This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: gcc 3.2.1 schedule?
- From: Janis Johnson <janis187 at us dot ibm dot com>
- To: Ulrich Weigand <weigand at immd1 dot informatik dot uni-erlangen dot de>
- Cc: gcc at gcc dot gnu dot org
- Date: Fri, 1 Nov 2002 08:56:10 -0800
- Subject: Re: gcc 3.2.1 schedule?
- References: <200210312238.XAA11214@faui11.informatik.uni-erlangen.de>
On Thu, Oct 31, 2002 at 11:38:41PM +0100, Ulrich Weigand wrote:
> Joe Buck wrote:
>
> >Now I'm confused. Doesn't stage 1 (where, in your case, 2.95.3 is being
> >built for the first time by 3.2) run without optimization at all? If so,
> >how can -fno-strict-aliasing have an effect?
>
> Hmm, good question. The case I had been debugging was a bit different
> in that it was a non-bootstrapped build of a gcc 2.x cross-compiler
> by a 3.x compiler. In that case the build did use optimization (-O2 IIRC).
>
> Are you sure the stage 1 builds run without optimization? (I'm currently
> on vacation and don't have the gcc sources installed here, so I can't
> readily check ...)
Paul asked if I had used STAGE1_CFLAGS when I built and saw the same
failure, and I'm fairly sure I had not. I'm bootstrapping today's 3.2
branch right now on i686-pc-linux-gnu and have a 3.2 compiler available
for comparison. I can try whatever bootstraps of 2.95.3 that would
provide useful information.
Janis