Re: [PATCH 3/5] btrfs: Do per-chunk degraded check for remount

[Qu Wenruo <quwenruo@cn.fujitsu.com>]

Thanks Anand,

I'm OK with both new patches.

Thanks for the modification.
Qu

Anand Jain wrote on 2015/09/25 16:30 +0800:
> Qu,
>
> Strictly speaking IMO it should be reported to the user on the cli
> terminal, and no logging in required. since its not that easy to get
> that at this point, I am ok with logging it as error. Since we are
> failing the task(mount), error is better.
>
> I have made that change this on top of the patch
>
>    [PATCH 1/1] Btrfs: consolidate btrfs_error() to btrfs_std_error()
>
> and sent them both.
>
> Thanks, Anand
>
>
>> Thanks for pointing this out.
>>
>> I was quite unsure about using btrfs_info/warn/error.
>>
>> In this case, I just wan't to output a dmesg info to let user know
>> exactly what caused the mount failed.
>> Original code output nothing but "failed to open chunk tree", which is
>> quite confusing for end user.
>>
>> I was planning to use btrfs_info, but at least this is really an error
>> message, but only to info user the real cause.
>>
>> Maybe btrfs_warn will be a better choice?
>>
>> Thanks,
>> Qu
>
>