[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <8e16a65d-bca2-c95a-aac5-0ba5411695ed@huawei.com>
Date: Mon, 10 May 2021 11:50:39 +0800
From: Zhiqiang Liu <liuzhiqiang26@...wei.com>
To: Haotian Li <lihaotian9@...wei.com>,
Ext4 Developers List <linux-ext4@...r.kernel.org>
CC: "Theodore Y. Ts'o" <tytso@....edu>,
linfeilong <linfeilong@...wei.com>
Subject: Re: [PATCH] e2fsck: try write_primary_superblock() again when it
failed
friendly ping...
On 2021/4/13 11:19, Haotian Li wrote:
> Function write_primary_superblock() has two ways to flush
> superblock, byte-by-byte as default. It may use
> io_channel_write_byte() many times. If some errors occur
> during these funcs, the superblock may become inconsistent
> and produce checksum error.
>
> Try write_primary_superblock() with whole-block way again
> when it failed on byte-by-byte way.
> ---
> lib/ext2fs/closefs.c | 4 +---
> 1 file changed, 1 insertion(+), 3 deletions(-)
>
> diff --git a/lib/ext2fs/closefs.c b/lib/ext2fs/closefs.c
> index 69cbdd8c..1fc27fb5 100644
> --- a/lib/ext2fs/closefs.c
> +++ b/lib/ext2fs/closefs.c
> @@ -223,10 +223,8 @@ static errcode_t write_primary_superblock(ext2_filsys fs,
> retval = io_channel_write_byte(fs->io,
> SUPERBLOCK_OFFSET + (2 * write_idx), size,
> new_super + write_idx);
> - if (retval == EXT2_ET_UNIMPLEMENTED)
> - goto fallback;
> if (retval)
> - return retval;
> + goto fallback;
> }
> memcpy(fs->orig_super, super, SUPERBLOCK_SIZE);
> return 0;
Powered by blists - more mailing lists