FAIL: t-perfpow, PPC darwin 9.8, gcc-4.2, ABI=mode32

[depp at moria.us]

Using gmp-5.0.1, the t-perfpow test fails.  No other test fails.  This  
test passes with ABI=mode64 and with ABI=32.  I wiped the compilation  
directory and recompiled in case the error was due to a memory error.   
No such luck.  Remind me to get ECC with my next system.  I *think*  
this test passed when I compiled GMP 4.2.4 for ABI=mode64, but that  
was a while ago.

$ ./configure CC=gcc-4.2 ABI=mode32
$ make
$ make check
n = 451620....omitted....50801
n was destroyed, but perfpow_p still believes n is a perfect power
/bin/sh: line 1: 12276 Abort trap              ${dir}$tst
FAIL: t-perfpow
$ uname -a
Darwin Gondor.local 9.8.0 Darwin Kernel Version 9.8.0: Wed Jul 15  
16:57:01 PDT 2009; root:xnu-1228.15.4~1/RELEASE_PPC Power Macintosh
$ gcc-4.2 --version
powerpc-apple-darwin9-gcc-4.2.1 (GCC) 4.2.1 (Apple Inc. build 5566)
Copyright (C) 2007 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There  
is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR  
PURPOSE.

I'm using gcc-4.2 because gcc-4.0 breaks GMP on OS X (it turns out  
that Apple backported some behavior from a later GCC, which breaks  
things that expect it to behave like GCC 4.0).  It would be nice if I  
could get 64-bit limbs and 32-bit pointers working together someday.   
In the meantime, I'll just use 32-bit limbs.  I can't go 64-bit  
because GHC doesn't support it on PPC/darwin.  Good news is five years  
from now, nobody will care if this bug ever got fixed.  Probably sooner.

--Dietrich Epp