[PATCH] D90822: [X86] use macros to split GFNI intrinsics into different kinds

Wed Nov 4 23:46:52 PST 2020

craig.topper added a comment.

In D90822#2375463 <https://reviews.llvm.org/D90822#2375463>, @FreddyYe wrote:

> In D90822#2375423 <https://reviews.llvm.org/D90822#2375423>, @pengfei wrote:
>
>> Craig's method sounds good.
>> @FreddyYe , Why we check AVX512BW instead of AVX512F. I saw SDM says it depends on AVX512F.
>
> I was referring to the old implement and test case. Seems like a error introduced before.

mmask64 requires avx512bw. And 32xi16 and 64xi8 aren’t well supported without avx512bw. We used to split them always, but we only split specific instructions now.

We also need avx512bw for the selectb and selectw builtins used for masking.

Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D90822/new/

https://reviews.llvm.org/D90822