1; RUN: llc < %s  -march=x86-64 -mtriple=x86_64-apple-darwin -mcpu=skx --show-mc-encoding | FileCheck %s
2
3define void @f256(<8 x float> %A, <8 x float> %AA, i8* %B, <4 x double> %C, <4 x double> %CC, i32 %D, <4 x i64> %E, <4 x i64> %EE) {
4; CHECK: vmovntps %ymm{{.*}} ## encoding: [0x62
5  %cast = bitcast i8* %B to <8 x float>*
6  %A2 = fadd <8 x float> %A, %AA
7  store <8 x float> %A2, <8 x float>* %cast, align 64, !nontemporal !0
8; CHECK: vmovntdq %ymm{{.*}} ## encoding: [0x62
9  %cast1 = bitcast i8* %B to <4 x i64>*
10  %E2 = add <4 x i64> %E, %EE
11  store <4 x i64> %E2, <4 x i64>* %cast1, align 64, !nontemporal !0
12; CHECK: vmovntpd %ymm{{.*}} ## encoding: [0x62
13  %cast2 = bitcast i8* %B to <4 x double>*
14  %C2 = fadd <4 x double> %C, %CC
15  store <4 x double> %C2, <4 x double>* %cast2, align 64, !nontemporal !0
16  ret void
17}
18
19define void @f128(<4 x float> %A, <4 x float> %AA, i8* %B, <2 x double> %C, <2 x double> %CC, i32 %D, <2 x i64> %E, <2 x i64> %EE) {
20; CHECK: vmovntps %xmm{{.*}} ## encoding: [0x62
21  %cast = bitcast i8* %B to <4 x float>*
22  %A2 = fadd <4 x float> %A, %AA
23  store <4 x float> %A2, <4 x float>* %cast, align 64, !nontemporal !0
24; CHECK: vmovntdq %xmm{{.*}} ## encoding: [0x62
25  %cast1 = bitcast i8* %B to <2 x i64>*
26  %E2 = add <2 x i64> %E, %EE
27  store <2 x i64> %E2, <2 x i64>* %cast1, align 64, !nontemporal !0
28; CHECK: vmovntpd %xmm{{.*}} ## encoding: [0x62
29  %cast2 = bitcast i8* %B to <2 x double>*
30  %C2 = fadd <2 x double> %C, %CC
31  store <2 x double> %C2, <2 x double>* %cast2, align 64, !nontemporal !0
32  ret void
33}
34!0 = !{i32 1}
35