From 3916c2642dcb23dff419ab1f8d998cfcf411b19b Mon Sep 17 00:00:00 2001
From: Chandler Carruth <chandlerc@gmail.com>
Date: Wed, 1 Oct 2014 20:41:36 +0000
Subject: [PATCH] [x86] Teach both sext and zext vector tests to cover a nice
 wide range of architectures: SSE2, SSSE3, SSE4.1, AVX, and AVX2.

Unfortunately, this exposses the absolute horror of the code we generate
for many of these patterns. Anyone wanting to familiarize themselves
with the x86 backend and improve performance could do a lot of good
sitting down and making these test cases not look so terrible. While the
new vector shuffle code I'm working on well help some, it won't fix all
of the crimes here.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@218807 91177308-0d34-0410-b5e6-96231b3b80d8
---
 test/CodeGen/X86/vector-sext.ll | 636 ++++++++++++++++++++++++--------
 test/CodeGen/X86/vector-zext.ll | 210 +++++++++--
 2 files changed, 662 insertions(+), 184 deletions(-)

diff --git a/test/CodeGen/X86/vector-sext.ll b/test/CodeGen/X86/vector-sext.ll
index 663c808a8b3..ead1b953dc5 100644
--- a/test/CodeGen/X86/vector-sext.ll
+++ b/test/CodeGen/X86/vector-sext.ll
@@ -1,20 +1,44 @@
-; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=pentium4 | FileCheck %s --check-prefix=SSE --check-prefix=SSE2
-; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=core2 | FileCheck %s --check-prefix=SSE --check-prefix=SSSE3
-; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=corei7-avx | FileCheck %s --check-prefix=AVX --check-prefix=AVX1
-; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=core-avx2 | FileCheck %s --check-prefix=AVX --check-prefix=AVX2
+; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=x86-64 -mattr=+sse2 | FileCheck %s --check-prefix=SSE --check-prefix=SSE2
+; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=x86-64 -mattr=+ssse3 | FileCheck %s --check-prefix=SSE --check-prefix=SSSE3
+; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=x86-64 -mattr=+sse4.1 | FileCheck %s --check-prefix=SSE --check-prefix=SSE41
+; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=x86-64 -mattr=+avx | FileCheck %s --check-prefix=AVX --check-prefix=AVX1
+; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=x86-64 -mattr=+avx2 | FileCheck %s --check-prefix=AVX --check-prefix=AVX2
 
 define <8 x i32> @sext_8i16_to_8i32(<8 x i16> %A) nounwind uwtable readnone ssp {
-; SSE-LABEL: sext_8i16_to_8i32:
-; SSE:       ## BB#0:
-; SSE-NEXT:    movdqa %xmm0, %xmm1
-; SSE-NEXT:      ## kill: XMM0<def> XMM1<kill>
-; SSE-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3]
-; SSE-NEXT:    pslld $16, %xmm0
-; SSE-NEXT:    psrad $16, %xmm0
-; SSE-NEXT:    punpckhwd {{.*#+}} xmm1 = xmm1[4],xmm0[4],xmm1[5],xmm0[5],xmm1[6],xmm0[6],xmm1[7],xmm0[7]
-; SSE-NEXT:    pslld $16, %xmm1
-; SSE-NEXT:    psrad $16, %xmm1
-; SSE-NEXT:    retq
+; SSE2-LABEL: sext_8i16_to_8i32:
+; SSE2:       ## BB#0:
+; SSE2-NEXT:    movdqa %xmm0, %xmm1
+; SSE2-NEXT:      ## kill: XMM0<def> XMM1<kill>
+; SSE2-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3]
+; SSE2-NEXT:    pslld $16, %xmm0
+; SSE2-NEXT:    psrad $16, %xmm0
+; SSE2-NEXT:    punpckhwd {{.*#+}} xmm1 = xmm1[4],xmm0[4],xmm1[5],xmm0[5],xmm1[6],xmm0[6],xmm1[7],xmm0[7]
+; SSE2-NEXT:    pslld $16, %xmm1
+; SSE2-NEXT:    psrad $16, %xmm1
+; SSE2-NEXT:    retq
+;
+; SSSE3-LABEL: sext_8i16_to_8i32:
+; SSSE3:       ## BB#0:
+; SSSE3-NEXT:    movdqa %xmm0, %xmm1
+; SSSE3-NEXT:      ## kill: XMM0<def> XMM1<kill>
+; SSSE3-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3]
+; SSSE3-NEXT:    pslld $16, %xmm0
+; SSSE3-NEXT:    psrad $16, %xmm0
+; SSSE3-NEXT:    punpckhwd {{.*#+}} xmm1 = xmm1[4],xmm0[4],xmm1[5],xmm0[5],xmm1[6],xmm0[6],xmm1[7],xmm0[7]
+; SSSE3-NEXT:    pslld $16, %xmm1
+; SSSE3-NEXT:    psrad $16, %xmm1
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: sext_8i16_to_8i32:
+; SSE41:       ## BB#0:
+; SSE41-NEXT:    movdqa %xmm0, %xmm1
+; SSE41-NEXT:    pmovzxwd %xmm1, %xmm0
+; SSE41-NEXT:    punpckhwd {{.*#+}} xmm1 = xmm1[4],xmm0[4],xmm1[5],xmm0[5],xmm1[6],xmm0[6],xmm1[7],xmm0[7]
+; SSE41-NEXT:    pslld $16, %xmm1
+; SSE41-NEXT:    psrad $16, %xmm1
+; SSE41-NEXT:    pslld $16, %xmm0
+; SSE41-NEXT:    psrad $16, %xmm0
+; SSE41-NEXT:    retq
 ;
 ; AVX1-LABEL: sext_8i16_to_8i32:
 ; AVX1:       ## BB#0:
@@ -34,28 +58,72 @@ define <8 x i32> @sext_8i16_to_8i32(<8 x i16> %A) nounwind uwtable readnone ssp
 }
 
 define <4 x i64> @sext_4i32_to_4i64(<4 x i32> %A) nounwind uwtable readnone ssp {
-; SSE-LABEL: sext_4i32_to_4i64:
-; SSE:       ## BB#0:
-; SSE-NEXT:    pshufd {{.*#+}} xmm1 = xmm0[0,0,1,0]
-; SSE-NEXT:    movd %xmm1, %rax
-; SSE-NEXT:    cltq
-; SSE-NEXT:    movd %rax, %xmm2
-; SSE-NEXT:    punpckhqdq {{.*#+}} xmm1 = xmm1[1,1]
-; SSE-NEXT:    movd %xmm1, %rax
-; SSE-NEXT:    cltq
-; SSE-NEXT:    movd %rax, %xmm1
-; SSE-NEXT:    punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm1[0]
-; SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,0,3,0]
-; SSE-NEXT:    movd %xmm0, %rax
-; SSE-NEXT:    cltq
-; SSE-NEXT:    movd %rax, %xmm1
-; SSE-NEXT:    punpckhqdq {{.*#+}} xmm0 = xmm0[1,1]
-; SSE-NEXT:    movd %xmm0, %rax
-; SSE-NEXT:    cltq
-; SSE-NEXT:    movd %rax, %xmm0
-; SSE-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
-; SSE-NEXT:    movdqa %xmm2, %xmm0
-; SSE-NEXT:    retq
+; SSE2-LABEL: sext_4i32_to_4i64:
+; SSE2:       ## BB#0:
+; SSE2-NEXT:    pshufd {{.*#+}} xmm1 = xmm0[0,0,1,0]
+; SSE2-NEXT:    movd %xmm1, %rax
+; SSE2-NEXT:    cltq
+; SSE2-NEXT:    movd %rax, %xmm2
+; SSE2-NEXT:    punpckhqdq {{.*#+}} xmm1 = xmm1[1,1]
+; SSE2-NEXT:    movd %xmm1, %rax
+; SSE2-NEXT:    cltq
+; SSE2-NEXT:    movd %rax, %xmm1
+; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm1[0]
+; SSE2-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,0,3,0]
+; SSE2-NEXT:    movd %xmm0, %rax
+; SSE2-NEXT:    cltq
+; SSE2-NEXT:    movd %rax, %xmm1
+; SSE2-NEXT:    punpckhqdq {{.*#+}} xmm0 = xmm0[1,1]
+; SSE2-NEXT:    movd %xmm0, %rax
+; SSE2-NEXT:    cltq
+; SSE2-NEXT:    movd %rax, %xmm0
+; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
+; SSE2-NEXT:    movdqa %xmm2, %xmm0
+; SSE2-NEXT:    retq
+;
+; SSSE3-LABEL: sext_4i32_to_4i64:
+; SSSE3:       ## BB#0:
+; SSSE3-NEXT:    pshufd {{.*#+}} xmm1 = xmm0[0,0,1,0]
+; SSSE3-NEXT:    movd %xmm1, %rax
+; SSSE3-NEXT:    cltq
+; SSSE3-NEXT:    movd %rax, %xmm2
+; SSSE3-NEXT:    punpckhqdq {{.*#+}} xmm1 = xmm1[1,1]
+; SSSE3-NEXT:    movd %xmm1, %rax
+; SSSE3-NEXT:    cltq
+; SSSE3-NEXT:    movd %rax, %xmm1
+; SSSE3-NEXT:    punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm1[0]
+; SSSE3-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,0,3,0]
+; SSSE3-NEXT:    movd %xmm0, %rax
+; SSSE3-NEXT:    cltq
+; SSSE3-NEXT:    movd %rax, %xmm1
+; SSSE3-NEXT:    punpckhqdq {{.*#+}} xmm0 = xmm0[1,1]
+; SSSE3-NEXT:    movd %xmm0, %rax
+; SSSE3-NEXT:    cltq
+; SSSE3-NEXT:    movd %rax, %xmm0
+; SSSE3-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
+; SSSE3-NEXT:    movdqa %xmm2, %xmm0
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: sext_4i32_to_4i64:
+; SSE41:       ## BB#0:
+; SSE41-NEXT:    pmovzxdq %xmm0, %xmm1
+; SSE41-NEXT:    pextrq $1, %xmm1, %rax
+; SSE41-NEXT:    cltq
+; SSE41-NEXT:    movd %rax, %xmm3
+; SSE41-NEXT:    movd %xmm1, %rax
+; SSE41-NEXT:    cltq
+; SSE41-NEXT:    movd %rax, %xmm2
+; SSE41-NEXT:    punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm3[0]
+; SSE41-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,0,3,0]
+; SSE41-NEXT:    pextrq $1, %xmm0, %rax
+; SSE41-NEXT:    cltq
+; SSE41-NEXT:    movd %rax, %xmm3
+; SSE41-NEXT:    movd %xmm0, %rax
+; SSE41-NEXT:    cltq
+; SSE41-NEXT:    movd %rax, %xmm1
+; SSE41-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm3[0]
+; SSE41-NEXT:    movdqa %xmm2, %xmm0
+; SSE41-NEXT:    retq
 ;
 ; AVX1-LABEL: sext_4i32_to_4i64:
 ; AVX1:       ## BB#0:
@@ -75,12 +143,24 @@ define <4 x i64> @sext_4i32_to_4i64(<4 x i32> %A) nounwind uwtable readnone ssp
 }
 
 define <4 x i32> @load_sext_test1(<4 x i16> *%ptr) {
-; SSE-LABEL: load_sext_test1:
-; SSE:       ## BB#0:
-; SSE-NEXT:    movq (%rdi), %xmm0
-; SSE-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3]
-; SSE-NEXT:    psrad $16, %xmm0
-; SSE-NEXT:    retq
+; SSE2-LABEL: load_sext_test1:
+; SSE2:       ## BB#0:
+; SSE2-NEXT:    movq (%rdi), %xmm0
+; SSE2-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3]
+; SSE2-NEXT:    psrad $16, %xmm0
+; SSE2-NEXT:    retq
+;
+; SSSE3-LABEL: load_sext_test1:
+; SSSE3:       ## BB#0:
+; SSSE3-NEXT:    movq (%rdi), %xmm0
+; SSSE3-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3]
+; SSSE3-NEXT:    psrad $16, %xmm0
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: load_sext_test1:
+; SSE41:       ## BB#0:
+; SSE41-NEXT:    pmovsxwd (%rdi), %xmm0
+; SSE41-NEXT:    retq
 ;
 ; AVX-LABEL: load_sext_test1:
 ; AVX:       ## BB#0:
@@ -116,6 +196,11 @@ define <4 x i32> @load_sext_test2(<4 x i8> *%ptr) {
 ; SSSE3-NEXT:    psrad $24, %xmm0
 ; SSSE3-NEXT:    retq
 ;
+; SSE41-LABEL: load_sext_test2:
+; SSE41:       ## BB#0:
+; SSE41-NEXT:    pmovsxbd (%rdi), %xmm0
+; SSE41-NEXT:    retq
+;
 ; AVX-LABEL: load_sext_test2:
 ; AVX:       ## BB#0:
 ; AVX-NEXT:    vpmovsxbd (%rdi), %xmm0
@@ -126,14 +211,28 @@ define <4 x i32> @load_sext_test2(<4 x i8> *%ptr) {
 }
 
 define <2 x i64> @load_sext_test3(<2 x i8> *%ptr) {
-; SSE-LABEL: load_sext_test3:
-; SSE:       ## BB#0:
-; SSE-NEXT:    movsbq 1(%rdi), %rax
-; SSE-NEXT:    movd %rax, %xmm1
-; SSE-NEXT:    movsbq (%rdi), %rax
-; SSE-NEXT:    movd %rax, %xmm0
-; SSE-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
-; SSE-NEXT:    retq
+; SSE2-LABEL: load_sext_test3:
+; SSE2:       ## BB#0:
+; SSE2-NEXT:    movsbq 1(%rdi), %rax
+; SSE2-NEXT:    movd %rax, %xmm1
+; SSE2-NEXT:    movsbq (%rdi), %rax
+; SSE2-NEXT:    movd %rax, %xmm0
+; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; SSE2-NEXT:    retq
+;
+; SSSE3-LABEL: load_sext_test3:
+; SSSE3:       ## BB#0:
+; SSSE3-NEXT:    movsbq 1(%rdi), %rax
+; SSSE3-NEXT:    movd %rax, %xmm1
+; SSSE3-NEXT:    movsbq (%rdi), %rax
+; SSSE3-NEXT:    movd %rax, %xmm0
+; SSSE3-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: load_sext_test3:
+; SSE41:       ## BB#0:
+; SSE41-NEXT:    pmovsxbq (%rdi), %xmm0
+; SSE41-NEXT:    retq
 ;
 ; AVX-LABEL: load_sext_test3:
 ; AVX:       ## BB#0:
@@ -145,14 +244,28 @@ define <2 x i64> @load_sext_test3(<2 x i8> *%ptr) {
 }
 
 define <2 x i64> @load_sext_test4(<2 x i16> *%ptr) {
-; SSE-LABEL: load_sext_test4:
-; SSE:       ## BB#0:
-; SSE-NEXT:    movswq 2(%rdi), %rax
-; SSE-NEXT:    movd %rax, %xmm1
-; SSE-NEXT:    movswq (%rdi), %rax
-; SSE-NEXT:    movd %rax, %xmm0
-; SSE-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
-; SSE-NEXT:    retq
+; SSE2-LABEL: load_sext_test4:
+; SSE2:       ## BB#0:
+; SSE2-NEXT:    movswq 2(%rdi), %rax
+; SSE2-NEXT:    movd %rax, %xmm1
+; SSE2-NEXT:    movswq (%rdi), %rax
+; SSE2-NEXT:    movd %rax, %xmm0
+; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; SSE2-NEXT:    retq
+;
+; SSSE3-LABEL: load_sext_test4:
+; SSSE3:       ## BB#0:
+; SSSE3-NEXT:    movswq 2(%rdi), %rax
+; SSSE3-NEXT:    movd %rax, %xmm1
+; SSSE3-NEXT:    movswq (%rdi), %rax
+; SSSE3-NEXT:    movd %rax, %xmm0
+; SSSE3-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: load_sext_test4:
+; SSE41:       ## BB#0:
+; SSE41-NEXT:    pmovsxwq (%rdi), %xmm0
+; SSE41-NEXT:    retq
 ;
 ; AVX-LABEL: load_sext_test4:
 ; AVX:       ## BB#0:
@@ -164,14 +277,28 @@ define <2 x i64> @load_sext_test4(<2 x i16> *%ptr) {
 }
 
 define <2 x i64> @load_sext_test5(<2 x i32> *%ptr) {
-; SSE-LABEL: load_sext_test5:
-; SSE:       ## BB#0:
-; SSE-NEXT:    movslq 4(%rdi), %rax
-; SSE-NEXT:    movd %rax, %xmm1
-; SSE-NEXT:    movslq (%rdi), %rax
-; SSE-NEXT:    movd %rax, %xmm0
-; SSE-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
-; SSE-NEXT:    retq
+; SSE2-LABEL: load_sext_test5:
+; SSE2:       ## BB#0:
+; SSE2-NEXT:    movslq 4(%rdi), %rax
+; SSE2-NEXT:    movd %rax, %xmm1
+; SSE2-NEXT:    movslq (%rdi), %rax
+; SSE2-NEXT:    movd %rax, %xmm0
+; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; SSE2-NEXT:    retq
+;
+; SSSE3-LABEL: load_sext_test5:
+; SSSE3:       ## BB#0:
+; SSSE3-NEXT:    movslq 4(%rdi), %rax
+; SSSE3-NEXT:    movd %rax, %xmm1
+; SSSE3-NEXT:    movslq (%rdi), %rax
+; SSSE3-NEXT:    movd %rax, %xmm0
+; SSSE3-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: load_sext_test5:
+; SSE41:       ## BB#0:
+; SSE41-NEXT:    pmovsxdq (%rdi), %xmm0
+; SSE41-NEXT:    retq
 ;
 ; AVX-LABEL: load_sext_test5:
 ; AVX:       ## BB#0:
@@ -183,12 +310,24 @@ define <2 x i64> @load_sext_test5(<2 x i32> *%ptr) {
 }
 
 define <8 x i16> @load_sext_test6(<8 x i8> *%ptr) {
-; SSE-LABEL: load_sext_test6:
-; SSE:       ## BB#0:
-; SSE-NEXT:    movq (%rdi), %xmm0
-; SSE-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
-; SSE-NEXT:    psraw $8, %xmm0
-; SSE-NEXT:    retq
+; SSE2-LABEL: load_sext_test6:
+; SSE2:       ## BB#0:
+; SSE2-NEXT:    movq (%rdi), %xmm0
+; SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; SSE2-NEXT:    psraw $8, %xmm0
+; SSE2-NEXT:    retq
+;
+; SSSE3-LABEL: load_sext_test6:
+; SSSE3:       ## BB#0:
+; SSSE3-NEXT:    movq (%rdi), %xmm0
+; SSSE3-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; SSSE3-NEXT:    psraw $8, %xmm0
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: load_sext_test6:
+; SSE41:       ## BB#0:
+; SSE41-NEXT:    pmovsxbw (%rdi), %xmm0
+; SSE41-NEXT:    retq
 ;
 ; AVX-LABEL: load_sext_test6:
 ; AVX:       ## BB#0:
@@ -200,30 +339,78 @@ define <8 x i16> @load_sext_test6(<8 x i8> *%ptr) {
 }
 
 define <4 x i64> @sext_4i1_to_4i64(<4 x i1> %mask) {
-; SSE-LABEL: sext_4i1_to_4i64:
-; SSE:       ## BB#0:
-; SSE-NEXT:    pslld $31, %xmm0
-; SSE-NEXT:    psrad $31, %xmm0
-; SSE-NEXT:    pshufd {{.*#+}} xmm1 = xmm0[0,0,1,0]
-; SSE-NEXT:    movd %xmm1, %rax
-; SSE-NEXT:    cltq
-; SSE-NEXT:    movd %rax, %xmm2
-; SSE-NEXT:    punpckhqdq {{.*#+}} xmm1 = xmm1[1,1]
-; SSE-NEXT:    movd %xmm1, %rax
-; SSE-NEXT:    cltq
-; SSE-NEXT:    movd %rax, %xmm1
-; SSE-NEXT:    punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm1[0]
-; SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,0,3,0]
-; SSE-NEXT:    movd %xmm0, %rax
-; SSE-NEXT:    cltq
-; SSE-NEXT:    movd %rax, %xmm1
-; SSE-NEXT:    punpckhqdq {{.*#+}} xmm0 = xmm0[1,1]
-; SSE-NEXT:    movd %xmm0, %rax
-; SSE-NEXT:    cltq
-; SSE-NEXT:    movd %rax, %xmm0
-; SSE-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
-; SSE-NEXT:    movdqa %xmm2, %xmm0
-; SSE-NEXT:    retq
+; SSE2-LABEL: sext_4i1_to_4i64:
+; SSE2:       ## BB#0:
+; SSE2-NEXT:    pslld $31, %xmm0
+; SSE2-NEXT:    psrad $31, %xmm0
+; SSE2-NEXT:    pshufd {{.*#+}} xmm1 = xmm0[0,0,1,0]
+; SSE2-NEXT:    movd %xmm1, %rax
+; SSE2-NEXT:    cltq
+; SSE2-NEXT:    movd %rax, %xmm2
+; SSE2-NEXT:    punpckhqdq {{.*#+}} xmm1 = xmm1[1,1]
+; SSE2-NEXT:    movd %xmm1, %rax
+; SSE2-NEXT:    cltq
+; SSE2-NEXT:    movd %rax, %xmm1
+; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm1[0]
+; SSE2-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,0,3,0]
+; SSE2-NEXT:    movd %xmm0, %rax
+; SSE2-NEXT:    cltq
+; SSE2-NEXT:    movd %rax, %xmm1
+; SSE2-NEXT:    punpckhqdq {{.*#+}} xmm0 = xmm0[1,1]
+; SSE2-NEXT:    movd %xmm0, %rax
+; SSE2-NEXT:    cltq
+; SSE2-NEXT:    movd %rax, %xmm0
+; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
+; SSE2-NEXT:    movdqa %xmm2, %xmm0
+; SSE2-NEXT:    retq
+;
+; SSSE3-LABEL: sext_4i1_to_4i64:
+; SSSE3:       ## BB#0:
+; SSSE3-NEXT:    pslld $31, %xmm0
+; SSSE3-NEXT:    psrad $31, %xmm0
+; SSSE3-NEXT:    pshufd {{.*#+}} xmm1 = xmm0[0,0,1,0]
+; SSSE3-NEXT:    movd %xmm1, %rax
+; SSSE3-NEXT:    cltq
+; SSSE3-NEXT:    movd %rax, %xmm2
+; SSSE3-NEXT:    punpckhqdq {{.*#+}} xmm1 = xmm1[1,1]
+; SSSE3-NEXT:    movd %xmm1, %rax
+; SSSE3-NEXT:    cltq
+; SSSE3-NEXT:    movd %rax, %xmm1
+; SSSE3-NEXT:    punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm1[0]
+; SSSE3-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,0,3,0]
+; SSSE3-NEXT:    movd %xmm0, %rax
+; SSSE3-NEXT:    cltq
+; SSSE3-NEXT:    movd %rax, %xmm1
+; SSSE3-NEXT:    punpckhqdq {{.*#+}} xmm0 = xmm0[1,1]
+; SSSE3-NEXT:    movd %xmm0, %rax
+; SSSE3-NEXT:    cltq
+; SSSE3-NEXT:    movd %rax, %xmm0
+; SSSE3-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
+; SSSE3-NEXT:    movdqa %xmm2, %xmm0
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: sext_4i1_to_4i64:
+; SSE41:       ## BB#0:
+; SSE41-NEXT:    pslld $31, %xmm0
+; SSE41-NEXT:    psrad $31, %xmm0
+; SSE41-NEXT:    pmovzxdq %xmm0, %xmm1
+; SSE41-NEXT:    pextrq $1, %xmm1, %rax
+; SSE41-NEXT:    cltq
+; SSE41-NEXT:    movd %rax, %xmm3
+; SSE41-NEXT:    movd %xmm1, %rax
+; SSE41-NEXT:    cltq
+; SSE41-NEXT:    movd %rax, %xmm2
+; SSE41-NEXT:    punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm3[0]
+; SSE41-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,0,3,0]
+; SSE41-NEXT:    pextrq $1, %xmm0, %rax
+; SSE41-NEXT:    cltq
+; SSE41-NEXT:    movd %rax, %xmm3
+; SSE41-NEXT:    movd %xmm0, %rax
+; SSE41-NEXT:    cltq
+; SSE41-NEXT:    movd %rax, %xmm1
+; SSE41-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm3[0]
+; SSE41-NEXT:    movdqa %xmm2, %xmm0
+; SSE41-NEXT:    retq
 ;
 ; AVX1-LABEL: sext_4i1_to_4i64:
 ; AVX1:       ## BB#0:
@@ -246,17 +433,40 @@ define <4 x i64> @sext_4i1_to_4i64(<4 x i1> %mask) {
 }
 
 define <16 x i16> @sext_16i8_to_16i16(<16 x i8> *%ptr) {
-; SSE-LABEL: sext_16i8_to_16i16:
-; SSE:       ## BB#0:
-; SSE-NEXT:    movdqa (%rdi), %xmm1
-; SSE-NEXT:    movdqa %xmm1, %xmm0
-; SSE-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
-; SSE-NEXT:    psllw $8, %xmm0
-; SSE-NEXT:    psraw $8, %xmm0
-; SSE-NEXT:    punpckhbw {{.*#+}} xmm1 = xmm1[8],xmm0[8],xmm1[9],xmm0[9],xmm1[10],xmm0[10],xmm1[11],xmm0[11],xmm1[12],xmm0[12],xmm1[13],xmm0[13],xmm1[14],xmm0[14],xmm1[15],xmm0[15]
-; SSE-NEXT:    psllw $8, %xmm1
-; SSE-NEXT:    psraw $8, %xmm1
-; SSE-NEXT:    retq
+; SSE2-LABEL: sext_16i8_to_16i16:
+; SSE2:       ## BB#0:
+; SSE2-NEXT:    movdqa (%rdi), %xmm1
+; SSE2-NEXT:    movdqa %xmm1, %xmm0
+; SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; SSE2-NEXT:    psllw $8, %xmm0
+; SSE2-NEXT:    psraw $8, %xmm0
+; SSE2-NEXT:    punpckhbw {{.*#+}} xmm1 = xmm1[8],xmm0[8],xmm1[9],xmm0[9],xmm1[10],xmm0[10],xmm1[11],xmm0[11],xmm1[12],xmm0[12],xmm1[13],xmm0[13],xmm1[14],xmm0[14],xmm1[15],xmm0[15]
+; SSE2-NEXT:    psllw $8, %xmm1
+; SSE2-NEXT:    psraw $8, %xmm1
+; SSE2-NEXT:    retq
+;
+; SSSE3-LABEL: sext_16i8_to_16i16:
+; SSSE3:       ## BB#0:
+; SSSE3-NEXT:    movdqa (%rdi), %xmm1
+; SSSE3-NEXT:    movdqa %xmm1, %xmm0
+; SSSE3-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; SSSE3-NEXT:    psllw $8, %xmm0
+; SSSE3-NEXT:    psraw $8, %xmm0
+; SSSE3-NEXT:    punpckhbw {{.*#+}} xmm1 = xmm1[8],xmm0[8],xmm1[9],xmm0[9],xmm1[10],xmm0[10],xmm1[11],xmm0[11],xmm1[12],xmm0[12],xmm1[13],xmm0[13],xmm1[14],xmm0[14],xmm1[15],xmm0[15]
+; SSSE3-NEXT:    psllw $8, %xmm1
+; SSSE3-NEXT:    psraw $8, %xmm1
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: sext_16i8_to_16i16:
+; SSE41:       ## BB#0:
+; SSE41-NEXT:    movdqa (%rdi), %xmm1
+; SSE41-NEXT:    pmovzxbw %xmm1, %xmm0
+; SSE41-NEXT:    punpckhbw {{.*#+}} xmm1 = xmm1[8],xmm0[8],xmm1[9],xmm0[9],xmm1[10],xmm0[10],xmm1[11],xmm0[11],xmm1[12],xmm0[12],xmm1[13],xmm0[13],xmm1[14],xmm0[14],xmm1[15],xmm0[15]
+; SSE41-NEXT:    psllw $8, %xmm1
+; SSE41-NEXT:    psraw $8, %xmm1
+; SSE41-NEXT:    psllw $8, %xmm0
+; SSE41-NEXT:    psraw $8, %xmm0
+; SSE41-NEXT:    retq
 ;
 ; AVX1-LABEL: sext_16i8_to_16i16:
 ; AVX1:       ## BB#0:
@@ -278,30 +488,78 @@ define <16 x i16> @sext_16i8_to_16i16(<16 x i8> *%ptr) {
 }
 
 define <4 x i64> @sext_4i8_to_4i64(<4 x i8> %mask) {
-; SSE-LABEL: sext_4i8_to_4i64:
-; SSE:       ## BB#0:
-; SSE-NEXT:    pslld $24, %xmm0
-; SSE-NEXT:    psrad $24, %xmm0
-; SSE-NEXT:    pshufd {{.*#+}} xmm1 = xmm0[0,0,1,0]
-; SSE-NEXT:    movd %xmm1, %rax
-; SSE-NEXT:    cltq
-; SSE-NEXT:    movd %rax, %xmm2
-; SSE-NEXT:    punpckhqdq {{.*#+}} xmm1 = xmm1[1,1]
-; SSE-NEXT:    movd %xmm1, %rax
-; SSE-NEXT:    cltq
-; SSE-NEXT:    movd %rax, %xmm1
-; SSE-NEXT:    punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm1[0]
-; SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,0,3,0]
-; SSE-NEXT:    movd %xmm0, %rax
-; SSE-NEXT:    cltq
-; SSE-NEXT:    movd %rax, %xmm1
-; SSE-NEXT:    punpckhqdq {{.*#+}} xmm0 = xmm0[1,1]
-; SSE-NEXT:    movd %xmm0, %rax
-; SSE-NEXT:    cltq
-; SSE-NEXT:    movd %rax, %xmm0
-; SSE-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
-; SSE-NEXT:    movdqa %xmm2, %xmm0
-; SSE-NEXT:    retq
+; SSE2-LABEL: sext_4i8_to_4i64:
+; SSE2:       ## BB#0:
+; SSE2-NEXT:    pslld $24, %xmm0
+; SSE2-NEXT:    psrad $24, %xmm0
+; SSE2-NEXT:    pshufd {{.*#+}} xmm1 = xmm0[0,0,1,0]
+; SSE2-NEXT:    movd %xmm1, %rax
+; SSE2-NEXT:    cltq
+; SSE2-NEXT:    movd %rax, %xmm2
+; SSE2-NEXT:    punpckhqdq {{.*#+}} xmm1 = xmm1[1,1]
+; SSE2-NEXT:    movd %xmm1, %rax
+; SSE2-NEXT:    cltq
+; SSE2-NEXT:    movd %rax, %xmm1
+; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm1[0]
+; SSE2-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,0,3,0]
+; SSE2-NEXT:    movd %xmm0, %rax
+; SSE2-NEXT:    cltq
+; SSE2-NEXT:    movd %rax, %xmm1
+; SSE2-NEXT:    punpckhqdq {{.*#+}} xmm0 = xmm0[1,1]
+; SSE2-NEXT:    movd %xmm0, %rax
+; SSE2-NEXT:    cltq
+; SSE2-NEXT:    movd %rax, %xmm0
+; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
+; SSE2-NEXT:    movdqa %xmm2, %xmm0
+; SSE2-NEXT:    retq
+;
+; SSSE3-LABEL: sext_4i8_to_4i64:
+; SSSE3:       ## BB#0:
+; SSSE3-NEXT:    pslld $24, %xmm0
+; SSSE3-NEXT:    psrad $24, %xmm0
+; SSSE3-NEXT:    pshufd {{.*#+}} xmm1 = xmm0[0,0,1,0]
+; SSSE3-NEXT:    movd %xmm1, %rax
+; SSSE3-NEXT:    cltq
+; SSSE3-NEXT:    movd %rax, %xmm2
+; SSSE3-NEXT:    punpckhqdq {{.*#+}} xmm1 = xmm1[1,1]
+; SSSE3-NEXT:    movd %xmm1, %rax
+; SSSE3-NEXT:    cltq
+; SSSE3-NEXT:    movd %rax, %xmm1
+; SSSE3-NEXT:    punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm1[0]
+; SSSE3-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,0,3,0]
+; SSSE3-NEXT:    movd %xmm0, %rax
+; SSSE3-NEXT:    cltq
+; SSSE3-NEXT:    movd %rax, %xmm1
+; SSSE3-NEXT:    punpckhqdq {{.*#+}} xmm0 = xmm0[1,1]
+; SSSE3-NEXT:    movd %xmm0, %rax
+; SSSE3-NEXT:    cltq
+; SSSE3-NEXT:    movd %rax, %xmm0
+; SSSE3-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
+; SSSE3-NEXT:    movdqa %xmm2, %xmm0
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: sext_4i8_to_4i64:
+; SSE41:       ## BB#0:
+; SSE41-NEXT:    pslld $24, %xmm0
+; SSE41-NEXT:    psrad $24, %xmm0
+; SSE41-NEXT:    pmovzxdq %xmm0, %xmm1
+; SSE41-NEXT:    pextrq $1, %xmm1, %rax
+; SSE41-NEXT:    cltq
+; SSE41-NEXT:    movd %rax, %xmm3
+; SSE41-NEXT:    movd %xmm1, %rax
+; SSE41-NEXT:    cltq
+; SSE41-NEXT:    movd %rax, %xmm2
+; SSE41-NEXT:    punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm3[0]
+; SSE41-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,0,3,0]
+; SSE41-NEXT:    pextrq $1, %xmm0, %rax
+; SSE41-NEXT:    cltq
+; SSE41-NEXT:    movd %rax, %xmm3
+; SSE41-NEXT:    movd %xmm0, %rax
+; SSE41-NEXT:    cltq
+; SSE41-NEXT:    movd %rax, %xmm1
+; SSE41-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm3[0]
+; SSE41-NEXT:    movdqa %xmm2, %xmm0
+; SSE41-NEXT:    retq
 ;
 ; AVX1-LABEL: sext_4i8_to_4i64:
 ; AVX1:       ## BB#0:
@@ -379,6 +637,28 @@ define <4 x i64> @load_sext_4i8_to_4i64(<4 x i8> *%ptr) {
 ; SSSE3-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm2[0]
 ; SSSE3-NEXT:    retq
 ;
+; SSE41-LABEL: load_sext_4i8_to_4i64:
+; SSE41:       ## BB#0:
+; SSE41-NEXT:    movd (%rdi), %xmm0
+; SSE41-NEXT:    pmovzxbd %xmm0, %xmm1
+; SSE41-NEXT:    pmovzxbq %xmm0, %xmm0
+; SSE41-NEXT:    pextrq $1, %xmm0, %rax
+; SSE41-NEXT:    movsbq %al, %rax
+; SSE41-NEXT:    movd %rax, %xmm2
+; SSE41-NEXT:    movd %xmm0, %rax
+; SSE41-NEXT:    movsbq %al, %rax
+; SSE41-NEXT:    movd %rax, %xmm0
+; SSE41-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm2[0]
+; SSE41-NEXT:    pshufd {{.*#+}} xmm1 = xmm1[2,0,3,0]
+; SSE41-NEXT:    pextrq $1, %xmm1, %rax
+; SSE41-NEXT:    movsbq %al, %rax
+; SSE41-NEXT:    movd %rax, %xmm2
+; SSE41-NEXT:    movd %xmm1, %rax
+; SSE41-NEXT:    movsbq %al, %rax
+; SSE41-NEXT:    movd %rax, %xmm1
+; SSE41-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm2[0]
+; SSE41-NEXT:    retq
+;
 ; AVX1-LABEL: load_sext_4i8_to_4i64:
 ; AVX1:       ## BB#0:
 ; AVX1-NEXT:    vpmovsxbd (%rdi), %xmm0
@@ -398,29 +678,75 @@ define <4 x i64> @load_sext_4i8_to_4i64(<4 x i8> *%ptr) {
 }
 
 define <4 x i64> @load_sext_4i16_to_4i64(<4 x i16> *%ptr) {
-; SSE-LABEL: load_sext_4i16_to_4i64:
-; SSE:       ## BB#0:
-; SSE-NEXT:    movq (%rdi), %xmm1
-; SSE-NEXT:    punpcklwd {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1],xmm1[2],xmm0[2],xmm1[3],xmm0[3]
-; SSE-NEXT:    pshufd {{.*#+}} xmm2 = xmm1[0,0,1,0]
-; SSE-NEXT:    movd %xmm2, %rax
-; SSE-NEXT:    movswq %ax, %rax
-; SSE-NEXT:    movd %rax, %xmm0
-; SSE-NEXT:    punpckhqdq {{.*#+}} xmm2 = xmm2[1,1]
-; SSE-NEXT:    movd %xmm2, %rax
-; SSE-NEXT:    movswq %ax, %rax
-; SSE-NEXT:    movd %rax, %xmm2
-; SSE-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm2[0]
-; SSE-NEXT:    pshufd {{.*#+}} xmm2 = xmm1[2,0,3,0]
-; SSE-NEXT:    movd %xmm2, %rax
-; SSE-NEXT:    movswq %ax, %rax
-; SSE-NEXT:    movd %rax, %xmm1
-; SSE-NEXT:    punpckhqdq {{.*#+}} xmm2 = xmm2[1,1]
-; SSE-NEXT:    movd %xmm2, %rax
-; SSE-NEXT:    movswq %ax, %rax
-; SSE-NEXT:    movd %rax, %xmm2
-; SSE-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm2[0]
-; SSE-NEXT:    retq
+; SSE2-LABEL: load_sext_4i16_to_4i64:
+; SSE2:       ## BB#0:
+; SSE2-NEXT:    movq (%rdi), %xmm1
+; SSE2-NEXT:    punpcklwd {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1],xmm1[2],xmm0[2],xmm1[3],xmm0[3]
+; SSE2-NEXT:    pshufd {{.*#+}} xmm2 = xmm1[0,0,1,0]
+; SSE2-NEXT:    movd %xmm2, %rax
+; SSE2-NEXT:    movswq %ax, %rax
+; SSE2-NEXT:    movd %rax, %xmm0
+; SSE2-NEXT:    punpckhqdq {{.*#+}} xmm2 = xmm2[1,1]
+; SSE2-NEXT:    movd %xmm2, %rax
+; SSE2-NEXT:    movswq %ax, %rax
+; SSE2-NEXT:    movd %rax, %xmm2
+; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm2[0]
+; SSE2-NEXT:    pshufd {{.*#+}} xmm2 = xmm1[2,0,3,0]
+; SSE2-NEXT:    movd %xmm2, %rax
+; SSE2-NEXT:    movswq %ax, %rax
+; SSE2-NEXT:    movd %rax, %xmm1
+; SSE2-NEXT:    punpckhqdq {{.*#+}} xmm2 = xmm2[1,1]
+; SSE2-NEXT:    movd %xmm2, %rax
+; SSE2-NEXT:    movswq %ax, %rax
+; SSE2-NEXT:    movd %rax, %xmm2
+; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm2[0]
+; SSE2-NEXT:    retq
+;
+; SSSE3-LABEL: load_sext_4i16_to_4i64:
+; SSSE3:       ## BB#0:
+; SSSE3-NEXT:    movq (%rdi), %xmm1
+; SSSE3-NEXT:    punpcklwd {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1],xmm1[2],xmm0[2],xmm1[3],xmm0[3]
+; SSSE3-NEXT:    pshufd {{.*#+}} xmm2 = xmm1[0,0,1,0]
+; SSSE3-NEXT:    movd %xmm2, %rax
+; SSSE3-NEXT:    movswq %ax, %rax
+; SSSE3-NEXT:    movd %rax, %xmm0
+; SSSE3-NEXT:    punpckhqdq {{.*#+}} xmm2 = xmm2[1,1]
+; SSSE3-NEXT:    movd %xmm2, %rax
+; SSSE3-NEXT:    movswq %ax, %rax
+; SSSE3-NEXT:    movd %rax, %xmm2
+; SSSE3-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm2[0]
+; SSSE3-NEXT:    pshufd {{.*#+}} xmm2 = xmm1[2,0,3,0]
+; SSSE3-NEXT:    movd %xmm2, %rax
+; SSSE3-NEXT:    movswq %ax, %rax
+; SSSE3-NEXT:    movd %rax, %xmm1
+; SSSE3-NEXT:    punpckhqdq {{.*#+}} xmm2 = xmm2[1,1]
+; SSSE3-NEXT:    movd %xmm2, %rax
+; SSSE3-NEXT:    movswq %ax, %rax
+; SSSE3-NEXT:    movd %rax, %xmm2
+; SSSE3-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm2[0]
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: load_sext_4i16_to_4i64:
+; SSE41:       ## BB#0:
+; SSE41-NEXT:    movq (%rdi), %xmm0
+; SSE41-NEXT:    pmovzxwd %xmm0, %xmm1
+; SSE41-NEXT:    pmovzxwq %xmm0, %xmm0
+; SSE41-NEXT:    pextrq $1, %xmm0, %rax
+; SSE41-NEXT:    movswq %ax, %rax
+; SSE41-NEXT:    movd %rax, %xmm2
+; SSE41-NEXT:    movd %xmm0, %rax
+; SSE41-NEXT:    movswq %ax, %rax
+; SSE41-NEXT:    movd %rax, %xmm0
+; SSE41-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm2[0]
+; SSE41-NEXT:    pshufd {{.*#+}} xmm1 = xmm1[2,0,3,0]
+; SSE41-NEXT:    pextrq $1, %xmm1, %rax
+; SSE41-NEXT:    movswq %ax, %rax
+; SSE41-NEXT:    movd %rax, %xmm2
+; SSE41-NEXT:    movd %xmm1, %rax
+; SSE41-NEXT:    movswq %ax, %rax
+; SSE41-NEXT:    movd %rax, %xmm1
+; SSE41-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm2[0]
+; SSE41-NEXT:    retq
 ;
 ; AVX1-LABEL: load_sext_4i16_to_4i64:
 ; AVX1:       ## BB#0:
diff --git a/test/CodeGen/X86/vector-zext.ll b/test/CodeGen/X86/vector-zext.ll
index 880be94ba2f..d47417a59fd 100644
--- a/test/CodeGen/X86/vector-zext.ll
+++ b/test/CodeGen/X86/vector-zext.ll
@@ -1,39 +1,154 @@
-; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=corei7-avx | FileCheck %s
+; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=x86-64 -mattr=+sse2 | FileCheck %s --check-prefix=SSE --check-prefix=SSE2
+; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=x86-64 -mattr=+ssse3 | FileCheck %s --check-prefix=SSE --check-prefix=SSSE3
+; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=x86-64 -mattr=+sse4.1 | FileCheck %s --check-prefix=SSE --check-prefix=SSE41
+; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=x86-64 -mattr=+avx | FileCheck %s --check-prefix=AVX --check-prefix=AVX1
+; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=x86-64 -mattr=+avx2 | FileCheck %s --check-prefix=AVX --check-prefix=AVX2
 
 define <8 x i32> @zext_8i16_to_8i32(<8 x i16> %A) nounwind uwtable readnone ssp {
-; CHECK-LABEL: zext_8i16_to_8i32:
-; CHECK:       ## BB#0: ## %entry
-; CHECK-NEXT:    vpxor %xmm1, %xmm1, %xmm1
-; CHECK-NEXT:    vpunpckhwd {{.*#+}} xmm2 = xmm0[4],xmm1[4],xmm0[5],xmm1[5],xmm0[6],xmm1[6],xmm0[7],xmm1[7]
-; CHECK-NEXT:    vpunpcklwd {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
-; CHECK-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
-; CHECK-NEXT:    retq
+; SSE2-LABEL: zext_8i16_to_8i32:
+; SSE2:       ## BB#0: ## %entry
+; SSE2-NEXT:    movdqa %xmm0, %xmm2
+; SSE2-NEXT:    punpcklwd {{.*#+}} xmm2 = xmm2[0],xmm0[0],xmm2[1],xmm0[1],xmm2[2],xmm0[2],xmm2[3],xmm0[3]
+; SSE2-NEXT:    movdqa {{.*#+}} xmm1 = [65535,65535,65535,65535]
+; SSE2-NEXT:    pand %xmm1, %xmm2
+; SSE2-NEXT:    punpckhwd {{.*#+}} xmm0 = xmm0[4,4,5,5,6,6,7,7]
+; SSE2-NEXT:    pand %xmm0, %xmm1
+; SSE2-NEXT:    movdqa %xmm2, %xmm0
+; SSE2-NEXT:    retq
+;
+; SSSE3-LABEL: zext_8i16_to_8i32:
+; SSSE3:       ## BB#0: ## %entry
+; SSSE3-NEXT:    movdqa %xmm0, %xmm2
+; SSSE3-NEXT:    punpcklwd {{.*#+}} xmm2 = xmm2[0],xmm0[0],xmm2[1],xmm0[1],xmm2[2],xmm0[2],xmm2[3],xmm0[3]
+; SSSE3-NEXT:    movdqa {{.*#+}} xmm1 = [65535,65535,65535,65535]
+; SSSE3-NEXT:    pand %xmm1, %xmm2
+; SSSE3-NEXT:    punpckhwd {{.*#+}} xmm0 = xmm0[4,4,5,5,6,6,7,7]
+; SSSE3-NEXT:    pand %xmm0, %xmm1
+; SSSE3-NEXT:    movdqa %xmm2, %xmm0
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: zext_8i16_to_8i32:
+; SSE41:       ## BB#0: ## %entry
+; SSE41-NEXT:    pmovzxwd %xmm0, %xmm2
+; SSE41-NEXT:    movdqa {{.*#+}} xmm1 = [65535,65535,65535,65535]
+; SSE41-NEXT:    pand %xmm1, %xmm2
+; SSE41-NEXT:    punpckhwd {{.*#+}} xmm0 = xmm0[4,4,5,5,6,6,7,7]
+; SSE41-NEXT:    pand %xmm0, %xmm1
+; SSE41-NEXT:    movdqa %xmm2, %xmm0
+; SSE41-NEXT:    retq
+;
+; AVX1-LABEL: zext_8i16_to_8i32:
+; AVX1:       ## BB#0: ## %entry
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpunpckhwd {{.*#+}} xmm2 = xmm0[4],xmm1[4],xmm0[5],xmm1[5],xmm0[6],xmm1[6],xmm0[7],xmm1[7]
+; AVX1-NEXT:    vpunpcklwd {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX1-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: zext_8i16_to_8i32:
+; AVX2:       ## BB#0: ## %entry
+; AVX2-NEXT:    vpmovzxwd %xmm0, %ymm0
+; AVX2-NEXT:    retq
 entry:
   %B = zext <8 x i16> %A to <8 x i32>
   ret <8 x i32>%B
 }
 
 define <4 x i64> @zext_4i32_to_4i64(<4 x i32> %A) nounwind uwtable readnone ssp {
-; CHECK-LABEL: zext_4i32_to_4i64:
-; CHECK:       ## BB#0: ## %entry
-; CHECK-NEXT:    vpxor %xmm1, %xmm1, %xmm1
-; CHECK-NEXT:    vpunpckhdq {{.*#+}} xmm2 = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
-; CHECK-NEXT:    vpunpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
-; CHECK-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
-; CHECK-NEXT:    retq
+; SSE2-LABEL: zext_4i32_to_4i64:
+; SSE2:       ## BB#0: ## %entry
+; SSE2-NEXT:    pshufd {{.*#+}} xmm2 = xmm0[0,0,1,0]
+; SSE2-NEXT:    movdqa {{.*#+}} xmm3 = [4294967295,4294967295]
+; SSE2-NEXT:    pand %xmm3, %xmm2
+; SSE2-NEXT:    pshufd {{.*#+}} xmm1 = xmm0[2,0,3,0]
+; SSE2-NEXT:    pand %xmm3, %xmm1
+; SSE2-NEXT:    movdqa %xmm2, %xmm0
+; SSE2-NEXT:    retq
+;
+; SSSE3-LABEL: zext_4i32_to_4i64:
+; SSSE3:       ## BB#0: ## %entry
+; SSSE3-NEXT:    pshufd {{.*#+}} xmm2 = xmm0[0,0,1,0]
+; SSSE3-NEXT:    movdqa {{.*#+}} xmm3 = [4294967295,4294967295]
+; SSSE3-NEXT:    pand %xmm3, %xmm2
+; SSSE3-NEXT:    pshufd {{.*#+}} xmm1 = xmm0[2,0,3,0]
+; SSSE3-NEXT:    pand %xmm3, %xmm1
+; SSSE3-NEXT:    movdqa %xmm2, %xmm0
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: zext_4i32_to_4i64:
+; SSE41:       ## BB#0: ## %entry
+; SSE41-NEXT:    pmovzxdq %xmm0, %xmm2
+; SSE41-NEXT:    movdqa {{.*#+}} xmm3 = [4294967295,4294967295]
+; SSE41-NEXT:    pand %xmm3, %xmm2
+; SSE41-NEXT:    pshufd {{.*#+}} xmm1 = xmm0[2,0,3,0]
+; SSE41-NEXT:    pand %xmm3, %xmm1
+; SSE41-NEXT:    movdqa %xmm2, %xmm0
+; SSE41-NEXT:    retq
+;
+; AVX1-LABEL: zext_4i32_to_4i64:
+; AVX1:       ## BB#0: ## %entry
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpunpckhdq {{.*#+}} xmm2 = xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX1-NEXT:    vpunpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; AVX1-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: zext_4i32_to_4i64:
+; AVX2:       ## BB#0: ## %entry
+; AVX2-NEXT:    vpmovzxdq %xmm0, %ymm0
+; AVX2-NEXT:    retq
 entry:
   %B = zext <4 x i32> %A to <4 x i64>
   ret <4 x i64>%B
 }
 
 define <8 x i32> @zext_8i8_to_8i32(<8 x i8> %z) {
-; CHECK-LABEL: zext_8i8_to_8i32:
-; CHECK:       ## BB#0: ## %entry
-; CHECK-NEXT:    vpunpckhwd {{.*#+}} xmm1 = xmm0[4,4,5,5,6,6,7,7]
-; CHECK-NEXT:    vpmovzxwd %xmm0, %xmm0
-; CHECK-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
-; CHECK-NEXT:    vandps {{.*}}, %ymm0, %ymm0
-; CHECK-NEXT:    retq
+; SSE2-LABEL: zext_8i8_to_8i32:
+; SSE2:       ## BB#0: ## %entry
+; SSE2-NEXT:    movdqa %xmm0, %xmm2
+; SSE2-NEXT:    punpcklwd {{.*#+}} xmm2 = xmm2[0],xmm0[0],xmm2[1],xmm0[1],xmm2[2],xmm0[2],xmm2[3],xmm0[3]
+; SSE2-NEXT:    movdqa {{.*#+}} xmm1 = [255,255,255,255]
+; SSE2-NEXT:    pand %xmm1, %xmm2
+; SSE2-NEXT:    punpckhwd {{.*#+}} xmm0 = xmm0[4,4,5,5,6,6,7,7]
+; SSE2-NEXT:    pand %xmm0, %xmm1
+; SSE2-NEXT:    movdqa %xmm2, %xmm0
+; SSE2-NEXT:    retq
+;
+; SSSE3-LABEL: zext_8i8_to_8i32:
+; SSSE3:       ## BB#0: ## %entry
+; SSSE3-NEXT:    movdqa %xmm0, %xmm2
+; SSSE3-NEXT:    punpcklwd {{.*#+}} xmm2 = xmm2[0],xmm0[0],xmm2[1],xmm0[1],xmm2[2],xmm0[2],xmm2[3],xmm0[3]
+; SSSE3-NEXT:    movdqa {{.*#+}} xmm1 = [255,255,255,255]
+; SSSE3-NEXT:    pand %xmm1, %xmm2
+; SSSE3-NEXT:    punpckhwd {{.*#+}} xmm0 = xmm0[4,4,5,5,6,6,7,7]
+; SSSE3-NEXT:    pand %xmm0, %xmm1
+; SSSE3-NEXT:    movdqa %xmm2, %xmm0
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: zext_8i8_to_8i32:
+; SSE41:       ## BB#0: ## %entry
+; SSE41-NEXT:    pmovzxwd %xmm0, %xmm2
+; SSE41-NEXT:    movdqa {{.*#+}} xmm1 = [255,255,255,255]
+; SSE41-NEXT:    pand %xmm1, %xmm2
+; SSE41-NEXT:    punpckhwd {{.*#+}} xmm0 = xmm0[4,4,5,5,6,6,7,7]
+; SSE41-NEXT:    pand %xmm0, %xmm1
+; SSE41-NEXT:    movdqa %xmm2, %xmm0
+; SSE41-NEXT:    retq
+;
+; AVX1-LABEL: zext_8i8_to_8i32:
+; AVX1:       ## BB#0: ## %entry
+; AVX1-NEXT:    vpunpckhwd {{.*#+}} xmm1 = xmm0[4,4,5,5,6,6,7,7]
+; AVX1-NEXT:    vpmovzxwd %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    vandps LCPI2_0(%rip), %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: zext_8i8_to_8i32:
+; AVX2:       ## BB#0: ## %entry
+; AVX2-NEXT:    vpmovzxwd %xmm0, %ymm0
+; AVX2-NEXT:    vpbroadcastd LCPI2_0(%rip), %ymm1
+; AVX2-NEXT:    vpand %ymm1, %ymm0, %ymm0
+; AVX2-NEXT:    retq
 entry:
   %t = zext <8 x i8> %z to <8 x i32>
   ret <8 x i32> %t
@@ -41,13 +156,50 @@ entry:
 
 ; PR17654
 define <16 x i16> @zext_16i8_to_16i16(<16 x i8> %z) {
-; CHECK-LABEL: zext_16i8_to_16i16:
-; CHECK:       ## BB#0: ## %entry
-; CHECK-NEXT:    vpxor %xmm1, %xmm1, %xmm1
-; CHECK-NEXT:    vpunpckhbw {{.*#+}} xmm2 = xmm0[8],xmm1[8],xmm0[9],xmm1[9],xmm0[10],xmm1[10],xmm0[11],xmm1[11],xmm0[12],xmm1[12],xmm0[13],xmm1[13],xmm0[14],xmm1[14],xmm0[15],xmm1[15]
-; CHECK-NEXT:    vpunpcklbw {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3],xmm0[4],xmm1[4],xmm0[5],xmm1[5],xmm0[6],xmm1[6],xmm0[7],xmm1[7]
-; CHECK-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
-; CHECK-NEXT:    retq
+; SSE2-LABEL: zext_16i8_to_16i16:
+; SSE2:       ## BB#0: ## %entry
+; SSE2-NEXT:    movdqa %xmm0, %xmm2
+; SSE2-NEXT:    punpcklbw {{.*#+}} xmm2 = xmm2[0],xmm0[0],xmm2[1],xmm0[1],xmm2[2],xmm0[2],xmm2[3],xmm0[3],xmm2[4],xmm0[4],xmm2[5],xmm0[5],xmm2[6],xmm0[6],xmm2[7],xmm0[7]
+; SSE2-NEXT:    movdqa {{.*#+}} xmm1 = [255,255,255,255,255,255,255,255]
+; SSE2-NEXT:    pand %xmm1, %xmm2
+; SSE2-NEXT:    punpckhbw {{.*#+}} xmm0 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
+; SSE2-NEXT:    pand %xmm0, %xmm1
+; SSE2-NEXT:    movdqa %xmm2, %xmm0
+; SSE2-NEXT:    retq
+;
+; SSSE3-LABEL: zext_16i8_to_16i16:
+; SSSE3:       ## BB#0: ## %entry
+; SSSE3-NEXT:    movdqa %xmm0, %xmm2
+; SSSE3-NEXT:    punpcklbw {{.*#+}} xmm2 = xmm2[0],xmm0[0],xmm2[1],xmm0[1],xmm2[2],xmm0[2],xmm2[3],xmm0[3],xmm2[4],xmm0[4],xmm2[5],xmm0[5],xmm2[6],xmm0[6],xmm2[7],xmm0[7]
+; SSSE3-NEXT:    movdqa {{.*#+}} xmm1 = [255,255,255,255,255,255,255,255]
+; SSSE3-NEXT:    pand %xmm1, %xmm2
+; SSSE3-NEXT:    punpckhbw {{.*#+}} xmm0 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
+; SSSE3-NEXT:    pand %xmm0, %xmm1
+; SSSE3-NEXT:    movdqa %xmm2, %xmm0
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: zext_16i8_to_16i16:
+; SSE41:       ## BB#0: ## %entry
+; SSE41-NEXT:    pmovzxbw %xmm0, %xmm2
+; SSE41-NEXT:    movdqa {{.*#+}} xmm1 = [255,255,255,255,255,255,255,255]
+; SSE41-NEXT:    pand %xmm1, %xmm2
+; SSE41-NEXT:    punpckhbw {{.*#+}} xmm0 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
+; SSE41-NEXT:    pand %xmm0, %xmm1
+; SSE41-NEXT:    movdqa %xmm2, %xmm0
+; SSE41-NEXT:    retq
+;
+; AVX1-LABEL: zext_16i8_to_16i16:
+; AVX1:       ## BB#0: ## %entry
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpunpckhbw {{.*#+}} xmm2 = xmm0[8],xmm1[8],xmm0[9],xmm1[9],xmm0[10],xmm1[10],xmm0[11],xmm1[11],xmm0[12],xmm1[12],xmm0[13],xmm1[13],xmm0[14],xmm1[14],xmm0[15],xmm1[15]
+; AVX1-NEXT:    vpunpcklbw {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3],xmm0[4],xmm1[4],xmm0[5],xmm1[5],xmm0[6],xmm1[6],xmm0[7],xmm1[7]
+; AVX1-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: zext_16i8_to_16i16:
+; AVX2:       ## BB#0: ## %entry
+; AVX2-NEXT:    vpmovzxbw %xmm0, %ymm0
+; AVX2-NEXT:    retq
 entry:
   %t = zext <16 x i8> %z to <16 x i16>
   ret <16 x i16> %t