On Wed 23-09-20 21:11:49, Zhang Xiaoxu wrote:
> If some errors has occurred on the device, and the orphan list not empty,
> then mount the device with 'ro', the bdev write error check will failed:
> ext4_check_bdev_write_error:193: comm mount: Error while async write back metadata
>
> Since the sbi->s_bdev_wb_err wouldn't be initialized when mount file system
> with 'ro', when clean up the orphan list and access the iloc buffer, bdev
> write error check will failed.
>
> So we should always initialize the sbi->s_bdev_wb_err even if mount the
> file system with 'ro'.
>
> Fixes: bc71726c7257 ("ext4: abort the filesystem if failed to async write metadata buffer")
> Signed-off-by: Zhang Xiaoxu <zhangxiaoxu5@xxxxxxxxxx>
Thanks for the patch! Good catch! I just think you should now remove the
errseq_check_and_advance() call in ext4_remount() because it isn't needed
anymore.
Honza
> diff --git a/fs/ext4/super.c b/fs/ext4/super.c
> index ea425b49b345..086439889869 100644
> --- a/fs/ext4/super.c
> +++ b/fs/ext4/super.c
> @@ -4814,9 +4814,8 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
> * used to detect the metadata async write error.
> */
> spin_lock_init(&sbi->s_bdev_wb_lock);
> - if (!sb_rdonly(sb))
> - errseq_check_and_advance(&sb->s_bdev->bd_inode->i_mapping->wb_err,
> - &sbi->s_bdev_wb_err);
> + errseq_check_and_advance(&sb->s_bdev->bd_inode->i_mapping->wb_err,
> + &sbi->s_bdev_wb_err);
> sb->s_bdev->bd_super = sb;
> EXT4_SB(sb)->s_mount_state |= EXT4_ORPHAN_FS;
> ext4_orphan_cleanup(sb, es);
> --
> 2.25.4
>
--
Jan Kara <jack@xxxxxxxx>
SUSE Labs, CR
