PATCH: libffi vs. SPARC (again)

[Boehm, Hans]

> -----Original Message-----
> From: Jeff Sturm [mailto:jsturm@one-point.com]
> That's ugly, for a library that's intended to be portable.  I think we
> could declare the return value "long" instead and it should 
> work almost
> anywhere.
...except MIPS N32 and win64 and possibly some others.

Is ptrdiff_t a better choice?  Hopefully that's defined to be "long long" on
win64.  That leaves MIPS N32.  (I would guess the HP/UX 32 bit ABI on
Itanium behaves like MIPS N32 in this regard.  The same may or may not be
true for the 32 bit ABI on recent PA-RISC machines.  I'm not sure.  None of
these are currently supported by  libffi.)

> It'd be better I guess if libffi supplied a 
> typedef for this.
Yes.

> 
> > Types that are equal/larger than int will be 
> > stored in the native way. As long as all 64-bit CPUs are 
> little endian 
> > then there is no problem! Or am I missing something?
> 
> Hans pointed out that sparc64 is big-endian.  (There may be 
> others... does
> anyone run ia64 in big-endian mode?)
HP/UX on Itanium is big-endian.  It's currently not supported by libffi,
though it would be nice if that changed eventually.

Hans