Skip to contentSkip to conversation

xfs: fix inode number overflow in ifree cluster helper

PATCH In review Brian Foster 3 participants 3 messages Apr 2, 2020 XFS filesystem

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 xfs: fix inode number overflow in ifree cluster helper

Qian Cai reports seemingly random buffer read verifier errors during filesystem writeback. This was isolated to a recent patch that factored out some inode cluster freeing code and happened to cast an unsigned inode number type to a signed value. If the inode number value overflows, we can skip marking in-core inodes associated with the underlying buffer stale at the time the physical inodes are freed. If such an inode happens to be dirty, xfsaild will eventually attempt to write it back over non-inode blocks. The invalidation of the underlying inode buffer causes writeback to read the buffer from disk. This fails the read verifier (preventing eventual corruption) if the buffer no longer looks like an inode cluster. Analysis by Dave Chinner.

Fix up the helper to use the proper type for inode number values.

Fixes 5806165a6663 ("xfs: factor inode lookup from xfs_ifree_cluster")

---

Fixes the problem described here[1]. I wasn't sure if we planned on fixing the original patch in for-next or wanted a separate patch. Feel free to commit standalone or fold into the original...

Brian

[1] https://lore.kernel.org/linux-xfs/[email protected]/

 fs/xfs/xfs_inode.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
fs/xfs/xfs_inode.c +1 -1
index 0cac0d37e3ae..ae86c870da92 100644
@@ -2511,7 +2511,7 @@ static struct xfs_inode *
 xfs_ifree_get_one_inode(
 	struct xfs_perag	*pag,
 	struct xfs_inode	*free_ip,
-	int			inum)
+	xfs_ino_t		inum)
 {
 	struct xfs_mount	*mp = pag->pag_mount;
 	struct xfs_inode	*ip;
Signature · 1 line
2.21.1
Reviewed-by on v1 1/1
#1 lore raw
Quoted 15 lines · Brian Foster

Qian Cai reports seemingly random buffer read verifier errors during filesystem writeback. This was isolated to a recent patch that factored out some inode cluster freeing code and happened to cast an unsigned inode number type to a signed value. If the inode number value overflows, we can skip marking in-core inodes associated with the underlying buffer stale at the time the physical inodes are freed. If such an inode happens to be dirty, xfsaild will eventually attempt to write it back over non-inode blocks. The invalidation of the underlying inode buffer causes writeback to read the buffer from disk. This fails the read verifier (preventing eventual corruption) if the buffer no longer looks like an inode cluster. Analysis by Dave Chinner.

Fix up the helper to use the proper type for inode number values.

Quoted 8 lines
Fixes 5806165a6663 ("xfs: factor inode lookup from xfs_ifree_cluster")

---

Fixes the problem described here[1]. I wasn't sure if we planned on fixing the original patch in for-next or wanted a separate patch. Feel free to commit standalone or fold into the original...

Looks good:

Reviewed-by on v1 1/1
#2 lore raw
In reply to Brian Foster
Quoted 20 lines · Brian Foster

Qian Cai reports seemingly random buffer read verifier errors during filesystem writeback. This was isolated to a recent patch that factored out some inode cluster freeing code and happened to cast an unsigned inode number type to a signed value. If the inode number value overflows, we can skip marking in-core inodes associated with the underlying buffer stale at the time the physical inodes are freed. If such an inode happens to be dirty, xfsaild will eventually attempt to write it back over non-inode blocks. The invalidation of the underlying inode buffer causes writeback to read the buffer from disk. This fails the read verifier (preventing eventual corruption) if the buffer no longer looks like an inode cluster. Analysis by Dave Chinner.

Fix up the helper to use the proper type for inode number values.

Fixes 5806165a6663 ("xfs: factor inode lookup from xfs_ifree_cluster")

---

Quoted 7 lines

Fixes the problem described here[1]. I wasn't sure if we planned on fixing the original patch in for-next or wanted a separate patch. Feel free to commit standalone or fold into the original...

Brian

[1] https://lore.kernel.org/linux-xfs/[email protected]/

Looks ok, I'll probably just attach it to the end rather than rebase for-next at this super late point...

--D

Quoted 20 lines
 fs/xfs/xfs_inode.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
fs/xfs/xfs_inode.c +1 -1
index 0cac0d37e3ae..ae86c870da92 100644
@@ -2511,7 +2511,7 @@ static struct xfs_inode *
 xfs_ifree_get_one_inode(
 	struct xfs_perag	*pag,
 	struct xfs_inode	*free_ip,
-	int			inum)
+	xfs_ino_t		inum)
 {
 	struct xfs_mount	*mp = pag->pag_mount;
 	struct xfs_inode	*ip;
Signature · 1 line
2.21.1