This is the mail archive of the fortran@gcc.gnu.org mailing list for the GNU Fortran project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: fortran.dg/default_format_denormal_1.f90


On Sun, Nov 16, 2008 at 08:59:19PM -0800, Steve Kargl wrote:
> On Sun, Nov 16, 2008 at 10:45:39PM -0500, Jack Howarth wrote:
> >    Starting with darwin10, the fortran.dg/default_format_denormal_1.f90
> > test case no longer fails on i686-apple-darwin10. Can we change the XFAIL
> > to something like...
> > 
> > Index: testsuite/gfortran.dg/default_format_denormal_1.f90
> > ===================================================================
> > --- testsuite/gfortran.dg/default_format_denormal_1.f90	(revision 141920)
> > +++ testsuite/gfortran.dg/default_format_denormal_1.f90	(working copy)
> > @@ -1,4 +1,4 @@
> > -! { dg-do run { xfail *-apple-darwin* *-*-freebsd* *-*-mingw* } }
> > +! { dg-do run { xfail *-apple-darwin[123456789] *-*-freebsd* *-*-mingw* } }
> >  ! Test XFAILed on these platforms because the system's printf() lacks
> >  ! proper support for denormals.
> >  !
> > 
> > ...for gcc trunk? This should cause darwin10 and later not to XFAIL.
> 
> Is there a 64-bit version of Darwin10?
> 
> PS: the regex is *-apple-darwin[1-9]
> 
> -- 
> Steve

Steve,
   One could build a x86_64-apple-darwin10 target just as one can
build a x86_64-apple-darwin9 target. The issue is rather academic
since if gfortran.dg/default_format_denormal_1.f90 now passes with
a i686-apple-darwin10 compiler at both -m32 and -m64, it should also
pass with a x86_64-apple-darwin10 compiler at -m64 and -m32 as
the same 32-bit and 64-bit routines for printf() are used from libc.
                     Jack
ps I'll post a revised patch to gcc-patches.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]