1 files changed, 43 insertions, 12 deletions
diff --git a/llvm/test/CodeGen/X86/fma4-intrinsics-x86_64-folded-load.ll b/llvm/test/CodeGen/X86/fma4-intrinsics-x86_64-folded-load.ll
index df4b48b769b..5b86a7d7eea 100644
--- a/llvm/test/CodeGen/X86/fma4-intrinsics-x86_64-folded-load.ll
+++ b/llvm/test/CodeGen/X86/fma4-intrinsics-x86_64-folded-load.ll
@@ -1,16 +1,23 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mcpu=corei7-avx -mattr=+fma4 | FileCheck %s
 ; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mcpu=bdver2 -mattr=+avx,-fma | FileCheck %s
 
 ; VFMADD
 define < 4 x float > @test_x86_fma_vfmadd_ss_load(< 4 x float > %a0, < 4 x float > %a1, float* %a2) {
-  ; CHECK: vfmaddss (%{{.*}})
+; CHECK-LABEL: test_x86_fma_vfmadd_ss_load:
+; CHECK:       # BB#0:
+; CHECK-NEXT:    vfmaddss (%rdi), %xmm1, %xmm0, %xmm0
+; CHECK-NEXT:    retq
   %x = load float , float *%a2
   %y = insertelement <4 x float> undef, float %x, i32 0
   %res = call < 4 x float > @llvm.x86.fma.vfmadd.ss(< 4 x float > %a0, < 4 x float > %a1, < 4 x float > %y)
   ret < 4 x float > %res
 }
 define < 4 x float > @test_x86_fma_vfmadd_ss_load2(< 4 x float > %a0, float* %a1, < 4 x float > %a2) {
-  ; CHECK: vfmaddss %{{.*}}, (%{{.*}})
+; CHECK-LABEL: test_x86_fma_vfmadd_ss_load2:
+; CHECK:       # BB#0:
+; CHECK-NEXT:    vfmaddss %xmm1, (%rdi), %xmm0, %xmm0
+; CHECK-NEXT:    retq
   %x = load float , float *%a1
   %y = insertelement <4 x float> undef, float %x, i32 0
   %res = call < 4 x float > @llvm.x86.fma.vfmadd.ss(< 4 x float > %a0, < 4 x float > %y, < 4 x float > %a2)
@@ -20,14 +27,20 @@ define < 4 x float > @test_x86_fma_vfmadd_ss_load2(< 4 x float > %a0, float* %a1
 declare < 4 x float > @llvm.x86.fma.vfmadd.ss(< 4 x float >, < 4 x float >, < 4 x float >) nounwind readnone
 
 define < 2 x double > @test_x86_fma_vfmadd_sd_load(< 2 x double > %a0, < 2 x double > %a1, double* %a2) {
-  ; CHECK: vfmaddsd (%{{.*}})
+; CHECK-LABEL: test_x86_fma_vfmadd_sd_load:
+; CHECK:       # BB#0:
+; CHECK-NEXT:    vfmaddsd (%rdi), %xmm1, %xmm0, %xmm0
+; CHECK-NEXT:    retq
   %x = load double , double *%a2
   %y = insertelement <2 x double> undef, double %x, i32 0
   %res = call < 2 x double > @llvm.x86.fma.vfmadd.sd(< 2 x double > %a0, < 2 x double > %a1, < 2 x double > %y)
   ret < 2 x double > %res
 }
 define < 2 x double > @test_x86_fma_vfmadd_sd_load2(< 2 x double > %a0, double* %a1, < 2 x double > %a2) {
-  ; CHECK: vfmaddsd %{{.*}}, (%{{.*}})
+; CHECK-LABEL: test_x86_fma_vfmadd_sd_load2:
+; CHECK:       # BB#0:
+; CHECK-NEXT:    vfmaddsd %xmm1, (%rdi), %xmm0, %xmm0
+; CHECK-NEXT:    retq
   %x = load double , double *%a1
   %y = insertelement <2 x double> undef, double %x, i32 0
   %res = call < 2 x double > @llvm.x86.fma.vfmadd.sd(< 2 x double > %a0, < 2 x double > %y, < 2 x double > %a2)
@@ -35,13 +48,19 @@ define < 2 x double > @test_x86_fma_vfmadd_sd_load2(< 2 x double > %a0, double*
 }
 declare < 2 x double > @llvm.x86.fma.vfmadd.sd(< 2 x double >, < 2 x double >, < 2 x double >) nounwind readnone
 define < 4 x float > @test_x86_fma_vfmadd_ps_load(< 4 x float > %a0, < 4 x float > %a1, < 4 x float >* %a2) {
-  ; CHECK: vfmaddps (%{{.*}})
+; CHECK-LABEL: test_x86_fma_vfmadd_ps_load:
+; CHECK:       # BB#0:
+; CHECK-NEXT:    vfmaddps (%rdi), %xmm1, %xmm0, %xmm0
+; CHECK-NEXT:    retq
   %x = load <4 x float>, <4 x float>* %a2
   %res = call < 4 x float > @llvm.x86.fma.vfmadd.ps(< 4 x float > %a0, < 4 x float > %a1, < 4 x float > %x)
   ret < 4 x float > %res
 }
 define < 4 x float > @test_x86_fma_vfmadd_ps_load2(< 4 x float > %a0, < 4 x float >* %a1, < 4 x float > %a2) {
-  ; CHECK: vfmaddps %{{.*}}, (%{{.*}})
+; CHECK-LABEL: test_x86_fma_vfmadd_ps_load2:
+; CHECK:       # BB#0:
+; CHECK-NEXT:    vfmaddps %xmm1, (%rdi), %xmm0, %xmm0
+; CHECK-NEXT:    retq
   %x = load <4 x float>, <4 x float>* %a1
   %res = call < 4 x float > @llvm.x86.fma.vfmadd.ps(< 4 x float > %a0, < 4 x float > %x, < 4 x float > %a2)
   ret < 4 x float > %res
@@ -50,8 +69,11 @@ declare < 4 x float > @llvm.x86.fma.vfmadd.ps(< 4 x float >, < 4 x float >, < 4
 
 ; To test execution dependency
 define < 4 x float > @test_x86_fma_vfmadd_ps_load3(< 4 x float >* %a0, < 4 x float >* %a1, < 4 x float > %a2) {
-  ; CHECK: vmovaps
-  ; CHECK: vfmaddps %{{.*}}, (%{{.*}})
+; CHECK-LABEL: test_x86_fma_vfmadd_ps_load3:
+; CHECK:       # BB#0:
+; CHECK-NEXT:    vmovaps (%rdi), %xmm1
+; CHECK-NEXT:    vfmaddps %xmm0, (%rsi), %xmm1, %xmm0
+; CHECK-NEXT:    retq
   %x = load <4 x float>, <4 x float>* %a0
   %y = load <4 x float>, <4 x float>* %a1
   %res = call < 4 x float > @llvm.x86.fma.vfmadd.ps(< 4 x float > %x, < 4 x float > %y, < 4 x float > %a2)
@@ -59,13 +81,19 @@ define < 4 x float > @test_x86_fma_vfmadd_ps_load3(< 4 x float >* %a0, < 4 x flo
 }
 
 define < 2 x double > @test_x86_fma_vfmadd_pd_load(< 2 x double > %a0, < 2 x double > %a1, < 2 x double >* %a2) {
-  ; CHECK: vfmaddpd (%{{.*}})
+; CHECK-LABEL: test_x86_fma_vfmadd_pd_load:
+; CHECK:       # BB#0:
+; CHECK-NEXT:    vfmaddpd (%rdi), %xmm1, %xmm0, %xmm0
+; CHECK-NEXT:    retq
   %x = load <2 x double>, <2 x double>* %a2
   %res = call < 2 x double > @llvm.x86.fma.vfmadd.pd(< 2 x double > %a0, < 2 x double > %a1, < 2 x double > %x)
   ret < 2 x double > %res
 }
 define < 2 x double > @test_x86_fma_vfmadd_pd_load2(< 2 x double > %a0, < 2 x double >* %a1, < 2 x double > %a2) {
-  ; CHECK: vfmaddpd %{{.*}}, (%{{.*}})
+; CHECK-LABEL: test_x86_fma_vfmadd_pd_load2:
+; CHECK:       # BB#0:
+; CHECK-NEXT:    vfmaddpd %xmm1, (%rdi), %xmm0, %xmm0
+; CHECK-NEXT:    retq
   %x = load <2 x double>, <2 x double>* %a1
   %res = call < 2 x double > @llvm.x86.fma.vfmadd.pd(< 2 x double > %a0, < 2 x double > %x, < 2 x double > %a2)
   ret < 2 x double > %res
@@ -74,8 +102,11 @@ declare < 2 x double > @llvm.x86.fma.vfmadd.pd(< 2 x double >, < 2 x double >, <
 
 ; To test execution dependency
 define < 2 x double > @test_x86_fma_vfmadd_pd_load3(< 2 x double >* %a0, < 2 x double >* %a1, < 2 x double > %a2) {
-  ; CHECK: vmovapd
-  ; CHECK: vfmaddpd %{{.*}}, (%{{.*}})
+; CHECK-LABEL: test_x86_fma_vfmadd_pd_load3:
+; CHECK:       # BB#0:
+; CHECK-NEXT:    vmovapd (%rdi), %xmm1
+; CHECK-NEXT:    vfmaddpd %xmm0, (%rsi), %xmm1, %xmm0
+; CHECK-NEXT:    retq
   %x = load <2 x double>, <2 x double>* %a0
   %y = load <2 x double>, <2 x double>* %a1
   %res = call < 2 x double > @llvm.x86.fma.vfmadd.pd(< 2 x double > %x, < 2 x double > %y, < 2 x double > %a2)