1 files changed, 46 insertions, 38 deletions
diff --git a/test/CodeGen/X86/sse_partial_update.ll b/test/CodeGen/X86/sse_partial_update.ll
index 8d61428420f6..51359d1790af 100644
--- a/test/CodeGen/X86/sse_partial_update.ll
+++ b/test/CodeGen/X86/sse_partial_update.ll
@@ -1,3 +1,4 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
 ; RUN: llc < %s -mtriple=x86_64-apple-macosx -mattr=+sse2 -mcpu=nehalem | FileCheck %s
 
 ; rdar: 12558838
@@ -8,14 +9,15 @@
 ; destination of each scalar unary op are the same.
 
 define void @rsqrtss(<4 x float> %a) nounwind uwtable ssp {
-entry:
 ; CHECK-LABEL: rsqrtss:
-; CHECK: rsqrtss %xmm0, %xmm0
-; CHECK-NEXT: cvtss2sd %xmm0
-; CHECK-NEXT: movshdup
-; CHECK-NEXT: cvtss2sd %xmm0
-; CHECK-NEXT: movap
-; CHECK-NEXT: jmp
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    rsqrtss %xmm0, %xmm0
+; CHECK-NEXT:    cvtss2sd %xmm0, %xmm2
+; CHECK-NEXT:    movshdup {{.*#+}} xmm0 = xmm0[1,1,3,3]
+; CHECK-NEXT:    cvtss2sd %xmm0, %xmm1
+; CHECK-NEXT:    movaps %xmm2, %xmm0
+; CHECK-NEXT:    jmp _callee ## TAILCALL
+entry:
 
   %0 = tail call <4 x float> @llvm.x86.sse.rsqrt.ss(<4 x float> %a) nounwind
   %a.addr.0.extract = extractelement <4 x float> %0, i32 0
@@ -29,14 +31,15 @@ declare void @callee(double, double)
 declare <4 x float> @llvm.x86.sse.rsqrt.ss(<4 x float>) nounwind readnone
 
 define void @rcpss(<4 x float> %a) nounwind uwtable ssp {
-entry:
 ; CHECK-LABEL: rcpss:
-; CHECK: rcpss %xmm0, %xmm0
-; CHECK-NEXT: cvtss2sd %xmm0
-; CHECK-NEXT: movshdup
-; CHECK-NEXT: cvtss2sd %xmm0
-; CHECK-NEXT: movap
-; CHECK-NEXT: jmp
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    rcpss %xmm0, %xmm0
+; CHECK-NEXT:    cvtss2sd %xmm0, %xmm2
+; CHECK-NEXT:    movshdup {{.*#+}} xmm0 = xmm0[1,1,3,3]
+; CHECK-NEXT:    cvtss2sd %xmm0, %xmm1
+; CHECK-NEXT:    movaps %xmm2, %xmm0
+; CHECK-NEXT:    jmp _callee ## TAILCALL
+entry:
 
   %0 = tail call <4 x float> @llvm.x86.sse.rcp.ss(<4 x float> %a) nounwind
   %a.addr.0.extract = extractelement <4 x float> %0, i32 0
@@ -49,14 +52,15 @@ entry:
 declare <4 x float> @llvm.x86.sse.rcp.ss(<4 x float>) nounwind readnone
 
 define void @sqrtss(<4 x float> %a) nounwind uwtable ssp {
-entry:
 ; CHECK-LABEL: sqrtss:
-; CHECK: sqrtss %xmm0, %xmm0
-; CHECK-NEXT: cvtss2sd %xmm0
-; CHECK-NEXT: movshdup
-; CHECK-NEXT: cvtss2sd %xmm0
-; CHECK-NEXT: movap
-; CHECK-NEXT: jmp
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    sqrtss %xmm0, %xmm0
+; CHECK-NEXT:    cvtss2sd %xmm0, %xmm2
+; CHECK-NEXT:    movshdup {{.*#+}} xmm0 = xmm0[1,1,3,3]
+; CHECK-NEXT:    cvtss2sd %xmm0, %xmm1
+; CHECK-NEXT:    movaps %xmm2, %xmm0
+; CHECK-NEXT:    jmp _callee ## TAILCALL
+entry:
 
   %0 = tail call <4 x float> @llvm.x86.sse.sqrt.ss(<4 x float> %a) nounwind
   %a.addr.0.extract = extractelement <4 x float> %0, i32 0
@@ -69,14 +73,15 @@ entry:
 declare <4 x float> @llvm.x86.sse.sqrt.ss(<4 x float>) nounwind readnone
 
 define void @sqrtsd(<2 x double> %a) nounwind uwtable ssp {
-entry:
 ; CHECK-LABEL: sqrtsd:
-; CHECK: sqrtsd %xmm0, %xmm0
-; CHECK-NEXT: cvtsd2ss %xmm0
-; CHECK-NEXT: shufpd
-; CHECK-NEXT: cvtsd2ss %xmm0
-; CHECK-NEXT: movap
-; CHECK-NEXT: jmp
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    sqrtsd %xmm0, %xmm0
+; CHECK-NEXT:    cvtsd2ss %xmm0, %xmm2
+; CHECK-NEXT:    shufpd {{.*#+}} xmm0 = xmm0[1,0]
+; CHECK-NEXT:    cvtsd2ss %xmm0, %xmm1
+; CHECK-NEXT:    movaps %xmm2, %xmm0
+; CHECK-NEXT:    jmp _callee2 ## TAILCALL
+entry:
 
  %0 = tail call <2 x double> @llvm.x86.sse2.sqrt.sd(<2 x double> %a) nounwind
  %a0 = extractelement <2 x double> %0, i32 0
@@ -92,10 +97,11 @@ declare <2 x double> @llvm.x86.sse2.sqrt.sd(<2 x double>) nounwind readnone
 
 define <2 x double> @load_fold_cvtss2sd_int(<4 x float> *%a) {
 ; CHECK-LABEL: load_fold_cvtss2sd_int:
-; CHECK:       movaps   (%rdi), %xmm1
-; CHECK-NEXT:  xorps    %xmm0, %xmm0
-; CHECK-NEXT:  cvtss2sd %xmm1, %xmm0
-; CHECK-NEXT:  retq
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    movaps (%rdi), %xmm1
+; CHECK-NEXT:    xorps %xmm0, %xmm0
+; CHECK-NEXT:    cvtss2sd %xmm1, %xmm0
+; CHECK-NEXT:    retq
   %ld = load <4 x float>, <4 x float> *%a
   %x = call <2 x double> @llvm.x86.sse2.cvtss2sd(<2 x double> <double 0x0, double 0x0>, <4 x float> %ld)
   ret <2 x double> %x
@@ -103,9 +109,10 @@ define <2 x double> @load_fold_cvtss2sd_int(<4 x float> *%a) {
 
 define <2 x double> @load_fold_cvtss2sd_int_optsize(<4 x float> *%a) optsize {
 ; CHECK-LABEL: load_fold_cvtss2sd_int_optsize:
-; CHECK:       xorps    %xmm0, %xmm0
-; CHECK-NEXT:  cvtss2sd (%rdi), %xmm0
-; CHECK-NEXT:  retq
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    xorps %xmm0, %xmm0
+; CHECK-NEXT:    cvtss2sd (%rdi), %xmm0
+; CHECK-NEXT:    retq
   %ld = load <4 x float>, <4 x float> *%a
   %x = call <2 x double> @llvm.x86.sse2.cvtss2sd(<2 x double> <double 0x0, double 0x0>, <4 x float> %ld)
   ret <2 x double> %x
@@ -113,9 +120,10 @@ define <2 x double> @load_fold_cvtss2sd_int_optsize(<4 x float> *%a) optsize {
 
 define <2 x double> @load_fold_cvtss2sd_int_minsize(<4 x float> *%a) minsize {
 ; CHECK-LABEL: load_fold_cvtss2sd_int_minsize:
-; CHECK:       xorps    %xmm0, %xmm0
-; CHECK-NEXT:  cvtss2sd (%rdi), %xmm0
-; CHECK-NEXT:  retq
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    xorps %xmm0, %xmm0
+; CHECK-NEXT:    cvtss2sd (%rdi), %xmm0
+; CHECK-NEXT:    retq
   %ld = load <4 x float>, <4 x float> *%a
   %x = call <2 x double> @llvm.x86.sse2.cvtss2sd(<2 x double> <double 0x0, double 0x0>, <4 x float> %ld)
   ret <2 x double> %x