https://gcc.gnu.org/bugzilla/show_bug.cgi?id=123841

--- Comment #30 from Iain Sandoe <iains at gcc dot gnu.org> ---
(In reply to [email protected] from comment #29)
> > --- Comment #28 from Iain Sandoe <iains at gcc dot gnu.org> ---
> > (In reply to Rainer Orth from comment #27)
> >> Created attachment 63525 [details]
> >> Revised patch
> >> 
> >> I'd started testing this revised patch which does the suggested merge, only
> >> handling the *-*-darwin*:no (native as) case.
> >
> > the revised patch is not correct for powerpc* darwin.
> >
> > powerpc64-apple-darwin* should have `-arch ppc64`
> >
> > I fully accept that these are legacy platforms, but we have always had a 
> > policy
> > of not breaking them if possible (even though they are very low priority
> > maintenance-wise).
> 
> Fully agreed.  However, on the gcc-15 branch there's
> 
>   powerpc*-*-darwin*)
>     dnl Always pass -arch ppc to assembler.
>     gcc_cv_as_flags="-arch ppc"
>     ;;
> 
> No ppc64 in sight.
> 
> I can certainly correct this, but for once this isn't my fault.

indeed - the configuration does not seem to account it (which supports the
observation that we probably do not do any arch-specific configuration probes
so far)

... Note the gcc specs must get it right or we'd not be able to build the 64b
multilib.

Given that we're changing this - we might as well correct the configure too?

Reply via email to