This is the mail archive of the
fortran@gcc.gnu.org
mailing list for the GNU Fortran project.
Re: nearest_1.f90 failure on i686-linux
- From: Steve Kargl <sgk at troutmask dot apl dot washington dot edu>
- To: Fran?ois-Xavier Coudert <fxcoudert at gmail dot com>
- Cc: Fortran List <fortran at gcc dot gnu dot org>
- Date: Fri, 29 Sep 2006 07:27:56 -0700
- Subject: Re: nearest_1.f90 failure on i686-linux
- References: <19c433eb0609290301p217cb31fwb7c5df4f2bc6bb0b@mail.gmail.com>
On Fri, Sep 29, 2006 at 12:01:53PM +0200, Fran?ois-Xavier Coudert wrote:
> Hi Steve,
>
> I see nearest_1.f90 failures at -O levels >= 1 on i686-linux, with a
> GMP 4.2.0 and MPFR 2.2.0. The last trees produced by -fdump-tree-all
> at -O0 and -O1 levels are attached for comparison. Is this supposed to
> happen with you recent patch for MPFR version checking, or is there a
> bug?
>
> I can give you additional info if you wish, just tell me what to look for.
>
Jerry and I have been looking at the problem. AFAICT,
the test fails on i686-linux due to the extra precision
of the 80-bit registers. Jerry confirmed the test passes
if one add -ffloat-store.
--
Steve