This is a note to let you know that I've just added the patch titled xfs: remove all COW fork extents when remounting readonly to the 5.10-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary The filename of the patch is: xfs-remove-all-cow-fork-extents-when-remounting-readonly.patch and it can be found in the queue-5.10 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable@xxxxxxxxxxxxxxx> know about it. >From foo@baz Thu Jun 30 01:20:25 PM CEST 2022 From: Amir Goldstein <amir73il@xxxxxxxxx> Date: Mon, 27 Jun 2022 09:51:39 +0300 Subject: xfs: remove all COW fork extents when remounting readonly To: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx> Cc: Sasha Levin <sashal@xxxxxxxxxx>, "Darrick J . Wong" <djwong@xxxxxxxxxx>, Leah Rumancik <leah.rumancik@xxxxxxxxx>, Chandan Babu R <chandan.babu@xxxxxxxxxx>, Luis Chamberlain <mcgrof@xxxxxxxxxx>, Adam Manzanares <a.manzanares@xxxxxxxxxxx>, linux-xfs@xxxxxxxxxxxxxxx, stable@xxxxxxxxxxxxxxx, Dave Chinner <dchinner@xxxxxxxxxx> Message-ID: <20220627065140.2798412-5-amir73il@xxxxxxxxx> From: "Darrick J. Wong" <djwong@xxxxxxxxxx> commit 089558bc7ba785c03815a49c89e28ad9b8de51f9 upstream. [backport xfs_icwalk -> xfs_eofblocks for 5.10.y] As part of multiple customer escalations due to file data corruption after copy on write operations, I wrote some fstests that use fsstress to hammer on COW to shake things loose. Regrettably, I caught some filesystem shutdowns due to incorrect rmap operations with the following loop: mount <filesystem> # (0) fsstress <run only readonly ops> & # (1) while true; do fsstress <run all ops> mount -o remount,ro # (2) fsstress <run only readonly ops> mount -o remount,rw # (3) done When (2) happens, notice that (1) is still running. xfs_remount_ro will call xfs_blockgc_stop to walk the inode cache to free all the COW extents, but the blockgc mechanism races with (1)'s reader threads to take IOLOCKs and loses, which means that it doesn't clean them all out. Call such a file (A). When (3) happens, xfs_remount_rw calls xfs_reflink_recover_cow, which walks the ondisk refcount btree and frees any COW extent that it finds. This function does not check the inode cache, which means that incore COW forks of inode (A) is now inconsistent with the ondisk metadata. If one of those former COW extents are allocated and mapped into another file (B) and someone triggers a COW to the stale reservation in (A), A's dirty data will be written into (B) and once that's done, those blocks will be transferred to (A)'s data fork without bumping the refcount. The results are catastrophic -- file (B) and the refcount btree are now corrupt. Solve this race by forcing the xfs_blockgc_free_space to run synchronously, which causes xfs_icwalk to return to inodes that were skipped because the blockgc code couldn't take the IOLOCK. This is safe to do here because the VFS has already prohibited new writer threads. Fixes: 10ddf64e420f ("xfs: remove leftover CoW reservations when remounting ro") Signed-off-by: Darrick J. Wong <djwong@xxxxxxxxxx> Reviewed-by: Dave Chinner <dchinner@xxxxxxxxxx> Reviewed-by: Chandan Babu R <chandan.babu@xxxxxxxxxx> Signed-off-by: Amir Goldstein <amir73il@xxxxxxxxx> Acked-by: Darrick J. Wong <djwong@xxxxxxxxxx> Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx> --- fs/xfs/xfs_super.c | 14 +++++++++++--- 1 file changed, 11 insertions(+), 3 deletions(-) --- a/fs/xfs/xfs_super.c +++ b/fs/xfs/xfs_super.c @@ -1695,7 +1695,10 @@ static int xfs_remount_ro( struct xfs_mount *mp) { - int error; + struct xfs_eofblocks eofb = { + .eof_flags = XFS_EOF_FLAGS_SYNC, + }; + int error; /* * Cancel background eofb scanning so it cannot race with the final @@ -1703,8 +1706,13 @@ xfs_remount_ro( */ xfs_stop_block_reaping(mp); - /* Get rid of any leftover CoW reservations... */ - error = xfs_icache_free_cowblocks(mp, NULL); + /* + * Clear out all remaining COW staging extents and speculative post-EOF + * preallocations so that we don't leave inodes requiring inactivation + * cleanups during reclaim on a read-only mount. We must process every + * cached inode, so this requires a synchronous cache scan. + */ + error = xfs_icache_free_cowblocks(mp, &eofb); if (error) { xfs_force_shutdown(mp, SHUTDOWN_CORRUPT_INCORE); return error; Patches currently in stable-queue which might be from amir73il@xxxxxxxxx are queue-5.10/xfs-fix-the-free-logic-of-state-in-xfs_attr_node_hasname.patch queue-5.10/xfs-punch-out-data-fork-delalloc-blocks-on-cow-writeback-failure.patch queue-5.10/xfs-use-kmem_cache_free-for-kmem_cache-objects.patch queue-5.10/xfs-remove-all-cow-fork-extents-when-remounting-readonly.patch queue-5.10/maintainers-add-amir-as-xfs-maintainer-for-5.10.y.patch queue-5.10/xfs-check-sb_meta_uuid-for-dabuf-buffer-recovery.patch