Re: [PATCH] btrfs-progs: image: fix compiler warning

From: Tsutomu Itoh <t-itoh@jp.fujitsu.com>
To: dsterba@suse.cz
Cc: linux-btrfs@vger.kernel.org
Subject: Re: [PATCH] btrfs-progs: image: fix compiler warning
Date: Fri, 7 Oct 2016 11:40:45 +0900	[thread overview]
Message-ID: <07aaf0c2-4130-c92a-c17d-29214d56cb31@jp.fujitsu.com> (raw)
In-Reply-To: <20161006160045.GK6576@twin.jikos.cz>

On 2016/10/07 1:00, David Sterba wrote:
> On Wed, Oct 05, 2016 at 05:07:48PM +0900, Tsutomu Itoh wrote:
>> In v4.8-rc1, gcc 5.3.1 gives following warning. Fixed it.
>>
>>     [CC]     btrfs-image.o
>> btrfs-image.c: In function 'flush_pending':
>> btrfs-image.c:708:17: warning: 'start' may be used uninitialized in this function [-Wmaybe-uninitialized]
>>   header->bytenr = cpu_to_le64(start);
>>                  ^
>> btrfs-image.c:927:6: note: 'start' was declared here
>>   u64 start;
>>       ^
> 
> So the patch makes the compiler warning go away, but is the code
> correct? AFAICS, the warning points to the case where flush_pending is
> called with done=1 (from create_metadump) and there's zero
> md->pending_size . Are you sure this is an expected state and that the
> function can proceed with state = 0 ?
> 

I think that this is a case where some errors occurred before calling
flush_pending.
Therefore, create_metadump returns the error to the caller. (creating
the image fails, I think.)

Thanks,
Tsutomu