commit 01fba45dea upstream.
With below scripts, it will trigger panic in f2fs:
mkfs.f2fs -f /dev/vdd
mount /dev/vdd /mnt/f2fs
touch /mnt/f2fs/foo
sync
echo 111 >> /mnt/f2fs/foo
f2fs_io fsync /mnt/f2fs/foo
f2fs_io shutdown 2 /mnt/f2fs
umount /mnt/f2fs
mount -o ro,norecovery /dev/vdd /mnt/f2fs
or
mount -o ro,disable_roll_forward /dev/vdd /mnt/f2fs
F2FS-fs (vdd): f2fs_recover_fsync_data: recovery fsync data, check_only: 0
F2FS-fs (vdd): Mounted with checkpoint version = 7f5c361f
F2FS-fs (vdd): Stopped filesystem due to reason: 0
F2FS-fs (vdd): f2fs_recover_fsync_data: recovery fsync data, check_only: 1
Filesystem f2fs get_tree() didn't set fc->root, returned 1
------------[ cut here ]------------
kernel BUG at fs/super.c:1761!
Oops: invalid opcode: 0000 [#1] SMP PTI
CPU: 3 UID: 0 PID: 722 Comm: mount Not tainted 6.18.0-rc2+ #721 PREEMPT(voluntary)
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.16.3-debian-1.16.3-2 04/01/2014
RIP: 0010:vfs_get_tree.cold+0x18/0x1a
Call Trace:
<TASK>
fc_mount+0x13/0xa0
path_mount+0x34e/0xc50
__x64_sys_mount+0x121/0x150
do_syscall_64+0x84/0x800
entry_SYSCALL_64_after_hwframe+0x76/0x7e
RIP: 0033:0x7fa6cc126cfe
The root cause is we missed to handle error number returned from
f2fs_recover_fsync_data() when mounting image w/ ro,norecovery or
ro,disable_roll_forward mount option, result in returning a positive
error number to vfs_get_tree(), fix it.
Cc: stable@kernel.org
Fixes: 6781eabba1 ("f2fs: give -EINVAL for norecovery and rw mount")
Signed-off-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 37345eae9d upstream.
w/ LFS mode, in get_left_section_blocks(), we should not account the
blocks which were used before and now are invalided, otherwise those
blocks will be counted as freed one in has_curseg_enough_space(), result
in missing to trigger GC in time.
Cc: stable@kernel.org
Fixes: 249ad438e1 ("f2fs: add a method for calculating the remaining blocks in the current segment in LFS mode.")
Fixes: bf34c93d26 ("f2fs: check curseg space before foreground GC")
Signed-off-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 68d05693f8 upstream.
mkfs.f2fs -f /dev/vdd
mount /dev/vdd /mnt/f2fs
touch /mnt/f2fs/foo
sync # avoid CP_UMOUNT_FLAG in last f2fs_checkpoint.ckpt_flags
touch /mnt/f2fs/bar
f2fs_io fsync /mnt/f2fs/bar
f2fs_io shutdown 2 /mnt/f2fs
umount /mnt/f2fs
blockdev --setro /dev/vdd
mount /dev/vdd /mnt/f2fs
mount: /mnt/f2fs: WARNING: source write-protected, mounted read-only.
For the case if we create and fsync a new inode before sudden power-cut,
without norecovery or disable_roll_forward mount option, the following
mount will succeed w/o recovering last fsynced inode.
The problem here is that we only check inode_list list after
find_fsync_dnodes() in f2fs_recover_fsync_data() to find out whether
there is recoverable data in the iamge, but there is a missed case, if
last fsynced inode is not existing in last checkpoint, then, we will
fail to get its inode due to nat of inode node is not existing in last
checkpoint, so the inode won't be linked in inode_list.
Let's detect such case in dyrun mode to fix this issue.
After this change, mount will fail as expected below:
mount: /mnt/f2fs: cannot mount /dev/vdd read-only.
dmesg(1) may have more information after failed mount system call.
demsg:
F2FS-fs (vdd): Need to recover fsync data, but write access unavailable, please try mount w/ disable_roll_forward or norecovery
Cc: stable@kernel.org
Fixes: 6781eabba1 ("f2fs: give -EINVAL for norecovery and rw mount")
Signed-off-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 392711ef18 upstream.
The one_time_gc field in struct victim_sel_policy is conditionally
initialized but unconditionally read, leading to undefined behavior
that triggers UBSAN warnings.
In f2fs_get_victim() at fs/f2fs/gc.c:774, the victim_sel_policy
structure is declared without initialization:
struct victim_sel_policy p;
The field p.one_time_gc is only assigned when the 'one_time' parameter
is true (line 789):
if (one_time) {
p.one_time_gc = one_time;
...
}
However, this field is unconditionally read in subsequent get_gc_cost()
at line 395:
if (p->one_time_gc && (valid_thresh_ratio < 100) && ...)
When one_time is false, p.one_time_gc contains uninitialized stack
memory. Hence p.one_time_gc is an invalid bool value.
UBSAN detects this invalid bool value:
UBSAN: invalid-load in fs/f2fs/gc.c:395:7
load of value 77 is not a valid value for type '_Bool'
CPU: 3 UID: 0 PID: 1297 Comm: f2fs_gc-252:16 Not tainted 6.18.0-rc3
#5 PREEMPT(voluntary)
Hardware name: OpenStack Foundation OpenStack Nova,
BIOS 1.13.0-1ubuntu1.1 04/01/2014
Call Trace:
<TASK>
dump_stack_lvl+0x70/0x90
dump_stack+0x14/0x20
__ubsan_handle_load_invalid_value+0xb3/0xf0
? dl_server_update+0x2e/0x40
? update_curr+0x147/0x170
f2fs_get_victim.cold+0x66/0x134 [f2fs]
? sched_balance_newidle+0x2ca/0x470
? finish_task_switch.isra.0+0x8d/0x2a0
f2fs_gc+0x2ba/0x8e0 [f2fs]
? _raw_spin_unlock_irqrestore+0x12/0x40
? __timer_delete_sync+0x80/0xe0
? timer_delete_sync+0x14/0x20
? schedule_timeout+0x82/0x100
gc_thread_func+0x38b/0x860 [f2fs]
? gc_thread_func+0x38b/0x860 [f2fs]
? __pfx_autoremove_wake_function+0x10/0x10
kthread+0x10b/0x220
? __pfx_gc_thread_func+0x10/0x10 [f2fs]
? _raw_spin_unlock_irq+0x12/0x40
? __pfx_kthread+0x10/0x10
ret_from_fork+0x11a/0x160
? __pfx_kthread+0x10/0x10
ret_from_fork_asm+0x1a/0x30
</TASK>
This issue is reliably reproducible with the following steps on a
100GB SSD /dev/vdb:
mkfs.f2fs -f /dev/vdb
mount /dev/vdb /mnt/f2fs_test
fio --name=gc --directory=/mnt/f2fs_test --rw=randwrite \
--bs=4k --size=8G --numjobs=12 --fsync=4 --runtime=10 \
--time_based
echo 1 > /sys/fs/f2fs/vdb/gc_urgent
The uninitialized value causes incorrect GC victim selection, leading
to unpredictable garbage collection behavior.
Fix by zero-initializing the entire victim_sel_policy structure to
ensure all fields have defined values.
Fixes: e791d00bd0 ("f2fs: add valid block ratio not to do excessive GC for one time GC")
Cc: stable@kernel.org
Signed-off-by: Xiaole He <hexiaole1994@126.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 1f27ef42bb upstream.
As Hong Yun reported in mailing list:
loop7: detected capacity change from 0 to 131072
------------[ cut here ]------------
kmem_cache of name 'f2fs_xattr_entry-7:7' already exists
WARNING: CPU: 0 PID: 24426 at mm/slab_common.c:110 kmem_cache_sanity_check mm/slab_common.c:109 [inline]
WARNING: CPU: 0 PID: 24426 at mm/slab_common.c:110 __kmem_cache_create_args+0xa6/0x320 mm/slab_common.c:307
CPU: 0 UID: 0 PID: 24426 Comm: syz.7.1370 Not tainted 6.17.0-rc4 #1 PREEMPT(full)
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.13.0-1ubuntu1.1 04/01/2014
RIP: 0010:kmem_cache_sanity_check mm/slab_common.c:109 [inline]
RIP: 0010:__kmem_cache_create_args+0xa6/0x320 mm/slab_common.c:307
Call Trace:
__kmem_cache_create include/linux/slab.h:353 [inline]
f2fs_kmem_cache_create fs/f2fs/f2fs.h:2943 [inline]
f2fs_init_xattr_caches+0xa5/0xe0 fs/f2fs/xattr.c:843
f2fs_fill_super+0x1645/0x2620 fs/f2fs/super.c:4918
get_tree_bdev_flags+0x1fb/0x260 fs/super.c:1692
vfs_get_tree+0x43/0x140 fs/super.c:1815
do_new_mount+0x201/0x550 fs/namespace.c:3808
do_mount fs/namespace.c:4136 [inline]
__do_sys_mount fs/namespace.c:4347 [inline]
__se_sys_mount+0x298/0x2f0 fs/namespace.c:4324
do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline]
do_syscall_64+0x8e/0x3a0 arch/x86/entry/syscall_64.c:94
entry_SYSCALL_64_after_hwframe+0x76/0x7e
The bug can be reproduced w/ below scripts:
- mount /dev/vdb /mnt1
- mount /dev/vdc /mnt2
- umount /mnt1
- mounnt /dev/vdb /mnt1
The reason is if we created two slab caches, named f2fs_xattr_entry-7:3
and f2fs_xattr_entry-7:7, and they have the same slab size. Actually,
slab system will only create one slab cache core structure which has
slab name of "f2fs_xattr_entry-7:3", and two slab caches share the same
structure and cache address.
So, if we destroy f2fs_xattr_entry-7:3 cache w/ cache address, it will
decrease reference count of slab cache, rather than release slab cache
entirely, since there is one more user has referenced the cache.
Then, if we try to create slab cache w/ name "f2fs_xattr_entry-7:3" again,
slab system will find that there is existed cache which has the same name
and trigger the warning.
Let's changes to use global inline_xattr_slab instead of per-sb slab cache
for fixing.
Fixes: a999150f4f ("f2fs: use kmem_cache pool during inline xattr lookups")
Cc: stable@kernel.org
Reported-by: Hong Yun <yhong@link.cuhk.edu.hk>
Tested-by: Hong Yun <yhong@link.cuhk.edu.hk>
Signed-off-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit d33f89b34a upstream.
F2FS can mount filesystems with corrupted directory depth values that
get runtime-clamped to MAX_DIR_HASH_DEPTH. When RENAME_WHITEOUT
operations are performed on such directories, f2fs_rename performs
directory modifications (updating target entry and deleting source
entry) before attempting to add the whiteout entry via f2fs_add_link.
If f2fs_add_link fails due to the corrupted directory structure, the
function returns an error to VFS, but the partial directory
modifications have already been committed to disk. VFS assumes the
entire rename operation failed and does not update the dentry cache,
leaving stale mappings.
In the error path, VFS does not call d_move() to update the dentry
cache. This results in new_dentry still pointing to the old inode
(new_inode) which has already had its i_nlink decremented to zero.
The stale cache causes subsequent operations to incorrectly reference
the freed inode.
This causes subsequent operations to use cached dentry information that
no longer matches the on-disk state. When a second rename targets the
same entry, VFS attempts to decrement i_nlink on the stale inode, which
may already have i_nlink=0, triggering a WARNING in drop_nlink().
Example sequence:
1. First rename (RENAME_WHITEOUT): file2 → file1
- f2fs updates file1 entry on disk (points to inode 8)
- f2fs deletes file2 entry on disk
- f2fs_add_link(whiteout) fails (corrupted directory)
- Returns error to VFS
- VFS does not call d_move() due to error
- VFS cache still has: file1 → inode 7 (stale!)
- inode 7 has i_nlink=0 (already decremented)
2. Second rename: file3 → file1
- VFS uses stale cache: file1 → inode 7
- Tries to drop_nlink on inode 7 (i_nlink already 0)
- WARNING in drop_nlink()
Fix this by explicitly invalidating old_dentry and new_dentry when
f2fs_add_link fails during whiteout creation. This forces VFS to
refresh from disk on subsequent operations, ensuring cache consistency
even when the rename partially succeeds.
Reproducer:
1. Mount F2FS image with corrupted i_current_depth
2. renameat2(file2, file1, RENAME_WHITEOUT)
3. renameat2(file3, file1, 0)
4. System triggers WARNING in drop_nlink()
Fixes: 7e01e7ad74 ("f2fs: support RENAME_WHITEOUT")
Reported-by: syzbot+632cf32276a9a564188d@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=632cf32276a9a564188d
Suggested-by: Chao Yu <chao@kernel.org>
Link: https://lore.kernel.org/all/20251022233349.102728-1-kartikey406@gmail.com/ [v1]
Cc: stable@vger.kernel.org
Signed-off-by: Deepanshu Kartikey <kartikey406@gmail.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit be112e7449 upstream.
In order to let userspace detect such error rather than suffering
silent failure.
Fixes: 4354994f09 ("f2fs: checkpoint disabling")
Cc: stable@kernel.org
Signed-off-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 297baa4aa2 upstream.
Xfstests generic/335, generic/336 sometimes crash with the following message:
F2FS-fs (dm-0): detect filesystem reference count leak during umount, type: 9, count: 1
------------[ cut here ]------------
kernel BUG at fs/f2fs/super.c:1939!
Oops: invalid opcode: 0000 [#1] SMP NOPTI
CPU: 1 UID: 0 PID: 609351 Comm: umount Tainted: G W 6.17.0-rc5-xfstests-g9dd1835ecda5 #1 PREEMPT(none)
Tainted: [W]=WARN
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.16.3-debian-1.16.3-2 04/01/2014
RIP: 0010:f2fs_put_super+0x3b3/0x3c0
Call Trace:
<TASK>
generic_shutdown_super+0x7e/0x190
kill_block_super+0x1a/0x40
kill_f2fs_super+0x9d/0x190
deactivate_locked_super+0x30/0xb0
cleanup_mnt+0xba/0x150
task_work_run+0x5c/0xa0
exit_to_user_mode_loop+0xb7/0xc0
do_syscall_64+0x1ae/0x1c0
entry_SYSCALL_64_after_hwframe+0x76/0x7e
</TASK>
---[ end trace 0000000000000000 ]---
It appears that sometimes it is possible that f2fs_put_super() is called before
all node page reads are completed.
Adding a call to f2fs_wait_on_all_pages() for F2FS_RD_NODE fixes the problem.
Cc: stable@kernel.org
Fixes: 20872584b8 ("f2fs: fix to drop all dirty meta/node pages during umount()")
Signed-off-by: Jan Prusakowski <jprusakowski@google.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 7ee8bc3942 ]
The recent increase in the number of Segment Summary Area (SSA) entries
from 512 to 2048 was an unintentional change in logic of 16kb block
support. This commit corrects the issue.
To better utilize the space available from the erroneous 2048-entry
calculation, we are implementing a solution to share the currently
unused SSA space with neighboring segments. This enhances overall
SSA utilization without impacting the established 8MB segment size.
Fixes: d7e9a9037d ("f2fs: Support Block Size == Page Size")
Signed-off-by: Daeho Jeong <daehojeong@google.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit e462fc48ce ]
The current version missed setting one time GC for normal zoned GC
cycle. So, valid threshold control is not working. Need to fix it to
prevent excessive GC for zoned devices.
Fixes: e791d00bd0 ("f2fs: add valid block ratio not to do excessive GC for one time GC")
Signed-off-by: Daeho Jeong <daehojeong@google.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Assuming the disk layout as below,
disk0: 0 --- 0x00035abfff
disk1: 0x00035ac000 --- 0x00037abfff
disk2: 0x00037ac000 --- 0x00037ebfff
and we want to read data from offset=13568 having len=128 across the block
devices, we can illustrate the block addresses like below.
0 .. 0x00037ac000 ------------------- 0x00037ebfff, 0x00037ec000 -------
| ^ ^ ^
| fofs 0 13568 13568+128
| ------------------------------------------------------
| LBA 0x37e8aa9 0x37ebfa9 0x37ec029
--- map 0x3caa9 0x3ffa9
In this example, we should give the relative map of the target block device
ranging from 0x3caa9 to 0x3ffa9 where the length should be calculated by
0x37ebfff + 1 - 0x37ebfa9.
In the below equation, however, map->m_pblk was supposed to be the original
address instead of the one from the target block address.
- map->m_len = min(map->m_len, dev->end_blk + 1 - map->m_pblk);
Cc: stable@vger.kernel.org
Fixes: 71f2c82062 ("f2fs: multidevice: support direct IO")
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
iput() calls the problematic routine, which does a ->i_count inc/dec
cycle. Undoing it with iput() recurses into the problem.
Note f2fs should not be playing games with the refcount to begin with,
but that will be handled later. Right now solve the immediate
regression.
Fixes: bc986b1d75 ("fs: stop accessing ->i_count directly in f2fs and gfs2")
Reported-by: kernel test robot <oliver.sang@intel.com>
Closes: https://lore.kernel.org/oe-lkp/202509301450.138b448f-lkp@intel.com
Signed-off-by: Mateusz Guzik <mjguzik@gmail.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
This release focuses on two primary updates for Android devices. First, it sets
hash-based file name lookup as the default method to improve performance, while
retaining an option to fall back to a linear lookup. Second, it resolves a
persistent issue with the checkpoint=enable feature. The update further boosts
performance by prefetching node blocks, merging FUA writes more efficiently, and
optimizing block allocation policies.
The release is rounded out by a comprehensive set of bug fixes that address
memory safety, data integrity, and potential system hangs, along with minor
documentation and code clean-ups.
Enhancement:
- add mount option and sysfs entry to tune the lookup mode
- dump more information and add a timeout when enabling/disabling checkpoints
- readahead node blocks in F2FS_GET_BLOCK_PRECACHE mode
- merge FUA command with the existing writes
- allocate HOT_DATA for IPU writes
- Use allocate_section_policy to control write priority in multi-devices setups
- add reserved nodes for privileged users
- Add bggc_io_aware to adjust the priority of BG_GC when issuing IO
- show the list of donation files
Bug fix:
- add missing dput() when printing the donation list
- fix UAF issue in f2fs_merge_page_bio()
- add sanity check on ei.len in __update_extent_tree_range()
- fix infinite loop in __insert_extent_tree()
- fix zero-sized extent for precache extents
- fix to mitigate overhead of f2fs_zero_post_eof_page()
- fix to avoid migrating empty section
- fix to truncate first page in error path of f2fs_truncate()
- fix to update map->m_next_extent correctly in f2fs_map_blocks()
- fix wrong layout information on 16KB page
- fix to do sanity check on node footer for non inode dnode
- fix to avoid NULL pointer dereference in f2fs_check_quota_consistency()
- fix to detect potential corrupted nid in free_nid_list
- fix to clear unusable_cap for checkpoint=enable
- fix to zero data after EOF for compressed file correctly
- fix to avoid overflow while left shift operation
- fix condition in __allow_reserved_blocks()
-----BEGIN PGP SIGNATURE-----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=cIQy
-----END PGP SIGNATURE-----
Merge tag 'f2fs-for-6.18-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk/f2fs
Pull f2fs updates from Jaegeuk Kim:
"This focuses on two primary updates for Android devices.
First, it sets hash-based file name lookup as the default method to
improve performance, while retaining an option to fall back to a
linear lookup.
Second, it resolves a persistent issue with the 'checkpoint=enable'
feature.
The update further boosts performance by prefetching node blocks,
merging FUA writes more efficiently, and optimizing block allocation
policies.
The release is rounded out by a comprehensive set of bug fixes that
address memory safety, data integrity, and potential system hangs,
along with minor documentation and code clean-ups.
Enhancements:
- add mount option and sysfs entry to tune the lookup mode
- dump more information and add a timeout when enabling/disabling
checkpoints
- readahead node blocks in F2FS_GET_BLOCK_PRECACHE mode
- merge FUA command with the existing writes
- allocate HOT_DATA for IPU writes
- Use allocate_section_policy to control write priority in
multi-devices setups
- add reserved nodes for privileged users
- Add bggc_io_aware to adjust the priority of BG_GC when issuing IO
- show the list of donation files
Bug fixes:
- add missing dput() when printing the donation list
- fix UAF issue in f2fs_merge_page_bio()
- add sanity check on ei.len in __update_extent_tree_range()
- fix infinite loop in __insert_extent_tree()
- fix zero-sized extent for precache extents
- fix to mitigate overhead of f2fs_zero_post_eof_page()
- fix to avoid migrating empty section
- fix to truncate first page in error path of f2fs_truncate()
- fix to update map->m_next_extent correctly in f2fs_map_blocks()
- fix wrong layout information on 16KB page
- fix to do sanity check on node footer for non inode dnode
- fix to avoid NULL pointer dereference in
f2fs_check_quota_consistency()
- fix to detect potential corrupted nid in free_nid_list
- fix to clear unusable_cap for checkpoint=enable
- fix to zero data after EOF for compressed file correctly
- fix to avoid overflow while left shift operation
- fix condition in __allow_reserved_blocks()"
* tag 'f2fs-for-6.18-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk/f2fs: (43 commits)
f2fs: add missing dput() when printing the donation list
f2fs: fix UAF issue in f2fs_merge_page_bio()
f2fs: readahead node blocks in F2FS_GET_BLOCK_PRECACHE mode
f2fs: add sanity check on ei.len in __update_extent_tree_range()
f2fs: fix infinite loop in __insert_extent_tree()
f2fs: fix zero-sized extent for precache extents
f2fs: fix to mitigate overhead of f2fs_zero_post_eof_page()
f2fs: fix to avoid migrating empty section
f2fs: fix to truncate first page in error path of f2fs_truncate()
f2fs: fix to update map->m_next_extent correctly in f2fs_map_blocks()
f2fs: fix wrong layout information on 16KB page
f2fs: clean up error handing of f2fs_submit_page_read()
f2fs: avoid unnecessary folio_clear_uptodate() for cleanup
f2fs: merge FUA command with the existing writes
f2fs: allocate HOT_DATA for IPU writes
f2fs: Use allocate_section_policy to control write priority in multi-devices setups
Documentation: f2fs: Reword title
Documentation: f2fs: Indent compression_mode option list
Documentation: f2fs: Wrap snippets in literal code blocks
Documentation: f2fs: Span write hint table section rows
...
We missed to call dput() on the grabbed dentry.
Fixes: f1a49c1b11 ("f2fs: show the list of donation files")
Reviewed-by: Daeho Jeong <daehojeong@google.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
-----BEGIN PGP SIGNATURE-----
iHUEABYKAB0WIQRAhzRXHqcMeLMyaSiRxhvAZXjcogUCaNZQQgAKCRCRxhvAZXjc
oud9AQD5IG4sNnzCjsvcTDpQkbX5eZW+LFIiAiiN+nztZ+OcRQEAvC2N7YovfqM3
TWpVoNDKvEPdtDc9ttFMUKqBZYvxvgE=
=sEaL
-----END PGP SIGNATURE-----
Merge tag 'vfs-6.18-rc1.inode' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs
Pull vfs inode updates from Christian Brauner:
"This contains a series I originally wrote and that Eric brought over
the finish line. It moves out the i_crypt_info and i_verity_info
pointers out of 'struct inode' and into the fs-specific part of the
inode.
So now the few filesytems that actually make use of this pay the price
in their own private inode storage instead of forcing it upon every
user of struct inode.
The pointer for the crypt and verity info is simply found by storing
an offset to its address in struct fsverity_operations and struct
fscrypt_operations. This shrinks struct inode by 16 bytes.
I hope to move a lot more out of it in the future so that struct inode
becomes really just about very core stuff that we need, much like
struct dentry and struct file, instead of the dumping ground it has
become over the years.
On top of this are a various changes associated with the ongoing inode
lifetime handling rework that multiple people are pushing forward:
- Stop accessing inode->i_count directly in f2fs and gfs2. They
simply should use the __iget() and iput() helpers
- Make the i_state flags an enum
- Rework the iput() logic
Currently, if we are the last iput, and we have the I_DIRTY_TIME
bit set, we will grab a reference on the inode again and then mark
it dirty and then redo the put. This is to make sure we delay the
time update for as long as possible
We can rework this logic to simply dec i_count if it is not 1, and
if it is do the time update while still holding the i_count
reference
Then we can replace the atomic_dec_and_lock with locking the
->i_lock and doing atomic_dec_and_test, since we did the
atomic_add_unless above
- Add an icount_read() helper and convert everyone that accesses
inode->i_count directly for this purpose to use the helper
- Expand dump_inode() to dump more information about an inode helping
in debugging
- Add some might_sleep() annotations to iput() and associated
helpers"
* tag 'vfs-6.18-rc1.inode' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs:
fs: add might_sleep() annotation to iput() and more
fs: expand dump_inode()
inode: fix whitespace issues
fs: add an icount_read helper
fs: rework iput logic
fs: make the i_state flags an enum
fs: stop accessing ->i_count directly in f2fs and gfs2
fsverity: check IS_VERITY() in fsverity_cleanup_inode()
fs: remove inode::i_verity_info
btrfs: move verity info pointer to fs-specific part of inode
f2fs: move verity info pointer to fs-specific part of inode
ext4: move verity info pointer to fs-specific part of inode
fsverity: add support for info in fs-specific part of inode
fs: remove inode::i_crypt_info
ceph: move crypt info pointer to fs-specific part of inode
ubifs: move crypt info pointer to fs-specific part of inode
f2fs: move crypt info pointer to fs-specific part of inode
ext4: move crypt info pointer to fs-specific part of inode
fscrypt: add support for info in fs-specific part of inode
fscrypt: replace raw loads of info pointer with helper function
In f2fs_precache_extents(), For large files, It requires reading many
node blocks. Instead of reading each node block with synchronous I/O,
this patch applies readahead so that node blocks can be fetched in
advance.
It reduces the overhead of repeated sync reads and improves efficiency
when precaching extents of large files.
I created a file with the same largest extent and executed the test.
For this experiment, I set the file's largest extent with an offset of 0
and a size of 1GB. I configured the remaining area with 100MB extents.
5GB test file:
dd if=/dev/urandom of=test1 bs=1m count=5120
cp test1 test2
fsync test1
dd if=test1 of=test2 bs=1m skip=1024 seek=1024 count=100 conv=notrunc
dd if=test1 of=test2 bs=1m skip=1224 seek=1224 count=100 conv=notrunc
...
dd if=test1 of=test2 bs=1m skip=5024 seek=5024 count=100 conv=notrunc
reboot
I also created 10GB and 20GB files with large extents using the same
method.
ioctl(F2FS_IOC_PRECACHE_EXTENTS) test results are as follows:
+-----------+---------+---------+-----------+
| File size | Before | After | Reduction |
+-----------+---------+---------+-----------+
| 5GB | 101.8ms | 37.0ms | 72.1% |
| 10GB | 222.9ms | 56.0ms | 74.9% |
| 20GB | 446.2ms | 116.4ms | 73.9% |
+-----------+---------+---------+-----------+
Tested on a 256GB mobile device with an SM8750 chipset.
Reviewed-by: Sungjong Seo <sj1557.seo@samsung.com>
Reviewed-by: Sunmin Jeong <s_min.jeong@samsung.com>
Signed-off-by: Yunji Kang <yunji0.kang@samsung.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Add a sanity check in __update_extent_tree_range() to detect any
zero-sized extent update.
Signed-off-by: wangzijie <wangzijie1@honor.com>
Signed-off-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
When we get wrong extent info data, and look up extent_node in rb tree,
it will cause infinite loop (CONFIG_F2FS_CHECK_FS=n). Avoiding this by
return NULL and print some kernel messages in that case.
Signed-off-by: wangzijie <wangzijie1@honor.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
f2fs_zero_post_eof_page() may cuase more overhead due to invalidate_lock
and page lookup, change as below to mitigate its overhead:
- check new_size before grabbing invalidate_lock
- lookup and invalidate pages only in range of [old_size, new_size]
Fixes: ba8dac350f ("f2fs: fix to zero post-eof page")
Signed-off-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
It reports a bug from device w/ zufs:
F2FS-fs (dm-64): Inconsistent segment (173822) type [1, 0] in SSA and SIT
F2FS-fs (dm-64): Stopped filesystem due to reason: 4
Thread A Thread B
- f2fs_expand_inode_data
- f2fs_allocate_pinning_section
- f2fs_gc_range
- do_garbage_collect w/ segno #x
- writepage
- f2fs_allocate_data_block
- new_curseg
- allocate segno #x
The root cause is: fallocate on pinning file may race w/ block allocation
as above, result in do_garbage_collect() from fallocate() may migrate
segment which is just allocated by a log, the log will update segment type
in its in-memory structure, however GC will get segment type from on-disk
SSA block, once segment type changes by log, we can detect such
inconsistency, then shutdown filesystem.
In this case, on-disk SSA shows type of segno #173822 is 1 (SUM_TYPE_NODE),
however segno #173822 was just allocated as data type segment, so in-memory
SIT shows type of segno #173822 is 0 (SUM_TYPE_DATA).
Change as below to fix this issue:
- check whether current section is empty before gc
- add sanity checks on do_garbage_collect() to avoid any race case, result
in migrating segment used by log.
- btw, it fixes misc issue in printed logs: "SSA and SIT" -> "SIT and SSA".
Fixes: 9703d69d9d ("f2fs: support file pinning for zoned devices")
Cc: Daeho Jeong <daehojeong@google.com>
Signed-off-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
syzbot reports a bug as below:
loop0: detected capacity change from 0 to 40427
F2FS-fs (loop0): Wrong SSA boundary, start(3584) end(4096) blocks(3072)
F2FS-fs (loop0): Can't find valid F2FS filesystem in 1th superblock
F2FS-fs (loop0): invalid crc value
F2FS-fs (loop0): f2fs_convert_inline_folio: corrupted inline inode ino=3, i_addr[0]:0x1601, run fsck to fix.
------------[ cut here ]------------
kernel BUG at fs/inode.c:753!
RIP: 0010:clear_inode+0x169/0x190 fs/inode.c:753
Call Trace:
<TASK>
evict+0x504/0x9c0 fs/inode.c:810
f2fs_fill_super+0x5612/0x6fa0 fs/f2fs/super.c:5047
get_tree_bdev_flags+0x40e/0x4d0 fs/super.c:1692
vfs_get_tree+0x8f/0x2b0 fs/super.c:1815
do_new_mount+0x2a2/0x9e0 fs/namespace.c:3808
do_mount fs/namespace.c:4136 [inline]
__do_sys_mount fs/namespace.c:4347 [inline]
__se_sys_mount+0x317/0x410 fs/namespace.c:4324
do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline]
do_syscall_64+0xfa/0x3b0 arch/x86/entry/syscall_64.c:94
entry_SYSCALL_64_after_hwframe+0x77/0x7f
During f2fs_evict_inode(), clear_inode() detects that we missed to truncate
all page cache before destorying inode, that is because in below path, we
will create page #0 in cache, but missed to drop it in error path, let's fix
it.
- evict
- f2fs_evict_inode
- f2fs_truncate
- f2fs_convert_inline_inode
- f2fs_grab_cache_folio
: create page #0 in cache
- f2fs_convert_inline_folio
: sanity check failed, return -EFSCORRUPTED
- clear_inode detects that inode->i_data.nrpages is not zero
Fixes: 92dffd0179 ("f2fs: convert inline_data when i_size becomes large")
Reported-by: syzbot+90266696fe5daacebd35@syzkaller.appspotmail.com
Closes: https://lore.kernel.org/linux-f2fs-devel/68c09802.050a0220.3c6139.000e.GAE@google.com
Signed-off-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
generic_delete_inode() is rather misleading for what the routine is
doing. inode_just_drop() should be much clearer.
The new naming is inconsistent with generic_drop_inode(), so rename that
one as well with inode_ as the suffix.
No functional changes.
Signed-off-by: Mateusz Guzik <mjguzik@gmail.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Signed-off-by: Christian Brauner <brauner@kernel.org>
Below two functions should never fail, clean up error handling in their
callers:
1) f2fs_grab_read_bio() in f2fs_submit_page_read()
2) bio_add_folio() in f2fs_submit_page_read()
Signed-off-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
In error path of __get_node_folio(), if the folio is not uptodate, let's
avoid unnecessary folio_clear_uptodate() for cleanup.
Signed-off-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Let's split IPU writes in hot data area to improve the GC efficiency.
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Introduces two new sys nodes: allocate_section_hint and
allocate_section_policy. The allocate_section_hint identifies the boundary
between devices, measured in sections; it defaults to the end of the device
for single storage setups, and the end of the first device for multiple
storage setups. The allocate_section_policy determines the write strategy,
with a default value of 0 for normal sequential write strategy. A value of
1 prioritizes writes before the allocate_section_hint, while a value of 2
prioritizes writes after it.
This strategy addresses the issue where, despite F2FS supporting multiple
devices, SOC vendors lack multi-devices support (currently only supporting
zoned devices). As a workaround, multiple storage devices are mapped to a
single dm device. Both this workaround and the F2FS multi-devices solution
may require prioritizing writing to certain devices, such as a device with
better performance or when switching is needed due to performance
degradation near a device's end. For scenarios with more than two devices,
sort them at mount time to utilize this feature.
When using this feature with a single storage device, it has almost no
impact. However, for configurations where multiple storage devices are
mapped to the same dm device using F2FS, utilizing this feature can provide
some optimization benefits. Therefore, I believe it should not be limited
to just multi-devices usage.
Signed-off-by: Liao Yuanhong <liaoyuanhong@vivo.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
As syzbot reported below:
------------[ cut here ]------------
kernel BUG at fs/f2fs/file.c:1243!
Oops: invalid opcode: 0000 [#1] SMP KASAN NOPTI
CPU: 0 UID: 0 PID: 5354 Comm: syz.0.0 Not tainted 6.17.0-rc1-syzkaller-00211-g90d970cade8e #0 PREEMPT(full)
RIP: 0010:f2fs_truncate_hole+0x69e/0x6c0 fs/f2fs/file.c:1243
Call Trace:
<TASK>
f2fs_punch_hole+0x2db/0x330 fs/f2fs/file.c:1306
f2fs_fallocate+0x546/0x990 fs/f2fs/file.c:2018
vfs_fallocate+0x666/0x7e0 fs/open.c:342
ksys_fallocate fs/open.c:366 [inline]
__do_sys_fallocate fs/open.c:371 [inline]
__se_sys_fallocate fs/open.c:369 [inline]
__x64_sys_fallocate+0xc0/0x110 fs/open.c:369
do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline]
do_syscall_64+0xfa/0x3b0 arch/x86/entry/syscall_64.c:94
entry_SYSCALL_64_after_hwframe+0x77/0x7f
RIP: 0033:0x7f1e65f8ebe9
w/ a fuzzed image, f2fs may encounter panic due to it detects inconsistent
truncation range in direct node in f2fs_truncate_hole().
The root cause is: a non-inode dnode may has the same footer.ino and
footer.nid, so the dnode will be parsed as an inode, then ADDRS_PER_PAGE()
may return wrong blkaddr count which may be 923 typically, by chance,
dn.ofs_in_node is equal to 923, then count can be calculated to 0 in below
statement, later it will trigger panic w/ f2fs_bug_on(, count == 0 || ...).
count = min(end_offset - dn.ofs_in_node, pg_end - pg_start);
This patch introduces a new node_type NODE_TYPE_NON_INODE, then allowing
passing the new_type to sanity_check_node_footer in f2fs_get_node_folio()
to detect corruption that a non-inode dnode has the same footer.ino and
footer.nid.
Scripts to reproduce:
mkfs.f2fs -f /dev/vdb
mount /dev/vdb /mnt/f2fs
touch /mnt/f2fs/foo
touch /mnt/f2fs/bar
dd if=/dev/zero of=/mnt/f2fs/foo bs=1M count=8
umount /mnt/f2fs
inject.f2fs --node --mb i_nid --nid 4 --idx 0 --val 5 /dev/vdb
mount /dev/vdb /mnt/f2fs
xfs_io /mnt/f2fs/foo -c "fpunch 6984k 4k"
Cc: stable@kernel.org
Reported-by: syzbot+b9c7ffd609c3f09416ab@syzkaller.appspotmail.com
Closes: https://lore.kernel.org/linux-f2fs-devel/68a68e27.050a0220.1a3988.0002.GAE@google.com
Signed-off-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Move the fsverity_info pointer into the filesystem-specific part of the
inode by adding the field f2fs_inode_info::i_verity_info and configuring
fsverity_operations::inode_info_offs accordingly.
This is a prerequisite for a later commit that removes
inode::i_verity_info, saving memory and improving cache efficiency on
filesystems that don't support fsverity.
Co-developed-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Eric Biggers <ebiggers@kernel.org>
Link: https://lore.kernel.org/20250810075706.172910-11-ebiggers@kernel.org
Signed-off-by: Christian Brauner <brauner@kernel.org>
Move the fscrypt_inode_info pointer into the filesystem-specific part of
the inode by adding the field f2fs_inode_info::i_crypt_info and
configuring fscrypt_operations::inode_info_offs accordingly.
This is a prerequisite for a later commit that removes
inode::i_crypt_info, saving memory and improving cache efficiency with
filesystems that don't support fscrypt.
Co-developed-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Eric Biggers <ebiggers@kernel.org>
Link: https://lore.kernel.org/20250810075706.172910-5-ebiggers@kernel.org
Signed-off-by: Christian Brauner <brauner@kernel.org>
This patch introduces a proc entry to show the currently enrolled donation
files.
- "File path" indicates a file.
- "Status"
a. "Donated" means the file is registed in the donation list by
fadvise(offset, length, POSIX_FADV_NOREUSE)
b. "Evicted" means the donated pages were reclaimed.
- "Offset (kb)" and "Length (kb) show the registered donation range.
- "Cached pages (kb)" shows the amount of cached pages in the inode page cache.
For example,
# of files : 2
File path Status Donation offset (kb) Donation size (kb) File cached size (kb)
---
/local/test2 Donated 0 1048576 2097152
/local/test Evicted 0 1048576 1048576
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Introduce get_left_section_blocks() for cleanup, no logic changes.
Signed-off-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
This patch allows privileged users to reserve nodes via the
'reserve_node' mount option, which is similar to the existing
'reserve_root' option.
"-o reserve_node=<N>" means <N> nodes are reserved for privileged
users only.
Signed-off-by: Chunhai Guo <guochunhai@vivo.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Currently, we have encountered some issues while testing ZUFS. In
situations near the storage limit (e.g., 50GB remaining), and after
simulating fragmentation by repeatedly writing and deleting data, we found
that application installation and startup tests conducted after idling for
a few minutes take significantly longer several times that of traditional
UFS. Tracing the operations revealed that the majority of I/Os were issued
by background GC, which blocks normal I/O operations.
Under normal circumstances, ZUFS indeed requires more background GC and
employs a more aggressive GC strategy. However, I aim to find a way to
minimize the impact on regular I/O operations under these near-limit
conditions. To address this, I have introduced a bggc_io_aware feature,
which controls the prioritization of background GC in the presence of I/Os.
This switch can be adjusted at the framework level to implement different
strategies. If set to AWARE_ALL_IO, all background GC operations will be
skipped during active I/O issuance. The default option remains consistent
with the current strategy, ensuring no change in behavior.
Signed-off-by: Liao Yuanhong <liaoyuanhong@vivo.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Changes as below:
- print more logs for f2fs_{enable,disable}_checkpoint()
- account and dump time stats for f2fs_enable_checkpoint()
Signed-off-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
During f2fs_enable_checkpoint() in remount(), if we flush a large
amount of dirty pages into slow device, it may take long time which
will block write IO, let's add a timeout machanism during dirty
pages flush to avoid long time block in f2fs_enable_checkpoint().
Signed-off-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
As reported, on-disk footer.ino and footer.nid is the same and
out-of-range, let's add sanity check on f2fs_alloc_nid() to detect
any potential corruption in free_nid_list.
Signed-off-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
mount -t f2fs -o checkpoint=disable:10% /dev/vdb /mnt/f2fs/
mount -t f2fs -o remount,checkpoint=enable /dev/vdb /mnt/f2fs/
kernel log:
F2FS-fs (vdb): Adjust unusable cap for checkpoint=disable = 204440 / 10%
If we has assigned checkpoint=enable mount option, unusable_cap{,_perc}
parameters of checkpoint=disable should be reset, then calculation and
log print could be avoid in adjust_unusable_cap_perc().
Fixes: 1ae18f71cb ("f2fs: fix checkpoint=disable:%u%%")
Signed-off-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
