This is the mail archive of the
fortran@gcc.gnu.org
mailing list for the GNU Fortran project.
Re: -ffast-math and gfortran
- From: Steve Kargl <sgk at troutmask dot apl dot washington dot edu>
- To: Jack Howarth <howarth at bromo dot msbb dot uc dot edu>
- Cc: toon at moene dot indiv dot nluug dot nl, fortran at gcc dot gnu dot org, fxcoudert at gmail dot com
- Date: Sun, 4 Jun 2006 13:03:09 -0700
- Subject: Re: -ffast-math and gfortran
- References: <20060604191239.80E7910028@bromo.msbb.uc.edu>
On Sun, Jun 04, 2006 at 03:12:39PM -0400, Jack Howarth wrote:
> Toon,
> What I would like to see though is some agreement on what sort of
> fortran code currently is failing with -ffast-math that shouldn't be.
> Even if we can't resolve these issues immediately, I think it would
> be a really good idea to have testcases for such breakage with
> -ffast-math introduced into the gfortran testsuite. At the very
> least this would dramatically increase the number of eyes on the
> problem.
I disagree with placing -ffast-math testcases in the gfortran
testsuite. It is already fairly large and those of us with
slow machines and who run the testsuite often can't afford
bloat.
I've answered way too many emails concerning the broken
behavior of -ffast-math. It has gotten to the point that I'm
considering disabling -ffast-math in favor of -frisky-math
in gfortran.
--
Steve