1 ; RUN: llc < %s -verify-machineinstrs -mtriple=arm64-none-linux-gnu -mattr=+neon | FileCheck %s
2 ; arm64 has a separate copy due to intrinsics
4 define <4 x i32> @copyTuple.QPair(i32* %a, i32* %b) {
5 ; CHECK-LABEL: copyTuple.QPair:
6 ; CHECK: mov v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
7 ; CHECK: mov v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
8 ; CHECK: ld2 { {{v[0-9]+}}.s, {{v[0-9]+}}.s }[{{[0-9]+}}], [x{{[0-9]+|sp}}]
10 %vld = tail call { <4 x i32>, <4 x i32> } @llvm.aarch64.neon.ld2lane.v4i32.p0i32(<4 x i32> <i32 -1, i32 -1, i32 -1, i32 -1>, <4 x i32> <i32 2, i32 2, i32 2, i32 2>, i64 1, i32* %a)
11 %extract = extractvalue { <4 x i32>, <4 x i32> } %vld, 0
12 %vld1 = tail call { <4 x i32>, <4 x i32> } @llvm.aarch64.neon.ld2lane.v4i32.p0i32(<4 x i32> %extract, <4 x i32> <i32 -1, i32 -1, i32 -1, i32 -1>, i64 1, i32* %b)
13 %vld1.fca.0.extract = extractvalue { <4 x i32>, <4 x i32> } %vld1, 0
14 ret <4 x i32> %vld1.fca.0.extract
17 define <4 x i32> @copyTuple.QTriple(i32* %a, i32* %b, <4 x i32> %c) {
18 ; CHECK-LABEL: copyTuple.QTriple:
19 ; CHECK: mov v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
20 ; CHECK: mov v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
21 ; CHECK: mov v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
22 ; CHECK: ld3 { {{v[0-9]+}}.s, {{v[0-9]+}}.s, {{v[0-9]+}}.s }[{{[0-9]+}}], [x{{[0-9]+|sp}}]
24 %vld = tail call { <4 x i32>, <4 x i32>, <4 x i32> } @llvm.aarch64.neon.ld3lane.v4i32.p0i32(<4 x i32> <i32 -1, i32 -1, i32 -1, i32 -1>, <4 x i32> %c, <4 x i32> %c, i64 1, i32* %a)
25 %extract = extractvalue { <4 x i32>, <4 x i32>, <4 x i32> } %vld, 0
26 %vld1 = tail call { <4 x i32>, <4 x i32>, <4 x i32> } @llvm.aarch64.neon.ld3lane.v4i32.p0i32(<4 x i32> %extract, <4 x i32> <i32 -1, i32 -1, i32 -1, i32 -1>, <4 x i32> %c, i64 1, i32* %b)
27 %vld1.fca.0.extract = extractvalue { <4 x i32>, <4 x i32>, <4 x i32> } %vld1, 0
28 ret <4 x i32> %vld1.fca.0.extract
31 define <4 x i32> @copyTuple.QQuad(i32* %a, i32* %b, <4 x i32> %c) {
32 ; CHECK-LABEL: copyTuple.QQuad:
33 ; CHECK: mov v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
34 ; CHECK: mov v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
35 ; CHECK: mov v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
36 ; CHECK: mov v{{[0-9]+}}.16b, v{{[0-9]+}}.16b
37 ; CHECK: ld4 { {{v[0-9]+}}.s, {{v[0-9]+}}.s, {{v[0-9]+}}.s, {{v[0-9]+}}.s }[{{[0-9]+}}], [x{{[0-9]+|sp}}]
39 %vld = tail call { <4 x i32>, <4 x i32>, <4 x i32>, <4 x i32> } @llvm.aarch64.neon.ld4lane.v4i32.p0i32(<4 x i32> <i32 -1, i32 -1, i32 -1, i32 -1>, <4 x i32> %c, <4 x i32> %c, <4 x i32> %c, i64 1, i32* %a)
40 %extract = extractvalue { <4 x i32>, <4 x i32>, <4 x i32>, <4 x i32> } %vld, 0
41 %vld1 = tail call { <4 x i32>, <4 x i32>, <4 x i32>, <4 x i32> } @llvm.aarch64.neon.ld4lane.v4i32.p0i32(<4 x i32> %extract, <4 x i32> <i32 -1, i32 -1, i32 -1, i32 -1>, <4 x i32> %c, <4 x i32> %c, i64 1, i32* %b)
42 %vld1.fca.0.extract = extractvalue { <4 x i32>, <4 x i32>, <4 x i32>, <4 x i32> } %vld1, 0
43 ret <4 x i32> %vld1.fca.0.extract
46 declare { <4 x i32>, <4 x i32> } @llvm.aarch64.neon.ld2lane.v4i32.p0i32(<4 x i32>, <4 x i32>, i64, i32*)
47 declare { <4 x i32>, <4 x i32>, <4 x i32> } @llvm.aarch64.neon.ld3lane.v4i32.p0i32(<4 x i32>, <4 x i32>, <4 x i32>, i64, i32*)
48 declare { <4 x i32>, <4 x i32>, <4 x i32>, <4 x i32> } @llvm.aarch64.neon.ld4lane.v4i32.p0i32(<4 x i32>, <4 x i32>, <4 x i32>, <4 x i32>, i64, i32*)