linux/fs/btrfs
Qu Wenruo e2b3859067 btrfs: ensure no dirty metadata is written back for an fs with errors
commit 2618849f31 upstream.

[BUG]
During development of a minor feature (make sure all btrfs_bio::end_io()
is called in task context), I noticed a crash in generic/388, where
metadata writes triggered new works after btrfs_stop_all_workers().

It turns out that it can even happen without any code modification, just
using RAID5 for metadata and the same workload from generic/388 is going
to trigger the use-after-free.

[CAUSE]
If btrfs hits an error, the fs is marked as error, no new
transaction is allowed thus metadata is in a frozen state.

But there are some metadata modifications before that error, and they are
still in the btree inode page cache.

Since there will be no real transaction commit, all those dirty folios
are just kept as is in the page cache, and they can not be invalidated
by invalidate_inode_pages2() call inside close_ctree(), because they are
dirty.

And finally after btrfs_stop_all_workers(), we call iput() on btree
inode, which triggers writeback of those dirty metadata.

And if the fs is using RAID56 metadata, this will trigger RMW and queue
new works into rmw_workers, which is already stopped, causing warning
from queue_work() and use-after-free.

[FIX]
Add a special handling for write_one_eb(), that if the fs is already in
an error state, immediately mark the bbio as failure, instead of really
submitting them.

Then during close_ctree(), iput() will just discard all those dirty
tree blocks without really writing them back, thus no more new jobs for
already stopped-and-freed workqueues.

The extra discard in write_one_eb() also acts as an extra safenet.
E.g. the transaction abort is triggered by some extent/free space
tree corruptions, and since extent/free space tree is already corrupted
some tree blocks may be allocated where they shouldn't be (overwriting
existing tree blocks). In that case writing them back will further
corrupting the fs.

CC: stable@vger.kernel.org # 6.6+
Reviewed-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
History 2025-11-13 15:34:36 -05:00
..
tests btrfs: use unsigned types for constants defined as bit shifts 2025-07-06 11:01:37 +02:00
accessors.c move asm/unaligned.h to linux/unaligned.h 2024-10-02 17:23:23 -04:00
accessors.h move asm/unaligned.h to linux/unaligned.h 2024-10-02 17:23:23 -04:00
acl.c btrfs: remove unused included headers 2024-03-04 16:24:46 +01:00
acl.h btrfs: add forward declarations and headers, part 1 2024-03-04 16:24:49 +01:00
async-thread.c btrfs: remove unused included headers 2024-03-04 16:24:46 +01:00
async-thread.h btrfs: add forward declarations and headers, part 1 2024-03-04 16:24:49 +01:00
backref.c btrfs: drop the backref cache during relocation if we commit 2024-10-01 19:10:26 +02:00
backref.h btrfs: use unsigned types for constants defined as bit shifts 2025-07-06 11:01:37 +02:00
bio.c btrfs: handle bio_split() errors 2025-01-09 13:33:44 +01:00
bio.h btrfs: fix error propagation of split bios 2024-10-23 18:17:43 +02:00
block-group.c btrfs: zoned: requeue to unused block group list if zone finish failed 2025-08-28 16:31:04 +02:00
block-group.h btrfs: constify more pointer parameters 2024-09-10 16:51:22 +02:00
block-rsv.c btrfs: constify more pointer parameters 2024-09-10 16:51:22 +02:00
block-rsv.h btrfs: constify more pointer parameters 2024-09-10 16:51:22 +02:00
btrfs_inode.h btrfs: fix race between setting last_dir_index_offset and inode logging 2025-09-09 18:58:02 +02:00
compression.c btrfs: compression: adjust cb->compressed_folios allocation type 2025-05-29 11:03:12 +02:00
compression.h btrfs: convert btrfs_decompress() to take a folio 2024-09-10 16:51:21 +02:00
ctree.c btrfs: abort transaction on unexpected eb generation at btrfs_copy_root() 2025-08-28 16:31:03 +02:00
ctree.h btrfs: fix improper generation check in snapshot delete 2024-12-27 14:02:14 +01:00
defrag.c btrfs: fix defrag not merging contiguous extents due to merged extent maps 2024-10-31 16:46:41 +01:00
defrag.h btrfs: drop transaction parameter from btrfs_add_inode_defrag() 2024-09-10 16:51:19 +02:00
delalloc-space.c btrfs: constify pointer parameters where applicable 2024-07-11 15:33:22 +02:00
delalloc-space.h btrfs: constify pointer parameters where applicable 2024-07-11 15:33:22 +02:00
delayed-inode.c btrfs: pass a btrfs_inode to btrfs_readdir_get_delayed_items() 2024-07-11 15:33:28 +02:00
delayed-inode.h btrfs: pass a btrfs_inode to btrfs_readdir_get_delayed_items() 2024-07-11 15:33:28 +02:00
delayed-ref.c btrfs: fix incorrect comparison for delayed refs 2024-11-14 16:11:02 +01:00
delayed-ref.h btrfs: use sector numbers as keys for the dirty extents xarray 2024-10-11 18:33:35 +02:00
dev-replace.c btrfs: don't take dev_replace rwsem on task already holding it 2024-12-14 20:03:36 +01:00
dev-replace.h btrfs: add forward declarations and headers, part 1 2024-03-04 16:24:49 +01:00
dir-item.c btrfs: fix passing 0 to ERR_PTR in btrfs_search_dir_index_item() 2024-10-22 16:10:55 +02:00
dir-item.h btrfs: constify pointer parameters where applicable 2024-07-11 15:33:22 +02:00
direct-io.c btrfs: use unsigned types for constants defined as bit shifts 2025-07-06 11:01:37 +02:00
direct-io.h btrfs: move the direct IO code into its own file 2024-07-11 15:33:29 +02:00
discard.c btrfs: make btrfs_discard_workfn() block_group ref explicit 2025-05-29 11:02:12 +02:00
discard.h btrfs: unexport btrfs_run_discard_work and make it static 2023-06-19 13:59:25 +02:00
disk-io.c btrfs: always drop log root tree reference in btrfs_replay_log() 2025-11-02 22:15:21 +09:00
disk-io.h btrfs: drop unused parameter options from open_ctree() 2024-12-14 20:04:12 +01:00
export.c btrfs: avoid potential out-of-bounds in btrfs_encode_fh() 2025-10-19 16:33:48 +02:00
export.h btrfs: add forward declarations and headers, part 1 2024-03-04 16:24:49 +01:00
extent-io-tree.c btrfs: exit after state split error at set_extent_bit() 2025-06-19 15:32:32 +02:00
extent-io-tree.h btrfs: introduce EXTENT_DIO_LOCKED 2024-09-10 16:51:20 +02:00
extent-tree.c btrfs: zoned: refine extent allocator hint selection 2025-11-02 22:15:21 +09:00
extent-tree.h btrfs: do not BUG_ON() when freeing tree block after error 2024-07-11 15:33:26 +02:00
extent_io.c btrfs: ensure no dirty metadata is written back for an fs with errors 2025-11-13 15:34:36 -05:00
extent_io.h btrfs: use unsigned types for constants defined as bit shifts 2025-07-06 11:01:37 +02:00
extent_map.c btrfs: fix use-after-free on inode when scanning root during em shrinking 2025-07-06 11:01:50 +02:00
extent_map.h btrfs: make the extent map shrinker run asynchronously as a work queue job 2025-07-06 11:01:49 +02:00
fiemap.c btrfs: initialize last_extent_end to fix -Wmaybe-uninitialized warning in extent_fiemap() 2024-08-26 16:58:13 +02:00
fiemap.h btrfs: move fiemap code into its own file 2024-07-11 15:33:20 +02:00
file-item.c btrfs: constify more pointer parameters 2024-09-10 16:51:22 +02:00
file-item.h btrfs: constify more pointer parameters 2024-09-10 16:51:22 +02:00
file.c btrfs: mark dirty extent range for out of bound prealloc extents 2025-11-13 15:34:33 -05:00
file.h btrfs: move the direct IO code into its own file 2024-07-11 15:33:29 +02:00
free-space-cache.c btrfs: add cancellation points to trim loops 2024-10-07 23:21:56 +02:00
free-space-cache.h btrfs: add cancellation points to trim loops 2024-10-07 23:21:56 +02:00
free-space-tree.c btrfs: do not assert we found block group item when creating free space tree 2025-10-23 16:20:20 +02:00
free-space-tree.h btrfs: add forward declarations and headers, part 2 2024-03-04 16:24:49 +01:00
fs.c btrfs: sysfs: update fs features directory asynchronously 2023-02-13 17:50:35 +01:00
fs.h btrfs: make the extent map shrinker run asynchronously as a work queue job 2025-07-06 11:01:49 +02:00
inode-item.c btrfs: constify more pointer parameters 2024-09-10 16:51:22 +02:00
inode-item.h btrfs: constify more pointer parameters 2024-09-10 16:51:22 +02:00
inode.c btrfs: zoned: return error from btrfs_zone_finish_endio() 2025-11-02 22:15:21 +09:00
ioctl.c btrfs: fix memory leak on duplicated memory in the qgroup assign ioctl 2025-10-23 16:20:20 +02:00
ioctl.h btrfs: constify pointer parameters where applicable 2024-07-11 15:33:22 +02:00
Kconfig btrfs: check-integrity: remove CONFIG_BTRFS_FS_CHECK_INTEGRITY option 2023-10-12 16:44:05 +02:00
locking.c btrfs: change root->root_key.objectid to btrfs_root_id() 2024-05-07 21:31:06 +02:00
locking.h btrfs: cleanup recursive include of the same header 2024-07-11 15:33:22 +02:00
lru_cache.c btrfs: fix typos found by codespell 2023-12-15 23:00:04 +01:00
lru_cache.h btrfs: cleanup recursive include of the same header 2024-07-11 15:33:22 +02:00
lzo.c btrfs: convert lzo_decompress() to take a folio 2024-09-10 16:51:21 +02:00
Makefile btrfs: move the direct IO code into its own file 2024-07-11 15:33:29 +02:00
messages.c btrfs: disable rate limiting when debug enabled 2024-10-01 19:29:41 +02:00
messages.h btrfs: constify fs_info parameter in __btrfs_panic() 2023-12-15 20:27:02 +01:00
misc.h btrfs: constify pointer parameters where applicable 2024-07-11 15:33:22 +02:00
ordered-data.c btrfs: fix qgroup reservation leak on failure to allocate ordered extent 2025-07-06 11:01:37 +02:00
ordered-data.h btrfs: convert btrfs_mark_ordered_io_finished() to take a folio 2024-09-10 16:51:14 +02:00
orphan.c btrfs: BTRFS_PATH_AUTO_FREE in orphan.c 2024-09-10 16:51:22 +02:00
orphan.h btrfs: add forward declarations and headers, part 1 2024-03-04 16:24:49 +01:00
print-tree.c btrfs: avoid using fixed char array size for tree names 2024-08-02 22:44:27 +02:00
print-tree.h btrfs: add forward declarations and headers, part 1 2024-03-04 16:24:49 +01:00
props.c btrfs: pass a btrfs_inode to btrfs_set_prop() 2024-07-11 15:33:29 +02:00
props.h btrfs: pass a btrfs_inode to btrfs_set_prop() 2024-07-11 15:33:29 +02:00
qgroup.c btrfs: fix memory leak of qgroup_list in btrfs_add_qgroup_relation 2025-11-13 15:34:33 -05:00
qgroup.h btrfs: qgroup: set a more sane default value for subtree drop threshold 2024-10-22 16:09:11 +02:00
raid-stripe-tree.c btrfs: change RST lookup error message level to debug 2024-09-10 16:51:17 +02:00
raid-stripe-tree.h btrfs: remove raid-stripe-tree encoding field from stripe_extent 2024-07-11 15:33:28 +02:00
raid56.c btrfs: use unsigned types for constants defined as bit shifts 2025-07-06 11:01:37 +02:00
raid56.h btrfs: add forward declarations and headers, part 2 2024-03-04 16:24:49 +01:00
rcu-string.h btrfs: add forward declarations and headers, part 1 2024-03-04 16:24:49 +01:00
ref-verify.c btrfs: ref-verify: handle damaged extent root tree 2025-10-12 12:57:17 +02:00
ref-verify.h btrfs: add forward declarations and headers, part 1 2024-03-04 16:24:49 +01:00
reflink.c btrfs: convert copy_inline_to_page() to use folio 2024-09-10 16:51:21 +02:00
reflink.h btrfs: add forward declarations and headers, part 1 2024-03-04 16:24:49 +01:00
relocation.c btrfs: fix clearing of BTRFS_FS_RELOC_RUNNING if relocation already running 2025-10-23 16:20:19 +02:00
relocation.h btrfs: add forward declarations and headers, part 1 2024-03-04 16:24:49 +01:00
root-tree.c btrfs: change root->root_key.objectid to btrfs_root_id() 2024-05-07 21:31:06 +02:00
root-tree.h btrfs: qgroup: fix qgroup prealloc rsv leak in subvolume operations 2024-04-02 19:18:23 +02:00
scrub.c btrfs: scrub: replace max_t()/min_t() with clamp() in scrub_throttle_dev_io() 2025-11-02 22:15:21 +09:00
scrub.h btrfs: add forward declarations and headers, part 1 2024-03-04 16:24:49 +01:00
send.c btrfs: send: make fs_path_len() inline and constify its argument 2025-08-28 16:31:04 +02:00
send.h btrfs: pass a btrfs_inode to btrfs_ioctl_send() 2024-07-11 15:33:28 +02:00
space-info.c btrfs: constify more pointer parameters 2024-09-10 16:51:22 +02:00
space-info.h btrfs: constify more pointer parameters 2024-09-10 16:51:22 +02:00
subpage.c btrfs: subpage: keep TOWRITE tag until folio is cleaned 2025-08-28 16:31:07 +02:00
subpage.h btrfs: rename btrfs_folio_(set|start|end)_writer_lock() 2025-02-27 04:30:14 -08:00
super.c btrfs: directly free partially initialized fs_info in btrfs_check_leaked_roots() 2025-10-29 14:08:56 +01:00
super.h btrfs: change BTRFS_MOUNT_* flags to 64bit type 2024-07-19 17:20:23 +02:00
sysfs.c btrfs: sysfs: fix direct super block member reads 2025-01-02 10:34:24 +01:00
sysfs.h btrfs: add forward declarations and headers, part 1 2024-03-04 16:24:49 +01:00
transaction.c btrfs: use smp_mb__after_atomic() when forcing COW in create_pending_snapshot() 2025-11-02 22:15:21 +09:00
transaction.h btrfs: fix race between direct IO write and fsync when using same fd 2024-09-03 20:29:55 +02:00
tree-checker.c btrfs: tree-checker: fix bounds check in check_inode_extref() 2025-11-02 22:15:23 +09:00
tree-checker.h btrfs: make sure that WRITTEN is set on all metadata blocks 2024-05-02 22:11:13 +02:00
tree-log.c btrfs: abort transaction if we fail to update inode in log replay dir fixup 2025-11-02 22:15:21 +09:00
tree-log.h btrfs: avoid transaction commit on any fsync after subvolume creation 2024-07-11 15:33:24 +02:00
tree-mod-log.c btrfs: constify more pointer parameters 2024-09-10 16:51:22 +02:00
tree-mod-log.h btrfs: constify more pointer parameters 2024-09-10 16:51:22 +02:00
ulist.c btrfs: preallocate ulist memory for qgroup rsv 2024-07-11 15:33:26 +02:00
ulist.h btrfs: preallocate ulist memory for qgroup rsv 2024-07-11 15:33:26 +02:00
uuid-tree.c move asm/unaligned.h to linux/unaligned.h 2024-10-02 17:23:23 -04:00
uuid-tree.h btrfs: move uuid tree related code to uuid-tree.[ch] 2024-09-10 16:51:12 +02:00
verity.c btrfs: add and use helper to verify the calling task has locked the inode 2024-09-10 16:51:22 +02:00
verity.h btrfs: add forward declarations and headers, part 1 2024-03-04 16:24:49 +01:00
volumes.c btrfs: don't allow adding block device of less than 1 MB 2025-10-02 13:44:08 +02:00
volumes.h btrfs: split remaining space to discard in chunks 2024-10-07 23:21:53 +02:00
xattr.c btrfs: add and use helper to verify the calling task has locked the inode 2024-09-10 16:51:22 +02:00
xattr.h btrfs: constify pointer parameters where applicable 2024-07-11 15:33:22 +02:00
zlib.c btrfs: zlib: fix avail_in bytes for s390 zlib HW compression path 2025-01-17 13:40:48 +01:00
zoned.c btrfs: zoned: return error from btrfs_zone_finish_endio() 2025-11-02 22:15:21 +09:00
zoned.h btrfs: zoned: return error from btrfs_zone_finish_endio() 2025-11-02 22:15:21 +09:00
zstd.c btrfs: use unsigned types for constants defined as bit shifts 2025-07-06 11:01:37 +02:00