1; RUN: llc < %s -mtriple=armv8-linux-gnueabi -verify-machineinstrs \
2; RUN:     -asm-verbose=false | FileCheck %s
3
4%struct.uint16x4x2_t = type { <4 x i16>, <4 x i16> }
5%struct.uint16x4x3_t = type { <4 x i16>, <4 x i16>, <4 x i16> }
6%struct.uint16x4x4_t = type { <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16> }
7
8%struct.uint32x2x2_t = type { <2 x i32>, <2 x i32> }
9%struct.uint32x2x3_t = type { <2 x i32>, <2 x i32>, <2 x i32> }
10%struct.uint32x2x4_t = type { <2 x i32>, <2 x i32>, <2 x i32>, <2 x i32> }
11
12%struct.uint64x1x2_t = type { <1 x i64>, <1 x i64> }
13%struct.uint64x1x3_t = type { <1 x i64>, <1 x i64>, <1 x i64> }
14%struct.uint64x1x4_t = type { <1 x i64>, <1 x i64>, <1 x i64>, <1 x i64> }
15
16%struct.uint8x8x2_t = type { <8 x i8>, <8 x i8> }
17%struct.uint8x8x3_t = type { <8 x i8>, <8 x i8>, <8 x i8> }
18%struct.uint8x8x4_t = type { <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8> }
19
20%struct.uint16x8x2_t = type { <8 x i16>, <8 x i16> }
21%struct.uint16x8x3_t = type { <8 x i16>, <8 x i16>, <8 x i16> }
22%struct.uint16x8x4_t = type { <8 x i16>, <8 x i16>, <8 x i16>, <8 x i16> }
23
24%struct.uint32x4x2_t = type { <4 x i32>, <4 x i32> }
25%struct.uint32x4x3_t = type { <4 x i32>, <4 x i32>, <4 x i32> }
26%struct.uint32x4x4_t = type { <4 x i32>, <4 x i32>, <4 x i32>, <4 x i32> }
27
28%struct.uint8x16x2_t = type { <16 x i8>, <16 x i8> }
29%struct.uint8x16x3_t = type { <16 x i8>, <16 x i8>, <16 x i8> }
30%struct.uint8x16x4_t = type { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> }
31
32declare %struct.uint8x8x2_t @llvm.arm.neon.vld2dup.v8i8.p0i8(i8*, i32)
33declare %struct.uint16x4x2_t @llvm.arm.neon.vld2dup.v4i16.p0i8(i8*, i32)
34declare %struct.uint32x2x2_t @llvm.arm.neon.vld2dup.v2i32.p0i8(i8*, i32)
35declare %struct.uint64x1x2_t @llvm.arm.neon.vld2dup.v1i64.p0i8(i8*, i32)
36
37declare %struct.uint8x8x3_t @llvm.arm.neon.vld3dup.v8i8.p0i8(i8*, i32)
38declare %struct.uint16x4x3_t @llvm.arm.neon.vld3dup.v4i16.p0i8(i8*, i32)
39declare %struct.uint32x2x3_t @llvm.arm.neon.vld3dup.v2i32.p0i8(i8*, i32)
40declare %struct.uint64x1x3_t @llvm.arm.neon.vld3dup.v1i64.p0i8(i8*, i32)
41
42declare %struct.uint8x8x4_t @llvm.arm.neon.vld4dup.v8i8.p0i8(i8*, i32)
43declare %struct.uint16x4x4_t @llvm.arm.neon.vld4dup.v4i16.p0i8(i8*, i32)
44declare %struct.uint32x2x4_t @llvm.arm.neon.vld4dup.v2i32.p0i8(i8*, i32)
45declare %struct.uint64x1x4_t @llvm.arm.neon.vld4dup.v1i64.p0i8(i8*, i32)
46
47declare %struct.uint8x16x2_t @llvm.arm.neon.vld2dup.v16i8.p0i8(i8*, i32)
48declare %struct.uint16x8x2_t @llvm.arm.neon.vld2dup.v8i16.p0i8(i8*, i32)
49declare %struct.uint32x4x2_t @llvm.arm.neon.vld2dup.v4i32.p0i8(i8*, i32)
50
51declare %struct.uint8x16x3_t @llvm.arm.neon.vld3dup.v16i8.p0i8(i8*, i32)
52declare %struct.uint16x8x3_t @llvm.arm.neon.vld3dup.v8i16.p0i8(i8*, i32)
53declare %struct.uint32x4x3_t @llvm.arm.neon.vld3dup.v4i32.p0i8(i8*, i32)
54
55declare %struct.uint8x16x4_t @llvm.arm.neon.vld4dup.v16i8.p0i8(i8*, i32)
56declare %struct.uint16x8x4_t @llvm.arm.neon.vld4dup.v8i16.p0i8(i8*, i32)
57declare %struct.uint32x4x4_t @llvm.arm.neon.vld4dup.v4i32.p0i8(i8*, i32)
58
59; CHECK-LABEL: test_vld2_dup_u16
60; CHECK: vld2.16 {d16[], d17[]}, [r0]
61define %struct.uint16x4x2_t @test_vld2_dup_u16(i8* %src) {
62entry:
63  %tmp = tail call %struct.uint16x4x2_t @llvm.arm.neon.vld2dup.v4i16.p0i8(i8* %src, i32 2)
64  ret %struct.uint16x4x2_t %tmp
65}
66
67; CHECK-LABEL: test_vld2_dup_u32
68; CHECK: vld2.32 {d16[], d17[]}, [r0]
69define %struct.uint32x2x2_t @test_vld2_dup_u32(i8* %src) {
70entry:
71  %tmp = tail call %struct.uint32x2x2_t @llvm.arm.neon.vld2dup.v2i32.p0i8(i8* %src, i32 4)
72  ret %struct.uint32x2x2_t %tmp
73}
74
75; CHECK-LABEL: test_vld2_dup_u64
76; CHECK: vld1.64 {d16, d17}, [r0:64]
77define %struct.uint64x1x2_t @test_vld2_dup_u64(i8* %src) {
78entry:
79  %tmp = tail call %struct.uint64x1x2_t @llvm.arm.neon.vld2dup.v1i64.p0i8(i8* %src, i32 8)
80  ret %struct.uint64x1x2_t %tmp
81}
82
83; CHECK-LABEL: test_vld2_dup_u8
84; CHECK: vld2.8 {d16[], d17[]}, [r0]
85define %struct.uint8x8x2_t @test_vld2_dup_u8(i8* %src) {
86entry:
87  %tmp = tail call %struct.uint8x8x2_t @llvm.arm.neon.vld2dup.v8i8.p0i8(i8* %src, i32 1)
88  ret %struct.uint8x8x2_t %tmp
89}
90
91; CHECK-LABEL: test_vld3_dup_u16
92; CHECK: vld3.16 {d16[], d17[], d18[]}, [r1]
93define %struct.uint16x4x3_t @test_vld3_dup_u16(i8* %src) {
94entry:
95  %tmp = tail call %struct.uint16x4x3_t @llvm.arm.neon.vld3dup.v4i16.p0i8(i8* %src, i32 2)
96  ret %struct.uint16x4x3_t %tmp
97}
98
99; CHECK-LABEL: test_vld3_dup_u32
100; CHECK: vld3.32 {d16[], d17[], d18[]}, [r1]
101define %struct.uint32x2x3_t @test_vld3_dup_u32(i8* %src) {
102entry:
103  %tmp = tail call %struct.uint32x2x3_t @llvm.arm.neon.vld3dup.v2i32.p0i8(i8* %src, i32 4)
104  ret %struct.uint32x2x3_t %tmp
105}
106
107; CHECK-LABEL: test_vld3_dup_u64
108; CHECK: vld1.64 {d16, d17, d18}, [r1]
109define %struct.uint64x1x3_t @test_vld3_dup_u64(i8* %src) {
110entry:
111  %tmp = tail call %struct.uint64x1x3_t @llvm.arm.neon.vld3dup.v1i64.p0i8(i8* %src, i32 8)
112  ret %struct.uint64x1x3_t %tmp
113}
114
115; CHECK-LABEL: test_vld3_dup_u8
116; CHECK: vld3.8 {d16[], d17[], d18[]}, [r1]
117define %struct.uint8x8x3_t @test_vld3_dup_u8(i8* %src) {
118entry:
119  %tmp = tail call %struct.uint8x8x3_t @llvm.arm.neon.vld3dup.v8i8.p0i8(i8* %src, i32 1)
120  ret %struct.uint8x8x3_t %tmp
121}
122
123; CHECK-LABEL: test_vld4_dup_u16
124; CHECK: vld4.16 {d16[], d17[], d18[], d19[]}, [r1]
125define %struct.uint16x4x4_t @test_vld4_dup_u16(i8* %src) {
126entry:
127  %tmp = tail call %struct.uint16x4x4_t @llvm.arm.neon.vld4dup.v4i16.p0i8(i8* %src, i32 2)
128  ret %struct.uint16x4x4_t %tmp
129}
130
131; CHECK-LABEL: test_vld4_dup_u32
132; CHECK: vld4.32 {d16[], d17[], d18[], d19[]}, [r1]
133define %struct.uint32x2x4_t @test_vld4_dup_u32(i8* %src) {
134entry:
135  %tmp = tail call %struct.uint32x2x4_t @llvm.arm.neon.vld4dup.v2i32.p0i8(i8* %src, i32 4)
136  ret %struct.uint32x2x4_t %tmp
137}
138
139; CHECK-LABEL: test_vld4_dup_u64
140; CHECK: vld1.64 {d16, d17, d18, d19}, [r1:64]
141define %struct.uint64x1x4_t @test_vld4_dup_u64(i8* %src) {
142entry:
143  %tmp = tail call %struct.uint64x1x4_t @llvm.arm.neon.vld4dup.v1i64.p0i8(i8* %src, i32 8)
144  ret %struct.uint64x1x4_t %tmp
145}
146
147; CHECK-LABEL: test_vld4_dup_u8
148; CHECK: vld4.8 {d16[], d17[], d18[], d19[]}, [r1]
149define %struct.uint8x8x4_t @test_vld4_dup_u8(i8* %src) {
150entry:
151  %tmp = tail call %struct.uint8x8x4_t @llvm.arm.neon.vld4dup.v8i8.p0i8(i8* %src, i32 1)
152  ret %struct.uint8x8x4_t %tmp
153}
154
155; CHECK-LABEL: test_vld2q_dup_u16
156; CHECK: vld2.16 {d16[], d18[]}, [r1]
157; CHECK: vld2.16 {d17[], d19[]}, [r1]
158define %struct.uint16x8x2_t @test_vld2q_dup_u16(i8* %src) {
159entry:
160  %tmp = tail call %struct.uint16x8x2_t @llvm.arm.neon.vld2dup.v8i16.p0i8(i8* %src, i32 2)
161  ret %struct.uint16x8x2_t %tmp
162}
163
164; CHECK-LABEL: test_vld2q_dup_u32
165; CHECK: vld2.32 {d16[], d18[]}, [r1]
166; CHECK: vld2.32 {d17[], d19[]}, [r1]
167define %struct.uint32x4x2_t @test_vld2q_dup_u32(i8* %src) {
168entry:
169  %tmp = tail call %struct.uint32x4x2_t @llvm.arm.neon.vld2dup.v4i32.p0i8(i8* %src, i32 4)
170  ret %struct.uint32x4x2_t %tmp
171}
172
173; CHECK-LABEL: test_vld2q_dup_u8
174; CHECK: vld2.8 {d16[], d18[]}, [r1]
175; CHECK: vld2.8 {d17[], d19[]}, [r1]
176define %struct.uint8x16x2_t @test_vld2q_dup_u8(i8* %src) {
177entry:
178  %tmp = tail call %struct.uint8x16x2_t @llvm.arm.neon.vld2dup.v16i8.p0i8(i8* %src, i32 1)
179  ret %struct.uint8x16x2_t %tmp
180}
181
182; CHECK-LABEL: test_vld3q_dup_u16
183; CHECK: vld3.16 {d16[], d18[], d20[]}, [r1]
184; CHECK: vld3.16 {d17[], d19[], d21[]}, [r1]
185define %struct.uint16x8x3_t @test_vld3q_dup_u16(i8* %src) {
186entry:
187  %tmp = tail call %struct.uint16x8x3_t @llvm.arm.neon.vld3dup.v8i16.p0i8(i8* %src, i32 2)
188  ret %struct.uint16x8x3_t %tmp
189}
190
191; CHECK-LABEL: test_vld3q_dup_u32
192; CHECK: vld3.32 {d16[], d18[], d20[]}, [r1]
193; CHECK: vld3.32 {d17[], d19[], d21[]}, [r1]
194define %struct.uint32x4x3_t @test_vld3q_dup_u32(i8* %src) {
195entry:
196  %tmp = tail call %struct.uint32x4x3_t @llvm.arm.neon.vld3dup.v4i32.p0i8(i8* %src, i32 4)
197  ret %struct.uint32x4x3_t %tmp
198}
199
200; CHECK-LABEL: test_vld3q_dup_u8
201; CHECK: vld3.8 {d16[], d18[], d20[]}, [r1]
202; CHECK: vld3.8 {d17[], d19[], d21[]}, [r1]
203define %struct.uint8x16x3_t @test_vld3q_dup_u8(i8* %src) {
204entry:
205  %tmp = tail call %struct.uint8x16x3_t @llvm.arm.neon.vld3dup.v16i8.p0i8(i8* %src, i32 1)
206  ret %struct.uint8x16x3_t %tmp
207}
208
209; CHECK-LABEL: test_vld4q_dup_u16
210; CHECK: vld4.16 {d16[], d18[], d20[], d22[]}, [r1]
211; CHECK: vld4.16 {d17[], d19[], d21[], d23[]}, [r1]
212define %struct.uint16x8x4_t @test_vld4q_dup_u16(i8* %src) {
213entry:
214  %tmp = tail call %struct.uint16x8x4_t @llvm.arm.neon.vld4dup.v8i16.p0i8(i8* %src, i32 2)
215  ret %struct.uint16x8x4_t %tmp
216}
217
218; CHECK-LABEL: test_vld4q_dup_u32
219; CHECK: vld4.32 {d16[], d18[], d20[], d22[]}, [r1]
220; CHECK: vld4.32 {d17[], d19[], d21[], d23[]}, [r1]
221define %struct.uint32x4x4_t @test_vld4q_dup_u32(i8* %src) {
222entry:
223  %tmp = tail call %struct.uint32x4x4_t @llvm.arm.neon.vld4dup.v4i32.p0i8(i8* %src, i32 4)
224  ret %struct.uint32x4x4_t %tmp
225}
226
227; CHECK-LABEL: test_vld4q_dup_u8
228; CHECK: vld4.8 {d16[], d18[], d20[], d22[]}, [r1]
229; CHECK: vld4.8 {d17[], d19[], d21[], d23[]}, [r1]
230define %struct.uint8x16x4_t @test_vld4q_dup_u8(i8* %src) {
231entry:
232  %tmp = tail call %struct.uint8x16x4_t @llvm.arm.neon.vld4dup.v16i8.p0i8(i8* %src, i32 1)
233  ret %struct.uint8x16x4_t %tmp
234}
235