1 ; RUN: llc < %s -march=arm -mattr=+neon | FileCheck %s
3 define <8 x i8> @vld1dupi8(i8* %A) nounwind {
5 ;Check the (default) alignment value.
6 ;CHECK: vld1.8 {d16[]}, [r0]
7 %tmp1 = load i8* %A, align 8
8 %tmp2 = insertelement <8 x i8> undef, i8 %tmp1, i32 0
9 %tmp3 = shufflevector <8 x i8> %tmp2, <8 x i8> undef, <8 x i32> zeroinitializer
13 define <4 x i16> @vld1dupi16(i16* %A) nounwind {
15 ;Check the alignment value. Max for this instruction is 16 bits:
16 ;CHECK: vld1.16 {d16[]}, [r0, :16]
17 %tmp1 = load i16* %A, align 8
18 %tmp2 = insertelement <4 x i16> undef, i16 %tmp1, i32 0
19 %tmp3 = shufflevector <4 x i16> %tmp2, <4 x i16> undef, <4 x i32> zeroinitializer
23 define <2 x i32> @vld1dupi32(i32* %A) nounwind {
25 ;Check the alignment value. Max for this instruction is 32 bits:
26 ;CHECK: vld1.32 {d16[]}, [r0, :32]
27 %tmp1 = load i32* %A, align 8
28 %tmp2 = insertelement <2 x i32> undef, i32 %tmp1, i32 0
29 %tmp3 = shufflevector <2 x i32> %tmp2, <2 x i32> undef, <2 x i32> zeroinitializer
33 define <16 x i8> @vld1dupQi8(i8* %A) nounwind {
35 ;Check the (default) alignment value.
36 ;CHECK: vld1.8 {d16[], d17[]}, [r0]
37 %tmp1 = load i8* %A, align 8
38 %tmp2 = insertelement <16 x i8> undef, i8 %tmp1, i32 0
39 %tmp3 = shufflevector <16 x i8> %tmp2, <16 x i8> undef, <16 x i32> zeroinitializer