[Openmcl-devel] What is the status of OpenMCL Altivec support

[Gary Byers]

On Fri, 20 Feb 2004, Gary King wrote:

> I'm interesting in trying to vectorize some algorithms and am wondering
> if anyone has used it to do anything interesting that they would be
> willing to share with the community.
>
> Thanks,
> --
> Gary Warren King, Lab Manager
> EKSL East, University of Massachusetts * 413 577 0176
>

What support there is is pretty much limited to the fact that the
assembler supports AltiVec instructions.  (Someone reported a bug in
the way that immediate operands were encoded in some vector instuctions
not too long ago, suggesting that they may have been trying to actually
use AltiVec.  I don't know exactly what they were doing or how far they
got.)

Aside from the fact that whatever you do's pretty much limited to what
you can do in LAP, it's hard to use AltiVec instructions with lisp
data in OpenMCL: lisp objects are allocated on 64-bit address boundaries,
and AltiVec wants to deal with data that's 128-bit aligned.  There's
a "vector permute" operation that can help to deal with misaligned
vector data, but whether something's aligned on a 128-bit boundary
or not can change from instruction to instruction (because of GC.)