diff options
Diffstat (limited to 'fs')
130 files changed, 3024 insertions, 3979 deletions
diff --git a/fs/Kconfig b/fs/Kconfig index 83eab52fb3f6..b0e42b6a96b9 100644 --- a/fs/Kconfig +++ b/fs/Kconfig @@ -39,6 +39,7 @@ config FS_DAX depends on MMU depends on !(ARM || MIPS || SPARC) select FS_IOMAP + select DAX help Direct Access (DAX) can be used on memory-backed block devices. If the block device supports DAX and the filesystem supports DAX, diff --git a/fs/bfs/inode.c b/fs/bfs/inode.c index f2deec0a62f0..25e312cb6071 100644 --- a/fs/bfs/inode.c +++ b/fs/bfs/inode.c @@ -1,7 +1,7 @@ /* * fs/bfs/inode.c * BFS superblock and inode operations. - * Copyright (C) 1999-2006 Tigran Aivazian <tigran@aivazian.fsnet.co.uk> + * Copyright (C) 1999-2006 Tigran Aivazian <aivazian.tigran@gmail.com> * From fs/minix, Copyright (C) 1991, 1992 Linus Torvalds. * * Made endianness-clean by Andrew Stribblehill <ads@wompom.org>, 2005. @@ -19,7 +19,7 @@ #include <linux/uaccess.h> #include "bfs.h" -MODULE_AUTHOR("Tigran Aivazian <tigran@aivazian.fsnet.co.uk>"); +MODULE_AUTHOR("Tigran Aivazian <aivazian.tigran@gmail.com>"); MODULE_DESCRIPTION("SCO UnixWare BFS filesystem for Linux"); MODULE_LICENSE("GPL"); diff --git a/fs/block_dev.c b/fs/block_dev.c index 2a305c1a2d88..519599dddd36 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -717,72 +717,6 @@ int bdev_write_page(struct block_device *bdev, sector_t sector, } EXPORT_SYMBOL_GPL(bdev_write_page); -int bdev_dax_pgoff(struct block_device *bdev, sector_t sector, size_t size, - pgoff_t *pgoff) -{ - phys_addr_t phys_off = (get_start_sect(bdev) + sector) * 512; - - if (pgoff) - *pgoff = PHYS_PFN(phys_off); - if (phys_off % PAGE_SIZE || size % PAGE_SIZE) - return -EINVAL; - return 0; -} -EXPORT_SYMBOL(bdev_dax_pgoff); - -/** - * bdev_dax_supported() - Check if the device supports dax for filesystem - * @sb: The superblock of the device - * @blocksize: The block size of the device - * - * This is a library function for filesystems to check if the block device - * can be mounted with dax option. - * - * Return: negative errno if unsupported, 0 if supported. - */ -int bdev_dax_supported(struct super_block *sb, int blocksize) -{ - struct block_device *bdev = sb->s_bdev; - struct dax_device *dax_dev; - pgoff_t pgoff; - int err, id; - void *kaddr; - pfn_t pfn; - long len; - - if (blocksize != PAGE_SIZE) { - vfs_msg(sb, KERN_ERR, "error: unsupported blocksize for dax"); - return -EINVAL; - } - - err = bdev_dax_pgoff(bdev, 0, PAGE_SIZE, &pgoff); - if (err) { - vfs_msg(sb, KERN_ERR, "error: unaligned partition for dax"); - return err; - } - - dax_dev = dax_get_by_host(bdev->bd_disk->disk_name); - if (!dax_dev) { - vfs_msg(sb, KERN_ERR, "error: device does not support dax"); - return -EOPNOTSUPP; - } - - id = dax_read_lock(); - len = dax_direct_access(dax_dev, pgoff, 1, &kaddr, &pfn); - dax_read_unlock(id); - - put_dax(dax_dev); - - if (len < 1) { - vfs_msg(sb, KERN_ERR, - "error: dax access failed (%ld)", len); - return len < 0 ? len : -EIO; - } - - return 0; -} -EXPORT_SYMBOL_GPL(bdev_dax_supported); - /* * pseudo-fs */ diff --git a/fs/btrfs/backref.c b/fs/btrfs/backref.c index 7699e16784d3..24865da63d8f 100644 --- a/fs/btrfs/backref.c +++ b/fs/btrfs/backref.c @@ -26,6 +26,11 @@ #include "delayed-ref.h" #include "locking.h" +enum merge_mode { + MERGE_IDENTICAL_KEYS = 1, + MERGE_IDENTICAL_PARENTS, +}; + /* Just an arbitrary number so we can be sure this happened */ #define BACKREF_FOUND_SHARED 6 @@ -533,7 +538,7 @@ static int add_all_parents(struct btrfs_root *root, struct btrfs_path *path, * slot==nritems. In that case, go to the next leaf before we continue. */ if (path->slots[0] >= btrfs_header_nritems(path->nodes[0])) { - if (time_seq == (u64)-1) + if (time_seq == SEQ_LAST) ret = btrfs_next_leaf(root, path); else ret = btrfs_next_old_leaf(root, path, time_seq); @@ -577,7 +582,7 @@ static int add_all_parents(struct btrfs_root *root, struct btrfs_path *path, eie = NULL; } next: - if (time_seq == (u64)-1) + if (time_seq == SEQ_LAST) ret = btrfs_next_item(root, path); else ret = btrfs_next_old_item(root, path, time_seq); @@ -629,7 +634,7 @@ static int __resolve_indirect_ref(struct btrfs_fs_info *fs_info, if (path->search_commit_root) root_level = btrfs_header_level(root->commit_root); - else if (time_seq == (u64)-1) + else if (time_seq == SEQ_LAST) root_level = btrfs_header_level(root->node); else root_level = btrfs_old_root_level(root, time_seq); @@ -640,7 +645,7 @@ static int __resolve_indirect_ref(struct btrfs_fs_info *fs_info, } path->lowest_level = level; - if (time_seq == (u64)-1) + if (time_seq == SEQ_LAST) ret = btrfs_search_slot(NULL, root, &ref->key_for_search, path, 0, 0); else @@ -809,14 +814,12 @@ static int __add_missing_keys(struct btrfs_fs_info *fs_info, /* * merge backrefs and adjust counts accordingly * - * mode = 1: merge identical keys, if key is set - * FIXME: if we add more keys in __add_prelim_ref, we can merge more here. - * additionally, we could even add a key range for the blocks we - * looked into to merge even more (-> replace unresolved refs by those - * having a parent). - * mode = 2: merge identical parents + * FIXME: For MERGE_IDENTICAL_KEYS, if we add more keys in __add_prelim_ref + * then we can merge more here. Additionally, we could even add a key + * range for the blocks we looked into to merge even more (-> replace + * unresolved refs by those having a parent). */ -static void __merge_refs(struct list_head *head, int mode) +static void __merge_refs(struct list_head *head, enum merge_mode mode) { struct __prelim_ref *pos1; @@ -829,7 +832,7 @@ static void __merge_refs(struct list_head *head, int mode) if (!ref_for_same_block(ref1, ref2)) continue; - if (mode == 1) { + if (mode == MERGE_IDENTICAL_KEYS) { if (!ref1->parent && ref2->parent) swap(ref1, ref2); } else { @@ -1196,7 +1199,7 @@ static int __add_keyed_refs(struct btrfs_fs_info *fs_info, * * NOTE: This can return values > 0 * - * If time_seq is set to (u64)-1, it will not search delayed_refs, and behave + * If time_seq is set to SEQ_LAST, it will not search delayed_refs, and behave * much like trans == NULL case, the difference only lies in it will not * commit root. * The special case is for qgroup to search roots in commit_transaction(). @@ -1243,7 +1246,7 @@ static int find_parent_nodes(struct btrfs_trans_handle *trans, path->skip_locking = 1; } - if (time_seq == (u64)-1) + if (time_seq == SEQ_LAST) path->skip_locking = 1; /* @@ -1273,9 +1276,9 @@ again: #ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS if (trans && likely(trans->type != __TRANS_DUMMY) && - time_seq != (u64)-1) { + time_seq != SEQ_LAST) { #else - if (trans && time_seq != (u64)-1) { + if (trans && time_seq != SEQ_LAST) { #endif /* * look if there are updates for this ref queued and lock the @@ -1286,7 +1289,7 @@ again: head = btrfs_find_delayed_ref_head(delayed_refs, bytenr); if (head) { if (!mutex_trylock(&head->mutex)) { - atomic_inc(&head->node.refs); + refcount_inc(&head->node.refs); spin_unlock(&delayed_refs->lock); btrfs_release_path(path); @@ -1374,7 +1377,7 @@ again: if (ret) goto out; - __merge_refs(&prefs, 1); + __merge_refs(&prefs, MERGE_IDENTICAL_KEYS); ret = __resolve_indirect_refs(fs_info, path, time_seq, &prefs, extent_item_pos, total_refs, @@ -1382,7 +1385,7 @@ again: if (ret) goto out; - __merge_refs(&prefs, 2); + __merge_refs(&prefs, MERGE_IDENTICAL_PARENTS); while (!list_empty(&prefs)) { ref = list_first_entry(&prefs, struct __prelim_ref, list); diff --git a/fs/btrfs/btrfs_inode.h b/fs/btrfs/btrfs_inode.h index 0c6baaba0651..b8622e4d1744 100644 --- a/fs/btrfs/btrfs_inode.h +++ b/fs/btrfs/btrfs_inode.h @@ -125,6 +125,13 @@ struct btrfs_inode { u64 delalloc_bytes; /* + * Total number of bytes pending delalloc that fall within a file + * range that is either a hole or beyond EOF (and no prealloc extent + * exists in the range). This is always <= delalloc_bytes. + */ + u64 new_delalloc_bytes; + + /* * total number of bytes pending defrag, used by stat to check whether * it needs COW. */ diff --git a/fs/btrfs/compression.c b/fs/btrfs/compression.c index c7721a6aa3bb..10e6b282d09d 100644 --- a/fs/btrfs/compression.c +++ b/fs/btrfs/compression.c @@ -44,7 +44,7 @@ struct compressed_bio { /* number of bios pending for this compressed extent */ - atomic_t pending_bios; + refcount_t pending_bios; /* the pages with the compressed data on them */ struct page **compressed_pages; @@ -161,7 +161,7 @@ static void end_compressed_bio_read(struct bio *bio) /* if there are more bios still pending for this compressed * extent, just exit */ - if (!atomic_dec_and_test(&cb->pending_bios)) + if (!refcount_dec_and_test(&cb->pending_bios)) goto out; inode = cb->inode; @@ -274,7 +274,7 @@ static void end_compressed_bio_write(struct bio *bio) /* if there are more bios still pending for this compressed * extent, just exit */ - if (!atomic_dec_and_test(&cb->pending_bios)) + if (!refcount_dec_and_test(&cb->pending_bios)) goto out; /* ok, we're the last bio for this extent, step one is to @@ -342,7 +342,7 @@ int btrfs_submit_compressed_write(struct inode *inode, u64 start, cb = kmalloc(compressed_bio_size(fs_info, compressed_len), GFP_NOFS); if (!cb) return -ENOMEM; - atomic_set(&cb->pending_bios, 0); + refcount_set(&cb->pending_bios, 0); cb->errors = 0; cb->inode = inode; cb->start = start; @@ -363,7 +363,7 @@ int btrfs_submit_compressed_write(struct inode *inode, u64 start, bio_set_op_attrs(bio, REQ_OP_WRITE, 0); bio->bi_private = cb; bio->bi_end_io = end_compressed_bio_write; - atomic_inc(&cb->pending_bios); + refcount_set(&cb->pending_bios, 1); /* create and submit bios for the compressed pages */ bytes_left = compressed_len; @@ -388,7 +388,7 @@ int btrfs_submit_compressed_write(struct inode *inode, u64 start, * we inc the count. Otherwise, the cb might get * freed before we're done setting it up */ - atomic_inc(&cb->pending_bios); + refcount_inc(&cb->pending_bios); ret = btrfs_bio_wq_end_io(fs_info, bio, BTRFS_WQ_ENDIO_DATA); BUG_ON(ret); /* -ENOMEM */ @@ -607,7 +607,7 @@ int btrfs_submit_compressed_read(struct inode *inode, struct bio *bio, if (!cb) goto out; - atomic_set(&cb->pending_bios, 0); + refcount_set(&cb->pending_bios, 0); cb->errors = 0; cb->inode = inode; cb->mirror_num = mirror_num; @@ -656,7 +656,7 @@ int btrfs_submit_compressed_read(struct inode *inode, struct bio *bio, bio_set_op_attrs (comp_bio, REQ_OP_READ, 0); comp_bio->bi_private = cb; comp_bio->bi_end_io = end_compressed_bio_read; - atomic_inc(&cb->pending_bios); + refcount_set(&cb->pending_bios, 1); for (pg_index = 0; pg_index < nr_pages; pg_index++) { page = cb->compressed_pages[pg_index]; @@ -685,7 +685,7 @@ int btrfs_submit_compressed_read(struct inode *inode, struct bio *bio, * we inc the count. Otherwise, the cb might get * freed before we're done setting it up */ - atomic_inc(&cb->pending_bios); + refcount_inc(&cb->pending_bios); if (!(BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM)) { ret = btrfs_lookup_bio_sums(inode, comp_bio, diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c index 1c3b6c54d5ee..a3a75f1de002 100644 --- a/fs/btrfs/ctree.c +++ b/fs/btrfs/ctree.c @@ -567,7 +567,7 @@ tree_mod_log_insert_key(struct btrfs_fs_info *fs_info, static noinline int tree_mod_log_insert_move(struct btrfs_fs_info *fs_info, struct extent_buffer *eb, int dst_slot, int src_slot, - int nr_items, gfp_t flags) + int nr_items) { struct tree_mod_elem *tm = NULL; struct tree_mod_elem **tm_list = NULL; @@ -578,11 +578,11 @@ tree_mod_log_insert_move(struct btrfs_fs_info *fs_info, if (!tree_mod_need_log(fs_info, eb)) return 0; - tm_list = kcalloc(nr_items, sizeof(struct tree_mod_elem *), flags); + tm_list = kcalloc(nr_items, sizeof(struct tree_mod_elem *), GFP_NOFS); if (!tm_list) return -ENOMEM; - tm = kzalloc(sizeof(*tm), flags); + tm = kzalloc(sizeof(*tm), GFP_NOFS); if (!tm) { ret = -ENOMEM; goto free_tms; @@ -596,7 +596,7 @@ tree_mod_log_insert_move(struct btrfs_fs_info *fs_info, for (i = 0; i + dst_slot < src_slot && i < nr_items; i++) { tm_list[i] = alloc_tree_mod_elem(eb, i + dst_slot, - MOD_LOG_KEY_REMOVE_WHILE_MOVING, flags); + MOD_LOG_KEY_REMOVE_WHILE_MOVING, GFP_NOFS); if (!tm_list[i]) { ret = -ENOMEM; goto free_tms; @@ -663,7 +663,7 @@ __tree_mod_log_free_eb(struct btrfs_fs_info *fs_info, static noinline int tree_mod_log_insert_root(struct btrfs_fs_info *fs_info, struct extent_buffer *old_root, - struct extent_buffer *new_root, gfp_t flags, + struct extent_buffer *new_root, int log_removal) { struct tree_mod_elem *tm = NULL; @@ -678,14 +678,14 @@ tree_mod_log_insert_root(struct btrfs_fs_info *fs_info, if (log_removal && btrfs_header_level(old_root) > 0) { nritems = btrfs_header_nritems(old_root); tm_list = kcalloc(nritems, sizeof(struct tree_mod_elem *), - flags); + GFP_NOFS); if (!tm_list) { ret = -ENOMEM; goto free_tms; } for (i = 0; i < nritems; i++) { tm_list[i] = alloc_tree_mod_elem(old_root, i, - MOD_LOG_KEY_REMOVE_WHILE_FREEING, flags); + MOD_LOG_KEY_REMOVE_WHILE_FREEING, GFP_NOFS); if (!tm_list[i]) { ret = -ENOMEM; goto free_tms; @@ -693,7 +693,7 @@ tree_mod_log_insert_root(struct btrfs_fs_info *fs_info, } } - tm = kzalloc(sizeof(*tm), flags); + tm = kzalloc(sizeof(*tm), GFP_NOFS); if (!tm) { ret = -ENOMEM; goto free_tms; @@ -873,7 +873,7 @@ tree_mod_log_eb_move(struct btrfs_fs_info *fs_info, struct extent_buffer *dst, { int ret; ret = tree_mod_log_insert_move(fs_info, dst, dst_offset, src_offset, - nr_items, GFP_NOFS); + nr_items); BUG_ON(ret < 0); } @@ -943,7 +943,7 @@ tree_mod_log_set_root_pointer(struct btrfs_root *root, { int ret; ret = tree_mod_log_insert_root(root->fs_info, root->node, - new_root_node, GFP_NOFS, log_removal); + new_root_node, log_removal); BUG_ON(ret < 0); } diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 3e21211e99c3..643c70d2b2e6 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -39,6 +39,7 @@ #include <linux/security.h> #include <linux/sizes.h> #include <linux/dynamic_debug.h> +#include <linux/refcount.h> #include "extent_io.h" #include "extent_map.h" #include "async-thread.h" @@ -518,7 +519,7 @@ struct btrfs_caching_control { struct btrfs_work work; struct btrfs_block_group_cache *block_group; u64 progress; - atomic_t count; + refcount_t count; }; /* Once caching_thread() finds this much free space, it will wake up waiters. */ @@ -538,6 +539,14 @@ struct btrfs_io_ctl { unsigned check_crcs:1; }; +/* + * Tree to record all locked full stripes of a RAID5/6 block group + */ +struct btrfs_full_stripe_locks_tree { + struct rb_root root; + struct mutex lock; +}; + struct btrfs_block_group_cache { struct btrfs_key key; struct btrfs_block_group_item item; @@ -648,6 +657,9 @@ struct btrfs_block_group_cache { * Protected by free_space_lock. */ int needs_free_space; + + /* Record locked full stripes for RAID5/6 block group */ + struct btrfs_full_stripe_locks_tree full_stripe_locks_root; }; /* delayed seq elem */ @@ -658,6 +670,8 @@ struct seq_list { #define SEQ_LIST_INIT(name) { .list = LIST_HEAD_INIT((name).list), .seq = 0 } +#define SEQ_LAST ((u64)-1) + enum btrfs_orphan_cleanup_state { ORPHAN_CLEANUP_STARTED = 1, ORPHAN_CLEANUP_DONE = 2, @@ -702,6 +716,11 @@ struct btrfs_delayed_root; #define BTRFS_FS_BTREE_ERR 11 #define BTRFS_FS_LOG1_ERR 12 #define BTRFS_FS_LOG2_ERR 13 +/* + * Indicate that a whole-filesystem exclusive operation is running + * (device replace, resize, device add/delete, balance) + */ +#define BTRFS_FS_EXCL_OP 14 struct btrfs_fs_info { u8 fsid[BTRFS_FSID_SIZE]; @@ -1066,8 +1085,6 @@ struct btrfs_fs_info { /* device replace state */ struct btrfs_dev_replace dev_replace; - atomic_t mutually_exclusive_operation_running; - struct percpu_counter bio_counter; wait_queue_head_t replace_wait; @@ -1220,7 +1237,7 @@ struct btrfs_root { dev_t anon_dev; spinlock_t root_item_lock; - atomic_t refs; + refcount_t refs; struct mutex delalloc_mutex; spinlock_t delalloc_lock; @@ -3646,6 +3663,12 @@ int btrfs_scrub_cancel_dev(struct btrfs_fs_info *info, struct btrfs_device *dev); int btrfs_scrub_progress(struct btrfs_fs_info *fs_info, u64 devid, struct btrfs_scrub_progress *progress); +static inline void btrfs_init_full_stripe_locks_tree( + struct btrfs_full_stripe_locks_tree *locks_root) +{ + locks_root->root = RB_ROOT; + mutex_init(&locks_root->lock); +} /* dev-replace.c */ void btrfs_bio_counter_inc_blocked(struct btrfs_fs_info *fs_info); @@ -3670,8 +3693,7 @@ struct reada_control *btrfs_reada_add(struct btrfs_root *root, struct btrfs_key *start, struct btrfs_key *end); int btrfs_reada_wait(void *handle); void btrfs_reada_detach(void *handle); -int btree_readahead_hook(struct btrfs_fs_info *fs_info, - struct extent_buffer *eb, int err); +int btree_readahead_hook(struct extent_buffer *eb, int err); static inline int is_fstree(u64 rootid) { diff --git a/fs/btrfs/delayed-inode.c b/fs/btrfs/delayed-inode.c index 1aff676f0e5b..8ae409b5a61d 100644 --- a/fs/btrfs/delayed-inode.c +++ b/fs/btrfs/delayed-inode.c @@ -52,7 +52,7 @@ static inline void btrfs_init_delayed_node( { delayed_node->root = root; delayed_node->inode_id = inode_id; - atomic_set(&delayed_node->refs, 0); + refcount_set(&delayed_node->refs, 0); delayed_node->ins_root = RB_ROOT; delayed_node->del_root = RB_ROOT; mutex_init(&delayed_node->mutex); @@ -81,7 +81,7 @@ static struct btrfs_delayed_node *btrfs_get_delayed_node( node = READ_ONCE(btrfs_inode->delayed_node); if (node) { - atomic_inc(&node->refs); + refcount_inc(&node->refs); return node; } @@ -89,14 +89,14 @@ static struct btrfs_delayed_node *btrfs_get_delayed_node( node = radix_tree_lookup(&root->delayed_nodes_tree, ino); if (node) { if (btrfs_inode->delayed_node) { - atomic_inc(&node->refs); /* can be accessed */ + refcount_inc(&node->refs); /* can be accessed */ BUG_ON(btrfs_inode->delayed_node != node); spin_unlock(&root->inode_lock); return node; } btrfs_inode->delayed_node = node; /* can be accessed and cached in the inode */ - atomic_add(2, &node->refs); + refcount_add(2, &node->refs); spin_unlock(&root->inode_lock); return node; } @@ -125,7 +125,7 @@ again: btrfs_init_delayed_node(node, root, ino); /* cached in the btrfs inode and can be accessed */ - atomic_add(2, &node->refs); + refcount_set(&node->refs, 2); ret = radix_tree_preload(GFP_NOFS); if (ret) { @@ -166,7 +166,7 @@ static void btrfs_queue_delayed_node(struct btrfs_delayed_root *root, } else { list_add_tail(&node->n_list, &root->node_list); list_add_tail(&node->p_list, &root->prepare_list); - atomic_inc(&node->refs); /* inserted into list */ + refcount_inc(&node->refs); /* inserted into list */ root->nodes++; set_bit(BTRFS_DELAYED_NODE_IN_LIST, &node->flags); } @@ -180,7 +180,7 @@ static void btrfs_dequeue_delayed_node(struct btrfs_delayed_root *root, spin_lock(&root->lock); if (test_bit(BTRFS_DELAYED_NODE_IN_LIST, &node->flags)) { root->nodes--; - atomic_dec(&node->refs); /* not in the list */ + refcount_dec(&node->refs); /* not in the list */ list_del_init(&node->n_list); if (!list_empty(&node->p_list)) list_del_init(&node->p_list); @@ -201,7 +201,7 @@ static struct btrfs_delayed_node *btrfs_first_delayed_node( p = delayed_root->node_list.next; node = list_entry(p, struct btrfs_delayed_node, n_list); - atomic_inc(&node->refs); + refcount_inc(&node->refs); out: spin_unlock(&delayed_root->lock); @@ -228,7 +228,7 @@ static struct btrfs_delayed_node *btrfs_next_delayed_node( p = node->n_list.next; next = list_entry(p, struct btrfs_delayed_node, n_list); - atomic_inc(&next->refs); + refcount_inc(&next->refs); out: spin_unlock(&delayed_root->lock); @@ -253,11 +253,11 @@ static void __btrfs_release_delayed_node( btrfs_dequeue_delayed_node(delayed_root, delayed_node); mutex_unlock(&delayed_node->mutex); - if (atomic_dec_and_test(&delayed_node->refs)) { + if (refcount_dec_and_test(&delayed_node->refs)) { bool free = false; struct btrfs_root *root = delayed_node->root; spin_lock(&root->inode_lock); - if (atomic_read(&delayed_node->refs) == 0) { + if (refcount_read(&delayed_node->refs) == 0) { radix_tree_delete(&root->delayed_nodes_tree, delayed_node->inode_id); free = true; @@ -286,7 +286,7 @@ static struct btrfs_delayed_node *btrfs_first_prepared_delayed_node( p = delayed_root->prepare_list.next; list_del_init(p); node = list_entry(p, struct btrfs_delayed_node, p_list); - atomic_inc(&node->refs); + refcount_inc(&node->refs); out: spin_unlock(&delayed_root->lock); @@ -308,7 +308,7 @@ static struct btrfs_delayed_item *btrfs_alloc_delayed_item(u32 data_len) item->ins_or_del = 0; item->bytes_reserved = 0; item->delayed_node = NULL; - atomic_set(&item->refs, 1); + refcount_set(&item->refs, 1); } return item; } @@ -483,7 +483,7 @@ static void btrfs_release_delayed_item(struct btrfs_delayed_item *item) { if (item) { __btrfs_remove_delayed_item(item); - if (atomic_dec_and_test(&item->refs)) + if (refcount_dec_and_test(&item->refs)) kfree(item); } } @@ -1600,14 +1600,14 @@ bool btrfs_readdir_get_delayed_items(struct inode *inode, mutex_lock(&delayed_node->mutex); item = __btrfs_first_delayed_insertion_item(delayed_node); while (item) { - atomic_inc(&item->refs); + refcount_inc(&item->refs); list_add_tail(&item->readdir_list, ins_list); item = __btrfs_next_delayed_item(item); } item = __btrfs_first_delayed_deletion_item(delayed_node); while (item) { - atomic_inc(&item->refs); + refcount_inc(&item->refs); list_add_tail(&item->readdir_list, del_list); item = __btrfs_next_delayed_item(item); } @@ -1621,7 +1621,7 @@ bool btrfs_readdir_get_delayed_items(struct inode *inode, * insert/delete delayed items in this period. So we also needn't * requeue or dequeue this delayed node. */ - atomic_dec(&delayed_node->refs); + refcount_dec(&delayed_node->refs); return true; } @@ -1634,13 +1634,13 @@ void btrfs_readdir_put_delayed_items(struct inode *inode, list_for_each_entry_safe(curr, next, ins_list, readdir_list) { list_del(&curr->readdir_list); - if (atomic_dec_and_test(&curr->refs)) + if (refcount_dec_and_test(&curr->refs)) kfree(curr); } list_for_each_entry_safe(curr, next, del_list, readdir_list) { list_del(&curr->readdir_list); - if (atomic_dec_and_test(&curr->refs)) + if (refcount_dec_and_test(&curr->refs)) kfree(curr); } @@ -1667,7 +1667,7 @@ int btrfs_should_delete_dir_index(struct list_head *del_list, list_del(&curr->readdir_list); ret = (curr->key.offset == index); - if (atomic_dec_and_test(&curr->refs)) + if (refcount_dec_and_test(&curr->refs)) kfree(curr); if (ret) @@ -1705,7 +1705,7 @@ int btrfs_readdir_delayed_dir_index(struct dir_context *ctx, list_del(&curr->readdir_list); if (curr->key.offset < ctx->pos) { - if (atomic_dec_and_test(&curr->refs)) + if (refcount_dec_and_test(&curr->refs)) kfree(curr); continue; } @@ -1722,7 +1722,7 @@ int btrfs_readdir_delayed_dir_index(struct dir_context *ctx, over = !dir_emit(ctx, name, name_len, location.objectid, d_type); - if (atomic_dec_and_test(&curr->refs)) + if (refcount_dec_and_test(&curr->refs)) kfree(curr); if (over) @@ -1963,7 +1963,7 @@ void btrfs_kill_all_delayed_nodes(struct btrfs_root *root) inode_id = delayed_nodes[n - 1]->inode_id + 1; for (i = 0; i < n; i++) - atomic_inc(&delayed_nodes[i]->refs); + refcount_inc(&delayed_nodes[i]->refs); spin_unlock(&root->inode_lock); for (i = 0; i < n; i++) { diff --git a/fs/btrfs/delayed-inode.h b/fs/btrfs/delayed-inode.h index 40327cc3b99a..c4189d495934 100644 --- a/fs/btrfs/delayed-inode.h +++ b/fs/btrfs/delayed-inode.h @@ -26,7 +26,7 @@ #include <linux/list.h> #include <linux/wait.h> #include <linux/atomic.h> - +#include <linux/refcount.h> #include "ctree.h" /* types of the delayed item */ @@ -67,7 +67,7 @@ struct btrfs_delayed_node { struct rb_root del_root; struct mutex mutex; struct btrfs_inode_item inode_item; - atomic_t refs; + refcount_t refs; u64 index_cnt; unsigned long flags; int count; @@ -80,7 +80,7 @@ struct btrfs_delayed_item { struct list_head readdir_list; /* used for readdir items */ u64 bytes_reserved; struct btrfs_delayed_node *delayed_node; - atomic_t refs; + refcount_t refs; int ins_or_del; u32 data_len; char data[0]; diff --git a/fs/btrfs/delayed-ref.c b/fs/btrfs/delayed-ref.c index 6eb80952efb3..be70d90dfee5 100644 --- a/fs/btrfs/delayed-ref.c +++ b/fs/btrfs/delayed-ref.c @@ -164,7 +164,7 @@ int btrfs_delayed_ref_lock(struct btrfs_trans_handle *trans, if (mutex_trylock(&head->mutex)) return 0; - atomic_inc(&head->node.refs); + refcount_inc(&head->node.refs); spin_unlock(&delayed_refs->lock); mutex_lock(&head->mutex); @@ -590,7 +590,7 @@ add_delayed_ref_head(struct btrfs_fs_info *fs_info, delayed_refs = &trans->transaction->delayed_refs; /* first set the basic ref node struct up */ - atomic_set(&ref->refs, 1); + refcount_set(&ref->refs, 1); ref->bytenr = bytenr; ref->num_bytes = num_bytes; ref->ref_mod = count_mod; @@ -682,7 +682,7 @@ add_delayed_tree_ref(struct btrfs_fs_info *fs_info, delayed_refs = &trans->transaction->delayed_refs; /* first set the basic ref node struct up */ - atomic_set(&ref->refs, 1); + refcount_set(&ref->refs, 1); ref->bytenr = bytenr; ref->num_bytes = num_bytes; ref->ref_mod = 1; @@ -739,7 +739,7 @@ add_delayed_data_ref(struct btrfs_fs_info *fs_info, seq = atomic64_read(&fs_info->tree_mod_seq); /* first set the basic ref node struct up */ - atomic_set(&ref->refs, 1); + refcount_set(&ref->refs, 1); ref->bytenr = bytenr; ref->num_bytes = num_bytes; ref->ref_mod = 1; diff --git a/fs/btrfs/delayed-ref.h b/fs/btrfs/delayed-ref.h index 0e537f98f1a1..c0264ff01b53 100644 --- a/fs/btrfs/delayed-ref.h +++ b/fs/btrfs/delayed-ref.h @@ -18,6 +18,8 @@ #ifndef __DELAYED_REF__ #define __DELAYED_REF__ +#include <linux/refcount.h> + /* these are the possible values of struct btrfs_delayed_ref_node->action */ #define BTRFS_ADD_DELAYED_REF 1 /* add one backref to the tree */ #define BTRFS_DROP_DELAYED_REF 2 /* delete one backref from the tree */ @@ -53,7 +55,7 @@ struct btrfs_delayed_ref_node { u64 seq; /* ref count on this data structure */ - atomic_t refs; + refcount_t refs; /* * how many refs is this entry adding or deleting. For @@ -220,8 +222,8 @@ btrfs_free_delayed_extent_op(struct btrfs_delayed_extent_op *op) static inline void btrfs_put_delayed_ref(struct btrfs_delayed_ref_node *ref) { - WARN_ON(atomic_read(&ref->refs) == 0); - if (atomic_dec_and_test(&ref->refs)) { + WARN_ON(refcount_read(&ref->refs) == 0); + if (refcount_dec_and_test(&ref->refs)) { WARN_ON(ref->in_tree); switch (ref->type) { case BTRFS_TREE_BLOCK_REF_KEY: diff --git a/fs/btrfs/dev-replace.c b/fs/btrfs/dev-replace.c index e653921f05d9..5fe1ca8abc70 100644 --- a/fs/btrfs/dev-replace.c +++ b/fs/btrfs/dev-replace.c @@ -546,8 +546,10 @@ static int btrfs_dev_replace_finishing(struct btrfs_fs_info *fs_info, mutex_unlock(&fs_info->chunk_mutex); mutex_unlock(&fs_info->fs_devices->device_list_mutex); mutex_unlock(&uuid_mutex); + btrfs_rm_dev_replace_blocked(fs_info); if (tgt_device) btrfs_destroy_dev_replace_tgtdev(fs_info, tgt_device); + btrfs_rm_dev_replace_unblocked(fs_info); mutex_unlock(&dev_replace->lock_finishing_cancel_unmount); return scrub_ret; @@ -665,7 +667,7 @@ void btrfs_dev_replace_status(struct btrfs_fs_info *fs_info, case BTRFS_IOCTL_DEV_REPLACE_STATE_STARTED: case BTRFS_IOCTL_DEV_REPLACE_STATE_SUSPENDED: srcdev = dev_replace->srcdev; - args->status.progress_1000 = div_u64(dev_replace->cursor_left, + args->status.progress_1000 = div64_u64(dev_replace->cursor_left, div_u64(btrfs_device_get_total_bytes(srcdev), 1000)); break; } @@ -784,8 +786,7 @@ int btrfs_resume_dev_replace_async(struct btrfs_fs_info *fs_info) } btrfs_dev_replace_unlock(dev_replace, 1); - WARN_ON(atomic_xchg( - &fs_info->mutually_exclusive_operation_running, 1)); + WARN_ON(test_and_set_bit(BTRFS_FS_EXCL_OP, &fs_info->flags)); task = kthread_run(btrfs_dev_replace_kthread, fs_info, "btrfs-devrepl"); return PTR_ERR_OR_ZERO(task); } @@ -814,7 +815,7 @@ static int btrfs_dev_replace_kthread(void *data) (unsigned int)progress); } btrfs_dev_replace_continue_on_mount(fs_info); - atomic_set(&fs_info->mutually_exclusive_operation_running, 0); + clear_bit(BTRFS_FS_EXCL_OP, &fs_info->flags); return 0; } diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 061c1d1f774f..8685d67185d0 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -762,7 +762,7 @@ static int btree_readpage_end_io_hook(struct btrfs_io_bio *io_bio, err: if (reads_done && test_and_clear_bit(EXTENT_BUFFER_READAHEAD, &eb->bflags)) - btree_readahead_hook(fs_info, eb, ret); + btree_readahead_hook(eb, ret); if (ret) { /* @@ -787,7 +787,7 @@ static int btree_io_failed_hook(struct page *page, int failed_mirror) eb->read_mirror = failed_mirror; atomic_dec(&eb->io_pages); if (test_and_clear_bit(EXTENT_BUFFER_READAHEAD, &eb->bflags)) - btree_readahead_hook(eb->fs_info, eb, -EIO); + btree_readahead_hook(eb, -EIO); return -EIO; /* we fixed nothing */ } @@ -1340,7 +1340,7 @@ static void __setup_root(struct btrfs_root *root, struct btrfs_fs_info *fs_info, atomic_set(&root->log_writers, 0); atomic_set(&root->log_batch, 0); atomic_set(&root->orphan_inodes, 0); - atomic_set(&root->refs, 1); + refcount_set(&root->refs, 1); atomic_set(&root->will_be_snapshoted, 0); atomic64_set(&root->qgroup_meta_rsv, 0); root->log_transid = 0; @@ -3497,10 +3497,11 @@ static void btrfs_end_empty_barrier(struct bio *bio) */ static int write_dev_flush(struct btrfs_device *device, int wait) { + struct request_queue *q = bdev_get_queue(device->bdev); struct bio *bio; int ret = 0; - if (device->nobarriers) + if (!test_bit(QUEUE_FLAG_WC, &q->queue_flags)) return 0; if (wait) { @@ -4321,7 +4322,7 @@ static int btrfs_destroy_delayed_refs(struct btrfs_transaction *trans, head = rb_entry(node, struct btrfs_delayed_ref_head, href_node); if (!mutex_trylock(&head->mutex)) { - atomic_inc(&head->node.refs); + refcount_inc(&head->node.refs); spin_unlock(&delayed_refs->lock); mutex_lock(&head->mutex); @@ -4593,7 +4594,7 @@ static int btrfs_cleanup_transaction(struct btrfs_fs_info *fs_info) t = list_first_entry(&fs_info->trans_list, struct btrfs_transaction, list); if (t->state >= TRANS_STATE_COMMIT_START) { - atomic_inc(&t->use_count); + refcount_inc(&t->use_count); spin_unlock(&fs_info->trans_lock); btrfs_wait_for_commit(fs_info, t->transid); btrfs_put_transaction(t); diff --git a/fs/btrfs/disk-io.h b/fs/btrfs/disk-io.h index 2e0ec29bfd69..21f1ceb85b76 100644 --- a/fs/btrfs/disk-io.h +++ b/fs/btrfs/disk-io.h @@ -101,14 +101,14 @@ struct btrfs_root *btrfs_alloc_dummy_root(struct btrfs_fs_info *fs_info); */ static inline struct btrfs_root *btrfs_grab_fs_root(struct btrfs_root *root) { - if (atomic_inc_not_zero(&root->refs)) + if (refcount_inc_not_zero(&root->refs)) return root; return NULL; } static inline void btrfs_put_fs_root(struct btrfs_root *root) { - if (atomic_dec_and_test(&root->refs)) + if (refcount_dec_and_test(&root->refs)) kfree(root); } diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index be5477676cc8..e390451c72e6 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -131,6 +131,16 @@ void btrfs_put_block_group(struct btrfs_block_group_cache *cache) if (atomic_dec_and_test(&cache->count)) { WARN_ON(cache->pinned > 0); WARN_ON(cache->reserved > 0); + + /* + * If not empty, someone is still holding mutex of + * full_stripe_lock, which can only be released by caller. + * And it will definitely cause use-after-free when caller + * tries to release full stripe lock. + * + * No better way to resolve, but only to warn. + */ + WARN_ON(!RB_EMPTY_ROOT(&cache->full_stripe_locks_root.root)); kfree(cache->free_space_ctl); kfree(cache); } @@ -316,14 +326,14 @@ get_caching_control(struct btrfs_block_group_cache *cache) } ctl = cache->caching_ctl; - atomic_inc(&ctl->count); + refcount_inc(&ctl->count); spin_unlock(&cache->lock); return ctl; } static void put_caching_control(struct btrfs_caching_control *ctl) { - if (atomic_dec_and_test(&ctl->count)) + if (refcount_dec_and_test(&ctl->count)) kfree(ctl); } @@ -599,7 +609,7 @@ static int cache_block_group(struct btrfs_block_group_cache *cache, init_waitqueue_head(&caching_ctl->wait); caching_ctl->block_group = cache; caching_ctl->progress = cache->key.objectid; - atomic_set(&caching_ctl->count, 1); + refcount_set(&caching_ctl->count, 1); btrfs_init_work(&caching_ctl->work, btrfs_cache_helper, caching_thread, NULL, NULL); @@ -620,7 +630,7 @@ static int cache_block_group(struct btrfs_block_group_cache *cache, struct btrfs_caching_control *ctl; ctl = cache->caching_ctl; - atomic_inc(&ctl->count); + refcount_inc(&ctl->count); prepare_to_wait(&ctl->wait, &wait, TASK_UNINTERRUPTIBLE); spin_unlock(&cache->lock); @@ -707,7 +717,7 @@ static int cache_block_group(struct btrfs_block_group_cache *cache, } down_write(&fs_info->commit_root_sem); - atomic_inc(&caching_ctl->count); + refcount_inc(&caching_ctl->count); list_add_tail(&caching_ctl->list, &fs_info->caching_block_groups); up_write(&fs_info->commit_root_sem); @@ -892,7 +902,7 @@ search_again: head = btrfs_find_delayed_ref_head(delayed_refs, bytenr); if (head) { if (!mutex_trylock(&head->mutex)) { - atomic_inc(&head->node.refs); + refcount_inc(&head->node.refs); spin_unlock(&delayed_refs->lock); btrfs_release_path(path); @@ -2980,7 +2990,7 @@ again: struct btrfs_delayed_ref_node *ref; ref = &head->node; - atomic_inc(&ref->refs); + refcount_inc(&ref->refs); spin_unlock(&delayed_refs->lock); /* @@ -3003,7 +3013,6 @@ again: goto again; } out: - assert_qgroups_uptodate(trans); trans->can_flush_pending_bgs = can_flush_pending_bgs; return 0; } @@ -3057,7 +3066,7 @@ static noinline int check_delayed_ref(struct btrfs_root *root, } if (!mutex_trylock(&head->mutex)) { - atomic_inc(&head->node.refs); + refcount_inc(&head->node.refs); spin_unlock(&delayed_refs->lock); btrfs_release_path(path); @@ -3443,7 +3452,8 @@ again: /* * don't bother trying to write stuff out _if_ * a) we're not cached, - * b) we're with nospace_cache mount option. + * b) we're with nospace_cache mount option, + * c) we're with v2 space_cache (FREE_SPACE_TREE). */ dcs = BTRFS_DC_WRITTEN; spin_unlock(&block_group->lock); @@ -9917,6 +9927,7 @@ btrfs_create_block_group_cache(struct btrfs_fs_info *fs_info, btrfs_init_free_space_ctl(cache); atomic_set(&cache->trimming, 0); mutex_init(&cache->free_space_lock); + btrfs_init_full_stripe_locks_tree(&cache->full_stripe_locks_root); return cache; } @@ -10416,7 +10427,7 @@ int btrfs_remove_block_group(struct btrfs_trans_handle *trans, &fs_info->caching_block_groups, list) if (ctl->block_group == block_group) { caching_ctl = ctl; - atomic_inc(&caching_ctl->count); + refcount_inc(&caching_ctl->count); break; } } @@ -10850,7 +10861,7 @@ static int btrfs_trim_free_extents(struct btrfs_device *device, spin_lock(&fs_info->trans_lock); trans = fs_info->running_transaction; if (trans) - atomic_inc(&trans->use_count); + refcount_inc(&trans->use_count); spin_unlock(&fs_info->trans_lock); ret = find_free_dev_extent_start(trans, device, minlen, start, diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index 27fdb250b446..d8da3edf2ac3 100644 --- a/fs/btrfs/extent_io.c +++ b/fs/btrfs/extent_io.c @@ -68,7 +68,7 @@ void btrfs_leak_debug_check(void) pr_err("BTRFS: state leak: start %llu end %llu state %u in tree %d refs %d\n", state->start, state->end, state->state, extent_state_in_tree(state), - atomic_read(&state->refs)); + refcount_read(&state->refs)); list_del(&state->leak_list); kmem_cache_free(extent_state_cache, state); } @@ -238,7 +238,7 @@ static struct extent_state *alloc_extent_state(gfp_t mask) state->failrec = NULL; RB_CLEAR_NODE(&state->rb_node); btrfs_leak_debug_add(&state->leak_list, &states); - atomic_set(&state->refs, 1); + refcount_set(&state->refs, 1); init_waitqueue_head(&state->wq); trace_alloc_extent_state(state, mask, _RET_IP_); return state; @@ -248,7 +248,7 @@ void free_extent_state(struct extent_state *state) { if (!state) return; - if (atomic_dec_and_test(&state->refs)) { + if (refcount_dec_and_test(&state->refs)) { WARN_ON(extent_state_in_tree(state)); btrfs_leak_debug_del(&state->leak_list); trace_free_extent_state(state, _RET_IP_); @@ -641,7 +641,7 @@ again: if (cached && extent_state_in_tree(cached) && cached->start <= start && cached->end > start) { if (clear) - atomic_dec(&cached->refs); + refcount_dec(&cached->refs); state = cached; goto hit_next; } @@ -793,7 +793,7 @@ process_node: if (state->state & bits) { start = state->start; - atomic_inc(&state->refs); + refcount_inc(&state->refs); wait_on_state(tree, state); free_extent_state(state); goto again; @@ -834,7 +834,7 @@ static void cache_state_if_flags(struct extent_state *state, if (cached_ptr && !(*cached_ptr)) { if (!flags || (state->state & flags)) { *cached_ptr = state; - atomic_inc(&state->refs); + refcount_inc(&state->refs); } } } @@ -1538,7 +1538,7 @@ static noinline u64 find_delalloc_range(struct extent_io_tree *tree, if (!found) { *start = state->start; *cached_state = state; - atomic_inc(&state->refs); + refcount_inc(&state->refs); } found++; *end = state->end; @@ -2004,16 +2004,11 @@ int repair_io_failure(struct btrfs_inode *inode, u64 start, u64 length, u64 map_length = 0; u64 sector; struct btrfs_bio *bbio = NULL; - struct btrfs_mapping_tree *map_tree = &fs_info->mapping_tree; int ret; ASSERT(!(fs_info->sb->s_flags & MS_RDONLY)); BUG_ON(!mirror_num); - /* we can't repair anything in raid56 yet */ - if (btrfs_is_parity_mirror(map_tree, logical, length, mirror_num)) - return 0; - bio = btrfs_io_bio_alloc(GFP_NOFS, 1); if (!bio) return -EIO; @@ -2026,17 +2021,35 @@ int repair_io_failure(struct btrfs_inode *inode, u64 start, u64 length, * read repair operation. */ btrfs_bio_counter_inc_blocked(fs_info); - ret = btrfs_map_block(fs_info, BTRFS_MAP_WRITE, logical, - &map_length, &bbio, mirror_num); - if (ret) { - btrfs_bio_counter_dec(fs_info); - bio_put(bio); - return -EIO; + if (btrfs_is_parity_mirror(fs_info, logical, length, mirror_num)) { + /* + * Note that we don't use BTRFS_MAP_WRITE because it's supposed + * to update all raid stripes, but here we just want to correct + * bad stripe, thus BTRFS_MAP_READ is abused to only get the bad + * stripe's dev and sector. + */ + ret = btrfs_map_block(fs_info, BTRFS_MAP_READ, logical, + &map_length, &bbio, 0); + if (ret) { + btrfs_bio_counter_dec(fs_info); + bio_put(bio); + return -EIO; + } + ASSERT(bbio->mirror_num == 1); + } else { + ret = btrfs_map_block(fs_info, BTRFS_MAP_WRITE, logical, + &map_length, &bbio, mirror_num); + if (ret) { + btrfs_bio_counter_dec(fs_info); + bio_put(bio); + return -EIO; + } + BUG_ON(mirror_num != bbio->mirror_num); } - BUG_ON(mirror_num != bbio->mirror_num); - sector = bbio->stripes[mirror_num-1].physical >> 9; + + sector = bbio->stripes[bbio->mirror_num - 1].physical >> 9; bio->bi_iter.bi_sector = sector; - dev = bbio->stripes[mirror_num-1].dev; + dev = bbio->stripes[bbio->mirror_num - 1].dev; btrfs_put_bbio(bbio); if (!dev || !dev->bdev || !dev->writeable) { btrfs_bio_counter_dec(fs_info); @@ -2859,7 +2872,7 @@ __get_extent_map(struct inode *inode, struct page *page, size_t pg_offset, em = *em_cached; if (extent_map_in_tree(em) && start >= em->start && start < extent_map_end(em)) { - atomic_inc(&em->refs); + refcount_inc(&em->refs); return em; } @@ -2870,7 +2883,7 @@ __get_extent_map(struct inode *inode, struct page *page, size_t pg_offset, em = get_extent(BTRFS_I(inode), page, pg_offset, start, len, 0); if (em_cached && !IS_ERR_OR_NULL(em)) { BUG_ON(*em_cached); - atomic_inc(&em->refs); + refcount_inc(&em->refs); *em_cached = em; } return em; diff --git a/fs/btrfs/extent_io.h b/fs/btrfs/extent_io.h index 3e4fad4a909d..1eafa2f0ede3 100644 --- a/fs/btrfs/extent_io.h +++ b/fs/btrfs/extent_io.h @@ -2,6 +2,7 @@ #define __EXTENTIO__ #include <linux/rbtree.h> +#include <linux/refcount.h> #include "ulist.h" /* bits for the extent state */ @@ -14,14 +15,17 @@ #define EXTENT_DEFRAG (1U << 6) #define EXTENT_BOUNDARY (1U << 9) #define EXTENT_NODATASUM (1U << 10) -#define EXTENT_DO_ACCOUNTING (1U << 11) +#define EXTENT_CLEAR_META_RESV (1U << 11) #define EXTENT_FIRST_DELALLOC (1U << 12) #define EXTENT_NEED_WAIT (1U << 13) #define EXTENT_DAMAGED (1U << 14) #define EXTENT_NORESERVE (1U << 15) #define EXTENT_QGROUP_RESERVED (1U << 16) #define EXTENT_CLEAR_DATA_RESV (1U << 17) +#define EXTENT_DELALLOC_NEW (1U << 18) #define EXTENT_IOBITS (EXTENT_LOCKED | EXTENT_WRITEBACK) +#define EXTENT_DO_ACCOUNTING (EXTENT_CLEAR_META_RESV | \ + EXTENT_CLEAR_DATA_RESV) #define EXTENT_CTLBITS (EXTENT_DO_ACCOUNTING | EXTENT_FIRST_DELALLOC) /* @@ -143,7 +147,7 @@ struct extent_state { /* ADD NEW ELEMENTS AFTER THIS */ wait_queue_head_t wq; - atomic_t refs; + refcount_t refs; unsigned state; struct io_failure_record *failrec; diff --git a/fs/btrfs/extent_map.c b/fs/btrfs/extent_map.c index 26f9ac719d20..69850155870c 100644 --- a/fs/btrfs/extent_map.c +++ b/fs/btrfs/extent_map.c @@ -55,7 +55,7 @@ struct extent_map *alloc_extent_map(void) em->flags = 0; em->compress_type = BTRFS_COMPRESS_NONE; em->generation = 0; - atomic_set(&em->refs, 1); + refcount_set(&em->refs, 1); INIT_LIST_HEAD(&em->list); return em; } @@ -71,8 +71,8 @@ void free_extent_map(struct extent_map *em) { if (!em) return; - WARN_ON(atomic_read(&em->refs) == 0); - if (atomic_dec_and_test(&em->refs)) { + WARN_ON(refcount_read(&em->refs) == 0); + if (refcount_dec_and_test(&em->refs)) { WARN_ON(extent_map_in_tree(em)); WARN_ON(!list_empty(&em->list)); if (test_bit(EXTENT_FLAG_FS_MAPPING, &em->flags)) @@ -322,7 +322,7 @@ static inline void setup_extent_mapping(struct extent_map_tree *tree, struct extent_map *em, int modified) { - atomic_inc(&em->refs); + refcount_inc(&em->refs); em->mod_start = em->start; em->mod_len = em->len; @@ -381,7 +381,7 @@ __lookup_extent_mapping(struct extent_map_tree *tree, if (strict && !(end > em->start && start < extent_map_end(em))) return NULL; - atomic_inc(&em->refs); + refcount_inc(&em->refs); return em; } diff --git a/fs/btrfs/extent_map.h b/fs/btrfs/extent_map.h index eb8b8fae036b..a67b2def5413 100644 --- a/fs/btrfs/extent_map.h +++ b/fs/btrfs/extent_map.h @@ -2,6 +2,7 @@ #define __EXTENTMAP__ #include <linux/rbtree.h> +#include <linux/refcount.h> #define EXTENT_MAP_LAST_BYTE ((u64)-4) #define EXTENT_MAP_HOLE ((u64)-3) @@ -41,7 +42,7 @@ struct extent_map { */ struct map_lookup *map_lookup; }; - atomic_t refs; + refcount_t refs; unsigned int compress_type; struct list_head list; }; diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index 520cb7230b2d..da1096eb1a40 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -1404,6 +1404,47 @@ fail: } +static int btrfs_find_new_delalloc_bytes(struct btrfs_inode *inode, + const u64 start, + const u64 len, + struct extent_state **cached_state) +{ + u64 search_start = start; + const u64 end = start + len - 1; + + while (search_start < end) { + const u64 search_len = end - search_start + 1; + struct extent_map *em; + u64 em_len; + int ret = 0; + + em = btrfs_get_extent(inode, NULL, 0, search_start, + search_len, 0); + if (IS_ERR(em)) + return PTR_ERR(em); + + if (em->block_start != EXTENT_MAP_HOLE) + goto next; + + em_len = em->len; + if (em->start < search_start) + em_len -= search_start - em->start; + if (em_len > search_len) + em_len = search_len; + + ret = set_extent_bit(&inode->io_tree, search_start, + search_start + em_len - 1, + EXTENT_DELALLOC_NEW, + NULL, cached_state, GFP_NOFS); +next: + search_start = extent_map_end(em); + free_extent_map(em); + if (ret) + return ret; + } + return 0; +} + /* * This function locks the extent and properly waits for data=ordered extents * to finish before allowing the pages to be modified if need. @@ -1432,8 +1473,11 @@ lock_and_cleanup_extent_if_need(struct btrfs_inode *inode, struct page **pages, + round_up(pos + write_bytes - start_pos, fs_info->sectorsize) - 1; - if (start_pos < inode->vfs_inode.i_size) { + if (start_pos < inode->vfs_inode.i_size || + (inode->flags & BTRFS_INODE_PREALLOC)) { struct btrfs_ordered_extent *ordered; + unsigned int clear_bits; + lock_extent_bits(&inode->io_tree, start_pos, last_pos, cached_state); ordered = btrfs_lookup_ordered_range(inode, start_pos, @@ -1454,11 +1498,19 @@ lock_and_cleanup_extent_if_need(struct btrfs_inode *inode, struct page **pages, } if (ordered) btrfs_put_ordered_extent(ordered); - + ret = btrfs_find_new_delalloc_bytes(inode, start_pos, + last_pos - start_pos + 1, + cached_state); + clear_bits = EXTENT_DIRTY | EXTENT_DELALLOC | + EXTENT_DO_ACCOUNTING | EXTENT_DEFRAG; + if (ret) + clear_bits |= EXTENT_DELALLOC_NEW | EXTENT_LOCKED; clear_extent_bit(&inode->io_tree, start_pos, - last_pos, EXTENT_DIRTY | EXTENT_DELALLOC | - EXTENT_DO_ACCOUNTING | EXTENT_DEFRAG, - 0, 0, cached_state, GFP_NOFS); + last_pos, clear_bits, + (clear_bits & EXTENT_LOCKED) ? 1 : 0, + 0, cached_state, GFP_NOFS); + if (ret) + return ret; *lockstart = start_pos; *lockend = last_pos; ret = 1; @@ -2342,13 +2394,8 @@ static int find_first_non_hole(struct inode *inode, u64 *start, u64 *len) int ret = 0; em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, *start, *len, 0); - if (IS_ERR_OR_NULL(em)) { - if (!em) - ret = -ENOMEM; - else - ret = PTR_ERR(em); - return ret; - } + if (IS_ERR(em)) + return PTR_ERR(em); /* Hole or vacuum extent(only exists in no-hole mode) */ if (em->block_start == EXTENT_MAP_HOLE) { @@ -2835,11 +2882,8 @@ static long btrfs_fallocate(struct file *file, int mode, while (1) { em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, cur_offset, alloc_end - cur_offset, 0); - if (IS_ERR_OR_NULL(em)) { - if (!em) - ret = -ENOMEM; - else - ret = PTR_ERR(em); + if (IS_ERR(em)) { + ret = PTR_ERR(em); break; } last_byte = min(extent_map_end(em), alloc_end); @@ -2856,8 +2900,10 @@ static long btrfs_fallocate(struct file *file, int mode, } ret = btrfs_qgroup_reserve_data(inode, cur_offset, last_byte - cur_offset); - if (ret < 0) + if (ret < 0) { + free_extent_map(em); break; + } } else { /* * Do not need to reserve unwritten extent for this diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c index da6841efac26..c5e6180cdb8c 100644 --- a/fs/btrfs/free-space-cache.c +++ b/fs/btrfs/free-space-cache.c @@ -355,7 +355,7 @@ static void io_ctl_map_page(struct btrfs_io_ctl *io_ctl, int clear) io_ctl->orig = io_ctl->cur; io_ctl->size = PAGE_SIZE; if (clear) - memset(io_ctl->cur, 0, PAGE_SIZE); + clear_page(io_ctl->cur); } static void io_ctl_drop_pages(struct btrfs_io_ctl *io_ctl) diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 5e71f1ea3391..17cbe9306faf 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -115,6 +115,31 @@ static struct extent_map *create_io_em(struct inode *inode, u64 start, u64 len, u64 ram_bytes, int compress_type, int type); +static void __endio_write_update_ordered(struct inode *inode, + const u64 offset, const u64 bytes, + const bool uptodate); + +/* + * Cleanup all submitted ordered extents in specified range to handle errors + * from the fill_dellaloc() callback. + * + * NOTE: caller must ensure that when an error happens, it can not call + * extent_clear_unlock_delalloc() to clear both the bits EXTENT_DO_ACCOUNTING + * and EXTENT_DELALLOC simultaneously, because that causes the reserved metadata + * to be released, which we want to happen only when finishing the ordered + * extent (btrfs_finish_ordered_io()). Also note that the caller of the + * fill_delalloc() callback already does proper cleanup for the first page of + * the range, that is, it invokes the callback writepage_end_io_hook() for the + * range of the first page. + */ +static inline void btrfs_cleanup_ordered_extents(struct inode *inode, + const u64 offset, + const u64 bytes) +{ + return __endio_write_update_ordered(inode, offset + PAGE_SIZE, + bytes - PAGE_SIZE, false); +} + static int btrfs_dirty_inode(struct inode *inode); #ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS @@ -547,7 +572,7 @@ cont: } if (ret <= 0) { unsigned long clear_flags = EXTENT_DELALLOC | - EXTENT_DEFRAG; + EXTENT_DELALLOC_NEW | EXTENT_DEFRAG; unsigned long page_error_op; clear_flags |= (ret < 0) ? EXTENT_DO_ACCOUNTING : 0; @@ -565,8 +590,10 @@ cont: PAGE_SET_WRITEBACK | page_error_op | PAGE_END_WRITEBACK); - btrfs_free_reserved_data_space_noquota(inode, start, - end - start + 1); + if (ret == 0) + btrfs_free_reserved_data_space_noquota(inode, + start, + end - start + 1); goto free_pages_out; } } @@ -852,6 +879,7 @@ out_free: async_extent->start + async_extent->ram_size - 1, NULL, EXTENT_LOCKED | EXTENT_DELALLOC | + EXTENT_DELALLOC_NEW | EXTENT_DEFRAG | EXTENT_DO_ACCOUNTING, PAGE_UNLOCK | PAGE_CLEAR_DIRTY | PAGE_SET_WRITEBACK | PAGE_END_WRITEBACK | @@ -918,10 +946,13 @@ static noinline int cow_file_range(struct inode *inode, u64 num_bytes; unsigned long ram_size; u64 disk_num_bytes; - u64 cur_alloc_size; + u64 cur_alloc_size = 0; u64 blocksize = fs_info->sectorsize; struct btrfs_key ins; struct extent_map *em; + unsigned clear_bits; + unsigned long page_ops; + bool extent_reserved = false; int ret = 0; if (btrfs_is_free_space_inode(BTRFS_I(inode))) { @@ -944,6 +975,7 @@ static noinline int cow_file_range(struct inode *inode, extent_clear_unlock_delalloc(inode, start, end, delalloc_end, NULL, EXTENT_LOCKED | EXTENT_DELALLOC | + EXTENT_DELALLOC_NEW | EXTENT_DEFRAG, PAGE_UNLOCK | PAGE_CLEAR_DIRTY | PAGE_SET_WRITEBACK | PAGE_END_WRITEBACK); @@ -966,14 +998,14 @@ static noinline int cow_file_range(struct inode *inode, start + num_bytes - 1, 0); while (disk_num_bytes > 0) { - unsigned long op; - cur_alloc_size = disk_num_bytes; ret = btrfs_reserve_extent(root, cur_alloc_size, cur_alloc_size, fs_info->sectorsize, 0, alloc_hint, &ins, 1, 1); if (ret < 0) goto out_unlock; + cur_alloc_size = ins.offset; + extent_reserved = true; ram_size = ins.offset; em = create_io_em(inode, start, ins.offset, /* len */ @@ -988,7 +1020,6 @@ static noinline int cow_file_range(struct inode *inode, goto out_reserve; free_extent_map(em); - cur_alloc_size = ins.offset; ret = btrfs_add_ordered_extent(inode, start, ins.objectid, ram_size, cur_alloc_size, 0); if (ret) @@ -998,15 +1029,24 @@ static noinline int cow_file_range(struct inode *inode, BTRFS_DATA_RELOC_TREE_OBJECTID) { ret = btrfs_reloc_clone_csums(inode, start, cur_alloc_size); + /* + * Only drop cache here, and process as normal. + * + * We must not allow extent_clear_unlock_delalloc() + * at out_unlock label to free meta of this ordered + * extent, as its meta should be freed by + * btrfs_finish_ordered_io(). + * + * So we must continue until @start is increased to + * skip current ordered extent. + */ if (ret) - goto out_drop_extent_cache; + btrfs_drop_extent_cache(BTRFS_I(inode), start, + start + ram_size - 1, 0); } btrfs_dec_block_group_reservations(fs_info, ins.objectid); - if (disk_num_bytes < cur_alloc_size) - break; - /* we're not doing compressed IO, don't unlock the first * page (which the caller expects to stay locked), don't * clear any dirty bits and don't set any writeback bits @@ -1014,18 +1054,30 @@ static noinline int cow_file_range(struct inode *inode, * Do set the Private2 bit so we know this page was properly * setup for writepage */ - op = unlock ? PAGE_UNLOCK : 0; - op |= PAGE_SET_PRIVATE2; + page_ops = unlock ? PAGE_UNLOCK : 0; + page_ops |= PAGE_SET_PRIVATE2; extent_clear_unlock_delalloc(inode, start, start + ram_size - 1, delalloc_end, locked_page, EXTENT_LOCKED | EXTENT_DELALLOC, - op); - disk_num_bytes -= cur_alloc_size; + page_ops); + if (disk_num_bytes < cur_alloc_size) + disk_num_bytes = 0; + else + disk_num_bytes -= cur_alloc_size; num_bytes -= cur_alloc_size; alloc_hint = ins.objectid + ins.offset; start += cur_alloc_size; + extent_reserved = false; + + /* + * btrfs_reloc_clone_csums() error, since start is increased + * extent_clear_unlock_delalloc() at out_unlock label won't + * free metadata of current ordered extent, we're OK to exit. + */ + if (ret) + goto out_unlock; } out: return ret; @@ -1036,12 +1088,35 @@ out_reserve: btrfs_dec_block_group_reservations(fs_info, ins.objectid); btrfs_free_reserved_extent(fs_info, ins.objectid, ins.offset, 1); out_unlock: + clear_bits = EXTENT_LOCKED | EXTENT_DELALLOC | EXTENT_DELALLOC_NEW | + EXTENT_DEFRAG | EXTENT_CLEAR_META_RESV; + page_ops = PAGE_UNLOCK | PAGE_CLEAR_DIRTY | PAGE_SET_WRITEBACK | + PAGE_END_WRITEBACK; + /* + * If we reserved an extent for our delalloc range (or a subrange) and + * failed to create the respective ordered extent, then it means that + * when we reserved the extent we decremented the extent's size from + * the data space_info's bytes_may_use counter and incremented the + * space_info's bytes_reserved counter by the same amount. We must make + * sure extent_clear_unlock_delalloc() does not try to decrement again + * the data space_info's bytes_may_use counter, therefore we do not pass + * it the flag EXTENT_CLEAR_DATA_RESV. + */ + if (extent_reserved) { + extent_clear_unlock_delalloc(inode, start, + start + cur_alloc_size, + start + cur_alloc_size, + locked_page, + clear_bits, + page_ops); + start += cur_alloc_size; + if (start >= end) + goto out; + } extent_clear_unlock_delalloc(inode, start, end, delalloc_end, locked_page, - EXTENT_LOCKED | EXTENT_DO_ACCOUNTING | - EXTENT_DELALLOC | EXTENT_DEFRAG, - PAGE_UNLOCK | PAGE_CLEAR_DIRTY | - PAGE_SET_WRITEBACK | PAGE_END_WRITEBACK); + clear_bits | EXTENT_CLEAR_DATA_RESV, + page_ops); goto out; } @@ -1414,15 +1489,14 @@ out_check: BUG_ON(ret); /* -ENOMEM */ if (root->root_key.objectid == - BTRFS_DATA_RELOC_TREE_OBJECTID) { + BTRFS_DATA_RELOC_TREE_OBJECTID) + /* + * Error handled later, as we must prevent + * extent_clear_unlock_delalloc() in error handler + * from freeing metadata of created ordered extent. + */ ret = btrfs_reloc_clone_csums(inode, cur_offset, num_bytes); - if (ret) { - if (!nolock && nocow) - btrfs_end_write_no_snapshoting(root); - goto error; - } - } extent_clear_unlock_delalloc(inode, cur_offset, cur_offset + num_bytes - 1, end, @@ -1434,6 +1508,14 @@ out_check: if (!nolock && nocow) btrfs_end_write_no_snapshoting(root); cur_offset = extent_end; + + /* + * btrfs_reloc_clone_csums() error, now we're OK to call error + * handler, as metadata for created ordered extent will only + * be freed by btrfs_finish_ordered_io(). + */ + if (ret) + goto error; if (cur_offset > end) break; } @@ -1509,6 +1591,8 @@ static int run_delalloc_range(struct inode *inode, struct page *locked_page, ret = cow_file_range_async(inode, locked_page, start, end, page_started, nr_written); } + if (ret) + btrfs_cleanup_ordered_extents(inode, start, end - start + 1); return ret; } @@ -1693,6 +1777,14 @@ static void btrfs_set_bit_hook(struct inode *inode, btrfs_add_delalloc_inodes(root, inode); spin_unlock(&BTRFS_I(inode)->lock); } + + if (!(state->state & EXTENT_DELALLOC_NEW) && + (*bits & EXTENT_DELALLOC_NEW)) { + spin_lock(&BTRFS_I(inode)->lock); + BTRFS_I(inode)->new_delalloc_bytes += state->end + 1 - + state->start; + spin_unlock(&BTRFS_I(inode)->lock); + } } /* @@ -1722,7 +1814,7 @@ static void btrfs_clear_bit_hook(struct btrfs_inode *inode, if (*bits & EXTENT_FIRST_DELALLOC) { *bits &= ~EXTENT_FIRST_DELALLOC; - } else if (!(*bits & EXTENT_DO_ACCOUNTING)) { + } else if (!(*bits & EXTENT_CLEAR_META_RESV)) { spin_lock(&inode->lock); inode->outstanding_extents -= num_extents; spin_unlock(&inode->lock); @@ -1733,7 +1825,7 @@ static void btrfs_clear_bit_hook(struct btrfs_inode *inode, * don't need to call dellalloc_release_metadata if there is an * error. */ - if (*bits & EXTENT_DO_ACCOUNTING && + if (*bits & EXTENT_CLEAR_META_RESV && root != fs_info->tree_root) btrfs_delalloc_release_metadata(inode, len); @@ -1741,10 +1833,9 @@ static void btrfs_clear_bit_hook(struct btrfs_inode *inode, if (btrfs_is_testing(fs_info)) return; - if (root->root_key.objectid != BTRFS_DATA_RELOC_TREE_OBJECTID - && do_list && !(state->state & EXTENT_NORESERVE) - && (*bits & (EXTENT_DO_ACCOUNTING | - EXTENT_CLEAR_DATA_RESV))) + if (root->root_key.objectid != BTRFS_DATA_RELOC_TREE_OBJECTID && + do_list && !(state->state & EXTENT_NORESERVE) && + (*bits & EXTENT_CLEAR_DATA_RESV)) btrfs_free_reserved_data_space_noquota( &inode->vfs_inode, state->start, len); @@ -1759,6 +1850,14 @@ static void btrfs_clear_bit_hook(struct btrfs_inode *inode, btrfs_del_delalloc_inode(root, inode); spin_unlock(&inode->lock); } + + if ((state->state & EXTENT_DELALLOC_NEW) && + (*bits & EXTENT_DELALLOC_NEW)) { + spin_lock(&inode->lock); + ASSERT(inode->new_delalloc_bytes >= len); + inode->new_delalloc_bytes -= len; + spin_unlock(&inode->lock); + } } /* @@ -2791,6 +2890,13 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent) u64 logical_len = ordered_extent->len; bool nolock; bool truncated = false; + bool range_locked = false; + bool clear_new_delalloc_bytes = false; + + if (!test_bit(BTRFS_ORDERED_NOCOW, &ordered_extent->flags) && + !test_bit(BTRFS_ORDERED_PREALLOC, &ordered_extent->flags) && + !test_bit(BTRFS_ORDERED_DIRECT, &ordered_extent->flags)) + clear_new_delalloc_bytes = true; nolock = btrfs_is_free_space_inode(BTRFS_I(inode)); @@ -2839,6 +2945,7 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent) goto out; } + range_locked = true; lock_extent_bits(io_tree, ordered_extent->file_offset, ordered_extent->file_offset + ordered_extent->len - 1, &cached_state); @@ -2864,7 +2971,7 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent) if (IS_ERR(trans)) { ret = PTR_ERR(trans); trans = NULL; - goto out_unlock; + goto out; } trans->block_rsv = &fs_info->delalloc_block_rsv; @@ -2896,7 +3003,7 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent) trans->transid); if (ret < 0) { btrfs_abort_transaction(trans, ret); - goto out_unlock; + goto out; } add_pending_csums(trans, inode, &ordered_extent->list); @@ -2905,14 +3012,26 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent) ret = btrfs_update_inode_fallback(trans, root, inode); if (ret) { /* -ENOMEM or corruption */ btrfs_abort_transaction(trans, ret); - goto out_unlock; + goto out; } ret = 0; -out_unlock: - unlock_extent_cached(io_tree, ordered_extent->file_offset, - ordered_extent->file_offset + - ordered_extent->len - 1, &cached_state, GFP_NOFS); out: + if (range_locked || clear_new_delalloc_bytes) { + unsigned int clear_bits = 0; + + if (range_locked) + clear_bits |= EXTENT_LOCKED; + if (clear_new_delalloc_bytes) + clear_bits |= EXTENT_DELALLOC_NEW; + clear_extent_bit(&BTRFS_I(inode)->io_tree, + ordered_extent->file_offset, + ordered_extent->file_offset + + ordered_extent->len - 1, + clear_bits, + (clear_bits & EXTENT_LOCKED) ? 1 : 0, + 0, &cached_state, GFP_NOFS); + } + if (root != fs_info->tree_root) btrfs_delalloc_release_metadata(BTRFS_I(inode), ordered_extent->len); @@ -4401,9 +4520,17 @@ search_again: if (extent_type != BTRFS_FILE_EXTENT_INLINE) { item_end += btrfs_file_extent_num_bytes(leaf, fi); + + trace_btrfs_truncate_show_fi_regular( + BTRFS_I(inode), leaf, fi, + found_key.offset); } else if (extent_type == BTRFS_FILE_EXTENT_INLINE) { item_end += btrfs_file_extent_inline_len(leaf, path->slots[0], fi); + + trace_btrfs_truncate_show_fi_inline( + BTRFS_I(inode), leaf, fi, path->slots[0], + found_key.offset); } item_end--; } @@ -4603,13 +4730,6 @@ error: btrfs_free_path(path); - if (err == 0) { - /* only inline file may have last_size != new_size */ - if (new_size >= fs_info->sectorsize || - new_size > fs_info->max_inline) - ASSERT(last_size == new_size); - } - if (be_nice && bytes_deleted > SZ_32M) { unsigned long updates = trans->delayed_ref_updates; if (updates) { @@ -6735,7 +6855,6 @@ static noinline int uncompress_inline(struct btrfs_path *path, * * This also copies inline extents directly into the page. */ - struct extent_map *btrfs_get_extent(struct btrfs_inode *inode, struct page *page, size_t pg_offset, u64 start, u64 len, @@ -6835,11 +6954,18 @@ again: found_type == BTRFS_FILE_EXTENT_PREALLOC) { extent_end = extent_start + btrfs_file_extent_num_bytes(leaf, item); + + trace_btrfs_get_extent_show_fi_regular(inode, leaf, item, + extent_start); } else if (found_type == BTRFS_FILE_EXTENT_INLINE) { size_t size; size = btrfs_file_extent_inline_len(leaf, path->slots[0], item); extent_end = ALIGN(extent_start + size, fs_info->sectorsize); + + trace_btrfs_get_extent_show_fi_inline(inode, leaf, item, + path->slots[0], + extent_start); } next: if (start >= extent_end) { @@ -7037,19 +7163,17 @@ struct extent_map *btrfs_get_extent_fiemap(struct btrfs_inode *inode, em = btrfs_get_extent(inode, page, pg_offset, start, len, create); if (IS_ERR(em)) return em; - if (em) { - /* - * if our em maps to - * - a hole or - * - a pre-alloc extent, - * there might actually be delalloc bytes behind it. - */ - if (em->block_start != EXTENT_MAP_HOLE && - !test_bit(EXTENT_FLAG_PREALLOC, &em->flags)) - return em; - else - hole_em = em; - } + /* + * If our em maps to: + * - a hole or + * - a pre-alloc extent, + * there might actually be delalloc bytes behind it. + */ + if (em->block_start != EXTENT_MAP_HOLE && + !test_bit(EXTENT_FLAG_PREALLOC, &em->flags)) + return em; + else + hole_em = em; /* check to see if we've wrapped (len == -1 or similar) */ end = start + len; @@ -8127,17 +8251,26 @@ static void btrfs_endio_direct_read(struct bio *bio) bio_put(bio); } -static void btrfs_endio_direct_write_update_ordered(struct inode *inode, - const u64 offset, - const u64 bytes, - const int uptodate) +static void __endio_write_update_ordered(struct inode *inode, + const u64 offset, const u64 bytes, + const bool uptodate) { struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb); struct btrfs_ordered_extent *ordered = NULL; + struct btrfs_workqueue *wq; + btrfs_work_func_t func; u64 ordered_offset = offset; u64 ordered_bytes = bytes; int ret; + if (btrfs_is_free_space_inode(BTRFS_I(inode))) { + wq = fs_info->endio_freespace_worker; + func = btrfs_freespace_write_helper; + } else { + wq = fs_info->endio_write_workers; + func = btrfs_endio_write_helper; + } + again: ret = btrfs_dec_test_first_ordered_pending(inode, &ordered, &ordered_offset, @@ -8146,9 +8279,8 @@ again: if (!ret) goto out_test; - btrfs_init_work(&ordered->work, btrfs_endio_write_helper, - finish_ordered_fn, NULL, NULL); - btrfs_queue_work(fs_info->endio_write_workers, &ordered->work); + btrfs_init_work(&ordered->work, func, finish_ordered_fn, NULL, NULL); + btrfs_queue_work(wq, &ordered->work); out_test: /* * our bio might span multiple ordered extents. If we haven't @@ -8166,10 +8298,8 @@ static void btrfs_endio_direct_write(struct bio *bio) struct btrfs_dio_private *dip = bio->bi_private; struct bio *dio_bio = dip->dio_bio; - btrfs_endio_direct_write_update_ordered(dip->inode, - dip->logical_offset, - dip->bytes, - !bio->bi_error); + __endio_write_update_ordered(dip->inode, dip->logical_offset, + dip->bytes, !bio->bi_error); kfree(dip); @@ -8530,10 +8660,10 @@ free_ordered: io_bio = NULL; } else { if (write) - btrfs_endio_direct_write_update_ordered(inode, + __endio_write_update_ordered(inode, file_offset, dio_bio->bi_iter.bi_size, - 0); + false); else unlock_extent(&BTRFS_I(inode)->io_tree, file_offset, file_offset + dio_bio->bi_iter.bi_size - 1); @@ -8668,11 +8798,11 @@ static ssize_t btrfs_direct_IO(struct kiocb *iocb, struct iov_iter *iter) */ if (dio_data.unsubmitted_oe_range_start < dio_data.unsubmitted_oe_range_end) - btrfs_endio_direct_write_update_ordered(inode, + __endio_write_update_ordered(inode, dio_data.unsubmitted_oe_range_start, dio_data.unsubmitted_oe_range_end - dio_data.unsubmitted_oe_range_start, - 0); + false); } else if (ret >= 0 && (size_t)ret < count) btrfs_delalloc_release_space(inode, offset, count - (size_t)ret); @@ -8819,6 +8949,7 @@ again: if (!inode_evicting) clear_extent_bit(tree, start, end, EXTENT_DIRTY | EXTENT_DELALLOC | + EXTENT_DELALLOC_NEW | EXTENT_LOCKED | EXTENT_DO_ACCOUNTING | EXTENT_DEFRAG, 1, 0, &cached_state, GFP_NOFS); @@ -8876,8 +9007,8 @@ again: if (!inode_evicting) { clear_extent_bit(tree, page_start, page_end, EXTENT_LOCKED | EXTENT_DIRTY | - EXTENT_DELALLOC | EXTENT_DO_ACCOUNTING | - EXTENT_DEFRAG, 1, 1, + EXTENT_DELALLOC | EXTENT_DELALLOC_NEW | + EXTENT_DO_ACCOUNTING | EXTENT_DEFRAG, 1, 1, &cached_state, GFP_NOFS); __btrfs_releasepage(page, GFP_NOFS); @@ -9248,6 +9379,7 @@ struct inode *btrfs_alloc_inode(struct super_block *sb) ei->last_sub_trans = 0; ei->logged_trans = 0; ei->delalloc_bytes = 0; + ei->new_delalloc_bytes = 0; ei->defrag_bytes = 0; ei->disk_i_size = 0; ei->flags = 0; @@ -9313,6 +9445,7 @@ void btrfs_destroy_inode(struct inode *inode) WARN_ON(BTRFS_I(inode)->outstanding_extents); WARN_ON(BTRFS_I(inode)->reserved_extents); WARN_ON(BTRFS_I(inode)->delalloc_bytes); + WARN_ON(BTRFS_I(inode)->new_delalloc_bytes); WARN_ON(BTRFS_I(inode)->csum_bytes); WARN_ON(BTRFS_I(inode)->defrag_bytes); @@ -9436,7 +9569,7 @@ static int btrfs_getattr(const struct path *path, struct kstat *stat, stat->dev = BTRFS_I(inode)->root->anon_dev; spin_lock(&BTRFS_I(inode)->lock); - delalloc_bytes = BTRFS_I(inode)->delalloc_bytes; + delalloc_bytes = BTRFS_I(inode)->new_delalloc_bytes; spin_unlock(&BTRFS_I(inode)->lock); stat->blocks = (ALIGN(inode_get_bytes(inode), blocksize) + ALIGN(delalloc_bytes, blocksize)) >> 9; diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index 922a66fce401..e176375f374f 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -1504,7 +1504,7 @@ static noinline int btrfs_ioctl_resize(struct file *file, if (ret) return ret; - if (atomic_xchg(&fs_info->mutually_exclusive_operation_running, 1)) { + if (test_and_set_bit(BTRFS_FS_EXCL_OP, &fs_info->flags)) { mnt_drop_write_file(file); return BTRFS_ERROR_DEV_EXCL_RUN_IN_PROGRESS; } @@ -1619,7 +1619,7 @@ out_free: kfree(vol_args); out: mutex_unlock(&fs_info->volume_mutex); - atomic_set(&fs_info->mutually_exclusive_operation_running, 0); + clear_bit(BTRFS_FS_EXCL_OP, &fs_info->flags); mnt_drop_write_file(file); return ret; } @@ -2661,7 +2661,7 @@ static long btrfs_ioctl_add_dev(struct btrfs_fs_info *fs_info, void __user *arg) if (!capable(CAP_SYS_ADMIN)) return -EPERM; - if (atomic_xchg(&fs_info->mutually_exclusive_operation_running, 1)) + if (test_and_set_bit(BTRFS_FS_EXCL_OP, &fs_info->flags)) return BTRFS_ERROR_DEV_EXCL_RUN_IN_PROGRESS; mutex_lock(&fs_info->volume_mutex); @@ -2680,7 +2680,7 @@ static long btrfs_ioctl_add_dev(struct btrfs_fs_info *fs_info, void __user *arg) kfree(vol_args); out: mutex_unlock(&fs_info->volume_mutex); - atomic_set(&fs_info->mutually_exclusive_operation_running, 0); + clear_bit(BTRFS_FS_EXCL_OP, &fs_info->flags); return ret; } @@ -2708,7 +2708,7 @@ static long btrfs_ioctl_rm_dev_v2(struct file *file, void __user *arg) if (vol_args->flags & ~BTRFS_VOL_ARG_V2_FLAGS_SUPPORTED) return -EOPNOTSUPP; - if (atomic_xchg(&fs_info->mutually_exclusive_operation_running, 1)) { + if (test_and_set_bit(BTRFS_FS_EXCL_OP, &fs_info->flags)) { ret = BTRFS_ERROR_DEV_EXCL_RUN_IN_PROGRESS; goto out; } @@ -2721,7 +2721,7 @@ static long btrfs_ioctl_rm_dev_v2(struct file *file, void __user *arg) ret = btrfs_rm_device(fs_info, vol_args->name, 0); } mutex_unlock(&fs_info->volume_mutex); - atomic_set(&fs_info->mutually_exclusive_operation_running, 0); + clear_bit(BTRFS_FS_EXCL_OP, &fs_info->flags); if (!ret) { if (vol_args->flags & BTRFS_DEVICE_SPEC_BY_ID) @@ -2752,7 +2752,7 @@ static long btrfs_ioctl_rm_dev(struct file *file, void __user *arg) if (ret) return ret; - if (atomic_xchg(&fs_info->mutually_exclusive_operation_running, 1)) { + if (test_and_set_bit(BTRFS_FS_EXCL_OP, &fs_info->flags)) { ret = BTRFS_ERROR_DEV_EXCL_RUN_IN_PROGRESS; goto out_drop_write; } @@ -2772,7 +2772,7 @@ static long btrfs_ioctl_rm_dev(struct file *file, void __user *arg) btrfs_info(fs_info, "disk deleted %s", vol_args->name); kfree(vol_args); out: - atomic_set(&fs_info->mutually_exclusive_operation_running, 0); + clear_bit(BTRFS_FS_EXCL_OP, &fs_info->flags); out_drop_write: mnt_drop_write_file(file); @@ -4439,13 +4439,11 @@ static long btrfs_ioctl_dev_replace(struct btrfs_fs_info *fs_info, ret = -EROFS; goto out; } - if (atomic_xchg( - &fs_info->mutually_exclusive_operation_running, 1)) { + if (test_and_set_bit(BTRFS_FS_EXCL_OP, &fs_info->flags)) { ret = BTRFS_ERROR_DEV_EXCL_RUN_IN_PROGRESS; } else { ret = btrfs_dev_replace_by_ioctl(fs_info, p); - atomic_set( - &fs_info->mutually_exclusive_operation_running, 0); + clear_bit(BTRFS_FS_EXCL_OP, &fs_info->flags); } break; case BTRFS_IOCTL_DEV_REPLACE_CMD_STATUS: @@ -4640,7 +4638,7 @@ static long btrfs_ioctl_balance(struct file *file, void __user *arg) return ret; again: - if (!atomic_xchg(&fs_info->mutually_exclusive_operation_running, 1)) { + if (!test_and_set_bit(BTRFS_FS_EXCL_OP, &fs_info->flags)) { mutex_lock(&fs_info->volume_mutex); mutex_lock(&fs_info->balance_mutex); need_unlock = true; @@ -4686,7 +4684,7 @@ again: } locked: - BUG_ON(!atomic_read(&fs_info->mutually_exclusive_operation_running)); + BUG_ON(!test_bit(BTRFS_FS_EXCL_OP, &fs_info->flags)); if (arg) { bargs = memdup_user(arg, sizeof(*bargs)); @@ -4742,11 +4740,10 @@ locked: do_balance: /* - * Ownership of bctl and mutually_exclusive_operation_running + * Ownership of bctl and filesystem flag BTRFS_FS_EXCL_OP * goes to to btrfs_balance. bctl is freed in __cancel_balance, * or, if restriper was paused all the way until unmount, in - * free_fs_info. mutually_exclusive_operation_running is - * cleared in __cancel_balance. + * free_fs_info. The flag is cleared in __cancel_balance. */ need_unlock = false; @@ -4766,7 +4763,7 @@ out_unlock: mutex_unlock(&fs_info->balance_mutex); mutex_unlock(&fs_info->volume_mutex); if (need_unlock) - atomic_set(&fs_info->mutually_exclusive_operation_running, 0); + clear_bit(BTRFS_FS_EXCL_OP, &fs_info->flags); out: mnt_drop_write_file(file); return ret; diff --git a/fs/btrfs/ordered-data.c b/fs/btrfs/ordered-data.c index 9a46878ba60f..7b40e2e7292a 100644 --- a/fs/btrfs/ordered-data.c +++ b/fs/btrfs/ordered-data.c @@ -212,7 +212,7 @@ static int __btrfs_add_ordered_extent(struct inode *inode, u64 file_offset, set_bit(BTRFS_ORDERED_DIRECT, &entry->flags); /* one ref for the tree */ - atomic_set(&entry->refs, 1); + refcount_set(&entry->refs, 1); init_waitqueue_head(&entry->wait); INIT_LIST_HEAD(&entry->list); INIT_LIST_HEAD(&entry->root_extent_list); @@ -358,7 +358,7 @@ int btrfs_dec_test_first_ordered_pending(struct inode *inode, out: if (!ret && cached && entry) { *cached = entry; - atomic_inc(&entry->refs); + refcount_inc(&entry->refs); } spin_unlock_irqrestore(&tree->lock, flags); return ret == 0; @@ -425,7 +425,7 @@ have_entry: out: if (!ret && cached && entry) { *cached = entry; - atomic_inc(&entry->refs); + refcount_inc(&entry->refs); } spin_unlock_irqrestore(&tree->lock, flags); return ret == 0; @@ -456,7 +456,7 @@ void btrfs_get_logged_extents(struct btrfs_inode *inode, if (test_and_set_bit(BTRFS_ORDERED_LOGGED, &ordered->flags)) continue; list_add(&ordered->log_list, logged_list); - atomic_inc(&ordered->refs); + refcount_inc(&ordered->refs); } spin_unlock_irq(&tree->lock); } @@ -565,7 +565,7 @@ void btrfs_put_ordered_extent(struct btrfs_ordered_extent *entry) trace_btrfs_ordered_extent_put(entry->inode, entry); - if (atomic_dec_and_test(&entry->refs)) { + if (refcount_dec_and_test(&entry->refs)) { ASSERT(list_empty(&entry->log_list)); ASSERT(list_empty(&entry->trans_list)); ASSERT(list_empty(&entry->root_extent_list)); @@ -623,7 +623,7 @@ void btrfs_remove_ordered_extent(struct inode *inode, spin_lock(&fs_info->trans_lock); trans = fs_info->running_transaction; if (trans) - atomic_inc(&trans->use_count); + refcount_inc(&trans->use_count); spin_unlock(&fs_info->trans_lock); ASSERT(trans); @@ -690,7 +690,7 @@ int btrfs_wait_ordered_extents(struct btrfs_root *root, int nr, list_move_tail(&ordered->root_extent_list, &root->ordered_extents); - atomic_inc(&ordered->refs); + refcount_inc(&ordered->refs); spin_unlock(&root->ordered_extent_lock); btrfs_init_work(&ordered->flush_work, @@ -870,7 +870,7 @@ struct btrfs_ordered_extent *btrfs_lookup_ordered_extent(struct inode *inode, if (!offset_in_entry(entry, file_offset)) entry = NULL; if (entry) - atomic_inc(&entry->refs); + refcount_inc(&entry->refs); out: spin_unlock_irq(&tree->lock); return entry; @@ -911,7 +911,7 @@ struct btrfs_ordered_extent *btrfs_lookup_ordered_range( } out: if (entry) - atomic_inc(&entry->refs); + refcount_inc(&entry->refs); spin_unlock_irq(&tree->lock); return entry; } @@ -948,7 +948,7 @@ btrfs_lookup_first_ordered_extent(struct inode *inode, u64 file_offset) goto out; entry = rb_entry(node, struct btrfs_ordered_extent, rb_node); - atomic_inc(&entry->refs); + refcount_inc(&entry->refs); out: spin_unlock_irq(&tree->lock); return entry; diff --git a/fs/btrfs/ordered-data.h b/fs/btrfs/ordered-data.h index 195c93b67fe0..e0c1d5b8d859 100644 --- a/fs/btrfs/ordered-data.h +++ b/fs/btrfs/ordered-data.h @@ -113,7 +113,7 @@ struct btrfs_ordered_extent { int compress_type; /* reference count */ - atomic_t refs; + refcount_t refs; /* the inode we belong to */ struct inode *inode; diff --git a/fs/btrfs/qgroup.c b/fs/btrfs/qgroup.c index afbea61d957e..deffbeb74a0b 100644 --- a/fs/btrfs/qgroup.c +++ b/fs/btrfs/qgroup.c @@ -47,50 +47,6 @@ * - check all ioctl parameters */ -/* - * one struct for each qgroup, organized in fs_info->qgroup_tree. - */ -struct btrfs_qgroup { - u64 qgroupid; - - /* - * state - */ - u64 rfer; /* referenced */ - u64 rfer_cmpr; /* referenced compressed */ - u64 excl; /* exclusive */ - u64 excl_cmpr; /* exclusive compressed */ - - /* - * limits - */ - u64 lim_flags; /* which limits are set */ - u64 max_rfer; - u64 max_excl; - u64 rsv_rfer; - u64 rsv_excl; - - /* - * reservation tracking - */ - u64 reserved; - - /* - * lists - */ - struct list_head groups; /* groups this group is member of */ - struct list_head members; /* groups that are members of this group */ - struct list_head dirty; /* dirty groups */ - struct rb_node node; /* tree of qgroups */ - - /* - * temp variables for accounting operations - * Refer to qgroup_shared_accounting() for details. - */ - u64 old_refcnt; - u64 new_refcnt; -}; - static void btrfs_qgroup_update_old_refcnt(struct btrfs_qgroup *qg, u64 seq, int mod) { @@ -1078,6 +1034,7 @@ static int __qgroup_excl_accounting(struct btrfs_fs_info *fs_info, qgroup->excl += sign * num_bytes; qgroup->excl_cmpr += sign * num_bytes; if (sign > 0) { + trace_qgroup_update_reserve(fs_info, qgroup, -(s64)num_bytes); if (qgroup->reserved < num_bytes) report_reserved_underflow(fs_info, qgroup, num_bytes); else @@ -1103,6 +1060,8 @@ static int __qgroup_excl_accounting(struct btrfs_fs_info *fs_info, WARN_ON(sign < 0 && qgroup->excl < num_bytes); qgroup->excl += sign * num_bytes; if (sign > 0) { + trace_qgroup_update_reserve(fs_info, qgroup, + -(s64)num_bytes); if (qgroup->reserved < num_bytes) report_reserved_underflow(fs_info, qgroup, num_bytes); @@ -2058,12 +2017,12 @@ int btrfs_qgroup_account_extents(struct btrfs_trans_handle *trans, if (!ret) { /* - * Use (u64)-1 as time_seq to do special search, which + * Use SEQ_LAST as time_seq to do special search, which * doesn't lock tree or delayed_refs and search current * root. It's safe inside commit_transaction(). */ ret = btrfs_find_all_roots(trans, fs_info, - record->bytenr, (u64)-1, &new_roots); + record->bytenr, SEQ_LAST, &new_roots); if (ret < 0) goto cleanup; if (qgroup_to_skip) @@ -2370,6 +2329,7 @@ static int qgroup_reserve(struct btrfs_root *root, u64 num_bytes, bool enforce) struct btrfs_fs_info *fs_info = root->fs_info; u64 ref_root = root->root_key.objectid; int ret = 0; + int retried = 0; struct ulist_node *unode; struct ulist_iterator uiter; @@ -2378,7 +2338,7 @@ static int qgroup_reserve(struct btrfs_root *root, u64 num_bytes, bool enforce) if (num_bytes == 0) return 0; - +retry: spin_lock(&fs_info->qgroup_lock); quota_root = fs_info->quota_root; if (!quota_root) @@ -2405,6 +2365,27 @@ static int qgroup_reserve(struct btrfs_root *root, u64 num_bytes, bool enforce) qg = unode_aux_to_qgroup(unode); if (enforce && !qgroup_check_limits(qg, num_bytes)) { + /* + * Commit the tree and retry, since we may have + * deletions which would free up space. + */ + if (!retried && qg->reserved > 0) { + struct btrfs_trans_handle *trans; + + spin_unlock(&fs_info->qgroup_lock); + ret = btrfs_start_delalloc_inodes(root, 0); + if (ret) + return ret; + btrfs_wait_ordered_extents(root, -1, 0, (u64)-1); + trans = btrfs_join_transaction(root); + if (IS_ERR(trans)) + return PTR_ERR(trans); + ret = btrfs_commit_transaction(trans); + if (ret) + return ret; + retried++; + goto retry; + } ret = -EDQUOT; goto out; } @@ -2427,6 +2408,7 @@ static int qgroup_reserve(struct btrfs_root *root, u64 num_bytes, bool enforce) qg = unode_aux_to_qgroup(unode); + trace_qgroup_update_reserve(fs_info, qg, num_bytes); qg->reserved += num_bytes; } @@ -2472,6 +2454,7 @@ void btrfs_qgroup_free_refroot(struct btrfs_fs_info *fs_info, qg = unode_aux_to_qgroup(unode); + trace_qgroup_update_reserve(fs_info, qg, -(s64)num_bytes); if (qg->reserved < num_bytes) report_reserved_underflow(fs_info, qg, num_bytes); else @@ -2490,18 +2473,6 @@ out: spin_unlock(&fs_info->qgroup_lock); } -void assert_qgroups_uptodate(struct btrfs_trans_handle *trans) -{ - if (list_empty(&trans->qgroup_ref_list) && !trans->delayed_ref_elem.seq) - return; - btrfs_err(trans->fs_info, - "qgroups not uptodate in trans handle %p: list is%s empty, seq is %#x.%x", - trans, list_empty(&trans->qgroup_ref_list) ? "" : " not", - (u32)(trans->delayed_ref_elem.seq >> 32), - (u32)trans->delayed_ref_elem.seq); - BUG(); -} - /* * returns < 0 on error, 0 when more leafs are to be scanned. * returns 1 when done. @@ -2889,14 +2860,14 @@ static int __btrfs_qgroup_release_data(struct inode *inode, u64 start, u64 len, if (ret < 0) goto out; - if (free) { - btrfs_qgroup_free_refroot(BTRFS_I(inode)->root->fs_info, - BTRFS_I(inode)->root->objectid, - changeset.bytes_changed); + if (free) trace_op = QGROUP_FREE; - } trace_btrfs_qgroup_release_data(inode, start, len, changeset.bytes_changed, trace_op); + if (free) + btrfs_qgroup_free_refroot(BTRFS_I(inode)->root->fs_info, + BTRFS_I(inode)->root->objectid, + changeset.bytes_changed); out: ulist_release(&changeset.range_changed); return ret; @@ -2948,6 +2919,7 @@ int btrfs_qgroup_reserve_meta(struct btrfs_root *root, int num_bytes, return 0; BUG_ON(num_bytes != round_down(num_bytes, fs_info->nodesize)); + trace_qgroup_meta_reserve(root, (s64)num_bytes); ret = qgroup_reserve(root, num_bytes, enforce); if (ret < 0) return ret; @@ -2967,6 +2939,7 @@ void btrfs_qgroup_free_meta_all(struct btrfs_root *root) reserved = atomic64_xchg(&root->qgroup_meta_rsv, 0); if (reserved == 0) return; + trace_qgroup_meta_reserve(root, -(s64)reserved); btrfs_qgroup_free_refroot(fs_info, root->objectid, reserved); } @@ -2981,6 +2954,7 @@ void btrfs_qgroup_free_meta(struct btrfs_root *root, int num_bytes) BUG_ON(num_bytes != round_down(num_bytes, fs_info->nodesize)); WARN_ON(atomic64_read(&root->qgroup_meta_rsv) < num_bytes); atomic64_sub(num_bytes, &root->qgroup_meta_rsv); + trace_qgroup_meta_reserve(root, -(s64)num_bytes); btrfs_qgroup_free_refroot(fs_info, root->objectid, num_bytes); } diff --git a/fs/btrfs/qgroup.h b/fs/btrfs/qgroup.h index 26932a8a1993..fe04d3f295c6 100644 --- a/fs/btrfs/qgroup.h +++ b/fs/btrfs/qgroup.h @@ -62,6 +62,50 @@ struct btrfs_qgroup_extent_record { }; /* + * one struct for each qgroup, organized in fs_info->qgroup_tree. + */ +struct btrfs_qgroup { + u64 qgroupid; + + /* + * state + */ + u64 rfer; /* referenced */ + u64 rfer_cmpr; /* referenced compressed */ + u64 excl; /* exclusive */ + u64 excl_cmpr; /* exclusive compressed */ + + /* + * limits + */ + u64 lim_flags; /* which limits are set */ + u64 max_rfer; + u64 max_excl; + u64 rsv_rfer; + u64 rsv_excl; + + /* + * reservation tracking + */ + u64 reserved; + + /* + * lists + */ + struct list_head groups; /* groups this group is member of */ + struct list_head members; /* groups that are members of this group */ + struct list_head dirty; /* dirty groups */ + struct rb_node node; /* tree of qgroups */ + + /* + * temp variables for accounting operations + * Refer to qgroup_shared_accounting() for details. + */ + u64 old_refcnt; + u64 new_refcnt; +}; + +/* * For qgroup event trace points only */ #define QGROUP_RESERVE (1<<0) @@ -186,17 +230,12 @@ int btrfs_qgroup_inherit(struct btrfs_trans_handle *trans, struct btrfs_qgroup_inherit *inherit); void btrfs_qgroup_free_refroot(struct btrfs_fs_info *fs_info, u64 ref_root, u64 num_bytes); -/* - * TODO: Add proper trace point for it, as btrfs_qgroup_free() is - * called by everywhere, can't provide good trace for delayed ref case. - */ static inline void btrfs_qgroup_free_delayed_ref(struct btrfs_fs_info *fs_info, u64 ref_root, u64 num_bytes) { - btrfs_qgroup_free_refroot(fs_info, ref_root, num_bytes); trace_btrfs_qgroup_free_delayed_ref(fs_info, ref_root, num_bytes); + btrfs_qgroup_free_refroot(fs_info, ref_root, num_bytes); } -void assert_qgroups_uptodate(struct btrfs_trans_handle *trans); #ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS int btrfs_verify_qgroup_counts(struct btrfs_fs_info *fs_info, u64 qgroupid, diff --git a/fs/btrfs/raid56.c b/fs/btrfs/raid56.c index 1571bf26dc07..d8ea0eb76325 100644 --- a/fs/btrfs/raid56.c +++ b/fs/btrfs/raid56.c @@ -149,7 +149,7 @@ struct btrfs_raid_bio { int generic_bio_cnt; - atomic_t refs; + refcount_t refs; atomic_t stripes_pending; @@ -389,7 +389,7 @@ static void __remove_rbio_from_cache(struct btrfs_raid_bio *rbio) if (bio_list_empty(&rbio->bio_list)) { if (!list_empty(&rbio->hash_list)) { list_del_init(&rbio->hash_list); - atomic_dec(&rbio->refs); + refcount_dec(&rbio->refs); BUG_ON(!list_empty(&rbio->plug_list)); } } @@ -480,7 +480,7 @@ static void cache_rbio(struct btrfs_raid_bio *rbio) /* bump our ref if we were not in the list before */ if (!test_and_set_bit(RBIO_CACHE_BIT, &rbio->flags)) - atomic_inc(&rbio->refs); + refcount_inc(&rbio->refs); if (!list_empty(&rbio->stripe_cache)){ list_move(&rbio->stripe_cache, &table->stripe_cache); @@ -689,7 +689,7 @@ static noinline int lock_stripe_add(struct btrfs_raid_bio *rbio) test_bit(RBIO_CACHE_BIT, &cur->flags) && !test_bit(RBIO_RMW_LOCKED_BIT, &cur->flags)) { list_del_init(&cur->hash_list); - atomic_dec(&cur->refs); + refcount_dec(&cur->refs); steal_rbio(cur, rbio); cache_drop = cur; @@ -738,7 +738,7 @@ static noinline int lock_stripe_add(struct btrfs_raid_bio *rbio) } } lockit: - atomic_inc(&rbio->refs); + refcount_inc(&rbio->refs); list_add(&rbio->hash_list, &h->hash_list); out: spin_unlock_irqrestore(&h->lock, flags); @@ -784,7 +784,7 @@ static noinline void unlock_stripe(struct btrfs_raid_bio *rbio) } list_del_init(&rbio->hash_list); - atomic_dec(&rbio->refs); + refcount_dec(&rbio->refs); /* * we use the plug list to hold all the rbios @@ -801,7 +801,7 @@ static noinline void unlock_stripe(struct btrfs_raid_bio *rbio) list_del_init(&rbio->plug_list); list_add(&next->hash_list, &h->hash_list); - atomic_inc(&next->refs); + refcount_inc(&next->refs); spin_unlock(&rbio->bio_list_lock); spin_unlock_irqrestore(&h->lock, flags); @@ -843,8 +843,7 @@ static void __free_raid_bio(struct btrfs_raid_bio *rbio) { int i; - WARN_ON(atomic_read(&rbio->refs) < 0); - if (!atomic_dec_and_test(&rbio->refs)) + if (!refcount_dec_and_test(&rbio->refs)) return; WARN_ON(!list_empty(&rbio->stripe_cache)); @@ -997,7 +996,7 @@ static struct btrfs_raid_bio *alloc_rbio(struct btrfs_fs_info *fs_info, rbio->stripe_npages = stripe_npages; rbio->faila = -1; rbio->failb = -1; - atomic_set(&rbio->refs, 1); + refcount_set(&rbio->refs, 1); atomic_set(&rbio->error, 0); atomic_set(&rbio->stripes_pending, 0); @@ -2118,6 +2117,11 @@ int raid56_parity_recover(struct btrfs_fs_info *fs_info, struct bio *bio, struct btrfs_raid_bio *rbio; int ret; + if (generic_io) { + ASSERT(bbio->mirror_num == mirror_num); + btrfs_io_bio(bio)->mirror_num = mirror_num; + } + rbio = alloc_rbio(fs_info, bbio, stripe_len); if (IS_ERR(rbio)) { if (generic_io) @@ -2194,6 +2198,8 @@ static void read_rebuild_work(struct btrfs_work *work) /* * The following code is used to scrub/replace the parity stripe * + * Caller must have already increased bio_counter for getting @bbio. + * * Note: We need make sure all the pages that add into the scrub/replace * raid bio are correct and not be changed during the scrub/replace. That * is those pages just hold metadata or file data with checksum. @@ -2231,6 +2237,12 @@ raid56_parity_alloc_scrub_rbio(struct btrfs_fs_info *fs_info, struct bio *bio, ASSERT(rbio->stripe_npages == stripe_nsectors); bitmap_copy(rbio->dbitmap, dbitmap, stripe_nsectors); + /* + * We have already increased bio_counter when getting bbio, record it + * so we can free it at rbio_orig_end_io(). + */ + rbio->generic_bio_cnt = 1; + return rbio; } @@ -2673,6 +2685,12 @@ raid56_alloc_missing_rbio(struct btrfs_fs_info *fs_info, struct bio *bio, return NULL; } + /* + * When we get bbio, we have already increased bio_counter, record it + * so we can free it at rbio_orig_end_io() + */ + rbio->generic_bio_cnt = 1; + return rbio; } diff --git a/fs/btrfs/reada.c b/fs/btrfs/reada.c index e88bca87f5d2..a17e775a4a89 100644 --- a/fs/btrfs/reada.c +++ b/fs/btrfs/reada.c @@ -209,9 +209,9 @@ cleanup: return; } -int btree_readahead_hook(struct btrfs_fs_info *fs_info, - struct extent_buffer *eb, int err) +int btree_readahead_hook(struct extent_buffer *eb, int err) { + struct btrfs_fs_info *fs_info = eb->fs_info; int ret = 0; struct reada_extent *re; @@ -235,10 +235,10 @@ start_machine: return ret; } -static struct reada_zone *reada_find_zone(struct btrfs_fs_info *fs_info, - struct btrfs_device *dev, u64 logical, +static struct reada_zone *reada_find_zone(struct btrfs_device *dev, u64 logical, struct btrfs_bio *bbio) { + struct btrfs_fs_info *fs_info = dev->fs_info; int ret; struct reada_zone *zone; struct btrfs_block_group_cache *cache = NULL; @@ -270,6 +270,12 @@ static struct reada_zone *reada_find_zone(struct btrfs_fs_info *fs_info, if (!zone) return NULL; + ret = radix_tree_preload(GFP_KERNEL); + if (ret) { + kfree(zone); + return NULL; + } + zone->start = start; zone->end = end; INIT_LIST_HEAD(&zone->list); @@ -299,6 +305,7 @@ static struct reada_zone *reada_find_zone(struct btrfs_fs_info *fs_info, zone = NULL; } spin_unlock(&fs_info->reada_lock); + radix_tree_preload_end(); return zone; } @@ -313,7 +320,6 @@ static struct reada_extent *reada_find_extent(struct btrfs_fs_info *fs_info, struct btrfs_bio *bbio = NULL; struct btrfs_device *dev; struct btrfs_device *prev_dev; - u32 blocksize; u64 length; int real_stripes; int nzones = 0; @@ -334,7 +340,6 @@ static struct reada_extent *reada_find_extent(struct btrfs_fs_info *fs_info, if (!re) return NULL; - blocksize = fs_info->nodesize; re->logical = logical; re->top = *top; INIT_LIST_HEAD(&re->extctl); @@ -344,10 +349,10 @@ static struct reada_extent *reada_find_extent(struct btrfs_fs_info *fs_info, /* * map block */ - length = blocksize; + length = fs_info->nodesize; ret = btrfs_map_block(fs_info, BTRFS_MAP_GET_READ_MIRRORS, logical, &length, &bbio, 0); - if (ret || !bbio || length < blocksize) + if (ret || !bbio || length < fs_info->nodesize) goto error; if (bbio->num_stripes > BTRFS_MAX_MIRRORS) { @@ -367,7 +372,7 @@ static struct reada_extent *reada_find_extent(struct btrfs_fs_info *fs_info, if (!dev->bdev) continue; - zone = reada_find_zone(fs_info, dev, logical, bbio); + zone = reada_find_zone(dev, logical, bbio); if (!zone) continue; @@ -386,6 +391,10 @@ static struct reada_extent *reada_find_extent(struct btrfs_fs_info *fs_info, goto error; } + ret = radix_tree_preload(GFP_KERNEL); + if (ret) + goto error; + /* insert extent in reada_tree + all per-device trees, all or nothing */ btrfs_dev_replace_lock(&fs_info->dev_replace, 0); spin_lock(&fs_info->reada_lock); @@ -395,13 +404,16 @@ static struct reada_extent *reada_find_extent(struct btrfs_fs_info *fs_info, re_exist->refcnt++; spin_unlock(&fs_info->reada_lock); btrfs_dev_replace_unlock(&fs_info->dev_replace, 0); + radix_tree_preload_end(); goto error; } if (ret) { spin_unlock(&fs_info->reada_lock); btrfs_dev_replace_unlock(&fs_info->dev_replace, 0); + radix_tree_preload_end(); goto error; } + radix_tree_preload_end(); prev_dev = NULL; dev_replace_is_ongoing = btrfs_dev_replace_is_ongoing( &fs_info->dev_replace); @@ -639,9 +651,9 @@ static int reada_pick_zone(struct btrfs_device *dev) return 1; } -static int reada_start_machine_dev(struct btrfs_fs_info *fs_info, - struct btrfs_device *dev) +static int reada_start_machine_dev(struct btrfs_device *dev) { + struct btrfs_fs_info *fs_info = dev->fs_info; struct reada_extent *re = NULL; int mirror_num = 0; struct extent_buffer *eb = NULL; @@ -754,8 +766,7 @@ static void __reada_start_machine(struct btrfs_fs_info *fs_info) list_for_each_entry(device, &fs_devices->devices, dev_list) { if (atomic_read(&device->reada_in_flight) < MAX_IN_FLIGHT) - enqueued += reada_start_machine_dev(fs_info, - device); + enqueued += reada_start_machine_dev(device); } mutex_unlock(&fs_devices->device_list_mutex); total += enqueued; diff --git a/fs/btrfs/root-tree.c b/fs/btrfs/root-tree.c index a08224eab8b4..7d6bc308bf43 100644 --- a/fs/btrfs/root-tree.c +++ b/fs/btrfs/root-tree.c @@ -501,8 +501,9 @@ void btrfs_update_root_times(struct btrfs_trans_handle *trans, struct btrfs_root *root) { struct btrfs_root_item *item = &root->root_item; - struct timespec ct = current_fs_time(root->fs_info->sb); + struct timespec ct; + ktime_get_real_ts(&ct); spin_lock(&root->root_item_lock); btrfs_set_root_ctransid(item, trans->transid); btrfs_set_stack_timespec_sec(&item->ctime, ct.tv_sec); diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c index b0251eb1239f..c7b45eb2403d 100644 --- a/fs/btrfs/scrub.c +++ b/fs/btrfs/scrub.c @@ -64,7 +64,7 @@ struct scrub_ctx; #define SCRUB_MAX_PAGES_PER_BLOCK 16 /* 64k per node/leaf/sector */ struct scrub_recover { - atomic_t refs; + refcount_t refs; struct btrfs_bio *bbio; u64 map_length; }; @@ -112,7 +112,7 @@ struct scrub_block { struct scrub_page *pagev[SCRUB_MAX_PAGES_PER_BLOCK]; int page_count; atomic_t outstanding_pages; - atomic_t refs; /* free mem on transition to zero */ + refcount_t refs; /* free mem on transition to zero */ struct scrub_ctx *sctx; struct scrub_parity *sparity; struct { @@ -140,9 +140,9 @@ struct scrub_parity { int nsectors; - int stripe_len; + u64 stripe_len; - atomic_t refs; + refcount_t refs; struct list_head spages; @@ -202,7 +202,7 @@ struct scrub_ctx { * doesn't free the scrub context before or while the workers are * doing the wakeup() call. */ - atomic_t refs; + refcount_t refs; }; struct scrub_fixup_nodatasum { @@ -240,6 +240,13 @@ struct scrub_warning { struct btrfs_device *dev; }; +struct full_stripe_lock { + struct rb_node node; + u64 logical; + u64 refs; + struct mutex mutex; +}; + static void scrub_pending_bio_inc(struct scrub_ctx *sctx); static void scrub_pending_bio_dec(struct scrub_ctx *sctx); static void scrub_pending_trans_workers_inc(struct scrub_ctx *sctx); @@ -305,7 +312,7 @@ static void scrub_put_ctx(struct scrub_ctx *sctx); static void scrub_pending_bio_inc(struct scrub_ctx *sctx) { - atomic_inc(&sctx->refs); + refcount_inc(&sctx->refs); atomic_inc(&sctx->bios_in_flight); } @@ -349,6 +356,222 @@ static void scrub_blocked_if_needed(struct btrfs_fs_info *fs_info) } /* + * Insert new full stripe lock into full stripe locks tree + * + * Return pointer to existing or newly inserted full_stripe_lock structure if + * everything works well. + * Return ERR_PTR(-ENOMEM) if we failed to allocate memory + * + * NOTE: caller must hold full_stripe_locks_root->lock before calling this + * function + */ +static struct full_stripe_lock *insert_full_stripe_lock( + struct btrfs_full_stripe_locks_tree *locks_root, + u64 fstripe_logical) +{ + struct rb_node **p; + struct rb_node *parent = NULL; + struct full_stripe_lock *entry; + struct full_stripe_lock *ret; + + WARN_ON(!mutex_is_locked(&locks_root->lock)); + + p = &locks_root->root.rb_node; + while (*p) { + parent = *p; + entry = rb_entry(parent, struct full_stripe_lock, node); + if (fstripe_logical < entry->logical) { + p = &(*p)->rb_left; + } else if (fstripe_logical > entry->logical) { + p = &(*p)->rb_right; + } else { + entry->refs++; + return entry; + } + } + + /* Insert new lock */ + ret = kmalloc(sizeof(*ret), GFP_KERNEL); + if (!ret) + return ERR_PTR(-ENOMEM); + ret->logical = fstripe_logical; + ret->refs = 1; + mutex_init(&ret->mutex); + + rb_link_node(&ret->node, parent, p); + rb_insert_color(&ret->node, &locks_root->root); + return ret; +} + +/* + * Search for a full stripe lock of a block group + * + * Return pointer to existing full stripe lock if found + * Return NULL if not found + */ +static struct full_stripe_lock *search_full_stripe_lock( + struct btrfs_full_stripe_locks_tree *locks_root, + u64 fstripe_logical) +{ + struct rb_node *node; + struct full_stripe_lock *entry; + + WARN_ON(!mutex_is_locked(&locks_root->lock)); + + node = locks_root->root.rb_node; + while (node) { + entry = rb_entry(node, struct full_stripe_lock, node); + if (fstripe_logical < entry->logical) + node = node->rb_left; + else if (fstripe_logical > entry->logical) + node = node->rb_right; + else + return entry; + } + return NULL; +} + +/* + * Helper to get full stripe logical from a normal bytenr. + * + * Caller must ensure @cache is a RAID56 block group. + */ +static u64 get_full_stripe_logical(struct btrfs_block_group_cache *cache, + u64 bytenr) +{ + u64 ret; + + /* + * Due to chunk item size limit, full stripe length should not be + * larger than U32_MAX. Just a sanity check here. + */ + WARN_ON_ONCE(cache->full_stripe_len >= U32_MAX); + + /* + * round_down() can only handle power of 2, while RAID56 full + * stripe length can be 64KiB * n, so we need to manually round down. + */ + ret = div64_u64(bytenr - cache->key.objectid, cache->full_stripe_len) * + cache->full_stripe_len + cache->key.objectid; + return ret; +} + +/* + * Lock a full stripe to avoid concurrency of recovery and read + * + * It's only used for profiles with parities (RAID5/6), for other profiles it + * does nothing. + * + * Return 0 if we locked full stripe covering @bytenr, with a mutex held. + * So caller must call unlock_full_stripe() at the same context. + * + * Return <0 if encounters error. + */ +static int lock_full_stripe(struct btrfs_fs_info *fs_info, u64 bytenr, + bool *locked_ret) +{ + struct btrfs_block_group_cache *bg_cache; + struct btrfs_full_stripe_locks_tree *locks_root; + struct full_stripe_lock *existing; + u64 fstripe_start; + int ret = 0; + + *locked_ret = false; + bg_cache = btrfs_lookup_block_group(fs_info, bytenr); + if (!bg_cache) { + ASSERT(0); + return -ENOENT; + } + + /* Profiles not based on parity don't need full stripe lock */ + if (!(bg_cache->flags & BTRFS_BLOCK_GROUP_RAID56_MASK)) + goto out; + locks_root = &bg_cache->full_stripe_locks_root; + + fstripe_start = get_full_stripe_logical(bg_cache, bytenr); + + /* Now insert the full stripe lock */ + mutex_lock(&locks_root->lock); + existing = insert_full_stripe_lock(locks_root, fstripe_start); + mutex_unlock(&locks_root->lock); + if (IS_ERR(existing)) { + ret = PTR_ERR(existing); + goto out; + } + mutex_lock(&existing->mutex); + *locked_ret = true; +out: + btrfs_put_block_group(bg_cache); + return ret; +} + +/* + * Unlock a full stripe. + * + * NOTE: Caller must ensure it's the same context calling corresponding + * lock_full_stripe(). + * + * Return 0 if we unlock full stripe without problem. + * Return <0 for error + */ +static int unlock_full_stripe(struct btrfs_fs_info *fs_info, u64 bytenr, + bool locked) +{ + struct btrfs_block_group_cache *bg_cache; + struct btrfs_full_stripe_locks_tree *locks_root; + struct full_stripe_lock *fstripe_lock; + u64 fstripe_start; + bool freeit = false; + int ret = 0; + + /* If we didn't acquire full stripe lock, no need to continue */ + if (!locked) + return 0; + + bg_cache = btrfs_lookup_block_group(fs_info, bytenr); + if (!bg_cache) { + ASSERT(0); + return -ENOENT; + } + if (!(bg_cache->flags & BTRFS_BLOCK_GROUP_RAID56_MASK)) + goto out; + + locks_root = &bg_cache->full_stripe_locks_root; + fstripe_start = get_full_stripe_logical(bg_cache, bytenr); + + mutex_lock(&locks_root->lock); + fstripe_lock = search_full_stripe_lock(locks_root, fstripe_start); + /* Unpaired unlock_full_stripe() detected */ + if (!fstripe_lock) { + WARN_ON(1); + ret = -ENOENT; + mutex_unlock(&locks_root->lock); + goto out; + } + + if (fstripe_lock->refs == 0) { + WARN_ON(1); + btrfs_warn(fs_info, "full stripe lock at %llu refcount underflow", + fstripe_lock->logical); + } else { + fstripe_lock->refs--; + } + + if (fstripe_lock->refs == 0) { + rb_erase(&fstripe_lock->node, &locks_root->root); + freeit = true; + } + mutex_unlock(&locks_root->lock); + + mutex_unlock(&fstripe_lock->mutex); + if (freeit) + kfree(fstripe_lock); +out: + btrfs_put_block_group(bg_cache); + return ret; +} + +/* * used for workers that require transaction commits (i.e., for the * NOCOW case) */ @@ -356,7 +579,7 @@ static void scrub_pending_trans_workers_inc(struct scrub_ctx *sctx) { struct btrfs_fs_info *fs_info = sctx->fs_info; - atomic_inc(&sctx->refs); + refcount_inc(&sctx->refs); /* * increment scrubs_running to prevent cancel requests from * completing as long as a worker is running. we must also @@ -447,7 +670,7 @@ static noinline_for_stack void scrub_free_ctx(struct scrub_ctx *sctx) static void scrub_put_ctx(struct scrub_ctx *sctx) { - if (atomic_dec_and_test(&sctx->refs)) + if (refcount_dec_and_test(&sctx->refs)) scrub_free_ctx(sctx); } @@ -462,7 +685,7 @@ struct scrub_ctx *scrub_setup_ctx(struct btrfs_device *dev, int is_dev_replace) sctx = kzalloc(sizeof(*sctx), GFP_KERNEL); if (!sctx) goto nomem; - atomic_set(&sctx->refs, 1); + refcount_set(&sctx->refs, 1); sctx->is_dev_replace = is_dev_replace; sctx->pages_per_rd_bio = SCRUB_PAGES_PER_RD_BIO; sctx->curr = -1; @@ -857,12 +1080,14 @@ out: static inline void scrub_get_recover(struct scrub_recover *recover) { - atomic_inc(&recover->refs); + refcount_inc(&recover->refs); } -static inline void scrub_put_recover(struct scrub_recover *recover) +static inline void scrub_put_recover(struct btrfs_fs_info *fs_info, + struct scrub_recover *recover) { - if (atomic_dec_and_test(&recover->refs)) { + if (refcount_dec_and_test(&recover->refs)) { + btrfs_bio_counter_dec(fs_info); btrfs_put_bbio(recover->bbio); kfree(recover); } @@ -892,6 +1117,7 @@ static int scrub_handle_errored_block(struct scrub_block *sblock_to_check) int mirror_index; int page_num; int success; + bool full_stripe_locked; static DEFINE_RATELIMIT_STATE(_rs, DEFAULT_RATELIMIT_INTERVAL, DEFAULT_RATELIMIT_BURST); @@ -917,6 +1143,24 @@ static int scrub_handle_errored_block(struct scrub_block *sblock_to_check) have_csum = sblock_to_check->pagev[0]->have_csum; dev = sblock_to_check->pagev[0]->dev; + /* + * For RAID5/6, race can happen for a different device scrub thread. + * For data corruption, Parity and Data threads will both try + * to recovery the data. + * Race can lead to doubly added csum error, or even unrecoverable + * error. + */ + ret = lock_full_stripe(fs_info, logical, &full_stripe_locked); + if (ret < 0) { + spin_lock(&sctx->stat_lock); + if (ret == -ENOMEM) + sctx->stat.malloc_errors++; + sctx->stat.read_errors++; + sctx->stat.uncorrectable_errors++; + spin_unlock(&sctx->stat_lock); + return ret; + } + if (sctx->is_dev_replace && !is_metadata && !have_csum) { sblocks_for_recheck = NULL; goto nodatasum_case; @@ -1241,7 +1485,7 @@ out: sblock->pagev[page_index]->sblock = NULL; recover = sblock->pagev[page_index]->recover; if (recover) { - scrub_put_recover(recover); + scrub_put_recover(fs_info, recover); sblock->pagev[page_index]->recover = NULL; } @@ -1251,6 +1495,9 @@ out: kfree(sblocks_for_recheck); } + ret = unlock_full_stripe(fs_info, logical, full_stripe_locked); + if (ret < 0) + return ret; return 0; } @@ -1330,20 +1577,23 @@ static int scrub_setup_recheck_block(struct scrub_block *original_sblock, * with a length of PAGE_SIZE, each returned stripe * represents one mirror */ + btrfs_bio_counter_inc_blocked(fs_info); ret = btrfs_map_sblock(fs_info, BTRFS_MAP_GET_READ_MIRRORS, - logical, &mapped_length, &bbio, 0, 1); + logical, &mapped_length, &bbio); if (ret || !bbio || mapped_length < sublen) { btrfs_put_bbio(bbio); + btrfs_bio_counter_dec(fs_info); return -EIO; } recover = kzalloc(sizeof(struct scrub_recover), GFP_NOFS); if (!recover) { btrfs_put_bbio(bbio); + btrfs_bio_counter_dec(fs_info); return -ENOMEM; } - atomic_set(&recover->refs, 1); + refcount_set(&recover->refs, 1); recover->bbio = bbio; recover->map_length = mapped_length; @@ -1365,7 +1615,7 @@ leave_nomem: spin_lock(&sctx->stat_lock); sctx->stat.malloc_errors++; spin_unlock(&sctx->stat_lock); - scrub_put_recover(recover); + scrub_put_recover(fs_info, recover); return -ENOMEM; } scrub_page_get(page); @@ -1407,7 +1657,7 @@ leave_nomem: scrub_get_recover(recover); page->recover = recover; } - scrub_put_recover(recover); + scrub_put_recover(fs_info, recover); length -= sublen; logical += sublen; page_index++; @@ -1497,14 +1747,18 @@ static void scrub_recheck_block(struct btrfs_fs_info *fs_info, bio_add_page(bio, page->page, PAGE_SIZE, 0); if (!retry_failed_mirror && scrub_is_page_on_raid56(page)) { - if (scrub_submit_raid56_bio_wait(fs_info, bio, page)) + if (scrub_submit_raid56_bio_wait(fs_info, bio, page)) { + page->io_error = 1; sblock->no_io_error_seen = 0; + } } else { bio->bi_iter.bi_sector = page->physical >> 9; bio_set_op_attrs(bio, REQ_OP_READ, 0); - if (btrfsic_submit_bio_wait(bio)) + if (btrfsic_submit_bio_wait(bio)) { + page->io_error = 1; sblock->no_io_error_seen = 0; + } } bio_put(bio); @@ -1634,7 +1888,7 @@ static int scrub_write_page_to_dev_replace(struct scrub_block *sblock, if (spage->io_error) { void *mapped_buffer = kmap_atomic(spage->page); - memset(mapped_buffer, 0, PAGE_SIZE); + clear_page(mapped_buffer); flush_dcache_page(spage->page); kunmap_atomic(mapped_buffer); } @@ -1998,12 +2252,12 @@ static int scrub_checksum_super(struct scrub_block *sblock) static void scrub_block_get(struct scrub_block *sblock) { - atomic_inc(&sblock->refs); + refcount_inc(&sblock->refs); } static void scrub_block_put(struct scrub_block *sblock) { - if (atomic_dec_and_test(&sblock->refs)) { + if (refcount_dec_and_test(&sblock->refs)) { int i; if (sblock->sparity) @@ -2187,8 +2441,9 @@ static void scrub_missing_raid56_pages(struct scrub_block *sblock) int ret; int i; + btrfs_bio_counter_inc_blocked(fs_info); ret = btrfs_map_sblock(fs_info, BTRFS_MAP_GET_READ_MIRRORS, logical, - &length, &bbio, 0, 1); + &length, &bbio); if (ret || !bbio || !bbio->raid_map) goto bbio_out; @@ -2231,6 +2486,7 @@ static void scrub_missing_raid56_pages(struct scrub_block *sblock) rbio_out: bio_put(bio); bbio_out: + btrfs_bio_counter_dec(fs_info); btrfs_put_bbio(bbio); spin_lock(&sctx->stat_lock); sctx->stat.malloc_errors++; @@ -2255,7 +2511,7 @@ static int scrub_pages(struct scrub_ctx *sctx, u64 logical, u64 len, /* one ref inside this function, plus one for each page added to * a bio later on */ - atomic_set(&sblock->refs, 1); + refcount_set(&sblock->refs, 1); sblock->sctx = sctx; sblock->no_io_error_seen = 1; @@ -2385,7 +2641,7 @@ static inline void __scrub_mark_bitmap(struct scrub_parity *sparity, unsigned long *bitmap, u64 start, u64 len) { - u32 offset; + u64 offset; int nsectors; int sectorsize = sparity->sctx->fs_info->sectorsize; @@ -2395,8 +2651,8 @@ static inline void __scrub_mark_bitmap(struct scrub_parity *sparity, } start -= sparity->logic_start; - start = div_u64_rem(start, sparity->stripe_len, &offset); - offset /= sectorsize; + start = div64_u64_rem(start, sparity->stripe_len, &offset); + offset = div_u64(offset, sectorsize); nsectors = (int)len / sectorsize; if (offset + nsectors <= sparity->nsectors) { @@ -2555,7 +2811,7 @@ static int scrub_pages_for_parity(struct scrub_parity *sparity, /* one ref inside this function, plus one for each page added to * a bio later on */ - atomic_set(&sblock->refs, 1); + refcount_set(&sblock->refs, 1); sblock->sctx = sctx; sblock->no_io_error_seen = 1; sblock->sparity = sparity; @@ -2694,7 +2950,7 @@ static int get_raid56_logic_offset(u64 physical, int num, for (i = 0; i < nr_data_stripes(map); i++) { *offset = last_offset + i * map->stripe_len; - stripe_nr = div_u64(*offset, map->stripe_len); + stripe_nr = div64_u64(*offset, map->stripe_len); stripe_nr = div_u64(stripe_nr, nr_data_stripes(map)); /* Work out the disk rotation on this stripe-set */ @@ -2765,7 +3021,6 @@ static void scrub_parity_check_and_repair(struct scrub_parity *sparity) struct btrfs_fs_info *fs_info = sctx->fs_info; struct bio *bio; struct btrfs_raid_bio *rbio; - struct scrub_page *spage; struct btrfs_bio *bbio = NULL; u64 length; int ret; @@ -2775,8 +3030,10 @@ static void scrub_parity_check_and_repair(struct scrub_parity *sparity) goto out; length = sparity->logic_end - sparity->logic_start; + + btrfs_bio_counter_inc_blocked(fs_info); ret = btrfs_map_sblock(fs_info, BTRFS_MAP_WRITE, sparity->logic_start, - &length, &bbio, 0, 1); + &length, &bbio); if (ret || !bbio || !bbio->raid_map) goto bbio_out; @@ -2795,9 +3052,6 @@ static void scrub_parity_check_and_repair(struct scrub_parity *sparity) if (!rbio) goto rbio_out; - list_for_each_entry(spage, &sparity->spages, list) - raid56_add_scrub_pages(rbio, spage->page, spage->logical); - scrub_pending_bio_inc(sctx); raid56_parity_submit_scrub_rbio(rbio); return; @@ -2805,6 +3059,7 @@ static void scrub_parity_check_and_repair(struct scrub_parity *sparity) rbio_out: bio_put(bio); bbio_out: + btrfs_bio_counter_dec(fs_info); btrfs_put_bbio(bbio); bitmap_or(sparity->ebitmap, sparity->ebitmap, sparity->dbitmap, sparity->nsectors); @@ -2822,12 +3077,12 @@ static inline int scrub_calc_parity_bitmap_len(int nsectors) static void scrub_parity_get(struct scrub_parity *sparity) { - atomic_inc(&sparity->refs); + refcount_inc(&sparity->refs); } static void scrub_parity_put(struct scrub_parity *sparity) { - if (!atomic_dec_and_test(&sparity->refs)) + if (!refcount_dec_and_test(&sparity->refs)) return; scrub_parity_check_and_repair(sparity); @@ -2879,7 +3134,7 @@ static noinline_for_stack int scrub_raid56_parity(struct scrub_ctx *sctx, sparity->scrub_dev = sdev; sparity->logic_start = logic_start; sparity->logic_end = logic_end; - atomic_set(&sparity->refs, 1); + refcount_set(&sparity->refs, 1); INIT_LIST_HEAD(&sparity->spages); sparity->dbitmap = sparity->bitmap; sparity->ebitmap = (void *)sparity->bitmap + bitmap_len; @@ -3098,7 +3353,7 @@ static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx, physical = map->stripes[num].physical; offset = 0; - nstripes = div_u64(length, map->stripe_len); + nstripes = div64_u64(length, map->stripe_len); if (map->type & BTRFS_BLOCK_GROUP_RAID0) { offset = map->stripe_len * num; increment = map->stripe_len * map->num_stripes; diff --git a/fs/btrfs/send.c b/fs/btrfs/send.c index 3f645cd67b54..fc496a6f842a 100644 --- a/fs/btrfs/send.c +++ b/fs/btrfs/send.c @@ -5184,13 +5184,19 @@ static int is_extent_unchanged(struct send_ctx *sctx, while (key.offset < ekey->offset + left_len) { ei = btrfs_item_ptr(eb, slot, struct btrfs_file_extent_item); right_type = btrfs_file_extent_type(eb, ei); - if (right_type != BTRFS_FILE_EXTENT_REG) { + if (right_type != BTRFS_FILE_EXTENT_REG && + right_type != BTRFS_FILE_EXTENT_INLINE) { ret = 0; goto out; } right_disknr = btrfs_file_extent_disk_bytenr(eb, ei); - right_len = btrfs_file_extent_num_bytes(eb, ei); + if (right_type == BTRFS_FILE_EXTENT_INLINE) { + right_len = btrfs_file_extent_inline_len(eb, slot, ei); + right_len = PAGE_ALIGN(right_len); + } else { + right_len = btrfs_file_extent_num_bytes(eb, ei); + } right_offset = btrfs_file_extent_offset(eb, ei); right_gen = btrfs_file_extent_generation(eb, ei); @@ -5204,6 +5210,19 @@ static int is_extent_unchanged(struct send_ctx *sctx, goto out; } + /* + * We just wanted to see if when we have an inline extent, what + * follows it is a regular extent (wanted to check the above + * condition for inline extents too). This should normally not + * happen but it's possible for example when we have an inline + * compressed extent representing data with a size matching + * the page size (currently the same as sector size). + */ + if (right_type == BTRFS_FILE_EXTENT_INLINE) { + ret = 0; + goto out; + } + left_offset_fixed = left_offset; if (key.offset < ekey->offset) { /* Fix the right offset for 2a and 7. */ diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index 72a053c9a7f0..4f1cdd5058f1 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -1795,8 +1795,7 @@ static int btrfs_remount(struct super_block *sb, int *flags, char *data) } if (fs_info->fs_devices->missing_devices > - fs_info->num_tolerated_disk_barrier_failures && - !(*flags & MS_RDONLY)) { + fs_info->num_tolerated_disk_barrier_failures) { btrfs_warn(fs_info, "too many missing devices, writeable remount is not allowed"); ret = -EACCES; diff --git a/fs/btrfs/tests/btrfs-tests.c b/fs/btrfs/tests/btrfs-tests.c index ea272432c930..b18ab8f327a5 100644 --- a/fs/btrfs/tests/btrfs-tests.c +++ b/fs/btrfs/tests/btrfs-tests.c @@ -237,7 +237,6 @@ void btrfs_init_dummy_trans(struct btrfs_trans_handle *trans) { memset(trans, 0, sizeof(*trans)); trans->transid = 1; - INIT_LIST_HEAD(&trans->qgroup_ref_list); trans->type = __TRANS_DUMMY; } diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 61b807de3e16..2168654c90a1 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -60,8 +60,8 @@ static const unsigned int btrfs_blocked_trans_types[TRANS_STATE_MAX] = { void btrfs_put_transaction(struct btrfs_transaction *transaction) { - WARN_ON(atomic_read(&transaction->use_count) == 0); - if (atomic_dec_and_test(&transaction->use_count)) { + WARN_ON(refcount_read(&transaction->use_count) == 0); + if (refcount_dec_and_test(&transaction->use_count)) { BUG_ON(!list_empty(&transaction->list)); WARN_ON(!RB_EMPTY_ROOT(&transaction->delayed_refs.href_root)); if (transaction->delayed_refs.pending_csums) @@ -207,7 +207,7 @@ loop: spin_unlock(&fs_info->trans_lock); return -EBUSY; } - atomic_inc(&cur_trans->use_count); + refcount_inc(&cur_trans->use_count); atomic_inc(&cur_trans->num_writers); extwriter_counter_inc(cur_trans, type); spin_unlock(&fs_info->trans_lock); @@ -257,7 +257,7 @@ loop: * One for this trans handle, one so it will live on until we * commit the transaction. */ - atomic_set(&cur_trans->use_count, 2); + refcount_set(&cur_trans->use_count, 2); atomic_set(&cur_trans->pending_ordered, 0); cur_trans->flags = 0; cur_trans->start_time = get_seconds(); @@ -432,7 +432,7 @@ static void wait_current_trans(struct btrfs_fs_info *fs_info) spin_lock(&fs_info->trans_lock); cur_trans = fs_info->running_transaction; if (cur_trans && is_transaction_blocked(cur_trans)) { - atomic_inc(&cur_trans->use_count); + refcount_inc(&cur_trans->use_count); spin_unlock(&fs_info->trans_lock); wait_event(fs_info->transaction_wait, @@ -572,7 +572,6 @@ again: h->type = type; h->can_flush_pending_bgs = true; - INIT_LIST_HEAD(&h->qgroup_ref_list); INIT_LIST_HEAD(&h->new_bgs); smp_mb(); @@ -744,7 +743,7 @@ int btrfs_wait_for_commit(struct btrfs_fs_info *fs_info, u64 transid) list_for_each_entry(t, &fs_info->trans_list, list) { if (t->transid == transid) { cur_trans = t; - atomic_inc(&cur_trans->use_count); + refcount_inc(&cur_trans->use_count); ret = 0; break; } @@ -773,7 +772,7 @@ int btrfs_wait_for_commit(struct btrfs_fs_info *fs_info, u64 transid) if (t->state == TRANS_STATE_COMPLETED) break; cur_trans = t; - atomic_inc(&cur_trans->use_count); + refcount_inc(&cur_trans->use_count); break; } } @@ -917,7 +916,6 @@ static int __btrfs_end_transaction(struct btrfs_trans_handle *trans, wake_up_process(info->transaction_kthread); err = -EIO; } - assert_qgroups_uptodate(trans); kmem_cache_free(btrfs_trans_handle_cachep, trans); if (must_run_delayed_refs) { @@ -1839,7 +1837,7 @@ int btrfs_commit_transaction_async(struct btrfs_trans_handle *trans, /* take transaction reference */ cur_trans = trans->transaction; - atomic_inc(&cur_trans->use_count); + refcount_inc(&cur_trans->use_count); btrfs_end_transaction(trans); @@ -2015,7 +2013,7 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans) spin_lock(&fs_info->trans_lock); if (cur_trans->state >= TRANS_STATE_COMMIT_START) { spin_unlock(&fs_info->trans_lock); - atomic_inc(&cur_trans->use_count); + refcount_inc(&cur_trans->use_count); ret = btrfs_end_transaction(trans); wait_for_commit(cur_trans); @@ -2035,7 +2033,7 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans) prev_trans = list_entry(cur_trans->list.prev, struct btrfs_transaction, list); if (prev_trans->state != TRANS_STATE_COMPLETED) { - atomic_inc(&prev_trans->use_count); + refcount_inc(&prev_trans->use_count); spin_unlock(&fs_info->trans_lock); wait_for_commit(prev_trans); @@ -2130,13 +2128,6 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans) goto scrub_continue; } - /* Reocrd old roots for later qgroup accounting */ - ret = btrfs_qgroup_prepare_account_extents(trans, fs_info); - if (ret) { - mutex_unlock(&fs_info->reloc_mutex); - goto scrub_continue; - } - /* * make sure none of the code above managed to slip in a * delayed item @@ -2179,6 +2170,24 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans) btrfs_free_log_root_tree(trans, fs_info); /* + * commit_fs_roots() can call btrfs_save_ino_cache(), which generates + * new delayed refs. Must handle them or qgroup can be wrong. + */ + ret = btrfs_run_delayed_refs(trans, fs_info, (unsigned long)-1); + if (ret) { + mutex_unlock(&fs_info->tree_log_mutex); + mutex_unlock(&fs_info->reloc_mutex); + goto scrub_continue; + } + + ret = btrfs_qgroup_prepare_account_extents(trans, fs_info); + if (ret) { + mutex_unlock(&fs_info->tree_log_mutex); + mutex_unlock(&fs_info->reloc_mutex); + goto scrub_continue; + } + + /* * Since fs roots are all committed, we can get a quite accurate * new_roots. So let's do quota accounting. */ @@ -2223,7 +2232,6 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans) switch_commit_roots(cur_trans, fs_info); - assert_qgroups_uptodate(trans); ASSERT(list_empty(&cur_trans->dirty_bgs)); ASSERT(list_empty(&cur_trans->io_bgs)); update_super_roots(fs_info); diff --git a/fs/btrfs/transaction.h b/fs/btrfs/transaction.h index 5dfb5590fff6..c55e44560103 100644 --- a/fs/btrfs/transaction.h +++ b/fs/btrfs/transaction.h @@ -18,6 +18,8 @@ #ifndef __BTRFS_TRANSACTION__ #define __BTRFS_TRANSACTION__ + +#include <linux/refcount.h> #include "btrfs_inode.h" #include "delayed-ref.h" #include "ctree.h" @@ -49,7 +51,7 @@ struct btrfs_transaction { * transaction can end */ atomic_t num_writers; - atomic_t use_count; + refcount_t use_count; atomic_t pending_ordered; unsigned long flags; @@ -125,8 +127,6 @@ struct btrfs_trans_handle { unsigned int type; struct btrfs_root *root; struct btrfs_fs_info *fs_info; - struct seq_list delayed_ref_elem; - struct list_head qgroup_ref_list; struct list_head new_bgs; }; diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c index a59674c3e69e..ccfe9fe7754a 100644 --- a/fs/btrfs/tree-log.c +++ b/fs/btrfs/tree-log.c @@ -4196,7 +4196,7 @@ static int btrfs_log_changed_extents(struct btrfs_trans_handle *trans, if (em->generation <= test_gen) continue; /* Need a ref to keep it from getting evicted from cache */ - atomic_inc(&em->refs); + refcount_inc(&em->refs); set_bit(EXTENT_FLAG_LOGGING, &em->flags); list_add_tail(&em->list, &extents); num++; diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index ab8a66d852f9..017b67daa3bb 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -139,6 +139,11 @@ static int btrfs_relocate_sys_chunks(struct btrfs_fs_info *fs_info); static void __btrfs_reset_dev_stats(struct btrfs_device *dev); static void btrfs_dev_stat_print_on_error(struct btrfs_device *dev); static void btrfs_dev_stat_print_on_load(struct btrfs_device *device); +static int __btrfs_map_block(struct btrfs_fs_info *fs_info, + enum btrfs_map_op op, + u64 logical, u64 *length, + struct btrfs_bio **bbio_ret, + int mirror_num, int need_raid_map); DEFINE_MUTEX(uuid_mutex); static LIST_HEAD(fs_uuids); @@ -1008,14 +1013,13 @@ static int __btrfs_open_devices(struct btrfs_fs_devices *fs_devices, q = bdev_get_queue(bdev); if (blk_queue_discard(q)) device->can_discard = 1; + if (!blk_queue_nonrot(q)) + fs_devices->rotating = 1; device->bdev = bdev; device->in_fs_metadata = 0; device->mode = flags; - if (!blk_queue_nonrot(bdev_get_queue(bdev))) - fs_devices->rotating = 1; - fs_devices->open_devices++; if (device->writeable && device->devid != BTRFS_DEV_REPLACE_DEVID) { @@ -2417,7 +2421,7 @@ int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path fs_info->free_chunk_space += device->total_bytes; spin_unlock(&fs_info->free_chunk_lock); - if (!blk_queue_nonrot(bdev_get_queue(bdev))) + if (!blk_queue_nonrot(q)) fs_info->fs_devices->rotating = 1; tmp = btrfs_super_total_bytes(fs_info->super_copy); @@ -2795,10 +2799,38 @@ static int btrfs_del_sys_chunk(struct btrfs_fs_info *fs_info, return ret; } +static struct extent_map *get_chunk_map(struct btrfs_fs_info *fs_info, + u64 logical, u64 length) +{ + struct extent_map_tree *em_tree; + struct extent_map *em; + + em_tree = &fs_info->mapping_tree.map_tree; + read_lock(&em_tree->lock); + em = lookup_extent_mapping(em_tree, logical, length); + read_unlock(&em_tree->lock); + + if (!em) { + btrfs_crit(fs_info, "unable to find logical %llu length %llu", + logical, length); + return ERR_PTR(-EINVAL); + } + + if (em->start > logical || em->start + em->len < logical) { + btrfs_crit(fs_info, + "found a bad mapping, wanted %llu-%llu, found %llu-%llu", + logical, length, em->start, em->start + em->len); + free_extent_map(em); + return ERR_PTR(-EINVAL); + } + + /* callers are responsible for dropping em's ref. */ + return em; +} + int btrfs_remove_chunk(struct btrfs_trans_handle *trans, struct btrfs_fs_info *fs_info, u64 chunk_offset) { - struct extent_map_tree *em_tree; struct extent_map *em; struct map_lookup *map; u64 dev_extent_len = 0; @@ -2806,23 +2838,15 @@ int btrfs_remove_chunk(struct btrfs_trans_handle *trans, int i, ret = 0; struct btrfs_fs_devices *fs_devices = fs_info->fs_devices; - em_tree = &fs_info->mapping_tree.map_tree; - - read_lock(&em_tree->lock); - em = lookup_extent_mapping(em_tree, chunk_offset, 1); - read_unlock(&em_tree->lock); - - if (!em || em->start > chunk_offset || - em->start + em->len < chunk_offset) { + em = get_chunk_map(fs_info, chunk_offset, 1); + if (IS_ERR(em)) { /* * This is a logic error, but we don't want to just rely on the * user having built with ASSERT enabled, so if ASSERT doesn't * do anything we still error out. */ ASSERT(0); - if (em) - free_extent_map(em); - return -EINVAL; + return PTR_ERR(em); } map = em->map_lookup; mutex_lock(&fs_info->chunk_mutex); @@ -3736,7 +3760,7 @@ static void __cancel_balance(struct btrfs_fs_info *fs_info) if (ret) btrfs_handle_fs_error(fs_info, ret, NULL); - atomic_set(&fs_info->mutually_exclusive_operation_running, 0); + clear_bit(BTRFS_FS_EXCL_OP, &fs_info->flags); } /* Non-zero return value signifies invalidity */ @@ -3755,6 +3779,7 @@ int btrfs_balance(struct btrfs_balance_control *bctl, struct btrfs_ioctl_balance_args *bargs) { struct btrfs_fs_info *fs_info = bctl->fs_info; + u64 meta_target, data_target; u64 allowed; int mixed = 0; int ret; @@ -3851,11 +3876,16 @@ int btrfs_balance(struct btrfs_balance_control *bctl, } } while (read_seqretry(&fs_info->profiles_lock, seq)); - if (btrfs_get_num_tolerated_disk_barrier_failures(bctl->meta.target) < - btrfs_get_num_tolerated_disk_barrier_failures(bctl->data.target)) { + /* if we're not converting, the target field is uninitialized */ + meta_target = (bctl->meta.flags & BTRFS_BALANCE_ARGS_CONVERT) ? + bctl->meta.target : fs_info->avail_metadata_alloc_bits; + data_target = (bctl->data.flags & BTRFS_BALANCE_ARGS_CONVERT) ? + bctl->data.target : fs_info->avail_data_alloc_bits; + if (btrfs_get_num_tolerated_disk_barrier_failures(meta_target) < + btrfs_get_num_tolerated_disk_barrier_failures(data_target)) { btrfs_warn(fs_info, "metadata profile 0x%llx has lower redundancy than data profile 0x%llx", - bctl->meta.target, bctl->data.target); + meta_target, data_target); } if (bctl->sys.flags & BTRFS_BALANCE_ARGS_CONVERT) { @@ -3910,7 +3940,7 @@ out: __cancel_balance(fs_info); else { kfree(bctl); - atomic_set(&fs_info->mutually_exclusive_operation_running, 0); + clear_bit(BTRFS_FS_EXCL_OP, &fs_info->flags); } return ret; } @@ -4000,7 +4030,7 @@ int btrfs_recover_balance(struct btrfs_fs_info *fs_info) btrfs_balance_sys(leaf, item, &disk_bargs); btrfs_disk_balance_args_to_cpu(&bctl->sys, &disk_bargs); - WARN_ON(atomic_xchg(&fs_info->mutually_exclusive_operation_running, 1)); + WARN_ON(test_and_set_bit(BTRFS_FS_EXCL_OP, &fs_info->flags)); mutex_lock(&fs_info->volume_mutex); mutex_lock(&fs_info->balance_mutex); @@ -4785,7 +4815,7 @@ static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans, stripe_size = div_u64(stripe_size, dev_stripes); /* align to BTRFS_STRIPE_LEN */ - stripe_size = div_u64(stripe_size, raid_stripe_len); + stripe_size = div64_u64(stripe_size, raid_stripe_len); stripe_size *= raid_stripe_len; map = kmalloc(map_lookup_size(num_stripes), GFP_NOFS); @@ -4833,7 +4863,7 @@ static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans, ret = add_extent_mapping(em_tree, em, 0); if (!ret) { list_add_tail(&em->list, &trans->transaction->pending_chunks); - atomic_inc(&em->refs); + refcount_inc(&em->refs); } write_unlock(&em_tree->lock); if (ret) { @@ -4888,7 +4918,6 @@ int btrfs_finish_chunk_alloc(struct btrfs_trans_handle *trans, struct btrfs_device *device; struct btrfs_chunk *chunk; struct btrfs_stripe *stripe; - struct extent_map_tree *em_tree; struct extent_map *em; struct map_lookup *map; size_t item_size; @@ -4897,24 +4926,9 @@ int btrfs_finish_chunk_alloc(struct btrfs_trans_handle *trans, int i = 0; int ret = 0; - em_tree = &fs_info->mapping_tree.map_tree; - read_lock(&em_tree->lock); - em = lookup_extent_mapping(em_tree, chunk_offset, chunk_size); - read_unlock(&em_tree->lock); - - if (!em) { - btrfs_crit(fs_info, "unable to find logical %Lu len %Lu", - chunk_offset, chunk_size); - return -EINVAL; - } - - if (em->start != chunk_offset || em->len != chunk_size) { - btrfs_crit(fs_info, - "found a bad mapping, wanted %Lu-%Lu, found %Lu-%Lu", - chunk_offset, chunk_size, em->start, em->len); - free_extent_map(em); - return -EINVAL; - } + em = get_chunk_map(fs_info, chunk_offset, chunk_size); + if (IS_ERR(em)) + return PTR_ERR(em); map = em->map_lookup; item_size = btrfs_chunk_item_size(map->num_stripes); @@ -5055,15 +5069,12 @@ int btrfs_chunk_readonly(struct btrfs_fs_info *fs_info, u64 chunk_offset) { struct extent_map *em; struct map_lookup *map; - struct btrfs_mapping_tree *map_tree = &fs_info->mapping_tree; int readonly = 0; int miss_ndevs = 0; int i; - read_lock(&map_tree->map_tree.lock); - em = lookup_extent_mapping(&map_tree->map_tree, chunk_offset, 1); - read_unlock(&map_tree->map_tree.lock); - if (!em) + em = get_chunk_map(fs_info, chunk_offset, 1); + if (IS_ERR(em)) return 1; map = em->map_lookup; @@ -5117,34 +5128,19 @@ void btrfs_mapping_tree_free(struct btrfs_mapping_tree *tree) int btrfs_num_copies(struct btrfs_fs_info *fs_info, u64 logical, u64 len) { - struct btrfs_mapping_tree *map_tree = &fs_info->mapping_tree; struct extent_map *em; struct map_lookup *map; - struct extent_map_tree *em_tree = &map_tree->map_tree; int ret; - read_lock(&em_tree->lock); - em = lookup_extent_mapping(em_tree, logical, len); - read_unlock(&em_tree->lock); - - /* - * We could return errors for these cases, but that could get ugly and - * we'd probably do the same thing which is just not do anything else - * and exit, so return 1 so the callers don't try to use other copies. - */ - if (!em) { - btrfs_crit(fs_info, "No mapping for %Lu-%Lu", logical, - logical+len); - return 1; - } - - if (em->start > logical || em->start + em->len < logical) { - btrfs_crit(fs_info, "Invalid mapping for %Lu-%Lu, got %Lu-%Lu", - logical, logical+len, em->start, - em->start + em->len); - free_extent_map(em); + em = get_chunk_map(fs_info, logical, len); + if (IS_ERR(em)) + /* + * We could return errors for these cases, but that could get + * ugly and we'd probably do the same thing which is just not do + * anything else and exit, so return 1 so the callers don't try + * to use other copies. + */ return 1; - } map = em->map_lookup; if (map->type & (BTRFS_BLOCK_GROUP_DUP | BTRFS_BLOCK_GROUP_RAID1)) @@ -5160,7 +5156,8 @@ int btrfs_num_copies(struct btrfs_fs_info *fs_info, u64 logical, u64 len) free_extent_map(em); btrfs_dev_replace_lock(&fs_info->dev_replace, 0); - if (btrfs_dev_replace_is_ongoing(&fs_info->dev_replace)) + if (btrfs_dev_replace_is_ongoing(&fs_info->dev_replace) && + fs_info->dev_replace.tgtdev) ret++; btrfs_dev_replace_unlock(&fs_info->dev_replace, 0); @@ -5173,15 +5170,11 @@ unsigned long btrfs_full_stripe_len(struct btrfs_fs_info *fs_info, { struct extent_map *em; struct map_lookup *map; - struct extent_map_tree *em_tree = &map_tree->map_tree; unsigned long len = fs_info->sectorsize; - read_lock(&em_tree->lock); - em = lookup_extent_mapping(em_tree, logical, len); - read_unlock(&em_tree->lock); - BUG_ON(!em); + em = get_chunk_map(fs_info, logical, len); + WARN_ON(IS_ERR(em)); - BUG_ON(em->start > logical || em->start + em->len < logical); map = em->map_lookup; if (map->type & BTRFS_BLOCK_GROUP_RAID56_MASK) len = map->stripe_len * nr_data_stripes(map); @@ -5189,20 +5182,16 @@ unsigned long btrfs_full_stripe_len(struct btrfs_fs_info *fs_info, return len; } -int btrfs_is_parity_mirror(struct btrfs_mapping_tree *map_tree, +int btrfs_is_parity_mirror(struct btrfs_fs_info *fs_info, u64 logical, u64 len, int mirror_num) { struct extent_map *em; struct map_lookup *map; - struct extent_map_tree *em_tree = &map_tree->map_tree; int ret = 0; - read_lock(&em_tree->lock); - em = lookup_extent_mapping(em_tree, logical, len); - read_unlock(&em_tree->lock); - BUG_ON(!em); + em = get_chunk_map(fs_info, logical, len); + WARN_ON(IS_ERR(em)); - BUG_ON(em->start > logical || em->start + em->len < logical); map = em->map_lookup; if (map->type & BTRFS_BLOCK_GROUP_RAID56_MASK) ret = 1; @@ -5295,25 +5284,353 @@ static struct btrfs_bio *alloc_btrfs_bio(int total_stripes, int real_stripes) GFP_NOFS|__GFP_NOFAIL); atomic_set(&bbio->error, 0); - atomic_set(&bbio->refs, 1); + refcount_set(&bbio->refs, 1); return bbio; } void btrfs_get_bbio(struct btrfs_bio *bbio) { - WARN_ON(!atomic_read(&bbio->refs)); - atomic_inc(&bbio->refs); + WARN_ON(!refcount_read(&bbio->refs)); + refcount_inc(&bbio->refs); } void btrfs_put_bbio(struct btrfs_bio *bbio) { if (!bbio) return; - if (atomic_dec_and_test(&bbio->refs)) + if (refcount_dec_and_test(&bbio->refs)) kfree(bbio); } +/* can REQ_OP_DISCARD be sent with other REQ like REQ_OP_WRITE? */ +/* + * Please note that, discard won't be sent to target device of device + * replace. + */ +static int __btrfs_map_block_for_discard(struct btrfs_fs_info *fs_info, + u64 logical, u64 length, + struct btrfs_bio **bbio_ret) +{ + struct extent_map *em; + struct map_lookup *map; + struct btrfs_bio *bbio; + u64 offset; + u64 stripe_nr; + u64 stripe_nr_end; + u64 stripe_end_offset; + u64 stripe_cnt; + u64 stripe_len; + u64 stripe_offset; + u64 num_stripes; + u32 stripe_index; + u32 factor = 0; + u32 sub_stripes = 0; + u64 stripes_per_dev = 0; + u32 remaining_stripes = 0; + u32 last_stripe = 0; + int ret = 0; + int i; + + /* discard always return a bbio */ + ASSERT(bbio_ret); + + em = get_chunk_map(fs_info, logical, length); + if (IS_ERR(em)) + return PTR_ERR(em); + + map = em->map_lookup; + /* we don't discard raid56 yet */ + if (map->type & BTRFS_BLOCK_GROUP_RAID56_MASK) { + ret = -EOPNOTSUPP; + goto out; + } + + offset = logical - em->start; + length = min_t(u64, em->len - offset, length); + + stripe_len = map->stripe_len; + /* + * stripe_nr counts the total number of stripes we have to stride + * to get to this block + */ + stripe_nr = div64_u64(offset, stripe_len); + + /* stripe_offset is the offset of this block in its stripe */ + stripe_offset = offset - stripe_nr * stripe_len; + + stripe_nr_end = round_up(offset + length, map->stripe_len); + stripe_nr_end = div64_u64(stripe_nr_end, map->stripe_len); + stripe_cnt = stripe_nr_end - stripe_nr; + stripe_end_offset = stripe_nr_end * map->stripe_len - + (offset + length); + /* + * after this, stripe_nr is the number of stripes on this + * device we have to walk to find the data, and stripe_index is + * the number of our device in the stripe array + */ + num_stripes = 1; + stripe_index = 0; + if (map->type & (BTRFS_BLOCK_GROUP_RAID0 | + BTRFS_BLOCK_GROUP_RAID10)) { + if (map->type & BTRFS_BLOCK_GROUP_RAID0) + sub_stripes = 1; + else + sub_stripes = map->sub_stripes; + + factor = map->num_stripes / sub_stripes; + num_stripes = min_t(u64, map->num_stripes, + sub_stripes * stripe_cnt); + stripe_nr = div_u64_rem(stripe_nr, factor, &stripe_index); + stripe_index *= sub_stripes; + stripes_per_dev = div_u64_rem(stripe_cnt, factor, + &remaining_stripes); + div_u64_rem(stripe_nr_end - 1, factor, &last_stripe); + last_stripe *= sub_stripes; + } else if (map->type & (BTRFS_BLOCK_GROUP_RAID1 | + BTRFS_BLOCK_GROUP_DUP)) { + num_stripes = map->num_stripes; + } else { + stripe_nr = div_u64_rem(stripe_nr, map->num_stripes, + &stripe_index); + } + + bbio = alloc_btrfs_bio(num_stripes, 0); + if (!bbio) { + ret = -ENOMEM; + goto out; + } + + for (i = 0; i < num_stripes; i++) { + bbio->stripes[i].physical = + map->stripes[stripe_index].physical + + stripe_offset + stripe_nr * map->stripe_len; + bbio->stripes[i].dev = map->stripes[stripe_index].dev; + + if (map->type & (BTRFS_BLOCK_GROUP_RAID0 | + BTRFS_BLOCK_GROUP_RAID10)) { + bbio->stripes[i].length = stripes_per_dev * + map->stripe_len; + + if (i / sub_stripes < remaining_stripes) + bbio->stripes[i].length += + map->stripe_len; + + /* + * Special for the first stripe and + * the last stripe: + * + * |-------|...|-------| + * |----------| + * off end_off + */ + if (i < sub_stripes) + bbio->stripes[i].length -= + stripe_offset; + + if (stripe_index >= last_stripe && + stripe_index <= (last_stripe + + sub_stripes - 1)) + bbio->stripes[i].length -= + stripe_end_offset; + + if (i == sub_stripes - 1) + stripe_offset = 0; + } else { + bbio->stripes[i].length = length; + } + + stripe_index++; + if (stripe_index == map->num_stripes) { + stripe_index = 0; + stripe_nr++; + } + } + + *bbio_ret = bbio; + bbio->map_type = map->type; + bbio->num_stripes = num_stripes; +out: + free_extent_map(em); + return ret; +} + +/* + * In dev-replace case, for repair case (that's the only case where the mirror + * is selected explicitly when calling btrfs_map_block), blocks left of the + * left cursor can also be read from the target drive. + * + * For REQ_GET_READ_MIRRORS, the target drive is added as the last one to the + * array of stripes. + * For READ, it also needs to be supported using the same mirror number. + * + * If the requested block is not left of the left cursor, EIO is returned. This + * can happen because btrfs_num_copies() returns one more in the dev-replace + * case. + */ +static int get_extra_mirror_from_replace(struct btrfs_fs_info *fs_info, + u64 logical, u64 length, + u64 srcdev_devid, int *mirror_num, + u64 *physical) +{ + struct btrfs_bio *bbio = NULL; + int num_stripes; + int index_srcdev = 0; + int found = 0; + u64 physical_of_found = 0; + int i; + int ret = 0; + + ret = __btrfs_map_block(fs_info, BTRFS_MAP_GET_READ_MIRRORS, + logical, &length, &bbio, 0, 0); + if (ret) { + ASSERT(bbio == NULL); + return ret; + } + + num_stripes = bbio->num_stripes; + if (*mirror_num > num_stripes) { + /* + * BTRFS_MAP_GET_READ_MIRRORS does not contain this mirror, + * that means that the requested area is not left of the left + * cursor + */ + btrfs_put_bbio(bbio); + return -EIO; + } + + /* + * process the rest of the function using the mirror_num of the source + * drive. Therefore look it up first. At the end, patch the device + * pointer to the one of the target drive. + */ + for (i = 0; i < num_stripes; i++) { + if (bbio->stripes[i].dev->devid != srcdev_devid) + continue; + + /* + * In case of DUP, in order to keep it simple, only add the + * mirror with the lowest physical address + */ + if (found && + physical_of_found <= bbio->stripes[i].physical) + continue; + + index_srcdev = i; + found = 1; + physical_of_found = bbio->stripes[i].physical; + } + + btrfs_put_bbio(bbio); + + ASSERT(found); + if (!found) + return -EIO; + + *mirror_num = index_srcdev + 1; + *physical = physical_of_found; + return ret; +} + +static void handle_ops_on_dev_replace(enum btrfs_map_op op, + struct btrfs_bio **bbio_ret, + struct btrfs_dev_replace *dev_replace, + int *num_stripes_ret, int *max_errors_ret) +{ + struct btrfs_bio *bbio = *bbio_ret; + u64 srcdev_devid = dev_replace->srcdev->devid; + int tgtdev_indexes = 0; + int num_stripes = *num_stripes_ret; + int max_errors = *max_errors_ret; + int i; + + if (op == BTRFS_MAP_WRITE) { + int index_where_to_add; + + /* + * duplicate the write operations while the dev replace + * procedure is running. Since the copying of the old disk to + * the new disk takes place at run time while the filesystem is + * mounted writable, the regular write operations to the old + * disk have to be duplicated to go to the new disk as well. + * + * Note that device->missing is handled by the caller, and that + * the write to the old disk is already set up in the stripes + * array. + */ + index_where_to_add = num_stripes; + for (i = 0; i < num_stripes; i++) { + if (bbio->stripes[i].dev->devid == srcdev_devid) { + /* write to new disk, too */ + struct btrfs_bio_stripe *new = + bbio->stripes + index_where_to_add; + struct btrfs_bio_stripe *old = + bbio->stripes + i; + + new->physical = old->physical; + new->length = old->length; + new->dev = dev_replace->tgtdev; + bbio->tgtdev_map[i] = index_where_to_add; + index_where_to_add++; + max_errors++; + tgtdev_indexes++; + } + } + num_stripes = index_where_to_add; + } else if (op == BTRFS_MAP_GET_READ_MIRRORS) { + int index_srcdev = 0; + int found = 0; + u64 physical_of_found = 0; + + /* + * During the dev-replace procedure, the target drive can also + * be used to read data in case it is needed to repair a corrupt + * block elsewhere. This is possible if the requested area is + * left of the left cursor. In this area, the target drive is a + * full copy of the source drive. + */ + for (i = 0; i < num_stripes; i++) { + if (bbio->stripes[i].dev->devid == srcdev_devid) { + /* + * In case of DUP, in order to keep it simple, + * only add the mirror with the lowest physical + * address + */ + if (found && + physical_of_found <= + bbio->stripes[i].physical) + continue; + index_srcdev = i; + found = 1; + physical_of_found = bbio->stripes[i].physical; + } + } + if (found) { + struct btrfs_bio_stripe *tgtdev_stripe = + bbio->stripes + num_stripes; + + tgtdev_stripe->physical = physical_of_found; + tgtdev_stripe->length = + bbio->stripes[index_srcdev].length; + tgtdev_stripe->dev = dev_replace->tgtdev; + bbio->tgtdev_map[index_srcdev] = num_stripes; + + tgtdev_indexes++; + num_stripes++; + } + } + + *num_stripes_ret = num_stripes; + *max_errors_ret = max_errors; + bbio->num_tgtdevs = tgtdev_indexes; + *bbio_ret = bbio; +} + +static bool need_full_stripe(enum btrfs_map_op op) +{ + return (op == BTRFS_MAP_WRITE || op == BTRFS_MAP_GET_READ_MIRRORS); +} + static int __btrfs_map_block(struct btrfs_fs_info *fs_info, enum btrfs_map_op op, u64 logical, u64 *length, @@ -5322,14 +5639,9 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, { struct extent_map *em; struct map_lookup *map; - struct btrfs_mapping_tree *map_tree = &fs_info->mapping_tree; - struct extent_map_tree *em_tree = &map_tree->map_tree; u64 offset; u64 stripe_offset; - u64 stripe_end_offset; u64 stripe_nr; - u64 stripe_nr_orig; - u64 stripe_nr_end; u64 stripe_len; u32 stripe_index; int i; @@ -5345,23 +5657,13 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, u64 physical_to_patch_in_first_stripe = 0; u64 raid56_full_stripe_start = (u64)-1; - read_lock(&em_tree->lock); - em = lookup_extent_mapping(em_tree, logical, *length); - read_unlock(&em_tree->lock); - - if (!em) { - btrfs_crit(fs_info, "unable to find logical %llu len %llu", - logical, *length); - return -EINVAL; - } + if (op == BTRFS_MAP_DISCARD) + return __btrfs_map_block_for_discard(fs_info, logical, + *length, bbio_ret); - if (em->start > logical || em->start + em->len < logical) { - btrfs_crit(fs_info, - "found a bad mapping, wanted %Lu, found %Lu-%Lu", - logical, em->start, em->start + em->len); - free_extent_map(em); - return -EINVAL; - } + em = get_chunk_map(fs_info, logical, *length); + if (IS_ERR(em)) + return PTR_ERR(em); map = em->map_lookup; offset = logical - em->start; @@ -5400,14 +5702,7 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, raid56_full_stripe_start *= full_stripe_len; } - if (op == BTRFS_MAP_DISCARD) { - /* we don't discard raid56 yet */ - if (map->type & BTRFS_BLOCK_GROUP_RAID56_MASK) { - ret = -EOPNOTSUPP; - goto out; - } - *length = min_t(u64, em->len - offset, *length); - } else if (map->type & BTRFS_BLOCK_GROUP_PROFILE_MASK) { + if (map->type & BTRFS_BLOCK_GROUP_PROFILE_MASK) { u64 max_len; /* For writes to RAID[56], allow a full stripeset across all disks. For other RAID types and for RAID[56] reads, just allow a single @@ -5438,105 +5733,28 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, btrfs_dev_replace_set_lock_blocking(dev_replace); if (dev_replace_is_ongoing && mirror_num == map->num_stripes + 1 && - op != BTRFS_MAP_WRITE && op != BTRFS_MAP_DISCARD && - op != BTRFS_MAP_GET_READ_MIRRORS && dev_replace->tgtdev != NULL) { - /* - * in dev-replace case, for repair case (that's the only - * case where the mirror is selected explicitly when - * calling btrfs_map_block), blocks left of the left cursor - * can also be read from the target drive. - * For REQ_GET_READ_MIRRORS, the target drive is added as - * the last one to the array of stripes. For READ, it also - * needs to be supported using the same mirror number. - * If the requested block is not left of the left cursor, - * EIO is returned. This can happen because btrfs_num_copies() - * returns one more in the dev-replace case. - */ - u64 tmp_length = *length; - struct btrfs_bio *tmp_bbio = NULL; - int tmp_num_stripes; - u64 srcdev_devid = dev_replace->srcdev->devid; - int index_srcdev = 0; - int found = 0; - u64 physical_of_found = 0; - - ret = __btrfs_map_block(fs_info, BTRFS_MAP_GET_READ_MIRRORS, - logical, &tmp_length, &tmp_bbio, 0, 0); - if (ret) { - WARN_ON(tmp_bbio != NULL); - goto out; - } - - tmp_num_stripes = tmp_bbio->num_stripes; - if (mirror_num > tmp_num_stripes) { - /* - * BTRFS_MAP_GET_READ_MIRRORS does not contain this - * mirror, that means that the requested area - * is not left of the left cursor - */ - ret = -EIO; - btrfs_put_bbio(tmp_bbio); - goto out; - } - - /* - * process the rest of the function using the mirror_num - * of the source drive. Therefore look it up first. - * At the end, patch the device pointer to the one of the - * target drive. - */ - for (i = 0; i < tmp_num_stripes; i++) { - if (tmp_bbio->stripes[i].dev->devid != srcdev_devid) - continue; - - /* - * In case of DUP, in order to keep it simple, only add - * the mirror with the lowest physical address - */ - if (found && - physical_of_found <= tmp_bbio->stripes[i].physical) - continue; - - index_srcdev = i; - found = 1; - physical_of_found = tmp_bbio->stripes[i].physical; - } - - btrfs_put_bbio(tmp_bbio); - - if (!found) { - WARN_ON(1); - ret = -EIO; + !need_full_stripe(op) && dev_replace->tgtdev != NULL) { + ret = get_extra_mirror_from_replace(fs_info, logical, *length, + dev_replace->srcdev->devid, + &mirror_num, + &physical_to_patch_in_first_stripe); + if (ret) goto out; - } - - mirror_num = index_srcdev + 1; - patch_the_first_stripe_for_dev_replace = 1; - physical_to_patch_in_first_stripe = physical_of_found; + else + patch_the_first_stripe_for_dev_replace = 1; } else if (mirror_num > map->num_stripes) { mirror_num = 0; } num_stripes = 1; stripe_index = 0; - stripe_nr_orig = stripe_nr; - stripe_nr_end = ALIGN(offset + *length, map->stripe_len); - stripe_nr_end = div_u64(stripe_nr_end, map->stripe_len); - stripe_end_offset = stripe_nr_end * map->stripe_len - - (offset + *length); - if (map->type & BTRFS_BLOCK_GROUP_RAID0) { - if (op == BTRFS_MAP_DISCARD) - num_stripes = min_t(u64, map->num_stripes, - stripe_nr_end - stripe_nr_orig); stripe_nr = div_u64_rem(stripe_nr, map->num_stripes, &stripe_index); - if (op != BTRFS_MAP_WRITE && op != BTRFS_MAP_DISCARD && - op != BTRFS_MAP_GET_READ_MIRRORS) + if (op != BTRFS_MAP_WRITE && op != BTRFS_MAP_GET_READ_MIRRORS) mirror_num = 1; } else if (map->type & BTRFS_BLOCK_GROUP_RAID1) { - if (op == BTRFS_MAP_WRITE || op == BTRFS_MAP_DISCARD || - op == BTRFS_MAP_GET_READ_MIRRORS) + if (op == BTRFS_MAP_WRITE || op == BTRFS_MAP_GET_READ_MIRRORS) num_stripes = map->num_stripes; else if (mirror_num) stripe_index = mirror_num - 1; @@ -5549,8 +5767,7 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, } } else if (map->type & BTRFS_BLOCK_GROUP_DUP) { - if (op == BTRFS_MAP_WRITE || op == BTRFS_MAP_DISCARD || - op == BTRFS_MAP_GET_READ_MIRRORS) { + if (op == BTRFS_MAP_WRITE || op == BTRFS_MAP_GET_READ_MIRRORS) { num_stripes = map->num_stripes; } else if (mirror_num) { stripe_index = mirror_num - 1; @@ -5566,10 +5783,6 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, if (op == BTRFS_MAP_WRITE || op == BTRFS_MAP_GET_READ_MIRRORS) num_stripes = map->sub_stripes; - else if (op == BTRFS_MAP_DISCARD) - num_stripes = min_t(u64, map->sub_stripes * - (stripe_nr_end - stripe_nr_orig), - map->num_stripes); else if (mirror_num) stripe_index += mirror_num - 1; else { @@ -5587,7 +5800,7 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, (op == BTRFS_MAP_WRITE || op == BTRFS_MAP_GET_READ_MIRRORS || mirror_num > 1)) { /* push stripe_nr back to the start of the full stripe */ - stripe_nr = div_u64(raid56_full_stripe_start, + stripe_nr = div64_u64(raid56_full_stripe_start, stripe_len * nr_data_stripes(map)); /* RAID[56] write or recovery. Return all stripes */ @@ -5612,8 +5825,9 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, /* We distribute the parity blocks across stripes */ div_u64_rem(stripe_nr + stripe_index, map->num_stripes, &stripe_index); - if ((op != BTRFS_MAP_WRITE && op != BTRFS_MAP_DISCARD && - op != BTRFS_MAP_GET_READ_MIRRORS) && mirror_num <= 1) + if ((op != BTRFS_MAP_WRITE && + op != BTRFS_MAP_GET_READ_MIRRORS) && + mirror_num <= 1) mirror_num = 1; } } else { @@ -5635,8 +5849,8 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, } num_alloc_stripes = num_stripes; - if (dev_replace_is_ongoing) { - if (op == BTRFS_MAP_WRITE || op == BTRFS_MAP_DISCARD) + if (dev_replace_is_ongoing && dev_replace->tgtdev != NULL) { + if (op == BTRFS_MAP_WRITE) num_alloc_stripes <<= 1; if (op == BTRFS_MAP_GET_READ_MIRRORS) num_alloc_stripes++; @@ -5648,14 +5862,12 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, ret = -ENOMEM; goto out; } - if (dev_replace_is_ongoing) + if (dev_replace_is_ongoing && dev_replace->tgtdev != NULL) bbio->tgtdev_map = (int *)(bbio->stripes + num_alloc_stripes); /* build raid_map */ - if (map->type & BTRFS_BLOCK_GROUP_RAID56_MASK && - need_raid_map && - ((op == BTRFS_MAP_WRITE || op == BTRFS_MAP_GET_READ_MIRRORS) || - mirror_num > 1)) { + if (map->type & BTRFS_BLOCK_GROUP_RAID56_MASK && need_raid_map && + (need_full_stripe(op) || mirror_num > 1)) { u64 tmp; unsigned rot; @@ -5679,173 +5891,27 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, RAID6_Q_STRIPE; } - if (op == BTRFS_MAP_DISCARD) { - u32 factor = 0; - u32 sub_stripes = 0; - u64 stripes_per_dev = 0; - u32 remaining_stripes = 0; - u32 last_stripe = 0; - if (map->type & - (BTRFS_BLOCK_GROUP_RAID0 | BTRFS_BLOCK_GROUP_RAID10)) { - if (map->type & BTRFS_BLOCK_GROUP_RAID0) - sub_stripes = 1; - else - sub_stripes = map->sub_stripes; - - factor = map->num_stripes / sub_stripes; - stripes_per_dev = div_u64_rem(stripe_nr_end - - stripe_nr_orig, - factor, - &remaining_stripes); - div_u64_rem(stripe_nr_end - 1, factor, &last_stripe); - last_stripe *= sub_stripes; - } - - for (i = 0; i < num_stripes; i++) { - bbio->stripes[i].physical = - map->stripes[stripe_index].physical + - stripe_offset + stripe_nr * map->stripe_len; - bbio->stripes[i].dev = map->stripes[stripe_index].dev; - - if (map->type & (BTRFS_BLOCK_GROUP_RAID0 | - BTRFS_BLOCK_GROUP_RAID10)) { - bbio->stripes[i].length = stripes_per_dev * - map->stripe_len; - - if (i / sub_stripes < remaining_stripes) - bbio->stripes[i].length += - map->stripe_len; - - /* - * Special for the first stripe and - * the last stripe: - * - * |-------|...|-------| - * |----------| - * off end_off - */ - if (i < sub_stripes) - bbio->stripes[i].length -= - stripe_offset; - - if (stripe_index >= last_stripe && - stripe_index <= (last_stripe + - sub_stripes - 1)) - bbio->stripes[i].length -= - stripe_end_offset; - - if (i == sub_stripes - 1) - stripe_offset = 0; - } else - bbio->stripes[i].length = *length; - - stripe_index++; - if (stripe_index == map->num_stripes) { - /* This could only happen for RAID0/10 */ - stripe_index = 0; - stripe_nr++; - } - } - } else { - for (i = 0; i < num_stripes; i++) { - bbio->stripes[i].physical = - map->stripes[stripe_index].physical + - stripe_offset + - stripe_nr * map->stripe_len; - bbio->stripes[i].dev = - map->stripes[stripe_index].dev; - stripe_index++; - } + for (i = 0; i < num_stripes; i++) { + bbio->stripes[i].physical = + map->stripes[stripe_index].physical + + stripe_offset + + stripe_nr * map->stripe_len; + bbio->stripes[i].dev = + map->stripes[stripe_index].dev; + stripe_index++; } - if (op == BTRFS_MAP_WRITE || op == BTRFS_MAP_GET_READ_MIRRORS) + if (need_full_stripe(op)) max_errors = btrfs_chunk_max_errors(map); if (bbio->raid_map) sort_parity_stripes(bbio, num_stripes); - tgtdev_indexes = 0; - if (dev_replace_is_ongoing && - (op == BTRFS_MAP_WRITE || op == BTRFS_MAP_DISCARD) && - dev_replace->tgtdev != NULL) { - int index_where_to_add; - u64 srcdev_devid = dev_replace->srcdev->devid; - - /* - * duplicate the write operations while the dev replace - * procedure is running. Since the copying of the old disk - * to the new disk takes place at run time while the - * filesystem is mounted writable, the regular write - * operations to the old disk have to be duplicated to go - * to the new disk as well. - * Note that device->missing is handled by the caller, and - * that the write to the old disk is already set up in the - * stripes array. - */ - index_where_to_add = num_stripes; - for (i = 0; i < num_stripes; i++) { - if (bbio->stripes[i].dev->devid == srcdev_devid) { - /* write to new disk, too */ - struct btrfs_bio_stripe *new = - bbio->stripes + index_where_to_add; - struct btrfs_bio_stripe *old = - bbio->stripes + i; - - new->physical = old->physical; - new->length = old->length; - new->dev = dev_replace->tgtdev; - bbio->tgtdev_map[i] = index_where_to_add; - index_where_to_add++; - max_errors++; - tgtdev_indexes++; - } - } - num_stripes = index_where_to_add; - } else if (dev_replace_is_ongoing && - op == BTRFS_MAP_GET_READ_MIRRORS && - dev_replace->tgtdev != NULL) { - u64 srcdev_devid = dev_replace->srcdev->devid; - int index_srcdev = 0; - int found = 0; - u64 physical_of_found = 0; - - /* - * During the dev-replace procedure, the target drive can - * also be used to read data in case it is needed to repair - * a corrupt block elsewhere. This is possible if the - * requested area is left of the left cursor. In this area, - * the target drive is a full copy of the source drive. - */ - for (i = 0; i < num_stripes; i++) { - if (bbio->stripes[i].dev->devid == srcdev_devid) { - /* - * In case of DUP, in order to keep it - * simple, only add the mirror with the - * lowest physical address - */ - if (found && - physical_of_found <= - bbio->stripes[i].physical) - continue; - index_srcdev = i; - found = 1; - physical_of_found = bbio->stripes[i].physical; - } - } - if (found) { - struct btrfs_bio_stripe *tgtdev_stripe = - bbio->stripes + num_stripes; - - tgtdev_stripe->physical = physical_of_found; - tgtdev_stripe->length = - bbio->stripes[index_srcdev].length; - tgtdev_stripe->dev = dev_replace->tgtdev; - bbio->tgtdev_map[index_srcdev] = num_stripes; - - tgtdev_indexes++; - num_stripes++; - } + if (dev_replace_is_ongoing && dev_replace->tgtdev != NULL && + need_full_stripe(op)) { + handle_ops_on_dev_replace(op, &bbio, dev_replace, &num_stripes, + &max_errors); } *bbio_ret = bbio; @@ -5853,7 +5919,6 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, bbio->num_stripes = num_stripes; bbio->max_errors = max_errors; bbio->mirror_num = mirror_num; - bbio->num_tgtdevs = tgtdev_indexes; /* * this is the case that REQ_READ && dev_replace_is_ongoing && @@ -5886,19 +5951,15 @@ int btrfs_map_block(struct btrfs_fs_info *fs_info, enum btrfs_map_op op, /* For Scrub/replace */ int btrfs_map_sblock(struct btrfs_fs_info *fs_info, enum btrfs_map_op op, u64 logical, u64 *length, - struct btrfs_bio **bbio_ret, int mirror_num, - int need_raid_map) + struct btrfs_bio **bbio_ret) { - return __btrfs_map_block(fs_info, op, logical, length, bbio_ret, - mirror_num, need_raid_map); + return __btrfs_map_block(fs_info, op, logical, length, bbio_ret, 0, 1); } int btrfs_rmap_block(struct btrfs_fs_info *fs_info, u64 chunk_start, u64 physical, u64 devid, u64 **logical, int *naddrs, int *stripe_len) { - struct btrfs_mapping_tree *map_tree = &fs_info->mapping_tree; - struct extent_map_tree *em_tree = &map_tree->map_tree; struct extent_map *em; struct map_lookup *map; u64 *buf; @@ -5908,24 +5969,11 @@ int btrfs_rmap_block(struct btrfs_fs_info *fs_info, u64 rmap_len; int i, j, nr = 0; - read_lock(&em_tree->lock); - em = lookup_extent_mapping(em_tree, chunk_start, 1); - read_unlock(&em_tree->lock); - - if (!em) { - btrfs_err(fs_info, "couldn't find em for chunk %Lu", - chunk_start); + em = get_chunk_map(fs_info, chunk_start, 1); + if (IS_ERR(em)) return -EIO; - } - if (em->start != chunk_start) { - btrfs_err(fs_info, "bad chunk start, em=%Lu, wanted=%Lu", - em->start, chunk_start); - free_extent_map(em); - return -EIO; - } map = em->map_lookup; - length = em->len; rmap_len = map->stripe_len; @@ -5949,7 +5997,7 @@ int btrfs_rmap_block(struct btrfs_fs_info *fs_info, continue; stripe_nr = physical - map->stripes[i].physical; - stripe_nr = div_u64(stripe_nr, map->stripe_len); + stripe_nr = div64_u64(stripe_nr, map->stripe_len); if (map->type & BTRFS_BLOCK_GROUP_RAID10) { stripe_nr = stripe_nr * map->num_stripes + i; diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h index 59be81206dd7..c7d0fbc915ca 100644 --- a/fs/btrfs/volumes.h +++ b/fs/btrfs/volumes.h @@ -123,7 +123,6 @@ struct btrfs_device { struct list_head resized_list; /* for sending down flush barriers */ - int nobarriers; struct bio *flush_bio; struct completion flush_wait; @@ -298,7 +297,7 @@ struct btrfs_bio; typedef void (btrfs_bio_end_io_t) (struct btrfs_bio *bio, int err); struct btrfs_bio { - atomic_t refs; + refcount_t refs; atomic_t stripes_pending; struct btrfs_fs_info *fs_info; u64 map_type; /* get from map_lookup->type */ @@ -400,8 +399,7 @@ int btrfs_map_block(struct btrfs_fs_info *fs_info, enum btrfs_map_op op, struct btrfs_bio **bbio_ret, int mirror_num); int btrfs_map_sblock(struct btrfs_fs_info *fs_info, enum btrfs_map_op op, u64 logical, u64 *length, - struct btrfs_bio **bbio_ret, int mirror_num, - int need_raid_map); + struct btrfs_bio **bbio_ret); int btrfs_rmap_block(struct btrfs_fs_info *fs_info, u64 chunk_start, u64 physical, u64 devid, u64 **logical, int *naddrs, int *stripe_len); @@ -475,7 +473,7 @@ void btrfs_destroy_dev_replace_tgtdev(struct btrfs_fs_info *fs_info, void btrfs_init_dev_replace_tgtdev_for_resume(struct btrfs_fs_info *fs_info, struct btrfs_device *tgtdev); void btrfs_scratch_superblocks(struct block_device *bdev, const char *device_path); -int btrfs_is_parity_mirror(struct btrfs_mapping_tree *map_tree, +int btrfs_is_parity_mirror(struct btrfs_fs_info *fs_info, u64 logical, u64 len, int mirror_num); unsigned long btrfs_full_stripe_len(struct btrfs_fs_info *fs_info, struct btrfs_mapping_tree *map_tree, diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c index 9ecb2fd348cb..1e71e6ca5ddf 100644 --- a/fs/ceph/addr.c +++ b/fs/ceph/addr.c @@ -670,8 +670,12 @@ static void writepages_finish(struct ceph_osd_request *req) bool remove_page; dout("writepages_finish %p rc %d\n", inode, rc); - if (rc < 0) + if (rc < 0) { mapping_set_error(mapping, rc); + ceph_set_error_write(ci); + } else { + ceph_clear_error_write(ci); + } /* * We lost the cache cap, need to truncate the page before @@ -703,9 +707,6 @@ static void writepages_finish(struct ceph_osd_request *req) clear_bdi_congested(inode_to_bdi(inode), BLK_RW_ASYNC); - if (rc < 0) - SetPageError(page); - ceph_put_snap_context(page_snap_context(page)); page->private = 0; ClearPagePrivate(page); @@ -1892,6 +1893,7 @@ static int __ceph_pool_perm_get(struct ceph_inode_info *ci, err = ceph_osdc_start_request(&fsc->client->osdc, rd_req, false); wr_req->r_mtime = ci->vfs_inode.i_mtime; + wr_req->r_abort_on_full = true; err2 = ceph_osdc_start_request(&fsc->client->osdc, wr_req, false); if (!err) diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c index 68c78be19d5b..a3ebb632294e 100644 --- a/fs/ceph/caps.c +++ b/fs/ceph/caps.c @@ -1015,6 +1015,7 @@ static int send_cap_msg(struct cap_msg_args *arg) void *p; size_t extra_len; struct timespec zerotime = {0}; + struct ceph_osd_client *osdc = &arg->session->s_mdsc->fsc->client->osdc; dout("send_cap_msg %s %llx %llx caps %s wanted %s dirty %s" " seq %u/%u tid %llu/%llu mseq %u follows %lld size %llu/%llu" @@ -1076,8 +1077,12 @@ static int send_cap_msg(struct cap_msg_args *arg) ceph_encode_64(&p, arg->inline_data ? 0 : CEPH_INLINE_NONE); /* inline data size */ ceph_encode_32(&p, 0); - /* osd_epoch_barrier (version 5) */ - ceph_encode_32(&p, 0); + /* + * osd_epoch_barrier (version 5) + * The epoch_barrier is protected osdc->lock, so READ_ONCE here in + * case it was recently changed + */ + ceph_encode_32(&p, READ_ONCE(osdc->epoch_barrier)); /* oldest_flush_tid (version 6) */ ceph_encode_64(&p, arg->oldest_flush_tid); @@ -1389,7 +1394,7 @@ static void __ceph_flush_snaps(struct ceph_inode_info *ci, first_tid = cf->tid + 1; capsnap = container_of(cf, struct ceph_cap_snap, cap_flush); - atomic_inc(&capsnap->nref); + refcount_inc(&capsnap->nref); spin_unlock(&ci->i_ceph_lock); dout("__flush_snaps %p capsnap %p tid %llu %s\n", @@ -2202,7 +2207,7 @@ static void __kick_flushing_caps(struct ceph_mds_client *mdsc, inode, capsnap, cf->tid, ceph_cap_string(capsnap->dirty)); - atomic_inc(&capsnap->nref); + refcount_inc(&capsnap->nref); spin_unlock(&ci->i_ceph_lock); ret = __send_flush_snap(inode, session, capsnap, cap->mseq, @@ -3633,13 +3638,19 @@ void ceph_handle_caps(struct ceph_mds_session *session, p += inline_len; } + if (le16_to_cpu(msg->hdr.version) >= 5) { + struct ceph_osd_client *osdc = &mdsc->fsc->client->osdc; + u32 epoch_barrier; + + ceph_decode_32_safe(&p, end, epoch_barrier, bad); + ceph_osdc_update_epoch_barrier(osdc, epoch_barrier); + } + if (le16_to_cpu(msg->hdr.version) >= 8) { u64 flush_tid; u32 caller_uid, caller_gid; - u32 osd_epoch_barrier; u32 pool_ns_len; - /* version >= 5 */ - ceph_decode_32_safe(&p, end, osd_epoch_barrier, bad); + /* version >= 6 */ ceph_decode_64_safe(&p, end, flush_tid, bad); /* version >= 7 */ diff --git a/fs/ceph/debugfs.c b/fs/ceph/debugfs.c index 3ef11bc8d728..4e2d112c982f 100644 --- a/fs/ceph/debugfs.c +++ b/fs/ceph/debugfs.c @@ -22,20 +22,19 @@ static int mdsmap_show(struct seq_file *s, void *p) { int i; struct ceph_fs_client *fsc = s->private; + struct ceph_mdsmap *mdsmap; if (fsc->mdsc == NULL || fsc->mdsc->mdsmap == NULL) return 0; - seq_printf(s, "epoch %d\n", fsc->mdsc->mdsmap->m_epoch); - seq_printf(s, "root %d\n", fsc->mdsc->mdsmap->m_root); - seq_printf(s, "session_timeout %d\n", - fsc->mdsc->mdsmap->m_session_timeout); - seq_printf(s, "session_autoclose %d\n", - fsc->mdsc->mdsmap->m_session_autoclose); - for (i = 0; i < fsc->mdsc->mdsmap->m_max_mds; i++) { - struct ceph_entity_addr *addr = - &fsc->mdsc->mdsmap->m_info[i].addr; - int state = fsc->mdsc->mdsmap->m_info[i].state; - + mdsmap = fsc->mdsc->mdsmap; + seq_printf(s, "epoch %d\n", mdsmap->m_epoch); + seq_printf(s, "root %d\n", mdsmap->m_root); + seq_printf(s, "max_mds %d\n", mdsmap->m_max_mds); + seq_printf(s, "session_timeout %d\n", mdsmap->m_session_timeout); + seq_printf(s, "session_autoclose %d\n", mdsmap->m_session_autoclose); + for (i = 0; i < mdsmap->m_num_mds; i++) { + 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_mds_state_name(state)); diff --git a/fs/ceph/dir.c b/fs/ceph/dir.c index 3e9ad501addf..e071d23f6148 100644 --- a/fs/ceph/dir.c +++ b/fs/ceph/dir.c @@ -294,7 +294,7 @@ static int ceph_readdir(struct file *file, struct dir_context *ctx) struct ceph_mds_client *mdsc = fsc->mdsc; int i; int err; - u32 ftype; + unsigned frag = -1; struct ceph_mds_reply_info_parsed *rinfo; dout("readdir %p file %p pos %llx\n", inode, file, ctx->pos); @@ -341,7 +341,6 @@ more: /* do we have the correct frag content buffered? */ if (need_send_readdir(fi, ctx->pos)) { struct ceph_mds_request *req; - unsigned frag; int op = ceph_snap(inode) == CEPH_SNAPDIR ? CEPH_MDS_OP_LSSNAP : CEPH_MDS_OP_READDIR; @@ -352,8 +351,11 @@ more: } if (is_hash_order(ctx->pos)) { - frag = ceph_choose_frag(ci, fpos_hash(ctx->pos), - NULL, NULL); + /* fragtree isn't always accurate. choose frag + * based on previous reply when possible. */ + if (frag == (unsigned)-1) + frag = ceph_choose_frag(ci, fpos_hash(ctx->pos), + NULL, NULL); } else { frag = fpos_frag(ctx->pos); } @@ -378,7 +380,11 @@ more: ceph_mdsc_put_request(req); return -ENOMEM; } + } else if (is_hash_order(ctx->pos)) { + req->r_args.readdir.offset_hash = + cpu_to_le32(fpos_hash(ctx->pos)); } + req->r_dir_release_cnt = fi->dir_release_count; req->r_dir_ordered_cnt = fi->dir_ordered_count; req->r_readdir_cache_idx = fi->readdir_cache_idx; @@ -476,6 +482,7 @@ more: struct ceph_mds_reply_dir_entry *rde = rinfo->dir_entries + i; struct ceph_vino vino; ino_t ino; + u32 ftype; BUG_ON(rde->offset < ctx->pos); @@ -498,15 +505,17 @@ more: ctx->pos++; } + ceph_mdsc_put_request(fi->last_readdir); + fi->last_readdir = NULL; + if (fi->next_offset > 2) { - ceph_mdsc_put_request(fi->last_readdir); - fi->last_readdir = NULL; + frag = fi->frag; goto more; } /* more frags? */ if (!ceph_frag_is_rightmost(fi->frag)) { - unsigned frag = ceph_frag_next(fi->frag); + frag = ceph_frag_next(fi->frag); if (is_hash_order(ctx->pos)) { loff_t new_pos = ceph_make_fpos(ceph_frag_value(frag), fi->next_offset, true); diff --git a/fs/ceph/file.c b/fs/ceph/file.c index 18c045e2ead6..3fdde0b283c9 100644 --- a/fs/ceph/file.c +++ b/fs/ceph/file.c @@ -13,6 +13,38 @@ #include "mds_client.h" #include "cache.h" +static __le32 ceph_flags_sys2wire(u32 flags) +{ + u32 wire_flags = 0; + + switch (flags & O_ACCMODE) { + case O_RDONLY: + wire_flags |= CEPH_O_RDONLY; + break; + case O_WRONLY: + wire_flags |= CEPH_O_WRONLY; + break; + case O_RDWR: + wire_flags |= CEPH_O_RDWR; + break; + } + +#define ceph_sys2wire(a) if (flags & a) { wire_flags |= CEPH_##a; flags &= ~a; } + + ceph_sys2wire(O_CREAT); + ceph_sys2wire(O_EXCL); + ceph_sys2wire(O_TRUNC); + ceph_sys2wire(O_DIRECTORY); + ceph_sys2wire(O_NOFOLLOW); + +#undef ceph_sys2wire + + if (flags) + dout("unused open flags: %x", flags); + + return cpu_to_le32(wire_flags); +} + /* * Ceph file operations * @@ -120,7 +152,7 @@ prepare_open_request(struct super_block *sb, int flags, int create_mode) if (IS_ERR(req)) goto out; req->r_fmode = ceph_flags_to_mode(flags); - req->r_args.open.flags = cpu_to_le32(flags); + req->r_args.open.flags = ceph_flags_sys2wire(flags); req->r_args.open.mode = cpu_to_le32(create_mode); out: return req; @@ -189,7 +221,7 @@ int ceph_renew_caps(struct inode *inode) spin_lock(&ci->i_ceph_lock); wanted = __ceph_caps_file_wanted(ci); if (__ceph_is_any_real_caps(ci) && - (!(wanted & CEPH_CAP_ANY_WR) == 0 || ci->i_auth_cap)) { + (!(wanted & CEPH_CAP_ANY_WR) || ci->i_auth_cap)) { int issued = __ceph_caps_issued(ci, NULL); spin_unlock(&ci->i_ceph_lock); dout("renew caps %p want %s issued %s updating mds_wanted\n", @@ -778,6 +810,7 @@ static void ceph_aio_retry_work(struct work_struct *work) req->r_callback = ceph_aio_complete_req; req->r_inode = inode; req->r_priv = aio_req; + req->r_abort_on_full = true; ret = ceph_osdc_start_request(req->r_osdc, req, false); out: @@ -1085,19 +1118,22 @@ ceph_sync_write(struct kiocb *iocb, struct iov_iter *from, loff_t pos, out: ceph_osdc_put_request(req); - if (ret == 0) { - pos += len; - written += len; - - if (pos > i_size_read(inode)) { - check_caps = ceph_inode_set_size(inode, pos); - if (check_caps) - ceph_check_caps(ceph_inode(inode), - CHECK_CAPS_AUTHONLY, - NULL); - } - } else + if (ret != 0) { + ceph_set_error_write(ci); break; + } + + ceph_clear_error_write(ci); + pos += len; + written += len; + if (pos > i_size_read(inode)) { + check_caps = ceph_inode_set_size(inode, pos); + if (check_caps) + ceph_check_caps(ceph_inode(inode), + CHECK_CAPS_AUTHONLY, + NULL); + } + } if (ret != -EOLDSNAPC && written > 0) { @@ -1303,6 +1339,7 @@ static ssize_t ceph_write_iter(struct kiocb *iocb, struct iov_iter *from) } retry_snap: + /* FIXME: not complete since it doesn't account for being at quota */ if (ceph_osdmap_flag(osdc, CEPH_OSDMAP_FULL)) { err = -ENOSPC; goto out; @@ -1324,7 +1361,8 @@ retry_snap: inode, ceph_vinop(inode), pos, count, ceph_cap_string(got)); if ((got & (CEPH_CAP_FILE_BUFFER|CEPH_CAP_FILE_LAZYIO)) == 0 || - (iocb->ki_flags & IOCB_DIRECT) || (fi->flags & CEPH_F_SYNC)) { + (iocb->ki_flags & IOCB_DIRECT) || (fi->flags & CEPH_F_SYNC) || + (ci->i_ceph_flags & CEPH_I_ERROR_WRITE)) { struct ceph_snap_context *snapc; struct iov_iter data; inode_unlock(inode); diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c index d3119fe3ab45..dcce79b84406 100644 --- a/fs/ceph/inode.c +++ b/fs/ceph/inode.c @@ -1482,10 +1482,17 @@ int ceph_readdir_prepopulate(struct ceph_mds_request *req, if (test_bit(CEPH_MDS_R_ABORTED, &req->r_req_flags)) return readdir_prepopulate_inodes_only(req, session); - if (rinfo->hash_order && req->r_path2) { - last_hash = ceph_str_hash(ci->i_dir_layout.dl_dir_hash, - req->r_path2, strlen(req->r_path2)); - last_hash = ceph_frag_value(last_hash); + if (rinfo->hash_order) { + if (req->r_path2) { + last_hash = ceph_str_hash(ci->i_dir_layout.dl_dir_hash, + req->r_path2, + strlen(req->r_path2)); + last_hash = ceph_frag_value(last_hash); + } else if (rinfo->offset_hash) { + /* mds understands offset_hash */ + WARN_ON_ONCE(req->r_readdir_offset != 2); + last_hash = le32_to_cpu(rhead->args.readdir.offset_hash); + } } if (rinfo->dir_dir && @@ -1510,7 +1517,7 @@ int ceph_readdir_prepopulate(struct ceph_mds_request *req, } if (ceph_frag_is_leftmost(frag) && req->r_readdir_offset == 2 && - !(rinfo->hash_order && req->r_path2)) { + !(rinfo->hash_order && last_hash)) { /* note dir version at start of readdir so we can tell * if any dentries get dropped */ req->r_dir_release_cnt = atomic64_read(&ci->i_release_count); diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c index 1d3fa90d40b9..f38e56fa9712 100644 --- a/fs/ceph/mds_client.c +++ b/fs/ceph/mds_client.c @@ -189,6 +189,7 @@ static int parse_reply_info_dir(void **p, void *end, info->dir_end = !!(flags & CEPH_READDIR_FRAG_END); info->dir_complete = !!(flags & CEPH_READDIR_FRAG_COMPLETE); info->hash_order = !!(flags & CEPH_READDIR_HASH_ORDER); + info->offset_hash = !!(flags & CEPH_READDIR_OFFSET_HASH); } if (num == 0) goto done; @@ -378,9 +379,9 @@ const char *ceph_session_state_name(int s) static struct ceph_mds_session *get_session(struct ceph_mds_session *s) { - if (atomic_inc_not_zero(&s->s_ref)) { + if (refcount_inc_not_zero(&s->s_ref)) { dout("mdsc get_session %p %d -> %d\n", s, - atomic_read(&s->s_ref)-1, atomic_read(&s->s_ref)); + refcount_read(&s->s_ref)-1, refcount_read(&s->s_ref)); return s; } else { dout("mdsc get_session %p 0 -- FAIL", s); @@ -391,8 +392,8 @@ static struct ceph_mds_session *get_session(struct ceph_mds_session *s) void ceph_put_mds_session(struct ceph_mds_session *s) { dout("mdsc put_session %p %d -> %d\n", s, - atomic_read(&s->s_ref), atomic_read(&s->s_ref)-1); - if (atomic_dec_and_test(&s->s_ref)) { + refcount_read(&s->s_ref), refcount_read(&s->s_ref)-1); + if (refcount_dec_and_test(&s->s_ref)) { if (s->s_auth.authorizer) ceph_auth_destroy_authorizer(s->s_auth.authorizer); kfree(s); @@ -411,7 +412,7 @@ struct ceph_mds_session *__ceph_lookup_mds_session(struct ceph_mds_client *mdsc, return NULL; session = mdsc->sessions[mds]; dout("lookup_mds_session %p %d\n", session, - atomic_read(&session->s_ref)); + refcount_read(&session->s_ref)); get_session(session); return session; } @@ -441,7 +442,7 @@ static struct ceph_mds_session *register_session(struct ceph_mds_client *mdsc, { struct ceph_mds_session *s; - if (mds >= mdsc->mdsmap->m_max_mds) + if (mds >= mdsc->mdsmap->m_num_mds) return ERR_PTR(-EINVAL); s = kzalloc(sizeof(*s), GFP_NOFS); @@ -466,7 +467,7 @@ static struct ceph_mds_session *register_session(struct ceph_mds_client *mdsc, INIT_LIST_HEAD(&s->s_caps); s->s_nr_caps = 0; s->s_trim_caps = 0; - atomic_set(&s->s_ref, 1); + refcount_set(&s->s_ref, 1); INIT_LIST_HEAD(&s->s_waiting); INIT_LIST_HEAD(&s->s_unsafe); s->s_num_cap_releases = 0; @@ -494,7 +495,7 @@ static struct ceph_mds_session *register_session(struct ceph_mds_client *mdsc, } mdsc->sessions[mds] = s; atomic_inc(&mdsc->num_sessions); - atomic_inc(&s->s_ref); /* one ref to sessions[], one to caller */ + refcount_inc(&s->s_ref); /* one ref to sessions[], one to caller */ ceph_con_open(&s->s_con, CEPH_ENTITY_TYPE_MDS, mds, ceph_mdsmap_get_addr(mdsc->mdsmap, mds)); @@ -1004,7 +1005,7 @@ static void __open_export_target_sessions(struct ceph_mds_client *mdsc, struct ceph_mds_session *ts; int i, mds = session->s_mds; - if (mds >= mdsc->mdsmap->m_max_mds) + if (mds >= mdsc->mdsmap->m_num_mds) return; mi = &mdsc->mdsmap->m_info[mds]; @@ -1551,9 +1552,15 @@ void ceph_send_cap_releases(struct ceph_mds_client *mdsc, struct ceph_msg *msg = NULL; struct ceph_mds_cap_release *head; struct ceph_mds_cap_item *item; + struct ceph_osd_client *osdc = &mdsc->fsc->client->osdc; struct ceph_cap *cap; LIST_HEAD(tmp_list); int num_cap_releases; + __le32 barrier, *cap_barrier; + + down_read(&osdc->lock); + barrier = cpu_to_le32(osdc->epoch_barrier); + up_read(&osdc->lock); spin_lock(&session->s_cap_lock); again: @@ -1571,7 +1578,11 @@ again: head = msg->front.iov_base; head->num = cpu_to_le32(0); msg->front.iov_len = sizeof(*head); + + msg->hdr.version = cpu_to_le16(2); + msg->hdr.compat_version = cpu_to_le16(1); } + cap = list_first_entry(&tmp_list, struct ceph_cap, session_caps); list_del(&cap->session_caps); @@ -1589,6 +1600,11 @@ again: ceph_put_cap(mdsc, cap); if (le32_to_cpu(head->num) == CEPH_CAPS_PER_RELEASE) { + // Append cap_barrier field + cap_barrier = msg->front.iov_base + msg->front.iov_len; + *cap_barrier = barrier; + msg->front.iov_len += sizeof(*cap_barrier); + msg->hdr.front_len = cpu_to_le32(msg->front.iov_len); dout("send_cap_releases mds%d %p\n", session->s_mds, msg); ceph_con_send(&session->s_con, msg); @@ -1604,6 +1620,11 @@ again: spin_unlock(&session->s_cap_lock); if (msg) { + // Append cap_barrier field + cap_barrier = msg->front.iov_base + msg->front.iov_len; + *cap_barrier = barrier; + msg->front.iov_len += sizeof(*cap_barrier); + msg->hdr.front_len = cpu_to_le32(msg->front.iov_len); dout("send_cap_releases mds%d %p\n", session->s_mds, msg); ceph_con_send(&session->s_con, msg); @@ -1993,7 +2014,7 @@ static struct ceph_msg *create_request_message(struct ceph_mds_client *mdsc, if (req->r_pagelist) { struct ceph_pagelist *pagelist = req->r_pagelist; - atomic_inc(&pagelist->refcnt); + refcount_inc(&pagelist->refcnt); ceph_msg_data_add_pagelist(msg, pagelist); msg->hdr.data_len = cpu_to_le32(pagelist->length); } else { @@ -2640,8 +2661,10 @@ static void handle_session(struct ceph_mds_session *session, seq = le64_to_cpu(h->seq); mutex_lock(&mdsc->mutex); - if (op == CEPH_SESSION_CLOSE) + if (op == CEPH_SESSION_CLOSE) { + get_session(session); __unregister_session(mdsc, session); + } /* FIXME: this ttl calculation is generous */ session->s_ttl = jiffies + HZ*mdsc->mdsmap->m_session_autoclose; mutex_unlock(&mdsc->mutex); @@ -2730,6 +2753,8 @@ static void handle_session(struct ceph_mds_session *session, kick_requests(mdsc, mds); mutex_unlock(&mdsc->mutex); } + if (op == CEPH_SESSION_CLOSE) + ceph_put_mds_session(session); return; bad: @@ -3109,7 +3134,7 @@ static void check_new_map(struct ceph_mds_client *mdsc, dout("check_new_map new %u old %u\n", newmap->m_epoch, oldmap->m_epoch); - for (i = 0; i < oldmap->m_max_mds && i < mdsc->max_sessions; i++) { + for (i = 0; i < oldmap->m_num_mds && i < mdsc->max_sessions; i++) { if (mdsc->sessions[i] == NULL) continue; s = mdsc->sessions[i]; @@ -3123,15 +3148,33 @@ static void check_new_map(struct ceph_mds_client *mdsc, ceph_mdsmap_is_laggy(newmap, i) ? " (laggy)" : "", ceph_session_state_name(s->s_state)); - if (i >= newmap->m_max_mds || + if (i >= newmap->m_num_mds || memcmp(ceph_mdsmap_get_addr(oldmap, i), ceph_mdsmap_get_addr(newmap, i), sizeof(struct ceph_entity_addr))) { if (s->s_state == CEPH_MDS_SESSION_OPENING) { /* the session never opened, just close it * out now */ + get_session(s); + __unregister_session(mdsc, s); __wake_requests(mdsc, &s->s_waiting); + ceph_put_mds_session(s); + } else if (i >= newmap->m_num_mds) { + /* force close session for stopped mds */ + get_session(s); __unregister_session(mdsc, s); + __wake_requests(mdsc, &s->s_waiting); + kick_requests(mdsc, i); + mutex_unlock(&mdsc->mutex); + + mutex_lock(&s->s_mutex); + cleanup_session_requests(mdsc, s); + remove_session_caps(s); + mutex_unlock(&s->s_mutex); + + ceph_put_mds_session(s); + + mutex_lock(&mdsc->mutex); } else { /* just close it */ mutex_unlock(&mdsc->mutex); @@ -3169,7 +3212,7 @@ static void check_new_map(struct ceph_mds_client *mdsc, } } - for (i = 0; i < newmap->m_max_mds && i < mdsc->max_sessions; i++) { + for (i = 0; i < newmap->m_num_mds && i < mdsc->max_sessions; i++) { s = mdsc->sessions[i]; if (!s) continue; @@ -3883,7 +3926,7 @@ static struct ceph_connection *con_get(struct ceph_connection *con) struct ceph_mds_session *s = con->private; if (get_session(s)) { - dout("mdsc con_get %p ok (%d)\n", s, atomic_read(&s->s_ref)); + dout("mdsc con_get %p ok (%d)\n", s, refcount_read(&s->s_ref)); return con; } dout("mdsc con_get %p FAIL\n", s); @@ -3894,7 +3937,7 @@ static void con_put(struct ceph_connection *con) { struct ceph_mds_session *s = con->private; - dout("mdsc con_put %p (%d)\n", s, atomic_read(&s->s_ref) - 1); + dout("mdsc con_put %p (%d)\n", s, refcount_read(&s->s_ref) - 1); ceph_put_mds_session(s); } diff --git a/fs/ceph/mds_client.h b/fs/ceph/mds_client.h index ac0475a2daa7..db57ae98ed34 100644 --- a/fs/ceph/mds_client.h +++ b/fs/ceph/mds_client.h @@ -7,6 +7,7 @@ #include <linux/mutex.h> #include <linux/rbtree.h> #include <linux/spinlock.h> +#include <linux/refcount.h> #include <linux/ceph/types.h> #include <linux/ceph/messenger.h> @@ -82,9 +83,10 @@ struct ceph_mds_reply_info_parsed { struct ceph_mds_reply_dirfrag *dir_dir; size_t dir_buf_size; int dir_nr; - bool dir_complete; bool dir_end; + bool dir_complete; bool hash_order; + bool offset_hash; struct ceph_mds_reply_dir_entry *dir_entries; }; @@ -104,10 +106,13 @@ struct ceph_mds_reply_info_parsed { /* * cap releases are batched and sent to the MDS en masse. + * + * Account for per-message overhead of mds_cap_release header + * and __le32 for osd epoch barrier trailing field. */ -#define CEPH_CAPS_PER_RELEASE ((PAGE_SIZE - \ +#define CEPH_CAPS_PER_RELEASE ((PAGE_SIZE - sizeof(u32) - \ sizeof(struct ceph_mds_cap_release)) / \ - sizeof(struct ceph_mds_cap_item)) + sizeof(struct ceph_mds_cap_item)) /* @@ -156,7 +161,7 @@ struct ceph_mds_session { unsigned long s_renew_requested; /* last time we sent a renew req */ u64 s_renew_seq; - atomic_t s_ref; + refcount_t s_ref; struct list_head s_waiting; /* waiting requests */ struct list_head s_unsafe; /* unsafe requests */ }; @@ -373,7 +378,7 @@ __ceph_lookup_mds_session(struct ceph_mds_client *, int mds); static inline struct ceph_mds_session * ceph_get_mds_session(struct ceph_mds_session *s) { - atomic_inc(&s->s_ref); + refcount_inc(&s->s_ref); return s; } diff --git a/fs/ceph/mdsmap.c b/fs/ceph/mdsmap.c index 5454e2327a5f..1a748cf88535 100644 --- a/fs/ceph/mdsmap.c +++ b/fs/ceph/mdsmap.c @@ -22,11 +22,11 @@ int ceph_mdsmap_get_random_mds(struct ceph_mdsmap *m) int i; /* special case for one mds */ - if (1 == m->m_max_mds && m->m_info[0].state > 0) + if (1 == m->m_num_mds && m->m_info[0].state > 0) return 0; /* count */ - for (i = 0; i < m->m_max_mds; i++) + for (i = 0; i < m->m_num_mds; i++) if (m->m_info[i].state > 0) n++; if (n == 0) @@ -135,8 +135,9 @@ struct ceph_mdsmap *ceph_mdsmap_decode(void **p, void *end) m->m_session_autoclose = ceph_decode_32(p); m->m_max_file_size = ceph_decode_64(p); m->m_max_mds = ceph_decode_32(p); + m->m_num_mds = m->m_max_mds; - m->m_info = kcalloc(m->m_max_mds, sizeof(*m->m_info), GFP_NOFS); + m->m_info = kcalloc(m->m_num_mds, sizeof(*m->m_info), GFP_NOFS); if (m->m_info == NULL) goto nomem; @@ -207,9 +208,20 @@ struct ceph_mdsmap *ceph_mdsmap_decode(void **p, void *end) ceph_pr_addr(&addr.in_addr), ceph_mds_state_name(state)); - if (mds < 0 || mds >= m->m_max_mds || state <= 0) + if (mds < 0 || state <= 0) continue; + if (mds >= m->m_num_mds) { + int new_num = max(mds + 1, m->m_num_mds * 2); + void *new_m_info = krealloc(m->m_info, + new_num * sizeof(*m->m_info), + GFP_NOFS | __GFP_ZERO); + if (!new_m_info) + goto nomem; + m->m_info = new_m_info; + m->m_num_mds = new_num; + } + info = &m->m_info[mds]; info->global_id = global_id; info->state = state; @@ -229,6 +241,14 @@ struct ceph_mdsmap *ceph_mdsmap_decode(void **p, void *end) info->export_targets = NULL; } } + if (m->m_num_mds > m->m_max_mds) { + /* find max up mds */ + for (i = m->m_num_mds; i >= m->m_max_mds; i--) { + if (i == 0 || m->m_info[i-1].state > 0) + break; + } + m->m_num_mds = i; + } /* pg_pools */ ceph_decode_32_safe(p, end, n, bad); @@ -270,12 +290,22 @@ struct ceph_mdsmap *ceph_mdsmap_decode(void **p, void *end) for (i = 0; i < n; i++) { s32 mds = ceph_decode_32(p); - if (mds >= 0 && mds < m->m_max_mds) { + if (mds >= 0 && mds < m->m_num_mds) { if (m->m_info[mds].laggy) num_laggy++; } } m->m_num_laggy = num_laggy; + + if (n > m->m_num_mds) { + void *new_m_info = krealloc(m->m_info, + n * sizeof(*m->m_info), + GFP_NOFS | __GFP_ZERO); + if (!new_m_info) + goto nomem; + m->m_info = new_m_info; + } + m->m_num_mds = n; } /* inc */ @@ -341,7 +371,7 @@ void ceph_mdsmap_destroy(struct ceph_mdsmap *m) { int i; - for (i = 0; i < m->m_max_mds; i++) + for (i = 0; i < m->m_num_mds; i++) kfree(m->m_info[i].export_targets); kfree(m->m_info); kfree(m->m_data_pg_pools); @@ -357,7 +387,7 @@ bool ceph_mdsmap_is_cluster_available(struct ceph_mdsmap *m) return false; if (m->m_num_laggy > 0) return false; - for (i = 0; i < m->m_max_mds; i++) { + for (i = 0; i < m->m_num_mds; i++) { if (m->m_info[i].state == CEPH_MDS_STATE_ACTIVE) nr_active++; } diff --git a/fs/ceph/snap.c b/fs/ceph/snap.c index 8f8b41c2ef0f..dab5d6732345 100644 --- a/fs/ceph/snap.c +++ b/fs/ceph/snap.c @@ -519,7 +519,7 @@ void ceph_queue_cap_snap(struct ceph_inode_info *ci) capsnap->need_flush ? "" : "no_flush"); ihold(inode); - atomic_set(&capsnap->nref, 1); + refcount_set(&capsnap->nref, 1); INIT_LIST_HEAD(&capsnap->ci_item); capsnap->follows = old_snapc->seq; diff --git a/fs/ceph/super.c b/fs/ceph/super.c index a8c81b2052ca..8d7918ce694a 100644 --- a/fs/ceph/super.c +++ b/fs/ceph/super.c @@ -544,10 +544,6 @@ static struct ceph_fs_client *create_fs_client(struct ceph_mount_options *fsopt, struct ceph_options *opt) { struct ceph_fs_client *fsc; - const u64 supported_features = - CEPH_FEATURE_FLOCK | CEPH_FEATURE_DIRLAYOUTHASH | - CEPH_FEATURE_MDSENC | CEPH_FEATURE_MDS_INLINE_DATA; - const u64 required_features = 0; int page_count; size_t size; int err = -ENOMEM; @@ -556,8 +552,7 @@ static struct ceph_fs_client *create_fs_client(struct ceph_mount_options *fsopt, if (!fsc) return ERR_PTR(-ENOMEM); - fsc->client = ceph_create_client(opt, fsc, supported_features, - required_features); + fsc->client = ceph_create_client(opt, fsc); if (IS_ERR(fsc->client)) { err = PTR_ERR(fsc->client); goto fail; diff --git a/fs/ceph/super.h b/fs/ceph/super.h index 176186b12457..a973acd8beaf 100644 --- a/fs/ceph/super.h +++ b/fs/ceph/super.h @@ -14,6 +14,7 @@ #include <linux/writeback.h> #include <linux/slab.h> #include <linux/posix_acl.h> +#include <linux/refcount.h> #include <linux/ceph/libceph.h> @@ -160,7 +161,7 @@ struct ceph_cap_flush { * data before flushing the snapped state (tracked here) back to the MDS. */ struct ceph_cap_snap { - atomic_t nref; + refcount_t nref; struct list_head ci_item; struct ceph_cap_flush cap_flush; @@ -189,7 +190,7 @@ struct ceph_cap_snap { static inline void ceph_put_cap_snap(struct ceph_cap_snap *capsnap) { - if (atomic_dec_and_test(&capsnap->nref)) { + if (refcount_dec_and_test(&capsnap->nref)) { if (capsnap->xattr_blob) ceph_buffer_put(capsnap->xattr_blob); kfree(capsnap); @@ -471,6 +472,32 @@ static inline struct inode *ceph_find_inode(struct super_block *sb, #define CEPH_I_CAP_DROPPED (1 << 8) /* caps were forcibly dropped */ #define CEPH_I_KICK_FLUSH (1 << 9) /* kick flushing caps */ #define CEPH_I_FLUSH_SNAPS (1 << 10) /* need flush snapss */ +#define CEPH_I_ERROR_WRITE (1 << 11) /* have seen write errors */ + +/* + * We set the ERROR_WRITE bit when we start seeing write errors on an inode + * and then clear it when they start succeeding. Note that we do a lockless + * check first, and only take the lock if it looks like it needs to be changed. + * The write submission code just takes this as a hint, so we're not too + * worried if a few slip through in either direction. + */ +static inline void ceph_set_error_write(struct ceph_inode_info *ci) +{ + if (!(READ_ONCE(ci->i_ceph_flags) & CEPH_I_ERROR_WRITE)) { + spin_lock(&ci->i_ceph_lock); + ci->i_ceph_flags |= CEPH_I_ERROR_WRITE; + spin_unlock(&ci->i_ceph_lock); + } +} + +static inline void ceph_clear_error_write(struct ceph_inode_info *ci) +{ + if (READ_ONCE(ci->i_ceph_flags) & CEPH_I_ERROR_WRITE) { + spin_lock(&ci->i_ceph_lock); + ci->i_ceph_flags &= ~CEPH_I_ERROR_WRITE; + spin_unlock(&ci->i_ceph_lock); + } +} static inline void __ceph_dir_set_complete(struct ceph_inode_info *ci, long long release_count, diff --git a/fs/ceph/xattr.c b/fs/ceph/xattr.c index febc28f9e2c2..75267cdd5dfd 100644 --- a/fs/ceph/xattr.c +++ b/fs/ceph/xattr.c @@ -392,6 +392,7 @@ static int __set_xattr(struct ceph_inode_info *ci, if (update_xattr) { int err = 0; + if (xattr && (flags & XATTR_CREATE)) err = -EEXIST; else if (!xattr && (flags & XATTR_REPLACE)) @@ -399,12 +400,14 @@ static int __set_xattr(struct ceph_inode_info *ci, if (err) { kfree(name); kfree(val); + kfree(*newxattr); return err; } if (update_xattr < 0) { if (xattr) __remove_xattr(ci, xattr); kfree(name); + kfree(*newxattr); return 0; } } @@ -461,35 +461,6 @@ int dax_delete_mapping_entry(struct address_space *mapping, pgoff_t index) } /* - * Invalidate exceptional DAX entry if easily possible. This handles DAX - * entries for invalidate_inode_pages() so we evict the entry only if we can - * do so without blocking. - */ -int dax_invalidate_mapping_entry(struct address_space *mapping, pgoff_t index) -{ - int ret = 0; - void *entry, **slot; - struct radix_tree_root *page_tree = &mapping->page_tree; - - spin_lock_irq(&mapping->tree_lock); - entry = __radix_tree_lookup(page_tree, index, NULL, &slot); - if (!entry || !radix_tree_exceptional_entry(entry) || - slot_locked(mapping, slot)) - goto out; - if (radix_tree_tag_get(page_tree, index, PAGECACHE_TAG_DIRTY) || - radix_tree_tag_get(page_tree, index, PAGECACHE_TAG_TOWRITE)) - goto out; - radix_tree_delete(page_tree, index); - mapping->nrexceptional--; - ret = 1; -out: - spin_unlock_irq(&mapping->tree_lock); - if (ret) - dax_wake_mapping_entry_waiter(mapping, index, entry, true); - return ret; -} - -/* * Invalidate exceptional DAX entry if it is clean. */ int dax_invalidate_mapping_entry_sync(struct address_space *mapping, @@ -993,12 +964,12 @@ int __dax_zero_page_range(struct block_device *bdev, void *kaddr; pfn_t pfn; - rc = bdev_dax_pgoff(bdev, sector, size, &pgoff); + rc = bdev_dax_pgoff(bdev, sector, PAGE_SIZE, &pgoff); if (rc) return rc; id = dax_read_lock(); - rc = dax_direct_access(dax_dev, pgoff, PHYS_PFN(size), &kaddr, + rc = dax_direct_access(dax_dev, pgoff, 1, &kaddr, &pfn); if (rc < 0) { dax_read_unlock(id); @@ -1044,7 +1015,7 @@ dax_iomap_actor(struct inode *inode, loff_t pos, loff_t length, void *data, * into page tables. We have to tear down these mappings so that data * written by write(2) is visible in mmap. */ - if ((iomap->flags & IOMAP_F_NEW) && inode->i_mapping->nrpages) { + if (iomap->flags & IOMAP_F_NEW) { invalidate_inode_pages2_range(inode->i_mapping, pos >> PAGE_SHIFT, (end - 1) >> PAGE_SHIFT); @@ -1177,6 +1148,12 @@ static int dax_iomap_pte_fault(struct vm_fault *vmf, if ((vmf->flags & FAULT_FLAG_WRITE) && !vmf->cow_page) flags |= IOMAP_WRITE; + entry = grab_mapping_entry(mapping, vmf->pgoff, 0); + if (IS_ERR(entry)) { + vmf_ret = dax_fault_return(PTR_ERR(entry)); + goto out; + } + /* * Note that we don't bother to use iomap_apply here: DAX required * the file system block size to be equal the page size, which means @@ -1185,17 +1162,11 @@ static int dax_iomap_pte_fault(struct vm_fault *vmf, error = ops->iomap_begin(inode, pos, PAGE_SIZE, flags, &iomap); if (error) { vmf_ret = dax_fault_return(error); - goto out; + goto unlock_entry; } if (WARN_ON_ONCE(iomap.offset + iomap.length < pos + PAGE_SIZE)) { - vmf_ret = dax_fault_return(-EIO); /* fs corruption? */ - goto finish_iomap; - } - - entry = grab_mapping_entry(mapping, vmf->pgoff, 0); - if (IS_ERR(entry)) { - vmf_ret = dax_fault_return(PTR_ERR(entry)); - goto finish_iomap; + error = -EIO; /* fs corruption? */ + goto error_finish_iomap; } sector = dax_iomap_sector(&iomap, pos); @@ -1217,13 +1188,13 @@ static int dax_iomap_pte_fault(struct vm_fault *vmf, } if (error) - goto error_unlock_entry; + goto error_finish_iomap; __SetPageUptodate(vmf->cow_page); vmf_ret = finish_fault(vmf); if (!vmf_ret) vmf_ret = VM_FAULT_DONE_COW; - goto unlock_entry; + goto finish_iomap; } switch (iomap.type) { @@ -1243,7 +1214,7 @@ static int dax_iomap_pte_fault(struct vm_fault *vmf, case IOMAP_HOLE: if (!(vmf->flags & FAULT_FLAG_WRITE)) { vmf_ret = dax_load_hole(mapping, &entry, vmf); - goto unlock_entry; + goto finish_iomap; } /*FALLTHRU*/ default: @@ -1252,10 +1223,8 @@ static int dax_iomap_pte_fault(struct vm_fault *vmf, break; } - error_unlock_entry: + error_finish_iomap: vmf_ret = dax_fault_return(error) | major; - unlock_entry: - put_locked_mapping_entry(mapping, vmf->pgoff, entry); finish_iomap: if (ops->iomap_end) { int copied = PAGE_SIZE; @@ -1270,7 +1239,9 @@ static int dax_iomap_pte_fault(struct vm_fault *vmf, */ ops->iomap_end(inode, pos, PAGE_SIZE, copied, flags, &iomap); } -out: + unlock_entry: + put_locked_mapping_entry(mapping, vmf->pgoff, entry); + out: trace_dax_pte_fault_done(inode, vmf, vmf_ret); return vmf_ret; } @@ -1417,6 +1388,16 @@ static int dax_iomap_pmd_fault(struct vm_fault *vmf, goto fallback; /* + * grab_mapping_entry() will make sure we get a 2M empty entry, a DAX + * PMD or a HZP entry. If it can't (because a 4k page is already in + * the tree, for instance), it will return -EEXIST and we just fall + * back to 4k entries. + */ + entry = grab_mapping_entry(mapping, pgoff, RADIX_DAX_PMD); + if (IS_ERR(entry)) + goto fallback; + + /* * Note that we don't use iomap_apply here. We aren't doing I/O, only * setting up a mapping, so really we're using iomap_begin() as a way * to look up our filesystem block. @@ -1424,21 +1405,11 @@ static int dax_iomap_pmd_fault(struct vm_fault *vmf, pos = (loff_t)pgoff << PAGE_SHIFT; error = ops->iomap_begin(inode, pos, PMD_SIZE, iomap_flags, &iomap); if (error) - goto fallback; + goto unlock_entry; if (iomap.offset + iomap.length < pos + PMD_SIZE) goto finish_iomap; - /* - * grab_mapping_entry() will make sure we get a 2M empty entry, a DAX - * PMD or a HZP entry. If it can't (because a 4k page is already in - * the tree, for instance), it will return -EEXIST and we just fall - * back to 4k entries. - */ - entry = grab_mapping_entry(mapping, pgoff, RADIX_DAX_PMD); - if (IS_ERR(entry)) - goto finish_iomap; - switch (iomap.type) { case IOMAP_MAPPED: result = dax_pmd_insert_mapping(vmf, &iomap, pos, &entry); @@ -1446,7 +1417,7 @@ static int dax_iomap_pmd_fault(struct vm_fault *vmf, case IOMAP_UNWRITTEN: case IOMAP_HOLE: if (WARN_ON_ONCE(write)) - goto unlock_entry; + break; result = dax_pmd_load_hole(vmf, &iomap, &entry); break; default: @@ -1454,8 +1425,6 @@ static int dax_iomap_pmd_fault(struct vm_fault *vmf, break; } - unlock_entry: - put_locked_mapping_entry(mapping, pgoff, entry); finish_iomap: if (ops->iomap_end) { int copied = PMD_SIZE; @@ -1471,6 +1440,8 @@ static int dax_iomap_pmd_fault(struct vm_fault *vmf, ops->iomap_end(inode, pos, PMD_SIZE, copied, iomap_flags, &iomap); } + unlock_entry: + put_locked_mapping_entry(mapping, pgoff, entry); fallback: if (result == VM_FAULT_FALLBACK) { split_huge_pmd(vma, vmf->pmd, vmf->address); diff --git a/fs/ext2/super.c b/fs/ext2/super.c index 8ac673c71a36..9c2028b50e5c 100644 --- a/fs/ext2/super.c +++ b/fs/ext2/super.c @@ -32,6 +32,7 @@ #include <linux/log2.h> #include <linux/quotaops.h> #include <linux/uaccess.h> +#include <linux/dax.h> #include "ext2.h" #include "xattr.h" #include "acl.h" diff --git a/fs/ext4/file.c b/fs/ext4/file.c index cefa9835f275..831fd6beebf0 100644 --- a/fs/ext4/file.c +++ b/fs/ext4/file.c @@ -257,6 +257,7 @@ static int ext4_dax_huge_fault(struct vm_fault *vmf, enum page_entry_size pe_size) { int result; + handle_t *handle = NULL; struct inode *inode = file_inode(vmf->vma->vm_file); struct super_block *sb = inode->i_sb; bool write = vmf->flags & FAULT_FLAG_WRITE; @@ -264,12 +265,24 @@ static int ext4_dax_huge_fault(struct vm_fault *vmf, if (write) { sb_start_pagefault(sb); file_update_time(vmf->vma->vm_file); + down_read(&EXT4_I(inode)->i_mmap_sem); + handle = ext4_journal_start_sb(sb, EXT4_HT_WRITE_PAGE, + EXT4_DATA_TRANS_BLOCKS(sb)); + } else { + down_read(&EXT4_I(inode)->i_mmap_sem); } - down_read(&EXT4_I(inode)->i_mmap_sem); - result = dax_iomap_fault(vmf, pe_size, &ext4_iomap_ops); - up_read(&EXT4_I(inode)->i_mmap_sem); - if (write) + if (!IS_ERR(handle)) + result = dax_iomap_fault(vmf, pe_size, &ext4_iomap_ops); + else + result = VM_FAULT_SIGBUS; + if (write) { + if (!IS_ERR(handle)) + ext4_journal_stop(handle); + up_read(&EXT4_I(inode)->i_mmap_sem); sb_end_pagefault(sb); + } else { + up_read(&EXT4_I(inode)->i_mmap_sem); + } return result; } diff --git a/fs/ext4/super.c b/fs/ext4/super.c index c90edf09b0c3..0b177da9ea82 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -37,6 +37,7 @@ #include <linux/ctype.h> #include <linux/log2.h> #include <linux/crc16.h> +#include <linux/dax.h> #include <linux/cleancache.h> #include <linux/uaccess.h> diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index c2d7f3a92679..c16d00e53264 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -20,6 +20,7 @@ #include <linux/pipe_fs_i.h> #include <linux/swap.h> #include <linux/splice.h> +#include <linux/sched.h> MODULE_ALIAS_MISCDEV(FUSE_MINOR); MODULE_ALIAS("devname:fuse"); @@ -45,7 +46,7 @@ static void fuse_request_init(struct fuse_req *req, struct page **pages, INIT_LIST_HEAD(&req->list); INIT_LIST_HEAD(&req->intr_entry); init_waitqueue_head(&req->waitq); - atomic_set(&req->count, 1); + refcount_set(&req->count, 1); req->pages = pages; req->page_descs = page_descs; req->max_pages = npages; @@ -102,21 +103,20 @@ void fuse_request_free(struct fuse_req *req) void __fuse_get_request(struct fuse_req *req) { - atomic_inc(&req->count); + refcount_inc(&req->count); } /* Must be called with > 1 refcount */ static void __fuse_put_request(struct fuse_req *req) { - BUG_ON(atomic_read(&req->count) < 2); - atomic_dec(&req->count); + refcount_dec(&req->count); } -static void fuse_req_init_context(struct fuse_req *req) +static void fuse_req_init_context(struct fuse_conn *fc, struct fuse_req *req) { req->in.h.uid = from_kuid_munged(&init_user_ns, current_fsuid()); req->in.h.gid = from_kgid_munged(&init_user_ns, current_fsgid()); - req->in.h.pid = current->pid; + req->in.h.pid = pid_nr_ns(task_pid(current), fc->pid_ns); } void fuse_set_initialized(struct fuse_conn *fc) @@ -163,7 +163,7 @@ static struct fuse_req *__fuse_get_req(struct fuse_conn *fc, unsigned npages, goto out; } - fuse_req_init_context(req); + fuse_req_init_context(fc, req); __set_bit(FR_WAITING, &req->flags); if (for_background) __set_bit(FR_BACKGROUND, &req->flags); @@ -256,7 +256,7 @@ struct fuse_req *fuse_get_req_nofail_nopages(struct fuse_conn *fc, if (!req) req = get_reserved_req(fc, file); - fuse_req_init_context(req); + fuse_req_init_context(fc, req); __set_bit(FR_WAITING, &req->flags); __clear_bit(FR_BACKGROUND, &req->flags); return req; @@ -264,7 +264,7 @@ struct fuse_req *fuse_get_req_nofail_nopages(struct fuse_conn *fc, void fuse_put_request(struct fuse_conn *fc, struct fuse_req *req) { - if (atomic_dec_and_test(&req->count)) { + if (refcount_dec_and_test(&req->count)) { if (test_bit(FR_BACKGROUND, &req->flags)) { /* * We get here in the unlikely case that a background @@ -1222,6 +1222,9 @@ static ssize_t fuse_dev_do_read(struct fuse_dev *fud, struct file *file, struct fuse_in *in; unsigned reqsize; + if (task_active_pid_ns(current) != fc->pid_ns) + return -EIO; + restart: spin_lock(&fiq->waitq.lock); err = -EAGAIN; @@ -1820,6 +1823,9 @@ static ssize_t fuse_dev_do_write(struct fuse_dev *fud, struct fuse_req *req; struct fuse_out_header oh; + if (task_active_pid_ns(current) != fc->pid_ns) + return -EIO; + if (nbytes < sizeof(struct fuse_out_header)) return -EINVAL; diff --git a/fs/fuse/file.c b/fs/fuse/file.c index ec238fb5a584..3ee4fdc3da9e 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -58,7 +58,7 @@ struct fuse_file *fuse_file_alloc(struct fuse_conn *fc) } INIT_LIST_HEAD(&ff->write_entry); - atomic_set(&ff->count, 1); + refcount_set(&ff->count, 1); RB_CLEAR_NODE(&ff->polled_node); init_waitqueue_head(&ff->poll_wait); @@ -77,7 +77,7 @@ void fuse_file_free(struct fuse_file *ff) static struct fuse_file *fuse_file_get(struct fuse_file *ff) { - atomic_inc(&ff->count); + refcount_inc(&ff->count); return ff; } @@ -88,7 +88,7 @@ static void fuse_release_end(struct fuse_conn *fc, struct fuse_req *req) static void fuse_file_put(struct fuse_file *ff, bool sync) { - if (atomic_dec_and_test(&ff->count)) { + if (refcount_dec_and_test(&ff->count)) { struct fuse_req *req = ff->reserved_req; if (ff->fc->no_open) { @@ -293,7 +293,7 @@ static int fuse_release(struct inode *inode, struct file *file) void fuse_sync_release(struct fuse_file *ff, int flags) { - WARN_ON(atomic_read(&ff->count) != 1); + WARN_ON(refcount_read(&ff->count) > 1); fuse_prepare_release(ff, flags, FUSE_RELEASE); /* * iput(NULL) is a no-op and since the refcount is 1 and everything's @@ -2083,7 +2083,8 @@ static int fuse_direct_mmap(struct file *file, struct vm_area_struct *vma) return generic_file_mmap(file, vma); } -static int convert_fuse_file_lock(const struct fuse_file_lock *ffl, +static int convert_fuse_file_lock(struct fuse_conn *fc, + const struct fuse_file_lock *ffl, struct file_lock *fl) { switch (ffl->type) { @@ -2098,7 +2099,14 @@ static int convert_fuse_file_lock(const struct fuse_file_lock *ffl, fl->fl_start = ffl->start; fl->fl_end = ffl->end; - fl->fl_pid = ffl->pid; + + /* + * Convert pid into the caller's pid namespace. If the pid + * does not map into the namespace fl_pid will get set to 0. + */ + rcu_read_lock(); + fl->fl_pid = pid_vnr(find_pid_ns(ffl->pid, fc->pid_ns)); + rcu_read_unlock(); break; default: @@ -2147,7 +2155,7 @@ static int fuse_getlk(struct file *file, struct file_lock *fl) args.out.args[0].value = &outarg; err = fuse_simple_request(fc, &args); if (!err) - err = convert_fuse_file_lock(&outarg.lk, fl); + err = convert_fuse_file_lock(fc, &outarg.lk, fl); return err; } @@ -2159,7 +2167,8 @@ static int fuse_setlk(struct file *file, struct file_lock *fl, int flock) FUSE_ARGS(args); struct fuse_lk_in inarg; int opcode = (fl->fl_flags & FL_SLEEP) ? FUSE_SETLKW : FUSE_SETLK; - pid_t pid = fl->fl_type != F_UNLCK ? current->tgid : 0; + struct pid *pid = fl->fl_type != F_UNLCK ? task_tgid(current) : NULL; + pid_t pid_nr = pid_nr_ns(pid, fc->pid_ns); int err; if (fl->fl_lmops && fl->fl_lmops->lm_grant) { @@ -2168,10 +2177,13 @@ static int fuse_setlk(struct file *file, struct file_lock *fl, int flock) } /* Unlock on close is handled by the flush method */ - if (fl->fl_flags & FL_CLOSE) + if ((fl->fl_flags & FL_CLOSE_POSIX) == FL_CLOSE_POSIX) return 0; - fuse_lk_fill(&args, file, fl, opcode, pid, flock, &inarg); + if (pid && pid_nr == 0) + return -EOVERFLOW; + + fuse_lk_fill(&args, file, fl, opcode, pid_nr, flock, &inarg); err = fuse_simple_request(fc, &args); /* locking is restartable */ diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h index f33341d9501a..1bd7ffdad593 100644 --- a/fs/fuse/fuse_i.h +++ b/fs/fuse/fuse_i.h @@ -24,6 +24,8 @@ #include <linux/workqueue.h> #include <linux/kref.h> #include <linux/xattr.h> +#include <linux/pid_namespace.h> +#include <linux/refcount.h> /** Max number of pages that can be used in a single read request */ #define FUSE_MAX_PAGES_PER_REQ 32 @@ -137,7 +139,7 @@ struct fuse_file { u64 nodeid; /** Refcount */ - atomic_t count; + refcount_t count; /** FOPEN_* flags returned by open */ u32 open_flags; @@ -306,7 +308,7 @@ struct fuse_req { struct list_head intr_entry; /** refcount */ - atomic_t count; + refcount_t count; /** Unique ID for the interrupt request */ u64 intr_unique; @@ -448,7 +450,7 @@ struct fuse_conn { spinlock_t lock; /** Refcount */ - atomic_t count; + refcount_t count; /** Number of fuse_dev's */ atomic_t dev_count; @@ -461,6 +463,9 @@ struct fuse_conn { /** The group id for this mount */ kgid_t group_id; + /** The pid namespace for this mount */ + struct pid_namespace *pid_ns; + /** Maximum read size */ unsigned max_read; diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index 73cf05135252..5a1b58f8fef4 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c @@ -21,6 +21,7 @@ #include <linux/sched.h> #include <linux/exportfs.h> #include <linux/posix_acl.h> +#include <linux/pid_namespace.h> MODULE_AUTHOR("Miklos Szeredi <miklos@szeredi.hu>"); MODULE_DESCRIPTION("Filesystem in Userspace"); @@ -601,7 +602,7 @@ void fuse_conn_init(struct fuse_conn *fc) memset(fc, 0, sizeof(*fc)); spin_lock_init(&fc->lock); init_rwsem(&fc->killsb); - atomic_set(&fc->count, 1); + refcount_set(&fc->count, 1); atomic_set(&fc->dev_count, 1); init_waitqueue_head(&fc->blocked_waitq); init_waitqueue_head(&fc->reserved_req_waitq); @@ -619,14 +620,16 @@ void fuse_conn_init(struct fuse_conn *fc) fc->connected = 1; fc->attr_version = 1; get_random_bytes(&fc->scramble_key, sizeof(fc->scramble_key)); + fc->pid_ns = get_pid_ns(task_active_pid_ns(current)); } EXPORT_SYMBOL_GPL(fuse_conn_init); void fuse_conn_put(struct fuse_conn *fc) { - if (atomic_dec_and_test(&fc->count)) { + if (refcount_dec_and_test(&fc->count)) { if (fc->destroy_req) fuse_request_free(fc->destroy_req); + put_pid_ns(fc->pid_ns); fc->release(fc); } } @@ -634,7 +637,7 @@ EXPORT_SYMBOL_GPL(fuse_conn_put); struct fuse_conn *fuse_conn_get(struct fuse_conn *fc) { - atomic_inc(&fc->count); + refcount_inc(&fc->count); return fc; } EXPORT_SYMBOL_GPL(fuse_conn_get); diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index 5a0245e36240..ebad34266bcf 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -2363,7 +2363,7 @@ static int jbd2_journal_init_journal_head_cache(void) jbd2_journal_head_cache = kmem_cache_create("jbd2_journal_head", sizeof(struct journal_head), 0, /* offset */ - SLAB_TEMPORARY | SLAB_DESTROY_BY_RCU, + SLAB_TEMPORARY | SLAB_TYPESAFE_BY_RCU, NULL); /* ctor */ retval = 0; if (!jbd2_journal_head_cache) { diff --git a/fs/jffs2/readinode.c b/fs/jffs2/readinode.c index 06a71dbd4833..389ea53ea487 100644 --- a/fs/jffs2/readinode.c +++ b/fs/jffs2/readinode.c @@ -1366,7 +1366,7 @@ int jffs2_do_read_inode(struct jffs2_sb_info *c, struct jffs2_inode_info *f, jffs2_add_ino_cache(c, f->inocache); } if (!f->inocache) { - JFFS2_ERROR("requestied to read an nonexistent ino %u\n", ino); + JFFS2_ERROR("requested to read a nonexistent ino %u\n", ino); return -ENOENT; } diff --git a/fs/lockd/clntlock.c b/fs/lockd/clntlock.c index 41e491b8e5d7..27d577dbe51a 100644 --- a/fs/lockd/clntlock.c +++ b/fs/lockd/clntlock.c @@ -69,6 +69,7 @@ struct nlm_host *nlmclnt_init(const struct nlmclnt_initdata *nlm_init) if (host->h_rpcclnt == NULL && nlm_bind_host(host) == NULL) goto out_nobind; + host->h_nlmclnt_ops = nlm_init->nlmclnt_ops; return host; out_nobind: nlmclnt_release_host(host); diff --git a/fs/lockd/clntproc.c b/fs/lockd/clntproc.c index 112952037933..066ac313ae5c 100644 --- a/fs/lockd/clntproc.c +++ b/fs/lockd/clntproc.c @@ -150,17 +150,22 @@ static void nlmclnt_release_lockargs(struct nlm_rqst *req) * @host: address of a valid nlm_host context representing the NLM server * @cmd: fcntl-style file lock operation to perform * @fl: address of arguments for the lock operation + * @data: address of data to be sent to callback operations * */ -int nlmclnt_proc(struct nlm_host *host, int cmd, struct file_lock *fl) +int nlmclnt_proc(struct nlm_host *host, int cmd, struct file_lock *fl, void *data) { struct nlm_rqst *call; int status; + const struct nlmclnt_operations *nlmclnt_ops = host->h_nlmclnt_ops; call = nlm_alloc_call(host); if (call == NULL) return -ENOMEM; + if (nlmclnt_ops && nlmclnt_ops->nlmclnt_alloc_call) + nlmclnt_ops->nlmclnt_alloc_call(data); + nlmclnt_locks_init_private(fl, host); if (!fl->fl_u.nfs_fl.owner) { /* lockowner allocation has failed */ @@ -169,6 +174,7 @@ int nlmclnt_proc(struct nlm_host *host, int cmd, struct file_lock *fl) } /* Set up the argument struct */ nlmclnt_setlockargs(call, fl); + call->a_callback_data = data; if (IS_SETLK(cmd) || IS_SETLKW(cmd)) { if (fl->fl_type != F_UNLCK) { @@ -214,8 +220,12 @@ struct nlm_rqst *nlm_alloc_call(struct nlm_host *host) void nlmclnt_release_call(struct nlm_rqst *call) { + const struct nlmclnt_operations *nlmclnt_ops = call->a_host->h_nlmclnt_ops; + if (!atomic_dec_and_test(&call->a_count)) return; + if (nlmclnt_ops && nlmclnt_ops->nlmclnt_release_call) + nlmclnt_ops->nlmclnt_release_call(call->a_callback_data); nlmclnt_release_host(call->a_host); nlmclnt_release_lockargs(call); kfree(call); @@ -687,6 +697,19 @@ out: return status; } +static void nlmclnt_unlock_prepare(struct rpc_task *task, void *data) +{ + struct nlm_rqst *req = data; + const struct nlmclnt_operations *nlmclnt_ops = req->a_host->h_nlmclnt_ops; + bool defer_call = false; + + if (nlmclnt_ops && nlmclnt_ops->nlmclnt_unlock_prepare) + defer_call = nlmclnt_ops->nlmclnt_unlock_prepare(task, req->a_callback_data); + + if (!defer_call) + rpc_call_start(task); +} + static void nlmclnt_unlock_callback(struct rpc_task *task, void *data) { struct nlm_rqst *req = data; @@ -720,6 +743,7 @@ die: } static const struct rpc_call_ops nlmclnt_unlock_ops = { + .rpc_call_prepare = nlmclnt_unlock_prepare, .rpc_call_done = nlmclnt_unlock_callback, .rpc_release = nlmclnt_rpc_release, }; diff --git a/fs/lockd/svc.c b/fs/lockd/svc.c index e7c8b9c76e48..5d481e8a1b5d 100644 --- a/fs/lockd/svc.c +++ b/fs/lockd/svc.c @@ -132,6 +132,8 @@ lockd(void *vrqstp) { int err = 0; struct svc_rqst *rqstp = vrqstp; + struct net *net = &init_net; + struct lockd_net *ln = net_generic(net, lockd_net_id); /* try_to_freeze() is called from svc_recv() */ set_freezable(); @@ -176,6 +178,8 @@ lockd(void *vrqstp) if (nlmsvc_ops) nlmsvc_invalidate_all(); nlm_shutdown_hosts(); + cancel_delayed_work_sync(&ln->grace_period_end); + locks_end_grace(&ln->lockd_manager); return 0; } @@ -270,8 +274,6 @@ static void lockd_down_net(struct svc_serv *serv, struct net *net) if (ln->nlmsvc_users) { if (--ln->nlmsvc_users == 0) { nlm_shutdown_hosts_net(net); - cancel_delayed_work_sync(&ln->grace_period_end); - locks_end_grace(&ln->lockd_manager); svc_shutdown_net(serv, net); dprintk("lockd_down_net: per-net data destroyed; net=%p\n", net); } diff --git a/fs/lockd/svclock.c b/fs/lockd/svclock.c index 5581e020644b..3507c80d1d4b 100644 --- a/fs/lockd/svclock.c +++ b/fs/lockd/svclock.c @@ -870,15 +870,15 @@ nlmsvc_grant_reply(struct nlm_cookie *cookie, __be32 status) if (!(block = nlmsvc_find_block(cookie))) return; - if (block) { - if (status == nlm_lck_denied_grace_period) { - /* Try again in a couple of seconds */ - nlmsvc_insert_block(block, 10 * HZ); - } else { - /* Lock is now held by client, or has been rejected. - * In both cases, the block should be removed. */ - nlmsvc_unlink_block(block); - } + if (status == nlm_lck_denied_grace_period) { + /* Try again in a couple of seconds */ + nlmsvc_insert_block(block, 10 * HZ); + } else { + /* + * Lock is now held by client, or has been rejected. + * In both cases, the block should be removed. + */ + nlmsvc_unlink_block(block); } nlmsvc_release_block(block); } diff --git a/fs/locks.c b/fs/locks.c index 26811321d39b..af2031a1fcff 100644 --- a/fs/locks.c +++ b/fs/locks.c @@ -2504,7 +2504,7 @@ locks_remove_flock(struct file *filp, struct file_lock_context *flctx) .fl_owner = filp, .fl_pid = current->tgid, .fl_file = filp, - .fl_flags = FL_FLOCK, + .fl_flags = FL_FLOCK | FL_CLOSE, .fl_type = F_UNLCK, .fl_end = OFFSET_MAX, }; diff --git a/fs/namei.c b/fs/namei.c index 7286f87ce863..6571a5f5112e 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -2142,7 +2142,6 @@ OK: static const char *path_init(struct nameidata *nd, unsigned flags) { - int retval = 0; const char *s = nd->name->name; if (!*s) @@ -2154,13 +2153,8 @@ static const char *path_init(struct nameidata *nd, unsigned flags) if (flags & LOOKUP_ROOT) { struct dentry *root = nd->root.dentry; struct inode *inode = root->d_inode; - if (*s) { - if (!d_can_lookup(root)) - return ERR_PTR(-ENOTDIR); - retval = inode_permission(inode, MAY_EXEC); - if (retval) - return ERR_PTR(retval); - } + if (*s && unlikely(!d_can_lookup(root))) + return ERR_PTR(-ENOTDIR); nd->path = nd->root; nd->inode = inode; if (flags & LOOKUP_RCU) { @@ -2258,6 +2252,35 @@ static inline int lookup_last(struct nameidata *nd) return walk_component(nd, 0); } +static int handle_lookup_down(struct nameidata *nd) +{ + struct path path = nd->path; + struct inode *inode = nd->inode; + unsigned seq = nd->seq; + int err; + + if (nd->flags & LOOKUP_RCU) { + /* + * don't bother with unlazy_walk on failure - we are + * at the very beginning of walk, so we lose nothing + * if we simply redo everything in non-RCU mode + */ + if (unlikely(!__follow_mount_rcu(nd, &path, &inode, &seq))) + return -ECHILD; + } else { + dget(path.dentry); + err = follow_managed(&path, nd); + if (unlikely(err < 0)) + return err; + inode = d_backing_inode(path.dentry); + seq = 0; + } + path_to_nameidata(&path, nd); + nd->inode = inode; + nd->seq = seq; + return 0; +} + /* Returns 0 and nd will be valid on success; Retuns error, otherwise. */ static int path_lookupat(struct nameidata *nd, unsigned flags, struct path *path) { @@ -2266,6 +2289,15 @@ static int path_lookupat(struct nameidata *nd, unsigned flags, struct path *path if (IS_ERR(s)) return PTR_ERR(s); + + if (unlikely(flags & LOOKUP_DOWN)) { + err = handle_lookup_down(nd); + if (unlikely(err < 0)) { + terminate_walk(nd); + return err; + } + } + while (!(err = link_path_walk(s, nd)) && ((err = lookup_last(nd)) > 0)) { s = trailing_symlink(nd); diff --git a/fs/namespace.c b/fs/namespace.c index b3b115bd4e1e..8bd3e4d448b9 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -3462,8 +3462,9 @@ static void mntns_put(struct ns_common *ns) static int mntns_install(struct nsproxy *nsproxy, struct ns_common *ns) { struct fs_struct *fs = current->fs; - struct mnt_namespace *mnt_ns = to_mnt_ns(ns); + struct mnt_namespace *mnt_ns = to_mnt_ns(ns), *old_mnt_ns; struct path root; + int err; if (!ns_capable(mnt_ns->user_ns, CAP_SYS_ADMIN) || !ns_capable(current_user_ns(), CAP_SYS_CHROOT) || @@ -3474,15 +3475,18 @@ static int mntns_install(struct nsproxy *nsproxy, struct ns_common *ns) return -EINVAL; get_mnt_ns(mnt_ns); - put_mnt_ns(nsproxy->mnt_ns); + old_mnt_ns = nsproxy->mnt_ns; nsproxy->mnt_ns = mnt_ns; /* Find the root */ - root.mnt = &mnt_ns->root->mnt; - root.dentry = mnt_ns->root->mnt.mnt_root; - path_get(&root); - while(d_mountpoint(root.dentry) && follow_down_one(&root)) - ; + err = vfs_path_lookup(mnt_ns->root->mnt.mnt_root, &mnt_ns->root->mnt, + "/", LOOKUP_DOWN, &root); + if (err) { + /* revert to old namespace */ + nsproxy->mnt_ns = old_mnt_ns; + put_mnt_ns(mnt_ns); + return err; + } /* Update the pwd and root */ set_fs_pwd(fs, &root); diff --git a/fs/nfs/Kconfig b/fs/nfs/Kconfig index f31fd0dd92c6..69d02cf8cf37 100644 --- a/fs/nfs/Kconfig +++ b/fs/nfs/Kconfig @@ -123,11 +123,6 @@ config PNFS_BLOCK depends on NFS_V4_1 && BLK_DEV_DM default NFS_V4 -config PNFS_OBJLAYOUT - tristate - depends on NFS_V4_1 && SCSI_OSD_ULD - default NFS_V4 - config PNFS_FLEXFILE_LAYOUT tristate depends on NFS_V4_1 && NFS_V3 diff --git a/fs/nfs/Makefile b/fs/nfs/Makefile index 6abdda209642..98f4e5728a67 100644 --- a/fs/nfs/Makefile +++ b/fs/nfs/Makefile @@ -31,6 +31,5 @@ nfsv4-$(CONFIG_NFS_V4_1) += pnfs.o pnfs_dev.o pnfs_nfs.o nfsv4-$(CONFIG_NFS_V4_2) += nfs42proc.o obj-$(CONFIG_PNFS_FILE_LAYOUT) += filelayout/ -obj-$(CONFIG_PNFS_OBJLAYOUT) += objlayout/ obj-$(CONFIG_PNFS_BLOCK) += blocklayout/ obj-$(CONFIG_PNFS_FLEXFILE_LAYOUT) += flexfilelayout/ diff --git a/fs/nfs/callback.c b/fs/nfs/callback.c index 773774531aff..73a1f928226c 100644 --- a/fs/nfs/callback.c +++ b/fs/nfs/callback.c @@ -76,7 +76,10 @@ nfs4_callback_svc(void *vrqstp) set_freezable(); - while (!kthread_should_stop()) { + while (!kthread_freezable_should_stop(NULL)) { + + if (signal_pending(current)) + flush_signals(current); /* * Listen for a request on the socket */ @@ -85,6 +88,8 @@ nfs4_callback_svc(void *vrqstp) continue; svc_process(rqstp); } + svc_exit_thread(rqstp); + module_put_and_exit(0); return 0; } @@ -103,9 +108,10 @@ nfs41_callback_svc(void *vrqstp) set_freezable(); - while (!kthread_should_stop()) { - if (try_to_freeze()) - continue; + while (!kthread_freezable_should_stop(NULL)) { + + if (signal_pending(current)) + flush_signals(current); prepare_to_wait(&serv->sv_cb_waitq, &wq, TASK_INTERRUPTIBLE); spin_lock_bh(&serv->sv_cb_lock); @@ -121,11 +127,13 @@ nfs41_callback_svc(void *vrqstp) error); } else { spin_unlock_bh(&serv->sv_cb_lock); - schedule(); + if (!kthread_should_stop()) + schedule(); finish_wait(&serv->sv_cb_waitq, &wq); } - flush_signals(current); } + svc_exit_thread(rqstp); + module_put_and_exit(0); return 0; } @@ -221,14 +229,14 @@ err_bind: static struct svc_serv_ops nfs40_cb_sv_ops = { .svo_function = nfs4_callback_svc, .svo_enqueue_xprt = svc_xprt_do_enqueue, - .svo_setup = svc_set_num_threads, + .svo_setup = svc_set_num_threads_sync, .svo_module = THIS_MODULE, }; #if defined(CONFIG_NFS_V4_1) static struct svc_serv_ops nfs41_cb_sv_ops = { .svo_function = nfs41_callback_svc, .svo_enqueue_xprt = svc_xprt_do_enqueue, - .svo_setup = svc_set_num_threads, + .svo_setup = svc_set_num_threads_sync, .svo_module = THIS_MODULE, }; @@ -280,7 +288,7 @@ static struct svc_serv *nfs_callback_create_svc(int minorversion) printk(KERN_WARNING "nfs_callback_create_svc: no kthread, %d users??\n", cb_info->users); - serv = svc_create(&nfs4_callback_program, NFS4_CALLBACK_BUFSIZE, sv_ops); + serv = svc_create_pooled(&nfs4_callback_program, NFS4_CALLBACK_BUFSIZE, sv_ops); if (!serv) { printk(KERN_ERR "nfs_callback_create_svc: create service failed\n"); return ERR_PTR(-ENOMEM); diff --git a/fs/nfs/callback_proc.c b/fs/nfs/callback_proc.c index f073a6d2c6a5..52479f180ea1 100644 --- a/fs/nfs/callback_proc.c +++ b/fs/nfs/callback_proc.c @@ -131,10 +131,11 @@ restart: if (!inode) continue; if (!nfs_sb_active(inode->i_sb)) { - rcu_read_lock(); + rcu_read_unlock(); spin_unlock(&clp->cl_lock); iput(inode); spin_lock(&clp->cl_lock); + rcu_read_lock(); goto restart; } return inode; @@ -170,10 +171,11 @@ restart: if (!inode) continue; if (!nfs_sb_active(inode->i_sb)) { - rcu_read_lock(); + rcu_read_unlock(); spin_unlock(&clp->cl_lock); iput(inode); spin_lock(&clp->cl_lock); + rcu_read_lock(); goto restart; } return inode; @@ -317,31 +319,18 @@ static u32 initiate_bulk_draining(struct nfs_client *clp, static u32 do_callback_layoutrecall(struct nfs_client *clp, struct cb_layoutrecallargs *args) { - u32 res; - - dprintk("%s enter, type=%i\n", __func__, args->cbl_recall_type); if (args->cbl_recall_type == RETURN_FILE) - res = initiate_file_draining(clp, args); - else - res = initiate_bulk_draining(clp, args); - dprintk("%s returning %i\n", __func__, res); - return res; - + return initiate_file_draining(clp, args); + return initiate_bulk_draining(clp, args); } __be32 nfs4_callback_layoutrecall(struct cb_layoutrecallargs *args, void *dummy, struct cb_process_state *cps) { - u32 res; - - dprintk("%s: -->\n", __func__); + u32 res = NFS4ERR_OP_NOT_IN_SESSION; if (cps->clp) res = do_callback_layoutrecall(cps->clp, args); - else - res = NFS4ERR_OP_NOT_IN_SESSION; - - dprintk("%s: exit with status = %d\n", __func__, res); return cpu_to_be32(res); } @@ -364,8 +353,6 @@ __be32 nfs4_callback_devicenotify(struct cb_devicenotifyargs *args, struct nfs_client *clp = cps->clp; struct nfs_server *server = NULL; - dprintk("%s: -->\n", __func__); - if (!clp) { res = cpu_to_be32(NFS4ERR_OP_NOT_IN_SESSION); goto out; @@ -384,8 +371,6 @@ __be32 nfs4_callback_devicenotify(struct cb_devicenotifyargs *args, goto found; } rcu_read_unlock(); - dprintk("%s: layout type %u not found\n", - __func__, dev->cbd_layout_type); continue; } @@ -395,8 +380,6 @@ __be32 nfs4_callback_devicenotify(struct cb_devicenotifyargs *args, out: kfree(args->devs); - dprintk("%s: exit with status = %u\n", - __func__, be32_to_cpu(res)); return res; } @@ -417,16 +400,11 @@ static __be32 validate_seqid(const struct nfs4_slot_table *tbl, const struct nfs4_slot *slot, const struct cb_sequenceargs * args) { - dprintk("%s enter. slotid %u seqid %u, slot table seqid: %u\n", - __func__, args->csa_slotid, args->csa_sequenceid, slot->seq_nr); - if (args->csa_slotid > tbl->server_highest_slotid) return htonl(NFS4ERR_BADSLOT); /* Replay */ if (args->csa_sequenceid == slot->seq_nr) { - dprintk("%s seqid %u is a replay\n", - __func__, args->csa_sequenceid); if (nfs4_test_locked_slot(tbl, slot->slot_nr)) return htonl(NFS4ERR_DELAY); /* Signal process_op to set this error on next op */ @@ -480,15 +458,6 @@ static bool referring_call_exists(struct nfs_client *clp, for (j = 0; j < rclist->rcl_nrefcalls; j++) { ref = &rclist->rcl_refcalls[j]; - - dprintk("%s: sessionid %x:%x:%x:%x sequenceid %u " - "slotid %u\n", __func__, - ((u32 *)&rclist->rcl_sessionid.data)[0], - ((u32 *)&rclist->rcl_sessionid.data)[1], - ((u32 *)&rclist->rcl_sessionid.data)[2], - ((u32 *)&rclist->rcl_sessionid.data)[3], - ref->rc_sequenceid, ref->rc_slotid); - status = nfs4_slot_wait_on_seqid(tbl, ref->rc_slotid, ref->rc_sequenceid, HZ >> 1) < 0; if (status) @@ -593,8 +562,6 @@ out: res->csr_status = status; trace_nfs4_cb_sequence(args, res, status); - dprintk("%s: exit with status = %d res->csr_status %d\n", __func__, - ntohl(status), ntohl(res->csr_status)); return status; } diff --git a/fs/nfs/callback_xdr.c b/fs/nfs/callback_xdr.c index d051fc3583a9..c14758e08d73 100644 --- a/fs/nfs/callback_xdr.c +++ b/fs/nfs/callback_xdr.c @@ -171,8 +171,6 @@ static __be32 decode_compound_hdr_arg(struct xdr_stream *xdr, struct cb_compound return htonl(NFS4ERR_MINOR_VERS_MISMATCH); } hdr->nops = ntohl(*p); - dprintk("%s: minorversion %d nops %d\n", __func__, - hdr->minorversion, hdr->nops); return 0; } @@ -192,11 +190,8 @@ static __be32 decode_getattr_args(struct svc_rqst *rqstp, struct xdr_stream *xdr status = decode_fh(xdr, &args->fh); if (unlikely(status != 0)) - goto out; - status = decode_bitmap(xdr, args->bitmap); -out: - dprintk("%s: exit with status = %d\n", __func__, ntohl(status)); - return status; + return status; + return decode_bitmap(xdr, args->bitmap); } static __be32 decode_recall_args(struct svc_rqst *rqstp, struct xdr_stream *xdr, struct cb_recallargs *args) @@ -206,17 +201,12 @@ static __be32 decode_recall_args(struct svc_rqst *rqstp, struct xdr_stream *xdr, status = decode_delegation_stateid(xdr, &args->stateid); if (unlikely(status != 0)) - goto out; + return status; p = read_buf(xdr, 4); - if (unlikely(p == NULL)) { - status = htonl(NFS4ERR_RESOURCE); - goto out; - } + if (unlikely(p == NULL)) + return htonl(NFS4ERR_RESOURCE); args->truncate = ntohl(*p); - status = decode_fh(xdr, &args->fh); -out: - dprintk("%s: exit with status = %d\n", __func__, ntohl(status)); - return status; + return decode_fh(xdr, &args->fh); } #if defined(CONFIG_NFS_V4_1) @@ -235,10 +225,8 @@ static __be32 decode_layoutrecall_args(struct svc_rqst *rqstp, uint32_t iomode; p = read_buf(xdr, 4 * sizeof(uint32_t)); - if (unlikely(p == NULL)) { - status = htonl(NFS4ERR_BADXDR); - goto out; - } + if (unlikely(p == NULL)) + return htonl(NFS4ERR_BADXDR); args->cbl_layout_type = ntohl(*p++); /* Depite the spec's xdr, iomode really belongs in the FILE switch, @@ -252,37 +240,23 @@ static __be32 decode_layoutrecall_args(struct svc_rqst *rqstp, args->cbl_range.iomode = iomode; status = decode_fh(xdr, &args->cbl_fh); if (unlikely(status != 0)) - goto out; + return status; p = read_buf(xdr, 2 * sizeof(uint64_t)); - if (unlikely(p == NULL)) { - status = htonl(NFS4ERR_BADXDR); - goto out; - } + if (unlikely(p == NULL)) + return htonl(NFS4ERR_BADXDR); p = xdr_decode_hyper(p, &args->cbl_range.offset); p = xdr_decode_hyper(p, &args->cbl_range.length); - status = decode_layout_stateid(xdr, &args->cbl_stateid); - if (unlikely(status != 0)) - goto out; + return decode_layout_stateid(xdr, &args->cbl_stateid); } else if (args->cbl_recall_type == RETURN_FSID) { p = read_buf(xdr, 2 * sizeof(uint64_t)); - if (unlikely(p == NULL)) { - status = htonl(NFS4ERR_BADXDR); - goto out; - } + if (unlikely(p == NULL)) + return htonl(NFS4ERR_BADXDR); p = xdr_decode_hyper(p, &args->cbl_fsid.major); p = xdr_decode_hyper(p, &args->cbl_fsid.minor); - } else if (args->cbl_recall_type != RETURN_ALL) { - status = htonl(NFS4ERR_BADXDR); - goto out; - } - dprintk("%s: ltype 0x%x iomode %d changed %d recall_type %d\n", - __func__, - args->cbl_layout_type, iomode, - args->cbl_layoutchanged, args->cbl_recall_type); -out: - dprintk("%s: exit with status = %d\n", __func__, ntohl(status)); - return status; + } else if (args->cbl_recall_type != RETURN_ALL) + return htonl(NFS4ERR_BADXDR); + return 0; } static @@ -437,12 +411,11 @@ static __be32 decode_cb_sequence_args(struct svc_rqst *rqstp, status = decode_sessionid(xdr, &args->csa_sessionid); if (status) - goto out; + return status; - status = htonl(NFS4ERR_RESOURCE); p = read_buf(xdr, 5 * sizeof(uint32_t)); if (unlikely(p == NULL)) - goto out; + return htonl(NFS4ERR_RESOURCE); args->csa_addr = svc_addr(rqstp); args->csa_sequenceid = ntohl(*p++); @@ -456,7 +429,7 @@ static __be32 decode_cb_sequence_args(struct svc_rqst *rqstp, sizeof(*args->csa_rclists), GFP_KERNEL); if (unlikely(args->csa_rclists == NULL)) - goto out; + return htonl(NFS4ERR_RESOURCE); for (i = 0; i < args->csa_nrclists; i++) { status = decode_rc_list(xdr, &args->csa_rclists[i]); @@ -466,27 +439,13 @@ static __be32 decode_cb_sequence_args(struct svc_rqst *rqstp, } } } - status = 0; - - dprintk("%s: sessionid %x:%x:%x:%x sequenceid %u slotid %u " - "highestslotid %u cachethis %d nrclists %u\n", - __func__, - ((u32 *)&args->csa_sessionid)[0], - ((u32 *)&args->csa_sessionid)[1], - ((u32 *)&args->csa_sessionid)[2], - ((u32 *)&args->csa_sessionid)[3], - args->csa_sequenceid, args->csa_slotid, - args->csa_highestslotid, args->csa_cachethis, - args->csa_nrclists); -out: - dprintk("%s: exit with status = %d\n", __func__, ntohl(status)); - return status; + return 0; out_free: for (i = 0; i < args->csa_nrclists; i++) kfree(args->csa_rclists[i].rcl_refcalls); kfree(args->csa_rclists); - goto out; + return status; } static __be32 decode_recallany_args(struct svc_rqst *rqstp, @@ -557,11 +516,8 @@ static __be32 decode_notify_lock_args(struct svc_rqst *rqstp, struct xdr_stream status = decode_fh(xdr, &args->cbnl_fh); if (unlikely(status != 0)) - goto out; - status = decode_lockowner(xdr, args); -out: - dprintk("%s: exit with status = %d\n", __func__, ntohl(status)); - return status; + return status; + return decode_lockowner(xdr, args); } #endif /* CONFIG_NFS_V4_1 */ @@ -707,7 +663,6 @@ static __be32 encode_getattr_res(struct svc_rqst *rqstp, struct xdr_stream *xdr, status = encode_attr_mtime(xdr, res->bitmap, &res->mtime); *savep = htonl((unsigned int)((char *)xdr->p - (char *)(savep+1))); out: - dprintk("%s: exit with status = %d\n", __func__, ntohl(status)); return status; } @@ -734,11 +689,11 @@ static __be32 encode_cb_sequence_res(struct svc_rqst *rqstp, __be32 status = res->csr_status; if (unlikely(status != 0)) - goto out; + return status; status = encode_sessionid(xdr, &res->csr_sessionid); if (status) - goto out; + return status; p = xdr_reserve_space(xdr, 4 * sizeof(uint32_t)); if (unlikely(p == NULL)) @@ -748,9 +703,7 @@ static __be32 encode_cb_sequence_res(struct svc_rqst *rqstp, *p++ = htonl(res->csr_slotid); *p++ = htonl(res->csr_highestslotid); *p++ = htonl(res->csr_target_highestslotid); -out: - dprintk("%s: exit with status = %d\n", __func__, ntohl(status)); - return status; + return 0; } static __be32 @@ -871,14 +824,10 @@ static __be32 process_op(int nop, struct svc_rqst *rqstp, long maxlen; __be32 res; - dprintk("%s: start\n", __func__); status = decode_op_hdr(xdr_in, &op_nr); if (unlikely(status)) return status; - dprintk("%s: minorversion=%d nop=%d op_nr=%u\n", - __func__, cps->minorversion, nop, op_nr); - switch (cps->minorversion) { case 0: status = preprocess_nfs4_op(op_nr, &op); @@ -917,7 +866,6 @@ encode_hdr: return res; if (op->encode_res != NULL && status == 0) status = op->encode_res(rqstp, xdr_out, resp); - dprintk("%s: done, status = %d\n", __func__, ntohl(status)); return status; } @@ -937,8 +885,6 @@ static __be32 nfs4_callback_compound(struct svc_rqst *rqstp, void *argp, void *r }; unsigned int nops = 0; - dprintk("%s: start\n", __func__); - xdr_init_decode(&xdr_in, &rqstp->rq_arg, rqstp->rq_arg.head[0].iov_base); p = (__be32*)((char *)rqstp->rq_res.head[0].iov_base + rqstp->rq_res.head[0].iov_len); @@ -977,7 +923,6 @@ static __be32 nfs4_callback_compound(struct svc_rqst *rqstp, void *argp, void *r *hdr_res.nops = htonl(nops); nfs4_cb_free_slot(&cps); nfs_put_client(cps.clp); - dprintk("%s: done, status = %u\n", __func__, ntohl(status)); return rpc_success; out_invalidcred: diff --git a/fs/nfs/client.c b/fs/nfs/client.c index 04d15a0045e3..ee5ddbd36088 100644 --- a/fs/nfs/client.c +++ b/fs/nfs/client.c @@ -218,6 +218,7 @@ static void nfs_cb_idr_remove_locked(struct nfs_client *clp) static void pnfs_init_server(struct nfs_server *server) { rpc_init_wait_queue(&server->roc_rpcwaitq, "pNFS ROC"); + rpc_init_wait_queue(&server->uoc_rpcwaitq, "NFS UOC"); } #else @@ -240,8 +241,6 @@ static void pnfs_init_server(struct nfs_server *server) */ void nfs_free_client(struct nfs_client *clp) { - dprintk("--> nfs_free_client(%u)\n", clp->rpc_ops->version); - nfs_fscache_release_client_cookie(clp); /* -EIO all pending I/O */ @@ -256,8 +255,6 @@ void nfs_free_client(struct nfs_client *clp) kfree(clp->cl_hostname); kfree(clp->cl_acceptor); kfree(clp); - - dprintk("<-- nfs_free_client()\n"); } EXPORT_SYMBOL_GPL(nfs_free_client); @@ -271,7 +268,6 @@ void nfs_put_client(struct nfs_client *clp) if (!clp) return; - dprintk("--> nfs_put_client({%d})\n", atomic_read(&clp->cl_count)); nn = net_generic(clp->cl_net, nfs_net_id); if (atomic_dec_and_lock(&clp->cl_count, &nn->nfs_client_lock)) { @@ -382,9 +378,6 @@ nfs_found_client(const struct nfs_client_initdata *cl_init, } smp_rmb(); - - dprintk("<-- %s found nfs_client %p for %s\n", - __func__, clp, cl_init->hostname ?: ""); return clp; } @@ -403,9 +396,6 @@ struct nfs_client *nfs_get_client(const struct nfs_client_initdata *cl_init) return NULL; } - dprintk("--> nfs_get_client(%s,v%u)\n", - cl_init->hostname, rpc_ops->version); - /* see if the client already exists */ do { spin_lock(&nn->nfs_client_lock); @@ -430,8 +420,6 @@ struct nfs_client *nfs_get_client(const struct nfs_client_initdata *cl_init) new = rpc_ops->alloc_client(cl_init); } while (!IS_ERR(new)); - dprintk("<-- nfs_get_client() Failed to find %s (%ld)\n", - cl_init->hostname, PTR_ERR(new)); return new; } EXPORT_SYMBOL_GPL(nfs_get_client); @@ -558,6 +546,7 @@ static int nfs_start_lockd(struct nfs_server *server) .noresvport = server->flags & NFS_MOUNT_NORESVPORT ? 1 : 0, .net = clp->cl_net, + .nlmclnt_ops = clp->cl_nfs_mod->rpc_ops->nlmclnt_ops, }; if (nlm_init.nfs_version > 3) @@ -624,27 +613,21 @@ struct nfs_client *nfs_init_client(struct nfs_client *clp, { int error; - if (clp->cl_cons_state == NFS_CS_READY) { - /* the client is already initialised */ - dprintk("<-- nfs_init_client() = 0 [already %p]\n", clp); + /* the client is already initialised */ + if (clp->cl_cons_state == NFS_CS_READY) return clp; - } /* * Create a client RPC handle for doing FSSTAT with UNIX auth only * - RFC 2623, sec 2.3.2 */ error = nfs_create_rpc_client(clp, cl_init, RPC_AUTH_UNIX); - if (error < 0) - goto error; - nfs_mark_client_ready(clp, NFS_CS_READY); + nfs_mark_client_ready(clp, error == 0 ? NFS_CS_READY : error); + if (error < 0) { + nfs_put_client(clp); + clp = ERR_PTR(error); + } return clp; - -error: - nfs_mark_client_ready(clp, error); - nfs_put_client(clp); - dprintk("<-- nfs_init_client() = xerror %d\n", error); - return ERR_PTR(error); } EXPORT_SYMBOL_GPL(nfs_init_client); @@ -668,8 +651,6 @@ static int nfs_init_server(struct nfs_server *server, struct nfs_client *clp; int error; - dprintk("--> nfs_init_server()\n"); - nfs_init_timeout_values(&timeparms, data->nfs_server.protocol, data->timeo, data->retrans); if (data->flags & NFS_MOUNT_NORESVPORT) @@ -677,10 +658,8 @@ static int nfs_init_server(struct nfs_server *server, /* Allocate or find a client reference we can use */ clp = nfs_get_client(&cl_init); - if (IS_ERR(clp)) { - dprintk("<-- nfs_init_server() = error %ld\n", PTR_ERR(clp)); + if (IS_ERR(clp)) return PTR_ERR(clp); - } server->nfs_client = clp; @@ -725,13 +704,11 @@ static int nfs_init_server(struct nfs_server *server, server->mountd_protocol = data->mount_server.protocol; server->namelen = data->namlen; - dprintk("<-- nfs_init_server() = 0 [new %p]\n", clp); return 0; error: server->nfs_client = NULL; nfs_put_client(clp); - dprintk("<-- nfs_init_server() = xerror %d\n", error); return error; } @@ -798,12 +775,10 @@ int nfs_probe_fsinfo(struct nfs_server *server, struct nfs_fh *mntfh, struct nfs struct nfs_client *clp = server->nfs_client; int error; - dprintk("--> nfs_probe_fsinfo()\n"); - if (clp->rpc_ops->set_capabilities != NULL) { error = clp->rpc_ops->set_capabilities(server, mntfh); if (error < 0) - goto out_error; + return error; } fsinfo.fattr = fattr; @@ -811,7 +786,7 @@ int nfs_probe_fsinfo(struct nfs_server *server, struct nfs_fh *mntfh, struct nfs memset(fsinfo.layouttype, 0, sizeof(fsinfo.layouttype)); error = clp->rpc_ops->fsinfo(server, mntfh, &fsinfo); if (error < 0) - goto out_error; + return error; nfs_server_set_fsinfo(server, &fsinfo); @@ -826,12 +801,7 @@ int nfs_probe_fsinfo(struct nfs_server *server, struct nfs_fh *mntfh, struct nfs server->namelen = pathinfo.max_namelen; } - dprintk("<-- nfs_probe_fsinfo() = 0\n"); return 0; - -out_error: - dprintk("nfs_probe_fsinfo: error = %d\n", -error); - return error; } EXPORT_SYMBOL_GPL(nfs_probe_fsinfo); @@ -927,8 +897,6 @@ EXPORT_SYMBOL_GPL(nfs_alloc_server); */ void nfs_free_server(struct nfs_server *server) { - dprintk("--> nfs_free_server()\n"); - nfs_server_remove_lists(server); if (server->destroy != NULL) @@ -946,7 +914,6 @@ void nfs_free_server(struct nfs_server *server) nfs_free_iostats(server->io_stats); kfree(server); nfs_release_automount_timer(); - dprintk("<-- nfs_free_server()\n"); } EXPORT_SYMBOL_GPL(nfs_free_server); @@ -1026,10 +993,6 @@ struct nfs_server *nfs_clone_server(struct nfs_server *source, struct nfs_fattr *fattr_fsinfo; int error; - dprintk("--> nfs_clone_server(,%llx:%llx,)\n", - (unsigned long long) fattr->fsid.major, - (unsigned long long) fattr->fsid.minor); - server = nfs_alloc_server(); if (!server) return ERR_PTR(-ENOMEM); @@ -1061,10 +1024,6 @@ struct nfs_server *nfs_clone_server(struct nfs_server *source, if (server->namelen == 0 || server->namelen > NFS4_MAXNAMLEN) server->namelen = NFS4_MAXNAMLEN; - dprintk("Cloned FSID: %llx:%llx\n", - (unsigned long long) server->fsid.major, - (unsigned long long) server->fsid.minor); - error = nfs_start_lockd(server); if (error < 0) goto out_free_server; @@ -1073,13 +1032,11 @@ struct nfs_server *nfs_clone_server(struct nfs_server *source, server->mount_time = jiffies; nfs_free_fattr(fattr_fsinfo); - dprintk("<-- nfs_clone_server() = %p\n", server); return server; out_free_server: nfs_free_fattr(fattr_fsinfo); nfs_free_server(server); - dprintk("<-- nfs_clone_server() = error %d\n", error); return ERR_PTR(error); } EXPORT_SYMBOL_GPL(nfs_clone_server); diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c index f92ba8d6c556..32ccd7754f8a 100644 --- a/fs/nfs/dir.c +++ b/fs/nfs/dir.c @@ -57,7 +57,7 @@ static void nfs_readdir_clear_array(struct page*); const struct file_operations nfs_dir_operations = { .llseek = nfs_llseek_dir, .read = generic_read_dir, - .iterate_shared = nfs_readdir, + .iterate = nfs_readdir, .open = nfs_opendir, .release = nfs_closedir, .fsync = nfs_fsync_dir, @@ -145,7 +145,6 @@ struct nfs_cache_array_entry { }; struct nfs_cache_array { - atomic_t refcount; int size; int eof_index; u64 last_cookie; @@ -171,27 +170,6 @@ typedef struct { } nfs_readdir_descriptor_t; /* - * The caller is responsible for calling nfs_readdir_release_array(page) - */ -static -struct nfs_cache_array *nfs_readdir_get_array(struct page *page) -{ - void *ptr; - if (page == NULL) - return ERR_PTR(-EIO); - ptr = kmap(page); - if (ptr == NULL) - return ERR_PTR(-ENOMEM); - return ptr; -} - -static -void nfs_readdir_release_array(struct page *page) -{ - kunmap(page); -} - -/* * we are freeing strings created by nfs_add_to_readdir_array() */ static @@ -201,18 +179,9 @@ void nfs_readdir_clear_array(struct page *page) int i; array = kmap_atomic(page); - if (atomic_dec_and_test(&array->refcount)) - for (i = 0; i < array->size; i++) - kfree(array->array[i].string.name); - kunmap_atomic(array); -} - -static bool grab_page(struct page *page) -{ - struct nfs_cache_array *array = kmap_atomic(page); - bool res = atomic_inc_not_zero(&array->refcount); + for (i = 0; i < array->size; i++) + kfree(array->array[i].string.name); kunmap_atomic(array); - return res; } /* @@ -239,13 +208,10 @@ int nfs_readdir_make_qstr(struct qstr *string, const char *name, unsigned int le static int nfs_readdir_add_to_array(struct nfs_entry *entry, struct page *page) { - struct nfs_cache_array *array = nfs_readdir_get_array(page); + struct nfs_cache_array *array = kmap(page); struct nfs_cache_array_entry *cache_entry; int ret; - if (IS_ERR(array)) - return PTR_ERR(array); - cache_entry = &array->array[array->size]; /* Check that this entry lies within the page bounds */ @@ -264,7 +230,7 @@ int nfs_readdir_add_to_array(struct nfs_entry *entry, struct page *page) if (entry->eof != 0) array->eof_index = array->size; out: - nfs_readdir_release_array(page); + kunmap(page); return ret; } @@ -353,11 +319,7 @@ int nfs_readdir_search_array(nfs_readdir_descriptor_t *desc) struct nfs_cache_array *array; int status; - array = nfs_readdir_get_array(desc->page); - if (IS_ERR(array)) { - status = PTR_ERR(array); - goto out; - } + array = kmap(desc->page); if (*desc->dir_cookie == 0) status = nfs_readdir_search_for_pos(array, desc); @@ -369,8 +331,7 @@ int nfs_readdir_search_array(nfs_readdir_descriptor_t *desc) desc->current_index += array->size; desc->page_index++; } - nfs_readdir_release_array(desc->page); -out: + kunmap(desc->page); return status; } @@ -606,13 +567,10 @@ int nfs_readdir_page_filler(nfs_readdir_descriptor_t *desc, struct nfs_entry *en out_nopages: if (count == 0 || (status == -EBADCOOKIE && entry->eof != 0)) { - array = nfs_readdir_get_array(page); - if (!IS_ERR(array)) { - array->eof_index = array->size; - status = 0; - nfs_readdir_release_array(page); - } else - status = PTR_ERR(array); + array = kmap(page); + array->eof_index = array->size; + status = 0; + kunmap(page); } put_page(scratch); @@ -674,13 +632,8 @@ int nfs_readdir_xdr_to_array(nfs_readdir_descriptor_t *desc, struct page *page, goto out; } - array = nfs_readdir_get_array(page); - if (IS_ERR(array)) { - status = PTR_ERR(array); - goto out_label_free; - } + array = kmap(page); memset(array, 0, sizeof(struct nfs_cache_array)); - atomic_set(&array->refcount, 1); array->eof_index = -1; status = nfs_readdir_alloc_pages(pages, array_size); @@ -703,8 +656,7 @@ int nfs_readdir_xdr_to_array(nfs_readdir_descriptor_t *desc, struct page *page, nfs_readdir_free_pages(pages, array_size); out_release_array: - nfs_readdir_release_array(page); -out_label_free: + kunmap(page); nfs4_label_free(entry.label); out: nfs_free_fattr(entry.fattr); @@ -743,7 +695,8 @@ int nfs_readdir_filler(nfs_readdir_descriptor_t *desc, struct page* page) static void cache_page_release(nfs_readdir_descriptor_t *desc) { - nfs_readdir_clear_array(desc->page); + if (!desc->page->mapping) + nfs_readdir_clear_array(desc->page); put_page(desc->page); desc->page = NULL; } @@ -751,16 +704,8 @@ void cache_page_release(nfs_readdir_descriptor_t *desc) static struct page *get_cache_page(nfs_readdir_descriptor_t *desc) { - struct page *page; - - for (;;) { - page = read_cache_page(desc->file->f_mapping, + return read_cache_page(desc->file->f_mapping, desc->page_index, (filler_t *)nfs_readdir_filler, desc); - if (IS_ERR(page) || grab_page(page)) - break; - put_page(page); - } - return page; } /* @@ -809,12 +754,7 @@ int nfs_do_filldir(nfs_readdir_descriptor_t *desc) struct nfs_cache_array *array = NULL; struct nfs_open_dir_context *ctx = file->private_data; - array = nfs_readdir_get_array(desc->page); - if (IS_ERR(array)) { - res = PTR_ERR(array); - goto out; - } - + array = kmap(desc->page); for (i = desc->cache_entry_index; i < array->size; i++) { struct nfs_cache_array_entry *ent; @@ -835,8 +775,7 @@ int nfs_do_filldir(nfs_readdir_descriptor_t *desc) if (array->eof_index >= 0) desc->eof = 1; - nfs_readdir_release_array(desc->page); -out: + kunmap(desc->page); cache_page_release(desc); dfprintk(DIRCACHE, "NFS: nfs_do_filldir() filling ended @ cookie %Lu; returning = %d\n", (unsigned long long)*desc->dir_cookie, res); @@ -966,11 +905,13 @@ out: static loff_t nfs_llseek_dir(struct file *filp, loff_t offset, int whence) { + struct inode *inode = file_inode(filp); struct nfs_open_dir_context *dir_ctx = filp->private_data; dfprintk(FILE, "NFS: llseek dir(%pD2, %lld, %d)\n", filp, offset, whence); + inode_lock(inode); switch (whence) { case 1: offset += filp->f_pos; @@ -978,13 +919,16 @@ static loff_t nfs_llseek_dir(struct file *filp, loff_t offset, int whence) if (offset >= 0) break; default: - return -EINVAL; + offset = -EINVAL; + goto out; } if (offset != filp->f_pos) { filp->f_pos = offset; dir_ctx->dir_cookie = 0; dir_ctx->duped = 0; } +out: + inode_unlock(inode); return offset; } diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c index c1b5fed7c863..6fb9fad2d1e6 100644 --- a/fs/nfs/direct.c +++ b/fs/nfs/direct.c @@ -392,16 +392,6 @@ static void nfs_direct_complete(struct nfs_direct_req *dreq) nfs_direct_req_release(dreq); } -static void nfs_direct_readpage_release(struct nfs_page *req) -{ - dprintk("NFS: direct read done (%s/%llu %d@%lld)\n", - req->wb_context->dentry->d_sb->s_id, - (unsigned long long)NFS_FILEID(d_inode(req->wb_context->dentry)), - req->wb_bytes, - (long long)req_offset(req)); - nfs_release_request(req); -} - static void nfs_direct_read_completion(struct nfs_pgio_header *hdr) { unsigned long bytes = 0; @@ -426,7 +416,7 @@ static void nfs_direct_read_completion(struct nfs_pgio_header *hdr) set_page_dirty(page); bytes += req->wb_bytes; nfs_list_remove_request(req); - nfs_direct_readpage_release(req); + nfs_release_request(req); } out_put: if (put_dreq(dreq)) @@ -700,16 +690,9 @@ static void nfs_direct_commit_complete(struct nfs_commit_data *data) int status = data->task.tk_status; nfs_init_cinfo_from_dreq(&cinfo, dreq); - if (status < 0) { - dprintk("NFS: %5u commit failed with error %d.\n", - data->task.tk_pid, status); - dreq->flags = NFS_ODIRECT_RESCHED_WRITES; - } else if (nfs_direct_cmp_commit_data_verf(dreq, data)) { - dprintk("NFS: %5u commit verify failed\n", data->task.tk_pid); + if (status < 0 || nfs_direct_cmp_commit_data_verf(dreq, data)) dreq->flags = NFS_ODIRECT_RESCHED_WRITES; - } - dprintk("NFS: %5u commit returned %d\n", data->task.tk_pid, status); while (!list_empty(&data->pages)) { req = nfs_list_entry(data->pages.next); nfs_list_remove_request(req); diff --git a/fs/nfs/file.c b/fs/nfs/file.c index 668213984d68..5713eb32a45e 100644 --- a/fs/nfs/file.c +++ b/fs/nfs/file.c @@ -482,7 +482,7 @@ static int nfs_launder_page(struct page *page) inode->i_ino, (long long)page_offset(page)); nfs_fscache_wait_on_page_write(nfsi, page); - return nfs_wb_launder_page(inode, page); + return nfs_wb_page(inode, page); } static int nfs_swap_activate(struct swap_info_struct *sis, struct file *file, @@ -697,14 +697,14 @@ do_unlk(struct file *filp, int cmd, struct file_lock *fl, int is_local) if (!IS_ERR(l_ctx)) { status = nfs_iocounter_wait(l_ctx); nfs_put_lock_context(l_ctx); - if (status < 0) + /* NOTE: special case + * If we're signalled while cleaning up locks on process exit, we + * still need to complete the unlock. + */ + if (status < 0 && !(fl->fl_flags & FL_CLOSE)) return status; } - /* NOTE: special case - * If we're signalled while cleaning up locks on process exit, we - * still need to complete the unlock. - */ /* * Use local locking if mounted with "-onolock" or with appropriate * "-olocal_lock=" @@ -820,9 +820,23 @@ int nfs_flock(struct file *filp, int cmd, struct file_lock *fl) if (NFS_SERVER(inode)->flags & NFS_MOUNT_LOCAL_FLOCK) is_local = 1; - /* We're simulating flock() locks using posix locks on the server */ - if (fl->fl_type == F_UNLCK) + /* + * VFS doesn't require the open mode to match a flock() lock's type. + * NFS, however, may simulate flock() locking with posix locking which + * requires the open mode to match the lock type. + */ + switch (fl->fl_type) { + case F_UNLCK: return do_unlk(filp, cmd, fl, is_local); + case F_RDLCK: + if (!(filp->f_mode & FMODE_READ)) + return -EBADF; + break; + case F_WRLCK: + if (!(filp->f_mode & FMODE_WRITE)) + return -EBADF; + } + return do_setlk(filp, cmd, fl, is_local); } EXPORT_SYMBOL_GPL(nfs_flock); diff --git a/fs/nfs/filelayout/filelayout.c b/fs/nfs/filelayout/filelayout.c index acd30baca461..1cf85d65b748 100644 --- a/fs/nfs/filelayout/filelayout.c +++ b/fs/nfs/filelayout/filelayout.c @@ -921,11 +921,11 @@ fl_pnfs_update_layout(struct inode *ino, fl = FILELAYOUT_LSEG(lseg); status = filelayout_check_deviceid(lo, fl, gfp_flags); - if (status) + if (status) { + pnfs_put_lseg(lseg); lseg = ERR_PTR(status); + } out: - if (IS_ERR(lseg)) - pnfs_put_lseg(lseg); return lseg; } @@ -933,6 +933,7 @@ static void filelayout_pg_init_read(struct nfs_pageio_descriptor *pgio, struct nfs_page *req) { + pnfs_generic_pg_check_layout(pgio); if (!pgio->pg_lseg) { pgio->pg_lseg = fl_pnfs_update_layout(pgio->pg_inode, req->wb_context, @@ -959,6 +960,7 @@ filelayout_pg_init_write(struct nfs_pageio_descriptor *pgio, struct nfs_commit_info cinfo; int status; + pnfs_generic_pg_check_layout(pgio); if (!pgio->pg_lseg) { pgio->pg_lseg = fl_pnfs_update_layout(pgio->pg_inode, req->wb_context, diff --git a/fs/nfs/flexfilelayout/flexfilelayout.c b/fs/nfs/flexfilelayout/flexfilelayout.c index 42dedf2d625f..f5714ee01000 100644 --- a/fs/nfs/flexfilelayout/flexfilelayout.c +++ b/fs/nfs/flexfilelayout/flexfilelayout.c @@ -846,6 +846,7 @@ ff_layout_pg_init_read(struct nfs_pageio_descriptor *pgio, int ds_idx; retry: + pnfs_generic_pg_check_layout(pgio); /* Use full layout for now */ if (!pgio->pg_lseg) ff_layout_pg_get_read(pgio, req, false); @@ -894,6 +895,7 @@ ff_layout_pg_init_write(struct nfs_pageio_descriptor *pgio, int status; retry: + pnfs_generic_pg_check_layout(pgio); if (!pgio->pg_lseg) { pgio->pg_lseg = pnfs_update_layout(pgio->pg_inode, req->wb_context, @@ -1800,16 +1802,16 @@ ff_layout_write_pagelist(struct nfs_pgio_header *hdr, int sync) ds = nfs4_ff_layout_prepare_ds(lseg, idx, true); if (!ds) - return PNFS_NOT_ATTEMPTED; + goto out_failed; ds_clnt = nfs4_ff_find_or_create_ds_client(lseg, idx, ds->ds_clp, hdr->inode); if (IS_ERR(ds_clnt)) - return PNFS_NOT_ATTEMPTED; + goto out_failed; ds_cred = ff_layout_get_ds_cred(lseg, idx, hdr->cred); if (!ds_cred) - return PNFS_NOT_ATTEMPTED; + goto out_failed; vers = nfs4_ff_layout_ds_version(lseg, idx); @@ -1839,6 +1841,11 @@ ff_layout_write_pagelist(struct nfs_pgio_header *hdr, int sync) sync, RPC_TASK_SOFTCONN); put_rpccred(ds_cred); return PNFS_ATTEMPTED; + +out_failed: + if (ff_layout_avoid_mds_available_ds(lseg)) + return PNFS_TRY_AGAIN; + return PNFS_NOT_ATTEMPTED; } static u32 calc_ds_index_from_commit(struct pnfs_layout_segment *lseg, u32 i) @@ -2354,10 +2361,21 @@ ff_layout_prepare_layoutstats(struct nfs42_layoutstat_args *args) return 0; } +static int +ff_layout_set_layoutdriver(struct nfs_server *server, + const struct nfs_fh *dummy) +{ +#if IS_ENABLED(CONFIG_NFS_V4_2) + server->caps |= NFS_CAP_LAYOUTSTATS; +#endif + return 0; +} + static struct pnfs_layoutdriver_type flexfilelayout_type = { .id = LAYOUT_FLEX_FILES, .name = "LAYOUT_FLEX_FILES", .owner = THIS_MODULE, + .set_layoutdriver = ff_layout_set_layoutdriver, .alloc_layout_hdr = ff_layout_alloc_layout_hdr, .free_layout_hdr = ff_layout_free_layout_hdr, .alloc_lseg = ff_layout_alloc_lseg, diff --git a/fs/nfs/flexfilelayout/flexfilelayoutdev.c b/fs/nfs/flexfilelayout/flexfilelayoutdev.c index 457cfeb1d5c1..6df7a0cf5660 100644 --- a/fs/nfs/flexfilelayout/flexfilelayoutdev.c +++ b/fs/nfs/flexfilelayout/flexfilelayoutdev.c @@ -119,7 +119,13 @@ nfs4_ff_alloc_deviceid_node(struct nfs_server *server, struct pnfs_device *pdev, if (ds_versions[i].wsize > NFS_MAX_FILE_IO_SIZE) ds_versions[i].wsize = NFS_MAX_FILE_IO_SIZE; - if (ds_versions[i].version != 3 || ds_versions[i].minor_version != 0) { + /* + * check for valid major/minor combination. + * currently we support dataserver which talk: + * v3, v4.0, v4.1, v4.2 + */ + if (!((ds_versions[i].version == 3 && ds_versions[i].minor_version == 0) || + (ds_versions[i].version == 4 && ds_versions[i].minor_version < 3))) { dprintk("%s: [%d] unsupported ds version %d-%d\n", __func__, i, ds_versions[i].version, ds_versions[i].minor_version); @@ -415,7 +421,7 @@ nfs4_ff_layout_prepare_ds(struct pnfs_layout_segment *lseg, u32 ds_idx, mirror->mirror_ds->ds_versions[0].minor_version); /* connect success, check rsize/wsize limit */ - if (ds->ds_clp) { + if (!status) { max_payload = nfs_block_size(rpc_max_payload(ds->ds_clp->cl_rpcclient), NULL); diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c index f489a5a71bd5..1de93ba78dc9 100644 --- a/fs/nfs/inode.c +++ b/fs/nfs/inode.c @@ -734,7 +734,10 @@ int nfs_getattr(const struct path *path, struct kstat *stat, if (need_atime || nfs_need_revalidate_inode(inode)) { struct nfs_server *server = NFS_SERVER(inode); - nfs_readdirplus_parent_cache_miss(path->dentry); + if (!(server->flags & NFS_MOUNT_NOAC)) + nfs_readdirplus_parent_cache_miss(path->dentry); + else + nfs_readdirplus_parent_cache_hit(path->dentry); err = __nfs_revalidate_inode(server, inode); } else nfs_readdirplus_parent_cache_hit(path->dentry); diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h index 7b38fedb7e03..e9b4c3320e37 100644 --- a/fs/nfs/internal.h +++ b/fs/nfs/internal.h @@ -495,7 +495,6 @@ void nfs_mark_request_commit(struct nfs_page *req, u32 ds_commit_idx); int nfs_write_need_commit(struct nfs_pgio_header *); void nfs_writeback_update_inode(struct nfs_pgio_header *hdr); -int nfs_commit_file(struct file *file, struct nfs_write_verifier *verf); int nfs_generic_commit_list(struct inode *inode, struct list_head *head, int how, struct nfs_commit_info *cinfo); void nfs_retry_commit(struct list_head *page_list, @@ -756,9 +755,13 @@ static inline bool nfs_error_is_fatal(int err) { switch (err) { case -ERESTARTSYS: + case -EACCES: + case -EDQUOT: + case -EFBIG: case -EIO: case -ENOSPC: case -EROFS: + case -ESTALE: case -E2BIG: return true; default: diff --git a/fs/nfs/namespace.c b/fs/nfs/namespace.c index 786f17580582..1a224a33a6c2 100644 --- a/fs/nfs/namespace.c +++ b/fs/nfs/namespace.c @@ -143,11 +143,8 @@ struct vfsmount *nfs_d_automount(struct path *path) struct nfs_fh *fh = NULL; struct nfs_fattr *fattr = NULL; - dprintk("--> nfs_d_automount()\n"); - - mnt = ERR_PTR(-ESTALE); if (IS_ROOT(path->dentry)) - goto out_nofree; + return ERR_PTR(-ESTALE); mnt = ERR_PTR(-ENOMEM); fh = nfs_alloc_fhandle(); @@ -155,13 +152,10 @@ struct vfsmount *nfs_d_automount(struct path *path) if (fh == NULL || fattr == NULL) goto out; - dprintk("%s: enter\n", __func__); - mnt = server->nfs_client->rpc_ops->submount(server, path->dentry, fh, fattr); if (IS_ERR(mnt)) goto out; - dprintk("%s: done, success\n", __func__); mntget(mnt); /* prevent immediate expiration */ mnt_set_expiry(mnt, &nfs_automount_list); schedule_delayed_work(&nfs_automount_task, nfs_mountpoint_expiry_timeout); @@ -169,11 +163,6 @@ struct vfsmount *nfs_d_automount(struct path *path) out: nfs_free_fattr(fattr); nfs_free_fhandle(fh); -out_nofree: - if (IS_ERR(mnt)) - dprintk("<-- %s(): error %ld\n", __func__, PTR_ERR(mnt)); - else - dprintk("<-- %s() = %p\n", __func__, mnt); return mnt; } @@ -248,27 +237,20 @@ struct vfsmount *nfs_do_submount(struct dentry *dentry, struct nfs_fh *fh, .fattr = fattr, .authflavor = authflavor, }; - struct vfsmount *mnt = ERR_PTR(-ENOMEM); + struct vfsmount *mnt; char *page = (char *) __get_free_page(GFP_USER); char *devname; - dprintk("--> nfs_do_submount()\n"); - - dprintk("%s: submounting on %pd2\n", __func__, - dentry); if (page == NULL) - goto out; + return ERR_PTR(-ENOMEM); + devname = nfs_devname(dentry, page, PAGE_SIZE); - mnt = (struct vfsmount *)devname; if (IS_ERR(devname)) - goto free_page; - mnt = nfs_do_clone_mount(NFS_SB(dentry->d_sb), devname, &mountdata); -free_page: - free_page((unsigned long)page); -out: - dprintk("%s: done\n", __func__); + mnt = (struct vfsmount *)devname; + else + mnt = nfs_do_clone_mount(NFS_SB(dentry->d_sb), devname, &mountdata); - dprintk("<-- nfs_do_submount() = %p\n", mnt); + free_page((unsigned long)page); return mnt; } EXPORT_SYMBOL_GPL(nfs_do_submount); diff --git a/fs/nfs/nfs3proc.c b/fs/nfs/nfs3proc.c index dc925b531f32..0c07b567118d 100644 --- a/fs/nfs/nfs3proc.c +++ b/fs/nfs/nfs3proc.c @@ -865,12 +865,63 @@ static void nfs3_proc_commit_setup(struct nfs_commit_data *data, struct rpc_mess msg->rpc_proc = &nfs3_procedures[NFS3PROC_COMMIT]; } +static void nfs3_nlm_alloc_call(void *data) +{ + struct nfs_lock_context *l_ctx = data; + if (l_ctx && test_bit(NFS_CONTEXT_UNLOCK, &l_ctx->open_context->flags)) { + get_nfs_open_context(l_ctx->open_context); + nfs_get_lock_context(l_ctx->open_context); + } +} + +static bool nfs3_nlm_unlock_prepare(struct rpc_task *task, void *data) +{ + struct nfs_lock_context *l_ctx = data; + if (l_ctx && test_bit(NFS_CONTEXT_UNLOCK, &l_ctx->open_context->flags)) + return nfs_async_iocounter_wait(task, l_ctx); + return false; + +} + +static void nfs3_nlm_release_call(void *data) +{ + struct nfs_lock_context *l_ctx = data; + struct nfs_open_context *ctx; + if (l_ctx && test_bit(NFS_CONTEXT_UNLOCK, &l_ctx->open_context->flags)) { + ctx = l_ctx->open_context; + nfs_put_lock_context(l_ctx); + put_nfs_open_context(ctx); + } +} + +const struct nlmclnt_operations nlmclnt_fl_close_lock_ops = { + .nlmclnt_alloc_call = nfs3_nlm_alloc_call, + .nlmclnt_unlock_prepare = nfs3_nlm_unlock_prepare, + .nlmclnt_release_call = nfs3_nlm_release_call, +}; + static int nfs3_proc_lock(struct file *filp, int cmd, struct file_lock *fl) { struct inode *inode = file_inode(filp); + struct nfs_lock_context *l_ctx = NULL; + struct nfs_open_context *ctx = nfs_file_open_context(filp); + int status; - return nlmclnt_proc(NFS_SERVER(inode)->nlm_host, cmd, fl); + if (fl->fl_flags & FL_CLOSE) { + l_ctx = nfs_get_lock_context(ctx); + if (IS_ERR(l_ctx)) + l_ctx = NULL; + else + set_bit(NFS_CONTEXT_UNLOCK, &ctx->flags); + } + + status = nlmclnt_proc(NFS_SERVER(inode)->nlm_host, cmd, fl, l_ctx); + + if (l_ctx) + nfs_put_lock_context(l_ctx); + + return status; } static int nfs3_have_delegation(struct inode *inode, fmode_t flags) @@ -921,6 +972,7 @@ const struct nfs_rpc_ops nfs_v3_clientops = { .dir_inode_ops = &nfs3_dir_inode_operations, .file_inode_ops = &nfs3_file_inode_operations, .file_ops = &nfs_file_operations, + .nlmclnt_ops = &nlmclnt_fl_close_lock_ops, .getroot = nfs3_proc_get_root, .submount = nfs_submount, .try_mount = nfs_try_mount, diff --git a/fs/nfs/nfs42proc.c b/fs/nfs/nfs42proc.c index 1e486c73ec94..929d09a5310a 100644 --- a/fs/nfs/nfs42proc.c +++ b/fs/nfs/nfs42proc.c @@ -167,23 +167,29 @@ static ssize_t _nfs42_proc_copy(struct file *src, if (status) return status; + res->commit_res.verf = kzalloc(sizeof(struct nfs_writeverf), GFP_NOFS); + if (!res->commit_res.verf) + return -ENOMEM; status = nfs4_call_sync(server->client, server, &msg, &args->seq_args, &res->seq_res, 0); if (status == -ENOTSUPP) server->caps &= ~NFS_CAP_COPY; if (status) - return status; + goto out; - if (res->write_res.verifier.committed != NFS_FILE_SYNC) { - status = nfs_commit_file(dst, &res->write_res.verifier.verifier); - if (status) - return status; + if (!nfs_write_verifier_cmp(&res->write_res.verifier.verifier, + &res->commit_res.verf->verifier)) { + status = -EAGAIN; + goto out; } truncate_pagecache_range(dst_inode, pos_dst, pos_dst + res->write_res.count); - return res->write_res.count; + status = res->write_res.count; +out: + kfree(res->commit_res.verf); + return status; } ssize_t nfs42_proc_copy(struct file *src, loff_t pos_src, @@ -240,6 +246,9 @@ ssize_t nfs42_proc_copy(struct file *src, loff_t pos_src, if (err == -ENOTSUPP) { err = -EOPNOTSUPP; break; + } if (err == -EAGAIN) { + dst_exception.retry = 1; + continue; } err2 = nfs4_handle_exception(server, err, &src_exception); @@ -379,6 +388,7 @@ nfs42_layoutstat_done(struct rpc_task *task, void *calldata) pnfs_mark_layout_stateid_invalid(lo, &head); spin_unlock(&inode->i_lock); pnfs_free_lseg_list(&head); + nfs_commit_inode(inode, 0); } else spin_unlock(&inode->i_lock); break; @@ -400,8 +410,6 @@ nfs42_layoutstat_done(struct rpc_task *task, void *calldata) case -EOPNOTSUPP: NFS_SERVER(inode)->caps &= ~NFS_CAP_LAYOUTSTATS; } - - dprintk("%s server returns %d\n", __func__, task->tk_status); } static void diff --git a/fs/nfs/nfs42xdr.c b/fs/nfs/nfs42xdr.c index 6c7296454bbc..528362f69cc1 100644 --- a/fs/nfs/nfs42xdr.c +++ b/fs/nfs/nfs42xdr.c @@ -66,12 +66,14 @@ encode_putfh_maxsz + \ encode_savefh_maxsz + \ encode_putfh_maxsz + \ - encode_copy_maxsz) + encode_copy_maxsz + \ + encode_commit_maxsz) #define NFS4_dec_copy_sz (compound_decode_hdr_maxsz + \ decode_putfh_maxsz + \ decode_savefh_maxsz + \ decode_putfh_maxsz + \ - decode_copy_maxsz) + decode_copy_maxsz + \ + decode_commit_maxsz) #define NFS4_enc_deallocate_sz (compound_encode_hdr_maxsz + \ encode_putfh_maxsz + \ encode_deallocate_maxsz + \ @@ -222,6 +224,18 @@ static void nfs4_xdr_enc_allocate(struct rpc_rqst *req, encode_nops(&hdr); } +static void encode_copy_commit(struct xdr_stream *xdr, + struct nfs42_copy_args *args, + struct compound_hdr *hdr) +{ + __be32 *p; + + encode_op_hdr(xdr, OP_COMMIT, decode_commit_maxsz, hdr); + p = reserve_space(xdr, 12); + p = xdr_encode_hyper(p, args->dst_pos); + *p = cpu_to_be32(args->count); +} + /* * Encode COPY request */ @@ -239,6 +253,7 @@ static void nfs4_xdr_enc_copy(struct rpc_rqst *req, encode_savefh(xdr, &hdr); encode_putfh(xdr, args->dst_fh, &hdr); encode_copy(xdr, args, &hdr); + encode_copy_commit(xdr, args, &hdr); encode_nops(&hdr); } @@ -481,6 +496,9 @@ static int nfs4_xdr_dec_copy(struct rpc_rqst *rqstp, if (status) goto out; status = decode_copy(xdr, res); + if (status) + goto out; + status = decode_commit(xdr, &res->commit_res); out: return status; } diff --git a/fs/nfs/nfs4client.c b/fs/nfs/nfs4client.c index 8346ccbf2d52..692a7a8bfc7a 100644 --- a/fs/nfs/nfs4client.c +++ b/fs/nfs/nfs4client.c @@ -359,11 +359,9 @@ struct nfs_client *nfs4_init_client(struct nfs_client *clp, struct nfs_client *old; int error; - if (clp->cl_cons_state == NFS_CS_READY) { + if (clp->cl_cons_state == NFS_CS_READY) /* the client is initialised already */ - dprintk("<-- nfs4_init_client() = 0 [already %p]\n", clp); return clp; - } /* Check NFS protocol revision and initialize RPC op vector */ clp->rpc_ops = &nfs_v4_clientops; @@ -421,7 +419,6 @@ struct nfs_client *nfs4_init_client(struct nfs_client *clp, error: nfs_mark_client_ready(clp, error); nfs_put_client(clp); - dprintk("<-- nfs4_init_client() = xerror %d\n", error); return ERR_PTR(error); } @@ -469,6 +466,50 @@ static bool nfs4_same_verifier(nfs4_verifier *v1, nfs4_verifier *v2) return memcmp(v1->data, v2->data, sizeof(v1->data)) == 0; } +static int nfs4_match_client(struct nfs_client *pos, struct nfs_client *new, + struct nfs_client **prev, struct nfs_net *nn) +{ + int status; + + if (pos->rpc_ops != new->rpc_ops) + return 1; + + if (pos->cl_minorversion != new->cl_minorversion) + return 1; + + /* If "pos" isn't marked ready, we can't trust the + * remaining fields in "pos", especially the client + * ID and serverowner fields. Wait for CREATE_SESSION + * to finish. */ + if (pos->cl_cons_state > NFS_CS_READY) { + atomic_inc(&pos->cl_count); + spin_unlock(&nn->nfs_client_lock); + + nfs_put_client(*prev); + *prev = pos; + + status = nfs_wait_client_init_complete(pos); + spin_lock(&nn->nfs_client_lock); + + if (status < 0) + return status; + } + + if (pos->cl_cons_state != NFS_CS_READY) + return 1; + + if (pos->cl_clientid != new->cl_clientid) + return 1; + + /* NFSv4.1 always uses the uniform string, however someone + * might switch the uniquifier string on us. + */ + if (!nfs4_match_client_owner_id(pos, new)) + return 1; + + return 0; +} + /** * nfs40_walk_client_list - Find server that recognizes a client ID * @@ -497,34 +538,10 @@ int nfs40_walk_client_list(struct nfs_client *new, spin_lock(&nn->nfs_client_lock); list_for_each_entry(pos, &nn->nfs_client_list, cl_share_link) { - if (pos->rpc_ops != new->rpc_ops) - continue; - - if (pos->cl_minorversion != new->cl_minorversion) - continue; - - /* If "pos" isn't marked ready, we can't trust the - * remaining fields in "pos" */ - if (pos->cl_cons_state > NFS_CS_READY) { - atomic_inc(&pos->cl_count); - spin_unlock(&nn->nfs_client_lock); - - nfs_put_client(prev); - prev = pos; - - status = nfs_wait_client_init_complete(pos); - if (status < 0) - goto out; - status = -NFS4ERR_STALE_CLIENTID; - spin_lock(&nn->nfs_client_lock); - } - if (pos->cl_cons_state != NFS_CS_READY) - continue; - - if (pos->cl_clientid != new->cl_clientid) - continue; - - if (!nfs4_match_client_owner_id(pos, new)) + status = nfs4_match_client(pos, new, &prev, nn); + if (status < 0) + goto out_unlock; + if (status != 0) continue; /* * We just sent a new SETCLIENTID, which should have @@ -557,8 +574,6 @@ int nfs40_walk_client_list(struct nfs_client *new, prev = NULL; *result = pos; - dprintk("NFS: <-- %s using nfs_client = %p ({%d})\n", - __func__, pos, atomic_read(&pos->cl_count)); goto out; case -ERESTARTSYS: case -ETIMEDOUT: @@ -567,37 +582,23 @@ int nfs40_walk_client_list(struct nfs_client *new, */ nfs4_schedule_path_down_recovery(pos); default: + spin_lock(&nn->nfs_client_lock); goto out; } spin_lock(&nn->nfs_client_lock); } +out_unlock: spin_unlock(&nn->nfs_client_lock); /* No match found. The server lost our clientid */ out: nfs_put_client(prev); - dprintk("NFS: <-- %s status = %d\n", __func__, status); return status; } #ifdef CONFIG_NFS_V4_1 /* - * Returns true if the client IDs match - */ -static bool nfs4_match_clientids(u64 a, u64 b) -{ - if (a != b) { - dprintk("NFS: --> %s client ID %llx does not match %llx\n", - __func__, a, b); - return false; - } - dprintk("NFS: --> %s client ID %llx matches %llx\n", - __func__, a, b); - return true; -} - -/* * Returns true if the server major ids match */ static bool @@ -605,36 +606,8 @@ nfs4_check_serverowner_major_id(struct nfs41_server_owner *o1, struct nfs41_server_owner *o2) { if (o1->major_id_sz != o2->major_id_sz) - goto out_major_mismatch; - if (memcmp(o1->major_id, o2->major_id, o1->major_id_sz) != 0) - goto out_major_mismatch; - - dprintk("NFS: --> %s server owner major IDs match\n", __func__); - return true; - -out_major_mismatch: - dprintk("NFS: --> %s server owner major IDs do not match\n", - __func__); - return false; -} - -/* - * Returns true if server minor ids match - */ -static bool -nfs4_check_serverowner_minor_id(struct nfs41_server_owner *o1, - struct nfs41_server_owner *o2) -{ - /* Check eir_server_owner so_minor_id */ - if (o1->minor_id != o2->minor_id) - goto out_minor_mismatch; - - dprintk("NFS: --> %s server owner minor IDs match\n", __func__); - return true; - -out_minor_mismatch: - dprintk("NFS: --> %s server owner minor IDs do not match\n", __func__); - return false; + return false; + return memcmp(o1->major_id, o2->major_id, o1->major_id_sz) == 0; } /* @@ -645,18 +618,9 @@ nfs4_check_server_scope(struct nfs41_server_scope *s1, struct nfs41_server_scope *s2) { if (s1->server_scope_sz != s2->server_scope_sz) - goto out_scope_mismatch; - if (memcmp(s1->server_scope, s2->server_scope, - s1->server_scope_sz) != 0) - goto out_scope_mismatch; - - dprintk("NFS: --> %s server scopes match\n", __func__); - return true; - -out_scope_mismatch: - dprintk("NFS: --> %s server scopes do not match\n", - __func__); - return false; + return false; + return memcmp(s1->server_scope, s2->server_scope, + s1->server_scope_sz) == 0; } /** @@ -680,7 +644,7 @@ int nfs4_detect_session_trunking(struct nfs_client *clp, struct rpc_xprt *xprt) { /* Check eir_clientid */ - if (!nfs4_match_clientids(clp->cl_clientid, res->clientid)) + if (clp->cl_clientid != res->clientid) goto out_err; /* Check eir_server_owner so_major_id */ @@ -689,8 +653,7 @@ int nfs4_detect_session_trunking(struct nfs_client *clp, goto out_err; /* Check eir_server_owner so_minor_id */ - if (!nfs4_check_serverowner_minor_id(clp->cl_serverowner, - res->server_owner)) + if (clp->cl_serverowner->minor_id != res->server_owner->minor_id) goto out_err; /* Check eir_server_scope */ @@ -739,33 +702,10 @@ int nfs41_walk_client_list(struct nfs_client *new, if (pos == new) goto found; - if (pos->rpc_ops != new->rpc_ops) - continue; - - if (pos->cl_minorversion != new->cl_minorversion) - continue; - - /* If "pos" isn't marked ready, we can't trust the - * remaining fields in "pos", especially the client - * ID and serverowner fields. Wait for CREATE_SESSION - * to finish. */ - if (pos->cl_cons_state > NFS_CS_READY) { - atomic_inc(&pos->cl_count); - spin_unlock(&nn->nfs_client_lock); - - nfs_put_client(prev); - prev = pos; - - status = nfs_wait_client_init_complete(pos); - spin_lock(&nn->nfs_client_lock); - if (status < 0) - break; - status = -NFS4ERR_STALE_CLIENTID; - } - if (pos->cl_cons_state != NFS_CS_READY) - continue; - - if (!nfs4_match_clientids(pos->cl_clientid, new->cl_clientid)) + status = nfs4_match_client(pos, new, &prev, nn); + if (status < 0) + goto out; + if (status != 0) continue; /* @@ -777,23 +717,15 @@ int nfs41_walk_client_list(struct nfs_client *new, new->cl_serverowner)) continue; - /* Unlike NFSv4.0, we know that NFSv4.1 always uses the - * uniform string, however someone might switch the - * uniquifier string on us. - */ - if (!nfs4_match_client_owner_id(pos, new)) - continue; found: atomic_inc(&pos->cl_count); *result = pos; status = 0; - dprintk("NFS: <-- %s using nfs_client = %p ({%d})\n", - __func__, pos, atomic_read(&pos->cl_count)); break; } +out: spin_unlock(&nn->nfs_client_lock); - dprintk("NFS: <-- %s status = %d\n", __func__, status); nfs_put_client(prev); return status; } @@ -916,9 +848,6 @@ static int nfs4_set_client(struct nfs_server *server, .timeparms = timeparms, }; struct nfs_client *clp; - int error; - - dprintk("--> nfs4_set_client()\n"); if (server->flags & NFS_MOUNT_NORESVPORT) set_bit(NFS_CS_NORESVPORT, &cl_init.init_flags); @@ -927,15 +856,11 @@ static int nfs4_set_client(struct nfs_server *server, /* Allocate or find a client reference we can use */ clp = nfs_get_client(&cl_init); - if (IS_ERR(clp)) { - error = PTR_ERR(clp); - goto error; - } + if (IS_ERR(clp)) + return PTR_ERR(clp); - if (server->nfs_client == clp) { - error = -ELOOP; - goto error; - } + if (server->nfs_client == clp) + return -ELOOP; /* * Query for the lease time on clientid setup or renewal @@ -947,11 +872,7 @@ static int nfs4_set_client(struct nfs_server *server, set_bit(NFS_CS_CHECK_LEASE_TIME, &clp->cl_res_state); server->nfs_client = clp; - dprintk("<-- nfs4_set_client() = 0 [new %p]\n", clp); return 0; -error: - dprintk("<-- nfs4_set_client() = xerror %d\n", error); - return error; } /* @@ -982,7 +903,6 @@ struct nfs_client *nfs4_set_ds_client(struct nfs_server *mds_srv, .net = mds_clp->cl_net, .timeparms = &ds_timeout, }; - struct nfs_client *clp; char buf[INET6_ADDRSTRLEN + 1]; if (rpc_ntop(ds_addr, buf, sizeof(buf)) <= 0) @@ -998,10 +918,7 @@ struct nfs_client *nfs4_set_ds_client(struct nfs_server *mds_srv, * (section 13.1 RFC 5661). */ nfs_init_timeout_values(&ds_timeout, ds_proto, ds_timeo, ds_retrans); - clp = nfs_get_client(&cl_init); - - dprintk("<-- %s %p\n", __func__, clp); - return clp; + return nfs_get_client(&cl_init); } EXPORT_SYMBOL_GPL(nfs4_set_ds_client); @@ -1098,8 +1015,6 @@ static int nfs4_init_server(struct nfs_server *server, struct rpc_timeout timeparms; int error; - dprintk("--> nfs4_init_server()\n"); - nfs_init_timeout_values(&timeparms, data->nfs_server.protocol, data->timeo, data->retrans); @@ -1127,7 +1042,7 @@ static int nfs4_init_server(struct nfs_server *server, data->minorversion, data->net); if (error < 0) - goto error; + return error; if (data->rsize) server->rsize = nfs_block_size(data->rsize, NULL); @@ -1138,16 +1053,10 @@ static int nfs4_init_server(struct nfs_server *server, server->acregmax = data->acregmax * HZ; server->acdirmin = data->acdirmin * HZ; server->acdirmax = data->acdirmax * HZ; + server->port = data->nfs_server.port; - server->port = data->nfs_server.port; - - error = nfs_init_server_rpcclient(server, &timeparms, - data->selected_flavor); - -error: - /* Done */ - dprintk("<-- nfs4_init_server() = %d\n", error); - return error; + return nfs_init_server_rpcclient(server, &timeparms, + data->selected_flavor); } /* @@ -1163,8 +1072,6 @@ struct nfs_server *nfs4_create_server(struct nfs_mount_info *mount_info, bool auth_probe; int error; - dprintk("--> nfs4_create_server()\n"); - server = nfs_alloc_server(); if (!server) return ERR_PTR(-ENOMEM); @@ -1180,12 +1087,10 @@ struct nfs_server *nfs4_create_server(struct nfs_mount_info *mount_info, if (error < 0) goto error; - dprintk("<-- nfs4_create_server() = %p\n", server); return server; error: nfs_free_server(server); - dprintk("<-- nfs4_create_server() = error %d\n", error); return ERR_PTR(error); } @@ -1200,8 +1105,6 @@ struct nfs_server *nfs4_create_referral_server(struct nfs_clone_mount *data, bool auth_probe; int error; - dprintk("--> nfs4_create_referral_server()\n"); - server = nfs_alloc_server(); if (!server) return ERR_PTR(-ENOMEM); @@ -1235,12 +1138,10 @@ struct nfs_server *nfs4_create_referral_server(struct nfs_clone_mount *data, if (error < 0) goto error; - dprintk("<-- nfs_create_referral_server() = %p\n", server); return server; error: nfs_free_server(server); - dprintk("<-- nfs4_create_referral_server() = error %d\n", error); return ERR_PTR(error); } @@ -1300,31 +1201,16 @@ int nfs4_update_server(struct nfs_server *server, const char *hostname, struct sockaddr *localaddr = (struct sockaddr *)&address; int error; - dprintk("--> %s: move FSID %llx:%llx to \"%s\")\n", __func__, - (unsigned long long)server->fsid.major, - (unsigned long long)server->fsid.minor, - hostname); - error = rpc_switch_client_transport(clnt, &xargs, clnt->cl_timeout); - if (error != 0) { - dprintk("<-- %s(): rpc_switch_client_transport returned %d\n", - __func__, error); - goto out; - } + if (error != 0) + return error; error = rpc_localaddr(clnt, localaddr, sizeof(address)); - if (error != 0) { - dprintk("<-- %s(): rpc_localaddr returned %d\n", - __func__, error); - goto out; - } + if (error != 0) + return error; - error = -EAFNOSUPPORT; - if (rpc_ntop(localaddr, buf, sizeof(buf)) == 0) { - dprintk("<-- %s(): rpc_ntop returned %d\n", - __func__, error); - goto out; - } + if (rpc_ntop(localaddr, buf, sizeof(buf)) == 0) + return -EAFNOSUPPORT; nfs_server_remove_lists(server); error = nfs4_set_client(server, hostname, sap, salen, buf, @@ -1333,21 +1219,12 @@ int nfs4_update_server(struct nfs_server *server, const char *hostname, nfs_put_client(clp); if (error != 0) { nfs_server_insert_lists(server); - dprintk("<-- %s(): nfs4_set_client returned %d\n", - __func__, error); - goto out; + return error; } if (server->nfs_client->cl_hostname == NULL) server->nfs_client->cl_hostname = kstrdup(hostname, GFP_KERNEL); nfs_server_insert_lists(server); - error = nfs_probe_destination(server); - if (error < 0) - goto out; - - dprintk("<-- %s() succeeded\n", __func__); - -out: - return error; + return nfs_probe_destination(server); } diff --git a/fs/nfs/nfs4getroot.c b/fs/nfs/nfs4getroot.c index 039b3eb6d834..ac8406018962 100644 --- a/fs/nfs/nfs4getroot.c +++ b/fs/nfs/nfs4getroot.c @@ -14,8 +14,6 @@ int nfs4_get_rootfh(struct nfs_server *server, struct nfs_fh *mntfh, bool auth_p struct nfs_fsinfo fsinfo; int ret = -ENOMEM; - dprintk("--> nfs4_get_rootfh()\n"); - fsinfo.fattr = nfs_alloc_fattr(); if (fsinfo.fattr == NULL) goto out; @@ -38,6 +36,5 @@ int nfs4_get_rootfh(struct nfs_server *server, struct nfs_fh *mntfh, bool auth_p memcpy(&server->fsid, &fsinfo.fattr->fsid, sizeof(server->fsid)); out: nfs_free_fattr(fsinfo.fattr); - dprintk("<-- nfs4_get_rootfh() = %d\n", ret); return ret; } diff --git a/fs/nfs/nfs4namespace.c b/fs/nfs/nfs4namespace.c index d8b040bd9814..7d531da1bae3 100644 --- a/fs/nfs/nfs4namespace.c +++ b/fs/nfs/nfs4namespace.c @@ -340,7 +340,6 @@ static struct vfsmount *nfs_follow_referral(struct dentry *dentry, out: free_page((unsigned long) page); free_page((unsigned long) page2); - dprintk("%s: done\n", __func__); return mnt; } @@ -358,11 +357,9 @@ static struct vfsmount *nfs_do_refmount(struct rpc_clnt *client, struct dentry * int err; /* BUG_ON(IS_ROOT(dentry)); */ - dprintk("%s: enter\n", __func__); - page = alloc_page(GFP_KERNEL); if (page == NULL) - goto out; + return mnt; fs_locations = kmalloc(sizeof(struct nfs4_fs_locations), GFP_KERNEL); if (fs_locations == NULL) @@ -386,8 +383,6 @@ static struct vfsmount *nfs_do_refmount(struct rpc_clnt *client, struct dentry * out_free: __free_page(page); kfree(fs_locations); -out: - dprintk("%s: done\n", __func__); return mnt; } diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 201ca3f2c4ba..c08c46a3b8cd 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -698,7 +698,8 @@ static int nfs41_sequence_process(struct rpc_task *task, session = slot->table->session; if (slot->interrupted) { - slot->interrupted = 0; + if (res->sr_status != -NFS4ERR_DELAY) + slot->interrupted = 0; interrupted = true; } @@ -2300,8 +2301,10 @@ static int _nfs4_proc_open(struct nfs4_opendata *data) if (status != 0) return status; } - if (!(o_res->f_attr->valid & NFS_ATTR_FATTR)) + if (!(o_res->f_attr->valid & NFS_ATTR_FATTR)) { + nfs4_sequence_free_slot(&o_res->seq_res); nfs4_proc_getattr(server, &o_res->fh, o_res->f_attr, o_res->f_label); + } return 0; } @@ -3265,6 +3268,7 @@ static int _nfs4_server_capabilities(struct nfs_server *server, struct nfs_fh *f .rpc_resp = &res, }; int status; + int i; bitmask[0] = FATTR4_WORD0_SUPPORTED_ATTRS | FATTR4_WORD0_FH_EXPIRE_TYPE | @@ -3330,8 +3334,13 @@ static int _nfs4_server_capabilities(struct nfs_server *server, struct nfs_fh *f server->cache_consistency_bitmask[0] &= FATTR4_WORD0_CHANGE|FATTR4_WORD0_SIZE; server->cache_consistency_bitmask[1] &= FATTR4_WORD1_TIME_METADATA|FATTR4_WORD1_TIME_MODIFY; server->cache_consistency_bitmask[2] = 0; + + /* Avoid a regression due to buggy server */ + for (i = 0; i < ARRAY_SIZE(res.exclcreat_bitmask); i++) + res.exclcreat_bitmask[i] &= res.attr_bitmask[i]; memcpy(server->exclcreat_bitmask, res.exclcreat_bitmask, sizeof(server->exclcreat_bitmask)); + server->acl_bitmask = res.acl_bitmask; server->fh_expire_type = res.fh_expire_type; } @@ -4610,7 +4619,7 @@ static int nfs4_proc_pgio_rpc_prepare(struct rpc_task *task, return 0; if (nfs4_set_rw_stateid(&hdr->args.stateid, hdr->args.context, hdr->args.lock_context, - hdr->rw_ops->rw_mode) == -EIO) + hdr->rw_mode) == -EIO) return -EIO; if (unlikely(test_bit(NFS_CONTEXT_BAD, &hdr->args.context->flags))) return -EIO; @@ -4804,8 +4813,10 @@ static int nfs4_proc_async_renew(struct nfs_client *clp, struct rpc_cred *cred, if (!atomic_inc_not_zero(&clp->cl_count)) return -EIO; data = kmalloc(sizeof(*data), GFP_NOFS); - if (data == NULL) + if (data == NULL) { + nfs_put_client(clp); return -ENOMEM; + } data->client = clp; data->timestamp = jiffies; return rpc_call_async(clp->cl_rpcclient, &msg, RPC_TASK_TIMEOUT, @@ -5782,6 +5793,7 @@ struct nfs4_unlockdata { struct nfs_locku_res res; struct nfs4_lock_state *lsp; struct nfs_open_context *ctx; + struct nfs_lock_context *l_ctx; struct file_lock fl; struct nfs_server *server; unsigned long timestamp; @@ -5806,6 +5818,7 @@ static struct nfs4_unlockdata *nfs4_alloc_unlockdata(struct file_lock *fl, atomic_inc(&lsp->ls_count); /* Ensure we don't close file until we're done freeing locks! */ p->ctx = get_nfs_open_context(ctx); + p->l_ctx = nfs_get_lock_context(ctx); memcpy(&p->fl, fl, sizeof(p->fl)); p->server = NFS_SERVER(inode); return p; @@ -5816,6 +5829,7 @@ static void nfs4_locku_release_calldata(void *data) struct nfs4_unlockdata *calldata = data; nfs_free_seqid(calldata->arg.seqid); nfs4_put_lock_state(calldata->lsp); + nfs_put_lock_context(calldata->l_ctx); put_nfs_open_context(calldata->ctx); kfree(calldata); } @@ -5857,6 +5871,10 @@ static void nfs4_locku_prepare(struct rpc_task *task, void *data) { struct nfs4_unlockdata *calldata = data; + if (test_bit(NFS_CONTEXT_UNLOCK, &calldata->l_ctx->open_context->flags) && + nfs_async_iocounter_wait(task, calldata->l_ctx)) + return; + if (nfs_wait_on_sequence(calldata->arg.seqid, task) != 0) goto out_wait; nfs4_stateid_copy(&calldata->arg.stateid, &calldata->lsp->ls_stateid); @@ -5908,6 +5926,8 @@ static struct rpc_task *nfs4_do_unlck(struct file_lock *fl, * canceled lock is passed in, and it won't be an unlock. */ fl->fl_type = F_UNLCK; + if (fl->fl_flags & FL_CLOSE) + set_bit(NFS_CONTEXT_UNLOCK, &ctx->flags); data = nfs4_alloc_unlockdata(fl, ctx, lsp, seqid); if (data == NULL) { @@ -6445,9 +6465,6 @@ nfs4_proc_lock(struct file *filp, int cmd, struct file_lock *request) ctx = nfs_file_open_context(filp); state = ctx->state; - if (request->fl_start < 0 || request->fl_end < 0) - return -EINVAL; - if (IS_GETLK(cmd)) { if (state != NULL) return nfs4_proc_getlk(state, F_GETLK, request); @@ -6470,20 +6487,6 @@ nfs4_proc_lock(struct file *filp, int cmd, struct file_lock *request) !test_bit(NFS_STATE_POSIX_LOCKS, &state->flags)) return -ENOLCK; - /* - * Don't rely on the VFS having checked the file open mode, - * since it won't do this for flock() locks. - */ - switch (request->fl_type) { - case F_RDLCK: - if (!(filp->f_mode & FMODE_READ)) - return -EBADF; - break; - case F_WRLCK: - if (!(filp->f_mode & FMODE_WRITE)) - return -EBADF; - } - status = nfs4_set_lock_state(state, request); if (status != 0) return status; @@ -7155,8 +7158,6 @@ int nfs4_proc_bind_one_conn_to_session(struct rpc_clnt *clnt, }; struct rpc_task *task; - dprintk("--> %s\n", __func__); - nfs4_copy_sessionid(&args.sessionid, &clp->cl_session->sess_id); if (!(clp->cl_session->flags & SESSION4_BACK_CHAN)) args.dir = NFS4_CDFC4_FORE; @@ -7176,24 +7177,20 @@ int nfs4_proc_bind_one_conn_to_session(struct rpc_clnt *clnt, if (memcmp(res.sessionid.data, clp->cl_session->sess_id.data, NFS4_MAX_SESSIONID_LEN)) { dprintk("NFS: %s: Session ID mismatch\n", __func__); - status = -EIO; - goto out; + return -EIO; } if ((res.dir & args.dir) != res.dir || res.dir == 0) { dprintk("NFS: %s: Unexpected direction from server\n", __func__); - status = -EIO; - goto out; + return -EIO; } if (res.use_conn_in_rdma_mode != args.use_conn_in_rdma_mode) { dprintk("NFS: %s: Server returned RDMA mode = true\n", __func__); - status = -EIO; - goto out; + return -EIO; } } -out: - dprintk("<-- %s status= %d\n", __func__, status); + return status; } @@ -7459,15 +7456,16 @@ static int _nfs4_proc_exchange_id(struct nfs_client *clp, struct rpc_cred *cred, }; struct nfs41_exchange_id_data *calldata; struct rpc_task *task; - int status = -EIO; + int status; if (!atomic_inc_not_zero(&clp->cl_count)) - goto out; + return -EIO; - status = -ENOMEM; calldata = kzalloc(sizeof(*calldata), GFP_NOFS); - if (!calldata) - goto out; + if (!calldata) { + nfs_put_client(clp); + return -ENOMEM; + } if (!xprt) nfs4_init_boot_verifier(clp, &verifier); @@ -7476,10 +7474,6 @@ static int _nfs4_proc_exchange_id(struct nfs_client *clp, struct rpc_cred *cred, if (status) goto out_calldata; - dprintk("NFS call exchange_id auth=%s, '%s'\n", - clp->cl_rpcclient->cl_auth->au_ops->au_name, - clp->cl_owner_id); - calldata->res.server_owner = kzalloc(sizeof(struct nfs41_server_owner), GFP_NOFS); status = -ENOMEM; @@ -7545,13 +7539,6 @@ static int _nfs4_proc_exchange_id(struct nfs_client *clp, struct rpc_cred *cred, rpc_put_task(task); out: - if (clp->cl_implid != NULL) - dprintk("NFS reply exchange_id: Server Implementation ID: " - "domain: %s, name: %s, date: %llu,%u\n", - clp->cl_implid->domain, clp->cl_implid->name, - clp->cl_implid->date.seconds, - clp->cl_implid->date.nseconds); - dprintk("NFS reply exchange_id: %d\n", status); return status; out_impl_id: @@ -7769,17 +7756,13 @@ int nfs4_proc_get_lease_time(struct nfs_client *clp, struct nfs_fsinfo *fsinfo) nfs4_init_sequence(&args.la_seq_args, &res.lr_seq_res, 0); nfs4_set_sequence_privileged(&args.la_seq_args); - dprintk("--> %s\n", __func__); task = rpc_run_task(&task_setup); if (IS_ERR(task)) - status = PTR_ERR(task); - else { - status = task->tk_status; - rpc_put_task(task); - } - dprintk("<-- %s return %d\n", __func__, status); + return PTR_ERR(task); + status = task->tk_status; + rpc_put_task(task); return status; } @@ -8180,6 +8163,12 @@ static int nfs41_reclaim_complete_handle_errors(struct rpc_task *task, struct nf /* fall through */ case -NFS4ERR_RETRY_UNCACHED_REP: return -EAGAIN; + case -NFS4ERR_BADSESSION: + case -NFS4ERR_DEADSESSION: + case -NFS4ERR_CONN_NOT_BOUND_TO_SESSION: + nfs4_schedule_session_recovery(clp->cl_session, + task->tk_status); + break; default: nfs4_schedule_lease_recovery(clp); } @@ -8258,7 +8247,6 @@ static int nfs41_proc_reclaim_complete(struct nfs_client *clp, if (status == 0) status = task->tk_status; rpc_put_task(task); - return 0; out: dprintk("<-- %s status=%d\n", __func__, status); return status; @@ -8357,6 +8345,7 @@ nfs4_layoutget_handle_exception(struct rpc_task *task, */ pnfs_mark_layout_stateid_invalid(lo, &head); spin_unlock(&inode->i_lock); + nfs_commit_inode(inode, 0); pnfs_free_lseg_list(&head); status = -EAGAIN; goto out; diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index 8156bad6b441..b34de036501b 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -1649,13 +1649,14 @@ static void nfs4_state_start_reclaim_reboot(struct nfs_client *clp) nfs4_state_mark_reclaim_helper(clp, nfs4_state_mark_reclaim_reboot); } -static void nfs4_reclaim_complete(struct nfs_client *clp, +static int nfs4_reclaim_complete(struct nfs_client *clp, const struct nfs4_state_recovery_ops *ops, struct rpc_cred *cred) { /* Notify the server we're done reclaiming our state */ if (ops->reclaim_complete) - (void)ops->reclaim_complete(clp, cred); + return ops->reclaim_complete(clp, cred); + return 0; } static void nfs4_clear_reclaim_server(struct nfs_server *server) @@ -1702,13 +1703,16 @@ static void nfs4_state_end_reclaim_reboot(struct nfs_client *clp) { const struct nfs4_state_recovery_ops *ops; struct rpc_cred *cred; + int err; if (!nfs4_state_clear_reclaim_reboot(clp)) return; ops = clp->cl_mvops->reboot_recovery_ops; cred = nfs4_get_clid_cred(clp); - nfs4_reclaim_complete(clp, ops, cred); + err = nfs4_reclaim_complete(clp, ops, cred); put_rpccred(cred); + if (err == -NFS4ERR_CONN_NOT_BOUND_TO_SESSION) + set_bit(NFS4CLNT_RECLAIM_REBOOT, &clp->cl_state); } static void nfs4_state_start_reclaim_nograce(struct nfs_client *clp) diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index 80ce289eea05..3aebfdc82b30 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -1000,8 +1000,9 @@ static void encode_nfs4_verifier(struct xdr_stream *xdr, const nfs4_verifier *ve static void encode_attrs(struct xdr_stream *xdr, const struct iattr *iap, const struct nfs4_label *label, + const umode_t *umask, const struct nfs_server *server, - bool excl_check, const umode_t *umask) + const uint32_t attrmask[]) { char owner_name[IDMAP_NAMESZ]; char owner_group[IDMAP_NAMESZ]; @@ -1016,22 +1017,20 @@ static void encode_attrs(struct xdr_stream *xdr, const struct iattr *iap, /* * We reserve enough space to write the entire attribute buffer at once. */ - if (iap->ia_valid & ATTR_SIZE) { + if ((iap->ia_valid & ATTR_SIZE) && (attrmask[0] & FATTR4_WORD0_SIZE)) { bmval[0] |= FATTR4_WORD0_SIZE; len += 8; } - if (!(server->attr_bitmask[2] & FATTR4_WORD2_MODE_UMASK)) - umask = NULL; if (iap->ia_valid & ATTR_MODE) { - if (umask) { + if (umask && (attrmask[2] & FATTR4_WORD2_MODE_UMASK)) { bmval[2] |= FATTR4_WORD2_MODE_UMASK; len += 8; - } else { + } else if (attrmask[1] & FATTR4_WORD1_MODE) { bmval[1] |= FATTR4_WORD1_MODE; len += 4; } } - if (iap->ia_valid & ATTR_UID) { + if ((iap->ia_valid & ATTR_UID) && (attrmask[1] & FATTR4_WORD1_OWNER)) { owner_namelen = nfs_map_uid_to_name(server, iap->ia_uid, owner_name, IDMAP_NAMESZ); if (owner_namelen < 0) { dprintk("nfs: couldn't resolve uid %d to string\n", @@ -1044,7 +1043,8 @@ static void encode_attrs(struct xdr_stream *xdr, const struct iattr *iap, bmval[1] |= FATTR4_WORD1_OWNER; len += 4 + (XDR_QUADLEN(owner_namelen) << 2); } - if (iap->ia_valid & ATTR_GID) { + if ((iap->ia_valid & ATTR_GID) && + (attrmask[1] & FATTR4_WORD1_OWNER_GROUP)) { owner_grouplen = nfs_map_gid_to_group(server, iap->ia_gid, owner_group, IDMAP_NAMESZ); if (owner_grouplen < 0) { dprintk("nfs: couldn't resolve gid %d to string\n", @@ -1056,32 +1056,26 @@ static void encode_attrs(struct xdr_stream *xdr, const struct iattr *iap, bmval[1] |= FATTR4_WORD1_OWNER_GROUP; len += 4 + (XDR_QUADLEN(owner_grouplen) << 2); } - if (iap->ia_valid & ATTR_ATIME_SET) { - bmval[1] |= FATTR4_WORD1_TIME_ACCESS_SET; - len += 16; - } else if (iap->ia_valid & ATTR_ATIME) { - bmval[1] |= FATTR4_WORD1_TIME_ACCESS_SET; - len += 4; - } - if (iap->ia_valid & ATTR_MTIME_SET) { - bmval[1] |= FATTR4_WORD1_TIME_MODIFY_SET; - len += 16; - } else if (iap->ia_valid & ATTR_MTIME) { - bmval[1] |= FATTR4_WORD1_TIME_MODIFY_SET; - len += 4; + if (attrmask[1] & FATTR4_WORD1_TIME_ACCESS_SET) { + if (iap->ia_valid & ATTR_ATIME_SET) { + bmval[1] |= FATTR4_WORD1_TIME_ACCESS_SET; + len += 16; + } else if (iap->ia_valid & ATTR_ATIME) { + bmval[1] |= FATTR4_WORD1_TIME_ACCESS_SET; + len += 4; + } } - - if (excl_check) { - const u32 *excl_bmval = server->exclcreat_bitmask; - bmval[0] &= excl_bmval[0]; - bmval[1] &= excl_bmval[1]; - bmval[2] &= excl_bmval[2]; - - if (!(excl_bmval[2] & FATTR4_WORD2_SECURITY_LABEL)) - label = NULL; + if (attrmask[1] & FATTR4_WORD1_TIME_MODIFY_SET) { + if (iap->ia_valid & ATTR_MTIME_SET) { + bmval[1] |= FATTR4_WORD1_TIME_MODIFY_SET; + len += 16; + } else if (iap->ia_valid & ATTR_MTIME) { + bmval[1] |= FATTR4_WORD1_TIME_MODIFY_SET; + len += 4; + } } - if (label) { + if (label && (attrmask[2] & FATTR4_WORD2_SECURITY_LABEL)) { len += 4 + 4 + 4 + (XDR_QUADLEN(label->len) << 2); bmval[2] |= FATTR4_WORD2_SECURITY_LABEL; } @@ -1188,8 +1182,8 @@ static void encode_create(struct xdr_stream *xdr, const struct nfs4_create_arg * } encode_string(xdr, create->name->len, create->name->name); - encode_attrs(xdr, create->attrs, create->label, create->server, false, - &create->umask); + encode_attrs(xdr, create->attrs, create->label, &create->umask, + create->server, create->server->attr_bitmask); } static void encode_getattr_one(struct xdr_stream *xdr, uint32_t bitmap, struct compound_hdr *hdr) @@ -1409,13 +1403,13 @@ static inline void encode_createmode(struct xdr_stream *xdr, const struct nfs_op switch(arg->createmode) { case NFS4_CREATE_UNCHECKED: *p = cpu_to_be32(NFS4_CREATE_UNCHECKED); - encode_attrs(xdr, arg->u.attrs, arg->label, arg->server, false, - &arg->umask); + encode_attrs(xdr, arg->u.attrs, arg->label, &arg->umask, + arg->server, arg->server->attr_bitmask); break; case NFS4_CREATE_GUARDED: *p = cpu_to_be32(NFS4_CREATE_GUARDED); - encode_attrs(xdr, arg->u.attrs, arg->label, arg->server, false, - &arg->umask); + encode_attrs(xdr, arg->u.attrs, arg->label, &arg->umask, + arg->server, arg->server->attr_bitmask); break; case NFS4_CREATE_EXCLUSIVE: *p = cpu_to_be32(NFS4_CREATE_EXCLUSIVE); @@ -1424,8 +1418,8 @@ static inline void encode_createmode(struct xdr_stream *xdr, const struct nfs_op case NFS4_CREATE_EXCLUSIVE4_1: *p = cpu_to_be32(NFS4_CREATE_EXCLUSIVE4_1); encode_nfs4_verifier(xdr, &arg->u.verifier); - encode_attrs(xdr, arg->u.attrs, arg->label, arg->server, true, - &arg->umask); + encode_attrs(xdr, arg->u.attrs, arg->label, &arg->umask, + arg->server, arg->server->exclcreat_bitmask); } } @@ -1681,7 +1675,8 @@ static void encode_setattr(struct xdr_stream *xdr, const struct nfs_setattrargs { encode_op_hdr(xdr, OP_SETATTR, decode_setattr_maxsz, hdr); encode_nfs4_stateid(xdr, &arg->stateid); - encode_attrs(xdr, arg->iap, arg->label, server, false, NULL); + encode_attrs(xdr, arg->iap, arg->label, NULL, server, + server->attr_bitmask); } static void encode_setclientid(struct xdr_stream *xdr, const struct nfs4_setclientid *setclientid, struct compound_hdr *hdr) @@ -2005,16 +2000,10 @@ encode_layoutcommit(struct xdr_stream *xdr, *p++ = cpu_to_be32(0); /* Never send time_modify_changed */ *p++ = cpu_to_be32(NFS_SERVER(args->inode)->pnfs_curr_ld->id);/* type */ - if (NFS_SERVER(inode)->pnfs_curr_ld->encode_layoutcommit) { - NFS_SERVER(inode)->pnfs_curr_ld->encode_layoutcommit( - NFS_I(inode)->layout, xdr, args); - } else { - encode_uint32(xdr, args->layoutupdate_len); - if (args->layoutupdate_pages) { - xdr_write_pages(xdr, args->layoutupdate_pages, 0, - args->layoutupdate_len); - } - } + encode_uint32(xdr, args->layoutupdate_len); + if (args->layoutupdate_pages) + xdr_write_pages(xdr, args->layoutupdate_pages, 0, + args->layoutupdate_len); return 0; } @@ -2024,7 +2013,6 @@ encode_layoutreturn(struct xdr_stream *xdr, const struct nfs4_layoutreturn_args *args, struct compound_hdr *hdr) { - const struct pnfs_layoutdriver_type *lr_ops = NFS_SERVER(args->inode)->pnfs_curr_ld; __be32 *p; encode_op_hdr(xdr, OP_LAYOUTRETURN, decode_layoutreturn_maxsz, hdr); @@ -2041,8 +2029,6 @@ encode_layoutreturn(struct xdr_stream *xdr, spin_unlock(&args->inode->i_lock); if (args->ld_private->ops && args->ld_private->ops->encode) args->ld_private->ops->encode(xdr, args, args->ld_private); - else if (lr_ops->encode_layoutreturn) - lr_ops->encode_layoutreturn(xdr, args); else encode_uint32(xdr, 0); } @@ -5579,6 +5565,8 @@ static int decode_op_map(struct xdr_stream *xdr, struct nfs4_op_map *op_map) unsigned int i; p = xdr_inline_decode(xdr, 4); + if (!p) + return -EIO; bitmap_words = be32_to_cpup(p++); if (bitmap_words > NFS4_OP_MAP_NUM_WORDS) return -EIO; diff --git a/fs/nfs/objlayout/Kbuild b/fs/nfs/objlayout/Kbuild deleted file mode 100644 index ed30ea072bb8..000000000000 --- a/fs/nfs/objlayout/Kbuild +++ /dev/null @@ -1,5 +0,0 @@ -# -# Makefile for the pNFS Objects Layout Driver kernel module -# -objlayoutdriver-y := objio_osd.o pnfs_osd_xdr_cli.o objlayout.o -obj-$(CONFIG_PNFS_OBJLAYOUT) += objlayoutdriver.o diff --git a/fs/nfs/objlayout/objio_osd.c b/fs/nfs/objlayout/objio_osd.c deleted file mode 100644 index 049c1b1f2932..000000000000 --- a/fs/nfs/objlayout/objio_osd.c +++ /dev/null @@ -1,675 +0,0 @@ -/* - * pNFS Objects layout implementation over open-osd initiator library - * - * Copyright (C) 2009 Panasas Inc. [year of first publication] - * All rights reserved. - * - * Benny Halevy <bhalevy@panasas.com> - * Boaz Harrosh <ooo@electrozaur.com> - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License version 2 - * See the file COPYING included with this distribution for more details. - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions - * are met: - * - * 1. Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * 2. Redistributions in binary form must reproduce the above copyright - * notice, this list of conditions and the following disclaimer in the - * documentation and/or other materials provided with the distribution. - * 3. Neither the name of the Panasas company nor the names of its - * contributors may be used to endorse or promote products derived - * from this software without specific prior written permission. - * - * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED - * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF - * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE - * DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE - * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR - * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF - * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR - * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF - * LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING - * NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS - * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - */ - -#include <linux/module.h> -#include <scsi/osd_ore.h> - -#include "objlayout.h" -#include "../internal.h" - -#define NFSDBG_FACILITY NFSDBG_PNFS_LD - -struct objio_dev_ent { - struct nfs4_deviceid_node id_node; - struct ore_dev od; -}; - -static void -objio_free_deviceid_node(struct nfs4_deviceid_node *d) -{ - struct objio_dev_ent *de = container_of(d, struct objio_dev_ent, id_node); - - dprintk("%s: free od=%p\n", __func__, de->od.od); - osduld_put_device(de->od.od); - kfree_rcu(d, rcu); -} - -struct objio_segment { - struct pnfs_layout_segment lseg; - - struct ore_layout layout; - struct ore_components oc; -}; - -static inline struct objio_segment * -OBJIO_LSEG(struct pnfs_layout_segment *lseg) -{ - return container_of(lseg, struct objio_segment, lseg); -} - -struct objio_state { - /* Generic layer */ - struct objlayout_io_res oir; - - bool sync; - /*FIXME: Support for extra_bytes at ore_get_rw_state() */ - struct ore_io_state *ios; -}; - -/* Send and wait for a get_device_info of devices in the layout, - then look them up with the osd_initiator library */ -struct nfs4_deviceid_node * -objio_alloc_deviceid_node(struct nfs_server *server, struct pnfs_device *pdev, - gfp_t gfp_flags) -{ - struct pnfs_osd_deviceaddr *deviceaddr; - struct objio_dev_ent *ode = NULL; - struct osd_dev *od; - struct osd_dev_info odi; - bool retry_flag = true; - __be32 *p; - int err; - - deviceaddr = kzalloc(sizeof(*deviceaddr), gfp_flags); - if (!deviceaddr) - return NULL; - - p = page_address(pdev->pages[0]); - pnfs_osd_xdr_decode_deviceaddr(deviceaddr, p); - - odi.systemid_len = deviceaddr->oda_systemid.len; - if (odi.systemid_len > sizeof(odi.systemid)) { - dprintk("%s: odi.systemid_len > sizeof(systemid=%zd)\n", - __func__, sizeof(odi.systemid)); - err = -EINVAL; - goto out; - } else if (odi.systemid_len) - memcpy(odi.systemid, deviceaddr->oda_systemid.data, - odi.systemid_len); - odi.osdname_len = deviceaddr->oda_osdname.len; - odi.osdname = (u8 *)deviceaddr->oda_osdname.data; - - if (!odi.osdname_len && !odi.systemid_len) { - dprintk("%s: !odi.osdname_len && !odi.systemid_len\n", - __func__); - err = -ENODEV; - goto out; - } - -retry_lookup: - od = osduld_info_lookup(&odi); - if (IS_ERR(od)) { - err = PTR_ERR(od); - dprintk("%s: osduld_info_lookup => %d\n", __func__, err); - if (err == -ENODEV && retry_flag) { - err = objlayout_autologin(deviceaddr); - if (likely(!err)) { - retry_flag = false; - goto retry_lookup; - } - } - goto out; - } - - dprintk("Adding new dev_id(%llx:%llx)\n", - _DEVID_LO(&pdev->dev_id), _DEVID_HI(&pdev->dev_id)); - - ode = kzalloc(sizeof(*ode), gfp_flags); - if (!ode) { - dprintk("%s: -ENOMEM od=%p\n", __func__, od); - goto out; - } - - nfs4_init_deviceid_node(&ode->id_node, server, &pdev->dev_id); - kfree(deviceaddr); - - ode->od.od = od; - return &ode->id_node; - -out: - kfree(deviceaddr); - return NULL; -} - -static void copy_single_comp(struct ore_components *oc, unsigned c, - struct pnfs_osd_object_cred *src_comp) -{ - struct ore_comp *ocomp = &oc->comps[c]; - - WARN_ON(src_comp->oc_cap_key.cred_len > 0); /* libosd is NO_SEC only */ - WARN_ON(src_comp->oc_cap.cred_len > sizeof(ocomp->cred)); - - ocomp->obj.partition = src_comp->oc_object_id.oid_partition_id; - ocomp->obj.id = src_comp->oc_object_id.oid_object_id; - - memcpy(ocomp->cred, src_comp->oc_cap.cred, sizeof(ocomp->cred)); -} - -static int __alloc_objio_seg(unsigned numdevs, gfp_t gfp_flags, - struct objio_segment **pseg) -{ -/* This is the in memory structure of the objio_segment - * - * struct __alloc_objio_segment { - * struct objio_segment olseg; - * struct ore_dev *ods[numdevs]; - * struct ore_comp comps[numdevs]; - * } *aolseg; - * NOTE: The code as above compiles and runs perfectly. It is elegant, - * type safe and compact. At some Past time Linus has decided he does not - * like variable length arrays, For the sake of this principal we uglify - * the code as below. - */ - struct objio_segment *lseg; - size_t lseg_size = sizeof(*lseg) + - numdevs * sizeof(lseg->oc.ods[0]) + - numdevs * sizeof(*lseg->oc.comps); - - lseg = kzalloc(lseg_size, gfp_flags); - if (unlikely(!lseg)) { - dprintk("%s: Failed allocation numdevs=%d size=%zd\n", __func__, - numdevs, lseg_size); - return -ENOMEM; - } - - lseg->oc.numdevs = numdevs; - lseg->oc.single_comp = EC_MULTPLE_COMPS; - lseg->oc.ods = (void *)(lseg + 1); - lseg->oc.comps = (void *)(lseg->oc.ods + numdevs); - - *pseg = lseg; - return 0; -} - -int objio_alloc_lseg(struct pnfs_layout_segment **outp, - struct pnfs_layout_hdr *pnfslay, - struct pnfs_layout_range *range, - struct xdr_stream *xdr, - gfp_t gfp_flags) -{ - struct nfs_server *server = NFS_SERVER(pnfslay->plh_inode); - struct objio_segment *objio_seg; - struct pnfs_osd_xdr_decode_layout_iter iter; - struct pnfs_osd_layout layout; - struct pnfs_osd_object_cred src_comp; - unsigned cur_comp; - int err; - - err = pnfs_osd_xdr_decode_layout_map(&layout, &iter, xdr); - if (unlikely(err)) - return err; - - err = __alloc_objio_seg(layout.olo_num_comps, gfp_flags, &objio_seg); - if (unlikely(err)) - return err; - - objio_seg->layout.stripe_unit = layout.olo_map.odm_stripe_unit; - objio_seg->layout.group_width = layout.olo_map.odm_group_width; - objio_seg->layout.group_depth = layout.olo_map.odm_group_depth; - objio_seg->layout.mirrors_p1 = layout.olo_map.odm_mirror_cnt + 1; - objio_seg->layout.raid_algorithm = layout.olo_map.odm_raid_algorithm; - - err = ore_verify_layout(layout.olo_map.odm_num_comps, - &objio_seg->layout); - if (unlikely(err)) - goto err; - - objio_seg->oc.first_dev = layout.olo_comps_index; - cur_comp = 0; - while (pnfs_osd_xdr_decode_layout_comp(&src_comp, &iter, xdr, &err)) { - struct nfs4_deviceid_node *d; - struct objio_dev_ent *ode; - - copy_single_comp(&objio_seg->oc, cur_comp, &src_comp); - - d = nfs4_find_get_deviceid(server, - &src_comp.oc_object_id.oid_device_id, - pnfslay->plh_lc_cred, gfp_flags); - if (!d) { - err = -ENXIO; - goto err; - } - - ode = container_of(d, struct objio_dev_ent, id_node); - objio_seg->oc.ods[cur_comp++] = &ode->od; - } - /* pnfs_osd_xdr_decode_layout_comp returns false on error */ - if (unlikely(err)) - goto err; - - *outp = &objio_seg->lseg; - return 0; - -err: - kfree(objio_seg); - dprintk("%s: Error: return %d\n", __func__, err); - *outp = NULL; - return err; -} - -void objio_free_lseg(struct pnfs_layout_segment *lseg) -{ - int i; - struct objio_segment *objio_seg = OBJIO_LSEG(lseg); - - for (i = 0; i < objio_seg->oc.numdevs; i++) { - struct ore_dev *od = objio_seg->oc.ods[i]; - struct objio_dev_ent *ode; - - if (!od) - break; - ode = container_of(od, typeof(*ode), od); - nfs4_put_deviceid_node(&ode->id_node); - } - kfree(objio_seg); -} - -static int -objio_alloc_io_state(struct pnfs_layout_hdr *pnfs_layout_type, bool is_reading, - struct pnfs_layout_segment *lseg, struct page **pages, unsigned pgbase, - loff_t offset, size_t count, void *rpcdata, gfp_t gfp_flags, - struct objio_state **outp) -{ - struct objio_segment *objio_seg = OBJIO_LSEG(lseg); - struct ore_io_state *ios; - int ret; - struct __alloc_objio_state { - struct objio_state objios; - struct pnfs_osd_ioerr ioerrs[objio_seg->oc.numdevs]; - } *aos; - - aos = kzalloc(sizeof(*aos), gfp_flags); - if (unlikely(!aos)) - return -ENOMEM; - - objlayout_init_ioerrs(&aos->objios.oir, objio_seg->oc.numdevs, - aos->ioerrs, rpcdata, pnfs_layout_type); - - ret = ore_get_rw_state(&objio_seg->layout, &objio_seg->oc, is_reading, - offset, count, &ios); - if (unlikely(ret)) { - kfree(aos); - return ret; - } - - ios->pages = pages; - ios->pgbase = pgbase; - ios->private = aos; - BUG_ON(ios->nr_pages > (pgbase + count + PAGE_SIZE - 1) >> PAGE_SHIFT); - - aos->objios.sync = 0; - aos->objios.ios = ios; - *outp = &aos->objios; - return 0; -} - -void objio_free_result(struct objlayout_io_res *oir) -{ - struct objio_state *objios = container_of(oir, struct objio_state, oir); - - ore_put_io_state(objios->ios); - kfree(objios); -} - -static enum pnfs_osd_errno osd_pri_2_pnfs_err(enum osd_err_priority oep) -{ - switch (oep) { - case OSD_ERR_PRI_NO_ERROR: - return (enum pnfs_osd_errno)0; - - case OSD_ERR_PRI_CLEAR_PAGES: - BUG_ON(1); - return 0; - - case OSD_ERR_PRI_RESOURCE: - return PNFS_OSD_ERR_RESOURCE; - case OSD_ERR_PRI_BAD_CRED: - return PNFS_OSD_ERR_BAD_CRED; - case OSD_ERR_PRI_NO_ACCESS: - return PNFS_OSD_ERR_NO_ACCESS; - case OSD_ERR_PRI_UNREACHABLE: - return PNFS_OSD_ERR_UNREACHABLE; - case OSD_ERR_PRI_NOT_FOUND: - return PNFS_OSD_ERR_NOT_FOUND; - case OSD_ERR_PRI_NO_SPACE: - return PNFS_OSD_ERR_NO_SPACE; - default: - WARN_ON(1); - /* fallthrough */ - case OSD_ERR_PRI_EIO: - return PNFS_OSD_ERR_EIO; - } -} - -static void __on_dev_error(struct ore_io_state *ios, - struct ore_dev *od, unsigned dev_index, enum osd_err_priority oep, - u64 dev_offset, u64 dev_len) -{ - struct objio_state *objios = ios->private; - struct pnfs_osd_objid pooid; - struct objio_dev_ent *ode = container_of(od, typeof(*ode), od); - /* FIXME: what to do with more-then-one-group layouts. We need to - * translate from ore_io_state index to oc->comps index - */ - unsigned comp = dev_index; - - pooid.oid_device_id = ode->id_node.deviceid; - pooid.oid_partition_id = ios->oc->comps[comp].obj.partition; - pooid.oid_object_id = ios->oc->comps[comp].obj.id; - - objlayout_io_set_result(&objios->oir, comp, - &pooid, osd_pri_2_pnfs_err(oep), - dev_offset, dev_len, !ios->reading); -} - -/* - * read - */ -static void _read_done(struct ore_io_state *ios, void *private) -{ - struct objio_state *objios = private; - ssize_t status; - int ret = ore_check_io(ios, &__on_dev_error); - - /* FIXME: _io_free(ios) can we dealocate the libosd resources; */ - - if (likely(!ret)) - status = ios->length; - else - status = ret; - - objlayout_read_done(&objios->oir, status, objios->sync); -} - -int objio_read_pagelist(struct nfs_pgio_header *hdr) -{ - struct objio_state *objios; - int ret; - - ret = objio_alloc_io_state(NFS_I(hdr->inode)->layout, true, - hdr->lseg, hdr->args.pages, hdr->args.pgbase, - hdr->args.offset, hdr->args.count, hdr, - GFP_KERNEL, &objios); - if (unlikely(ret)) - return ret; - - objios->ios->done = _read_done; - dprintk("%s: offset=0x%llx length=0x%x\n", __func__, - hdr->args.offset, hdr->args.count); - ret = ore_read(objios->ios); - if (unlikely(ret)) - objio_free_result(&objios->oir); - return ret; -} - -/* - * write - */ -static void _write_done(struct ore_io_state *ios, void *private) -{ - struct objio_state *objios = private; - ssize_t status; - int ret = ore_check_io(ios, &__on_dev_error); - - /* FIXME: _io_free(ios) can we dealocate the libosd resources; */ - - if (likely(!ret)) { - /* FIXME: should be based on the OSD's persistence model - * See OSD2r05 Section 4.13 Data persistence model */ - objios->oir.committed = NFS_FILE_SYNC; - status = ios->length; - } else { - status = ret; - } - - objlayout_write_done(&objios->oir, status, objios->sync); -} - -static struct page *__r4w_get_page(void *priv, u64 offset, bool *uptodate) -{ - struct objio_state *objios = priv; - struct nfs_pgio_header *hdr = objios->oir.rpcdata; - struct address_space *mapping = hdr->inode->i_mapping; - pgoff_t index = offset / PAGE_SIZE; - struct page *page; - loff_t i_size = i_size_read(hdr->inode); - - if (offset >= i_size) { - *uptodate = true; - dprintk("%s: g_zero_page index=0x%lx\n", __func__, index); - return ZERO_PAGE(0); - } - - page = find_get_page(mapping, index); - if (!page) { - page = find_or_create_page(mapping, index, GFP_NOFS); - if (unlikely(!page)) { - dprintk("%s: grab_cache_page Failed index=0x%lx\n", - __func__, index); - return NULL; - } - unlock_page(page); - } - *uptodate = PageUptodate(page); - dprintk("%s: index=0x%lx uptodate=%d\n", __func__, index, *uptodate); - return page; -} - -static void __r4w_put_page(void *priv, struct page *page) -{ - dprintk("%s: index=0x%lx\n", __func__, - (page == ZERO_PAGE(0)) ? -1UL : page->index); - if (ZERO_PAGE(0) != page) - put_page(page); - return; -} - -static const struct _ore_r4w_op _r4w_op = { - .get_page = &__r4w_get_page, - .put_page = &__r4w_put_page, -}; - -int objio_write_pagelist(struct nfs_pgio_header *hdr, int how) -{ - struct objio_state *objios; - int ret; - - ret = objio_alloc_io_state(NFS_I(hdr->inode)->layout, false, - hdr->lseg, hdr->args.pages, hdr->args.pgbase, - hdr->args.offset, hdr->args.count, hdr, GFP_NOFS, - &objios); - if (unlikely(ret)) - return ret; - - objios->sync = 0 != (how & FLUSH_SYNC); - objios->ios->r4w = &_r4w_op; - - if (!objios->sync) - objios->ios->done = _write_done; - - dprintk("%s: offset=0x%llx length=0x%x\n", __func__, - hdr->args.offset, hdr->args.count); - ret = ore_write(objios->ios); - if (unlikely(ret)) { - objio_free_result(&objios->oir); - return ret; - } - - if (objios->sync) - _write_done(objios->ios, objios); - - return 0; -} - -/* - * Return 0 if @req cannot be coalesced into @pgio, otherwise return the number - * of bytes (maximum @req->wb_bytes) that can be coalesced. - */ -static size_t objio_pg_test(struct nfs_pageio_descriptor *pgio, - struct nfs_page *prev, struct nfs_page *req) -{ - struct nfs_pgio_mirror *mirror = nfs_pgio_current_mirror(pgio); - unsigned int size; - - size = pnfs_generic_pg_test(pgio, prev, req); - - if (!size || mirror->pg_count + req->wb_bytes > - (unsigned long)pgio->pg_layout_private) - return 0; - - return min(size, req->wb_bytes); -} - -static void objio_init_read(struct nfs_pageio_descriptor *pgio, struct nfs_page *req) -{ - pnfs_generic_pg_init_read(pgio, req); - if (unlikely(pgio->pg_lseg == NULL)) - return; /* Not pNFS */ - - pgio->pg_layout_private = (void *) - OBJIO_LSEG(pgio->pg_lseg)->layout.max_io_length; -} - -static bool aligned_on_raid_stripe(u64 offset, struct ore_layout *layout, - unsigned long *stripe_end) -{ - u32 stripe_off; - unsigned stripe_size; - - if (layout->raid_algorithm == PNFS_OSD_RAID_0) - return true; - - stripe_size = layout->stripe_unit * - (layout->group_width - layout->parity); - - div_u64_rem(offset, stripe_size, &stripe_off); - if (!stripe_off) - return true; - - *stripe_end = stripe_size - stripe_off; - return false; -} - -static void objio_init_write(struct nfs_pageio_descriptor *pgio, struct nfs_page *req) -{ - unsigned long stripe_end = 0; - u64 wb_size; - - if (pgio->pg_dreq == NULL) - wb_size = i_size_read(pgio->pg_inode) - req_offset(req); - else - wb_size = nfs_dreq_bytes_left(pgio->pg_dreq); - - pnfs_generic_pg_init_write(pgio, req, wb_size); - if (unlikely(pgio->pg_lseg == NULL)) - return; /* Not pNFS */ - - if (req->wb_offset || - !aligned_on_raid_stripe(req->wb_index * PAGE_SIZE, - &OBJIO_LSEG(pgio->pg_lseg)->layout, - &stripe_end)) { - pgio->pg_layout_private = (void *)stripe_end; - } else { - pgio->pg_layout_private = (void *) - OBJIO_LSEG(pgio->pg_lseg)->layout.max_io_length; - } -} - -static const struct nfs_pageio_ops objio_pg_read_ops = { - .pg_init = objio_init_read, - .pg_test = objio_pg_test, - .pg_doio = pnfs_generic_pg_readpages, - .pg_cleanup = pnfs_generic_pg_cleanup, -}; - -static const struct nfs_pageio_ops objio_pg_write_ops = { - .pg_init = objio_init_write, - .pg_test = objio_pg_test, - .pg_doio = pnfs_generic_pg_writepages, - .pg_cleanup = pnfs_generic_pg_cleanup, -}; - -static struct pnfs_layoutdriver_type objlayout_type = { - .id = LAYOUT_OSD2_OBJECTS, - .name = "LAYOUT_OSD2_OBJECTS", - .flags = PNFS_LAYOUTRET_ON_SETATTR | - PNFS_LAYOUTRET_ON_ERROR, - - .max_deviceinfo_size = PAGE_SIZE, - .owner = THIS_MODULE, - .alloc_layout_hdr = objlayout_alloc_layout_hdr, - .free_layout_hdr = objlayout_free_layout_hdr, - - .alloc_lseg = objlayout_alloc_lseg, - .free_lseg = objlayout_free_lseg, - - .read_pagelist = objlayout_read_pagelist, - .write_pagelist = objlayout_write_pagelist, - .pg_read_ops = &objio_pg_read_ops, - .pg_write_ops = &objio_pg_write_ops, - - .sync = pnfs_generic_sync, - - .free_deviceid_node = objio_free_deviceid_node, - - .encode_layoutcommit = objlayout_encode_layoutcommit, - .encode_layoutreturn = objlayout_encode_layoutreturn, -}; - -MODULE_DESCRIPTION("pNFS Layout Driver for OSD2 objects"); -MODULE_AUTHOR("Benny Halevy <bhalevy@panasas.com>"); -MODULE_LICENSE("GPL"); - -static int __init -objlayout_init(void) -{ - int ret = pnfs_register_layoutdriver(&objlayout_type); - - if (ret) - printk(KERN_INFO - "NFS: %s: Registering OSD pNFS Layout Driver failed: error=%d\n", - __func__, ret); - else - printk(KERN_INFO "NFS: %s: Registered OSD pNFS Layout Driver\n", - __func__); - return ret; -} - -static void __exit -objlayout_exit(void) -{ - pnfs_unregister_layoutdriver(&objlayout_type); - printk(KERN_INFO "NFS: %s: Unregistered OSD pNFS Layout Driver\n", - __func__); -} - -MODULE_ALIAS("nfs-layouttype4-2"); - -module_init(objlayout_init); -module_exit(objlayout_exit); diff --git a/fs/nfs/objlayout/objlayout.c b/fs/nfs/objlayout/objlayout.c deleted file mode 100644 index 8f3d2acb81c3..000000000000 --- a/fs/nfs/objlayout/objlayout.c +++ /dev/null @@ -1,706 +0,0 @@ -/* - * pNFS Objects layout driver high level definitions - * - * Copyright (C) 2007 Panasas Inc. [year of first publication] - * All rights reserved. - * - * Benny Halevy <bhalevy@panasas.com> - * Boaz Harrosh <ooo@electrozaur.com> - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License version 2 - * See the file COPYING included with this distribution for more details. - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions - * are met: - * - * 1. Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * 2. Redistributions in binary form must reproduce the above copyright - * notice, this list of conditions and the following disclaimer in the - * documentation and/or other materials provided with the distribution. - * 3. Neither the name of the Panasas company nor the names of its - * contributors may be used to endorse or promote products derived - * from this software without specific prior written permission. - * - * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED - * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF - * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE - * DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE - * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR - * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF - * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR - * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF - * LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING - * NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS - * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - */ - -#include <linux/kmod.h> -#include <linux/moduleparam.h> -#include <linux/ratelimit.h> -#include <scsi/osd_initiator.h> -#include "objlayout.h" - -#define NFSDBG_FACILITY NFSDBG_PNFS_LD -/* - * Create a objlayout layout structure for the given inode and return it. - */ -struct pnfs_layout_hdr * -objlayout_alloc_layout_hdr(struct inode *inode, gfp_t gfp_flags) -{ - struct objlayout *objlay; - - objlay = kzalloc(sizeof(struct objlayout), gfp_flags); - if (!objlay) - return NULL; - spin_lock_init(&objlay->lock); - INIT_LIST_HEAD(&objlay->err_list); - dprintk("%s: Return %p\n", __func__, objlay); - return &objlay->pnfs_layout; -} - -/* - * Free an objlayout layout structure - */ -void -objlayout_free_layout_hdr(struct pnfs_layout_hdr *lo) -{ - struct objlayout *objlay = OBJLAYOUT(lo); - - dprintk("%s: objlay %p\n", __func__, objlay); - - WARN_ON(!list_empty(&objlay->err_list)); - kfree(objlay); -} - -/* - * Unmarshall layout and store it in pnfslay. - */ -struct pnfs_layout_segment * -objlayout_alloc_lseg(struct pnfs_layout_hdr *pnfslay, - struct nfs4_layoutget_res *lgr, - gfp_t gfp_flags) -{ - int status = -ENOMEM; - struct xdr_stream stream; - struct xdr_buf buf = { - .pages = lgr->layoutp->pages, - .page_len = lgr->layoutp->len, - .buflen = lgr->layoutp->len, - .len = lgr->layoutp->len, - }; - struct page *scratch; - struct pnfs_layout_segment *lseg; - - dprintk("%s: Begin pnfslay %p\n", __func__, pnfslay); - - scratch = alloc_page(gfp_flags); - if (!scratch) - goto err_nofree; - - xdr_init_decode(&stream, &buf, NULL); - xdr_set_scratch_buffer(&stream, page_address(scratch), PAGE_SIZE); - - status = objio_alloc_lseg(&lseg, pnfslay, &lgr->range, &stream, gfp_flags); - if (unlikely(status)) { - dprintk("%s: objio_alloc_lseg Return err %d\n", __func__, - status); - goto err; - } - - __free_page(scratch); - - dprintk("%s: Return %p\n", __func__, lseg); - return lseg; - -err: - __free_page(scratch); -err_nofree: - dprintk("%s: Err Return=>%d\n", __func__, status); - return ERR_PTR(status); -} - -/* - * Free a layout segement - */ -void -objlayout_free_lseg(struct pnfs_layout_segment *lseg) -{ - dprintk("%s: freeing layout segment %p\n", __func__, lseg); - - if (unlikely(!lseg)) - return; - - objio_free_lseg(lseg); -} - -/* - * I/O Operations - */ -static inline u64 -end_offset(u64 start, u64 len) -{ - u64 end; - - end = start + len; - return end >= start ? end : NFS4_MAX_UINT64; -} - -static void _fix_verify_io_params(struct pnfs_layout_segment *lseg, - struct page ***p_pages, unsigned *p_pgbase, - u64 offset, unsigned long count) -{ - u64 lseg_end_offset; - - BUG_ON(offset < lseg->pls_range.offset); - lseg_end_offset = end_offset(lseg->pls_range.offset, - lseg->pls_range.length); - BUG_ON(offset >= lseg_end_offset); - WARN_ON(offset + count > lseg_end_offset); - - if (*p_pgbase > PAGE_SIZE) { - dprintk("%s: pgbase(0x%x) > PAGE_SIZE\n", __func__, *p_pgbase); - *p_pages += *p_pgbase >> PAGE_SHIFT; - *p_pgbase &= ~PAGE_MASK; - } -} - -/* - * I/O done common code - */ -static void -objlayout_iodone(struct objlayout_io_res *oir) -{ - if (likely(oir->status >= 0)) { - objio_free_result(oir); - } else { - struct objlayout *objlay = oir->objlay; - - spin_lock(&objlay->lock); - objlay->delta_space_valid = OBJ_DSU_INVALID; - list_add(&objlay->err_list, &oir->err_list); - spin_unlock(&objlay->lock); - } -} - -/* - * objlayout_io_set_result - Set an osd_error code on a specific osd comp. - * - * The @index component IO failed (error returned from target). Register - * the error for later reporting at layout-return. - */ -void -objlayout_io_set_result(struct objlayout_io_res *oir, unsigned index, - struct pnfs_osd_objid *pooid, int osd_error, - u64 offset, u64 length, bool is_write) -{ - struct pnfs_osd_ioerr *ioerr = &oir->ioerrs[index]; - - BUG_ON(index >= oir->num_comps); - if (osd_error) { - ioerr->oer_component = *pooid; - ioerr->oer_comp_offset = offset; - ioerr->oer_comp_length = length; - ioerr->oer_iswrite = is_write; - ioerr->oer_errno = osd_error; - - dprintk("%s: err[%d]: errno=%d is_write=%d dev(%llx:%llx) " - "par=0x%llx obj=0x%llx offset=0x%llx length=0x%llx\n", - __func__, index, ioerr->oer_errno, - ioerr->oer_iswrite, - _DEVID_LO(&ioerr->oer_component.oid_device_id), - _DEVID_HI(&ioerr->oer_component.oid_device_id), - ioerr->oer_component.oid_partition_id, - ioerr->oer_component.oid_object_id, - ioerr->oer_comp_offset, - ioerr->oer_comp_length); - } else { - /* User need not call if no error is reported */ - ioerr->oer_errno = 0; - } -} - -/* Function scheduled on rpc workqueue to call ->nfs_readlist_complete(). - * This is because the osd completion is called with ints-off from - * the block layer - */ -static void _rpc_read_complete(struct work_struct *work) -{ - struct rpc_task *task; - struct nfs_pgio_header *hdr; - - dprintk("%s enter\n", __func__); - task = container_of(work, struct rpc_task, u.tk_work); - hdr = container_of(task, struct nfs_pgio_header, task); - - pnfs_ld_read_done(hdr); -} - -void -objlayout_read_done(struct objlayout_io_res *oir, ssize_t status, bool sync) -{ - struct nfs_pgio_header *hdr = oir->rpcdata; - - oir->status = hdr->task.tk_status = status; - if (status >= 0) - hdr->res.count = status; - else - hdr->pnfs_error = status; - objlayout_iodone(oir); - /* must not use oir after this point */ - - dprintk("%s: Return status=%zd eof=%d sync=%d\n", __func__, - status, hdr->res.eof, sync); - - if (sync) - pnfs_ld_read_done(hdr); - else { - INIT_WORK(&hdr->task.u.tk_work, _rpc_read_complete); - schedule_work(&hdr->task.u.tk_work); - } -} - -/* - * Perform sync or async reads. - */ -enum pnfs_try_status -objlayout_read_pagelist(struct nfs_pgio_header *hdr) -{ - struct inode *inode = hdr->inode; - loff_t offset = hdr->args.offset; - size_t count = hdr->args.count; - int err; - loff_t eof; - - eof = i_size_read(inode); - if (unlikely(offset + count > eof)) { - if (offset >= eof) { - err = 0; - hdr->res.count = 0; - hdr->res.eof = 1; - /*FIXME: do we need to call pnfs_ld_read_done() */ - goto out; - } - count = eof - offset; - } - - hdr->res.eof = (offset + count) >= eof; - _fix_verify_io_params(hdr->lseg, &hdr->args.pages, - &hdr->args.pgbase, - hdr->args.offset, hdr->args.count); - - dprintk("%s: inode(%lx) offset 0x%llx count 0x%zx eof=%d\n", - __func__, inode->i_ino, offset, count, hdr->res.eof); - - err = objio_read_pagelist(hdr); - out: - if (unlikely(err)) { - hdr->pnfs_error = err; - dprintk("%s: Returned Error %d\n", __func__, err); - return PNFS_NOT_ATTEMPTED; - } - return PNFS_ATTEMPTED; -} - -/* Function scheduled on rpc workqueue to call ->nfs_writelist_complete(). - * This is because the osd completion is called with ints-off from - * the block layer - */ -static void _rpc_write_complete(struct work_struct *work) -{ - struct rpc_task *task; - struct nfs_pgio_header *hdr; - - dprintk("%s enter\n", __func__); - task = container_of(work, struct rpc_task, u.tk_work); - hdr = container_of(task, struct nfs_pgio_header, task); - - pnfs_ld_write_done(hdr); -} - -void -objlayout_write_done(struct objlayout_io_res *oir, ssize_t status, bool sync) -{ - struct nfs_pgio_header *hdr = oir->rpcdata; - - oir->status = hdr->task.tk_status = status; - if (status >= 0) { - hdr->res.count = status; - hdr->verf.committed = oir->committed; - } else { - hdr->pnfs_error = status; - } - objlayout_iodone(oir); - /* must not use oir after this point */ - - dprintk("%s: Return status %zd committed %d sync=%d\n", __func__, - status, hdr->verf.committed, sync); - - if (sync) - pnfs_ld_write_done(hdr); - else { - INIT_WORK(&hdr->task.u.tk_work, _rpc_write_complete); - schedule_work(&hdr->task.u.tk_work); - } -} - -/* - * Perform sync or async writes. - */ -enum pnfs_try_status -objlayout_write_pagelist(struct nfs_pgio_header *hdr, int how) -{ - int err; - - _fix_verify_io_params(hdr->lseg, &hdr->args.pages, - &hdr->args.pgbase, - hdr->args.offset, hdr->args.count); - - err = objio_write_pagelist(hdr, how); - if (unlikely(err)) { - hdr->pnfs_error = err; - dprintk("%s: Returned Error %d\n", __func__, err); - return PNFS_NOT_ATTEMPTED; - } - return PNFS_ATTEMPTED; -} - -void -objlayout_encode_layoutcommit(struct pnfs_layout_hdr *pnfslay, - struct xdr_stream *xdr, - const struct nfs4_layoutcommit_args *args) -{ - struct objlayout *objlay = OBJLAYOUT(pnfslay); - struct pnfs_osd_layoutupdate lou; - __be32 *start; - - dprintk("%s: Begin\n", __func__); - - spin_lock(&objlay->lock); - lou.dsu_valid = (objlay->delta_space_valid == OBJ_DSU_VALID); - lou.dsu_delta = objlay->delta_space_used; - objlay->delta_space_used = 0; - objlay->delta_space_valid = OBJ_DSU_INIT; - lou.olu_ioerr_flag = !list_empty(&objlay->err_list); - spin_unlock(&objlay->lock); - - start = xdr_reserve_space(xdr, 4); - - BUG_ON(pnfs_osd_xdr_encode_layoutupdate(xdr, &lou)); - - *start = cpu_to_be32((xdr->p - start - 1) * 4); - - dprintk("%s: Return delta_space_used %lld err %d\n", __func__, - lou.dsu_delta, lou.olu_ioerr_flag); -} - -static int -err_prio(u32 oer_errno) -{ - switch (oer_errno) { - case 0: - return 0; - - case PNFS_OSD_ERR_RESOURCE: - return OSD_ERR_PRI_RESOURCE; - case PNFS_OSD_ERR_BAD_CRED: - return OSD_ERR_PRI_BAD_CRED; - case PNFS_OSD_ERR_NO_ACCESS: - return OSD_ERR_PRI_NO_ACCESS; - case PNFS_OSD_ERR_UNREACHABLE: - return OSD_ERR_PRI_UNREACHABLE; - case PNFS_OSD_ERR_NOT_FOUND: - return OSD_ERR_PRI_NOT_FOUND; - case PNFS_OSD_ERR_NO_SPACE: - return OSD_ERR_PRI_NO_SPACE; - default: - WARN_ON(1); - /* fallthrough */ - case PNFS_OSD_ERR_EIO: - return OSD_ERR_PRI_EIO; - } -} - -static void -merge_ioerr(struct pnfs_osd_ioerr *dest_err, - const struct pnfs_osd_ioerr *src_err) -{ - u64 dest_end, src_end; - - if (!dest_err->oer_errno) { - *dest_err = *src_err; - /* accumulated device must be blank */ - memset(&dest_err->oer_component.oid_device_id, 0, - sizeof(dest_err->oer_component.oid_device_id)); - - return; - } - - if (dest_err->oer_component.oid_partition_id != - src_err->oer_component.oid_partition_id) - dest_err->oer_component.oid_partition_id = 0; - - if (dest_err->oer_component.oid_object_id != - src_err->oer_component.oid_object_id) - dest_err->oer_component.oid_object_id = 0; - - if (dest_err->oer_comp_offset > src_err->oer_comp_offset) - dest_err->oer_comp_offset = src_err->oer_comp_offset; - - dest_end = end_offset(dest_err->oer_comp_offset, - dest_err->oer_comp_length); - src_end = end_offset(src_err->oer_comp_offset, - src_err->oer_comp_length); - if (dest_end < src_end) - dest_end = src_end; - - dest_err->oer_comp_length = dest_end - dest_err->oer_comp_offset; - - if ((src_err->oer_iswrite == dest_err->oer_iswrite) && - (err_prio(src_err->oer_errno) > err_prio(dest_err->oer_errno))) { - dest_err->oer_errno = src_err->oer_errno; - } else if (src_err->oer_iswrite) { - dest_err->oer_iswrite = true; - dest_err->oer_errno = src_err->oer_errno; - } -} - -static void -encode_accumulated_error(struct objlayout *objlay, __be32 *p) -{ - struct objlayout_io_res *oir, *tmp; - struct pnfs_osd_ioerr accumulated_err = {.oer_errno = 0}; - - list_for_each_entry_safe(oir, tmp, &objlay->err_list, err_list) { - unsigned i; - - for (i = 0; i < oir->num_comps; i++) { - struct pnfs_osd_ioerr *ioerr = &oir->ioerrs[i]; - - if (!ioerr->oer_errno) - continue; - - printk(KERN_ERR "NFS: %s: err[%d]: errno=%d " - "is_write=%d dev(%llx:%llx) par=0x%llx " - "obj=0x%llx offset=0x%llx length=0x%llx\n", - __func__, i, ioerr->oer_errno, - ioerr->oer_iswrite, - _DEVID_LO(&ioerr->oer_component.oid_device_id), - _DEVID_HI(&ioerr->oer_component.oid_device_id), - ioerr->oer_component.oid_partition_id, - ioerr->oer_component.oid_object_id, - ioerr->oer_comp_offset, - ioerr->oer_comp_length); - - merge_ioerr(&accumulated_err, ioerr); - } - list_del(&oir->err_list); - objio_free_result(oir); - } - - pnfs_osd_xdr_encode_ioerr(p, &accumulated_err); -} - -void -objlayout_encode_layoutreturn(struct xdr_stream *xdr, - const struct nfs4_layoutreturn_args *args) -{ - struct pnfs_layout_hdr *pnfslay = args->layout; - struct objlayout *objlay = OBJLAYOUT(pnfslay); - struct objlayout_io_res *oir, *tmp; - __be32 *start; - - dprintk("%s: Begin\n", __func__); - start = xdr_reserve_space(xdr, 4); - BUG_ON(!start); - - spin_lock(&objlay->lock); - - list_for_each_entry_safe(oir, tmp, &objlay->err_list, err_list) { - __be32 *last_xdr = NULL, *p; - unsigned i; - int res = 0; - - for (i = 0; i < oir->num_comps; i++) { - struct pnfs_osd_ioerr *ioerr = &oir->ioerrs[i]; - - if (!ioerr->oer_errno) - continue; - - dprintk("%s: err[%d]: errno=%d is_write=%d " - "dev(%llx:%llx) par=0x%llx obj=0x%llx " - "offset=0x%llx length=0x%llx\n", - __func__, i, ioerr->oer_errno, - ioerr->oer_iswrite, - _DEVID_LO(&ioerr->oer_component.oid_device_id), - _DEVID_HI(&ioerr->oer_component.oid_device_id), - ioerr->oer_component.oid_partition_id, - ioerr->oer_component.oid_object_id, - ioerr->oer_comp_offset, - ioerr->oer_comp_length); - - p = pnfs_osd_xdr_ioerr_reserve_space(xdr); - if (unlikely(!p)) { - res = -E2BIG; - break; /* accumulated_error */ - } - - last_xdr = p; - pnfs_osd_xdr_encode_ioerr(p, &oir->ioerrs[i]); - } - - /* TODO: use xdr_write_pages */ - if (unlikely(res)) { - /* no space for even one error descriptor */ - BUG_ON(!last_xdr); - - /* we've encountered a situation with lots and lots of - * errors and no space to encode them all. Use the last - * available slot to report the union of all the - * remaining errors. - */ - encode_accumulated_error(objlay, last_xdr); - goto loop_done; - } - list_del(&oir->err_list); - objio_free_result(oir); - } -loop_done: - spin_unlock(&objlay->lock); - - *start = cpu_to_be32((xdr->p - start - 1) * 4); - dprintk("%s: Return\n", __func__); -} - -enum { - OBJLAYOUT_MAX_URI_LEN = 256, OBJLAYOUT_MAX_OSDNAME_LEN = 64, - OBJLAYOUT_MAX_SYSID_HEX_LEN = OSD_SYSTEMID_LEN * 2 + 1, - OSD_LOGIN_UPCALL_PATHLEN = 256 -}; - -static char osd_login_prog[OSD_LOGIN_UPCALL_PATHLEN] = "/sbin/osd_login"; - -module_param_string(osd_login_prog, osd_login_prog, sizeof(osd_login_prog), - 0600); -MODULE_PARM_DESC(osd_login_prog, "Path to the osd_login upcall program"); - -struct __auto_login { - char uri[OBJLAYOUT_MAX_URI_LEN]; - char osdname[OBJLAYOUT_MAX_OSDNAME_LEN]; - char systemid_hex[OBJLAYOUT_MAX_SYSID_HEX_LEN]; -}; - -static int __objlayout_upcall(struct __auto_login *login) -{ - static char *envp[] = { "HOME=/", - "TERM=linux", - "PATH=/sbin:/usr/sbin:/bin:/usr/bin", - NULL - }; - char *argv[8]; - int ret; - - if (unlikely(!osd_login_prog[0])) { - dprintk("%s: osd_login_prog is disabled\n", __func__); - return -EACCES; - } - - dprintk("%s uri: %s\n", __func__, login->uri); - dprintk("%s osdname %s\n", __func__, login->osdname); - dprintk("%s systemid_hex %s\n", __func__, login->systemid_hex); - - argv[0] = (char *)osd_login_prog; - argv[1] = "-u"; - argv[2] = login->uri; - argv[3] = "-o"; - argv[4] = login->osdname; - argv[5] = "-s"; - argv[6] = login->systemid_hex; - argv[7] = NULL; - - ret = call_usermodehelper(argv[0], argv, envp, UMH_WAIT_PROC); - /* - * Disable the upcall mechanism if we're getting an ENOENT or - * EACCES error. The admin can re-enable it on the fly by using - * sysfs to set the objlayoutdriver.osd_login_prog module parameter once - * the problem has been fixed. - */ - if (ret == -ENOENT || ret == -EACCES) { - printk(KERN_ERR "PNFS-OBJ: %s was not found please set " - "objlayoutdriver.osd_login_prog kernel parameter!\n", - osd_login_prog); - osd_login_prog[0] = '\0'; - } - dprintk("%s %s return value: %d\n", __func__, osd_login_prog, ret); - - return ret; -} - -/* Assume dest is all zeros */ -static void __copy_nfsS_and_zero_terminate(struct nfs4_string s, - char *dest, int max_len, - const char *var_name) -{ - if (!s.len) - return; - - if (s.len >= max_len) { - pr_warn_ratelimited( - "objlayout_autologin: %s: s.len(%d) >= max_len(%d)", - var_name, s.len, max_len); - s.len = max_len - 1; /* space for null terminator */ - } - - memcpy(dest, s.data, s.len); -} - -/* Assume sysid is all zeros */ -static void _sysid_2_hex(struct nfs4_string s, - char sysid[OBJLAYOUT_MAX_SYSID_HEX_LEN]) -{ - int i; - char *cur; - - if (!s.len) - return; - - if (s.len != OSD_SYSTEMID_LEN) { - pr_warn_ratelimited( - "objlayout_autologin: systemid_len(%d) != OSD_SYSTEMID_LEN", - s.len); - if (s.len > OSD_SYSTEMID_LEN) - s.len = OSD_SYSTEMID_LEN; - } - - cur = sysid; - for (i = 0; i < s.len; i++) - cur = hex_byte_pack(cur, s.data[i]); -} - -int objlayout_autologin(struct pnfs_osd_deviceaddr *deviceaddr) -{ - int rc; - struct __auto_login login; - - if (!deviceaddr->oda_targetaddr.ota_netaddr.r_addr.len) - return -ENODEV; - - memset(&login, 0, sizeof(login)); - __copy_nfsS_and_zero_terminate( - deviceaddr->oda_targetaddr.ota_netaddr.r_addr, - login.uri, sizeof(login.uri), "URI"); - - __copy_nfsS_and_zero_terminate( - deviceaddr->oda_osdname, - login.osdname, sizeof(login.osdname), "OSDNAME"); - - _sysid_2_hex(deviceaddr->oda_systemid, login.systemid_hex); - - rc = __objlayout_upcall(&login); - if (rc > 0) /* script returns positive values */ - rc = -ENODEV; - - return rc; -} diff --git a/fs/nfs/objlayout/objlayout.h b/fs/nfs/objlayout/objlayout.h deleted file mode 100644 index fc94a5872ed4..000000000000 --- a/fs/nfs/objlayout/objlayout.h +++ /dev/null @@ -1,183 +0,0 @@ -/* - * Data types and function declerations for interfacing with the - * pNFS standard object layout driver. - * - * Copyright (C) 2007 Panasas Inc. [year of first publication] - * All rights reserved. - * - * Benny Halevy <bhalevy@panasas.com> - * Boaz Harrosh <ooo@electrozaur.com> - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License version 2 - * See the file COPYING included with this distribution for more details. - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions - * are met: - * - * 1. Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * 2. Redistributions in binary form must reproduce the above copyright - * notice, this list of conditions and the following disclaimer in the - * documentation and/or other materials provided with the distribution. - * 3. Neither the name of the Panasas company nor the names of its - * contributors may be used to endorse or promote products derived - * from this software without specific prior written permission. - * - * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED - * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF - * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE - * DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE - * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR - * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF - * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR - * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF - * LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING - * NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS - * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - */ - -#ifndef _OBJLAYOUT_H -#define _OBJLAYOUT_H - -#include <linux/nfs_fs.h> -#include <linux/pnfs_osd_xdr.h> -#include "../pnfs.h" - -/* - * per-inode layout - */ -struct objlayout { - struct pnfs_layout_hdr pnfs_layout; - - /* for layout_commit */ - enum osd_delta_space_valid_enum { - OBJ_DSU_INIT = 0, - OBJ_DSU_VALID, - OBJ_DSU_INVALID, - } delta_space_valid; - s64 delta_space_used; /* consumed by write ops */ - - /* for layout_return */ - spinlock_t lock; - struct list_head err_list; -}; - -static inline struct objlayout * -OBJLAYOUT(struct pnfs_layout_hdr *lo) -{ - return container_of(lo, struct objlayout, pnfs_layout); -} - -/* - * per-I/O operation state - * embedded in objects provider io_state data structure - */ -struct objlayout_io_res { - struct objlayout *objlay; - - void *rpcdata; - int status; /* res */ - int committed; /* res */ - - /* Error reporting (layout_return) */ - struct list_head err_list; - unsigned num_comps; - /* Pointer to array of error descriptors of size num_comps. - * It should contain as many entries as devices in the osd_layout - * that participate in the I/O. It is up to the io_engine to allocate - * needed space and set num_comps. - */ - struct pnfs_osd_ioerr *ioerrs; -}; - -static inline -void objlayout_init_ioerrs(struct objlayout_io_res *oir, unsigned num_comps, - struct pnfs_osd_ioerr *ioerrs, void *rpcdata, - struct pnfs_layout_hdr *pnfs_layout_type) -{ - oir->objlay = OBJLAYOUT(pnfs_layout_type); - oir->rpcdata = rpcdata; - INIT_LIST_HEAD(&oir->err_list); - oir->num_comps = num_comps; - oir->ioerrs = ioerrs; -} - -/* - * Raid engine I/O API - */ -extern int objio_alloc_lseg(struct pnfs_layout_segment **outp, - struct pnfs_layout_hdr *pnfslay, - struct pnfs_layout_range *range, - struct xdr_stream *xdr, - gfp_t gfp_flags); -extern void objio_free_lseg(struct pnfs_layout_segment *lseg); - -/* objio_free_result will free these @oir structs received from - * objlayout_{read,write}_done - */ -extern void objio_free_result(struct objlayout_io_res *oir); - -extern int objio_read_pagelist(struct nfs_pgio_header *rdata); -extern int objio_write_pagelist(struct nfs_pgio_header *wdata, int how); - -/* - * callback API - */ -extern void objlayout_io_set_result(struct objlayout_io_res *oir, - unsigned index, struct pnfs_osd_objid *pooid, - int osd_error, u64 offset, u64 length, bool is_write); - -static inline void -objlayout_add_delta_space_used(struct objlayout *objlay, s64 space_used) -{ - /* If one of the I/Os errored out and the delta_space_used was - * invalid we render the complete report as invalid. Protocol mandate - * the DSU be accurate or not reported. - */ - spin_lock(&objlay->lock); - if (objlay->delta_space_valid != OBJ_DSU_INVALID) { - objlay->delta_space_valid = OBJ_DSU_VALID; - objlay->delta_space_used += space_used; - } - spin_unlock(&objlay->lock); -} - -extern void objlayout_read_done(struct objlayout_io_res *oir, - ssize_t status, bool sync); -extern void objlayout_write_done(struct objlayout_io_res *oir, - ssize_t status, bool sync); - -/* - * exported generic objects function vectors - */ - -extern struct pnfs_layout_hdr *objlayout_alloc_layout_hdr(struct inode *, gfp_t gfp_flags); -extern void objlayout_free_layout_hdr(struct pnfs_layout_hdr *); - -extern struct pnfs_layout_segment *objlayout_alloc_lseg( - struct pnfs_layout_hdr *, - struct nfs4_layoutget_res *, - gfp_t gfp_flags); -extern void objlayout_free_lseg(struct pnfs_layout_segment *); - -extern enum pnfs_try_status objlayout_read_pagelist( - struct nfs_pgio_header *); - -extern enum pnfs_try_status objlayout_write_pagelist( - struct nfs_pgio_header *, - int how); - -extern void objlayout_encode_layoutcommit( - struct pnfs_layout_hdr *, - struct xdr_stream *, - const struct nfs4_layoutcommit_args *); - -extern void objlayout_encode_layoutreturn( - struct xdr_stream *, - const struct nfs4_layoutreturn_args *); - -extern int objlayout_autologin(struct pnfs_osd_deviceaddr *deviceaddr); - -#endif /* _OBJLAYOUT_H */ diff --git a/fs/nfs/objlayout/pnfs_osd_xdr_cli.c b/fs/nfs/objlayout/pnfs_osd_xdr_cli.c deleted file mode 100644 index f093c7ec983b..000000000000 --- a/fs/nfs/objlayout/pnfs_osd_xdr_cli.c +++ /dev/null @@ -1,415 +0,0 @@ -/* - * Object-Based pNFS Layout XDR layer - * - * Copyright (C) 2007 Panasas Inc. [year of first publication] - * All rights reserved. - * - * Benny Halevy <bhalevy@panasas.com> - * Boaz Harrosh <ooo@electrozaur.com> - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License version 2 - * See the file COPYING included with this distribution for more details. - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions - * are met: - * - * 1. Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * 2. Redistributions in binary form must reproduce the above copyright - * notice, this list of conditions and the following disclaimer in the - * documentation and/or other materials provided with the distribution. - * 3. Neither the name of the Panasas company nor the names of its - * contributors may be used to endorse or promote products derived - * from this software without specific prior written permission. - * - * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED - * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF - * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE - * DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE - * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR - * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF - * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR - * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF - * LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING - * NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS - * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - */ - -#include <linux/pnfs_osd_xdr.h> - -#define NFSDBG_FACILITY NFSDBG_PNFS_LD - -/* - * The following implementation is based on RFC5664 - */ - -/* - * struct pnfs_osd_objid { - * struct nfs4_deviceid oid_device_id; - * u64 oid_partition_id; - * u64 oid_object_id; - * }; // xdr size 32 bytes - */ -static __be32 * -_osd_xdr_decode_objid(__be32 *p, struct pnfs_osd_objid *objid) -{ - p = xdr_decode_opaque_fixed(p, objid->oid_device_id.data, - sizeof(objid->oid_device_id.data)); - - p = xdr_decode_hyper(p, &objid->oid_partition_id); - p = xdr_decode_hyper(p, &objid->oid_object_id); - return p; -} -/* - * struct pnfs_osd_opaque_cred { - * u32 cred_len; - * void *cred; - * }; // xdr size [variable] - * The return pointers are from the xdr buffer - */ -static int -_osd_xdr_decode_opaque_cred(struct pnfs_osd_opaque_cred *opaque_cred, - struct xdr_stream *xdr) -{ - __be32 *p = xdr_inline_decode(xdr, 1); - - if (!p) - return -EINVAL; - - opaque_cred->cred_len = be32_to_cpu(*p++); - - p = xdr_inline_decode(xdr, opaque_cred->cred_len); - if (!p) - return -EINVAL; - - opaque_cred->cred = p; - return 0; -} - -/* - * struct pnfs_osd_object_cred { - * struct pnfs_osd_objid oc_object_id; - * u32 oc_osd_version; - * u32 oc_cap_key_sec; - * struct pnfs_osd_opaque_cred oc_cap_key - * struct pnfs_osd_opaque_cred oc_cap; - * }; // xdr size 32 + 4 + 4 + [variable] + [variable] - */ -static int -_osd_xdr_decode_object_cred(struct pnfs_osd_object_cred *comp, - struct xdr_stream *xdr) -{ - __be32 *p = xdr_inline_decode(xdr, 32 + 4 + 4); - int ret; - - if (!p) - return -EIO; - - p = _osd_xdr_decode_objid(p, &comp->oc_object_id); - comp->oc_osd_version = be32_to_cpup(p++); - comp->oc_cap_key_sec = be32_to_cpup(p); - - ret = _osd_xdr_decode_opaque_cred(&comp->oc_cap_key, xdr); - if (unlikely(ret)) - return ret; - - ret = _osd_xdr_decode_opaque_cred(&comp->oc_cap, xdr); - return ret; -} - -/* - * struct pnfs_osd_data_map { - * u32 odm_num_comps; - * u64 odm_stripe_unit; - * u32 odm_group_width; - * u32 odm_group_depth; - * u32 odm_mirror_cnt; - * u32 odm_raid_algorithm; - * }; // xdr size 4 + 8 + 4 + 4 + 4 + 4 - */ -static inline int -_osd_data_map_xdr_sz(void) -{ - return 4 + 8 + 4 + 4 + 4 + 4; -} - -static __be32 * -_osd_xdr_decode_data_map(__be32 *p, struct pnfs_osd_data_map *data_map) -{ - data_map->odm_num_comps = be32_to_cpup(p++); - p = xdr_decode_hyper(p, &data_map->odm_stripe_unit); - data_map->odm_group_width = be32_to_cpup(p++); - data_map->odm_group_depth = be32_to_cpup(p++); - data_map->odm_mirror_cnt = be32_to_cpup(p++); - data_map->odm_raid_algorithm = be32_to_cpup(p++); - dprintk("%s: odm_num_comps=%u odm_stripe_unit=%llu odm_group_width=%u " - "odm_group_depth=%u odm_mirror_cnt=%u odm_raid_algorithm=%u\n", - __func__, - data_map->odm_num_comps, - (unsigned long long)data_map->odm_stripe_unit, - data_map->odm_group_width, - data_map->odm_group_depth, - data_map->odm_mirror_cnt, - data_map->odm_raid_algorithm); - return p; -} - -int pnfs_osd_xdr_decode_layout_map(struct pnfs_osd_layout *layout, - struct pnfs_osd_xdr_decode_layout_iter *iter, struct xdr_stream *xdr) -{ - __be32 *p; - - memset(iter, 0, sizeof(*iter)); - - p = xdr_inline_decode(xdr, _osd_data_map_xdr_sz() + 4 + 4); - if (unlikely(!p)) - return -EINVAL; - - p = _osd_xdr_decode_data_map(p, &layout->olo_map); - layout->olo_comps_index = be32_to_cpup(p++); - layout->olo_num_comps = be32_to_cpup(p++); - dprintk("%s: olo_comps_index=%d olo_num_comps=%d\n", __func__, - layout->olo_comps_index, layout->olo_num_comps); - - iter->total_comps = layout->olo_num_comps; - return 0; -} - -bool pnfs_osd_xdr_decode_layout_comp(struct pnfs_osd_object_cred *comp, - struct pnfs_osd_xdr_decode_layout_iter *iter, struct xdr_stream *xdr, - int *err) -{ - BUG_ON(iter->decoded_comps > iter->total_comps); - if (iter->decoded_comps == iter->total_comps) - return false; - - *err = _osd_xdr_decode_object_cred(comp, xdr); - if (unlikely(*err)) { - dprintk("%s: _osd_xdr_decode_object_cred=>%d decoded_comps=%d " - "total_comps=%d\n", __func__, *err, - iter->decoded_comps, iter->total_comps); - return false; /* stop the loop */ - } - dprintk("%s: dev(%llx:%llx) par=0x%llx obj=0x%llx " - "key_len=%u cap_len=%u\n", - __func__, - _DEVID_LO(&comp->oc_object_id.oid_device_id), - _DEVID_HI(&comp->oc_object_id.oid_device_id), - comp->oc_object_id.oid_partition_id, - comp->oc_object_id.oid_object_id, - comp->oc_cap_key.cred_len, comp->oc_cap.cred_len); - - iter->decoded_comps++; - return true; -} - -/* - * Get Device Information Decoding - * - * Note: since Device Information is currently done synchronously, all - * variable strings fields are left inside the rpc buffer and are only - * pointed to by the pnfs_osd_deviceaddr members. So the read buffer - * should not be freed while the returned information is in use. - */ -/* - *struct nfs4_string { - * unsigned int len; - * char *data; - *}; // size [variable] - * NOTE: Returned string points to inside the XDR buffer - */ -static __be32 * -__read_u8_opaque(__be32 *p, struct nfs4_string *str) -{ - str->len = be32_to_cpup(p++); - str->data = (char *)p; - - p += XDR_QUADLEN(str->len); - return p; -} - -/* - * struct pnfs_osd_targetid { - * u32 oti_type; - * struct nfs4_string oti_scsi_device_id; - * };// size 4 + [variable] - */ -static __be32 * -__read_targetid(__be32 *p, struct pnfs_osd_targetid* targetid) -{ - u32 oti_type; - - oti_type = be32_to_cpup(p++); - targetid->oti_type = oti_type; - - switch (oti_type) { - case OBJ_TARGET_SCSI_NAME: - case OBJ_TARGET_SCSI_DEVICE_ID: - p = __read_u8_opaque(p, &targetid->oti_scsi_device_id); - } - - return p; -} - -/* - * struct pnfs_osd_net_addr { - * struct nfs4_string r_netid; - * struct nfs4_string r_addr; - * }; - */ -static __be32 * -__read_net_addr(__be32 *p, struct pnfs_osd_net_addr* netaddr) -{ - p = __read_u8_opaque(p, &netaddr->r_netid); - p = __read_u8_opaque(p, &netaddr->r_addr); - - return p; -} - -/* - * struct pnfs_osd_targetaddr { - * u32 ota_available; - * struct pnfs_osd_net_addr ota_netaddr; - * }; - */ -static __be32 * -__read_targetaddr(__be32 *p, struct pnfs_osd_targetaddr *targetaddr) -{ - u32 ota_available; - - ota_available = be32_to_cpup(p++); - targetaddr->ota_available = ota_available; - - if (ota_available) - p = __read_net_addr(p, &targetaddr->ota_netaddr); - - - return p; -} - -/* - * struct pnfs_osd_deviceaddr { - * struct pnfs_osd_targetid oda_targetid; - * struct pnfs_osd_targetaddr oda_targetaddr; - * u8 oda_lun[8]; - * struct nfs4_string oda_systemid; - * struct pnfs_osd_object_cred oda_root_obj_cred; - * struct nfs4_string oda_osdname; - * }; - */ - -/* We need this version for the pnfs_osd_xdr_decode_deviceaddr which does - * not have an xdr_stream - */ -static __be32 * -__read_opaque_cred(__be32 *p, - struct pnfs_osd_opaque_cred *opaque_cred) -{ - opaque_cred->cred_len = be32_to_cpu(*p++); - opaque_cred->cred = p; - return p + XDR_QUADLEN(opaque_cred->cred_len); -} - -static __be32 * -__read_object_cred(__be32 *p, struct pnfs_osd_object_cred *comp) -{ - p = _osd_xdr_decode_objid(p, &comp->oc_object_id); - comp->oc_osd_version = be32_to_cpup(p++); - comp->oc_cap_key_sec = be32_to_cpup(p++); - - p = __read_opaque_cred(p, &comp->oc_cap_key); - p = __read_opaque_cred(p, &comp->oc_cap); - return p; -} - -void pnfs_osd_xdr_decode_deviceaddr( - struct pnfs_osd_deviceaddr *deviceaddr, __be32 *p) -{ - p = __read_targetid(p, &deviceaddr->oda_targetid); - - p = __read_targetaddr(p, &deviceaddr->oda_targetaddr); - - p = xdr_decode_opaque_fixed(p, deviceaddr->oda_lun, - sizeof(deviceaddr->oda_lun)); - - p = __read_u8_opaque(p, &deviceaddr->oda_systemid); - - p = __read_object_cred(p, &deviceaddr->oda_root_obj_cred); - - p = __read_u8_opaque(p, &deviceaddr->oda_osdname); - - /* libosd likes this terminated in dbg. It's last, so no problems */ - deviceaddr->oda_osdname.data[deviceaddr->oda_osdname.len] = 0; -} - -/* - * struct pnfs_osd_layoutupdate { - * u32 dsu_valid; - * s64 dsu_delta; - * u32 olu_ioerr_flag; - * }; xdr size 4 + 8 + 4 - */ -int -pnfs_osd_xdr_encode_layoutupdate(struct xdr_stream *xdr, - struct pnfs_osd_layoutupdate *lou) -{ - __be32 *p = xdr_reserve_space(xdr, 4 + 8 + 4); - - if (!p) - return -E2BIG; - - *p++ = cpu_to_be32(lou->dsu_valid); - if (lou->dsu_valid) - p = xdr_encode_hyper(p, lou->dsu_delta); - *p++ = cpu_to_be32(lou->olu_ioerr_flag); - return 0; -} - -/* - * struct pnfs_osd_objid { - * struct nfs4_deviceid oid_device_id; - * u64 oid_partition_id; - * u64 oid_object_id; - * }; // xdr size 32 bytes - */ -static inline __be32 * -pnfs_osd_xdr_encode_objid(__be32 *p, struct pnfs_osd_objid *object_id) -{ - p = xdr_encode_opaque_fixed(p, &object_id->oid_device_id.data, - sizeof(object_id->oid_device_id.data)); - p = xdr_encode_hyper(p, object_id->oid_partition_id); - p = xdr_encode_hyper(p, object_id->oid_object_id); - - return p; -} - -/* - * struct pnfs_osd_ioerr { - * struct pnfs_osd_objid oer_component; - * u64 oer_comp_offset; - * u64 oer_comp_length; - * u32 oer_iswrite; - * u32 oer_errno; - * }; // xdr size 32 + 24 bytes - */ -void pnfs_osd_xdr_encode_ioerr(__be32 *p, struct pnfs_osd_ioerr *ioerr) -{ - p = pnfs_osd_xdr_encode_objid(p, &ioerr->oer_component); - p = xdr_encode_hyper(p, ioerr->oer_comp_offset); - p = xdr_encode_hyper(p, ioerr->oer_comp_length); - *p++ = cpu_to_be32(ioerr->oer_iswrite); - *p = cpu_to_be32(ioerr->oer_errno); -} - -__be32 *pnfs_osd_xdr_ioerr_reserve_space(struct xdr_stream *xdr) -{ - __be32 *p; - - p = xdr_reserve_space(xdr, 32 + 24); - if (unlikely(!p)) - dprintk("%s: out of xdr space\n", __func__); - - return p; -} diff --git a/fs/nfs/pagelist.c b/fs/nfs/pagelist.c index 6e629b856a00..ad92b401326c 100644 --- a/fs/nfs/pagelist.c +++ b/fs/nfs/pagelist.c @@ -29,19 +29,6 @@ static struct kmem_cache *nfs_page_cachep; static const struct rpc_call_ops nfs_pgio_common_ops; -static bool nfs_pgarray_set(struct nfs_page_array *p, unsigned int pagecount) -{ - p->npages = pagecount; - if (pagecount <= ARRAY_SIZE(p->page_array)) - p->pagevec = p->page_array; - else { - p->pagevec = kcalloc(pagecount, sizeof(struct page *), GFP_KERNEL); - if (!p->pagevec) - p->npages = 0; - } - return p->pagevec != NULL; -} - struct nfs_pgio_mirror * nfs_pgio_current_mirror(struct nfs_pageio_descriptor *desc) { @@ -115,6 +102,35 @@ nfs_iocounter_wait(struct nfs_lock_context *l_ctx) TASK_KILLABLE); } +/** + * nfs_async_iocounter_wait - wait on a rpc_waitqueue for I/O + * to complete + * @task: the rpc_task that should wait + * @l_ctx: nfs_lock_context with io_counter to check + * + * Returns true if there is outstanding I/O to wait on and the + * task has been put to sleep. + */ +bool +nfs_async_iocounter_wait(struct rpc_task *task, struct nfs_lock_context *l_ctx) +{ + struct inode *inode = d_inode(l_ctx->open_context->dentry); + bool ret = false; + + if (atomic_read(&l_ctx->io_count) > 0) { + rpc_sleep_on(&NFS_SERVER(inode)->uoc_rpcwaitq, task, NULL); + ret = true; + } + + if (atomic_read(&l_ctx->io_count) == 0) { + rpc_wake_up_queued_task(&NFS_SERVER(inode)->uoc_rpcwaitq, task); + ret = false; + } + + return ret; +} +EXPORT_SYMBOL_GPL(nfs_async_iocounter_wait); + /* * nfs_page_group_lock - lock the head of the page group * @req - request in group that is to be locked @@ -398,8 +414,11 @@ static void nfs_clear_request(struct nfs_page *req) req->wb_page = NULL; } if (l_ctx != NULL) { - if (atomic_dec_and_test(&l_ctx->io_count)) + if (atomic_dec_and_test(&l_ctx->io_count)) { wake_up_atomic_t(&l_ctx->io_count); + if (test_bit(NFS_CONTEXT_UNLOCK, &ctx->flags)) + rpc_wake_up(&NFS_SERVER(d_inode(ctx->dentry))->uoc_rpcwaitq); + } nfs_put_lock_context(l_ctx); req->wb_lock_context = NULL; } @@ -677,7 +696,8 @@ void nfs_pageio_init(struct nfs_pageio_descriptor *desc, const struct nfs_pgio_completion_ops *compl_ops, const struct nfs_rw_ops *rw_ops, size_t bsize, - int io_flags) + int io_flags, + gfp_t gfp_flags) { struct nfs_pgio_mirror *new; int i; @@ -701,7 +721,7 @@ void nfs_pageio_init(struct nfs_pageio_descriptor *desc, /* until we have a request, we don't have an lseg and no * idea how many mirrors there will be */ new = kcalloc(NFS_PAGEIO_DESCRIPTOR_MIRROR_MAX, - sizeof(struct nfs_pgio_mirror), GFP_KERNEL); + sizeof(struct nfs_pgio_mirror), gfp_flags); desc->pg_mirrors_dynamic = new; desc->pg_mirrors = new; @@ -754,13 +774,24 @@ int nfs_generic_pgio(struct nfs_pageio_descriptor *desc, *last_page; struct list_head *head = &mirror->pg_list; struct nfs_commit_info cinfo; + struct nfs_page_array *pg_array = &hdr->page_array; unsigned int pagecount, pageused; + gfp_t gfp_flags = GFP_KERNEL; pagecount = nfs_page_array_len(mirror->pg_base, mirror->pg_count); - if (!nfs_pgarray_set(&hdr->page_array, pagecount)) { - nfs_pgio_error(hdr); - desc->pg_error = -ENOMEM; - return desc->pg_error; + + if (pagecount <= ARRAY_SIZE(pg_array->page_array)) + pg_array->pagevec = pg_array->page_array; + else { + if (hdr->rw_mode == FMODE_WRITE) + gfp_flags = GFP_NOIO; + pg_array->pagevec = kcalloc(pagecount, sizeof(struct page *), gfp_flags); + if (!pg_array->pagevec) { + pg_array->npages = 0; + nfs_pgio_error(hdr); + desc->pg_error = -ENOMEM; + return desc->pg_error; + } } nfs_init_cinfo(&cinfo, desc->pg_inode, desc->pg_dreq); @@ -1256,8 +1287,10 @@ void nfs_pageio_cond_complete(struct nfs_pageio_descriptor *desc, pgoff_t index) mirror = &desc->pg_mirrors[midx]; if (!list_empty(&mirror->pg_list)) { prev = nfs_list_entry(mirror->pg_list.prev); - if (index != prev->wb_index + 1) - nfs_pageio_complete_mirror(desc, midx); + if (index != prev->wb_index + 1) { + nfs_pageio_complete(desc); + break; + } } } } diff --git a/fs/nfs/pnfs.c b/fs/nfs/pnfs.c index dd042498ce7c..adc6ec28d4b5 100644 --- a/fs/nfs/pnfs.c +++ b/fs/nfs/pnfs.c @@ -322,9 +322,15 @@ pnfs_set_plh_return_info(struct pnfs_layout_hdr *lo, enum pnfs_iomode iomode, static void pnfs_clear_layoutreturn_info(struct pnfs_layout_hdr *lo) { + struct pnfs_layout_segment *lseg; lo->plh_return_iomode = 0; lo->plh_return_seq = 0; clear_bit(NFS_LAYOUT_RETURN_REQUESTED, &lo->plh_flags); + list_for_each_entry(lseg, &lo->plh_segs, pls_list) { + if (!test_bit(NFS_LSEG_LAYOUTRETURN, &lseg->pls_flags)) + continue; + pnfs_set_plh_return_info(lo, lseg->pls_range.iomode, 0); + } } static void pnfs_clear_layoutreturn_waitbit(struct pnfs_layout_hdr *lo) @@ -367,9 +373,9 @@ pnfs_mark_layout_stateid_invalid(struct pnfs_layout_hdr *lo, struct pnfs_layout_segment *lseg, *next; set_bit(NFS_LAYOUT_INVALID_STID, &lo->plh_flags); - pnfs_clear_layoutreturn_info(lo); list_for_each_entry_safe(lseg, next, &lo->plh_segs, pls_list) pnfs_clear_lseg_state(lseg, lseg_list); + pnfs_clear_layoutreturn_info(lo); pnfs_free_returned_lsegs(lo, lseg_list, &range, 0); if (test_bit(NFS_LAYOUT_RETURN, &lo->plh_flags) && !test_and_set_bit(NFS_LAYOUT_RETURN_LOCK, &lo->plh_flags)) @@ -563,7 +569,6 @@ pnfs_put_lseg_locked(struct pnfs_layout_segment *lseg) } } } -EXPORT_SYMBOL_GPL(pnfs_put_lseg_locked); /* * is l2 fully contained in l1? @@ -728,6 +733,7 @@ pnfs_destroy_layout(struct nfs_inode *nfsi) pnfs_layout_clear_fail_bit(lo, NFS_LAYOUT_RW_FAILED); spin_unlock(&nfsi->vfs_inode.i_lock); pnfs_free_lseg_list(&tmp_list); + nfs_commit_inode(&nfsi->vfs_inode, 0); pnfs_put_layout_hdr(lo); } else spin_unlock(&nfsi->vfs_inode.i_lock); @@ -1209,7 +1215,6 @@ out: dprintk("<-- %s status: %d\n", __func__, status); return status; } -EXPORT_SYMBOL_GPL(_pnfs_return_layout); int pnfs_commit_and_return_layout(struct inode *inode) @@ -1991,6 +1996,8 @@ out_forget: spin_unlock(&ino->i_lock); lseg->pls_layout = lo; NFS_SERVER(ino)->pnfs_curr_ld->free_lseg(lseg); + if (!pnfs_layout_is_valid(lo)) + nfs_commit_inode(ino, 0); return ERR_PTR(-EAGAIN); } @@ -2051,9 +2058,11 @@ void pnfs_error_mark_layout_for_return(struct inode *inode, bool return_now = false; spin_lock(&inode->i_lock); + if (!pnfs_layout_is_valid(lo)) { + spin_unlock(&inode->i_lock); + return; + } pnfs_set_plh_return_info(lo, range.iomode, 0); - /* Block LAYOUTGET */ - set_bit(NFS_LAYOUT_RETURN, &lo->plh_flags); /* * mark all matching lsegs so that we are sure to have no live * segments at hand when sending layoutreturn. See pnfs_put_lseg() @@ -2075,10 +2084,22 @@ void pnfs_error_mark_layout_for_return(struct inode *inode, EXPORT_SYMBOL_GPL(pnfs_error_mark_layout_for_return); void +pnfs_generic_pg_check_layout(struct nfs_pageio_descriptor *pgio) +{ + if (pgio->pg_lseg == NULL || + test_bit(NFS_LSEG_VALID, &pgio->pg_lseg->pls_flags)) + return; + pnfs_put_lseg(pgio->pg_lseg); + pgio->pg_lseg = NULL; +} +EXPORT_SYMBOL_GPL(pnfs_generic_pg_check_layout); + +void pnfs_generic_pg_init_read(struct nfs_pageio_descriptor *pgio, struct nfs_page *req) { u64 rd_size = req->wb_bytes; + pnfs_generic_pg_check_layout(pgio); if (pgio->pg_lseg == NULL) { if (pgio->pg_dreq == NULL) rd_size = i_size_read(pgio->pg_inode) - req_offset(req); @@ -2109,6 +2130,7 @@ void pnfs_generic_pg_init_write(struct nfs_pageio_descriptor *pgio, struct nfs_page *req, u64 wb_size) { + pnfs_generic_pg_check_layout(pgio); if (pgio->pg_lseg == NULL) { pgio->pg_lseg = pnfs_update_layout(pgio->pg_inode, req->wb_context, @@ -2277,8 +2299,20 @@ pnfs_do_write(struct nfs_pageio_descriptor *desc, enum pnfs_try_status trypnfs; trypnfs = pnfs_try_to_write_data(hdr, call_ops, lseg, how); - if (trypnfs == PNFS_NOT_ATTEMPTED) + switch (trypnfs) { + case PNFS_NOT_ATTEMPTED: pnfs_write_through_mds(desc, hdr); + case PNFS_ATTEMPTED: + break; + case PNFS_TRY_AGAIN: + /* cleanup hdr and prepare to redo pnfs */ + if (!test_and_set_bit(NFS_IOHDR_REDO, &hdr->flags)) { + struct nfs_pgio_mirror *mirror = nfs_pgio_current_mirror(desc); + list_splice_init(&hdr->pages, &mirror->pg_list); + mirror->pg_recoalesce = 1; + } + hdr->mds_ops->rpc_release(hdr); + } } static void pnfs_writehdr_free(struct nfs_pgio_header *hdr) @@ -2408,10 +2442,20 @@ pnfs_do_read(struct nfs_pageio_descriptor *desc, struct nfs_pgio_header *hdr) enum pnfs_try_status trypnfs; trypnfs = pnfs_try_to_read_data(hdr, call_ops, lseg); - if (trypnfs == PNFS_TRY_AGAIN) - pnfs_read_resend_pnfs(hdr); - if (trypnfs == PNFS_NOT_ATTEMPTED || hdr->task.tk_status) + switch (trypnfs) { + case PNFS_NOT_ATTEMPTED: pnfs_read_through_mds(desc, hdr); + case PNFS_ATTEMPTED: + break; + case PNFS_TRY_AGAIN: + /* cleanup hdr and prepare to redo pnfs */ + if (!test_and_set_bit(NFS_IOHDR_REDO, &hdr->flags)) { + struct nfs_pgio_mirror *mirror = nfs_pgio_current_mirror(desc); + list_splice_init(&hdr->pages, &mirror->pg_list); + mirror->pg_recoalesce = 1; + } + hdr->mds_ops->rpc_release(hdr); + } } static void pnfs_readhdr_free(struct nfs_pgio_header *hdr) diff --git a/fs/nfs/pnfs.h b/fs/nfs/pnfs.h index 590e1e35781f..2d05b756a8d6 100644 --- a/fs/nfs/pnfs.h +++ b/fs/nfs/pnfs.h @@ -173,14 +173,9 @@ struct pnfs_layoutdriver_type { gfp_t gfp_flags); int (*prepare_layoutreturn) (struct nfs4_layoutreturn_args *); - void (*encode_layoutreturn) (struct xdr_stream *xdr, - const struct nfs4_layoutreturn_args *args); void (*cleanup_layoutcommit) (struct nfs4_layoutcommit_data *data); int (*prepare_layoutcommit) (struct nfs4_layoutcommit_args *args); - void (*encode_layoutcommit) (struct pnfs_layout_hdr *lo, - struct xdr_stream *xdr, - const struct nfs4_layoutcommit_args *args); int (*prepare_layoutstats) (struct nfs42_layoutstat_args *args); }; @@ -239,6 +234,7 @@ void pnfs_put_lseg_locked(struct pnfs_layout_segment *lseg); void set_pnfs_layoutdriver(struct nfs_server *, const struct nfs_fh *, struct nfs_fsinfo *); void unset_pnfs_layoutdriver(struct nfs_server *); +void pnfs_generic_pg_check_layout(struct nfs_pageio_descriptor *pgio); void pnfs_generic_pg_init_read(struct nfs_pageio_descriptor *, struct nfs_page *); int pnfs_generic_pg_readpages(struct nfs_pageio_descriptor *desc); void pnfs_generic_pg_init_write(struct nfs_pageio_descriptor *pgio, diff --git a/fs/nfs/pnfs_nfs.c b/fs/nfs/pnfs_nfs.c index 7250b95549ec..d40755a0984b 100644 --- a/fs/nfs/pnfs_nfs.c +++ b/fs/nfs/pnfs_nfs.c @@ -217,7 +217,14 @@ pnfs_generic_alloc_ds_commits(struct nfs_commit_info *cinfo, for (i = 0; i < fl_cinfo->nbuckets; i++, bucket++) { if (list_empty(&bucket->committing)) continue; - data = nfs_commitdata_alloc(); + /* + * If the layout segment is invalid, then let + * pnfs_generic_retry_commit() clean up the bucket. + */ + if (bucket->clseg && !pnfs_is_valid_lseg(bucket->clseg) && + !test_bit(NFS_LSEG_LAYOUTRETURN, &bucket->clseg->pls_flags)) + break; + data = nfs_commitdata_alloc(false); if (!data) break; data->ds_commit_index = i; @@ -283,16 +290,10 @@ pnfs_generic_commit_pagelist(struct inode *inode, struct list_head *mds_pages, unsigned int nreq = 0; if (!list_empty(mds_pages)) { - data = nfs_commitdata_alloc(); - if (data != NULL) { - data->ds_commit_index = -1; - list_add(&data->pages, &list); - nreq++; - } else { - nfs_retry_commit(mds_pages, NULL, cinfo, 0); - pnfs_generic_retry_commit(cinfo, 0); - return -ENOMEM; - } + data = nfs_commitdata_alloc(true); + data->ds_commit_index = -1; + list_add(&data->pages, &list); + nreq++; } nreq += pnfs_generic_alloc_ds_commits(cinfo, &list); @@ -619,7 +620,6 @@ void nfs4_pnfs_v3_ds_connect_unload(void) get_v3_ds_connect = NULL; } } -EXPORT_SYMBOL_GPL(nfs4_pnfs_v3_ds_connect_unload); static int _nfs4_pnfs_v3_ds_connect(struct nfs_server *mds_srv, struct nfs4_pnfs_ds *ds, diff --git a/fs/nfs/proc.c b/fs/nfs/proc.c index b7bca8303989..9872cf676a50 100644 --- a/fs/nfs/proc.c +++ b/fs/nfs/proc.c @@ -638,7 +638,7 @@ nfs_proc_lock(struct file *filp, int cmd, struct file_lock *fl) { struct inode *inode = file_inode(filp); - return nlmclnt_proc(NFS_SERVER(inode)->nlm_host, cmd, fl); + return nlmclnt_proc(NFS_SERVER(inode)->nlm_host, cmd, fl, NULL); } /* Helper functions for NFS lock bounds checking */ diff --git a/fs/nfs/read.c b/fs/nfs/read.c index defc9233e985..a8421d9dab6a 100644 --- a/fs/nfs/read.c +++ b/fs/nfs/read.c @@ -35,7 +35,11 @@ static struct kmem_cache *nfs_rdata_cachep; static struct nfs_pgio_header *nfs_readhdr_alloc(void) { - return kmem_cache_zalloc(nfs_rdata_cachep, GFP_KERNEL); + struct nfs_pgio_header *p = kmem_cache_zalloc(nfs_rdata_cachep, GFP_KERNEL); + + if (p) + p->rw_mode = FMODE_READ; + return p; } static void nfs_readhdr_free(struct nfs_pgio_header *rhdr) @@ -64,7 +68,7 @@ void nfs_pageio_init_read(struct nfs_pageio_descriptor *pgio, pg_ops = server->pnfs_curr_ld->pg_read_ops; #endif nfs_pageio_init(pgio, inode, pg_ops, compl_ops, &nfs_rw_read_ops, - server->rsize, 0); + server->rsize, 0, GFP_KERNEL); } EXPORT_SYMBOL_GPL(nfs_pageio_init_read); @@ -451,7 +455,6 @@ void nfs_destroy_readpagecache(void) } static const struct nfs_rw_ops nfs_rw_read_ops = { - .rw_mode = FMODE_READ, .rw_alloc_header = nfs_readhdr_alloc, .rw_free_header = nfs_readhdr_free, .rw_done = nfs_readpage_done, diff --git a/fs/nfs/write.c b/fs/nfs/write.c index cc341fc7fd44..db7ba542559e 100644 --- a/fs/nfs/write.c +++ b/fs/nfs/write.c @@ -60,14 +60,28 @@ static mempool_t *nfs_wdata_mempool; static struct kmem_cache *nfs_cdata_cachep; static mempool_t *nfs_commit_mempool; -struct nfs_commit_data *nfs_commitdata_alloc(void) +struct nfs_commit_data *nfs_commitdata_alloc(bool never_fail) { - struct nfs_commit_data *p = mempool_alloc(nfs_commit_mempool, GFP_NOIO); + struct nfs_commit_data *p; - if (p) { - memset(p, 0, sizeof(*p)); - INIT_LIST_HEAD(&p->pages); + if (never_fail) + p = mempool_alloc(nfs_commit_mempool, GFP_NOIO); + else { + /* It is OK to do some reclaim, not no safe to wait + * for anything to be returned to the pool. + * mempool_alloc() cannot handle that particular combination, + * so we need two separate attempts. + */ + p = mempool_alloc(nfs_commit_mempool, GFP_NOWAIT); + if (!p) + p = kmem_cache_alloc(nfs_cdata_cachep, GFP_NOIO | + __GFP_NOWARN | __GFP_NORETRY); + if (!p) + return NULL; } + + memset(p, 0, sizeof(*p)); + INIT_LIST_HEAD(&p->pages); return p; } EXPORT_SYMBOL_GPL(nfs_commitdata_alloc); @@ -82,8 +96,10 @@ static struct nfs_pgio_header *nfs_writehdr_alloc(void) { struct nfs_pgio_header *p = mempool_alloc(nfs_wdata_mempool, GFP_NOIO); - if (p) + if (p) { memset(p, 0, sizeof(*p)); + p->rw_mode = FMODE_WRITE; + } return p; } @@ -547,9 +563,21 @@ static void nfs_write_error_remove_page(struct nfs_page *req) { nfs_unlock_request(req); nfs_end_page_writeback(req); - nfs_release_request(req); generic_error_remove_page(page_file_mapping(req->wb_page), req->wb_page); + nfs_release_request(req); +} + +static bool +nfs_error_is_fatal_on_server(int err) +{ + switch (err) { + case 0: + case -ERESTARTSYS: + case -EINTR: + return false; + } + return nfs_error_is_fatal(err); } /* @@ -557,8 +585,7 @@ static void nfs_write_error_remove_page(struct nfs_page *req) * May return an error if the user signalled nfs_wait_on_request(). */ static int nfs_page_async_flush(struct nfs_pageio_descriptor *pgio, - struct page *page, bool nonblock, - bool launder) + struct page *page, bool nonblock) { struct nfs_page *req; int ret = 0; @@ -574,19 +601,19 @@ static int nfs_page_async_flush(struct nfs_pageio_descriptor *pgio, WARN_ON_ONCE(test_bit(PG_CLEAN, &req->wb_flags)); ret = 0; + /* If there is a fatal error that covers this write, just exit */ + if (nfs_error_is_fatal_on_server(req->wb_context->error)) + goto out_launder; + if (!nfs_pageio_add_request(pgio, req)) { ret = pgio->pg_error; /* - * Remove the problematic req upon fatal errors - * in launder case, while other dirty pages can - * still be around until they get flushed. + * Remove the problematic req upon fatal errors on the server */ if (nfs_error_is_fatal(ret)) { nfs_context_set_write_error(req->wb_context, ret); - if (launder) { - nfs_write_error_remove_page(req); - goto out; - } + if (nfs_error_is_fatal_on_server(ret)) + goto out_launder; } nfs_redirty_request(req); ret = -EAGAIN; @@ -595,16 +622,18 @@ static int nfs_page_async_flush(struct nfs_pageio_descriptor *pgio, NFSIOS_WRITEPAGES, 1); out: return ret; +out_launder: + nfs_write_error_remove_page(req); + return ret; } static int nfs_do_writepage(struct page *page, struct writeback_control *wbc, - struct nfs_pageio_descriptor *pgio, bool launder) + struct nfs_pageio_descriptor *pgio) { int ret; nfs_pageio_cond_complete(pgio, page_index(page)); - ret = nfs_page_async_flush(pgio, page, wbc->sync_mode == WB_SYNC_NONE, - launder); + ret = nfs_page_async_flush(pgio, page, wbc->sync_mode == WB_SYNC_NONE); if (ret == -EAGAIN) { redirty_page_for_writepage(wbc, page); ret = 0; @@ -616,8 +645,7 @@ static int nfs_do_writepage(struct page *page, struct writeback_control *wbc, * Write an mmapped page to the server. */ static int nfs_writepage_locked(struct page *page, - struct writeback_control *wbc, - bool launder) + struct writeback_control *wbc) { struct nfs_pageio_descriptor pgio; struct inode *inode = page_file_mapping(page)->host; @@ -626,7 +654,7 @@ static int nfs_writepage_locked(struct page *page, nfs_inc_stats(inode, NFSIOS_VFSWRITEPAGE); nfs_pageio_init_write(&pgio, inode, 0, false, &nfs_async_write_completion_ops); - err = nfs_do_writepage(page, wbc, &pgio, launder); + err = nfs_do_writepage(page, wbc, &pgio); nfs_pageio_complete(&pgio); if (err < 0) return err; @@ -639,7 +667,7 @@ int nfs_writepage(struct page *page, struct writeback_control *wbc) { int ret; - ret = nfs_writepage_locked(page, wbc, false); + ret = nfs_writepage_locked(page, wbc); unlock_page(page); return ret; } @@ -648,7 +676,7 @@ static int nfs_writepages_callback(struct page *page, struct writeback_control * { int ret; - ret = nfs_do_writepage(page, wbc, data, false); + ret = nfs_do_writepage(page, wbc, data); unlock_page(page); return ret; } @@ -1367,7 +1395,7 @@ void nfs_pageio_init_write(struct nfs_pageio_descriptor *pgio, pg_ops = server->pnfs_curr_ld->pg_write_ops; #endif nfs_pageio_init(pgio, inode, pg_ops, compl_ops, &nfs_rw_write_ops, - server->wsize, ioflags); + server->wsize, ioflags, GFP_NOIO); } EXPORT_SYMBOL_GPL(nfs_pageio_init_write); @@ -1704,50 +1732,14 @@ nfs_commit_list(struct inode *inode, struct list_head *head, int how, if (list_empty(head)) return 0; - data = nfs_commitdata_alloc(); - - if (!data) - goto out_bad; + data = nfs_commitdata_alloc(true); /* Set up the argument struct */ nfs_init_commit(data, head, NULL, cinfo); atomic_inc(&cinfo->mds->rpcs_out); return nfs_initiate_commit(NFS_CLIENT(inode), data, NFS_PROTO(inode), data->mds_ops, how, 0); - out_bad: - nfs_retry_commit(head, NULL, cinfo, 0); - return -ENOMEM; -} - -int nfs_commit_file(struct file *file, struct nfs_write_verifier *verf) -{ - struct inode *inode = file_inode(file); - struct nfs_open_context *open; - struct nfs_commit_info cinfo; - struct nfs_page *req; - int ret; - - open = get_nfs_open_context(nfs_file_open_context(file)); - req = nfs_create_request(open, NULL, NULL, 0, i_size_read(inode)); - if (IS_ERR(req)) { - ret = PTR_ERR(req); - goto out_put; - } - - nfs_init_cinfo_from_inode(&cinfo, inode); - - memcpy(&req->wb_verf, verf, sizeof(struct nfs_write_verifier)); - nfs_request_add_commit_list(req, &cinfo); - ret = nfs_commit_inode(inode, FLUSH_SYNC); - if (ret > 0) - ret = 0; - - nfs_free_request(req); -out_put: - put_nfs_open_context(open); - return ret; } -EXPORT_SYMBOL_GPL(nfs_commit_file); /* * COMMIT call returned @@ -1985,7 +1977,7 @@ int nfs_wb_page_cancel(struct inode *inode, struct page *page) /* * Write back all requests on one page - we do this before reading it. */ -int nfs_wb_single_page(struct inode *inode, struct page *page, bool launder) +int nfs_wb_page(struct inode *inode, struct page *page) { loff_t range_start = page_file_offset(page); loff_t range_end = range_start + (loff_t)(PAGE_SIZE - 1); @@ -2002,7 +1994,7 @@ int nfs_wb_single_page(struct inode *inode, struct page *page, bool launder) for (;;) { wait_on_page_writeback(page); if (clear_page_dirty_for_io(page)) { - ret = nfs_writepage_locked(page, &wbc, launder); + ret = nfs_writepage_locked(page, &wbc); if (ret < 0) goto out_error; continue; @@ -2107,7 +2099,6 @@ void nfs_destroy_writepagecache(void) } static const struct nfs_rw_ops nfs_rw_write_ops = { - .rw_mode = FMODE_WRITE, .rw_alloc_header = nfs_writehdr_alloc, .rw_free_header = nfs_writehdr_free, .rw_done = nfs_writeback_done, diff --git a/fs/nfsd/nfs3xdr.c b/fs/nfsd/nfs3xdr.c index 452334694a5d..12feac6ee2fd 100644 --- a/fs/nfsd/nfs3xdr.c +++ b/fs/nfsd/nfs3xdr.c @@ -334,8 +334,11 @@ nfs3svc_decode_readargs(struct svc_rqst *rqstp, __be32 *p, if (!p) return 0; p = xdr_decode_hyper(p, &args->offset); - args->count = ntohl(*p++); + + if (!xdr_argsize_check(rqstp, p)) + return 0; + len = min(args->count, max_blocksize); /* set up the kvec */ @@ -349,7 +352,7 @@ nfs3svc_decode_readargs(struct svc_rqst *rqstp, __be32 *p, v++; } args->vlen = v; - return xdr_argsize_check(rqstp, p); + return 1; } int @@ -541,9 +544,11 @@ nfs3svc_decode_readlinkargs(struct svc_rqst *rqstp, __be32 *p, p = decode_fh(p, &args->fh); if (!p) return 0; + if (!xdr_argsize_check(rqstp, p)) + return 0; args->buffer = page_address(*(rqstp->rq_next_page++)); - return xdr_argsize_check(rqstp, p); + return 1; } int @@ -569,10 +574,14 @@ nfs3svc_decode_readdirargs(struct svc_rqst *rqstp, __be32 *p, args->verf = p; p += 2; args->dircount = ~0; args->count = ntohl(*p++); + + if (!xdr_argsize_check(rqstp, p)) + return 0; + args->count = min_t(u32, args->count, PAGE_SIZE); args->buffer = page_address(*(rqstp->rq_next_page++)); - return xdr_argsize_check(rqstp, p); + return 1; } int @@ -590,6 +599,9 @@ nfs3svc_decode_readdirplusargs(struct svc_rqst *rqstp, __be32 *p, args->dircount = ntohl(*p++); args->count = ntohl(*p++); + if (!xdr_argsize_check(rqstp, p)) + return 0; + len = args->count = min(args->count, max_blocksize); while (len > 0) { struct page *p = *(rqstp->rq_next_page++); @@ -597,8 +609,7 @@ nfs3svc_decode_readdirplusargs(struct svc_rqst *rqstp, __be32 *p, args->buffer = page_address(p); len -= PAGE_SIZE; } - - return xdr_argsize_check(rqstp, p); + return 1; } int diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index d86031b6ad79..c453a1998e00 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -1259,7 +1259,8 @@ nfsd4_layout_verify(struct svc_export *exp, unsigned int layout_type) return NULL; } - if (!(exp->ex_layout_types & (1 << layout_type))) { + if (layout_type >= LAYOUT_TYPE_MAX || + !(exp->ex_layout_types & (1 << layout_type))) { dprintk("%s: layout type %d not supported\n", __func__, layout_type); return NULL; diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index e9ef50addddb..22002fb75a18 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -1912,28 +1912,15 @@ static void copy_clid(struct nfs4_client *target, struct nfs4_client *source) target->cl_clientid.cl_id = source->cl_clientid.cl_id; } -int strdup_if_nonnull(char **target, char *source) -{ - if (source) { - *target = kstrdup(source, GFP_KERNEL); - if (!*target) - return -ENOMEM; - } else - *target = NULL; - return 0; -} - static int copy_cred(struct svc_cred *target, struct svc_cred *source) { - int ret; + target->cr_principal = kstrdup(source->cr_principal, GFP_KERNEL); + target->cr_raw_principal = kstrdup(source->cr_raw_principal, + GFP_KERNEL); + if ((source->cr_principal && ! target->cr_principal) || + (source->cr_raw_principal && ! target->cr_raw_principal)) + return -ENOMEM; - ret = strdup_if_nonnull(&target->cr_principal, source->cr_principal); - if (ret) - return ret; - ret = strdup_if_nonnull(&target->cr_raw_principal, - source->cr_raw_principal); - if (ret) - return ret; target->cr_flavor = source->cr_flavor; target->cr_uid = source->cr_uid; target->cr_gid = source->cr_gid; diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index 33017d652b1d..26780d53a6f9 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -2831,9 +2831,14 @@ out_acl: } #endif /* CONFIG_NFSD_PNFS */ if (bmval2 & FATTR4_WORD2_SUPPATTR_EXCLCREAT) { - status = nfsd4_encode_bitmap(xdr, NFSD_SUPPATTR_EXCLCREAT_WORD0, - NFSD_SUPPATTR_EXCLCREAT_WORD1, - NFSD_SUPPATTR_EXCLCREAT_WORD2); + u32 supp[3]; + + memcpy(supp, nfsd_suppattrs[minorversion], sizeof(supp)); + supp[0] &= NFSD_SUPPATTR_EXCLCREAT_WORD0; + supp[1] &= NFSD_SUPPATTR_EXCLCREAT_WORD1; + supp[2] &= NFSD_SUPPATTR_EXCLCREAT_WORD2; + + status = nfsd4_encode_bitmap(xdr, supp[0], supp[1], supp[2]); if (status) goto out; } @@ -4119,8 +4124,7 @@ nfsd4_encode_getdeviceinfo(struct nfsd4_compoundres *resp, __be32 nfserr, struct nfsd4_getdeviceinfo *gdev) { struct xdr_stream *xdr = &resp->xdr; - const struct nfsd4_layout_ops *ops = - nfsd4_layout_ops[gdev->gd_layout_type]; + const struct nfsd4_layout_ops *ops; u32 starting_len = xdr->buf->len, needed_len; __be32 *p; @@ -4137,6 +4141,7 @@ nfsd4_encode_getdeviceinfo(struct nfsd4_compoundres *resp, __be32 nfserr, /* If maxcount is 0 then just update notifications */ if (gdev->gd_maxcount != 0) { + ops = nfsd4_layout_ops[gdev->gd_layout_type]; nfserr = ops->encode_getdeviceinfo(xdr, gdev); if (nfserr) { /* @@ -4189,8 +4194,7 @@ nfsd4_encode_layoutget(struct nfsd4_compoundres *resp, __be32 nfserr, struct nfsd4_layoutget *lgp) { struct xdr_stream *xdr = &resp->xdr; - const struct nfsd4_layout_ops *ops = - nfsd4_layout_ops[lgp->lg_layout_type]; + const struct nfsd4_layout_ops *ops; __be32 *p; dprintk("%s: err %d\n", __func__, nfserr); @@ -4213,6 +4217,7 @@ nfsd4_encode_layoutget(struct nfsd4_compoundres *resp, __be32 nfserr, *p++ = cpu_to_be32(lgp->lg_seg.iomode); *p++ = cpu_to_be32(lgp->lg_layout_type); + ops = nfsd4_layout_ops[lgp->lg_layout_type]; nfserr = ops->encode_layoutget(xdr, lgp); out: kfree(lgp->lg_content); diff --git a/fs/nfsd/nfsxdr.c b/fs/nfsd/nfsxdr.c index de07ff625777..6a4947a3f4fa 100644 --- a/fs/nfsd/nfsxdr.c +++ b/fs/nfsd/nfsxdr.c @@ -257,6 +257,9 @@ nfssvc_decode_readargs(struct svc_rqst *rqstp, __be32 *p, len = args->count = ntohl(*p++); p++; /* totalcount - unused */ + if (!xdr_argsize_check(rqstp, p)) + return 0; + len = min_t(unsigned int, len, NFSSVC_MAXBLKSIZE_V2); /* set up somewhere to store response. @@ -272,7 +275,7 @@ nfssvc_decode_readargs(struct svc_rqst *rqstp, __be32 *p, v++; } args->vlen = v; - return xdr_argsize_check(rqstp, p); + return 1; } int @@ -362,9 +365,11 @@ nfssvc_decode_readlinkargs(struct svc_rqst *rqstp, __be32 *p, struct nfsd_readli p = decode_fh(p, &args->fh); if (!p) return 0; + if (!xdr_argsize_check(rqstp, p)) + return 0; args->buffer = page_address(*(rqstp->rq_next_page++)); - return xdr_argsize_check(rqstp, p); + return 1; } int @@ -402,9 +407,11 @@ nfssvc_decode_readdirargs(struct svc_rqst *rqstp, __be32 *p, args->cookie = ntohl(*p++); args->count = ntohl(*p++); args->count = min_t(u32, args->count, PAGE_SIZE); + if (!xdr_argsize_check(rqstp, p)) + return 0; args->buffer = page_address(*(rqstp->rq_next_page++)); - return xdr_argsize_check(rqstp, p); + return 1; } /* diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c index 9aaf6ca77569..2be32955d7f2 100644 --- a/fs/nfsd/vfs.c +++ b/fs/nfsd/vfs.c @@ -94,6 +94,12 @@ nfsd_cross_mnt(struct svc_rqst *rqstp, struct dentry **dpp, err = follow_down(&path); if (err < 0) goto out; + if (path.mnt == exp->ex_path.mnt && path.dentry == dentry && + nfsd_mountpoint(dentry, exp) == 2) { + /* This is only a mountpoint in some other namespace */ + path_put(&path); + goto out; + } exp2 = rqst_exp_get_by_name(rqstp, &path); if (IS_ERR(exp2)) { @@ -167,16 +173,26 @@ static int nfsd_lookup_parent(struct svc_rqst *rqstp, struct dentry *dparent, st /* * For nfsd purposes, we treat V4ROOT exports as though there was an * export at *every* directory. + * We return: + * '1' if this dentry *must* be an export point, + * '2' if it might be, if there is really a mount here, and + * '0' if there is no chance of an export point here. */ int nfsd_mountpoint(struct dentry *dentry, struct svc_export *exp) { - if (d_mountpoint(dentry)) + if (!d_inode(dentry)) + return 0; + if (exp->ex_flags & NFSEXP_V4ROOT) return 1; if (nfsd4_is_junction(dentry)) return 1; - if (!(exp->ex_flags & NFSEXP_V4ROOT)) - return 0; - return d_inode(dentry) != NULL; + if (d_mountpoint(dentry)) + /* + * Might only be a mountpoint in a different namespace, + * but we need to check. + */ + return 2; + return 0; } __be32 diff --git a/fs/open.c b/fs/open.c index 373787afd638..cd0c5be8d012 100644 --- a/fs/open.c +++ b/fs/open.c @@ -193,7 +193,8 @@ static long do_sys_ftruncate(unsigned int fd, loff_t length, int small) goto out_putf; error = -EPERM; - if (IS_APPEND(inode)) + /* Check IS_APPEND on real upper inode */ + if (IS_APPEND(file_inode(f.file))) goto out_putf; sb_start_write(inode->i_sb); @@ -459,20 +460,17 @@ out: SYSCALL_DEFINE1(fchdir, unsigned int, fd) { struct fd f = fdget_raw(fd); - struct inode *inode; - int error = -EBADF; + int error; error = -EBADF; if (!f.file) goto out; - inode = file_inode(f.file); - error = -ENOTDIR; - if (!S_ISDIR(inode->i_mode)) + if (!d_can_lookup(f.file->f_path.dentry)) goto out_putf; - error = inode_permission(inode, MAY_EXEC | MAY_CHDIR); + error = inode_permission(file_inode(f.file), MAY_EXEC | MAY_CHDIR); if (!error) set_fs_pwd(current->fs, &f.file->f_path); out_putf: diff --git a/fs/overlayfs/copy_up.c b/fs/overlayfs/copy_up.c index 906ea6c93260..9008ab9fbd2e 100644 --- a/fs/overlayfs/copy_up.c +++ b/fs/overlayfs/copy_up.c @@ -20,6 +20,7 @@ #include <linux/namei.h> #include <linux/fdtable.h> #include <linux/ratelimit.h> +#include <linux/exportfs.h> #include "overlayfs.h" #include "ovl_entry.h" @@ -232,6 +233,79 @@ int ovl_set_attr(struct dentry *upperdentry, struct kstat *stat) return err; } +static struct ovl_fh *ovl_encode_fh(struct dentry *lower, uuid_be *uuid) +{ + struct ovl_fh *fh; + int fh_type, fh_len, dwords; + void *buf; + int buflen = MAX_HANDLE_SZ; + + buf = kmalloc(buflen, GFP_TEMPORARY); + if (!buf) + return ERR_PTR(-ENOMEM); + + /* + * We encode a non-connectable file handle for non-dir, because we + * only need to find the lower inode number and we don't want to pay + * the price or reconnecting the dentry. + */ + dwords = buflen >> 2; + fh_type = exportfs_encode_fh(lower, buf, &dwords, 0); + buflen = (dwords << 2); + + fh = ERR_PTR(-EIO); + if (WARN_ON(fh_type < 0) || + WARN_ON(buflen > MAX_HANDLE_SZ) || + WARN_ON(fh_type == FILEID_INVALID)) + goto out; + + BUILD_BUG_ON(MAX_HANDLE_SZ + offsetof(struct ovl_fh, fid) > 255); + fh_len = offsetof(struct ovl_fh, fid) + buflen; + fh = kmalloc(fh_len, GFP_KERNEL); + if (!fh) { + fh = ERR_PTR(-ENOMEM); + goto out; + } + + fh->version = OVL_FH_VERSION; + fh->magic = OVL_FH_MAGIC; + fh->type = fh_type; + fh->flags = OVL_FH_FLAG_CPU_ENDIAN; + fh->len = fh_len; + fh->uuid = *uuid; + memcpy(fh->fid, buf, buflen); + +out: + kfree(buf); + return fh; +} + +static int ovl_set_origin(struct dentry *dentry, struct dentry *lower, + struct dentry *upper) +{ + struct super_block *sb = lower->d_sb; + uuid_be *uuid = (uuid_be *) &sb->s_uuid; + const struct ovl_fh *fh = NULL; + int err; + + /* + * When lower layer doesn't support export operations store a 'null' fh, + * so we can use the overlay.origin xattr to distignuish between a copy + * up and a pure upper inode. + */ + if (sb->s_export_op && sb->s_export_op->fh_to_dentry && + uuid_be_cmp(*uuid, NULL_UUID_BE)) { + fh = ovl_encode_fh(lower, uuid); + if (IS_ERR(fh)) + return PTR_ERR(fh); + } + + err = ovl_do_setxattr(upper, OVL_XATTR_ORIGIN, fh, fh ? fh->len : 0, 0); + kfree(fh); + + return err; +} + static int ovl_copy_up_locked(struct dentry *workdir, struct dentry *upperdir, struct dentry *dentry, struct path *lowerpath, struct kstat *stat, const char *link, @@ -316,6 +390,14 @@ static int ovl_copy_up_locked(struct dentry *workdir, struct dentry *upperdir, if (err) goto out_cleanup; + /* + * Store identifier of lower inode in upper inode xattr to + * allow lookup of the copy up origin inode. + */ + err = ovl_set_origin(dentry, lowerpath->dentry, temp); + if (err) + goto out_cleanup; + if (tmpfile) err = ovl_do_link(temp, udir, upper, true); else diff --git a/fs/overlayfs/dir.c b/fs/overlayfs/dir.c index 6515796460df..723b98b90698 100644 --- a/fs/overlayfs/dir.c +++ b/fs/overlayfs/dir.c @@ -138,36 +138,6 @@ static int ovl_set_opaque(struct dentry *dentry, struct dentry *upperdentry) return err; } -static int ovl_dir_getattr(const struct path *path, struct kstat *stat, - u32 request_mask, unsigned int flags) -{ - struct dentry *dentry = path->dentry; - int err; - enum ovl_path_type type; - struct path realpath; - const struct cred *old_cred; - - type = ovl_path_real(dentry, &realpath); - old_cred = ovl_override_creds(dentry->d_sb); - err = vfs_getattr(&realpath, stat, request_mask, flags); - revert_creds(old_cred); - if (err) - return err; - - stat->dev = dentry->d_sb->s_dev; - stat->ino = dentry->d_inode->i_ino; - - /* - * It's probably not worth it to count subdirs to get the - * correct link count. nlink=1 seems to pacify 'find' and - * other utilities. - */ - if (OVL_TYPE_MERGE(type)) - stat->nlink = 1; - - return 0; -} - /* Common operations required to be done after creation of file on upper */ static void ovl_instantiate(struct dentry *dentry, struct inode *inode, struct dentry *newdentry, bool hardlink) @@ -182,6 +152,9 @@ static void ovl_instantiate(struct dentry *dentry, struct inode *inode, inc_nlink(inode); } d_instantiate(dentry, inode); + /* Force lookup of new upper hardlink to find its lower */ + if (hardlink) + d_drop(dentry); } static bool ovl_type_merge(struct dentry *dentry) @@ -210,7 +183,7 @@ static int ovl_create_upper(struct dentry *dentry, struct inode *inode, if (err) goto out_dput; - if (ovl_type_merge(dentry->d_parent)) { + if (ovl_type_merge(dentry->d_parent) && d_is_dir(newdentry)) { /* Setting opaque here is just an optimization, allow to fail */ ovl_set_opaque(dentry, newdentry); } @@ -1070,7 +1043,7 @@ const struct inode_operations ovl_dir_inode_operations = { .create = ovl_create, .mknod = ovl_mknod, .permission = ovl_permission, - .getattr = ovl_dir_getattr, + .getattr = ovl_getattr, .listxattr = ovl_listxattr, .get_acl = ovl_get_acl, .update_time = ovl_update_time, diff --git a/fs/overlayfs/inode.c b/fs/overlayfs/inode.c index f8fe6bf2036d..ad9547f82da5 100644 --- a/fs/overlayfs/inode.c +++ b/fs/overlayfs/inode.c @@ -57,18 +57,78 @@ out: return err; } -static int ovl_getattr(const struct path *path, struct kstat *stat, - u32 request_mask, unsigned int flags) +int ovl_getattr(const struct path *path, struct kstat *stat, + u32 request_mask, unsigned int flags) { struct dentry *dentry = path->dentry; + enum ovl_path_type type; struct path realpath; const struct cred *old_cred; + bool is_dir = S_ISDIR(dentry->d_inode->i_mode); int err; - ovl_path_real(dentry, &realpath); + type = ovl_path_real(dentry, &realpath); old_cred = ovl_override_creds(dentry->d_sb); err = vfs_getattr(&realpath, stat, request_mask, flags); + if (err) + goto out; + + /* + * When all layers are on the same fs, all real inode number are + * unique, so we use the overlay st_dev, which is friendly to du -x. + * + * We also use st_ino of the copy up origin, if we know it. + * This guaranties constant st_dev/st_ino across copy up. + * + * If filesystem supports NFS export ops, this also guaranties + * persistent st_ino across mount cycle. + */ + if (ovl_same_sb(dentry->d_sb)) { + if (OVL_TYPE_ORIGIN(type)) { + struct kstat lowerstat; + u32 lowermask = STATX_INO | (!is_dir ? STATX_NLINK : 0); + + ovl_path_lower(dentry, &realpath); + err = vfs_getattr(&realpath, &lowerstat, + lowermask, flags); + if (err) + goto out; + + WARN_ON_ONCE(stat->dev != lowerstat.dev); + /* + * Lower hardlinks are broken on copy up to different + * upper files, so we cannot use the lower origin st_ino + * for those different files, even for the same fs case. + */ + if (is_dir || lowerstat.nlink == 1) + stat->ino = lowerstat.ino; + } + stat->dev = dentry->d_sb->s_dev; + } else if (is_dir) { + /* + * If not all layers are on the same fs the pair {real st_ino; + * overlay st_dev} is not unique, so use the non persistent + * overlay st_ino. + * + * Always use the overlay st_dev for directories, so 'find + * -xdev' will scan the entire overlay mount and won't cross the + * overlay mount boundaries. + */ + stat->dev = dentry->d_sb->s_dev; + stat->ino = dentry->d_inode->i_ino; + } + + /* + * It's probably not worth it to count subdirs to get the + * correct link count. nlink=1 seems to pacify 'find' and + * other utilities. + */ + if (is_dir && OVL_TYPE_MERGE(type)) + stat->nlink = 1; + +out: revert_creds(old_cred); + return err; } @@ -303,6 +363,41 @@ static const struct inode_operations ovl_symlink_inode_operations = { .update_time = ovl_update_time, }; +/* + * It is possible to stack overlayfs instance on top of another + * overlayfs instance as lower layer. We need to annonate the + * stackable i_mutex locks according to stack level of the super + * block instance. An overlayfs instance can never be in stack + * depth 0 (there is always a real fs below it). An overlayfs + * inode lock will use the lockdep annotaion ovl_i_mutex_key[depth]. + * + * For example, here is a snip from /proc/lockdep_chains after + * dir_iterate of nested overlayfs: + * + * [...] &ovl_i_mutex_dir_key[depth] (stack_depth=2) + * [...] &ovl_i_mutex_dir_key[depth]#2 (stack_depth=1) + * [...] &type->i_mutex_dir_key (stack_depth=0) + */ +#define OVL_MAX_NESTING FILESYSTEM_MAX_STACK_DEPTH + +static inline void ovl_lockdep_annotate_inode_mutex_key(struct inode *inode) +{ +#ifdef CONFIG_LOCKDEP + static struct lock_class_key ovl_i_mutex_key[OVL_MAX_NESTING]; + static struct lock_class_key ovl_i_mutex_dir_key[OVL_MAX_NESTING]; + + int depth = inode->i_sb->s_stack_depth - 1; + + if (WARN_ON_ONCE(depth < 0 || depth >= OVL_MAX_NESTING)) + depth = 0; + + if (S_ISDIR(inode->i_mode)) + lockdep_set_class(&inode->i_rwsem, &ovl_i_mutex_dir_key[depth]); + else + lockdep_set_class(&inode->i_rwsem, &ovl_i_mutex_key[depth]); +#endif +} + static void ovl_fill_inode(struct inode *inode, umode_t mode, dev_t rdev) { inode->i_ino = get_next_ino(); @@ -312,6 +407,8 @@ static void ovl_fill_inode(struct inode *inode, umode_t mode, dev_t rdev) inode->i_acl = inode->i_default_acl = ACL_DONT_CACHE; #endif + ovl_lockdep_annotate_inode_mutex_key(inode); + switch (mode & S_IFMT) { case S_IFREG: inode->i_op = &ovl_file_inode_operations; diff --git a/fs/overlayfs/namei.c b/fs/overlayfs/namei.c index b8b077821fb0..bad0f665a635 100644 --- a/fs/overlayfs/namei.c +++ b/fs/overlayfs/namei.c @@ -12,6 +12,8 @@ #include <linux/namei.h> #include <linux/xattr.h> #include <linux/ratelimit.h> +#include <linux/mount.h> +#include <linux/exportfs.h> #include "overlayfs.h" #include "ovl_entry.h" @@ -81,6 +83,90 @@ invalid: goto err_free; } +static int ovl_acceptable(void *ctx, struct dentry *dentry) +{ + return 1; +} + +static struct dentry *ovl_get_origin(struct dentry *dentry, + struct vfsmount *mnt) +{ + int res; + struct ovl_fh *fh = NULL; + struct dentry *origin = NULL; + int bytes; + + res = vfs_getxattr(dentry, OVL_XATTR_ORIGIN, NULL, 0); + if (res < 0) { + if (res == -ENODATA || res == -EOPNOTSUPP) + return NULL; + goto fail; + } + /* Zero size value means "copied up but origin unknown" */ + if (res == 0) + return NULL; + + fh = kzalloc(res, GFP_TEMPORARY); + if (!fh) + return ERR_PTR(-ENOMEM); + + res = vfs_getxattr(dentry, OVL_XATTR_ORIGIN, fh, res); + if (res < 0) + goto fail; + + if (res < sizeof(struct ovl_fh) || res < fh->len) + goto invalid; + + if (fh->magic != OVL_FH_MAGIC) + goto invalid; + + /* Treat larger version and unknown flags as "origin unknown" */ + if (fh->version > OVL_FH_VERSION || fh->flags & ~OVL_FH_FLAG_ALL) + goto out; + + /* Treat endianness mismatch as "origin unknown" */ + if (!(fh->flags & OVL_FH_FLAG_ANY_ENDIAN) && + (fh->flags & OVL_FH_FLAG_BIG_ENDIAN) != OVL_FH_FLAG_CPU_ENDIAN) + goto out; + + bytes = (fh->len - offsetof(struct ovl_fh, fid)); + + /* + * Make sure that the stored uuid matches the uuid of the lower + * layer where file handle will be decoded. + */ + if (uuid_be_cmp(fh->uuid, *(uuid_be *) &mnt->mnt_sb->s_uuid)) + goto out; + + origin = exportfs_decode_fh(mnt, (struct fid *)fh->fid, + bytes >> 2, (int)fh->type, + ovl_acceptable, NULL); + if (IS_ERR(origin)) { + /* Treat stale file handle as "origin unknown" */ + if (origin == ERR_PTR(-ESTALE)) + origin = NULL; + goto out; + } + + if (ovl_dentry_weird(origin) || + ((d_inode(origin)->i_mode ^ d_inode(dentry)->i_mode) & S_IFMT)) { + dput(origin); + origin = NULL; + goto invalid; + } + +out: + kfree(fh); + return origin; + +fail: + pr_warn_ratelimited("overlayfs: failed to get origin (%i)\n", res); + goto out; +invalid: + pr_warn_ratelimited("overlayfs: invalid origin (%*phN)\n", res, fh); + goto out; +} + static bool ovl_is_opaquedir(struct dentry *dentry) { int res; @@ -192,6 +278,45 @@ static int ovl_lookup_layer(struct dentry *base, struct ovl_lookup_data *d, return 0; } + +static int ovl_check_origin(struct dentry *dentry, struct dentry *upperdentry, + struct path **stackp, unsigned int *ctrp) +{ + struct super_block *same_sb = ovl_same_sb(dentry->d_sb); + struct ovl_entry *roe = dentry->d_sb->s_root->d_fsdata; + struct vfsmount *mnt; + struct dentry *origin; + + if (!same_sb || !roe->numlower) + return 0; + + /* + * Since all layers are on the same fs, we use the first layer for + * decoding the file handle. We may get a disconnected dentry, + * which is fine, because we only need to hold the origin inode in + * cache and use its inode number. We may even get a connected dentry, + * that is not under the first layer's root. That is also fine for + * using it's inode number - it's the same as if we held a reference + * to a dentry in first layer that was moved under us. + */ + mnt = roe->lowerstack[0].mnt; + + origin = ovl_get_origin(upperdentry, mnt); + if (IS_ERR_OR_NULL(origin)) + return PTR_ERR(origin); + + BUG_ON(*stackp || *ctrp); + *stackp = kmalloc(sizeof(struct path), GFP_TEMPORARY); + if (!*stackp) { + dput(origin); + return -ENOMEM; + } + **stackp = (struct path) { .dentry = origin, .mnt = mnt }; + *ctrp = 1; + + return 0; +} + /* * Returns next layer in stack starting from top. * Returns -1 if this is the last layer. @@ -220,6 +345,7 @@ struct dentry *ovl_lookup(struct inode *dir, struct dentry *dentry, const struct cred *old_cred; struct ovl_fs *ofs = dentry->d_sb->s_fs_info; struct ovl_entry *poe = dentry->d_parent->d_fsdata; + struct ovl_entry *roe = dentry->d_sb->s_root->d_fsdata; struct path *stack = NULL; struct dentry *upperdir, *upperdentry = NULL; unsigned int ctr = 0; @@ -253,13 +379,20 @@ struct dentry *ovl_lookup(struct inode *dir, struct dentry *dentry, err = -EREMOTE; goto out; } + if (upperdentry && !d.is_dir) { + BUG_ON(!d.stop || d.redirect); + err = ovl_check_origin(dentry, upperdentry, + &stack, &ctr); + if (err) + goto out; + } if (d.redirect) { upperredirect = kstrdup(d.redirect, GFP_KERNEL); if (!upperredirect) goto out_put_upper; if (d.redirect[0] == '/') - poe = dentry->d_sb->s_root->d_fsdata; + poe = roe; } upperopaque = d.opaque; } @@ -290,10 +423,8 @@ struct dentry *ovl_lookup(struct inode *dir, struct dentry *dentry, if (d.stop) break; - if (d.redirect && - d.redirect[0] == '/' && - poe != dentry->d_sb->s_root->d_fsdata) { - poe = dentry->d_sb->s_root->d_fsdata; + if (d.redirect && d.redirect[0] == '/' && poe != roe) { + poe = roe; /* Find the current layer on the root dentry */ for (i = 0; i < poe->numlower; i++) diff --git a/fs/overlayfs/overlayfs.h b/fs/overlayfs/overlayfs.h index 741dc0b6931f..caa36cb9c46d 100644 --- a/fs/overlayfs/overlayfs.h +++ b/fs/overlayfs/overlayfs.h @@ -8,18 +8,56 @@ */ #include <linux/kernel.h> +#include <linux/uuid.h> enum ovl_path_type { __OVL_PATH_UPPER = (1 << 0), __OVL_PATH_MERGE = (1 << 1), + __OVL_PATH_ORIGIN = (1 << 2), }; #define OVL_TYPE_UPPER(type) ((type) & __OVL_PATH_UPPER) #define OVL_TYPE_MERGE(type) ((type) & __OVL_PATH_MERGE) +#define OVL_TYPE_ORIGIN(type) ((type) & __OVL_PATH_ORIGIN) #define OVL_XATTR_PREFIX XATTR_TRUSTED_PREFIX "overlay." #define OVL_XATTR_OPAQUE OVL_XATTR_PREFIX "opaque" #define OVL_XATTR_REDIRECT OVL_XATTR_PREFIX "redirect" +#define OVL_XATTR_ORIGIN OVL_XATTR_PREFIX "origin" + +/* + * The tuple (fh,uuid) is a universal unique identifier for a copy up origin, + * where: + * origin.fh - exported file handle of the lower file + * origin.uuid - uuid of the lower filesystem + */ +#define OVL_FH_VERSION 0 +#define OVL_FH_MAGIC 0xfb + +/* CPU byte order required for fid decoding: */ +#define OVL_FH_FLAG_BIG_ENDIAN (1 << 0) +#define OVL_FH_FLAG_ANY_ENDIAN (1 << 1) + +#define OVL_FH_FLAG_ALL (OVL_FH_FLAG_BIG_ENDIAN | OVL_FH_FLAG_ANY_ENDIAN) + +#if defined(__LITTLE_ENDIAN) +#define OVL_FH_FLAG_CPU_ENDIAN 0 +#elif defined(__BIG_ENDIAN) +#define OVL_FH_FLAG_CPU_ENDIAN OVL_FH_FLAG_BIG_ENDIAN +#else +#error Endianness not defined +#endif + +/* On-disk and in-memeory format for redirect by file handle */ +struct ovl_fh { + u8 version; /* 0 */ + u8 magic; /* 0xfb */ + u8 len; /* size of this header + size of fid */ + u8 flags; /* OVL_FH_FLAG_* */ + u8 type; /* fid_type of fid */ + uuid_be uuid; /* uuid of filesystem */ + u8 fid[0]; /* file identifier */ +} __packed; #define OVL_ISUPPER_MASK 1UL @@ -151,6 +189,7 @@ int ovl_want_write(struct dentry *dentry); void ovl_drop_write(struct dentry *dentry); struct dentry *ovl_workdir(struct dentry *dentry); const struct cred *ovl_override_creds(struct super_block *sb); +struct super_block *ovl_same_sb(struct super_block *sb); struct ovl_entry *ovl_alloc_entry(unsigned int numlower); bool ovl_dentry_remote(struct dentry *dentry); bool ovl_dentry_weird(struct dentry *dentry); @@ -197,6 +236,8 @@ void ovl_workdir_cleanup(struct inode *dir, struct vfsmount *mnt, /* inode.c */ int ovl_setattr(struct dentry *dentry, struct iattr *attr); +int ovl_getattr(const struct path *path, struct kstat *stat, + u32 request_mask, unsigned int flags); int ovl_permission(struct inode *inode, int mask); int ovl_xattr_set(struct dentry *dentry, const char *name, const void *value, size_t size, int flags); diff --git a/fs/overlayfs/ovl_entry.h b/fs/overlayfs/ovl_entry.h index 59614faa14c3..b2023ddb8532 100644 --- a/fs/overlayfs/ovl_entry.h +++ b/fs/overlayfs/ovl_entry.h @@ -29,6 +29,8 @@ struct ovl_fs { const struct cred *creator_cred; bool tmpfile; wait_queue_head_t copyup_wq; + /* sb common to all layers */ + struct super_block *same_sb; }; /* private information held for every overlayfs dentry */ diff --git a/fs/overlayfs/super.c b/fs/overlayfs/super.c index c9e70d39c1ea..9828b7de8999 100644 --- a/fs/overlayfs/super.c +++ b/fs/overlayfs/super.c @@ -49,11 +49,28 @@ static void ovl_dentry_release(struct dentry *dentry) } } +static int ovl_check_append_only(struct inode *inode, int flag) +{ + /* + * This test was moot in vfs may_open() because overlay inode does + * not have the S_APPEND flag, so re-check on real upper inode + */ + if (IS_APPEND(inode)) { + if ((flag & O_ACCMODE) != O_RDONLY && !(flag & O_APPEND)) + return -EPERM; + if (flag & O_TRUNC) + return -EPERM; + } + + return 0; +} + static struct dentry *ovl_d_real(struct dentry *dentry, const struct inode *inode, unsigned int open_flags) { struct dentry *real; + int err; if (!d_is_reg(dentry)) { if (!inode || inode == d_inode(dentry)) @@ -65,15 +82,20 @@ static struct dentry *ovl_d_real(struct dentry *dentry, return dentry; if (open_flags) { - int err = ovl_open_maybe_copy_up(dentry, open_flags); - + err = ovl_open_maybe_copy_up(dentry, open_flags); if (err) return ERR_PTR(err); } real = ovl_dentry_upper(dentry); - if (real && (!inode || inode == d_inode(real))) + if (real && (!inode || inode == d_inode(real))) { + if (!inode) { + err = ovl_check_append_only(d_inode(real), open_flags); + if (err) + return ERR_PTR(err); + } return real; + } real = ovl_dentry_lower(dentry); if (!real) @@ -709,8 +731,8 @@ static const struct xattr_handler *ovl_xattr_handlers[] = { static int ovl_fill_super(struct super_block *sb, void *data, int silent) { - struct path upperpath = { NULL, NULL }; - struct path workpath = { NULL, NULL }; + struct path upperpath = { }; + struct path workpath = { }; struct dentry *root_dentry; struct inode *realinode; struct ovl_entry *oe; @@ -892,11 +914,19 @@ static int ovl_fill_super(struct super_block *sb, void *data, int silent) ufs->lower_mnt[ufs->numlower] = mnt; ufs->numlower++; + + /* Check if all lower layers are on same sb */ + if (i == 0) + ufs->same_sb = mnt->mnt_sb; + else if (ufs->same_sb != mnt->mnt_sb) + ufs->same_sb = NULL; } /* If the upper fs is nonexistent, we mark overlayfs r/o too */ if (!ufs->upper_mnt) sb->s_flags |= MS_RDONLY; + else if (ufs->upper_mnt->mnt_sb != ufs->same_sb) + ufs->same_sb = NULL; if (remote) sb->s_d_op = &ovl_reval_dentry_operations; diff --git a/fs/overlayfs/util.c b/fs/overlayfs/util.c index 6e610a205e15..cfdea47313a1 100644 --- a/fs/overlayfs/util.c +++ b/fs/overlayfs/util.c @@ -40,6 +40,13 @@ const struct cred *ovl_override_creds(struct super_block *sb) return override_creds(ofs->creator_cred); } +struct super_block *ovl_same_sb(struct super_block *sb) +{ + struct ovl_fs *ofs = sb->s_fs_info; + + return ofs->same_sb; +} + struct ovl_entry *ovl_alloc_entry(unsigned int numlower) { size_t size = offsetof(struct ovl_entry, lowerstack[numlower]); @@ -75,11 +82,13 @@ enum ovl_path_type ovl_path_type(struct dentry *dentry) type = __OVL_PATH_UPPER; /* - * Non-dir dentry can hold lower dentry from previous - * location. + * Non-dir dentry can hold lower dentry of its copy up origin. */ - if (oe->numlower && d_is_dir(dentry)) - type |= __OVL_PATH_MERGE; + if (oe->numlower) { + type |= __OVL_PATH_ORIGIN; + if (d_is_dir(dentry)) + type |= __OVL_PATH_MERGE; + } } else { if (oe->numlower > 1) type |= __OVL_PATH_MERGE; @@ -100,7 +109,7 @@ void ovl_path_lower(struct dentry *dentry, struct path *path) { struct ovl_entry *oe = dentry->d_fsdata; - *path = oe->numlower ? oe->lowerstack[0] : (struct path) { NULL, NULL }; + *path = oe->numlower ? oe->lowerstack[0] : (struct path) { }; } enum ovl_path_type ovl_path_real(struct dentry *dentry, struct path *path) diff --git a/fs/pstore/ram.c b/fs/pstore/ram.c index 5523df7f17ef..5cb022c8cd33 100644 --- a/fs/pstore/ram.c +++ b/fs/pstore/ram.c @@ -58,7 +58,7 @@ module_param_named(pmsg_size, ramoops_pmsg_size, ulong, 0400); MODULE_PARM_DESC(pmsg_size, "size of user space message log"); static unsigned long long mem_address; -module_param(mem_address, ullong, 0400); +module_param_hw(mem_address, ullong, other, 0400); MODULE_PARM_DESC(mem_address, "start of reserved RAM used to store oops/panic logs"); diff --git a/fs/signalfd.c b/fs/signalfd.c index 270221fcef42..7e3d71109f51 100644 --- a/fs/signalfd.c +++ b/fs/signalfd.c @@ -38,7 +38,7 @@ void signalfd_cleanup(struct sighand_struct *sighand) /* * The lockless check can race with remove_wait_queue() in progress, * but in this case its caller should run under rcu_read_lock() and - * sighand_cachep is SLAB_DESTROY_BY_RCU, we can safely return. + * sighand_cachep is SLAB_TYPESAFE_BY_RCU, we can safely return. */ if (likely(!waitqueue_active(wqh))) return; diff --git a/fs/ubifs/Kconfig b/fs/ubifs/Kconfig index b0d0623c83ed..83a961bf7280 100644 --- a/fs/ubifs/Kconfig +++ b/fs/ubifs/Kconfig @@ -61,3 +61,16 @@ config UBIFS_FS_ENCRYPTION feature is similar to ecryptfs, but it is more memory efficient since it avoids caching the encrypted and decrypted pages in the page cache. + +config UBIFS_FS_SECURITY + bool "UBIFS Security Labels" + depends on UBIFS_FS + default y + help + Security labels provide an access control facility to support Linux + Security Models (LSMs) accepted by AppArmor, SELinux, Smack and TOMOYO + Linux. This option enables an extended attribute handler for file + security labels in the ubifs filesystem, so that it requires enabling + the extended attribute support in advance. + + If you are not using a security module, say N. diff --git a/fs/ubifs/debug.c b/fs/ubifs/debug.c index 718b749fa11a..7cd8a7b95299 100644 --- a/fs/ubifs/debug.c +++ b/fs/ubifs/debug.c @@ -2391,8 +2391,8 @@ int dbg_check_nondata_nodes_order(struct ubifs_info *c, struct list_head *head) ubifs_dump_node(c, sa->node); return -EINVAL; } - if (sa->type != UBIFS_INO_NODE && sa->type != UBIFS_DENT_NODE && - sa->type != UBIFS_XENT_NODE) { + if (sb->type != UBIFS_INO_NODE && sb->type != UBIFS_DENT_NODE && + sb->type != UBIFS_XENT_NODE) { ubifs_err(c, "bad node type %d", sb->type); ubifs_dump_node(c, sb->node); return -EINVAL; diff --git a/fs/ubifs/ioctl.c b/fs/ubifs/ioctl.c index 12b9eb5005ff..fdc311246807 100644 --- a/fs/ubifs/ioctl.c +++ b/fs/ubifs/ioctl.c @@ -53,7 +53,7 @@ void ubifs_set_inode_flags(struct inode *inode) * ioctl2ubifs - convert ioctl inode flags to UBIFS inode flags. * @ioctl_flags: flags to convert * - * This function convert ioctl flags (@FS_COMPR_FL, etc) to UBIFS inode flags + * This function converts ioctl flags (@FS_COMPR_FL, etc) to UBIFS inode flags * (@UBIFS_COMPR_FL, etc). */ static int ioctl2ubifs(int ioctl_flags) @@ -78,8 +78,8 @@ static int ioctl2ubifs(int ioctl_flags) * ubifs2ioctl - convert UBIFS inode flags to ioctl inode flags. * @ubifs_flags: flags to convert * - * This function convert UBIFS (@UBIFS_COMPR_FL, etc) to ioctl flags - * (@FS_COMPR_FL, etc). + * This function converts UBIFS inode flags (@UBIFS_COMPR_FL, etc) to ioctl + * flags (@FS_COMPR_FL, etc). */ static int ubifs2ioctl(int ubifs_flags) { diff --git a/fs/ubifs/recovery.c b/fs/ubifs/recovery.c index 586d59347fff..3af4472061cc 100644 --- a/fs/ubifs/recovery.c +++ b/fs/ubifs/recovery.c @@ -442,7 +442,6 @@ static void clean_buf(const struct ubifs_info *c, void **buf, int lnum, { int empty_offs, pad_len; - lnum = lnum; dbg_rcvry("cleaning corruption at %d:%d", lnum, *offs); ubifs_assert(!(*offs & 7)); diff --git a/fs/ubifs/ubifs.h b/fs/ubifs/ubifs.h index 4da10a6d702a..298b4d89eee9 100644 --- a/fs/ubifs/ubifs.h +++ b/fs/ubifs/ubifs.h @@ -1753,13 +1753,23 @@ int ubifs_check_dir_empty(struct inode *dir); /* xattr.c */ extern const struct xattr_handler *ubifs_xattr_handlers[]; ssize_t ubifs_listxattr(struct dentry *dentry, char *buffer, size_t size); -int ubifs_init_security(struct inode *dentry, struct inode *inode, - const struct qstr *qstr); int ubifs_xattr_set(struct inode *host, const char *name, const void *value, size_t size, int flags); ssize_t ubifs_xattr_get(struct inode *host, const char *name, void *buf, size_t size); +#ifdef CONFIG_UBIFS_FS_SECURITY +extern int ubifs_init_security(struct inode *dentry, struct inode *inode, + const struct qstr *qstr); +#else +static inline int ubifs_init_security(struct inode *dentry, + struct inode *inode, const struct qstr *qstr) +{ + return 0; +} +#endif + + /* super.c */ struct inode *ubifs_iget(struct super_block *sb, unsigned long inum); diff --git a/fs/ubifs/xattr.c b/fs/ubifs/xattr.c index 3e53fdbf7997..6c9e62c2ef55 100644 --- a/fs/ubifs/xattr.c +++ b/fs/ubifs/xattr.c @@ -559,6 +559,7 @@ out_free: return err; } +#ifdef CONFIG_UBIFS_FS_SECURITY static int init_xattrs(struct inode *inode, const struct xattr *xattr_array, void *fs_info) { @@ -599,6 +600,7 @@ int ubifs_init_security(struct inode *dentry, struct inode *inode, } return err; } +#endif static int xattr_get(const struct xattr_handler *handler, struct dentry *dentry, struct inode *inode, @@ -639,15 +641,19 @@ static const struct xattr_handler ubifs_trusted_xattr_handler = { .set = xattr_set, }; +#ifdef CONFIG_UBIFS_FS_SECURITY static const struct xattr_handler ubifs_security_xattr_handler = { .prefix = XATTR_SECURITY_PREFIX, .get = xattr_get, .set = xattr_set, }; +#endif const struct xattr_handler *ubifs_xattr_handlers[] = { &ubifs_user_xattr_handler, &ubifs_trusted_xattr_handler, +#ifdef CONFIG_UBIFS_FS_SECURITY &ubifs_security_xattr_handler, +#endif NULL }; diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c index 47d239dcf3f4..455a575f101d 100644 --- a/fs/xfs/xfs_super.c +++ b/fs/xfs/xfs_super.c @@ -52,6 +52,7 @@ #include "xfs_reflink.h" #include <linux/namei.h> +#include <linux/dax.h> #include <linux/init.h> #include <linux/slab.h> #include <linux/mount.h> |