1 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
2 ; RUN: llc -mtriple=arm-eabi -mattr=+neon %s -o - | FileCheck %s
4 define <8 x i8> @vzipi8(<8 x i8>* %A, <8 x i8>* %B) nounwind {
7 ; CHECK-NEXT: vldr d16, [r1]
8 ; CHECK-NEXT: vldr d17, [r0]
9 ; CHECK-NEXT: vzip.8 d17, d16
10 ; CHECK-NEXT: vadd.i8 d16, d17, d16
11 ; CHECK-NEXT: vmov r0, r1, d16
12 ; CHECK-NEXT: mov pc, lr
13 %tmp1 = load <8 x i8>, <8 x i8>* %A
14 %tmp2 = load <8 x i8>, <8 x i8>* %B
15 %tmp3 = shufflevector <8 x i8> %tmp1, <8 x i8> %tmp2, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 2, i32 10, i32 3, i32 11>
16 %tmp4 = shufflevector <8 x i8> %tmp1, <8 x i8> %tmp2, <8 x i32> <i32 4, i32 12, i32 5, i32 13, i32 6, i32 14, i32 7, i32 15>
17 %tmp5 = add <8 x i8> %tmp3, %tmp4
21 define <16 x i8> @vzipi8_Qres(<8 x i8>* %A, <8 x i8>* %B) nounwind {
22 ; CHECK-LABEL: vzipi8_Qres:
24 ; CHECK-NEXT: vldr d17, [r1]
25 ; CHECK-NEXT: vldr d16, [r0]
26 ; CHECK-NEXT: vzip.8 d16, d17
27 ; CHECK-NEXT: vmov r0, r1, d16
28 ; CHECK-NEXT: vmov r2, r3, d17
29 ; CHECK-NEXT: mov pc, lr
30 %tmp1 = load <8 x i8>, <8 x i8>* %A
31 %tmp2 = load <8 x i8>, <8 x i8>* %B
32 %tmp3 = shufflevector <8 x i8> %tmp1, <8 x i8> %tmp2, <16 x i32> <i32 0, i32 8, i32 1, i32 9, i32 2, i32 10, i32 3, i32 11, i32 4, i32 12, i32 5, i32 13, i32 6, i32 14, i32 7, i32 15>
36 define <4 x i16> @vzipi16(<4 x i16>* %A, <4 x i16>* %B) nounwind {
37 ; CHECK-LABEL: vzipi16:
39 ; CHECK-NEXT: vldr d16, [r1]
40 ; CHECK-NEXT: vldr d17, [r0]
41 ; CHECK-NEXT: vzip.16 d17, d16
42 ; CHECK-NEXT: vadd.i16 d16, d17, d16
43 ; CHECK-NEXT: vmov r0, r1, d16
44 ; CHECK-NEXT: mov pc, lr
45 %tmp1 = load <4 x i16>, <4 x i16>* %A
46 %tmp2 = load <4 x i16>, <4 x i16>* %B
47 %tmp3 = shufflevector <4 x i16> %tmp1, <4 x i16> %tmp2, <4 x i32> <i32 0, i32 4, i32 1, i32 5>
48 %tmp4 = shufflevector <4 x i16> %tmp1, <4 x i16> %tmp2, <4 x i32> <i32 2, i32 6, i32 3, i32 7>
49 %tmp5 = add <4 x i16> %tmp3, %tmp4
53 define <8 x i16> @vzipi16_Qres(<4 x i16>* %A, <4 x i16>* %B) nounwind {
54 ; CHECK-LABEL: vzipi16_Qres:
56 ; CHECK-NEXT: vldr d17, [r1]
57 ; CHECK-NEXT: vldr d16, [r0]
58 ; CHECK-NEXT: vzip.16 d16, d17
59 ; CHECK-NEXT: vmov r0, r1, d16
60 ; CHECK-NEXT: vmov r2, r3, d17
61 ; CHECK-NEXT: mov pc, lr
62 %tmp1 = load <4 x i16>, <4 x i16>* %A
63 %tmp2 = load <4 x i16>, <4 x i16>* %B
64 %tmp3 = shufflevector <4 x i16> %tmp1, <4 x i16> %tmp2, <8 x i32> <i32 0, i32 4, i32 1, i32 5, i32 2, i32 6, i32 3, i32 7>
68 ; VZIP.32 is equivalent to VTRN.32 for 64-bit vectors.
70 define <16 x i8> @vzipQi8(<16 x i8>* %A, <16 x i8>* %B) nounwind {
71 ; CHECK-LABEL: vzipQi8:
73 ; CHECK-NEXT: vld1.64 {d16, d17}, [r1]
74 ; CHECK-NEXT: vld1.64 {d18, d19}, [r0]
75 ; CHECK-NEXT: vzip.8 q9, q8
76 ; CHECK-NEXT: vadd.i8 q8, q9, q8
77 ; CHECK-NEXT: vmov r0, r1, d16
78 ; CHECK-NEXT: vmov r2, r3, d17
79 ; CHECK-NEXT: mov pc, lr
80 %tmp1 = load <16 x i8>, <16 x i8>* %A
81 %tmp2 = load <16 x i8>, <16 x i8>* %B
82 %tmp3 = shufflevector <16 x i8> %tmp1, <16 x i8> %tmp2, <16 x i32> <i32 0, i32 16, i32 1, i32 17, i32 2, i32 18, i32 3, i32 19, i32 4, i32 20, i32 5, i32 21, i32 6, i32 22, i32 7, i32 23>
83 %tmp4 = shufflevector <16 x i8> %tmp1, <16 x i8> %tmp2, <16 x i32> <i32 8, i32 24, i32 9, i32 25, i32 10, i32 26, i32 11, i32 27, i32 12, i32 28, i32 13, i32 29, i32 14, i32 30, i32 15, i32 31>
84 %tmp5 = add <16 x i8> %tmp3, %tmp4
88 define <32 x i8> @vzipQi8_QQres(<16 x i8>* %A, <16 x i8>* %B) nounwind {
89 ; CHECK-LABEL: vzipQi8_QQres:
91 ; CHECK-NEXT: vld1.64 {d16, d17}, [r2]
92 ; CHECK-NEXT: vld1.64 {d18, d19}, [r1]
93 ; CHECK-NEXT: vzip.8 q9, q8
94 ; CHECK-NEXT: vst1.8 {d18, d19}, [r0:128]!
95 ; CHECK-NEXT: vst1.64 {d16, d17}, [r0:128]
96 ; CHECK-NEXT: mov pc, lr
97 %tmp1 = load <16 x i8>, <16 x i8>* %A
98 %tmp2 = load <16 x i8>, <16 x i8>* %B
99 %tmp3 = shufflevector <16 x i8> %tmp1, <16 x i8> %tmp2, <32 x i32> <i32 0, i32 16, i32 1, i32 17, i32 2, i32 18, i32 3, i32 19, i32 4, i32 20, i32 5, i32 21, i32 6, i32 22, i32 7, i32 23, i32 8, i32 24, i32 9, i32 25, i32 10, i32 26, i32 11, i32 27, i32 12, i32 28, i32 13, i32 29, i32 14, i32 30, i32 15, i32 31>
103 define <8 x i16> @vzipQi16(<8 x i16>* %A, <8 x i16>* %B) nounwind {
104 ; CHECK-LABEL: vzipQi16:
106 ; CHECK-NEXT: vld1.64 {d16, d17}, [r1]
107 ; CHECK-NEXT: vld1.64 {d18, d19}, [r0]
108 ; CHECK-NEXT: vzip.16 q9, q8
109 ; CHECK-NEXT: vadd.i16 q8, q9, q8
110 ; CHECK-NEXT: vmov r0, r1, d16
111 ; CHECK-NEXT: vmov r2, r3, d17
112 ; CHECK-NEXT: mov pc, lr
113 %tmp1 = load <8 x i16>, <8 x i16>* %A
114 %tmp2 = load <8 x i16>, <8 x i16>* %B
115 %tmp3 = shufflevector <8 x i16> %tmp1, <8 x i16> %tmp2, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 2, i32 10, i32 3, i32 11>
116 %tmp4 = shufflevector <8 x i16> %tmp1, <8 x i16> %tmp2, <8 x i32> <i32 4, i32 12, i32 5, i32 13, i32 6, i32 14, i32 7, i32 15>
117 %tmp5 = add <8 x i16> %tmp3, %tmp4
121 define <16 x i16> @vzipQi16_QQres(<8 x i16>* %A, <8 x i16>* %B) nounwind {
122 ; CHECK-LABEL: vzipQi16_QQres:
124 ; CHECK-NEXT: vld1.64 {d16, d17}, [r2]
125 ; CHECK-NEXT: vld1.64 {d18, d19}, [r1]
126 ; CHECK-NEXT: vzip.16 q9, q8
127 ; CHECK-NEXT: vst1.16 {d18, d19}, [r0:128]!
128 ; CHECK-NEXT: vst1.64 {d16, d17}, [r0:128]
129 ; CHECK-NEXT: mov pc, lr
130 %tmp1 = load <8 x i16>, <8 x i16>* %A
131 %tmp2 = load <8 x i16>, <8 x i16>* %B
132 %tmp3 = shufflevector <8 x i16> %tmp1, <8 x i16> %tmp2, <16 x i32> <i32 0, i32 8, i32 1, i32 9, i32 2, i32 10, i32 3, i32 11, i32 4, i32 12, i32 5, i32 13, i32 6, i32 14, i32 7, i32 15>
136 define <4 x i32> @vzipQi32(<4 x i32>* %A, <4 x i32>* %B) nounwind {
137 ; CHECK-LABEL: vzipQi32:
139 ; CHECK-NEXT: vld1.64 {d16, d17}, [r1]
140 ; CHECK-NEXT: vld1.64 {d18, d19}, [r0]
141 ; CHECK-NEXT: vzip.32 q9, q8
142 ; CHECK-NEXT: vadd.i32 q8, q9, q8
143 ; CHECK-NEXT: vmov r0, r1, d16
144 ; CHECK-NEXT: vmov r2, r3, d17
145 ; CHECK-NEXT: mov pc, lr
146 %tmp1 = load <4 x i32>, <4 x i32>* %A
147 %tmp2 = load <4 x i32>, <4 x i32>* %B
148 %tmp3 = shufflevector <4 x i32> %tmp1, <4 x i32> %tmp2, <4 x i32> <i32 0, i32 4, i32 1, i32 5>
149 %tmp4 = shufflevector <4 x i32> %tmp1, <4 x i32> %tmp2, <4 x i32> <i32 2, i32 6, i32 3, i32 7>
150 %tmp5 = add <4 x i32> %tmp3, %tmp4
154 define <8 x i32> @vzipQi32_QQres(<4 x i32>* %A, <4 x i32>* %B) nounwind {
155 ; CHECK-LABEL: vzipQi32_QQres:
157 ; CHECK-NEXT: vld1.64 {d16, d17}, [r2]
158 ; CHECK-NEXT: vld1.64 {d18, d19}, [r1]
159 ; CHECK-NEXT: vzip.32 q9, q8
160 ; CHECK-NEXT: vst1.32 {d18, d19}, [r0:128]!
161 ; CHECK-NEXT: vst1.64 {d16, d17}, [r0:128]
162 ; CHECK-NEXT: mov pc, lr
163 %tmp1 = load <4 x i32>, <4 x i32>* %A
164 %tmp2 = load <4 x i32>, <4 x i32>* %B
165 %tmp3 = shufflevector <4 x i32> %tmp1, <4 x i32> %tmp2, <8 x i32> <i32 0, i32 4, i32 1, i32 5, i32 2, i32 6, i32 3, i32 7>
169 define <4 x float> @vzipQf(<4 x float>* %A, <4 x float>* %B) nounwind {
170 ; CHECK-LABEL: vzipQf:
172 ; CHECK-NEXT: vld1.64 {d16, d17}, [r1]
173 ; CHECK-NEXT: vld1.64 {d18, d19}, [r0]
174 ; CHECK-NEXT: vzip.32 q9, q8
175 ; CHECK-NEXT: vadd.f32 q8, q9, q8
176 ; CHECK-NEXT: vmov r0, r1, d16
177 ; CHECK-NEXT: vmov r2, r3, d17
178 ; CHECK-NEXT: mov pc, lr
179 %tmp1 = load <4 x float>, <4 x float>* %A
180 %tmp2 = load <4 x float>, <4 x float>* %B
181 %tmp3 = shufflevector <4 x float> %tmp1, <4 x float> %tmp2, <4 x i32> <i32 0, i32 4, i32 1, i32 5>
182 %tmp4 = shufflevector <4 x float> %tmp1, <4 x float> %tmp2, <4 x i32> <i32 2, i32 6, i32 3, i32 7>
183 %tmp5 = fadd <4 x float> %tmp3, %tmp4
184 ret <4 x float> %tmp5
187 define <8 x float> @vzipQf_QQres(<4 x float>* %A, <4 x float>* %B) nounwind {
188 ; CHECK-LABEL: vzipQf_QQres:
190 ; CHECK-NEXT: vld1.64 {d16, d17}, [r2]
191 ; CHECK-NEXT: vld1.64 {d18, d19}, [r1]
192 ; CHECK-NEXT: vzip.32 q9, q8
193 ; CHECK-NEXT: vst1.32 {d18, d19}, [r0:128]!
194 ; CHECK-NEXT: vst1.64 {d16, d17}, [r0:128]
195 ; CHECK-NEXT: mov pc, lr
196 %tmp1 = load <4 x float>, <4 x float>* %A
197 %tmp2 = load <4 x float>, <4 x float>* %B
198 %tmp3 = shufflevector <4 x float> %tmp1, <4 x float> %tmp2, <8 x i32> <i32 0, i32 4, i32 1, i32 5, i32 2, i32 6, i32 3, i32 7>
199 ret <8 x float> %tmp3
202 ; Undef shuffle indices should not prevent matching to VZIP:
204 define <8 x i8> @vzipi8_undef(<8 x i8>* %A, <8 x i8>* %B) nounwind {
205 ; CHECK-LABEL: vzipi8_undef:
207 ; CHECK-NEXT: vldr d16, [r1]
208 ; CHECK-NEXT: vldr d17, [r0]
209 ; CHECK-NEXT: vzip.8 d17, d16
210 ; CHECK-NEXT: vadd.i8 d16, d17, d16
211 ; CHECK-NEXT: vmov r0, r1, d16
212 ; CHECK-NEXT: mov pc, lr
213 %tmp1 = load <8 x i8>, <8 x i8>* %A
214 %tmp2 = load <8 x i8>, <8 x i8>* %B
215 %tmp3 = shufflevector <8 x i8> %tmp1, <8 x i8> %tmp2, <8 x i32> <i32 0, i32 undef, i32 1, i32 9, i32 undef, i32 10, i32 3, i32 11>
216 %tmp4 = shufflevector <8 x i8> %tmp1, <8 x i8> %tmp2, <8 x i32> <i32 4, i32 12, i32 5, i32 13, i32 6, i32 undef, i32 undef, i32 15>
217 %tmp5 = add <8 x i8> %tmp3, %tmp4
221 define <16 x i8> @vzipi8_undef_Qres(<8 x i8>* %A, <8 x i8>* %B) nounwind {
222 ; CHECK-LABEL: vzipi8_undef_Qres:
224 ; CHECK-NEXT: vldr d17, [r1]
225 ; CHECK-NEXT: vldr d16, [r0]
226 ; CHECK-NEXT: vzip.8 d16, d17
227 ; CHECK-NEXT: vmov r0, r1, d16
228 ; CHECK-NEXT: vmov r2, r3, d17
229 ; CHECK-NEXT: mov pc, lr
230 %tmp1 = load <8 x i8>, <8 x i8>* %A
231 %tmp2 = load <8 x i8>, <8 x i8>* %B
232 %tmp3 = shufflevector <8 x i8> %tmp1, <8 x i8> %tmp2, <16 x i32> <i32 0, i32 undef, i32 1, i32 9, i32 undef, i32 10, i32 3, i32 11, i32 4, i32 12, i32 5, i32 13, i32 6, i32 undef, i32 undef, i32 15>
236 define <16 x i8> @vzipQi8_undef(<16 x i8>* %A, <16 x i8>* %B) nounwind {
237 ; CHECK-LABEL: vzipQi8_undef:
239 ; CHECK-NEXT: vld1.64 {d16, d17}, [r1]
240 ; CHECK-NEXT: vld1.64 {d18, d19}, [r0]
241 ; CHECK-NEXT: vzip.8 q9, q8
242 ; CHECK-NEXT: vadd.i8 q8, q9, q8
243 ; CHECK-NEXT: vmov r0, r1, d16
244 ; CHECK-NEXT: vmov r2, r3, d17
245 ; CHECK-NEXT: mov pc, lr
246 %tmp1 = load <16 x i8>, <16 x i8>* %A
247 %tmp2 = load <16 x i8>, <16 x i8>* %B
248 %tmp3 = shufflevector <16 x i8> %tmp1, <16 x i8> %tmp2, <16 x i32> <i32 0, i32 16, i32 1, i32 undef, i32 undef, i32 undef, i32 3, i32 19, i32 4, i32 20, i32 5, i32 21, i32 6, i32 22, i32 7, i32 23>
249 %tmp4 = shufflevector <16 x i8> %tmp1, <16 x i8> %tmp2, <16 x i32> <i32 8, i32 24, i32 9, i32 undef, i32 10, i32 26, i32 11, i32 27, i32 12, i32 28, i32 13, i32 undef, i32 14, i32 30, i32 undef, i32 31>
250 %tmp5 = add <16 x i8> %tmp3, %tmp4
254 define <32 x i8> @vzipQi8_undef_QQres(<16 x i8>* %A, <16 x i8>* %B) nounwind {
255 ; CHECK-LABEL: vzipQi8_undef_QQres:
257 ; CHECK-NEXT: vld1.64 {d16, d17}, [r2]
258 ; CHECK-NEXT: vld1.64 {d18, d19}, [r1]
259 ; CHECK-NEXT: vzip.8 q9, q8
260 ; CHECK-NEXT: vst1.8 {d18, d19}, [r0:128]!
261 ; CHECK-NEXT: vst1.64 {d16, d17}, [r0:128]
262 ; CHECK-NEXT: mov pc, lr
263 %tmp1 = load <16 x i8>, <16 x i8>* %A
264 %tmp2 = load <16 x i8>, <16 x i8>* %B
265 %tmp3 = shufflevector <16 x i8> %tmp1, <16 x i8> %tmp2, <32 x i32> <i32 0, i32 16, i32 1, i32 undef, i32 undef, i32 undef, i32 3, i32 19, i32 4, i32 20, i32 5, i32 21, i32 6, i32 22, i32 7, i32 23, i32 8, i32 24, i32 9, i32 undef, i32 10, i32 26, i32 11, i32 27, i32 12, i32 28, i32 13, i32 undef, i32 14, i32 30, i32 undef, i32 31>
269 define <8 x i16> @vzip_lower_shufflemask_undef(<4 x i16>* %A, <4 x i16>* %B) {
270 ; CHECK-LABEL: vzip_lower_shufflemask_undef:
271 ; CHECK: @ %bb.0: @ %entry
272 ; CHECK-NEXT: vldr d17, [r1]
273 ; CHECK-NEXT: vldr d18, [r0]
274 ; CHECK-NEXT: vzip.16 d18, d17
275 ; CHECK-NEXT: vmov r0, r1, d16
276 ; CHECK-NEXT: vmov r2, r3, d17
277 ; CHECK-NEXT: mov pc, lr
279 %tmp1 = load <4 x i16>, <4 x i16>* %A
280 %tmp2 = load <4 x i16>, <4 x i16>* %B
281 %0 = shufflevector <4 x i16> %tmp1, <4 x i16> %tmp2, <8 x i32> <i32 undef, i32 undef, i32 undef, i32 undef, i32 2, i32 6, i32 3, i32 7>
285 ; NOTE: The mask here looks like something that could be done with a vzip,
286 ; but which the current handling of two-result vzip can't do - thus ending up
288 define <8 x i16> @vzip_lower_shufflemask_undef_rev(<4 x i16>* %A, <4 x i16>* %B) {
289 ; CHECK-LABEL: vzip_lower_shufflemask_undef_rev:
290 ; CHECK: @ %bb.0: @ %entry
291 ; CHECK-NEXT: vldr d16, [r1]
292 ; CHECK-NEXT: vldr d19, [r0]
293 ; CHECK-NEXT: vtrn.16 d19, d16
294 ; CHECK-NEXT: vmov r0, r1, d18
295 ; CHECK-NEXT: vmov r2, r3, d19
296 ; CHECK-NEXT: mov pc, lr
298 %tmp1 = load <4 x i16>, <4 x i16>* %A
299 %tmp2 = load <4 x i16>, <4 x i16>* %B
300 %0 = shufflevector <4 x i16> %tmp1, <4 x i16> %tmp2, <8 x i32> <i32 undef, i32 undef, i32 undef, i32 undef, i32 0, i32 4, i32 undef, i32 undef>
304 define <4 x i32> @vzip_lower_shufflemask_zeroed(<2 x i32>* %A) {
305 ; CHECK-LABEL: vzip_lower_shufflemask_zeroed:
306 ; CHECK: @ %bb.0: @ %entry
307 ; CHECK-NEXT: vldr d16, [r0]
308 ; CHECK-NEXT: vdup.32 q9, d16[0]
309 ; CHECK-NEXT: vzip.32 q8, q9
310 ; CHECK-NEXT: vmov r0, r1, d16
311 ; CHECK-NEXT: vmov r2, r3, d17
312 ; CHECK-NEXT: mov pc, lr
314 %tmp1 = load <2 x i32>, <2 x i32>* %A
315 %0 = shufflevector <2 x i32> %tmp1, <2 x i32> %tmp1, <4 x i32> <i32 0, i32 0, i32 1, i32 0>
319 define <4 x i32> @vzip_lower_shufflemask_vuzp(<2 x i32>* %A) {
320 ; CHECK-LABEL: vzip_lower_shufflemask_vuzp:
321 ; CHECK: @ %bb.0: @ %entry
322 ; CHECK-NEXT: vldr d16, [r0]
323 ; CHECK-NEXT: vdup.32 q9, d16[0]
324 ; CHECK-NEXT: vzip.32 q8, q9
325 ; CHECK-NEXT: vmov r0, r1, d16
326 ; CHECK-NEXT: vmov r2, r3, d17
327 ; CHECK-NEXT: mov pc, lr
329 %tmp1 = load <2 x i32>, <2 x i32>* %A
330 %0 = shufflevector <2 x i32> %tmp1, <2 x i32> %tmp1, <4 x i32> <i32 0, i32 2, i32 1, i32 0>
334 define void @vzip_undef_rev_shufflemask_vtrn(<2 x i32>* %A, <4 x i32>* %B) {
335 ; CHECK-LABEL: vzip_undef_rev_shufflemask_vtrn:
336 ; CHECK: @ %bb.0: @ %entry
337 ; CHECK-NEXT: vldr d16, [r0]
338 ; CHECK-NEXT: vorr q9, q8, q8
339 ; CHECK-NEXT: vzip.32 q8, q9
340 ; CHECK-NEXT: vext.32 q8, q8, q8, #2
341 ; CHECK-NEXT: vst1.64 {d16, d17}, [r1]
342 ; CHECK-NEXT: mov pc, lr
344 %tmp1 = load <2 x i32>, <2 x i32>* %A
345 %0 = shufflevector <2 x i32> %tmp1, <2 x i32> undef, <4 x i32> <i32 1, i32 1, i32 0, i32 0>
346 store <4 x i32> %0, <4 x i32>* %B
350 define void @vzip_vext_factor(<8 x i16>* %A, <4 x i16>* %B) {
351 ; CHECK-LABEL: vzip_vext_factor:
352 ; CHECK: @ %bb.0: @ %entry
353 ; CHECK-NEXT: vld1.64 {d16, d17}, [r0]
354 ; CHECK-NEXT: vext.16 d18, d16, d17, #1
355 ; CHECK-NEXT: vext.16 d16, d18, d17, #2
356 ; CHECK-NEXT: vext.16 d16, d16, d16, #1
357 ; CHECK-NEXT: vstr d16, [r1]
358 ; CHECK-NEXT: mov pc, lr
360 %tmp1 = load <8 x i16>, <8 x i16>* %A
361 %0 = shufflevector <8 x i16> %tmp1, <8 x i16> undef, <4 x i32> <i32 4, i32 4, i32 5, i32 3>
362 store <4 x i16> %0, <4 x i16>* %B
366 define <8 x i8> @vdup_zip(i8* nocapture readonly %x, i8* nocapture readonly %y) {
367 ; CHECK-LABEL: vdup_zip:
368 ; CHECK: @ %bb.0: @ %entry
369 ; CHECK-NEXT: vld1.8 {d16[]}, [r1]
370 ; CHECK-NEXT: vld1.8 {d17[]}, [r0]
371 ; CHECK-NEXT: vzip.8 d17, d16
372 ; CHECK-NEXT: vmov r0, r1, d17
373 ; CHECK-NEXT: mov pc, lr
375 %0 = load i8, i8* %x, align 1
376 %1 = insertelement <8 x i8> undef, i8 %0, i32 0
377 %lane = shufflevector <8 x i8> %1, <8 x i8> undef, <8 x i32> <i32 0, i32 0, i32 0, i32 0, i32 undef, i32 undef, i32 undef, i32 undef>
378 %2 = load i8, i8* %y, align 1
379 %3 = insertelement <8 x i8> undef, i8 %2, i32 0
380 %lane3 = shufflevector <8 x i8> %3, <8 x i8> undef, <8 x i32> <i32 0, i32 0, i32 0, i32 0, i32 undef, i32 undef, i32 undef, i32 undef>
381 %vzip.i = shufflevector <8 x i8> %lane, <8 x i8> %lane3, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 2, i32 10, i32 3, i32 11>