[PATCH 3.13.y-ckt 44/96] btrfs: fix use after free iterating extrefs

Kamal Mostafa kamal at canonical.com
Fri Nov 13 21:48:58 UTC 2015


3.13.11-ckt30 -stable review patch.  If anyone has any objections, please let me know.

------------------

From: Chris Mason <clm at fb.com>

commit dc6c5fb3b514221f2e9d21ee626a9d95d3418dff upstream.

The code for btrfs inode-resolve has never worked properly for
files with enough hard links to trigger extrefs.  It was trying to
get the leaf out of a path after freeing the path:

	btrfs_release_path(path);
	leaf = path->nodes[0];
	item_size = btrfs_item_size_nr(leaf, slot);

The fix here is to use the extent buffer we cloned just a little higher
up to avoid deadlocks caused by using the leaf in the path.

Signed-off-by: Chris Mason <clm at fb.com>
cc: Mark Fasheh <mfasheh at suse.de>
Reviewed-by: Filipe Manana <fdmanana at suse.com>
Reviewed-by: Mark Fasheh <mfasheh at suse.de>
[ kamal: backport to 3.13-stable: context ]
Signed-off-by: Kamal Mostafa <kamal at canonical.com>
---
 fs/btrfs/backref.c | 8 +++-----
 1 file changed, 3 insertions(+), 5 deletions(-)

diff --git a/fs/btrfs/backref.c b/fs/btrfs/backref.c
index 5ee99e3..e6b6cc0 100644
--- a/fs/btrfs/backref.c
+++ b/fs/btrfs/backref.c
@@ -1656,7 +1656,6 @@ static int iterate_inode_extrefs(u64 inum, struct btrfs_root *fs_root,
 	int found = 0;
 	struct extent_buffer *eb;
 	struct btrfs_inode_extref *extref;
-	struct extent_buffer *leaf;
 	u32 item_size;
 	u32 cur_offset;
 	unsigned long ptr;
@@ -1681,9 +1680,8 @@ static int iterate_inode_extrefs(u64 inum, struct btrfs_root *fs_root,
 		btrfs_set_lock_blocking_rw(eb, BTRFS_READ_LOCK);
 		btrfs_release_path(path);
 
-		leaf = path->nodes[0];
-		item_size = btrfs_item_size_nr(leaf, path->slots[0]);
-		ptr = btrfs_item_ptr_offset(leaf, path->slots[0]);
+		item_size = btrfs_item_size_nr(eb, path->slots[0]);
+		ptr = btrfs_item_ptr_offset(eb, path->slots[0]);
 		cur_offset = 0;
 
 		while (cur_offset < item_size) {
@@ -1697,7 +1695,7 @@ static int iterate_inode_extrefs(u64 inum, struct btrfs_root *fs_root,
 			if (ret)
 				break;
 
-			cur_offset += btrfs_inode_extref_name_len(leaf, extref);
+			cur_offset += btrfs_inode_extref_name_len(eb, extref);
 			cur_offset += sizeof(*extref);
 		}
 		btrfs_tree_read_unlock_blocking(eb);
-- 
1.9.1





More information about the kernel-team mailing list