On Saturday 19 March 2011 23:35:40 Jason wrote:
> On Saturday 19 March 2011 18:17:14 Jeff Gilchrist wrote:
> > On Thu, Mar 17, 2011 at 8:32 PM, Jason <ja...@njkfrudils.plus.com> wrote:
> > > Yep , it was just a hunch that was wrong , could you try putting some
> > > printf("here\n");fflush(0);
> > > between he tests in t-constants.c , so we can work out which one is the
> > > problem.
> >
> > It stops inside this section:
> >
> > #ifdef PP_INVERTED
> >
> > {
> >
> > mp_limb_t pp_inverted_calc;
> > invert_limb (pp_inverted_calc, PP);
> > CHECK_LIMB (PP_INVERTED, pp_inverted_calc);
> >
> > }
> >
> > #endif
> >
> > It never completes: invert_limb (pp_inverted_calc, PP);
>
> I know we have a problem with invert_limb on ia64 with icc and no assembler
> , perhaps this is the "same" C-code , I'll have a look in longlong.h
>
> Jason
The problem is in the files mpn/x86_68/k8/udiv.as and umul.as , they are both
wrong. This problem only bites on a AMD 64bit with a non-gcc compiler (not
MSVC) as there is no inline asm . I think the best solution is just to delete
the files as any modern compiler will be able to generate better code anyway
(certainly for umul , and I presume udiv).
Thanks
Jason
--
You received this message because you are subscribed to the Google Groups
"mpir-devel" group.
To post to this group, send email to mpir-devel@googlegroups.com.
To unsubscribe from this group, send email to
mpir-devel+unsubscr...@googlegroups.com.
For more options, visit this group at
http://groups.google.com/group/mpir-devel?hl=en.
