Btrfs: Fix bookend extent race v2

When dropping middle part of an extent, btrfs_drop_extents truncates
the extent at first, then inserts a bookend extent.

Since truncation and insertion can't be done atomically, there is a small
period that the bookend extent isn't in the tree. This causes problem for
functions that search the tree for file extent item. The way to fix this is
lock the range of the bookend extent before truncation.

Signed-off-by: Yan Zheng <zheng.yan@oracle.com>

4 files changed, 43 insertions, 32 deletions

diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index fada9c2..535cee47 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -3379,11 +3379,13 @@ static int noinline relocate_data_extent(struct inode *reloc_inode,
 	struct btrfs_root *root = BTRFS_I(reloc_inode)->root;
 	struct extent_map_tree *em_tree = &BTRFS_I(reloc_inode)->extent_tree;
 	struct extent_map *em;
+	u64 start = extent_key->objectid - offset;
+	u64 end = start + extent_key->offset - 1;
 
 	em = alloc_extent_map(GFP_NOFS);
 	BUG_ON(!em || IS_ERR(em));
 
-	em->start = extent_key->objectid - offset;
+	em->start = start;
 	em->len = extent_key->offset;
 	em->block_len = extent_key->offset;
 	em->block_start = extent_key->objectid;
@@ -3391,7 +3393,7 @@ static int noinline relocate_data_extent(struct inode *reloc_inode,
 	set_bit(EXTENT_FLAG_PINNED, &em->flags);
 
 	/* setup extent map to cheat btrfs_readpage */
-	mutex_lock(&BTRFS_I(reloc_inode)->extent_mutex);
+	lock_extent(&BTRFS_I(reloc_inode)->io_tree, start, end, GFP_NOFS);
 	while (1) {
 		int ret;
 		spin_lock(&em_tree->lock);
@@ -3401,13 +3403,11 @@ static int noinline relocate_data_extent(struct inode *reloc_inode,
 			free_extent_map(em);
 			break;
 		}
-		btrfs_drop_extent_cache(reloc_inode, em->start,
-					em->start + em->len - 1, 0);
+		btrfs_drop_extent_cache(reloc_inode, start, end, 0);
 	}
-	mutex_unlock(&BTRFS_I(reloc_inode)->extent_mutex);
+	unlock_extent(&BTRFS_I(reloc_inode)->io_tree, start, end, GFP_NOFS);
 
-	return relocate_inode_pages(reloc_inode, extent_key->objectid - offset,
-				    extent_key->offset);
+	return relocate_inode_pages(reloc_inode, start, extent_key->offset);
 }
 
 struct btrfs_ref_path {
@@ -3831,7 +3831,6 @@ next:
 			 * the file extent item was modified by someone
 			 * before the extent got locked.
 			 */
-			mutex_unlock(&BTRFS_I(inode)->extent_mutex);
 			unlock_extent(&BTRFS_I(inode)->io_tree, lock_start,
 				      lock_end, GFP_NOFS);
 			extent_locked = 0;
@@ -3896,8 +3895,12 @@ next:
 			lock_start = key.offset;
 			lock_end = lock_start + num_bytes - 1;
 		} else {
-			BUG_ON(lock_start != key.offset);
-			BUG_ON(lock_end - lock_start + 1 < num_bytes);
+			if (lock_start > key.offset ||
+			    lock_end + 1 < key.offset + num_bytes) {
+				unlock_extent(&BTRFS_I(inode)->io_tree,
+					      lock_start, lock_end, GFP_NOFS);
+				extent_locked = 0;
+			}
 		}
 
 		if (!inode) {
@@ -3951,7 +3954,6 @@ next:
 			if (ordered)
 				btrfs_put_ordered_extent(ordered);
 
-			mutex_lock(&BTRFS_I(inode)->extent_mutex);
 			extent_locked = 1;
 			continue;
 		}
@@ -4073,7 +4075,6 @@ next:
 		}
 
 		if (extent_locked) {
-			mutex_unlock(&BTRFS_I(inode)->extent_mutex);
 			unlock_extent(&BTRFS_I(inode)->io_tree, lock_start,
 				      lock_end, GFP_NOFS);
 			extent_locked = 0;
@@ -4091,7 +4092,6 @@ out:
 	if (inode) {
 		mutex_unlock(&inode->i_mutex);
 		if (extent_locked) {
-			mutex_unlock(&BTRFS_I(inode)->extent_mutex);
 			unlock_extent(&BTRFS_I(inode)->io_tree, lock_start,
 				      lock_end, GFP_NOFS);
 		}
@@ -4180,10 +4180,8 @@ static int noinline invalidate_extent_cache(struct btrfs_root *root,
 
 		lock_extent(&BTRFS_I(inode)->io_tree, key.offset,
 			    key.offset + num_bytes - 1, GFP_NOFS);
-		mutex_lock(&BTRFS_I(inode)->extent_mutex);
 		btrfs_drop_extent_cache(inode, key.offset,
 					key.offset + num_bytes - 1, 1);
-		mutex_unlock(&BTRFS_I(inode)->extent_mutex);
 		unlock_extent(&BTRFS_I(inode)->io_tree, key.offset,
 			      key.offset + num_bytes - 1, GFP_NOFS);
 		cond_resched();
diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c
index 7503bd4..65a0583 100644
--- a/fs/btrfs/extent_io.c
+++ b/fs/btrfs/extent_io.c
@@ -946,8 +946,12 @@ int try_lock_extent(struct extent_io_tree *tree, u64 start, u64 end,
 
 	err = set_extent_bit(tree, start, end, EXTENT_LOCKED, 1,
 			     &failed_start, mask);
-	if (err == -EEXIST)
+	if (err == -EEXIST) {
+		if (failed_start > start)
+			clear_extent_bit(tree, start, failed_start - 1,
+					 EXTENT_LOCKED, 1, 0, mask);
 		return 0;
+	}
 	return 1;
 }
 EXPORT_SYMBOL(try_lock_extent);
diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index b8a7637..1a0510a 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -364,6 +364,7 @@ int noinline btrfs_drop_extents(struct btrfs_trans_handle *trans,
 		       u64 start, u64 end, u64 inline_limit, u64 *hint_byte)
 {
 	u64 extent_end = 0;
+	u64 locked_end = end;
 	u64 search_start = start;
 	u64 leaf_start;
 	u64 ram_bytes = 0;
@@ -479,12 +480,6 @@ next_slot:
 			goto next_slot;
 		}
 
-		if (found_inline) {
-			u64 mask = root->sectorsize - 1;
-			search_start = (extent_end + mask) & ~mask;
-		} else
-			search_start = extent_end;
-
 		if (end <= extent_end && start >= key.offset && found_inline)
 			*hint_byte = EXTENT_MAP_INLINE;
 
@@ -501,6 +496,26 @@ next_slot:
 			if (found_inline && start <= key.offset)
 				keep = 1;
 		}
+
+		if (bookend && found_extent && locked_end < extent_end) {
+			ret = try_lock_extent(&BTRFS_I(inode)->io_tree,
+					locked_end, extent_end - 1, GFP_NOFS);
+			if (!ret) {
+				btrfs_release_path(root, path);
+				lock_extent(&BTRFS_I(inode)->io_tree,
+					locked_end, extent_end - 1, GFP_NOFS);
+				locked_end = extent_end;
+				continue;
+			}
+			locked_end = extent_end;
+		}
+
+		if (found_inline) {
+			u64 mask = root->sectorsize - 1;
+			search_start = (extent_end + mask) & ~mask;
+		} else
+			search_start = extent_end;
+
 		/* truncate existing extent */
 		if (start > key.offset) {
 			u64 new_num;
@@ -638,6 +653,10 @@ next_slot:
 	}
 out:
 	btrfs_free_path(path);
+	if (locked_end > end) {
+		unlock_extent(&BTRFS_I(inode)->io_tree, end, locked_end - 1,
+			      GFP_NOFS);
+	}
 	btrfs_check_file(root, inode);
 	return ret;
 }
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 8254d6f..e8511d1 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -246,7 +246,6 @@ static int cow_file_range_inline(struct btrfs_trans_handle *trans,
 		return 1;
 	}
 
-	mutex_lock(&BTRFS_I(inode)->extent_mutex);
 	ret = btrfs_drop_extents(trans, root, inode, start,
 				 aligned_end, aligned_end, &hint_byte);
 	BUG_ON(ret);
@@ -258,7 +257,6 @@ static int cow_file_range_inline(struct btrfs_trans_handle *trans,
 				   compressed_pages);
 	BUG_ON(ret);
 	btrfs_drop_extent_cache(inode, start, aligned_end, 0);
-	mutex_unlock(&BTRFS_I(inode)->extent_mutex);
 	return 0;
 }
 
@@ -437,9 +435,7 @@ again:
 	BUG_ON(disk_num_bytes >
 	       btrfs_super_total_bytes(&root->fs_info->super_copy));
 
-	mutex_lock(&BTRFS_I(inode)->extent_mutex);
 	btrfs_drop_extent_cache(inode, start, start + num_bytes - 1, 0);
-	mutex_unlock(&BTRFS_I(inode)->extent_mutex);
 
 	while(disk_num_bytes > 0) {
 		unsigned long min_bytes;
@@ -477,8 +473,6 @@ again:
 		em->block_start = ins.objectid;
 		em->block_len = ins.offset;
 		em->bdev = root->fs_info->fs_devices->latest_bdev;
-
-		mutex_lock(&BTRFS_I(inode)->extent_mutex);
 		set_bit(EXTENT_FLAG_PINNED, &em->flags);
 
 		if (will_compress)
@@ -495,7 +489,6 @@ again:
 			btrfs_drop_extent_cache(inode, start,
 						start + ram_size - 1, 0);
 		}
-		mutex_unlock(&BTRFS_I(inode)->extent_mutex);
 
 		cur_alloc_size = ins.offset;
 		ret = btrfs_add_ordered_extent(inode, start, ins.objectid,
@@ -1016,8 +1009,6 @@ static int btrfs_finish_ordered_io(struct inode *inode, u64 start, u64 end)
 
 	INIT_LIST_HEAD(&list);
 
-	mutex_lock(&BTRFS_I(inode)->extent_mutex);
-
 	ret = btrfs_drop_extents(trans, root, inode,
 				 ordered_extent->file_offset,
 				 ordered_extent->file_offset +
@@ -1059,7 +1050,6 @@ static int btrfs_finish_ordered_io(struct inode *inode, u64 start, u64 end)
 	btrfs_drop_extent_cache(inode, ordered_extent->file_offset,
 				ordered_extent->file_offset +
 				ordered_extent->len - 1, 0);
-	mutex_unlock(&BTRFS_I(inode)->extent_mutex);
 
 	ins.objectid = ordered_extent->start;
 	ins.offset = ordered_extent->disk_len;