diff --git a/test/CodeGen/X86/pmul.ll b/test/CodeGen/X86/pmul.ll
index dbe5bd646c7..37b6fdf7cfe 100644
--- a/test/CodeGen/X86/pmul.ll
+++ b/test/CodeGen/X86/pmul.ll
@@ -1,6 +1,7 @@
-; RUN: llc < %s -mtriple=x86_64-unknown-unknown | FileCheck %s --check-prefix=ALL --check-prefix=SSE2
-; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=sse4.1 | FileCheck %s --check-prefix=ALL --check-prefix=SSE41
-; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mcpu=core-avx2 | FileCheck %s --check-prefix=AVX2
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
+; RUN: llc < %s -mtriple=x86_64-unknown-unknown | FileCheck %s --check-prefix=SSE --check-prefix=SSE2
+; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+sse4.1 | FileCheck %s --check-prefix=SSE --check-prefix=SSE41
+; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx2 | FileCheck %s --check-prefix=AVX2
 
 define <16 x i8> @mul8c(<16 x i8> %i) nounwind  {
 ; SSE2-LABEL: mul8c:
@@ -34,16 +35,34 @@ define <16 x i8> @mul8c(<16 x i8> %i) nounwind  {
 ; SSE41-NEXT:    packuswb %xmm0, %xmm1
 ; SSE41-NEXT:    movdqa %xmm1, %xmm0
 ; SSE41-NEXT:    retq
+;
+; AVX2-LABEL: mul8c:
+; AVX2:       # BB#0: # %entry
+; AVX2-NEXT:    vpmovsxbw %xmm0, %ymm0
+; AVX2-NEXT:    vpmovsxbw {{.*}}(%rip), %ymm1
+; AVX2-NEXT:    vpmullw %ymm1, %ymm0, %ymm0
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vmovdqa {{.*#+}} xmm2 = <0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u>
+; AVX2-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX2-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; AVX2-NEXT:    vzeroupper
+; AVX2-NEXT:    retq
 entry:
   %A = mul <16 x i8> %i, < i8 117, i8 117, i8 117, i8 117, i8 117, i8 117, i8 117, i8 117, i8 117, i8 117, i8 117, i8 117, i8 117, i8 117, i8 117, i8 117 >
   ret <16 x i8> %A
 }
 
 define <8 x i16> @mul16c(<8 x i16> %i) nounwind  {
-; ALL-LABEL: mul16c:
-; ALL:       # BB#0: # %entry
-; ALL-NEXT:    pmullw {{.*}}(%rip), %xmm0
-; ALL-NEXT:    retq
+; SSE-LABEL: mul16c:
+; SSE:       # BB#0: # %entry
+; SSE-NEXT:    pmullw {{.*}}(%rip), %xmm0
+; SSE-NEXT:    retq
+;
+; AVX2-LABEL: mul16c:
+; AVX2:       # BB#0: # %entry
+; AVX2-NEXT:    vpmullw {{.*}}(%rip), %xmm0, %xmm0
+; AVX2-NEXT:    retq
 entry:
   %A = mul <8 x i16> %i, < i16 117, i16 117, i16 117, i16 117, i16 117, i16 117, i16 117, i16 117 >
   ret <8 x i16> %A
@@ -65,22 +84,38 @@ define <4 x i32> @a(<4 x i32> %i) nounwind  {
 ; SSE41:       # BB#0: # %entry
 ; SSE41-NEXT:    pmulld {{.*}}(%rip), %xmm0
 ; SSE41-NEXT:    retq
+;
+; AVX2-LABEL: a:
+; AVX2:       # BB#0: # %entry
+; AVX2-NEXT:    vpbroadcastd {{.*}}(%rip), %xmm1
+; AVX2-NEXT:    vpmulld %xmm1, %xmm0, %xmm0
+; AVX2-NEXT:    retq
 entry:
   %A = mul <4 x i32> %i, < i32 117, i32 117, i32 117, i32 117 >
   ret <4 x i32> %A
 }
 
 define <2 x i64> @b(<2 x i64> %i) nounwind  {
-; ALL-LABEL: b:
-; ALL:       # BB#0: # %entry
-; ALL-NEXT:    movdqa {{.*#+}} xmm1 = [117,117]
-; ALL-NEXT:    movdqa %xmm0, %xmm2
-; ALL-NEXT:    pmuludq %xmm1, %xmm2
-; ALL-NEXT:    psrlq $32, %xmm0
-; ALL-NEXT:    pmuludq %xmm1, %xmm0
-; ALL-NEXT:    psllq $32, %xmm0
-; ALL-NEXT:    paddq %xmm2, %xmm0
-; ALL-NEXT:    retq
+; SSE-LABEL: b:
+; SSE:       # BB#0: # %entry
+; SSE-NEXT:    movdqa {{.*#+}} xmm1 = [117,117]
+; SSE-NEXT:    movdqa %xmm0, %xmm2
+; SSE-NEXT:    pmuludq %xmm1, %xmm2
+; SSE-NEXT:    psrlq $32, %xmm0
+; SSE-NEXT:    pmuludq %xmm1, %xmm0
+; SSE-NEXT:    psllq $32, %xmm0
+; SSE-NEXT:    paddq %xmm2, %xmm0
+; SSE-NEXT:    retq
+;
+; AVX2-LABEL: b:
+; AVX2:       # BB#0: # %entry
+; AVX2-NEXT:    vmovdqa {{.*#+}} xmm1 = [117,117]
+; AVX2-NEXT:    vpmuludq %xmm1, %xmm0, %xmm2
+; AVX2-NEXT:    vpsrlq $32, %xmm0, %xmm0
+; AVX2-NEXT:    vpmuludq %xmm1, %xmm0, %xmm0
+; AVX2-NEXT:    vpsllq $32, %xmm0, %xmm0
+; AVX2-NEXT:    vpaddq %xmm0, %xmm2, %xmm0
+; AVX2-NEXT:    retq
 entry:
   %A = mul <2 x i64> %i, < i64 117, i64 117 >
   ret <2 x i64> %A
@@ -123,16 +158,34 @@ define <16 x i8> @mul8(<16 x i8> %i, <16 x i8> %j) nounwind  {
 ; SSE41-NEXT:    packuswb %xmm0, %xmm2
 ; SSE41-NEXT:    movdqa %xmm2, %xmm0
 ; SSE41-NEXT:    retq
+;
+; AVX2-LABEL: mul8:
+; AVX2:       # BB#0: # %entry
+; AVX2-NEXT:    vpmovsxbw %xmm1, %ymm1
+; AVX2-NEXT:    vpmovsxbw %xmm0, %ymm0
+; AVX2-NEXT:    vpmullw %ymm1, %ymm0, %ymm0
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT:    vmovdqa {{.*#+}} xmm2 = <0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u>
+; AVX2-NEXT:    vpshufb %xmm2, %xmm1, %xmm1
+; AVX2-NEXT:    vpshufb %xmm2, %xmm0, %xmm0
+; AVX2-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; AVX2-NEXT:    vzeroupper
+; AVX2-NEXT:    retq
 entry:
   %A = mul <16 x i8> %i, %j
   ret <16 x i8> %A
 }
 
 define <8 x i16> @mul16(<8 x i16> %i, <8 x i16> %j) nounwind  {
-; ALL-LABEL: mul16:
-; ALL:       # BB#0: # %entry
-; ALL-NEXT:    pmullw %xmm1, %xmm0
-; ALL-NEXT:    retq
+; SSE-LABEL: mul16:
+; SSE:       # BB#0: # %entry
+; SSE-NEXT:    pmullw %xmm1, %xmm0
+; SSE-NEXT:    retq
+;
+; AVX2-LABEL: mul16:
+; AVX2:       # BB#0: # %entry
+; AVX2-NEXT:    vpmullw %xmm1, %xmm0, %xmm0
+; AVX2-NEXT:    retq
 entry:
   %A = mul <8 x i16> %i, %j
   ret <8 x i16> %A
@@ -154,26 +207,44 @@ define <4 x i32> @c(<4 x i32> %i, <4 x i32> %j) nounwind  {
 ; SSE41:       # BB#0: # %entry
 ; SSE41-NEXT:    pmulld %xmm1, %xmm0
 ; SSE41-NEXT:    retq
+;
+; AVX2-LABEL: c:
+; AVX2:       # BB#0: # %entry
+; AVX2-NEXT:    vpmulld %xmm1, %xmm0, %xmm0
+; AVX2-NEXT:    retq
 entry:
   %A = mul <4 x i32> %i, %j
   ret <4 x i32> %A
 }
 
 define <2 x i64> @d(<2 x i64> %i, <2 x i64> %j) nounwind  {
-; ALL-LABEL: d:
-; ALL:       # BB#0: # %entry
-; ALL-NEXT:    movdqa %xmm0, %xmm2
-; ALL-NEXT:    pmuludq %xmm1, %xmm2
-; ALL-NEXT:    movdqa %xmm1, %xmm3
-; ALL-NEXT:    psrlq $32, %xmm3
-; ALL-NEXT:    pmuludq %xmm0, %xmm3
-; ALL-NEXT:    psllq $32, %xmm3
-; ALL-NEXT:    paddq %xmm3, %xmm2
-; ALL-NEXT:    psrlq $32, %xmm0
-; ALL-NEXT:    pmuludq %xmm1, %xmm0
-; ALL-NEXT:    psllq $32, %xmm0
-; ALL-NEXT:    paddq %xmm2, %xmm0
-; ALL-NEXT:    retq
+; SSE-LABEL: d:
+; SSE:       # BB#0: # %entry
+; SSE-NEXT:    movdqa %xmm0, %xmm2
+; SSE-NEXT:    pmuludq %xmm1, %xmm2
+; SSE-NEXT:    movdqa %xmm1, %xmm3
+; SSE-NEXT:    psrlq $32, %xmm3
+; SSE-NEXT:    pmuludq %xmm0, %xmm3
+; SSE-NEXT:    psllq $32, %xmm3
+; SSE-NEXT:    paddq %xmm3, %xmm2
+; SSE-NEXT:    psrlq $32, %xmm0
+; SSE-NEXT:    pmuludq %xmm1, %xmm0
+; SSE-NEXT:    psllq $32, %xmm0
+; SSE-NEXT:    paddq %xmm2, %xmm0
+; SSE-NEXT:    retq
+;
+; AVX2-LABEL: d:
+; AVX2:       # BB#0: # %entry
+; AVX2-NEXT:    vpmuludq %xmm1, %xmm0, %xmm2
+; AVX2-NEXT:    vpsrlq $32, %xmm1, %xmm3
+; AVX2-NEXT:    vpmuludq %xmm3, %xmm0, %xmm3
+; AVX2-NEXT:    vpsllq $32, %xmm3, %xmm3
+; AVX2-NEXT:    vpaddq %xmm3, %xmm2, %xmm2
+; AVX2-NEXT:    vpsrlq $32, %xmm0, %xmm0
+; AVX2-NEXT:    vpmuludq %xmm1, %xmm0, %xmm0
+; AVX2-NEXT:    vpsllq $32, %xmm0, %xmm0
+; AVX2-NEXT:    vpaddq %xmm0, %xmm2, %xmm0
+; AVX2-NEXT:    retq
 entry:
   %A = mul <2 x i64> %i, %j
   ret <2 x i64> %A
@@ -210,6 +281,17 @@ define <4 x i32> @e(<4 x i32> %i, <4 x i32> %j) nounwind  {
 ; SSE41-NEXT:    pmulld {{[0-9]+}}(%rsp), %xmm0 # 16-byte Folded Reload
 ; SSE41-NEXT:    addq $40, %rsp
 ; SSE41-NEXT:    retq
+;
+; AVX2-LABEL: e:
+; AVX2:       # BB#0: # %entry
+; AVX2-NEXT:    subq $40, %rsp
+; AVX2-NEXT:    vmovaps %xmm1, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX2-NEXT:    vmovaps %xmm0, (%rsp) # 16-byte Spill
+; AVX2-NEXT:    callq foo
+; AVX2-NEXT:    vmovdqa (%rsp), %xmm0 # 16-byte Reload
+; AVX2-NEXT:    vpmulld {{[0-9]+}}(%rsp), %xmm0, %xmm0 # 16-byte Folded Reload
+; AVX2-NEXT:    addq $40, %rsp
+; AVX2-NEXT:    retq
 entry:
   ; Use a call to force spills.
   call void @foo()
@@ -218,27 +300,47 @@ entry:
 }
 
 define <2 x i64> @f(<2 x i64> %i, <2 x i64> %j) nounwind  {
-; ALL-LABEL: f:
-; ALL:       # BB#0: # %entry
-; ALL-NEXT:    subq $40, %rsp
-; ALL-NEXT:    movaps %xmm1, {{[0-9]+}}(%rsp) # 16-byte Spill
-; ALL-NEXT:    movaps %xmm0, (%rsp) # 16-byte Spill
-; ALL-NEXT:    callq foo
-; ALL-NEXT:    movdqa (%rsp), %xmm0 # 16-byte Reload
-; ALL-NEXT:    movdqa %xmm0, %xmm2
-; ALL-NEXT:    movdqa {{[0-9]+}}(%rsp), %xmm3 # 16-byte Reload
-; ALL-NEXT:    pmuludq %xmm3, %xmm2
-; ALL-NEXT:    movdqa %xmm3, %xmm1
-; ALL-NEXT:    psrlq $32, %xmm1
-; ALL-NEXT:    pmuludq %xmm0, %xmm1
-; ALL-NEXT:    psllq $32, %xmm1
-; ALL-NEXT:    paddq %xmm1, %xmm2
-; ALL-NEXT:    psrlq $32, %xmm0
-; ALL-NEXT:    pmuludq %xmm3, %xmm0
-; ALL-NEXT:    psllq $32, %xmm0
-; ALL-NEXT:    paddq %xmm2, %xmm0
-; ALL-NEXT:    addq $40, %rsp
-; ALL-NEXT:    retq
+; SSE-LABEL: f:
+; SSE:       # BB#0: # %entry
+; SSE-NEXT:    subq $40, %rsp
+; SSE-NEXT:    movaps %xmm1, {{[0-9]+}}(%rsp) # 16-byte Spill
+; SSE-NEXT:    movaps %xmm0, (%rsp) # 16-byte Spill
+; SSE-NEXT:    callq foo
+; SSE-NEXT:    movdqa (%rsp), %xmm0 # 16-byte Reload
+; SSE-NEXT:    movdqa %xmm0, %xmm2
+; SSE-NEXT:    movdqa {{[0-9]+}}(%rsp), %xmm3 # 16-byte Reload
+; SSE-NEXT:    pmuludq %xmm3, %xmm2
+; SSE-NEXT:    movdqa %xmm3, %xmm1
+; SSE-NEXT:    psrlq $32, %xmm1
+; SSE-NEXT:    pmuludq %xmm0, %xmm1
+; SSE-NEXT:    psllq $32, %xmm1
+; SSE-NEXT:    paddq %xmm1, %xmm2
+; SSE-NEXT:    psrlq $32, %xmm0
+; SSE-NEXT:    pmuludq %xmm3, %xmm0
+; SSE-NEXT:    psllq $32, %xmm0
+; SSE-NEXT:    paddq %xmm2, %xmm0
+; SSE-NEXT:    addq $40, %rsp
+; SSE-NEXT:    retq
+;
+; AVX2-LABEL: f:
+; AVX2:       # BB#0: # %entry
+; AVX2-NEXT:    subq $40, %rsp
+; AVX2-NEXT:    vmovaps %xmm1, {{[0-9]+}}(%rsp) # 16-byte Spill
+; AVX2-NEXT:    vmovaps %xmm0, (%rsp) # 16-byte Spill
+; AVX2-NEXT:    callq foo
+; AVX2-NEXT:    vmovdqa {{[0-9]+}}(%rsp), %xmm2 # 16-byte Reload
+; AVX2-NEXT:    vmovdqa (%rsp), %xmm3 # 16-byte Reload
+; AVX2-NEXT:    vpmuludq %xmm2, %xmm3, %xmm0
+; AVX2-NEXT:    vpsrlq $32, %xmm2, %xmm1
+; AVX2-NEXT:    vpmuludq %xmm1, %xmm3, %xmm1
+; AVX2-NEXT:    vpsllq $32, %xmm1, %xmm1
+; AVX2-NEXT:    vpaddq %xmm1, %xmm0, %xmm0
+; AVX2-NEXT:    vpsrlq $32, %xmm3, %xmm1
+; AVX2-NEXT:    vpmuludq %xmm2, %xmm1, %xmm1
+; AVX2-NEXT:    vpsllq $32, %xmm1, %xmm1
+; AVX2-NEXT:    vpaddq %xmm1, %xmm0, %xmm0
+; AVX2-NEXT:    addq $40, %rsp
+; AVX2-NEXT:    retq
 entry:
   ; Use a call to force spills.
   call void @foo()
@@ -247,31 +349,76 @@ entry:
 }
 
 define <4 x i64> @b1(<4 x i64> %i) nounwind  {
-; AVX2-LABEL: @b1
-; AVX2: vpbroadcastq
-; AVX2-NEXT: vpmuludq
-; AVX2-NEXT: vpsrlq  $32 
-; AVX2-NEXT: vpmuludq
-; AVX2-NEXT: vpsllq  $32
-; AVX2-NEXT: vpaddq
-; AVX2-NEXT: retq
+; SSE-LABEL: b1:
+; SSE:       # BB#0: # %entry
+; SSE-NEXT:    movdqa {{.*#+}} xmm2 = [117,117]
+; SSE-NEXT:    movdqa %xmm0, %xmm3
+; SSE-NEXT:    pmuludq %xmm2, %xmm3
+; SSE-NEXT:    psrlq $32, %xmm0
+; SSE-NEXT:    pmuludq %xmm2, %xmm0
+; SSE-NEXT:    psllq $32, %xmm0
+; SSE-NEXT:    paddq %xmm3, %xmm0
+; SSE-NEXT:    movdqa %xmm1, %xmm3
+; SSE-NEXT:    pmuludq %xmm2, %xmm3
+; SSE-NEXT:    psrlq $32, %xmm1
+; SSE-NEXT:    pmuludq %xmm2, %xmm1
+; SSE-NEXT:    psllq $32, %xmm1
+; SSE-NEXT:    paddq %xmm3, %xmm1
+; SSE-NEXT:    retq
+;
+; AVX2-LABEL: b1:
+; AVX2:       # BB#0: # %entry
+; AVX2-NEXT:    vpbroadcastq {{.*}}(%rip), %ymm1
+; AVX2-NEXT:    vpmuludq %ymm1, %ymm0, %ymm2
+; AVX2-NEXT:    vpsrlq $32, %ymm0, %ymm0
+; AVX2-NEXT:    vpmuludq %ymm1, %ymm0, %ymm0
+; AVX2-NEXT:    vpsllq $32, %ymm0, %ymm0
+; AVX2-NEXT:    vpaddq %ymm0, %ymm2, %ymm0
+; AVX2-NEXT:    retq
 entry:
   %A = mul <4 x i64> %i, < i64 117, i64 117, i64 117, i64 117 >
   ret <4 x i64> %A
 }
 
 define <4 x i64> @b2(<4 x i64> %i, <4 x i64> %j) nounwind  {
-; AVX2-LABEL: @b2
-; AVX2:  vpmuludq
-; AVX2-NEXT: vpsrlq  $32
-; AVX2-NEXT: vpmuludq
-; AVX2-NEXT: vpsllq  $32
-; AVX2-NEXT: vpaddq
-; AVX2-NEXT: vpsrlq  $32
-; AVX2-NEXT: vpmuludq
-; AVX2-NEXT: vpsllq  $32
-; AVX2-NEXT: vpaddq
-; AVX2-NEXT: retq
+; SSE-LABEL: b2:
+; SSE:       # BB#0: # %entry
+; SSE-NEXT:    movdqa %xmm0, %xmm4
+; SSE-NEXT:    pmuludq %xmm2, %xmm4
+; SSE-NEXT:    movdqa %xmm2, %xmm5
+; SSE-NEXT:    psrlq $32, %xmm5
+; SSE-NEXT:    pmuludq %xmm0, %xmm5
+; SSE-NEXT:    psllq $32, %xmm5
+; SSE-NEXT:    paddq %xmm5, %xmm4
+; SSE-NEXT:    psrlq $32, %xmm0
+; SSE-NEXT:    pmuludq %xmm2, %xmm0
+; SSE-NEXT:    psllq $32, %xmm0
+; SSE-NEXT:    paddq %xmm4, %xmm0
+; SSE-NEXT:    movdqa %xmm1, %xmm2
+; SSE-NEXT:    pmuludq %xmm3, %xmm2
+; SSE-NEXT:    movdqa %xmm3, %xmm4
+; SSE-NEXT:    psrlq $32, %xmm4
+; SSE-NEXT:    pmuludq %xmm1, %xmm4
+; SSE-NEXT:    psllq $32, %xmm4
+; SSE-NEXT:    paddq %xmm4, %xmm2
+; SSE-NEXT:    psrlq $32, %xmm1
+; SSE-NEXT:    pmuludq %xmm3, %xmm1
+; SSE-NEXT:    psllq $32, %xmm1
+; SSE-NEXT:    paddq %xmm2, %xmm1
+; SSE-NEXT:    retq
+;
+; AVX2-LABEL: b2:
+; AVX2:       # BB#0: # %entry
+; AVX2-NEXT:    vpmuludq %ymm1, %ymm0, %ymm2
+; AVX2-NEXT:    vpsrlq $32, %ymm1, %ymm3
+; AVX2-NEXT:    vpmuludq %ymm3, %ymm0, %ymm3
+; AVX2-NEXT:    vpsllq $32, %ymm3, %ymm3
+; AVX2-NEXT:    vpaddq %ymm3, %ymm2, %ymm2
+; AVX2-NEXT:    vpsrlq $32, %ymm0, %ymm0
+; AVX2-NEXT:    vpmuludq %ymm1, %ymm0, %ymm0
+; AVX2-NEXT:    vpsllq $32, %ymm0, %ymm0
+; AVX2-NEXT:    vpaddq %ymm0, %ymm2, %ymm0
+; AVX2-NEXT:    retq
 entry:
   %A = mul <4 x i64> %i, %j
   ret <4 x i64> %A