[X86][AVX] Add test case for PR51281

Author: RKSimon

llvm/test/CodeGen/X86/pr51281.ll

@@ -0,0 +1,39 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --no_x86_scrub_mem_shuffle
+; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx | FileCheck %s
+
+; CHECK: .LCPI0_0:
+; CHECK-NEXT: .long   0x00000000
+; CHECK-NEXT: .zero   4
+; CHECK-NEXT: .long   0x3eb5dbc6
+; CHECK-NEXT: .long   0x3eb5dbc6
+; CHECK-NEXT: .zero   4
+; CHECK-NEXT: .long   0x3eb5dbc6
+; CHECK-NEXT: .zero   4
+; CHECK-NEXT: .zero   4
+
+; CHECK: .LCPI0_1:
+; CHECK-NEXT: .long   3
+; CHECK-NEXT: .long   3
+; CHECK-NEXT: .long   1
+; CHECK-NEXT: .long   1
+; CHECK-NEXT: .long   6
+; CHECK-NEXT: .long   7
+; CHECK-NEXT: .long   6
+; CHECK-NEXT: .long   4
+
+define <16 x float> @PR51281(<8 x float> %a0) {
+; CHECK-LABEL: PR51281:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; CHECK-NEXT:    vshufps {{.*#+}} xmm2 = xmm0[1,0],xmm1[3,0]
+; CHECK-NEXT:    vshufps {{.*#+}} xmm1 = xmm2[2,0],xmm1[1,3]
+; CHECK-NEXT:    vmovsldup {{.*#+}} xmm2 = xmm0[0,0,2,2]
+; CHECK-NEXT:    vinsertf128 $1, %xmm1, %ymm2, %ymm1
+; CHECK-NEXT:    vblendps $141, {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %ymm1, %ymm1 # ymm1 = mem[0],ymm1[1],mem[2,3],ymm1[4,5,6],mem[7]
+; CHECK-NEXT:    vperm2f128 {{.*#+}} ymm2 = ymm0[2,3,0,1]
+; CHECK-NEXT:    vblendps {{.*#+}} ymm0 = ymm2[0],ymm0[1],ymm2[2,3,4,5],ymm0[6],ymm2[7]
+; CHECK-NEXT:    vpermilps {{.*#+}} ymm0 = ymm0[3,3,1,1,6,7,6,4]
+; CHECK-NEXT:    retq
+  %shuffle = shufflevector <8 x float> %a0, <8 x float> <float poison, float 0x3FD6BB78C0000000, float poison, float poison, float poison, float poison, float 0.000000e+00, float poison>, <16 x i32> <i32 7, i32 7, i32 1, i32 1, i32 6, i32 3, i32 6, i32 0, i32 14, i32 0, i32 9, i32 9, i32 7, i32 1, i32 5, i32 9>
+  ret <16 x float> %shuffle
+}