From e6dad29b16bd340937985131a0368214052dff62 Mon Sep 17 00:00:00 2001
From: Simon Pilgrim <llvm-dev@redking.me.uk>
Date: Sat, 25 Jul 2015 14:07:20 +0000
Subject: [PATCH] [X86][SSE] Added additional vector sign/zero load extension
 tests.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@243216 91177308-0d34-0410-b5e6-96231b3b80d8
---
 test/CodeGen/X86/vector-sext.ll | 721 ++++++++++++++++++++------------
 test/CodeGen/X86/vector-zext.ll | 403 ++++++++++++++++--
 2 files changed, 808 insertions(+), 316 deletions(-)

diff --git a/test/CodeGen/X86/vector-sext.ll b/test/CodeGen/X86/vector-sext.ll
index 5938b56acb5..8d931a3109a 100644
--- a/test/CodeGen/X86/vector-sext.ll
+++ b/test/CodeGen/X86/vector-sext.ll
@@ -589,170 +589,296 @@ define <4 x i64> @sext_4i32_to_4i64(<4 x i32> %A) nounwind uwtable readnone ssp
 ; X32-SSE41-NEXT:    retl
 entry:
   %B = sext <4 x i32> %A to <4 x i64>
-  ret <4 x i64>%B
+  ret <4 x i64> %B
 }
 
-define i32 @sext_2i8_to_i32(<16 x i8> %A) nounwind uwtable readnone ssp {
-; SSE2-LABEL: sext_2i8_to_i32:
+define <2 x i64> @load_sext_2i8_to_2i64(<2 x i8> *%ptr) {
+; SSE2-LABEL: load_sext_2i8_to_2i64:
 ; SSE2:       # BB#0: # %entry
+; SSE2-NEXT:    movzwl (%rdi), %eax
+; SSE2-NEXT:    movd %eax, %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:    movd %xmm0, %eax
+; SSE2-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3]
+; SSE2-NEXT:    movdqa %xmm0, %xmm1
+; SSE2-NEXT:    psrad $31, %xmm1
+; SSE2-NEXT:    psrad $24, %xmm0
+; SSE2-NEXT:    punpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
 ; SSE2-NEXT:    retq
 ;
-; SSSE3-LABEL: sext_2i8_to_i32:
+; SSSE3-LABEL: load_sext_2i8_to_2i64:
 ; SSSE3:       # BB#0: # %entry
+; SSSE3-NEXT:    movzwl (%rdi), %eax
+; SSSE3-NEXT:    movd %eax, %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:    movd %xmm0, %eax
+; SSSE3-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3]
+; SSSE3-NEXT:    movdqa %xmm0, %xmm1
+; SSSE3-NEXT:    psrad $31, %xmm1
+; SSSE3-NEXT:    psrad $24, %xmm0
+; SSSE3-NEXT:    punpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
 ; SSSE3-NEXT:    retq
 ;
-; SSE41-LABEL: sext_2i8_to_i32:
+; SSE41-LABEL: load_sext_2i8_to_2i64:
 ; SSE41:       # BB#0: # %entry
-; SSE41-NEXT:    pmovsxbw %xmm0, %xmm0
-; SSE41-NEXT:    movd %xmm0, %eax
+; SSE41-NEXT:    pmovsxbq (%rdi), %xmm0
 ; SSE41-NEXT:    retq
 ;
-; AVX-LABEL: sext_2i8_to_i32:
+; AVX-LABEL: load_sext_2i8_to_2i64:
 ; AVX:       # BB#0: # %entry
-; AVX-NEXT:    vpmovsxbw %xmm0, %xmm0
-; AVX-NEXT:    vmovd %xmm0, %eax
+; AVX-NEXT:    vpmovsxbq (%rdi), %xmm0
 ; AVX-NEXT:    retq
 ;
-; X32-SSE41-LABEL: sext_2i8_to_i32:
+; X32-SSE41-LABEL: load_sext_2i8_to_2i64:
 ; X32-SSE41:       # BB#0: # %entry
-; X32-SSE41-NEXT:    pushl %eax
-; X32-SSE41-NEXT:  .Ltmp0:
-; X32-SSE41-NEXT:    .cfi_def_cfa_offset 8
-; X32-SSE41-NEXT:    pmovsxbw %xmm0, %xmm0
-; X32-SSE41-NEXT:    movd %xmm0, %eax
-; X32-SSE41-NEXT:    popl %edx
+; X32-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X32-SSE41-NEXT:    pmovsxbq (%eax), %xmm0
 ; X32-SSE41-NEXT:    retl
 entry:
-  %Shuf = shufflevector <16 x i8> %A, <16 x i8> undef, <2 x i32> <i32 0, i32 1>
-  %Ex = sext <2 x i8> %Shuf to <2 x i16>
-  %Bc = bitcast <2 x i16> %Ex to i32
-  ret i32 %Bc
+ %X = load <2 x i8>, <2 x i8>* %ptr
+ %Y = sext <2 x i8> %X to <2 x i64>
+ ret <2 x i64> %Y
 }
 
-define <4 x i32> @load_sext_4i16_to_4i32(<4 x i16> *%ptr) {
-; SSE2-LABEL: load_sext_4i16_to_4i32:
+define <4 x i32> @load_sext_4i8_to_4i32(<4 x i8> *%ptr) {
+; SSE2-LABEL: load_sext_4i8_to_4i32:
 ; SSE2:       # BB#0: # %entry
-; SSE2-NEXT:    movq {{.*#+}} xmm0 = mem[0],zero
+; SSE2-NEXT:    movd {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
 ; SSE2-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3]
-; SSE2-NEXT:    psrad $16, %xmm0
+; SSE2-NEXT:    psrad $24, %xmm0
 ; SSE2-NEXT:    retq
 ;
-; SSSE3-LABEL: load_sext_4i16_to_4i32:
+; SSSE3-LABEL: load_sext_4i8_to_4i32:
 ; SSSE3:       # BB#0: # %entry
-; SSSE3-NEXT:    movq {{.*#+}} xmm0 = mem[0],zero
+; SSSE3-NEXT:    movd {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; SSSE3-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
 ; SSSE3-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3]
-; SSSE3-NEXT:    psrad $16, %xmm0
+; SSSE3-NEXT:    psrad $24, %xmm0
 ; SSSE3-NEXT:    retq
 ;
-; SSE41-LABEL: load_sext_4i16_to_4i32:
+; SSE41-LABEL: load_sext_4i8_to_4i32:
 ; SSE41:       # BB#0: # %entry
-; SSE41-NEXT:    pmovsxwd (%rdi), %xmm0
+; SSE41-NEXT:    pmovsxbd (%rdi), %xmm0
 ; SSE41-NEXT:    retq
 ;
-; AVX-LABEL: load_sext_4i16_to_4i32:
+; AVX-LABEL: load_sext_4i8_to_4i32:
 ; AVX:       # BB#0: # %entry
-; AVX-NEXT:    vpmovsxwd (%rdi), %xmm0
+; AVX-NEXT:    vpmovsxbd (%rdi), %xmm0
 ; AVX-NEXT:    retq
 ;
-; X32-SSE41-LABEL: load_sext_4i16_to_4i32:
+; X32-SSE41-LABEL: load_sext_4i8_to_4i32:
 ; X32-SSE41:       # BB#0: # %entry
 ; X32-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %eax
-; X32-SSE41-NEXT:    pmovsxwd (%eax), %xmm0
+; X32-SSE41-NEXT:    pmovsxbd (%eax), %xmm0
 ; X32-SSE41-NEXT:    retl
 entry:
- %X = load <4 x i16>, <4 x i16>* %ptr
- %Y = sext <4 x i16> %X to <4 x i32>
- ret <4 x i32>%Y
+ %X = load <4 x i8>, <4 x i8>* %ptr
+ %Y = sext <4 x i8> %X to <4 x i32>
+ ret <4 x i32> %Y
 }
 
-define <4 x i32> @load_sext_4i8_to_4i32(<4 x i8> *%ptr) {
-; SSE2-LABEL: load_sext_4i8_to_4i32:
+define <4 x i64> @load_sext_4i8_to_4i64(<4 x i8> *%ptr) {
+; SSE2-LABEL: load_sext_4i8_to_4i64:
+; SSE2:       # BB#0: # %entry
+; 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:    movsbq 3(%rdi), %rax
+; SSE2-NEXT:    movd %rax, %xmm2
+; SSE2-NEXT:    movsbq 2(%rdi), %rax
+; SSE2-NEXT:    movd %rax, %xmm1
+; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm2[0]
+; SSE2-NEXT:    retq
+;
+; SSSE3-LABEL: load_sext_4i8_to_4i64:
+; SSSE3:       # BB#0: # %entry
+; 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:    movsbq 3(%rdi), %rax
+; SSSE3-NEXT:    movd %rax, %xmm2
+; SSSE3-NEXT:    movsbq 2(%rdi), %rax
+; SSSE3-NEXT:    movd %rax, %xmm1
+; SSSE3-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm2[0]
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: load_sext_4i8_to_4i64:
+; SSE41:       # BB#0: # %entry
+; SSE41-NEXT:    pmovsxbq (%rdi), %xmm0
+; SSE41-NEXT:    pmovsxbq 2(%rdi), %xmm1
+; SSE41-NEXT:    retq
+;
+; AVX1-LABEL: load_sext_4i8_to_4i64:
+; AVX1:       # BB#0: # %entry
+; AVX1-NEXT:    vpmovsxbd (%rdi), %xmm0
+; AVX1-NEXT:    vpmovsxdq %xmm0, %xmm1
+; AVX1-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[2,3,0,1]
+; AVX1-NEXT:    vpmovsxdq %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: load_sext_4i8_to_4i64:
+; AVX2:       # BB#0: # %entry
+; AVX2-NEXT:    vpmovsxbq (%rdi), %ymm0
+; AVX2-NEXT:    retq
+;
+; X32-SSE41-LABEL: load_sext_4i8_to_4i64:
+; X32-SSE41:       # BB#0: # %entry
+; X32-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X32-SSE41-NEXT:    pmovsxbq (%eax), %xmm0
+; X32-SSE41-NEXT:    pmovsxbq 2(%eax), %xmm1
+; X32-SSE41-NEXT:    retl
+entry:
+ %X = load <4 x i8>, <4 x i8>* %ptr
+ %Y = sext <4 x i8> %X to <4 x i64>
+ ret <4 x i64> %Y
+}
+
+define <8 x i16> @load_sext_8i8_to_8i16(<8 x i8> *%ptr) {
+; SSE2-LABEL: load_sext_8i8_to_8i16:
+; SSE2:       # BB#0: # %entry
+; SSE2-NEXT:    movq {{.*#+}} xmm0 = mem[0],zero
+; 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_8i8_to_8i16:
+; SSSE3:       # BB#0: # %entry
+; SSSE3-NEXT:    movq {{.*#+}} xmm0 = mem[0],zero
+; 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_8i8_to_8i16:
+; SSE41:       # BB#0: # %entry
+; SSE41-NEXT:    pmovsxbw (%rdi), %xmm0
+; SSE41-NEXT:    retq
+;
+; AVX-LABEL: load_sext_8i8_to_8i16:
+; AVX:       # BB#0: # %entry
+; AVX-NEXT:    vpmovsxbw (%rdi), %xmm0
+; AVX-NEXT:    retq
+;
+; X32-SSE41-LABEL: load_sext_8i8_to_8i16:
+; X32-SSE41:       # BB#0: # %entry
+; X32-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X32-SSE41-NEXT:    pmovsxbw (%eax), %xmm0
+; X32-SSE41-NEXT:    retl
+entry:
+ %X = load <8 x i8>, <8 x i8>* %ptr
+ %Y = sext <8 x i8> %X to <8 x i16>
+ ret <8 x i16> %Y
+}
+
+define <8 x i32> @load_sext_8i8_to_8i32(<8 x i8> *%ptr) {
+; SSE2-LABEL: load_sext_8i8_to_8i32:
 ; SSE2:       # BB#0: # %entry
 ; SSE2-NEXT:    movd {{.*#+}} xmm0 = mem[0],zero,zero,zero
 ; SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
 ; SSE2-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3]
 ; SSE2-NEXT:    psrad $24, %xmm0
+; SSE2-NEXT:    movd {{.*#+}} xmm1 = mem[0],zero,zero,zero
+; SSE2-NEXT:    punpcklbw {{.*#+}} xmm1 = xmm1[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; SSE2-NEXT:    punpcklwd {{.*#+}} xmm1 = xmm1[0,0,1,1,2,2,3,3]
+; SSE2-NEXT:    psrad $24, %xmm1
 ; SSE2-NEXT:    retq
 ;
-; SSSE3-LABEL: load_sext_4i8_to_4i32:
+; SSSE3-LABEL: load_sext_8i8_to_8i32:
 ; SSSE3:       # BB#0: # %entry
 ; SSSE3-NEXT:    movd {{.*#+}} xmm0 = mem[0],zero,zero,zero
 ; SSSE3-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
 ; SSSE3-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3]
 ; SSSE3-NEXT:    psrad $24, %xmm0
+; SSSE3-NEXT:    movd {{.*#+}} xmm1 = mem[0],zero,zero,zero
+; SSSE3-NEXT:    punpcklbw {{.*#+}} xmm1 = xmm1[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; SSSE3-NEXT:    punpcklwd {{.*#+}} xmm1 = xmm1[0,0,1,1,2,2,3,3]
+; SSSE3-NEXT:    psrad $24, %xmm1
 ; SSSE3-NEXT:    retq
 ;
-; SSE41-LABEL: load_sext_4i8_to_4i32:
+; SSE41-LABEL: load_sext_8i8_to_8i32:
 ; SSE41:       # BB#0: # %entry
 ; SSE41-NEXT:    pmovsxbd (%rdi), %xmm0
+; SSE41-NEXT:    pmovsxbd 4(%rdi), %xmm1
 ; SSE41-NEXT:    retq
 ;
-; AVX-LABEL: load_sext_4i8_to_4i32:
-; AVX:       # BB#0: # %entry
-; AVX-NEXT:    vpmovsxbd (%rdi), %xmm0
-; AVX-NEXT:    retq
+; AVX1-LABEL: load_sext_8i8_to_8i32:
+; AVX1:       # BB#0: # %entry
+; AVX1-NEXT:    vpmovsxbw (%rdi), %xmm0
+; AVX1-NEXT:    vpmovsxwd %xmm0, %xmm1
+; AVX1-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[2,3,0,1]
+; AVX1-NEXT:    vpmovsxwd %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
 ;
-; X32-SSE41-LABEL: load_sext_4i8_to_4i32:
+; AVX2-LABEL: load_sext_8i8_to_8i32:
+; AVX2:       # BB#0: # %entry
+; AVX2-NEXT:    vpmovsxbd (%rdi), %ymm0
+; AVX2-NEXT:    retq
+;
+; X32-SSE41-LABEL: load_sext_8i8_to_8i32:
 ; X32-SSE41:       # BB#0: # %entry
 ; X32-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %eax
 ; X32-SSE41-NEXT:    pmovsxbd (%eax), %xmm0
+; X32-SSE41-NEXT:    pmovsxbd 4(%eax), %xmm1
 ; X32-SSE41-NEXT:    retl
 entry:
- %X = load <4 x i8>, <4 x i8>* %ptr
- %Y = sext <4 x i8> %X to <4 x i32>
- ret <4 x i32>%Y
+ %X = load <8 x i8>, <8 x i8>* %ptr
+ %Y = sext <8 x i8> %X to <8 x i32>
+ ret <8 x i32> %Y
 }
 
-define <2 x i64> @load_sext_2i8_to_2i64(<2 x i8> *%ptr) {
-; SSE2-LABEL: load_sext_2i8_to_2i64:
+define <16 x i16> @load_sext_16i8_to_16i16(<16 x i8> *%ptr) {
+; SSE2-LABEL: load_sext_16i8_to_16i16:
 ; SSE2:       # BB#0: # %entry
-; SSE2-NEXT:    movzwl (%rdi), %eax
-; SSE2-NEXT:    movd %eax, %xmm0
+; SSE2-NEXT:    movq {{.*#+}} xmm0 = mem[0],zero
 ; SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
-; SSE2-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3]
-; SSE2-NEXT:    movdqa %xmm0, %xmm1
-; SSE2-NEXT:    psrad $31, %xmm1
-; SSE2-NEXT:    psrad $24, %xmm0
-; SSE2-NEXT:    punpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; SSE2-NEXT:    psraw $8, %xmm0
+; SSE2-NEXT:    movq {{.*#+}} xmm1 = mem[0],zero
+; SSE2-NEXT:    punpcklbw {{.*#+}} xmm1 = xmm1[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; SSE2-NEXT:    psraw $8, %xmm1
 ; SSE2-NEXT:    retq
 ;
-; SSSE3-LABEL: load_sext_2i8_to_2i64:
+; SSSE3-LABEL: load_sext_16i8_to_16i16:
 ; SSSE3:       # BB#0: # %entry
-; SSSE3-NEXT:    movzwl (%rdi), %eax
-; SSSE3-NEXT:    movd %eax, %xmm0
+; SSSE3-NEXT:    movq {{.*#+}} xmm0 = mem[0],zero
 ; SSSE3-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
-; SSSE3-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3]
-; SSSE3-NEXT:    movdqa %xmm0, %xmm1
-; SSSE3-NEXT:    psrad $31, %xmm1
-; SSSE3-NEXT:    psrad $24, %xmm0
-; SSSE3-NEXT:    punpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; SSSE3-NEXT:    psraw $8, %xmm0
+; SSSE3-NEXT:    movq {{.*#+}} xmm1 = mem[0],zero
+; SSSE3-NEXT:    punpcklbw {{.*#+}} xmm1 = xmm1[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; SSSE3-NEXT:    psraw $8, %xmm1
 ; SSSE3-NEXT:    retq
 ;
-; SSE41-LABEL: load_sext_2i8_to_2i64:
+; SSE41-LABEL: load_sext_16i8_to_16i16:
 ; SSE41:       # BB#0: # %entry
-; SSE41-NEXT:    pmovsxbq (%rdi), %xmm0
+; SSE41-NEXT:    pmovsxbw (%rdi), %xmm0
+; SSE41-NEXT:    pmovsxbw 8(%rdi), %xmm1
 ; SSE41-NEXT:    retq
 ;
-; AVX-LABEL: load_sext_2i8_to_2i64:
-; AVX:       # BB#0: # %entry
-; AVX-NEXT:    vpmovsxbq (%rdi), %xmm0
-; AVX-NEXT:    retq
+; AVX1-LABEL: load_sext_16i8_to_16i16:
+; AVX1:       # BB#0: # %entry
+; AVX1-NEXT:    vpmovsxbw (%rdi), %xmm0
+; AVX1-NEXT:    vpmovsxbw 8(%rdi), %xmm1
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
 ;
-; X32-SSE41-LABEL: load_sext_2i8_to_2i64:
+; AVX2-LABEL: load_sext_16i8_to_16i16:
+; AVX2:       # BB#0: # %entry
+; AVX2-NEXT:    vpmovsxbw (%rdi), %ymm0
+; AVX2-NEXT:    retq
+;
+; X32-SSE41-LABEL: load_sext_16i8_to_16i16:
 ; X32-SSE41:       # BB#0: # %entry
 ; X32-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %eax
-; X32-SSE41-NEXT:    pmovsxbq (%eax), %xmm0
+; X32-SSE41-NEXT:    pmovsxbw (%eax), %xmm0
+; X32-SSE41-NEXT:    pmovsxbw 8(%eax), %xmm1
 ; X32-SSE41-NEXT:    retl
 entry:
- %X = load <2 x i8>, <2 x i8>* %ptr
- %Y = sext <2 x i8> %X to <2 x i64>
- ret <2 x i64>%Y
+ %X = load <16 x i8>, <16 x i8>* %ptr
+ %Y = sext <16 x i8> %X to <16 x i16>
+ ret <16 x i16> %Y
 }
 
 define <2 x i64> @load_sext_2i16_to_2i64(<2 x i16> *%ptr) {
@@ -781,20 +907,168 @@ define <2 x i64> @load_sext_2i16_to_2i64(<2 x i16> *%ptr) {
 ; SSE41-NEXT:    pmovsxwq (%rdi), %xmm0
 ; SSE41-NEXT:    retq
 ;
-; AVX-LABEL: load_sext_2i16_to_2i64:
-; AVX:       # BB#0: # %entry
-; AVX-NEXT:    vpmovsxwq (%rdi), %xmm0
-; AVX-NEXT:    retq
+; AVX-LABEL: load_sext_2i16_to_2i64:
+; AVX:       # BB#0: # %entry
+; AVX-NEXT:    vpmovsxwq (%rdi), %xmm0
+; AVX-NEXT:    retq
+;
+; X32-SSE41-LABEL: load_sext_2i16_to_2i64:
+; X32-SSE41:       # BB#0: # %entry
+; X32-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X32-SSE41-NEXT:    pmovsxwq (%eax), %xmm0
+; X32-SSE41-NEXT:    retl
+entry:
+ %X = load <2 x i16>, <2 x i16>* %ptr
+ %Y = sext <2 x i16> %X to <2 x i64>
+ ret <2 x i64> %Y
+}
+
+define <4 x i32> @load_sext_4i16_to_4i32(<4 x i16> *%ptr) {
+; SSE2-LABEL: load_sext_4i16_to_4i32:
+; SSE2:       # BB#0: # %entry
+; SSE2-NEXT:    movq {{.*#+}} xmm0 = mem[0],zero
+; 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_4i16_to_4i32:
+; SSSE3:       # BB#0: # %entry
+; SSSE3-NEXT:    movq {{.*#+}} xmm0 = mem[0],zero
+; 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_4i16_to_4i32:
+; SSE41:       # BB#0: # %entry
+; SSE41-NEXT:    pmovsxwd (%rdi), %xmm0
+; SSE41-NEXT:    retq
+;
+; AVX-LABEL: load_sext_4i16_to_4i32:
+; AVX:       # BB#0: # %entry
+; AVX-NEXT:    vpmovsxwd (%rdi), %xmm0
+; AVX-NEXT:    retq
+;
+; X32-SSE41-LABEL: load_sext_4i16_to_4i32:
+; X32-SSE41:       # BB#0: # %entry
+; X32-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X32-SSE41-NEXT:    pmovsxwd (%eax), %xmm0
+; X32-SSE41-NEXT:    retl
+entry:
+ %X = load <4 x i16>, <4 x i16>* %ptr
+ %Y = sext <4 x i16> %X to <4 x i32>
+ ret <4 x i32> %Y
+}
+
+define <4 x i64> @load_sext_4i16_to_4i64(<4 x i16> *%ptr) {
+; SSE2-LABEL: load_sext_4i16_to_4i64:
+; SSE2:       # BB#0: # %entry
+; 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:    movswq 6(%rdi), %rax
+; SSE2-NEXT:    movd %rax, %xmm2
+; SSE2-NEXT:    movswq 4(%rdi), %rax
+; SSE2-NEXT:    movd %rax, %xmm1
+; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm2[0]
+; SSE2-NEXT:    retq
+;
+; SSSE3-LABEL: load_sext_4i16_to_4i64:
+; SSSE3:       # BB#0: # %entry
+; 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:    movswq 6(%rdi), %rax
+; SSSE3-NEXT:    movd %rax, %xmm2
+; SSSE3-NEXT:    movswq 4(%rdi), %rax
+; SSSE3-NEXT:    movd %rax, %xmm1
+; SSSE3-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm2[0]
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: load_sext_4i16_to_4i64:
+; SSE41:       # BB#0: # %entry
+; SSE41-NEXT:    pmovsxwq (%rdi), %xmm0
+; SSE41-NEXT:    pmovsxwq 4(%rdi), %xmm1
+; SSE41-NEXT:    retq
+;
+; AVX1-LABEL: load_sext_4i16_to_4i64:
+; AVX1:       # BB#0: # %entry
+; AVX1-NEXT:    vpmovsxwd (%rdi), %xmm0
+; AVX1-NEXT:    vpmovsxdq %xmm0, %xmm1
+; AVX1-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[2,3,0,1]
+; AVX1-NEXT:    vpmovsxdq %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: load_sext_4i16_to_4i64:
+; AVX2:       # BB#0: # %entry
+; AVX2-NEXT:    vpmovsxwq (%rdi), %ymm0
+; AVX2-NEXT:    retq
+;
+; X32-SSE41-LABEL: load_sext_4i16_to_4i64:
+; X32-SSE41:       # BB#0: # %entry
+; X32-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X32-SSE41-NEXT:    pmovsxwq (%eax), %xmm0
+; X32-SSE41-NEXT:    pmovsxwq 4(%eax), %xmm1
+; X32-SSE41-NEXT:    retl
+entry:
+ %X = load <4 x i16>, <4 x i16>* %ptr
+ %Y = sext <4 x i16> %X to <4 x i64>
+ ret <4 x i64> %Y
+}
+
+define <8 x i32> @load_sext_8i16_to_8i32(<8 x i16> *%ptr) {
+; SSE2-LABEL: load_sext_8i16_to_8i32:
+; SSE2:       # BB#0: # %entry
+; SSE2-NEXT:    movq {{.*#+}} xmm0 = mem[0],zero
+; SSE2-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3]
+; SSE2-NEXT:    psrad $16, %xmm0
+; SSE2-NEXT:    movq {{.*#+}} xmm1 = mem[0],zero
+; SSE2-NEXT:    punpcklwd {{.*#+}} xmm1 = xmm1[0,0,1,1,2,2,3,3]
+; SSE2-NEXT:    psrad $16, %xmm1
+; SSE2-NEXT:    retq
+;
+; SSSE3-LABEL: load_sext_8i16_to_8i32:
+; SSSE3:       # BB#0: # %entry
+; SSSE3-NEXT:    movq {{.*#+}} xmm0 = mem[0],zero
+; SSSE3-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3]
+; SSSE3-NEXT:    psrad $16, %xmm0
+; SSSE3-NEXT:    movq {{.*#+}} xmm1 = mem[0],zero
+; SSSE3-NEXT:    punpcklwd {{.*#+}} xmm1 = xmm1[0,0,1,1,2,2,3,3]
+; SSSE3-NEXT:    psrad $16, %xmm1
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: load_sext_8i16_to_8i32:
+; SSE41:       # BB#0: # %entry
+; SSE41-NEXT:    pmovsxwd (%rdi), %xmm0
+; SSE41-NEXT:    pmovsxwd 8(%rdi), %xmm1
+; SSE41-NEXT:    retq
+;
+; AVX1-LABEL: load_sext_8i16_to_8i32:
+; AVX1:       # BB#0: # %entry
+; AVX1-NEXT:    vpmovsxwd (%rdi), %xmm0
+; AVX1-NEXT:    vpmovsxwd 8(%rdi), %xmm1
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: load_sext_8i16_to_8i32:
+; AVX2:       # BB#0: # %entry
+; AVX2-NEXT:    vpmovsxwd (%rdi), %ymm0
+; AVX2-NEXT:    retq
 ;
-; X32-SSE41-LABEL: load_sext_2i16_to_2i64:
+; X32-SSE41-LABEL: load_sext_8i16_to_8i32:
 ; X32-SSE41:       # BB#0: # %entry
 ; X32-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %eax
-; X32-SSE41-NEXT:    pmovsxwq (%eax), %xmm0
+; X32-SSE41-NEXT:    pmovsxwd (%eax), %xmm0
+; X32-SSE41-NEXT:    pmovsxwd 8(%eax), %xmm1
 ; X32-SSE41-NEXT:    retl
 entry:
- %X = load <2 x i16>, <2 x i16>* %ptr
- %Y = sext <2 x i16> %X to <2 x i64>
- ret <2 x i64>%Y
+ %X = load <8 x i16>, <8 x i16>* %ptr
+ %Y = sext <8 x i16> %X to <8 x i32>
+ ret <8 x i32> %Y
 }
 
 define <2 x i64> @load_sext_2i32_to_2i64(<2 x i32> *%ptr) {
@@ -832,43 +1106,105 @@ define <2 x i64> @load_sext_2i32_to_2i64(<2 x i32> *%ptr) {
 entry:
  %X = load <2 x i32>, <2 x i32>* %ptr
  %Y = sext <2 x i32> %X to <2 x i64>
- ret <2 x i64>%Y
+ ret <2 x i64> %Y
 }
 
-define <8 x i16> @load_sext_8i8_to_8i16(<8 x i8> *%ptr) {
-; SSE2-LABEL: load_sext_8i8_to_8i16:
+define <4 x i64> @load_sext_4i32_to_4i64(<4 x i32> *%ptr) {
+; SSE2-LABEL: load_sext_4i32_to_4i64:
+; SSE2:       # BB#0: # %entry
+; SSE2-NEXT:    movdqa (%rdi), %xmm0
+; SSE2-NEXT:    movdqa %xmm0, %xmm2
+; SSE2-NEXT:    psrad $31, %xmm2
+; SSE2-NEXT:    pshufd {{.*#+}} xmm1 = xmm0[2,3,0,1]
+; SSE2-NEXT:    punpckldq {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1]
+; SSE2-NEXT:    movdqa %xmm1, %xmm2
+; SSE2-NEXT:    psrad $31, %xmm2
+; SSE2-NEXT:    punpckldq {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1]
+; SSE2-NEXT:    retq
+;
+; SSSE3-LABEL: load_sext_4i32_to_4i64:
+; SSSE3:       # BB#0: # %entry
+; SSSE3-NEXT:    movdqa (%rdi), %xmm0
+; SSSE3-NEXT:    movdqa %xmm0, %xmm2
+; SSSE3-NEXT:    psrad $31, %xmm2
+; SSSE3-NEXT:    pshufd {{.*#+}} xmm1 = xmm0[2,3,0,1]
+; SSSE3-NEXT:    punpckldq {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1]
+; SSSE3-NEXT:    movdqa %xmm1, %xmm2
+; SSSE3-NEXT:    psrad $31, %xmm2
+; SSSE3-NEXT:    punpckldq {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1]
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: load_sext_4i32_to_4i64:
+; SSE41:       # BB#0: # %entry
+; SSE41-NEXT:    pmovsxdq (%rdi), %xmm0
+; SSE41-NEXT:    pmovsxdq 8(%rdi), %xmm1
+; SSE41-NEXT:    retq
+;
+; AVX1-LABEL: load_sext_4i32_to_4i64:
+; AVX1:       # BB#0: # %entry
+; AVX1-NEXT:    vpmovsxdq (%rdi), %xmm0
+; AVX1-NEXT:    vpmovsxdq 8(%rdi), %xmm1
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: load_sext_4i32_to_4i64:
+; AVX2:       # BB#0: # %entry
+; AVX2-NEXT:    vpmovsxdq (%rdi), %ymm0
+; AVX2-NEXT:    retq
+;
+; X32-SSE41-LABEL: load_sext_4i32_to_4i64:
+; X32-SSE41:       # BB#0: # %entry
+; X32-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X32-SSE41-NEXT:    pmovsxdq (%eax), %xmm0
+; X32-SSE41-NEXT:    pmovsxdq 8(%eax), %xmm1
+; X32-SSE41-NEXT:    retl
+entry:
+ %X = load <4 x i32>, <4 x i32>* %ptr
+ %Y = sext <4 x i32> %X to <4 x i64>
+ ret <4 x i64> %Y
+}
+
+define i32 @sext_2i8_to_i32(<16 x i8> %A) nounwind uwtable readnone ssp {
+; SSE2-LABEL: sext_2i8_to_i32:
 ; SSE2:       # BB#0: # %entry
-; SSE2-NEXT:    movq {{.*#+}} xmm0 = mem[0],zero
 ; 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:    movd %xmm0, %eax
 ; SSE2-NEXT:    retq
 ;
-; SSSE3-LABEL: load_sext_8i8_to_8i16:
+; SSSE3-LABEL: sext_2i8_to_i32:
 ; SSSE3:       # BB#0: # %entry
-; SSSE3-NEXT:    movq {{.*#+}} xmm0 = mem[0],zero
 ; 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:    movd %xmm0, %eax
 ; SSSE3-NEXT:    retq
 ;
-; SSE41-LABEL: load_sext_8i8_to_8i16:
+; SSE41-LABEL: sext_2i8_to_i32:
 ; SSE41:       # BB#0: # %entry
-; SSE41-NEXT:    pmovsxbw (%rdi), %xmm0
+; SSE41-NEXT:    pmovsxbw %xmm0, %xmm0
+; SSE41-NEXT:    movd %xmm0, %eax
 ; SSE41-NEXT:    retq
 ;
-; AVX-LABEL: load_sext_8i8_to_8i16:
+; AVX-LABEL: sext_2i8_to_i32:
 ; AVX:       # BB#0: # %entry
-; AVX-NEXT:    vpmovsxbw (%rdi), %xmm0
+; AVX-NEXT:    vpmovsxbw %xmm0, %xmm0
+; AVX-NEXT:    vmovd %xmm0, %eax
 ; AVX-NEXT:    retq
 ;
-; X32-SSE41-LABEL: load_sext_8i8_to_8i16:
+; X32-SSE41-LABEL: sext_2i8_to_i32:
 ; X32-SSE41:       # BB#0: # %entry
-; X32-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %eax
-; X32-SSE41-NEXT:    pmovsxbw (%eax), %xmm0
+; X32-SSE41-NEXT:    pushl %eax
+; X32-SSE41-NEXT:  .Ltmp0:
+; X32-SSE41-NEXT:    .cfi_def_cfa_offset 8
+; X32-SSE41-NEXT:    pmovsxbw %xmm0, %xmm0
+; X32-SSE41-NEXT:    movd %xmm0, %eax
+; X32-SSE41-NEXT:    popl %edx
 ; X32-SSE41-NEXT:    retl
 entry:
- %X = load <8 x i8>, <8 x i8>* %ptr
- %Y = sext <8 x i8> %X to <8 x i16>
- ret <8 x i16>%Y
+  %Shuf = shufflevector <16 x i8> %A, <16 x i8> undef, <2 x i32> <i32 0, i32 1>
+  %Ex = sext <2 x i8> %Shuf to <2 x i16>
+  %Bc = bitcast <2 x i16> %Ex to i32
+  ret i32 %Bc
 }
 
 define <4 x i64> @sext_4i1_to_4i64(<4 x i1> %mask) {
@@ -938,57 +1274,6 @@ define <4 x i64> @sext_4i1_to_4i64(<4 x i1> %mask) {
   ret <4 x i64> %extmask
 }
 
-define <16 x i16> @load_sext_16i8_to_16i16(<16 x i8> *%ptr) {
-; SSE2-LABEL: load_sext_16i8_to_16i16:
-; SSE2:       # BB#0: # %entry
-; SSE2-NEXT:    movq {{.*#+}} xmm0 = mem[0],zero
-; 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:    movq {{.*#+}} xmm1 = mem[0],zero
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm1 = xmm1[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
-; SSE2-NEXT:    psraw $8, %xmm1
-; SSE2-NEXT:    retq
-;
-; SSSE3-LABEL: load_sext_16i8_to_16i16:
-; SSSE3:       # BB#0: # %entry
-; SSSE3-NEXT:    movq {{.*#+}} xmm0 = mem[0],zero
-; 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:    movq {{.*#+}} xmm1 = mem[0],zero
-; SSSE3-NEXT:    punpcklbw {{.*#+}} xmm1 = xmm1[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
-; SSSE3-NEXT:    psraw $8, %xmm1
-; SSSE3-NEXT:    retq
-;
-; SSE41-LABEL: load_sext_16i8_to_16i16:
-; SSE41:       # BB#0: # %entry
-; SSE41-NEXT:    pmovsxbw (%rdi), %xmm0
-; SSE41-NEXT:    pmovsxbw 8(%rdi), %xmm1
-; SSE41-NEXT:    retq
-;
-; AVX1-LABEL: load_sext_16i8_to_16i16:
-; AVX1:       # BB#0: # %entry
-; AVX1-NEXT:    vpmovsxbw (%rdi), %xmm0
-; AVX1-NEXT:    vpmovsxbw 8(%rdi), %xmm1
-; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
-; AVX1-NEXT:    retq
-;
-; AVX2-LABEL: load_sext_16i8_to_16i16:
-; AVX2:       # BB#0: # %entry
-; AVX2-NEXT:    vpmovsxbw (%rdi), %ymm0
-; AVX2-NEXT:    retq
-;
-; X32-SSE41-LABEL: load_sext_16i8_to_16i16:
-; X32-SSE41:       # BB#0: # %entry
-; X32-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %eax
-; X32-SSE41-NEXT:    pmovsxbw (%eax), %xmm0
-; X32-SSE41-NEXT:    pmovsxbw 8(%eax), %xmm1
-; X32-SSE41-NEXT:    retl
-entry:
- %X = load <16 x i8>, <16 x i8>* %ptr
- %Y = sext <16 x i8> %X to <16 x i16>
- ret <16 x i16> %Y
-}
-
 define <4 x i64> @sext_4i8_to_4i64(<4 x i8> %mask) {
 ; SSE2-LABEL: sext_4i8_to_4i64:
 ; SSE2:       # BB#0:
@@ -1055,125 +1340,3 @@ define <4 x i64> @sext_4i8_to_4i64(<4 x i8> %mask) {
   %extmask = sext <4 x i8> %mask to <4 x i64>
   ret <4 x i64> %extmask
 }
-
-define <4 x i64> @load_sext_4i8_to_4i64(<4 x i8> *%ptr) {
-; SSE2-LABEL: load_sext_4i8_to_4i64:
-; SSE2:       # BB#0: # %entry
-; 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:    movsbq 3(%rdi), %rax
-; SSE2-NEXT:    movd %rax, %xmm2
-; SSE2-NEXT:    movsbq 2(%rdi), %rax
-; SSE2-NEXT:    movd %rax, %xmm1
-; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm2[0]
-; SSE2-NEXT:    retq
-;
-; SSSE3-LABEL: load_sext_4i8_to_4i64:
-; SSSE3:       # BB#0: # %entry
-; 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:    movsbq 3(%rdi), %rax
-; SSSE3-NEXT:    movd %rax, %xmm2
-; SSSE3-NEXT:    movsbq 2(%rdi), %rax
-; SSSE3-NEXT:    movd %rax, %xmm1
-; SSSE3-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm2[0]
-; SSSE3-NEXT:    retq
-;
-; SSE41-LABEL: load_sext_4i8_to_4i64:
-; SSE41:       # BB#0: # %entry
-; SSE41-NEXT:    pmovsxbq (%rdi), %xmm0
-; SSE41-NEXT:    pmovsxbq 2(%rdi), %xmm1
-; SSE41-NEXT:    retq
-;
-; AVX1-LABEL: load_sext_4i8_to_4i64:
-; AVX1:       # BB#0: # %entry
-; AVX1-NEXT:    vpmovsxbd (%rdi), %xmm0
-; AVX1-NEXT:    vpmovsxdq %xmm0, %xmm1
-; AVX1-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[2,3,0,1]
-; AVX1-NEXT:    vpmovsxdq %xmm0, %xmm0
-; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
-; AVX1-NEXT:    retq
-;
-; AVX2-LABEL: load_sext_4i8_to_4i64:
-; AVX2:       # BB#0: # %entry
-; AVX2-NEXT:    vpmovsxbq (%rdi), %ymm0
-; AVX2-NEXT:    retq
-;
-; X32-SSE41-LABEL: load_sext_4i8_to_4i64:
-; X32-SSE41:       # BB#0: # %entry
-; X32-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %eax
-; X32-SSE41-NEXT:    pmovsxbq (%eax), %xmm0
-; X32-SSE41-NEXT:    pmovsxbq 2(%eax), %xmm1
-; X32-SSE41-NEXT:    retl
-entry:
- %X = load <4 x i8>, <4 x i8>* %ptr
- %Y = sext <4 x i8> %X to <4 x i64>
- ret <4 x i64>%Y
-}
-
-define <4 x i64> @load_sext_4i16_to_4i64(<4 x i16> *%ptr) {
-; SSE2-LABEL: load_sext_4i16_to_4i64:
-; SSE2:       # BB#0: # %entry
-; 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:    movswq 6(%rdi), %rax
-; SSE2-NEXT:    movd %rax, %xmm2
-; SSE2-NEXT:    movswq 4(%rdi), %rax
-; SSE2-NEXT:    movd %rax, %xmm1
-; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm2[0]
-; SSE2-NEXT:    retq
-;
-; SSSE3-LABEL: load_sext_4i16_to_4i64:
-; SSSE3:       # BB#0: # %entry
-; 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:    movswq 6(%rdi), %rax
-; SSSE3-NEXT:    movd %rax, %xmm2
-; SSSE3-NEXT:    movswq 4(%rdi), %rax
-; SSSE3-NEXT:    movd %rax, %xmm1
-; SSSE3-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm2[0]
-; SSSE3-NEXT:    retq
-;
-; SSE41-LABEL: load_sext_4i16_to_4i64:
-; SSE41:       # BB#0: # %entry
-; SSE41-NEXT:    pmovsxwq (%rdi), %xmm0
-; SSE41-NEXT:    pmovsxwq 4(%rdi), %xmm1
-; SSE41-NEXT:    retq
-;
-; AVX1-LABEL: load_sext_4i16_to_4i64:
-; AVX1:       # BB#0: # %entry
-; AVX1-NEXT:    vpmovsxwd (%rdi), %xmm0
-; AVX1-NEXT:    vpmovsxdq %xmm0, %xmm1
-; AVX1-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[2,3,0,1]
-; AVX1-NEXT:    vpmovsxdq %xmm0, %xmm0
-; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
-; AVX1-NEXT:    retq
-;
-; AVX2-LABEL: load_sext_4i16_to_4i64:
-; AVX2:       # BB#0: # %entry
-; AVX2-NEXT:    vpmovsxwq (%rdi), %ymm0
-; AVX2-NEXT:    retq
-;
-; X32-SSE41-LABEL: load_sext_4i16_to_4i64:
-; X32-SSE41:       # BB#0: # %entry
-; X32-SSE41-NEXT:    movl {{[0-9]+}}(%esp), %eax
-; X32-SSE41-NEXT:    pmovsxwq (%eax), %xmm0
-; X32-SSE41-NEXT:    pmovsxwq 4(%eax), %xmm1
-; X32-SSE41-NEXT:    retl
-entry:
- %X = load <4 x i16>, <4 x i16>* %ptr
- %Y = sext <4 x i16> %X to <4 x i64>
- ret <4 x i64>%Y
-}
diff --git a/test/CodeGen/X86/vector-zext.ll b/test/CodeGen/X86/vector-zext.ll
index b17799fd8c8..77435fd3cac 100644
--- a/test/CodeGen/X86/vector-zext.ll
+++ b/test/CodeGen/X86/vector-zext.ll
@@ -489,56 +489,195 @@ entry:
   ret <4 x i64>%B
 }
 
-define <8 x i32> @zext_8i8_to_8i32(<8 x i8> %z) {
-; SSE2-LABEL: zext_8i8_to_8i32:
+define <2 x i64> @load_zext_2i8_to_2i64(<2 x i8> *%ptr) {
+; SSE2-LABEL: load_zext_2i8_to_2i64:
 ; 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:    movzwl (%rdi), %eax
+; SSE2-NEXT:    movd %eax, %xmm0
+; SSE2-NEXT:    pxor %xmm1, %xmm1
+; SSE2-NEXT:    punpcklbw {{.*#+}} 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]
+; SSE2-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; SSE2-NEXT:    punpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
 ; SSE2-NEXT:    retq
 ;
-; SSSE3-LABEL: zext_8i8_to_8i32:
+; SSSE3-LABEL: load_zext_2i8_to_2i64:
 ; 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:    movzwl (%rdi), %eax
+; SSSE3-NEXT:    movd %eax, %xmm0
+; SSSE3-NEXT:    pshufb {{.*#+}} xmm0 = xmm0[0],zero,zero,zero,zero,zero,zero,zero,xmm0[1],zero,zero,zero,zero,zero,zero,zero
 ; SSSE3-NEXT:    retq
 ;
-; SSE41-LABEL: zext_8i8_to_8i32:
+; SSE41-LABEL: load_zext_2i8_to_2i64:
 ; SSE41:       # BB#0: # %entry
-; SSE41-NEXT:    pmovzxwd {{.*#+}} xmm2 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero
-; 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:    pmovzxbq {{.*#+}} xmm0 = mem[0],zero,zero,zero,zero,zero,zero,zero,mem[1],zero,zero,zero,zero,zero,zero,zero
 ; SSE41-NEXT:    retq
 ;
-; AVX1-LABEL: zext_8i8_to_8i32:
+; AVX-LABEL: load_zext_2i8_to_2i64:
+; AVX:       # BB#0: # %entry
+; AVX-NEXT:    vpmovzxbq {{.*#+}} xmm0 = mem[0],zero,zero,zero,zero,zero,zero,zero,mem[1],zero,zero,zero,zero,zero,zero,zero
+; AVX-NEXT:    retq
+entry:
+ %X = load <2 x i8>, <2 x i8>* %ptr
+ %Y = zext <2 x i8> %X to <2 x i64>
+ ret <2 x i64> %Y
+}
+
+define <4 x i32> @load_zext_4i8_to_4i32(<4 x i8> *%ptr) {
+; SSE2-LABEL: load_zext_4i8_to_4i32:
+; SSE2:       # BB#0: # %entry
+; SSE2-NEXT:    movd {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; SSE2-NEXT:    pxor %xmm1, %xmm1
+; SSE2-NEXT:    punpcklbw {{.*#+}} 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]
+; SSE2-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; SSE2-NEXT:    retq
+;
+; SSSE3-LABEL: load_zext_4i8_to_4i32:
+; SSSE3:       # BB#0: # %entry
+; SSSE3-NEXT:    movd {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; SSSE3-NEXT:    pxor %xmm1, %xmm1
+; SSSE3-NEXT:    punpcklbw {{.*#+}} 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]
+; SSSE3-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: load_zext_4i8_to_4i32:
+; SSE41:       # BB#0: # %entry
+; SSE41-NEXT:    pmovzxbd {{.*#+}} xmm0 = mem[0],zero,zero,zero,mem[1],zero,zero,zero,mem[2],zero,zero,zero,mem[3],zero,zero,zero
+; SSE41-NEXT:    retq
+;
+; AVX-LABEL: load_zext_4i8_to_4i32:
+; AVX:       # BB#0: # %entry
+; AVX-NEXT:    vpmovzxbd {{.*#+}} xmm0 = mem[0],zero,zero,zero,mem[1],zero,zero,zero,mem[2],zero,zero,zero,mem[3],zero,zero,zero
+; AVX-NEXT:    retq
+entry:
+ %X = load <4 x i8>, <4 x i8>* %ptr
+ %Y = zext <4 x i8> %X to <4 x i32>
+ ret <4 x i32> %Y
+}
+
+define <4 x i64> @load_zext_4i8_to_4i64(<4 x i8> *%ptr) {
+; SSE2-LABEL: load_zext_4i8_to_4i64:
+; SSE2:       # BB#0: # %entry
+; SSE2-NEXT:    movd {{.*#+}} xmm1 = mem[0],zero,zero,zero
+; SSE2-NEXT:    punpcklbw {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1],xmm1[2],xmm0[2],xmm1[3],xmm0[3],xmm1[4],xmm0[4],xmm1[5],xmm0[5],xmm1[6],xmm0[6],xmm1[7],xmm0[7]
+; SSE2-NEXT:    punpcklwd {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1],xmm1[2],xmm0[2],xmm1[3],xmm0[3]
+; SSE2-NEXT:    pshufd {{.*#+}} xmm0 = xmm1[0,1,1,3]
+; SSE2-NEXT:    movdqa {{.*#+}} xmm2 = [255,255]
+; SSE2-NEXT:    pand %xmm2, %xmm0
+; SSE2-NEXT:    pshufd {{.*#+}} xmm1 = xmm1[2,2,3,3]
+; SSE2-NEXT:    pand %xmm2, %xmm1
+; SSE2-NEXT:    retq
+;
+; SSSE3-LABEL: load_zext_4i8_to_4i64:
+; SSSE3:       # BB#0: # %entry
+; SSSE3-NEXT:    movd {{.*#+}} xmm1 = mem[0],zero,zero,zero
+; SSSE3-NEXT:    punpcklbw {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1],xmm1[2],xmm0[2],xmm1[3],xmm0[3],xmm1[4],xmm0[4],xmm1[5],xmm0[5],xmm1[6],xmm0[6],xmm1[7],xmm0[7]
+; SSSE3-NEXT:    punpcklwd {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1],xmm1[2],xmm0[2],xmm1[3],xmm0[3]
+; SSSE3-NEXT:    pshufd {{.*#+}} xmm0 = xmm1[0,1,1,3]
+; SSSE3-NEXT:    movdqa {{.*#+}} xmm2 = [255,255]
+; SSSE3-NEXT:    pand %xmm2, %xmm0
+; SSSE3-NEXT:    pshufd {{.*#+}} xmm1 = xmm1[2,2,3,3]
+; SSSE3-NEXT:    pand %xmm2, %xmm1
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: load_zext_4i8_to_4i64:
+; SSE41:       # BB#0: # %entry
+; SSE41-NEXT:    pmovzxbq {{.*#+}} xmm0 = mem[0],zero,zero,zero,zero,zero,zero,zero,mem[1],zero,zero,zero,zero,zero,zero,zero
+; SSE41-NEXT:    pmovzxbq {{.*#+}} xmm1 = mem[0],zero,zero,zero,zero,zero,zero,zero,mem[1],zero,zero,zero,zero,zero,zero,zero
+; SSE41-NEXT:    retq
+;
+; AVX1-LABEL: load_zext_4i8_to_4i64:
 ; AVX1:       # BB#0: # %entry
-; AVX1-NEXT:    vpmovzxwd {{.*#+}} xmm1 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero
-; AVX1-NEXT:    vpunpckhwd {{.*#+}} xmm0 = xmm0[4,4,5,5,6,6,7,7]
-; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
-; AVX1-NEXT:    vandps {{.*}}(%rip), %ymm0, %ymm0
+; AVX1-NEXT:    vpmovzxbq {{.*#+}} xmm0 = mem[0],zero,zero,zero,zero,zero,zero,zero,mem[1],zero,zero,zero,zero,zero,zero,zero
+; AVX1-NEXT:    vpmovzxbq {{.*#+}} xmm1 = mem[0],zero,zero,zero,zero,zero,zero,zero,mem[1],zero,zero,zero,zero,zero,zero,zero
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
 ; AVX1-NEXT:    retq
 ;
-; AVX2-LABEL: zext_8i8_to_8i32:
+; AVX2-LABEL: load_zext_4i8_to_4i64:
 ; AVX2:       # BB#0: # %entry
-; AVX2-NEXT:    vpmovzxwd {{.*#+}} ymm0 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero,xmm0[4],zero,xmm0[5],zero,xmm0[6],zero,xmm0[7],zero
-; AVX2-NEXT:    vpbroadcastd {{.*}}(%rip), %ymm1
-; AVX2-NEXT:    vpand %ymm1, %ymm0, %ymm0
+; AVX2-NEXT:    vpmovzxbq {{.*#+}} ymm0 = mem[0],zero,zero,zero,zero,zero,zero,zero,mem[1],zero,zero,zero,zero,zero,zero,zero,mem[2],zero,zero,zero,zero,zero,zero,zero,mem[3],zero,zero,zero,zero,zero,zero,zero
 ; AVX2-NEXT:    retq
 entry:
-  %t = zext <8 x i8> %z to <8 x i32>
-  ret <8 x i32> %t
+ %X = load <4 x i8>, <4 x i8>* %ptr
+ %Y = zext <4 x i8> %X to <4 x i64>
+ ret <4 x i64> %Y
+}
+
+define <8 x i16> @load_zext_8i8_to_8i16(<8 x i8> *%ptr) {
+; SSE2-LABEL: load_zext_8i8_to_8i16:
+; SSE2:       # BB#0: # %entry
+; SSE2-NEXT:    movq {{.*#+}} xmm0 = mem[0],zero
+; SSE2-NEXT:    pxor %xmm1, %xmm1
+; SSE2-NEXT:    punpcklbw {{.*#+}} 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]
+; SSE2-NEXT:    retq
+;
+; SSSE3-LABEL: load_zext_8i8_to_8i16:
+; SSSE3:       # BB#0: # %entry
+; SSSE3-NEXT:    movq {{.*#+}} xmm0 = mem[0],zero
+; SSSE3-NEXT:    pxor %xmm1, %xmm1
+; SSSE3-NEXT:    punpcklbw {{.*#+}} 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]
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: load_zext_8i8_to_8i16:
+; SSE41:       # BB#0: # %entry
+; SSE41-NEXT:    pmovzxbw {{.*#+}} xmm0 = mem[0],zero,mem[1],zero,mem[2],zero,mem[3],zero,mem[4],zero,mem[5],zero,mem[6],zero,mem[7],zero
+; SSE41-NEXT:    retq
+;
+; AVX-LABEL: load_zext_8i8_to_8i16:
+; AVX:       # BB#0: # %entry
+; AVX-NEXT:    vpmovzxbw {{.*#+}} xmm0 = mem[0],zero,mem[1],zero,mem[2],zero,mem[3],zero,mem[4],zero,mem[5],zero,mem[6],zero,mem[7],zero
+; AVX-NEXT:    retq
+entry:
+ %X = load <8 x i8>, <8 x i8>* %ptr
+ %Y = zext <8 x i8> %X to <8 x i16>
+ ret <8 x i16> %Y
+}
+
+define <8 x i32> @load_zext_8i8_to_8i32(<8 x i8> *%ptr) {
+; SSE2-LABEL: load_zext_8i8_to_8i32:
+; SSE2:       # BB#0: # %entry
+; SSE2-NEXT:    movq {{.*#+}} xmm1 = mem[0],zero
+; SSE2-NEXT:    punpcklbw {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1],xmm1[2],xmm0[2],xmm1[3],xmm0[3],xmm1[4],xmm0[4],xmm1[5],xmm0[5],xmm1[6],xmm0[6],xmm1[7],xmm0[7]
+; SSE2-NEXT:    movdqa %xmm1, %xmm0
+; SSE2-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3]
+; SSE2-NEXT:    movdqa {{.*#+}} xmm2 = [255,255,255,255]
+; SSE2-NEXT:    pand %xmm2, %xmm0
+; SSE2-NEXT:    punpckhwd {{.*#+}} xmm1 = xmm1[4,4,5,5,6,6,7,7]
+; SSE2-NEXT:    pand %xmm2, %xmm1
+; SSE2-NEXT:    retq
+;
+; SSSE3-LABEL: load_zext_8i8_to_8i32:
+; SSSE3:       # BB#0: # %entry
+; SSSE3-NEXT:    movq {{.*#+}} xmm1 = mem[0],zero
+; SSSE3-NEXT:    punpcklbw {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1],xmm1[2],xmm0[2],xmm1[3],xmm0[3],xmm1[4],xmm0[4],xmm1[5],xmm0[5],xmm1[6],xmm0[6],xmm1[7],xmm0[7]
+; SSSE3-NEXT:    movdqa %xmm1, %xmm0
+; SSSE3-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3]
+; SSSE3-NEXT:    movdqa {{.*#+}} xmm2 = [255,255,255,255]
+; SSSE3-NEXT:    pand %xmm2, %xmm0
+; SSSE3-NEXT:    punpckhwd {{.*#+}} xmm1 = xmm1[4,4,5,5,6,6,7,7]
+; SSSE3-NEXT:    pand %xmm2, %xmm1
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: load_zext_8i8_to_8i32:
+; SSE41:       # BB#0: # %entry
+; SSE41-NEXT:    pmovzxbd {{.*#+}} xmm0 = mem[0],zero,zero,zero,mem[1],zero,zero,zero,mem[2],zero,zero,zero,mem[3],zero,zero,zero
+; SSE41-NEXT:    pmovzxbd {{.*#+}} xmm1 = mem[0],zero,zero,zero,mem[1],zero,zero,zero,mem[2],zero,zero,zero,mem[3],zero,zero,zero
+; SSE41-NEXT:    retq
+;
+; AVX1-LABEL: load_zext_8i8_to_8i32:
+; AVX1:       # BB#0: # %entry
+; AVX1-NEXT:    vpmovzxbd {{.*#+}} xmm0 = mem[0],zero,zero,zero,mem[1],zero,zero,zero,mem[2],zero,zero,zero,mem[3],zero,zero,zero
+; AVX1-NEXT:    vpmovzxbd {{.*#+}} xmm1 = mem[0],zero,zero,zero,mem[1],zero,zero,zero,mem[2],zero,zero,zero,mem[3],zero,zero,zero
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: load_zext_8i8_to_8i32:
+; AVX2:       # BB#0: # %entry
+; AVX2-NEXT:    vpmovzxbd {{.*#+}} ymm0 = mem[0],zero,zero,zero,mem[1],zero,zero,zero,mem[2],zero,zero,zero,mem[3],zero,zero,zero,mem[4],zero,zero,zero,mem[5],zero,zero,zero,mem[6],zero,zero,zero,mem[7],zero,zero,zero
+; AVX2-NEXT:    retq
+entry:
+ %X = load <8 x i8>, <8 x i8>* %ptr
+ %Y = zext <8 x i8> %X to <8 x i32>
+ ret <8 x i32> %Y
 }
 
 define <16 x i16> @load_zext_16i8_to_16i16(<16 x i8> *%ptr) {
@@ -585,6 +724,114 @@ entry:
  ret <16 x i16> %Y
 }
 
+define <2 x i64> @load_zext_2i16_to_2i64(<2 x i16> *%ptr) {
+; SSE2-LABEL: load_zext_2i16_to_2i64:
+; SSE2:       # BB#0: # %entry
+; SSE2-NEXT:    movd {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; SSE2-NEXT:    pxor %xmm1, %xmm1
+; SSE2-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; SSE2-NEXT:    punpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; SSE2-NEXT:    retq
+;
+; SSSE3-LABEL: load_zext_2i16_to_2i64:
+; SSSE3:       # BB#0: # %entry
+; SSSE3-NEXT:    movd {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; SSSE3-NEXT:    pxor %xmm1, %xmm1
+; SSSE3-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; SSSE3-NEXT:    punpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: load_zext_2i16_to_2i64:
+; SSE41:       # BB#0: # %entry
+; SSE41-NEXT:    pmovzxwq {{.*#+}} xmm0 = mem[0],zero,zero,zero,mem[1],zero,zero,zero
+; SSE41-NEXT:    retq
+;
+; AVX-LABEL: load_zext_2i16_to_2i64:
+; AVX:       # BB#0: # %entry
+; AVX-NEXT:    vpmovzxwq {{.*#+}} xmm0 = mem[0],zero,zero,zero,mem[1],zero,zero,zero
+; AVX-NEXT:    retq
+entry:
+ %X = load <2 x i16>, <2 x i16>* %ptr
+ %Y = zext <2 x i16> %X to <2 x i64>
+ ret <2 x i64> %Y
+}
+
+define <4 x i32> @load_zext_4i16_to_4i32(<4 x i16> *%ptr) {
+; SSE2-LABEL: load_zext_4i16_to_4i32:
+; SSE2:       # BB#0: # %entry
+; SSE2-NEXT:    movq {{.*#+}} xmm0 = mem[0],zero
+; SSE2-NEXT:    pxor %xmm1, %xmm1
+; SSE2-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; SSE2-NEXT:    retq
+;
+; SSSE3-LABEL: load_zext_4i16_to_4i32:
+; SSSE3:       # BB#0: # %entry
+; SSSE3-NEXT:    movq {{.*#+}} xmm0 = mem[0],zero
+; SSSE3-NEXT:    pxor %xmm1, %xmm1
+; SSSE3-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: load_zext_4i16_to_4i32:
+; SSE41:       # BB#0: # %entry
+; SSE41-NEXT:    pmovzxwd {{.*#+}} xmm0 = mem[0],zero,mem[1],zero,mem[2],zero,mem[3],zero
+; SSE41-NEXT:    retq
+;
+; AVX-LABEL: load_zext_4i16_to_4i32:
+; AVX:       # BB#0: # %entry
+; AVX-NEXT:    vpmovzxwd {{.*#+}} xmm0 = mem[0],zero,mem[1],zero,mem[2],zero,mem[3],zero
+; AVX-NEXT:    retq
+entry:
+ %X = load <4 x i16>, <4 x i16>* %ptr
+ %Y = zext <4 x i16> %X to <4 x i32>
+ ret <4 x i32> %Y
+}
+
+define <4 x i64> @load_zext_4i16_to_4i64(<4 x i16> *%ptr) {
+; SSE2-LABEL: load_zext_4i16_to_4i64:
+; SSE2:       # BB#0: # %entry
+; SSE2-NEXT:    movq {{.*#+}} xmm1 = mem[0],zero
+; SSE2-NEXT:    punpcklwd {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1],xmm1[2],xmm0[2],xmm1[3],xmm0[3]
+; SSE2-NEXT:    pshufd {{.*#+}} xmm0 = xmm1[0,1,1,3]
+; SSE2-NEXT:    movdqa {{.*#+}} xmm2 = [65535,65535]
+; SSE2-NEXT:    pand %xmm2, %xmm0
+; SSE2-NEXT:    pshufd {{.*#+}} xmm1 = xmm1[2,2,3,3]
+; SSE2-NEXT:    pand %xmm2, %xmm1
+; SSE2-NEXT:    retq
+;
+; SSSE3-LABEL: load_zext_4i16_to_4i64:
+; SSSE3:       # BB#0: # %entry
+; SSSE3-NEXT:    movq {{.*#+}} xmm1 = mem[0],zero
+; SSSE3-NEXT:    punpcklwd {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1],xmm1[2],xmm0[2],xmm1[3],xmm0[3]
+; SSSE3-NEXT:    pshufd {{.*#+}} xmm0 = xmm1[0,1,1,3]
+; SSSE3-NEXT:    movdqa {{.*#+}} xmm2 = [65535,65535]
+; SSSE3-NEXT:    pand %xmm2, %xmm0
+; SSSE3-NEXT:    pshufd {{.*#+}} xmm1 = xmm1[2,2,3,3]
+; SSSE3-NEXT:    pand %xmm2, %xmm1
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: load_zext_4i16_to_4i64:
+; SSE41:       # BB#0: # %entry
+; SSE41-NEXT:    pmovzxwq {{.*#+}} xmm0 = mem[0],zero,zero,zero,mem[1],zero,zero,zero
+; SSE41-NEXT:    pmovzxwq {{.*#+}} xmm1 = mem[0],zero,zero,zero,mem[1],zero,zero,zero
+; SSE41-NEXT:    retq
+;
+; AVX1-LABEL: load_zext_4i16_to_4i64:
+; AVX1:       # BB#0: # %entry
+; AVX1-NEXT:    vpmovzxwq {{.*#+}} xmm0 = mem[0],zero,zero,zero,mem[1],zero,zero,zero
+; AVX1-NEXT:    vpmovzxwq {{.*#+}} xmm1 = mem[0],zero,zero,zero,mem[1],zero,zero,zero
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: load_zext_4i16_to_4i64:
+; AVX2:       # BB#0: # %entry
+; AVX2-NEXT:    vpmovzxwq {{.*#+}} ymm0 = mem[0],zero,zero,zero,mem[1],zero,zero,zero,mem[2],zero,zero,zero,mem[3],zero,zero,zero
+; AVX2-NEXT:    retq
+entry:
+ %X = load <4 x i16>, <4 x i16>* %ptr
+ %Y = zext <4 x i16> %X to <4 x i64>
+ ret <4 x i64> %Y
+}
+
 define <8 x i32> @load_zext_8i16_to_8i32(<8 x i16> *%ptr) {
 ; SSE2-LABEL: load_zext_8i16_to_8i32:
 ; SSE2:       # BB#0: # %entry
@@ -626,7 +873,37 @@ define <8 x i32> @load_zext_8i16_to_8i32(<8 x i16> *%ptr) {
 entry:
  %X = load <8 x i16>, <8 x i16>* %ptr
  %Y = zext <8 x i16> %X to <8 x i32>
- ret <8 x i32>%Y
+ ret <8 x i32> %Y
+}
+
+define <2 x i64> @load_zext_2i32_to_2i64(<2 x i32> *%ptr) {
+; SSE2-LABEL: load_zext_2i32_to_2i64:
+; SSE2:       # BB#0: # %entry
+; SSE2-NEXT:    movq {{.*#+}} xmm0 = mem[0],zero
+; SSE2-NEXT:    pxor %xmm1, %xmm1
+; SSE2-NEXT:    punpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; SSE2-NEXT:    retq
+;
+; SSSE3-LABEL: load_zext_2i32_to_2i64:
+; SSSE3:       # BB#0: # %entry
+; SSSE3-NEXT:    movq {{.*#+}} xmm0 = mem[0],zero
+; SSSE3-NEXT:    pxor %xmm1, %xmm1
+; SSSE3-NEXT:    punpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: load_zext_2i32_to_2i64:
+; SSE41:       # BB#0: # %entry
+; SSE41-NEXT:    pmovzxdq {{.*#+}} xmm0 = mem[0],zero,mem[1],zero
+; SSE41-NEXT:    retq
+;
+; AVX-LABEL: load_zext_2i32_to_2i64:
+; AVX:       # BB#0: # %entry
+; AVX-NEXT:    vpmovzxdq {{.*#+}} xmm0 = mem[0],zero,mem[1],zero
+; AVX-NEXT:    retq
+entry:
+ %X = load <2 x i32>, <2 x i32>* %ptr
+ %Y = zext <2 x i32> %X to <2 x i64>
+ ret <2 x i64> %Y
 }
 
 define <4 x i64> @load_zext_4i32_to_4i64(<4 x i32> *%ptr) {
@@ -670,7 +947,59 @@ define <4 x i64> @load_zext_4i32_to_4i64(<4 x i32> *%ptr) {
 entry:
  %X = load <4 x i32>, <4 x i32>* %ptr
  %Y = zext <4 x i32> %X to <4 x i64>
- ret <4 x i64>%Y
+ ret <4 x i64> %Y
+}
+
+define <8 x i32> @zext_8i8_to_8i32(<8 x i8> %z) {
+; 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 {{.*#+}} xmm2 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero
+; 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:    vpmovzxwd {{.*#+}} xmm1 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero
+; AVX1-NEXT:    vpunpckhwd {{.*#+}} xmm0 = xmm0[4,4,5,5,6,6,7,7]
+; AVX1-NEXT:    vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; AVX1-NEXT:    vandps {{.*}}(%rip), %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: zext_8i8_to_8i32:
+; AVX2:       # BB#0: # %entry
+; AVX2-NEXT:    vpmovzxwd {{.*#+}} ymm0 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero,xmm0[4],zero,xmm0[5],zero,xmm0[6],zero,xmm0[7],zero
+; AVX2-NEXT:    vpbroadcastd {{.*}}(%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
 }
 
 define <8 x i32> @shuf_zext_8i16_to_8i32(<8 x i16> %A) nounwind uwtable readnone ssp {
-- 
2.34.1