[x86] Add some important, missing test coverage for blending from one

vector to a zero vector for the v2 cases and fix the v4 integer cases to
actually blend from a vector.

There are already seprate tests for the case of inserting from a scalar.

These cases cover a lot of the regressions I've seen in the regression
test suite for the new vector shuffle lowering and specifically cover
the reported lack of using various zext-ing instruction patterns. My
next patch should fix a big chunk of this, but wanted to get a nice
baseline for these patterns in the test cases first.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@218976 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/test/CodeGen/X86/vector-shuffle-128-v2.ll b/test/CodeGen/X86/vector-shuffle-128-v2.ll
index c0bdc06138a219a4a8128b01bda08722a9f68d5f..1fc1b287ef7b75269daf14163033c2c6a85a2149 100644 (file)
--- a/test/CodeGen/X86/vector-shuffle-128-v2.ll
+++ b/test/CodeGen/X86/vector-shuffle-128-v2.ll
@@ -651,6 +651,227 @@ define <2 x i64> @shuffle_v2i64_31_copy(<2 x i64> %nonce, <2 x i64> %a, <2 x i64
   ret <2 x i64> %shuffle
 }
 
+define <2 x i64> @shuffle_v2i64_0z(<2 x i64> %a) {
+; SSE2-LABEL: shuffle_v2i64_0z:
+; SSE2:       # BB#0:
+; SSE2-NEXT:    xorpd %xmm1, %xmm1
+; SSE2-NEXT:    shufpd {{.*#+}} xmm0 = xmm0[0],xmm1[1]
+; SSE2-NEXT:    retq
+;
+; SSE3-LABEL: shuffle_v2i64_0z:
+; SSE3:       # BB#0:
+; SSE3-NEXT:    xorpd %xmm1, %xmm1
+; SSE3-NEXT:    shufpd {{.*#+}} xmm0 = xmm0[0],xmm1[1]
+; SSE3-NEXT:    retq
+;
+; SSSE3-LABEL: shuffle_v2i64_0z:
+; SSSE3:       # BB#0:
+; SSSE3-NEXT:    xorpd %xmm1, %xmm1
+; SSSE3-NEXT:    shufpd {{.*#+}} xmm0 = xmm0[0],xmm1[1]
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: shuffle_v2i64_0z:
+; SSE41:       # BB#0:
+; SSE41-NEXT:    pxor %xmm1, %xmm1
+; SSE41-NEXT:    pblendw {{.*#+}} xmm0 = xmm0[0,1,2,3],xmm1[4,5,6,7]
+; SSE41-NEXT:    retq
+;
+; AVX1-LABEL: shuffle_v2i64_0z:
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpblendw {{.*#+}} xmm0 = xmm0[0,1,2,3],xmm1[4,5,6,7]
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: shuffle_v2i64_0z:
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX2-NEXT:    vpblendd {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,3]
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <2 x i64> %a, <2 x i64> zeroinitializer, <2 x i32> <i32 0, i32 3>
+  ret <2 x i64> %shuffle
+}
+
+define <2 x i64> @shuffle_v2i64_1z(<2 x i64> %a) {
+; SSE-LABEL: shuffle_v2i64_1z:
+; SSE:       # BB#0:
+; SSE-NEXT:    pxor %xmm1, %xmm1
+; SSE-NEXT:    punpckhqdq {{.*#+}} xmm0 = xmm0[1],xmm1[1]
+; SSE-NEXT:    retq
+;
+; AVX-LABEL: shuffle_v2i64_1z:
+; AVX:       # BB#0:
+; AVX-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX-NEXT:    vpunpckhqdq {{.*#+}} xmm0 = xmm0[1],xmm1[1]
+; AVX-NEXT:    retq
+  %shuffle = shufflevector <2 x i64> %a, <2 x i64> zeroinitializer, <2 x i32> <i32 1, i32 3>
+  ret <2 x i64> %shuffle
+}
+
+define <2 x i64> @shuffle_v2i64_z0(<2 x i64> %a) {
+; SSE-LABEL: shuffle_v2i64_z0:
+; SSE:       # BB#0:
+; SSE-NEXT:    pxor %xmm1, %xmm1
+; SSE-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
+; SSE-NEXT:    movdqa %xmm1, %xmm0
+; SSE-NEXT:    retq
+;
+; AVX-LABEL: shuffle_v2i64_z0:
+; AVX:       # BB#0:
+; AVX-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm1[0],xmm0[0]
+; AVX-NEXT:    retq
+  %shuffle = shufflevector <2 x i64> %a, <2 x i64> zeroinitializer, <2 x i32> <i32 2, i32 0>
+  ret <2 x i64> %shuffle
+}
+
+define <2 x i64> @shuffle_v2i64_z1(<2 x i64> %a) {
+; SSE2-LABEL: shuffle_v2i64_z1:
+; SSE2:       # BB#0:
+; SSE2-NEXT:    xorpd %xmm1, %xmm1
+; SSE2-NEXT:    shufpd {{.*#+}} xmm1 = xmm1[0],xmm0[1]
+; SSE2-NEXT:    movapd %xmm1, %xmm0
+; SSE2-NEXT:    retq
+;
+; SSE3-LABEL: shuffle_v2i64_z1:
+; SSE3:       # BB#0:
+; SSE3-NEXT:    xorpd %xmm1, %xmm1
+; SSE3-NEXT:    shufpd {{.*#+}} xmm1 = xmm1[0],xmm0[1]
+; SSE3-NEXT:    movapd %xmm1, %xmm0
+; SSE3-NEXT:    retq
+;
+; SSSE3-LABEL: shuffle_v2i64_z1:
+; SSSE3:       # BB#0:
+; SSSE3-NEXT:    xorpd %xmm1, %xmm1
+; SSSE3-NEXT:    shufpd {{.*#+}} xmm1 = xmm1[0],xmm0[1]
+; SSSE3-NEXT:    movapd %xmm1, %xmm0
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: shuffle_v2i64_z1:
+; SSE41:       # BB#0:
+; SSE41-NEXT:    pxor %xmm1, %xmm1
+; SSE41-NEXT:    pblendw {{.*#+}} xmm1 = xmm1[0,1,2,3],xmm0[4,5,6,7]
+; SSE41-NEXT:    movdqa %xmm1, %xmm0
+; SSE41-NEXT:    retq
+;
+; AVX1-LABEL: shuffle_v2i64_z1:
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpblendw {{.*#+}} xmm0 = xmm1[0,1,2,3],xmm0[4,5,6,7]
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: shuffle_v2i64_z1:
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX2-NEXT:    vpblendd {{.*#+}} xmm0 = xmm1[0,1],xmm0[2,3]
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <2 x i64> %a, <2 x i64> zeroinitializer, <2 x i32> <i32 2, i32 1>
+  ret <2 x i64> %shuffle
+}
+
+define <2 x double> @shuffle_v2f64_0z(<2 x double> %a) {
+; SSE2-LABEL: shuffle_v2f64_0z:
+; SSE2:       # BB#0:
+; SSE2-NEXT:    xorpd %xmm1, %xmm1
+; SSE2-NEXT:    shufpd {{.*#+}} xmm0 = xmm0[0],xmm1[1]
+; SSE2-NEXT:    retq
+;
+; SSE3-LABEL: shuffle_v2f64_0z:
+; SSE3:       # BB#0:
+; SSE3-NEXT:    xorpd %xmm1, %xmm1
+; SSE3-NEXT:    shufpd {{.*#+}} xmm0 = xmm0[0],xmm1[1]
+; SSE3-NEXT:    retq
+;
+; SSSE3-LABEL: shuffle_v2f64_0z:
+; SSSE3:       # BB#0:
+; SSSE3-NEXT:    xorpd %xmm1, %xmm1
+; SSSE3-NEXT:    shufpd {{.*#+}} xmm0 = xmm0[0],xmm1[1]
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: shuffle_v2f64_0z:
+; SSE41:       # BB#0:
+; SSE41-NEXT:    xorpd %xmm1, %xmm1
+; SSE41-NEXT:    blendpd {{.*#+}} xmm0 = xmm0[0],xmm1[1]
+; SSE41-NEXT:    retq
+;
+; AVX-LABEL: shuffle_v2f64_0z:
+; AVX:       # BB#0:
+; AVX-NEXT:    vxorpd %xmm1, %xmm1, %xmm1
+; AVX-NEXT:    vblendpd {{.*#+}} xmm0 = xmm0[0],xmm1[1]
+; AVX-NEXT:    retq
+  %shuffle = shufflevector <2 x double> %a, <2 x double> zeroinitializer, <2 x i32> <i32 0, i32 3>
+  ret <2 x double> %shuffle
+}
+
+define <2 x double> @shuffle_v2f64_1z(<2 x double> %a) {
+; SSE-LABEL: shuffle_v2f64_1z:
+; SSE:       # BB#0:
+; SSE-NEXT:    xorpd %xmm1, %xmm1
+; SSE-NEXT:    unpckhpd {{.*#+}} xmm0 = xmm0[1],xmm1[1]
+; SSE-NEXT:    retq
+;
+; AVX-LABEL: shuffle_v2f64_1z:
+; AVX:       # BB#0:
+; AVX-NEXT:    vxorpd %xmm1, %xmm1, %xmm1
+; AVX-NEXT:    vunpckhpd {{.*#+}} xmm0 = xmm0[1],xmm1[1]
+; AVX-NEXT:    retq
+  %shuffle = shufflevector <2 x double> %a, <2 x double> zeroinitializer, <2 x i32> <i32 1, i32 3>
+  ret <2 x double> %shuffle
+}
+
+define <2 x double> @shuffle_v2f64_z0(<2 x double> %a) {
+; SSE-LABEL: shuffle_v2f64_z0:
+; SSE:       # BB#0:
+; SSE-NEXT:    xorpd %xmm1, %xmm1
+; SSE-NEXT:    unpcklpd {{.*#+}} xmm1 = xmm1[0],xmm0[0]
+; SSE-NEXT:    movapd %xmm1, %xmm0
+; SSE-NEXT:    retq
+;
+; AVX-LABEL: shuffle_v2f64_z0:
+; AVX:       # BB#0:
+; AVX-NEXT:    vxorpd %xmm1, %xmm1, %xmm1
+; AVX-NEXT:    vunpcklpd {{.*#+}} xmm0 = xmm1[0],xmm0[0]
+; AVX-NEXT:    retq
+  %shuffle = shufflevector <2 x double> %a, <2 x double> zeroinitializer, <2 x i32> <i32 2, i32 0>
+  ret <2 x double> %shuffle
+}
+
+define <2 x double> @shuffle_v2f64_z1(<2 x double> %a) {
+; SSE2-LABEL: shuffle_v2f64_z1:
+; SSE2:       # BB#0:
+; SSE2-NEXT:    xorpd %xmm1, %xmm1
+; SSE2-NEXT:    shufpd {{.*#+}} xmm1 = xmm1[0],xmm0[1]
+; SSE2-NEXT:    movapd %xmm1, %xmm0
+; SSE2-NEXT:    retq
+;
+; SSE3-LABEL: shuffle_v2f64_z1:
+; SSE3:       # BB#0:
+; SSE3-NEXT:    xorpd %xmm1, %xmm1
+; SSE3-NEXT:    shufpd {{.*#+}} xmm1 = xmm1[0],xmm0[1]
+; SSE3-NEXT:    movapd %xmm1, %xmm0
+; SSE3-NEXT:    retq
+;
+; SSSE3-LABEL: shuffle_v2f64_z1:
+; SSSE3:       # BB#0:
+; SSSE3-NEXT:    xorpd %xmm1, %xmm1
+; SSSE3-NEXT:    shufpd {{.*#+}} xmm1 = xmm1[0],xmm0[1]
+; SSSE3-NEXT:    movapd %xmm1, %xmm0
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: shuffle_v2f64_z1:
+; SSE41:       # BB#0:
+; SSE41-NEXT:    xorpd %xmm1, %xmm1
+; SSE41-NEXT:    blendpd {{.*#+}} xmm1 = xmm1[0],xmm0[1]
+; SSE41-NEXT:    movapd %xmm1, %xmm0
+; SSE41-NEXT:    retq
+;
+; AVX-LABEL: shuffle_v2f64_z1:
+; AVX:       # BB#0:
+; AVX-NEXT:    vxorpd %xmm1, %xmm1, %xmm1
+; AVX-NEXT:    vblendpd {{.*#+}} xmm0 = xmm1[0],xmm0[1]
+; AVX-NEXT:    retq
+  %shuffle = shufflevector <2 x double> %a, <2 x double> zeroinitializer, <2 x i32> <i32 2, i32 1>
+  ret <2 x double> %shuffle
+}
 
 define <2 x i64> @insert_reg_and_zero_v2i64(i64 %a) {
 ; SSE-LABEL: insert_reg_and_zero_v2i64:

diff --git a/test/CodeGen/X86/vector-shuffle-128-v4.ll b/test/CodeGen/X86/vector-shuffle-128-v4.ll
index a359ce7b1b8678363af6335b23052f408440dc95..3e278d23f8af0e6b65104ea4c923694bafd75d58 100644 (file)
--- a/test/CodeGen/X86/vector-shuffle-128-v4.ll
+++ b/test/CodeGen/X86/vector-shuffle-128-v4.ll
@@ -659,85 +659,189 @@ define <4 x float> @shuffle_v4f32_z6zz(<4 x float> %a) {
   ret <4 x float> %shuffle
 }
 
-define <4 x i32> @shuffle_v4i32_4zzz(i32 %i) {
-; SSE-LABEL: shuffle_v4i32_4zzz:
-; SSE:       # BB#0:
-; SSE-NEXT:    movd %edi, %xmm0
-; SSE-NEXT:    retq
+define <4 x i32> @shuffle_v4i32_4zzz(<4 x i32> %a) {
+; SSE2-LABEL: shuffle_v4i32_4zzz:
+; SSE2:       # BB#0:
+; SSE2-NEXT:    xorps %xmm1, %xmm1
+; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,0],xmm1[1,0]
+; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[2,3]
+; SSE2-NEXT:    retq
 ;
-; AVX-LABEL: shuffle_v4i32_4zzz:
-; AVX:       # BB#0:
-; AVX-NEXT:    vmovd %edi, %xmm0
-; AVX-NEXT:    retq
-  %a = insertelement <4 x i32> undef, i32 %i, i32 0
+; SSE3-LABEL: shuffle_v4i32_4zzz:
+; SSE3:       # BB#0:
+; SSE3-NEXT:    xorps %xmm1, %xmm1
+; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,0],xmm1[1,0]
+; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[2,3]
+; SSE3-NEXT:    retq
+;
+; SSSE3-LABEL: shuffle_v4i32_4zzz:
+; SSSE3:       # BB#0:
+; SSSE3-NEXT:    xorps %xmm1, %xmm1
+; SSSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,0],xmm1[1,0]
+; SSSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[2,3]
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: shuffle_v4i32_4zzz:
+; SSE41:       # BB#0:
+; SSE41-NEXT:    pxor %xmm1, %xmm1
+; SSE41-NEXT:    pblendw {{.*#+}} xmm1 = xmm0[0,1],xmm1[2,3,4,5,6,7]
+; SSE41-NEXT:    movdqa %xmm1, %xmm0
+; SSE41-NEXT:    retq
+;
+; AVX1-LABEL: shuffle_v4i32_4zzz:
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpblendw {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,3,4,5,6,7]
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: shuffle_v4i32_4zzz:
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX2-NEXT:    vpblendd {{.*#+}} xmm0 = xmm0[0],xmm1[1,2,3]
+; AVX2-NEXT:    retq
   %shuffle = shufflevector <4 x i32> zeroinitializer, <4 x i32> %a, <4 x i32> <i32 4, i32 1, i32 2, i32 3>
   ret <4 x i32> %shuffle
 }
 
-define <4 x i32> @shuffle_v4i32_z4zz(i32 %i) {
-; SSE-LABEL: shuffle_v4i32_z4zz:
-; SSE:       # BB#0:
-; SSE-NEXT:    movd %edi, %xmm0
-; SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[1,0,1,1]
-; SSE-NEXT:    retq
+define <4 x i32> @shuffle_v4i32_z4zz(<4 x i32> %a) {
+; SSE2-LABEL: shuffle_v4i32_z4zz:
+; SSE2:       # BB#0:
+; SSE2-NEXT:    xorps %xmm1, %xmm1
+; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,0],xmm1[2,0]
+; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[2,0],xmm1[3,0]
+; SSE2-NEXT:    retq
+;
+; SSE3-LABEL: shuffle_v4i32_z4zz:
+; SSE3:       # BB#0:
+; SSE3-NEXT:    xorps %xmm1, %xmm1
+; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,0],xmm1[2,0]
+; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[2,0],xmm1[3,0]
+; SSE3-NEXT:    retq
+;
+; SSSE3-LABEL: shuffle_v4i32_z4zz:
+; SSSE3:       # BB#0:
+; SSSE3-NEXT:    xorps %xmm1, %xmm1
+; SSSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,0],xmm1[2,0]
+; SSSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[2,0],xmm1[3,0]
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: shuffle_v4i32_z4zz:
+; SSE41:       # BB#0:
+; SSE41-NEXT:    insertps {{.*#+}} xmm0 = zero,xmm0[0],zero,zero
+; SSE41-NEXT:    retq
 ;
 ; AVX-LABEL: shuffle_v4i32_z4zz:
 ; AVX:       # BB#0:
-; AVX-NEXT:    vmovd %edi, %xmm0
-; AVX-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[1,0,1,1]
+; AVX-NEXT:    vinsertps {{.*#+}} xmm0 = zero,xmm0[0],zero,zero
 ; AVX-NEXT:    retq
-  %a = insertelement <4 x i32> undef, i32 %i, i32 0
   %shuffle = shufflevector <4 x i32> zeroinitializer, <4 x i32> %a, <4 x i32> <i32 2, i32 4, i32 3, i32 0>
   ret <4 x i32> %shuffle
 }
 
-define <4 x i32> @shuffle_v4i32_zz4z(i32 %i) {
-; SSE-LABEL: shuffle_v4i32_zz4z:
-; SSE:       # BB#0:
-; SSE-NEXT:    movd %edi, %xmm0
-; SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[1,1,0,1]
-; SSE-NEXT:    retq
+define <4 x i32> @shuffle_v4i32_zz4z(<4 x i32> %a) {
+; SSE2-LABEL: shuffle_v4i32_zz4z:
+; SSE2:       # BB#0:
+; SSE2-NEXT:    xorps %xmm1, %xmm1
+; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,0],xmm1[0,0]
+; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,0],xmm0[0,2]
+; SSE2-NEXT:    movaps %xmm1, %xmm0
+; SSE2-NEXT:    retq
+;
+; SSE3-LABEL: shuffle_v4i32_zz4z:
+; SSE3:       # BB#0:
+; SSE3-NEXT:    xorps %xmm1, %xmm1
+; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,0],xmm1[0,0]
+; SSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,0],xmm0[0,2]
+; SSE3-NEXT:    movaps %xmm1, %xmm0
+; SSE3-NEXT:    retq
+;
+; SSSE3-LABEL: shuffle_v4i32_zz4z:
+; SSSE3:       # BB#0:
+; SSSE3-NEXT:    xorps %xmm1, %xmm1
+; SSSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,0],xmm1[0,0]
+; SSSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,0],xmm0[0,2]
+; SSSE3-NEXT:    movaps %xmm1, %xmm0
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: shuffle_v4i32_zz4z:
+; SSE41:       # BB#0:
+; SSE41-NEXT:    insertps {{.*#+}} xmm0 = zero,zero,xmm0[0],zero
+; SSE41-NEXT:    retq
 ;
 ; AVX-LABEL: shuffle_v4i32_zz4z:
 ; AVX:       # BB#0:
-; AVX-NEXT:    vmovd %edi, %xmm0
-; AVX-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[1,1,0,1]
+; AVX-NEXT:    vinsertps {{.*#+}} xmm0 = zero,zero,xmm0[0],zero
 ; AVX-NEXT:    retq
-  %a = insertelement <4 x i32> undef, i32 %i, i32 0
   %shuffle = shufflevector <4 x i32> zeroinitializer, <4 x i32> %a, <4 x i32> <i32 0, i32 0, i32 4, i32 0>
   ret <4 x i32> %shuffle
 }
 
-define <4 x i32> @shuffle_v4i32_zuu4(i32 %i) {
-; SSE-LABEL: shuffle_v4i32_zuu4:
-; SSE:       # BB#0:
-; SSE-NEXT:    movd %edi, %xmm0
-; SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[1,1,1,0]
-; SSE-NEXT:    retq
+define <4 x i32> @shuffle_v4i32_zuu4(<4 x i32> %a) {
+; SSE2-LABEL: shuffle_v4i32_zuu4:
+; SSE2:       # BB#0:
+; SSE2-NEXT:    xorps %xmm1, %xmm1
+; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,1],xmm0[2,0]
+; SSE2-NEXT:    movaps %xmm1, %xmm0
+; SSE2-NEXT:    retq
+;
+; SSE3-LABEL: shuffle_v4i32_zuu4:
+; SSE3:       # BB#0:
+; SSE3-NEXT:    xorps %xmm1, %xmm1
+; SSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,1],xmm0[2,0]
+; SSE3-NEXT:    movaps %xmm1, %xmm0
+; SSE3-NEXT:    retq
+;
+; SSSE3-LABEL: shuffle_v4i32_zuu4:
+; SSSE3:       # BB#0:
+; SSSE3-NEXT:    xorps %xmm1, %xmm1
+; SSSE3-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,1],xmm0[2,0]
+; SSSE3-NEXT:    movaps %xmm1, %xmm0
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: shuffle_v4i32_zuu4:
+; SSE41:       # BB#0:
+; SSE41-NEXT:    insertps {{.*#+}} xmm0 = zero,zero,zero,xmm0[0]
+; SSE41-NEXT:    retq
 ;
 ; AVX-LABEL: shuffle_v4i32_zuu4:
 ; AVX:       # BB#0:
-; AVX-NEXT:    vmovd %edi, %xmm0
-; AVX-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[1,1,1,0]
+; AVX-NEXT:    vinsertps {{.*#+}} xmm0 = zero,zero,zero,xmm0[0]
 ; AVX-NEXT:    retq
-  %a = insertelement <4 x i32> undef, i32 %i, i32 0
   %shuffle = shufflevector <4 x i32> zeroinitializer, <4 x i32> %a, <4 x i32> <i32 0, i32 undef, i32 undef, i32 4>
   ret <4 x i32> %shuffle
 }
 
-define <4 x i32> @shuffle_v4i32_z6zz(i32 %i) {
-; SSE-LABEL: shuffle_v4i32_z6zz:
-; SSE:       # BB#0:
-; SSE-NEXT:    movd %edi, %xmm0
-; SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[1,0,1,1]
-; SSE-NEXT:    retq
+define <4 x i32> @shuffle_v4i32_z6zz(<4 x i32> %a) {
+; SSE2-LABEL: shuffle_v4i32_z6zz:
+; SSE2:       # BB#0:
+; SSE2-NEXT:    xorps %xmm1, %xmm1
+; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[2,0],xmm1[0,0]
+; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[2,0],xmm1[2,3]
+; SSE2-NEXT:    retq
+;
+; SSE3-LABEL: shuffle_v4i32_z6zz:
+; SSE3:       # BB#0:
+; SSE3-NEXT:    xorps %xmm1, %xmm1
+; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[2,0],xmm1[0,0]
+; SSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[2,0],xmm1[2,3]
+; SSE3-NEXT:    retq
+;
+; SSSE3-LABEL: shuffle_v4i32_z6zz:
+; SSSE3:       # BB#0:
+; SSSE3-NEXT:    xorps %xmm1, %xmm1
+; SSSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[2,0],xmm1[0,0]
+; SSSE3-NEXT:    shufps {{.*#+}} xmm0 = xmm0[2,0],xmm1[2,3]
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: shuffle_v4i32_z6zz:
+; SSE41:       # BB#0:
+; SSE41-NEXT:    insertps {{.*#+}} xmm0 = zero,xmm0[2],zero,zero
+; SSE41-NEXT:    retq
 ;
 ; AVX-LABEL: shuffle_v4i32_z6zz:
 ; AVX:       # BB#0:
-; AVX-NEXT:    vmovd %edi, %xmm0
-; AVX-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[1,0,1,1]
+; AVX-NEXT:    vinsertps {{.*#+}} xmm0 = zero,xmm0[2],zero,zero
 ; AVX-NEXT:    retq
-  %a = insertelement <4 x i32> undef, i32 %i, i32 2
   %shuffle = shufflevector <4 x i32> zeroinitializer, <4 x i32> %a, <4 x i32> <i32 0, i32 6, i32 2, i32 3>
   ret <4 x i32> %shuffle
 }