Re: What's cooking in git.git (Dec 2012, #05; Tue, 18)

From: Jeff King <peff@peff.net>
Date: Dec 20, 2012, 14:59 UTC
Message-ID: <20121220145941.GC27211@sigill.intra.peff.net>
In-Reply-To: <7v4njjf6fk.fsf@alter.siamese.dyndns.org>

On Tue, Dec 18, 2012 at 12:31:43PM -0800, Junio C Hamano wrote:

Show 9 quoted lines

> * jk/error-const-return (2012-12-15) 2 commits
>  - silence some -Wuninitialized false positives
>  - make error()'s constant return value more visible
> 
>  Help compilers' flow analysis by making it more explicit that
>  error() always returns -1, to reduce false "variable used
>  uninitialized" warnings.
> 
>  This is still an RFC.

What's your opinion on this? The last round I posted (and what you have in pu) is about as good as this technique is going to get. So it is really a taste thing. On the one hand, I really like that it is not just papering over the problem, but rather giving the compiler more data to do its analysis. On the other hand, it makes the source kind of ugly.

I don't think there's a rush on it, but I'm just sifting through my half-done topics.

-Peff