gmp-5.0.4 fails FAIL: t-invert on 2.6.32-5-powerpc64

[Torbjorn Granlund]

I don't understand.

Did you get a linkage error and then a miscomputation?

How did you get past the linkage error?

(The __mull function is not part of GMP.  It is a GCC/system library
issue.)

I am surprised that -mpowerpc64 (or similar) option is not part of the
compiler command.  Did configure attempt to pass any such option?

(I have no powerpc64 system running GNU/Linux, and will not be able to
try to reproduce this problem.)

-- 
Torbjörn