Fix a broadcast related regression on the vector shuffle lowering.

[oota-llvm.git] / test / CodeGen / X86 / break-sse-dep.ll

diff --git a/test/CodeGen/X86/break-sse-dep.ll b/test/CodeGen/X86/break-sse-dep.ll
index 027d2f1dafc1da321954c338bf4038bd743a7e65..8124d6f522633e1e63cbf8fb9a6e5341d6a93a78 100644 (file)
--- a/test/CodeGen/X86/break-sse-dep.ll
+++ b/test/CodeGen/X86/break-sse-dep.ll
@@ -1,9 +1,10 @@
-; RUN: llc < %s -march=x86-64 -mattr=+sse2 | FileCheck %s
+; RUN: llc < %s -mtriple=x86_64-linux -mattr=+sse2 -mcpu=nehalem | FileCheck %s
+; RUN: llc < %s -mtriple=x86_64-win32 -mattr=+sse2 -mcpu=nehalem | FileCheck %s
 
 define double @t1(float* nocapture %x) nounwind readonly ssp {
 entry:
-; CHECK: t1:
-; CHECK: movss (%rdi), %xmm0
+; CHECK-LABEL: t1:
+; CHECK: movss ([[A0:%rdi|%rcx]]), %xmm0
 ; CHECK: cvtss2sd %xmm0, %xmm0
 
   %0 = load float* %x, align 4
@@ -13,9 +14,49 @@ entry:
 
 define float @t2(double* nocapture %x) nounwind readonly ssp optsize {
 entry:
-; CHECK: t2:
-; CHECK: cvtsd2ss (%rdi), %xmm0
+; CHECK-LABEL: t2:
+; CHECK: cvtsd2ss ([[A0]]), %xmm0
   %0 = load double* %x, align 8
   %1 = fptrunc double %0 to float
   ret float %1
 }
+
+define float @squirtf(float* %x) nounwind {
+entry:
+; CHECK-LABEL: squirtf:
+; CHECK: movss ([[A0]]), %xmm0
+; CHECK: sqrtss %xmm0, %xmm0
+  %z = load float* %x
+  %t = call float @llvm.sqrt.f32(float %z)
+  ret float %t
+}
+
+define double @squirt(double* %x) nounwind {
+entry:
+; CHECK-LABEL: squirt:
+; CHECK: sqrtsd ([[A0]]), %xmm0
+  %z = load double* %x
+  %t = call double @llvm.sqrt.f64(double %z)
+  ret double %t
+}
+
+define float @squirtf_size(float* %x) nounwind optsize {
+entry:
+; CHECK-LABEL: squirtf_size:
+; CHECK: sqrtss ([[A0]]), %xmm0
+  %z = load float* %x
+  %t = call float @llvm.sqrt.f32(float %z)
+  ret float %t
+}
+
+define double @squirt_size(double* %x) nounwind optsize {
+entry:
+; CHECK-LABEL: squirt_size:
+; CHECK: sqrtsd ([[A0]]), %xmm0
+  %z = load double* %x
+  %t = call double @llvm.sqrt.f64(double %z)
+  ret double %t
+}
+
+declare float @llvm.sqrt.f32(float)
+declare double @llvm.sqrt.f64(double)