suyu/dynarmic
28
1
Fork 1
Code Issues Pull requests Projects Releases Packages Wiki Activity

emit_x64_vector: Simplify AVX-512 codepath in EmitVectorMultiply64

Browse source 
I realized I introduced a helper for simple AVX operation emitting, so
use that instead of writing it all out long-form.
This commit is contained in:
Lioncash 2018-05-14 23:02:38 -04:00 committed by MerryMage
parent cb456f914b
commit 054549da35
1 changed files with 1 additions and 7 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

8
src/backend_x64/emit_x64_vector.cpp
View file

@ -1460,13 +1460,7 @@ void EmitX64::EmitVectorMultiply32(EmitContext& ctx, IR::Inst* inst) {
void EmitX64::EmitVectorMultiply64(EmitContext& ctx, IR::Inst* inst) {
if (code.DoesCpuSupport(Xbyak::util::Cpu::tAVX512DQ) && code.DoesCpuSupport(Xbyak::util::Cpu::tAVX512VL)) {
auto args = ctx.reg_alloc.GetArgumentInfo(inst);
const Xbyak::Xmm a = ctx.reg_alloc.UseScratchXmm(args[0]);
const Xbyak::Xmm b = ctx.reg_alloc.UseXmm(args[1]);
code.vpmullq(a, a, b);
ctx.reg_alloc.DefineValue(inst, a);
EmitAVXVectorOperation(code, ctx, inst, &Xbyak::CodeGenerator::vpmullq);
return;
}