This is a note to let you know that I've just added the patch titled
xfs: punching delalloc extents on write failure is racy
to the 6.1-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-punching-delalloc-extents-on-write-failure-is-racy.patch
and it can be found in the queue-6.1 subdirectory.
If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@xxxxxxxxxxxxxxx> know about it.
>From stable+bounces-42891-greg=kroah.com@xxxxxxxxxxxxxxx Wed May 1 20:41:26 2024
From: Leah Rumancik <leah.rumancik@xxxxxxxxx>
Date: Wed, 1 May 2024 11:40:50 -0700
Subject: xfs: punching delalloc extents on write failure is racy
To: stable@xxxxxxxxxxxxxxx
Cc: linux-xfs@xxxxxxxxxxxxxxx, amir73il@xxxxxxxxx, chandan.babu@xxxxxxxxxx, fred@xxxxxxxxxxxxxx, Dave Chinner <dchinner@xxxxxxxxxx>, Christoph Hellwig <hch@xxxxxx>, "Darrick J . Wong" <djwong@xxxxxxxxxx>, Leah Rumancik <leah.rumancik@xxxxxxxxx>
Message-ID: <20240501184112.3799035-2-leah.rumancik@xxxxxxxxx>
From: Dave Chinner <dchinner@xxxxxxxxxx>
[ Upstream commit 198dd8aedee6a7d2de0dfa739f9a008a938f6848 ]
xfs_buffered_write_iomap_end() has a comment about the safety of
punching delalloc extents based holding the IOLOCK_EXCL. This
comment is wrong, and punching delalloc extents is not race free.
When we punch out a delalloc extent after a write failure in
xfs_buffered_write_iomap_end(), we punch out the page cache with
truncate_pagecache_range() before we punch out the delalloc extents.
At this point, we only hold the IOLOCK_EXCL, so there is nothing
stopping mmap() write faults racing with this cleanup operation,
reinstantiating a folio over the range we are about to punch and
hence requiring the delalloc extent to be kept.
If this race condition is hit, we can end up with a dirty page in
the page cache that has no delalloc extent or space reservation
backing it. This leads to bad things happening at writeback time.
To avoid this race condition, we need the page cache truncation to
be atomic w.r.t. the extent manipulation. We can do this by holding
the mapping->invalidate_lock exclusively across this operation -
this will prevent new pages from being inserted into the page cache
whilst we are removing the pages and the backing extent and space
reservation.
Taking the mapping->invalidate_lock exclusively in the buffered
write IO path is safe - it naturally nests inside the IOLOCK (see
truncate and fallocate paths). iomap_zero_range() can be called from
under the mapping->invalidate_lock (from the truncate path via
either xfs_zero_eof() or xfs_truncate_page(), but iomap_zero_iter()
will not instantiate new delalloc pages (because it skips holes) and
hence will not ever need to punch out delalloc extents on failure.
Fix the locking issue, and clean up the code logic a little to avoid
unnecessary work if we didn't allocate the delalloc extent or wrote
the entire region we allocated.
Signed-off-by: Dave Chinner <dchinner@xxxxxxxxxx>
Reviewed-by: Christoph Hellwig <hch@xxxxxx>
Reviewed-by: Darrick J. Wong <djwong@xxxxxxxxxx>
Signed-off-by: Leah Rumancik <leah.rumancik@xxxxxxxxx>
Acked-by: Darrick J. Wong <djwong@xxxxxxxxxx>
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
---
fs/xfs/xfs_iomap.c | 41 +++++++++++++++++++++++------------------
1 file changed, 23 insertions(+), 18 deletions(-)
--- a/fs/xfs/xfs_iomap.c
+++ b/fs/xfs/xfs_iomap.c
@@ -1147,6 +1147,10 @@ xfs_buffered_write_iomap_end(
written = 0;
}
+ /* If we didn't reserve the blocks, we're not allowed to punch them. */
+ if (!(iomap->flags & IOMAP_F_NEW))
+ return 0;
+
/*
* start_fsb refers to the first unused block after a short write. If
* nothing was written, round offset down to point at the first block in
@@ -1158,27 +1162,28 @@ xfs_buffered_write_iomap_end(
start_fsb = XFS_B_TO_FSB(mp, offset + written);
end_fsb = XFS_B_TO_FSB(mp, offset + length);
+ /* Nothing to do if we've written the entire delalloc extent */
+ if (start_fsb >= end_fsb)
+ return 0;
+
/*
- * Trim delalloc blocks if they were allocated by this write and we
- * didn't manage to write the whole range.
- *
- * We don't need to care about racing delalloc as we hold i_mutex
- * across the reserve/allocate/unreserve calls. If there are delalloc
- * blocks in the range, they are ours.
+ * Lock the mapping to avoid races with page faults re-instantiating
+ * folios and dirtying them via ->page_mkwrite between the page cache
+ * truncation and the delalloc extent removal. Failing to do this can
+ * leave dirty pages with no space reservation in the cache.
*/
- if ((iomap->flags & IOMAP_F_NEW) && start_fsb < end_fsb) {
- truncate_pagecache_range(VFS_I(ip), XFS_FSB_TO_B(mp, start_fsb),
- XFS_FSB_TO_B(mp, end_fsb) - 1);
-
- error = xfs_bmap_punch_delalloc_range(ip, start_fsb,
- end_fsb - start_fsb);
- if (error && !xfs_is_shutdown(mp)) {
- xfs_alert(mp, "%s: unable to clean up ino %lld",
- __func__, ip->i_ino);
- return error;
- }
- }
+ filemap_invalidate_lock(inode->i_mapping);
+ truncate_pagecache_range(VFS_I(ip), XFS_FSB_TO_B(mp, start_fsb),
+ XFS_FSB_TO_B(mp, end_fsb) - 1);
+ error = xfs_bmap_punch_delalloc_range(ip, start_fsb,
+ end_fsb - start_fsb);
+ filemap_invalidate_unlock(inode->i_mapping);
+ if (error && !xfs_is_shutdown(mp)) {
+ xfs_alert(mp, "%s: unable to clean up ino %lld",
+ __func__, ip->i_ino);
+ return error;
+ }
return 0;
}
Patches currently in stable-queue which might be from kroah.com@xxxxxxxxxxxxxxx are
queue-6.1/xfs-iomap-move-delalloc-punching-to-iomap.patch
queue-6.1/xfs-fix-off-by-one-block-in-xfs_discard_folio.patch
queue-6.1/xfs-invalidate-block-device-page-cache-during-unmount.patch
queue-6.1/xfs-drop-write-error-injection-is-unfixable-remove-it.patch
queue-6.1/iomap-buffered-write-failure-should-not-truncate-the-page-cache.patch
queue-6.1/xfs-fix-super-block-buf-log-item-uaf-during-force-shutdown.patch
queue-6.1/xfs-fix-incorrect-i_nlink-caused-by-inode-racing.patch
queue-6.1/xfs-estimate-post-merge-refcounts-correctly.patch
queue-6.1/xfs-fix-log-recovery-when-unknown-rocompat-bits-are-set.patch
queue-6.1/xfs-punching-delalloc-extents-on-write-failure-is-racy.patch
queue-6.1/xfs-allow-inode-inactivation-during-a-ro-mount-log-recovery.patch
queue-6.1/iomap-write-iomap-validity-checks.patch
queue-6.1/xfs-attach-dquots-to-inode-before-reading-data-cow-fork-mappings.patch
queue-6.1/xfs-fix-sb-write-verify-for-lazysbcount.patch
queue-6.1/xfs-wait-iclog-complete-before-tearing-down-ail.patch
queue-6.1/xfs-use-byte-ranges-for-write-cleanup-ranges.patch
queue-6.1/xfs-xfs_bmap_punch_delalloc_range-should-take-a-byte-range.patch
queue-6.1/xfs-write-page-faults-in-iomap-are-not-buffered-writes.patch
queue-6.1/xfs-short-circuit-xfs_growfs_data_private-if-delta-is-zero.patch
queue-6.1/xfs-fix-incorrect-error-out-in-xfs_remove.patch
queue-6.1/xfs-invalidate-xfs_bufs-when-allocating-cow-extents.patch
queue-6.1/xfs-hoist-refcount-record-merge-predicates.patch
queue-6.1/xfs-get-root-inode-correctly-at-bulkstat.patch
queue-6.1/xfs-use-iomap_valid-method-to-detect-stale-cached-iomaps.patch
