Bignum support still broken on powerpc (ping: Jerry James)

[robert delius royar]

Mon, 10 Apr 2006 (16:35 -0600 UTC) Jerry James wrote:

> robert delius royar <xemacs at frinabulax.org> wrote:
>> OK, I simply removed the 'if (__GMP_ULONG_MAX <= GMP_NUMB_MASK)' from line
>> 1595 of gmp.h.  That stopped the warnings.  Now, my loadup/dump fails with the
>> same type error
>> Requiring disp-table.el...*** Error in XEmacs initialization
>> (wrong-type-argument integer-or-char-p 225)
>
> I wonder if you are triggering a similar bug to the icc bug that bit me.
> Let's see.  I am attaching the test case I sent Intel a couple of years
> ago.  You'll have to edit the Makefile a bit, but try it out and see if
> the test program reports that you are getting bignum results.  You
> shouldn't be.

I mistyped in my last reply.  I did NOT see "a bignum" in the output, 
and I should NOT have.  So it looks like the problem is something to do 
with my --enable-union-type setting.  On the G5, if I remove that, I get 
a working bignum-enabled XEmacs.  I haven't tried again on my G4, but 
the compilation I tried last summer (gcc4.00) failed with any 
configuration using bignum.  The G5 uses gcc4.01, which warns infewer 
places and fails less often than did gcc4.00.

-- 
Dr. Robert Delius Royar                   Associate Professor of English
Morehead State University                             Morehead, Kentucky