From e18fca734278784bd6591de63ca148cc27344ca9 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Mon, 18 Jun 2012 07:23:18 -0600 Subject: Btrfs: add a missing spin_lock When fixing up the locking in the delayed ref destruction work I accidently broke the locking myself ;(. Add back a spin_lock that should be there and we are now all set. Thanks, Btrfs: add a missing spin_lock When fixing up the locking in the delayed ref destruction work I accidently broke the locking myself ;(. Add back a spin_lock that should be there and we are now all set. Thanks, Reported-by: Dan Carpenter Signed-off-by: Josef Bacik Signed-off-by: Chris Mason --- fs/btrfs/disk-io.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs/btrfs/disk-io.c') diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index e22c5bbf0223..7d7bc8eace86 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -3426,6 +3426,7 @@ int btrfs_destroy_delayed_refs(struct btrfs_transaction *trans, mutex_unlock(&head->mutex); btrfs_put_delayed_ref(ref); + spin_lock(&delayed_refs->lock); continue; } -- cgit v1.2.3 From 68310a5e42f93c2242ec1836c3b18d531e0065e2 Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Fri, 22 Jun 2012 12:24:12 -0600 Subject: Btrfs: restore restriper state on all mounts Fix a bug that triggered asserts in btrfs_balance() in both normal and resume modes -- restriper state was not properly restored on read-only mounts. This factors out resuming code from btrfs_restore_balance(), which is now also called earlier in the mount sequence to avoid the problem of some early writes getting the old profile. Signed-off-by: Ilya Dryomov --- fs/btrfs/disk-io.c | 10 ++++++---- fs/btrfs/volumes.c | 39 +++++++++++++++++++-------------------- fs/btrfs/volumes.h | 2 +- 3 files changed, 26 insertions(+), 25 deletions(-) (limited to 'fs/btrfs/disk-io.c') diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 7d7bc8eace86..3a7961ba161e 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -2354,12 +2354,17 @@ retry_root_backup: BTRFS_CSUM_TREE_OBJECTID, csum_root); if (ret) goto recovery_tree_root; - csum_root->track_dirty = 1; fs_info->generation = generation; fs_info->last_trans_committed = generation; + ret = btrfs_recover_balance(fs_info); + if (ret) { + printk(KERN_WARNING "btrfs: failed to recover balance\n"); + goto fail_block_groups; + } + ret = btrfs_init_dev_stats(fs_info); if (ret) { printk(KERN_ERR "btrfs: failed to init dev_stats: %d\n", @@ -2492,9 +2497,6 @@ retry_root_backup: err = btrfs_orphan_cleanup(fs_info->tree_root); up_read(&fs_info->cleanup_work_sem); - if (!err) - err = btrfs_recover_balance(fs_info->tree_root); - if (err) { close_ctree(tree_root); return err; diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 3f292cf693a7..48943d0f861a 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -2867,9 +2867,8 @@ static int balance_kthread(void *data) return ret; } -int btrfs_recover_balance(struct btrfs_root *tree_root) +int btrfs_recover_balance(struct btrfs_fs_info *fs_info) { - struct task_struct *tsk; struct btrfs_balance_control *bctl; struct btrfs_balance_item *item; struct btrfs_disk_balance_args disk_bargs; @@ -2882,29 +2881,30 @@ int btrfs_recover_balance(struct btrfs_root *tree_root) if (!path) return -ENOMEM; - bctl = kzalloc(sizeof(*bctl), GFP_NOFS); - if (!bctl) { - ret = -ENOMEM; - goto out; - } - key.objectid = BTRFS_BALANCE_OBJECTID; key.type = BTRFS_BALANCE_ITEM_KEY; key.offset = 0; - ret = btrfs_search_slot(NULL, tree_root, &key, path, 0, 0); + ret = btrfs_search_slot(NULL, fs_info->tree_root, &key, path, 0, 0); if (ret < 0) - goto out_bctl; + goto out; if (ret > 0) { /* ret = -ENOENT; */ ret = 0; - goto out_bctl; + goto out; + } + + bctl = kzalloc(sizeof(*bctl), GFP_NOFS); + if (!bctl) { + ret = -ENOMEM; + goto out; } leaf = path->nodes[0]; item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_balance_item); - bctl->fs_info = tree_root->fs_info; - bctl->flags = btrfs_balance_flags(leaf, item) | BTRFS_BALANCE_RESUME; + bctl->fs_info = fs_info; + bctl->flags = btrfs_balance_flags(leaf, item); + bctl->flags |= BTRFS_BALANCE_RESUME; btrfs_balance_data(leaf, item, &disk_bargs); btrfs_disk_balance_args_to_cpu(&bctl->data, &disk_bargs); @@ -2913,14 +2913,13 @@ int btrfs_recover_balance(struct btrfs_root *tree_root) btrfs_balance_sys(leaf, item, &disk_bargs); btrfs_disk_balance_args_to_cpu(&bctl->sys, &disk_bargs); - tsk = kthread_run(balance_kthread, bctl, "btrfs-balance"); - if (IS_ERR(tsk)) - ret = PTR_ERR(tsk); - else - goto out; + mutex_lock(&fs_info->volume_mutex); + mutex_lock(&fs_info->balance_mutex); -out_bctl: - kfree(bctl); + set_balance_control(bctl); + + mutex_unlock(&fs_info->balance_mutex); + mutex_unlock(&fs_info->volume_mutex); out: btrfs_free_path(path); return ret; diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h index 74366f27a76b..e1b1a649fc5a 100644 --- a/fs/btrfs/volumes.h +++ b/fs/btrfs/volumes.h @@ -281,7 +281,7 @@ int btrfs_shrink_device(struct btrfs_device *device, u64 new_size); int btrfs_init_new_device(struct btrfs_root *root, char *path); int btrfs_balance(struct btrfs_balance_control *bctl, struct btrfs_ioctl_balance_args *bargs); -int btrfs_recover_balance(struct btrfs_root *tree_root); +int btrfs_recover_balance(struct btrfs_fs_info *fs_info); int btrfs_pause_balance(struct btrfs_fs_info *fs_info); int btrfs_cancel_balance(struct btrfs_fs_info *fs_info); int btrfs_chunk_readonly(struct btrfs_root *root, u64 chunk_offset); -- cgit v1.2.3 From 2b6ba629b5aac51e7099efbb43e2b403213aa7fb Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Fri, 22 Jun 2012 12:24:13 -0600 Subject: Btrfs: resume balance on rw (re)mounts properly This introduces btrfs_resume_balance_async(), which, given that restriper state was recovered earlier by btrfs_recover_balance(), resumes balance in btrfs-balance kthread. Signed-off-by: Ilya Dryomov --- fs/btrfs/disk-io.c | 24 +++++++++++++++--------- fs/btrfs/super.c | 4 ++++ fs/btrfs/volumes.c | 36 +++++++++++++++++++++++++++--------- fs/btrfs/volumes.h | 1 + 4 files changed, 47 insertions(+), 18 deletions(-) (limited to 'fs/btrfs/disk-io.c') diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 3a7961ba161e..8cc47103a32e 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -2490,17 +2490,23 @@ retry_root_backup: goto fail_trans_kthread; } - if (!(sb->s_flags & MS_RDONLY)) { - down_read(&fs_info->cleanup_work_sem); - err = btrfs_orphan_cleanup(fs_info->fs_root); - if (!err) - err = btrfs_orphan_cleanup(fs_info->tree_root); + if (sb->s_flags & MS_RDONLY) + return 0; + + down_read(&fs_info->cleanup_work_sem); + if ((ret = btrfs_orphan_cleanup(fs_info->fs_root)) || + (ret = btrfs_orphan_cleanup(fs_info->tree_root))) { up_read(&fs_info->cleanup_work_sem); + close_ctree(tree_root); + return ret; + } + up_read(&fs_info->cleanup_work_sem); - if (err) { - close_ctree(tree_root); - return err; - } + ret = btrfs_resume_balance_async(fs_info); + if (ret) { + printk(KERN_WARNING "btrfs: failed to resume balance\n"); + close_ctree(tree_root); + return ret; } return 0; diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index 0eb9a4da069e..e23991574fdf 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -1187,6 +1187,10 @@ static int btrfs_remount(struct super_block *sb, int *flags, char *data) if (ret) goto restore; + ret = btrfs_resume_balance_async(fs_info); + if (ret) + goto restore; + sb->s_flags &= ~MS_RDONLY; } diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 48943d0f861a..ecaad40e7ef4 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -2845,28 +2845,46 @@ out: static int balance_kthread(void *data) { - struct btrfs_balance_control *bctl = - (struct btrfs_balance_control *)data; - struct btrfs_fs_info *fs_info = bctl->fs_info; + struct btrfs_fs_info *fs_info = data; int ret = 0; mutex_lock(&fs_info->volume_mutex); mutex_lock(&fs_info->balance_mutex); - set_balance_control(bctl); - - if (btrfs_test_opt(fs_info->tree_root, SKIP_BALANCE)) { - printk(KERN_INFO "btrfs: force skipping balance\n"); - } else { + if (fs_info->balance_ctl) { printk(KERN_INFO "btrfs: continuing balance\n"); - ret = btrfs_balance(bctl, NULL); + ret = btrfs_balance(fs_info->balance_ctl, NULL); } mutex_unlock(&fs_info->balance_mutex); mutex_unlock(&fs_info->volume_mutex); + return ret; } +int btrfs_resume_balance_async(struct btrfs_fs_info *fs_info) +{ + struct task_struct *tsk; + + spin_lock(&fs_info->balance_lock); + if (!fs_info->balance_ctl) { + spin_unlock(&fs_info->balance_lock); + return 0; + } + spin_unlock(&fs_info->balance_lock); + + if (btrfs_test_opt(fs_info->tree_root, SKIP_BALANCE)) { + printk(KERN_INFO "btrfs: force skipping balance\n"); + return 0; + } + + tsk = kthread_run(balance_kthread, fs_info, "btrfs-balance"); + if (IS_ERR(tsk)) + return PTR_ERR(tsk); + + return 0; +} + int btrfs_recover_balance(struct btrfs_fs_info *fs_info) { struct btrfs_balance_control *bctl; diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h index e1b1a649fc5a..95f6637614db 100644 --- a/fs/btrfs/volumes.h +++ b/fs/btrfs/volumes.h @@ -281,6 +281,7 @@ int btrfs_shrink_device(struct btrfs_device *device, u64 new_size); int btrfs_init_new_device(struct btrfs_root *root, char *path); int btrfs_balance(struct btrfs_balance_control *bctl, struct btrfs_ioctl_balance_args *bargs); +int btrfs_resume_balance_async(struct btrfs_fs_info *fs_info); int btrfs_recover_balance(struct btrfs_fs_info *fs_info); int btrfs_pause_balance(struct btrfs_fs_info *fs_info); int btrfs_cancel_balance(struct btrfs_fs_info *fs_info); -- cgit v1.2.3 From 097b8a7c9e48e2cb50fd0eb9315791921beaf484 Mon Sep 17 00:00:00 2001 From: Jan Schmidt Date: Thu, 21 Jun 2012 11:08:04 +0200 Subject: Btrfs: join tree mod log code with the code holding back delayed refs We've got two mechanisms both required for reliable backref resolving (tree mod log and holding back delayed refs). You cannot make use of one without the other. So instead of requiring the user of this mechanism to setup both correctly, we join them into a single interface. Additionally, we stop inserting non-blockers into fs_info->tree_mod_seq_list as we did before, which was of no value. Signed-off-by: Jan Schmidt --- fs/btrfs/backref.c | 30 ++---- fs/btrfs/backref.h | 3 +- fs/btrfs/ctree.c | 275 ++++++++++++++++++++++++++++++------------------- fs/btrfs/ctree.h | 31 +++--- fs/btrfs/delayed-ref.c | 44 ++++---- fs/btrfs/delayed-ref.h | 49 +-------- fs/btrfs/disk-io.c | 2 + fs/btrfs/extent-tree.c | 21 ++-- fs/btrfs/transaction.c | 4 - 9 files changed, 240 insertions(+), 219 deletions(-) (limited to 'fs/btrfs/disk-io.c') diff --git a/fs/btrfs/backref.c b/fs/btrfs/backref.c index a383c18e74e8..7d80ddd8f544 100644 --- a/fs/btrfs/backref.c +++ b/fs/btrfs/backref.c @@ -773,9 +773,8 @@ static int __add_keyed_refs(struct btrfs_fs_info *fs_info, */ static int find_parent_nodes(struct btrfs_trans_handle *trans, struct btrfs_fs_info *fs_info, u64 bytenr, - u64 delayed_ref_seq, u64 time_seq, - struct ulist *refs, struct ulist *roots, - const u64 *extent_item_pos) + u64 time_seq, struct ulist *refs, + struct ulist *roots, const u64 *extent_item_pos) { struct btrfs_key key; struct btrfs_path *path; @@ -837,7 +836,7 @@ again: btrfs_put_delayed_ref(&head->node); goto again; } - ret = __add_delayed_refs(head, delayed_ref_seq, + ret = __add_delayed_refs(head, time_seq, &prefs_delayed); mutex_unlock(&head->mutex); if (ret) { @@ -981,8 +980,7 @@ static void free_leaf_list(struct ulist *blocks) */ static int btrfs_find_all_leafs(struct btrfs_trans_handle *trans, struct btrfs_fs_info *fs_info, u64 bytenr, - u64 delayed_ref_seq, u64 time_seq, - struct ulist **leafs, + u64 time_seq, struct ulist **leafs, const u64 *extent_item_pos) { struct ulist *tmp; @@ -997,7 +995,7 @@ static int btrfs_find_all_leafs(struct btrfs_trans_handle *trans, return -ENOMEM; } - ret = find_parent_nodes(trans, fs_info, bytenr, delayed_ref_seq, + ret = find_parent_nodes(trans, fs_info, bytenr, time_seq, *leafs, tmp, extent_item_pos); ulist_free(tmp); @@ -1024,8 +1022,7 @@ static int btrfs_find_all_leafs(struct btrfs_trans_handle *trans, */ int btrfs_find_all_roots(struct btrfs_trans_handle *trans, struct btrfs_fs_info *fs_info, u64 bytenr, - u64 delayed_ref_seq, u64 time_seq, - struct ulist **roots) + u64 time_seq, struct ulist **roots) { struct ulist *tmp; struct ulist_node *node = NULL; @@ -1043,7 +1040,7 @@ int btrfs_find_all_roots(struct btrfs_trans_handle *trans, ULIST_ITER_INIT(&uiter); while (1) { - ret = find_parent_nodes(trans, fs_info, bytenr, delayed_ref_seq, + ret = find_parent_nodes(trans, fs_info, bytenr, time_seq, tmp, *roots, NULL); if (ret < 0 && ret != -ENOENT) { ulist_free(tmp); @@ -1376,11 +1373,9 @@ int iterate_extent_inodes(struct btrfs_fs_info *fs_info, struct ulist *roots = NULL; struct ulist_node *ref_node = NULL; struct ulist_node *root_node = NULL; - struct seq_list seq_elem = {}; struct seq_list tree_mod_seq_elem = {}; struct ulist_iterator ref_uiter; struct ulist_iterator root_uiter; - struct btrfs_delayed_ref_root *delayed_refs = NULL; pr_debug("resolving all inodes for extent %llu\n", extent_item_objectid); @@ -1391,16 +1386,11 @@ int iterate_extent_inodes(struct btrfs_fs_info *fs_info, trans = btrfs_join_transaction(fs_info->extent_root); if (IS_ERR(trans)) return PTR_ERR(trans); - - delayed_refs = &trans->transaction->delayed_refs; - spin_lock(&delayed_refs->lock); - btrfs_get_delayed_seq(delayed_refs, &seq_elem); - spin_unlock(&delayed_refs->lock); btrfs_get_tree_mod_seq(fs_info, &tree_mod_seq_elem); } ret = btrfs_find_all_leafs(trans, fs_info, extent_item_objectid, - seq_elem.seq, tree_mod_seq_elem.seq, &refs, + tree_mod_seq_elem.seq, &refs, &extent_item_pos); if (ret) goto out; @@ -1408,8 +1398,7 @@ int iterate_extent_inodes(struct btrfs_fs_info *fs_info, ULIST_ITER_INIT(&ref_uiter); while (!ret && (ref_node = ulist_next(refs, &ref_uiter))) { ret = btrfs_find_all_roots(trans, fs_info, ref_node->val, - seq_elem.seq, - tree_mod_seq_elem.seq, &roots); + tree_mod_seq_elem.seq, &roots); if (ret) break; ULIST_ITER_INIT(&root_uiter); @@ -1431,7 +1420,6 @@ int iterate_extent_inodes(struct btrfs_fs_info *fs_info, out: if (!search_commit_root) { btrfs_put_tree_mod_seq(fs_info, &tree_mod_seq_elem); - btrfs_put_delayed_seq(delayed_refs, &seq_elem); btrfs_end_transaction(trans, fs_info->extent_root); } diff --git a/fs/btrfs/backref.h b/fs/btrfs/backref.h index c18d8ac7b795..3a1ad3e2dcb0 100644 --- a/fs/btrfs/backref.h +++ b/fs/btrfs/backref.h @@ -58,8 +58,7 @@ int paths_from_inode(u64 inum, struct inode_fs_paths *ipath); int btrfs_find_all_roots(struct btrfs_trans_handle *trans, struct btrfs_fs_info *fs_info, u64 bytenr, - u64 delayed_ref_seq, u64 time_seq, - struct ulist **roots); + u64 time_seq, struct ulist **roots); struct btrfs_data_container *init_data_container(u32 total_bytes); struct inode_fs_paths *init_ipath(s32 total_bytes, struct btrfs_root *fs_root, diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c index 67fe46fdee6f..bef68ab32204 100644 --- a/fs/btrfs/ctree.c +++ b/fs/btrfs/ctree.c @@ -321,7 +321,7 @@ struct tree_mod_root { struct tree_mod_elem { struct rb_node node; u64 index; /* shifted logical */ - struct seq_list elem; + u64 seq; enum mod_log_op op; /* this is used for MOD_LOG_KEY_* and MOD_LOG_MOVE_KEYS operations */ @@ -341,20 +341,50 @@ struct tree_mod_elem { struct tree_mod_root old_root; }; -static inline void -__get_tree_mod_seq(struct btrfs_fs_info *fs_info, struct seq_list *elem) +static inline void tree_mod_log_read_lock(struct btrfs_fs_info *fs_info) { - elem->seq = atomic_inc_return(&fs_info->tree_mod_seq); - list_add_tail(&elem->list, &fs_info->tree_mod_seq_list); + read_lock(&fs_info->tree_mod_log_lock); } -void btrfs_get_tree_mod_seq(struct btrfs_fs_info *fs_info, - struct seq_list *elem) +static inline void tree_mod_log_read_unlock(struct btrfs_fs_info *fs_info) +{ + read_unlock(&fs_info->tree_mod_log_lock); +} + +static inline void tree_mod_log_write_lock(struct btrfs_fs_info *fs_info) +{ + write_lock(&fs_info->tree_mod_log_lock); +} + +static inline void tree_mod_log_write_unlock(struct btrfs_fs_info *fs_info) { - elem->flags = 1; + write_unlock(&fs_info->tree_mod_log_lock); +} + +/* + * This adds a new blocker to the tree mod log's blocker list if the @elem + * passed does not already have a sequence number set. So when a caller expects + * to record tree modifications, it should ensure to set elem->seq to zero + * before calling btrfs_get_tree_mod_seq. + * Returns a fresh, unused tree log modification sequence number, even if no new + * blocker was added. + */ +u64 btrfs_get_tree_mod_seq(struct btrfs_fs_info *fs_info, + struct seq_list *elem) +{ + u64 seq; + + tree_mod_log_write_lock(fs_info); spin_lock(&fs_info->tree_mod_seq_lock); - __get_tree_mod_seq(fs_info, elem); + if (!elem->seq) { + elem->seq = btrfs_inc_tree_mod_seq(fs_info); + list_add_tail(&elem->list, &fs_info->tree_mod_seq_list); + } + seq = btrfs_inc_tree_mod_seq(fs_info); spin_unlock(&fs_info->tree_mod_seq_lock); + tree_mod_log_write_unlock(fs_info); + + return seq; } void btrfs_put_tree_mod_seq(struct btrfs_fs_info *fs_info, @@ -371,41 +401,46 @@ void btrfs_put_tree_mod_seq(struct btrfs_fs_info *fs_info, if (!seq_putting) return; - BUG_ON(!(elem->flags & 1)); spin_lock(&fs_info->tree_mod_seq_lock); list_del(&elem->list); + elem->seq = 0; list_for_each_entry(cur_elem, &fs_info->tree_mod_seq_list, list) { - if ((cur_elem->flags & 1) && cur_elem->seq < min_seq) { + if (cur_elem->seq < min_seq) { if (seq_putting > cur_elem->seq) { /* * blocker with lower sequence number exists, we * cannot remove anything from the log */ - goto out; + spin_unlock(&fs_info->tree_mod_seq_lock); + return; } min_seq = cur_elem->seq; } } + spin_unlock(&fs_info->tree_mod_seq_lock); + + /* + * we removed the lowest blocker from the blocker list, so there may be + * more processible delayed refs. + */ + wake_up(&fs_info->tree_mod_seq_wait); /* * anything that's lower than the lowest existing (read: blocked) * sequence number can be removed from the tree. */ - write_lock(&fs_info->tree_mod_log_lock); + tree_mod_log_write_lock(fs_info); tm_root = &fs_info->tree_mod_log; for (node = rb_first(tm_root); node; node = next) { next = rb_next(node); tm = container_of(node, struct tree_mod_elem, node); - if (tm->elem.seq > min_seq) + if (tm->seq > min_seq) continue; rb_erase(node, tm_root); - list_del(&tm->elem.list); kfree(tm); } - write_unlock(&fs_info->tree_mod_log_lock); -out: - spin_unlock(&fs_info->tree_mod_seq_lock); + tree_mod_log_write_unlock(fs_info); } /* @@ -423,11 +458,9 @@ __tree_mod_log_insert(struct btrfs_fs_info *fs_info, struct tree_mod_elem *tm) struct rb_node **new; struct rb_node *parent = NULL; struct tree_mod_elem *cur; - int ret = 0; - BUG_ON(!tm || !tm->elem.seq); + BUG_ON(!tm || !tm->seq); - write_lock(&fs_info->tree_mod_log_lock); tm_root = &fs_info->tree_mod_log; new = &tm_root->rb_node; while (*new) { @@ -437,88 +470,81 @@ __tree_mod_log_insert(struct btrfs_fs_info *fs_info, struct tree_mod_elem *tm) new = &((*new)->rb_left); else if (cur->index > tm->index) new = &((*new)->rb_right); - else if (cur->elem.seq < tm->elem.seq) + else if (cur->seq < tm->seq) new = &((*new)->rb_left); - else if (cur->elem.seq > tm->elem.seq) + else if (cur->seq > tm->seq) new = &((*new)->rb_right); else { kfree(tm); - ret = -EEXIST; - goto unlock; + return -EEXIST; } } rb_link_node(&tm->node, parent, new); rb_insert_color(&tm->node, tm_root); -unlock: - write_unlock(&fs_info->tree_mod_log_lock); - return ret; + return 0; } +/* + * Determines if logging can be omitted. Returns 1 if it can. Otherwise, it + * returns zero with the tree_mod_log_lock acquired. The caller must hold + * this until all tree mod log insertions are recorded in the rb tree and then + * call tree_mod_log_write_unlock() to release. + */ static inline int tree_mod_dont_log(struct btrfs_fs_info *fs_info, struct extent_buffer *eb) { smp_mb(); if (list_empty(&(fs_info)->tree_mod_seq_list)) return 1; - if (!eb) - return 0; - if (btrfs_header_level(eb) == 0) + if (eb && btrfs_header_level(eb) == 0) return 1; + + tree_mod_log_write_lock(fs_info); + if (list_empty(&fs_info->tree_mod_seq_list)) { + /* + * someone emptied the list while we were waiting for the lock. + * we must not add to the list when no blocker exists. + */ + tree_mod_log_write_unlock(fs_info); + return 1; + } + return 0; } /* - * This allocates memory and gets a tree modification sequence number when - * needed. + * This allocates memory and gets a tree modification sequence number. * - * Returns 0 when no sequence number is needed, < 0 on error. - * Returns 1 when a sequence number was added. In this case, - * fs_info->tree_mod_seq_lock was acquired and must be released by the caller - * after inserting into the rb tree. + * Returns <0 on error. + * Returns >0 (the added sequence number) on success. */ static inline int tree_mod_alloc(struct btrfs_fs_info *fs_info, gfp_t flags, struct tree_mod_elem **tm_ret) { struct tree_mod_elem *tm; - int seq; - if (tree_mod_dont_log(fs_info, NULL)) - return 0; - - tm = *tm_ret = kzalloc(sizeof(*tm), flags); + /* + * once we switch from spin locks to something different, we should + * honor the flags parameter here. + */ + tm = *tm_ret = kzalloc(sizeof(*tm), GFP_ATOMIC); if (!tm) return -ENOMEM; - tm->elem.flags = 0; - spin_lock(&fs_info->tree_mod_seq_lock); - if (list_empty(&fs_info->tree_mod_seq_list)) { - /* - * someone emptied the list while we were waiting for the lock. - * we must not add to the list, because no blocker exists. items - * are removed from the list only when the existing blocker is - * removed from the list. - */ - kfree(tm); - seq = 0; - spin_unlock(&fs_info->tree_mod_seq_lock); - } else { - __get_tree_mod_seq(fs_info, &tm->elem); - seq = tm->elem.seq; - } - - return seq; + tm->seq = btrfs_inc_tree_mod_seq(fs_info); + return tm->seq; } -static noinline int -tree_mod_log_insert_key_mask(struct btrfs_fs_info *fs_info, - struct extent_buffer *eb, int slot, - enum mod_log_op op, gfp_t flags) +static inline int +__tree_mod_log_insert_key(struct btrfs_fs_info *fs_info, + struct extent_buffer *eb, int slot, + enum mod_log_op op, gfp_t flags) { - struct tree_mod_elem *tm; int ret; + struct tree_mod_elem *tm; ret = tree_mod_alloc(fs_info, flags, &tm); - if (ret <= 0) + if (ret < 0) return ret; tm->index = eb->start >> PAGE_CACHE_SHIFT; @@ -530,8 +556,22 @@ tree_mod_log_insert_key_mask(struct btrfs_fs_info *fs_info, tm->slot = slot; tm->generation = btrfs_node_ptr_generation(eb, slot); - ret = __tree_mod_log_insert(fs_info, tm); - spin_unlock(&fs_info->tree_mod_seq_lock); + return __tree_mod_log_insert(fs_info, tm); +} + +static noinline int +tree_mod_log_insert_key_mask(struct btrfs_fs_info *fs_info, + struct extent_buffer *eb, int slot, + enum mod_log_op op, gfp_t flags) +{ + int ret; + + if (tree_mod_dont_log(fs_info, eb)) + return 0; + + ret = __tree_mod_log_insert_key(fs_info, eb, slot, op, flags); + + tree_mod_log_write_unlock(fs_info); return ret; } @@ -542,6 +582,14 @@ tree_mod_log_insert_key(struct btrfs_fs_info *fs_info, struct extent_buffer *eb, return tree_mod_log_insert_key_mask(fs_info, eb, slot, op, GFP_NOFS); } +static noinline int +tree_mod_log_insert_key_locked(struct btrfs_fs_info *fs_info, + struct extent_buffer *eb, int slot, + enum mod_log_op op) +{ + return __tree_mod_log_insert_key(fs_info, eb, slot, op, GFP_NOFS); +} + static noinline int tree_mod_log_insert_move(struct btrfs_fs_info *fs_info, struct extent_buffer *eb, int dst_slot, int src_slot, @@ -555,14 +603,14 @@ tree_mod_log_insert_move(struct btrfs_fs_info *fs_info, return 0; for (i = 0; i + dst_slot < src_slot && i < nr_items; i++) { - ret = tree_mod_log_insert_key(fs_info, eb, i + dst_slot, + ret = tree_mod_log_insert_key_locked(fs_info, eb, i + dst_slot, MOD_LOG_KEY_REMOVE_WHILE_MOVING); BUG_ON(ret < 0); } ret = tree_mod_alloc(fs_info, flags, &tm); - if (ret <= 0) - return ret; + if (ret < 0) + goto out; tm->index = eb->start >> PAGE_CACHE_SHIFT; tm->slot = src_slot; @@ -571,10 +619,26 @@ tree_mod_log_insert_move(struct btrfs_fs_info *fs_info, tm->op = MOD_LOG_MOVE_KEYS; ret = __tree_mod_log_insert(fs_info, tm); - spin_unlock(&fs_info->tree_mod_seq_lock); +out: + tree_mod_log_write_unlock(fs_info); return ret; } +static inline void +__tree_mod_log_free_eb(struct btrfs_fs_info *fs_info, struct extent_buffer *eb) +{ + int i; + u32 nritems; + int ret; + + nritems = btrfs_header_nritems(eb); + for (i = nritems - 1; i >= 0; i--) { + ret = tree_mod_log_insert_key_locked(fs_info, eb, i, + MOD_LOG_KEY_REMOVE_WHILE_FREEING); + BUG_ON(ret < 0); + } +} + static noinline int tree_mod_log_insert_root(struct btrfs_fs_info *fs_info, struct extent_buffer *old_root, @@ -583,9 +647,14 @@ tree_mod_log_insert_root(struct btrfs_fs_info *fs_info, struct tree_mod_elem *tm; int ret; + if (tree_mod_dont_log(fs_info, NULL)) + return 0; + + __tree_mod_log_free_eb(fs_info, old_root); + ret = tree_mod_alloc(fs_info, flags, &tm); - if (ret <= 0) - return ret; + if (ret < 0) + goto out; tm->index = new_root->start >> PAGE_CACHE_SHIFT; tm->old_root.logical = old_root->start; @@ -594,7 +663,8 @@ tree_mod_log_insert_root(struct btrfs_fs_info *fs_info, tm->op = MOD_LOG_ROOT_REPLACE; ret = __tree_mod_log_insert(fs_info, tm); - spin_unlock(&fs_info->tree_mod_seq_lock); +out: + tree_mod_log_write_unlock(fs_info); return ret; } @@ -608,7 +678,7 @@ __tree_mod_log_search(struct btrfs_fs_info *fs_info, u64 start, u64 min_seq, struct tree_mod_elem *found = NULL; u64 index = start >> PAGE_CACHE_SHIFT; - read_lock(&fs_info->tree_mod_log_lock); + tree_mod_log_read_lock(fs_info); tm_root = &fs_info->tree_mod_log; node = tm_root->rb_node; while (node) { @@ -617,18 +687,18 @@ __tree_mod_log_search(struct btrfs_fs_info *fs_info, u64 start, u64 min_seq, node = node->rb_left; } else if (cur->index > index) { node = node->rb_right; - } else if (cur->elem.seq < min_seq) { + } else if (cur->seq < min_seq) { node = node->rb_left; } else if (!smallest) { /* we want the node with the highest seq */ if (found) - BUG_ON(found->elem.seq > cur->elem.seq); + BUG_ON(found->seq > cur->seq); found = cur; node = node->rb_left; - } else if (cur->elem.seq > min_seq) { + } else if (cur->seq > min_seq) { /* we want the node with the smallest seq */ if (found) - BUG_ON(found->elem.seq < cur->elem.seq); + BUG_ON(found->seq < cur->seq); found = cur; node = node->rb_right; } else { @@ -636,7 +706,7 @@ __tree_mod_log_search(struct btrfs_fs_info *fs_info, u64 start, u64 min_seq, break; } } - read_unlock(&fs_info->tree_mod_log_lock); + tree_mod_log_read_unlock(fs_info); return found; } @@ -664,7 +734,7 @@ tree_mod_log_search(struct btrfs_fs_info *fs_info, u64 start, u64 min_seq) return __tree_mod_log_search(fs_info, start, min_seq, 0); } -static inline void +static noinline void tree_mod_log_eb_copy(struct btrfs_fs_info *fs_info, struct extent_buffer *dst, struct extent_buffer *src, unsigned long dst_offset, unsigned long src_offset, int nr_items) @@ -675,18 +745,23 @@ tree_mod_log_eb_copy(struct btrfs_fs_info *fs_info, struct extent_buffer *dst, if (tree_mod_dont_log(fs_info, NULL)) return; - if (btrfs_header_level(dst) == 0 && btrfs_header_level(src) == 0) + if (btrfs_header_level(dst) == 0 && btrfs_header_level(src) == 0) { + tree_mod_log_write_unlock(fs_info); return; + } - /* speed this up by single seq for all operations? */ for (i = 0; i < nr_items; i++) { - ret = tree_mod_log_insert_key(fs_info, src, i + src_offset, - MOD_LOG_KEY_REMOVE); + ret = tree_mod_log_insert_key_locked(fs_info, src, + i + src_offset, + MOD_LOG_KEY_REMOVE); BUG_ON(ret < 0); - ret = tree_mod_log_insert_key(fs_info, dst, i + dst_offset, - MOD_LOG_KEY_ADD); + ret = tree_mod_log_insert_key_locked(fs_info, dst, + i + dst_offset, + MOD_LOG_KEY_ADD); BUG_ON(ret < 0); } + + tree_mod_log_write_unlock(fs_info); } static inline void @@ -699,7 +774,7 @@ tree_mod_log_eb_move(struct btrfs_fs_info *fs_info, struct extent_buffer *dst, BUG_ON(ret < 0); } -static inline void +static noinline void tree_mod_log_set_node_key(struct btrfs_fs_info *fs_info, struct extent_buffer *eb, struct btrfs_disk_key *disk_key, int slot, int atomic) @@ -712,30 +787,22 @@ tree_mod_log_set_node_key(struct btrfs_fs_info *fs_info, BUG_ON(ret < 0); } -static void tree_mod_log_free_eb(struct btrfs_fs_info *fs_info, - struct extent_buffer *eb) +static noinline void +tree_mod_log_free_eb(struct btrfs_fs_info *fs_info, struct extent_buffer *eb) { - int i; - int ret; - u32 nritems; - if (tree_mod_dont_log(fs_info, eb)) return; - nritems = btrfs_header_nritems(eb); - for (i = nritems - 1; i >= 0; i--) { - ret = tree_mod_log_insert_key(fs_info, eb, i, - MOD_LOG_KEY_REMOVE_WHILE_FREEING); - BUG_ON(ret < 0); - } + __tree_mod_log_free_eb(fs_info, eb); + + tree_mod_log_write_unlock(fs_info); } -static inline void +static noinline void tree_mod_log_set_root_pointer(struct btrfs_root *root, struct extent_buffer *new_root_node) { int ret; - tree_mod_log_free_eb(root->fs_info, root->node); ret = tree_mod_log_insert_root(root->fs_info, root->node, new_root_node, GFP_NOFS); BUG_ON(ret < 0); @@ -1069,7 +1136,7 @@ __tree_mod_log_rewind(struct extent_buffer *eb, u64 time_seq, unsigned long p_size = sizeof(struct btrfs_key_ptr); n = btrfs_header_nritems(eb); - while (tm && tm->elem.seq >= time_seq) { + while (tm && tm->seq >= time_seq) { /* * all the operations are recorded with the operator used for * the modification. as we're going backwards, we do the diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 84ac723f58f8..8f8dc46f44e7 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -1030,6 +1030,13 @@ struct btrfs_block_group_cache { struct list_head cluster_list; }; +/* delayed seq elem */ +struct seq_list { + struct list_head list; + u64 seq; +}; + +/* fs_info */ struct reloc_control; struct btrfs_device; struct btrfs_fs_devices; @@ -1144,6 +1151,8 @@ struct btrfs_fs_info { spinlock_t tree_mod_seq_lock; atomic_t tree_mod_seq; struct list_head tree_mod_seq_list; + struct seq_list tree_mod_seq_elem; + wait_queue_head_t tree_mod_seq_wait; /* this protects tree_mod_log */ rwlock_t tree_mod_log_lock; @@ -2798,6 +2807,16 @@ static inline void free_fs_info(struct btrfs_fs_info *fs_info) kfree(fs_info); } +/* tree mod log functions from ctree.c */ +u64 btrfs_get_tree_mod_seq(struct btrfs_fs_info *fs_info, + struct seq_list *elem); +void btrfs_put_tree_mod_seq(struct btrfs_fs_info *fs_info, + struct seq_list *elem); +static inline u64 btrfs_inc_tree_mod_seq(struct btrfs_fs_info *fs_info) +{ + return atomic_inc_return(&fs_info->tree_mod_seq); +} + /* root-item.c */ int btrfs_find_root_ref(struct btrfs_root *tree_root, struct btrfs_path *path, @@ -3157,18 +3176,6 @@ void btrfs_reada_detach(void *handle); int btree_readahead_hook(struct btrfs_root *root, struct extent_buffer *eb, u64 start, int err); -/* delayed seq elem */ -struct seq_list { - struct list_head list; - u64 seq; - u32 flags; -}; - -void btrfs_get_tree_mod_seq(struct btrfs_fs_info *fs_info, - struct seq_list *elem); -void btrfs_put_tree_mod_seq(struct btrfs_fs_info *fs_info, - struct seq_list *elem); - static inline int is_fstree(u64 rootid) { if (rootid == BTRFS_FS_TREE_OBJECTID || diff --git a/fs/btrfs/delayed-ref.c b/fs/btrfs/delayed-ref.c index 13ae7b04790e..21a757717637 100644 --- a/fs/btrfs/delayed-ref.c +++ b/fs/btrfs/delayed-ref.c @@ -233,22 +233,26 @@ int btrfs_delayed_ref_lock(struct btrfs_trans_handle *trans, return 0; } -int btrfs_check_delayed_seq(struct btrfs_delayed_ref_root *delayed_refs, +int btrfs_check_delayed_seq(struct btrfs_fs_info *fs_info, + struct btrfs_delayed_ref_root *delayed_refs, u64 seq) { struct seq_list *elem; - - assert_spin_locked(&delayed_refs->lock); - if (list_empty(&delayed_refs->seq_head)) - return 0; - - elem = list_first_entry(&delayed_refs->seq_head, struct seq_list, list); - if (seq >= elem->seq) { - pr_debug("holding back delayed_ref %llu, lowest is %llu (%p)\n", - seq, elem->seq, delayed_refs); - return 1; + int ret = 0; + + spin_lock(&fs_info->tree_mod_seq_lock); + if (!list_empty(&fs_info->tree_mod_seq_list)) { + elem = list_first_entry(&fs_info->tree_mod_seq_list, + struct seq_list, list); + if (seq >= elem->seq) { + pr_debug("holding back delayed_ref %llu, lowest is " + "%llu (%p)\n", seq, elem->seq, delayed_refs); + ret = 1; + } } - return 0; + + spin_unlock(&fs_info->tree_mod_seq_lock); + return ret; } int btrfs_find_ref_cluster(struct btrfs_trans_handle *trans, @@ -526,7 +530,7 @@ static noinline void add_delayed_tree_ref(struct btrfs_fs_info *fs_info, ref->in_tree = 1; if (is_fstree(ref_root)) - seq = inc_delayed_seq(delayed_refs); + seq = btrfs_inc_tree_mod_seq(fs_info); ref->seq = seq; full_ref = btrfs_delayed_node_to_tree_ref(ref); @@ -585,7 +589,7 @@ static noinline void add_delayed_data_ref(struct btrfs_fs_info *fs_info, ref->in_tree = 1; if (is_fstree(ref_root)) - seq = inc_delayed_seq(delayed_refs); + seq = btrfs_inc_tree_mod_seq(fs_info); ref->seq = seq; full_ref = btrfs_delayed_node_to_data_ref(ref); @@ -659,8 +663,8 @@ int btrfs_add_delayed_tree_ref(struct btrfs_fs_info *fs_info, num_bytes, parent, ref_root, level, action, for_cow); if (!is_fstree(ref_root) && - waitqueue_active(&delayed_refs->seq_wait)) - wake_up(&delayed_refs->seq_wait); + waitqueue_active(&fs_info->tree_mod_seq_wait)) + wake_up(&fs_info->tree_mod_seq_wait); spin_unlock(&delayed_refs->lock); return 0; @@ -708,8 +712,8 @@ int btrfs_add_delayed_data_ref(struct btrfs_fs_info *fs_info, num_bytes, parent, ref_root, owner, offset, action, for_cow); if (!is_fstree(ref_root) && - waitqueue_active(&delayed_refs->seq_wait)) - wake_up(&delayed_refs->seq_wait); + waitqueue_active(&fs_info->tree_mod_seq_wait)) + wake_up(&fs_info->tree_mod_seq_wait); spin_unlock(&delayed_refs->lock); return 0; @@ -736,8 +740,8 @@ int btrfs_add_delayed_extent_op(struct btrfs_fs_info *fs_info, num_bytes, BTRFS_UPDATE_DELAYED_HEAD, extent_op->is_data); - if (waitqueue_active(&delayed_refs->seq_wait)) - wake_up(&delayed_refs->seq_wait); + if (waitqueue_active(&fs_info->tree_mod_seq_wait)) + wake_up(&fs_info->tree_mod_seq_wait); spin_unlock(&delayed_refs->lock); return 0; } diff --git a/fs/btrfs/delayed-ref.h b/fs/btrfs/delayed-ref.h index 413927fb9957..2b5cb27f9861 100644 --- a/fs/btrfs/delayed-ref.h +++ b/fs/btrfs/delayed-ref.h @@ -139,26 +139,6 @@ struct btrfs_delayed_ref_root { int flushing; u64 run_delayed_start; - - /* - * seq number of delayed refs. We need to know if a backref was being - * added before the currently processed ref or afterwards. - */ - u64 seq; - - /* - * seq_list holds a list of all seq numbers that are currently being - * added to the list. While walking backrefs (btrfs_find_all_roots, - * qgroups), which might take some time, no newer ref must be processed, - * as it might influence the outcome of the walk. - */ - struct list_head seq_head; - - /* - * when the only refs we have in the list must not be processed, we want - * to wait for more refs to show up or for the end of backref walking. - */ - wait_queue_head_t seq_wait; }; static inline void btrfs_put_delayed_ref(struct btrfs_delayed_ref_node *ref) @@ -195,33 +175,8 @@ int btrfs_delayed_ref_lock(struct btrfs_trans_handle *trans, int btrfs_find_ref_cluster(struct btrfs_trans_handle *trans, struct list_head *cluster, u64 search_start); -static inline u64 inc_delayed_seq(struct btrfs_delayed_ref_root *delayed_refs) -{ - assert_spin_locked(&delayed_refs->lock); - ++delayed_refs->seq; - return delayed_refs->seq; -} - -static inline void -btrfs_get_delayed_seq(struct btrfs_delayed_ref_root *delayed_refs, - struct seq_list *elem) -{ - assert_spin_locked(&delayed_refs->lock); - elem->seq = delayed_refs->seq; - list_add_tail(&elem->list, &delayed_refs->seq_head); -} - -static inline void -btrfs_put_delayed_seq(struct btrfs_delayed_ref_root *delayed_refs, - struct seq_list *elem) -{ - spin_lock(&delayed_refs->lock); - list_del(&elem->list); - wake_up(&delayed_refs->seq_wait); - spin_unlock(&delayed_refs->lock); -} - -int btrfs_check_delayed_seq(struct btrfs_delayed_ref_root *delayed_refs, +int btrfs_check_delayed_seq(struct btrfs_fs_info *fs_info, + struct btrfs_delayed_ref_root *delayed_refs, u64 seq); /* diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 8cc47103a32e..19a39e10d6f5 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -1944,6 +1944,8 @@ int open_ctree(struct super_block *sb, fs_info->free_chunk_space = 0; fs_info->tree_mod_log = RB_ROOT; + init_waitqueue_head(&fs_info->tree_mod_seq_wait); + /* readahead state */ INIT_RADIX_TREE(&fs_info->reada_tree, GFP_NOFS & ~__GFP_WAIT); spin_lock_init(&fs_info->reada_lock); diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 6e1d36702ff7..94ce79f76e5f 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -2217,6 +2217,7 @@ static noinline int run_clustered_refs(struct btrfs_trans_handle *trans, struct btrfs_delayed_ref_node *ref; struct btrfs_delayed_ref_head *locked_ref = NULL; struct btrfs_delayed_extent_op *extent_op; + struct btrfs_fs_info *fs_info = root->fs_info; int ret; int count = 0; int must_insert_reserved = 0; @@ -2255,7 +2256,7 @@ static noinline int run_clustered_refs(struct btrfs_trans_handle *trans, ref = select_delayed_ref(locked_ref); if (ref && ref->seq && - btrfs_check_delayed_seq(delayed_refs, ref->seq)) { + btrfs_check_delayed_seq(fs_info, delayed_refs, ref->seq)) { /* * there are still refs with lower seq numbers in the * process of being added. Don't run this ref yet. @@ -2337,7 +2338,7 @@ static noinline int run_clustered_refs(struct btrfs_trans_handle *trans, } next: - do_chunk_alloc(trans, root->fs_info->extent_root, + do_chunk_alloc(trans, fs_info->extent_root, 2 * 1024 * 1024, btrfs_get_alloc_profile(root, 0), CHUNK_ALLOC_NO_FORCE); @@ -2347,18 +2348,19 @@ next: return count; } -static void wait_for_more_refs(struct btrfs_delayed_ref_root *delayed_refs, +static void wait_for_more_refs(struct btrfs_fs_info *fs_info, + struct btrfs_delayed_ref_root *delayed_refs, unsigned long num_refs, struct list_head *first_seq) { spin_unlock(&delayed_refs->lock); pr_debug("waiting for more refs (num %ld, first %p)\n", num_refs, first_seq); - wait_event(delayed_refs->seq_wait, + wait_event(fs_info->tree_mod_seq_wait, num_refs != delayed_refs->num_entries || - delayed_refs->seq_head.next != first_seq); + fs_info->tree_mod_seq_list.next != first_seq); pr_debug("done waiting for more refs (num %ld, first %p)\n", - delayed_refs->num_entries, delayed_refs->seq_head.next); + delayed_refs->num_entries, fs_info->tree_mod_seq_list.next); spin_lock(&delayed_refs->lock); } @@ -2403,6 +2405,7 @@ int btrfs_run_delayed_refs(struct btrfs_trans_handle *trans, again: consider_waiting = 0; spin_lock(&delayed_refs->lock); + if (count == 0) { count = delayed_refs->num_entries * 2; run_most = 1; @@ -2437,7 +2440,7 @@ again: num_refs = delayed_refs->num_entries; first_seq = root->fs_info->tree_mod_seq_list.next; } else { - wait_for_more_refs(delayed_refs, + wait_for_more_refs(root->fs_info, delayed_refs, num_refs, first_seq); /* * after waiting, things have changed. we @@ -5190,8 +5193,8 @@ static noinline int check_ref_cleanup(struct btrfs_trans_handle *trans, rb_erase(&head->node.rb_node, &delayed_refs->root); delayed_refs->num_entries--; - if (waitqueue_active(&delayed_refs->seq_wait)) - wake_up(&delayed_refs->seq_wait); + if (waitqueue_active(&root->fs_info->tree_mod_seq_wait)) + wake_up(&root->fs_info->tree_mod_seq_wait); /* * we don't take a ref on the node because we're removing it from the diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index b72b068183ec..621c8dc48fb6 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -38,7 +38,6 @@ void put_transaction(struct btrfs_transaction *transaction) if (atomic_dec_and_test(&transaction->use_count)) { BUG_ON(!list_empty(&transaction->list)); WARN_ON(transaction->delayed_refs.root.rb_node); - WARN_ON(!list_empty(&transaction->delayed_refs.seq_head)); memset(transaction, 0, sizeof(*transaction)); kmem_cache_free(btrfs_transaction_cachep, transaction); } @@ -126,7 +125,6 @@ loop: cur_trans->delayed_refs.num_heads = 0; cur_trans->delayed_refs.flushing = 0; cur_trans->delayed_refs.run_delayed_start = 0; - cur_trans->delayed_refs.seq = 1; /* * although the tree mod log is per file system and not per transaction, @@ -145,10 +143,8 @@ loop: } atomic_set(&fs_info->tree_mod_seq, 0); - init_waitqueue_head(&cur_trans->delayed_refs.seq_wait); spin_lock_init(&cur_trans->commit_lock); spin_lock_init(&cur_trans->delayed_refs.lock); - INIT_LIST_HEAD(&cur_trans->delayed_refs.seq_head); INIT_LIST_HEAD(&cur_trans->pending_snapshots); list_add_tail(&cur_trans->list, &fs_info->trans_list); -- cgit v1.2.3 From 20897f5c86b9d2b77baea1d48eda7fa4ac217279 Mon Sep 17 00:00:00 2001 From: Arne Jansen Date: Tue, 13 Sep 2011 12:44:20 +0200 Subject: Btrfs: added helper to create new trees This creates a brand new tree. Will be used to create the quota tree. Signed-off-by: Arne Jansen --- fs/btrfs/disk-io.c | 78 +++++++++++++++++++++++++++++++++++++++++++++++++++++- fs/btrfs/disk-io.h | 6 +++++ 2 files changed, 83 insertions(+), 1 deletion(-) (limited to 'fs/btrfs/disk-io.c') diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 19a39e10d6f5..6fc243eccffa 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -1225,6 +1225,82 @@ static struct btrfs_root *btrfs_alloc_root(struct btrfs_fs_info *fs_info) return root; } +struct btrfs_root *btrfs_create_tree(struct btrfs_trans_handle *trans, + struct btrfs_fs_info *fs_info, + u64 objectid) +{ + struct extent_buffer *leaf; + struct btrfs_root *tree_root = fs_info->tree_root; + struct btrfs_root *root; + struct btrfs_key key; + int ret = 0; + u64 bytenr; + + root = btrfs_alloc_root(fs_info); + if (!root) + return ERR_PTR(-ENOMEM); + + __setup_root(tree_root->nodesize, tree_root->leafsize, + tree_root->sectorsize, tree_root->stripesize, + root, fs_info, objectid); + root->root_key.objectid = objectid; + root->root_key.type = BTRFS_ROOT_ITEM_KEY; + root->root_key.offset = 0; + + leaf = btrfs_alloc_free_block(trans, root, root->leafsize, + 0, objectid, NULL, 0, 0, 0); + if (IS_ERR(leaf)) { + ret = PTR_ERR(leaf); + goto fail; + } + + bytenr = leaf->start; + memset_extent_buffer(leaf, 0, 0, sizeof(struct btrfs_header)); + btrfs_set_header_bytenr(leaf, leaf->start); + btrfs_set_header_generation(leaf, trans->transid); + btrfs_set_header_backref_rev(leaf, BTRFS_MIXED_BACKREF_REV); + btrfs_set_header_owner(leaf, objectid); + root->node = leaf; + + write_extent_buffer(leaf, fs_info->fsid, + (unsigned long)btrfs_header_fsid(leaf), + BTRFS_FSID_SIZE); + write_extent_buffer(leaf, fs_info->chunk_tree_uuid, + (unsigned long)btrfs_header_chunk_tree_uuid(leaf), + BTRFS_UUID_SIZE); + btrfs_mark_buffer_dirty(leaf); + + root->commit_root = btrfs_root_node(root); + root->track_dirty = 1; + + + root->root_item.flags = 0; + root->root_item.byte_limit = 0; + btrfs_set_root_bytenr(&root->root_item, leaf->start); + btrfs_set_root_generation(&root->root_item, trans->transid); + btrfs_set_root_level(&root->root_item, 0); + btrfs_set_root_refs(&root->root_item, 1); + btrfs_set_root_used(&root->root_item, leaf->len); + btrfs_set_root_last_snapshot(&root->root_item, 0); + btrfs_set_root_dirid(&root->root_item, 0); + root->root_item.drop_level = 0; + + key.objectid = objectid; + key.type = BTRFS_ROOT_ITEM_KEY; + key.offset = 0; + ret = btrfs_insert_root(trans, tree_root, &key, &root->root_item); + if (ret) + goto fail; + + btrfs_tree_unlock(leaf); + +fail: + if (ret) + return ERR_PTR(ret); + + return root; +} + static struct btrfs_root *alloc_log_tree(struct btrfs_trans_handle *trans, struct btrfs_fs_info *fs_info) { @@ -3260,7 +3336,7 @@ int btrfs_read_buffer(struct extent_buffer *buf, u64 parent_transid) return btree_read_extent_buffer_pages(root, buf, 0, parent_transid); } -static int btree_lock_page_hook(struct page *page, void *data, +int btree_lock_page_hook(struct page *page, void *data, void (*flush_fn)(void *)) { struct inode *inode = page->mapping->host; diff --git a/fs/btrfs/disk-io.h b/fs/btrfs/disk-io.h index 05b3fab39f7e..95e147eea239 100644 --- a/fs/btrfs/disk-io.h +++ b/fs/btrfs/disk-io.h @@ -89,6 +89,12 @@ int btrfs_add_log_tree(struct btrfs_trans_handle *trans, int btrfs_cleanup_transaction(struct btrfs_root *root); void btrfs_cleanup_one_transaction(struct btrfs_transaction *trans, struct btrfs_root *root); +void btrfs_abort_devices(struct btrfs_root *root); +struct btrfs_root *btrfs_create_tree(struct btrfs_trans_handle *trans, + struct btrfs_fs_info *fs_info, + u64 objectid); +int btree_lock_page_hook(struct page *page, void *data, + void (*flush_fn)(void *)); #ifdef CONFIG_DEBUG_LOCK_ALLOC void btrfs_init_lockdep(void); -- cgit v1.2.3 From 416ac51da90e98daaac17e1f359a6c5591f7f5bd Mon Sep 17 00:00:00 2001 From: Arne Jansen Date: Tue, 13 Sep 2011 12:56:09 +0200 Subject: Btrfs: qgroup state and initialization Add state to fs_info. Signed-off-by: Arne Jansen --- fs/btrfs/ctree.h | 24 ++++++++++++++++++++++++ fs/btrfs/disk-io.c | 7 +++++++ 2 files changed, 31 insertions(+) (limited to 'fs/btrfs/disk-io.c') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 27cf995564ed..a5269d4a164f 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -1120,6 +1120,7 @@ struct btrfs_fs_info { struct btrfs_root *dev_root; struct btrfs_root *fs_root; struct btrfs_root *csum_root; + struct btrfs_root *quota_root; /* the log root tree is a directory of all the other log roots */ struct btrfs_root *log_root_tree; @@ -1374,6 +1375,29 @@ struct btrfs_fs_info { #ifdef CONFIG_BTRFS_FS_CHECK_INTEGRITY u32 check_integrity_print_mask; #endif + /* + * quota information + */ + unsigned int quota_enabled:1; + + /* + * quota_enabled only changes state after a commit. This holds the + * next state. + */ + unsigned int pending_quota_state:1; + + /* is qgroup tracking in a consistent state? */ + u64 qgroup_flags; + + /* holds configuration and tracking. Protected by qgroup_lock */ + struct rb_root qgroup_tree; + spinlock_t qgroup_lock; + + /* list of dirty qgroups to be written at next commit */ + struct list_head dirty_qgroups; + + /* used by btrfs_qgroup_record_ref for an efficient tree traversal */ + u64 qgroup_seq; /* filesystem state */ u64 fs_state; diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 6fc243eccffa..eca054974425 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -2110,6 +2110,13 @@ int open_ctree(struct super_block *sb, init_rwsem(&fs_info->cleanup_work_sem); init_rwsem(&fs_info->subvol_sem); + spin_lock_init(&fs_info->qgroup_lock); + fs_info->qgroup_tree = RB_ROOT; + INIT_LIST_HEAD(&fs_info->dirty_qgroups); + fs_info->qgroup_seq = 1; + fs_info->quota_enabled = 0; + fs_info->pending_quota_state = 0; + btrfs_init_free_cluster(&fs_info->meta_alloc_cluster); btrfs_init_free_cluster(&fs_info->data_alloc_cluster); -- cgit v1.2.3 From bcef60f249034f69e89e544461cbfecb68975595 Mon Sep 17 00:00:00 2001 From: Arne Jansen Date: Tue, 13 Sep 2011 15:23:30 +0200 Subject: Btrfs: quota tree support and startup Init the quota tree along with the others on open_ctree and close_ctree. Add the quota tree to the list of well known trees in btrfs_read_fs_root_no_name. Signed-off-by: Arne Jansen --- fs/btrfs/ctree.h | 1 + fs/btrfs/disk-io.c | 47 +++++++++++++++++++++++++++++++++++++++++------ 2 files changed, 42 insertions(+), 6 deletions(-) (limited to 'fs/btrfs/disk-io.c') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index ccba9b684c96..2ba03b96fbe0 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -2967,6 +2967,7 @@ static inline void free_fs_info(struct btrfs_fs_info *fs_info) kfree(fs_info->chunk_root); kfree(fs_info->dev_root); kfree(fs_info->csum_root); + kfree(fs_info->quota_root); kfree(fs_info->super_copy); kfree(fs_info->super_for_commit); kfree(fs_info); diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index eca054974425..87d9391c0576 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -1472,6 +1472,9 @@ struct btrfs_root *btrfs_read_fs_root_no_name(struct btrfs_fs_info *fs_info, return fs_info->dev_root; if (location->objectid == BTRFS_CSUM_TREE_OBJECTID) return fs_info->csum_root; + if (location->objectid == BTRFS_QUOTA_TREE_OBJECTID) + return fs_info->quota_root ? fs_info->quota_root : + ERR_PTR(-ENOENT); again: spin_lock(&fs_info->fs_roots_radix_lock); root = radix_tree_lookup(&fs_info->fs_roots_radix, @@ -1899,6 +1902,10 @@ static void free_root_pointers(struct btrfs_fs_info *info, int chunk_root) free_extent_buffer(info->extent_root->commit_root); free_extent_buffer(info->csum_root->node); free_extent_buffer(info->csum_root->commit_root); + if (info->quota_root) { + free_extent_buffer(info->quota_root->node); + free_extent_buffer(info->quota_root->commit_root); + } info->tree_root->node = NULL; info->tree_root->commit_root = NULL; @@ -1908,6 +1915,10 @@ static void free_root_pointers(struct btrfs_fs_info *info, int chunk_root) info->extent_root->commit_root = NULL; info->csum_root->node = NULL; info->csum_root->commit_root = NULL; + if (info->quota_root) { + info->quota_root->node = NULL; + info->quota_root->commit_root = NULL; + } if (chunk_root) { free_extent_buffer(info->chunk_root->node); @@ -1938,6 +1949,7 @@ int open_ctree(struct super_block *sb, struct btrfs_root *csum_root; struct btrfs_root *chunk_root; struct btrfs_root *dev_root; + struct btrfs_root *quota_root; struct btrfs_root *log_tree_root; int ret; int err = -EINVAL; @@ -1949,9 +1961,10 @@ int open_ctree(struct super_block *sb, csum_root = fs_info->csum_root = btrfs_alloc_root(fs_info); chunk_root = fs_info->chunk_root = btrfs_alloc_root(fs_info); dev_root = fs_info->dev_root = btrfs_alloc_root(fs_info); + quota_root = fs_info->quota_root = btrfs_alloc_root(fs_info); if (!tree_root || !extent_root || !csum_root || - !chunk_root || !dev_root) { + !chunk_root || !dev_root || !quota_root) { err = -ENOMEM; goto fail; } @@ -2441,6 +2454,17 @@ retry_root_backup: goto recovery_tree_root; csum_root->track_dirty = 1; + ret = find_and_setup_root(tree_root, fs_info, + BTRFS_QUOTA_TREE_OBJECTID, quota_root); + if (ret) { + kfree(quota_root); + quota_root = fs_info->quota_root = NULL; + } else { + quota_root->track_dirty = 1; + fs_info->quota_enabled = 1; + fs_info->pending_quota_state = 1; + } + fs_info->generation = generation; fs_info->last_trans_committed = generation; @@ -2500,6 +2524,9 @@ retry_root_backup: " integrity check module %s\n", sb->s_id); } #endif + ret = btrfs_read_qgroup_config(fs_info); + if (ret) + goto fail_trans_kthread; /* do not make disk changes in broken FS */ if (btrfs_super_log_root(disk_super) != 0 && @@ -2510,7 +2537,7 @@ retry_root_backup: printk(KERN_WARNING "Btrfs log replay required " "on RO media\n"); err = -EIO; - goto fail_trans_kthread; + goto fail_qgroup; } blocksize = btrfs_level_size(tree_root, @@ -2519,7 +2546,7 @@ retry_root_backup: log_tree_root = btrfs_alloc_root(fs_info); if (!log_tree_root) { err = -ENOMEM; - goto fail_trans_kthread; + goto fail_qgroup; } __setup_root(nodesize, leafsize, sectorsize, stripesize, @@ -2559,7 +2586,7 @@ retry_root_backup: printk(KERN_WARNING "btrfs: failed to recover relocation\n"); err = -EINVAL; - goto fail_trans_kthread; + goto fail_qgroup; } } @@ -2569,10 +2596,10 @@ retry_root_backup: fs_info->fs_root = btrfs_read_fs_root_no_name(fs_info, &location); if (!fs_info->fs_root) - goto fail_trans_kthread; + goto fail_qgroup; if (IS_ERR(fs_info->fs_root)) { err = PTR_ERR(fs_info->fs_root); - goto fail_trans_kthread; + goto fail_qgroup; } if (sb->s_flags & MS_RDONLY) @@ -2596,6 +2623,8 @@ retry_root_backup: return 0; +fail_qgroup: + btrfs_free_qgroup_config(fs_info); fail_trans_kthread: kthread_stop(fs_info->transaction_kthread); fail_cleaner: @@ -3194,6 +3223,8 @@ int close_ctree(struct btrfs_root *root) fs_info->closing = 2; smp_mb(); + btrfs_free_qgroup_config(root->fs_info); + if (fs_info->delalloc_bytes) { printk(KERN_INFO "btrfs: at unmount delalloc count %llu\n", (unsigned long long)fs_info->delalloc_bytes); @@ -3213,6 +3244,10 @@ int close_ctree(struct btrfs_root *root) free_extent_buffer(fs_info->dev_root->commit_root); free_extent_buffer(fs_info->csum_root->node); free_extent_buffer(fs_info->csum_root->commit_root); + if (fs_info->quota_root) { + free_extent_buffer(fs_info->quota_root->node); + free_extent_buffer(fs_info->quota_root->commit_root); + } btrfs_free_block_groups(fs_info); -- cgit v1.2.3 From fed425c742cb1262ce90a41f2d3d211bac099533 Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Fri, 22 Jun 2012 12:13:01 -0600 Subject: Btrfs: do not return EINVAL instead of ENOMEM from open_ctree() When bailing from open_ctree() err is returned, not ret. Signed-off-by: Ilya Dryomov --- fs/btrfs/disk-io.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs/btrfs/disk-io.c') diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 2936ca49b3b4..fd216d9369fa 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -2244,7 +2244,7 @@ int open_ctree(struct super_block *sb, ret |= btrfs_start_workers(&fs_info->caching_workers); ret |= btrfs_start_workers(&fs_info->readahead_workers); if (ret) { - ret = -ENOMEM; + err = -ENOMEM; goto fail_sb_buffer; } -- cgit v1.2.3 From 44c44af2f4a6dc1595f1711cf307bd01062fd129 Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Fri, 22 Jun 2012 12:14:13 -0600 Subject: Btrfs: do not ignore errors from btrfs_cleanup_fs_roots() when mounting There used to be a BUG_ON(ret) there before EH patch (79787eaa) went in. Bail out with EINVAL. Cc: David Sterba Signed-off-by: Ilya Dryomov --- fs/btrfs/disk-io.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs/btrfs/disk-io.c') diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index fd216d9369fa..dd6676b446f6 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -2466,8 +2466,8 @@ retry_root_backup: if (!(sb->s_flags & MS_RDONLY)) { ret = btrfs_cleanup_fs_roots(fs_info); - if (ret) { - } + if (ret) + goto fail_trans_kthread; ret = btrfs_recover_relocation(tree_root); if (ret < 0) { -- cgit v1.2.3 From c0901581ad077004145c9ee80e843fba71c100b8 Mon Sep 17 00:00:00 2001 From: Stefan Behrens Date: Tue, 10 Jul 2012 07:30:17 -0600 Subject: Btrfs: avoid I/O repair BUG() from btree_read_extent_buffer_pages() From btree_read_extent_buffer_pages(), currently repair_io_failure() can be called with mirror_num being zero when submit_one_bio() returned an error before. This used to cause a BUG_ON(!mirror_num) in repair_io_failure() and indeed this is not a case that needs the I/O repair code to rewrite disk blocks. This commit prevents calling repair_io_failure() in this case and thus avoids the BUG_ON() and malfunction. Signed-off-by: Stefan Behrens Signed-off-by: Josef Bacik --- fs/btrfs/disk-io.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs/btrfs/disk-io.c') diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index dd6676b446f6..1a4a2a975926 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -407,7 +407,7 @@ static int btree_read_extent_buffer_pages(struct btrfs_root *root, break; } - if (failed && !ret) + if (failed && !ret && failed_mirror) repair_eb_io_failure(root, eb, failed_mirror); return ret; -- cgit v1.2.3 From 8ea05e3a4262b9e6871c349fa3486bcfc72ffd1a Mon Sep 17 00:00:00 2001 From: Alexander Block Date: Wed, 25 Jul 2012 17:35:53 +0200 Subject: Btrfs: introduce subvol uuids and times This patch introduces uuids for subvolumes. Each subvolume has it's own uuid. In case it was snapshotted, it also contains parent_uuid. In case it was received, it also contains received_uuid. It also introduces subvolume ctime/otime/stime/rtime. The first two are comparable to the times found in inodes. otime is the origin/creation time and ctime is the change time. stime/rtime are only valid on received subvolumes. stime is the time of the subvolume when it was sent. rtime is the time of the subvolume when it was received. Additionally to the times, we have a transid for each time. They are updated at the same place as the times. btrfs receive uses stransid and rtransid to find out if a received subvolume changed in the meantime. If an older kernel mounts a filesystem with the extented fields, all fields become invalid. The next mount with a new kernel will detect this and reset the fields. Signed-off-by: Alexander Block Reviewed-by: David Sterba Reviewed-by: Arne Jansen Reviewed-by: Jan Schmidt Reviewed-by: Alex Lyakas --- fs/btrfs/check-integrity.c | 7 +-- fs/btrfs/ctree.h | 47 ++++++++++++++++++++ fs/btrfs/disk-io.c | 8 ++-- fs/btrfs/inode.c | 4 ++ fs/btrfs/ioctl.c | 100 ++++++++++++++++++++++++++++++++++++++++-- fs/btrfs/ioctl.h | 17 +++++++ fs/btrfs/root-tree.c | 107 ++++++++++++++++++++++++++++++++++++++++++--- fs/btrfs/transaction.c | 17 +++++++ 8 files changed, 292 insertions(+), 15 deletions(-) (limited to 'fs/btrfs/disk-io.c') diff --git a/fs/btrfs/check-integrity.c b/fs/btrfs/check-integrity.c index da6e9364a5e3..9197e2e33407 100644 --- a/fs/btrfs/check-integrity.c +++ b/fs/btrfs/check-integrity.c @@ -1032,6 +1032,7 @@ continue_with_current_leaf_stack_frame: struct btrfs_disk_key *disk_key; u8 type; u32 item_offset; + u32 item_size; if (disk_item_offset + sizeof(struct btrfs_item) > sf->block_ctx->len) { @@ -1047,6 +1048,7 @@ leaf_item_out_of_bounce_error: disk_item_offset, sizeof(struct btrfs_item)); item_offset = le32_to_cpu(disk_item.offset); + item_size = le32_to_cpu(disk_item.size); disk_key = &disk_item.key; type = disk_key->type; @@ -1057,14 +1059,13 @@ leaf_item_out_of_bounce_error: root_item_offset = item_offset + offsetof(struct btrfs_leaf, items); - if (root_item_offset + - sizeof(struct btrfs_root_item) > + if (root_item_offset + item_size > sf->block_ctx->len) goto leaf_item_out_of_bounce_error; btrfsic_read_from_block_data( sf->block_ctx, &root_item, root_item_offset, - sizeof(struct btrfs_root_item)); + item_size); next_bytenr = le64_to_cpu(root_item.bytenr); sf->error = diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 8cfde9326dd6..d5f6d7458676 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -709,6 +709,36 @@ struct btrfs_root_item { struct btrfs_disk_key drop_progress; u8 drop_level; u8 level; + + /* + * The following fields appear after subvol_uuids+subvol_times + * were introduced. + */ + + /* + * This generation number is used to test if the new fields are valid + * and up to date while reading the root item. Everytime the root item + * is written out, the "generation" field is copied into this field. If + * anyone ever mounted the fs with an older kernel, we will have + * mismatching generation values here and thus must invalidate the + * new fields. See btrfs_update_root and btrfs_find_last_root for + * details. + * the offset of generation_v2 is also used as the start for the memset + * when invalidating the fields. + */ + __le64 generation_v2; + u8 uuid[BTRFS_UUID_SIZE]; + u8 parent_uuid[BTRFS_UUID_SIZE]; + u8 received_uuid[BTRFS_UUID_SIZE]; + __le64 ctransid; /* updated when an inode changes */ + __le64 otransid; /* trans when created */ + __le64 stransid; /* trans when sent. non-zero for received subvol */ + __le64 rtransid; /* trans when received. non-zero for received subvol */ + struct btrfs_timespec ctime; + struct btrfs_timespec otime; + struct btrfs_timespec stime; + struct btrfs_timespec rtime; + __le64 reserved[8]; /* for future */ } __attribute__ ((__packed__)); /* @@ -1416,6 +1446,8 @@ struct btrfs_root { dev_t anon_dev; int force_cow; + + spinlock_t root_times_lock; }; struct btrfs_ioctl_defrag_range_args { @@ -2189,6 +2221,16 @@ BTRFS_SETGET_STACK_FUNCS(root_used, struct btrfs_root_item, bytes_used, 64); BTRFS_SETGET_STACK_FUNCS(root_limit, struct btrfs_root_item, byte_limit, 64); BTRFS_SETGET_STACK_FUNCS(root_last_snapshot, struct btrfs_root_item, last_snapshot, 64); +BTRFS_SETGET_STACK_FUNCS(root_generation_v2, struct btrfs_root_item, + generation_v2, 64); +BTRFS_SETGET_STACK_FUNCS(root_ctransid, struct btrfs_root_item, + ctransid, 64); +BTRFS_SETGET_STACK_FUNCS(root_otransid, struct btrfs_root_item, + otransid, 64); +BTRFS_SETGET_STACK_FUNCS(root_stransid, struct btrfs_root_item, + stransid, 64); +BTRFS_SETGET_STACK_FUNCS(root_rtransid, struct btrfs_root_item, + rtransid, 64); static inline bool btrfs_root_readonly(struct btrfs_root *root) { @@ -2822,6 +2864,9 @@ int __must_check btrfs_update_root(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct btrfs_key *key, struct btrfs_root_item *item); +void btrfs_read_root_item(struct btrfs_root *root, + struct extent_buffer *eb, int slot, + struct btrfs_root_item *item); int btrfs_find_last_root(struct btrfs_root *root, u64 objectid, struct btrfs_root_item *item, struct btrfs_key *key); int btrfs_find_dead_roots(struct btrfs_root *root, u64 objectid); @@ -2829,6 +2874,8 @@ int btrfs_find_orphan_roots(struct btrfs_root *tree_root); void btrfs_set_root_node(struct btrfs_root_item *item, struct extent_buffer *node); void btrfs_check_and_init_root_item(struct btrfs_root_item *item); +void btrfs_update_root_times(struct btrfs_trans_handle *trans, + struct btrfs_root *root); /* dir-item.c */ int btrfs_insert_dir_item(struct btrfs_trans_handle *trans, diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 2936ca49b3b4..c39eb71fae31 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -1182,6 +1182,8 @@ static void __setup_root(u32 nodesize, u32 leafsize, u32 sectorsize, root->defrag_running = 0; root->root_key.objectid = objectid; root->anon_dev = 0; + + spin_lock_init(&root->root_times_lock); } static int __must_check find_and_setup_root(struct btrfs_root *tree_root, @@ -1326,6 +1328,7 @@ struct btrfs_root *btrfs_read_fs_root_no_radix(struct btrfs_root *tree_root, u64 generation; u32 blocksize; int ret = 0; + int slot; root = btrfs_alloc_root(fs_info); if (!root) @@ -1352,9 +1355,8 @@ struct btrfs_root *btrfs_read_fs_root_no_radix(struct btrfs_root *tree_root, ret = btrfs_search_slot(NULL, tree_root, location, path, 0, 0); if (ret == 0) { l = path->nodes[0]; - read_extent_buffer(l, &root->root_item, - btrfs_item_ptr_offset(l, path->slots[0]), - sizeof(root->root_item)); + slot = path->slots[0]; + btrfs_read_root_item(tree_root, l, slot, &root->root_item); memcpy(&root->root_key, location, sizeof(*location)); } btrfs_free_path(path); diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index a7d1921ac76b..4ffc87389545 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -2734,6 +2734,8 @@ noinline int btrfs_update_inode(struct btrfs_trans_handle *trans, */ if (!btrfs_is_free_space_inode(root, inode) && root->root_key.objectid != BTRFS_DATA_RELOC_TREE_OBJECTID) { + btrfs_update_root_times(trans, root); + ret = btrfs_delayed_update_inode(trans, root, inode); if (!ret) btrfs_set_inode_last_trans(trans, inode); @@ -4723,6 +4725,8 @@ static struct inode *btrfs_new_inode(struct btrfs_trans_handle *trans, trace_btrfs_inode_new(inode); btrfs_set_inode_last_trans(trans, inode); + btrfs_update_root_times(trans, root); + return inode; fail: if (dir) diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index 7011871c45b8..99fe2ce7f721 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -41,6 +41,7 @@ #include #include #include +#include #include "compat.h" #include "ctree.h" #include "disk-io.h" @@ -346,11 +347,13 @@ static noinline int create_subvol(struct btrfs_root *root, struct btrfs_root *new_root; struct dentry *parent = dentry->d_parent; struct inode *dir; + struct timespec cur_time = CURRENT_TIME; int ret; int err; u64 objectid; u64 new_dirid = BTRFS_FIRST_FREE_OBJECTID; u64 index = 0; + uuid_le new_uuid; ret = btrfs_find_free_objectid(root->fs_info->tree_root, &objectid); if (ret) @@ -389,8 +392,9 @@ static noinline int create_subvol(struct btrfs_root *root, BTRFS_UUID_SIZE); btrfs_mark_buffer_dirty(leaf); + memset(&root_item, 0, sizeof(root_item)); + inode_item = &root_item.inode; - memset(inode_item, 0, sizeof(*inode_item)); inode_item->generation = cpu_to_le64(1); inode_item->size = cpu_to_le64(3); inode_item->nlink = cpu_to_le32(1); @@ -408,8 +412,15 @@ static noinline int create_subvol(struct btrfs_root *root, btrfs_set_root_used(&root_item, leaf->len); btrfs_set_root_last_snapshot(&root_item, 0); - memset(&root_item.drop_progress, 0, sizeof(root_item.drop_progress)); - root_item.drop_level = 0; + btrfs_set_root_generation_v2(&root_item, + btrfs_root_generation(&root_item)); + uuid_le_gen(&new_uuid); + memcpy(root_item.uuid, new_uuid.b, BTRFS_UUID_SIZE); + root_item.otime.sec = cpu_to_le64(cur_time.tv_sec); + root_item.otime.nsec = cpu_to_le64(cur_time.tv_nsec); + root_item.ctime = root_item.otime; + btrfs_set_root_ctransid(&root_item, trans->transid); + btrfs_set_root_otransid(&root_item, trans->transid); btrfs_tree_unlock(leaf); free_extent_buffer(leaf); @@ -3395,6 +3406,87 @@ out: return ret; } +static long btrfs_ioctl_set_received_subvol(struct file *file, + void __user *arg) +{ + struct btrfs_ioctl_received_subvol_args *sa = NULL; + struct inode *inode = fdentry(file)->d_inode; + struct btrfs_root *root = BTRFS_I(inode)->root; + struct btrfs_root_item *root_item = &root->root_item; + struct btrfs_trans_handle *trans; + struct timespec ct = CURRENT_TIME; + int ret = 0; + + ret = mnt_want_write_file(file); + if (ret < 0) + return ret; + + down_write(&root->fs_info->subvol_sem); + + if (btrfs_ino(inode) != BTRFS_FIRST_FREE_OBJECTID) { + ret = -EINVAL; + goto out; + } + + if (btrfs_root_readonly(root)) { + ret = -EROFS; + goto out; + } + + if (!inode_owner_or_capable(inode)) { + ret = -EACCES; + goto out; + } + + sa = memdup_user(arg, sizeof(*sa)); + if (IS_ERR(sa)) { + ret = PTR_ERR(sa); + sa = NULL; + goto out; + } + + trans = btrfs_start_transaction(root, 1); + if (IS_ERR(trans)) { + ret = PTR_ERR(trans); + trans = NULL; + goto out; + } + + sa->rtransid = trans->transid; + sa->rtime.sec = ct.tv_sec; + sa->rtime.nsec = ct.tv_nsec; + + memcpy(root_item->received_uuid, sa->uuid, BTRFS_UUID_SIZE); + btrfs_set_root_stransid(root_item, sa->stransid); + btrfs_set_root_rtransid(root_item, sa->rtransid); + root_item->stime.sec = cpu_to_le64(sa->stime.sec); + root_item->stime.nsec = cpu_to_le32(sa->stime.nsec); + root_item->rtime.sec = cpu_to_le64(sa->rtime.sec); + root_item->rtime.nsec = cpu_to_le32(sa->rtime.nsec); + + ret = btrfs_update_root(trans, root->fs_info->tree_root, + &root->root_key, &root->root_item); + if (ret < 0) { + btrfs_end_transaction(trans, root); + trans = NULL; + goto out; + } else { + ret = btrfs_commit_transaction(trans, root); + if (ret < 0) + goto out; + } + + ret = copy_to_user(arg, sa, sizeof(*sa)); + if (ret) + ret = -EFAULT; + +out: + kfree(sa); + up_write(&root->fs_info->subvol_sem); + mnt_drop_write_file(file); + return ret; +} + long btrfs_ioctl(struct file *file, unsigned int cmd, unsigned long arg) { @@ -3477,6 +3569,8 @@ long btrfs_ioctl(struct file *file, unsigned int return btrfs_ioctl_balance_ctl(root, arg); case BTRFS_IOC_BALANCE_PROGRESS: return btrfs_ioctl_balance_progress(root, argp); + case BTRFS_IOC_SET_RECEIVED_SUBVOL: + return btrfs_ioctl_set_received_subvol(file, argp); case BTRFS_IOC_GET_DEV_STATS: return btrfs_ioctl_get_dev_stats(root, argp, 0); case BTRFS_IOC_GET_AND_RESET_DEV_STATS: diff --git a/fs/btrfs/ioctl.h b/fs/btrfs/ioctl.h index e440aa653c30..0c505d7ff8ed 100644 --- a/fs/btrfs/ioctl.h +++ b/fs/btrfs/ioctl.h @@ -295,6 +295,21 @@ struct btrfs_ioctl_get_dev_stats { __u64 unused[128 - 2 - BTRFS_DEV_STAT_VALUES_MAX]; /* pad to 1k */ }; +struct btrfs_ioctl_timespec { + __u64 sec; + __u32 nsec; +}; + +struct btrfs_ioctl_received_subvol_args { + char uuid[BTRFS_UUID_SIZE]; /* in */ + __u64 stransid; /* in */ + __u64 rtransid; /* out */ + struct btrfs_ioctl_timespec stime; /* in */ + struct btrfs_ioctl_timespec rtime; /* out */ + __u64 flags; /* in */ + __u64 reserved[16]; /* in */ +}; + #define BTRFS_IOC_SNAP_CREATE _IOW(BTRFS_IOCTL_MAGIC, 1, \ struct btrfs_ioctl_vol_args) #define BTRFS_IOC_DEFRAG _IOW(BTRFS_IOCTL_MAGIC, 2, \ @@ -359,6 +374,8 @@ struct btrfs_ioctl_get_dev_stats { struct btrfs_ioctl_ino_path_args) #define BTRFS_IOC_LOGICAL_INO _IOWR(BTRFS_IOCTL_MAGIC, 36, \ struct btrfs_ioctl_ino_path_args) +#define BTRFS_IOC_SET_RECEIVED_SUBVOL _IOWR(BTRFS_IOCTL_MAGIC, 37, \ + struct btrfs_ioctl_received_subvol_args) #define BTRFS_IOC_GET_DEV_STATS _IOWR(BTRFS_IOCTL_MAGIC, 52, \ struct btrfs_ioctl_get_dev_stats) #define BTRFS_IOC_GET_AND_RESET_DEV_STATS _IOWR(BTRFS_IOCTL_MAGIC, 53, \ diff --git a/fs/btrfs/root-tree.c b/fs/btrfs/root-tree.c index 24fb8ce4e071..6bb465cca20f 100644 --- a/fs/btrfs/root-tree.c +++ b/fs/btrfs/root-tree.c @@ -16,11 +16,54 @@ * Boston, MA 021110-1307, USA. */ +#include #include "ctree.h" #include "transaction.h" #include "disk-io.h" #include "print-tree.h" +/* + * Read a root item from the tree. In case we detect a root item smaller then + * sizeof(root_item), we know it's an old version of the root structure and + * initialize all new fields to zero. The same happens if we detect mismatching + * generation numbers as then we know the root was once mounted with an older + * kernel that was not aware of the root item structure change. + */ +void btrfs_read_root_item(struct btrfs_root *root, + struct extent_buffer *eb, int slot, + struct btrfs_root_item *item) +{ + uuid_le uuid; + int len; + int need_reset = 0; + + len = btrfs_item_size_nr(eb, slot); + read_extent_buffer(eb, item, btrfs_item_ptr_offset(eb, slot), + min_t(int, len, (int)sizeof(*item))); + if (len < sizeof(*item)) + need_reset = 1; + if (!need_reset && btrfs_root_generation(item) + != btrfs_root_generation_v2(item)) { + if (btrfs_root_generation_v2(item) != 0) { + printk(KERN_WARNING "btrfs: mismatching " + "generation and generation_v2 " + "found in root item. This root " + "was probably mounted with an " + "older kernel. Resetting all " + "new fields.\n"); + } + need_reset = 1; + } + if (need_reset) { + memset(&item->generation_v2, 0, + sizeof(*item) - offsetof(struct btrfs_root_item, + generation_v2)); + + uuid_le_gen(&uuid); + memcpy(item->uuid, uuid.b, BTRFS_UUID_SIZE); + } +} + /* * lookup the root with the highest offset for a given objectid. The key we do * find is copied into 'key'. If we find something return 0, otherwise 1, < 0 @@ -61,10 +104,10 @@ int btrfs_find_last_root(struct btrfs_root *root, u64 objectid, goto out; } if (item) - read_extent_buffer(l, item, btrfs_item_ptr_offset(l, slot), - sizeof(*item)); + btrfs_read_root_item(root, l, slot, item); if (key) memcpy(key, &found_key, sizeof(found_key)); + ret = 0; out: btrfs_free_path(path); @@ -91,16 +134,15 @@ int btrfs_update_root(struct btrfs_trans_handle *trans, struct btrfs_root int ret; int slot; unsigned long ptr; + int old_len; path = btrfs_alloc_path(); if (!path) return -ENOMEM; ret = btrfs_search_slot(trans, root, key, path, 0, 1); - if (ret < 0) { - btrfs_abort_transaction(trans, root, ret); - goto out; - } + if (ret < 0) + goto out_abort; if (ret != 0) { btrfs_print_leaf(root, path->nodes[0]); @@ -113,16 +155,56 @@ int btrfs_update_root(struct btrfs_trans_handle *trans, struct btrfs_root l = path->nodes[0]; slot = path->slots[0]; ptr = btrfs_item_ptr_offset(l, slot); + old_len = btrfs_item_size_nr(l, slot); + + /* + * If this is the first time we update the root item which originated + * from an older kernel, we need to enlarge the item size to make room + * for the added fields. + */ + if (old_len < sizeof(*item)) { + btrfs_release_path(path); + ret = btrfs_search_slot(trans, root, key, path, + -1, 1); + if (ret < 0) + goto out_abort; + ret = btrfs_del_item(trans, root, path); + if (ret < 0) + goto out_abort; + btrfs_release_path(path); + ret = btrfs_insert_empty_item(trans, root, path, + key, sizeof(*item)); + if (ret < 0) + goto out_abort; + l = path->nodes[0]; + slot = path->slots[0]; + ptr = btrfs_item_ptr_offset(l, slot); + } + + /* + * Update generation_v2 so at the next mount we know the new root + * fields are valid. + */ + btrfs_set_root_generation_v2(item, btrfs_root_generation(item)); + write_extent_buffer(l, item, ptr, sizeof(*item)); btrfs_mark_buffer_dirty(path->nodes[0]); out: btrfs_free_path(path); return ret; + +out_abort: + btrfs_abort_transaction(trans, root, ret); + goto out; } int btrfs_insert_root(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct btrfs_key *key, struct btrfs_root_item *item) { + /* + * Make sure generation v1 and v2 match. See update_root for details. + */ + btrfs_set_root_generation_v2(item, btrfs_root_generation(item)); return btrfs_insert_item(trans, root, key, item, sizeof(*item)); } @@ -454,3 +536,16 @@ void btrfs_check_and_init_root_item(struct btrfs_root_item *root_item) root_item->byte_limit = 0; } } + +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_TIME; + + spin_lock(&root->root_times_lock); + item->ctransid = trans->transid; + item->ctime.sec = cpu_to_le64(ct.tv_sec); + item->ctime.nsec = cpu_to_le64(ct.tv_nsec); + spin_unlock(&root->root_times_lock); +} diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index b72b068183ec..a21f3085a334 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -22,6 +22,7 @@ #include #include #include +#include #include "ctree.h" #include "disk-io.h" #include "transaction.h" @@ -926,11 +927,13 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans, struct dentry *dentry; struct extent_buffer *tmp; struct extent_buffer *old; + struct timespec cur_time = CURRENT_TIME; int ret; u64 to_reserve = 0; u64 index = 0; u64 objectid; u64 root_flags; + uuid_le new_uuid; rsv = trans->block_rsv; @@ -1016,6 +1019,20 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans, root_flags &= ~BTRFS_ROOT_SUBVOL_RDONLY; btrfs_set_root_flags(new_root_item, root_flags); + btrfs_set_root_generation_v2(new_root_item, + trans->transid); + uuid_le_gen(&new_uuid); + memcpy(new_root_item->uuid, new_uuid.b, BTRFS_UUID_SIZE); + memcpy(new_root_item->parent_uuid, root->root_item.uuid, + BTRFS_UUID_SIZE); + new_root_item->otime.sec = cpu_to_le64(cur_time.tv_sec); + new_root_item->otime.nsec = cpu_to_le64(cur_time.tv_nsec); + btrfs_set_root_otransid(new_root_item, trans->transid); + memset(&new_root_item->stime, 0, sizeof(new_root_item->stime)); + memset(&new_root_item->rtime, 0, sizeof(new_root_item->rtime)); + btrfs_set_root_stransid(new_root_item, 0); + btrfs_set_root_rtransid(new_root_item, 0); + old = btrfs_lock_root_node(root); ret = btrfs_cow_block(trans, root, old, NULL, 0, &old); if (ret) { -- cgit v1.2.3 From 533574c6bc30cf526cc1c41bde050c854a945efb Mon Sep 17 00:00:00 2001 From: Joe Perches Date: Mon, 30 Jul 2012 14:40:13 -0700 Subject: btrfs: use printk_get_level and printk_skip_level, add __printf, fix fallout Use the generic printk_get_level() to search a message for a kern_level. Add __printf to verify format and arguments. Fix a few messages that had mismatches in format and arguments. Add #ifdef CONFIG_PRINTK blocks to shrink the object size a bit when not using printk. [akpm@linux-foundation.org: whitespace tweak] Signed-off-by: Joe Perches Cc: Kay Sievers Cc: Chris Mason Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/btrfs/ctree.h | 13 +++++++++++++ fs/btrfs/disk-io.c | 2 +- fs/btrfs/relocation.c | 2 +- fs/btrfs/super.c | 41 +++++++++++++++++++++++++++++++++++------ 4 files changed, 50 insertions(+), 8 deletions(-) (limited to 'fs/btrfs/disk-io.c') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index adb1cd7ceb9b..4bab807227ad 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -3342,10 +3342,22 @@ ssize_t btrfs_listxattr(struct dentry *dentry, char *buffer, size_t size); /* super.c */ int btrfs_parse_options(struct btrfs_root *root, char *options); int btrfs_sync_fs(struct super_block *sb, int wait); + +#ifdef CONFIG_PRINTK +__printf(2, 3) void btrfs_printk(struct btrfs_fs_info *fs_info, const char *fmt, ...); +#else +static inline __printf(2, 3) +void btrfs_printk(struct btrfs_fs_info *fs_info, const char *fmt, ...) +{ +} +#endif + +__printf(5, 6) void __btrfs_std_error(struct btrfs_fs_info *fs_info, const char *function, unsigned int line, int errno, const char *fmt, ...); + void __btrfs_abort_transaction(struct btrfs_trans_handle *trans, struct btrfs_root *root, const char *function, unsigned int line, int errno); @@ -3386,6 +3398,7 @@ do { \ (errno), fmt, ##args); \ } while (0) +__printf(5, 6) void __btrfs_panic(struct btrfs_fs_info *fs_info, const char *function, unsigned int line, int errno, const char *fmt, ...); diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 502b20c56e84..fadeba6a5db9 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -1114,7 +1114,7 @@ void clean_tree_block(struct btrfs_trans_handle *trans, struct btrfs_root *root, spin_unlock(&root->fs_info->delalloc_lock); btrfs_panic(root->fs_info, -EOVERFLOW, "Can't clear %lu bytes from " - " dirty_mdatadata_bytes (%lu)", + " dirty_mdatadata_bytes (%llu)", buf->len, root->fs_info->dirty_metadata_bytes); } diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c index c5dbd9149679..4da08652004d 100644 --- a/fs/btrfs/relocation.c +++ b/fs/btrfs/relocation.c @@ -1241,7 +1241,7 @@ static int __must_check __add_reloc_root(struct btrfs_root *root) if (rb_node) { btrfs_panic(root->fs_info, -EEXIST, "Duplicate root found " "for start=%llu while inserting into relocation " - "tree\n"); + "tree\n", node->bytenr); kfree(node); return -EEXIST; } diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index fa61ef59cd61..8c6e61d6eed5 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -125,6 +125,7 @@ static void btrfs_handle_error(struct btrfs_fs_info *fs_info) } } +#ifdef CONFIG_PRINTK /* * __btrfs_std_error decodes expected errors from the caller and * invokes the approciate error response. @@ -167,7 +168,7 @@ void __btrfs_std_error(struct btrfs_fs_info *fs_info, const char *function, va_end(args); } -const char *logtypes[] = { +static const char * const logtypes[] = { "emergency", "alert", "critical", @@ -185,22 +186,50 @@ void btrfs_printk(struct btrfs_fs_info *fs_info, const char *fmt, ...) struct va_format vaf; va_list args; const char *type = logtypes[4]; + int kern_level; va_start(args, fmt); - if (fmt[0] == '<' && isdigit(fmt[1]) && fmt[2] == '>') { - memcpy(lvl, fmt, 3); - lvl[3] = '\0'; - fmt += 3; - type = logtypes[fmt[1] - '0']; + kern_level = printk_get_level(fmt); + if (kern_level) { + size_t size = printk_skip_level(fmt) - fmt; + memcpy(lvl, fmt, size); + lvl[size] = '\0'; + fmt += size; + type = logtypes[kern_level - '0']; } else *lvl = '\0'; vaf.fmt = fmt; vaf.va = &args; + printk("%sBTRFS %s (device %s): %pV", lvl, type, sb->s_id, &vaf); + + va_end(args); } +#else + +void __btrfs_std_error(struct btrfs_fs_info *fs_info, const char *function, + unsigned int line, int errno, const char *fmt, ...) +{ + struct super_block *sb = fs_info->sb; + + /* + * Special case: if the error is EROFS, and we're already + * under MS_RDONLY, then it is safe here. + */ + if (errno == -EROFS && (sb->s_flags & MS_RDONLY)) + return; + + /* Don't go through full error handling during mount */ + if (sb->s_flags & MS_BORN) { + save_error_info(fs_info); + btrfs_handle_error(fs_info); + } +} +#endif + /* * We only mark the transaction aborted and then set the file system read-only. * This will prevent new transactions from starting or trying to join this -- cgit v1.2.3 From b2b5ef5c8e89f19b68c174bf246f3ca212dbf0bc Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Tue, 12 Jun 2012 16:20:45 +0200 Subject: btrfs: Convert to new freezing mechanism We convert btrfs_file_aio_write() to use new freeze check. We also add proper freeze protection to btrfs_page_mkwrite(). We also add freeze protection to the transaction mechanism to avoid starting transactions on frozen filesystem. At minimum this is necessary to stop iput() of unlinked file to change frozen filesystem during truncation. Checks in cleaner_kthread() and transaction_kthread() can be safely removed since btrfs_freeze() will lock the mutexes and thus block the threads (and they shouldn't have anything to do anyway). CC: linux-btrfs@vger.kernel.org CC: Chris Mason Signed-off-by: Jan Kara Signed-off-by: Al Viro --- fs/btrfs/disk-io.c | 3 --- fs/btrfs/file.c | 3 ++- fs/btrfs/inode.c | 6 +++++- fs/btrfs/transaction.c | 7 +++++++ 4 files changed, 14 insertions(+), 5 deletions(-) (limited to 'fs/btrfs/disk-io.c') diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 2936ca49b3b4..20c49b16b759 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -1533,8 +1533,6 @@ static int cleaner_kthread(void *arg) struct btrfs_root *root = arg; do { - vfs_check_frozen(root->fs_info->sb, SB_FREEZE_WRITE); - if (!(root->fs_info->sb->s_flags & MS_RDONLY) && mutex_trylock(&root->fs_info->cleaner_mutex)) { btrfs_run_delayed_iputs(root); @@ -1566,7 +1564,6 @@ static int transaction_kthread(void *arg) do { cannot_commit = false; delay = HZ * 30; - vfs_check_frozen(root->fs_info->sb, SB_FREEZE_WRITE); mutex_lock(&root->fs_info->transaction_kthread_mutex); spin_lock(&root->fs_info->trans_lock); diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index 9aa01ec2138d..5caf285c6e4d 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -1379,7 +1379,7 @@ static ssize_t btrfs_file_aio_write(struct kiocb *iocb, ssize_t err = 0; size_t count, ocount; - vfs_check_frozen(inode->i_sb, SB_FREEZE_WRITE); + sb_start_write(inode->i_sb); mutex_lock(&inode->i_mutex); @@ -1469,6 +1469,7 @@ static ssize_t btrfs_file_aio_write(struct kiocb *iocb, num_written = err; } out: + sb_end_write(inode->i_sb); current->backing_dev_info = NULL; return num_written ? num_written : err; } diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index fb8d671d00e6..f4d901786b3c 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -6620,6 +6620,7 @@ int btrfs_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) u64 page_start; u64 page_end; + sb_start_pagefault(inode->i_sb); ret = btrfs_delalloc_reserve_space(inode, PAGE_CACHE_SIZE); if (!ret) { ret = file_update_time(vma->vm_file); @@ -6709,12 +6710,15 @@ again: unlock_extent_cached(io_tree, page_start, page_end, &cached_state, GFP_NOFS); out_unlock: - if (!ret) + if (!ret) { + sb_end_pagefault(inode->i_sb); return VM_FAULT_LOCKED; + } unlock_page(page); out: btrfs_delalloc_release_space(inode, PAGE_CACHE_SIZE); out_noreserve: + sb_end_pagefault(inode->i_sb); return ret; } diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index b72b068183ec..fa67ba51516e 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -329,6 +329,8 @@ again: if (!h) return ERR_PTR(-ENOMEM); + sb_start_intwrite(root->fs_info->sb); + if (may_wait_transaction(root, type)) wait_current_trans(root); @@ -339,6 +341,7 @@ again: } while (ret == -EBUSY); if (ret < 0) { + sb_end_intwrite(root->fs_info->sb); kmem_cache_free(btrfs_trans_handle_cachep, h); return ERR_PTR(ret); } @@ -528,6 +531,8 @@ static int __btrfs_end_transaction(struct btrfs_trans_handle *trans, count++; } + sb_end_intwrite(root->fs_info->sb); + if (lock && !atomic_read(&root->fs_info->open_ioctl_trans) && should_end_transaction(trans, root)) { trans->transaction->blocked = 1; @@ -1517,6 +1522,8 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans, put_transaction(cur_trans); put_transaction(cur_trans); + sb_end_intwrite(root->fs_info->sb); + trace_btrfs_transaction_commit(root); btrfs_scrub_continue(root); -- cgit v1.2.3