diff options
Diffstat (limited to 'fs/btrfs/ordered-data.c')
| -rw-r--r-- | fs/btrfs/ordered-data.c | 332 |
1 files changed, 225 insertions, 107 deletions
diff --git a/fs/btrfs/ordered-data.c b/fs/btrfs/ordered-data.c index b749ba45da2b..9212ce110cde 100644 --- a/fs/btrfs/ordered-data.c +++ b/fs/btrfs/ordered-data.c @@ -19,6 +19,7 @@ #include "qgroup.h" #include "subpage.h" #include "file.h" +#include "block-group.h" static struct kmem_cache *btrfs_ordered_extent_cache; @@ -110,8 +111,8 @@ static struct rb_node *__tree_search(struct rb_root *root, u64 file_offset, return NULL; } -static int range_overlaps(struct btrfs_ordered_extent *entry, u64 file_offset, - u64 len) +static int btrfs_range_overlaps(struct btrfs_ordered_extent *entry, u64 file_offset, + u64 len) { if (file_offset + len <= entry->file_offset || entry->file_offset + entry->num_bytes <= file_offset) @@ -152,25 +153,30 @@ static struct btrfs_ordered_extent *alloc_ordered_extent( struct btrfs_ordered_extent *entry; int ret; u64 qgroup_rsv = 0; + const bool is_nocow = (flags & + ((1U << BTRFS_ORDERED_NOCOW) | (1U << BTRFS_ORDERED_PREALLOC))); - if (flags & - ((1 << BTRFS_ORDERED_NOCOW) | (1 << BTRFS_ORDERED_PREALLOC))) { - /* For nocow write, we can release the qgroup rsv right now */ + /* + * For a NOCOW write we can free the qgroup reserve right now. For a COW + * one we transfer the reserved space from the inode's iotree into the + * ordered extent by calling btrfs_qgroup_release_data() and tracking + * the qgroup reserved amount in the ordered extent, so that later after + * completing the ordered extent, when running the data delayed ref it + * creates, we free the reserved data with btrfs_qgroup_free_refroot(). + */ + if (is_nocow) ret = btrfs_qgroup_free_data(inode, NULL, file_offset, num_bytes, &qgroup_rsv); - if (ret < 0) - return ERR_PTR(ret); - } else { - /* - * The ordered extent has reserved qgroup space, release now - * and pass the reserved number for qgroup_record to free. - */ + else ret = btrfs_qgroup_release_data(inode, file_offset, num_bytes, &qgroup_rsv); - if (ret < 0) - return ERR_PTR(ret); - } + + if (ret < 0) + return ERR_PTR(ret); + entry = kmem_cache_zalloc(btrfs_ordered_extent_cache, GFP_NOFS); - if (!entry) - return ERR_PTR(-ENOMEM); + if (!entry) { + entry = ERR_PTR(-ENOMEM); + goto out; + } entry->file_offset = file_offset; entry->num_bytes = num_bytes; @@ -179,7 +185,12 @@ static struct btrfs_ordered_extent *alloc_ordered_extent( entry->disk_num_bytes = disk_num_bytes; entry->offset = offset; entry->bytes_left = num_bytes; - entry->inode = igrab(&inode->vfs_inode); + if (WARN_ON_ONCE(!igrab(&inode->vfs_inode))) { + kmem_cache_free(btrfs_ordered_extent_cache, entry); + entry = ERR_PTR(-ESTALE); + goto out; + } + entry->inode = inode; entry->compress_type = compress_type; entry->truncated_len = (u64)-1; entry->qgroup_rsv = qgroup_rsv; @@ -202,12 +213,18 @@ static struct btrfs_ordered_extent *alloc_ordered_extent( btrfs_mod_outstanding_extents(inode, 1); spin_unlock(&inode->lock); +out: + if (IS_ERR(entry) && !is_nocow) + btrfs_qgroup_free_refroot(inode->root->fs_info, + btrfs_root_id(inode->root), + qgroup_rsv, BTRFS_QGROUP_RSV_DATA); + return entry; } static void insert_ordered_extent(struct btrfs_ordered_extent *entry) { - struct btrfs_inode *inode = BTRFS_I(entry->inode); + struct btrfs_inode *inode = entry->inode; struct btrfs_root *root = inode->root; struct btrfs_fs_info *fs_info = root->fs_info; struct rb_node *node; @@ -223,7 +240,7 @@ static void insert_ordered_extent(struct btrfs_ordered_extent *entry) spin_lock_irq(&inode->ordered_tree_lock); node = tree_insert(&inode->ordered_tree, entry->file_offset, &entry->rb_node); - if (node) + if (unlikely(node)) btrfs_panic(fs_info, -EEXIST, "inconsistency in ordered tree at offset %llu", entry->file_offset); @@ -252,7 +269,7 @@ static void insert_ordered_extent(struct btrfs_ordered_extent *entry) * @disk_bytenr: Offset of extent on disk. * @disk_num_bytes: Size of extent on disk. * @offset: Offset into unencoded data where file data starts. - * @flags: Flags specifying type of extent (1 << BTRFS_ORDERED_*). + * @flags: Flags specifying type of extent (1U << BTRFS_ORDERED_*). * @compress_type: Compression algorithm used for data. * * Most of these parameters correspond to &struct btrfs_file_extent_item. The @@ -263,17 +280,39 @@ static void insert_ordered_extent(struct btrfs_ordered_extent *entry) */ struct btrfs_ordered_extent *btrfs_alloc_ordered_extent( struct btrfs_inode *inode, u64 file_offset, - u64 num_bytes, u64 ram_bytes, u64 disk_bytenr, - u64 disk_num_bytes, u64 offset, unsigned long flags, - int compress_type) + const struct btrfs_file_extent *file_extent, unsigned long flags) { struct btrfs_ordered_extent *entry; ASSERT((flags & ~BTRFS_ORDERED_TYPE_FLAGS) == 0); - entry = alloc_ordered_extent(inode, file_offset, num_bytes, ram_bytes, - disk_bytenr, disk_num_bytes, offset, flags, - compress_type); + /* + * For regular writes, we just use the members in @file_extent. + * + * For NOCOW, we don't really care about the numbers except @start and + * file_extent->num_bytes, as we won't insert a file extent item at all. + * + * For PREALLOC, we do not use ordered extent members, but + * btrfs_mark_extent_written() handles everything. + * + * So here we always pass 0 as offset for NOCOW/PREALLOC ordered extents, + * or btrfs_split_ordered_extent() cannot handle it correctly. + */ + if (flags & ((1U << BTRFS_ORDERED_NOCOW) | (1U << BTRFS_ORDERED_PREALLOC))) + entry = alloc_ordered_extent(inode, file_offset, + file_extent->num_bytes, + file_extent->num_bytes, + file_extent->disk_bytenr + file_extent->offset, + file_extent->num_bytes, 0, flags, + file_extent->compression); + else + entry = alloc_ordered_extent(inode, file_offset, + file_extent->num_bytes, + file_extent->ram_bytes, + file_extent->disk_bytenr, + file_extent->disk_num_bytes, + file_extent->offset, flags, + file_extent->compression); if (!IS_ERR(entry)) insert_ordered_extent(entry); return entry; @@ -287,13 +326,19 @@ struct btrfs_ordered_extent *btrfs_alloc_ordered_extent( void btrfs_add_ordered_sum(struct btrfs_ordered_extent *entry, struct btrfs_ordered_sum *sum) { - struct btrfs_inode *inode = BTRFS_I(entry->inode); + struct btrfs_inode *inode = entry->inode; spin_lock_irq(&inode->ordered_tree_lock); list_add_tail(&sum->list, &entry->list); spin_unlock_irq(&inode->ordered_tree_lock); } +void btrfs_mark_ordered_extent_error(struct btrfs_ordered_extent *ordered) +{ + if (!test_and_set_bit(BTRFS_ORDERED_IOERR, &ordered->flags)) + mapping_set_error(ordered->inode->vfs_inode.i_mapping, -EIO); +} + static void finish_ordered_fn(struct btrfs_work *work) { struct btrfs_ordered_extent *ordered_extent; @@ -303,36 +348,35 @@ static void finish_ordered_fn(struct btrfs_work *work) } static bool can_finish_ordered_extent(struct btrfs_ordered_extent *ordered, - struct page *page, u64 file_offset, + struct folio *folio, u64 file_offset, u64 len, bool uptodate) { - struct btrfs_inode *inode = BTRFS_I(ordered->inode); + struct btrfs_inode *inode = ordered->inode; struct btrfs_fs_info *fs_info = inode->root->fs_info; lockdep_assert_held(&inode->ordered_tree_lock); - if (page) { - ASSERT(page->mapping); - ASSERT(page_offset(page) <= file_offset); - ASSERT(file_offset + len <= page_offset(page) + PAGE_SIZE); + if (folio) { + ASSERT(folio->mapping); + ASSERT(folio_pos(folio) <= file_offset); + ASSERT(file_offset + len <= folio_pos(folio) + folio_size(folio)); /* - * Ordered (Private2) bit indicates whether we still have + * Ordered flag indicates whether we still have * pending io unfinished for the ordered extent. * - * If there's no such bit, we need to skip to next range. + * If it's not set, we need to skip to next range. */ - if (!btrfs_folio_test_ordered(fs_info, page_folio(page), - file_offset, len)) + if (!btrfs_folio_test_ordered(fs_info, folio, file_offset, len)) return false; - btrfs_folio_clear_ordered(fs_info, page_folio(page), file_offset, len); + btrfs_folio_clear_ordered(fs_info, folio, file_offset, len); } /* Now we're fine to update the accounting. */ if (WARN_ON_ONCE(len > ordered->bytes_left)) { btrfs_crit(fs_info, "bad ordered extent accounting, root=%llu ino=%llu OE offset=%llu OE len=%llu to_dec=%llu left=%llu", - inode->root->root_key.objectid, btrfs_ino(inode), + btrfs_root_id(inode->root), btrfs_ino(inode), ordered->file_offset, ordered->num_bytes, len, ordered->bytes_left); ordered->bytes_left = 0; @@ -359,7 +403,7 @@ static bool can_finish_ordered_extent(struct btrfs_ordered_extent *ordered, static void btrfs_queue_ordered_fn(struct btrfs_ordered_extent *ordered) { - struct btrfs_inode *inode = BTRFS_I(ordered->inode); + struct btrfs_inode *inode = ordered->inode; struct btrfs_fs_info *fs_info = inode->root->fs_info; struct btrfs_workqueue *wq = btrfs_is_free_space_inode(inode) ? fs_info->endio_freespace_worker : fs_info->endio_write_workers; @@ -368,30 +412,61 @@ static void btrfs_queue_ordered_fn(struct btrfs_ordered_extent *ordered) btrfs_queue_work(wq, &ordered->work); } -bool btrfs_finish_ordered_extent(struct btrfs_ordered_extent *ordered, - struct page *page, u64 file_offset, u64 len, +void btrfs_finish_ordered_extent(struct btrfs_ordered_extent *ordered, + struct folio *folio, u64 file_offset, u64 len, bool uptodate) { - struct btrfs_inode *inode = BTRFS_I(ordered->inode); + struct btrfs_inode *inode = ordered->inode; unsigned long flags; bool ret; trace_btrfs_finish_ordered_extent(inode, file_offset, len, uptodate); spin_lock_irqsave(&inode->ordered_tree_lock, flags); - ret = can_finish_ordered_extent(ordered, page, file_offset, len, uptodate); + ret = can_finish_ordered_extent(ordered, folio, file_offset, len, + uptodate); spin_unlock_irqrestore(&inode->ordered_tree_lock, flags); + /* + * If this is a COW write it means we created new extent maps for the + * range and they point to unwritten locations if we got an error either + * before submitting a bio or during IO. + * + * We have marked the ordered extent with BTRFS_ORDERED_IOERR, and we + * are queuing its completion below. During completion, at + * btrfs_finish_one_ordered(), we will drop the extent maps for the + * unwritten extents. + * + * However because completion runs in a work queue we can end up having + * a fast fsync running before that. In the case of direct IO, once we + * unlock the inode the fsync might start, and we queue the completion + * before unlocking the inode. In the case of buffered IO when writeback + * finishes (end_bbio_data_write()) we queue the completion, so if the + * writeback was triggered by a fast fsync, the fsync might start + * logging before ordered extent completion runs in the work queue. + * + * The fast fsync will log file extent items based on the extent maps it + * finds, so if by the time it collects extent maps the ordered extent + * completion didn't happen yet, it will log file extent items that + * point to unwritten extents, resulting in a corruption if a crash + * happens and the log tree is replayed. Note that a fast fsync does not + * wait for completion of ordered extents in order to reduce latency. + * + * Set a flag in the inode so that the next fast fsync will wait for + * ordered extents to complete before starting to log. + */ + if (!uptodate && !test_bit(BTRFS_ORDERED_NOCOW, &ordered->flags)) + set_bit(BTRFS_INODE_COW_WRITE_ERROR, &inode->runtime_flags); + if (ret) btrfs_queue_ordered_fn(ordered); - return ret; } /* * Mark all ordered extents io inside the specified range finished. * - * @page: The involved page for the operation. - * For uncompressed buffered IO, the page status also needs to be + * @folio: The involved folio for the operation. + * For uncompressed buffered IO, the folio status also needs to be * updated to indicate whether the pending ordered io is finished. * Can be NULL for direct IO and compressed write. * For these cases, callers are ensured they won't execute the @@ -401,7 +476,7 @@ bool btrfs_finish_ordered_extent(struct btrfs_ordered_extent *ordered, * extent(s) covering it. */ void btrfs_mark_ordered_io_finished(struct btrfs_inode *inode, - struct page *page, u64 file_offset, + struct folio *folio, u64 file_offset, u64 num_bytes, bool uptodate) { struct rb_node *node; @@ -465,7 +540,7 @@ void btrfs_mark_ordered_io_finished(struct btrfs_inode *inode, ASSERT(end + 1 - cur < U32_MAX); len = end + 1 - cur; - if (can_finish_ordered_extent(entry, page, cur, len, uptodate)) { + if (can_finish_ordered_extent(entry, folio, cur, len, uptodate)) { spin_unlock_irqrestore(&inode->ordered_tree_lock, flags); btrfs_queue_ordered_fn(entry); spin_lock_irqsave(&inode->ordered_tree_lock, flags); @@ -548,23 +623,18 @@ out: */ void btrfs_put_ordered_extent(struct btrfs_ordered_extent *entry) { - struct list_head *cur; - struct btrfs_ordered_sum *sum; - - trace_btrfs_ordered_extent_put(BTRFS_I(entry->inode), entry); + trace_btrfs_ordered_extent_put(entry->inode, entry); if (refcount_dec_and_test(&entry->refs)) { + struct btrfs_ordered_sum *sum; + struct btrfs_ordered_sum *tmp; + ASSERT(list_empty(&entry->root_extent_list)); ASSERT(list_empty(&entry->log_list)); ASSERT(RB_EMPTY_NODE(&entry->rb_node)); - if (entry->inode) - btrfs_add_delayed_iput(BTRFS_I(entry->inode)); - while (!list_empty(&entry->list)) { - cur = entry->list.next; - sum = list_entry(cur, struct btrfs_ordered_sum, list); - list_del(&sum->list); + btrfs_add_delayed_iput(entry->inode); + list_for_each_entry_safe(sum, tmp, &entry->list, list) kvfree(sum); - } kmem_cache_free(btrfs_ordered_extent_cache, entry); } } @@ -589,7 +659,7 @@ void btrfs_remove_ordered_extent(struct btrfs_inode *btrfs_inode, freespace_inode = btrfs_is_free_space_inode(btrfs_inode); btrfs_lockdep_acquire(fs_info, btrfs_trans_pending_ordered); - /* This is paired with btrfs_alloc_ordered_extent. */ + /* This is paired with alloc_ordered_extent(). */ spin_lock(&btrfs_inode->lock); btrfs_mod_outstanding_extents(btrfs_inode, -1); spin_unlock(&btrfs_inode->lock); @@ -675,11 +745,11 @@ static void btrfs_run_ordered_extent_work(struct btrfs_work *work) } /* - * wait for all the ordered extents in a root. This is done when balancing - * space between drives. + * Wait for all the ordered extents in a root. Use @bg as range or do whole + * range if it's NULL. */ u64 btrfs_wait_ordered_extents(struct btrfs_root *root, u64 nr, - const u64 range_start, const u64 range_len) + const struct btrfs_block_group *bg) { struct btrfs_fs_info *fs_info = root->fs_info; LIST_HEAD(splice); @@ -687,7 +757,17 @@ u64 btrfs_wait_ordered_extents(struct btrfs_root *root, u64 nr, LIST_HEAD(works); struct btrfs_ordered_extent *ordered, *next; u64 count = 0; - const u64 range_end = range_start + range_len; + u64 range_start, range_len; + u64 range_end; + + if (bg) { + range_start = bg->start; + range_len = bg->length; + } else { + range_start = 0; + range_len = U64_MAX; + } + range_end = range_start + range_len; mutex_lock(&root->ordered_extent_mutex); spin_lock(&root->ordered_extent_lock); @@ -714,10 +794,10 @@ u64 btrfs_wait_ordered_extents(struct btrfs_root *root, u64 nr, btrfs_queue_work(fs_info->flush_workers, &ordered->flush_work); cond_resched(); - spin_lock(&root->ordered_extent_lock); if (nr != U64_MAX) nr--; count++; + spin_lock(&root->ordered_extent_lock); } list_splice_tail(&skipped, &root->ordered_extents); list_splice_tail(&splice, &root->ordered_extents); @@ -734,8 +814,12 @@ u64 btrfs_wait_ordered_extents(struct btrfs_root *root, u64 nr, return count; } +/* + * Wait for @nr ordered extents that intersect the @bg, or the whole range of + * the filesystem if @bg is NULL. + */ void btrfs_wait_ordered_roots(struct btrfs_fs_info *fs_info, u64 nr, - const u64 range_start, const u64 range_len) + const struct btrfs_block_group *bg) { struct btrfs_root *root; LIST_HEAD(splice); @@ -753,14 +837,13 @@ void btrfs_wait_ordered_roots(struct btrfs_fs_info *fs_info, u64 nr, &fs_info->ordered_roots); spin_unlock(&fs_info->ordered_root_lock); - done = btrfs_wait_ordered_extents(root, nr, - range_start, range_len); + done = btrfs_wait_ordered_extents(root, nr, bg); btrfs_put_root(root); - spin_lock(&fs_info->ordered_root_lock); - if (nr != U64_MAX) { + if (nr != U64_MAX) nr -= done; - } + + spin_lock(&fs_info->ordered_root_lock); } list_splice_tail(&splice, &fs_info->ordered_roots); spin_unlock(&fs_info->ordered_root_lock); @@ -770,14 +853,16 @@ void btrfs_wait_ordered_roots(struct btrfs_fs_info *fs_info, u64 nr, /* * Start IO and wait for a given ordered extent to finish. * - * Wait on page writeback for all the pages in the extent and the IO completion - * code to insert metadata into the btree corresponding to the extent. + * Wait on page writeback for all the pages in the extent but not in + * [@nowriteback_start, @nowriteback_start + @nowriteback_len) and the + * IO completion code to insert metadata into the btree corresponding to the extent. */ -void btrfs_start_ordered_extent(struct btrfs_ordered_extent *entry) +void btrfs_start_ordered_extent_nowriteback(struct btrfs_ordered_extent *entry, + u64 nowriteback_start, u32 nowriteback_len) { u64 start = entry->file_offset; u64 end = start + entry->num_bytes - 1; - struct btrfs_inode *inode = BTRFS_I(entry->inode); + struct btrfs_inode *inode = entry->inode; bool freespace_inode; trace_btrfs_ordered_extent_start(inode, entry); @@ -793,8 +878,19 @@ void btrfs_start_ordered_extent(struct btrfs_ordered_extent *entry) * start IO on any dirty ones so the wait doesn't stall waiting * for the flusher thread to find them */ - if (!test_bit(BTRFS_ORDERED_DIRECT, &entry->flags)) - filemap_fdatawrite_range(inode->vfs_inode.i_mapping, start, end); + if (!test_bit(BTRFS_ORDERED_DIRECT, &entry->flags)) { + if (!nowriteback_len) { + filemap_fdatawrite_range(inode->vfs_inode.i_mapping, start, end); + } else { + if (start < nowriteback_start) + filemap_fdatawrite_range(inode->vfs_inode.i_mapping, start, + nowriteback_start - 1); + if (nowriteback_start + nowriteback_len < end) + filemap_fdatawrite_range(inode->vfs_inode.i_mapping, + nowriteback_start + nowriteback_len, + end); + } + } if (!freespace_inode) btrfs_might_wait_for_event(inode->root->fs_info, btrfs_ordered_extent); @@ -804,7 +900,7 @@ void btrfs_start_ordered_extent(struct btrfs_ordered_extent *entry) /* * Used to wait on ordered extents across a large range of bytes. */ -int btrfs_wait_ordered_range(struct inode *inode, u64 start, u64 len) +int btrfs_wait_ordered_range(struct btrfs_inode *inode, u64 start, u64 len) { int ret = 0; int ret_wb = 0; @@ -834,11 +930,11 @@ int btrfs_wait_ordered_range(struct inode *inode, u64 start, u64 len) * before the ordered extents complete - to avoid failures (-EEXIST) * when adding the new ordered extents to the ordered tree. */ - ret_wb = filemap_fdatawait_range(inode->i_mapping, start, orig_end); + ret_wb = filemap_fdatawait_range(inode->vfs_inode.i_mapping, start, orig_end); end = orig_end; while (1) { - ordered = btrfs_lookup_first_ordered_extent(BTRFS_I(inode), end); + ordered = btrfs_lookup_first_ordered_extent(inode, end); if (!ordered) break; if (ordered->file_offset > orig_end) { @@ -913,7 +1009,7 @@ struct btrfs_ordered_extent *btrfs_lookup_ordered_range( while (1) { entry = rb_entry(node, struct btrfs_ordered_extent, rb_node); - if (range_overlaps(entry, file_offset, len)) + if (btrfs_range_overlaps(entry, file_offset, len)) break; if (entry->file_offset >= file_offset + len) { @@ -943,7 +1039,7 @@ void btrfs_get_ordered_extents_for_logging(struct btrfs_inode *inode, { struct rb_node *n; - ASSERT(inode_is_locked(&inode->vfs_inode)); + btrfs_assert_inode_locked(inode); spin_lock_irq(&inode->ordered_tree_lock); for (n = rb_first(&inode->ordered_tree); n; n = rb_next(n)) { @@ -1042,12 +1138,12 @@ struct btrfs_ordered_extent *btrfs_lookup_first_ordered_range( } if (prev) { entry = rb_entry(prev, struct btrfs_ordered_extent, rb_node); - if (range_overlaps(entry, file_offset, len)) + if (btrfs_range_overlaps(entry, file_offset, len)) goto out; } if (next) { entry = rb_entry(next, struct btrfs_ordered_extent, rb_node); - if (range_overlaps(entry, file_offset, len)) + if (btrfs_range_overlaps(entry, file_offset, len)) goto out; } /* No ordered extent in the range */ @@ -1088,7 +1184,7 @@ void btrfs_lock_and_flush_ordered_range(struct btrfs_inode *inode, u64 start, cachedp = cached_state; while (1) { - lock_extent(&inode->io_tree, start, end, cachedp); + btrfs_lock_extent(&inode->io_tree, start, end, cachedp); ordered = btrfs_lookup_ordered_range(inode, start, end - start + 1); if (!ordered) { @@ -1101,7 +1197,7 @@ void btrfs_lock_and_flush_ordered_range(struct btrfs_inode *inode, u64 start, refcount_dec(&cache->refs); break; } - unlock_extent(&inode->io_tree, start, end, cachedp); + btrfs_unlock_extent(&inode->io_tree, start, end, cachedp); btrfs_start_ordered_extent(ordered); btrfs_put_ordered_extent(ordered); } @@ -1119,7 +1215,7 @@ bool btrfs_try_lock_ordered_range(struct btrfs_inode *inode, u64 start, u64 end, { struct btrfs_ordered_extent *ordered; - if (!try_lock_extent(&inode->io_tree, start, end, cached_state)) + if (!btrfs_try_lock_extent(&inode->io_tree, start, end, cached_state)) return false; ordered = btrfs_lookup_ordered_range(inode, start, end - start + 1); @@ -1127,7 +1223,7 @@ bool btrfs_try_lock_ordered_range(struct btrfs_inode *inode, u64 start, u64 end, return true; btrfs_put_ordered_extent(ordered); - unlock_extent(&inode->io_tree, start, end, cached_state); + btrfs_unlock_extent(&inode->io_tree, start, end, cached_state); return false; } @@ -1136,7 +1232,7 @@ bool btrfs_try_lock_ordered_range(struct btrfs_inode *inode, u64 start, u64 end, struct btrfs_ordered_extent *btrfs_split_ordered_extent( struct btrfs_ordered_extent *ordered, u64 len) { - struct btrfs_inode *inode = BTRFS_I(ordered->inode); + struct btrfs_inode *inode = ordered->inode; struct btrfs_root *root = inode->root; struct btrfs_fs_info *fs_info = root->fs_info; u64 file_offset = ordered->file_offset; @@ -1157,6 +1253,18 @@ struct btrfs_ordered_extent *btrfs_split_ordered_extent( */ if (WARN_ON_ONCE(len >= ordered->num_bytes)) return ERR_PTR(-EINVAL); + /* + * If our ordered extent had an error there's no point in continuing. + * The error may have come from a transaction abort done either by this + * task or some other concurrent task, and the transaction abort path + * iterates over all existing ordered extents and sets the flag + * BTRFS_ORDERED_IOERR on them. + */ + if (unlikely(flags & (1U << BTRFS_ORDERED_IOERR))) { + const int fs_error = BTRFS_FS_ERROR(fs_info); + + return fs_error ? ERR_PTR(fs_error) : ERR_PTR(-EIO); + } /* We cannot split partially completed ordered extents. */ if (ordered->bytes_left) { ASSERT(!(flags & ~BTRFS_ORDERED_TYPE_FLAGS)); @@ -1175,19 +1283,37 @@ struct btrfs_ordered_extent *btrfs_split_ordered_extent( /* One ref for the tree. */ refcount_inc(&new->refs); + /* + * Take the root's ordered_extent_lock to avoid a race with + * btrfs_wait_ordered_extents() when updating the disk_bytenr and + * disk_num_bytes fields of the ordered extent below. And we disable + * IRQs because the inode's ordered_tree_lock is used in IRQ context + * elsewhere. + * + * There's no concern about a previous caller of + * btrfs_wait_ordered_extents() getting the trimmed ordered extent + * before we insert the new one, because even if it gets the ordered + * extent before it's trimmed and the new one inserted, right before it + * uses it or during its use, the ordered extent might have been + * trimmed in the meanwhile, and it missed the new ordered extent. + * There's no way around this and it's harmless for current use cases, + * so we take the root's ordered_extent_lock to fix that race during + * trimming and silence tools like KCSAN. + */ spin_lock_irq(&root->ordered_extent_lock); spin_lock(&inode->ordered_tree_lock); - /* Remove from tree once */ - node = &ordered->rb_node; - rb_erase(node, &inode->ordered_tree); - RB_CLEAR_NODE(node); - if (inode->ordered_tree_last == node) - inode->ordered_tree_last = NULL; + /* + * We don't have overlapping ordered extents (that would imply double + * allocation of extents) and we checked above that the split length + * does not cross the ordered extent's num_bytes field, so there's + * no need to remove it and re-insert it in the tree. + */ ordered->file_offset += len; ordered->disk_bytenr += len; ordered->num_bytes -= len; ordered->disk_num_bytes -= len; + ordered->ram_bytes -= len; if (test_bit(BTRFS_ORDERED_IO_DONE, &ordered->flags)) { ASSERT(ordered->bytes_left == 0); @@ -1212,18 +1338,10 @@ struct btrfs_ordered_extent *btrfs_split_ordered_extent( offset += sum->len; } - /* Re-insert the node */ - node = tree_insert(&inode->ordered_tree, ordered->file_offset, - &ordered->rb_node); - if (node) - btrfs_panic(fs_info, -EEXIST, - "zoned: inconsistency in ordered tree at offset %llu", - ordered->file_offset); - node = tree_insert(&inode->ordered_tree, new->file_offset, &new->rb_node); - if (node) + if (unlikely(node)) btrfs_panic(fs_info, -EEXIST, - "zoned: inconsistency in ordered tree at offset %llu", + "inconsistency in ordered tree at offset %llu after split", new->file_offset); spin_unlock(&inode->ordered_tree_lock); |