1 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
2 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+sse4.1 | FileCheck %s --check-prefix=SSE
3 ; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx | FileCheck %s --check-prefix=AVX
5 define <8 x i16> @blend_packusdw(<4 x i32> %a0, <4 x i32> %a1, <4 x i32> %a2, <4 x i32> %a3) {
6 ; SSE-LABEL: blend_packusdw:
8 ; SSE-NEXT: packusdw %xmm2, %xmm0
11 ; AVX-LABEL: blend_packusdw:
13 ; AVX-NEXT: vpackusdw %xmm2, %xmm0, %xmm0
15 %p0 = call <8 x i16> @llvm.x86.sse41.packusdw(<4 x i32> %a0, <4 x i32> %a1)
16 %p1 = call <8 x i16> @llvm.x86.sse41.packusdw(<4 x i32> %a2, <4 x i32> %a3)
17 %s0 = shufflevector <8 x i16> %p0, <8 x i16> %p1, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 8, i32 9, i32 10, i32 11>
21 define <16 x i8> @blend_packuswb(<8 x i16> %a0, <8 x i16> %a1, <8 x i16> %a2, <8 x i16> %a3) {
22 ; SSE-LABEL: blend_packuswb:
24 ; SSE-NEXT: packuswb %xmm2, %xmm0
27 ; AVX-LABEL: blend_packuswb:
29 ; AVX-NEXT: vpackuswb %xmm2, %xmm0, %xmm0
31 %p0 = call <16 x i8> @llvm.x86.sse2.packuswb.128(<8 x i16> %a0, <8 x i16> %a1)
32 %p1 = call <16 x i8> @llvm.x86.sse2.packuswb.128(<8 x i16> %a2, <8 x i16> %a3)
33 %s0 = shufflevector <16 x i8> %p0, <16 x i8> %p1, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 16, i32 17, i32 18, i32 19, i32 20, i32 21, i32 22, i32 23>
37 define <8 x i16> @blend_packusdw_packuswb(<4 x i32> %a0, <4 x i32> %a1, <8 x i16> %a2, <8 x i16> %a3) {
38 ; SSE-LABEL: blend_packusdw_packuswb:
40 ; SSE-NEXT: packusdw %xmm0, %xmm0
41 ; SSE-NEXT: packuswb %xmm2, %xmm2
42 ; SSE-NEXT: punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm2[0]
45 ; AVX-LABEL: blend_packusdw_packuswb:
47 ; AVX-NEXT: vpackusdw %xmm0, %xmm0, %xmm0
48 ; AVX-NEXT: vpackuswb %xmm2, %xmm2, %xmm1
49 ; AVX-NEXT: vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
51 %p0 = call <8 x i16> @llvm.x86.sse41.packusdw(<4 x i32> %a0, <4 x i32> %a1)
52 %p1 = call <16 x i8> @llvm.x86.sse2.packuswb.128(<8 x i16> %a2, <8 x i16> %a3)
53 %b1 = bitcast <16 x i8> %p1 to <8 x i16>
54 %s0 = shufflevector <8 x i16> %p0, <8 x i16> %b1, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 8, i32 9, i32 10, i32 11>
58 declare <16 x i8> @llvm.x86.sse2.packuswb.128(<8 x i16>, <8 x i16>)
59 declare <8 x i16> @llvm.x86.sse41.packusdw(<4 x i32>, <4 x i32>)