Re: [Evolution-hackers] gcc 4.4 may be causing a number of bugs in Evolution

[Gilles Dartiguelongue <gilles dartiguelongue esiee org>]

Le 1 févr. 2010 à 17:52, Jeffrey Stedfast a écrit :

> 
> This weekend I discovered a particularly nasty bug in gcc 4.4 where gcc
> would mistakenly optimize out important sections of code
> when it encountered a particular trick used in a ton of places inside
> Evolution (EDList and pretty much everywhere custom single-linked lists
> are used inside at least Camel and likely other places as well).
> 
> A temporary solution is to pass the -fno-strict-aliasing argument to gcc.
> 
> Unfortunately, the gcc developers claim that this is not a bug:
> http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42907

I'm sure a lot of subscribers already know, but this has been a "problem" since at least gcc 3.4 (iirc) where it started complaining about strict aliasing. If you let gcc build non compliant code, it'll indeed optimize the code wrongly. gcc just got a lot less fool proof in the last releases in order to force developers into coding more easily optimizable code (probably not the way they would put it but that's the wanted result afaik).

-- 
Gilles Dartiguelongue
gilles dartiguelongue esiee org