[f2fs-dev] [PATCH v2] f2fs: invalidate block device page cache on umount

[f2fs-dev] [PATCH v2] f2fs: invalidate block device page cache on umount

[Yongpeng Yang]

From: Yongpeng Yang <yangyongpeng@xiaomi.com>

Neither F2FS nor VFS invalidates the block device page cache, which
results in reading stale metadata. An example scenario is shown below:

Terminal A                  Terminal B
mount /dev/vdb /mnt/f2fs
touch mx // ino = 4
sync
dump.f2fs -i 4 /dev/vdb// block on "[Y/N]"
                            touch mx2 // ino = 5
                            sync
                            umount /mnt/f2fs
                            dump.f2fs -i 5 /dev/vdb // block addr is 0

After umount, the block device page cache is not purged, causing
`dump.f2fs -i 5 /dev/vdb` to read stale metadata and see inode 5 with
block address 0.

Btrfs has encountered a similar issue before, the solution there was to
call sync_blockdev() and invalidate_bdev() when the device is closed:

mail-archive.com/linux-btrfs@vger.kernel.org/msg54188.html

For the root user, the f2fs kernel calls sync_blockdev() on umount to
flush all cached data to disk, and f2fs-tools can release the page cache
by issuing ioctl(fd, BLKFLSBUF) when accessing the device. However,
non-root users are not permitted to drop the page cache, and may still
observe stale data.

This patch calls sync_blockdev() and invalidate_bdev() during umount to
invalidate the block device page cache, thereby preventing stale
metadata from being read.

Note that this may result in an extra sync_blockdev() call on the first
device, in both f2fs_put_super() and kill_block_super(). The second call
do nothing, as there are no dirty pages left to flush. It ensures that
non-root users do not observe stale data.

Signed-off-by: Yongpeng Yang <yangyongpeng@xiaomi.com>
---
v2:
- Add commit message explaining why sync_blockdev is called redundantly
for the first device.
---
 fs/f2fs/super.c | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c
index 40079fd7886b..b624ed253a95 100644
--- a/fs/f2fs/super.c
+++ b/fs/f2fs/super.c
@@ -2088,6 +2088,12 @@ static void f2fs_put_super(struct super_block *sb)
 #if IS_ENABLED(CONFIG_UNICODE)
 	utf8_unload(sb->s_encoding);
 #endif
+	sync_blockdev(sb->s_bdev);
+	invalidate_bdev(sb->s_bdev);
+	for (i = 1; i < sbi->s_ndevs; i++) {
+		sync_blockdev(FDEV(i).bdev);
+		invalidate_bdev(FDEV(i).bdev);
+	}
 }
 
 int f2fs_sync_fs(struct super_block *sb, int sync)
-- 
2.43.0



_______________________________________________
Linux-f2fs-devel mailing list
Linux-f2fs-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linux-f2fs-devel

Re: [f2fs-dev] [PATCH v2] f2fs: invalidate block device page cache on umount

[Chao Yu via Linux-f2fs-devel]

On 3/24/26 17:47, Yongpeng Yang wrote:
> From: Yongpeng Yang <yangyongpeng@xiaomi.com>
> 
> Neither F2FS nor VFS invalidates the block device page cache, which
> results in reading stale metadata. An example scenario is shown below:
> 
> Terminal A                  Terminal B
> mount /dev/vdb /mnt/f2fs
> touch mx // ino = 4
> sync
> dump.f2fs -i 4 /dev/vdb// block on "[Y/N]"
>                             touch mx2 // ino = 5
>                             sync
>                             umount /mnt/f2fs
>                             dump.f2fs -i 5 /dev/vdb // block addr is 0
> 
> After umount, the block device page cache is not purged, causing
> `dump.f2fs -i 5 /dev/vdb` to read stale metadata and see inode 5 with
> block address 0.
> 
> Btrfs has encountered a similar issue before, the solution there was to
> call sync_blockdev() and invalidate_bdev() when the device is closed:
> 
> mail-archive.com/linux-btrfs@vger.kernel.org/msg54188.html
> 
> For the root user, the f2fs kernel calls sync_blockdev() on umount to
> flush all cached data to disk, and f2fs-tools can release the page cache
> by issuing ioctl(fd, BLKFLSBUF) when accessing the device. However,
> non-root users are not permitted to drop the page cache, and may still
> observe stale data.
> 
> This patch calls sync_blockdev() and invalidate_bdev() during umount to
> invalidate the block device page cache, thereby preventing stale
> metadata from being read.
> 
> Note that this may result in an extra sync_blockdev() call on the first
> device, in both f2fs_put_super() and kill_block_super(). The second call
> do nothing, as there are no dirty pages left to flush. It ensures that
> non-root users do not observe stale data.
> 
> Signed-off-by: Yongpeng Yang <yangyongpeng@xiaomi.com>

Reviewed-by: Chao Yu <chao@kernel.org>

Thanks,


_______________________________________________
Linux-f2fs-devel mailing list
Linux-f2fs-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linux-f2fs-devel

Re: [f2fs-dev] [PATCH v2] f2fs: invalidate block device page cache on umount

[patchwork-bot+f2fs--- via Linux-f2fs-devel]

Hello:

This patch was applied to jaegeuk/f2fs.git (dev)
by Jaegeuk Kim <jaegeuk@kernel.org>:

On Tue, 24 Mar 2026 17:47:08 +0800 you wrote:
> From: Yongpeng Yang <yangyongpeng@xiaomi.com>
> 
> Neither F2FS nor VFS invalidates the block device page cache, which
> results in reading stale metadata. An example scenario is shown below:
> 
> Terminal A                  Terminal B
> mount /dev/vdb /mnt/f2fs
> touch mx // ino = 4
> sync
> dump.f2fs -i 4 /dev/vdb// block on "[Y/N]"
>                             touch mx2 // ino = 5
>                             sync
>                             umount /mnt/f2fs
>                             dump.f2fs -i 5 /dev/vdb // block addr is 0
> 
> [...]

Here is the summary with links:
  - [f2fs-dev,v2] f2fs: invalidate block device page cache on umount
    https://git.kernel.org/jaegeuk/f2fs/c/8979bc3d2a25

You are awesome, thank you!
-- 
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html




_______________________________________________
Linux-f2fs-devel mailing list
Linux-f2fs-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linux-f2fs-devel