From 56c7cfe41f0364a5ac15db17ec4876b09acc012f Mon Sep 17 00:00:00 2001
From: Chandler Carruth <chandlerc@gmail.com>
Date: Mon, 22 Sep 2014 20:25:08 +0000
Subject: [PATCH] [x86] Introduce tests covering the gamut of 256-bit vector
 shuffling.

These are just test cases, no actual code yet. This establishes the
baseline fallback strategy we're starting from on AVX2 and the expected
lowering we use on AVX1.

Also, these test cases are very much generated. I've manually crafted
the specific pattern set that I'm hoping will be useful at exercising
the lowering code, but I've not (and could not) manually verify *all* of
these. I've spot checked and they seem legit to me.

As with the rest of vector shuffling, at a certain point the only really
useful way to check the correctness of this stuff is through fuzz
testing.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@218267 91177308-0d34-0410-b5e6-96231b3b80d8
---
 test/CodeGen/X86/vector-shuffle-256-v16.ll | 1334 +++++++++++++++
 test/CodeGen/X86/vector-shuffle-256-v32.ll | 1732 ++++++++++++++++++++
 test/CodeGen/X86/vector-shuffle-256-v8.ll  |  575 +++++++
 3 files changed, 3641 insertions(+)
 create mode 100644 test/CodeGen/X86/vector-shuffle-256-v16.ll
 create mode 100644 test/CodeGen/X86/vector-shuffle-256-v32.ll

diff --git a/test/CodeGen/X86/vector-shuffle-256-v16.ll b/test/CodeGen/X86/vector-shuffle-256-v16.ll
new file mode 100644
index 00000000000..c567ac1dc62
--- /dev/null
+++ b/test/CodeGen/X86/vector-shuffle-256-v16.ll
@@ -0,0 +1,1334 @@
+; RUN: llc < %s -mcpu=x86-64 -mattr=+avx -x86-experimental-vector-shuffle-lowering | FileCheck %s --check-prefix=ALL --check-prefix=AVX1
+; RUN: llc < %s -mcpu=x86-64 -mattr=+avx2 -x86-experimental-vector-shuffle-lowering | FileCheck %s --check-prefix=ALL --check-prefix=AVX2
+
+target triple = "x86_64-unknown-unknown"
+
+define <16 x i16> @shuffle_v16i16_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00
+; 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:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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:    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
+}
+
+define <16 x i16> @shuffle_v16i16_00_00_00_00_00_00_00_00_00_00_00_00_00_00_01_00(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_00_00_00_00_00_00_00_00_00_00_00_00_00_01_00
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufd {{.*}} # xmm0 = xmm0[0,1,0,3]
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm0 = xmm0[0,0,0,0,4,5,6,7]
+; AVX1-NEXT:    vpshufhw {{.*}} # xmm1 = xmm0[0,1,2,3,4,4,4,4]
+; AVX1-NEXT:    vpshufhw {{.*}} # xmm0 = xmm0[0,1,2,3,4,4,5,4]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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:    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>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_00_00_00_00_00_00_00_00_00_00_00_00_02_00_00(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_00_00_00_00_00_00_00_00_00_00_00_00_02_00_00
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,1,0,1,0,1,0,1,4,5,0,1,0,1]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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:    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
+  %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 2, i32 0, i32 0>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_00_00_00_00_00_00_00_00_00_00_00_03_00_00_00(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_00_00_00_00_00_00_00_00_00_00_00_03_00_00_00
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,1,0,1,0,1,6,7,0,1,0,1,0,1]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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:    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
+  %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 3, i32 0, i32 0, i32 0>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_00_00_00_00_00_00_00_00_00_00_04_00_00_00_00(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_00_00_00_00_00_00_00_00_00_00_04_00_00_00_00
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,1,0,1,8,9,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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:    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
+  %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 4, i32 0, i32 0, i32 0, i32 0>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_00_00_00_00_00_00_00_00_00_05_00_00_00_00_00(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_00_00_00_00_00_00_00_00_00_05_00_00_00_00_00
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,1,10,11,0,1,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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:    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
+  %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 5, i32 0, i32 0, i32 0, i32 0, i32 0>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_00_00_00_00_00_00_00_00_06_00_00_00_00_00_00(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_00_00_00_00_00_00_00_00_06_00_00_00_00_00_00
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,12,13,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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:    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
+  %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 6, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_00_00_00_00_00_00_00_07_00_00_00_00_00_00_00(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_00_00_00_00_00_00_00_07_00_00_00_00_00_00_00
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[14,15,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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:    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
+  %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 7, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_00_00_00_00_00_00_08_00_00_00_00_00_00_00_00(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_00_00_00_00_00_00_08_00_00_00_00_00_00_00_00
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT:    vpunpcklwd {{.*}} # xmm1 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm1 = xmm1[0,0,0,1,4,5,6,7]
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm2 = xmm0[0,0,0,0,4,5,6,7]
+; AVX1-NEXT:    vpunpcklqdq {{.*}} # xmm1 = xmm2[0],xmm1[0]
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_00_00_00_00_00_00_08_00_00_00_00_00_00_00_00
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vpunpcklwd {{.*}} # xmm1 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm1 = xmm1[0,0,0,1,4,5,6,7]
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm2 = xmm0[0,0,0,0,4,5,6,7]
+; AVX2-NEXT:    vpunpcklqdq {{.*}} # xmm1 = xmm2[0],xmm1[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, %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 8, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_00_00_00_00_00_09_00_00_00_00_00_00_00_00_00(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_00_00_00_00_00_09_00_00_00_00_00_00_00_00_00
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT:    vpunpcklwd {{.*}} # xmm1 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm1 = xmm1[0,0,3,0,4,5,6,7]
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm2 = xmm0[0,0,0,0,4,5,6,7]
+; AVX1-NEXT:    vpunpcklqdq {{.*}} # xmm1 = xmm2[0],xmm1[0]
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_00_00_00_00_00_09_00_00_00_00_00_00_00_00_00
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vpunpcklwd {{.*}} # xmm1 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm1 = xmm1[0,0,3,0,4,5,6,7]
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm2 = xmm0[0,0,0,0,4,5,6,7]
+; AVX2-NEXT:    vpunpcklqdq {{.*}} # xmm1 = xmm2[0],xmm1[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, %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 9, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_00_00_00_00_10_00_00_00_00_00_00_00_00_00_00(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_00_00_00_00_10_00_00_00_00_00_00_00_00_00_00
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT:    vpunpcklwd {{.*}} # xmm1 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX1-NEXT:    vpshufd {{.*}} # xmm1 = xmm1[0,2,2,3]
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm1 = xmm1[0,3,0,0,4,5,6,7]
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm2 = xmm0[0,0,0,0,4,5,6,7]
+; AVX1-NEXT:    vpunpcklqdq {{.*}} # xmm1 = xmm2[0],xmm1[0]
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_00_00_00_00_10_00_00_00_00_00_00_00_00_00_00
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vpunpcklwd {{.*}} # xmm1 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX2-NEXT:    vpshufd {{.*}} # xmm1 = xmm1[0,2,2,3]
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm1 = xmm1[0,3,0,0,4,5,6,7]
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm2 = xmm0[0,0,0,0,4,5,6,7]
+; AVX2-NEXT:    vpunpcklqdq {{.*}} # xmm1 = xmm2[0],xmm1[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, %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 10, 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
+}
+
+define <16 x i16> @shuffle_v16i16_00_00_00_00_11_00_00_00_00_00_00_00_00_00_00_00(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_00_00_00_11_00_00_00_00_00_00_00_00_00_00_00
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT:    vpunpcklwd {{.*}} # xmm1 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX1-NEXT:    vpshufd {{.*}} # xmm1 = xmm1[0,3,2,3]
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm1 = xmm1[3,0,0,0,4,5,6,7]
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm2 = xmm0[0,0,0,0,4,5,6,7]
+; AVX1-NEXT:    vpunpcklqdq {{.*}} # xmm1 = xmm2[0],xmm1[0]
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_00_00_00_11_00_00_00_00_00_00_00_00_00_00_00
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vpunpcklwd {{.*}} # xmm1 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX2-NEXT:    vpshufd {{.*}} # xmm1 = xmm1[0,3,2,3]
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm1 = xmm1[3,0,0,0,4,5,6,7]
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm2 = xmm0[0,0,0,0,4,5,6,7]
+; AVX2-NEXT:    vpunpcklqdq {{.*}} # xmm1 = xmm2[0],xmm1[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, %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 11, 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
+}
+
+define <16 x i16> @shuffle_v16i16_00_00_00_12_00_00_00_00_00_00_00_00_00_00_00_00(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_00_00_12_00_00_00_00_00_00_00_00_00_00_00_00
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT:    vpshufd {{.*}} # xmm1 = xmm1[2,3,0,1]
+; AVX1-NEXT:    vpunpcklwd {{.*}} # xmm1 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm1 = xmm1[0,0,0,1,4,5,6,7]
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm2 = xmm0[0,0,0,0,4,5,6,7]
+; AVX1-NEXT:    vpunpcklqdq {{.*}} # xmm1 = xmm1[0],xmm2[0]
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_00_00_12_00_00_00_00_00_00_00_00_00_00_00_00
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vpshufd {{.*}} # xmm1 = xmm1[2,3,0,1]
+; AVX2-NEXT:    vpunpcklwd {{.*}} # xmm1 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm1 = xmm1[0,0,0,1,4,5,6,7]
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm2 = xmm0[0,0,0,0,4,5,6,7]
+; AVX2-NEXT:    vpunpcklqdq {{.*}} # xmm1 = xmm1[0],xmm2[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, %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 12, 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
+}
+
+define <16 x i16> @shuffle_v16i16_00_00_13_00_00_00_00_00_00_00_00_00_00_00_00_00(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_00_13_00_00_00_00_00_00_00_00_00_00_00_00_00
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT:    vpshufd {{.*}} # xmm1 = xmm1[2,3,0,1]
+; AVX1-NEXT:    vpunpcklwd {{.*}} # xmm1 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm1 = xmm1[0,0,3,0,4,5,6,7]
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm2 = xmm0[0,0,0,0,4,5,6,7]
+; AVX1-NEXT:    vpunpcklqdq {{.*}} # xmm1 = xmm1[0],xmm2[0]
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_00_13_00_00_00_00_00_00_00_00_00_00_00_00_00
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vpshufd {{.*}} # xmm1 = xmm1[2,3,0,1]
+; AVX2-NEXT:    vpunpcklwd {{.*}} # xmm1 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm1 = xmm1[0,0,3,0,4,5,6,7]
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm2 = xmm0[0,0,0,0,4,5,6,7]
+; AVX2-NEXT:    vpunpcklqdq {{.*}} # xmm1 = xmm1[0],xmm2[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, %ymm1, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <i32 0, i32 0, i32 13, 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
+}
+
+define <16 x i16> @shuffle_v16i16_00_14_00_00_00_00_00_00_00_00_00_00_00_00_00_00(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_14_00_00_00_00_00_00_00_00_00_00_00_00_00_00
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT:    vpshufd {{.*}} # xmm1 = xmm1[2,3,0,1]
+; AVX1-NEXT:    vpunpcklwd {{.*}} # xmm1 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX1-NEXT:    vpshufd {{.*}} # xmm1 = xmm1[0,2,2,3]
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm1 = xmm1[0,3,0,0,4,5,6,7]
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm2 = xmm0[0,0,0,0,4,5,6,7]
+; AVX1-NEXT:    vpunpcklqdq {{.*}} # xmm1 = xmm1[0],xmm2[0]
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_14_00_00_00_00_00_00_00_00_00_00_00_00_00_00
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vpshufd {{.*}} # xmm1 = xmm1[2,3,0,1]
+; AVX2-NEXT:    vpunpcklwd {{.*}} # xmm1 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX2-NEXT:    vpshufd {{.*}} # xmm1 = xmm1[0,2,2,3]
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm1 = xmm1[0,3,0,0,4,5,6,7]
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm2 = xmm0[0,0,0,0,4,5,6,7]
+; AVX2-NEXT:    vpunpcklqdq {{.*}} # xmm1 = xmm1[0],xmm2[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, %ymm1, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <i32 0, i32 14, 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
+}
+
+define <16 x i16> @shuffle_v16i16_15_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_15_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT:    vpshufd {{.*}} # xmm1 = xmm1[2,3,0,1]
+; AVX1-NEXT:    vpunpcklwd {{.*}} # xmm1 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX1-NEXT:    vpshufd {{.*}} # xmm1 = xmm1[0,3,2,3]
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm1 = xmm1[3,0,0,0,4,5,6,7]
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm2 = xmm0[0,0,0,0,4,5,6,7]
+; AVX1-NEXT:    vpunpcklqdq {{.*}} # xmm1 = xmm1[0],xmm2[0]
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_15_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vpshufd {{.*}} # xmm1 = xmm1[2,3,0,1]
+; AVX2-NEXT:    vpunpcklwd {{.*}} # xmm1 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX2-NEXT:    vpshufd {{.*}} # xmm1 = xmm1[0,3,2,3]
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm1 = xmm1[3,0,0,0,4,5,6,7]
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm2 = xmm0[0,0,0,0,4,5,6,7]
+; AVX2-NEXT:    vpunpcklqdq {{.*}} # xmm1 = xmm1[0],xmm2[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, %ymm1, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <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>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_00_00_00_00_00_00_00_08_08_08_08_08_08_08_08(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_00_00_00_00_00_00_00_08_08_08_08_08_08_08_08
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT:    vmovdqa .LCPI16_0(%rip), %xmm2
+; AVX1-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_00_00_00_00_00_00_00_08_08_08_08_08_08_08_08
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vmovdqa .LCPI16_0(%rip), %xmm2
+; AVX2-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX2-NEXT:    vinserti128 $1, %xmm1, %ymm0, %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 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_07_07_07_07_07_07_07_07_15_15_15_15_15_15_15_15(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_07_07_07_07_07_07_07_07_15_15_15_15_15_15_15_15
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT:    vmovdqa .LCPI17_0(%rip), %xmm2
+; AVX1-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_07_07_07_07_07_07_07_07_15_15_15_15_15_15_15_15
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vmovdqa .LCPI17_0(%rip), %xmm2
+; AVX2-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX2-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <i32 7, i32 7, i32 7, i32 7, i32 7, i32 7, i32 7, i32 7, i32 15, i32 15, i32 15, i32 15, i32 15, i32 15, i32 15, i32 15>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_00_00_00_04_04_04_04_08_08_08_08_12_12_12_12(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_00_00_00_04_04_04_04_08_08_08_08_12_12_12_12
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm1 = xmm0[0,0,0,0,4,5,6,7]
+; AVX1-NEXT:    vpshufhw {{.*}} # xmm1 = xmm1[0,1,2,3,4,4,4,4]
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm0 = xmm0[0,0,0,0,4,5,6,7]
+; AVX1-NEXT:    vpshufhw {{.*}} # xmm0 = xmm0[0,1,2,3,4,4,4,4]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_00_00_00_04_04_04_04_08_08_08_08_12_12_12_12
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm1 = xmm0[0,0,0,0,4,5,6,7]
+; AVX2-NEXT:    vpshufhw {{.*}} # xmm1 = xmm1[0,1,2,3,4,4,4,4]
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm0
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm0 = xmm0[0,0,0,0,4,5,6,7]
+; AVX2-NEXT:    vpshufhw {{.*}} # xmm0 = xmm0[0,1,2,3,4,4,4,4]
+; 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 4, i32 4, i32 4, i32 4, i32 8, i32 8, i32 8, i32 8, i32 12, i32 12, i32 12, i32 12>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_03_03_03_03_07_07_07_07_11_11_11_11_15_15_15_15(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_03_03_03_03_07_07_07_07_11_11_11_11_15_15_15_15
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm1 = xmm0[3,3,3,3,4,5,6,7]
+; AVX1-NEXT:    vpshufhw {{.*}} # xmm1 = xmm1[0,1,2,3,7,7,7,7]
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm0 = xmm0[3,3,3,3,4,5,6,7]
+; AVX1-NEXT:    vpshufhw {{.*}} # xmm0 = xmm0[0,1,2,3,7,7,7,7]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_03_03_03_03_07_07_07_07_11_11_11_11_15_15_15_15
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm1 = xmm0[3,3,3,3,4,5,6,7]
+; AVX2-NEXT:    vpshufhw {{.*}} # xmm1 = xmm1[0,1,2,3,7,7,7,7]
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm0
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm0 = xmm0[3,3,3,3,4,5,6,7]
+; AVX2-NEXT:    vpshufhw {{.*}} # xmm0 = xmm0[0,1,2,3,7,7,7,7]
+; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <i32 3, i32 3, i32 3, i32 3, i32 7, i32 7, i32 7, i32 7, i32 11, i32 11, i32 11, i32 11, i32 15, i32 15, i32 15, i32 15>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_00_02_02_04_04_06_06_08_08_10_10_12_12_14_14(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_00_02_02_04_04_06_06_08_08_10_10_12_12_14_14
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm1 = xmm0[0,0,2,2,4,5,6,7]
+; AVX1-NEXT:    vpshufhw {{.*}} # xmm1 = xmm1[0,1,2,3,4,4,6,6]
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm0 = xmm0[0,0,2,2,4,5,6,7]
+; AVX1-NEXT:    vpshufhw {{.*}} # xmm0 = xmm0[0,1,2,3,4,4,6,6]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_00_02_02_04_04_06_06_08_08_10_10_12_12_14_14
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm1 = xmm0[0,0,2,2,4,5,6,7]
+; AVX2-NEXT:    vpshufhw {{.*}} # xmm1 = xmm1[0,1,2,3,4,4,6,6]
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm0
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm0 = xmm0[0,0,2,2,4,5,6,7]
+; AVX2-NEXT:    vpshufhw {{.*}} # xmm0 = xmm0[0,1,2,3,4,4,6,6]
+; 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 2, i32 2, i32 4, i32 4, i32 6, i32 6, i32 8, i32 8, i32 10, i32 10, i32 12, i32 12, i32 14, i32 14>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_01_01_03_03_05_05_07_07_09_09_11_11_13_13_15_15(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_01_01_03_03_05_05_07_07_09_09_11_11_13_13_15_15
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm1 = xmm0[1,1,3,3,4,5,6,7]
+; AVX1-NEXT:    vpshufhw {{.*}} # xmm1 = xmm1[0,1,2,3,5,5,7,7]
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm0 = xmm0[1,1,3,3,4,5,6,7]
+; AVX1-NEXT:    vpshufhw {{.*}} # xmm0 = xmm0[0,1,2,3,5,5,7,7]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_01_01_03_03_05_05_07_07_09_09_11_11_13_13_15_15
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm1 = xmm0[1,1,3,3,4,5,6,7]
+; AVX2-NEXT:    vpshufhw {{.*}} # xmm1 = xmm1[0,1,2,3,5,5,7,7]
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm0
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm0 = xmm0[1,1,3,3,4,5,6,7]
+; AVX2-NEXT:    vpshufhw {{.*}} # xmm0 = xmm0[0,1,2,3,5,5,7,7]
+; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <i32 1, i32 1, i32 3, i32 3, i32 5, i32 5, i32 7, i32 7, i32 9, i32 9, i32 11, i32 11, i32 13, i32 13, i32 15, i32 15>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_00_00_00_00_00_01_00_00_00_00_00_00_00_01_00(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_00_00_00_00_00_01_00_00_00_00_00_00_00_01_00
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,2,3,0,1]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_00_00_00_00_00_01_00_00_00_00_00_00_00_01_00
+; AVX2:       # BB#0:
+; 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, %ymm0, %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 1, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 1, i32 0>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_00_00_00_00_02_00_00_00_00_00_00_00_02_00_00(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_00_00_00_00_02_00_00_00_00_00_00_00_02_00_00
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,1,0,1,0,1,0,1,4,5,0,1,0,1]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_00_00_00_00_02_00_00_00_00_00_00_00_02_00_00
+; AVX2:       # BB#0:
+; 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, %ymm0, %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 2, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 2, i32 0, i32 0>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_00_00_00_03_00_00_00_00_00_00_00_03_00_00_00(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_00_00_00_03_00_00_00_00_00_00_00_03_00_00_00
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,1,0,1,0,1,6,7,0,1,0,1,0,1]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_00_00_00_03_00_00_00_00_00_00_00_03_00_00_00
+; AVX2:       # BB#0:
+; 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, %ymm0, %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 3, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 3, i32 0, i32 0, i32 0>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_00_00_04_00_00_00_00_00_00_00_04_00_00_00_00(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_00_00_04_00_00_00_00_00_00_00_04_00_00_00_00
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,1,0,1,8,9,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_00_00_04_00_00_00_00_00_00_00_04_00_00_00_00
+; AVX2:       # BB#0:
+; 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, %ymm0, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <i32 0, i32 0, i32 0, i32 4, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 4, i32 0, i32 0, i32 0, i32 0>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_00_05_00_00_00_00_00_00_00_05_00_00_00_00_00(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_00_05_00_00_00_00_00_00_00_05_00_00_00_00_00
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,1,10,11,0,1,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_00_05_00_00_00_00_00_00_00_05_00_00_00_00_00
+; AVX2:       # BB#0:
+; 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, %ymm0, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <i32 0, i32 0, i32 5, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 5, i32 0, i32 0, i32 0, i32 0, i32 0>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_06_00_00_00_00_00_00_00_06_00_00_00_00_00_00(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_06_00_00_00_00_00_00_00_06_00_00_00_00_00_00
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,12,13,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_06_00_00_00_00_00_00_00_06_00_00_00_00_00_00
+; AVX2:       # BB#0:
+; 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, %ymm0, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <i32 0, i32 6, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 6, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_07_00_00_00_00_00_00_00_07_00_00_00_00_00_00_00(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_07_00_00_00_00_00_00_00_07_00_00_00_00_00_00_00
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[14,15,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_07_00_00_00_00_00_00_00_07_00_00_00_00_00_00_00
+; AVX2:       # BB#0:
+; 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, %ymm0, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <i32 7, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 7, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_16_01_18_03_20_05_22_07_24_09_26_11_28_13_30_15(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_16_01_18_03_20_05_22_07_24_09_26_11_28_13_30_15
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm2
+; AVX1-NEXT:    vextractf128 $1, %ymm1, %xmm3
+; AVX1-NEXT:    vpblendw {{.*}} # xmm2 = xmm3[0],xmm2[1],xmm3[2],xmm2[3],xmm3[4],xmm2[5],xmm3[6],xmm2[7]
+; AVX1-NEXT:    vpblendw {{.*}} # xmm0 = xmm1[0],xmm0[1],xmm1[2],xmm0[3],xmm1[4],xmm0[5],xmm1[6],xmm0[7]
+; AVX1-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_16_01_18_03_20_05_22_07_24_09_26_11_28_13_30_15
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm2
+; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm3
+; AVX2-NEXT:    vpblendw {{.*}} # xmm2 = xmm3[0],xmm2[1],xmm3[2],xmm2[3],xmm3[4],xmm2[5],xmm3[6],xmm2[7]
+; AVX2-NEXT:    vpblendw {{.*}} # xmm0 = xmm1[0],xmm0[1],xmm1[2],xmm0[3],xmm1[4],xmm0[5],xmm1[6],xmm0[7]
+; AVX2-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <i32 16, i32 1, i32 18, i32 3, i32 20, i32 5, i32 22, i32 7, i32 24, i32 9, i32 26, i32 11, i32 28, i32 13, i32 30, i32 15>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_16_00_16_00_16_00_16_00_16_00_16_00_16_00_16(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_16_00_16_00_16_00_16_00_16_00_16_00_16_00_16
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm1 = xmm1[0,1,0,1,4,5,0,1,0,1,0,1,12,13,0,1]
+; AVX1-NEXT:    vpshufd {{.*}} # xmm0 = xmm0[0,0,0,0]
+; AVX1-NEXT:    vpblendw {{.*}} # xmm0 = xmm0[0],xmm1[1],xmm0[2],xmm1[3],xmm0[4],xmm1[5],xmm0[6],xmm1[7]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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:    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
+  %shuffle = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <i32 0, i32 16, i32 0, i32 16, i32 0, i32 16, i32 0, i32 16, i32 0, i32 16, i32 0, i32 16, i32 0, i32 16, i32 0, i32 16>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_16_00_16_00_16_00_16_08_24_08_24_08_24_08_24(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_16_00_16_00_16_00_16_08_24_08_24_08_24_08_24
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm1, %xmm2
+; AVX1-NEXT:    vmovdqa .LCPI31_0(%rip), %xmm3
+; AVX1-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm4
+; AVX1-NEXT:    vpshufd {{.*}} # xmm4 = xmm4[0,0,0,0]
+; AVX1-NEXT:    vpblendw {{.*}} # xmm2 = xmm4[0],xmm2[1],xmm4[2],xmm2[3],xmm4[4],xmm2[5],xmm4[6],xmm2[7]
+; AVX1-NEXT:    vpshufb %xmm3, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufd {{.*}} # xmm0 = xmm0[0,0,0,0]
+; AVX1-NEXT:    vpblendw {{.*}} # xmm0 = xmm0[0],xmm1[1],xmm0[2],xmm1[3],xmm0[4],xmm1[5],xmm0[6],xmm1[7]
+; AVX1-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_16_00_16_00_16_00_16_08_24_08_24_08_24_08_24
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm2
+; AVX2-NEXT:    vmovdqa .LCPI31_0(%rip), %xmm3
+; AVX2-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm4
+; AVX2-NEXT:    vpshufd {{.*}} # xmm4 = xmm4[0,0,0,0]
+; AVX2-NEXT:    vpblendw {{.*}} # xmm2 = xmm4[0],xmm2[1],xmm4[2],xmm2[3],xmm4[4],xmm2[5],xmm4[6],xmm2[7]
+; AVX2-NEXT:    vpshufb %xmm3, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufd {{.*}} # xmm0 = xmm0[0,0,0,0]
+; AVX2-NEXT:    vpblendw {{.*}} # xmm0 = xmm0[0],xmm1[1],xmm0[2],xmm1[3],xmm0[4],xmm1[5],xmm0[6],xmm1[7]
+; AVX2-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <i32 0, i32 16, i32 0, i32 16, i32 0, i32 16, i32 0, i32 16, i32 8, i32 24, i32 8, i32 24, i32 8, i32 24, i32 8, i32 24>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_16_16_16_16_04_05_06_07_24_24_24_24_12_13_14_15(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_16_16_16_16_04_05_06_07_24_24_24_24_12_13_14_15
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm2
+; AVX1-NEXT:    vextractf128 $1, %ymm1, %xmm3
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm3 = xmm3[0,0,0,0,4,5,6,7]
+; AVX1-NEXT:    vpblendw {{.*}} # xmm2 = xmm3[0,1,2,3],xmm2[4,5,6,7]
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm1 = xmm1[0,0,0,0,4,5,6,7]
+; AVX1-NEXT:    vpblendw {{.*}} # xmm0 = xmm1[0,1,2,3],xmm0[4,5,6,7]
+; AVX1-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_16_16_16_16_04_05_06_07_24_24_24_24_12_13_14_15
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm2
+; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm3
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm3 = xmm3[0,0,0,0,4,5,6,7]
+; AVX2-NEXT:    vpblendw {{.*}} # xmm2 = xmm3[0,1,2,3],xmm2[4,5,6,7]
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm1 = xmm1[0,0,0,0,4,5,6,7]
+; AVX2-NEXT:    vpblendw {{.*}} # xmm0 = xmm1[0,1,2,3],xmm0[4,5,6,7]
+; AVX2-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <i32 16, i32 16, i32 16, i32 16, i32 4, i32 5, i32 6, i32 7, i32 24, i32 24, i32 24, i32 24, i32 12, i32 13, i32 14, i32 15>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_19_18_17_16_07_06_05_04_27_26_25_24_15_14_13_12(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_19_18_17_16_07_06_05_04_27_26_25_24_15_14_13_12
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm2
+; AVX1-NEXT:    vpshufhw {{.*}} # xmm2 = xmm2[0,1,2,3,7,6,5,4]
+; AVX1-NEXT:    vextractf128 $1, %ymm1, %xmm3
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm3 = xmm3[3,2,1,0,4,5,6,7]
+; AVX1-NEXT:    vpblendw {{.*}} # xmm2 = xmm3[0,1,2,3],xmm2[4,5,6,7]
+; AVX1-NEXT:    vpshufhw {{.*}} # xmm0 = xmm0[0,1,2,3,7,6,5,4]
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm1 = xmm1[3,2,1,0,4,5,6,7]
+; AVX1-NEXT:    vpblendw {{.*}} # xmm0 = xmm1[0,1,2,3],xmm0[4,5,6,7]
+; AVX1-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_19_18_17_16_07_06_05_04_27_26_25_24_15_14_13_12
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm2
+; AVX2-NEXT:    vpshufhw {{.*}} # xmm2 = xmm2[0,1,2,3,7,6,5,4]
+; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm3
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm3 = xmm3[3,2,1,0,4,5,6,7]
+; AVX2-NEXT:    vpblendw {{.*}} # xmm2 = xmm3[0,1,2,3],xmm2[4,5,6,7]
+; AVX2-NEXT:    vpshufhw {{.*}} # xmm0 = xmm0[0,1,2,3,7,6,5,4]
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm1 = xmm1[3,2,1,0,4,5,6,7]
+; AVX2-NEXT:    vpblendw {{.*}} # xmm0 = xmm1[0,1,2,3],xmm0[4,5,6,7]
+; AVX2-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <i32 19, i32 18, i32 17, i32 16, i32 7, i32 6, i32 5, i32 4, i32 27, i32 26, i32 25, i32 24, i32 15, i32 14, i32 13, i32 12>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_19_18_17_16_03_02_01_00_27_26_25_24_11_10_09_08(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_19_18_17_16_03_02_01_00_27_26_25_24_11_10_09_08
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm1, %xmm2
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm2 = xmm2[3,2,1,0,4,5,6,7]
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm3
+; AVX1-NEXT:    vpshufd {{.*}} # xmm3 = xmm3[0,1,0,1]
+; AVX1-NEXT:    vpshufhw {{.*}} # xmm3 = xmm3[0,1,2,3,7,6,5,4]
+; AVX1-NEXT:    vpblendw {{.*}} # xmm2 = xmm2[0,1,2,3],xmm3[4,5,6,7]
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm1 = xmm1[3,2,1,0,4,5,6,7]
+; AVX1-NEXT:    vpshufd {{.*}} # xmm0 = xmm0[0,1,0,1]
+; AVX1-NEXT:    vpshufhw {{.*}} # xmm0 = xmm0[0,1,2,3,7,6,5,4]
+; AVX1-NEXT:    vpblendw {{.*}} # xmm0 = xmm1[0,1,2,3],xmm0[4,5,6,7]
+; AVX1-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_19_18_17_16_03_02_01_00_27_26_25_24_11_10_09_08
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm2
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm2 = xmm2[3,2,1,0,4,5,6,7]
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm3
+; AVX2-NEXT:    vpshufd {{.*}} # xmm3 = xmm3[0,1,0,1]
+; AVX2-NEXT:    vpshufhw {{.*}} # xmm3 = xmm3[0,1,2,3,7,6,5,4]
+; AVX2-NEXT:    vpblendw {{.*}} # xmm2 = xmm2[0,1,2,3],xmm3[4,5,6,7]
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm1 = xmm1[3,2,1,0,4,5,6,7]
+; AVX2-NEXT:    vpshufd {{.*}} # xmm0 = xmm0[0,1,0,1]
+; AVX2-NEXT:    vpshufhw {{.*}} # xmm0 = xmm0[0,1,2,3,7,6,5,4]
+; AVX2-NEXT:    vpblendw {{.*}} # xmm0 = xmm1[0,1,2,3],xmm0[4,5,6,7]
+; AVX2-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <i32 19, i32 18, i32 17, i32 16, i32 3, i32 2, i32 1, i32 0, i32 27, i32 26, i32 25, i32 24, i32 11, i32 10, i32 9, i32 8>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_00_00_00_00_00_01_00_08_08_08_08_08_08_09_08(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_00_00_00_00_00_01_00_08_08_08_08_08_08_09_08
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT:    vmovdqa .LCPI35_0(%rip), %xmm2
+; AVX1-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_00_00_00_00_00_01_00_08_08_08_08_08_08_09_08
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vmovdqa .LCPI35_0(%rip), %xmm2
+; AVX2-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX2-NEXT:    vinserti128 $1, %xmm1, %ymm0, %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 1, i32 0, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 9, i32 8>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_00_00_00_00_02_00_00_08_08_08_08_08_10_08_08(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_00_00_00_00_02_00_00_08_08_08_08_08_10_08_08
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT:    vmovdqa .LCPI36_0(%rip), %xmm2
+; AVX1-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_00_00_00_00_02_00_00_08_08_08_08_08_10_08_08
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vmovdqa .LCPI36_0(%rip), %xmm2
+; AVX2-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX2-NEXT:    vinserti128 $1, %xmm1, %ymm0, %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 2, i32 0, i32 0, i32 8, i32 8, i32 8, i32 8, i32 8, i32 10, i32 8, i32 8>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_00_00_00_03_00_00_00_08_08_08_08_11_08_08_08(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_00_00_00_03_00_00_00_08_08_08_08_11_08_08_08
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT:    vmovdqa .LCPI37_0(%rip), %xmm2
+; AVX1-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_00_00_00_03_00_00_00_08_08_08_08_11_08_08_08
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vmovdqa .LCPI37_0(%rip), %xmm2
+; AVX2-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX2-NEXT:    vinserti128 $1, %xmm1, %ymm0, %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 3, i32 0, i32 0, i32 0, i32 8, i32 8, i32 8, i32 8, i32 11, i32 8, i32 8, i32 8>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_00_00_04_00_00_00_00_08_08_08_12_08_08_08_08(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_00_00_04_00_00_00_00_08_08_08_12_08_08_08_08
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT:    vmovdqa .LCPI38_0(%rip), %xmm2
+; AVX1-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_00_00_04_00_00_00_00_08_08_08_12_08_08_08_08
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vmovdqa .LCPI38_0(%rip), %xmm2
+; AVX2-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX2-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <i32 0, i32 0, i32 0, i32 4, i32 0, i32 0, i32 0, i32 0, i32 8, i32 8, i32 8, i32 12, i32 8, i32 8, i32 8, i32 8>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_00_05_00_00_00_00_00_08_08_13_08_08_08_08_08(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_00_05_00_00_00_00_00_08_08_13_08_08_08_08_08
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT:    vmovdqa .LCPI39_0(%rip), %xmm2
+; AVX1-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_00_05_00_00_00_00_00_08_08_13_08_08_08_08_08
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vmovdqa .LCPI39_0(%rip), %xmm2
+; AVX2-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX2-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <i32 0, i32 0, i32 5, i32 0, i32 0, i32 0, i32 0, i32 0, i32 8, i32 8, i32 13, i32 8, i32 8, i32 8, i32 8, i32 8>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_06_00_00_00_00_00_00_08_14_08_08_08_08_08_08(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_06_00_00_00_00_00_00_08_14_08_08_08_08_08_08
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT:    vmovdqa .LCPI40_0(%rip), %xmm2
+; AVX1-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_06_00_00_00_00_00_00_08_14_08_08_08_08_08_08
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vmovdqa .LCPI40_0(%rip), %xmm2
+; AVX2-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX2-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <i32 0, i32 6, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 8, i32 14, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_07_00_00_00_00_00_00_00_15_08_08_08_08_08_08_08(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_07_00_00_00_00_00_00_00_15_08_08_08_08_08_08_08
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT:    vmovdqa .LCPI41_0(%rip), %xmm2
+; AVX1-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_07_00_00_00_00_00_00_00_15_08_08_08_08_08_08_08
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vmovdqa .LCPI41_0(%rip), %xmm2
+; AVX2-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX2-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <i32 7, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 15, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_16_01_17_02_18_03_19_08_24_09_25_10_26_11_27(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_16_01_17_02_18_03_19_08_24_09_25_10_26_11_27
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm1, %xmm2
+; AVX1-NEXT:    vpunpcklwd {{.*}} # xmm2 = xmm2[0,0,1,1,2,2,3,3]
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm3
+; AVX1-NEXT:    vpmovzxwd %xmm3, %xmm3
+; AVX1-NEXT:    vpblendw {{.*}} # xmm2 = xmm3[0],xmm2[1],xmm3[2],xmm2[3],xmm3[4],xmm2[5],xmm3[6],xmm2[7]
+; AVX1-NEXT:    vpunpcklwd {{.*}} # xmm1 = xmm1[0,0,1,1,2,2,3,3]
+; AVX1-NEXT:    vpmovzxwd %xmm0, %xmm0
+; AVX1-NEXT:    vpblendw {{.*}} # xmm0 = xmm0[0],xmm1[1],xmm0[2],xmm1[3],xmm0[4],xmm1[5],xmm0[6],xmm1[7]
+; AVX1-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_16_01_17_02_18_03_19_08_24_09_25_10_26_11_27
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm2
+; AVX2-NEXT:    vpunpcklwd {{.*}} # xmm2 = xmm2[0,0,1,1,2,2,3,3]
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm3
+; AVX2-NEXT:    vpmovzxwd %xmm3, %xmm3
+; AVX2-NEXT:    vpblendw {{.*}} # xmm2 = xmm3[0],xmm2[1],xmm3[2],xmm2[3],xmm3[4],xmm2[5],xmm3[6],xmm2[7]
+; AVX2-NEXT:    vpunpcklwd {{.*}} # xmm1 = xmm1[0,0,1,1,2,2,3,3]
+; AVX2-NEXT:    vpmovzxwd %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, %xmm2, %ymm0, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <i32 0, i32 16, i32 1, i32 17, i32 2, i32 18, i32 3, i32 19, i32 8, i32 24, i32 9, i32 25, i32 10, i32 26, i32 11, i32 27>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_04_20_05_21_06_22_07_23_12_28_13_29_14_30_15_31(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_04_20_05_21_06_22_07_23_12_28_13_29_14_30_15_31
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm1, %xmm2
+; AVX1-NEXT:    vpunpckhwd {{.*}} # xmm2 = xmm2[4,4,5,5,6,6,7,7]
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm3
+; AVX1-NEXT:    vpunpckhwd {{.*}} # xmm3 = xmm3[4,4,5,5,6,6,7,7]
+; AVX1-NEXT:    vpblendw {{.*}} # xmm2 = xmm3[0],xmm2[1],xmm3[2],xmm2[3],xmm3[4],xmm2[5],xmm3[6],xmm2[7]
+; AVX1-NEXT:    vpunpckhwd {{.*}} # xmm1 = xmm1[4,4,5,5,6,6,7,7]
+; AVX1-NEXT:    vpunpckhwd {{.*}} # xmm0 = xmm0[4,4,5,5,6,6,7,7]
+; AVX1-NEXT:    vpblendw {{.*}} # xmm0 = xmm0[0],xmm1[1],xmm0[2],xmm1[3],xmm0[4],xmm1[5],xmm0[6],xmm1[7]
+; AVX1-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_04_20_05_21_06_22_07_23_12_28_13_29_14_30_15_31
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm2
+; AVX2-NEXT:    vpunpckhwd {{.*}} # xmm2 = xmm2[4,4,5,5,6,6,7,7]
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm3
+; AVX2-NEXT:    vpunpckhwd {{.*}} # xmm3 = xmm3[4,4,5,5,6,6,7,7]
+; AVX2-NEXT:    vpblendw {{.*}} # xmm2 = xmm3[0],xmm2[1],xmm3[2],xmm2[3],xmm3[4],xmm2[5],xmm3[6],xmm2[7]
+; AVX2-NEXT:    vpunpckhwd {{.*}} # xmm1 = xmm1[4,4,5,5,6,6,7,7]
+; AVX2-NEXT:    vpunpckhwd {{.*}} # xmm0 = xmm0[4,4,5,5,6,6,7,7]
+; AVX2-NEXT:    vpblendw {{.*}} # xmm0 = xmm0[0],xmm1[1],xmm0[2],xmm1[3],xmm0[4],xmm1[5],xmm0[6],xmm1[7]
+; AVX2-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <i32 4, i32 20, i32 5, i32 21, i32 6, i32 22, i32 7, i32 23, i32 12, i32 28, i32 13, i32 29, i32 14, i32 30, i32 15, i32 31>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_16_01_17_02_18_03_19_12_28_13_29_14_30_15_31(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_16_01_17_02_18_03_19_12_28_13_29_14_30_15_31
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm1, %xmm2
+; AVX1-NEXT:    vpunpckhwd {{.*}} # xmm2 = xmm2[4,4,5,5,6,6,7,7]
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm3
+; AVX1-NEXT:    vpunpckhwd {{.*}} # xmm3 = xmm3[4,4,5,5,6,6,7,7]
+; AVX1-NEXT:    vpblendw {{.*}} # xmm2 = xmm3[0],xmm2[1],xmm3[2],xmm2[3],xmm3[4],xmm2[5],xmm3[6],xmm2[7]
+; AVX1-NEXT:    vpunpcklwd {{.*}} # xmm1 = xmm1[0,0,1,1,2,2,3,3]
+; AVX1-NEXT:    vpmovzxwd %xmm0, %xmm0
+; AVX1-NEXT:    vpblendw {{.*}} # xmm0 = xmm0[0],xmm1[1],xmm0[2],xmm1[3],xmm0[4],xmm1[5],xmm0[6],xmm1[7]
+; AVX1-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_16_01_17_02_18_03_19_12_28_13_29_14_30_15_31
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm2
+; AVX2-NEXT:    vpunpckhwd {{.*}} # xmm2 = xmm2[4,4,5,5,6,6,7,7]
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm3
+; AVX2-NEXT:    vpunpckhwd {{.*}} # xmm3 = xmm3[4,4,5,5,6,6,7,7]
+; AVX2-NEXT:    vpblendw {{.*}} # xmm2 = xmm3[0],xmm2[1],xmm3[2],xmm2[3],xmm3[4],xmm2[5],xmm3[6],xmm2[7]
+; AVX2-NEXT:    vpunpcklwd {{.*}} # xmm1 = xmm1[0,0,1,1,2,2,3,3]
+; AVX2-NEXT:    vpmovzxwd %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, %xmm2, %ymm0, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <i32 0, i32 16, i32 1, i32 17, i32 2, i32 18, i32 3, i32 19, i32 12, i32 28, i32 13, i32 29, i32 14, i32 30, i32 15, i32 31>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_04_20_05_21_06_22_07_23_08_24_09_25_10_26_11_27(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_04_20_05_21_06_22_07_23_08_24_09_25_10_26_11_27
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm1, %xmm2
+; AVX1-NEXT:    vpunpcklwd {{.*}} # xmm2 = xmm2[0,0,1,1,2,2,3,3]
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm3
+; AVX1-NEXT:    vpmovzxwd %xmm3, %xmm3
+; AVX1-NEXT:    vpblendw {{.*}} # xmm2 = xmm3[0],xmm2[1],xmm3[2],xmm2[3],xmm3[4],xmm2[5],xmm3[6],xmm2[7]
+; AVX1-NEXT:    vpunpckhwd {{.*}} # xmm1 = xmm1[4,4,5,5,6,6,7,7]
+; AVX1-NEXT:    vpunpckhwd {{.*}} # xmm0 = xmm0[4,4,5,5,6,6,7,7]
+; AVX1-NEXT:    vpblendw {{.*}} # xmm0 = xmm0[0],xmm1[1],xmm0[2],xmm1[3],xmm0[4],xmm1[5],xmm0[6],xmm1[7]
+; AVX1-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_04_20_05_21_06_22_07_23_08_24_09_25_10_26_11_27
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm2
+; AVX2-NEXT:    vpunpcklwd {{.*}} # xmm2 = xmm2[0,0,1,1,2,2,3,3]
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm3
+; AVX2-NEXT:    vpmovzxwd %xmm3, %xmm3
+; AVX2-NEXT:    vpblendw {{.*}} # xmm2 = xmm3[0],xmm2[1],xmm3[2],xmm2[3],xmm3[4],xmm2[5],xmm3[6],xmm2[7]
+; AVX2-NEXT:    vpunpckhwd {{.*}} # xmm1 = xmm1[4,4,5,5,6,6,7,7]
+; AVX2-NEXT:    vpunpckhwd {{.*}} # xmm0 = xmm0[4,4,5,5,6,6,7,7]
+; AVX2-NEXT:    vpblendw {{.*}} # xmm0 = xmm0[0],xmm1[1],xmm0[2],xmm1[3],xmm0[4],xmm1[5],xmm0[6],xmm1[7]
+; AVX2-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <i32 4, i32 20, i32 5, i32 21, i32 6, i32 22, i32 7, i32 23, i32 8, i32 24, i32 9, i32 25, i32 10, i32 26, i32 11, i32 27>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_00_00_00_00_00_01_00_08_09_08_08_08_08_08_08(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_00_00_00_00_00_01_00_08_09_08_08_08_08_08_08
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,2,3,0,1]
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,2,3,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_00_00_00_00_00_01_00_08_09_08_08_08_08_08_08
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,2,3,0,1]
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm0
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,2,3,0,1,0,1,0,1,0,1,0,1,0,1]
+; 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 1, i32 0, i32 8, i32 9, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_00_00_00_00_02_00_00_08_08_10_08_08_08_08_08(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_00_00_00_00_02_00_00_08_08_10_08_08_08_08_08
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,1,0,1,0,1,0,1,0,1,4,5,0,1,0,1]
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,1,4,5,0,1,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_00_00_00_00_02_00_00_08_08_10_08_08_08_08_08
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,1,0,1,0,1,0,1,0,1,4,5,0,1,0,1]
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm0
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,1,4,5,0,1,0,1,0,1,0,1,0,1]
+; 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 2, i32 0, i32 0, i32 8, i32 8, i32 10, i32 8, i32 8, i32 8, i32 8, i32 8>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_00_00_00_03_00_00_00_08_08_08_11_08_08_08_08(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_00_00_00_03_00_00_00_08_08_08_11_08_08_08_08
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,1,0,1,0,1,0,1,6,7,0,1,0,1,0,1]
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,1,0,1,6,7,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_00_00_00_03_00_00_00_08_08_08_11_08_08_08_08
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,1,0,1,0,1,0,1,6,7,0,1,0,1,0,1]
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm0
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,1,0,1,6,7,0,1,0,1,0,1,0,1]
+; 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 3, i32 0, i32 0, i32 0, i32 8, i32 8, i32 8, i32 11, i32 8, i32 8, i32 8, i32 8>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_00_00_04_00_00_00_00_08_08_08_08_12_08_08_08(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_00_00_04_00_00_00_00_08_08_08_08_12_08_08_08
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,1,0,1,0,1,8,9,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,1,0,1,0,1,8,9,0,1,0,1,0,1]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_00_00_04_00_00_00_00_08_08_08_08_12_08_08_08
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,1,0,1,0,1,8,9,0,1,0,1,0,1,0,1]
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm0
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,1,0,1,0,1,8,9,0,1,0,1,0,1]
+; 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 4, i32 0, i32 0, i32 0, i32 0, i32 8, i32 8, i32 8, i32 8, i32 12, i32 8, i32 8, i32 8>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_00_05_00_00_00_00_00_08_08_08_08_08_13_08_08(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_00_05_00_00_00_00_00_08_08_08_08_08_13_08_08
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,1,0,1,10,11,0,1,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,1,0,1,0,1,0,1,10,11,0,1,0,1]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_00_05_00_00_00_00_00_08_08_08_08_08_13_08_08
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,1,0,1,10,11,0,1,0,1,0,1,0,1,0,1]
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm0
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,1,0,1,0,1,0,1,10,11,0,1,0,1]
+; 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 5, i32 0, i32 0, i32 0, i32 0, i32 0, i32 8, i32 8, i32 8, i32 8, i32 8, i32 13, i32 8, i32 8>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_06_00_00_00_00_00_00_08_08_08_08_08_08_14_08(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_06_00_00_00_00_00_00_08_08_08_08_08_08_14_08
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,1,12,13,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,12,13,0,1]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_06_00_00_00_00_00_00_08_08_08_08_08_08_14_08
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,1,12,13,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm0
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,12,13,0,1]
+; 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 6, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 14, i32 8>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_07_00_00_00_00_00_00_00_08_08_08_08_08_08_08_15(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_07_00_00_00_00_00_00_00_08_08_08_08_08_08_08_15
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[14,15,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,14,15]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_07_00_00_00_00_00_00_00_08_08_08_08_08_08_08_15
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[14,15,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm0
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,14,15]
+; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <i32 7, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 15>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_00_02_02_04_04_06_06_14_14_12_12_10_10_08_08(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_00_02_02_04_04_06_06_14_14_12_12_10_10_08_08
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm1 = xmm0[0,0,2,2,4,5,6,7]
+; AVX1-NEXT:    vpshufhw {{.*}} # xmm1 = xmm1[0,1,2,3,4,4,6,6]
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[12,13,12,13,8,9,8,9,4,5,4,5,0,1,0,1]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_00_02_02_04_04_06_06_14_14_12_12_10_10_08_08
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm1 = xmm0[0,0,2,2,4,5,6,7]
+; AVX2-NEXT:    vpshufhw {{.*}} # xmm1 = xmm1[0,1,2,3,4,4,6,6]
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm0
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[12,13,12,13,8,9,8,9,4,5,4,5,0,1,0,1]
+; 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 2, i32 2, i32 4, i32 4, i32 6, i32 6, i32 14, i32 14, i32 12, i32 12, i32 10, i32 10, i32 8, i32 8>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_04_04_04_04_00_00_00_00_08_08_08_08_12_12_12_12(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_04_04_04_04_00_00_00_00_08_08_08_08_12_12_12_12
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[8,9,8,9,8,9,8,9,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm0 = xmm0[0,0,0,0,4,5,6,7]
+; AVX1-NEXT:    vpshufhw {{.*}} # xmm0 = xmm0[0,1,2,3,4,4,4,4]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_04_04_04_04_00_00_00_00_08_08_08_08_12_12_12_12
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[8,9,8,9,8,9,8,9,0,1,0,1,0,1,0,1]
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm0
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm0 = xmm0[0,0,0,0,4,5,6,7]
+; AVX2-NEXT:    vpshufhw {{.*}} # xmm0 = xmm0[0,1,2,3,4,4,4,4]
+; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <i32 4, i32 4, i32 4, i32 4, i32 0, i32 0, i32 0, i32 0, i32 8, i32 8, i32 8, i32 8, i32 12, i32 12, i32 12, i32 12>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_uu_uu_00_00_00_00_00_08_08_uu_uu_08_08_14_08(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_uu_uu_00_00_00_00_00_08_08_uu_uu_08_08_14_08
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,1,2,3,4,5,0,1,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,1,4,5,6,7,0,1,0,1,12,13,0,1]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_uu_uu_00_00_00_00_00_08_08_uu_uu_08_08_14_08
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,1,2,3,4,5,0,1,0,1,0,1,0,1,0,1]
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm0
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,1,4,5,6,7,0,1,0,1,12,13,0,1]
+; 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 undef, i32 undef, i32 0, i32 0, i32 0, i32 0, i32 0, i32 8, i32 8, i32 undef, i32 undef, i32 8, i32 8, i32 14, i32 8>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_07_uu_00_00_00_00_00_00_08_08_uu_uu_08_08_08_15(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_07_uu_00_00_00_00_00_00_08_08_uu_uu_08_08_08_15
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[14,15,2,3,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,1,4,5,6,7,0,1,0,1,0,1,14,15]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_07_uu_00_00_00_00_00_00_08_08_uu_uu_08_08_08_15
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[14,15,2,3,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm0
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,1,4,5,6,7,0,1,0,1,0,1,14,15]
+; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <i32 7, i32 undef, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 8, i32 8, i32 undef, i32 undef, i32 8, i32 8, i32 8, i32 15>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_00_uu_uu_02_04_04_uu_06_14_14_uu_12_10_10_08_08(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_00_uu_uu_02_04_04_uu_06_14_14_uu_12_10_10_08_08
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm1 = xmm0[0,1,2,2,4,5,6,7]
+; AVX1-NEXT:    vpshufhw {{.*}} # xmm1 = xmm1[0,1,2,3,4,4,6,6]
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[12,13,12,13,12,13,8,9,4,5,4,5,0,1,0,1]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_00_uu_uu_02_04_04_uu_06_14_14_uu_12_10_10_08_08
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm1 = xmm0[0,1,2,2,4,5,6,7]
+; AVX2-NEXT:    vpshufhw {{.*}} # xmm1 = xmm1[0,1,2,3,4,4,6,6]
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm0
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[12,13,12,13,12,13,8,9,4,5,4,5,0,1,0,1]
+; 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 undef, i32 undef, i32 2, i32 4, i32 4, i32 undef, i32 6, i32 14, i32 14, i32 undef, i32 12, i32 10, i32 10, i32 8, i32 8>
+  ret <16 x i16> %shuffle
+}
+
+define <16 x i16> @shuffle_v16i16_04_04_04_04_uu_uu_uu_uu_08_08_08_uu_uu_12_12_12(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: @shuffle_v16i16_04_04_04_04_uu_uu_uu_uu_08_08_08_uu_uu_12_12_12
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufd {{.*}} # xmm1 = xmm0[2,1,2,3]
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm1 = xmm1[0,0,0,0,4,5,6,7]
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm0 = xmm0[0,0,0,3,4,5,6,7]
+; AVX1-NEXT:    vpshufhw {{.*}} # xmm0 = xmm0[0,1,2,3,4,4,4,4]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v16i16_04_04_04_04_uu_uu_uu_uu_08_08_08_uu_uu_12_12_12
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpshufd {{.*}} # xmm1 = xmm0[2,1,2,3]
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm1 = xmm1[0,0,0,0,4,5,6,7]
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm0
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm0 = xmm0[0,0,0,3,4,5,6,7]
+; AVX2-NEXT:    vpshufhw {{.*}} # xmm0 = xmm0[0,1,2,3,4,4,4,4]
+; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <i32 4, i32 4, i32 4, i32 4, i32 undef, i32 undef, i32 undef, i32 undef, i32 8, i32 8, i32 8, i32 undef, i32 undef, i32 12, i32 12, i32 12>
+  ret <16 x i16> %shuffle
+}
diff --git a/test/CodeGen/X86/vector-shuffle-256-v32.ll b/test/CodeGen/X86/vector-shuffle-256-v32.ll
new file mode 100644
index 00000000000..9851e4d786d
--- /dev/null
+++ b/test/CodeGen/X86/vector-shuffle-256-v32.ll
@@ -0,0 +1,1732 @@
+; RUN: llc < %s -mcpu=x86-64 -mattr=+avx -x86-experimental-vector-shuffle-lowering | FileCheck %s --check-prefix=ALL --check-prefix=AVX --check-prefix=AVX1
+; RUN: llc < %s -mcpu=x86-64 -mattr=+avx2 -x86-experimental-vector-shuffle-lowering | FileCheck %s --check-prefix=ALL --check-prefix=AVX --check-prefix=AVX2
+
+target triple = "x86_64-unknown-unknown"
+
+define <32 x i8> @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(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-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
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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:    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
+}
+
+define <32 x i8> @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(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-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
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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:    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
+  %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 1, i32 0>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @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(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-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
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0,0,0,0,0,0,2,0,0]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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:    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
+  %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 2, i32 0, i32 0>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @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(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-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
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0,0,0,0,0,3,0,0,0]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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:    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
+  %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 3, i32 0, i32 0, i32 0>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @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(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-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
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0,0,0,0,4,0,0,0,0]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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:    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
+  %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 4, i32 0, i32 0, i32 0, i32 0>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @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(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-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
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0,0,0,5,0,0,0,0,0]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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:    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
+  %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 5, i32 0, i32 0, i32 0, i32 0, i32 0>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @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(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-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
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0,0,6,0,0,0,0,0,0]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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:    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
+  %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 6, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @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(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-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
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0,7,0,0,0,0,0,0,0]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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:    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
+  %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 7, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @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(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-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
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,8,0,0,0,0,0,0,0,0]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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:    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
+  %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 8, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @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(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-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
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,9,0,0,0,0,0,0,0,0,0]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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:    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
+  %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 9, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @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(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-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
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,10,0,0,0,0,0,0,0,0,0,0]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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:    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
+  %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 10, 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
+}
+
+define <32 x i8> @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(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-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
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,11,0,0,0,0,0,0,0,0,0,0,0]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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:    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
+  %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 11, 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
+}
+
+define <32 x i8> @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(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-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
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,12,0,0,0,0,0,0,0,0,0,0,0,0]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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:    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
+  %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 12, 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
+}
+
+define <32 x i8> @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(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-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
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,13,0,0,0,0,0,0,0,0,0,0,0,0,0]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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:    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
+  %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 13, 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
+}
+
+define <32 x i8> @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(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-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
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,14,0,0,0,0,0,0,0,0,0,0,0,0,0,0]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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:    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
+  %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 14, 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
+}
+
+define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_15_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_15_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00
+; AVX1:       # BB#0:
+; AVX1-NEXT:    movl $15, %eax
+; AVX1-NEXT:    vmovd %eax, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX1-NEXT:    vpxor %xmm2, %xmm2, %xmm2
+; AVX1-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_15_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00
+; AVX2:       # BB#0:
+; 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:    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>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @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(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-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
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm2
+; AVX1-NEXT:    vpshufb {{.*}} # xmm2 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm2[0]
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0],zero
+; AVX1-NEXT:    vpor %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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 %xmm1, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm2
+; AVX2-NEXT:    vpshufb {{.*}} # xmm2 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm2[0]
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0],zero
+; AVX2-NEXT:    vpor %xmm2, %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 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
+}
+
+define <32 x i8> @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(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-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
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm2
+; AVX1-NEXT:    vpshufb {{.*}} # xmm2 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm2[1],zero
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0,0,0,0,0,0,0],zero,xmm0[0]
+; AVX1-NEXT:    vpor %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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 %xmm1, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm2
+; AVX2-NEXT:    vpshufb {{.*}} # xmm2 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm2[1],zero
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0,0,0,0,0,0,0],zero,xmm0[0]
+; AVX2-NEXT:    vpor %xmm2, %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 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
+}
+
+define <32 x i8> @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(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-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
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm2
+; AVX1-NEXT:    vpshufb {{.*}} # xmm2 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm2[2],zero,zero
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0,0,0,0,0,0],zero,xmm0[0,0]
+; AVX1-NEXT:    vpor %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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 %xmm1, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm2
+; AVX2-NEXT:    vpshufb {{.*}} # xmm2 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm2[2],zero,zero
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0,0,0,0,0,0],zero,xmm0[0,0]
+; AVX2-NEXT:    vpor %xmm2, %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 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
+}
+
+define <32 x i8> @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(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-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
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm2
+; AVX1-NEXT:    vpshufb {{.*}} # xmm2 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm2[3],zero,zero,zero
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0,0,0,0,0],zero,xmm0[0,0,0]
+; AVX1-NEXT:    vpor %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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 %xmm1, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm2
+; AVX2-NEXT:    vpshufb {{.*}} # xmm2 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm2[3],zero,zero,zero
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0,0,0,0,0],zero,xmm0[0,0,0]
+; AVX2-NEXT:    vpor %xmm2, %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 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
+}
+
+define <32 x i8> @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(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-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
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm2
+; AVX1-NEXT:    vpshufb {{.*}} # xmm2 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm2[4],zero,zero,zero,zero
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0,0,0,0],zero,xmm0[0,0,0,0]
+; AVX1-NEXT:    vpor %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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 %xmm1, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm2
+; AVX2-NEXT:    vpshufb {{.*}} # xmm2 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm2[4],zero,zero,zero,zero
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0,0,0,0],zero,xmm0[0,0,0,0]
+; AVX2-NEXT:    vpor %xmm2, %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 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
+}
+
+define <32 x i8> @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(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-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
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm2
+; AVX1-NEXT:    vpshufb {{.*}} # xmm2 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm2[5],zero,zero,zero,zero,zero
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0,0,0],zero,xmm0[0,0,0,0,0]
+; AVX1-NEXT:    vpor %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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 %xmm1, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm2
+; AVX2-NEXT:    vpshufb {{.*}} # xmm2 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm2[5],zero,zero,zero,zero,zero
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0,0,0],zero,xmm0[0,0,0,0,0]
+; AVX2-NEXT:    vpor %xmm2, %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 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
+}
+
+define <32 x i8> @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(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-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
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm2
+; AVX1-NEXT:    vpshufb {{.*}} # xmm2 = zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm2[6],zero,zero,zero,zero,zero,zero
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0,0],zero,xmm0[0,0,0,0,0,0]
+; AVX1-NEXT:    vpor %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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 %xmm1, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm2
+; AVX2-NEXT:    vpshufb {{.*}} # xmm2 = zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm2[6],zero,zero,zero,zero,zero,zero
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0,0],zero,xmm0[0,0,0,0,0,0]
+; AVX2-NEXT:    vpor %xmm2, %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 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
+}
+
+define <32 x i8> @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(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-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
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm2
+; AVX1-NEXT:    vpshufb {{.*}} # xmm2 = zero,zero,zero,zero,zero,zero,zero,zero,xmm2[7],zero,zero,zero,zero,zero,zero,zero
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0],zero,xmm0[0,0,0,0,0,0,0]
+; AVX1-NEXT:    vpor %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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 %xmm1, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm2
+; AVX2-NEXT:    vpshufb {{.*}} # xmm2 = zero,zero,zero,zero,zero,zero,zero,zero,xmm2[7],zero,zero,zero,zero,zero,zero,zero
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0],zero,xmm0[0,0,0,0,0,0,0]
+; AVX2-NEXT:    vpor %xmm2, %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 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
+}
+
+define <32 x i8> @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(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-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
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm2
+; AVX1-NEXT:    vpshufb {{.*}} # xmm2 = zero,zero,zero,zero,zero,zero,zero,xmm2[8],zero,zero,zero,zero,zero,zero,zero,zero
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0],zero,xmm0[0,0,0,0,0,0,0,0]
+; AVX1-NEXT:    vpor %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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 %xmm1, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm2
+; AVX2-NEXT:    vpshufb {{.*}} # xmm2 = zero,zero,zero,zero,zero,zero,zero,xmm2[8],zero,zero,zero,zero,zero,zero,zero,zero
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0],zero,xmm0[0,0,0,0,0,0,0,0]
+; AVX2-NEXT:    vpor %xmm2, %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 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
+}
+
+define <32 x i8> @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(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-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
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm2
+; AVX1-NEXT:    vpshufb {{.*}} # xmm2 = zero,zero,zero,zero,zero,zero,xmm2[9],zero,zero,zero,zero,zero,zero,zero,zero,zero
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0],zero,xmm0[0,0,0,0,0,0,0,0,0]
+; AVX1-NEXT:    vpor %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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 %xmm1, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm2
+; AVX2-NEXT:    vpshufb {{.*}} # xmm2 = zero,zero,zero,zero,zero,zero,xmm2[9],zero,zero,zero,zero,zero,zero,zero,zero,zero
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0],zero,xmm0[0,0,0,0,0,0,0,0,0]
+; AVX2-NEXT:    vpor %xmm2, %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 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
+}
+
+define <32 x i8> @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(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-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
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm2
+; AVX1-NEXT:    vpshufb {{.*}} # xmm2 = zero,zero,zero,zero,zero,xmm2[10],zero,zero,zero,zero,zero,zero,zero,zero,zero,zero
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0],zero,xmm0[0,0,0,0,0,0,0,0,0,0]
+; AVX1-NEXT:    vpor %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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 %xmm1, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm2
+; AVX2-NEXT:    vpshufb {{.*}} # xmm2 = zero,zero,zero,zero,zero,xmm2[10],zero,zero,zero,zero,zero,zero,zero,zero,zero,zero
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0],zero,xmm0[0,0,0,0,0,0,0,0,0,0]
+; AVX2-NEXT:    vpor %xmm2, %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 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
+}
+
+define <32 x i8> @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(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-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
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm2
+; AVX1-NEXT:    vpshufb {{.*}} # xmm2 = zero,zero,zero,zero,xmm2[11],zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0],zero,xmm0[0,0,0,0,0,0,0,0,0,0,0]
+; AVX1-NEXT:    vpor %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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 %xmm1, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm2
+; AVX2-NEXT:    vpshufb {{.*}} # xmm2 = zero,zero,zero,zero,xmm2[11],zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0],zero,xmm0[0,0,0,0,0,0,0,0,0,0,0]
+; AVX2-NEXT:    vpor %xmm2, %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 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
+}
+
+define <32 x i8> @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(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-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
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm2
+; AVX1-NEXT:    vpshufb {{.*}} # xmm2 = zero,zero,zero,xmm2[12],zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0],zero,xmm0[0,0,0,0,0,0,0,0,0,0,0,0]
+; AVX1-NEXT:    vpor %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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 %xmm1, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm2
+; AVX2-NEXT:    vpshufb {{.*}} # xmm2 = zero,zero,zero,xmm2[12],zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0],zero,xmm0[0,0,0,0,0,0,0,0,0,0,0,0]
+; AVX2-NEXT:    vpor %xmm2, %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 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
+}
+
+define <32 x i8> @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(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-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
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm2
+; AVX1-NEXT:    vpshufb {{.*}} # xmm2 = zero,zero,xmm2[13],zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0],zero,xmm0[0,0,0,0,0,0,0,0,0,0,0,0,0]
+; AVX1-NEXT:    vpor %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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 %xmm1, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm2
+; AVX2-NEXT:    vpshufb {{.*}} # xmm2 = zero,zero,xmm2[13],zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0],zero,xmm0[0,0,0,0,0,0,0,0,0,0,0,0,0]
+; AVX2-NEXT:    vpor %xmm2, %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 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
+}
+
+define <32 x i8> @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(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-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
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm2
+; AVX1-NEXT:    vpshufb {{.*}} # xmm2 = zero,xmm2[14],zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0],zero,xmm0[0,0,0,0,0,0,0,0,0,0,0,0,0,0]
+; AVX1-NEXT:    vpor %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; 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 %xmm1, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm2
+; AVX2-NEXT:    vpshufb {{.*}} # xmm2 = zero,xmm2[14],zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0],zero,xmm0[0,0,0,0,0,0,0,0,0,0,0,0,0,0]
+; AVX2-NEXT:    vpor %xmm2, %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 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
+}
+
+define <32 x i8> @shuffle_v32i8_31_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(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_31_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
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX1-NEXT:    movl $128, %eax
+; AVX1-NEXT:    vmovd %eax, %xmm2
+; AVX1-NEXT:    vpshufb %xmm2, %xmm0, %xmm2
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[15],zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero
+; AVX1-NEXT:    vpor %xmm0, %xmm2, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_31_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, %xmm1
+; AVX2-NEXT:    movl $128, %eax
+; AVX2-NEXT:    vmovd %eax, %xmm2
+; AVX2-NEXT:    vpshufb %xmm2, %xmm0, %xmm2
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm0
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[15],zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero
+; AVX2-NEXT:    vpor %xmm0, %xmm2, %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 31, 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
+}
+
+define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_16_16_16_16_16_16_16_16_16_16_16_16_16_16_16_16(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_16_16_16_16_16_16_16_16_16_16_16_16_16_16_16_16
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT:    vpxor %xmm2, %xmm2, %xmm2
+; AVX1-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_16_16_16_16_16_16_16_16_16_16_16_16_16_16_16_16
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vpxor %xmm2, %xmm2, %xmm2
+; AVX2-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm2, %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 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_15_15_15_15_15_15_15_15_15_15_15_15_15_15_15_15_31_31_31_31_31_31_31_31_31_31_31_31_31_31_31_31(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_15_15_15_15_15_15_15_15_15_15_15_15_15_15_15_15_31_31_31_31_31_31_31_31_31_31_31_31_31_31_31_31
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT:    vmovdqa .LCPI33_0(%rip), %xmm2
+; AVX1-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_15_15_15_15_15_15_15_15_15_15_15_15_15_15_15_15_31_31_31_31_31_31_31_31_31_31_31_31_31_31_31_31
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vmovdqa .LCPI33_0(%rip), %xmm2
+; AVX2-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm2, %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 15, i32 15, i32 15, i32 15, i32 15, i32 15, i32 15, i32 15, i32 15, i32 15, i32 15, i32 15, i32 15, i32 15, i32 15, i32 15, i32 31, i32 31, i32 31, i32 31, i32 31, i32 31, i32 31, i32 31, i32 31, i32 31, i32 31, i32 31, i32 31, i32 31, i32 31, i32 31>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_08_08_08_08_08_08_08_08_16_16_16_16_16_16_16_16_24_24_24_24_24_24_24_24(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_00_00_00_00_00_00_00_00_08_08_08_08_08_08_08_08_16_16_16_16_16_16_16_16_24_24_24_24_24_24_24_24
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT:    vmovdqa .LCPI34_0(%rip), %xmm2
+; AVX1-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_00_00_00_00_00_00_00_00_08_08_08_08_08_08_08_08_16_16_16_16_16_16_16_16_24_24_24_24_24_24_24_24
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vmovdqa .LCPI34_0(%rip), %xmm2
+; AVX2-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm2, %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 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 24, i32 24, i32 24, i32 24, i32 24, i32 24, i32 24, i32 24>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_07_07_07_07_07_07_07_07_15_15_15_15_15_15_15_15_23_23_23_23_23_23_23_23_31_31_31_31_31_31_31_31(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_07_07_07_07_07_07_07_07_15_15_15_15_15_15_15_15_23_23_23_23_23_23_23_23_31_31_31_31_31_31_31_31
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT:    vmovdqa .LCPI35_0(%rip), %xmm2
+; AVX1-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_07_07_07_07_07_07_07_07_15_15_15_15_15_15_15_15_23_23_23_23_23_23_23_23_31_31_31_31_31_31_31_31
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vmovdqa .LCPI35_0(%rip), %xmm2
+; AVX2-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm2, %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 7, i32 7, i32 7, i32 7, i32 7, i32 7, i32 7, i32 7, i32 15, i32 15, i32 15, i32 15, i32 15, i32 15, i32 15, i32 15, i32 23, i32 23, i32 23, i32 23, i32 23, i32 23, i32 23, i32 23, i32 31, i32 31, i32 31, i32 31, i32 31, i32 31, i32 31, i32 31>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_00_00_00_00_04_04_04_04_08_08_08_08_12_12_12_12_16_16_16_16_20_20_20_20_24_24_24_24_28_28_28_28(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_00_00_00_00_04_04_04_04_08_08_08_08_12_12_12_12_16_16_16_16_20_20_20_20_24_24_24_24_28_28_28_28
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT:    vmovdqa .LCPI36_0(%rip), %xmm2
+; AVX1-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_00_00_00_00_04_04_04_04_08_08_08_08_12_12_12_12_16_16_16_16_20_20_20_20_24_24_24_24_28_28_28_28
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vmovdqa .LCPI36_0(%rip), %xmm2
+; AVX2-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm2, %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 4, i32 4, i32 4, i32 4, i32 8, i32 8, i32 8, i32 8, i32 12, i32 12, i32 12, i32 12, i32 16, i32 16, i32 16, i32 16, i32 20, i32 20, i32 20, i32 20, i32 24, i32 24, i32 24, i32 24, i32 28, i32 28, i32 28, i32 28>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_03_03_03_03_07_07_07_07_11_11_11_11_15_15_15_15_19_19_19_19_23_23_23_23_27_27_27_27_31_31_31_31(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_03_03_03_03_07_07_07_07_11_11_11_11_15_15_15_15_19_19_19_19_23_23_23_23_27_27_27_27_31_31_31_31
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT:    vmovdqa .LCPI37_0(%rip), %xmm2
+; AVX1-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_03_03_03_03_07_07_07_07_11_11_11_11_15_15_15_15_19_19_19_19_23_23_23_23_27_27_27_27_31_31_31_31
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vmovdqa .LCPI37_0(%rip), %xmm2
+; AVX2-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm2, %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 3, i32 3, i32 3, i32 3, i32 7, i32 7, i32 7, i32 7, i32 11, i32 11, i32 11, i32 11, i32 15, i32 15, i32 15, i32 15, i32 19, i32 19, i32 19, i32 19, i32 23, i32 23, i32 23, i32 23, i32 27, i32 27, i32 27, i32 27, i32 31, i32 31, i32 31, i32 31>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_00_00_02_02_04_04_06_06_08_08_10_10_12_12_14_14_16_16_18_18_20_20_22_22_24_24_26_26_28_28_30_30(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_00_00_02_02_04_04_06_06_08_08_10_10_12_12_14_14_16_16_18_18_20_20_22_22_24_24_26_26_28_28_30_30
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT:    vmovdqa .LCPI38_0(%rip), %xmm2
+; AVX1-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_00_00_02_02_04_04_06_06_08_08_10_10_12_12_14_14_16_16_18_18_20_20_22_22_24_24_26_26_28_28_30_30
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vmovdqa .LCPI38_0(%rip), %xmm2
+; AVX2-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm2, %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 2, i32 2, i32 4, i32 4, i32 6, i32 6, i32 8, i32 8, i32 10, i32 10, i32 12, i32 12, i32 14, i32 14, i32 16, i32 16, i32 18, i32 18, i32 20, i32 20, i32 22, i32 22, i32 24, i32 24, i32 26, i32 26, i32 28, i32 28, i32 30, i32 30>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_01_01_03_03_05_05_07_07_09_09_11_11_13_13_15_15_17_17_19_19_21_21_23_23_25_25_27_27_29_29_31_31(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_01_01_03_03_05_05_07_07_09_09_11_11_13_13_15_15_17_17_19_19_21_21_23_23_25_25_27_27_29_29_31_31
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT:    vmovdqa .LCPI39_0(%rip), %xmm2
+; AVX1-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_01_01_03_03_05_05_07_07_09_09_11_11_13_13_15_15_17_17_19_19_21_21_23_23_25_25_27_27_29_29_31_31
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vmovdqa .LCPI39_0(%rip), %xmm2
+; AVX2-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm2, %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 1, i32 1, i32 3, i32 3, i32 5, i32 5, i32 7, i32 7, i32 9, i32 9, i32 11, i32 11, i32 13, i32 13, i32 15, i32 15, i32 17, i32 17, i32 19, i32 19, i32 21, i32 21, i32 23, i32 23, i32 25, i32 25, i32 27, i32 27, i32 29, i32 29, i32 31, i32 31>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_01_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_01_00(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_01_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_01_00
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_01_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_01_00
+; AVX2:       # BB#0:
+; 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, %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 1, 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 1, i32 0>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_02_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_02_00_00(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_02_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_02_00_00
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0,0,0,0,0,0,2,0,0]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_02_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_02_00_00
+; AVX2:       # BB#0:
+; 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, %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 2, 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 2, i32 0, i32 0>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_07_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_07_00_00_00_00_00_00_00(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_00_00_00_00_00_00_00_00_07_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_07_00_00_00_00_00_00_00
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0,7,0,0,0,0,0,0,0]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_00_00_00_00_00_00_00_00_07_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:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0,7,0,0,0,0,0,0,0]
+; AVX2-NEXT:    vinserti128 $1, %xmm0, %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 7, 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 7, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_08_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_08_00_00_00_00_00_00_00_00(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_00_00_00_00_00_00_00_08_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_08_00_00_00_00_00_00_00_00
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,8,0,0,0,0,0,0,0,0]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_00_00_00_00_00_00_00_08_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:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,8,0,0,0,0,0,0,0,0]
+; AVX2-NEXT:    vinserti128 $1, %xmm0, %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 8, 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 8, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_00_14_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(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_00_14_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
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,14,0,0,0,0,0,0,0,0,0,0,0,0,0,0]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_00_14_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:    vpshufb {{.*}} # xmm0 = xmm0[0,14,0,0,0,0,0,0,0,0,0,0,0,0,0,0]
+; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <i32 0, i32 14, 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 14, 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
+}
+
+define <32 x i8> @shuffle_v32i8_15_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_15_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_15_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_15_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00
+; AVX1:       # BB#0:
+; AVX1-NEXT:    movl $15, %eax
+; AVX1-NEXT:    vmovd %eax, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_15_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_15_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00
+; AVX2:       # BB#0:
+; AVX2-NEXT:    movl $15, %eax
+; AVX2-NEXT:    vmovd %eax, %xmm1
+; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm0
+; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <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, 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>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_32_32_02_02_36_36_06_06_40_40_10_10_44_44_14_14_48_48_18_18_52_52_22_22_56_56_26_26_60_60_30_30(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_32_32_02_02_36_36_06_06_40_40_10_10_44_44_14_14_48_48_18_18_52_52_22_22_56_56_26_26_60_60_30_30
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm2
+; AVX1-NEXT:    vmovdqa .LCPI46_0(%rip), %xmm3
+; AVX1-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
+; AVX1-NEXT:    vextractf128 $1, %ymm1, %xmm4
+; AVX1-NEXT:    vmovdqa .LCPI46_1(%rip), %xmm5
+; AVX1-NEXT:    vpshufb %xmm5, %xmm4, %xmm4
+; AVX1-NEXT:    vpblendw {{.*}} # xmm2 = xmm4[0],xmm2[1],xmm4[2],xmm2[3],xmm4[4],xmm2[5],xmm4[6],xmm2[7]
+; AVX1-NEXT:    vpshufb %xmm3, %xmm0, %xmm0
+; AVX1-NEXT:    vpshufb %xmm5, %xmm1, %xmm1
+; AVX1-NEXT:    vpblendw {{.*}} # xmm0 = xmm1[0],xmm0[1],xmm1[2],xmm0[3],xmm1[4],xmm0[5],xmm1[6],xmm0[7]
+; AVX1-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_32_32_02_02_36_36_06_06_40_40_10_10_44_44_14_14_48_48_18_18_52_52_22_22_56_56_26_26_60_60_30_30
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm2
+; AVX2-NEXT:    vmovdqa .LCPI46_0(%rip), %xmm3
+; AVX2-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
+; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm4
+; AVX2-NEXT:    vmovdqa .LCPI46_1(%rip), %xmm5
+; AVX2-NEXT:    vpshufb %xmm5, %xmm4, %xmm4
+; AVX2-NEXT:    vpblendw {{.*}} # xmm2 = xmm4[0],xmm2[1],xmm4[2],xmm2[3],xmm4[4],xmm2[5],xmm4[6],xmm2[7]
+; AVX2-NEXT:    vpshufb %xmm3, %xmm0, %xmm0
+; AVX2-NEXT:    vpshufb %xmm5, %xmm1, %xmm1
+; AVX2-NEXT:    vpblendw {{.*}} # xmm0 = xmm1[0],xmm0[1],xmm1[2],xmm0[3],xmm1[4],xmm0[5],xmm1[6],xmm0[7]
+; AVX2-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <i32 32, i32 32, i32 2, i32 2, i32 36, i32 36, i32 6, i32 6, i32 40, i32 40, i32 10, i32 10, i32 44, i32 44, i32 14, i32 14, i32 48, i32 48, i32 18, i32 18, i32 52, i32 52, i32 22, i32 22, i32 56, i32 56, i32 26, i32 26, i32 60, i32 60, i32 30, i32 30>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_00_00_32_32_00_00_32_32_00_00_32_32_00_00_32_32_00_00_32_32_00_00_32_32_00_00_32_32_00_00_32_32(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_00_00_32_32_00_00_32_32_00_00_32_32_00_00_32_32_00_00_32_32_00_00_32_32_00_00_32_32_00_00_32_32
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm1 = xmm1[0,0,0,0,2,2,0,0,0,0,0,0,6,6,0,0]
+; AVX1-NEXT:    vpunpcklbw {{.*}} # xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; AVX1-NEXT:    vpshufd {{.*}} # xmm0 = xmm0[0,0,0,0]
+; AVX1-NEXT:    vpblendw {{.*}} # xmm0 = xmm0[0],xmm1[1],xmm0[2],xmm1[3],xmm0[4],xmm1[5],xmm0[6],xmm1[7]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_00_00_32_32_00_00_32_32_00_00_32_32_00_00_32_32_00_00_32_32_00_00_32_32_00_00_32_32_00_00_32_32
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpshufb {{.*}} # xmm1 = xmm1[0,0,0,0,2,2,0,0,0,0,0,0,6,6,0,0]
+; AVX2-NEXT:    vpunpcklbw {{.*}} # xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; AVX2-NEXT:    vpshufd {{.*}} # xmm0 = xmm0[0,0,0,0]
+; 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
+  %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <i32 0, i32 0, i32 32, i32 32, i32 0, i32 0, i32 32, i32 32, i32 0, i32 0, i32 32, i32 32, i32 0, i32 0, i32 32, i32 32, i32 0, i32 0, i32 32, i32 32, i32 0, i32 0, i32 32, i32 32, i32 0, i32 0, i32 32, i32 32, i32 0, i32 0, i32 32, i32 32>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_00_00_32_32_00_00_32_32_00_00_32_32_00_00_32_32_16_16_48_48_16_16_48_48_16_16_48_48_16_16_48_48(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_00_00_32_32_00_00_32_32_00_00_32_32_00_00_32_32_16_16_48_48_16_16_48_48_16_16_48_48_16_16_48_48
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm1, %xmm2
+; AVX1-NEXT:    vmovdqa .LCPI48_0(%rip), %xmm3
+; AVX1-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm4
+; AVX1-NEXT:    vpunpcklbw {{.*}} # xmm4 = xmm4[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; AVX1-NEXT:    vpshufd {{.*}} # xmm4 = xmm4[0,0,0,0]
+; AVX1-NEXT:    vpblendw {{.*}} # xmm2 = xmm4[0],xmm2[1],xmm4[2],xmm2[3],xmm4[4],xmm2[5],xmm4[6],xmm2[7]
+; AVX1-NEXT:    vpshufb %xmm3, %xmm1, %xmm1
+; AVX1-NEXT:    vpunpcklbw {{.*}} # xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; AVX1-NEXT:    vpshufd {{.*}} # xmm0 = xmm0[0,0,0,0]
+; AVX1-NEXT:    vpblendw {{.*}} # xmm0 = xmm0[0],xmm1[1],xmm0[2],xmm1[3],xmm0[4],xmm1[5],xmm0[6],xmm1[7]
+; AVX1-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_00_00_32_32_00_00_32_32_00_00_32_32_00_00_32_32_16_16_48_48_16_16_48_48_16_16_48_48_16_16_48_48
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm2
+; AVX2-NEXT:    vmovdqa .LCPI48_0(%rip), %xmm3
+; AVX2-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm4
+; AVX2-NEXT:    vpunpcklbw {{.*}} # xmm4 = xmm4[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; AVX2-NEXT:    vpshufd {{.*}} # xmm4 = xmm4[0,0,0,0]
+; AVX2-NEXT:    vpblendw {{.*}} # xmm2 = xmm4[0],xmm2[1],xmm4[2],xmm2[3],xmm4[4],xmm2[5],xmm4[6],xmm2[7]
+; AVX2-NEXT:    vpshufb %xmm3, %xmm1, %xmm1
+; AVX2-NEXT:    vpunpcklbw {{.*}} # xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; AVX2-NEXT:    vpshufd {{.*}} # xmm0 = xmm0[0,0,0,0]
+; AVX2-NEXT:    vpblendw {{.*}} # xmm0 = xmm0[0],xmm1[1],xmm0[2],xmm1[3],xmm0[4],xmm1[5],xmm0[6],xmm1[7]
+; AVX2-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <i32 0, i32 0, i32 32, i32 32, i32 0, i32 0, i32 32, i32 32, i32 0, i32 0, i32 32, i32 32, i32 0, i32 0, i32 32, i32 32, i32 16, i32 16, i32 48, i32 48, i32 16, i32 16, i32 48, i32 48, i32 16, i32 16, i32 48, i32 48, i32 16, i32 16, i32 48, i32 48>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_32_32_32_32_32_32_32_32_08_08_10_10_12_12_14_14_48_48_48_48_48_48_48_48_24_24_26_26_28_28_30_30(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_32_32_32_32_32_32_32_32_08_08_10_10_12_12_14_14_48_48_48_48_48_48_48_48_24_24_26_26_28_28_30_30
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm2
+; AVX1-NEXT:    vmovdqa .LCPI49_0(%rip), %xmm3
+; AVX1-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
+; AVX1-NEXT:    vextractf128 $1, %ymm1, %xmm4
+; AVX1-NEXT:    vpunpcklbw {{.*}} # xmm4 = xmm4[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm4 = xmm4[0,0,0,0,4,5,6,7]
+; AVX1-NEXT:    vpblendw {{.*}} # xmm2 = xmm4[0,1,2,3],xmm2[4,5,6,7]
+; AVX1-NEXT:    vpshufb %xmm3, %xmm0, %xmm0
+; 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:    vpblendw {{.*}} # xmm0 = xmm1[0,1,2,3],xmm0[4,5,6,7]
+; AVX1-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_32_32_32_32_32_32_32_32_08_08_10_10_12_12_14_14_48_48_48_48_48_48_48_48_24_24_26_26_28_28_30_30
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm2
+; AVX2-NEXT:    vmovdqa .LCPI49_0(%rip), %xmm3
+; AVX2-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
+; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm4
+; AVX2-NEXT:    vpunpcklbw {{.*}} # xmm4 = xmm4[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm4 = xmm4[0,0,0,0,4,5,6,7]
+; AVX2-NEXT:    vpblendw {{.*}} # xmm2 = xmm4[0,1,2,3],xmm2[4,5,6,7]
+; AVX2-NEXT:    vpshufb %xmm3, %xmm0, %xmm0
+; AVX2-NEXT:    vpunpcklbw {{.*}} # xmm1 = xmm1[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm1 = xmm1[0,0,0,0,4,5,6,7]
+; AVX2-NEXT:    vpblendw {{.*}} # xmm0 = xmm1[0,1,2,3],xmm0[4,5,6,7]
+; AVX2-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <i32 32, i32 32, i32 32, i32 32, i32 32, i32 32, i32 32, i32 32, i32 8, i32 8, i32 10, i32 10, i32 12, i32 12, i32 14, i32 14, i32 48, i32 48, i32 48, i32 48, i32 48, i32 48, i32 48, i32 48, i32 24, i32 24, i32 26, i32 26, i32 28, i32 28, i32 30, i32 30>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_38_38_36_36_34_34_32_32_14_14_12_12_10_10_08_08_54_54_52_52_50_50_48_48_30_30_28_28_26_26_24_24(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_38_38_36_36_34_34_32_32_14_14_12_12_10_10_08_08_54_54_52_52_50_50_48_48_30_30_28_28_26_26_24_24
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm2
+; AVX1-NEXT:    vmovdqa .LCPI50_0(%rip), %xmm3
+; AVX1-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
+; AVX1-NEXT:    vextractf128 $1, %ymm1, %xmm4
+; AVX1-NEXT:    vmovdqa .LCPI50_1(%rip), %xmm5
+; AVX1-NEXT:    vpshufb %xmm5, %xmm4, %xmm4
+; AVX1-NEXT:    vpblendw {{.*}} # xmm2 = xmm4[0,1,2,3],xmm2[4,5,6,7]
+; AVX1-NEXT:    vpshufb %xmm3, %xmm0, %xmm0
+; AVX1-NEXT:    vpshufb %xmm5, %xmm1, %xmm1
+; AVX1-NEXT:    vpblendw {{.*}} # xmm0 = xmm1[0,1,2,3],xmm0[4,5,6,7]
+; AVX1-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_38_38_36_36_34_34_32_32_14_14_12_12_10_10_08_08_54_54_52_52_50_50_48_48_30_30_28_28_26_26_24_24
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm2
+; AVX2-NEXT:    vmovdqa .LCPI50_0(%rip), %xmm3
+; AVX2-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
+; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm4
+; AVX2-NEXT:    vmovdqa .LCPI50_1(%rip), %xmm5
+; AVX2-NEXT:    vpshufb %xmm5, %xmm4, %xmm4
+; AVX2-NEXT:    vpblendw {{.*}} # xmm2 = xmm4[0,1,2,3],xmm2[4,5,6,7]
+; AVX2-NEXT:    vpshufb %xmm3, %xmm0, %xmm0
+; AVX2-NEXT:    vpshufb %xmm5, %xmm1, %xmm1
+; AVX2-NEXT:    vpblendw {{.*}} # xmm0 = xmm1[0,1,2,3],xmm0[4,5,6,7]
+; AVX2-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <i32 38, i32 38, i32 36, i32 36, i32 34, i32 34, i32 32, i32 32, i32 14, i32 14, i32 12, i32 12, i32 10, i32 10, i32 8, i32 8, i32 54, i32 54, i32 52, i32 52, i32 50, i32 50, i32 48, i32 48, i32 30, i32 30, i32 28, i32 28, i32 26, i32 26, i32 24, i32 24>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_38_38_36_36_34_34_32_32_06_06_04_04_02_02_00_00_54_54_52_52_50_50_48_48_22_22_20_20_18_18_16_16(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_38_38_36_36_34_34_32_32_06_06_04_04_02_02_00_00_54_54_52_52_50_50_48_48_22_22_20_20_18_18_16_16
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm2
+; AVX1-NEXT:    vmovdqa .LCPI51_0(%rip), %xmm3
+; AVX1-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
+; AVX1-NEXT:    vextractf128 $1, %ymm1, %xmm4
+; AVX1-NEXT:    vmovdqa .LCPI51_1(%rip), %xmm5
+; AVX1-NEXT:    vpshufb %xmm5, %xmm4, %xmm4
+; AVX1-NEXT:    vpblendw {{.*}} # xmm2 = xmm4[0,1,2,3],xmm2[4,5,6,7]
+; AVX1-NEXT:    vpshufb %xmm3, %xmm0, %xmm0
+; AVX1-NEXT:    vpshufb %xmm5, %xmm1, %xmm1
+; AVX1-NEXT:    vpblendw {{.*}} # xmm0 = xmm1[0,1,2,3],xmm0[4,5,6,7]
+; AVX1-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_38_38_36_36_34_34_32_32_06_06_04_04_02_02_00_00_54_54_52_52_50_50_48_48_22_22_20_20_18_18_16_16
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm2
+; AVX2-NEXT:    vmovdqa .LCPI51_0(%rip), %xmm3
+; AVX2-NEXT:    vpshufb %xmm3, %xmm2, %xmm2
+; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm4
+; AVX2-NEXT:    vmovdqa .LCPI51_1(%rip), %xmm5
+; AVX2-NEXT:    vpshufb %xmm5, %xmm4, %xmm4
+; AVX2-NEXT:    vpblendw {{.*}} # xmm2 = xmm4[0,1,2,3],xmm2[4,5,6,7]
+; AVX2-NEXT:    vpshufb %xmm3, %xmm0, %xmm0
+; AVX2-NEXT:    vpshufb %xmm5, %xmm1, %xmm1
+; AVX2-NEXT:    vpblendw {{.*}} # xmm0 = xmm1[0,1,2,3],xmm0[4,5,6,7]
+; AVX2-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <i32 38, i32 38, i32 36, i32 36, i32 34, i32 34, i32 32, i32 32, i32 6, i32 6, i32 4, i32 4, i32 2, i32 2, i32 0, i32 0, i32 54, i32 54, i32 52, i32 52, i32 50, i32 50, i32 48, i32 48, i32 22, i32 22, i32 20, i32 20, i32 18, i32 18, i32 16, i32 16>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_01_00_16_16_16_16_16_16_16_16_16_16_16_16_16_16_17_16(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_01_00_16_16_16_16_16_16_16_16_16_16_16_16_16_16_17_16
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT:    vmovdqa .LCPI52_0(%rip), %xmm2
+; AVX1-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_01_00_16_16_16_16_16_16_16_16_16_16_16_16_16_16_17_16
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vmovdqa .LCPI52_0(%rip), %xmm2
+; AVX2-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm2, %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 1, i32 0, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 17, i32 16>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_02_00_00_16_16_16_16_16_16_16_16_16_16_16_16_16_18_16_16(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_02_00_00_16_16_16_16_16_16_16_16_16_16_16_16_16_18_16_16
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT:    vmovdqa .LCPI53_0(%rip), %xmm2
+; AVX1-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_02_00_00_16_16_16_16_16_16_16_16_16_16_16_16_16_18_16_16
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vmovdqa .LCPI53_0(%rip), %xmm2
+; AVX2-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm2, %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 2, i32 0, i32 0, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 18, i32 16, i32 16>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_07_00_00_00_00_00_00_00_16_16_16_16_16_16_16_16_23_16_16_16_16_16_16_16(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_00_00_00_00_00_00_00_00_07_00_00_00_00_00_00_00_16_16_16_16_16_16_16_16_23_16_16_16_16_16_16_16
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT:    vmovdqa .LCPI54_0(%rip), %xmm2
+; AVX1-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_00_00_00_00_00_00_00_00_07_00_00_00_00_00_00_00_16_16_16_16_16_16_16_16_23_16_16_16_16_16_16_16
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vmovdqa .LCPI54_0(%rip), %xmm2
+; AVX2-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm2, %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 7, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 23, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_08_00_00_00_00_00_00_00_00_16_16_16_16_16_16_16_24_16_16_16_16_16_16_16_16(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_00_00_00_00_00_00_00_08_00_00_00_00_00_00_00_00_16_16_16_16_16_16_16_24_16_16_16_16_16_16_16_16
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT:    vmovdqa .LCPI55_0(%rip), %xmm2
+; AVX1-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_00_00_00_00_00_00_00_08_00_00_00_00_00_00_00_00_16_16_16_16_16_16_16_24_16_16_16_16_16_16_16_16
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vmovdqa .LCPI55_0(%rip), %xmm2
+; AVX2-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm2, %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 8, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 24, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_00_14_00_00_00_00_00_00_00_00_00_00_00_00_00_00_16_30_16_16_16_16_16_16_16_16_16_16_16_16_16_16(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_00_14_00_00_00_00_00_00_00_00_00_00_00_00_00_00_16_30_16_16_16_16_16_16_16_16_16_16_16_16_16_16
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT:    vmovdqa .LCPI56_0(%rip), %xmm2
+; AVX1-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_00_14_00_00_00_00_00_00_00_00_00_00_00_00_00_00_16_30_16_16_16_16_16_16_16_16_16_16_16_16_16_16
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vmovdqa .LCPI56_0(%rip), %xmm2
+; AVX2-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm2, %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 14, 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 30, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_15_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_31_16_16_16_16_16_16_16_16_16_16_16_16_16_16_16(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_15_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_31_16_16_16_16_16_16_16_16_16_16_16_16_16_16_16
+; AVX1:       # BB#0:
+; AVX1-NEXT:    movl $15, %eax
+; AVX1-NEXT:    vmovd %eax, %xmm1
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm2
+; AVX1-NEXT:    vpshufb %xmm1, %xmm2, %xmm2
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_15_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_31_16_16_16_16_16_16_16_16_16_16_16_16_16_16_16
+; AVX2:       # BB#0:
+; AVX2-NEXT:    movl $15, %eax
+; AVX2-NEXT:    vmovd %eax, %xmm1
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm2
+; AVX2-NEXT:    vpshufb %xmm1, %xmm2, %xmm2
+; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm0
+; AVX2-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <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, i32 31, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_00_00_32_32_02_02_34_34_04_04_36_36_06_06_38_38_16_16_48_48_18_18_50_50_20_20_52_52_22_22_54_54(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_00_00_32_32_02_02_34_34_04_04_36_36_06_06_38_38_16_16_48_48_18_18_50_50_20_20_52_52_22_22_54_54
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm2
+; AVX1-NEXT:    vpunpcklbw {{.*}} # xmm2 = xmm2[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; AVX1-NEXT:    vextractf128 $1, %ymm1, %xmm3
+; AVX1-NEXT:    vpunpcklbw {{.*}} # xmm3 = xmm3[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; AVX1-NEXT:    vpalignr {{.*}} # xmm3 = xmm3[14,15,0,1,2,3,4,5,6,7,8,9,10,11,12,13]
+; AVX1-NEXT:    vpblendw {{.*}} # xmm2 = xmm2[0],xmm3[1],xmm2[2],xmm3[3],xmm2[4],xmm3[5],xmm2[6],xmm3[7]
+; AVX1-NEXT:    vpunpcklbw {{.*}} # xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; AVX1-NEXT:    vpunpcklbw {{.*}} # xmm1 = xmm1[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; AVX1-NEXT:    vpalignr {{.*}} # xmm1 = xmm1[14,15,0,1,2,3,4,5,6,7,8,9,10,11,12,13]
+; AVX1-NEXT:    vpblendw {{.*}} # xmm0 = xmm0[0],xmm1[1],xmm0[2],xmm1[3],xmm0[4],xmm1[5],xmm0[6],xmm1[7]
+; AVX1-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_00_00_32_32_02_02_34_34_04_04_36_36_06_06_38_38_16_16_48_48_18_18_50_50_20_20_52_52_22_22_54_54
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm2
+; AVX2-NEXT:    vpunpcklbw {{.*}} # xmm2 = xmm2[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm3
+; AVX2-NEXT:    vpunpcklbw {{.*}} # xmm3 = xmm3[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; AVX2-NEXT:    vpalignr {{.*}} # xmm3 = xmm3[14,15,0,1,2,3,4,5,6,7,8,9,10,11,12,13]
+; AVX2-NEXT:    vpblendw {{.*}} # xmm2 = xmm2[0],xmm3[1],xmm2[2],xmm3[3],xmm2[4],xmm3[5],xmm2[6],xmm3[7]
+; AVX2-NEXT:    vpunpcklbw {{.*}} # xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; AVX2-NEXT:    vpunpcklbw {{.*}} # xmm1 = xmm1[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; AVX2-NEXT:    vpalignr {{.*}} # xmm1 = xmm1[14,15,0,1,2,3,4,5,6,7,8,9,10,11,12,13]
+; AVX2-NEXT:    vpblendw {{.*}} # xmm0 = xmm0[0],xmm1[1],xmm0[2],xmm1[3],xmm0[4],xmm1[5],xmm0[6],xmm1[7]
+; AVX2-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <i32 0, i32 0, i32 32, i32 32, i32 2, i32 2, i32 34, i32 34, i32 4, i32 4, i32 36, i32 36, i32 6, i32 6, i32 38, i32 38, i32 16, i32 16, i32 48, i32 48, i32 18, i32 18, i32 50, i32 50, i32 20, i32 20, i32 52, i32 52, i32 22, i32 22, i32 54, i32 54>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_08_08_40_40_10_10_42_42_12_12_44_44_14_14_46_46_24_24_56_56_26_26_58_58_28_28_60_60_30_30_62_62(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_08_08_40_40_10_10_42_42_12_12_44_44_14_14_46_46_24_24_56_56_26_26_58_58_28_28_60_60_30_30_62_62
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm2
+; AVX1-NEXT:    vpunpckhbw {{.*}} # xmm2 = xmm2[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
+; AVX1-NEXT:    vextractf128 $1, %ymm1, %xmm3
+; AVX1-NEXT:    vpunpckhbw {{.*}} # xmm3 = xmm3[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
+; AVX1-NEXT:    vpalignr {{.*}} # xmm3 = xmm3[14,15,0,1,2,3,4,5,6,7,8,9,10,11,12,13]
+; AVX1-NEXT:    vpblendw {{.*}} # xmm2 = xmm2[0],xmm3[1],xmm2[2],xmm3[3],xmm2[4],xmm3[5],xmm2[6],xmm3[7]
+; AVX1-NEXT:    vpunpckhbw {{.*}} # xmm0 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
+; AVX1-NEXT:    vpunpckhbw {{.*}} # xmm1 = xmm1[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
+; AVX1-NEXT:    vpalignr {{.*}} # xmm1 = xmm1[14,15,0,1,2,3,4,5,6,7,8,9,10,11,12,13]
+; AVX1-NEXT:    vpblendw {{.*}} # xmm0 = xmm0[0],xmm1[1],xmm0[2],xmm1[3],xmm0[4],xmm1[5],xmm0[6],xmm1[7]
+; AVX1-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_08_08_40_40_10_10_42_42_12_12_44_44_14_14_46_46_24_24_56_56_26_26_58_58_28_28_60_60_30_30_62_62
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm2
+; AVX2-NEXT:    vpunpckhbw {{.*}} # xmm2 = xmm2[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
+; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm3
+; AVX2-NEXT:    vpunpckhbw {{.*}} # xmm3 = xmm3[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
+; AVX2-NEXT:    vpalignr {{.*}} # xmm3 = xmm3[14,15,0,1,2,3,4,5,6,7,8,9,10,11,12,13]
+; AVX2-NEXT:    vpblendw {{.*}} # xmm2 = xmm2[0],xmm3[1],xmm2[2],xmm3[3],xmm2[4],xmm3[5],xmm2[6],xmm3[7]
+; AVX2-NEXT:    vpunpckhbw {{.*}} # xmm0 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
+; AVX2-NEXT:    vpunpckhbw {{.*}} # xmm1 = xmm1[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
+; AVX2-NEXT:    vpalignr {{.*}} # xmm1 = xmm1[14,15,0,1,2,3,4,5,6,7,8,9,10,11,12,13]
+; AVX2-NEXT:    vpblendw {{.*}} # xmm0 = xmm0[0],xmm1[1],xmm0[2],xmm1[3],xmm0[4],xmm1[5],xmm0[6],xmm1[7]
+; AVX2-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <i32 8, i32 8, i32 40, i32 40, i32 10, i32 10, i32 42, i32 42, i32 12, i32 12, i32 44, i32 44, i32 14, i32 14, i32 46, i32 46, i32 24, i32 24, i32 56, i32 56, i32 26, i32 26, i32 58, i32 58, i32 28, i32 28, i32 60, i32 60, i32 30, i32 30, i32 62, i32 62>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_00_00_32_32_02_02_34_34_04_04_36_36_06_06_38_38_24_24_56_56_26_26_58_58_28_28_60_60_30_30_62_62(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_00_00_32_32_02_02_34_34_04_04_36_36_06_06_38_38_24_24_56_56_26_26_58_58_28_28_60_60_30_30_62_62
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm2
+; AVX1-NEXT:    vpunpckhbw {{.*}} # xmm2 = xmm2[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
+; AVX1-NEXT:    vextractf128 $1, %ymm1, %xmm3
+; AVX1-NEXT:    vpunpckhbw {{.*}} # xmm3 = xmm3[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
+; AVX1-NEXT:    vpalignr {{.*}} # xmm3 = xmm3[14,15,0,1,2,3,4,5,6,7,8,9,10,11,12,13]
+; AVX1-NEXT:    vpblendw {{.*}} # xmm2 = xmm2[0],xmm3[1],xmm2[2],xmm3[3],xmm2[4],xmm3[5],xmm2[6],xmm3[7]
+; AVX1-NEXT:    vpunpcklbw {{.*}} # xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; AVX1-NEXT:    vpunpcklbw {{.*}} # xmm1 = xmm1[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; AVX1-NEXT:    vpalignr {{.*}} # xmm1 = xmm1[14,15,0,1,2,3,4,5,6,7,8,9,10,11,12,13]
+; AVX1-NEXT:    vpblendw {{.*}} # xmm0 = xmm0[0],xmm1[1],xmm0[2],xmm1[3],xmm0[4],xmm1[5],xmm0[6],xmm1[7]
+; AVX1-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_00_00_32_32_02_02_34_34_04_04_36_36_06_06_38_38_24_24_56_56_26_26_58_58_28_28_60_60_30_30_62_62
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm2
+; AVX2-NEXT:    vpunpckhbw {{.*}} # xmm2 = xmm2[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
+; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm3
+; AVX2-NEXT:    vpunpckhbw {{.*}} # xmm3 = xmm3[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
+; AVX2-NEXT:    vpalignr {{.*}} # xmm3 = xmm3[14,15,0,1,2,3,4,5,6,7,8,9,10,11,12,13]
+; AVX2-NEXT:    vpblendw {{.*}} # xmm2 = xmm2[0],xmm3[1],xmm2[2],xmm3[3],xmm2[4],xmm3[5],xmm2[6],xmm3[7]
+; AVX2-NEXT:    vpunpcklbw {{.*}} # xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; AVX2-NEXT:    vpunpcklbw {{.*}} # xmm1 = xmm1[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; AVX2-NEXT:    vpalignr {{.*}} # xmm1 = xmm1[14,15,0,1,2,3,4,5,6,7,8,9,10,11,12,13]
+; AVX2-NEXT:    vpblendw {{.*}} # xmm0 = xmm0[0],xmm1[1],xmm0[2],xmm1[3],xmm0[4],xmm1[5],xmm0[6],xmm1[7]
+; AVX2-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <i32 0, i32 0, i32 32, i32 32, i32 2, i32 2, i32 34, i32 34, i32 4, i32 4, i32 36, i32 36, i32 6, i32 6, i32 38, i32 38, i32 24, i32 24, i32 56, i32 56, i32 26, i32 26, i32 58, i32 58, i32 28, i32 28, i32 60, i32 60, i32 30, i32 30, i32 62, i32 62>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_08_08_40_40_10_10_42_42_12_12_44_44_14_14_46_46_16_16_48_48_18_18_50_50_20_20_52_52_22_22_54_54(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_08_08_40_40_10_10_42_42_12_12_44_44_14_14_46_46_16_16_48_48_18_18_50_50_20_20_52_52_22_22_54_54
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm2
+; AVX1-NEXT:    vpunpcklbw {{.*}} # xmm2 = xmm2[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; AVX1-NEXT:    vextractf128 $1, %ymm1, %xmm3
+; AVX1-NEXT:    vpunpcklbw {{.*}} # xmm3 = xmm3[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; AVX1-NEXT:    vpalignr {{.*}} # xmm3 = xmm3[14,15,0,1,2,3,4,5,6,7,8,9,10,11,12,13]
+; AVX1-NEXT:    vpblendw {{.*}} # xmm2 = xmm2[0],xmm3[1],xmm2[2],xmm3[3],xmm2[4],xmm3[5],xmm2[6],xmm3[7]
+; AVX1-NEXT:    vpunpckhbw {{.*}} # xmm0 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
+; AVX1-NEXT:    vpunpckhbw {{.*}} # xmm1 = xmm1[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
+; AVX1-NEXT:    vpalignr {{.*}} # xmm1 = xmm1[14,15,0,1,2,3,4,5,6,7,8,9,10,11,12,13]
+; AVX1-NEXT:    vpblendw {{.*}} # xmm0 = xmm0[0],xmm1[1],xmm0[2],xmm1[3],xmm0[4],xmm1[5],xmm0[6],xmm1[7]
+; AVX1-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_08_08_40_40_10_10_42_42_12_12_44_44_14_14_46_46_16_16_48_48_18_18_50_50_20_20_52_52_22_22_54_54
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm2
+; AVX2-NEXT:    vpunpcklbw {{.*}} # xmm2 = xmm2[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm3
+; AVX2-NEXT:    vpunpcklbw {{.*}} # xmm3 = xmm3[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; AVX2-NEXT:    vpalignr {{.*}} # xmm3 = xmm3[14,15,0,1,2,3,4,5,6,7,8,9,10,11,12,13]
+; AVX2-NEXT:    vpblendw {{.*}} # xmm2 = xmm2[0],xmm3[1],xmm2[2],xmm3[3],xmm2[4],xmm3[5],xmm2[6],xmm3[7]
+; AVX2-NEXT:    vpunpckhbw {{.*}} # xmm0 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
+; AVX2-NEXT:    vpunpckhbw {{.*}} # xmm1 = xmm1[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
+; AVX2-NEXT:    vpalignr {{.*}} # xmm1 = xmm1[14,15,0,1,2,3,4,5,6,7,8,9,10,11,12,13]
+; AVX2-NEXT:    vpblendw {{.*}} # xmm0 = xmm0[0],xmm1[1],xmm0[2],xmm1[3],xmm0[4],xmm1[5],xmm0[6],xmm1[7]
+; AVX2-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <i32 8, i32 8, i32 40, i32 40, i32 10, i32 10, i32 42, i32 42, i32 12, i32 12, i32 44, i32 44, i32 14, i32 14, i32 46, i32 46, i32 16, i32 16, i32 48, i32 48, i32 18, i32 18, i32 50, i32 50, i32 20, i32 20, i32 52, i32 52, i32 22, i32 22, i32 54, i32 54>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_01_00_16_17_16_16_16_16_16_16_16_16_16_16_16_16_16_16(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_01_00_16_17_16_16_16_16_16_16_16_16_16_16_16_16_16_16
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0]
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_00_01_00_16_17_16_16_16_16_16_16_16_16_16_16_16_16_16_16
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0]
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm0
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0]
+; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %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 1, i32 0, i32 16, i32 17, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_02_00_00_16_16_18_16_16_16_16_16_16_16_16_16_16_16_16_16(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_02_00_00_16_16_18_16_16_16_16_16_16_16_16_16_16_16_16_16
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,0,0,0,0,0,0,0,0,0,0,0,0,2,0,0]
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,2,0,0,0,0,0,0,0,0,0,0,0,0,0]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_00_00_00_00_00_00_00_00_00_00_00_00_00_02_00_00_16_16_18_16_16_16_16_16_16_16_16_16_16_16_16_16
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,0,0,0,0,0,0,0,0,0,0,0,0,2,0,0]
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm0
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,2,0,0,0,0,0,0,0,0,0,0,0,0,0]
+; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %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 2, i32 0, i32 0, i32 16, i32 16, i32 18, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_00_07_00_00_00_00_00_00_00_16_16_16_16_16_16_16_23_16_16_16_16_16_16_16_16(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_00_00_00_00_00_00_00_00_07_00_00_00_00_00_00_00_16_16_16_16_16_16_16_23_16_16_16_16_16_16_16_16
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,0,0,0,0,0,0,0,7,0,0,0,0,0,0,0]
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,7,0,0,0,0,0,0,0,0]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_00_00_00_00_00_00_00_00_07_00_00_00_00_00_00_00_16_16_16_16_16_16_16_23_16_16_16_16_16_16_16_16
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,0,0,0,0,0,0,0,7,0,0,0,0,0,0,0]
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm0
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,7,0,0,0,0,0,0,0,0]
+; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %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 7, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 23, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_00_00_00_00_00_00_00_08_00_00_00_00_00_00_00_00_16_16_16_16_16_16_16_16_24_16_16_16_16_16_16_16(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_00_00_00_00_00_00_00_08_00_00_00_00_00_00_00_00_16_16_16_16_16_16_16_16_24_16_16_16_16_16_16_16
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,0,0,0,0,0,0,8,0,0,0,0,0,0,0,0]
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0,8,0,0,0,0,0,0,0]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_00_00_00_00_00_00_00_08_00_00_00_00_00_00_00_00_16_16_16_16_16_16_16_16_24_16_16_16_16_16_16_16
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,0,0,0,0,0,0,8,0,0,0,0,0,0,0,0]
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm0
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0,8,0,0,0,0,0,0,0]
+; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %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 8, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 24, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_00_14_00_00_00_00_00_00_00_00_00_00_00_00_00_00_16_16_16_16_16_16_16_16_16_16_16_16_16_16_30_16(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_00_14_00_00_00_00_00_00_00_00_00_00_00_00_00_00_16_16_16_16_16_16_16_16_16_16_16_16_16_16_30_16
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,14,0,0,0,0,0,0,0,0,0,0,0,0,0,0]
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0,0,0,0,0,0,0,14,0]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_00_14_00_00_00_00_00_00_00_00_00_00_00_00_00_00_16_16_16_16_16_16_16_16_16_16_16_16_16_16_30_16
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,14,0,0,0,0,0,0,0,0,0,0,0,0,0,0]
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm0
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0,0,0,0,0,0,0,14,0]
+; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <i32 0, i32 14, 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 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 30, i32 16>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_15_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_16_16_16_16_16_16_16_16_16_16_16_16_16_16_16_31(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_15_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_16_16_16_16_16_16_16_16_16_16_16_16_16_16_16_31
+; AVX1:       # BB#0:
+; AVX1-NEXT:    movl $15, %eax
+; AVX1-NEXT:    vmovd %eax, %xmm1
+; AVX1-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,15]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_15_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_16_16_16_16_16_16_16_16_16_16_16_16_16_16_16_31
+; AVX2:       # BB#0:
+; AVX2-NEXT:    movl $15, %eax
+; AVX2-NEXT:    vmovd %eax, %xmm1
+; AVX2-NEXT:    vpshufb %xmm1, %xmm0, %xmm1
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm0
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,15]
+; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <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, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 31>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_00_00_00_00_04_04_04_04_08_08_08_08_12_12_12_12_28_28_28_28_24_24_24_24_20_20_20_20_16_16_16_16(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_00_00_00_00_04_04_04_04_08_08_08_08_12_12_12_12_28_28_28_28_24_24_24_24_20_20_20_20_16_16_16_16
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,0,0,0,4,4,4,4,8,8,8,8,12,12,12,12]
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[12,12,12,12,8,8,8,8,4,4,4,4,0,0,0,0]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_00_00_00_00_04_04_04_04_08_08_08_08_12_12_12_12_28_28_28_28_24_24_24_24_20_20_20_20_16_16_16_16
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,0,0,0,4,4,4,4,8,8,8,8,12,12,12,12]
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm0
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[12,12,12,12,8,8,8,8,4,4,4,4,0,0,0,0]
+; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %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 4, i32 4, i32 4, i32 4, i32 8, i32 8, i32 8, i32 8, i32 12, i32 12, i32 12, i32 12, i32 28, i32 28, i32 28, i32 28, i32 24, i32 24, i32 24, i32 24, i32 20, i32 20, i32 20, i32 20, i32 16, i32 16, i32 16, i32 16>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_08_08_08_08_08_08_08_08_00_00_00_00_00_00_00_00_16_16_16_16_16_16_16_16_24_24_24_24_24_24_24_24(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_08_08_08_08_08_08_08_08_00_00_00_00_00_00_00_00_16_16_16_16_16_16_16_16_24_24_24_24_24_24_24_24
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[8,8,8,8,8,8,8,8,0,0,0,0,0,0,0,0]
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0,8,8,8,8,8,8,8,8]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_08_08_08_08_08_08_08_08_00_00_00_00_00_00_00_00_16_16_16_16_16_16_16_16_24_24_24_24_24_24_24_24
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[8,8,8,8,8,8,8,8,0,0,0,0,0,0,0,0]
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm0
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,0,0,0,0,8,8,8,8,8,8,8,8]
+; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 24, i32 24, i32 24, i32 24, i32 24, i32 24, i32 24, i32 24>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_00_uu_uu_uu_uu_uu_uu_uu_uu_uu_uu_uu_uu_uu_uu_uu_16_16_16_16_uu_uu_uu_uu_uu_16_16_16_28_28_16_16(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_00_uu_uu_uu_uu_uu_uu_uu_uu_uu_uu_uu_uu_uu_uu_uu_16_16_16_16_uu_uu_uu_uu_uu_16_16_16_28_28_16_16
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT:    vpshufb {{.*}} # xmm1 = xmm1[0,0,0,0,12,12,13,13,0,0,0,0,12,12,0,0]
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_00_uu_uu_uu_uu_uu_uu_uu_uu_uu_uu_uu_uu_uu_uu_uu_16_16_16_16_uu_uu_uu_uu_uu_16_16_16_28_28_16_16
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vpshufb {{.*}} # xmm1 = xmm1[0,0,0,0,12,12,13,13,0,0,0,0,12,12,0,0]
+; 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 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 16, i32 16, i32 16, i32 16, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 16, i32 16, i32 16, i32 28, i32 28, i32 16, i32 16>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_uu_14_uu_uu_00_00_00_00_00_00_00_00_00_00_00_00_16_16_uu_16_uu_uu_uu_uu_16_16_16_16_16_16_30_30(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_uu_14_uu_uu_00_00_00_00_00_00_00_00_00_00_00_00_16_16_uu_16_uu_uu_uu_uu_16_16_16_16_16_16_30_30
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0]
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,1,1,14,14,15,15,0,0,0,0,0,0,14,14]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_uu_14_uu_uu_00_00_00_00_00_00_00_00_00_00_00_00_16_16_uu_16_uu_uu_uu_uu_16_16_16_16_16_16_30_30
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0]
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm0
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,1,1,14,14,15,15,0,0,0,0,0,0,14,14]
+; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <i32 undef, i32 14, i32 undef, i32 undef, 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 16, i32 undef, i32 16, i32 undef, i32 undef, i32 undef, i32 undef, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 30, i32 30>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_00_00_00_uu_uu_uu_04_04_08_08_08_08_uu_uu_12_uu_28_28_28_28_uu_uu_24_24_20_20_20_20_16_16_16_16(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_00_00_00_uu_uu_uu_04_04_08_08_08_08_uu_uu_12_uu_28_28_28_28_uu_uu_24_24_20_20_20_20_16_16_16_16
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,0,0,0,4,4,4,4,8,8,8,8,12,12,12,12]
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[12,12,12,12,8,8,8,8,4,4,4,4,0,0,0,0]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_00_00_00_uu_uu_uu_04_04_08_08_08_08_uu_uu_12_uu_28_28_28_28_uu_uu_24_24_20_20_20_20_16_16_16_16
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpshufb {{.*}} # xmm1 = xmm0[0,0,0,0,4,4,4,4,8,8,8,8,12,12,12,12]
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm0
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[12,12,12,12,8,8,8,8,4,4,4,4,0,0,0,0]
+; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <i32 0, i32 0, i32 0, i32 undef, i32 undef, i32 undef, i32 4, i32 4, i32 8, i32 8, i32 8, i32 8, i32 undef, i32 undef, i32 12, i32 undef, i32 28, i32 28, i32 28, i32 28, i32 undef, i32 undef, i32 24, i32 24, i32 20, i32 20, i32 20, i32 20, i32 16, i32 16, i32 16, i32 16>
+  ret <32 x i8> %shuffle
+}
+
+define <32 x i8> @shuffle_v32i8_08_08_08_08_08_08_08_08_uu_uu_uu_uu_uu_uu_uu_uu_16_16_16_uu_uu_uu_uu_uu_uu_uu_24_24_24_24_24_24(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: @shuffle_v32i8_08_08_08_08_08_08_08_08_uu_uu_uu_uu_uu_uu_uu_uu_16_16_16_uu_uu_uu_uu_uu_uu_uu_24_24_24_24_24_24
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpunpckhbw {{.*}} # xmm1 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
+; AVX1-NEXT:    vpshuflw {{.*}} # xmm1 = xmm1[0,0,0,0,4,5,6,7]
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm0
+; AVX1-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,8,8,9,9,8,8,8,8,8,8,8,8]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: @shuffle_v32i8_08_08_08_08_08_08_08_08_uu_uu_uu_uu_uu_uu_uu_uu_16_16_16_uu_uu_uu_uu_uu_uu_uu_24_24_24_24_24_24
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpunpckhbw {{.*}} # xmm1 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
+; AVX2-NEXT:    vpshuflw {{.*}} # xmm1 = xmm1[0,0,0,0,4,5,6,7]
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm0
+; AVX2-NEXT:    vpshufb {{.*}} # xmm0 = xmm0[0,0,0,0,8,8,9,9,8,8,8,8,8,8,8,8]
+; AVX2-NEXT:    vinserti128 $1, %xmm0, %ymm1, %ymm0
+; AVX2-NEXT:    retq
+  %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 16, i32 16, i32 16, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 24, i32 24, i32 24, i32 24, i32 24, i32 24>
+  ret <32 x i8> %shuffle
+}
diff --git a/test/CodeGen/X86/vector-shuffle-256-v8.ll b/test/CodeGen/X86/vector-shuffle-256-v8.ll
index 1335a53a32a..e2f731b2af5 100644
--- a/test/CodeGen/X86/vector-shuffle-256-v8.ll
+++ b/test/CodeGen/X86/vector-shuffle-256-v8.ll
@@ -577,3 +577,578 @@ define <8 x float> @shuffle_v8f32_uuu3uu66(<8 x float> %a, <8 x float> %b) {
   %shuffle = shufflevector <8 x float> %a, <8 x float> %b, <8 x i32> <i32 undef, i32 undef, i32 undef, i32 3, i32 undef, i32 undef, i32 6, i32 6>
   ret <8 x float> %shuffle
 }
+
+define <8 x i32> @shuffle_v8i32_00000000(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_00000000
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # xmm0 = xmm0[0,0,0,0]
+; ALL-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; ALL-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
+}
+
+define <8 x i32> @shuffle_v8i32_00000010(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_00000010
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # xmm1 = xmm0[0,0,0,0]
+; ALL-NEXT:    vpermilps {{.*}} # xmm0 = xmm0[0,0,1,0]
+; ALL-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; ALL-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 1, i32 0>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_00000200(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_00000200
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # xmm1 = xmm0[0,0,0,0]
+; ALL-NEXT:    vpermilps {{.*}} # xmm0 = xmm0[0,2,0,0]
+; ALL-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; ALL-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 2, i32 0, i32 0>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_00003000(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_00003000
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # xmm1 = xmm0[0,0,0,0]
+; ALL-NEXT:    vpermilps {{.*}} # xmm0 = xmm0[3,0,0,0]
+; ALL-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 0, i32 0, i32 0, i32 0, i32 3, i32 0, i32 0, i32 0>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_00040000(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_00040000
+; ALL:       # BB#0:
+; ALL-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; ALL-NEXT:    vshufps {{.*}} # xmm1 = xmm1[0,0],xmm0[0,0]
+; ALL-NEXT:    vshufps {{.*}} # xmm1 = xmm0[0,0],xmm1[2,0]
+; ALL-NEXT:    vpermilps {{.*}} # xmm0 = xmm0[0,0,0,0]
+; ALL-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 0, i32 0, i32 0, i32 4, i32 0, i32 0, i32 0, i32 0>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_00500000(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_00500000
+; ALL:       # BB#0:
+; ALL-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; ALL-NEXT:    vshufps {{.*}} # xmm1 = xmm1[1,0],xmm0[0,0]
+; ALL-NEXT:    vshufps {{.*}} # xmm1 = xmm0[0,0],xmm1[0,2]
+; ALL-NEXT:    vpermilps {{.*}} # xmm0 = xmm0[0,0,0,0]
+; ALL-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 0, i32 0, i32 5, i32 0, i32 0, i32 0, i32 0, i32 0>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_06000000(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_06000000
+; ALL:       # BB#0:
+; ALL-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; ALL-NEXT:    vshufps {{.*}} # xmm1 = xmm1[2,0],xmm0[0,0]
+; ALL-NEXT:    vshufps {{.*}} # xmm1 = xmm1[2,0],xmm0[0,0]
+; ALL-NEXT:    vpermilps {{.*}} # xmm0 = xmm0[0,0,0,0]
+; ALL-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 0, i32 6, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_70000000(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_70000000
+; ALL:       # BB#0:
+; ALL-NEXT:    vextractf128 $1, %ymm0, %xmm1
+; ALL-NEXT:    vshufps {{.*}} # xmm1 = xmm1[3,0],xmm0[0,0]
+; ALL-NEXT:    vshufps {{.*}} # xmm1 = xmm1[0,2],xmm0[0,0]
+; ALL-NEXT:    vpermilps {{.*}} # xmm0 = xmm0[0,0,0,0]
+; ALL-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 7, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_01014545(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_01014545
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilpd {{.*}} # ymm0 = ymm0[0,0,2,2]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 0, i32 1, i32 0, i32 1, i32 4, i32 5, i32 4, i32 5>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_00112233(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_00112233
+; ALL:       # BB#0:
+; ALL-NEXT:    vunpcklps {{.*}} # xmm1 = xmm0[0,0,1,1]
+; ALL-NEXT:    vunpckhps {{.*}} # xmm0 = xmm0[2,2,3,3]
+; ALL-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 0, i32 0, i32 1, i32 1, i32 2, i32 2, i32 3, i32 3>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_00001111(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_00001111
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # xmm1 = xmm0[0,0,0,0]
+; ALL-NEXT:    vpermilps {{.*}} # xmm0 = xmm0[1,1,1,1]
+; ALL-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 0, i32 0, i32 0, i32 0, i32 1, i32 1, i32 1, i32 1>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_81a3c5e7(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_81a3c5e7
+; ALL:       # BB#0:
+; ALL-NEXT:    vblendps {{.*}} # ymm0 = ymm1[0],ymm0[1],ymm1[2],ymm0[3],ymm1[4],ymm0[5],ymm1[6],ymm0[7]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 8, i32 1, i32 10, i32 3, i32 12, i32 5, i32 14, i32 7>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_08080808(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_08080808
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # xmm1 = xmm1[0,0,2,0]
+; ALL-NEXT:    vpermilps {{.*}} # xmm0 = xmm0[0,1,0,3]
+; ALL-NEXT:    vblendps {{.*}} # xmm0 = xmm0[0],xmm1[1],xmm0[2],xmm1[3]
+; ALL-NEXT:    vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; ALL-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>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_08084c4c(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_08084c4c
+; ALL:       # BB#0:
+; ALL-NEXT:    vshufps {{.*}} # ymm0 = ymm0[0,0],ymm1[0,0],ymm0[4,4],ymm1[4,4]
+; ALL-NEXT:    vshufps {{.*}} # ymm0 = ymm0[0,2,1,3,4,6,5,7]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 0, i32 8, i32 0, i32 8, i32 4, i32 12, i32 4, i32 12>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_8823cc67(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_8823cc67
+; ALL:       # BB#0:
+; ALL-NEXT:    vshufps {{.*}} # ymm0 = ymm1[0,0],ymm0[2,3],ymm1[4,4],ymm0[6,7]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 8, i32 8, i32 2, i32 3, i32 12, i32 12, i32 6, i32 7>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_9832dc76(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_9832dc76
+; ALL:       # BB#0:
+; ALL-NEXT:    vshufps {{.*}} # ymm0 = ymm1[1,0],ymm0[3,2],ymm1[5,4],ymm0[7,6]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 9, i32 8, i32 3, i32 2, i32 13, i32 12, i32 7, i32 6>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_9810dc54(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_9810dc54
+; ALL:       # BB#0:
+; ALL-NEXT:    vshufps {{.*}} # ymm0 = ymm1[1,0],ymm0[1,0],ymm1[5,4],ymm0[5,4]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 9, i32 8, i32 1, i32 0, i32 13, i32 12, i32 5, i32 4>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_08194c5d(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_08194c5d
+; ALL:       # BB#0:
+; ALL-NEXT:    vunpcklps {{.*}} # ymm0 = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[4],ymm1[4],ymm0[5],ymm1[5]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 4, i32 12, i32 5, i32 13>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_2a3b6e7f(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_2a3b6e7f
+; ALL:       # BB#0:
+; ALL-NEXT:    vunpckhps {{.*}} # ymm0 = ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[6],ymm1[6],ymm0[7],ymm1[7]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 2, i32 10, i32 3, i32 11, i32 6, i32 14, i32 7, i32 15>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_08192a3b(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_08192a3b
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # xmm2 = xmm1[0,2,2,3]
+; ALL-NEXT:    vpermilps {{.*}} # xmm3 = xmm0[2,1,3,3]
+; ALL-NEXT:    vblendps {{.*}} # xmm2 = xmm3[0],xmm2[1],xmm3[2],xmm2[3]
+; ALL-NEXT:    vpermilps {{.*}} # xmm1 = xmm1[0,0,2,1]
+; ALL-NEXT:    vpermilps {{.*}} # xmm0 = xmm0[0,1,1,3]
+; ALL-NEXT:    vblendps {{.*}} # xmm0 = xmm0[0],xmm1[1],xmm0[2],xmm1[3]
+; ALL-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 2, i32 10, i32 3, i32 11>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_08991abb(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_08991abb
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # xmm2 = xmm0[1,1,2,3]
+; ALL-NEXT:    vpermilps {{.*}} # xmm3 = xmm1[0,2,3,3]
+; ALL-NEXT:    vblendps {{.*}} # xmm2 = xmm2[0],xmm3[1,2,3]
+; ALL-NEXT:    vunpcklps {{.*}} # xmm1 = xmm1[0,0,1,1]
+; ALL-NEXT:    vblendps {{.*}} # xmm0 = xmm0[0],xmm1[1,2,3]
+; ALL-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 0, i32 8, i32 9, i32 9, i32 1, i32 10, i32 11, i32 11>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_091b2d3f(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_091b2d3f
+; ALL:       # BB#0:
+; ALL-NEXT:    vextractf128 $1, %ymm1, %xmm2
+; ALL-NEXT:    vpermilps {{.*}} # xmm3 = xmm0[2,1,3,3]
+; ALL-NEXT:    vblendps {{.*}} # xmm2 = xmm3[0],xmm2[1],xmm3[2],xmm2[3]
+; ALL-NEXT:    vpermilps {{.*}} # xmm0 = xmm0[0,1,1,3]
+; ALL-NEXT:    vblendps {{.*}} # xmm0 = xmm0[0],xmm1[1],xmm0[2],xmm1[3]
+; ALL-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 0, i32 9, i32 1, i32 11, i32 2, i32 13, i32 3, i32 15>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_09ab1def(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_09ab1def
+; ALL:       # BB#0:
+; ALL-NEXT:    vextractf128 $1, %ymm1, %xmm2
+; ALL-NEXT:    vpermilps {{.*}} # xmm3 = xmm0[1,1,2,3]
+; ALL-NEXT:    vblendps {{.*}} # xmm2 = xmm3[0],xmm2[1,2,3]
+; ALL-NEXT:    vblendps {{.*}} # xmm0 = xmm0[0],xmm1[1,2,3]
+; ALL-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 0, i32 9, i32 10, i32 11, i32 1, i32 13, i32 14, i32 15>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_00014445(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_00014445
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # ymm0 = ymm0[0,0,0,1,4,4,4,5]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 0, i32 0, i32 0, i32 1, i32 4, i32 4, i32 4, i32 5>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_00204464(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_00204464
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # ymm0 = ymm0[0,0,2,0,4,4,6,4]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 0, i32 0, i32 2, i32 0, i32 4, i32 4, i32 6, i32 4>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_03004744(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_03004744
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # ymm0 = ymm0[0,3,0,0,4,7,4,4]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 0, i32 3, i32 0, i32 0, i32 4, i32 7, i32 4, i32 4>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_10005444(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_10005444
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # ymm0 = ymm0[1,0,0,0,5,4,4,4]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 1, i32 0, i32 0, i32 0, i32 5, i32 4, i32 4, i32 4>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_22006644(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_22006644
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # ymm0 = ymm0[2,2,0,0,6,6,4,4]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 2, i32 2, i32 0, i32 0, i32 6, i32 6, i32 4, i32 4>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_33307774(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_33307774
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # ymm0 = ymm0[3,3,3,0,7,7,7,4]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 3, i32 3, i32 3, i32 0, i32 7, i32 7, i32 7, i32 4>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_32107654(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_32107654
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # ymm0 = ymm0[3,2,1,0,7,6,5,4]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 3, i32 2, i32 1, i32 0, i32 7, i32 6, i32 5, i32 4>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_00234467(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_00234467
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # ymm0 = ymm0[0,0,2,3,4,4,6,7]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 0, i32 0, i32 2, i32 3, i32 4, i32 4, i32 6, i32 7>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_00224466(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_00224466
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # ymm0 = ymm0[0,0,2,2,4,4,6,6]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 0, i32 0, i32 2, i32 2, i32 4, i32 4, i32 6, i32 6>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_10325476(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_10325476
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # ymm0 = ymm0[1,0,3,2,5,4,7,6]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 1, i32 0, i32 3, i32 2, i32 5, i32 4, i32 7, i32 6>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_11335577(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_11335577
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # ymm0 = ymm0[1,1,3,3,5,5,7,7]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 1, i32 1, i32 3, i32 3, i32 5, i32 5, i32 7, i32 7>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_10235467(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_10235467
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # ymm0 = ymm0[1,0,2,3,5,4,6,7]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 1, i32 0, i32 2, i32 3, i32 5, i32 4, i32 6, i32 7>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_10225466(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_10225466
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # ymm0 = ymm0[1,0,2,2,5,4,6,6]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 1, i32 0, i32 2, i32 2, i32 5, i32 4, i32 6, i32 6>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_00015444(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_00015444
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # ymm1 = ymm0[1,0,0,0,5,4,4,4]
+; ALL-NEXT:    vpermilps {{.*}} # ymm0 = ymm0[0,0,0,1,4,4,4,5]
+; ALL-NEXT:    vblendps {{.*}} # ymm0 = ymm0[0,1,2,3],ymm1[4,5,6,7]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 0, i32 0, i32 0, i32 1, i32 5, i32 4, i32 4, i32 4>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_00204644(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_00204644
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # ymm1 = ymm0[0,2,0,0,4,6,4,4]
+; ALL-NEXT:    vpermilps {{.*}} # ymm0 = ymm0[0,0,2,0,4,4,6,4]
+; ALL-NEXT:    vblendps {{.*}} # ymm0 = ymm0[0,1,2,3],ymm1[4,5,6,7]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 0, i32 0, i32 2, i32 0, i32 4, i32 6, i32 4, i32 4>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_03004474(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_03004474
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # ymm1 = ymm0[0,0,3,0,4,4,7,4]
+; ALL-NEXT:    vpermilps {{.*}} # ymm0 = ymm0[0,3,0,0,4,7,4,4]
+; ALL-NEXT:    vblendps {{.*}} # ymm0 = ymm0[0,1,2,3],ymm1[4,5,6,7]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 0, i32 3, i32 0, i32 0, i32 4, i32 4, i32 7, i32 4>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_10004444(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_10004444
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # ymm1 = ymm0[0,0,0,0,4,4,4,4]
+; ALL-NEXT:    vpermilps {{.*}} # ymm0 = ymm0[1,0,0,0,5,4,4,4]
+; ALL-NEXT:    vblendps {{.*}} # ymm0 = ymm0[0,1,2,3],ymm1[4,5,6,7]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 1, i32 0, i32 0, i32 0, i32 4, i32 4, i32 4, i32 4>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_22006446(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_22006446
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # ymm1 = ymm0[2,0,0,2,6,4,4,6]
+; ALL-NEXT:    vpermilps {{.*}} # ymm0 = ymm0[2,2,0,0,6,6,4,4]
+; ALL-NEXT:    vblendps {{.*}} # ymm0 = ymm0[0,1,2,3],ymm1[4,5,6,7]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 2, i32 2, i32 0, i32 0, i32 6, i32 4, i32 4, i32 6>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_33307474(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_33307474
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # ymm1 = ymm0[3,0,3,0,7,4,7,4]
+; ALL-NEXT:    vpermilps {{.*}} # ymm0 = ymm0[3,3,3,0,7,7,7,4]
+; ALL-NEXT:    vblendps {{.*}} # ymm0 = ymm0[0,1,2,3],ymm1[4,5,6,7]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 3, i32 3, i32 3, i32 0, i32 7, i32 4, i32 7, i32 4>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_32104567(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_32104567
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # ymm1 = ymm0[3,2,1,0,7,6,5,4]
+; ALL-NEXT:    vblendps {{.*}} # ymm0 = ymm1[0,1,2,3],ymm0[4,5,6,7]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 3, i32 2, i32 1, i32 0, i32 4, i32 5, i32 6, i32 7>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_00236744(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_00236744
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # ymm1 = ymm0[2,3,0,0,6,7,4,4]
+; ALL-NEXT:    vpermilps {{.*}} # ymm0 = ymm0[0,0,2,3,4,4,6,7]
+; ALL-NEXT:    vblendps {{.*}} # ymm0 = ymm0[0,1,2,3],ymm1[4,5,6,7]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 0, i32 0, i32 2, i32 3, i32 6, i32 7, i32 4, i32 4>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_00226644(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_00226644
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # ymm1 = ymm0[2,2,0,0,6,6,4,4]
+; ALL-NEXT:    vpermilps {{.*}} # ymm0 = ymm0[0,0,2,2,4,4,6,6]
+; ALL-NEXT:    vblendps {{.*}} # ymm0 = ymm0[0,1,2,3],ymm1[4,5,6,7]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 0, i32 0, i32 2, i32 2, i32 6, i32 6, i32 4, i32 4>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_10324567(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_10324567
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # ymm1 = ymm0[1,0,3,2,5,4,7,6]
+; ALL-NEXT:    vblendps {{.*}} # ymm0 = ymm1[0,1,2,3],ymm0[4,5,6,7]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 1, i32 0, i32 3, i32 2, i32 4, i32 5, i32 6, i32 7>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_11334567(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_11334567
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # ymm1 = ymm0[1,1,3,3,5,5,7,7]
+; ALL-NEXT:    vblendps {{.*}} # ymm0 = ymm1[0,1,2,3],ymm0[4,5,6,7]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 1, i32 1, i32 3, i32 3, i32 4, i32 5, i32 6, i32 7>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_01235467(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_01235467
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # ymm1 = ymm0[1,0,2,3,5,4,6,7]
+; ALL-NEXT:    vblendps {{.*}} # ymm0 = ymm0[0,1,2,3],ymm1[4,5,6,7]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 5, i32 4, i32 6, i32 7>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_01235466(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_01235466
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # ymm1 = ymm0[1,0,2,2,5,4,6,6]
+; ALL-NEXT:    vblendps {{.*}} # ymm0 = ymm0[0,1,2,3],ymm1[4,5,6,7]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 5, i32 4, i32 6, i32 6>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_002u6u44(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_002u6u44
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # ymm1 = ymm0[2,1,0,0,6,5,4,4]
+; ALL-NEXT:    vpermilps {{.*}} # ymm0 = ymm0[0,0,2,3,4,4,6,7]
+; ALL-NEXT:    vblendps {{.*}} # ymm0 = ymm0[0,1,2,3],ymm1[4,5,6,7]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 0, i32 0, i32 2, i32 undef, i32 6, i32 undef, i32 4, i32 4>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_00uu66uu(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_00uu66uu
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # ymm1 = ymm0[2,2,2,3,6,6,6,7]
+; ALL-NEXT:    vpermilps {{.*}} # ymm0 = ymm0[0,0,2,3,4,4,6,7]
+; ALL-NEXT:    vblendps {{.*}} # ymm0 = ymm0[0,1,2,3],ymm1[4,5,6,7]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 0, i32 0, i32 undef, i32 undef, i32 6, i32 6, i32 undef, i32 undef>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_103245uu(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_103245uu
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # ymm1 = ymm0[1,0,3,2,5,4,7,6]
+; ALL-NEXT:    vblendps {{.*}} # ymm0 = ymm1[0,1,2,3],ymm0[4,5,6,7]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 1, i32 0, i32 3, i32 2, i32 4, i32 5, i32 undef, i32 undef>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_1133uu67(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_1133uu67
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # ymm1 = ymm0[1,1,3,3,5,5,7,7]
+; ALL-NEXT:    vblendps {{.*}} # ymm0 = ymm1[0,1,2,3],ymm0[4,5,6,7]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 1, i32 1, i32 3, i32 3, i32 undef, i32 undef, i32 6, i32 7>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_0uu354uu(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_0uu354uu
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # ymm1 = ymm0[1,0,2,3,5,4,6,7]
+; ALL-NEXT:    vblendps {{.*}} # ymm0 = ymm0[0,1,2,3],ymm1[4,5,6,7]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 0, i32 undef, i32 undef, i32 3, i32 5, i32 4, i32 undef, i32 undef>
+  ret <8 x i32> %shuffle
+}
+
+define <8 x i32> @shuffle_v8i32_uuu3uu66(<8 x i32> %a, <8 x i32> %b) {
+; ALL-LABEL: @shuffle_v8i32_uuu3uu66
+; ALL:       # BB#0:
+; ALL-NEXT:    vpermilps {{.*}} # ymm1 = ymm0[0,1,2,2,4,5,6,6]
+; ALL-NEXT:    vblendps {{.*}} # ymm0 = ymm0[0,1,2,3],ymm1[4,5,6,7]
+; ALL-NEXT:    retq
+  %shuffle = shufflevector <8 x i32> %a, <8 x i32> %b, <8 x i32> <i32 undef, i32 undef, i32 undef, i32 3, i32 undef, i32 undef, i32 6, i32 6>
+  ret <8 x i32> %shuffle
+}