1 ; RUN: llc < %s -march=x86 -mcpu=core2 -o %t
2 ; RUN: grep movq %t | count 1
3 ; RUN: grep pshufd %t | count 1
4 ; RUN: grep movupd %t | count 1
5 ; RUN: grep pshufhw %t | count 1
7 define void @test_v4sf(<4 x float>* %P, float %X, float %Y) nounwind {
8 %tmp = insertelement <4 x float> zeroinitializer, float %X, i32 0 ; <<4 x float>> [#uses=1]
9 %tmp2 = insertelement <4 x float> %tmp, float %X, i32 1 ; <<4 x float>> [#uses=1]
10 %tmp4 = insertelement <4 x float> %tmp2, float %Y, i32 2 ; <<4 x float>> [#uses=1]
11 %tmp6 = insertelement <4 x float> %tmp4, float %Y, i32 3 ; <<4 x float>> [#uses=1]
12 store <4 x float> %tmp6, <4 x float>* %P
16 define void @test_v2sd(<2 x double>* %P, double %X, double %Y) nounwind {
17 %tmp = insertelement <2 x double> zeroinitializer, double %X, i32 0 ; <<2 x double>> [#uses=1]
18 %tmp2 = insertelement <2 x double> %tmp, double %Y, i32 1 ; <<2 x double>> [#uses=1]
19 store <2 x double> %tmp2, <2 x double>* %P
23 define void @test_v8i16(<2 x i64>* %res, <2 x i64>* %A) nounwind {
24 %tmp = load <2 x i64>* %A ; <<2 x i64>> [#uses=1]
25 %tmp.upgrd.1 = bitcast <2 x i64> %tmp to <8 x i16> ; <<8 x i16>> [#uses=8]
26 %tmp.upgrd.2 = extractelement <8 x i16> %tmp.upgrd.1, i32 0 ; <i16> [#uses=1]
27 %tmp1 = extractelement <8 x i16> %tmp.upgrd.1, i32 1 ; <i16> [#uses=1]
28 %tmp2 = extractelement <8 x i16> %tmp.upgrd.1, i32 2 ; <i16> [#uses=1]
29 %tmp3 = extractelement <8 x i16> %tmp.upgrd.1, i32 3 ; <i16> [#uses=1]
30 %tmp4 = extractelement <8 x i16> %tmp.upgrd.1, i32 6 ; <i16> [#uses=1]
31 %tmp5 = extractelement <8 x i16> %tmp.upgrd.1, i32 5 ; <i16> [#uses=1]
32 %tmp6 = extractelement <8 x i16> %tmp.upgrd.1, i32 4 ; <i16> [#uses=1]
33 %tmp7 = extractelement <8 x i16> %tmp.upgrd.1, i32 7 ; <i16> [#uses=1]
34 %tmp8 = insertelement <8 x i16> undef, i16 %tmp.upgrd.2, i32 0 ; <<8 x i16>> [#uses=1]
35 %tmp9 = insertelement <8 x i16> %tmp8, i16 %tmp1, i32 1 ; <<8 x i16>> [#uses=1]
36 %tmp10 = insertelement <8 x i16> %tmp9, i16 %tmp2, i32 2 ; <<8 x i16>> [#uses=1]
37 %tmp11 = insertelement <8 x i16> %tmp10, i16 %tmp3, i32 3 ; <<8 x i16>> [#uses=1]
38 %tmp12 = insertelement <8 x i16> %tmp11, i16 %tmp4, i32 4 ; <<8 x i16>> [#uses=1]
39 %tmp13 = insertelement <8 x i16> %tmp12, i16 %tmp5, i32 5 ; <<8 x i16>> [#uses=1]
40 %tmp14 = insertelement <8 x i16> %tmp13, i16 %tmp6, i32 6 ; <<8 x i16>> [#uses=1]
41 %tmp15 = insertelement <8 x i16> %tmp14, i16 %tmp7, i32 7 ; <<8 x i16>> [#uses=1]
42 %tmp15.upgrd.3 = bitcast <8 x i16> %tmp15 to <2 x i64> ; <<2 x i64>> [#uses=1]
43 store <2 x i64> %tmp15.upgrd.3, <2 x i64>* %res