-; RUN: llvm-as < %s | llc -march=x86-64 | grep unpck | count 2
-; RUN: llvm-as < %s | llc -march=x86-64 | grep shuf | count 2
-; RUN: llvm-as < %s | llc -march=x86-64 | grep ps | count 4
-; RUN: llvm-as < %s | llc -march=x86-64 | grep pd | count 4
-; RUN: llvm-as < %s | llc -march=x86-64 | grep movup | count 4
+; RUN: llc < %s -march=x86-64 | FileCheck %s
-define <4 x float> @a(<4 x float>* %y)
-{
+; CHECK: a:
+; CHECK: movdqu
+; CHECK: pshufd
+define <4 x float> @a(<4 x float>* %y) nounwind {
%x = load <4 x float>* %y, align 4
%a = extractelement <4 x float> %x, i32 0
%b = extractelement <4 x float> %x, i32 1
%s = insertelement <4 x float> %r, float %a, i32 3
ret <4 x float> %s
}
-define <4 x float> @b(<4 x float>* %y, <4 x float> %z)
-{
+
+; CHECK: b:
+; CHECK: movups
+; CHECK: unpckhps
+define <4 x float> @b(<4 x float>* %y, <4 x float> %z) nounwind {
%x = load <4 x float>* %y, align 4
%a = extractelement <4 x float> %x, i32 2
%b = extractelement <4 x float> %x, i32 3
%s = insertelement <4 x float> %r, float %b, i32 3
ret <4 x float> %s
}
-define <2 x double> @c(<2 x double>* %y)
-{
+
+; CHECK: c:
+; CHECK: movupd
+; CHECK: shufpd
+define <2 x double> @c(<2 x double>* %y) nounwind {
%x = load <2 x double>* %y, align 8
%a = extractelement <2 x double> %x, i32 0
%c = extractelement <2 x double> %x, i32 1
%r = insertelement <2 x double> %p, double %a, i32 1
ret <2 x double> %r
}
-define <2 x double> @d(<2 x double>* %y, <2 x double> %z)
-{
+
+; CHECK: d:
+; CHECK: movupd
+; CHECK: unpckhpd
+define <2 x double> @d(<2 x double>* %y, <2 x double> %z) nounwind {
%x = load <2 x double>* %y, align 8
%a = extractelement <2 x double> %x, i32 1
%c = extractelement <2 x double> %z, i32 1