equivalence really broken :-)
Andrew Pinski
pinskia@physics.uc.edu
Fri Oct 14 17:13:00 GMT 2005
On Oct 14, 2005, at 12:40 PM, Paul Thomas wrote:
> François-Xavier,
>
>>
>> PS: we should really work on an automated regtesting procedure that
>> includes code that isn't in the testsuite (mainly for copyright
>> reasons)
>>
> I agree - Polyhedron, Meissner and NIST all come to mind as candidates.
SPEC also comes to mind, even though most people don't have access to
that.
SPEC was also broken the same way as seen on
http://people.redhat.com/dnovillo/spec2000/gcc/
-- Pinski
More information about the Fortran
mailing list