Re: [PATCH v2] Detailed diagnosis when parsing an object name fails.

From: Matthieu Moy <matthieu.moy@grenoble-inp.fr>
Date: Dec 7, 2009, 10:07 UTC
Message-ID: <vpqws0zcd1c.fsf@bauges.imag.fr>
In-Reply-To: <7vk4wz9rbt.fsf@alter.siamese.dyndns.org>

Junio C Hamano <gitster@pobox.com> writes:

> Matthieu Moy <Matthieu.Moy@grenoble-inp.fr> writes:
>
> The second round still had a few compilation warnings (turned to errors in
> my environment) so I fixed them up somewhat.

Thanks. For some reasons, I had CFLAGS=-g and I didn't see the warnings :-(

Show 5 quoted lines

> I stopped before getting a
> clean compile, though---you will still get:
>
>   sha1_name.c: In function 'get_sha1_with_mode_1':
>   sha1_name.c:956: error: 'object_name' may be used uninitialized in this function

This one is strange. I don't git it with gcc 4.4, even with -Wall -Werror. And indeed, object_name is initialized when !gently, which is precisely when we use it.

I did

-               char *object_name;
+               char *object_name = NULL;

to make sure the warning goes away on your machine, but I don't understand what's going on.

>  cache.h     |    4 ++--
>  sha1_name.c |    2 +-

Squashed into v3, together with initialisation of object_name, and two s/int/unsigned/ to please gcc -Wextra.

-- 
Matthieu Moy
http://www-verimag.imag.fr/~moy/