This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
RE: Patch for fortran/16290
- From: "Billinghurst, David (CALCRTS)" <david dot billinghurst at comalco dot riotinto dot com dot au>
- To: "Steven Bosscher" <stevenb at suse dot de>
- Cc: <fortran at gcc dot gnu dot org>,<gcc-patches at gcc dot gnu dot org>
- Date: Thu, 1 Jul 2004 08:33:23 +1000
- Subject: RE: Patch for fortran/16290
> From: Billinghurst, David (CALCRTS)
>> From: Steven Bosscher
>>
>> On Wednesday 30 June 2004 06:57, Billinghurst, David (CALCRTS) wrote:
>>> Test intrinsic_set_exponent.f90 fails on irix6.5 as it is
>>> using exponents smaller than minexponent(x). Fixed thus.
>>>
>>> OK to commit?
>>>
>>> 2004-06-30 David Billinghurst (David.Billinghurst@riotinto.com)
>>>
>>> PR fortran/16290
>>> * gfortran.fortran-torture/execute/intrinsic_set_exponent.f90
>>> Don't use values of n smaller than minexponent(x).
>>
>> I'd suggest you xfail it instead.
>>
>> Gr
>> Steven
>
> I'd rather not. I will see if I think the IRIX behaviour conforms to
> the standards then make a counter offer.
I withdraw this patch. It would be much better to fix set_exponent() for
the cases in question, as there is a legitimate return value. I wonder if
the existing implementation relies on additional intermediate precision?
I won't XFAIL it yet, as a reminder.