This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: 4.0/4.1 PATCH: Fix Solaris 10/x86 bootstrap failure with nativeas
Joseph,
> > This bootstrap failure is a regression from 3.4.
>
> I must doubt that it is a regression given that vanilla 3.4 doesn't
> support Solaris 10 and csl-sol210-3_4-branch doesn't have this patch.
it did since this patch:
2003-09-29 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
* config.gcc (sparc-*-solaris2*): Handle Solaris 10 and up like
Solaris 7-9.
The CSL branch only added amd64 support, which isn't necessary (or useful)
on machines without amd64 processors.
Btw., I've brought up this issue here:
http://gcc.gnu.org/ml/gcc-patches/2004-11/msg00908.html
when it only affected the csl-sol210-branch, but now it affects mainline as
well. I'll try to get gcc working on Solaris 10/amd64 with the native as
later, but this will require help from Sun.
> > Testresults are somewhat reasonable; the fortran failures are largely due
> > to <iso/math_c99.h> defining
> >
> > #define isfinite(x) __builtin_isfinite(x)
> >
> > which gcc doesn't currently support.
>
> The issues with <iso/math_c99.h> have been reported to Sun; they may put
> appropriate __GNUC__ conditionals in it in future. I'll put details of
> all the issues in bug 19933.
Ok, thanks for that. Beyond the libgfortran issue, there have been reports
on mailing lists about missing support for fpclassify with the gcc 3.4.3
included in S10.
> I can't approve the patch for mainline or 4.0 but if it's OK there then
> it's OK for csl-sol210-3_4-branch and I can check it in there if you don't
> have a checkout of that branch.
That would be nice since I'd like to concentrate my limited efforts to
mainline and 4.0.
Rainer
-----------------------------------------------------------------------------
Rainer Orth, Faculty of Technology, Bielefeld University