Displaying 1 result from an estimated 1 matches for "simd_mul".
2014 Sep 03
2
[LLVMdev] Questions on the llvm 'vector' types and resulting SIMD instructions
If I generate IR using 'vector' types, for example, if my code assembles IR like this:
define <4 x float> @simd_mul(<4 x float>, <4 x float>) {
%3 = fmul <4 x float> %0, %1
ret <4 x float> %3
}
I assume that when I JIT, it will generates the best SIMD instructions available on the host it's running on? For example, when running on a machine supporting SSE, it does see...