that bit * diff --git a/include/net/af_rxrpc.h b/include/net/af_rxrpc.h index 78c856cba4f5..93358bfc0e1b 100644 --- a/include/net/af_rxrpc.h +++ b/include/net/af_rxrpc.h @@ -45,6 +45,7 @@ struct rxrpc_call *rxrpc_kernel_begin_call(struct socket *, gfp_t, rxrpc_notify_rx_t, bool, + bool, unsigned int); int rxrpc_kernel_send_data(struct socket *, struct rxrpc_call *, struct msghdr *, size_t, @@ -68,5 +69,7 @@ u32 rxrpc_kernel_get_epoch(struct socket *, struct rxrpc_call *); bool rxrpc_kernel_get_reply_time(struct socket *, struct rxrpc_call *, ktime_t *); bool rxrpc_kernel_call_is_complete(struct rxrpc_call *); +void rxrpc_kernel_set_max_life(struct socket *, struct rxrpc_call *, + unsigned long); #endif /* _NET_RXRPC_H */ diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index fdfa173651eb..eae43952902e 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -542,10 +542,6 @@ config DEBUG_SLAB allocation as well as poisoning memory on free to catch use of freed memory. This can make kmalloc/kfree-intensive workloads much slower. -config DEBUG_SLAB_LEAK - bool "Memory leak debugging" - depends on DEBUG_SLAB - config SLUB_DEBUG_ON bool "SLUB debugging on by default" depends on SLUB && SLUB_DEBUG diff --git a/mm/slab.c b/mm/slab.c index 2915d912e89a..f7117ad9b3a3 100644 --- a/mm/slab.c +++ b/mm/slab.c @@ -362,29 +362,6 @@ static void **dbg_userword(struct kmem_cache *cachep, void *objp) #endif -#ifdef CONFIG_DEBUG_SLAB_LEAK - -static inline bool is_store_user_clean(struct kmem_cache *cachep) -{ - return atomic_read(&cachep->store_user_clean) == 1; -} - -static inline void set_store_user_clean(struct kmem_cache *cachep) -{ - atomic_set(&cachep->store_user_clean, 1); -} - -static inline void set_store_user_dirty(struct kmem_cache *cachep) -{ - if (is_store_user_clean(cachep)) - atomic_set(&cachep->store_user_clean, 0); -} - -#else -static inline void set_store_user_dirty(struct kmem_cache *cachep) {} - -#endif - /* * Do not go above this order unless 0 objects fit into the slab or * overridden on the command line. @@ -2552,11 +2529,6 @@ static void *slab_get_obj(struct kmem_cache *cachep, struct page *page) objp = index_to_obj(cachep, page, get_free_obj(page, page->active)); page->active++; -#if DEBUG - if (cachep->flags & SLAB_STORE_USER) - set_store_user_dirty(cachep); -#endif - return objp; } @@ -2762,10 +2734,8 @@ static void *cache_free_debugcheck(struct kmem_cache *cachep, void *objp, *dbg_redzone1(cachep, objp) = RED_INACTIVE; *dbg_redzone2(cachep, objp) = RED_INACTIVE; } - if (cachep->flags & SLAB_STORE_USER) { - set_store_user_dirty(cachep); + if (cachep->flags & SLAB_STORE_USER) *dbg_userword(cachep, objp) = (void *)caller; - } objnr = obj_to_index(cachep, page, objp); @@ -4184,200 +4154,6 @@ ssize_t slabinfo_write(struct file *file, const char __user *buffer, return res; } -#ifdef CONFIG_DEBUG_SLAB_LEAK - -static inline int add_caller(unsigned long *n, unsigned long v) -{ - unsigned long *p; - int l; - if (!v) - return 1; - l = n[1]; - p = n + 2; - while (l) { - int i = l/2; - unsigned long *q = p + 2 * i; - if (*q == v) { - q[1]++; - return 1; - } - if (*q > v) { - l = i; - } else { - p = q + 2; - l -= i + 1; - } - } - if (++n[1] == n[0]) - return 0; - memmove(p + 2, p, n[1] * 2 * sizeof(unsigned long) - ((void *)p - (void *)n)); - p[0] = v; - p[1] = 1; - return 1; -} - -static void handle_slab(unsigned long *n, struct kmem_cache *c, - struct page *page) -{ - void *p; - int i, j; - unsigned long v; - - if (n[0] == n[1]) - return; - for (i = 0, p = page->s_mem; i < c->num; i++, p += c->size) { - bool active = true; - - for (j = page->active; j < c->num; j++) { - if (get_free_obj(page, j) == i) { - active = false; - break; - } - } - - if (!active) - continue; - - /* - * probe_kernel_read() is used for DEBUG_PAGEALLOC. page table - * mapping is established when actual object allocation and - * we could mistakenly access the unmapped object in the cpu - * cache. - */ - if (probe_kernel_read(&v, dbg_userword(c, p), sizeof(v))) - continue; - - if (!add_caller(n, v)) - return; - } -} - -static void show_symbol(struct seq_file *m, unsigned long address) -{ -#ifdef CONFIG_KALLSYMS - unsigned long offset, size; - char modname[MODULE_NAME_LEN], name[KSYM_NAME_LEN]; - - if (lookup_symbol_attrs(address, &size, &offset, modname, name) == 0) { - seq_printf(m, "%s+%#lx/%#lx", name, offset, size); - if (modname[0]) - seq_printf(m, " [%s]", modname); - return; - } -#endif - seq_printf(m, "%px", (void *)address); -} - -static int leaks_show(struct seq_file *m, void *p) -{ - struct kmem_cache *cachep = list_entry(p, struct kmem_cache, - root_caches_node); - struct page *page; - struct kmem_cache_node *n; - const char *name; - unsigned long *x = m->private; - int node; - int i; - - if (!(cachep->flags & SLAB_STORE_USER)) - return 0; - if (!(cachep->flags & SLAB_RED_ZONE)) - return 0; - - /* - * Set store_user_clean and start to grab stored user information - * for all objects on this cache. If some alloc/free requests comes - * during the processing, information would be wrong so restart - * whole processing. - */ - do { - drain_cpu_caches(cachep); - /* - * drain_cpu_caches() could make kmemleak_object and - * debug_objects_cache dirty, so reset afterwards. - */ - set_store_user_clean(cachep); - - x[1] = 0; - - for_each_kmem_cache_node(cachep, node, n) { - - check_irq_on(); - spin_lock_irq(&n->list_lock); - - list_for_each_entry(page, &n->slabs_full, slab_list) - handle_slab(x, cachep, page); - list_for_each_entry(page, &n->slabs_partial, slab_list) - handle_slab(x, cachep, page); - spin_unlock_irq(&n->list_lock); - } - } while (!is_store_user_clean(cachep)); - - name = cachep->name; - if (x[0] == x[1]) { - /* Increase the buffer size */ - mutex_unlock(&slab_mutex); - m->private = kcalloc(x[0] * 4, sizeof(unsigned long), - GFP_KERNEL); - if (!m->private) { - /* Too bad, we are really out */ - m->private = x; - mutex_lock(&slab_mutex); - return -ENOMEM; - } - *(unsigned long *)m->private = x[0] * 2; - kfree(x); - mutex_lock(&slab_mutex); - /* Now make sure this entry will be retried */ - m->count = m->size; - return 0; - } - for (i = 0; i < x[1]; i++) { - seq_printf(m, "%s: %lu ", name, x[2*i+3]); - show_symbol(m, x[2*i+2]); - seq_putc(m, '\n'); - } - - return 0; -} - -static const struct seq_operations slabstats_op = { - .start = slab_start, - .next = slab_next, - .stop = slab_stop, - .show = leaks_show, -}; - -static int slabstats_open(struct inode *inode, struct file *file) -{ - unsigned long *n; - - n = __seq_open_private(file, &slabstats_op, PAGE_SIZE); - if (!n) - return -ENOMEM; - - *n = PAGE_SIZE / (2 * sizeof(unsigned long)); - - return 0; -} - -static const struct file_operations proc_slabstats_operations = { - .open = slabstats_open, - .read = seq_read, - .llseek = seq_lseek, - .release = seq_release_private, -}; -#endif - -static int __init slab_proc_init(void) -{ -#ifdef CONFIG_DEBUG_SLAB_LEAK - proc_create("slab_allocators", 0, NULL, &proc_slabstats_operations); -#endif - return 0; -} -module_init(slab_proc_init); - #ifdef CONFIG_HARDENED_USERCOPY /* * Rejects incorrectly sized objects and objects that are to be copied diff --git a/net/ceph/cls_lock_client.c b/net/ceph/cls_lock_client.c index 2105a6eaa66c..4cc28541281b 100644 --- a/net/ceph/cls_lock_client.c +++ b/net/ceph/cls_lock_client.c @@ -271,7 +271,7 @@ static int decode_locker(void **p, void *end, struct ceph_locker *locker) dout("%s %s%llu cookie %s addr %s\n", __func__, ENTITY_NAME(locker->id.name), locker->id.cookie, - ceph_pr_addr(&locker->info.addr.in_addr)); + ceph_pr_addr(&locker->info.addr)); return 0; } diff --git a/net/ceph/debugfs.c b/net/ceph/debugfs.c index 46f65709a6ff..63aef9915f75 100644 --- a/net/ceph/debugfs.c +++ b/net/ceph/debugfs.c @@ -46,7 +46,7 @@ static int monmap_show(struct seq_file *s, void *p) seq_printf(s, "\t%s%lld\t%s\n", ENTITY_NAME(inst->name), - ceph_pr_addr(&inst->addr.in_addr)); + ceph_pr_addr(&inst->addr)); } return 0; } @@ -82,7 +82,7 @@ static int osdmap_show(struct seq_file *s, void *p) char sb[64]; seq_printf(s, "osd%d\t%s\t%3d%%\t(%s)\t%3d%%\n", - i, ceph_pr_addr(&addr->in_addr), + i, ceph_pr_addr(addr), ((map->osd_weight[i]*100) >> 16), ceph_osdmap_state_str(sb, sizeof(sb), state), ((ceph_get_primary_affinity(map, i)*100) >> 16)); diff --git a/net/ceph/messenger.c b/net/ceph/messenger.c index 3083988ce729..cd0b094468b6 100644 --- a/net/ceph/messenger.c +++ b/net/ceph/messenger.c @@ -186,17 +186,18 @@ static atomic_t addr_str_seq = ATOMIC_INIT(0); static struct page *zero_page; /* used in certain error cases */ -const char *ceph_pr_addr(const struct sockaddr_storage *ss) +const char *ceph_pr_addr(const struct ceph_entity_addr *addr) { int i; char *s; - struct sockaddr_in *in4 = (struct sockaddr_in *) ss; - struct sockaddr_in6 *in6 = (struct sockaddr_in6 *) ss; + struct sockaddr_storage ss = addr->in_addr; /* align */ + struct sockaddr_in *in4 = (struct sockaddr_in *)&ss; + struct sockaddr_in6 *in6 = (struct sockaddr_in6 *)&ss; i = atomic_inc_return(&addr_str_seq) & ADDR_STR_COUNT_MASK; s = addr_str[i]; - switch (ss->ss_family) { + switch (ss.ss_family) { case AF_INET: snprintf(s, MAX_ADDR_STR_LEN, "%pI4:%hu", &in4->sin_addr, ntohs(in4->sin_port)); @@ -209,7 +210,7 @@ const char *ceph_pr_addr(const struct sockaddr_storage *ss) default: snprintf(s, MAX_ADDR_STR_LEN, "(unknown sockaddr family %hu)", - ss->ss_family); + ss.ss_family); } return s; @@ -449,7 +450,7 @@ static void set_sock_callbacks(struct socket *sock, */ static int ceph_tcp_connect(struct ceph_connection *con) { - struct sockaddr_storage *paddr = &con->peer_addr.in_addr; + struct sockaddr_storage ss = con->peer_addr.in_addr; /* align */ struct socket *sock; unsigned int noio_flag; int ret; @@ -458,7 +459,7 @@ static int ceph_tcp_connect(struct ceph_connection *con) /* sock_create_kern() allocates with GFP_KERNEL */ noio_flag = memalloc_noio_save(); - ret = sock_create_kern(read_pnet(&con->msgr->net), paddr->ss_family, + ret = sock_create_kern(read_pnet(&con->msgr->net), ss.ss_family, SOCK_STREAM, IPPROTO_TCP, &sock); memalloc_noio_restore(noio_flag); if (ret) @@ -471,18 +472,18 @@ static int ceph_tcp_connect(struct ceph_connection *con) set_sock_callbacks(sock, con); - dout("connect %s\n", ceph_pr_addr(&con->peer_addr.in_addr)); + dout("connect %s\n", ceph_pr_addr(&con->peer_addr)); con_sock_state_connecting(con); - ret = sock->ops->connect(sock, (struct sockaddr *)paddr, sizeof(*paddr), + ret = sock->ops->connect(sock, (struct sockaddr *)&ss, sizeof(ss), O_NONBLOCK); if (ret == -EINPROGRESS) { dout("connect %s EINPROGRESS sk_state = %u\n", - ceph_pr_addr(&con->peer_addr.in_addr), + ceph_pr_addr(&con->peer_addr), sock->sk->sk_state); } else if (ret < 0) { pr_err("connect %s error %d\n", - ceph_pr_addr(&con->peer_addr.in_addr), ret); + ceph_pr_addr(&con->peer_addr), ret); sock_release(sock); return ret; } @@ -669,8 +670,7 @@ static void reset_connection(struct ceph_connection *con) void ceph_con_close(struct ceph_connection *con) { mutex_lock(&con->mutex); - dout("con_close %p peer %s\n", con, - ceph_pr_addr(&con->peer_addr.in_addr)); + dout("con_close %p peer %s\n", con, ceph_pr_addr(&con->peer_addr)); con->state = CON_STATE_CLOSED; con_flag_clear(con, CON_FLAG_LOSSYTX); /* so we retry next connect */ @@ -694,7 +694,7 @@ void ceph_con_open(struct ceph_connection *con, struct ceph_entity_addr *addr) { mutex_lock(&con->mutex); - dout("con_open %p %s\n", con, ceph_pr_addr(&addr->in_addr)); + dout("con_open %p %s\n", con, ceph_pr_addr(addr)); WARN_ON(con->state != CON_STATE_CLOSED); con->state = CON_STATE_PREOPEN; @@ -1788,21 +1788,22 @@ static int verify_hello(struct ceph_connection *con) { if (memcmp(con->in_banner, CEPH_BANNER, strlen(CEPH_BANNER))) { pr_err("connect to %s got bad banner\n", - ceph_pr_addr(&con->peer_addr.in_addr)); + ceph_pr_addr(&con->peer_addr)); con->error_msg = "protocol error, bad banner"; return -1; } return 0; } -static bool addr_is_blank(struct sockaddr_storage *ss) +static bool addr_is_blank(struct ceph_entity_addr *addr) { - struct in_addr *addr = &((struct sockaddr_in *)ss)->sin_addr; - struct in6_addr *addr6 = &((struct sockaddr_in6 *)ss)->sin6_addr; + struct sockaddr_storage ss = addr->in_addr; /* align */ + struct in_addr *addr4 = &((struct sockaddr_in *)&ss)->sin_addr; + struct in6_addr *addr6 = &((struct sockaddr_in6 *)&ss)->sin6_addr; - switch (ss->ss_family) { + switch (ss.ss_family) { case AF_INET: - return addr->s_addr == htonl(INADDR_ANY); + return addr4->s_addr == htonl(INADDR_ANY); case AF_INET6: return ipv6_addr_any(addr6); default: @@ -1810,25 +1811,25 @@ static bool addr_is_blank(struct sockaddr_storage *ss) } } -static int addr_port(struct sockaddr_storage *ss) +static int addr_port(struct ceph_entity_addr *addr) { - switch (ss->ss_family) { + switch (get_unaligned(&addr->in_addr.ss_family)) { case AF_INET: - return ntohs(((struct sockaddr_in *)ss)->sin_port); + return ntohs(get_unaligned(&((struct sockaddr_in *)&addr->in_addr)->sin_port)); case AF_INET6: - return ntohs(((struct sockaddr_in6 *)ss)->sin6_port); + return ntohs(get_unaligned(&((struct sockaddr_in6 *)&addr->in_addr)->sin6_port)); } return 0; } -static void addr_set_port(struct sockaddr_storage *ss, int p) +static void addr_set_port(struct ceph_entity_addr *addr, int p) { - switch (ss->ss_family) { + switch (get_unaligned(&addr->in_addr.ss_family)) { case AF_INET: - ((struct sockaddr_in *)ss)->sin_port = htons(p); + put_unaligned(htons(p), &((struct sockaddr_in *)&addr->in_addr)->sin_port); break; case AF_INET6: - ((struct sockaddr_in6 *)ss)->sin6_port = htons(p); + put_unaligned(htons(p), &((struct sockaddr_in6 *)&addr->in_addr)->sin6_port); break; } } @@ -1836,21 +1837,18 @@ static void addr_set_port(struct sockaddr_storage *ss, int p) /* * Unlike other *_pton function semantics, zero indicates success. */ -static int ceph_pton(const char *str, size_t len, struct sockaddr_storage *ss, +static int ceph_pton(const char *str, size_t len, struct ceph_entity_addr *addr, char delim, const char **ipend) { - struct sockaddr_in *in4 = (struct sockaddr_in *) ss; - struct sockaddr_in6 *in6 = (struct sockaddr_in6 *) ss; + memset(&addr->in_addr, 0, sizeof(addr->in_addr)); - memset(ss, 0, sizeof(*ss)); - - if (in4_pton(str, len, (u8 *)&in4->sin_addr.s_addr, delim, ipend)) { - ss->ss_family = AF_INET; + if (in4_pton(str, len, (u8 *)&((struct sockaddr_in *)&addr->in_addr)->sin_addr.s_addr, delim, ipend)) { + put_unaligned(AF_INET, &addr->in_addr.ss_family); return 0; } - if (in6_pton(str, len, (u8 *)&in6->sin6_addr.s6_addr, delim, ipend)) { - ss->ss_family = AF_INET6; + if (in6_pton(str, len, (u8 *)&((struct sockaddr_in6 *)&addr->in_addr)->sin6_addr.s6_addr, delim, ipend)) { + put_unaligned(AF_INET6, &addr->in_addr.ss_family); return 0; } @@ -1862,7 +1860,7 @@ static int ceph_pton(const char *str, size_t len, struct sockaddr_storage *ss, */ #ifdef CONFIG_CEPH_LIB_USE_DNS_RESOLVER static int ceph_dns_resolve_name(const char *name, size_t namelen, - struct sockaddr_storage *ss, char delim, const char **ipend) + struct ceph_entity_addr *addr, char delim, const char **ipend) { const char *end, *delim_p; char *colon_p, *ip_addr = NULL; @@ -1889,9 +1887,9 @@ static int ceph_dns_resolve_name(const char *name, size_t namelen, return -EINVAL; /* do dns_resolve upcall */ - ip_len = dns_query(NULL, name, end - name, NULL, &ip_addr, NULL); + ip_len = dns_query(NULL, name, end - name, NULL, &ip_addr, NULL, false); if (ip_len > 0) - ret = ceph_pton(ip_addr, ip_len, ss, -1, NULL); + ret = ceph_pton(ip_addr, ip_len, addr, -1, NULL); else ret = -ESRCH; @@ -1900,13 +1898,13 @@ static int ceph_dns_resolve_name(const char *name, size_t namelen, *ipend = end; pr_info("resolve '%.*s' (ret=%d): %s\n", (int)(end - name), name, - ret, ret ? "failed" : ceph_pr_addr(ss)); + ret, ret ? "failed" : ceph_pr_addr(addr)); return ret; } #else static inline int ceph_dns_resolve_name(const char *name, size_t namelen, - struct sockaddr_storage *ss, char delim, const char **ipend) + struct ceph_entity_addr *addr, char delim, const char **ipend) { return -EINVAL; } @@ -1917,13 +1915,13 @@ static inline int ceph_dns_resolve_name(const char *name, size_t namelen, * then try to extract a hostname to resolve using userspace DNS upcall. */ static int ceph_parse_server_name(const char *name, size_t namelen, - struct sockaddr_storage *ss, char delim, const char **ipend) + struct ceph_entity_addr *addr, char delim, const char **ipend) { int ret; - ret = ceph_pton(name, namelen, ss, delim, ipend); + ret = ceph_pton(name, namelen, addr, delim, ipend); if (ret) - ret = ceph_dns_resolve_name(name, namelen, ss, delim, ipend); + ret = ceph_dns_resolve_name(name, namelen, addr, delim, ipend); return ret; } @@ -1942,7 +1940,6 @@ int ceph_parse_ips(const char *c, const char *end, dout("parse_ips on '%.*s'\n", (int)(end-c), c); for (i = 0; i < max_count; i++) { const char *ipend; - struct sockaddr_storage *ss = &addr[i].in_addr; int port; char delim = ','; @@ -1951,7 +1948,7 @@ int ceph_parse_ips(const char *c, const char *end, p++; } - ret = ceph_parse_server_name(p, end - p, ss, delim, &ipend); + ret = ceph_parse_server_name(p, end - p, &addr[i], delim, &ipend); if (ret) goto bad; ret = -EINVAL; @@ -1982,9 +1979,9 @@ int ceph_parse_ips(const char *c, const char *end, port = CEPH_MON_PORT; } - addr_set_port(ss, port); + addr_set_port(&addr[i], port); - dout("parse_ips got %s\n", ceph_pr_addr(ss)); + dout("parse_ips got %s\n", ceph_pr_addr(&addr[i])); if (p == end) break; @@ -2023,12 +2020,12 @@ static int process_banner(struct ceph_connection *con) */ if (memcmp(&con->peer_addr, &con->actual_peer_addr, sizeof(con->peer_addr)) != 0 && - !(addr_is_blank(&con->actual_peer_addr.in_addr) && + !(addr_is_blank(&con->actual_peer_addr) && con->actual_peer_addr.nonce == con->peer_addr.nonce)) { pr_warn("wrong peer, want %s/%d, got %s/%d\n", - ceph_pr_addr(&con->peer_addr.in_addr), + ceph_pr_addr(&con->peer_addr), (int)le32_to_cpu(con->peer_addr.nonce), - ceph_pr_addr(&con->actual_peer_addr.in_addr), + ceph_pr_addr(&con->actual_peer_addr), (int)le32_to_cpu(con->actual_peer_addr.nonce)); con->error_msg = "wrong peer at address"; return -1; @@ -2037,16 +2034,16 @@ static int process_banner(struct ceph_connection *con) /* * did we learn our address? */ - if (addr_is_blank(&con->msgr->inst.addr.in_addr)) { - int port = addr_port(&con->msgr->inst.addr.in_addr); + if (addr_is_blank(&con->msgr->inst.addr)) { + int port = addr_port(&con->msgr->inst.addr); memcpy(&con->msgr->inst.addr.in_addr, &con->peer_addr_for_me.in_addr, sizeof(con->peer_addr_for_me.in_addr)); - addr_set_port(&con->msgr->inst.addr.in_addr, port); + addr_set_port(&con->msgr->inst.addr, port); encode_my_addr(con->msgr); dout("process_banner learned my addr is %s\n", - ceph_pr_addr(&con->msgr->inst.addr.in_addr)); + ceph_pr_addr(&con->msgr->inst.addr)); } return 0; @@ -2097,7 +2094,7 @@ static int process_connect(struct ceph_connection *con) pr_err("%s%lld %s feature set mismatch," " my %llx < server's %llx, missing %llx\n", ENTITY_NAME(con->peer_name), - ceph_pr_addr(&con->peer_addr.in_addr), + ceph_pr_addr(&con->peer_addr), sup_feat, server_feat, server_feat & ~sup_feat); con->error_msg = "missing required protocol features"; reset_connection(con); @@ -2107,7 +2104,7 @@ static int process_connect(struct ceph_connection *con) pr_err("%s%lld %s protocol version mismatch," " my %d != server's %d\n", ENTITY_NAME(con->peer_name), - ceph_pr_addr(&con->peer_addr.in_addr), + ceph_pr_addr(&con->peer_addr), le32_to_cpu(con->out_connect.protocol_version), le32_to_cpu(con->in_reply.protocol_version)); con->error_msg = "protocol version mismatch"; @@ -2141,7 +2138,7 @@ static int process_connect(struct ceph_connection *con) le32_to_cpu(con->in_reply.connect_seq)); pr_err("%s%lld %s connection reset\n", ENTITY_NAME(con->peer_name), - ceph_pr_addr(&con->peer_addr.in_addr)); + ceph_pr_addr(&con->peer_addr)); reset_connection(con); con_out_kvec_reset(con); ret = prepare_write_connect(con); @@ -2198,7 +2195,7 @@ static int process_connect(struct ceph_connection *con) pr_err("%s%lld %s protocol feature mismatch," " my required %llx > server's %llx, need %llx\n", ENTITY_NAME(con->peer_name), - ceph_pr_addr(&con->peer_addr.in_addr), + ceph_pr_addr(&con->peer_addr), req_feat, server_feat, req_feat & ~server_feat); con->error_msg = "missing required protocol features"; reset_connection(con); @@ -2405,7 +2402,7 @@ static int read_partial_message(struct ceph_connection *con) if ((s64)seq - (s64)con->in_seq < 1) { pr_info("skipping %s%lld %s seq %lld expected %lld\n", ENTITY_NAME(con->peer_name), - ceph_pr_addr(&con->peer_addr.in_addr), + ceph_pr_addr(&con->peer_addr), seq, con->in_seq + 1); con->in_base_pos = -front_len - middle_len - data_len - sizeof_footer(con); @@ -2984,10 +2981,10 @@ static void ceph_con_workfn(struct work_struct *work) static void con_fault(struct ceph_connection *con) { dout("fault %p state %lu to peer %s\n", - con, con->state, ceph_pr_addr(&con->peer_addr.in_addr)); + con, con->state, ceph_pr_addr(&con->peer_addr)); pr_warn("%s%lld %s %s\n", ENTITY_NAME(con->peer_name), - ceph_pr_addr(&con->peer_addr.in_addr), con->error_msg); + ceph_pr_addr(&con->peer_addr), con->error_msg); con->error_msg = NULL; WARN_ON(con->state != CON_STATE_CONNECTING && diff --git a/net/ceph/mon_client.c b/net/ceph/mon_client.c index a53e4fbb6319..895679d3529b 100644 --- a/net/ceph/mon_client.c +++ b/net/ceph/mon_client.c @@ -76,7 +76,7 @@ struct ceph_monmap *ceph_monmap_decode(void *p, void *end) m->num_mon); for (i = 0; i < m->num_mon; i++) dout("monmap_decode mon%d is %s\n", i, - ceph_pr_addr(&m->mon_inst[i].addr.in_addr)); + ceph_pr_addr(&m->mon_inst[i].addr)); return m; bad: @@ -203,7 +203,7 @@ static void reopen_session(struct ceph_mon_client *monc) { if (!monc->hunting) pr_info("mon%d %s session lost, hunting for new mon\n", - monc->cur_mon, ceph_pr_addr(&monc->con.peer_addr.in_addr)); + monc->cur_mon, ceph_pr_addr(&monc->con.peer_addr)); __close_session(monc); __open_session(monc); @@ -1178,7 +1178,7 @@ static void handle_auth_reply(struct ceph_mon_client *monc, __resend_generic_request(monc); pr_info("mon%d %s session established\n", monc->cur_mon, - ceph_pr_addr(&monc->con.peer_addr.in_addr)); + ceph_pr_addr(&monc->con.peer_addr)); } out: diff --git a/net/ceph/osd_client.c b/net/ceph/osd_client.c index 6f739de28918..9a8eca5eda65 100644 --- a/net/ceph/osd_client.c +++ b/net/ceph/osd_client.c @@ -4926,7 +4926,7 @@ static int decode_watcher(void **p, void *end, struct ceph_watch_item *item) dout("%s %s%llu cookie %llu addr %s\n", __func__, ENTITY_NAME(item->name), item->cookie, - ceph_pr_addr(&item->addr.in_addr)); + ceph_pr_addr(&item->addr)); return 0; } diff --git a/net/dns_resolver/dns_query.c b/net/dns_resolver/dns_query.c index 19aa32fc1802..2d260432b3be 100644 --- a/net/dns_resolver/dns_query.c +++ b/net/dns_resolver/dns_query.c @@ -54,6 +54,7 @@ * @options: Request options (or NULL if no options) * @_result: Where to place the returned data (or NULL) * @_expiry: Where to store the result expiry time (or NULL) + * @invalidate: Always invalidate the key after use * * The data will be returned in the pointer at *result, if provided, and the * caller is responsible for freeing it. @@ -69,7 +70,8 @@ * Returns the size of the result on success, -ve error code otherwise. */ int dns_query(const char *type, const char *name, size_t namelen, - const char *options, char **_result, time64_t *_expiry) + const char *options, char **_result, time64_t *_expiry, + bool invalidate) { struct key *rkey; struct user_key_payload *upayload; @@ -157,6 +159,8 @@ int dns_query(const char *type, const char *name, size_t namelen, ret = len; put: up_read(&rkey->sem); + if (invalidate) + key_invalidate(rkey); key_put(rkey); out: kleave(" = %d", ret); diff --git a/net/rxrpc/af_rxrpc.c b/net/rxrpc/af_rxrpc.c index ae8c5d7f3bf1..ffde5b187f5d 100644 --- a/net/rxrpc/af_rxrpc.c +++ b/net/rxrpc/af_rxrpc.c @@ -270,6 +270,7 @@ static int rxrpc_listen(struct socket *sock, int backlog) * @gfp: The allocation constraints * @notify_rx: Where to send notifications instead of socket queue * @upgrade: Request service upgrade for call + * @intr: The call is interruptible * @debug_id: The debug ID for tracing to be assigned to the call * * Allow a kernel service to begin a call on the nominated socket. This just @@ -287,6 +288,7 @@ struct rxrpc_call *rxrpc_kernel_begin_call(struct socket *sock, gfp_t gfp, rxrpc_notify_rx_t notify_rx, bool upgrade, + bool intr, unsigned int debug_id) { struct rxrpc_conn_parameters cp; @@ -311,6 +313,7 @@ struct rxrpc_call *rxrpc_kernel_begin_call(struct socket *sock, memset(&p, 0, sizeof(p)); p.user_call_ID = user_call_ID; p.tx_total_len = tx_total_len; + p.intr = intr; memset(&cp, 0, sizeof(cp)); cp.local = rx->local; @@ -443,6 +446,31 @@ void rxrpc_kernel_new_call_notification( } EXPORT_SYMBOL(rxrpc_kernel_new_call_notification); +/** + * rxrpc_kernel_set_max_life - Set maximum lifespan on a call + * @sock: The socket the call is on + * @call: The call to configure + * @hard_timeout: The maximum lifespan of the call in jiffies + * + * Set the maximum lifespan of a call. The call will end with ETIME or + * ETIMEDOUT if it takes longer than this. + */ +void rxrpc_kernel_set_max_life(struct socket *sock, struct rxrpc_call *call, + unsigned long hard_timeout) +{ + unsigned long now; + + mutex_lock(&call->user_mutex); + + now = jiffies; + hard_timeout += now; + WRITE_ONCE(call->expect_term_by, hard_timeout); + rxrpc_reduce_call_timer(call, hard_timeout, now, rxrpc_timer_set_for_hard); + + mutex_unlock(&call->user_mutex); +} +EXPORT_SYMBOL(rxrpc_kernel_set_max_life); + /* * connect an RxRPC socket * - this just targets it at a specific destination; no actual connection diff --git a/net/rxrpc/ar-internal.h b/net/rxrpc/ar-internal.h index 062ca9dc29b8..07fc1dfa4878 100644 --- a/net/rxrpc/ar-internal.h +++ b/net/rxrpc/ar-internal.h @@ -482,6 +482,7 @@ enum rxrpc_call_flag { RXRPC_CALL_BEGAN_RX_TIMER, /* We began the expect_rx_by timer */ RXRPC_CALL_RX_HEARD, /* The peer responded at least once to this call */ RXRPC_CALL_RX_UNDERRUN, /* Got data underrun */ + RXRPC_CALL_IS_INTR, /* The call is interruptible */ }; /* @@ -711,6 +712,7 @@ struct rxrpc_call_params { u32 normal; /* Max time since last call packet (msec) */ } timeouts; u8 nr_timeouts; /* Number of timeouts specified */ + bool intr; /* The call is interruptible */ }; struct rxrpc_send_params { diff --git a/net/rxrpc/call_object.c b/net/rxrpc/call_object.c index fe96881a334d..d0ca98d7aef5 100644 --- a/net/rxrpc/call_object.c +++ b/net/rxrpc/call_object.c @@ -241,6 +241,8 @@ struct rxrpc_call *rxrpc_new_client_call(struct rxrpc_sock *rx, return call; } + if (p->intr) + __set_bit(RXRPC_CALL_IS_INTR, &call->flags); call->tx_total_len = p->tx_total_len; trace_rxrpc_call(call, rxrpc_call_new_client, atomic_read(&call->usage), here, (const void *)p->user_call_ID); diff --git a/net/rxrpc/conn_client.c b/net/rxrpc/conn_client.c index 83797b3949e2..5cf5595a14d8 100644 --- a/net/rxrpc/conn_client.c +++ b/net/rxrpc/conn_client.c @@ -656,10 +656,14 @@ static int rxrpc_wait_for_channel(struct rxrpc_call *call, gfp_t gfp) add_wait_queue_exclusive(&call->waitq, &myself); for (;;) { - set_current_state(TASK_INTERRUPTIBLE); + if (test_bit(RXRPC_CALL_IS_INTR, &call->flags)) + set_current_state(TASK_INTERRUPTIBLE); + else + set_current_state(TASK_UNINTERRUPTIBLE); if (call->call_id) break; - if (signal_pending(current)) { + if (test_bit(RXRPC_CALL_IS_INTR, &call->flags) && + signal_pending(current)) { ret = -ERESTARTSYS; break; } diff --git a/net/rxrpc/sendmsg.c b/net/rxrpc/sendmsg.c index bec64deb7b0a..45a05d9a27fa 100644 --- a/net/rxrpc/sendmsg.c +++ b/net/rxrpc/sendmsg.c @@ -80,7 +80,8 @@ static int rxrpc_wait_for_tx_window_nonintr(struct rxrpc_sock *rx, if (call->state >= RXRPC_CALL_COMPLETE) return call->error; - if (timeout == 0 && + if (test_bit(RXRPC_CALL_IS_INTR, &call->flags) && + timeout == 0 && tx_win == tx_start && signal_pending(current)) return -EINTR; @@ -620,6 +621,7 @@ int rxrpc_do_sendmsg(struct rxrpc_sock *rx, struct msghdr *msg, size_t len) .call.tx_total_len = -1, .call.user_call_ID = 0, .call.nr_timeouts = 0, + .call.intr = true, .abort_code = 0, .command = RXRPC_CMD_SEND_DATA, .exclusive = false, |