[x86] Teach the new vector shuffle lowering about VBROADCAST and

VPBROADCAST.

This has the somewhat expected pervasive impact. I don't know why
I forgot about this. Everything seems good with lots of significant
improvements in the tests.

llvm-svn: 218724

8 files changed, 310 insertions, 263 deletions

diff --git a/llvm/test/CodeGen/X86/vector-shuffle-128-v16.ll b/llvm/test/CodeGen/X86/vector-shuffle-128-v16.ll
index 306ef8a220f..d7373602431 100644
--- a/llvm/test/CodeGen/X86/vector-shuffle-128-v16.ll
+++ b/llvm/test/CodeGen/X86/vector-shuffle-128-v16.ll
@@ -36,11 +36,16 @@ define <16 x i8> @shuffle_v16i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00(
 ; SSE41-NEXT:    pshufb %xmm1, %xmm0
 ; SSE41-NEXT:    retq
 ;
-; AVX-LABEL: shuffle_v16i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00:
-; AVX:       # BB#0:
-; AVX-NEXT:    vpxor %xmm1, %xmm1, %xmm1
-; AVX-NEXT:    vpshufb %xmm1, %xmm0, %xmm0
-; AVX-NEXT:    retq
+; AVX1-LABEL: shuffle_v16i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00:
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: shuffle_v16i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00:
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpbroadcastb %xmm0, %xmm0
+; AVX2-NEXT:    retq
   %shuffle = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <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 0, i32 0>
   ret <16 x i8> %shuffle
 }
@@ -200,10 +205,15 @@ define <16 x i8> @shuffle_v16i8_0101010101010101(<16 x i8> %a, <16 x i8> %b) {
 ; SSE41-NEXT:    pshufb {{.*#+}} xmm0 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
 ; SSE41-NEXT:    retq
 ;
-; AVX-LABEL: shuffle_v16i8_0101010101010101:
-; AVX:       # BB#0:
-; AVX-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
-; AVX-NEXT:    retq
+; AVX1-LABEL: shuffle_v16i8_0101010101010101:
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: shuffle_v16i8_0101010101010101:
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpbroadcastw %xmm0, %xmm0
+; AVX2-NEXT:    retq
   %shuffle = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 0, i32 1, i32 0, i32 1, i32 0, i32 1, i32 0, i32 1, i32 0, i32 1, i32 0, i32 1, i32 0, i32 1, i32 0, i32 1>
   ret <16 x i8> %shuffle
 }
@@ -231,12 +241,18 @@ define <16 x i8> @shuffle_v16i8_16_00_16_01_16_02_16_03_16_04_16_05_16_06_16_07(
 ; SSE-NEXT:    movdqa %xmm1, %xmm0
 ; SSE-NEXT:    retq
 ;
-; AVX-LABEL: shuffle_v16i8_16_00_16_01_16_02_16_03_16_04_16_05_16_06_16_07:
-; AVX:       # BB#0:
-; AVX-NEXT:    vpunpcklbw {{.*#+}} xmm1 = xmm1[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
-; AVX-NEXT:    vpshuflw {{.*#+}} xmm1 = xmm1[0,0,0,0,4,5,6,7]
-; AVX-NEXT:    vpunpcklbw {{.*#+}} xmm0 = xmm1[0],xmm0[0],xmm1[1],xmm0[1],xmm1[2],xmm0[2],xmm1[3],xmm0[3],xmm1[4],xmm0[4],xmm1[5],xmm0[5],xmm1[6],xmm0[6],xmm1[7],xmm0[7]
-; AVX-NEXT:    retq
+; AVX1-LABEL: shuffle_v16i8_16_00_16_01_16_02_16_03_16_04_16_05_16_06_16_07:
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpunpcklbw {{.*#+}} xmm1 = xmm1[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; AVX1-NEXT:    vpshuflw {{.*#+}} xmm1 = xmm1[0,0,0,0,4,5,6,7]
+; AVX1-NEXT:    vpunpcklbw {{.*#+}} xmm0 = xmm1[0],xmm0[0],xmm1[1],xmm0[1],xmm1[2],xmm0[2],xmm1[3],xmm0[3],xmm1[4],xmm0[4],xmm1[5],xmm0[5],xmm1[6],xmm0[6],xmm1[7],xmm0[7]
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: shuffle_v16i8_16_00_16_01_16_02_16_03_16_04_16_05_16_06_16_07:
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpbroadcastb %xmm1, %xmm1
+; AVX2-NEXT:    vpunpcklbw {{.*#+}} xmm0 = xmm1[0],xmm0[0],xmm1[1],xmm0[1],xmm1[2],xmm0[2],xmm1[3],xmm0[3],xmm1[4],xmm0[4],xmm1[5],xmm0[5],xmm1[6],xmm0[6],xmm1[7],xmm0[7]
+; AVX2-NEXT:    retq
   %shuffle = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 16, i32 0, i32 16, i32 1, i32 16, i32 2, i32 16, i32 3, i32 16, i32 4, i32 16, i32 5, i32 16, i32 6, i32 16, i32 7>
   ret <16 x i8> %shuffle
 }
diff --git a/llvm/test/CodeGen/X86/vector-shuffle-128-v2.ll b/llvm/test/CodeGen/X86/vector-shuffle-128-v2.ll
index 6f891c12505..67723dd07b2 100644
--- a/llvm/test/CodeGen/X86/vector-shuffle-128-v2.ll
+++ b/llvm/test/CodeGen/X86/vector-shuffle-128-v2.ll
@@ -14,10 +14,15 @@ define <2 x i64> @shuffle_v2i64_00(<2 x i64> %a, <2 x i64> %b) {
 ; SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,1,0,1]
 ; SSE-NEXT:    retq
 ;
-; AVX-LABEL: shuffle_v2i64_00:
-; AVX:       # BB#0:
-; AVX-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[0,1,0,1]
-; AVX-NEXT:    retq
+; AVX1-LABEL: shuffle_v2i64_00:
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[0,1,0,1]
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: shuffle_v2i64_00:
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpbroadcastq %xmm0, %xmm0
+; AVX2-NEXT:    retq
   %shuffle = shufflevector <2 x i64> %a, <2 x i64> %b, <2 x i32> <i32 0, i32 0>
   ret <2 x i64> %shuffle
 }
@@ -53,10 +58,15 @@ define <2 x i64> @shuffle_v2i64_22(<2 x i64> %a, <2 x i64> %b) {
 ; SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm1[0,1,0,1]
 ; SSE-NEXT:    retq
 ;
-; AVX-LABEL: shuffle_v2i64_22:
-; AVX:       # BB#0:
-; AVX-NEXT:    vpshufd {{.*#+}} xmm0 = xmm1[0,1,0,1]
-; AVX-NEXT:    retq
+; AVX1-LABEL: shuffle_v2i64_22:
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufd {{.*#+}} xmm0 = xmm1[0,1,0,1]
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: shuffle_v2i64_22:
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpbroadcastq %xmm1, %xmm0
+; AVX2-NEXT:    retq
   %shuffle = shufflevector <2 x i64> %a, <2 x i64> %b, <2 x i32> <i32 2, i32 2>
   ret <2 x i64> %shuffle
 }
diff --git a/llvm/test/CodeGen/X86/vector-shuffle-128-v8.ll b/llvm/test/CodeGen/X86/vector-shuffle-128-v8.ll
index d0c4a3f9d36..7d6e2bb1124 100644
--- a/llvm/test/CodeGen/X86/vector-shuffle-128-v8.ll
+++ b/llvm/test/CodeGen/X86/vector-shuffle-128-v8.ll
@@ -77,10 +77,15 @@ define <8 x i16> @shuffle_v8i16_00000000(<8 x i16> %a, <8 x i16> %b) {
 ; SSE41-NEXT:    pshufb {{.*#+}} xmm0 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
 ; SSE41-NEXT:    retq
 ;
-; AVX-LABEL: shuffle_v8i16_00000000:
-; AVX:       # BB#0:
-; AVX-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
-; AVX-NEXT:    retq
+; AVX1-LABEL: shuffle_v8i16_00000000:
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: shuffle_v8i16_00000000:
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpbroadcastw %xmm0, %xmm0
+; AVX2-NEXT:    retq
   %shuffle = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0>
   ret <8 x i16> %shuffle
 }
@@ -1169,12 +1174,19 @@ define <8 x i16> @shuffle_v8i16_XXXXcde3(<8 x i16> %a, <8 x i16> %b) {
 ; SSE41-NEXT:    movdqa %xmm1, %xmm0
 ; SSE41-NEXT:    retq
 ;
-; AVX-LABEL: shuffle_v8i16_XXXXcde3:
-; AVX:       # BB#0:
-; AVX-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[0,1,0,1]
-; AVX-NEXT:    vpunpckhwd {{.*#+}} xmm0 = xmm1[4],xmm0[4],xmm1[5],xmm0[5],xmm1[6],xmm0[6],xmm1[7],xmm0[7]
-; AVX-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,1,4,5,4,5,6,7,0,1,4,5,8,9,14,15]
-; AVX-NEXT:    retq
+; AVX1-LABEL: shuffle_v8i16_XXXXcde3:
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[0,1,0,1]
+; AVX1-NEXT:    vpunpckhwd {{.*#+}} xmm0 = xmm1[4],xmm0[4],xmm1[5],xmm0[5],xmm1[6],xmm0[6],xmm1[7],xmm0[7]
+; AVX1-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,1,4,5,4,5,6,7,0,1,4,5,8,9,14,15]
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: shuffle_v8i16_XXXXcde3:
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpbroadcastq %xmm0, %xmm0
+; AVX2-NEXT:    vpunpckhwd {{.*#+}} xmm0 = xmm1[4],xmm0[4],xmm1[5],xmm0[5],xmm1[6],xmm0[6],xmm1[7],xmm0[7]
+; AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,1,4,5,4,5,6,7,0,1,4,5,8,9,14,15]
+; AVX2-NEXT:    retq
   %shuffle = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 undef, i32 undef, i32 undef, i32 undef, i32 12, i32 13, i32 14, i32 3>
   ret <8 x i16> %shuffle
 }
@@ -1205,12 +1217,19 @@ define <8 x i16> @shuffle_v8i16_cde3XXXX(<8 x i16> %a, <8 x i16> %b) {
 ; SSE41-NEXT:    movdqa %xmm1, %xmm0
 ; SSE41-NEXT:    retq
 ;
-; AVX-LABEL: shuffle_v8i16_cde3XXXX:
-; AVX:       # BB#0:
-; AVX-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[0,1,0,1]
-; AVX-NEXT:    vpunpckhwd {{.*#+}} xmm0 = xmm1[4],xmm0[4],xmm1[5],xmm0[5],xmm1[6],xmm0[6],xmm1[7],xmm0[7]
-; AVX-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,1,4,5,8,9,14,15,8,9,14,15,12,13,14,15]
-; AVX-NEXT:    retq
+; AVX1-LABEL: shuffle_v8i16_cde3XXXX:
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[0,1,0,1]
+; AVX1-NEXT:    vpunpckhwd {{.*#+}} xmm0 = xmm1[4],xmm0[4],xmm1[5],xmm0[5],xmm1[6],xmm0[6],xmm1[7],xmm0[7]
+; AVX1-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,1,4,5,8,9,14,15,8,9,14,15,12,13,14,15]
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: shuffle_v8i16_cde3XXXX:
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpbroadcastq %xmm0, %xmm0
+; AVX2-NEXT:    vpunpckhwd {{.*#+}} xmm0 = xmm1[4],xmm0[4],xmm1[5],xmm0[5],xmm1[6],xmm0[6],xmm1[7],xmm0[7]
+; AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,1,4,5,8,9,14,15,8,9,14,15,12,13,14,15]
+; AVX2-NEXT:    retq
   %shuffle = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 12, i32 13, i32 14, i32 3, i32 undef, i32 undef, i32 undef, i32 undef>
   ret <8 x i16> %shuffle
 }
@@ -1254,16 +1273,27 @@ define <8 x i16> @shuffle_v8i16_012dcde3(<8 x i16> %a, <8 x i16> %b) {
 ; SSE41-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
 ; SSE41-NEXT:    retq
 ;
-; AVX-LABEL: shuffle_v8i16_012dcde3:
-; AVX:       # BB#0:
-; AVX-NEXT:    vpshufd {{.*#+}} xmm2 = xmm0[0,1,0,1]
-; AVX-NEXT:    vpunpckhwd {{.*#+}} xmm2 = xmm1[4],xmm2[4],xmm1[5],xmm2[5],xmm1[6],xmm2[6],xmm1[7],xmm2[7]
-; AVX-NEXT:    vpshufb {{.*#+}} xmm2 = xmm2[0,1,4,5,8,9,14,15,8,9,14,15,12,13,14,15]
-; AVX-NEXT:    vpshufd {{.*#+}} xmm1 = xmm1[2,3,0,1]
-; AVX-NEXT:    vpunpcklwd {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
-; AVX-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,1,4,5,8,9,6,7,8,9,0,1,0,1,2,3]
-; AVX-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm2[0]
-; AVX-NEXT:    retq
+; AVX1-LABEL: shuffle_v8i16_012dcde3:
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufd {{.*#+}} xmm2 = xmm0[0,1,0,1]
+; AVX1-NEXT:    vpunpckhwd {{.*#+}} xmm2 = xmm1[4],xmm2[4],xmm1[5],xmm2[5],xmm1[6],xmm2[6],xmm1[7],xmm2[7]
+; AVX1-NEXT:    vpshufb {{.*#+}} xmm2 = xmm2[0,1,4,5,8,9,14,15,8,9,14,15,12,13,14,15]
+; AVX1-NEXT:    vpshufd {{.*#+}} xmm1 = xmm1[2,3,0,1]
+; AVX1-NEXT:    vpunpcklwd {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX1-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,1,4,5,8,9,6,7,8,9,0,1,0,1,2,3]
+; AVX1-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm2[0]
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: shuffle_v8i16_012dcde3:
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpbroadcastq %xmm0, %xmm2
+; AVX2-NEXT:    vpunpckhwd {{.*#+}} xmm2 = xmm1[4],xmm2[4],xmm1[5],xmm2[5],xmm1[6],xmm2[6],xmm1[7],xmm2[7]
+; AVX2-NEXT:    vpshufb {{.*#+}} xmm2 = xmm2[0,1,4,5,8,9,14,15,8,9,14,15,12,13,14,15]
+; AVX2-NEXT:    vpshufd {{.*#+}} xmm1 = xmm1[2,3,0,1]
+; AVX2-NEXT:    vpunpcklwd {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,1,4,5,8,9,6,7,8,9,0,1,0,1,2,3]
+; AVX2-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm2[0]
+; AVX2-NEXT:    retq
   %shuffle = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 0, i32 1, i32 2, i32 13, i32 12, i32 13, i32 14, i32 3>
   ret <8 x i16> %shuffle
 }
diff --git a/llvm/test/CodeGen/X86/vector-shuffle-256-v16.ll b/llvm/test/CodeGen/X86/vector-shuffle-256-v16.ll
index 0cb73521503..cc449899110 100644
--- a/llvm/test/CodeGen/X86/vector-shuffle-256-v16.ll
+++ b/llvm/test/CodeGen/X86/vector-shuffle-256-v16.ll
@@ -12,8 +12,7 @@ define <16 x i16> @shuffle_v16i16_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_0
 ;
 ; AVX2-LABEL: shuffle_v16i16_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
-; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; AVX2-NEXT:    vpbroadcastw %xmm0, %ymm0
 ; AVX2-NEXT:    retq
   %shuffle = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <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 0, i32 0>
   ret <16 x i16> %shuffle
@@ -31,10 +30,8 @@ define <16 x i16> @shuffle_v16i16_00_00_00_00_00_00_00_00_00_00_00_00_00_00_01_0
 ;
 ; AVX2-LABEL: shuffle_v16i16_00_00_00_00_00_00_00_00_00_00_00_00_00_00_01_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[0,1,0,3]
-; AVX2-NEXT:    vpshuflw {{.*#+}} xmm0 = xmm0[0,0,0,0,4,5,6,7]
-; AVX2-NEXT:    vpshufhw {{.*#+}} xmm1 = xmm0[0,1,2,3,4,4,4,4]
-; AVX2-NEXT:    vpshufhw {{.*#+}} xmm0 = xmm0[0,1,2,3,4,4,5,4]
+; AVX2-NEXT:    vpbroadcastw %xmm0, %xmm1
+; AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,2,3,0,1]
 ; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
 ; AVX2-NEXT:    retq
   %shuffle = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <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 1, i32 0>
@@ -51,7 +48,7 @@ define <16 x i16> @shuffle_v16i16_00_00_00_00_00_00_00_00_00_00_00_00_00_02_00_0
 ;
 ; AVX2-LABEL: shuffle_v16i16_00_00_00_00_00_00_00_00_00_00_00_00_00_02_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpshufb {{.*#+}} xmm1 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX2-NEXT:    vpbroadcastw %xmm0, %xmm1
 ; AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,1,0,1,0,1,0,1,0,1,4,5,0,1,0,1]
 ; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
 ; AVX2-NEXT:    retq
@@ -69,7 +66,7 @@ define <16 x i16> @shuffle_v16i16_00_00_00_00_00_00_00_00_00_00_00_00_03_00_00_0
 ;
 ; AVX2-LABEL: shuffle_v16i16_00_00_00_00_00_00_00_00_00_00_00_00_03_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpshufb {{.*#+}} xmm1 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX2-NEXT:    vpbroadcastw %xmm0, %xmm1
 ; AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,1,0,1,0,1,0,1,6,7,0,1,0,1,0,1]
 ; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
 ; AVX2-NEXT:    retq
@@ -87,7 +84,7 @@ define <16 x i16> @shuffle_v16i16_00_00_00_00_00_00_00_00_00_00_00_04_00_00_00_0
 ;
 ; AVX2-LABEL: shuffle_v16i16_00_00_00_00_00_00_00_00_00_00_00_04_00_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpshufb {{.*#+}} xmm1 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX2-NEXT:    vpbroadcastw %xmm0, %xmm1
 ; AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,1,0,1,0,1,8,9,0,1,0,1,0,1,0,1]
 ; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
 ; AVX2-NEXT:    retq
@@ -105,7 +102,7 @@ define <16 x i16> @shuffle_v16i16_00_00_00_00_00_00_00_00_00_00_05_00_00_00_00_0
 ;
 ; AVX2-LABEL: shuffle_v16i16_00_00_00_00_00_00_00_00_00_00_05_00_00_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpshufb {{.*#+}} xmm1 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX2-NEXT:    vpbroadcastw %xmm0, %xmm1
 ; AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,1,0,1,10,11,0,1,0,1,0,1,0,1,0,1]
 ; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
 ; AVX2-NEXT:    retq
@@ -123,7 +120,7 @@ define <16 x i16> @shuffle_v16i16_00_00_00_00_00_00_00_00_00_06_00_00_00_00_00_0
 ;
 ; AVX2-LABEL: shuffle_v16i16_00_00_00_00_00_00_00_00_00_06_00_00_00_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpshufb {{.*#+}} xmm1 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX2-NEXT:    vpbroadcastw %xmm0, %xmm1
 ; AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,1,12,13,0,1,0,1,0,1,0,1,0,1,0,1]
 ; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
 ; AVX2-NEXT:    retq
@@ -141,7 +138,7 @@ define <16 x i16> @shuffle_v16i16_00_00_00_00_00_00_00_00_07_00_00_00_00_00_00_0
 ;
 ; AVX2-LABEL: shuffle_v16i16_00_00_00_00_00_00_00_00_07_00_00_00_00_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpshufb {{.*#+}} xmm1 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX2-NEXT:    vpbroadcastw %xmm0, %xmm1
 ; AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[14,15,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
 ; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
 ; AVX2-NEXT:    retq
@@ -165,7 +162,7 @@ define <16 x i16> @shuffle_v16i16_00_00_00_00_00_00_00_08_00_00_00_00_00_00_00_0
 ; AVX2:       # BB#0:
 ; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm1 = ymm0[2,3,0,1]
 ; AVX2-NEXT:    vpshufb {{.*#+}} ymm1 = ymm1[u,u,u,u,u,u,u,u,u,u,u,u,u,u,0,1,16,17,16,17,16,17,16,17,16,17,16,17,16,17,16,17]
-; AVX2-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u]
+; AVX2-NEXT:    vpbroadcastw %xmm0, %ymm0
 ; AVX2-NEXT:    vmovdqa {{.*#+}} ymm2 = [0,0,0,0,0,0,0,0,0,0,0,0,0,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
 ; AVX2-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
 ; AVX2-NEXT:    retq
@@ -189,7 +186,7 @@ define <16 x i16> @shuffle_v16i16_00_00_00_00_00_00_09_00_00_00_00_00_00_00_00_0
 ; AVX2:       # BB#0:
 ; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm1 = ymm0[2,3,0,1]
 ; AVX2-NEXT:    vpshufb {{.*#+}} ymm1 = ymm1[u,u,u,u,u,u,u,u,u,u,u,u,2,3,u,u,16,17,16,17,16,17,16,17,16,17,16,17,16,17,16,17]
-; AVX2-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[0,1,0,1,0,1,0,1,0,1,0,1,u,u,0,1,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u]
+; AVX2-NEXT:    vpbroadcastw %xmm0, %ymm0
 ; AVX2-NEXT:    vmovdqa {{.*#+}} ymm2 = [0,0,0,0,0,0,0,0,0,0,0,0,255,255,0,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
 ; AVX2-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
 ; AVX2-NEXT:    retq
@@ -214,7 +211,7 @@ define <16 x i16> @shuffle_v16i16_00_00_00_00_00_10_00_00_00_00_00_00_00_00_00_0
 ; AVX2:       # BB#0:
 ; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm1 = ymm0[2,3,0,1]
 ; AVX2-NEXT:    vpshufb {{.*#+}} ymm1 = ymm1[u,u,u,u,u,u,u,u,u,u,4,5,u,u,u,u,16,17,16,17,16,17,16,17,16,17,16,17,16,17,16,17]
-; AVX2-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[0,1,0,1,0,1,0,1,0,1,u,u,0,1,0,1,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u]
+; AVX2-NEXT:    vpbroadcastw %xmm0, %ymm0
 ; AVX2-NEXT:    vmovdqa {{.*#+}} ymm2 = [0,0,0,0,0,0,0,0,0,0,255,255,0,0,0,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
 ; AVX2-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
 ; AVX2-NEXT:    retq
@@ -239,7 +236,7 @@ define <16 x i16> @shuffle_v16i16_00_00_00_00_11_00_00_00_00_00_00_00_00_00_00_0
 ; AVX2:       # BB#0:
 ; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm1 = ymm0[2,3,0,1]
 ; AVX2-NEXT:    vpshufb {{.*#+}} ymm1 = ymm1[u,u,u,u,u,u,u,u,6,7,u,u,u,u,u,u,16,17,16,17,16,17,16,17,16,17,16,17,16,17,16,17]
-; AVX2-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[0,1,0,1,0,1,0,1,u,u,0,1,0,1,0,1,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u]
+; AVX2-NEXT:    vpbroadcastw %xmm0, %ymm0
 ; AVX2-NEXT:    vmovdqa {{.*#+}} ymm2 = [0,0,0,0,0,0,0,0,255,255,0,0,0,0,0,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
 ; AVX2-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
 ; AVX2-NEXT:    retq
@@ -264,7 +261,7 @@ define <16 x i16> @shuffle_v16i16_00_00_00_12_00_00_00_00_00_00_00_00_00_00_00_0
 ; AVX2:       # BB#0:
 ; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm1 = ymm0[2,3,0,1]
 ; AVX2-NEXT:    vpshufb {{.*#+}} ymm1 = ymm1[u,u,u,u,u,u,8,9,u,u,u,u,u,u,u,u,16,17,16,17,16,17,16,17,16,17,16,17,16,17,16,17]
-; AVX2-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[0,1,0,1,0,1,u,u,0,1,0,1,0,1,0,1,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u]
+; AVX2-NEXT:    vpbroadcastw %xmm0, %ymm0
 ; AVX2-NEXT:    vmovdqa {{.*#+}} ymm2 = [0,0,0,0,0,0,255,255,0,0,0,0,0,0,0,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
 ; AVX2-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
 ; AVX2-NEXT:    retq
@@ -289,7 +286,7 @@ define <16 x i16> @shuffle_v16i16_00_00_13_00_00_00_00_00_00_00_00_00_00_00_00_0
 ; AVX2:       # BB#0:
 ; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm1 = ymm0[2,3,0,1]
 ; AVX2-NEXT:    vpshufb {{.*#+}} ymm1 = ymm1[u,u,u,u,10,11,u,u,u,u,u,u,u,u,u,u,16,17,16,17,16,17,16,17,16,17,16,17,16,17,16,17]
-; AVX2-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[0,1,0,1,u,u,0,1,0,1,0,1,0,1,0,1,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u]
+; AVX2-NEXT:    vpbroadcastw %xmm0, %ymm0
 ; AVX2-NEXT:    vmovdqa {{.*#+}} ymm2 = [0,0,0,0,255,255,0,0,0,0,0,0,0,0,0,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
 ; AVX2-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
 ; AVX2-NEXT:    retq
@@ -315,7 +312,7 @@ define <16 x i16> @shuffle_v16i16_00_14_00_00_00_00_00_00_00_00_00_00_00_00_00_0
 ; AVX2:       # BB#0:
 ; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm1 = ymm0[2,3,0,1]
 ; AVX2-NEXT:    vpshufb {{.*#+}} ymm1 = ymm1[u,u,12,13,u,u,u,u,u,u,u,u,u,u,u,u,16,17,16,17,16,17,16,17,16,17,16,17,16,17,16,17]
-; AVX2-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[0,1,u,u,0,1,0,1,0,1,0,1,0,1,0,1,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u]
+; AVX2-NEXT:    vpbroadcastw %xmm0, %ymm0
 ; AVX2-NEXT:    vmovdqa {{.*#+}} ymm2 = [0,0,255,255,0,0,0,0,0,0,0,0,0,0,0,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
 ; AVX2-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
 ; AVX2-NEXT:    retq
@@ -341,7 +338,7 @@ define <16 x i16> @shuffle_v16i16_15_00_00_00_00_00_00_00_00_00_00_00_00_00_00_0
 ; AVX2:       # BB#0:
 ; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm1 = ymm0[2,3,0,1]
 ; AVX2-NEXT:    vpshufb {{.*#+}} ymm1 = ymm1[14,15,u,u,u,u,u,u,u,u,u,u,u,u,u,u,16,17,16,17,16,17,16,17,16,17,16,17,16,17,16,17]
-; AVX2-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[u,u,0,1,0,1,0,1,0,1,0,1,0,1,0,1,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u,u]
+; AVX2-NEXT:    vpbroadcastw %xmm0, %ymm0
 ; AVX2-NEXT:    vmovdqa {{.*#+}} ymm2 = [255,255,0,0,0,0,0,0,0,0,0,0,0,0,0,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
 ; AVX2-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
 ; AVX2-NEXT:    retq
@@ -735,8 +732,8 @@ define <16 x i16> @shuffle_v16i16_00_16_00_16_00_16_00_16_00_16_00_16_00_16_00_1
 ;
 ; AVX2-LABEL: shuffle_v16i16_00_16_00_16_00_16_00_16_00_16_00_16_00_16_00_16:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpshufb {{.*#+}} xmm1 = xmm1[0,1,0,1,4,5,0,1,0,1,0,1,12,13,0,1]
-; AVX2-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[0,0,0,0]
+; AVX2-NEXT:    vpbroadcastw %xmm1, %xmm1
+; AVX2-NEXT:    vpbroadcastd %xmm0, %xmm0
 ; AVX2-NEXT:    vpblendw {{.*#+}} xmm0 = xmm0[0],xmm1[1],xmm0[2],xmm1[3],xmm0[4],xmm1[5],xmm0[6],xmm1[7]
 ; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
 ; AVX2-NEXT:    retq
diff --git a/llvm/test/CodeGen/X86/vector-shuffle-256-v32.ll b/llvm/test/CodeGen/X86/vector-shuffle-256-v32.ll
index fab36b2b273..687928f0a7c 100644
--- a/llvm/test/CodeGen/X86/vector-shuffle-256-v32.ll
+++ b/llvm/test/CodeGen/X86/vector-shuffle-256-v32.ll
@@ -13,9 +13,7 @@ define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_
 ;
 ; AVX2-LABEL: shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %xmm1, %xmm1, %xmm1
-; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm0
-; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; AVX2-NEXT:    vpbroadcastb %xmm0, %ymm0
 ; AVX2-NEXT:    retq
   %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <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 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 0, i32 0, i32 0, i32 0>
   ret <32 x i8> %shuffle
@@ -32,8 +30,7 @@ define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_
 ;
 ; AVX2-LABEL: shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_01_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %xmm1, %xmm1, %xmm1
-; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX2-NEXT:    vpbroadcastb %xmm0, %xmm1
 ; AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0]
 ; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
 ; AVX2-NEXT:    retq
@@ -52,8 +49,7 @@ define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_
 ;
 ; AVX2-LABEL: shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_02_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %xmm1, %xmm1, %xmm1
-; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX2-NEXT:    vpbroadcastb %xmm0, %xmm1
 ; AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,0,0,0,0,0,0,0,0,0,0,0,0,2,0,0]
 ; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
 ; AVX2-NEXT:    retq
@@ -72,8 +68,7 @@ define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_
 ;
 ; AVX2-LABEL: shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_03_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %xmm1, %xmm1, %xmm1
-; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX2-NEXT:    vpbroadcastb %xmm0, %xmm1
 ; AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,0,0,0,0,0,0,0,0,0,0,0,3,0,0,0]
 ; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
 ; AVX2-NEXT:    retq
@@ -92,8 +87,7 @@ define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_
 ;
 ; AVX2-LABEL: shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_04_00_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %xmm1, %xmm1, %xmm1
-; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX2-NEXT:    vpbroadcastb %xmm0, %xmm1
 ; AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,0,0,0,0,0,0,0,0,0,0,4,0,0,0,0]
 ; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
 ; AVX2-NEXT:    retq
@@ -112,8 +106,7 @@ define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_
 ;
 ; AVX2-LABEL: shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_05_00_00_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %xmm1, %xmm1, %xmm1
-; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX2-NEXT:    vpbroadcastb %xmm0, %xmm1
 ; AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,0,0,0,0,0,0,0,0,0,5,0,0,0,0,0]
 ; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
 ; AVX2-NEXT:    retq
@@ -132,8 +125,7 @@ define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_
 ;
 ; AVX2-LABEL: shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_06_00_00_00_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %xmm1, %xmm1, %xmm1
-; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX2-NEXT:    vpbroadcastb %xmm0, %xmm1
 ; AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,0,0,0,0,0,0,0,0,6,0,0,0,0,0,0]
 ; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
 ; AVX2-NEXT:    retq
@@ -152,8 +144,7 @@ define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_
 ;
 ; AVX2-LABEL: shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_07_00_00_00_00_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %xmm1, %xmm1, %xmm1
-; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX2-NEXT:    vpbroadcastb %xmm0, %xmm1
 ; AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,0,0,0,0,0,0,0,7,0,0,0,0,0,0,0]
 ; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
 ; AVX2-NEXT:    retq
@@ -172,8 +163,7 @@ define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_
 ;
 ; AVX2-LABEL: shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_08_00_00_00_00_00_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %xmm1, %xmm1, %xmm1
-; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX2-NEXT:    vpbroadcastb %xmm0, %xmm1
 ; AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,0,0,0,0,0,0,8,0,0,0,0,0,0,0,0]
 ; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
 ; AVX2-NEXT:    retq
@@ -192,8 +182,7 @@ define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_
 ;
 ; AVX2-LABEL: shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_09_00_00_00_00_00_00_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %xmm1, %xmm1, %xmm1
-; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX2-NEXT:    vpbroadcastb %xmm0, %xmm1
 ; AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,0,0,0,0,0,9,0,0,0,0,0,0,0,0,0]
 ; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
 ; AVX2-NEXT:    retq
@@ -212,8 +201,7 @@ define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_
 ;
 ; AVX2-LABEL: shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_10_00_00_00_00_00_00_00_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %xmm1, %xmm1, %xmm1
-; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX2-NEXT:    vpbroadcastb %xmm0, %xmm1
 ; AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,0,0,0,0,10,0,0,0,0,0,0,0,0,0,0]
 ; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
 ; AVX2-NEXT:    retq
@@ -232,8 +220,7 @@ define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_
 ;
 ; AVX2-LABEL: shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_11_00_00_00_00_00_00_00_00_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %xmm1, %xmm1, %xmm1
-; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX2-NEXT:    vpbroadcastb %xmm0, %xmm1
 ; AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,0,0,0,11,0,0,0,0,0,0,0,0,0,0,0]
 ; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
 ; AVX2-NEXT:    retq
@@ -252,8 +239,7 @@ define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_
 ;
 ; AVX2-LABEL: shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_12_00_00_00_00_00_00_00_00_00_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %xmm1, %xmm1, %xmm1
-; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX2-NEXT:    vpbroadcastb %xmm0, %xmm1
 ; AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,0,0,12,0,0,0,0,0,0,0,0,0,0,0,0]
 ; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
 ; AVX2-NEXT:    retq
@@ -272,8 +258,7 @@ define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_
 ;
 ; AVX2-LABEL: shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_13_00_00_00_00_00_00_00_00_00_00_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %xmm1, %xmm1, %xmm1
-; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX2-NEXT:    vpbroadcastb %xmm0, %xmm1
 ; AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,0,13,0,0,0,0,0,0,0,0,0,0,0,0,0]
 ; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
 ; AVX2-NEXT:    retq
@@ -292,8 +277,7 @@ define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_
 ;
 ; AVX2-LABEL: shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_14_00_00_00_00_00_00_00_00_00_00_00_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %xmm1, %xmm1, %xmm1
-; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX2-NEXT:    vpbroadcastb %xmm0, %xmm1
 ; AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,14,0,0,0,0,0,0,0,0,0,0,0,0,0,0]
 ; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
 ; AVX2-NEXT:    retq
@@ -317,8 +301,7 @@ define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_
 ; AVX2-NEXT:    movl $15, %eax
 ; AVX2-NEXT:    vmovd %eax, %xmm1
 ; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
-; AVX2-NEXT:    vpxor %xmm2, %xmm2, %xmm2
-; AVX2-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX2-NEXT:    vpbroadcastb %xmm0, %xmm0
 ; AVX2-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm0
 ; AVX2-NEXT:    retq
   %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <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 0, i32 0, i32 15, 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 0>
@@ -339,12 +322,12 @@ define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_16_
 ;
 ; AVX2-LABEL: shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_16_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %ymm1, %ymm1, %ymm1
-; AVX2-NEXT:    vpshufb %ymm1, %ymm0, %ymm2
-; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm0 = ymm0[2,3,0,1]
-; AVX2-NEXT:    vpshufb %ymm1, %ymm0, %ymm0
-; AVX2-NEXT:    vmovdqa {{.*#+}} ymm1 = [0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
-; AVX2-NEXT:    vpblendvb %ymm1, %ymm0, %ymm2, %ymm0
+; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm1 = ymm0[2,3,0,1]
+; AVX2-NEXT:    vpxor %ymm2, %ymm2, %ymm2
+; AVX2-NEXT:    vpshufb %ymm2, %ymm1, %ymm1
+; AVX2-NEXT:    vpbroadcastb %xmm0, %ymm0
+; AVX2-NEXT:    vmovdqa {{.*#+}} ymm2 = [0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; AVX2-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
 ; AVX2-NEXT:    retq
   %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <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 0, i32 16, 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 0, i32 0>
   ret <32 x i8> %shuffle
@@ -364,12 +347,11 @@ define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_17_00_
 ;
 ; AVX2-LABEL: shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_17_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %ymm1, %ymm1, %ymm1
-; AVX2-NEXT:    vpshufb %ymm1, %ymm0, %ymm1
-; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm0 = ymm0[2,3,0,1]
-; AVX2-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[u,u,u,u,u,u,u,u,u,u,u,u,u,u,1,u,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16]
+; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm1 = ymm0[2,3,0,1]
+; AVX2-NEXT:    vpshufb {{.*#+}} ymm1 = ymm1[u,u,u,u,u,u,u,u,u,u,u,u,u,u,1,u,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16]
+; AVX2-NEXT:    vpbroadcastb %xmm0, %ymm0
 ; AVX2-NEXT:    vmovdqa {{.*#+}} ymm2 = [0,0,0,0,0,0,0,0,0,0,0,0,0,0,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
-; AVX2-NEXT:    vpblendvb %ymm2, %ymm0, %ymm1, %ymm0
+; AVX2-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
 ; AVX2-NEXT:    retq
   %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <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 17, 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 0, i32 0, i32 0>
   ret <32 x i8> %shuffle
@@ -389,12 +371,11 @@ define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_18_00_00_
 ;
 ; AVX2-LABEL: shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_18_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %ymm1, %ymm1, %ymm1
-; AVX2-NEXT:    vpshufb %ymm1, %ymm0, %ymm1
-; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm0 = ymm0[2,3,0,1]
-; AVX2-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[u,u,u,u,u,u,u,u,u,u,u,u,u,2,u,u,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16]
+; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm1 = ymm0[2,3,0,1]
+; AVX2-NEXT:    vpshufb {{.*#+}} ymm1 = ymm1[u,u,u,u,u,u,u,u,u,u,u,u,u,2,u,u,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16]
+; AVX2-NEXT:    vpbroadcastb %xmm0, %ymm0
 ; AVX2-NEXT:    vmovdqa {{.*#+}} ymm2 = [0,0,0,0,0,0,0,0,0,0,0,0,0,255,0,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
-; AVX2-NEXT:    vpblendvb %ymm2, %ymm0, %ymm1, %ymm0
+; AVX2-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
 ; AVX2-NEXT:    retq
   %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <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 18, 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 0, i32 0, i32 0, i32 0>
   ret <32 x i8> %shuffle
@@ -414,12 +395,11 @@ define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_19_00_00_00_
 ;
 ; AVX2-LABEL: shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_19_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %ymm1, %ymm1, %ymm1
-; AVX2-NEXT:    vpshufb %ymm1, %ymm0, %ymm1
-; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm0 = ymm0[2,3,0,1]
-; AVX2-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[u,u,u,u,u,u,u,u,u,u,u,u,3,u,u,u,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16]
+; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm1 = ymm0[2,3,0,1]
+; AVX2-NEXT:    vpshufb {{.*#+}} ymm1 = ymm1[u,u,u,u,u,u,u,u,u,u,u,u,3,u,u,u,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16]
+; AVX2-NEXT:    vpbroadcastb %xmm0, %ymm0
 ; AVX2-NEXT:    vmovdqa {{.*#+}} ymm2 = [0,0,0,0,0,0,0,0,0,0,0,0,255,0,0,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
-; AVX2-NEXT:    vpblendvb %ymm2, %ymm0, %ymm1, %ymm0
+; AVX2-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
 ; AVX2-NEXT:    retq
   %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <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 19, 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 0, i32 0, i32 0, i32 0, i32 0>
   ret <32 x i8> %shuffle
@@ -439,12 +419,11 @@ define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_20_00_00_00_00_
 ;
 ; AVX2-LABEL: shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_20_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %ymm1, %ymm1, %ymm1
-; AVX2-NEXT:    vpshufb %ymm1, %ymm0, %ymm1
-; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm0 = ymm0[2,3,0,1]
-; AVX2-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[u,u,u,u,u,u,u,u,u,u,u,4,u,u,u,u,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16]
+; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm1 = ymm0[2,3,0,1]
+; AVX2-NEXT:    vpshufb {{.*#+}} ymm1 = ymm1[u,u,u,u,u,u,u,u,u,u,u,4,u,u,u,u,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16]
+; AVX2-NEXT:    vpbroadcastb %xmm0, %ymm0
 ; AVX2-NEXT:    vmovdqa {{.*#+}} ymm2 = [0,0,0,0,0,0,0,0,0,0,0,255,0,0,0,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
-; AVX2-NEXT:    vpblendvb %ymm2, %ymm0, %ymm1, %ymm0
+; AVX2-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
 ; AVX2-NEXT:    retq
   %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 20, 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 0, i32 0, i32 0, i32 0, i32 0, i32 0>
   ret <32 x i8> %shuffle
@@ -464,12 +443,11 @@ define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_21_00_00_00_00_00_
 ;
 ; AVX2-LABEL: shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_21_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %ymm1, %ymm1, %ymm1
-; AVX2-NEXT:    vpshufb %ymm1, %ymm0, %ymm1
-; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm0 = ymm0[2,3,0,1]
-; AVX2-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[u,u,u,u,u,u,u,u,u,u,5,u,u,u,u,u,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16]
+; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm1 = ymm0[2,3,0,1]
+; AVX2-NEXT:    vpshufb {{.*#+}} ymm1 = ymm1[u,u,u,u,u,u,u,u,u,u,5,u,u,u,u,u,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16]
+; AVX2-NEXT:    vpbroadcastb %xmm0, %ymm0
 ; AVX2-NEXT:    vmovdqa {{.*#+}} ymm2 = [0,0,0,0,0,0,0,0,0,0,255,0,0,0,0,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
-; AVX2-NEXT:    vpblendvb %ymm2, %ymm0, %ymm1, %ymm0
+; AVX2-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
 ; AVX2-NEXT:    retq
   %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 21, 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 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0>
   ret <32 x i8> %shuffle
@@ -489,12 +467,11 @@ define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_00_22_00_00_00_00_00_00_
 ;
 ; AVX2-LABEL: shuffle_v32i8_00_00_00_00_00_00_00_00_00_22_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %ymm1, %ymm1, %ymm1
-; AVX2-NEXT:    vpshufb %ymm1, %ymm0, %ymm1
-; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm0 = ymm0[2,3,0,1]
-; AVX2-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[u,u,u,u,u,u,u,u,u,6,u,u,u,u,u,u,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16]
+; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm1 = ymm0[2,3,0,1]
+; AVX2-NEXT:    vpshufb {{.*#+}} ymm1 = ymm1[u,u,u,u,u,u,u,u,u,6,u,u,u,u,u,u,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16]
+; AVX2-NEXT:    vpbroadcastb %xmm0, %ymm0
 ; AVX2-NEXT:    vmovdqa {{.*#+}} ymm2 = [0,0,0,0,0,0,0,0,0,255,0,0,0,0,0,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
-; AVX2-NEXT:    vpblendvb %ymm2, %ymm0, %ymm1, %ymm0
+; AVX2-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
 ; AVX2-NEXT:    retq
   %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 22, 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 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0>
   ret <32 x i8> %shuffle
@@ -514,12 +491,11 @@ define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_23_00_00_00_00_00_00_00_
 ;
 ; AVX2-LABEL: shuffle_v32i8_00_00_00_00_00_00_00_00_23_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %ymm1, %ymm1, %ymm1
-; AVX2-NEXT:    vpshufb %ymm1, %ymm0, %ymm1
-; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm0 = ymm0[2,3,0,1]
-; AVX2-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[u,u,u,u,u,u,u,u,7,u,u,u,u,u,u,u,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16]
+; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm1 = ymm0[2,3,0,1]
+; AVX2-NEXT:    vpshufb {{.*#+}} ymm1 = ymm1[u,u,u,u,u,u,u,u,7,u,u,u,u,u,u,u,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16]
+; AVX2-NEXT:    vpbroadcastb %xmm0, %ymm0
 ; AVX2-NEXT:    vmovdqa {{.*#+}} ymm2 = [0,0,0,0,0,0,0,0,255,0,0,0,0,0,0,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
-; AVX2-NEXT:    vpblendvb %ymm2, %ymm0, %ymm1, %ymm0
+; AVX2-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
 ; AVX2-NEXT:    retq
   %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 23, 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 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0>
   ret <32 x i8> %shuffle
@@ -539,12 +515,11 @@ define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_24_00_00_00_00_00_00_00_00_
 ;
 ; AVX2-LABEL: shuffle_v32i8_00_00_00_00_00_00_00_24_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %ymm1, %ymm1, %ymm1
-; AVX2-NEXT:    vpshufb %ymm1, %ymm0, %ymm1
-; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm0 = ymm0[2,3,0,1]
-; AVX2-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[u,u,u,u,u,u,u,8,u,u,u,u,u,u,u,u,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16]
+; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm1 = ymm0[2,3,0,1]
+; AVX2-NEXT:    vpshufb {{.*#+}} ymm1 = ymm1[u,u,u,u,u,u,u,8,u,u,u,u,u,u,u,u,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16]
+; AVX2-NEXT:    vpbroadcastb %xmm0, %ymm0
 ; AVX2-NEXT:    vmovdqa {{.*#+}} ymm2 = [0,0,0,0,0,0,0,255,0,0,0,0,0,0,0,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
-; AVX2-NEXT:    vpblendvb %ymm2, %ymm0, %ymm1, %ymm0
+; AVX2-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
 ; AVX2-NEXT:    retq
   %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 24, 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 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0>
   ret <32 x i8> %shuffle
@@ -564,12 +539,11 @@ define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_25_00_00_00_00_00_00_00_00_00_
 ;
 ; AVX2-LABEL: shuffle_v32i8_00_00_00_00_00_00_25_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %ymm1, %ymm1, %ymm1
-; AVX2-NEXT:    vpshufb %ymm1, %ymm0, %ymm1
-; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm0 = ymm0[2,3,0,1]
-; AVX2-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[u,u,u,u,u,u,9,u,u,u,u,u,u,u,u,u,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16]
+; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm1 = ymm0[2,3,0,1]
+; AVX2-NEXT:    vpshufb {{.*#+}} ymm1 = ymm1[u,u,u,u,u,u,9,u,u,u,u,u,u,u,u,u,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16]
+; AVX2-NEXT:    vpbroadcastb %xmm0, %ymm0
 ; AVX2-NEXT:    vmovdqa {{.*#+}} ymm2 = [0,0,0,0,0,0,255,0,0,0,0,0,0,0,0,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
-; AVX2-NEXT:    vpblendvb %ymm2, %ymm0, %ymm1, %ymm0
+; AVX2-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
 ; AVX2-NEXT:    retq
   %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 25, 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 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0>
   ret <32 x i8> %shuffle
@@ -589,12 +563,11 @@ define <32 x i8> @shuffle_v32i8_00_00_00_00_00_26_00_00_00_00_00_00_00_00_00_00_
 ;
 ; AVX2-LABEL: shuffle_v32i8_00_00_00_00_00_26_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %ymm1, %ymm1, %ymm1
-; AVX2-NEXT:    vpshufb %ymm1, %ymm0, %ymm1
-; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm0 = ymm0[2,3,0,1]
-; AVX2-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[u,u,u,u,u,10,u,u,u,u,u,u,u,u,u,u,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16]
+; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm1 = ymm0[2,3,0,1]
+; AVX2-NEXT:    vpshufb {{.*#+}} ymm1 = ymm1[u,u,u,u,u,10,u,u,u,u,u,u,u,u,u,u,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16]
+; AVX2-NEXT:    vpbroadcastb %xmm0, %ymm0
 ; AVX2-NEXT:    vmovdqa {{.*#+}} ymm2 = [0,0,0,0,0,255,0,0,0,0,0,0,0,0,0,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
-; AVX2-NEXT:    vpblendvb %ymm2, %ymm0, %ymm1, %ymm0
+; AVX2-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
 ; AVX2-NEXT:    retq
   %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <i32 0, i32 0, i32 0, i32 0, i32 0, i32 26, 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 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0>
   ret <32 x i8> %shuffle
@@ -614,12 +587,11 @@ define <32 x i8> @shuffle_v32i8_00_00_00_00_27_00_00_00_00_00_00_00_00_00_00_00_
 ;
 ; AVX2-LABEL: shuffle_v32i8_00_00_00_00_27_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %ymm1, %ymm1, %ymm1
-; AVX2-NEXT:    vpshufb %ymm1, %ymm0, %ymm1
-; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm0 = ymm0[2,3,0,1]
-; AVX2-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[u,u,u,u,11,u,u,u,u,u,u,u,u,u,u,u,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16]
+; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm1 = ymm0[2,3,0,1]
+; AVX2-NEXT:    vpshufb {{.*#+}} ymm1 = ymm1[u,u,u,u,11,u,u,u,u,u,u,u,u,u,u,u,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16]
+; AVX2-NEXT:    vpbroadcastb %xmm0, %ymm0
 ; AVX2-NEXT:    vmovdqa {{.*#+}} ymm2 = [0,0,0,0,255,0,0,0,0,0,0,0,0,0,0,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
-; AVX2-NEXT:    vpblendvb %ymm2, %ymm0, %ymm1, %ymm0
+; AVX2-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
 ; AVX2-NEXT:    retq
   %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <i32 0, i32 0, i32 0, i32 0, i32 27, 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 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>
   ret <32 x i8> %shuffle
@@ -639,12 +611,11 @@ define <32 x i8> @shuffle_v32i8_00_00_00_28_00_00_00_00_00_00_00_00_00_00_00_00_
 ;
 ; AVX2-LABEL: shuffle_v32i8_00_00_00_28_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %ymm1, %ymm1, %ymm1
-; AVX2-NEXT:    vpshufb %ymm1, %ymm0, %ymm1
-; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm0 = ymm0[2,3,0,1]
-; AVX2-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[u,u,u,12,u,u,u,u,u,u,u,u,u,u,u,u,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16]
+; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm1 = ymm0[2,3,0,1]
+; AVX2-NEXT:    vpshufb {{.*#+}} ymm1 = ymm1[u,u,u,12,u,u,u,u,u,u,u,u,u,u,u,u,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16]
+; AVX2-NEXT:    vpbroadcastb %xmm0, %ymm0
 ; AVX2-NEXT:    vmovdqa {{.*#+}} ymm2 = [0,0,0,255,0,0,0,0,0,0,0,0,0,0,0,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
-; AVX2-NEXT:    vpblendvb %ymm2, %ymm0, %ymm1, %ymm0
+; AVX2-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
 ; AVX2-NEXT:    retq
   %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <i32 0, i32 0, i32 0, i32 28, 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 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>
   ret <32 x i8> %shuffle
@@ -664,12 +635,11 @@ define <32 x i8> @shuffle_v32i8_00_00_29_00_00_00_00_00_00_00_00_00_00_00_00_00_
 ;
 ; AVX2-LABEL: shuffle_v32i8_00_00_29_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %ymm1, %ymm1, %ymm1
-; AVX2-NEXT:    vpshufb %ymm1, %ymm0, %ymm1
-; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm0 = ymm0[2,3,0,1]
-; AVX2-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[u,u,13,u,u,u,u,u,u,u,u,u,u,u,u,u,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16]
+; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm1 = ymm0[2,3,0,1]
+; AVX2-NEXT:    vpshufb {{.*#+}} ymm1 = ymm1[u,u,13,u,u,u,u,u,u,u,u,u,u,u,u,u,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16]
+; AVX2-NEXT:    vpbroadcastb %xmm0, %ymm0
 ; AVX2-NEXT:    vmovdqa {{.*#+}} ymm2 = [0,0,255,0,0,0,0,0,0,0,0,0,0,0,0,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
-; AVX2-NEXT:    vpblendvb %ymm2, %ymm0, %ymm1, %ymm0
+; AVX2-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
 ; AVX2-NEXT:    retq
   %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <i32 0, i32 0, i32 29, 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 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 0>
   ret <32 x i8> %shuffle
@@ -689,12 +659,11 @@ define <32 x i8> @shuffle_v32i8_00_30_00_00_00_00_00_00_00_00_00_00_00_00_00_00_
 ;
 ; AVX2-LABEL: shuffle_v32i8_00_30_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %ymm1, %ymm1, %ymm1
-; AVX2-NEXT:    vpshufb %ymm1, %ymm0, %ymm1
-; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm0 = ymm0[2,3,0,1]
-; AVX2-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[u,14,u,u,u,u,u,u,u,u,u,u,u,u,u,u,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16]
+; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm1 = ymm0[2,3,0,1]
+; AVX2-NEXT:    vpshufb {{.*#+}} ymm1 = ymm1[u,14,u,u,u,u,u,u,u,u,u,u,u,u,u,u,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16]
+; AVX2-NEXT:    vpbroadcastb %xmm0, %ymm0
 ; AVX2-NEXT:    vmovdqa {{.*#+}} ymm2 = [0,255,0,0,0,0,0,0,0,0,0,0,0,0,0,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
-; AVX2-NEXT:    vpblendvb %ymm2, %ymm0, %ymm1, %ymm0
+; AVX2-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
 ; AVX2-NEXT:    retq
   %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <i32 0, i32 30, 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 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 0, i32 0>
   ret <32 x i8> %shuffle
@@ -722,7 +691,7 @@ define <32 x i8> @shuffle_v32i8_31_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_
 ; AVX2-NEXT:    vpxor %ymm3, %ymm3, %ymm3
 ; AVX2-NEXT:    vinserti128 $0, %xmm2, %ymm3, %ymm2
 ; AVX2-NEXT:    vpshufb %ymm2, %ymm1, %ymm1
-; AVX2-NEXT:    vpshufb %ymm3, %ymm0, %ymm0
+; AVX2-NEXT:    vpbroadcastb %xmm0, %ymm0
 ; AVX2-NEXT:    vmovdqa {{.*#+}} ymm2 = [255,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
 ; AVX2-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
 ; AVX2-NEXT:    retq
@@ -1037,9 +1006,10 @@ define <32 x i8> @shuffle_v32i8_00_32_00_32_00_32_00_32_00_32_00_32_00_32_00_32_
 ;
 ; AVX2-LABEL: shuffle_v32i8_00_32_00_32_00_32_00_32_00_32_00_32_00_32_00_32_00_32_00_32_00_32_00_32_00_32_00_32_00_32_00_32:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %xmm2, %xmm2, %xmm2
-; AVX2-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
-; AVX2-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX2-NEXT:    vpbroadcastb %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb {{.*#+}} xmm1 = xmm1[1,3,5,7,9,11,13,15,u,u,u,u,u,u,u,u]
+; AVX2-NEXT:    vpbroadcastw %xmm0, %xmm0
+; AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u]
 ; AVX2-NEXT:    vpunpcklbw {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3],xmm0[4],xmm1[4],xmm0[5],xmm1[5],xmm0[6],xmm1[6],xmm0[7],xmm1[7]
 ; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
 ; AVX2-NEXT:    retq
diff --git a/llvm/test/CodeGen/X86/vector-shuffle-256-v4.ll b/llvm/test/CodeGen/X86/vector-shuffle-256-v4.ll
index 74662ac82e4..5c5bc21a48f 100644
--- a/llvm/test/CodeGen/X86/vector-shuffle-256-v4.ll
+++ b/llvm/test/CodeGen/X86/vector-shuffle-256-v4.ll
@@ -12,7 +12,7 @@ define <4 x double> @shuffle_v4f64_0000(<4 x double> %a, <4 x double> %b) {
 ;
 ; AVX2-LABEL: shuffle_v4f64_0000:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,0,0,0]
+; AVX2-NEXT:    vbroadcastsd %xmm0, %ymm0
 ; AVX2-NEXT:    retq
   %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 0, i32 0, i32 0, i32 0>
   ret <4 x double> %shuffle
@@ -184,11 +184,17 @@ define <4 x double> @shuffle_v4f64_1022(<4 x double> %a, <4 x double> %b) {
 }
 
 define <4 x double> @shuffle_v4f64_0423(<4 x double> %a, <4 x double> %b) {
-; ALL-LABEL: shuffle_v4f64_0423:
-; ALL:       # BB#0:
-; ALL-NEXT:    vpermilpd {{.*#+}} ymm1 = ymm1[0,0,2,2]
-; ALL-NEXT:    vblendpd {{.*#+}} ymm0 = ymm0[0],ymm1[1],ymm0[2,3]
-; ALL-NEXT:    retq
+; AVX1-LABEL: shuffle_v4f64_0423:
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpermilpd {{.*#+}} ymm1 = ymm1[0,0,2,2]
+; AVX1-NEXT:    vblendpd {{.*#+}} ymm0 = ymm0[0],ymm1[1],ymm0[2,3]
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: shuffle_v4f64_0423:
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vbroadcastsd %xmm1, %ymm1
+; AVX2-NEXT:    vblendpd {{.*#+}} ymm0 = ymm0[0],ymm1[1],ymm0[2,3]
+; AVX2-NEXT:    retq
   %shuffle = shufflevector <4 x double> %a, <4 x double> %b, <4 x i32> <i32 0, i32 4, i32 2, i32 3>
   ret <4 x double> %shuffle
 }
@@ -317,7 +323,7 @@ define <4 x i64> @shuffle_v4i64_0000(<4 x i64> %a, <4 x i64> %b) {
 ;
 ; AVX2-LABEL: shuffle_v4i64_0000:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpermq {{.*#+}} ymm0 = ymm0[0,0,0,0]
+; AVX2-NEXT:    vbroadcastsd %xmm0, %ymm0
 ; AVX2-NEXT:    retq
   %shuffle = shufflevector <4 x i64> %a, <4 x i64> %b, <4 x i32> <i32 0, i32 0, i32 0, i32 0>
   ret <4 x i64> %shuffle
@@ -460,7 +466,7 @@ define <4 x i64> @shuffle_v4i64_0124(<4 x i64> %a, <4 x i64> %b) {
 ;
 ; AVX2-LABEL: shuffle_v4i64_0124:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpermq {{.*#+}} ymm1 = ymm1[0,1,2,0]
+; AVX2-NEXT:    vpbroadcastq %xmm1, %ymm1
 ; AVX2-NEXT:    vpblendd {{.*#+}} ymm0 = ymm0[0,1,2,3,4,5],ymm1[6,7]
 ; AVX2-NEXT:    retq
   %shuffle = shufflevector <4 x i64> %a, <4 x i64> %b, <4 x i32> <i32 0, i32 1, i32 2, i32 4>
@@ -477,8 +483,8 @@ define <4 x i64> @shuffle_v4i64_0142(<4 x i64> %a, <4 x i64> %b) {
 ;
 ; AVX2-LABEL: shuffle_v4i64_0142:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpermq {{.*#+}} ymm1 = ymm1[0,1,0,3]
 ; AVX2-NEXT:    vpermq {{.*#+}} ymm0 = ymm0[0,1,2,2]
+; AVX2-NEXT:    vpbroadcastq %xmm1, %ymm1
 ; AVX2-NEXT:    vpblendd {{.*#+}} ymm0 = ymm0[0,1,2,3],ymm1[4,5],ymm0[6,7]
 ; AVX2-NEXT:    retq
   %shuffle = shufflevector <4 x i64> %a, <4 x i64> %b, <4 x i32> <i32 0, i32 1, i32 4, i32 2>
@@ -497,8 +503,8 @@ define <4 x i64> @shuffle_v4i64_0412(<4 x i64> %a, <4 x i64> %b) {
 ;
 ; AVX2-LABEL: shuffle_v4i64_0412:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpshufd {{.*#+}} ymm1 = ymm1[0,1,0,1,4,5,4,5]
 ; AVX2-NEXT:    vpermq {{.*#+}} ymm0 = ymm0[0,1,1,2]
+; AVX2-NEXT:    vpbroadcastq %xmm1, %ymm1
 ; AVX2-NEXT:    vpblendd {{.*#+}} ymm0 = ymm0[0,1],ymm1[2,3],ymm0[4,5,6,7]
 ; AVX2-NEXT:    retq
   %shuffle = shufflevector <4 x i64> %a, <4 x i64> %b, <4 x i32> <i32 0, i32 4, i32 1, i32 2>
@@ -731,3 +737,31 @@ define <4 x double> @insert_mem_and_zero_v4f64(double* %ptr) {
   %shuffle = shufflevector <4 x double> %v, <4 x double> zeroinitializer, <4 x i32> <i32 0, i32 5, i32 6, i32 7>
   ret <4 x double> %shuffle
 }
+
+define <4 x double> @splat_mem_v4f64(double* %ptr) {
+; ALL-LABEL: splat_mem_v4f64:
+; ALL:       # BB#0:
+; ALL-NEXT:    vbroadcastsd (%rdi), %ymm0
+; ALL-NEXT:    retq
+  %a = load double* %ptr
+  %v = insertelement <4 x double> undef, double %a, i32 0
+  %shuffle = shufflevector <4 x double> %v, <4 x double> undef, <4 x i32> <i32 0, i32 0, i32 0, i32 0>
+  ret <4 x double> %shuffle
+}
+
+define <4 x i64> @splat_mem_v4i64(i64* %ptr) {
+; AVX1-LABEL: splat_mem_v4i64:
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vmovddup (%rdi), %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: splat_mem_v4i64:
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vbroadcastsd (%rdi), %ymm0
+; AVX2-NEXT:    retq
+  %a = load i64* %ptr
+  %v = insertelement <4 x i64> undef, i64 %a, i64 0
+  %shuffle = shufflevector <4 x i64> %v, <4 x i64> undef, <4 x i32> <i32 0, i32 0, i32 0, i32 0>
+  ret <4 x i64> %shuffle
+}
diff --git a/llvm/test/CodeGen/X86/vector-shuffle-256-v8.ll b/llvm/test/CodeGen/X86/vector-shuffle-256-v8.ll
index b9a2de0298b..f11554cc2fc 100644
--- a/llvm/test/CodeGen/X86/vector-shuffle-256-v8.ll
+++ b/llvm/test/CodeGen/X86/vector-shuffle-256-v8.ll
@@ -12,8 +12,7 @@ define <8 x float> @shuffle_v8f32_00000000(<8 x float> %a, <8 x float> %b) {
 ;
 ; AVX2-LABEL: shuffle_v8f32_00000000:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vxorps %ymm1, %ymm1, %ymm1
-; AVX2-NEXT:    vpermps %ymm0, %ymm1, %ymm0
+; AVX2-NEXT:    vbroadcastss %xmm0, %ymm0
 ; AVX2-NEXT:    retq
   %shuffle = shufflevector <8 x float> %a, <8 x float> %b, <8 x i32> <i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0>
   ret <8 x float> %shuffle
@@ -210,9 +209,8 @@ define <8 x float> @shuffle_v8f32_08080808(<8 x float> %a, <8 x float> %b) {
 ;
 ; AVX2-LABEL: shuffle_v8f32_08080808:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vxorps %ymm2, %ymm2, %ymm2
-; AVX2-NEXT:    vpermps %ymm1, %ymm2, %ymm1
-; AVX2-NEXT:    vpermps %ymm0, %ymm2, %ymm0
+; AVX2-NEXT:    vbroadcastss %xmm1, %ymm1
+; AVX2-NEXT:    vbroadcastss %xmm0, %ymm0
 ; AVX2-NEXT:    vblendps {{.*#+}} ymm0 = ymm0[0],ymm1[1],ymm0[2],ymm1[3],ymm0[4],ymm1[5],ymm0[6],ymm1[7]
 ; AVX2-NEXT:    retq
   %shuffle = shufflevector <8 x float> %a, <8 x float> %b, <8 x i32> <i32 0, i32 8, i32 0, i32 8, i32 0, i32 8, i32 0, i32 8>
@@ -702,8 +700,7 @@ define <8 x i32> @shuffle_v8i32_00000000(<8 x i32> %a, <8 x i32> %b) {
 ;
 ; AVX2-LABEL: shuffle_v8i32_00000000:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %ymm1, %ymm1, %ymm1
-; AVX2-NEXT:    vpermd %ymm0, %ymm1, %ymm0
+; AVX2-NEXT:    vbroadcastss %xmm0, %ymm0
 ; AVX2-NEXT:    retq
   %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0>
   ret <8 x i32> %shuffle
@@ -910,9 +907,8 @@ define <8 x i32> @shuffle_v8i32_08080808(<8 x i32> %a, <8 x i32> %b) {
 ;
 ; AVX2-LABEL: shuffle_v8i32_08080808:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vpxor %ymm2, %ymm2, %ymm2
-; AVX2-NEXT:    vpermd %ymm1, %ymm2, %ymm1
-; AVX2-NEXT:    vpermq {{.*#+}} ymm0 = ymm0[0,0,0,0]
+; AVX2-NEXT:    vpbroadcastd %xmm1, %ymm1
+; AVX2-NEXT:    vpbroadcastq %xmm0, %ymm0
 ; AVX2-NEXT:    vpblendd {{.*#+}} ymm0 = ymm0[0],ymm1[1],ymm0[2],ymm1[3],ymm0[4],ymm1[5],ymm0[6],ymm1[7]
 ; AVX2-NEXT:    retq
   %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 0, i32 8, i32 0, i32 8, i32 0, i32 8, i32 0, i32 8>
diff --git a/llvm/test/CodeGen/X86/vector-shuffle-512-v8.ll b/llvm/test/CodeGen/X86/vector-shuffle-512-v8.ll
index 09dde3132dd..2f02f2fc08f 100644
--- a/llvm/test/CodeGen/X86/vector-shuffle-512-v8.ll
+++ b/llvm/test/CodeGen/X86/vector-shuffle-512-v8.ll
@@ -6,7 +6,7 @@ target triple = "x86_64-unknown-unknown"
 define <8 x double> @shuffle_v8f64_00000000(<8 x double> %a, <8 x double> %b) {
 ; ALL-LABEL: shuffle_v8f64_00000000:
 ; ALL:       # BB#0:
-; ALL-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,0,0,0]
+; ALL-NEXT:    vbroadcastsd %xmm0, %ymm0
 ; ALL-NEXT:    vinsertf64x4 $1, %ymm0, %zmm0, %zmm0
 ; ALL-NEXT:    retq
   %shuffle = shufflevector <8 x double> %a, <8 x double> %b, <8 x i32> <i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0>
@@ -16,7 +16,7 @@ define <8 x double> @shuffle_v8f64_00000000(<8 x double> %a, <8 x double> %b) {
 define <8 x double> @shuffle_v8f64_00000010(<8 x double> %a, <8 x double> %b) {
 ; ALL-LABEL: shuffle_v8f64_00000010:
 ; ALL:       # BB#0:
-; ALL-NEXT:    vpermpd {{.*#+}} ymm1 = ymm0[0,0,0,0]
+; ALL-NEXT:    vbroadcastsd %xmm0, %ymm1
 ; ALL-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,0,1,0]
 ; ALL-NEXT:    vinsertf64x4 $1, %ymm0, %zmm1, %zmm0
 ; ALL-NEXT:    retq
@@ -27,7 +27,7 @@ define <8 x double> @shuffle_v8f64_00000010(<8 x double> %a, <8 x double> %b) {
 define <8 x double> @shuffle_v8f64_00000200(<8 x double> %a, <8 x double> %b) {
 ; ALL-LABEL: shuffle_v8f64_00000200:
 ; ALL:       # BB#0:
-; ALL-NEXT:    vpermpd {{.*#+}} ymm1 = ymm0[0,0,0,0]
+; ALL-NEXT:    vbroadcastsd %xmm0, %ymm1
 ; ALL-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,2,0,0]
 ; ALL-NEXT:    vinsertf64x4 $1, %ymm0, %zmm1, %zmm0
 ; ALL-NEXT:    retq
@@ -38,7 +38,7 @@ define <8 x double> @shuffle_v8f64_00000200(<8 x double> %a, <8 x double> %b) {
 define <8 x double> @shuffle_v8f64_00003000(<8 x double> %a, <8 x double> %b) {
 ; ALL-LABEL: shuffle_v8f64_00003000:
 ; ALL:       # BB#0:
-; ALL-NEXT:    vpermpd {{.*#+}} ymm1 = ymm0[0,0,0,0]
+; ALL-NEXT:    vbroadcastsd %xmm0, %ymm1
 ; ALL-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[3,0,0,0]
 ; ALL-NEXT:    vinsertf64x4 $1, %ymm0, %zmm1, %zmm0
 ; ALL-NEXT:    retq
@@ -50,10 +50,9 @@ define <8 x double> @shuffle_v8f64_00040000(<8 x double> %a, <8 x double> %b) {
 ; ALL-LABEL: shuffle_v8f64_00040000:
 ; ALL:       # BB#0:
 ; ALL-NEXT:    vextractf64x4 $1, %zmm0, %ymm1
-; ALL-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[0,1,2,0]
-; ALL-NEXT:    vpermpd {{.*#+}} ymm2 = ymm0[0,0,0,3]
-; ALL-NEXT:    vblendpd {{.*#+}} ymm1 = ymm2[0,1,2],ymm1[3]
-; ALL-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,0,0,0]
+; ALL-NEXT:    vbroadcastsd %xmm1, %ymm1
+; ALL-NEXT:    vbroadcastsd %xmm0, %ymm0
+; ALL-NEXT:    vblendpd {{.*#+}} ymm1 = ymm0[0,1,2],ymm1[3]
 ; ALL-NEXT:    vinsertf64x4 $1, %ymm0, %zmm1, %zmm0
 ; ALL-NEXT:    retq
   %shuffle = shufflevector <8 x double> %a, <8 x double> %b, <8 x i32> <i32 0, i32 0, i32 0, i32 4, i32 0, i32 0, i32 0, i32 0>
@@ -65,9 +64,8 @@ define <8 x double> @shuffle_v8f64_00500000(<8 x double> %a, <8 x double> %b) {
 ; ALL:       # BB#0:
 ; ALL-NEXT:    vextractf64x4 $1, %zmm0, %ymm1
 ; ALL-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[0,1,1,3]
-; ALL-NEXT:    vpermpd {{.*#+}} ymm2 = ymm0[0,0,2,0]
-; ALL-NEXT:    vblendpd {{.*#+}} ymm1 = ymm2[0,1],ymm1[2],ymm2[3]
-; ALL-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,0,0,0]
+; ALL-NEXT:    vbroadcastsd %xmm0, %ymm0
+; ALL-NEXT:    vblendpd {{.*#+}} ymm1 = ymm0[0,1],ymm1[2],ymm0[3]
 ; ALL-NEXT:    vinsertf64x4 $1, %ymm0, %zmm1, %zmm0
 ; ALL-NEXT:    retq
   %shuffle = shufflevector <8 x double> %a, <8 x double> %b, <8 x i32> <i32 0, i32 0, i32 5, i32 0, i32 0, i32 0, i32 0, i32 0>
@@ -79,9 +77,8 @@ define <8 x double> @shuffle_v8f64_06000000(<8 x double> %a, <8 x double> %b) {
 ; ALL:       # BB#0:
 ; ALL-NEXT:    vextractf64x4 $1, %zmm0, %ymm1
 ; ALL-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[0,2,2,3]
-; ALL-NEXT:    vpermpd {{.*#+}} ymm2 = ymm0[0,1,0,0]
-; ALL-NEXT:    vblendpd {{.*#+}} ymm1 = ymm2[0],ymm1[1],ymm2[2,3]
-; ALL-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,0,0,0]
+; ALL-NEXT:    vbroadcastsd %xmm0, %ymm0
+; ALL-NEXT:    vblendpd {{.*#+}} ymm1 = ymm0[0],ymm1[1],ymm0[2,3]
 ; ALL-NEXT:    vinsertf64x4 $1, %ymm0, %zmm1, %zmm0
 ; ALL-NEXT:    retq
   %shuffle = shufflevector <8 x double> %a, <8 x double> %b, <8 x i32> <i32 0, i32 6, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0>
@@ -93,7 +90,7 @@ define <8 x double> @shuffle_v8f64_70000000(<8 x double> %a, <8 x double> %b) {
 ; ALL:       # BB#0:
 ; ALL-NEXT:    vextractf64x4 $1, %zmm0, %ymm1
 ; ALL-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[3,1,2,3]
-; ALL-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,0,0,0]
+; ALL-NEXT:    vbroadcastsd %xmm0, %ymm0
 ; ALL-NEXT:    vblendpd {{.*#+}} ymm1 = ymm1[0],ymm0[1,2,3]
 ; ALL-NEXT:    vinsertf64x4 $1, %ymm0, %zmm1, %zmm0
 ; ALL-NEXT:    retq
@@ -127,7 +124,7 @@ define <8 x double> @shuffle_v8f64_00112233(<8 x double> %a, <8 x double> %b) {
 define <8 x double> @shuffle_v8f64_00001111(<8 x double> %a, <8 x double> %b) {
 ; ALL-LABEL: shuffle_v8f64_00001111:
 ; ALL:       # BB#0:
-; ALL-NEXT:    vpermpd {{.*#+}} ymm1 = ymm0[0,0,0,0]
+; ALL-NEXT:    vbroadcastsd %xmm0, %ymm1
 ; ALL-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[1,1,1,1]
 ; ALL-NEXT:    vinsertf64x4 $1, %ymm0, %zmm1, %zmm0
 ; ALL-NEXT:    retq
@@ -151,8 +148,8 @@ define <8 x double> @shuffle_v8f64_81a3c5e7(<8 x double> %a, <8 x double> %b) {
 define <8 x double> @shuffle_v8f64_08080808(<8 x double> %a, <8 x double> %b) {
 ; ALL-LABEL: shuffle_v8f64_08080808:
 ; ALL:       # BB#0:
-; ALL-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[0,0,2,0]
-; ALL-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,1,0,3]
+; ALL-NEXT:    vbroadcastsd %xmm1, %ymm1
+; ALL-NEXT:    vbroadcastsd %xmm0, %ymm0
 ; ALL-NEXT:    vblendpd {{.*#+}} ymm0 = ymm0[0],ymm1[1],ymm0[2],ymm1[3]
 ; ALL-NEXT:    vinsertf64x4 $1, %ymm0, %zmm0, %zmm0
 ; ALL-NEXT:    retq
@@ -164,12 +161,12 @@ define <8 x double> @shuffle_v8f64_08084c4c(<8 x double> %a, <8 x double> %b) {
 ; ALL-LABEL: shuffle_v8f64_08084c4c:
 ; ALL:       # BB#0:
 ; ALL-NEXT:    vextractf64x4 $1, %zmm1, %ymm2
-; ALL-NEXT:    vpermpd {{.*#+}} ymm2 = ymm2[0,0,2,0]
+; ALL-NEXT:    vbroadcastsd %xmm2, %ymm2
 ; ALL-NEXT:    vextractf64x4 $1, %zmm0, %ymm3
-; ALL-NEXT:    vpermpd {{.*#+}} ymm3 = ymm3[0,1,0,3]
+; ALL-NEXT:    vbroadcastsd %xmm3, %ymm3
 ; ALL-NEXT:    vblendpd {{.*#+}} ymm2 = ymm3[0],ymm2[1],ymm3[2],ymm2[3]
-; ALL-NEXT:    vpermpd {{.*#+}} ymm1 = ymm1[0,0,2,0]
-; ALL-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,1,0,3]
+; ALL-NEXT:    vbroadcastsd %xmm1, %ymm1
+; ALL-NEXT:    vbroadcastsd %xmm0, %ymm0
 ; ALL-NEXT:    vblendpd {{.*#+}} ymm0 = ymm0[0],ymm1[1],ymm0[2],ymm1[3]
 ; ALL-NEXT:    vinsertf64x4 $1, %ymm2, %zmm0, %zmm0
 ; ALL-NEXT:    retq
@@ -182,9 +179,9 @@ define <8 x double> @shuffle_v8f64_8823cc67(<8 x double> %a, <8 x double> %b) {
 ; ALL:       # BB#0:
 ; ALL-NEXT:    vextractf64x4 $1, %zmm0, %ymm2
 ; ALL-NEXT:    vextractf64x4 $1, %zmm1, %ymm3
-; ALL-NEXT:    vpermilpd {{.*#+}} ymm3 = ymm3[0,0,2,2]
+; ALL-NEXT:    vbroadcastsd %xmm3, %ymm3
 ; ALL-NEXT:    vblendpd {{.*#+}} ymm2 = ymm3[0,1],ymm2[2,3]
-; ALL-NEXT:    vpermilpd {{.*#+}} ymm1 = ymm1[0,0,2,2]
+; ALL-NEXT:    vbroadcastsd %xmm1, %ymm1
 ; ALL-NEXT:    vblendpd {{.*#+}} ymm0 = ymm1[0,1],ymm0[2,3]
 ; ALL-NEXT:    vinsertf64x4 $1, %ymm2, %zmm0, %zmm0
 ; ALL-NEXT:    retq
@@ -513,7 +510,7 @@ define <8 x double> @shuffle_v8f64_10004444(<8 x double> %a, <8 x double> %b) {
 ; ALL:       # BB#0:
 ; ALL-NEXT:    vpermpd {{.*#+}} ymm1 = ymm0[1,0,0,0]
 ; ALL-NEXT:    vextractf64x4 $1, %zmm0, %ymm0
-; ALL-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,0,0,0]
+; ALL-NEXT:    vbroadcastsd %xmm0, %ymm0
 ; ALL-NEXT:    vinsertf64x4 $1, %ymm0, %zmm1, %zmm0
 ; ALL-NEXT:    retq
   %shuffle = shufflevector <8 x double> %a, <8 x double> %b, <8 x i32> <i32 1, i32 0, i32 0, i32 0, i32 4, i32 4, i32 4, i32 4>
@@ -638,7 +635,7 @@ define <8 x double> @shuffle_v8f64_002u6u44(<8 x double> %a, <8 x double> %b) {
 define <8 x double> @shuffle_v8f64_00uu66uu(<8 x double> %a, <8 x double> %b) {
 ; ALL-LABEL: shuffle_v8f64_00uu66uu:
 ; ALL:       # BB#0:
-; ALL-NEXT:    vpermilpd {{.*#+}} ymm1 = ymm0[0,0,2,2]
+; ALL-NEXT:    vbroadcastsd %xmm0, %ymm1
 ; ALL-NEXT:    vextractf64x4 $1, %zmm0, %ymm0
 ; ALL-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[2,2,2,3]
 ; ALL-NEXT:    vinsertf64x4 $1, %ymm0, %zmm1, %zmm0
@@ -694,16 +691,16 @@ define <8 x double> @shuffle_v8f64_uuu3uu66(<8 x double> %a, <8 x double> %b) {
 define <8 x double> @shuffle_v8f64_c348cda0(<8 x double> %a, <8 x double> %b) {
 ; ALL-LABEL: shuffle_v8f64_c348cda0:
 ; ALL:       # BB#0:
-; ALL-NEXT:    vextractf64x4 $1, %zmm0, %ymm2
-; ALL-NEXT:    vpermpd {{.*#+}} ymm2 = ymm2[0,1,0,3]
-; ALL-NEXT:    vpermpd {{.*#+}} ymm3 = ymm0[0,3,2,3]
-; ALL-NEXT:    vblendpd {{.*#+}} ymm2 = ymm3[0,1],ymm2[2],ymm3[3]
+; ALL-NEXT:    vpermpd {{.*#+}} ymm2 = ymm0[0,3,2,3]
+; ALL-NEXT:    vextractf64x4 $1, %zmm0, %ymm3
+; ALL-NEXT:    vbroadcastsd %xmm3, %ymm3
+; ALL-NEXT:    vblendpd {{.*#+}} ymm2 = ymm2[0,1],ymm3[2],ymm2[3]
 ; ALL-NEXT:    vextractf64x4 $1, %zmm1, %ymm3
-; ALL-NEXT:    vpermpd {{.*#+}} ymm4 = ymm1[0,1,2,0]
+; ALL-NEXT:    vbroadcastsd %xmm1, %ymm4
 ; ALL-NEXT:    vblendpd {{.*#+}} ymm4 = ymm3[0,1,2],ymm4[3]
 ; ALL-NEXT:    vblendpd {{.*#+}} ymm2 = ymm4[0],ymm2[1,2],ymm4[3]
 ; ALL-NEXT:    vblendpd {{.*#+}} ymm1 = ymm3[0,1],ymm1[2],ymm3[3]
-; ALL-NEXT:    vpermpd {{.*#+}} ymm0 = ymm0[0,1,2,0]
+; ALL-NEXT:    vbroadcastsd %xmm0, %ymm0
 ; ALL-NEXT:    vblendpd {{.*#+}} ymm0 = ymm1[0,1,2],ymm0[3]
 ; ALL-NEXT:    vinsertf64x4 $1, %ymm0, %zmm2, %zmm0
 ; ALL-NEXT:    retq
@@ -734,7 +731,7 @@ define <8 x double> @shuffle_v8f64_f511235a(<8 x double> %a, <8 x double> %b) {
 define <8 x i64> @shuffle_v8i64_00000000(<8 x i64> %a, <8 x i64> %b) {
 ; ALL-LABEL: shuffle_v8i64_00000000:
 ; ALL:       # BB#0:
-; ALL-NEXT:    vpermq {{.*#+}} ymm0 = ymm0[0,0,0,0]
+; ALL-NEXT:    vpbroadcastq %xmm0, %ymm0
 ; ALL-NEXT:    vinserti64x4 $1, %ymm0, %zmm0, %zmm0
 ; ALL-NEXT:    retq
   %shuffle = shufflevector <8 x i64> %a, <8 x i64> %b, <8 x i32> <i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0>
@@ -744,7 +741,7 @@ define <8 x i64> @shuffle_v8i64_00000000(<8 x i64> %a, <8 x i64> %b) {
 define <8 x i64> @shuffle_v8i64_00000010(<8 x i64> %a, <8 x i64> %b) {
 ; ALL-LABEL: shuffle_v8i64_00000010:
 ; ALL:       # BB#0:
-; ALL-NEXT:    vpermq {{.*#+}} ymm1 = ymm0[0,0,0,0]
+; ALL-NEXT:    vpbroadcastq %xmm0, %ymm1
 ; ALL-NEXT:    vpermq {{.*#+}} ymm0 = ymm0[0,0,1,0]
 ; ALL-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
 ; ALL-NEXT:    retq
@@ -755,7 +752,7 @@ define <8 x i64> @shuffle_v8i64_00000010(<8 x i64> %a, <8 x i64> %b) {
 define <8 x i64> @shuffle_v8i64_00000200(<8 x i64> %a, <8 x i64> %b) {
 ; ALL-LABEL: shuffle_v8i64_00000200:
 ; ALL:       # BB#0:
-; ALL-NEXT:    vpermq {{.*#+}} ymm1 = ymm0[0,0,0,0]
+; ALL-NEXT:    vpbroadcastq %xmm0, %ymm1
 ; ALL-NEXT:    vpermq {{.*#+}} ymm0 = ymm0[0,2,0,0]
 ; ALL-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
 ; ALL-NEXT:    retq
@@ -766,7 +763,7 @@ define <8 x i64> @shuffle_v8i64_00000200(<8 x i64> %a, <8 x i64> %b) {
 define <8 x i64> @shuffle_v8i64_00003000(<8 x i64> %a, <8 x i64> %b) {
 ; ALL-LABEL: shuffle_v8i64_00003000:
 ; ALL:       # BB#0:
-; ALL-NEXT:    vpermq {{.*#+}} ymm1 = ymm0[0,0,0,0]
+; ALL-NEXT:    vpbroadcastq %xmm0, %ymm1
 ; ALL-NEXT:    vpermq {{.*#+}} ymm0 = ymm0[3,0,0,0]
 ; ALL-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
 ; ALL-NEXT:    retq
@@ -778,10 +775,9 @@ define <8 x i64> @shuffle_v8i64_00040000(<8 x i64> %a, <8 x i64> %b) {
 ; ALL-LABEL: shuffle_v8i64_00040000:
 ; ALL:       # BB#0:
 ; ALL-NEXT:    vextracti64x4 $1, %zmm0, %ymm1
-; ALL-NEXT:    vpermq {{.*#+}} ymm1 = ymm1[0,1,2,0]
-; ALL-NEXT:    vpermq {{.*#+}} ymm2 = ymm0[0,0,0,3]
-; ALL-NEXT:    vpblendd {{.*#+}} ymm1 = ymm2[0,1,2,3,4,5],ymm1[6,7]
-; ALL-NEXT:    vpermq {{.*#+}} ymm0 = ymm0[0,0,0,0]
+; ALL-NEXT:    vpbroadcastq %xmm1, %ymm1
+; ALL-NEXT:    vpbroadcastq %xmm0, %ymm0
+; ALL-NEXT:    vpblendd {{.*#+}} ymm1 = ymm0[0,1,2,3,4,5],ymm1[6,7]
 ; ALL-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
 ; ALL-NEXT:    retq
   %shuffle = shufflevector <8 x i64> %a, <8 x i64> %b, <8 x i32> <i32 0, i32 0, i32 0, i32 4, i32 0, i32 0, i32 0, i32 0>
@@ -793,9 +789,8 @@ define <8 x i64> @shuffle_v8i64_00500000(<8 x i64> %a, <8 x i64> %b) {
 ; ALL:       # BB#0:
 ; ALL-NEXT:    vextracti64x4 $1, %zmm0, %ymm1
 ; ALL-NEXT:    vpermq {{.*#+}} ymm1 = ymm1[0,1,1,3]
-; ALL-NEXT:    vpermq {{.*#+}} ymm2 = ymm0[0,0,2,0]
-; ALL-NEXT:    vpblendd {{.*#+}} ymm1 = ymm2[0,1,2,3],ymm1[4,5],ymm2[6,7]
-; ALL-NEXT:    vpermq {{.*#+}} ymm0 = ymm0[0,0,0,0]
+; ALL-NEXT:    vpbroadcastq %xmm0, %ymm0
+; ALL-NEXT:    vpblendd {{.*#+}} ymm1 = ymm0[0,1,2,3],ymm1[4,5],ymm0[6,7]
 ; ALL-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
 ; ALL-NEXT:    retq
   %shuffle = shufflevector <8 x i64> %a, <8 x i64> %b, <8 x i32> <i32 0, i32 0, i32 5, i32 0, i32 0, i32 0, i32 0, i32 0>
@@ -807,9 +802,8 @@ define <8 x i64> @shuffle_v8i64_06000000(<8 x i64> %a, <8 x i64> %b) {
 ; ALL:       # BB#0:
 ; ALL-NEXT:    vextracti64x4 $1, %zmm0, %ymm1
 ; ALL-NEXT:    vpermq {{.*#+}} ymm1 = ymm1[0,2,2,3]
-; ALL-NEXT:    vpermq {{.*#+}} ymm2 = ymm0[0,1,0,0]
-; ALL-NEXT:    vpblendd {{.*#+}} ymm1 = ymm2[0,1],ymm1[2,3],ymm2[4,5,6,7]
-; ALL-NEXT:    vpermq {{.*#+}} ymm0 = ymm0[0,0,0,0]
+; ALL-NEXT:    vpbroadcastq %xmm0, %ymm0
+; ALL-NEXT:    vpblendd {{.*#+}} ymm1 = ymm0[0,1],ymm1[2,3],ymm0[4,5,6,7]
 ; ALL-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
 ; ALL-NEXT:    retq
   %shuffle = shufflevector <8 x i64> %a, <8 x i64> %b, <8 x i32> <i32 0, i32 6, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0>
@@ -821,7 +815,7 @@ define <8 x i64> @shuffle_v8i64_70000000(<8 x i64> %a, <8 x i64> %b) {
 ; ALL:       # BB#0:
 ; ALL-NEXT:    vextracti64x4 $1, %zmm0, %ymm1
 ; ALL-NEXT:    vpermq {{.*#+}} ymm1 = ymm1[3,1,2,3]
-; ALL-NEXT:    vpermq {{.*#+}} ymm0 = ymm0[0,0,0,0]
+; ALL-NEXT:    vpbroadcastq %xmm0, %ymm0
 ; ALL-NEXT:    vpblendd {{.*#+}} ymm1 = ymm1[0,1],ymm0[2,3,4,5,6,7]
 ; ALL-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
 ; ALL-NEXT:    retq
@@ -855,7 +849,7 @@ define <8 x i64> @shuffle_v8i64_00112233(<8 x i64> %a, <8 x i64> %b) {
 define <8 x i64> @shuffle_v8i64_00001111(<8 x i64> %a, <8 x i64> %b) {
 ; ALL-LABEL: shuffle_v8i64_00001111:
 ; ALL:       # BB#0:
-; ALL-NEXT:    vpermq {{.*#+}} ymm1 = ymm0[0,0,0,0]
+; ALL-NEXT:    vpbroadcastq %xmm0, %ymm1
 ; ALL-NEXT:    vpermq {{.*#+}} ymm0 = ymm0[1,1,1,1]
 ; ALL-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
 ; ALL-NEXT:    retq
@@ -879,8 +873,8 @@ define <8 x i64> @shuffle_v8i64_81a3c5e7(<8 x i64> %a, <8 x i64> %b) {
 define <8 x i64> @shuffle_v8i64_08080808(<8 x i64> %a, <8 x i64> %b) {
 ; ALL-LABEL: shuffle_v8i64_08080808:
 ; ALL:       # BB#0:
-; ALL-NEXT:    vpermq {{.*#+}} ymm1 = ymm1[0,0,2,0]
-; ALL-NEXT:    vpermq {{.*#+}} ymm0 = ymm0[0,1,0,3]
+; ALL-NEXT:    vpbroadcastq %xmm1, %ymm1
+; ALL-NEXT:    vpbroadcastq %xmm0, %ymm0
 ; ALL-NEXT:    vpblendd {{.*#+}} ymm0 = ymm0[0,1],ymm1[2,3],ymm0[4,5],ymm1[6,7]
 ; ALL-NEXT:    vinserti64x4 $1, %ymm0, %zmm0, %zmm0
 ; ALL-NEXT:    retq
@@ -892,12 +886,12 @@ define <8 x i64> @shuffle_v8i64_08084c4c(<8 x i64> %a, <8 x i64> %b) {
 ; ALL-LABEL: shuffle_v8i64_08084c4c:
 ; ALL:       # BB#0:
 ; ALL-NEXT:    vextracti64x4 $1, %zmm1, %ymm2
-; ALL-NEXT:    vpermq {{.*#+}} ymm2 = ymm2[0,0,2,0]
+; ALL-NEXT:    vpbroadcastq %xmm2, %ymm2
 ; ALL-NEXT:    vextracti64x4 $1, %zmm0, %ymm3
-; ALL-NEXT:    vpermq {{.*#+}} ymm3 = ymm3[0,1,0,3]
+; ALL-NEXT:    vpbroadcastq %xmm3, %ymm3
 ; ALL-NEXT:    vpblendd {{.*#+}} ymm2 = ymm3[0,1],ymm2[2,3],ymm3[4,5],ymm2[6,7]
-; ALL-NEXT:    vpermq {{.*#+}} ymm1 = ymm1[0,0,2,0]
-; ALL-NEXT:    vpermq {{.*#+}} ymm0 = ymm0[0,1,0,3]
+; ALL-NEXT:    vpbroadcastq %xmm1, %ymm1
+; ALL-NEXT:    vpbroadcastq %xmm0, %ymm0
 ; ALL-NEXT:    vpblendd {{.*#+}} ymm0 = ymm0[0,1],ymm1[2,3],ymm0[4,5],ymm1[6,7]
 ; ALL-NEXT:    vinserti64x4 $1, %ymm2, %zmm0, %zmm0
 ; ALL-NEXT:    retq
@@ -910,9 +904,9 @@ define <8 x i64> @shuffle_v8i64_8823cc67(<8 x i64> %a, <8 x i64> %b) {
 ; ALL:       # BB#0:
 ; ALL-NEXT:    vextracti64x4 $1, %zmm0, %ymm2
 ; ALL-NEXT:    vextracti64x4 $1, %zmm1, %ymm3
-; ALL-NEXT:    vpshufd {{.*#+}} ymm3 = ymm3[0,1,0,1,4,5,4,5]
+; ALL-NEXT:    vpbroadcastq %xmm3, %ymm3
 ; ALL-NEXT:    vpblendd {{.*#+}} ymm2 = ymm3[0,1,2,3],ymm2[4,5,6,7]
-; ALL-NEXT:    vpshufd {{.*#+}} ymm1 = ymm1[0,1,0,1,4,5,4,5]
+; ALL-NEXT:    vpbroadcastq %xmm1, %ymm1
 ; ALL-NEXT:    vpblendd {{.*#+}} ymm0 = ymm1[0,1,2,3],ymm0[4,5,6,7]
 ; ALL-NEXT:    vinserti64x4 $1, %ymm2, %zmm0, %zmm0
 ; ALL-NEXT:    retq
@@ -1241,7 +1235,7 @@ define <8 x i64> @shuffle_v8i64_10004444(<8 x i64> %a, <8 x i64> %b) {
 ; ALL:       # BB#0:
 ; ALL-NEXT:    vpermq {{.*#+}} ymm1 = ymm0[1,0,0,0]
 ; ALL-NEXT:    vextracti64x4 $1, %zmm0, %ymm0
-; ALL-NEXT:    vpermq {{.*#+}} ymm0 = ymm0[0,0,0,0]
+; ALL-NEXT:    vpbroadcastq %xmm0, %ymm0
 ; ALL-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
 ; ALL-NEXT:    retq
   %shuffle = shufflevector <8 x i64> %a, <8 x i64> %b, <8 x i32> <i32 1, i32 0, i32 0, i32 0, i32 4, i32 4, i32 4, i32 4>
@@ -1366,7 +1360,7 @@ define <8 x i64> @shuffle_v8i64_002u6u44(<8 x i64> %a, <8 x i64> %b) {
 define <8 x i64> @shuffle_v8i64_00uu66uu(<8 x i64> %a, <8 x i64> %b) {
 ; ALL-LABEL: shuffle_v8i64_00uu66uu:
 ; ALL:       # BB#0:
-; ALL-NEXT:    vpshufd {{.*#+}} ymm1 = ymm0[0,1,0,1,4,5,4,5]
+; ALL-NEXT:    vpbroadcastq %xmm0, %ymm1
 ; ALL-NEXT:    vextracti64x4 $1, %zmm0, %ymm0
 ; ALL-NEXT:    vpermq {{.*#+}} ymm0 = ymm0[2,2,2,3]
 ; ALL-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
@@ -1427,8 +1421,8 @@ define <8 x i64> @shuffle_v8i64_6caa87e5(<8 x i64> %a, <8 x i64> %b) {
 ; ALL-NEXT:    vextracti64x4 $1, %zmm0, %ymm0
 ; ALL-NEXT:    vpermq {{.*#+}} ymm4 = ymm0[0,3,2,1]
 ; ALL-NEXT:    vpblendd {{.*#+}} ymm3 = ymm3[0,1],ymm4[2,3],ymm3[4,5],ymm4[6,7]
-; ALL-NEXT:    vpshufd {{.*#+}} ymm2 = ymm2[0,1,0,1,4,5,4,5]
 ; ALL-NEXT:    vpshufd {{.*#+}} ymm1 = ymm1[0,1,0,1,4,5,4,5]
+; ALL-NEXT:    vpbroadcastq %xmm2, %ymm2
 ; ALL-NEXT:    vpblendd {{.*#+}} ymm1 = ymm1[0,1],ymm2[2,3],ymm1[4,5,6,7]
 ; ALL-NEXT:    vpermq {{.*#+}} ymm0 = ymm0[2,1,2,3]
 ; ALL-NEXT:    vpblendd {{.*#+}} ymm0 = ymm0[0,1],ymm1[2,3,4,5,6,7]