[X86][MMX] Regenerate mmx load folding tests

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@297470 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/test/CodeGen/X86/mmx-fold-load.ll b/test/CodeGen/X86/mmx-fold-load.ll
index 2b9d30f59fd50ff5cbf2c91c76a19a6b9e387587..9505adbabf8c6ef567fd119f46fe04867318d2fb 100644 (file)
--- a/test/CodeGen/X86/mmx-fold-load.ll
+++ b/test/CodeGen/X86/mmx-fold-load.ll
@@ -1,12 +1,31 @@
-; RUN: llc < %s -march=x86-64 -mattr=+mmx,+sse2 | FileCheck %s
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
+; RUN: llc < %s -mtriple=i686-unknown -mattr=+mmx,+sse2 | FileCheck %s --check-prefix=X86
+; RUN: llc < %s -mtriple=x86_64-unknown -mattr=+mmx,+sse2 | FileCheck %s --check-prefix=X64
 
-define i64 @t0(<1 x i64>* %a, i32* %b) {
-; CHECK-LABEL: t0:
-; CHECK:       # BB#0:{{.*}} %entry
-; CHECK:         movq (%[[REG1:[a-z]+]]), %mm0
-; CHECK-NEXT:    psllq (%[[REG2:[a-z]+]]), %mm0
-; CHECK-NEXT:    movd %mm0, %rax
-; CHECK-NEXT:    retq
+define i64 @t0(<1 x i64>* %a, i32* %b) nounwind {
+; X86-LABEL: t0:
+; X86:       # BB#0: # %entry
+; X86-NEXT:    pushl %ebp
+; X86-NEXT:    movl %esp, %ebp
+; X86-NEXT:    andl $-8, %esp
+; X86-NEXT:    subl $8, %esp
+; X86-NEXT:    movl 12(%ebp), %eax
+; X86-NEXT:    movl 8(%ebp), %ecx
+; X86-NEXT:    movq (%ecx), %mm0
+; X86-NEXT:    psllq (%eax), %mm0
+; X86-NEXT:    movq %mm0, (%esp)
+; X86-NEXT:    movl (%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %edx
+; X86-NEXT:    movl %ebp, %esp
+; X86-NEXT:    popl %ebp
+; X86-NEXT:    retl
+;
+; X64-LABEL: t0:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    movq (%rdi), %mm0
+; X64-NEXT:    psllq (%rsi), %mm0
+; X64-NEXT:    movd %mm0, %rax
+; X64-NEXT:    retq
 entry:
   %0 = bitcast <1 x i64>* %a to x86_mmx*
   %1 = load x86_mmx, x86_mmx* %0, align 8
@@ -17,13 +36,30 @@ entry:
 }
 declare x86_mmx @llvm.x86.mmx.pslli.q(x86_mmx, i32)
 
-define i64 @t1(<1 x i64>* %a, i32* %b) {
-; CHECK-LABEL: t1:
-; CHECK:       # BB#0:{{.*}} %entry
-; CHECK:         movq (%[[REG1]]), %mm0
-; CHECK-NEXT:    psrlq (%[[REG2]]), %mm0
-; CHECK-NEXT:    movd %mm0, %rax
-; CHECK-NEXT:    retq
+define i64 @t1(<1 x i64>* %a, i32* %b) nounwind {
+; X86-LABEL: t1:
+; X86:       # BB#0: # %entry
+; X86-NEXT:    pushl %ebp
+; X86-NEXT:    movl %esp, %ebp
+; X86-NEXT:    andl $-8, %esp
+; X86-NEXT:    subl $8, %esp
+; X86-NEXT:    movl 12(%ebp), %eax
+; X86-NEXT:    movl 8(%ebp), %ecx
+; X86-NEXT:    movq (%ecx), %mm0
+; X86-NEXT:    psrlq (%eax), %mm0
+; X86-NEXT:    movq %mm0, (%esp)
+; X86-NEXT:    movl (%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %edx
+; X86-NEXT:    movl %ebp, %esp
+; X86-NEXT:    popl %ebp
+; X86-NEXT:    retl
+;
+; X64-LABEL: t1:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    movq (%rdi), %mm0
+; X64-NEXT:    psrlq (%rsi), %mm0
+; X64-NEXT:    movd %mm0, %rax
+; X64-NEXT:    retq
 entry:
   %0 = bitcast <1 x i64>* %a to x86_mmx*
   %1 = load x86_mmx, x86_mmx* %0, align 8
@@ -34,13 +70,30 @@ entry:
 }
 declare x86_mmx @llvm.x86.mmx.psrli.q(x86_mmx, i32)
 
-define i64 @t2(<1 x i64>* %a, i32* %b) {
-; CHECK-LABEL: t2:
-; CHECK:       # BB#0:{{.*}} %entry
-; CHECK:         movq (%[[REG1]]), %mm0
-; CHECK-NEXT:    psllw (%[[REG2]]), %mm0
-; CHECK-NEXT:    movd %mm0, %rax
-; CHECK-NEXT:    retq
+define i64 @t2(<1 x i64>* %a, i32* %b) nounwind {
+; X86-LABEL: t2:
+; X86:       # BB#0: # %entry
+; X86-NEXT:    pushl %ebp
+; X86-NEXT:    movl %esp, %ebp
+; X86-NEXT:    andl $-8, %esp
+; X86-NEXT:    subl $8, %esp
+; X86-NEXT:    movl 12(%ebp), %eax
+; X86-NEXT:    movl 8(%ebp), %ecx
+; X86-NEXT:    movq (%ecx), %mm0
+; X86-NEXT:    psllw (%eax), %mm0
+; X86-NEXT:    movq %mm0, (%esp)
+; X86-NEXT:    movl (%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %edx
+; X86-NEXT:    movl %ebp, %esp
+; X86-NEXT:    popl %ebp
+; X86-NEXT:    retl
+;
+; X64-LABEL: t2:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    movq (%rdi), %mm0
+; X64-NEXT:    psllw (%rsi), %mm0
+; X64-NEXT:    movd %mm0, %rax
+; X64-NEXT:    retq
 entry:
   %0 = bitcast <1 x i64>* %a to x86_mmx*
   %1 = load x86_mmx, x86_mmx* %0, align 8
@@ -51,13 +104,30 @@ entry:
 }
 declare x86_mmx @llvm.x86.mmx.pslli.w(x86_mmx, i32)
 
-define i64 @t3(<1 x i64>* %a, i32* %b) {
-; CHECK-LABEL: t3:
-; CHECK:       # BB#0:{{.*}} %entry
-; CHECK:         movq (%[[REG1]]), %mm0
-; CHECK-NEXT:    psrlw (%[[REG2]]), %mm0
-; CHECK-NEXT:    movd %mm0, %rax
-; CHECK-NEXT:    retq
+define i64 @t3(<1 x i64>* %a, i32* %b) nounwind {
+; X86-LABEL: t3:
+; X86:       # BB#0: # %entry
+; X86-NEXT:    pushl %ebp
+; X86-NEXT:    movl %esp, %ebp
+; X86-NEXT:    andl $-8, %esp
+; X86-NEXT:    subl $8, %esp
+; X86-NEXT:    movl 12(%ebp), %eax
+; X86-NEXT:    movl 8(%ebp), %ecx
+; X86-NEXT:    movq (%ecx), %mm0
+; X86-NEXT:    psrlw (%eax), %mm0
+; X86-NEXT:    movq %mm0, (%esp)
+; X86-NEXT:    movl (%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %edx
+; X86-NEXT:    movl %ebp, %esp
+; X86-NEXT:    popl %ebp
+; X86-NEXT:    retl
+;
+; X64-LABEL: t3:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    movq (%rdi), %mm0
+; X64-NEXT:    psrlw (%rsi), %mm0
+; X64-NEXT:    movd %mm0, %rax
+; X64-NEXT:    retq
 entry:
   %0 = bitcast <1 x i64>* %a to x86_mmx*
   %1 = load x86_mmx, x86_mmx* %0, align 8
@@ -68,13 +138,30 @@ entry:
 }
 declare x86_mmx @llvm.x86.mmx.psrli.w(x86_mmx, i32)
 
-define i64 @t4(<1 x i64>* %a, i32* %b) {
-; CHECK-LABEL: t4:
-; CHECK:       # BB#0:{{.*}} %entry
-; CHECK:         movq (%[[REG1]]), %mm0
-; CHECK-NEXT:    pslld (%[[REG2]]), %mm0
-; CHECK-NEXT:    movd %mm0, %rax
-; CHECK-NEXT:    retq
+define i64 @t4(<1 x i64>* %a, i32* %b) nounwind {
+; X86-LABEL: t4:
+; X86:       # BB#0: # %entry
+; X86-NEXT:    pushl %ebp
+; X86-NEXT:    movl %esp, %ebp
+; X86-NEXT:    andl $-8, %esp
+; X86-NEXT:    subl $8, %esp
+; X86-NEXT:    movl 12(%ebp), %eax
+; X86-NEXT:    movl 8(%ebp), %ecx
+; X86-NEXT:    movq (%ecx), %mm0
+; X86-NEXT:    pslld (%eax), %mm0
+; X86-NEXT:    movq %mm0, (%esp)
+; X86-NEXT:    movl (%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %edx
+; X86-NEXT:    movl %ebp, %esp
+; X86-NEXT:    popl %ebp
+; X86-NEXT:    retl
+;
+; X64-LABEL: t4:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    movq (%rdi), %mm0
+; X64-NEXT:    pslld (%rsi), %mm0
+; X64-NEXT:    movd %mm0, %rax
+; X64-NEXT:    retq
 entry:
   %0 = bitcast <1 x i64>* %a to x86_mmx*
   %1 = load x86_mmx, x86_mmx* %0, align 8
@@ -85,13 +172,30 @@ entry:
 }
 declare x86_mmx @llvm.x86.mmx.pslli.d(x86_mmx, i32)
 
-define i64 @t5(<1 x i64>* %a, i32* %b) {
-; CHECK-LABEL: t5:
-; CHECK:       # BB#0:{{.*}} %entry
-; CHECK:         movq (%[[REG1]]), %mm0
-; CHECK-NEXT:    psrld (%[[REG2]]), %mm0
-; CHECK-NEXT:    movd %mm0, %rax
-; CHECK-NEXT:    retq
+define i64 @t5(<1 x i64>* %a, i32* %b) nounwind {
+; X86-LABEL: t5:
+; X86:       # BB#0: # %entry
+; X86-NEXT:    pushl %ebp
+; X86-NEXT:    movl %esp, %ebp
+; X86-NEXT:    andl $-8, %esp
+; X86-NEXT:    subl $8, %esp
+; X86-NEXT:    movl 12(%ebp), %eax
+; X86-NEXT:    movl 8(%ebp), %ecx
+; X86-NEXT:    movq (%ecx), %mm0
+; X86-NEXT:    psrld (%eax), %mm0
+; X86-NEXT:    movq %mm0, (%esp)
+; X86-NEXT:    movl (%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %edx
+; X86-NEXT:    movl %ebp, %esp
+; X86-NEXT:    popl %ebp
+; X86-NEXT:    retl
+;
+; X64-LABEL: t5:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    movq (%rdi), %mm0
+; X64-NEXT:    psrld (%rsi), %mm0
+; X64-NEXT:    movd %mm0, %rax
+; X64-NEXT:    retq
 entry:
   %0 = bitcast <1 x i64>* %a to x86_mmx*
   %1 = load x86_mmx, x86_mmx* %0, align 8
@@ -102,13 +206,30 @@ entry:
 }
 declare x86_mmx @llvm.x86.mmx.psrli.d(x86_mmx, i32)
 
-define i64 @t6(<1 x i64>* %a, i32* %b) {
-; CHECK-LABEL: t6:
-; CHECK:       # BB#0:{{.*}} %entry
-; CHECK:         movq (%[[REG1]]), %mm0
-; CHECK-NEXT:    psraw (%[[REG2]]), %mm0
-; CHECK-NEXT:    movd %mm0, %rax
-; CHECK-NEXT:    retq
+define i64 @t6(<1 x i64>* %a, i32* %b) nounwind {
+; X86-LABEL: t6:
+; X86:       # BB#0: # %entry
+; X86-NEXT:    pushl %ebp
+; X86-NEXT:    movl %esp, %ebp
+; X86-NEXT:    andl $-8, %esp
+; X86-NEXT:    subl $8, %esp
+; X86-NEXT:    movl 12(%ebp), %eax
+; X86-NEXT:    movl 8(%ebp), %ecx
+; X86-NEXT:    movq (%ecx), %mm0
+; X86-NEXT:    psraw (%eax), %mm0
+; X86-NEXT:    movq %mm0, (%esp)
+; X86-NEXT:    movl (%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %edx
+; X86-NEXT:    movl %ebp, %esp
+; X86-NEXT:    popl %ebp
+; X86-NEXT:    retl
+;
+; X64-LABEL: t6:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    movq (%rdi), %mm0
+; X64-NEXT:    psraw (%rsi), %mm0
+; X64-NEXT:    movd %mm0, %rax
+; X64-NEXT:    retq
 entry:
   %0 = bitcast <1 x i64>* %a to x86_mmx*
   %1 = load x86_mmx, x86_mmx* %0, align 8
@@ -119,13 +240,30 @@ entry:
 }
 declare x86_mmx @llvm.x86.mmx.psrai.w(x86_mmx, i32)
 
-define i64 @t7(<1 x i64>* %a, i32* %b) {
-; CHECK-LABEL: t7:
-; CHECK:       # BB#0:{{.*}} %entry
-; CHECK:         movq (%[[REG1]]), %mm0
-; CHECK-NEXT:    psrad (%[[REG2]]), %mm0
-; CHECK-NEXT:    movd %mm0, %rax
-; CHECK-NEXT:    retq
+define i64 @t7(<1 x i64>* %a, i32* %b) nounwind {
+; X86-LABEL: t7:
+; X86:       # BB#0: # %entry
+; X86-NEXT:    pushl %ebp
+; X86-NEXT:    movl %esp, %ebp
+; X86-NEXT:    andl $-8, %esp
+; X86-NEXT:    subl $8, %esp
+; X86-NEXT:    movl 12(%ebp), %eax
+; X86-NEXT:    movl 8(%ebp), %ecx
+; X86-NEXT:    movq (%ecx), %mm0
+; X86-NEXT:    psrad (%eax), %mm0
+; X86-NEXT:    movq %mm0, (%esp)
+; X86-NEXT:    movl (%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %edx
+; X86-NEXT:    movl %ebp, %esp
+; X86-NEXT:    popl %ebp
+; X86-NEXT:    retl
+;
+; X64-LABEL: t7:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    movq (%rdi), %mm0
+; X64-NEXT:    psrad (%rsi), %mm0
+; X64-NEXT:    movd %mm0, %rax
+; X64-NEXT:    retq
 entry:
   %0 = bitcast <1 x i64>* %a to x86_mmx*
   %1 = load x86_mmx, x86_mmx* %0, align 8
@@ -136,13 +274,29 @@ entry:
 }
 declare x86_mmx @llvm.x86.mmx.psrai.d(x86_mmx, i32)
 
-define i64 @tt0(x86_mmx %t, x86_mmx* %q) {
-; CHECK-LABEL: tt0:
-; CHECK:       # BB#0:{{.*}} %entry
-; CHECK:    paddb (%[[REG3:[a-z]+]]), %mm0
-; CHECK-NEXT:    movd %mm0, %rax
-; CHECK-NEXT:    emms
-; CHECK-NEXT:    retq
+define i64 @tt0(x86_mmx %t, x86_mmx* %q) nounwind {
+; X86-LABEL: tt0:
+; X86:       # BB#0: # %entry
+; X86-NEXT:    pushl %ebp
+; X86-NEXT:    movl %esp, %ebp
+; X86-NEXT:    andl $-8, %esp
+; X86-NEXT:    subl $8, %esp
+; X86-NEXT:    movl 8(%ebp), %eax
+; X86-NEXT:    paddb (%eax), %mm0
+; X86-NEXT:    movq %mm0, (%esp)
+; X86-NEXT:    movl (%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %edx
+; X86-NEXT:    emms
+; X86-NEXT:    movl %ebp, %esp
+; X86-NEXT:    popl %ebp
+; X86-NEXT:    retl
+;
+; X64-LABEL: tt0:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    paddb (%rdi), %mm0
+; X64-NEXT:    movd %mm0, %rax
+; X64-NEXT:    emms
+; X64-NEXT:    retq
 entry:
   %v = load x86_mmx, x86_mmx* %q
   %u = tail call x86_mmx @llvm.x86.mmx.padd.b(x86_mmx %t, x86_mmx %v)
@@ -153,13 +307,29 @@ entry:
 declare x86_mmx @llvm.x86.mmx.padd.b(x86_mmx, x86_mmx)
 declare void @llvm.x86.mmx.emms()
 
-define i64 @tt1(x86_mmx %t, x86_mmx* %q) {
-; CHECK-LABEL: tt1:
-; CHECK:       # BB#0:{{.*}} %entry
-; CHECK:    paddw (%[[REG3]]), %mm0
-; CHECK-NEXT:    movd %mm0, %rax
-; CHECK-NEXT:    emms
-; CHECK-NEXT:    retq
+define i64 @tt1(x86_mmx %t, x86_mmx* %q) nounwind {
+; X86-LABEL: tt1:
+; X86:       # BB#0: # %entry
+; X86-NEXT:    pushl %ebp
+; X86-NEXT:    movl %esp, %ebp
+; X86-NEXT:    andl $-8, %esp
+; X86-NEXT:    subl $8, %esp
+; X86-NEXT:    movl 8(%ebp), %eax
+; X86-NEXT:    paddw (%eax), %mm0
+; X86-NEXT:    movq %mm0, (%esp)
+; X86-NEXT:    movl (%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %edx
+; X86-NEXT:    emms
+; X86-NEXT:    movl %ebp, %esp
+; X86-NEXT:    popl %ebp
+; X86-NEXT:    retl
+;
+; X64-LABEL: tt1:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    paddw (%rdi), %mm0
+; X64-NEXT:    movd %mm0, %rax
+; X64-NEXT:    emms
+; X64-NEXT:    retq
 entry:
   %v = load x86_mmx, x86_mmx* %q
   %u = tail call x86_mmx @llvm.x86.mmx.padd.w(x86_mmx %t, x86_mmx %v)
@@ -169,13 +339,29 @@ entry:
 }
 declare x86_mmx @llvm.x86.mmx.padd.w(x86_mmx, x86_mmx)
 
-define i64 @tt2(x86_mmx %t, x86_mmx* %q) {
-; CHECK-LABEL: tt2:
-; CHECK:       # BB#0:{{.*}} %entry
-; CHECK:    paddd (%[[REG3]]), %mm0
-; CHECK-NEXT:    movd %mm0, %rax
-; CHECK-NEXT:    emms
-; CHECK-NEXT:    retq
+define i64 @tt2(x86_mmx %t, x86_mmx* %q) nounwind {
+; X86-LABEL: tt2:
+; X86:       # BB#0: # %entry
+; X86-NEXT:    pushl %ebp
+; X86-NEXT:    movl %esp, %ebp
+; X86-NEXT:    andl $-8, %esp
+; X86-NEXT:    subl $8, %esp
+; X86-NEXT:    movl 8(%ebp), %eax
+; X86-NEXT:    paddd (%eax), %mm0
+; X86-NEXT:    movq %mm0, (%esp)
+; X86-NEXT:    movl (%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %edx
+; X86-NEXT:    emms
+; X86-NEXT:    movl %ebp, %esp
+; X86-NEXT:    popl %ebp
+; X86-NEXT:    retl
+;
+; X64-LABEL: tt2:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    paddd (%rdi), %mm0
+; X64-NEXT:    movd %mm0, %rax
+; X64-NEXT:    emms
+; X64-NEXT:    retq
 entry:
   %v = load x86_mmx, x86_mmx* %q
   %u = tail call x86_mmx @llvm.x86.mmx.padd.d(x86_mmx %t, x86_mmx %v)
@@ -185,13 +371,29 @@ entry:
 }
 declare x86_mmx @llvm.x86.mmx.padd.d(x86_mmx, x86_mmx)
 
-define i64 @tt3(x86_mmx %t, x86_mmx* %q) {
-; CHECK-LABEL: tt3:
-; CHECK:       # BB#0:{{.*}} %entry
-; CHECK:    paddq (%[[REG3]]), %mm0
-; CHECK-NEXT:    movd %mm0, %rax
-; CHECK-NEXT:    emms
-; CHECK-NEXT:    retq
+define i64 @tt3(x86_mmx %t, x86_mmx* %q) nounwind {
+; X86-LABEL: tt3:
+; X86:       # BB#0: # %entry
+; X86-NEXT:    pushl %ebp
+; X86-NEXT:    movl %esp, %ebp
+; X86-NEXT:    andl $-8, %esp
+; X86-NEXT:    subl $8, %esp
+; X86-NEXT:    movl 8(%ebp), %eax
+; X86-NEXT:    paddq (%eax), %mm0
+; X86-NEXT:    movq %mm0, (%esp)
+; X86-NEXT:    movl (%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %edx
+; X86-NEXT:    emms
+; X86-NEXT:    movl %ebp, %esp
+; X86-NEXT:    popl %ebp
+; X86-NEXT:    retl
+;
+; X64-LABEL: tt3:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    paddq (%rdi), %mm0
+; X64-NEXT:    movd %mm0, %rax
+; X64-NEXT:    emms
+; X64-NEXT:    retq
 entry:
   %v = load x86_mmx, x86_mmx* %q
   %u = tail call x86_mmx @llvm.x86.mmx.padd.q(x86_mmx %t, x86_mmx %v)
@@ -201,13 +403,29 @@ entry:
 }
 declare x86_mmx @llvm.x86.mmx.padd.q(x86_mmx, x86_mmx)
 
-define i64 @tt4(x86_mmx %t, x86_mmx* %q) {
-; CHECK-LABEL: tt4:
-; CHECK:       # BB#0:{{.*}} %entry
-; CHECK:    paddusb (%[[REG3]]), %mm0
-; CHECK-NEXT:    movd %mm0, %rax
-; CHECK-NEXT:    emms
-; CHECK-NEXT:    retq
+define i64 @tt4(x86_mmx %t, x86_mmx* %q) nounwind {
+; X86-LABEL: tt4:
+; X86:       # BB#0: # %entry
+; X86-NEXT:    pushl %ebp
+; X86-NEXT:    movl %esp, %ebp
+; X86-NEXT:    andl $-8, %esp
+; X86-NEXT:    subl $8, %esp
+; X86-NEXT:    movl 8(%ebp), %eax
+; X86-NEXT:    paddusb (%eax), %mm0
+; X86-NEXT:    movq %mm0, (%esp)
+; X86-NEXT:    movl (%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %edx
+; X86-NEXT:    emms
+; X86-NEXT:    movl %ebp, %esp
+; X86-NEXT:    popl %ebp
+; X86-NEXT:    retl
+;
+; X64-LABEL: tt4:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    paddusb (%rdi), %mm0
+; X64-NEXT:    movd %mm0, %rax
+; X64-NEXT:    emms
+; X64-NEXT:    retq
 entry:
   %v = load x86_mmx, x86_mmx* %q
   %u = tail call x86_mmx @llvm.x86.mmx.paddus.b(x86_mmx %t, x86_mmx %v)
@@ -217,13 +435,29 @@ entry:
 }
 declare x86_mmx @llvm.x86.mmx.paddus.b(x86_mmx, x86_mmx)
 
-define i64 @tt5(x86_mmx %t, x86_mmx* %q) {
-; CHECK-LABEL: tt5:
-; CHECK:       # BB#0:{{.*}} %entry
-; CHECK:    paddusw (%[[REG3]]), %mm0
-; CHECK-NEXT:    movd %mm0, %rax
-; CHECK-NEXT:    emms
-; CHECK-NEXT:    retq
+define i64 @tt5(x86_mmx %t, x86_mmx* %q) nounwind {
+; X86-LABEL: tt5:
+; X86:       # BB#0: # %entry
+; X86-NEXT:    pushl %ebp
+; X86-NEXT:    movl %esp, %ebp
+; X86-NEXT:    andl $-8, %esp
+; X86-NEXT:    subl $8, %esp
+; X86-NEXT:    movl 8(%ebp), %eax
+; X86-NEXT:    paddusw (%eax), %mm0
+; X86-NEXT:    movq %mm0, (%esp)
+; X86-NEXT:    movl (%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %edx
+; X86-NEXT:    emms
+; X86-NEXT:    movl %ebp, %esp
+; X86-NEXT:    popl %ebp
+; X86-NEXT:    retl
+;
+; X64-LABEL: tt5:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    paddusw (%rdi), %mm0
+; X64-NEXT:    movd %mm0, %rax
+; X64-NEXT:    emms
+; X64-NEXT:    retq
 entry:
   %v = load x86_mmx, x86_mmx* %q
   %u = tail call x86_mmx @llvm.x86.mmx.paddus.w(x86_mmx %t, x86_mmx %v)
@@ -233,13 +467,29 @@ entry:
 }
 declare x86_mmx @llvm.x86.mmx.paddus.w(x86_mmx, x86_mmx)
 
-define i64 @tt6(x86_mmx %t, x86_mmx* %q) {
-; CHECK-LABEL: tt6:
-; CHECK:       # BB#0:{{.*}} %entry
-; CHECK:    psrlw (%[[REG3]]), %mm0
-; CHECK-NEXT:    movd %mm0, %rax
-; CHECK-NEXT:    emms
-; CHECK-NEXT:    retq
+define i64 @tt6(x86_mmx %t, x86_mmx* %q) nounwind {
+; X86-LABEL: tt6:
+; X86:       # BB#0: # %entry
+; X86-NEXT:    pushl %ebp
+; X86-NEXT:    movl %esp, %ebp
+; X86-NEXT:    andl $-8, %esp
+; X86-NEXT:    subl $8, %esp
+; X86-NEXT:    movl 8(%ebp), %eax
+; X86-NEXT:    psrlw (%eax), %mm0
+; X86-NEXT:    movq %mm0, (%esp)
+; X86-NEXT:    movl (%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %edx
+; X86-NEXT:    emms
+; X86-NEXT:    movl %ebp, %esp
+; X86-NEXT:    popl %ebp
+; X86-NEXT:    retl
+;
+; X64-LABEL: tt6:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    psrlw (%rdi), %mm0
+; X64-NEXT:    movd %mm0, %rax
+; X64-NEXT:    emms
+; X64-NEXT:    retq
 entry:
   %v = load x86_mmx, x86_mmx* %q
   %u = tail call x86_mmx @llvm.x86.mmx.psrl.w(x86_mmx %t, x86_mmx %v)
@@ -249,13 +499,29 @@ entry:
 }
 declare x86_mmx @llvm.x86.mmx.psrl.w(x86_mmx, x86_mmx)
 
-define i64 @tt7(x86_mmx %t, x86_mmx* %q) {
-; CHECK-LABEL: tt7:
-; CHECK:       # BB#0:{{.*}} %entry
-; CHECK:    psrld (%[[REG3]]), %mm0
-; CHECK-NEXT:    movd %mm0, %rax
-; CHECK-NEXT:    emms
-; CHECK-NEXT:    retq
+define i64 @tt7(x86_mmx %t, x86_mmx* %q) nounwind {
+; X86-LABEL: tt7:
+; X86:       # BB#0: # %entry
+; X86-NEXT:    pushl %ebp
+; X86-NEXT:    movl %esp, %ebp
+; X86-NEXT:    andl $-8, %esp
+; X86-NEXT:    subl $8, %esp
+; X86-NEXT:    movl 8(%ebp), %eax
+; X86-NEXT:    psrld (%eax), %mm0
+; X86-NEXT:    movq %mm0, (%esp)
+; X86-NEXT:    movl (%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %edx
+; X86-NEXT:    emms
+; X86-NEXT:    movl %ebp, %esp
+; X86-NEXT:    popl %ebp
+; X86-NEXT:    retl
+;
+; X64-LABEL: tt7:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    psrld (%rdi), %mm0
+; X64-NEXT:    movd %mm0, %rax
+; X64-NEXT:    emms
+; X64-NEXT:    retq
 entry:
   %v = load x86_mmx, x86_mmx* %q
   %u = tail call x86_mmx @llvm.x86.mmx.psrl.d(x86_mmx %t, x86_mmx %v)
@@ -265,13 +531,29 @@ entry:
 }
 declare x86_mmx @llvm.x86.mmx.psrl.d(x86_mmx, x86_mmx)
 
-define i64 @tt8(x86_mmx %t, x86_mmx* %q) {
-; CHECK-LABEL: tt8:
-; CHECK:       # BB#0:{{.*}} %entry
-; CHECK:    psrlq (%[[REG3]]), %mm0
-; CHECK-NEXT:    movd %mm0, %rax
-; CHECK-NEXT:    emms
-; CHECK-NEXT:    retq
+define i64 @tt8(x86_mmx %t, x86_mmx* %q) nounwind {
+; X86-LABEL: tt8:
+; X86:       # BB#0: # %entry
+; X86-NEXT:    pushl %ebp
+; X86-NEXT:    movl %esp, %ebp
+; X86-NEXT:    andl $-8, %esp
+; X86-NEXT:    subl $8, %esp
+; X86-NEXT:    movl 8(%ebp), %eax
+; X86-NEXT:    psrlq (%eax), %mm0
+; X86-NEXT:    movq %mm0, (%esp)
+; X86-NEXT:    movl (%esp), %eax
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %edx
+; X86-NEXT:    emms
+; X86-NEXT:    movl %ebp, %esp
+; X86-NEXT:    popl %ebp
+; X86-NEXT:    retl
+;
+; X64-LABEL: tt8:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    psrlq (%rdi), %mm0
+; X64-NEXT:    movd %mm0, %rax
+; X64-NEXT:    emms
+; X64-NEXT:    retq
 entry:
   %v = load x86_mmx, x86_mmx* %q
   %u = tail call x86_mmx @llvm.x86.mmx.psrl.q(x86_mmx %t, x86_mmx %v)