emit_x64_vector: Changes to VectorSignedSaturatedDoublingMultiply

* Return both the upper and lower parts of the multiply if required
* SSE2 does not support the pmuldq instruction, do sign correction to an unsigned result instead
* Improve port utilisation where possible (punpck instructions were a bottleneck)
This commit is contained in:
MerryMage
2018-09-15 09:04:19 +01:00
parent 08c0e017a5
commit 06b31448aa
8 changed files with 233 additions and 70 deletions

View File

@@ -432,7 +432,7 @@ bool TranslatorVisitor::SQDMULH_vec_2(bool Q, Imm<2> size, Vec Vm, Vec Vn, Vec V
const IR::U128 operand1 = V(datasize, Vn);
const IR::U128 operand2 = V(datasize, Vm);
const IR::U128 result = ir.VectorSignedSaturatedDoublingMultiplyReturnHigh(esize, operand1, operand2);
const IR::U128 result = ir.VectorSignedSaturatedDoublingMultiply(esize, operand1, operand2).upper;
V(datasize, Vd, result);
return true;