On 2019/1/17 下午10:08, Johannes Thumshirn wrote:
> On 17/01/2019 14:42, Qu Wenruo wrote:
>>
>> To show when the error happens.
>>
>> This patch only handles the read time error, thus we have two same error
>> message.
>>
>> But the next patch will introduce new write time validation check, thus
>> we need to distinguish then by the extra message.
>
> OK, so if we really want the same message printed for all read time
> errors, how about this:
>
> if (found_level == 0 && btrfs_check_leaf_full(fs_info, eb)) {
> set_bit(EXTENT_BUFFER_CORRUPT, &eb->bflags);
> ret = -EIO;
> }
>
> if (found_level > 0 && btrfs_check_node(fs_info, eb))
> ret = -EIO;
>
> + if (ret == -EIO)
> + btrfs_err(fs_info,
> + "read time tree block corruption detected");
> + else if (!ret)
> - if (!ret)
> set_extent_buffer_uptodate(eb);
>
>
> So we really only have one message printed and not duplicating the prints.
Works for me, I'll use this one in next version.
Thanks,
Qu
>
> Byte,
> Johannes
>
