1 ; RUN: llc < %s -march=arm -mattr=+neon > %t
2 ; RUN: grep {vmov\\.8} %t | count 2
3 ; RUN: grep {vmov\\.16} %t | count 2
4 ; RUN: grep {vmov\\.32} %t | count 2
5 ; RUN: grep {fcpys} %t | count 2
7 define <8 x i8> @vset_lane8(<8 x i8>* %A, i8 %B) nounwind {
8 %tmp1 = load <8 x i8>* %A
9 %tmp2 = insertelement <8 x i8> %tmp1, i8 %B, i32 1
13 define <4 x i16> @vset_lane16(<4 x i16>* %A, i16 %B) nounwind {
14 %tmp1 = load <4 x i16>* %A
15 %tmp2 = insertelement <4 x i16> %tmp1, i16 %B, i32 1
19 define <2 x i32> @vset_lane32(<2 x i32>* %A, i32 %B) nounwind {
20 %tmp1 = load <2 x i32>* %A
21 %tmp2 = insertelement <2 x i32> %tmp1, i32 %B, i32 1
25 define <16 x i8> @vsetQ_lane8(<16 x i8>* %A, i8 %B) nounwind {
26 %tmp1 = load <16 x i8>* %A
27 %tmp2 = insertelement <16 x i8> %tmp1, i8 %B, i32 1
31 define <8 x i16> @vsetQ_lane16(<8 x i16>* %A, i16 %B) nounwind {
32 %tmp1 = load <8 x i16>* %A
33 %tmp2 = insertelement <8 x i16> %tmp1, i16 %B, i32 1
37 define <4 x i32> @vsetQ_lane32(<4 x i32>* %A, i32 %B) nounwind {
38 %tmp1 = load <4 x i32>* %A
39 %tmp2 = insertelement <4 x i32> %tmp1, i32 %B, i32 1
43 define arm_aapcs_vfpcc <2 x float> @test_vset_lanef32(float %arg0_float32_t, <2 x float> %arg1_float32x2_t) nounwind {
45 %0 = insertelement <2 x float> %arg1_float32x2_t, float %arg0_float32_t, i32 1 ; <<2 x float>> [#uses=1]