Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <f3fd41e1-ea76-a7d7-4890-ff66ea87b7ab@kernel.org>
Date:   Sun, 19 Jun 2022 08:10:52 +0800
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:91.0) Gecko/20100101
 Thunderbird/91.10.0
Subject: Re: [PATCH 2/2] f2fs-tools: fix to check free space before grow
Content-Language: en-US
To:     qixiaoyu1 <qxy65535@gmail.com>, jaegeuk@kernel.org
Cc:     linux-f2fs-devel@lists.sourceforge.net,
        linux-kernel@vger.kernel.org, qixiaoyu1 <qixiaoyu1@xiaomi.com>
References: <20220614114929.6897-1-qixiaoyu1@xiaomi.com>
 <20220614114929.6897-2-qixiaoyu1@xiaomi.com>
From:   Chao Yu <chao@kernel.org>
In-Reply-To: <20220614114929.6897-2-qixiaoyu1@xiaomi.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Precedence: bulk

On 2022/6/14 19:49, qixiaoyu1 wrote:
> Otherwise, after grow, kernel may report below error message
> when we mount the image if -o parameter is specified during resize:
> 
> F2FS-fs (loop0): invalid crc_offset: 0
> F2FS-fs (loop0): Wrong valid_user_blocks: 16404, user_block_count: 13312
> F2FS-fs (loop0): Failed to get valid F2FS checkpoint
> mount(2) system call failed: Structure needs cleaning.
> 
> Signed-off-by: qixiaoyu1 <qixiaoyu1@xiaomi.com>

It looks this patch should be merged into previous one, otherwise
-o option support is broken for resize.f2fs.

Thanks,

> ---
>   fsck/resize.c | 36 ++++++++++++++++++++++++------------
>   1 file changed, 24 insertions(+), 12 deletions(-)
> 
> diff --git a/fsck/resize.c b/fsck/resize.c
> index d19c6fa..e135b66 100644
> --- a/fsck/resize.c
> +++ b/fsck/resize.c
> @@ -599,6 +599,26 @@ static void rebuild_checkpoint(struct f2fs_sb_info *sbi,
>   	DBG(0, "Info: Done to rebuild checkpoint blocks\n");
>   }
>   
> +static int f2fs_resize_check(struct f2fs_sb_info *sbi, struct f2fs_super_block *new_sb)
> +{
> +	struct f2fs_checkpoint *cp = F2FS_CKPT(sbi);
> +	block_t user_block_count;
> +	unsigned int overprov_segment_count;
> +
> +	overprov_segment_count = (get_newsb(segment_count_main) -
> +			c.new_reserved_segments) *
> +			c.new_overprovision / 100;
> +	overprov_segment_count += c.new_reserved_segments;
> +
> +	user_block_count = (get_newsb(segment_count_main) -
> +			overprov_segment_count) * c.blks_per_seg;
> +
> +	if (get_cp(valid_block_count) > user_block_count)
> +		return -1;
> +
> +	return 0;
> +}
> +
>   static int f2fs_resize_grow(struct f2fs_sb_info *sbi)
>   {
>   	struct f2fs_super_block *sb = F2FS_RAW_SUPER(sbi);
> @@ -616,6 +636,9 @@ static int f2fs_resize_grow(struct f2fs_sb_info *sbi)
>   	if (get_new_sb(new_sb))
>   		return -1;
>   
> +	if (f2fs_resize_check(sbi, new_sb) < 0)
> +		return -1;
> +
>   	/* check nat availability */
>   	if (get_sb(segment_count_nat) > get_newsb(segment_count_nat)) {
>   		err = shrink_nats(sbi, new_sb);
> @@ -659,11 +682,8 @@ static int f2fs_resize_shrink(struct f2fs_sb_info *sbi)
>   	struct f2fs_super_block *sb = F2FS_RAW_SUPER(sbi);
>   	struct f2fs_super_block new_sb_raw;
>   	struct f2fs_super_block *new_sb = &new_sb_raw;
> -	struct f2fs_checkpoint *cp = F2FS_CKPT(sbi);
>   	block_t old_end_blkaddr, old_main_blkaddr;
>   	block_t new_end_blkaddr, new_main_blkaddr, tmp_end_blkaddr;
> -	block_t user_block_count;
> -	unsigned int overprov_segment_count;
>   	unsigned int offset;
>   	int err = -1;
>   
> @@ -674,15 +694,7 @@ static int f2fs_resize_shrink(struct f2fs_sb_info *sbi)
>   	if (get_new_sb(new_sb))
>   		return -1;
>   
> -	overprov_segment_count = (get_newsb(segment_count_main) -
> -			c.new_reserved_segments) *
> -			c.new_overprovision / 100;
> -	overprov_segment_count += c.new_reserved_segments;
> -
> -	user_block_count = (get_newsb(segment_count_main) -
> -			overprov_segment_count) * c.blks_per_seg;
> -
> -	if (get_cp(valid_block_count) > user_block_count)
> +	if (f2fs_resize_check(sbi, new_sb) < 0)
>   		return -1;
>   
>   	/* check nat availability */