Even if the slot is already read out, we may still need to re-balance
the tree, thus it can cause error in that btrfs_del_item() call.
Link: https://lore.kernel.org/linux-btrfs/9b45f60f-c849-e072-64ef-298937e6a8a3@xxxxxxxx/T/#t
Signed-off-by: void0red <void0red@xxxxxxxxx>
---
fs/btrfs/delayed-inode.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/fs/btrfs/delayed-inode.c b/fs/btrfs/delayed-inode.c
index cac5169eaf8d..88a5d893972d 100644
--- a/fs/btrfs/delayed-inode.c
+++ b/fs/btrfs/delayed-inode.c
@@ -1043,7 +1043,7 @@ static int __btrfs_update_delayed_inode(struct btrfs_trans_handle *trans,
* so there is only one iref. The case that several irefs are
* in the same item doesn't exist.
*/
- btrfs_del_item(trans, root, path);
+ ret = btrfs_del_item(trans, root, path);
out:
btrfs_release_delayed_iref(node);
btrfs_release_path(path);