[X86][SSE] Added v4f32 shuffle with zero tests

This is mainly test cases for improvements to insertps matching, but pre-SSE41 shuffles could be improved as well

llvm-svn: 256705

1 files changed, 139 insertions, 0 deletions

diff --git a/llvm/test/CodeGen/X86/vector-shuffle-128-v4.ll b/llvm/test/CodeGen/X86/vector-shuffle-128-v4.ll
index 35c3b708fd0..53dbb32235a 100644
--- a/llvm/test/CodeGen/X86/vector-shuffle-128-v4.ll
+++ b/llvm/test/CodeGen/X86/vector-shuffle-128-v4.ll
@@ -1005,6 +1005,145 @@ define <4 x float> @shuffle_v4f32_u051(<4 x float> %a, <4 x float> %b) {
   ret <4 x float> %shuffle
 }
 
+define <4 x float> @shuffle_v4f32_0zz4(<4 x float> %a, <4 x float> %b) {
+; SSE2-LABEL: shuffle_v4f32_0zz4:
+; SSE2:       # BB#0:
+; SSE2-NEXT:    xorps %xmm2, %xmm2
+; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,0],xmm2[2,0]
+; SSE2-NEXT:    shufps {{.*#+}} xmm2 = xmm2[0,1],xmm1[2,0]
+; SSE2-NEXT:    movss {{.*#+}} xmm2 = xmm0[0],xmm2[1,2,3]
+; SSE2-NEXT:    movaps %xmm2, %xmm0
+; SSE2-NEXT:    retq
+;
+; SSE3-LABEL: shuffle_v4f32_0zz4:
+; SSE3:       # BB#0:
+; SSE3-NEXT:    xorps %xmm2, %xmm2
+; SSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,0],xmm2[2,0]
+; SSE3-NEXT:    shufps {{.*#+}} xmm2 = xmm2[0,1],xmm1[2,0]
+; SSE3-NEXT:    movss {{.*#+}} xmm2 = xmm0[0],xmm2[1,2,3]
+; SSE3-NEXT:    movaps %xmm2, %xmm0
+; SSE3-NEXT:    retq
+;
+; SSSE3-LABEL: shuffle_v4f32_0zz4:
+; SSSE3:       # BB#0:
+; SSSE3-NEXT:    xorps %xmm2, %xmm2
+; SSSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,0],xmm2[2,0]
+; SSSE3-NEXT:    shufps {{.*#+}} xmm2 = xmm2[0,1],xmm1[2,0]
+; SSSE3-NEXT:    movss {{.*#+}} xmm2 = xmm0[0],xmm2[1,2,3]
+; SSSE3-NEXT:    movaps %xmm2, %xmm0
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: shuffle_v4f32_0zz4:
+; SSE41:       # BB#0:
+; SSE41-NEXT:    insertps {{.*#+}} xmm1 = zero,zero,zero,xmm1[0]
+; SSE41-NEXT:    blendps {{.*#+}} xmm0 = xmm0[0],xmm1[1,2,3]
+; SSE41-NEXT:    retq
+;
+; AVX-LABEL: shuffle_v4f32_0zz4:
+; AVX:       # BB#0:
+; AVX-NEXT:    vinsertps {{.*#+}} xmm1 = zero,zero,zero,xmm1[0]
+; AVX-NEXT:    vblendps {{.*#+}} xmm0 = xmm0[0],xmm1[1,2,3]
+; AVX-NEXT:    retq
+  %shuffle = shufflevector <4 x float> %b, <4 x float> zeroinitializer, <4 x i32> <i32 undef, i32 5, i32 6, i32 0>
+  %shuffle1 = shufflevector <4 x float> %a, <4 x float> %shuffle, <4 x i32> <i32 0, i32 5, i32 6, i32 7>
+  ret <4 x float> %shuffle1
+}
+
+define <4 x float> @shuffle_v4f32_0zz6(<4 x float> %a, <4 x float> %b) {
+; SSE2-LABEL: shuffle_v4f32_0zz6:
+; SSE2:       # BB#0:
+; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,2]
+; SSE2-NEXT:    xorps %xmm1, %xmm1
+; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[1,2],xmm0[0,3]
+; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[2,0,1,3]
+; SSE2-NEXT:    movaps %xmm1, %xmm0
+; SSE2-NEXT:    retq
+;
+; SSE3-LABEL: shuffle_v4f32_0zz6:
+; SSE3:       # BB#0:
+; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,2]
+; SSE3-NEXT:    xorps %xmm1, %xmm1
+; SSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[1,2],xmm0[0,3]
+; SSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[2,0,1,3]
+; SSE3-NEXT:    movaps %xmm1, %xmm0
+; SSE3-NEXT:    retq
+;
+; SSSE3-LABEL: shuffle_v4f32_0zz6:
+; SSSE3:       # BB#0:
+; SSSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,2]
+; SSSE3-NEXT:    xorps %xmm1, %xmm1
+; SSSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[1,2],xmm0[0,3]
+; SSSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[2,0,1,3]
+; SSSE3-NEXT:    movaps %xmm1, %xmm0
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: shuffle_v4f32_0zz6:
+; SSE41:       # BB#0:
+; SSE41-NEXT:    insertps {{.*#+}} xmm0 = xmm0[0],zero,zero,xmm1[2]
+; SSE41-NEXT:    xorps %xmm1, %xmm1
+; SSE41-NEXT:    blendps {{.*#+}} xmm0 = xmm0[0],xmm1[1,2],xmm0[3]
+; SSE41-NEXT:    retq
+;
+; AVX-LABEL: shuffle_v4f32_0zz6:
+; AVX:       # BB#0:
+; AVX-NEXT:    vinsertps {{.*#+}} xmm0 = xmm0[0],zero,zero,xmm1[2]
+; AVX-NEXT:    vxorps %xmm1, %xmm1, %xmm1
+; AVX-NEXT:    vblendps {{.*#+}} xmm0 = xmm0[0],xmm1[1,2],xmm0[3]
+; AVX-NEXT:    retq
+  %shuffle = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 0, i32 undef, i32 undef, i32 6>
+  %shuffle1 = shufflevector <4 x float> zeroinitializer, <4 x float> %shuffle, <4 x i32> <i32 4, i32 1, i32 2, i32 7>
+  ret <4 x float> %shuffle1
+}
+
+define <4 x float> @shuffle_v4f32_0z24(<4 x float> %a, <4 x float> %b) {
+; SSE2-LABEL: shuffle_v4f32_0z24:
+; SSE2:       # BB#0:
+; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,0],xmm0[2,0]
+; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,0]
+; SSE2-NEXT:    xorps %xmm1, %xmm1
+; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[1,0],xmm0[0,0]
+; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[2,0],xmm0[2,3]
+; SSE2-NEXT:    movaps %xmm1, %xmm0
+; SSE2-NEXT:    retq
+;
+; SSE3-LABEL: shuffle_v4f32_0z24:
+; SSE3:       # BB#0:
+; SSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,0],xmm0[2,0]
+; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,0]
+; SSE3-NEXT:    xorps %xmm1, %xmm1
+; SSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[1,0],xmm0[0,0]
+; SSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[2,0],xmm0[2,3]
+; SSE3-NEXT:    movaps %xmm1, %xmm0
+; SSE3-NEXT:    retq
+;
+; SSSE3-LABEL: shuffle_v4f32_0z24:
+; SSSE3:       # BB#0:
+; SSSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,0],xmm0[2,0]
+; SSSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,0]
+; SSSE3-NEXT:    xorps %xmm1, %xmm1
+; SSSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[1,0],xmm0[0,0]
+; SSSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[2,0],xmm0[2,3]
+; SSSE3-NEXT:    movaps %xmm1, %xmm0
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: shuffle_v4f32_0z24:
+; SSE41:       # BB#0:
+; SSE41-NEXT:    insertps {{.*#+}} xmm0 = xmm0[0],zero,xmm0[2],xmm1[0]
+; SSE41-NEXT:    xorps %xmm1, %xmm1
+; SSE41-NEXT:    blendps {{.*#+}} xmm0 = xmm0[0],xmm1[1],xmm0[2,3]
+; SSE41-NEXT:    retq
+;
+; AVX-LABEL: shuffle_v4f32_0z24:
+; AVX:       # BB#0:
+; AVX-NEXT:    vinsertps {{.*#+}} xmm0 = xmm0[0],zero,xmm0[2],xmm1[0]
+; AVX-NEXT:    vxorps %xmm1, %xmm1, %xmm1
+; AVX-NEXT:    vblendps {{.*#+}} xmm0 = xmm0[0],xmm1[1],xmm0[2,3]
+; AVX-NEXT:    retq
+  %shuffle = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 0, i32 undef, i32 2, i32 4>
+  %shuffle1 = shufflevector <4 x float> zeroinitializer, <4 x float> %shuffle, <4 x i32> <i32 4, i32 1, i32 6, i32 7>
+  ret <4 x float> %shuffle1
+}
+
 define <4 x i32> @shuffle_v4i32_4zzz(<4 x i32> %a) {
 ; SSE2-LABEL: shuffle_v4i32_4zzz:
 ; SSE2:       # BB#0: