Re: [f2fs-dev] [PATCH 5/5] f2fs: enforce fsync_mode=strict for renamed directory

[Chao Yu]

On 2018/4/25 13:46, Jaegeuk Kim wrote:
> This is to give a option for user to be able to recover B/foo in the below
> case.
> 
> mkdir A
> sync()
> rename(A, B)
> creat (B/foo)
> fsync (B/foo)
> ---crash---

That makes sense, IMO, it will be better to cover cross rename case as well?

Thanks,

> 
> Sugessted-by: Velayudhan Pillai <vijay@xxxxxxxxxxxxx>
> Signed-off-by: Jaegeuk Kim <jaegeuk@xxxxxxxxxx>
> ---
>  fs/f2fs/namei.c | 5 ++++-
>  1 file changed, 4 insertions(+), 1 deletion(-)
> 
> diff --git a/fs/f2fs/namei.c b/fs/f2fs/namei.c
> index b5f404674cad..fef6e3ab2135 100644
> --- a/fs/f2fs/namei.c
> +++ b/fs/f2fs/namei.c
> @@ -973,8 +973,11 @@ static int f2fs_rename(struct inode *old_dir, struct dentry *old_dentry,
>  			f2fs_put_page(old_dir_page, 0);
>  		f2fs_i_links_write(old_dir, false);
>  	}
> -	if (F2FS_OPTION(sbi).fsync_mode == FSYNC_MODE_STRICT)
> +	if (F2FS_OPTION(sbi).fsync_mode == FSYNC_MODE_STRICT) {
>  		add_ino_entry(sbi, new_dir->i_ino, TRANS_DIR_INO);
> +		if (S_ISDIR(old_inode->i_mode))
> +			add_ino_entry(sbi, old_inode->i_ino, TRANS_DIR_INO);
> +	}
>  
>  	f2fs_unlock_op(sbi);
>  
>