This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: [tree-ssa] New regressions as of 2003-11-04
- From: Zdenek Dvorak <rakdver at atrey dot karlin dot mff dot cuni dot cz>
- To: law at redhat dot com
- Cc: Andrew MacLeod <amacleod at redhat dot com>, Jan Hubicka <jh at suse dot cz>,Diego Novillo <dnovillo at redhat dot com>,Daniel Berlin <dberlin at dberlin dot org>,gcc mailing list <gcc at gcc dot gnu dot org>
- Date: Mon, 10 Nov 2003 22:53:37 +0100
- Subject: Re: [tree-ssa] New regressions as of 2003-11-04
- References: <20031110203548.GA17744@atrey.karlin.mff.cuni.cz> <200311102053.hAAKrbGO020601@speedy.slc.redhat.com>
Hello,
> >> >And finally, I already have it done -- see
> >> >http://gcc.gnu.org/ml/gcc-patches/2003-11/msg00388.html.
> >> And have you compared the timings before and after with your patch?
> >
> >tested compilation of combine.c:
> >
> >without the patch: TOTAL : 16.62 0.91 17.53
> >with the patch: TOTAL : 16.08 0.96 17.04
> >
> >Before you ask -- yes, the results are consistent; 10 consecutive runs
> >without the patch:
> Excellent. Thank you.
>
> FWIW, what I do is I keep a .i file for all the gcc source files and
> time compiling all of those. It's a slightly broader test than picking
> a single source file.
I timed make stage2_build, which should be basically the same (except
for some additional irrelevant overhead); here are the results:
without:
real 16m8.511s
user 15m43.420s
sys 0m18.050s
with:
real 15m55.366s
user 15m30.850s
sys 0m17.800s
Zdenek