From 1ac40fdf4d0bbb8c9d6a86328c49f0ec2bd588e1 Mon Sep 17 00:00:00 2001
From: Simon Pilgrim <llvm-dev@redking.me.uk>
Date: Wed, 27 Mar 2019 11:21:09 +0000
Subject: [PATCH] [X86][SSE] Add shuffle test case for PR41249

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@357062 91177308-0d34-0410-b5e6-96231b3b80d8
---
 test/CodeGen/X86/vector-shuffle-128-v4.ll | 37 +++++++++++++++++++++++
 1 file changed, 37 insertions(+)

diff --git a/test/CodeGen/X86/vector-shuffle-128-v4.ll b/test/CodeGen/X86/vector-shuffle-128-v4.ll
index 211c57c63ac..d4bb1b2fa68 100644
--- a/test/CodeGen/X86/vector-shuffle-128-v4.ll
+++ b/test/CodeGen/X86/vector-shuffle-128-v4.ll
@@ -2339,6 +2339,43 @@ define <4 x i32> @insert_dup_mem_v4i32(i32* %ptr) {
   ret <4 x i32> %tmp2
 }
 
+; PR41249
+define <4 x float> @shuffle_mem_pmovzx_v4f32(<2 x float>* %p0, <4 x float>* %p1) {
+; SSE-LABEL: shuffle_mem_pmovzx_v4f32:
+; SSE:       # %bb.0:
+; SSE-NEXT:    movsd {{.*#+}} xmm0 = mem[0],zero
+; SSE-NEXT:    xorps %xmm1, %xmm1
+; SSE-NEXT:    movaps %xmm0, %xmm2
+; SSE-NEXT:    unpcklps {{.*#+}} xmm2 = xmm2[0],xmm1[0],xmm2[1],xmm1[1]
+; SSE-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,0,0,0]
+; SSE-NEXT:    movaps %xmm2, (%rsi)
+; SSE-NEXT:    retq
+;
+; AVX1-LABEL: shuffle_mem_pmovzx_v4f32:
+; AVX1:       # %bb.0:
+; AVX1-NEXT:    vmovsd {{.*#+}} xmm0 = mem[0],zero
+; AVX1-NEXT:    vxorps %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vunpcklps {{.*#+}} xmm1 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; AVX1-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[0,0,0,0]
+; AVX1-NEXT:    vmovaps %xmm1, (%rsi)
+; AVX1-NEXT:    retq
+;
+; AVX2OR512VL-LABEL: shuffle_mem_pmovzx_v4f32:
+; AVX2OR512VL:       # %bb.0:
+; AVX2OR512VL-NEXT:    vmovsd {{.*#+}} xmm0 = mem[0],zero
+; AVX2OR512VL-NEXT:    vxorps %xmm1, %xmm1, %xmm1
+; AVX2OR512VL-NEXT:    vunpcklps {{.*#+}} xmm1 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; AVX2OR512VL-NEXT:    vbroadcastss %xmm0, %xmm0
+; AVX2OR512VL-NEXT:    vmovaps %xmm1, (%rsi)
+; AVX2OR512VL-NEXT:    retq
+  %1 = load <2 x float>, <2 x float>* %p0
+  %2 = shufflevector <2 x float> %1, <2 x float> undef, <4 x i32> <i32 undef, i32 undef, i32 0, i32 1>
+  %3 = shufflevector <4 x float> %2, <4 x float> <float undef, float undef, float 0.000000e+00, float 0.000000e+00>, <4 x i32> <i32 2, i32 6, i32 3, i32 7>
+  %4 = shufflevector <2 x float> %1, <2 x float> undef, <4 x i32> zeroinitializer
+  store <4 x float> %3, <4 x float>* %p1
+  ret <4 x float> %4
+}
+
 ;
 ; Shuffle to logical bit shifts
 ;
-- 
2.40.0