xref: /llvm-project/llvm/test/CodeGen/X86/2012-01-12-extract-sv.ll (revision 69a322fed19b977d15be9500d8653496b73673e9)
1; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
2; RUN: llc < %s -mattr=+avx -mtriple=i686-pc-win32 | FileCheck %s --check-prefixes=AVX1
3; RUN: llc < %s -mattr=+avx2 -mtriple=i686-pc-win32 | FileCheck %s --check-prefixes=AVX2
4
5define void @endless_loop() {
6; AVX1-LABEL: endless_loop:
7; AVX1:       # %bb.0: # %entry
8; AVX1-NEXT:    vmovaps (%eax), %xmm0
9; AVX1-NEXT:    vshufps {{.*#+}} xmm1 = xmm0[0,0,0,0]
10; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm1
11; AVX1-NEXT:    vxorps %xmm2, %xmm2, %xmm2
12; AVX1-NEXT:    vblendps {{.*#+}} ymm1 = ymm2[0,1,2,3,4,5,6],ymm1[7]
13; AVX1-NEXT:    vxorps %xmm2, %xmm2, %xmm2
14; AVX1-NEXT:    vblendps {{.*#+}} xmm0 = xmm0[0],xmm2[1,2,3]
15; AVX1-NEXT:    vmovaps %ymm0, (%eax)
16; AVX1-NEXT:    vmovaps %ymm1, (%eax)
17; AVX1-NEXT:    vzeroupper
18; AVX1-NEXT:    retl
19;
20; AVX2-LABEL: endless_loop:
21; AVX2:       # %bb.0: # %entry
22; AVX2-NEXT:    vbroadcastss (%eax), %xmm0
23; AVX2-NEXT:    vxorps %xmm1, %xmm1, %xmm1
24; AVX2-NEXT:    vblendps {{.*#+}} xmm1 = xmm0[0],xmm1[1,2,3]
25; AVX2-NEXT:    vbroadcastss %xmm0, %ymm0
26; AVX2-NEXT:    vxorps %xmm2, %xmm2, %xmm2
27; AVX2-NEXT:    vblendps {{.*#+}} ymm0 = ymm2[0,1,2,3,4,5,6],ymm0[7]
28; AVX2-NEXT:    vmovaps %ymm0, (%eax)
29; AVX2-NEXT:    vmovaps %ymm1, (%eax)
30; AVX2-NEXT:    vzeroupper
31; AVX2-NEXT:    retl
32entry:
33  %0 = load <8 x i32>, ptr addrspace(1) undef, align 32
34  %1 = shufflevector <8 x i32> %0, <8 x i32> undef, <16 x i32> <i32 4, i32 4, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
35  %2 = shufflevector <16 x i32> <i32 undef, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 undef>, <16 x i32> %1, <16 x i32> <i32 16, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 17>
36  store <16 x i32> %2, ptr addrspace(1) undef, align 64
37  ret void
38}
39