This is the mail archive of the
mailing list for the GCC project.
Re: cprop testing
- From: Richard Henderson <rth at redhat dot com>
- To: David Edelsohn <dje at watson dot ibm dot com>
- Cc: gcc-patches at gcc dot gnu dot org
- Date: Sun, 16 Dec 2001 08:48:00 -0800
- Subject: Re: cprop testing
- References: <firstname.lastname@example.org> <200112160538.AAA28638@makai.watson.ibm.com>
On Sun, Dec 16, 2001 at 12:38:00AM -0500, David Edelsohn wrote:
> With the current version of the cprop optimization in CVS, my
> bootstrap on AIX has a miscompare of every object file. If I disable the
> new cprop optimization, things work fine.
Yes, the scheduler has been miscompiled. I've been working on
it, but it's irritatingly difficult to find what went wrong.
> I would like to ask that toplev.c be change to not include this
> optimization algorithm in any default -Ox optimization levels until it is
> fixed. Having AIX bootstrap broken for four days with a known culprit is
> too long.