-; RUN: llvm-as < %s | llc -march=x86 -mattr=+sse2 | grep pshufhw | wc -l | grep 1 &&
-; RUN: llvm-as < %s | llc -march=x86 -mattr=+sse2 | grep pshuflw | wc -l | grep 1 &&
-; RUN: llvm-as < %s | llc -march=x86 -mattr=+sse2 | grep movhps | wc -l | grep 1
+; RUN: llvm-as < %s | llc -march=x86 -mattr=+sse2 -o %t -f
+; RUN: grep pshufhw %t | count 1
+; RUN: grep pshuflw %t | count 1
+; RUN: grep movhps %t | count 1
-void %test1(<2 x long>* %res, <2 x long>* %A) {
- %tmp = load <2 x long>* %A
- %tmp = cast <2 x long> %tmp to <8 x short>
- %tmp0 = extractelement <8 x short> %tmp, uint 0
- %tmp1 = extractelement <8 x short> %tmp, uint 1
- %tmp2 = extractelement <8 x short> %tmp, uint 2
- %tmp3 = extractelement <8 x short> %tmp, uint 3
- %tmp4 = extractelement <8 x short> %tmp, uint 4
- %tmp5 = extractelement <8 x short> %tmp, uint 5
- %tmp6 = extractelement <8 x short> %tmp, uint 6
- %tmp7 = extractelement <8 x short> %tmp, uint 7
- %tmp8 = insertelement <8 x short> undef, short %tmp2, uint 0
- %tmp9 = insertelement <8 x short> %tmp8, short %tmp1, uint 1
- %tmp10 = insertelement <8 x short> %tmp9, short %tmp0, uint 2
- %tmp11 = insertelement <8 x short> %tmp10, short %tmp3, uint 3
- %tmp12 = insertelement <8 x short> %tmp11, short %tmp6, uint 4
- %tmp13 = insertelement <8 x short> %tmp12, short %tmp5, uint 5
- %tmp14 = insertelement <8 x short> %tmp13, short %tmp4, uint 6
- %tmp15 = insertelement <8 x short> %tmp14, short %tmp7, uint 7
- %tmp15 = cast <8 x short> %tmp15 to <2 x long>
- store <2 x long> %tmp15, <2 x long>* %res
+define void @test1(<2 x i64>* %res, <2 x i64>* %A) nounwind {
+ %tmp = load <2 x i64>* %A ; <<2 x i64>> [#uses=1]
+ %tmp.upgrd.1 = bitcast <2 x i64> %tmp to <8 x i16> ; <<8 x i16>> [#uses=8]
+ %tmp0 = extractelement <8 x i16> %tmp.upgrd.1, i32 0 ; <i16> [#uses=1]
+ %tmp1 = extractelement <8 x i16> %tmp.upgrd.1, i32 1 ; <i16> [#uses=1]
+ %tmp2 = extractelement <8 x i16> %tmp.upgrd.1, i32 2 ; <i16> [#uses=1]
+ %tmp3 = extractelement <8 x i16> %tmp.upgrd.1, i32 3 ; <i16> [#uses=1]
+ %tmp4 = extractelement <8 x i16> %tmp.upgrd.1, i32 4 ; <i16> [#uses=1]
+ %tmp5 = extractelement <8 x i16> %tmp.upgrd.1, i32 5 ; <i16> [#uses=1]
+ %tmp6 = extractelement <8 x i16> %tmp.upgrd.1, i32 6 ; <i16> [#uses=1]
+ %tmp7 = extractelement <8 x i16> %tmp.upgrd.1, i32 7 ; <i16> [#uses=1]
+ %tmp8 = insertelement <8 x i16> undef, i16 %tmp2, i32 0 ; <<8 x i16>> [#uses=1]
+ %tmp9 = insertelement <8 x i16> %tmp8, i16 %tmp1, i32 1 ; <<8 x i16>> [#uses=1]
+ %tmp10 = insertelement <8 x i16> %tmp9, i16 %tmp0, i32 2 ; <<8 x i16>> [#uses=1]
+ %tmp11 = insertelement <8 x i16> %tmp10, i16 %tmp3, i32 3 ; <<8 x i16>> [#uses=1]
+ %tmp12 = insertelement <8 x i16> %tmp11, i16 %tmp6, i32 4 ; <<8 x i16>> [#uses=1]
+ %tmp13 = insertelement <8 x i16> %tmp12, i16 %tmp5, i32 5 ; <<8 x i16>> [#uses=1]
+ %tmp14 = insertelement <8 x i16> %tmp13, i16 %tmp4, i32 6 ; <<8 x i16>> [#uses=1]
+ %tmp15 = insertelement <8 x i16> %tmp14, i16 %tmp7, i32 7 ; <<8 x i16>> [#uses=1]
+ %tmp15.upgrd.2 = bitcast <8 x i16> %tmp15 to <2 x i64> ; <<2 x i64>> [#uses=1]
+ store <2 x i64> %tmp15.upgrd.2, <2 x i64>* %res
ret void
}
-void %test2(<4 x float>* %r, <2 x int>* %A) {
- %tmp = load <4 x float>* %r
- %tmp = cast <2 x int>* %A to double*
- %tmp = load double* %tmp
- %tmp = insertelement <2 x double> undef, double %tmp, uint 0
- %tmp5 = insertelement <2 x double> %tmp, double undef, uint 1
- %tmp6 = cast <2 x double> %tmp5 to <4 x float>
- %tmp = extractelement <4 x float> %tmp, uint 0
- %tmp7 = extractelement <4 x float> %tmp, uint 1
- %tmp8 = extractelement <4 x float> %tmp6, uint 0
- %tmp9 = extractelement <4 x float> %tmp6, uint 1
- %tmp10 = insertelement <4 x float> undef, float %tmp, uint 0
- %tmp11 = insertelement <4 x float> %tmp10, float %tmp7, uint 1
- %tmp12 = insertelement <4 x float> %tmp11, float %tmp8, uint 2
- %tmp13 = insertelement <4 x float> %tmp12, float %tmp9, uint 3
+define void @test2(<4 x float>* %r, <2 x i32>* %A) nounwind {
+ %tmp = load <4 x float>* %r ; <<4 x float>> [#uses=2]
+ %tmp.upgrd.3 = bitcast <2 x i32>* %A to double* ; <double*> [#uses=1]
+ %tmp.upgrd.4 = load double* %tmp.upgrd.3 ; <double> [#uses=1]
+ %tmp.upgrd.5 = insertelement <2 x double> undef, double %tmp.upgrd.4, i32 0 ; <<2 x double>> [#uses=1]
+ %tmp5 = insertelement <2 x double> %tmp.upgrd.5, double undef, i32 1 ; <<2 x double>> [#uses=1]
+ %tmp6 = bitcast <2 x double> %tmp5 to <4 x float> ; <<4 x float>> [#uses=2]
+ %tmp.upgrd.6 = extractelement <4 x float> %tmp, i32 0 ; <float> [#uses=1]
+ %tmp7 = extractelement <4 x float> %tmp, i32 1 ; <float> [#uses=1]
+ %tmp8 = extractelement <4 x float> %tmp6, i32 0 ; <float> [#uses=1]
+ %tmp9 = extractelement <4 x float> %tmp6, i32 1 ; <float> [#uses=1]
+ %tmp10 = insertelement <4 x float> undef, float %tmp.upgrd.6, i32 0 ; <<4 x float>> [#uses=1]
+ %tmp11 = insertelement <4 x float> %tmp10, float %tmp7, i32 1 ; <<4 x float>> [#uses=1]
+ %tmp12 = insertelement <4 x float> %tmp11, float %tmp8, i32 2 ; <<4 x float>> [#uses=1]
+ %tmp13 = insertelement <4 x float> %tmp12, float %tmp9, i32 3 ; <<4 x float>> [#uses=1]
store <4 x float> %tmp13, <4 x float>* %r
ret void
}