1 /*===---- arm_neon.h - ARM Neon intrinsics ---------------------------------===
2 *
3 * Permission is hereby granted, free of charge, to any person obtaining a copy
4 * of this software and associated documentation files (the "Software"), to deal
5 * in the Software without restriction, including without limitation the rights
6 * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
7 * copies of the Software, and to permit persons to whom the Software is
8 * furnished to do so, subject to the following conditions:
9 *
10 * The above copyright notice and this permission notice shall be included in
11 * all copies or substantial portions of the Software.
12 *
13 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
14 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
15 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
16 * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
17 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
18 * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
19 * THE SOFTWARE.
20 *
21 *===-----------------------------------------------------------------------===
22 */
23
24 #ifndef __ARM_NEON_H
25 #define __ARM_NEON_H
26
27 #ifndef __ARM_FP
28 #error "NEON intrinsics not available with the soft-float ABI. Please use -mfloat-abi=softfp or -mfloat-abi=hard"
29 #else
30
31 #if !defined(__ARM_NEON)
32 #error "NEON support not enabled"
33 #else
34
35 #include <stdint.h>
36
37 #include <arm_bf16.h>
38 #include <arm_vector_types.h>
39 #ifdef __aarch64__
40 typedef uint8_t poly8_t;
41 typedef uint16_t poly16_t;
42 typedef uint64_t poly64_t;
43 typedef __uint128_t poly128_t;
44 #else
45 typedef int8_t poly8_t;
46 typedef int16_t poly16_t;
47 typedef int64_t poly64_t;
48 #endif
49 typedef __attribute__((neon_polyvector_type(8))) poly8_t poly8x8_t;
50 typedef __attribute__((neon_polyvector_type(16))) poly8_t poly8x16_t;
51 typedef __attribute__((neon_polyvector_type(4))) poly16_t poly16x4_t;
52 typedef __attribute__((neon_polyvector_type(8))) poly16_t poly16x8_t;
53 typedef __attribute__((neon_polyvector_type(1))) poly64_t poly64x1_t;
54 typedef __attribute__((neon_polyvector_type(2))) poly64_t poly64x2_t;
55
56 typedef struct poly8x8x2_t {
57 poly8x8_t val[2];
58 } poly8x8x2_t;
59
60 typedef struct poly8x16x2_t {
61 poly8x16_t val[2];
62 } poly8x16x2_t;
63
64 typedef struct poly16x4x2_t {
65 poly16x4_t val[2];
66 } poly16x4x2_t;
67
68 typedef struct poly16x8x2_t {
69 poly16x8_t val[2];
70 } poly16x8x2_t;
71
72 typedef struct poly64x1x2_t {
73 poly64x1_t val[2];
74 } poly64x1x2_t;
75
76 typedef struct poly64x2x2_t {
77 poly64x2_t val[2];
78 } poly64x2x2_t;
79
80 typedef struct poly8x8x3_t {
81 poly8x8_t val[3];
82 } poly8x8x3_t;
83
84 typedef struct poly8x16x3_t {
85 poly8x16_t val[3];
86 } poly8x16x3_t;
87
88 typedef struct poly16x4x3_t {
89 poly16x4_t val[3];
90 } poly16x4x3_t;
91
92 typedef struct poly16x8x3_t {
93 poly16x8_t val[3];
94 } poly16x8x3_t;
95
96 typedef struct poly64x1x3_t {
97 poly64x1_t val[3];
98 } poly64x1x3_t;
99
100 typedef struct poly64x2x3_t {
101 poly64x2_t val[3];
102 } poly64x2x3_t;
103
104 typedef struct poly8x8x4_t {
105 poly8x8_t val[4];
106 } poly8x8x4_t;
107
108 typedef struct poly8x16x4_t {
109 poly8x16_t val[4];
110 } poly8x16x4_t;
111
112 typedef struct poly16x4x4_t {
113 poly16x4_t val[4];
114 } poly16x4x4_t;
115
116 typedef struct poly16x8x4_t {
117 poly16x8_t val[4];
118 } poly16x8x4_t;
119
120 typedef struct poly64x1x4_t {
121 poly64x1_t val[4];
122 } poly64x1x4_t;
123
124 typedef struct poly64x2x4_t {
125 poly64x2_t val[4];
126 } poly64x2x4_t;
127
128 #define __ai static __inline__ __attribute__((__always_inline__, __nodebug__))
129
130 #ifdef __LITTLE_ENDIAN__
131 #define splat_lane_p8(__p0, __p1) __extension__ ({ \
132 poly8x8_t __ret; \
133 poly8x8_t __s0 = __p0; \
134 __ret = (poly8x8_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 4); \
135 __ret; \
136 })
137 #else
138 #define splat_lane_p8(__p0, __p1) __extension__ ({ \
139 poly8x8_t __ret; \
140 poly8x8_t __s0 = __p0; \
141 poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
142 __ret = (poly8x8_t) __builtin_neon_splat_lane_v((int8x8_t)__rev0, __p1, 4); \
143 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
144 __ret; \
145 })
146 #define __noswap_splat_lane_p8(__p0, __p1) __extension__ ({ \
147 poly8x8_t __ret; \
148 poly8x8_t __s0 = __p0; \
149 __ret = (poly8x8_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 4); \
150 __ret; \
151 })
152 #endif
153
154 #define splat_lane_p64(__p0, __p1) __extension__ ({ \
155 poly64x1_t __ret; \
156 poly64x1_t __s0 = __p0; \
157 __ret = (poly64x1_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 6); \
158 __ret; \
159 })
160 #ifdef __LITTLE_ENDIAN__
161 #define splat_lane_p16(__p0, __p1) __extension__ ({ \
162 poly16x4_t __ret; \
163 poly16x4_t __s0 = __p0; \
164 __ret = (poly16x4_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 5); \
165 __ret; \
166 })
167 #else
168 #define splat_lane_p16(__p0, __p1) __extension__ ({ \
169 poly16x4_t __ret; \
170 poly16x4_t __s0 = __p0; \
171 poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
172 __ret = (poly16x4_t) __builtin_neon_splat_lane_v((int8x8_t)__rev0, __p1, 5); \
173 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
174 __ret; \
175 })
176 #define __noswap_splat_lane_p16(__p0, __p1) __extension__ ({ \
177 poly16x4_t __ret; \
178 poly16x4_t __s0 = __p0; \
179 __ret = (poly16x4_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 5); \
180 __ret; \
181 })
182 #endif
183
184 #ifdef __LITTLE_ENDIAN__
185 #define splatq_lane_p8(__p0, __p1) __extension__ ({ \
186 poly8x16_t __ret; \
187 poly8x8_t __s0 = __p0; \
188 __ret = (poly8x16_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 4); \
189 __ret; \
190 })
191 #else
192 #define splatq_lane_p8(__p0, __p1) __extension__ ({ \
193 poly8x16_t __ret; \
194 poly8x8_t __s0 = __p0; \
195 poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
196 __ret = (poly8x16_t) __builtin_neon_splatq_lane_v((int8x8_t)__rev0, __p1, 4); \
197 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
198 __ret; \
199 })
200 #define __noswap_splatq_lane_p8(__p0, __p1) __extension__ ({ \
201 poly8x16_t __ret; \
202 poly8x8_t __s0 = __p0; \
203 __ret = (poly8x16_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 4); \
204 __ret; \
205 })
206 #endif
207
208 #ifdef __LITTLE_ENDIAN__
209 #define splatq_lane_p64(__p0, __p1) __extension__ ({ \
210 poly64x2_t __ret; \
211 poly64x1_t __s0 = __p0; \
212 __ret = (poly64x2_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 6); \
213 __ret; \
214 })
215 #else
216 #define splatq_lane_p64(__p0, __p1) __extension__ ({ \
217 poly64x2_t __ret; \
218 poly64x1_t __s0 = __p0; \
219 __ret = (poly64x2_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 6); \
220 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
221 __ret; \
222 })
223 #define __noswap_splatq_lane_p64(__p0, __p1) __extension__ ({ \
224 poly64x2_t __ret; \
225 poly64x1_t __s0 = __p0; \
226 __ret = (poly64x2_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 6); \
227 __ret; \
228 })
229 #endif
230
231 #ifdef __LITTLE_ENDIAN__
232 #define splatq_lane_p16(__p0, __p1) __extension__ ({ \
233 poly16x8_t __ret; \
234 poly16x4_t __s0 = __p0; \
235 __ret = (poly16x8_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 5); \
236 __ret; \
237 })
238 #else
239 #define splatq_lane_p16(__p0, __p1) __extension__ ({ \
240 poly16x8_t __ret; \
241 poly16x4_t __s0 = __p0; \
242 poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
243 __ret = (poly16x8_t) __builtin_neon_splatq_lane_v((int8x8_t)__rev0, __p1, 5); \
244 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
245 __ret; \
246 })
247 #define __noswap_splatq_lane_p16(__p0, __p1) __extension__ ({ \
248 poly16x8_t __ret; \
249 poly16x4_t __s0 = __p0; \
250 __ret = (poly16x8_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 5); \
251 __ret; \
252 })
253 #endif
254
255 #ifdef __LITTLE_ENDIAN__
256 #define splatq_lane_u8(__p0, __p1) __extension__ ({ \
257 uint8x16_t __ret; \
258 uint8x8_t __s0 = __p0; \
259 __ret = (uint8x16_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 16); \
260 __ret; \
261 })
262 #else
263 #define splatq_lane_u8(__p0, __p1) __extension__ ({ \
264 uint8x16_t __ret; \
265 uint8x8_t __s0 = __p0; \
266 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
267 __ret = (uint8x16_t) __builtin_neon_splatq_lane_v((int8x8_t)__rev0, __p1, 16); \
268 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
269 __ret; \
270 })
271 #define __noswap_splatq_lane_u8(__p0, __p1) __extension__ ({ \
272 uint8x16_t __ret; \
273 uint8x8_t __s0 = __p0; \
274 __ret = (uint8x16_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 16); \
275 __ret; \
276 })
277 #endif
278
279 #ifdef __LITTLE_ENDIAN__
280 #define splatq_lane_u32(__p0, __p1) __extension__ ({ \
281 uint32x4_t __ret; \
282 uint32x2_t __s0 = __p0; \
283 __ret = (uint32x4_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 18); \
284 __ret; \
285 })
286 #else
287 #define splatq_lane_u32(__p0, __p1) __extension__ ({ \
288 uint32x4_t __ret; \
289 uint32x2_t __s0 = __p0; \
290 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
291 __ret = (uint32x4_t) __builtin_neon_splatq_lane_v((int8x8_t)__rev0, __p1, 18); \
292 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
293 __ret; \
294 })
295 #define __noswap_splatq_lane_u32(__p0, __p1) __extension__ ({ \
296 uint32x4_t __ret; \
297 uint32x2_t __s0 = __p0; \
298 __ret = (uint32x4_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 18); \
299 __ret; \
300 })
301 #endif
302
303 #ifdef __LITTLE_ENDIAN__
304 #define splatq_lane_u64(__p0, __p1) __extension__ ({ \
305 uint64x2_t __ret; \
306 uint64x1_t __s0 = __p0; \
307 __ret = (uint64x2_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 19); \
308 __ret; \
309 })
310 #else
311 #define splatq_lane_u64(__p0, __p1) __extension__ ({ \
312 uint64x2_t __ret; \
313 uint64x1_t __s0 = __p0; \
314 __ret = (uint64x2_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 19); \
315 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
316 __ret; \
317 })
318 #define __noswap_splatq_lane_u64(__p0, __p1) __extension__ ({ \
319 uint64x2_t __ret; \
320 uint64x1_t __s0 = __p0; \
321 __ret = (uint64x2_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 19); \
322 __ret; \
323 })
324 #endif
325
326 #ifdef __LITTLE_ENDIAN__
327 #define splatq_lane_u16(__p0, __p1) __extension__ ({ \
328 uint16x8_t __ret; \
329 uint16x4_t __s0 = __p0; \
330 __ret = (uint16x8_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 17); \
331 __ret; \
332 })
333 #else
334 #define splatq_lane_u16(__p0, __p1) __extension__ ({ \
335 uint16x8_t __ret; \
336 uint16x4_t __s0 = __p0; \
337 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
338 __ret = (uint16x8_t) __builtin_neon_splatq_lane_v((int8x8_t)__rev0, __p1, 17); \
339 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
340 __ret; \
341 })
342 #define __noswap_splatq_lane_u16(__p0, __p1) __extension__ ({ \
343 uint16x8_t __ret; \
344 uint16x4_t __s0 = __p0; \
345 __ret = (uint16x8_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 17); \
346 __ret; \
347 })
348 #endif
349
350 #ifdef __LITTLE_ENDIAN__
351 #define splatq_lane_s8(__p0, __p1) __extension__ ({ \
352 int8x16_t __ret; \
353 int8x8_t __s0 = __p0; \
354 __ret = (int8x16_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 0); \
355 __ret; \
356 })
357 #else
358 #define splatq_lane_s8(__p0, __p1) __extension__ ({ \
359 int8x16_t __ret; \
360 int8x8_t __s0 = __p0; \
361 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
362 __ret = (int8x16_t) __builtin_neon_splatq_lane_v((int8x8_t)__rev0, __p1, 0); \
363 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
364 __ret; \
365 })
366 #define __noswap_splatq_lane_s8(__p0, __p1) __extension__ ({ \
367 int8x16_t __ret; \
368 int8x8_t __s0 = __p0; \
369 __ret = (int8x16_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 0); \
370 __ret; \
371 })
372 #endif
373
374 #ifdef __LITTLE_ENDIAN__
375 #define splatq_lane_f64(__p0, __p1) __extension__ ({ \
376 float64x2_t __ret; \
377 float64x1_t __s0 = __p0; \
378 __ret = (float64x2_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 10); \
379 __ret; \
380 })
381 #else
382 #define splatq_lane_f64(__p0, __p1) __extension__ ({ \
383 float64x2_t __ret; \
384 float64x1_t __s0 = __p0; \
385 __ret = (float64x2_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 10); \
386 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
387 __ret; \
388 })
389 #define __noswap_splatq_lane_f64(__p0, __p1) __extension__ ({ \
390 float64x2_t __ret; \
391 float64x1_t __s0 = __p0; \
392 __ret = (float64x2_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 10); \
393 __ret; \
394 })
395 #endif
396
397 #ifdef __LITTLE_ENDIAN__
398 #define splatq_lane_f32(__p0, __p1) __extension__ ({ \
399 float32x4_t __ret; \
400 float32x2_t __s0 = __p0; \
401 __ret = (float32x4_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 9); \
402 __ret; \
403 })
404 #else
405 #define splatq_lane_f32(__p0, __p1) __extension__ ({ \
406 float32x4_t __ret; \
407 float32x2_t __s0 = __p0; \
408 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
409 __ret = (float32x4_t) __builtin_neon_splatq_lane_v((int8x8_t)__rev0, __p1, 9); \
410 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
411 __ret; \
412 })
413 #define __noswap_splatq_lane_f32(__p0, __p1) __extension__ ({ \
414 float32x4_t __ret; \
415 float32x2_t __s0 = __p0; \
416 __ret = (float32x4_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 9); \
417 __ret; \
418 })
419 #endif
420
421 #ifdef __LITTLE_ENDIAN__
422 #define splatq_lane_f16(__p0, __p1) __extension__ ({ \
423 float16x8_t __ret; \
424 float16x4_t __s0 = __p0; \
425 __ret = (float16x8_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 8); \
426 __ret; \
427 })
428 #else
429 #define splatq_lane_f16(__p0, __p1) __extension__ ({ \
430 float16x8_t __ret; \
431 float16x4_t __s0 = __p0; \
432 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
433 __ret = (float16x8_t) __builtin_neon_splatq_lane_v((int8x8_t)__rev0, __p1, 8); \
434 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
435 __ret; \
436 })
437 #define __noswap_splatq_lane_f16(__p0, __p1) __extension__ ({ \
438 float16x8_t __ret; \
439 float16x4_t __s0 = __p0; \
440 __ret = (float16x8_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 8); \
441 __ret; \
442 })
443 #endif
444
445 #ifdef __LITTLE_ENDIAN__
446 #define splatq_lane_s32(__p0, __p1) __extension__ ({ \
447 int32x4_t __ret; \
448 int32x2_t __s0 = __p0; \
449 __ret = (int32x4_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 2); \
450 __ret; \
451 })
452 #else
453 #define splatq_lane_s32(__p0, __p1) __extension__ ({ \
454 int32x4_t __ret; \
455 int32x2_t __s0 = __p0; \
456 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
457 __ret = (int32x4_t) __builtin_neon_splatq_lane_v((int8x8_t)__rev0, __p1, 2); \
458 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
459 __ret; \
460 })
461 #define __noswap_splatq_lane_s32(__p0, __p1) __extension__ ({ \
462 int32x4_t __ret; \
463 int32x2_t __s0 = __p0; \
464 __ret = (int32x4_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 2); \
465 __ret; \
466 })
467 #endif
468
469 #ifdef __LITTLE_ENDIAN__
470 #define splatq_lane_s64(__p0, __p1) __extension__ ({ \
471 int64x2_t __ret; \
472 int64x1_t __s0 = __p0; \
473 __ret = (int64x2_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 3); \
474 __ret; \
475 })
476 #else
477 #define splatq_lane_s64(__p0, __p1) __extension__ ({ \
478 int64x2_t __ret; \
479 int64x1_t __s0 = __p0; \
480 __ret = (int64x2_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 3); \
481 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
482 __ret; \
483 })
484 #define __noswap_splatq_lane_s64(__p0, __p1) __extension__ ({ \
485 int64x2_t __ret; \
486 int64x1_t __s0 = __p0; \
487 __ret = (int64x2_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 3); \
488 __ret; \
489 })
490 #endif
491
492 #ifdef __LITTLE_ENDIAN__
493 #define splatq_lane_s16(__p0, __p1) __extension__ ({ \
494 int16x8_t __ret; \
495 int16x4_t __s0 = __p0; \
496 __ret = (int16x8_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 1); \
497 __ret; \
498 })
499 #else
500 #define splatq_lane_s16(__p0, __p1) __extension__ ({ \
501 int16x8_t __ret; \
502 int16x4_t __s0 = __p0; \
503 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
504 __ret = (int16x8_t) __builtin_neon_splatq_lane_v((int8x8_t)__rev0, __p1, 1); \
505 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
506 __ret; \
507 })
508 #define __noswap_splatq_lane_s16(__p0, __p1) __extension__ ({ \
509 int16x8_t __ret; \
510 int16x4_t __s0 = __p0; \
511 __ret = (int16x8_t) __builtin_neon_splatq_lane_v((int8x8_t)__s0, __p1, 1); \
512 __ret; \
513 })
514 #endif
515
516 #ifdef __LITTLE_ENDIAN__
517 #define splat_lane_u8(__p0, __p1) __extension__ ({ \
518 uint8x8_t __ret; \
519 uint8x8_t __s0 = __p0; \
520 __ret = (uint8x8_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 16); \
521 __ret; \
522 })
523 #else
524 #define splat_lane_u8(__p0, __p1) __extension__ ({ \
525 uint8x8_t __ret; \
526 uint8x8_t __s0 = __p0; \
527 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
528 __ret = (uint8x8_t) __builtin_neon_splat_lane_v((int8x8_t)__rev0, __p1, 16); \
529 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
530 __ret; \
531 })
532 #define __noswap_splat_lane_u8(__p0, __p1) __extension__ ({ \
533 uint8x8_t __ret; \
534 uint8x8_t __s0 = __p0; \
535 __ret = (uint8x8_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 16); \
536 __ret; \
537 })
538 #endif
539
540 #ifdef __LITTLE_ENDIAN__
541 #define splat_lane_u32(__p0, __p1) __extension__ ({ \
542 uint32x2_t __ret; \
543 uint32x2_t __s0 = __p0; \
544 __ret = (uint32x2_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 18); \
545 __ret; \
546 })
547 #else
548 #define splat_lane_u32(__p0, __p1) __extension__ ({ \
549 uint32x2_t __ret; \
550 uint32x2_t __s0 = __p0; \
551 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
552 __ret = (uint32x2_t) __builtin_neon_splat_lane_v((int8x8_t)__rev0, __p1, 18); \
553 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
554 __ret; \
555 })
556 #define __noswap_splat_lane_u32(__p0, __p1) __extension__ ({ \
557 uint32x2_t __ret; \
558 uint32x2_t __s0 = __p0; \
559 __ret = (uint32x2_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 18); \
560 __ret; \
561 })
562 #endif
563
564 #define splat_lane_u64(__p0, __p1) __extension__ ({ \
565 uint64x1_t __ret; \
566 uint64x1_t __s0 = __p0; \
567 __ret = (uint64x1_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 19); \
568 __ret; \
569 })
570 #ifdef __LITTLE_ENDIAN__
571 #define splat_lane_u16(__p0, __p1) __extension__ ({ \
572 uint16x4_t __ret; \
573 uint16x4_t __s0 = __p0; \
574 __ret = (uint16x4_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 17); \
575 __ret; \
576 })
577 #else
578 #define splat_lane_u16(__p0, __p1) __extension__ ({ \
579 uint16x4_t __ret; \
580 uint16x4_t __s0 = __p0; \
581 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
582 __ret = (uint16x4_t) __builtin_neon_splat_lane_v((int8x8_t)__rev0, __p1, 17); \
583 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
584 __ret; \
585 })
586 #define __noswap_splat_lane_u16(__p0, __p1) __extension__ ({ \
587 uint16x4_t __ret; \
588 uint16x4_t __s0 = __p0; \
589 __ret = (uint16x4_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 17); \
590 __ret; \
591 })
592 #endif
593
594 #ifdef __LITTLE_ENDIAN__
595 #define splat_lane_s8(__p0, __p1) __extension__ ({ \
596 int8x8_t __ret; \
597 int8x8_t __s0 = __p0; \
598 __ret = (int8x8_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 0); \
599 __ret; \
600 })
601 #else
602 #define splat_lane_s8(__p0, __p1) __extension__ ({ \
603 int8x8_t __ret; \
604 int8x8_t __s0 = __p0; \
605 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
606 __ret = (int8x8_t) __builtin_neon_splat_lane_v((int8x8_t)__rev0, __p1, 0); \
607 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
608 __ret; \
609 })
610 #define __noswap_splat_lane_s8(__p0, __p1) __extension__ ({ \
611 int8x8_t __ret; \
612 int8x8_t __s0 = __p0; \
613 __ret = (int8x8_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 0); \
614 __ret; \
615 })
616 #endif
617
618 #define splat_lane_f64(__p0, __p1) __extension__ ({ \
619 float64x1_t __ret; \
620 float64x1_t __s0 = __p0; \
621 __ret = (float64x1_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 10); \
622 __ret; \
623 })
624 #ifdef __LITTLE_ENDIAN__
625 #define splat_lane_f32(__p0, __p1) __extension__ ({ \
626 float32x2_t __ret; \
627 float32x2_t __s0 = __p0; \
628 __ret = (float32x2_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 9); \
629 __ret; \
630 })
631 #else
632 #define splat_lane_f32(__p0, __p1) __extension__ ({ \
633 float32x2_t __ret; \
634 float32x2_t __s0 = __p0; \
635 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
636 __ret = (float32x2_t) __builtin_neon_splat_lane_v((int8x8_t)__rev0, __p1, 9); \
637 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
638 __ret; \
639 })
640 #define __noswap_splat_lane_f32(__p0, __p1) __extension__ ({ \
641 float32x2_t __ret; \
642 float32x2_t __s0 = __p0; \
643 __ret = (float32x2_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 9); \
644 __ret; \
645 })
646 #endif
647
648 #ifdef __LITTLE_ENDIAN__
649 #define splat_lane_f16(__p0, __p1) __extension__ ({ \
650 float16x4_t __ret; \
651 float16x4_t __s0 = __p0; \
652 __ret = (float16x4_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 8); \
653 __ret; \
654 })
655 #else
656 #define splat_lane_f16(__p0, __p1) __extension__ ({ \
657 float16x4_t __ret; \
658 float16x4_t __s0 = __p0; \
659 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
660 __ret = (float16x4_t) __builtin_neon_splat_lane_v((int8x8_t)__rev0, __p1, 8); \
661 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
662 __ret; \
663 })
664 #define __noswap_splat_lane_f16(__p0, __p1) __extension__ ({ \
665 float16x4_t __ret; \
666 float16x4_t __s0 = __p0; \
667 __ret = (float16x4_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 8); \
668 __ret; \
669 })
670 #endif
671
672 #ifdef __LITTLE_ENDIAN__
673 #define splat_lane_s32(__p0, __p1) __extension__ ({ \
674 int32x2_t __ret; \
675 int32x2_t __s0 = __p0; \
676 __ret = (int32x2_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 2); \
677 __ret; \
678 })
679 #else
680 #define splat_lane_s32(__p0, __p1) __extension__ ({ \
681 int32x2_t __ret; \
682 int32x2_t __s0 = __p0; \
683 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
684 __ret = (int32x2_t) __builtin_neon_splat_lane_v((int8x8_t)__rev0, __p1, 2); \
685 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
686 __ret; \
687 })
688 #define __noswap_splat_lane_s32(__p0, __p1) __extension__ ({ \
689 int32x2_t __ret; \
690 int32x2_t __s0 = __p0; \
691 __ret = (int32x2_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 2); \
692 __ret; \
693 })
694 #endif
695
696 #define splat_lane_s64(__p0, __p1) __extension__ ({ \
697 int64x1_t __ret; \
698 int64x1_t __s0 = __p0; \
699 __ret = (int64x1_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 3); \
700 __ret; \
701 })
702 #ifdef __LITTLE_ENDIAN__
703 #define splat_lane_s16(__p0, __p1) __extension__ ({ \
704 int16x4_t __ret; \
705 int16x4_t __s0 = __p0; \
706 __ret = (int16x4_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 1); \
707 __ret; \
708 })
709 #else
710 #define splat_lane_s16(__p0, __p1) __extension__ ({ \
711 int16x4_t __ret; \
712 int16x4_t __s0 = __p0; \
713 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
714 __ret = (int16x4_t) __builtin_neon_splat_lane_v((int8x8_t)__rev0, __p1, 1); \
715 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
716 __ret; \
717 })
718 #define __noswap_splat_lane_s16(__p0, __p1) __extension__ ({ \
719 int16x4_t __ret; \
720 int16x4_t __s0 = __p0; \
721 __ret = (int16x4_t) __builtin_neon_splat_lane_v((int8x8_t)__s0, __p1, 1); \
722 __ret; \
723 })
724 #endif
725
726 #ifdef __LITTLE_ENDIAN__
727 #define splat_laneq_p8(__p0, __p1) __extension__ ({ \
728 poly8x8_t __ret; \
729 poly8x16_t __s0 = __p0; \
730 __ret = (poly8x8_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 36); \
731 __ret; \
732 })
733 #else
734 #define splat_laneq_p8(__p0, __p1) __extension__ ({ \
735 poly8x8_t __ret; \
736 poly8x16_t __s0 = __p0; \
737 poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
738 __ret = (poly8x8_t) __builtin_neon_splat_laneq_v((int8x16_t)__rev0, __p1, 36); \
739 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
740 __ret; \
741 })
742 #define __noswap_splat_laneq_p8(__p0, __p1) __extension__ ({ \
743 poly8x8_t __ret; \
744 poly8x16_t __s0 = __p0; \
745 __ret = (poly8x8_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 36); \
746 __ret; \
747 })
748 #endif
749
750 #ifdef __LITTLE_ENDIAN__
751 #define splat_laneq_p64(__p0, __p1) __extension__ ({ \
752 poly64x1_t __ret; \
753 poly64x2_t __s0 = __p0; \
754 __ret = (poly64x1_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 38); \
755 __ret; \
756 })
757 #else
758 #define splat_laneq_p64(__p0, __p1) __extension__ ({ \
759 poly64x1_t __ret; \
760 poly64x2_t __s0 = __p0; \
761 poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
762 __ret = (poly64x1_t) __builtin_neon_splat_laneq_v((int8x16_t)__rev0, __p1, 38); \
763 __ret; \
764 })
765 #define __noswap_splat_laneq_p64(__p0, __p1) __extension__ ({ \
766 poly64x1_t __ret; \
767 poly64x2_t __s0 = __p0; \
768 __ret = (poly64x1_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 38); \
769 __ret; \
770 })
771 #endif
772
773 #ifdef __LITTLE_ENDIAN__
774 #define splat_laneq_p16(__p0, __p1) __extension__ ({ \
775 poly16x4_t __ret; \
776 poly16x8_t __s0 = __p0; \
777 __ret = (poly16x4_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 37); \
778 __ret; \
779 })
780 #else
781 #define splat_laneq_p16(__p0, __p1) __extension__ ({ \
782 poly16x4_t __ret; \
783 poly16x8_t __s0 = __p0; \
784 poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
785 __ret = (poly16x4_t) __builtin_neon_splat_laneq_v((int8x16_t)__rev0, __p1, 37); \
786 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
787 __ret; \
788 })
789 #define __noswap_splat_laneq_p16(__p0, __p1) __extension__ ({ \
790 poly16x4_t __ret; \
791 poly16x8_t __s0 = __p0; \
792 __ret = (poly16x4_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 37); \
793 __ret; \
794 })
795 #endif
796
797 #ifdef __LITTLE_ENDIAN__
798 #define splatq_laneq_p8(__p0, __p1) __extension__ ({ \
799 poly8x16_t __ret; \
800 poly8x16_t __s0 = __p0; \
801 __ret = (poly8x16_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 36); \
802 __ret; \
803 })
804 #else
805 #define splatq_laneq_p8(__p0, __p1) __extension__ ({ \
806 poly8x16_t __ret; \
807 poly8x16_t __s0 = __p0; \
808 poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
809 __ret = (poly8x16_t) __builtin_neon_splatq_laneq_v((int8x16_t)__rev0, __p1, 36); \
810 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
811 __ret; \
812 })
813 #define __noswap_splatq_laneq_p8(__p0, __p1) __extension__ ({ \
814 poly8x16_t __ret; \
815 poly8x16_t __s0 = __p0; \
816 __ret = (poly8x16_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 36); \
817 __ret; \
818 })
819 #endif
820
821 #ifdef __LITTLE_ENDIAN__
822 #define splatq_laneq_p64(__p0, __p1) __extension__ ({ \
823 poly64x2_t __ret; \
824 poly64x2_t __s0 = __p0; \
825 __ret = (poly64x2_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 38); \
826 __ret; \
827 })
828 #else
829 #define splatq_laneq_p64(__p0, __p1) __extension__ ({ \
830 poly64x2_t __ret; \
831 poly64x2_t __s0 = __p0; \
832 poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
833 __ret = (poly64x2_t) __builtin_neon_splatq_laneq_v((int8x16_t)__rev0, __p1, 38); \
834 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
835 __ret; \
836 })
837 #define __noswap_splatq_laneq_p64(__p0, __p1) __extension__ ({ \
838 poly64x2_t __ret; \
839 poly64x2_t __s0 = __p0; \
840 __ret = (poly64x2_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 38); \
841 __ret; \
842 })
843 #endif
844
845 #ifdef __LITTLE_ENDIAN__
846 #define splatq_laneq_p16(__p0, __p1) __extension__ ({ \
847 poly16x8_t __ret; \
848 poly16x8_t __s0 = __p0; \
849 __ret = (poly16x8_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 37); \
850 __ret; \
851 })
852 #else
853 #define splatq_laneq_p16(__p0, __p1) __extension__ ({ \
854 poly16x8_t __ret; \
855 poly16x8_t __s0 = __p0; \
856 poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
857 __ret = (poly16x8_t) __builtin_neon_splatq_laneq_v((int8x16_t)__rev0, __p1, 37); \
858 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
859 __ret; \
860 })
861 #define __noswap_splatq_laneq_p16(__p0, __p1) __extension__ ({ \
862 poly16x8_t __ret; \
863 poly16x8_t __s0 = __p0; \
864 __ret = (poly16x8_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 37); \
865 __ret; \
866 })
867 #endif
868
869 #ifdef __LITTLE_ENDIAN__
870 #define splatq_laneq_u8(__p0, __p1) __extension__ ({ \
871 uint8x16_t __ret; \
872 uint8x16_t __s0 = __p0; \
873 __ret = (uint8x16_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 48); \
874 __ret; \
875 })
876 #else
877 #define splatq_laneq_u8(__p0, __p1) __extension__ ({ \
878 uint8x16_t __ret; \
879 uint8x16_t __s0 = __p0; \
880 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
881 __ret = (uint8x16_t) __builtin_neon_splatq_laneq_v((int8x16_t)__rev0, __p1, 48); \
882 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
883 __ret; \
884 })
885 #define __noswap_splatq_laneq_u8(__p0, __p1) __extension__ ({ \
886 uint8x16_t __ret; \
887 uint8x16_t __s0 = __p0; \
888 __ret = (uint8x16_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 48); \
889 __ret; \
890 })
891 #endif
892
893 #ifdef __LITTLE_ENDIAN__
894 #define splatq_laneq_u32(__p0, __p1) __extension__ ({ \
895 uint32x4_t __ret; \
896 uint32x4_t __s0 = __p0; \
897 __ret = (uint32x4_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 50); \
898 __ret; \
899 })
900 #else
901 #define splatq_laneq_u32(__p0, __p1) __extension__ ({ \
902 uint32x4_t __ret; \
903 uint32x4_t __s0 = __p0; \
904 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
905 __ret = (uint32x4_t) __builtin_neon_splatq_laneq_v((int8x16_t)__rev0, __p1, 50); \
906 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
907 __ret; \
908 })
909 #define __noswap_splatq_laneq_u32(__p0, __p1) __extension__ ({ \
910 uint32x4_t __ret; \
911 uint32x4_t __s0 = __p0; \
912 __ret = (uint32x4_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 50); \
913 __ret; \
914 })
915 #endif
916
917 #ifdef __LITTLE_ENDIAN__
918 #define splatq_laneq_u64(__p0, __p1) __extension__ ({ \
919 uint64x2_t __ret; \
920 uint64x2_t __s0 = __p0; \
921 __ret = (uint64x2_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 51); \
922 __ret; \
923 })
924 #else
925 #define splatq_laneq_u64(__p0, __p1) __extension__ ({ \
926 uint64x2_t __ret; \
927 uint64x2_t __s0 = __p0; \
928 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
929 __ret = (uint64x2_t) __builtin_neon_splatq_laneq_v((int8x16_t)__rev0, __p1, 51); \
930 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
931 __ret; \
932 })
933 #define __noswap_splatq_laneq_u64(__p0, __p1) __extension__ ({ \
934 uint64x2_t __ret; \
935 uint64x2_t __s0 = __p0; \
936 __ret = (uint64x2_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 51); \
937 __ret; \
938 })
939 #endif
940
941 #ifdef __LITTLE_ENDIAN__
942 #define splatq_laneq_u16(__p0, __p1) __extension__ ({ \
943 uint16x8_t __ret; \
944 uint16x8_t __s0 = __p0; \
945 __ret = (uint16x8_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 49); \
946 __ret; \
947 })
948 #else
949 #define splatq_laneq_u16(__p0, __p1) __extension__ ({ \
950 uint16x8_t __ret; \
951 uint16x8_t __s0 = __p0; \
952 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
953 __ret = (uint16x8_t) __builtin_neon_splatq_laneq_v((int8x16_t)__rev0, __p1, 49); \
954 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
955 __ret; \
956 })
957 #define __noswap_splatq_laneq_u16(__p0, __p1) __extension__ ({ \
958 uint16x8_t __ret; \
959 uint16x8_t __s0 = __p0; \
960 __ret = (uint16x8_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 49); \
961 __ret; \
962 })
963 #endif
964
965 #ifdef __LITTLE_ENDIAN__
966 #define splatq_laneq_s8(__p0, __p1) __extension__ ({ \
967 int8x16_t __ret; \
968 int8x16_t __s0 = __p0; \
969 __ret = (int8x16_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 32); \
970 __ret; \
971 })
972 #else
973 #define splatq_laneq_s8(__p0, __p1) __extension__ ({ \
974 int8x16_t __ret; \
975 int8x16_t __s0 = __p0; \
976 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
977 __ret = (int8x16_t) __builtin_neon_splatq_laneq_v((int8x16_t)__rev0, __p1, 32); \
978 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
979 __ret; \
980 })
981 #define __noswap_splatq_laneq_s8(__p0, __p1) __extension__ ({ \
982 int8x16_t __ret; \
983 int8x16_t __s0 = __p0; \
984 __ret = (int8x16_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 32); \
985 __ret; \
986 })
987 #endif
988
989 #ifdef __LITTLE_ENDIAN__
990 #define splatq_laneq_f64(__p0, __p1) __extension__ ({ \
991 float64x2_t __ret; \
992 float64x2_t __s0 = __p0; \
993 __ret = (float64x2_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 42); \
994 __ret; \
995 })
996 #else
997 #define splatq_laneq_f64(__p0, __p1) __extension__ ({ \
998 float64x2_t __ret; \
999 float64x2_t __s0 = __p0; \
1000 float64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
1001 __ret = (float64x2_t) __builtin_neon_splatq_laneq_v((int8x16_t)__rev0, __p1, 42); \
1002 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
1003 __ret; \
1004 })
1005 #define __noswap_splatq_laneq_f64(__p0, __p1) __extension__ ({ \
1006 float64x2_t __ret; \
1007 float64x2_t __s0 = __p0; \
1008 __ret = (float64x2_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 42); \
1009 __ret; \
1010 })
1011 #endif
1012
1013 #ifdef __LITTLE_ENDIAN__
1014 #define splatq_laneq_f32(__p0, __p1) __extension__ ({ \
1015 float32x4_t __ret; \
1016 float32x4_t __s0 = __p0; \
1017 __ret = (float32x4_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 41); \
1018 __ret; \
1019 })
1020 #else
1021 #define splatq_laneq_f32(__p0, __p1) __extension__ ({ \
1022 float32x4_t __ret; \
1023 float32x4_t __s0 = __p0; \
1024 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
1025 __ret = (float32x4_t) __builtin_neon_splatq_laneq_v((int8x16_t)__rev0, __p1, 41); \
1026 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
1027 __ret; \
1028 })
1029 #define __noswap_splatq_laneq_f32(__p0, __p1) __extension__ ({ \
1030 float32x4_t __ret; \
1031 float32x4_t __s0 = __p0; \
1032 __ret = (float32x4_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 41); \
1033 __ret; \
1034 })
1035 #endif
1036
1037 #ifdef __LITTLE_ENDIAN__
1038 #define splatq_laneq_f16(__p0, __p1) __extension__ ({ \
1039 float16x8_t __ret; \
1040 float16x8_t __s0 = __p0; \
1041 __ret = (float16x8_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 40); \
1042 __ret; \
1043 })
1044 #else
1045 #define splatq_laneq_f16(__p0, __p1) __extension__ ({ \
1046 float16x8_t __ret; \
1047 float16x8_t __s0 = __p0; \
1048 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
1049 __ret = (float16x8_t) __builtin_neon_splatq_laneq_v((int8x16_t)__rev0, __p1, 40); \
1050 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
1051 __ret; \
1052 })
1053 #define __noswap_splatq_laneq_f16(__p0, __p1) __extension__ ({ \
1054 float16x8_t __ret; \
1055 float16x8_t __s0 = __p0; \
1056 __ret = (float16x8_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 40); \
1057 __ret; \
1058 })
1059 #endif
1060
1061 #ifdef __LITTLE_ENDIAN__
1062 #define splatq_laneq_s32(__p0, __p1) __extension__ ({ \
1063 int32x4_t __ret; \
1064 int32x4_t __s0 = __p0; \
1065 __ret = (int32x4_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 34); \
1066 __ret; \
1067 })
1068 #else
1069 #define splatq_laneq_s32(__p0, __p1) __extension__ ({ \
1070 int32x4_t __ret; \
1071 int32x4_t __s0 = __p0; \
1072 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
1073 __ret = (int32x4_t) __builtin_neon_splatq_laneq_v((int8x16_t)__rev0, __p1, 34); \
1074 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
1075 __ret; \
1076 })
1077 #define __noswap_splatq_laneq_s32(__p0, __p1) __extension__ ({ \
1078 int32x4_t __ret; \
1079 int32x4_t __s0 = __p0; \
1080 __ret = (int32x4_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 34); \
1081 __ret; \
1082 })
1083 #endif
1084
1085 #ifdef __LITTLE_ENDIAN__
1086 #define splatq_laneq_s64(__p0, __p1) __extension__ ({ \
1087 int64x2_t __ret; \
1088 int64x2_t __s0 = __p0; \
1089 __ret = (int64x2_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 35); \
1090 __ret; \
1091 })
1092 #else
1093 #define splatq_laneq_s64(__p0, __p1) __extension__ ({ \
1094 int64x2_t __ret; \
1095 int64x2_t __s0 = __p0; \
1096 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
1097 __ret = (int64x2_t) __builtin_neon_splatq_laneq_v((int8x16_t)__rev0, __p1, 35); \
1098 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
1099 __ret; \
1100 })
1101 #define __noswap_splatq_laneq_s64(__p0, __p1) __extension__ ({ \
1102 int64x2_t __ret; \
1103 int64x2_t __s0 = __p0; \
1104 __ret = (int64x2_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 35); \
1105 __ret; \
1106 })
1107 #endif
1108
1109 #ifdef __LITTLE_ENDIAN__
1110 #define splatq_laneq_s16(__p0, __p1) __extension__ ({ \
1111 int16x8_t __ret; \
1112 int16x8_t __s0 = __p0; \
1113 __ret = (int16x8_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 33); \
1114 __ret; \
1115 })
1116 #else
1117 #define splatq_laneq_s16(__p0, __p1) __extension__ ({ \
1118 int16x8_t __ret; \
1119 int16x8_t __s0 = __p0; \
1120 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
1121 __ret = (int16x8_t) __builtin_neon_splatq_laneq_v((int8x16_t)__rev0, __p1, 33); \
1122 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
1123 __ret; \
1124 })
1125 #define __noswap_splatq_laneq_s16(__p0, __p1) __extension__ ({ \
1126 int16x8_t __ret; \
1127 int16x8_t __s0 = __p0; \
1128 __ret = (int16x8_t) __builtin_neon_splatq_laneq_v((int8x16_t)__s0, __p1, 33); \
1129 __ret; \
1130 })
1131 #endif
1132
1133 #ifdef __LITTLE_ENDIAN__
1134 #define splat_laneq_u8(__p0, __p1) __extension__ ({ \
1135 uint8x8_t __ret; \
1136 uint8x16_t __s0 = __p0; \
1137 __ret = (uint8x8_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 48); \
1138 __ret; \
1139 })
1140 #else
1141 #define splat_laneq_u8(__p0, __p1) __extension__ ({ \
1142 uint8x8_t __ret; \
1143 uint8x16_t __s0 = __p0; \
1144 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
1145 __ret = (uint8x8_t) __builtin_neon_splat_laneq_v((int8x16_t)__rev0, __p1, 48); \
1146 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
1147 __ret; \
1148 })
1149 #define __noswap_splat_laneq_u8(__p0, __p1) __extension__ ({ \
1150 uint8x8_t __ret; \
1151 uint8x16_t __s0 = __p0; \
1152 __ret = (uint8x8_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 48); \
1153 __ret; \
1154 })
1155 #endif
1156
1157 #ifdef __LITTLE_ENDIAN__
1158 #define splat_laneq_u32(__p0, __p1) __extension__ ({ \
1159 uint32x2_t __ret; \
1160 uint32x4_t __s0 = __p0; \
1161 __ret = (uint32x2_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 50); \
1162 __ret; \
1163 })
1164 #else
1165 #define splat_laneq_u32(__p0, __p1) __extension__ ({ \
1166 uint32x2_t __ret; \
1167 uint32x4_t __s0 = __p0; \
1168 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
1169 __ret = (uint32x2_t) __builtin_neon_splat_laneq_v((int8x16_t)__rev0, __p1, 50); \
1170 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
1171 __ret; \
1172 })
1173 #define __noswap_splat_laneq_u32(__p0, __p1) __extension__ ({ \
1174 uint32x2_t __ret; \
1175 uint32x4_t __s0 = __p0; \
1176 __ret = (uint32x2_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 50); \
1177 __ret; \
1178 })
1179 #endif
1180
1181 #ifdef __LITTLE_ENDIAN__
1182 #define splat_laneq_u64(__p0, __p1) __extension__ ({ \
1183 uint64x1_t __ret; \
1184 uint64x2_t __s0 = __p0; \
1185 __ret = (uint64x1_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 51); \
1186 __ret; \
1187 })
1188 #else
1189 #define splat_laneq_u64(__p0, __p1) __extension__ ({ \
1190 uint64x1_t __ret; \
1191 uint64x2_t __s0 = __p0; \
1192 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
1193 __ret = (uint64x1_t) __builtin_neon_splat_laneq_v((int8x16_t)__rev0, __p1, 51); \
1194 __ret; \
1195 })
1196 #define __noswap_splat_laneq_u64(__p0, __p1) __extension__ ({ \
1197 uint64x1_t __ret; \
1198 uint64x2_t __s0 = __p0; \
1199 __ret = (uint64x1_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 51); \
1200 __ret; \
1201 })
1202 #endif
1203
1204 #ifdef __LITTLE_ENDIAN__
1205 #define splat_laneq_u16(__p0, __p1) __extension__ ({ \
1206 uint16x4_t __ret; \
1207 uint16x8_t __s0 = __p0; \
1208 __ret = (uint16x4_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 49); \
1209 __ret; \
1210 })
1211 #else
1212 #define splat_laneq_u16(__p0, __p1) __extension__ ({ \
1213 uint16x4_t __ret; \
1214 uint16x8_t __s0 = __p0; \
1215 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
1216 __ret = (uint16x4_t) __builtin_neon_splat_laneq_v((int8x16_t)__rev0, __p1, 49); \
1217 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
1218 __ret; \
1219 })
1220 #define __noswap_splat_laneq_u16(__p0, __p1) __extension__ ({ \
1221 uint16x4_t __ret; \
1222 uint16x8_t __s0 = __p0; \
1223 __ret = (uint16x4_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 49); \
1224 __ret; \
1225 })
1226 #endif
1227
1228 #ifdef __LITTLE_ENDIAN__
1229 #define splat_laneq_s8(__p0, __p1) __extension__ ({ \
1230 int8x8_t __ret; \
1231 int8x16_t __s0 = __p0; \
1232 __ret = (int8x8_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 32); \
1233 __ret; \
1234 })
1235 #else
1236 #define splat_laneq_s8(__p0, __p1) __extension__ ({ \
1237 int8x8_t __ret; \
1238 int8x16_t __s0 = __p0; \
1239 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
1240 __ret = (int8x8_t) __builtin_neon_splat_laneq_v((int8x16_t)__rev0, __p1, 32); \
1241 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
1242 __ret; \
1243 })
1244 #define __noswap_splat_laneq_s8(__p0, __p1) __extension__ ({ \
1245 int8x8_t __ret; \
1246 int8x16_t __s0 = __p0; \
1247 __ret = (int8x8_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 32); \
1248 __ret; \
1249 })
1250 #endif
1251
1252 #ifdef __LITTLE_ENDIAN__
1253 #define splat_laneq_f64(__p0, __p1) __extension__ ({ \
1254 float64x1_t __ret; \
1255 float64x2_t __s0 = __p0; \
1256 __ret = (float64x1_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 42); \
1257 __ret; \
1258 })
1259 #else
1260 #define splat_laneq_f64(__p0, __p1) __extension__ ({ \
1261 float64x1_t __ret; \
1262 float64x2_t __s0 = __p0; \
1263 float64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
1264 __ret = (float64x1_t) __builtin_neon_splat_laneq_v((int8x16_t)__rev0, __p1, 42); \
1265 __ret; \
1266 })
1267 #define __noswap_splat_laneq_f64(__p0, __p1) __extension__ ({ \
1268 float64x1_t __ret; \
1269 float64x2_t __s0 = __p0; \
1270 __ret = (float64x1_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 42); \
1271 __ret; \
1272 })
1273 #endif
1274
1275 #ifdef __LITTLE_ENDIAN__
1276 #define splat_laneq_f32(__p0, __p1) __extension__ ({ \
1277 float32x2_t __ret; \
1278 float32x4_t __s0 = __p0; \
1279 __ret = (float32x2_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 41); \
1280 __ret; \
1281 })
1282 #else
1283 #define splat_laneq_f32(__p0, __p1) __extension__ ({ \
1284 float32x2_t __ret; \
1285 float32x4_t __s0 = __p0; \
1286 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
1287 __ret = (float32x2_t) __builtin_neon_splat_laneq_v((int8x16_t)__rev0, __p1, 41); \
1288 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
1289 __ret; \
1290 })
1291 #define __noswap_splat_laneq_f32(__p0, __p1) __extension__ ({ \
1292 float32x2_t __ret; \
1293 float32x4_t __s0 = __p0; \
1294 __ret = (float32x2_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 41); \
1295 __ret; \
1296 })
1297 #endif
1298
1299 #ifdef __LITTLE_ENDIAN__
1300 #define splat_laneq_f16(__p0, __p1) __extension__ ({ \
1301 float16x4_t __ret; \
1302 float16x8_t __s0 = __p0; \
1303 __ret = (float16x4_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 40); \
1304 __ret; \
1305 })
1306 #else
1307 #define splat_laneq_f16(__p0, __p1) __extension__ ({ \
1308 float16x4_t __ret; \
1309 float16x8_t __s0 = __p0; \
1310 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
1311 __ret = (float16x4_t) __builtin_neon_splat_laneq_v((int8x16_t)__rev0, __p1, 40); \
1312 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
1313 __ret; \
1314 })
1315 #define __noswap_splat_laneq_f16(__p0, __p1) __extension__ ({ \
1316 float16x4_t __ret; \
1317 float16x8_t __s0 = __p0; \
1318 __ret = (float16x4_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 40); \
1319 __ret; \
1320 })
1321 #endif
1322
1323 #ifdef __LITTLE_ENDIAN__
1324 #define splat_laneq_s32(__p0, __p1) __extension__ ({ \
1325 int32x2_t __ret; \
1326 int32x4_t __s0 = __p0; \
1327 __ret = (int32x2_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 34); \
1328 __ret; \
1329 })
1330 #else
1331 #define splat_laneq_s32(__p0, __p1) __extension__ ({ \
1332 int32x2_t __ret; \
1333 int32x4_t __s0 = __p0; \
1334 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
1335 __ret = (int32x2_t) __builtin_neon_splat_laneq_v((int8x16_t)__rev0, __p1, 34); \
1336 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
1337 __ret; \
1338 })
1339 #define __noswap_splat_laneq_s32(__p0, __p1) __extension__ ({ \
1340 int32x2_t __ret; \
1341 int32x4_t __s0 = __p0; \
1342 __ret = (int32x2_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 34); \
1343 __ret; \
1344 })
1345 #endif
1346
1347 #ifdef __LITTLE_ENDIAN__
1348 #define splat_laneq_s64(__p0, __p1) __extension__ ({ \
1349 int64x1_t __ret; \
1350 int64x2_t __s0 = __p0; \
1351 __ret = (int64x1_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 35); \
1352 __ret; \
1353 })
1354 #else
1355 #define splat_laneq_s64(__p0, __p1) __extension__ ({ \
1356 int64x1_t __ret; \
1357 int64x2_t __s0 = __p0; \
1358 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
1359 __ret = (int64x1_t) __builtin_neon_splat_laneq_v((int8x16_t)__rev0, __p1, 35); \
1360 __ret; \
1361 })
1362 #define __noswap_splat_laneq_s64(__p0, __p1) __extension__ ({ \
1363 int64x1_t __ret; \
1364 int64x2_t __s0 = __p0; \
1365 __ret = (int64x1_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 35); \
1366 __ret; \
1367 })
1368 #endif
1369
1370 #ifdef __LITTLE_ENDIAN__
1371 #define splat_laneq_s16(__p0, __p1) __extension__ ({ \
1372 int16x4_t __ret; \
1373 int16x8_t __s0 = __p0; \
1374 __ret = (int16x4_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 33); \
1375 __ret; \
1376 })
1377 #else
1378 #define splat_laneq_s16(__p0, __p1) __extension__ ({ \
1379 int16x4_t __ret; \
1380 int16x8_t __s0 = __p0; \
1381 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
1382 __ret = (int16x4_t) __builtin_neon_splat_laneq_v((int8x16_t)__rev0, __p1, 33); \
1383 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
1384 __ret; \
1385 })
1386 #define __noswap_splat_laneq_s16(__p0, __p1) __extension__ ({ \
1387 int16x4_t __ret; \
1388 int16x8_t __s0 = __p0; \
1389 __ret = (int16x4_t) __builtin_neon_splat_laneq_v((int8x16_t)__s0, __p1, 33); \
1390 __ret; \
1391 })
1392 #endif
1393
1394 #ifdef __LITTLE_ENDIAN__
vabdq_u8(uint8x16_t __p0,uint8x16_t __p1)1395 __ai uint8x16_t vabdq_u8(uint8x16_t __p0, uint8x16_t __p1) {
1396 uint8x16_t __ret;
1397 __ret = (uint8x16_t) __builtin_neon_vabdq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
1398 return __ret;
1399 }
1400 #else
vabdq_u8(uint8x16_t __p0,uint8x16_t __p1)1401 __ai uint8x16_t vabdq_u8(uint8x16_t __p0, uint8x16_t __p1) {
1402 uint8x16_t __ret;
1403 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
1404 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
1405 __ret = (uint8x16_t) __builtin_neon_vabdq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
1406 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
1407 return __ret;
1408 }
__noswap_vabdq_u8(uint8x16_t __p0,uint8x16_t __p1)1409 __ai uint8x16_t __noswap_vabdq_u8(uint8x16_t __p0, uint8x16_t __p1) {
1410 uint8x16_t __ret;
1411 __ret = (uint8x16_t) __builtin_neon_vabdq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
1412 return __ret;
1413 }
1414 #endif
1415
1416 #ifdef __LITTLE_ENDIAN__
vabdq_u32(uint32x4_t __p0,uint32x4_t __p1)1417 __ai uint32x4_t vabdq_u32(uint32x4_t __p0, uint32x4_t __p1) {
1418 uint32x4_t __ret;
1419 __ret = (uint32x4_t) __builtin_neon_vabdq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
1420 return __ret;
1421 }
1422 #else
vabdq_u32(uint32x4_t __p0,uint32x4_t __p1)1423 __ai uint32x4_t vabdq_u32(uint32x4_t __p0, uint32x4_t __p1) {
1424 uint32x4_t __ret;
1425 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
1426 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
1427 __ret = (uint32x4_t) __builtin_neon_vabdq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
1428 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
1429 return __ret;
1430 }
__noswap_vabdq_u32(uint32x4_t __p0,uint32x4_t __p1)1431 __ai uint32x4_t __noswap_vabdq_u32(uint32x4_t __p0, uint32x4_t __p1) {
1432 uint32x4_t __ret;
1433 __ret = (uint32x4_t) __builtin_neon_vabdq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
1434 return __ret;
1435 }
1436 #endif
1437
1438 #ifdef __LITTLE_ENDIAN__
vabdq_u16(uint16x8_t __p0,uint16x8_t __p1)1439 __ai uint16x8_t vabdq_u16(uint16x8_t __p0, uint16x8_t __p1) {
1440 uint16x8_t __ret;
1441 __ret = (uint16x8_t) __builtin_neon_vabdq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
1442 return __ret;
1443 }
1444 #else
vabdq_u16(uint16x8_t __p0,uint16x8_t __p1)1445 __ai uint16x8_t vabdq_u16(uint16x8_t __p0, uint16x8_t __p1) {
1446 uint16x8_t __ret;
1447 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
1448 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
1449 __ret = (uint16x8_t) __builtin_neon_vabdq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
1450 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
1451 return __ret;
1452 }
__noswap_vabdq_u16(uint16x8_t __p0,uint16x8_t __p1)1453 __ai uint16x8_t __noswap_vabdq_u16(uint16x8_t __p0, uint16x8_t __p1) {
1454 uint16x8_t __ret;
1455 __ret = (uint16x8_t) __builtin_neon_vabdq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
1456 return __ret;
1457 }
1458 #endif
1459
1460 #ifdef __LITTLE_ENDIAN__
vabdq_s8(int8x16_t __p0,int8x16_t __p1)1461 __ai int8x16_t vabdq_s8(int8x16_t __p0, int8x16_t __p1) {
1462 int8x16_t __ret;
1463 __ret = (int8x16_t) __builtin_neon_vabdq_v((int8x16_t)__p0, (int8x16_t)__p1, 32);
1464 return __ret;
1465 }
1466 #else
vabdq_s8(int8x16_t __p0,int8x16_t __p1)1467 __ai int8x16_t vabdq_s8(int8x16_t __p0, int8x16_t __p1) {
1468 int8x16_t __ret;
1469 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
1470 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
1471 __ret = (int8x16_t) __builtin_neon_vabdq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 32);
1472 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
1473 return __ret;
1474 }
__noswap_vabdq_s8(int8x16_t __p0,int8x16_t __p1)1475 __ai int8x16_t __noswap_vabdq_s8(int8x16_t __p0, int8x16_t __p1) {
1476 int8x16_t __ret;
1477 __ret = (int8x16_t) __builtin_neon_vabdq_v((int8x16_t)__p0, (int8x16_t)__p1, 32);
1478 return __ret;
1479 }
1480 #endif
1481
1482 #ifdef __LITTLE_ENDIAN__
vabdq_f32(float32x4_t __p0,float32x4_t __p1)1483 __ai float32x4_t vabdq_f32(float32x4_t __p0, float32x4_t __p1) {
1484 float32x4_t __ret;
1485 __ret = (float32x4_t) __builtin_neon_vabdq_v((int8x16_t)__p0, (int8x16_t)__p1, 41);
1486 return __ret;
1487 }
1488 #else
vabdq_f32(float32x4_t __p0,float32x4_t __p1)1489 __ai float32x4_t vabdq_f32(float32x4_t __p0, float32x4_t __p1) {
1490 float32x4_t __ret;
1491 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
1492 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
1493 __ret = (float32x4_t) __builtin_neon_vabdq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 41);
1494 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
1495 return __ret;
1496 }
1497 #endif
1498
1499 #ifdef __LITTLE_ENDIAN__
vabdq_s32(int32x4_t __p0,int32x4_t __p1)1500 __ai int32x4_t vabdq_s32(int32x4_t __p0, int32x4_t __p1) {
1501 int32x4_t __ret;
1502 __ret = (int32x4_t) __builtin_neon_vabdq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
1503 return __ret;
1504 }
1505 #else
vabdq_s32(int32x4_t __p0,int32x4_t __p1)1506 __ai int32x4_t vabdq_s32(int32x4_t __p0, int32x4_t __p1) {
1507 int32x4_t __ret;
1508 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
1509 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
1510 __ret = (int32x4_t) __builtin_neon_vabdq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
1511 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
1512 return __ret;
1513 }
__noswap_vabdq_s32(int32x4_t __p0,int32x4_t __p1)1514 __ai int32x4_t __noswap_vabdq_s32(int32x4_t __p0, int32x4_t __p1) {
1515 int32x4_t __ret;
1516 __ret = (int32x4_t) __builtin_neon_vabdq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
1517 return __ret;
1518 }
1519 #endif
1520
1521 #ifdef __LITTLE_ENDIAN__
vabdq_s16(int16x8_t __p0,int16x8_t __p1)1522 __ai int16x8_t vabdq_s16(int16x8_t __p0, int16x8_t __p1) {
1523 int16x8_t __ret;
1524 __ret = (int16x8_t) __builtin_neon_vabdq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
1525 return __ret;
1526 }
1527 #else
vabdq_s16(int16x8_t __p0,int16x8_t __p1)1528 __ai int16x8_t vabdq_s16(int16x8_t __p0, int16x8_t __p1) {
1529 int16x8_t __ret;
1530 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
1531 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
1532 __ret = (int16x8_t) __builtin_neon_vabdq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
1533 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
1534 return __ret;
1535 }
__noswap_vabdq_s16(int16x8_t __p0,int16x8_t __p1)1536 __ai int16x8_t __noswap_vabdq_s16(int16x8_t __p0, int16x8_t __p1) {
1537 int16x8_t __ret;
1538 __ret = (int16x8_t) __builtin_neon_vabdq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
1539 return __ret;
1540 }
1541 #endif
1542
1543 #ifdef __LITTLE_ENDIAN__
vabd_u8(uint8x8_t __p0,uint8x8_t __p1)1544 __ai uint8x8_t vabd_u8(uint8x8_t __p0, uint8x8_t __p1) {
1545 uint8x8_t __ret;
1546 __ret = (uint8x8_t) __builtin_neon_vabd_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
1547 return __ret;
1548 }
1549 #else
vabd_u8(uint8x8_t __p0,uint8x8_t __p1)1550 __ai uint8x8_t vabd_u8(uint8x8_t __p0, uint8x8_t __p1) {
1551 uint8x8_t __ret;
1552 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
1553 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
1554 __ret = (uint8x8_t) __builtin_neon_vabd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
1555 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
1556 return __ret;
1557 }
__noswap_vabd_u8(uint8x8_t __p0,uint8x8_t __p1)1558 __ai uint8x8_t __noswap_vabd_u8(uint8x8_t __p0, uint8x8_t __p1) {
1559 uint8x8_t __ret;
1560 __ret = (uint8x8_t) __builtin_neon_vabd_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
1561 return __ret;
1562 }
1563 #endif
1564
1565 #ifdef __LITTLE_ENDIAN__
vabd_u32(uint32x2_t __p0,uint32x2_t __p1)1566 __ai uint32x2_t vabd_u32(uint32x2_t __p0, uint32x2_t __p1) {
1567 uint32x2_t __ret;
1568 __ret = (uint32x2_t) __builtin_neon_vabd_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
1569 return __ret;
1570 }
1571 #else
vabd_u32(uint32x2_t __p0,uint32x2_t __p1)1572 __ai uint32x2_t vabd_u32(uint32x2_t __p0, uint32x2_t __p1) {
1573 uint32x2_t __ret;
1574 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
1575 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
1576 __ret = (uint32x2_t) __builtin_neon_vabd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
1577 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
1578 return __ret;
1579 }
__noswap_vabd_u32(uint32x2_t __p0,uint32x2_t __p1)1580 __ai uint32x2_t __noswap_vabd_u32(uint32x2_t __p0, uint32x2_t __p1) {
1581 uint32x2_t __ret;
1582 __ret = (uint32x2_t) __builtin_neon_vabd_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
1583 return __ret;
1584 }
1585 #endif
1586
1587 #ifdef __LITTLE_ENDIAN__
vabd_u16(uint16x4_t __p0,uint16x4_t __p1)1588 __ai uint16x4_t vabd_u16(uint16x4_t __p0, uint16x4_t __p1) {
1589 uint16x4_t __ret;
1590 __ret = (uint16x4_t) __builtin_neon_vabd_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
1591 return __ret;
1592 }
1593 #else
vabd_u16(uint16x4_t __p0,uint16x4_t __p1)1594 __ai uint16x4_t vabd_u16(uint16x4_t __p0, uint16x4_t __p1) {
1595 uint16x4_t __ret;
1596 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
1597 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
1598 __ret = (uint16x4_t) __builtin_neon_vabd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
1599 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
1600 return __ret;
1601 }
__noswap_vabd_u16(uint16x4_t __p0,uint16x4_t __p1)1602 __ai uint16x4_t __noswap_vabd_u16(uint16x4_t __p0, uint16x4_t __p1) {
1603 uint16x4_t __ret;
1604 __ret = (uint16x4_t) __builtin_neon_vabd_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
1605 return __ret;
1606 }
1607 #endif
1608
1609 #ifdef __LITTLE_ENDIAN__
vabd_s8(int8x8_t __p0,int8x8_t __p1)1610 __ai int8x8_t vabd_s8(int8x8_t __p0, int8x8_t __p1) {
1611 int8x8_t __ret;
1612 __ret = (int8x8_t) __builtin_neon_vabd_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
1613 return __ret;
1614 }
1615 #else
vabd_s8(int8x8_t __p0,int8x8_t __p1)1616 __ai int8x8_t vabd_s8(int8x8_t __p0, int8x8_t __p1) {
1617 int8x8_t __ret;
1618 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
1619 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
1620 __ret = (int8x8_t) __builtin_neon_vabd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 0);
1621 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
1622 return __ret;
1623 }
__noswap_vabd_s8(int8x8_t __p0,int8x8_t __p1)1624 __ai int8x8_t __noswap_vabd_s8(int8x8_t __p0, int8x8_t __p1) {
1625 int8x8_t __ret;
1626 __ret = (int8x8_t) __builtin_neon_vabd_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
1627 return __ret;
1628 }
1629 #endif
1630
1631 #ifdef __LITTLE_ENDIAN__
vabd_f32(float32x2_t __p0,float32x2_t __p1)1632 __ai float32x2_t vabd_f32(float32x2_t __p0, float32x2_t __p1) {
1633 float32x2_t __ret;
1634 __ret = (float32x2_t) __builtin_neon_vabd_v((int8x8_t)__p0, (int8x8_t)__p1, 9);
1635 return __ret;
1636 }
1637 #else
vabd_f32(float32x2_t __p0,float32x2_t __p1)1638 __ai float32x2_t vabd_f32(float32x2_t __p0, float32x2_t __p1) {
1639 float32x2_t __ret;
1640 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
1641 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
1642 __ret = (float32x2_t) __builtin_neon_vabd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 9);
1643 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
1644 return __ret;
1645 }
1646 #endif
1647
1648 #ifdef __LITTLE_ENDIAN__
vabd_s32(int32x2_t __p0,int32x2_t __p1)1649 __ai int32x2_t vabd_s32(int32x2_t __p0, int32x2_t __p1) {
1650 int32x2_t __ret;
1651 __ret = (int32x2_t) __builtin_neon_vabd_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
1652 return __ret;
1653 }
1654 #else
vabd_s32(int32x2_t __p0,int32x2_t __p1)1655 __ai int32x2_t vabd_s32(int32x2_t __p0, int32x2_t __p1) {
1656 int32x2_t __ret;
1657 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
1658 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
1659 __ret = (int32x2_t) __builtin_neon_vabd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
1660 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
1661 return __ret;
1662 }
__noswap_vabd_s32(int32x2_t __p0,int32x2_t __p1)1663 __ai int32x2_t __noswap_vabd_s32(int32x2_t __p0, int32x2_t __p1) {
1664 int32x2_t __ret;
1665 __ret = (int32x2_t) __builtin_neon_vabd_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
1666 return __ret;
1667 }
1668 #endif
1669
1670 #ifdef __LITTLE_ENDIAN__
vabd_s16(int16x4_t __p0,int16x4_t __p1)1671 __ai int16x4_t vabd_s16(int16x4_t __p0, int16x4_t __p1) {
1672 int16x4_t __ret;
1673 __ret = (int16x4_t) __builtin_neon_vabd_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
1674 return __ret;
1675 }
1676 #else
vabd_s16(int16x4_t __p0,int16x4_t __p1)1677 __ai int16x4_t vabd_s16(int16x4_t __p0, int16x4_t __p1) {
1678 int16x4_t __ret;
1679 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
1680 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
1681 __ret = (int16x4_t) __builtin_neon_vabd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
1682 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
1683 return __ret;
1684 }
__noswap_vabd_s16(int16x4_t __p0,int16x4_t __p1)1685 __ai int16x4_t __noswap_vabd_s16(int16x4_t __p0, int16x4_t __p1) {
1686 int16x4_t __ret;
1687 __ret = (int16x4_t) __builtin_neon_vabd_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
1688 return __ret;
1689 }
1690 #endif
1691
1692 #ifdef __LITTLE_ENDIAN__
vabsq_s8(int8x16_t __p0)1693 __ai int8x16_t vabsq_s8(int8x16_t __p0) {
1694 int8x16_t __ret;
1695 __ret = (int8x16_t) __builtin_neon_vabsq_v((int8x16_t)__p0, 32);
1696 return __ret;
1697 }
1698 #else
vabsq_s8(int8x16_t __p0)1699 __ai int8x16_t vabsq_s8(int8x16_t __p0) {
1700 int8x16_t __ret;
1701 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
1702 __ret = (int8x16_t) __builtin_neon_vabsq_v((int8x16_t)__rev0, 32);
1703 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
1704 return __ret;
1705 }
1706 #endif
1707
1708 #ifdef __LITTLE_ENDIAN__
vabsq_f32(float32x4_t __p0)1709 __ai float32x4_t vabsq_f32(float32x4_t __p0) {
1710 float32x4_t __ret;
1711 __ret = (float32x4_t) __builtin_neon_vabsq_v((int8x16_t)__p0, 41);
1712 return __ret;
1713 }
1714 #else
vabsq_f32(float32x4_t __p0)1715 __ai float32x4_t vabsq_f32(float32x4_t __p0) {
1716 float32x4_t __ret;
1717 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
1718 __ret = (float32x4_t) __builtin_neon_vabsq_v((int8x16_t)__rev0, 41);
1719 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
1720 return __ret;
1721 }
1722 #endif
1723
1724 #ifdef __LITTLE_ENDIAN__
vabsq_s32(int32x4_t __p0)1725 __ai int32x4_t vabsq_s32(int32x4_t __p0) {
1726 int32x4_t __ret;
1727 __ret = (int32x4_t) __builtin_neon_vabsq_v((int8x16_t)__p0, 34);
1728 return __ret;
1729 }
1730 #else
vabsq_s32(int32x4_t __p0)1731 __ai int32x4_t vabsq_s32(int32x4_t __p0) {
1732 int32x4_t __ret;
1733 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
1734 __ret = (int32x4_t) __builtin_neon_vabsq_v((int8x16_t)__rev0, 34);
1735 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
1736 return __ret;
1737 }
1738 #endif
1739
1740 #ifdef __LITTLE_ENDIAN__
vabsq_s16(int16x8_t __p0)1741 __ai int16x8_t vabsq_s16(int16x8_t __p0) {
1742 int16x8_t __ret;
1743 __ret = (int16x8_t) __builtin_neon_vabsq_v((int8x16_t)__p0, 33);
1744 return __ret;
1745 }
1746 #else
vabsq_s16(int16x8_t __p0)1747 __ai int16x8_t vabsq_s16(int16x8_t __p0) {
1748 int16x8_t __ret;
1749 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
1750 __ret = (int16x8_t) __builtin_neon_vabsq_v((int8x16_t)__rev0, 33);
1751 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
1752 return __ret;
1753 }
1754 #endif
1755
1756 #ifdef __LITTLE_ENDIAN__
vabs_s8(int8x8_t __p0)1757 __ai int8x8_t vabs_s8(int8x8_t __p0) {
1758 int8x8_t __ret;
1759 __ret = (int8x8_t) __builtin_neon_vabs_v((int8x8_t)__p0, 0);
1760 return __ret;
1761 }
1762 #else
vabs_s8(int8x8_t __p0)1763 __ai int8x8_t vabs_s8(int8x8_t __p0) {
1764 int8x8_t __ret;
1765 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
1766 __ret = (int8x8_t) __builtin_neon_vabs_v((int8x8_t)__rev0, 0);
1767 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
1768 return __ret;
1769 }
1770 #endif
1771
1772 #ifdef __LITTLE_ENDIAN__
vabs_f32(float32x2_t __p0)1773 __ai float32x2_t vabs_f32(float32x2_t __p0) {
1774 float32x2_t __ret;
1775 __ret = (float32x2_t) __builtin_neon_vabs_v((int8x8_t)__p0, 9);
1776 return __ret;
1777 }
1778 #else
vabs_f32(float32x2_t __p0)1779 __ai float32x2_t vabs_f32(float32x2_t __p0) {
1780 float32x2_t __ret;
1781 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
1782 __ret = (float32x2_t) __builtin_neon_vabs_v((int8x8_t)__rev0, 9);
1783 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
1784 return __ret;
1785 }
1786 #endif
1787
1788 #ifdef __LITTLE_ENDIAN__
vabs_s32(int32x2_t __p0)1789 __ai int32x2_t vabs_s32(int32x2_t __p0) {
1790 int32x2_t __ret;
1791 __ret = (int32x2_t) __builtin_neon_vabs_v((int8x8_t)__p0, 2);
1792 return __ret;
1793 }
1794 #else
vabs_s32(int32x2_t __p0)1795 __ai int32x2_t vabs_s32(int32x2_t __p0) {
1796 int32x2_t __ret;
1797 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
1798 __ret = (int32x2_t) __builtin_neon_vabs_v((int8x8_t)__rev0, 2);
1799 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
1800 return __ret;
1801 }
1802 #endif
1803
1804 #ifdef __LITTLE_ENDIAN__
vabs_s16(int16x4_t __p0)1805 __ai int16x4_t vabs_s16(int16x4_t __p0) {
1806 int16x4_t __ret;
1807 __ret = (int16x4_t) __builtin_neon_vabs_v((int8x8_t)__p0, 1);
1808 return __ret;
1809 }
1810 #else
vabs_s16(int16x4_t __p0)1811 __ai int16x4_t vabs_s16(int16x4_t __p0) {
1812 int16x4_t __ret;
1813 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
1814 __ret = (int16x4_t) __builtin_neon_vabs_v((int8x8_t)__rev0, 1);
1815 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
1816 return __ret;
1817 }
1818 #endif
1819
1820 #ifdef __LITTLE_ENDIAN__
vaddq_u8(uint8x16_t __p0,uint8x16_t __p1)1821 __ai uint8x16_t vaddq_u8(uint8x16_t __p0, uint8x16_t __p1) {
1822 uint8x16_t __ret;
1823 __ret = __p0 + __p1;
1824 return __ret;
1825 }
1826 #else
vaddq_u8(uint8x16_t __p0,uint8x16_t __p1)1827 __ai uint8x16_t vaddq_u8(uint8x16_t __p0, uint8x16_t __p1) {
1828 uint8x16_t __ret;
1829 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
1830 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
1831 __ret = __rev0 + __rev1;
1832 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
1833 return __ret;
1834 }
1835 #endif
1836
1837 #ifdef __LITTLE_ENDIAN__
vaddq_u32(uint32x4_t __p0,uint32x4_t __p1)1838 __ai uint32x4_t vaddq_u32(uint32x4_t __p0, uint32x4_t __p1) {
1839 uint32x4_t __ret;
1840 __ret = __p0 + __p1;
1841 return __ret;
1842 }
1843 #else
vaddq_u32(uint32x4_t __p0,uint32x4_t __p1)1844 __ai uint32x4_t vaddq_u32(uint32x4_t __p0, uint32x4_t __p1) {
1845 uint32x4_t __ret;
1846 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
1847 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
1848 __ret = __rev0 + __rev1;
1849 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
1850 return __ret;
1851 }
1852 #endif
1853
1854 #ifdef __LITTLE_ENDIAN__
vaddq_u64(uint64x2_t __p0,uint64x2_t __p1)1855 __ai uint64x2_t vaddq_u64(uint64x2_t __p0, uint64x2_t __p1) {
1856 uint64x2_t __ret;
1857 __ret = __p0 + __p1;
1858 return __ret;
1859 }
1860 #else
vaddq_u64(uint64x2_t __p0,uint64x2_t __p1)1861 __ai uint64x2_t vaddq_u64(uint64x2_t __p0, uint64x2_t __p1) {
1862 uint64x2_t __ret;
1863 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
1864 uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
1865 __ret = __rev0 + __rev1;
1866 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
1867 return __ret;
1868 }
1869 #endif
1870
1871 #ifdef __LITTLE_ENDIAN__
vaddq_u16(uint16x8_t __p0,uint16x8_t __p1)1872 __ai uint16x8_t vaddq_u16(uint16x8_t __p0, uint16x8_t __p1) {
1873 uint16x8_t __ret;
1874 __ret = __p0 + __p1;
1875 return __ret;
1876 }
1877 #else
vaddq_u16(uint16x8_t __p0,uint16x8_t __p1)1878 __ai uint16x8_t vaddq_u16(uint16x8_t __p0, uint16x8_t __p1) {
1879 uint16x8_t __ret;
1880 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
1881 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
1882 __ret = __rev0 + __rev1;
1883 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
1884 return __ret;
1885 }
1886 #endif
1887
1888 #ifdef __LITTLE_ENDIAN__
vaddq_s8(int8x16_t __p0,int8x16_t __p1)1889 __ai int8x16_t vaddq_s8(int8x16_t __p0, int8x16_t __p1) {
1890 int8x16_t __ret;
1891 __ret = __p0 + __p1;
1892 return __ret;
1893 }
1894 #else
vaddq_s8(int8x16_t __p0,int8x16_t __p1)1895 __ai int8x16_t vaddq_s8(int8x16_t __p0, int8x16_t __p1) {
1896 int8x16_t __ret;
1897 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
1898 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
1899 __ret = __rev0 + __rev1;
1900 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
1901 return __ret;
1902 }
1903 #endif
1904
1905 #ifdef __LITTLE_ENDIAN__
vaddq_f32(float32x4_t __p0,float32x4_t __p1)1906 __ai float32x4_t vaddq_f32(float32x4_t __p0, float32x4_t __p1) {
1907 float32x4_t __ret;
1908 __ret = __p0 + __p1;
1909 return __ret;
1910 }
1911 #else
vaddq_f32(float32x4_t __p0,float32x4_t __p1)1912 __ai float32x4_t vaddq_f32(float32x4_t __p0, float32x4_t __p1) {
1913 float32x4_t __ret;
1914 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
1915 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
1916 __ret = __rev0 + __rev1;
1917 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
1918 return __ret;
1919 }
1920 #endif
1921
1922 #ifdef __LITTLE_ENDIAN__
vaddq_s32(int32x4_t __p0,int32x4_t __p1)1923 __ai int32x4_t vaddq_s32(int32x4_t __p0, int32x4_t __p1) {
1924 int32x4_t __ret;
1925 __ret = __p0 + __p1;
1926 return __ret;
1927 }
1928 #else
vaddq_s32(int32x4_t __p0,int32x4_t __p1)1929 __ai int32x4_t vaddq_s32(int32x4_t __p0, int32x4_t __p1) {
1930 int32x4_t __ret;
1931 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
1932 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
1933 __ret = __rev0 + __rev1;
1934 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
1935 return __ret;
1936 }
1937 #endif
1938
1939 #ifdef __LITTLE_ENDIAN__
vaddq_s64(int64x2_t __p0,int64x2_t __p1)1940 __ai int64x2_t vaddq_s64(int64x2_t __p0, int64x2_t __p1) {
1941 int64x2_t __ret;
1942 __ret = __p0 + __p1;
1943 return __ret;
1944 }
1945 #else
vaddq_s64(int64x2_t __p0,int64x2_t __p1)1946 __ai int64x2_t vaddq_s64(int64x2_t __p0, int64x2_t __p1) {
1947 int64x2_t __ret;
1948 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
1949 int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
1950 __ret = __rev0 + __rev1;
1951 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
1952 return __ret;
1953 }
1954 #endif
1955
1956 #ifdef __LITTLE_ENDIAN__
vaddq_s16(int16x8_t __p0,int16x8_t __p1)1957 __ai int16x8_t vaddq_s16(int16x8_t __p0, int16x8_t __p1) {
1958 int16x8_t __ret;
1959 __ret = __p0 + __p1;
1960 return __ret;
1961 }
1962 #else
vaddq_s16(int16x8_t __p0,int16x8_t __p1)1963 __ai int16x8_t vaddq_s16(int16x8_t __p0, int16x8_t __p1) {
1964 int16x8_t __ret;
1965 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
1966 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
1967 __ret = __rev0 + __rev1;
1968 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
1969 return __ret;
1970 }
1971 #endif
1972
1973 #ifdef __LITTLE_ENDIAN__
vadd_u8(uint8x8_t __p0,uint8x8_t __p1)1974 __ai uint8x8_t vadd_u8(uint8x8_t __p0, uint8x8_t __p1) {
1975 uint8x8_t __ret;
1976 __ret = __p0 + __p1;
1977 return __ret;
1978 }
1979 #else
vadd_u8(uint8x8_t __p0,uint8x8_t __p1)1980 __ai uint8x8_t vadd_u8(uint8x8_t __p0, uint8x8_t __p1) {
1981 uint8x8_t __ret;
1982 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
1983 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
1984 __ret = __rev0 + __rev1;
1985 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
1986 return __ret;
1987 }
1988 #endif
1989
1990 #ifdef __LITTLE_ENDIAN__
vadd_u32(uint32x2_t __p0,uint32x2_t __p1)1991 __ai uint32x2_t vadd_u32(uint32x2_t __p0, uint32x2_t __p1) {
1992 uint32x2_t __ret;
1993 __ret = __p0 + __p1;
1994 return __ret;
1995 }
1996 #else
vadd_u32(uint32x2_t __p0,uint32x2_t __p1)1997 __ai uint32x2_t vadd_u32(uint32x2_t __p0, uint32x2_t __p1) {
1998 uint32x2_t __ret;
1999 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
2000 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
2001 __ret = __rev0 + __rev1;
2002 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
2003 return __ret;
2004 }
2005 #endif
2006
vadd_u64(uint64x1_t __p0,uint64x1_t __p1)2007 __ai uint64x1_t vadd_u64(uint64x1_t __p0, uint64x1_t __p1) {
2008 uint64x1_t __ret;
2009 __ret = __p0 + __p1;
2010 return __ret;
2011 }
2012 #ifdef __LITTLE_ENDIAN__
vadd_u16(uint16x4_t __p0,uint16x4_t __p1)2013 __ai uint16x4_t vadd_u16(uint16x4_t __p0, uint16x4_t __p1) {
2014 uint16x4_t __ret;
2015 __ret = __p0 + __p1;
2016 return __ret;
2017 }
2018 #else
vadd_u16(uint16x4_t __p0,uint16x4_t __p1)2019 __ai uint16x4_t vadd_u16(uint16x4_t __p0, uint16x4_t __p1) {
2020 uint16x4_t __ret;
2021 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2022 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
2023 __ret = __rev0 + __rev1;
2024 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
2025 return __ret;
2026 }
2027 #endif
2028
2029 #ifdef __LITTLE_ENDIAN__
vadd_s8(int8x8_t __p0,int8x8_t __p1)2030 __ai int8x8_t vadd_s8(int8x8_t __p0, int8x8_t __p1) {
2031 int8x8_t __ret;
2032 __ret = __p0 + __p1;
2033 return __ret;
2034 }
2035 #else
vadd_s8(int8x8_t __p0,int8x8_t __p1)2036 __ai int8x8_t vadd_s8(int8x8_t __p0, int8x8_t __p1) {
2037 int8x8_t __ret;
2038 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2039 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
2040 __ret = __rev0 + __rev1;
2041 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
2042 return __ret;
2043 }
2044 #endif
2045
2046 #ifdef __LITTLE_ENDIAN__
vadd_f32(float32x2_t __p0,float32x2_t __p1)2047 __ai float32x2_t vadd_f32(float32x2_t __p0, float32x2_t __p1) {
2048 float32x2_t __ret;
2049 __ret = __p0 + __p1;
2050 return __ret;
2051 }
2052 #else
vadd_f32(float32x2_t __p0,float32x2_t __p1)2053 __ai float32x2_t vadd_f32(float32x2_t __p0, float32x2_t __p1) {
2054 float32x2_t __ret;
2055 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
2056 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
2057 __ret = __rev0 + __rev1;
2058 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
2059 return __ret;
2060 }
2061 #endif
2062
2063 #ifdef __LITTLE_ENDIAN__
vadd_s32(int32x2_t __p0,int32x2_t __p1)2064 __ai int32x2_t vadd_s32(int32x2_t __p0, int32x2_t __p1) {
2065 int32x2_t __ret;
2066 __ret = __p0 + __p1;
2067 return __ret;
2068 }
2069 #else
vadd_s32(int32x2_t __p0,int32x2_t __p1)2070 __ai int32x2_t vadd_s32(int32x2_t __p0, int32x2_t __p1) {
2071 int32x2_t __ret;
2072 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
2073 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
2074 __ret = __rev0 + __rev1;
2075 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
2076 return __ret;
2077 }
2078 #endif
2079
vadd_s64(int64x1_t __p0,int64x1_t __p1)2080 __ai int64x1_t vadd_s64(int64x1_t __p0, int64x1_t __p1) {
2081 int64x1_t __ret;
2082 __ret = __p0 + __p1;
2083 return __ret;
2084 }
2085 #ifdef __LITTLE_ENDIAN__
vadd_s16(int16x4_t __p0,int16x4_t __p1)2086 __ai int16x4_t vadd_s16(int16x4_t __p0, int16x4_t __p1) {
2087 int16x4_t __ret;
2088 __ret = __p0 + __p1;
2089 return __ret;
2090 }
2091 #else
vadd_s16(int16x4_t __p0,int16x4_t __p1)2092 __ai int16x4_t vadd_s16(int16x4_t __p0, int16x4_t __p1) {
2093 int16x4_t __ret;
2094 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2095 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
2096 __ret = __rev0 + __rev1;
2097 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
2098 return __ret;
2099 }
2100 #endif
2101
2102 #ifdef __LITTLE_ENDIAN__
vadd_p8(poly8x8_t __p0,poly8x8_t __p1)2103 __ai poly8x8_t vadd_p8(poly8x8_t __p0, poly8x8_t __p1) {
2104 poly8x8_t __ret;
2105 __ret = (poly8x8_t) __builtin_neon_vadd_v((int8x8_t)__p0, (int8x8_t)__p1, 4);
2106 return __ret;
2107 }
2108 #else
vadd_p8(poly8x8_t __p0,poly8x8_t __p1)2109 __ai poly8x8_t vadd_p8(poly8x8_t __p0, poly8x8_t __p1) {
2110 poly8x8_t __ret;
2111 poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2112 poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
2113 __ret = (poly8x8_t) __builtin_neon_vadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 4);
2114 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
2115 return __ret;
2116 }
2117 #endif
2118
vadd_p64(poly64x1_t __p0,poly64x1_t __p1)2119 __ai poly64x1_t vadd_p64(poly64x1_t __p0, poly64x1_t __p1) {
2120 poly64x1_t __ret;
2121 __ret = (poly64x1_t) __builtin_neon_vadd_v((int8x8_t)__p0, (int8x8_t)__p1, 6);
2122 return __ret;
2123 }
2124 #ifdef __LITTLE_ENDIAN__
vadd_p16(poly16x4_t __p0,poly16x4_t __p1)2125 __ai poly16x4_t vadd_p16(poly16x4_t __p0, poly16x4_t __p1) {
2126 poly16x4_t __ret;
2127 __ret = (poly16x4_t) __builtin_neon_vadd_v((int8x8_t)__p0, (int8x8_t)__p1, 5);
2128 return __ret;
2129 }
2130 #else
vadd_p16(poly16x4_t __p0,poly16x4_t __p1)2131 __ai poly16x4_t vadd_p16(poly16x4_t __p0, poly16x4_t __p1) {
2132 poly16x4_t __ret;
2133 poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2134 poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
2135 __ret = (poly16x4_t) __builtin_neon_vadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 5);
2136 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
2137 return __ret;
2138 }
2139 #endif
2140
2141 #ifdef __LITTLE_ENDIAN__
vaddq_p8(poly8x16_t __p0,poly8x16_t __p1)2142 __ai poly8x16_t vaddq_p8(poly8x16_t __p0, poly8x16_t __p1) {
2143 poly8x16_t __ret;
2144 __ret = (poly8x16_t) __builtin_neon_vaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 36);
2145 return __ret;
2146 }
2147 #else
vaddq_p8(poly8x16_t __p0,poly8x16_t __p1)2148 __ai poly8x16_t vaddq_p8(poly8x16_t __p0, poly8x16_t __p1) {
2149 poly8x16_t __ret;
2150 poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
2151 poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
2152 __ret = (poly8x16_t) __builtin_neon_vaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 36);
2153 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
2154 return __ret;
2155 }
2156 #endif
2157
2158 #ifdef __LITTLE_ENDIAN__
vaddq_p64(poly64x2_t __p0,poly64x2_t __p1)2159 __ai poly64x2_t vaddq_p64(poly64x2_t __p0, poly64x2_t __p1) {
2160 poly64x2_t __ret;
2161 __ret = (poly64x2_t) __builtin_neon_vaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 38);
2162 return __ret;
2163 }
2164 #else
vaddq_p64(poly64x2_t __p0,poly64x2_t __p1)2165 __ai poly64x2_t vaddq_p64(poly64x2_t __p0, poly64x2_t __p1) {
2166 poly64x2_t __ret;
2167 poly64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
2168 poly64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
2169 __ret = (poly64x2_t) __builtin_neon_vaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 38);
2170 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
2171 return __ret;
2172 }
2173 #endif
2174
2175 #ifdef __LITTLE_ENDIAN__
vaddq_p16(poly16x8_t __p0,poly16x8_t __p1)2176 __ai poly16x8_t vaddq_p16(poly16x8_t __p0, poly16x8_t __p1) {
2177 poly16x8_t __ret;
2178 __ret = (poly16x8_t) __builtin_neon_vaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 37);
2179 return __ret;
2180 }
2181 #else
vaddq_p16(poly16x8_t __p0,poly16x8_t __p1)2182 __ai poly16x8_t vaddq_p16(poly16x8_t __p0, poly16x8_t __p1) {
2183 poly16x8_t __ret;
2184 poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2185 poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
2186 __ret = (poly16x8_t) __builtin_neon_vaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 37);
2187 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
2188 return __ret;
2189 }
2190 #endif
2191
2192 #ifdef __LITTLE_ENDIAN__
vaddhn_u32(uint32x4_t __p0,uint32x4_t __p1)2193 __ai uint16x4_t vaddhn_u32(uint32x4_t __p0, uint32x4_t __p1) {
2194 uint16x4_t __ret;
2195 __ret = (uint16x4_t) __builtin_neon_vaddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 17);
2196 return __ret;
2197 }
2198 #else
vaddhn_u32(uint32x4_t __p0,uint32x4_t __p1)2199 __ai uint16x4_t vaddhn_u32(uint32x4_t __p0, uint32x4_t __p1) {
2200 uint16x4_t __ret;
2201 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2202 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
2203 __ret = (uint16x4_t) __builtin_neon_vaddhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 17);
2204 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
2205 return __ret;
2206 }
__noswap_vaddhn_u32(uint32x4_t __p0,uint32x4_t __p1)2207 __ai uint16x4_t __noswap_vaddhn_u32(uint32x4_t __p0, uint32x4_t __p1) {
2208 uint16x4_t __ret;
2209 __ret = (uint16x4_t) __builtin_neon_vaddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 17);
2210 return __ret;
2211 }
2212 #endif
2213
2214 #ifdef __LITTLE_ENDIAN__
vaddhn_u64(uint64x2_t __p0,uint64x2_t __p1)2215 __ai uint32x2_t vaddhn_u64(uint64x2_t __p0, uint64x2_t __p1) {
2216 uint32x2_t __ret;
2217 __ret = (uint32x2_t) __builtin_neon_vaddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 18);
2218 return __ret;
2219 }
2220 #else
vaddhn_u64(uint64x2_t __p0,uint64x2_t __p1)2221 __ai uint32x2_t vaddhn_u64(uint64x2_t __p0, uint64x2_t __p1) {
2222 uint32x2_t __ret;
2223 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
2224 uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
2225 __ret = (uint32x2_t) __builtin_neon_vaddhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 18);
2226 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
2227 return __ret;
2228 }
__noswap_vaddhn_u64(uint64x2_t __p0,uint64x2_t __p1)2229 __ai uint32x2_t __noswap_vaddhn_u64(uint64x2_t __p0, uint64x2_t __p1) {
2230 uint32x2_t __ret;
2231 __ret = (uint32x2_t) __builtin_neon_vaddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 18);
2232 return __ret;
2233 }
2234 #endif
2235
2236 #ifdef __LITTLE_ENDIAN__
vaddhn_u16(uint16x8_t __p0,uint16x8_t __p1)2237 __ai uint8x8_t vaddhn_u16(uint16x8_t __p0, uint16x8_t __p1) {
2238 uint8x8_t __ret;
2239 __ret = (uint8x8_t) __builtin_neon_vaddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 16);
2240 return __ret;
2241 }
2242 #else
vaddhn_u16(uint16x8_t __p0,uint16x8_t __p1)2243 __ai uint8x8_t vaddhn_u16(uint16x8_t __p0, uint16x8_t __p1) {
2244 uint8x8_t __ret;
2245 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2246 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
2247 __ret = (uint8x8_t) __builtin_neon_vaddhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 16);
2248 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
2249 return __ret;
2250 }
__noswap_vaddhn_u16(uint16x8_t __p0,uint16x8_t __p1)2251 __ai uint8x8_t __noswap_vaddhn_u16(uint16x8_t __p0, uint16x8_t __p1) {
2252 uint8x8_t __ret;
2253 __ret = (uint8x8_t) __builtin_neon_vaddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 16);
2254 return __ret;
2255 }
2256 #endif
2257
2258 #ifdef __LITTLE_ENDIAN__
vaddhn_s32(int32x4_t __p0,int32x4_t __p1)2259 __ai int16x4_t vaddhn_s32(int32x4_t __p0, int32x4_t __p1) {
2260 int16x4_t __ret;
2261 __ret = (int16x4_t) __builtin_neon_vaddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 1);
2262 return __ret;
2263 }
2264 #else
vaddhn_s32(int32x4_t __p0,int32x4_t __p1)2265 __ai int16x4_t vaddhn_s32(int32x4_t __p0, int32x4_t __p1) {
2266 int16x4_t __ret;
2267 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2268 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
2269 __ret = (int16x4_t) __builtin_neon_vaddhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 1);
2270 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
2271 return __ret;
2272 }
__noswap_vaddhn_s32(int32x4_t __p0,int32x4_t __p1)2273 __ai int16x4_t __noswap_vaddhn_s32(int32x4_t __p0, int32x4_t __p1) {
2274 int16x4_t __ret;
2275 __ret = (int16x4_t) __builtin_neon_vaddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 1);
2276 return __ret;
2277 }
2278 #endif
2279
2280 #ifdef __LITTLE_ENDIAN__
vaddhn_s64(int64x2_t __p0,int64x2_t __p1)2281 __ai int32x2_t vaddhn_s64(int64x2_t __p0, int64x2_t __p1) {
2282 int32x2_t __ret;
2283 __ret = (int32x2_t) __builtin_neon_vaddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 2);
2284 return __ret;
2285 }
2286 #else
vaddhn_s64(int64x2_t __p0,int64x2_t __p1)2287 __ai int32x2_t vaddhn_s64(int64x2_t __p0, int64x2_t __p1) {
2288 int32x2_t __ret;
2289 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
2290 int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
2291 __ret = (int32x2_t) __builtin_neon_vaddhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 2);
2292 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
2293 return __ret;
2294 }
__noswap_vaddhn_s64(int64x2_t __p0,int64x2_t __p1)2295 __ai int32x2_t __noswap_vaddhn_s64(int64x2_t __p0, int64x2_t __p1) {
2296 int32x2_t __ret;
2297 __ret = (int32x2_t) __builtin_neon_vaddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 2);
2298 return __ret;
2299 }
2300 #endif
2301
2302 #ifdef __LITTLE_ENDIAN__
vaddhn_s16(int16x8_t __p0,int16x8_t __p1)2303 __ai int8x8_t vaddhn_s16(int16x8_t __p0, int16x8_t __p1) {
2304 int8x8_t __ret;
2305 __ret = (int8x8_t) __builtin_neon_vaddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 0);
2306 return __ret;
2307 }
2308 #else
vaddhn_s16(int16x8_t __p0,int16x8_t __p1)2309 __ai int8x8_t vaddhn_s16(int16x8_t __p0, int16x8_t __p1) {
2310 int8x8_t __ret;
2311 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2312 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
2313 __ret = (int8x8_t) __builtin_neon_vaddhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 0);
2314 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
2315 return __ret;
2316 }
__noswap_vaddhn_s16(int16x8_t __p0,int16x8_t __p1)2317 __ai int8x8_t __noswap_vaddhn_s16(int16x8_t __p0, int16x8_t __p1) {
2318 int8x8_t __ret;
2319 __ret = (int8x8_t) __builtin_neon_vaddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 0);
2320 return __ret;
2321 }
2322 #endif
2323
2324 #ifdef __LITTLE_ENDIAN__
vandq_u8(uint8x16_t __p0,uint8x16_t __p1)2325 __ai uint8x16_t vandq_u8(uint8x16_t __p0, uint8x16_t __p1) {
2326 uint8x16_t __ret;
2327 __ret = __p0 & __p1;
2328 return __ret;
2329 }
2330 #else
vandq_u8(uint8x16_t __p0,uint8x16_t __p1)2331 __ai uint8x16_t vandq_u8(uint8x16_t __p0, uint8x16_t __p1) {
2332 uint8x16_t __ret;
2333 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
2334 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
2335 __ret = __rev0 & __rev1;
2336 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
2337 return __ret;
2338 }
2339 #endif
2340
2341 #ifdef __LITTLE_ENDIAN__
vandq_u32(uint32x4_t __p0,uint32x4_t __p1)2342 __ai uint32x4_t vandq_u32(uint32x4_t __p0, uint32x4_t __p1) {
2343 uint32x4_t __ret;
2344 __ret = __p0 & __p1;
2345 return __ret;
2346 }
2347 #else
vandq_u32(uint32x4_t __p0,uint32x4_t __p1)2348 __ai uint32x4_t vandq_u32(uint32x4_t __p0, uint32x4_t __p1) {
2349 uint32x4_t __ret;
2350 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2351 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
2352 __ret = __rev0 & __rev1;
2353 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
2354 return __ret;
2355 }
2356 #endif
2357
2358 #ifdef __LITTLE_ENDIAN__
vandq_u64(uint64x2_t __p0,uint64x2_t __p1)2359 __ai uint64x2_t vandq_u64(uint64x2_t __p0, uint64x2_t __p1) {
2360 uint64x2_t __ret;
2361 __ret = __p0 & __p1;
2362 return __ret;
2363 }
2364 #else
vandq_u64(uint64x2_t __p0,uint64x2_t __p1)2365 __ai uint64x2_t vandq_u64(uint64x2_t __p0, uint64x2_t __p1) {
2366 uint64x2_t __ret;
2367 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
2368 uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
2369 __ret = __rev0 & __rev1;
2370 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
2371 return __ret;
2372 }
2373 #endif
2374
2375 #ifdef __LITTLE_ENDIAN__
vandq_u16(uint16x8_t __p0,uint16x8_t __p1)2376 __ai uint16x8_t vandq_u16(uint16x8_t __p0, uint16x8_t __p1) {
2377 uint16x8_t __ret;
2378 __ret = __p0 & __p1;
2379 return __ret;
2380 }
2381 #else
vandq_u16(uint16x8_t __p0,uint16x8_t __p1)2382 __ai uint16x8_t vandq_u16(uint16x8_t __p0, uint16x8_t __p1) {
2383 uint16x8_t __ret;
2384 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2385 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
2386 __ret = __rev0 & __rev1;
2387 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
2388 return __ret;
2389 }
2390 #endif
2391
2392 #ifdef __LITTLE_ENDIAN__
vandq_s8(int8x16_t __p0,int8x16_t __p1)2393 __ai int8x16_t vandq_s8(int8x16_t __p0, int8x16_t __p1) {
2394 int8x16_t __ret;
2395 __ret = __p0 & __p1;
2396 return __ret;
2397 }
2398 #else
vandq_s8(int8x16_t __p0,int8x16_t __p1)2399 __ai int8x16_t vandq_s8(int8x16_t __p0, int8x16_t __p1) {
2400 int8x16_t __ret;
2401 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
2402 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
2403 __ret = __rev0 & __rev1;
2404 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
2405 return __ret;
2406 }
2407 #endif
2408
2409 #ifdef __LITTLE_ENDIAN__
vandq_s32(int32x4_t __p0,int32x4_t __p1)2410 __ai int32x4_t vandq_s32(int32x4_t __p0, int32x4_t __p1) {
2411 int32x4_t __ret;
2412 __ret = __p0 & __p1;
2413 return __ret;
2414 }
2415 #else
vandq_s32(int32x4_t __p0,int32x4_t __p1)2416 __ai int32x4_t vandq_s32(int32x4_t __p0, int32x4_t __p1) {
2417 int32x4_t __ret;
2418 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2419 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
2420 __ret = __rev0 & __rev1;
2421 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
2422 return __ret;
2423 }
2424 #endif
2425
2426 #ifdef __LITTLE_ENDIAN__
vandq_s64(int64x2_t __p0,int64x2_t __p1)2427 __ai int64x2_t vandq_s64(int64x2_t __p0, int64x2_t __p1) {
2428 int64x2_t __ret;
2429 __ret = __p0 & __p1;
2430 return __ret;
2431 }
2432 #else
vandq_s64(int64x2_t __p0,int64x2_t __p1)2433 __ai int64x2_t vandq_s64(int64x2_t __p0, int64x2_t __p1) {
2434 int64x2_t __ret;
2435 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
2436 int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
2437 __ret = __rev0 & __rev1;
2438 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
2439 return __ret;
2440 }
2441 #endif
2442
2443 #ifdef __LITTLE_ENDIAN__
vandq_s16(int16x8_t __p0,int16x8_t __p1)2444 __ai int16x8_t vandq_s16(int16x8_t __p0, int16x8_t __p1) {
2445 int16x8_t __ret;
2446 __ret = __p0 & __p1;
2447 return __ret;
2448 }
2449 #else
vandq_s16(int16x8_t __p0,int16x8_t __p1)2450 __ai int16x8_t vandq_s16(int16x8_t __p0, int16x8_t __p1) {
2451 int16x8_t __ret;
2452 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2453 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
2454 __ret = __rev0 & __rev1;
2455 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
2456 return __ret;
2457 }
2458 #endif
2459
2460 #ifdef __LITTLE_ENDIAN__
vand_u8(uint8x8_t __p0,uint8x8_t __p1)2461 __ai uint8x8_t vand_u8(uint8x8_t __p0, uint8x8_t __p1) {
2462 uint8x8_t __ret;
2463 __ret = __p0 & __p1;
2464 return __ret;
2465 }
2466 #else
vand_u8(uint8x8_t __p0,uint8x8_t __p1)2467 __ai uint8x8_t vand_u8(uint8x8_t __p0, uint8x8_t __p1) {
2468 uint8x8_t __ret;
2469 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2470 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
2471 __ret = __rev0 & __rev1;
2472 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
2473 return __ret;
2474 }
2475 #endif
2476
2477 #ifdef __LITTLE_ENDIAN__
vand_u32(uint32x2_t __p0,uint32x2_t __p1)2478 __ai uint32x2_t vand_u32(uint32x2_t __p0, uint32x2_t __p1) {
2479 uint32x2_t __ret;
2480 __ret = __p0 & __p1;
2481 return __ret;
2482 }
2483 #else
vand_u32(uint32x2_t __p0,uint32x2_t __p1)2484 __ai uint32x2_t vand_u32(uint32x2_t __p0, uint32x2_t __p1) {
2485 uint32x2_t __ret;
2486 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
2487 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
2488 __ret = __rev0 & __rev1;
2489 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
2490 return __ret;
2491 }
2492 #endif
2493
vand_u64(uint64x1_t __p0,uint64x1_t __p1)2494 __ai uint64x1_t vand_u64(uint64x1_t __p0, uint64x1_t __p1) {
2495 uint64x1_t __ret;
2496 __ret = __p0 & __p1;
2497 return __ret;
2498 }
2499 #ifdef __LITTLE_ENDIAN__
vand_u16(uint16x4_t __p0,uint16x4_t __p1)2500 __ai uint16x4_t vand_u16(uint16x4_t __p0, uint16x4_t __p1) {
2501 uint16x4_t __ret;
2502 __ret = __p0 & __p1;
2503 return __ret;
2504 }
2505 #else
vand_u16(uint16x4_t __p0,uint16x4_t __p1)2506 __ai uint16x4_t vand_u16(uint16x4_t __p0, uint16x4_t __p1) {
2507 uint16x4_t __ret;
2508 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2509 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
2510 __ret = __rev0 & __rev1;
2511 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
2512 return __ret;
2513 }
2514 #endif
2515
2516 #ifdef __LITTLE_ENDIAN__
vand_s8(int8x8_t __p0,int8x8_t __p1)2517 __ai int8x8_t vand_s8(int8x8_t __p0, int8x8_t __p1) {
2518 int8x8_t __ret;
2519 __ret = __p0 & __p1;
2520 return __ret;
2521 }
2522 #else
vand_s8(int8x8_t __p0,int8x8_t __p1)2523 __ai int8x8_t vand_s8(int8x8_t __p0, int8x8_t __p1) {
2524 int8x8_t __ret;
2525 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2526 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
2527 __ret = __rev0 & __rev1;
2528 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
2529 return __ret;
2530 }
2531 #endif
2532
2533 #ifdef __LITTLE_ENDIAN__
vand_s32(int32x2_t __p0,int32x2_t __p1)2534 __ai int32x2_t vand_s32(int32x2_t __p0, int32x2_t __p1) {
2535 int32x2_t __ret;
2536 __ret = __p0 & __p1;
2537 return __ret;
2538 }
2539 #else
vand_s32(int32x2_t __p0,int32x2_t __p1)2540 __ai int32x2_t vand_s32(int32x2_t __p0, int32x2_t __p1) {
2541 int32x2_t __ret;
2542 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
2543 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
2544 __ret = __rev0 & __rev1;
2545 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
2546 return __ret;
2547 }
2548 #endif
2549
vand_s64(int64x1_t __p0,int64x1_t __p1)2550 __ai int64x1_t vand_s64(int64x1_t __p0, int64x1_t __p1) {
2551 int64x1_t __ret;
2552 __ret = __p0 & __p1;
2553 return __ret;
2554 }
2555 #ifdef __LITTLE_ENDIAN__
vand_s16(int16x4_t __p0,int16x4_t __p1)2556 __ai int16x4_t vand_s16(int16x4_t __p0, int16x4_t __p1) {
2557 int16x4_t __ret;
2558 __ret = __p0 & __p1;
2559 return __ret;
2560 }
2561 #else
vand_s16(int16x4_t __p0,int16x4_t __p1)2562 __ai int16x4_t vand_s16(int16x4_t __p0, int16x4_t __p1) {
2563 int16x4_t __ret;
2564 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2565 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
2566 __ret = __rev0 & __rev1;
2567 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
2568 return __ret;
2569 }
2570 #endif
2571
2572 #ifdef __LITTLE_ENDIAN__
vbicq_u8(uint8x16_t __p0,uint8x16_t __p1)2573 __ai uint8x16_t vbicq_u8(uint8x16_t __p0, uint8x16_t __p1) {
2574 uint8x16_t __ret;
2575 __ret = __p0 & ~__p1;
2576 return __ret;
2577 }
2578 #else
vbicq_u8(uint8x16_t __p0,uint8x16_t __p1)2579 __ai uint8x16_t vbicq_u8(uint8x16_t __p0, uint8x16_t __p1) {
2580 uint8x16_t __ret;
2581 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
2582 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
2583 __ret = __rev0 & ~__rev1;
2584 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
2585 return __ret;
2586 }
2587 #endif
2588
2589 #ifdef __LITTLE_ENDIAN__
vbicq_u32(uint32x4_t __p0,uint32x4_t __p1)2590 __ai uint32x4_t vbicq_u32(uint32x4_t __p0, uint32x4_t __p1) {
2591 uint32x4_t __ret;
2592 __ret = __p0 & ~__p1;
2593 return __ret;
2594 }
2595 #else
vbicq_u32(uint32x4_t __p0,uint32x4_t __p1)2596 __ai uint32x4_t vbicq_u32(uint32x4_t __p0, uint32x4_t __p1) {
2597 uint32x4_t __ret;
2598 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2599 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
2600 __ret = __rev0 & ~__rev1;
2601 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
2602 return __ret;
2603 }
2604 #endif
2605
2606 #ifdef __LITTLE_ENDIAN__
vbicq_u64(uint64x2_t __p0,uint64x2_t __p1)2607 __ai uint64x2_t vbicq_u64(uint64x2_t __p0, uint64x2_t __p1) {
2608 uint64x2_t __ret;
2609 __ret = __p0 & ~__p1;
2610 return __ret;
2611 }
2612 #else
vbicq_u64(uint64x2_t __p0,uint64x2_t __p1)2613 __ai uint64x2_t vbicq_u64(uint64x2_t __p0, uint64x2_t __p1) {
2614 uint64x2_t __ret;
2615 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
2616 uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
2617 __ret = __rev0 & ~__rev1;
2618 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
2619 return __ret;
2620 }
2621 #endif
2622
2623 #ifdef __LITTLE_ENDIAN__
vbicq_u16(uint16x8_t __p0,uint16x8_t __p1)2624 __ai uint16x8_t vbicq_u16(uint16x8_t __p0, uint16x8_t __p1) {
2625 uint16x8_t __ret;
2626 __ret = __p0 & ~__p1;
2627 return __ret;
2628 }
2629 #else
vbicq_u16(uint16x8_t __p0,uint16x8_t __p1)2630 __ai uint16x8_t vbicq_u16(uint16x8_t __p0, uint16x8_t __p1) {
2631 uint16x8_t __ret;
2632 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2633 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
2634 __ret = __rev0 & ~__rev1;
2635 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
2636 return __ret;
2637 }
2638 #endif
2639
2640 #ifdef __LITTLE_ENDIAN__
vbicq_s8(int8x16_t __p0,int8x16_t __p1)2641 __ai int8x16_t vbicq_s8(int8x16_t __p0, int8x16_t __p1) {
2642 int8x16_t __ret;
2643 __ret = __p0 & ~__p1;
2644 return __ret;
2645 }
2646 #else
vbicq_s8(int8x16_t __p0,int8x16_t __p1)2647 __ai int8x16_t vbicq_s8(int8x16_t __p0, int8x16_t __p1) {
2648 int8x16_t __ret;
2649 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
2650 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
2651 __ret = __rev0 & ~__rev1;
2652 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
2653 return __ret;
2654 }
2655 #endif
2656
2657 #ifdef __LITTLE_ENDIAN__
vbicq_s32(int32x4_t __p0,int32x4_t __p1)2658 __ai int32x4_t vbicq_s32(int32x4_t __p0, int32x4_t __p1) {
2659 int32x4_t __ret;
2660 __ret = __p0 & ~__p1;
2661 return __ret;
2662 }
2663 #else
vbicq_s32(int32x4_t __p0,int32x4_t __p1)2664 __ai int32x4_t vbicq_s32(int32x4_t __p0, int32x4_t __p1) {
2665 int32x4_t __ret;
2666 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2667 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
2668 __ret = __rev0 & ~__rev1;
2669 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
2670 return __ret;
2671 }
2672 #endif
2673
2674 #ifdef __LITTLE_ENDIAN__
vbicq_s64(int64x2_t __p0,int64x2_t __p1)2675 __ai int64x2_t vbicq_s64(int64x2_t __p0, int64x2_t __p1) {
2676 int64x2_t __ret;
2677 __ret = __p0 & ~__p1;
2678 return __ret;
2679 }
2680 #else
vbicq_s64(int64x2_t __p0,int64x2_t __p1)2681 __ai int64x2_t vbicq_s64(int64x2_t __p0, int64x2_t __p1) {
2682 int64x2_t __ret;
2683 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
2684 int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
2685 __ret = __rev0 & ~__rev1;
2686 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
2687 return __ret;
2688 }
2689 #endif
2690
2691 #ifdef __LITTLE_ENDIAN__
vbicq_s16(int16x8_t __p0,int16x8_t __p1)2692 __ai int16x8_t vbicq_s16(int16x8_t __p0, int16x8_t __p1) {
2693 int16x8_t __ret;
2694 __ret = __p0 & ~__p1;
2695 return __ret;
2696 }
2697 #else
vbicq_s16(int16x8_t __p0,int16x8_t __p1)2698 __ai int16x8_t vbicq_s16(int16x8_t __p0, int16x8_t __p1) {
2699 int16x8_t __ret;
2700 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2701 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
2702 __ret = __rev0 & ~__rev1;
2703 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
2704 return __ret;
2705 }
2706 #endif
2707
2708 #ifdef __LITTLE_ENDIAN__
vbic_u8(uint8x8_t __p0,uint8x8_t __p1)2709 __ai uint8x8_t vbic_u8(uint8x8_t __p0, uint8x8_t __p1) {
2710 uint8x8_t __ret;
2711 __ret = __p0 & ~__p1;
2712 return __ret;
2713 }
2714 #else
vbic_u8(uint8x8_t __p0,uint8x8_t __p1)2715 __ai uint8x8_t vbic_u8(uint8x8_t __p0, uint8x8_t __p1) {
2716 uint8x8_t __ret;
2717 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2718 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
2719 __ret = __rev0 & ~__rev1;
2720 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
2721 return __ret;
2722 }
2723 #endif
2724
2725 #ifdef __LITTLE_ENDIAN__
vbic_u32(uint32x2_t __p0,uint32x2_t __p1)2726 __ai uint32x2_t vbic_u32(uint32x2_t __p0, uint32x2_t __p1) {
2727 uint32x2_t __ret;
2728 __ret = __p0 & ~__p1;
2729 return __ret;
2730 }
2731 #else
vbic_u32(uint32x2_t __p0,uint32x2_t __p1)2732 __ai uint32x2_t vbic_u32(uint32x2_t __p0, uint32x2_t __p1) {
2733 uint32x2_t __ret;
2734 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
2735 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
2736 __ret = __rev0 & ~__rev1;
2737 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
2738 return __ret;
2739 }
2740 #endif
2741
vbic_u64(uint64x1_t __p0,uint64x1_t __p1)2742 __ai uint64x1_t vbic_u64(uint64x1_t __p0, uint64x1_t __p1) {
2743 uint64x1_t __ret;
2744 __ret = __p0 & ~__p1;
2745 return __ret;
2746 }
2747 #ifdef __LITTLE_ENDIAN__
vbic_u16(uint16x4_t __p0,uint16x4_t __p1)2748 __ai uint16x4_t vbic_u16(uint16x4_t __p0, uint16x4_t __p1) {
2749 uint16x4_t __ret;
2750 __ret = __p0 & ~__p1;
2751 return __ret;
2752 }
2753 #else
vbic_u16(uint16x4_t __p0,uint16x4_t __p1)2754 __ai uint16x4_t vbic_u16(uint16x4_t __p0, uint16x4_t __p1) {
2755 uint16x4_t __ret;
2756 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2757 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
2758 __ret = __rev0 & ~__rev1;
2759 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
2760 return __ret;
2761 }
2762 #endif
2763
2764 #ifdef __LITTLE_ENDIAN__
vbic_s8(int8x8_t __p0,int8x8_t __p1)2765 __ai int8x8_t vbic_s8(int8x8_t __p0, int8x8_t __p1) {
2766 int8x8_t __ret;
2767 __ret = __p0 & ~__p1;
2768 return __ret;
2769 }
2770 #else
vbic_s8(int8x8_t __p0,int8x8_t __p1)2771 __ai int8x8_t vbic_s8(int8x8_t __p0, int8x8_t __p1) {
2772 int8x8_t __ret;
2773 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2774 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
2775 __ret = __rev0 & ~__rev1;
2776 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
2777 return __ret;
2778 }
2779 #endif
2780
2781 #ifdef __LITTLE_ENDIAN__
vbic_s32(int32x2_t __p0,int32x2_t __p1)2782 __ai int32x2_t vbic_s32(int32x2_t __p0, int32x2_t __p1) {
2783 int32x2_t __ret;
2784 __ret = __p0 & ~__p1;
2785 return __ret;
2786 }
2787 #else
vbic_s32(int32x2_t __p0,int32x2_t __p1)2788 __ai int32x2_t vbic_s32(int32x2_t __p0, int32x2_t __p1) {
2789 int32x2_t __ret;
2790 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
2791 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
2792 __ret = __rev0 & ~__rev1;
2793 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
2794 return __ret;
2795 }
2796 #endif
2797
vbic_s64(int64x1_t __p0,int64x1_t __p1)2798 __ai int64x1_t vbic_s64(int64x1_t __p0, int64x1_t __p1) {
2799 int64x1_t __ret;
2800 __ret = __p0 & ~__p1;
2801 return __ret;
2802 }
2803 #ifdef __LITTLE_ENDIAN__
vbic_s16(int16x4_t __p0,int16x4_t __p1)2804 __ai int16x4_t vbic_s16(int16x4_t __p0, int16x4_t __p1) {
2805 int16x4_t __ret;
2806 __ret = __p0 & ~__p1;
2807 return __ret;
2808 }
2809 #else
vbic_s16(int16x4_t __p0,int16x4_t __p1)2810 __ai int16x4_t vbic_s16(int16x4_t __p0, int16x4_t __p1) {
2811 int16x4_t __ret;
2812 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2813 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
2814 __ret = __rev0 & ~__rev1;
2815 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
2816 return __ret;
2817 }
2818 #endif
2819
2820 #ifdef __LITTLE_ENDIAN__
vbsl_p8(uint8x8_t __p0,poly8x8_t __p1,poly8x8_t __p2)2821 __ai poly8x8_t vbsl_p8(uint8x8_t __p0, poly8x8_t __p1, poly8x8_t __p2) {
2822 poly8x8_t __ret;
2823 __ret = (poly8x8_t) __builtin_neon_vbsl_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 4);
2824 return __ret;
2825 }
2826 #else
vbsl_p8(uint8x8_t __p0,poly8x8_t __p1,poly8x8_t __p2)2827 __ai poly8x8_t vbsl_p8(uint8x8_t __p0, poly8x8_t __p1, poly8x8_t __p2) {
2828 poly8x8_t __ret;
2829 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2830 poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
2831 poly8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
2832 __ret = (poly8x8_t) __builtin_neon_vbsl_v((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 4);
2833 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
2834 return __ret;
2835 }
2836 #endif
2837
2838 #ifdef __LITTLE_ENDIAN__
vbsl_p16(uint16x4_t __p0,poly16x4_t __p1,poly16x4_t __p2)2839 __ai poly16x4_t vbsl_p16(uint16x4_t __p0, poly16x4_t __p1, poly16x4_t __p2) {
2840 poly16x4_t __ret;
2841 __ret = (poly16x4_t) __builtin_neon_vbsl_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 5);
2842 return __ret;
2843 }
2844 #else
vbsl_p16(uint16x4_t __p0,poly16x4_t __p1,poly16x4_t __p2)2845 __ai poly16x4_t vbsl_p16(uint16x4_t __p0, poly16x4_t __p1, poly16x4_t __p2) {
2846 poly16x4_t __ret;
2847 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2848 poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
2849 poly16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
2850 __ret = (poly16x4_t) __builtin_neon_vbsl_v((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 5);
2851 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
2852 return __ret;
2853 }
2854 #endif
2855
2856 #ifdef __LITTLE_ENDIAN__
vbslq_p8(uint8x16_t __p0,poly8x16_t __p1,poly8x16_t __p2)2857 __ai poly8x16_t vbslq_p8(uint8x16_t __p0, poly8x16_t __p1, poly8x16_t __p2) {
2858 poly8x16_t __ret;
2859 __ret = (poly8x16_t) __builtin_neon_vbslq_v((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 36);
2860 return __ret;
2861 }
2862 #else
vbslq_p8(uint8x16_t __p0,poly8x16_t __p1,poly8x16_t __p2)2863 __ai poly8x16_t vbslq_p8(uint8x16_t __p0, poly8x16_t __p1, poly8x16_t __p2) {
2864 poly8x16_t __ret;
2865 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
2866 poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
2867 poly8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
2868 __ret = (poly8x16_t) __builtin_neon_vbslq_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 36);
2869 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
2870 return __ret;
2871 }
2872 #endif
2873
2874 #ifdef __LITTLE_ENDIAN__
vbslq_p16(uint16x8_t __p0,poly16x8_t __p1,poly16x8_t __p2)2875 __ai poly16x8_t vbslq_p16(uint16x8_t __p0, poly16x8_t __p1, poly16x8_t __p2) {
2876 poly16x8_t __ret;
2877 __ret = (poly16x8_t) __builtin_neon_vbslq_v((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 37);
2878 return __ret;
2879 }
2880 #else
vbslq_p16(uint16x8_t __p0,poly16x8_t __p1,poly16x8_t __p2)2881 __ai poly16x8_t vbslq_p16(uint16x8_t __p0, poly16x8_t __p1, poly16x8_t __p2) {
2882 poly16x8_t __ret;
2883 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2884 poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
2885 poly16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
2886 __ret = (poly16x8_t) __builtin_neon_vbslq_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 37);
2887 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
2888 return __ret;
2889 }
2890 #endif
2891
2892 #ifdef __LITTLE_ENDIAN__
vbslq_u8(uint8x16_t __p0,uint8x16_t __p1,uint8x16_t __p2)2893 __ai uint8x16_t vbslq_u8(uint8x16_t __p0, uint8x16_t __p1, uint8x16_t __p2) {
2894 uint8x16_t __ret;
2895 __ret = (uint8x16_t) __builtin_neon_vbslq_v((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 48);
2896 return __ret;
2897 }
2898 #else
vbslq_u8(uint8x16_t __p0,uint8x16_t __p1,uint8x16_t __p2)2899 __ai uint8x16_t vbslq_u8(uint8x16_t __p0, uint8x16_t __p1, uint8x16_t __p2) {
2900 uint8x16_t __ret;
2901 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
2902 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
2903 uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
2904 __ret = (uint8x16_t) __builtin_neon_vbslq_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 48);
2905 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
2906 return __ret;
2907 }
2908 #endif
2909
2910 #ifdef __LITTLE_ENDIAN__
vbslq_u32(uint32x4_t __p0,uint32x4_t __p1,uint32x4_t __p2)2911 __ai uint32x4_t vbslq_u32(uint32x4_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
2912 uint32x4_t __ret;
2913 __ret = (uint32x4_t) __builtin_neon_vbslq_v((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 50);
2914 return __ret;
2915 }
2916 #else
vbslq_u32(uint32x4_t __p0,uint32x4_t __p1,uint32x4_t __p2)2917 __ai uint32x4_t vbslq_u32(uint32x4_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
2918 uint32x4_t __ret;
2919 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2920 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
2921 uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
2922 __ret = (uint32x4_t) __builtin_neon_vbslq_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 50);
2923 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
2924 return __ret;
2925 }
2926 #endif
2927
2928 #ifdef __LITTLE_ENDIAN__
vbslq_u64(uint64x2_t __p0,uint64x2_t __p1,uint64x2_t __p2)2929 __ai uint64x2_t vbslq_u64(uint64x2_t __p0, uint64x2_t __p1, uint64x2_t __p2) {
2930 uint64x2_t __ret;
2931 __ret = (uint64x2_t) __builtin_neon_vbslq_v((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 51);
2932 return __ret;
2933 }
2934 #else
vbslq_u64(uint64x2_t __p0,uint64x2_t __p1,uint64x2_t __p2)2935 __ai uint64x2_t vbslq_u64(uint64x2_t __p0, uint64x2_t __p1, uint64x2_t __p2) {
2936 uint64x2_t __ret;
2937 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
2938 uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
2939 uint64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
2940 __ret = (uint64x2_t) __builtin_neon_vbslq_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 51);
2941 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
2942 return __ret;
2943 }
2944 #endif
2945
2946 #ifdef __LITTLE_ENDIAN__
vbslq_u16(uint16x8_t __p0,uint16x8_t __p1,uint16x8_t __p2)2947 __ai uint16x8_t vbslq_u16(uint16x8_t __p0, uint16x8_t __p1, uint16x8_t __p2) {
2948 uint16x8_t __ret;
2949 __ret = (uint16x8_t) __builtin_neon_vbslq_v((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 49);
2950 return __ret;
2951 }
2952 #else
vbslq_u16(uint16x8_t __p0,uint16x8_t __p1,uint16x8_t __p2)2953 __ai uint16x8_t vbslq_u16(uint16x8_t __p0, uint16x8_t __p1, uint16x8_t __p2) {
2954 uint16x8_t __ret;
2955 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
2956 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
2957 uint16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
2958 __ret = (uint16x8_t) __builtin_neon_vbslq_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 49);
2959 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
2960 return __ret;
2961 }
2962 #endif
2963
2964 #ifdef __LITTLE_ENDIAN__
vbslq_s8(uint8x16_t __p0,int8x16_t __p1,int8x16_t __p2)2965 __ai int8x16_t vbslq_s8(uint8x16_t __p0, int8x16_t __p1, int8x16_t __p2) {
2966 int8x16_t __ret;
2967 __ret = (int8x16_t) __builtin_neon_vbslq_v((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 32);
2968 return __ret;
2969 }
2970 #else
vbslq_s8(uint8x16_t __p0,int8x16_t __p1,int8x16_t __p2)2971 __ai int8x16_t vbslq_s8(uint8x16_t __p0, int8x16_t __p1, int8x16_t __p2) {
2972 int8x16_t __ret;
2973 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
2974 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
2975 int8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
2976 __ret = (int8x16_t) __builtin_neon_vbslq_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 32);
2977 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
2978 return __ret;
2979 }
2980 #endif
2981
2982 #ifdef __LITTLE_ENDIAN__
vbslq_f32(uint32x4_t __p0,float32x4_t __p1,float32x4_t __p2)2983 __ai float32x4_t vbslq_f32(uint32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
2984 float32x4_t __ret;
2985 __ret = (float32x4_t) __builtin_neon_vbslq_v((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
2986 return __ret;
2987 }
2988 #else
vbslq_f32(uint32x4_t __p0,float32x4_t __p1,float32x4_t __p2)2989 __ai float32x4_t vbslq_f32(uint32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
2990 float32x4_t __ret;
2991 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
2992 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
2993 float32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
2994 __ret = (float32x4_t) __builtin_neon_vbslq_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 41);
2995 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
2996 return __ret;
2997 }
2998 #endif
2999
3000 #ifdef __LITTLE_ENDIAN__
vbslq_s32(uint32x4_t __p0,int32x4_t __p1,int32x4_t __p2)3001 __ai int32x4_t vbslq_s32(uint32x4_t __p0, int32x4_t __p1, int32x4_t __p2) {
3002 int32x4_t __ret;
3003 __ret = (int32x4_t) __builtin_neon_vbslq_v((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 34);
3004 return __ret;
3005 }
3006 #else
vbslq_s32(uint32x4_t __p0,int32x4_t __p1,int32x4_t __p2)3007 __ai int32x4_t vbslq_s32(uint32x4_t __p0, int32x4_t __p1, int32x4_t __p2) {
3008 int32x4_t __ret;
3009 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
3010 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
3011 int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
3012 __ret = (int32x4_t) __builtin_neon_vbslq_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 34);
3013 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
3014 return __ret;
3015 }
3016 #endif
3017
3018 #ifdef __LITTLE_ENDIAN__
vbslq_s64(uint64x2_t __p0,int64x2_t __p1,int64x2_t __p2)3019 __ai int64x2_t vbslq_s64(uint64x2_t __p0, int64x2_t __p1, int64x2_t __p2) {
3020 int64x2_t __ret;
3021 __ret = (int64x2_t) __builtin_neon_vbslq_v((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 35);
3022 return __ret;
3023 }
3024 #else
vbslq_s64(uint64x2_t __p0,int64x2_t __p1,int64x2_t __p2)3025 __ai int64x2_t vbslq_s64(uint64x2_t __p0, int64x2_t __p1, int64x2_t __p2) {
3026 int64x2_t __ret;
3027 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
3028 int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
3029 int64x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
3030 __ret = (int64x2_t) __builtin_neon_vbslq_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 35);
3031 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
3032 return __ret;
3033 }
3034 #endif
3035
3036 #ifdef __LITTLE_ENDIAN__
vbslq_s16(uint16x8_t __p0,int16x8_t __p1,int16x8_t __p2)3037 __ai int16x8_t vbslq_s16(uint16x8_t __p0, int16x8_t __p1, int16x8_t __p2) {
3038 int16x8_t __ret;
3039 __ret = (int16x8_t) __builtin_neon_vbslq_v((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 33);
3040 return __ret;
3041 }
3042 #else
vbslq_s16(uint16x8_t __p0,int16x8_t __p1,int16x8_t __p2)3043 __ai int16x8_t vbslq_s16(uint16x8_t __p0, int16x8_t __p1, int16x8_t __p2) {
3044 int16x8_t __ret;
3045 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
3046 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
3047 int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
3048 __ret = (int16x8_t) __builtin_neon_vbslq_v((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 33);
3049 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
3050 return __ret;
3051 }
3052 #endif
3053
3054 #ifdef __LITTLE_ENDIAN__
vbsl_u8(uint8x8_t __p0,uint8x8_t __p1,uint8x8_t __p2)3055 __ai uint8x8_t vbsl_u8(uint8x8_t __p0, uint8x8_t __p1, uint8x8_t __p2) {
3056 uint8x8_t __ret;
3057 __ret = (uint8x8_t) __builtin_neon_vbsl_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 16);
3058 return __ret;
3059 }
3060 #else
vbsl_u8(uint8x8_t __p0,uint8x8_t __p1,uint8x8_t __p2)3061 __ai uint8x8_t vbsl_u8(uint8x8_t __p0, uint8x8_t __p1, uint8x8_t __p2) {
3062 uint8x8_t __ret;
3063 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
3064 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
3065 uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
3066 __ret = (uint8x8_t) __builtin_neon_vbsl_v((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 16);
3067 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
3068 return __ret;
3069 }
3070 #endif
3071
3072 #ifdef __LITTLE_ENDIAN__
vbsl_u32(uint32x2_t __p0,uint32x2_t __p1,uint32x2_t __p2)3073 __ai uint32x2_t vbsl_u32(uint32x2_t __p0, uint32x2_t __p1, uint32x2_t __p2) {
3074 uint32x2_t __ret;
3075 __ret = (uint32x2_t) __builtin_neon_vbsl_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 18);
3076 return __ret;
3077 }
3078 #else
vbsl_u32(uint32x2_t __p0,uint32x2_t __p1,uint32x2_t __p2)3079 __ai uint32x2_t vbsl_u32(uint32x2_t __p0, uint32x2_t __p1, uint32x2_t __p2) {
3080 uint32x2_t __ret;
3081 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
3082 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
3083 uint32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
3084 __ret = (uint32x2_t) __builtin_neon_vbsl_v((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 18);
3085 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
3086 return __ret;
3087 }
3088 #endif
3089
vbsl_u64(uint64x1_t __p0,uint64x1_t __p1,uint64x1_t __p2)3090 __ai uint64x1_t vbsl_u64(uint64x1_t __p0, uint64x1_t __p1, uint64x1_t __p2) {
3091 uint64x1_t __ret;
3092 __ret = (uint64x1_t) __builtin_neon_vbsl_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 19);
3093 return __ret;
3094 }
3095 #ifdef __LITTLE_ENDIAN__
vbsl_u16(uint16x4_t __p0,uint16x4_t __p1,uint16x4_t __p2)3096 __ai uint16x4_t vbsl_u16(uint16x4_t __p0, uint16x4_t __p1, uint16x4_t __p2) {
3097 uint16x4_t __ret;
3098 __ret = (uint16x4_t) __builtin_neon_vbsl_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 17);
3099 return __ret;
3100 }
3101 #else
vbsl_u16(uint16x4_t __p0,uint16x4_t __p1,uint16x4_t __p2)3102 __ai uint16x4_t vbsl_u16(uint16x4_t __p0, uint16x4_t __p1, uint16x4_t __p2) {
3103 uint16x4_t __ret;
3104 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
3105 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
3106 uint16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
3107 __ret = (uint16x4_t) __builtin_neon_vbsl_v((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 17);
3108 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
3109 return __ret;
3110 }
3111 #endif
3112
3113 #ifdef __LITTLE_ENDIAN__
vbsl_s8(uint8x8_t __p0,int8x8_t __p1,int8x8_t __p2)3114 __ai int8x8_t vbsl_s8(uint8x8_t __p0, int8x8_t __p1, int8x8_t __p2) {
3115 int8x8_t __ret;
3116 __ret = (int8x8_t) __builtin_neon_vbsl_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 0);
3117 return __ret;
3118 }
3119 #else
vbsl_s8(uint8x8_t __p0,int8x8_t __p1,int8x8_t __p2)3120 __ai int8x8_t vbsl_s8(uint8x8_t __p0, int8x8_t __p1, int8x8_t __p2) {
3121 int8x8_t __ret;
3122 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
3123 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
3124 int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
3125 __ret = (int8x8_t) __builtin_neon_vbsl_v((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 0);
3126 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
3127 return __ret;
3128 }
3129 #endif
3130
3131 #ifdef __LITTLE_ENDIAN__
vbsl_f32(uint32x2_t __p0,float32x2_t __p1,float32x2_t __p2)3132 __ai float32x2_t vbsl_f32(uint32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
3133 float32x2_t __ret;
3134 __ret = (float32x2_t) __builtin_neon_vbsl_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 9);
3135 return __ret;
3136 }
3137 #else
vbsl_f32(uint32x2_t __p0,float32x2_t __p1,float32x2_t __p2)3138 __ai float32x2_t vbsl_f32(uint32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
3139 float32x2_t __ret;
3140 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
3141 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
3142 float32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
3143 __ret = (float32x2_t) __builtin_neon_vbsl_v((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 9);
3144 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
3145 return __ret;
3146 }
3147 #endif
3148
3149 #ifdef __LITTLE_ENDIAN__
vbsl_s32(uint32x2_t __p0,int32x2_t __p1,int32x2_t __p2)3150 __ai int32x2_t vbsl_s32(uint32x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
3151 int32x2_t __ret;
3152 __ret = (int32x2_t) __builtin_neon_vbsl_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 2);
3153 return __ret;
3154 }
3155 #else
vbsl_s32(uint32x2_t __p0,int32x2_t __p1,int32x2_t __p2)3156 __ai int32x2_t vbsl_s32(uint32x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
3157 int32x2_t __ret;
3158 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
3159 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
3160 int32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
3161 __ret = (int32x2_t) __builtin_neon_vbsl_v((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 2);
3162 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
3163 return __ret;
3164 }
3165 #endif
3166
vbsl_s64(uint64x1_t __p0,int64x1_t __p1,int64x1_t __p2)3167 __ai int64x1_t vbsl_s64(uint64x1_t __p0, int64x1_t __p1, int64x1_t __p2) {
3168 int64x1_t __ret;
3169 __ret = (int64x1_t) __builtin_neon_vbsl_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 3);
3170 return __ret;
3171 }
3172 #ifdef __LITTLE_ENDIAN__
vbsl_s16(uint16x4_t __p0,int16x4_t __p1,int16x4_t __p2)3173 __ai int16x4_t vbsl_s16(uint16x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
3174 int16x4_t __ret;
3175 __ret = (int16x4_t) __builtin_neon_vbsl_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 1);
3176 return __ret;
3177 }
3178 #else
vbsl_s16(uint16x4_t __p0,int16x4_t __p1,int16x4_t __p2)3179 __ai int16x4_t vbsl_s16(uint16x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
3180 int16x4_t __ret;
3181 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
3182 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
3183 int16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
3184 __ret = (int16x4_t) __builtin_neon_vbsl_v((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 1);
3185 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
3186 return __ret;
3187 }
3188 #endif
3189
3190 #ifdef __LITTLE_ENDIAN__
vcageq_f32(float32x4_t __p0,float32x4_t __p1)3191 __ai uint32x4_t vcageq_f32(float32x4_t __p0, float32x4_t __p1) {
3192 uint32x4_t __ret;
3193 __ret = (uint32x4_t) __builtin_neon_vcageq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
3194 return __ret;
3195 }
3196 #else
vcageq_f32(float32x4_t __p0,float32x4_t __p1)3197 __ai uint32x4_t vcageq_f32(float32x4_t __p0, float32x4_t __p1) {
3198 uint32x4_t __ret;
3199 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
3200 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
3201 __ret = (uint32x4_t) __builtin_neon_vcageq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
3202 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
3203 return __ret;
3204 }
3205 #endif
3206
3207 #ifdef __LITTLE_ENDIAN__
vcage_f32(float32x2_t __p0,float32x2_t __p1)3208 __ai uint32x2_t vcage_f32(float32x2_t __p0, float32x2_t __p1) {
3209 uint32x2_t __ret;
3210 __ret = (uint32x2_t) __builtin_neon_vcage_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
3211 return __ret;
3212 }
3213 #else
vcage_f32(float32x2_t __p0,float32x2_t __p1)3214 __ai uint32x2_t vcage_f32(float32x2_t __p0, float32x2_t __p1) {
3215 uint32x2_t __ret;
3216 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
3217 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
3218 __ret = (uint32x2_t) __builtin_neon_vcage_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
3219 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
3220 return __ret;
3221 }
3222 #endif
3223
3224 #ifdef __LITTLE_ENDIAN__
vcagtq_f32(float32x4_t __p0,float32x4_t __p1)3225 __ai uint32x4_t vcagtq_f32(float32x4_t __p0, float32x4_t __p1) {
3226 uint32x4_t __ret;
3227 __ret = (uint32x4_t) __builtin_neon_vcagtq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
3228 return __ret;
3229 }
3230 #else
vcagtq_f32(float32x4_t __p0,float32x4_t __p1)3231 __ai uint32x4_t vcagtq_f32(float32x4_t __p0, float32x4_t __p1) {
3232 uint32x4_t __ret;
3233 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
3234 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
3235 __ret = (uint32x4_t) __builtin_neon_vcagtq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
3236 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
3237 return __ret;
3238 }
3239 #endif
3240
3241 #ifdef __LITTLE_ENDIAN__
vcagt_f32(float32x2_t __p0,float32x2_t __p1)3242 __ai uint32x2_t vcagt_f32(float32x2_t __p0, float32x2_t __p1) {
3243 uint32x2_t __ret;
3244 __ret = (uint32x2_t) __builtin_neon_vcagt_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
3245 return __ret;
3246 }
3247 #else
vcagt_f32(float32x2_t __p0,float32x2_t __p1)3248 __ai uint32x2_t vcagt_f32(float32x2_t __p0, float32x2_t __p1) {
3249 uint32x2_t __ret;
3250 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
3251 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
3252 __ret = (uint32x2_t) __builtin_neon_vcagt_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
3253 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
3254 return __ret;
3255 }
3256 #endif
3257
3258 #ifdef __LITTLE_ENDIAN__
vcaleq_f32(float32x4_t __p0,float32x4_t __p1)3259 __ai uint32x4_t vcaleq_f32(float32x4_t __p0, float32x4_t __p1) {
3260 uint32x4_t __ret;
3261 __ret = (uint32x4_t) __builtin_neon_vcaleq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
3262 return __ret;
3263 }
3264 #else
vcaleq_f32(float32x4_t __p0,float32x4_t __p1)3265 __ai uint32x4_t vcaleq_f32(float32x4_t __p0, float32x4_t __p1) {
3266 uint32x4_t __ret;
3267 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
3268 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
3269 __ret = (uint32x4_t) __builtin_neon_vcaleq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
3270 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
3271 return __ret;
3272 }
3273 #endif
3274
3275 #ifdef __LITTLE_ENDIAN__
vcale_f32(float32x2_t __p0,float32x2_t __p1)3276 __ai uint32x2_t vcale_f32(float32x2_t __p0, float32x2_t __p1) {
3277 uint32x2_t __ret;
3278 __ret = (uint32x2_t) __builtin_neon_vcale_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
3279 return __ret;
3280 }
3281 #else
vcale_f32(float32x2_t __p0,float32x2_t __p1)3282 __ai uint32x2_t vcale_f32(float32x2_t __p0, float32x2_t __p1) {
3283 uint32x2_t __ret;
3284 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
3285 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
3286 __ret = (uint32x2_t) __builtin_neon_vcale_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
3287 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
3288 return __ret;
3289 }
3290 #endif
3291
3292 #ifdef __LITTLE_ENDIAN__
vcaltq_f32(float32x4_t __p0,float32x4_t __p1)3293 __ai uint32x4_t vcaltq_f32(float32x4_t __p0, float32x4_t __p1) {
3294 uint32x4_t __ret;
3295 __ret = (uint32x4_t) __builtin_neon_vcaltq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
3296 return __ret;
3297 }
3298 #else
vcaltq_f32(float32x4_t __p0,float32x4_t __p1)3299 __ai uint32x4_t vcaltq_f32(float32x4_t __p0, float32x4_t __p1) {
3300 uint32x4_t __ret;
3301 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
3302 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
3303 __ret = (uint32x4_t) __builtin_neon_vcaltq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
3304 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
3305 return __ret;
3306 }
3307 #endif
3308
3309 #ifdef __LITTLE_ENDIAN__
vcalt_f32(float32x2_t __p0,float32x2_t __p1)3310 __ai uint32x2_t vcalt_f32(float32x2_t __p0, float32x2_t __p1) {
3311 uint32x2_t __ret;
3312 __ret = (uint32x2_t) __builtin_neon_vcalt_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
3313 return __ret;
3314 }
3315 #else
vcalt_f32(float32x2_t __p0,float32x2_t __p1)3316 __ai uint32x2_t vcalt_f32(float32x2_t __p0, float32x2_t __p1) {
3317 uint32x2_t __ret;
3318 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
3319 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
3320 __ret = (uint32x2_t) __builtin_neon_vcalt_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
3321 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
3322 return __ret;
3323 }
3324 #endif
3325
3326 #ifdef __LITTLE_ENDIAN__
vceq_p8(poly8x8_t __p0,poly8x8_t __p1)3327 __ai uint8x8_t vceq_p8(poly8x8_t __p0, poly8x8_t __p1) {
3328 uint8x8_t __ret;
3329 __ret = (uint8x8_t)(__p0 == __p1);
3330 return __ret;
3331 }
3332 #else
vceq_p8(poly8x8_t __p0,poly8x8_t __p1)3333 __ai uint8x8_t vceq_p8(poly8x8_t __p0, poly8x8_t __p1) {
3334 uint8x8_t __ret;
3335 poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
3336 poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
3337 __ret = (uint8x8_t)(__rev0 == __rev1);
3338 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
3339 return __ret;
3340 }
3341 #endif
3342
3343 #ifdef __LITTLE_ENDIAN__
vceqq_p8(poly8x16_t __p0,poly8x16_t __p1)3344 __ai uint8x16_t vceqq_p8(poly8x16_t __p0, poly8x16_t __p1) {
3345 uint8x16_t __ret;
3346 __ret = (uint8x16_t)(__p0 == __p1);
3347 return __ret;
3348 }
3349 #else
vceqq_p8(poly8x16_t __p0,poly8x16_t __p1)3350 __ai uint8x16_t vceqq_p8(poly8x16_t __p0, poly8x16_t __p1) {
3351 uint8x16_t __ret;
3352 poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
3353 poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
3354 __ret = (uint8x16_t)(__rev0 == __rev1);
3355 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
3356 return __ret;
3357 }
3358 #endif
3359
3360 #ifdef __LITTLE_ENDIAN__
vceqq_u8(uint8x16_t __p0,uint8x16_t __p1)3361 __ai uint8x16_t vceqq_u8(uint8x16_t __p0, uint8x16_t __p1) {
3362 uint8x16_t __ret;
3363 __ret = (uint8x16_t)(__p0 == __p1);
3364 return __ret;
3365 }
3366 #else
vceqq_u8(uint8x16_t __p0,uint8x16_t __p1)3367 __ai uint8x16_t vceqq_u8(uint8x16_t __p0, uint8x16_t __p1) {
3368 uint8x16_t __ret;
3369 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
3370 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
3371 __ret = (uint8x16_t)(__rev0 == __rev1);
3372 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
3373 return __ret;
3374 }
3375 #endif
3376
3377 #ifdef __LITTLE_ENDIAN__
vceqq_u32(uint32x4_t __p0,uint32x4_t __p1)3378 __ai uint32x4_t vceqq_u32(uint32x4_t __p0, uint32x4_t __p1) {
3379 uint32x4_t __ret;
3380 __ret = (uint32x4_t)(__p0 == __p1);
3381 return __ret;
3382 }
3383 #else
vceqq_u32(uint32x4_t __p0,uint32x4_t __p1)3384 __ai uint32x4_t vceqq_u32(uint32x4_t __p0, uint32x4_t __p1) {
3385 uint32x4_t __ret;
3386 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
3387 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
3388 __ret = (uint32x4_t)(__rev0 == __rev1);
3389 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
3390 return __ret;
3391 }
3392 #endif
3393
3394 #ifdef __LITTLE_ENDIAN__
vceqq_u16(uint16x8_t __p0,uint16x8_t __p1)3395 __ai uint16x8_t vceqq_u16(uint16x8_t __p0, uint16x8_t __p1) {
3396 uint16x8_t __ret;
3397 __ret = (uint16x8_t)(__p0 == __p1);
3398 return __ret;
3399 }
3400 #else
vceqq_u16(uint16x8_t __p0,uint16x8_t __p1)3401 __ai uint16x8_t vceqq_u16(uint16x8_t __p0, uint16x8_t __p1) {
3402 uint16x8_t __ret;
3403 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
3404 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
3405 __ret = (uint16x8_t)(__rev0 == __rev1);
3406 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
3407 return __ret;
3408 }
3409 #endif
3410
3411 #ifdef __LITTLE_ENDIAN__
vceqq_s8(int8x16_t __p0,int8x16_t __p1)3412 __ai uint8x16_t vceqq_s8(int8x16_t __p0, int8x16_t __p1) {
3413 uint8x16_t __ret;
3414 __ret = (uint8x16_t)(__p0 == __p1);
3415 return __ret;
3416 }
3417 #else
vceqq_s8(int8x16_t __p0,int8x16_t __p1)3418 __ai uint8x16_t vceqq_s8(int8x16_t __p0, int8x16_t __p1) {
3419 uint8x16_t __ret;
3420 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
3421 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
3422 __ret = (uint8x16_t)(__rev0 == __rev1);
3423 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
3424 return __ret;
3425 }
3426 #endif
3427
3428 #ifdef __LITTLE_ENDIAN__
vceqq_f32(float32x4_t __p0,float32x4_t __p1)3429 __ai uint32x4_t vceqq_f32(float32x4_t __p0, float32x4_t __p1) {
3430 uint32x4_t __ret;
3431 __ret = (uint32x4_t)(__p0 == __p1);
3432 return __ret;
3433 }
3434 #else
vceqq_f32(float32x4_t __p0,float32x4_t __p1)3435 __ai uint32x4_t vceqq_f32(float32x4_t __p0, float32x4_t __p1) {
3436 uint32x4_t __ret;
3437 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
3438 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
3439 __ret = (uint32x4_t)(__rev0 == __rev1);
3440 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
3441 return __ret;
3442 }
3443 #endif
3444
3445 #ifdef __LITTLE_ENDIAN__
vceqq_s32(int32x4_t __p0,int32x4_t __p1)3446 __ai uint32x4_t vceqq_s32(int32x4_t __p0, int32x4_t __p1) {
3447 uint32x4_t __ret;
3448 __ret = (uint32x4_t)(__p0 == __p1);
3449 return __ret;
3450 }
3451 #else
vceqq_s32(int32x4_t __p0,int32x4_t __p1)3452 __ai uint32x4_t vceqq_s32(int32x4_t __p0, int32x4_t __p1) {
3453 uint32x4_t __ret;
3454 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
3455 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
3456 __ret = (uint32x4_t)(__rev0 == __rev1);
3457 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
3458 return __ret;
3459 }
3460 #endif
3461
3462 #ifdef __LITTLE_ENDIAN__
vceqq_s16(int16x8_t __p0,int16x8_t __p1)3463 __ai uint16x8_t vceqq_s16(int16x8_t __p0, int16x8_t __p1) {
3464 uint16x8_t __ret;
3465 __ret = (uint16x8_t)(__p0 == __p1);
3466 return __ret;
3467 }
3468 #else
vceqq_s16(int16x8_t __p0,int16x8_t __p1)3469 __ai uint16x8_t vceqq_s16(int16x8_t __p0, int16x8_t __p1) {
3470 uint16x8_t __ret;
3471 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
3472 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
3473 __ret = (uint16x8_t)(__rev0 == __rev1);
3474 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
3475 return __ret;
3476 }
3477 #endif
3478
3479 #ifdef __LITTLE_ENDIAN__
vceq_u8(uint8x8_t __p0,uint8x8_t __p1)3480 __ai uint8x8_t vceq_u8(uint8x8_t __p0, uint8x8_t __p1) {
3481 uint8x8_t __ret;
3482 __ret = (uint8x8_t)(__p0 == __p1);
3483 return __ret;
3484 }
3485 #else
vceq_u8(uint8x8_t __p0,uint8x8_t __p1)3486 __ai uint8x8_t vceq_u8(uint8x8_t __p0, uint8x8_t __p1) {
3487 uint8x8_t __ret;
3488 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
3489 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
3490 __ret = (uint8x8_t)(__rev0 == __rev1);
3491 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
3492 return __ret;
3493 }
3494 #endif
3495
3496 #ifdef __LITTLE_ENDIAN__
vceq_u32(uint32x2_t __p0,uint32x2_t __p1)3497 __ai uint32x2_t vceq_u32(uint32x2_t __p0, uint32x2_t __p1) {
3498 uint32x2_t __ret;
3499 __ret = (uint32x2_t)(__p0 == __p1);
3500 return __ret;
3501 }
3502 #else
vceq_u32(uint32x2_t __p0,uint32x2_t __p1)3503 __ai uint32x2_t vceq_u32(uint32x2_t __p0, uint32x2_t __p1) {
3504 uint32x2_t __ret;
3505 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
3506 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
3507 __ret = (uint32x2_t)(__rev0 == __rev1);
3508 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
3509 return __ret;
3510 }
3511 #endif
3512
3513 #ifdef __LITTLE_ENDIAN__
vceq_u16(uint16x4_t __p0,uint16x4_t __p1)3514 __ai uint16x4_t vceq_u16(uint16x4_t __p0, uint16x4_t __p1) {
3515 uint16x4_t __ret;
3516 __ret = (uint16x4_t)(__p0 == __p1);
3517 return __ret;
3518 }
3519 #else
vceq_u16(uint16x4_t __p0,uint16x4_t __p1)3520 __ai uint16x4_t vceq_u16(uint16x4_t __p0, uint16x4_t __p1) {
3521 uint16x4_t __ret;
3522 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
3523 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
3524 __ret = (uint16x4_t)(__rev0 == __rev1);
3525 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
3526 return __ret;
3527 }
3528 #endif
3529
3530 #ifdef __LITTLE_ENDIAN__
vceq_s8(int8x8_t __p0,int8x8_t __p1)3531 __ai uint8x8_t vceq_s8(int8x8_t __p0, int8x8_t __p1) {
3532 uint8x8_t __ret;
3533 __ret = (uint8x8_t)(__p0 == __p1);
3534 return __ret;
3535 }
3536 #else
vceq_s8(int8x8_t __p0,int8x8_t __p1)3537 __ai uint8x8_t vceq_s8(int8x8_t __p0, int8x8_t __p1) {
3538 uint8x8_t __ret;
3539 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
3540 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
3541 __ret = (uint8x8_t)(__rev0 == __rev1);
3542 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
3543 return __ret;
3544 }
3545 #endif
3546
3547 #ifdef __LITTLE_ENDIAN__
vceq_f32(float32x2_t __p0,float32x2_t __p1)3548 __ai uint32x2_t vceq_f32(float32x2_t __p0, float32x2_t __p1) {
3549 uint32x2_t __ret;
3550 __ret = (uint32x2_t)(__p0 == __p1);
3551 return __ret;
3552 }
3553 #else
vceq_f32(float32x2_t __p0,float32x2_t __p1)3554 __ai uint32x2_t vceq_f32(float32x2_t __p0, float32x2_t __p1) {
3555 uint32x2_t __ret;
3556 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
3557 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
3558 __ret = (uint32x2_t)(__rev0 == __rev1);
3559 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
3560 return __ret;
3561 }
3562 #endif
3563
3564 #ifdef __LITTLE_ENDIAN__
vceq_s32(int32x2_t __p0,int32x2_t __p1)3565 __ai uint32x2_t vceq_s32(int32x2_t __p0, int32x2_t __p1) {
3566 uint32x2_t __ret;
3567 __ret = (uint32x2_t)(__p0 == __p1);
3568 return __ret;
3569 }
3570 #else
vceq_s32(int32x2_t __p0,int32x2_t __p1)3571 __ai uint32x2_t vceq_s32(int32x2_t __p0, int32x2_t __p1) {
3572 uint32x2_t __ret;
3573 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
3574 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
3575 __ret = (uint32x2_t)(__rev0 == __rev1);
3576 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
3577 return __ret;
3578 }
3579 #endif
3580
3581 #ifdef __LITTLE_ENDIAN__
vceq_s16(int16x4_t __p0,int16x4_t __p1)3582 __ai uint16x4_t vceq_s16(int16x4_t __p0, int16x4_t __p1) {
3583 uint16x4_t __ret;
3584 __ret = (uint16x4_t)(__p0 == __p1);
3585 return __ret;
3586 }
3587 #else
vceq_s16(int16x4_t __p0,int16x4_t __p1)3588 __ai uint16x4_t vceq_s16(int16x4_t __p0, int16x4_t __p1) {
3589 uint16x4_t __ret;
3590 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
3591 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
3592 __ret = (uint16x4_t)(__rev0 == __rev1);
3593 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
3594 return __ret;
3595 }
3596 #endif
3597
3598 #ifdef __LITTLE_ENDIAN__
vcgeq_u8(uint8x16_t __p0,uint8x16_t __p1)3599 __ai uint8x16_t vcgeq_u8(uint8x16_t __p0, uint8x16_t __p1) {
3600 uint8x16_t __ret;
3601 __ret = (uint8x16_t)(__p0 >= __p1);
3602 return __ret;
3603 }
3604 #else
vcgeq_u8(uint8x16_t __p0,uint8x16_t __p1)3605 __ai uint8x16_t vcgeq_u8(uint8x16_t __p0, uint8x16_t __p1) {
3606 uint8x16_t __ret;
3607 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
3608 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
3609 __ret = (uint8x16_t)(__rev0 >= __rev1);
3610 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
3611 return __ret;
3612 }
3613 #endif
3614
3615 #ifdef __LITTLE_ENDIAN__
vcgeq_u32(uint32x4_t __p0,uint32x4_t __p1)3616 __ai uint32x4_t vcgeq_u32(uint32x4_t __p0, uint32x4_t __p1) {
3617 uint32x4_t __ret;
3618 __ret = (uint32x4_t)(__p0 >= __p1);
3619 return __ret;
3620 }
3621 #else
vcgeq_u32(uint32x4_t __p0,uint32x4_t __p1)3622 __ai uint32x4_t vcgeq_u32(uint32x4_t __p0, uint32x4_t __p1) {
3623 uint32x4_t __ret;
3624 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
3625 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
3626 __ret = (uint32x4_t)(__rev0 >= __rev1);
3627 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
3628 return __ret;
3629 }
3630 #endif
3631
3632 #ifdef __LITTLE_ENDIAN__
vcgeq_u16(uint16x8_t __p0,uint16x8_t __p1)3633 __ai uint16x8_t vcgeq_u16(uint16x8_t __p0, uint16x8_t __p1) {
3634 uint16x8_t __ret;
3635 __ret = (uint16x8_t)(__p0 >= __p1);
3636 return __ret;
3637 }
3638 #else
vcgeq_u16(uint16x8_t __p0,uint16x8_t __p1)3639 __ai uint16x8_t vcgeq_u16(uint16x8_t __p0, uint16x8_t __p1) {
3640 uint16x8_t __ret;
3641 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
3642 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
3643 __ret = (uint16x8_t)(__rev0 >= __rev1);
3644 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
3645 return __ret;
3646 }
3647 #endif
3648
3649 #ifdef __LITTLE_ENDIAN__
vcgeq_s8(int8x16_t __p0,int8x16_t __p1)3650 __ai uint8x16_t vcgeq_s8(int8x16_t __p0, int8x16_t __p1) {
3651 uint8x16_t __ret;
3652 __ret = (uint8x16_t)(__p0 >= __p1);
3653 return __ret;
3654 }
3655 #else
vcgeq_s8(int8x16_t __p0,int8x16_t __p1)3656 __ai uint8x16_t vcgeq_s8(int8x16_t __p0, int8x16_t __p1) {
3657 uint8x16_t __ret;
3658 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
3659 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
3660 __ret = (uint8x16_t)(__rev0 >= __rev1);
3661 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
3662 return __ret;
3663 }
3664 #endif
3665
3666 #ifdef __LITTLE_ENDIAN__
vcgeq_f32(float32x4_t __p0,float32x4_t __p1)3667 __ai uint32x4_t vcgeq_f32(float32x4_t __p0, float32x4_t __p1) {
3668 uint32x4_t __ret;
3669 __ret = (uint32x4_t)(__p0 >= __p1);
3670 return __ret;
3671 }
3672 #else
vcgeq_f32(float32x4_t __p0,float32x4_t __p1)3673 __ai uint32x4_t vcgeq_f32(float32x4_t __p0, float32x4_t __p1) {
3674 uint32x4_t __ret;
3675 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
3676 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
3677 __ret = (uint32x4_t)(__rev0 >= __rev1);
3678 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
3679 return __ret;
3680 }
3681 #endif
3682
3683 #ifdef __LITTLE_ENDIAN__
vcgeq_s32(int32x4_t __p0,int32x4_t __p1)3684 __ai uint32x4_t vcgeq_s32(int32x4_t __p0, int32x4_t __p1) {
3685 uint32x4_t __ret;
3686 __ret = (uint32x4_t)(__p0 >= __p1);
3687 return __ret;
3688 }
3689 #else
vcgeq_s32(int32x4_t __p0,int32x4_t __p1)3690 __ai uint32x4_t vcgeq_s32(int32x4_t __p0, int32x4_t __p1) {
3691 uint32x4_t __ret;
3692 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
3693 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
3694 __ret = (uint32x4_t)(__rev0 >= __rev1);
3695 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
3696 return __ret;
3697 }
3698 #endif
3699
3700 #ifdef __LITTLE_ENDIAN__
vcgeq_s16(int16x8_t __p0,int16x8_t __p1)3701 __ai uint16x8_t vcgeq_s16(int16x8_t __p0, int16x8_t __p1) {
3702 uint16x8_t __ret;
3703 __ret = (uint16x8_t)(__p0 >= __p1);
3704 return __ret;
3705 }
3706 #else
vcgeq_s16(int16x8_t __p0,int16x8_t __p1)3707 __ai uint16x8_t vcgeq_s16(int16x8_t __p0, int16x8_t __p1) {
3708 uint16x8_t __ret;
3709 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
3710 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
3711 __ret = (uint16x8_t)(__rev0 >= __rev1);
3712 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
3713 return __ret;
3714 }
3715 #endif
3716
3717 #ifdef __LITTLE_ENDIAN__
vcge_u8(uint8x8_t __p0,uint8x8_t __p1)3718 __ai uint8x8_t vcge_u8(uint8x8_t __p0, uint8x8_t __p1) {
3719 uint8x8_t __ret;
3720 __ret = (uint8x8_t)(__p0 >= __p1);
3721 return __ret;
3722 }
3723 #else
vcge_u8(uint8x8_t __p0,uint8x8_t __p1)3724 __ai uint8x8_t vcge_u8(uint8x8_t __p0, uint8x8_t __p1) {
3725 uint8x8_t __ret;
3726 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
3727 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
3728 __ret = (uint8x8_t)(__rev0 >= __rev1);
3729 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
3730 return __ret;
3731 }
3732 #endif
3733
3734 #ifdef __LITTLE_ENDIAN__
vcge_u32(uint32x2_t __p0,uint32x2_t __p1)3735 __ai uint32x2_t vcge_u32(uint32x2_t __p0, uint32x2_t __p1) {
3736 uint32x2_t __ret;
3737 __ret = (uint32x2_t)(__p0 >= __p1);
3738 return __ret;
3739 }
3740 #else
vcge_u32(uint32x2_t __p0,uint32x2_t __p1)3741 __ai uint32x2_t vcge_u32(uint32x2_t __p0, uint32x2_t __p1) {
3742 uint32x2_t __ret;
3743 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
3744 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
3745 __ret = (uint32x2_t)(__rev0 >= __rev1);
3746 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
3747 return __ret;
3748 }
3749 #endif
3750
3751 #ifdef __LITTLE_ENDIAN__
vcge_u16(uint16x4_t __p0,uint16x4_t __p1)3752 __ai uint16x4_t vcge_u16(uint16x4_t __p0, uint16x4_t __p1) {
3753 uint16x4_t __ret;
3754 __ret = (uint16x4_t)(__p0 >= __p1);
3755 return __ret;
3756 }
3757 #else
vcge_u16(uint16x4_t __p0,uint16x4_t __p1)3758 __ai uint16x4_t vcge_u16(uint16x4_t __p0, uint16x4_t __p1) {
3759 uint16x4_t __ret;
3760 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
3761 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
3762 __ret = (uint16x4_t)(__rev0 >= __rev1);
3763 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
3764 return __ret;
3765 }
3766 #endif
3767
3768 #ifdef __LITTLE_ENDIAN__
vcge_s8(int8x8_t __p0,int8x8_t __p1)3769 __ai uint8x8_t vcge_s8(int8x8_t __p0, int8x8_t __p1) {
3770 uint8x8_t __ret;
3771 __ret = (uint8x8_t)(__p0 >= __p1);
3772 return __ret;
3773 }
3774 #else
vcge_s8(int8x8_t __p0,int8x8_t __p1)3775 __ai uint8x8_t vcge_s8(int8x8_t __p0, int8x8_t __p1) {
3776 uint8x8_t __ret;
3777 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
3778 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
3779 __ret = (uint8x8_t)(__rev0 >= __rev1);
3780 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
3781 return __ret;
3782 }
3783 #endif
3784
3785 #ifdef __LITTLE_ENDIAN__
vcge_f32(float32x2_t __p0,float32x2_t __p1)3786 __ai uint32x2_t vcge_f32(float32x2_t __p0, float32x2_t __p1) {
3787 uint32x2_t __ret;
3788 __ret = (uint32x2_t)(__p0 >= __p1);
3789 return __ret;
3790 }
3791 #else
vcge_f32(float32x2_t __p0,float32x2_t __p1)3792 __ai uint32x2_t vcge_f32(float32x2_t __p0, float32x2_t __p1) {
3793 uint32x2_t __ret;
3794 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
3795 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
3796 __ret = (uint32x2_t)(__rev0 >= __rev1);
3797 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
3798 return __ret;
3799 }
3800 #endif
3801
3802 #ifdef __LITTLE_ENDIAN__
vcge_s32(int32x2_t __p0,int32x2_t __p1)3803 __ai uint32x2_t vcge_s32(int32x2_t __p0, int32x2_t __p1) {
3804 uint32x2_t __ret;
3805 __ret = (uint32x2_t)(__p0 >= __p1);
3806 return __ret;
3807 }
3808 #else
vcge_s32(int32x2_t __p0,int32x2_t __p1)3809 __ai uint32x2_t vcge_s32(int32x2_t __p0, int32x2_t __p1) {
3810 uint32x2_t __ret;
3811 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
3812 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
3813 __ret = (uint32x2_t)(__rev0 >= __rev1);
3814 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
3815 return __ret;
3816 }
3817 #endif
3818
3819 #ifdef __LITTLE_ENDIAN__
vcge_s16(int16x4_t __p0,int16x4_t __p1)3820 __ai uint16x4_t vcge_s16(int16x4_t __p0, int16x4_t __p1) {
3821 uint16x4_t __ret;
3822 __ret = (uint16x4_t)(__p0 >= __p1);
3823 return __ret;
3824 }
3825 #else
vcge_s16(int16x4_t __p0,int16x4_t __p1)3826 __ai uint16x4_t vcge_s16(int16x4_t __p0, int16x4_t __p1) {
3827 uint16x4_t __ret;
3828 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
3829 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
3830 __ret = (uint16x4_t)(__rev0 >= __rev1);
3831 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
3832 return __ret;
3833 }
3834 #endif
3835
3836 #ifdef __LITTLE_ENDIAN__
vcgtq_u8(uint8x16_t __p0,uint8x16_t __p1)3837 __ai uint8x16_t vcgtq_u8(uint8x16_t __p0, uint8x16_t __p1) {
3838 uint8x16_t __ret;
3839 __ret = (uint8x16_t)(__p0 > __p1);
3840 return __ret;
3841 }
3842 #else
vcgtq_u8(uint8x16_t __p0,uint8x16_t __p1)3843 __ai uint8x16_t vcgtq_u8(uint8x16_t __p0, uint8x16_t __p1) {
3844 uint8x16_t __ret;
3845 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
3846 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
3847 __ret = (uint8x16_t)(__rev0 > __rev1);
3848 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
3849 return __ret;
3850 }
3851 #endif
3852
3853 #ifdef __LITTLE_ENDIAN__
vcgtq_u32(uint32x4_t __p0,uint32x4_t __p1)3854 __ai uint32x4_t vcgtq_u32(uint32x4_t __p0, uint32x4_t __p1) {
3855 uint32x4_t __ret;
3856 __ret = (uint32x4_t)(__p0 > __p1);
3857 return __ret;
3858 }
3859 #else
vcgtq_u32(uint32x4_t __p0,uint32x4_t __p1)3860 __ai uint32x4_t vcgtq_u32(uint32x4_t __p0, uint32x4_t __p1) {
3861 uint32x4_t __ret;
3862 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
3863 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
3864 __ret = (uint32x4_t)(__rev0 > __rev1);
3865 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
3866 return __ret;
3867 }
3868 #endif
3869
3870 #ifdef __LITTLE_ENDIAN__
vcgtq_u16(uint16x8_t __p0,uint16x8_t __p1)3871 __ai uint16x8_t vcgtq_u16(uint16x8_t __p0, uint16x8_t __p1) {
3872 uint16x8_t __ret;
3873 __ret = (uint16x8_t)(__p0 > __p1);
3874 return __ret;
3875 }
3876 #else
vcgtq_u16(uint16x8_t __p0,uint16x8_t __p1)3877 __ai uint16x8_t vcgtq_u16(uint16x8_t __p0, uint16x8_t __p1) {
3878 uint16x8_t __ret;
3879 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
3880 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
3881 __ret = (uint16x8_t)(__rev0 > __rev1);
3882 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
3883 return __ret;
3884 }
3885 #endif
3886
3887 #ifdef __LITTLE_ENDIAN__
vcgtq_s8(int8x16_t __p0,int8x16_t __p1)3888 __ai uint8x16_t vcgtq_s8(int8x16_t __p0, int8x16_t __p1) {
3889 uint8x16_t __ret;
3890 __ret = (uint8x16_t)(__p0 > __p1);
3891 return __ret;
3892 }
3893 #else
vcgtq_s8(int8x16_t __p0,int8x16_t __p1)3894 __ai uint8x16_t vcgtq_s8(int8x16_t __p0, int8x16_t __p1) {
3895 uint8x16_t __ret;
3896 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
3897 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
3898 __ret = (uint8x16_t)(__rev0 > __rev1);
3899 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
3900 return __ret;
3901 }
3902 #endif
3903
3904 #ifdef __LITTLE_ENDIAN__
vcgtq_f32(float32x4_t __p0,float32x4_t __p1)3905 __ai uint32x4_t vcgtq_f32(float32x4_t __p0, float32x4_t __p1) {
3906 uint32x4_t __ret;
3907 __ret = (uint32x4_t)(__p0 > __p1);
3908 return __ret;
3909 }
3910 #else
vcgtq_f32(float32x4_t __p0,float32x4_t __p1)3911 __ai uint32x4_t vcgtq_f32(float32x4_t __p0, float32x4_t __p1) {
3912 uint32x4_t __ret;
3913 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
3914 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
3915 __ret = (uint32x4_t)(__rev0 > __rev1);
3916 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
3917 return __ret;
3918 }
3919 #endif
3920
3921 #ifdef __LITTLE_ENDIAN__
vcgtq_s32(int32x4_t __p0,int32x4_t __p1)3922 __ai uint32x4_t vcgtq_s32(int32x4_t __p0, int32x4_t __p1) {
3923 uint32x4_t __ret;
3924 __ret = (uint32x4_t)(__p0 > __p1);
3925 return __ret;
3926 }
3927 #else
vcgtq_s32(int32x4_t __p0,int32x4_t __p1)3928 __ai uint32x4_t vcgtq_s32(int32x4_t __p0, int32x4_t __p1) {
3929 uint32x4_t __ret;
3930 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
3931 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
3932 __ret = (uint32x4_t)(__rev0 > __rev1);
3933 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
3934 return __ret;
3935 }
3936 #endif
3937
3938 #ifdef __LITTLE_ENDIAN__
vcgtq_s16(int16x8_t __p0,int16x8_t __p1)3939 __ai uint16x8_t vcgtq_s16(int16x8_t __p0, int16x8_t __p1) {
3940 uint16x8_t __ret;
3941 __ret = (uint16x8_t)(__p0 > __p1);
3942 return __ret;
3943 }
3944 #else
vcgtq_s16(int16x8_t __p0,int16x8_t __p1)3945 __ai uint16x8_t vcgtq_s16(int16x8_t __p0, int16x8_t __p1) {
3946 uint16x8_t __ret;
3947 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
3948 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
3949 __ret = (uint16x8_t)(__rev0 > __rev1);
3950 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
3951 return __ret;
3952 }
3953 #endif
3954
3955 #ifdef __LITTLE_ENDIAN__
vcgt_u8(uint8x8_t __p0,uint8x8_t __p1)3956 __ai uint8x8_t vcgt_u8(uint8x8_t __p0, uint8x8_t __p1) {
3957 uint8x8_t __ret;
3958 __ret = (uint8x8_t)(__p0 > __p1);
3959 return __ret;
3960 }
3961 #else
vcgt_u8(uint8x8_t __p0,uint8x8_t __p1)3962 __ai uint8x8_t vcgt_u8(uint8x8_t __p0, uint8x8_t __p1) {
3963 uint8x8_t __ret;
3964 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
3965 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
3966 __ret = (uint8x8_t)(__rev0 > __rev1);
3967 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
3968 return __ret;
3969 }
3970 #endif
3971
3972 #ifdef __LITTLE_ENDIAN__
vcgt_u32(uint32x2_t __p0,uint32x2_t __p1)3973 __ai uint32x2_t vcgt_u32(uint32x2_t __p0, uint32x2_t __p1) {
3974 uint32x2_t __ret;
3975 __ret = (uint32x2_t)(__p0 > __p1);
3976 return __ret;
3977 }
3978 #else
vcgt_u32(uint32x2_t __p0,uint32x2_t __p1)3979 __ai uint32x2_t vcgt_u32(uint32x2_t __p0, uint32x2_t __p1) {
3980 uint32x2_t __ret;
3981 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
3982 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
3983 __ret = (uint32x2_t)(__rev0 > __rev1);
3984 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
3985 return __ret;
3986 }
3987 #endif
3988
3989 #ifdef __LITTLE_ENDIAN__
vcgt_u16(uint16x4_t __p0,uint16x4_t __p1)3990 __ai uint16x4_t vcgt_u16(uint16x4_t __p0, uint16x4_t __p1) {
3991 uint16x4_t __ret;
3992 __ret = (uint16x4_t)(__p0 > __p1);
3993 return __ret;
3994 }
3995 #else
vcgt_u16(uint16x4_t __p0,uint16x4_t __p1)3996 __ai uint16x4_t vcgt_u16(uint16x4_t __p0, uint16x4_t __p1) {
3997 uint16x4_t __ret;
3998 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
3999 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
4000 __ret = (uint16x4_t)(__rev0 > __rev1);
4001 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
4002 return __ret;
4003 }
4004 #endif
4005
4006 #ifdef __LITTLE_ENDIAN__
vcgt_s8(int8x8_t __p0,int8x8_t __p1)4007 __ai uint8x8_t vcgt_s8(int8x8_t __p0, int8x8_t __p1) {
4008 uint8x8_t __ret;
4009 __ret = (uint8x8_t)(__p0 > __p1);
4010 return __ret;
4011 }
4012 #else
vcgt_s8(int8x8_t __p0,int8x8_t __p1)4013 __ai uint8x8_t vcgt_s8(int8x8_t __p0, int8x8_t __p1) {
4014 uint8x8_t __ret;
4015 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
4016 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
4017 __ret = (uint8x8_t)(__rev0 > __rev1);
4018 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
4019 return __ret;
4020 }
4021 #endif
4022
4023 #ifdef __LITTLE_ENDIAN__
vcgt_f32(float32x2_t __p0,float32x2_t __p1)4024 __ai uint32x2_t vcgt_f32(float32x2_t __p0, float32x2_t __p1) {
4025 uint32x2_t __ret;
4026 __ret = (uint32x2_t)(__p0 > __p1);
4027 return __ret;
4028 }
4029 #else
vcgt_f32(float32x2_t __p0,float32x2_t __p1)4030 __ai uint32x2_t vcgt_f32(float32x2_t __p0, float32x2_t __p1) {
4031 uint32x2_t __ret;
4032 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
4033 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
4034 __ret = (uint32x2_t)(__rev0 > __rev1);
4035 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
4036 return __ret;
4037 }
4038 #endif
4039
4040 #ifdef __LITTLE_ENDIAN__
vcgt_s32(int32x2_t __p0,int32x2_t __p1)4041 __ai uint32x2_t vcgt_s32(int32x2_t __p0, int32x2_t __p1) {
4042 uint32x2_t __ret;
4043 __ret = (uint32x2_t)(__p0 > __p1);
4044 return __ret;
4045 }
4046 #else
vcgt_s32(int32x2_t __p0,int32x2_t __p1)4047 __ai uint32x2_t vcgt_s32(int32x2_t __p0, int32x2_t __p1) {
4048 uint32x2_t __ret;
4049 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
4050 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
4051 __ret = (uint32x2_t)(__rev0 > __rev1);
4052 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
4053 return __ret;
4054 }
4055 #endif
4056
4057 #ifdef __LITTLE_ENDIAN__
vcgt_s16(int16x4_t __p0,int16x4_t __p1)4058 __ai uint16x4_t vcgt_s16(int16x4_t __p0, int16x4_t __p1) {
4059 uint16x4_t __ret;
4060 __ret = (uint16x4_t)(__p0 > __p1);
4061 return __ret;
4062 }
4063 #else
vcgt_s16(int16x4_t __p0,int16x4_t __p1)4064 __ai uint16x4_t vcgt_s16(int16x4_t __p0, int16x4_t __p1) {
4065 uint16x4_t __ret;
4066 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
4067 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
4068 __ret = (uint16x4_t)(__rev0 > __rev1);
4069 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
4070 return __ret;
4071 }
4072 #endif
4073
4074 #ifdef __LITTLE_ENDIAN__
vcleq_u8(uint8x16_t __p0,uint8x16_t __p1)4075 __ai uint8x16_t vcleq_u8(uint8x16_t __p0, uint8x16_t __p1) {
4076 uint8x16_t __ret;
4077 __ret = (uint8x16_t)(__p0 <= __p1);
4078 return __ret;
4079 }
4080 #else
vcleq_u8(uint8x16_t __p0,uint8x16_t __p1)4081 __ai uint8x16_t vcleq_u8(uint8x16_t __p0, uint8x16_t __p1) {
4082 uint8x16_t __ret;
4083 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
4084 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
4085 __ret = (uint8x16_t)(__rev0 <= __rev1);
4086 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
4087 return __ret;
4088 }
4089 #endif
4090
4091 #ifdef __LITTLE_ENDIAN__
vcleq_u32(uint32x4_t __p0,uint32x4_t __p1)4092 __ai uint32x4_t vcleq_u32(uint32x4_t __p0, uint32x4_t __p1) {
4093 uint32x4_t __ret;
4094 __ret = (uint32x4_t)(__p0 <= __p1);
4095 return __ret;
4096 }
4097 #else
vcleq_u32(uint32x4_t __p0,uint32x4_t __p1)4098 __ai uint32x4_t vcleq_u32(uint32x4_t __p0, uint32x4_t __p1) {
4099 uint32x4_t __ret;
4100 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
4101 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
4102 __ret = (uint32x4_t)(__rev0 <= __rev1);
4103 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
4104 return __ret;
4105 }
4106 #endif
4107
4108 #ifdef __LITTLE_ENDIAN__
vcleq_u16(uint16x8_t __p0,uint16x8_t __p1)4109 __ai uint16x8_t vcleq_u16(uint16x8_t __p0, uint16x8_t __p1) {
4110 uint16x8_t __ret;
4111 __ret = (uint16x8_t)(__p0 <= __p1);
4112 return __ret;
4113 }
4114 #else
vcleq_u16(uint16x8_t __p0,uint16x8_t __p1)4115 __ai uint16x8_t vcleq_u16(uint16x8_t __p0, uint16x8_t __p1) {
4116 uint16x8_t __ret;
4117 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
4118 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
4119 __ret = (uint16x8_t)(__rev0 <= __rev1);
4120 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
4121 return __ret;
4122 }
4123 #endif
4124
4125 #ifdef __LITTLE_ENDIAN__
vcleq_s8(int8x16_t __p0,int8x16_t __p1)4126 __ai uint8x16_t vcleq_s8(int8x16_t __p0, int8x16_t __p1) {
4127 uint8x16_t __ret;
4128 __ret = (uint8x16_t)(__p0 <= __p1);
4129 return __ret;
4130 }
4131 #else
vcleq_s8(int8x16_t __p0,int8x16_t __p1)4132 __ai uint8x16_t vcleq_s8(int8x16_t __p0, int8x16_t __p1) {
4133 uint8x16_t __ret;
4134 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
4135 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
4136 __ret = (uint8x16_t)(__rev0 <= __rev1);
4137 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
4138 return __ret;
4139 }
4140 #endif
4141
4142 #ifdef __LITTLE_ENDIAN__
vcleq_f32(float32x4_t __p0,float32x4_t __p1)4143 __ai uint32x4_t vcleq_f32(float32x4_t __p0, float32x4_t __p1) {
4144 uint32x4_t __ret;
4145 __ret = (uint32x4_t)(__p0 <= __p1);
4146 return __ret;
4147 }
4148 #else
vcleq_f32(float32x4_t __p0,float32x4_t __p1)4149 __ai uint32x4_t vcleq_f32(float32x4_t __p0, float32x4_t __p1) {
4150 uint32x4_t __ret;
4151 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
4152 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
4153 __ret = (uint32x4_t)(__rev0 <= __rev1);
4154 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
4155 return __ret;
4156 }
4157 #endif
4158
4159 #ifdef __LITTLE_ENDIAN__
vcleq_s32(int32x4_t __p0,int32x4_t __p1)4160 __ai uint32x4_t vcleq_s32(int32x4_t __p0, int32x4_t __p1) {
4161 uint32x4_t __ret;
4162 __ret = (uint32x4_t)(__p0 <= __p1);
4163 return __ret;
4164 }
4165 #else
vcleq_s32(int32x4_t __p0,int32x4_t __p1)4166 __ai uint32x4_t vcleq_s32(int32x4_t __p0, int32x4_t __p1) {
4167 uint32x4_t __ret;
4168 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
4169 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
4170 __ret = (uint32x4_t)(__rev0 <= __rev1);
4171 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
4172 return __ret;
4173 }
4174 #endif
4175
4176 #ifdef __LITTLE_ENDIAN__
vcleq_s16(int16x8_t __p0,int16x8_t __p1)4177 __ai uint16x8_t vcleq_s16(int16x8_t __p0, int16x8_t __p1) {
4178 uint16x8_t __ret;
4179 __ret = (uint16x8_t)(__p0 <= __p1);
4180 return __ret;
4181 }
4182 #else
vcleq_s16(int16x8_t __p0,int16x8_t __p1)4183 __ai uint16x8_t vcleq_s16(int16x8_t __p0, int16x8_t __p1) {
4184 uint16x8_t __ret;
4185 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
4186 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
4187 __ret = (uint16x8_t)(__rev0 <= __rev1);
4188 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
4189 return __ret;
4190 }
4191 #endif
4192
4193 #ifdef __LITTLE_ENDIAN__
vcle_u8(uint8x8_t __p0,uint8x8_t __p1)4194 __ai uint8x8_t vcle_u8(uint8x8_t __p0, uint8x8_t __p1) {
4195 uint8x8_t __ret;
4196 __ret = (uint8x8_t)(__p0 <= __p1);
4197 return __ret;
4198 }
4199 #else
vcle_u8(uint8x8_t __p0,uint8x8_t __p1)4200 __ai uint8x8_t vcle_u8(uint8x8_t __p0, uint8x8_t __p1) {
4201 uint8x8_t __ret;
4202 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
4203 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
4204 __ret = (uint8x8_t)(__rev0 <= __rev1);
4205 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
4206 return __ret;
4207 }
4208 #endif
4209
4210 #ifdef __LITTLE_ENDIAN__
vcle_u32(uint32x2_t __p0,uint32x2_t __p1)4211 __ai uint32x2_t vcle_u32(uint32x2_t __p0, uint32x2_t __p1) {
4212 uint32x2_t __ret;
4213 __ret = (uint32x2_t)(__p0 <= __p1);
4214 return __ret;
4215 }
4216 #else
vcle_u32(uint32x2_t __p0,uint32x2_t __p1)4217 __ai uint32x2_t vcle_u32(uint32x2_t __p0, uint32x2_t __p1) {
4218 uint32x2_t __ret;
4219 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
4220 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
4221 __ret = (uint32x2_t)(__rev0 <= __rev1);
4222 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
4223 return __ret;
4224 }
4225 #endif
4226
4227 #ifdef __LITTLE_ENDIAN__
vcle_u16(uint16x4_t __p0,uint16x4_t __p1)4228 __ai uint16x4_t vcle_u16(uint16x4_t __p0, uint16x4_t __p1) {
4229 uint16x4_t __ret;
4230 __ret = (uint16x4_t)(__p0 <= __p1);
4231 return __ret;
4232 }
4233 #else
vcle_u16(uint16x4_t __p0,uint16x4_t __p1)4234 __ai uint16x4_t vcle_u16(uint16x4_t __p0, uint16x4_t __p1) {
4235 uint16x4_t __ret;
4236 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
4237 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
4238 __ret = (uint16x4_t)(__rev0 <= __rev1);
4239 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
4240 return __ret;
4241 }
4242 #endif
4243
4244 #ifdef __LITTLE_ENDIAN__
vcle_s8(int8x8_t __p0,int8x8_t __p1)4245 __ai uint8x8_t vcle_s8(int8x8_t __p0, int8x8_t __p1) {
4246 uint8x8_t __ret;
4247 __ret = (uint8x8_t)(__p0 <= __p1);
4248 return __ret;
4249 }
4250 #else
vcle_s8(int8x8_t __p0,int8x8_t __p1)4251 __ai uint8x8_t vcle_s8(int8x8_t __p0, int8x8_t __p1) {
4252 uint8x8_t __ret;
4253 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
4254 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
4255 __ret = (uint8x8_t)(__rev0 <= __rev1);
4256 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
4257 return __ret;
4258 }
4259 #endif
4260
4261 #ifdef __LITTLE_ENDIAN__
vcle_f32(float32x2_t __p0,float32x2_t __p1)4262 __ai uint32x2_t vcle_f32(float32x2_t __p0, float32x2_t __p1) {
4263 uint32x2_t __ret;
4264 __ret = (uint32x2_t)(__p0 <= __p1);
4265 return __ret;
4266 }
4267 #else
vcle_f32(float32x2_t __p0,float32x2_t __p1)4268 __ai uint32x2_t vcle_f32(float32x2_t __p0, float32x2_t __p1) {
4269 uint32x2_t __ret;
4270 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
4271 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
4272 __ret = (uint32x2_t)(__rev0 <= __rev1);
4273 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
4274 return __ret;
4275 }
4276 #endif
4277
4278 #ifdef __LITTLE_ENDIAN__
vcle_s32(int32x2_t __p0,int32x2_t __p1)4279 __ai uint32x2_t vcle_s32(int32x2_t __p0, int32x2_t __p1) {
4280 uint32x2_t __ret;
4281 __ret = (uint32x2_t)(__p0 <= __p1);
4282 return __ret;
4283 }
4284 #else
vcle_s32(int32x2_t __p0,int32x2_t __p1)4285 __ai uint32x2_t vcle_s32(int32x2_t __p0, int32x2_t __p1) {
4286 uint32x2_t __ret;
4287 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
4288 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
4289 __ret = (uint32x2_t)(__rev0 <= __rev1);
4290 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
4291 return __ret;
4292 }
4293 #endif
4294
4295 #ifdef __LITTLE_ENDIAN__
vcle_s16(int16x4_t __p0,int16x4_t __p1)4296 __ai uint16x4_t vcle_s16(int16x4_t __p0, int16x4_t __p1) {
4297 uint16x4_t __ret;
4298 __ret = (uint16x4_t)(__p0 <= __p1);
4299 return __ret;
4300 }
4301 #else
vcle_s16(int16x4_t __p0,int16x4_t __p1)4302 __ai uint16x4_t vcle_s16(int16x4_t __p0, int16x4_t __p1) {
4303 uint16x4_t __ret;
4304 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
4305 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
4306 __ret = (uint16x4_t)(__rev0 <= __rev1);
4307 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
4308 return __ret;
4309 }
4310 #endif
4311
4312 #ifdef __LITTLE_ENDIAN__
vclsq_u8(uint8x16_t __p0)4313 __ai int8x16_t vclsq_u8(uint8x16_t __p0) {
4314 int8x16_t __ret;
4315 __ret = (int8x16_t) __builtin_neon_vclsq_v((int8x16_t)__p0, 32);
4316 return __ret;
4317 }
4318 #else
vclsq_u8(uint8x16_t __p0)4319 __ai int8x16_t vclsq_u8(uint8x16_t __p0) {
4320 int8x16_t __ret;
4321 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
4322 __ret = (int8x16_t) __builtin_neon_vclsq_v((int8x16_t)__rev0, 32);
4323 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
4324 return __ret;
4325 }
4326 #endif
4327
4328 #ifdef __LITTLE_ENDIAN__
vclsq_u32(uint32x4_t __p0)4329 __ai int32x4_t vclsq_u32(uint32x4_t __p0) {
4330 int32x4_t __ret;
4331 __ret = (int32x4_t) __builtin_neon_vclsq_v((int8x16_t)__p0, 34);
4332 return __ret;
4333 }
4334 #else
vclsq_u32(uint32x4_t __p0)4335 __ai int32x4_t vclsq_u32(uint32x4_t __p0) {
4336 int32x4_t __ret;
4337 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
4338 __ret = (int32x4_t) __builtin_neon_vclsq_v((int8x16_t)__rev0, 34);
4339 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
4340 return __ret;
4341 }
4342 #endif
4343
4344 #ifdef __LITTLE_ENDIAN__
vclsq_u16(uint16x8_t __p0)4345 __ai int16x8_t vclsq_u16(uint16x8_t __p0) {
4346 int16x8_t __ret;
4347 __ret = (int16x8_t) __builtin_neon_vclsq_v((int8x16_t)__p0, 33);
4348 return __ret;
4349 }
4350 #else
vclsq_u16(uint16x8_t __p0)4351 __ai int16x8_t vclsq_u16(uint16x8_t __p0) {
4352 int16x8_t __ret;
4353 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
4354 __ret = (int16x8_t) __builtin_neon_vclsq_v((int8x16_t)__rev0, 33);
4355 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
4356 return __ret;
4357 }
4358 #endif
4359
4360 #ifdef __LITTLE_ENDIAN__
vclsq_s8(int8x16_t __p0)4361 __ai int8x16_t vclsq_s8(int8x16_t __p0) {
4362 int8x16_t __ret;
4363 __ret = (int8x16_t) __builtin_neon_vclsq_v((int8x16_t)__p0, 32);
4364 return __ret;
4365 }
4366 #else
vclsq_s8(int8x16_t __p0)4367 __ai int8x16_t vclsq_s8(int8x16_t __p0) {
4368 int8x16_t __ret;
4369 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
4370 __ret = (int8x16_t) __builtin_neon_vclsq_v((int8x16_t)__rev0, 32);
4371 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
4372 return __ret;
4373 }
4374 #endif
4375
4376 #ifdef __LITTLE_ENDIAN__
vclsq_s32(int32x4_t __p0)4377 __ai int32x4_t vclsq_s32(int32x4_t __p0) {
4378 int32x4_t __ret;
4379 __ret = (int32x4_t) __builtin_neon_vclsq_v((int8x16_t)__p0, 34);
4380 return __ret;
4381 }
4382 #else
vclsq_s32(int32x4_t __p0)4383 __ai int32x4_t vclsq_s32(int32x4_t __p0) {
4384 int32x4_t __ret;
4385 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
4386 __ret = (int32x4_t) __builtin_neon_vclsq_v((int8x16_t)__rev0, 34);
4387 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
4388 return __ret;
4389 }
4390 #endif
4391
4392 #ifdef __LITTLE_ENDIAN__
vclsq_s16(int16x8_t __p0)4393 __ai int16x8_t vclsq_s16(int16x8_t __p0) {
4394 int16x8_t __ret;
4395 __ret = (int16x8_t) __builtin_neon_vclsq_v((int8x16_t)__p0, 33);
4396 return __ret;
4397 }
4398 #else
vclsq_s16(int16x8_t __p0)4399 __ai int16x8_t vclsq_s16(int16x8_t __p0) {
4400 int16x8_t __ret;
4401 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
4402 __ret = (int16x8_t) __builtin_neon_vclsq_v((int8x16_t)__rev0, 33);
4403 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
4404 return __ret;
4405 }
4406 #endif
4407
4408 #ifdef __LITTLE_ENDIAN__
vcls_u8(uint8x8_t __p0)4409 __ai int8x8_t vcls_u8(uint8x8_t __p0) {
4410 int8x8_t __ret;
4411 __ret = (int8x8_t) __builtin_neon_vcls_v((int8x8_t)__p0, 0);
4412 return __ret;
4413 }
4414 #else
vcls_u8(uint8x8_t __p0)4415 __ai int8x8_t vcls_u8(uint8x8_t __p0) {
4416 int8x8_t __ret;
4417 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
4418 __ret = (int8x8_t) __builtin_neon_vcls_v((int8x8_t)__rev0, 0);
4419 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
4420 return __ret;
4421 }
4422 #endif
4423
4424 #ifdef __LITTLE_ENDIAN__
vcls_u32(uint32x2_t __p0)4425 __ai int32x2_t vcls_u32(uint32x2_t __p0) {
4426 int32x2_t __ret;
4427 __ret = (int32x2_t) __builtin_neon_vcls_v((int8x8_t)__p0, 2);
4428 return __ret;
4429 }
4430 #else
vcls_u32(uint32x2_t __p0)4431 __ai int32x2_t vcls_u32(uint32x2_t __p0) {
4432 int32x2_t __ret;
4433 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
4434 __ret = (int32x2_t) __builtin_neon_vcls_v((int8x8_t)__rev0, 2);
4435 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
4436 return __ret;
4437 }
4438 #endif
4439
4440 #ifdef __LITTLE_ENDIAN__
vcls_u16(uint16x4_t __p0)4441 __ai int16x4_t vcls_u16(uint16x4_t __p0) {
4442 int16x4_t __ret;
4443 __ret = (int16x4_t) __builtin_neon_vcls_v((int8x8_t)__p0, 1);
4444 return __ret;
4445 }
4446 #else
vcls_u16(uint16x4_t __p0)4447 __ai int16x4_t vcls_u16(uint16x4_t __p0) {
4448 int16x4_t __ret;
4449 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
4450 __ret = (int16x4_t) __builtin_neon_vcls_v((int8x8_t)__rev0, 1);
4451 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
4452 return __ret;
4453 }
4454 #endif
4455
4456 #ifdef __LITTLE_ENDIAN__
vcls_s8(int8x8_t __p0)4457 __ai int8x8_t vcls_s8(int8x8_t __p0) {
4458 int8x8_t __ret;
4459 __ret = (int8x8_t) __builtin_neon_vcls_v((int8x8_t)__p0, 0);
4460 return __ret;
4461 }
4462 #else
vcls_s8(int8x8_t __p0)4463 __ai int8x8_t vcls_s8(int8x8_t __p0) {
4464 int8x8_t __ret;
4465 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
4466 __ret = (int8x8_t) __builtin_neon_vcls_v((int8x8_t)__rev0, 0);
4467 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
4468 return __ret;
4469 }
4470 #endif
4471
4472 #ifdef __LITTLE_ENDIAN__
vcls_s32(int32x2_t __p0)4473 __ai int32x2_t vcls_s32(int32x2_t __p0) {
4474 int32x2_t __ret;
4475 __ret = (int32x2_t) __builtin_neon_vcls_v((int8x8_t)__p0, 2);
4476 return __ret;
4477 }
4478 #else
vcls_s32(int32x2_t __p0)4479 __ai int32x2_t vcls_s32(int32x2_t __p0) {
4480 int32x2_t __ret;
4481 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
4482 __ret = (int32x2_t) __builtin_neon_vcls_v((int8x8_t)__rev0, 2);
4483 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
4484 return __ret;
4485 }
4486 #endif
4487
4488 #ifdef __LITTLE_ENDIAN__
vcls_s16(int16x4_t __p0)4489 __ai int16x4_t vcls_s16(int16x4_t __p0) {
4490 int16x4_t __ret;
4491 __ret = (int16x4_t) __builtin_neon_vcls_v((int8x8_t)__p0, 1);
4492 return __ret;
4493 }
4494 #else
vcls_s16(int16x4_t __p0)4495 __ai int16x4_t vcls_s16(int16x4_t __p0) {
4496 int16x4_t __ret;
4497 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
4498 __ret = (int16x4_t) __builtin_neon_vcls_v((int8x8_t)__rev0, 1);
4499 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
4500 return __ret;
4501 }
4502 #endif
4503
4504 #ifdef __LITTLE_ENDIAN__
vcltq_u8(uint8x16_t __p0,uint8x16_t __p1)4505 __ai uint8x16_t vcltq_u8(uint8x16_t __p0, uint8x16_t __p1) {
4506 uint8x16_t __ret;
4507 __ret = (uint8x16_t)(__p0 < __p1);
4508 return __ret;
4509 }
4510 #else
vcltq_u8(uint8x16_t __p0,uint8x16_t __p1)4511 __ai uint8x16_t vcltq_u8(uint8x16_t __p0, uint8x16_t __p1) {
4512 uint8x16_t __ret;
4513 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
4514 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
4515 __ret = (uint8x16_t)(__rev0 < __rev1);
4516 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
4517 return __ret;
4518 }
4519 #endif
4520
4521 #ifdef __LITTLE_ENDIAN__
vcltq_u32(uint32x4_t __p0,uint32x4_t __p1)4522 __ai uint32x4_t vcltq_u32(uint32x4_t __p0, uint32x4_t __p1) {
4523 uint32x4_t __ret;
4524 __ret = (uint32x4_t)(__p0 < __p1);
4525 return __ret;
4526 }
4527 #else
vcltq_u32(uint32x4_t __p0,uint32x4_t __p1)4528 __ai uint32x4_t vcltq_u32(uint32x4_t __p0, uint32x4_t __p1) {
4529 uint32x4_t __ret;
4530 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
4531 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
4532 __ret = (uint32x4_t)(__rev0 < __rev1);
4533 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
4534 return __ret;
4535 }
4536 #endif
4537
4538 #ifdef __LITTLE_ENDIAN__
vcltq_u16(uint16x8_t __p0,uint16x8_t __p1)4539 __ai uint16x8_t vcltq_u16(uint16x8_t __p0, uint16x8_t __p1) {
4540 uint16x8_t __ret;
4541 __ret = (uint16x8_t)(__p0 < __p1);
4542 return __ret;
4543 }
4544 #else
vcltq_u16(uint16x8_t __p0,uint16x8_t __p1)4545 __ai uint16x8_t vcltq_u16(uint16x8_t __p0, uint16x8_t __p1) {
4546 uint16x8_t __ret;
4547 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
4548 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
4549 __ret = (uint16x8_t)(__rev0 < __rev1);
4550 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
4551 return __ret;
4552 }
4553 #endif
4554
4555 #ifdef __LITTLE_ENDIAN__
vcltq_s8(int8x16_t __p0,int8x16_t __p1)4556 __ai uint8x16_t vcltq_s8(int8x16_t __p0, int8x16_t __p1) {
4557 uint8x16_t __ret;
4558 __ret = (uint8x16_t)(__p0 < __p1);
4559 return __ret;
4560 }
4561 #else
vcltq_s8(int8x16_t __p0,int8x16_t __p1)4562 __ai uint8x16_t vcltq_s8(int8x16_t __p0, int8x16_t __p1) {
4563 uint8x16_t __ret;
4564 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
4565 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
4566 __ret = (uint8x16_t)(__rev0 < __rev1);
4567 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
4568 return __ret;
4569 }
4570 #endif
4571
4572 #ifdef __LITTLE_ENDIAN__
vcltq_f32(float32x4_t __p0,float32x4_t __p1)4573 __ai uint32x4_t vcltq_f32(float32x4_t __p0, float32x4_t __p1) {
4574 uint32x4_t __ret;
4575 __ret = (uint32x4_t)(__p0 < __p1);
4576 return __ret;
4577 }
4578 #else
vcltq_f32(float32x4_t __p0,float32x4_t __p1)4579 __ai uint32x4_t vcltq_f32(float32x4_t __p0, float32x4_t __p1) {
4580 uint32x4_t __ret;
4581 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
4582 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
4583 __ret = (uint32x4_t)(__rev0 < __rev1);
4584 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
4585 return __ret;
4586 }
4587 #endif
4588
4589 #ifdef __LITTLE_ENDIAN__
vcltq_s32(int32x4_t __p0,int32x4_t __p1)4590 __ai uint32x4_t vcltq_s32(int32x4_t __p0, int32x4_t __p1) {
4591 uint32x4_t __ret;
4592 __ret = (uint32x4_t)(__p0 < __p1);
4593 return __ret;
4594 }
4595 #else
vcltq_s32(int32x4_t __p0,int32x4_t __p1)4596 __ai uint32x4_t vcltq_s32(int32x4_t __p0, int32x4_t __p1) {
4597 uint32x4_t __ret;
4598 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
4599 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
4600 __ret = (uint32x4_t)(__rev0 < __rev1);
4601 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
4602 return __ret;
4603 }
4604 #endif
4605
4606 #ifdef __LITTLE_ENDIAN__
vcltq_s16(int16x8_t __p0,int16x8_t __p1)4607 __ai uint16x8_t vcltq_s16(int16x8_t __p0, int16x8_t __p1) {
4608 uint16x8_t __ret;
4609 __ret = (uint16x8_t)(__p0 < __p1);
4610 return __ret;
4611 }
4612 #else
vcltq_s16(int16x8_t __p0,int16x8_t __p1)4613 __ai uint16x8_t vcltq_s16(int16x8_t __p0, int16x8_t __p1) {
4614 uint16x8_t __ret;
4615 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
4616 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
4617 __ret = (uint16x8_t)(__rev0 < __rev1);
4618 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
4619 return __ret;
4620 }
4621 #endif
4622
4623 #ifdef __LITTLE_ENDIAN__
vclt_u8(uint8x8_t __p0,uint8x8_t __p1)4624 __ai uint8x8_t vclt_u8(uint8x8_t __p0, uint8x8_t __p1) {
4625 uint8x8_t __ret;
4626 __ret = (uint8x8_t)(__p0 < __p1);
4627 return __ret;
4628 }
4629 #else
vclt_u8(uint8x8_t __p0,uint8x8_t __p1)4630 __ai uint8x8_t vclt_u8(uint8x8_t __p0, uint8x8_t __p1) {
4631 uint8x8_t __ret;
4632 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
4633 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
4634 __ret = (uint8x8_t)(__rev0 < __rev1);
4635 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
4636 return __ret;
4637 }
4638 #endif
4639
4640 #ifdef __LITTLE_ENDIAN__
vclt_u32(uint32x2_t __p0,uint32x2_t __p1)4641 __ai uint32x2_t vclt_u32(uint32x2_t __p0, uint32x2_t __p1) {
4642 uint32x2_t __ret;
4643 __ret = (uint32x2_t)(__p0 < __p1);
4644 return __ret;
4645 }
4646 #else
vclt_u32(uint32x2_t __p0,uint32x2_t __p1)4647 __ai uint32x2_t vclt_u32(uint32x2_t __p0, uint32x2_t __p1) {
4648 uint32x2_t __ret;
4649 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
4650 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
4651 __ret = (uint32x2_t)(__rev0 < __rev1);
4652 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
4653 return __ret;
4654 }
4655 #endif
4656
4657 #ifdef __LITTLE_ENDIAN__
vclt_u16(uint16x4_t __p0,uint16x4_t __p1)4658 __ai uint16x4_t vclt_u16(uint16x4_t __p0, uint16x4_t __p1) {
4659 uint16x4_t __ret;
4660 __ret = (uint16x4_t)(__p0 < __p1);
4661 return __ret;
4662 }
4663 #else
vclt_u16(uint16x4_t __p0,uint16x4_t __p1)4664 __ai uint16x4_t vclt_u16(uint16x4_t __p0, uint16x4_t __p1) {
4665 uint16x4_t __ret;
4666 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
4667 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
4668 __ret = (uint16x4_t)(__rev0 < __rev1);
4669 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
4670 return __ret;
4671 }
4672 #endif
4673
4674 #ifdef __LITTLE_ENDIAN__
vclt_s8(int8x8_t __p0,int8x8_t __p1)4675 __ai uint8x8_t vclt_s8(int8x8_t __p0, int8x8_t __p1) {
4676 uint8x8_t __ret;
4677 __ret = (uint8x8_t)(__p0 < __p1);
4678 return __ret;
4679 }
4680 #else
vclt_s8(int8x8_t __p0,int8x8_t __p1)4681 __ai uint8x8_t vclt_s8(int8x8_t __p0, int8x8_t __p1) {
4682 uint8x8_t __ret;
4683 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
4684 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
4685 __ret = (uint8x8_t)(__rev0 < __rev1);
4686 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
4687 return __ret;
4688 }
4689 #endif
4690
4691 #ifdef __LITTLE_ENDIAN__
vclt_f32(float32x2_t __p0,float32x2_t __p1)4692 __ai uint32x2_t vclt_f32(float32x2_t __p0, float32x2_t __p1) {
4693 uint32x2_t __ret;
4694 __ret = (uint32x2_t)(__p0 < __p1);
4695 return __ret;
4696 }
4697 #else
vclt_f32(float32x2_t __p0,float32x2_t __p1)4698 __ai uint32x2_t vclt_f32(float32x2_t __p0, float32x2_t __p1) {
4699 uint32x2_t __ret;
4700 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
4701 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
4702 __ret = (uint32x2_t)(__rev0 < __rev1);
4703 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
4704 return __ret;
4705 }
4706 #endif
4707
4708 #ifdef __LITTLE_ENDIAN__
vclt_s32(int32x2_t __p0,int32x2_t __p1)4709 __ai uint32x2_t vclt_s32(int32x2_t __p0, int32x2_t __p1) {
4710 uint32x2_t __ret;
4711 __ret = (uint32x2_t)(__p0 < __p1);
4712 return __ret;
4713 }
4714 #else
vclt_s32(int32x2_t __p0,int32x2_t __p1)4715 __ai uint32x2_t vclt_s32(int32x2_t __p0, int32x2_t __p1) {
4716 uint32x2_t __ret;
4717 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
4718 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
4719 __ret = (uint32x2_t)(__rev0 < __rev1);
4720 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
4721 return __ret;
4722 }
4723 #endif
4724
4725 #ifdef __LITTLE_ENDIAN__
vclt_s16(int16x4_t __p0,int16x4_t __p1)4726 __ai uint16x4_t vclt_s16(int16x4_t __p0, int16x4_t __p1) {
4727 uint16x4_t __ret;
4728 __ret = (uint16x4_t)(__p0 < __p1);
4729 return __ret;
4730 }
4731 #else
vclt_s16(int16x4_t __p0,int16x4_t __p1)4732 __ai uint16x4_t vclt_s16(int16x4_t __p0, int16x4_t __p1) {
4733 uint16x4_t __ret;
4734 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
4735 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
4736 __ret = (uint16x4_t)(__rev0 < __rev1);
4737 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
4738 return __ret;
4739 }
4740 #endif
4741
4742 #ifdef __LITTLE_ENDIAN__
vclzq_u8(uint8x16_t __p0)4743 __ai uint8x16_t vclzq_u8(uint8x16_t __p0) {
4744 uint8x16_t __ret;
4745 __ret = (uint8x16_t) __builtin_neon_vclzq_v((int8x16_t)__p0, 48);
4746 return __ret;
4747 }
4748 #else
vclzq_u8(uint8x16_t __p0)4749 __ai uint8x16_t vclzq_u8(uint8x16_t __p0) {
4750 uint8x16_t __ret;
4751 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
4752 __ret = (uint8x16_t) __builtin_neon_vclzq_v((int8x16_t)__rev0, 48);
4753 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
4754 return __ret;
4755 }
4756 #endif
4757
4758 #ifdef __LITTLE_ENDIAN__
vclzq_u32(uint32x4_t __p0)4759 __ai uint32x4_t vclzq_u32(uint32x4_t __p0) {
4760 uint32x4_t __ret;
4761 __ret = (uint32x4_t) __builtin_neon_vclzq_v((int8x16_t)__p0, 50);
4762 return __ret;
4763 }
4764 #else
vclzq_u32(uint32x4_t __p0)4765 __ai uint32x4_t vclzq_u32(uint32x4_t __p0) {
4766 uint32x4_t __ret;
4767 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
4768 __ret = (uint32x4_t) __builtin_neon_vclzq_v((int8x16_t)__rev0, 50);
4769 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
4770 return __ret;
4771 }
4772 #endif
4773
4774 #ifdef __LITTLE_ENDIAN__
vclzq_u16(uint16x8_t __p0)4775 __ai uint16x8_t vclzq_u16(uint16x8_t __p0) {
4776 uint16x8_t __ret;
4777 __ret = (uint16x8_t) __builtin_neon_vclzq_v((int8x16_t)__p0, 49);
4778 return __ret;
4779 }
4780 #else
vclzq_u16(uint16x8_t __p0)4781 __ai uint16x8_t vclzq_u16(uint16x8_t __p0) {
4782 uint16x8_t __ret;
4783 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
4784 __ret = (uint16x8_t) __builtin_neon_vclzq_v((int8x16_t)__rev0, 49);
4785 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
4786 return __ret;
4787 }
4788 #endif
4789
4790 #ifdef __LITTLE_ENDIAN__
vclzq_s8(int8x16_t __p0)4791 __ai int8x16_t vclzq_s8(int8x16_t __p0) {
4792 int8x16_t __ret;
4793 __ret = (int8x16_t) __builtin_neon_vclzq_v((int8x16_t)__p0, 32);
4794 return __ret;
4795 }
4796 #else
vclzq_s8(int8x16_t __p0)4797 __ai int8x16_t vclzq_s8(int8x16_t __p0) {
4798 int8x16_t __ret;
4799 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
4800 __ret = (int8x16_t) __builtin_neon_vclzq_v((int8x16_t)__rev0, 32);
4801 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
4802 return __ret;
4803 }
4804 #endif
4805
4806 #ifdef __LITTLE_ENDIAN__
vclzq_s32(int32x4_t __p0)4807 __ai int32x4_t vclzq_s32(int32x4_t __p0) {
4808 int32x4_t __ret;
4809 __ret = (int32x4_t) __builtin_neon_vclzq_v((int8x16_t)__p0, 34);
4810 return __ret;
4811 }
4812 #else
vclzq_s32(int32x4_t __p0)4813 __ai int32x4_t vclzq_s32(int32x4_t __p0) {
4814 int32x4_t __ret;
4815 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
4816 __ret = (int32x4_t) __builtin_neon_vclzq_v((int8x16_t)__rev0, 34);
4817 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
4818 return __ret;
4819 }
4820 #endif
4821
4822 #ifdef __LITTLE_ENDIAN__
vclzq_s16(int16x8_t __p0)4823 __ai int16x8_t vclzq_s16(int16x8_t __p0) {
4824 int16x8_t __ret;
4825 __ret = (int16x8_t) __builtin_neon_vclzq_v((int8x16_t)__p0, 33);
4826 return __ret;
4827 }
4828 #else
vclzq_s16(int16x8_t __p0)4829 __ai int16x8_t vclzq_s16(int16x8_t __p0) {
4830 int16x8_t __ret;
4831 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
4832 __ret = (int16x8_t) __builtin_neon_vclzq_v((int8x16_t)__rev0, 33);
4833 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
4834 return __ret;
4835 }
4836 #endif
4837
4838 #ifdef __LITTLE_ENDIAN__
vclz_u8(uint8x8_t __p0)4839 __ai uint8x8_t vclz_u8(uint8x8_t __p0) {
4840 uint8x8_t __ret;
4841 __ret = (uint8x8_t) __builtin_neon_vclz_v((int8x8_t)__p0, 16);
4842 return __ret;
4843 }
4844 #else
vclz_u8(uint8x8_t __p0)4845 __ai uint8x8_t vclz_u8(uint8x8_t __p0) {
4846 uint8x8_t __ret;
4847 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
4848 __ret = (uint8x8_t) __builtin_neon_vclz_v((int8x8_t)__rev0, 16);
4849 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
4850 return __ret;
4851 }
4852 #endif
4853
4854 #ifdef __LITTLE_ENDIAN__
vclz_u32(uint32x2_t __p0)4855 __ai uint32x2_t vclz_u32(uint32x2_t __p0) {
4856 uint32x2_t __ret;
4857 __ret = (uint32x2_t) __builtin_neon_vclz_v((int8x8_t)__p0, 18);
4858 return __ret;
4859 }
4860 #else
vclz_u32(uint32x2_t __p0)4861 __ai uint32x2_t vclz_u32(uint32x2_t __p0) {
4862 uint32x2_t __ret;
4863 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
4864 __ret = (uint32x2_t) __builtin_neon_vclz_v((int8x8_t)__rev0, 18);
4865 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
4866 return __ret;
4867 }
4868 #endif
4869
4870 #ifdef __LITTLE_ENDIAN__
vclz_u16(uint16x4_t __p0)4871 __ai uint16x4_t vclz_u16(uint16x4_t __p0) {
4872 uint16x4_t __ret;
4873 __ret = (uint16x4_t) __builtin_neon_vclz_v((int8x8_t)__p0, 17);
4874 return __ret;
4875 }
4876 #else
vclz_u16(uint16x4_t __p0)4877 __ai uint16x4_t vclz_u16(uint16x4_t __p0) {
4878 uint16x4_t __ret;
4879 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
4880 __ret = (uint16x4_t) __builtin_neon_vclz_v((int8x8_t)__rev0, 17);
4881 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
4882 return __ret;
4883 }
4884 #endif
4885
4886 #ifdef __LITTLE_ENDIAN__
vclz_s8(int8x8_t __p0)4887 __ai int8x8_t vclz_s8(int8x8_t __p0) {
4888 int8x8_t __ret;
4889 __ret = (int8x8_t) __builtin_neon_vclz_v((int8x8_t)__p0, 0);
4890 return __ret;
4891 }
4892 #else
vclz_s8(int8x8_t __p0)4893 __ai int8x8_t vclz_s8(int8x8_t __p0) {
4894 int8x8_t __ret;
4895 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
4896 __ret = (int8x8_t) __builtin_neon_vclz_v((int8x8_t)__rev0, 0);
4897 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
4898 return __ret;
4899 }
4900 #endif
4901
4902 #ifdef __LITTLE_ENDIAN__
vclz_s32(int32x2_t __p0)4903 __ai int32x2_t vclz_s32(int32x2_t __p0) {
4904 int32x2_t __ret;
4905 __ret = (int32x2_t) __builtin_neon_vclz_v((int8x8_t)__p0, 2);
4906 return __ret;
4907 }
4908 #else
vclz_s32(int32x2_t __p0)4909 __ai int32x2_t vclz_s32(int32x2_t __p0) {
4910 int32x2_t __ret;
4911 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
4912 __ret = (int32x2_t) __builtin_neon_vclz_v((int8x8_t)__rev0, 2);
4913 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
4914 return __ret;
4915 }
4916 #endif
4917
4918 #ifdef __LITTLE_ENDIAN__
vclz_s16(int16x4_t __p0)4919 __ai int16x4_t vclz_s16(int16x4_t __p0) {
4920 int16x4_t __ret;
4921 __ret = (int16x4_t) __builtin_neon_vclz_v((int8x8_t)__p0, 1);
4922 return __ret;
4923 }
4924 #else
vclz_s16(int16x4_t __p0)4925 __ai int16x4_t vclz_s16(int16x4_t __p0) {
4926 int16x4_t __ret;
4927 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
4928 __ret = (int16x4_t) __builtin_neon_vclz_v((int8x8_t)__rev0, 1);
4929 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
4930 return __ret;
4931 }
4932 #endif
4933
4934 #ifdef __LITTLE_ENDIAN__
vcnt_p8(poly8x8_t __p0)4935 __ai poly8x8_t vcnt_p8(poly8x8_t __p0) {
4936 poly8x8_t __ret;
4937 __ret = (poly8x8_t) __builtin_neon_vcnt_v((int8x8_t)__p0, 4);
4938 return __ret;
4939 }
4940 #else
vcnt_p8(poly8x8_t __p0)4941 __ai poly8x8_t vcnt_p8(poly8x8_t __p0) {
4942 poly8x8_t __ret;
4943 poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
4944 __ret = (poly8x8_t) __builtin_neon_vcnt_v((int8x8_t)__rev0, 4);
4945 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
4946 return __ret;
4947 }
4948 #endif
4949
4950 #ifdef __LITTLE_ENDIAN__
vcntq_p8(poly8x16_t __p0)4951 __ai poly8x16_t vcntq_p8(poly8x16_t __p0) {
4952 poly8x16_t __ret;
4953 __ret = (poly8x16_t) __builtin_neon_vcntq_v((int8x16_t)__p0, 36);
4954 return __ret;
4955 }
4956 #else
vcntq_p8(poly8x16_t __p0)4957 __ai poly8x16_t vcntq_p8(poly8x16_t __p0) {
4958 poly8x16_t __ret;
4959 poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
4960 __ret = (poly8x16_t) __builtin_neon_vcntq_v((int8x16_t)__rev0, 36);
4961 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
4962 return __ret;
4963 }
4964 #endif
4965
4966 #ifdef __LITTLE_ENDIAN__
vcntq_u8(uint8x16_t __p0)4967 __ai uint8x16_t vcntq_u8(uint8x16_t __p0) {
4968 uint8x16_t __ret;
4969 __ret = (uint8x16_t) __builtin_neon_vcntq_v((int8x16_t)__p0, 48);
4970 return __ret;
4971 }
4972 #else
vcntq_u8(uint8x16_t __p0)4973 __ai uint8x16_t vcntq_u8(uint8x16_t __p0) {
4974 uint8x16_t __ret;
4975 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
4976 __ret = (uint8x16_t) __builtin_neon_vcntq_v((int8x16_t)__rev0, 48);
4977 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
4978 return __ret;
4979 }
4980 #endif
4981
4982 #ifdef __LITTLE_ENDIAN__
vcntq_s8(int8x16_t __p0)4983 __ai int8x16_t vcntq_s8(int8x16_t __p0) {
4984 int8x16_t __ret;
4985 __ret = (int8x16_t) __builtin_neon_vcntq_v((int8x16_t)__p0, 32);
4986 return __ret;
4987 }
4988 #else
vcntq_s8(int8x16_t __p0)4989 __ai int8x16_t vcntq_s8(int8x16_t __p0) {
4990 int8x16_t __ret;
4991 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
4992 __ret = (int8x16_t) __builtin_neon_vcntq_v((int8x16_t)__rev0, 32);
4993 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
4994 return __ret;
4995 }
4996 #endif
4997
4998 #ifdef __LITTLE_ENDIAN__
vcnt_u8(uint8x8_t __p0)4999 __ai uint8x8_t vcnt_u8(uint8x8_t __p0) {
5000 uint8x8_t __ret;
5001 __ret = (uint8x8_t) __builtin_neon_vcnt_v((int8x8_t)__p0, 16);
5002 return __ret;
5003 }
5004 #else
vcnt_u8(uint8x8_t __p0)5005 __ai uint8x8_t vcnt_u8(uint8x8_t __p0) {
5006 uint8x8_t __ret;
5007 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
5008 __ret = (uint8x8_t) __builtin_neon_vcnt_v((int8x8_t)__rev0, 16);
5009 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
5010 return __ret;
5011 }
5012 #endif
5013
5014 #ifdef __LITTLE_ENDIAN__
vcnt_s8(int8x8_t __p0)5015 __ai int8x8_t vcnt_s8(int8x8_t __p0) {
5016 int8x8_t __ret;
5017 __ret = (int8x8_t) __builtin_neon_vcnt_v((int8x8_t)__p0, 0);
5018 return __ret;
5019 }
5020 #else
vcnt_s8(int8x8_t __p0)5021 __ai int8x8_t vcnt_s8(int8x8_t __p0) {
5022 int8x8_t __ret;
5023 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
5024 __ret = (int8x8_t) __builtin_neon_vcnt_v((int8x8_t)__rev0, 0);
5025 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
5026 return __ret;
5027 }
5028 #endif
5029
5030 #ifdef __LITTLE_ENDIAN__
vcombine_p8(poly8x8_t __p0,poly8x8_t __p1)5031 __ai poly8x16_t vcombine_p8(poly8x8_t __p0, poly8x8_t __p1) {
5032 poly8x16_t __ret;
5033 __ret = __builtin_shufflevector(__p0, __p1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15);
5034 return __ret;
5035 }
5036 #else
vcombine_p8(poly8x8_t __p0,poly8x8_t __p1)5037 __ai poly8x16_t vcombine_p8(poly8x8_t __p0, poly8x8_t __p1) {
5038 poly8x16_t __ret;
5039 poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
5040 poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
5041 __ret = __builtin_shufflevector(__rev0, __rev1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15);
5042 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
5043 return __ret;
5044 }
5045 #endif
5046
5047 #ifdef __LITTLE_ENDIAN__
vcombine_p16(poly16x4_t __p0,poly16x4_t __p1)5048 __ai poly16x8_t vcombine_p16(poly16x4_t __p0, poly16x4_t __p1) {
5049 poly16x8_t __ret;
5050 __ret = __builtin_shufflevector(__p0, __p1, 0, 1, 2, 3, 4, 5, 6, 7);
5051 return __ret;
5052 }
5053 #else
vcombine_p16(poly16x4_t __p0,poly16x4_t __p1)5054 __ai poly16x8_t vcombine_p16(poly16x4_t __p0, poly16x4_t __p1) {
5055 poly16x8_t __ret;
5056 poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
5057 poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
5058 __ret = __builtin_shufflevector(__rev0, __rev1, 0, 1, 2, 3, 4, 5, 6, 7);
5059 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
5060 return __ret;
5061 }
5062 #endif
5063
5064 #ifdef __LITTLE_ENDIAN__
vcombine_u8(uint8x8_t __p0,uint8x8_t __p1)5065 __ai uint8x16_t vcombine_u8(uint8x8_t __p0, uint8x8_t __p1) {
5066 uint8x16_t __ret;
5067 __ret = __builtin_shufflevector(__p0, __p1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15);
5068 return __ret;
5069 }
5070 #else
vcombine_u8(uint8x8_t __p0,uint8x8_t __p1)5071 __ai uint8x16_t vcombine_u8(uint8x8_t __p0, uint8x8_t __p1) {
5072 uint8x16_t __ret;
5073 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
5074 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
5075 __ret = __builtin_shufflevector(__rev0, __rev1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15);
5076 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
5077 return __ret;
5078 }
__noswap_vcombine_u8(uint8x8_t __p0,uint8x8_t __p1)5079 __ai uint8x16_t __noswap_vcombine_u8(uint8x8_t __p0, uint8x8_t __p1) {
5080 uint8x16_t __ret;
5081 __ret = __builtin_shufflevector(__p0, __p1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15);
5082 return __ret;
5083 }
5084 #endif
5085
5086 #ifdef __LITTLE_ENDIAN__
vcombine_u32(uint32x2_t __p0,uint32x2_t __p1)5087 __ai uint32x4_t vcombine_u32(uint32x2_t __p0, uint32x2_t __p1) {
5088 uint32x4_t __ret;
5089 __ret = __builtin_shufflevector(__p0, __p1, 0, 1, 2, 3);
5090 return __ret;
5091 }
5092 #else
vcombine_u32(uint32x2_t __p0,uint32x2_t __p1)5093 __ai uint32x4_t vcombine_u32(uint32x2_t __p0, uint32x2_t __p1) {
5094 uint32x4_t __ret;
5095 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
5096 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
5097 __ret = __builtin_shufflevector(__rev0, __rev1, 0, 1, 2, 3);
5098 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
5099 return __ret;
5100 }
__noswap_vcombine_u32(uint32x2_t __p0,uint32x2_t __p1)5101 __ai uint32x4_t __noswap_vcombine_u32(uint32x2_t __p0, uint32x2_t __p1) {
5102 uint32x4_t __ret;
5103 __ret = __builtin_shufflevector(__p0, __p1, 0, 1, 2, 3);
5104 return __ret;
5105 }
5106 #endif
5107
5108 #ifdef __LITTLE_ENDIAN__
vcombine_u64(uint64x1_t __p0,uint64x1_t __p1)5109 __ai uint64x2_t vcombine_u64(uint64x1_t __p0, uint64x1_t __p1) {
5110 uint64x2_t __ret;
5111 __ret = __builtin_shufflevector(__p0, __p1, 0, 1);
5112 return __ret;
5113 }
5114 #else
vcombine_u64(uint64x1_t __p0,uint64x1_t __p1)5115 __ai uint64x2_t vcombine_u64(uint64x1_t __p0, uint64x1_t __p1) {
5116 uint64x2_t __ret;
5117 __ret = __builtin_shufflevector(__p0, __p1, 0, 1);
5118 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
5119 return __ret;
5120 }
5121 #endif
5122
5123 #ifdef __LITTLE_ENDIAN__
vcombine_u16(uint16x4_t __p0,uint16x4_t __p1)5124 __ai uint16x8_t vcombine_u16(uint16x4_t __p0, uint16x4_t __p1) {
5125 uint16x8_t __ret;
5126 __ret = __builtin_shufflevector(__p0, __p1, 0, 1, 2, 3, 4, 5, 6, 7);
5127 return __ret;
5128 }
5129 #else
vcombine_u16(uint16x4_t __p0,uint16x4_t __p1)5130 __ai uint16x8_t vcombine_u16(uint16x4_t __p0, uint16x4_t __p1) {
5131 uint16x8_t __ret;
5132 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
5133 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
5134 __ret = __builtin_shufflevector(__rev0, __rev1, 0, 1, 2, 3, 4, 5, 6, 7);
5135 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
5136 return __ret;
5137 }
__noswap_vcombine_u16(uint16x4_t __p0,uint16x4_t __p1)5138 __ai uint16x8_t __noswap_vcombine_u16(uint16x4_t __p0, uint16x4_t __p1) {
5139 uint16x8_t __ret;
5140 __ret = __builtin_shufflevector(__p0, __p1, 0, 1, 2, 3, 4, 5, 6, 7);
5141 return __ret;
5142 }
5143 #endif
5144
5145 #ifdef __LITTLE_ENDIAN__
vcombine_s8(int8x8_t __p0,int8x8_t __p1)5146 __ai int8x16_t vcombine_s8(int8x8_t __p0, int8x8_t __p1) {
5147 int8x16_t __ret;
5148 __ret = __builtin_shufflevector(__p0, __p1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15);
5149 return __ret;
5150 }
5151 #else
vcombine_s8(int8x8_t __p0,int8x8_t __p1)5152 __ai int8x16_t vcombine_s8(int8x8_t __p0, int8x8_t __p1) {
5153 int8x16_t __ret;
5154 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
5155 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
5156 __ret = __builtin_shufflevector(__rev0, __rev1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15);
5157 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
5158 return __ret;
5159 }
__noswap_vcombine_s8(int8x8_t __p0,int8x8_t __p1)5160 __ai int8x16_t __noswap_vcombine_s8(int8x8_t __p0, int8x8_t __p1) {
5161 int8x16_t __ret;
5162 __ret = __builtin_shufflevector(__p0, __p1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15);
5163 return __ret;
5164 }
5165 #endif
5166
5167 #ifdef __LITTLE_ENDIAN__
vcombine_f32(float32x2_t __p0,float32x2_t __p1)5168 __ai float32x4_t vcombine_f32(float32x2_t __p0, float32x2_t __p1) {
5169 float32x4_t __ret;
5170 __ret = __builtin_shufflevector(__p0, __p1, 0, 1, 2, 3);
5171 return __ret;
5172 }
5173 #else
vcombine_f32(float32x2_t __p0,float32x2_t __p1)5174 __ai float32x4_t vcombine_f32(float32x2_t __p0, float32x2_t __p1) {
5175 float32x4_t __ret;
5176 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
5177 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
5178 __ret = __builtin_shufflevector(__rev0, __rev1, 0, 1, 2, 3);
5179 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
5180 return __ret;
5181 }
__noswap_vcombine_f32(float32x2_t __p0,float32x2_t __p1)5182 __ai float32x4_t __noswap_vcombine_f32(float32x2_t __p0, float32x2_t __p1) {
5183 float32x4_t __ret;
5184 __ret = __builtin_shufflevector(__p0, __p1, 0, 1, 2, 3);
5185 return __ret;
5186 }
5187 #endif
5188
5189 #ifdef __LITTLE_ENDIAN__
vcombine_f16(float16x4_t __p0,float16x4_t __p1)5190 __ai float16x8_t vcombine_f16(float16x4_t __p0, float16x4_t __p1) {
5191 float16x8_t __ret;
5192 __ret = __builtin_shufflevector(__p0, __p1, 0, 1, 2, 3, 4, 5, 6, 7);
5193 return __ret;
5194 }
5195 #else
vcombine_f16(float16x4_t __p0,float16x4_t __p1)5196 __ai float16x8_t vcombine_f16(float16x4_t __p0, float16x4_t __p1) {
5197 float16x8_t __ret;
5198 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
5199 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
5200 __ret = __builtin_shufflevector(__rev0, __rev1, 0, 1, 2, 3, 4, 5, 6, 7);
5201 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
5202 return __ret;
5203 }
__noswap_vcombine_f16(float16x4_t __p0,float16x4_t __p1)5204 __ai float16x8_t __noswap_vcombine_f16(float16x4_t __p0, float16x4_t __p1) {
5205 float16x8_t __ret;
5206 __ret = __builtin_shufflevector(__p0, __p1, 0, 1, 2, 3, 4, 5, 6, 7);
5207 return __ret;
5208 }
5209 #endif
5210
5211 #ifdef __LITTLE_ENDIAN__
vcombine_s32(int32x2_t __p0,int32x2_t __p1)5212 __ai int32x4_t vcombine_s32(int32x2_t __p0, int32x2_t __p1) {
5213 int32x4_t __ret;
5214 __ret = __builtin_shufflevector(__p0, __p1, 0, 1, 2, 3);
5215 return __ret;
5216 }
5217 #else
vcombine_s32(int32x2_t __p0,int32x2_t __p1)5218 __ai int32x4_t vcombine_s32(int32x2_t __p0, int32x2_t __p1) {
5219 int32x4_t __ret;
5220 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
5221 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
5222 __ret = __builtin_shufflevector(__rev0, __rev1, 0, 1, 2, 3);
5223 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
5224 return __ret;
5225 }
__noswap_vcombine_s32(int32x2_t __p0,int32x2_t __p1)5226 __ai int32x4_t __noswap_vcombine_s32(int32x2_t __p0, int32x2_t __p1) {
5227 int32x4_t __ret;
5228 __ret = __builtin_shufflevector(__p0, __p1, 0, 1, 2, 3);
5229 return __ret;
5230 }
5231 #endif
5232
5233 #ifdef __LITTLE_ENDIAN__
vcombine_s64(int64x1_t __p0,int64x1_t __p1)5234 __ai int64x2_t vcombine_s64(int64x1_t __p0, int64x1_t __p1) {
5235 int64x2_t __ret;
5236 __ret = __builtin_shufflevector(__p0, __p1, 0, 1);
5237 return __ret;
5238 }
5239 #else
vcombine_s64(int64x1_t __p0,int64x1_t __p1)5240 __ai int64x2_t vcombine_s64(int64x1_t __p0, int64x1_t __p1) {
5241 int64x2_t __ret;
5242 __ret = __builtin_shufflevector(__p0, __p1, 0, 1);
5243 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
5244 return __ret;
5245 }
5246 #endif
5247
5248 #ifdef __LITTLE_ENDIAN__
vcombine_s16(int16x4_t __p0,int16x4_t __p1)5249 __ai int16x8_t vcombine_s16(int16x4_t __p0, int16x4_t __p1) {
5250 int16x8_t __ret;
5251 __ret = __builtin_shufflevector(__p0, __p1, 0, 1, 2, 3, 4, 5, 6, 7);
5252 return __ret;
5253 }
5254 #else
vcombine_s16(int16x4_t __p0,int16x4_t __p1)5255 __ai int16x8_t vcombine_s16(int16x4_t __p0, int16x4_t __p1) {
5256 int16x8_t __ret;
5257 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
5258 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
5259 __ret = __builtin_shufflevector(__rev0, __rev1, 0, 1, 2, 3, 4, 5, 6, 7);
5260 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
5261 return __ret;
5262 }
__noswap_vcombine_s16(int16x4_t __p0,int16x4_t __p1)5263 __ai int16x8_t __noswap_vcombine_s16(int16x4_t __p0, int16x4_t __p1) {
5264 int16x8_t __ret;
5265 __ret = __builtin_shufflevector(__p0, __p1, 0, 1, 2, 3, 4, 5, 6, 7);
5266 return __ret;
5267 }
5268 #endif
5269
5270 #define vcreate_p8(__p0) __extension__ ({ \
5271 poly8x8_t __ret; \
5272 uint64_t __promote = __p0; \
5273 __ret = (poly8x8_t)(__promote); \
5274 __ret; \
5275 })
5276 #define vcreate_p16(__p0) __extension__ ({ \
5277 poly16x4_t __ret; \
5278 uint64_t __promote = __p0; \
5279 __ret = (poly16x4_t)(__promote); \
5280 __ret; \
5281 })
5282 #define vcreate_u8(__p0) __extension__ ({ \
5283 uint8x8_t __ret; \
5284 uint64_t __promote = __p0; \
5285 __ret = (uint8x8_t)(__promote); \
5286 __ret; \
5287 })
5288 #define vcreate_u32(__p0) __extension__ ({ \
5289 uint32x2_t __ret; \
5290 uint64_t __promote = __p0; \
5291 __ret = (uint32x2_t)(__promote); \
5292 __ret; \
5293 })
5294 #define vcreate_u64(__p0) __extension__ ({ \
5295 uint64x1_t __ret; \
5296 uint64_t __promote = __p0; \
5297 __ret = (uint64x1_t)(__promote); \
5298 __ret; \
5299 })
5300 #define vcreate_u16(__p0) __extension__ ({ \
5301 uint16x4_t __ret; \
5302 uint64_t __promote = __p0; \
5303 __ret = (uint16x4_t)(__promote); \
5304 __ret; \
5305 })
5306 #define vcreate_s8(__p0) __extension__ ({ \
5307 int8x8_t __ret; \
5308 uint64_t __promote = __p0; \
5309 __ret = (int8x8_t)(__promote); \
5310 __ret; \
5311 })
5312 #define vcreate_f32(__p0) __extension__ ({ \
5313 float32x2_t __ret; \
5314 uint64_t __promote = __p0; \
5315 __ret = (float32x2_t)(__promote); \
5316 __ret; \
5317 })
5318 #define vcreate_f16(__p0) __extension__ ({ \
5319 float16x4_t __ret; \
5320 uint64_t __promote = __p0; \
5321 __ret = (float16x4_t)(__promote); \
5322 __ret; \
5323 })
5324 #define vcreate_s32(__p0) __extension__ ({ \
5325 int32x2_t __ret; \
5326 uint64_t __promote = __p0; \
5327 __ret = (int32x2_t)(__promote); \
5328 __ret; \
5329 })
5330 #define vcreate_s64(__p0) __extension__ ({ \
5331 int64x1_t __ret; \
5332 uint64_t __promote = __p0; \
5333 __ret = (int64x1_t)(__promote); \
5334 __ret; \
5335 })
5336 #define vcreate_s16(__p0) __extension__ ({ \
5337 int16x4_t __ret; \
5338 uint64_t __promote = __p0; \
5339 __ret = (int16x4_t)(__promote); \
5340 __ret; \
5341 })
5342 #ifdef __LITTLE_ENDIAN__
vcvtq_f32_u32(uint32x4_t __p0)5343 __ai float32x4_t vcvtq_f32_u32(uint32x4_t __p0) {
5344 float32x4_t __ret;
5345 __ret = (float32x4_t) __builtin_neon_vcvtq_f32_v((int8x16_t)__p0, 50);
5346 return __ret;
5347 }
5348 #else
vcvtq_f32_u32(uint32x4_t __p0)5349 __ai float32x4_t vcvtq_f32_u32(uint32x4_t __p0) {
5350 float32x4_t __ret;
5351 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
5352 __ret = (float32x4_t) __builtin_neon_vcvtq_f32_v((int8x16_t)__rev0, 50);
5353 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
5354 return __ret;
5355 }
5356 #endif
5357
5358 #ifdef __LITTLE_ENDIAN__
vcvtq_f32_s32(int32x4_t __p0)5359 __ai float32x4_t vcvtq_f32_s32(int32x4_t __p0) {
5360 float32x4_t __ret;
5361 __ret = (float32x4_t) __builtin_neon_vcvtq_f32_v((int8x16_t)__p0, 34);
5362 return __ret;
5363 }
5364 #else
vcvtq_f32_s32(int32x4_t __p0)5365 __ai float32x4_t vcvtq_f32_s32(int32x4_t __p0) {
5366 float32x4_t __ret;
5367 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
5368 __ret = (float32x4_t) __builtin_neon_vcvtq_f32_v((int8x16_t)__rev0, 34);
5369 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
5370 return __ret;
5371 }
5372 #endif
5373
5374 #ifdef __LITTLE_ENDIAN__
vcvt_f32_u32(uint32x2_t __p0)5375 __ai float32x2_t vcvt_f32_u32(uint32x2_t __p0) {
5376 float32x2_t __ret;
5377 __ret = (float32x2_t) __builtin_neon_vcvt_f32_v((int8x8_t)__p0, 18);
5378 return __ret;
5379 }
5380 #else
vcvt_f32_u32(uint32x2_t __p0)5381 __ai float32x2_t vcvt_f32_u32(uint32x2_t __p0) {
5382 float32x2_t __ret;
5383 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
5384 __ret = (float32x2_t) __builtin_neon_vcvt_f32_v((int8x8_t)__rev0, 18);
5385 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
5386 return __ret;
5387 }
5388 #endif
5389
5390 #ifdef __LITTLE_ENDIAN__
vcvt_f32_s32(int32x2_t __p0)5391 __ai float32x2_t vcvt_f32_s32(int32x2_t __p0) {
5392 float32x2_t __ret;
5393 __ret = (float32x2_t) __builtin_neon_vcvt_f32_v((int8x8_t)__p0, 2);
5394 return __ret;
5395 }
5396 #else
vcvt_f32_s32(int32x2_t __p0)5397 __ai float32x2_t vcvt_f32_s32(int32x2_t __p0) {
5398 float32x2_t __ret;
5399 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
5400 __ret = (float32x2_t) __builtin_neon_vcvt_f32_v((int8x8_t)__rev0, 2);
5401 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
5402 return __ret;
5403 }
5404 #endif
5405
5406 #ifdef __LITTLE_ENDIAN__
5407 #define vcvtq_n_f32_u32(__p0, __p1) __extension__ ({ \
5408 float32x4_t __ret; \
5409 uint32x4_t __s0 = __p0; \
5410 __ret = (float32x4_t) __builtin_neon_vcvtq_n_f32_v((int8x16_t)__s0, __p1, 50); \
5411 __ret; \
5412 })
5413 #else
5414 #define vcvtq_n_f32_u32(__p0, __p1) __extension__ ({ \
5415 float32x4_t __ret; \
5416 uint32x4_t __s0 = __p0; \
5417 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
5418 __ret = (float32x4_t) __builtin_neon_vcvtq_n_f32_v((int8x16_t)__rev0, __p1, 50); \
5419 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
5420 __ret; \
5421 })
5422 #endif
5423
5424 #ifdef __LITTLE_ENDIAN__
5425 #define vcvtq_n_f32_s32(__p0, __p1) __extension__ ({ \
5426 float32x4_t __ret; \
5427 int32x4_t __s0 = __p0; \
5428 __ret = (float32x4_t) __builtin_neon_vcvtq_n_f32_v((int8x16_t)__s0, __p1, 34); \
5429 __ret; \
5430 })
5431 #else
5432 #define vcvtq_n_f32_s32(__p0, __p1) __extension__ ({ \
5433 float32x4_t __ret; \
5434 int32x4_t __s0 = __p0; \
5435 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
5436 __ret = (float32x4_t) __builtin_neon_vcvtq_n_f32_v((int8x16_t)__rev0, __p1, 34); \
5437 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
5438 __ret; \
5439 })
5440 #endif
5441
5442 #ifdef __LITTLE_ENDIAN__
5443 #define vcvt_n_f32_u32(__p0, __p1) __extension__ ({ \
5444 float32x2_t __ret; \
5445 uint32x2_t __s0 = __p0; \
5446 __ret = (float32x2_t) __builtin_neon_vcvt_n_f32_v((int8x8_t)__s0, __p1, 18); \
5447 __ret; \
5448 })
5449 #else
5450 #define vcvt_n_f32_u32(__p0, __p1) __extension__ ({ \
5451 float32x2_t __ret; \
5452 uint32x2_t __s0 = __p0; \
5453 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
5454 __ret = (float32x2_t) __builtin_neon_vcvt_n_f32_v((int8x8_t)__rev0, __p1, 18); \
5455 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
5456 __ret; \
5457 })
5458 #endif
5459
5460 #ifdef __LITTLE_ENDIAN__
5461 #define vcvt_n_f32_s32(__p0, __p1) __extension__ ({ \
5462 float32x2_t __ret; \
5463 int32x2_t __s0 = __p0; \
5464 __ret = (float32x2_t) __builtin_neon_vcvt_n_f32_v((int8x8_t)__s0, __p1, 2); \
5465 __ret; \
5466 })
5467 #else
5468 #define vcvt_n_f32_s32(__p0, __p1) __extension__ ({ \
5469 float32x2_t __ret; \
5470 int32x2_t __s0 = __p0; \
5471 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
5472 __ret = (float32x2_t) __builtin_neon_vcvt_n_f32_v((int8x8_t)__rev0, __p1, 2); \
5473 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
5474 __ret; \
5475 })
5476 #endif
5477
5478 #ifdef __LITTLE_ENDIAN__
5479 #define vcvtq_n_s32_f32(__p0, __p1) __extension__ ({ \
5480 int32x4_t __ret; \
5481 float32x4_t __s0 = __p0; \
5482 __ret = (int32x4_t) __builtin_neon_vcvtq_n_s32_v((int8x16_t)__s0, __p1, 34); \
5483 __ret; \
5484 })
5485 #else
5486 #define vcvtq_n_s32_f32(__p0, __p1) __extension__ ({ \
5487 int32x4_t __ret; \
5488 float32x4_t __s0 = __p0; \
5489 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
5490 __ret = (int32x4_t) __builtin_neon_vcvtq_n_s32_v((int8x16_t)__rev0, __p1, 34); \
5491 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
5492 __ret; \
5493 })
5494 #endif
5495
5496 #ifdef __LITTLE_ENDIAN__
5497 #define vcvt_n_s32_f32(__p0, __p1) __extension__ ({ \
5498 int32x2_t __ret; \
5499 float32x2_t __s0 = __p0; \
5500 __ret = (int32x2_t) __builtin_neon_vcvt_n_s32_v((int8x8_t)__s0, __p1, 2); \
5501 __ret; \
5502 })
5503 #else
5504 #define vcvt_n_s32_f32(__p0, __p1) __extension__ ({ \
5505 int32x2_t __ret; \
5506 float32x2_t __s0 = __p0; \
5507 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
5508 __ret = (int32x2_t) __builtin_neon_vcvt_n_s32_v((int8x8_t)__rev0, __p1, 2); \
5509 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
5510 __ret; \
5511 })
5512 #endif
5513
5514 #ifdef __LITTLE_ENDIAN__
5515 #define vcvtq_n_u32_f32(__p0, __p1) __extension__ ({ \
5516 uint32x4_t __ret; \
5517 float32x4_t __s0 = __p0; \
5518 __ret = (uint32x4_t) __builtin_neon_vcvtq_n_u32_v((int8x16_t)__s0, __p1, 50); \
5519 __ret; \
5520 })
5521 #else
5522 #define vcvtq_n_u32_f32(__p0, __p1) __extension__ ({ \
5523 uint32x4_t __ret; \
5524 float32x4_t __s0 = __p0; \
5525 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
5526 __ret = (uint32x4_t) __builtin_neon_vcvtq_n_u32_v((int8x16_t)__rev0, __p1, 50); \
5527 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
5528 __ret; \
5529 })
5530 #endif
5531
5532 #ifdef __LITTLE_ENDIAN__
5533 #define vcvt_n_u32_f32(__p0, __p1) __extension__ ({ \
5534 uint32x2_t __ret; \
5535 float32x2_t __s0 = __p0; \
5536 __ret = (uint32x2_t) __builtin_neon_vcvt_n_u32_v((int8x8_t)__s0, __p1, 18); \
5537 __ret; \
5538 })
5539 #else
5540 #define vcvt_n_u32_f32(__p0, __p1) __extension__ ({ \
5541 uint32x2_t __ret; \
5542 float32x2_t __s0 = __p0; \
5543 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
5544 __ret = (uint32x2_t) __builtin_neon_vcvt_n_u32_v((int8x8_t)__rev0, __p1, 18); \
5545 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
5546 __ret; \
5547 })
5548 #endif
5549
5550 #ifdef __LITTLE_ENDIAN__
vcvtq_s32_f32(float32x4_t __p0)5551 __ai int32x4_t vcvtq_s32_f32(float32x4_t __p0) {
5552 int32x4_t __ret;
5553 __ret = (int32x4_t) __builtin_neon_vcvtq_s32_v((int8x16_t)__p0, 34);
5554 return __ret;
5555 }
5556 #else
vcvtq_s32_f32(float32x4_t __p0)5557 __ai int32x4_t vcvtq_s32_f32(float32x4_t __p0) {
5558 int32x4_t __ret;
5559 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
5560 __ret = (int32x4_t) __builtin_neon_vcvtq_s32_v((int8x16_t)__rev0, 34);
5561 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
5562 return __ret;
5563 }
5564 #endif
5565
5566 #ifdef __LITTLE_ENDIAN__
vcvt_s32_f32(float32x2_t __p0)5567 __ai int32x2_t vcvt_s32_f32(float32x2_t __p0) {
5568 int32x2_t __ret;
5569 __ret = (int32x2_t) __builtin_neon_vcvt_s32_v((int8x8_t)__p0, 2);
5570 return __ret;
5571 }
5572 #else
vcvt_s32_f32(float32x2_t __p0)5573 __ai int32x2_t vcvt_s32_f32(float32x2_t __p0) {
5574 int32x2_t __ret;
5575 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
5576 __ret = (int32x2_t) __builtin_neon_vcvt_s32_v((int8x8_t)__rev0, 2);
5577 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
5578 return __ret;
5579 }
5580 #endif
5581
5582 #ifdef __LITTLE_ENDIAN__
vcvtq_u32_f32(float32x4_t __p0)5583 __ai uint32x4_t vcvtq_u32_f32(float32x4_t __p0) {
5584 uint32x4_t __ret;
5585 __ret = (uint32x4_t) __builtin_neon_vcvtq_u32_v((int8x16_t)__p0, 50);
5586 return __ret;
5587 }
5588 #else
vcvtq_u32_f32(float32x4_t __p0)5589 __ai uint32x4_t vcvtq_u32_f32(float32x4_t __p0) {
5590 uint32x4_t __ret;
5591 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
5592 __ret = (uint32x4_t) __builtin_neon_vcvtq_u32_v((int8x16_t)__rev0, 50);
5593 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
5594 return __ret;
5595 }
5596 #endif
5597
5598 #ifdef __LITTLE_ENDIAN__
vcvt_u32_f32(float32x2_t __p0)5599 __ai uint32x2_t vcvt_u32_f32(float32x2_t __p0) {
5600 uint32x2_t __ret;
5601 __ret = (uint32x2_t) __builtin_neon_vcvt_u32_v((int8x8_t)__p0, 18);
5602 return __ret;
5603 }
5604 #else
vcvt_u32_f32(float32x2_t __p0)5605 __ai uint32x2_t vcvt_u32_f32(float32x2_t __p0) {
5606 uint32x2_t __ret;
5607 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
5608 __ret = (uint32x2_t) __builtin_neon_vcvt_u32_v((int8x8_t)__rev0, 18);
5609 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
5610 return __ret;
5611 }
5612 #endif
5613
5614 #ifdef __LITTLE_ENDIAN__
5615 #define vdup_lane_p8(__p0_0, __p1_0) __extension__ ({ \
5616 poly8x8_t __ret_0; \
5617 poly8x8_t __s0_0 = __p0_0; \
5618 __ret_0 = splat_lane_p8(__s0_0, __p1_0); \
5619 __ret_0; \
5620 })
5621 #else
5622 #define vdup_lane_p8(__p0_1, __p1_1) __extension__ ({ \
5623 poly8x8_t __ret_1; \
5624 poly8x8_t __s0_1 = __p0_1; \
5625 poly8x8_t __rev0_1; __rev0_1 = __builtin_shufflevector(__s0_1, __s0_1, 7, 6, 5, 4, 3, 2, 1, 0); \
5626 __ret_1 = __noswap_splat_lane_p8(__rev0_1, __p1_1); \
5627 __ret_1 = __builtin_shufflevector(__ret_1, __ret_1, 7, 6, 5, 4, 3, 2, 1, 0); \
5628 __ret_1; \
5629 })
5630 #endif
5631
5632 #ifdef __LITTLE_ENDIAN__
5633 #define vdup_lane_p16(__p0_2, __p1_2) __extension__ ({ \
5634 poly16x4_t __ret_2; \
5635 poly16x4_t __s0_2 = __p0_2; \
5636 __ret_2 = splat_lane_p16(__s0_2, __p1_2); \
5637 __ret_2; \
5638 })
5639 #else
5640 #define vdup_lane_p16(__p0_3, __p1_3) __extension__ ({ \
5641 poly16x4_t __ret_3; \
5642 poly16x4_t __s0_3 = __p0_3; \
5643 poly16x4_t __rev0_3; __rev0_3 = __builtin_shufflevector(__s0_3, __s0_3, 3, 2, 1, 0); \
5644 __ret_3 = __noswap_splat_lane_p16(__rev0_3, __p1_3); \
5645 __ret_3 = __builtin_shufflevector(__ret_3, __ret_3, 3, 2, 1, 0); \
5646 __ret_3; \
5647 })
5648 #endif
5649
5650 #ifdef __LITTLE_ENDIAN__
5651 #define vdupq_lane_p8(__p0_4, __p1_4) __extension__ ({ \
5652 poly8x16_t __ret_4; \
5653 poly8x8_t __s0_4 = __p0_4; \
5654 __ret_4 = splatq_lane_p8(__s0_4, __p1_4); \
5655 __ret_4; \
5656 })
5657 #else
5658 #define vdupq_lane_p8(__p0_5, __p1_5) __extension__ ({ \
5659 poly8x16_t __ret_5; \
5660 poly8x8_t __s0_5 = __p0_5; \
5661 poly8x8_t __rev0_5; __rev0_5 = __builtin_shufflevector(__s0_5, __s0_5, 7, 6, 5, 4, 3, 2, 1, 0); \
5662 __ret_5 = __noswap_splatq_lane_p8(__rev0_5, __p1_5); \
5663 __ret_5 = __builtin_shufflevector(__ret_5, __ret_5, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
5664 __ret_5; \
5665 })
5666 #endif
5667
5668 #ifdef __LITTLE_ENDIAN__
5669 #define vdupq_lane_p16(__p0_6, __p1_6) __extension__ ({ \
5670 poly16x8_t __ret_6; \
5671 poly16x4_t __s0_6 = __p0_6; \
5672 __ret_6 = splatq_lane_p16(__s0_6, __p1_6); \
5673 __ret_6; \
5674 })
5675 #else
5676 #define vdupq_lane_p16(__p0_7, __p1_7) __extension__ ({ \
5677 poly16x8_t __ret_7; \
5678 poly16x4_t __s0_7 = __p0_7; \
5679 poly16x4_t __rev0_7; __rev0_7 = __builtin_shufflevector(__s0_7, __s0_7, 3, 2, 1, 0); \
5680 __ret_7 = __noswap_splatq_lane_p16(__rev0_7, __p1_7); \
5681 __ret_7 = __builtin_shufflevector(__ret_7, __ret_7, 7, 6, 5, 4, 3, 2, 1, 0); \
5682 __ret_7; \
5683 })
5684 #endif
5685
5686 #ifdef __LITTLE_ENDIAN__
5687 #define vdupq_lane_u8(__p0_8, __p1_8) __extension__ ({ \
5688 uint8x16_t __ret_8; \
5689 uint8x8_t __s0_8 = __p0_8; \
5690 __ret_8 = splatq_lane_u8(__s0_8, __p1_8); \
5691 __ret_8; \
5692 })
5693 #else
5694 #define vdupq_lane_u8(__p0_9, __p1_9) __extension__ ({ \
5695 uint8x16_t __ret_9; \
5696 uint8x8_t __s0_9 = __p0_9; \
5697 uint8x8_t __rev0_9; __rev0_9 = __builtin_shufflevector(__s0_9, __s0_9, 7, 6, 5, 4, 3, 2, 1, 0); \
5698 __ret_9 = __noswap_splatq_lane_u8(__rev0_9, __p1_9); \
5699 __ret_9 = __builtin_shufflevector(__ret_9, __ret_9, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
5700 __ret_9; \
5701 })
5702 #endif
5703
5704 #ifdef __LITTLE_ENDIAN__
5705 #define vdupq_lane_u32(__p0_10, __p1_10) __extension__ ({ \
5706 uint32x4_t __ret_10; \
5707 uint32x2_t __s0_10 = __p0_10; \
5708 __ret_10 = splatq_lane_u32(__s0_10, __p1_10); \
5709 __ret_10; \
5710 })
5711 #else
5712 #define vdupq_lane_u32(__p0_11, __p1_11) __extension__ ({ \
5713 uint32x4_t __ret_11; \
5714 uint32x2_t __s0_11 = __p0_11; \
5715 uint32x2_t __rev0_11; __rev0_11 = __builtin_shufflevector(__s0_11, __s0_11, 1, 0); \
5716 __ret_11 = __noswap_splatq_lane_u32(__rev0_11, __p1_11); \
5717 __ret_11 = __builtin_shufflevector(__ret_11, __ret_11, 3, 2, 1, 0); \
5718 __ret_11; \
5719 })
5720 #endif
5721
5722 #ifdef __LITTLE_ENDIAN__
5723 #define vdupq_lane_u64(__p0_12, __p1_12) __extension__ ({ \
5724 uint64x2_t __ret_12; \
5725 uint64x1_t __s0_12 = __p0_12; \
5726 __ret_12 = splatq_lane_u64(__s0_12, __p1_12); \
5727 __ret_12; \
5728 })
5729 #else
5730 #define vdupq_lane_u64(__p0_13, __p1_13) __extension__ ({ \
5731 uint64x2_t __ret_13; \
5732 uint64x1_t __s0_13 = __p0_13; \
5733 __ret_13 = __noswap_splatq_lane_u64(__s0_13, __p1_13); \
5734 __ret_13 = __builtin_shufflevector(__ret_13, __ret_13, 1, 0); \
5735 __ret_13; \
5736 })
5737 #endif
5738
5739 #ifdef __LITTLE_ENDIAN__
5740 #define vdupq_lane_u16(__p0_14, __p1_14) __extension__ ({ \
5741 uint16x8_t __ret_14; \
5742 uint16x4_t __s0_14 = __p0_14; \
5743 __ret_14 = splatq_lane_u16(__s0_14, __p1_14); \
5744 __ret_14; \
5745 })
5746 #else
5747 #define vdupq_lane_u16(__p0_15, __p1_15) __extension__ ({ \
5748 uint16x8_t __ret_15; \
5749 uint16x4_t __s0_15 = __p0_15; \
5750 uint16x4_t __rev0_15; __rev0_15 = __builtin_shufflevector(__s0_15, __s0_15, 3, 2, 1, 0); \
5751 __ret_15 = __noswap_splatq_lane_u16(__rev0_15, __p1_15); \
5752 __ret_15 = __builtin_shufflevector(__ret_15, __ret_15, 7, 6, 5, 4, 3, 2, 1, 0); \
5753 __ret_15; \
5754 })
5755 #endif
5756
5757 #ifdef __LITTLE_ENDIAN__
5758 #define vdupq_lane_s8(__p0_16, __p1_16) __extension__ ({ \
5759 int8x16_t __ret_16; \
5760 int8x8_t __s0_16 = __p0_16; \
5761 __ret_16 = splatq_lane_s8(__s0_16, __p1_16); \
5762 __ret_16; \
5763 })
5764 #else
5765 #define vdupq_lane_s8(__p0_17, __p1_17) __extension__ ({ \
5766 int8x16_t __ret_17; \
5767 int8x8_t __s0_17 = __p0_17; \
5768 int8x8_t __rev0_17; __rev0_17 = __builtin_shufflevector(__s0_17, __s0_17, 7, 6, 5, 4, 3, 2, 1, 0); \
5769 __ret_17 = __noswap_splatq_lane_s8(__rev0_17, __p1_17); \
5770 __ret_17 = __builtin_shufflevector(__ret_17, __ret_17, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
5771 __ret_17; \
5772 })
5773 #endif
5774
5775 #ifdef __LITTLE_ENDIAN__
5776 #define vdupq_lane_f32(__p0_18, __p1_18) __extension__ ({ \
5777 float32x4_t __ret_18; \
5778 float32x2_t __s0_18 = __p0_18; \
5779 __ret_18 = splatq_lane_f32(__s0_18, __p1_18); \
5780 __ret_18; \
5781 })
5782 #else
5783 #define vdupq_lane_f32(__p0_19, __p1_19) __extension__ ({ \
5784 float32x4_t __ret_19; \
5785 float32x2_t __s0_19 = __p0_19; \
5786 float32x2_t __rev0_19; __rev0_19 = __builtin_shufflevector(__s0_19, __s0_19, 1, 0); \
5787 __ret_19 = __noswap_splatq_lane_f32(__rev0_19, __p1_19); \
5788 __ret_19 = __builtin_shufflevector(__ret_19, __ret_19, 3, 2, 1, 0); \
5789 __ret_19; \
5790 })
5791 #endif
5792
5793 #ifdef __LITTLE_ENDIAN__
5794 #define vdupq_lane_f16(__p0_20, __p1_20) __extension__ ({ \
5795 float16x8_t __ret_20; \
5796 float16x4_t __s0_20 = __p0_20; \
5797 __ret_20 = splatq_lane_f16(__s0_20, __p1_20); \
5798 __ret_20; \
5799 })
5800 #else
5801 #define vdupq_lane_f16(__p0_21, __p1_21) __extension__ ({ \
5802 float16x8_t __ret_21; \
5803 float16x4_t __s0_21 = __p0_21; \
5804 float16x4_t __rev0_21; __rev0_21 = __builtin_shufflevector(__s0_21, __s0_21, 3, 2, 1, 0); \
5805 __ret_21 = __noswap_splatq_lane_f16(__rev0_21, __p1_21); \
5806 __ret_21 = __builtin_shufflevector(__ret_21, __ret_21, 7, 6, 5, 4, 3, 2, 1, 0); \
5807 __ret_21; \
5808 })
5809 #endif
5810
5811 #ifdef __LITTLE_ENDIAN__
5812 #define vdupq_lane_s32(__p0_22, __p1_22) __extension__ ({ \
5813 int32x4_t __ret_22; \
5814 int32x2_t __s0_22 = __p0_22; \
5815 __ret_22 = splatq_lane_s32(__s0_22, __p1_22); \
5816 __ret_22; \
5817 })
5818 #else
5819 #define vdupq_lane_s32(__p0_23, __p1_23) __extension__ ({ \
5820 int32x4_t __ret_23; \
5821 int32x2_t __s0_23 = __p0_23; \
5822 int32x2_t __rev0_23; __rev0_23 = __builtin_shufflevector(__s0_23, __s0_23, 1, 0); \
5823 __ret_23 = __noswap_splatq_lane_s32(__rev0_23, __p1_23); \
5824 __ret_23 = __builtin_shufflevector(__ret_23, __ret_23, 3, 2, 1, 0); \
5825 __ret_23; \
5826 })
5827 #endif
5828
5829 #ifdef __LITTLE_ENDIAN__
5830 #define vdupq_lane_s64(__p0_24, __p1_24) __extension__ ({ \
5831 int64x2_t __ret_24; \
5832 int64x1_t __s0_24 = __p0_24; \
5833 __ret_24 = splatq_lane_s64(__s0_24, __p1_24); \
5834 __ret_24; \
5835 })
5836 #else
5837 #define vdupq_lane_s64(__p0_25, __p1_25) __extension__ ({ \
5838 int64x2_t __ret_25; \
5839 int64x1_t __s0_25 = __p0_25; \
5840 __ret_25 = __noswap_splatq_lane_s64(__s0_25, __p1_25); \
5841 __ret_25 = __builtin_shufflevector(__ret_25, __ret_25, 1, 0); \
5842 __ret_25; \
5843 })
5844 #endif
5845
5846 #ifdef __LITTLE_ENDIAN__
5847 #define vdupq_lane_s16(__p0_26, __p1_26) __extension__ ({ \
5848 int16x8_t __ret_26; \
5849 int16x4_t __s0_26 = __p0_26; \
5850 __ret_26 = splatq_lane_s16(__s0_26, __p1_26); \
5851 __ret_26; \
5852 })
5853 #else
5854 #define vdupq_lane_s16(__p0_27, __p1_27) __extension__ ({ \
5855 int16x8_t __ret_27; \
5856 int16x4_t __s0_27 = __p0_27; \
5857 int16x4_t __rev0_27; __rev0_27 = __builtin_shufflevector(__s0_27, __s0_27, 3, 2, 1, 0); \
5858 __ret_27 = __noswap_splatq_lane_s16(__rev0_27, __p1_27); \
5859 __ret_27 = __builtin_shufflevector(__ret_27, __ret_27, 7, 6, 5, 4, 3, 2, 1, 0); \
5860 __ret_27; \
5861 })
5862 #endif
5863
5864 #ifdef __LITTLE_ENDIAN__
5865 #define vdup_lane_u8(__p0_28, __p1_28) __extension__ ({ \
5866 uint8x8_t __ret_28; \
5867 uint8x8_t __s0_28 = __p0_28; \
5868 __ret_28 = splat_lane_u8(__s0_28, __p1_28); \
5869 __ret_28; \
5870 })
5871 #else
5872 #define vdup_lane_u8(__p0_29, __p1_29) __extension__ ({ \
5873 uint8x8_t __ret_29; \
5874 uint8x8_t __s0_29 = __p0_29; \
5875 uint8x8_t __rev0_29; __rev0_29 = __builtin_shufflevector(__s0_29, __s0_29, 7, 6, 5, 4, 3, 2, 1, 0); \
5876 __ret_29 = __noswap_splat_lane_u8(__rev0_29, __p1_29); \
5877 __ret_29 = __builtin_shufflevector(__ret_29, __ret_29, 7, 6, 5, 4, 3, 2, 1, 0); \
5878 __ret_29; \
5879 })
5880 #endif
5881
5882 #ifdef __LITTLE_ENDIAN__
5883 #define vdup_lane_u32(__p0_30, __p1_30) __extension__ ({ \
5884 uint32x2_t __ret_30; \
5885 uint32x2_t __s0_30 = __p0_30; \
5886 __ret_30 = splat_lane_u32(__s0_30, __p1_30); \
5887 __ret_30; \
5888 })
5889 #else
5890 #define vdup_lane_u32(__p0_31, __p1_31) __extension__ ({ \
5891 uint32x2_t __ret_31; \
5892 uint32x2_t __s0_31 = __p0_31; \
5893 uint32x2_t __rev0_31; __rev0_31 = __builtin_shufflevector(__s0_31, __s0_31, 1, 0); \
5894 __ret_31 = __noswap_splat_lane_u32(__rev0_31, __p1_31); \
5895 __ret_31 = __builtin_shufflevector(__ret_31, __ret_31, 1, 0); \
5896 __ret_31; \
5897 })
5898 #endif
5899
5900 #define vdup_lane_u64(__p0_32, __p1_32) __extension__ ({ \
5901 uint64x1_t __ret_32; \
5902 uint64x1_t __s0_32 = __p0_32; \
5903 __ret_32 = splat_lane_u64(__s0_32, __p1_32); \
5904 __ret_32; \
5905 })
5906 #ifdef __LITTLE_ENDIAN__
5907 #define vdup_lane_u16(__p0_33, __p1_33) __extension__ ({ \
5908 uint16x4_t __ret_33; \
5909 uint16x4_t __s0_33 = __p0_33; \
5910 __ret_33 = splat_lane_u16(__s0_33, __p1_33); \
5911 __ret_33; \
5912 })
5913 #else
5914 #define vdup_lane_u16(__p0_34, __p1_34) __extension__ ({ \
5915 uint16x4_t __ret_34; \
5916 uint16x4_t __s0_34 = __p0_34; \
5917 uint16x4_t __rev0_34; __rev0_34 = __builtin_shufflevector(__s0_34, __s0_34, 3, 2, 1, 0); \
5918 __ret_34 = __noswap_splat_lane_u16(__rev0_34, __p1_34); \
5919 __ret_34 = __builtin_shufflevector(__ret_34, __ret_34, 3, 2, 1, 0); \
5920 __ret_34; \
5921 })
5922 #endif
5923
5924 #ifdef __LITTLE_ENDIAN__
5925 #define vdup_lane_s8(__p0_35, __p1_35) __extension__ ({ \
5926 int8x8_t __ret_35; \
5927 int8x8_t __s0_35 = __p0_35; \
5928 __ret_35 = splat_lane_s8(__s0_35, __p1_35); \
5929 __ret_35; \
5930 })
5931 #else
5932 #define vdup_lane_s8(__p0_36, __p1_36) __extension__ ({ \
5933 int8x8_t __ret_36; \
5934 int8x8_t __s0_36 = __p0_36; \
5935 int8x8_t __rev0_36; __rev0_36 = __builtin_shufflevector(__s0_36, __s0_36, 7, 6, 5, 4, 3, 2, 1, 0); \
5936 __ret_36 = __noswap_splat_lane_s8(__rev0_36, __p1_36); \
5937 __ret_36 = __builtin_shufflevector(__ret_36, __ret_36, 7, 6, 5, 4, 3, 2, 1, 0); \
5938 __ret_36; \
5939 })
5940 #endif
5941
5942 #ifdef __LITTLE_ENDIAN__
5943 #define vdup_lane_f32(__p0_37, __p1_37) __extension__ ({ \
5944 float32x2_t __ret_37; \
5945 float32x2_t __s0_37 = __p0_37; \
5946 __ret_37 = splat_lane_f32(__s0_37, __p1_37); \
5947 __ret_37; \
5948 })
5949 #else
5950 #define vdup_lane_f32(__p0_38, __p1_38) __extension__ ({ \
5951 float32x2_t __ret_38; \
5952 float32x2_t __s0_38 = __p0_38; \
5953 float32x2_t __rev0_38; __rev0_38 = __builtin_shufflevector(__s0_38, __s0_38, 1, 0); \
5954 __ret_38 = __noswap_splat_lane_f32(__rev0_38, __p1_38); \
5955 __ret_38 = __builtin_shufflevector(__ret_38, __ret_38, 1, 0); \
5956 __ret_38; \
5957 })
5958 #endif
5959
5960 #ifdef __LITTLE_ENDIAN__
5961 #define vdup_lane_f16(__p0_39, __p1_39) __extension__ ({ \
5962 float16x4_t __ret_39; \
5963 float16x4_t __s0_39 = __p0_39; \
5964 __ret_39 = splat_lane_f16(__s0_39, __p1_39); \
5965 __ret_39; \
5966 })
5967 #else
5968 #define vdup_lane_f16(__p0_40, __p1_40) __extension__ ({ \
5969 float16x4_t __ret_40; \
5970 float16x4_t __s0_40 = __p0_40; \
5971 float16x4_t __rev0_40; __rev0_40 = __builtin_shufflevector(__s0_40, __s0_40, 3, 2, 1, 0); \
5972 __ret_40 = __noswap_splat_lane_f16(__rev0_40, __p1_40); \
5973 __ret_40 = __builtin_shufflevector(__ret_40, __ret_40, 3, 2, 1, 0); \
5974 __ret_40; \
5975 })
5976 #endif
5977
5978 #ifdef __LITTLE_ENDIAN__
5979 #define vdup_lane_s32(__p0_41, __p1_41) __extension__ ({ \
5980 int32x2_t __ret_41; \
5981 int32x2_t __s0_41 = __p0_41; \
5982 __ret_41 = splat_lane_s32(__s0_41, __p1_41); \
5983 __ret_41; \
5984 })
5985 #else
5986 #define vdup_lane_s32(__p0_42, __p1_42) __extension__ ({ \
5987 int32x2_t __ret_42; \
5988 int32x2_t __s0_42 = __p0_42; \
5989 int32x2_t __rev0_42; __rev0_42 = __builtin_shufflevector(__s0_42, __s0_42, 1, 0); \
5990 __ret_42 = __noswap_splat_lane_s32(__rev0_42, __p1_42); \
5991 __ret_42 = __builtin_shufflevector(__ret_42, __ret_42, 1, 0); \
5992 __ret_42; \
5993 })
5994 #endif
5995
5996 #define vdup_lane_s64(__p0_43, __p1_43) __extension__ ({ \
5997 int64x1_t __ret_43; \
5998 int64x1_t __s0_43 = __p0_43; \
5999 __ret_43 = splat_lane_s64(__s0_43, __p1_43); \
6000 __ret_43; \
6001 })
6002 #ifdef __LITTLE_ENDIAN__
6003 #define vdup_lane_s16(__p0_44, __p1_44) __extension__ ({ \
6004 int16x4_t __ret_44; \
6005 int16x4_t __s0_44 = __p0_44; \
6006 __ret_44 = splat_lane_s16(__s0_44, __p1_44); \
6007 __ret_44; \
6008 })
6009 #else
6010 #define vdup_lane_s16(__p0_45, __p1_45) __extension__ ({ \
6011 int16x4_t __ret_45; \
6012 int16x4_t __s0_45 = __p0_45; \
6013 int16x4_t __rev0_45; __rev0_45 = __builtin_shufflevector(__s0_45, __s0_45, 3, 2, 1, 0); \
6014 __ret_45 = __noswap_splat_lane_s16(__rev0_45, __p1_45); \
6015 __ret_45 = __builtin_shufflevector(__ret_45, __ret_45, 3, 2, 1, 0); \
6016 __ret_45; \
6017 })
6018 #endif
6019
6020 #ifdef __LITTLE_ENDIAN__
vdup_n_p8(poly8_t __p0)6021 __ai poly8x8_t vdup_n_p8(poly8_t __p0) {
6022 poly8x8_t __ret;
6023 __ret = (poly8x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
6024 return __ret;
6025 }
6026 #else
vdup_n_p8(poly8_t __p0)6027 __ai poly8x8_t vdup_n_p8(poly8_t __p0) {
6028 poly8x8_t __ret;
6029 __ret = (poly8x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
6030 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
6031 return __ret;
6032 }
6033 #endif
6034
6035 #ifdef __LITTLE_ENDIAN__
vdup_n_p16(poly16_t __p0)6036 __ai poly16x4_t vdup_n_p16(poly16_t __p0) {
6037 poly16x4_t __ret;
6038 __ret = (poly16x4_t) {__p0, __p0, __p0, __p0};
6039 return __ret;
6040 }
6041 #else
vdup_n_p16(poly16_t __p0)6042 __ai poly16x4_t vdup_n_p16(poly16_t __p0) {
6043 poly16x4_t __ret;
6044 __ret = (poly16x4_t) {__p0, __p0, __p0, __p0};
6045 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
6046 return __ret;
6047 }
6048 #endif
6049
6050 #ifdef __LITTLE_ENDIAN__
vdupq_n_p8(poly8_t __p0)6051 __ai poly8x16_t vdupq_n_p8(poly8_t __p0) {
6052 poly8x16_t __ret;
6053 __ret = (poly8x16_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
6054 return __ret;
6055 }
6056 #else
vdupq_n_p8(poly8_t __p0)6057 __ai poly8x16_t vdupq_n_p8(poly8_t __p0) {
6058 poly8x16_t __ret;
6059 __ret = (poly8x16_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
6060 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
6061 return __ret;
6062 }
6063 #endif
6064
6065 #ifdef __LITTLE_ENDIAN__
vdupq_n_p16(poly16_t __p0)6066 __ai poly16x8_t vdupq_n_p16(poly16_t __p0) {
6067 poly16x8_t __ret;
6068 __ret = (poly16x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
6069 return __ret;
6070 }
6071 #else
vdupq_n_p16(poly16_t __p0)6072 __ai poly16x8_t vdupq_n_p16(poly16_t __p0) {
6073 poly16x8_t __ret;
6074 __ret = (poly16x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
6075 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
6076 return __ret;
6077 }
6078 #endif
6079
6080 #ifdef __LITTLE_ENDIAN__
vdupq_n_u8(uint8_t __p0)6081 __ai uint8x16_t vdupq_n_u8(uint8_t __p0) {
6082 uint8x16_t __ret;
6083 __ret = (uint8x16_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
6084 return __ret;
6085 }
6086 #else
vdupq_n_u8(uint8_t __p0)6087 __ai uint8x16_t vdupq_n_u8(uint8_t __p0) {
6088 uint8x16_t __ret;
6089 __ret = (uint8x16_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
6090 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
6091 return __ret;
6092 }
6093 #endif
6094
6095 #ifdef __LITTLE_ENDIAN__
vdupq_n_u32(uint32_t __p0)6096 __ai uint32x4_t vdupq_n_u32(uint32_t __p0) {
6097 uint32x4_t __ret;
6098 __ret = (uint32x4_t) {__p0, __p0, __p0, __p0};
6099 return __ret;
6100 }
6101 #else
vdupq_n_u32(uint32_t __p0)6102 __ai uint32x4_t vdupq_n_u32(uint32_t __p0) {
6103 uint32x4_t __ret;
6104 __ret = (uint32x4_t) {__p0, __p0, __p0, __p0};
6105 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
6106 return __ret;
6107 }
6108 #endif
6109
6110 #ifdef __LITTLE_ENDIAN__
vdupq_n_u64(uint64_t __p0)6111 __ai uint64x2_t vdupq_n_u64(uint64_t __p0) {
6112 uint64x2_t __ret;
6113 __ret = (uint64x2_t) {__p0, __p0};
6114 return __ret;
6115 }
6116 #else
vdupq_n_u64(uint64_t __p0)6117 __ai uint64x2_t vdupq_n_u64(uint64_t __p0) {
6118 uint64x2_t __ret;
6119 __ret = (uint64x2_t) {__p0, __p0};
6120 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
6121 return __ret;
6122 }
6123 #endif
6124
6125 #ifdef __LITTLE_ENDIAN__
vdupq_n_u16(uint16_t __p0)6126 __ai uint16x8_t vdupq_n_u16(uint16_t __p0) {
6127 uint16x8_t __ret;
6128 __ret = (uint16x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
6129 return __ret;
6130 }
6131 #else
vdupq_n_u16(uint16_t __p0)6132 __ai uint16x8_t vdupq_n_u16(uint16_t __p0) {
6133 uint16x8_t __ret;
6134 __ret = (uint16x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
6135 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
6136 return __ret;
6137 }
6138 #endif
6139
6140 #ifdef __LITTLE_ENDIAN__
vdupq_n_s8(int8_t __p0)6141 __ai int8x16_t vdupq_n_s8(int8_t __p0) {
6142 int8x16_t __ret;
6143 __ret = (int8x16_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
6144 return __ret;
6145 }
6146 #else
vdupq_n_s8(int8_t __p0)6147 __ai int8x16_t vdupq_n_s8(int8_t __p0) {
6148 int8x16_t __ret;
6149 __ret = (int8x16_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
6150 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
6151 return __ret;
6152 }
6153 #endif
6154
6155 #ifdef __LITTLE_ENDIAN__
vdupq_n_f32(float32_t __p0)6156 __ai float32x4_t vdupq_n_f32(float32_t __p0) {
6157 float32x4_t __ret;
6158 __ret = (float32x4_t) {__p0, __p0, __p0, __p0};
6159 return __ret;
6160 }
6161 #else
vdupq_n_f32(float32_t __p0)6162 __ai float32x4_t vdupq_n_f32(float32_t __p0) {
6163 float32x4_t __ret;
6164 __ret = (float32x4_t) {__p0, __p0, __p0, __p0};
6165 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
6166 return __ret;
6167 }
6168 #endif
6169
6170 #ifdef __LITTLE_ENDIAN__
6171 #define vdupq_n_f16(__p0) __extension__ ({ \
6172 float16x8_t __ret; \
6173 float16_t __s0 = __p0; \
6174 __ret = (float16x8_t) {__s0, __s0, __s0, __s0, __s0, __s0, __s0, __s0}; \
6175 __ret; \
6176 })
6177 #else
6178 #define vdupq_n_f16(__p0) __extension__ ({ \
6179 float16x8_t __ret; \
6180 float16_t __s0 = __p0; \
6181 __ret = (float16x8_t) {__s0, __s0, __s0, __s0, __s0, __s0, __s0, __s0}; \
6182 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
6183 __ret; \
6184 })
6185 #endif
6186
6187 #ifdef __LITTLE_ENDIAN__
vdupq_n_s32(int32_t __p0)6188 __ai int32x4_t vdupq_n_s32(int32_t __p0) {
6189 int32x4_t __ret;
6190 __ret = (int32x4_t) {__p0, __p0, __p0, __p0};
6191 return __ret;
6192 }
6193 #else
vdupq_n_s32(int32_t __p0)6194 __ai int32x4_t vdupq_n_s32(int32_t __p0) {
6195 int32x4_t __ret;
6196 __ret = (int32x4_t) {__p0, __p0, __p0, __p0};
6197 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
6198 return __ret;
6199 }
6200 #endif
6201
6202 #ifdef __LITTLE_ENDIAN__
vdupq_n_s64(int64_t __p0)6203 __ai int64x2_t vdupq_n_s64(int64_t __p0) {
6204 int64x2_t __ret;
6205 __ret = (int64x2_t) {__p0, __p0};
6206 return __ret;
6207 }
6208 #else
vdupq_n_s64(int64_t __p0)6209 __ai int64x2_t vdupq_n_s64(int64_t __p0) {
6210 int64x2_t __ret;
6211 __ret = (int64x2_t) {__p0, __p0};
6212 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
6213 return __ret;
6214 }
6215 #endif
6216
6217 #ifdef __LITTLE_ENDIAN__
vdupq_n_s16(int16_t __p0)6218 __ai int16x8_t vdupq_n_s16(int16_t __p0) {
6219 int16x8_t __ret;
6220 __ret = (int16x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
6221 return __ret;
6222 }
6223 #else
vdupq_n_s16(int16_t __p0)6224 __ai int16x8_t vdupq_n_s16(int16_t __p0) {
6225 int16x8_t __ret;
6226 __ret = (int16x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
6227 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
6228 return __ret;
6229 }
6230 #endif
6231
6232 #ifdef __LITTLE_ENDIAN__
vdup_n_u8(uint8_t __p0)6233 __ai uint8x8_t vdup_n_u8(uint8_t __p0) {
6234 uint8x8_t __ret;
6235 __ret = (uint8x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
6236 return __ret;
6237 }
6238 #else
vdup_n_u8(uint8_t __p0)6239 __ai uint8x8_t vdup_n_u8(uint8_t __p0) {
6240 uint8x8_t __ret;
6241 __ret = (uint8x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
6242 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
6243 return __ret;
6244 }
6245 #endif
6246
6247 #ifdef __LITTLE_ENDIAN__
vdup_n_u32(uint32_t __p0)6248 __ai uint32x2_t vdup_n_u32(uint32_t __p0) {
6249 uint32x2_t __ret;
6250 __ret = (uint32x2_t) {__p0, __p0};
6251 return __ret;
6252 }
6253 #else
vdup_n_u32(uint32_t __p0)6254 __ai uint32x2_t vdup_n_u32(uint32_t __p0) {
6255 uint32x2_t __ret;
6256 __ret = (uint32x2_t) {__p0, __p0};
6257 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
6258 return __ret;
6259 }
6260 #endif
6261
vdup_n_u64(uint64_t __p0)6262 __ai uint64x1_t vdup_n_u64(uint64_t __p0) {
6263 uint64x1_t __ret;
6264 __ret = (uint64x1_t) {__p0};
6265 return __ret;
6266 }
6267 #ifdef __LITTLE_ENDIAN__
vdup_n_u16(uint16_t __p0)6268 __ai uint16x4_t vdup_n_u16(uint16_t __p0) {
6269 uint16x4_t __ret;
6270 __ret = (uint16x4_t) {__p0, __p0, __p0, __p0};
6271 return __ret;
6272 }
6273 #else
vdup_n_u16(uint16_t __p0)6274 __ai uint16x4_t vdup_n_u16(uint16_t __p0) {
6275 uint16x4_t __ret;
6276 __ret = (uint16x4_t) {__p0, __p0, __p0, __p0};
6277 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
6278 return __ret;
6279 }
6280 #endif
6281
6282 #ifdef __LITTLE_ENDIAN__
vdup_n_s8(int8_t __p0)6283 __ai int8x8_t vdup_n_s8(int8_t __p0) {
6284 int8x8_t __ret;
6285 __ret = (int8x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
6286 return __ret;
6287 }
6288 #else
vdup_n_s8(int8_t __p0)6289 __ai int8x8_t vdup_n_s8(int8_t __p0) {
6290 int8x8_t __ret;
6291 __ret = (int8x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
6292 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
6293 return __ret;
6294 }
6295 #endif
6296
6297 #ifdef __LITTLE_ENDIAN__
vdup_n_f32(float32_t __p0)6298 __ai float32x2_t vdup_n_f32(float32_t __p0) {
6299 float32x2_t __ret;
6300 __ret = (float32x2_t) {__p0, __p0};
6301 return __ret;
6302 }
6303 #else
vdup_n_f32(float32_t __p0)6304 __ai float32x2_t vdup_n_f32(float32_t __p0) {
6305 float32x2_t __ret;
6306 __ret = (float32x2_t) {__p0, __p0};
6307 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
6308 return __ret;
6309 }
6310 #endif
6311
6312 #ifdef __LITTLE_ENDIAN__
6313 #define vdup_n_f16(__p0) __extension__ ({ \
6314 float16x4_t __ret; \
6315 float16_t __s0 = __p0; \
6316 __ret = (float16x4_t) {__s0, __s0, __s0, __s0}; \
6317 __ret; \
6318 })
6319 #else
6320 #define vdup_n_f16(__p0) __extension__ ({ \
6321 float16x4_t __ret; \
6322 float16_t __s0 = __p0; \
6323 __ret = (float16x4_t) {__s0, __s0, __s0, __s0}; \
6324 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
6325 __ret; \
6326 })
6327 #endif
6328
6329 #ifdef __LITTLE_ENDIAN__
vdup_n_s32(int32_t __p0)6330 __ai int32x2_t vdup_n_s32(int32_t __p0) {
6331 int32x2_t __ret;
6332 __ret = (int32x2_t) {__p0, __p0};
6333 return __ret;
6334 }
6335 #else
vdup_n_s32(int32_t __p0)6336 __ai int32x2_t vdup_n_s32(int32_t __p0) {
6337 int32x2_t __ret;
6338 __ret = (int32x2_t) {__p0, __p0};
6339 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
6340 return __ret;
6341 }
6342 #endif
6343
vdup_n_s64(int64_t __p0)6344 __ai int64x1_t vdup_n_s64(int64_t __p0) {
6345 int64x1_t __ret;
6346 __ret = (int64x1_t) {__p0};
6347 return __ret;
6348 }
6349 #ifdef __LITTLE_ENDIAN__
vdup_n_s16(int16_t __p0)6350 __ai int16x4_t vdup_n_s16(int16_t __p0) {
6351 int16x4_t __ret;
6352 __ret = (int16x4_t) {__p0, __p0, __p0, __p0};
6353 return __ret;
6354 }
6355 #else
vdup_n_s16(int16_t __p0)6356 __ai int16x4_t vdup_n_s16(int16_t __p0) {
6357 int16x4_t __ret;
6358 __ret = (int16x4_t) {__p0, __p0, __p0, __p0};
6359 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
6360 return __ret;
6361 }
6362 #endif
6363
6364 #ifdef __LITTLE_ENDIAN__
veorq_u8(uint8x16_t __p0,uint8x16_t __p1)6365 __ai uint8x16_t veorq_u8(uint8x16_t __p0, uint8x16_t __p1) {
6366 uint8x16_t __ret;
6367 __ret = __p0 ^ __p1;
6368 return __ret;
6369 }
6370 #else
veorq_u8(uint8x16_t __p0,uint8x16_t __p1)6371 __ai uint8x16_t veorq_u8(uint8x16_t __p0, uint8x16_t __p1) {
6372 uint8x16_t __ret;
6373 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
6374 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
6375 __ret = __rev0 ^ __rev1;
6376 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
6377 return __ret;
6378 }
6379 #endif
6380
6381 #ifdef __LITTLE_ENDIAN__
veorq_u32(uint32x4_t __p0,uint32x4_t __p1)6382 __ai uint32x4_t veorq_u32(uint32x4_t __p0, uint32x4_t __p1) {
6383 uint32x4_t __ret;
6384 __ret = __p0 ^ __p1;
6385 return __ret;
6386 }
6387 #else
veorq_u32(uint32x4_t __p0,uint32x4_t __p1)6388 __ai uint32x4_t veorq_u32(uint32x4_t __p0, uint32x4_t __p1) {
6389 uint32x4_t __ret;
6390 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
6391 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
6392 __ret = __rev0 ^ __rev1;
6393 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
6394 return __ret;
6395 }
6396 #endif
6397
6398 #ifdef __LITTLE_ENDIAN__
veorq_u64(uint64x2_t __p0,uint64x2_t __p1)6399 __ai uint64x2_t veorq_u64(uint64x2_t __p0, uint64x2_t __p1) {
6400 uint64x2_t __ret;
6401 __ret = __p0 ^ __p1;
6402 return __ret;
6403 }
6404 #else
veorq_u64(uint64x2_t __p0,uint64x2_t __p1)6405 __ai uint64x2_t veorq_u64(uint64x2_t __p0, uint64x2_t __p1) {
6406 uint64x2_t __ret;
6407 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
6408 uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
6409 __ret = __rev0 ^ __rev1;
6410 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
6411 return __ret;
6412 }
6413 #endif
6414
6415 #ifdef __LITTLE_ENDIAN__
veorq_u16(uint16x8_t __p0,uint16x8_t __p1)6416 __ai uint16x8_t veorq_u16(uint16x8_t __p0, uint16x8_t __p1) {
6417 uint16x8_t __ret;
6418 __ret = __p0 ^ __p1;
6419 return __ret;
6420 }
6421 #else
veorq_u16(uint16x8_t __p0,uint16x8_t __p1)6422 __ai uint16x8_t veorq_u16(uint16x8_t __p0, uint16x8_t __p1) {
6423 uint16x8_t __ret;
6424 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
6425 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
6426 __ret = __rev0 ^ __rev1;
6427 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
6428 return __ret;
6429 }
6430 #endif
6431
6432 #ifdef __LITTLE_ENDIAN__
veorq_s8(int8x16_t __p0,int8x16_t __p1)6433 __ai int8x16_t veorq_s8(int8x16_t __p0, int8x16_t __p1) {
6434 int8x16_t __ret;
6435 __ret = __p0 ^ __p1;
6436 return __ret;
6437 }
6438 #else
veorq_s8(int8x16_t __p0,int8x16_t __p1)6439 __ai int8x16_t veorq_s8(int8x16_t __p0, int8x16_t __p1) {
6440 int8x16_t __ret;
6441 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
6442 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
6443 __ret = __rev0 ^ __rev1;
6444 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
6445 return __ret;
6446 }
6447 #endif
6448
6449 #ifdef __LITTLE_ENDIAN__
veorq_s32(int32x4_t __p0,int32x4_t __p1)6450 __ai int32x4_t veorq_s32(int32x4_t __p0, int32x4_t __p1) {
6451 int32x4_t __ret;
6452 __ret = __p0 ^ __p1;
6453 return __ret;
6454 }
6455 #else
veorq_s32(int32x4_t __p0,int32x4_t __p1)6456 __ai int32x4_t veorq_s32(int32x4_t __p0, int32x4_t __p1) {
6457 int32x4_t __ret;
6458 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
6459 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
6460 __ret = __rev0 ^ __rev1;
6461 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
6462 return __ret;
6463 }
6464 #endif
6465
6466 #ifdef __LITTLE_ENDIAN__
veorq_s64(int64x2_t __p0,int64x2_t __p1)6467 __ai int64x2_t veorq_s64(int64x2_t __p0, int64x2_t __p1) {
6468 int64x2_t __ret;
6469 __ret = __p0 ^ __p1;
6470 return __ret;
6471 }
6472 #else
veorq_s64(int64x2_t __p0,int64x2_t __p1)6473 __ai int64x2_t veorq_s64(int64x2_t __p0, int64x2_t __p1) {
6474 int64x2_t __ret;
6475 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
6476 int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
6477 __ret = __rev0 ^ __rev1;
6478 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
6479 return __ret;
6480 }
6481 #endif
6482
6483 #ifdef __LITTLE_ENDIAN__
veorq_s16(int16x8_t __p0,int16x8_t __p1)6484 __ai int16x8_t veorq_s16(int16x8_t __p0, int16x8_t __p1) {
6485 int16x8_t __ret;
6486 __ret = __p0 ^ __p1;
6487 return __ret;
6488 }
6489 #else
veorq_s16(int16x8_t __p0,int16x8_t __p1)6490 __ai int16x8_t veorq_s16(int16x8_t __p0, int16x8_t __p1) {
6491 int16x8_t __ret;
6492 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
6493 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
6494 __ret = __rev0 ^ __rev1;
6495 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
6496 return __ret;
6497 }
6498 #endif
6499
6500 #ifdef __LITTLE_ENDIAN__
veor_u8(uint8x8_t __p0,uint8x8_t __p1)6501 __ai uint8x8_t veor_u8(uint8x8_t __p0, uint8x8_t __p1) {
6502 uint8x8_t __ret;
6503 __ret = __p0 ^ __p1;
6504 return __ret;
6505 }
6506 #else
veor_u8(uint8x8_t __p0,uint8x8_t __p1)6507 __ai uint8x8_t veor_u8(uint8x8_t __p0, uint8x8_t __p1) {
6508 uint8x8_t __ret;
6509 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
6510 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
6511 __ret = __rev0 ^ __rev1;
6512 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
6513 return __ret;
6514 }
6515 #endif
6516
6517 #ifdef __LITTLE_ENDIAN__
veor_u32(uint32x2_t __p0,uint32x2_t __p1)6518 __ai uint32x2_t veor_u32(uint32x2_t __p0, uint32x2_t __p1) {
6519 uint32x2_t __ret;
6520 __ret = __p0 ^ __p1;
6521 return __ret;
6522 }
6523 #else
veor_u32(uint32x2_t __p0,uint32x2_t __p1)6524 __ai uint32x2_t veor_u32(uint32x2_t __p0, uint32x2_t __p1) {
6525 uint32x2_t __ret;
6526 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
6527 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
6528 __ret = __rev0 ^ __rev1;
6529 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
6530 return __ret;
6531 }
6532 #endif
6533
veor_u64(uint64x1_t __p0,uint64x1_t __p1)6534 __ai uint64x1_t veor_u64(uint64x1_t __p0, uint64x1_t __p1) {
6535 uint64x1_t __ret;
6536 __ret = __p0 ^ __p1;
6537 return __ret;
6538 }
6539 #ifdef __LITTLE_ENDIAN__
veor_u16(uint16x4_t __p0,uint16x4_t __p1)6540 __ai uint16x4_t veor_u16(uint16x4_t __p0, uint16x4_t __p1) {
6541 uint16x4_t __ret;
6542 __ret = __p0 ^ __p1;
6543 return __ret;
6544 }
6545 #else
veor_u16(uint16x4_t __p0,uint16x4_t __p1)6546 __ai uint16x4_t veor_u16(uint16x4_t __p0, uint16x4_t __p1) {
6547 uint16x4_t __ret;
6548 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
6549 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
6550 __ret = __rev0 ^ __rev1;
6551 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
6552 return __ret;
6553 }
6554 #endif
6555
6556 #ifdef __LITTLE_ENDIAN__
veor_s8(int8x8_t __p0,int8x8_t __p1)6557 __ai int8x8_t veor_s8(int8x8_t __p0, int8x8_t __p1) {
6558 int8x8_t __ret;
6559 __ret = __p0 ^ __p1;
6560 return __ret;
6561 }
6562 #else
veor_s8(int8x8_t __p0,int8x8_t __p1)6563 __ai int8x8_t veor_s8(int8x8_t __p0, int8x8_t __p1) {
6564 int8x8_t __ret;
6565 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
6566 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
6567 __ret = __rev0 ^ __rev1;
6568 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
6569 return __ret;
6570 }
6571 #endif
6572
6573 #ifdef __LITTLE_ENDIAN__
veor_s32(int32x2_t __p0,int32x2_t __p1)6574 __ai int32x2_t veor_s32(int32x2_t __p0, int32x2_t __p1) {
6575 int32x2_t __ret;
6576 __ret = __p0 ^ __p1;
6577 return __ret;
6578 }
6579 #else
veor_s32(int32x2_t __p0,int32x2_t __p1)6580 __ai int32x2_t veor_s32(int32x2_t __p0, int32x2_t __p1) {
6581 int32x2_t __ret;
6582 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
6583 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
6584 __ret = __rev0 ^ __rev1;
6585 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
6586 return __ret;
6587 }
6588 #endif
6589
veor_s64(int64x1_t __p0,int64x1_t __p1)6590 __ai int64x1_t veor_s64(int64x1_t __p0, int64x1_t __p1) {
6591 int64x1_t __ret;
6592 __ret = __p0 ^ __p1;
6593 return __ret;
6594 }
6595 #ifdef __LITTLE_ENDIAN__
veor_s16(int16x4_t __p0,int16x4_t __p1)6596 __ai int16x4_t veor_s16(int16x4_t __p0, int16x4_t __p1) {
6597 int16x4_t __ret;
6598 __ret = __p0 ^ __p1;
6599 return __ret;
6600 }
6601 #else
veor_s16(int16x4_t __p0,int16x4_t __p1)6602 __ai int16x4_t veor_s16(int16x4_t __p0, int16x4_t __p1) {
6603 int16x4_t __ret;
6604 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
6605 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
6606 __ret = __rev0 ^ __rev1;
6607 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
6608 return __ret;
6609 }
6610 #endif
6611
6612 #ifdef __LITTLE_ENDIAN__
6613 #define vext_p8(__p0, __p1, __p2) __extension__ ({ \
6614 poly8x8_t __ret; \
6615 poly8x8_t __s0 = __p0; \
6616 poly8x8_t __s1 = __p1; \
6617 __ret = (poly8x8_t) __builtin_neon_vext_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 4); \
6618 __ret; \
6619 })
6620 #else
6621 #define vext_p8(__p0, __p1, __p2) __extension__ ({ \
6622 poly8x8_t __ret; \
6623 poly8x8_t __s0 = __p0; \
6624 poly8x8_t __s1 = __p1; \
6625 poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
6626 poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
6627 __ret = (poly8x8_t) __builtin_neon_vext_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 4); \
6628 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
6629 __ret; \
6630 })
6631 #endif
6632
6633 #ifdef __LITTLE_ENDIAN__
6634 #define vext_p16(__p0, __p1, __p2) __extension__ ({ \
6635 poly16x4_t __ret; \
6636 poly16x4_t __s0 = __p0; \
6637 poly16x4_t __s1 = __p1; \
6638 __ret = (poly16x4_t) __builtin_neon_vext_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 5); \
6639 __ret; \
6640 })
6641 #else
6642 #define vext_p16(__p0, __p1, __p2) __extension__ ({ \
6643 poly16x4_t __ret; \
6644 poly16x4_t __s0 = __p0; \
6645 poly16x4_t __s1 = __p1; \
6646 poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
6647 poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
6648 __ret = (poly16x4_t) __builtin_neon_vext_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 5); \
6649 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
6650 __ret; \
6651 })
6652 #endif
6653
6654 #ifdef __LITTLE_ENDIAN__
6655 #define vextq_p8(__p0, __p1, __p2) __extension__ ({ \
6656 poly8x16_t __ret; \
6657 poly8x16_t __s0 = __p0; \
6658 poly8x16_t __s1 = __p1; \
6659 __ret = (poly8x16_t) __builtin_neon_vextq_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 36); \
6660 __ret; \
6661 })
6662 #else
6663 #define vextq_p8(__p0, __p1, __p2) __extension__ ({ \
6664 poly8x16_t __ret; \
6665 poly8x16_t __s0 = __p0; \
6666 poly8x16_t __s1 = __p1; \
6667 poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
6668 poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
6669 __ret = (poly8x16_t) __builtin_neon_vextq_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 36); \
6670 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
6671 __ret; \
6672 })
6673 #endif
6674
6675 #ifdef __LITTLE_ENDIAN__
6676 #define vextq_p16(__p0, __p1, __p2) __extension__ ({ \
6677 poly16x8_t __ret; \
6678 poly16x8_t __s0 = __p0; \
6679 poly16x8_t __s1 = __p1; \
6680 __ret = (poly16x8_t) __builtin_neon_vextq_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 37); \
6681 __ret; \
6682 })
6683 #else
6684 #define vextq_p16(__p0, __p1, __p2) __extension__ ({ \
6685 poly16x8_t __ret; \
6686 poly16x8_t __s0 = __p0; \
6687 poly16x8_t __s1 = __p1; \
6688 poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
6689 poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
6690 __ret = (poly16x8_t) __builtin_neon_vextq_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 37); \
6691 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
6692 __ret; \
6693 })
6694 #endif
6695
6696 #ifdef __LITTLE_ENDIAN__
6697 #define vextq_u8(__p0, __p1, __p2) __extension__ ({ \
6698 uint8x16_t __ret; \
6699 uint8x16_t __s0 = __p0; \
6700 uint8x16_t __s1 = __p1; \
6701 __ret = (uint8x16_t) __builtin_neon_vextq_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 48); \
6702 __ret; \
6703 })
6704 #else
6705 #define vextq_u8(__p0, __p1, __p2) __extension__ ({ \
6706 uint8x16_t __ret; \
6707 uint8x16_t __s0 = __p0; \
6708 uint8x16_t __s1 = __p1; \
6709 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
6710 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
6711 __ret = (uint8x16_t) __builtin_neon_vextq_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 48); \
6712 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
6713 __ret; \
6714 })
6715 #endif
6716
6717 #ifdef __LITTLE_ENDIAN__
6718 #define vextq_u32(__p0, __p1, __p2) __extension__ ({ \
6719 uint32x4_t __ret; \
6720 uint32x4_t __s0 = __p0; \
6721 uint32x4_t __s1 = __p1; \
6722 __ret = (uint32x4_t) __builtin_neon_vextq_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 50); \
6723 __ret; \
6724 })
6725 #else
6726 #define vextq_u32(__p0, __p1, __p2) __extension__ ({ \
6727 uint32x4_t __ret; \
6728 uint32x4_t __s0 = __p0; \
6729 uint32x4_t __s1 = __p1; \
6730 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
6731 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
6732 __ret = (uint32x4_t) __builtin_neon_vextq_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 50); \
6733 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
6734 __ret; \
6735 })
6736 #endif
6737
6738 #ifdef __LITTLE_ENDIAN__
6739 #define vextq_u64(__p0, __p1, __p2) __extension__ ({ \
6740 uint64x2_t __ret; \
6741 uint64x2_t __s0 = __p0; \
6742 uint64x2_t __s1 = __p1; \
6743 __ret = (uint64x2_t) __builtin_neon_vextq_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 51); \
6744 __ret; \
6745 })
6746 #else
6747 #define vextq_u64(__p0, __p1, __p2) __extension__ ({ \
6748 uint64x2_t __ret; \
6749 uint64x2_t __s0 = __p0; \
6750 uint64x2_t __s1 = __p1; \
6751 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
6752 uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
6753 __ret = (uint64x2_t) __builtin_neon_vextq_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 51); \
6754 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
6755 __ret; \
6756 })
6757 #endif
6758
6759 #ifdef __LITTLE_ENDIAN__
6760 #define vextq_u16(__p0, __p1, __p2) __extension__ ({ \
6761 uint16x8_t __ret; \
6762 uint16x8_t __s0 = __p0; \
6763 uint16x8_t __s1 = __p1; \
6764 __ret = (uint16x8_t) __builtin_neon_vextq_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 49); \
6765 __ret; \
6766 })
6767 #else
6768 #define vextq_u16(__p0, __p1, __p2) __extension__ ({ \
6769 uint16x8_t __ret; \
6770 uint16x8_t __s0 = __p0; \
6771 uint16x8_t __s1 = __p1; \
6772 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
6773 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
6774 __ret = (uint16x8_t) __builtin_neon_vextq_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 49); \
6775 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
6776 __ret; \
6777 })
6778 #endif
6779
6780 #ifdef __LITTLE_ENDIAN__
6781 #define vextq_s8(__p0, __p1, __p2) __extension__ ({ \
6782 int8x16_t __ret; \
6783 int8x16_t __s0 = __p0; \
6784 int8x16_t __s1 = __p1; \
6785 __ret = (int8x16_t) __builtin_neon_vextq_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 32); \
6786 __ret; \
6787 })
6788 #else
6789 #define vextq_s8(__p0, __p1, __p2) __extension__ ({ \
6790 int8x16_t __ret; \
6791 int8x16_t __s0 = __p0; \
6792 int8x16_t __s1 = __p1; \
6793 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
6794 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
6795 __ret = (int8x16_t) __builtin_neon_vextq_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 32); \
6796 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
6797 __ret; \
6798 })
6799 #endif
6800
6801 #ifdef __LITTLE_ENDIAN__
6802 #define vextq_f32(__p0, __p1, __p2) __extension__ ({ \
6803 float32x4_t __ret; \
6804 float32x4_t __s0 = __p0; \
6805 float32x4_t __s1 = __p1; \
6806 __ret = (float32x4_t) __builtin_neon_vextq_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 41); \
6807 __ret; \
6808 })
6809 #else
6810 #define vextq_f32(__p0, __p1, __p2) __extension__ ({ \
6811 float32x4_t __ret; \
6812 float32x4_t __s0 = __p0; \
6813 float32x4_t __s1 = __p1; \
6814 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
6815 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
6816 __ret = (float32x4_t) __builtin_neon_vextq_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 41); \
6817 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
6818 __ret; \
6819 })
6820 #endif
6821
6822 #ifdef __LITTLE_ENDIAN__
6823 #define vextq_s32(__p0, __p1, __p2) __extension__ ({ \
6824 int32x4_t __ret; \
6825 int32x4_t __s0 = __p0; \
6826 int32x4_t __s1 = __p1; \
6827 __ret = (int32x4_t) __builtin_neon_vextq_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 34); \
6828 __ret; \
6829 })
6830 #else
6831 #define vextq_s32(__p0, __p1, __p2) __extension__ ({ \
6832 int32x4_t __ret; \
6833 int32x4_t __s0 = __p0; \
6834 int32x4_t __s1 = __p1; \
6835 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
6836 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
6837 __ret = (int32x4_t) __builtin_neon_vextq_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 34); \
6838 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
6839 __ret; \
6840 })
6841 #endif
6842
6843 #ifdef __LITTLE_ENDIAN__
6844 #define vextq_s64(__p0, __p1, __p2) __extension__ ({ \
6845 int64x2_t __ret; \
6846 int64x2_t __s0 = __p0; \
6847 int64x2_t __s1 = __p1; \
6848 __ret = (int64x2_t) __builtin_neon_vextq_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 35); \
6849 __ret; \
6850 })
6851 #else
6852 #define vextq_s64(__p0, __p1, __p2) __extension__ ({ \
6853 int64x2_t __ret; \
6854 int64x2_t __s0 = __p0; \
6855 int64x2_t __s1 = __p1; \
6856 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
6857 int64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
6858 __ret = (int64x2_t) __builtin_neon_vextq_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 35); \
6859 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
6860 __ret; \
6861 })
6862 #endif
6863
6864 #ifdef __LITTLE_ENDIAN__
6865 #define vextq_s16(__p0, __p1, __p2) __extension__ ({ \
6866 int16x8_t __ret; \
6867 int16x8_t __s0 = __p0; \
6868 int16x8_t __s1 = __p1; \
6869 __ret = (int16x8_t) __builtin_neon_vextq_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 33); \
6870 __ret; \
6871 })
6872 #else
6873 #define vextq_s16(__p0, __p1, __p2) __extension__ ({ \
6874 int16x8_t __ret; \
6875 int16x8_t __s0 = __p0; \
6876 int16x8_t __s1 = __p1; \
6877 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
6878 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
6879 __ret = (int16x8_t) __builtin_neon_vextq_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 33); \
6880 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
6881 __ret; \
6882 })
6883 #endif
6884
6885 #ifdef __LITTLE_ENDIAN__
6886 #define vext_u8(__p0, __p1, __p2) __extension__ ({ \
6887 uint8x8_t __ret; \
6888 uint8x8_t __s0 = __p0; \
6889 uint8x8_t __s1 = __p1; \
6890 __ret = (uint8x8_t) __builtin_neon_vext_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 16); \
6891 __ret; \
6892 })
6893 #else
6894 #define vext_u8(__p0, __p1, __p2) __extension__ ({ \
6895 uint8x8_t __ret; \
6896 uint8x8_t __s0 = __p0; \
6897 uint8x8_t __s1 = __p1; \
6898 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
6899 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
6900 __ret = (uint8x8_t) __builtin_neon_vext_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 16); \
6901 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
6902 __ret; \
6903 })
6904 #endif
6905
6906 #ifdef __LITTLE_ENDIAN__
6907 #define vext_u32(__p0, __p1, __p2) __extension__ ({ \
6908 uint32x2_t __ret; \
6909 uint32x2_t __s0 = __p0; \
6910 uint32x2_t __s1 = __p1; \
6911 __ret = (uint32x2_t) __builtin_neon_vext_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 18); \
6912 __ret; \
6913 })
6914 #else
6915 #define vext_u32(__p0, __p1, __p2) __extension__ ({ \
6916 uint32x2_t __ret; \
6917 uint32x2_t __s0 = __p0; \
6918 uint32x2_t __s1 = __p1; \
6919 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
6920 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
6921 __ret = (uint32x2_t) __builtin_neon_vext_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 18); \
6922 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
6923 __ret; \
6924 })
6925 #endif
6926
6927 #define vext_u64(__p0, __p1, __p2) __extension__ ({ \
6928 uint64x1_t __ret; \
6929 uint64x1_t __s0 = __p0; \
6930 uint64x1_t __s1 = __p1; \
6931 __ret = (uint64x1_t) __builtin_neon_vext_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 19); \
6932 __ret; \
6933 })
6934 #ifdef __LITTLE_ENDIAN__
6935 #define vext_u16(__p0, __p1, __p2) __extension__ ({ \
6936 uint16x4_t __ret; \
6937 uint16x4_t __s0 = __p0; \
6938 uint16x4_t __s1 = __p1; \
6939 __ret = (uint16x4_t) __builtin_neon_vext_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 17); \
6940 __ret; \
6941 })
6942 #else
6943 #define vext_u16(__p0, __p1, __p2) __extension__ ({ \
6944 uint16x4_t __ret; \
6945 uint16x4_t __s0 = __p0; \
6946 uint16x4_t __s1 = __p1; \
6947 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
6948 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
6949 __ret = (uint16x4_t) __builtin_neon_vext_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 17); \
6950 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
6951 __ret; \
6952 })
6953 #endif
6954
6955 #ifdef __LITTLE_ENDIAN__
6956 #define vext_s8(__p0, __p1, __p2) __extension__ ({ \
6957 int8x8_t __ret; \
6958 int8x8_t __s0 = __p0; \
6959 int8x8_t __s1 = __p1; \
6960 __ret = (int8x8_t) __builtin_neon_vext_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 0); \
6961 __ret; \
6962 })
6963 #else
6964 #define vext_s8(__p0, __p1, __p2) __extension__ ({ \
6965 int8x8_t __ret; \
6966 int8x8_t __s0 = __p0; \
6967 int8x8_t __s1 = __p1; \
6968 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
6969 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
6970 __ret = (int8x8_t) __builtin_neon_vext_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 0); \
6971 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
6972 __ret; \
6973 })
6974 #endif
6975
6976 #ifdef __LITTLE_ENDIAN__
6977 #define vext_f32(__p0, __p1, __p2) __extension__ ({ \
6978 float32x2_t __ret; \
6979 float32x2_t __s0 = __p0; \
6980 float32x2_t __s1 = __p1; \
6981 __ret = (float32x2_t) __builtin_neon_vext_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 9); \
6982 __ret; \
6983 })
6984 #else
6985 #define vext_f32(__p0, __p1, __p2) __extension__ ({ \
6986 float32x2_t __ret; \
6987 float32x2_t __s0 = __p0; \
6988 float32x2_t __s1 = __p1; \
6989 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
6990 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
6991 __ret = (float32x2_t) __builtin_neon_vext_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 9); \
6992 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
6993 __ret; \
6994 })
6995 #endif
6996
6997 #ifdef __LITTLE_ENDIAN__
6998 #define vext_s32(__p0, __p1, __p2) __extension__ ({ \
6999 int32x2_t __ret; \
7000 int32x2_t __s0 = __p0; \
7001 int32x2_t __s1 = __p1; \
7002 __ret = (int32x2_t) __builtin_neon_vext_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 2); \
7003 __ret; \
7004 })
7005 #else
7006 #define vext_s32(__p0, __p1, __p2) __extension__ ({ \
7007 int32x2_t __ret; \
7008 int32x2_t __s0 = __p0; \
7009 int32x2_t __s1 = __p1; \
7010 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
7011 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
7012 __ret = (int32x2_t) __builtin_neon_vext_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 2); \
7013 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
7014 __ret; \
7015 })
7016 #endif
7017
7018 #define vext_s64(__p0, __p1, __p2) __extension__ ({ \
7019 int64x1_t __ret; \
7020 int64x1_t __s0 = __p0; \
7021 int64x1_t __s1 = __p1; \
7022 __ret = (int64x1_t) __builtin_neon_vext_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 3); \
7023 __ret; \
7024 })
7025 #ifdef __LITTLE_ENDIAN__
7026 #define vext_s16(__p0, __p1, __p2) __extension__ ({ \
7027 int16x4_t __ret; \
7028 int16x4_t __s0 = __p0; \
7029 int16x4_t __s1 = __p1; \
7030 __ret = (int16x4_t) __builtin_neon_vext_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 1); \
7031 __ret; \
7032 })
7033 #else
7034 #define vext_s16(__p0, __p1, __p2) __extension__ ({ \
7035 int16x4_t __ret; \
7036 int16x4_t __s0 = __p0; \
7037 int16x4_t __s1 = __p1; \
7038 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
7039 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
7040 __ret = (int16x4_t) __builtin_neon_vext_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 1); \
7041 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
7042 __ret; \
7043 })
7044 #endif
7045
7046 #ifdef __LITTLE_ENDIAN__
vget_high_p8(poly8x16_t __p0)7047 __ai poly8x8_t vget_high_p8(poly8x16_t __p0) {
7048 poly8x8_t __ret;
7049 __ret = __builtin_shufflevector(__p0, __p0, 8, 9, 10, 11, 12, 13, 14, 15);
7050 return __ret;
7051 }
7052 #else
vget_high_p8(poly8x16_t __p0)7053 __ai poly8x8_t vget_high_p8(poly8x16_t __p0) {
7054 poly8x8_t __ret;
7055 poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
7056 __ret = __builtin_shufflevector(__rev0, __rev0, 8, 9, 10, 11, 12, 13, 14, 15);
7057 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
7058 return __ret;
7059 }
__noswap_vget_high_p8(poly8x16_t __p0)7060 __ai poly8x8_t __noswap_vget_high_p8(poly8x16_t __p0) {
7061 poly8x8_t __ret;
7062 __ret = __builtin_shufflevector(__p0, __p0, 8, 9, 10, 11, 12, 13, 14, 15);
7063 return __ret;
7064 }
7065 #endif
7066
7067 #ifdef __LITTLE_ENDIAN__
vget_high_p16(poly16x8_t __p0)7068 __ai poly16x4_t vget_high_p16(poly16x8_t __p0) {
7069 poly16x4_t __ret;
7070 __ret = __builtin_shufflevector(__p0, __p0, 4, 5, 6, 7);
7071 return __ret;
7072 }
7073 #else
vget_high_p16(poly16x8_t __p0)7074 __ai poly16x4_t vget_high_p16(poly16x8_t __p0) {
7075 poly16x4_t __ret;
7076 poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
7077 __ret = __builtin_shufflevector(__rev0, __rev0, 4, 5, 6, 7);
7078 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
7079 return __ret;
7080 }
7081 #endif
7082
7083 #ifdef __LITTLE_ENDIAN__
vget_high_u8(uint8x16_t __p0)7084 __ai uint8x8_t vget_high_u8(uint8x16_t __p0) {
7085 uint8x8_t __ret;
7086 __ret = __builtin_shufflevector(__p0, __p0, 8, 9, 10, 11, 12, 13, 14, 15);
7087 return __ret;
7088 }
7089 #else
vget_high_u8(uint8x16_t __p0)7090 __ai uint8x8_t vget_high_u8(uint8x16_t __p0) {
7091 uint8x8_t __ret;
7092 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
7093 __ret = __builtin_shufflevector(__rev0, __rev0, 8, 9, 10, 11, 12, 13, 14, 15);
7094 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
7095 return __ret;
7096 }
__noswap_vget_high_u8(uint8x16_t __p0)7097 __ai uint8x8_t __noswap_vget_high_u8(uint8x16_t __p0) {
7098 uint8x8_t __ret;
7099 __ret = __builtin_shufflevector(__p0, __p0, 8, 9, 10, 11, 12, 13, 14, 15);
7100 return __ret;
7101 }
7102 #endif
7103
7104 #ifdef __LITTLE_ENDIAN__
vget_high_u32(uint32x4_t __p0)7105 __ai uint32x2_t vget_high_u32(uint32x4_t __p0) {
7106 uint32x2_t __ret;
7107 __ret = __builtin_shufflevector(__p0, __p0, 2, 3);
7108 return __ret;
7109 }
7110 #else
vget_high_u32(uint32x4_t __p0)7111 __ai uint32x2_t vget_high_u32(uint32x4_t __p0) {
7112 uint32x2_t __ret;
7113 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
7114 __ret = __builtin_shufflevector(__rev0, __rev0, 2, 3);
7115 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
7116 return __ret;
7117 }
__noswap_vget_high_u32(uint32x4_t __p0)7118 __ai uint32x2_t __noswap_vget_high_u32(uint32x4_t __p0) {
7119 uint32x2_t __ret;
7120 __ret = __builtin_shufflevector(__p0, __p0, 2, 3);
7121 return __ret;
7122 }
7123 #endif
7124
7125 #ifdef __LITTLE_ENDIAN__
vget_high_u64(uint64x2_t __p0)7126 __ai uint64x1_t vget_high_u64(uint64x2_t __p0) {
7127 uint64x1_t __ret;
7128 __ret = __builtin_shufflevector(__p0, __p0, 1);
7129 return __ret;
7130 }
7131 #else
vget_high_u64(uint64x2_t __p0)7132 __ai uint64x1_t vget_high_u64(uint64x2_t __p0) {
7133 uint64x1_t __ret;
7134 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
7135 __ret = __builtin_shufflevector(__rev0, __rev0, 1);
7136 return __ret;
7137 }
7138 #endif
7139
7140 #ifdef __LITTLE_ENDIAN__
vget_high_u16(uint16x8_t __p0)7141 __ai uint16x4_t vget_high_u16(uint16x8_t __p0) {
7142 uint16x4_t __ret;
7143 __ret = __builtin_shufflevector(__p0, __p0, 4, 5, 6, 7);
7144 return __ret;
7145 }
7146 #else
vget_high_u16(uint16x8_t __p0)7147 __ai uint16x4_t vget_high_u16(uint16x8_t __p0) {
7148 uint16x4_t __ret;
7149 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
7150 __ret = __builtin_shufflevector(__rev0, __rev0, 4, 5, 6, 7);
7151 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
7152 return __ret;
7153 }
__noswap_vget_high_u16(uint16x8_t __p0)7154 __ai uint16x4_t __noswap_vget_high_u16(uint16x8_t __p0) {
7155 uint16x4_t __ret;
7156 __ret = __builtin_shufflevector(__p0, __p0, 4, 5, 6, 7);
7157 return __ret;
7158 }
7159 #endif
7160
7161 #ifdef __LITTLE_ENDIAN__
vget_high_s8(int8x16_t __p0)7162 __ai int8x8_t vget_high_s8(int8x16_t __p0) {
7163 int8x8_t __ret;
7164 __ret = __builtin_shufflevector(__p0, __p0, 8, 9, 10, 11, 12, 13, 14, 15);
7165 return __ret;
7166 }
7167 #else
vget_high_s8(int8x16_t __p0)7168 __ai int8x8_t vget_high_s8(int8x16_t __p0) {
7169 int8x8_t __ret;
7170 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
7171 __ret = __builtin_shufflevector(__rev0, __rev0, 8, 9, 10, 11, 12, 13, 14, 15);
7172 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
7173 return __ret;
7174 }
__noswap_vget_high_s8(int8x16_t __p0)7175 __ai int8x8_t __noswap_vget_high_s8(int8x16_t __p0) {
7176 int8x8_t __ret;
7177 __ret = __builtin_shufflevector(__p0, __p0, 8, 9, 10, 11, 12, 13, 14, 15);
7178 return __ret;
7179 }
7180 #endif
7181
7182 #ifdef __LITTLE_ENDIAN__
vget_high_f32(float32x4_t __p0)7183 __ai float32x2_t vget_high_f32(float32x4_t __p0) {
7184 float32x2_t __ret;
7185 __ret = __builtin_shufflevector(__p0, __p0, 2, 3);
7186 return __ret;
7187 }
7188 #else
vget_high_f32(float32x4_t __p0)7189 __ai float32x2_t vget_high_f32(float32x4_t __p0) {
7190 float32x2_t __ret;
7191 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
7192 __ret = __builtin_shufflevector(__rev0, __rev0, 2, 3);
7193 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
7194 return __ret;
7195 }
__noswap_vget_high_f32(float32x4_t __p0)7196 __ai float32x2_t __noswap_vget_high_f32(float32x4_t __p0) {
7197 float32x2_t __ret;
7198 __ret = __builtin_shufflevector(__p0, __p0, 2, 3);
7199 return __ret;
7200 }
7201 #endif
7202
7203 #ifdef __LITTLE_ENDIAN__
vget_high_f16(float16x8_t __p0)7204 __ai float16x4_t vget_high_f16(float16x8_t __p0) {
7205 float16x4_t __ret;
7206 __ret = __builtin_shufflevector(__p0, __p0, 4, 5, 6, 7);
7207 return __ret;
7208 }
7209 #else
vget_high_f16(float16x8_t __p0)7210 __ai float16x4_t vget_high_f16(float16x8_t __p0) {
7211 float16x4_t __ret;
7212 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
7213 __ret = __builtin_shufflevector(__rev0, __rev0, 4, 5, 6, 7);
7214 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
7215 return __ret;
7216 }
__noswap_vget_high_f16(float16x8_t __p0)7217 __ai float16x4_t __noswap_vget_high_f16(float16x8_t __p0) {
7218 float16x4_t __ret;
7219 __ret = __builtin_shufflevector(__p0, __p0, 4, 5, 6, 7);
7220 return __ret;
7221 }
7222 #endif
7223
7224 #ifdef __LITTLE_ENDIAN__
vget_high_s32(int32x4_t __p0)7225 __ai int32x2_t vget_high_s32(int32x4_t __p0) {
7226 int32x2_t __ret;
7227 __ret = __builtin_shufflevector(__p0, __p0, 2, 3);
7228 return __ret;
7229 }
7230 #else
vget_high_s32(int32x4_t __p0)7231 __ai int32x2_t vget_high_s32(int32x4_t __p0) {
7232 int32x2_t __ret;
7233 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
7234 __ret = __builtin_shufflevector(__rev0, __rev0, 2, 3);
7235 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
7236 return __ret;
7237 }
__noswap_vget_high_s32(int32x4_t __p0)7238 __ai int32x2_t __noswap_vget_high_s32(int32x4_t __p0) {
7239 int32x2_t __ret;
7240 __ret = __builtin_shufflevector(__p0, __p0, 2, 3);
7241 return __ret;
7242 }
7243 #endif
7244
7245 #ifdef __LITTLE_ENDIAN__
vget_high_s64(int64x2_t __p0)7246 __ai int64x1_t vget_high_s64(int64x2_t __p0) {
7247 int64x1_t __ret;
7248 __ret = __builtin_shufflevector(__p0, __p0, 1);
7249 return __ret;
7250 }
7251 #else
vget_high_s64(int64x2_t __p0)7252 __ai int64x1_t vget_high_s64(int64x2_t __p0) {
7253 int64x1_t __ret;
7254 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
7255 __ret = __builtin_shufflevector(__rev0, __rev0, 1);
7256 return __ret;
7257 }
7258 #endif
7259
7260 #ifdef __LITTLE_ENDIAN__
vget_high_s16(int16x8_t __p0)7261 __ai int16x4_t vget_high_s16(int16x8_t __p0) {
7262 int16x4_t __ret;
7263 __ret = __builtin_shufflevector(__p0, __p0, 4, 5, 6, 7);
7264 return __ret;
7265 }
7266 #else
vget_high_s16(int16x8_t __p0)7267 __ai int16x4_t vget_high_s16(int16x8_t __p0) {
7268 int16x4_t __ret;
7269 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
7270 __ret = __builtin_shufflevector(__rev0, __rev0, 4, 5, 6, 7);
7271 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
7272 return __ret;
7273 }
__noswap_vget_high_s16(int16x8_t __p0)7274 __ai int16x4_t __noswap_vget_high_s16(int16x8_t __p0) {
7275 int16x4_t __ret;
7276 __ret = __builtin_shufflevector(__p0, __p0, 4, 5, 6, 7);
7277 return __ret;
7278 }
7279 #endif
7280
7281 #ifdef __LITTLE_ENDIAN__
7282 #define vget_lane_p8(__p0, __p1) __extension__ ({ \
7283 poly8_t __ret; \
7284 poly8x8_t __s0 = __p0; \
7285 __ret = (poly8_t) __builtin_neon_vget_lane_i8((poly8x8_t)__s0, __p1); \
7286 __ret; \
7287 })
7288 #else
7289 #define vget_lane_p8(__p0, __p1) __extension__ ({ \
7290 poly8_t __ret; \
7291 poly8x8_t __s0 = __p0; \
7292 poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
7293 __ret = (poly8_t) __builtin_neon_vget_lane_i8((poly8x8_t)__rev0, __p1); \
7294 __ret; \
7295 })
7296 #define __noswap_vget_lane_p8(__p0, __p1) __extension__ ({ \
7297 poly8_t __ret; \
7298 poly8x8_t __s0 = __p0; \
7299 __ret = (poly8_t) __builtin_neon_vget_lane_i8((poly8x8_t)__s0, __p1); \
7300 __ret; \
7301 })
7302 #endif
7303
7304 #ifdef __LITTLE_ENDIAN__
7305 #define vget_lane_p16(__p0, __p1) __extension__ ({ \
7306 poly16_t __ret; \
7307 poly16x4_t __s0 = __p0; \
7308 __ret = (poly16_t) __builtin_neon_vget_lane_i16((poly16x4_t)__s0, __p1); \
7309 __ret; \
7310 })
7311 #else
7312 #define vget_lane_p16(__p0, __p1) __extension__ ({ \
7313 poly16_t __ret; \
7314 poly16x4_t __s0 = __p0; \
7315 poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
7316 __ret = (poly16_t) __builtin_neon_vget_lane_i16((poly16x4_t)__rev0, __p1); \
7317 __ret; \
7318 })
7319 #define __noswap_vget_lane_p16(__p0, __p1) __extension__ ({ \
7320 poly16_t __ret; \
7321 poly16x4_t __s0 = __p0; \
7322 __ret = (poly16_t) __builtin_neon_vget_lane_i16((poly16x4_t)__s0, __p1); \
7323 __ret; \
7324 })
7325 #endif
7326
7327 #ifdef __LITTLE_ENDIAN__
7328 #define vgetq_lane_p8(__p0, __p1) __extension__ ({ \
7329 poly8_t __ret; \
7330 poly8x16_t __s0 = __p0; \
7331 __ret = (poly8_t) __builtin_neon_vgetq_lane_i8((poly8x16_t)__s0, __p1); \
7332 __ret; \
7333 })
7334 #else
7335 #define vgetq_lane_p8(__p0, __p1) __extension__ ({ \
7336 poly8_t __ret; \
7337 poly8x16_t __s0 = __p0; \
7338 poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
7339 __ret = (poly8_t) __builtin_neon_vgetq_lane_i8((poly8x16_t)__rev0, __p1); \
7340 __ret; \
7341 })
7342 #define __noswap_vgetq_lane_p8(__p0, __p1) __extension__ ({ \
7343 poly8_t __ret; \
7344 poly8x16_t __s0 = __p0; \
7345 __ret = (poly8_t) __builtin_neon_vgetq_lane_i8((poly8x16_t)__s0, __p1); \
7346 __ret; \
7347 })
7348 #endif
7349
7350 #ifdef __LITTLE_ENDIAN__
7351 #define vgetq_lane_p16(__p0, __p1) __extension__ ({ \
7352 poly16_t __ret; \
7353 poly16x8_t __s0 = __p0; \
7354 __ret = (poly16_t) __builtin_neon_vgetq_lane_i16((poly16x8_t)__s0, __p1); \
7355 __ret; \
7356 })
7357 #else
7358 #define vgetq_lane_p16(__p0, __p1) __extension__ ({ \
7359 poly16_t __ret; \
7360 poly16x8_t __s0 = __p0; \
7361 poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
7362 __ret = (poly16_t) __builtin_neon_vgetq_lane_i16((poly16x8_t)__rev0, __p1); \
7363 __ret; \
7364 })
7365 #define __noswap_vgetq_lane_p16(__p0, __p1) __extension__ ({ \
7366 poly16_t __ret; \
7367 poly16x8_t __s0 = __p0; \
7368 __ret = (poly16_t) __builtin_neon_vgetq_lane_i16((poly16x8_t)__s0, __p1); \
7369 __ret; \
7370 })
7371 #endif
7372
7373 #ifdef __LITTLE_ENDIAN__
7374 #define vgetq_lane_u8(__p0, __p1) __extension__ ({ \
7375 uint8_t __ret; \
7376 uint8x16_t __s0 = __p0; \
7377 __ret = (uint8_t) __builtin_neon_vgetq_lane_i8((int8x16_t)__s0, __p1); \
7378 __ret; \
7379 })
7380 #else
7381 #define vgetq_lane_u8(__p0, __p1) __extension__ ({ \
7382 uint8_t __ret; \
7383 uint8x16_t __s0 = __p0; \
7384 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
7385 __ret = (uint8_t) __builtin_neon_vgetq_lane_i8((int8x16_t)__rev0, __p1); \
7386 __ret; \
7387 })
7388 #define __noswap_vgetq_lane_u8(__p0, __p1) __extension__ ({ \
7389 uint8_t __ret; \
7390 uint8x16_t __s0 = __p0; \
7391 __ret = (uint8_t) __builtin_neon_vgetq_lane_i8((int8x16_t)__s0, __p1); \
7392 __ret; \
7393 })
7394 #endif
7395
7396 #ifdef __LITTLE_ENDIAN__
7397 #define vgetq_lane_u32(__p0, __p1) __extension__ ({ \
7398 uint32_t __ret; \
7399 uint32x4_t __s0 = __p0; \
7400 __ret = (uint32_t) __builtin_neon_vgetq_lane_i32((int32x4_t)__s0, __p1); \
7401 __ret; \
7402 })
7403 #else
7404 #define vgetq_lane_u32(__p0, __p1) __extension__ ({ \
7405 uint32_t __ret; \
7406 uint32x4_t __s0 = __p0; \
7407 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
7408 __ret = (uint32_t) __builtin_neon_vgetq_lane_i32((int32x4_t)__rev0, __p1); \
7409 __ret; \
7410 })
7411 #define __noswap_vgetq_lane_u32(__p0, __p1) __extension__ ({ \
7412 uint32_t __ret; \
7413 uint32x4_t __s0 = __p0; \
7414 __ret = (uint32_t) __builtin_neon_vgetq_lane_i32((int32x4_t)__s0, __p1); \
7415 __ret; \
7416 })
7417 #endif
7418
7419 #ifdef __LITTLE_ENDIAN__
7420 #define vgetq_lane_u64(__p0, __p1) __extension__ ({ \
7421 uint64_t __ret; \
7422 uint64x2_t __s0 = __p0; \
7423 __ret = (uint64_t) __builtin_neon_vgetq_lane_i64((int64x2_t)__s0, __p1); \
7424 __ret; \
7425 })
7426 #else
7427 #define vgetq_lane_u64(__p0, __p1) __extension__ ({ \
7428 uint64_t __ret; \
7429 uint64x2_t __s0 = __p0; \
7430 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
7431 __ret = (uint64_t) __builtin_neon_vgetq_lane_i64((int64x2_t)__rev0, __p1); \
7432 __ret; \
7433 })
7434 #define __noswap_vgetq_lane_u64(__p0, __p1) __extension__ ({ \
7435 uint64_t __ret; \
7436 uint64x2_t __s0 = __p0; \
7437 __ret = (uint64_t) __builtin_neon_vgetq_lane_i64((int64x2_t)__s0, __p1); \
7438 __ret; \
7439 })
7440 #endif
7441
7442 #ifdef __LITTLE_ENDIAN__
7443 #define vgetq_lane_u16(__p0, __p1) __extension__ ({ \
7444 uint16_t __ret; \
7445 uint16x8_t __s0 = __p0; \
7446 __ret = (uint16_t) __builtin_neon_vgetq_lane_i16((int16x8_t)__s0, __p1); \
7447 __ret; \
7448 })
7449 #else
7450 #define vgetq_lane_u16(__p0, __p1) __extension__ ({ \
7451 uint16_t __ret; \
7452 uint16x8_t __s0 = __p0; \
7453 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
7454 __ret = (uint16_t) __builtin_neon_vgetq_lane_i16((int16x8_t)__rev0, __p1); \
7455 __ret; \
7456 })
7457 #define __noswap_vgetq_lane_u16(__p0, __p1) __extension__ ({ \
7458 uint16_t __ret; \
7459 uint16x8_t __s0 = __p0; \
7460 __ret = (uint16_t) __builtin_neon_vgetq_lane_i16((int16x8_t)__s0, __p1); \
7461 __ret; \
7462 })
7463 #endif
7464
7465 #ifdef __LITTLE_ENDIAN__
7466 #define vgetq_lane_s8(__p0, __p1) __extension__ ({ \
7467 int8_t __ret; \
7468 int8x16_t __s0 = __p0; \
7469 __ret = (int8_t) __builtin_neon_vgetq_lane_i8((int8x16_t)__s0, __p1); \
7470 __ret; \
7471 })
7472 #else
7473 #define vgetq_lane_s8(__p0, __p1) __extension__ ({ \
7474 int8_t __ret; \
7475 int8x16_t __s0 = __p0; \
7476 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
7477 __ret = (int8_t) __builtin_neon_vgetq_lane_i8((int8x16_t)__rev0, __p1); \
7478 __ret; \
7479 })
7480 #define __noswap_vgetq_lane_s8(__p0, __p1) __extension__ ({ \
7481 int8_t __ret; \
7482 int8x16_t __s0 = __p0; \
7483 __ret = (int8_t) __builtin_neon_vgetq_lane_i8((int8x16_t)__s0, __p1); \
7484 __ret; \
7485 })
7486 #endif
7487
7488 #ifdef __LITTLE_ENDIAN__
7489 #define vgetq_lane_f32(__p0, __p1) __extension__ ({ \
7490 float32_t __ret; \
7491 float32x4_t __s0 = __p0; \
7492 __ret = (float32_t) __builtin_neon_vgetq_lane_f32((float32x4_t)__s0, __p1); \
7493 __ret; \
7494 })
7495 #else
7496 #define vgetq_lane_f32(__p0, __p1) __extension__ ({ \
7497 float32_t __ret; \
7498 float32x4_t __s0 = __p0; \
7499 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
7500 __ret = (float32_t) __builtin_neon_vgetq_lane_f32((float32x4_t)__rev0, __p1); \
7501 __ret; \
7502 })
7503 #define __noswap_vgetq_lane_f32(__p0, __p1) __extension__ ({ \
7504 float32_t __ret; \
7505 float32x4_t __s0 = __p0; \
7506 __ret = (float32_t) __builtin_neon_vgetq_lane_f32((float32x4_t)__s0, __p1); \
7507 __ret; \
7508 })
7509 #endif
7510
7511 #ifdef __LITTLE_ENDIAN__
7512 #define vgetq_lane_s32(__p0, __p1) __extension__ ({ \
7513 int32_t __ret; \
7514 int32x4_t __s0 = __p0; \
7515 __ret = (int32_t) __builtin_neon_vgetq_lane_i32((int32x4_t)__s0, __p1); \
7516 __ret; \
7517 })
7518 #else
7519 #define vgetq_lane_s32(__p0, __p1) __extension__ ({ \
7520 int32_t __ret; \
7521 int32x4_t __s0 = __p0; \
7522 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
7523 __ret = (int32_t) __builtin_neon_vgetq_lane_i32((int32x4_t)__rev0, __p1); \
7524 __ret; \
7525 })
7526 #define __noswap_vgetq_lane_s32(__p0, __p1) __extension__ ({ \
7527 int32_t __ret; \
7528 int32x4_t __s0 = __p0; \
7529 __ret = (int32_t) __builtin_neon_vgetq_lane_i32((int32x4_t)__s0, __p1); \
7530 __ret; \
7531 })
7532 #endif
7533
7534 #ifdef __LITTLE_ENDIAN__
7535 #define vgetq_lane_s64(__p0, __p1) __extension__ ({ \
7536 int64_t __ret; \
7537 int64x2_t __s0 = __p0; \
7538 __ret = (int64_t) __builtin_neon_vgetq_lane_i64((int64x2_t)__s0, __p1); \
7539 __ret; \
7540 })
7541 #else
7542 #define vgetq_lane_s64(__p0, __p1) __extension__ ({ \
7543 int64_t __ret; \
7544 int64x2_t __s0 = __p0; \
7545 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
7546 __ret = (int64_t) __builtin_neon_vgetq_lane_i64((int64x2_t)__rev0, __p1); \
7547 __ret; \
7548 })
7549 #define __noswap_vgetq_lane_s64(__p0, __p1) __extension__ ({ \
7550 int64_t __ret; \
7551 int64x2_t __s0 = __p0; \
7552 __ret = (int64_t) __builtin_neon_vgetq_lane_i64((int64x2_t)__s0, __p1); \
7553 __ret; \
7554 })
7555 #endif
7556
7557 #ifdef __LITTLE_ENDIAN__
7558 #define vgetq_lane_s16(__p0, __p1) __extension__ ({ \
7559 int16_t __ret; \
7560 int16x8_t __s0 = __p0; \
7561 __ret = (int16_t) __builtin_neon_vgetq_lane_i16((int16x8_t)__s0, __p1); \
7562 __ret; \
7563 })
7564 #else
7565 #define vgetq_lane_s16(__p0, __p1) __extension__ ({ \
7566 int16_t __ret; \
7567 int16x8_t __s0 = __p0; \
7568 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
7569 __ret = (int16_t) __builtin_neon_vgetq_lane_i16((int16x8_t)__rev0, __p1); \
7570 __ret; \
7571 })
7572 #define __noswap_vgetq_lane_s16(__p0, __p1) __extension__ ({ \
7573 int16_t __ret; \
7574 int16x8_t __s0 = __p0; \
7575 __ret = (int16_t) __builtin_neon_vgetq_lane_i16((int16x8_t)__s0, __p1); \
7576 __ret; \
7577 })
7578 #endif
7579
7580 #ifdef __LITTLE_ENDIAN__
7581 #define vget_lane_u8(__p0, __p1) __extension__ ({ \
7582 uint8_t __ret; \
7583 uint8x8_t __s0 = __p0; \
7584 __ret = (uint8_t) __builtin_neon_vget_lane_i8((int8x8_t)__s0, __p1); \
7585 __ret; \
7586 })
7587 #else
7588 #define vget_lane_u8(__p0, __p1) __extension__ ({ \
7589 uint8_t __ret; \
7590 uint8x8_t __s0 = __p0; \
7591 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
7592 __ret = (uint8_t) __builtin_neon_vget_lane_i8((int8x8_t)__rev0, __p1); \
7593 __ret; \
7594 })
7595 #define __noswap_vget_lane_u8(__p0, __p1) __extension__ ({ \
7596 uint8_t __ret; \
7597 uint8x8_t __s0 = __p0; \
7598 __ret = (uint8_t) __builtin_neon_vget_lane_i8((int8x8_t)__s0, __p1); \
7599 __ret; \
7600 })
7601 #endif
7602
7603 #ifdef __LITTLE_ENDIAN__
7604 #define vget_lane_u32(__p0, __p1) __extension__ ({ \
7605 uint32_t __ret; \
7606 uint32x2_t __s0 = __p0; \
7607 __ret = (uint32_t) __builtin_neon_vget_lane_i32((int32x2_t)__s0, __p1); \
7608 __ret; \
7609 })
7610 #else
7611 #define vget_lane_u32(__p0, __p1) __extension__ ({ \
7612 uint32_t __ret; \
7613 uint32x2_t __s0 = __p0; \
7614 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
7615 __ret = (uint32_t) __builtin_neon_vget_lane_i32((int32x2_t)__rev0, __p1); \
7616 __ret; \
7617 })
7618 #define __noswap_vget_lane_u32(__p0, __p1) __extension__ ({ \
7619 uint32_t __ret; \
7620 uint32x2_t __s0 = __p0; \
7621 __ret = (uint32_t) __builtin_neon_vget_lane_i32((int32x2_t)__s0, __p1); \
7622 __ret; \
7623 })
7624 #endif
7625
7626 #define vget_lane_u64(__p0, __p1) __extension__ ({ \
7627 uint64_t __ret; \
7628 uint64x1_t __s0 = __p0; \
7629 __ret = (uint64_t) __builtin_neon_vget_lane_i64((int64x1_t)__s0, __p1); \
7630 __ret; \
7631 })
7632 #ifdef __LITTLE_ENDIAN__
7633 #define vget_lane_u16(__p0, __p1) __extension__ ({ \
7634 uint16_t __ret; \
7635 uint16x4_t __s0 = __p0; \
7636 __ret = (uint16_t) __builtin_neon_vget_lane_i16((int16x4_t)__s0, __p1); \
7637 __ret; \
7638 })
7639 #else
7640 #define vget_lane_u16(__p0, __p1) __extension__ ({ \
7641 uint16_t __ret; \
7642 uint16x4_t __s0 = __p0; \
7643 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
7644 __ret = (uint16_t) __builtin_neon_vget_lane_i16((int16x4_t)__rev0, __p1); \
7645 __ret; \
7646 })
7647 #define __noswap_vget_lane_u16(__p0, __p1) __extension__ ({ \
7648 uint16_t __ret; \
7649 uint16x4_t __s0 = __p0; \
7650 __ret = (uint16_t) __builtin_neon_vget_lane_i16((int16x4_t)__s0, __p1); \
7651 __ret; \
7652 })
7653 #endif
7654
7655 #ifdef __LITTLE_ENDIAN__
7656 #define vget_lane_s8(__p0, __p1) __extension__ ({ \
7657 int8_t __ret; \
7658 int8x8_t __s0 = __p0; \
7659 __ret = (int8_t) __builtin_neon_vget_lane_i8((int8x8_t)__s0, __p1); \
7660 __ret; \
7661 })
7662 #else
7663 #define vget_lane_s8(__p0, __p1) __extension__ ({ \
7664 int8_t __ret; \
7665 int8x8_t __s0 = __p0; \
7666 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
7667 __ret = (int8_t) __builtin_neon_vget_lane_i8((int8x8_t)__rev0, __p1); \
7668 __ret; \
7669 })
7670 #define __noswap_vget_lane_s8(__p0, __p1) __extension__ ({ \
7671 int8_t __ret; \
7672 int8x8_t __s0 = __p0; \
7673 __ret = (int8_t) __builtin_neon_vget_lane_i8((int8x8_t)__s0, __p1); \
7674 __ret; \
7675 })
7676 #endif
7677
7678 #ifdef __LITTLE_ENDIAN__
7679 #define vget_lane_f32(__p0, __p1) __extension__ ({ \
7680 float32_t __ret; \
7681 float32x2_t __s0 = __p0; \
7682 __ret = (float32_t) __builtin_neon_vget_lane_f32((float32x2_t)__s0, __p1); \
7683 __ret; \
7684 })
7685 #else
7686 #define vget_lane_f32(__p0, __p1) __extension__ ({ \
7687 float32_t __ret; \
7688 float32x2_t __s0 = __p0; \
7689 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
7690 __ret = (float32_t) __builtin_neon_vget_lane_f32((float32x2_t)__rev0, __p1); \
7691 __ret; \
7692 })
7693 #define __noswap_vget_lane_f32(__p0, __p1) __extension__ ({ \
7694 float32_t __ret; \
7695 float32x2_t __s0 = __p0; \
7696 __ret = (float32_t) __builtin_neon_vget_lane_f32((float32x2_t)__s0, __p1); \
7697 __ret; \
7698 })
7699 #endif
7700
7701 #ifdef __LITTLE_ENDIAN__
7702 #define vget_lane_s32(__p0, __p1) __extension__ ({ \
7703 int32_t __ret; \
7704 int32x2_t __s0 = __p0; \
7705 __ret = (int32_t) __builtin_neon_vget_lane_i32((int32x2_t)__s0, __p1); \
7706 __ret; \
7707 })
7708 #else
7709 #define vget_lane_s32(__p0, __p1) __extension__ ({ \
7710 int32_t __ret; \
7711 int32x2_t __s0 = __p0; \
7712 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
7713 __ret = (int32_t) __builtin_neon_vget_lane_i32((int32x2_t)__rev0, __p1); \
7714 __ret; \
7715 })
7716 #define __noswap_vget_lane_s32(__p0, __p1) __extension__ ({ \
7717 int32_t __ret; \
7718 int32x2_t __s0 = __p0; \
7719 __ret = (int32_t) __builtin_neon_vget_lane_i32((int32x2_t)__s0, __p1); \
7720 __ret; \
7721 })
7722 #endif
7723
7724 #define vget_lane_s64(__p0, __p1) __extension__ ({ \
7725 int64_t __ret; \
7726 int64x1_t __s0 = __p0; \
7727 __ret = (int64_t) __builtin_neon_vget_lane_i64((int64x1_t)__s0, __p1); \
7728 __ret; \
7729 })
7730 #ifdef __LITTLE_ENDIAN__
7731 #define vget_lane_s16(__p0, __p1) __extension__ ({ \
7732 int16_t __ret; \
7733 int16x4_t __s0 = __p0; \
7734 __ret = (int16_t) __builtin_neon_vget_lane_i16((int16x4_t)__s0, __p1); \
7735 __ret; \
7736 })
7737 #else
7738 #define vget_lane_s16(__p0, __p1) __extension__ ({ \
7739 int16_t __ret; \
7740 int16x4_t __s0 = __p0; \
7741 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
7742 __ret = (int16_t) __builtin_neon_vget_lane_i16((int16x4_t)__rev0, __p1); \
7743 __ret; \
7744 })
7745 #define __noswap_vget_lane_s16(__p0, __p1) __extension__ ({ \
7746 int16_t __ret; \
7747 int16x4_t __s0 = __p0; \
7748 __ret = (int16_t) __builtin_neon_vget_lane_i16((int16x4_t)__s0, __p1); \
7749 __ret; \
7750 })
7751 #endif
7752
7753 #ifdef __LITTLE_ENDIAN__
vget_low_p8(poly8x16_t __p0)7754 __ai poly8x8_t vget_low_p8(poly8x16_t __p0) {
7755 poly8x8_t __ret;
7756 __ret = __builtin_shufflevector(__p0, __p0, 0, 1, 2, 3, 4, 5, 6, 7);
7757 return __ret;
7758 }
7759 #else
vget_low_p8(poly8x16_t __p0)7760 __ai poly8x8_t vget_low_p8(poly8x16_t __p0) {
7761 poly8x8_t __ret;
7762 poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
7763 __ret = __builtin_shufflevector(__rev0, __rev0, 0, 1, 2, 3, 4, 5, 6, 7);
7764 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
7765 return __ret;
7766 }
7767 #endif
7768
7769 #ifdef __LITTLE_ENDIAN__
vget_low_p16(poly16x8_t __p0)7770 __ai poly16x4_t vget_low_p16(poly16x8_t __p0) {
7771 poly16x4_t __ret;
7772 __ret = __builtin_shufflevector(__p0, __p0, 0, 1, 2, 3);
7773 return __ret;
7774 }
7775 #else
vget_low_p16(poly16x8_t __p0)7776 __ai poly16x4_t vget_low_p16(poly16x8_t __p0) {
7777 poly16x4_t __ret;
7778 poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
7779 __ret = __builtin_shufflevector(__rev0, __rev0, 0, 1, 2, 3);
7780 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
7781 return __ret;
7782 }
7783 #endif
7784
7785 #ifdef __LITTLE_ENDIAN__
vget_low_u8(uint8x16_t __p0)7786 __ai uint8x8_t vget_low_u8(uint8x16_t __p0) {
7787 uint8x8_t __ret;
7788 __ret = __builtin_shufflevector(__p0, __p0, 0, 1, 2, 3, 4, 5, 6, 7);
7789 return __ret;
7790 }
7791 #else
vget_low_u8(uint8x16_t __p0)7792 __ai uint8x8_t vget_low_u8(uint8x16_t __p0) {
7793 uint8x8_t __ret;
7794 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
7795 __ret = __builtin_shufflevector(__rev0, __rev0, 0, 1, 2, 3, 4, 5, 6, 7);
7796 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
7797 return __ret;
7798 }
7799 #endif
7800
7801 #ifdef __LITTLE_ENDIAN__
vget_low_u32(uint32x4_t __p0)7802 __ai uint32x2_t vget_low_u32(uint32x4_t __p0) {
7803 uint32x2_t __ret;
7804 __ret = __builtin_shufflevector(__p0, __p0, 0, 1);
7805 return __ret;
7806 }
7807 #else
vget_low_u32(uint32x4_t __p0)7808 __ai uint32x2_t vget_low_u32(uint32x4_t __p0) {
7809 uint32x2_t __ret;
7810 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
7811 __ret = __builtin_shufflevector(__rev0, __rev0, 0, 1);
7812 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
7813 return __ret;
7814 }
7815 #endif
7816
7817 #ifdef __LITTLE_ENDIAN__
vget_low_u64(uint64x2_t __p0)7818 __ai uint64x1_t vget_low_u64(uint64x2_t __p0) {
7819 uint64x1_t __ret;
7820 __ret = __builtin_shufflevector(__p0, __p0, 0);
7821 return __ret;
7822 }
7823 #else
vget_low_u64(uint64x2_t __p0)7824 __ai uint64x1_t vget_low_u64(uint64x2_t __p0) {
7825 uint64x1_t __ret;
7826 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
7827 __ret = __builtin_shufflevector(__rev0, __rev0, 0);
7828 return __ret;
7829 }
7830 #endif
7831
7832 #ifdef __LITTLE_ENDIAN__
vget_low_u16(uint16x8_t __p0)7833 __ai uint16x4_t vget_low_u16(uint16x8_t __p0) {
7834 uint16x4_t __ret;
7835 __ret = __builtin_shufflevector(__p0, __p0, 0, 1, 2, 3);
7836 return __ret;
7837 }
7838 #else
vget_low_u16(uint16x8_t __p0)7839 __ai uint16x4_t vget_low_u16(uint16x8_t __p0) {
7840 uint16x4_t __ret;
7841 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
7842 __ret = __builtin_shufflevector(__rev0, __rev0, 0, 1, 2, 3);
7843 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
7844 return __ret;
7845 }
7846 #endif
7847
7848 #ifdef __LITTLE_ENDIAN__
vget_low_s8(int8x16_t __p0)7849 __ai int8x8_t vget_low_s8(int8x16_t __p0) {
7850 int8x8_t __ret;
7851 __ret = __builtin_shufflevector(__p0, __p0, 0, 1, 2, 3, 4, 5, 6, 7);
7852 return __ret;
7853 }
7854 #else
vget_low_s8(int8x16_t __p0)7855 __ai int8x8_t vget_low_s8(int8x16_t __p0) {
7856 int8x8_t __ret;
7857 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
7858 __ret = __builtin_shufflevector(__rev0, __rev0, 0, 1, 2, 3, 4, 5, 6, 7);
7859 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
7860 return __ret;
7861 }
7862 #endif
7863
7864 #ifdef __LITTLE_ENDIAN__
vget_low_f32(float32x4_t __p0)7865 __ai float32x2_t vget_low_f32(float32x4_t __p0) {
7866 float32x2_t __ret;
7867 __ret = __builtin_shufflevector(__p0, __p0, 0, 1);
7868 return __ret;
7869 }
7870 #else
vget_low_f32(float32x4_t __p0)7871 __ai float32x2_t vget_low_f32(float32x4_t __p0) {
7872 float32x2_t __ret;
7873 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
7874 __ret = __builtin_shufflevector(__rev0, __rev0, 0, 1);
7875 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
7876 return __ret;
7877 }
7878 #endif
7879
7880 #ifdef __LITTLE_ENDIAN__
vget_low_f16(float16x8_t __p0)7881 __ai float16x4_t vget_low_f16(float16x8_t __p0) {
7882 float16x4_t __ret;
7883 __ret = __builtin_shufflevector(__p0, __p0, 0, 1, 2, 3);
7884 return __ret;
7885 }
7886 #else
vget_low_f16(float16x8_t __p0)7887 __ai float16x4_t vget_low_f16(float16x8_t __p0) {
7888 float16x4_t __ret;
7889 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
7890 __ret = __builtin_shufflevector(__rev0, __rev0, 0, 1, 2, 3);
7891 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
7892 return __ret;
7893 }
7894 #endif
7895
7896 #ifdef __LITTLE_ENDIAN__
vget_low_s32(int32x4_t __p0)7897 __ai int32x2_t vget_low_s32(int32x4_t __p0) {
7898 int32x2_t __ret;
7899 __ret = __builtin_shufflevector(__p0, __p0, 0, 1);
7900 return __ret;
7901 }
7902 #else
vget_low_s32(int32x4_t __p0)7903 __ai int32x2_t vget_low_s32(int32x4_t __p0) {
7904 int32x2_t __ret;
7905 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
7906 __ret = __builtin_shufflevector(__rev0, __rev0, 0, 1);
7907 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
7908 return __ret;
7909 }
7910 #endif
7911
7912 #ifdef __LITTLE_ENDIAN__
vget_low_s64(int64x2_t __p0)7913 __ai int64x1_t vget_low_s64(int64x2_t __p0) {
7914 int64x1_t __ret;
7915 __ret = __builtin_shufflevector(__p0, __p0, 0);
7916 return __ret;
7917 }
7918 #else
vget_low_s64(int64x2_t __p0)7919 __ai int64x1_t vget_low_s64(int64x2_t __p0) {
7920 int64x1_t __ret;
7921 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
7922 __ret = __builtin_shufflevector(__rev0, __rev0, 0);
7923 return __ret;
7924 }
7925 #endif
7926
7927 #ifdef __LITTLE_ENDIAN__
vget_low_s16(int16x8_t __p0)7928 __ai int16x4_t vget_low_s16(int16x8_t __p0) {
7929 int16x4_t __ret;
7930 __ret = __builtin_shufflevector(__p0, __p0, 0, 1, 2, 3);
7931 return __ret;
7932 }
7933 #else
vget_low_s16(int16x8_t __p0)7934 __ai int16x4_t vget_low_s16(int16x8_t __p0) {
7935 int16x4_t __ret;
7936 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
7937 __ret = __builtin_shufflevector(__rev0, __rev0, 0, 1, 2, 3);
7938 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
7939 return __ret;
7940 }
7941 #endif
7942
7943 #ifdef __LITTLE_ENDIAN__
vhaddq_u8(uint8x16_t __p0,uint8x16_t __p1)7944 __ai uint8x16_t vhaddq_u8(uint8x16_t __p0, uint8x16_t __p1) {
7945 uint8x16_t __ret;
7946 __ret = (uint8x16_t) __builtin_neon_vhaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
7947 return __ret;
7948 }
7949 #else
vhaddq_u8(uint8x16_t __p0,uint8x16_t __p1)7950 __ai uint8x16_t vhaddq_u8(uint8x16_t __p0, uint8x16_t __p1) {
7951 uint8x16_t __ret;
7952 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
7953 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
7954 __ret = (uint8x16_t) __builtin_neon_vhaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
7955 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
7956 return __ret;
7957 }
7958 #endif
7959
7960 #ifdef __LITTLE_ENDIAN__
vhaddq_u32(uint32x4_t __p0,uint32x4_t __p1)7961 __ai uint32x4_t vhaddq_u32(uint32x4_t __p0, uint32x4_t __p1) {
7962 uint32x4_t __ret;
7963 __ret = (uint32x4_t) __builtin_neon_vhaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
7964 return __ret;
7965 }
7966 #else
vhaddq_u32(uint32x4_t __p0,uint32x4_t __p1)7967 __ai uint32x4_t vhaddq_u32(uint32x4_t __p0, uint32x4_t __p1) {
7968 uint32x4_t __ret;
7969 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
7970 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
7971 __ret = (uint32x4_t) __builtin_neon_vhaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
7972 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
7973 return __ret;
7974 }
7975 #endif
7976
7977 #ifdef __LITTLE_ENDIAN__
vhaddq_u16(uint16x8_t __p0,uint16x8_t __p1)7978 __ai uint16x8_t vhaddq_u16(uint16x8_t __p0, uint16x8_t __p1) {
7979 uint16x8_t __ret;
7980 __ret = (uint16x8_t) __builtin_neon_vhaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
7981 return __ret;
7982 }
7983 #else
vhaddq_u16(uint16x8_t __p0,uint16x8_t __p1)7984 __ai uint16x8_t vhaddq_u16(uint16x8_t __p0, uint16x8_t __p1) {
7985 uint16x8_t __ret;
7986 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
7987 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
7988 __ret = (uint16x8_t) __builtin_neon_vhaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
7989 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
7990 return __ret;
7991 }
7992 #endif
7993
7994 #ifdef __LITTLE_ENDIAN__
vhaddq_s8(int8x16_t __p0,int8x16_t __p1)7995 __ai int8x16_t vhaddq_s8(int8x16_t __p0, int8x16_t __p1) {
7996 int8x16_t __ret;
7997 __ret = (int8x16_t) __builtin_neon_vhaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 32);
7998 return __ret;
7999 }
8000 #else
vhaddq_s8(int8x16_t __p0,int8x16_t __p1)8001 __ai int8x16_t vhaddq_s8(int8x16_t __p0, int8x16_t __p1) {
8002 int8x16_t __ret;
8003 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
8004 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
8005 __ret = (int8x16_t) __builtin_neon_vhaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 32);
8006 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
8007 return __ret;
8008 }
8009 #endif
8010
8011 #ifdef __LITTLE_ENDIAN__
vhaddq_s32(int32x4_t __p0,int32x4_t __p1)8012 __ai int32x4_t vhaddq_s32(int32x4_t __p0, int32x4_t __p1) {
8013 int32x4_t __ret;
8014 __ret = (int32x4_t) __builtin_neon_vhaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
8015 return __ret;
8016 }
8017 #else
vhaddq_s32(int32x4_t __p0,int32x4_t __p1)8018 __ai int32x4_t vhaddq_s32(int32x4_t __p0, int32x4_t __p1) {
8019 int32x4_t __ret;
8020 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
8021 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
8022 __ret = (int32x4_t) __builtin_neon_vhaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
8023 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
8024 return __ret;
8025 }
8026 #endif
8027
8028 #ifdef __LITTLE_ENDIAN__
vhaddq_s16(int16x8_t __p0,int16x8_t __p1)8029 __ai int16x8_t vhaddq_s16(int16x8_t __p0, int16x8_t __p1) {
8030 int16x8_t __ret;
8031 __ret = (int16x8_t) __builtin_neon_vhaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
8032 return __ret;
8033 }
8034 #else
vhaddq_s16(int16x8_t __p0,int16x8_t __p1)8035 __ai int16x8_t vhaddq_s16(int16x8_t __p0, int16x8_t __p1) {
8036 int16x8_t __ret;
8037 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
8038 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
8039 __ret = (int16x8_t) __builtin_neon_vhaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
8040 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
8041 return __ret;
8042 }
8043 #endif
8044
8045 #ifdef __LITTLE_ENDIAN__
vhadd_u8(uint8x8_t __p0,uint8x8_t __p1)8046 __ai uint8x8_t vhadd_u8(uint8x8_t __p0, uint8x8_t __p1) {
8047 uint8x8_t __ret;
8048 __ret = (uint8x8_t) __builtin_neon_vhadd_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
8049 return __ret;
8050 }
8051 #else
vhadd_u8(uint8x8_t __p0,uint8x8_t __p1)8052 __ai uint8x8_t vhadd_u8(uint8x8_t __p0, uint8x8_t __p1) {
8053 uint8x8_t __ret;
8054 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
8055 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
8056 __ret = (uint8x8_t) __builtin_neon_vhadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
8057 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
8058 return __ret;
8059 }
8060 #endif
8061
8062 #ifdef __LITTLE_ENDIAN__
vhadd_u32(uint32x2_t __p0,uint32x2_t __p1)8063 __ai uint32x2_t vhadd_u32(uint32x2_t __p0, uint32x2_t __p1) {
8064 uint32x2_t __ret;
8065 __ret = (uint32x2_t) __builtin_neon_vhadd_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
8066 return __ret;
8067 }
8068 #else
vhadd_u32(uint32x2_t __p0,uint32x2_t __p1)8069 __ai uint32x2_t vhadd_u32(uint32x2_t __p0, uint32x2_t __p1) {
8070 uint32x2_t __ret;
8071 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
8072 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
8073 __ret = (uint32x2_t) __builtin_neon_vhadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
8074 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
8075 return __ret;
8076 }
8077 #endif
8078
8079 #ifdef __LITTLE_ENDIAN__
vhadd_u16(uint16x4_t __p0,uint16x4_t __p1)8080 __ai uint16x4_t vhadd_u16(uint16x4_t __p0, uint16x4_t __p1) {
8081 uint16x4_t __ret;
8082 __ret = (uint16x4_t) __builtin_neon_vhadd_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
8083 return __ret;
8084 }
8085 #else
vhadd_u16(uint16x4_t __p0,uint16x4_t __p1)8086 __ai uint16x4_t vhadd_u16(uint16x4_t __p0, uint16x4_t __p1) {
8087 uint16x4_t __ret;
8088 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
8089 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
8090 __ret = (uint16x4_t) __builtin_neon_vhadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
8091 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
8092 return __ret;
8093 }
8094 #endif
8095
8096 #ifdef __LITTLE_ENDIAN__
vhadd_s8(int8x8_t __p0,int8x8_t __p1)8097 __ai int8x8_t vhadd_s8(int8x8_t __p0, int8x8_t __p1) {
8098 int8x8_t __ret;
8099 __ret = (int8x8_t) __builtin_neon_vhadd_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
8100 return __ret;
8101 }
8102 #else
vhadd_s8(int8x8_t __p0,int8x8_t __p1)8103 __ai int8x8_t vhadd_s8(int8x8_t __p0, int8x8_t __p1) {
8104 int8x8_t __ret;
8105 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
8106 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
8107 __ret = (int8x8_t) __builtin_neon_vhadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 0);
8108 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
8109 return __ret;
8110 }
8111 #endif
8112
8113 #ifdef __LITTLE_ENDIAN__
vhadd_s32(int32x2_t __p0,int32x2_t __p1)8114 __ai int32x2_t vhadd_s32(int32x2_t __p0, int32x2_t __p1) {
8115 int32x2_t __ret;
8116 __ret = (int32x2_t) __builtin_neon_vhadd_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
8117 return __ret;
8118 }
8119 #else
vhadd_s32(int32x2_t __p0,int32x2_t __p1)8120 __ai int32x2_t vhadd_s32(int32x2_t __p0, int32x2_t __p1) {
8121 int32x2_t __ret;
8122 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
8123 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
8124 __ret = (int32x2_t) __builtin_neon_vhadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
8125 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
8126 return __ret;
8127 }
8128 #endif
8129
8130 #ifdef __LITTLE_ENDIAN__
vhadd_s16(int16x4_t __p0,int16x4_t __p1)8131 __ai int16x4_t vhadd_s16(int16x4_t __p0, int16x4_t __p1) {
8132 int16x4_t __ret;
8133 __ret = (int16x4_t) __builtin_neon_vhadd_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
8134 return __ret;
8135 }
8136 #else
vhadd_s16(int16x4_t __p0,int16x4_t __p1)8137 __ai int16x4_t vhadd_s16(int16x4_t __p0, int16x4_t __p1) {
8138 int16x4_t __ret;
8139 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
8140 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
8141 __ret = (int16x4_t) __builtin_neon_vhadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
8142 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
8143 return __ret;
8144 }
8145 #endif
8146
8147 #ifdef __LITTLE_ENDIAN__
vhsubq_u8(uint8x16_t __p0,uint8x16_t __p1)8148 __ai uint8x16_t vhsubq_u8(uint8x16_t __p0, uint8x16_t __p1) {
8149 uint8x16_t __ret;
8150 __ret = (uint8x16_t) __builtin_neon_vhsubq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
8151 return __ret;
8152 }
8153 #else
vhsubq_u8(uint8x16_t __p0,uint8x16_t __p1)8154 __ai uint8x16_t vhsubq_u8(uint8x16_t __p0, uint8x16_t __p1) {
8155 uint8x16_t __ret;
8156 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
8157 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
8158 __ret = (uint8x16_t) __builtin_neon_vhsubq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
8159 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
8160 return __ret;
8161 }
8162 #endif
8163
8164 #ifdef __LITTLE_ENDIAN__
vhsubq_u32(uint32x4_t __p0,uint32x4_t __p1)8165 __ai uint32x4_t vhsubq_u32(uint32x4_t __p0, uint32x4_t __p1) {
8166 uint32x4_t __ret;
8167 __ret = (uint32x4_t) __builtin_neon_vhsubq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
8168 return __ret;
8169 }
8170 #else
vhsubq_u32(uint32x4_t __p0,uint32x4_t __p1)8171 __ai uint32x4_t vhsubq_u32(uint32x4_t __p0, uint32x4_t __p1) {
8172 uint32x4_t __ret;
8173 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
8174 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
8175 __ret = (uint32x4_t) __builtin_neon_vhsubq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
8176 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
8177 return __ret;
8178 }
8179 #endif
8180
8181 #ifdef __LITTLE_ENDIAN__
vhsubq_u16(uint16x8_t __p0,uint16x8_t __p1)8182 __ai uint16x8_t vhsubq_u16(uint16x8_t __p0, uint16x8_t __p1) {
8183 uint16x8_t __ret;
8184 __ret = (uint16x8_t) __builtin_neon_vhsubq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
8185 return __ret;
8186 }
8187 #else
vhsubq_u16(uint16x8_t __p0,uint16x8_t __p1)8188 __ai uint16x8_t vhsubq_u16(uint16x8_t __p0, uint16x8_t __p1) {
8189 uint16x8_t __ret;
8190 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
8191 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
8192 __ret = (uint16x8_t) __builtin_neon_vhsubq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
8193 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
8194 return __ret;
8195 }
8196 #endif
8197
8198 #ifdef __LITTLE_ENDIAN__
vhsubq_s8(int8x16_t __p0,int8x16_t __p1)8199 __ai int8x16_t vhsubq_s8(int8x16_t __p0, int8x16_t __p1) {
8200 int8x16_t __ret;
8201 __ret = (int8x16_t) __builtin_neon_vhsubq_v((int8x16_t)__p0, (int8x16_t)__p1, 32);
8202 return __ret;
8203 }
8204 #else
vhsubq_s8(int8x16_t __p0,int8x16_t __p1)8205 __ai int8x16_t vhsubq_s8(int8x16_t __p0, int8x16_t __p1) {
8206 int8x16_t __ret;
8207 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
8208 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
8209 __ret = (int8x16_t) __builtin_neon_vhsubq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 32);
8210 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
8211 return __ret;
8212 }
8213 #endif
8214
8215 #ifdef __LITTLE_ENDIAN__
vhsubq_s32(int32x4_t __p0,int32x4_t __p1)8216 __ai int32x4_t vhsubq_s32(int32x4_t __p0, int32x4_t __p1) {
8217 int32x4_t __ret;
8218 __ret = (int32x4_t) __builtin_neon_vhsubq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
8219 return __ret;
8220 }
8221 #else
vhsubq_s32(int32x4_t __p0,int32x4_t __p1)8222 __ai int32x4_t vhsubq_s32(int32x4_t __p0, int32x4_t __p1) {
8223 int32x4_t __ret;
8224 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
8225 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
8226 __ret = (int32x4_t) __builtin_neon_vhsubq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
8227 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
8228 return __ret;
8229 }
8230 #endif
8231
8232 #ifdef __LITTLE_ENDIAN__
vhsubq_s16(int16x8_t __p0,int16x8_t __p1)8233 __ai int16x8_t vhsubq_s16(int16x8_t __p0, int16x8_t __p1) {
8234 int16x8_t __ret;
8235 __ret = (int16x8_t) __builtin_neon_vhsubq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
8236 return __ret;
8237 }
8238 #else
vhsubq_s16(int16x8_t __p0,int16x8_t __p1)8239 __ai int16x8_t vhsubq_s16(int16x8_t __p0, int16x8_t __p1) {
8240 int16x8_t __ret;
8241 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
8242 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
8243 __ret = (int16x8_t) __builtin_neon_vhsubq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
8244 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
8245 return __ret;
8246 }
8247 #endif
8248
8249 #ifdef __LITTLE_ENDIAN__
vhsub_u8(uint8x8_t __p0,uint8x8_t __p1)8250 __ai uint8x8_t vhsub_u8(uint8x8_t __p0, uint8x8_t __p1) {
8251 uint8x8_t __ret;
8252 __ret = (uint8x8_t) __builtin_neon_vhsub_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
8253 return __ret;
8254 }
8255 #else
vhsub_u8(uint8x8_t __p0,uint8x8_t __p1)8256 __ai uint8x8_t vhsub_u8(uint8x8_t __p0, uint8x8_t __p1) {
8257 uint8x8_t __ret;
8258 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
8259 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
8260 __ret = (uint8x8_t) __builtin_neon_vhsub_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
8261 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
8262 return __ret;
8263 }
8264 #endif
8265
8266 #ifdef __LITTLE_ENDIAN__
vhsub_u32(uint32x2_t __p0,uint32x2_t __p1)8267 __ai uint32x2_t vhsub_u32(uint32x2_t __p0, uint32x2_t __p1) {
8268 uint32x2_t __ret;
8269 __ret = (uint32x2_t) __builtin_neon_vhsub_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
8270 return __ret;
8271 }
8272 #else
vhsub_u32(uint32x2_t __p0,uint32x2_t __p1)8273 __ai uint32x2_t vhsub_u32(uint32x2_t __p0, uint32x2_t __p1) {
8274 uint32x2_t __ret;
8275 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
8276 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
8277 __ret = (uint32x2_t) __builtin_neon_vhsub_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
8278 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
8279 return __ret;
8280 }
8281 #endif
8282
8283 #ifdef __LITTLE_ENDIAN__
vhsub_u16(uint16x4_t __p0,uint16x4_t __p1)8284 __ai uint16x4_t vhsub_u16(uint16x4_t __p0, uint16x4_t __p1) {
8285 uint16x4_t __ret;
8286 __ret = (uint16x4_t) __builtin_neon_vhsub_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
8287 return __ret;
8288 }
8289 #else
vhsub_u16(uint16x4_t __p0,uint16x4_t __p1)8290 __ai uint16x4_t vhsub_u16(uint16x4_t __p0, uint16x4_t __p1) {
8291 uint16x4_t __ret;
8292 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
8293 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
8294 __ret = (uint16x4_t) __builtin_neon_vhsub_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
8295 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
8296 return __ret;
8297 }
8298 #endif
8299
8300 #ifdef __LITTLE_ENDIAN__
vhsub_s8(int8x8_t __p0,int8x8_t __p1)8301 __ai int8x8_t vhsub_s8(int8x8_t __p0, int8x8_t __p1) {
8302 int8x8_t __ret;
8303 __ret = (int8x8_t) __builtin_neon_vhsub_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
8304 return __ret;
8305 }
8306 #else
vhsub_s8(int8x8_t __p0,int8x8_t __p1)8307 __ai int8x8_t vhsub_s8(int8x8_t __p0, int8x8_t __p1) {
8308 int8x8_t __ret;
8309 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
8310 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
8311 __ret = (int8x8_t) __builtin_neon_vhsub_v((int8x8_t)__rev0, (int8x8_t)__rev1, 0);
8312 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
8313 return __ret;
8314 }
8315 #endif
8316
8317 #ifdef __LITTLE_ENDIAN__
vhsub_s32(int32x2_t __p0,int32x2_t __p1)8318 __ai int32x2_t vhsub_s32(int32x2_t __p0, int32x2_t __p1) {
8319 int32x2_t __ret;
8320 __ret = (int32x2_t) __builtin_neon_vhsub_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
8321 return __ret;
8322 }
8323 #else
vhsub_s32(int32x2_t __p0,int32x2_t __p1)8324 __ai int32x2_t vhsub_s32(int32x2_t __p0, int32x2_t __p1) {
8325 int32x2_t __ret;
8326 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
8327 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
8328 __ret = (int32x2_t) __builtin_neon_vhsub_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
8329 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
8330 return __ret;
8331 }
8332 #endif
8333
8334 #ifdef __LITTLE_ENDIAN__
vhsub_s16(int16x4_t __p0,int16x4_t __p1)8335 __ai int16x4_t vhsub_s16(int16x4_t __p0, int16x4_t __p1) {
8336 int16x4_t __ret;
8337 __ret = (int16x4_t) __builtin_neon_vhsub_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
8338 return __ret;
8339 }
8340 #else
vhsub_s16(int16x4_t __p0,int16x4_t __p1)8341 __ai int16x4_t vhsub_s16(int16x4_t __p0, int16x4_t __p1) {
8342 int16x4_t __ret;
8343 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
8344 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
8345 __ret = (int16x4_t) __builtin_neon_vhsub_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
8346 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
8347 return __ret;
8348 }
8349 #endif
8350
8351 #ifdef __LITTLE_ENDIAN__
8352 #define vld1_p8(__p0) __extension__ ({ \
8353 poly8x8_t __ret; \
8354 __ret = (poly8x8_t) __builtin_neon_vld1_v(__p0, 4); \
8355 __ret; \
8356 })
8357 #else
8358 #define vld1_p8(__p0) __extension__ ({ \
8359 poly8x8_t __ret; \
8360 __ret = (poly8x8_t) __builtin_neon_vld1_v(__p0, 4); \
8361 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
8362 __ret; \
8363 })
8364 #endif
8365
8366 #ifdef __LITTLE_ENDIAN__
8367 #define vld1_p16(__p0) __extension__ ({ \
8368 poly16x4_t __ret; \
8369 __ret = (poly16x4_t) __builtin_neon_vld1_v(__p0, 5); \
8370 __ret; \
8371 })
8372 #else
8373 #define vld1_p16(__p0) __extension__ ({ \
8374 poly16x4_t __ret; \
8375 __ret = (poly16x4_t) __builtin_neon_vld1_v(__p0, 5); \
8376 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
8377 __ret; \
8378 })
8379 #endif
8380
8381 #ifdef __LITTLE_ENDIAN__
8382 #define vld1q_p8(__p0) __extension__ ({ \
8383 poly8x16_t __ret; \
8384 __ret = (poly8x16_t) __builtin_neon_vld1q_v(__p0, 36); \
8385 __ret; \
8386 })
8387 #else
8388 #define vld1q_p8(__p0) __extension__ ({ \
8389 poly8x16_t __ret; \
8390 __ret = (poly8x16_t) __builtin_neon_vld1q_v(__p0, 36); \
8391 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
8392 __ret; \
8393 })
8394 #endif
8395
8396 #ifdef __LITTLE_ENDIAN__
8397 #define vld1q_p16(__p0) __extension__ ({ \
8398 poly16x8_t __ret; \
8399 __ret = (poly16x8_t) __builtin_neon_vld1q_v(__p0, 37); \
8400 __ret; \
8401 })
8402 #else
8403 #define vld1q_p16(__p0) __extension__ ({ \
8404 poly16x8_t __ret; \
8405 __ret = (poly16x8_t) __builtin_neon_vld1q_v(__p0, 37); \
8406 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
8407 __ret; \
8408 })
8409 #endif
8410
8411 #ifdef __LITTLE_ENDIAN__
8412 #define vld1q_u8(__p0) __extension__ ({ \
8413 uint8x16_t __ret; \
8414 __ret = (uint8x16_t) __builtin_neon_vld1q_v(__p0, 48); \
8415 __ret; \
8416 })
8417 #else
8418 #define vld1q_u8(__p0) __extension__ ({ \
8419 uint8x16_t __ret; \
8420 __ret = (uint8x16_t) __builtin_neon_vld1q_v(__p0, 48); \
8421 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
8422 __ret; \
8423 })
8424 #endif
8425
8426 #ifdef __LITTLE_ENDIAN__
8427 #define vld1q_u32(__p0) __extension__ ({ \
8428 uint32x4_t __ret; \
8429 __ret = (uint32x4_t) __builtin_neon_vld1q_v(__p0, 50); \
8430 __ret; \
8431 })
8432 #else
8433 #define vld1q_u32(__p0) __extension__ ({ \
8434 uint32x4_t __ret; \
8435 __ret = (uint32x4_t) __builtin_neon_vld1q_v(__p0, 50); \
8436 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
8437 __ret; \
8438 })
8439 #endif
8440
8441 #ifdef __LITTLE_ENDIAN__
8442 #define vld1q_u64(__p0) __extension__ ({ \
8443 uint64x2_t __ret; \
8444 __ret = (uint64x2_t) __builtin_neon_vld1q_v(__p0, 51); \
8445 __ret; \
8446 })
8447 #else
8448 #define vld1q_u64(__p0) __extension__ ({ \
8449 uint64x2_t __ret; \
8450 __ret = (uint64x2_t) __builtin_neon_vld1q_v(__p0, 51); \
8451 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
8452 __ret; \
8453 })
8454 #endif
8455
8456 #ifdef __LITTLE_ENDIAN__
8457 #define vld1q_u16(__p0) __extension__ ({ \
8458 uint16x8_t __ret; \
8459 __ret = (uint16x8_t) __builtin_neon_vld1q_v(__p0, 49); \
8460 __ret; \
8461 })
8462 #else
8463 #define vld1q_u16(__p0) __extension__ ({ \
8464 uint16x8_t __ret; \
8465 __ret = (uint16x8_t) __builtin_neon_vld1q_v(__p0, 49); \
8466 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
8467 __ret; \
8468 })
8469 #endif
8470
8471 #ifdef __LITTLE_ENDIAN__
8472 #define vld1q_s8(__p0) __extension__ ({ \
8473 int8x16_t __ret; \
8474 __ret = (int8x16_t) __builtin_neon_vld1q_v(__p0, 32); \
8475 __ret; \
8476 })
8477 #else
8478 #define vld1q_s8(__p0) __extension__ ({ \
8479 int8x16_t __ret; \
8480 __ret = (int8x16_t) __builtin_neon_vld1q_v(__p0, 32); \
8481 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
8482 __ret; \
8483 })
8484 #endif
8485
8486 #ifdef __LITTLE_ENDIAN__
8487 #define vld1q_f32(__p0) __extension__ ({ \
8488 float32x4_t __ret; \
8489 __ret = (float32x4_t) __builtin_neon_vld1q_v(__p0, 41); \
8490 __ret; \
8491 })
8492 #else
8493 #define vld1q_f32(__p0) __extension__ ({ \
8494 float32x4_t __ret; \
8495 __ret = (float32x4_t) __builtin_neon_vld1q_v(__p0, 41); \
8496 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
8497 __ret; \
8498 })
8499 #endif
8500
8501 #ifdef __LITTLE_ENDIAN__
8502 #define vld1q_s32(__p0) __extension__ ({ \
8503 int32x4_t __ret; \
8504 __ret = (int32x4_t) __builtin_neon_vld1q_v(__p0, 34); \
8505 __ret; \
8506 })
8507 #else
8508 #define vld1q_s32(__p0) __extension__ ({ \
8509 int32x4_t __ret; \
8510 __ret = (int32x4_t) __builtin_neon_vld1q_v(__p0, 34); \
8511 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
8512 __ret; \
8513 })
8514 #endif
8515
8516 #ifdef __LITTLE_ENDIAN__
8517 #define vld1q_s64(__p0) __extension__ ({ \
8518 int64x2_t __ret; \
8519 __ret = (int64x2_t) __builtin_neon_vld1q_v(__p0, 35); \
8520 __ret; \
8521 })
8522 #else
8523 #define vld1q_s64(__p0) __extension__ ({ \
8524 int64x2_t __ret; \
8525 __ret = (int64x2_t) __builtin_neon_vld1q_v(__p0, 35); \
8526 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
8527 __ret; \
8528 })
8529 #endif
8530
8531 #ifdef __LITTLE_ENDIAN__
8532 #define vld1q_s16(__p0) __extension__ ({ \
8533 int16x8_t __ret; \
8534 __ret = (int16x8_t) __builtin_neon_vld1q_v(__p0, 33); \
8535 __ret; \
8536 })
8537 #else
8538 #define vld1q_s16(__p0) __extension__ ({ \
8539 int16x8_t __ret; \
8540 __ret = (int16x8_t) __builtin_neon_vld1q_v(__p0, 33); \
8541 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
8542 __ret; \
8543 })
8544 #endif
8545
8546 #ifdef __LITTLE_ENDIAN__
8547 #define vld1_u8(__p0) __extension__ ({ \
8548 uint8x8_t __ret; \
8549 __ret = (uint8x8_t) __builtin_neon_vld1_v(__p0, 16); \
8550 __ret; \
8551 })
8552 #else
8553 #define vld1_u8(__p0) __extension__ ({ \
8554 uint8x8_t __ret; \
8555 __ret = (uint8x8_t) __builtin_neon_vld1_v(__p0, 16); \
8556 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
8557 __ret; \
8558 })
8559 #endif
8560
8561 #ifdef __LITTLE_ENDIAN__
8562 #define vld1_u32(__p0) __extension__ ({ \
8563 uint32x2_t __ret; \
8564 __ret = (uint32x2_t) __builtin_neon_vld1_v(__p0, 18); \
8565 __ret; \
8566 })
8567 #else
8568 #define vld1_u32(__p0) __extension__ ({ \
8569 uint32x2_t __ret; \
8570 __ret = (uint32x2_t) __builtin_neon_vld1_v(__p0, 18); \
8571 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
8572 __ret; \
8573 })
8574 #endif
8575
8576 #define vld1_u64(__p0) __extension__ ({ \
8577 uint64x1_t __ret; \
8578 __ret = (uint64x1_t) __builtin_neon_vld1_v(__p0, 19); \
8579 __ret; \
8580 })
8581 #ifdef __LITTLE_ENDIAN__
8582 #define vld1_u16(__p0) __extension__ ({ \
8583 uint16x4_t __ret; \
8584 __ret = (uint16x4_t) __builtin_neon_vld1_v(__p0, 17); \
8585 __ret; \
8586 })
8587 #else
8588 #define vld1_u16(__p0) __extension__ ({ \
8589 uint16x4_t __ret; \
8590 __ret = (uint16x4_t) __builtin_neon_vld1_v(__p0, 17); \
8591 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
8592 __ret; \
8593 })
8594 #endif
8595
8596 #ifdef __LITTLE_ENDIAN__
8597 #define vld1_s8(__p0) __extension__ ({ \
8598 int8x8_t __ret; \
8599 __ret = (int8x8_t) __builtin_neon_vld1_v(__p0, 0); \
8600 __ret; \
8601 })
8602 #else
8603 #define vld1_s8(__p0) __extension__ ({ \
8604 int8x8_t __ret; \
8605 __ret = (int8x8_t) __builtin_neon_vld1_v(__p0, 0); \
8606 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
8607 __ret; \
8608 })
8609 #endif
8610
8611 #ifdef __LITTLE_ENDIAN__
8612 #define vld1_f32(__p0) __extension__ ({ \
8613 float32x2_t __ret; \
8614 __ret = (float32x2_t) __builtin_neon_vld1_v(__p0, 9); \
8615 __ret; \
8616 })
8617 #else
8618 #define vld1_f32(__p0) __extension__ ({ \
8619 float32x2_t __ret; \
8620 __ret = (float32x2_t) __builtin_neon_vld1_v(__p0, 9); \
8621 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
8622 __ret; \
8623 })
8624 #endif
8625
8626 #ifdef __LITTLE_ENDIAN__
8627 #define vld1_s32(__p0) __extension__ ({ \
8628 int32x2_t __ret; \
8629 __ret = (int32x2_t) __builtin_neon_vld1_v(__p0, 2); \
8630 __ret; \
8631 })
8632 #else
8633 #define vld1_s32(__p0) __extension__ ({ \
8634 int32x2_t __ret; \
8635 __ret = (int32x2_t) __builtin_neon_vld1_v(__p0, 2); \
8636 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
8637 __ret; \
8638 })
8639 #endif
8640
8641 #define vld1_s64(__p0) __extension__ ({ \
8642 int64x1_t __ret; \
8643 __ret = (int64x1_t) __builtin_neon_vld1_v(__p0, 3); \
8644 __ret; \
8645 })
8646 #ifdef __LITTLE_ENDIAN__
8647 #define vld1_s16(__p0) __extension__ ({ \
8648 int16x4_t __ret; \
8649 __ret = (int16x4_t) __builtin_neon_vld1_v(__p0, 1); \
8650 __ret; \
8651 })
8652 #else
8653 #define vld1_s16(__p0) __extension__ ({ \
8654 int16x4_t __ret; \
8655 __ret = (int16x4_t) __builtin_neon_vld1_v(__p0, 1); \
8656 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
8657 __ret; \
8658 })
8659 #endif
8660
8661 #ifdef __LITTLE_ENDIAN__
8662 #define vld1_dup_p8(__p0) __extension__ ({ \
8663 poly8x8_t __ret; \
8664 __ret = (poly8x8_t) __builtin_neon_vld1_dup_v(__p0, 4); \
8665 __ret; \
8666 })
8667 #else
8668 #define vld1_dup_p8(__p0) __extension__ ({ \
8669 poly8x8_t __ret; \
8670 __ret = (poly8x8_t) __builtin_neon_vld1_dup_v(__p0, 4); \
8671 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
8672 __ret; \
8673 })
8674 #endif
8675
8676 #ifdef __LITTLE_ENDIAN__
8677 #define vld1_dup_p16(__p0) __extension__ ({ \
8678 poly16x4_t __ret; \
8679 __ret = (poly16x4_t) __builtin_neon_vld1_dup_v(__p0, 5); \
8680 __ret; \
8681 })
8682 #else
8683 #define vld1_dup_p16(__p0) __extension__ ({ \
8684 poly16x4_t __ret; \
8685 __ret = (poly16x4_t) __builtin_neon_vld1_dup_v(__p0, 5); \
8686 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
8687 __ret; \
8688 })
8689 #endif
8690
8691 #ifdef __LITTLE_ENDIAN__
8692 #define vld1q_dup_p8(__p0) __extension__ ({ \
8693 poly8x16_t __ret; \
8694 __ret = (poly8x16_t) __builtin_neon_vld1q_dup_v(__p0, 36); \
8695 __ret; \
8696 })
8697 #else
8698 #define vld1q_dup_p8(__p0) __extension__ ({ \
8699 poly8x16_t __ret; \
8700 __ret = (poly8x16_t) __builtin_neon_vld1q_dup_v(__p0, 36); \
8701 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
8702 __ret; \
8703 })
8704 #endif
8705
8706 #ifdef __LITTLE_ENDIAN__
8707 #define vld1q_dup_p16(__p0) __extension__ ({ \
8708 poly16x8_t __ret; \
8709 __ret = (poly16x8_t) __builtin_neon_vld1q_dup_v(__p0, 37); \
8710 __ret; \
8711 })
8712 #else
8713 #define vld1q_dup_p16(__p0) __extension__ ({ \
8714 poly16x8_t __ret; \
8715 __ret = (poly16x8_t) __builtin_neon_vld1q_dup_v(__p0, 37); \
8716 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
8717 __ret; \
8718 })
8719 #endif
8720
8721 #ifdef __LITTLE_ENDIAN__
8722 #define vld1q_dup_u8(__p0) __extension__ ({ \
8723 uint8x16_t __ret; \
8724 __ret = (uint8x16_t) __builtin_neon_vld1q_dup_v(__p0, 48); \
8725 __ret; \
8726 })
8727 #else
8728 #define vld1q_dup_u8(__p0) __extension__ ({ \
8729 uint8x16_t __ret; \
8730 __ret = (uint8x16_t) __builtin_neon_vld1q_dup_v(__p0, 48); \
8731 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
8732 __ret; \
8733 })
8734 #endif
8735
8736 #ifdef __LITTLE_ENDIAN__
8737 #define vld1q_dup_u32(__p0) __extension__ ({ \
8738 uint32x4_t __ret; \
8739 __ret = (uint32x4_t) __builtin_neon_vld1q_dup_v(__p0, 50); \
8740 __ret; \
8741 })
8742 #else
8743 #define vld1q_dup_u32(__p0) __extension__ ({ \
8744 uint32x4_t __ret; \
8745 __ret = (uint32x4_t) __builtin_neon_vld1q_dup_v(__p0, 50); \
8746 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
8747 __ret; \
8748 })
8749 #endif
8750
8751 #ifdef __LITTLE_ENDIAN__
8752 #define vld1q_dup_u64(__p0) __extension__ ({ \
8753 uint64x2_t __ret; \
8754 __ret = (uint64x2_t) __builtin_neon_vld1q_dup_v(__p0, 51); \
8755 __ret; \
8756 })
8757 #else
8758 #define vld1q_dup_u64(__p0) __extension__ ({ \
8759 uint64x2_t __ret; \
8760 __ret = (uint64x2_t) __builtin_neon_vld1q_dup_v(__p0, 51); \
8761 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
8762 __ret; \
8763 })
8764 #endif
8765
8766 #ifdef __LITTLE_ENDIAN__
8767 #define vld1q_dup_u16(__p0) __extension__ ({ \
8768 uint16x8_t __ret; \
8769 __ret = (uint16x8_t) __builtin_neon_vld1q_dup_v(__p0, 49); \
8770 __ret; \
8771 })
8772 #else
8773 #define vld1q_dup_u16(__p0) __extension__ ({ \
8774 uint16x8_t __ret; \
8775 __ret = (uint16x8_t) __builtin_neon_vld1q_dup_v(__p0, 49); \
8776 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
8777 __ret; \
8778 })
8779 #endif
8780
8781 #ifdef __LITTLE_ENDIAN__
8782 #define vld1q_dup_s8(__p0) __extension__ ({ \
8783 int8x16_t __ret; \
8784 __ret = (int8x16_t) __builtin_neon_vld1q_dup_v(__p0, 32); \
8785 __ret; \
8786 })
8787 #else
8788 #define vld1q_dup_s8(__p0) __extension__ ({ \
8789 int8x16_t __ret; \
8790 __ret = (int8x16_t) __builtin_neon_vld1q_dup_v(__p0, 32); \
8791 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
8792 __ret; \
8793 })
8794 #endif
8795
8796 #ifdef __LITTLE_ENDIAN__
8797 #define vld1q_dup_f32(__p0) __extension__ ({ \
8798 float32x4_t __ret; \
8799 __ret = (float32x4_t) __builtin_neon_vld1q_dup_v(__p0, 41); \
8800 __ret; \
8801 })
8802 #else
8803 #define vld1q_dup_f32(__p0) __extension__ ({ \
8804 float32x4_t __ret; \
8805 __ret = (float32x4_t) __builtin_neon_vld1q_dup_v(__p0, 41); \
8806 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
8807 __ret; \
8808 })
8809 #endif
8810
8811 #ifdef __LITTLE_ENDIAN__
8812 #define vld1q_dup_s32(__p0) __extension__ ({ \
8813 int32x4_t __ret; \
8814 __ret = (int32x4_t) __builtin_neon_vld1q_dup_v(__p0, 34); \
8815 __ret; \
8816 })
8817 #else
8818 #define vld1q_dup_s32(__p0) __extension__ ({ \
8819 int32x4_t __ret; \
8820 __ret = (int32x4_t) __builtin_neon_vld1q_dup_v(__p0, 34); \
8821 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
8822 __ret; \
8823 })
8824 #endif
8825
8826 #ifdef __LITTLE_ENDIAN__
8827 #define vld1q_dup_s64(__p0) __extension__ ({ \
8828 int64x2_t __ret; \
8829 __ret = (int64x2_t) __builtin_neon_vld1q_dup_v(__p0, 35); \
8830 __ret; \
8831 })
8832 #else
8833 #define vld1q_dup_s64(__p0) __extension__ ({ \
8834 int64x2_t __ret; \
8835 __ret = (int64x2_t) __builtin_neon_vld1q_dup_v(__p0, 35); \
8836 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
8837 __ret; \
8838 })
8839 #endif
8840
8841 #ifdef __LITTLE_ENDIAN__
8842 #define vld1q_dup_s16(__p0) __extension__ ({ \
8843 int16x8_t __ret; \
8844 __ret = (int16x8_t) __builtin_neon_vld1q_dup_v(__p0, 33); \
8845 __ret; \
8846 })
8847 #else
8848 #define vld1q_dup_s16(__p0) __extension__ ({ \
8849 int16x8_t __ret; \
8850 __ret = (int16x8_t) __builtin_neon_vld1q_dup_v(__p0, 33); \
8851 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
8852 __ret; \
8853 })
8854 #endif
8855
8856 #ifdef __LITTLE_ENDIAN__
8857 #define vld1_dup_u8(__p0) __extension__ ({ \
8858 uint8x8_t __ret; \
8859 __ret = (uint8x8_t) __builtin_neon_vld1_dup_v(__p0, 16); \
8860 __ret; \
8861 })
8862 #else
8863 #define vld1_dup_u8(__p0) __extension__ ({ \
8864 uint8x8_t __ret; \
8865 __ret = (uint8x8_t) __builtin_neon_vld1_dup_v(__p0, 16); \
8866 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
8867 __ret; \
8868 })
8869 #endif
8870
8871 #ifdef __LITTLE_ENDIAN__
8872 #define vld1_dup_u32(__p0) __extension__ ({ \
8873 uint32x2_t __ret; \
8874 __ret = (uint32x2_t) __builtin_neon_vld1_dup_v(__p0, 18); \
8875 __ret; \
8876 })
8877 #else
8878 #define vld1_dup_u32(__p0) __extension__ ({ \
8879 uint32x2_t __ret; \
8880 __ret = (uint32x2_t) __builtin_neon_vld1_dup_v(__p0, 18); \
8881 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
8882 __ret; \
8883 })
8884 #endif
8885
8886 #define vld1_dup_u64(__p0) __extension__ ({ \
8887 uint64x1_t __ret; \
8888 __ret = (uint64x1_t) __builtin_neon_vld1_dup_v(__p0, 19); \
8889 __ret; \
8890 })
8891 #ifdef __LITTLE_ENDIAN__
8892 #define vld1_dup_u16(__p0) __extension__ ({ \
8893 uint16x4_t __ret; \
8894 __ret = (uint16x4_t) __builtin_neon_vld1_dup_v(__p0, 17); \
8895 __ret; \
8896 })
8897 #else
8898 #define vld1_dup_u16(__p0) __extension__ ({ \
8899 uint16x4_t __ret; \
8900 __ret = (uint16x4_t) __builtin_neon_vld1_dup_v(__p0, 17); \
8901 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
8902 __ret; \
8903 })
8904 #endif
8905
8906 #ifdef __LITTLE_ENDIAN__
8907 #define vld1_dup_s8(__p0) __extension__ ({ \
8908 int8x8_t __ret; \
8909 __ret = (int8x8_t) __builtin_neon_vld1_dup_v(__p0, 0); \
8910 __ret; \
8911 })
8912 #else
8913 #define vld1_dup_s8(__p0) __extension__ ({ \
8914 int8x8_t __ret; \
8915 __ret = (int8x8_t) __builtin_neon_vld1_dup_v(__p0, 0); \
8916 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
8917 __ret; \
8918 })
8919 #endif
8920
8921 #ifdef __LITTLE_ENDIAN__
8922 #define vld1_dup_f32(__p0) __extension__ ({ \
8923 float32x2_t __ret; \
8924 __ret = (float32x2_t) __builtin_neon_vld1_dup_v(__p0, 9); \
8925 __ret; \
8926 })
8927 #else
8928 #define vld1_dup_f32(__p0) __extension__ ({ \
8929 float32x2_t __ret; \
8930 __ret = (float32x2_t) __builtin_neon_vld1_dup_v(__p0, 9); \
8931 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
8932 __ret; \
8933 })
8934 #endif
8935
8936 #ifdef __LITTLE_ENDIAN__
8937 #define vld1_dup_s32(__p0) __extension__ ({ \
8938 int32x2_t __ret; \
8939 __ret = (int32x2_t) __builtin_neon_vld1_dup_v(__p0, 2); \
8940 __ret; \
8941 })
8942 #else
8943 #define vld1_dup_s32(__p0) __extension__ ({ \
8944 int32x2_t __ret; \
8945 __ret = (int32x2_t) __builtin_neon_vld1_dup_v(__p0, 2); \
8946 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
8947 __ret; \
8948 })
8949 #endif
8950
8951 #define vld1_dup_s64(__p0) __extension__ ({ \
8952 int64x1_t __ret; \
8953 __ret = (int64x1_t) __builtin_neon_vld1_dup_v(__p0, 3); \
8954 __ret; \
8955 })
8956 #ifdef __LITTLE_ENDIAN__
8957 #define vld1_dup_s16(__p0) __extension__ ({ \
8958 int16x4_t __ret; \
8959 __ret = (int16x4_t) __builtin_neon_vld1_dup_v(__p0, 1); \
8960 __ret; \
8961 })
8962 #else
8963 #define vld1_dup_s16(__p0) __extension__ ({ \
8964 int16x4_t __ret; \
8965 __ret = (int16x4_t) __builtin_neon_vld1_dup_v(__p0, 1); \
8966 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
8967 __ret; \
8968 })
8969 #endif
8970
8971 #ifdef __LITTLE_ENDIAN__
8972 #define vld1_lane_p8(__p0, __p1, __p2) __extension__ ({ \
8973 poly8x8_t __ret; \
8974 poly8x8_t __s1 = __p1; \
8975 __ret = (poly8x8_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__s1, __p2, 4); \
8976 __ret; \
8977 })
8978 #else
8979 #define vld1_lane_p8(__p0, __p1, __p2) __extension__ ({ \
8980 poly8x8_t __ret; \
8981 poly8x8_t __s1 = __p1; \
8982 poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
8983 __ret = (poly8x8_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__rev1, __p2, 4); \
8984 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
8985 __ret; \
8986 })
8987 #endif
8988
8989 #ifdef __LITTLE_ENDIAN__
8990 #define vld1_lane_p16(__p0, __p1, __p2) __extension__ ({ \
8991 poly16x4_t __ret; \
8992 poly16x4_t __s1 = __p1; \
8993 __ret = (poly16x4_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__s1, __p2, 5); \
8994 __ret; \
8995 })
8996 #else
8997 #define vld1_lane_p16(__p0, __p1, __p2) __extension__ ({ \
8998 poly16x4_t __ret; \
8999 poly16x4_t __s1 = __p1; \
9000 poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
9001 __ret = (poly16x4_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__rev1, __p2, 5); \
9002 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
9003 __ret; \
9004 })
9005 #endif
9006
9007 #ifdef __LITTLE_ENDIAN__
9008 #define vld1q_lane_p8(__p0, __p1, __p2) __extension__ ({ \
9009 poly8x16_t __ret; \
9010 poly8x16_t __s1 = __p1; \
9011 __ret = (poly8x16_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__s1, __p2, 36); \
9012 __ret; \
9013 })
9014 #else
9015 #define vld1q_lane_p8(__p0, __p1, __p2) __extension__ ({ \
9016 poly8x16_t __ret; \
9017 poly8x16_t __s1 = __p1; \
9018 poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
9019 __ret = (poly8x16_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 36); \
9020 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
9021 __ret; \
9022 })
9023 #endif
9024
9025 #ifdef __LITTLE_ENDIAN__
9026 #define vld1q_lane_p16(__p0, __p1, __p2) __extension__ ({ \
9027 poly16x8_t __ret; \
9028 poly16x8_t __s1 = __p1; \
9029 __ret = (poly16x8_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__s1, __p2, 37); \
9030 __ret; \
9031 })
9032 #else
9033 #define vld1q_lane_p16(__p0, __p1, __p2) __extension__ ({ \
9034 poly16x8_t __ret; \
9035 poly16x8_t __s1 = __p1; \
9036 poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
9037 __ret = (poly16x8_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 37); \
9038 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
9039 __ret; \
9040 })
9041 #endif
9042
9043 #ifdef __LITTLE_ENDIAN__
9044 #define vld1q_lane_u8(__p0, __p1, __p2) __extension__ ({ \
9045 uint8x16_t __ret; \
9046 uint8x16_t __s1 = __p1; \
9047 __ret = (uint8x16_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__s1, __p2, 48); \
9048 __ret; \
9049 })
9050 #else
9051 #define vld1q_lane_u8(__p0, __p1, __p2) __extension__ ({ \
9052 uint8x16_t __ret; \
9053 uint8x16_t __s1 = __p1; \
9054 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
9055 __ret = (uint8x16_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 48); \
9056 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
9057 __ret; \
9058 })
9059 #endif
9060
9061 #ifdef __LITTLE_ENDIAN__
9062 #define vld1q_lane_u32(__p0, __p1, __p2) __extension__ ({ \
9063 uint32x4_t __ret; \
9064 uint32x4_t __s1 = __p1; \
9065 __ret = (uint32x4_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__s1, __p2, 50); \
9066 __ret; \
9067 })
9068 #else
9069 #define vld1q_lane_u32(__p0, __p1, __p2) __extension__ ({ \
9070 uint32x4_t __ret; \
9071 uint32x4_t __s1 = __p1; \
9072 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
9073 __ret = (uint32x4_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 50); \
9074 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
9075 __ret; \
9076 })
9077 #endif
9078
9079 #ifdef __LITTLE_ENDIAN__
9080 #define vld1q_lane_u64(__p0, __p1, __p2) __extension__ ({ \
9081 uint64x2_t __ret; \
9082 uint64x2_t __s1 = __p1; \
9083 __ret = (uint64x2_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__s1, __p2, 51); \
9084 __ret; \
9085 })
9086 #else
9087 #define vld1q_lane_u64(__p0, __p1, __p2) __extension__ ({ \
9088 uint64x2_t __ret; \
9089 uint64x2_t __s1 = __p1; \
9090 uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
9091 __ret = (uint64x2_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 51); \
9092 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
9093 __ret; \
9094 })
9095 #endif
9096
9097 #ifdef __LITTLE_ENDIAN__
9098 #define vld1q_lane_u16(__p0, __p1, __p2) __extension__ ({ \
9099 uint16x8_t __ret; \
9100 uint16x8_t __s1 = __p1; \
9101 __ret = (uint16x8_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__s1, __p2, 49); \
9102 __ret; \
9103 })
9104 #else
9105 #define vld1q_lane_u16(__p0, __p1, __p2) __extension__ ({ \
9106 uint16x8_t __ret; \
9107 uint16x8_t __s1 = __p1; \
9108 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
9109 __ret = (uint16x8_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 49); \
9110 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
9111 __ret; \
9112 })
9113 #endif
9114
9115 #ifdef __LITTLE_ENDIAN__
9116 #define vld1q_lane_s8(__p0, __p1, __p2) __extension__ ({ \
9117 int8x16_t __ret; \
9118 int8x16_t __s1 = __p1; \
9119 __ret = (int8x16_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__s1, __p2, 32); \
9120 __ret; \
9121 })
9122 #else
9123 #define vld1q_lane_s8(__p0, __p1, __p2) __extension__ ({ \
9124 int8x16_t __ret; \
9125 int8x16_t __s1 = __p1; \
9126 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
9127 __ret = (int8x16_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 32); \
9128 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
9129 __ret; \
9130 })
9131 #endif
9132
9133 #ifdef __LITTLE_ENDIAN__
9134 #define vld1q_lane_f32(__p0, __p1, __p2) __extension__ ({ \
9135 float32x4_t __ret; \
9136 float32x4_t __s1 = __p1; \
9137 __ret = (float32x4_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__s1, __p2, 41); \
9138 __ret; \
9139 })
9140 #else
9141 #define vld1q_lane_f32(__p0, __p1, __p2) __extension__ ({ \
9142 float32x4_t __ret; \
9143 float32x4_t __s1 = __p1; \
9144 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
9145 __ret = (float32x4_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 41); \
9146 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
9147 __ret; \
9148 })
9149 #endif
9150
9151 #ifdef __LITTLE_ENDIAN__
9152 #define vld1q_lane_s32(__p0, __p1, __p2) __extension__ ({ \
9153 int32x4_t __ret; \
9154 int32x4_t __s1 = __p1; \
9155 __ret = (int32x4_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__s1, __p2, 34); \
9156 __ret; \
9157 })
9158 #else
9159 #define vld1q_lane_s32(__p0, __p1, __p2) __extension__ ({ \
9160 int32x4_t __ret; \
9161 int32x4_t __s1 = __p1; \
9162 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
9163 __ret = (int32x4_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 34); \
9164 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
9165 __ret; \
9166 })
9167 #endif
9168
9169 #ifdef __LITTLE_ENDIAN__
9170 #define vld1q_lane_s64(__p0, __p1, __p2) __extension__ ({ \
9171 int64x2_t __ret; \
9172 int64x2_t __s1 = __p1; \
9173 __ret = (int64x2_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__s1, __p2, 35); \
9174 __ret; \
9175 })
9176 #else
9177 #define vld1q_lane_s64(__p0, __p1, __p2) __extension__ ({ \
9178 int64x2_t __ret; \
9179 int64x2_t __s1 = __p1; \
9180 int64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
9181 __ret = (int64x2_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 35); \
9182 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
9183 __ret; \
9184 })
9185 #endif
9186
9187 #ifdef __LITTLE_ENDIAN__
9188 #define vld1q_lane_s16(__p0, __p1, __p2) __extension__ ({ \
9189 int16x8_t __ret; \
9190 int16x8_t __s1 = __p1; \
9191 __ret = (int16x8_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__s1, __p2, 33); \
9192 __ret; \
9193 })
9194 #else
9195 #define vld1q_lane_s16(__p0, __p1, __p2) __extension__ ({ \
9196 int16x8_t __ret; \
9197 int16x8_t __s1 = __p1; \
9198 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
9199 __ret = (int16x8_t) __builtin_neon_vld1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 33); \
9200 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
9201 __ret; \
9202 })
9203 #endif
9204
9205 #ifdef __LITTLE_ENDIAN__
9206 #define vld1_lane_u8(__p0, __p1, __p2) __extension__ ({ \
9207 uint8x8_t __ret; \
9208 uint8x8_t __s1 = __p1; \
9209 __ret = (uint8x8_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__s1, __p2, 16); \
9210 __ret; \
9211 })
9212 #else
9213 #define vld1_lane_u8(__p0, __p1, __p2) __extension__ ({ \
9214 uint8x8_t __ret; \
9215 uint8x8_t __s1 = __p1; \
9216 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
9217 __ret = (uint8x8_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__rev1, __p2, 16); \
9218 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
9219 __ret; \
9220 })
9221 #endif
9222
9223 #ifdef __LITTLE_ENDIAN__
9224 #define vld1_lane_u32(__p0, __p1, __p2) __extension__ ({ \
9225 uint32x2_t __ret; \
9226 uint32x2_t __s1 = __p1; \
9227 __ret = (uint32x2_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__s1, __p2, 18); \
9228 __ret; \
9229 })
9230 #else
9231 #define vld1_lane_u32(__p0, __p1, __p2) __extension__ ({ \
9232 uint32x2_t __ret; \
9233 uint32x2_t __s1 = __p1; \
9234 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
9235 __ret = (uint32x2_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__rev1, __p2, 18); \
9236 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
9237 __ret; \
9238 })
9239 #endif
9240
9241 #define vld1_lane_u64(__p0, __p1, __p2) __extension__ ({ \
9242 uint64x1_t __ret; \
9243 uint64x1_t __s1 = __p1; \
9244 __ret = (uint64x1_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__s1, __p2, 19); \
9245 __ret; \
9246 })
9247 #ifdef __LITTLE_ENDIAN__
9248 #define vld1_lane_u16(__p0, __p1, __p2) __extension__ ({ \
9249 uint16x4_t __ret; \
9250 uint16x4_t __s1 = __p1; \
9251 __ret = (uint16x4_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__s1, __p2, 17); \
9252 __ret; \
9253 })
9254 #else
9255 #define vld1_lane_u16(__p0, __p1, __p2) __extension__ ({ \
9256 uint16x4_t __ret; \
9257 uint16x4_t __s1 = __p1; \
9258 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
9259 __ret = (uint16x4_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__rev1, __p2, 17); \
9260 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
9261 __ret; \
9262 })
9263 #endif
9264
9265 #ifdef __LITTLE_ENDIAN__
9266 #define vld1_lane_s8(__p0, __p1, __p2) __extension__ ({ \
9267 int8x8_t __ret; \
9268 int8x8_t __s1 = __p1; \
9269 __ret = (int8x8_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__s1, __p2, 0); \
9270 __ret; \
9271 })
9272 #else
9273 #define vld1_lane_s8(__p0, __p1, __p2) __extension__ ({ \
9274 int8x8_t __ret; \
9275 int8x8_t __s1 = __p1; \
9276 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
9277 __ret = (int8x8_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__rev1, __p2, 0); \
9278 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
9279 __ret; \
9280 })
9281 #endif
9282
9283 #ifdef __LITTLE_ENDIAN__
9284 #define vld1_lane_f32(__p0, __p1, __p2) __extension__ ({ \
9285 float32x2_t __ret; \
9286 float32x2_t __s1 = __p1; \
9287 __ret = (float32x2_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__s1, __p2, 9); \
9288 __ret; \
9289 })
9290 #else
9291 #define vld1_lane_f32(__p0, __p1, __p2) __extension__ ({ \
9292 float32x2_t __ret; \
9293 float32x2_t __s1 = __p1; \
9294 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
9295 __ret = (float32x2_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__rev1, __p2, 9); \
9296 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
9297 __ret; \
9298 })
9299 #endif
9300
9301 #ifdef __LITTLE_ENDIAN__
9302 #define vld1_lane_s32(__p0, __p1, __p2) __extension__ ({ \
9303 int32x2_t __ret; \
9304 int32x2_t __s1 = __p1; \
9305 __ret = (int32x2_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__s1, __p2, 2); \
9306 __ret; \
9307 })
9308 #else
9309 #define vld1_lane_s32(__p0, __p1, __p2) __extension__ ({ \
9310 int32x2_t __ret; \
9311 int32x2_t __s1 = __p1; \
9312 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
9313 __ret = (int32x2_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__rev1, __p2, 2); \
9314 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
9315 __ret; \
9316 })
9317 #endif
9318
9319 #define vld1_lane_s64(__p0, __p1, __p2) __extension__ ({ \
9320 int64x1_t __ret; \
9321 int64x1_t __s1 = __p1; \
9322 __ret = (int64x1_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__s1, __p2, 3); \
9323 __ret; \
9324 })
9325 #ifdef __LITTLE_ENDIAN__
9326 #define vld1_lane_s16(__p0, __p1, __p2) __extension__ ({ \
9327 int16x4_t __ret; \
9328 int16x4_t __s1 = __p1; \
9329 __ret = (int16x4_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__s1, __p2, 1); \
9330 __ret; \
9331 })
9332 #else
9333 #define vld1_lane_s16(__p0, __p1, __p2) __extension__ ({ \
9334 int16x4_t __ret; \
9335 int16x4_t __s1 = __p1; \
9336 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
9337 __ret = (int16x4_t) __builtin_neon_vld1_lane_v(__p0, (int8x8_t)__rev1, __p2, 1); \
9338 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
9339 __ret; \
9340 })
9341 #endif
9342
9343 #ifdef __LITTLE_ENDIAN__
9344 #define vld1_p8_x2(__p0) __extension__ ({ \
9345 poly8x8x2_t __ret; \
9346 __builtin_neon_vld1_x2_v(&__ret, __p0, 4); \
9347 __ret; \
9348 })
9349 #else
9350 #define vld1_p8_x2(__p0) __extension__ ({ \
9351 poly8x8x2_t __ret; \
9352 __builtin_neon_vld1_x2_v(&__ret, __p0, 4); \
9353 \
9354 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
9355 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
9356 __ret; \
9357 })
9358 #endif
9359
9360 #ifdef __LITTLE_ENDIAN__
9361 #define vld1_p16_x2(__p0) __extension__ ({ \
9362 poly16x4x2_t __ret; \
9363 __builtin_neon_vld1_x2_v(&__ret, __p0, 5); \
9364 __ret; \
9365 })
9366 #else
9367 #define vld1_p16_x2(__p0) __extension__ ({ \
9368 poly16x4x2_t __ret; \
9369 __builtin_neon_vld1_x2_v(&__ret, __p0, 5); \
9370 \
9371 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
9372 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
9373 __ret; \
9374 })
9375 #endif
9376
9377 #ifdef __LITTLE_ENDIAN__
9378 #define vld1q_p8_x2(__p0) __extension__ ({ \
9379 poly8x16x2_t __ret; \
9380 __builtin_neon_vld1q_x2_v(&__ret, __p0, 36); \
9381 __ret; \
9382 })
9383 #else
9384 #define vld1q_p8_x2(__p0) __extension__ ({ \
9385 poly8x16x2_t __ret; \
9386 __builtin_neon_vld1q_x2_v(&__ret, __p0, 36); \
9387 \
9388 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
9389 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
9390 __ret; \
9391 })
9392 #endif
9393
9394 #ifdef __LITTLE_ENDIAN__
9395 #define vld1q_p16_x2(__p0) __extension__ ({ \
9396 poly16x8x2_t __ret; \
9397 __builtin_neon_vld1q_x2_v(&__ret, __p0, 37); \
9398 __ret; \
9399 })
9400 #else
9401 #define vld1q_p16_x2(__p0) __extension__ ({ \
9402 poly16x8x2_t __ret; \
9403 __builtin_neon_vld1q_x2_v(&__ret, __p0, 37); \
9404 \
9405 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
9406 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
9407 __ret; \
9408 })
9409 #endif
9410
9411 #ifdef __LITTLE_ENDIAN__
9412 #define vld1q_u8_x2(__p0) __extension__ ({ \
9413 uint8x16x2_t __ret; \
9414 __builtin_neon_vld1q_x2_v(&__ret, __p0, 48); \
9415 __ret; \
9416 })
9417 #else
9418 #define vld1q_u8_x2(__p0) __extension__ ({ \
9419 uint8x16x2_t __ret; \
9420 __builtin_neon_vld1q_x2_v(&__ret, __p0, 48); \
9421 \
9422 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
9423 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
9424 __ret; \
9425 })
9426 #endif
9427
9428 #ifdef __LITTLE_ENDIAN__
9429 #define vld1q_u32_x2(__p0) __extension__ ({ \
9430 uint32x4x2_t __ret; \
9431 __builtin_neon_vld1q_x2_v(&__ret, __p0, 50); \
9432 __ret; \
9433 })
9434 #else
9435 #define vld1q_u32_x2(__p0) __extension__ ({ \
9436 uint32x4x2_t __ret; \
9437 __builtin_neon_vld1q_x2_v(&__ret, __p0, 50); \
9438 \
9439 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
9440 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
9441 __ret; \
9442 })
9443 #endif
9444
9445 #ifdef __LITTLE_ENDIAN__
9446 #define vld1q_u64_x2(__p0) __extension__ ({ \
9447 uint64x2x2_t __ret; \
9448 __builtin_neon_vld1q_x2_v(&__ret, __p0, 51); \
9449 __ret; \
9450 })
9451 #else
9452 #define vld1q_u64_x2(__p0) __extension__ ({ \
9453 uint64x2x2_t __ret; \
9454 __builtin_neon_vld1q_x2_v(&__ret, __p0, 51); \
9455 \
9456 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
9457 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
9458 __ret; \
9459 })
9460 #endif
9461
9462 #ifdef __LITTLE_ENDIAN__
9463 #define vld1q_u16_x2(__p0) __extension__ ({ \
9464 uint16x8x2_t __ret; \
9465 __builtin_neon_vld1q_x2_v(&__ret, __p0, 49); \
9466 __ret; \
9467 })
9468 #else
9469 #define vld1q_u16_x2(__p0) __extension__ ({ \
9470 uint16x8x2_t __ret; \
9471 __builtin_neon_vld1q_x2_v(&__ret, __p0, 49); \
9472 \
9473 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
9474 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
9475 __ret; \
9476 })
9477 #endif
9478
9479 #ifdef __LITTLE_ENDIAN__
9480 #define vld1q_s8_x2(__p0) __extension__ ({ \
9481 int8x16x2_t __ret; \
9482 __builtin_neon_vld1q_x2_v(&__ret, __p0, 32); \
9483 __ret; \
9484 })
9485 #else
9486 #define vld1q_s8_x2(__p0) __extension__ ({ \
9487 int8x16x2_t __ret; \
9488 __builtin_neon_vld1q_x2_v(&__ret, __p0, 32); \
9489 \
9490 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
9491 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
9492 __ret; \
9493 })
9494 #endif
9495
9496 #ifdef __LITTLE_ENDIAN__
9497 #define vld1q_f32_x2(__p0) __extension__ ({ \
9498 float32x4x2_t __ret; \
9499 __builtin_neon_vld1q_x2_v(&__ret, __p0, 41); \
9500 __ret; \
9501 })
9502 #else
9503 #define vld1q_f32_x2(__p0) __extension__ ({ \
9504 float32x4x2_t __ret; \
9505 __builtin_neon_vld1q_x2_v(&__ret, __p0, 41); \
9506 \
9507 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
9508 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
9509 __ret; \
9510 })
9511 #endif
9512
9513 #ifdef __LITTLE_ENDIAN__
9514 #define vld1q_s32_x2(__p0) __extension__ ({ \
9515 int32x4x2_t __ret; \
9516 __builtin_neon_vld1q_x2_v(&__ret, __p0, 34); \
9517 __ret; \
9518 })
9519 #else
9520 #define vld1q_s32_x2(__p0) __extension__ ({ \
9521 int32x4x2_t __ret; \
9522 __builtin_neon_vld1q_x2_v(&__ret, __p0, 34); \
9523 \
9524 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
9525 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
9526 __ret; \
9527 })
9528 #endif
9529
9530 #ifdef __LITTLE_ENDIAN__
9531 #define vld1q_s64_x2(__p0) __extension__ ({ \
9532 int64x2x2_t __ret; \
9533 __builtin_neon_vld1q_x2_v(&__ret, __p0, 35); \
9534 __ret; \
9535 })
9536 #else
9537 #define vld1q_s64_x2(__p0) __extension__ ({ \
9538 int64x2x2_t __ret; \
9539 __builtin_neon_vld1q_x2_v(&__ret, __p0, 35); \
9540 \
9541 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
9542 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
9543 __ret; \
9544 })
9545 #endif
9546
9547 #ifdef __LITTLE_ENDIAN__
9548 #define vld1q_s16_x2(__p0) __extension__ ({ \
9549 int16x8x2_t __ret; \
9550 __builtin_neon_vld1q_x2_v(&__ret, __p0, 33); \
9551 __ret; \
9552 })
9553 #else
9554 #define vld1q_s16_x2(__p0) __extension__ ({ \
9555 int16x8x2_t __ret; \
9556 __builtin_neon_vld1q_x2_v(&__ret, __p0, 33); \
9557 \
9558 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
9559 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
9560 __ret; \
9561 })
9562 #endif
9563
9564 #ifdef __LITTLE_ENDIAN__
9565 #define vld1_u8_x2(__p0) __extension__ ({ \
9566 uint8x8x2_t __ret; \
9567 __builtin_neon_vld1_x2_v(&__ret, __p0, 16); \
9568 __ret; \
9569 })
9570 #else
9571 #define vld1_u8_x2(__p0) __extension__ ({ \
9572 uint8x8x2_t __ret; \
9573 __builtin_neon_vld1_x2_v(&__ret, __p0, 16); \
9574 \
9575 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
9576 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
9577 __ret; \
9578 })
9579 #endif
9580
9581 #ifdef __LITTLE_ENDIAN__
9582 #define vld1_u32_x2(__p0) __extension__ ({ \
9583 uint32x2x2_t __ret; \
9584 __builtin_neon_vld1_x2_v(&__ret, __p0, 18); \
9585 __ret; \
9586 })
9587 #else
9588 #define vld1_u32_x2(__p0) __extension__ ({ \
9589 uint32x2x2_t __ret; \
9590 __builtin_neon_vld1_x2_v(&__ret, __p0, 18); \
9591 \
9592 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
9593 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
9594 __ret; \
9595 })
9596 #endif
9597
9598 #define vld1_u64_x2(__p0) __extension__ ({ \
9599 uint64x1x2_t __ret; \
9600 __builtin_neon_vld1_x2_v(&__ret, __p0, 19); \
9601 __ret; \
9602 })
9603 #ifdef __LITTLE_ENDIAN__
9604 #define vld1_u16_x2(__p0) __extension__ ({ \
9605 uint16x4x2_t __ret; \
9606 __builtin_neon_vld1_x2_v(&__ret, __p0, 17); \
9607 __ret; \
9608 })
9609 #else
9610 #define vld1_u16_x2(__p0) __extension__ ({ \
9611 uint16x4x2_t __ret; \
9612 __builtin_neon_vld1_x2_v(&__ret, __p0, 17); \
9613 \
9614 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
9615 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
9616 __ret; \
9617 })
9618 #endif
9619
9620 #ifdef __LITTLE_ENDIAN__
9621 #define vld1_s8_x2(__p0) __extension__ ({ \
9622 int8x8x2_t __ret; \
9623 __builtin_neon_vld1_x2_v(&__ret, __p0, 0); \
9624 __ret; \
9625 })
9626 #else
9627 #define vld1_s8_x2(__p0) __extension__ ({ \
9628 int8x8x2_t __ret; \
9629 __builtin_neon_vld1_x2_v(&__ret, __p0, 0); \
9630 \
9631 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
9632 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
9633 __ret; \
9634 })
9635 #endif
9636
9637 #ifdef __LITTLE_ENDIAN__
9638 #define vld1_f32_x2(__p0) __extension__ ({ \
9639 float32x2x2_t __ret; \
9640 __builtin_neon_vld1_x2_v(&__ret, __p0, 9); \
9641 __ret; \
9642 })
9643 #else
9644 #define vld1_f32_x2(__p0) __extension__ ({ \
9645 float32x2x2_t __ret; \
9646 __builtin_neon_vld1_x2_v(&__ret, __p0, 9); \
9647 \
9648 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
9649 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
9650 __ret; \
9651 })
9652 #endif
9653
9654 #ifdef __LITTLE_ENDIAN__
9655 #define vld1_s32_x2(__p0) __extension__ ({ \
9656 int32x2x2_t __ret; \
9657 __builtin_neon_vld1_x2_v(&__ret, __p0, 2); \
9658 __ret; \
9659 })
9660 #else
9661 #define vld1_s32_x2(__p0) __extension__ ({ \
9662 int32x2x2_t __ret; \
9663 __builtin_neon_vld1_x2_v(&__ret, __p0, 2); \
9664 \
9665 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
9666 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
9667 __ret; \
9668 })
9669 #endif
9670
9671 #define vld1_s64_x2(__p0) __extension__ ({ \
9672 int64x1x2_t __ret; \
9673 __builtin_neon_vld1_x2_v(&__ret, __p0, 3); \
9674 __ret; \
9675 })
9676 #ifdef __LITTLE_ENDIAN__
9677 #define vld1_s16_x2(__p0) __extension__ ({ \
9678 int16x4x2_t __ret; \
9679 __builtin_neon_vld1_x2_v(&__ret, __p0, 1); \
9680 __ret; \
9681 })
9682 #else
9683 #define vld1_s16_x2(__p0) __extension__ ({ \
9684 int16x4x2_t __ret; \
9685 __builtin_neon_vld1_x2_v(&__ret, __p0, 1); \
9686 \
9687 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
9688 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
9689 __ret; \
9690 })
9691 #endif
9692
9693 #ifdef __LITTLE_ENDIAN__
9694 #define vld1_p8_x3(__p0) __extension__ ({ \
9695 poly8x8x3_t __ret; \
9696 __builtin_neon_vld1_x3_v(&__ret, __p0, 4); \
9697 __ret; \
9698 })
9699 #else
9700 #define vld1_p8_x3(__p0) __extension__ ({ \
9701 poly8x8x3_t __ret; \
9702 __builtin_neon_vld1_x3_v(&__ret, __p0, 4); \
9703 \
9704 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
9705 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
9706 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
9707 __ret; \
9708 })
9709 #endif
9710
9711 #ifdef __LITTLE_ENDIAN__
9712 #define vld1_p16_x3(__p0) __extension__ ({ \
9713 poly16x4x3_t __ret; \
9714 __builtin_neon_vld1_x3_v(&__ret, __p0, 5); \
9715 __ret; \
9716 })
9717 #else
9718 #define vld1_p16_x3(__p0) __extension__ ({ \
9719 poly16x4x3_t __ret; \
9720 __builtin_neon_vld1_x3_v(&__ret, __p0, 5); \
9721 \
9722 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
9723 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
9724 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
9725 __ret; \
9726 })
9727 #endif
9728
9729 #ifdef __LITTLE_ENDIAN__
9730 #define vld1q_p8_x3(__p0) __extension__ ({ \
9731 poly8x16x3_t __ret; \
9732 __builtin_neon_vld1q_x3_v(&__ret, __p0, 36); \
9733 __ret; \
9734 })
9735 #else
9736 #define vld1q_p8_x3(__p0) __extension__ ({ \
9737 poly8x16x3_t __ret; \
9738 __builtin_neon_vld1q_x3_v(&__ret, __p0, 36); \
9739 \
9740 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
9741 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
9742 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
9743 __ret; \
9744 })
9745 #endif
9746
9747 #ifdef __LITTLE_ENDIAN__
9748 #define vld1q_p16_x3(__p0) __extension__ ({ \
9749 poly16x8x3_t __ret; \
9750 __builtin_neon_vld1q_x3_v(&__ret, __p0, 37); \
9751 __ret; \
9752 })
9753 #else
9754 #define vld1q_p16_x3(__p0) __extension__ ({ \
9755 poly16x8x3_t __ret; \
9756 __builtin_neon_vld1q_x3_v(&__ret, __p0, 37); \
9757 \
9758 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
9759 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
9760 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
9761 __ret; \
9762 })
9763 #endif
9764
9765 #ifdef __LITTLE_ENDIAN__
9766 #define vld1q_u8_x3(__p0) __extension__ ({ \
9767 uint8x16x3_t __ret; \
9768 __builtin_neon_vld1q_x3_v(&__ret, __p0, 48); \
9769 __ret; \
9770 })
9771 #else
9772 #define vld1q_u8_x3(__p0) __extension__ ({ \
9773 uint8x16x3_t __ret; \
9774 __builtin_neon_vld1q_x3_v(&__ret, __p0, 48); \
9775 \
9776 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
9777 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
9778 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
9779 __ret; \
9780 })
9781 #endif
9782
9783 #ifdef __LITTLE_ENDIAN__
9784 #define vld1q_u32_x3(__p0) __extension__ ({ \
9785 uint32x4x3_t __ret; \
9786 __builtin_neon_vld1q_x3_v(&__ret, __p0, 50); \
9787 __ret; \
9788 })
9789 #else
9790 #define vld1q_u32_x3(__p0) __extension__ ({ \
9791 uint32x4x3_t __ret; \
9792 __builtin_neon_vld1q_x3_v(&__ret, __p0, 50); \
9793 \
9794 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
9795 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
9796 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
9797 __ret; \
9798 })
9799 #endif
9800
9801 #ifdef __LITTLE_ENDIAN__
9802 #define vld1q_u64_x3(__p0) __extension__ ({ \
9803 uint64x2x3_t __ret; \
9804 __builtin_neon_vld1q_x3_v(&__ret, __p0, 51); \
9805 __ret; \
9806 })
9807 #else
9808 #define vld1q_u64_x3(__p0) __extension__ ({ \
9809 uint64x2x3_t __ret; \
9810 __builtin_neon_vld1q_x3_v(&__ret, __p0, 51); \
9811 \
9812 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
9813 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
9814 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
9815 __ret; \
9816 })
9817 #endif
9818
9819 #ifdef __LITTLE_ENDIAN__
9820 #define vld1q_u16_x3(__p0) __extension__ ({ \
9821 uint16x8x3_t __ret; \
9822 __builtin_neon_vld1q_x3_v(&__ret, __p0, 49); \
9823 __ret; \
9824 })
9825 #else
9826 #define vld1q_u16_x3(__p0) __extension__ ({ \
9827 uint16x8x3_t __ret; \
9828 __builtin_neon_vld1q_x3_v(&__ret, __p0, 49); \
9829 \
9830 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
9831 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
9832 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
9833 __ret; \
9834 })
9835 #endif
9836
9837 #ifdef __LITTLE_ENDIAN__
9838 #define vld1q_s8_x3(__p0) __extension__ ({ \
9839 int8x16x3_t __ret; \
9840 __builtin_neon_vld1q_x3_v(&__ret, __p0, 32); \
9841 __ret; \
9842 })
9843 #else
9844 #define vld1q_s8_x3(__p0) __extension__ ({ \
9845 int8x16x3_t __ret; \
9846 __builtin_neon_vld1q_x3_v(&__ret, __p0, 32); \
9847 \
9848 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
9849 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
9850 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
9851 __ret; \
9852 })
9853 #endif
9854
9855 #ifdef __LITTLE_ENDIAN__
9856 #define vld1q_f32_x3(__p0) __extension__ ({ \
9857 float32x4x3_t __ret; \
9858 __builtin_neon_vld1q_x3_v(&__ret, __p0, 41); \
9859 __ret; \
9860 })
9861 #else
9862 #define vld1q_f32_x3(__p0) __extension__ ({ \
9863 float32x4x3_t __ret; \
9864 __builtin_neon_vld1q_x3_v(&__ret, __p0, 41); \
9865 \
9866 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
9867 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
9868 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
9869 __ret; \
9870 })
9871 #endif
9872
9873 #ifdef __LITTLE_ENDIAN__
9874 #define vld1q_s32_x3(__p0) __extension__ ({ \
9875 int32x4x3_t __ret; \
9876 __builtin_neon_vld1q_x3_v(&__ret, __p0, 34); \
9877 __ret; \
9878 })
9879 #else
9880 #define vld1q_s32_x3(__p0) __extension__ ({ \
9881 int32x4x3_t __ret; \
9882 __builtin_neon_vld1q_x3_v(&__ret, __p0, 34); \
9883 \
9884 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
9885 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
9886 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
9887 __ret; \
9888 })
9889 #endif
9890
9891 #ifdef __LITTLE_ENDIAN__
9892 #define vld1q_s64_x3(__p0) __extension__ ({ \
9893 int64x2x3_t __ret; \
9894 __builtin_neon_vld1q_x3_v(&__ret, __p0, 35); \
9895 __ret; \
9896 })
9897 #else
9898 #define vld1q_s64_x3(__p0) __extension__ ({ \
9899 int64x2x3_t __ret; \
9900 __builtin_neon_vld1q_x3_v(&__ret, __p0, 35); \
9901 \
9902 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
9903 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
9904 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
9905 __ret; \
9906 })
9907 #endif
9908
9909 #ifdef __LITTLE_ENDIAN__
9910 #define vld1q_s16_x3(__p0) __extension__ ({ \
9911 int16x8x3_t __ret; \
9912 __builtin_neon_vld1q_x3_v(&__ret, __p0, 33); \
9913 __ret; \
9914 })
9915 #else
9916 #define vld1q_s16_x3(__p0) __extension__ ({ \
9917 int16x8x3_t __ret; \
9918 __builtin_neon_vld1q_x3_v(&__ret, __p0, 33); \
9919 \
9920 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
9921 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
9922 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
9923 __ret; \
9924 })
9925 #endif
9926
9927 #ifdef __LITTLE_ENDIAN__
9928 #define vld1_u8_x3(__p0) __extension__ ({ \
9929 uint8x8x3_t __ret; \
9930 __builtin_neon_vld1_x3_v(&__ret, __p0, 16); \
9931 __ret; \
9932 })
9933 #else
9934 #define vld1_u8_x3(__p0) __extension__ ({ \
9935 uint8x8x3_t __ret; \
9936 __builtin_neon_vld1_x3_v(&__ret, __p0, 16); \
9937 \
9938 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
9939 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
9940 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
9941 __ret; \
9942 })
9943 #endif
9944
9945 #ifdef __LITTLE_ENDIAN__
9946 #define vld1_u32_x3(__p0) __extension__ ({ \
9947 uint32x2x3_t __ret; \
9948 __builtin_neon_vld1_x3_v(&__ret, __p0, 18); \
9949 __ret; \
9950 })
9951 #else
9952 #define vld1_u32_x3(__p0) __extension__ ({ \
9953 uint32x2x3_t __ret; \
9954 __builtin_neon_vld1_x3_v(&__ret, __p0, 18); \
9955 \
9956 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
9957 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
9958 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
9959 __ret; \
9960 })
9961 #endif
9962
9963 #define vld1_u64_x3(__p0) __extension__ ({ \
9964 uint64x1x3_t __ret; \
9965 __builtin_neon_vld1_x3_v(&__ret, __p0, 19); \
9966 __ret; \
9967 })
9968 #ifdef __LITTLE_ENDIAN__
9969 #define vld1_u16_x3(__p0) __extension__ ({ \
9970 uint16x4x3_t __ret; \
9971 __builtin_neon_vld1_x3_v(&__ret, __p0, 17); \
9972 __ret; \
9973 })
9974 #else
9975 #define vld1_u16_x3(__p0) __extension__ ({ \
9976 uint16x4x3_t __ret; \
9977 __builtin_neon_vld1_x3_v(&__ret, __p0, 17); \
9978 \
9979 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
9980 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
9981 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
9982 __ret; \
9983 })
9984 #endif
9985
9986 #ifdef __LITTLE_ENDIAN__
9987 #define vld1_s8_x3(__p0) __extension__ ({ \
9988 int8x8x3_t __ret; \
9989 __builtin_neon_vld1_x3_v(&__ret, __p0, 0); \
9990 __ret; \
9991 })
9992 #else
9993 #define vld1_s8_x3(__p0) __extension__ ({ \
9994 int8x8x3_t __ret; \
9995 __builtin_neon_vld1_x3_v(&__ret, __p0, 0); \
9996 \
9997 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
9998 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
9999 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
10000 __ret; \
10001 })
10002 #endif
10003
10004 #ifdef __LITTLE_ENDIAN__
10005 #define vld1_f32_x3(__p0) __extension__ ({ \
10006 float32x2x3_t __ret; \
10007 __builtin_neon_vld1_x3_v(&__ret, __p0, 9); \
10008 __ret; \
10009 })
10010 #else
10011 #define vld1_f32_x3(__p0) __extension__ ({ \
10012 float32x2x3_t __ret; \
10013 __builtin_neon_vld1_x3_v(&__ret, __p0, 9); \
10014 \
10015 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
10016 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
10017 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
10018 __ret; \
10019 })
10020 #endif
10021
10022 #ifdef __LITTLE_ENDIAN__
10023 #define vld1_s32_x3(__p0) __extension__ ({ \
10024 int32x2x3_t __ret; \
10025 __builtin_neon_vld1_x3_v(&__ret, __p0, 2); \
10026 __ret; \
10027 })
10028 #else
10029 #define vld1_s32_x3(__p0) __extension__ ({ \
10030 int32x2x3_t __ret; \
10031 __builtin_neon_vld1_x3_v(&__ret, __p0, 2); \
10032 \
10033 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
10034 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
10035 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
10036 __ret; \
10037 })
10038 #endif
10039
10040 #define vld1_s64_x3(__p0) __extension__ ({ \
10041 int64x1x3_t __ret; \
10042 __builtin_neon_vld1_x3_v(&__ret, __p0, 3); \
10043 __ret; \
10044 })
10045 #ifdef __LITTLE_ENDIAN__
10046 #define vld1_s16_x3(__p0) __extension__ ({ \
10047 int16x4x3_t __ret; \
10048 __builtin_neon_vld1_x3_v(&__ret, __p0, 1); \
10049 __ret; \
10050 })
10051 #else
10052 #define vld1_s16_x3(__p0) __extension__ ({ \
10053 int16x4x3_t __ret; \
10054 __builtin_neon_vld1_x3_v(&__ret, __p0, 1); \
10055 \
10056 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
10057 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
10058 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
10059 __ret; \
10060 })
10061 #endif
10062
10063 #ifdef __LITTLE_ENDIAN__
10064 #define vld1_p8_x4(__p0) __extension__ ({ \
10065 poly8x8x4_t __ret; \
10066 __builtin_neon_vld1_x4_v(&__ret, __p0, 4); \
10067 __ret; \
10068 })
10069 #else
10070 #define vld1_p8_x4(__p0) __extension__ ({ \
10071 poly8x8x4_t __ret; \
10072 __builtin_neon_vld1_x4_v(&__ret, __p0, 4); \
10073 \
10074 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
10075 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
10076 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
10077 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
10078 __ret; \
10079 })
10080 #endif
10081
10082 #ifdef __LITTLE_ENDIAN__
10083 #define vld1_p16_x4(__p0) __extension__ ({ \
10084 poly16x4x4_t __ret; \
10085 __builtin_neon_vld1_x4_v(&__ret, __p0, 5); \
10086 __ret; \
10087 })
10088 #else
10089 #define vld1_p16_x4(__p0) __extension__ ({ \
10090 poly16x4x4_t __ret; \
10091 __builtin_neon_vld1_x4_v(&__ret, __p0, 5); \
10092 \
10093 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
10094 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
10095 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
10096 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
10097 __ret; \
10098 })
10099 #endif
10100
10101 #ifdef __LITTLE_ENDIAN__
10102 #define vld1q_p8_x4(__p0) __extension__ ({ \
10103 poly8x16x4_t __ret; \
10104 __builtin_neon_vld1q_x4_v(&__ret, __p0, 36); \
10105 __ret; \
10106 })
10107 #else
10108 #define vld1q_p8_x4(__p0) __extension__ ({ \
10109 poly8x16x4_t __ret; \
10110 __builtin_neon_vld1q_x4_v(&__ret, __p0, 36); \
10111 \
10112 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
10113 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
10114 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
10115 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
10116 __ret; \
10117 })
10118 #endif
10119
10120 #ifdef __LITTLE_ENDIAN__
10121 #define vld1q_p16_x4(__p0) __extension__ ({ \
10122 poly16x8x4_t __ret; \
10123 __builtin_neon_vld1q_x4_v(&__ret, __p0, 37); \
10124 __ret; \
10125 })
10126 #else
10127 #define vld1q_p16_x4(__p0) __extension__ ({ \
10128 poly16x8x4_t __ret; \
10129 __builtin_neon_vld1q_x4_v(&__ret, __p0, 37); \
10130 \
10131 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
10132 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
10133 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
10134 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
10135 __ret; \
10136 })
10137 #endif
10138
10139 #ifdef __LITTLE_ENDIAN__
10140 #define vld1q_u8_x4(__p0) __extension__ ({ \
10141 uint8x16x4_t __ret; \
10142 __builtin_neon_vld1q_x4_v(&__ret, __p0, 48); \
10143 __ret; \
10144 })
10145 #else
10146 #define vld1q_u8_x4(__p0) __extension__ ({ \
10147 uint8x16x4_t __ret; \
10148 __builtin_neon_vld1q_x4_v(&__ret, __p0, 48); \
10149 \
10150 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
10151 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
10152 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
10153 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
10154 __ret; \
10155 })
10156 #endif
10157
10158 #ifdef __LITTLE_ENDIAN__
10159 #define vld1q_u32_x4(__p0) __extension__ ({ \
10160 uint32x4x4_t __ret; \
10161 __builtin_neon_vld1q_x4_v(&__ret, __p0, 50); \
10162 __ret; \
10163 })
10164 #else
10165 #define vld1q_u32_x4(__p0) __extension__ ({ \
10166 uint32x4x4_t __ret; \
10167 __builtin_neon_vld1q_x4_v(&__ret, __p0, 50); \
10168 \
10169 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
10170 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
10171 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
10172 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
10173 __ret; \
10174 })
10175 #endif
10176
10177 #ifdef __LITTLE_ENDIAN__
10178 #define vld1q_u64_x4(__p0) __extension__ ({ \
10179 uint64x2x4_t __ret; \
10180 __builtin_neon_vld1q_x4_v(&__ret, __p0, 51); \
10181 __ret; \
10182 })
10183 #else
10184 #define vld1q_u64_x4(__p0) __extension__ ({ \
10185 uint64x2x4_t __ret; \
10186 __builtin_neon_vld1q_x4_v(&__ret, __p0, 51); \
10187 \
10188 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
10189 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
10190 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
10191 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
10192 __ret; \
10193 })
10194 #endif
10195
10196 #ifdef __LITTLE_ENDIAN__
10197 #define vld1q_u16_x4(__p0) __extension__ ({ \
10198 uint16x8x4_t __ret; \
10199 __builtin_neon_vld1q_x4_v(&__ret, __p0, 49); \
10200 __ret; \
10201 })
10202 #else
10203 #define vld1q_u16_x4(__p0) __extension__ ({ \
10204 uint16x8x4_t __ret; \
10205 __builtin_neon_vld1q_x4_v(&__ret, __p0, 49); \
10206 \
10207 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
10208 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
10209 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
10210 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
10211 __ret; \
10212 })
10213 #endif
10214
10215 #ifdef __LITTLE_ENDIAN__
10216 #define vld1q_s8_x4(__p0) __extension__ ({ \
10217 int8x16x4_t __ret; \
10218 __builtin_neon_vld1q_x4_v(&__ret, __p0, 32); \
10219 __ret; \
10220 })
10221 #else
10222 #define vld1q_s8_x4(__p0) __extension__ ({ \
10223 int8x16x4_t __ret; \
10224 __builtin_neon_vld1q_x4_v(&__ret, __p0, 32); \
10225 \
10226 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
10227 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
10228 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
10229 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
10230 __ret; \
10231 })
10232 #endif
10233
10234 #ifdef __LITTLE_ENDIAN__
10235 #define vld1q_f32_x4(__p0) __extension__ ({ \
10236 float32x4x4_t __ret; \
10237 __builtin_neon_vld1q_x4_v(&__ret, __p0, 41); \
10238 __ret; \
10239 })
10240 #else
10241 #define vld1q_f32_x4(__p0) __extension__ ({ \
10242 float32x4x4_t __ret; \
10243 __builtin_neon_vld1q_x4_v(&__ret, __p0, 41); \
10244 \
10245 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
10246 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
10247 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
10248 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
10249 __ret; \
10250 })
10251 #endif
10252
10253 #ifdef __LITTLE_ENDIAN__
10254 #define vld1q_s32_x4(__p0) __extension__ ({ \
10255 int32x4x4_t __ret; \
10256 __builtin_neon_vld1q_x4_v(&__ret, __p0, 34); \
10257 __ret; \
10258 })
10259 #else
10260 #define vld1q_s32_x4(__p0) __extension__ ({ \
10261 int32x4x4_t __ret; \
10262 __builtin_neon_vld1q_x4_v(&__ret, __p0, 34); \
10263 \
10264 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
10265 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
10266 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
10267 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
10268 __ret; \
10269 })
10270 #endif
10271
10272 #ifdef __LITTLE_ENDIAN__
10273 #define vld1q_s64_x4(__p0) __extension__ ({ \
10274 int64x2x4_t __ret; \
10275 __builtin_neon_vld1q_x4_v(&__ret, __p0, 35); \
10276 __ret; \
10277 })
10278 #else
10279 #define vld1q_s64_x4(__p0) __extension__ ({ \
10280 int64x2x4_t __ret; \
10281 __builtin_neon_vld1q_x4_v(&__ret, __p0, 35); \
10282 \
10283 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
10284 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
10285 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
10286 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
10287 __ret; \
10288 })
10289 #endif
10290
10291 #ifdef __LITTLE_ENDIAN__
10292 #define vld1q_s16_x4(__p0) __extension__ ({ \
10293 int16x8x4_t __ret; \
10294 __builtin_neon_vld1q_x4_v(&__ret, __p0, 33); \
10295 __ret; \
10296 })
10297 #else
10298 #define vld1q_s16_x4(__p0) __extension__ ({ \
10299 int16x8x4_t __ret; \
10300 __builtin_neon_vld1q_x4_v(&__ret, __p0, 33); \
10301 \
10302 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
10303 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
10304 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
10305 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
10306 __ret; \
10307 })
10308 #endif
10309
10310 #ifdef __LITTLE_ENDIAN__
10311 #define vld1_u8_x4(__p0) __extension__ ({ \
10312 uint8x8x4_t __ret; \
10313 __builtin_neon_vld1_x4_v(&__ret, __p0, 16); \
10314 __ret; \
10315 })
10316 #else
10317 #define vld1_u8_x4(__p0) __extension__ ({ \
10318 uint8x8x4_t __ret; \
10319 __builtin_neon_vld1_x4_v(&__ret, __p0, 16); \
10320 \
10321 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
10322 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
10323 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
10324 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
10325 __ret; \
10326 })
10327 #endif
10328
10329 #ifdef __LITTLE_ENDIAN__
10330 #define vld1_u32_x4(__p0) __extension__ ({ \
10331 uint32x2x4_t __ret; \
10332 __builtin_neon_vld1_x4_v(&__ret, __p0, 18); \
10333 __ret; \
10334 })
10335 #else
10336 #define vld1_u32_x4(__p0) __extension__ ({ \
10337 uint32x2x4_t __ret; \
10338 __builtin_neon_vld1_x4_v(&__ret, __p0, 18); \
10339 \
10340 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
10341 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
10342 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
10343 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
10344 __ret; \
10345 })
10346 #endif
10347
10348 #define vld1_u64_x4(__p0) __extension__ ({ \
10349 uint64x1x4_t __ret; \
10350 __builtin_neon_vld1_x4_v(&__ret, __p0, 19); \
10351 __ret; \
10352 })
10353 #ifdef __LITTLE_ENDIAN__
10354 #define vld1_u16_x4(__p0) __extension__ ({ \
10355 uint16x4x4_t __ret; \
10356 __builtin_neon_vld1_x4_v(&__ret, __p0, 17); \
10357 __ret; \
10358 })
10359 #else
10360 #define vld1_u16_x4(__p0) __extension__ ({ \
10361 uint16x4x4_t __ret; \
10362 __builtin_neon_vld1_x4_v(&__ret, __p0, 17); \
10363 \
10364 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
10365 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
10366 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
10367 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
10368 __ret; \
10369 })
10370 #endif
10371
10372 #ifdef __LITTLE_ENDIAN__
10373 #define vld1_s8_x4(__p0) __extension__ ({ \
10374 int8x8x4_t __ret; \
10375 __builtin_neon_vld1_x4_v(&__ret, __p0, 0); \
10376 __ret; \
10377 })
10378 #else
10379 #define vld1_s8_x4(__p0) __extension__ ({ \
10380 int8x8x4_t __ret; \
10381 __builtin_neon_vld1_x4_v(&__ret, __p0, 0); \
10382 \
10383 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
10384 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
10385 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
10386 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
10387 __ret; \
10388 })
10389 #endif
10390
10391 #ifdef __LITTLE_ENDIAN__
10392 #define vld1_f32_x4(__p0) __extension__ ({ \
10393 float32x2x4_t __ret; \
10394 __builtin_neon_vld1_x4_v(&__ret, __p0, 9); \
10395 __ret; \
10396 })
10397 #else
10398 #define vld1_f32_x4(__p0) __extension__ ({ \
10399 float32x2x4_t __ret; \
10400 __builtin_neon_vld1_x4_v(&__ret, __p0, 9); \
10401 \
10402 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
10403 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
10404 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
10405 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
10406 __ret; \
10407 })
10408 #endif
10409
10410 #ifdef __LITTLE_ENDIAN__
10411 #define vld1_s32_x4(__p0) __extension__ ({ \
10412 int32x2x4_t __ret; \
10413 __builtin_neon_vld1_x4_v(&__ret, __p0, 2); \
10414 __ret; \
10415 })
10416 #else
10417 #define vld1_s32_x4(__p0) __extension__ ({ \
10418 int32x2x4_t __ret; \
10419 __builtin_neon_vld1_x4_v(&__ret, __p0, 2); \
10420 \
10421 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
10422 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
10423 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
10424 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
10425 __ret; \
10426 })
10427 #endif
10428
10429 #define vld1_s64_x4(__p0) __extension__ ({ \
10430 int64x1x4_t __ret; \
10431 __builtin_neon_vld1_x4_v(&__ret, __p0, 3); \
10432 __ret; \
10433 })
10434 #ifdef __LITTLE_ENDIAN__
10435 #define vld1_s16_x4(__p0) __extension__ ({ \
10436 int16x4x4_t __ret; \
10437 __builtin_neon_vld1_x4_v(&__ret, __p0, 1); \
10438 __ret; \
10439 })
10440 #else
10441 #define vld1_s16_x4(__p0) __extension__ ({ \
10442 int16x4x4_t __ret; \
10443 __builtin_neon_vld1_x4_v(&__ret, __p0, 1); \
10444 \
10445 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
10446 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
10447 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
10448 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
10449 __ret; \
10450 })
10451 #endif
10452
10453 #ifdef __LITTLE_ENDIAN__
10454 #define vld2_p8(__p0) __extension__ ({ \
10455 poly8x8x2_t __ret; \
10456 __builtin_neon_vld2_v(&__ret, __p0, 4); \
10457 __ret; \
10458 })
10459 #else
10460 #define vld2_p8(__p0) __extension__ ({ \
10461 poly8x8x2_t __ret; \
10462 __builtin_neon_vld2_v(&__ret, __p0, 4); \
10463 \
10464 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
10465 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
10466 __ret; \
10467 })
10468 #endif
10469
10470 #ifdef __LITTLE_ENDIAN__
10471 #define vld2_p16(__p0) __extension__ ({ \
10472 poly16x4x2_t __ret; \
10473 __builtin_neon_vld2_v(&__ret, __p0, 5); \
10474 __ret; \
10475 })
10476 #else
10477 #define vld2_p16(__p0) __extension__ ({ \
10478 poly16x4x2_t __ret; \
10479 __builtin_neon_vld2_v(&__ret, __p0, 5); \
10480 \
10481 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
10482 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
10483 __ret; \
10484 })
10485 #endif
10486
10487 #ifdef __LITTLE_ENDIAN__
10488 #define vld2q_p8(__p0) __extension__ ({ \
10489 poly8x16x2_t __ret; \
10490 __builtin_neon_vld2q_v(&__ret, __p0, 36); \
10491 __ret; \
10492 })
10493 #else
10494 #define vld2q_p8(__p0) __extension__ ({ \
10495 poly8x16x2_t __ret; \
10496 __builtin_neon_vld2q_v(&__ret, __p0, 36); \
10497 \
10498 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
10499 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
10500 __ret; \
10501 })
10502 #endif
10503
10504 #ifdef __LITTLE_ENDIAN__
10505 #define vld2q_p16(__p0) __extension__ ({ \
10506 poly16x8x2_t __ret; \
10507 __builtin_neon_vld2q_v(&__ret, __p0, 37); \
10508 __ret; \
10509 })
10510 #else
10511 #define vld2q_p16(__p0) __extension__ ({ \
10512 poly16x8x2_t __ret; \
10513 __builtin_neon_vld2q_v(&__ret, __p0, 37); \
10514 \
10515 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
10516 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
10517 __ret; \
10518 })
10519 #endif
10520
10521 #ifdef __LITTLE_ENDIAN__
10522 #define vld2q_u8(__p0) __extension__ ({ \
10523 uint8x16x2_t __ret; \
10524 __builtin_neon_vld2q_v(&__ret, __p0, 48); \
10525 __ret; \
10526 })
10527 #else
10528 #define vld2q_u8(__p0) __extension__ ({ \
10529 uint8x16x2_t __ret; \
10530 __builtin_neon_vld2q_v(&__ret, __p0, 48); \
10531 \
10532 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
10533 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
10534 __ret; \
10535 })
10536 #endif
10537
10538 #ifdef __LITTLE_ENDIAN__
10539 #define vld2q_u32(__p0) __extension__ ({ \
10540 uint32x4x2_t __ret; \
10541 __builtin_neon_vld2q_v(&__ret, __p0, 50); \
10542 __ret; \
10543 })
10544 #else
10545 #define vld2q_u32(__p0) __extension__ ({ \
10546 uint32x4x2_t __ret; \
10547 __builtin_neon_vld2q_v(&__ret, __p0, 50); \
10548 \
10549 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
10550 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
10551 __ret; \
10552 })
10553 #endif
10554
10555 #ifdef __LITTLE_ENDIAN__
10556 #define vld2q_u16(__p0) __extension__ ({ \
10557 uint16x8x2_t __ret; \
10558 __builtin_neon_vld2q_v(&__ret, __p0, 49); \
10559 __ret; \
10560 })
10561 #else
10562 #define vld2q_u16(__p0) __extension__ ({ \
10563 uint16x8x2_t __ret; \
10564 __builtin_neon_vld2q_v(&__ret, __p0, 49); \
10565 \
10566 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
10567 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
10568 __ret; \
10569 })
10570 #endif
10571
10572 #ifdef __LITTLE_ENDIAN__
10573 #define vld2q_s8(__p0) __extension__ ({ \
10574 int8x16x2_t __ret; \
10575 __builtin_neon_vld2q_v(&__ret, __p0, 32); \
10576 __ret; \
10577 })
10578 #else
10579 #define vld2q_s8(__p0) __extension__ ({ \
10580 int8x16x2_t __ret; \
10581 __builtin_neon_vld2q_v(&__ret, __p0, 32); \
10582 \
10583 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
10584 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
10585 __ret; \
10586 })
10587 #endif
10588
10589 #ifdef __LITTLE_ENDIAN__
10590 #define vld2q_f32(__p0) __extension__ ({ \
10591 float32x4x2_t __ret; \
10592 __builtin_neon_vld2q_v(&__ret, __p0, 41); \
10593 __ret; \
10594 })
10595 #else
10596 #define vld2q_f32(__p0) __extension__ ({ \
10597 float32x4x2_t __ret; \
10598 __builtin_neon_vld2q_v(&__ret, __p0, 41); \
10599 \
10600 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
10601 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
10602 __ret; \
10603 })
10604 #endif
10605
10606 #ifdef __LITTLE_ENDIAN__
10607 #define vld2q_s32(__p0) __extension__ ({ \
10608 int32x4x2_t __ret; \
10609 __builtin_neon_vld2q_v(&__ret, __p0, 34); \
10610 __ret; \
10611 })
10612 #else
10613 #define vld2q_s32(__p0) __extension__ ({ \
10614 int32x4x2_t __ret; \
10615 __builtin_neon_vld2q_v(&__ret, __p0, 34); \
10616 \
10617 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
10618 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
10619 __ret; \
10620 })
10621 #endif
10622
10623 #ifdef __LITTLE_ENDIAN__
10624 #define vld2q_s16(__p0) __extension__ ({ \
10625 int16x8x2_t __ret; \
10626 __builtin_neon_vld2q_v(&__ret, __p0, 33); \
10627 __ret; \
10628 })
10629 #else
10630 #define vld2q_s16(__p0) __extension__ ({ \
10631 int16x8x2_t __ret; \
10632 __builtin_neon_vld2q_v(&__ret, __p0, 33); \
10633 \
10634 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
10635 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
10636 __ret; \
10637 })
10638 #endif
10639
10640 #ifdef __LITTLE_ENDIAN__
10641 #define vld2_u8(__p0) __extension__ ({ \
10642 uint8x8x2_t __ret; \
10643 __builtin_neon_vld2_v(&__ret, __p0, 16); \
10644 __ret; \
10645 })
10646 #else
10647 #define vld2_u8(__p0) __extension__ ({ \
10648 uint8x8x2_t __ret; \
10649 __builtin_neon_vld2_v(&__ret, __p0, 16); \
10650 \
10651 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
10652 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
10653 __ret; \
10654 })
10655 #endif
10656
10657 #ifdef __LITTLE_ENDIAN__
10658 #define vld2_u32(__p0) __extension__ ({ \
10659 uint32x2x2_t __ret; \
10660 __builtin_neon_vld2_v(&__ret, __p0, 18); \
10661 __ret; \
10662 })
10663 #else
10664 #define vld2_u32(__p0) __extension__ ({ \
10665 uint32x2x2_t __ret; \
10666 __builtin_neon_vld2_v(&__ret, __p0, 18); \
10667 \
10668 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
10669 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
10670 __ret; \
10671 })
10672 #endif
10673
10674 #define vld2_u64(__p0) __extension__ ({ \
10675 uint64x1x2_t __ret; \
10676 __builtin_neon_vld2_v(&__ret, __p0, 19); \
10677 __ret; \
10678 })
10679 #ifdef __LITTLE_ENDIAN__
10680 #define vld2_u16(__p0) __extension__ ({ \
10681 uint16x4x2_t __ret; \
10682 __builtin_neon_vld2_v(&__ret, __p0, 17); \
10683 __ret; \
10684 })
10685 #else
10686 #define vld2_u16(__p0) __extension__ ({ \
10687 uint16x4x2_t __ret; \
10688 __builtin_neon_vld2_v(&__ret, __p0, 17); \
10689 \
10690 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
10691 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
10692 __ret; \
10693 })
10694 #endif
10695
10696 #ifdef __LITTLE_ENDIAN__
10697 #define vld2_s8(__p0) __extension__ ({ \
10698 int8x8x2_t __ret; \
10699 __builtin_neon_vld2_v(&__ret, __p0, 0); \
10700 __ret; \
10701 })
10702 #else
10703 #define vld2_s8(__p0) __extension__ ({ \
10704 int8x8x2_t __ret; \
10705 __builtin_neon_vld2_v(&__ret, __p0, 0); \
10706 \
10707 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
10708 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
10709 __ret; \
10710 })
10711 #endif
10712
10713 #ifdef __LITTLE_ENDIAN__
10714 #define vld2_f32(__p0) __extension__ ({ \
10715 float32x2x2_t __ret; \
10716 __builtin_neon_vld2_v(&__ret, __p0, 9); \
10717 __ret; \
10718 })
10719 #else
10720 #define vld2_f32(__p0) __extension__ ({ \
10721 float32x2x2_t __ret; \
10722 __builtin_neon_vld2_v(&__ret, __p0, 9); \
10723 \
10724 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
10725 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
10726 __ret; \
10727 })
10728 #endif
10729
10730 #ifdef __LITTLE_ENDIAN__
10731 #define vld2_s32(__p0) __extension__ ({ \
10732 int32x2x2_t __ret; \
10733 __builtin_neon_vld2_v(&__ret, __p0, 2); \
10734 __ret; \
10735 })
10736 #else
10737 #define vld2_s32(__p0) __extension__ ({ \
10738 int32x2x2_t __ret; \
10739 __builtin_neon_vld2_v(&__ret, __p0, 2); \
10740 \
10741 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
10742 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
10743 __ret; \
10744 })
10745 #endif
10746
10747 #define vld2_s64(__p0) __extension__ ({ \
10748 int64x1x2_t __ret; \
10749 __builtin_neon_vld2_v(&__ret, __p0, 3); \
10750 __ret; \
10751 })
10752 #ifdef __LITTLE_ENDIAN__
10753 #define vld2_s16(__p0) __extension__ ({ \
10754 int16x4x2_t __ret; \
10755 __builtin_neon_vld2_v(&__ret, __p0, 1); \
10756 __ret; \
10757 })
10758 #else
10759 #define vld2_s16(__p0) __extension__ ({ \
10760 int16x4x2_t __ret; \
10761 __builtin_neon_vld2_v(&__ret, __p0, 1); \
10762 \
10763 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
10764 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
10765 __ret; \
10766 })
10767 #endif
10768
10769 #ifdef __LITTLE_ENDIAN__
10770 #define vld2_dup_p8(__p0) __extension__ ({ \
10771 poly8x8x2_t __ret; \
10772 __builtin_neon_vld2_dup_v(&__ret, __p0, 4); \
10773 __ret; \
10774 })
10775 #else
10776 #define vld2_dup_p8(__p0) __extension__ ({ \
10777 poly8x8x2_t __ret; \
10778 __builtin_neon_vld2_dup_v(&__ret, __p0, 4); \
10779 \
10780 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
10781 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
10782 __ret; \
10783 })
10784 #endif
10785
10786 #ifdef __LITTLE_ENDIAN__
10787 #define vld2_dup_p16(__p0) __extension__ ({ \
10788 poly16x4x2_t __ret; \
10789 __builtin_neon_vld2_dup_v(&__ret, __p0, 5); \
10790 __ret; \
10791 })
10792 #else
10793 #define vld2_dup_p16(__p0) __extension__ ({ \
10794 poly16x4x2_t __ret; \
10795 __builtin_neon_vld2_dup_v(&__ret, __p0, 5); \
10796 \
10797 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
10798 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
10799 __ret; \
10800 })
10801 #endif
10802
10803 #ifdef __LITTLE_ENDIAN__
10804 #define vld2q_dup_p8(__p0) __extension__ ({ \
10805 poly8x16x2_t __ret; \
10806 __builtin_neon_vld2q_dup_v(&__ret, __p0, 36); \
10807 __ret; \
10808 })
10809 #else
10810 #define vld2q_dup_p8(__p0) __extension__ ({ \
10811 poly8x16x2_t __ret; \
10812 __builtin_neon_vld2q_dup_v(&__ret, __p0, 36); \
10813 \
10814 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
10815 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
10816 __ret; \
10817 })
10818 #endif
10819
10820 #ifdef __LITTLE_ENDIAN__
10821 #define vld2q_dup_p16(__p0) __extension__ ({ \
10822 poly16x8x2_t __ret; \
10823 __builtin_neon_vld2q_dup_v(&__ret, __p0, 37); \
10824 __ret; \
10825 })
10826 #else
10827 #define vld2q_dup_p16(__p0) __extension__ ({ \
10828 poly16x8x2_t __ret; \
10829 __builtin_neon_vld2q_dup_v(&__ret, __p0, 37); \
10830 \
10831 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
10832 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
10833 __ret; \
10834 })
10835 #endif
10836
10837 #ifdef __LITTLE_ENDIAN__
10838 #define vld2q_dup_u8(__p0) __extension__ ({ \
10839 uint8x16x2_t __ret; \
10840 __builtin_neon_vld2q_dup_v(&__ret, __p0, 48); \
10841 __ret; \
10842 })
10843 #else
10844 #define vld2q_dup_u8(__p0) __extension__ ({ \
10845 uint8x16x2_t __ret; \
10846 __builtin_neon_vld2q_dup_v(&__ret, __p0, 48); \
10847 \
10848 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
10849 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
10850 __ret; \
10851 })
10852 #endif
10853
10854 #ifdef __LITTLE_ENDIAN__
10855 #define vld2q_dup_u32(__p0) __extension__ ({ \
10856 uint32x4x2_t __ret; \
10857 __builtin_neon_vld2q_dup_v(&__ret, __p0, 50); \
10858 __ret; \
10859 })
10860 #else
10861 #define vld2q_dup_u32(__p0) __extension__ ({ \
10862 uint32x4x2_t __ret; \
10863 __builtin_neon_vld2q_dup_v(&__ret, __p0, 50); \
10864 \
10865 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
10866 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
10867 __ret; \
10868 })
10869 #endif
10870
10871 #ifdef __LITTLE_ENDIAN__
10872 #define vld2q_dup_u64(__p0) __extension__ ({ \
10873 uint64x2x2_t __ret; \
10874 __builtin_neon_vld2q_dup_v(&__ret, __p0, 51); \
10875 __ret; \
10876 })
10877 #else
10878 #define vld2q_dup_u64(__p0) __extension__ ({ \
10879 uint64x2x2_t __ret; \
10880 __builtin_neon_vld2q_dup_v(&__ret, __p0, 51); \
10881 \
10882 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
10883 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
10884 __ret; \
10885 })
10886 #endif
10887
10888 #ifdef __LITTLE_ENDIAN__
10889 #define vld2q_dup_u16(__p0) __extension__ ({ \
10890 uint16x8x2_t __ret; \
10891 __builtin_neon_vld2q_dup_v(&__ret, __p0, 49); \
10892 __ret; \
10893 })
10894 #else
10895 #define vld2q_dup_u16(__p0) __extension__ ({ \
10896 uint16x8x2_t __ret; \
10897 __builtin_neon_vld2q_dup_v(&__ret, __p0, 49); \
10898 \
10899 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
10900 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
10901 __ret; \
10902 })
10903 #endif
10904
10905 #ifdef __LITTLE_ENDIAN__
10906 #define vld2q_dup_s8(__p0) __extension__ ({ \
10907 int8x16x2_t __ret; \
10908 __builtin_neon_vld2q_dup_v(&__ret, __p0, 32); \
10909 __ret; \
10910 })
10911 #else
10912 #define vld2q_dup_s8(__p0) __extension__ ({ \
10913 int8x16x2_t __ret; \
10914 __builtin_neon_vld2q_dup_v(&__ret, __p0, 32); \
10915 \
10916 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
10917 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
10918 __ret; \
10919 })
10920 #endif
10921
10922 #ifdef __LITTLE_ENDIAN__
10923 #define vld2q_dup_f32(__p0) __extension__ ({ \
10924 float32x4x2_t __ret; \
10925 __builtin_neon_vld2q_dup_v(&__ret, __p0, 41); \
10926 __ret; \
10927 })
10928 #else
10929 #define vld2q_dup_f32(__p0) __extension__ ({ \
10930 float32x4x2_t __ret; \
10931 __builtin_neon_vld2q_dup_v(&__ret, __p0, 41); \
10932 \
10933 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
10934 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
10935 __ret; \
10936 })
10937 #endif
10938
10939 #ifdef __LITTLE_ENDIAN__
10940 #define vld2q_dup_s32(__p0) __extension__ ({ \
10941 int32x4x2_t __ret; \
10942 __builtin_neon_vld2q_dup_v(&__ret, __p0, 34); \
10943 __ret; \
10944 })
10945 #else
10946 #define vld2q_dup_s32(__p0) __extension__ ({ \
10947 int32x4x2_t __ret; \
10948 __builtin_neon_vld2q_dup_v(&__ret, __p0, 34); \
10949 \
10950 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
10951 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
10952 __ret; \
10953 })
10954 #endif
10955
10956 #ifdef __LITTLE_ENDIAN__
10957 #define vld2q_dup_s64(__p0) __extension__ ({ \
10958 int64x2x2_t __ret; \
10959 __builtin_neon_vld2q_dup_v(&__ret, __p0, 35); \
10960 __ret; \
10961 })
10962 #else
10963 #define vld2q_dup_s64(__p0) __extension__ ({ \
10964 int64x2x2_t __ret; \
10965 __builtin_neon_vld2q_dup_v(&__ret, __p0, 35); \
10966 \
10967 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
10968 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
10969 __ret; \
10970 })
10971 #endif
10972
10973 #ifdef __LITTLE_ENDIAN__
10974 #define vld2q_dup_s16(__p0) __extension__ ({ \
10975 int16x8x2_t __ret; \
10976 __builtin_neon_vld2q_dup_v(&__ret, __p0, 33); \
10977 __ret; \
10978 })
10979 #else
10980 #define vld2q_dup_s16(__p0) __extension__ ({ \
10981 int16x8x2_t __ret; \
10982 __builtin_neon_vld2q_dup_v(&__ret, __p0, 33); \
10983 \
10984 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
10985 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
10986 __ret; \
10987 })
10988 #endif
10989
10990 #ifdef __LITTLE_ENDIAN__
10991 #define vld2_dup_u8(__p0) __extension__ ({ \
10992 uint8x8x2_t __ret; \
10993 __builtin_neon_vld2_dup_v(&__ret, __p0, 16); \
10994 __ret; \
10995 })
10996 #else
10997 #define vld2_dup_u8(__p0) __extension__ ({ \
10998 uint8x8x2_t __ret; \
10999 __builtin_neon_vld2_dup_v(&__ret, __p0, 16); \
11000 \
11001 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
11002 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
11003 __ret; \
11004 })
11005 #endif
11006
11007 #ifdef __LITTLE_ENDIAN__
11008 #define vld2_dup_u32(__p0) __extension__ ({ \
11009 uint32x2x2_t __ret; \
11010 __builtin_neon_vld2_dup_v(&__ret, __p0, 18); \
11011 __ret; \
11012 })
11013 #else
11014 #define vld2_dup_u32(__p0) __extension__ ({ \
11015 uint32x2x2_t __ret; \
11016 __builtin_neon_vld2_dup_v(&__ret, __p0, 18); \
11017 \
11018 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
11019 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
11020 __ret; \
11021 })
11022 #endif
11023
11024 #define vld2_dup_u64(__p0) __extension__ ({ \
11025 uint64x1x2_t __ret; \
11026 __builtin_neon_vld2_dup_v(&__ret, __p0, 19); \
11027 __ret; \
11028 })
11029 #ifdef __LITTLE_ENDIAN__
11030 #define vld2_dup_u16(__p0) __extension__ ({ \
11031 uint16x4x2_t __ret; \
11032 __builtin_neon_vld2_dup_v(&__ret, __p0, 17); \
11033 __ret; \
11034 })
11035 #else
11036 #define vld2_dup_u16(__p0) __extension__ ({ \
11037 uint16x4x2_t __ret; \
11038 __builtin_neon_vld2_dup_v(&__ret, __p0, 17); \
11039 \
11040 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
11041 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
11042 __ret; \
11043 })
11044 #endif
11045
11046 #ifdef __LITTLE_ENDIAN__
11047 #define vld2_dup_s8(__p0) __extension__ ({ \
11048 int8x8x2_t __ret; \
11049 __builtin_neon_vld2_dup_v(&__ret, __p0, 0); \
11050 __ret; \
11051 })
11052 #else
11053 #define vld2_dup_s8(__p0) __extension__ ({ \
11054 int8x8x2_t __ret; \
11055 __builtin_neon_vld2_dup_v(&__ret, __p0, 0); \
11056 \
11057 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
11058 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
11059 __ret; \
11060 })
11061 #endif
11062
11063 #ifdef __LITTLE_ENDIAN__
11064 #define vld2_dup_f32(__p0) __extension__ ({ \
11065 float32x2x2_t __ret; \
11066 __builtin_neon_vld2_dup_v(&__ret, __p0, 9); \
11067 __ret; \
11068 })
11069 #else
11070 #define vld2_dup_f32(__p0) __extension__ ({ \
11071 float32x2x2_t __ret; \
11072 __builtin_neon_vld2_dup_v(&__ret, __p0, 9); \
11073 \
11074 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
11075 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
11076 __ret; \
11077 })
11078 #endif
11079
11080 #ifdef __LITTLE_ENDIAN__
11081 #define vld2_dup_s32(__p0) __extension__ ({ \
11082 int32x2x2_t __ret; \
11083 __builtin_neon_vld2_dup_v(&__ret, __p0, 2); \
11084 __ret; \
11085 })
11086 #else
11087 #define vld2_dup_s32(__p0) __extension__ ({ \
11088 int32x2x2_t __ret; \
11089 __builtin_neon_vld2_dup_v(&__ret, __p0, 2); \
11090 \
11091 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
11092 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
11093 __ret; \
11094 })
11095 #endif
11096
11097 #define vld2_dup_s64(__p0) __extension__ ({ \
11098 int64x1x2_t __ret; \
11099 __builtin_neon_vld2_dup_v(&__ret, __p0, 3); \
11100 __ret; \
11101 })
11102 #ifdef __LITTLE_ENDIAN__
11103 #define vld2_dup_s16(__p0) __extension__ ({ \
11104 int16x4x2_t __ret; \
11105 __builtin_neon_vld2_dup_v(&__ret, __p0, 1); \
11106 __ret; \
11107 })
11108 #else
11109 #define vld2_dup_s16(__p0) __extension__ ({ \
11110 int16x4x2_t __ret; \
11111 __builtin_neon_vld2_dup_v(&__ret, __p0, 1); \
11112 \
11113 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
11114 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
11115 __ret; \
11116 })
11117 #endif
11118
11119 #ifdef __LITTLE_ENDIAN__
11120 #define vld2_lane_p8(__p0, __p1, __p2) __extension__ ({ \
11121 poly8x8x2_t __ret; \
11122 poly8x8x2_t __s1 = __p1; \
11123 __builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 4); \
11124 __ret; \
11125 })
11126 #else
11127 #define vld2_lane_p8(__p0, __p1, __p2) __extension__ ({ \
11128 poly8x8x2_t __ret; \
11129 poly8x8x2_t __s1 = __p1; \
11130 poly8x8x2_t __rev1; \
11131 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
11132 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
11133 __builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 4); \
11134 \
11135 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
11136 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
11137 __ret; \
11138 })
11139 #endif
11140
11141 #ifdef __LITTLE_ENDIAN__
11142 #define vld2_lane_p16(__p0, __p1, __p2) __extension__ ({ \
11143 poly16x4x2_t __ret; \
11144 poly16x4x2_t __s1 = __p1; \
11145 __builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 5); \
11146 __ret; \
11147 })
11148 #else
11149 #define vld2_lane_p16(__p0, __p1, __p2) __extension__ ({ \
11150 poly16x4x2_t __ret; \
11151 poly16x4x2_t __s1 = __p1; \
11152 poly16x4x2_t __rev1; \
11153 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
11154 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
11155 __builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 5); \
11156 \
11157 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
11158 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
11159 __ret; \
11160 })
11161 #endif
11162
11163 #ifdef __LITTLE_ENDIAN__
11164 #define vld2q_lane_p16(__p0, __p1, __p2) __extension__ ({ \
11165 poly16x8x2_t __ret; \
11166 poly16x8x2_t __s1 = __p1; \
11167 __builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 37); \
11168 __ret; \
11169 })
11170 #else
11171 #define vld2q_lane_p16(__p0, __p1, __p2) __extension__ ({ \
11172 poly16x8x2_t __ret; \
11173 poly16x8x2_t __s1 = __p1; \
11174 poly16x8x2_t __rev1; \
11175 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
11176 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
11177 __builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 37); \
11178 \
11179 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
11180 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
11181 __ret; \
11182 })
11183 #endif
11184
11185 #ifdef __LITTLE_ENDIAN__
11186 #define vld2q_lane_u32(__p0, __p1, __p2) __extension__ ({ \
11187 uint32x4x2_t __ret; \
11188 uint32x4x2_t __s1 = __p1; \
11189 __builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 50); \
11190 __ret; \
11191 })
11192 #else
11193 #define vld2q_lane_u32(__p0, __p1, __p2) __extension__ ({ \
11194 uint32x4x2_t __ret; \
11195 uint32x4x2_t __s1 = __p1; \
11196 uint32x4x2_t __rev1; \
11197 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
11198 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
11199 __builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 50); \
11200 \
11201 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
11202 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
11203 __ret; \
11204 })
11205 #endif
11206
11207 #ifdef __LITTLE_ENDIAN__
11208 #define vld2q_lane_u16(__p0, __p1, __p2) __extension__ ({ \
11209 uint16x8x2_t __ret; \
11210 uint16x8x2_t __s1 = __p1; \
11211 __builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 49); \
11212 __ret; \
11213 })
11214 #else
11215 #define vld2q_lane_u16(__p0, __p1, __p2) __extension__ ({ \
11216 uint16x8x2_t __ret; \
11217 uint16x8x2_t __s1 = __p1; \
11218 uint16x8x2_t __rev1; \
11219 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
11220 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
11221 __builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 49); \
11222 \
11223 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
11224 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
11225 __ret; \
11226 })
11227 #endif
11228
11229 #ifdef __LITTLE_ENDIAN__
11230 #define vld2q_lane_f32(__p0, __p1, __p2) __extension__ ({ \
11231 float32x4x2_t __ret; \
11232 float32x4x2_t __s1 = __p1; \
11233 __builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 41); \
11234 __ret; \
11235 })
11236 #else
11237 #define vld2q_lane_f32(__p0, __p1, __p2) __extension__ ({ \
11238 float32x4x2_t __ret; \
11239 float32x4x2_t __s1 = __p1; \
11240 float32x4x2_t __rev1; \
11241 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
11242 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
11243 __builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 41); \
11244 \
11245 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
11246 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
11247 __ret; \
11248 })
11249 #endif
11250
11251 #ifdef __LITTLE_ENDIAN__
11252 #define vld2q_lane_s32(__p0, __p1, __p2) __extension__ ({ \
11253 int32x4x2_t __ret; \
11254 int32x4x2_t __s1 = __p1; \
11255 __builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 34); \
11256 __ret; \
11257 })
11258 #else
11259 #define vld2q_lane_s32(__p0, __p1, __p2) __extension__ ({ \
11260 int32x4x2_t __ret; \
11261 int32x4x2_t __s1 = __p1; \
11262 int32x4x2_t __rev1; \
11263 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
11264 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
11265 __builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 34); \
11266 \
11267 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
11268 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
11269 __ret; \
11270 })
11271 #endif
11272
11273 #ifdef __LITTLE_ENDIAN__
11274 #define vld2q_lane_s16(__p0, __p1, __p2) __extension__ ({ \
11275 int16x8x2_t __ret; \
11276 int16x8x2_t __s1 = __p1; \
11277 __builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 33); \
11278 __ret; \
11279 })
11280 #else
11281 #define vld2q_lane_s16(__p0, __p1, __p2) __extension__ ({ \
11282 int16x8x2_t __ret; \
11283 int16x8x2_t __s1 = __p1; \
11284 int16x8x2_t __rev1; \
11285 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
11286 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
11287 __builtin_neon_vld2q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 33); \
11288 \
11289 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
11290 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
11291 __ret; \
11292 })
11293 #endif
11294
11295 #ifdef __LITTLE_ENDIAN__
11296 #define vld2_lane_u8(__p0, __p1, __p2) __extension__ ({ \
11297 uint8x8x2_t __ret; \
11298 uint8x8x2_t __s1 = __p1; \
11299 __builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 16); \
11300 __ret; \
11301 })
11302 #else
11303 #define vld2_lane_u8(__p0, __p1, __p2) __extension__ ({ \
11304 uint8x8x2_t __ret; \
11305 uint8x8x2_t __s1 = __p1; \
11306 uint8x8x2_t __rev1; \
11307 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
11308 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
11309 __builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 16); \
11310 \
11311 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
11312 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
11313 __ret; \
11314 })
11315 #endif
11316
11317 #ifdef __LITTLE_ENDIAN__
11318 #define vld2_lane_u32(__p0, __p1, __p2) __extension__ ({ \
11319 uint32x2x2_t __ret; \
11320 uint32x2x2_t __s1 = __p1; \
11321 __builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 18); \
11322 __ret; \
11323 })
11324 #else
11325 #define vld2_lane_u32(__p0, __p1, __p2) __extension__ ({ \
11326 uint32x2x2_t __ret; \
11327 uint32x2x2_t __s1 = __p1; \
11328 uint32x2x2_t __rev1; \
11329 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
11330 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
11331 __builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 18); \
11332 \
11333 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
11334 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
11335 __ret; \
11336 })
11337 #endif
11338
11339 #ifdef __LITTLE_ENDIAN__
11340 #define vld2_lane_u16(__p0, __p1, __p2) __extension__ ({ \
11341 uint16x4x2_t __ret; \
11342 uint16x4x2_t __s1 = __p1; \
11343 __builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 17); \
11344 __ret; \
11345 })
11346 #else
11347 #define vld2_lane_u16(__p0, __p1, __p2) __extension__ ({ \
11348 uint16x4x2_t __ret; \
11349 uint16x4x2_t __s1 = __p1; \
11350 uint16x4x2_t __rev1; \
11351 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
11352 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
11353 __builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 17); \
11354 \
11355 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
11356 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
11357 __ret; \
11358 })
11359 #endif
11360
11361 #ifdef __LITTLE_ENDIAN__
11362 #define vld2_lane_s8(__p0, __p1, __p2) __extension__ ({ \
11363 int8x8x2_t __ret; \
11364 int8x8x2_t __s1 = __p1; \
11365 __builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 0); \
11366 __ret; \
11367 })
11368 #else
11369 #define vld2_lane_s8(__p0, __p1, __p2) __extension__ ({ \
11370 int8x8x2_t __ret; \
11371 int8x8x2_t __s1 = __p1; \
11372 int8x8x2_t __rev1; \
11373 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
11374 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
11375 __builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 0); \
11376 \
11377 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
11378 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
11379 __ret; \
11380 })
11381 #endif
11382
11383 #ifdef __LITTLE_ENDIAN__
11384 #define vld2_lane_f32(__p0, __p1, __p2) __extension__ ({ \
11385 float32x2x2_t __ret; \
11386 float32x2x2_t __s1 = __p1; \
11387 __builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 9); \
11388 __ret; \
11389 })
11390 #else
11391 #define vld2_lane_f32(__p0, __p1, __p2) __extension__ ({ \
11392 float32x2x2_t __ret; \
11393 float32x2x2_t __s1 = __p1; \
11394 float32x2x2_t __rev1; \
11395 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
11396 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
11397 __builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 9); \
11398 \
11399 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
11400 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
11401 __ret; \
11402 })
11403 #endif
11404
11405 #ifdef __LITTLE_ENDIAN__
11406 #define vld2_lane_s32(__p0, __p1, __p2) __extension__ ({ \
11407 int32x2x2_t __ret; \
11408 int32x2x2_t __s1 = __p1; \
11409 __builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 2); \
11410 __ret; \
11411 })
11412 #else
11413 #define vld2_lane_s32(__p0, __p1, __p2) __extension__ ({ \
11414 int32x2x2_t __ret; \
11415 int32x2x2_t __s1 = __p1; \
11416 int32x2x2_t __rev1; \
11417 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
11418 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
11419 __builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 2); \
11420 \
11421 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
11422 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
11423 __ret; \
11424 })
11425 #endif
11426
11427 #ifdef __LITTLE_ENDIAN__
11428 #define vld2_lane_s16(__p0, __p1, __p2) __extension__ ({ \
11429 int16x4x2_t __ret; \
11430 int16x4x2_t __s1 = __p1; \
11431 __builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 1); \
11432 __ret; \
11433 })
11434 #else
11435 #define vld2_lane_s16(__p0, __p1, __p2) __extension__ ({ \
11436 int16x4x2_t __ret; \
11437 int16x4x2_t __s1 = __p1; \
11438 int16x4x2_t __rev1; \
11439 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
11440 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
11441 __builtin_neon_vld2_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 1); \
11442 \
11443 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
11444 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
11445 __ret; \
11446 })
11447 #endif
11448
11449 #ifdef __LITTLE_ENDIAN__
11450 #define vld3_p8(__p0) __extension__ ({ \
11451 poly8x8x3_t __ret; \
11452 __builtin_neon_vld3_v(&__ret, __p0, 4); \
11453 __ret; \
11454 })
11455 #else
11456 #define vld3_p8(__p0) __extension__ ({ \
11457 poly8x8x3_t __ret; \
11458 __builtin_neon_vld3_v(&__ret, __p0, 4); \
11459 \
11460 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
11461 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
11462 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
11463 __ret; \
11464 })
11465 #endif
11466
11467 #ifdef __LITTLE_ENDIAN__
11468 #define vld3_p16(__p0) __extension__ ({ \
11469 poly16x4x3_t __ret; \
11470 __builtin_neon_vld3_v(&__ret, __p0, 5); \
11471 __ret; \
11472 })
11473 #else
11474 #define vld3_p16(__p0) __extension__ ({ \
11475 poly16x4x3_t __ret; \
11476 __builtin_neon_vld3_v(&__ret, __p0, 5); \
11477 \
11478 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
11479 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
11480 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
11481 __ret; \
11482 })
11483 #endif
11484
11485 #ifdef __LITTLE_ENDIAN__
11486 #define vld3q_p8(__p0) __extension__ ({ \
11487 poly8x16x3_t __ret; \
11488 __builtin_neon_vld3q_v(&__ret, __p0, 36); \
11489 __ret; \
11490 })
11491 #else
11492 #define vld3q_p8(__p0) __extension__ ({ \
11493 poly8x16x3_t __ret; \
11494 __builtin_neon_vld3q_v(&__ret, __p0, 36); \
11495 \
11496 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
11497 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
11498 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
11499 __ret; \
11500 })
11501 #endif
11502
11503 #ifdef __LITTLE_ENDIAN__
11504 #define vld3q_p16(__p0) __extension__ ({ \
11505 poly16x8x3_t __ret; \
11506 __builtin_neon_vld3q_v(&__ret, __p0, 37); \
11507 __ret; \
11508 })
11509 #else
11510 #define vld3q_p16(__p0) __extension__ ({ \
11511 poly16x8x3_t __ret; \
11512 __builtin_neon_vld3q_v(&__ret, __p0, 37); \
11513 \
11514 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
11515 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
11516 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
11517 __ret; \
11518 })
11519 #endif
11520
11521 #ifdef __LITTLE_ENDIAN__
11522 #define vld3q_u8(__p0) __extension__ ({ \
11523 uint8x16x3_t __ret; \
11524 __builtin_neon_vld3q_v(&__ret, __p0, 48); \
11525 __ret; \
11526 })
11527 #else
11528 #define vld3q_u8(__p0) __extension__ ({ \
11529 uint8x16x3_t __ret; \
11530 __builtin_neon_vld3q_v(&__ret, __p0, 48); \
11531 \
11532 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
11533 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
11534 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
11535 __ret; \
11536 })
11537 #endif
11538
11539 #ifdef __LITTLE_ENDIAN__
11540 #define vld3q_u32(__p0) __extension__ ({ \
11541 uint32x4x3_t __ret; \
11542 __builtin_neon_vld3q_v(&__ret, __p0, 50); \
11543 __ret; \
11544 })
11545 #else
11546 #define vld3q_u32(__p0) __extension__ ({ \
11547 uint32x4x3_t __ret; \
11548 __builtin_neon_vld3q_v(&__ret, __p0, 50); \
11549 \
11550 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
11551 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
11552 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
11553 __ret; \
11554 })
11555 #endif
11556
11557 #ifdef __LITTLE_ENDIAN__
11558 #define vld3q_u16(__p0) __extension__ ({ \
11559 uint16x8x3_t __ret; \
11560 __builtin_neon_vld3q_v(&__ret, __p0, 49); \
11561 __ret; \
11562 })
11563 #else
11564 #define vld3q_u16(__p0) __extension__ ({ \
11565 uint16x8x3_t __ret; \
11566 __builtin_neon_vld3q_v(&__ret, __p0, 49); \
11567 \
11568 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
11569 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
11570 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
11571 __ret; \
11572 })
11573 #endif
11574
11575 #ifdef __LITTLE_ENDIAN__
11576 #define vld3q_s8(__p0) __extension__ ({ \
11577 int8x16x3_t __ret; \
11578 __builtin_neon_vld3q_v(&__ret, __p0, 32); \
11579 __ret; \
11580 })
11581 #else
11582 #define vld3q_s8(__p0) __extension__ ({ \
11583 int8x16x3_t __ret; \
11584 __builtin_neon_vld3q_v(&__ret, __p0, 32); \
11585 \
11586 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
11587 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
11588 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
11589 __ret; \
11590 })
11591 #endif
11592
11593 #ifdef __LITTLE_ENDIAN__
11594 #define vld3q_f32(__p0) __extension__ ({ \
11595 float32x4x3_t __ret; \
11596 __builtin_neon_vld3q_v(&__ret, __p0, 41); \
11597 __ret; \
11598 })
11599 #else
11600 #define vld3q_f32(__p0) __extension__ ({ \
11601 float32x4x3_t __ret; \
11602 __builtin_neon_vld3q_v(&__ret, __p0, 41); \
11603 \
11604 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
11605 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
11606 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
11607 __ret; \
11608 })
11609 #endif
11610
11611 #ifdef __LITTLE_ENDIAN__
11612 #define vld3q_s32(__p0) __extension__ ({ \
11613 int32x4x3_t __ret; \
11614 __builtin_neon_vld3q_v(&__ret, __p0, 34); \
11615 __ret; \
11616 })
11617 #else
11618 #define vld3q_s32(__p0) __extension__ ({ \
11619 int32x4x3_t __ret; \
11620 __builtin_neon_vld3q_v(&__ret, __p0, 34); \
11621 \
11622 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
11623 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
11624 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
11625 __ret; \
11626 })
11627 #endif
11628
11629 #ifdef __LITTLE_ENDIAN__
11630 #define vld3q_s16(__p0) __extension__ ({ \
11631 int16x8x3_t __ret; \
11632 __builtin_neon_vld3q_v(&__ret, __p0, 33); \
11633 __ret; \
11634 })
11635 #else
11636 #define vld3q_s16(__p0) __extension__ ({ \
11637 int16x8x3_t __ret; \
11638 __builtin_neon_vld3q_v(&__ret, __p0, 33); \
11639 \
11640 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
11641 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
11642 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
11643 __ret; \
11644 })
11645 #endif
11646
11647 #ifdef __LITTLE_ENDIAN__
11648 #define vld3_u8(__p0) __extension__ ({ \
11649 uint8x8x3_t __ret; \
11650 __builtin_neon_vld3_v(&__ret, __p0, 16); \
11651 __ret; \
11652 })
11653 #else
11654 #define vld3_u8(__p0) __extension__ ({ \
11655 uint8x8x3_t __ret; \
11656 __builtin_neon_vld3_v(&__ret, __p0, 16); \
11657 \
11658 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
11659 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
11660 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
11661 __ret; \
11662 })
11663 #endif
11664
11665 #ifdef __LITTLE_ENDIAN__
11666 #define vld3_u32(__p0) __extension__ ({ \
11667 uint32x2x3_t __ret; \
11668 __builtin_neon_vld3_v(&__ret, __p0, 18); \
11669 __ret; \
11670 })
11671 #else
11672 #define vld3_u32(__p0) __extension__ ({ \
11673 uint32x2x3_t __ret; \
11674 __builtin_neon_vld3_v(&__ret, __p0, 18); \
11675 \
11676 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
11677 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
11678 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
11679 __ret; \
11680 })
11681 #endif
11682
11683 #define vld3_u64(__p0) __extension__ ({ \
11684 uint64x1x3_t __ret; \
11685 __builtin_neon_vld3_v(&__ret, __p0, 19); \
11686 __ret; \
11687 })
11688 #ifdef __LITTLE_ENDIAN__
11689 #define vld3_u16(__p0) __extension__ ({ \
11690 uint16x4x3_t __ret; \
11691 __builtin_neon_vld3_v(&__ret, __p0, 17); \
11692 __ret; \
11693 })
11694 #else
11695 #define vld3_u16(__p0) __extension__ ({ \
11696 uint16x4x3_t __ret; \
11697 __builtin_neon_vld3_v(&__ret, __p0, 17); \
11698 \
11699 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
11700 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
11701 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
11702 __ret; \
11703 })
11704 #endif
11705
11706 #ifdef __LITTLE_ENDIAN__
11707 #define vld3_s8(__p0) __extension__ ({ \
11708 int8x8x3_t __ret; \
11709 __builtin_neon_vld3_v(&__ret, __p0, 0); \
11710 __ret; \
11711 })
11712 #else
11713 #define vld3_s8(__p0) __extension__ ({ \
11714 int8x8x3_t __ret; \
11715 __builtin_neon_vld3_v(&__ret, __p0, 0); \
11716 \
11717 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
11718 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
11719 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
11720 __ret; \
11721 })
11722 #endif
11723
11724 #ifdef __LITTLE_ENDIAN__
11725 #define vld3_f32(__p0) __extension__ ({ \
11726 float32x2x3_t __ret; \
11727 __builtin_neon_vld3_v(&__ret, __p0, 9); \
11728 __ret; \
11729 })
11730 #else
11731 #define vld3_f32(__p0) __extension__ ({ \
11732 float32x2x3_t __ret; \
11733 __builtin_neon_vld3_v(&__ret, __p0, 9); \
11734 \
11735 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
11736 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
11737 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
11738 __ret; \
11739 })
11740 #endif
11741
11742 #ifdef __LITTLE_ENDIAN__
11743 #define vld3_s32(__p0) __extension__ ({ \
11744 int32x2x3_t __ret; \
11745 __builtin_neon_vld3_v(&__ret, __p0, 2); \
11746 __ret; \
11747 })
11748 #else
11749 #define vld3_s32(__p0) __extension__ ({ \
11750 int32x2x3_t __ret; \
11751 __builtin_neon_vld3_v(&__ret, __p0, 2); \
11752 \
11753 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
11754 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
11755 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
11756 __ret; \
11757 })
11758 #endif
11759
11760 #define vld3_s64(__p0) __extension__ ({ \
11761 int64x1x3_t __ret; \
11762 __builtin_neon_vld3_v(&__ret, __p0, 3); \
11763 __ret; \
11764 })
11765 #ifdef __LITTLE_ENDIAN__
11766 #define vld3_s16(__p0) __extension__ ({ \
11767 int16x4x3_t __ret; \
11768 __builtin_neon_vld3_v(&__ret, __p0, 1); \
11769 __ret; \
11770 })
11771 #else
11772 #define vld3_s16(__p0) __extension__ ({ \
11773 int16x4x3_t __ret; \
11774 __builtin_neon_vld3_v(&__ret, __p0, 1); \
11775 \
11776 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
11777 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
11778 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
11779 __ret; \
11780 })
11781 #endif
11782
11783 #ifdef __LITTLE_ENDIAN__
11784 #define vld3_dup_p8(__p0) __extension__ ({ \
11785 poly8x8x3_t __ret; \
11786 __builtin_neon_vld3_dup_v(&__ret, __p0, 4); \
11787 __ret; \
11788 })
11789 #else
11790 #define vld3_dup_p8(__p0) __extension__ ({ \
11791 poly8x8x3_t __ret; \
11792 __builtin_neon_vld3_dup_v(&__ret, __p0, 4); \
11793 \
11794 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
11795 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
11796 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
11797 __ret; \
11798 })
11799 #endif
11800
11801 #ifdef __LITTLE_ENDIAN__
11802 #define vld3_dup_p16(__p0) __extension__ ({ \
11803 poly16x4x3_t __ret; \
11804 __builtin_neon_vld3_dup_v(&__ret, __p0, 5); \
11805 __ret; \
11806 })
11807 #else
11808 #define vld3_dup_p16(__p0) __extension__ ({ \
11809 poly16x4x3_t __ret; \
11810 __builtin_neon_vld3_dup_v(&__ret, __p0, 5); \
11811 \
11812 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
11813 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
11814 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
11815 __ret; \
11816 })
11817 #endif
11818
11819 #ifdef __LITTLE_ENDIAN__
11820 #define vld3q_dup_p8(__p0) __extension__ ({ \
11821 poly8x16x3_t __ret; \
11822 __builtin_neon_vld3q_dup_v(&__ret, __p0, 36); \
11823 __ret; \
11824 })
11825 #else
11826 #define vld3q_dup_p8(__p0) __extension__ ({ \
11827 poly8x16x3_t __ret; \
11828 __builtin_neon_vld3q_dup_v(&__ret, __p0, 36); \
11829 \
11830 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
11831 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
11832 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
11833 __ret; \
11834 })
11835 #endif
11836
11837 #ifdef __LITTLE_ENDIAN__
11838 #define vld3q_dup_p16(__p0) __extension__ ({ \
11839 poly16x8x3_t __ret; \
11840 __builtin_neon_vld3q_dup_v(&__ret, __p0, 37); \
11841 __ret; \
11842 })
11843 #else
11844 #define vld3q_dup_p16(__p0) __extension__ ({ \
11845 poly16x8x3_t __ret; \
11846 __builtin_neon_vld3q_dup_v(&__ret, __p0, 37); \
11847 \
11848 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
11849 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
11850 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
11851 __ret; \
11852 })
11853 #endif
11854
11855 #ifdef __LITTLE_ENDIAN__
11856 #define vld3q_dup_u8(__p0) __extension__ ({ \
11857 uint8x16x3_t __ret; \
11858 __builtin_neon_vld3q_dup_v(&__ret, __p0, 48); \
11859 __ret; \
11860 })
11861 #else
11862 #define vld3q_dup_u8(__p0) __extension__ ({ \
11863 uint8x16x3_t __ret; \
11864 __builtin_neon_vld3q_dup_v(&__ret, __p0, 48); \
11865 \
11866 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
11867 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
11868 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
11869 __ret; \
11870 })
11871 #endif
11872
11873 #ifdef __LITTLE_ENDIAN__
11874 #define vld3q_dup_u32(__p0) __extension__ ({ \
11875 uint32x4x3_t __ret; \
11876 __builtin_neon_vld3q_dup_v(&__ret, __p0, 50); \
11877 __ret; \
11878 })
11879 #else
11880 #define vld3q_dup_u32(__p0) __extension__ ({ \
11881 uint32x4x3_t __ret; \
11882 __builtin_neon_vld3q_dup_v(&__ret, __p0, 50); \
11883 \
11884 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
11885 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
11886 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
11887 __ret; \
11888 })
11889 #endif
11890
11891 #ifdef __LITTLE_ENDIAN__
11892 #define vld3q_dup_u64(__p0) __extension__ ({ \
11893 uint64x2x3_t __ret; \
11894 __builtin_neon_vld3q_dup_v(&__ret, __p0, 51); \
11895 __ret; \
11896 })
11897 #else
11898 #define vld3q_dup_u64(__p0) __extension__ ({ \
11899 uint64x2x3_t __ret; \
11900 __builtin_neon_vld3q_dup_v(&__ret, __p0, 51); \
11901 \
11902 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
11903 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
11904 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
11905 __ret; \
11906 })
11907 #endif
11908
11909 #ifdef __LITTLE_ENDIAN__
11910 #define vld3q_dup_u16(__p0) __extension__ ({ \
11911 uint16x8x3_t __ret; \
11912 __builtin_neon_vld3q_dup_v(&__ret, __p0, 49); \
11913 __ret; \
11914 })
11915 #else
11916 #define vld3q_dup_u16(__p0) __extension__ ({ \
11917 uint16x8x3_t __ret; \
11918 __builtin_neon_vld3q_dup_v(&__ret, __p0, 49); \
11919 \
11920 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
11921 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
11922 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
11923 __ret; \
11924 })
11925 #endif
11926
11927 #ifdef __LITTLE_ENDIAN__
11928 #define vld3q_dup_s8(__p0) __extension__ ({ \
11929 int8x16x3_t __ret; \
11930 __builtin_neon_vld3q_dup_v(&__ret, __p0, 32); \
11931 __ret; \
11932 })
11933 #else
11934 #define vld3q_dup_s8(__p0) __extension__ ({ \
11935 int8x16x3_t __ret; \
11936 __builtin_neon_vld3q_dup_v(&__ret, __p0, 32); \
11937 \
11938 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
11939 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
11940 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
11941 __ret; \
11942 })
11943 #endif
11944
11945 #ifdef __LITTLE_ENDIAN__
11946 #define vld3q_dup_f32(__p0) __extension__ ({ \
11947 float32x4x3_t __ret; \
11948 __builtin_neon_vld3q_dup_v(&__ret, __p0, 41); \
11949 __ret; \
11950 })
11951 #else
11952 #define vld3q_dup_f32(__p0) __extension__ ({ \
11953 float32x4x3_t __ret; \
11954 __builtin_neon_vld3q_dup_v(&__ret, __p0, 41); \
11955 \
11956 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
11957 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
11958 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
11959 __ret; \
11960 })
11961 #endif
11962
11963 #ifdef __LITTLE_ENDIAN__
11964 #define vld3q_dup_s32(__p0) __extension__ ({ \
11965 int32x4x3_t __ret; \
11966 __builtin_neon_vld3q_dup_v(&__ret, __p0, 34); \
11967 __ret; \
11968 })
11969 #else
11970 #define vld3q_dup_s32(__p0) __extension__ ({ \
11971 int32x4x3_t __ret; \
11972 __builtin_neon_vld3q_dup_v(&__ret, __p0, 34); \
11973 \
11974 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
11975 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
11976 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
11977 __ret; \
11978 })
11979 #endif
11980
11981 #ifdef __LITTLE_ENDIAN__
11982 #define vld3q_dup_s64(__p0) __extension__ ({ \
11983 int64x2x3_t __ret; \
11984 __builtin_neon_vld3q_dup_v(&__ret, __p0, 35); \
11985 __ret; \
11986 })
11987 #else
11988 #define vld3q_dup_s64(__p0) __extension__ ({ \
11989 int64x2x3_t __ret; \
11990 __builtin_neon_vld3q_dup_v(&__ret, __p0, 35); \
11991 \
11992 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
11993 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
11994 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
11995 __ret; \
11996 })
11997 #endif
11998
11999 #ifdef __LITTLE_ENDIAN__
12000 #define vld3q_dup_s16(__p0) __extension__ ({ \
12001 int16x8x3_t __ret; \
12002 __builtin_neon_vld3q_dup_v(&__ret, __p0, 33); \
12003 __ret; \
12004 })
12005 #else
12006 #define vld3q_dup_s16(__p0) __extension__ ({ \
12007 int16x8x3_t __ret; \
12008 __builtin_neon_vld3q_dup_v(&__ret, __p0, 33); \
12009 \
12010 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
12011 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
12012 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
12013 __ret; \
12014 })
12015 #endif
12016
12017 #ifdef __LITTLE_ENDIAN__
12018 #define vld3_dup_u8(__p0) __extension__ ({ \
12019 uint8x8x3_t __ret; \
12020 __builtin_neon_vld3_dup_v(&__ret, __p0, 16); \
12021 __ret; \
12022 })
12023 #else
12024 #define vld3_dup_u8(__p0) __extension__ ({ \
12025 uint8x8x3_t __ret; \
12026 __builtin_neon_vld3_dup_v(&__ret, __p0, 16); \
12027 \
12028 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
12029 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
12030 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
12031 __ret; \
12032 })
12033 #endif
12034
12035 #ifdef __LITTLE_ENDIAN__
12036 #define vld3_dup_u32(__p0) __extension__ ({ \
12037 uint32x2x3_t __ret; \
12038 __builtin_neon_vld3_dup_v(&__ret, __p0, 18); \
12039 __ret; \
12040 })
12041 #else
12042 #define vld3_dup_u32(__p0) __extension__ ({ \
12043 uint32x2x3_t __ret; \
12044 __builtin_neon_vld3_dup_v(&__ret, __p0, 18); \
12045 \
12046 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
12047 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
12048 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
12049 __ret; \
12050 })
12051 #endif
12052
12053 #define vld3_dup_u64(__p0) __extension__ ({ \
12054 uint64x1x3_t __ret; \
12055 __builtin_neon_vld3_dup_v(&__ret, __p0, 19); \
12056 __ret; \
12057 })
12058 #ifdef __LITTLE_ENDIAN__
12059 #define vld3_dup_u16(__p0) __extension__ ({ \
12060 uint16x4x3_t __ret; \
12061 __builtin_neon_vld3_dup_v(&__ret, __p0, 17); \
12062 __ret; \
12063 })
12064 #else
12065 #define vld3_dup_u16(__p0) __extension__ ({ \
12066 uint16x4x3_t __ret; \
12067 __builtin_neon_vld3_dup_v(&__ret, __p0, 17); \
12068 \
12069 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
12070 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
12071 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
12072 __ret; \
12073 })
12074 #endif
12075
12076 #ifdef __LITTLE_ENDIAN__
12077 #define vld3_dup_s8(__p0) __extension__ ({ \
12078 int8x8x3_t __ret; \
12079 __builtin_neon_vld3_dup_v(&__ret, __p0, 0); \
12080 __ret; \
12081 })
12082 #else
12083 #define vld3_dup_s8(__p0) __extension__ ({ \
12084 int8x8x3_t __ret; \
12085 __builtin_neon_vld3_dup_v(&__ret, __p0, 0); \
12086 \
12087 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
12088 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
12089 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
12090 __ret; \
12091 })
12092 #endif
12093
12094 #ifdef __LITTLE_ENDIAN__
12095 #define vld3_dup_f32(__p0) __extension__ ({ \
12096 float32x2x3_t __ret; \
12097 __builtin_neon_vld3_dup_v(&__ret, __p0, 9); \
12098 __ret; \
12099 })
12100 #else
12101 #define vld3_dup_f32(__p0) __extension__ ({ \
12102 float32x2x3_t __ret; \
12103 __builtin_neon_vld3_dup_v(&__ret, __p0, 9); \
12104 \
12105 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
12106 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
12107 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
12108 __ret; \
12109 })
12110 #endif
12111
12112 #ifdef __LITTLE_ENDIAN__
12113 #define vld3_dup_s32(__p0) __extension__ ({ \
12114 int32x2x3_t __ret; \
12115 __builtin_neon_vld3_dup_v(&__ret, __p0, 2); \
12116 __ret; \
12117 })
12118 #else
12119 #define vld3_dup_s32(__p0) __extension__ ({ \
12120 int32x2x3_t __ret; \
12121 __builtin_neon_vld3_dup_v(&__ret, __p0, 2); \
12122 \
12123 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
12124 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
12125 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
12126 __ret; \
12127 })
12128 #endif
12129
12130 #define vld3_dup_s64(__p0) __extension__ ({ \
12131 int64x1x3_t __ret; \
12132 __builtin_neon_vld3_dup_v(&__ret, __p0, 3); \
12133 __ret; \
12134 })
12135 #ifdef __LITTLE_ENDIAN__
12136 #define vld3_dup_s16(__p0) __extension__ ({ \
12137 int16x4x3_t __ret; \
12138 __builtin_neon_vld3_dup_v(&__ret, __p0, 1); \
12139 __ret; \
12140 })
12141 #else
12142 #define vld3_dup_s16(__p0) __extension__ ({ \
12143 int16x4x3_t __ret; \
12144 __builtin_neon_vld3_dup_v(&__ret, __p0, 1); \
12145 \
12146 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
12147 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
12148 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
12149 __ret; \
12150 })
12151 #endif
12152
12153 #ifdef __LITTLE_ENDIAN__
12154 #define vld3_lane_p8(__p0, __p1, __p2) __extension__ ({ \
12155 poly8x8x3_t __ret; \
12156 poly8x8x3_t __s1 = __p1; \
12157 __builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 4); \
12158 __ret; \
12159 })
12160 #else
12161 #define vld3_lane_p8(__p0, __p1, __p2) __extension__ ({ \
12162 poly8x8x3_t __ret; \
12163 poly8x8x3_t __s1 = __p1; \
12164 poly8x8x3_t __rev1; \
12165 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
12166 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
12167 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
12168 __builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 4); \
12169 \
12170 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
12171 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
12172 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
12173 __ret; \
12174 })
12175 #endif
12176
12177 #ifdef __LITTLE_ENDIAN__
12178 #define vld3_lane_p16(__p0, __p1, __p2) __extension__ ({ \
12179 poly16x4x3_t __ret; \
12180 poly16x4x3_t __s1 = __p1; \
12181 __builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 5); \
12182 __ret; \
12183 })
12184 #else
12185 #define vld3_lane_p16(__p0, __p1, __p2) __extension__ ({ \
12186 poly16x4x3_t __ret; \
12187 poly16x4x3_t __s1 = __p1; \
12188 poly16x4x3_t __rev1; \
12189 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
12190 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
12191 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
12192 __builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 5); \
12193 \
12194 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
12195 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
12196 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
12197 __ret; \
12198 })
12199 #endif
12200
12201 #ifdef __LITTLE_ENDIAN__
12202 #define vld3q_lane_p16(__p0, __p1, __p2) __extension__ ({ \
12203 poly16x8x3_t __ret; \
12204 poly16x8x3_t __s1 = __p1; \
12205 __builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 37); \
12206 __ret; \
12207 })
12208 #else
12209 #define vld3q_lane_p16(__p0, __p1, __p2) __extension__ ({ \
12210 poly16x8x3_t __ret; \
12211 poly16x8x3_t __s1 = __p1; \
12212 poly16x8x3_t __rev1; \
12213 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
12214 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
12215 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
12216 __builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 37); \
12217 \
12218 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
12219 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
12220 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
12221 __ret; \
12222 })
12223 #endif
12224
12225 #ifdef __LITTLE_ENDIAN__
12226 #define vld3q_lane_u32(__p0, __p1, __p2) __extension__ ({ \
12227 uint32x4x3_t __ret; \
12228 uint32x4x3_t __s1 = __p1; \
12229 __builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 50); \
12230 __ret; \
12231 })
12232 #else
12233 #define vld3q_lane_u32(__p0, __p1, __p2) __extension__ ({ \
12234 uint32x4x3_t __ret; \
12235 uint32x4x3_t __s1 = __p1; \
12236 uint32x4x3_t __rev1; \
12237 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
12238 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
12239 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
12240 __builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 50); \
12241 \
12242 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
12243 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
12244 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
12245 __ret; \
12246 })
12247 #endif
12248
12249 #ifdef __LITTLE_ENDIAN__
12250 #define vld3q_lane_u16(__p0, __p1, __p2) __extension__ ({ \
12251 uint16x8x3_t __ret; \
12252 uint16x8x3_t __s1 = __p1; \
12253 __builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 49); \
12254 __ret; \
12255 })
12256 #else
12257 #define vld3q_lane_u16(__p0, __p1, __p2) __extension__ ({ \
12258 uint16x8x3_t __ret; \
12259 uint16x8x3_t __s1 = __p1; \
12260 uint16x8x3_t __rev1; \
12261 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
12262 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
12263 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
12264 __builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 49); \
12265 \
12266 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
12267 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
12268 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
12269 __ret; \
12270 })
12271 #endif
12272
12273 #ifdef __LITTLE_ENDIAN__
12274 #define vld3q_lane_f32(__p0, __p1, __p2) __extension__ ({ \
12275 float32x4x3_t __ret; \
12276 float32x4x3_t __s1 = __p1; \
12277 __builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 41); \
12278 __ret; \
12279 })
12280 #else
12281 #define vld3q_lane_f32(__p0, __p1, __p2) __extension__ ({ \
12282 float32x4x3_t __ret; \
12283 float32x4x3_t __s1 = __p1; \
12284 float32x4x3_t __rev1; \
12285 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
12286 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
12287 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
12288 __builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 41); \
12289 \
12290 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
12291 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
12292 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
12293 __ret; \
12294 })
12295 #endif
12296
12297 #ifdef __LITTLE_ENDIAN__
12298 #define vld3q_lane_s32(__p0, __p1, __p2) __extension__ ({ \
12299 int32x4x3_t __ret; \
12300 int32x4x3_t __s1 = __p1; \
12301 __builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 34); \
12302 __ret; \
12303 })
12304 #else
12305 #define vld3q_lane_s32(__p0, __p1, __p2) __extension__ ({ \
12306 int32x4x3_t __ret; \
12307 int32x4x3_t __s1 = __p1; \
12308 int32x4x3_t __rev1; \
12309 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
12310 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
12311 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
12312 __builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 34); \
12313 \
12314 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
12315 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
12316 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
12317 __ret; \
12318 })
12319 #endif
12320
12321 #ifdef __LITTLE_ENDIAN__
12322 #define vld3q_lane_s16(__p0, __p1, __p2) __extension__ ({ \
12323 int16x8x3_t __ret; \
12324 int16x8x3_t __s1 = __p1; \
12325 __builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 33); \
12326 __ret; \
12327 })
12328 #else
12329 #define vld3q_lane_s16(__p0, __p1, __p2) __extension__ ({ \
12330 int16x8x3_t __ret; \
12331 int16x8x3_t __s1 = __p1; \
12332 int16x8x3_t __rev1; \
12333 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
12334 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
12335 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
12336 __builtin_neon_vld3q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 33); \
12337 \
12338 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
12339 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
12340 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
12341 __ret; \
12342 })
12343 #endif
12344
12345 #ifdef __LITTLE_ENDIAN__
12346 #define vld3_lane_u8(__p0, __p1, __p2) __extension__ ({ \
12347 uint8x8x3_t __ret; \
12348 uint8x8x3_t __s1 = __p1; \
12349 __builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 16); \
12350 __ret; \
12351 })
12352 #else
12353 #define vld3_lane_u8(__p0, __p1, __p2) __extension__ ({ \
12354 uint8x8x3_t __ret; \
12355 uint8x8x3_t __s1 = __p1; \
12356 uint8x8x3_t __rev1; \
12357 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
12358 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
12359 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
12360 __builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 16); \
12361 \
12362 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
12363 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
12364 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
12365 __ret; \
12366 })
12367 #endif
12368
12369 #ifdef __LITTLE_ENDIAN__
12370 #define vld3_lane_u32(__p0, __p1, __p2) __extension__ ({ \
12371 uint32x2x3_t __ret; \
12372 uint32x2x3_t __s1 = __p1; \
12373 __builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 18); \
12374 __ret; \
12375 })
12376 #else
12377 #define vld3_lane_u32(__p0, __p1, __p2) __extension__ ({ \
12378 uint32x2x3_t __ret; \
12379 uint32x2x3_t __s1 = __p1; \
12380 uint32x2x3_t __rev1; \
12381 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
12382 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
12383 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
12384 __builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 18); \
12385 \
12386 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
12387 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
12388 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
12389 __ret; \
12390 })
12391 #endif
12392
12393 #ifdef __LITTLE_ENDIAN__
12394 #define vld3_lane_u16(__p0, __p1, __p2) __extension__ ({ \
12395 uint16x4x3_t __ret; \
12396 uint16x4x3_t __s1 = __p1; \
12397 __builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 17); \
12398 __ret; \
12399 })
12400 #else
12401 #define vld3_lane_u16(__p0, __p1, __p2) __extension__ ({ \
12402 uint16x4x3_t __ret; \
12403 uint16x4x3_t __s1 = __p1; \
12404 uint16x4x3_t __rev1; \
12405 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
12406 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
12407 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
12408 __builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 17); \
12409 \
12410 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
12411 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
12412 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
12413 __ret; \
12414 })
12415 #endif
12416
12417 #ifdef __LITTLE_ENDIAN__
12418 #define vld3_lane_s8(__p0, __p1, __p2) __extension__ ({ \
12419 int8x8x3_t __ret; \
12420 int8x8x3_t __s1 = __p1; \
12421 __builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 0); \
12422 __ret; \
12423 })
12424 #else
12425 #define vld3_lane_s8(__p0, __p1, __p2) __extension__ ({ \
12426 int8x8x3_t __ret; \
12427 int8x8x3_t __s1 = __p1; \
12428 int8x8x3_t __rev1; \
12429 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
12430 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
12431 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
12432 __builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 0); \
12433 \
12434 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
12435 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
12436 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
12437 __ret; \
12438 })
12439 #endif
12440
12441 #ifdef __LITTLE_ENDIAN__
12442 #define vld3_lane_f32(__p0, __p1, __p2) __extension__ ({ \
12443 float32x2x3_t __ret; \
12444 float32x2x3_t __s1 = __p1; \
12445 __builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 9); \
12446 __ret; \
12447 })
12448 #else
12449 #define vld3_lane_f32(__p0, __p1, __p2) __extension__ ({ \
12450 float32x2x3_t __ret; \
12451 float32x2x3_t __s1 = __p1; \
12452 float32x2x3_t __rev1; \
12453 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
12454 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
12455 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
12456 __builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 9); \
12457 \
12458 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
12459 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
12460 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
12461 __ret; \
12462 })
12463 #endif
12464
12465 #ifdef __LITTLE_ENDIAN__
12466 #define vld3_lane_s32(__p0, __p1, __p2) __extension__ ({ \
12467 int32x2x3_t __ret; \
12468 int32x2x3_t __s1 = __p1; \
12469 __builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 2); \
12470 __ret; \
12471 })
12472 #else
12473 #define vld3_lane_s32(__p0, __p1, __p2) __extension__ ({ \
12474 int32x2x3_t __ret; \
12475 int32x2x3_t __s1 = __p1; \
12476 int32x2x3_t __rev1; \
12477 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
12478 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
12479 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
12480 __builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 2); \
12481 \
12482 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
12483 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
12484 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
12485 __ret; \
12486 })
12487 #endif
12488
12489 #ifdef __LITTLE_ENDIAN__
12490 #define vld3_lane_s16(__p0, __p1, __p2) __extension__ ({ \
12491 int16x4x3_t __ret; \
12492 int16x4x3_t __s1 = __p1; \
12493 __builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 1); \
12494 __ret; \
12495 })
12496 #else
12497 #define vld3_lane_s16(__p0, __p1, __p2) __extension__ ({ \
12498 int16x4x3_t __ret; \
12499 int16x4x3_t __s1 = __p1; \
12500 int16x4x3_t __rev1; \
12501 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
12502 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
12503 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
12504 __builtin_neon_vld3_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 1); \
12505 \
12506 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
12507 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
12508 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
12509 __ret; \
12510 })
12511 #endif
12512
12513 #ifdef __LITTLE_ENDIAN__
12514 #define vld4_p8(__p0) __extension__ ({ \
12515 poly8x8x4_t __ret; \
12516 __builtin_neon_vld4_v(&__ret, __p0, 4); \
12517 __ret; \
12518 })
12519 #else
12520 #define vld4_p8(__p0) __extension__ ({ \
12521 poly8x8x4_t __ret; \
12522 __builtin_neon_vld4_v(&__ret, __p0, 4); \
12523 \
12524 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
12525 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
12526 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
12527 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
12528 __ret; \
12529 })
12530 #endif
12531
12532 #ifdef __LITTLE_ENDIAN__
12533 #define vld4_p16(__p0) __extension__ ({ \
12534 poly16x4x4_t __ret; \
12535 __builtin_neon_vld4_v(&__ret, __p0, 5); \
12536 __ret; \
12537 })
12538 #else
12539 #define vld4_p16(__p0) __extension__ ({ \
12540 poly16x4x4_t __ret; \
12541 __builtin_neon_vld4_v(&__ret, __p0, 5); \
12542 \
12543 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
12544 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
12545 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
12546 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
12547 __ret; \
12548 })
12549 #endif
12550
12551 #ifdef __LITTLE_ENDIAN__
12552 #define vld4q_p8(__p0) __extension__ ({ \
12553 poly8x16x4_t __ret; \
12554 __builtin_neon_vld4q_v(&__ret, __p0, 36); \
12555 __ret; \
12556 })
12557 #else
12558 #define vld4q_p8(__p0) __extension__ ({ \
12559 poly8x16x4_t __ret; \
12560 __builtin_neon_vld4q_v(&__ret, __p0, 36); \
12561 \
12562 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
12563 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
12564 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
12565 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
12566 __ret; \
12567 })
12568 #endif
12569
12570 #ifdef __LITTLE_ENDIAN__
12571 #define vld4q_p16(__p0) __extension__ ({ \
12572 poly16x8x4_t __ret; \
12573 __builtin_neon_vld4q_v(&__ret, __p0, 37); \
12574 __ret; \
12575 })
12576 #else
12577 #define vld4q_p16(__p0) __extension__ ({ \
12578 poly16x8x4_t __ret; \
12579 __builtin_neon_vld4q_v(&__ret, __p0, 37); \
12580 \
12581 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
12582 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
12583 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
12584 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
12585 __ret; \
12586 })
12587 #endif
12588
12589 #ifdef __LITTLE_ENDIAN__
12590 #define vld4q_u8(__p0) __extension__ ({ \
12591 uint8x16x4_t __ret; \
12592 __builtin_neon_vld4q_v(&__ret, __p0, 48); \
12593 __ret; \
12594 })
12595 #else
12596 #define vld4q_u8(__p0) __extension__ ({ \
12597 uint8x16x4_t __ret; \
12598 __builtin_neon_vld4q_v(&__ret, __p0, 48); \
12599 \
12600 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
12601 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
12602 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
12603 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
12604 __ret; \
12605 })
12606 #endif
12607
12608 #ifdef __LITTLE_ENDIAN__
12609 #define vld4q_u32(__p0) __extension__ ({ \
12610 uint32x4x4_t __ret; \
12611 __builtin_neon_vld4q_v(&__ret, __p0, 50); \
12612 __ret; \
12613 })
12614 #else
12615 #define vld4q_u32(__p0) __extension__ ({ \
12616 uint32x4x4_t __ret; \
12617 __builtin_neon_vld4q_v(&__ret, __p0, 50); \
12618 \
12619 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
12620 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
12621 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
12622 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
12623 __ret; \
12624 })
12625 #endif
12626
12627 #ifdef __LITTLE_ENDIAN__
12628 #define vld4q_u16(__p0) __extension__ ({ \
12629 uint16x8x4_t __ret; \
12630 __builtin_neon_vld4q_v(&__ret, __p0, 49); \
12631 __ret; \
12632 })
12633 #else
12634 #define vld4q_u16(__p0) __extension__ ({ \
12635 uint16x8x4_t __ret; \
12636 __builtin_neon_vld4q_v(&__ret, __p0, 49); \
12637 \
12638 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
12639 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
12640 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
12641 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
12642 __ret; \
12643 })
12644 #endif
12645
12646 #ifdef __LITTLE_ENDIAN__
12647 #define vld4q_s8(__p0) __extension__ ({ \
12648 int8x16x4_t __ret; \
12649 __builtin_neon_vld4q_v(&__ret, __p0, 32); \
12650 __ret; \
12651 })
12652 #else
12653 #define vld4q_s8(__p0) __extension__ ({ \
12654 int8x16x4_t __ret; \
12655 __builtin_neon_vld4q_v(&__ret, __p0, 32); \
12656 \
12657 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
12658 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
12659 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
12660 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
12661 __ret; \
12662 })
12663 #endif
12664
12665 #ifdef __LITTLE_ENDIAN__
12666 #define vld4q_f32(__p0) __extension__ ({ \
12667 float32x4x4_t __ret; \
12668 __builtin_neon_vld4q_v(&__ret, __p0, 41); \
12669 __ret; \
12670 })
12671 #else
12672 #define vld4q_f32(__p0) __extension__ ({ \
12673 float32x4x4_t __ret; \
12674 __builtin_neon_vld4q_v(&__ret, __p0, 41); \
12675 \
12676 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
12677 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
12678 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
12679 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
12680 __ret; \
12681 })
12682 #endif
12683
12684 #ifdef __LITTLE_ENDIAN__
12685 #define vld4q_s32(__p0) __extension__ ({ \
12686 int32x4x4_t __ret; \
12687 __builtin_neon_vld4q_v(&__ret, __p0, 34); \
12688 __ret; \
12689 })
12690 #else
12691 #define vld4q_s32(__p0) __extension__ ({ \
12692 int32x4x4_t __ret; \
12693 __builtin_neon_vld4q_v(&__ret, __p0, 34); \
12694 \
12695 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
12696 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
12697 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
12698 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
12699 __ret; \
12700 })
12701 #endif
12702
12703 #ifdef __LITTLE_ENDIAN__
12704 #define vld4q_s16(__p0) __extension__ ({ \
12705 int16x8x4_t __ret; \
12706 __builtin_neon_vld4q_v(&__ret, __p0, 33); \
12707 __ret; \
12708 })
12709 #else
12710 #define vld4q_s16(__p0) __extension__ ({ \
12711 int16x8x4_t __ret; \
12712 __builtin_neon_vld4q_v(&__ret, __p0, 33); \
12713 \
12714 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
12715 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
12716 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
12717 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
12718 __ret; \
12719 })
12720 #endif
12721
12722 #ifdef __LITTLE_ENDIAN__
12723 #define vld4_u8(__p0) __extension__ ({ \
12724 uint8x8x4_t __ret; \
12725 __builtin_neon_vld4_v(&__ret, __p0, 16); \
12726 __ret; \
12727 })
12728 #else
12729 #define vld4_u8(__p0) __extension__ ({ \
12730 uint8x8x4_t __ret; \
12731 __builtin_neon_vld4_v(&__ret, __p0, 16); \
12732 \
12733 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
12734 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
12735 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
12736 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
12737 __ret; \
12738 })
12739 #endif
12740
12741 #ifdef __LITTLE_ENDIAN__
12742 #define vld4_u32(__p0) __extension__ ({ \
12743 uint32x2x4_t __ret; \
12744 __builtin_neon_vld4_v(&__ret, __p0, 18); \
12745 __ret; \
12746 })
12747 #else
12748 #define vld4_u32(__p0) __extension__ ({ \
12749 uint32x2x4_t __ret; \
12750 __builtin_neon_vld4_v(&__ret, __p0, 18); \
12751 \
12752 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
12753 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
12754 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
12755 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
12756 __ret; \
12757 })
12758 #endif
12759
12760 #define vld4_u64(__p0) __extension__ ({ \
12761 uint64x1x4_t __ret; \
12762 __builtin_neon_vld4_v(&__ret, __p0, 19); \
12763 __ret; \
12764 })
12765 #ifdef __LITTLE_ENDIAN__
12766 #define vld4_u16(__p0) __extension__ ({ \
12767 uint16x4x4_t __ret; \
12768 __builtin_neon_vld4_v(&__ret, __p0, 17); \
12769 __ret; \
12770 })
12771 #else
12772 #define vld4_u16(__p0) __extension__ ({ \
12773 uint16x4x4_t __ret; \
12774 __builtin_neon_vld4_v(&__ret, __p0, 17); \
12775 \
12776 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
12777 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
12778 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
12779 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
12780 __ret; \
12781 })
12782 #endif
12783
12784 #ifdef __LITTLE_ENDIAN__
12785 #define vld4_s8(__p0) __extension__ ({ \
12786 int8x8x4_t __ret; \
12787 __builtin_neon_vld4_v(&__ret, __p0, 0); \
12788 __ret; \
12789 })
12790 #else
12791 #define vld4_s8(__p0) __extension__ ({ \
12792 int8x8x4_t __ret; \
12793 __builtin_neon_vld4_v(&__ret, __p0, 0); \
12794 \
12795 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
12796 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
12797 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
12798 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
12799 __ret; \
12800 })
12801 #endif
12802
12803 #ifdef __LITTLE_ENDIAN__
12804 #define vld4_f32(__p0) __extension__ ({ \
12805 float32x2x4_t __ret; \
12806 __builtin_neon_vld4_v(&__ret, __p0, 9); \
12807 __ret; \
12808 })
12809 #else
12810 #define vld4_f32(__p0) __extension__ ({ \
12811 float32x2x4_t __ret; \
12812 __builtin_neon_vld4_v(&__ret, __p0, 9); \
12813 \
12814 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
12815 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
12816 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
12817 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
12818 __ret; \
12819 })
12820 #endif
12821
12822 #ifdef __LITTLE_ENDIAN__
12823 #define vld4_s32(__p0) __extension__ ({ \
12824 int32x2x4_t __ret; \
12825 __builtin_neon_vld4_v(&__ret, __p0, 2); \
12826 __ret; \
12827 })
12828 #else
12829 #define vld4_s32(__p0) __extension__ ({ \
12830 int32x2x4_t __ret; \
12831 __builtin_neon_vld4_v(&__ret, __p0, 2); \
12832 \
12833 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
12834 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
12835 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
12836 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
12837 __ret; \
12838 })
12839 #endif
12840
12841 #define vld4_s64(__p0) __extension__ ({ \
12842 int64x1x4_t __ret; \
12843 __builtin_neon_vld4_v(&__ret, __p0, 3); \
12844 __ret; \
12845 })
12846 #ifdef __LITTLE_ENDIAN__
12847 #define vld4_s16(__p0) __extension__ ({ \
12848 int16x4x4_t __ret; \
12849 __builtin_neon_vld4_v(&__ret, __p0, 1); \
12850 __ret; \
12851 })
12852 #else
12853 #define vld4_s16(__p0) __extension__ ({ \
12854 int16x4x4_t __ret; \
12855 __builtin_neon_vld4_v(&__ret, __p0, 1); \
12856 \
12857 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
12858 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
12859 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
12860 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
12861 __ret; \
12862 })
12863 #endif
12864
12865 #ifdef __LITTLE_ENDIAN__
12866 #define vld4_dup_p8(__p0) __extension__ ({ \
12867 poly8x8x4_t __ret; \
12868 __builtin_neon_vld4_dup_v(&__ret, __p0, 4); \
12869 __ret; \
12870 })
12871 #else
12872 #define vld4_dup_p8(__p0) __extension__ ({ \
12873 poly8x8x4_t __ret; \
12874 __builtin_neon_vld4_dup_v(&__ret, __p0, 4); \
12875 \
12876 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
12877 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
12878 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
12879 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
12880 __ret; \
12881 })
12882 #endif
12883
12884 #ifdef __LITTLE_ENDIAN__
12885 #define vld4_dup_p16(__p0) __extension__ ({ \
12886 poly16x4x4_t __ret; \
12887 __builtin_neon_vld4_dup_v(&__ret, __p0, 5); \
12888 __ret; \
12889 })
12890 #else
12891 #define vld4_dup_p16(__p0) __extension__ ({ \
12892 poly16x4x4_t __ret; \
12893 __builtin_neon_vld4_dup_v(&__ret, __p0, 5); \
12894 \
12895 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
12896 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
12897 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
12898 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
12899 __ret; \
12900 })
12901 #endif
12902
12903 #ifdef __LITTLE_ENDIAN__
12904 #define vld4q_dup_p8(__p0) __extension__ ({ \
12905 poly8x16x4_t __ret; \
12906 __builtin_neon_vld4q_dup_v(&__ret, __p0, 36); \
12907 __ret; \
12908 })
12909 #else
12910 #define vld4q_dup_p8(__p0) __extension__ ({ \
12911 poly8x16x4_t __ret; \
12912 __builtin_neon_vld4q_dup_v(&__ret, __p0, 36); \
12913 \
12914 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
12915 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
12916 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
12917 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
12918 __ret; \
12919 })
12920 #endif
12921
12922 #ifdef __LITTLE_ENDIAN__
12923 #define vld4q_dup_p16(__p0) __extension__ ({ \
12924 poly16x8x4_t __ret; \
12925 __builtin_neon_vld4q_dup_v(&__ret, __p0, 37); \
12926 __ret; \
12927 })
12928 #else
12929 #define vld4q_dup_p16(__p0) __extension__ ({ \
12930 poly16x8x4_t __ret; \
12931 __builtin_neon_vld4q_dup_v(&__ret, __p0, 37); \
12932 \
12933 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
12934 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
12935 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
12936 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
12937 __ret; \
12938 })
12939 #endif
12940
12941 #ifdef __LITTLE_ENDIAN__
12942 #define vld4q_dup_u8(__p0) __extension__ ({ \
12943 uint8x16x4_t __ret; \
12944 __builtin_neon_vld4q_dup_v(&__ret, __p0, 48); \
12945 __ret; \
12946 })
12947 #else
12948 #define vld4q_dup_u8(__p0) __extension__ ({ \
12949 uint8x16x4_t __ret; \
12950 __builtin_neon_vld4q_dup_v(&__ret, __p0, 48); \
12951 \
12952 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
12953 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
12954 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
12955 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
12956 __ret; \
12957 })
12958 #endif
12959
12960 #ifdef __LITTLE_ENDIAN__
12961 #define vld4q_dup_u32(__p0) __extension__ ({ \
12962 uint32x4x4_t __ret; \
12963 __builtin_neon_vld4q_dup_v(&__ret, __p0, 50); \
12964 __ret; \
12965 })
12966 #else
12967 #define vld4q_dup_u32(__p0) __extension__ ({ \
12968 uint32x4x4_t __ret; \
12969 __builtin_neon_vld4q_dup_v(&__ret, __p0, 50); \
12970 \
12971 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
12972 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
12973 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
12974 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
12975 __ret; \
12976 })
12977 #endif
12978
12979 #ifdef __LITTLE_ENDIAN__
12980 #define vld4q_dup_u64(__p0) __extension__ ({ \
12981 uint64x2x4_t __ret; \
12982 __builtin_neon_vld4q_dup_v(&__ret, __p0, 51); \
12983 __ret; \
12984 })
12985 #else
12986 #define vld4q_dup_u64(__p0) __extension__ ({ \
12987 uint64x2x4_t __ret; \
12988 __builtin_neon_vld4q_dup_v(&__ret, __p0, 51); \
12989 \
12990 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
12991 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
12992 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
12993 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
12994 __ret; \
12995 })
12996 #endif
12997
12998 #ifdef __LITTLE_ENDIAN__
12999 #define vld4q_dup_u16(__p0) __extension__ ({ \
13000 uint16x8x4_t __ret; \
13001 __builtin_neon_vld4q_dup_v(&__ret, __p0, 49); \
13002 __ret; \
13003 })
13004 #else
13005 #define vld4q_dup_u16(__p0) __extension__ ({ \
13006 uint16x8x4_t __ret; \
13007 __builtin_neon_vld4q_dup_v(&__ret, __p0, 49); \
13008 \
13009 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
13010 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
13011 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
13012 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
13013 __ret; \
13014 })
13015 #endif
13016
13017 #ifdef __LITTLE_ENDIAN__
13018 #define vld4q_dup_s8(__p0) __extension__ ({ \
13019 int8x16x4_t __ret; \
13020 __builtin_neon_vld4q_dup_v(&__ret, __p0, 32); \
13021 __ret; \
13022 })
13023 #else
13024 #define vld4q_dup_s8(__p0) __extension__ ({ \
13025 int8x16x4_t __ret; \
13026 __builtin_neon_vld4q_dup_v(&__ret, __p0, 32); \
13027 \
13028 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
13029 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
13030 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
13031 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
13032 __ret; \
13033 })
13034 #endif
13035
13036 #ifdef __LITTLE_ENDIAN__
13037 #define vld4q_dup_f32(__p0) __extension__ ({ \
13038 float32x4x4_t __ret; \
13039 __builtin_neon_vld4q_dup_v(&__ret, __p0, 41); \
13040 __ret; \
13041 })
13042 #else
13043 #define vld4q_dup_f32(__p0) __extension__ ({ \
13044 float32x4x4_t __ret; \
13045 __builtin_neon_vld4q_dup_v(&__ret, __p0, 41); \
13046 \
13047 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
13048 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
13049 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
13050 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
13051 __ret; \
13052 })
13053 #endif
13054
13055 #ifdef __LITTLE_ENDIAN__
13056 #define vld4q_dup_s32(__p0) __extension__ ({ \
13057 int32x4x4_t __ret; \
13058 __builtin_neon_vld4q_dup_v(&__ret, __p0, 34); \
13059 __ret; \
13060 })
13061 #else
13062 #define vld4q_dup_s32(__p0) __extension__ ({ \
13063 int32x4x4_t __ret; \
13064 __builtin_neon_vld4q_dup_v(&__ret, __p0, 34); \
13065 \
13066 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
13067 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
13068 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
13069 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
13070 __ret; \
13071 })
13072 #endif
13073
13074 #ifdef __LITTLE_ENDIAN__
13075 #define vld4q_dup_s64(__p0) __extension__ ({ \
13076 int64x2x4_t __ret; \
13077 __builtin_neon_vld4q_dup_v(&__ret, __p0, 35); \
13078 __ret; \
13079 })
13080 #else
13081 #define vld4q_dup_s64(__p0) __extension__ ({ \
13082 int64x2x4_t __ret; \
13083 __builtin_neon_vld4q_dup_v(&__ret, __p0, 35); \
13084 \
13085 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
13086 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
13087 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
13088 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
13089 __ret; \
13090 })
13091 #endif
13092
13093 #ifdef __LITTLE_ENDIAN__
13094 #define vld4q_dup_s16(__p0) __extension__ ({ \
13095 int16x8x4_t __ret; \
13096 __builtin_neon_vld4q_dup_v(&__ret, __p0, 33); \
13097 __ret; \
13098 })
13099 #else
13100 #define vld4q_dup_s16(__p0) __extension__ ({ \
13101 int16x8x4_t __ret; \
13102 __builtin_neon_vld4q_dup_v(&__ret, __p0, 33); \
13103 \
13104 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
13105 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
13106 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
13107 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
13108 __ret; \
13109 })
13110 #endif
13111
13112 #ifdef __LITTLE_ENDIAN__
13113 #define vld4_dup_u8(__p0) __extension__ ({ \
13114 uint8x8x4_t __ret; \
13115 __builtin_neon_vld4_dup_v(&__ret, __p0, 16); \
13116 __ret; \
13117 })
13118 #else
13119 #define vld4_dup_u8(__p0) __extension__ ({ \
13120 uint8x8x4_t __ret; \
13121 __builtin_neon_vld4_dup_v(&__ret, __p0, 16); \
13122 \
13123 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
13124 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
13125 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
13126 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
13127 __ret; \
13128 })
13129 #endif
13130
13131 #ifdef __LITTLE_ENDIAN__
13132 #define vld4_dup_u32(__p0) __extension__ ({ \
13133 uint32x2x4_t __ret; \
13134 __builtin_neon_vld4_dup_v(&__ret, __p0, 18); \
13135 __ret; \
13136 })
13137 #else
13138 #define vld4_dup_u32(__p0) __extension__ ({ \
13139 uint32x2x4_t __ret; \
13140 __builtin_neon_vld4_dup_v(&__ret, __p0, 18); \
13141 \
13142 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
13143 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
13144 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
13145 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
13146 __ret; \
13147 })
13148 #endif
13149
13150 #define vld4_dup_u64(__p0) __extension__ ({ \
13151 uint64x1x4_t __ret; \
13152 __builtin_neon_vld4_dup_v(&__ret, __p0, 19); \
13153 __ret; \
13154 })
13155 #ifdef __LITTLE_ENDIAN__
13156 #define vld4_dup_u16(__p0) __extension__ ({ \
13157 uint16x4x4_t __ret; \
13158 __builtin_neon_vld4_dup_v(&__ret, __p0, 17); \
13159 __ret; \
13160 })
13161 #else
13162 #define vld4_dup_u16(__p0) __extension__ ({ \
13163 uint16x4x4_t __ret; \
13164 __builtin_neon_vld4_dup_v(&__ret, __p0, 17); \
13165 \
13166 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
13167 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
13168 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
13169 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
13170 __ret; \
13171 })
13172 #endif
13173
13174 #ifdef __LITTLE_ENDIAN__
13175 #define vld4_dup_s8(__p0) __extension__ ({ \
13176 int8x8x4_t __ret; \
13177 __builtin_neon_vld4_dup_v(&__ret, __p0, 0); \
13178 __ret; \
13179 })
13180 #else
13181 #define vld4_dup_s8(__p0) __extension__ ({ \
13182 int8x8x4_t __ret; \
13183 __builtin_neon_vld4_dup_v(&__ret, __p0, 0); \
13184 \
13185 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
13186 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
13187 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
13188 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
13189 __ret; \
13190 })
13191 #endif
13192
13193 #ifdef __LITTLE_ENDIAN__
13194 #define vld4_dup_f32(__p0) __extension__ ({ \
13195 float32x2x4_t __ret; \
13196 __builtin_neon_vld4_dup_v(&__ret, __p0, 9); \
13197 __ret; \
13198 })
13199 #else
13200 #define vld4_dup_f32(__p0) __extension__ ({ \
13201 float32x2x4_t __ret; \
13202 __builtin_neon_vld4_dup_v(&__ret, __p0, 9); \
13203 \
13204 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
13205 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
13206 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
13207 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
13208 __ret; \
13209 })
13210 #endif
13211
13212 #ifdef __LITTLE_ENDIAN__
13213 #define vld4_dup_s32(__p0) __extension__ ({ \
13214 int32x2x4_t __ret; \
13215 __builtin_neon_vld4_dup_v(&__ret, __p0, 2); \
13216 __ret; \
13217 })
13218 #else
13219 #define vld4_dup_s32(__p0) __extension__ ({ \
13220 int32x2x4_t __ret; \
13221 __builtin_neon_vld4_dup_v(&__ret, __p0, 2); \
13222 \
13223 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
13224 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
13225 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
13226 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
13227 __ret; \
13228 })
13229 #endif
13230
13231 #define vld4_dup_s64(__p0) __extension__ ({ \
13232 int64x1x4_t __ret; \
13233 __builtin_neon_vld4_dup_v(&__ret, __p0, 3); \
13234 __ret; \
13235 })
13236 #ifdef __LITTLE_ENDIAN__
13237 #define vld4_dup_s16(__p0) __extension__ ({ \
13238 int16x4x4_t __ret; \
13239 __builtin_neon_vld4_dup_v(&__ret, __p0, 1); \
13240 __ret; \
13241 })
13242 #else
13243 #define vld4_dup_s16(__p0) __extension__ ({ \
13244 int16x4x4_t __ret; \
13245 __builtin_neon_vld4_dup_v(&__ret, __p0, 1); \
13246 \
13247 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
13248 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
13249 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
13250 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
13251 __ret; \
13252 })
13253 #endif
13254
13255 #ifdef __LITTLE_ENDIAN__
13256 #define vld4_lane_p8(__p0, __p1, __p2) __extension__ ({ \
13257 poly8x8x4_t __ret; \
13258 poly8x8x4_t __s1 = __p1; \
13259 __builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 4); \
13260 __ret; \
13261 })
13262 #else
13263 #define vld4_lane_p8(__p0, __p1, __p2) __extension__ ({ \
13264 poly8x8x4_t __ret; \
13265 poly8x8x4_t __s1 = __p1; \
13266 poly8x8x4_t __rev1; \
13267 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
13268 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
13269 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
13270 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
13271 __builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 4); \
13272 \
13273 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
13274 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
13275 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
13276 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
13277 __ret; \
13278 })
13279 #endif
13280
13281 #ifdef __LITTLE_ENDIAN__
13282 #define vld4_lane_p16(__p0, __p1, __p2) __extension__ ({ \
13283 poly16x4x4_t __ret; \
13284 poly16x4x4_t __s1 = __p1; \
13285 __builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 5); \
13286 __ret; \
13287 })
13288 #else
13289 #define vld4_lane_p16(__p0, __p1, __p2) __extension__ ({ \
13290 poly16x4x4_t __ret; \
13291 poly16x4x4_t __s1 = __p1; \
13292 poly16x4x4_t __rev1; \
13293 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
13294 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
13295 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
13296 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
13297 __builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 5); \
13298 \
13299 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
13300 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
13301 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
13302 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
13303 __ret; \
13304 })
13305 #endif
13306
13307 #ifdef __LITTLE_ENDIAN__
13308 #define vld4q_lane_p16(__p0, __p1, __p2) __extension__ ({ \
13309 poly16x8x4_t __ret; \
13310 poly16x8x4_t __s1 = __p1; \
13311 __builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 37); \
13312 __ret; \
13313 })
13314 #else
13315 #define vld4q_lane_p16(__p0, __p1, __p2) __extension__ ({ \
13316 poly16x8x4_t __ret; \
13317 poly16x8x4_t __s1 = __p1; \
13318 poly16x8x4_t __rev1; \
13319 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
13320 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
13321 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
13322 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
13323 __builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 37); \
13324 \
13325 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
13326 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
13327 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
13328 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
13329 __ret; \
13330 })
13331 #endif
13332
13333 #ifdef __LITTLE_ENDIAN__
13334 #define vld4q_lane_u32(__p0, __p1, __p2) __extension__ ({ \
13335 uint32x4x4_t __ret; \
13336 uint32x4x4_t __s1 = __p1; \
13337 __builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 50); \
13338 __ret; \
13339 })
13340 #else
13341 #define vld4q_lane_u32(__p0, __p1, __p2) __extension__ ({ \
13342 uint32x4x4_t __ret; \
13343 uint32x4x4_t __s1 = __p1; \
13344 uint32x4x4_t __rev1; \
13345 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
13346 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
13347 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
13348 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
13349 __builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 50); \
13350 \
13351 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
13352 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
13353 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
13354 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
13355 __ret; \
13356 })
13357 #endif
13358
13359 #ifdef __LITTLE_ENDIAN__
13360 #define vld4q_lane_u16(__p0, __p1, __p2) __extension__ ({ \
13361 uint16x8x4_t __ret; \
13362 uint16x8x4_t __s1 = __p1; \
13363 __builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 49); \
13364 __ret; \
13365 })
13366 #else
13367 #define vld4q_lane_u16(__p0, __p1, __p2) __extension__ ({ \
13368 uint16x8x4_t __ret; \
13369 uint16x8x4_t __s1 = __p1; \
13370 uint16x8x4_t __rev1; \
13371 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
13372 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
13373 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
13374 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
13375 __builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 49); \
13376 \
13377 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
13378 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
13379 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
13380 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
13381 __ret; \
13382 })
13383 #endif
13384
13385 #ifdef __LITTLE_ENDIAN__
13386 #define vld4q_lane_f32(__p0, __p1, __p2) __extension__ ({ \
13387 float32x4x4_t __ret; \
13388 float32x4x4_t __s1 = __p1; \
13389 __builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 41); \
13390 __ret; \
13391 })
13392 #else
13393 #define vld4q_lane_f32(__p0, __p1, __p2) __extension__ ({ \
13394 float32x4x4_t __ret; \
13395 float32x4x4_t __s1 = __p1; \
13396 float32x4x4_t __rev1; \
13397 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
13398 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
13399 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
13400 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
13401 __builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 41); \
13402 \
13403 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
13404 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
13405 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
13406 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
13407 __ret; \
13408 })
13409 #endif
13410
13411 #ifdef __LITTLE_ENDIAN__
13412 #define vld4q_lane_s32(__p0, __p1, __p2) __extension__ ({ \
13413 int32x4x4_t __ret; \
13414 int32x4x4_t __s1 = __p1; \
13415 __builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 34); \
13416 __ret; \
13417 })
13418 #else
13419 #define vld4q_lane_s32(__p0, __p1, __p2) __extension__ ({ \
13420 int32x4x4_t __ret; \
13421 int32x4x4_t __s1 = __p1; \
13422 int32x4x4_t __rev1; \
13423 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
13424 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
13425 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
13426 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
13427 __builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 34); \
13428 \
13429 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
13430 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
13431 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
13432 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
13433 __ret; \
13434 })
13435 #endif
13436
13437 #ifdef __LITTLE_ENDIAN__
13438 #define vld4q_lane_s16(__p0, __p1, __p2) __extension__ ({ \
13439 int16x8x4_t __ret; \
13440 int16x8x4_t __s1 = __p1; \
13441 __builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 33); \
13442 __ret; \
13443 })
13444 #else
13445 #define vld4q_lane_s16(__p0, __p1, __p2) __extension__ ({ \
13446 int16x8x4_t __ret; \
13447 int16x8x4_t __s1 = __p1; \
13448 int16x8x4_t __rev1; \
13449 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
13450 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
13451 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
13452 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
13453 __builtin_neon_vld4q_lane_v(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 33); \
13454 \
13455 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
13456 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
13457 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
13458 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
13459 __ret; \
13460 })
13461 #endif
13462
13463 #ifdef __LITTLE_ENDIAN__
13464 #define vld4_lane_u8(__p0, __p1, __p2) __extension__ ({ \
13465 uint8x8x4_t __ret; \
13466 uint8x8x4_t __s1 = __p1; \
13467 __builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 16); \
13468 __ret; \
13469 })
13470 #else
13471 #define vld4_lane_u8(__p0, __p1, __p2) __extension__ ({ \
13472 uint8x8x4_t __ret; \
13473 uint8x8x4_t __s1 = __p1; \
13474 uint8x8x4_t __rev1; \
13475 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
13476 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
13477 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
13478 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
13479 __builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 16); \
13480 \
13481 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
13482 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
13483 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
13484 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
13485 __ret; \
13486 })
13487 #endif
13488
13489 #ifdef __LITTLE_ENDIAN__
13490 #define vld4_lane_u32(__p0, __p1, __p2) __extension__ ({ \
13491 uint32x2x4_t __ret; \
13492 uint32x2x4_t __s1 = __p1; \
13493 __builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 18); \
13494 __ret; \
13495 })
13496 #else
13497 #define vld4_lane_u32(__p0, __p1, __p2) __extension__ ({ \
13498 uint32x2x4_t __ret; \
13499 uint32x2x4_t __s1 = __p1; \
13500 uint32x2x4_t __rev1; \
13501 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
13502 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
13503 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
13504 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
13505 __builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 18); \
13506 \
13507 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
13508 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
13509 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
13510 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
13511 __ret; \
13512 })
13513 #endif
13514
13515 #ifdef __LITTLE_ENDIAN__
13516 #define vld4_lane_u16(__p0, __p1, __p2) __extension__ ({ \
13517 uint16x4x4_t __ret; \
13518 uint16x4x4_t __s1 = __p1; \
13519 __builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 17); \
13520 __ret; \
13521 })
13522 #else
13523 #define vld4_lane_u16(__p0, __p1, __p2) __extension__ ({ \
13524 uint16x4x4_t __ret; \
13525 uint16x4x4_t __s1 = __p1; \
13526 uint16x4x4_t __rev1; \
13527 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
13528 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
13529 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
13530 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
13531 __builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 17); \
13532 \
13533 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
13534 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
13535 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
13536 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
13537 __ret; \
13538 })
13539 #endif
13540
13541 #ifdef __LITTLE_ENDIAN__
13542 #define vld4_lane_s8(__p0, __p1, __p2) __extension__ ({ \
13543 int8x8x4_t __ret; \
13544 int8x8x4_t __s1 = __p1; \
13545 __builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 0); \
13546 __ret; \
13547 })
13548 #else
13549 #define vld4_lane_s8(__p0, __p1, __p2) __extension__ ({ \
13550 int8x8x4_t __ret; \
13551 int8x8x4_t __s1 = __p1; \
13552 int8x8x4_t __rev1; \
13553 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
13554 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
13555 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
13556 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
13557 __builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 0); \
13558 \
13559 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
13560 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
13561 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
13562 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
13563 __ret; \
13564 })
13565 #endif
13566
13567 #ifdef __LITTLE_ENDIAN__
13568 #define vld4_lane_f32(__p0, __p1, __p2) __extension__ ({ \
13569 float32x2x4_t __ret; \
13570 float32x2x4_t __s1 = __p1; \
13571 __builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 9); \
13572 __ret; \
13573 })
13574 #else
13575 #define vld4_lane_f32(__p0, __p1, __p2) __extension__ ({ \
13576 float32x2x4_t __ret; \
13577 float32x2x4_t __s1 = __p1; \
13578 float32x2x4_t __rev1; \
13579 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
13580 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
13581 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
13582 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
13583 __builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 9); \
13584 \
13585 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
13586 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
13587 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
13588 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
13589 __ret; \
13590 })
13591 #endif
13592
13593 #ifdef __LITTLE_ENDIAN__
13594 #define vld4_lane_s32(__p0, __p1, __p2) __extension__ ({ \
13595 int32x2x4_t __ret; \
13596 int32x2x4_t __s1 = __p1; \
13597 __builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 2); \
13598 __ret; \
13599 })
13600 #else
13601 #define vld4_lane_s32(__p0, __p1, __p2) __extension__ ({ \
13602 int32x2x4_t __ret; \
13603 int32x2x4_t __s1 = __p1; \
13604 int32x2x4_t __rev1; \
13605 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
13606 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
13607 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
13608 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
13609 __builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 2); \
13610 \
13611 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0); \
13612 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0); \
13613 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 1, 0); \
13614 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 1, 0); \
13615 __ret; \
13616 })
13617 #endif
13618
13619 #ifdef __LITTLE_ENDIAN__
13620 #define vld4_lane_s16(__p0, __p1, __p2) __extension__ ({ \
13621 int16x4x4_t __ret; \
13622 int16x4x4_t __s1 = __p1; \
13623 __builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 1); \
13624 __ret; \
13625 })
13626 #else
13627 #define vld4_lane_s16(__p0, __p1, __p2) __extension__ ({ \
13628 int16x4x4_t __ret; \
13629 int16x4x4_t __s1 = __p1; \
13630 int16x4x4_t __rev1; \
13631 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
13632 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
13633 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
13634 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
13635 __builtin_neon_vld4_lane_v(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 1); \
13636 \
13637 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
13638 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
13639 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
13640 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
13641 __ret; \
13642 })
13643 #endif
13644
13645 #ifdef __LITTLE_ENDIAN__
vmaxq_u8(uint8x16_t __p0,uint8x16_t __p1)13646 __ai uint8x16_t vmaxq_u8(uint8x16_t __p0, uint8x16_t __p1) {
13647 uint8x16_t __ret;
13648 __ret = (uint8x16_t) __builtin_neon_vmaxq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
13649 return __ret;
13650 }
13651 #else
vmaxq_u8(uint8x16_t __p0,uint8x16_t __p1)13652 __ai uint8x16_t vmaxq_u8(uint8x16_t __p0, uint8x16_t __p1) {
13653 uint8x16_t __ret;
13654 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
13655 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
13656 __ret = (uint8x16_t) __builtin_neon_vmaxq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
13657 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
13658 return __ret;
13659 }
13660 #endif
13661
13662 #ifdef __LITTLE_ENDIAN__
vmaxq_u32(uint32x4_t __p0,uint32x4_t __p1)13663 __ai uint32x4_t vmaxq_u32(uint32x4_t __p0, uint32x4_t __p1) {
13664 uint32x4_t __ret;
13665 __ret = (uint32x4_t) __builtin_neon_vmaxq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
13666 return __ret;
13667 }
13668 #else
vmaxq_u32(uint32x4_t __p0,uint32x4_t __p1)13669 __ai uint32x4_t vmaxq_u32(uint32x4_t __p0, uint32x4_t __p1) {
13670 uint32x4_t __ret;
13671 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
13672 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
13673 __ret = (uint32x4_t) __builtin_neon_vmaxq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
13674 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
13675 return __ret;
13676 }
13677 #endif
13678
13679 #ifdef __LITTLE_ENDIAN__
vmaxq_u16(uint16x8_t __p0,uint16x8_t __p1)13680 __ai uint16x8_t vmaxq_u16(uint16x8_t __p0, uint16x8_t __p1) {
13681 uint16x8_t __ret;
13682 __ret = (uint16x8_t) __builtin_neon_vmaxq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
13683 return __ret;
13684 }
13685 #else
vmaxq_u16(uint16x8_t __p0,uint16x8_t __p1)13686 __ai uint16x8_t vmaxq_u16(uint16x8_t __p0, uint16x8_t __p1) {
13687 uint16x8_t __ret;
13688 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
13689 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
13690 __ret = (uint16x8_t) __builtin_neon_vmaxq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
13691 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
13692 return __ret;
13693 }
13694 #endif
13695
13696 #ifdef __LITTLE_ENDIAN__
vmaxq_s8(int8x16_t __p0,int8x16_t __p1)13697 __ai int8x16_t vmaxq_s8(int8x16_t __p0, int8x16_t __p1) {
13698 int8x16_t __ret;
13699 __ret = (int8x16_t) __builtin_neon_vmaxq_v((int8x16_t)__p0, (int8x16_t)__p1, 32);
13700 return __ret;
13701 }
13702 #else
vmaxq_s8(int8x16_t __p0,int8x16_t __p1)13703 __ai int8x16_t vmaxq_s8(int8x16_t __p0, int8x16_t __p1) {
13704 int8x16_t __ret;
13705 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
13706 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
13707 __ret = (int8x16_t) __builtin_neon_vmaxq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 32);
13708 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
13709 return __ret;
13710 }
13711 #endif
13712
13713 #ifdef __LITTLE_ENDIAN__
vmaxq_f32(float32x4_t __p0,float32x4_t __p1)13714 __ai float32x4_t vmaxq_f32(float32x4_t __p0, float32x4_t __p1) {
13715 float32x4_t __ret;
13716 __ret = (float32x4_t) __builtin_neon_vmaxq_v((int8x16_t)__p0, (int8x16_t)__p1, 41);
13717 return __ret;
13718 }
13719 #else
vmaxq_f32(float32x4_t __p0,float32x4_t __p1)13720 __ai float32x4_t vmaxq_f32(float32x4_t __p0, float32x4_t __p1) {
13721 float32x4_t __ret;
13722 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
13723 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
13724 __ret = (float32x4_t) __builtin_neon_vmaxq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 41);
13725 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
13726 return __ret;
13727 }
13728 #endif
13729
13730 #ifdef __LITTLE_ENDIAN__
vmaxq_s32(int32x4_t __p0,int32x4_t __p1)13731 __ai int32x4_t vmaxq_s32(int32x4_t __p0, int32x4_t __p1) {
13732 int32x4_t __ret;
13733 __ret = (int32x4_t) __builtin_neon_vmaxq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
13734 return __ret;
13735 }
13736 #else
vmaxq_s32(int32x4_t __p0,int32x4_t __p1)13737 __ai int32x4_t vmaxq_s32(int32x4_t __p0, int32x4_t __p1) {
13738 int32x4_t __ret;
13739 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
13740 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
13741 __ret = (int32x4_t) __builtin_neon_vmaxq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
13742 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
13743 return __ret;
13744 }
13745 #endif
13746
13747 #ifdef __LITTLE_ENDIAN__
vmaxq_s16(int16x8_t __p0,int16x8_t __p1)13748 __ai int16x8_t vmaxq_s16(int16x8_t __p0, int16x8_t __p1) {
13749 int16x8_t __ret;
13750 __ret = (int16x8_t) __builtin_neon_vmaxq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
13751 return __ret;
13752 }
13753 #else
vmaxq_s16(int16x8_t __p0,int16x8_t __p1)13754 __ai int16x8_t vmaxq_s16(int16x8_t __p0, int16x8_t __p1) {
13755 int16x8_t __ret;
13756 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
13757 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
13758 __ret = (int16x8_t) __builtin_neon_vmaxq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
13759 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
13760 return __ret;
13761 }
13762 #endif
13763
13764 #ifdef __LITTLE_ENDIAN__
vmax_u8(uint8x8_t __p0,uint8x8_t __p1)13765 __ai uint8x8_t vmax_u8(uint8x8_t __p0, uint8x8_t __p1) {
13766 uint8x8_t __ret;
13767 __ret = (uint8x8_t) __builtin_neon_vmax_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
13768 return __ret;
13769 }
13770 #else
vmax_u8(uint8x8_t __p0,uint8x8_t __p1)13771 __ai uint8x8_t vmax_u8(uint8x8_t __p0, uint8x8_t __p1) {
13772 uint8x8_t __ret;
13773 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
13774 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
13775 __ret = (uint8x8_t) __builtin_neon_vmax_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
13776 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
13777 return __ret;
13778 }
13779 #endif
13780
13781 #ifdef __LITTLE_ENDIAN__
vmax_u32(uint32x2_t __p0,uint32x2_t __p1)13782 __ai uint32x2_t vmax_u32(uint32x2_t __p0, uint32x2_t __p1) {
13783 uint32x2_t __ret;
13784 __ret = (uint32x2_t) __builtin_neon_vmax_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
13785 return __ret;
13786 }
13787 #else
vmax_u32(uint32x2_t __p0,uint32x2_t __p1)13788 __ai uint32x2_t vmax_u32(uint32x2_t __p0, uint32x2_t __p1) {
13789 uint32x2_t __ret;
13790 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
13791 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
13792 __ret = (uint32x2_t) __builtin_neon_vmax_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
13793 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
13794 return __ret;
13795 }
13796 #endif
13797
13798 #ifdef __LITTLE_ENDIAN__
vmax_u16(uint16x4_t __p0,uint16x4_t __p1)13799 __ai uint16x4_t vmax_u16(uint16x4_t __p0, uint16x4_t __p1) {
13800 uint16x4_t __ret;
13801 __ret = (uint16x4_t) __builtin_neon_vmax_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
13802 return __ret;
13803 }
13804 #else
vmax_u16(uint16x4_t __p0,uint16x4_t __p1)13805 __ai uint16x4_t vmax_u16(uint16x4_t __p0, uint16x4_t __p1) {
13806 uint16x4_t __ret;
13807 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
13808 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
13809 __ret = (uint16x4_t) __builtin_neon_vmax_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
13810 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
13811 return __ret;
13812 }
13813 #endif
13814
13815 #ifdef __LITTLE_ENDIAN__
vmax_s8(int8x8_t __p0,int8x8_t __p1)13816 __ai int8x8_t vmax_s8(int8x8_t __p0, int8x8_t __p1) {
13817 int8x8_t __ret;
13818 __ret = (int8x8_t) __builtin_neon_vmax_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
13819 return __ret;
13820 }
13821 #else
vmax_s8(int8x8_t __p0,int8x8_t __p1)13822 __ai int8x8_t vmax_s8(int8x8_t __p0, int8x8_t __p1) {
13823 int8x8_t __ret;
13824 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
13825 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
13826 __ret = (int8x8_t) __builtin_neon_vmax_v((int8x8_t)__rev0, (int8x8_t)__rev1, 0);
13827 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
13828 return __ret;
13829 }
13830 #endif
13831
13832 #ifdef __LITTLE_ENDIAN__
vmax_f32(float32x2_t __p0,float32x2_t __p1)13833 __ai float32x2_t vmax_f32(float32x2_t __p0, float32x2_t __p1) {
13834 float32x2_t __ret;
13835 __ret = (float32x2_t) __builtin_neon_vmax_v((int8x8_t)__p0, (int8x8_t)__p1, 9);
13836 return __ret;
13837 }
13838 #else
vmax_f32(float32x2_t __p0,float32x2_t __p1)13839 __ai float32x2_t vmax_f32(float32x2_t __p0, float32x2_t __p1) {
13840 float32x2_t __ret;
13841 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
13842 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
13843 __ret = (float32x2_t) __builtin_neon_vmax_v((int8x8_t)__rev0, (int8x8_t)__rev1, 9);
13844 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
13845 return __ret;
13846 }
13847 #endif
13848
13849 #ifdef __LITTLE_ENDIAN__
vmax_s32(int32x2_t __p0,int32x2_t __p1)13850 __ai int32x2_t vmax_s32(int32x2_t __p0, int32x2_t __p1) {
13851 int32x2_t __ret;
13852 __ret = (int32x2_t) __builtin_neon_vmax_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
13853 return __ret;
13854 }
13855 #else
vmax_s32(int32x2_t __p0,int32x2_t __p1)13856 __ai int32x2_t vmax_s32(int32x2_t __p0, int32x2_t __p1) {
13857 int32x2_t __ret;
13858 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
13859 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
13860 __ret = (int32x2_t) __builtin_neon_vmax_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
13861 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
13862 return __ret;
13863 }
13864 #endif
13865
13866 #ifdef __LITTLE_ENDIAN__
vmax_s16(int16x4_t __p0,int16x4_t __p1)13867 __ai int16x4_t vmax_s16(int16x4_t __p0, int16x4_t __p1) {
13868 int16x4_t __ret;
13869 __ret = (int16x4_t) __builtin_neon_vmax_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
13870 return __ret;
13871 }
13872 #else
vmax_s16(int16x4_t __p0,int16x4_t __p1)13873 __ai int16x4_t vmax_s16(int16x4_t __p0, int16x4_t __p1) {
13874 int16x4_t __ret;
13875 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
13876 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
13877 __ret = (int16x4_t) __builtin_neon_vmax_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
13878 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
13879 return __ret;
13880 }
13881 #endif
13882
13883 #ifdef __LITTLE_ENDIAN__
vminq_u8(uint8x16_t __p0,uint8x16_t __p1)13884 __ai uint8x16_t vminq_u8(uint8x16_t __p0, uint8x16_t __p1) {
13885 uint8x16_t __ret;
13886 __ret = (uint8x16_t) __builtin_neon_vminq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
13887 return __ret;
13888 }
13889 #else
vminq_u8(uint8x16_t __p0,uint8x16_t __p1)13890 __ai uint8x16_t vminq_u8(uint8x16_t __p0, uint8x16_t __p1) {
13891 uint8x16_t __ret;
13892 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
13893 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
13894 __ret = (uint8x16_t) __builtin_neon_vminq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
13895 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
13896 return __ret;
13897 }
13898 #endif
13899
13900 #ifdef __LITTLE_ENDIAN__
vminq_u32(uint32x4_t __p0,uint32x4_t __p1)13901 __ai uint32x4_t vminq_u32(uint32x4_t __p0, uint32x4_t __p1) {
13902 uint32x4_t __ret;
13903 __ret = (uint32x4_t) __builtin_neon_vminq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
13904 return __ret;
13905 }
13906 #else
vminq_u32(uint32x4_t __p0,uint32x4_t __p1)13907 __ai uint32x4_t vminq_u32(uint32x4_t __p0, uint32x4_t __p1) {
13908 uint32x4_t __ret;
13909 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
13910 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
13911 __ret = (uint32x4_t) __builtin_neon_vminq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
13912 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
13913 return __ret;
13914 }
13915 #endif
13916
13917 #ifdef __LITTLE_ENDIAN__
vminq_u16(uint16x8_t __p0,uint16x8_t __p1)13918 __ai uint16x8_t vminq_u16(uint16x8_t __p0, uint16x8_t __p1) {
13919 uint16x8_t __ret;
13920 __ret = (uint16x8_t) __builtin_neon_vminq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
13921 return __ret;
13922 }
13923 #else
vminq_u16(uint16x8_t __p0,uint16x8_t __p1)13924 __ai uint16x8_t vminq_u16(uint16x8_t __p0, uint16x8_t __p1) {
13925 uint16x8_t __ret;
13926 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
13927 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
13928 __ret = (uint16x8_t) __builtin_neon_vminq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
13929 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
13930 return __ret;
13931 }
13932 #endif
13933
13934 #ifdef __LITTLE_ENDIAN__
vminq_s8(int8x16_t __p0,int8x16_t __p1)13935 __ai int8x16_t vminq_s8(int8x16_t __p0, int8x16_t __p1) {
13936 int8x16_t __ret;
13937 __ret = (int8x16_t) __builtin_neon_vminq_v((int8x16_t)__p0, (int8x16_t)__p1, 32);
13938 return __ret;
13939 }
13940 #else
vminq_s8(int8x16_t __p0,int8x16_t __p1)13941 __ai int8x16_t vminq_s8(int8x16_t __p0, int8x16_t __p1) {
13942 int8x16_t __ret;
13943 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
13944 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
13945 __ret = (int8x16_t) __builtin_neon_vminq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 32);
13946 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
13947 return __ret;
13948 }
13949 #endif
13950
13951 #ifdef __LITTLE_ENDIAN__
vminq_f32(float32x4_t __p0,float32x4_t __p1)13952 __ai float32x4_t vminq_f32(float32x4_t __p0, float32x4_t __p1) {
13953 float32x4_t __ret;
13954 __ret = (float32x4_t) __builtin_neon_vminq_v((int8x16_t)__p0, (int8x16_t)__p1, 41);
13955 return __ret;
13956 }
13957 #else
vminq_f32(float32x4_t __p0,float32x4_t __p1)13958 __ai float32x4_t vminq_f32(float32x4_t __p0, float32x4_t __p1) {
13959 float32x4_t __ret;
13960 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
13961 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
13962 __ret = (float32x4_t) __builtin_neon_vminq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 41);
13963 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
13964 return __ret;
13965 }
13966 #endif
13967
13968 #ifdef __LITTLE_ENDIAN__
vminq_s32(int32x4_t __p0,int32x4_t __p1)13969 __ai int32x4_t vminq_s32(int32x4_t __p0, int32x4_t __p1) {
13970 int32x4_t __ret;
13971 __ret = (int32x4_t) __builtin_neon_vminq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
13972 return __ret;
13973 }
13974 #else
vminq_s32(int32x4_t __p0,int32x4_t __p1)13975 __ai int32x4_t vminq_s32(int32x4_t __p0, int32x4_t __p1) {
13976 int32x4_t __ret;
13977 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
13978 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
13979 __ret = (int32x4_t) __builtin_neon_vminq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
13980 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
13981 return __ret;
13982 }
13983 #endif
13984
13985 #ifdef __LITTLE_ENDIAN__
vminq_s16(int16x8_t __p0,int16x8_t __p1)13986 __ai int16x8_t vminq_s16(int16x8_t __p0, int16x8_t __p1) {
13987 int16x8_t __ret;
13988 __ret = (int16x8_t) __builtin_neon_vminq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
13989 return __ret;
13990 }
13991 #else
vminq_s16(int16x8_t __p0,int16x8_t __p1)13992 __ai int16x8_t vminq_s16(int16x8_t __p0, int16x8_t __p1) {
13993 int16x8_t __ret;
13994 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
13995 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
13996 __ret = (int16x8_t) __builtin_neon_vminq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
13997 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
13998 return __ret;
13999 }
14000 #endif
14001
14002 #ifdef __LITTLE_ENDIAN__
vmin_u8(uint8x8_t __p0,uint8x8_t __p1)14003 __ai uint8x8_t vmin_u8(uint8x8_t __p0, uint8x8_t __p1) {
14004 uint8x8_t __ret;
14005 __ret = (uint8x8_t) __builtin_neon_vmin_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
14006 return __ret;
14007 }
14008 #else
vmin_u8(uint8x8_t __p0,uint8x8_t __p1)14009 __ai uint8x8_t vmin_u8(uint8x8_t __p0, uint8x8_t __p1) {
14010 uint8x8_t __ret;
14011 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
14012 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
14013 __ret = (uint8x8_t) __builtin_neon_vmin_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
14014 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
14015 return __ret;
14016 }
14017 #endif
14018
14019 #ifdef __LITTLE_ENDIAN__
vmin_u32(uint32x2_t __p0,uint32x2_t __p1)14020 __ai uint32x2_t vmin_u32(uint32x2_t __p0, uint32x2_t __p1) {
14021 uint32x2_t __ret;
14022 __ret = (uint32x2_t) __builtin_neon_vmin_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
14023 return __ret;
14024 }
14025 #else
vmin_u32(uint32x2_t __p0,uint32x2_t __p1)14026 __ai uint32x2_t vmin_u32(uint32x2_t __p0, uint32x2_t __p1) {
14027 uint32x2_t __ret;
14028 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
14029 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
14030 __ret = (uint32x2_t) __builtin_neon_vmin_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
14031 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
14032 return __ret;
14033 }
14034 #endif
14035
14036 #ifdef __LITTLE_ENDIAN__
vmin_u16(uint16x4_t __p0,uint16x4_t __p1)14037 __ai uint16x4_t vmin_u16(uint16x4_t __p0, uint16x4_t __p1) {
14038 uint16x4_t __ret;
14039 __ret = (uint16x4_t) __builtin_neon_vmin_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
14040 return __ret;
14041 }
14042 #else
vmin_u16(uint16x4_t __p0,uint16x4_t __p1)14043 __ai uint16x4_t vmin_u16(uint16x4_t __p0, uint16x4_t __p1) {
14044 uint16x4_t __ret;
14045 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
14046 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
14047 __ret = (uint16x4_t) __builtin_neon_vmin_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
14048 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
14049 return __ret;
14050 }
14051 #endif
14052
14053 #ifdef __LITTLE_ENDIAN__
vmin_s8(int8x8_t __p0,int8x8_t __p1)14054 __ai int8x8_t vmin_s8(int8x8_t __p0, int8x8_t __p1) {
14055 int8x8_t __ret;
14056 __ret = (int8x8_t) __builtin_neon_vmin_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
14057 return __ret;
14058 }
14059 #else
vmin_s8(int8x8_t __p0,int8x8_t __p1)14060 __ai int8x8_t vmin_s8(int8x8_t __p0, int8x8_t __p1) {
14061 int8x8_t __ret;
14062 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
14063 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
14064 __ret = (int8x8_t) __builtin_neon_vmin_v((int8x8_t)__rev0, (int8x8_t)__rev1, 0);
14065 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
14066 return __ret;
14067 }
14068 #endif
14069
14070 #ifdef __LITTLE_ENDIAN__
vmin_f32(float32x2_t __p0,float32x2_t __p1)14071 __ai float32x2_t vmin_f32(float32x2_t __p0, float32x2_t __p1) {
14072 float32x2_t __ret;
14073 __ret = (float32x2_t) __builtin_neon_vmin_v((int8x8_t)__p0, (int8x8_t)__p1, 9);
14074 return __ret;
14075 }
14076 #else
vmin_f32(float32x2_t __p0,float32x2_t __p1)14077 __ai float32x2_t vmin_f32(float32x2_t __p0, float32x2_t __p1) {
14078 float32x2_t __ret;
14079 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
14080 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
14081 __ret = (float32x2_t) __builtin_neon_vmin_v((int8x8_t)__rev0, (int8x8_t)__rev1, 9);
14082 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
14083 return __ret;
14084 }
14085 #endif
14086
14087 #ifdef __LITTLE_ENDIAN__
vmin_s32(int32x2_t __p0,int32x2_t __p1)14088 __ai int32x2_t vmin_s32(int32x2_t __p0, int32x2_t __p1) {
14089 int32x2_t __ret;
14090 __ret = (int32x2_t) __builtin_neon_vmin_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
14091 return __ret;
14092 }
14093 #else
vmin_s32(int32x2_t __p0,int32x2_t __p1)14094 __ai int32x2_t vmin_s32(int32x2_t __p0, int32x2_t __p1) {
14095 int32x2_t __ret;
14096 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
14097 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
14098 __ret = (int32x2_t) __builtin_neon_vmin_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
14099 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
14100 return __ret;
14101 }
14102 #endif
14103
14104 #ifdef __LITTLE_ENDIAN__
vmin_s16(int16x4_t __p0,int16x4_t __p1)14105 __ai int16x4_t vmin_s16(int16x4_t __p0, int16x4_t __p1) {
14106 int16x4_t __ret;
14107 __ret = (int16x4_t) __builtin_neon_vmin_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
14108 return __ret;
14109 }
14110 #else
vmin_s16(int16x4_t __p0,int16x4_t __p1)14111 __ai int16x4_t vmin_s16(int16x4_t __p0, int16x4_t __p1) {
14112 int16x4_t __ret;
14113 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
14114 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
14115 __ret = (int16x4_t) __builtin_neon_vmin_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
14116 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
14117 return __ret;
14118 }
14119 #endif
14120
14121 #ifdef __LITTLE_ENDIAN__
vmlaq_u8(uint8x16_t __p0,uint8x16_t __p1,uint8x16_t __p2)14122 __ai uint8x16_t vmlaq_u8(uint8x16_t __p0, uint8x16_t __p1, uint8x16_t __p2) {
14123 uint8x16_t __ret;
14124 __ret = __p0 + __p1 * __p2;
14125 return __ret;
14126 }
14127 #else
vmlaq_u8(uint8x16_t __p0,uint8x16_t __p1,uint8x16_t __p2)14128 __ai uint8x16_t vmlaq_u8(uint8x16_t __p0, uint8x16_t __p1, uint8x16_t __p2) {
14129 uint8x16_t __ret;
14130 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
14131 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
14132 uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
14133 __ret = __rev0 + __rev1 * __rev2;
14134 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
14135 return __ret;
14136 }
14137 #endif
14138
14139 #ifdef __LITTLE_ENDIAN__
vmlaq_u32(uint32x4_t __p0,uint32x4_t __p1,uint32x4_t __p2)14140 __ai uint32x4_t vmlaq_u32(uint32x4_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
14141 uint32x4_t __ret;
14142 __ret = __p0 + __p1 * __p2;
14143 return __ret;
14144 }
14145 #else
vmlaq_u32(uint32x4_t __p0,uint32x4_t __p1,uint32x4_t __p2)14146 __ai uint32x4_t vmlaq_u32(uint32x4_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
14147 uint32x4_t __ret;
14148 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
14149 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
14150 uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
14151 __ret = __rev0 + __rev1 * __rev2;
14152 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
14153 return __ret;
14154 }
14155 #endif
14156
14157 #ifdef __LITTLE_ENDIAN__
vmlaq_u16(uint16x8_t __p0,uint16x8_t __p1,uint16x8_t __p2)14158 __ai uint16x8_t vmlaq_u16(uint16x8_t __p0, uint16x8_t __p1, uint16x8_t __p2) {
14159 uint16x8_t __ret;
14160 __ret = __p0 + __p1 * __p2;
14161 return __ret;
14162 }
14163 #else
vmlaq_u16(uint16x8_t __p0,uint16x8_t __p1,uint16x8_t __p2)14164 __ai uint16x8_t vmlaq_u16(uint16x8_t __p0, uint16x8_t __p1, uint16x8_t __p2) {
14165 uint16x8_t __ret;
14166 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
14167 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
14168 uint16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
14169 __ret = __rev0 + __rev1 * __rev2;
14170 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
14171 return __ret;
14172 }
14173 #endif
14174
14175 #ifdef __LITTLE_ENDIAN__
vmlaq_s8(int8x16_t __p0,int8x16_t __p1,int8x16_t __p2)14176 __ai int8x16_t vmlaq_s8(int8x16_t __p0, int8x16_t __p1, int8x16_t __p2) {
14177 int8x16_t __ret;
14178 __ret = __p0 + __p1 * __p2;
14179 return __ret;
14180 }
14181 #else
vmlaq_s8(int8x16_t __p0,int8x16_t __p1,int8x16_t __p2)14182 __ai int8x16_t vmlaq_s8(int8x16_t __p0, int8x16_t __p1, int8x16_t __p2) {
14183 int8x16_t __ret;
14184 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
14185 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
14186 int8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
14187 __ret = __rev0 + __rev1 * __rev2;
14188 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
14189 return __ret;
14190 }
14191 #endif
14192
14193 #ifdef __LITTLE_ENDIAN__
vmlaq_f32(float32x4_t __p0,float32x4_t __p1,float32x4_t __p2)14194 __ai float32x4_t vmlaq_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
14195 float32x4_t __ret;
14196 __ret = __p0 + __p1 * __p2;
14197 return __ret;
14198 }
14199 #else
vmlaq_f32(float32x4_t __p0,float32x4_t __p1,float32x4_t __p2)14200 __ai float32x4_t vmlaq_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
14201 float32x4_t __ret;
14202 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
14203 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
14204 float32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
14205 __ret = __rev0 + __rev1 * __rev2;
14206 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
14207 return __ret;
14208 }
14209 #endif
14210
14211 #ifdef __LITTLE_ENDIAN__
vmlaq_s32(int32x4_t __p0,int32x4_t __p1,int32x4_t __p2)14212 __ai int32x4_t vmlaq_s32(int32x4_t __p0, int32x4_t __p1, int32x4_t __p2) {
14213 int32x4_t __ret;
14214 __ret = __p0 + __p1 * __p2;
14215 return __ret;
14216 }
14217 #else
vmlaq_s32(int32x4_t __p0,int32x4_t __p1,int32x4_t __p2)14218 __ai int32x4_t vmlaq_s32(int32x4_t __p0, int32x4_t __p1, int32x4_t __p2) {
14219 int32x4_t __ret;
14220 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
14221 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
14222 int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
14223 __ret = __rev0 + __rev1 * __rev2;
14224 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
14225 return __ret;
14226 }
14227 #endif
14228
14229 #ifdef __LITTLE_ENDIAN__
vmlaq_s16(int16x8_t __p0,int16x8_t __p1,int16x8_t __p2)14230 __ai int16x8_t vmlaq_s16(int16x8_t __p0, int16x8_t __p1, int16x8_t __p2) {
14231 int16x8_t __ret;
14232 __ret = __p0 + __p1 * __p2;
14233 return __ret;
14234 }
14235 #else
vmlaq_s16(int16x8_t __p0,int16x8_t __p1,int16x8_t __p2)14236 __ai int16x8_t vmlaq_s16(int16x8_t __p0, int16x8_t __p1, int16x8_t __p2) {
14237 int16x8_t __ret;
14238 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
14239 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
14240 int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
14241 __ret = __rev0 + __rev1 * __rev2;
14242 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
14243 return __ret;
14244 }
14245 #endif
14246
14247 #ifdef __LITTLE_ENDIAN__
vmla_u8(uint8x8_t __p0,uint8x8_t __p1,uint8x8_t __p2)14248 __ai uint8x8_t vmla_u8(uint8x8_t __p0, uint8x8_t __p1, uint8x8_t __p2) {
14249 uint8x8_t __ret;
14250 __ret = __p0 + __p1 * __p2;
14251 return __ret;
14252 }
14253 #else
vmla_u8(uint8x8_t __p0,uint8x8_t __p1,uint8x8_t __p2)14254 __ai uint8x8_t vmla_u8(uint8x8_t __p0, uint8x8_t __p1, uint8x8_t __p2) {
14255 uint8x8_t __ret;
14256 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
14257 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
14258 uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
14259 __ret = __rev0 + __rev1 * __rev2;
14260 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
14261 return __ret;
14262 }
14263 #endif
14264
14265 #ifdef __LITTLE_ENDIAN__
vmla_u32(uint32x2_t __p0,uint32x2_t __p1,uint32x2_t __p2)14266 __ai uint32x2_t vmla_u32(uint32x2_t __p0, uint32x2_t __p1, uint32x2_t __p2) {
14267 uint32x2_t __ret;
14268 __ret = __p0 + __p1 * __p2;
14269 return __ret;
14270 }
14271 #else
vmla_u32(uint32x2_t __p0,uint32x2_t __p1,uint32x2_t __p2)14272 __ai uint32x2_t vmla_u32(uint32x2_t __p0, uint32x2_t __p1, uint32x2_t __p2) {
14273 uint32x2_t __ret;
14274 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
14275 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
14276 uint32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
14277 __ret = __rev0 + __rev1 * __rev2;
14278 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
14279 return __ret;
14280 }
14281 #endif
14282
14283 #ifdef __LITTLE_ENDIAN__
vmla_u16(uint16x4_t __p0,uint16x4_t __p1,uint16x4_t __p2)14284 __ai uint16x4_t vmla_u16(uint16x4_t __p0, uint16x4_t __p1, uint16x4_t __p2) {
14285 uint16x4_t __ret;
14286 __ret = __p0 + __p1 * __p2;
14287 return __ret;
14288 }
14289 #else
vmla_u16(uint16x4_t __p0,uint16x4_t __p1,uint16x4_t __p2)14290 __ai uint16x4_t vmla_u16(uint16x4_t __p0, uint16x4_t __p1, uint16x4_t __p2) {
14291 uint16x4_t __ret;
14292 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
14293 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
14294 uint16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
14295 __ret = __rev0 + __rev1 * __rev2;
14296 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
14297 return __ret;
14298 }
14299 #endif
14300
14301 #ifdef __LITTLE_ENDIAN__
vmla_s8(int8x8_t __p0,int8x8_t __p1,int8x8_t __p2)14302 __ai int8x8_t vmla_s8(int8x8_t __p0, int8x8_t __p1, int8x8_t __p2) {
14303 int8x8_t __ret;
14304 __ret = __p0 + __p1 * __p2;
14305 return __ret;
14306 }
14307 #else
vmla_s8(int8x8_t __p0,int8x8_t __p1,int8x8_t __p2)14308 __ai int8x8_t vmla_s8(int8x8_t __p0, int8x8_t __p1, int8x8_t __p2) {
14309 int8x8_t __ret;
14310 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
14311 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
14312 int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
14313 __ret = __rev0 + __rev1 * __rev2;
14314 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
14315 return __ret;
14316 }
14317 #endif
14318
14319 #ifdef __LITTLE_ENDIAN__
vmla_f32(float32x2_t __p0,float32x2_t __p1,float32x2_t __p2)14320 __ai float32x2_t vmla_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
14321 float32x2_t __ret;
14322 __ret = __p0 + __p1 * __p2;
14323 return __ret;
14324 }
14325 #else
vmla_f32(float32x2_t __p0,float32x2_t __p1,float32x2_t __p2)14326 __ai float32x2_t vmla_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
14327 float32x2_t __ret;
14328 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
14329 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
14330 float32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
14331 __ret = __rev0 + __rev1 * __rev2;
14332 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
14333 return __ret;
14334 }
14335 #endif
14336
14337 #ifdef __LITTLE_ENDIAN__
vmla_s32(int32x2_t __p0,int32x2_t __p1,int32x2_t __p2)14338 __ai int32x2_t vmla_s32(int32x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
14339 int32x2_t __ret;
14340 __ret = __p0 + __p1 * __p2;
14341 return __ret;
14342 }
14343 #else
vmla_s32(int32x2_t __p0,int32x2_t __p1,int32x2_t __p2)14344 __ai int32x2_t vmla_s32(int32x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
14345 int32x2_t __ret;
14346 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
14347 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
14348 int32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
14349 __ret = __rev0 + __rev1 * __rev2;
14350 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
14351 return __ret;
14352 }
14353 #endif
14354
14355 #ifdef __LITTLE_ENDIAN__
vmla_s16(int16x4_t __p0,int16x4_t __p1,int16x4_t __p2)14356 __ai int16x4_t vmla_s16(int16x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
14357 int16x4_t __ret;
14358 __ret = __p0 + __p1 * __p2;
14359 return __ret;
14360 }
14361 #else
vmla_s16(int16x4_t __p0,int16x4_t __p1,int16x4_t __p2)14362 __ai int16x4_t vmla_s16(int16x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
14363 int16x4_t __ret;
14364 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
14365 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
14366 int16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
14367 __ret = __rev0 + __rev1 * __rev2;
14368 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
14369 return __ret;
14370 }
14371 #endif
14372
14373 #ifdef __LITTLE_ENDIAN__
14374 #define vmlaq_lane_u32(__p0_46, __p1_46, __p2_46, __p3_46) __extension__ ({ \
14375 uint32x4_t __ret_46; \
14376 uint32x4_t __s0_46 = __p0_46; \
14377 uint32x4_t __s1_46 = __p1_46; \
14378 uint32x2_t __s2_46 = __p2_46; \
14379 __ret_46 = __s0_46 + __s1_46 * splatq_lane_u32(__s2_46, __p3_46); \
14380 __ret_46; \
14381 })
14382 #else
14383 #define vmlaq_lane_u32(__p0_47, __p1_47, __p2_47, __p3_47) __extension__ ({ \
14384 uint32x4_t __ret_47; \
14385 uint32x4_t __s0_47 = __p0_47; \
14386 uint32x4_t __s1_47 = __p1_47; \
14387 uint32x2_t __s2_47 = __p2_47; \
14388 uint32x4_t __rev0_47; __rev0_47 = __builtin_shufflevector(__s0_47, __s0_47, 3, 2, 1, 0); \
14389 uint32x4_t __rev1_47; __rev1_47 = __builtin_shufflevector(__s1_47, __s1_47, 3, 2, 1, 0); \
14390 uint32x2_t __rev2_47; __rev2_47 = __builtin_shufflevector(__s2_47, __s2_47, 1, 0); \
14391 __ret_47 = __rev0_47 + __rev1_47 * __noswap_splatq_lane_u32(__rev2_47, __p3_47); \
14392 __ret_47 = __builtin_shufflevector(__ret_47, __ret_47, 3, 2, 1, 0); \
14393 __ret_47; \
14394 })
14395 #endif
14396
14397 #ifdef __LITTLE_ENDIAN__
14398 #define vmlaq_lane_u16(__p0_48, __p1_48, __p2_48, __p3_48) __extension__ ({ \
14399 uint16x8_t __ret_48; \
14400 uint16x8_t __s0_48 = __p0_48; \
14401 uint16x8_t __s1_48 = __p1_48; \
14402 uint16x4_t __s2_48 = __p2_48; \
14403 __ret_48 = __s0_48 + __s1_48 * splatq_lane_u16(__s2_48, __p3_48); \
14404 __ret_48; \
14405 })
14406 #else
14407 #define vmlaq_lane_u16(__p0_49, __p1_49, __p2_49, __p3_49) __extension__ ({ \
14408 uint16x8_t __ret_49; \
14409 uint16x8_t __s0_49 = __p0_49; \
14410 uint16x8_t __s1_49 = __p1_49; \
14411 uint16x4_t __s2_49 = __p2_49; \
14412 uint16x8_t __rev0_49; __rev0_49 = __builtin_shufflevector(__s0_49, __s0_49, 7, 6, 5, 4, 3, 2, 1, 0); \
14413 uint16x8_t __rev1_49; __rev1_49 = __builtin_shufflevector(__s1_49, __s1_49, 7, 6, 5, 4, 3, 2, 1, 0); \
14414 uint16x4_t __rev2_49; __rev2_49 = __builtin_shufflevector(__s2_49, __s2_49, 3, 2, 1, 0); \
14415 __ret_49 = __rev0_49 + __rev1_49 * __noswap_splatq_lane_u16(__rev2_49, __p3_49); \
14416 __ret_49 = __builtin_shufflevector(__ret_49, __ret_49, 7, 6, 5, 4, 3, 2, 1, 0); \
14417 __ret_49; \
14418 })
14419 #endif
14420
14421 #ifdef __LITTLE_ENDIAN__
14422 #define vmlaq_lane_f32(__p0_50, __p1_50, __p2_50, __p3_50) __extension__ ({ \
14423 float32x4_t __ret_50; \
14424 float32x4_t __s0_50 = __p0_50; \
14425 float32x4_t __s1_50 = __p1_50; \
14426 float32x2_t __s2_50 = __p2_50; \
14427 __ret_50 = __s0_50 + __s1_50 * splatq_lane_f32(__s2_50, __p3_50); \
14428 __ret_50; \
14429 })
14430 #else
14431 #define vmlaq_lane_f32(__p0_51, __p1_51, __p2_51, __p3_51) __extension__ ({ \
14432 float32x4_t __ret_51; \
14433 float32x4_t __s0_51 = __p0_51; \
14434 float32x4_t __s1_51 = __p1_51; \
14435 float32x2_t __s2_51 = __p2_51; \
14436 float32x4_t __rev0_51; __rev0_51 = __builtin_shufflevector(__s0_51, __s0_51, 3, 2, 1, 0); \
14437 float32x4_t __rev1_51; __rev1_51 = __builtin_shufflevector(__s1_51, __s1_51, 3, 2, 1, 0); \
14438 float32x2_t __rev2_51; __rev2_51 = __builtin_shufflevector(__s2_51, __s2_51, 1, 0); \
14439 __ret_51 = __rev0_51 + __rev1_51 * __noswap_splatq_lane_f32(__rev2_51, __p3_51); \
14440 __ret_51 = __builtin_shufflevector(__ret_51, __ret_51, 3, 2, 1, 0); \
14441 __ret_51; \
14442 })
14443 #endif
14444
14445 #ifdef __LITTLE_ENDIAN__
14446 #define vmlaq_lane_s32(__p0_52, __p1_52, __p2_52, __p3_52) __extension__ ({ \
14447 int32x4_t __ret_52; \
14448 int32x4_t __s0_52 = __p0_52; \
14449 int32x4_t __s1_52 = __p1_52; \
14450 int32x2_t __s2_52 = __p2_52; \
14451 __ret_52 = __s0_52 + __s1_52 * splatq_lane_s32(__s2_52, __p3_52); \
14452 __ret_52; \
14453 })
14454 #else
14455 #define vmlaq_lane_s32(__p0_53, __p1_53, __p2_53, __p3_53) __extension__ ({ \
14456 int32x4_t __ret_53; \
14457 int32x4_t __s0_53 = __p0_53; \
14458 int32x4_t __s1_53 = __p1_53; \
14459 int32x2_t __s2_53 = __p2_53; \
14460 int32x4_t __rev0_53; __rev0_53 = __builtin_shufflevector(__s0_53, __s0_53, 3, 2, 1, 0); \
14461 int32x4_t __rev1_53; __rev1_53 = __builtin_shufflevector(__s1_53, __s1_53, 3, 2, 1, 0); \
14462 int32x2_t __rev2_53; __rev2_53 = __builtin_shufflevector(__s2_53, __s2_53, 1, 0); \
14463 __ret_53 = __rev0_53 + __rev1_53 * __noswap_splatq_lane_s32(__rev2_53, __p3_53); \
14464 __ret_53 = __builtin_shufflevector(__ret_53, __ret_53, 3, 2, 1, 0); \
14465 __ret_53; \
14466 })
14467 #endif
14468
14469 #ifdef __LITTLE_ENDIAN__
14470 #define vmlaq_lane_s16(__p0_54, __p1_54, __p2_54, __p3_54) __extension__ ({ \
14471 int16x8_t __ret_54; \
14472 int16x8_t __s0_54 = __p0_54; \
14473 int16x8_t __s1_54 = __p1_54; \
14474 int16x4_t __s2_54 = __p2_54; \
14475 __ret_54 = __s0_54 + __s1_54 * splatq_lane_s16(__s2_54, __p3_54); \
14476 __ret_54; \
14477 })
14478 #else
14479 #define vmlaq_lane_s16(__p0_55, __p1_55, __p2_55, __p3_55) __extension__ ({ \
14480 int16x8_t __ret_55; \
14481 int16x8_t __s0_55 = __p0_55; \
14482 int16x8_t __s1_55 = __p1_55; \
14483 int16x4_t __s2_55 = __p2_55; \
14484 int16x8_t __rev0_55; __rev0_55 = __builtin_shufflevector(__s0_55, __s0_55, 7, 6, 5, 4, 3, 2, 1, 0); \
14485 int16x8_t __rev1_55; __rev1_55 = __builtin_shufflevector(__s1_55, __s1_55, 7, 6, 5, 4, 3, 2, 1, 0); \
14486 int16x4_t __rev2_55; __rev2_55 = __builtin_shufflevector(__s2_55, __s2_55, 3, 2, 1, 0); \
14487 __ret_55 = __rev0_55 + __rev1_55 * __noswap_splatq_lane_s16(__rev2_55, __p3_55); \
14488 __ret_55 = __builtin_shufflevector(__ret_55, __ret_55, 7, 6, 5, 4, 3, 2, 1, 0); \
14489 __ret_55; \
14490 })
14491 #endif
14492
14493 #ifdef __LITTLE_ENDIAN__
14494 #define vmla_lane_u32(__p0_56, __p1_56, __p2_56, __p3_56) __extension__ ({ \
14495 uint32x2_t __ret_56; \
14496 uint32x2_t __s0_56 = __p0_56; \
14497 uint32x2_t __s1_56 = __p1_56; \
14498 uint32x2_t __s2_56 = __p2_56; \
14499 __ret_56 = __s0_56 + __s1_56 * splat_lane_u32(__s2_56, __p3_56); \
14500 __ret_56; \
14501 })
14502 #else
14503 #define vmla_lane_u32(__p0_57, __p1_57, __p2_57, __p3_57) __extension__ ({ \
14504 uint32x2_t __ret_57; \
14505 uint32x2_t __s0_57 = __p0_57; \
14506 uint32x2_t __s1_57 = __p1_57; \
14507 uint32x2_t __s2_57 = __p2_57; \
14508 uint32x2_t __rev0_57; __rev0_57 = __builtin_shufflevector(__s0_57, __s0_57, 1, 0); \
14509 uint32x2_t __rev1_57; __rev1_57 = __builtin_shufflevector(__s1_57, __s1_57, 1, 0); \
14510 uint32x2_t __rev2_57; __rev2_57 = __builtin_shufflevector(__s2_57, __s2_57, 1, 0); \
14511 __ret_57 = __rev0_57 + __rev1_57 * __noswap_splat_lane_u32(__rev2_57, __p3_57); \
14512 __ret_57 = __builtin_shufflevector(__ret_57, __ret_57, 1, 0); \
14513 __ret_57; \
14514 })
14515 #endif
14516
14517 #ifdef __LITTLE_ENDIAN__
14518 #define vmla_lane_u16(__p0_58, __p1_58, __p2_58, __p3_58) __extension__ ({ \
14519 uint16x4_t __ret_58; \
14520 uint16x4_t __s0_58 = __p0_58; \
14521 uint16x4_t __s1_58 = __p1_58; \
14522 uint16x4_t __s2_58 = __p2_58; \
14523 __ret_58 = __s0_58 + __s1_58 * splat_lane_u16(__s2_58, __p3_58); \
14524 __ret_58; \
14525 })
14526 #else
14527 #define vmla_lane_u16(__p0_59, __p1_59, __p2_59, __p3_59) __extension__ ({ \
14528 uint16x4_t __ret_59; \
14529 uint16x4_t __s0_59 = __p0_59; \
14530 uint16x4_t __s1_59 = __p1_59; \
14531 uint16x4_t __s2_59 = __p2_59; \
14532 uint16x4_t __rev0_59; __rev0_59 = __builtin_shufflevector(__s0_59, __s0_59, 3, 2, 1, 0); \
14533 uint16x4_t __rev1_59; __rev1_59 = __builtin_shufflevector(__s1_59, __s1_59, 3, 2, 1, 0); \
14534 uint16x4_t __rev2_59; __rev2_59 = __builtin_shufflevector(__s2_59, __s2_59, 3, 2, 1, 0); \
14535 __ret_59 = __rev0_59 + __rev1_59 * __noswap_splat_lane_u16(__rev2_59, __p3_59); \
14536 __ret_59 = __builtin_shufflevector(__ret_59, __ret_59, 3, 2, 1, 0); \
14537 __ret_59; \
14538 })
14539 #endif
14540
14541 #ifdef __LITTLE_ENDIAN__
14542 #define vmla_lane_f32(__p0_60, __p1_60, __p2_60, __p3_60) __extension__ ({ \
14543 float32x2_t __ret_60; \
14544 float32x2_t __s0_60 = __p0_60; \
14545 float32x2_t __s1_60 = __p1_60; \
14546 float32x2_t __s2_60 = __p2_60; \
14547 __ret_60 = __s0_60 + __s1_60 * splat_lane_f32(__s2_60, __p3_60); \
14548 __ret_60; \
14549 })
14550 #else
14551 #define vmla_lane_f32(__p0_61, __p1_61, __p2_61, __p3_61) __extension__ ({ \
14552 float32x2_t __ret_61; \
14553 float32x2_t __s0_61 = __p0_61; \
14554 float32x2_t __s1_61 = __p1_61; \
14555 float32x2_t __s2_61 = __p2_61; \
14556 float32x2_t __rev0_61; __rev0_61 = __builtin_shufflevector(__s0_61, __s0_61, 1, 0); \
14557 float32x2_t __rev1_61; __rev1_61 = __builtin_shufflevector(__s1_61, __s1_61, 1, 0); \
14558 float32x2_t __rev2_61; __rev2_61 = __builtin_shufflevector(__s2_61, __s2_61, 1, 0); \
14559 __ret_61 = __rev0_61 + __rev1_61 * __noswap_splat_lane_f32(__rev2_61, __p3_61); \
14560 __ret_61 = __builtin_shufflevector(__ret_61, __ret_61, 1, 0); \
14561 __ret_61; \
14562 })
14563 #endif
14564
14565 #ifdef __LITTLE_ENDIAN__
14566 #define vmla_lane_s32(__p0_62, __p1_62, __p2_62, __p3_62) __extension__ ({ \
14567 int32x2_t __ret_62; \
14568 int32x2_t __s0_62 = __p0_62; \
14569 int32x2_t __s1_62 = __p1_62; \
14570 int32x2_t __s2_62 = __p2_62; \
14571 __ret_62 = __s0_62 + __s1_62 * splat_lane_s32(__s2_62, __p3_62); \
14572 __ret_62; \
14573 })
14574 #else
14575 #define vmla_lane_s32(__p0_63, __p1_63, __p2_63, __p3_63) __extension__ ({ \
14576 int32x2_t __ret_63; \
14577 int32x2_t __s0_63 = __p0_63; \
14578 int32x2_t __s1_63 = __p1_63; \
14579 int32x2_t __s2_63 = __p2_63; \
14580 int32x2_t __rev0_63; __rev0_63 = __builtin_shufflevector(__s0_63, __s0_63, 1, 0); \
14581 int32x2_t __rev1_63; __rev1_63 = __builtin_shufflevector(__s1_63, __s1_63, 1, 0); \
14582 int32x2_t __rev2_63; __rev2_63 = __builtin_shufflevector(__s2_63, __s2_63, 1, 0); \
14583 __ret_63 = __rev0_63 + __rev1_63 * __noswap_splat_lane_s32(__rev2_63, __p3_63); \
14584 __ret_63 = __builtin_shufflevector(__ret_63, __ret_63, 1, 0); \
14585 __ret_63; \
14586 })
14587 #endif
14588
14589 #ifdef __LITTLE_ENDIAN__
14590 #define vmla_lane_s16(__p0_64, __p1_64, __p2_64, __p3_64) __extension__ ({ \
14591 int16x4_t __ret_64; \
14592 int16x4_t __s0_64 = __p0_64; \
14593 int16x4_t __s1_64 = __p1_64; \
14594 int16x4_t __s2_64 = __p2_64; \
14595 __ret_64 = __s0_64 + __s1_64 * splat_lane_s16(__s2_64, __p3_64); \
14596 __ret_64; \
14597 })
14598 #else
14599 #define vmla_lane_s16(__p0_65, __p1_65, __p2_65, __p3_65) __extension__ ({ \
14600 int16x4_t __ret_65; \
14601 int16x4_t __s0_65 = __p0_65; \
14602 int16x4_t __s1_65 = __p1_65; \
14603 int16x4_t __s2_65 = __p2_65; \
14604 int16x4_t __rev0_65; __rev0_65 = __builtin_shufflevector(__s0_65, __s0_65, 3, 2, 1, 0); \
14605 int16x4_t __rev1_65; __rev1_65 = __builtin_shufflevector(__s1_65, __s1_65, 3, 2, 1, 0); \
14606 int16x4_t __rev2_65; __rev2_65 = __builtin_shufflevector(__s2_65, __s2_65, 3, 2, 1, 0); \
14607 __ret_65 = __rev0_65 + __rev1_65 * __noswap_splat_lane_s16(__rev2_65, __p3_65); \
14608 __ret_65 = __builtin_shufflevector(__ret_65, __ret_65, 3, 2, 1, 0); \
14609 __ret_65; \
14610 })
14611 #endif
14612
14613 #ifdef __LITTLE_ENDIAN__
vmlaq_n_u32(uint32x4_t __p0,uint32x4_t __p1,uint32_t __p2)14614 __ai uint32x4_t vmlaq_n_u32(uint32x4_t __p0, uint32x4_t __p1, uint32_t __p2) {
14615 uint32x4_t __ret;
14616 __ret = __p0 + __p1 * (uint32x4_t) {__p2, __p2, __p2, __p2};
14617 return __ret;
14618 }
14619 #else
vmlaq_n_u32(uint32x4_t __p0,uint32x4_t __p1,uint32_t __p2)14620 __ai uint32x4_t vmlaq_n_u32(uint32x4_t __p0, uint32x4_t __p1, uint32_t __p2) {
14621 uint32x4_t __ret;
14622 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
14623 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
14624 __ret = __rev0 + __rev1 * (uint32x4_t) {__p2, __p2, __p2, __p2};
14625 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
14626 return __ret;
14627 }
14628 #endif
14629
14630 #ifdef __LITTLE_ENDIAN__
vmlaq_n_u16(uint16x8_t __p0,uint16x8_t __p1,uint16_t __p2)14631 __ai uint16x8_t vmlaq_n_u16(uint16x8_t __p0, uint16x8_t __p1, uint16_t __p2) {
14632 uint16x8_t __ret;
14633 __ret = __p0 + __p1 * (uint16x8_t) {__p2, __p2, __p2, __p2, __p2, __p2, __p2, __p2};
14634 return __ret;
14635 }
14636 #else
vmlaq_n_u16(uint16x8_t __p0,uint16x8_t __p1,uint16_t __p2)14637 __ai uint16x8_t vmlaq_n_u16(uint16x8_t __p0, uint16x8_t __p1, uint16_t __p2) {
14638 uint16x8_t __ret;
14639 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
14640 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
14641 __ret = __rev0 + __rev1 * (uint16x8_t) {__p2, __p2, __p2, __p2, __p2, __p2, __p2, __p2};
14642 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
14643 return __ret;
14644 }
14645 #endif
14646
14647 #ifdef __LITTLE_ENDIAN__
vmlaq_n_f32(float32x4_t __p0,float32x4_t __p1,float32_t __p2)14648 __ai float32x4_t vmlaq_n_f32(float32x4_t __p0, float32x4_t __p1, float32_t __p2) {
14649 float32x4_t __ret;
14650 __ret = __p0 + __p1 * (float32x4_t) {__p2, __p2, __p2, __p2};
14651 return __ret;
14652 }
14653 #else
vmlaq_n_f32(float32x4_t __p0,float32x4_t __p1,float32_t __p2)14654 __ai float32x4_t vmlaq_n_f32(float32x4_t __p0, float32x4_t __p1, float32_t __p2) {
14655 float32x4_t __ret;
14656 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
14657 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
14658 __ret = __rev0 + __rev1 * (float32x4_t) {__p2, __p2, __p2, __p2};
14659 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
14660 return __ret;
14661 }
14662 #endif
14663
14664 #ifdef __LITTLE_ENDIAN__
vmlaq_n_s32(int32x4_t __p0,int32x4_t __p1,int32_t __p2)14665 __ai int32x4_t vmlaq_n_s32(int32x4_t __p0, int32x4_t __p1, int32_t __p2) {
14666 int32x4_t __ret;
14667 __ret = __p0 + __p1 * (int32x4_t) {__p2, __p2, __p2, __p2};
14668 return __ret;
14669 }
14670 #else
vmlaq_n_s32(int32x4_t __p0,int32x4_t __p1,int32_t __p2)14671 __ai int32x4_t vmlaq_n_s32(int32x4_t __p0, int32x4_t __p1, int32_t __p2) {
14672 int32x4_t __ret;
14673 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
14674 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
14675 __ret = __rev0 + __rev1 * (int32x4_t) {__p2, __p2, __p2, __p2};
14676 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
14677 return __ret;
14678 }
14679 #endif
14680
14681 #ifdef __LITTLE_ENDIAN__
vmlaq_n_s16(int16x8_t __p0,int16x8_t __p1,int16_t __p2)14682 __ai int16x8_t vmlaq_n_s16(int16x8_t __p0, int16x8_t __p1, int16_t __p2) {
14683 int16x8_t __ret;
14684 __ret = __p0 + __p1 * (int16x8_t) {__p2, __p2, __p2, __p2, __p2, __p2, __p2, __p2};
14685 return __ret;
14686 }
14687 #else
vmlaq_n_s16(int16x8_t __p0,int16x8_t __p1,int16_t __p2)14688 __ai int16x8_t vmlaq_n_s16(int16x8_t __p0, int16x8_t __p1, int16_t __p2) {
14689 int16x8_t __ret;
14690 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
14691 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
14692 __ret = __rev0 + __rev1 * (int16x8_t) {__p2, __p2, __p2, __p2, __p2, __p2, __p2, __p2};
14693 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
14694 return __ret;
14695 }
14696 #endif
14697
14698 #ifdef __LITTLE_ENDIAN__
vmla_n_u32(uint32x2_t __p0,uint32x2_t __p1,uint32_t __p2)14699 __ai uint32x2_t vmla_n_u32(uint32x2_t __p0, uint32x2_t __p1, uint32_t __p2) {
14700 uint32x2_t __ret;
14701 __ret = __p0 + __p1 * (uint32x2_t) {__p2, __p2};
14702 return __ret;
14703 }
14704 #else
vmla_n_u32(uint32x2_t __p0,uint32x2_t __p1,uint32_t __p2)14705 __ai uint32x2_t vmla_n_u32(uint32x2_t __p0, uint32x2_t __p1, uint32_t __p2) {
14706 uint32x2_t __ret;
14707 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
14708 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
14709 __ret = __rev0 + __rev1 * (uint32x2_t) {__p2, __p2};
14710 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
14711 return __ret;
14712 }
14713 #endif
14714
14715 #ifdef __LITTLE_ENDIAN__
vmla_n_u16(uint16x4_t __p0,uint16x4_t __p1,uint16_t __p2)14716 __ai uint16x4_t vmla_n_u16(uint16x4_t __p0, uint16x4_t __p1, uint16_t __p2) {
14717 uint16x4_t __ret;
14718 __ret = __p0 + __p1 * (uint16x4_t) {__p2, __p2, __p2, __p2};
14719 return __ret;
14720 }
14721 #else
vmla_n_u16(uint16x4_t __p0,uint16x4_t __p1,uint16_t __p2)14722 __ai uint16x4_t vmla_n_u16(uint16x4_t __p0, uint16x4_t __p1, uint16_t __p2) {
14723 uint16x4_t __ret;
14724 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
14725 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
14726 __ret = __rev0 + __rev1 * (uint16x4_t) {__p2, __p2, __p2, __p2};
14727 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
14728 return __ret;
14729 }
14730 #endif
14731
14732 #ifdef __LITTLE_ENDIAN__
vmla_n_f32(float32x2_t __p0,float32x2_t __p1,float32_t __p2)14733 __ai float32x2_t vmla_n_f32(float32x2_t __p0, float32x2_t __p1, float32_t __p2) {
14734 float32x2_t __ret;
14735 __ret = __p0 + __p1 * (float32x2_t) {__p2, __p2};
14736 return __ret;
14737 }
14738 #else
vmla_n_f32(float32x2_t __p0,float32x2_t __p1,float32_t __p2)14739 __ai float32x2_t vmla_n_f32(float32x2_t __p0, float32x2_t __p1, float32_t __p2) {
14740 float32x2_t __ret;
14741 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
14742 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
14743 __ret = __rev0 + __rev1 * (float32x2_t) {__p2, __p2};
14744 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
14745 return __ret;
14746 }
14747 #endif
14748
14749 #ifdef __LITTLE_ENDIAN__
vmla_n_s32(int32x2_t __p0,int32x2_t __p1,int32_t __p2)14750 __ai int32x2_t vmla_n_s32(int32x2_t __p0, int32x2_t __p1, int32_t __p2) {
14751 int32x2_t __ret;
14752 __ret = __p0 + __p1 * (int32x2_t) {__p2, __p2};
14753 return __ret;
14754 }
14755 #else
vmla_n_s32(int32x2_t __p0,int32x2_t __p1,int32_t __p2)14756 __ai int32x2_t vmla_n_s32(int32x2_t __p0, int32x2_t __p1, int32_t __p2) {
14757 int32x2_t __ret;
14758 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
14759 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
14760 __ret = __rev0 + __rev1 * (int32x2_t) {__p2, __p2};
14761 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
14762 return __ret;
14763 }
14764 #endif
14765
14766 #ifdef __LITTLE_ENDIAN__
vmla_n_s16(int16x4_t __p0,int16x4_t __p1,int16_t __p2)14767 __ai int16x4_t vmla_n_s16(int16x4_t __p0, int16x4_t __p1, int16_t __p2) {
14768 int16x4_t __ret;
14769 __ret = __p0 + __p1 * (int16x4_t) {__p2, __p2, __p2, __p2};
14770 return __ret;
14771 }
14772 #else
vmla_n_s16(int16x4_t __p0,int16x4_t __p1,int16_t __p2)14773 __ai int16x4_t vmla_n_s16(int16x4_t __p0, int16x4_t __p1, int16_t __p2) {
14774 int16x4_t __ret;
14775 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
14776 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
14777 __ret = __rev0 + __rev1 * (int16x4_t) {__p2, __p2, __p2, __p2};
14778 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
14779 return __ret;
14780 }
14781 #endif
14782
14783 #ifdef __LITTLE_ENDIAN__
vmlsq_u8(uint8x16_t __p0,uint8x16_t __p1,uint8x16_t __p2)14784 __ai uint8x16_t vmlsq_u8(uint8x16_t __p0, uint8x16_t __p1, uint8x16_t __p2) {
14785 uint8x16_t __ret;
14786 __ret = __p0 - __p1 * __p2;
14787 return __ret;
14788 }
14789 #else
vmlsq_u8(uint8x16_t __p0,uint8x16_t __p1,uint8x16_t __p2)14790 __ai uint8x16_t vmlsq_u8(uint8x16_t __p0, uint8x16_t __p1, uint8x16_t __p2) {
14791 uint8x16_t __ret;
14792 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
14793 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
14794 uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
14795 __ret = __rev0 - __rev1 * __rev2;
14796 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
14797 return __ret;
14798 }
14799 #endif
14800
14801 #ifdef __LITTLE_ENDIAN__
vmlsq_u32(uint32x4_t __p0,uint32x4_t __p1,uint32x4_t __p2)14802 __ai uint32x4_t vmlsq_u32(uint32x4_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
14803 uint32x4_t __ret;
14804 __ret = __p0 - __p1 * __p2;
14805 return __ret;
14806 }
14807 #else
vmlsq_u32(uint32x4_t __p0,uint32x4_t __p1,uint32x4_t __p2)14808 __ai uint32x4_t vmlsq_u32(uint32x4_t __p0, uint32x4_t __p1, uint32x4_t __p2) {
14809 uint32x4_t __ret;
14810 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
14811 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
14812 uint32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
14813 __ret = __rev0 - __rev1 * __rev2;
14814 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
14815 return __ret;
14816 }
14817 #endif
14818
14819 #ifdef __LITTLE_ENDIAN__
vmlsq_u16(uint16x8_t __p0,uint16x8_t __p1,uint16x8_t __p2)14820 __ai uint16x8_t vmlsq_u16(uint16x8_t __p0, uint16x8_t __p1, uint16x8_t __p2) {
14821 uint16x8_t __ret;
14822 __ret = __p0 - __p1 * __p2;
14823 return __ret;
14824 }
14825 #else
vmlsq_u16(uint16x8_t __p0,uint16x8_t __p1,uint16x8_t __p2)14826 __ai uint16x8_t vmlsq_u16(uint16x8_t __p0, uint16x8_t __p1, uint16x8_t __p2) {
14827 uint16x8_t __ret;
14828 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
14829 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
14830 uint16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
14831 __ret = __rev0 - __rev1 * __rev2;
14832 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
14833 return __ret;
14834 }
14835 #endif
14836
14837 #ifdef __LITTLE_ENDIAN__
vmlsq_s8(int8x16_t __p0,int8x16_t __p1,int8x16_t __p2)14838 __ai int8x16_t vmlsq_s8(int8x16_t __p0, int8x16_t __p1, int8x16_t __p2) {
14839 int8x16_t __ret;
14840 __ret = __p0 - __p1 * __p2;
14841 return __ret;
14842 }
14843 #else
vmlsq_s8(int8x16_t __p0,int8x16_t __p1,int8x16_t __p2)14844 __ai int8x16_t vmlsq_s8(int8x16_t __p0, int8x16_t __p1, int8x16_t __p2) {
14845 int8x16_t __ret;
14846 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
14847 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
14848 int8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
14849 __ret = __rev0 - __rev1 * __rev2;
14850 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
14851 return __ret;
14852 }
14853 #endif
14854
14855 #ifdef __LITTLE_ENDIAN__
vmlsq_f32(float32x4_t __p0,float32x4_t __p1,float32x4_t __p2)14856 __ai float32x4_t vmlsq_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
14857 float32x4_t __ret;
14858 __ret = __p0 - __p1 * __p2;
14859 return __ret;
14860 }
14861 #else
vmlsq_f32(float32x4_t __p0,float32x4_t __p1,float32x4_t __p2)14862 __ai float32x4_t vmlsq_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
14863 float32x4_t __ret;
14864 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
14865 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
14866 float32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
14867 __ret = __rev0 - __rev1 * __rev2;
14868 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
14869 return __ret;
14870 }
14871 #endif
14872
14873 #ifdef __LITTLE_ENDIAN__
vmlsq_s32(int32x4_t __p0,int32x4_t __p1,int32x4_t __p2)14874 __ai int32x4_t vmlsq_s32(int32x4_t __p0, int32x4_t __p1, int32x4_t __p2) {
14875 int32x4_t __ret;
14876 __ret = __p0 - __p1 * __p2;
14877 return __ret;
14878 }
14879 #else
vmlsq_s32(int32x4_t __p0,int32x4_t __p1,int32x4_t __p2)14880 __ai int32x4_t vmlsq_s32(int32x4_t __p0, int32x4_t __p1, int32x4_t __p2) {
14881 int32x4_t __ret;
14882 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
14883 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
14884 int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
14885 __ret = __rev0 - __rev1 * __rev2;
14886 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
14887 return __ret;
14888 }
14889 #endif
14890
14891 #ifdef __LITTLE_ENDIAN__
vmlsq_s16(int16x8_t __p0,int16x8_t __p1,int16x8_t __p2)14892 __ai int16x8_t vmlsq_s16(int16x8_t __p0, int16x8_t __p1, int16x8_t __p2) {
14893 int16x8_t __ret;
14894 __ret = __p0 - __p1 * __p2;
14895 return __ret;
14896 }
14897 #else
vmlsq_s16(int16x8_t __p0,int16x8_t __p1,int16x8_t __p2)14898 __ai int16x8_t vmlsq_s16(int16x8_t __p0, int16x8_t __p1, int16x8_t __p2) {
14899 int16x8_t __ret;
14900 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
14901 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
14902 int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
14903 __ret = __rev0 - __rev1 * __rev2;
14904 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
14905 return __ret;
14906 }
14907 #endif
14908
14909 #ifdef __LITTLE_ENDIAN__
vmls_u8(uint8x8_t __p0,uint8x8_t __p1,uint8x8_t __p2)14910 __ai uint8x8_t vmls_u8(uint8x8_t __p0, uint8x8_t __p1, uint8x8_t __p2) {
14911 uint8x8_t __ret;
14912 __ret = __p0 - __p1 * __p2;
14913 return __ret;
14914 }
14915 #else
vmls_u8(uint8x8_t __p0,uint8x8_t __p1,uint8x8_t __p2)14916 __ai uint8x8_t vmls_u8(uint8x8_t __p0, uint8x8_t __p1, uint8x8_t __p2) {
14917 uint8x8_t __ret;
14918 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
14919 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
14920 uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
14921 __ret = __rev0 - __rev1 * __rev2;
14922 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
14923 return __ret;
14924 }
14925 #endif
14926
14927 #ifdef __LITTLE_ENDIAN__
vmls_u32(uint32x2_t __p0,uint32x2_t __p1,uint32x2_t __p2)14928 __ai uint32x2_t vmls_u32(uint32x2_t __p0, uint32x2_t __p1, uint32x2_t __p2) {
14929 uint32x2_t __ret;
14930 __ret = __p0 - __p1 * __p2;
14931 return __ret;
14932 }
14933 #else
vmls_u32(uint32x2_t __p0,uint32x2_t __p1,uint32x2_t __p2)14934 __ai uint32x2_t vmls_u32(uint32x2_t __p0, uint32x2_t __p1, uint32x2_t __p2) {
14935 uint32x2_t __ret;
14936 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
14937 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
14938 uint32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
14939 __ret = __rev0 - __rev1 * __rev2;
14940 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
14941 return __ret;
14942 }
14943 #endif
14944
14945 #ifdef __LITTLE_ENDIAN__
vmls_u16(uint16x4_t __p0,uint16x4_t __p1,uint16x4_t __p2)14946 __ai uint16x4_t vmls_u16(uint16x4_t __p0, uint16x4_t __p1, uint16x4_t __p2) {
14947 uint16x4_t __ret;
14948 __ret = __p0 - __p1 * __p2;
14949 return __ret;
14950 }
14951 #else
vmls_u16(uint16x4_t __p0,uint16x4_t __p1,uint16x4_t __p2)14952 __ai uint16x4_t vmls_u16(uint16x4_t __p0, uint16x4_t __p1, uint16x4_t __p2) {
14953 uint16x4_t __ret;
14954 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
14955 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
14956 uint16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
14957 __ret = __rev0 - __rev1 * __rev2;
14958 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
14959 return __ret;
14960 }
14961 #endif
14962
14963 #ifdef __LITTLE_ENDIAN__
vmls_s8(int8x8_t __p0,int8x8_t __p1,int8x8_t __p2)14964 __ai int8x8_t vmls_s8(int8x8_t __p0, int8x8_t __p1, int8x8_t __p2) {
14965 int8x8_t __ret;
14966 __ret = __p0 - __p1 * __p2;
14967 return __ret;
14968 }
14969 #else
vmls_s8(int8x8_t __p0,int8x8_t __p1,int8x8_t __p2)14970 __ai int8x8_t vmls_s8(int8x8_t __p0, int8x8_t __p1, int8x8_t __p2) {
14971 int8x8_t __ret;
14972 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
14973 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
14974 int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
14975 __ret = __rev0 - __rev1 * __rev2;
14976 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
14977 return __ret;
14978 }
14979 #endif
14980
14981 #ifdef __LITTLE_ENDIAN__
vmls_f32(float32x2_t __p0,float32x2_t __p1,float32x2_t __p2)14982 __ai float32x2_t vmls_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
14983 float32x2_t __ret;
14984 __ret = __p0 - __p1 * __p2;
14985 return __ret;
14986 }
14987 #else
vmls_f32(float32x2_t __p0,float32x2_t __p1,float32x2_t __p2)14988 __ai float32x2_t vmls_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
14989 float32x2_t __ret;
14990 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
14991 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
14992 float32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
14993 __ret = __rev0 - __rev1 * __rev2;
14994 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
14995 return __ret;
14996 }
14997 #endif
14998
14999 #ifdef __LITTLE_ENDIAN__
vmls_s32(int32x2_t __p0,int32x2_t __p1,int32x2_t __p2)15000 __ai int32x2_t vmls_s32(int32x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
15001 int32x2_t __ret;
15002 __ret = __p0 - __p1 * __p2;
15003 return __ret;
15004 }
15005 #else
vmls_s32(int32x2_t __p0,int32x2_t __p1,int32x2_t __p2)15006 __ai int32x2_t vmls_s32(int32x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
15007 int32x2_t __ret;
15008 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
15009 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
15010 int32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
15011 __ret = __rev0 - __rev1 * __rev2;
15012 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
15013 return __ret;
15014 }
15015 #endif
15016
15017 #ifdef __LITTLE_ENDIAN__
vmls_s16(int16x4_t __p0,int16x4_t __p1,int16x4_t __p2)15018 __ai int16x4_t vmls_s16(int16x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
15019 int16x4_t __ret;
15020 __ret = __p0 - __p1 * __p2;
15021 return __ret;
15022 }
15023 #else
vmls_s16(int16x4_t __p0,int16x4_t __p1,int16x4_t __p2)15024 __ai int16x4_t vmls_s16(int16x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
15025 int16x4_t __ret;
15026 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
15027 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
15028 int16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
15029 __ret = __rev0 - __rev1 * __rev2;
15030 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
15031 return __ret;
15032 }
15033 #endif
15034
15035 #ifdef __LITTLE_ENDIAN__
15036 #define vmlsq_lane_u32(__p0_66, __p1_66, __p2_66, __p3_66) __extension__ ({ \
15037 uint32x4_t __ret_66; \
15038 uint32x4_t __s0_66 = __p0_66; \
15039 uint32x4_t __s1_66 = __p1_66; \
15040 uint32x2_t __s2_66 = __p2_66; \
15041 __ret_66 = __s0_66 - __s1_66 * splatq_lane_u32(__s2_66, __p3_66); \
15042 __ret_66; \
15043 })
15044 #else
15045 #define vmlsq_lane_u32(__p0_67, __p1_67, __p2_67, __p3_67) __extension__ ({ \
15046 uint32x4_t __ret_67; \
15047 uint32x4_t __s0_67 = __p0_67; \
15048 uint32x4_t __s1_67 = __p1_67; \
15049 uint32x2_t __s2_67 = __p2_67; \
15050 uint32x4_t __rev0_67; __rev0_67 = __builtin_shufflevector(__s0_67, __s0_67, 3, 2, 1, 0); \
15051 uint32x4_t __rev1_67; __rev1_67 = __builtin_shufflevector(__s1_67, __s1_67, 3, 2, 1, 0); \
15052 uint32x2_t __rev2_67; __rev2_67 = __builtin_shufflevector(__s2_67, __s2_67, 1, 0); \
15053 __ret_67 = __rev0_67 - __rev1_67 * __noswap_splatq_lane_u32(__rev2_67, __p3_67); \
15054 __ret_67 = __builtin_shufflevector(__ret_67, __ret_67, 3, 2, 1, 0); \
15055 __ret_67; \
15056 })
15057 #endif
15058
15059 #ifdef __LITTLE_ENDIAN__
15060 #define vmlsq_lane_u16(__p0_68, __p1_68, __p2_68, __p3_68) __extension__ ({ \
15061 uint16x8_t __ret_68; \
15062 uint16x8_t __s0_68 = __p0_68; \
15063 uint16x8_t __s1_68 = __p1_68; \
15064 uint16x4_t __s2_68 = __p2_68; \
15065 __ret_68 = __s0_68 - __s1_68 * splatq_lane_u16(__s2_68, __p3_68); \
15066 __ret_68; \
15067 })
15068 #else
15069 #define vmlsq_lane_u16(__p0_69, __p1_69, __p2_69, __p3_69) __extension__ ({ \
15070 uint16x8_t __ret_69; \
15071 uint16x8_t __s0_69 = __p0_69; \
15072 uint16x8_t __s1_69 = __p1_69; \
15073 uint16x4_t __s2_69 = __p2_69; \
15074 uint16x8_t __rev0_69; __rev0_69 = __builtin_shufflevector(__s0_69, __s0_69, 7, 6, 5, 4, 3, 2, 1, 0); \
15075 uint16x8_t __rev1_69; __rev1_69 = __builtin_shufflevector(__s1_69, __s1_69, 7, 6, 5, 4, 3, 2, 1, 0); \
15076 uint16x4_t __rev2_69; __rev2_69 = __builtin_shufflevector(__s2_69, __s2_69, 3, 2, 1, 0); \
15077 __ret_69 = __rev0_69 - __rev1_69 * __noswap_splatq_lane_u16(__rev2_69, __p3_69); \
15078 __ret_69 = __builtin_shufflevector(__ret_69, __ret_69, 7, 6, 5, 4, 3, 2, 1, 0); \
15079 __ret_69; \
15080 })
15081 #endif
15082
15083 #ifdef __LITTLE_ENDIAN__
15084 #define vmlsq_lane_f32(__p0_70, __p1_70, __p2_70, __p3_70) __extension__ ({ \
15085 float32x4_t __ret_70; \
15086 float32x4_t __s0_70 = __p0_70; \
15087 float32x4_t __s1_70 = __p1_70; \
15088 float32x2_t __s2_70 = __p2_70; \
15089 __ret_70 = __s0_70 - __s1_70 * splatq_lane_f32(__s2_70, __p3_70); \
15090 __ret_70; \
15091 })
15092 #else
15093 #define vmlsq_lane_f32(__p0_71, __p1_71, __p2_71, __p3_71) __extension__ ({ \
15094 float32x4_t __ret_71; \
15095 float32x4_t __s0_71 = __p0_71; \
15096 float32x4_t __s1_71 = __p1_71; \
15097 float32x2_t __s2_71 = __p2_71; \
15098 float32x4_t __rev0_71; __rev0_71 = __builtin_shufflevector(__s0_71, __s0_71, 3, 2, 1, 0); \
15099 float32x4_t __rev1_71; __rev1_71 = __builtin_shufflevector(__s1_71, __s1_71, 3, 2, 1, 0); \
15100 float32x2_t __rev2_71; __rev2_71 = __builtin_shufflevector(__s2_71, __s2_71, 1, 0); \
15101 __ret_71 = __rev0_71 - __rev1_71 * __noswap_splatq_lane_f32(__rev2_71, __p3_71); \
15102 __ret_71 = __builtin_shufflevector(__ret_71, __ret_71, 3, 2, 1, 0); \
15103 __ret_71; \
15104 })
15105 #endif
15106
15107 #ifdef __LITTLE_ENDIAN__
15108 #define vmlsq_lane_s32(__p0_72, __p1_72, __p2_72, __p3_72) __extension__ ({ \
15109 int32x4_t __ret_72; \
15110 int32x4_t __s0_72 = __p0_72; \
15111 int32x4_t __s1_72 = __p1_72; \
15112 int32x2_t __s2_72 = __p2_72; \
15113 __ret_72 = __s0_72 - __s1_72 * splatq_lane_s32(__s2_72, __p3_72); \
15114 __ret_72; \
15115 })
15116 #else
15117 #define vmlsq_lane_s32(__p0_73, __p1_73, __p2_73, __p3_73) __extension__ ({ \
15118 int32x4_t __ret_73; \
15119 int32x4_t __s0_73 = __p0_73; \
15120 int32x4_t __s1_73 = __p1_73; \
15121 int32x2_t __s2_73 = __p2_73; \
15122 int32x4_t __rev0_73; __rev0_73 = __builtin_shufflevector(__s0_73, __s0_73, 3, 2, 1, 0); \
15123 int32x4_t __rev1_73; __rev1_73 = __builtin_shufflevector(__s1_73, __s1_73, 3, 2, 1, 0); \
15124 int32x2_t __rev2_73; __rev2_73 = __builtin_shufflevector(__s2_73, __s2_73, 1, 0); \
15125 __ret_73 = __rev0_73 - __rev1_73 * __noswap_splatq_lane_s32(__rev2_73, __p3_73); \
15126 __ret_73 = __builtin_shufflevector(__ret_73, __ret_73, 3, 2, 1, 0); \
15127 __ret_73; \
15128 })
15129 #endif
15130
15131 #ifdef __LITTLE_ENDIAN__
15132 #define vmlsq_lane_s16(__p0_74, __p1_74, __p2_74, __p3_74) __extension__ ({ \
15133 int16x8_t __ret_74; \
15134 int16x8_t __s0_74 = __p0_74; \
15135 int16x8_t __s1_74 = __p1_74; \
15136 int16x4_t __s2_74 = __p2_74; \
15137 __ret_74 = __s0_74 - __s1_74 * splatq_lane_s16(__s2_74, __p3_74); \
15138 __ret_74; \
15139 })
15140 #else
15141 #define vmlsq_lane_s16(__p0_75, __p1_75, __p2_75, __p3_75) __extension__ ({ \
15142 int16x8_t __ret_75; \
15143 int16x8_t __s0_75 = __p0_75; \
15144 int16x8_t __s1_75 = __p1_75; \
15145 int16x4_t __s2_75 = __p2_75; \
15146 int16x8_t __rev0_75; __rev0_75 = __builtin_shufflevector(__s0_75, __s0_75, 7, 6, 5, 4, 3, 2, 1, 0); \
15147 int16x8_t __rev1_75; __rev1_75 = __builtin_shufflevector(__s1_75, __s1_75, 7, 6, 5, 4, 3, 2, 1, 0); \
15148 int16x4_t __rev2_75; __rev2_75 = __builtin_shufflevector(__s2_75, __s2_75, 3, 2, 1, 0); \
15149 __ret_75 = __rev0_75 - __rev1_75 * __noswap_splatq_lane_s16(__rev2_75, __p3_75); \
15150 __ret_75 = __builtin_shufflevector(__ret_75, __ret_75, 7, 6, 5, 4, 3, 2, 1, 0); \
15151 __ret_75; \
15152 })
15153 #endif
15154
15155 #ifdef __LITTLE_ENDIAN__
15156 #define vmls_lane_u32(__p0_76, __p1_76, __p2_76, __p3_76) __extension__ ({ \
15157 uint32x2_t __ret_76; \
15158 uint32x2_t __s0_76 = __p0_76; \
15159 uint32x2_t __s1_76 = __p1_76; \
15160 uint32x2_t __s2_76 = __p2_76; \
15161 __ret_76 = __s0_76 - __s1_76 * splat_lane_u32(__s2_76, __p3_76); \
15162 __ret_76; \
15163 })
15164 #else
15165 #define vmls_lane_u32(__p0_77, __p1_77, __p2_77, __p3_77) __extension__ ({ \
15166 uint32x2_t __ret_77; \
15167 uint32x2_t __s0_77 = __p0_77; \
15168 uint32x2_t __s1_77 = __p1_77; \
15169 uint32x2_t __s2_77 = __p2_77; \
15170 uint32x2_t __rev0_77; __rev0_77 = __builtin_shufflevector(__s0_77, __s0_77, 1, 0); \
15171 uint32x2_t __rev1_77; __rev1_77 = __builtin_shufflevector(__s1_77, __s1_77, 1, 0); \
15172 uint32x2_t __rev2_77; __rev2_77 = __builtin_shufflevector(__s2_77, __s2_77, 1, 0); \
15173 __ret_77 = __rev0_77 - __rev1_77 * __noswap_splat_lane_u32(__rev2_77, __p3_77); \
15174 __ret_77 = __builtin_shufflevector(__ret_77, __ret_77, 1, 0); \
15175 __ret_77; \
15176 })
15177 #endif
15178
15179 #ifdef __LITTLE_ENDIAN__
15180 #define vmls_lane_u16(__p0_78, __p1_78, __p2_78, __p3_78) __extension__ ({ \
15181 uint16x4_t __ret_78; \
15182 uint16x4_t __s0_78 = __p0_78; \
15183 uint16x4_t __s1_78 = __p1_78; \
15184 uint16x4_t __s2_78 = __p2_78; \
15185 __ret_78 = __s0_78 - __s1_78 * splat_lane_u16(__s2_78, __p3_78); \
15186 __ret_78; \
15187 })
15188 #else
15189 #define vmls_lane_u16(__p0_79, __p1_79, __p2_79, __p3_79) __extension__ ({ \
15190 uint16x4_t __ret_79; \
15191 uint16x4_t __s0_79 = __p0_79; \
15192 uint16x4_t __s1_79 = __p1_79; \
15193 uint16x4_t __s2_79 = __p2_79; \
15194 uint16x4_t __rev0_79; __rev0_79 = __builtin_shufflevector(__s0_79, __s0_79, 3, 2, 1, 0); \
15195 uint16x4_t __rev1_79; __rev1_79 = __builtin_shufflevector(__s1_79, __s1_79, 3, 2, 1, 0); \
15196 uint16x4_t __rev2_79; __rev2_79 = __builtin_shufflevector(__s2_79, __s2_79, 3, 2, 1, 0); \
15197 __ret_79 = __rev0_79 - __rev1_79 * __noswap_splat_lane_u16(__rev2_79, __p3_79); \
15198 __ret_79 = __builtin_shufflevector(__ret_79, __ret_79, 3, 2, 1, 0); \
15199 __ret_79; \
15200 })
15201 #endif
15202
15203 #ifdef __LITTLE_ENDIAN__
15204 #define vmls_lane_f32(__p0_80, __p1_80, __p2_80, __p3_80) __extension__ ({ \
15205 float32x2_t __ret_80; \
15206 float32x2_t __s0_80 = __p0_80; \
15207 float32x2_t __s1_80 = __p1_80; \
15208 float32x2_t __s2_80 = __p2_80; \
15209 __ret_80 = __s0_80 - __s1_80 * splat_lane_f32(__s2_80, __p3_80); \
15210 __ret_80; \
15211 })
15212 #else
15213 #define vmls_lane_f32(__p0_81, __p1_81, __p2_81, __p3_81) __extension__ ({ \
15214 float32x2_t __ret_81; \
15215 float32x2_t __s0_81 = __p0_81; \
15216 float32x2_t __s1_81 = __p1_81; \
15217 float32x2_t __s2_81 = __p2_81; \
15218 float32x2_t __rev0_81; __rev0_81 = __builtin_shufflevector(__s0_81, __s0_81, 1, 0); \
15219 float32x2_t __rev1_81; __rev1_81 = __builtin_shufflevector(__s1_81, __s1_81, 1, 0); \
15220 float32x2_t __rev2_81; __rev2_81 = __builtin_shufflevector(__s2_81, __s2_81, 1, 0); \
15221 __ret_81 = __rev0_81 - __rev1_81 * __noswap_splat_lane_f32(__rev2_81, __p3_81); \
15222 __ret_81 = __builtin_shufflevector(__ret_81, __ret_81, 1, 0); \
15223 __ret_81; \
15224 })
15225 #endif
15226
15227 #ifdef __LITTLE_ENDIAN__
15228 #define vmls_lane_s32(__p0_82, __p1_82, __p2_82, __p3_82) __extension__ ({ \
15229 int32x2_t __ret_82; \
15230 int32x2_t __s0_82 = __p0_82; \
15231 int32x2_t __s1_82 = __p1_82; \
15232 int32x2_t __s2_82 = __p2_82; \
15233 __ret_82 = __s0_82 - __s1_82 * splat_lane_s32(__s2_82, __p3_82); \
15234 __ret_82; \
15235 })
15236 #else
15237 #define vmls_lane_s32(__p0_83, __p1_83, __p2_83, __p3_83) __extension__ ({ \
15238 int32x2_t __ret_83; \
15239 int32x2_t __s0_83 = __p0_83; \
15240 int32x2_t __s1_83 = __p1_83; \
15241 int32x2_t __s2_83 = __p2_83; \
15242 int32x2_t __rev0_83; __rev0_83 = __builtin_shufflevector(__s0_83, __s0_83, 1, 0); \
15243 int32x2_t __rev1_83; __rev1_83 = __builtin_shufflevector(__s1_83, __s1_83, 1, 0); \
15244 int32x2_t __rev2_83; __rev2_83 = __builtin_shufflevector(__s2_83, __s2_83, 1, 0); \
15245 __ret_83 = __rev0_83 - __rev1_83 * __noswap_splat_lane_s32(__rev2_83, __p3_83); \
15246 __ret_83 = __builtin_shufflevector(__ret_83, __ret_83, 1, 0); \
15247 __ret_83; \
15248 })
15249 #endif
15250
15251 #ifdef __LITTLE_ENDIAN__
15252 #define vmls_lane_s16(__p0_84, __p1_84, __p2_84, __p3_84) __extension__ ({ \
15253 int16x4_t __ret_84; \
15254 int16x4_t __s0_84 = __p0_84; \
15255 int16x4_t __s1_84 = __p1_84; \
15256 int16x4_t __s2_84 = __p2_84; \
15257 __ret_84 = __s0_84 - __s1_84 * splat_lane_s16(__s2_84, __p3_84); \
15258 __ret_84; \
15259 })
15260 #else
15261 #define vmls_lane_s16(__p0_85, __p1_85, __p2_85, __p3_85) __extension__ ({ \
15262 int16x4_t __ret_85; \
15263 int16x4_t __s0_85 = __p0_85; \
15264 int16x4_t __s1_85 = __p1_85; \
15265 int16x4_t __s2_85 = __p2_85; \
15266 int16x4_t __rev0_85; __rev0_85 = __builtin_shufflevector(__s0_85, __s0_85, 3, 2, 1, 0); \
15267 int16x4_t __rev1_85; __rev1_85 = __builtin_shufflevector(__s1_85, __s1_85, 3, 2, 1, 0); \
15268 int16x4_t __rev2_85; __rev2_85 = __builtin_shufflevector(__s2_85, __s2_85, 3, 2, 1, 0); \
15269 __ret_85 = __rev0_85 - __rev1_85 * __noswap_splat_lane_s16(__rev2_85, __p3_85); \
15270 __ret_85 = __builtin_shufflevector(__ret_85, __ret_85, 3, 2, 1, 0); \
15271 __ret_85; \
15272 })
15273 #endif
15274
15275 #ifdef __LITTLE_ENDIAN__
vmlsq_n_u32(uint32x4_t __p0,uint32x4_t __p1,uint32_t __p2)15276 __ai uint32x4_t vmlsq_n_u32(uint32x4_t __p0, uint32x4_t __p1, uint32_t __p2) {
15277 uint32x4_t __ret;
15278 __ret = __p0 - __p1 * (uint32x4_t) {__p2, __p2, __p2, __p2};
15279 return __ret;
15280 }
15281 #else
vmlsq_n_u32(uint32x4_t __p0,uint32x4_t __p1,uint32_t __p2)15282 __ai uint32x4_t vmlsq_n_u32(uint32x4_t __p0, uint32x4_t __p1, uint32_t __p2) {
15283 uint32x4_t __ret;
15284 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
15285 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
15286 __ret = __rev0 - __rev1 * (uint32x4_t) {__p2, __p2, __p2, __p2};
15287 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
15288 return __ret;
15289 }
15290 #endif
15291
15292 #ifdef __LITTLE_ENDIAN__
vmlsq_n_u16(uint16x8_t __p0,uint16x8_t __p1,uint16_t __p2)15293 __ai uint16x8_t vmlsq_n_u16(uint16x8_t __p0, uint16x8_t __p1, uint16_t __p2) {
15294 uint16x8_t __ret;
15295 __ret = __p0 - __p1 * (uint16x8_t) {__p2, __p2, __p2, __p2, __p2, __p2, __p2, __p2};
15296 return __ret;
15297 }
15298 #else
vmlsq_n_u16(uint16x8_t __p0,uint16x8_t __p1,uint16_t __p2)15299 __ai uint16x8_t vmlsq_n_u16(uint16x8_t __p0, uint16x8_t __p1, uint16_t __p2) {
15300 uint16x8_t __ret;
15301 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
15302 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
15303 __ret = __rev0 - __rev1 * (uint16x8_t) {__p2, __p2, __p2, __p2, __p2, __p2, __p2, __p2};
15304 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
15305 return __ret;
15306 }
15307 #endif
15308
15309 #ifdef __LITTLE_ENDIAN__
vmlsq_n_f32(float32x4_t __p0,float32x4_t __p1,float32_t __p2)15310 __ai float32x4_t vmlsq_n_f32(float32x4_t __p0, float32x4_t __p1, float32_t __p2) {
15311 float32x4_t __ret;
15312 __ret = __p0 - __p1 * (float32x4_t) {__p2, __p2, __p2, __p2};
15313 return __ret;
15314 }
15315 #else
vmlsq_n_f32(float32x4_t __p0,float32x4_t __p1,float32_t __p2)15316 __ai float32x4_t vmlsq_n_f32(float32x4_t __p0, float32x4_t __p1, float32_t __p2) {
15317 float32x4_t __ret;
15318 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
15319 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
15320 __ret = __rev0 - __rev1 * (float32x4_t) {__p2, __p2, __p2, __p2};
15321 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
15322 return __ret;
15323 }
15324 #endif
15325
15326 #ifdef __LITTLE_ENDIAN__
vmlsq_n_s32(int32x4_t __p0,int32x4_t __p1,int32_t __p2)15327 __ai int32x4_t vmlsq_n_s32(int32x4_t __p0, int32x4_t __p1, int32_t __p2) {
15328 int32x4_t __ret;
15329 __ret = __p0 - __p1 * (int32x4_t) {__p2, __p2, __p2, __p2};
15330 return __ret;
15331 }
15332 #else
vmlsq_n_s32(int32x4_t __p0,int32x4_t __p1,int32_t __p2)15333 __ai int32x4_t vmlsq_n_s32(int32x4_t __p0, int32x4_t __p1, int32_t __p2) {
15334 int32x4_t __ret;
15335 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
15336 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
15337 __ret = __rev0 - __rev1 * (int32x4_t) {__p2, __p2, __p2, __p2};
15338 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
15339 return __ret;
15340 }
15341 #endif
15342
15343 #ifdef __LITTLE_ENDIAN__
vmlsq_n_s16(int16x8_t __p0,int16x8_t __p1,int16_t __p2)15344 __ai int16x8_t vmlsq_n_s16(int16x8_t __p0, int16x8_t __p1, int16_t __p2) {
15345 int16x8_t __ret;
15346 __ret = __p0 - __p1 * (int16x8_t) {__p2, __p2, __p2, __p2, __p2, __p2, __p2, __p2};
15347 return __ret;
15348 }
15349 #else
vmlsq_n_s16(int16x8_t __p0,int16x8_t __p1,int16_t __p2)15350 __ai int16x8_t vmlsq_n_s16(int16x8_t __p0, int16x8_t __p1, int16_t __p2) {
15351 int16x8_t __ret;
15352 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
15353 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
15354 __ret = __rev0 - __rev1 * (int16x8_t) {__p2, __p2, __p2, __p2, __p2, __p2, __p2, __p2};
15355 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
15356 return __ret;
15357 }
15358 #endif
15359
15360 #ifdef __LITTLE_ENDIAN__
vmls_n_u32(uint32x2_t __p0,uint32x2_t __p1,uint32_t __p2)15361 __ai uint32x2_t vmls_n_u32(uint32x2_t __p0, uint32x2_t __p1, uint32_t __p2) {
15362 uint32x2_t __ret;
15363 __ret = __p0 - __p1 * (uint32x2_t) {__p2, __p2};
15364 return __ret;
15365 }
15366 #else
vmls_n_u32(uint32x2_t __p0,uint32x2_t __p1,uint32_t __p2)15367 __ai uint32x2_t vmls_n_u32(uint32x2_t __p0, uint32x2_t __p1, uint32_t __p2) {
15368 uint32x2_t __ret;
15369 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
15370 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
15371 __ret = __rev0 - __rev1 * (uint32x2_t) {__p2, __p2};
15372 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
15373 return __ret;
15374 }
15375 #endif
15376
15377 #ifdef __LITTLE_ENDIAN__
vmls_n_u16(uint16x4_t __p0,uint16x4_t __p1,uint16_t __p2)15378 __ai uint16x4_t vmls_n_u16(uint16x4_t __p0, uint16x4_t __p1, uint16_t __p2) {
15379 uint16x4_t __ret;
15380 __ret = __p0 - __p1 * (uint16x4_t) {__p2, __p2, __p2, __p2};
15381 return __ret;
15382 }
15383 #else
vmls_n_u16(uint16x4_t __p0,uint16x4_t __p1,uint16_t __p2)15384 __ai uint16x4_t vmls_n_u16(uint16x4_t __p0, uint16x4_t __p1, uint16_t __p2) {
15385 uint16x4_t __ret;
15386 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
15387 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
15388 __ret = __rev0 - __rev1 * (uint16x4_t) {__p2, __p2, __p2, __p2};
15389 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
15390 return __ret;
15391 }
15392 #endif
15393
15394 #ifdef __LITTLE_ENDIAN__
vmls_n_f32(float32x2_t __p0,float32x2_t __p1,float32_t __p2)15395 __ai float32x2_t vmls_n_f32(float32x2_t __p0, float32x2_t __p1, float32_t __p2) {
15396 float32x2_t __ret;
15397 __ret = __p0 - __p1 * (float32x2_t) {__p2, __p2};
15398 return __ret;
15399 }
15400 #else
vmls_n_f32(float32x2_t __p0,float32x2_t __p1,float32_t __p2)15401 __ai float32x2_t vmls_n_f32(float32x2_t __p0, float32x2_t __p1, float32_t __p2) {
15402 float32x2_t __ret;
15403 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
15404 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
15405 __ret = __rev0 - __rev1 * (float32x2_t) {__p2, __p2};
15406 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
15407 return __ret;
15408 }
15409 #endif
15410
15411 #ifdef __LITTLE_ENDIAN__
vmls_n_s32(int32x2_t __p0,int32x2_t __p1,int32_t __p2)15412 __ai int32x2_t vmls_n_s32(int32x2_t __p0, int32x2_t __p1, int32_t __p2) {
15413 int32x2_t __ret;
15414 __ret = __p0 - __p1 * (int32x2_t) {__p2, __p2};
15415 return __ret;
15416 }
15417 #else
vmls_n_s32(int32x2_t __p0,int32x2_t __p1,int32_t __p2)15418 __ai int32x2_t vmls_n_s32(int32x2_t __p0, int32x2_t __p1, int32_t __p2) {
15419 int32x2_t __ret;
15420 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
15421 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
15422 __ret = __rev0 - __rev1 * (int32x2_t) {__p2, __p2};
15423 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
15424 return __ret;
15425 }
15426 #endif
15427
15428 #ifdef __LITTLE_ENDIAN__
vmls_n_s16(int16x4_t __p0,int16x4_t __p1,int16_t __p2)15429 __ai int16x4_t vmls_n_s16(int16x4_t __p0, int16x4_t __p1, int16_t __p2) {
15430 int16x4_t __ret;
15431 __ret = __p0 - __p1 * (int16x4_t) {__p2, __p2, __p2, __p2};
15432 return __ret;
15433 }
15434 #else
vmls_n_s16(int16x4_t __p0,int16x4_t __p1,int16_t __p2)15435 __ai int16x4_t vmls_n_s16(int16x4_t __p0, int16x4_t __p1, int16_t __p2) {
15436 int16x4_t __ret;
15437 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
15438 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
15439 __ret = __rev0 - __rev1 * (int16x4_t) {__p2, __p2, __p2, __p2};
15440 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
15441 return __ret;
15442 }
15443 #endif
15444
15445 #ifdef __LITTLE_ENDIAN__
vmov_n_p8(poly8_t __p0)15446 __ai poly8x8_t vmov_n_p8(poly8_t __p0) {
15447 poly8x8_t __ret;
15448 __ret = (poly8x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
15449 return __ret;
15450 }
15451 #else
vmov_n_p8(poly8_t __p0)15452 __ai poly8x8_t vmov_n_p8(poly8_t __p0) {
15453 poly8x8_t __ret;
15454 __ret = (poly8x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
15455 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
15456 return __ret;
15457 }
15458 #endif
15459
15460 #ifdef __LITTLE_ENDIAN__
vmov_n_p16(poly16_t __p0)15461 __ai poly16x4_t vmov_n_p16(poly16_t __p0) {
15462 poly16x4_t __ret;
15463 __ret = (poly16x4_t) {__p0, __p0, __p0, __p0};
15464 return __ret;
15465 }
15466 #else
vmov_n_p16(poly16_t __p0)15467 __ai poly16x4_t vmov_n_p16(poly16_t __p0) {
15468 poly16x4_t __ret;
15469 __ret = (poly16x4_t) {__p0, __p0, __p0, __p0};
15470 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
15471 return __ret;
15472 }
15473 #endif
15474
15475 #ifdef __LITTLE_ENDIAN__
vmovq_n_p8(poly8_t __p0)15476 __ai poly8x16_t vmovq_n_p8(poly8_t __p0) {
15477 poly8x16_t __ret;
15478 __ret = (poly8x16_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
15479 return __ret;
15480 }
15481 #else
vmovq_n_p8(poly8_t __p0)15482 __ai poly8x16_t vmovq_n_p8(poly8_t __p0) {
15483 poly8x16_t __ret;
15484 __ret = (poly8x16_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
15485 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
15486 return __ret;
15487 }
15488 #endif
15489
15490 #ifdef __LITTLE_ENDIAN__
vmovq_n_p16(poly16_t __p0)15491 __ai poly16x8_t vmovq_n_p16(poly16_t __p0) {
15492 poly16x8_t __ret;
15493 __ret = (poly16x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
15494 return __ret;
15495 }
15496 #else
vmovq_n_p16(poly16_t __p0)15497 __ai poly16x8_t vmovq_n_p16(poly16_t __p0) {
15498 poly16x8_t __ret;
15499 __ret = (poly16x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
15500 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
15501 return __ret;
15502 }
15503 #endif
15504
15505 #ifdef __LITTLE_ENDIAN__
vmovq_n_u8(uint8_t __p0)15506 __ai uint8x16_t vmovq_n_u8(uint8_t __p0) {
15507 uint8x16_t __ret;
15508 __ret = (uint8x16_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
15509 return __ret;
15510 }
15511 #else
vmovq_n_u8(uint8_t __p0)15512 __ai uint8x16_t vmovq_n_u8(uint8_t __p0) {
15513 uint8x16_t __ret;
15514 __ret = (uint8x16_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
15515 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
15516 return __ret;
15517 }
15518 #endif
15519
15520 #ifdef __LITTLE_ENDIAN__
vmovq_n_u32(uint32_t __p0)15521 __ai uint32x4_t vmovq_n_u32(uint32_t __p0) {
15522 uint32x4_t __ret;
15523 __ret = (uint32x4_t) {__p0, __p0, __p0, __p0};
15524 return __ret;
15525 }
15526 #else
vmovq_n_u32(uint32_t __p0)15527 __ai uint32x4_t vmovq_n_u32(uint32_t __p0) {
15528 uint32x4_t __ret;
15529 __ret = (uint32x4_t) {__p0, __p0, __p0, __p0};
15530 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
15531 return __ret;
15532 }
15533 #endif
15534
15535 #ifdef __LITTLE_ENDIAN__
vmovq_n_u64(uint64_t __p0)15536 __ai uint64x2_t vmovq_n_u64(uint64_t __p0) {
15537 uint64x2_t __ret;
15538 __ret = (uint64x2_t) {__p0, __p0};
15539 return __ret;
15540 }
15541 #else
vmovq_n_u64(uint64_t __p0)15542 __ai uint64x2_t vmovq_n_u64(uint64_t __p0) {
15543 uint64x2_t __ret;
15544 __ret = (uint64x2_t) {__p0, __p0};
15545 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
15546 return __ret;
15547 }
15548 #endif
15549
15550 #ifdef __LITTLE_ENDIAN__
vmovq_n_u16(uint16_t __p0)15551 __ai uint16x8_t vmovq_n_u16(uint16_t __p0) {
15552 uint16x8_t __ret;
15553 __ret = (uint16x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
15554 return __ret;
15555 }
15556 #else
vmovq_n_u16(uint16_t __p0)15557 __ai uint16x8_t vmovq_n_u16(uint16_t __p0) {
15558 uint16x8_t __ret;
15559 __ret = (uint16x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
15560 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
15561 return __ret;
15562 }
15563 #endif
15564
15565 #ifdef __LITTLE_ENDIAN__
vmovq_n_s8(int8_t __p0)15566 __ai int8x16_t vmovq_n_s8(int8_t __p0) {
15567 int8x16_t __ret;
15568 __ret = (int8x16_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
15569 return __ret;
15570 }
15571 #else
vmovq_n_s8(int8_t __p0)15572 __ai int8x16_t vmovq_n_s8(int8_t __p0) {
15573 int8x16_t __ret;
15574 __ret = (int8x16_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
15575 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
15576 return __ret;
15577 }
15578 #endif
15579
15580 #ifdef __LITTLE_ENDIAN__
vmovq_n_f32(float32_t __p0)15581 __ai float32x4_t vmovq_n_f32(float32_t __p0) {
15582 float32x4_t __ret;
15583 __ret = (float32x4_t) {__p0, __p0, __p0, __p0};
15584 return __ret;
15585 }
15586 #else
vmovq_n_f32(float32_t __p0)15587 __ai float32x4_t vmovq_n_f32(float32_t __p0) {
15588 float32x4_t __ret;
15589 __ret = (float32x4_t) {__p0, __p0, __p0, __p0};
15590 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
15591 return __ret;
15592 }
15593 #endif
15594
15595 #ifdef __LITTLE_ENDIAN__
15596 #define vmovq_n_f16(__p0) __extension__ ({ \
15597 float16x8_t __ret; \
15598 float16_t __s0 = __p0; \
15599 __ret = (float16x8_t) {__s0, __s0, __s0, __s0, __s0, __s0, __s0, __s0}; \
15600 __ret; \
15601 })
15602 #else
15603 #define vmovq_n_f16(__p0) __extension__ ({ \
15604 float16x8_t __ret; \
15605 float16_t __s0 = __p0; \
15606 __ret = (float16x8_t) {__s0, __s0, __s0, __s0, __s0, __s0, __s0, __s0}; \
15607 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
15608 __ret; \
15609 })
15610 #endif
15611
15612 #ifdef __LITTLE_ENDIAN__
vmovq_n_s32(int32_t __p0)15613 __ai int32x4_t vmovq_n_s32(int32_t __p0) {
15614 int32x4_t __ret;
15615 __ret = (int32x4_t) {__p0, __p0, __p0, __p0};
15616 return __ret;
15617 }
15618 #else
vmovq_n_s32(int32_t __p0)15619 __ai int32x4_t vmovq_n_s32(int32_t __p0) {
15620 int32x4_t __ret;
15621 __ret = (int32x4_t) {__p0, __p0, __p0, __p0};
15622 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
15623 return __ret;
15624 }
15625 #endif
15626
15627 #ifdef __LITTLE_ENDIAN__
vmovq_n_s64(int64_t __p0)15628 __ai int64x2_t vmovq_n_s64(int64_t __p0) {
15629 int64x2_t __ret;
15630 __ret = (int64x2_t) {__p0, __p0};
15631 return __ret;
15632 }
15633 #else
vmovq_n_s64(int64_t __p0)15634 __ai int64x2_t vmovq_n_s64(int64_t __p0) {
15635 int64x2_t __ret;
15636 __ret = (int64x2_t) {__p0, __p0};
15637 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
15638 return __ret;
15639 }
15640 #endif
15641
15642 #ifdef __LITTLE_ENDIAN__
vmovq_n_s16(int16_t __p0)15643 __ai int16x8_t vmovq_n_s16(int16_t __p0) {
15644 int16x8_t __ret;
15645 __ret = (int16x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
15646 return __ret;
15647 }
15648 #else
vmovq_n_s16(int16_t __p0)15649 __ai int16x8_t vmovq_n_s16(int16_t __p0) {
15650 int16x8_t __ret;
15651 __ret = (int16x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
15652 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
15653 return __ret;
15654 }
15655 #endif
15656
15657 #ifdef __LITTLE_ENDIAN__
vmov_n_u8(uint8_t __p0)15658 __ai uint8x8_t vmov_n_u8(uint8_t __p0) {
15659 uint8x8_t __ret;
15660 __ret = (uint8x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
15661 return __ret;
15662 }
15663 #else
vmov_n_u8(uint8_t __p0)15664 __ai uint8x8_t vmov_n_u8(uint8_t __p0) {
15665 uint8x8_t __ret;
15666 __ret = (uint8x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
15667 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
15668 return __ret;
15669 }
15670 #endif
15671
15672 #ifdef __LITTLE_ENDIAN__
vmov_n_u32(uint32_t __p0)15673 __ai uint32x2_t vmov_n_u32(uint32_t __p0) {
15674 uint32x2_t __ret;
15675 __ret = (uint32x2_t) {__p0, __p0};
15676 return __ret;
15677 }
15678 #else
vmov_n_u32(uint32_t __p0)15679 __ai uint32x2_t vmov_n_u32(uint32_t __p0) {
15680 uint32x2_t __ret;
15681 __ret = (uint32x2_t) {__p0, __p0};
15682 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
15683 return __ret;
15684 }
15685 #endif
15686
vmov_n_u64(uint64_t __p0)15687 __ai uint64x1_t vmov_n_u64(uint64_t __p0) {
15688 uint64x1_t __ret;
15689 __ret = (uint64x1_t) {__p0};
15690 return __ret;
15691 }
15692 #ifdef __LITTLE_ENDIAN__
vmov_n_u16(uint16_t __p0)15693 __ai uint16x4_t vmov_n_u16(uint16_t __p0) {
15694 uint16x4_t __ret;
15695 __ret = (uint16x4_t) {__p0, __p0, __p0, __p0};
15696 return __ret;
15697 }
15698 #else
vmov_n_u16(uint16_t __p0)15699 __ai uint16x4_t vmov_n_u16(uint16_t __p0) {
15700 uint16x4_t __ret;
15701 __ret = (uint16x4_t) {__p0, __p0, __p0, __p0};
15702 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
15703 return __ret;
15704 }
15705 #endif
15706
15707 #ifdef __LITTLE_ENDIAN__
vmov_n_s8(int8_t __p0)15708 __ai int8x8_t vmov_n_s8(int8_t __p0) {
15709 int8x8_t __ret;
15710 __ret = (int8x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
15711 return __ret;
15712 }
15713 #else
vmov_n_s8(int8_t __p0)15714 __ai int8x8_t vmov_n_s8(int8_t __p0) {
15715 int8x8_t __ret;
15716 __ret = (int8x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
15717 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
15718 return __ret;
15719 }
15720 #endif
15721
15722 #ifdef __LITTLE_ENDIAN__
vmov_n_f32(float32_t __p0)15723 __ai float32x2_t vmov_n_f32(float32_t __p0) {
15724 float32x2_t __ret;
15725 __ret = (float32x2_t) {__p0, __p0};
15726 return __ret;
15727 }
15728 #else
vmov_n_f32(float32_t __p0)15729 __ai float32x2_t vmov_n_f32(float32_t __p0) {
15730 float32x2_t __ret;
15731 __ret = (float32x2_t) {__p0, __p0};
15732 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
15733 return __ret;
15734 }
15735 #endif
15736
15737 #ifdef __LITTLE_ENDIAN__
15738 #define vmov_n_f16(__p0) __extension__ ({ \
15739 float16x4_t __ret; \
15740 float16_t __s0 = __p0; \
15741 __ret = (float16x4_t) {__s0, __s0, __s0, __s0}; \
15742 __ret; \
15743 })
15744 #else
15745 #define vmov_n_f16(__p0) __extension__ ({ \
15746 float16x4_t __ret; \
15747 float16_t __s0 = __p0; \
15748 __ret = (float16x4_t) {__s0, __s0, __s0, __s0}; \
15749 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
15750 __ret; \
15751 })
15752 #endif
15753
15754 #ifdef __LITTLE_ENDIAN__
vmov_n_s32(int32_t __p0)15755 __ai int32x2_t vmov_n_s32(int32_t __p0) {
15756 int32x2_t __ret;
15757 __ret = (int32x2_t) {__p0, __p0};
15758 return __ret;
15759 }
15760 #else
vmov_n_s32(int32_t __p0)15761 __ai int32x2_t vmov_n_s32(int32_t __p0) {
15762 int32x2_t __ret;
15763 __ret = (int32x2_t) {__p0, __p0};
15764 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
15765 return __ret;
15766 }
15767 #endif
15768
vmov_n_s64(int64_t __p0)15769 __ai int64x1_t vmov_n_s64(int64_t __p0) {
15770 int64x1_t __ret;
15771 __ret = (int64x1_t) {__p0};
15772 return __ret;
15773 }
15774 #ifdef __LITTLE_ENDIAN__
vmov_n_s16(int16_t __p0)15775 __ai int16x4_t vmov_n_s16(int16_t __p0) {
15776 int16x4_t __ret;
15777 __ret = (int16x4_t) {__p0, __p0, __p0, __p0};
15778 return __ret;
15779 }
15780 #else
vmov_n_s16(int16_t __p0)15781 __ai int16x4_t vmov_n_s16(int16_t __p0) {
15782 int16x4_t __ret;
15783 __ret = (int16x4_t) {__p0, __p0, __p0, __p0};
15784 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
15785 return __ret;
15786 }
15787 #endif
15788
15789 #ifdef __LITTLE_ENDIAN__
vmovl_u8(uint8x8_t __p0)15790 __ai uint16x8_t vmovl_u8(uint8x8_t __p0) {
15791 uint16x8_t __ret;
15792 __ret = (uint16x8_t) __builtin_neon_vmovl_v((int8x8_t)__p0, 49);
15793 return __ret;
15794 }
15795 #else
vmovl_u8(uint8x8_t __p0)15796 __ai uint16x8_t vmovl_u8(uint8x8_t __p0) {
15797 uint16x8_t __ret;
15798 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
15799 __ret = (uint16x8_t) __builtin_neon_vmovl_v((int8x8_t)__rev0, 49);
15800 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
15801 return __ret;
15802 }
__noswap_vmovl_u8(uint8x8_t __p0)15803 __ai uint16x8_t __noswap_vmovl_u8(uint8x8_t __p0) {
15804 uint16x8_t __ret;
15805 __ret = (uint16x8_t) __builtin_neon_vmovl_v((int8x8_t)__p0, 49);
15806 return __ret;
15807 }
15808 #endif
15809
15810 #ifdef __LITTLE_ENDIAN__
vmovl_u32(uint32x2_t __p0)15811 __ai uint64x2_t vmovl_u32(uint32x2_t __p0) {
15812 uint64x2_t __ret;
15813 __ret = (uint64x2_t) __builtin_neon_vmovl_v((int8x8_t)__p0, 51);
15814 return __ret;
15815 }
15816 #else
vmovl_u32(uint32x2_t __p0)15817 __ai uint64x2_t vmovl_u32(uint32x2_t __p0) {
15818 uint64x2_t __ret;
15819 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
15820 __ret = (uint64x2_t) __builtin_neon_vmovl_v((int8x8_t)__rev0, 51);
15821 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
15822 return __ret;
15823 }
__noswap_vmovl_u32(uint32x2_t __p0)15824 __ai uint64x2_t __noswap_vmovl_u32(uint32x2_t __p0) {
15825 uint64x2_t __ret;
15826 __ret = (uint64x2_t) __builtin_neon_vmovl_v((int8x8_t)__p0, 51);
15827 return __ret;
15828 }
15829 #endif
15830
15831 #ifdef __LITTLE_ENDIAN__
vmovl_u16(uint16x4_t __p0)15832 __ai uint32x4_t vmovl_u16(uint16x4_t __p0) {
15833 uint32x4_t __ret;
15834 __ret = (uint32x4_t) __builtin_neon_vmovl_v((int8x8_t)__p0, 50);
15835 return __ret;
15836 }
15837 #else
vmovl_u16(uint16x4_t __p0)15838 __ai uint32x4_t vmovl_u16(uint16x4_t __p0) {
15839 uint32x4_t __ret;
15840 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
15841 __ret = (uint32x4_t) __builtin_neon_vmovl_v((int8x8_t)__rev0, 50);
15842 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
15843 return __ret;
15844 }
__noswap_vmovl_u16(uint16x4_t __p0)15845 __ai uint32x4_t __noswap_vmovl_u16(uint16x4_t __p0) {
15846 uint32x4_t __ret;
15847 __ret = (uint32x4_t) __builtin_neon_vmovl_v((int8x8_t)__p0, 50);
15848 return __ret;
15849 }
15850 #endif
15851
15852 #ifdef __LITTLE_ENDIAN__
vmovl_s8(int8x8_t __p0)15853 __ai int16x8_t vmovl_s8(int8x8_t __p0) {
15854 int16x8_t __ret;
15855 __ret = (int16x8_t) __builtin_neon_vmovl_v((int8x8_t)__p0, 33);
15856 return __ret;
15857 }
15858 #else
vmovl_s8(int8x8_t __p0)15859 __ai int16x8_t vmovl_s8(int8x8_t __p0) {
15860 int16x8_t __ret;
15861 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
15862 __ret = (int16x8_t) __builtin_neon_vmovl_v((int8x8_t)__rev0, 33);
15863 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
15864 return __ret;
15865 }
__noswap_vmovl_s8(int8x8_t __p0)15866 __ai int16x8_t __noswap_vmovl_s8(int8x8_t __p0) {
15867 int16x8_t __ret;
15868 __ret = (int16x8_t) __builtin_neon_vmovl_v((int8x8_t)__p0, 33);
15869 return __ret;
15870 }
15871 #endif
15872
15873 #ifdef __LITTLE_ENDIAN__
vmovl_s32(int32x2_t __p0)15874 __ai int64x2_t vmovl_s32(int32x2_t __p0) {
15875 int64x2_t __ret;
15876 __ret = (int64x2_t) __builtin_neon_vmovl_v((int8x8_t)__p0, 35);
15877 return __ret;
15878 }
15879 #else
vmovl_s32(int32x2_t __p0)15880 __ai int64x2_t vmovl_s32(int32x2_t __p0) {
15881 int64x2_t __ret;
15882 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
15883 __ret = (int64x2_t) __builtin_neon_vmovl_v((int8x8_t)__rev0, 35);
15884 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
15885 return __ret;
15886 }
__noswap_vmovl_s32(int32x2_t __p0)15887 __ai int64x2_t __noswap_vmovl_s32(int32x2_t __p0) {
15888 int64x2_t __ret;
15889 __ret = (int64x2_t) __builtin_neon_vmovl_v((int8x8_t)__p0, 35);
15890 return __ret;
15891 }
15892 #endif
15893
15894 #ifdef __LITTLE_ENDIAN__
vmovl_s16(int16x4_t __p0)15895 __ai int32x4_t vmovl_s16(int16x4_t __p0) {
15896 int32x4_t __ret;
15897 __ret = (int32x4_t) __builtin_neon_vmovl_v((int8x8_t)__p0, 34);
15898 return __ret;
15899 }
15900 #else
vmovl_s16(int16x4_t __p0)15901 __ai int32x4_t vmovl_s16(int16x4_t __p0) {
15902 int32x4_t __ret;
15903 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
15904 __ret = (int32x4_t) __builtin_neon_vmovl_v((int8x8_t)__rev0, 34);
15905 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
15906 return __ret;
15907 }
__noswap_vmovl_s16(int16x4_t __p0)15908 __ai int32x4_t __noswap_vmovl_s16(int16x4_t __p0) {
15909 int32x4_t __ret;
15910 __ret = (int32x4_t) __builtin_neon_vmovl_v((int8x8_t)__p0, 34);
15911 return __ret;
15912 }
15913 #endif
15914
15915 #ifdef __LITTLE_ENDIAN__
vmovn_u32(uint32x4_t __p0)15916 __ai uint16x4_t vmovn_u32(uint32x4_t __p0) {
15917 uint16x4_t __ret;
15918 __ret = (uint16x4_t) __builtin_neon_vmovn_v((int8x16_t)__p0, 17);
15919 return __ret;
15920 }
15921 #else
vmovn_u32(uint32x4_t __p0)15922 __ai uint16x4_t vmovn_u32(uint32x4_t __p0) {
15923 uint16x4_t __ret;
15924 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
15925 __ret = (uint16x4_t) __builtin_neon_vmovn_v((int8x16_t)__rev0, 17);
15926 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
15927 return __ret;
15928 }
__noswap_vmovn_u32(uint32x4_t __p0)15929 __ai uint16x4_t __noswap_vmovn_u32(uint32x4_t __p0) {
15930 uint16x4_t __ret;
15931 __ret = (uint16x4_t) __builtin_neon_vmovn_v((int8x16_t)__p0, 17);
15932 return __ret;
15933 }
15934 #endif
15935
15936 #ifdef __LITTLE_ENDIAN__
vmovn_u64(uint64x2_t __p0)15937 __ai uint32x2_t vmovn_u64(uint64x2_t __p0) {
15938 uint32x2_t __ret;
15939 __ret = (uint32x2_t) __builtin_neon_vmovn_v((int8x16_t)__p0, 18);
15940 return __ret;
15941 }
15942 #else
vmovn_u64(uint64x2_t __p0)15943 __ai uint32x2_t vmovn_u64(uint64x2_t __p0) {
15944 uint32x2_t __ret;
15945 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
15946 __ret = (uint32x2_t) __builtin_neon_vmovn_v((int8x16_t)__rev0, 18);
15947 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
15948 return __ret;
15949 }
__noswap_vmovn_u64(uint64x2_t __p0)15950 __ai uint32x2_t __noswap_vmovn_u64(uint64x2_t __p0) {
15951 uint32x2_t __ret;
15952 __ret = (uint32x2_t) __builtin_neon_vmovn_v((int8x16_t)__p0, 18);
15953 return __ret;
15954 }
15955 #endif
15956
15957 #ifdef __LITTLE_ENDIAN__
vmovn_u16(uint16x8_t __p0)15958 __ai uint8x8_t vmovn_u16(uint16x8_t __p0) {
15959 uint8x8_t __ret;
15960 __ret = (uint8x8_t) __builtin_neon_vmovn_v((int8x16_t)__p0, 16);
15961 return __ret;
15962 }
15963 #else
vmovn_u16(uint16x8_t __p0)15964 __ai uint8x8_t vmovn_u16(uint16x8_t __p0) {
15965 uint8x8_t __ret;
15966 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
15967 __ret = (uint8x8_t) __builtin_neon_vmovn_v((int8x16_t)__rev0, 16);
15968 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
15969 return __ret;
15970 }
__noswap_vmovn_u16(uint16x8_t __p0)15971 __ai uint8x8_t __noswap_vmovn_u16(uint16x8_t __p0) {
15972 uint8x8_t __ret;
15973 __ret = (uint8x8_t) __builtin_neon_vmovn_v((int8x16_t)__p0, 16);
15974 return __ret;
15975 }
15976 #endif
15977
15978 #ifdef __LITTLE_ENDIAN__
vmovn_s32(int32x4_t __p0)15979 __ai int16x4_t vmovn_s32(int32x4_t __p0) {
15980 int16x4_t __ret;
15981 __ret = (int16x4_t) __builtin_neon_vmovn_v((int8x16_t)__p0, 1);
15982 return __ret;
15983 }
15984 #else
vmovn_s32(int32x4_t __p0)15985 __ai int16x4_t vmovn_s32(int32x4_t __p0) {
15986 int16x4_t __ret;
15987 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
15988 __ret = (int16x4_t) __builtin_neon_vmovn_v((int8x16_t)__rev0, 1);
15989 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
15990 return __ret;
15991 }
__noswap_vmovn_s32(int32x4_t __p0)15992 __ai int16x4_t __noswap_vmovn_s32(int32x4_t __p0) {
15993 int16x4_t __ret;
15994 __ret = (int16x4_t) __builtin_neon_vmovn_v((int8x16_t)__p0, 1);
15995 return __ret;
15996 }
15997 #endif
15998
15999 #ifdef __LITTLE_ENDIAN__
vmovn_s64(int64x2_t __p0)16000 __ai int32x2_t vmovn_s64(int64x2_t __p0) {
16001 int32x2_t __ret;
16002 __ret = (int32x2_t) __builtin_neon_vmovn_v((int8x16_t)__p0, 2);
16003 return __ret;
16004 }
16005 #else
vmovn_s64(int64x2_t __p0)16006 __ai int32x2_t vmovn_s64(int64x2_t __p0) {
16007 int32x2_t __ret;
16008 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
16009 __ret = (int32x2_t) __builtin_neon_vmovn_v((int8x16_t)__rev0, 2);
16010 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
16011 return __ret;
16012 }
__noswap_vmovn_s64(int64x2_t __p0)16013 __ai int32x2_t __noswap_vmovn_s64(int64x2_t __p0) {
16014 int32x2_t __ret;
16015 __ret = (int32x2_t) __builtin_neon_vmovn_v((int8x16_t)__p0, 2);
16016 return __ret;
16017 }
16018 #endif
16019
16020 #ifdef __LITTLE_ENDIAN__
vmovn_s16(int16x8_t __p0)16021 __ai int8x8_t vmovn_s16(int16x8_t __p0) {
16022 int8x8_t __ret;
16023 __ret = (int8x8_t) __builtin_neon_vmovn_v((int8x16_t)__p0, 0);
16024 return __ret;
16025 }
16026 #else
vmovn_s16(int16x8_t __p0)16027 __ai int8x8_t vmovn_s16(int16x8_t __p0) {
16028 int8x8_t __ret;
16029 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
16030 __ret = (int8x8_t) __builtin_neon_vmovn_v((int8x16_t)__rev0, 0);
16031 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
16032 return __ret;
16033 }
__noswap_vmovn_s16(int16x8_t __p0)16034 __ai int8x8_t __noswap_vmovn_s16(int16x8_t __p0) {
16035 int8x8_t __ret;
16036 __ret = (int8x8_t) __builtin_neon_vmovn_v((int8x16_t)__p0, 0);
16037 return __ret;
16038 }
16039 #endif
16040
16041 #ifdef __LITTLE_ENDIAN__
vmulq_u8(uint8x16_t __p0,uint8x16_t __p1)16042 __ai uint8x16_t vmulq_u8(uint8x16_t __p0, uint8x16_t __p1) {
16043 uint8x16_t __ret;
16044 __ret = __p0 * __p1;
16045 return __ret;
16046 }
16047 #else
vmulq_u8(uint8x16_t __p0,uint8x16_t __p1)16048 __ai uint8x16_t vmulq_u8(uint8x16_t __p0, uint8x16_t __p1) {
16049 uint8x16_t __ret;
16050 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
16051 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
16052 __ret = __rev0 * __rev1;
16053 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
16054 return __ret;
16055 }
16056 #endif
16057
16058 #ifdef __LITTLE_ENDIAN__
vmulq_u32(uint32x4_t __p0,uint32x4_t __p1)16059 __ai uint32x4_t vmulq_u32(uint32x4_t __p0, uint32x4_t __p1) {
16060 uint32x4_t __ret;
16061 __ret = __p0 * __p1;
16062 return __ret;
16063 }
16064 #else
vmulq_u32(uint32x4_t __p0,uint32x4_t __p1)16065 __ai uint32x4_t vmulq_u32(uint32x4_t __p0, uint32x4_t __p1) {
16066 uint32x4_t __ret;
16067 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
16068 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
16069 __ret = __rev0 * __rev1;
16070 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
16071 return __ret;
16072 }
16073 #endif
16074
16075 #ifdef __LITTLE_ENDIAN__
vmulq_u16(uint16x8_t __p0,uint16x8_t __p1)16076 __ai uint16x8_t vmulq_u16(uint16x8_t __p0, uint16x8_t __p1) {
16077 uint16x8_t __ret;
16078 __ret = __p0 * __p1;
16079 return __ret;
16080 }
16081 #else
vmulq_u16(uint16x8_t __p0,uint16x8_t __p1)16082 __ai uint16x8_t vmulq_u16(uint16x8_t __p0, uint16x8_t __p1) {
16083 uint16x8_t __ret;
16084 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
16085 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
16086 __ret = __rev0 * __rev1;
16087 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
16088 return __ret;
16089 }
16090 #endif
16091
16092 #ifdef __LITTLE_ENDIAN__
vmulq_s8(int8x16_t __p0,int8x16_t __p1)16093 __ai int8x16_t vmulq_s8(int8x16_t __p0, int8x16_t __p1) {
16094 int8x16_t __ret;
16095 __ret = __p0 * __p1;
16096 return __ret;
16097 }
16098 #else
vmulq_s8(int8x16_t __p0,int8x16_t __p1)16099 __ai int8x16_t vmulq_s8(int8x16_t __p0, int8x16_t __p1) {
16100 int8x16_t __ret;
16101 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
16102 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
16103 __ret = __rev0 * __rev1;
16104 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
16105 return __ret;
16106 }
16107 #endif
16108
16109 #ifdef __LITTLE_ENDIAN__
vmulq_f32(float32x4_t __p0,float32x4_t __p1)16110 __ai float32x4_t vmulq_f32(float32x4_t __p0, float32x4_t __p1) {
16111 float32x4_t __ret;
16112 __ret = __p0 * __p1;
16113 return __ret;
16114 }
16115 #else
vmulq_f32(float32x4_t __p0,float32x4_t __p1)16116 __ai float32x4_t vmulq_f32(float32x4_t __p0, float32x4_t __p1) {
16117 float32x4_t __ret;
16118 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
16119 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
16120 __ret = __rev0 * __rev1;
16121 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
16122 return __ret;
16123 }
16124 #endif
16125
16126 #ifdef __LITTLE_ENDIAN__
vmulq_s32(int32x4_t __p0,int32x4_t __p1)16127 __ai int32x4_t vmulq_s32(int32x4_t __p0, int32x4_t __p1) {
16128 int32x4_t __ret;
16129 __ret = __p0 * __p1;
16130 return __ret;
16131 }
16132 #else
vmulq_s32(int32x4_t __p0,int32x4_t __p1)16133 __ai int32x4_t vmulq_s32(int32x4_t __p0, int32x4_t __p1) {
16134 int32x4_t __ret;
16135 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
16136 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
16137 __ret = __rev0 * __rev1;
16138 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
16139 return __ret;
16140 }
16141 #endif
16142
16143 #ifdef __LITTLE_ENDIAN__
vmulq_s16(int16x8_t __p0,int16x8_t __p1)16144 __ai int16x8_t vmulq_s16(int16x8_t __p0, int16x8_t __p1) {
16145 int16x8_t __ret;
16146 __ret = __p0 * __p1;
16147 return __ret;
16148 }
16149 #else
vmulq_s16(int16x8_t __p0,int16x8_t __p1)16150 __ai int16x8_t vmulq_s16(int16x8_t __p0, int16x8_t __p1) {
16151 int16x8_t __ret;
16152 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
16153 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
16154 __ret = __rev0 * __rev1;
16155 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
16156 return __ret;
16157 }
16158 #endif
16159
16160 #ifdef __LITTLE_ENDIAN__
vmul_u8(uint8x8_t __p0,uint8x8_t __p1)16161 __ai uint8x8_t vmul_u8(uint8x8_t __p0, uint8x8_t __p1) {
16162 uint8x8_t __ret;
16163 __ret = __p0 * __p1;
16164 return __ret;
16165 }
16166 #else
vmul_u8(uint8x8_t __p0,uint8x8_t __p1)16167 __ai uint8x8_t vmul_u8(uint8x8_t __p0, uint8x8_t __p1) {
16168 uint8x8_t __ret;
16169 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
16170 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
16171 __ret = __rev0 * __rev1;
16172 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
16173 return __ret;
16174 }
16175 #endif
16176
16177 #ifdef __LITTLE_ENDIAN__
vmul_u32(uint32x2_t __p0,uint32x2_t __p1)16178 __ai uint32x2_t vmul_u32(uint32x2_t __p0, uint32x2_t __p1) {
16179 uint32x2_t __ret;
16180 __ret = __p0 * __p1;
16181 return __ret;
16182 }
16183 #else
vmul_u32(uint32x2_t __p0,uint32x2_t __p1)16184 __ai uint32x2_t vmul_u32(uint32x2_t __p0, uint32x2_t __p1) {
16185 uint32x2_t __ret;
16186 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
16187 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
16188 __ret = __rev0 * __rev1;
16189 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
16190 return __ret;
16191 }
16192 #endif
16193
16194 #ifdef __LITTLE_ENDIAN__
vmul_u16(uint16x4_t __p0,uint16x4_t __p1)16195 __ai uint16x4_t vmul_u16(uint16x4_t __p0, uint16x4_t __p1) {
16196 uint16x4_t __ret;
16197 __ret = __p0 * __p1;
16198 return __ret;
16199 }
16200 #else
vmul_u16(uint16x4_t __p0,uint16x4_t __p1)16201 __ai uint16x4_t vmul_u16(uint16x4_t __p0, uint16x4_t __p1) {
16202 uint16x4_t __ret;
16203 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
16204 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
16205 __ret = __rev0 * __rev1;
16206 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
16207 return __ret;
16208 }
16209 #endif
16210
16211 #ifdef __LITTLE_ENDIAN__
vmul_s8(int8x8_t __p0,int8x8_t __p1)16212 __ai int8x8_t vmul_s8(int8x8_t __p0, int8x8_t __p1) {
16213 int8x8_t __ret;
16214 __ret = __p0 * __p1;
16215 return __ret;
16216 }
16217 #else
vmul_s8(int8x8_t __p0,int8x8_t __p1)16218 __ai int8x8_t vmul_s8(int8x8_t __p0, int8x8_t __p1) {
16219 int8x8_t __ret;
16220 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
16221 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
16222 __ret = __rev0 * __rev1;
16223 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
16224 return __ret;
16225 }
16226 #endif
16227
16228 #ifdef __LITTLE_ENDIAN__
vmul_f32(float32x2_t __p0,float32x2_t __p1)16229 __ai float32x2_t vmul_f32(float32x2_t __p0, float32x2_t __p1) {
16230 float32x2_t __ret;
16231 __ret = __p0 * __p1;
16232 return __ret;
16233 }
16234 #else
vmul_f32(float32x2_t __p0,float32x2_t __p1)16235 __ai float32x2_t vmul_f32(float32x2_t __p0, float32x2_t __p1) {
16236 float32x2_t __ret;
16237 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
16238 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
16239 __ret = __rev0 * __rev1;
16240 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
16241 return __ret;
16242 }
16243 #endif
16244
16245 #ifdef __LITTLE_ENDIAN__
vmul_s32(int32x2_t __p0,int32x2_t __p1)16246 __ai int32x2_t vmul_s32(int32x2_t __p0, int32x2_t __p1) {
16247 int32x2_t __ret;
16248 __ret = __p0 * __p1;
16249 return __ret;
16250 }
16251 #else
vmul_s32(int32x2_t __p0,int32x2_t __p1)16252 __ai int32x2_t vmul_s32(int32x2_t __p0, int32x2_t __p1) {
16253 int32x2_t __ret;
16254 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
16255 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
16256 __ret = __rev0 * __rev1;
16257 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
16258 return __ret;
16259 }
16260 #endif
16261
16262 #ifdef __LITTLE_ENDIAN__
vmul_s16(int16x4_t __p0,int16x4_t __p1)16263 __ai int16x4_t vmul_s16(int16x4_t __p0, int16x4_t __p1) {
16264 int16x4_t __ret;
16265 __ret = __p0 * __p1;
16266 return __ret;
16267 }
16268 #else
vmul_s16(int16x4_t __p0,int16x4_t __p1)16269 __ai int16x4_t vmul_s16(int16x4_t __p0, int16x4_t __p1) {
16270 int16x4_t __ret;
16271 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
16272 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
16273 __ret = __rev0 * __rev1;
16274 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
16275 return __ret;
16276 }
16277 #endif
16278
16279 #ifdef __LITTLE_ENDIAN__
vmul_p8(poly8x8_t __p0,poly8x8_t __p1)16280 __ai poly8x8_t vmul_p8(poly8x8_t __p0, poly8x8_t __p1) {
16281 poly8x8_t __ret;
16282 __ret = (poly8x8_t) __builtin_neon_vmul_v((int8x8_t)__p0, (int8x8_t)__p1, 4);
16283 return __ret;
16284 }
16285 #else
vmul_p8(poly8x8_t __p0,poly8x8_t __p1)16286 __ai poly8x8_t vmul_p8(poly8x8_t __p0, poly8x8_t __p1) {
16287 poly8x8_t __ret;
16288 poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
16289 poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
16290 __ret = (poly8x8_t) __builtin_neon_vmul_v((int8x8_t)__rev0, (int8x8_t)__rev1, 4);
16291 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
16292 return __ret;
16293 }
16294 #endif
16295
16296 #ifdef __LITTLE_ENDIAN__
vmulq_p8(poly8x16_t __p0,poly8x16_t __p1)16297 __ai poly8x16_t vmulq_p8(poly8x16_t __p0, poly8x16_t __p1) {
16298 poly8x16_t __ret;
16299 __ret = (poly8x16_t) __builtin_neon_vmulq_v((int8x16_t)__p0, (int8x16_t)__p1, 36);
16300 return __ret;
16301 }
16302 #else
vmulq_p8(poly8x16_t __p0,poly8x16_t __p1)16303 __ai poly8x16_t vmulq_p8(poly8x16_t __p0, poly8x16_t __p1) {
16304 poly8x16_t __ret;
16305 poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
16306 poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
16307 __ret = (poly8x16_t) __builtin_neon_vmulq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 36);
16308 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
16309 return __ret;
16310 }
16311 #endif
16312
16313 #ifdef __LITTLE_ENDIAN__
16314 #define vmulq_lane_u32(__p0_86, __p1_86, __p2_86) __extension__ ({ \
16315 uint32x4_t __ret_86; \
16316 uint32x4_t __s0_86 = __p0_86; \
16317 uint32x2_t __s1_86 = __p1_86; \
16318 __ret_86 = __s0_86 * splatq_lane_u32(__s1_86, __p2_86); \
16319 __ret_86; \
16320 })
16321 #else
16322 #define vmulq_lane_u32(__p0_87, __p1_87, __p2_87) __extension__ ({ \
16323 uint32x4_t __ret_87; \
16324 uint32x4_t __s0_87 = __p0_87; \
16325 uint32x2_t __s1_87 = __p1_87; \
16326 uint32x4_t __rev0_87; __rev0_87 = __builtin_shufflevector(__s0_87, __s0_87, 3, 2, 1, 0); \
16327 uint32x2_t __rev1_87; __rev1_87 = __builtin_shufflevector(__s1_87, __s1_87, 1, 0); \
16328 __ret_87 = __rev0_87 * __noswap_splatq_lane_u32(__rev1_87, __p2_87); \
16329 __ret_87 = __builtin_shufflevector(__ret_87, __ret_87, 3, 2, 1, 0); \
16330 __ret_87; \
16331 })
16332 #endif
16333
16334 #ifdef __LITTLE_ENDIAN__
16335 #define vmulq_lane_u16(__p0_88, __p1_88, __p2_88) __extension__ ({ \
16336 uint16x8_t __ret_88; \
16337 uint16x8_t __s0_88 = __p0_88; \
16338 uint16x4_t __s1_88 = __p1_88; \
16339 __ret_88 = __s0_88 * splatq_lane_u16(__s1_88, __p2_88); \
16340 __ret_88; \
16341 })
16342 #else
16343 #define vmulq_lane_u16(__p0_89, __p1_89, __p2_89) __extension__ ({ \
16344 uint16x8_t __ret_89; \
16345 uint16x8_t __s0_89 = __p0_89; \
16346 uint16x4_t __s1_89 = __p1_89; \
16347 uint16x8_t __rev0_89; __rev0_89 = __builtin_shufflevector(__s0_89, __s0_89, 7, 6, 5, 4, 3, 2, 1, 0); \
16348 uint16x4_t __rev1_89; __rev1_89 = __builtin_shufflevector(__s1_89, __s1_89, 3, 2, 1, 0); \
16349 __ret_89 = __rev0_89 * __noswap_splatq_lane_u16(__rev1_89, __p2_89); \
16350 __ret_89 = __builtin_shufflevector(__ret_89, __ret_89, 7, 6, 5, 4, 3, 2, 1, 0); \
16351 __ret_89; \
16352 })
16353 #endif
16354
16355 #ifdef __LITTLE_ENDIAN__
16356 #define vmulq_lane_f32(__p0_90, __p1_90, __p2_90) __extension__ ({ \
16357 float32x4_t __ret_90; \
16358 float32x4_t __s0_90 = __p0_90; \
16359 float32x2_t __s1_90 = __p1_90; \
16360 __ret_90 = __s0_90 * splatq_lane_f32(__s1_90, __p2_90); \
16361 __ret_90; \
16362 })
16363 #else
16364 #define vmulq_lane_f32(__p0_91, __p1_91, __p2_91) __extension__ ({ \
16365 float32x4_t __ret_91; \
16366 float32x4_t __s0_91 = __p0_91; \
16367 float32x2_t __s1_91 = __p1_91; \
16368 float32x4_t __rev0_91; __rev0_91 = __builtin_shufflevector(__s0_91, __s0_91, 3, 2, 1, 0); \
16369 float32x2_t __rev1_91; __rev1_91 = __builtin_shufflevector(__s1_91, __s1_91, 1, 0); \
16370 __ret_91 = __rev0_91 * __noswap_splatq_lane_f32(__rev1_91, __p2_91); \
16371 __ret_91 = __builtin_shufflevector(__ret_91, __ret_91, 3, 2, 1, 0); \
16372 __ret_91; \
16373 })
16374 #endif
16375
16376 #ifdef __LITTLE_ENDIAN__
16377 #define vmulq_lane_s32(__p0_92, __p1_92, __p2_92) __extension__ ({ \
16378 int32x4_t __ret_92; \
16379 int32x4_t __s0_92 = __p0_92; \
16380 int32x2_t __s1_92 = __p1_92; \
16381 __ret_92 = __s0_92 * splatq_lane_s32(__s1_92, __p2_92); \
16382 __ret_92; \
16383 })
16384 #else
16385 #define vmulq_lane_s32(__p0_93, __p1_93, __p2_93) __extension__ ({ \
16386 int32x4_t __ret_93; \
16387 int32x4_t __s0_93 = __p0_93; \
16388 int32x2_t __s1_93 = __p1_93; \
16389 int32x4_t __rev0_93; __rev0_93 = __builtin_shufflevector(__s0_93, __s0_93, 3, 2, 1, 0); \
16390 int32x2_t __rev1_93; __rev1_93 = __builtin_shufflevector(__s1_93, __s1_93, 1, 0); \
16391 __ret_93 = __rev0_93 * __noswap_splatq_lane_s32(__rev1_93, __p2_93); \
16392 __ret_93 = __builtin_shufflevector(__ret_93, __ret_93, 3, 2, 1, 0); \
16393 __ret_93; \
16394 })
16395 #endif
16396
16397 #ifdef __LITTLE_ENDIAN__
16398 #define vmulq_lane_s16(__p0_94, __p1_94, __p2_94) __extension__ ({ \
16399 int16x8_t __ret_94; \
16400 int16x8_t __s0_94 = __p0_94; \
16401 int16x4_t __s1_94 = __p1_94; \
16402 __ret_94 = __s0_94 * splatq_lane_s16(__s1_94, __p2_94); \
16403 __ret_94; \
16404 })
16405 #else
16406 #define vmulq_lane_s16(__p0_95, __p1_95, __p2_95) __extension__ ({ \
16407 int16x8_t __ret_95; \
16408 int16x8_t __s0_95 = __p0_95; \
16409 int16x4_t __s1_95 = __p1_95; \
16410 int16x8_t __rev0_95; __rev0_95 = __builtin_shufflevector(__s0_95, __s0_95, 7, 6, 5, 4, 3, 2, 1, 0); \
16411 int16x4_t __rev1_95; __rev1_95 = __builtin_shufflevector(__s1_95, __s1_95, 3, 2, 1, 0); \
16412 __ret_95 = __rev0_95 * __noswap_splatq_lane_s16(__rev1_95, __p2_95); \
16413 __ret_95 = __builtin_shufflevector(__ret_95, __ret_95, 7, 6, 5, 4, 3, 2, 1, 0); \
16414 __ret_95; \
16415 })
16416 #endif
16417
16418 #ifdef __LITTLE_ENDIAN__
16419 #define vmul_lane_u32(__p0_96, __p1_96, __p2_96) __extension__ ({ \
16420 uint32x2_t __ret_96; \
16421 uint32x2_t __s0_96 = __p0_96; \
16422 uint32x2_t __s1_96 = __p1_96; \
16423 __ret_96 = __s0_96 * splat_lane_u32(__s1_96, __p2_96); \
16424 __ret_96; \
16425 })
16426 #else
16427 #define vmul_lane_u32(__p0_97, __p1_97, __p2_97) __extension__ ({ \
16428 uint32x2_t __ret_97; \
16429 uint32x2_t __s0_97 = __p0_97; \
16430 uint32x2_t __s1_97 = __p1_97; \
16431 uint32x2_t __rev0_97; __rev0_97 = __builtin_shufflevector(__s0_97, __s0_97, 1, 0); \
16432 uint32x2_t __rev1_97; __rev1_97 = __builtin_shufflevector(__s1_97, __s1_97, 1, 0); \
16433 __ret_97 = __rev0_97 * __noswap_splat_lane_u32(__rev1_97, __p2_97); \
16434 __ret_97 = __builtin_shufflevector(__ret_97, __ret_97, 1, 0); \
16435 __ret_97; \
16436 })
16437 #endif
16438
16439 #ifdef __LITTLE_ENDIAN__
16440 #define vmul_lane_u16(__p0_98, __p1_98, __p2_98) __extension__ ({ \
16441 uint16x4_t __ret_98; \
16442 uint16x4_t __s0_98 = __p0_98; \
16443 uint16x4_t __s1_98 = __p1_98; \
16444 __ret_98 = __s0_98 * splat_lane_u16(__s1_98, __p2_98); \
16445 __ret_98; \
16446 })
16447 #else
16448 #define vmul_lane_u16(__p0_99, __p1_99, __p2_99) __extension__ ({ \
16449 uint16x4_t __ret_99; \
16450 uint16x4_t __s0_99 = __p0_99; \
16451 uint16x4_t __s1_99 = __p1_99; \
16452 uint16x4_t __rev0_99; __rev0_99 = __builtin_shufflevector(__s0_99, __s0_99, 3, 2, 1, 0); \
16453 uint16x4_t __rev1_99; __rev1_99 = __builtin_shufflevector(__s1_99, __s1_99, 3, 2, 1, 0); \
16454 __ret_99 = __rev0_99 * __noswap_splat_lane_u16(__rev1_99, __p2_99); \
16455 __ret_99 = __builtin_shufflevector(__ret_99, __ret_99, 3, 2, 1, 0); \
16456 __ret_99; \
16457 })
16458 #endif
16459
16460 #ifdef __LITTLE_ENDIAN__
16461 #define vmul_lane_f32(__p0_100, __p1_100, __p2_100) __extension__ ({ \
16462 float32x2_t __ret_100; \
16463 float32x2_t __s0_100 = __p0_100; \
16464 float32x2_t __s1_100 = __p1_100; \
16465 __ret_100 = __s0_100 * splat_lane_f32(__s1_100, __p2_100); \
16466 __ret_100; \
16467 })
16468 #else
16469 #define vmul_lane_f32(__p0_101, __p1_101, __p2_101) __extension__ ({ \
16470 float32x2_t __ret_101; \
16471 float32x2_t __s0_101 = __p0_101; \
16472 float32x2_t __s1_101 = __p1_101; \
16473 float32x2_t __rev0_101; __rev0_101 = __builtin_shufflevector(__s0_101, __s0_101, 1, 0); \
16474 float32x2_t __rev1_101; __rev1_101 = __builtin_shufflevector(__s1_101, __s1_101, 1, 0); \
16475 __ret_101 = __rev0_101 * __noswap_splat_lane_f32(__rev1_101, __p2_101); \
16476 __ret_101 = __builtin_shufflevector(__ret_101, __ret_101, 1, 0); \
16477 __ret_101; \
16478 })
16479 #endif
16480
16481 #ifdef __LITTLE_ENDIAN__
16482 #define vmul_lane_s32(__p0_102, __p1_102, __p2_102) __extension__ ({ \
16483 int32x2_t __ret_102; \
16484 int32x2_t __s0_102 = __p0_102; \
16485 int32x2_t __s1_102 = __p1_102; \
16486 __ret_102 = __s0_102 * splat_lane_s32(__s1_102, __p2_102); \
16487 __ret_102; \
16488 })
16489 #else
16490 #define vmul_lane_s32(__p0_103, __p1_103, __p2_103) __extension__ ({ \
16491 int32x2_t __ret_103; \
16492 int32x2_t __s0_103 = __p0_103; \
16493 int32x2_t __s1_103 = __p1_103; \
16494 int32x2_t __rev0_103; __rev0_103 = __builtin_shufflevector(__s0_103, __s0_103, 1, 0); \
16495 int32x2_t __rev1_103; __rev1_103 = __builtin_shufflevector(__s1_103, __s1_103, 1, 0); \
16496 __ret_103 = __rev0_103 * __noswap_splat_lane_s32(__rev1_103, __p2_103); \
16497 __ret_103 = __builtin_shufflevector(__ret_103, __ret_103, 1, 0); \
16498 __ret_103; \
16499 })
16500 #endif
16501
16502 #ifdef __LITTLE_ENDIAN__
16503 #define vmul_lane_s16(__p0_104, __p1_104, __p2_104) __extension__ ({ \
16504 int16x4_t __ret_104; \
16505 int16x4_t __s0_104 = __p0_104; \
16506 int16x4_t __s1_104 = __p1_104; \
16507 __ret_104 = __s0_104 * splat_lane_s16(__s1_104, __p2_104); \
16508 __ret_104; \
16509 })
16510 #else
16511 #define vmul_lane_s16(__p0_105, __p1_105, __p2_105) __extension__ ({ \
16512 int16x4_t __ret_105; \
16513 int16x4_t __s0_105 = __p0_105; \
16514 int16x4_t __s1_105 = __p1_105; \
16515 int16x4_t __rev0_105; __rev0_105 = __builtin_shufflevector(__s0_105, __s0_105, 3, 2, 1, 0); \
16516 int16x4_t __rev1_105; __rev1_105 = __builtin_shufflevector(__s1_105, __s1_105, 3, 2, 1, 0); \
16517 __ret_105 = __rev0_105 * __noswap_splat_lane_s16(__rev1_105, __p2_105); \
16518 __ret_105 = __builtin_shufflevector(__ret_105, __ret_105, 3, 2, 1, 0); \
16519 __ret_105; \
16520 })
16521 #endif
16522
16523 #ifdef __LITTLE_ENDIAN__
vmulq_n_u32(uint32x4_t __p0,uint32_t __p1)16524 __ai uint32x4_t vmulq_n_u32(uint32x4_t __p0, uint32_t __p1) {
16525 uint32x4_t __ret;
16526 __ret = __p0 * (uint32x4_t) {__p1, __p1, __p1, __p1};
16527 return __ret;
16528 }
16529 #else
vmulq_n_u32(uint32x4_t __p0,uint32_t __p1)16530 __ai uint32x4_t vmulq_n_u32(uint32x4_t __p0, uint32_t __p1) {
16531 uint32x4_t __ret;
16532 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
16533 __ret = __rev0 * (uint32x4_t) {__p1, __p1, __p1, __p1};
16534 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
16535 return __ret;
16536 }
16537 #endif
16538
16539 #ifdef __LITTLE_ENDIAN__
vmulq_n_u16(uint16x8_t __p0,uint16_t __p1)16540 __ai uint16x8_t vmulq_n_u16(uint16x8_t __p0, uint16_t __p1) {
16541 uint16x8_t __ret;
16542 __ret = __p0 * (uint16x8_t) {__p1, __p1, __p1, __p1, __p1, __p1, __p1, __p1};
16543 return __ret;
16544 }
16545 #else
vmulq_n_u16(uint16x8_t __p0,uint16_t __p1)16546 __ai uint16x8_t vmulq_n_u16(uint16x8_t __p0, uint16_t __p1) {
16547 uint16x8_t __ret;
16548 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
16549 __ret = __rev0 * (uint16x8_t) {__p1, __p1, __p1, __p1, __p1, __p1, __p1, __p1};
16550 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
16551 return __ret;
16552 }
16553 #endif
16554
16555 #ifdef __LITTLE_ENDIAN__
vmulq_n_f32(float32x4_t __p0,float32_t __p1)16556 __ai float32x4_t vmulq_n_f32(float32x4_t __p0, float32_t __p1) {
16557 float32x4_t __ret;
16558 __ret = __p0 * (float32x4_t) {__p1, __p1, __p1, __p1};
16559 return __ret;
16560 }
16561 #else
vmulq_n_f32(float32x4_t __p0,float32_t __p1)16562 __ai float32x4_t vmulq_n_f32(float32x4_t __p0, float32_t __p1) {
16563 float32x4_t __ret;
16564 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
16565 __ret = __rev0 * (float32x4_t) {__p1, __p1, __p1, __p1};
16566 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
16567 return __ret;
16568 }
16569 #endif
16570
16571 #ifdef __LITTLE_ENDIAN__
vmulq_n_s32(int32x4_t __p0,int32_t __p1)16572 __ai int32x4_t vmulq_n_s32(int32x4_t __p0, int32_t __p1) {
16573 int32x4_t __ret;
16574 __ret = __p0 * (int32x4_t) {__p1, __p1, __p1, __p1};
16575 return __ret;
16576 }
16577 #else
vmulq_n_s32(int32x4_t __p0,int32_t __p1)16578 __ai int32x4_t vmulq_n_s32(int32x4_t __p0, int32_t __p1) {
16579 int32x4_t __ret;
16580 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
16581 __ret = __rev0 * (int32x4_t) {__p1, __p1, __p1, __p1};
16582 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
16583 return __ret;
16584 }
16585 #endif
16586
16587 #ifdef __LITTLE_ENDIAN__
vmulq_n_s16(int16x8_t __p0,int16_t __p1)16588 __ai int16x8_t vmulq_n_s16(int16x8_t __p0, int16_t __p1) {
16589 int16x8_t __ret;
16590 __ret = __p0 * (int16x8_t) {__p1, __p1, __p1, __p1, __p1, __p1, __p1, __p1};
16591 return __ret;
16592 }
16593 #else
vmulq_n_s16(int16x8_t __p0,int16_t __p1)16594 __ai int16x8_t vmulq_n_s16(int16x8_t __p0, int16_t __p1) {
16595 int16x8_t __ret;
16596 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
16597 __ret = __rev0 * (int16x8_t) {__p1, __p1, __p1, __p1, __p1, __p1, __p1, __p1};
16598 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
16599 return __ret;
16600 }
16601 #endif
16602
16603 #ifdef __LITTLE_ENDIAN__
vmul_n_u32(uint32x2_t __p0,uint32_t __p1)16604 __ai uint32x2_t vmul_n_u32(uint32x2_t __p0, uint32_t __p1) {
16605 uint32x2_t __ret;
16606 __ret = __p0 * (uint32x2_t) {__p1, __p1};
16607 return __ret;
16608 }
16609 #else
vmul_n_u32(uint32x2_t __p0,uint32_t __p1)16610 __ai uint32x2_t vmul_n_u32(uint32x2_t __p0, uint32_t __p1) {
16611 uint32x2_t __ret;
16612 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
16613 __ret = __rev0 * (uint32x2_t) {__p1, __p1};
16614 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
16615 return __ret;
16616 }
16617 #endif
16618
16619 #ifdef __LITTLE_ENDIAN__
vmul_n_u16(uint16x4_t __p0,uint16_t __p1)16620 __ai uint16x4_t vmul_n_u16(uint16x4_t __p0, uint16_t __p1) {
16621 uint16x4_t __ret;
16622 __ret = __p0 * (uint16x4_t) {__p1, __p1, __p1, __p1};
16623 return __ret;
16624 }
16625 #else
vmul_n_u16(uint16x4_t __p0,uint16_t __p1)16626 __ai uint16x4_t vmul_n_u16(uint16x4_t __p0, uint16_t __p1) {
16627 uint16x4_t __ret;
16628 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
16629 __ret = __rev0 * (uint16x4_t) {__p1, __p1, __p1, __p1};
16630 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
16631 return __ret;
16632 }
16633 #endif
16634
16635 #ifdef __LITTLE_ENDIAN__
vmul_n_f32(float32x2_t __p0,float32_t __p1)16636 __ai float32x2_t vmul_n_f32(float32x2_t __p0, float32_t __p1) {
16637 float32x2_t __ret;
16638 __ret = __p0 * (float32x2_t) {__p1, __p1};
16639 return __ret;
16640 }
16641 #else
vmul_n_f32(float32x2_t __p0,float32_t __p1)16642 __ai float32x2_t vmul_n_f32(float32x2_t __p0, float32_t __p1) {
16643 float32x2_t __ret;
16644 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
16645 __ret = __rev0 * (float32x2_t) {__p1, __p1};
16646 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
16647 return __ret;
16648 }
16649 #endif
16650
16651 #ifdef __LITTLE_ENDIAN__
vmul_n_s32(int32x2_t __p0,int32_t __p1)16652 __ai int32x2_t vmul_n_s32(int32x2_t __p0, int32_t __p1) {
16653 int32x2_t __ret;
16654 __ret = __p0 * (int32x2_t) {__p1, __p1};
16655 return __ret;
16656 }
16657 #else
vmul_n_s32(int32x2_t __p0,int32_t __p1)16658 __ai int32x2_t vmul_n_s32(int32x2_t __p0, int32_t __p1) {
16659 int32x2_t __ret;
16660 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
16661 __ret = __rev0 * (int32x2_t) {__p1, __p1};
16662 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
16663 return __ret;
16664 }
16665 #endif
16666
16667 #ifdef __LITTLE_ENDIAN__
vmul_n_s16(int16x4_t __p0,int16_t __p1)16668 __ai int16x4_t vmul_n_s16(int16x4_t __p0, int16_t __p1) {
16669 int16x4_t __ret;
16670 __ret = __p0 * (int16x4_t) {__p1, __p1, __p1, __p1};
16671 return __ret;
16672 }
16673 #else
vmul_n_s16(int16x4_t __p0,int16_t __p1)16674 __ai int16x4_t vmul_n_s16(int16x4_t __p0, int16_t __p1) {
16675 int16x4_t __ret;
16676 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
16677 __ret = __rev0 * (int16x4_t) {__p1, __p1, __p1, __p1};
16678 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
16679 return __ret;
16680 }
16681 #endif
16682
16683 #ifdef __LITTLE_ENDIAN__
vmull_p8(poly8x8_t __p0,poly8x8_t __p1)16684 __ai poly16x8_t vmull_p8(poly8x8_t __p0, poly8x8_t __p1) {
16685 poly16x8_t __ret;
16686 __ret = (poly16x8_t) __builtin_neon_vmull_v((int8x8_t)__p0, (int8x8_t)__p1, 37);
16687 return __ret;
16688 }
16689 #else
vmull_p8(poly8x8_t __p0,poly8x8_t __p1)16690 __ai poly16x8_t vmull_p8(poly8x8_t __p0, poly8x8_t __p1) {
16691 poly16x8_t __ret;
16692 poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
16693 poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
16694 __ret = (poly16x8_t) __builtin_neon_vmull_v((int8x8_t)__rev0, (int8x8_t)__rev1, 37);
16695 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
16696 return __ret;
16697 }
__noswap_vmull_p8(poly8x8_t __p0,poly8x8_t __p1)16698 __ai poly16x8_t __noswap_vmull_p8(poly8x8_t __p0, poly8x8_t __p1) {
16699 poly16x8_t __ret;
16700 __ret = (poly16x8_t) __builtin_neon_vmull_v((int8x8_t)__p0, (int8x8_t)__p1, 37);
16701 return __ret;
16702 }
16703 #endif
16704
16705 #ifdef __LITTLE_ENDIAN__
vmull_u8(uint8x8_t __p0,uint8x8_t __p1)16706 __ai uint16x8_t vmull_u8(uint8x8_t __p0, uint8x8_t __p1) {
16707 uint16x8_t __ret;
16708 __ret = (uint16x8_t) __builtin_neon_vmull_v((int8x8_t)__p0, (int8x8_t)__p1, 49);
16709 return __ret;
16710 }
16711 #else
vmull_u8(uint8x8_t __p0,uint8x8_t __p1)16712 __ai uint16x8_t vmull_u8(uint8x8_t __p0, uint8x8_t __p1) {
16713 uint16x8_t __ret;
16714 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
16715 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
16716 __ret = (uint16x8_t) __builtin_neon_vmull_v((int8x8_t)__rev0, (int8x8_t)__rev1, 49);
16717 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
16718 return __ret;
16719 }
__noswap_vmull_u8(uint8x8_t __p0,uint8x8_t __p1)16720 __ai uint16x8_t __noswap_vmull_u8(uint8x8_t __p0, uint8x8_t __p1) {
16721 uint16x8_t __ret;
16722 __ret = (uint16x8_t) __builtin_neon_vmull_v((int8x8_t)__p0, (int8x8_t)__p1, 49);
16723 return __ret;
16724 }
16725 #endif
16726
16727 #ifdef __LITTLE_ENDIAN__
vmull_u32(uint32x2_t __p0,uint32x2_t __p1)16728 __ai uint64x2_t vmull_u32(uint32x2_t __p0, uint32x2_t __p1) {
16729 uint64x2_t __ret;
16730 __ret = (uint64x2_t) __builtin_neon_vmull_v((int8x8_t)__p0, (int8x8_t)__p1, 51);
16731 return __ret;
16732 }
16733 #else
vmull_u32(uint32x2_t __p0,uint32x2_t __p1)16734 __ai uint64x2_t vmull_u32(uint32x2_t __p0, uint32x2_t __p1) {
16735 uint64x2_t __ret;
16736 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
16737 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
16738 __ret = (uint64x2_t) __builtin_neon_vmull_v((int8x8_t)__rev0, (int8x8_t)__rev1, 51);
16739 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
16740 return __ret;
16741 }
__noswap_vmull_u32(uint32x2_t __p0,uint32x2_t __p1)16742 __ai uint64x2_t __noswap_vmull_u32(uint32x2_t __p0, uint32x2_t __p1) {
16743 uint64x2_t __ret;
16744 __ret = (uint64x2_t) __builtin_neon_vmull_v((int8x8_t)__p0, (int8x8_t)__p1, 51);
16745 return __ret;
16746 }
16747 #endif
16748
16749 #ifdef __LITTLE_ENDIAN__
vmull_u16(uint16x4_t __p0,uint16x4_t __p1)16750 __ai uint32x4_t vmull_u16(uint16x4_t __p0, uint16x4_t __p1) {
16751 uint32x4_t __ret;
16752 __ret = (uint32x4_t) __builtin_neon_vmull_v((int8x8_t)__p0, (int8x8_t)__p1, 50);
16753 return __ret;
16754 }
16755 #else
vmull_u16(uint16x4_t __p0,uint16x4_t __p1)16756 __ai uint32x4_t vmull_u16(uint16x4_t __p0, uint16x4_t __p1) {
16757 uint32x4_t __ret;
16758 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
16759 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
16760 __ret = (uint32x4_t) __builtin_neon_vmull_v((int8x8_t)__rev0, (int8x8_t)__rev1, 50);
16761 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
16762 return __ret;
16763 }
__noswap_vmull_u16(uint16x4_t __p0,uint16x4_t __p1)16764 __ai uint32x4_t __noswap_vmull_u16(uint16x4_t __p0, uint16x4_t __p1) {
16765 uint32x4_t __ret;
16766 __ret = (uint32x4_t) __builtin_neon_vmull_v((int8x8_t)__p0, (int8x8_t)__p1, 50);
16767 return __ret;
16768 }
16769 #endif
16770
16771 #ifdef __LITTLE_ENDIAN__
vmull_s8(int8x8_t __p0,int8x8_t __p1)16772 __ai int16x8_t vmull_s8(int8x8_t __p0, int8x8_t __p1) {
16773 int16x8_t __ret;
16774 __ret = (int16x8_t) __builtin_neon_vmull_v((int8x8_t)__p0, (int8x8_t)__p1, 33);
16775 return __ret;
16776 }
16777 #else
vmull_s8(int8x8_t __p0,int8x8_t __p1)16778 __ai int16x8_t vmull_s8(int8x8_t __p0, int8x8_t __p1) {
16779 int16x8_t __ret;
16780 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
16781 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
16782 __ret = (int16x8_t) __builtin_neon_vmull_v((int8x8_t)__rev0, (int8x8_t)__rev1, 33);
16783 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
16784 return __ret;
16785 }
__noswap_vmull_s8(int8x8_t __p0,int8x8_t __p1)16786 __ai int16x8_t __noswap_vmull_s8(int8x8_t __p0, int8x8_t __p1) {
16787 int16x8_t __ret;
16788 __ret = (int16x8_t) __builtin_neon_vmull_v((int8x8_t)__p0, (int8x8_t)__p1, 33);
16789 return __ret;
16790 }
16791 #endif
16792
16793 #ifdef __LITTLE_ENDIAN__
vmull_s32(int32x2_t __p0,int32x2_t __p1)16794 __ai int64x2_t vmull_s32(int32x2_t __p0, int32x2_t __p1) {
16795 int64x2_t __ret;
16796 __ret = (int64x2_t) __builtin_neon_vmull_v((int8x8_t)__p0, (int8x8_t)__p1, 35);
16797 return __ret;
16798 }
16799 #else
vmull_s32(int32x2_t __p0,int32x2_t __p1)16800 __ai int64x2_t vmull_s32(int32x2_t __p0, int32x2_t __p1) {
16801 int64x2_t __ret;
16802 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
16803 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
16804 __ret = (int64x2_t) __builtin_neon_vmull_v((int8x8_t)__rev0, (int8x8_t)__rev1, 35);
16805 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
16806 return __ret;
16807 }
__noswap_vmull_s32(int32x2_t __p0,int32x2_t __p1)16808 __ai int64x2_t __noswap_vmull_s32(int32x2_t __p0, int32x2_t __p1) {
16809 int64x2_t __ret;
16810 __ret = (int64x2_t) __builtin_neon_vmull_v((int8x8_t)__p0, (int8x8_t)__p1, 35);
16811 return __ret;
16812 }
16813 #endif
16814
16815 #ifdef __LITTLE_ENDIAN__
vmull_s16(int16x4_t __p0,int16x4_t __p1)16816 __ai int32x4_t vmull_s16(int16x4_t __p0, int16x4_t __p1) {
16817 int32x4_t __ret;
16818 __ret = (int32x4_t) __builtin_neon_vmull_v((int8x8_t)__p0, (int8x8_t)__p1, 34);
16819 return __ret;
16820 }
16821 #else
vmull_s16(int16x4_t __p0,int16x4_t __p1)16822 __ai int32x4_t vmull_s16(int16x4_t __p0, int16x4_t __p1) {
16823 int32x4_t __ret;
16824 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
16825 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
16826 __ret = (int32x4_t) __builtin_neon_vmull_v((int8x8_t)__rev0, (int8x8_t)__rev1, 34);
16827 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
16828 return __ret;
16829 }
__noswap_vmull_s16(int16x4_t __p0,int16x4_t __p1)16830 __ai int32x4_t __noswap_vmull_s16(int16x4_t __p0, int16x4_t __p1) {
16831 int32x4_t __ret;
16832 __ret = (int32x4_t) __builtin_neon_vmull_v((int8x8_t)__p0, (int8x8_t)__p1, 34);
16833 return __ret;
16834 }
16835 #endif
16836
16837 #ifdef __LITTLE_ENDIAN__
16838 #define vmull_lane_u32(__p0_106, __p1_106, __p2_106) __extension__ ({ \
16839 uint64x2_t __ret_106; \
16840 uint32x2_t __s0_106 = __p0_106; \
16841 uint32x2_t __s1_106 = __p1_106; \
16842 __ret_106 = vmull_u32(__s0_106, splat_lane_u32(__s1_106, __p2_106)); \
16843 __ret_106; \
16844 })
16845 #else
16846 #define vmull_lane_u32(__p0_107, __p1_107, __p2_107) __extension__ ({ \
16847 uint64x2_t __ret_107; \
16848 uint32x2_t __s0_107 = __p0_107; \
16849 uint32x2_t __s1_107 = __p1_107; \
16850 uint32x2_t __rev0_107; __rev0_107 = __builtin_shufflevector(__s0_107, __s0_107, 1, 0); \
16851 uint32x2_t __rev1_107; __rev1_107 = __builtin_shufflevector(__s1_107, __s1_107, 1, 0); \
16852 __ret_107 = __noswap_vmull_u32(__rev0_107, __noswap_splat_lane_u32(__rev1_107, __p2_107)); \
16853 __ret_107 = __builtin_shufflevector(__ret_107, __ret_107, 1, 0); \
16854 __ret_107; \
16855 })
16856 #endif
16857
16858 #ifdef __LITTLE_ENDIAN__
16859 #define vmull_lane_u16(__p0_108, __p1_108, __p2_108) __extension__ ({ \
16860 uint32x4_t __ret_108; \
16861 uint16x4_t __s0_108 = __p0_108; \
16862 uint16x4_t __s1_108 = __p1_108; \
16863 __ret_108 = vmull_u16(__s0_108, splat_lane_u16(__s1_108, __p2_108)); \
16864 __ret_108; \
16865 })
16866 #else
16867 #define vmull_lane_u16(__p0_109, __p1_109, __p2_109) __extension__ ({ \
16868 uint32x4_t __ret_109; \
16869 uint16x4_t __s0_109 = __p0_109; \
16870 uint16x4_t __s1_109 = __p1_109; \
16871 uint16x4_t __rev0_109; __rev0_109 = __builtin_shufflevector(__s0_109, __s0_109, 3, 2, 1, 0); \
16872 uint16x4_t __rev1_109; __rev1_109 = __builtin_shufflevector(__s1_109, __s1_109, 3, 2, 1, 0); \
16873 __ret_109 = __noswap_vmull_u16(__rev0_109, __noswap_splat_lane_u16(__rev1_109, __p2_109)); \
16874 __ret_109 = __builtin_shufflevector(__ret_109, __ret_109, 3, 2, 1, 0); \
16875 __ret_109; \
16876 })
16877 #endif
16878
16879 #ifdef __LITTLE_ENDIAN__
16880 #define vmull_lane_s32(__p0_110, __p1_110, __p2_110) __extension__ ({ \
16881 int64x2_t __ret_110; \
16882 int32x2_t __s0_110 = __p0_110; \
16883 int32x2_t __s1_110 = __p1_110; \
16884 __ret_110 = vmull_s32(__s0_110, splat_lane_s32(__s1_110, __p2_110)); \
16885 __ret_110; \
16886 })
16887 #else
16888 #define vmull_lane_s32(__p0_111, __p1_111, __p2_111) __extension__ ({ \
16889 int64x2_t __ret_111; \
16890 int32x2_t __s0_111 = __p0_111; \
16891 int32x2_t __s1_111 = __p1_111; \
16892 int32x2_t __rev0_111; __rev0_111 = __builtin_shufflevector(__s0_111, __s0_111, 1, 0); \
16893 int32x2_t __rev1_111; __rev1_111 = __builtin_shufflevector(__s1_111, __s1_111, 1, 0); \
16894 __ret_111 = __noswap_vmull_s32(__rev0_111, __noswap_splat_lane_s32(__rev1_111, __p2_111)); \
16895 __ret_111 = __builtin_shufflevector(__ret_111, __ret_111, 1, 0); \
16896 __ret_111; \
16897 })
16898 #endif
16899
16900 #ifdef __LITTLE_ENDIAN__
16901 #define vmull_lane_s16(__p0_112, __p1_112, __p2_112) __extension__ ({ \
16902 int32x4_t __ret_112; \
16903 int16x4_t __s0_112 = __p0_112; \
16904 int16x4_t __s1_112 = __p1_112; \
16905 __ret_112 = vmull_s16(__s0_112, splat_lane_s16(__s1_112, __p2_112)); \
16906 __ret_112; \
16907 })
16908 #else
16909 #define vmull_lane_s16(__p0_113, __p1_113, __p2_113) __extension__ ({ \
16910 int32x4_t __ret_113; \
16911 int16x4_t __s0_113 = __p0_113; \
16912 int16x4_t __s1_113 = __p1_113; \
16913 int16x4_t __rev0_113; __rev0_113 = __builtin_shufflevector(__s0_113, __s0_113, 3, 2, 1, 0); \
16914 int16x4_t __rev1_113; __rev1_113 = __builtin_shufflevector(__s1_113, __s1_113, 3, 2, 1, 0); \
16915 __ret_113 = __noswap_vmull_s16(__rev0_113, __noswap_splat_lane_s16(__rev1_113, __p2_113)); \
16916 __ret_113 = __builtin_shufflevector(__ret_113, __ret_113, 3, 2, 1, 0); \
16917 __ret_113; \
16918 })
16919 #endif
16920
16921 #ifdef __LITTLE_ENDIAN__
vmull_n_u32(uint32x2_t __p0,uint32_t __p1)16922 __ai uint64x2_t vmull_n_u32(uint32x2_t __p0, uint32_t __p1) {
16923 uint64x2_t __ret;
16924 __ret = vmull_u32(__p0, (uint32x2_t) {__p1, __p1});
16925 return __ret;
16926 }
16927 #else
vmull_n_u32(uint32x2_t __p0,uint32_t __p1)16928 __ai uint64x2_t vmull_n_u32(uint32x2_t __p0, uint32_t __p1) {
16929 uint64x2_t __ret;
16930 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
16931 __ret = __noswap_vmull_u32(__rev0, (uint32x2_t) {__p1, __p1});
16932 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
16933 return __ret;
16934 }
__noswap_vmull_n_u32(uint32x2_t __p0,uint32_t __p1)16935 __ai uint64x2_t __noswap_vmull_n_u32(uint32x2_t __p0, uint32_t __p1) {
16936 uint64x2_t __ret;
16937 __ret = __noswap_vmull_u32(__p0, (uint32x2_t) {__p1, __p1});
16938 return __ret;
16939 }
16940 #endif
16941
16942 #ifdef __LITTLE_ENDIAN__
vmull_n_u16(uint16x4_t __p0,uint16_t __p1)16943 __ai uint32x4_t vmull_n_u16(uint16x4_t __p0, uint16_t __p1) {
16944 uint32x4_t __ret;
16945 __ret = vmull_u16(__p0, (uint16x4_t) {__p1, __p1, __p1, __p1});
16946 return __ret;
16947 }
16948 #else
vmull_n_u16(uint16x4_t __p0,uint16_t __p1)16949 __ai uint32x4_t vmull_n_u16(uint16x4_t __p0, uint16_t __p1) {
16950 uint32x4_t __ret;
16951 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
16952 __ret = __noswap_vmull_u16(__rev0, (uint16x4_t) {__p1, __p1, __p1, __p1});
16953 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
16954 return __ret;
16955 }
__noswap_vmull_n_u16(uint16x4_t __p0,uint16_t __p1)16956 __ai uint32x4_t __noswap_vmull_n_u16(uint16x4_t __p0, uint16_t __p1) {
16957 uint32x4_t __ret;
16958 __ret = __noswap_vmull_u16(__p0, (uint16x4_t) {__p1, __p1, __p1, __p1});
16959 return __ret;
16960 }
16961 #endif
16962
16963 #ifdef __LITTLE_ENDIAN__
vmull_n_s32(int32x2_t __p0,int32_t __p1)16964 __ai int64x2_t vmull_n_s32(int32x2_t __p0, int32_t __p1) {
16965 int64x2_t __ret;
16966 __ret = vmull_s32(__p0, (int32x2_t) {__p1, __p1});
16967 return __ret;
16968 }
16969 #else
vmull_n_s32(int32x2_t __p0,int32_t __p1)16970 __ai int64x2_t vmull_n_s32(int32x2_t __p0, int32_t __p1) {
16971 int64x2_t __ret;
16972 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
16973 __ret = __noswap_vmull_s32(__rev0, (int32x2_t) {__p1, __p1});
16974 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
16975 return __ret;
16976 }
__noswap_vmull_n_s32(int32x2_t __p0,int32_t __p1)16977 __ai int64x2_t __noswap_vmull_n_s32(int32x2_t __p0, int32_t __p1) {
16978 int64x2_t __ret;
16979 __ret = __noswap_vmull_s32(__p0, (int32x2_t) {__p1, __p1});
16980 return __ret;
16981 }
16982 #endif
16983
16984 #ifdef __LITTLE_ENDIAN__
vmull_n_s16(int16x4_t __p0,int16_t __p1)16985 __ai int32x4_t vmull_n_s16(int16x4_t __p0, int16_t __p1) {
16986 int32x4_t __ret;
16987 __ret = vmull_s16(__p0, (int16x4_t) {__p1, __p1, __p1, __p1});
16988 return __ret;
16989 }
16990 #else
vmull_n_s16(int16x4_t __p0,int16_t __p1)16991 __ai int32x4_t vmull_n_s16(int16x4_t __p0, int16_t __p1) {
16992 int32x4_t __ret;
16993 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
16994 __ret = __noswap_vmull_s16(__rev0, (int16x4_t) {__p1, __p1, __p1, __p1});
16995 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
16996 return __ret;
16997 }
__noswap_vmull_n_s16(int16x4_t __p0,int16_t __p1)16998 __ai int32x4_t __noswap_vmull_n_s16(int16x4_t __p0, int16_t __p1) {
16999 int32x4_t __ret;
17000 __ret = __noswap_vmull_s16(__p0, (int16x4_t) {__p1, __p1, __p1, __p1});
17001 return __ret;
17002 }
17003 #endif
17004
17005 #ifdef __LITTLE_ENDIAN__
vmvn_p8(poly8x8_t __p0)17006 __ai poly8x8_t vmvn_p8(poly8x8_t __p0) {
17007 poly8x8_t __ret;
17008 __ret = ~__p0;
17009 return __ret;
17010 }
17011 #else
vmvn_p8(poly8x8_t __p0)17012 __ai poly8x8_t vmvn_p8(poly8x8_t __p0) {
17013 poly8x8_t __ret;
17014 poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
17015 __ret = ~__rev0;
17016 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
17017 return __ret;
17018 }
17019 #endif
17020
17021 #ifdef __LITTLE_ENDIAN__
vmvnq_p8(poly8x16_t __p0)17022 __ai poly8x16_t vmvnq_p8(poly8x16_t __p0) {
17023 poly8x16_t __ret;
17024 __ret = ~__p0;
17025 return __ret;
17026 }
17027 #else
vmvnq_p8(poly8x16_t __p0)17028 __ai poly8x16_t vmvnq_p8(poly8x16_t __p0) {
17029 poly8x16_t __ret;
17030 poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
17031 __ret = ~__rev0;
17032 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
17033 return __ret;
17034 }
17035 #endif
17036
17037 #ifdef __LITTLE_ENDIAN__
vmvnq_u8(uint8x16_t __p0)17038 __ai uint8x16_t vmvnq_u8(uint8x16_t __p0) {
17039 uint8x16_t __ret;
17040 __ret = ~__p0;
17041 return __ret;
17042 }
17043 #else
vmvnq_u8(uint8x16_t __p0)17044 __ai uint8x16_t vmvnq_u8(uint8x16_t __p0) {
17045 uint8x16_t __ret;
17046 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
17047 __ret = ~__rev0;
17048 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
17049 return __ret;
17050 }
17051 #endif
17052
17053 #ifdef __LITTLE_ENDIAN__
vmvnq_u32(uint32x4_t __p0)17054 __ai uint32x4_t vmvnq_u32(uint32x4_t __p0) {
17055 uint32x4_t __ret;
17056 __ret = ~__p0;
17057 return __ret;
17058 }
17059 #else
vmvnq_u32(uint32x4_t __p0)17060 __ai uint32x4_t vmvnq_u32(uint32x4_t __p0) {
17061 uint32x4_t __ret;
17062 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17063 __ret = ~__rev0;
17064 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
17065 return __ret;
17066 }
17067 #endif
17068
17069 #ifdef __LITTLE_ENDIAN__
vmvnq_u16(uint16x8_t __p0)17070 __ai uint16x8_t vmvnq_u16(uint16x8_t __p0) {
17071 uint16x8_t __ret;
17072 __ret = ~__p0;
17073 return __ret;
17074 }
17075 #else
vmvnq_u16(uint16x8_t __p0)17076 __ai uint16x8_t vmvnq_u16(uint16x8_t __p0) {
17077 uint16x8_t __ret;
17078 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
17079 __ret = ~__rev0;
17080 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
17081 return __ret;
17082 }
17083 #endif
17084
17085 #ifdef __LITTLE_ENDIAN__
vmvnq_s8(int8x16_t __p0)17086 __ai int8x16_t vmvnq_s8(int8x16_t __p0) {
17087 int8x16_t __ret;
17088 __ret = ~__p0;
17089 return __ret;
17090 }
17091 #else
vmvnq_s8(int8x16_t __p0)17092 __ai int8x16_t vmvnq_s8(int8x16_t __p0) {
17093 int8x16_t __ret;
17094 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
17095 __ret = ~__rev0;
17096 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
17097 return __ret;
17098 }
17099 #endif
17100
17101 #ifdef __LITTLE_ENDIAN__
vmvnq_s32(int32x4_t __p0)17102 __ai int32x4_t vmvnq_s32(int32x4_t __p0) {
17103 int32x4_t __ret;
17104 __ret = ~__p0;
17105 return __ret;
17106 }
17107 #else
vmvnq_s32(int32x4_t __p0)17108 __ai int32x4_t vmvnq_s32(int32x4_t __p0) {
17109 int32x4_t __ret;
17110 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17111 __ret = ~__rev0;
17112 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
17113 return __ret;
17114 }
17115 #endif
17116
17117 #ifdef __LITTLE_ENDIAN__
vmvnq_s16(int16x8_t __p0)17118 __ai int16x8_t vmvnq_s16(int16x8_t __p0) {
17119 int16x8_t __ret;
17120 __ret = ~__p0;
17121 return __ret;
17122 }
17123 #else
vmvnq_s16(int16x8_t __p0)17124 __ai int16x8_t vmvnq_s16(int16x8_t __p0) {
17125 int16x8_t __ret;
17126 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
17127 __ret = ~__rev0;
17128 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
17129 return __ret;
17130 }
17131 #endif
17132
17133 #ifdef __LITTLE_ENDIAN__
vmvn_u8(uint8x8_t __p0)17134 __ai uint8x8_t vmvn_u8(uint8x8_t __p0) {
17135 uint8x8_t __ret;
17136 __ret = ~__p0;
17137 return __ret;
17138 }
17139 #else
vmvn_u8(uint8x8_t __p0)17140 __ai uint8x8_t vmvn_u8(uint8x8_t __p0) {
17141 uint8x8_t __ret;
17142 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
17143 __ret = ~__rev0;
17144 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
17145 return __ret;
17146 }
17147 #endif
17148
17149 #ifdef __LITTLE_ENDIAN__
vmvn_u32(uint32x2_t __p0)17150 __ai uint32x2_t vmvn_u32(uint32x2_t __p0) {
17151 uint32x2_t __ret;
17152 __ret = ~__p0;
17153 return __ret;
17154 }
17155 #else
vmvn_u32(uint32x2_t __p0)17156 __ai uint32x2_t vmvn_u32(uint32x2_t __p0) {
17157 uint32x2_t __ret;
17158 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
17159 __ret = ~__rev0;
17160 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
17161 return __ret;
17162 }
17163 #endif
17164
17165 #ifdef __LITTLE_ENDIAN__
vmvn_u16(uint16x4_t __p0)17166 __ai uint16x4_t vmvn_u16(uint16x4_t __p0) {
17167 uint16x4_t __ret;
17168 __ret = ~__p0;
17169 return __ret;
17170 }
17171 #else
vmvn_u16(uint16x4_t __p0)17172 __ai uint16x4_t vmvn_u16(uint16x4_t __p0) {
17173 uint16x4_t __ret;
17174 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17175 __ret = ~__rev0;
17176 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
17177 return __ret;
17178 }
17179 #endif
17180
17181 #ifdef __LITTLE_ENDIAN__
vmvn_s8(int8x8_t __p0)17182 __ai int8x8_t vmvn_s8(int8x8_t __p0) {
17183 int8x8_t __ret;
17184 __ret = ~__p0;
17185 return __ret;
17186 }
17187 #else
vmvn_s8(int8x8_t __p0)17188 __ai int8x8_t vmvn_s8(int8x8_t __p0) {
17189 int8x8_t __ret;
17190 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
17191 __ret = ~__rev0;
17192 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
17193 return __ret;
17194 }
17195 #endif
17196
17197 #ifdef __LITTLE_ENDIAN__
vmvn_s32(int32x2_t __p0)17198 __ai int32x2_t vmvn_s32(int32x2_t __p0) {
17199 int32x2_t __ret;
17200 __ret = ~__p0;
17201 return __ret;
17202 }
17203 #else
vmvn_s32(int32x2_t __p0)17204 __ai int32x2_t vmvn_s32(int32x2_t __p0) {
17205 int32x2_t __ret;
17206 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
17207 __ret = ~__rev0;
17208 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
17209 return __ret;
17210 }
17211 #endif
17212
17213 #ifdef __LITTLE_ENDIAN__
vmvn_s16(int16x4_t __p0)17214 __ai int16x4_t vmvn_s16(int16x4_t __p0) {
17215 int16x4_t __ret;
17216 __ret = ~__p0;
17217 return __ret;
17218 }
17219 #else
vmvn_s16(int16x4_t __p0)17220 __ai int16x4_t vmvn_s16(int16x4_t __p0) {
17221 int16x4_t __ret;
17222 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17223 __ret = ~__rev0;
17224 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
17225 return __ret;
17226 }
17227 #endif
17228
17229 #ifdef __LITTLE_ENDIAN__
vnegq_s8(int8x16_t __p0)17230 __ai int8x16_t vnegq_s8(int8x16_t __p0) {
17231 int8x16_t __ret;
17232 __ret = -__p0;
17233 return __ret;
17234 }
17235 #else
vnegq_s8(int8x16_t __p0)17236 __ai int8x16_t vnegq_s8(int8x16_t __p0) {
17237 int8x16_t __ret;
17238 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
17239 __ret = -__rev0;
17240 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
17241 return __ret;
17242 }
17243 #endif
17244
17245 #ifdef __LITTLE_ENDIAN__
vnegq_f32(float32x4_t __p0)17246 __ai float32x4_t vnegq_f32(float32x4_t __p0) {
17247 float32x4_t __ret;
17248 __ret = -__p0;
17249 return __ret;
17250 }
17251 #else
vnegq_f32(float32x4_t __p0)17252 __ai float32x4_t vnegq_f32(float32x4_t __p0) {
17253 float32x4_t __ret;
17254 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17255 __ret = -__rev0;
17256 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
17257 return __ret;
17258 }
17259 #endif
17260
17261 #ifdef __LITTLE_ENDIAN__
vnegq_s32(int32x4_t __p0)17262 __ai int32x4_t vnegq_s32(int32x4_t __p0) {
17263 int32x4_t __ret;
17264 __ret = -__p0;
17265 return __ret;
17266 }
17267 #else
vnegq_s32(int32x4_t __p0)17268 __ai int32x4_t vnegq_s32(int32x4_t __p0) {
17269 int32x4_t __ret;
17270 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17271 __ret = -__rev0;
17272 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
17273 return __ret;
17274 }
17275 #endif
17276
17277 #ifdef __LITTLE_ENDIAN__
vnegq_s16(int16x8_t __p0)17278 __ai int16x8_t vnegq_s16(int16x8_t __p0) {
17279 int16x8_t __ret;
17280 __ret = -__p0;
17281 return __ret;
17282 }
17283 #else
vnegq_s16(int16x8_t __p0)17284 __ai int16x8_t vnegq_s16(int16x8_t __p0) {
17285 int16x8_t __ret;
17286 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
17287 __ret = -__rev0;
17288 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
17289 return __ret;
17290 }
17291 #endif
17292
17293 #ifdef __LITTLE_ENDIAN__
vneg_s8(int8x8_t __p0)17294 __ai int8x8_t vneg_s8(int8x8_t __p0) {
17295 int8x8_t __ret;
17296 __ret = -__p0;
17297 return __ret;
17298 }
17299 #else
vneg_s8(int8x8_t __p0)17300 __ai int8x8_t vneg_s8(int8x8_t __p0) {
17301 int8x8_t __ret;
17302 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
17303 __ret = -__rev0;
17304 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
17305 return __ret;
17306 }
17307 #endif
17308
17309 #ifdef __LITTLE_ENDIAN__
vneg_f32(float32x2_t __p0)17310 __ai float32x2_t vneg_f32(float32x2_t __p0) {
17311 float32x2_t __ret;
17312 __ret = -__p0;
17313 return __ret;
17314 }
17315 #else
vneg_f32(float32x2_t __p0)17316 __ai float32x2_t vneg_f32(float32x2_t __p0) {
17317 float32x2_t __ret;
17318 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
17319 __ret = -__rev0;
17320 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
17321 return __ret;
17322 }
17323 #endif
17324
17325 #ifdef __LITTLE_ENDIAN__
vneg_s32(int32x2_t __p0)17326 __ai int32x2_t vneg_s32(int32x2_t __p0) {
17327 int32x2_t __ret;
17328 __ret = -__p0;
17329 return __ret;
17330 }
17331 #else
vneg_s32(int32x2_t __p0)17332 __ai int32x2_t vneg_s32(int32x2_t __p0) {
17333 int32x2_t __ret;
17334 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
17335 __ret = -__rev0;
17336 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
17337 return __ret;
17338 }
17339 #endif
17340
17341 #ifdef __LITTLE_ENDIAN__
vneg_s16(int16x4_t __p0)17342 __ai int16x4_t vneg_s16(int16x4_t __p0) {
17343 int16x4_t __ret;
17344 __ret = -__p0;
17345 return __ret;
17346 }
17347 #else
vneg_s16(int16x4_t __p0)17348 __ai int16x4_t vneg_s16(int16x4_t __p0) {
17349 int16x4_t __ret;
17350 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17351 __ret = -__rev0;
17352 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
17353 return __ret;
17354 }
17355 #endif
17356
17357 #ifdef __LITTLE_ENDIAN__
vornq_u8(uint8x16_t __p0,uint8x16_t __p1)17358 __ai uint8x16_t vornq_u8(uint8x16_t __p0, uint8x16_t __p1) {
17359 uint8x16_t __ret;
17360 __ret = __p0 | ~__p1;
17361 return __ret;
17362 }
17363 #else
vornq_u8(uint8x16_t __p0,uint8x16_t __p1)17364 __ai uint8x16_t vornq_u8(uint8x16_t __p0, uint8x16_t __p1) {
17365 uint8x16_t __ret;
17366 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
17367 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
17368 __ret = __rev0 | ~__rev1;
17369 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
17370 return __ret;
17371 }
17372 #endif
17373
17374 #ifdef __LITTLE_ENDIAN__
vornq_u32(uint32x4_t __p0,uint32x4_t __p1)17375 __ai uint32x4_t vornq_u32(uint32x4_t __p0, uint32x4_t __p1) {
17376 uint32x4_t __ret;
17377 __ret = __p0 | ~__p1;
17378 return __ret;
17379 }
17380 #else
vornq_u32(uint32x4_t __p0,uint32x4_t __p1)17381 __ai uint32x4_t vornq_u32(uint32x4_t __p0, uint32x4_t __p1) {
17382 uint32x4_t __ret;
17383 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17384 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
17385 __ret = __rev0 | ~__rev1;
17386 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
17387 return __ret;
17388 }
17389 #endif
17390
17391 #ifdef __LITTLE_ENDIAN__
vornq_u64(uint64x2_t __p0,uint64x2_t __p1)17392 __ai uint64x2_t vornq_u64(uint64x2_t __p0, uint64x2_t __p1) {
17393 uint64x2_t __ret;
17394 __ret = __p0 | ~__p1;
17395 return __ret;
17396 }
17397 #else
vornq_u64(uint64x2_t __p0,uint64x2_t __p1)17398 __ai uint64x2_t vornq_u64(uint64x2_t __p0, uint64x2_t __p1) {
17399 uint64x2_t __ret;
17400 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
17401 uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
17402 __ret = __rev0 | ~__rev1;
17403 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
17404 return __ret;
17405 }
17406 #endif
17407
17408 #ifdef __LITTLE_ENDIAN__
vornq_u16(uint16x8_t __p0,uint16x8_t __p1)17409 __ai uint16x8_t vornq_u16(uint16x8_t __p0, uint16x8_t __p1) {
17410 uint16x8_t __ret;
17411 __ret = __p0 | ~__p1;
17412 return __ret;
17413 }
17414 #else
vornq_u16(uint16x8_t __p0,uint16x8_t __p1)17415 __ai uint16x8_t vornq_u16(uint16x8_t __p0, uint16x8_t __p1) {
17416 uint16x8_t __ret;
17417 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
17418 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
17419 __ret = __rev0 | ~__rev1;
17420 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
17421 return __ret;
17422 }
17423 #endif
17424
17425 #ifdef __LITTLE_ENDIAN__
vornq_s8(int8x16_t __p0,int8x16_t __p1)17426 __ai int8x16_t vornq_s8(int8x16_t __p0, int8x16_t __p1) {
17427 int8x16_t __ret;
17428 __ret = __p0 | ~__p1;
17429 return __ret;
17430 }
17431 #else
vornq_s8(int8x16_t __p0,int8x16_t __p1)17432 __ai int8x16_t vornq_s8(int8x16_t __p0, int8x16_t __p1) {
17433 int8x16_t __ret;
17434 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
17435 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
17436 __ret = __rev0 | ~__rev1;
17437 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
17438 return __ret;
17439 }
17440 #endif
17441
17442 #ifdef __LITTLE_ENDIAN__
vornq_s32(int32x4_t __p0,int32x4_t __p1)17443 __ai int32x4_t vornq_s32(int32x4_t __p0, int32x4_t __p1) {
17444 int32x4_t __ret;
17445 __ret = __p0 | ~__p1;
17446 return __ret;
17447 }
17448 #else
vornq_s32(int32x4_t __p0,int32x4_t __p1)17449 __ai int32x4_t vornq_s32(int32x4_t __p0, int32x4_t __p1) {
17450 int32x4_t __ret;
17451 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17452 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
17453 __ret = __rev0 | ~__rev1;
17454 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
17455 return __ret;
17456 }
17457 #endif
17458
17459 #ifdef __LITTLE_ENDIAN__
vornq_s64(int64x2_t __p0,int64x2_t __p1)17460 __ai int64x2_t vornq_s64(int64x2_t __p0, int64x2_t __p1) {
17461 int64x2_t __ret;
17462 __ret = __p0 | ~__p1;
17463 return __ret;
17464 }
17465 #else
vornq_s64(int64x2_t __p0,int64x2_t __p1)17466 __ai int64x2_t vornq_s64(int64x2_t __p0, int64x2_t __p1) {
17467 int64x2_t __ret;
17468 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
17469 int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
17470 __ret = __rev0 | ~__rev1;
17471 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
17472 return __ret;
17473 }
17474 #endif
17475
17476 #ifdef __LITTLE_ENDIAN__
vornq_s16(int16x8_t __p0,int16x8_t __p1)17477 __ai int16x8_t vornq_s16(int16x8_t __p0, int16x8_t __p1) {
17478 int16x8_t __ret;
17479 __ret = __p0 | ~__p1;
17480 return __ret;
17481 }
17482 #else
vornq_s16(int16x8_t __p0,int16x8_t __p1)17483 __ai int16x8_t vornq_s16(int16x8_t __p0, int16x8_t __p1) {
17484 int16x8_t __ret;
17485 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
17486 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
17487 __ret = __rev0 | ~__rev1;
17488 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
17489 return __ret;
17490 }
17491 #endif
17492
17493 #ifdef __LITTLE_ENDIAN__
vorn_u8(uint8x8_t __p0,uint8x8_t __p1)17494 __ai uint8x8_t vorn_u8(uint8x8_t __p0, uint8x8_t __p1) {
17495 uint8x8_t __ret;
17496 __ret = __p0 | ~__p1;
17497 return __ret;
17498 }
17499 #else
vorn_u8(uint8x8_t __p0,uint8x8_t __p1)17500 __ai uint8x8_t vorn_u8(uint8x8_t __p0, uint8x8_t __p1) {
17501 uint8x8_t __ret;
17502 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
17503 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
17504 __ret = __rev0 | ~__rev1;
17505 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
17506 return __ret;
17507 }
17508 #endif
17509
17510 #ifdef __LITTLE_ENDIAN__
vorn_u32(uint32x2_t __p0,uint32x2_t __p1)17511 __ai uint32x2_t vorn_u32(uint32x2_t __p0, uint32x2_t __p1) {
17512 uint32x2_t __ret;
17513 __ret = __p0 | ~__p1;
17514 return __ret;
17515 }
17516 #else
vorn_u32(uint32x2_t __p0,uint32x2_t __p1)17517 __ai uint32x2_t vorn_u32(uint32x2_t __p0, uint32x2_t __p1) {
17518 uint32x2_t __ret;
17519 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
17520 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
17521 __ret = __rev0 | ~__rev1;
17522 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
17523 return __ret;
17524 }
17525 #endif
17526
vorn_u64(uint64x1_t __p0,uint64x1_t __p1)17527 __ai uint64x1_t vorn_u64(uint64x1_t __p0, uint64x1_t __p1) {
17528 uint64x1_t __ret;
17529 __ret = __p0 | ~__p1;
17530 return __ret;
17531 }
17532 #ifdef __LITTLE_ENDIAN__
vorn_u16(uint16x4_t __p0,uint16x4_t __p1)17533 __ai uint16x4_t vorn_u16(uint16x4_t __p0, uint16x4_t __p1) {
17534 uint16x4_t __ret;
17535 __ret = __p0 | ~__p1;
17536 return __ret;
17537 }
17538 #else
vorn_u16(uint16x4_t __p0,uint16x4_t __p1)17539 __ai uint16x4_t vorn_u16(uint16x4_t __p0, uint16x4_t __p1) {
17540 uint16x4_t __ret;
17541 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17542 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
17543 __ret = __rev0 | ~__rev1;
17544 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
17545 return __ret;
17546 }
17547 #endif
17548
17549 #ifdef __LITTLE_ENDIAN__
vorn_s8(int8x8_t __p0,int8x8_t __p1)17550 __ai int8x8_t vorn_s8(int8x8_t __p0, int8x8_t __p1) {
17551 int8x8_t __ret;
17552 __ret = __p0 | ~__p1;
17553 return __ret;
17554 }
17555 #else
vorn_s8(int8x8_t __p0,int8x8_t __p1)17556 __ai int8x8_t vorn_s8(int8x8_t __p0, int8x8_t __p1) {
17557 int8x8_t __ret;
17558 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
17559 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
17560 __ret = __rev0 | ~__rev1;
17561 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
17562 return __ret;
17563 }
17564 #endif
17565
17566 #ifdef __LITTLE_ENDIAN__
vorn_s32(int32x2_t __p0,int32x2_t __p1)17567 __ai int32x2_t vorn_s32(int32x2_t __p0, int32x2_t __p1) {
17568 int32x2_t __ret;
17569 __ret = __p0 | ~__p1;
17570 return __ret;
17571 }
17572 #else
vorn_s32(int32x2_t __p0,int32x2_t __p1)17573 __ai int32x2_t vorn_s32(int32x2_t __p0, int32x2_t __p1) {
17574 int32x2_t __ret;
17575 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
17576 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
17577 __ret = __rev0 | ~__rev1;
17578 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
17579 return __ret;
17580 }
17581 #endif
17582
vorn_s64(int64x1_t __p0,int64x1_t __p1)17583 __ai int64x1_t vorn_s64(int64x1_t __p0, int64x1_t __p1) {
17584 int64x1_t __ret;
17585 __ret = __p0 | ~__p1;
17586 return __ret;
17587 }
17588 #ifdef __LITTLE_ENDIAN__
vorn_s16(int16x4_t __p0,int16x4_t __p1)17589 __ai int16x4_t vorn_s16(int16x4_t __p0, int16x4_t __p1) {
17590 int16x4_t __ret;
17591 __ret = __p0 | ~__p1;
17592 return __ret;
17593 }
17594 #else
vorn_s16(int16x4_t __p0,int16x4_t __p1)17595 __ai int16x4_t vorn_s16(int16x4_t __p0, int16x4_t __p1) {
17596 int16x4_t __ret;
17597 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17598 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
17599 __ret = __rev0 | ~__rev1;
17600 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
17601 return __ret;
17602 }
17603 #endif
17604
17605 #ifdef __LITTLE_ENDIAN__
vorrq_u8(uint8x16_t __p0,uint8x16_t __p1)17606 __ai uint8x16_t vorrq_u8(uint8x16_t __p0, uint8x16_t __p1) {
17607 uint8x16_t __ret;
17608 __ret = __p0 | __p1;
17609 return __ret;
17610 }
17611 #else
vorrq_u8(uint8x16_t __p0,uint8x16_t __p1)17612 __ai uint8x16_t vorrq_u8(uint8x16_t __p0, uint8x16_t __p1) {
17613 uint8x16_t __ret;
17614 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
17615 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
17616 __ret = __rev0 | __rev1;
17617 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
17618 return __ret;
17619 }
17620 #endif
17621
17622 #ifdef __LITTLE_ENDIAN__
vorrq_u32(uint32x4_t __p0,uint32x4_t __p1)17623 __ai uint32x4_t vorrq_u32(uint32x4_t __p0, uint32x4_t __p1) {
17624 uint32x4_t __ret;
17625 __ret = __p0 | __p1;
17626 return __ret;
17627 }
17628 #else
vorrq_u32(uint32x4_t __p0,uint32x4_t __p1)17629 __ai uint32x4_t vorrq_u32(uint32x4_t __p0, uint32x4_t __p1) {
17630 uint32x4_t __ret;
17631 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17632 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
17633 __ret = __rev0 | __rev1;
17634 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
17635 return __ret;
17636 }
17637 #endif
17638
17639 #ifdef __LITTLE_ENDIAN__
vorrq_u64(uint64x2_t __p0,uint64x2_t __p1)17640 __ai uint64x2_t vorrq_u64(uint64x2_t __p0, uint64x2_t __p1) {
17641 uint64x2_t __ret;
17642 __ret = __p0 | __p1;
17643 return __ret;
17644 }
17645 #else
vorrq_u64(uint64x2_t __p0,uint64x2_t __p1)17646 __ai uint64x2_t vorrq_u64(uint64x2_t __p0, uint64x2_t __p1) {
17647 uint64x2_t __ret;
17648 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
17649 uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
17650 __ret = __rev0 | __rev1;
17651 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
17652 return __ret;
17653 }
17654 #endif
17655
17656 #ifdef __LITTLE_ENDIAN__
vorrq_u16(uint16x8_t __p0,uint16x8_t __p1)17657 __ai uint16x8_t vorrq_u16(uint16x8_t __p0, uint16x8_t __p1) {
17658 uint16x8_t __ret;
17659 __ret = __p0 | __p1;
17660 return __ret;
17661 }
17662 #else
vorrq_u16(uint16x8_t __p0,uint16x8_t __p1)17663 __ai uint16x8_t vorrq_u16(uint16x8_t __p0, uint16x8_t __p1) {
17664 uint16x8_t __ret;
17665 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
17666 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
17667 __ret = __rev0 | __rev1;
17668 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
17669 return __ret;
17670 }
17671 #endif
17672
17673 #ifdef __LITTLE_ENDIAN__
vorrq_s8(int8x16_t __p0,int8x16_t __p1)17674 __ai int8x16_t vorrq_s8(int8x16_t __p0, int8x16_t __p1) {
17675 int8x16_t __ret;
17676 __ret = __p0 | __p1;
17677 return __ret;
17678 }
17679 #else
vorrq_s8(int8x16_t __p0,int8x16_t __p1)17680 __ai int8x16_t vorrq_s8(int8x16_t __p0, int8x16_t __p1) {
17681 int8x16_t __ret;
17682 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
17683 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
17684 __ret = __rev0 | __rev1;
17685 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
17686 return __ret;
17687 }
17688 #endif
17689
17690 #ifdef __LITTLE_ENDIAN__
vorrq_s32(int32x4_t __p0,int32x4_t __p1)17691 __ai int32x4_t vorrq_s32(int32x4_t __p0, int32x4_t __p1) {
17692 int32x4_t __ret;
17693 __ret = __p0 | __p1;
17694 return __ret;
17695 }
17696 #else
vorrq_s32(int32x4_t __p0,int32x4_t __p1)17697 __ai int32x4_t vorrq_s32(int32x4_t __p0, int32x4_t __p1) {
17698 int32x4_t __ret;
17699 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17700 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
17701 __ret = __rev0 | __rev1;
17702 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
17703 return __ret;
17704 }
17705 #endif
17706
17707 #ifdef __LITTLE_ENDIAN__
vorrq_s64(int64x2_t __p0,int64x2_t __p1)17708 __ai int64x2_t vorrq_s64(int64x2_t __p0, int64x2_t __p1) {
17709 int64x2_t __ret;
17710 __ret = __p0 | __p1;
17711 return __ret;
17712 }
17713 #else
vorrq_s64(int64x2_t __p0,int64x2_t __p1)17714 __ai int64x2_t vorrq_s64(int64x2_t __p0, int64x2_t __p1) {
17715 int64x2_t __ret;
17716 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
17717 int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
17718 __ret = __rev0 | __rev1;
17719 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
17720 return __ret;
17721 }
17722 #endif
17723
17724 #ifdef __LITTLE_ENDIAN__
vorrq_s16(int16x8_t __p0,int16x8_t __p1)17725 __ai int16x8_t vorrq_s16(int16x8_t __p0, int16x8_t __p1) {
17726 int16x8_t __ret;
17727 __ret = __p0 | __p1;
17728 return __ret;
17729 }
17730 #else
vorrq_s16(int16x8_t __p0,int16x8_t __p1)17731 __ai int16x8_t vorrq_s16(int16x8_t __p0, int16x8_t __p1) {
17732 int16x8_t __ret;
17733 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
17734 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
17735 __ret = __rev0 | __rev1;
17736 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
17737 return __ret;
17738 }
17739 #endif
17740
17741 #ifdef __LITTLE_ENDIAN__
vorr_u8(uint8x8_t __p0,uint8x8_t __p1)17742 __ai uint8x8_t vorr_u8(uint8x8_t __p0, uint8x8_t __p1) {
17743 uint8x8_t __ret;
17744 __ret = __p0 | __p1;
17745 return __ret;
17746 }
17747 #else
vorr_u8(uint8x8_t __p0,uint8x8_t __p1)17748 __ai uint8x8_t vorr_u8(uint8x8_t __p0, uint8x8_t __p1) {
17749 uint8x8_t __ret;
17750 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
17751 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
17752 __ret = __rev0 | __rev1;
17753 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
17754 return __ret;
17755 }
17756 #endif
17757
17758 #ifdef __LITTLE_ENDIAN__
vorr_u32(uint32x2_t __p0,uint32x2_t __p1)17759 __ai uint32x2_t vorr_u32(uint32x2_t __p0, uint32x2_t __p1) {
17760 uint32x2_t __ret;
17761 __ret = __p0 | __p1;
17762 return __ret;
17763 }
17764 #else
vorr_u32(uint32x2_t __p0,uint32x2_t __p1)17765 __ai uint32x2_t vorr_u32(uint32x2_t __p0, uint32x2_t __p1) {
17766 uint32x2_t __ret;
17767 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
17768 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
17769 __ret = __rev0 | __rev1;
17770 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
17771 return __ret;
17772 }
17773 #endif
17774
vorr_u64(uint64x1_t __p0,uint64x1_t __p1)17775 __ai uint64x1_t vorr_u64(uint64x1_t __p0, uint64x1_t __p1) {
17776 uint64x1_t __ret;
17777 __ret = __p0 | __p1;
17778 return __ret;
17779 }
17780 #ifdef __LITTLE_ENDIAN__
vorr_u16(uint16x4_t __p0,uint16x4_t __p1)17781 __ai uint16x4_t vorr_u16(uint16x4_t __p0, uint16x4_t __p1) {
17782 uint16x4_t __ret;
17783 __ret = __p0 | __p1;
17784 return __ret;
17785 }
17786 #else
vorr_u16(uint16x4_t __p0,uint16x4_t __p1)17787 __ai uint16x4_t vorr_u16(uint16x4_t __p0, uint16x4_t __p1) {
17788 uint16x4_t __ret;
17789 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17790 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
17791 __ret = __rev0 | __rev1;
17792 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
17793 return __ret;
17794 }
17795 #endif
17796
17797 #ifdef __LITTLE_ENDIAN__
vorr_s8(int8x8_t __p0,int8x8_t __p1)17798 __ai int8x8_t vorr_s8(int8x8_t __p0, int8x8_t __p1) {
17799 int8x8_t __ret;
17800 __ret = __p0 | __p1;
17801 return __ret;
17802 }
17803 #else
vorr_s8(int8x8_t __p0,int8x8_t __p1)17804 __ai int8x8_t vorr_s8(int8x8_t __p0, int8x8_t __p1) {
17805 int8x8_t __ret;
17806 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
17807 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
17808 __ret = __rev0 | __rev1;
17809 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
17810 return __ret;
17811 }
17812 #endif
17813
17814 #ifdef __LITTLE_ENDIAN__
vorr_s32(int32x2_t __p0,int32x2_t __p1)17815 __ai int32x2_t vorr_s32(int32x2_t __p0, int32x2_t __p1) {
17816 int32x2_t __ret;
17817 __ret = __p0 | __p1;
17818 return __ret;
17819 }
17820 #else
vorr_s32(int32x2_t __p0,int32x2_t __p1)17821 __ai int32x2_t vorr_s32(int32x2_t __p0, int32x2_t __p1) {
17822 int32x2_t __ret;
17823 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
17824 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
17825 __ret = __rev0 | __rev1;
17826 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
17827 return __ret;
17828 }
17829 #endif
17830
vorr_s64(int64x1_t __p0,int64x1_t __p1)17831 __ai int64x1_t vorr_s64(int64x1_t __p0, int64x1_t __p1) {
17832 int64x1_t __ret;
17833 __ret = __p0 | __p1;
17834 return __ret;
17835 }
17836 #ifdef __LITTLE_ENDIAN__
vorr_s16(int16x4_t __p0,int16x4_t __p1)17837 __ai int16x4_t vorr_s16(int16x4_t __p0, int16x4_t __p1) {
17838 int16x4_t __ret;
17839 __ret = __p0 | __p1;
17840 return __ret;
17841 }
17842 #else
vorr_s16(int16x4_t __p0,int16x4_t __p1)17843 __ai int16x4_t vorr_s16(int16x4_t __p0, int16x4_t __p1) {
17844 int16x4_t __ret;
17845 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17846 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
17847 __ret = __rev0 | __rev1;
17848 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
17849 return __ret;
17850 }
17851 #endif
17852
17853 #ifdef __LITTLE_ENDIAN__
vpadalq_u8(uint16x8_t __p0,uint8x16_t __p1)17854 __ai uint16x8_t vpadalq_u8(uint16x8_t __p0, uint8x16_t __p1) {
17855 uint16x8_t __ret;
17856 __ret = (uint16x8_t) __builtin_neon_vpadalq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
17857 return __ret;
17858 }
17859 #else
vpadalq_u8(uint16x8_t __p0,uint8x16_t __p1)17860 __ai uint16x8_t vpadalq_u8(uint16x8_t __p0, uint8x16_t __p1) {
17861 uint16x8_t __ret;
17862 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
17863 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
17864 __ret = (uint16x8_t) __builtin_neon_vpadalq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
17865 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
17866 return __ret;
17867 }
17868 #endif
17869
17870 #ifdef __LITTLE_ENDIAN__
vpadalq_u32(uint64x2_t __p0,uint32x4_t __p1)17871 __ai uint64x2_t vpadalq_u32(uint64x2_t __p0, uint32x4_t __p1) {
17872 uint64x2_t __ret;
17873 __ret = (uint64x2_t) __builtin_neon_vpadalq_v((int8x16_t)__p0, (int8x16_t)__p1, 51);
17874 return __ret;
17875 }
17876 #else
vpadalq_u32(uint64x2_t __p0,uint32x4_t __p1)17877 __ai uint64x2_t vpadalq_u32(uint64x2_t __p0, uint32x4_t __p1) {
17878 uint64x2_t __ret;
17879 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
17880 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
17881 __ret = (uint64x2_t) __builtin_neon_vpadalq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 51);
17882 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
17883 return __ret;
17884 }
17885 #endif
17886
17887 #ifdef __LITTLE_ENDIAN__
vpadalq_u16(uint32x4_t __p0,uint16x8_t __p1)17888 __ai uint32x4_t vpadalq_u16(uint32x4_t __p0, uint16x8_t __p1) {
17889 uint32x4_t __ret;
17890 __ret = (uint32x4_t) __builtin_neon_vpadalq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
17891 return __ret;
17892 }
17893 #else
vpadalq_u16(uint32x4_t __p0,uint16x8_t __p1)17894 __ai uint32x4_t vpadalq_u16(uint32x4_t __p0, uint16x8_t __p1) {
17895 uint32x4_t __ret;
17896 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17897 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
17898 __ret = (uint32x4_t) __builtin_neon_vpadalq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
17899 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
17900 return __ret;
17901 }
17902 #endif
17903
17904 #ifdef __LITTLE_ENDIAN__
vpadalq_s8(int16x8_t __p0,int8x16_t __p1)17905 __ai int16x8_t vpadalq_s8(int16x8_t __p0, int8x16_t __p1) {
17906 int16x8_t __ret;
17907 __ret = (int16x8_t) __builtin_neon_vpadalq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
17908 return __ret;
17909 }
17910 #else
vpadalq_s8(int16x8_t __p0,int8x16_t __p1)17911 __ai int16x8_t vpadalq_s8(int16x8_t __p0, int8x16_t __p1) {
17912 int16x8_t __ret;
17913 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
17914 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
17915 __ret = (int16x8_t) __builtin_neon_vpadalq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
17916 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
17917 return __ret;
17918 }
17919 #endif
17920
17921 #ifdef __LITTLE_ENDIAN__
vpadalq_s32(int64x2_t __p0,int32x4_t __p1)17922 __ai int64x2_t vpadalq_s32(int64x2_t __p0, int32x4_t __p1) {
17923 int64x2_t __ret;
17924 __ret = (int64x2_t) __builtin_neon_vpadalq_v((int8x16_t)__p0, (int8x16_t)__p1, 35);
17925 return __ret;
17926 }
17927 #else
vpadalq_s32(int64x2_t __p0,int32x4_t __p1)17928 __ai int64x2_t vpadalq_s32(int64x2_t __p0, int32x4_t __p1) {
17929 int64x2_t __ret;
17930 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
17931 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
17932 __ret = (int64x2_t) __builtin_neon_vpadalq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 35);
17933 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
17934 return __ret;
17935 }
17936 #endif
17937
17938 #ifdef __LITTLE_ENDIAN__
vpadalq_s16(int32x4_t __p0,int16x8_t __p1)17939 __ai int32x4_t vpadalq_s16(int32x4_t __p0, int16x8_t __p1) {
17940 int32x4_t __ret;
17941 __ret = (int32x4_t) __builtin_neon_vpadalq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
17942 return __ret;
17943 }
17944 #else
vpadalq_s16(int32x4_t __p0,int16x8_t __p1)17945 __ai int32x4_t vpadalq_s16(int32x4_t __p0, int16x8_t __p1) {
17946 int32x4_t __ret;
17947 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17948 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
17949 __ret = (int32x4_t) __builtin_neon_vpadalq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
17950 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
17951 return __ret;
17952 }
17953 #endif
17954
17955 #ifdef __LITTLE_ENDIAN__
vpadal_u8(uint16x4_t __p0,uint8x8_t __p1)17956 __ai uint16x4_t vpadal_u8(uint16x4_t __p0, uint8x8_t __p1) {
17957 uint16x4_t __ret;
17958 __ret = (uint16x4_t) __builtin_neon_vpadal_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
17959 return __ret;
17960 }
17961 #else
vpadal_u8(uint16x4_t __p0,uint8x8_t __p1)17962 __ai uint16x4_t vpadal_u8(uint16x4_t __p0, uint8x8_t __p1) {
17963 uint16x4_t __ret;
17964 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
17965 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
17966 __ret = (uint16x4_t) __builtin_neon_vpadal_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
17967 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
17968 return __ret;
17969 }
17970 #endif
17971
17972 #ifdef __LITTLE_ENDIAN__
vpadal_u32(uint64x1_t __p0,uint32x2_t __p1)17973 __ai uint64x1_t vpadal_u32(uint64x1_t __p0, uint32x2_t __p1) {
17974 uint64x1_t __ret;
17975 __ret = (uint64x1_t) __builtin_neon_vpadal_v((int8x8_t)__p0, (int8x8_t)__p1, 19);
17976 return __ret;
17977 }
17978 #else
vpadal_u32(uint64x1_t __p0,uint32x2_t __p1)17979 __ai uint64x1_t vpadal_u32(uint64x1_t __p0, uint32x2_t __p1) {
17980 uint64x1_t __ret;
17981 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
17982 __ret = (uint64x1_t) __builtin_neon_vpadal_v((int8x8_t)__p0, (int8x8_t)__rev1, 19);
17983 return __ret;
17984 }
17985 #endif
17986
17987 #ifdef __LITTLE_ENDIAN__
vpadal_u16(uint32x2_t __p0,uint16x4_t __p1)17988 __ai uint32x2_t vpadal_u16(uint32x2_t __p0, uint16x4_t __p1) {
17989 uint32x2_t __ret;
17990 __ret = (uint32x2_t) __builtin_neon_vpadal_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
17991 return __ret;
17992 }
17993 #else
vpadal_u16(uint32x2_t __p0,uint16x4_t __p1)17994 __ai uint32x2_t vpadal_u16(uint32x2_t __p0, uint16x4_t __p1) {
17995 uint32x2_t __ret;
17996 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
17997 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
17998 __ret = (uint32x2_t) __builtin_neon_vpadal_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
17999 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
18000 return __ret;
18001 }
18002 #endif
18003
18004 #ifdef __LITTLE_ENDIAN__
vpadal_s8(int16x4_t __p0,int8x8_t __p1)18005 __ai int16x4_t vpadal_s8(int16x4_t __p0, int8x8_t __p1) {
18006 int16x4_t __ret;
18007 __ret = (int16x4_t) __builtin_neon_vpadal_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
18008 return __ret;
18009 }
18010 #else
vpadal_s8(int16x4_t __p0,int8x8_t __p1)18011 __ai int16x4_t vpadal_s8(int16x4_t __p0, int8x8_t __p1) {
18012 int16x4_t __ret;
18013 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
18014 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
18015 __ret = (int16x4_t) __builtin_neon_vpadal_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
18016 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
18017 return __ret;
18018 }
18019 #endif
18020
18021 #ifdef __LITTLE_ENDIAN__
vpadal_s32(int64x1_t __p0,int32x2_t __p1)18022 __ai int64x1_t vpadal_s32(int64x1_t __p0, int32x2_t __p1) {
18023 int64x1_t __ret;
18024 __ret = (int64x1_t) __builtin_neon_vpadal_v((int8x8_t)__p0, (int8x8_t)__p1, 3);
18025 return __ret;
18026 }
18027 #else
vpadal_s32(int64x1_t __p0,int32x2_t __p1)18028 __ai int64x1_t vpadal_s32(int64x1_t __p0, int32x2_t __p1) {
18029 int64x1_t __ret;
18030 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
18031 __ret = (int64x1_t) __builtin_neon_vpadal_v((int8x8_t)__p0, (int8x8_t)__rev1, 3);
18032 return __ret;
18033 }
18034 #endif
18035
18036 #ifdef __LITTLE_ENDIAN__
vpadal_s16(int32x2_t __p0,int16x4_t __p1)18037 __ai int32x2_t vpadal_s16(int32x2_t __p0, int16x4_t __p1) {
18038 int32x2_t __ret;
18039 __ret = (int32x2_t) __builtin_neon_vpadal_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
18040 return __ret;
18041 }
18042 #else
vpadal_s16(int32x2_t __p0,int16x4_t __p1)18043 __ai int32x2_t vpadal_s16(int32x2_t __p0, int16x4_t __p1) {
18044 int32x2_t __ret;
18045 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
18046 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
18047 __ret = (int32x2_t) __builtin_neon_vpadal_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
18048 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
18049 return __ret;
18050 }
18051 #endif
18052
18053 #ifdef __LITTLE_ENDIAN__
vpadd_u8(uint8x8_t __p0,uint8x8_t __p1)18054 __ai uint8x8_t vpadd_u8(uint8x8_t __p0, uint8x8_t __p1) {
18055 uint8x8_t __ret;
18056 __ret = (uint8x8_t) __builtin_neon_vpadd_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
18057 return __ret;
18058 }
18059 #else
vpadd_u8(uint8x8_t __p0,uint8x8_t __p1)18060 __ai uint8x8_t vpadd_u8(uint8x8_t __p0, uint8x8_t __p1) {
18061 uint8x8_t __ret;
18062 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
18063 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
18064 __ret = (uint8x8_t) __builtin_neon_vpadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
18065 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
18066 return __ret;
18067 }
18068 #endif
18069
18070 #ifdef __LITTLE_ENDIAN__
vpadd_u32(uint32x2_t __p0,uint32x2_t __p1)18071 __ai uint32x2_t vpadd_u32(uint32x2_t __p0, uint32x2_t __p1) {
18072 uint32x2_t __ret;
18073 __ret = (uint32x2_t) __builtin_neon_vpadd_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
18074 return __ret;
18075 }
18076 #else
vpadd_u32(uint32x2_t __p0,uint32x2_t __p1)18077 __ai uint32x2_t vpadd_u32(uint32x2_t __p0, uint32x2_t __p1) {
18078 uint32x2_t __ret;
18079 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
18080 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
18081 __ret = (uint32x2_t) __builtin_neon_vpadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
18082 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
18083 return __ret;
18084 }
18085 #endif
18086
18087 #ifdef __LITTLE_ENDIAN__
vpadd_u16(uint16x4_t __p0,uint16x4_t __p1)18088 __ai uint16x4_t vpadd_u16(uint16x4_t __p0, uint16x4_t __p1) {
18089 uint16x4_t __ret;
18090 __ret = (uint16x4_t) __builtin_neon_vpadd_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
18091 return __ret;
18092 }
18093 #else
vpadd_u16(uint16x4_t __p0,uint16x4_t __p1)18094 __ai uint16x4_t vpadd_u16(uint16x4_t __p0, uint16x4_t __p1) {
18095 uint16x4_t __ret;
18096 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
18097 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
18098 __ret = (uint16x4_t) __builtin_neon_vpadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
18099 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
18100 return __ret;
18101 }
18102 #endif
18103
18104 #ifdef __LITTLE_ENDIAN__
vpadd_s8(int8x8_t __p0,int8x8_t __p1)18105 __ai int8x8_t vpadd_s8(int8x8_t __p0, int8x8_t __p1) {
18106 int8x8_t __ret;
18107 __ret = (int8x8_t) __builtin_neon_vpadd_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
18108 return __ret;
18109 }
18110 #else
vpadd_s8(int8x8_t __p0,int8x8_t __p1)18111 __ai int8x8_t vpadd_s8(int8x8_t __p0, int8x8_t __p1) {
18112 int8x8_t __ret;
18113 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
18114 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
18115 __ret = (int8x8_t) __builtin_neon_vpadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 0);
18116 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
18117 return __ret;
18118 }
18119 #endif
18120
18121 #ifdef __LITTLE_ENDIAN__
vpadd_f32(float32x2_t __p0,float32x2_t __p1)18122 __ai float32x2_t vpadd_f32(float32x2_t __p0, float32x2_t __p1) {
18123 float32x2_t __ret;
18124 __ret = (float32x2_t) __builtin_neon_vpadd_v((int8x8_t)__p0, (int8x8_t)__p1, 9);
18125 return __ret;
18126 }
18127 #else
vpadd_f32(float32x2_t __p0,float32x2_t __p1)18128 __ai float32x2_t vpadd_f32(float32x2_t __p0, float32x2_t __p1) {
18129 float32x2_t __ret;
18130 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
18131 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
18132 __ret = (float32x2_t) __builtin_neon_vpadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 9);
18133 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
18134 return __ret;
18135 }
18136 #endif
18137
18138 #ifdef __LITTLE_ENDIAN__
vpadd_s32(int32x2_t __p0,int32x2_t __p1)18139 __ai int32x2_t vpadd_s32(int32x2_t __p0, int32x2_t __p1) {
18140 int32x2_t __ret;
18141 __ret = (int32x2_t) __builtin_neon_vpadd_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
18142 return __ret;
18143 }
18144 #else
vpadd_s32(int32x2_t __p0,int32x2_t __p1)18145 __ai int32x2_t vpadd_s32(int32x2_t __p0, int32x2_t __p1) {
18146 int32x2_t __ret;
18147 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
18148 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
18149 __ret = (int32x2_t) __builtin_neon_vpadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
18150 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
18151 return __ret;
18152 }
18153 #endif
18154
18155 #ifdef __LITTLE_ENDIAN__
vpadd_s16(int16x4_t __p0,int16x4_t __p1)18156 __ai int16x4_t vpadd_s16(int16x4_t __p0, int16x4_t __p1) {
18157 int16x4_t __ret;
18158 __ret = (int16x4_t) __builtin_neon_vpadd_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
18159 return __ret;
18160 }
18161 #else
vpadd_s16(int16x4_t __p0,int16x4_t __p1)18162 __ai int16x4_t vpadd_s16(int16x4_t __p0, int16x4_t __p1) {
18163 int16x4_t __ret;
18164 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
18165 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
18166 __ret = (int16x4_t) __builtin_neon_vpadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
18167 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
18168 return __ret;
18169 }
18170 #endif
18171
18172 #ifdef __LITTLE_ENDIAN__
vpaddlq_u8(uint8x16_t __p0)18173 __ai uint16x8_t vpaddlq_u8(uint8x16_t __p0) {
18174 uint16x8_t __ret;
18175 __ret = (uint16x8_t) __builtin_neon_vpaddlq_v((int8x16_t)__p0, 49);
18176 return __ret;
18177 }
18178 #else
vpaddlq_u8(uint8x16_t __p0)18179 __ai uint16x8_t vpaddlq_u8(uint8x16_t __p0) {
18180 uint16x8_t __ret;
18181 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
18182 __ret = (uint16x8_t) __builtin_neon_vpaddlq_v((int8x16_t)__rev0, 49);
18183 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
18184 return __ret;
18185 }
18186 #endif
18187
18188 #ifdef __LITTLE_ENDIAN__
vpaddlq_u32(uint32x4_t __p0)18189 __ai uint64x2_t vpaddlq_u32(uint32x4_t __p0) {
18190 uint64x2_t __ret;
18191 __ret = (uint64x2_t) __builtin_neon_vpaddlq_v((int8x16_t)__p0, 51);
18192 return __ret;
18193 }
18194 #else
vpaddlq_u32(uint32x4_t __p0)18195 __ai uint64x2_t vpaddlq_u32(uint32x4_t __p0) {
18196 uint64x2_t __ret;
18197 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
18198 __ret = (uint64x2_t) __builtin_neon_vpaddlq_v((int8x16_t)__rev0, 51);
18199 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
18200 return __ret;
18201 }
18202 #endif
18203
18204 #ifdef __LITTLE_ENDIAN__
vpaddlq_u16(uint16x8_t __p0)18205 __ai uint32x4_t vpaddlq_u16(uint16x8_t __p0) {
18206 uint32x4_t __ret;
18207 __ret = (uint32x4_t) __builtin_neon_vpaddlq_v((int8x16_t)__p0, 50);
18208 return __ret;
18209 }
18210 #else
vpaddlq_u16(uint16x8_t __p0)18211 __ai uint32x4_t vpaddlq_u16(uint16x8_t __p0) {
18212 uint32x4_t __ret;
18213 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
18214 __ret = (uint32x4_t) __builtin_neon_vpaddlq_v((int8x16_t)__rev0, 50);
18215 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
18216 return __ret;
18217 }
18218 #endif
18219
18220 #ifdef __LITTLE_ENDIAN__
vpaddlq_s8(int8x16_t __p0)18221 __ai int16x8_t vpaddlq_s8(int8x16_t __p0) {
18222 int16x8_t __ret;
18223 __ret = (int16x8_t) __builtin_neon_vpaddlq_v((int8x16_t)__p0, 33);
18224 return __ret;
18225 }
18226 #else
vpaddlq_s8(int8x16_t __p0)18227 __ai int16x8_t vpaddlq_s8(int8x16_t __p0) {
18228 int16x8_t __ret;
18229 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
18230 __ret = (int16x8_t) __builtin_neon_vpaddlq_v((int8x16_t)__rev0, 33);
18231 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
18232 return __ret;
18233 }
18234 #endif
18235
18236 #ifdef __LITTLE_ENDIAN__
vpaddlq_s32(int32x4_t __p0)18237 __ai int64x2_t vpaddlq_s32(int32x4_t __p0) {
18238 int64x2_t __ret;
18239 __ret = (int64x2_t) __builtin_neon_vpaddlq_v((int8x16_t)__p0, 35);
18240 return __ret;
18241 }
18242 #else
vpaddlq_s32(int32x4_t __p0)18243 __ai int64x2_t vpaddlq_s32(int32x4_t __p0) {
18244 int64x2_t __ret;
18245 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
18246 __ret = (int64x2_t) __builtin_neon_vpaddlq_v((int8x16_t)__rev0, 35);
18247 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
18248 return __ret;
18249 }
18250 #endif
18251
18252 #ifdef __LITTLE_ENDIAN__
vpaddlq_s16(int16x8_t __p0)18253 __ai int32x4_t vpaddlq_s16(int16x8_t __p0) {
18254 int32x4_t __ret;
18255 __ret = (int32x4_t) __builtin_neon_vpaddlq_v((int8x16_t)__p0, 34);
18256 return __ret;
18257 }
18258 #else
vpaddlq_s16(int16x8_t __p0)18259 __ai int32x4_t vpaddlq_s16(int16x8_t __p0) {
18260 int32x4_t __ret;
18261 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
18262 __ret = (int32x4_t) __builtin_neon_vpaddlq_v((int8x16_t)__rev0, 34);
18263 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
18264 return __ret;
18265 }
18266 #endif
18267
18268 #ifdef __LITTLE_ENDIAN__
vpaddl_u8(uint8x8_t __p0)18269 __ai uint16x4_t vpaddl_u8(uint8x8_t __p0) {
18270 uint16x4_t __ret;
18271 __ret = (uint16x4_t) __builtin_neon_vpaddl_v((int8x8_t)__p0, 17);
18272 return __ret;
18273 }
18274 #else
vpaddl_u8(uint8x8_t __p0)18275 __ai uint16x4_t vpaddl_u8(uint8x8_t __p0) {
18276 uint16x4_t __ret;
18277 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
18278 __ret = (uint16x4_t) __builtin_neon_vpaddl_v((int8x8_t)__rev0, 17);
18279 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
18280 return __ret;
18281 }
18282 #endif
18283
18284 #ifdef __LITTLE_ENDIAN__
vpaddl_u32(uint32x2_t __p0)18285 __ai uint64x1_t vpaddl_u32(uint32x2_t __p0) {
18286 uint64x1_t __ret;
18287 __ret = (uint64x1_t) __builtin_neon_vpaddl_v((int8x8_t)__p0, 19);
18288 return __ret;
18289 }
18290 #else
vpaddl_u32(uint32x2_t __p0)18291 __ai uint64x1_t vpaddl_u32(uint32x2_t __p0) {
18292 uint64x1_t __ret;
18293 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
18294 __ret = (uint64x1_t) __builtin_neon_vpaddl_v((int8x8_t)__rev0, 19);
18295 return __ret;
18296 }
18297 #endif
18298
18299 #ifdef __LITTLE_ENDIAN__
vpaddl_u16(uint16x4_t __p0)18300 __ai uint32x2_t vpaddl_u16(uint16x4_t __p0) {
18301 uint32x2_t __ret;
18302 __ret = (uint32x2_t) __builtin_neon_vpaddl_v((int8x8_t)__p0, 18);
18303 return __ret;
18304 }
18305 #else
vpaddl_u16(uint16x4_t __p0)18306 __ai uint32x2_t vpaddl_u16(uint16x4_t __p0) {
18307 uint32x2_t __ret;
18308 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
18309 __ret = (uint32x2_t) __builtin_neon_vpaddl_v((int8x8_t)__rev0, 18);
18310 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
18311 return __ret;
18312 }
18313 #endif
18314
18315 #ifdef __LITTLE_ENDIAN__
vpaddl_s8(int8x8_t __p0)18316 __ai int16x4_t vpaddl_s8(int8x8_t __p0) {
18317 int16x4_t __ret;
18318 __ret = (int16x4_t) __builtin_neon_vpaddl_v((int8x8_t)__p0, 1);
18319 return __ret;
18320 }
18321 #else
vpaddl_s8(int8x8_t __p0)18322 __ai int16x4_t vpaddl_s8(int8x8_t __p0) {
18323 int16x4_t __ret;
18324 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
18325 __ret = (int16x4_t) __builtin_neon_vpaddl_v((int8x8_t)__rev0, 1);
18326 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
18327 return __ret;
18328 }
18329 #endif
18330
18331 #ifdef __LITTLE_ENDIAN__
vpaddl_s32(int32x2_t __p0)18332 __ai int64x1_t vpaddl_s32(int32x2_t __p0) {
18333 int64x1_t __ret;
18334 __ret = (int64x1_t) __builtin_neon_vpaddl_v((int8x8_t)__p0, 3);
18335 return __ret;
18336 }
18337 #else
vpaddl_s32(int32x2_t __p0)18338 __ai int64x1_t vpaddl_s32(int32x2_t __p0) {
18339 int64x1_t __ret;
18340 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
18341 __ret = (int64x1_t) __builtin_neon_vpaddl_v((int8x8_t)__rev0, 3);
18342 return __ret;
18343 }
18344 #endif
18345
18346 #ifdef __LITTLE_ENDIAN__
vpaddl_s16(int16x4_t __p0)18347 __ai int32x2_t vpaddl_s16(int16x4_t __p0) {
18348 int32x2_t __ret;
18349 __ret = (int32x2_t) __builtin_neon_vpaddl_v((int8x8_t)__p0, 2);
18350 return __ret;
18351 }
18352 #else
vpaddl_s16(int16x4_t __p0)18353 __ai int32x2_t vpaddl_s16(int16x4_t __p0) {
18354 int32x2_t __ret;
18355 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
18356 __ret = (int32x2_t) __builtin_neon_vpaddl_v((int8x8_t)__rev0, 2);
18357 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
18358 return __ret;
18359 }
18360 #endif
18361
18362 #ifdef __LITTLE_ENDIAN__
vpmax_u8(uint8x8_t __p0,uint8x8_t __p1)18363 __ai uint8x8_t vpmax_u8(uint8x8_t __p0, uint8x8_t __p1) {
18364 uint8x8_t __ret;
18365 __ret = (uint8x8_t) __builtin_neon_vpmax_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
18366 return __ret;
18367 }
18368 #else
vpmax_u8(uint8x8_t __p0,uint8x8_t __p1)18369 __ai uint8x8_t vpmax_u8(uint8x8_t __p0, uint8x8_t __p1) {
18370 uint8x8_t __ret;
18371 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
18372 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
18373 __ret = (uint8x8_t) __builtin_neon_vpmax_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
18374 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
18375 return __ret;
18376 }
18377 #endif
18378
18379 #ifdef __LITTLE_ENDIAN__
vpmax_u32(uint32x2_t __p0,uint32x2_t __p1)18380 __ai uint32x2_t vpmax_u32(uint32x2_t __p0, uint32x2_t __p1) {
18381 uint32x2_t __ret;
18382 __ret = (uint32x2_t) __builtin_neon_vpmax_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
18383 return __ret;
18384 }
18385 #else
vpmax_u32(uint32x2_t __p0,uint32x2_t __p1)18386 __ai uint32x2_t vpmax_u32(uint32x2_t __p0, uint32x2_t __p1) {
18387 uint32x2_t __ret;
18388 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
18389 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
18390 __ret = (uint32x2_t) __builtin_neon_vpmax_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
18391 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
18392 return __ret;
18393 }
18394 #endif
18395
18396 #ifdef __LITTLE_ENDIAN__
vpmax_u16(uint16x4_t __p0,uint16x4_t __p1)18397 __ai uint16x4_t vpmax_u16(uint16x4_t __p0, uint16x4_t __p1) {
18398 uint16x4_t __ret;
18399 __ret = (uint16x4_t) __builtin_neon_vpmax_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
18400 return __ret;
18401 }
18402 #else
vpmax_u16(uint16x4_t __p0,uint16x4_t __p1)18403 __ai uint16x4_t vpmax_u16(uint16x4_t __p0, uint16x4_t __p1) {
18404 uint16x4_t __ret;
18405 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
18406 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
18407 __ret = (uint16x4_t) __builtin_neon_vpmax_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
18408 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
18409 return __ret;
18410 }
18411 #endif
18412
18413 #ifdef __LITTLE_ENDIAN__
vpmax_s8(int8x8_t __p0,int8x8_t __p1)18414 __ai int8x8_t vpmax_s8(int8x8_t __p0, int8x8_t __p1) {
18415 int8x8_t __ret;
18416 __ret = (int8x8_t) __builtin_neon_vpmax_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
18417 return __ret;
18418 }
18419 #else
vpmax_s8(int8x8_t __p0,int8x8_t __p1)18420 __ai int8x8_t vpmax_s8(int8x8_t __p0, int8x8_t __p1) {
18421 int8x8_t __ret;
18422 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
18423 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
18424 __ret = (int8x8_t) __builtin_neon_vpmax_v((int8x8_t)__rev0, (int8x8_t)__rev1, 0);
18425 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
18426 return __ret;
18427 }
18428 #endif
18429
18430 #ifdef __LITTLE_ENDIAN__
vpmax_f32(float32x2_t __p0,float32x2_t __p1)18431 __ai float32x2_t vpmax_f32(float32x2_t __p0, float32x2_t __p1) {
18432 float32x2_t __ret;
18433 __ret = (float32x2_t) __builtin_neon_vpmax_v((int8x8_t)__p0, (int8x8_t)__p1, 9);
18434 return __ret;
18435 }
18436 #else
vpmax_f32(float32x2_t __p0,float32x2_t __p1)18437 __ai float32x2_t vpmax_f32(float32x2_t __p0, float32x2_t __p1) {
18438 float32x2_t __ret;
18439 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
18440 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
18441 __ret = (float32x2_t) __builtin_neon_vpmax_v((int8x8_t)__rev0, (int8x8_t)__rev1, 9);
18442 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
18443 return __ret;
18444 }
18445 #endif
18446
18447 #ifdef __LITTLE_ENDIAN__
vpmax_s32(int32x2_t __p0,int32x2_t __p1)18448 __ai int32x2_t vpmax_s32(int32x2_t __p0, int32x2_t __p1) {
18449 int32x2_t __ret;
18450 __ret = (int32x2_t) __builtin_neon_vpmax_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
18451 return __ret;
18452 }
18453 #else
vpmax_s32(int32x2_t __p0,int32x2_t __p1)18454 __ai int32x2_t vpmax_s32(int32x2_t __p0, int32x2_t __p1) {
18455 int32x2_t __ret;
18456 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
18457 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
18458 __ret = (int32x2_t) __builtin_neon_vpmax_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
18459 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
18460 return __ret;
18461 }
18462 #endif
18463
18464 #ifdef __LITTLE_ENDIAN__
vpmax_s16(int16x4_t __p0,int16x4_t __p1)18465 __ai int16x4_t vpmax_s16(int16x4_t __p0, int16x4_t __p1) {
18466 int16x4_t __ret;
18467 __ret = (int16x4_t) __builtin_neon_vpmax_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
18468 return __ret;
18469 }
18470 #else
vpmax_s16(int16x4_t __p0,int16x4_t __p1)18471 __ai int16x4_t vpmax_s16(int16x4_t __p0, int16x4_t __p1) {
18472 int16x4_t __ret;
18473 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
18474 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
18475 __ret = (int16x4_t) __builtin_neon_vpmax_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
18476 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
18477 return __ret;
18478 }
18479 #endif
18480
18481 #ifdef __LITTLE_ENDIAN__
vpmin_u8(uint8x8_t __p0,uint8x8_t __p1)18482 __ai uint8x8_t vpmin_u8(uint8x8_t __p0, uint8x8_t __p1) {
18483 uint8x8_t __ret;
18484 __ret = (uint8x8_t) __builtin_neon_vpmin_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
18485 return __ret;
18486 }
18487 #else
vpmin_u8(uint8x8_t __p0,uint8x8_t __p1)18488 __ai uint8x8_t vpmin_u8(uint8x8_t __p0, uint8x8_t __p1) {
18489 uint8x8_t __ret;
18490 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
18491 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
18492 __ret = (uint8x8_t) __builtin_neon_vpmin_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
18493 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
18494 return __ret;
18495 }
18496 #endif
18497
18498 #ifdef __LITTLE_ENDIAN__
vpmin_u32(uint32x2_t __p0,uint32x2_t __p1)18499 __ai uint32x2_t vpmin_u32(uint32x2_t __p0, uint32x2_t __p1) {
18500 uint32x2_t __ret;
18501 __ret = (uint32x2_t) __builtin_neon_vpmin_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
18502 return __ret;
18503 }
18504 #else
vpmin_u32(uint32x2_t __p0,uint32x2_t __p1)18505 __ai uint32x2_t vpmin_u32(uint32x2_t __p0, uint32x2_t __p1) {
18506 uint32x2_t __ret;
18507 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
18508 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
18509 __ret = (uint32x2_t) __builtin_neon_vpmin_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
18510 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
18511 return __ret;
18512 }
18513 #endif
18514
18515 #ifdef __LITTLE_ENDIAN__
vpmin_u16(uint16x4_t __p0,uint16x4_t __p1)18516 __ai uint16x4_t vpmin_u16(uint16x4_t __p0, uint16x4_t __p1) {
18517 uint16x4_t __ret;
18518 __ret = (uint16x4_t) __builtin_neon_vpmin_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
18519 return __ret;
18520 }
18521 #else
vpmin_u16(uint16x4_t __p0,uint16x4_t __p1)18522 __ai uint16x4_t vpmin_u16(uint16x4_t __p0, uint16x4_t __p1) {
18523 uint16x4_t __ret;
18524 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
18525 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
18526 __ret = (uint16x4_t) __builtin_neon_vpmin_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
18527 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
18528 return __ret;
18529 }
18530 #endif
18531
18532 #ifdef __LITTLE_ENDIAN__
vpmin_s8(int8x8_t __p0,int8x8_t __p1)18533 __ai int8x8_t vpmin_s8(int8x8_t __p0, int8x8_t __p1) {
18534 int8x8_t __ret;
18535 __ret = (int8x8_t) __builtin_neon_vpmin_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
18536 return __ret;
18537 }
18538 #else
vpmin_s8(int8x8_t __p0,int8x8_t __p1)18539 __ai int8x8_t vpmin_s8(int8x8_t __p0, int8x8_t __p1) {
18540 int8x8_t __ret;
18541 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
18542 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
18543 __ret = (int8x8_t) __builtin_neon_vpmin_v((int8x8_t)__rev0, (int8x8_t)__rev1, 0);
18544 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
18545 return __ret;
18546 }
18547 #endif
18548
18549 #ifdef __LITTLE_ENDIAN__
vpmin_f32(float32x2_t __p0,float32x2_t __p1)18550 __ai float32x2_t vpmin_f32(float32x2_t __p0, float32x2_t __p1) {
18551 float32x2_t __ret;
18552 __ret = (float32x2_t) __builtin_neon_vpmin_v((int8x8_t)__p0, (int8x8_t)__p1, 9);
18553 return __ret;
18554 }
18555 #else
vpmin_f32(float32x2_t __p0,float32x2_t __p1)18556 __ai float32x2_t vpmin_f32(float32x2_t __p0, float32x2_t __p1) {
18557 float32x2_t __ret;
18558 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
18559 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
18560 __ret = (float32x2_t) __builtin_neon_vpmin_v((int8x8_t)__rev0, (int8x8_t)__rev1, 9);
18561 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
18562 return __ret;
18563 }
18564 #endif
18565
18566 #ifdef __LITTLE_ENDIAN__
vpmin_s32(int32x2_t __p0,int32x2_t __p1)18567 __ai int32x2_t vpmin_s32(int32x2_t __p0, int32x2_t __p1) {
18568 int32x2_t __ret;
18569 __ret = (int32x2_t) __builtin_neon_vpmin_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
18570 return __ret;
18571 }
18572 #else
vpmin_s32(int32x2_t __p0,int32x2_t __p1)18573 __ai int32x2_t vpmin_s32(int32x2_t __p0, int32x2_t __p1) {
18574 int32x2_t __ret;
18575 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
18576 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
18577 __ret = (int32x2_t) __builtin_neon_vpmin_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
18578 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
18579 return __ret;
18580 }
18581 #endif
18582
18583 #ifdef __LITTLE_ENDIAN__
vpmin_s16(int16x4_t __p0,int16x4_t __p1)18584 __ai int16x4_t vpmin_s16(int16x4_t __p0, int16x4_t __p1) {
18585 int16x4_t __ret;
18586 __ret = (int16x4_t) __builtin_neon_vpmin_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
18587 return __ret;
18588 }
18589 #else
vpmin_s16(int16x4_t __p0,int16x4_t __p1)18590 __ai int16x4_t vpmin_s16(int16x4_t __p0, int16x4_t __p1) {
18591 int16x4_t __ret;
18592 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
18593 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
18594 __ret = (int16x4_t) __builtin_neon_vpmin_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
18595 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
18596 return __ret;
18597 }
18598 #endif
18599
18600 #ifdef __LITTLE_ENDIAN__
vqabsq_s8(int8x16_t __p0)18601 __ai int8x16_t vqabsq_s8(int8x16_t __p0) {
18602 int8x16_t __ret;
18603 __ret = (int8x16_t) __builtin_neon_vqabsq_v((int8x16_t)__p0, 32);
18604 return __ret;
18605 }
18606 #else
vqabsq_s8(int8x16_t __p0)18607 __ai int8x16_t vqabsq_s8(int8x16_t __p0) {
18608 int8x16_t __ret;
18609 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
18610 __ret = (int8x16_t) __builtin_neon_vqabsq_v((int8x16_t)__rev0, 32);
18611 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
18612 return __ret;
18613 }
18614 #endif
18615
18616 #ifdef __LITTLE_ENDIAN__
vqabsq_s32(int32x4_t __p0)18617 __ai int32x4_t vqabsq_s32(int32x4_t __p0) {
18618 int32x4_t __ret;
18619 __ret = (int32x4_t) __builtin_neon_vqabsq_v((int8x16_t)__p0, 34);
18620 return __ret;
18621 }
18622 #else
vqabsq_s32(int32x4_t __p0)18623 __ai int32x4_t vqabsq_s32(int32x4_t __p0) {
18624 int32x4_t __ret;
18625 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
18626 __ret = (int32x4_t) __builtin_neon_vqabsq_v((int8x16_t)__rev0, 34);
18627 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
18628 return __ret;
18629 }
18630 #endif
18631
18632 #ifdef __LITTLE_ENDIAN__
vqabsq_s16(int16x8_t __p0)18633 __ai int16x8_t vqabsq_s16(int16x8_t __p0) {
18634 int16x8_t __ret;
18635 __ret = (int16x8_t) __builtin_neon_vqabsq_v((int8x16_t)__p0, 33);
18636 return __ret;
18637 }
18638 #else
vqabsq_s16(int16x8_t __p0)18639 __ai int16x8_t vqabsq_s16(int16x8_t __p0) {
18640 int16x8_t __ret;
18641 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
18642 __ret = (int16x8_t) __builtin_neon_vqabsq_v((int8x16_t)__rev0, 33);
18643 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
18644 return __ret;
18645 }
18646 #endif
18647
18648 #ifdef __LITTLE_ENDIAN__
vqabs_s8(int8x8_t __p0)18649 __ai int8x8_t vqabs_s8(int8x8_t __p0) {
18650 int8x8_t __ret;
18651 __ret = (int8x8_t) __builtin_neon_vqabs_v((int8x8_t)__p0, 0);
18652 return __ret;
18653 }
18654 #else
vqabs_s8(int8x8_t __p0)18655 __ai int8x8_t vqabs_s8(int8x8_t __p0) {
18656 int8x8_t __ret;
18657 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
18658 __ret = (int8x8_t) __builtin_neon_vqabs_v((int8x8_t)__rev0, 0);
18659 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
18660 return __ret;
18661 }
18662 #endif
18663
18664 #ifdef __LITTLE_ENDIAN__
vqabs_s32(int32x2_t __p0)18665 __ai int32x2_t vqabs_s32(int32x2_t __p0) {
18666 int32x2_t __ret;
18667 __ret = (int32x2_t) __builtin_neon_vqabs_v((int8x8_t)__p0, 2);
18668 return __ret;
18669 }
18670 #else
vqabs_s32(int32x2_t __p0)18671 __ai int32x2_t vqabs_s32(int32x2_t __p0) {
18672 int32x2_t __ret;
18673 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
18674 __ret = (int32x2_t) __builtin_neon_vqabs_v((int8x8_t)__rev0, 2);
18675 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
18676 return __ret;
18677 }
18678 #endif
18679
18680 #ifdef __LITTLE_ENDIAN__
vqabs_s16(int16x4_t __p0)18681 __ai int16x4_t vqabs_s16(int16x4_t __p0) {
18682 int16x4_t __ret;
18683 __ret = (int16x4_t) __builtin_neon_vqabs_v((int8x8_t)__p0, 1);
18684 return __ret;
18685 }
18686 #else
vqabs_s16(int16x4_t __p0)18687 __ai int16x4_t vqabs_s16(int16x4_t __p0) {
18688 int16x4_t __ret;
18689 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
18690 __ret = (int16x4_t) __builtin_neon_vqabs_v((int8x8_t)__rev0, 1);
18691 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
18692 return __ret;
18693 }
18694 #endif
18695
18696 #ifdef __LITTLE_ENDIAN__
vqaddq_u8(uint8x16_t __p0,uint8x16_t __p1)18697 __ai uint8x16_t vqaddq_u8(uint8x16_t __p0, uint8x16_t __p1) {
18698 uint8x16_t __ret;
18699 __ret = (uint8x16_t) __builtin_neon_vqaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
18700 return __ret;
18701 }
18702 #else
vqaddq_u8(uint8x16_t __p0,uint8x16_t __p1)18703 __ai uint8x16_t vqaddq_u8(uint8x16_t __p0, uint8x16_t __p1) {
18704 uint8x16_t __ret;
18705 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
18706 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
18707 __ret = (uint8x16_t) __builtin_neon_vqaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
18708 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
18709 return __ret;
18710 }
18711 #endif
18712
18713 #ifdef __LITTLE_ENDIAN__
vqaddq_u32(uint32x4_t __p0,uint32x4_t __p1)18714 __ai uint32x4_t vqaddq_u32(uint32x4_t __p0, uint32x4_t __p1) {
18715 uint32x4_t __ret;
18716 __ret = (uint32x4_t) __builtin_neon_vqaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
18717 return __ret;
18718 }
18719 #else
vqaddq_u32(uint32x4_t __p0,uint32x4_t __p1)18720 __ai uint32x4_t vqaddq_u32(uint32x4_t __p0, uint32x4_t __p1) {
18721 uint32x4_t __ret;
18722 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
18723 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
18724 __ret = (uint32x4_t) __builtin_neon_vqaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
18725 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
18726 return __ret;
18727 }
18728 #endif
18729
18730 #ifdef __LITTLE_ENDIAN__
vqaddq_u64(uint64x2_t __p0,uint64x2_t __p1)18731 __ai uint64x2_t vqaddq_u64(uint64x2_t __p0, uint64x2_t __p1) {
18732 uint64x2_t __ret;
18733 __ret = (uint64x2_t) __builtin_neon_vqaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 51);
18734 return __ret;
18735 }
18736 #else
vqaddq_u64(uint64x2_t __p0,uint64x2_t __p1)18737 __ai uint64x2_t vqaddq_u64(uint64x2_t __p0, uint64x2_t __p1) {
18738 uint64x2_t __ret;
18739 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
18740 uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
18741 __ret = (uint64x2_t) __builtin_neon_vqaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 51);
18742 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
18743 return __ret;
18744 }
18745 #endif
18746
18747 #ifdef __LITTLE_ENDIAN__
vqaddq_u16(uint16x8_t __p0,uint16x8_t __p1)18748 __ai uint16x8_t vqaddq_u16(uint16x8_t __p0, uint16x8_t __p1) {
18749 uint16x8_t __ret;
18750 __ret = (uint16x8_t) __builtin_neon_vqaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
18751 return __ret;
18752 }
18753 #else
vqaddq_u16(uint16x8_t __p0,uint16x8_t __p1)18754 __ai uint16x8_t vqaddq_u16(uint16x8_t __p0, uint16x8_t __p1) {
18755 uint16x8_t __ret;
18756 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
18757 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
18758 __ret = (uint16x8_t) __builtin_neon_vqaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
18759 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
18760 return __ret;
18761 }
18762 #endif
18763
18764 #ifdef __LITTLE_ENDIAN__
vqaddq_s8(int8x16_t __p0,int8x16_t __p1)18765 __ai int8x16_t vqaddq_s8(int8x16_t __p0, int8x16_t __p1) {
18766 int8x16_t __ret;
18767 __ret = (int8x16_t) __builtin_neon_vqaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 32);
18768 return __ret;
18769 }
18770 #else
vqaddq_s8(int8x16_t __p0,int8x16_t __p1)18771 __ai int8x16_t vqaddq_s8(int8x16_t __p0, int8x16_t __p1) {
18772 int8x16_t __ret;
18773 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
18774 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
18775 __ret = (int8x16_t) __builtin_neon_vqaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 32);
18776 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
18777 return __ret;
18778 }
18779 #endif
18780
18781 #ifdef __LITTLE_ENDIAN__
vqaddq_s32(int32x4_t __p0,int32x4_t __p1)18782 __ai int32x4_t vqaddq_s32(int32x4_t __p0, int32x4_t __p1) {
18783 int32x4_t __ret;
18784 __ret = (int32x4_t) __builtin_neon_vqaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
18785 return __ret;
18786 }
18787 #else
vqaddq_s32(int32x4_t __p0,int32x4_t __p1)18788 __ai int32x4_t vqaddq_s32(int32x4_t __p0, int32x4_t __p1) {
18789 int32x4_t __ret;
18790 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
18791 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
18792 __ret = (int32x4_t) __builtin_neon_vqaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
18793 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
18794 return __ret;
18795 }
18796 #endif
18797
18798 #ifdef __LITTLE_ENDIAN__
vqaddq_s64(int64x2_t __p0,int64x2_t __p1)18799 __ai int64x2_t vqaddq_s64(int64x2_t __p0, int64x2_t __p1) {
18800 int64x2_t __ret;
18801 __ret = (int64x2_t) __builtin_neon_vqaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 35);
18802 return __ret;
18803 }
18804 #else
vqaddq_s64(int64x2_t __p0,int64x2_t __p1)18805 __ai int64x2_t vqaddq_s64(int64x2_t __p0, int64x2_t __p1) {
18806 int64x2_t __ret;
18807 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
18808 int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
18809 __ret = (int64x2_t) __builtin_neon_vqaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 35);
18810 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
18811 return __ret;
18812 }
18813 #endif
18814
18815 #ifdef __LITTLE_ENDIAN__
vqaddq_s16(int16x8_t __p0,int16x8_t __p1)18816 __ai int16x8_t vqaddq_s16(int16x8_t __p0, int16x8_t __p1) {
18817 int16x8_t __ret;
18818 __ret = (int16x8_t) __builtin_neon_vqaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
18819 return __ret;
18820 }
18821 #else
vqaddq_s16(int16x8_t __p0,int16x8_t __p1)18822 __ai int16x8_t vqaddq_s16(int16x8_t __p0, int16x8_t __p1) {
18823 int16x8_t __ret;
18824 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
18825 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
18826 __ret = (int16x8_t) __builtin_neon_vqaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
18827 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
18828 return __ret;
18829 }
18830 #endif
18831
18832 #ifdef __LITTLE_ENDIAN__
vqadd_u8(uint8x8_t __p0,uint8x8_t __p1)18833 __ai uint8x8_t vqadd_u8(uint8x8_t __p0, uint8x8_t __p1) {
18834 uint8x8_t __ret;
18835 __ret = (uint8x8_t) __builtin_neon_vqadd_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
18836 return __ret;
18837 }
18838 #else
vqadd_u8(uint8x8_t __p0,uint8x8_t __p1)18839 __ai uint8x8_t vqadd_u8(uint8x8_t __p0, uint8x8_t __p1) {
18840 uint8x8_t __ret;
18841 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
18842 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
18843 __ret = (uint8x8_t) __builtin_neon_vqadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
18844 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
18845 return __ret;
18846 }
18847 #endif
18848
18849 #ifdef __LITTLE_ENDIAN__
vqadd_u32(uint32x2_t __p0,uint32x2_t __p1)18850 __ai uint32x2_t vqadd_u32(uint32x2_t __p0, uint32x2_t __p1) {
18851 uint32x2_t __ret;
18852 __ret = (uint32x2_t) __builtin_neon_vqadd_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
18853 return __ret;
18854 }
18855 #else
vqadd_u32(uint32x2_t __p0,uint32x2_t __p1)18856 __ai uint32x2_t vqadd_u32(uint32x2_t __p0, uint32x2_t __p1) {
18857 uint32x2_t __ret;
18858 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
18859 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
18860 __ret = (uint32x2_t) __builtin_neon_vqadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
18861 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
18862 return __ret;
18863 }
18864 #endif
18865
vqadd_u64(uint64x1_t __p0,uint64x1_t __p1)18866 __ai uint64x1_t vqadd_u64(uint64x1_t __p0, uint64x1_t __p1) {
18867 uint64x1_t __ret;
18868 __ret = (uint64x1_t) __builtin_neon_vqadd_v((int8x8_t)__p0, (int8x8_t)__p1, 19);
18869 return __ret;
18870 }
18871 #ifdef __LITTLE_ENDIAN__
vqadd_u16(uint16x4_t __p0,uint16x4_t __p1)18872 __ai uint16x4_t vqadd_u16(uint16x4_t __p0, uint16x4_t __p1) {
18873 uint16x4_t __ret;
18874 __ret = (uint16x4_t) __builtin_neon_vqadd_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
18875 return __ret;
18876 }
18877 #else
vqadd_u16(uint16x4_t __p0,uint16x4_t __p1)18878 __ai uint16x4_t vqadd_u16(uint16x4_t __p0, uint16x4_t __p1) {
18879 uint16x4_t __ret;
18880 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
18881 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
18882 __ret = (uint16x4_t) __builtin_neon_vqadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
18883 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
18884 return __ret;
18885 }
18886 #endif
18887
18888 #ifdef __LITTLE_ENDIAN__
vqadd_s8(int8x8_t __p0,int8x8_t __p1)18889 __ai int8x8_t vqadd_s8(int8x8_t __p0, int8x8_t __p1) {
18890 int8x8_t __ret;
18891 __ret = (int8x8_t) __builtin_neon_vqadd_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
18892 return __ret;
18893 }
18894 #else
vqadd_s8(int8x8_t __p0,int8x8_t __p1)18895 __ai int8x8_t vqadd_s8(int8x8_t __p0, int8x8_t __p1) {
18896 int8x8_t __ret;
18897 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
18898 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
18899 __ret = (int8x8_t) __builtin_neon_vqadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 0);
18900 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
18901 return __ret;
18902 }
18903 #endif
18904
18905 #ifdef __LITTLE_ENDIAN__
vqadd_s32(int32x2_t __p0,int32x2_t __p1)18906 __ai int32x2_t vqadd_s32(int32x2_t __p0, int32x2_t __p1) {
18907 int32x2_t __ret;
18908 __ret = (int32x2_t) __builtin_neon_vqadd_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
18909 return __ret;
18910 }
18911 #else
vqadd_s32(int32x2_t __p0,int32x2_t __p1)18912 __ai int32x2_t vqadd_s32(int32x2_t __p0, int32x2_t __p1) {
18913 int32x2_t __ret;
18914 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
18915 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
18916 __ret = (int32x2_t) __builtin_neon_vqadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
18917 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
18918 return __ret;
18919 }
18920 #endif
18921
vqadd_s64(int64x1_t __p0,int64x1_t __p1)18922 __ai int64x1_t vqadd_s64(int64x1_t __p0, int64x1_t __p1) {
18923 int64x1_t __ret;
18924 __ret = (int64x1_t) __builtin_neon_vqadd_v((int8x8_t)__p0, (int8x8_t)__p1, 3);
18925 return __ret;
18926 }
18927 #ifdef __LITTLE_ENDIAN__
vqadd_s16(int16x4_t __p0,int16x4_t __p1)18928 __ai int16x4_t vqadd_s16(int16x4_t __p0, int16x4_t __p1) {
18929 int16x4_t __ret;
18930 __ret = (int16x4_t) __builtin_neon_vqadd_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
18931 return __ret;
18932 }
18933 #else
vqadd_s16(int16x4_t __p0,int16x4_t __p1)18934 __ai int16x4_t vqadd_s16(int16x4_t __p0, int16x4_t __p1) {
18935 int16x4_t __ret;
18936 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
18937 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
18938 __ret = (int16x4_t) __builtin_neon_vqadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
18939 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
18940 return __ret;
18941 }
18942 #endif
18943
18944 #ifdef __LITTLE_ENDIAN__
vqdmlal_s32(int64x2_t __p0,int32x2_t __p1,int32x2_t __p2)18945 __ai int64x2_t vqdmlal_s32(int64x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
18946 int64x2_t __ret;
18947 __ret = (int64x2_t) __builtin_neon_vqdmlal_v((int8x16_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 35);
18948 return __ret;
18949 }
18950 #else
vqdmlal_s32(int64x2_t __p0,int32x2_t __p1,int32x2_t __p2)18951 __ai int64x2_t vqdmlal_s32(int64x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
18952 int64x2_t __ret;
18953 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
18954 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
18955 int32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
18956 __ret = (int64x2_t) __builtin_neon_vqdmlal_v((int8x16_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 35);
18957 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
18958 return __ret;
18959 }
__noswap_vqdmlal_s32(int64x2_t __p0,int32x2_t __p1,int32x2_t __p2)18960 __ai int64x2_t __noswap_vqdmlal_s32(int64x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
18961 int64x2_t __ret;
18962 __ret = (int64x2_t) __builtin_neon_vqdmlal_v((int8x16_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 35);
18963 return __ret;
18964 }
18965 #endif
18966
18967 #ifdef __LITTLE_ENDIAN__
vqdmlal_s16(int32x4_t __p0,int16x4_t __p1,int16x4_t __p2)18968 __ai int32x4_t vqdmlal_s16(int32x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
18969 int32x4_t __ret;
18970 __ret = (int32x4_t) __builtin_neon_vqdmlal_v((int8x16_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 34);
18971 return __ret;
18972 }
18973 #else
vqdmlal_s16(int32x4_t __p0,int16x4_t __p1,int16x4_t __p2)18974 __ai int32x4_t vqdmlal_s16(int32x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
18975 int32x4_t __ret;
18976 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
18977 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
18978 int16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
18979 __ret = (int32x4_t) __builtin_neon_vqdmlal_v((int8x16_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 34);
18980 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
18981 return __ret;
18982 }
__noswap_vqdmlal_s16(int32x4_t __p0,int16x4_t __p1,int16x4_t __p2)18983 __ai int32x4_t __noswap_vqdmlal_s16(int32x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
18984 int32x4_t __ret;
18985 __ret = (int32x4_t) __builtin_neon_vqdmlal_v((int8x16_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 34);
18986 return __ret;
18987 }
18988 #endif
18989
18990 #ifdef __LITTLE_ENDIAN__
18991 #define vqdmlal_lane_s32(__p0_114, __p1_114, __p2_114, __p3_114) __extension__ ({ \
18992 int64x2_t __ret_114; \
18993 int64x2_t __s0_114 = __p0_114; \
18994 int32x2_t __s1_114 = __p1_114; \
18995 int32x2_t __s2_114 = __p2_114; \
18996 __ret_114 = vqdmlal_s32(__s0_114, __s1_114, splat_lane_s32(__s2_114, __p3_114)); \
18997 __ret_114; \
18998 })
18999 #else
19000 #define vqdmlal_lane_s32(__p0_115, __p1_115, __p2_115, __p3_115) __extension__ ({ \
19001 int64x2_t __ret_115; \
19002 int64x2_t __s0_115 = __p0_115; \
19003 int32x2_t __s1_115 = __p1_115; \
19004 int32x2_t __s2_115 = __p2_115; \
19005 int64x2_t __rev0_115; __rev0_115 = __builtin_shufflevector(__s0_115, __s0_115, 1, 0); \
19006 int32x2_t __rev1_115; __rev1_115 = __builtin_shufflevector(__s1_115, __s1_115, 1, 0); \
19007 int32x2_t __rev2_115; __rev2_115 = __builtin_shufflevector(__s2_115, __s2_115, 1, 0); \
19008 __ret_115 = __noswap_vqdmlal_s32(__rev0_115, __rev1_115, __noswap_splat_lane_s32(__rev2_115, __p3_115)); \
19009 __ret_115 = __builtin_shufflevector(__ret_115, __ret_115, 1, 0); \
19010 __ret_115; \
19011 })
19012 #endif
19013
19014 #ifdef __LITTLE_ENDIAN__
19015 #define vqdmlal_lane_s16(__p0_116, __p1_116, __p2_116, __p3_116) __extension__ ({ \
19016 int32x4_t __ret_116; \
19017 int32x4_t __s0_116 = __p0_116; \
19018 int16x4_t __s1_116 = __p1_116; \
19019 int16x4_t __s2_116 = __p2_116; \
19020 __ret_116 = vqdmlal_s16(__s0_116, __s1_116, splat_lane_s16(__s2_116, __p3_116)); \
19021 __ret_116; \
19022 })
19023 #else
19024 #define vqdmlal_lane_s16(__p0_117, __p1_117, __p2_117, __p3_117) __extension__ ({ \
19025 int32x4_t __ret_117; \
19026 int32x4_t __s0_117 = __p0_117; \
19027 int16x4_t __s1_117 = __p1_117; \
19028 int16x4_t __s2_117 = __p2_117; \
19029 int32x4_t __rev0_117; __rev0_117 = __builtin_shufflevector(__s0_117, __s0_117, 3, 2, 1, 0); \
19030 int16x4_t __rev1_117; __rev1_117 = __builtin_shufflevector(__s1_117, __s1_117, 3, 2, 1, 0); \
19031 int16x4_t __rev2_117; __rev2_117 = __builtin_shufflevector(__s2_117, __s2_117, 3, 2, 1, 0); \
19032 __ret_117 = __noswap_vqdmlal_s16(__rev0_117, __rev1_117, __noswap_splat_lane_s16(__rev2_117, __p3_117)); \
19033 __ret_117 = __builtin_shufflevector(__ret_117, __ret_117, 3, 2, 1, 0); \
19034 __ret_117; \
19035 })
19036 #endif
19037
19038 #ifdef __LITTLE_ENDIAN__
vqdmlal_n_s32(int64x2_t __p0,int32x2_t __p1,int32_t __p2)19039 __ai int64x2_t vqdmlal_n_s32(int64x2_t __p0, int32x2_t __p1, int32_t __p2) {
19040 int64x2_t __ret;
19041 __ret = vqdmlal_s32(__p0, __p1, (int32x2_t) {__p2, __p2});
19042 return __ret;
19043 }
19044 #else
vqdmlal_n_s32(int64x2_t __p0,int32x2_t __p1,int32_t __p2)19045 __ai int64x2_t vqdmlal_n_s32(int64x2_t __p0, int32x2_t __p1, int32_t __p2) {
19046 int64x2_t __ret;
19047 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
19048 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
19049 __ret = __noswap_vqdmlal_s32(__rev0, __rev1, (int32x2_t) {__p2, __p2});
19050 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
19051 return __ret;
19052 }
__noswap_vqdmlal_n_s32(int64x2_t __p0,int32x2_t __p1,int32_t __p2)19053 __ai int64x2_t __noswap_vqdmlal_n_s32(int64x2_t __p0, int32x2_t __p1, int32_t __p2) {
19054 int64x2_t __ret;
19055 __ret = __noswap_vqdmlal_s32(__p0, __p1, (int32x2_t) {__p2, __p2});
19056 return __ret;
19057 }
19058 #endif
19059
19060 #ifdef __LITTLE_ENDIAN__
vqdmlal_n_s16(int32x4_t __p0,int16x4_t __p1,int16_t __p2)19061 __ai int32x4_t vqdmlal_n_s16(int32x4_t __p0, int16x4_t __p1, int16_t __p2) {
19062 int32x4_t __ret;
19063 __ret = vqdmlal_s16(__p0, __p1, (int16x4_t) {__p2, __p2, __p2, __p2});
19064 return __ret;
19065 }
19066 #else
vqdmlal_n_s16(int32x4_t __p0,int16x4_t __p1,int16_t __p2)19067 __ai int32x4_t vqdmlal_n_s16(int32x4_t __p0, int16x4_t __p1, int16_t __p2) {
19068 int32x4_t __ret;
19069 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
19070 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
19071 __ret = __noswap_vqdmlal_s16(__rev0, __rev1, (int16x4_t) {__p2, __p2, __p2, __p2});
19072 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
19073 return __ret;
19074 }
__noswap_vqdmlal_n_s16(int32x4_t __p0,int16x4_t __p1,int16_t __p2)19075 __ai int32x4_t __noswap_vqdmlal_n_s16(int32x4_t __p0, int16x4_t __p1, int16_t __p2) {
19076 int32x4_t __ret;
19077 __ret = __noswap_vqdmlal_s16(__p0, __p1, (int16x4_t) {__p2, __p2, __p2, __p2});
19078 return __ret;
19079 }
19080 #endif
19081
19082 #ifdef __LITTLE_ENDIAN__
vqdmlsl_s32(int64x2_t __p0,int32x2_t __p1,int32x2_t __p2)19083 __ai int64x2_t vqdmlsl_s32(int64x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
19084 int64x2_t __ret;
19085 __ret = (int64x2_t) __builtin_neon_vqdmlsl_v((int8x16_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 35);
19086 return __ret;
19087 }
19088 #else
vqdmlsl_s32(int64x2_t __p0,int32x2_t __p1,int32x2_t __p2)19089 __ai int64x2_t vqdmlsl_s32(int64x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
19090 int64x2_t __ret;
19091 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
19092 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
19093 int32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
19094 __ret = (int64x2_t) __builtin_neon_vqdmlsl_v((int8x16_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 35);
19095 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
19096 return __ret;
19097 }
__noswap_vqdmlsl_s32(int64x2_t __p0,int32x2_t __p1,int32x2_t __p2)19098 __ai int64x2_t __noswap_vqdmlsl_s32(int64x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
19099 int64x2_t __ret;
19100 __ret = (int64x2_t) __builtin_neon_vqdmlsl_v((int8x16_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 35);
19101 return __ret;
19102 }
19103 #endif
19104
19105 #ifdef __LITTLE_ENDIAN__
vqdmlsl_s16(int32x4_t __p0,int16x4_t __p1,int16x4_t __p2)19106 __ai int32x4_t vqdmlsl_s16(int32x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
19107 int32x4_t __ret;
19108 __ret = (int32x4_t) __builtin_neon_vqdmlsl_v((int8x16_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 34);
19109 return __ret;
19110 }
19111 #else
vqdmlsl_s16(int32x4_t __p0,int16x4_t __p1,int16x4_t __p2)19112 __ai int32x4_t vqdmlsl_s16(int32x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
19113 int32x4_t __ret;
19114 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
19115 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
19116 int16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
19117 __ret = (int32x4_t) __builtin_neon_vqdmlsl_v((int8x16_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 34);
19118 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
19119 return __ret;
19120 }
__noswap_vqdmlsl_s16(int32x4_t __p0,int16x4_t __p1,int16x4_t __p2)19121 __ai int32x4_t __noswap_vqdmlsl_s16(int32x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
19122 int32x4_t __ret;
19123 __ret = (int32x4_t) __builtin_neon_vqdmlsl_v((int8x16_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 34);
19124 return __ret;
19125 }
19126 #endif
19127
19128 #ifdef __LITTLE_ENDIAN__
19129 #define vqdmlsl_lane_s32(__p0_118, __p1_118, __p2_118, __p3_118) __extension__ ({ \
19130 int64x2_t __ret_118; \
19131 int64x2_t __s0_118 = __p0_118; \
19132 int32x2_t __s1_118 = __p1_118; \
19133 int32x2_t __s2_118 = __p2_118; \
19134 __ret_118 = vqdmlsl_s32(__s0_118, __s1_118, splat_lane_s32(__s2_118, __p3_118)); \
19135 __ret_118; \
19136 })
19137 #else
19138 #define vqdmlsl_lane_s32(__p0_119, __p1_119, __p2_119, __p3_119) __extension__ ({ \
19139 int64x2_t __ret_119; \
19140 int64x2_t __s0_119 = __p0_119; \
19141 int32x2_t __s1_119 = __p1_119; \
19142 int32x2_t __s2_119 = __p2_119; \
19143 int64x2_t __rev0_119; __rev0_119 = __builtin_shufflevector(__s0_119, __s0_119, 1, 0); \
19144 int32x2_t __rev1_119; __rev1_119 = __builtin_shufflevector(__s1_119, __s1_119, 1, 0); \
19145 int32x2_t __rev2_119; __rev2_119 = __builtin_shufflevector(__s2_119, __s2_119, 1, 0); \
19146 __ret_119 = __noswap_vqdmlsl_s32(__rev0_119, __rev1_119, __noswap_splat_lane_s32(__rev2_119, __p3_119)); \
19147 __ret_119 = __builtin_shufflevector(__ret_119, __ret_119, 1, 0); \
19148 __ret_119; \
19149 })
19150 #endif
19151
19152 #ifdef __LITTLE_ENDIAN__
19153 #define vqdmlsl_lane_s16(__p0_120, __p1_120, __p2_120, __p3_120) __extension__ ({ \
19154 int32x4_t __ret_120; \
19155 int32x4_t __s0_120 = __p0_120; \
19156 int16x4_t __s1_120 = __p1_120; \
19157 int16x4_t __s2_120 = __p2_120; \
19158 __ret_120 = vqdmlsl_s16(__s0_120, __s1_120, splat_lane_s16(__s2_120, __p3_120)); \
19159 __ret_120; \
19160 })
19161 #else
19162 #define vqdmlsl_lane_s16(__p0_121, __p1_121, __p2_121, __p3_121) __extension__ ({ \
19163 int32x4_t __ret_121; \
19164 int32x4_t __s0_121 = __p0_121; \
19165 int16x4_t __s1_121 = __p1_121; \
19166 int16x4_t __s2_121 = __p2_121; \
19167 int32x4_t __rev0_121; __rev0_121 = __builtin_shufflevector(__s0_121, __s0_121, 3, 2, 1, 0); \
19168 int16x4_t __rev1_121; __rev1_121 = __builtin_shufflevector(__s1_121, __s1_121, 3, 2, 1, 0); \
19169 int16x4_t __rev2_121; __rev2_121 = __builtin_shufflevector(__s2_121, __s2_121, 3, 2, 1, 0); \
19170 __ret_121 = __noswap_vqdmlsl_s16(__rev0_121, __rev1_121, __noswap_splat_lane_s16(__rev2_121, __p3_121)); \
19171 __ret_121 = __builtin_shufflevector(__ret_121, __ret_121, 3, 2, 1, 0); \
19172 __ret_121; \
19173 })
19174 #endif
19175
19176 #ifdef __LITTLE_ENDIAN__
vqdmlsl_n_s32(int64x2_t __p0,int32x2_t __p1,int32_t __p2)19177 __ai int64x2_t vqdmlsl_n_s32(int64x2_t __p0, int32x2_t __p1, int32_t __p2) {
19178 int64x2_t __ret;
19179 __ret = vqdmlsl_s32(__p0, __p1, (int32x2_t) {__p2, __p2});
19180 return __ret;
19181 }
19182 #else
vqdmlsl_n_s32(int64x2_t __p0,int32x2_t __p1,int32_t __p2)19183 __ai int64x2_t vqdmlsl_n_s32(int64x2_t __p0, int32x2_t __p1, int32_t __p2) {
19184 int64x2_t __ret;
19185 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
19186 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
19187 __ret = __noswap_vqdmlsl_s32(__rev0, __rev1, (int32x2_t) {__p2, __p2});
19188 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
19189 return __ret;
19190 }
__noswap_vqdmlsl_n_s32(int64x2_t __p0,int32x2_t __p1,int32_t __p2)19191 __ai int64x2_t __noswap_vqdmlsl_n_s32(int64x2_t __p0, int32x2_t __p1, int32_t __p2) {
19192 int64x2_t __ret;
19193 __ret = __noswap_vqdmlsl_s32(__p0, __p1, (int32x2_t) {__p2, __p2});
19194 return __ret;
19195 }
19196 #endif
19197
19198 #ifdef __LITTLE_ENDIAN__
vqdmlsl_n_s16(int32x4_t __p0,int16x4_t __p1,int16_t __p2)19199 __ai int32x4_t vqdmlsl_n_s16(int32x4_t __p0, int16x4_t __p1, int16_t __p2) {
19200 int32x4_t __ret;
19201 __ret = vqdmlsl_s16(__p0, __p1, (int16x4_t) {__p2, __p2, __p2, __p2});
19202 return __ret;
19203 }
19204 #else
vqdmlsl_n_s16(int32x4_t __p0,int16x4_t __p1,int16_t __p2)19205 __ai int32x4_t vqdmlsl_n_s16(int32x4_t __p0, int16x4_t __p1, int16_t __p2) {
19206 int32x4_t __ret;
19207 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
19208 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
19209 __ret = __noswap_vqdmlsl_s16(__rev0, __rev1, (int16x4_t) {__p2, __p2, __p2, __p2});
19210 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
19211 return __ret;
19212 }
__noswap_vqdmlsl_n_s16(int32x4_t __p0,int16x4_t __p1,int16_t __p2)19213 __ai int32x4_t __noswap_vqdmlsl_n_s16(int32x4_t __p0, int16x4_t __p1, int16_t __p2) {
19214 int32x4_t __ret;
19215 __ret = __noswap_vqdmlsl_s16(__p0, __p1, (int16x4_t) {__p2, __p2, __p2, __p2});
19216 return __ret;
19217 }
19218 #endif
19219
19220 #ifdef __LITTLE_ENDIAN__
vqdmulhq_s32(int32x4_t __p0,int32x4_t __p1)19221 __ai int32x4_t vqdmulhq_s32(int32x4_t __p0, int32x4_t __p1) {
19222 int32x4_t __ret;
19223 __ret = (int32x4_t) __builtin_neon_vqdmulhq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
19224 return __ret;
19225 }
19226 #else
vqdmulhq_s32(int32x4_t __p0,int32x4_t __p1)19227 __ai int32x4_t vqdmulhq_s32(int32x4_t __p0, int32x4_t __p1) {
19228 int32x4_t __ret;
19229 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
19230 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
19231 __ret = (int32x4_t) __builtin_neon_vqdmulhq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
19232 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
19233 return __ret;
19234 }
__noswap_vqdmulhq_s32(int32x4_t __p0,int32x4_t __p1)19235 __ai int32x4_t __noswap_vqdmulhq_s32(int32x4_t __p0, int32x4_t __p1) {
19236 int32x4_t __ret;
19237 __ret = (int32x4_t) __builtin_neon_vqdmulhq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
19238 return __ret;
19239 }
19240 #endif
19241
19242 #ifdef __LITTLE_ENDIAN__
vqdmulhq_s16(int16x8_t __p0,int16x8_t __p1)19243 __ai int16x8_t vqdmulhq_s16(int16x8_t __p0, int16x8_t __p1) {
19244 int16x8_t __ret;
19245 __ret = (int16x8_t) __builtin_neon_vqdmulhq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
19246 return __ret;
19247 }
19248 #else
vqdmulhq_s16(int16x8_t __p0,int16x8_t __p1)19249 __ai int16x8_t vqdmulhq_s16(int16x8_t __p0, int16x8_t __p1) {
19250 int16x8_t __ret;
19251 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
19252 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
19253 __ret = (int16x8_t) __builtin_neon_vqdmulhq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
19254 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
19255 return __ret;
19256 }
__noswap_vqdmulhq_s16(int16x8_t __p0,int16x8_t __p1)19257 __ai int16x8_t __noswap_vqdmulhq_s16(int16x8_t __p0, int16x8_t __p1) {
19258 int16x8_t __ret;
19259 __ret = (int16x8_t) __builtin_neon_vqdmulhq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
19260 return __ret;
19261 }
19262 #endif
19263
19264 #ifdef __LITTLE_ENDIAN__
vqdmulh_s32(int32x2_t __p0,int32x2_t __p1)19265 __ai int32x2_t vqdmulh_s32(int32x2_t __p0, int32x2_t __p1) {
19266 int32x2_t __ret;
19267 __ret = (int32x2_t) __builtin_neon_vqdmulh_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
19268 return __ret;
19269 }
19270 #else
vqdmulh_s32(int32x2_t __p0,int32x2_t __p1)19271 __ai int32x2_t vqdmulh_s32(int32x2_t __p0, int32x2_t __p1) {
19272 int32x2_t __ret;
19273 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
19274 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
19275 __ret = (int32x2_t) __builtin_neon_vqdmulh_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
19276 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
19277 return __ret;
19278 }
__noswap_vqdmulh_s32(int32x2_t __p0,int32x2_t __p1)19279 __ai int32x2_t __noswap_vqdmulh_s32(int32x2_t __p0, int32x2_t __p1) {
19280 int32x2_t __ret;
19281 __ret = (int32x2_t) __builtin_neon_vqdmulh_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
19282 return __ret;
19283 }
19284 #endif
19285
19286 #ifdef __LITTLE_ENDIAN__
vqdmulh_s16(int16x4_t __p0,int16x4_t __p1)19287 __ai int16x4_t vqdmulh_s16(int16x4_t __p0, int16x4_t __p1) {
19288 int16x4_t __ret;
19289 __ret = (int16x4_t) __builtin_neon_vqdmulh_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
19290 return __ret;
19291 }
19292 #else
vqdmulh_s16(int16x4_t __p0,int16x4_t __p1)19293 __ai int16x4_t vqdmulh_s16(int16x4_t __p0, int16x4_t __p1) {
19294 int16x4_t __ret;
19295 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
19296 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
19297 __ret = (int16x4_t) __builtin_neon_vqdmulh_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
19298 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
19299 return __ret;
19300 }
__noswap_vqdmulh_s16(int16x4_t __p0,int16x4_t __p1)19301 __ai int16x4_t __noswap_vqdmulh_s16(int16x4_t __p0, int16x4_t __p1) {
19302 int16x4_t __ret;
19303 __ret = (int16x4_t) __builtin_neon_vqdmulh_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
19304 return __ret;
19305 }
19306 #endif
19307
19308 #ifdef __LITTLE_ENDIAN__
vqdmulhq_n_s32(int32x4_t __p0,int32_t __p1)19309 __ai int32x4_t vqdmulhq_n_s32(int32x4_t __p0, int32_t __p1) {
19310 int32x4_t __ret;
19311 __ret = vqdmulhq_s32(__p0, (int32x4_t) {__p1, __p1, __p1, __p1});
19312 return __ret;
19313 }
19314 #else
vqdmulhq_n_s32(int32x4_t __p0,int32_t __p1)19315 __ai int32x4_t vqdmulhq_n_s32(int32x4_t __p0, int32_t __p1) {
19316 int32x4_t __ret;
19317 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
19318 __ret = __noswap_vqdmulhq_s32(__rev0, (int32x4_t) {__p1, __p1, __p1, __p1});
19319 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
19320 return __ret;
19321 }
19322 #endif
19323
19324 #ifdef __LITTLE_ENDIAN__
vqdmulhq_n_s16(int16x8_t __p0,int16_t __p1)19325 __ai int16x8_t vqdmulhq_n_s16(int16x8_t __p0, int16_t __p1) {
19326 int16x8_t __ret;
19327 __ret = vqdmulhq_s16(__p0, (int16x8_t) {__p1, __p1, __p1, __p1, __p1, __p1, __p1, __p1});
19328 return __ret;
19329 }
19330 #else
vqdmulhq_n_s16(int16x8_t __p0,int16_t __p1)19331 __ai int16x8_t vqdmulhq_n_s16(int16x8_t __p0, int16_t __p1) {
19332 int16x8_t __ret;
19333 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
19334 __ret = __noswap_vqdmulhq_s16(__rev0, (int16x8_t) {__p1, __p1, __p1, __p1, __p1, __p1, __p1, __p1});
19335 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
19336 return __ret;
19337 }
19338 #endif
19339
19340 #ifdef __LITTLE_ENDIAN__
vqdmulh_n_s32(int32x2_t __p0,int32_t __p1)19341 __ai int32x2_t vqdmulh_n_s32(int32x2_t __p0, int32_t __p1) {
19342 int32x2_t __ret;
19343 __ret = vqdmulh_s32(__p0, (int32x2_t) {__p1, __p1});
19344 return __ret;
19345 }
19346 #else
vqdmulh_n_s32(int32x2_t __p0,int32_t __p1)19347 __ai int32x2_t vqdmulh_n_s32(int32x2_t __p0, int32_t __p1) {
19348 int32x2_t __ret;
19349 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
19350 __ret = __noswap_vqdmulh_s32(__rev0, (int32x2_t) {__p1, __p1});
19351 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
19352 return __ret;
19353 }
19354 #endif
19355
19356 #ifdef __LITTLE_ENDIAN__
vqdmulh_n_s16(int16x4_t __p0,int16_t __p1)19357 __ai int16x4_t vqdmulh_n_s16(int16x4_t __p0, int16_t __p1) {
19358 int16x4_t __ret;
19359 __ret = vqdmulh_s16(__p0, (int16x4_t) {__p1, __p1, __p1, __p1});
19360 return __ret;
19361 }
19362 #else
vqdmulh_n_s16(int16x4_t __p0,int16_t __p1)19363 __ai int16x4_t vqdmulh_n_s16(int16x4_t __p0, int16_t __p1) {
19364 int16x4_t __ret;
19365 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
19366 __ret = __noswap_vqdmulh_s16(__rev0, (int16x4_t) {__p1, __p1, __p1, __p1});
19367 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
19368 return __ret;
19369 }
19370 #endif
19371
19372 #ifdef __LITTLE_ENDIAN__
vqdmull_s32(int32x2_t __p0,int32x2_t __p1)19373 __ai int64x2_t vqdmull_s32(int32x2_t __p0, int32x2_t __p1) {
19374 int64x2_t __ret;
19375 __ret = (int64x2_t) __builtin_neon_vqdmull_v((int8x8_t)__p0, (int8x8_t)__p1, 35);
19376 return __ret;
19377 }
19378 #else
vqdmull_s32(int32x2_t __p0,int32x2_t __p1)19379 __ai int64x2_t vqdmull_s32(int32x2_t __p0, int32x2_t __p1) {
19380 int64x2_t __ret;
19381 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
19382 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
19383 __ret = (int64x2_t) __builtin_neon_vqdmull_v((int8x8_t)__rev0, (int8x8_t)__rev1, 35);
19384 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
19385 return __ret;
19386 }
__noswap_vqdmull_s32(int32x2_t __p0,int32x2_t __p1)19387 __ai int64x2_t __noswap_vqdmull_s32(int32x2_t __p0, int32x2_t __p1) {
19388 int64x2_t __ret;
19389 __ret = (int64x2_t) __builtin_neon_vqdmull_v((int8x8_t)__p0, (int8x8_t)__p1, 35);
19390 return __ret;
19391 }
19392 #endif
19393
19394 #ifdef __LITTLE_ENDIAN__
vqdmull_s16(int16x4_t __p0,int16x4_t __p1)19395 __ai int32x4_t vqdmull_s16(int16x4_t __p0, int16x4_t __p1) {
19396 int32x4_t __ret;
19397 __ret = (int32x4_t) __builtin_neon_vqdmull_v((int8x8_t)__p0, (int8x8_t)__p1, 34);
19398 return __ret;
19399 }
19400 #else
vqdmull_s16(int16x4_t __p0,int16x4_t __p1)19401 __ai int32x4_t vqdmull_s16(int16x4_t __p0, int16x4_t __p1) {
19402 int32x4_t __ret;
19403 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
19404 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
19405 __ret = (int32x4_t) __builtin_neon_vqdmull_v((int8x8_t)__rev0, (int8x8_t)__rev1, 34);
19406 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
19407 return __ret;
19408 }
__noswap_vqdmull_s16(int16x4_t __p0,int16x4_t __p1)19409 __ai int32x4_t __noswap_vqdmull_s16(int16x4_t __p0, int16x4_t __p1) {
19410 int32x4_t __ret;
19411 __ret = (int32x4_t) __builtin_neon_vqdmull_v((int8x8_t)__p0, (int8x8_t)__p1, 34);
19412 return __ret;
19413 }
19414 #endif
19415
19416 #ifdef __LITTLE_ENDIAN__
19417 #define vqdmull_lane_s32(__p0_122, __p1_122, __p2_122) __extension__ ({ \
19418 int64x2_t __ret_122; \
19419 int32x2_t __s0_122 = __p0_122; \
19420 int32x2_t __s1_122 = __p1_122; \
19421 __ret_122 = vqdmull_s32(__s0_122, splat_lane_s32(__s1_122, __p2_122)); \
19422 __ret_122; \
19423 })
19424 #else
19425 #define vqdmull_lane_s32(__p0_123, __p1_123, __p2_123) __extension__ ({ \
19426 int64x2_t __ret_123; \
19427 int32x2_t __s0_123 = __p0_123; \
19428 int32x2_t __s1_123 = __p1_123; \
19429 int32x2_t __rev0_123; __rev0_123 = __builtin_shufflevector(__s0_123, __s0_123, 1, 0); \
19430 int32x2_t __rev1_123; __rev1_123 = __builtin_shufflevector(__s1_123, __s1_123, 1, 0); \
19431 __ret_123 = __noswap_vqdmull_s32(__rev0_123, __noswap_splat_lane_s32(__rev1_123, __p2_123)); \
19432 __ret_123 = __builtin_shufflevector(__ret_123, __ret_123, 1, 0); \
19433 __ret_123; \
19434 })
19435 #endif
19436
19437 #ifdef __LITTLE_ENDIAN__
19438 #define vqdmull_lane_s16(__p0_124, __p1_124, __p2_124) __extension__ ({ \
19439 int32x4_t __ret_124; \
19440 int16x4_t __s0_124 = __p0_124; \
19441 int16x4_t __s1_124 = __p1_124; \
19442 __ret_124 = vqdmull_s16(__s0_124, splat_lane_s16(__s1_124, __p2_124)); \
19443 __ret_124; \
19444 })
19445 #else
19446 #define vqdmull_lane_s16(__p0_125, __p1_125, __p2_125) __extension__ ({ \
19447 int32x4_t __ret_125; \
19448 int16x4_t __s0_125 = __p0_125; \
19449 int16x4_t __s1_125 = __p1_125; \
19450 int16x4_t __rev0_125; __rev0_125 = __builtin_shufflevector(__s0_125, __s0_125, 3, 2, 1, 0); \
19451 int16x4_t __rev1_125; __rev1_125 = __builtin_shufflevector(__s1_125, __s1_125, 3, 2, 1, 0); \
19452 __ret_125 = __noswap_vqdmull_s16(__rev0_125, __noswap_splat_lane_s16(__rev1_125, __p2_125)); \
19453 __ret_125 = __builtin_shufflevector(__ret_125, __ret_125, 3, 2, 1, 0); \
19454 __ret_125; \
19455 })
19456 #endif
19457
19458 #ifdef __LITTLE_ENDIAN__
vqdmull_n_s32(int32x2_t __p0,int32_t __p1)19459 __ai int64x2_t vqdmull_n_s32(int32x2_t __p0, int32_t __p1) {
19460 int64x2_t __ret;
19461 __ret = vqdmull_s32(__p0, (int32x2_t) {__p1, __p1});
19462 return __ret;
19463 }
19464 #else
vqdmull_n_s32(int32x2_t __p0,int32_t __p1)19465 __ai int64x2_t vqdmull_n_s32(int32x2_t __p0, int32_t __p1) {
19466 int64x2_t __ret;
19467 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
19468 __ret = __noswap_vqdmull_s32(__rev0, (int32x2_t) {__p1, __p1});
19469 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
19470 return __ret;
19471 }
__noswap_vqdmull_n_s32(int32x2_t __p0,int32_t __p1)19472 __ai int64x2_t __noswap_vqdmull_n_s32(int32x2_t __p0, int32_t __p1) {
19473 int64x2_t __ret;
19474 __ret = __noswap_vqdmull_s32(__p0, (int32x2_t) {__p1, __p1});
19475 return __ret;
19476 }
19477 #endif
19478
19479 #ifdef __LITTLE_ENDIAN__
vqdmull_n_s16(int16x4_t __p0,int16_t __p1)19480 __ai int32x4_t vqdmull_n_s16(int16x4_t __p0, int16_t __p1) {
19481 int32x4_t __ret;
19482 __ret = vqdmull_s16(__p0, (int16x4_t) {__p1, __p1, __p1, __p1});
19483 return __ret;
19484 }
19485 #else
vqdmull_n_s16(int16x4_t __p0,int16_t __p1)19486 __ai int32x4_t vqdmull_n_s16(int16x4_t __p0, int16_t __p1) {
19487 int32x4_t __ret;
19488 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
19489 __ret = __noswap_vqdmull_s16(__rev0, (int16x4_t) {__p1, __p1, __p1, __p1});
19490 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
19491 return __ret;
19492 }
__noswap_vqdmull_n_s16(int16x4_t __p0,int16_t __p1)19493 __ai int32x4_t __noswap_vqdmull_n_s16(int16x4_t __p0, int16_t __p1) {
19494 int32x4_t __ret;
19495 __ret = __noswap_vqdmull_s16(__p0, (int16x4_t) {__p1, __p1, __p1, __p1});
19496 return __ret;
19497 }
19498 #endif
19499
19500 #ifdef __LITTLE_ENDIAN__
vqmovn_u32(uint32x4_t __p0)19501 __ai uint16x4_t vqmovn_u32(uint32x4_t __p0) {
19502 uint16x4_t __ret;
19503 __ret = (uint16x4_t) __builtin_neon_vqmovn_v((int8x16_t)__p0, 17);
19504 return __ret;
19505 }
19506 #else
vqmovn_u32(uint32x4_t __p0)19507 __ai uint16x4_t vqmovn_u32(uint32x4_t __p0) {
19508 uint16x4_t __ret;
19509 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
19510 __ret = (uint16x4_t) __builtin_neon_vqmovn_v((int8x16_t)__rev0, 17);
19511 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
19512 return __ret;
19513 }
__noswap_vqmovn_u32(uint32x4_t __p0)19514 __ai uint16x4_t __noswap_vqmovn_u32(uint32x4_t __p0) {
19515 uint16x4_t __ret;
19516 __ret = (uint16x4_t) __builtin_neon_vqmovn_v((int8x16_t)__p0, 17);
19517 return __ret;
19518 }
19519 #endif
19520
19521 #ifdef __LITTLE_ENDIAN__
vqmovn_u64(uint64x2_t __p0)19522 __ai uint32x2_t vqmovn_u64(uint64x2_t __p0) {
19523 uint32x2_t __ret;
19524 __ret = (uint32x2_t) __builtin_neon_vqmovn_v((int8x16_t)__p0, 18);
19525 return __ret;
19526 }
19527 #else
vqmovn_u64(uint64x2_t __p0)19528 __ai uint32x2_t vqmovn_u64(uint64x2_t __p0) {
19529 uint32x2_t __ret;
19530 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
19531 __ret = (uint32x2_t) __builtin_neon_vqmovn_v((int8x16_t)__rev0, 18);
19532 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
19533 return __ret;
19534 }
__noswap_vqmovn_u64(uint64x2_t __p0)19535 __ai uint32x2_t __noswap_vqmovn_u64(uint64x2_t __p0) {
19536 uint32x2_t __ret;
19537 __ret = (uint32x2_t) __builtin_neon_vqmovn_v((int8x16_t)__p0, 18);
19538 return __ret;
19539 }
19540 #endif
19541
19542 #ifdef __LITTLE_ENDIAN__
vqmovn_u16(uint16x8_t __p0)19543 __ai uint8x8_t vqmovn_u16(uint16x8_t __p0) {
19544 uint8x8_t __ret;
19545 __ret = (uint8x8_t) __builtin_neon_vqmovn_v((int8x16_t)__p0, 16);
19546 return __ret;
19547 }
19548 #else
vqmovn_u16(uint16x8_t __p0)19549 __ai uint8x8_t vqmovn_u16(uint16x8_t __p0) {
19550 uint8x8_t __ret;
19551 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
19552 __ret = (uint8x8_t) __builtin_neon_vqmovn_v((int8x16_t)__rev0, 16);
19553 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
19554 return __ret;
19555 }
__noswap_vqmovn_u16(uint16x8_t __p0)19556 __ai uint8x8_t __noswap_vqmovn_u16(uint16x8_t __p0) {
19557 uint8x8_t __ret;
19558 __ret = (uint8x8_t) __builtin_neon_vqmovn_v((int8x16_t)__p0, 16);
19559 return __ret;
19560 }
19561 #endif
19562
19563 #ifdef __LITTLE_ENDIAN__
vqmovn_s32(int32x4_t __p0)19564 __ai int16x4_t vqmovn_s32(int32x4_t __p0) {
19565 int16x4_t __ret;
19566 __ret = (int16x4_t) __builtin_neon_vqmovn_v((int8x16_t)__p0, 1);
19567 return __ret;
19568 }
19569 #else
vqmovn_s32(int32x4_t __p0)19570 __ai int16x4_t vqmovn_s32(int32x4_t __p0) {
19571 int16x4_t __ret;
19572 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
19573 __ret = (int16x4_t) __builtin_neon_vqmovn_v((int8x16_t)__rev0, 1);
19574 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
19575 return __ret;
19576 }
__noswap_vqmovn_s32(int32x4_t __p0)19577 __ai int16x4_t __noswap_vqmovn_s32(int32x4_t __p0) {
19578 int16x4_t __ret;
19579 __ret = (int16x4_t) __builtin_neon_vqmovn_v((int8x16_t)__p0, 1);
19580 return __ret;
19581 }
19582 #endif
19583
19584 #ifdef __LITTLE_ENDIAN__
vqmovn_s64(int64x2_t __p0)19585 __ai int32x2_t vqmovn_s64(int64x2_t __p0) {
19586 int32x2_t __ret;
19587 __ret = (int32x2_t) __builtin_neon_vqmovn_v((int8x16_t)__p0, 2);
19588 return __ret;
19589 }
19590 #else
vqmovn_s64(int64x2_t __p0)19591 __ai int32x2_t vqmovn_s64(int64x2_t __p0) {
19592 int32x2_t __ret;
19593 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
19594 __ret = (int32x2_t) __builtin_neon_vqmovn_v((int8x16_t)__rev0, 2);
19595 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
19596 return __ret;
19597 }
__noswap_vqmovn_s64(int64x2_t __p0)19598 __ai int32x2_t __noswap_vqmovn_s64(int64x2_t __p0) {
19599 int32x2_t __ret;
19600 __ret = (int32x2_t) __builtin_neon_vqmovn_v((int8x16_t)__p0, 2);
19601 return __ret;
19602 }
19603 #endif
19604
19605 #ifdef __LITTLE_ENDIAN__
vqmovn_s16(int16x8_t __p0)19606 __ai int8x8_t vqmovn_s16(int16x8_t __p0) {
19607 int8x8_t __ret;
19608 __ret = (int8x8_t) __builtin_neon_vqmovn_v((int8x16_t)__p0, 0);
19609 return __ret;
19610 }
19611 #else
vqmovn_s16(int16x8_t __p0)19612 __ai int8x8_t vqmovn_s16(int16x8_t __p0) {
19613 int8x8_t __ret;
19614 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
19615 __ret = (int8x8_t) __builtin_neon_vqmovn_v((int8x16_t)__rev0, 0);
19616 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
19617 return __ret;
19618 }
__noswap_vqmovn_s16(int16x8_t __p0)19619 __ai int8x8_t __noswap_vqmovn_s16(int16x8_t __p0) {
19620 int8x8_t __ret;
19621 __ret = (int8x8_t) __builtin_neon_vqmovn_v((int8x16_t)__p0, 0);
19622 return __ret;
19623 }
19624 #endif
19625
19626 #ifdef __LITTLE_ENDIAN__
vqmovun_s32(int32x4_t __p0)19627 __ai uint16x4_t vqmovun_s32(int32x4_t __p0) {
19628 uint16x4_t __ret;
19629 __ret = (uint16x4_t) __builtin_neon_vqmovun_v((int8x16_t)__p0, 17);
19630 return __ret;
19631 }
19632 #else
vqmovun_s32(int32x4_t __p0)19633 __ai uint16x4_t vqmovun_s32(int32x4_t __p0) {
19634 uint16x4_t __ret;
19635 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
19636 __ret = (uint16x4_t) __builtin_neon_vqmovun_v((int8x16_t)__rev0, 17);
19637 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
19638 return __ret;
19639 }
__noswap_vqmovun_s32(int32x4_t __p0)19640 __ai uint16x4_t __noswap_vqmovun_s32(int32x4_t __p0) {
19641 uint16x4_t __ret;
19642 __ret = (uint16x4_t) __builtin_neon_vqmovun_v((int8x16_t)__p0, 17);
19643 return __ret;
19644 }
19645 #endif
19646
19647 #ifdef __LITTLE_ENDIAN__
vqmovun_s64(int64x2_t __p0)19648 __ai uint32x2_t vqmovun_s64(int64x2_t __p0) {
19649 uint32x2_t __ret;
19650 __ret = (uint32x2_t) __builtin_neon_vqmovun_v((int8x16_t)__p0, 18);
19651 return __ret;
19652 }
19653 #else
vqmovun_s64(int64x2_t __p0)19654 __ai uint32x2_t vqmovun_s64(int64x2_t __p0) {
19655 uint32x2_t __ret;
19656 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
19657 __ret = (uint32x2_t) __builtin_neon_vqmovun_v((int8x16_t)__rev0, 18);
19658 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
19659 return __ret;
19660 }
__noswap_vqmovun_s64(int64x2_t __p0)19661 __ai uint32x2_t __noswap_vqmovun_s64(int64x2_t __p0) {
19662 uint32x2_t __ret;
19663 __ret = (uint32x2_t) __builtin_neon_vqmovun_v((int8x16_t)__p0, 18);
19664 return __ret;
19665 }
19666 #endif
19667
19668 #ifdef __LITTLE_ENDIAN__
vqmovun_s16(int16x8_t __p0)19669 __ai uint8x8_t vqmovun_s16(int16x8_t __p0) {
19670 uint8x8_t __ret;
19671 __ret = (uint8x8_t) __builtin_neon_vqmovun_v((int8x16_t)__p0, 16);
19672 return __ret;
19673 }
19674 #else
vqmovun_s16(int16x8_t __p0)19675 __ai uint8x8_t vqmovun_s16(int16x8_t __p0) {
19676 uint8x8_t __ret;
19677 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
19678 __ret = (uint8x8_t) __builtin_neon_vqmovun_v((int8x16_t)__rev0, 16);
19679 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
19680 return __ret;
19681 }
__noswap_vqmovun_s16(int16x8_t __p0)19682 __ai uint8x8_t __noswap_vqmovun_s16(int16x8_t __p0) {
19683 uint8x8_t __ret;
19684 __ret = (uint8x8_t) __builtin_neon_vqmovun_v((int8x16_t)__p0, 16);
19685 return __ret;
19686 }
19687 #endif
19688
19689 #ifdef __LITTLE_ENDIAN__
vqnegq_s8(int8x16_t __p0)19690 __ai int8x16_t vqnegq_s8(int8x16_t __p0) {
19691 int8x16_t __ret;
19692 __ret = (int8x16_t) __builtin_neon_vqnegq_v((int8x16_t)__p0, 32);
19693 return __ret;
19694 }
19695 #else
vqnegq_s8(int8x16_t __p0)19696 __ai int8x16_t vqnegq_s8(int8x16_t __p0) {
19697 int8x16_t __ret;
19698 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
19699 __ret = (int8x16_t) __builtin_neon_vqnegq_v((int8x16_t)__rev0, 32);
19700 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
19701 return __ret;
19702 }
19703 #endif
19704
19705 #ifdef __LITTLE_ENDIAN__
vqnegq_s32(int32x4_t __p0)19706 __ai int32x4_t vqnegq_s32(int32x4_t __p0) {
19707 int32x4_t __ret;
19708 __ret = (int32x4_t) __builtin_neon_vqnegq_v((int8x16_t)__p0, 34);
19709 return __ret;
19710 }
19711 #else
vqnegq_s32(int32x4_t __p0)19712 __ai int32x4_t vqnegq_s32(int32x4_t __p0) {
19713 int32x4_t __ret;
19714 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
19715 __ret = (int32x4_t) __builtin_neon_vqnegq_v((int8x16_t)__rev0, 34);
19716 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
19717 return __ret;
19718 }
19719 #endif
19720
19721 #ifdef __LITTLE_ENDIAN__
vqnegq_s16(int16x8_t __p0)19722 __ai int16x8_t vqnegq_s16(int16x8_t __p0) {
19723 int16x8_t __ret;
19724 __ret = (int16x8_t) __builtin_neon_vqnegq_v((int8x16_t)__p0, 33);
19725 return __ret;
19726 }
19727 #else
vqnegq_s16(int16x8_t __p0)19728 __ai int16x8_t vqnegq_s16(int16x8_t __p0) {
19729 int16x8_t __ret;
19730 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
19731 __ret = (int16x8_t) __builtin_neon_vqnegq_v((int8x16_t)__rev0, 33);
19732 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
19733 return __ret;
19734 }
19735 #endif
19736
19737 #ifdef __LITTLE_ENDIAN__
vqneg_s8(int8x8_t __p0)19738 __ai int8x8_t vqneg_s8(int8x8_t __p0) {
19739 int8x8_t __ret;
19740 __ret = (int8x8_t) __builtin_neon_vqneg_v((int8x8_t)__p0, 0);
19741 return __ret;
19742 }
19743 #else
vqneg_s8(int8x8_t __p0)19744 __ai int8x8_t vqneg_s8(int8x8_t __p0) {
19745 int8x8_t __ret;
19746 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
19747 __ret = (int8x8_t) __builtin_neon_vqneg_v((int8x8_t)__rev0, 0);
19748 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
19749 return __ret;
19750 }
19751 #endif
19752
19753 #ifdef __LITTLE_ENDIAN__
vqneg_s32(int32x2_t __p0)19754 __ai int32x2_t vqneg_s32(int32x2_t __p0) {
19755 int32x2_t __ret;
19756 __ret = (int32x2_t) __builtin_neon_vqneg_v((int8x8_t)__p0, 2);
19757 return __ret;
19758 }
19759 #else
vqneg_s32(int32x2_t __p0)19760 __ai int32x2_t vqneg_s32(int32x2_t __p0) {
19761 int32x2_t __ret;
19762 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
19763 __ret = (int32x2_t) __builtin_neon_vqneg_v((int8x8_t)__rev0, 2);
19764 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
19765 return __ret;
19766 }
19767 #endif
19768
19769 #ifdef __LITTLE_ENDIAN__
vqneg_s16(int16x4_t __p0)19770 __ai int16x4_t vqneg_s16(int16x4_t __p0) {
19771 int16x4_t __ret;
19772 __ret = (int16x4_t) __builtin_neon_vqneg_v((int8x8_t)__p0, 1);
19773 return __ret;
19774 }
19775 #else
vqneg_s16(int16x4_t __p0)19776 __ai int16x4_t vqneg_s16(int16x4_t __p0) {
19777 int16x4_t __ret;
19778 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
19779 __ret = (int16x4_t) __builtin_neon_vqneg_v((int8x8_t)__rev0, 1);
19780 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
19781 return __ret;
19782 }
19783 #endif
19784
19785 #ifdef __LITTLE_ENDIAN__
vqrdmulhq_s32(int32x4_t __p0,int32x4_t __p1)19786 __ai int32x4_t vqrdmulhq_s32(int32x4_t __p0, int32x4_t __p1) {
19787 int32x4_t __ret;
19788 __ret = (int32x4_t) __builtin_neon_vqrdmulhq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
19789 return __ret;
19790 }
19791 #else
vqrdmulhq_s32(int32x4_t __p0,int32x4_t __p1)19792 __ai int32x4_t vqrdmulhq_s32(int32x4_t __p0, int32x4_t __p1) {
19793 int32x4_t __ret;
19794 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
19795 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
19796 __ret = (int32x4_t) __builtin_neon_vqrdmulhq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
19797 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
19798 return __ret;
19799 }
__noswap_vqrdmulhq_s32(int32x4_t __p0,int32x4_t __p1)19800 __ai int32x4_t __noswap_vqrdmulhq_s32(int32x4_t __p0, int32x4_t __p1) {
19801 int32x4_t __ret;
19802 __ret = (int32x4_t) __builtin_neon_vqrdmulhq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
19803 return __ret;
19804 }
19805 #endif
19806
19807 #ifdef __LITTLE_ENDIAN__
vqrdmulhq_s16(int16x8_t __p0,int16x8_t __p1)19808 __ai int16x8_t vqrdmulhq_s16(int16x8_t __p0, int16x8_t __p1) {
19809 int16x8_t __ret;
19810 __ret = (int16x8_t) __builtin_neon_vqrdmulhq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
19811 return __ret;
19812 }
19813 #else
vqrdmulhq_s16(int16x8_t __p0,int16x8_t __p1)19814 __ai int16x8_t vqrdmulhq_s16(int16x8_t __p0, int16x8_t __p1) {
19815 int16x8_t __ret;
19816 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
19817 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
19818 __ret = (int16x8_t) __builtin_neon_vqrdmulhq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
19819 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
19820 return __ret;
19821 }
__noswap_vqrdmulhq_s16(int16x8_t __p0,int16x8_t __p1)19822 __ai int16x8_t __noswap_vqrdmulhq_s16(int16x8_t __p0, int16x8_t __p1) {
19823 int16x8_t __ret;
19824 __ret = (int16x8_t) __builtin_neon_vqrdmulhq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
19825 return __ret;
19826 }
19827 #endif
19828
19829 #ifdef __LITTLE_ENDIAN__
vqrdmulh_s32(int32x2_t __p0,int32x2_t __p1)19830 __ai int32x2_t vqrdmulh_s32(int32x2_t __p0, int32x2_t __p1) {
19831 int32x2_t __ret;
19832 __ret = (int32x2_t) __builtin_neon_vqrdmulh_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
19833 return __ret;
19834 }
19835 #else
vqrdmulh_s32(int32x2_t __p0,int32x2_t __p1)19836 __ai int32x2_t vqrdmulh_s32(int32x2_t __p0, int32x2_t __p1) {
19837 int32x2_t __ret;
19838 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
19839 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
19840 __ret = (int32x2_t) __builtin_neon_vqrdmulh_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
19841 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
19842 return __ret;
19843 }
__noswap_vqrdmulh_s32(int32x2_t __p0,int32x2_t __p1)19844 __ai int32x2_t __noswap_vqrdmulh_s32(int32x2_t __p0, int32x2_t __p1) {
19845 int32x2_t __ret;
19846 __ret = (int32x2_t) __builtin_neon_vqrdmulh_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
19847 return __ret;
19848 }
19849 #endif
19850
19851 #ifdef __LITTLE_ENDIAN__
vqrdmulh_s16(int16x4_t __p0,int16x4_t __p1)19852 __ai int16x4_t vqrdmulh_s16(int16x4_t __p0, int16x4_t __p1) {
19853 int16x4_t __ret;
19854 __ret = (int16x4_t) __builtin_neon_vqrdmulh_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
19855 return __ret;
19856 }
19857 #else
vqrdmulh_s16(int16x4_t __p0,int16x4_t __p1)19858 __ai int16x4_t vqrdmulh_s16(int16x4_t __p0, int16x4_t __p1) {
19859 int16x4_t __ret;
19860 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
19861 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
19862 __ret = (int16x4_t) __builtin_neon_vqrdmulh_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
19863 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
19864 return __ret;
19865 }
__noswap_vqrdmulh_s16(int16x4_t __p0,int16x4_t __p1)19866 __ai int16x4_t __noswap_vqrdmulh_s16(int16x4_t __p0, int16x4_t __p1) {
19867 int16x4_t __ret;
19868 __ret = (int16x4_t) __builtin_neon_vqrdmulh_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
19869 return __ret;
19870 }
19871 #endif
19872
19873 #ifdef __LITTLE_ENDIAN__
vqrdmulhq_n_s32(int32x4_t __p0,int32_t __p1)19874 __ai int32x4_t vqrdmulhq_n_s32(int32x4_t __p0, int32_t __p1) {
19875 int32x4_t __ret;
19876 __ret = vqrdmulhq_s32(__p0, (int32x4_t) {__p1, __p1, __p1, __p1});
19877 return __ret;
19878 }
19879 #else
vqrdmulhq_n_s32(int32x4_t __p0,int32_t __p1)19880 __ai int32x4_t vqrdmulhq_n_s32(int32x4_t __p0, int32_t __p1) {
19881 int32x4_t __ret;
19882 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
19883 __ret = __noswap_vqrdmulhq_s32(__rev0, (int32x4_t) {__p1, __p1, __p1, __p1});
19884 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
19885 return __ret;
19886 }
19887 #endif
19888
19889 #ifdef __LITTLE_ENDIAN__
vqrdmulhq_n_s16(int16x8_t __p0,int16_t __p1)19890 __ai int16x8_t vqrdmulhq_n_s16(int16x8_t __p0, int16_t __p1) {
19891 int16x8_t __ret;
19892 __ret = vqrdmulhq_s16(__p0, (int16x8_t) {__p1, __p1, __p1, __p1, __p1, __p1, __p1, __p1});
19893 return __ret;
19894 }
19895 #else
vqrdmulhq_n_s16(int16x8_t __p0,int16_t __p1)19896 __ai int16x8_t vqrdmulhq_n_s16(int16x8_t __p0, int16_t __p1) {
19897 int16x8_t __ret;
19898 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
19899 __ret = __noswap_vqrdmulhq_s16(__rev0, (int16x8_t) {__p1, __p1, __p1, __p1, __p1, __p1, __p1, __p1});
19900 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
19901 return __ret;
19902 }
19903 #endif
19904
19905 #ifdef __LITTLE_ENDIAN__
vqrdmulh_n_s32(int32x2_t __p0,int32_t __p1)19906 __ai int32x2_t vqrdmulh_n_s32(int32x2_t __p0, int32_t __p1) {
19907 int32x2_t __ret;
19908 __ret = vqrdmulh_s32(__p0, (int32x2_t) {__p1, __p1});
19909 return __ret;
19910 }
19911 #else
vqrdmulh_n_s32(int32x2_t __p0,int32_t __p1)19912 __ai int32x2_t vqrdmulh_n_s32(int32x2_t __p0, int32_t __p1) {
19913 int32x2_t __ret;
19914 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
19915 __ret = __noswap_vqrdmulh_s32(__rev0, (int32x2_t) {__p1, __p1});
19916 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
19917 return __ret;
19918 }
19919 #endif
19920
19921 #ifdef __LITTLE_ENDIAN__
vqrdmulh_n_s16(int16x4_t __p0,int16_t __p1)19922 __ai int16x4_t vqrdmulh_n_s16(int16x4_t __p0, int16_t __p1) {
19923 int16x4_t __ret;
19924 __ret = vqrdmulh_s16(__p0, (int16x4_t) {__p1, __p1, __p1, __p1});
19925 return __ret;
19926 }
19927 #else
vqrdmulh_n_s16(int16x4_t __p0,int16_t __p1)19928 __ai int16x4_t vqrdmulh_n_s16(int16x4_t __p0, int16_t __p1) {
19929 int16x4_t __ret;
19930 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
19931 __ret = __noswap_vqrdmulh_s16(__rev0, (int16x4_t) {__p1, __p1, __p1, __p1});
19932 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
19933 return __ret;
19934 }
19935 #endif
19936
19937 #ifdef __LITTLE_ENDIAN__
vqrshlq_u8(uint8x16_t __p0,int8x16_t __p1)19938 __ai uint8x16_t vqrshlq_u8(uint8x16_t __p0, int8x16_t __p1) {
19939 uint8x16_t __ret;
19940 __ret = (uint8x16_t) __builtin_neon_vqrshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
19941 return __ret;
19942 }
19943 #else
vqrshlq_u8(uint8x16_t __p0,int8x16_t __p1)19944 __ai uint8x16_t vqrshlq_u8(uint8x16_t __p0, int8x16_t __p1) {
19945 uint8x16_t __ret;
19946 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
19947 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
19948 __ret = (uint8x16_t) __builtin_neon_vqrshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
19949 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
19950 return __ret;
19951 }
19952 #endif
19953
19954 #ifdef __LITTLE_ENDIAN__
vqrshlq_u32(uint32x4_t __p0,int32x4_t __p1)19955 __ai uint32x4_t vqrshlq_u32(uint32x4_t __p0, int32x4_t __p1) {
19956 uint32x4_t __ret;
19957 __ret = (uint32x4_t) __builtin_neon_vqrshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
19958 return __ret;
19959 }
19960 #else
vqrshlq_u32(uint32x4_t __p0,int32x4_t __p1)19961 __ai uint32x4_t vqrshlq_u32(uint32x4_t __p0, int32x4_t __p1) {
19962 uint32x4_t __ret;
19963 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
19964 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
19965 __ret = (uint32x4_t) __builtin_neon_vqrshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
19966 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
19967 return __ret;
19968 }
19969 #endif
19970
19971 #ifdef __LITTLE_ENDIAN__
vqrshlq_u64(uint64x2_t __p0,int64x2_t __p1)19972 __ai uint64x2_t vqrshlq_u64(uint64x2_t __p0, int64x2_t __p1) {
19973 uint64x2_t __ret;
19974 __ret = (uint64x2_t) __builtin_neon_vqrshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 51);
19975 return __ret;
19976 }
19977 #else
vqrshlq_u64(uint64x2_t __p0,int64x2_t __p1)19978 __ai uint64x2_t vqrshlq_u64(uint64x2_t __p0, int64x2_t __p1) {
19979 uint64x2_t __ret;
19980 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
19981 int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
19982 __ret = (uint64x2_t) __builtin_neon_vqrshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 51);
19983 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
19984 return __ret;
19985 }
19986 #endif
19987
19988 #ifdef __LITTLE_ENDIAN__
vqrshlq_u16(uint16x8_t __p0,int16x8_t __p1)19989 __ai uint16x8_t vqrshlq_u16(uint16x8_t __p0, int16x8_t __p1) {
19990 uint16x8_t __ret;
19991 __ret = (uint16x8_t) __builtin_neon_vqrshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
19992 return __ret;
19993 }
19994 #else
vqrshlq_u16(uint16x8_t __p0,int16x8_t __p1)19995 __ai uint16x8_t vqrshlq_u16(uint16x8_t __p0, int16x8_t __p1) {
19996 uint16x8_t __ret;
19997 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
19998 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
19999 __ret = (uint16x8_t) __builtin_neon_vqrshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
20000 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
20001 return __ret;
20002 }
20003 #endif
20004
20005 #ifdef __LITTLE_ENDIAN__
vqrshlq_s8(int8x16_t __p0,int8x16_t __p1)20006 __ai int8x16_t vqrshlq_s8(int8x16_t __p0, int8x16_t __p1) {
20007 int8x16_t __ret;
20008 __ret = (int8x16_t) __builtin_neon_vqrshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 32);
20009 return __ret;
20010 }
20011 #else
vqrshlq_s8(int8x16_t __p0,int8x16_t __p1)20012 __ai int8x16_t vqrshlq_s8(int8x16_t __p0, int8x16_t __p1) {
20013 int8x16_t __ret;
20014 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
20015 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
20016 __ret = (int8x16_t) __builtin_neon_vqrshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 32);
20017 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
20018 return __ret;
20019 }
20020 #endif
20021
20022 #ifdef __LITTLE_ENDIAN__
vqrshlq_s32(int32x4_t __p0,int32x4_t __p1)20023 __ai int32x4_t vqrshlq_s32(int32x4_t __p0, int32x4_t __p1) {
20024 int32x4_t __ret;
20025 __ret = (int32x4_t) __builtin_neon_vqrshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
20026 return __ret;
20027 }
20028 #else
vqrshlq_s32(int32x4_t __p0,int32x4_t __p1)20029 __ai int32x4_t vqrshlq_s32(int32x4_t __p0, int32x4_t __p1) {
20030 int32x4_t __ret;
20031 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
20032 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
20033 __ret = (int32x4_t) __builtin_neon_vqrshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
20034 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
20035 return __ret;
20036 }
20037 #endif
20038
20039 #ifdef __LITTLE_ENDIAN__
vqrshlq_s64(int64x2_t __p0,int64x2_t __p1)20040 __ai int64x2_t vqrshlq_s64(int64x2_t __p0, int64x2_t __p1) {
20041 int64x2_t __ret;
20042 __ret = (int64x2_t) __builtin_neon_vqrshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 35);
20043 return __ret;
20044 }
20045 #else
vqrshlq_s64(int64x2_t __p0,int64x2_t __p1)20046 __ai int64x2_t vqrshlq_s64(int64x2_t __p0, int64x2_t __p1) {
20047 int64x2_t __ret;
20048 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
20049 int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
20050 __ret = (int64x2_t) __builtin_neon_vqrshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 35);
20051 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
20052 return __ret;
20053 }
20054 #endif
20055
20056 #ifdef __LITTLE_ENDIAN__
vqrshlq_s16(int16x8_t __p0,int16x8_t __p1)20057 __ai int16x8_t vqrshlq_s16(int16x8_t __p0, int16x8_t __p1) {
20058 int16x8_t __ret;
20059 __ret = (int16x8_t) __builtin_neon_vqrshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
20060 return __ret;
20061 }
20062 #else
vqrshlq_s16(int16x8_t __p0,int16x8_t __p1)20063 __ai int16x8_t vqrshlq_s16(int16x8_t __p0, int16x8_t __p1) {
20064 int16x8_t __ret;
20065 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
20066 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
20067 __ret = (int16x8_t) __builtin_neon_vqrshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
20068 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
20069 return __ret;
20070 }
20071 #endif
20072
20073 #ifdef __LITTLE_ENDIAN__
vqrshl_u8(uint8x8_t __p0,int8x8_t __p1)20074 __ai uint8x8_t vqrshl_u8(uint8x8_t __p0, int8x8_t __p1) {
20075 uint8x8_t __ret;
20076 __ret = (uint8x8_t) __builtin_neon_vqrshl_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
20077 return __ret;
20078 }
20079 #else
vqrshl_u8(uint8x8_t __p0,int8x8_t __p1)20080 __ai uint8x8_t vqrshl_u8(uint8x8_t __p0, int8x8_t __p1) {
20081 uint8x8_t __ret;
20082 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
20083 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
20084 __ret = (uint8x8_t) __builtin_neon_vqrshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
20085 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
20086 return __ret;
20087 }
20088 #endif
20089
20090 #ifdef __LITTLE_ENDIAN__
vqrshl_u32(uint32x2_t __p0,int32x2_t __p1)20091 __ai uint32x2_t vqrshl_u32(uint32x2_t __p0, int32x2_t __p1) {
20092 uint32x2_t __ret;
20093 __ret = (uint32x2_t) __builtin_neon_vqrshl_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
20094 return __ret;
20095 }
20096 #else
vqrshl_u32(uint32x2_t __p0,int32x2_t __p1)20097 __ai uint32x2_t vqrshl_u32(uint32x2_t __p0, int32x2_t __p1) {
20098 uint32x2_t __ret;
20099 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
20100 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
20101 __ret = (uint32x2_t) __builtin_neon_vqrshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
20102 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
20103 return __ret;
20104 }
20105 #endif
20106
vqrshl_u64(uint64x1_t __p0,int64x1_t __p1)20107 __ai uint64x1_t vqrshl_u64(uint64x1_t __p0, int64x1_t __p1) {
20108 uint64x1_t __ret;
20109 __ret = (uint64x1_t) __builtin_neon_vqrshl_v((int8x8_t)__p0, (int8x8_t)__p1, 19);
20110 return __ret;
20111 }
20112 #ifdef __LITTLE_ENDIAN__
vqrshl_u16(uint16x4_t __p0,int16x4_t __p1)20113 __ai uint16x4_t vqrshl_u16(uint16x4_t __p0, int16x4_t __p1) {
20114 uint16x4_t __ret;
20115 __ret = (uint16x4_t) __builtin_neon_vqrshl_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
20116 return __ret;
20117 }
20118 #else
vqrshl_u16(uint16x4_t __p0,int16x4_t __p1)20119 __ai uint16x4_t vqrshl_u16(uint16x4_t __p0, int16x4_t __p1) {
20120 uint16x4_t __ret;
20121 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
20122 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
20123 __ret = (uint16x4_t) __builtin_neon_vqrshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
20124 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
20125 return __ret;
20126 }
20127 #endif
20128
20129 #ifdef __LITTLE_ENDIAN__
vqrshl_s8(int8x8_t __p0,int8x8_t __p1)20130 __ai int8x8_t vqrshl_s8(int8x8_t __p0, int8x8_t __p1) {
20131 int8x8_t __ret;
20132 __ret = (int8x8_t) __builtin_neon_vqrshl_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
20133 return __ret;
20134 }
20135 #else
vqrshl_s8(int8x8_t __p0,int8x8_t __p1)20136 __ai int8x8_t vqrshl_s8(int8x8_t __p0, int8x8_t __p1) {
20137 int8x8_t __ret;
20138 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
20139 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
20140 __ret = (int8x8_t) __builtin_neon_vqrshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 0);
20141 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
20142 return __ret;
20143 }
20144 #endif
20145
20146 #ifdef __LITTLE_ENDIAN__
vqrshl_s32(int32x2_t __p0,int32x2_t __p1)20147 __ai int32x2_t vqrshl_s32(int32x2_t __p0, int32x2_t __p1) {
20148 int32x2_t __ret;
20149 __ret = (int32x2_t) __builtin_neon_vqrshl_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
20150 return __ret;
20151 }
20152 #else
vqrshl_s32(int32x2_t __p0,int32x2_t __p1)20153 __ai int32x2_t vqrshl_s32(int32x2_t __p0, int32x2_t __p1) {
20154 int32x2_t __ret;
20155 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
20156 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
20157 __ret = (int32x2_t) __builtin_neon_vqrshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
20158 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
20159 return __ret;
20160 }
20161 #endif
20162
vqrshl_s64(int64x1_t __p0,int64x1_t __p1)20163 __ai int64x1_t vqrshl_s64(int64x1_t __p0, int64x1_t __p1) {
20164 int64x1_t __ret;
20165 __ret = (int64x1_t) __builtin_neon_vqrshl_v((int8x8_t)__p0, (int8x8_t)__p1, 3);
20166 return __ret;
20167 }
20168 #ifdef __LITTLE_ENDIAN__
vqrshl_s16(int16x4_t __p0,int16x4_t __p1)20169 __ai int16x4_t vqrshl_s16(int16x4_t __p0, int16x4_t __p1) {
20170 int16x4_t __ret;
20171 __ret = (int16x4_t) __builtin_neon_vqrshl_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
20172 return __ret;
20173 }
20174 #else
vqrshl_s16(int16x4_t __p0,int16x4_t __p1)20175 __ai int16x4_t vqrshl_s16(int16x4_t __p0, int16x4_t __p1) {
20176 int16x4_t __ret;
20177 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
20178 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
20179 __ret = (int16x4_t) __builtin_neon_vqrshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
20180 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
20181 return __ret;
20182 }
20183 #endif
20184
20185 #ifdef __LITTLE_ENDIAN__
20186 #define vqrshrn_n_u32(__p0, __p1) __extension__ ({ \
20187 uint16x4_t __ret; \
20188 uint32x4_t __s0 = __p0; \
20189 __ret = (uint16x4_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__s0, __p1, 17); \
20190 __ret; \
20191 })
20192 #else
20193 #define vqrshrn_n_u32(__p0, __p1) __extension__ ({ \
20194 uint16x4_t __ret; \
20195 uint32x4_t __s0 = __p0; \
20196 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
20197 __ret = (uint16x4_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__rev0, __p1, 17); \
20198 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
20199 __ret; \
20200 })
20201 #define __noswap_vqrshrn_n_u32(__p0, __p1) __extension__ ({ \
20202 uint16x4_t __ret; \
20203 uint32x4_t __s0 = __p0; \
20204 __ret = (uint16x4_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__s0, __p1, 17); \
20205 __ret; \
20206 })
20207 #endif
20208
20209 #ifdef __LITTLE_ENDIAN__
20210 #define vqrshrn_n_u64(__p0, __p1) __extension__ ({ \
20211 uint32x2_t __ret; \
20212 uint64x2_t __s0 = __p0; \
20213 __ret = (uint32x2_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__s0, __p1, 18); \
20214 __ret; \
20215 })
20216 #else
20217 #define vqrshrn_n_u64(__p0, __p1) __extension__ ({ \
20218 uint32x2_t __ret; \
20219 uint64x2_t __s0 = __p0; \
20220 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
20221 __ret = (uint32x2_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__rev0, __p1, 18); \
20222 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
20223 __ret; \
20224 })
20225 #define __noswap_vqrshrn_n_u64(__p0, __p1) __extension__ ({ \
20226 uint32x2_t __ret; \
20227 uint64x2_t __s0 = __p0; \
20228 __ret = (uint32x2_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__s0, __p1, 18); \
20229 __ret; \
20230 })
20231 #endif
20232
20233 #ifdef __LITTLE_ENDIAN__
20234 #define vqrshrn_n_u16(__p0, __p1) __extension__ ({ \
20235 uint8x8_t __ret; \
20236 uint16x8_t __s0 = __p0; \
20237 __ret = (uint8x8_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__s0, __p1, 16); \
20238 __ret; \
20239 })
20240 #else
20241 #define vqrshrn_n_u16(__p0, __p1) __extension__ ({ \
20242 uint8x8_t __ret; \
20243 uint16x8_t __s0 = __p0; \
20244 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
20245 __ret = (uint8x8_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__rev0, __p1, 16); \
20246 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
20247 __ret; \
20248 })
20249 #define __noswap_vqrshrn_n_u16(__p0, __p1) __extension__ ({ \
20250 uint8x8_t __ret; \
20251 uint16x8_t __s0 = __p0; \
20252 __ret = (uint8x8_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__s0, __p1, 16); \
20253 __ret; \
20254 })
20255 #endif
20256
20257 #ifdef __LITTLE_ENDIAN__
20258 #define vqrshrn_n_s32(__p0, __p1) __extension__ ({ \
20259 int16x4_t __ret; \
20260 int32x4_t __s0 = __p0; \
20261 __ret = (int16x4_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__s0, __p1, 1); \
20262 __ret; \
20263 })
20264 #else
20265 #define vqrshrn_n_s32(__p0, __p1) __extension__ ({ \
20266 int16x4_t __ret; \
20267 int32x4_t __s0 = __p0; \
20268 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
20269 __ret = (int16x4_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__rev0, __p1, 1); \
20270 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
20271 __ret; \
20272 })
20273 #define __noswap_vqrshrn_n_s32(__p0, __p1) __extension__ ({ \
20274 int16x4_t __ret; \
20275 int32x4_t __s0 = __p0; \
20276 __ret = (int16x4_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__s0, __p1, 1); \
20277 __ret; \
20278 })
20279 #endif
20280
20281 #ifdef __LITTLE_ENDIAN__
20282 #define vqrshrn_n_s64(__p0, __p1) __extension__ ({ \
20283 int32x2_t __ret; \
20284 int64x2_t __s0 = __p0; \
20285 __ret = (int32x2_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__s0, __p1, 2); \
20286 __ret; \
20287 })
20288 #else
20289 #define vqrshrn_n_s64(__p0, __p1) __extension__ ({ \
20290 int32x2_t __ret; \
20291 int64x2_t __s0 = __p0; \
20292 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
20293 __ret = (int32x2_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__rev0, __p1, 2); \
20294 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
20295 __ret; \
20296 })
20297 #define __noswap_vqrshrn_n_s64(__p0, __p1) __extension__ ({ \
20298 int32x2_t __ret; \
20299 int64x2_t __s0 = __p0; \
20300 __ret = (int32x2_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__s0, __p1, 2); \
20301 __ret; \
20302 })
20303 #endif
20304
20305 #ifdef __LITTLE_ENDIAN__
20306 #define vqrshrn_n_s16(__p0, __p1) __extension__ ({ \
20307 int8x8_t __ret; \
20308 int16x8_t __s0 = __p0; \
20309 __ret = (int8x8_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__s0, __p1, 0); \
20310 __ret; \
20311 })
20312 #else
20313 #define vqrshrn_n_s16(__p0, __p1) __extension__ ({ \
20314 int8x8_t __ret; \
20315 int16x8_t __s0 = __p0; \
20316 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
20317 __ret = (int8x8_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__rev0, __p1, 0); \
20318 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
20319 __ret; \
20320 })
20321 #define __noswap_vqrshrn_n_s16(__p0, __p1) __extension__ ({ \
20322 int8x8_t __ret; \
20323 int16x8_t __s0 = __p0; \
20324 __ret = (int8x8_t) __builtin_neon_vqrshrn_n_v((int8x16_t)__s0, __p1, 0); \
20325 __ret; \
20326 })
20327 #endif
20328
20329 #ifdef __LITTLE_ENDIAN__
20330 #define vqrshrun_n_s32(__p0, __p1) __extension__ ({ \
20331 uint16x4_t __ret; \
20332 int32x4_t __s0 = __p0; \
20333 __ret = (uint16x4_t) __builtin_neon_vqrshrun_n_v((int8x16_t)__s0, __p1, 17); \
20334 __ret; \
20335 })
20336 #else
20337 #define vqrshrun_n_s32(__p0, __p1) __extension__ ({ \
20338 uint16x4_t __ret; \
20339 int32x4_t __s0 = __p0; \
20340 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
20341 __ret = (uint16x4_t) __builtin_neon_vqrshrun_n_v((int8x16_t)__rev0, __p1, 17); \
20342 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
20343 __ret; \
20344 })
20345 #define __noswap_vqrshrun_n_s32(__p0, __p1) __extension__ ({ \
20346 uint16x4_t __ret; \
20347 int32x4_t __s0 = __p0; \
20348 __ret = (uint16x4_t) __builtin_neon_vqrshrun_n_v((int8x16_t)__s0, __p1, 17); \
20349 __ret; \
20350 })
20351 #endif
20352
20353 #ifdef __LITTLE_ENDIAN__
20354 #define vqrshrun_n_s64(__p0, __p1) __extension__ ({ \
20355 uint32x2_t __ret; \
20356 int64x2_t __s0 = __p0; \
20357 __ret = (uint32x2_t) __builtin_neon_vqrshrun_n_v((int8x16_t)__s0, __p1, 18); \
20358 __ret; \
20359 })
20360 #else
20361 #define vqrshrun_n_s64(__p0, __p1) __extension__ ({ \
20362 uint32x2_t __ret; \
20363 int64x2_t __s0 = __p0; \
20364 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
20365 __ret = (uint32x2_t) __builtin_neon_vqrshrun_n_v((int8x16_t)__rev0, __p1, 18); \
20366 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
20367 __ret; \
20368 })
20369 #define __noswap_vqrshrun_n_s64(__p0, __p1) __extension__ ({ \
20370 uint32x2_t __ret; \
20371 int64x2_t __s0 = __p0; \
20372 __ret = (uint32x2_t) __builtin_neon_vqrshrun_n_v((int8x16_t)__s0, __p1, 18); \
20373 __ret; \
20374 })
20375 #endif
20376
20377 #ifdef __LITTLE_ENDIAN__
20378 #define vqrshrun_n_s16(__p0, __p1) __extension__ ({ \
20379 uint8x8_t __ret; \
20380 int16x8_t __s0 = __p0; \
20381 __ret = (uint8x8_t) __builtin_neon_vqrshrun_n_v((int8x16_t)__s0, __p1, 16); \
20382 __ret; \
20383 })
20384 #else
20385 #define vqrshrun_n_s16(__p0, __p1) __extension__ ({ \
20386 uint8x8_t __ret; \
20387 int16x8_t __s0 = __p0; \
20388 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
20389 __ret = (uint8x8_t) __builtin_neon_vqrshrun_n_v((int8x16_t)__rev0, __p1, 16); \
20390 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
20391 __ret; \
20392 })
20393 #define __noswap_vqrshrun_n_s16(__p0, __p1) __extension__ ({ \
20394 uint8x8_t __ret; \
20395 int16x8_t __s0 = __p0; \
20396 __ret = (uint8x8_t) __builtin_neon_vqrshrun_n_v((int8x16_t)__s0, __p1, 16); \
20397 __ret; \
20398 })
20399 #endif
20400
20401 #ifdef __LITTLE_ENDIAN__
vqshlq_u8(uint8x16_t __p0,int8x16_t __p1)20402 __ai uint8x16_t vqshlq_u8(uint8x16_t __p0, int8x16_t __p1) {
20403 uint8x16_t __ret;
20404 __ret = (uint8x16_t) __builtin_neon_vqshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
20405 return __ret;
20406 }
20407 #else
vqshlq_u8(uint8x16_t __p0,int8x16_t __p1)20408 __ai uint8x16_t vqshlq_u8(uint8x16_t __p0, int8x16_t __p1) {
20409 uint8x16_t __ret;
20410 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
20411 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
20412 __ret = (uint8x16_t) __builtin_neon_vqshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
20413 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
20414 return __ret;
20415 }
20416 #endif
20417
20418 #ifdef __LITTLE_ENDIAN__
vqshlq_u32(uint32x4_t __p0,int32x4_t __p1)20419 __ai uint32x4_t vqshlq_u32(uint32x4_t __p0, int32x4_t __p1) {
20420 uint32x4_t __ret;
20421 __ret = (uint32x4_t) __builtin_neon_vqshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
20422 return __ret;
20423 }
20424 #else
vqshlq_u32(uint32x4_t __p0,int32x4_t __p1)20425 __ai uint32x4_t vqshlq_u32(uint32x4_t __p0, int32x4_t __p1) {
20426 uint32x4_t __ret;
20427 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
20428 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
20429 __ret = (uint32x4_t) __builtin_neon_vqshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
20430 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
20431 return __ret;
20432 }
20433 #endif
20434
20435 #ifdef __LITTLE_ENDIAN__
vqshlq_u64(uint64x2_t __p0,int64x2_t __p1)20436 __ai uint64x2_t vqshlq_u64(uint64x2_t __p0, int64x2_t __p1) {
20437 uint64x2_t __ret;
20438 __ret = (uint64x2_t) __builtin_neon_vqshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 51);
20439 return __ret;
20440 }
20441 #else
vqshlq_u64(uint64x2_t __p0,int64x2_t __p1)20442 __ai uint64x2_t vqshlq_u64(uint64x2_t __p0, int64x2_t __p1) {
20443 uint64x2_t __ret;
20444 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
20445 int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
20446 __ret = (uint64x2_t) __builtin_neon_vqshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 51);
20447 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
20448 return __ret;
20449 }
20450 #endif
20451
20452 #ifdef __LITTLE_ENDIAN__
vqshlq_u16(uint16x8_t __p0,int16x8_t __p1)20453 __ai uint16x8_t vqshlq_u16(uint16x8_t __p0, int16x8_t __p1) {
20454 uint16x8_t __ret;
20455 __ret = (uint16x8_t) __builtin_neon_vqshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
20456 return __ret;
20457 }
20458 #else
vqshlq_u16(uint16x8_t __p0,int16x8_t __p1)20459 __ai uint16x8_t vqshlq_u16(uint16x8_t __p0, int16x8_t __p1) {
20460 uint16x8_t __ret;
20461 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
20462 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
20463 __ret = (uint16x8_t) __builtin_neon_vqshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
20464 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
20465 return __ret;
20466 }
20467 #endif
20468
20469 #ifdef __LITTLE_ENDIAN__
vqshlq_s8(int8x16_t __p0,int8x16_t __p1)20470 __ai int8x16_t vqshlq_s8(int8x16_t __p0, int8x16_t __p1) {
20471 int8x16_t __ret;
20472 __ret = (int8x16_t) __builtin_neon_vqshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 32);
20473 return __ret;
20474 }
20475 #else
vqshlq_s8(int8x16_t __p0,int8x16_t __p1)20476 __ai int8x16_t vqshlq_s8(int8x16_t __p0, int8x16_t __p1) {
20477 int8x16_t __ret;
20478 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
20479 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
20480 __ret = (int8x16_t) __builtin_neon_vqshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 32);
20481 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
20482 return __ret;
20483 }
20484 #endif
20485
20486 #ifdef __LITTLE_ENDIAN__
vqshlq_s32(int32x4_t __p0,int32x4_t __p1)20487 __ai int32x4_t vqshlq_s32(int32x4_t __p0, int32x4_t __p1) {
20488 int32x4_t __ret;
20489 __ret = (int32x4_t) __builtin_neon_vqshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
20490 return __ret;
20491 }
20492 #else
vqshlq_s32(int32x4_t __p0,int32x4_t __p1)20493 __ai int32x4_t vqshlq_s32(int32x4_t __p0, int32x4_t __p1) {
20494 int32x4_t __ret;
20495 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
20496 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
20497 __ret = (int32x4_t) __builtin_neon_vqshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
20498 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
20499 return __ret;
20500 }
20501 #endif
20502
20503 #ifdef __LITTLE_ENDIAN__
vqshlq_s64(int64x2_t __p0,int64x2_t __p1)20504 __ai int64x2_t vqshlq_s64(int64x2_t __p0, int64x2_t __p1) {
20505 int64x2_t __ret;
20506 __ret = (int64x2_t) __builtin_neon_vqshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 35);
20507 return __ret;
20508 }
20509 #else
vqshlq_s64(int64x2_t __p0,int64x2_t __p1)20510 __ai int64x2_t vqshlq_s64(int64x2_t __p0, int64x2_t __p1) {
20511 int64x2_t __ret;
20512 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
20513 int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
20514 __ret = (int64x2_t) __builtin_neon_vqshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 35);
20515 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
20516 return __ret;
20517 }
20518 #endif
20519
20520 #ifdef __LITTLE_ENDIAN__
vqshlq_s16(int16x8_t __p0,int16x8_t __p1)20521 __ai int16x8_t vqshlq_s16(int16x8_t __p0, int16x8_t __p1) {
20522 int16x8_t __ret;
20523 __ret = (int16x8_t) __builtin_neon_vqshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
20524 return __ret;
20525 }
20526 #else
vqshlq_s16(int16x8_t __p0,int16x8_t __p1)20527 __ai int16x8_t vqshlq_s16(int16x8_t __p0, int16x8_t __p1) {
20528 int16x8_t __ret;
20529 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
20530 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
20531 __ret = (int16x8_t) __builtin_neon_vqshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
20532 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
20533 return __ret;
20534 }
20535 #endif
20536
20537 #ifdef __LITTLE_ENDIAN__
vqshl_u8(uint8x8_t __p0,int8x8_t __p1)20538 __ai uint8x8_t vqshl_u8(uint8x8_t __p0, int8x8_t __p1) {
20539 uint8x8_t __ret;
20540 __ret = (uint8x8_t) __builtin_neon_vqshl_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
20541 return __ret;
20542 }
20543 #else
vqshl_u8(uint8x8_t __p0,int8x8_t __p1)20544 __ai uint8x8_t vqshl_u8(uint8x8_t __p0, int8x8_t __p1) {
20545 uint8x8_t __ret;
20546 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
20547 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
20548 __ret = (uint8x8_t) __builtin_neon_vqshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
20549 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
20550 return __ret;
20551 }
20552 #endif
20553
20554 #ifdef __LITTLE_ENDIAN__
vqshl_u32(uint32x2_t __p0,int32x2_t __p1)20555 __ai uint32x2_t vqshl_u32(uint32x2_t __p0, int32x2_t __p1) {
20556 uint32x2_t __ret;
20557 __ret = (uint32x2_t) __builtin_neon_vqshl_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
20558 return __ret;
20559 }
20560 #else
vqshl_u32(uint32x2_t __p0,int32x2_t __p1)20561 __ai uint32x2_t vqshl_u32(uint32x2_t __p0, int32x2_t __p1) {
20562 uint32x2_t __ret;
20563 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
20564 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
20565 __ret = (uint32x2_t) __builtin_neon_vqshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
20566 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
20567 return __ret;
20568 }
20569 #endif
20570
vqshl_u64(uint64x1_t __p0,int64x1_t __p1)20571 __ai uint64x1_t vqshl_u64(uint64x1_t __p0, int64x1_t __p1) {
20572 uint64x1_t __ret;
20573 __ret = (uint64x1_t) __builtin_neon_vqshl_v((int8x8_t)__p0, (int8x8_t)__p1, 19);
20574 return __ret;
20575 }
20576 #ifdef __LITTLE_ENDIAN__
vqshl_u16(uint16x4_t __p0,int16x4_t __p1)20577 __ai uint16x4_t vqshl_u16(uint16x4_t __p0, int16x4_t __p1) {
20578 uint16x4_t __ret;
20579 __ret = (uint16x4_t) __builtin_neon_vqshl_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
20580 return __ret;
20581 }
20582 #else
vqshl_u16(uint16x4_t __p0,int16x4_t __p1)20583 __ai uint16x4_t vqshl_u16(uint16x4_t __p0, int16x4_t __p1) {
20584 uint16x4_t __ret;
20585 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
20586 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
20587 __ret = (uint16x4_t) __builtin_neon_vqshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
20588 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
20589 return __ret;
20590 }
20591 #endif
20592
20593 #ifdef __LITTLE_ENDIAN__
vqshl_s8(int8x8_t __p0,int8x8_t __p1)20594 __ai int8x8_t vqshl_s8(int8x8_t __p0, int8x8_t __p1) {
20595 int8x8_t __ret;
20596 __ret = (int8x8_t) __builtin_neon_vqshl_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
20597 return __ret;
20598 }
20599 #else
vqshl_s8(int8x8_t __p0,int8x8_t __p1)20600 __ai int8x8_t vqshl_s8(int8x8_t __p0, int8x8_t __p1) {
20601 int8x8_t __ret;
20602 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
20603 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
20604 __ret = (int8x8_t) __builtin_neon_vqshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 0);
20605 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
20606 return __ret;
20607 }
20608 #endif
20609
20610 #ifdef __LITTLE_ENDIAN__
vqshl_s32(int32x2_t __p0,int32x2_t __p1)20611 __ai int32x2_t vqshl_s32(int32x2_t __p0, int32x2_t __p1) {
20612 int32x2_t __ret;
20613 __ret = (int32x2_t) __builtin_neon_vqshl_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
20614 return __ret;
20615 }
20616 #else
vqshl_s32(int32x2_t __p0,int32x2_t __p1)20617 __ai int32x2_t vqshl_s32(int32x2_t __p0, int32x2_t __p1) {
20618 int32x2_t __ret;
20619 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
20620 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
20621 __ret = (int32x2_t) __builtin_neon_vqshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
20622 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
20623 return __ret;
20624 }
20625 #endif
20626
vqshl_s64(int64x1_t __p0,int64x1_t __p1)20627 __ai int64x1_t vqshl_s64(int64x1_t __p0, int64x1_t __p1) {
20628 int64x1_t __ret;
20629 __ret = (int64x1_t) __builtin_neon_vqshl_v((int8x8_t)__p0, (int8x8_t)__p1, 3);
20630 return __ret;
20631 }
20632 #ifdef __LITTLE_ENDIAN__
vqshl_s16(int16x4_t __p0,int16x4_t __p1)20633 __ai int16x4_t vqshl_s16(int16x4_t __p0, int16x4_t __p1) {
20634 int16x4_t __ret;
20635 __ret = (int16x4_t) __builtin_neon_vqshl_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
20636 return __ret;
20637 }
20638 #else
vqshl_s16(int16x4_t __p0,int16x4_t __p1)20639 __ai int16x4_t vqshl_s16(int16x4_t __p0, int16x4_t __p1) {
20640 int16x4_t __ret;
20641 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
20642 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
20643 __ret = (int16x4_t) __builtin_neon_vqshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
20644 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
20645 return __ret;
20646 }
20647 #endif
20648
20649 #ifdef __LITTLE_ENDIAN__
20650 #define vqshlq_n_u8(__p0, __p1) __extension__ ({ \
20651 uint8x16_t __ret; \
20652 uint8x16_t __s0 = __p0; \
20653 __ret = (uint8x16_t) __builtin_neon_vqshlq_n_v((int8x16_t)__s0, __p1, 48); \
20654 __ret; \
20655 })
20656 #else
20657 #define vqshlq_n_u8(__p0, __p1) __extension__ ({ \
20658 uint8x16_t __ret; \
20659 uint8x16_t __s0 = __p0; \
20660 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
20661 __ret = (uint8x16_t) __builtin_neon_vqshlq_n_v((int8x16_t)__rev0, __p1, 48); \
20662 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
20663 __ret; \
20664 })
20665 #endif
20666
20667 #ifdef __LITTLE_ENDIAN__
20668 #define vqshlq_n_u32(__p0, __p1) __extension__ ({ \
20669 uint32x4_t __ret; \
20670 uint32x4_t __s0 = __p0; \
20671 __ret = (uint32x4_t) __builtin_neon_vqshlq_n_v((int8x16_t)__s0, __p1, 50); \
20672 __ret; \
20673 })
20674 #else
20675 #define vqshlq_n_u32(__p0, __p1) __extension__ ({ \
20676 uint32x4_t __ret; \
20677 uint32x4_t __s0 = __p0; \
20678 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
20679 __ret = (uint32x4_t) __builtin_neon_vqshlq_n_v((int8x16_t)__rev0, __p1, 50); \
20680 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
20681 __ret; \
20682 })
20683 #endif
20684
20685 #ifdef __LITTLE_ENDIAN__
20686 #define vqshlq_n_u64(__p0, __p1) __extension__ ({ \
20687 uint64x2_t __ret; \
20688 uint64x2_t __s0 = __p0; \
20689 __ret = (uint64x2_t) __builtin_neon_vqshlq_n_v((int8x16_t)__s0, __p1, 51); \
20690 __ret; \
20691 })
20692 #else
20693 #define vqshlq_n_u64(__p0, __p1) __extension__ ({ \
20694 uint64x2_t __ret; \
20695 uint64x2_t __s0 = __p0; \
20696 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
20697 __ret = (uint64x2_t) __builtin_neon_vqshlq_n_v((int8x16_t)__rev0, __p1, 51); \
20698 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
20699 __ret; \
20700 })
20701 #endif
20702
20703 #ifdef __LITTLE_ENDIAN__
20704 #define vqshlq_n_u16(__p0, __p1) __extension__ ({ \
20705 uint16x8_t __ret; \
20706 uint16x8_t __s0 = __p0; \
20707 __ret = (uint16x8_t) __builtin_neon_vqshlq_n_v((int8x16_t)__s0, __p1, 49); \
20708 __ret; \
20709 })
20710 #else
20711 #define vqshlq_n_u16(__p0, __p1) __extension__ ({ \
20712 uint16x8_t __ret; \
20713 uint16x8_t __s0 = __p0; \
20714 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
20715 __ret = (uint16x8_t) __builtin_neon_vqshlq_n_v((int8x16_t)__rev0, __p1, 49); \
20716 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
20717 __ret; \
20718 })
20719 #endif
20720
20721 #ifdef __LITTLE_ENDIAN__
20722 #define vqshlq_n_s8(__p0, __p1) __extension__ ({ \
20723 int8x16_t __ret; \
20724 int8x16_t __s0 = __p0; \
20725 __ret = (int8x16_t) __builtin_neon_vqshlq_n_v((int8x16_t)__s0, __p1, 32); \
20726 __ret; \
20727 })
20728 #else
20729 #define vqshlq_n_s8(__p0, __p1) __extension__ ({ \
20730 int8x16_t __ret; \
20731 int8x16_t __s0 = __p0; \
20732 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
20733 __ret = (int8x16_t) __builtin_neon_vqshlq_n_v((int8x16_t)__rev0, __p1, 32); \
20734 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
20735 __ret; \
20736 })
20737 #endif
20738
20739 #ifdef __LITTLE_ENDIAN__
20740 #define vqshlq_n_s32(__p0, __p1) __extension__ ({ \
20741 int32x4_t __ret; \
20742 int32x4_t __s0 = __p0; \
20743 __ret = (int32x4_t) __builtin_neon_vqshlq_n_v((int8x16_t)__s0, __p1, 34); \
20744 __ret; \
20745 })
20746 #else
20747 #define vqshlq_n_s32(__p0, __p1) __extension__ ({ \
20748 int32x4_t __ret; \
20749 int32x4_t __s0 = __p0; \
20750 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
20751 __ret = (int32x4_t) __builtin_neon_vqshlq_n_v((int8x16_t)__rev0, __p1, 34); \
20752 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
20753 __ret; \
20754 })
20755 #endif
20756
20757 #ifdef __LITTLE_ENDIAN__
20758 #define vqshlq_n_s64(__p0, __p1) __extension__ ({ \
20759 int64x2_t __ret; \
20760 int64x2_t __s0 = __p0; \
20761 __ret = (int64x2_t) __builtin_neon_vqshlq_n_v((int8x16_t)__s0, __p1, 35); \
20762 __ret; \
20763 })
20764 #else
20765 #define vqshlq_n_s64(__p0, __p1) __extension__ ({ \
20766 int64x2_t __ret; \
20767 int64x2_t __s0 = __p0; \
20768 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
20769 __ret = (int64x2_t) __builtin_neon_vqshlq_n_v((int8x16_t)__rev0, __p1, 35); \
20770 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
20771 __ret; \
20772 })
20773 #endif
20774
20775 #ifdef __LITTLE_ENDIAN__
20776 #define vqshlq_n_s16(__p0, __p1) __extension__ ({ \
20777 int16x8_t __ret; \
20778 int16x8_t __s0 = __p0; \
20779 __ret = (int16x8_t) __builtin_neon_vqshlq_n_v((int8x16_t)__s0, __p1, 33); \
20780 __ret; \
20781 })
20782 #else
20783 #define vqshlq_n_s16(__p0, __p1) __extension__ ({ \
20784 int16x8_t __ret; \
20785 int16x8_t __s0 = __p0; \
20786 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
20787 __ret = (int16x8_t) __builtin_neon_vqshlq_n_v((int8x16_t)__rev0, __p1, 33); \
20788 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
20789 __ret; \
20790 })
20791 #endif
20792
20793 #ifdef __LITTLE_ENDIAN__
20794 #define vqshl_n_u8(__p0, __p1) __extension__ ({ \
20795 uint8x8_t __ret; \
20796 uint8x8_t __s0 = __p0; \
20797 __ret = (uint8x8_t) __builtin_neon_vqshl_n_v((int8x8_t)__s0, __p1, 16); \
20798 __ret; \
20799 })
20800 #else
20801 #define vqshl_n_u8(__p0, __p1) __extension__ ({ \
20802 uint8x8_t __ret; \
20803 uint8x8_t __s0 = __p0; \
20804 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
20805 __ret = (uint8x8_t) __builtin_neon_vqshl_n_v((int8x8_t)__rev0, __p1, 16); \
20806 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
20807 __ret; \
20808 })
20809 #endif
20810
20811 #ifdef __LITTLE_ENDIAN__
20812 #define vqshl_n_u32(__p0, __p1) __extension__ ({ \
20813 uint32x2_t __ret; \
20814 uint32x2_t __s0 = __p0; \
20815 __ret = (uint32x2_t) __builtin_neon_vqshl_n_v((int8x8_t)__s0, __p1, 18); \
20816 __ret; \
20817 })
20818 #else
20819 #define vqshl_n_u32(__p0, __p1) __extension__ ({ \
20820 uint32x2_t __ret; \
20821 uint32x2_t __s0 = __p0; \
20822 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
20823 __ret = (uint32x2_t) __builtin_neon_vqshl_n_v((int8x8_t)__rev0, __p1, 18); \
20824 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
20825 __ret; \
20826 })
20827 #endif
20828
20829 #define vqshl_n_u64(__p0, __p1) __extension__ ({ \
20830 uint64x1_t __ret; \
20831 uint64x1_t __s0 = __p0; \
20832 __ret = (uint64x1_t) __builtin_neon_vqshl_n_v((int8x8_t)__s0, __p1, 19); \
20833 __ret; \
20834 })
20835 #ifdef __LITTLE_ENDIAN__
20836 #define vqshl_n_u16(__p0, __p1) __extension__ ({ \
20837 uint16x4_t __ret; \
20838 uint16x4_t __s0 = __p0; \
20839 __ret = (uint16x4_t) __builtin_neon_vqshl_n_v((int8x8_t)__s0, __p1, 17); \
20840 __ret; \
20841 })
20842 #else
20843 #define vqshl_n_u16(__p0, __p1) __extension__ ({ \
20844 uint16x4_t __ret; \
20845 uint16x4_t __s0 = __p0; \
20846 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
20847 __ret = (uint16x4_t) __builtin_neon_vqshl_n_v((int8x8_t)__rev0, __p1, 17); \
20848 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
20849 __ret; \
20850 })
20851 #endif
20852
20853 #ifdef __LITTLE_ENDIAN__
20854 #define vqshl_n_s8(__p0, __p1) __extension__ ({ \
20855 int8x8_t __ret; \
20856 int8x8_t __s0 = __p0; \
20857 __ret = (int8x8_t) __builtin_neon_vqshl_n_v((int8x8_t)__s0, __p1, 0); \
20858 __ret; \
20859 })
20860 #else
20861 #define vqshl_n_s8(__p0, __p1) __extension__ ({ \
20862 int8x8_t __ret; \
20863 int8x8_t __s0 = __p0; \
20864 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
20865 __ret = (int8x8_t) __builtin_neon_vqshl_n_v((int8x8_t)__rev0, __p1, 0); \
20866 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
20867 __ret; \
20868 })
20869 #endif
20870
20871 #ifdef __LITTLE_ENDIAN__
20872 #define vqshl_n_s32(__p0, __p1) __extension__ ({ \
20873 int32x2_t __ret; \
20874 int32x2_t __s0 = __p0; \
20875 __ret = (int32x2_t) __builtin_neon_vqshl_n_v((int8x8_t)__s0, __p1, 2); \
20876 __ret; \
20877 })
20878 #else
20879 #define vqshl_n_s32(__p0, __p1) __extension__ ({ \
20880 int32x2_t __ret; \
20881 int32x2_t __s0 = __p0; \
20882 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
20883 __ret = (int32x2_t) __builtin_neon_vqshl_n_v((int8x8_t)__rev0, __p1, 2); \
20884 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
20885 __ret; \
20886 })
20887 #endif
20888
20889 #define vqshl_n_s64(__p0, __p1) __extension__ ({ \
20890 int64x1_t __ret; \
20891 int64x1_t __s0 = __p0; \
20892 __ret = (int64x1_t) __builtin_neon_vqshl_n_v((int8x8_t)__s0, __p1, 3); \
20893 __ret; \
20894 })
20895 #ifdef __LITTLE_ENDIAN__
20896 #define vqshl_n_s16(__p0, __p1) __extension__ ({ \
20897 int16x4_t __ret; \
20898 int16x4_t __s0 = __p0; \
20899 __ret = (int16x4_t) __builtin_neon_vqshl_n_v((int8x8_t)__s0, __p1, 1); \
20900 __ret; \
20901 })
20902 #else
20903 #define vqshl_n_s16(__p0, __p1) __extension__ ({ \
20904 int16x4_t __ret; \
20905 int16x4_t __s0 = __p0; \
20906 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
20907 __ret = (int16x4_t) __builtin_neon_vqshl_n_v((int8x8_t)__rev0, __p1, 1); \
20908 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
20909 __ret; \
20910 })
20911 #endif
20912
20913 #ifdef __LITTLE_ENDIAN__
20914 #define vqshluq_n_s8(__p0, __p1) __extension__ ({ \
20915 uint8x16_t __ret; \
20916 int8x16_t __s0 = __p0; \
20917 __ret = (uint8x16_t) __builtin_neon_vqshluq_n_v((int8x16_t)__s0, __p1, 48); \
20918 __ret; \
20919 })
20920 #else
20921 #define vqshluq_n_s8(__p0, __p1) __extension__ ({ \
20922 uint8x16_t __ret; \
20923 int8x16_t __s0 = __p0; \
20924 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
20925 __ret = (uint8x16_t) __builtin_neon_vqshluq_n_v((int8x16_t)__rev0, __p1, 48); \
20926 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
20927 __ret; \
20928 })
20929 #endif
20930
20931 #ifdef __LITTLE_ENDIAN__
20932 #define vqshluq_n_s32(__p0, __p1) __extension__ ({ \
20933 uint32x4_t __ret; \
20934 int32x4_t __s0 = __p0; \
20935 __ret = (uint32x4_t) __builtin_neon_vqshluq_n_v((int8x16_t)__s0, __p1, 50); \
20936 __ret; \
20937 })
20938 #else
20939 #define vqshluq_n_s32(__p0, __p1) __extension__ ({ \
20940 uint32x4_t __ret; \
20941 int32x4_t __s0 = __p0; \
20942 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
20943 __ret = (uint32x4_t) __builtin_neon_vqshluq_n_v((int8x16_t)__rev0, __p1, 50); \
20944 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
20945 __ret; \
20946 })
20947 #endif
20948
20949 #ifdef __LITTLE_ENDIAN__
20950 #define vqshluq_n_s64(__p0, __p1) __extension__ ({ \
20951 uint64x2_t __ret; \
20952 int64x2_t __s0 = __p0; \
20953 __ret = (uint64x2_t) __builtin_neon_vqshluq_n_v((int8x16_t)__s0, __p1, 51); \
20954 __ret; \
20955 })
20956 #else
20957 #define vqshluq_n_s64(__p0, __p1) __extension__ ({ \
20958 uint64x2_t __ret; \
20959 int64x2_t __s0 = __p0; \
20960 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
20961 __ret = (uint64x2_t) __builtin_neon_vqshluq_n_v((int8x16_t)__rev0, __p1, 51); \
20962 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
20963 __ret; \
20964 })
20965 #endif
20966
20967 #ifdef __LITTLE_ENDIAN__
20968 #define vqshluq_n_s16(__p0, __p1) __extension__ ({ \
20969 uint16x8_t __ret; \
20970 int16x8_t __s0 = __p0; \
20971 __ret = (uint16x8_t) __builtin_neon_vqshluq_n_v((int8x16_t)__s0, __p1, 49); \
20972 __ret; \
20973 })
20974 #else
20975 #define vqshluq_n_s16(__p0, __p1) __extension__ ({ \
20976 uint16x8_t __ret; \
20977 int16x8_t __s0 = __p0; \
20978 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
20979 __ret = (uint16x8_t) __builtin_neon_vqshluq_n_v((int8x16_t)__rev0, __p1, 49); \
20980 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
20981 __ret; \
20982 })
20983 #endif
20984
20985 #ifdef __LITTLE_ENDIAN__
20986 #define vqshlu_n_s8(__p0, __p1) __extension__ ({ \
20987 uint8x8_t __ret; \
20988 int8x8_t __s0 = __p0; \
20989 __ret = (uint8x8_t) __builtin_neon_vqshlu_n_v((int8x8_t)__s0, __p1, 16); \
20990 __ret; \
20991 })
20992 #else
20993 #define vqshlu_n_s8(__p0, __p1) __extension__ ({ \
20994 uint8x8_t __ret; \
20995 int8x8_t __s0 = __p0; \
20996 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
20997 __ret = (uint8x8_t) __builtin_neon_vqshlu_n_v((int8x8_t)__rev0, __p1, 16); \
20998 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
20999 __ret; \
21000 })
21001 #endif
21002
21003 #ifdef __LITTLE_ENDIAN__
21004 #define vqshlu_n_s32(__p0, __p1) __extension__ ({ \
21005 uint32x2_t __ret; \
21006 int32x2_t __s0 = __p0; \
21007 __ret = (uint32x2_t) __builtin_neon_vqshlu_n_v((int8x8_t)__s0, __p1, 18); \
21008 __ret; \
21009 })
21010 #else
21011 #define vqshlu_n_s32(__p0, __p1) __extension__ ({ \
21012 uint32x2_t __ret; \
21013 int32x2_t __s0 = __p0; \
21014 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
21015 __ret = (uint32x2_t) __builtin_neon_vqshlu_n_v((int8x8_t)__rev0, __p1, 18); \
21016 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
21017 __ret; \
21018 })
21019 #endif
21020
21021 #define vqshlu_n_s64(__p0, __p1) __extension__ ({ \
21022 uint64x1_t __ret; \
21023 int64x1_t __s0 = __p0; \
21024 __ret = (uint64x1_t) __builtin_neon_vqshlu_n_v((int8x8_t)__s0, __p1, 19); \
21025 __ret; \
21026 })
21027 #ifdef __LITTLE_ENDIAN__
21028 #define vqshlu_n_s16(__p0, __p1) __extension__ ({ \
21029 uint16x4_t __ret; \
21030 int16x4_t __s0 = __p0; \
21031 __ret = (uint16x4_t) __builtin_neon_vqshlu_n_v((int8x8_t)__s0, __p1, 17); \
21032 __ret; \
21033 })
21034 #else
21035 #define vqshlu_n_s16(__p0, __p1) __extension__ ({ \
21036 uint16x4_t __ret; \
21037 int16x4_t __s0 = __p0; \
21038 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
21039 __ret = (uint16x4_t) __builtin_neon_vqshlu_n_v((int8x8_t)__rev0, __p1, 17); \
21040 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
21041 __ret; \
21042 })
21043 #endif
21044
21045 #ifdef __LITTLE_ENDIAN__
21046 #define vqshrn_n_u32(__p0, __p1) __extension__ ({ \
21047 uint16x4_t __ret; \
21048 uint32x4_t __s0 = __p0; \
21049 __ret = (uint16x4_t) __builtin_neon_vqshrn_n_v((int8x16_t)__s0, __p1, 17); \
21050 __ret; \
21051 })
21052 #else
21053 #define vqshrn_n_u32(__p0, __p1) __extension__ ({ \
21054 uint16x4_t __ret; \
21055 uint32x4_t __s0 = __p0; \
21056 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
21057 __ret = (uint16x4_t) __builtin_neon_vqshrn_n_v((int8x16_t)__rev0, __p1, 17); \
21058 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
21059 __ret; \
21060 })
21061 #define __noswap_vqshrn_n_u32(__p0, __p1) __extension__ ({ \
21062 uint16x4_t __ret; \
21063 uint32x4_t __s0 = __p0; \
21064 __ret = (uint16x4_t) __builtin_neon_vqshrn_n_v((int8x16_t)__s0, __p1, 17); \
21065 __ret; \
21066 })
21067 #endif
21068
21069 #ifdef __LITTLE_ENDIAN__
21070 #define vqshrn_n_u64(__p0, __p1) __extension__ ({ \
21071 uint32x2_t __ret; \
21072 uint64x2_t __s0 = __p0; \
21073 __ret = (uint32x2_t) __builtin_neon_vqshrn_n_v((int8x16_t)__s0, __p1, 18); \
21074 __ret; \
21075 })
21076 #else
21077 #define vqshrn_n_u64(__p0, __p1) __extension__ ({ \
21078 uint32x2_t __ret; \
21079 uint64x2_t __s0 = __p0; \
21080 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
21081 __ret = (uint32x2_t) __builtin_neon_vqshrn_n_v((int8x16_t)__rev0, __p1, 18); \
21082 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
21083 __ret; \
21084 })
21085 #define __noswap_vqshrn_n_u64(__p0, __p1) __extension__ ({ \
21086 uint32x2_t __ret; \
21087 uint64x2_t __s0 = __p0; \
21088 __ret = (uint32x2_t) __builtin_neon_vqshrn_n_v((int8x16_t)__s0, __p1, 18); \
21089 __ret; \
21090 })
21091 #endif
21092
21093 #ifdef __LITTLE_ENDIAN__
21094 #define vqshrn_n_u16(__p0, __p1) __extension__ ({ \
21095 uint8x8_t __ret; \
21096 uint16x8_t __s0 = __p0; \
21097 __ret = (uint8x8_t) __builtin_neon_vqshrn_n_v((int8x16_t)__s0, __p1, 16); \
21098 __ret; \
21099 })
21100 #else
21101 #define vqshrn_n_u16(__p0, __p1) __extension__ ({ \
21102 uint8x8_t __ret; \
21103 uint16x8_t __s0 = __p0; \
21104 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
21105 __ret = (uint8x8_t) __builtin_neon_vqshrn_n_v((int8x16_t)__rev0, __p1, 16); \
21106 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
21107 __ret; \
21108 })
21109 #define __noswap_vqshrn_n_u16(__p0, __p1) __extension__ ({ \
21110 uint8x8_t __ret; \
21111 uint16x8_t __s0 = __p0; \
21112 __ret = (uint8x8_t) __builtin_neon_vqshrn_n_v((int8x16_t)__s0, __p1, 16); \
21113 __ret; \
21114 })
21115 #endif
21116
21117 #ifdef __LITTLE_ENDIAN__
21118 #define vqshrn_n_s32(__p0, __p1) __extension__ ({ \
21119 int16x4_t __ret; \
21120 int32x4_t __s0 = __p0; \
21121 __ret = (int16x4_t) __builtin_neon_vqshrn_n_v((int8x16_t)__s0, __p1, 1); \
21122 __ret; \
21123 })
21124 #else
21125 #define vqshrn_n_s32(__p0, __p1) __extension__ ({ \
21126 int16x4_t __ret; \
21127 int32x4_t __s0 = __p0; \
21128 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
21129 __ret = (int16x4_t) __builtin_neon_vqshrn_n_v((int8x16_t)__rev0, __p1, 1); \
21130 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
21131 __ret; \
21132 })
21133 #define __noswap_vqshrn_n_s32(__p0, __p1) __extension__ ({ \
21134 int16x4_t __ret; \
21135 int32x4_t __s0 = __p0; \
21136 __ret = (int16x4_t) __builtin_neon_vqshrn_n_v((int8x16_t)__s0, __p1, 1); \
21137 __ret; \
21138 })
21139 #endif
21140
21141 #ifdef __LITTLE_ENDIAN__
21142 #define vqshrn_n_s64(__p0, __p1) __extension__ ({ \
21143 int32x2_t __ret; \
21144 int64x2_t __s0 = __p0; \
21145 __ret = (int32x2_t) __builtin_neon_vqshrn_n_v((int8x16_t)__s0, __p1, 2); \
21146 __ret; \
21147 })
21148 #else
21149 #define vqshrn_n_s64(__p0, __p1) __extension__ ({ \
21150 int32x2_t __ret; \
21151 int64x2_t __s0 = __p0; \
21152 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
21153 __ret = (int32x2_t) __builtin_neon_vqshrn_n_v((int8x16_t)__rev0, __p1, 2); \
21154 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
21155 __ret; \
21156 })
21157 #define __noswap_vqshrn_n_s64(__p0, __p1) __extension__ ({ \
21158 int32x2_t __ret; \
21159 int64x2_t __s0 = __p0; \
21160 __ret = (int32x2_t) __builtin_neon_vqshrn_n_v((int8x16_t)__s0, __p1, 2); \
21161 __ret; \
21162 })
21163 #endif
21164
21165 #ifdef __LITTLE_ENDIAN__
21166 #define vqshrn_n_s16(__p0, __p1) __extension__ ({ \
21167 int8x8_t __ret; \
21168 int16x8_t __s0 = __p0; \
21169 __ret = (int8x8_t) __builtin_neon_vqshrn_n_v((int8x16_t)__s0, __p1, 0); \
21170 __ret; \
21171 })
21172 #else
21173 #define vqshrn_n_s16(__p0, __p1) __extension__ ({ \
21174 int8x8_t __ret; \
21175 int16x8_t __s0 = __p0; \
21176 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
21177 __ret = (int8x8_t) __builtin_neon_vqshrn_n_v((int8x16_t)__rev0, __p1, 0); \
21178 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
21179 __ret; \
21180 })
21181 #define __noswap_vqshrn_n_s16(__p0, __p1) __extension__ ({ \
21182 int8x8_t __ret; \
21183 int16x8_t __s0 = __p0; \
21184 __ret = (int8x8_t) __builtin_neon_vqshrn_n_v((int8x16_t)__s0, __p1, 0); \
21185 __ret; \
21186 })
21187 #endif
21188
21189 #ifdef __LITTLE_ENDIAN__
21190 #define vqshrun_n_s32(__p0, __p1) __extension__ ({ \
21191 uint16x4_t __ret; \
21192 int32x4_t __s0 = __p0; \
21193 __ret = (uint16x4_t) __builtin_neon_vqshrun_n_v((int8x16_t)__s0, __p1, 17); \
21194 __ret; \
21195 })
21196 #else
21197 #define vqshrun_n_s32(__p0, __p1) __extension__ ({ \
21198 uint16x4_t __ret; \
21199 int32x4_t __s0 = __p0; \
21200 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
21201 __ret = (uint16x4_t) __builtin_neon_vqshrun_n_v((int8x16_t)__rev0, __p1, 17); \
21202 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
21203 __ret; \
21204 })
21205 #define __noswap_vqshrun_n_s32(__p0, __p1) __extension__ ({ \
21206 uint16x4_t __ret; \
21207 int32x4_t __s0 = __p0; \
21208 __ret = (uint16x4_t) __builtin_neon_vqshrun_n_v((int8x16_t)__s0, __p1, 17); \
21209 __ret; \
21210 })
21211 #endif
21212
21213 #ifdef __LITTLE_ENDIAN__
21214 #define vqshrun_n_s64(__p0, __p1) __extension__ ({ \
21215 uint32x2_t __ret; \
21216 int64x2_t __s0 = __p0; \
21217 __ret = (uint32x2_t) __builtin_neon_vqshrun_n_v((int8x16_t)__s0, __p1, 18); \
21218 __ret; \
21219 })
21220 #else
21221 #define vqshrun_n_s64(__p0, __p1) __extension__ ({ \
21222 uint32x2_t __ret; \
21223 int64x2_t __s0 = __p0; \
21224 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
21225 __ret = (uint32x2_t) __builtin_neon_vqshrun_n_v((int8x16_t)__rev0, __p1, 18); \
21226 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
21227 __ret; \
21228 })
21229 #define __noswap_vqshrun_n_s64(__p0, __p1) __extension__ ({ \
21230 uint32x2_t __ret; \
21231 int64x2_t __s0 = __p0; \
21232 __ret = (uint32x2_t) __builtin_neon_vqshrun_n_v((int8x16_t)__s0, __p1, 18); \
21233 __ret; \
21234 })
21235 #endif
21236
21237 #ifdef __LITTLE_ENDIAN__
21238 #define vqshrun_n_s16(__p0, __p1) __extension__ ({ \
21239 uint8x8_t __ret; \
21240 int16x8_t __s0 = __p0; \
21241 __ret = (uint8x8_t) __builtin_neon_vqshrun_n_v((int8x16_t)__s0, __p1, 16); \
21242 __ret; \
21243 })
21244 #else
21245 #define vqshrun_n_s16(__p0, __p1) __extension__ ({ \
21246 uint8x8_t __ret; \
21247 int16x8_t __s0 = __p0; \
21248 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
21249 __ret = (uint8x8_t) __builtin_neon_vqshrun_n_v((int8x16_t)__rev0, __p1, 16); \
21250 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
21251 __ret; \
21252 })
21253 #define __noswap_vqshrun_n_s16(__p0, __p1) __extension__ ({ \
21254 uint8x8_t __ret; \
21255 int16x8_t __s0 = __p0; \
21256 __ret = (uint8x8_t) __builtin_neon_vqshrun_n_v((int8x16_t)__s0, __p1, 16); \
21257 __ret; \
21258 })
21259 #endif
21260
21261 #ifdef __LITTLE_ENDIAN__
vqsubq_u8(uint8x16_t __p0,uint8x16_t __p1)21262 __ai uint8x16_t vqsubq_u8(uint8x16_t __p0, uint8x16_t __p1) {
21263 uint8x16_t __ret;
21264 __ret = (uint8x16_t) __builtin_neon_vqsubq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
21265 return __ret;
21266 }
21267 #else
vqsubq_u8(uint8x16_t __p0,uint8x16_t __p1)21268 __ai uint8x16_t vqsubq_u8(uint8x16_t __p0, uint8x16_t __p1) {
21269 uint8x16_t __ret;
21270 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
21271 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
21272 __ret = (uint8x16_t) __builtin_neon_vqsubq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
21273 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
21274 return __ret;
21275 }
21276 #endif
21277
21278 #ifdef __LITTLE_ENDIAN__
vqsubq_u32(uint32x4_t __p0,uint32x4_t __p1)21279 __ai uint32x4_t vqsubq_u32(uint32x4_t __p0, uint32x4_t __p1) {
21280 uint32x4_t __ret;
21281 __ret = (uint32x4_t) __builtin_neon_vqsubq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
21282 return __ret;
21283 }
21284 #else
vqsubq_u32(uint32x4_t __p0,uint32x4_t __p1)21285 __ai uint32x4_t vqsubq_u32(uint32x4_t __p0, uint32x4_t __p1) {
21286 uint32x4_t __ret;
21287 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
21288 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
21289 __ret = (uint32x4_t) __builtin_neon_vqsubq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
21290 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
21291 return __ret;
21292 }
21293 #endif
21294
21295 #ifdef __LITTLE_ENDIAN__
vqsubq_u64(uint64x2_t __p0,uint64x2_t __p1)21296 __ai uint64x2_t vqsubq_u64(uint64x2_t __p0, uint64x2_t __p1) {
21297 uint64x2_t __ret;
21298 __ret = (uint64x2_t) __builtin_neon_vqsubq_v((int8x16_t)__p0, (int8x16_t)__p1, 51);
21299 return __ret;
21300 }
21301 #else
vqsubq_u64(uint64x2_t __p0,uint64x2_t __p1)21302 __ai uint64x2_t vqsubq_u64(uint64x2_t __p0, uint64x2_t __p1) {
21303 uint64x2_t __ret;
21304 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
21305 uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
21306 __ret = (uint64x2_t) __builtin_neon_vqsubq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 51);
21307 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
21308 return __ret;
21309 }
21310 #endif
21311
21312 #ifdef __LITTLE_ENDIAN__
vqsubq_u16(uint16x8_t __p0,uint16x8_t __p1)21313 __ai uint16x8_t vqsubq_u16(uint16x8_t __p0, uint16x8_t __p1) {
21314 uint16x8_t __ret;
21315 __ret = (uint16x8_t) __builtin_neon_vqsubq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
21316 return __ret;
21317 }
21318 #else
vqsubq_u16(uint16x8_t __p0,uint16x8_t __p1)21319 __ai uint16x8_t vqsubq_u16(uint16x8_t __p0, uint16x8_t __p1) {
21320 uint16x8_t __ret;
21321 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
21322 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
21323 __ret = (uint16x8_t) __builtin_neon_vqsubq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
21324 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
21325 return __ret;
21326 }
21327 #endif
21328
21329 #ifdef __LITTLE_ENDIAN__
vqsubq_s8(int8x16_t __p0,int8x16_t __p1)21330 __ai int8x16_t vqsubq_s8(int8x16_t __p0, int8x16_t __p1) {
21331 int8x16_t __ret;
21332 __ret = (int8x16_t) __builtin_neon_vqsubq_v((int8x16_t)__p0, (int8x16_t)__p1, 32);
21333 return __ret;
21334 }
21335 #else
vqsubq_s8(int8x16_t __p0,int8x16_t __p1)21336 __ai int8x16_t vqsubq_s8(int8x16_t __p0, int8x16_t __p1) {
21337 int8x16_t __ret;
21338 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
21339 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
21340 __ret = (int8x16_t) __builtin_neon_vqsubq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 32);
21341 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
21342 return __ret;
21343 }
21344 #endif
21345
21346 #ifdef __LITTLE_ENDIAN__
vqsubq_s32(int32x4_t __p0,int32x4_t __p1)21347 __ai int32x4_t vqsubq_s32(int32x4_t __p0, int32x4_t __p1) {
21348 int32x4_t __ret;
21349 __ret = (int32x4_t) __builtin_neon_vqsubq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
21350 return __ret;
21351 }
21352 #else
vqsubq_s32(int32x4_t __p0,int32x4_t __p1)21353 __ai int32x4_t vqsubq_s32(int32x4_t __p0, int32x4_t __p1) {
21354 int32x4_t __ret;
21355 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
21356 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
21357 __ret = (int32x4_t) __builtin_neon_vqsubq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
21358 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
21359 return __ret;
21360 }
21361 #endif
21362
21363 #ifdef __LITTLE_ENDIAN__
vqsubq_s64(int64x2_t __p0,int64x2_t __p1)21364 __ai int64x2_t vqsubq_s64(int64x2_t __p0, int64x2_t __p1) {
21365 int64x2_t __ret;
21366 __ret = (int64x2_t) __builtin_neon_vqsubq_v((int8x16_t)__p0, (int8x16_t)__p1, 35);
21367 return __ret;
21368 }
21369 #else
vqsubq_s64(int64x2_t __p0,int64x2_t __p1)21370 __ai int64x2_t vqsubq_s64(int64x2_t __p0, int64x2_t __p1) {
21371 int64x2_t __ret;
21372 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
21373 int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
21374 __ret = (int64x2_t) __builtin_neon_vqsubq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 35);
21375 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
21376 return __ret;
21377 }
21378 #endif
21379
21380 #ifdef __LITTLE_ENDIAN__
vqsubq_s16(int16x8_t __p0,int16x8_t __p1)21381 __ai int16x8_t vqsubq_s16(int16x8_t __p0, int16x8_t __p1) {
21382 int16x8_t __ret;
21383 __ret = (int16x8_t) __builtin_neon_vqsubq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
21384 return __ret;
21385 }
21386 #else
vqsubq_s16(int16x8_t __p0,int16x8_t __p1)21387 __ai int16x8_t vqsubq_s16(int16x8_t __p0, int16x8_t __p1) {
21388 int16x8_t __ret;
21389 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
21390 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
21391 __ret = (int16x8_t) __builtin_neon_vqsubq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
21392 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
21393 return __ret;
21394 }
21395 #endif
21396
21397 #ifdef __LITTLE_ENDIAN__
vqsub_u8(uint8x8_t __p0,uint8x8_t __p1)21398 __ai uint8x8_t vqsub_u8(uint8x8_t __p0, uint8x8_t __p1) {
21399 uint8x8_t __ret;
21400 __ret = (uint8x8_t) __builtin_neon_vqsub_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
21401 return __ret;
21402 }
21403 #else
vqsub_u8(uint8x8_t __p0,uint8x8_t __p1)21404 __ai uint8x8_t vqsub_u8(uint8x8_t __p0, uint8x8_t __p1) {
21405 uint8x8_t __ret;
21406 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
21407 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
21408 __ret = (uint8x8_t) __builtin_neon_vqsub_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
21409 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
21410 return __ret;
21411 }
21412 #endif
21413
21414 #ifdef __LITTLE_ENDIAN__
vqsub_u32(uint32x2_t __p0,uint32x2_t __p1)21415 __ai uint32x2_t vqsub_u32(uint32x2_t __p0, uint32x2_t __p1) {
21416 uint32x2_t __ret;
21417 __ret = (uint32x2_t) __builtin_neon_vqsub_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
21418 return __ret;
21419 }
21420 #else
vqsub_u32(uint32x2_t __p0,uint32x2_t __p1)21421 __ai uint32x2_t vqsub_u32(uint32x2_t __p0, uint32x2_t __p1) {
21422 uint32x2_t __ret;
21423 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
21424 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
21425 __ret = (uint32x2_t) __builtin_neon_vqsub_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
21426 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
21427 return __ret;
21428 }
21429 #endif
21430
vqsub_u64(uint64x1_t __p0,uint64x1_t __p1)21431 __ai uint64x1_t vqsub_u64(uint64x1_t __p0, uint64x1_t __p1) {
21432 uint64x1_t __ret;
21433 __ret = (uint64x1_t) __builtin_neon_vqsub_v((int8x8_t)__p0, (int8x8_t)__p1, 19);
21434 return __ret;
21435 }
21436 #ifdef __LITTLE_ENDIAN__
vqsub_u16(uint16x4_t __p0,uint16x4_t __p1)21437 __ai uint16x4_t vqsub_u16(uint16x4_t __p0, uint16x4_t __p1) {
21438 uint16x4_t __ret;
21439 __ret = (uint16x4_t) __builtin_neon_vqsub_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
21440 return __ret;
21441 }
21442 #else
vqsub_u16(uint16x4_t __p0,uint16x4_t __p1)21443 __ai uint16x4_t vqsub_u16(uint16x4_t __p0, uint16x4_t __p1) {
21444 uint16x4_t __ret;
21445 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
21446 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
21447 __ret = (uint16x4_t) __builtin_neon_vqsub_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
21448 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
21449 return __ret;
21450 }
21451 #endif
21452
21453 #ifdef __LITTLE_ENDIAN__
vqsub_s8(int8x8_t __p0,int8x8_t __p1)21454 __ai int8x8_t vqsub_s8(int8x8_t __p0, int8x8_t __p1) {
21455 int8x8_t __ret;
21456 __ret = (int8x8_t) __builtin_neon_vqsub_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
21457 return __ret;
21458 }
21459 #else
vqsub_s8(int8x8_t __p0,int8x8_t __p1)21460 __ai int8x8_t vqsub_s8(int8x8_t __p0, int8x8_t __p1) {
21461 int8x8_t __ret;
21462 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
21463 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
21464 __ret = (int8x8_t) __builtin_neon_vqsub_v((int8x8_t)__rev0, (int8x8_t)__rev1, 0);
21465 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
21466 return __ret;
21467 }
21468 #endif
21469
21470 #ifdef __LITTLE_ENDIAN__
vqsub_s32(int32x2_t __p0,int32x2_t __p1)21471 __ai int32x2_t vqsub_s32(int32x2_t __p0, int32x2_t __p1) {
21472 int32x2_t __ret;
21473 __ret = (int32x2_t) __builtin_neon_vqsub_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
21474 return __ret;
21475 }
21476 #else
vqsub_s32(int32x2_t __p0,int32x2_t __p1)21477 __ai int32x2_t vqsub_s32(int32x2_t __p0, int32x2_t __p1) {
21478 int32x2_t __ret;
21479 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
21480 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
21481 __ret = (int32x2_t) __builtin_neon_vqsub_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
21482 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
21483 return __ret;
21484 }
21485 #endif
21486
vqsub_s64(int64x1_t __p0,int64x1_t __p1)21487 __ai int64x1_t vqsub_s64(int64x1_t __p0, int64x1_t __p1) {
21488 int64x1_t __ret;
21489 __ret = (int64x1_t) __builtin_neon_vqsub_v((int8x8_t)__p0, (int8x8_t)__p1, 3);
21490 return __ret;
21491 }
21492 #ifdef __LITTLE_ENDIAN__
vqsub_s16(int16x4_t __p0,int16x4_t __p1)21493 __ai int16x4_t vqsub_s16(int16x4_t __p0, int16x4_t __p1) {
21494 int16x4_t __ret;
21495 __ret = (int16x4_t) __builtin_neon_vqsub_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
21496 return __ret;
21497 }
21498 #else
vqsub_s16(int16x4_t __p0,int16x4_t __p1)21499 __ai int16x4_t vqsub_s16(int16x4_t __p0, int16x4_t __p1) {
21500 int16x4_t __ret;
21501 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
21502 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
21503 __ret = (int16x4_t) __builtin_neon_vqsub_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
21504 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
21505 return __ret;
21506 }
21507 #endif
21508
21509 #ifdef __LITTLE_ENDIAN__
vraddhn_u32(uint32x4_t __p0,uint32x4_t __p1)21510 __ai uint16x4_t vraddhn_u32(uint32x4_t __p0, uint32x4_t __p1) {
21511 uint16x4_t __ret;
21512 __ret = (uint16x4_t) __builtin_neon_vraddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 17);
21513 return __ret;
21514 }
21515 #else
vraddhn_u32(uint32x4_t __p0,uint32x4_t __p1)21516 __ai uint16x4_t vraddhn_u32(uint32x4_t __p0, uint32x4_t __p1) {
21517 uint16x4_t __ret;
21518 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
21519 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
21520 __ret = (uint16x4_t) __builtin_neon_vraddhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 17);
21521 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
21522 return __ret;
21523 }
__noswap_vraddhn_u32(uint32x4_t __p0,uint32x4_t __p1)21524 __ai uint16x4_t __noswap_vraddhn_u32(uint32x4_t __p0, uint32x4_t __p1) {
21525 uint16x4_t __ret;
21526 __ret = (uint16x4_t) __builtin_neon_vraddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 17);
21527 return __ret;
21528 }
21529 #endif
21530
21531 #ifdef __LITTLE_ENDIAN__
vraddhn_u64(uint64x2_t __p0,uint64x2_t __p1)21532 __ai uint32x2_t vraddhn_u64(uint64x2_t __p0, uint64x2_t __p1) {
21533 uint32x2_t __ret;
21534 __ret = (uint32x2_t) __builtin_neon_vraddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 18);
21535 return __ret;
21536 }
21537 #else
vraddhn_u64(uint64x2_t __p0,uint64x2_t __p1)21538 __ai uint32x2_t vraddhn_u64(uint64x2_t __p0, uint64x2_t __p1) {
21539 uint32x2_t __ret;
21540 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
21541 uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
21542 __ret = (uint32x2_t) __builtin_neon_vraddhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 18);
21543 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
21544 return __ret;
21545 }
__noswap_vraddhn_u64(uint64x2_t __p0,uint64x2_t __p1)21546 __ai uint32x2_t __noswap_vraddhn_u64(uint64x2_t __p0, uint64x2_t __p1) {
21547 uint32x2_t __ret;
21548 __ret = (uint32x2_t) __builtin_neon_vraddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 18);
21549 return __ret;
21550 }
21551 #endif
21552
21553 #ifdef __LITTLE_ENDIAN__
vraddhn_u16(uint16x8_t __p0,uint16x8_t __p1)21554 __ai uint8x8_t vraddhn_u16(uint16x8_t __p0, uint16x8_t __p1) {
21555 uint8x8_t __ret;
21556 __ret = (uint8x8_t) __builtin_neon_vraddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 16);
21557 return __ret;
21558 }
21559 #else
vraddhn_u16(uint16x8_t __p0,uint16x8_t __p1)21560 __ai uint8x8_t vraddhn_u16(uint16x8_t __p0, uint16x8_t __p1) {
21561 uint8x8_t __ret;
21562 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
21563 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
21564 __ret = (uint8x8_t) __builtin_neon_vraddhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 16);
21565 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
21566 return __ret;
21567 }
__noswap_vraddhn_u16(uint16x8_t __p0,uint16x8_t __p1)21568 __ai uint8x8_t __noswap_vraddhn_u16(uint16x8_t __p0, uint16x8_t __p1) {
21569 uint8x8_t __ret;
21570 __ret = (uint8x8_t) __builtin_neon_vraddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 16);
21571 return __ret;
21572 }
21573 #endif
21574
21575 #ifdef __LITTLE_ENDIAN__
vraddhn_s32(int32x4_t __p0,int32x4_t __p1)21576 __ai int16x4_t vraddhn_s32(int32x4_t __p0, int32x4_t __p1) {
21577 int16x4_t __ret;
21578 __ret = (int16x4_t) __builtin_neon_vraddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 1);
21579 return __ret;
21580 }
21581 #else
vraddhn_s32(int32x4_t __p0,int32x4_t __p1)21582 __ai int16x4_t vraddhn_s32(int32x4_t __p0, int32x4_t __p1) {
21583 int16x4_t __ret;
21584 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
21585 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
21586 __ret = (int16x4_t) __builtin_neon_vraddhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 1);
21587 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
21588 return __ret;
21589 }
__noswap_vraddhn_s32(int32x4_t __p0,int32x4_t __p1)21590 __ai int16x4_t __noswap_vraddhn_s32(int32x4_t __p0, int32x4_t __p1) {
21591 int16x4_t __ret;
21592 __ret = (int16x4_t) __builtin_neon_vraddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 1);
21593 return __ret;
21594 }
21595 #endif
21596
21597 #ifdef __LITTLE_ENDIAN__
vraddhn_s64(int64x2_t __p0,int64x2_t __p1)21598 __ai int32x2_t vraddhn_s64(int64x2_t __p0, int64x2_t __p1) {
21599 int32x2_t __ret;
21600 __ret = (int32x2_t) __builtin_neon_vraddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 2);
21601 return __ret;
21602 }
21603 #else
vraddhn_s64(int64x2_t __p0,int64x2_t __p1)21604 __ai int32x2_t vraddhn_s64(int64x2_t __p0, int64x2_t __p1) {
21605 int32x2_t __ret;
21606 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
21607 int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
21608 __ret = (int32x2_t) __builtin_neon_vraddhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 2);
21609 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
21610 return __ret;
21611 }
__noswap_vraddhn_s64(int64x2_t __p0,int64x2_t __p1)21612 __ai int32x2_t __noswap_vraddhn_s64(int64x2_t __p0, int64x2_t __p1) {
21613 int32x2_t __ret;
21614 __ret = (int32x2_t) __builtin_neon_vraddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 2);
21615 return __ret;
21616 }
21617 #endif
21618
21619 #ifdef __LITTLE_ENDIAN__
vraddhn_s16(int16x8_t __p0,int16x8_t __p1)21620 __ai int8x8_t vraddhn_s16(int16x8_t __p0, int16x8_t __p1) {
21621 int8x8_t __ret;
21622 __ret = (int8x8_t) __builtin_neon_vraddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 0);
21623 return __ret;
21624 }
21625 #else
vraddhn_s16(int16x8_t __p0,int16x8_t __p1)21626 __ai int8x8_t vraddhn_s16(int16x8_t __p0, int16x8_t __p1) {
21627 int8x8_t __ret;
21628 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
21629 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
21630 __ret = (int8x8_t) __builtin_neon_vraddhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 0);
21631 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
21632 return __ret;
21633 }
__noswap_vraddhn_s16(int16x8_t __p0,int16x8_t __p1)21634 __ai int8x8_t __noswap_vraddhn_s16(int16x8_t __p0, int16x8_t __p1) {
21635 int8x8_t __ret;
21636 __ret = (int8x8_t) __builtin_neon_vraddhn_v((int8x16_t)__p0, (int8x16_t)__p1, 0);
21637 return __ret;
21638 }
21639 #endif
21640
21641 #ifdef __LITTLE_ENDIAN__
vrecpeq_u32(uint32x4_t __p0)21642 __ai uint32x4_t vrecpeq_u32(uint32x4_t __p0) {
21643 uint32x4_t __ret;
21644 __ret = (uint32x4_t) __builtin_neon_vrecpeq_v((int8x16_t)__p0, 50);
21645 return __ret;
21646 }
21647 #else
vrecpeq_u32(uint32x4_t __p0)21648 __ai uint32x4_t vrecpeq_u32(uint32x4_t __p0) {
21649 uint32x4_t __ret;
21650 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
21651 __ret = (uint32x4_t) __builtin_neon_vrecpeq_v((int8x16_t)__rev0, 50);
21652 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
21653 return __ret;
21654 }
21655 #endif
21656
21657 #ifdef __LITTLE_ENDIAN__
vrecpeq_f32(float32x4_t __p0)21658 __ai float32x4_t vrecpeq_f32(float32x4_t __p0) {
21659 float32x4_t __ret;
21660 __ret = (float32x4_t) __builtin_neon_vrecpeq_v((int8x16_t)__p0, 41);
21661 return __ret;
21662 }
21663 #else
vrecpeq_f32(float32x4_t __p0)21664 __ai float32x4_t vrecpeq_f32(float32x4_t __p0) {
21665 float32x4_t __ret;
21666 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
21667 __ret = (float32x4_t) __builtin_neon_vrecpeq_v((int8x16_t)__rev0, 41);
21668 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
21669 return __ret;
21670 }
21671 #endif
21672
21673 #ifdef __LITTLE_ENDIAN__
vrecpe_u32(uint32x2_t __p0)21674 __ai uint32x2_t vrecpe_u32(uint32x2_t __p0) {
21675 uint32x2_t __ret;
21676 __ret = (uint32x2_t) __builtin_neon_vrecpe_v((int8x8_t)__p0, 18);
21677 return __ret;
21678 }
21679 #else
vrecpe_u32(uint32x2_t __p0)21680 __ai uint32x2_t vrecpe_u32(uint32x2_t __p0) {
21681 uint32x2_t __ret;
21682 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
21683 __ret = (uint32x2_t) __builtin_neon_vrecpe_v((int8x8_t)__rev0, 18);
21684 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
21685 return __ret;
21686 }
21687 #endif
21688
21689 #ifdef __LITTLE_ENDIAN__
vrecpe_f32(float32x2_t __p0)21690 __ai float32x2_t vrecpe_f32(float32x2_t __p0) {
21691 float32x2_t __ret;
21692 __ret = (float32x2_t) __builtin_neon_vrecpe_v((int8x8_t)__p0, 9);
21693 return __ret;
21694 }
21695 #else
vrecpe_f32(float32x2_t __p0)21696 __ai float32x2_t vrecpe_f32(float32x2_t __p0) {
21697 float32x2_t __ret;
21698 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
21699 __ret = (float32x2_t) __builtin_neon_vrecpe_v((int8x8_t)__rev0, 9);
21700 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
21701 return __ret;
21702 }
21703 #endif
21704
21705 #ifdef __LITTLE_ENDIAN__
vrecpsq_f32(float32x4_t __p0,float32x4_t __p1)21706 __ai float32x4_t vrecpsq_f32(float32x4_t __p0, float32x4_t __p1) {
21707 float32x4_t __ret;
21708 __ret = (float32x4_t) __builtin_neon_vrecpsq_v((int8x16_t)__p0, (int8x16_t)__p1, 41);
21709 return __ret;
21710 }
21711 #else
vrecpsq_f32(float32x4_t __p0,float32x4_t __p1)21712 __ai float32x4_t vrecpsq_f32(float32x4_t __p0, float32x4_t __p1) {
21713 float32x4_t __ret;
21714 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
21715 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
21716 __ret = (float32x4_t) __builtin_neon_vrecpsq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 41);
21717 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
21718 return __ret;
21719 }
21720 #endif
21721
21722 #ifdef __LITTLE_ENDIAN__
vrecps_f32(float32x2_t __p0,float32x2_t __p1)21723 __ai float32x2_t vrecps_f32(float32x2_t __p0, float32x2_t __p1) {
21724 float32x2_t __ret;
21725 __ret = (float32x2_t) __builtin_neon_vrecps_v((int8x8_t)__p0, (int8x8_t)__p1, 9);
21726 return __ret;
21727 }
21728 #else
vrecps_f32(float32x2_t __p0,float32x2_t __p1)21729 __ai float32x2_t vrecps_f32(float32x2_t __p0, float32x2_t __p1) {
21730 float32x2_t __ret;
21731 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
21732 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
21733 __ret = (float32x2_t) __builtin_neon_vrecps_v((int8x8_t)__rev0, (int8x8_t)__rev1, 9);
21734 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
21735 return __ret;
21736 }
21737 #endif
21738
21739 #ifdef __LITTLE_ENDIAN__
vrev16_p8(poly8x8_t __p0)21740 __ai poly8x8_t vrev16_p8(poly8x8_t __p0) {
21741 poly8x8_t __ret;
21742 __ret = __builtin_shufflevector(__p0, __p0, 1, 0, 3, 2, 5, 4, 7, 6);
21743 return __ret;
21744 }
21745 #else
vrev16_p8(poly8x8_t __p0)21746 __ai poly8x8_t vrev16_p8(poly8x8_t __p0) {
21747 poly8x8_t __ret;
21748 poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
21749 __ret = __builtin_shufflevector(__rev0, __rev0, 1, 0, 3, 2, 5, 4, 7, 6);
21750 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
21751 return __ret;
21752 }
21753 #endif
21754
21755 #ifdef __LITTLE_ENDIAN__
vrev16q_p8(poly8x16_t __p0)21756 __ai poly8x16_t vrev16q_p8(poly8x16_t __p0) {
21757 poly8x16_t __ret;
21758 __ret = __builtin_shufflevector(__p0, __p0, 1, 0, 3, 2, 5, 4, 7, 6, 9, 8, 11, 10, 13, 12, 15, 14);
21759 return __ret;
21760 }
21761 #else
vrev16q_p8(poly8x16_t __p0)21762 __ai poly8x16_t vrev16q_p8(poly8x16_t __p0) {
21763 poly8x16_t __ret;
21764 poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
21765 __ret = __builtin_shufflevector(__rev0, __rev0, 1, 0, 3, 2, 5, 4, 7, 6, 9, 8, 11, 10, 13, 12, 15, 14);
21766 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
21767 return __ret;
21768 }
21769 #endif
21770
21771 #ifdef __LITTLE_ENDIAN__
vrev16q_u8(uint8x16_t __p0)21772 __ai uint8x16_t vrev16q_u8(uint8x16_t __p0) {
21773 uint8x16_t __ret;
21774 __ret = __builtin_shufflevector(__p0, __p0, 1, 0, 3, 2, 5, 4, 7, 6, 9, 8, 11, 10, 13, 12, 15, 14);
21775 return __ret;
21776 }
21777 #else
vrev16q_u8(uint8x16_t __p0)21778 __ai uint8x16_t vrev16q_u8(uint8x16_t __p0) {
21779 uint8x16_t __ret;
21780 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
21781 __ret = __builtin_shufflevector(__rev0, __rev0, 1, 0, 3, 2, 5, 4, 7, 6, 9, 8, 11, 10, 13, 12, 15, 14);
21782 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
21783 return __ret;
21784 }
21785 #endif
21786
21787 #ifdef __LITTLE_ENDIAN__
vrev16q_s8(int8x16_t __p0)21788 __ai int8x16_t vrev16q_s8(int8x16_t __p0) {
21789 int8x16_t __ret;
21790 __ret = __builtin_shufflevector(__p0, __p0, 1, 0, 3, 2, 5, 4, 7, 6, 9, 8, 11, 10, 13, 12, 15, 14);
21791 return __ret;
21792 }
21793 #else
vrev16q_s8(int8x16_t __p0)21794 __ai int8x16_t vrev16q_s8(int8x16_t __p0) {
21795 int8x16_t __ret;
21796 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
21797 __ret = __builtin_shufflevector(__rev0, __rev0, 1, 0, 3, 2, 5, 4, 7, 6, 9, 8, 11, 10, 13, 12, 15, 14);
21798 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
21799 return __ret;
21800 }
21801 #endif
21802
21803 #ifdef __LITTLE_ENDIAN__
vrev16_u8(uint8x8_t __p0)21804 __ai uint8x8_t vrev16_u8(uint8x8_t __p0) {
21805 uint8x8_t __ret;
21806 __ret = __builtin_shufflevector(__p0, __p0, 1, 0, 3, 2, 5, 4, 7, 6);
21807 return __ret;
21808 }
21809 #else
vrev16_u8(uint8x8_t __p0)21810 __ai uint8x8_t vrev16_u8(uint8x8_t __p0) {
21811 uint8x8_t __ret;
21812 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
21813 __ret = __builtin_shufflevector(__rev0, __rev0, 1, 0, 3, 2, 5, 4, 7, 6);
21814 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
21815 return __ret;
21816 }
21817 #endif
21818
21819 #ifdef __LITTLE_ENDIAN__
vrev16_s8(int8x8_t __p0)21820 __ai int8x8_t vrev16_s8(int8x8_t __p0) {
21821 int8x8_t __ret;
21822 __ret = __builtin_shufflevector(__p0, __p0, 1, 0, 3, 2, 5, 4, 7, 6);
21823 return __ret;
21824 }
21825 #else
vrev16_s8(int8x8_t __p0)21826 __ai int8x8_t vrev16_s8(int8x8_t __p0) {
21827 int8x8_t __ret;
21828 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
21829 __ret = __builtin_shufflevector(__rev0, __rev0, 1, 0, 3, 2, 5, 4, 7, 6);
21830 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
21831 return __ret;
21832 }
21833 #endif
21834
21835 #ifdef __LITTLE_ENDIAN__
vrev32_p8(poly8x8_t __p0)21836 __ai poly8x8_t vrev32_p8(poly8x8_t __p0) {
21837 poly8x8_t __ret;
21838 __ret = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0, 7, 6, 5, 4);
21839 return __ret;
21840 }
21841 #else
vrev32_p8(poly8x8_t __p0)21842 __ai poly8x8_t vrev32_p8(poly8x8_t __p0) {
21843 poly8x8_t __ret;
21844 poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
21845 __ret = __builtin_shufflevector(__rev0, __rev0, 3, 2, 1, 0, 7, 6, 5, 4);
21846 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
21847 return __ret;
21848 }
21849 #endif
21850
21851 #ifdef __LITTLE_ENDIAN__
vrev32_p16(poly16x4_t __p0)21852 __ai poly16x4_t vrev32_p16(poly16x4_t __p0) {
21853 poly16x4_t __ret;
21854 __ret = __builtin_shufflevector(__p0, __p0, 1, 0, 3, 2);
21855 return __ret;
21856 }
21857 #else
vrev32_p16(poly16x4_t __p0)21858 __ai poly16x4_t vrev32_p16(poly16x4_t __p0) {
21859 poly16x4_t __ret;
21860 poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
21861 __ret = __builtin_shufflevector(__rev0, __rev0, 1, 0, 3, 2);
21862 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
21863 return __ret;
21864 }
21865 #endif
21866
21867 #ifdef __LITTLE_ENDIAN__
vrev32q_p8(poly8x16_t __p0)21868 __ai poly8x16_t vrev32q_p8(poly8x16_t __p0) {
21869 poly8x16_t __ret;
21870 __ret = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0, 7, 6, 5, 4, 11, 10, 9, 8, 15, 14, 13, 12);
21871 return __ret;
21872 }
21873 #else
vrev32q_p8(poly8x16_t __p0)21874 __ai poly8x16_t vrev32q_p8(poly8x16_t __p0) {
21875 poly8x16_t __ret;
21876 poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
21877 __ret = __builtin_shufflevector(__rev0, __rev0, 3, 2, 1, 0, 7, 6, 5, 4, 11, 10, 9, 8, 15, 14, 13, 12);
21878 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
21879 return __ret;
21880 }
21881 #endif
21882
21883 #ifdef __LITTLE_ENDIAN__
vrev32q_p16(poly16x8_t __p0)21884 __ai poly16x8_t vrev32q_p16(poly16x8_t __p0) {
21885 poly16x8_t __ret;
21886 __ret = __builtin_shufflevector(__p0, __p0, 1, 0, 3, 2, 5, 4, 7, 6);
21887 return __ret;
21888 }
21889 #else
vrev32q_p16(poly16x8_t __p0)21890 __ai poly16x8_t vrev32q_p16(poly16x8_t __p0) {
21891 poly16x8_t __ret;
21892 poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
21893 __ret = __builtin_shufflevector(__rev0, __rev0, 1, 0, 3, 2, 5, 4, 7, 6);
21894 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
21895 return __ret;
21896 }
21897 #endif
21898
21899 #ifdef __LITTLE_ENDIAN__
vrev32q_u8(uint8x16_t __p0)21900 __ai uint8x16_t vrev32q_u8(uint8x16_t __p0) {
21901 uint8x16_t __ret;
21902 __ret = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0, 7, 6, 5, 4, 11, 10, 9, 8, 15, 14, 13, 12);
21903 return __ret;
21904 }
21905 #else
vrev32q_u8(uint8x16_t __p0)21906 __ai uint8x16_t vrev32q_u8(uint8x16_t __p0) {
21907 uint8x16_t __ret;
21908 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
21909 __ret = __builtin_shufflevector(__rev0, __rev0, 3, 2, 1, 0, 7, 6, 5, 4, 11, 10, 9, 8, 15, 14, 13, 12);
21910 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
21911 return __ret;
21912 }
21913 #endif
21914
21915 #ifdef __LITTLE_ENDIAN__
vrev32q_u16(uint16x8_t __p0)21916 __ai uint16x8_t vrev32q_u16(uint16x8_t __p0) {
21917 uint16x8_t __ret;
21918 __ret = __builtin_shufflevector(__p0, __p0, 1, 0, 3, 2, 5, 4, 7, 6);
21919 return __ret;
21920 }
21921 #else
vrev32q_u16(uint16x8_t __p0)21922 __ai uint16x8_t vrev32q_u16(uint16x8_t __p0) {
21923 uint16x8_t __ret;
21924 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
21925 __ret = __builtin_shufflevector(__rev0, __rev0, 1, 0, 3, 2, 5, 4, 7, 6);
21926 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
21927 return __ret;
21928 }
21929 #endif
21930
21931 #ifdef __LITTLE_ENDIAN__
vrev32q_s8(int8x16_t __p0)21932 __ai int8x16_t vrev32q_s8(int8x16_t __p0) {
21933 int8x16_t __ret;
21934 __ret = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0, 7, 6, 5, 4, 11, 10, 9, 8, 15, 14, 13, 12);
21935 return __ret;
21936 }
21937 #else
vrev32q_s8(int8x16_t __p0)21938 __ai int8x16_t vrev32q_s8(int8x16_t __p0) {
21939 int8x16_t __ret;
21940 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
21941 __ret = __builtin_shufflevector(__rev0, __rev0, 3, 2, 1, 0, 7, 6, 5, 4, 11, 10, 9, 8, 15, 14, 13, 12);
21942 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
21943 return __ret;
21944 }
21945 #endif
21946
21947 #ifdef __LITTLE_ENDIAN__
vrev32q_s16(int16x8_t __p0)21948 __ai int16x8_t vrev32q_s16(int16x8_t __p0) {
21949 int16x8_t __ret;
21950 __ret = __builtin_shufflevector(__p0, __p0, 1, 0, 3, 2, 5, 4, 7, 6);
21951 return __ret;
21952 }
21953 #else
vrev32q_s16(int16x8_t __p0)21954 __ai int16x8_t vrev32q_s16(int16x8_t __p0) {
21955 int16x8_t __ret;
21956 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
21957 __ret = __builtin_shufflevector(__rev0, __rev0, 1, 0, 3, 2, 5, 4, 7, 6);
21958 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
21959 return __ret;
21960 }
21961 #endif
21962
21963 #ifdef __LITTLE_ENDIAN__
vrev32_u8(uint8x8_t __p0)21964 __ai uint8x8_t vrev32_u8(uint8x8_t __p0) {
21965 uint8x8_t __ret;
21966 __ret = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0, 7, 6, 5, 4);
21967 return __ret;
21968 }
21969 #else
vrev32_u8(uint8x8_t __p0)21970 __ai uint8x8_t vrev32_u8(uint8x8_t __p0) {
21971 uint8x8_t __ret;
21972 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
21973 __ret = __builtin_shufflevector(__rev0, __rev0, 3, 2, 1, 0, 7, 6, 5, 4);
21974 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
21975 return __ret;
21976 }
21977 #endif
21978
21979 #ifdef __LITTLE_ENDIAN__
vrev32_u16(uint16x4_t __p0)21980 __ai uint16x4_t vrev32_u16(uint16x4_t __p0) {
21981 uint16x4_t __ret;
21982 __ret = __builtin_shufflevector(__p0, __p0, 1, 0, 3, 2);
21983 return __ret;
21984 }
21985 #else
vrev32_u16(uint16x4_t __p0)21986 __ai uint16x4_t vrev32_u16(uint16x4_t __p0) {
21987 uint16x4_t __ret;
21988 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
21989 __ret = __builtin_shufflevector(__rev0, __rev0, 1, 0, 3, 2);
21990 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
21991 return __ret;
21992 }
21993 #endif
21994
21995 #ifdef __LITTLE_ENDIAN__
vrev32_s8(int8x8_t __p0)21996 __ai int8x8_t vrev32_s8(int8x8_t __p0) {
21997 int8x8_t __ret;
21998 __ret = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0, 7, 6, 5, 4);
21999 return __ret;
22000 }
22001 #else
vrev32_s8(int8x8_t __p0)22002 __ai int8x8_t vrev32_s8(int8x8_t __p0) {
22003 int8x8_t __ret;
22004 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
22005 __ret = __builtin_shufflevector(__rev0, __rev0, 3, 2, 1, 0, 7, 6, 5, 4);
22006 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
22007 return __ret;
22008 }
22009 #endif
22010
22011 #ifdef __LITTLE_ENDIAN__
vrev32_s16(int16x4_t __p0)22012 __ai int16x4_t vrev32_s16(int16x4_t __p0) {
22013 int16x4_t __ret;
22014 __ret = __builtin_shufflevector(__p0, __p0, 1, 0, 3, 2);
22015 return __ret;
22016 }
22017 #else
vrev32_s16(int16x4_t __p0)22018 __ai int16x4_t vrev32_s16(int16x4_t __p0) {
22019 int16x4_t __ret;
22020 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
22021 __ret = __builtin_shufflevector(__rev0, __rev0, 1, 0, 3, 2);
22022 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
22023 return __ret;
22024 }
22025 #endif
22026
22027 #ifdef __LITTLE_ENDIAN__
vrev64_p8(poly8x8_t __p0)22028 __ai poly8x8_t vrev64_p8(poly8x8_t __p0) {
22029 poly8x8_t __ret;
22030 __ret = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
22031 return __ret;
22032 }
22033 #else
vrev64_p8(poly8x8_t __p0)22034 __ai poly8x8_t vrev64_p8(poly8x8_t __p0) {
22035 poly8x8_t __ret;
22036 poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
22037 __ret = __builtin_shufflevector(__rev0, __rev0, 7, 6, 5, 4, 3, 2, 1, 0);
22038 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
22039 return __ret;
22040 }
22041 #endif
22042
22043 #ifdef __LITTLE_ENDIAN__
vrev64_p16(poly16x4_t __p0)22044 __ai poly16x4_t vrev64_p16(poly16x4_t __p0) {
22045 poly16x4_t __ret;
22046 __ret = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
22047 return __ret;
22048 }
22049 #else
vrev64_p16(poly16x4_t __p0)22050 __ai poly16x4_t vrev64_p16(poly16x4_t __p0) {
22051 poly16x4_t __ret;
22052 poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
22053 __ret = __builtin_shufflevector(__rev0, __rev0, 3, 2, 1, 0);
22054 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
22055 return __ret;
22056 }
22057 #endif
22058
22059 #ifdef __LITTLE_ENDIAN__
vrev64q_p8(poly8x16_t __p0)22060 __ai poly8x16_t vrev64q_p8(poly8x16_t __p0) {
22061 poly8x16_t __ret;
22062 __ret = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0, 15, 14, 13, 12, 11, 10, 9, 8);
22063 return __ret;
22064 }
22065 #else
vrev64q_p8(poly8x16_t __p0)22066 __ai poly8x16_t vrev64q_p8(poly8x16_t __p0) {
22067 poly8x16_t __ret;
22068 poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
22069 __ret = __builtin_shufflevector(__rev0, __rev0, 7, 6, 5, 4, 3, 2, 1, 0, 15, 14, 13, 12, 11, 10, 9, 8);
22070 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
22071 return __ret;
22072 }
22073 #endif
22074
22075 #ifdef __LITTLE_ENDIAN__
vrev64q_p16(poly16x8_t __p0)22076 __ai poly16x8_t vrev64q_p16(poly16x8_t __p0) {
22077 poly16x8_t __ret;
22078 __ret = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0, 7, 6, 5, 4);
22079 return __ret;
22080 }
22081 #else
vrev64q_p16(poly16x8_t __p0)22082 __ai poly16x8_t vrev64q_p16(poly16x8_t __p0) {
22083 poly16x8_t __ret;
22084 poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
22085 __ret = __builtin_shufflevector(__rev0, __rev0, 3, 2, 1, 0, 7, 6, 5, 4);
22086 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
22087 return __ret;
22088 }
22089 #endif
22090
22091 #ifdef __LITTLE_ENDIAN__
vrev64q_u8(uint8x16_t __p0)22092 __ai uint8x16_t vrev64q_u8(uint8x16_t __p0) {
22093 uint8x16_t __ret;
22094 __ret = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0, 15, 14, 13, 12, 11, 10, 9, 8);
22095 return __ret;
22096 }
22097 #else
vrev64q_u8(uint8x16_t __p0)22098 __ai uint8x16_t vrev64q_u8(uint8x16_t __p0) {
22099 uint8x16_t __ret;
22100 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
22101 __ret = __builtin_shufflevector(__rev0, __rev0, 7, 6, 5, 4, 3, 2, 1, 0, 15, 14, 13, 12, 11, 10, 9, 8);
22102 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
22103 return __ret;
22104 }
22105 #endif
22106
22107 #ifdef __LITTLE_ENDIAN__
vrev64q_u32(uint32x4_t __p0)22108 __ai uint32x4_t vrev64q_u32(uint32x4_t __p0) {
22109 uint32x4_t __ret;
22110 __ret = __builtin_shufflevector(__p0, __p0, 1, 0, 3, 2);
22111 return __ret;
22112 }
22113 #else
vrev64q_u32(uint32x4_t __p0)22114 __ai uint32x4_t vrev64q_u32(uint32x4_t __p0) {
22115 uint32x4_t __ret;
22116 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
22117 __ret = __builtin_shufflevector(__rev0, __rev0, 1, 0, 3, 2);
22118 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
22119 return __ret;
22120 }
22121 #endif
22122
22123 #ifdef __LITTLE_ENDIAN__
vrev64q_u16(uint16x8_t __p0)22124 __ai uint16x8_t vrev64q_u16(uint16x8_t __p0) {
22125 uint16x8_t __ret;
22126 __ret = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0, 7, 6, 5, 4);
22127 return __ret;
22128 }
22129 #else
vrev64q_u16(uint16x8_t __p0)22130 __ai uint16x8_t vrev64q_u16(uint16x8_t __p0) {
22131 uint16x8_t __ret;
22132 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
22133 __ret = __builtin_shufflevector(__rev0, __rev0, 3, 2, 1, 0, 7, 6, 5, 4);
22134 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
22135 return __ret;
22136 }
22137 #endif
22138
22139 #ifdef __LITTLE_ENDIAN__
vrev64q_s8(int8x16_t __p0)22140 __ai int8x16_t vrev64q_s8(int8x16_t __p0) {
22141 int8x16_t __ret;
22142 __ret = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0, 15, 14, 13, 12, 11, 10, 9, 8);
22143 return __ret;
22144 }
22145 #else
vrev64q_s8(int8x16_t __p0)22146 __ai int8x16_t vrev64q_s8(int8x16_t __p0) {
22147 int8x16_t __ret;
22148 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
22149 __ret = __builtin_shufflevector(__rev0, __rev0, 7, 6, 5, 4, 3, 2, 1, 0, 15, 14, 13, 12, 11, 10, 9, 8);
22150 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
22151 return __ret;
22152 }
22153 #endif
22154
22155 #ifdef __LITTLE_ENDIAN__
vrev64q_f32(float32x4_t __p0)22156 __ai float32x4_t vrev64q_f32(float32x4_t __p0) {
22157 float32x4_t __ret;
22158 __ret = __builtin_shufflevector(__p0, __p0, 1, 0, 3, 2);
22159 return __ret;
22160 }
22161 #else
vrev64q_f32(float32x4_t __p0)22162 __ai float32x4_t vrev64q_f32(float32x4_t __p0) {
22163 float32x4_t __ret;
22164 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
22165 __ret = __builtin_shufflevector(__rev0, __rev0, 1, 0, 3, 2);
22166 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
22167 return __ret;
22168 }
22169 #endif
22170
22171 #ifdef __LITTLE_ENDIAN__
vrev64q_s32(int32x4_t __p0)22172 __ai int32x4_t vrev64q_s32(int32x4_t __p0) {
22173 int32x4_t __ret;
22174 __ret = __builtin_shufflevector(__p0, __p0, 1, 0, 3, 2);
22175 return __ret;
22176 }
22177 #else
vrev64q_s32(int32x4_t __p0)22178 __ai int32x4_t vrev64q_s32(int32x4_t __p0) {
22179 int32x4_t __ret;
22180 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
22181 __ret = __builtin_shufflevector(__rev0, __rev0, 1, 0, 3, 2);
22182 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
22183 return __ret;
22184 }
22185 #endif
22186
22187 #ifdef __LITTLE_ENDIAN__
vrev64q_s16(int16x8_t __p0)22188 __ai int16x8_t vrev64q_s16(int16x8_t __p0) {
22189 int16x8_t __ret;
22190 __ret = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0, 7, 6, 5, 4);
22191 return __ret;
22192 }
22193 #else
vrev64q_s16(int16x8_t __p0)22194 __ai int16x8_t vrev64q_s16(int16x8_t __p0) {
22195 int16x8_t __ret;
22196 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
22197 __ret = __builtin_shufflevector(__rev0, __rev0, 3, 2, 1, 0, 7, 6, 5, 4);
22198 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
22199 return __ret;
22200 }
22201 #endif
22202
22203 #ifdef __LITTLE_ENDIAN__
vrev64_u8(uint8x8_t __p0)22204 __ai uint8x8_t vrev64_u8(uint8x8_t __p0) {
22205 uint8x8_t __ret;
22206 __ret = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
22207 return __ret;
22208 }
22209 #else
vrev64_u8(uint8x8_t __p0)22210 __ai uint8x8_t vrev64_u8(uint8x8_t __p0) {
22211 uint8x8_t __ret;
22212 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
22213 __ret = __builtin_shufflevector(__rev0, __rev0, 7, 6, 5, 4, 3, 2, 1, 0);
22214 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
22215 return __ret;
22216 }
22217 #endif
22218
22219 #ifdef __LITTLE_ENDIAN__
vrev64_u32(uint32x2_t __p0)22220 __ai uint32x2_t vrev64_u32(uint32x2_t __p0) {
22221 uint32x2_t __ret;
22222 __ret = __builtin_shufflevector(__p0, __p0, 1, 0);
22223 return __ret;
22224 }
22225 #else
vrev64_u32(uint32x2_t __p0)22226 __ai uint32x2_t vrev64_u32(uint32x2_t __p0) {
22227 uint32x2_t __ret;
22228 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
22229 __ret = __builtin_shufflevector(__rev0, __rev0, 1, 0);
22230 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
22231 return __ret;
22232 }
22233 #endif
22234
22235 #ifdef __LITTLE_ENDIAN__
vrev64_u16(uint16x4_t __p0)22236 __ai uint16x4_t vrev64_u16(uint16x4_t __p0) {
22237 uint16x4_t __ret;
22238 __ret = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
22239 return __ret;
22240 }
22241 #else
vrev64_u16(uint16x4_t __p0)22242 __ai uint16x4_t vrev64_u16(uint16x4_t __p0) {
22243 uint16x4_t __ret;
22244 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
22245 __ret = __builtin_shufflevector(__rev0, __rev0, 3, 2, 1, 0);
22246 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
22247 return __ret;
22248 }
22249 #endif
22250
22251 #ifdef __LITTLE_ENDIAN__
vrev64_s8(int8x8_t __p0)22252 __ai int8x8_t vrev64_s8(int8x8_t __p0) {
22253 int8x8_t __ret;
22254 __ret = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
22255 return __ret;
22256 }
22257 #else
vrev64_s8(int8x8_t __p0)22258 __ai int8x8_t vrev64_s8(int8x8_t __p0) {
22259 int8x8_t __ret;
22260 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
22261 __ret = __builtin_shufflevector(__rev0, __rev0, 7, 6, 5, 4, 3, 2, 1, 0);
22262 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
22263 return __ret;
22264 }
22265 #endif
22266
22267 #ifdef __LITTLE_ENDIAN__
vrev64_f32(float32x2_t __p0)22268 __ai float32x2_t vrev64_f32(float32x2_t __p0) {
22269 float32x2_t __ret;
22270 __ret = __builtin_shufflevector(__p0, __p0, 1, 0);
22271 return __ret;
22272 }
22273 #else
vrev64_f32(float32x2_t __p0)22274 __ai float32x2_t vrev64_f32(float32x2_t __p0) {
22275 float32x2_t __ret;
22276 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
22277 __ret = __builtin_shufflevector(__rev0, __rev0, 1, 0);
22278 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
22279 return __ret;
22280 }
22281 #endif
22282
22283 #ifdef __LITTLE_ENDIAN__
vrev64_s32(int32x2_t __p0)22284 __ai int32x2_t vrev64_s32(int32x2_t __p0) {
22285 int32x2_t __ret;
22286 __ret = __builtin_shufflevector(__p0, __p0, 1, 0);
22287 return __ret;
22288 }
22289 #else
vrev64_s32(int32x2_t __p0)22290 __ai int32x2_t vrev64_s32(int32x2_t __p0) {
22291 int32x2_t __ret;
22292 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
22293 __ret = __builtin_shufflevector(__rev0, __rev0, 1, 0);
22294 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
22295 return __ret;
22296 }
22297 #endif
22298
22299 #ifdef __LITTLE_ENDIAN__
vrev64_s16(int16x4_t __p0)22300 __ai int16x4_t vrev64_s16(int16x4_t __p0) {
22301 int16x4_t __ret;
22302 __ret = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
22303 return __ret;
22304 }
22305 #else
vrev64_s16(int16x4_t __p0)22306 __ai int16x4_t vrev64_s16(int16x4_t __p0) {
22307 int16x4_t __ret;
22308 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
22309 __ret = __builtin_shufflevector(__rev0, __rev0, 3, 2, 1, 0);
22310 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
22311 return __ret;
22312 }
22313 #endif
22314
22315 #ifdef __LITTLE_ENDIAN__
vrhaddq_u8(uint8x16_t __p0,uint8x16_t __p1)22316 __ai uint8x16_t vrhaddq_u8(uint8x16_t __p0, uint8x16_t __p1) {
22317 uint8x16_t __ret;
22318 __ret = (uint8x16_t) __builtin_neon_vrhaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
22319 return __ret;
22320 }
22321 #else
vrhaddq_u8(uint8x16_t __p0,uint8x16_t __p1)22322 __ai uint8x16_t vrhaddq_u8(uint8x16_t __p0, uint8x16_t __p1) {
22323 uint8x16_t __ret;
22324 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
22325 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
22326 __ret = (uint8x16_t) __builtin_neon_vrhaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
22327 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
22328 return __ret;
22329 }
22330 #endif
22331
22332 #ifdef __LITTLE_ENDIAN__
vrhaddq_u32(uint32x4_t __p0,uint32x4_t __p1)22333 __ai uint32x4_t vrhaddq_u32(uint32x4_t __p0, uint32x4_t __p1) {
22334 uint32x4_t __ret;
22335 __ret = (uint32x4_t) __builtin_neon_vrhaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
22336 return __ret;
22337 }
22338 #else
vrhaddq_u32(uint32x4_t __p0,uint32x4_t __p1)22339 __ai uint32x4_t vrhaddq_u32(uint32x4_t __p0, uint32x4_t __p1) {
22340 uint32x4_t __ret;
22341 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
22342 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
22343 __ret = (uint32x4_t) __builtin_neon_vrhaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
22344 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
22345 return __ret;
22346 }
22347 #endif
22348
22349 #ifdef __LITTLE_ENDIAN__
vrhaddq_u16(uint16x8_t __p0,uint16x8_t __p1)22350 __ai uint16x8_t vrhaddq_u16(uint16x8_t __p0, uint16x8_t __p1) {
22351 uint16x8_t __ret;
22352 __ret = (uint16x8_t) __builtin_neon_vrhaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
22353 return __ret;
22354 }
22355 #else
vrhaddq_u16(uint16x8_t __p0,uint16x8_t __p1)22356 __ai uint16x8_t vrhaddq_u16(uint16x8_t __p0, uint16x8_t __p1) {
22357 uint16x8_t __ret;
22358 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
22359 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
22360 __ret = (uint16x8_t) __builtin_neon_vrhaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
22361 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
22362 return __ret;
22363 }
22364 #endif
22365
22366 #ifdef __LITTLE_ENDIAN__
vrhaddq_s8(int8x16_t __p0,int8x16_t __p1)22367 __ai int8x16_t vrhaddq_s8(int8x16_t __p0, int8x16_t __p1) {
22368 int8x16_t __ret;
22369 __ret = (int8x16_t) __builtin_neon_vrhaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 32);
22370 return __ret;
22371 }
22372 #else
vrhaddq_s8(int8x16_t __p0,int8x16_t __p1)22373 __ai int8x16_t vrhaddq_s8(int8x16_t __p0, int8x16_t __p1) {
22374 int8x16_t __ret;
22375 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
22376 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
22377 __ret = (int8x16_t) __builtin_neon_vrhaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 32);
22378 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
22379 return __ret;
22380 }
22381 #endif
22382
22383 #ifdef __LITTLE_ENDIAN__
vrhaddq_s32(int32x4_t __p0,int32x4_t __p1)22384 __ai int32x4_t vrhaddq_s32(int32x4_t __p0, int32x4_t __p1) {
22385 int32x4_t __ret;
22386 __ret = (int32x4_t) __builtin_neon_vrhaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
22387 return __ret;
22388 }
22389 #else
vrhaddq_s32(int32x4_t __p0,int32x4_t __p1)22390 __ai int32x4_t vrhaddq_s32(int32x4_t __p0, int32x4_t __p1) {
22391 int32x4_t __ret;
22392 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
22393 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
22394 __ret = (int32x4_t) __builtin_neon_vrhaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
22395 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
22396 return __ret;
22397 }
22398 #endif
22399
22400 #ifdef __LITTLE_ENDIAN__
vrhaddq_s16(int16x8_t __p0,int16x8_t __p1)22401 __ai int16x8_t vrhaddq_s16(int16x8_t __p0, int16x8_t __p1) {
22402 int16x8_t __ret;
22403 __ret = (int16x8_t) __builtin_neon_vrhaddq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
22404 return __ret;
22405 }
22406 #else
vrhaddq_s16(int16x8_t __p0,int16x8_t __p1)22407 __ai int16x8_t vrhaddq_s16(int16x8_t __p0, int16x8_t __p1) {
22408 int16x8_t __ret;
22409 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
22410 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
22411 __ret = (int16x8_t) __builtin_neon_vrhaddq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
22412 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
22413 return __ret;
22414 }
22415 #endif
22416
22417 #ifdef __LITTLE_ENDIAN__
vrhadd_u8(uint8x8_t __p0,uint8x8_t __p1)22418 __ai uint8x8_t vrhadd_u8(uint8x8_t __p0, uint8x8_t __p1) {
22419 uint8x8_t __ret;
22420 __ret = (uint8x8_t) __builtin_neon_vrhadd_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
22421 return __ret;
22422 }
22423 #else
vrhadd_u8(uint8x8_t __p0,uint8x8_t __p1)22424 __ai uint8x8_t vrhadd_u8(uint8x8_t __p0, uint8x8_t __p1) {
22425 uint8x8_t __ret;
22426 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
22427 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
22428 __ret = (uint8x8_t) __builtin_neon_vrhadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
22429 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
22430 return __ret;
22431 }
22432 #endif
22433
22434 #ifdef __LITTLE_ENDIAN__
vrhadd_u32(uint32x2_t __p0,uint32x2_t __p1)22435 __ai uint32x2_t vrhadd_u32(uint32x2_t __p0, uint32x2_t __p1) {
22436 uint32x2_t __ret;
22437 __ret = (uint32x2_t) __builtin_neon_vrhadd_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
22438 return __ret;
22439 }
22440 #else
vrhadd_u32(uint32x2_t __p0,uint32x2_t __p1)22441 __ai uint32x2_t vrhadd_u32(uint32x2_t __p0, uint32x2_t __p1) {
22442 uint32x2_t __ret;
22443 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
22444 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
22445 __ret = (uint32x2_t) __builtin_neon_vrhadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
22446 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
22447 return __ret;
22448 }
22449 #endif
22450
22451 #ifdef __LITTLE_ENDIAN__
vrhadd_u16(uint16x4_t __p0,uint16x4_t __p1)22452 __ai uint16x4_t vrhadd_u16(uint16x4_t __p0, uint16x4_t __p1) {
22453 uint16x4_t __ret;
22454 __ret = (uint16x4_t) __builtin_neon_vrhadd_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
22455 return __ret;
22456 }
22457 #else
vrhadd_u16(uint16x4_t __p0,uint16x4_t __p1)22458 __ai uint16x4_t vrhadd_u16(uint16x4_t __p0, uint16x4_t __p1) {
22459 uint16x4_t __ret;
22460 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
22461 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
22462 __ret = (uint16x4_t) __builtin_neon_vrhadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
22463 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
22464 return __ret;
22465 }
22466 #endif
22467
22468 #ifdef __LITTLE_ENDIAN__
vrhadd_s8(int8x8_t __p0,int8x8_t __p1)22469 __ai int8x8_t vrhadd_s8(int8x8_t __p0, int8x8_t __p1) {
22470 int8x8_t __ret;
22471 __ret = (int8x8_t) __builtin_neon_vrhadd_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
22472 return __ret;
22473 }
22474 #else
vrhadd_s8(int8x8_t __p0,int8x8_t __p1)22475 __ai int8x8_t vrhadd_s8(int8x8_t __p0, int8x8_t __p1) {
22476 int8x8_t __ret;
22477 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
22478 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
22479 __ret = (int8x8_t) __builtin_neon_vrhadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 0);
22480 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
22481 return __ret;
22482 }
22483 #endif
22484
22485 #ifdef __LITTLE_ENDIAN__
vrhadd_s32(int32x2_t __p0,int32x2_t __p1)22486 __ai int32x2_t vrhadd_s32(int32x2_t __p0, int32x2_t __p1) {
22487 int32x2_t __ret;
22488 __ret = (int32x2_t) __builtin_neon_vrhadd_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
22489 return __ret;
22490 }
22491 #else
vrhadd_s32(int32x2_t __p0,int32x2_t __p1)22492 __ai int32x2_t vrhadd_s32(int32x2_t __p0, int32x2_t __p1) {
22493 int32x2_t __ret;
22494 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
22495 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
22496 __ret = (int32x2_t) __builtin_neon_vrhadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
22497 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
22498 return __ret;
22499 }
22500 #endif
22501
22502 #ifdef __LITTLE_ENDIAN__
vrhadd_s16(int16x4_t __p0,int16x4_t __p1)22503 __ai int16x4_t vrhadd_s16(int16x4_t __p0, int16x4_t __p1) {
22504 int16x4_t __ret;
22505 __ret = (int16x4_t) __builtin_neon_vrhadd_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
22506 return __ret;
22507 }
22508 #else
vrhadd_s16(int16x4_t __p0,int16x4_t __p1)22509 __ai int16x4_t vrhadd_s16(int16x4_t __p0, int16x4_t __p1) {
22510 int16x4_t __ret;
22511 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
22512 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
22513 __ret = (int16x4_t) __builtin_neon_vrhadd_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
22514 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
22515 return __ret;
22516 }
22517 #endif
22518
22519 #ifdef __LITTLE_ENDIAN__
vrshlq_u8(uint8x16_t __p0,int8x16_t __p1)22520 __ai uint8x16_t vrshlq_u8(uint8x16_t __p0, int8x16_t __p1) {
22521 uint8x16_t __ret;
22522 __ret = (uint8x16_t) __builtin_neon_vrshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
22523 return __ret;
22524 }
22525 #else
vrshlq_u8(uint8x16_t __p0,int8x16_t __p1)22526 __ai uint8x16_t vrshlq_u8(uint8x16_t __p0, int8x16_t __p1) {
22527 uint8x16_t __ret;
22528 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
22529 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
22530 __ret = (uint8x16_t) __builtin_neon_vrshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
22531 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
22532 return __ret;
22533 }
22534 #endif
22535
22536 #ifdef __LITTLE_ENDIAN__
vrshlq_u32(uint32x4_t __p0,int32x4_t __p1)22537 __ai uint32x4_t vrshlq_u32(uint32x4_t __p0, int32x4_t __p1) {
22538 uint32x4_t __ret;
22539 __ret = (uint32x4_t) __builtin_neon_vrshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
22540 return __ret;
22541 }
22542 #else
vrshlq_u32(uint32x4_t __p0,int32x4_t __p1)22543 __ai uint32x4_t vrshlq_u32(uint32x4_t __p0, int32x4_t __p1) {
22544 uint32x4_t __ret;
22545 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
22546 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
22547 __ret = (uint32x4_t) __builtin_neon_vrshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
22548 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
22549 return __ret;
22550 }
22551 #endif
22552
22553 #ifdef __LITTLE_ENDIAN__
vrshlq_u64(uint64x2_t __p0,int64x2_t __p1)22554 __ai uint64x2_t vrshlq_u64(uint64x2_t __p0, int64x2_t __p1) {
22555 uint64x2_t __ret;
22556 __ret = (uint64x2_t) __builtin_neon_vrshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 51);
22557 return __ret;
22558 }
22559 #else
vrshlq_u64(uint64x2_t __p0,int64x2_t __p1)22560 __ai uint64x2_t vrshlq_u64(uint64x2_t __p0, int64x2_t __p1) {
22561 uint64x2_t __ret;
22562 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
22563 int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
22564 __ret = (uint64x2_t) __builtin_neon_vrshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 51);
22565 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
22566 return __ret;
22567 }
22568 #endif
22569
22570 #ifdef __LITTLE_ENDIAN__
vrshlq_u16(uint16x8_t __p0,int16x8_t __p1)22571 __ai uint16x8_t vrshlq_u16(uint16x8_t __p0, int16x8_t __p1) {
22572 uint16x8_t __ret;
22573 __ret = (uint16x8_t) __builtin_neon_vrshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
22574 return __ret;
22575 }
22576 #else
vrshlq_u16(uint16x8_t __p0,int16x8_t __p1)22577 __ai uint16x8_t vrshlq_u16(uint16x8_t __p0, int16x8_t __p1) {
22578 uint16x8_t __ret;
22579 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
22580 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
22581 __ret = (uint16x8_t) __builtin_neon_vrshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
22582 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
22583 return __ret;
22584 }
22585 #endif
22586
22587 #ifdef __LITTLE_ENDIAN__
vrshlq_s8(int8x16_t __p0,int8x16_t __p1)22588 __ai int8x16_t vrshlq_s8(int8x16_t __p0, int8x16_t __p1) {
22589 int8x16_t __ret;
22590 __ret = (int8x16_t) __builtin_neon_vrshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 32);
22591 return __ret;
22592 }
22593 #else
vrshlq_s8(int8x16_t __p0,int8x16_t __p1)22594 __ai int8x16_t vrshlq_s8(int8x16_t __p0, int8x16_t __p1) {
22595 int8x16_t __ret;
22596 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
22597 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
22598 __ret = (int8x16_t) __builtin_neon_vrshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 32);
22599 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
22600 return __ret;
22601 }
22602 #endif
22603
22604 #ifdef __LITTLE_ENDIAN__
vrshlq_s32(int32x4_t __p0,int32x4_t __p1)22605 __ai int32x4_t vrshlq_s32(int32x4_t __p0, int32x4_t __p1) {
22606 int32x4_t __ret;
22607 __ret = (int32x4_t) __builtin_neon_vrshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
22608 return __ret;
22609 }
22610 #else
vrshlq_s32(int32x4_t __p0,int32x4_t __p1)22611 __ai int32x4_t vrshlq_s32(int32x4_t __p0, int32x4_t __p1) {
22612 int32x4_t __ret;
22613 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
22614 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
22615 __ret = (int32x4_t) __builtin_neon_vrshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
22616 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
22617 return __ret;
22618 }
22619 #endif
22620
22621 #ifdef __LITTLE_ENDIAN__
vrshlq_s64(int64x2_t __p0,int64x2_t __p1)22622 __ai int64x2_t vrshlq_s64(int64x2_t __p0, int64x2_t __p1) {
22623 int64x2_t __ret;
22624 __ret = (int64x2_t) __builtin_neon_vrshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 35);
22625 return __ret;
22626 }
22627 #else
vrshlq_s64(int64x2_t __p0,int64x2_t __p1)22628 __ai int64x2_t vrshlq_s64(int64x2_t __p0, int64x2_t __p1) {
22629 int64x2_t __ret;
22630 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
22631 int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
22632 __ret = (int64x2_t) __builtin_neon_vrshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 35);
22633 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
22634 return __ret;
22635 }
22636 #endif
22637
22638 #ifdef __LITTLE_ENDIAN__
vrshlq_s16(int16x8_t __p0,int16x8_t __p1)22639 __ai int16x8_t vrshlq_s16(int16x8_t __p0, int16x8_t __p1) {
22640 int16x8_t __ret;
22641 __ret = (int16x8_t) __builtin_neon_vrshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
22642 return __ret;
22643 }
22644 #else
vrshlq_s16(int16x8_t __p0,int16x8_t __p1)22645 __ai int16x8_t vrshlq_s16(int16x8_t __p0, int16x8_t __p1) {
22646 int16x8_t __ret;
22647 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
22648 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
22649 __ret = (int16x8_t) __builtin_neon_vrshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
22650 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
22651 return __ret;
22652 }
22653 #endif
22654
22655 #ifdef __LITTLE_ENDIAN__
vrshl_u8(uint8x8_t __p0,int8x8_t __p1)22656 __ai uint8x8_t vrshl_u8(uint8x8_t __p0, int8x8_t __p1) {
22657 uint8x8_t __ret;
22658 __ret = (uint8x8_t) __builtin_neon_vrshl_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
22659 return __ret;
22660 }
22661 #else
vrshl_u8(uint8x8_t __p0,int8x8_t __p1)22662 __ai uint8x8_t vrshl_u8(uint8x8_t __p0, int8x8_t __p1) {
22663 uint8x8_t __ret;
22664 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
22665 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
22666 __ret = (uint8x8_t) __builtin_neon_vrshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
22667 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
22668 return __ret;
22669 }
22670 #endif
22671
22672 #ifdef __LITTLE_ENDIAN__
vrshl_u32(uint32x2_t __p0,int32x2_t __p1)22673 __ai uint32x2_t vrshl_u32(uint32x2_t __p0, int32x2_t __p1) {
22674 uint32x2_t __ret;
22675 __ret = (uint32x2_t) __builtin_neon_vrshl_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
22676 return __ret;
22677 }
22678 #else
vrshl_u32(uint32x2_t __p0,int32x2_t __p1)22679 __ai uint32x2_t vrshl_u32(uint32x2_t __p0, int32x2_t __p1) {
22680 uint32x2_t __ret;
22681 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
22682 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
22683 __ret = (uint32x2_t) __builtin_neon_vrshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
22684 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
22685 return __ret;
22686 }
22687 #endif
22688
vrshl_u64(uint64x1_t __p0,int64x1_t __p1)22689 __ai uint64x1_t vrshl_u64(uint64x1_t __p0, int64x1_t __p1) {
22690 uint64x1_t __ret;
22691 __ret = (uint64x1_t) __builtin_neon_vrshl_v((int8x8_t)__p0, (int8x8_t)__p1, 19);
22692 return __ret;
22693 }
22694 #ifdef __LITTLE_ENDIAN__
vrshl_u16(uint16x4_t __p0,int16x4_t __p1)22695 __ai uint16x4_t vrshl_u16(uint16x4_t __p0, int16x4_t __p1) {
22696 uint16x4_t __ret;
22697 __ret = (uint16x4_t) __builtin_neon_vrshl_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
22698 return __ret;
22699 }
22700 #else
vrshl_u16(uint16x4_t __p0,int16x4_t __p1)22701 __ai uint16x4_t vrshl_u16(uint16x4_t __p0, int16x4_t __p1) {
22702 uint16x4_t __ret;
22703 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
22704 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
22705 __ret = (uint16x4_t) __builtin_neon_vrshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
22706 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
22707 return __ret;
22708 }
22709 #endif
22710
22711 #ifdef __LITTLE_ENDIAN__
vrshl_s8(int8x8_t __p0,int8x8_t __p1)22712 __ai int8x8_t vrshl_s8(int8x8_t __p0, int8x8_t __p1) {
22713 int8x8_t __ret;
22714 __ret = (int8x8_t) __builtin_neon_vrshl_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
22715 return __ret;
22716 }
22717 #else
vrshl_s8(int8x8_t __p0,int8x8_t __p1)22718 __ai int8x8_t vrshl_s8(int8x8_t __p0, int8x8_t __p1) {
22719 int8x8_t __ret;
22720 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
22721 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
22722 __ret = (int8x8_t) __builtin_neon_vrshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 0);
22723 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
22724 return __ret;
22725 }
22726 #endif
22727
22728 #ifdef __LITTLE_ENDIAN__
vrshl_s32(int32x2_t __p0,int32x2_t __p1)22729 __ai int32x2_t vrshl_s32(int32x2_t __p0, int32x2_t __p1) {
22730 int32x2_t __ret;
22731 __ret = (int32x2_t) __builtin_neon_vrshl_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
22732 return __ret;
22733 }
22734 #else
vrshl_s32(int32x2_t __p0,int32x2_t __p1)22735 __ai int32x2_t vrshl_s32(int32x2_t __p0, int32x2_t __p1) {
22736 int32x2_t __ret;
22737 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
22738 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
22739 __ret = (int32x2_t) __builtin_neon_vrshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
22740 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
22741 return __ret;
22742 }
22743 #endif
22744
vrshl_s64(int64x1_t __p0,int64x1_t __p1)22745 __ai int64x1_t vrshl_s64(int64x1_t __p0, int64x1_t __p1) {
22746 int64x1_t __ret;
22747 __ret = (int64x1_t) __builtin_neon_vrshl_v((int8x8_t)__p0, (int8x8_t)__p1, 3);
22748 return __ret;
22749 }
22750 #ifdef __LITTLE_ENDIAN__
vrshl_s16(int16x4_t __p0,int16x4_t __p1)22751 __ai int16x4_t vrshl_s16(int16x4_t __p0, int16x4_t __p1) {
22752 int16x4_t __ret;
22753 __ret = (int16x4_t) __builtin_neon_vrshl_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
22754 return __ret;
22755 }
22756 #else
vrshl_s16(int16x4_t __p0,int16x4_t __p1)22757 __ai int16x4_t vrshl_s16(int16x4_t __p0, int16x4_t __p1) {
22758 int16x4_t __ret;
22759 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
22760 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
22761 __ret = (int16x4_t) __builtin_neon_vrshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
22762 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
22763 return __ret;
22764 }
22765 #endif
22766
22767 #ifdef __LITTLE_ENDIAN__
22768 #define vrshrq_n_u8(__p0, __p1) __extension__ ({ \
22769 uint8x16_t __ret; \
22770 uint8x16_t __s0 = __p0; \
22771 __ret = (uint8x16_t) __builtin_neon_vrshrq_n_v((int8x16_t)__s0, __p1, 48); \
22772 __ret; \
22773 })
22774 #else
22775 #define vrshrq_n_u8(__p0, __p1) __extension__ ({ \
22776 uint8x16_t __ret; \
22777 uint8x16_t __s0 = __p0; \
22778 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
22779 __ret = (uint8x16_t) __builtin_neon_vrshrq_n_v((int8x16_t)__rev0, __p1, 48); \
22780 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
22781 __ret; \
22782 })
22783 #endif
22784
22785 #ifdef __LITTLE_ENDIAN__
22786 #define vrshrq_n_u32(__p0, __p1) __extension__ ({ \
22787 uint32x4_t __ret; \
22788 uint32x4_t __s0 = __p0; \
22789 __ret = (uint32x4_t) __builtin_neon_vrshrq_n_v((int8x16_t)__s0, __p1, 50); \
22790 __ret; \
22791 })
22792 #else
22793 #define vrshrq_n_u32(__p0, __p1) __extension__ ({ \
22794 uint32x4_t __ret; \
22795 uint32x4_t __s0 = __p0; \
22796 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
22797 __ret = (uint32x4_t) __builtin_neon_vrshrq_n_v((int8x16_t)__rev0, __p1, 50); \
22798 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
22799 __ret; \
22800 })
22801 #endif
22802
22803 #ifdef __LITTLE_ENDIAN__
22804 #define vrshrq_n_u64(__p0, __p1) __extension__ ({ \
22805 uint64x2_t __ret; \
22806 uint64x2_t __s0 = __p0; \
22807 __ret = (uint64x2_t) __builtin_neon_vrshrq_n_v((int8x16_t)__s0, __p1, 51); \
22808 __ret; \
22809 })
22810 #else
22811 #define vrshrq_n_u64(__p0, __p1) __extension__ ({ \
22812 uint64x2_t __ret; \
22813 uint64x2_t __s0 = __p0; \
22814 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
22815 __ret = (uint64x2_t) __builtin_neon_vrshrq_n_v((int8x16_t)__rev0, __p1, 51); \
22816 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
22817 __ret; \
22818 })
22819 #endif
22820
22821 #ifdef __LITTLE_ENDIAN__
22822 #define vrshrq_n_u16(__p0, __p1) __extension__ ({ \
22823 uint16x8_t __ret; \
22824 uint16x8_t __s0 = __p0; \
22825 __ret = (uint16x8_t) __builtin_neon_vrshrq_n_v((int8x16_t)__s0, __p1, 49); \
22826 __ret; \
22827 })
22828 #else
22829 #define vrshrq_n_u16(__p0, __p1) __extension__ ({ \
22830 uint16x8_t __ret; \
22831 uint16x8_t __s0 = __p0; \
22832 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
22833 __ret = (uint16x8_t) __builtin_neon_vrshrq_n_v((int8x16_t)__rev0, __p1, 49); \
22834 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
22835 __ret; \
22836 })
22837 #endif
22838
22839 #ifdef __LITTLE_ENDIAN__
22840 #define vrshrq_n_s8(__p0, __p1) __extension__ ({ \
22841 int8x16_t __ret; \
22842 int8x16_t __s0 = __p0; \
22843 __ret = (int8x16_t) __builtin_neon_vrshrq_n_v((int8x16_t)__s0, __p1, 32); \
22844 __ret; \
22845 })
22846 #else
22847 #define vrshrq_n_s8(__p0, __p1) __extension__ ({ \
22848 int8x16_t __ret; \
22849 int8x16_t __s0 = __p0; \
22850 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
22851 __ret = (int8x16_t) __builtin_neon_vrshrq_n_v((int8x16_t)__rev0, __p1, 32); \
22852 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
22853 __ret; \
22854 })
22855 #endif
22856
22857 #ifdef __LITTLE_ENDIAN__
22858 #define vrshrq_n_s32(__p0, __p1) __extension__ ({ \
22859 int32x4_t __ret; \
22860 int32x4_t __s0 = __p0; \
22861 __ret = (int32x4_t) __builtin_neon_vrshrq_n_v((int8x16_t)__s0, __p1, 34); \
22862 __ret; \
22863 })
22864 #else
22865 #define vrshrq_n_s32(__p0, __p1) __extension__ ({ \
22866 int32x4_t __ret; \
22867 int32x4_t __s0 = __p0; \
22868 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
22869 __ret = (int32x4_t) __builtin_neon_vrshrq_n_v((int8x16_t)__rev0, __p1, 34); \
22870 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
22871 __ret; \
22872 })
22873 #endif
22874
22875 #ifdef __LITTLE_ENDIAN__
22876 #define vrshrq_n_s64(__p0, __p1) __extension__ ({ \
22877 int64x2_t __ret; \
22878 int64x2_t __s0 = __p0; \
22879 __ret = (int64x2_t) __builtin_neon_vrshrq_n_v((int8x16_t)__s0, __p1, 35); \
22880 __ret; \
22881 })
22882 #else
22883 #define vrshrq_n_s64(__p0, __p1) __extension__ ({ \
22884 int64x2_t __ret; \
22885 int64x2_t __s0 = __p0; \
22886 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
22887 __ret = (int64x2_t) __builtin_neon_vrshrq_n_v((int8x16_t)__rev0, __p1, 35); \
22888 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
22889 __ret; \
22890 })
22891 #endif
22892
22893 #ifdef __LITTLE_ENDIAN__
22894 #define vrshrq_n_s16(__p0, __p1) __extension__ ({ \
22895 int16x8_t __ret; \
22896 int16x8_t __s0 = __p0; \
22897 __ret = (int16x8_t) __builtin_neon_vrshrq_n_v((int8x16_t)__s0, __p1, 33); \
22898 __ret; \
22899 })
22900 #else
22901 #define vrshrq_n_s16(__p0, __p1) __extension__ ({ \
22902 int16x8_t __ret; \
22903 int16x8_t __s0 = __p0; \
22904 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
22905 __ret = (int16x8_t) __builtin_neon_vrshrq_n_v((int8x16_t)__rev0, __p1, 33); \
22906 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
22907 __ret; \
22908 })
22909 #endif
22910
22911 #ifdef __LITTLE_ENDIAN__
22912 #define vrshr_n_u8(__p0, __p1) __extension__ ({ \
22913 uint8x8_t __ret; \
22914 uint8x8_t __s0 = __p0; \
22915 __ret = (uint8x8_t) __builtin_neon_vrshr_n_v((int8x8_t)__s0, __p1, 16); \
22916 __ret; \
22917 })
22918 #else
22919 #define vrshr_n_u8(__p0, __p1) __extension__ ({ \
22920 uint8x8_t __ret; \
22921 uint8x8_t __s0 = __p0; \
22922 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
22923 __ret = (uint8x8_t) __builtin_neon_vrshr_n_v((int8x8_t)__rev0, __p1, 16); \
22924 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
22925 __ret; \
22926 })
22927 #endif
22928
22929 #ifdef __LITTLE_ENDIAN__
22930 #define vrshr_n_u32(__p0, __p1) __extension__ ({ \
22931 uint32x2_t __ret; \
22932 uint32x2_t __s0 = __p0; \
22933 __ret = (uint32x2_t) __builtin_neon_vrshr_n_v((int8x8_t)__s0, __p1, 18); \
22934 __ret; \
22935 })
22936 #else
22937 #define vrshr_n_u32(__p0, __p1) __extension__ ({ \
22938 uint32x2_t __ret; \
22939 uint32x2_t __s0 = __p0; \
22940 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
22941 __ret = (uint32x2_t) __builtin_neon_vrshr_n_v((int8x8_t)__rev0, __p1, 18); \
22942 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
22943 __ret; \
22944 })
22945 #endif
22946
22947 #define vrshr_n_u64(__p0, __p1) __extension__ ({ \
22948 uint64x1_t __ret; \
22949 uint64x1_t __s0 = __p0; \
22950 __ret = (uint64x1_t) __builtin_neon_vrshr_n_v((int8x8_t)__s0, __p1, 19); \
22951 __ret; \
22952 })
22953 #ifdef __LITTLE_ENDIAN__
22954 #define vrshr_n_u16(__p0, __p1) __extension__ ({ \
22955 uint16x4_t __ret; \
22956 uint16x4_t __s0 = __p0; \
22957 __ret = (uint16x4_t) __builtin_neon_vrshr_n_v((int8x8_t)__s0, __p1, 17); \
22958 __ret; \
22959 })
22960 #else
22961 #define vrshr_n_u16(__p0, __p1) __extension__ ({ \
22962 uint16x4_t __ret; \
22963 uint16x4_t __s0 = __p0; \
22964 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
22965 __ret = (uint16x4_t) __builtin_neon_vrshr_n_v((int8x8_t)__rev0, __p1, 17); \
22966 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
22967 __ret; \
22968 })
22969 #endif
22970
22971 #ifdef __LITTLE_ENDIAN__
22972 #define vrshr_n_s8(__p0, __p1) __extension__ ({ \
22973 int8x8_t __ret; \
22974 int8x8_t __s0 = __p0; \
22975 __ret = (int8x8_t) __builtin_neon_vrshr_n_v((int8x8_t)__s0, __p1, 0); \
22976 __ret; \
22977 })
22978 #else
22979 #define vrshr_n_s8(__p0, __p1) __extension__ ({ \
22980 int8x8_t __ret; \
22981 int8x8_t __s0 = __p0; \
22982 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
22983 __ret = (int8x8_t) __builtin_neon_vrshr_n_v((int8x8_t)__rev0, __p1, 0); \
22984 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
22985 __ret; \
22986 })
22987 #endif
22988
22989 #ifdef __LITTLE_ENDIAN__
22990 #define vrshr_n_s32(__p0, __p1) __extension__ ({ \
22991 int32x2_t __ret; \
22992 int32x2_t __s0 = __p0; \
22993 __ret = (int32x2_t) __builtin_neon_vrshr_n_v((int8x8_t)__s0, __p1, 2); \
22994 __ret; \
22995 })
22996 #else
22997 #define vrshr_n_s32(__p0, __p1) __extension__ ({ \
22998 int32x2_t __ret; \
22999 int32x2_t __s0 = __p0; \
23000 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
23001 __ret = (int32x2_t) __builtin_neon_vrshr_n_v((int8x8_t)__rev0, __p1, 2); \
23002 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
23003 __ret; \
23004 })
23005 #endif
23006
23007 #define vrshr_n_s64(__p0, __p1) __extension__ ({ \
23008 int64x1_t __ret; \
23009 int64x1_t __s0 = __p0; \
23010 __ret = (int64x1_t) __builtin_neon_vrshr_n_v((int8x8_t)__s0, __p1, 3); \
23011 __ret; \
23012 })
23013 #ifdef __LITTLE_ENDIAN__
23014 #define vrshr_n_s16(__p0, __p1) __extension__ ({ \
23015 int16x4_t __ret; \
23016 int16x4_t __s0 = __p0; \
23017 __ret = (int16x4_t) __builtin_neon_vrshr_n_v((int8x8_t)__s0, __p1, 1); \
23018 __ret; \
23019 })
23020 #else
23021 #define vrshr_n_s16(__p0, __p1) __extension__ ({ \
23022 int16x4_t __ret; \
23023 int16x4_t __s0 = __p0; \
23024 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
23025 __ret = (int16x4_t) __builtin_neon_vrshr_n_v((int8x8_t)__rev0, __p1, 1); \
23026 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
23027 __ret; \
23028 })
23029 #endif
23030
23031 #ifdef __LITTLE_ENDIAN__
23032 #define vrshrn_n_u32(__p0, __p1) __extension__ ({ \
23033 uint16x4_t __ret; \
23034 uint32x4_t __s0 = __p0; \
23035 __ret = (uint16x4_t) __builtin_neon_vrshrn_n_v((int8x16_t)__s0, __p1, 17); \
23036 __ret; \
23037 })
23038 #else
23039 #define vrshrn_n_u32(__p0, __p1) __extension__ ({ \
23040 uint16x4_t __ret; \
23041 uint32x4_t __s0 = __p0; \
23042 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
23043 __ret = (uint16x4_t) __builtin_neon_vrshrn_n_v((int8x16_t)__rev0, __p1, 17); \
23044 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
23045 __ret; \
23046 })
23047 #define __noswap_vrshrn_n_u32(__p0, __p1) __extension__ ({ \
23048 uint16x4_t __ret; \
23049 uint32x4_t __s0 = __p0; \
23050 __ret = (uint16x4_t) __builtin_neon_vrshrn_n_v((int8x16_t)__s0, __p1, 17); \
23051 __ret; \
23052 })
23053 #endif
23054
23055 #ifdef __LITTLE_ENDIAN__
23056 #define vrshrn_n_u64(__p0, __p1) __extension__ ({ \
23057 uint32x2_t __ret; \
23058 uint64x2_t __s0 = __p0; \
23059 __ret = (uint32x2_t) __builtin_neon_vrshrn_n_v((int8x16_t)__s0, __p1, 18); \
23060 __ret; \
23061 })
23062 #else
23063 #define vrshrn_n_u64(__p0, __p1) __extension__ ({ \
23064 uint32x2_t __ret; \
23065 uint64x2_t __s0 = __p0; \
23066 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
23067 __ret = (uint32x2_t) __builtin_neon_vrshrn_n_v((int8x16_t)__rev0, __p1, 18); \
23068 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
23069 __ret; \
23070 })
23071 #define __noswap_vrshrn_n_u64(__p0, __p1) __extension__ ({ \
23072 uint32x2_t __ret; \
23073 uint64x2_t __s0 = __p0; \
23074 __ret = (uint32x2_t) __builtin_neon_vrshrn_n_v((int8x16_t)__s0, __p1, 18); \
23075 __ret; \
23076 })
23077 #endif
23078
23079 #ifdef __LITTLE_ENDIAN__
23080 #define vrshrn_n_u16(__p0, __p1) __extension__ ({ \
23081 uint8x8_t __ret; \
23082 uint16x8_t __s0 = __p0; \
23083 __ret = (uint8x8_t) __builtin_neon_vrshrn_n_v((int8x16_t)__s0, __p1, 16); \
23084 __ret; \
23085 })
23086 #else
23087 #define vrshrn_n_u16(__p0, __p1) __extension__ ({ \
23088 uint8x8_t __ret; \
23089 uint16x8_t __s0 = __p0; \
23090 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
23091 __ret = (uint8x8_t) __builtin_neon_vrshrn_n_v((int8x16_t)__rev0, __p1, 16); \
23092 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
23093 __ret; \
23094 })
23095 #define __noswap_vrshrn_n_u16(__p0, __p1) __extension__ ({ \
23096 uint8x8_t __ret; \
23097 uint16x8_t __s0 = __p0; \
23098 __ret = (uint8x8_t) __builtin_neon_vrshrn_n_v((int8x16_t)__s0, __p1, 16); \
23099 __ret; \
23100 })
23101 #endif
23102
23103 #ifdef __LITTLE_ENDIAN__
23104 #define vrshrn_n_s32(__p0, __p1) __extension__ ({ \
23105 int16x4_t __ret; \
23106 int32x4_t __s0 = __p0; \
23107 __ret = (int16x4_t) __builtin_neon_vrshrn_n_v((int8x16_t)__s0, __p1, 1); \
23108 __ret; \
23109 })
23110 #else
23111 #define vrshrn_n_s32(__p0, __p1) __extension__ ({ \
23112 int16x4_t __ret; \
23113 int32x4_t __s0 = __p0; \
23114 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
23115 __ret = (int16x4_t) __builtin_neon_vrshrn_n_v((int8x16_t)__rev0, __p1, 1); \
23116 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
23117 __ret; \
23118 })
23119 #define __noswap_vrshrn_n_s32(__p0, __p1) __extension__ ({ \
23120 int16x4_t __ret; \
23121 int32x4_t __s0 = __p0; \
23122 __ret = (int16x4_t) __builtin_neon_vrshrn_n_v((int8x16_t)__s0, __p1, 1); \
23123 __ret; \
23124 })
23125 #endif
23126
23127 #ifdef __LITTLE_ENDIAN__
23128 #define vrshrn_n_s64(__p0, __p1) __extension__ ({ \
23129 int32x2_t __ret; \
23130 int64x2_t __s0 = __p0; \
23131 __ret = (int32x2_t) __builtin_neon_vrshrn_n_v((int8x16_t)__s0, __p1, 2); \
23132 __ret; \
23133 })
23134 #else
23135 #define vrshrn_n_s64(__p0, __p1) __extension__ ({ \
23136 int32x2_t __ret; \
23137 int64x2_t __s0 = __p0; \
23138 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
23139 __ret = (int32x2_t) __builtin_neon_vrshrn_n_v((int8x16_t)__rev0, __p1, 2); \
23140 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
23141 __ret; \
23142 })
23143 #define __noswap_vrshrn_n_s64(__p0, __p1) __extension__ ({ \
23144 int32x2_t __ret; \
23145 int64x2_t __s0 = __p0; \
23146 __ret = (int32x2_t) __builtin_neon_vrshrn_n_v((int8x16_t)__s0, __p1, 2); \
23147 __ret; \
23148 })
23149 #endif
23150
23151 #ifdef __LITTLE_ENDIAN__
23152 #define vrshrn_n_s16(__p0, __p1) __extension__ ({ \
23153 int8x8_t __ret; \
23154 int16x8_t __s0 = __p0; \
23155 __ret = (int8x8_t) __builtin_neon_vrshrn_n_v((int8x16_t)__s0, __p1, 0); \
23156 __ret; \
23157 })
23158 #else
23159 #define vrshrn_n_s16(__p0, __p1) __extension__ ({ \
23160 int8x8_t __ret; \
23161 int16x8_t __s0 = __p0; \
23162 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
23163 __ret = (int8x8_t) __builtin_neon_vrshrn_n_v((int8x16_t)__rev0, __p1, 0); \
23164 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
23165 __ret; \
23166 })
23167 #define __noswap_vrshrn_n_s16(__p0, __p1) __extension__ ({ \
23168 int8x8_t __ret; \
23169 int16x8_t __s0 = __p0; \
23170 __ret = (int8x8_t) __builtin_neon_vrshrn_n_v((int8x16_t)__s0, __p1, 0); \
23171 __ret; \
23172 })
23173 #endif
23174
23175 #ifdef __LITTLE_ENDIAN__
vrsqrteq_u32(uint32x4_t __p0)23176 __ai uint32x4_t vrsqrteq_u32(uint32x4_t __p0) {
23177 uint32x4_t __ret;
23178 __ret = (uint32x4_t) __builtin_neon_vrsqrteq_v((int8x16_t)__p0, 50);
23179 return __ret;
23180 }
23181 #else
vrsqrteq_u32(uint32x4_t __p0)23182 __ai uint32x4_t vrsqrteq_u32(uint32x4_t __p0) {
23183 uint32x4_t __ret;
23184 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
23185 __ret = (uint32x4_t) __builtin_neon_vrsqrteq_v((int8x16_t)__rev0, 50);
23186 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
23187 return __ret;
23188 }
23189 #endif
23190
23191 #ifdef __LITTLE_ENDIAN__
vrsqrteq_f32(float32x4_t __p0)23192 __ai float32x4_t vrsqrteq_f32(float32x4_t __p0) {
23193 float32x4_t __ret;
23194 __ret = (float32x4_t) __builtin_neon_vrsqrteq_v((int8x16_t)__p0, 41);
23195 return __ret;
23196 }
23197 #else
vrsqrteq_f32(float32x4_t __p0)23198 __ai float32x4_t vrsqrteq_f32(float32x4_t __p0) {
23199 float32x4_t __ret;
23200 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
23201 __ret = (float32x4_t) __builtin_neon_vrsqrteq_v((int8x16_t)__rev0, 41);
23202 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
23203 return __ret;
23204 }
23205 #endif
23206
23207 #ifdef __LITTLE_ENDIAN__
vrsqrte_u32(uint32x2_t __p0)23208 __ai uint32x2_t vrsqrte_u32(uint32x2_t __p0) {
23209 uint32x2_t __ret;
23210 __ret = (uint32x2_t) __builtin_neon_vrsqrte_v((int8x8_t)__p0, 18);
23211 return __ret;
23212 }
23213 #else
vrsqrte_u32(uint32x2_t __p0)23214 __ai uint32x2_t vrsqrte_u32(uint32x2_t __p0) {
23215 uint32x2_t __ret;
23216 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
23217 __ret = (uint32x2_t) __builtin_neon_vrsqrte_v((int8x8_t)__rev0, 18);
23218 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
23219 return __ret;
23220 }
23221 #endif
23222
23223 #ifdef __LITTLE_ENDIAN__
vrsqrte_f32(float32x2_t __p0)23224 __ai float32x2_t vrsqrte_f32(float32x2_t __p0) {
23225 float32x2_t __ret;
23226 __ret = (float32x2_t) __builtin_neon_vrsqrte_v((int8x8_t)__p0, 9);
23227 return __ret;
23228 }
23229 #else
vrsqrte_f32(float32x2_t __p0)23230 __ai float32x2_t vrsqrte_f32(float32x2_t __p0) {
23231 float32x2_t __ret;
23232 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
23233 __ret = (float32x2_t) __builtin_neon_vrsqrte_v((int8x8_t)__rev0, 9);
23234 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
23235 return __ret;
23236 }
23237 #endif
23238
23239 #ifdef __LITTLE_ENDIAN__
vrsqrtsq_f32(float32x4_t __p0,float32x4_t __p1)23240 __ai float32x4_t vrsqrtsq_f32(float32x4_t __p0, float32x4_t __p1) {
23241 float32x4_t __ret;
23242 __ret = (float32x4_t) __builtin_neon_vrsqrtsq_v((int8x16_t)__p0, (int8x16_t)__p1, 41);
23243 return __ret;
23244 }
23245 #else
vrsqrtsq_f32(float32x4_t __p0,float32x4_t __p1)23246 __ai float32x4_t vrsqrtsq_f32(float32x4_t __p0, float32x4_t __p1) {
23247 float32x4_t __ret;
23248 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
23249 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
23250 __ret = (float32x4_t) __builtin_neon_vrsqrtsq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 41);
23251 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
23252 return __ret;
23253 }
23254 #endif
23255
23256 #ifdef __LITTLE_ENDIAN__
vrsqrts_f32(float32x2_t __p0,float32x2_t __p1)23257 __ai float32x2_t vrsqrts_f32(float32x2_t __p0, float32x2_t __p1) {
23258 float32x2_t __ret;
23259 __ret = (float32x2_t) __builtin_neon_vrsqrts_v((int8x8_t)__p0, (int8x8_t)__p1, 9);
23260 return __ret;
23261 }
23262 #else
vrsqrts_f32(float32x2_t __p0,float32x2_t __p1)23263 __ai float32x2_t vrsqrts_f32(float32x2_t __p0, float32x2_t __p1) {
23264 float32x2_t __ret;
23265 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
23266 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
23267 __ret = (float32x2_t) __builtin_neon_vrsqrts_v((int8x8_t)__rev0, (int8x8_t)__rev1, 9);
23268 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
23269 return __ret;
23270 }
23271 #endif
23272
23273 #ifdef __LITTLE_ENDIAN__
23274 #define vrsraq_n_u8(__p0, __p1, __p2) __extension__ ({ \
23275 uint8x16_t __ret; \
23276 uint8x16_t __s0 = __p0; \
23277 uint8x16_t __s1 = __p1; \
23278 __ret = (uint8x16_t) __builtin_neon_vrsraq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 48); \
23279 __ret; \
23280 })
23281 #else
23282 #define vrsraq_n_u8(__p0, __p1, __p2) __extension__ ({ \
23283 uint8x16_t __ret; \
23284 uint8x16_t __s0 = __p0; \
23285 uint8x16_t __s1 = __p1; \
23286 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
23287 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
23288 __ret = (uint8x16_t) __builtin_neon_vrsraq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 48); \
23289 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
23290 __ret; \
23291 })
23292 #endif
23293
23294 #ifdef __LITTLE_ENDIAN__
23295 #define vrsraq_n_u32(__p0, __p1, __p2) __extension__ ({ \
23296 uint32x4_t __ret; \
23297 uint32x4_t __s0 = __p0; \
23298 uint32x4_t __s1 = __p1; \
23299 __ret = (uint32x4_t) __builtin_neon_vrsraq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 50); \
23300 __ret; \
23301 })
23302 #else
23303 #define vrsraq_n_u32(__p0, __p1, __p2) __extension__ ({ \
23304 uint32x4_t __ret; \
23305 uint32x4_t __s0 = __p0; \
23306 uint32x4_t __s1 = __p1; \
23307 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
23308 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
23309 __ret = (uint32x4_t) __builtin_neon_vrsraq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 50); \
23310 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
23311 __ret; \
23312 })
23313 #endif
23314
23315 #ifdef __LITTLE_ENDIAN__
23316 #define vrsraq_n_u64(__p0, __p1, __p2) __extension__ ({ \
23317 uint64x2_t __ret; \
23318 uint64x2_t __s0 = __p0; \
23319 uint64x2_t __s1 = __p1; \
23320 __ret = (uint64x2_t) __builtin_neon_vrsraq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 51); \
23321 __ret; \
23322 })
23323 #else
23324 #define vrsraq_n_u64(__p0, __p1, __p2) __extension__ ({ \
23325 uint64x2_t __ret; \
23326 uint64x2_t __s0 = __p0; \
23327 uint64x2_t __s1 = __p1; \
23328 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
23329 uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
23330 __ret = (uint64x2_t) __builtin_neon_vrsraq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 51); \
23331 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
23332 __ret; \
23333 })
23334 #endif
23335
23336 #ifdef __LITTLE_ENDIAN__
23337 #define vrsraq_n_u16(__p0, __p1, __p2) __extension__ ({ \
23338 uint16x8_t __ret; \
23339 uint16x8_t __s0 = __p0; \
23340 uint16x8_t __s1 = __p1; \
23341 __ret = (uint16x8_t) __builtin_neon_vrsraq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 49); \
23342 __ret; \
23343 })
23344 #else
23345 #define vrsraq_n_u16(__p0, __p1, __p2) __extension__ ({ \
23346 uint16x8_t __ret; \
23347 uint16x8_t __s0 = __p0; \
23348 uint16x8_t __s1 = __p1; \
23349 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
23350 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
23351 __ret = (uint16x8_t) __builtin_neon_vrsraq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 49); \
23352 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
23353 __ret; \
23354 })
23355 #endif
23356
23357 #ifdef __LITTLE_ENDIAN__
23358 #define vrsraq_n_s8(__p0, __p1, __p2) __extension__ ({ \
23359 int8x16_t __ret; \
23360 int8x16_t __s0 = __p0; \
23361 int8x16_t __s1 = __p1; \
23362 __ret = (int8x16_t) __builtin_neon_vrsraq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 32); \
23363 __ret; \
23364 })
23365 #else
23366 #define vrsraq_n_s8(__p0, __p1, __p2) __extension__ ({ \
23367 int8x16_t __ret; \
23368 int8x16_t __s0 = __p0; \
23369 int8x16_t __s1 = __p1; \
23370 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
23371 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
23372 __ret = (int8x16_t) __builtin_neon_vrsraq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 32); \
23373 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
23374 __ret; \
23375 })
23376 #endif
23377
23378 #ifdef __LITTLE_ENDIAN__
23379 #define vrsraq_n_s32(__p0, __p1, __p2) __extension__ ({ \
23380 int32x4_t __ret; \
23381 int32x4_t __s0 = __p0; \
23382 int32x4_t __s1 = __p1; \
23383 __ret = (int32x4_t) __builtin_neon_vrsraq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 34); \
23384 __ret; \
23385 })
23386 #else
23387 #define vrsraq_n_s32(__p0, __p1, __p2) __extension__ ({ \
23388 int32x4_t __ret; \
23389 int32x4_t __s0 = __p0; \
23390 int32x4_t __s1 = __p1; \
23391 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
23392 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
23393 __ret = (int32x4_t) __builtin_neon_vrsraq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 34); \
23394 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
23395 __ret; \
23396 })
23397 #endif
23398
23399 #ifdef __LITTLE_ENDIAN__
23400 #define vrsraq_n_s64(__p0, __p1, __p2) __extension__ ({ \
23401 int64x2_t __ret; \
23402 int64x2_t __s0 = __p0; \
23403 int64x2_t __s1 = __p1; \
23404 __ret = (int64x2_t) __builtin_neon_vrsraq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 35); \
23405 __ret; \
23406 })
23407 #else
23408 #define vrsraq_n_s64(__p0, __p1, __p2) __extension__ ({ \
23409 int64x2_t __ret; \
23410 int64x2_t __s0 = __p0; \
23411 int64x2_t __s1 = __p1; \
23412 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
23413 int64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
23414 __ret = (int64x2_t) __builtin_neon_vrsraq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 35); \
23415 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
23416 __ret; \
23417 })
23418 #endif
23419
23420 #ifdef __LITTLE_ENDIAN__
23421 #define vrsraq_n_s16(__p0, __p1, __p2) __extension__ ({ \
23422 int16x8_t __ret; \
23423 int16x8_t __s0 = __p0; \
23424 int16x8_t __s1 = __p1; \
23425 __ret = (int16x8_t) __builtin_neon_vrsraq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 33); \
23426 __ret; \
23427 })
23428 #else
23429 #define vrsraq_n_s16(__p0, __p1, __p2) __extension__ ({ \
23430 int16x8_t __ret; \
23431 int16x8_t __s0 = __p0; \
23432 int16x8_t __s1 = __p1; \
23433 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
23434 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
23435 __ret = (int16x8_t) __builtin_neon_vrsraq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 33); \
23436 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
23437 __ret; \
23438 })
23439 #endif
23440
23441 #ifdef __LITTLE_ENDIAN__
23442 #define vrsra_n_u8(__p0, __p1, __p2) __extension__ ({ \
23443 uint8x8_t __ret; \
23444 uint8x8_t __s0 = __p0; \
23445 uint8x8_t __s1 = __p1; \
23446 __ret = (uint8x8_t) __builtin_neon_vrsra_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 16); \
23447 __ret; \
23448 })
23449 #else
23450 #define vrsra_n_u8(__p0, __p1, __p2) __extension__ ({ \
23451 uint8x8_t __ret; \
23452 uint8x8_t __s0 = __p0; \
23453 uint8x8_t __s1 = __p1; \
23454 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
23455 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
23456 __ret = (uint8x8_t) __builtin_neon_vrsra_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 16); \
23457 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
23458 __ret; \
23459 })
23460 #endif
23461
23462 #ifdef __LITTLE_ENDIAN__
23463 #define vrsra_n_u32(__p0, __p1, __p2) __extension__ ({ \
23464 uint32x2_t __ret; \
23465 uint32x2_t __s0 = __p0; \
23466 uint32x2_t __s1 = __p1; \
23467 __ret = (uint32x2_t) __builtin_neon_vrsra_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 18); \
23468 __ret; \
23469 })
23470 #else
23471 #define vrsra_n_u32(__p0, __p1, __p2) __extension__ ({ \
23472 uint32x2_t __ret; \
23473 uint32x2_t __s0 = __p0; \
23474 uint32x2_t __s1 = __p1; \
23475 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
23476 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
23477 __ret = (uint32x2_t) __builtin_neon_vrsra_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 18); \
23478 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
23479 __ret; \
23480 })
23481 #endif
23482
23483 #define vrsra_n_u64(__p0, __p1, __p2) __extension__ ({ \
23484 uint64x1_t __ret; \
23485 uint64x1_t __s0 = __p0; \
23486 uint64x1_t __s1 = __p1; \
23487 __ret = (uint64x1_t) __builtin_neon_vrsra_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 19); \
23488 __ret; \
23489 })
23490 #ifdef __LITTLE_ENDIAN__
23491 #define vrsra_n_u16(__p0, __p1, __p2) __extension__ ({ \
23492 uint16x4_t __ret; \
23493 uint16x4_t __s0 = __p0; \
23494 uint16x4_t __s1 = __p1; \
23495 __ret = (uint16x4_t) __builtin_neon_vrsra_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 17); \
23496 __ret; \
23497 })
23498 #else
23499 #define vrsra_n_u16(__p0, __p1, __p2) __extension__ ({ \
23500 uint16x4_t __ret; \
23501 uint16x4_t __s0 = __p0; \
23502 uint16x4_t __s1 = __p1; \
23503 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
23504 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
23505 __ret = (uint16x4_t) __builtin_neon_vrsra_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 17); \
23506 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
23507 __ret; \
23508 })
23509 #endif
23510
23511 #ifdef __LITTLE_ENDIAN__
23512 #define vrsra_n_s8(__p0, __p1, __p2) __extension__ ({ \
23513 int8x8_t __ret; \
23514 int8x8_t __s0 = __p0; \
23515 int8x8_t __s1 = __p1; \
23516 __ret = (int8x8_t) __builtin_neon_vrsra_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 0); \
23517 __ret; \
23518 })
23519 #else
23520 #define vrsra_n_s8(__p0, __p1, __p2) __extension__ ({ \
23521 int8x8_t __ret; \
23522 int8x8_t __s0 = __p0; \
23523 int8x8_t __s1 = __p1; \
23524 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
23525 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
23526 __ret = (int8x8_t) __builtin_neon_vrsra_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 0); \
23527 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
23528 __ret; \
23529 })
23530 #endif
23531
23532 #ifdef __LITTLE_ENDIAN__
23533 #define vrsra_n_s32(__p0, __p1, __p2) __extension__ ({ \
23534 int32x2_t __ret; \
23535 int32x2_t __s0 = __p0; \
23536 int32x2_t __s1 = __p1; \
23537 __ret = (int32x2_t) __builtin_neon_vrsra_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 2); \
23538 __ret; \
23539 })
23540 #else
23541 #define vrsra_n_s32(__p0, __p1, __p2) __extension__ ({ \
23542 int32x2_t __ret; \
23543 int32x2_t __s0 = __p0; \
23544 int32x2_t __s1 = __p1; \
23545 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
23546 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
23547 __ret = (int32x2_t) __builtin_neon_vrsra_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 2); \
23548 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
23549 __ret; \
23550 })
23551 #endif
23552
23553 #define vrsra_n_s64(__p0, __p1, __p2) __extension__ ({ \
23554 int64x1_t __ret; \
23555 int64x1_t __s0 = __p0; \
23556 int64x1_t __s1 = __p1; \
23557 __ret = (int64x1_t) __builtin_neon_vrsra_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 3); \
23558 __ret; \
23559 })
23560 #ifdef __LITTLE_ENDIAN__
23561 #define vrsra_n_s16(__p0, __p1, __p2) __extension__ ({ \
23562 int16x4_t __ret; \
23563 int16x4_t __s0 = __p0; \
23564 int16x4_t __s1 = __p1; \
23565 __ret = (int16x4_t) __builtin_neon_vrsra_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 1); \
23566 __ret; \
23567 })
23568 #else
23569 #define vrsra_n_s16(__p0, __p1, __p2) __extension__ ({ \
23570 int16x4_t __ret; \
23571 int16x4_t __s0 = __p0; \
23572 int16x4_t __s1 = __p1; \
23573 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
23574 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
23575 __ret = (int16x4_t) __builtin_neon_vrsra_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 1); \
23576 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
23577 __ret; \
23578 })
23579 #endif
23580
23581 #ifdef __LITTLE_ENDIAN__
vrsubhn_u32(uint32x4_t __p0,uint32x4_t __p1)23582 __ai uint16x4_t vrsubhn_u32(uint32x4_t __p0, uint32x4_t __p1) {
23583 uint16x4_t __ret;
23584 __ret = (uint16x4_t) __builtin_neon_vrsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 17);
23585 return __ret;
23586 }
23587 #else
vrsubhn_u32(uint32x4_t __p0,uint32x4_t __p1)23588 __ai uint16x4_t vrsubhn_u32(uint32x4_t __p0, uint32x4_t __p1) {
23589 uint16x4_t __ret;
23590 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
23591 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
23592 __ret = (uint16x4_t) __builtin_neon_vrsubhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 17);
23593 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
23594 return __ret;
23595 }
__noswap_vrsubhn_u32(uint32x4_t __p0,uint32x4_t __p1)23596 __ai uint16x4_t __noswap_vrsubhn_u32(uint32x4_t __p0, uint32x4_t __p1) {
23597 uint16x4_t __ret;
23598 __ret = (uint16x4_t) __builtin_neon_vrsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 17);
23599 return __ret;
23600 }
23601 #endif
23602
23603 #ifdef __LITTLE_ENDIAN__
vrsubhn_u64(uint64x2_t __p0,uint64x2_t __p1)23604 __ai uint32x2_t vrsubhn_u64(uint64x2_t __p0, uint64x2_t __p1) {
23605 uint32x2_t __ret;
23606 __ret = (uint32x2_t) __builtin_neon_vrsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 18);
23607 return __ret;
23608 }
23609 #else
vrsubhn_u64(uint64x2_t __p0,uint64x2_t __p1)23610 __ai uint32x2_t vrsubhn_u64(uint64x2_t __p0, uint64x2_t __p1) {
23611 uint32x2_t __ret;
23612 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
23613 uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
23614 __ret = (uint32x2_t) __builtin_neon_vrsubhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 18);
23615 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
23616 return __ret;
23617 }
__noswap_vrsubhn_u64(uint64x2_t __p0,uint64x2_t __p1)23618 __ai uint32x2_t __noswap_vrsubhn_u64(uint64x2_t __p0, uint64x2_t __p1) {
23619 uint32x2_t __ret;
23620 __ret = (uint32x2_t) __builtin_neon_vrsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 18);
23621 return __ret;
23622 }
23623 #endif
23624
23625 #ifdef __LITTLE_ENDIAN__
vrsubhn_u16(uint16x8_t __p0,uint16x8_t __p1)23626 __ai uint8x8_t vrsubhn_u16(uint16x8_t __p0, uint16x8_t __p1) {
23627 uint8x8_t __ret;
23628 __ret = (uint8x8_t) __builtin_neon_vrsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 16);
23629 return __ret;
23630 }
23631 #else
vrsubhn_u16(uint16x8_t __p0,uint16x8_t __p1)23632 __ai uint8x8_t vrsubhn_u16(uint16x8_t __p0, uint16x8_t __p1) {
23633 uint8x8_t __ret;
23634 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
23635 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
23636 __ret = (uint8x8_t) __builtin_neon_vrsubhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 16);
23637 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
23638 return __ret;
23639 }
__noswap_vrsubhn_u16(uint16x8_t __p0,uint16x8_t __p1)23640 __ai uint8x8_t __noswap_vrsubhn_u16(uint16x8_t __p0, uint16x8_t __p1) {
23641 uint8x8_t __ret;
23642 __ret = (uint8x8_t) __builtin_neon_vrsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 16);
23643 return __ret;
23644 }
23645 #endif
23646
23647 #ifdef __LITTLE_ENDIAN__
vrsubhn_s32(int32x4_t __p0,int32x4_t __p1)23648 __ai int16x4_t vrsubhn_s32(int32x4_t __p0, int32x4_t __p1) {
23649 int16x4_t __ret;
23650 __ret = (int16x4_t) __builtin_neon_vrsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 1);
23651 return __ret;
23652 }
23653 #else
vrsubhn_s32(int32x4_t __p0,int32x4_t __p1)23654 __ai int16x4_t vrsubhn_s32(int32x4_t __p0, int32x4_t __p1) {
23655 int16x4_t __ret;
23656 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
23657 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
23658 __ret = (int16x4_t) __builtin_neon_vrsubhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 1);
23659 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
23660 return __ret;
23661 }
__noswap_vrsubhn_s32(int32x4_t __p0,int32x4_t __p1)23662 __ai int16x4_t __noswap_vrsubhn_s32(int32x4_t __p0, int32x4_t __p1) {
23663 int16x4_t __ret;
23664 __ret = (int16x4_t) __builtin_neon_vrsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 1);
23665 return __ret;
23666 }
23667 #endif
23668
23669 #ifdef __LITTLE_ENDIAN__
vrsubhn_s64(int64x2_t __p0,int64x2_t __p1)23670 __ai int32x2_t vrsubhn_s64(int64x2_t __p0, int64x2_t __p1) {
23671 int32x2_t __ret;
23672 __ret = (int32x2_t) __builtin_neon_vrsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 2);
23673 return __ret;
23674 }
23675 #else
vrsubhn_s64(int64x2_t __p0,int64x2_t __p1)23676 __ai int32x2_t vrsubhn_s64(int64x2_t __p0, int64x2_t __p1) {
23677 int32x2_t __ret;
23678 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
23679 int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
23680 __ret = (int32x2_t) __builtin_neon_vrsubhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 2);
23681 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
23682 return __ret;
23683 }
__noswap_vrsubhn_s64(int64x2_t __p0,int64x2_t __p1)23684 __ai int32x2_t __noswap_vrsubhn_s64(int64x2_t __p0, int64x2_t __p1) {
23685 int32x2_t __ret;
23686 __ret = (int32x2_t) __builtin_neon_vrsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 2);
23687 return __ret;
23688 }
23689 #endif
23690
23691 #ifdef __LITTLE_ENDIAN__
vrsubhn_s16(int16x8_t __p0,int16x8_t __p1)23692 __ai int8x8_t vrsubhn_s16(int16x8_t __p0, int16x8_t __p1) {
23693 int8x8_t __ret;
23694 __ret = (int8x8_t) __builtin_neon_vrsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 0);
23695 return __ret;
23696 }
23697 #else
vrsubhn_s16(int16x8_t __p0,int16x8_t __p1)23698 __ai int8x8_t vrsubhn_s16(int16x8_t __p0, int16x8_t __p1) {
23699 int8x8_t __ret;
23700 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
23701 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
23702 __ret = (int8x8_t) __builtin_neon_vrsubhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 0);
23703 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
23704 return __ret;
23705 }
__noswap_vrsubhn_s16(int16x8_t __p0,int16x8_t __p1)23706 __ai int8x8_t __noswap_vrsubhn_s16(int16x8_t __p0, int16x8_t __p1) {
23707 int8x8_t __ret;
23708 __ret = (int8x8_t) __builtin_neon_vrsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 0);
23709 return __ret;
23710 }
23711 #endif
23712
23713 #ifdef __LITTLE_ENDIAN__
23714 #define vset_lane_p8(__p0, __p1, __p2) __extension__ ({ \
23715 poly8x8_t __ret; \
23716 poly8_t __s0 = __p0; \
23717 poly8x8_t __s1 = __p1; \
23718 __ret = (poly8x8_t) __builtin_neon_vset_lane_i8(__s0, (poly8x8_t)__s1, __p2); \
23719 __ret; \
23720 })
23721 #else
23722 #define vset_lane_p8(__p0, __p1, __p2) __extension__ ({ \
23723 poly8x8_t __ret; \
23724 poly8_t __s0 = __p0; \
23725 poly8x8_t __s1 = __p1; \
23726 poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
23727 __ret = (poly8x8_t) __builtin_neon_vset_lane_i8(__s0, (poly8x8_t)__rev1, __p2); \
23728 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
23729 __ret; \
23730 })
23731 #define __noswap_vset_lane_p8(__p0, __p1, __p2) __extension__ ({ \
23732 poly8x8_t __ret; \
23733 poly8_t __s0 = __p0; \
23734 poly8x8_t __s1 = __p1; \
23735 __ret = (poly8x8_t) __builtin_neon_vset_lane_i8(__s0, (poly8x8_t)__s1, __p2); \
23736 __ret; \
23737 })
23738 #endif
23739
23740 #ifdef __LITTLE_ENDIAN__
23741 #define vset_lane_p16(__p0, __p1, __p2) __extension__ ({ \
23742 poly16x4_t __ret; \
23743 poly16_t __s0 = __p0; \
23744 poly16x4_t __s1 = __p1; \
23745 __ret = (poly16x4_t) __builtin_neon_vset_lane_i16(__s0, (poly16x4_t)__s1, __p2); \
23746 __ret; \
23747 })
23748 #else
23749 #define vset_lane_p16(__p0, __p1, __p2) __extension__ ({ \
23750 poly16x4_t __ret; \
23751 poly16_t __s0 = __p0; \
23752 poly16x4_t __s1 = __p1; \
23753 poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
23754 __ret = (poly16x4_t) __builtin_neon_vset_lane_i16(__s0, (poly16x4_t)__rev1, __p2); \
23755 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
23756 __ret; \
23757 })
23758 #define __noswap_vset_lane_p16(__p0, __p1, __p2) __extension__ ({ \
23759 poly16x4_t __ret; \
23760 poly16_t __s0 = __p0; \
23761 poly16x4_t __s1 = __p1; \
23762 __ret = (poly16x4_t) __builtin_neon_vset_lane_i16(__s0, (poly16x4_t)__s1, __p2); \
23763 __ret; \
23764 })
23765 #endif
23766
23767 #ifdef __LITTLE_ENDIAN__
23768 #define vsetq_lane_p8(__p0, __p1, __p2) __extension__ ({ \
23769 poly8x16_t __ret; \
23770 poly8_t __s0 = __p0; \
23771 poly8x16_t __s1 = __p1; \
23772 __ret = (poly8x16_t) __builtin_neon_vsetq_lane_i8(__s0, (poly8x16_t)__s1, __p2); \
23773 __ret; \
23774 })
23775 #else
23776 #define vsetq_lane_p8(__p0, __p1, __p2) __extension__ ({ \
23777 poly8x16_t __ret; \
23778 poly8_t __s0 = __p0; \
23779 poly8x16_t __s1 = __p1; \
23780 poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
23781 __ret = (poly8x16_t) __builtin_neon_vsetq_lane_i8(__s0, (poly8x16_t)__rev1, __p2); \
23782 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
23783 __ret; \
23784 })
23785 #define __noswap_vsetq_lane_p8(__p0, __p1, __p2) __extension__ ({ \
23786 poly8x16_t __ret; \
23787 poly8_t __s0 = __p0; \
23788 poly8x16_t __s1 = __p1; \
23789 __ret = (poly8x16_t) __builtin_neon_vsetq_lane_i8(__s0, (poly8x16_t)__s1, __p2); \
23790 __ret; \
23791 })
23792 #endif
23793
23794 #ifdef __LITTLE_ENDIAN__
23795 #define vsetq_lane_p16(__p0, __p1, __p2) __extension__ ({ \
23796 poly16x8_t __ret; \
23797 poly16_t __s0 = __p0; \
23798 poly16x8_t __s1 = __p1; \
23799 __ret = (poly16x8_t) __builtin_neon_vsetq_lane_i16(__s0, (poly16x8_t)__s1, __p2); \
23800 __ret; \
23801 })
23802 #else
23803 #define vsetq_lane_p16(__p0, __p1, __p2) __extension__ ({ \
23804 poly16x8_t __ret; \
23805 poly16_t __s0 = __p0; \
23806 poly16x8_t __s1 = __p1; \
23807 poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
23808 __ret = (poly16x8_t) __builtin_neon_vsetq_lane_i16(__s0, (poly16x8_t)__rev1, __p2); \
23809 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
23810 __ret; \
23811 })
23812 #define __noswap_vsetq_lane_p16(__p0, __p1, __p2) __extension__ ({ \
23813 poly16x8_t __ret; \
23814 poly16_t __s0 = __p0; \
23815 poly16x8_t __s1 = __p1; \
23816 __ret = (poly16x8_t) __builtin_neon_vsetq_lane_i16(__s0, (poly16x8_t)__s1, __p2); \
23817 __ret; \
23818 })
23819 #endif
23820
23821 #ifdef __LITTLE_ENDIAN__
23822 #define vsetq_lane_u8(__p0, __p1, __p2) __extension__ ({ \
23823 uint8x16_t __ret; \
23824 uint8_t __s0 = __p0; \
23825 uint8x16_t __s1 = __p1; \
23826 __ret = (uint8x16_t) __builtin_neon_vsetq_lane_i8(__s0, (int8x16_t)__s1, __p2); \
23827 __ret; \
23828 })
23829 #else
23830 #define vsetq_lane_u8(__p0, __p1, __p2) __extension__ ({ \
23831 uint8x16_t __ret; \
23832 uint8_t __s0 = __p0; \
23833 uint8x16_t __s1 = __p1; \
23834 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
23835 __ret = (uint8x16_t) __builtin_neon_vsetq_lane_i8(__s0, (int8x16_t)__rev1, __p2); \
23836 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
23837 __ret; \
23838 })
23839 #define __noswap_vsetq_lane_u8(__p0, __p1, __p2) __extension__ ({ \
23840 uint8x16_t __ret; \
23841 uint8_t __s0 = __p0; \
23842 uint8x16_t __s1 = __p1; \
23843 __ret = (uint8x16_t) __builtin_neon_vsetq_lane_i8(__s0, (int8x16_t)__s1, __p2); \
23844 __ret; \
23845 })
23846 #endif
23847
23848 #ifdef __LITTLE_ENDIAN__
23849 #define vsetq_lane_u32(__p0, __p1, __p2) __extension__ ({ \
23850 uint32x4_t __ret; \
23851 uint32_t __s0 = __p0; \
23852 uint32x4_t __s1 = __p1; \
23853 __ret = (uint32x4_t) __builtin_neon_vsetq_lane_i32(__s0, (int32x4_t)__s1, __p2); \
23854 __ret; \
23855 })
23856 #else
23857 #define vsetq_lane_u32(__p0, __p1, __p2) __extension__ ({ \
23858 uint32x4_t __ret; \
23859 uint32_t __s0 = __p0; \
23860 uint32x4_t __s1 = __p1; \
23861 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
23862 __ret = (uint32x4_t) __builtin_neon_vsetq_lane_i32(__s0, (int32x4_t)__rev1, __p2); \
23863 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
23864 __ret; \
23865 })
23866 #define __noswap_vsetq_lane_u32(__p0, __p1, __p2) __extension__ ({ \
23867 uint32x4_t __ret; \
23868 uint32_t __s0 = __p0; \
23869 uint32x4_t __s1 = __p1; \
23870 __ret = (uint32x4_t) __builtin_neon_vsetq_lane_i32(__s0, (int32x4_t)__s1, __p2); \
23871 __ret; \
23872 })
23873 #endif
23874
23875 #ifdef __LITTLE_ENDIAN__
23876 #define vsetq_lane_u64(__p0, __p1, __p2) __extension__ ({ \
23877 uint64x2_t __ret; \
23878 uint64_t __s0 = __p0; \
23879 uint64x2_t __s1 = __p1; \
23880 __ret = (uint64x2_t) __builtin_neon_vsetq_lane_i64(__s0, (int64x2_t)__s1, __p2); \
23881 __ret; \
23882 })
23883 #else
23884 #define vsetq_lane_u64(__p0, __p1, __p2) __extension__ ({ \
23885 uint64x2_t __ret; \
23886 uint64_t __s0 = __p0; \
23887 uint64x2_t __s1 = __p1; \
23888 uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
23889 __ret = (uint64x2_t) __builtin_neon_vsetq_lane_i64(__s0, (int64x2_t)__rev1, __p2); \
23890 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
23891 __ret; \
23892 })
23893 #define __noswap_vsetq_lane_u64(__p0, __p1, __p2) __extension__ ({ \
23894 uint64x2_t __ret; \
23895 uint64_t __s0 = __p0; \
23896 uint64x2_t __s1 = __p1; \
23897 __ret = (uint64x2_t) __builtin_neon_vsetq_lane_i64(__s0, (int64x2_t)__s1, __p2); \
23898 __ret; \
23899 })
23900 #endif
23901
23902 #ifdef __LITTLE_ENDIAN__
23903 #define vsetq_lane_u16(__p0, __p1, __p2) __extension__ ({ \
23904 uint16x8_t __ret; \
23905 uint16_t __s0 = __p0; \
23906 uint16x8_t __s1 = __p1; \
23907 __ret = (uint16x8_t) __builtin_neon_vsetq_lane_i16(__s0, (int16x8_t)__s1, __p2); \
23908 __ret; \
23909 })
23910 #else
23911 #define vsetq_lane_u16(__p0, __p1, __p2) __extension__ ({ \
23912 uint16x8_t __ret; \
23913 uint16_t __s0 = __p0; \
23914 uint16x8_t __s1 = __p1; \
23915 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
23916 __ret = (uint16x8_t) __builtin_neon_vsetq_lane_i16(__s0, (int16x8_t)__rev1, __p2); \
23917 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
23918 __ret; \
23919 })
23920 #define __noswap_vsetq_lane_u16(__p0, __p1, __p2) __extension__ ({ \
23921 uint16x8_t __ret; \
23922 uint16_t __s0 = __p0; \
23923 uint16x8_t __s1 = __p1; \
23924 __ret = (uint16x8_t) __builtin_neon_vsetq_lane_i16(__s0, (int16x8_t)__s1, __p2); \
23925 __ret; \
23926 })
23927 #endif
23928
23929 #ifdef __LITTLE_ENDIAN__
23930 #define vsetq_lane_s8(__p0, __p1, __p2) __extension__ ({ \
23931 int8x16_t __ret; \
23932 int8_t __s0 = __p0; \
23933 int8x16_t __s1 = __p1; \
23934 __ret = (int8x16_t) __builtin_neon_vsetq_lane_i8(__s0, (int8x16_t)__s1, __p2); \
23935 __ret; \
23936 })
23937 #else
23938 #define vsetq_lane_s8(__p0, __p1, __p2) __extension__ ({ \
23939 int8x16_t __ret; \
23940 int8_t __s0 = __p0; \
23941 int8x16_t __s1 = __p1; \
23942 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
23943 __ret = (int8x16_t) __builtin_neon_vsetq_lane_i8(__s0, (int8x16_t)__rev1, __p2); \
23944 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
23945 __ret; \
23946 })
23947 #define __noswap_vsetq_lane_s8(__p0, __p1, __p2) __extension__ ({ \
23948 int8x16_t __ret; \
23949 int8_t __s0 = __p0; \
23950 int8x16_t __s1 = __p1; \
23951 __ret = (int8x16_t) __builtin_neon_vsetq_lane_i8(__s0, (int8x16_t)__s1, __p2); \
23952 __ret; \
23953 })
23954 #endif
23955
23956 #ifdef __LITTLE_ENDIAN__
23957 #define vsetq_lane_f32(__p0, __p1, __p2) __extension__ ({ \
23958 float32x4_t __ret; \
23959 float32_t __s0 = __p0; \
23960 float32x4_t __s1 = __p1; \
23961 __ret = (float32x4_t) __builtin_neon_vsetq_lane_f32(__s0, (float32x4_t)__s1, __p2); \
23962 __ret; \
23963 })
23964 #else
23965 #define vsetq_lane_f32(__p0, __p1, __p2) __extension__ ({ \
23966 float32x4_t __ret; \
23967 float32_t __s0 = __p0; \
23968 float32x4_t __s1 = __p1; \
23969 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
23970 __ret = (float32x4_t) __builtin_neon_vsetq_lane_f32(__s0, (float32x4_t)__rev1, __p2); \
23971 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
23972 __ret; \
23973 })
23974 #define __noswap_vsetq_lane_f32(__p0, __p1, __p2) __extension__ ({ \
23975 float32x4_t __ret; \
23976 float32_t __s0 = __p0; \
23977 float32x4_t __s1 = __p1; \
23978 __ret = (float32x4_t) __builtin_neon_vsetq_lane_f32(__s0, (float32x4_t)__s1, __p2); \
23979 __ret; \
23980 })
23981 #endif
23982
23983 #ifdef __LITTLE_ENDIAN__
23984 #define vsetq_lane_s32(__p0, __p1, __p2) __extension__ ({ \
23985 int32x4_t __ret; \
23986 int32_t __s0 = __p0; \
23987 int32x4_t __s1 = __p1; \
23988 __ret = (int32x4_t) __builtin_neon_vsetq_lane_i32(__s0, (int32x4_t)__s1, __p2); \
23989 __ret; \
23990 })
23991 #else
23992 #define vsetq_lane_s32(__p0, __p1, __p2) __extension__ ({ \
23993 int32x4_t __ret; \
23994 int32_t __s0 = __p0; \
23995 int32x4_t __s1 = __p1; \
23996 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
23997 __ret = (int32x4_t) __builtin_neon_vsetq_lane_i32(__s0, (int32x4_t)__rev1, __p2); \
23998 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
23999 __ret; \
24000 })
24001 #define __noswap_vsetq_lane_s32(__p0, __p1, __p2) __extension__ ({ \
24002 int32x4_t __ret; \
24003 int32_t __s0 = __p0; \
24004 int32x4_t __s1 = __p1; \
24005 __ret = (int32x4_t) __builtin_neon_vsetq_lane_i32(__s0, (int32x4_t)__s1, __p2); \
24006 __ret; \
24007 })
24008 #endif
24009
24010 #ifdef __LITTLE_ENDIAN__
24011 #define vsetq_lane_s64(__p0, __p1, __p2) __extension__ ({ \
24012 int64x2_t __ret; \
24013 int64_t __s0 = __p0; \
24014 int64x2_t __s1 = __p1; \
24015 __ret = (int64x2_t) __builtin_neon_vsetq_lane_i64(__s0, (int64x2_t)__s1, __p2); \
24016 __ret; \
24017 })
24018 #else
24019 #define vsetq_lane_s64(__p0, __p1, __p2) __extension__ ({ \
24020 int64x2_t __ret; \
24021 int64_t __s0 = __p0; \
24022 int64x2_t __s1 = __p1; \
24023 int64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
24024 __ret = (int64x2_t) __builtin_neon_vsetq_lane_i64(__s0, (int64x2_t)__rev1, __p2); \
24025 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
24026 __ret; \
24027 })
24028 #define __noswap_vsetq_lane_s64(__p0, __p1, __p2) __extension__ ({ \
24029 int64x2_t __ret; \
24030 int64_t __s0 = __p0; \
24031 int64x2_t __s1 = __p1; \
24032 __ret = (int64x2_t) __builtin_neon_vsetq_lane_i64(__s0, (int64x2_t)__s1, __p2); \
24033 __ret; \
24034 })
24035 #endif
24036
24037 #ifdef __LITTLE_ENDIAN__
24038 #define vsetq_lane_s16(__p0, __p1, __p2) __extension__ ({ \
24039 int16x8_t __ret; \
24040 int16_t __s0 = __p0; \
24041 int16x8_t __s1 = __p1; \
24042 __ret = (int16x8_t) __builtin_neon_vsetq_lane_i16(__s0, (int16x8_t)__s1, __p2); \
24043 __ret; \
24044 })
24045 #else
24046 #define vsetq_lane_s16(__p0, __p1, __p2) __extension__ ({ \
24047 int16x8_t __ret; \
24048 int16_t __s0 = __p0; \
24049 int16x8_t __s1 = __p1; \
24050 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
24051 __ret = (int16x8_t) __builtin_neon_vsetq_lane_i16(__s0, (int16x8_t)__rev1, __p2); \
24052 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
24053 __ret; \
24054 })
24055 #define __noswap_vsetq_lane_s16(__p0, __p1, __p2) __extension__ ({ \
24056 int16x8_t __ret; \
24057 int16_t __s0 = __p0; \
24058 int16x8_t __s1 = __p1; \
24059 __ret = (int16x8_t) __builtin_neon_vsetq_lane_i16(__s0, (int16x8_t)__s1, __p2); \
24060 __ret; \
24061 })
24062 #endif
24063
24064 #ifdef __LITTLE_ENDIAN__
24065 #define vset_lane_u8(__p0, __p1, __p2) __extension__ ({ \
24066 uint8x8_t __ret; \
24067 uint8_t __s0 = __p0; \
24068 uint8x8_t __s1 = __p1; \
24069 __ret = (uint8x8_t) __builtin_neon_vset_lane_i8(__s0, (int8x8_t)__s1, __p2); \
24070 __ret; \
24071 })
24072 #else
24073 #define vset_lane_u8(__p0, __p1, __p2) __extension__ ({ \
24074 uint8x8_t __ret; \
24075 uint8_t __s0 = __p0; \
24076 uint8x8_t __s1 = __p1; \
24077 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
24078 __ret = (uint8x8_t) __builtin_neon_vset_lane_i8(__s0, (int8x8_t)__rev1, __p2); \
24079 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
24080 __ret; \
24081 })
24082 #define __noswap_vset_lane_u8(__p0, __p1, __p2) __extension__ ({ \
24083 uint8x8_t __ret; \
24084 uint8_t __s0 = __p0; \
24085 uint8x8_t __s1 = __p1; \
24086 __ret = (uint8x8_t) __builtin_neon_vset_lane_i8(__s0, (int8x8_t)__s1, __p2); \
24087 __ret; \
24088 })
24089 #endif
24090
24091 #ifdef __LITTLE_ENDIAN__
24092 #define vset_lane_u32(__p0, __p1, __p2) __extension__ ({ \
24093 uint32x2_t __ret; \
24094 uint32_t __s0 = __p0; \
24095 uint32x2_t __s1 = __p1; \
24096 __ret = (uint32x2_t) __builtin_neon_vset_lane_i32(__s0, (int32x2_t)__s1, __p2); \
24097 __ret; \
24098 })
24099 #else
24100 #define vset_lane_u32(__p0, __p1, __p2) __extension__ ({ \
24101 uint32x2_t __ret; \
24102 uint32_t __s0 = __p0; \
24103 uint32x2_t __s1 = __p1; \
24104 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
24105 __ret = (uint32x2_t) __builtin_neon_vset_lane_i32(__s0, (int32x2_t)__rev1, __p2); \
24106 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
24107 __ret; \
24108 })
24109 #define __noswap_vset_lane_u32(__p0, __p1, __p2) __extension__ ({ \
24110 uint32x2_t __ret; \
24111 uint32_t __s0 = __p0; \
24112 uint32x2_t __s1 = __p1; \
24113 __ret = (uint32x2_t) __builtin_neon_vset_lane_i32(__s0, (int32x2_t)__s1, __p2); \
24114 __ret; \
24115 })
24116 #endif
24117
24118 #define vset_lane_u64(__p0, __p1, __p2) __extension__ ({ \
24119 uint64x1_t __ret; \
24120 uint64_t __s0 = __p0; \
24121 uint64x1_t __s1 = __p1; \
24122 __ret = (uint64x1_t) __builtin_neon_vset_lane_i64(__s0, (int64x1_t)__s1, __p2); \
24123 __ret; \
24124 })
24125 #ifdef __LITTLE_ENDIAN__
24126 #define vset_lane_u16(__p0, __p1, __p2) __extension__ ({ \
24127 uint16x4_t __ret; \
24128 uint16_t __s0 = __p0; \
24129 uint16x4_t __s1 = __p1; \
24130 __ret = (uint16x4_t) __builtin_neon_vset_lane_i16(__s0, (int16x4_t)__s1, __p2); \
24131 __ret; \
24132 })
24133 #else
24134 #define vset_lane_u16(__p0, __p1, __p2) __extension__ ({ \
24135 uint16x4_t __ret; \
24136 uint16_t __s0 = __p0; \
24137 uint16x4_t __s1 = __p1; \
24138 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
24139 __ret = (uint16x4_t) __builtin_neon_vset_lane_i16(__s0, (int16x4_t)__rev1, __p2); \
24140 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
24141 __ret; \
24142 })
24143 #define __noswap_vset_lane_u16(__p0, __p1, __p2) __extension__ ({ \
24144 uint16x4_t __ret; \
24145 uint16_t __s0 = __p0; \
24146 uint16x4_t __s1 = __p1; \
24147 __ret = (uint16x4_t) __builtin_neon_vset_lane_i16(__s0, (int16x4_t)__s1, __p2); \
24148 __ret; \
24149 })
24150 #endif
24151
24152 #ifdef __LITTLE_ENDIAN__
24153 #define vset_lane_s8(__p0, __p1, __p2) __extension__ ({ \
24154 int8x8_t __ret; \
24155 int8_t __s0 = __p0; \
24156 int8x8_t __s1 = __p1; \
24157 __ret = (int8x8_t) __builtin_neon_vset_lane_i8(__s0, (int8x8_t)__s1, __p2); \
24158 __ret; \
24159 })
24160 #else
24161 #define vset_lane_s8(__p0, __p1, __p2) __extension__ ({ \
24162 int8x8_t __ret; \
24163 int8_t __s0 = __p0; \
24164 int8x8_t __s1 = __p1; \
24165 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
24166 __ret = (int8x8_t) __builtin_neon_vset_lane_i8(__s0, (int8x8_t)__rev1, __p2); \
24167 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
24168 __ret; \
24169 })
24170 #define __noswap_vset_lane_s8(__p0, __p1, __p2) __extension__ ({ \
24171 int8x8_t __ret; \
24172 int8_t __s0 = __p0; \
24173 int8x8_t __s1 = __p1; \
24174 __ret = (int8x8_t) __builtin_neon_vset_lane_i8(__s0, (int8x8_t)__s1, __p2); \
24175 __ret; \
24176 })
24177 #endif
24178
24179 #ifdef __LITTLE_ENDIAN__
24180 #define vset_lane_f32(__p0, __p1, __p2) __extension__ ({ \
24181 float32x2_t __ret; \
24182 float32_t __s0 = __p0; \
24183 float32x2_t __s1 = __p1; \
24184 __ret = (float32x2_t) __builtin_neon_vset_lane_f32(__s0, (float32x2_t)__s1, __p2); \
24185 __ret; \
24186 })
24187 #else
24188 #define vset_lane_f32(__p0, __p1, __p2) __extension__ ({ \
24189 float32x2_t __ret; \
24190 float32_t __s0 = __p0; \
24191 float32x2_t __s1 = __p1; \
24192 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
24193 __ret = (float32x2_t) __builtin_neon_vset_lane_f32(__s0, (float32x2_t)__rev1, __p2); \
24194 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
24195 __ret; \
24196 })
24197 #define __noswap_vset_lane_f32(__p0, __p1, __p2) __extension__ ({ \
24198 float32x2_t __ret; \
24199 float32_t __s0 = __p0; \
24200 float32x2_t __s1 = __p1; \
24201 __ret = (float32x2_t) __builtin_neon_vset_lane_f32(__s0, (float32x2_t)__s1, __p2); \
24202 __ret; \
24203 })
24204 #endif
24205
24206 #ifdef __LITTLE_ENDIAN__
24207 #define vset_lane_s32(__p0, __p1, __p2) __extension__ ({ \
24208 int32x2_t __ret; \
24209 int32_t __s0 = __p0; \
24210 int32x2_t __s1 = __p1; \
24211 __ret = (int32x2_t) __builtin_neon_vset_lane_i32(__s0, (int32x2_t)__s1, __p2); \
24212 __ret; \
24213 })
24214 #else
24215 #define vset_lane_s32(__p0, __p1, __p2) __extension__ ({ \
24216 int32x2_t __ret; \
24217 int32_t __s0 = __p0; \
24218 int32x2_t __s1 = __p1; \
24219 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
24220 __ret = (int32x2_t) __builtin_neon_vset_lane_i32(__s0, (int32x2_t)__rev1, __p2); \
24221 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
24222 __ret; \
24223 })
24224 #define __noswap_vset_lane_s32(__p0, __p1, __p2) __extension__ ({ \
24225 int32x2_t __ret; \
24226 int32_t __s0 = __p0; \
24227 int32x2_t __s1 = __p1; \
24228 __ret = (int32x2_t) __builtin_neon_vset_lane_i32(__s0, (int32x2_t)__s1, __p2); \
24229 __ret; \
24230 })
24231 #endif
24232
24233 #define vset_lane_s64(__p0, __p1, __p2) __extension__ ({ \
24234 int64x1_t __ret; \
24235 int64_t __s0 = __p0; \
24236 int64x1_t __s1 = __p1; \
24237 __ret = (int64x1_t) __builtin_neon_vset_lane_i64(__s0, (int64x1_t)__s1, __p2); \
24238 __ret; \
24239 })
24240 #ifdef __LITTLE_ENDIAN__
24241 #define vset_lane_s16(__p0, __p1, __p2) __extension__ ({ \
24242 int16x4_t __ret; \
24243 int16_t __s0 = __p0; \
24244 int16x4_t __s1 = __p1; \
24245 __ret = (int16x4_t) __builtin_neon_vset_lane_i16(__s0, (int16x4_t)__s1, __p2); \
24246 __ret; \
24247 })
24248 #else
24249 #define vset_lane_s16(__p0, __p1, __p2) __extension__ ({ \
24250 int16x4_t __ret; \
24251 int16_t __s0 = __p0; \
24252 int16x4_t __s1 = __p1; \
24253 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
24254 __ret = (int16x4_t) __builtin_neon_vset_lane_i16(__s0, (int16x4_t)__rev1, __p2); \
24255 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
24256 __ret; \
24257 })
24258 #define __noswap_vset_lane_s16(__p0, __p1, __p2) __extension__ ({ \
24259 int16x4_t __ret; \
24260 int16_t __s0 = __p0; \
24261 int16x4_t __s1 = __p1; \
24262 __ret = (int16x4_t) __builtin_neon_vset_lane_i16(__s0, (int16x4_t)__s1, __p2); \
24263 __ret; \
24264 })
24265 #endif
24266
24267 #ifdef __LITTLE_ENDIAN__
vshlq_u8(uint8x16_t __p0,int8x16_t __p1)24268 __ai uint8x16_t vshlq_u8(uint8x16_t __p0, int8x16_t __p1) {
24269 uint8x16_t __ret;
24270 __ret = (uint8x16_t) __builtin_neon_vshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
24271 return __ret;
24272 }
24273 #else
vshlq_u8(uint8x16_t __p0,int8x16_t __p1)24274 __ai uint8x16_t vshlq_u8(uint8x16_t __p0, int8x16_t __p1) {
24275 uint8x16_t __ret;
24276 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
24277 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
24278 __ret = (uint8x16_t) __builtin_neon_vshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
24279 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
24280 return __ret;
24281 }
24282 #endif
24283
24284 #ifdef __LITTLE_ENDIAN__
vshlq_u32(uint32x4_t __p0,int32x4_t __p1)24285 __ai uint32x4_t vshlq_u32(uint32x4_t __p0, int32x4_t __p1) {
24286 uint32x4_t __ret;
24287 __ret = (uint32x4_t) __builtin_neon_vshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
24288 return __ret;
24289 }
24290 #else
vshlq_u32(uint32x4_t __p0,int32x4_t __p1)24291 __ai uint32x4_t vshlq_u32(uint32x4_t __p0, int32x4_t __p1) {
24292 uint32x4_t __ret;
24293 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
24294 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
24295 __ret = (uint32x4_t) __builtin_neon_vshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
24296 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
24297 return __ret;
24298 }
24299 #endif
24300
24301 #ifdef __LITTLE_ENDIAN__
vshlq_u64(uint64x2_t __p0,int64x2_t __p1)24302 __ai uint64x2_t vshlq_u64(uint64x2_t __p0, int64x2_t __p1) {
24303 uint64x2_t __ret;
24304 __ret = (uint64x2_t) __builtin_neon_vshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 51);
24305 return __ret;
24306 }
24307 #else
vshlq_u64(uint64x2_t __p0,int64x2_t __p1)24308 __ai uint64x2_t vshlq_u64(uint64x2_t __p0, int64x2_t __p1) {
24309 uint64x2_t __ret;
24310 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
24311 int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
24312 __ret = (uint64x2_t) __builtin_neon_vshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 51);
24313 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
24314 return __ret;
24315 }
24316 #endif
24317
24318 #ifdef __LITTLE_ENDIAN__
vshlq_u16(uint16x8_t __p0,int16x8_t __p1)24319 __ai uint16x8_t vshlq_u16(uint16x8_t __p0, int16x8_t __p1) {
24320 uint16x8_t __ret;
24321 __ret = (uint16x8_t) __builtin_neon_vshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
24322 return __ret;
24323 }
24324 #else
vshlq_u16(uint16x8_t __p0,int16x8_t __p1)24325 __ai uint16x8_t vshlq_u16(uint16x8_t __p0, int16x8_t __p1) {
24326 uint16x8_t __ret;
24327 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
24328 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
24329 __ret = (uint16x8_t) __builtin_neon_vshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
24330 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
24331 return __ret;
24332 }
24333 #endif
24334
24335 #ifdef __LITTLE_ENDIAN__
vshlq_s8(int8x16_t __p0,int8x16_t __p1)24336 __ai int8x16_t vshlq_s8(int8x16_t __p0, int8x16_t __p1) {
24337 int8x16_t __ret;
24338 __ret = (int8x16_t) __builtin_neon_vshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 32);
24339 return __ret;
24340 }
24341 #else
vshlq_s8(int8x16_t __p0,int8x16_t __p1)24342 __ai int8x16_t vshlq_s8(int8x16_t __p0, int8x16_t __p1) {
24343 int8x16_t __ret;
24344 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
24345 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
24346 __ret = (int8x16_t) __builtin_neon_vshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 32);
24347 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
24348 return __ret;
24349 }
24350 #endif
24351
24352 #ifdef __LITTLE_ENDIAN__
vshlq_s32(int32x4_t __p0,int32x4_t __p1)24353 __ai int32x4_t vshlq_s32(int32x4_t __p0, int32x4_t __p1) {
24354 int32x4_t __ret;
24355 __ret = (int32x4_t) __builtin_neon_vshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 34);
24356 return __ret;
24357 }
24358 #else
vshlq_s32(int32x4_t __p0,int32x4_t __p1)24359 __ai int32x4_t vshlq_s32(int32x4_t __p0, int32x4_t __p1) {
24360 int32x4_t __ret;
24361 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
24362 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
24363 __ret = (int32x4_t) __builtin_neon_vshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 34);
24364 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
24365 return __ret;
24366 }
24367 #endif
24368
24369 #ifdef __LITTLE_ENDIAN__
vshlq_s64(int64x2_t __p0,int64x2_t __p1)24370 __ai int64x2_t vshlq_s64(int64x2_t __p0, int64x2_t __p1) {
24371 int64x2_t __ret;
24372 __ret = (int64x2_t) __builtin_neon_vshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 35);
24373 return __ret;
24374 }
24375 #else
vshlq_s64(int64x2_t __p0,int64x2_t __p1)24376 __ai int64x2_t vshlq_s64(int64x2_t __p0, int64x2_t __p1) {
24377 int64x2_t __ret;
24378 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
24379 int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
24380 __ret = (int64x2_t) __builtin_neon_vshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 35);
24381 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
24382 return __ret;
24383 }
24384 #endif
24385
24386 #ifdef __LITTLE_ENDIAN__
vshlq_s16(int16x8_t __p0,int16x8_t __p1)24387 __ai int16x8_t vshlq_s16(int16x8_t __p0, int16x8_t __p1) {
24388 int16x8_t __ret;
24389 __ret = (int16x8_t) __builtin_neon_vshlq_v((int8x16_t)__p0, (int8x16_t)__p1, 33);
24390 return __ret;
24391 }
24392 #else
vshlq_s16(int16x8_t __p0,int16x8_t __p1)24393 __ai int16x8_t vshlq_s16(int16x8_t __p0, int16x8_t __p1) {
24394 int16x8_t __ret;
24395 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
24396 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
24397 __ret = (int16x8_t) __builtin_neon_vshlq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 33);
24398 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
24399 return __ret;
24400 }
24401 #endif
24402
24403 #ifdef __LITTLE_ENDIAN__
vshl_u8(uint8x8_t __p0,int8x8_t __p1)24404 __ai uint8x8_t vshl_u8(uint8x8_t __p0, int8x8_t __p1) {
24405 uint8x8_t __ret;
24406 __ret = (uint8x8_t) __builtin_neon_vshl_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
24407 return __ret;
24408 }
24409 #else
vshl_u8(uint8x8_t __p0,int8x8_t __p1)24410 __ai uint8x8_t vshl_u8(uint8x8_t __p0, int8x8_t __p1) {
24411 uint8x8_t __ret;
24412 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
24413 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
24414 __ret = (uint8x8_t) __builtin_neon_vshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
24415 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
24416 return __ret;
24417 }
24418 #endif
24419
24420 #ifdef __LITTLE_ENDIAN__
vshl_u32(uint32x2_t __p0,int32x2_t __p1)24421 __ai uint32x2_t vshl_u32(uint32x2_t __p0, int32x2_t __p1) {
24422 uint32x2_t __ret;
24423 __ret = (uint32x2_t) __builtin_neon_vshl_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
24424 return __ret;
24425 }
24426 #else
vshl_u32(uint32x2_t __p0,int32x2_t __p1)24427 __ai uint32x2_t vshl_u32(uint32x2_t __p0, int32x2_t __p1) {
24428 uint32x2_t __ret;
24429 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
24430 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
24431 __ret = (uint32x2_t) __builtin_neon_vshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
24432 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
24433 return __ret;
24434 }
24435 #endif
24436
vshl_u64(uint64x1_t __p0,int64x1_t __p1)24437 __ai uint64x1_t vshl_u64(uint64x1_t __p0, int64x1_t __p1) {
24438 uint64x1_t __ret;
24439 __ret = (uint64x1_t) __builtin_neon_vshl_v((int8x8_t)__p0, (int8x8_t)__p1, 19);
24440 return __ret;
24441 }
24442 #ifdef __LITTLE_ENDIAN__
vshl_u16(uint16x4_t __p0,int16x4_t __p1)24443 __ai uint16x4_t vshl_u16(uint16x4_t __p0, int16x4_t __p1) {
24444 uint16x4_t __ret;
24445 __ret = (uint16x4_t) __builtin_neon_vshl_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
24446 return __ret;
24447 }
24448 #else
vshl_u16(uint16x4_t __p0,int16x4_t __p1)24449 __ai uint16x4_t vshl_u16(uint16x4_t __p0, int16x4_t __p1) {
24450 uint16x4_t __ret;
24451 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
24452 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
24453 __ret = (uint16x4_t) __builtin_neon_vshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
24454 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
24455 return __ret;
24456 }
24457 #endif
24458
24459 #ifdef __LITTLE_ENDIAN__
vshl_s8(int8x8_t __p0,int8x8_t __p1)24460 __ai int8x8_t vshl_s8(int8x8_t __p0, int8x8_t __p1) {
24461 int8x8_t __ret;
24462 __ret = (int8x8_t) __builtin_neon_vshl_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
24463 return __ret;
24464 }
24465 #else
vshl_s8(int8x8_t __p0,int8x8_t __p1)24466 __ai int8x8_t vshl_s8(int8x8_t __p0, int8x8_t __p1) {
24467 int8x8_t __ret;
24468 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
24469 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
24470 __ret = (int8x8_t) __builtin_neon_vshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 0);
24471 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
24472 return __ret;
24473 }
24474 #endif
24475
24476 #ifdef __LITTLE_ENDIAN__
vshl_s32(int32x2_t __p0,int32x2_t __p1)24477 __ai int32x2_t vshl_s32(int32x2_t __p0, int32x2_t __p1) {
24478 int32x2_t __ret;
24479 __ret = (int32x2_t) __builtin_neon_vshl_v((int8x8_t)__p0, (int8x8_t)__p1, 2);
24480 return __ret;
24481 }
24482 #else
vshl_s32(int32x2_t __p0,int32x2_t __p1)24483 __ai int32x2_t vshl_s32(int32x2_t __p0, int32x2_t __p1) {
24484 int32x2_t __ret;
24485 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
24486 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
24487 __ret = (int32x2_t) __builtin_neon_vshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 2);
24488 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
24489 return __ret;
24490 }
24491 #endif
24492
vshl_s64(int64x1_t __p0,int64x1_t __p1)24493 __ai int64x1_t vshl_s64(int64x1_t __p0, int64x1_t __p1) {
24494 int64x1_t __ret;
24495 __ret = (int64x1_t) __builtin_neon_vshl_v((int8x8_t)__p0, (int8x8_t)__p1, 3);
24496 return __ret;
24497 }
24498 #ifdef __LITTLE_ENDIAN__
vshl_s16(int16x4_t __p0,int16x4_t __p1)24499 __ai int16x4_t vshl_s16(int16x4_t __p0, int16x4_t __p1) {
24500 int16x4_t __ret;
24501 __ret = (int16x4_t) __builtin_neon_vshl_v((int8x8_t)__p0, (int8x8_t)__p1, 1);
24502 return __ret;
24503 }
24504 #else
vshl_s16(int16x4_t __p0,int16x4_t __p1)24505 __ai int16x4_t vshl_s16(int16x4_t __p0, int16x4_t __p1) {
24506 int16x4_t __ret;
24507 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
24508 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
24509 __ret = (int16x4_t) __builtin_neon_vshl_v((int8x8_t)__rev0, (int8x8_t)__rev1, 1);
24510 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
24511 return __ret;
24512 }
24513 #endif
24514
24515 #ifdef __LITTLE_ENDIAN__
24516 #define vshlq_n_u8(__p0, __p1) __extension__ ({ \
24517 uint8x16_t __ret; \
24518 uint8x16_t __s0 = __p0; \
24519 __ret = (uint8x16_t) __builtin_neon_vshlq_n_v((int8x16_t)__s0, __p1, 48); \
24520 __ret; \
24521 })
24522 #else
24523 #define vshlq_n_u8(__p0, __p1) __extension__ ({ \
24524 uint8x16_t __ret; \
24525 uint8x16_t __s0 = __p0; \
24526 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
24527 __ret = (uint8x16_t) __builtin_neon_vshlq_n_v((int8x16_t)__rev0, __p1, 48); \
24528 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
24529 __ret; \
24530 })
24531 #endif
24532
24533 #ifdef __LITTLE_ENDIAN__
24534 #define vshlq_n_u32(__p0, __p1) __extension__ ({ \
24535 uint32x4_t __ret; \
24536 uint32x4_t __s0 = __p0; \
24537 __ret = (uint32x4_t) __builtin_neon_vshlq_n_v((int8x16_t)__s0, __p1, 50); \
24538 __ret; \
24539 })
24540 #else
24541 #define vshlq_n_u32(__p0, __p1) __extension__ ({ \
24542 uint32x4_t __ret; \
24543 uint32x4_t __s0 = __p0; \
24544 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
24545 __ret = (uint32x4_t) __builtin_neon_vshlq_n_v((int8x16_t)__rev0, __p1, 50); \
24546 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
24547 __ret; \
24548 })
24549 #endif
24550
24551 #ifdef __LITTLE_ENDIAN__
24552 #define vshlq_n_u64(__p0, __p1) __extension__ ({ \
24553 uint64x2_t __ret; \
24554 uint64x2_t __s0 = __p0; \
24555 __ret = (uint64x2_t) __builtin_neon_vshlq_n_v((int8x16_t)__s0, __p1, 51); \
24556 __ret; \
24557 })
24558 #else
24559 #define vshlq_n_u64(__p0, __p1) __extension__ ({ \
24560 uint64x2_t __ret; \
24561 uint64x2_t __s0 = __p0; \
24562 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
24563 __ret = (uint64x2_t) __builtin_neon_vshlq_n_v((int8x16_t)__rev0, __p1, 51); \
24564 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
24565 __ret; \
24566 })
24567 #endif
24568
24569 #ifdef __LITTLE_ENDIAN__
24570 #define vshlq_n_u16(__p0, __p1) __extension__ ({ \
24571 uint16x8_t __ret; \
24572 uint16x8_t __s0 = __p0; \
24573 __ret = (uint16x8_t) __builtin_neon_vshlq_n_v((int8x16_t)__s0, __p1, 49); \
24574 __ret; \
24575 })
24576 #else
24577 #define vshlq_n_u16(__p0, __p1) __extension__ ({ \
24578 uint16x8_t __ret; \
24579 uint16x8_t __s0 = __p0; \
24580 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
24581 __ret = (uint16x8_t) __builtin_neon_vshlq_n_v((int8x16_t)__rev0, __p1, 49); \
24582 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
24583 __ret; \
24584 })
24585 #endif
24586
24587 #ifdef __LITTLE_ENDIAN__
24588 #define vshlq_n_s8(__p0, __p1) __extension__ ({ \
24589 int8x16_t __ret; \
24590 int8x16_t __s0 = __p0; \
24591 __ret = (int8x16_t) __builtin_neon_vshlq_n_v((int8x16_t)__s0, __p1, 32); \
24592 __ret; \
24593 })
24594 #else
24595 #define vshlq_n_s8(__p0, __p1) __extension__ ({ \
24596 int8x16_t __ret; \
24597 int8x16_t __s0 = __p0; \
24598 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
24599 __ret = (int8x16_t) __builtin_neon_vshlq_n_v((int8x16_t)__rev0, __p1, 32); \
24600 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
24601 __ret; \
24602 })
24603 #endif
24604
24605 #ifdef __LITTLE_ENDIAN__
24606 #define vshlq_n_s32(__p0, __p1) __extension__ ({ \
24607 int32x4_t __ret; \
24608 int32x4_t __s0 = __p0; \
24609 __ret = (int32x4_t) __builtin_neon_vshlq_n_v((int8x16_t)__s0, __p1, 34); \
24610 __ret; \
24611 })
24612 #else
24613 #define vshlq_n_s32(__p0, __p1) __extension__ ({ \
24614 int32x4_t __ret; \
24615 int32x4_t __s0 = __p0; \
24616 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
24617 __ret = (int32x4_t) __builtin_neon_vshlq_n_v((int8x16_t)__rev0, __p1, 34); \
24618 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
24619 __ret; \
24620 })
24621 #endif
24622
24623 #ifdef __LITTLE_ENDIAN__
24624 #define vshlq_n_s64(__p0, __p1) __extension__ ({ \
24625 int64x2_t __ret; \
24626 int64x2_t __s0 = __p0; \
24627 __ret = (int64x2_t) __builtin_neon_vshlq_n_v((int8x16_t)__s0, __p1, 35); \
24628 __ret; \
24629 })
24630 #else
24631 #define vshlq_n_s64(__p0, __p1) __extension__ ({ \
24632 int64x2_t __ret; \
24633 int64x2_t __s0 = __p0; \
24634 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
24635 __ret = (int64x2_t) __builtin_neon_vshlq_n_v((int8x16_t)__rev0, __p1, 35); \
24636 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
24637 __ret; \
24638 })
24639 #endif
24640
24641 #ifdef __LITTLE_ENDIAN__
24642 #define vshlq_n_s16(__p0, __p1) __extension__ ({ \
24643 int16x8_t __ret; \
24644 int16x8_t __s0 = __p0; \
24645 __ret = (int16x8_t) __builtin_neon_vshlq_n_v((int8x16_t)__s0, __p1, 33); \
24646 __ret; \
24647 })
24648 #else
24649 #define vshlq_n_s16(__p0, __p1) __extension__ ({ \
24650 int16x8_t __ret; \
24651 int16x8_t __s0 = __p0; \
24652 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
24653 __ret = (int16x8_t) __builtin_neon_vshlq_n_v((int8x16_t)__rev0, __p1, 33); \
24654 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
24655 __ret; \
24656 })
24657 #endif
24658
24659 #ifdef __LITTLE_ENDIAN__
24660 #define vshl_n_u8(__p0, __p1) __extension__ ({ \
24661 uint8x8_t __ret; \
24662 uint8x8_t __s0 = __p0; \
24663 __ret = (uint8x8_t) __builtin_neon_vshl_n_v((int8x8_t)__s0, __p1, 16); \
24664 __ret; \
24665 })
24666 #else
24667 #define vshl_n_u8(__p0, __p1) __extension__ ({ \
24668 uint8x8_t __ret; \
24669 uint8x8_t __s0 = __p0; \
24670 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
24671 __ret = (uint8x8_t) __builtin_neon_vshl_n_v((int8x8_t)__rev0, __p1, 16); \
24672 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
24673 __ret; \
24674 })
24675 #endif
24676
24677 #ifdef __LITTLE_ENDIAN__
24678 #define vshl_n_u32(__p0, __p1) __extension__ ({ \
24679 uint32x2_t __ret; \
24680 uint32x2_t __s0 = __p0; \
24681 __ret = (uint32x2_t) __builtin_neon_vshl_n_v((int8x8_t)__s0, __p1, 18); \
24682 __ret; \
24683 })
24684 #else
24685 #define vshl_n_u32(__p0, __p1) __extension__ ({ \
24686 uint32x2_t __ret; \
24687 uint32x2_t __s0 = __p0; \
24688 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
24689 __ret = (uint32x2_t) __builtin_neon_vshl_n_v((int8x8_t)__rev0, __p1, 18); \
24690 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
24691 __ret; \
24692 })
24693 #endif
24694
24695 #define vshl_n_u64(__p0, __p1) __extension__ ({ \
24696 uint64x1_t __ret; \
24697 uint64x1_t __s0 = __p0; \
24698 __ret = (uint64x1_t) __builtin_neon_vshl_n_v((int8x8_t)__s0, __p1, 19); \
24699 __ret; \
24700 })
24701 #ifdef __LITTLE_ENDIAN__
24702 #define vshl_n_u16(__p0, __p1) __extension__ ({ \
24703 uint16x4_t __ret; \
24704 uint16x4_t __s0 = __p0; \
24705 __ret = (uint16x4_t) __builtin_neon_vshl_n_v((int8x8_t)__s0, __p1, 17); \
24706 __ret; \
24707 })
24708 #else
24709 #define vshl_n_u16(__p0, __p1) __extension__ ({ \
24710 uint16x4_t __ret; \
24711 uint16x4_t __s0 = __p0; \
24712 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
24713 __ret = (uint16x4_t) __builtin_neon_vshl_n_v((int8x8_t)__rev0, __p1, 17); \
24714 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
24715 __ret; \
24716 })
24717 #endif
24718
24719 #ifdef __LITTLE_ENDIAN__
24720 #define vshl_n_s8(__p0, __p1) __extension__ ({ \
24721 int8x8_t __ret; \
24722 int8x8_t __s0 = __p0; \
24723 __ret = (int8x8_t) __builtin_neon_vshl_n_v((int8x8_t)__s0, __p1, 0); \
24724 __ret; \
24725 })
24726 #else
24727 #define vshl_n_s8(__p0, __p1) __extension__ ({ \
24728 int8x8_t __ret; \
24729 int8x8_t __s0 = __p0; \
24730 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
24731 __ret = (int8x8_t) __builtin_neon_vshl_n_v((int8x8_t)__rev0, __p1, 0); \
24732 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
24733 __ret; \
24734 })
24735 #endif
24736
24737 #ifdef __LITTLE_ENDIAN__
24738 #define vshl_n_s32(__p0, __p1) __extension__ ({ \
24739 int32x2_t __ret; \
24740 int32x2_t __s0 = __p0; \
24741 __ret = (int32x2_t) __builtin_neon_vshl_n_v((int8x8_t)__s0, __p1, 2); \
24742 __ret; \
24743 })
24744 #else
24745 #define vshl_n_s32(__p0, __p1) __extension__ ({ \
24746 int32x2_t __ret; \
24747 int32x2_t __s0 = __p0; \
24748 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
24749 __ret = (int32x2_t) __builtin_neon_vshl_n_v((int8x8_t)__rev0, __p1, 2); \
24750 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
24751 __ret; \
24752 })
24753 #endif
24754
24755 #define vshl_n_s64(__p0, __p1) __extension__ ({ \
24756 int64x1_t __ret; \
24757 int64x1_t __s0 = __p0; \
24758 __ret = (int64x1_t) __builtin_neon_vshl_n_v((int8x8_t)__s0, __p1, 3); \
24759 __ret; \
24760 })
24761 #ifdef __LITTLE_ENDIAN__
24762 #define vshl_n_s16(__p0, __p1) __extension__ ({ \
24763 int16x4_t __ret; \
24764 int16x4_t __s0 = __p0; \
24765 __ret = (int16x4_t) __builtin_neon_vshl_n_v((int8x8_t)__s0, __p1, 1); \
24766 __ret; \
24767 })
24768 #else
24769 #define vshl_n_s16(__p0, __p1) __extension__ ({ \
24770 int16x4_t __ret; \
24771 int16x4_t __s0 = __p0; \
24772 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
24773 __ret = (int16x4_t) __builtin_neon_vshl_n_v((int8x8_t)__rev0, __p1, 1); \
24774 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
24775 __ret; \
24776 })
24777 #endif
24778
24779 #ifdef __LITTLE_ENDIAN__
24780 #define vshll_n_u8(__p0, __p1) __extension__ ({ \
24781 uint16x8_t __ret; \
24782 uint8x8_t __s0 = __p0; \
24783 __ret = (uint16x8_t) __builtin_neon_vshll_n_v((int8x8_t)__s0, __p1, 49); \
24784 __ret; \
24785 })
24786 #else
24787 #define vshll_n_u8(__p0, __p1) __extension__ ({ \
24788 uint16x8_t __ret; \
24789 uint8x8_t __s0 = __p0; \
24790 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
24791 __ret = (uint16x8_t) __builtin_neon_vshll_n_v((int8x8_t)__rev0, __p1, 49); \
24792 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
24793 __ret; \
24794 })
24795 #define __noswap_vshll_n_u8(__p0, __p1) __extension__ ({ \
24796 uint16x8_t __ret; \
24797 uint8x8_t __s0 = __p0; \
24798 __ret = (uint16x8_t) __builtin_neon_vshll_n_v((int8x8_t)__s0, __p1, 49); \
24799 __ret; \
24800 })
24801 #endif
24802
24803 #ifdef __LITTLE_ENDIAN__
24804 #define vshll_n_u32(__p0, __p1) __extension__ ({ \
24805 uint64x2_t __ret; \
24806 uint32x2_t __s0 = __p0; \
24807 __ret = (uint64x2_t) __builtin_neon_vshll_n_v((int8x8_t)__s0, __p1, 51); \
24808 __ret; \
24809 })
24810 #else
24811 #define vshll_n_u32(__p0, __p1) __extension__ ({ \
24812 uint64x2_t __ret; \
24813 uint32x2_t __s0 = __p0; \
24814 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
24815 __ret = (uint64x2_t) __builtin_neon_vshll_n_v((int8x8_t)__rev0, __p1, 51); \
24816 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
24817 __ret; \
24818 })
24819 #define __noswap_vshll_n_u32(__p0, __p1) __extension__ ({ \
24820 uint64x2_t __ret; \
24821 uint32x2_t __s0 = __p0; \
24822 __ret = (uint64x2_t) __builtin_neon_vshll_n_v((int8x8_t)__s0, __p1, 51); \
24823 __ret; \
24824 })
24825 #endif
24826
24827 #ifdef __LITTLE_ENDIAN__
24828 #define vshll_n_u16(__p0, __p1) __extension__ ({ \
24829 uint32x4_t __ret; \
24830 uint16x4_t __s0 = __p0; \
24831 __ret = (uint32x4_t) __builtin_neon_vshll_n_v((int8x8_t)__s0, __p1, 50); \
24832 __ret; \
24833 })
24834 #else
24835 #define vshll_n_u16(__p0, __p1) __extension__ ({ \
24836 uint32x4_t __ret; \
24837 uint16x4_t __s0 = __p0; \
24838 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
24839 __ret = (uint32x4_t) __builtin_neon_vshll_n_v((int8x8_t)__rev0, __p1, 50); \
24840 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
24841 __ret; \
24842 })
24843 #define __noswap_vshll_n_u16(__p0, __p1) __extension__ ({ \
24844 uint32x4_t __ret; \
24845 uint16x4_t __s0 = __p0; \
24846 __ret = (uint32x4_t) __builtin_neon_vshll_n_v((int8x8_t)__s0, __p1, 50); \
24847 __ret; \
24848 })
24849 #endif
24850
24851 #ifdef __LITTLE_ENDIAN__
24852 #define vshll_n_s8(__p0, __p1) __extension__ ({ \
24853 int16x8_t __ret; \
24854 int8x8_t __s0 = __p0; \
24855 __ret = (int16x8_t) __builtin_neon_vshll_n_v((int8x8_t)__s0, __p1, 33); \
24856 __ret; \
24857 })
24858 #else
24859 #define vshll_n_s8(__p0, __p1) __extension__ ({ \
24860 int16x8_t __ret; \
24861 int8x8_t __s0 = __p0; \
24862 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
24863 __ret = (int16x8_t) __builtin_neon_vshll_n_v((int8x8_t)__rev0, __p1, 33); \
24864 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
24865 __ret; \
24866 })
24867 #define __noswap_vshll_n_s8(__p0, __p1) __extension__ ({ \
24868 int16x8_t __ret; \
24869 int8x8_t __s0 = __p0; \
24870 __ret = (int16x8_t) __builtin_neon_vshll_n_v((int8x8_t)__s0, __p1, 33); \
24871 __ret; \
24872 })
24873 #endif
24874
24875 #ifdef __LITTLE_ENDIAN__
24876 #define vshll_n_s32(__p0, __p1) __extension__ ({ \
24877 int64x2_t __ret; \
24878 int32x2_t __s0 = __p0; \
24879 __ret = (int64x2_t) __builtin_neon_vshll_n_v((int8x8_t)__s0, __p1, 35); \
24880 __ret; \
24881 })
24882 #else
24883 #define vshll_n_s32(__p0, __p1) __extension__ ({ \
24884 int64x2_t __ret; \
24885 int32x2_t __s0 = __p0; \
24886 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
24887 __ret = (int64x2_t) __builtin_neon_vshll_n_v((int8x8_t)__rev0, __p1, 35); \
24888 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
24889 __ret; \
24890 })
24891 #define __noswap_vshll_n_s32(__p0, __p1) __extension__ ({ \
24892 int64x2_t __ret; \
24893 int32x2_t __s0 = __p0; \
24894 __ret = (int64x2_t) __builtin_neon_vshll_n_v((int8x8_t)__s0, __p1, 35); \
24895 __ret; \
24896 })
24897 #endif
24898
24899 #ifdef __LITTLE_ENDIAN__
24900 #define vshll_n_s16(__p0, __p1) __extension__ ({ \
24901 int32x4_t __ret; \
24902 int16x4_t __s0 = __p0; \
24903 __ret = (int32x4_t) __builtin_neon_vshll_n_v((int8x8_t)__s0, __p1, 34); \
24904 __ret; \
24905 })
24906 #else
24907 #define vshll_n_s16(__p0, __p1) __extension__ ({ \
24908 int32x4_t __ret; \
24909 int16x4_t __s0 = __p0; \
24910 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
24911 __ret = (int32x4_t) __builtin_neon_vshll_n_v((int8x8_t)__rev0, __p1, 34); \
24912 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
24913 __ret; \
24914 })
24915 #define __noswap_vshll_n_s16(__p0, __p1) __extension__ ({ \
24916 int32x4_t __ret; \
24917 int16x4_t __s0 = __p0; \
24918 __ret = (int32x4_t) __builtin_neon_vshll_n_v((int8x8_t)__s0, __p1, 34); \
24919 __ret; \
24920 })
24921 #endif
24922
24923 #ifdef __LITTLE_ENDIAN__
24924 #define vshrq_n_u8(__p0, __p1) __extension__ ({ \
24925 uint8x16_t __ret; \
24926 uint8x16_t __s0 = __p0; \
24927 __ret = (uint8x16_t) __builtin_neon_vshrq_n_v((int8x16_t)__s0, __p1, 48); \
24928 __ret; \
24929 })
24930 #else
24931 #define vshrq_n_u8(__p0, __p1) __extension__ ({ \
24932 uint8x16_t __ret; \
24933 uint8x16_t __s0 = __p0; \
24934 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
24935 __ret = (uint8x16_t) __builtin_neon_vshrq_n_v((int8x16_t)__rev0, __p1, 48); \
24936 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
24937 __ret; \
24938 })
24939 #endif
24940
24941 #ifdef __LITTLE_ENDIAN__
24942 #define vshrq_n_u32(__p0, __p1) __extension__ ({ \
24943 uint32x4_t __ret; \
24944 uint32x4_t __s0 = __p0; \
24945 __ret = (uint32x4_t) __builtin_neon_vshrq_n_v((int8x16_t)__s0, __p1, 50); \
24946 __ret; \
24947 })
24948 #else
24949 #define vshrq_n_u32(__p0, __p1) __extension__ ({ \
24950 uint32x4_t __ret; \
24951 uint32x4_t __s0 = __p0; \
24952 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
24953 __ret = (uint32x4_t) __builtin_neon_vshrq_n_v((int8x16_t)__rev0, __p1, 50); \
24954 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
24955 __ret; \
24956 })
24957 #endif
24958
24959 #ifdef __LITTLE_ENDIAN__
24960 #define vshrq_n_u64(__p0, __p1) __extension__ ({ \
24961 uint64x2_t __ret; \
24962 uint64x2_t __s0 = __p0; \
24963 __ret = (uint64x2_t) __builtin_neon_vshrq_n_v((int8x16_t)__s0, __p1, 51); \
24964 __ret; \
24965 })
24966 #else
24967 #define vshrq_n_u64(__p0, __p1) __extension__ ({ \
24968 uint64x2_t __ret; \
24969 uint64x2_t __s0 = __p0; \
24970 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
24971 __ret = (uint64x2_t) __builtin_neon_vshrq_n_v((int8x16_t)__rev0, __p1, 51); \
24972 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
24973 __ret; \
24974 })
24975 #endif
24976
24977 #ifdef __LITTLE_ENDIAN__
24978 #define vshrq_n_u16(__p0, __p1) __extension__ ({ \
24979 uint16x8_t __ret; \
24980 uint16x8_t __s0 = __p0; \
24981 __ret = (uint16x8_t) __builtin_neon_vshrq_n_v((int8x16_t)__s0, __p1, 49); \
24982 __ret; \
24983 })
24984 #else
24985 #define vshrq_n_u16(__p0, __p1) __extension__ ({ \
24986 uint16x8_t __ret; \
24987 uint16x8_t __s0 = __p0; \
24988 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
24989 __ret = (uint16x8_t) __builtin_neon_vshrq_n_v((int8x16_t)__rev0, __p1, 49); \
24990 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
24991 __ret; \
24992 })
24993 #endif
24994
24995 #ifdef __LITTLE_ENDIAN__
24996 #define vshrq_n_s8(__p0, __p1) __extension__ ({ \
24997 int8x16_t __ret; \
24998 int8x16_t __s0 = __p0; \
24999 __ret = (int8x16_t) __builtin_neon_vshrq_n_v((int8x16_t)__s0, __p1, 32); \
25000 __ret; \
25001 })
25002 #else
25003 #define vshrq_n_s8(__p0, __p1) __extension__ ({ \
25004 int8x16_t __ret; \
25005 int8x16_t __s0 = __p0; \
25006 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
25007 __ret = (int8x16_t) __builtin_neon_vshrq_n_v((int8x16_t)__rev0, __p1, 32); \
25008 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
25009 __ret; \
25010 })
25011 #endif
25012
25013 #ifdef __LITTLE_ENDIAN__
25014 #define vshrq_n_s32(__p0, __p1) __extension__ ({ \
25015 int32x4_t __ret; \
25016 int32x4_t __s0 = __p0; \
25017 __ret = (int32x4_t) __builtin_neon_vshrq_n_v((int8x16_t)__s0, __p1, 34); \
25018 __ret; \
25019 })
25020 #else
25021 #define vshrq_n_s32(__p0, __p1) __extension__ ({ \
25022 int32x4_t __ret; \
25023 int32x4_t __s0 = __p0; \
25024 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
25025 __ret = (int32x4_t) __builtin_neon_vshrq_n_v((int8x16_t)__rev0, __p1, 34); \
25026 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
25027 __ret; \
25028 })
25029 #endif
25030
25031 #ifdef __LITTLE_ENDIAN__
25032 #define vshrq_n_s64(__p0, __p1) __extension__ ({ \
25033 int64x2_t __ret; \
25034 int64x2_t __s0 = __p0; \
25035 __ret = (int64x2_t) __builtin_neon_vshrq_n_v((int8x16_t)__s0, __p1, 35); \
25036 __ret; \
25037 })
25038 #else
25039 #define vshrq_n_s64(__p0, __p1) __extension__ ({ \
25040 int64x2_t __ret; \
25041 int64x2_t __s0 = __p0; \
25042 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
25043 __ret = (int64x2_t) __builtin_neon_vshrq_n_v((int8x16_t)__rev0, __p1, 35); \
25044 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
25045 __ret; \
25046 })
25047 #endif
25048
25049 #ifdef __LITTLE_ENDIAN__
25050 #define vshrq_n_s16(__p0, __p1) __extension__ ({ \
25051 int16x8_t __ret; \
25052 int16x8_t __s0 = __p0; \
25053 __ret = (int16x8_t) __builtin_neon_vshrq_n_v((int8x16_t)__s0, __p1, 33); \
25054 __ret; \
25055 })
25056 #else
25057 #define vshrq_n_s16(__p0, __p1) __extension__ ({ \
25058 int16x8_t __ret; \
25059 int16x8_t __s0 = __p0; \
25060 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
25061 __ret = (int16x8_t) __builtin_neon_vshrq_n_v((int8x16_t)__rev0, __p1, 33); \
25062 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
25063 __ret; \
25064 })
25065 #endif
25066
25067 #ifdef __LITTLE_ENDIAN__
25068 #define vshr_n_u8(__p0, __p1) __extension__ ({ \
25069 uint8x8_t __ret; \
25070 uint8x8_t __s0 = __p0; \
25071 __ret = (uint8x8_t) __builtin_neon_vshr_n_v((int8x8_t)__s0, __p1, 16); \
25072 __ret; \
25073 })
25074 #else
25075 #define vshr_n_u8(__p0, __p1) __extension__ ({ \
25076 uint8x8_t __ret; \
25077 uint8x8_t __s0 = __p0; \
25078 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
25079 __ret = (uint8x8_t) __builtin_neon_vshr_n_v((int8x8_t)__rev0, __p1, 16); \
25080 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
25081 __ret; \
25082 })
25083 #endif
25084
25085 #ifdef __LITTLE_ENDIAN__
25086 #define vshr_n_u32(__p0, __p1) __extension__ ({ \
25087 uint32x2_t __ret; \
25088 uint32x2_t __s0 = __p0; \
25089 __ret = (uint32x2_t) __builtin_neon_vshr_n_v((int8x8_t)__s0, __p1, 18); \
25090 __ret; \
25091 })
25092 #else
25093 #define vshr_n_u32(__p0, __p1) __extension__ ({ \
25094 uint32x2_t __ret; \
25095 uint32x2_t __s0 = __p0; \
25096 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
25097 __ret = (uint32x2_t) __builtin_neon_vshr_n_v((int8x8_t)__rev0, __p1, 18); \
25098 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
25099 __ret; \
25100 })
25101 #endif
25102
25103 #define vshr_n_u64(__p0, __p1) __extension__ ({ \
25104 uint64x1_t __ret; \
25105 uint64x1_t __s0 = __p0; \
25106 __ret = (uint64x1_t) __builtin_neon_vshr_n_v((int8x8_t)__s0, __p1, 19); \
25107 __ret; \
25108 })
25109 #ifdef __LITTLE_ENDIAN__
25110 #define vshr_n_u16(__p0, __p1) __extension__ ({ \
25111 uint16x4_t __ret; \
25112 uint16x4_t __s0 = __p0; \
25113 __ret = (uint16x4_t) __builtin_neon_vshr_n_v((int8x8_t)__s0, __p1, 17); \
25114 __ret; \
25115 })
25116 #else
25117 #define vshr_n_u16(__p0, __p1) __extension__ ({ \
25118 uint16x4_t __ret; \
25119 uint16x4_t __s0 = __p0; \
25120 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
25121 __ret = (uint16x4_t) __builtin_neon_vshr_n_v((int8x8_t)__rev0, __p1, 17); \
25122 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
25123 __ret; \
25124 })
25125 #endif
25126
25127 #ifdef __LITTLE_ENDIAN__
25128 #define vshr_n_s8(__p0, __p1) __extension__ ({ \
25129 int8x8_t __ret; \
25130 int8x8_t __s0 = __p0; \
25131 __ret = (int8x8_t) __builtin_neon_vshr_n_v((int8x8_t)__s0, __p1, 0); \
25132 __ret; \
25133 })
25134 #else
25135 #define vshr_n_s8(__p0, __p1) __extension__ ({ \
25136 int8x8_t __ret; \
25137 int8x8_t __s0 = __p0; \
25138 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
25139 __ret = (int8x8_t) __builtin_neon_vshr_n_v((int8x8_t)__rev0, __p1, 0); \
25140 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
25141 __ret; \
25142 })
25143 #endif
25144
25145 #ifdef __LITTLE_ENDIAN__
25146 #define vshr_n_s32(__p0, __p1) __extension__ ({ \
25147 int32x2_t __ret; \
25148 int32x2_t __s0 = __p0; \
25149 __ret = (int32x2_t) __builtin_neon_vshr_n_v((int8x8_t)__s0, __p1, 2); \
25150 __ret; \
25151 })
25152 #else
25153 #define vshr_n_s32(__p0, __p1) __extension__ ({ \
25154 int32x2_t __ret; \
25155 int32x2_t __s0 = __p0; \
25156 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
25157 __ret = (int32x2_t) __builtin_neon_vshr_n_v((int8x8_t)__rev0, __p1, 2); \
25158 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
25159 __ret; \
25160 })
25161 #endif
25162
25163 #define vshr_n_s64(__p0, __p1) __extension__ ({ \
25164 int64x1_t __ret; \
25165 int64x1_t __s0 = __p0; \
25166 __ret = (int64x1_t) __builtin_neon_vshr_n_v((int8x8_t)__s0, __p1, 3); \
25167 __ret; \
25168 })
25169 #ifdef __LITTLE_ENDIAN__
25170 #define vshr_n_s16(__p0, __p1) __extension__ ({ \
25171 int16x4_t __ret; \
25172 int16x4_t __s0 = __p0; \
25173 __ret = (int16x4_t) __builtin_neon_vshr_n_v((int8x8_t)__s0, __p1, 1); \
25174 __ret; \
25175 })
25176 #else
25177 #define vshr_n_s16(__p0, __p1) __extension__ ({ \
25178 int16x4_t __ret; \
25179 int16x4_t __s0 = __p0; \
25180 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
25181 __ret = (int16x4_t) __builtin_neon_vshr_n_v((int8x8_t)__rev0, __p1, 1); \
25182 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
25183 __ret; \
25184 })
25185 #endif
25186
25187 #ifdef __LITTLE_ENDIAN__
25188 #define vshrn_n_u32(__p0, __p1) __extension__ ({ \
25189 uint16x4_t __ret; \
25190 uint32x4_t __s0 = __p0; \
25191 __ret = (uint16x4_t) __builtin_neon_vshrn_n_v((int8x16_t)__s0, __p1, 17); \
25192 __ret; \
25193 })
25194 #else
25195 #define vshrn_n_u32(__p0, __p1) __extension__ ({ \
25196 uint16x4_t __ret; \
25197 uint32x4_t __s0 = __p0; \
25198 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
25199 __ret = (uint16x4_t) __builtin_neon_vshrn_n_v((int8x16_t)__rev0, __p1, 17); \
25200 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
25201 __ret; \
25202 })
25203 #define __noswap_vshrn_n_u32(__p0, __p1) __extension__ ({ \
25204 uint16x4_t __ret; \
25205 uint32x4_t __s0 = __p0; \
25206 __ret = (uint16x4_t) __builtin_neon_vshrn_n_v((int8x16_t)__s0, __p1, 17); \
25207 __ret; \
25208 })
25209 #endif
25210
25211 #ifdef __LITTLE_ENDIAN__
25212 #define vshrn_n_u64(__p0, __p1) __extension__ ({ \
25213 uint32x2_t __ret; \
25214 uint64x2_t __s0 = __p0; \
25215 __ret = (uint32x2_t) __builtin_neon_vshrn_n_v((int8x16_t)__s0, __p1, 18); \
25216 __ret; \
25217 })
25218 #else
25219 #define vshrn_n_u64(__p0, __p1) __extension__ ({ \
25220 uint32x2_t __ret; \
25221 uint64x2_t __s0 = __p0; \
25222 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
25223 __ret = (uint32x2_t) __builtin_neon_vshrn_n_v((int8x16_t)__rev0, __p1, 18); \
25224 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
25225 __ret; \
25226 })
25227 #define __noswap_vshrn_n_u64(__p0, __p1) __extension__ ({ \
25228 uint32x2_t __ret; \
25229 uint64x2_t __s0 = __p0; \
25230 __ret = (uint32x2_t) __builtin_neon_vshrn_n_v((int8x16_t)__s0, __p1, 18); \
25231 __ret; \
25232 })
25233 #endif
25234
25235 #ifdef __LITTLE_ENDIAN__
25236 #define vshrn_n_u16(__p0, __p1) __extension__ ({ \
25237 uint8x8_t __ret; \
25238 uint16x8_t __s0 = __p0; \
25239 __ret = (uint8x8_t) __builtin_neon_vshrn_n_v((int8x16_t)__s0, __p1, 16); \
25240 __ret; \
25241 })
25242 #else
25243 #define vshrn_n_u16(__p0, __p1) __extension__ ({ \
25244 uint8x8_t __ret; \
25245 uint16x8_t __s0 = __p0; \
25246 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
25247 __ret = (uint8x8_t) __builtin_neon_vshrn_n_v((int8x16_t)__rev0, __p1, 16); \
25248 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
25249 __ret; \
25250 })
25251 #define __noswap_vshrn_n_u16(__p0, __p1) __extension__ ({ \
25252 uint8x8_t __ret; \
25253 uint16x8_t __s0 = __p0; \
25254 __ret = (uint8x8_t) __builtin_neon_vshrn_n_v((int8x16_t)__s0, __p1, 16); \
25255 __ret; \
25256 })
25257 #endif
25258
25259 #ifdef __LITTLE_ENDIAN__
25260 #define vshrn_n_s32(__p0, __p1) __extension__ ({ \
25261 int16x4_t __ret; \
25262 int32x4_t __s0 = __p0; \
25263 __ret = (int16x4_t) __builtin_neon_vshrn_n_v((int8x16_t)__s0, __p1, 1); \
25264 __ret; \
25265 })
25266 #else
25267 #define vshrn_n_s32(__p0, __p1) __extension__ ({ \
25268 int16x4_t __ret; \
25269 int32x4_t __s0 = __p0; \
25270 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
25271 __ret = (int16x4_t) __builtin_neon_vshrn_n_v((int8x16_t)__rev0, __p1, 1); \
25272 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
25273 __ret; \
25274 })
25275 #define __noswap_vshrn_n_s32(__p0, __p1) __extension__ ({ \
25276 int16x4_t __ret; \
25277 int32x4_t __s0 = __p0; \
25278 __ret = (int16x4_t) __builtin_neon_vshrn_n_v((int8x16_t)__s0, __p1, 1); \
25279 __ret; \
25280 })
25281 #endif
25282
25283 #ifdef __LITTLE_ENDIAN__
25284 #define vshrn_n_s64(__p0, __p1) __extension__ ({ \
25285 int32x2_t __ret; \
25286 int64x2_t __s0 = __p0; \
25287 __ret = (int32x2_t) __builtin_neon_vshrn_n_v((int8x16_t)__s0, __p1, 2); \
25288 __ret; \
25289 })
25290 #else
25291 #define vshrn_n_s64(__p0, __p1) __extension__ ({ \
25292 int32x2_t __ret; \
25293 int64x2_t __s0 = __p0; \
25294 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
25295 __ret = (int32x2_t) __builtin_neon_vshrn_n_v((int8x16_t)__rev0, __p1, 2); \
25296 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
25297 __ret; \
25298 })
25299 #define __noswap_vshrn_n_s64(__p0, __p1) __extension__ ({ \
25300 int32x2_t __ret; \
25301 int64x2_t __s0 = __p0; \
25302 __ret = (int32x2_t) __builtin_neon_vshrn_n_v((int8x16_t)__s0, __p1, 2); \
25303 __ret; \
25304 })
25305 #endif
25306
25307 #ifdef __LITTLE_ENDIAN__
25308 #define vshrn_n_s16(__p0, __p1) __extension__ ({ \
25309 int8x8_t __ret; \
25310 int16x8_t __s0 = __p0; \
25311 __ret = (int8x8_t) __builtin_neon_vshrn_n_v((int8x16_t)__s0, __p1, 0); \
25312 __ret; \
25313 })
25314 #else
25315 #define vshrn_n_s16(__p0, __p1) __extension__ ({ \
25316 int8x8_t __ret; \
25317 int16x8_t __s0 = __p0; \
25318 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
25319 __ret = (int8x8_t) __builtin_neon_vshrn_n_v((int8x16_t)__rev0, __p1, 0); \
25320 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
25321 __ret; \
25322 })
25323 #define __noswap_vshrn_n_s16(__p0, __p1) __extension__ ({ \
25324 int8x8_t __ret; \
25325 int16x8_t __s0 = __p0; \
25326 __ret = (int8x8_t) __builtin_neon_vshrn_n_v((int8x16_t)__s0, __p1, 0); \
25327 __ret; \
25328 })
25329 #endif
25330
25331 #ifdef __LITTLE_ENDIAN__
25332 #define vsli_n_p8(__p0, __p1, __p2) __extension__ ({ \
25333 poly8x8_t __ret; \
25334 poly8x8_t __s0 = __p0; \
25335 poly8x8_t __s1 = __p1; \
25336 __ret = (poly8x8_t) __builtin_neon_vsli_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 4); \
25337 __ret; \
25338 })
25339 #else
25340 #define vsli_n_p8(__p0, __p1, __p2) __extension__ ({ \
25341 poly8x8_t __ret; \
25342 poly8x8_t __s0 = __p0; \
25343 poly8x8_t __s1 = __p1; \
25344 poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
25345 poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
25346 __ret = (poly8x8_t) __builtin_neon_vsli_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 4); \
25347 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
25348 __ret; \
25349 })
25350 #endif
25351
25352 #ifdef __LITTLE_ENDIAN__
25353 #define vsli_n_p16(__p0, __p1, __p2) __extension__ ({ \
25354 poly16x4_t __ret; \
25355 poly16x4_t __s0 = __p0; \
25356 poly16x4_t __s1 = __p1; \
25357 __ret = (poly16x4_t) __builtin_neon_vsli_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 5); \
25358 __ret; \
25359 })
25360 #else
25361 #define vsli_n_p16(__p0, __p1, __p2) __extension__ ({ \
25362 poly16x4_t __ret; \
25363 poly16x4_t __s0 = __p0; \
25364 poly16x4_t __s1 = __p1; \
25365 poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
25366 poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
25367 __ret = (poly16x4_t) __builtin_neon_vsli_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 5); \
25368 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
25369 __ret; \
25370 })
25371 #endif
25372
25373 #ifdef __LITTLE_ENDIAN__
25374 #define vsliq_n_p8(__p0, __p1, __p2) __extension__ ({ \
25375 poly8x16_t __ret; \
25376 poly8x16_t __s0 = __p0; \
25377 poly8x16_t __s1 = __p1; \
25378 __ret = (poly8x16_t) __builtin_neon_vsliq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 36); \
25379 __ret; \
25380 })
25381 #else
25382 #define vsliq_n_p8(__p0, __p1, __p2) __extension__ ({ \
25383 poly8x16_t __ret; \
25384 poly8x16_t __s0 = __p0; \
25385 poly8x16_t __s1 = __p1; \
25386 poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
25387 poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
25388 __ret = (poly8x16_t) __builtin_neon_vsliq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 36); \
25389 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
25390 __ret; \
25391 })
25392 #endif
25393
25394 #ifdef __LITTLE_ENDIAN__
25395 #define vsliq_n_p16(__p0, __p1, __p2) __extension__ ({ \
25396 poly16x8_t __ret; \
25397 poly16x8_t __s0 = __p0; \
25398 poly16x8_t __s1 = __p1; \
25399 __ret = (poly16x8_t) __builtin_neon_vsliq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 37); \
25400 __ret; \
25401 })
25402 #else
25403 #define vsliq_n_p16(__p0, __p1, __p2) __extension__ ({ \
25404 poly16x8_t __ret; \
25405 poly16x8_t __s0 = __p0; \
25406 poly16x8_t __s1 = __p1; \
25407 poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
25408 poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
25409 __ret = (poly16x8_t) __builtin_neon_vsliq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 37); \
25410 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
25411 __ret; \
25412 })
25413 #endif
25414
25415 #ifdef __LITTLE_ENDIAN__
25416 #define vsliq_n_u8(__p0, __p1, __p2) __extension__ ({ \
25417 uint8x16_t __ret; \
25418 uint8x16_t __s0 = __p0; \
25419 uint8x16_t __s1 = __p1; \
25420 __ret = (uint8x16_t) __builtin_neon_vsliq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 48); \
25421 __ret; \
25422 })
25423 #else
25424 #define vsliq_n_u8(__p0, __p1, __p2) __extension__ ({ \
25425 uint8x16_t __ret; \
25426 uint8x16_t __s0 = __p0; \
25427 uint8x16_t __s1 = __p1; \
25428 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
25429 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
25430 __ret = (uint8x16_t) __builtin_neon_vsliq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 48); \
25431 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
25432 __ret; \
25433 })
25434 #endif
25435
25436 #ifdef __LITTLE_ENDIAN__
25437 #define vsliq_n_u32(__p0, __p1, __p2) __extension__ ({ \
25438 uint32x4_t __ret; \
25439 uint32x4_t __s0 = __p0; \
25440 uint32x4_t __s1 = __p1; \
25441 __ret = (uint32x4_t) __builtin_neon_vsliq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 50); \
25442 __ret; \
25443 })
25444 #else
25445 #define vsliq_n_u32(__p0, __p1, __p2) __extension__ ({ \
25446 uint32x4_t __ret; \
25447 uint32x4_t __s0 = __p0; \
25448 uint32x4_t __s1 = __p1; \
25449 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
25450 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
25451 __ret = (uint32x4_t) __builtin_neon_vsliq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 50); \
25452 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
25453 __ret; \
25454 })
25455 #endif
25456
25457 #ifdef __LITTLE_ENDIAN__
25458 #define vsliq_n_u64(__p0, __p1, __p2) __extension__ ({ \
25459 uint64x2_t __ret; \
25460 uint64x2_t __s0 = __p0; \
25461 uint64x2_t __s1 = __p1; \
25462 __ret = (uint64x2_t) __builtin_neon_vsliq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 51); \
25463 __ret; \
25464 })
25465 #else
25466 #define vsliq_n_u64(__p0, __p1, __p2) __extension__ ({ \
25467 uint64x2_t __ret; \
25468 uint64x2_t __s0 = __p0; \
25469 uint64x2_t __s1 = __p1; \
25470 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
25471 uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
25472 __ret = (uint64x2_t) __builtin_neon_vsliq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 51); \
25473 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
25474 __ret; \
25475 })
25476 #endif
25477
25478 #ifdef __LITTLE_ENDIAN__
25479 #define vsliq_n_u16(__p0, __p1, __p2) __extension__ ({ \
25480 uint16x8_t __ret; \
25481 uint16x8_t __s0 = __p0; \
25482 uint16x8_t __s1 = __p1; \
25483 __ret = (uint16x8_t) __builtin_neon_vsliq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 49); \
25484 __ret; \
25485 })
25486 #else
25487 #define vsliq_n_u16(__p0, __p1, __p2) __extension__ ({ \
25488 uint16x8_t __ret; \
25489 uint16x8_t __s0 = __p0; \
25490 uint16x8_t __s1 = __p1; \
25491 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
25492 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
25493 __ret = (uint16x8_t) __builtin_neon_vsliq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 49); \
25494 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
25495 __ret; \
25496 })
25497 #endif
25498
25499 #ifdef __LITTLE_ENDIAN__
25500 #define vsliq_n_s8(__p0, __p1, __p2) __extension__ ({ \
25501 int8x16_t __ret; \
25502 int8x16_t __s0 = __p0; \
25503 int8x16_t __s1 = __p1; \
25504 __ret = (int8x16_t) __builtin_neon_vsliq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 32); \
25505 __ret; \
25506 })
25507 #else
25508 #define vsliq_n_s8(__p0, __p1, __p2) __extension__ ({ \
25509 int8x16_t __ret; \
25510 int8x16_t __s0 = __p0; \
25511 int8x16_t __s1 = __p1; \
25512 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
25513 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
25514 __ret = (int8x16_t) __builtin_neon_vsliq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 32); \
25515 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
25516 __ret; \
25517 })
25518 #endif
25519
25520 #ifdef __LITTLE_ENDIAN__
25521 #define vsliq_n_s32(__p0, __p1, __p2) __extension__ ({ \
25522 int32x4_t __ret; \
25523 int32x4_t __s0 = __p0; \
25524 int32x4_t __s1 = __p1; \
25525 __ret = (int32x4_t) __builtin_neon_vsliq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 34); \
25526 __ret; \
25527 })
25528 #else
25529 #define vsliq_n_s32(__p0, __p1, __p2) __extension__ ({ \
25530 int32x4_t __ret; \
25531 int32x4_t __s0 = __p0; \
25532 int32x4_t __s1 = __p1; \
25533 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
25534 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
25535 __ret = (int32x4_t) __builtin_neon_vsliq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 34); \
25536 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
25537 __ret; \
25538 })
25539 #endif
25540
25541 #ifdef __LITTLE_ENDIAN__
25542 #define vsliq_n_s64(__p0, __p1, __p2) __extension__ ({ \
25543 int64x2_t __ret; \
25544 int64x2_t __s0 = __p0; \
25545 int64x2_t __s1 = __p1; \
25546 __ret = (int64x2_t) __builtin_neon_vsliq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 35); \
25547 __ret; \
25548 })
25549 #else
25550 #define vsliq_n_s64(__p0, __p1, __p2) __extension__ ({ \
25551 int64x2_t __ret; \
25552 int64x2_t __s0 = __p0; \
25553 int64x2_t __s1 = __p1; \
25554 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
25555 int64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
25556 __ret = (int64x2_t) __builtin_neon_vsliq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 35); \
25557 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
25558 __ret; \
25559 })
25560 #endif
25561
25562 #ifdef __LITTLE_ENDIAN__
25563 #define vsliq_n_s16(__p0, __p1, __p2) __extension__ ({ \
25564 int16x8_t __ret; \
25565 int16x8_t __s0 = __p0; \
25566 int16x8_t __s1 = __p1; \
25567 __ret = (int16x8_t) __builtin_neon_vsliq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 33); \
25568 __ret; \
25569 })
25570 #else
25571 #define vsliq_n_s16(__p0, __p1, __p2) __extension__ ({ \
25572 int16x8_t __ret; \
25573 int16x8_t __s0 = __p0; \
25574 int16x8_t __s1 = __p1; \
25575 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
25576 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
25577 __ret = (int16x8_t) __builtin_neon_vsliq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 33); \
25578 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
25579 __ret; \
25580 })
25581 #endif
25582
25583 #ifdef __LITTLE_ENDIAN__
25584 #define vsli_n_u8(__p0, __p1, __p2) __extension__ ({ \
25585 uint8x8_t __ret; \
25586 uint8x8_t __s0 = __p0; \
25587 uint8x8_t __s1 = __p1; \
25588 __ret = (uint8x8_t) __builtin_neon_vsli_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 16); \
25589 __ret; \
25590 })
25591 #else
25592 #define vsli_n_u8(__p0, __p1, __p2) __extension__ ({ \
25593 uint8x8_t __ret; \
25594 uint8x8_t __s0 = __p0; \
25595 uint8x8_t __s1 = __p1; \
25596 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
25597 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
25598 __ret = (uint8x8_t) __builtin_neon_vsli_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 16); \
25599 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
25600 __ret; \
25601 })
25602 #endif
25603
25604 #ifdef __LITTLE_ENDIAN__
25605 #define vsli_n_u32(__p0, __p1, __p2) __extension__ ({ \
25606 uint32x2_t __ret; \
25607 uint32x2_t __s0 = __p0; \
25608 uint32x2_t __s1 = __p1; \
25609 __ret = (uint32x2_t) __builtin_neon_vsli_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 18); \
25610 __ret; \
25611 })
25612 #else
25613 #define vsli_n_u32(__p0, __p1, __p2) __extension__ ({ \
25614 uint32x2_t __ret; \
25615 uint32x2_t __s0 = __p0; \
25616 uint32x2_t __s1 = __p1; \
25617 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
25618 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
25619 __ret = (uint32x2_t) __builtin_neon_vsli_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 18); \
25620 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
25621 __ret; \
25622 })
25623 #endif
25624
25625 #define vsli_n_u64(__p0, __p1, __p2) __extension__ ({ \
25626 uint64x1_t __ret; \
25627 uint64x1_t __s0 = __p0; \
25628 uint64x1_t __s1 = __p1; \
25629 __ret = (uint64x1_t) __builtin_neon_vsli_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 19); \
25630 __ret; \
25631 })
25632 #ifdef __LITTLE_ENDIAN__
25633 #define vsli_n_u16(__p0, __p1, __p2) __extension__ ({ \
25634 uint16x4_t __ret; \
25635 uint16x4_t __s0 = __p0; \
25636 uint16x4_t __s1 = __p1; \
25637 __ret = (uint16x4_t) __builtin_neon_vsli_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 17); \
25638 __ret; \
25639 })
25640 #else
25641 #define vsli_n_u16(__p0, __p1, __p2) __extension__ ({ \
25642 uint16x4_t __ret; \
25643 uint16x4_t __s0 = __p0; \
25644 uint16x4_t __s1 = __p1; \
25645 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
25646 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
25647 __ret = (uint16x4_t) __builtin_neon_vsli_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 17); \
25648 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
25649 __ret; \
25650 })
25651 #endif
25652
25653 #ifdef __LITTLE_ENDIAN__
25654 #define vsli_n_s8(__p0, __p1, __p2) __extension__ ({ \
25655 int8x8_t __ret; \
25656 int8x8_t __s0 = __p0; \
25657 int8x8_t __s1 = __p1; \
25658 __ret = (int8x8_t) __builtin_neon_vsli_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 0); \
25659 __ret; \
25660 })
25661 #else
25662 #define vsli_n_s8(__p0, __p1, __p2) __extension__ ({ \
25663 int8x8_t __ret; \
25664 int8x8_t __s0 = __p0; \
25665 int8x8_t __s1 = __p1; \
25666 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
25667 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
25668 __ret = (int8x8_t) __builtin_neon_vsli_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 0); \
25669 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
25670 __ret; \
25671 })
25672 #endif
25673
25674 #ifdef __LITTLE_ENDIAN__
25675 #define vsli_n_s32(__p0, __p1, __p2) __extension__ ({ \
25676 int32x2_t __ret; \
25677 int32x2_t __s0 = __p0; \
25678 int32x2_t __s1 = __p1; \
25679 __ret = (int32x2_t) __builtin_neon_vsli_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 2); \
25680 __ret; \
25681 })
25682 #else
25683 #define vsli_n_s32(__p0, __p1, __p2) __extension__ ({ \
25684 int32x2_t __ret; \
25685 int32x2_t __s0 = __p0; \
25686 int32x2_t __s1 = __p1; \
25687 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
25688 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
25689 __ret = (int32x2_t) __builtin_neon_vsli_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 2); \
25690 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
25691 __ret; \
25692 })
25693 #endif
25694
25695 #define vsli_n_s64(__p0, __p1, __p2) __extension__ ({ \
25696 int64x1_t __ret; \
25697 int64x1_t __s0 = __p0; \
25698 int64x1_t __s1 = __p1; \
25699 __ret = (int64x1_t) __builtin_neon_vsli_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 3); \
25700 __ret; \
25701 })
25702 #ifdef __LITTLE_ENDIAN__
25703 #define vsli_n_s16(__p0, __p1, __p2) __extension__ ({ \
25704 int16x4_t __ret; \
25705 int16x4_t __s0 = __p0; \
25706 int16x4_t __s1 = __p1; \
25707 __ret = (int16x4_t) __builtin_neon_vsli_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 1); \
25708 __ret; \
25709 })
25710 #else
25711 #define vsli_n_s16(__p0, __p1, __p2) __extension__ ({ \
25712 int16x4_t __ret; \
25713 int16x4_t __s0 = __p0; \
25714 int16x4_t __s1 = __p1; \
25715 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
25716 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
25717 __ret = (int16x4_t) __builtin_neon_vsli_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 1); \
25718 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
25719 __ret; \
25720 })
25721 #endif
25722
25723 #ifdef __LITTLE_ENDIAN__
25724 #define vsraq_n_u8(__p0, __p1, __p2) __extension__ ({ \
25725 uint8x16_t __ret; \
25726 uint8x16_t __s0 = __p0; \
25727 uint8x16_t __s1 = __p1; \
25728 __ret = (uint8x16_t) __builtin_neon_vsraq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 48); \
25729 __ret; \
25730 })
25731 #else
25732 #define vsraq_n_u8(__p0, __p1, __p2) __extension__ ({ \
25733 uint8x16_t __ret; \
25734 uint8x16_t __s0 = __p0; \
25735 uint8x16_t __s1 = __p1; \
25736 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
25737 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
25738 __ret = (uint8x16_t) __builtin_neon_vsraq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 48); \
25739 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
25740 __ret; \
25741 })
25742 #endif
25743
25744 #ifdef __LITTLE_ENDIAN__
25745 #define vsraq_n_u32(__p0, __p1, __p2) __extension__ ({ \
25746 uint32x4_t __ret; \
25747 uint32x4_t __s0 = __p0; \
25748 uint32x4_t __s1 = __p1; \
25749 __ret = (uint32x4_t) __builtin_neon_vsraq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 50); \
25750 __ret; \
25751 })
25752 #else
25753 #define vsraq_n_u32(__p0, __p1, __p2) __extension__ ({ \
25754 uint32x4_t __ret; \
25755 uint32x4_t __s0 = __p0; \
25756 uint32x4_t __s1 = __p1; \
25757 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
25758 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
25759 __ret = (uint32x4_t) __builtin_neon_vsraq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 50); \
25760 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
25761 __ret; \
25762 })
25763 #endif
25764
25765 #ifdef __LITTLE_ENDIAN__
25766 #define vsraq_n_u64(__p0, __p1, __p2) __extension__ ({ \
25767 uint64x2_t __ret; \
25768 uint64x2_t __s0 = __p0; \
25769 uint64x2_t __s1 = __p1; \
25770 __ret = (uint64x2_t) __builtin_neon_vsraq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 51); \
25771 __ret; \
25772 })
25773 #else
25774 #define vsraq_n_u64(__p0, __p1, __p2) __extension__ ({ \
25775 uint64x2_t __ret; \
25776 uint64x2_t __s0 = __p0; \
25777 uint64x2_t __s1 = __p1; \
25778 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
25779 uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
25780 __ret = (uint64x2_t) __builtin_neon_vsraq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 51); \
25781 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
25782 __ret; \
25783 })
25784 #endif
25785
25786 #ifdef __LITTLE_ENDIAN__
25787 #define vsraq_n_u16(__p0, __p1, __p2) __extension__ ({ \
25788 uint16x8_t __ret; \
25789 uint16x8_t __s0 = __p0; \
25790 uint16x8_t __s1 = __p1; \
25791 __ret = (uint16x8_t) __builtin_neon_vsraq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 49); \
25792 __ret; \
25793 })
25794 #else
25795 #define vsraq_n_u16(__p0, __p1, __p2) __extension__ ({ \
25796 uint16x8_t __ret; \
25797 uint16x8_t __s0 = __p0; \
25798 uint16x8_t __s1 = __p1; \
25799 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
25800 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
25801 __ret = (uint16x8_t) __builtin_neon_vsraq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 49); \
25802 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
25803 __ret; \
25804 })
25805 #endif
25806
25807 #ifdef __LITTLE_ENDIAN__
25808 #define vsraq_n_s8(__p0, __p1, __p2) __extension__ ({ \
25809 int8x16_t __ret; \
25810 int8x16_t __s0 = __p0; \
25811 int8x16_t __s1 = __p1; \
25812 __ret = (int8x16_t) __builtin_neon_vsraq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 32); \
25813 __ret; \
25814 })
25815 #else
25816 #define vsraq_n_s8(__p0, __p1, __p2) __extension__ ({ \
25817 int8x16_t __ret; \
25818 int8x16_t __s0 = __p0; \
25819 int8x16_t __s1 = __p1; \
25820 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
25821 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
25822 __ret = (int8x16_t) __builtin_neon_vsraq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 32); \
25823 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
25824 __ret; \
25825 })
25826 #endif
25827
25828 #ifdef __LITTLE_ENDIAN__
25829 #define vsraq_n_s32(__p0, __p1, __p2) __extension__ ({ \
25830 int32x4_t __ret; \
25831 int32x4_t __s0 = __p0; \
25832 int32x4_t __s1 = __p1; \
25833 __ret = (int32x4_t) __builtin_neon_vsraq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 34); \
25834 __ret; \
25835 })
25836 #else
25837 #define vsraq_n_s32(__p0, __p1, __p2) __extension__ ({ \
25838 int32x4_t __ret; \
25839 int32x4_t __s0 = __p0; \
25840 int32x4_t __s1 = __p1; \
25841 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
25842 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
25843 __ret = (int32x4_t) __builtin_neon_vsraq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 34); \
25844 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
25845 __ret; \
25846 })
25847 #endif
25848
25849 #ifdef __LITTLE_ENDIAN__
25850 #define vsraq_n_s64(__p0, __p1, __p2) __extension__ ({ \
25851 int64x2_t __ret; \
25852 int64x2_t __s0 = __p0; \
25853 int64x2_t __s1 = __p1; \
25854 __ret = (int64x2_t) __builtin_neon_vsraq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 35); \
25855 __ret; \
25856 })
25857 #else
25858 #define vsraq_n_s64(__p0, __p1, __p2) __extension__ ({ \
25859 int64x2_t __ret; \
25860 int64x2_t __s0 = __p0; \
25861 int64x2_t __s1 = __p1; \
25862 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
25863 int64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
25864 __ret = (int64x2_t) __builtin_neon_vsraq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 35); \
25865 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
25866 __ret; \
25867 })
25868 #endif
25869
25870 #ifdef __LITTLE_ENDIAN__
25871 #define vsraq_n_s16(__p0, __p1, __p2) __extension__ ({ \
25872 int16x8_t __ret; \
25873 int16x8_t __s0 = __p0; \
25874 int16x8_t __s1 = __p1; \
25875 __ret = (int16x8_t) __builtin_neon_vsraq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 33); \
25876 __ret; \
25877 })
25878 #else
25879 #define vsraq_n_s16(__p0, __p1, __p2) __extension__ ({ \
25880 int16x8_t __ret; \
25881 int16x8_t __s0 = __p0; \
25882 int16x8_t __s1 = __p1; \
25883 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
25884 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
25885 __ret = (int16x8_t) __builtin_neon_vsraq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 33); \
25886 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
25887 __ret; \
25888 })
25889 #endif
25890
25891 #ifdef __LITTLE_ENDIAN__
25892 #define vsra_n_u8(__p0, __p1, __p2) __extension__ ({ \
25893 uint8x8_t __ret; \
25894 uint8x8_t __s0 = __p0; \
25895 uint8x8_t __s1 = __p1; \
25896 __ret = (uint8x8_t) __builtin_neon_vsra_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 16); \
25897 __ret; \
25898 })
25899 #else
25900 #define vsra_n_u8(__p0, __p1, __p2) __extension__ ({ \
25901 uint8x8_t __ret; \
25902 uint8x8_t __s0 = __p0; \
25903 uint8x8_t __s1 = __p1; \
25904 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
25905 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
25906 __ret = (uint8x8_t) __builtin_neon_vsra_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 16); \
25907 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
25908 __ret; \
25909 })
25910 #endif
25911
25912 #ifdef __LITTLE_ENDIAN__
25913 #define vsra_n_u32(__p0, __p1, __p2) __extension__ ({ \
25914 uint32x2_t __ret; \
25915 uint32x2_t __s0 = __p0; \
25916 uint32x2_t __s1 = __p1; \
25917 __ret = (uint32x2_t) __builtin_neon_vsra_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 18); \
25918 __ret; \
25919 })
25920 #else
25921 #define vsra_n_u32(__p0, __p1, __p2) __extension__ ({ \
25922 uint32x2_t __ret; \
25923 uint32x2_t __s0 = __p0; \
25924 uint32x2_t __s1 = __p1; \
25925 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
25926 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
25927 __ret = (uint32x2_t) __builtin_neon_vsra_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 18); \
25928 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
25929 __ret; \
25930 })
25931 #endif
25932
25933 #define vsra_n_u64(__p0, __p1, __p2) __extension__ ({ \
25934 uint64x1_t __ret; \
25935 uint64x1_t __s0 = __p0; \
25936 uint64x1_t __s1 = __p1; \
25937 __ret = (uint64x1_t) __builtin_neon_vsra_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 19); \
25938 __ret; \
25939 })
25940 #ifdef __LITTLE_ENDIAN__
25941 #define vsra_n_u16(__p0, __p1, __p2) __extension__ ({ \
25942 uint16x4_t __ret; \
25943 uint16x4_t __s0 = __p0; \
25944 uint16x4_t __s1 = __p1; \
25945 __ret = (uint16x4_t) __builtin_neon_vsra_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 17); \
25946 __ret; \
25947 })
25948 #else
25949 #define vsra_n_u16(__p0, __p1, __p2) __extension__ ({ \
25950 uint16x4_t __ret; \
25951 uint16x4_t __s0 = __p0; \
25952 uint16x4_t __s1 = __p1; \
25953 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
25954 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
25955 __ret = (uint16x4_t) __builtin_neon_vsra_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 17); \
25956 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
25957 __ret; \
25958 })
25959 #endif
25960
25961 #ifdef __LITTLE_ENDIAN__
25962 #define vsra_n_s8(__p0, __p1, __p2) __extension__ ({ \
25963 int8x8_t __ret; \
25964 int8x8_t __s0 = __p0; \
25965 int8x8_t __s1 = __p1; \
25966 __ret = (int8x8_t) __builtin_neon_vsra_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 0); \
25967 __ret; \
25968 })
25969 #else
25970 #define vsra_n_s8(__p0, __p1, __p2) __extension__ ({ \
25971 int8x8_t __ret; \
25972 int8x8_t __s0 = __p0; \
25973 int8x8_t __s1 = __p1; \
25974 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
25975 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
25976 __ret = (int8x8_t) __builtin_neon_vsra_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 0); \
25977 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
25978 __ret; \
25979 })
25980 #endif
25981
25982 #ifdef __LITTLE_ENDIAN__
25983 #define vsra_n_s32(__p0, __p1, __p2) __extension__ ({ \
25984 int32x2_t __ret; \
25985 int32x2_t __s0 = __p0; \
25986 int32x2_t __s1 = __p1; \
25987 __ret = (int32x2_t) __builtin_neon_vsra_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 2); \
25988 __ret; \
25989 })
25990 #else
25991 #define vsra_n_s32(__p0, __p1, __p2) __extension__ ({ \
25992 int32x2_t __ret; \
25993 int32x2_t __s0 = __p0; \
25994 int32x2_t __s1 = __p1; \
25995 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
25996 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
25997 __ret = (int32x2_t) __builtin_neon_vsra_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 2); \
25998 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
25999 __ret; \
26000 })
26001 #endif
26002
26003 #define vsra_n_s64(__p0, __p1, __p2) __extension__ ({ \
26004 int64x1_t __ret; \
26005 int64x1_t __s0 = __p0; \
26006 int64x1_t __s1 = __p1; \
26007 __ret = (int64x1_t) __builtin_neon_vsra_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 3); \
26008 __ret; \
26009 })
26010 #ifdef __LITTLE_ENDIAN__
26011 #define vsra_n_s16(__p0, __p1, __p2) __extension__ ({ \
26012 int16x4_t __ret; \
26013 int16x4_t __s0 = __p0; \
26014 int16x4_t __s1 = __p1; \
26015 __ret = (int16x4_t) __builtin_neon_vsra_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 1); \
26016 __ret; \
26017 })
26018 #else
26019 #define vsra_n_s16(__p0, __p1, __p2) __extension__ ({ \
26020 int16x4_t __ret; \
26021 int16x4_t __s0 = __p0; \
26022 int16x4_t __s1 = __p1; \
26023 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
26024 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
26025 __ret = (int16x4_t) __builtin_neon_vsra_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 1); \
26026 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
26027 __ret; \
26028 })
26029 #endif
26030
26031 #ifdef __LITTLE_ENDIAN__
26032 #define vsri_n_p8(__p0, __p1, __p2) __extension__ ({ \
26033 poly8x8_t __ret; \
26034 poly8x8_t __s0 = __p0; \
26035 poly8x8_t __s1 = __p1; \
26036 __ret = (poly8x8_t) __builtin_neon_vsri_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 4); \
26037 __ret; \
26038 })
26039 #else
26040 #define vsri_n_p8(__p0, __p1, __p2) __extension__ ({ \
26041 poly8x8_t __ret; \
26042 poly8x8_t __s0 = __p0; \
26043 poly8x8_t __s1 = __p1; \
26044 poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
26045 poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
26046 __ret = (poly8x8_t) __builtin_neon_vsri_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 4); \
26047 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
26048 __ret; \
26049 })
26050 #endif
26051
26052 #ifdef __LITTLE_ENDIAN__
26053 #define vsri_n_p16(__p0, __p1, __p2) __extension__ ({ \
26054 poly16x4_t __ret; \
26055 poly16x4_t __s0 = __p0; \
26056 poly16x4_t __s1 = __p1; \
26057 __ret = (poly16x4_t) __builtin_neon_vsri_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 5); \
26058 __ret; \
26059 })
26060 #else
26061 #define vsri_n_p16(__p0, __p1, __p2) __extension__ ({ \
26062 poly16x4_t __ret; \
26063 poly16x4_t __s0 = __p0; \
26064 poly16x4_t __s1 = __p1; \
26065 poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
26066 poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
26067 __ret = (poly16x4_t) __builtin_neon_vsri_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 5); \
26068 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
26069 __ret; \
26070 })
26071 #endif
26072
26073 #ifdef __LITTLE_ENDIAN__
26074 #define vsriq_n_p8(__p0, __p1, __p2) __extension__ ({ \
26075 poly8x16_t __ret; \
26076 poly8x16_t __s0 = __p0; \
26077 poly8x16_t __s1 = __p1; \
26078 __ret = (poly8x16_t) __builtin_neon_vsriq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 36); \
26079 __ret; \
26080 })
26081 #else
26082 #define vsriq_n_p8(__p0, __p1, __p2) __extension__ ({ \
26083 poly8x16_t __ret; \
26084 poly8x16_t __s0 = __p0; \
26085 poly8x16_t __s1 = __p1; \
26086 poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
26087 poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
26088 __ret = (poly8x16_t) __builtin_neon_vsriq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 36); \
26089 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
26090 __ret; \
26091 })
26092 #endif
26093
26094 #ifdef __LITTLE_ENDIAN__
26095 #define vsriq_n_p16(__p0, __p1, __p2) __extension__ ({ \
26096 poly16x8_t __ret; \
26097 poly16x8_t __s0 = __p0; \
26098 poly16x8_t __s1 = __p1; \
26099 __ret = (poly16x8_t) __builtin_neon_vsriq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 37); \
26100 __ret; \
26101 })
26102 #else
26103 #define vsriq_n_p16(__p0, __p1, __p2) __extension__ ({ \
26104 poly16x8_t __ret; \
26105 poly16x8_t __s0 = __p0; \
26106 poly16x8_t __s1 = __p1; \
26107 poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
26108 poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
26109 __ret = (poly16x8_t) __builtin_neon_vsriq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 37); \
26110 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
26111 __ret; \
26112 })
26113 #endif
26114
26115 #ifdef __LITTLE_ENDIAN__
26116 #define vsriq_n_u8(__p0, __p1, __p2) __extension__ ({ \
26117 uint8x16_t __ret; \
26118 uint8x16_t __s0 = __p0; \
26119 uint8x16_t __s1 = __p1; \
26120 __ret = (uint8x16_t) __builtin_neon_vsriq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 48); \
26121 __ret; \
26122 })
26123 #else
26124 #define vsriq_n_u8(__p0, __p1, __p2) __extension__ ({ \
26125 uint8x16_t __ret; \
26126 uint8x16_t __s0 = __p0; \
26127 uint8x16_t __s1 = __p1; \
26128 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
26129 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
26130 __ret = (uint8x16_t) __builtin_neon_vsriq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 48); \
26131 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
26132 __ret; \
26133 })
26134 #endif
26135
26136 #ifdef __LITTLE_ENDIAN__
26137 #define vsriq_n_u32(__p0, __p1, __p2) __extension__ ({ \
26138 uint32x4_t __ret; \
26139 uint32x4_t __s0 = __p0; \
26140 uint32x4_t __s1 = __p1; \
26141 __ret = (uint32x4_t) __builtin_neon_vsriq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 50); \
26142 __ret; \
26143 })
26144 #else
26145 #define vsriq_n_u32(__p0, __p1, __p2) __extension__ ({ \
26146 uint32x4_t __ret; \
26147 uint32x4_t __s0 = __p0; \
26148 uint32x4_t __s1 = __p1; \
26149 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
26150 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
26151 __ret = (uint32x4_t) __builtin_neon_vsriq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 50); \
26152 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
26153 __ret; \
26154 })
26155 #endif
26156
26157 #ifdef __LITTLE_ENDIAN__
26158 #define vsriq_n_u64(__p0, __p1, __p2) __extension__ ({ \
26159 uint64x2_t __ret; \
26160 uint64x2_t __s0 = __p0; \
26161 uint64x2_t __s1 = __p1; \
26162 __ret = (uint64x2_t) __builtin_neon_vsriq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 51); \
26163 __ret; \
26164 })
26165 #else
26166 #define vsriq_n_u64(__p0, __p1, __p2) __extension__ ({ \
26167 uint64x2_t __ret; \
26168 uint64x2_t __s0 = __p0; \
26169 uint64x2_t __s1 = __p1; \
26170 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
26171 uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
26172 __ret = (uint64x2_t) __builtin_neon_vsriq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 51); \
26173 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
26174 __ret; \
26175 })
26176 #endif
26177
26178 #ifdef __LITTLE_ENDIAN__
26179 #define vsriq_n_u16(__p0, __p1, __p2) __extension__ ({ \
26180 uint16x8_t __ret; \
26181 uint16x8_t __s0 = __p0; \
26182 uint16x8_t __s1 = __p1; \
26183 __ret = (uint16x8_t) __builtin_neon_vsriq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 49); \
26184 __ret; \
26185 })
26186 #else
26187 #define vsriq_n_u16(__p0, __p1, __p2) __extension__ ({ \
26188 uint16x8_t __ret; \
26189 uint16x8_t __s0 = __p0; \
26190 uint16x8_t __s1 = __p1; \
26191 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
26192 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
26193 __ret = (uint16x8_t) __builtin_neon_vsriq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 49); \
26194 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
26195 __ret; \
26196 })
26197 #endif
26198
26199 #ifdef __LITTLE_ENDIAN__
26200 #define vsriq_n_s8(__p0, __p1, __p2) __extension__ ({ \
26201 int8x16_t __ret; \
26202 int8x16_t __s0 = __p0; \
26203 int8x16_t __s1 = __p1; \
26204 __ret = (int8x16_t) __builtin_neon_vsriq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 32); \
26205 __ret; \
26206 })
26207 #else
26208 #define vsriq_n_s8(__p0, __p1, __p2) __extension__ ({ \
26209 int8x16_t __ret; \
26210 int8x16_t __s0 = __p0; \
26211 int8x16_t __s1 = __p1; \
26212 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
26213 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
26214 __ret = (int8x16_t) __builtin_neon_vsriq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 32); \
26215 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
26216 __ret; \
26217 })
26218 #endif
26219
26220 #ifdef __LITTLE_ENDIAN__
26221 #define vsriq_n_s32(__p0, __p1, __p2) __extension__ ({ \
26222 int32x4_t __ret; \
26223 int32x4_t __s0 = __p0; \
26224 int32x4_t __s1 = __p1; \
26225 __ret = (int32x4_t) __builtin_neon_vsriq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 34); \
26226 __ret; \
26227 })
26228 #else
26229 #define vsriq_n_s32(__p0, __p1, __p2) __extension__ ({ \
26230 int32x4_t __ret; \
26231 int32x4_t __s0 = __p0; \
26232 int32x4_t __s1 = __p1; \
26233 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
26234 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
26235 __ret = (int32x4_t) __builtin_neon_vsriq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 34); \
26236 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
26237 __ret; \
26238 })
26239 #endif
26240
26241 #ifdef __LITTLE_ENDIAN__
26242 #define vsriq_n_s64(__p0, __p1, __p2) __extension__ ({ \
26243 int64x2_t __ret; \
26244 int64x2_t __s0 = __p0; \
26245 int64x2_t __s1 = __p1; \
26246 __ret = (int64x2_t) __builtin_neon_vsriq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 35); \
26247 __ret; \
26248 })
26249 #else
26250 #define vsriq_n_s64(__p0, __p1, __p2) __extension__ ({ \
26251 int64x2_t __ret; \
26252 int64x2_t __s0 = __p0; \
26253 int64x2_t __s1 = __p1; \
26254 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
26255 int64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
26256 __ret = (int64x2_t) __builtin_neon_vsriq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 35); \
26257 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
26258 __ret; \
26259 })
26260 #endif
26261
26262 #ifdef __LITTLE_ENDIAN__
26263 #define vsriq_n_s16(__p0, __p1, __p2) __extension__ ({ \
26264 int16x8_t __ret; \
26265 int16x8_t __s0 = __p0; \
26266 int16x8_t __s1 = __p1; \
26267 __ret = (int16x8_t) __builtin_neon_vsriq_n_v((int8x16_t)__s0, (int8x16_t)__s1, __p2, 33); \
26268 __ret; \
26269 })
26270 #else
26271 #define vsriq_n_s16(__p0, __p1, __p2) __extension__ ({ \
26272 int16x8_t __ret; \
26273 int16x8_t __s0 = __p0; \
26274 int16x8_t __s1 = __p1; \
26275 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
26276 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
26277 __ret = (int16x8_t) __builtin_neon_vsriq_n_v((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 33); \
26278 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
26279 __ret; \
26280 })
26281 #endif
26282
26283 #ifdef __LITTLE_ENDIAN__
26284 #define vsri_n_u8(__p0, __p1, __p2) __extension__ ({ \
26285 uint8x8_t __ret; \
26286 uint8x8_t __s0 = __p0; \
26287 uint8x8_t __s1 = __p1; \
26288 __ret = (uint8x8_t) __builtin_neon_vsri_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 16); \
26289 __ret; \
26290 })
26291 #else
26292 #define vsri_n_u8(__p0, __p1, __p2) __extension__ ({ \
26293 uint8x8_t __ret; \
26294 uint8x8_t __s0 = __p0; \
26295 uint8x8_t __s1 = __p1; \
26296 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
26297 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
26298 __ret = (uint8x8_t) __builtin_neon_vsri_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 16); \
26299 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
26300 __ret; \
26301 })
26302 #endif
26303
26304 #ifdef __LITTLE_ENDIAN__
26305 #define vsri_n_u32(__p0, __p1, __p2) __extension__ ({ \
26306 uint32x2_t __ret; \
26307 uint32x2_t __s0 = __p0; \
26308 uint32x2_t __s1 = __p1; \
26309 __ret = (uint32x2_t) __builtin_neon_vsri_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 18); \
26310 __ret; \
26311 })
26312 #else
26313 #define vsri_n_u32(__p0, __p1, __p2) __extension__ ({ \
26314 uint32x2_t __ret; \
26315 uint32x2_t __s0 = __p0; \
26316 uint32x2_t __s1 = __p1; \
26317 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
26318 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
26319 __ret = (uint32x2_t) __builtin_neon_vsri_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 18); \
26320 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
26321 __ret; \
26322 })
26323 #endif
26324
26325 #define vsri_n_u64(__p0, __p1, __p2) __extension__ ({ \
26326 uint64x1_t __ret; \
26327 uint64x1_t __s0 = __p0; \
26328 uint64x1_t __s1 = __p1; \
26329 __ret = (uint64x1_t) __builtin_neon_vsri_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 19); \
26330 __ret; \
26331 })
26332 #ifdef __LITTLE_ENDIAN__
26333 #define vsri_n_u16(__p0, __p1, __p2) __extension__ ({ \
26334 uint16x4_t __ret; \
26335 uint16x4_t __s0 = __p0; \
26336 uint16x4_t __s1 = __p1; \
26337 __ret = (uint16x4_t) __builtin_neon_vsri_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 17); \
26338 __ret; \
26339 })
26340 #else
26341 #define vsri_n_u16(__p0, __p1, __p2) __extension__ ({ \
26342 uint16x4_t __ret; \
26343 uint16x4_t __s0 = __p0; \
26344 uint16x4_t __s1 = __p1; \
26345 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
26346 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
26347 __ret = (uint16x4_t) __builtin_neon_vsri_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 17); \
26348 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
26349 __ret; \
26350 })
26351 #endif
26352
26353 #ifdef __LITTLE_ENDIAN__
26354 #define vsri_n_s8(__p0, __p1, __p2) __extension__ ({ \
26355 int8x8_t __ret; \
26356 int8x8_t __s0 = __p0; \
26357 int8x8_t __s1 = __p1; \
26358 __ret = (int8x8_t) __builtin_neon_vsri_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 0); \
26359 __ret; \
26360 })
26361 #else
26362 #define vsri_n_s8(__p0, __p1, __p2) __extension__ ({ \
26363 int8x8_t __ret; \
26364 int8x8_t __s0 = __p0; \
26365 int8x8_t __s1 = __p1; \
26366 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
26367 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
26368 __ret = (int8x8_t) __builtin_neon_vsri_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 0); \
26369 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
26370 __ret; \
26371 })
26372 #endif
26373
26374 #ifdef __LITTLE_ENDIAN__
26375 #define vsri_n_s32(__p0, __p1, __p2) __extension__ ({ \
26376 int32x2_t __ret; \
26377 int32x2_t __s0 = __p0; \
26378 int32x2_t __s1 = __p1; \
26379 __ret = (int32x2_t) __builtin_neon_vsri_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 2); \
26380 __ret; \
26381 })
26382 #else
26383 #define vsri_n_s32(__p0, __p1, __p2) __extension__ ({ \
26384 int32x2_t __ret; \
26385 int32x2_t __s0 = __p0; \
26386 int32x2_t __s1 = __p1; \
26387 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 1, 0); \
26388 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
26389 __ret = (int32x2_t) __builtin_neon_vsri_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 2); \
26390 __ret = __builtin_shufflevector(__ret, __ret, 1, 0); \
26391 __ret; \
26392 })
26393 #endif
26394
26395 #define vsri_n_s64(__p0, __p1, __p2) __extension__ ({ \
26396 int64x1_t __ret; \
26397 int64x1_t __s0 = __p0; \
26398 int64x1_t __s1 = __p1; \
26399 __ret = (int64x1_t) __builtin_neon_vsri_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 3); \
26400 __ret; \
26401 })
26402 #ifdef __LITTLE_ENDIAN__
26403 #define vsri_n_s16(__p0, __p1, __p2) __extension__ ({ \
26404 int16x4_t __ret; \
26405 int16x4_t __s0 = __p0; \
26406 int16x4_t __s1 = __p1; \
26407 __ret = (int16x4_t) __builtin_neon_vsri_n_v((int8x8_t)__s0, (int8x8_t)__s1, __p2, 1); \
26408 __ret; \
26409 })
26410 #else
26411 #define vsri_n_s16(__p0, __p1, __p2) __extension__ ({ \
26412 int16x4_t __ret; \
26413 int16x4_t __s0 = __p0; \
26414 int16x4_t __s1 = __p1; \
26415 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
26416 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
26417 __ret = (int16x4_t) __builtin_neon_vsri_n_v((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 1); \
26418 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
26419 __ret; \
26420 })
26421 #endif
26422
26423 #ifdef __LITTLE_ENDIAN__
26424 #define vst1_p8(__p0, __p1) __extension__ ({ \
26425 poly8x8_t __s1 = __p1; \
26426 __builtin_neon_vst1_v(__p0, (int8x8_t)__s1, 4); \
26427 })
26428 #else
26429 #define vst1_p8(__p0, __p1) __extension__ ({ \
26430 poly8x8_t __s1 = __p1; \
26431 poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
26432 __builtin_neon_vst1_v(__p0, (int8x8_t)__rev1, 4); \
26433 })
26434 #endif
26435
26436 #ifdef __LITTLE_ENDIAN__
26437 #define vst1_p16(__p0, __p1) __extension__ ({ \
26438 poly16x4_t __s1 = __p1; \
26439 __builtin_neon_vst1_v(__p0, (int8x8_t)__s1, 5); \
26440 })
26441 #else
26442 #define vst1_p16(__p0, __p1) __extension__ ({ \
26443 poly16x4_t __s1 = __p1; \
26444 poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
26445 __builtin_neon_vst1_v(__p0, (int8x8_t)__rev1, 5); \
26446 })
26447 #endif
26448
26449 #ifdef __LITTLE_ENDIAN__
26450 #define vst1q_p8(__p0, __p1) __extension__ ({ \
26451 poly8x16_t __s1 = __p1; \
26452 __builtin_neon_vst1q_v(__p0, (int8x16_t)__s1, 36); \
26453 })
26454 #else
26455 #define vst1q_p8(__p0, __p1) __extension__ ({ \
26456 poly8x16_t __s1 = __p1; \
26457 poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
26458 __builtin_neon_vst1q_v(__p0, (int8x16_t)__rev1, 36); \
26459 })
26460 #endif
26461
26462 #ifdef __LITTLE_ENDIAN__
26463 #define vst1q_p16(__p0, __p1) __extension__ ({ \
26464 poly16x8_t __s1 = __p1; \
26465 __builtin_neon_vst1q_v(__p0, (int8x16_t)__s1, 37); \
26466 })
26467 #else
26468 #define vst1q_p16(__p0, __p1) __extension__ ({ \
26469 poly16x8_t __s1 = __p1; \
26470 poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
26471 __builtin_neon_vst1q_v(__p0, (int8x16_t)__rev1, 37); \
26472 })
26473 #endif
26474
26475 #ifdef __LITTLE_ENDIAN__
26476 #define vst1q_u8(__p0, __p1) __extension__ ({ \
26477 uint8x16_t __s1 = __p1; \
26478 __builtin_neon_vst1q_v(__p0, (int8x16_t)__s1, 48); \
26479 })
26480 #else
26481 #define vst1q_u8(__p0, __p1) __extension__ ({ \
26482 uint8x16_t __s1 = __p1; \
26483 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
26484 __builtin_neon_vst1q_v(__p0, (int8x16_t)__rev1, 48); \
26485 })
26486 #endif
26487
26488 #ifdef __LITTLE_ENDIAN__
26489 #define vst1q_u32(__p0, __p1) __extension__ ({ \
26490 uint32x4_t __s1 = __p1; \
26491 __builtin_neon_vst1q_v(__p0, (int8x16_t)__s1, 50); \
26492 })
26493 #else
26494 #define vst1q_u32(__p0, __p1) __extension__ ({ \
26495 uint32x4_t __s1 = __p1; \
26496 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
26497 __builtin_neon_vst1q_v(__p0, (int8x16_t)__rev1, 50); \
26498 })
26499 #endif
26500
26501 #ifdef __LITTLE_ENDIAN__
26502 #define vst1q_u64(__p0, __p1) __extension__ ({ \
26503 uint64x2_t __s1 = __p1; \
26504 __builtin_neon_vst1q_v(__p0, (int8x16_t)__s1, 51); \
26505 })
26506 #else
26507 #define vst1q_u64(__p0, __p1) __extension__ ({ \
26508 uint64x2_t __s1 = __p1; \
26509 uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
26510 __builtin_neon_vst1q_v(__p0, (int8x16_t)__rev1, 51); \
26511 })
26512 #endif
26513
26514 #ifdef __LITTLE_ENDIAN__
26515 #define vst1q_u16(__p0, __p1) __extension__ ({ \
26516 uint16x8_t __s1 = __p1; \
26517 __builtin_neon_vst1q_v(__p0, (int8x16_t)__s1, 49); \
26518 })
26519 #else
26520 #define vst1q_u16(__p0, __p1) __extension__ ({ \
26521 uint16x8_t __s1 = __p1; \
26522 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
26523 __builtin_neon_vst1q_v(__p0, (int8x16_t)__rev1, 49); \
26524 })
26525 #endif
26526
26527 #ifdef __LITTLE_ENDIAN__
26528 #define vst1q_s8(__p0, __p1) __extension__ ({ \
26529 int8x16_t __s1 = __p1; \
26530 __builtin_neon_vst1q_v(__p0, (int8x16_t)__s1, 32); \
26531 })
26532 #else
26533 #define vst1q_s8(__p0, __p1) __extension__ ({ \
26534 int8x16_t __s1 = __p1; \
26535 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
26536 __builtin_neon_vst1q_v(__p0, (int8x16_t)__rev1, 32); \
26537 })
26538 #endif
26539
26540 #ifdef __LITTLE_ENDIAN__
26541 #define vst1q_f32(__p0, __p1) __extension__ ({ \
26542 float32x4_t __s1 = __p1; \
26543 __builtin_neon_vst1q_v(__p0, (int8x16_t)__s1, 41); \
26544 })
26545 #else
26546 #define vst1q_f32(__p0, __p1) __extension__ ({ \
26547 float32x4_t __s1 = __p1; \
26548 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
26549 __builtin_neon_vst1q_v(__p0, (int8x16_t)__rev1, 41); \
26550 })
26551 #endif
26552
26553 #ifdef __LITTLE_ENDIAN__
26554 #define vst1q_s32(__p0, __p1) __extension__ ({ \
26555 int32x4_t __s1 = __p1; \
26556 __builtin_neon_vst1q_v(__p0, (int8x16_t)__s1, 34); \
26557 })
26558 #else
26559 #define vst1q_s32(__p0, __p1) __extension__ ({ \
26560 int32x4_t __s1 = __p1; \
26561 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
26562 __builtin_neon_vst1q_v(__p0, (int8x16_t)__rev1, 34); \
26563 })
26564 #endif
26565
26566 #ifdef __LITTLE_ENDIAN__
26567 #define vst1q_s64(__p0, __p1) __extension__ ({ \
26568 int64x2_t __s1 = __p1; \
26569 __builtin_neon_vst1q_v(__p0, (int8x16_t)__s1, 35); \
26570 })
26571 #else
26572 #define vst1q_s64(__p0, __p1) __extension__ ({ \
26573 int64x2_t __s1 = __p1; \
26574 int64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
26575 __builtin_neon_vst1q_v(__p0, (int8x16_t)__rev1, 35); \
26576 })
26577 #endif
26578
26579 #ifdef __LITTLE_ENDIAN__
26580 #define vst1q_s16(__p0, __p1) __extension__ ({ \
26581 int16x8_t __s1 = __p1; \
26582 __builtin_neon_vst1q_v(__p0, (int8x16_t)__s1, 33); \
26583 })
26584 #else
26585 #define vst1q_s16(__p0, __p1) __extension__ ({ \
26586 int16x8_t __s1 = __p1; \
26587 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
26588 __builtin_neon_vst1q_v(__p0, (int8x16_t)__rev1, 33); \
26589 })
26590 #endif
26591
26592 #ifdef __LITTLE_ENDIAN__
26593 #define vst1_u8(__p0, __p1) __extension__ ({ \
26594 uint8x8_t __s1 = __p1; \
26595 __builtin_neon_vst1_v(__p0, (int8x8_t)__s1, 16); \
26596 })
26597 #else
26598 #define vst1_u8(__p0, __p1) __extension__ ({ \
26599 uint8x8_t __s1 = __p1; \
26600 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
26601 __builtin_neon_vst1_v(__p0, (int8x8_t)__rev1, 16); \
26602 })
26603 #endif
26604
26605 #ifdef __LITTLE_ENDIAN__
26606 #define vst1_u32(__p0, __p1) __extension__ ({ \
26607 uint32x2_t __s1 = __p1; \
26608 __builtin_neon_vst1_v(__p0, (int8x8_t)__s1, 18); \
26609 })
26610 #else
26611 #define vst1_u32(__p0, __p1) __extension__ ({ \
26612 uint32x2_t __s1 = __p1; \
26613 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
26614 __builtin_neon_vst1_v(__p0, (int8x8_t)__rev1, 18); \
26615 })
26616 #endif
26617
26618 #define vst1_u64(__p0, __p1) __extension__ ({ \
26619 uint64x1_t __s1 = __p1; \
26620 __builtin_neon_vst1_v(__p0, (int8x8_t)__s1, 19); \
26621 })
26622 #ifdef __LITTLE_ENDIAN__
26623 #define vst1_u16(__p0, __p1) __extension__ ({ \
26624 uint16x4_t __s1 = __p1; \
26625 __builtin_neon_vst1_v(__p0, (int8x8_t)__s1, 17); \
26626 })
26627 #else
26628 #define vst1_u16(__p0, __p1) __extension__ ({ \
26629 uint16x4_t __s1 = __p1; \
26630 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
26631 __builtin_neon_vst1_v(__p0, (int8x8_t)__rev1, 17); \
26632 })
26633 #endif
26634
26635 #ifdef __LITTLE_ENDIAN__
26636 #define vst1_s8(__p0, __p1) __extension__ ({ \
26637 int8x8_t __s1 = __p1; \
26638 __builtin_neon_vst1_v(__p0, (int8x8_t)__s1, 0); \
26639 })
26640 #else
26641 #define vst1_s8(__p0, __p1) __extension__ ({ \
26642 int8x8_t __s1 = __p1; \
26643 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
26644 __builtin_neon_vst1_v(__p0, (int8x8_t)__rev1, 0); \
26645 })
26646 #endif
26647
26648 #ifdef __LITTLE_ENDIAN__
26649 #define vst1_f32(__p0, __p1) __extension__ ({ \
26650 float32x2_t __s1 = __p1; \
26651 __builtin_neon_vst1_v(__p0, (int8x8_t)__s1, 9); \
26652 })
26653 #else
26654 #define vst1_f32(__p0, __p1) __extension__ ({ \
26655 float32x2_t __s1 = __p1; \
26656 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
26657 __builtin_neon_vst1_v(__p0, (int8x8_t)__rev1, 9); \
26658 })
26659 #endif
26660
26661 #ifdef __LITTLE_ENDIAN__
26662 #define vst1_s32(__p0, __p1) __extension__ ({ \
26663 int32x2_t __s1 = __p1; \
26664 __builtin_neon_vst1_v(__p0, (int8x8_t)__s1, 2); \
26665 })
26666 #else
26667 #define vst1_s32(__p0, __p1) __extension__ ({ \
26668 int32x2_t __s1 = __p1; \
26669 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
26670 __builtin_neon_vst1_v(__p0, (int8x8_t)__rev1, 2); \
26671 })
26672 #endif
26673
26674 #define vst1_s64(__p0, __p1) __extension__ ({ \
26675 int64x1_t __s1 = __p1; \
26676 __builtin_neon_vst1_v(__p0, (int8x8_t)__s1, 3); \
26677 })
26678 #ifdef __LITTLE_ENDIAN__
26679 #define vst1_s16(__p0, __p1) __extension__ ({ \
26680 int16x4_t __s1 = __p1; \
26681 __builtin_neon_vst1_v(__p0, (int8x8_t)__s1, 1); \
26682 })
26683 #else
26684 #define vst1_s16(__p0, __p1) __extension__ ({ \
26685 int16x4_t __s1 = __p1; \
26686 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
26687 __builtin_neon_vst1_v(__p0, (int8x8_t)__rev1, 1); \
26688 })
26689 #endif
26690
26691 #ifdef __LITTLE_ENDIAN__
26692 #define vst1_lane_p8(__p0, __p1, __p2) __extension__ ({ \
26693 poly8x8_t __s1 = __p1; \
26694 __builtin_neon_vst1_lane_v(__p0, (int8x8_t)__s1, __p2, 4); \
26695 })
26696 #else
26697 #define vst1_lane_p8(__p0, __p1, __p2) __extension__ ({ \
26698 poly8x8_t __s1 = __p1; \
26699 poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
26700 __builtin_neon_vst1_lane_v(__p0, (int8x8_t)__rev1, __p2, 4); \
26701 })
26702 #endif
26703
26704 #ifdef __LITTLE_ENDIAN__
26705 #define vst1_lane_p16(__p0, __p1, __p2) __extension__ ({ \
26706 poly16x4_t __s1 = __p1; \
26707 __builtin_neon_vst1_lane_v(__p0, (int8x8_t)__s1, __p2, 5); \
26708 })
26709 #else
26710 #define vst1_lane_p16(__p0, __p1, __p2) __extension__ ({ \
26711 poly16x4_t __s1 = __p1; \
26712 poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
26713 __builtin_neon_vst1_lane_v(__p0, (int8x8_t)__rev1, __p2, 5); \
26714 })
26715 #endif
26716
26717 #ifdef __LITTLE_ENDIAN__
26718 #define vst1q_lane_p8(__p0, __p1, __p2) __extension__ ({ \
26719 poly8x16_t __s1 = __p1; \
26720 __builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__s1, __p2, 36); \
26721 })
26722 #else
26723 #define vst1q_lane_p8(__p0, __p1, __p2) __extension__ ({ \
26724 poly8x16_t __s1 = __p1; \
26725 poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
26726 __builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 36); \
26727 })
26728 #endif
26729
26730 #ifdef __LITTLE_ENDIAN__
26731 #define vst1q_lane_p16(__p0, __p1, __p2) __extension__ ({ \
26732 poly16x8_t __s1 = __p1; \
26733 __builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__s1, __p2, 37); \
26734 })
26735 #else
26736 #define vst1q_lane_p16(__p0, __p1, __p2) __extension__ ({ \
26737 poly16x8_t __s1 = __p1; \
26738 poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
26739 __builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 37); \
26740 })
26741 #endif
26742
26743 #ifdef __LITTLE_ENDIAN__
26744 #define vst1q_lane_u8(__p0, __p1, __p2) __extension__ ({ \
26745 uint8x16_t __s1 = __p1; \
26746 __builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__s1, __p2, 48); \
26747 })
26748 #else
26749 #define vst1q_lane_u8(__p0, __p1, __p2) __extension__ ({ \
26750 uint8x16_t __s1 = __p1; \
26751 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
26752 __builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 48); \
26753 })
26754 #endif
26755
26756 #ifdef __LITTLE_ENDIAN__
26757 #define vst1q_lane_u32(__p0, __p1, __p2) __extension__ ({ \
26758 uint32x4_t __s1 = __p1; \
26759 __builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__s1, __p2, 50); \
26760 })
26761 #else
26762 #define vst1q_lane_u32(__p0, __p1, __p2) __extension__ ({ \
26763 uint32x4_t __s1 = __p1; \
26764 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
26765 __builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 50); \
26766 })
26767 #endif
26768
26769 #ifdef __LITTLE_ENDIAN__
26770 #define vst1q_lane_u64(__p0, __p1, __p2) __extension__ ({ \
26771 uint64x2_t __s1 = __p1; \
26772 __builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__s1, __p2, 51); \
26773 })
26774 #else
26775 #define vst1q_lane_u64(__p0, __p1, __p2) __extension__ ({ \
26776 uint64x2_t __s1 = __p1; \
26777 uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
26778 __builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 51); \
26779 })
26780 #endif
26781
26782 #ifdef __LITTLE_ENDIAN__
26783 #define vst1q_lane_u16(__p0, __p1, __p2) __extension__ ({ \
26784 uint16x8_t __s1 = __p1; \
26785 __builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__s1, __p2, 49); \
26786 })
26787 #else
26788 #define vst1q_lane_u16(__p0, __p1, __p2) __extension__ ({ \
26789 uint16x8_t __s1 = __p1; \
26790 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
26791 __builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 49); \
26792 })
26793 #endif
26794
26795 #ifdef __LITTLE_ENDIAN__
26796 #define vst1q_lane_s8(__p0, __p1, __p2) __extension__ ({ \
26797 int8x16_t __s1 = __p1; \
26798 __builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__s1, __p2, 32); \
26799 })
26800 #else
26801 #define vst1q_lane_s8(__p0, __p1, __p2) __extension__ ({ \
26802 int8x16_t __s1 = __p1; \
26803 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
26804 __builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 32); \
26805 })
26806 #endif
26807
26808 #ifdef __LITTLE_ENDIAN__
26809 #define vst1q_lane_f32(__p0, __p1, __p2) __extension__ ({ \
26810 float32x4_t __s1 = __p1; \
26811 __builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__s1, __p2, 41); \
26812 })
26813 #else
26814 #define vst1q_lane_f32(__p0, __p1, __p2) __extension__ ({ \
26815 float32x4_t __s1 = __p1; \
26816 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
26817 __builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 41); \
26818 })
26819 #endif
26820
26821 #ifdef __LITTLE_ENDIAN__
26822 #define vst1q_lane_s32(__p0, __p1, __p2) __extension__ ({ \
26823 int32x4_t __s1 = __p1; \
26824 __builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__s1, __p2, 34); \
26825 })
26826 #else
26827 #define vst1q_lane_s32(__p0, __p1, __p2) __extension__ ({ \
26828 int32x4_t __s1 = __p1; \
26829 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
26830 __builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 34); \
26831 })
26832 #endif
26833
26834 #ifdef __LITTLE_ENDIAN__
26835 #define vst1q_lane_s64(__p0, __p1, __p2) __extension__ ({ \
26836 int64x2_t __s1 = __p1; \
26837 __builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__s1, __p2, 35); \
26838 })
26839 #else
26840 #define vst1q_lane_s64(__p0, __p1, __p2) __extension__ ({ \
26841 int64x2_t __s1 = __p1; \
26842 int64x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
26843 __builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 35); \
26844 })
26845 #endif
26846
26847 #ifdef __LITTLE_ENDIAN__
26848 #define vst1q_lane_s16(__p0, __p1, __p2) __extension__ ({ \
26849 int16x8_t __s1 = __p1; \
26850 __builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__s1, __p2, 33); \
26851 })
26852 #else
26853 #define vst1q_lane_s16(__p0, __p1, __p2) __extension__ ({ \
26854 int16x8_t __s1 = __p1; \
26855 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
26856 __builtin_neon_vst1q_lane_v(__p0, (int8x16_t)__rev1, __p2, 33); \
26857 })
26858 #endif
26859
26860 #ifdef __LITTLE_ENDIAN__
26861 #define vst1_lane_u8(__p0, __p1, __p2) __extension__ ({ \
26862 uint8x8_t __s1 = __p1; \
26863 __builtin_neon_vst1_lane_v(__p0, (int8x8_t)__s1, __p2, 16); \
26864 })
26865 #else
26866 #define vst1_lane_u8(__p0, __p1, __p2) __extension__ ({ \
26867 uint8x8_t __s1 = __p1; \
26868 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
26869 __builtin_neon_vst1_lane_v(__p0, (int8x8_t)__rev1, __p2, 16); \
26870 })
26871 #endif
26872
26873 #ifdef __LITTLE_ENDIAN__
26874 #define vst1_lane_u32(__p0, __p1, __p2) __extension__ ({ \
26875 uint32x2_t __s1 = __p1; \
26876 __builtin_neon_vst1_lane_v(__p0, (int8x8_t)__s1, __p2, 18); \
26877 })
26878 #else
26879 #define vst1_lane_u32(__p0, __p1, __p2) __extension__ ({ \
26880 uint32x2_t __s1 = __p1; \
26881 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
26882 __builtin_neon_vst1_lane_v(__p0, (int8x8_t)__rev1, __p2, 18); \
26883 })
26884 #endif
26885
26886 #define vst1_lane_u64(__p0, __p1, __p2) __extension__ ({ \
26887 uint64x1_t __s1 = __p1; \
26888 __builtin_neon_vst1_lane_v(__p0, (int8x8_t)__s1, __p2, 19); \
26889 })
26890 #ifdef __LITTLE_ENDIAN__
26891 #define vst1_lane_u16(__p0, __p1, __p2) __extension__ ({ \
26892 uint16x4_t __s1 = __p1; \
26893 __builtin_neon_vst1_lane_v(__p0, (int8x8_t)__s1, __p2, 17); \
26894 })
26895 #else
26896 #define vst1_lane_u16(__p0, __p1, __p2) __extension__ ({ \
26897 uint16x4_t __s1 = __p1; \
26898 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
26899 __builtin_neon_vst1_lane_v(__p0, (int8x8_t)__rev1, __p2, 17); \
26900 })
26901 #endif
26902
26903 #ifdef __LITTLE_ENDIAN__
26904 #define vst1_lane_s8(__p0, __p1, __p2) __extension__ ({ \
26905 int8x8_t __s1 = __p1; \
26906 __builtin_neon_vst1_lane_v(__p0, (int8x8_t)__s1, __p2, 0); \
26907 })
26908 #else
26909 #define vst1_lane_s8(__p0, __p1, __p2) __extension__ ({ \
26910 int8x8_t __s1 = __p1; \
26911 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
26912 __builtin_neon_vst1_lane_v(__p0, (int8x8_t)__rev1, __p2, 0); \
26913 })
26914 #endif
26915
26916 #ifdef __LITTLE_ENDIAN__
26917 #define vst1_lane_f32(__p0, __p1, __p2) __extension__ ({ \
26918 float32x2_t __s1 = __p1; \
26919 __builtin_neon_vst1_lane_v(__p0, (int8x8_t)__s1, __p2, 9); \
26920 })
26921 #else
26922 #define vst1_lane_f32(__p0, __p1, __p2) __extension__ ({ \
26923 float32x2_t __s1 = __p1; \
26924 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
26925 __builtin_neon_vst1_lane_v(__p0, (int8x8_t)__rev1, __p2, 9); \
26926 })
26927 #endif
26928
26929 #ifdef __LITTLE_ENDIAN__
26930 #define vst1_lane_s32(__p0, __p1, __p2) __extension__ ({ \
26931 int32x2_t __s1 = __p1; \
26932 __builtin_neon_vst1_lane_v(__p0, (int8x8_t)__s1, __p2, 2); \
26933 })
26934 #else
26935 #define vst1_lane_s32(__p0, __p1, __p2) __extension__ ({ \
26936 int32x2_t __s1 = __p1; \
26937 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 1, 0); \
26938 __builtin_neon_vst1_lane_v(__p0, (int8x8_t)__rev1, __p2, 2); \
26939 })
26940 #endif
26941
26942 #define vst1_lane_s64(__p0, __p1, __p2) __extension__ ({ \
26943 int64x1_t __s1 = __p1; \
26944 __builtin_neon_vst1_lane_v(__p0, (int8x8_t)__s1, __p2, 3); \
26945 })
26946 #ifdef __LITTLE_ENDIAN__
26947 #define vst1_lane_s16(__p0, __p1, __p2) __extension__ ({ \
26948 int16x4_t __s1 = __p1; \
26949 __builtin_neon_vst1_lane_v(__p0, (int8x8_t)__s1, __p2, 1); \
26950 })
26951 #else
26952 #define vst1_lane_s16(__p0, __p1, __p2) __extension__ ({ \
26953 int16x4_t __s1 = __p1; \
26954 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
26955 __builtin_neon_vst1_lane_v(__p0, (int8x8_t)__rev1, __p2, 1); \
26956 })
26957 #endif
26958
26959 #ifdef __LITTLE_ENDIAN__
26960 #define vst1_p8_x2(__p0, __p1) __extension__ ({ \
26961 poly8x8x2_t __s1 = __p1; \
26962 __builtin_neon_vst1_x2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 4); \
26963 })
26964 #else
26965 #define vst1_p8_x2(__p0, __p1) __extension__ ({ \
26966 poly8x8x2_t __s1 = __p1; \
26967 poly8x8x2_t __rev1; \
26968 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
26969 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
26970 __builtin_neon_vst1_x2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 4); \
26971 })
26972 #endif
26973
26974 #ifdef __LITTLE_ENDIAN__
26975 #define vst1_p16_x2(__p0, __p1) __extension__ ({ \
26976 poly16x4x2_t __s1 = __p1; \
26977 __builtin_neon_vst1_x2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 5); \
26978 })
26979 #else
26980 #define vst1_p16_x2(__p0, __p1) __extension__ ({ \
26981 poly16x4x2_t __s1 = __p1; \
26982 poly16x4x2_t __rev1; \
26983 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
26984 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
26985 __builtin_neon_vst1_x2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 5); \
26986 })
26987 #endif
26988
26989 #ifdef __LITTLE_ENDIAN__
26990 #define vst1q_p8_x2(__p0, __p1) __extension__ ({ \
26991 poly8x16x2_t __s1 = __p1; \
26992 __builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 36); \
26993 })
26994 #else
26995 #define vst1q_p8_x2(__p0, __p1) __extension__ ({ \
26996 poly8x16x2_t __s1 = __p1; \
26997 poly8x16x2_t __rev1; \
26998 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
26999 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
27000 __builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 36); \
27001 })
27002 #endif
27003
27004 #ifdef __LITTLE_ENDIAN__
27005 #define vst1q_p16_x2(__p0, __p1) __extension__ ({ \
27006 poly16x8x2_t __s1 = __p1; \
27007 __builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 37); \
27008 })
27009 #else
27010 #define vst1q_p16_x2(__p0, __p1) __extension__ ({ \
27011 poly16x8x2_t __s1 = __p1; \
27012 poly16x8x2_t __rev1; \
27013 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
27014 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
27015 __builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 37); \
27016 })
27017 #endif
27018
27019 #ifdef __LITTLE_ENDIAN__
27020 #define vst1q_u8_x2(__p0, __p1) __extension__ ({ \
27021 uint8x16x2_t __s1 = __p1; \
27022 __builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 48); \
27023 })
27024 #else
27025 #define vst1q_u8_x2(__p0, __p1) __extension__ ({ \
27026 uint8x16x2_t __s1 = __p1; \
27027 uint8x16x2_t __rev1; \
27028 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
27029 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
27030 __builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 48); \
27031 })
27032 #endif
27033
27034 #ifdef __LITTLE_ENDIAN__
27035 #define vst1q_u32_x2(__p0, __p1) __extension__ ({ \
27036 uint32x4x2_t __s1 = __p1; \
27037 __builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 50); \
27038 })
27039 #else
27040 #define vst1q_u32_x2(__p0, __p1) __extension__ ({ \
27041 uint32x4x2_t __s1 = __p1; \
27042 uint32x4x2_t __rev1; \
27043 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
27044 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
27045 __builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 50); \
27046 })
27047 #endif
27048
27049 #ifdef __LITTLE_ENDIAN__
27050 #define vst1q_u64_x2(__p0, __p1) __extension__ ({ \
27051 uint64x2x2_t __s1 = __p1; \
27052 __builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 51); \
27053 })
27054 #else
27055 #define vst1q_u64_x2(__p0, __p1) __extension__ ({ \
27056 uint64x2x2_t __s1 = __p1; \
27057 uint64x2x2_t __rev1; \
27058 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
27059 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
27060 __builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 51); \
27061 })
27062 #endif
27063
27064 #ifdef __LITTLE_ENDIAN__
27065 #define vst1q_u16_x2(__p0, __p1) __extension__ ({ \
27066 uint16x8x2_t __s1 = __p1; \
27067 __builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 49); \
27068 })
27069 #else
27070 #define vst1q_u16_x2(__p0, __p1) __extension__ ({ \
27071 uint16x8x2_t __s1 = __p1; \
27072 uint16x8x2_t __rev1; \
27073 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
27074 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
27075 __builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 49); \
27076 })
27077 #endif
27078
27079 #ifdef __LITTLE_ENDIAN__
27080 #define vst1q_s8_x2(__p0, __p1) __extension__ ({ \
27081 int8x16x2_t __s1 = __p1; \
27082 __builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 32); \
27083 })
27084 #else
27085 #define vst1q_s8_x2(__p0, __p1) __extension__ ({ \
27086 int8x16x2_t __s1 = __p1; \
27087 int8x16x2_t __rev1; \
27088 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
27089 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
27090 __builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 32); \
27091 })
27092 #endif
27093
27094 #ifdef __LITTLE_ENDIAN__
27095 #define vst1q_f32_x2(__p0, __p1) __extension__ ({ \
27096 float32x4x2_t __s1 = __p1; \
27097 __builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 41); \
27098 })
27099 #else
27100 #define vst1q_f32_x2(__p0, __p1) __extension__ ({ \
27101 float32x4x2_t __s1 = __p1; \
27102 float32x4x2_t __rev1; \
27103 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
27104 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
27105 __builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 41); \
27106 })
27107 #endif
27108
27109 #ifdef __LITTLE_ENDIAN__
27110 #define vst1q_s32_x2(__p0, __p1) __extension__ ({ \
27111 int32x4x2_t __s1 = __p1; \
27112 __builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 34); \
27113 })
27114 #else
27115 #define vst1q_s32_x2(__p0, __p1) __extension__ ({ \
27116 int32x4x2_t __s1 = __p1; \
27117 int32x4x2_t __rev1; \
27118 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
27119 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
27120 __builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 34); \
27121 })
27122 #endif
27123
27124 #ifdef __LITTLE_ENDIAN__
27125 #define vst1q_s64_x2(__p0, __p1) __extension__ ({ \
27126 int64x2x2_t __s1 = __p1; \
27127 __builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 35); \
27128 })
27129 #else
27130 #define vst1q_s64_x2(__p0, __p1) __extension__ ({ \
27131 int64x2x2_t __s1 = __p1; \
27132 int64x2x2_t __rev1; \
27133 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
27134 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
27135 __builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 35); \
27136 })
27137 #endif
27138
27139 #ifdef __LITTLE_ENDIAN__
27140 #define vst1q_s16_x2(__p0, __p1) __extension__ ({ \
27141 int16x8x2_t __s1 = __p1; \
27142 __builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 33); \
27143 })
27144 #else
27145 #define vst1q_s16_x2(__p0, __p1) __extension__ ({ \
27146 int16x8x2_t __s1 = __p1; \
27147 int16x8x2_t __rev1; \
27148 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
27149 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
27150 __builtin_neon_vst1q_x2_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 33); \
27151 })
27152 #endif
27153
27154 #ifdef __LITTLE_ENDIAN__
27155 #define vst1_u8_x2(__p0, __p1) __extension__ ({ \
27156 uint8x8x2_t __s1 = __p1; \
27157 __builtin_neon_vst1_x2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 16); \
27158 })
27159 #else
27160 #define vst1_u8_x2(__p0, __p1) __extension__ ({ \
27161 uint8x8x2_t __s1 = __p1; \
27162 uint8x8x2_t __rev1; \
27163 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
27164 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
27165 __builtin_neon_vst1_x2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 16); \
27166 })
27167 #endif
27168
27169 #ifdef __LITTLE_ENDIAN__
27170 #define vst1_u32_x2(__p0, __p1) __extension__ ({ \
27171 uint32x2x2_t __s1 = __p1; \
27172 __builtin_neon_vst1_x2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 18); \
27173 })
27174 #else
27175 #define vst1_u32_x2(__p0, __p1) __extension__ ({ \
27176 uint32x2x2_t __s1 = __p1; \
27177 uint32x2x2_t __rev1; \
27178 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
27179 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
27180 __builtin_neon_vst1_x2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 18); \
27181 })
27182 #endif
27183
27184 #define vst1_u64_x2(__p0, __p1) __extension__ ({ \
27185 uint64x1x2_t __s1 = __p1; \
27186 __builtin_neon_vst1_x2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 19); \
27187 })
27188 #ifdef __LITTLE_ENDIAN__
27189 #define vst1_u16_x2(__p0, __p1) __extension__ ({ \
27190 uint16x4x2_t __s1 = __p1; \
27191 __builtin_neon_vst1_x2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 17); \
27192 })
27193 #else
27194 #define vst1_u16_x2(__p0, __p1) __extension__ ({ \
27195 uint16x4x2_t __s1 = __p1; \
27196 uint16x4x2_t __rev1; \
27197 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
27198 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
27199 __builtin_neon_vst1_x2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 17); \
27200 })
27201 #endif
27202
27203 #ifdef __LITTLE_ENDIAN__
27204 #define vst1_s8_x2(__p0, __p1) __extension__ ({ \
27205 int8x8x2_t __s1 = __p1; \
27206 __builtin_neon_vst1_x2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 0); \
27207 })
27208 #else
27209 #define vst1_s8_x2(__p0, __p1) __extension__ ({ \
27210 int8x8x2_t __s1 = __p1; \
27211 int8x8x2_t __rev1; \
27212 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
27213 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
27214 __builtin_neon_vst1_x2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 0); \
27215 })
27216 #endif
27217
27218 #ifdef __LITTLE_ENDIAN__
27219 #define vst1_f32_x2(__p0, __p1) __extension__ ({ \
27220 float32x2x2_t __s1 = __p1; \
27221 __builtin_neon_vst1_x2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 9); \
27222 })
27223 #else
27224 #define vst1_f32_x2(__p0, __p1) __extension__ ({ \
27225 float32x2x2_t __s1 = __p1; \
27226 float32x2x2_t __rev1; \
27227 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
27228 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
27229 __builtin_neon_vst1_x2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 9); \
27230 })
27231 #endif
27232
27233 #ifdef __LITTLE_ENDIAN__
27234 #define vst1_s32_x2(__p0, __p1) __extension__ ({ \
27235 int32x2x2_t __s1 = __p1; \
27236 __builtin_neon_vst1_x2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 2); \
27237 })
27238 #else
27239 #define vst1_s32_x2(__p0, __p1) __extension__ ({ \
27240 int32x2x2_t __s1 = __p1; \
27241 int32x2x2_t __rev1; \
27242 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
27243 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
27244 __builtin_neon_vst1_x2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 2); \
27245 })
27246 #endif
27247
27248 #define vst1_s64_x2(__p0, __p1) __extension__ ({ \
27249 int64x1x2_t __s1 = __p1; \
27250 __builtin_neon_vst1_x2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 3); \
27251 })
27252 #ifdef __LITTLE_ENDIAN__
27253 #define vst1_s16_x2(__p0, __p1) __extension__ ({ \
27254 int16x4x2_t __s1 = __p1; \
27255 __builtin_neon_vst1_x2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 1); \
27256 })
27257 #else
27258 #define vst1_s16_x2(__p0, __p1) __extension__ ({ \
27259 int16x4x2_t __s1 = __p1; \
27260 int16x4x2_t __rev1; \
27261 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
27262 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
27263 __builtin_neon_vst1_x2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 1); \
27264 })
27265 #endif
27266
27267 #ifdef __LITTLE_ENDIAN__
27268 #define vst1_p8_x3(__p0, __p1) __extension__ ({ \
27269 poly8x8x3_t __s1 = __p1; \
27270 __builtin_neon_vst1_x3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 4); \
27271 })
27272 #else
27273 #define vst1_p8_x3(__p0, __p1) __extension__ ({ \
27274 poly8x8x3_t __s1 = __p1; \
27275 poly8x8x3_t __rev1; \
27276 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
27277 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
27278 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
27279 __builtin_neon_vst1_x3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 4); \
27280 })
27281 #endif
27282
27283 #ifdef __LITTLE_ENDIAN__
27284 #define vst1_p16_x3(__p0, __p1) __extension__ ({ \
27285 poly16x4x3_t __s1 = __p1; \
27286 __builtin_neon_vst1_x3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 5); \
27287 })
27288 #else
27289 #define vst1_p16_x3(__p0, __p1) __extension__ ({ \
27290 poly16x4x3_t __s1 = __p1; \
27291 poly16x4x3_t __rev1; \
27292 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
27293 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
27294 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
27295 __builtin_neon_vst1_x3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 5); \
27296 })
27297 #endif
27298
27299 #ifdef __LITTLE_ENDIAN__
27300 #define vst1q_p8_x3(__p0, __p1) __extension__ ({ \
27301 poly8x16x3_t __s1 = __p1; \
27302 __builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 36); \
27303 })
27304 #else
27305 #define vst1q_p8_x3(__p0, __p1) __extension__ ({ \
27306 poly8x16x3_t __s1 = __p1; \
27307 poly8x16x3_t __rev1; \
27308 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
27309 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
27310 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
27311 __builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 36); \
27312 })
27313 #endif
27314
27315 #ifdef __LITTLE_ENDIAN__
27316 #define vst1q_p16_x3(__p0, __p1) __extension__ ({ \
27317 poly16x8x3_t __s1 = __p1; \
27318 __builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 37); \
27319 })
27320 #else
27321 #define vst1q_p16_x3(__p0, __p1) __extension__ ({ \
27322 poly16x8x3_t __s1 = __p1; \
27323 poly16x8x3_t __rev1; \
27324 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
27325 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
27326 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
27327 __builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 37); \
27328 })
27329 #endif
27330
27331 #ifdef __LITTLE_ENDIAN__
27332 #define vst1q_u8_x3(__p0, __p1) __extension__ ({ \
27333 uint8x16x3_t __s1 = __p1; \
27334 __builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 48); \
27335 })
27336 #else
27337 #define vst1q_u8_x3(__p0, __p1) __extension__ ({ \
27338 uint8x16x3_t __s1 = __p1; \
27339 uint8x16x3_t __rev1; \
27340 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
27341 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
27342 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
27343 __builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 48); \
27344 })
27345 #endif
27346
27347 #ifdef __LITTLE_ENDIAN__
27348 #define vst1q_u32_x3(__p0, __p1) __extension__ ({ \
27349 uint32x4x3_t __s1 = __p1; \
27350 __builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 50); \
27351 })
27352 #else
27353 #define vst1q_u32_x3(__p0, __p1) __extension__ ({ \
27354 uint32x4x3_t __s1 = __p1; \
27355 uint32x4x3_t __rev1; \
27356 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
27357 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
27358 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
27359 __builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 50); \
27360 })
27361 #endif
27362
27363 #ifdef __LITTLE_ENDIAN__
27364 #define vst1q_u64_x3(__p0, __p1) __extension__ ({ \
27365 uint64x2x3_t __s1 = __p1; \
27366 __builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 51); \
27367 })
27368 #else
27369 #define vst1q_u64_x3(__p0, __p1) __extension__ ({ \
27370 uint64x2x3_t __s1 = __p1; \
27371 uint64x2x3_t __rev1; \
27372 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
27373 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
27374 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
27375 __builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 51); \
27376 })
27377 #endif
27378
27379 #ifdef __LITTLE_ENDIAN__
27380 #define vst1q_u16_x3(__p0, __p1) __extension__ ({ \
27381 uint16x8x3_t __s1 = __p1; \
27382 __builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 49); \
27383 })
27384 #else
27385 #define vst1q_u16_x3(__p0, __p1) __extension__ ({ \
27386 uint16x8x3_t __s1 = __p1; \
27387 uint16x8x3_t __rev1; \
27388 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
27389 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
27390 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
27391 __builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 49); \
27392 })
27393 #endif
27394
27395 #ifdef __LITTLE_ENDIAN__
27396 #define vst1q_s8_x3(__p0, __p1) __extension__ ({ \
27397 int8x16x3_t __s1 = __p1; \
27398 __builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 32); \
27399 })
27400 #else
27401 #define vst1q_s8_x3(__p0, __p1) __extension__ ({ \
27402 int8x16x3_t __s1 = __p1; \
27403 int8x16x3_t __rev1; \
27404 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
27405 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
27406 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
27407 __builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 32); \
27408 })
27409 #endif
27410
27411 #ifdef __LITTLE_ENDIAN__
27412 #define vst1q_f32_x3(__p0, __p1) __extension__ ({ \
27413 float32x4x3_t __s1 = __p1; \
27414 __builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 41); \
27415 })
27416 #else
27417 #define vst1q_f32_x3(__p0, __p1) __extension__ ({ \
27418 float32x4x3_t __s1 = __p1; \
27419 float32x4x3_t __rev1; \
27420 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
27421 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
27422 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
27423 __builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 41); \
27424 })
27425 #endif
27426
27427 #ifdef __LITTLE_ENDIAN__
27428 #define vst1q_s32_x3(__p0, __p1) __extension__ ({ \
27429 int32x4x3_t __s1 = __p1; \
27430 __builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 34); \
27431 })
27432 #else
27433 #define vst1q_s32_x3(__p0, __p1) __extension__ ({ \
27434 int32x4x3_t __s1 = __p1; \
27435 int32x4x3_t __rev1; \
27436 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
27437 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
27438 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
27439 __builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 34); \
27440 })
27441 #endif
27442
27443 #ifdef __LITTLE_ENDIAN__
27444 #define vst1q_s64_x3(__p0, __p1) __extension__ ({ \
27445 int64x2x3_t __s1 = __p1; \
27446 __builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 35); \
27447 })
27448 #else
27449 #define vst1q_s64_x3(__p0, __p1) __extension__ ({ \
27450 int64x2x3_t __s1 = __p1; \
27451 int64x2x3_t __rev1; \
27452 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
27453 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
27454 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
27455 __builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 35); \
27456 })
27457 #endif
27458
27459 #ifdef __LITTLE_ENDIAN__
27460 #define vst1q_s16_x3(__p0, __p1) __extension__ ({ \
27461 int16x8x3_t __s1 = __p1; \
27462 __builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 33); \
27463 })
27464 #else
27465 #define vst1q_s16_x3(__p0, __p1) __extension__ ({ \
27466 int16x8x3_t __s1 = __p1; \
27467 int16x8x3_t __rev1; \
27468 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
27469 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
27470 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
27471 __builtin_neon_vst1q_x3_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 33); \
27472 })
27473 #endif
27474
27475 #ifdef __LITTLE_ENDIAN__
27476 #define vst1_u8_x3(__p0, __p1) __extension__ ({ \
27477 uint8x8x3_t __s1 = __p1; \
27478 __builtin_neon_vst1_x3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 16); \
27479 })
27480 #else
27481 #define vst1_u8_x3(__p0, __p1) __extension__ ({ \
27482 uint8x8x3_t __s1 = __p1; \
27483 uint8x8x3_t __rev1; \
27484 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
27485 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
27486 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
27487 __builtin_neon_vst1_x3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 16); \
27488 })
27489 #endif
27490
27491 #ifdef __LITTLE_ENDIAN__
27492 #define vst1_u32_x3(__p0, __p1) __extension__ ({ \
27493 uint32x2x3_t __s1 = __p1; \
27494 __builtin_neon_vst1_x3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 18); \
27495 })
27496 #else
27497 #define vst1_u32_x3(__p0, __p1) __extension__ ({ \
27498 uint32x2x3_t __s1 = __p1; \
27499 uint32x2x3_t __rev1; \
27500 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
27501 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
27502 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
27503 __builtin_neon_vst1_x3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 18); \
27504 })
27505 #endif
27506
27507 #define vst1_u64_x3(__p0, __p1) __extension__ ({ \
27508 uint64x1x3_t __s1 = __p1; \
27509 __builtin_neon_vst1_x3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 19); \
27510 })
27511 #ifdef __LITTLE_ENDIAN__
27512 #define vst1_u16_x3(__p0, __p1) __extension__ ({ \
27513 uint16x4x3_t __s1 = __p1; \
27514 __builtin_neon_vst1_x3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 17); \
27515 })
27516 #else
27517 #define vst1_u16_x3(__p0, __p1) __extension__ ({ \
27518 uint16x4x3_t __s1 = __p1; \
27519 uint16x4x3_t __rev1; \
27520 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
27521 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
27522 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
27523 __builtin_neon_vst1_x3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 17); \
27524 })
27525 #endif
27526
27527 #ifdef __LITTLE_ENDIAN__
27528 #define vst1_s8_x3(__p0, __p1) __extension__ ({ \
27529 int8x8x3_t __s1 = __p1; \
27530 __builtin_neon_vst1_x3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 0); \
27531 })
27532 #else
27533 #define vst1_s8_x3(__p0, __p1) __extension__ ({ \
27534 int8x8x3_t __s1 = __p1; \
27535 int8x8x3_t __rev1; \
27536 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
27537 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
27538 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
27539 __builtin_neon_vst1_x3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 0); \
27540 })
27541 #endif
27542
27543 #ifdef __LITTLE_ENDIAN__
27544 #define vst1_f32_x3(__p0, __p1) __extension__ ({ \
27545 float32x2x3_t __s1 = __p1; \
27546 __builtin_neon_vst1_x3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 9); \
27547 })
27548 #else
27549 #define vst1_f32_x3(__p0, __p1) __extension__ ({ \
27550 float32x2x3_t __s1 = __p1; \
27551 float32x2x3_t __rev1; \
27552 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
27553 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
27554 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
27555 __builtin_neon_vst1_x3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 9); \
27556 })
27557 #endif
27558
27559 #ifdef __LITTLE_ENDIAN__
27560 #define vst1_s32_x3(__p0, __p1) __extension__ ({ \
27561 int32x2x3_t __s1 = __p1; \
27562 __builtin_neon_vst1_x3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 2); \
27563 })
27564 #else
27565 #define vst1_s32_x3(__p0, __p1) __extension__ ({ \
27566 int32x2x3_t __s1 = __p1; \
27567 int32x2x3_t __rev1; \
27568 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
27569 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
27570 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
27571 __builtin_neon_vst1_x3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 2); \
27572 })
27573 #endif
27574
27575 #define vst1_s64_x3(__p0, __p1) __extension__ ({ \
27576 int64x1x3_t __s1 = __p1; \
27577 __builtin_neon_vst1_x3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 3); \
27578 })
27579 #ifdef __LITTLE_ENDIAN__
27580 #define vst1_s16_x3(__p0, __p1) __extension__ ({ \
27581 int16x4x3_t __s1 = __p1; \
27582 __builtin_neon_vst1_x3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 1); \
27583 })
27584 #else
27585 #define vst1_s16_x3(__p0, __p1) __extension__ ({ \
27586 int16x4x3_t __s1 = __p1; \
27587 int16x4x3_t __rev1; \
27588 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
27589 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
27590 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
27591 __builtin_neon_vst1_x3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 1); \
27592 })
27593 #endif
27594
27595 #ifdef __LITTLE_ENDIAN__
27596 #define vst1_p8_x4(__p0, __p1) __extension__ ({ \
27597 poly8x8x4_t __s1 = __p1; \
27598 __builtin_neon_vst1_x4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 4); \
27599 })
27600 #else
27601 #define vst1_p8_x4(__p0, __p1) __extension__ ({ \
27602 poly8x8x4_t __s1 = __p1; \
27603 poly8x8x4_t __rev1; \
27604 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
27605 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
27606 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
27607 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
27608 __builtin_neon_vst1_x4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 4); \
27609 })
27610 #endif
27611
27612 #ifdef __LITTLE_ENDIAN__
27613 #define vst1_p16_x4(__p0, __p1) __extension__ ({ \
27614 poly16x4x4_t __s1 = __p1; \
27615 __builtin_neon_vst1_x4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 5); \
27616 })
27617 #else
27618 #define vst1_p16_x4(__p0, __p1) __extension__ ({ \
27619 poly16x4x4_t __s1 = __p1; \
27620 poly16x4x4_t __rev1; \
27621 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
27622 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
27623 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
27624 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
27625 __builtin_neon_vst1_x4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 5); \
27626 })
27627 #endif
27628
27629 #ifdef __LITTLE_ENDIAN__
27630 #define vst1q_p8_x4(__p0, __p1) __extension__ ({ \
27631 poly8x16x4_t __s1 = __p1; \
27632 __builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 36); \
27633 })
27634 #else
27635 #define vst1q_p8_x4(__p0, __p1) __extension__ ({ \
27636 poly8x16x4_t __s1 = __p1; \
27637 poly8x16x4_t __rev1; \
27638 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
27639 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
27640 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
27641 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
27642 __builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 36); \
27643 })
27644 #endif
27645
27646 #ifdef __LITTLE_ENDIAN__
27647 #define vst1q_p16_x4(__p0, __p1) __extension__ ({ \
27648 poly16x8x4_t __s1 = __p1; \
27649 __builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 37); \
27650 })
27651 #else
27652 #define vst1q_p16_x4(__p0, __p1) __extension__ ({ \
27653 poly16x8x4_t __s1 = __p1; \
27654 poly16x8x4_t __rev1; \
27655 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
27656 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
27657 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
27658 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
27659 __builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 37); \
27660 })
27661 #endif
27662
27663 #ifdef __LITTLE_ENDIAN__
27664 #define vst1q_u8_x4(__p0, __p1) __extension__ ({ \
27665 uint8x16x4_t __s1 = __p1; \
27666 __builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 48); \
27667 })
27668 #else
27669 #define vst1q_u8_x4(__p0, __p1) __extension__ ({ \
27670 uint8x16x4_t __s1 = __p1; \
27671 uint8x16x4_t __rev1; \
27672 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
27673 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
27674 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
27675 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
27676 __builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 48); \
27677 })
27678 #endif
27679
27680 #ifdef __LITTLE_ENDIAN__
27681 #define vst1q_u32_x4(__p0, __p1) __extension__ ({ \
27682 uint32x4x4_t __s1 = __p1; \
27683 __builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 50); \
27684 })
27685 #else
27686 #define vst1q_u32_x4(__p0, __p1) __extension__ ({ \
27687 uint32x4x4_t __s1 = __p1; \
27688 uint32x4x4_t __rev1; \
27689 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
27690 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
27691 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
27692 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
27693 __builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 50); \
27694 })
27695 #endif
27696
27697 #ifdef __LITTLE_ENDIAN__
27698 #define vst1q_u64_x4(__p0, __p1) __extension__ ({ \
27699 uint64x2x4_t __s1 = __p1; \
27700 __builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 51); \
27701 })
27702 #else
27703 #define vst1q_u64_x4(__p0, __p1) __extension__ ({ \
27704 uint64x2x4_t __s1 = __p1; \
27705 uint64x2x4_t __rev1; \
27706 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
27707 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
27708 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
27709 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
27710 __builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 51); \
27711 })
27712 #endif
27713
27714 #ifdef __LITTLE_ENDIAN__
27715 #define vst1q_u16_x4(__p0, __p1) __extension__ ({ \
27716 uint16x8x4_t __s1 = __p1; \
27717 __builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 49); \
27718 })
27719 #else
27720 #define vst1q_u16_x4(__p0, __p1) __extension__ ({ \
27721 uint16x8x4_t __s1 = __p1; \
27722 uint16x8x4_t __rev1; \
27723 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
27724 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
27725 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
27726 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
27727 __builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 49); \
27728 })
27729 #endif
27730
27731 #ifdef __LITTLE_ENDIAN__
27732 #define vst1q_s8_x4(__p0, __p1) __extension__ ({ \
27733 int8x16x4_t __s1 = __p1; \
27734 __builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 32); \
27735 })
27736 #else
27737 #define vst1q_s8_x4(__p0, __p1) __extension__ ({ \
27738 int8x16x4_t __s1 = __p1; \
27739 int8x16x4_t __rev1; \
27740 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
27741 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
27742 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
27743 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
27744 __builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 32); \
27745 })
27746 #endif
27747
27748 #ifdef __LITTLE_ENDIAN__
27749 #define vst1q_f32_x4(__p0, __p1) __extension__ ({ \
27750 float32x4x4_t __s1 = __p1; \
27751 __builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 41); \
27752 })
27753 #else
27754 #define vst1q_f32_x4(__p0, __p1) __extension__ ({ \
27755 float32x4x4_t __s1 = __p1; \
27756 float32x4x4_t __rev1; \
27757 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
27758 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
27759 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
27760 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
27761 __builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 41); \
27762 })
27763 #endif
27764
27765 #ifdef __LITTLE_ENDIAN__
27766 #define vst1q_s32_x4(__p0, __p1) __extension__ ({ \
27767 int32x4x4_t __s1 = __p1; \
27768 __builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 34); \
27769 })
27770 #else
27771 #define vst1q_s32_x4(__p0, __p1) __extension__ ({ \
27772 int32x4x4_t __s1 = __p1; \
27773 int32x4x4_t __rev1; \
27774 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
27775 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
27776 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
27777 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
27778 __builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 34); \
27779 })
27780 #endif
27781
27782 #ifdef __LITTLE_ENDIAN__
27783 #define vst1q_s64_x4(__p0, __p1) __extension__ ({ \
27784 int64x2x4_t __s1 = __p1; \
27785 __builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 35); \
27786 })
27787 #else
27788 #define vst1q_s64_x4(__p0, __p1) __extension__ ({ \
27789 int64x2x4_t __s1 = __p1; \
27790 int64x2x4_t __rev1; \
27791 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
27792 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
27793 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
27794 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
27795 __builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 35); \
27796 })
27797 #endif
27798
27799 #ifdef __LITTLE_ENDIAN__
27800 #define vst1q_s16_x4(__p0, __p1) __extension__ ({ \
27801 int16x8x4_t __s1 = __p1; \
27802 __builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 33); \
27803 })
27804 #else
27805 #define vst1q_s16_x4(__p0, __p1) __extension__ ({ \
27806 int16x8x4_t __s1 = __p1; \
27807 int16x8x4_t __rev1; \
27808 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
27809 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
27810 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
27811 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
27812 __builtin_neon_vst1q_x4_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 33); \
27813 })
27814 #endif
27815
27816 #ifdef __LITTLE_ENDIAN__
27817 #define vst1_u8_x4(__p0, __p1) __extension__ ({ \
27818 uint8x8x4_t __s1 = __p1; \
27819 __builtin_neon_vst1_x4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 16); \
27820 })
27821 #else
27822 #define vst1_u8_x4(__p0, __p1) __extension__ ({ \
27823 uint8x8x4_t __s1 = __p1; \
27824 uint8x8x4_t __rev1; \
27825 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
27826 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
27827 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
27828 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
27829 __builtin_neon_vst1_x4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 16); \
27830 })
27831 #endif
27832
27833 #ifdef __LITTLE_ENDIAN__
27834 #define vst1_u32_x4(__p0, __p1) __extension__ ({ \
27835 uint32x2x4_t __s1 = __p1; \
27836 __builtin_neon_vst1_x4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 18); \
27837 })
27838 #else
27839 #define vst1_u32_x4(__p0, __p1) __extension__ ({ \
27840 uint32x2x4_t __s1 = __p1; \
27841 uint32x2x4_t __rev1; \
27842 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
27843 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
27844 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
27845 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
27846 __builtin_neon_vst1_x4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 18); \
27847 })
27848 #endif
27849
27850 #define vst1_u64_x4(__p0, __p1) __extension__ ({ \
27851 uint64x1x4_t __s1 = __p1; \
27852 __builtin_neon_vst1_x4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 19); \
27853 })
27854 #ifdef __LITTLE_ENDIAN__
27855 #define vst1_u16_x4(__p0, __p1) __extension__ ({ \
27856 uint16x4x4_t __s1 = __p1; \
27857 __builtin_neon_vst1_x4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 17); \
27858 })
27859 #else
27860 #define vst1_u16_x4(__p0, __p1) __extension__ ({ \
27861 uint16x4x4_t __s1 = __p1; \
27862 uint16x4x4_t __rev1; \
27863 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
27864 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
27865 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
27866 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
27867 __builtin_neon_vst1_x4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 17); \
27868 })
27869 #endif
27870
27871 #ifdef __LITTLE_ENDIAN__
27872 #define vst1_s8_x4(__p0, __p1) __extension__ ({ \
27873 int8x8x4_t __s1 = __p1; \
27874 __builtin_neon_vst1_x4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 0); \
27875 })
27876 #else
27877 #define vst1_s8_x4(__p0, __p1) __extension__ ({ \
27878 int8x8x4_t __s1 = __p1; \
27879 int8x8x4_t __rev1; \
27880 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
27881 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
27882 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
27883 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
27884 __builtin_neon_vst1_x4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 0); \
27885 })
27886 #endif
27887
27888 #ifdef __LITTLE_ENDIAN__
27889 #define vst1_f32_x4(__p0, __p1) __extension__ ({ \
27890 float32x2x4_t __s1 = __p1; \
27891 __builtin_neon_vst1_x4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 9); \
27892 })
27893 #else
27894 #define vst1_f32_x4(__p0, __p1) __extension__ ({ \
27895 float32x2x4_t __s1 = __p1; \
27896 float32x2x4_t __rev1; \
27897 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
27898 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
27899 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
27900 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
27901 __builtin_neon_vst1_x4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 9); \
27902 })
27903 #endif
27904
27905 #ifdef __LITTLE_ENDIAN__
27906 #define vst1_s32_x4(__p0, __p1) __extension__ ({ \
27907 int32x2x4_t __s1 = __p1; \
27908 __builtin_neon_vst1_x4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 2); \
27909 })
27910 #else
27911 #define vst1_s32_x4(__p0, __p1) __extension__ ({ \
27912 int32x2x4_t __s1 = __p1; \
27913 int32x2x4_t __rev1; \
27914 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
27915 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
27916 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
27917 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
27918 __builtin_neon_vst1_x4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 2); \
27919 })
27920 #endif
27921
27922 #define vst1_s64_x4(__p0, __p1) __extension__ ({ \
27923 int64x1x4_t __s1 = __p1; \
27924 __builtin_neon_vst1_x4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 3); \
27925 })
27926 #ifdef __LITTLE_ENDIAN__
27927 #define vst1_s16_x4(__p0, __p1) __extension__ ({ \
27928 int16x4x4_t __s1 = __p1; \
27929 __builtin_neon_vst1_x4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 1); \
27930 })
27931 #else
27932 #define vst1_s16_x4(__p0, __p1) __extension__ ({ \
27933 int16x4x4_t __s1 = __p1; \
27934 int16x4x4_t __rev1; \
27935 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
27936 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
27937 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
27938 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
27939 __builtin_neon_vst1_x4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 1); \
27940 })
27941 #endif
27942
27943 #ifdef __LITTLE_ENDIAN__
27944 #define vst2_p8(__p0, __p1) __extension__ ({ \
27945 poly8x8x2_t __s1 = __p1; \
27946 __builtin_neon_vst2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 4); \
27947 })
27948 #else
27949 #define vst2_p8(__p0, __p1) __extension__ ({ \
27950 poly8x8x2_t __s1 = __p1; \
27951 poly8x8x2_t __rev1; \
27952 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
27953 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
27954 __builtin_neon_vst2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 4); \
27955 })
27956 #endif
27957
27958 #ifdef __LITTLE_ENDIAN__
27959 #define vst2_p16(__p0, __p1) __extension__ ({ \
27960 poly16x4x2_t __s1 = __p1; \
27961 __builtin_neon_vst2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 5); \
27962 })
27963 #else
27964 #define vst2_p16(__p0, __p1) __extension__ ({ \
27965 poly16x4x2_t __s1 = __p1; \
27966 poly16x4x2_t __rev1; \
27967 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
27968 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
27969 __builtin_neon_vst2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 5); \
27970 })
27971 #endif
27972
27973 #ifdef __LITTLE_ENDIAN__
27974 #define vst2q_p8(__p0, __p1) __extension__ ({ \
27975 poly8x16x2_t __s1 = __p1; \
27976 __builtin_neon_vst2q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 36); \
27977 })
27978 #else
27979 #define vst2q_p8(__p0, __p1) __extension__ ({ \
27980 poly8x16x2_t __s1 = __p1; \
27981 poly8x16x2_t __rev1; \
27982 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
27983 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
27984 __builtin_neon_vst2q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 36); \
27985 })
27986 #endif
27987
27988 #ifdef __LITTLE_ENDIAN__
27989 #define vst2q_p16(__p0, __p1) __extension__ ({ \
27990 poly16x8x2_t __s1 = __p1; \
27991 __builtin_neon_vst2q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 37); \
27992 })
27993 #else
27994 #define vst2q_p16(__p0, __p1) __extension__ ({ \
27995 poly16x8x2_t __s1 = __p1; \
27996 poly16x8x2_t __rev1; \
27997 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
27998 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
27999 __builtin_neon_vst2q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 37); \
28000 })
28001 #endif
28002
28003 #ifdef __LITTLE_ENDIAN__
28004 #define vst2q_u8(__p0, __p1) __extension__ ({ \
28005 uint8x16x2_t __s1 = __p1; \
28006 __builtin_neon_vst2q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 48); \
28007 })
28008 #else
28009 #define vst2q_u8(__p0, __p1) __extension__ ({ \
28010 uint8x16x2_t __s1 = __p1; \
28011 uint8x16x2_t __rev1; \
28012 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
28013 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
28014 __builtin_neon_vst2q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 48); \
28015 })
28016 #endif
28017
28018 #ifdef __LITTLE_ENDIAN__
28019 #define vst2q_u32(__p0, __p1) __extension__ ({ \
28020 uint32x4x2_t __s1 = __p1; \
28021 __builtin_neon_vst2q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 50); \
28022 })
28023 #else
28024 #define vst2q_u32(__p0, __p1) __extension__ ({ \
28025 uint32x4x2_t __s1 = __p1; \
28026 uint32x4x2_t __rev1; \
28027 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
28028 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
28029 __builtin_neon_vst2q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 50); \
28030 })
28031 #endif
28032
28033 #ifdef __LITTLE_ENDIAN__
28034 #define vst2q_u16(__p0, __p1) __extension__ ({ \
28035 uint16x8x2_t __s1 = __p1; \
28036 __builtin_neon_vst2q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 49); \
28037 })
28038 #else
28039 #define vst2q_u16(__p0, __p1) __extension__ ({ \
28040 uint16x8x2_t __s1 = __p1; \
28041 uint16x8x2_t __rev1; \
28042 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
28043 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
28044 __builtin_neon_vst2q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 49); \
28045 })
28046 #endif
28047
28048 #ifdef __LITTLE_ENDIAN__
28049 #define vst2q_s8(__p0, __p1) __extension__ ({ \
28050 int8x16x2_t __s1 = __p1; \
28051 __builtin_neon_vst2q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 32); \
28052 })
28053 #else
28054 #define vst2q_s8(__p0, __p1) __extension__ ({ \
28055 int8x16x2_t __s1 = __p1; \
28056 int8x16x2_t __rev1; \
28057 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
28058 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
28059 __builtin_neon_vst2q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 32); \
28060 })
28061 #endif
28062
28063 #ifdef __LITTLE_ENDIAN__
28064 #define vst2q_f32(__p0, __p1) __extension__ ({ \
28065 float32x4x2_t __s1 = __p1; \
28066 __builtin_neon_vst2q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 41); \
28067 })
28068 #else
28069 #define vst2q_f32(__p0, __p1) __extension__ ({ \
28070 float32x4x2_t __s1 = __p1; \
28071 float32x4x2_t __rev1; \
28072 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
28073 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
28074 __builtin_neon_vst2q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 41); \
28075 })
28076 #endif
28077
28078 #ifdef __LITTLE_ENDIAN__
28079 #define vst2q_s32(__p0, __p1) __extension__ ({ \
28080 int32x4x2_t __s1 = __p1; \
28081 __builtin_neon_vst2q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 34); \
28082 })
28083 #else
28084 #define vst2q_s32(__p0, __p1) __extension__ ({ \
28085 int32x4x2_t __s1 = __p1; \
28086 int32x4x2_t __rev1; \
28087 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
28088 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
28089 __builtin_neon_vst2q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 34); \
28090 })
28091 #endif
28092
28093 #ifdef __LITTLE_ENDIAN__
28094 #define vst2q_s16(__p0, __p1) __extension__ ({ \
28095 int16x8x2_t __s1 = __p1; \
28096 __builtin_neon_vst2q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 33); \
28097 })
28098 #else
28099 #define vst2q_s16(__p0, __p1) __extension__ ({ \
28100 int16x8x2_t __s1 = __p1; \
28101 int16x8x2_t __rev1; \
28102 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
28103 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
28104 __builtin_neon_vst2q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 33); \
28105 })
28106 #endif
28107
28108 #ifdef __LITTLE_ENDIAN__
28109 #define vst2_u8(__p0, __p1) __extension__ ({ \
28110 uint8x8x2_t __s1 = __p1; \
28111 __builtin_neon_vst2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 16); \
28112 })
28113 #else
28114 #define vst2_u8(__p0, __p1) __extension__ ({ \
28115 uint8x8x2_t __s1 = __p1; \
28116 uint8x8x2_t __rev1; \
28117 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
28118 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
28119 __builtin_neon_vst2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 16); \
28120 })
28121 #endif
28122
28123 #ifdef __LITTLE_ENDIAN__
28124 #define vst2_u32(__p0, __p1) __extension__ ({ \
28125 uint32x2x2_t __s1 = __p1; \
28126 __builtin_neon_vst2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 18); \
28127 })
28128 #else
28129 #define vst2_u32(__p0, __p1) __extension__ ({ \
28130 uint32x2x2_t __s1 = __p1; \
28131 uint32x2x2_t __rev1; \
28132 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
28133 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
28134 __builtin_neon_vst2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 18); \
28135 })
28136 #endif
28137
28138 #define vst2_u64(__p0, __p1) __extension__ ({ \
28139 uint64x1x2_t __s1 = __p1; \
28140 __builtin_neon_vst2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 19); \
28141 })
28142 #ifdef __LITTLE_ENDIAN__
28143 #define vst2_u16(__p0, __p1) __extension__ ({ \
28144 uint16x4x2_t __s1 = __p1; \
28145 __builtin_neon_vst2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 17); \
28146 })
28147 #else
28148 #define vst2_u16(__p0, __p1) __extension__ ({ \
28149 uint16x4x2_t __s1 = __p1; \
28150 uint16x4x2_t __rev1; \
28151 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
28152 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
28153 __builtin_neon_vst2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 17); \
28154 })
28155 #endif
28156
28157 #ifdef __LITTLE_ENDIAN__
28158 #define vst2_s8(__p0, __p1) __extension__ ({ \
28159 int8x8x2_t __s1 = __p1; \
28160 __builtin_neon_vst2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 0); \
28161 })
28162 #else
28163 #define vst2_s8(__p0, __p1) __extension__ ({ \
28164 int8x8x2_t __s1 = __p1; \
28165 int8x8x2_t __rev1; \
28166 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
28167 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
28168 __builtin_neon_vst2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 0); \
28169 })
28170 #endif
28171
28172 #ifdef __LITTLE_ENDIAN__
28173 #define vst2_f32(__p0, __p1) __extension__ ({ \
28174 float32x2x2_t __s1 = __p1; \
28175 __builtin_neon_vst2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 9); \
28176 })
28177 #else
28178 #define vst2_f32(__p0, __p1) __extension__ ({ \
28179 float32x2x2_t __s1 = __p1; \
28180 float32x2x2_t __rev1; \
28181 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
28182 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
28183 __builtin_neon_vst2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 9); \
28184 })
28185 #endif
28186
28187 #ifdef __LITTLE_ENDIAN__
28188 #define vst2_s32(__p0, __p1) __extension__ ({ \
28189 int32x2x2_t __s1 = __p1; \
28190 __builtin_neon_vst2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 2); \
28191 })
28192 #else
28193 #define vst2_s32(__p0, __p1) __extension__ ({ \
28194 int32x2x2_t __s1 = __p1; \
28195 int32x2x2_t __rev1; \
28196 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
28197 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
28198 __builtin_neon_vst2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 2); \
28199 })
28200 #endif
28201
28202 #define vst2_s64(__p0, __p1) __extension__ ({ \
28203 int64x1x2_t __s1 = __p1; \
28204 __builtin_neon_vst2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 3); \
28205 })
28206 #ifdef __LITTLE_ENDIAN__
28207 #define vst2_s16(__p0, __p1) __extension__ ({ \
28208 int16x4x2_t __s1 = __p1; \
28209 __builtin_neon_vst2_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 1); \
28210 })
28211 #else
28212 #define vst2_s16(__p0, __p1) __extension__ ({ \
28213 int16x4x2_t __s1 = __p1; \
28214 int16x4x2_t __rev1; \
28215 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
28216 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
28217 __builtin_neon_vst2_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 1); \
28218 })
28219 #endif
28220
28221 #ifdef __LITTLE_ENDIAN__
28222 #define vst2_lane_p8(__p0, __p1, __p2) __extension__ ({ \
28223 poly8x8x2_t __s1 = __p1; \
28224 __builtin_neon_vst2_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 4); \
28225 })
28226 #else
28227 #define vst2_lane_p8(__p0, __p1, __p2) __extension__ ({ \
28228 poly8x8x2_t __s1 = __p1; \
28229 poly8x8x2_t __rev1; \
28230 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
28231 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
28232 __builtin_neon_vst2_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 4); \
28233 })
28234 #endif
28235
28236 #ifdef __LITTLE_ENDIAN__
28237 #define vst2_lane_p16(__p0, __p1, __p2) __extension__ ({ \
28238 poly16x4x2_t __s1 = __p1; \
28239 __builtin_neon_vst2_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 5); \
28240 })
28241 #else
28242 #define vst2_lane_p16(__p0, __p1, __p2) __extension__ ({ \
28243 poly16x4x2_t __s1 = __p1; \
28244 poly16x4x2_t __rev1; \
28245 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
28246 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
28247 __builtin_neon_vst2_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 5); \
28248 })
28249 #endif
28250
28251 #ifdef __LITTLE_ENDIAN__
28252 #define vst2q_lane_p16(__p0, __p1, __p2) __extension__ ({ \
28253 poly16x8x2_t __s1 = __p1; \
28254 __builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 37); \
28255 })
28256 #else
28257 #define vst2q_lane_p16(__p0, __p1, __p2) __extension__ ({ \
28258 poly16x8x2_t __s1 = __p1; \
28259 poly16x8x2_t __rev1; \
28260 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
28261 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
28262 __builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 37); \
28263 })
28264 #endif
28265
28266 #ifdef __LITTLE_ENDIAN__
28267 #define vst2q_lane_u32(__p0, __p1, __p2) __extension__ ({ \
28268 uint32x4x2_t __s1 = __p1; \
28269 __builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 50); \
28270 })
28271 #else
28272 #define vst2q_lane_u32(__p0, __p1, __p2) __extension__ ({ \
28273 uint32x4x2_t __s1 = __p1; \
28274 uint32x4x2_t __rev1; \
28275 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
28276 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
28277 __builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 50); \
28278 })
28279 #endif
28280
28281 #ifdef __LITTLE_ENDIAN__
28282 #define vst2q_lane_u16(__p0, __p1, __p2) __extension__ ({ \
28283 uint16x8x2_t __s1 = __p1; \
28284 __builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 49); \
28285 })
28286 #else
28287 #define vst2q_lane_u16(__p0, __p1, __p2) __extension__ ({ \
28288 uint16x8x2_t __s1 = __p1; \
28289 uint16x8x2_t __rev1; \
28290 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
28291 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
28292 __builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 49); \
28293 })
28294 #endif
28295
28296 #ifdef __LITTLE_ENDIAN__
28297 #define vst2q_lane_f32(__p0, __p1, __p2) __extension__ ({ \
28298 float32x4x2_t __s1 = __p1; \
28299 __builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 41); \
28300 })
28301 #else
28302 #define vst2q_lane_f32(__p0, __p1, __p2) __extension__ ({ \
28303 float32x4x2_t __s1 = __p1; \
28304 float32x4x2_t __rev1; \
28305 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
28306 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
28307 __builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 41); \
28308 })
28309 #endif
28310
28311 #ifdef __LITTLE_ENDIAN__
28312 #define vst2q_lane_s32(__p0, __p1, __p2) __extension__ ({ \
28313 int32x4x2_t __s1 = __p1; \
28314 __builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 34); \
28315 })
28316 #else
28317 #define vst2q_lane_s32(__p0, __p1, __p2) __extension__ ({ \
28318 int32x4x2_t __s1 = __p1; \
28319 int32x4x2_t __rev1; \
28320 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
28321 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
28322 __builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 34); \
28323 })
28324 #endif
28325
28326 #ifdef __LITTLE_ENDIAN__
28327 #define vst2q_lane_s16(__p0, __p1, __p2) __extension__ ({ \
28328 int16x8x2_t __s1 = __p1; \
28329 __builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 33); \
28330 })
28331 #else
28332 #define vst2q_lane_s16(__p0, __p1, __p2) __extension__ ({ \
28333 int16x8x2_t __s1 = __p1; \
28334 int16x8x2_t __rev1; \
28335 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
28336 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
28337 __builtin_neon_vst2q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 33); \
28338 })
28339 #endif
28340
28341 #ifdef __LITTLE_ENDIAN__
28342 #define vst2_lane_u8(__p0, __p1, __p2) __extension__ ({ \
28343 uint8x8x2_t __s1 = __p1; \
28344 __builtin_neon_vst2_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 16); \
28345 })
28346 #else
28347 #define vst2_lane_u8(__p0, __p1, __p2) __extension__ ({ \
28348 uint8x8x2_t __s1 = __p1; \
28349 uint8x8x2_t __rev1; \
28350 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
28351 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
28352 __builtin_neon_vst2_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 16); \
28353 })
28354 #endif
28355
28356 #ifdef __LITTLE_ENDIAN__
28357 #define vst2_lane_u32(__p0, __p1, __p2) __extension__ ({ \
28358 uint32x2x2_t __s1 = __p1; \
28359 __builtin_neon_vst2_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 18); \
28360 })
28361 #else
28362 #define vst2_lane_u32(__p0, __p1, __p2) __extension__ ({ \
28363 uint32x2x2_t __s1 = __p1; \
28364 uint32x2x2_t __rev1; \
28365 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
28366 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
28367 __builtin_neon_vst2_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 18); \
28368 })
28369 #endif
28370
28371 #ifdef __LITTLE_ENDIAN__
28372 #define vst2_lane_u16(__p0, __p1, __p2) __extension__ ({ \
28373 uint16x4x2_t __s1 = __p1; \
28374 __builtin_neon_vst2_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 17); \
28375 })
28376 #else
28377 #define vst2_lane_u16(__p0, __p1, __p2) __extension__ ({ \
28378 uint16x4x2_t __s1 = __p1; \
28379 uint16x4x2_t __rev1; \
28380 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
28381 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
28382 __builtin_neon_vst2_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 17); \
28383 })
28384 #endif
28385
28386 #ifdef __LITTLE_ENDIAN__
28387 #define vst2_lane_s8(__p0, __p1, __p2) __extension__ ({ \
28388 int8x8x2_t __s1 = __p1; \
28389 __builtin_neon_vst2_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 0); \
28390 })
28391 #else
28392 #define vst2_lane_s8(__p0, __p1, __p2) __extension__ ({ \
28393 int8x8x2_t __s1 = __p1; \
28394 int8x8x2_t __rev1; \
28395 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
28396 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
28397 __builtin_neon_vst2_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 0); \
28398 })
28399 #endif
28400
28401 #ifdef __LITTLE_ENDIAN__
28402 #define vst2_lane_f32(__p0, __p1, __p2) __extension__ ({ \
28403 float32x2x2_t __s1 = __p1; \
28404 __builtin_neon_vst2_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 9); \
28405 })
28406 #else
28407 #define vst2_lane_f32(__p0, __p1, __p2) __extension__ ({ \
28408 float32x2x2_t __s1 = __p1; \
28409 float32x2x2_t __rev1; \
28410 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
28411 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
28412 __builtin_neon_vst2_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 9); \
28413 })
28414 #endif
28415
28416 #ifdef __LITTLE_ENDIAN__
28417 #define vst2_lane_s32(__p0, __p1, __p2) __extension__ ({ \
28418 int32x2x2_t __s1 = __p1; \
28419 __builtin_neon_vst2_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 2); \
28420 })
28421 #else
28422 #define vst2_lane_s32(__p0, __p1, __p2) __extension__ ({ \
28423 int32x2x2_t __s1 = __p1; \
28424 int32x2x2_t __rev1; \
28425 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
28426 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
28427 __builtin_neon_vst2_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 2); \
28428 })
28429 #endif
28430
28431 #ifdef __LITTLE_ENDIAN__
28432 #define vst2_lane_s16(__p0, __p1, __p2) __extension__ ({ \
28433 int16x4x2_t __s1 = __p1; \
28434 __builtin_neon_vst2_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 1); \
28435 })
28436 #else
28437 #define vst2_lane_s16(__p0, __p1, __p2) __extension__ ({ \
28438 int16x4x2_t __s1 = __p1; \
28439 int16x4x2_t __rev1; \
28440 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
28441 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
28442 __builtin_neon_vst2_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 1); \
28443 })
28444 #endif
28445
28446 #ifdef __LITTLE_ENDIAN__
28447 #define vst3_p8(__p0, __p1) __extension__ ({ \
28448 poly8x8x3_t __s1 = __p1; \
28449 __builtin_neon_vst3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 4); \
28450 })
28451 #else
28452 #define vst3_p8(__p0, __p1) __extension__ ({ \
28453 poly8x8x3_t __s1 = __p1; \
28454 poly8x8x3_t __rev1; \
28455 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
28456 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
28457 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
28458 __builtin_neon_vst3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 4); \
28459 })
28460 #endif
28461
28462 #ifdef __LITTLE_ENDIAN__
28463 #define vst3_p16(__p0, __p1) __extension__ ({ \
28464 poly16x4x3_t __s1 = __p1; \
28465 __builtin_neon_vst3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 5); \
28466 })
28467 #else
28468 #define vst3_p16(__p0, __p1) __extension__ ({ \
28469 poly16x4x3_t __s1 = __p1; \
28470 poly16x4x3_t __rev1; \
28471 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
28472 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
28473 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
28474 __builtin_neon_vst3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 5); \
28475 })
28476 #endif
28477
28478 #ifdef __LITTLE_ENDIAN__
28479 #define vst3q_p8(__p0, __p1) __extension__ ({ \
28480 poly8x16x3_t __s1 = __p1; \
28481 __builtin_neon_vst3q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 36); \
28482 })
28483 #else
28484 #define vst3q_p8(__p0, __p1) __extension__ ({ \
28485 poly8x16x3_t __s1 = __p1; \
28486 poly8x16x3_t __rev1; \
28487 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
28488 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
28489 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
28490 __builtin_neon_vst3q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 36); \
28491 })
28492 #endif
28493
28494 #ifdef __LITTLE_ENDIAN__
28495 #define vst3q_p16(__p0, __p1) __extension__ ({ \
28496 poly16x8x3_t __s1 = __p1; \
28497 __builtin_neon_vst3q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 37); \
28498 })
28499 #else
28500 #define vst3q_p16(__p0, __p1) __extension__ ({ \
28501 poly16x8x3_t __s1 = __p1; \
28502 poly16x8x3_t __rev1; \
28503 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
28504 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
28505 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
28506 __builtin_neon_vst3q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 37); \
28507 })
28508 #endif
28509
28510 #ifdef __LITTLE_ENDIAN__
28511 #define vst3q_u8(__p0, __p1) __extension__ ({ \
28512 uint8x16x3_t __s1 = __p1; \
28513 __builtin_neon_vst3q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 48); \
28514 })
28515 #else
28516 #define vst3q_u8(__p0, __p1) __extension__ ({ \
28517 uint8x16x3_t __s1 = __p1; \
28518 uint8x16x3_t __rev1; \
28519 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
28520 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
28521 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
28522 __builtin_neon_vst3q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 48); \
28523 })
28524 #endif
28525
28526 #ifdef __LITTLE_ENDIAN__
28527 #define vst3q_u32(__p0, __p1) __extension__ ({ \
28528 uint32x4x3_t __s1 = __p1; \
28529 __builtin_neon_vst3q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 50); \
28530 })
28531 #else
28532 #define vst3q_u32(__p0, __p1) __extension__ ({ \
28533 uint32x4x3_t __s1 = __p1; \
28534 uint32x4x3_t __rev1; \
28535 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
28536 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
28537 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
28538 __builtin_neon_vst3q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 50); \
28539 })
28540 #endif
28541
28542 #ifdef __LITTLE_ENDIAN__
28543 #define vst3q_u16(__p0, __p1) __extension__ ({ \
28544 uint16x8x3_t __s1 = __p1; \
28545 __builtin_neon_vst3q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 49); \
28546 })
28547 #else
28548 #define vst3q_u16(__p0, __p1) __extension__ ({ \
28549 uint16x8x3_t __s1 = __p1; \
28550 uint16x8x3_t __rev1; \
28551 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
28552 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
28553 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
28554 __builtin_neon_vst3q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 49); \
28555 })
28556 #endif
28557
28558 #ifdef __LITTLE_ENDIAN__
28559 #define vst3q_s8(__p0, __p1) __extension__ ({ \
28560 int8x16x3_t __s1 = __p1; \
28561 __builtin_neon_vst3q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 32); \
28562 })
28563 #else
28564 #define vst3q_s8(__p0, __p1) __extension__ ({ \
28565 int8x16x3_t __s1 = __p1; \
28566 int8x16x3_t __rev1; \
28567 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
28568 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
28569 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
28570 __builtin_neon_vst3q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 32); \
28571 })
28572 #endif
28573
28574 #ifdef __LITTLE_ENDIAN__
28575 #define vst3q_f32(__p0, __p1) __extension__ ({ \
28576 float32x4x3_t __s1 = __p1; \
28577 __builtin_neon_vst3q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 41); \
28578 })
28579 #else
28580 #define vst3q_f32(__p0, __p1) __extension__ ({ \
28581 float32x4x3_t __s1 = __p1; \
28582 float32x4x3_t __rev1; \
28583 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
28584 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
28585 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
28586 __builtin_neon_vst3q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 41); \
28587 })
28588 #endif
28589
28590 #ifdef __LITTLE_ENDIAN__
28591 #define vst3q_s32(__p0, __p1) __extension__ ({ \
28592 int32x4x3_t __s1 = __p1; \
28593 __builtin_neon_vst3q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 34); \
28594 })
28595 #else
28596 #define vst3q_s32(__p0, __p1) __extension__ ({ \
28597 int32x4x3_t __s1 = __p1; \
28598 int32x4x3_t __rev1; \
28599 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
28600 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
28601 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
28602 __builtin_neon_vst3q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 34); \
28603 })
28604 #endif
28605
28606 #ifdef __LITTLE_ENDIAN__
28607 #define vst3q_s16(__p0, __p1) __extension__ ({ \
28608 int16x8x3_t __s1 = __p1; \
28609 __builtin_neon_vst3q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 33); \
28610 })
28611 #else
28612 #define vst3q_s16(__p0, __p1) __extension__ ({ \
28613 int16x8x3_t __s1 = __p1; \
28614 int16x8x3_t __rev1; \
28615 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
28616 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
28617 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
28618 __builtin_neon_vst3q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 33); \
28619 })
28620 #endif
28621
28622 #ifdef __LITTLE_ENDIAN__
28623 #define vst3_u8(__p0, __p1) __extension__ ({ \
28624 uint8x8x3_t __s1 = __p1; \
28625 __builtin_neon_vst3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 16); \
28626 })
28627 #else
28628 #define vst3_u8(__p0, __p1) __extension__ ({ \
28629 uint8x8x3_t __s1 = __p1; \
28630 uint8x8x3_t __rev1; \
28631 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
28632 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
28633 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
28634 __builtin_neon_vst3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 16); \
28635 })
28636 #endif
28637
28638 #ifdef __LITTLE_ENDIAN__
28639 #define vst3_u32(__p0, __p1) __extension__ ({ \
28640 uint32x2x3_t __s1 = __p1; \
28641 __builtin_neon_vst3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 18); \
28642 })
28643 #else
28644 #define vst3_u32(__p0, __p1) __extension__ ({ \
28645 uint32x2x3_t __s1 = __p1; \
28646 uint32x2x3_t __rev1; \
28647 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
28648 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
28649 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
28650 __builtin_neon_vst3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 18); \
28651 })
28652 #endif
28653
28654 #define vst3_u64(__p0, __p1) __extension__ ({ \
28655 uint64x1x3_t __s1 = __p1; \
28656 __builtin_neon_vst3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 19); \
28657 })
28658 #ifdef __LITTLE_ENDIAN__
28659 #define vst3_u16(__p0, __p1) __extension__ ({ \
28660 uint16x4x3_t __s1 = __p1; \
28661 __builtin_neon_vst3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 17); \
28662 })
28663 #else
28664 #define vst3_u16(__p0, __p1) __extension__ ({ \
28665 uint16x4x3_t __s1 = __p1; \
28666 uint16x4x3_t __rev1; \
28667 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
28668 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
28669 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
28670 __builtin_neon_vst3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 17); \
28671 })
28672 #endif
28673
28674 #ifdef __LITTLE_ENDIAN__
28675 #define vst3_s8(__p0, __p1) __extension__ ({ \
28676 int8x8x3_t __s1 = __p1; \
28677 __builtin_neon_vst3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 0); \
28678 })
28679 #else
28680 #define vst3_s8(__p0, __p1) __extension__ ({ \
28681 int8x8x3_t __s1 = __p1; \
28682 int8x8x3_t __rev1; \
28683 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
28684 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
28685 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
28686 __builtin_neon_vst3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 0); \
28687 })
28688 #endif
28689
28690 #ifdef __LITTLE_ENDIAN__
28691 #define vst3_f32(__p0, __p1) __extension__ ({ \
28692 float32x2x3_t __s1 = __p1; \
28693 __builtin_neon_vst3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 9); \
28694 })
28695 #else
28696 #define vst3_f32(__p0, __p1) __extension__ ({ \
28697 float32x2x3_t __s1 = __p1; \
28698 float32x2x3_t __rev1; \
28699 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
28700 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
28701 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
28702 __builtin_neon_vst3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 9); \
28703 })
28704 #endif
28705
28706 #ifdef __LITTLE_ENDIAN__
28707 #define vst3_s32(__p0, __p1) __extension__ ({ \
28708 int32x2x3_t __s1 = __p1; \
28709 __builtin_neon_vst3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 2); \
28710 })
28711 #else
28712 #define vst3_s32(__p0, __p1) __extension__ ({ \
28713 int32x2x3_t __s1 = __p1; \
28714 int32x2x3_t __rev1; \
28715 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
28716 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
28717 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
28718 __builtin_neon_vst3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 2); \
28719 })
28720 #endif
28721
28722 #define vst3_s64(__p0, __p1) __extension__ ({ \
28723 int64x1x3_t __s1 = __p1; \
28724 __builtin_neon_vst3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 3); \
28725 })
28726 #ifdef __LITTLE_ENDIAN__
28727 #define vst3_s16(__p0, __p1) __extension__ ({ \
28728 int16x4x3_t __s1 = __p1; \
28729 __builtin_neon_vst3_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 1); \
28730 })
28731 #else
28732 #define vst3_s16(__p0, __p1) __extension__ ({ \
28733 int16x4x3_t __s1 = __p1; \
28734 int16x4x3_t __rev1; \
28735 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
28736 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
28737 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
28738 __builtin_neon_vst3_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 1); \
28739 })
28740 #endif
28741
28742 #ifdef __LITTLE_ENDIAN__
28743 #define vst3_lane_p8(__p0, __p1, __p2) __extension__ ({ \
28744 poly8x8x3_t __s1 = __p1; \
28745 __builtin_neon_vst3_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 4); \
28746 })
28747 #else
28748 #define vst3_lane_p8(__p0, __p1, __p2) __extension__ ({ \
28749 poly8x8x3_t __s1 = __p1; \
28750 poly8x8x3_t __rev1; \
28751 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
28752 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
28753 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
28754 __builtin_neon_vst3_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 4); \
28755 })
28756 #endif
28757
28758 #ifdef __LITTLE_ENDIAN__
28759 #define vst3_lane_p16(__p0, __p1, __p2) __extension__ ({ \
28760 poly16x4x3_t __s1 = __p1; \
28761 __builtin_neon_vst3_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 5); \
28762 })
28763 #else
28764 #define vst3_lane_p16(__p0, __p1, __p2) __extension__ ({ \
28765 poly16x4x3_t __s1 = __p1; \
28766 poly16x4x3_t __rev1; \
28767 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
28768 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
28769 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
28770 __builtin_neon_vst3_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 5); \
28771 })
28772 #endif
28773
28774 #ifdef __LITTLE_ENDIAN__
28775 #define vst3q_lane_p16(__p0, __p1, __p2) __extension__ ({ \
28776 poly16x8x3_t __s1 = __p1; \
28777 __builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 37); \
28778 })
28779 #else
28780 #define vst3q_lane_p16(__p0, __p1, __p2) __extension__ ({ \
28781 poly16x8x3_t __s1 = __p1; \
28782 poly16x8x3_t __rev1; \
28783 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
28784 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
28785 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
28786 __builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 37); \
28787 })
28788 #endif
28789
28790 #ifdef __LITTLE_ENDIAN__
28791 #define vst3q_lane_u32(__p0, __p1, __p2) __extension__ ({ \
28792 uint32x4x3_t __s1 = __p1; \
28793 __builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 50); \
28794 })
28795 #else
28796 #define vst3q_lane_u32(__p0, __p1, __p2) __extension__ ({ \
28797 uint32x4x3_t __s1 = __p1; \
28798 uint32x4x3_t __rev1; \
28799 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
28800 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
28801 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
28802 __builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 50); \
28803 })
28804 #endif
28805
28806 #ifdef __LITTLE_ENDIAN__
28807 #define vst3q_lane_u16(__p0, __p1, __p2) __extension__ ({ \
28808 uint16x8x3_t __s1 = __p1; \
28809 __builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 49); \
28810 })
28811 #else
28812 #define vst3q_lane_u16(__p0, __p1, __p2) __extension__ ({ \
28813 uint16x8x3_t __s1 = __p1; \
28814 uint16x8x3_t __rev1; \
28815 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
28816 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
28817 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
28818 __builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 49); \
28819 })
28820 #endif
28821
28822 #ifdef __LITTLE_ENDIAN__
28823 #define vst3q_lane_f32(__p0, __p1, __p2) __extension__ ({ \
28824 float32x4x3_t __s1 = __p1; \
28825 __builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 41); \
28826 })
28827 #else
28828 #define vst3q_lane_f32(__p0, __p1, __p2) __extension__ ({ \
28829 float32x4x3_t __s1 = __p1; \
28830 float32x4x3_t __rev1; \
28831 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
28832 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
28833 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
28834 __builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 41); \
28835 })
28836 #endif
28837
28838 #ifdef __LITTLE_ENDIAN__
28839 #define vst3q_lane_s32(__p0, __p1, __p2) __extension__ ({ \
28840 int32x4x3_t __s1 = __p1; \
28841 __builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 34); \
28842 })
28843 #else
28844 #define vst3q_lane_s32(__p0, __p1, __p2) __extension__ ({ \
28845 int32x4x3_t __s1 = __p1; \
28846 int32x4x3_t __rev1; \
28847 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
28848 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
28849 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
28850 __builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 34); \
28851 })
28852 #endif
28853
28854 #ifdef __LITTLE_ENDIAN__
28855 #define vst3q_lane_s16(__p0, __p1, __p2) __extension__ ({ \
28856 int16x8x3_t __s1 = __p1; \
28857 __builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 33); \
28858 })
28859 #else
28860 #define vst3q_lane_s16(__p0, __p1, __p2) __extension__ ({ \
28861 int16x8x3_t __s1 = __p1; \
28862 int16x8x3_t __rev1; \
28863 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
28864 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
28865 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
28866 __builtin_neon_vst3q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 33); \
28867 })
28868 #endif
28869
28870 #ifdef __LITTLE_ENDIAN__
28871 #define vst3_lane_u8(__p0, __p1, __p2) __extension__ ({ \
28872 uint8x8x3_t __s1 = __p1; \
28873 __builtin_neon_vst3_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 16); \
28874 })
28875 #else
28876 #define vst3_lane_u8(__p0, __p1, __p2) __extension__ ({ \
28877 uint8x8x3_t __s1 = __p1; \
28878 uint8x8x3_t __rev1; \
28879 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
28880 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
28881 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
28882 __builtin_neon_vst3_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 16); \
28883 })
28884 #endif
28885
28886 #ifdef __LITTLE_ENDIAN__
28887 #define vst3_lane_u32(__p0, __p1, __p2) __extension__ ({ \
28888 uint32x2x3_t __s1 = __p1; \
28889 __builtin_neon_vst3_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 18); \
28890 })
28891 #else
28892 #define vst3_lane_u32(__p0, __p1, __p2) __extension__ ({ \
28893 uint32x2x3_t __s1 = __p1; \
28894 uint32x2x3_t __rev1; \
28895 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
28896 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
28897 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
28898 __builtin_neon_vst3_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 18); \
28899 })
28900 #endif
28901
28902 #ifdef __LITTLE_ENDIAN__
28903 #define vst3_lane_u16(__p0, __p1, __p2) __extension__ ({ \
28904 uint16x4x3_t __s1 = __p1; \
28905 __builtin_neon_vst3_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 17); \
28906 })
28907 #else
28908 #define vst3_lane_u16(__p0, __p1, __p2) __extension__ ({ \
28909 uint16x4x3_t __s1 = __p1; \
28910 uint16x4x3_t __rev1; \
28911 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
28912 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
28913 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
28914 __builtin_neon_vst3_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 17); \
28915 })
28916 #endif
28917
28918 #ifdef __LITTLE_ENDIAN__
28919 #define vst3_lane_s8(__p0, __p1, __p2) __extension__ ({ \
28920 int8x8x3_t __s1 = __p1; \
28921 __builtin_neon_vst3_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 0); \
28922 })
28923 #else
28924 #define vst3_lane_s8(__p0, __p1, __p2) __extension__ ({ \
28925 int8x8x3_t __s1 = __p1; \
28926 int8x8x3_t __rev1; \
28927 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
28928 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
28929 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
28930 __builtin_neon_vst3_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 0); \
28931 })
28932 #endif
28933
28934 #ifdef __LITTLE_ENDIAN__
28935 #define vst3_lane_f32(__p0, __p1, __p2) __extension__ ({ \
28936 float32x2x3_t __s1 = __p1; \
28937 __builtin_neon_vst3_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 9); \
28938 })
28939 #else
28940 #define vst3_lane_f32(__p0, __p1, __p2) __extension__ ({ \
28941 float32x2x3_t __s1 = __p1; \
28942 float32x2x3_t __rev1; \
28943 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
28944 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
28945 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
28946 __builtin_neon_vst3_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 9); \
28947 })
28948 #endif
28949
28950 #ifdef __LITTLE_ENDIAN__
28951 #define vst3_lane_s32(__p0, __p1, __p2) __extension__ ({ \
28952 int32x2x3_t __s1 = __p1; \
28953 __builtin_neon_vst3_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 2); \
28954 })
28955 #else
28956 #define vst3_lane_s32(__p0, __p1, __p2) __extension__ ({ \
28957 int32x2x3_t __s1 = __p1; \
28958 int32x2x3_t __rev1; \
28959 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
28960 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
28961 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
28962 __builtin_neon_vst3_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 2); \
28963 })
28964 #endif
28965
28966 #ifdef __LITTLE_ENDIAN__
28967 #define vst3_lane_s16(__p0, __p1, __p2) __extension__ ({ \
28968 int16x4x3_t __s1 = __p1; \
28969 __builtin_neon_vst3_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 1); \
28970 })
28971 #else
28972 #define vst3_lane_s16(__p0, __p1, __p2) __extension__ ({ \
28973 int16x4x3_t __s1 = __p1; \
28974 int16x4x3_t __rev1; \
28975 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
28976 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
28977 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
28978 __builtin_neon_vst3_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 1); \
28979 })
28980 #endif
28981
28982 #ifdef __LITTLE_ENDIAN__
28983 #define vst4_p8(__p0, __p1) __extension__ ({ \
28984 poly8x8x4_t __s1 = __p1; \
28985 __builtin_neon_vst4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 4); \
28986 })
28987 #else
28988 #define vst4_p8(__p0, __p1) __extension__ ({ \
28989 poly8x8x4_t __s1 = __p1; \
28990 poly8x8x4_t __rev1; \
28991 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
28992 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
28993 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
28994 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
28995 __builtin_neon_vst4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 4); \
28996 })
28997 #endif
28998
28999 #ifdef __LITTLE_ENDIAN__
29000 #define vst4_p16(__p0, __p1) __extension__ ({ \
29001 poly16x4x4_t __s1 = __p1; \
29002 __builtin_neon_vst4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 5); \
29003 })
29004 #else
29005 #define vst4_p16(__p0, __p1) __extension__ ({ \
29006 poly16x4x4_t __s1 = __p1; \
29007 poly16x4x4_t __rev1; \
29008 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
29009 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
29010 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
29011 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
29012 __builtin_neon_vst4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 5); \
29013 })
29014 #endif
29015
29016 #ifdef __LITTLE_ENDIAN__
29017 #define vst4q_p8(__p0, __p1) __extension__ ({ \
29018 poly8x16x4_t __s1 = __p1; \
29019 __builtin_neon_vst4q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 36); \
29020 })
29021 #else
29022 #define vst4q_p8(__p0, __p1) __extension__ ({ \
29023 poly8x16x4_t __s1 = __p1; \
29024 poly8x16x4_t __rev1; \
29025 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
29026 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
29027 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
29028 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
29029 __builtin_neon_vst4q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 36); \
29030 })
29031 #endif
29032
29033 #ifdef __LITTLE_ENDIAN__
29034 #define vst4q_p16(__p0, __p1) __extension__ ({ \
29035 poly16x8x4_t __s1 = __p1; \
29036 __builtin_neon_vst4q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 37); \
29037 })
29038 #else
29039 #define vst4q_p16(__p0, __p1) __extension__ ({ \
29040 poly16x8x4_t __s1 = __p1; \
29041 poly16x8x4_t __rev1; \
29042 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
29043 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
29044 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
29045 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
29046 __builtin_neon_vst4q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 37); \
29047 })
29048 #endif
29049
29050 #ifdef __LITTLE_ENDIAN__
29051 #define vst4q_u8(__p0, __p1) __extension__ ({ \
29052 uint8x16x4_t __s1 = __p1; \
29053 __builtin_neon_vst4q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 48); \
29054 })
29055 #else
29056 #define vst4q_u8(__p0, __p1) __extension__ ({ \
29057 uint8x16x4_t __s1 = __p1; \
29058 uint8x16x4_t __rev1; \
29059 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
29060 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
29061 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
29062 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
29063 __builtin_neon_vst4q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 48); \
29064 })
29065 #endif
29066
29067 #ifdef __LITTLE_ENDIAN__
29068 #define vst4q_u32(__p0, __p1) __extension__ ({ \
29069 uint32x4x4_t __s1 = __p1; \
29070 __builtin_neon_vst4q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 50); \
29071 })
29072 #else
29073 #define vst4q_u32(__p0, __p1) __extension__ ({ \
29074 uint32x4x4_t __s1 = __p1; \
29075 uint32x4x4_t __rev1; \
29076 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
29077 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
29078 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
29079 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
29080 __builtin_neon_vst4q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 50); \
29081 })
29082 #endif
29083
29084 #ifdef __LITTLE_ENDIAN__
29085 #define vst4q_u16(__p0, __p1) __extension__ ({ \
29086 uint16x8x4_t __s1 = __p1; \
29087 __builtin_neon_vst4q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 49); \
29088 })
29089 #else
29090 #define vst4q_u16(__p0, __p1) __extension__ ({ \
29091 uint16x8x4_t __s1 = __p1; \
29092 uint16x8x4_t __rev1; \
29093 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
29094 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
29095 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
29096 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
29097 __builtin_neon_vst4q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 49); \
29098 })
29099 #endif
29100
29101 #ifdef __LITTLE_ENDIAN__
29102 #define vst4q_s8(__p0, __p1) __extension__ ({ \
29103 int8x16x4_t __s1 = __p1; \
29104 __builtin_neon_vst4q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 32); \
29105 })
29106 #else
29107 #define vst4q_s8(__p0, __p1) __extension__ ({ \
29108 int8x16x4_t __s1 = __p1; \
29109 int8x16x4_t __rev1; \
29110 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
29111 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
29112 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
29113 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
29114 __builtin_neon_vst4q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 32); \
29115 })
29116 #endif
29117
29118 #ifdef __LITTLE_ENDIAN__
29119 #define vst4q_f32(__p0, __p1) __extension__ ({ \
29120 float32x4x4_t __s1 = __p1; \
29121 __builtin_neon_vst4q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 41); \
29122 })
29123 #else
29124 #define vst4q_f32(__p0, __p1) __extension__ ({ \
29125 float32x4x4_t __s1 = __p1; \
29126 float32x4x4_t __rev1; \
29127 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
29128 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
29129 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
29130 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
29131 __builtin_neon_vst4q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 41); \
29132 })
29133 #endif
29134
29135 #ifdef __LITTLE_ENDIAN__
29136 #define vst4q_s32(__p0, __p1) __extension__ ({ \
29137 int32x4x4_t __s1 = __p1; \
29138 __builtin_neon_vst4q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 34); \
29139 })
29140 #else
29141 #define vst4q_s32(__p0, __p1) __extension__ ({ \
29142 int32x4x4_t __s1 = __p1; \
29143 int32x4x4_t __rev1; \
29144 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
29145 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
29146 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
29147 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
29148 __builtin_neon_vst4q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 34); \
29149 })
29150 #endif
29151
29152 #ifdef __LITTLE_ENDIAN__
29153 #define vst4q_s16(__p0, __p1) __extension__ ({ \
29154 int16x8x4_t __s1 = __p1; \
29155 __builtin_neon_vst4q_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 33); \
29156 })
29157 #else
29158 #define vst4q_s16(__p0, __p1) __extension__ ({ \
29159 int16x8x4_t __s1 = __p1; \
29160 int16x8x4_t __rev1; \
29161 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
29162 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
29163 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
29164 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
29165 __builtin_neon_vst4q_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 33); \
29166 })
29167 #endif
29168
29169 #ifdef __LITTLE_ENDIAN__
29170 #define vst4_u8(__p0, __p1) __extension__ ({ \
29171 uint8x8x4_t __s1 = __p1; \
29172 __builtin_neon_vst4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 16); \
29173 })
29174 #else
29175 #define vst4_u8(__p0, __p1) __extension__ ({ \
29176 uint8x8x4_t __s1 = __p1; \
29177 uint8x8x4_t __rev1; \
29178 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
29179 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
29180 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
29181 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
29182 __builtin_neon_vst4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 16); \
29183 })
29184 #endif
29185
29186 #ifdef __LITTLE_ENDIAN__
29187 #define vst4_u32(__p0, __p1) __extension__ ({ \
29188 uint32x2x4_t __s1 = __p1; \
29189 __builtin_neon_vst4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 18); \
29190 })
29191 #else
29192 #define vst4_u32(__p0, __p1) __extension__ ({ \
29193 uint32x2x4_t __s1 = __p1; \
29194 uint32x2x4_t __rev1; \
29195 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
29196 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
29197 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
29198 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
29199 __builtin_neon_vst4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 18); \
29200 })
29201 #endif
29202
29203 #define vst4_u64(__p0, __p1) __extension__ ({ \
29204 uint64x1x4_t __s1 = __p1; \
29205 __builtin_neon_vst4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 19); \
29206 })
29207 #ifdef __LITTLE_ENDIAN__
29208 #define vst4_u16(__p0, __p1) __extension__ ({ \
29209 uint16x4x4_t __s1 = __p1; \
29210 __builtin_neon_vst4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 17); \
29211 })
29212 #else
29213 #define vst4_u16(__p0, __p1) __extension__ ({ \
29214 uint16x4x4_t __s1 = __p1; \
29215 uint16x4x4_t __rev1; \
29216 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
29217 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
29218 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
29219 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
29220 __builtin_neon_vst4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 17); \
29221 })
29222 #endif
29223
29224 #ifdef __LITTLE_ENDIAN__
29225 #define vst4_s8(__p0, __p1) __extension__ ({ \
29226 int8x8x4_t __s1 = __p1; \
29227 __builtin_neon_vst4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 0); \
29228 })
29229 #else
29230 #define vst4_s8(__p0, __p1) __extension__ ({ \
29231 int8x8x4_t __s1 = __p1; \
29232 int8x8x4_t __rev1; \
29233 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
29234 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
29235 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
29236 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
29237 __builtin_neon_vst4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 0); \
29238 })
29239 #endif
29240
29241 #ifdef __LITTLE_ENDIAN__
29242 #define vst4_f32(__p0, __p1) __extension__ ({ \
29243 float32x2x4_t __s1 = __p1; \
29244 __builtin_neon_vst4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 9); \
29245 })
29246 #else
29247 #define vst4_f32(__p0, __p1) __extension__ ({ \
29248 float32x2x4_t __s1 = __p1; \
29249 float32x2x4_t __rev1; \
29250 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
29251 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
29252 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
29253 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
29254 __builtin_neon_vst4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 9); \
29255 })
29256 #endif
29257
29258 #ifdef __LITTLE_ENDIAN__
29259 #define vst4_s32(__p0, __p1) __extension__ ({ \
29260 int32x2x4_t __s1 = __p1; \
29261 __builtin_neon_vst4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 2); \
29262 })
29263 #else
29264 #define vst4_s32(__p0, __p1) __extension__ ({ \
29265 int32x2x4_t __s1 = __p1; \
29266 int32x2x4_t __rev1; \
29267 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
29268 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
29269 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
29270 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
29271 __builtin_neon_vst4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 2); \
29272 })
29273 #endif
29274
29275 #define vst4_s64(__p0, __p1) __extension__ ({ \
29276 int64x1x4_t __s1 = __p1; \
29277 __builtin_neon_vst4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 3); \
29278 })
29279 #ifdef __LITTLE_ENDIAN__
29280 #define vst4_s16(__p0, __p1) __extension__ ({ \
29281 int16x4x4_t __s1 = __p1; \
29282 __builtin_neon_vst4_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 1); \
29283 })
29284 #else
29285 #define vst4_s16(__p0, __p1) __extension__ ({ \
29286 int16x4x4_t __s1 = __p1; \
29287 int16x4x4_t __rev1; \
29288 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
29289 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
29290 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
29291 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
29292 __builtin_neon_vst4_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 1); \
29293 })
29294 #endif
29295
29296 #ifdef __LITTLE_ENDIAN__
29297 #define vst4_lane_p8(__p0, __p1, __p2) __extension__ ({ \
29298 poly8x8x4_t __s1 = __p1; \
29299 __builtin_neon_vst4_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 4); \
29300 })
29301 #else
29302 #define vst4_lane_p8(__p0, __p1, __p2) __extension__ ({ \
29303 poly8x8x4_t __s1 = __p1; \
29304 poly8x8x4_t __rev1; \
29305 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
29306 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
29307 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
29308 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
29309 __builtin_neon_vst4_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 4); \
29310 })
29311 #endif
29312
29313 #ifdef __LITTLE_ENDIAN__
29314 #define vst4_lane_p16(__p0, __p1, __p2) __extension__ ({ \
29315 poly16x4x4_t __s1 = __p1; \
29316 __builtin_neon_vst4_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 5); \
29317 })
29318 #else
29319 #define vst4_lane_p16(__p0, __p1, __p2) __extension__ ({ \
29320 poly16x4x4_t __s1 = __p1; \
29321 poly16x4x4_t __rev1; \
29322 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
29323 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
29324 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
29325 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
29326 __builtin_neon_vst4_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 5); \
29327 })
29328 #endif
29329
29330 #ifdef __LITTLE_ENDIAN__
29331 #define vst4q_lane_p16(__p0, __p1, __p2) __extension__ ({ \
29332 poly16x8x4_t __s1 = __p1; \
29333 __builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 37); \
29334 })
29335 #else
29336 #define vst4q_lane_p16(__p0, __p1, __p2) __extension__ ({ \
29337 poly16x8x4_t __s1 = __p1; \
29338 poly16x8x4_t __rev1; \
29339 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
29340 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
29341 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
29342 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
29343 __builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 37); \
29344 })
29345 #endif
29346
29347 #ifdef __LITTLE_ENDIAN__
29348 #define vst4q_lane_u32(__p0, __p1, __p2) __extension__ ({ \
29349 uint32x4x4_t __s1 = __p1; \
29350 __builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 50); \
29351 })
29352 #else
29353 #define vst4q_lane_u32(__p0, __p1, __p2) __extension__ ({ \
29354 uint32x4x4_t __s1 = __p1; \
29355 uint32x4x4_t __rev1; \
29356 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
29357 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
29358 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
29359 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
29360 __builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 50); \
29361 })
29362 #endif
29363
29364 #ifdef __LITTLE_ENDIAN__
29365 #define vst4q_lane_u16(__p0, __p1, __p2) __extension__ ({ \
29366 uint16x8x4_t __s1 = __p1; \
29367 __builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 49); \
29368 })
29369 #else
29370 #define vst4q_lane_u16(__p0, __p1, __p2) __extension__ ({ \
29371 uint16x8x4_t __s1 = __p1; \
29372 uint16x8x4_t __rev1; \
29373 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
29374 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
29375 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
29376 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
29377 __builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 49); \
29378 })
29379 #endif
29380
29381 #ifdef __LITTLE_ENDIAN__
29382 #define vst4q_lane_f32(__p0, __p1, __p2) __extension__ ({ \
29383 float32x4x4_t __s1 = __p1; \
29384 __builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 41); \
29385 })
29386 #else
29387 #define vst4q_lane_f32(__p0, __p1, __p2) __extension__ ({ \
29388 float32x4x4_t __s1 = __p1; \
29389 float32x4x4_t __rev1; \
29390 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
29391 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
29392 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
29393 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
29394 __builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 41); \
29395 })
29396 #endif
29397
29398 #ifdef __LITTLE_ENDIAN__
29399 #define vst4q_lane_s32(__p0, __p1, __p2) __extension__ ({ \
29400 int32x4x4_t __s1 = __p1; \
29401 __builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 34); \
29402 })
29403 #else
29404 #define vst4q_lane_s32(__p0, __p1, __p2) __extension__ ({ \
29405 int32x4x4_t __s1 = __p1; \
29406 int32x4x4_t __rev1; \
29407 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
29408 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
29409 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
29410 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
29411 __builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 34); \
29412 })
29413 #endif
29414
29415 #ifdef __LITTLE_ENDIAN__
29416 #define vst4q_lane_s16(__p0, __p1, __p2) __extension__ ({ \
29417 int16x8x4_t __s1 = __p1; \
29418 __builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 33); \
29419 })
29420 #else
29421 #define vst4q_lane_s16(__p0, __p1, __p2) __extension__ ({ \
29422 int16x8x4_t __s1 = __p1; \
29423 int16x8x4_t __rev1; \
29424 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
29425 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
29426 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
29427 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
29428 __builtin_neon_vst4q_lane_v(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 33); \
29429 })
29430 #endif
29431
29432 #ifdef __LITTLE_ENDIAN__
29433 #define vst4_lane_u8(__p0, __p1, __p2) __extension__ ({ \
29434 uint8x8x4_t __s1 = __p1; \
29435 __builtin_neon_vst4_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 16); \
29436 })
29437 #else
29438 #define vst4_lane_u8(__p0, __p1, __p2) __extension__ ({ \
29439 uint8x8x4_t __s1 = __p1; \
29440 uint8x8x4_t __rev1; \
29441 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
29442 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
29443 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
29444 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
29445 __builtin_neon_vst4_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 16); \
29446 })
29447 #endif
29448
29449 #ifdef __LITTLE_ENDIAN__
29450 #define vst4_lane_u32(__p0, __p1, __p2) __extension__ ({ \
29451 uint32x2x4_t __s1 = __p1; \
29452 __builtin_neon_vst4_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 18); \
29453 })
29454 #else
29455 #define vst4_lane_u32(__p0, __p1, __p2) __extension__ ({ \
29456 uint32x2x4_t __s1 = __p1; \
29457 uint32x2x4_t __rev1; \
29458 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
29459 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
29460 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
29461 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
29462 __builtin_neon_vst4_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 18); \
29463 })
29464 #endif
29465
29466 #ifdef __LITTLE_ENDIAN__
29467 #define vst4_lane_u16(__p0, __p1, __p2) __extension__ ({ \
29468 uint16x4x4_t __s1 = __p1; \
29469 __builtin_neon_vst4_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 17); \
29470 })
29471 #else
29472 #define vst4_lane_u16(__p0, __p1, __p2) __extension__ ({ \
29473 uint16x4x4_t __s1 = __p1; \
29474 uint16x4x4_t __rev1; \
29475 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
29476 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
29477 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
29478 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
29479 __builtin_neon_vst4_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 17); \
29480 })
29481 #endif
29482
29483 #ifdef __LITTLE_ENDIAN__
29484 #define vst4_lane_s8(__p0, __p1, __p2) __extension__ ({ \
29485 int8x8x4_t __s1 = __p1; \
29486 __builtin_neon_vst4_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 0); \
29487 })
29488 #else
29489 #define vst4_lane_s8(__p0, __p1, __p2) __extension__ ({ \
29490 int8x8x4_t __s1 = __p1; \
29491 int8x8x4_t __rev1; \
29492 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
29493 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
29494 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
29495 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
29496 __builtin_neon_vst4_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 0); \
29497 })
29498 #endif
29499
29500 #ifdef __LITTLE_ENDIAN__
29501 #define vst4_lane_f32(__p0, __p1, __p2) __extension__ ({ \
29502 float32x2x4_t __s1 = __p1; \
29503 __builtin_neon_vst4_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 9); \
29504 })
29505 #else
29506 #define vst4_lane_f32(__p0, __p1, __p2) __extension__ ({ \
29507 float32x2x4_t __s1 = __p1; \
29508 float32x2x4_t __rev1; \
29509 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
29510 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
29511 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
29512 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
29513 __builtin_neon_vst4_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 9); \
29514 })
29515 #endif
29516
29517 #ifdef __LITTLE_ENDIAN__
29518 #define vst4_lane_s32(__p0, __p1, __p2) __extension__ ({ \
29519 int32x2x4_t __s1 = __p1; \
29520 __builtin_neon_vst4_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 2); \
29521 })
29522 #else
29523 #define vst4_lane_s32(__p0, __p1, __p2) __extension__ ({ \
29524 int32x2x4_t __s1 = __p1; \
29525 int32x2x4_t __rev1; \
29526 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 1, 0); \
29527 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 1, 0); \
29528 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 1, 0); \
29529 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 1, 0); \
29530 __builtin_neon_vst4_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 2); \
29531 })
29532 #endif
29533
29534 #ifdef __LITTLE_ENDIAN__
29535 #define vst4_lane_s16(__p0, __p1, __p2) __extension__ ({ \
29536 int16x4x4_t __s1 = __p1; \
29537 __builtin_neon_vst4_lane_v(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 1); \
29538 })
29539 #else
29540 #define vst4_lane_s16(__p0, __p1, __p2) __extension__ ({ \
29541 int16x4x4_t __s1 = __p1; \
29542 int16x4x4_t __rev1; \
29543 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
29544 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
29545 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
29546 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
29547 __builtin_neon_vst4_lane_v(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 1); \
29548 })
29549 #endif
29550
29551 #ifdef __LITTLE_ENDIAN__
vsubq_u8(uint8x16_t __p0,uint8x16_t __p1)29552 __ai uint8x16_t vsubq_u8(uint8x16_t __p0, uint8x16_t __p1) {
29553 uint8x16_t __ret;
29554 __ret = __p0 - __p1;
29555 return __ret;
29556 }
29557 #else
vsubq_u8(uint8x16_t __p0,uint8x16_t __p1)29558 __ai uint8x16_t vsubq_u8(uint8x16_t __p0, uint8x16_t __p1) {
29559 uint8x16_t __ret;
29560 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
29561 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
29562 __ret = __rev0 - __rev1;
29563 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
29564 return __ret;
29565 }
29566 #endif
29567
29568 #ifdef __LITTLE_ENDIAN__
vsubq_u32(uint32x4_t __p0,uint32x4_t __p1)29569 __ai uint32x4_t vsubq_u32(uint32x4_t __p0, uint32x4_t __p1) {
29570 uint32x4_t __ret;
29571 __ret = __p0 - __p1;
29572 return __ret;
29573 }
29574 #else
vsubq_u32(uint32x4_t __p0,uint32x4_t __p1)29575 __ai uint32x4_t vsubq_u32(uint32x4_t __p0, uint32x4_t __p1) {
29576 uint32x4_t __ret;
29577 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
29578 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
29579 __ret = __rev0 - __rev1;
29580 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
29581 return __ret;
29582 }
29583 #endif
29584
29585 #ifdef __LITTLE_ENDIAN__
vsubq_u64(uint64x2_t __p0,uint64x2_t __p1)29586 __ai uint64x2_t vsubq_u64(uint64x2_t __p0, uint64x2_t __p1) {
29587 uint64x2_t __ret;
29588 __ret = __p0 - __p1;
29589 return __ret;
29590 }
29591 #else
vsubq_u64(uint64x2_t __p0,uint64x2_t __p1)29592 __ai uint64x2_t vsubq_u64(uint64x2_t __p0, uint64x2_t __p1) {
29593 uint64x2_t __ret;
29594 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
29595 uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
29596 __ret = __rev0 - __rev1;
29597 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
29598 return __ret;
29599 }
29600 #endif
29601
29602 #ifdef __LITTLE_ENDIAN__
vsubq_u16(uint16x8_t __p0,uint16x8_t __p1)29603 __ai uint16x8_t vsubq_u16(uint16x8_t __p0, uint16x8_t __p1) {
29604 uint16x8_t __ret;
29605 __ret = __p0 - __p1;
29606 return __ret;
29607 }
29608 #else
vsubq_u16(uint16x8_t __p0,uint16x8_t __p1)29609 __ai uint16x8_t vsubq_u16(uint16x8_t __p0, uint16x8_t __p1) {
29610 uint16x8_t __ret;
29611 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
29612 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
29613 __ret = __rev0 - __rev1;
29614 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
29615 return __ret;
29616 }
29617 #endif
29618
29619 #ifdef __LITTLE_ENDIAN__
vsubq_s8(int8x16_t __p0,int8x16_t __p1)29620 __ai int8x16_t vsubq_s8(int8x16_t __p0, int8x16_t __p1) {
29621 int8x16_t __ret;
29622 __ret = __p0 - __p1;
29623 return __ret;
29624 }
29625 #else
vsubq_s8(int8x16_t __p0,int8x16_t __p1)29626 __ai int8x16_t vsubq_s8(int8x16_t __p0, int8x16_t __p1) {
29627 int8x16_t __ret;
29628 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
29629 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
29630 __ret = __rev0 - __rev1;
29631 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
29632 return __ret;
29633 }
29634 #endif
29635
29636 #ifdef __LITTLE_ENDIAN__
vsubq_f32(float32x4_t __p0,float32x4_t __p1)29637 __ai float32x4_t vsubq_f32(float32x4_t __p0, float32x4_t __p1) {
29638 float32x4_t __ret;
29639 __ret = __p0 - __p1;
29640 return __ret;
29641 }
29642 #else
vsubq_f32(float32x4_t __p0,float32x4_t __p1)29643 __ai float32x4_t vsubq_f32(float32x4_t __p0, float32x4_t __p1) {
29644 float32x4_t __ret;
29645 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
29646 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
29647 __ret = __rev0 - __rev1;
29648 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
29649 return __ret;
29650 }
29651 #endif
29652
29653 #ifdef __LITTLE_ENDIAN__
vsubq_s32(int32x4_t __p0,int32x4_t __p1)29654 __ai int32x4_t vsubq_s32(int32x4_t __p0, int32x4_t __p1) {
29655 int32x4_t __ret;
29656 __ret = __p0 - __p1;
29657 return __ret;
29658 }
29659 #else
vsubq_s32(int32x4_t __p0,int32x4_t __p1)29660 __ai int32x4_t vsubq_s32(int32x4_t __p0, int32x4_t __p1) {
29661 int32x4_t __ret;
29662 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
29663 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
29664 __ret = __rev0 - __rev1;
29665 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
29666 return __ret;
29667 }
29668 #endif
29669
29670 #ifdef __LITTLE_ENDIAN__
vsubq_s64(int64x2_t __p0,int64x2_t __p1)29671 __ai int64x2_t vsubq_s64(int64x2_t __p0, int64x2_t __p1) {
29672 int64x2_t __ret;
29673 __ret = __p0 - __p1;
29674 return __ret;
29675 }
29676 #else
vsubq_s64(int64x2_t __p0,int64x2_t __p1)29677 __ai int64x2_t vsubq_s64(int64x2_t __p0, int64x2_t __p1) {
29678 int64x2_t __ret;
29679 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
29680 int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
29681 __ret = __rev0 - __rev1;
29682 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
29683 return __ret;
29684 }
29685 #endif
29686
29687 #ifdef __LITTLE_ENDIAN__
vsubq_s16(int16x8_t __p0,int16x8_t __p1)29688 __ai int16x8_t vsubq_s16(int16x8_t __p0, int16x8_t __p1) {
29689 int16x8_t __ret;
29690 __ret = __p0 - __p1;
29691 return __ret;
29692 }
29693 #else
vsubq_s16(int16x8_t __p0,int16x8_t __p1)29694 __ai int16x8_t vsubq_s16(int16x8_t __p0, int16x8_t __p1) {
29695 int16x8_t __ret;
29696 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
29697 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
29698 __ret = __rev0 - __rev1;
29699 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
29700 return __ret;
29701 }
29702 #endif
29703
29704 #ifdef __LITTLE_ENDIAN__
vsub_u8(uint8x8_t __p0,uint8x8_t __p1)29705 __ai uint8x8_t vsub_u8(uint8x8_t __p0, uint8x8_t __p1) {
29706 uint8x8_t __ret;
29707 __ret = __p0 - __p1;
29708 return __ret;
29709 }
29710 #else
vsub_u8(uint8x8_t __p0,uint8x8_t __p1)29711 __ai uint8x8_t vsub_u8(uint8x8_t __p0, uint8x8_t __p1) {
29712 uint8x8_t __ret;
29713 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
29714 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
29715 __ret = __rev0 - __rev1;
29716 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
29717 return __ret;
29718 }
29719 #endif
29720
29721 #ifdef __LITTLE_ENDIAN__
vsub_u32(uint32x2_t __p0,uint32x2_t __p1)29722 __ai uint32x2_t vsub_u32(uint32x2_t __p0, uint32x2_t __p1) {
29723 uint32x2_t __ret;
29724 __ret = __p0 - __p1;
29725 return __ret;
29726 }
29727 #else
vsub_u32(uint32x2_t __p0,uint32x2_t __p1)29728 __ai uint32x2_t vsub_u32(uint32x2_t __p0, uint32x2_t __p1) {
29729 uint32x2_t __ret;
29730 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
29731 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
29732 __ret = __rev0 - __rev1;
29733 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
29734 return __ret;
29735 }
29736 #endif
29737
vsub_u64(uint64x1_t __p0,uint64x1_t __p1)29738 __ai uint64x1_t vsub_u64(uint64x1_t __p0, uint64x1_t __p1) {
29739 uint64x1_t __ret;
29740 __ret = __p0 - __p1;
29741 return __ret;
29742 }
29743 #ifdef __LITTLE_ENDIAN__
vsub_u16(uint16x4_t __p0,uint16x4_t __p1)29744 __ai uint16x4_t vsub_u16(uint16x4_t __p0, uint16x4_t __p1) {
29745 uint16x4_t __ret;
29746 __ret = __p0 - __p1;
29747 return __ret;
29748 }
29749 #else
vsub_u16(uint16x4_t __p0,uint16x4_t __p1)29750 __ai uint16x4_t vsub_u16(uint16x4_t __p0, uint16x4_t __p1) {
29751 uint16x4_t __ret;
29752 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
29753 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
29754 __ret = __rev0 - __rev1;
29755 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
29756 return __ret;
29757 }
29758 #endif
29759
29760 #ifdef __LITTLE_ENDIAN__
vsub_s8(int8x8_t __p0,int8x8_t __p1)29761 __ai int8x8_t vsub_s8(int8x8_t __p0, int8x8_t __p1) {
29762 int8x8_t __ret;
29763 __ret = __p0 - __p1;
29764 return __ret;
29765 }
29766 #else
vsub_s8(int8x8_t __p0,int8x8_t __p1)29767 __ai int8x8_t vsub_s8(int8x8_t __p0, int8x8_t __p1) {
29768 int8x8_t __ret;
29769 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
29770 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
29771 __ret = __rev0 - __rev1;
29772 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
29773 return __ret;
29774 }
29775 #endif
29776
29777 #ifdef __LITTLE_ENDIAN__
vsub_f32(float32x2_t __p0,float32x2_t __p1)29778 __ai float32x2_t vsub_f32(float32x2_t __p0, float32x2_t __p1) {
29779 float32x2_t __ret;
29780 __ret = __p0 - __p1;
29781 return __ret;
29782 }
29783 #else
vsub_f32(float32x2_t __p0,float32x2_t __p1)29784 __ai float32x2_t vsub_f32(float32x2_t __p0, float32x2_t __p1) {
29785 float32x2_t __ret;
29786 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
29787 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
29788 __ret = __rev0 - __rev1;
29789 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
29790 return __ret;
29791 }
29792 #endif
29793
29794 #ifdef __LITTLE_ENDIAN__
vsub_s32(int32x2_t __p0,int32x2_t __p1)29795 __ai int32x2_t vsub_s32(int32x2_t __p0, int32x2_t __p1) {
29796 int32x2_t __ret;
29797 __ret = __p0 - __p1;
29798 return __ret;
29799 }
29800 #else
vsub_s32(int32x2_t __p0,int32x2_t __p1)29801 __ai int32x2_t vsub_s32(int32x2_t __p0, int32x2_t __p1) {
29802 int32x2_t __ret;
29803 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
29804 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
29805 __ret = __rev0 - __rev1;
29806 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
29807 return __ret;
29808 }
29809 #endif
29810
vsub_s64(int64x1_t __p0,int64x1_t __p1)29811 __ai int64x1_t vsub_s64(int64x1_t __p0, int64x1_t __p1) {
29812 int64x1_t __ret;
29813 __ret = __p0 - __p1;
29814 return __ret;
29815 }
29816 #ifdef __LITTLE_ENDIAN__
vsub_s16(int16x4_t __p0,int16x4_t __p1)29817 __ai int16x4_t vsub_s16(int16x4_t __p0, int16x4_t __p1) {
29818 int16x4_t __ret;
29819 __ret = __p0 - __p1;
29820 return __ret;
29821 }
29822 #else
vsub_s16(int16x4_t __p0,int16x4_t __p1)29823 __ai int16x4_t vsub_s16(int16x4_t __p0, int16x4_t __p1) {
29824 int16x4_t __ret;
29825 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
29826 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
29827 __ret = __rev0 - __rev1;
29828 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
29829 return __ret;
29830 }
29831 #endif
29832
29833 #ifdef __LITTLE_ENDIAN__
vsubhn_u32(uint32x4_t __p0,uint32x4_t __p1)29834 __ai uint16x4_t vsubhn_u32(uint32x4_t __p0, uint32x4_t __p1) {
29835 uint16x4_t __ret;
29836 __ret = (uint16x4_t) __builtin_neon_vsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 17);
29837 return __ret;
29838 }
29839 #else
vsubhn_u32(uint32x4_t __p0,uint32x4_t __p1)29840 __ai uint16x4_t vsubhn_u32(uint32x4_t __p0, uint32x4_t __p1) {
29841 uint16x4_t __ret;
29842 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
29843 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
29844 __ret = (uint16x4_t) __builtin_neon_vsubhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 17);
29845 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
29846 return __ret;
29847 }
__noswap_vsubhn_u32(uint32x4_t __p0,uint32x4_t __p1)29848 __ai uint16x4_t __noswap_vsubhn_u32(uint32x4_t __p0, uint32x4_t __p1) {
29849 uint16x4_t __ret;
29850 __ret = (uint16x4_t) __builtin_neon_vsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 17);
29851 return __ret;
29852 }
29853 #endif
29854
29855 #ifdef __LITTLE_ENDIAN__
vsubhn_u64(uint64x2_t __p0,uint64x2_t __p1)29856 __ai uint32x2_t vsubhn_u64(uint64x2_t __p0, uint64x2_t __p1) {
29857 uint32x2_t __ret;
29858 __ret = (uint32x2_t) __builtin_neon_vsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 18);
29859 return __ret;
29860 }
29861 #else
vsubhn_u64(uint64x2_t __p0,uint64x2_t __p1)29862 __ai uint32x2_t vsubhn_u64(uint64x2_t __p0, uint64x2_t __p1) {
29863 uint32x2_t __ret;
29864 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
29865 uint64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
29866 __ret = (uint32x2_t) __builtin_neon_vsubhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 18);
29867 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
29868 return __ret;
29869 }
__noswap_vsubhn_u64(uint64x2_t __p0,uint64x2_t __p1)29870 __ai uint32x2_t __noswap_vsubhn_u64(uint64x2_t __p0, uint64x2_t __p1) {
29871 uint32x2_t __ret;
29872 __ret = (uint32x2_t) __builtin_neon_vsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 18);
29873 return __ret;
29874 }
29875 #endif
29876
29877 #ifdef __LITTLE_ENDIAN__
vsubhn_u16(uint16x8_t __p0,uint16x8_t __p1)29878 __ai uint8x8_t vsubhn_u16(uint16x8_t __p0, uint16x8_t __p1) {
29879 uint8x8_t __ret;
29880 __ret = (uint8x8_t) __builtin_neon_vsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 16);
29881 return __ret;
29882 }
29883 #else
vsubhn_u16(uint16x8_t __p0,uint16x8_t __p1)29884 __ai uint8x8_t vsubhn_u16(uint16x8_t __p0, uint16x8_t __p1) {
29885 uint8x8_t __ret;
29886 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
29887 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
29888 __ret = (uint8x8_t) __builtin_neon_vsubhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 16);
29889 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
29890 return __ret;
29891 }
__noswap_vsubhn_u16(uint16x8_t __p0,uint16x8_t __p1)29892 __ai uint8x8_t __noswap_vsubhn_u16(uint16x8_t __p0, uint16x8_t __p1) {
29893 uint8x8_t __ret;
29894 __ret = (uint8x8_t) __builtin_neon_vsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 16);
29895 return __ret;
29896 }
29897 #endif
29898
29899 #ifdef __LITTLE_ENDIAN__
vsubhn_s32(int32x4_t __p0,int32x4_t __p1)29900 __ai int16x4_t vsubhn_s32(int32x4_t __p0, int32x4_t __p1) {
29901 int16x4_t __ret;
29902 __ret = (int16x4_t) __builtin_neon_vsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 1);
29903 return __ret;
29904 }
29905 #else
vsubhn_s32(int32x4_t __p0,int32x4_t __p1)29906 __ai int16x4_t vsubhn_s32(int32x4_t __p0, int32x4_t __p1) {
29907 int16x4_t __ret;
29908 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
29909 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
29910 __ret = (int16x4_t) __builtin_neon_vsubhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 1);
29911 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
29912 return __ret;
29913 }
__noswap_vsubhn_s32(int32x4_t __p0,int32x4_t __p1)29914 __ai int16x4_t __noswap_vsubhn_s32(int32x4_t __p0, int32x4_t __p1) {
29915 int16x4_t __ret;
29916 __ret = (int16x4_t) __builtin_neon_vsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 1);
29917 return __ret;
29918 }
29919 #endif
29920
29921 #ifdef __LITTLE_ENDIAN__
vsubhn_s64(int64x2_t __p0,int64x2_t __p1)29922 __ai int32x2_t vsubhn_s64(int64x2_t __p0, int64x2_t __p1) {
29923 int32x2_t __ret;
29924 __ret = (int32x2_t) __builtin_neon_vsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 2);
29925 return __ret;
29926 }
29927 #else
vsubhn_s64(int64x2_t __p0,int64x2_t __p1)29928 __ai int32x2_t vsubhn_s64(int64x2_t __p0, int64x2_t __p1) {
29929 int32x2_t __ret;
29930 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
29931 int64x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
29932 __ret = (int32x2_t) __builtin_neon_vsubhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 2);
29933 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
29934 return __ret;
29935 }
__noswap_vsubhn_s64(int64x2_t __p0,int64x2_t __p1)29936 __ai int32x2_t __noswap_vsubhn_s64(int64x2_t __p0, int64x2_t __p1) {
29937 int32x2_t __ret;
29938 __ret = (int32x2_t) __builtin_neon_vsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 2);
29939 return __ret;
29940 }
29941 #endif
29942
29943 #ifdef __LITTLE_ENDIAN__
vsubhn_s16(int16x8_t __p0,int16x8_t __p1)29944 __ai int8x8_t vsubhn_s16(int16x8_t __p0, int16x8_t __p1) {
29945 int8x8_t __ret;
29946 __ret = (int8x8_t) __builtin_neon_vsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 0);
29947 return __ret;
29948 }
29949 #else
vsubhn_s16(int16x8_t __p0,int16x8_t __p1)29950 __ai int8x8_t vsubhn_s16(int16x8_t __p0, int16x8_t __p1) {
29951 int8x8_t __ret;
29952 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
29953 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
29954 __ret = (int8x8_t) __builtin_neon_vsubhn_v((int8x16_t)__rev0, (int8x16_t)__rev1, 0);
29955 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
29956 return __ret;
29957 }
__noswap_vsubhn_s16(int16x8_t __p0,int16x8_t __p1)29958 __ai int8x8_t __noswap_vsubhn_s16(int16x8_t __p0, int16x8_t __p1) {
29959 int8x8_t __ret;
29960 __ret = (int8x8_t) __builtin_neon_vsubhn_v((int8x16_t)__p0, (int8x16_t)__p1, 0);
29961 return __ret;
29962 }
29963 #endif
29964
29965 #ifdef __LITTLE_ENDIAN__
vsubl_u8(uint8x8_t __p0,uint8x8_t __p1)29966 __ai uint16x8_t vsubl_u8(uint8x8_t __p0, uint8x8_t __p1) {
29967 uint16x8_t __ret;
29968 __ret = vmovl_u8(__p0) - vmovl_u8(__p1);
29969 return __ret;
29970 }
29971 #else
vsubl_u8(uint8x8_t __p0,uint8x8_t __p1)29972 __ai uint16x8_t vsubl_u8(uint8x8_t __p0, uint8x8_t __p1) {
29973 uint16x8_t __ret;
29974 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
29975 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
29976 __ret = __noswap_vmovl_u8(__rev0) - __noswap_vmovl_u8(__rev1);
29977 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
29978 return __ret;
29979 }
29980 #endif
29981
29982 #ifdef __LITTLE_ENDIAN__
vsubl_u32(uint32x2_t __p0,uint32x2_t __p1)29983 __ai uint64x2_t vsubl_u32(uint32x2_t __p0, uint32x2_t __p1) {
29984 uint64x2_t __ret;
29985 __ret = vmovl_u32(__p0) - vmovl_u32(__p1);
29986 return __ret;
29987 }
29988 #else
vsubl_u32(uint32x2_t __p0,uint32x2_t __p1)29989 __ai uint64x2_t vsubl_u32(uint32x2_t __p0, uint32x2_t __p1) {
29990 uint64x2_t __ret;
29991 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
29992 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
29993 __ret = __noswap_vmovl_u32(__rev0) - __noswap_vmovl_u32(__rev1);
29994 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
29995 return __ret;
29996 }
29997 #endif
29998
29999 #ifdef __LITTLE_ENDIAN__
vsubl_u16(uint16x4_t __p0,uint16x4_t __p1)30000 __ai uint32x4_t vsubl_u16(uint16x4_t __p0, uint16x4_t __p1) {
30001 uint32x4_t __ret;
30002 __ret = vmovl_u16(__p0) - vmovl_u16(__p1);
30003 return __ret;
30004 }
30005 #else
vsubl_u16(uint16x4_t __p0,uint16x4_t __p1)30006 __ai uint32x4_t vsubl_u16(uint16x4_t __p0, uint16x4_t __p1) {
30007 uint32x4_t __ret;
30008 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
30009 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
30010 __ret = __noswap_vmovl_u16(__rev0) - __noswap_vmovl_u16(__rev1);
30011 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
30012 return __ret;
30013 }
30014 #endif
30015
30016 #ifdef __LITTLE_ENDIAN__
vsubl_s8(int8x8_t __p0,int8x8_t __p1)30017 __ai int16x8_t vsubl_s8(int8x8_t __p0, int8x8_t __p1) {
30018 int16x8_t __ret;
30019 __ret = vmovl_s8(__p0) - vmovl_s8(__p1);
30020 return __ret;
30021 }
30022 #else
vsubl_s8(int8x8_t __p0,int8x8_t __p1)30023 __ai int16x8_t vsubl_s8(int8x8_t __p0, int8x8_t __p1) {
30024 int16x8_t __ret;
30025 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
30026 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
30027 __ret = __noswap_vmovl_s8(__rev0) - __noswap_vmovl_s8(__rev1);
30028 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
30029 return __ret;
30030 }
30031 #endif
30032
30033 #ifdef __LITTLE_ENDIAN__
vsubl_s32(int32x2_t __p0,int32x2_t __p1)30034 __ai int64x2_t vsubl_s32(int32x2_t __p0, int32x2_t __p1) {
30035 int64x2_t __ret;
30036 __ret = vmovl_s32(__p0) - vmovl_s32(__p1);
30037 return __ret;
30038 }
30039 #else
vsubl_s32(int32x2_t __p0,int32x2_t __p1)30040 __ai int64x2_t vsubl_s32(int32x2_t __p0, int32x2_t __p1) {
30041 int64x2_t __ret;
30042 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
30043 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
30044 __ret = __noswap_vmovl_s32(__rev0) - __noswap_vmovl_s32(__rev1);
30045 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
30046 return __ret;
30047 }
30048 #endif
30049
30050 #ifdef __LITTLE_ENDIAN__
vsubl_s16(int16x4_t __p0,int16x4_t __p1)30051 __ai int32x4_t vsubl_s16(int16x4_t __p0, int16x4_t __p1) {
30052 int32x4_t __ret;
30053 __ret = vmovl_s16(__p0) - vmovl_s16(__p1);
30054 return __ret;
30055 }
30056 #else
vsubl_s16(int16x4_t __p0,int16x4_t __p1)30057 __ai int32x4_t vsubl_s16(int16x4_t __p0, int16x4_t __p1) {
30058 int32x4_t __ret;
30059 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
30060 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
30061 __ret = __noswap_vmovl_s16(__rev0) - __noswap_vmovl_s16(__rev1);
30062 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
30063 return __ret;
30064 }
30065 #endif
30066
30067 #ifdef __LITTLE_ENDIAN__
vsubw_u8(uint16x8_t __p0,uint8x8_t __p1)30068 __ai uint16x8_t vsubw_u8(uint16x8_t __p0, uint8x8_t __p1) {
30069 uint16x8_t __ret;
30070 __ret = __p0 - vmovl_u8(__p1);
30071 return __ret;
30072 }
30073 #else
vsubw_u8(uint16x8_t __p0,uint8x8_t __p1)30074 __ai uint16x8_t vsubw_u8(uint16x8_t __p0, uint8x8_t __p1) {
30075 uint16x8_t __ret;
30076 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
30077 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
30078 __ret = __rev0 - __noswap_vmovl_u8(__rev1);
30079 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
30080 return __ret;
30081 }
30082 #endif
30083
30084 #ifdef __LITTLE_ENDIAN__
vsubw_u32(uint64x2_t __p0,uint32x2_t __p1)30085 __ai uint64x2_t vsubw_u32(uint64x2_t __p0, uint32x2_t __p1) {
30086 uint64x2_t __ret;
30087 __ret = __p0 - vmovl_u32(__p1);
30088 return __ret;
30089 }
30090 #else
vsubw_u32(uint64x2_t __p0,uint32x2_t __p1)30091 __ai uint64x2_t vsubw_u32(uint64x2_t __p0, uint32x2_t __p1) {
30092 uint64x2_t __ret;
30093 uint64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
30094 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
30095 __ret = __rev0 - __noswap_vmovl_u32(__rev1);
30096 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
30097 return __ret;
30098 }
30099 #endif
30100
30101 #ifdef __LITTLE_ENDIAN__
vsubw_u16(uint32x4_t __p0,uint16x4_t __p1)30102 __ai uint32x4_t vsubw_u16(uint32x4_t __p0, uint16x4_t __p1) {
30103 uint32x4_t __ret;
30104 __ret = __p0 - vmovl_u16(__p1);
30105 return __ret;
30106 }
30107 #else
vsubw_u16(uint32x4_t __p0,uint16x4_t __p1)30108 __ai uint32x4_t vsubw_u16(uint32x4_t __p0, uint16x4_t __p1) {
30109 uint32x4_t __ret;
30110 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
30111 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
30112 __ret = __rev0 - __noswap_vmovl_u16(__rev1);
30113 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
30114 return __ret;
30115 }
30116 #endif
30117
30118 #ifdef __LITTLE_ENDIAN__
vsubw_s8(int16x8_t __p0,int8x8_t __p1)30119 __ai int16x8_t vsubw_s8(int16x8_t __p0, int8x8_t __p1) {
30120 int16x8_t __ret;
30121 __ret = __p0 - vmovl_s8(__p1);
30122 return __ret;
30123 }
30124 #else
vsubw_s8(int16x8_t __p0,int8x8_t __p1)30125 __ai int16x8_t vsubw_s8(int16x8_t __p0, int8x8_t __p1) {
30126 int16x8_t __ret;
30127 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
30128 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
30129 __ret = __rev0 - __noswap_vmovl_s8(__rev1);
30130 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
30131 return __ret;
30132 }
30133 #endif
30134
30135 #ifdef __LITTLE_ENDIAN__
vsubw_s32(int64x2_t __p0,int32x2_t __p1)30136 __ai int64x2_t vsubw_s32(int64x2_t __p0, int32x2_t __p1) {
30137 int64x2_t __ret;
30138 __ret = __p0 - vmovl_s32(__p1);
30139 return __ret;
30140 }
30141 #else
vsubw_s32(int64x2_t __p0,int32x2_t __p1)30142 __ai int64x2_t vsubw_s32(int64x2_t __p0, int32x2_t __p1) {
30143 int64x2_t __ret;
30144 int64x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
30145 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
30146 __ret = __rev0 - __noswap_vmovl_s32(__rev1);
30147 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
30148 return __ret;
30149 }
30150 #endif
30151
30152 #ifdef __LITTLE_ENDIAN__
vsubw_s16(int32x4_t __p0,int16x4_t __p1)30153 __ai int32x4_t vsubw_s16(int32x4_t __p0, int16x4_t __p1) {
30154 int32x4_t __ret;
30155 __ret = __p0 - vmovl_s16(__p1);
30156 return __ret;
30157 }
30158 #else
vsubw_s16(int32x4_t __p0,int16x4_t __p1)30159 __ai int32x4_t vsubw_s16(int32x4_t __p0, int16x4_t __p1) {
30160 int32x4_t __ret;
30161 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
30162 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
30163 __ret = __rev0 - __noswap_vmovl_s16(__rev1);
30164 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
30165 return __ret;
30166 }
30167 #endif
30168
30169 #ifdef __LITTLE_ENDIAN__
vtbl1_p8(poly8x8_t __p0,uint8x8_t __p1)30170 __ai poly8x8_t vtbl1_p8(poly8x8_t __p0, uint8x8_t __p1) {
30171 poly8x8_t __ret;
30172 __ret = (poly8x8_t) __builtin_neon_vtbl1_v((int8x8_t)__p0, (int8x8_t)__p1, 4);
30173 return __ret;
30174 }
30175 #else
vtbl1_p8(poly8x8_t __p0,uint8x8_t __p1)30176 __ai poly8x8_t vtbl1_p8(poly8x8_t __p0, uint8x8_t __p1) {
30177 poly8x8_t __ret;
30178 poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
30179 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
30180 __ret = (poly8x8_t) __builtin_neon_vtbl1_v((int8x8_t)__rev0, (int8x8_t)__rev1, 4);
30181 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
30182 return __ret;
30183 }
30184 #endif
30185
30186 #ifdef __LITTLE_ENDIAN__
vtbl1_u8(uint8x8_t __p0,uint8x8_t __p1)30187 __ai uint8x8_t vtbl1_u8(uint8x8_t __p0, uint8x8_t __p1) {
30188 uint8x8_t __ret;
30189 __ret = (uint8x8_t) __builtin_neon_vtbl1_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
30190 return __ret;
30191 }
30192 #else
vtbl1_u8(uint8x8_t __p0,uint8x8_t __p1)30193 __ai uint8x8_t vtbl1_u8(uint8x8_t __p0, uint8x8_t __p1) {
30194 uint8x8_t __ret;
30195 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
30196 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
30197 __ret = (uint8x8_t) __builtin_neon_vtbl1_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
30198 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
30199 return __ret;
30200 }
30201 #endif
30202
30203 #ifdef __LITTLE_ENDIAN__
vtbl1_s8(int8x8_t __p0,int8x8_t __p1)30204 __ai int8x8_t vtbl1_s8(int8x8_t __p0, int8x8_t __p1) {
30205 int8x8_t __ret;
30206 __ret = (int8x8_t) __builtin_neon_vtbl1_v((int8x8_t)__p0, (int8x8_t)__p1, 0);
30207 return __ret;
30208 }
30209 #else
vtbl1_s8(int8x8_t __p0,int8x8_t __p1)30210 __ai int8x8_t vtbl1_s8(int8x8_t __p0, int8x8_t __p1) {
30211 int8x8_t __ret;
30212 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
30213 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
30214 __ret = (int8x8_t) __builtin_neon_vtbl1_v((int8x8_t)__rev0, (int8x8_t)__rev1, 0);
30215 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
30216 return __ret;
30217 }
30218 #endif
30219
30220 #ifdef __LITTLE_ENDIAN__
vtbl2_p8(poly8x8x2_t __p0,uint8x8_t __p1)30221 __ai poly8x8_t vtbl2_p8(poly8x8x2_t __p0, uint8x8_t __p1) {
30222 poly8x8_t __ret;
30223 __ret = (poly8x8_t) __builtin_neon_vtbl2_v((int8x8_t)__p0.val[0], (int8x8_t)__p0.val[1], (int8x8_t)__p1, 4);
30224 return __ret;
30225 }
30226 #else
vtbl2_p8(poly8x8x2_t __p0,uint8x8_t __p1)30227 __ai poly8x8_t vtbl2_p8(poly8x8x2_t __p0, uint8x8_t __p1) {
30228 poly8x8_t __ret;
30229 poly8x8x2_t __rev0;
30230 __rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
30231 __rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
30232 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
30233 __ret = (poly8x8_t) __builtin_neon_vtbl2_v((int8x8_t)__rev0.val[0], (int8x8_t)__rev0.val[1], (int8x8_t)__rev1, 4);
30234 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
30235 return __ret;
30236 }
30237 #endif
30238
30239 #ifdef __LITTLE_ENDIAN__
vtbl2_u8(uint8x8x2_t __p0,uint8x8_t __p1)30240 __ai uint8x8_t vtbl2_u8(uint8x8x2_t __p0, uint8x8_t __p1) {
30241 uint8x8_t __ret;
30242 __ret = (uint8x8_t) __builtin_neon_vtbl2_v((int8x8_t)__p0.val[0], (int8x8_t)__p0.val[1], (int8x8_t)__p1, 16);
30243 return __ret;
30244 }
30245 #else
vtbl2_u8(uint8x8x2_t __p0,uint8x8_t __p1)30246 __ai uint8x8_t vtbl2_u8(uint8x8x2_t __p0, uint8x8_t __p1) {
30247 uint8x8_t __ret;
30248 uint8x8x2_t __rev0;
30249 __rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
30250 __rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
30251 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
30252 __ret = (uint8x8_t) __builtin_neon_vtbl2_v((int8x8_t)__rev0.val[0], (int8x8_t)__rev0.val[1], (int8x8_t)__rev1, 16);
30253 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
30254 return __ret;
30255 }
30256 #endif
30257
30258 #ifdef __LITTLE_ENDIAN__
vtbl2_s8(int8x8x2_t __p0,int8x8_t __p1)30259 __ai int8x8_t vtbl2_s8(int8x8x2_t __p0, int8x8_t __p1) {
30260 int8x8_t __ret;
30261 __ret = (int8x8_t) __builtin_neon_vtbl2_v((int8x8_t)__p0.val[0], (int8x8_t)__p0.val[1], (int8x8_t)__p1, 0);
30262 return __ret;
30263 }
30264 #else
vtbl2_s8(int8x8x2_t __p0,int8x8_t __p1)30265 __ai int8x8_t vtbl2_s8(int8x8x2_t __p0, int8x8_t __p1) {
30266 int8x8_t __ret;
30267 int8x8x2_t __rev0;
30268 __rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
30269 __rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
30270 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
30271 __ret = (int8x8_t) __builtin_neon_vtbl2_v((int8x8_t)__rev0.val[0], (int8x8_t)__rev0.val[1], (int8x8_t)__rev1, 0);
30272 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
30273 return __ret;
30274 }
30275 #endif
30276
30277 #ifdef __LITTLE_ENDIAN__
vtbl3_p8(poly8x8x3_t __p0,uint8x8_t __p1)30278 __ai poly8x8_t vtbl3_p8(poly8x8x3_t __p0, uint8x8_t __p1) {
30279 poly8x8_t __ret;
30280 __ret = (poly8x8_t) __builtin_neon_vtbl3_v((int8x8_t)__p0.val[0], (int8x8_t)__p0.val[1], (int8x8_t)__p0.val[2], (int8x8_t)__p1, 4);
30281 return __ret;
30282 }
30283 #else
vtbl3_p8(poly8x8x3_t __p0,uint8x8_t __p1)30284 __ai poly8x8_t vtbl3_p8(poly8x8x3_t __p0, uint8x8_t __p1) {
30285 poly8x8_t __ret;
30286 poly8x8x3_t __rev0;
30287 __rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
30288 __rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
30289 __rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], 7, 6, 5, 4, 3, 2, 1, 0);
30290 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
30291 __ret = (poly8x8_t) __builtin_neon_vtbl3_v((int8x8_t)__rev0.val[0], (int8x8_t)__rev0.val[1], (int8x8_t)__rev0.val[2], (int8x8_t)__rev1, 4);
30292 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
30293 return __ret;
30294 }
30295 #endif
30296
30297 #ifdef __LITTLE_ENDIAN__
vtbl3_u8(uint8x8x3_t __p0,uint8x8_t __p1)30298 __ai uint8x8_t vtbl3_u8(uint8x8x3_t __p0, uint8x8_t __p1) {
30299 uint8x8_t __ret;
30300 __ret = (uint8x8_t) __builtin_neon_vtbl3_v((int8x8_t)__p0.val[0], (int8x8_t)__p0.val[1], (int8x8_t)__p0.val[2], (int8x8_t)__p1, 16);
30301 return __ret;
30302 }
30303 #else
vtbl3_u8(uint8x8x3_t __p0,uint8x8_t __p1)30304 __ai uint8x8_t vtbl3_u8(uint8x8x3_t __p0, uint8x8_t __p1) {
30305 uint8x8_t __ret;
30306 uint8x8x3_t __rev0;
30307 __rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
30308 __rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
30309 __rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], 7, 6, 5, 4, 3, 2, 1, 0);
30310 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
30311 __ret = (uint8x8_t) __builtin_neon_vtbl3_v((int8x8_t)__rev0.val[0], (int8x8_t)__rev0.val[1], (int8x8_t)__rev0.val[2], (int8x8_t)__rev1, 16);
30312 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
30313 return __ret;
30314 }
30315 #endif
30316
30317 #ifdef __LITTLE_ENDIAN__
vtbl3_s8(int8x8x3_t __p0,int8x8_t __p1)30318 __ai int8x8_t vtbl3_s8(int8x8x3_t __p0, int8x8_t __p1) {
30319 int8x8_t __ret;
30320 __ret = (int8x8_t) __builtin_neon_vtbl3_v((int8x8_t)__p0.val[0], (int8x8_t)__p0.val[1], (int8x8_t)__p0.val[2], (int8x8_t)__p1, 0);
30321 return __ret;
30322 }
30323 #else
vtbl3_s8(int8x8x3_t __p0,int8x8_t __p1)30324 __ai int8x8_t vtbl3_s8(int8x8x3_t __p0, int8x8_t __p1) {
30325 int8x8_t __ret;
30326 int8x8x3_t __rev0;
30327 __rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
30328 __rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
30329 __rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], 7, 6, 5, 4, 3, 2, 1, 0);
30330 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
30331 __ret = (int8x8_t) __builtin_neon_vtbl3_v((int8x8_t)__rev0.val[0], (int8x8_t)__rev0.val[1], (int8x8_t)__rev0.val[2], (int8x8_t)__rev1, 0);
30332 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
30333 return __ret;
30334 }
30335 #endif
30336
30337 #ifdef __LITTLE_ENDIAN__
vtbl4_p8(poly8x8x4_t __p0,uint8x8_t __p1)30338 __ai poly8x8_t vtbl4_p8(poly8x8x4_t __p0, uint8x8_t __p1) {
30339 poly8x8_t __ret;
30340 __ret = (poly8x8_t) __builtin_neon_vtbl4_v((int8x8_t)__p0.val[0], (int8x8_t)__p0.val[1], (int8x8_t)__p0.val[2], (int8x8_t)__p0.val[3], (int8x8_t)__p1, 4);
30341 return __ret;
30342 }
30343 #else
vtbl4_p8(poly8x8x4_t __p0,uint8x8_t __p1)30344 __ai poly8x8_t vtbl4_p8(poly8x8x4_t __p0, uint8x8_t __p1) {
30345 poly8x8_t __ret;
30346 poly8x8x4_t __rev0;
30347 __rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
30348 __rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
30349 __rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], 7, 6, 5, 4, 3, 2, 1, 0);
30350 __rev0.val[3] = __builtin_shufflevector(__p0.val[3], __p0.val[3], 7, 6, 5, 4, 3, 2, 1, 0);
30351 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
30352 __ret = (poly8x8_t) __builtin_neon_vtbl4_v((int8x8_t)__rev0.val[0], (int8x8_t)__rev0.val[1], (int8x8_t)__rev0.val[2], (int8x8_t)__rev0.val[3], (int8x8_t)__rev1, 4);
30353 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
30354 return __ret;
30355 }
30356 #endif
30357
30358 #ifdef __LITTLE_ENDIAN__
vtbl4_u8(uint8x8x4_t __p0,uint8x8_t __p1)30359 __ai uint8x8_t vtbl4_u8(uint8x8x4_t __p0, uint8x8_t __p1) {
30360 uint8x8_t __ret;
30361 __ret = (uint8x8_t) __builtin_neon_vtbl4_v((int8x8_t)__p0.val[0], (int8x8_t)__p0.val[1], (int8x8_t)__p0.val[2], (int8x8_t)__p0.val[3], (int8x8_t)__p1, 16);
30362 return __ret;
30363 }
30364 #else
vtbl4_u8(uint8x8x4_t __p0,uint8x8_t __p1)30365 __ai uint8x8_t vtbl4_u8(uint8x8x4_t __p0, uint8x8_t __p1) {
30366 uint8x8_t __ret;
30367 uint8x8x4_t __rev0;
30368 __rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
30369 __rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
30370 __rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], 7, 6, 5, 4, 3, 2, 1, 0);
30371 __rev0.val[3] = __builtin_shufflevector(__p0.val[3], __p0.val[3], 7, 6, 5, 4, 3, 2, 1, 0);
30372 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
30373 __ret = (uint8x8_t) __builtin_neon_vtbl4_v((int8x8_t)__rev0.val[0], (int8x8_t)__rev0.val[1], (int8x8_t)__rev0.val[2], (int8x8_t)__rev0.val[3], (int8x8_t)__rev1, 16);
30374 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
30375 return __ret;
30376 }
30377 #endif
30378
30379 #ifdef __LITTLE_ENDIAN__
vtbl4_s8(int8x8x4_t __p0,int8x8_t __p1)30380 __ai int8x8_t vtbl4_s8(int8x8x4_t __p0, int8x8_t __p1) {
30381 int8x8_t __ret;
30382 __ret = (int8x8_t) __builtin_neon_vtbl4_v((int8x8_t)__p0.val[0], (int8x8_t)__p0.val[1], (int8x8_t)__p0.val[2], (int8x8_t)__p0.val[3], (int8x8_t)__p1, 0);
30383 return __ret;
30384 }
30385 #else
vtbl4_s8(int8x8x4_t __p0,int8x8_t __p1)30386 __ai int8x8_t vtbl4_s8(int8x8x4_t __p0, int8x8_t __p1) {
30387 int8x8_t __ret;
30388 int8x8x4_t __rev0;
30389 __rev0.val[0] = __builtin_shufflevector(__p0.val[0], __p0.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
30390 __rev0.val[1] = __builtin_shufflevector(__p0.val[1], __p0.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
30391 __rev0.val[2] = __builtin_shufflevector(__p0.val[2], __p0.val[2], 7, 6, 5, 4, 3, 2, 1, 0);
30392 __rev0.val[3] = __builtin_shufflevector(__p0.val[3], __p0.val[3], 7, 6, 5, 4, 3, 2, 1, 0);
30393 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
30394 __ret = (int8x8_t) __builtin_neon_vtbl4_v((int8x8_t)__rev0.val[0], (int8x8_t)__rev0.val[1], (int8x8_t)__rev0.val[2], (int8x8_t)__rev0.val[3], (int8x8_t)__rev1, 0);
30395 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
30396 return __ret;
30397 }
30398 #endif
30399
30400 #ifdef __LITTLE_ENDIAN__
vtbx1_p8(poly8x8_t __p0,poly8x8_t __p1,uint8x8_t __p2)30401 __ai poly8x8_t vtbx1_p8(poly8x8_t __p0, poly8x8_t __p1, uint8x8_t __p2) {
30402 poly8x8_t __ret;
30403 __ret = (poly8x8_t) __builtin_neon_vtbx1_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 4);
30404 return __ret;
30405 }
30406 #else
vtbx1_p8(poly8x8_t __p0,poly8x8_t __p1,uint8x8_t __p2)30407 __ai poly8x8_t vtbx1_p8(poly8x8_t __p0, poly8x8_t __p1, uint8x8_t __p2) {
30408 poly8x8_t __ret;
30409 poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
30410 poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
30411 uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
30412 __ret = (poly8x8_t) __builtin_neon_vtbx1_v((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 4);
30413 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
30414 return __ret;
30415 }
30416 #endif
30417
30418 #ifdef __LITTLE_ENDIAN__
vtbx1_u8(uint8x8_t __p0,uint8x8_t __p1,uint8x8_t __p2)30419 __ai uint8x8_t vtbx1_u8(uint8x8_t __p0, uint8x8_t __p1, uint8x8_t __p2) {
30420 uint8x8_t __ret;
30421 __ret = (uint8x8_t) __builtin_neon_vtbx1_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 16);
30422 return __ret;
30423 }
30424 #else
vtbx1_u8(uint8x8_t __p0,uint8x8_t __p1,uint8x8_t __p2)30425 __ai uint8x8_t vtbx1_u8(uint8x8_t __p0, uint8x8_t __p1, uint8x8_t __p2) {
30426 uint8x8_t __ret;
30427 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
30428 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
30429 uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
30430 __ret = (uint8x8_t) __builtin_neon_vtbx1_v((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 16);
30431 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
30432 return __ret;
30433 }
30434 #endif
30435
30436 #ifdef __LITTLE_ENDIAN__
vtbx1_s8(int8x8_t __p0,int8x8_t __p1,int8x8_t __p2)30437 __ai int8x8_t vtbx1_s8(int8x8_t __p0, int8x8_t __p1, int8x8_t __p2) {
30438 int8x8_t __ret;
30439 __ret = (int8x8_t) __builtin_neon_vtbx1_v((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 0);
30440 return __ret;
30441 }
30442 #else
vtbx1_s8(int8x8_t __p0,int8x8_t __p1,int8x8_t __p2)30443 __ai int8x8_t vtbx1_s8(int8x8_t __p0, int8x8_t __p1, int8x8_t __p2) {
30444 int8x8_t __ret;
30445 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
30446 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
30447 int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
30448 __ret = (int8x8_t) __builtin_neon_vtbx1_v((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 0);
30449 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
30450 return __ret;
30451 }
30452 #endif
30453
30454 #ifdef __LITTLE_ENDIAN__
vtbx2_p8(poly8x8_t __p0,poly8x8x2_t __p1,uint8x8_t __p2)30455 __ai poly8x8_t vtbx2_p8(poly8x8_t __p0, poly8x8x2_t __p1, uint8x8_t __p2) {
30456 poly8x8_t __ret;
30457 __ret = (poly8x8_t) __builtin_neon_vtbx2_v((int8x8_t)__p0, (int8x8_t)__p1.val[0], (int8x8_t)__p1.val[1], (int8x8_t)__p2, 4);
30458 return __ret;
30459 }
30460 #else
vtbx2_p8(poly8x8_t __p0,poly8x8x2_t __p1,uint8x8_t __p2)30461 __ai poly8x8_t vtbx2_p8(poly8x8_t __p0, poly8x8x2_t __p1, uint8x8_t __p2) {
30462 poly8x8_t __ret;
30463 poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
30464 poly8x8x2_t __rev1;
30465 __rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
30466 __rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
30467 uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
30468 __ret = (poly8x8_t) __builtin_neon_vtbx2_v((int8x8_t)__rev0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev2, 4);
30469 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
30470 return __ret;
30471 }
30472 #endif
30473
30474 #ifdef __LITTLE_ENDIAN__
vtbx2_u8(uint8x8_t __p0,uint8x8x2_t __p1,uint8x8_t __p2)30475 __ai uint8x8_t vtbx2_u8(uint8x8_t __p0, uint8x8x2_t __p1, uint8x8_t __p2) {
30476 uint8x8_t __ret;
30477 __ret = (uint8x8_t) __builtin_neon_vtbx2_v((int8x8_t)__p0, (int8x8_t)__p1.val[0], (int8x8_t)__p1.val[1], (int8x8_t)__p2, 16);
30478 return __ret;
30479 }
30480 #else
vtbx2_u8(uint8x8_t __p0,uint8x8x2_t __p1,uint8x8_t __p2)30481 __ai uint8x8_t vtbx2_u8(uint8x8_t __p0, uint8x8x2_t __p1, uint8x8_t __p2) {
30482 uint8x8_t __ret;
30483 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
30484 uint8x8x2_t __rev1;
30485 __rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
30486 __rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
30487 uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
30488 __ret = (uint8x8_t) __builtin_neon_vtbx2_v((int8x8_t)__rev0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev2, 16);
30489 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
30490 return __ret;
30491 }
30492 #endif
30493
30494 #ifdef __LITTLE_ENDIAN__
vtbx2_s8(int8x8_t __p0,int8x8x2_t __p1,int8x8_t __p2)30495 __ai int8x8_t vtbx2_s8(int8x8_t __p0, int8x8x2_t __p1, int8x8_t __p2) {
30496 int8x8_t __ret;
30497 __ret = (int8x8_t) __builtin_neon_vtbx2_v((int8x8_t)__p0, (int8x8_t)__p1.val[0], (int8x8_t)__p1.val[1], (int8x8_t)__p2, 0);
30498 return __ret;
30499 }
30500 #else
vtbx2_s8(int8x8_t __p0,int8x8x2_t __p1,int8x8_t __p2)30501 __ai int8x8_t vtbx2_s8(int8x8_t __p0, int8x8x2_t __p1, int8x8_t __p2) {
30502 int8x8_t __ret;
30503 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
30504 int8x8x2_t __rev1;
30505 __rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
30506 __rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
30507 int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
30508 __ret = (int8x8_t) __builtin_neon_vtbx2_v((int8x8_t)__rev0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev2, 0);
30509 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
30510 return __ret;
30511 }
30512 #endif
30513
30514 #ifdef __LITTLE_ENDIAN__
vtbx3_p8(poly8x8_t __p0,poly8x8x3_t __p1,uint8x8_t __p2)30515 __ai poly8x8_t vtbx3_p8(poly8x8_t __p0, poly8x8x3_t __p1, uint8x8_t __p2) {
30516 poly8x8_t __ret;
30517 __ret = (poly8x8_t) __builtin_neon_vtbx3_v((int8x8_t)__p0, (int8x8_t)__p1.val[0], (int8x8_t)__p1.val[1], (int8x8_t)__p1.val[2], (int8x8_t)__p2, 4);
30518 return __ret;
30519 }
30520 #else
vtbx3_p8(poly8x8_t __p0,poly8x8x3_t __p1,uint8x8_t __p2)30521 __ai poly8x8_t vtbx3_p8(poly8x8_t __p0, poly8x8x3_t __p1, uint8x8_t __p2) {
30522 poly8x8_t __ret;
30523 poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
30524 poly8x8x3_t __rev1;
30525 __rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
30526 __rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
30527 __rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], 7, 6, 5, 4, 3, 2, 1, 0);
30528 uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
30529 __ret = (poly8x8_t) __builtin_neon_vtbx3_v((int8x8_t)__rev0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev2, 4);
30530 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
30531 return __ret;
30532 }
30533 #endif
30534
30535 #ifdef __LITTLE_ENDIAN__
vtbx3_u8(uint8x8_t __p0,uint8x8x3_t __p1,uint8x8_t __p2)30536 __ai uint8x8_t vtbx3_u8(uint8x8_t __p0, uint8x8x3_t __p1, uint8x8_t __p2) {
30537 uint8x8_t __ret;
30538 __ret = (uint8x8_t) __builtin_neon_vtbx3_v((int8x8_t)__p0, (int8x8_t)__p1.val[0], (int8x8_t)__p1.val[1], (int8x8_t)__p1.val[2], (int8x8_t)__p2, 16);
30539 return __ret;
30540 }
30541 #else
vtbx3_u8(uint8x8_t __p0,uint8x8x3_t __p1,uint8x8_t __p2)30542 __ai uint8x8_t vtbx3_u8(uint8x8_t __p0, uint8x8x3_t __p1, uint8x8_t __p2) {
30543 uint8x8_t __ret;
30544 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
30545 uint8x8x3_t __rev1;
30546 __rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
30547 __rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
30548 __rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], 7, 6, 5, 4, 3, 2, 1, 0);
30549 uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
30550 __ret = (uint8x8_t) __builtin_neon_vtbx3_v((int8x8_t)__rev0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev2, 16);
30551 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
30552 return __ret;
30553 }
30554 #endif
30555
30556 #ifdef __LITTLE_ENDIAN__
vtbx3_s8(int8x8_t __p0,int8x8x3_t __p1,int8x8_t __p2)30557 __ai int8x8_t vtbx3_s8(int8x8_t __p0, int8x8x3_t __p1, int8x8_t __p2) {
30558 int8x8_t __ret;
30559 __ret = (int8x8_t) __builtin_neon_vtbx3_v((int8x8_t)__p0, (int8x8_t)__p1.val[0], (int8x8_t)__p1.val[1], (int8x8_t)__p1.val[2], (int8x8_t)__p2, 0);
30560 return __ret;
30561 }
30562 #else
vtbx3_s8(int8x8_t __p0,int8x8x3_t __p1,int8x8_t __p2)30563 __ai int8x8_t vtbx3_s8(int8x8_t __p0, int8x8x3_t __p1, int8x8_t __p2) {
30564 int8x8_t __ret;
30565 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
30566 int8x8x3_t __rev1;
30567 __rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
30568 __rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
30569 __rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], 7, 6, 5, 4, 3, 2, 1, 0);
30570 int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
30571 __ret = (int8x8_t) __builtin_neon_vtbx3_v((int8x8_t)__rev0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev2, 0);
30572 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
30573 return __ret;
30574 }
30575 #endif
30576
30577 #ifdef __LITTLE_ENDIAN__
vtbx4_p8(poly8x8_t __p0,poly8x8x4_t __p1,uint8x8_t __p2)30578 __ai poly8x8_t vtbx4_p8(poly8x8_t __p0, poly8x8x4_t __p1, uint8x8_t __p2) {
30579 poly8x8_t __ret;
30580 __ret = (poly8x8_t) __builtin_neon_vtbx4_v((int8x8_t)__p0, (int8x8_t)__p1.val[0], (int8x8_t)__p1.val[1], (int8x8_t)__p1.val[2], (int8x8_t)__p1.val[3], (int8x8_t)__p2, 4);
30581 return __ret;
30582 }
30583 #else
vtbx4_p8(poly8x8_t __p0,poly8x8x4_t __p1,uint8x8_t __p2)30584 __ai poly8x8_t vtbx4_p8(poly8x8_t __p0, poly8x8x4_t __p1, uint8x8_t __p2) {
30585 poly8x8_t __ret;
30586 poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
30587 poly8x8x4_t __rev1;
30588 __rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
30589 __rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
30590 __rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], 7, 6, 5, 4, 3, 2, 1, 0);
30591 __rev1.val[3] = __builtin_shufflevector(__p1.val[3], __p1.val[3], 7, 6, 5, 4, 3, 2, 1, 0);
30592 uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
30593 __ret = (poly8x8_t) __builtin_neon_vtbx4_v((int8x8_t)__rev0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], (int8x8_t)__rev2, 4);
30594 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
30595 return __ret;
30596 }
30597 #endif
30598
30599 #ifdef __LITTLE_ENDIAN__
vtbx4_u8(uint8x8_t __p0,uint8x8x4_t __p1,uint8x8_t __p2)30600 __ai uint8x8_t vtbx4_u8(uint8x8_t __p0, uint8x8x4_t __p1, uint8x8_t __p2) {
30601 uint8x8_t __ret;
30602 __ret = (uint8x8_t) __builtin_neon_vtbx4_v((int8x8_t)__p0, (int8x8_t)__p1.val[0], (int8x8_t)__p1.val[1], (int8x8_t)__p1.val[2], (int8x8_t)__p1.val[3], (int8x8_t)__p2, 16);
30603 return __ret;
30604 }
30605 #else
vtbx4_u8(uint8x8_t __p0,uint8x8x4_t __p1,uint8x8_t __p2)30606 __ai uint8x8_t vtbx4_u8(uint8x8_t __p0, uint8x8x4_t __p1, uint8x8_t __p2) {
30607 uint8x8_t __ret;
30608 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
30609 uint8x8x4_t __rev1;
30610 __rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
30611 __rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
30612 __rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], 7, 6, 5, 4, 3, 2, 1, 0);
30613 __rev1.val[3] = __builtin_shufflevector(__p1.val[3], __p1.val[3], 7, 6, 5, 4, 3, 2, 1, 0);
30614 uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
30615 __ret = (uint8x8_t) __builtin_neon_vtbx4_v((int8x8_t)__rev0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], (int8x8_t)__rev2, 16);
30616 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
30617 return __ret;
30618 }
30619 #endif
30620
30621 #ifdef __LITTLE_ENDIAN__
vtbx4_s8(int8x8_t __p0,int8x8x4_t __p1,int8x8_t __p2)30622 __ai int8x8_t vtbx4_s8(int8x8_t __p0, int8x8x4_t __p1, int8x8_t __p2) {
30623 int8x8_t __ret;
30624 __ret = (int8x8_t) __builtin_neon_vtbx4_v((int8x8_t)__p0, (int8x8_t)__p1.val[0], (int8x8_t)__p1.val[1], (int8x8_t)__p1.val[2], (int8x8_t)__p1.val[3], (int8x8_t)__p2, 0);
30625 return __ret;
30626 }
30627 #else
vtbx4_s8(int8x8_t __p0,int8x8x4_t __p1,int8x8_t __p2)30628 __ai int8x8_t vtbx4_s8(int8x8_t __p0, int8x8x4_t __p1, int8x8_t __p2) {
30629 int8x8_t __ret;
30630 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
30631 int8x8x4_t __rev1;
30632 __rev1.val[0] = __builtin_shufflevector(__p1.val[0], __p1.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
30633 __rev1.val[1] = __builtin_shufflevector(__p1.val[1], __p1.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
30634 __rev1.val[2] = __builtin_shufflevector(__p1.val[2], __p1.val[2], 7, 6, 5, 4, 3, 2, 1, 0);
30635 __rev1.val[3] = __builtin_shufflevector(__p1.val[3], __p1.val[3], 7, 6, 5, 4, 3, 2, 1, 0);
30636 int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
30637 __ret = (int8x8_t) __builtin_neon_vtbx4_v((int8x8_t)__rev0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], (int8x8_t)__rev2, 0);
30638 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
30639 return __ret;
30640 }
30641 #endif
30642
30643 #ifdef __LITTLE_ENDIAN__
vtrn_p8(poly8x8_t __p0,poly8x8_t __p1)30644 __ai poly8x8x2_t vtrn_p8(poly8x8_t __p0, poly8x8_t __p1) {
30645 poly8x8x2_t __ret;
30646 __builtin_neon_vtrn_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 4);
30647 return __ret;
30648 }
30649 #else
vtrn_p8(poly8x8_t __p0,poly8x8_t __p1)30650 __ai poly8x8x2_t vtrn_p8(poly8x8_t __p0, poly8x8_t __p1) {
30651 poly8x8x2_t __ret;
30652 poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
30653 poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
30654 __builtin_neon_vtrn_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 4);
30655
30656 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
30657 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
30658 return __ret;
30659 }
30660 #endif
30661
30662 #ifdef __LITTLE_ENDIAN__
vtrn_p16(poly16x4_t __p0,poly16x4_t __p1)30663 __ai poly16x4x2_t vtrn_p16(poly16x4_t __p0, poly16x4_t __p1) {
30664 poly16x4x2_t __ret;
30665 __builtin_neon_vtrn_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 5);
30666 return __ret;
30667 }
30668 #else
vtrn_p16(poly16x4_t __p0,poly16x4_t __p1)30669 __ai poly16x4x2_t vtrn_p16(poly16x4_t __p0, poly16x4_t __p1) {
30670 poly16x4x2_t __ret;
30671 poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
30672 poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
30673 __builtin_neon_vtrn_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 5);
30674
30675 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
30676 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
30677 return __ret;
30678 }
30679 #endif
30680
30681 #ifdef __LITTLE_ENDIAN__
vtrnq_p8(poly8x16_t __p0,poly8x16_t __p1)30682 __ai poly8x16x2_t vtrnq_p8(poly8x16_t __p0, poly8x16_t __p1) {
30683 poly8x16x2_t __ret;
30684 __builtin_neon_vtrnq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 36);
30685 return __ret;
30686 }
30687 #else
vtrnq_p8(poly8x16_t __p0,poly8x16_t __p1)30688 __ai poly8x16x2_t vtrnq_p8(poly8x16_t __p0, poly8x16_t __p1) {
30689 poly8x16x2_t __ret;
30690 poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
30691 poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
30692 __builtin_neon_vtrnq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 36);
30693
30694 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
30695 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
30696 return __ret;
30697 }
30698 #endif
30699
30700 #ifdef __LITTLE_ENDIAN__
vtrnq_p16(poly16x8_t __p0,poly16x8_t __p1)30701 __ai poly16x8x2_t vtrnq_p16(poly16x8_t __p0, poly16x8_t __p1) {
30702 poly16x8x2_t __ret;
30703 __builtin_neon_vtrnq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 37);
30704 return __ret;
30705 }
30706 #else
vtrnq_p16(poly16x8_t __p0,poly16x8_t __p1)30707 __ai poly16x8x2_t vtrnq_p16(poly16x8_t __p0, poly16x8_t __p1) {
30708 poly16x8x2_t __ret;
30709 poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
30710 poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
30711 __builtin_neon_vtrnq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 37);
30712
30713 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
30714 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
30715 return __ret;
30716 }
30717 #endif
30718
30719 #ifdef __LITTLE_ENDIAN__
vtrnq_u8(uint8x16_t __p0,uint8x16_t __p1)30720 __ai uint8x16x2_t vtrnq_u8(uint8x16_t __p0, uint8x16_t __p1) {
30721 uint8x16x2_t __ret;
30722 __builtin_neon_vtrnq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 48);
30723 return __ret;
30724 }
30725 #else
vtrnq_u8(uint8x16_t __p0,uint8x16_t __p1)30726 __ai uint8x16x2_t vtrnq_u8(uint8x16_t __p0, uint8x16_t __p1) {
30727 uint8x16x2_t __ret;
30728 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
30729 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
30730 __builtin_neon_vtrnq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 48);
30731
30732 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
30733 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
30734 return __ret;
30735 }
30736 #endif
30737
30738 #ifdef __LITTLE_ENDIAN__
vtrnq_u32(uint32x4_t __p0,uint32x4_t __p1)30739 __ai uint32x4x2_t vtrnq_u32(uint32x4_t __p0, uint32x4_t __p1) {
30740 uint32x4x2_t __ret;
30741 __builtin_neon_vtrnq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 50);
30742 return __ret;
30743 }
30744 #else
vtrnq_u32(uint32x4_t __p0,uint32x4_t __p1)30745 __ai uint32x4x2_t vtrnq_u32(uint32x4_t __p0, uint32x4_t __p1) {
30746 uint32x4x2_t __ret;
30747 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
30748 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
30749 __builtin_neon_vtrnq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 50);
30750
30751 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
30752 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
30753 return __ret;
30754 }
30755 #endif
30756
30757 #ifdef __LITTLE_ENDIAN__
vtrnq_u16(uint16x8_t __p0,uint16x8_t __p1)30758 __ai uint16x8x2_t vtrnq_u16(uint16x8_t __p0, uint16x8_t __p1) {
30759 uint16x8x2_t __ret;
30760 __builtin_neon_vtrnq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 49);
30761 return __ret;
30762 }
30763 #else
vtrnq_u16(uint16x8_t __p0,uint16x8_t __p1)30764 __ai uint16x8x2_t vtrnq_u16(uint16x8_t __p0, uint16x8_t __p1) {
30765 uint16x8x2_t __ret;
30766 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
30767 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
30768 __builtin_neon_vtrnq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 49);
30769
30770 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
30771 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
30772 return __ret;
30773 }
30774 #endif
30775
30776 #ifdef __LITTLE_ENDIAN__
vtrnq_s8(int8x16_t __p0,int8x16_t __p1)30777 __ai int8x16x2_t vtrnq_s8(int8x16_t __p0, int8x16_t __p1) {
30778 int8x16x2_t __ret;
30779 __builtin_neon_vtrnq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 32);
30780 return __ret;
30781 }
30782 #else
vtrnq_s8(int8x16_t __p0,int8x16_t __p1)30783 __ai int8x16x2_t vtrnq_s8(int8x16_t __p0, int8x16_t __p1) {
30784 int8x16x2_t __ret;
30785 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
30786 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
30787 __builtin_neon_vtrnq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 32);
30788
30789 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
30790 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
30791 return __ret;
30792 }
30793 #endif
30794
30795 #ifdef __LITTLE_ENDIAN__
vtrnq_f32(float32x4_t __p0,float32x4_t __p1)30796 __ai float32x4x2_t vtrnq_f32(float32x4_t __p0, float32x4_t __p1) {
30797 float32x4x2_t __ret;
30798 __builtin_neon_vtrnq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 41);
30799 return __ret;
30800 }
30801 #else
vtrnq_f32(float32x4_t __p0,float32x4_t __p1)30802 __ai float32x4x2_t vtrnq_f32(float32x4_t __p0, float32x4_t __p1) {
30803 float32x4x2_t __ret;
30804 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
30805 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
30806 __builtin_neon_vtrnq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 41);
30807
30808 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
30809 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
30810 return __ret;
30811 }
30812 #endif
30813
30814 #ifdef __LITTLE_ENDIAN__
vtrnq_s32(int32x4_t __p0,int32x4_t __p1)30815 __ai int32x4x2_t vtrnq_s32(int32x4_t __p0, int32x4_t __p1) {
30816 int32x4x2_t __ret;
30817 __builtin_neon_vtrnq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 34);
30818 return __ret;
30819 }
30820 #else
vtrnq_s32(int32x4_t __p0,int32x4_t __p1)30821 __ai int32x4x2_t vtrnq_s32(int32x4_t __p0, int32x4_t __p1) {
30822 int32x4x2_t __ret;
30823 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
30824 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
30825 __builtin_neon_vtrnq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 34);
30826
30827 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
30828 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
30829 return __ret;
30830 }
30831 #endif
30832
30833 #ifdef __LITTLE_ENDIAN__
vtrnq_s16(int16x8_t __p0,int16x8_t __p1)30834 __ai int16x8x2_t vtrnq_s16(int16x8_t __p0, int16x8_t __p1) {
30835 int16x8x2_t __ret;
30836 __builtin_neon_vtrnq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 33);
30837 return __ret;
30838 }
30839 #else
vtrnq_s16(int16x8_t __p0,int16x8_t __p1)30840 __ai int16x8x2_t vtrnq_s16(int16x8_t __p0, int16x8_t __p1) {
30841 int16x8x2_t __ret;
30842 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
30843 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
30844 __builtin_neon_vtrnq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 33);
30845
30846 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
30847 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
30848 return __ret;
30849 }
30850 #endif
30851
30852 #ifdef __LITTLE_ENDIAN__
vtrn_u8(uint8x8_t __p0,uint8x8_t __p1)30853 __ai uint8x8x2_t vtrn_u8(uint8x8_t __p0, uint8x8_t __p1) {
30854 uint8x8x2_t __ret;
30855 __builtin_neon_vtrn_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 16);
30856 return __ret;
30857 }
30858 #else
vtrn_u8(uint8x8_t __p0,uint8x8_t __p1)30859 __ai uint8x8x2_t vtrn_u8(uint8x8_t __p0, uint8x8_t __p1) {
30860 uint8x8x2_t __ret;
30861 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
30862 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
30863 __builtin_neon_vtrn_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 16);
30864
30865 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
30866 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
30867 return __ret;
30868 }
30869 #endif
30870
30871 #ifdef __LITTLE_ENDIAN__
vtrn_u32(uint32x2_t __p0,uint32x2_t __p1)30872 __ai uint32x2x2_t vtrn_u32(uint32x2_t __p0, uint32x2_t __p1) {
30873 uint32x2x2_t __ret;
30874 __builtin_neon_vtrn_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 18);
30875 return __ret;
30876 }
30877 #else
vtrn_u32(uint32x2_t __p0,uint32x2_t __p1)30878 __ai uint32x2x2_t vtrn_u32(uint32x2_t __p0, uint32x2_t __p1) {
30879 uint32x2x2_t __ret;
30880 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
30881 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
30882 __builtin_neon_vtrn_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 18);
30883
30884 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0);
30885 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0);
30886 return __ret;
30887 }
30888 #endif
30889
30890 #ifdef __LITTLE_ENDIAN__
vtrn_u16(uint16x4_t __p0,uint16x4_t __p1)30891 __ai uint16x4x2_t vtrn_u16(uint16x4_t __p0, uint16x4_t __p1) {
30892 uint16x4x2_t __ret;
30893 __builtin_neon_vtrn_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 17);
30894 return __ret;
30895 }
30896 #else
vtrn_u16(uint16x4_t __p0,uint16x4_t __p1)30897 __ai uint16x4x2_t vtrn_u16(uint16x4_t __p0, uint16x4_t __p1) {
30898 uint16x4x2_t __ret;
30899 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
30900 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
30901 __builtin_neon_vtrn_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 17);
30902
30903 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
30904 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
30905 return __ret;
30906 }
30907 #endif
30908
30909 #ifdef __LITTLE_ENDIAN__
vtrn_s8(int8x8_t __p0,int8x8_t __p1)30910 __ai int8x8x2_t vtrn_s8(int8x8_t __p0, int8x8_t __p1) {
30911 int8x8x2_t __ret;
30912 __builtin_neon_vtrn_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 0);
30913 return __ret;
30914 }
30915 #else
vtrn_s8(int8x8_t __p0,int8x8_t __p1)30916 __ai int8x8x2_t vtrn_s8(int8x8_t __p0, int8x8_t __p1) {
30917 int8x8x2_t __ret;
30918 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
30919 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
30920 __builtin_neon_vtrn_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 0);
30921
30922 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
30923 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
30924 return __ret;
30925 }
30926 #endif
30927
30928 #ifdef __LITTLE_ENDIAN__
vtrn_f32(float32x2_t __p0,float32x2_t __p1)30929 __ai float32x2x2_t vtrn_f32(float32x2_t __p0, float32x2_t __p1) {
30930 float32x2x2_t __ret;
30931 __builtin_neon_vtrn_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 9);
30932 return __ret;
30933 }
30934 #else
vtrn_f32(float32x2_t __p0,float32x2_t __p1)30935 __ai float32x2x2_t vtrn_f32(float32x2_t __p0, float32x2_t __p1) {
30936 float32x2x2_t __ret;
30937 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
30938 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
30939 __builtin_neon_vtrn_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 9);
30940
30941 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0);
30942 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0);
30943 return __ret;
30944 }
30945 #endif
30946
30947 #ifdef __LITTLE_ENDIAN__
vtrn_s32(int32x2_t __p0,int32x2_t __p1)30948 __ai int32x2x2_t vtrn_s32(int32x2_t __p0, int32x2_t __p1) {
30949 int32x2x2_t __ret;
30950 __builtin_neon_vtrn_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 2);
30951 return __ret;
30952 }
30953 #else
vtrn_s32(int32x2_t __p0,int32x2_t __p1)30954 __ai int32x2x2_t vtrn_s32(int32x2_t __p0, int32x2_t __p1) {
30955 int32x2x2_t __ret;
30956 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
30957 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
30958 __builtin_neon_vtrn_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 2);
30959
30960 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0);
30961 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0);
30962 return __ret;
30963 }
30964 #endif
30965
30966 #ifdef __LITTLE_ENDIAN__
vtrn_s16(int16x4_t __p0,int16x4_t __p1)30967 __ai int16x4x2_t vtrn_s16(int16x4_t __p0, int16x4_t __p1) {
30968 int16x4x2_t __ret;
30969 __builtin_neon_vtrn_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 1);
30970 return __ret;
30971 }
30972 #else
vtrn_s16(int16x4_t __p0,int16x4_t __p1)30973 __ai int16x4x2_t vtrn_s16(int16x4_t __p0, int16x4_t __p1) {
30974 int16x4x2_t __ret;
30975 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
30976 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
30977 __builtin_neon_vtrn_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 1);
30978
30979 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
30980 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
30981 return __ret;
30982 }
30983 #endif
30984
30985 #ifdef __LITTLE_ENDIAN__
vtst_p8(poly8x8_t __p0,poly8x8_t __p1)30986 __ai uint8x8_t vtst_p8(poly8x8_t __p0, poly8x8_t __p1) {
30987 uint8x8_t __ret;
30988 __ret = (uint8x8_t) __builtin_neon_vtst_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
30989 return __ret;
30990 }
30991 #else
vtst_p8(poly8x8_t __p0,poly8x8_t __p1)30992 __ai uint8x8_t vtst_p8(poly8x8_t __p0, poly8x8_t __p1) {
30993 uint8x8_t __ret;
30994 poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
30995 poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
30996 __ret = (uint8x8_t) __builtin_neon_vtst_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
30997 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
30998 return __ret;
30999 }
31000 #endif
31001
31002 #ifdef __LITTLE_ENDIAN__
vtst_p16(poly16x4_t __p0,poly16x4_t __p1)31003 __ai uint16x4_t vtst_p16(poly16x4_t __p0, poly16x4_t __p1) {
31004 uint16x4_t __ret;
31005 __ret = (uint16x4_t) __builtin_neon_vtst_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
31006 return __ret;
31007 }
31008 #else
vtst_p16(poly16x4_t __p0,poly16x4_t __p1)31009 __ai uint16x4_t vtst_p16(poly16x4_t __p0, poly16x4_t __p1) {
31010 uint16x4_t __ret;
31011 poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
31012 poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
31013 __ret = (uint16x4_t) __builtin_neon_vtst_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
31014 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
31015 return __ret;
31016 }
31017 #endif
31018
31019 #ifdef __LITTLE_ENDIAN__
vtstq_p8(poly8x16_t __p0,poly8x16_t __p1)31020 __ai uint8x16_t vtstq_p8(poly8x16_t __p0, poly8x16_t __p1) {
31021 uint8x16_t __ret;
31022 __ret = (uint8x16_t) __builtin_neon_vtstq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
31023 return __ret;
31024 }
31025 #else
vtstq_p8(poly8x16_t __p0,poly8x16_t __p1)31026 __ai uint8x16_t vtstq_p8(poly8x16_t __p0, poly8x16_t __p1) {
31027 uint8x16_t __ret;
31028 poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
31029 poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
31030 __ret = (uint8x16_t) __builtin_neon_vtstq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
31031 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
31032 return __ret;
31033 }
31034 #endif
31035
31036 #ifdef __LITTLE_ENDIAN__
vtstq_p16(poly16x8_t __p0,poly16x8_t __p1)31037 __ai uint16x8_t vtstq_p16(poly16x8_t __p0, poly16x8_t __p1) {
31038 uint16x8_t __ret;
31039 __ret = (uint16x8_t) __builtin_neon_vtstq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
31040 return __ret;
31041 }
31042 #else
vtstq_p16(poly16x8_t __p0,poly16x8_t __p1)31043 __ai uint16x8_t vtstq_p16(poly16x8_t __p0, poly16x8_t __p1) {
31044 uint16x8_t __ret;
31045 poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
31046 poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
31047 __ret = (uint16x8_t) __builtin_neon_vtstq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
31048 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
31049 return __ret;
31050 }
31051 #endif
31052
31053 #ifdef __LITTLE_ENDIAN__
vtstq_u8(uint8x16_t __p0,uint8x16_t __p1)31054 __ai uint8x16_t vtstq_u8(uint8x16_t __p0, uint8x16_t __p1) {
31055 uint8x16_t __ret;
31056 __ret = (uint8x16_t) __builtin_neon_vtstq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
31057 return __ret;
31058 }
31059 #else
vtstq_u8(uint8x16_t __p0,uint8x16_t __p1)31060 __ai uint8x16_t vtstq_u8(uint8x16_t __p0, uint8x16_t __p1) {
31061 uint8x16_t __ret;
31062 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
31063 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
31064 __ret = (uint8x16_t) __builtin_neon_vtstq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
31065 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
31066 return __ret;
31067 }
31068 #endif
31069
31070 #ifdef __LITTLE_ENDIAN__
vtstq_u32(uint32x4_t __p0,uint32x4_t __p1)31071 __ai uint32x4_t vtstq_u32(uint32x4_t __p0, uint32x4_t __p1) {
31072 uint32x4_t __ret;
31073 __ret = (uint32x4_t) __builtin_neon_vtstq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
31074 return __ret;
31075 }
31076 #else
vtstq_u32(uint32x4_t __p0,uint32x4_t __p1)31077 __ai uint32x4_t vtstq_u32(uint32x4_t __p0, uint32x4_t __p1) {
31078 uint32x4_t __ret;
31079 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
31080 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
31081 __ret = (uint32x4_t) __builtin_neon_vtstq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
31082 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
31083 return __ret;
31084 }
31085 #endif
31086
31087 #ifdef __LITTLE_ENDIAN__
vtstq_u16(uint16x8_t __p0,uint16x8_t __p1)31088 __ai uint16x8_t vtstq_u16(uint16x8_t __p0, uint16x8_t __p1) {
31089 uint16x8_t __ret;
31090 __ret = (uint16x8_t) __builtin_neon_vtstq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
31091 return __ret;
31092 }
31093 #else
vtstq_u16(uint16x8_t __p0,uint16x8_t __p1)31094 __ai uint16x8_t vtstq_u16(uint16x8_t __p0, uint16x8_t __p1) {
31095 uint16x8_t __ret;
31096 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
31097 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
31098 __ret = (uint16x8_t) __builtin_neon_vtstq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
31099 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
31100 return __ret;
31101 }
31102 #endif
31103
31104 #ifdef __LITTLE_ENDIAN__
vtstq_s8(int8x16_t __p0,int8x16_t __p1)31105 __ai uint8x16_t vtstq_s8(int8x16_t __p0, int8x16_t __p1) {
31106 uint8x16_t __ret;
31107 __ret = (uint8x16_t) __builtin_neon_vtstq_v((int8x16_t)__p0, (int8x16_t)__p1, 48);
31108 return __ret;
31109 }
31110 #else
vtstq_s8(int8x16_t __p0,int8x16_t __p1)31111 __ai uint8x16_t vtstq_s8(int8x16_t __p0, int8x16_t __p1) {
31112 uint8x16_t __ret;
31113 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
31114 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
31115 __ret = (uint8x16_t) __builtin_neon_vtstq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 48);
31116 __ret = __builtin_shufflevector(__ret, __ret, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
31117 return __ret;
31118 }
31119 #endif
31120
31121 #ifdef __LITTLE_ENDIAN__
vtstq_s32(int32x4_t __p0,int32x4_t __p1)31122 __ai uint32x4_t vtstq_s32(int32x4_t __p0, int32x4_t __p1) {
31123 uint32x4_t __ret;
31124 __ret = (uint32x4_t) __builtin_neon_vtstq_v((int8x16_t)__p0, (int8x16_t)__p1, 50);
31125 return __ret;
31126 }
31127 #else
vtstq_s32(int32x4_t __p0,int32x4_t __p1)31128 __ai uint32x4_t vtstq_s32(int32x4_t __p0, int32x4_t __p1) {
31129 uint32x4_t __ret;
31130 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
31131 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
31132 __ret = (uint32x4_t) __builtin_neon_vtstq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 50);
31133 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
31134 return __ret;
31135 }
31136 #endif
31137
31138 #ifdef __LITTLE_ENDIAN__
vtstq_s16(int16x8_t __p0,int16x8_t __p1)31139 __ai uint16x8_t vtstq_s16(int16x8_t __p0, int16x8_t __p1) {
31140 uint16x8_t __ret;
31141 __ret = (uint16x8_t) __builtin_neon_vtstq_v((int8x16_t)__p0, (int8x16_t)__p1, 49);
31142 return __ret;
31143 }
31144 #else
vtstq_s16(int16x8_t __p0,int16x8_t __p1)31145 __ai uint16x8_t vtstq_s16(int16x8_t __p0, int16x8_t __p1) {
31146 uint16x8_t __ret;
31147 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
31148 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
31149 __ret = (uint16x8_t) __builtin_neon_vtstq_v((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
31150 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
31151 return __ret;
31152 }
31153 #endif
31154
31155 #ifdef __LITTLE_ENDIAN__
vtst_u8(uint8x8_t __p0,uint8x8_t __p1)31156 __ai uint8x8_t vtst_u8(uint8x8_t __p0, uint8x8_t __p1) {
31157 uint8x8_t __ret;
31158 __ret = (uint8x8_t) __builtin_neon_vtst_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
31159 return __ret;
31160 }
31161 #else
vtst_u8(uint8x8_t __p0,uint8x8_t __p1)31162 __ai uint8x8_t vtst_u8(uint8x8_t __p0, uint8x8_t __p1) {
31163 uint8x8_t __ret;
31164 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
31165 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
31166 __ret = (uint8x8_t) __builtin_neon_vtst_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
31167 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
31168 return __ret;
31169 }
31170 #endif
31171
31172 #ifdef __LITTLE_ENDIAN__
vtst_u32(uint32x2_t __p0,uint32x2_t __p1)31173 __ai uint32x2_t vtst_u32(uint32x2_t __p0, uint32x2_t __p1) {
31174 uint32x2_t __ret;
31175 __ret = (uint32x2_t) __builtin_neon_vtst_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
31176 return __ret;
31177 }
31178 #else
vtst_u32(uint32x2_t __p0,uint32x2_t __p1)31179 __ai uint32x2_t vtst_u32(uint32x2_t __p0, uint32x2_t __p1) {
31180 uint32x2_t __ret;
31181 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
31182 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
31183 __ret = (uint32x2_t) __builtin_neon_vtst_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
31184 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
31185 return __ret;
31186 }
31187 #endif
31188
31189 #ifdef __LITTLE_ENDIAN__
vtst_u16(uint16x4_t __p0,uint16x4_t __p1)31190 __ai uint16x4_t vtst_u16(uint16x4_t __p0, uint16x4_t __p1) {
31191 uint16x4_t __ret;
31192 __ret = (uint16x4_t) __builtin_neon_vtst_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
31193 return __ret;
31194 }
31195 #else
vtst_u16(uint16x4_t __p0,uint16x4_t __p1)31196 __ai uint16x4_t vtst_u16(uint16x4_t __p0, uint16x4_t __p1) {
31197 uint16x4_t __ret;
31198 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
31199 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
31200 __ret = (uint16x4_t) __builtin_neon_vtst_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
31201 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
31202 return __ret;
31203 }
31204 #endif
31205
31206 #ifdef __LITTLE_ENDIAN__
vtst_s8(int8x8_t __p0,int8x8_t __p1)31207 __ai uint8x8_t vtst_s8(int8x8_t __p0, int8x8_t __p1) {
31208 uint8x8_t __ret;
31209 __ret = (uint8x8_t) __builtin_neon_vtst_v((int8x8_t)__p0, (int8x8_t)__p1, 16);
31210 return __ret;
31211 }
31212 #else
vtst_s8(int8x8_t __p0,int8x8_t __p1)31213 __ai uint8x8_t vtst_s8(int8x8_t __p0, int8x8_t __p1) {
31214 uint8x8_t __ret;
31215 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
31216 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
31217 __ret = (uint8x8_t) __builtin_neon_vtst_v((int8x8_t)__rev0, (int8x8_t)__rev1, 16);
31218 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
31219 return __ret;
31220 }
31221 #endif
31222
31223 #ifdef __LITTLE_ENDIAN__
vtst_s32(int32x2_t __p0,int32x2_t __p1)31224 __ai uint32x2_t vtst_s32(int32x2_t __p0, int32x2_t __p1) {
31225 uint32x2_t __ret;
31226 __ret = (uint32x2_t) __builtin_neon_vtst_v((int8x8_t)__p0, (int8x8_t)__p1, 18);
31227 return __ret;
31228 }
31229 #else
vtst_s32(int32x2_t __p0,int32x2_t __p1)31230 __ai uint32x2_t vtst_s32(int32x2_t __p0, int32x2_t __p1) {
31231 uint32x2_t __ret;
31232 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
31233 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
31234 __ret = (uint32x2_t) __builtin_neon_vtst_v((int8x8_t)__rev0, (int8x8_t)__rev1, 18);
31235 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
31236 return __ret;
31237 }
31238 #endif
31239
31240 #ifdef __LITTLE_ENDIAN__
vtst_s16(int16x4_t __p0,int16x4_t __p1)31241 __ai uint16x4_t vtst_s16(int16x4_t __p0, int16x4_t __p1) {
31242 uint16x4_t __ret;
31243 __ret = (uint16x4_t) __builtin_neon_vtst_v((int8x8_t)__p0, (int8x8_t)__p1, 17);
31244 return __ret;
31245 }
31246 #else
vtst_s16(int16x4_t __p0,int16x4_t __p1)31247 __ai uint16x4_t vtst_s16(int16x4_t __p0, int16x4_t __p1) {
31248 uint16x4_t __ret;
31249 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
31250 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
31251 __ret = (uint16x4_t) __builtin_neon_vtst_v((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
31252 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
31253 return __ret;
31254 }
31255 #endif
31256
31257 #ifdef __LITTLE_ENDIAN__
vuzp_p8(poly8x8_t __p0,poly8x8_t __p1)31258 __ai poly8x8x2_t vuzp_p8(poly8x8_t __p0, poly8x8_t __p1) {
31259 poly8x8x2_t __ret;
31260 __builtin_neon_vuzp_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 4);
31261 return __ret;
31262 }
31263 #else
vuzp_p8(poly8x8_t __p0,poly8x8_t __p1)31264 __ai poly8x8x2_t vuzp_p8(poly8x8_t __p0, poly8x8_t __p1) {
31265 poly8x8x2_t __ret;
31266 poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
31267 poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
31268 __builtin_neon_vuzp_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 4);
31269
31270 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
31271 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
31272 return __ret;
31273 }
31274 #endif
31275
31276 #ifdef __LITTLE_ENDIAN__
vuzp_p16(poly16x4_t __p0,poly16x4_t __p1)31277 __ai poly16x4x2_t vuzp_p16(poly16x4_t __p0, poly16x4_t __p1) {
31278 poly16x4x2_t __ret;
31279 __builtin_neon_vuzp_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 5);
31280 return __ret;
31281 }
31282 #else
vuzp_p16(poly16x4_t __p0,poly16x4_t __p1)31283 __ai poly16x4x2_t vuzp_p16(poly16x4_t __p0, poly16x4_t __p1) {
31284 poly16x4x2_t __ret;
31285 poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
31286 poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
31287 __builtin_neon_vuzp_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 5);
31288
31289 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
31290 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
31291 return __ret;
31292 }
31293 #endif
31294
31295 #ifdef __LITTLE_ENDIAN__
vuzpq_p8(poly8x16_t __p0,poly8x16_t __p1)31296 __ai poly8x16x2_t vuzpq_p8(poly8x16_t __p0, poly8x16_t __p1) {
31297 poly8x16x2_t __ret;
31298 __builtin_neon_vuzpq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 36);
31299 return __ret;
31300 }
31301 #else
vuzpq_p8(poly8x16_t __p0,poly8x16_t __p1)31302 __ai poly8x16x2_t vuzpq_p8(poly8x16_t __p0, poly8x16_t __p1) {
31303 poly8x16x2_t __ret;
31304 poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
31305 poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
31306 __builtin_neon_vuzpq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 36);
31307
31308 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
31309 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
31310 return __ret;
31311 }
31312 #endif
31313
31314 #ifdef __LITTLE_ENDIAN__
vuzpq_p16(poly16x8_t __p0,poly16x8_t __p1)31315 __ai poly16x8x2_t vuzpq_p16(poly16x8_t __p0, poly16x8_t __p1) {
31316 poly16x8x2_t __ret;
31317 __builtin_neon_vuzpq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 37);
31318 return __ret;
31319 }
31320 #else
vuzpq_p16(poly16x8_t __p0,poly16x8_t __p1)31321 __ai poly16x8x2_t vuzpq_p16(poly16x8_t __p0, poly16x8_t __p1) {
31322 poly16x8x2_t __ret;
31323 poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
31324 poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
31325 __builtin_neon_vuzpq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 37);
31326
31327 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
31328 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
31329 return __ret;
31330 }
31331 #endif
31332
31333 #ifdef __LITTLE_ENDIAN__
vuzpq_u8(uint8x16_t __p0,uint8x16_t __p1)31334 __ai uint8x16x2_t vuzpq_u8(uint8x16_t __p0, uint8x16_t __p1) {
31335 uint8x16x2_t __ret;
31336 __builtin_neon_vuzpq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 48);
31337 return __ret;
31338 }
31339 #else
vuzpq_u8(uint8x16_t __p0,uint8x16_t __p1)31340 __ai uint8x16x2_t vuzpq_u8(uint8x16_t __p0, uint8x16_t __p1) {
31341 uint8x16x2_t __ret;
31342 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
31343 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
31344 __builtin_neon_vuzpq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 48);
31345
31346 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
31347 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
31348 return __ret;
31349 }
31350 #endif
31351
31352 #ifdef __LITTLE_ENDIAN__
vuzpq_u32(uint32x4_t __p0,uint32x4_t __p1)31353 __ai uint32x4x2_t vuzpq_u32(uint32x4_t __p0, uint32x4_t __p1) {
31354 uint32x4x2_t __ret;
31355 __builtin_neon_vuzpq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 50);
31356 return __ret;
31357 }
31358 #else
vuzpq_u32(uint32x4_t __p0,uint32x4_t __p1)31359 __ai uint32x4x2_t vuzpq_u32(uint32x4_t __p0, uint32x4_t __p1) {
31360 uint32x4x2_t __ret;
31361 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
31362 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
31363 __builtin_neon_vuzpq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 50);
31364
31365 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
31366 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
31367 return __ret;
31368 }
31369 #endif
31370
31371 #ifdef __LITTLE_ENDIAN__
vuzpq_u16(uint16x8_t __p0,uint16x8_t __p1)31372 __ai uint16x8x2_t vuzpq_u16(uint16x8_t __p0, uint16x8_t __p1) {
31373 uint16x8x2_t __ret;
31374 __builtin_neon_vuzpq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 49);
31375 return __ret;
31376 }
31377 #else
vuzpq_u16(uint16x8_t __p0,uint16x8_t __p1)31378 __ai uint16x8x2_t vuzpq_u16(uint16x8_t __p0, uint16x8_t __p1) {
31379 uint16x8x2_t __ret;
31380 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
31381 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
31382 __builtin_neon_vuzpq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 49);
31383
31384 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
31385 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
31386 return __ret;
31387 }
31388 #endif
31389
31390 #ifdef __LITTLE_ENDIAN__
vuzpq_s8(int8x16_t __p0,int8x16_t __p1)31391 __ai int8x16x2_t vuzpq_s8(int8x16_t __p0, int8x16_t __p1) {
31392 int8x16x2_t __ret;
31393 __builtin_neon_vuzpq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 32);
31394 return __ret;
31395 }
31396 #else
vuzpq_s8(int8x16_t __p0,int8x16_t __p1)31397 __ai int8x16x2_t vuzpq_s8(int8x16_t __p0, int8x16_t __p1) {
31398 int8x16x2_t __ret;
31399 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
31400 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
31401 __builtin_neon_vuzpq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 32);
31402
31403 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
31404 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
31405 return __ret;
31406 }
31407 #endif
31408
31409 #ifdef __LITTLE_ENDIAN__
vuzpq_f32(float32x4_t __p0,float32x4_t __p1)31410 __ai float32x4x2_t vuzpq_f32(float32x4_t __p0, float32x4_t __p1) {
31411 float32x4x2_t __ret;
31412 __builtin_neon_vuzpq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 41);
31413 return __ret;
31414 }
31415 #else
vuzpq_f32(float32x4_t __p0,float32x4_t __p1)31416 __ai float32x4x2_t vuzpq_f32(float32x4_t __p0, float32x4_t __p1) {
31417 float32x4x2_t __ret;
31418 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
31419 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
31420 __builtin_neon_vuzpq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 41);
31421
31422 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
31423 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
31424 return __ret;
31425 }
31426 #endif
31427
31428 #ifdef __LITTLE_ENDIAN__
vuzpq_s32(int32x4_t __p0,int32x4_t __p1)31429 __ai int32x4x2_t vuzpq_s32(int32x4_t __p0, int32x4_t __p1) {
31430 int32x4x2_t __ret;
31431 __builtin_neon_vuzpq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 34);
31432 return __ret;
31433 }
31434 #else
vuzpq_s32(int32x4_t __p0,int32x4_t __p1)31435 __ai int32x4x2_t vuzpq_s32(int32x4_t __p0, int32x4_t __p1) {
31436 int32x4x2_t __ret;
31437 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
31438 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
31439 __builtin_neon_vuzpq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 34);
31440
31441 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
31442 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
31443 return __ret;
31444 }
31445 #endif
31446
31447 #ifdef __LITTLE_ENDIAN__
vuzpq_s16(int16x8_t __p0,int16x8_t __p1)31448 __ai int16x8x2_t vuzpq_s16(int16x8_t __p0, int16x8_t __p1) {
31449 int16x8x2_t __ret;
31450 __builtin_neon_vuzpq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 33);
31451 return __ret;
31452 }
31453 #else
vuzpq_s16(int16x8_t __p0,int16x8_t __p1)31454 __ai int16x8x2_t vuzpq_s16(int16x8_t __p0, int16x8_t __p1) {
31455 int16x8x2_t __ret;
31456 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
31457 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
31458 __builtin_neon_vuzpq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 33);
31459
31460 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
31461 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
31462 return __ret;
31463 }
31464 #endif
31465
31466 #ifdef __LITTLE_ENDIAN__
vuzp_u8(uint8x8_t __p0,uint8x8_t __p1)31467 __ai uint8x8x2_t vuzp_u8(uint8x8_t __p0, uint8x8_t __p1) {
31468 uint8x8x2_t __ret;
31469 __builtin_neon_vuzp_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 16);
31470 return __ret;
31471 }
31472 #else
vuzp_u8(uint8x8_t __p0,uint8x8_t __p1)31473 __ai uint8x8x2_t vuzp_u8(uint8x8_t __p0, uint8x8_t __p1) {
31474 uint8x8x2_t __ret;
31475 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
31476 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
31477 __builtin_neon_vuzp_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 16);
31478
31479 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
31480 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
31481 return __ret;
31482 }
31483 #endif
31484
31485 #ifdef __LITTLE_ENDIAN__
vuzp_u32(uint32x2_t __p0,uint32x2_t __p1)31486 __ai uint32x2x2_t vuzp_u32(uint32x2_t __p0, uint32x2_t __p1) {
31487 uint32x2x2_t __ret;
31488 __builtin_neon_vuzp_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 18);
31489 return __ret;
31490 }
31491 #else
vuzp_u32(uint32x2_t __p0,uint32x2_t __p1)31492 __ai uint32x2x2_t vuzp_u32(uint32x2_t __p0, uint32x2_t __p1) {
31493 uint32x2x2_t __ret;
31494 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
31495 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
31496 __builtin_neon_vuzp_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 18);
31497
31498 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0);
31499 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0);
31500 return __ret;
31501 }
31502 #endif
31503
31504 #ifdef __LITTLE_ENDIAN__
vuzp_u16(uint16x4_t __p0,uint16x4_t __p1)31505 __ai uint16x4x2_t vuzp_u16(uint16x4_t __p0, uint16x4_t __p1) {
31506 uint16x4x2_t __ret;
31507 __builtin_neon_vuzp_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 17);
31508 return __ret;
31509 }
31510 #else
vuzp_u16(uint16x4_t __p0,uint16x4_t __p1)31511 __ai uint16x4x2_t vuzp_u16(uint16x4_t __p0, uint16x4_t __p1) {
31512 uint16x4x2_t __ret;
31513 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
31514 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
31515 __builtin_neon_vuzp_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 17);
31516
31517 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
31518 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
31519 return __ret;
31520 }
31521 #endif
31522
31523 #ifdef __LITTLE_ENDIAN__
vuzp_s8(int8x8_t __p0,int8x8_t __p1)31524 __ai int8x8x2_t vuzp_s8(int8x8_t __p0, int8x8_t __p1) {
31525 int8x8x2_t __ret;
31526 __builtin_neon_vuzp_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 0);
31527 return __ret;
31528 }
31529 #else
vuzp_s8(int8x8_t __p0,int8x8_t __p1)31530 __ai int8x8x2_t vuzp_s8(int8x8_t __p0, int8x8_t __p1) {
31531 int8x8x2_t __ret;
31532 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
31533 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
31534 __builtin_neon_vuzp_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 0);
31535
31536 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
31537 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
31538 return __ret;
31539 }
31540 #endif
31541
31542 #ifdef __LITTLE_ENDIAN__
vuzp_f32(float32x2_t __p0,float32x2_t __p1)31543 __ai float32x2x2_t vuzp_f32(float32x2_t __p0, float32x2_t __p1) {
31544 float32x2x2_t __ret;
31545 __builtin_neon_vuzp_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 9);
31546 return __ret;
31547 }
31548 #else
vuzp_f32(float32x2_t __p0,float32x2_t __p1)31549 __ai float32x2x2_t vuzp_f32(float32x2_t __p0, float32x2_t __p1) {
31550 float32x2x2_t __ret;
31551 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
31552 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
31553 __builtin_neon_vuzp_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 9);
31554
31555 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0);
31556 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0);
31557 return __ret;
31558 }
31559 #endif
31560
31561 #ifdef __LITTLE_ENDIAN__
vuzp_s32(int32x2_t __p0,int32x2_t __p1)31562 __ai int32x2x2_t vuzp_s32(int32x2_t __p0, int32x2_t __p1) {
31563 int32x2x2_t __ret;
31564 __builtin_neon_vuzp_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 2);
31565 return __ret;
31566 }
31567 #else
vuzp_s32(int32x2_t __p0,int32x2_t __p1)31568 __ai int32x2x2_t vuzp_s32(int32x2_t __p0, int32x2_t __p1) {
31569 int32x2x2_t __ret;
31570 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
31571 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
31572 __builtin_neon_vuzp_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 2);
31573
31574 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0);
31575 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0);
31576 return __ret;
31577 }
31578 #endif
31579
31580 #ifdef __LITTLE_ENDIAN__
vuzp_s16(int16x4_t __p0,int16x4_t __p1)31581 __ai int16x4x2_t vuzp_s16(int16x4_t __p0, int16x4_t __p1) {
31582 int16x4x2_t __ret;
31583 __builtin_neon_vuzp_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 1);
31584 return __ret;
31585 }
31586 #else
vuzp_s16(int16x4_t __p0,int16x4_t __p1)31587 __ai int16x4x2_t vuzp_s16(int16x4_t __p0, int16x4_t __p1) {
31588 int16x4x2_t __ret;
31589 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
31590 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
31591 __builtin_neon_vuzp_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 1);
31592
31593 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
31594 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
31595 return __ret;
31596 }
31597 #endif
31598
31599 #ifdef __LITTLE_ENDIAN__
vzip_p8(poly8x8_t __p0,poly8x8_t __p1)31600 __ai poly8x8x2_t vzip_p8(poly8x8_t __p0, poly8x8_t __p1) {
31601 poly8x8x2_t __ret;
31602 __builtin_neon_vzip_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 4);
31603 return __ret;
31604 }
31605 #else
vzip_p8(poly8x8_t __p0,poly8x8_t __p1)31606 __ai poly8x8x2_t vzip_p8(poly8x8_t __p0, poly8x8_t __p1) {
31607 poly8x8x2_t __ret;
31608 poly8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
31609 poly8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
31610 __builtin_neon_vzip_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 4);
31611
31612 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
31613 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
31614 return __ret;
31615 }
31616 #endif
31617
31618 #ifdef __LITTLE_ENDIAN__
vzip_p16(poly16x4_t __p0,poly16x4_t __p1)31619 __ai poly16x4x2_t vzip_p16(poly16x4_t __p0, poly16x4_t __p1) {
31620 poly16x4x2_t __ret;
31621 __builtin_neon_vzip_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 5);
31622 return __ret;
31623 }
31624 #else
vzip_p16(poly16x4_t __p0,poly16x4_t __p1)31625 __ai poly16x4x2_t vzip_p16(poly16x4_t __p0, poly16x4_t __p1) {
31626 poly16x4x2_t __ret;
31627 poly16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
31628 poly16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
31629 __builtin_neon_vzip_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 5);
31630
31631 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
31632 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
31633 return __ret;
31634 }
31635 #endif
31636
31637 #ifdef __LITTLE_ENDIAN__
vzipq_p8(poly8x16_t __p0,poly8x16_t __p1)31638 __ai poly8x16x2_t vzipq_p8(poly8x16_t __p0, poly8x16_t __p1) {
31639 poly8x16x2_t __ret;
31640 __builtin_neon_vzipq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 36);
31641 return __ret;
31642 }
31643 #else
vzipq_p8(poly8x16_t __p0,poly8x16_t __p1)31644 __ai poly8x16x2_t vzipq_p8(poly8x16_t __p0, poly8x16_t __p1) {
31645 poly8x16x2_t __ret;
31646 poly8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
31647 poly8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
31648 __builtin_neon_vzipq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 36);
31649
31650 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
31651 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
31652 return __ret;
31653 }
31654 #endif
31655
31656 #ifdef __LITTLE_ENDIAN__
vzipq_p16(poly16x8_t __p0,poly16x8_t __p1)31657 __ai poly16x8x2_t vzipq_p16(poly16x8_t __p0, poly16x8_t __p1) {
31658 poly16x8x2_t __ret;
31659 __builtin_neon_vzipq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 37);
31660 return __ret;
31661 }
31662 #else
vzipq_p16(poly16x8_t __p0,poly16x8_t __p1)31663 __ai poly16x8x2_t vzipq_p16(poly16x8_t __p0, poly16x8_t __p1) {
31664 poly16x8x2_t __ret;
31665 poly16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
31666 poly16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
31667 __builtin_neon_vzipq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 37);
31668
31669 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
31670 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
31671 return __ret;
31672 }
31673 #endif
31674
31675 #ifdef __LITTLE_ENDIAN__
vzipq_u8(uint8x16_t __p0,uint8x16_t __p1)31676 __ai uint8x16x2_t vzipq_u8(uint8x16_t __p0, uint8x16_t __p1) {
31677 uint8x16x2_t __ret;
31678 __builtin_neon_vzipq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 48);
31679 return __ret;
31680 }
31681 #else
vzipq_u8(uint8x16_t __p0,uint8x16_t __p1)31682 __ai uint8x16x2_t vzipq_u8(uint8x16_t __p0, uint8x16_t __p1) {
31683 uint8x16x2_t __ret;
31684 uint8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
31685 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
31686 __builtin_neon_vzipq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 48);
31687
31688 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
31689 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
31690 return __ret;
31691 }
31692 #endif
31693
31694 #ifdef __LITTLE_ENDIAN__
vzipq_u32(uint32x4_t __p0,uint32x4_t __p1)31695 __ai uint32x4x2_t vzipq_u32(uint32x4_t __p0, uint32x4_t __p1) {
31696 uint32x4x2_t __ret;
31697 __builtin_neon_vzipq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 50);
31698 return __ret;
31699 }
31700 #else
vzipq_u32(uint32x4_t __p0,uint32x4_t __p1)31701 __ai uint32x4x2_t vzipq_u32(uint32x4_t __p0, uint32x4_t __p1) {
31702 uint32x4x2_t __ret;
31703 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
31704 uint32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
31705 __builtin_neon_vzipq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 50);
31706
31707 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
31708 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
31709 return __ret;
31710 }
31711 #endif
31712
31713 #ifdef __LITTLE_ENDIAN__
vzipq_u16(uint16x8_t __p0,uint16x8_t __p1)31714 __ai uint16x8x2_t vzipq_u16(uint16x8_t __p0, uint16x8_t __p1) {
31715 uint16x8x2_t __ret;
31716 __builtin_neon_vzipq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 49);
31717 return __ret;
31718 }
31719 #else
vzipq_u16(uint16x8_t __p0,uint16x8_t __p1)31720 __ai uint16x8x2_t vzipq_u16(uint16x8_t __p0, uint16x8_t __p1) {
31721 uint16x8x2_t __ret;
31722 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
31723 uint16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
31724 __builtin_neon_vzipq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 49);
31725
31726 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
31727 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
31728 return __ret;
31729 }
31730 #endif
31731
31732 #ifdef __LITTLE_ENDIAN__
vzipq_s8(int8x16_t __p0,int8x16_t __p1)31733 __ai int8x16x2_t vzipq_s8(int8x16_t __p0, int8x16_t __p1) {
31734 int8x16x2_t __ret;
31735 __builtin_neon_vzipq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 32);
31736 return __ret;
31737 }
31738 #else
vzipq_s8(int8x16_t __p0,int8x16_t __p1)31739 __ai int8x16x2_t vzipq_s8(int8x16_t __p0, int8x16_t __p1) {
31740 int8x16x2_t __ret;
31741 int8x16_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
31742 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
31743 __builtin_neon_vzipq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 32);
31744
31745 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
31746 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
31747 return __ret;
31748 }
31749 #endif
31750
31751 #ifdef __LITTLE_ENDIAN__
vzipq_f32(float32x4_t __p0,float32x4_t __p1)31752 __ai float32x4x2_t vzipq_f32(float32x4_t __p0, float32x4_t __p1) {
31753 float32x4x2_t __ret;
31754 __builtin_neon_vzipq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 41);
31755 return __ret;
31756 }
31757 #else
vzipq_f32(float32x4_t __p0,float32x4_t __p1)31758 __ai float32x4x2_t vzipq_f32(float32x4_t __p0, float32x4_t __p1) {
31759 float32x4x2_t __ret;
31760 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
31761 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
31762 __builtin_neon_vzipq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 41);
31763
31764 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
31765 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
31766 return __ret;
31767 }
31768 #endif
31769
31770 #ifdef __LITTLE_ENDIAN__
vzipq_s32(int32x4_t __p0,int32x4_t __p1)31771 __ai int32x4x2_t vzipq_s32(int32x4_t __p0, int32x4_t __p1) {
31772 int32x4x2_t __ret;
31773 __builtin_neon_vzipq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 34);
31774 return __ret;
31775 }
31776 #else
vzipq_s32(int32x4_t __p0,int32x4_t __p1)31777 __ai int32x4x2_t vzipq_s32(int32x4_t __p0, int32x4_t __p1) {
31778 int32x4x2_t __ret;
31779 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
31780 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
31781 __builtin_neon_vzipq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 34);
31782
31783 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
31784 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
31785 return __ret;
31786 }
31787 #endif
31788
31789 #ifdef __LITTLE_ENDIAN__
vzipq_s16(int16x8_t __p0,int16x8_t __p1)31790 __ai int16x8x2_t vzipq_s16(int16x8_t __p0, int16x8_t __p1) {
31791 int16x8x2_t __ret;
31792 __builtin_neon_vzipq_v(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 33);
31793 return __ret;
31794 }
31795 #else
vzipq_s16(int16x8_t __p0,int16x8_t __p1)31796 __ai int16x8x2_t vzipq_s16(int16x8_t __p0, int16x8_t __p1) {
31797 int16x8x2_t __ret;
31798 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
31799 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
31800 __builtin_neon_vzipq_v(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 33);
31801
31802 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
31803 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
31804 return __ret;
31805 }
31806 #endif
31807
31808 #ifdef __LITTLE_ENDIAN__
vzip_u8(uint8x8_t __p0,uint8x8_t __p1)31809 __ai uint8x8x2_t vzip_u8(uint8x8_t __p0, uint8x8_t __p1) {
31810 uint8x8x2_t __ret;
31811 __builtin_neon_vzip_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 16);
31812 return __ret;
31813 }
31814 #else
vzip_u8(uint8x8_t __p0,uint8x8_t __p1)31815 __ai uint8x8x2_t vzip_u8(uint8x8_t __p0, uint8x8_t __p1) {
31816 uint8x8x2_t __ret;
31817 uint8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
31818 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
31819 __builtin_neon_vzip_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 16);
31820
31821 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
31822 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
31823 return __ret;
31824 }
31825 #endif
31826
31827 #ifdef __LITTLE_ENDIAN__
vzip_u32(uint32x2_t __p0,uint32x2_t __p1)31828 __ai uint32x2x2_t vzip_u32(uint32x2_t __p0, uint32x2_t __p1) {
31829 uint32x2x2_t __ret;
31830 __builtin_neon_vzip_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 18);
31831 return __ret;
31832 }
31833 #else
vzip_u32(uint32x2_t __p0,uint32x2_t __p1)31834 __ai uint32x2x2_t vzip_u32(uint32x2_t __p0, uint32x2_t __p1) {
31835 uint32x2x2_t __ret;
31836 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
31837 uint32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
31838 __builtin_neon_vzip_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 18);
31839
31840 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0);
31841 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0);
31842 return __ret;
31843 }
31844 #endif
31845
31846 #ifdef __LITTLE_ENDIAN__
vzip_u16(uint16x4_t __p0,uint16x4_t __p1)31847 __ai uint16x4x2_t vzip_u16(uint16x4_t __p0, uint16x4_t __p1) {
31848 uint16x4x2_t __ret;
31849 __builtin_neon_vzip_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 17);
31850 return __ret;
31851 }
31852 #else
vzip_u16(uint16x4_t __p0,uint16x4_t __p1)31853 __ai uint16x4x2_t vzip_u16(uint16x4_t __p0, uint16x4_t __p1) {
31854 uint16x4x2_t __ret;
31855 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
31856 uint16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
31857 __builtin_neon_vzip_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 17);
31858
31859 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
31860 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
31861 return __ret;
31862 }
31863 #endif
31864
31865 #ifdef __LITTLE_ENDIAN__
vzip_s8(int8x8_t __p0,int8x8_t __p1)31866 __ai int8x8x2_t vzip_s8(int8x8_t __p0, int8x8_t __p1) {
31867 int8x8x2_t __ret;
31868 __builtin_neon_vzip_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 0);
31869 return __ret;
31870 }
31871 #else
vzip_s8(int8x8_t __p0,int8x8_t __p1)31872 __ai int8x8x2_t vzip_s8(int8x8_t __p0, int8x8_t __p1) {
31873 int8x8x2_t __ret;
31874 int8x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
31875 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
31876 __builtin_neon_vzip_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 0);
31877
31878 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
31879 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
31880 return __ret;
31881 }
31882 #endif
31883
31884 #ifdef __LITTLE_ENDIAN__
vzip_f32(float32x2_t __p0,float32x2_t __p1)31885 __ai float32x2x2_t vzip_f32(float32x2_t __p0, float32x2_t __p1) {
31886 float32x2x2_t __ret;
31887 __builtin_neon_vzip_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 9);
31888 return __ret;
31889 }
31890 #else
vzip_f32(float32x2_t __p0,float32x2_t __p1)31891 __ai float32x2x2_t vzip_f32(float32x2_t __p0, float32x2_t __p1) {
31892 float32x2x2_t __ret;
31893 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
31894 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
31895 __builtin_neon_vzip_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 9);
31896
31897 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0);
31898 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0);
31899 return __ret;
31900 }
31901 #endif
31902
31903 #ifdef __LITTLE_ENDIAN__
vzip_s32(int32x2_t __p0,int32x2_t __p1)31904 __ai int32x2x2_t vzip_s32(int32x2_t __p0, int32x2_t __p1) {
31905 int32x2x2_t __ret;
31906 __builtin_neon_vzip_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 2);
31907 return __ret;
31908 }
31909 #else
vzip_s32(int32x2_t __p0,int32x2_t __p1)31910 __ai int32x2x2_t vzip_s32(int32x2_t __p0, int32x2_t __p1) {
31911 int32x2x2_t __ret;
31912 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
31913 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
31914 __builtin_neon_vzip_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 2);
31915
31916 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 1, 0);
31917 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 1, 0);
31918 return __ret;
31919 }
31920 #endif
31921
31922 #ifdef __LITTLE_ENDIAN__
vzip_s16(int16x4_t __p0,int16x4_t __p1)31923 __ai int16x4x2_t vzip_s16(int16x4_t __p0, int16x4_t __p1) {
31924 int16x4x2_t __ret;
31925 __builtin_neon_vzip_v(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 1);
31926 return __ret;
31927 }
31928 #else
vzip_s16(int16x4_t __p0,int16x4_t __p1)31929 __ai int16x4x2_t vzip_s16(int16x4_t __p0, int16x4_t __p1) {
31930 int16x4x2_t __ret;
31931 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
31932 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
31933 __builtin_neon_vzip_v(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 1);
31934
31935 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
31936 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
31937 return __ret;
31938 }
31939 #endif
31940
31941 #ifdef __LITTLE_ENDIAN__
31942 #define splatq_lane_bf16(__p0, __p1) __extension__ ({ \
31943 bfloat16x8_t __ret; \
31944 bfloat16x4_t __s0 = __p0; \
31945 __ret = (bfloat16x8_t) __builtin_neon_splatq_lane_bf16((int8x8_t)__s0, __p1, 11); \
31946 __ret; \
31947 })
31948 #else
31949 #define splatq_lane_bf16(__p0, __p1) __extension__ ({ \
31950 bfloat16x8_t __ret; \
31951 bfloat16x4_t __s0 = __p0; \
31952 bfloat16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
31953 __ret = (bfloat16x8_t) __builtin_neon_splatq_lane_bf16((int8x8_t)__rev0, __p1, 11); \
31954 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
31955 __ret; \
31956 })
31957 #define __noswap_splatq_lane_bf16(__p0, __p1) __extension__ ({ \
31958 bfloat16x8_t __ret; \
31959 bfloat16x4_t __s0 = __p0; \
31960 __ret = (bfloat16x8_t) __builtin_neon_splatq_lane_bf16((int8x8_t)__s0, __p1, 11); \
31961 __ret; \
31962 })
31963 #endif
31964
31965 #ifdef __LITTLE_ENDIAN__
31966 #define splat_lane_bf16(__p0, __p1) __extension__ ({ \
31967 bfloat16x4_t __ret; \
31968 bfloat16x4_t __s0 = __p0; \
31969 __ret = (bfloat16x4_t) __builtin_neon_splat_lane_bf16((int8x8_t)__s0, __p1, 11); \
31970 __ret; \
31971 })
31972 #else
31973 #define splat_lane_bf16(__p0, __p1) __extension__ ({ \
31974 bfloat16x4_t __ret; \
31975 bfloat16x4_t __s0 = __p0; \
31976 bfloat16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
31977 __ret = (bfloat16x4_t) __builtin_neon_splat_lane_bf16((int8x8_t)__rev0, __p1, 11); \
31978 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
31979 __ret; \
31980 })
31981 #define __noswap_splat_lane_bf16(__p0, __p1) __extension__ ({ \
31982 bfloat16x4_t __ret; \
31983 bfloat16x4_t __s0 = __p0; \
31984 __ret = (bfloat16x4_t) __builtin_neon_splat_lane_bf16((int8x8_t)__s0, __p1, 11); \
31985 __ret; \
31986 })
31987 #endif
31988
31989 #ifdef __LITTLE_ENDIAN__
31990 #define splatq_laneq_bf16(__p0, __p1) __extension__ ({ \
31991 bfloat16x8_t __ret; \
31992 bfloat16x8_t __s0 = __p0; \
31993 __ret = (bfloat16x8_t) __builtin_neon_splatq_laneq_bf16((int8x16_t)__s0, __p1, 43); \
31994 __ret; \
31995 })
31996 #else
31997 #define splatq_laneq_bf16(__p0, __p1) __extension__ ({ \
31998 bfloat16x8_t __ret; \
31999 bfloat16x8_t __s0 = __p0; \
32000 bfloat16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
32001 __ret = (bfloat16x8_t) __builtin_neon_splatq_laneq_bf16((int8x16_t)__rev0, __p1, 43); \
32002 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
32003 __ret; \
32004 })
32005 #define __noswap_splatq_laneq_bf16(__p0, __p1) __extension__ ({ \
32006 bfloat16x8_t __ret; \
32007 bfloat16x8_t __s0 = __p0; \
32008 __ret = (bfloat16x8_t) __builtin_neon_splatq_laneq_bf16((int8x16_t)__s0, __p1, 43); \
32009 __ret; \
32010 })
32011 #endif
32012
32013 #ifdef __LITTLE_ENDIAN__
32014 #define splat_laneq_bf16(__p0, __p1) __extension__ ({ \
32015 bfloat16x4_t __ret; \
32016 bfloat16x8_t __s0 = __p0; \
32017 __ret = (bfloat16x4_t) __builtin_neon_splat_laneq_bf16((int8x16_t)__s0, __p1, 43); \
32018 __ret; \
32019 })
32020 #else
32021 #define splat_laneq_bf16(__p0, __p1) __extension__ ({ \
32022 bfloat16x4_t __ret; \
32023 bfloat16x8_t __s0 = __p0; \
32024 bfloat16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
32025 __ret = (bfloat16x4_t) __builtin_neon_splat_laneq_bf16((int8x16_t)__rev0, __p1, 43); \
32026 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
32027 __ret; \
32028 })
32029 #define __noswap_splat_laneq_bf16(__p0, __p1) __extension__ ({ \
32030 bfloat16x4_t __ret; \
32031 bfloat16x8_t __s0 = __p0; \
32032 __ret = (bfloat16x4_t) __builtin_neon_splat_laneq_bf16((int8x16_t)__s0, __p1, 43); \
32033 __ret; \
32034 })
32035 #endif
32036
32037 #ifdef __LITTLE_ENDIAN__
vbfdotq_f32(float32x4_t __p0,bfloat16x8_t __p1,bfloat16x8_t __p2)32038 __ai __attribute__((target("bf16"))) float32x4_t vbfdotq_f32(float32x4_t __p0, bfloat16x8_t __p1, bfloat16x8_t __p2) {
32039 float32x4_t __ret;
32040 __ret = (float32x4_t) __builtin_neon_vbfdotq_f32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
32041 return __ret;
32042 }
32043 #else
vbfdotq_f32(float32x4_t __p0,bfloat16x8_t __p1,bfloat16x8_t __p2)32044 __ai __attribute__((target("bf16"))) float32x4_t vbfdotq_f32(float32x4_t __p0, bfloat16x8_t __p1, bfloat16x8_t __p2) {
32045 float32x4_t __ret;
32046 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
32047 bfloat16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
32048 bfloat16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
32049 __ret = (float32x4_t) __builtin_neon_vbfdotq_f32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 41);
32050 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
32051 return __ret;
32052 }
__noswap_vbfdotq_f32(float32x4_t __p0,bfloat16x8_t __p1,bfloat16x8_t __p2)32053 __ai __attribute__((target("bf16"))) float32x4_t __noswap_vbfdotq_f32(float32x4_t __p0, bfloat16x8_t __p1, bfloat16x8_t __p2) {
32054 float32x4_t __ret;
32055 __ret = (float32x4_t) __builtin_neon_vbfdotq_f32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
32056 return __ret;
32057 }
32058 #endif
32059
32060 #ifdef __LITTLE_ENDIAN__
vbfdot_f32(float32x2_t __p0,bfloat16x4_t __p1,bfloat16x4_t __p2)32061 __ai __attribute__((target("bf16"))) float32x2_t vbfdot_f32(float32x2_t __p0, bfloat16x4_t __p1, bfloat16x4_t __p2) {
32062 float32x2_t __ret;
32063 __ret = (float32x2_t) __builtin_neon_vbfdot_f32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 9);
32064 return __ret;
32065 }
32066 #else
vbfdot_f32(float32x2_t __p0,bfloat16x4_t __p1,bfloat16x4_t __p2)32067 __ai __attribute__((target("bf16"))) float32x2_t vbfdot_f32(float32x2_t __p0, bfloat16x4_t __p1, bfloat16x4_t __p2) {
32068 float32x2_t __ret;
32069 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
32070 bfloat16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
32071 bfloat16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
32072 __ret = (float32x2_t) __builtin_neon_vbfdot_f32((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 9);
32073 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
32074 return __ret;
32075 }
__noswap_vbfdot_f32(float32x2_t __p0,bfloat16x4_t __p1,bfloat16x4_t __p2)32076 __ai __attribute__((target("bf16"))) float32x2_t __noswap_vbfdot_f32(float32x2_t __p0, bfloat16x4_t __p1, bfloat16x4_t __p2) {
32077 float32x2_t __ret;
32078 __ret = (float32x2_t) __builtin_neon_vbfdot_f32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 9);
32079 return __ret;
32080 }
32081 #endif
32082
32083 #ifdef __LITTLE_ENDIAN__
32084 #define vbfdotq_lane_f32(__p0_126, __p1_126, __p2_126, __p3_126) __extension__ ({ \
32085 float32x4_t __ret_126; \
32086 float32x4_t __s0_126 = __p0_126; \
32087 bfloat16x8_t __s1_126 = __p1_126; \
32088 bfloat16x4_t __s2_126 = __p2_126; \
32089 bfloat16x4_t __reint_126 = __s2_126; \
32090 float32x4_t __reint1_126 = splatq_lane_f32(*(float32x2_t *) &__reint_126, __p3_126); \
32091 __ret_126 = vbfdotq_f32(__s0_126, __s1_126, *(bfloat16x8_t *) &__reint1_126); \
32092 __ret_126; \
32093 })
32094 #else
32095 #define vbfdotq_lane_f32(__p0_127, __p1_127, __p2_127, __p3_127) __extension__ ({ \
32096 float32x4_t __ret_127; \
32097 float32x4_t __s0_127 = __p0_127; \
32098 bfloat16x8_t __s1_127 = __p1_127; \
32099 bfloat16x4_t __s2_127 = __p2_127; \
32100 float32x4_t __rev0_127; __rev0_127 = __builtin_shufflevector(__s0_127, __s0_127, 3, 2, 1, 0); \
32101 bfloat16x8_t __rev1_127; __rev1_127 = __builtin_shufflevector(__s1_127, __s1_127, 7, 6, 5, 4, 3, 2, 1, 0); \
32102 bfloat16x4_t __rev2_127; __rev2_127 = __builtin_shufflevector(__s2_127, __s2_127, 3, 2, 1, 0); \
32103 bfloat16x4_t __reint_127 = __rev2_127; \
32104 float32x4_t __reint1_127 = __noswap_splatq_lane_f32(*(float32x2_t *) &__reint_127, __p3_127); \
32105 __ret_127 = __noswap_vbfdotq_f32(__rev0_127, __rev1_127, *(bfloat16x8_t *) &__reint1_127); \
32106 __ret_127 = __builtin_shufflevector(__ret_127, __ret_127, 3, 2, 1, 0); \
32107 __ret_127; \
32108 })
32109 #endif
32110
32111 #ifdef __LITTLE_ENDIAN__
32112 #define vbfdot_lane_f32(__p0_128, __p1_128, __p2_128, __p3_128) __extension__ ({ \
32113 float32x2_t __ret_128; \
32114 float32x2_t __s0_128 = __p0_128; \
32115 bfloat16x4_t __s1_128 = __p1_128; \
32116 bfloat16x4_t __s2_128 = __p2_128; \
32117 bfloat16x4_t __reint_128 = __s2_128; \
32118 float32x2_t __reint1_128 = splat_lane_f32(*(float32x2_t *) &__reint_128, __p3_128); \
32119 __ret_128 = vbfdot_f32(__s0_128, __s1_128, *(bfloat16x4_t *) &__reint1_128); \
32120 __ret_128; \
32121 })
32122 #else
32123 #define vbfdot_lane_f32(__p0_129, __p1_129, __p2_129, __p3_129) __extension__ ({ \
32124 float32x2_t __ret_129; \
32125 float32x2_t __s0_129 = __p0_129; \
32126 bfloat16x4_t __s1_129 = __p1_129; \
32127 bfloat16x4_t __s2_129 = __p2_129; \
32128 float32x2_t __rev0_129; __rev0_129 = __builtin_shufflevector(__s0_129, __s0_129, 1, 0); \
32129 bfloat16x4_t __rev1_129; __rev1_129 = __builtin_shufflevector(__s1_129, __s1_129, 3, 2, 1, 0); \
32130 bfloat16x4_t __rev2_129; __rev2_129 = __builtin_shufflevector(__s2_129, __s2_129, 3, 2, 1, 0); \
32131 bfloat16x4_t __reint_129 = __rev2_129; \
32132 float32x2_t __reint1_129 = __noswap_splat_lane_f32(*(float32x2_t *) &__reint_129, __p3_129); \
32133 __ret_129 = __noswap_vbfdot_f32(__rev0_129, __rev1_129, *(bfloat16x4_t *) &__reint1_129); \
32134 __ret_129 = __builtin_shufflevector(__ret_129, __ret_129, 1, 0); \
32135 __ret_129; \
32136 })
32137 #endif
32138
32139 #ifdef __LITTLE_ENDIAN__
32140 #define vbfdotq_laneq_f32(__p0_130, __p1_130, __p2_130, __p3_130) __extension__ ({ \
32141 float32x4_t __ret_130; \
32142 float32x4_t __s0_130 = __p0_130; \
32143 bfloat16x8_t __s1_130 = __p1_130; \
32144 bfloat16x8_t __s2_130 = __p2_130; \
32145 bfloat16x8_t __reint_130 = __s2_130; \
32146 float32x4_t __reint1_130 = splatq_laneq_f32(*(float32x4_t *) &__reint_130, __p3_130); \
32147 __ret_130 = vbfdotq_f32(__s0_130, __s1_130, *(bfloat16x8_t *) &__reint1_130); \
32148 __ret_130; \
32149 })
32150 #else
32151 #define vbfdotq_laneq_f32(__p0_131, __p1_131, __p2_131, __p3_131) __extension__ ({ \
32152 float32x4_t __ret_131; \
32153 float32x4_t __s0_131 = __p0_131; \
32154 bfloat16x8_t __s1_131 = __p1_131; \
32155 bfloat16x8_t __s2_131 = __p2_131; \
32156 float32x4_t __rev0_131; __rev0_131 = __builtin_shufflevector(__s0_131, __s0_131, 3, 2, 1, 0); \
32157 bfloat16x8_t __rev1_131; __rev1_131 = __builtin_shufflevector(__s1_131, __s1_131, 7, 6, 5, 4, 3, 2, 1, 0); \
32158 bfloat16x8_t __rev2_131; __rev2_131 = __builtin_shufflevector(__s2_131, __s2_131, 7, 6, 5, 4, 3, 2, 1, 0); \
32159 bfloat16x8_t __reint_131 = __rev2_131; \
32160 float32x4_t __reint1_131 = __noswap_splatq_laneq_f32(*(float32x4_t *) &__reint_131, __p3_131); \
32161 __ret_131 = __noswap_vbfdotq_f32(__rev0_131, __rev1_131, *(bfloat16x8_t *) &__reint1_131); \
32162 __ret_131 = __builtin_shufflevector(__ret_131, __ret_131, 3, 2, 1, 0); \
32163 __ret_131; \
32164 })
32165 #endif
32166
32167 #ifdef __LITTLE_ENDIAN__
32168 #define vbfdot_laneq_f32(__p0_132, __p1_132, __p2_132, __p3_132) __extension__ ({ \
32169 float32x2_t __ret_132; \
32170 float32x2_t __s0_132 = __p0_132; \
32171 bfloat16x4_t __s1_132 = __p1_132; \
32172 bfloat16x8_t __s2_132 = __p2_132; \
32173 bfloat16x8_t __reint_132 = __s2_132; \
32174 float32x2_t __reint1_132 = splat_laneq_f32(*(float32x4_t *) &__reint_132, __p3_132); \
32175 __ret_132 = vbfdot_f32(__s0_132, __s1_132, *(bfloat16x4_t *) &__reint1_132); \
32176 __ret_132; \
32177 })
32178 #else
32179 #define vbfdot_laneq_f32(__p0_133, __p1_133, __p2_133, __p3_133) __extension__ ({ \
32180 float32x2_t __ret_133; \
32181 float32x2_t __s0_133 = __p0_133; \
32182 bfloat16x4_t __s1_133 = __p1_133; \
32183 bfloat16x8_t __s2_133 = __p2_133; \
32184 float32x2_t __rev0_133; __rev0_133 = __builtin_shufflevector(__s0_133, __s0_133, 1, 0); \
32185 bfloat16x4_t __rev1_133; __rev1_133 = __builtin_shufflevector(__s1_133, __s1_133, 3, 2, 1, 0); \
32186 bfloat16x8_t __rev2_133; __rev2_133 = __builtin_shufflevector(__s2_133, __s2_133, 7, 6, 5, 4, 3, 2, 1, 0); \
32187 bfloat16x8_t __reint_133 = __rev2_133; \
32188 float32x2_t __reint1_133 = __noswap_splat_laneq_f32(*(float32x4_t *) &__reint_133, __p3_133); \
32189 __ret_133 = __noswap_vbfdot_f32(__rev0_133, __rev1_133, *(bfloat16x4_t *) &__reint1_133); \
32190 __ret_133 = __builtin_shufflevector(__ret_133, __ret_133, 1, 0); \
32191 __ret_133; \
32192 })
32193 #endif
32194
32195 #ifdef __LITTLE_ENDIAN__
vbfmlalbq_f32(float32x4_t __p0,bfloat16x8_t __p1,bfloat16x8_t __p2)32196 __ai __attribute__((target("bf16"))) float32x4_t vbfmlalbq_f32(float32x4_t __p0, bfloat16x8_t __p1, bfloat16x8_t __p2) {
32197 float32x4_t __ret;
32198 __ret = (float32x4_t) __builtin_neon_vbfmlalbq_f32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
32199 return __ret;
32200 }
32201 #else
vbfmlalbq_f32(float32x4_t __p0,bfloat16x8_t __p1,bfloat16x8_t __p2)32202 __ai __attribute__((target("bf16"))) float32x4_t vbfmlalbq_f32(float32x4_t __p0, bfloat16x8_t __p1, bfloat16x8_t __p2) {
32203 float32x4_t __ret;
32204 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
32205 bfloat16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
32206 bfloat16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
32207 __ret = (float32x4_t) __builtin_neon_vbfmlalbq_f32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 41);
32208 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
32209 return __ret;
32210 }
__noswap_vbfmlalbq_f32(float32x4_t __p0,bfloat16x8_t __p1,bfloat16x8_t __p2)32211 __ai __attribute__((target("bf16"))) float32x4_t __noswap_vbfmlalbq_f32(float32x4_t __p0, bfloat16x8_t __p1, bfloat16x8_t __p2) {
32212 float32x4_t __ret;
32213 __ret = (float32x4_t) __builtin_neon_vbfmlalbq_f32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
32214 return __ret;
32215 }
32216 #endif
32217
32218 #ifdef __LITTLE_ENDIAN__
vbfmlaltq_f32(float32x4_t __p0,bfloat16x8_t __p1,bfloat16x8_t __p2)32219 __ai __attribute__((target("bf16"))) float32x4_t vbfmlaltq_f32(float32x4_t __p0, bfloat16x8_t __p1, bfloat16x8_t __p2) {
32220 float32x4_t __ret;
32221 __ret = (float32x4_t) __builtin_neon_vbfmlaltq_f32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
32222 return __ret;
32223 }
32224 #else
vbfmlaltq_f32(float32x4_t __p0,bfloat16x8_t __p1,bfloat16x8_t __p2)32225 __ai __attribute__((target("bf16"))) float32x4_t vbfmlaltq_f32(float32x4_t __p0, bfloat16x8_t __p1, bfloat16x8_t __p2) {
32226 float32x4_t __ret;
32227 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
32228 bfloat16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
32229 bfloat16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
32230 __ret = (float32x4_t) __builtin_neon_vbfmlaltq_f32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 41);
32231 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
32232 return __ret;
32233 }
__noswap_vbfmlaltq_f32(float32x4_t __p0,bfloat16x8_t __p1,bfloat16x8_t __p2)32234 __ai __attribute__((target("bf16"))) float32x4_t __noswap_vbfmlaltq_f32(float32x4_t __p0, bfloat16x8_t __p1, bfloat16x8_t __p2) {
32235 float32x4_t __ret;
32236 __ret = (float32x4_t) __builtin_neon_vbfmlaltq_f32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
32237 return __ret;
32238 }
32239 #endif
32240
32241 #ifdef __LITTLE_ENDIAN__
vbfmmlaq_f32(float32x4_t __p0,bfloat16x8_t __p1,bfloat16x8_t __p2)32242 __ai __attribute__((target("bf16"))) float32x4_t vbfmmlaq_f32(float32x4_t __p0, bfloat16x8_t __p1, bfloat16x8_t __p2) {
32243 float32x4_t __ret;
32244 __ret = (float32x4_t) __builtin_neon_vbfmmlaq_f32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
32245 return __ret;
32246 }
32247 #else
vbfmmlaq_f32(float32x4_t __p0,bfloat16x8_t __p1,bfloat16x8_t __p2)32248 __ai __attribute__((target("bf16"))) float32x4_t vbfmmlaq_f32(float32x4_t __p0, bfloat16x8_t __p1, bfloat16x8_t __p2) {
32249 float32x4_t __ret;
32250 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
32251 bfloat16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
32252 bfloat16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
32253 __ret = (float32x4_t) __builtin_neon_vbfmmlaq_f32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 41);
32254 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
32255 return __ret;
32256 }
32257 #endif
32258
32259 #ifdef __LITTLE_ENDIAN__
vcombine_bf16(bfloat16x4_t __p0,bfloat16x4_t __p1)32260 __ai __attribute__((target("bf16"))) bfloat16x8_t vcombine_bf16(bfloat16x4_t __p0, bfloat16x4_t __p1) {
32261 bfloat16x8_t __ret;
32262 __ret = __builtin_shufflevector(__p0, __p1, 0, 1, 2, 3, 4, 5, 6, 7);
32263 return __ret;
32264 }
32265 #else
vcombine_bf16(bfloat16x4_t __p0,bfloat16x4_t __p1)32266 __ai __attribute__((target("bf16"))) bfloat16x8_t vcombine_bf16(bfloat16x4_t __p0, bfloat16x4_t __p1) {
32267 bfloat16x8_t __ret;
32268 bfloat16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
32269 bfloat16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
32270 __ret = __builtin_shufflevector(__rev0, __rev1, 0, 1, 2, 3, 4, 5, 6, 7);
32271 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
32272 return __ret;
32273 }
__noswap_vcombine_bf16(bfloat16x4_t __p0,bfloat16x4_t __p1)32274 __ai __attribute__((target("bf16"))) bfloat16x8_t __noswap_vcombine_bf16(bfloat16x4_t __p0, bfloat16x4_t __p1) {
32275 bfloat16x8_t __ret;
32276 __ret = __builtin_shufflevector(__p0, __p1, 0, 1, 2, 3, 4, 5, 6, 7);
32277 return __ret;
32278 }
32279 #endif
32280
32281 #define vcreate_bf16(__p0) __extension__ ({ \
32282 bfloat16x4_t __ret; \
32283 uint64_t __promote = __p0; \
32284 __ret = (bfloat16x4_t)(__promote); \
32285 __ret; \
32286 })
32287 #ifdef __LITTLE_ENDIAN__
vcvt_f32_bf16(bfloat16x4_t __p0_134)32288 __ai __attribute__((target("bf16"))) float32x4_t vcvt_f32_bf16(bfloat16x4_t __p0_134) {
32289 float32x4_t __ret_134;
32290 bfloat16x4_t __reint_134 = __p0_134;
32291 int32x4_t __reint1_134 = vshll_n_s16(*(int16x4_t *) &__reint_134, 16);
32292 __ret_134 = *(float32x4_t *) &__reint1_134;
32293 return __ret_134;
32294 }
32295 #else
vcvt_f32_bf16(bfloat16x4_t __p0_135)32296 __ai __attribute__((target("bf16"))) float32x4_t vcvt_f32_bf16(bfloat16x4_t __p0_135) {
32297 float32x4_t __ret_135;
32298 bfloat16x4_t __rev0_135; __rev0_135 = __builtin_shufflevector(__p0_135, __p0_135, 3, 2, 1, 0);
32299 bfloat16x4_t __reint_135 = __rev0_135;
32300 int32x4_t __reint1_135 = __noswap_vshll_n_s16(*(int16x4_t *) &__reint_135, 16);
32301 __ret_135 = *(float32x4_t *) &__reint1_135;
32302 __ret_135 = __builtin_shufflevector(__ret_135, __ret_135, 3, 2, 1, 0);
32303 return __ret_135;
32304 }
__noswap_vcvt_f32_bf16(bfloat16x4_t __p0_136)32305 __ai __attribute__((target("bf16"))) float32x4_t __noswap_vcvt_f32_bf16(bfloat16x4_t __p0_136) {
32306 float32x4_t __ret_136;
32307 bfloat16x4_t __reint_136 = __p0_136;
32308 int32x4_t __reint1_136 = __noswap_vshll_n_s16(*(int16x4_t *) &__reint_136, 16);
32309 __ret_136 = *(float32x4_t *) &__reint1_136;
32310 return __ret_136;
32311 }
32312 #endif
32313
vcvtah_f32_bf16(bfloat16_t __p0)32314 __ai __attribute__((target("bf16"))) float32_t vcvtah_f32_bf16(bfloat16_t __p0) {
32315 float32_t __ret;
32316 bfloat16_t __reint = __p0;
32317 int32_t __reint1 = *(int32_t *) &__reint << 16;
32318 __ret = *(float32_t *) &__reint1;
32319 return __ret;
32320 }
vcvth_bf16_f32(float32_t __p0)32321 __ai __attribute__((target("bf16"))) bfloat16_t vcvth_bf16_f32(float32_t __p0) {
32322 bfloat16_t __ret;
32323 __ret = (bfloat16_t) __builtin_neon_vcvth_bf16_f32(__p0);
32324 return __ret;
32325 }
32326 #ifdef __LITTLE_ENDIAN__
32327 #define vduph_lane_bf16(__p0, __p1) __extension__ ({ \
32328 bfloat16_t __ret; \
32329 bfloat16x4_t __s0 = __p0; \
32330 __ret = (bfloat16_t) __builtin_neon_vduph_lane_bf16((bfloat16x4_t)__s0, __p1); \
32331 __ret; \
32332 })
32333 #else
32334 #define vduph_lane_bf16(__p0, __p1) __extension__ ({ \
32335 bfloat16_t __ret; \
32336 bfloat16x4_t __s0 = __p0; \
32337 bfloat16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
32338 __ret = (bfloat16_t) __builtin_neon_vduph_lane_bf16((bfloat16x4_t)__rev0, __p1); \
32339 __ret; \
32340 })
32341 #endif
32342
32343 #ifdef __LITTLE_ENDIAN__
32344 #define vdupq_lane_bf16(__p0_137, __p1_137) __extension__ ({ \
32345 bfloat16x8_t __ret_137; \
32346 bfloat16x4_t __s0_137 = __p0_137; \
32347 __ret_137 = splatq_lane_bf16(__s0_137, __p1_137); \
32348 __ret_137; \
32349 })
32350 #else
32351 #define vdupq_lane_bf16(__p0_138, __p1_138) __extension__ ({ \
32352 bfloat16x8_t __ret_138; \
32353 bfloat16x4_t __s0_138 = __p0_138; \
32354 bfloat16x4_t __rev0_138; __rev0_138 = __builtin_shufflevector(__s0_138, __s0_138, 3, 2, 1, 0); \
32355 __ret_138 = __noswap_splatq_lane_bf16(__rev0_138, __p1_138); \
32356 __ret_138 = __builtin_shufflevector(__ret_138, __ret_138, 7, 6, 5, 4, 3, 2, 1, 0); \
32357 __ret_138; \
32358 })
32359 #endif
32360
32361 #ifdef __LITTLE_ENDIAN__
32362 #define vdup_lane_bf16(__p0_139, __p1_139) __extension__ ({ \
32363 bfloat16x4_t __ret_139; \
32364 bfloat16x4_t __s0_139 = __p0_139; \
32365 __ret_139 = splat_lane_bf16(__s0_139, __p1_139); \
32366 __ret_139; \
32367 })
32368 #else
32369 #define vdup_lane_bf16(__p0_140, __p1_140) __extension__ ({ \
32370 bfloat16x4_t __ret_140; \
32371 bfloat16x4_t __s0_140 = __p0_140; \
32372 bfloat16x4_t __rev0_140; __rev0_140 = __builtin_shufflevector(__s0_140, __s0_140, 3, 2, 1, 0); \
32373 __ret_140 = __noswap_splat_lane_bf16(__rev0_140, __p1_140); \
32374 __ret_140 = __builtin_shufflevector(__ret_140, __ret_140, 3, 2, 1, 0); \
32375 __ret_140; \
32376 })
32377 #endif
32378
32379 #ifdef __LITTLE_ENDIAN__
32380 #define vduph_laneq_bf16(__p0, __p1) __extension__ ({ \
32381 bfloat16_t __ret; \
32382 bfloat16x8_t __s0 = __p0; \
32383 __ret = (bfloat16_t) __builtin_neon_vduph_laneq_bf16((bfloat16x8_t)__s0, __p1); \
32384 __ret; \
32385 })
32386 #else
32387 #define vduph_laneq_bf16(__p0, __p1) __extension__ ({ \
32388 bfloat16_t __ret; \
32389 bfloat16x8_t __s0 = __p0; \
32390 bfloat16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
32391 __ret = (bfloat16_t) __builtin_neon_vduph_laneq_bf16((bfloat16x8_t)__rev0, __p1); \
32392 __ret; \
32393 })
32394 #endif
32395
32396 #ifdef __LITTLE_ENDIAN__
32397 #define vdupq_laneq_bf16(__p0_141, __p1_141) __extension__ ({ \
32398 bfloat16x8_t __ret_141; \
32399 bfloat16x8_t __s0_141 = __p0_141; \
32400 __ret_141 = splatq_laneq_bf16(__s0_141, __p1_141); \
32401 __ret_141; \
32402 })
32403 #else
32404 #define vdupq_laneq_bf16(__p0_142, __p1_142) __extension__ ({ \
32405 bfloat16x8_t __ret_142; \
32406 bfloat16x8_t __s0_142 = __p0_142; \
32407 bfloat16x8_t __rev0_142; __rev0_142 = __builtin_shufflevector(__s0_142, __s0_142, 7, 6, 5, 4, 3, 2, 1, 0); \
32408 __ret_142 = __noswap_splatq_laneq_bf16(__rev0_142, __p1_142); \
32409 __ret_142 = __builtin_shufflevector(__ret_142, __ret_142, 7, 6, 5, 4, 3, 2, 1, 0); \
32410 __ret_142; \
32411 })
32412 #endif
32413
32414 #ifdef __LITTLE_ENDIAN__
32415 #define vdup_laneq_bf16(__p0_143, __p1_143) __extension__ ({ \
32416 bfloat16x4_t __ret_143; \
32417 bfloat16x8_t __s0_143 = __p0_143; \
32418 __ret_143 = splat_laneq_bf16(__s0_143, __p1_143); \
32419 __ret_143; \
32420 })
32421 #else
32422 #define vdup_laneq_bf16(__p0_144, __p1_144) __extension__ ({ \
32423 bfloat16x4_t __ret_144; \
32424 bfloat16x8_t __s0_144 = __p0_144; \
32425 bfloat16x8_t __rev0_144; __rev0_144 = __builtin_shufflevector(__s0_144, __s0_144, 7, 6, 5, 4, 3, 2, 1, 0); \
32426 __ret_144 = __noswap_splat_laneq_bf16(__rev0_144, __p1_144); \
32427 __ret_144 = __builtin_shufflevector(__ret_144, __ret_144, 3, 2, 1, 0); \
32428 __ret_144; \
32429 })
32430 #endif
32431
32432 #ifdef __LITTLE_ENDIAN__
vdupq_n_bf16(bfloat16_t __p0)32433 __ai __attribute__((target("bf16"))) bfloat16x8_t vdupq_n_bf16(bfloat16_t __p0) {
32434 bfloat16x8_t __ret;
32435 __ret = (bfloat16x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
32436 return __ret;
32437 }
32438 #else
vdupq_n_bf16(bfloat16_t __p0)32439 __ai __attribute__((target("bf16"))) bfloat16x8_t vdupq_n_bf16(bfloat16_t __p0) {
32440 bfloat16x8_t __ret;
32441 __ret = (bfloat16x8_t) {__p0, __p0, __p0, __p0, __p0, __p0, __p0, __p0};
32442 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
32443 return __ret;
32444 }
32445 #endif
32446
32447 #ifdef __LITTLE_ENDIAN__
vdup_n_bf16(bfloat16_t __p0)32448 __ai __attribute__((target("bf16"))) bfloat16x4_t vdup_n_bf16(bfloat16_t __p0) {
32449 bfloat16x4_t __ret;
32450 __ret = (bfloat16x4_t) {__p0, __p0, __p0, __p0};
32451 return __ret;
32452 }
32453 #else
vdup_n_bf16(bfloat16_t __p0)32454 __ai __attribute__((target("bf16"))) bfloat16x4_t vdup_n_bf16(bfloat16_t __p0) {
32455 bfloat16x4_t __ret;
32456 __ret = (bfloat16x4_t) {__p0, __p0, __p0, __p0};
32457 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
32458 return __ret;
32459 }
32460 #endif
32461
32462 #ifdef __LITTLE_ENDIAN__
vget_high_bf16(bfloat16x8_t __p0)32463 __ai __attribute__((target("bf16"))) bfloat16x4_t vget_high_bf16(bfloat16x8_t __p0) {
32464 bfloat16x4_t __ret;
32465 __ret = __builtin_shufflevector(__p0, __p0, 4, 5, 6, 7);
32466 return __ret;
32467 }
32468 #else
vget_high_bf16(bfloat16x8_t __p0)32469 __ai __attribute__((target("bf16"))) bfloat16x4_t vget_high_bf16(bfloat16x8_t __p0) {
32470 bfloat16x4_t __ret;
32471 bfloat16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
32472 __ret = __builtin_shufflevector(__rev0, __rev0, 4, 5, 6, 7);
32473 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
32474 return __ret;
32475 }
__noswap_vget_high_bf16(bfloat16x8_t __p0)32476 __ai __attribute__((target("bf16"))) bfloat16x4_t __noswap_vget_high_bf16(bfloat16x8_t __p0) {
32477 bfloat16x4_t __ret;
32478 __ret = __builtin_shufflevector(__p0, __p0, 4, 5, 6, 7);
32479 return __ret;
32480 }
32481 #endif
32482
32483 #ifdef __LITTLE_ENDIAN__
32484 #define vgetq_lane_bf16(__p0, __p1) __extension__ ({ \
32485 bfloat16_t __ret; \
32486 bfloat16x8_t __s0 = __p0; \
32487 __ret = (bfloat16_t) __builtin_neon_vgetq_lane_bf16((bfloat16x8_t)__s0, __p1); \
32488 __ret; \
32489 })
32490 #else
32491 #define vgetq_lane_bf16(__p0, __p1) __extension__ ({ \
32492 bfloat16_t __ret; \
32493 bfloat16x8_t __s0 = __p0; \
32494 bfloat16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
32495 __ret = (bfloat16_t) __builtin_neon_vgetq_lane_bf16((bfloat16x8_t)__rev0, __p1); \
32496 __ret; \
32497 })
32498 #define __noswap_vgetq_lane_bf16(__p0, __p1) __extension__ ({ \
32499 bfloat16_t __ret; \
32500 bfloat16x8_t __s0 = __p0; \
32501 __ret = (bfloat16_t) __builtin_neon_vgetq_lane_bf16((bfloat16x8_t)__s0, __p1); \
32502 __ret; \
32503 })
32504 #endif
32505
32506 #ifdef __LITTLE_ENDIAN__
32507 #define vget_lane_bf16(__p0, __p1) __extension__ ({ \
32508 bfloat16_t __ret; \
32509 bfloat16x4_t __s0 = __p0; \
32510 __ret = (bfloat16_t) __builtin_neon_vget_lane_bf16((bfloat16x4_t)__s0, __p1); \
32511 __ret; \
32512 })
32513 #else
32514 #define vget_lane_bf16(__p0, __p1) __extension__ ({ \
32515 bfloat16_t __ret; \
32516 bfloat16x4_t __s0 = __p0; \
32517 bfloat16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
32518 __ret = (bfloat16_t) __builtin_neon_vget_lane_bf16((bfloat16x4_t)__rev0, __p1); \
32519 __ret; \
32520 })
32521 #define __noswap_vget_lane_bf16(__p0, __p1) __extension__ ({ \
32522 bfloat16_t __ret; \
32523 bfloat16x4_t __s0 = __p0; \
32524 __ret = (bfloat16_t) __builtin_neon_vget_lane_bf16((bfloat16x4_t)__s0, __p1); \
32525 __ret; \
32526 })
32527 #endif
32528
32529 #ifdef __LITTLE_ENDIAN__
vget_low_bf16(bfloat16x8_t __p0)32530 __ai __attribute__((target("bf16"))) bfloat16x4_t vget_low_bf16(bfloat16x8_t __p0) {
32531 bfloat16x4_t __ret;
32532 __ret = __builtin_shufflevector(__p0, __p0, 0, 1, 2, 3);
32533 return __ret;
32534 }
32535 #else
vget_low_bf16(bfloat16x8_t __p0)32536 __ai __attribute__((target("bf16"))) bfloat16x4_t vget_low_bf16(bfloat16x8_t __p0) {
32537 bfloat16x4_t __ret;
32538 bfloat16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
32539 __ret = __builtin_shufflevector(__rev0, __rev0, 0, 1, 2, 3);
32540 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
32541 return __ret;
32542 }
__noswap_vget_low_bf16(bfloat16x8_t __p0)32543 __ai __attribute__((target("bf16"))) bfloat16x4_t __noswap_vget_low_bf16(bfloat16x8_t __p0) {
32544 bfloat16x4_t __ret;
32545 __ret = __builtin_shufflevector(__p0, __p0, 0, 1, 2, 3);
32546 return __ret;
32547 }
32548 #endif
32549
32550 #ifdef __LITTLE_ENDIAN__
32551 #define vld1q_bf16(__p0) __extension__ ({ \
32552 bfloat16x8_t __ret; \
32553 __ret = (bfloat16x8_t) __builtin_neon_vld1q_bf16(__p0, 43); \
32554 __ret; \
32555 })
32556 #else
32557 #define vld1q_bf16(__p0) __extension__ ({ \
32558 bfloat16x8_t __ret; \
32559 __ret = (bfloat16x8_t) __builtin_neon_vld1q_bf16(__p0, 43); \
32560 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
32561 __ret; \
32562 })
32563 #endif
32564
32565 #ifdef __LITTLE_ENDIAN__
32566 #define vld1_bf16(__p0) __extension__ ({ \
32567 bfloat16x4_t __ret; \
32568 __ret = (bfloat16x4_t) __builtin_neon_vld1_bf16(__p0, 11); \
32569 __ret; \
32570 })
32571 #else
32572 #define vld1_bf16(__p0) __extension__ ({ \
32573 bfloat16x4_t __ret; \
32574 __ret = (bfloat16x4_t) __builtin_neon_vld1_bf16(__p0, 11); \
32575 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
32576 __ret; \
32577 })
32578 #endif
32579
32580 #ifdef __LITTLE_ENDIAN__
32581 #define vld1q_dup_bf16(__p0) __extension__ ({ \
32582 bfloat16x8_t __ret; \
32583 __ret = (bfloat16x8_t) __builtin_neon_vld1q_dup_bf16(__p0, 43); \
32584 __ret; \
32585 })
32586 #else
32587 #define vld1q_dup_bf16(__p0) __extension__ ({ \
32588 bfloat16x8_t __ret; \
32589 __ret = (bfloat16x8_t) __builtin_neon_vld1q_dup_bf16(__p0, 43); \
32590 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
32591 __ret; \
32592 })
32593 #endif
32594
32595 #ifdef __LITTLE_ENDIAN__
32596 #define vld1_dup_bf16(__p0) __extension__ ({ \
32597 bfloat16x4_t __ret; \
32598 __ret = (bfloat16x4_t) __builtin_neon_vld1_dup_bf16(__p0, 11); \
32599 __ret; \
32600 })
32601 #else
32602 #define vld1_dup_bf16(__p0) __extension__ ({ \
32603 bfloat16x4_t __ret; \
32604 __ret = (bfloat16x4_t) __builtin_neon_vld1_dup_bf16(__p0, 11); \
32605 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
32606 __ret; \
32607 })
32608 #endif
32609
32610 #ifdef __LITTLE_ENDIAN__
32611 #define vld1q_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
32612 bfloat16x8_t __ret; \
32613 bfloat16x8_t __s1 = __p1; \
32614 __ret = (bfloat16x8_t) __builtin_neon_vld1q_lane_bf16(__p0, (int8x16_t)__s1, __p2, 43); \
32615 __ret; \
32616 })
32617 #else
32618 #define vld1q_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
32619 bfloat16x8_t __ret; \
32620 bfloat16x8_t __s1 = __p1; \
32621 bfloat16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
32622 __ret = (bfloat16x8_t) __builtin_neon_vld1q_lane_bf16(__p0, (int8x16_t)__rev1, __p2, 43); \
32623 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
32624 __ret; \
32625 })
32626 #endif
32627
32628 #ifdef __LITTLE_ENDIAN__
32629 #define vld1_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
32630 bfloat16x4_t __ret; \
32631 bfloat16x4_t __s1 = __p1; \
32632 __ret = (bfloat16x4_t) __builtin_neon_vld1_lane_bf16(__p0, (int8x8_t)__s1, __p2, 11); \
32633 __ret; \
32634 })
32635 #else
32636 #define vld1_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
32637 bfloat16x4_t __ret; \
32638 bfloat16x4_t __s1 = __p1; \
32639 bfloat16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
32640 __ret = (bfloat16x4_t) __builtin_neon_vld1_lane_bf16(__p0, (int8x8_t)__rev1, __p2, 11); \
32641 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
32642 __ret; \
32643 })
32644 #endif
32645
32646 #ifdef __LITTLE_ENDIAN__
32647 #define vld1q_bf16_x2(__p0) __extension__ ({ \
32648 bfloat16x8x2_t __ret; \
32649 __builtin_neon_vld1q_bf16_x2(&__ret, __p0, 43); \
32650 __ret; \
32651 })
32652 #else
32653 #define vld1q_bf16_x2(__p0) __extension__ ({ \
32654 bfloat16x8x2_t __ret; \
32655 __builtin_neon_vld1q_bf16_x2(&__ret, __p0, 43); \
32656 \
32657 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
32658 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
32659 __ret; \
32660 })
32661 #endif
32662
32663 #ifdef __LITTLE_ENDIAN__
32664 #define vld1_bf16_x2(__p0) __extension__ ({ \
32665 bfloat16x4x2_t __ret; \
32666 __builtin_neon_vld1_bf16_x2(&__ret, __p0, 11); \
32667 __ret; \
32668 })
32669 #else
32670 #define vld1_bf16_x2(__p0) __extension__ ({ \
32671 bfloat16x4x2_t __ret; \
32672 __builtin_neon_vld1_bf16_x2(&__ret, __p0, 11); \
32673 \
32674 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
32675 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
32676 __ret; \
32677 })
32678 #endif
32679
32680 #ifdef __LITTLE_ENDIAN__
32681 #define vld1q_bf16_x3(__p0) __extension__ ({ \
32682 bfloat16x8x3_t __ret; \
32683 __builtin_neon_vld1q_bf16_x3(&__ret, __p0, 43); \
32684 __ret; \
32685 })
32686 #else
32687 #define vld1q_bf16_x3(__p0) __extension__ ({ \
32688 bfloat16x8x3_t __ret; \
32689 __builtin_neon_vld1q_bf16_x3(&__ret, __p0, 43); \
32690 \
32691 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
32692 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
32693 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
32694 __ret; \
32695 })
32696 #endif
32697
32698 #ifdef __LITTLE_ENDIAN__
32699 #define vld1_bf16_x3(__p0) __extension__ ({ \
32700 bfloat16x4x3_t __ret; \
32701 __builtin_neon_vld1_bf16_x3(&__ret, __p0, 11); \
32702 __ret; \
32703 })
32704 #else
32705 #define vld1_bf16_x3(__p0) __extension__ ({ \
32706 bfloat16x4x3_t __ret; \
32707 __builtin_neon_vld1_bf16_x3(&__ret, __p0, 11); \
32708 \
32709 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
32710 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
32711 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
32712 __ret; \
32713 })
32714 #endif
32715
32716 #ifdef __LITTLE_ENDIAN__
32717 #define vld1q_bf16_x4(__p0) __extension__ ({ \
32718 bfloat16x8x4_t __ret; \
32719 __builtin_neon_vld1q_bf16_x4(&__ret, __p0, 43); \
32720 __ret; \
32721 })
32722 #else
32723 #define vld1q_bf16_x4(__p0) __extension__ ({ \
32724 bfloat16x8x4_t __ret; \
32725 __builtin_neon_vld1q_bf16_x4(&__ret, __p0, 43); \
32726 \
32727 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
32728 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
32729 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
32730 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
32731 __ret; \
32732 })
32733 #endif
32734
32735 #ifdef __LITTLE_ENDIAN__
32736 #define vld1_bf16_x4(__p0) __extension__ ({ \
32737 bfloat16x4x4_t __ret; \
32738 __builtin_neon_vld1_bf16_x4(&__ret, __p0, 11); \
32739 __ret; \
32740 })
32741 #else
32742 #define vld1_bf16_x4(__p0) __extension__ ({ \
32743 bfloat16x4x4_t __ret; \
32744 __builtin_neon_vld1_bf16_x4(&__ret, __p0, 11); \
32745 \
32746 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
32747 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
32748 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
32749 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
32750 __ret; \
32751 })
32752 #endif
32753
32754 #ifdef __LITTLE_ENDIAN__
32755 #define vld2q_bf16(__p0) __extension__ ({ \
32756 bfloat16x8x2_t __ret; \
32757 __builtin_neon_vld2q_bf16(&__ret, __p0, 43); \
32758 __ret; \
32759 })
32760 #else
32761 #define vld2q_bf16(__p0) __extension__ ({ \
32762 bfloat16x8x2_t __ret; \
32763 __builtin_neon_vld2q_bf16(&__ret, __p0, 43); \
32764 \
32765 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
32766 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
32767 __ret; \
32768 })
32769 #endif
32770
32771 #ifdef __LITTLE_ENDIAN__
32772 #define vld2_bf16(__p0) __extension__ ({ \
32773 bfloat16x4x2_t __ret; \
32774 __builtin_neon_vld2_bf16(&__ret, __p0, 11); \
32775 __ret; \
32776 })
32777 #else
32778 #define vld2_bf16(__p0) __extension__ ({ \
32779 bfloat16x4x2_t __ret; \
32780 __builtin_neon_vld2_bf16(&__ret, __p0, 11); \
32781 \
32782 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
32783 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
32784 __ret; \
32785 })
32786 #endif
32787
32788 #ifdef __LITTLE_ENDIAN__
32789 #define vld2q_dup_bf16(__p0) __extension__ ({ \
32790 bfloat16x8x2_t __ret; \
32791 __builtin_neon_vld2q_dup_bf16(&__ret, __p0, 43); \
32792 __ret; \
32793 })
32794 #else
32795 #define vld2q_dup_bf16(__p0) __extension__ ({ \
32796 bfloat16x8x2_t __ret; \
32797 __builtin_neon_vld2q_dup_bf16(&__ret, __p0, 43); \
32798 \
32799 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
32800 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
32801 __ret; \
32802 })
32803 #endif
32804
32805 #ifdef __LITTLE_ENDIAN__
32806 #define vld2_dup_bf16(__p0) __extension__ ({ \
32807 bfloat16x4x2_t __ret; \
32808 __builtin_neon_vld2_dup_bf16(&__ret, __p0, 11); \
32809 __ret; \
32810 })
32811 #else
32812 #define vld2_dup_bf16(__p0) __extension__ ({ \
32813 bfloat16x4x2_t __ret; \
32814 __builtin_neon_vld2_dup_bf16(&__ret, __p0, 11); \
32815 \
32816 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
32817 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
32818 __ret; \
32819 })
32820 #endif
32821
32822 #ifdef __LITTLE_ENDIAN__
32823 #define vld2q_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
32824 bfloat16x8x2_t __ret; \
32825 bfloat16x8x2_t __s1 = __p1; \
32826 __builtin_neon_vld2q_lane_bf16(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 43); \
32827 __ret; \
32828 })
32829 #else
32830 #define vld2q_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
32831 bfloat16x8x2_t __ret; \
32832 bfloat16x8x2_t __s1 = __p1; \
32833 bfloat16x8x2_t __rev1; \
32834 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
32835 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
32836 __builtin_neon_vld2q_lane_bf16(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 43); \
32837 \
32838 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
32839 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
32840 __ret; \
32841 })
32842 #endif
32843
32844 #ifdef __LITTLE_ENDIAN__
32845 #define vld2_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
32846 bfloat16x4x2_t __ret; \
32847 bfloat16x4x2_t __s1 = __p1; \
32848 __builtin_neon_vld2_lane_bf16(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 11); \
32849 __ret; \
32850 })
32851 #else
32852 #define vld2_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
32853 bfloat16x4x2_t __ret; \
32854 bfloat16x4x2_t __s1 = __p1; \
32855 bfloat16x4x2_t __rev1; \
32856 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
32857 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
32858 __builtin_neon_vld2_lane_bf16(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 11); \
32859 \
32860 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
32861 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
32862 __ret; \
32863 })
32864 #endif
32865
32866 #ifdef __LITTLE_ENDIAN__
32867 #define vld3q_bf16(__p0) __extension__ ({ \
32868 bfloat16x8x3_t __ret; \
32869 __builtin_neon_vld3q_bf16(&__ret, __p0, 43); \
32870 __ret; \
32871 })
32872 #else
32873 #define vld3q_bf16(__p0) __extension__ ({ \
32874 bfloat16x8x3_t __ret; \
32875 __builtin_neon_vld3q_bf16(&__ret, __p0, 43); \
32876 \
32877 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
32878 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
32879 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
32880 __ret; \
32881 })
32882 #endif
32883
32884 #ifdef __LITTLE_ENDIAN__
32885 #define vld3_bf16(__p0) __extension__ ({ \
32886 bfloat16x4x3_t __ret; \
32887 __builtin_neon_vld3_bf16(&__ret, __p0, 11); \
32888 __ret; \
32889 })
32890 #else
32891 #define vld3_bf16(__p0) __extension__ ({ \
32892 bfloat16x4x3_t __ret; \
32893 __builtin_neon_vld3_bf16(&__ret, __p0, 11); \
32894 \
32895 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
32896 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
32897 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
32898 __ret; \
32899 })
32900 #endif
32901
32902 #ifdef __LITTLE_ENDIAN__
32903 #define vld3q_dup_bf16(__p0) __extension__ ({ \
32904 bfloat16x8x3_t __ret; \
32905 __builtin_neon_vld3q_dup_bf16(&__ret, __p0, 43); \
32906 __ret; \
32907 })
32908 #else
32909 #define vld3q_dup_bf16(__p0) __extension__ ({ \
32910 bfloat16x8x3_t __ret; \
32911 __builtin_neon_vld3q_dup_bf16(&__ret, __p0, 43); \
32912 \
32913 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
32914 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
32915 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
32916 __ret; \
32917 })
32918 #endif
32919
32920 #ifdef __LITTLE_ENDIAN__
32921 #define vld3_dup_bf16(__p0) __extension__ ({ \
32922 bfloat16x4x3_t __ret; \
32923 __builtin_neon_vld3_dup_bf16(&__ret, __p0, 11); \
32924 __ret; \
32925 })
32926 #else
32927 #define vld3_dup_bf16(__p0) __extension__ ({ \
32928 bfloat16x4x3_t __ret; \
32929 __builtin_neon_vld3_dup_bf16(&__ret, __p0, 11); \
32930 \
32931 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
32932 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
32933 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
32934 __ret; \
32935 })
32936 #endif
32937
32938 #ifdef __LITTLE_ENDIAN__
32939 #define vld3q_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
32940 bfloat16x8x3_t __ret; \
32941 bfloat16x8x3_t __s1 = __p1; \
32942 __builtin_neon_vld3q_lane_bf16(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 43); \
32943 __ret; \
32944 })
32945 #else
32946 #define vld3q_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
32947 bfloat16x8x3_t __ret; \
32948 bfloat16x8x3_t __s1 = __p1; \
32949 bfloat16x8x3_t __rev1; \
32950 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
32951 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
32952 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
32953 __builtin_neon_vld3q_lane_bf16(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 43); \
32954 \
32955 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
32956 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
32957 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
32958 __ret; \
32959 })
32960 #endif
32961
32962 #ifdef __LITTLE_ENDIAN__
32963 #define vld3_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
32964 bfloat16x4x3_t __ret; \
32965 bfloat16x4x3_t __s1 = __p1; \
32966 __builtin_neon_vld3_lane_bf16(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 11); \
32967 __ret; \
32968 })
32969 #else
32970 #define vld3_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
32971 bfloat16x4x3_t __ret; \
32972 bfloat16x4x3_t __s1 = __p1; \
32973 bfloat16x4x3_t __rev1; \
32974 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
32975 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
32976 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
32977 __builtin_neon_vld3_lane_bf16(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 11); \
32978 \
32979 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
32980 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
32981 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
32982 __ret; \
32983 })
32984 #endif
32985
32986 #ifdef __LITTLE_ENDIAN__
32987 #define vld4q_bf16(__p0) __extension__ ({ \
32988 bfloat16x8x4_t __ret; \
32989 __builtin_neon_vld4q_bf16(&__ret, __p0, 43); \
32990 __ret; \
32991 })
32992 #else
32993 #define vld4q_bf16(__p0) __extension__ ({ \
32994 bfloat16x8x4_t __ret; \
32995 __builtin_neon_vld4q_bf16(&__ret, __p0, 43); \
32996 \
32997 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
32998 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
32999 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
33000 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
33001 __ret; \
33002 })
33003 #endif
33004
33005 #ifdef __LITTLE_ENDIAN__
33006 #define vld4_bf16(__p0) __extension__ ({ \
33007 bfloat16x4x4_t __ret; \
33008 __builtin_neon_vld4_bf16(&__ret, __p0, 11); \
33009 __ret; \
33010 })
33011 #else
33012 #define vld4_bf16(__p0) __extension__ ({ \
33013 bfloat16x4x4_t __ret; \
33014 __builtin_neon_vld4_bf16(&__ret, __p0, 11); \
33015 \
33016 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
33017 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
33018 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
33019 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
33020 __ret; \
33021 })
33022 #endif
33023
33024 #ifdef __LITTLE_ENDIAN__
33025 #define vld4q_dup_bf16(__p0) __extension__ ({ \
33026 bfloat16x8x4_t __ret; \
33027 __builtin_neon_vld4q_dup_bf16(&__ret, __p0, 43); \
33028 __ret; \
33029 })
33030 #else
33031 #define vld4q_dup_bf16(__p0) __extension__ ({ \
33032 bfloat16x8x4_t __ret; \
33033 __builtin_neon_vld4q_dup_bf16(&__ret, __p0, 43); \
33034 \
33035 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
33036 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
33037 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
33038 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
33039 __ret; \
33040 })
33041 #endif
33042
33043 #ifdef __LITTLE_ENDIAN__
33044 #define vld4_dup_bf16(__p0) __extension__ ({ \
33045 bfloat16x4x4_t __ret; \
33046 __builtin_neon_vld4_dup_bf16(&__ret, __p0, 11); \
33047 __ret; \
33048 })
33049 #else
33050 #define vld4_dup_bf16(__p0) __extension__ ({ \
33051 bfloat16x4x4_t __ret; \
33052 __builtin_neon_vld4_dup_bf16(&__ret, __p0, 11); \
33053 \
33054 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
33055 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
33056 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
33057 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
33058 __ret; \
33059 })
33060 #endif
33061
33062 #ifdef __LITTLE_ENDIAN__
33063 #define vld4q_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
33064 bfloat16x8x4_t __ret; \
33065 bfloat16x8x4_t __s1 = __p1; \
33066 __builtin_neon_vld4q_lane_bf16(&__ret, __p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 43); \
33067 __ret; \
33068 })
33069 #else
33070 #define vld4q_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
33071 bfloat16x8x4_t __ret; \
33072 bfloat16x8x4_t __s1 = __p1; \
33073 bfloat16x8x4_t __rev1; \
33074 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
33075 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
33076 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
33077 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
33078 __builtin_neon_vld4q_lane_bf16(&__ret, __p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 43); \
33079 \
33080 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
33081 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
33082 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
33083 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
33084 __ret; \
33085 })
33086 #endif
33087
33088 #ifdef __LITTLE_ENDIAN__
33089 #define vld4_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
33090 bfloat16x4x4_t __ret; \
33091 bfloat16x4x4_t __s1 = __p1; \
33092 __builtin_neon_vld4_lane_bf16(&__ret, __p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 11); \
33093 __ret; \
33094 })
33095 #else
33096 #define vld4_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
33097 bfloat16x4x4_t __ret; \
33098 bfloat16x4x4_t __s1 = __p1; \
33099 bfloat16x4x4_t __rev1; \
33100 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
33101 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
33102 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
33103 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
33104 __builtin_neon_vld4_lane_bf16(&__ret, __p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 11); \
33105 \
33106 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0); \
33107 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0); \
33108 __ret.val[2] = __builtin_shufflevector(__ret.val[2], __ret.val[2], 3, 2, 1, 0); \
33109 __ret.val[3] = __builtin_shufflevector(__ret.val[3], __ret.val[3], 3, 2, 1, 0); \
33110 __ret; \
33111 })
33112 #endif
33113
33114 #ifdef __LITTLE_ENDIAN__
33115 #define vsetq_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
33116 bfloat16x8_t __ret; \
33117 bfloat16_t __s0 = __p0; \
33118 bfloat16x8_t __s1 = __p1; \
33119 __ret = (bfloat16x8_t) __builtin_neon_vsetq_lane_bf16(__s0, (bfloat16x8_t)__s1, __p2); \
33120 __ret; \
33121 })
33122 #else
33123 #define vsetq_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
33124 bfloat16x8_t __ret; \
33125 bfloat16_t __s0 = __p0; \
33126 bfloat16x8_t __s1 = __p1; \
33127 bfloat16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
33128 __ret = (bfloat16x8_t) __builtin_neon_vsetq_lane_bf16(__s0, (bfloat16x8_t)__rev1, __p2); \
33129 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
33130 __ret; \
33131 })
33132 #define __noswap_vsetq_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
33133 bfloat16x8_t __ret; \
33134 bfloat16_t __s0 = __p0; \
33135 bfloat16x8_t __s1 = __p1; \
33136 __ret = (bfloat16x8_t) __builtin_neon_vsetq_lane_bf16(__s0, (bfloat16x8_t)__s1, __p2); \
33137 __ret; \
33138 })
33139 #endif
33140
33141 #ifdef __LITTLE_ENDIAN__
33142 #define vset_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
33143 bfloat16x4_t __ret; \
33144 bfloat16_t __s0 = __p0; \
33145 bfloat16x4_t __s1 = __p1; \
33146 __ret = (bfloat16x4_t) __builtin_neon_vset_lane_bf16(__s0, (bfloat16x4_t)__s1, __p2); \
33147 __ret; \
33148 })
33149 #else
33150 #define vset_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
33151 bfloat16x4_t __ret; \
33152 bfloat16_t __s0 = __p0; \
33153 bfloat16x4_t __s1 = __p1; \
33154 bfloat16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
33155 __ret = (bfloat16x4_t) __builtin_neon_vset_lane_bf16(__s0, (bfloat16x4_t)__rev1, __p2); \
33156 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
33157 __ret; \
33158 })
33159 #define __noswap_vset_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
33160 bfloat16x4_t __ret; \
33161 bfloat16_t __s0 = __p0; \
33162 bfloat16x4_t __s1 = __p1; \
33163 __ret = (bfloat16x4_t) __builtin_neon_vset_lane_bf16(__s0, (bfloat16x4_t)__s1, __p2); \
33164 __ret; \
33165 })
33166 #endif
33167
33168 #ifdef __LITTLE_ENDIAN__
33169 #define vst1q_bf16(__p0, __p1) __extension__ ({ \
33170 bfloat16x8_t __s1 = __p1; \
33171 __builtin_neon_vst1q_bf16(__p0, (int8x16_t)__s1, 43); \
33172 })
33173 #else
33174 #define vst1q_bf16(__p0, __p1) __extension__ ({ \
33175 bfloat16x8_t __s1 = __p1; \
33176 bfloat16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
33177 __builtin_neon_vst1q_bf16(__p0, (int8x16_t)__rev1, 43); \
33178 })
33179 #endif
33180
33181 #ifdef __LITTLE_ENDIAN__
33182 #define vst1_bf16(__p0, __p1) __extension__ ({ \
33183 bfloat16x4_t __s1 = __p1; \
33184 __builtin_neon_vst1_bf16(__p0, (int8x8_t)__s1, 11); \
33185 })
33186 #else
33187 #define vst1_bf16(__p0, __p1) __extension__ ({ \
33188 bfloat16x4_t __s1 = __p1; \
33189 bfloat16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
33190 __builtin_neon_vst1_bf16(__p0, (int8x8_t)__rev1, 11); \
33191 })
33192 #endif
33193
33194 #ifdef __LITTLE_ENDIAN__
33195 #define vst1q_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
33196 bfloat16x8_t __s1 = __p1; \
33197 __builtin_neon_vst1q_lane_bf16(__p0, (int8x16_t)__s1, __p2, 43); \
33198 })
33199 #else
33200 #define vst1q_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
33201 bfloat16x8_t __s1 = __p1; \
33202 bfloat16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
33203 __builtin_neon_vst1q_lane_bf16(__p0, (int8x16_t)__rev1, __p2, 43); \
33204 })
33205 #endif
33206
33207 #ifdef __LITTLE_ENDIAN__
33208 #define vst1_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
33209 bfloat16x4_t __s1 = __p1; \
33210 __builtin_neon_vst1_lane_bf16(__p0, (int8x8_t)__s1, __p2, 11); \
33211 })
33212 #else
33213 #define vst1_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
33214 bfloat16x4_t __s1 = __p1; \
33215 bfloat16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
33216 __builtin_neon_vst1_lane_bf16(__p0, (int8x8_t)__rev1, __p2, 11); \
33217 })
33218 #endif
33219
33220 #ifdef __LITTLE_ENDIAN__
33221 #define vst1q_bf16_x2(__p0, __p1) __extension__ ({ \
33222 bfloat16x8x2_t __s1 = __p1; \
33223 __builtin_neon_vst1q_bf16_x2(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 43); \
33224 })
33225 #else
33226 #define vst1q_bf16_x2(__p0, __p1) __extension__ ({ \
33227 bfloat16x8x2_t __s1 = __p1; \
33228 bfloat16x8x2_t __rev1; \
33229 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
33230 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
33231 __builtin_neon_vst1q_bf16_x2(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 43); \
33232 })
33233 #endif
33234
33235 #ifdef __LITTLE_ENDIAN__
33236 #define vst1_bf16_x2(__p0, __p1) __extension__ ({ \
33237 bfloat16x4x2_t __s1 = __p1; \
33238 __builtin_neon_vst1_bf16_x2(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 11); \
33239 })
33240 #else
33241 #define vst1_bf16_x2(__p0, __p1) __extension__ ({ \
33242 bfloat16x4x2_t __s1 = __p1; \
33243 bfloat16x4x2_t __rev1; \
33244 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
33245 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
33246 __builtin_neon_vst1_bf16_x2(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 11); \
33247 })
33248 #endif
33249
33250 #ifdef __LITTLE_ENDIAN__
33251 #define vst1q_bf16_x3(__p0, __p1) __extension__ ({ \
33252 bfloat16x8x3_t __s1 = __p1; \
33253 __builtin_neon_vst1q_bf16_x3(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 43); \
33254 })
33255 #else
33256 #define vst1q_bf16_x3(__p0, __p1) __extension__ ({ \
33257 bfloat16x8x3_t __s1 = __p1; \
33258 bfloat16x8x3_t __rev1; \
33259 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
33260 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
33261 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
33262 __builtin_neon_vst1q_bf16_x3(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 43); \
33263 })
33264 #endif
33265
33266 #ifdef __LITTLE_ENDIAN__
33267 #define vst1_bf16_x3(__p0, __p1) __extension__ ({ \
33268 bfloat16x4x3_t __s1 = __p1; \
33269 __builtin_neon_vst1_bf16_x3(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 11); \
33270 })
33271 #else
33272 #define vst1_bf16_x3(__p0, __p1) __extension__ ({ \
33273 bfloat16x4x3_t __s1 = __p1; \
33274 bfloat16x4x3_t __rev1; \
33275 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
33276 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
33277 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
33278 __builtin_neon_vst1_bf16_x3(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 11); \
33279 })
33280 #endif
33281
33282 #ifdef __LITTLE_ENDIAN__
33283 #define vst1q_bf16_x4(__p0, __p1) __extension__ ({ \
33284 bfloat16x8x4_t __s1 = __p1; \
33285 __builtin_neon_vst1q_bf16_x4(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 43); \
33286 })
33287 #else
33288 #define vst1q_bf16_x4(__p0, __p1) __extension__ ({ \
33289 bfloat16x8x4_t __s1 = __p1; \
33290 bfloat16x8x4_t __rev1; \
33291 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
33292 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
33293 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
33294 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
33295 __builtin_neon_vst1q_bf16_x4(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 43); \
33296 })
33297 #endif
33298
33299 #ifdef __LITTLE_ENDIAN__
33300 #define vst1_bf16_x4(__p0, __p1) __extension__ ({ \
33301 bfloat16x4x4_t __s1 = __p1; \
33302 __builtin_neon_vst1_bf16_x4(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 11); \
33303 })
33304 #else
33305 #define vst1_bf16_x4(__p0, __p1) __extension__ ({ \
33306 bfloat16x4x4_t __s1 = __p1; \
33307 bfloat16x4x4_t __rev1; \
33308 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
33309 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
33310 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
33311 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
33312 __builtin_neon_vst1_bf16_x4(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 11); \
33313 })
33314 #endif
33315
33316 #ifdef __LITTLE_ENDIAN__
33317 #define vst2q_bf16(__p0, __p1) __extension__ ({ \
33318 bfloat16x8x2_t __s1 = __p1; \
33319 __builtin_neon_vst2q_bf16(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], 43); \
33320 })
33321 #else
33322 #define vst2q_bf16(__p0, __p1) __extension__ ({ \
33323 bfloat16x8x2_t __s1 = __p1; \
33324 bfloat16x8x2_t __rev1; \
33325 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
33326 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
33327 __builtin_neon_vst2q_bf16(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], 43); \
33328 })
33329 #endif
33330
33331 #ifdef __LITTLE_ENDIAN__
33332 #define vst2_bf16(__p0, __p1) __extension__ ({ \
33333 bfloat16x4x2_t __s1 = __p1; \
33334 __builtin_neon_vst2_bf16(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], 11); \
33335 })
33336 #else
33337 #define vst2_bf16(__p0, __p1) __extension__ ({ \
33338 bfloat16x4x2_t __s1 = __p1; \
33339 bfloat16x4x2_t __rev1; \
33340 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
33341 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
33342 __builtin_neon_vst2_bf16(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], 11); \
33343 })
33344 #endif
33345
33346 #ifdef __LITTLE_ENDIAN__
33347 #define vst2q_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
33348 bfloat16x8x2_t __s1 = __p1; \
33349 __builtin_neon_vst2q_lane_bf16(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], __p2, 43); \
33350 })
33351 #else
33352 #define vst2q_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
33353 bfloat16x8x2_t __s1 = __p1; \
33354 bfloat16x8x2_t __rev1; \
33355 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
33356 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
33357 __builtin_neon_vst2q_lane_bf16(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], __p2, 43); \
33358 })
33359 #endif
33360
33361 #ifdef __LITTLE_ENDIAN__
33362 #define vst2_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
33363 bfloat16x4x2_t __s1 = __p1; \
33364 __builtin_neon_vst2_lane_bf16(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], __p2, 11); \
33365 })
33366 #else
33367 #define vst2_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
33368 bfloat16x4x2_t __s1 = __p1; \
33369 bfloat16x4x2_t __rev1; \
33370 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
33371 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
33372 __builtin_neon_vst2_lane_bf16(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], __p2, 11); \
33373 })
33374 #endif
33375
33376 #ifdef __LITTLE_ENDIAN__
33377 #define vst3q_bf16(__p0, __p1) __extension__ ({ \
33378 bfloat16x8x3_t __s1 = __p1; \
33379 __builtin_neon_vst3q_bf16(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], 43); \
33380 })
33381 #else
33382 #define vst3q_bf16(__p0, __p1) __extension__ ({ \
33383 bfloat16x8x3_t __s1 = __p1; \
33384 bfloat16x8x3_t __rev1; \
33385 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
33386 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
33387 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
33388 __builtin_neon_vst3q_bf16(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], 43); \
33389 })
33390 #endif
33391
33392 #ifdef __LITTLE_ENDIAN__
33393 #define vst3_bf16(__p0, __p1) __extension__ ({ \
33394 bfloat16x4x3_t __s1 = __p1; \
33395 __builtin_neon_vst3_bf16(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], 11); \
33396 })
33397 #else
33398 #define vst3_bf16(__p0, __p1) __extension__ ({ \
33399 bfloat16x4x3_t __s1 = __p1; \
33400 bfloat16x4x3_t __rev1; \
33401 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
33402 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
33403 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
33404 __builtin_neon_vst3_bf16(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], 11); \
33405 })
33406 #endif
33407
33408 #ifdef __LITTLE_ENDIAN__
33409 #define vst3q_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
33410 bfloat16x8x3_t __s1 = __p1; \
33411 __builtin_neon_vst3q_lane_bf16(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], __p2, 43); \
33412 })
33413 #else
33414 #define vst3q_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
33415 bfloat16x8x3_t __s1 = __p1; \
33416 bfloat16x8x3_t __rev1; \
33417 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
33418 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
33419 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
33420 __builtin_neon_vst3q_lane_bf16(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], __p2, 43); \
33421 })
33422 #endif
33423
33424 #ifdef __LITTLE_ENDIAN__
33425 #define vst3_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
33426 bfloat16x4x3_t __s1 = __p1; \
33427 __builtin_neon_vst3_lane_bf16(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], __p2, 11); \
33428 })
33429 #else
33430 #define vst3_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
33431 bfloat16x4x3_t __s1 = __p1; \
33432 bfloat16x4x3_t __rev1; \
33433 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
33434 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
33435 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
33436 __builtin_neon_vst3_lane_bf16(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], __p2, 11); \
33437 })
33438 #endif
33439
33440 #ifdef __LITTLE_ENDIAN__
33441 #define vst4q_bf16(__p0, __p1) __extension__ ({ \
33442 bfloat16x8x4_t __s1 = __p1; \
33443 __builtin_neon_vst4q_bf16(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], 43); \
33444 })
33445 #else
33446 #define vst4q_bf16(__p0, __p1) __extension__ ({ \
33447 bfloat16x8x4_t __s1 = __p1; \
33448 bfloat16x8x4_t __rev1; \
33449 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
33450 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
33451 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
33452 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
33453 __builtin_neon_vst4q_bf16(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], 43); \
33454 })
33455 #endif
33456
33457 #ifdef __LITTLE_ENDIAN__
33458 #define vst4_bf16(__p0, __p1) __extension__ ({ \
33459 bfloat16x4x4_t __s1 = __p1; \
33460 __builtin_neon_vst4_bf16(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], 11); \
33461 })
33462 #else
33463 #define vst4_bf16(__p0, __p1) __extension__ ({ \
33464 bfloat16x4x4_t __s1 = __p1; \
33465 bfloat16x4x4_t __rev1; \
33466 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
33467 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
33468 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
33469 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
33470 __builtin_neon_vst4_bf16(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], 11); \
33471 })
33472 #endif
33473
33474 #ifdef __LITTLE_ENDIAN__
33475 #define vst4q_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
33476 bfloat16x8x4_t __s1 = __p1; \
33477 __builtin_neon_vst4q_lane_bf16(__p0, (int8x16_t)__s1.val[0], (int8x16_t)__s1.val[1], (int8x16_t)__s1.val[2], (int8x16_t)__s1.val[3], __p2, 43); \
33478 })
33479 #else
33480 #define vst4q_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
33481 bfloat16x8x4_t __s1 = __p1; \
33482 bfloat16x8x4_t __rev1; \
33483 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 7, 6, 5, 4, 3, 2, 1, 0); \
33484 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 7, 6, 5, 4, 3, 2, 1, 0); \
33485 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 7, 6, 5, 4, 3, 2, 1, 0); \
33486 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 7, 6, 5, 4, 3, 2, 1, 0); \
33487 __builtin_neon_vst4q_lane_bf16(__p0, (int8x16_t)__rev1.val[0], (int8x16_t)__rev1.val[1], (int8x16_t)__rev1.val[2], (int8x16_t)__rev1.val[3], __p2, 43); \
33488 })
33489 #endif
33490
33491 #ifdef __LITTLE_ENDIAN__
33492 #define vst4_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
33493 bfloat16x4x4_t __s1 = __p1; \
33494 __builtin_neon_vst4_lane_bf16(__p0, (int8x8_t)__s1.val[0], (int8x8_t)__s1.val[1], (int8x8_t)__s1.val[2], (int8x8_t)__s1.val[3], __p2, 11); \
33495 })
33496 #else
33497 #define vst4_lane_bf16(__p0, __p1, __p2) __extension__ ({ \
33498 bfloat16x4x4_t __s1 = __p1; \
33499 bfloat16x4x4_t __rev1; \
33500 __rev1.val[0] = __builtin_shufflevector(__s1.val[0], __s1.val[0], 3, 2, 1, 0); \
33501 __rev1.val[1] = __builtin_shufflevector(__s1.val[1], __s1.val[1], 3, 2, 1, 0); \
33502 __rev1.val[2] = __builtin_shufflevector(__s1.val[2], __s1.val[2], 3, 2, 1, 0); \
33503 __rev1.val[3] = __builtin_shufflevector(__s1.val[3], __s1.val[3], 3, 2, 1, 0); \
33504 __builtin_neon_vst4_lane_bf16(__p0, (int8x8_t)__rev1.val[0], (int8x8_t)__rev1.val[1], (int8x8_t)__rev1.val[2], (int8x8_t)__rev1.val[3], __p2, 11); \
33505 })
33506 #endif
33507
33508 #ifdef __LITTLE_ENDIAN__
vdotq_u32(uint32x4_t __p0,uint8x16_t __p1,uint8x16_t __p2)33509 __ai __attribute__((target("dotprod"))) uint32x4_t vdotq_u32(uint32x4_t __p0, uint8x16_t __p1, uint8x16_t __p2) {
33510 uint32x4_t __ret;
33511 __ret = (uint32x4_t) __builtin_neon_vdotq_u32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 50);
33512 return __ret;
33513 }
33514 #else
vdotq_u32(uint32x4_t __p0,uint8x16_t __p1,uint8x16_t __p2)33515 __ai __attribute__((target("dotprod"))) uint32x4_t vdotq_u32(uint32x4_t __p0, uint8x16_t __p1, uint8x16_t __p2) {
33516 uint32x4_t __ret;
33517 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
33518 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
33519 uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
33520 __ret = (uint32x4_t) __builtin_neon_vdotq_u32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 50);
33521 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
33522 return __ret;
33523 }
__noswap_vdotq_u32(uint32x4_t __p0,uint8x16_t __p1,uint8x16_t __p2)33524 __ai __attribute__((target("dotprod"))) uint32x4_t __noswap_vdotq_u32(uint32x4_t __p0, uint8x16_t __p1, uint8x16_t __p2) {
33525 uint32x4_t __ret;
33526 __ret = (uint32x4_t) __builtin_neon_vdotq_u32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 50);
33527 return __ret;
33528 }
33529 #endif
33530
33531 #ifdef __LITTLE_ENDIAN__
vdotq_s32(int32x4_t __p0,int8x16_t __p1,int8x16_t __p2)33532 __ai __attribute__((target("dotprod"))) int32x4_t vdotq_s32(int32x4_t __p0, int8x16_t __p1, int8x16_t __p2) {
33533 int32x4_t __ret;
33534 __ret = (int32x4_t) __builtin_neon_vdotq_s32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 34);
33535 return __ret;
33536 }
33537 #else
vdotq_s32(int32x4_t __p0,int8x16_t __p1,int8x16_t __p2)33538 __ai __attribute__((target("dotprod"))) int32x4_t vdotq_s32(int32x4_t __p0, int8x16_t __p1, int8x16_t __p2) {
33539 int32x4_t __ret;
33540 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
33541 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
33542 int8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
33543 __ret = (int32x4_t) __builtin_neon_vdotq_s32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 34);
33544 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
33545 return __ret;
33546 }
__noswap_vdotq_s32(int32x4_t __p0,int8x16_t __p1,int8x16_t __p2)33547 __ai __attribute__((target("dotprod"))) int32x4_t __noswap_vdotq_s32(int32x4_t __p0, int8x16_t __p1, int8x16_t __p2) {
33548 int32x4_t __ret;
33549 __ret = (int32x4_t) __builtin_neon_vdotq_s32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 34);
33550 return __ret;
33551 }
33552 #endif
33553
33554 #ifdef __LITTLE_ENDIAN__
vdot_u32(uint32x2_t __p0,uint8x8_t __p1,uint8x8_t __p2)33555 __ai __attribute__((target("dotprod"))) uint32x2_t vdot_u32(uint32x2_t __p0, uint8x8_t __p1, uint8x8_t __p2) {
33556 uint32x2_t __ret;
33557 __ret = (uint32x2_t) __builtin_neon_vdot_u32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 18);
33558 return __ret;
33559 }
33560 #else
vdot_u32(uint32x2_t __p0,uint8x8_t __p1,uint8x8_t __p2)33561 __ai __attribute__((target("dotprod"))) uint32x2_t vdot_u32(uint32x2_t __p0, uint8x8_t __p1, uint8x8_t __p2) {
33562 uint32x2_t __ret;
33563 uint32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
33564 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
33565 uint8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
33566 __ret = (uint32x2_t) __builtin_neon_vdot_u32((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 18);
33567 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
33568 return __ret;
33569 }
__noswap_vdot_u32(uint32x2_t __p0,uint8x8_t __p1,uint8x8_t __p2)33570 __ai __attribute__((target("dotprod"))) uint32x2_t __noswap_vdot_u32(uint32x2_t __p0, uint8x8_t __p1, uint8x8_t __p2) {
33571 uint32x2_t __ret;
33572 __ret = (uint32x2_t) __builtin_neon_vdot_u32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 18);
33573 return __ret;
33574 }
33575 #endif
33576
33577 #ifdef __LITTLE_ENDIAN__
vdot_s32(int32x2_t __p0,int8x8_t __p1,int8x8_t __p2)33578 __ai __attribute__((target("dotprod"))) int32x2_t vdot_s32(int32x2_t __p0, int8x8_t __p1, int8x8_t __p2) {
33579 int32x2_t __ret;
33580 __ret = (int32x2_t) __builtin_neon_vdot_s32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 2);
33581 return __ret;
33582 }
33583 #else
vdot_s32(int32x2_t __p0,int8x8_t __p1,int8x8_t __p2)33584 __ai __attribute__((target("dotprod"))) int32x2_t vdot_s32(int32x2_t __p0, int8x8_t __p1, int8x8_t __p2) {
33585 int32x2_t __ret;
33586 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
33587 int8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
33588 int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
33589 __ret = (int32x2_t) __builtin_neon_vdot_s32((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 2);
33590 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
33591 return __ret;
33592 }
__noswap_vdot_s32(int32x2_t __p0,int8x8_t __p1,int8x8_t __p2)33593 __ai __attribute__((target("dotprod"))) int32x2_t __noswap_vdot_s32(int32x2_t __p0, int8x8_t __p1, int8x8_t __p2) {
33594 int32x2_t __ret;
33595 __ret = (int32x2_t) __builtin_neon_vdot_s32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 2);
33596 return __ret;
33597 }
33598 #endif
33599
33600 #ifdef __LITTLE_ENDIAN__
33601 #define vdotq_lane_u32(__p0_145, __p1_145, __p2_145, __p3_145) __extension__ ({ \
33602 uint32x4_t __ret_145; \
33603 uint32x4_t __s0_145 = __p0_145; \
33604 uint8x16_t __s1_145 = __p1_145; \
33605 uint8x8_t __s2_145 = __p2_145; \
33606 uint8x8_t __reint_145 = __s2_145; \
33607 uint32x4_t __reint1_145 = splatq_lane_u32(*(uint32x2_t *) &__reint_145, __p3_145); \
33608 __ret_145 = vdotq_u32(__s0_145, __s1_145, *(uint8x16_t *) &__reint1_145); \
33609 __ret_145; \
33610 })
33611 #else
33612 #define vdotq_lane_u32(__p0_146, __p1_146, __p2_146, __p3_146) __extension__ ({ \
33613 uint32x4_t __ret_146; \
33614 uint32x4_t __s0_146 = __p0_146; \
33615 uint8x16_t __s1_146 = __p1_146; \
33616 uint8x8_t __s2_146 = __p2_146; \
33617 uint32x4_t __rev0_146; __rev0_146 = __builtin_shufflevector(__s0_146, __s0_146, 3, 2, 1, 0); \
33618 uint8x16_t __rev1_146; __rev1_146 = __builtin_shufflevector(__s1_146, __s1_146, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
33619 uint8x8_t __rev2_146; __rev2_146 = __builtin_shufflevector(__s2_146, __s2_146, 7, 6, 5, 4, 3, 2, 1, 0); \
33620 uint8x8_t __reint_146 = __rev2_146; \
33621 uint32x4_t __reint1_146 = __noswap_splatq_lane_u32(*(uint32x2_t *) &__reint_146, __p3_146); \
33622 __ret_146 = __noswap_vdotq_u32(__rev0_146, __rev1_146, *(uint8x16_t *) &__reint1_146); \
33623 __ret_146 = __builtin_shufflevector(__ret_146, __ret_146, 3, 2, 1, 0); \
33624 __ret_146; \
33625 })
33626 #endif
33627
33628 #ifdef __LITTLE_ENDIAN__
33629 #define vdotq_lane_s32(__p0_147, __p1_147, __p2_147, __p3_147) __extension__ ({ \
33630 int32x4_t __ret_147; \
33631 int32x4_t __s0_147 = __p0_147; \
33632 int8x16_t __s1_147 = __p1_147; \
33633 int8x8_t __s2_147 = __p2_147; \
33634 int8x8_t __reint_147 = __s2_147; \
33635 int32x4_t __reint1_147 = splatq_lane_s32(*(int32x2_t *) &__reint_147, __p3_147); \
33636 __ret_147 = vdotq_s32(__s0_147, __s1_147, *(int8x16_t *) &__reint1_147); \
33637 __ret_147; \
33638 })
33639 #else
33640 #define vdotq_lane_s32(__p0_148, __p1_148, __p2_148, __p3_148) __extension__ ({ \
33641 int32x4_t __ret_148; \
33642 int32x4_t __s0_148 = __p0_148; \
33643 int8x16_t __s1_148 = __p1_148; \
33644 int8x8_t __s2_148 = __p2_148; \
33645 int32x4_t __rev0_148; __rev0_148 = __builtin_shufflevector(__s0_148, __s0_148, 3, 2, 1, 0); \
33646 int8x16_t __rev1_148; __rev1_148 = __builtin_shufflevector(__s1_148, __s1_148, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
33647 int8x8_t __rev2_148; __rev2_148 = __builtin_shufflevector(__s2_148, __s2_148, 7, 6, 5, 4, 3, 2, 1, 0); \
33648 int8x8_t __reint_148 = __rev2_148; \
33649 int32x4_t __reint1_148 = __noswap_splatq_lane_s32(*(int32x2_t *) &__reint_148, __p3_148); \
33650 __ret_148 = __noswap_vdotq_s32(__rev0_148, __rev1_148, *(int8x16_t *) &__reint1_148); \
33651 __ret_148 = __builtin_shufflevector(__ret_148, __ret_148, 3, 2, 1, 0); \
33652 __ret_148; \
33653 })
33654 #endif
33655
33656 #ifdef __LITTLE_ENDIAN__
33657 #define vdot_lane_u32(__p0_149, __p1_149, __p2_149, __p3_149) __extension__ ({ \
33658 uint32x2_t __ret_149; \
33659 uint32x2_t __s0_149 = __p0_149; \
33660 uint8x8_t __s1_149 = __p1_149; \
33661 uint8x8_t __s2_149 = __p2_149; \
33662 uint8x8_t __reint_149 = __s2_149; \
33663 uint32x2_t __reint1_149 = splat_lane_u32(*(uint32x2_t *) &__reint_149, __p3_149); \
33664 __ret_149 = vdot_u32(__s0_149, __s1_149, *(uint8x8_t *) &__reint1_149); \
33665 __ret_149; \
33666 })
33667 #else
33668 #define vdot_lane_u32(__p0_150, __p1_150, __p2_150, __p3_150) __extension__ ({ \
33669 uint32x2_t __ret_150; \
33670 uint32x2_t __s0_150 = __p0_150; \
33671 uint8x8_t __s1_150 = __p1_150; \
33672 uint8x8_t __s2_150 = __p2_150; \
33673 uint32x2_t __rev0_150; __rev0_150 = __builtin_shufflevector(__s0_150, __s0_150, 1, 0); \
33674 uint8x8_t __rev1_150; __rev1_150 = __builtin_shufflevector(__s1_150, __s1_150, 7, 6, 5, 4, 3, 2, 1, 0); \
33675 uint8x8_t __rev2_150; __rev2_150 = __builtin_shufflevector(__s2_150, __s2_150, 7, 6, 5, 4, 3, 2, 1, 0); \
33676 uint8x8_t __reint_150 = __rev2_150; \
33677 uint32x2_t __reint1_150 = __noswap_splat_lane_u32(*(uint32x2_t *) &__reint_150, __p3_150); \
33678 __ret_150 = __noswap_vdot_u32(__rev0_150, __rev1_150, *(uint8x8_t *) &__reint1_150); \
33679 __ret_150 = __builtin_shufflevector(__ret_150, __ret_150, 1, 0); \
33680 __ret_150; \
33681 })
33682 #endif
33683
33684 #ifdef __LITTLE_ENDIAN__
33685 #define vdot_lane_s32(__p0_151, __p1_151, __p2_151, __p3_151) __extension__ ({ \
33686 int32x2_t __ret_151; \
33687 int32x2_t __s0_151 = __p0_151; \
33688 int8x8_t __s1_151 = __p1_151; \
33689 int8x8_t __s2_151 = __p2_151; \
33690 int8x8_t __reint_151 = __s2_151; \
33691 int32x2_t __reint1_151 = splat_lane_s32(*(int32x2_t *) &__reint_151, __p3_151); \
33692 __ret_151 = vdot_s32(__s0_151, __s1_151, *(int8x8_t *) &__reint1_151); \
33693 __ret_151; \
33694 })
33695 #else
33696 #define vdot_lane_s32(__p0_152, __p1_152, __p2_152, __p3_152) __extension__ ({ \
33697 int32x2_t __ret_152; \
33698 int32x2_t __s0_152 = __p0_152; \
33699 int8x8_t __s1_152 = __p1_152; \
33700 int8x8_t __s2_152 = __p2_152; \
33701 int32x2_t __rev0_152; __rev0_152 = __builtin_shufflevector(__s0_152, __s0_152, 1, 0); \
33702 int8x8_t __rev1_152; __rev1_152 = __builtin_shufflevector(__s1_152, __s1_152, 7, 6, 5, 4, 3, 2, 1, 0); \
33703 int8x8_t __rev2_152; __rev2_152 = __builtin_shufflevector(__s2_152, __s2_152, 7, 6, 5, 4, 3, 2, 1, 0); \
33704 int8x8_t __reint_152 = __rev2_152; \
33705 int32x2_t __reint1_152 = __noswap_splat_lane_s32(*(int32x2_t *) &__reint_152, __p3_152); \
33706 __ret_152 = __noswap_vdot_s32(__rev0_152, __rev1_152, *(int8x8_t *) &__reint1_152); \
33707 __ret_152 = __builtin_shufflevector(__ret_152, __ret_152, 1, 0); \
33708 __ret_152; \
33709 })
33710 #endif
33711
33712 #ifdef __LITTLE_ENDIAN__
vabdq_f16(float16x8_t __p0,float16x8_t __p1)33713 __ai __attribute__((target("fullfp16"))) float16x8_t vabdq_f16(float16x8_t __p0, float16x8_t __p1) {
33714 float16x8_t __ret;
33715 __ret = (float16x8_t) __builtin_neon_vabdq_f16((int8x16_t)__p0, (int8x16_t)__p1, 40);
33716 return __ret;
33717 }
33718 #else
vabdq_f16(float16x8_t __p0,float16x8_t __p1)33719 __ai __attribute__((target("fullfp16"))) float16x8_t vabdq_f16(float16x8_t __p0, float16x8_t __p1) {
33720 float16x8_t __ret;
33721 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
33722 float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
33723 __ret = (float16x8_t) __builtin_neon_vabdq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, 40);
33724 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
33725 return __ret;
33726 }
33727 #endif
33728
33729 #ifdef __LITTLE_ENDIAN__
vabd_f16(float16x4_t __p0,float16x4_t __p1)33730 __ai __attribute__((target("fullfp16"))) float16x4_t vabd_f16(float16x4_t __p0, float16x4_t __p1) {
33731 float16x4_t __ret;
33732 __ret = (float16x4_t) __builtin_neon_vabd_f16((int8x8_t)__p0, (int8x8_t)__p1, 8);
33733 return __ret;
33734 }
33735 #else
vabd_f16(float16x4_t __p0,float16x4_t __p1)33736 __ai __attribute__((target("fullfp16"))) float16x4_t vabd_f16(float16x4_t __p0, float16x4_t __p1) {
33737 float16x4_t __ret;
33738 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
33739 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
33740 __ret = (float16x4_t) __builtin_neon_vabd_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 8);
33741 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
33742 return __ret;
33743 }
33744 #endif
33745
33746 #ifdef __LITTLE_ENDIAN__
vabsq_f16(float16x8_t __p0)33747 __ai __attribute__((target("fullfp16"))) float16x8_t vabsq_f16(float16x8_t __p0) {
33748 float16x8_t __ret;
33749 __ret = (float16x8_t) __builtin_neon_vabsq_f16((int8x16_t)__p0, 40);
33750 return __ret;
33751 }
33752 #else
vabsq_f16(float16x8_t __p0)33753 __ai __attribute__((target("fullfp16"))) float16x8_t vabsq_f16(float16x8_t __p0) {
33754 float16x8_t __ret;
33755 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
33756 __ret = (float16x8_t) __builtin_neon_vabsq_f16((int8x16_t)__rev0, 40);
33757 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
33758 return __ret;
33759 }
33760 #endif
33761
33762 #ifdef __LITTLE_ENDIAN__
vabs_f16(float16x4_t __p0)33763 __ai __attribute__((target("fullfp16"))) float16x4_t vabs_f16(float16x4_t __p0) {
33764 float16x4_t __ret;
33765 __ret = (float16x4_t) __builtin_neon_vabs_f16((int8x8_t)__p0, 8);
33766 return __ret;
33767 }
33768 #else
vabs_f16(float16x4_t __p0)33769 __ai __attribute__((target("fullfp16"))) float16x4_t vabs_f16(float16x4_t __p0) {
33770 float16x4_t __ret;
33771 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
33772 __ret = (float16x4_t) __builtin_neon_vabs_f16((int8x8_t)__rev0, 8);
33773 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
33774 return __ret;
33775 }
33776 #endif
33777
33778 #ifdef __LITTLE_ENDIAN__
vaddq_f16(float16x8_t __p0,float16x8_t __p1)33779 __ai __attribute__((target("fullfp16"))) float16x8_t vaddq_f16(float16x8_t __p0, float16x8_t __p1) {
33780 float16x8_t __ret;
33781 __ret = __p0 + __p1;
33782 return __ret;
33783 }
33784 #else
vaddq_f16(float16x8_t __p0,float16x8_t __p1)33785 __ai __attribute__((target("fullfp16"))) float16x8_t vaddq_f16(float16x8_t __p0, float16x8_t __p1) {
33786 float16x8_t __ret;
33787 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
33788 float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
33789 __ret = __rev0 + __rev1;
33790 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
33791 return __ret;
33792 }
33793 #endif
33794
33795 #ifdef __LITTLE_ENDIAN__
vadd_f16(float16x4_t __p0,float16x4_t __p1)33796 __ai __attribute__((target("fullfp16"))) float16x4_t vadd_f16(float16x4_t __p0, float16x4_t __p1) {
33797 float16x4_t __ret;
33798 __ret = __p0 + __p1;
33799 return __ret;
33800 }
33801 #else
vadd_f16(float16x4_t __p0,float16x4_t __p1)33802 __ai __attribute__((target("fullfp16"))) float16x4_t vadd_f16(float16x4_t __p0, float16x4_t __p1) {
33803 float16x4_t __ret;
33804 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
33805 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
33806 __ret = __rev0 + __rev1;
33807 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
33808 return __ret;
33809 }
33810 #endif
33811
33812 #ifdef __LITTLE_ENDIAN__
vbslq_f16(uint16x8_t __p0,float16x8_t __p1,float16x8_t __p2)33813 __ai __attribute__((target("fullfp16"))) float16x8_t vbslq_f16(uint16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
33814 float16x8_t __ret;
33815 __ret = (float16x8_t) __builtin_neon_vbslq_f16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 40);
33816 return __ret;
33817 }
33818 #else
vbslq_f16(uint16x8_t __p0,float16x8_t __p1,float16x8_t __p2)33819 __ai __attribute__((target("fullfp16"))) float16x8_t vbslq_f16(uint16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
33820 float16x8_t __ret;
33821 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
33822 float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
33823 float16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
33824 __ret = (float16x8_t) __builtin_neon_vbslq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 40);
33825 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
33826 return __ret;
33827 }
33828 #endif
33829
33830 #ifdef __LITTLE_ENDIAN__
vbsl_f16(uint16x4_t __p0,float16x4_t __p1,float16x4_t __p2)33831 __ai __attribute__((target("fullfp16"))) float16x4_t vbsl_f16(uint16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
33832 float16x4_t __ret;
33833 __ret = (float16x4_t) __builtin_neon_vbsl_f16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 8);
33834 return __ret;
33835 }
33836 #else
vbsl_f16(uint16x4_t __p0,float16x4_t __p1,float16x4_t __p2)33837 __ai __attribute__((target("fullfp16"))) float16x4_t vbsl_f16(uint16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
33838 float16x4_t __ret;
33839 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
33840 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
33841 float16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
33842 __ret = (float16x4_t) __builtin_neon_vbsl_f16((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 8);
33843 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
33844 return __ret;
33845 }
33846 #endif
33847
33848 #ifdef __LITTLE_ENDIAN__
vcageq_f16(float16x8_t __p0,float16x8_t __p1)33849 __ai __attribute__((target("fullfp16"))) uint16x8_t vcageq_f16(float16x8_t __p0, float16x8_t __p1) {
33850 uint16x8_t __ret;
33851 __ret = (uint16x8_t) __builtin_neon_vcageq_f16((int8x16_t)__p0, (int8x16_t)__p1, 49);
33852 return __ret;
33853 }
33854 #else
vcageq_f16(float16x8_t __p0,float16x8_t __p1)33855 __ai __attribute__((target("fullfp16"))) uint16x8_t vcageq_f16(float16x8_t __p0, float16x8_t __p1) {
33856 uint16x8_t __ret;
33857 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
33858 float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
33859 __ret = (uint16x8_t) __builtin_neon_vcageq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
33860 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
33861 return __ret;
33862 }
33863 #endif
33864
33865 #ifdef __LITTLE_ENDIAN__
vcage_f16(float16x4_t __p0,float16x4_t __p1)33866 __ai __attribute__((target("fullfp16"))) uint16x4_t vcage_f16(float16x4_t __p0, float16x4_t __p1) {
33867 uint16x4_t __ret;
33868 __ret = (uint16x4_t) __builtin_neon_vcage_f16((int8x8_t)__p0, (int8x8_t)__p1, 17);
33869 return __ret;
33870 }
33871 #else
vcage_f16(float16x4_t __p0,float16x4_t __p1)33872 __ai __attribute__((target("fullfp16"))) uint16x4_t vcage_f16(float16x4_t __p0, float16x4_t __p1) {
33873 uint16x4_t __ret;
33874 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
33875 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
33876 __ret = (uint16x4_t) __builtin_neon_vcage_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
33877 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
33878 return __ret;
33879 }
33880 #endif
33881
33882 #ifdef __LITTLE_ENDIAN__
vcagtq_f16(float16x8_t __p0,float16x8_t __p1)33883 __ai __attribute__((target("fullfp16"))) uint16x8_t vcagtq_f16(float16x8_t __p0, float16x8_t __p1) {
33884 uint16x8_t __ret;
33885 __ret = (uint16x8_t) __builtin_neon_vcagtq_f16((int8x16_t)__p0, (int8x16_t)__p1, 49);
33886 return __ret;
33887 }
33888 #else
vcagtq_f16(float16x8_t __p0,float16x8_t __p1)33889 __ai __attribute__((target("fullfp16"))) uint16x8_t vcagtq_f16(float16x8_t __p0, float16x8_t __p1) {
33890 uint16x8_t __ret;
33891 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
33892 float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
33893 __ret = (uint16x8_t) __builtin_neon_vcagtq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
33894 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
33895 return __ret;
33896 }
33897 #endif
33898
33899 #ifdef __LITTLE_ENDIAN__
vcagt_f16(float16x4_t __p0,float16x4_t __p1)33900 __ai __attribute__((target("fullfp16"))) uint16x4_t vcagt_f16(float16x4_t __p0, float16x4_t __p1) {
33901 uint16x4_t __ret;
33902 __ret = (uint16x4_t) __builtin_neon_vcagt_f16((int8x8_t)__p0, (int8x8_t)__p1, 17);
33903 return __ret;
33904 }
33905 #else
vcagt_f16(float16x4_t __p0,float16x4_t __p1)33906 __ai __attribute__((target("fullfp16"))) uint16x4_t vcagt_f16(float16x4_t __p0, float16x4_t __p1) {
33907 uint16x4_t __ret;
33908 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
33909 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
33910 __ret = (uint16x4_t) __builtin_neon_vcagt_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
33911 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
33912 return __ret;
33913 }
33914 #endif
33915
33916 #ifdef __LITTLE_ENDIAN__
vcaleq_f16(float16x8_t __p0,float16x8_t __p1)33917 __ai __attribute__((target("fullfp16"))) uint16x8_t vcaleq_f16(float16x8_t __p0, float16x8_t __p1) {
33918 uint16x8_t __ret;
33919 __ret = (uint16x8_t) __builtin_neon_vcaleq_f16((int8x16_t)__p0, (int8x16_t)__p1, 49);
33920 return __ret;
33921 }
33922 #else
vcaleq_f16(float16x8_t __p0,float16x8_t __p1)33923 __ai __attribute__((target("fullfp16"))) uint16x8_t vcaleq_f16(float16x8_t __p0, float16x8_t __p1) {
33924 uint16x8_t __ret;
33925 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
33926 float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
33927 __ret = (uint16x8_t) __builtin_neon_vcaleq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
33928 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
33929 return __ret;
33930 }
33931 #endif
33932
33933 #ifdef __LITTLE_ENDIAN__
vcale_f16(float16x4_t __p0,float16x4_t __p1)33934 __ai __attribute__((target("fullfp16"))) uint16x4_t vcale_f16(float16x4_t __p0, float16x4_t __p1) {
33935 uint16x4_t __ret;
33936 __ret = (uint16x4_t) __builtin_neon_vcale_f16((int8x8_t)__p0, (int8x8_t)__p1, 17);
33937 return __ret;
33938 }
33939 #else
vcale_f16(float16x4_t __p0,float16x4_t __p1)33940 __ai __attribute__((target("fullfp16"))) uint16x4_t vcale_f16(float16x4_t __p0, float16x4_t __p1) {
33941 uint16x4_t __ret;
33942 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
33943 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
33944 __ret = (uint16x4_t) __builtin_neon_vcale_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
33945 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
33946 return __ret;
33947 }
33948 #endif
33949
33950 #ifdef __LITTLE_ENDIAN__
vcaltq_f16(float16x8_t __p0,float16x8_t __p1)33951 __ai __attribute__((target("fullfp16"))) uint16x8_t vcaltq_f16(float16x8_t __p0, float16x8_t __p1) {
33952 uint16x8_t __ret;
33953 __ret = (uint16x8_t) __builtin_neon_vcaltq_f16((int8x16_t)__p0, (int8x16_t)__p1, 49);
33954 return __ret;
33955 }
33956 #else
vcaltq_f16(float16x8_t __p0,float16x8_t __p1)33957 __ai __attribute__((target("fullfp16"))) uint16x8_t vcaltq_f16(float16x8_t __p0, float16x8_t __p1) {
33958 uint16x8_t __ret;
33959 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
33960 float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
33961 __ret = (uint16x8_t) __builtin_neon_vcaltq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, 49);
33962 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
33963 return __ret;
33964 }
33965 #endif
33966
33967 #ifdef __LITTLE_ENDIAN__
vcalt_f16(float16x4_t __p0,float16x4_t __p1)33968 __ai __attribute__((target("fullfp16"))) uint16x4_t vcalt_f16(float16x4_t __p0, float16x4_t __p1) {
33969 uint16x4_t __ret;
33970 __ret = (uint16x4_t) __builtin_neon_vcalt_f16((int8x8_t)__p0, (int8x8_t)__p1, 17);
33971 return __ret;
33972 }
33973 #else
vcalt_f16(float16x4_t __p0,float16x4_t __p1)33974 __ai __attribute__((target("fullfp16"))) uint16x4_t vcalt_f16(float16x4_t __p0, float16x4_t __p1) {
33975 uint16x4_t __ret;
33976 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
33977 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
33978 __ret = (uint16x4_t) __builtin_neon_vcalt_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 17);
33979 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
33980 return __ret;
33981 }
33982 #endif
33983
33984 #ifdef __LITTLE_ENDIAN__
vceqq_f16(float16x8_t __p0,float16x8_t __p1)33985 __ai __attribute__((target("fullfp16"))) uint16x8_t vceqq_f16(float16x8_t __p0, float16x8_t __p1) {
33986 uint16x8_t __ret;
33987 __ret = (uint16x8_t)(__p0 == __p1);
33988 return __ret;
33989 }
33990 #else
vceqq_f16(float16x8_t __p0,float16x8_t __p1)33991 __ai __attribute__((target("fullfp16"))) uint16x8_t vceqq_f16(float16x8_t __p0, float16x8_t __p1) {
33992 uint16x8_t __ret;
33993 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
33994 float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
33995 __ret = (uint16x8_t)(__rev0 == __rev1);
33996 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
33997 return __ret;
33998 }
33999 #endif
34000
34001 #ifdef __LITTLE_ENDIAN__
vceq_f16(float16x4_t __p0,float16x4_t __p1)34002 __ai __attribute__((target("fullfp16"))) uint16x4_t vceq_f16(float16x4_t __p0, float16x4_t __p1) {
34003 uint16x4_t __ret;
34004 __ret = (uint16x4_t)(__p0 == __p1);
34005 return __ret;
34006 }
34007 #else
vceq_f16(float16x4_t __p0,float16x4_t __p1)34008 __ai __attribute__((target("fullfp16"))) uint16x4_t vceq_f16(float16x4_t __p0, float16x4_t __p1) {
34009 uint16x4_t __ret;
34010 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34011 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
34012 __ret = (uint16x4_t)(__rev0 == __rev1);
34013 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
34014 return __ret;
34015 }
34016 #endif
34017
34018 #ifdef __LITTLE_ENDIAN__
vceqzq_f16(float16x8_t __p0)34019 __ai __attribute__((target("fullfp16"))) uint16x8_t vceqzq_f16(float16x8_t __p0) {
34020 uint16x8_t __ret;
34021 __ret = (uint16x8_t) __builtin_neon_vceqzq_f16((int8x16_t)__p0, 49);
34022 return __ret;
34023 }
34024 #else
vceqzq_f16(float16x8_t __p0)34025 __ai __attribute__((target("fullfp16"))) uint16x8_t vceqzq_f16(float16x8_t __p0) {
34026 uint16x8_t __ret;
34027 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34028 __ret = (uint16x8_t) __builtin_neon_vceqzq_f16((int8x16_t)__rev0, 49);
34029 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
34030 return __ret;
34031 }
34032 #endif
34033
34034 #ifdef __LITTLE_ENDIAN__
vceqz_f16(float16x4_t __p0)34035 __ai __attribute__((target("fullfp16"))) uint16x4_t vceqz_f16(float16x4_t __p0) {
34036 uint16x4_t __ret;
34037 __ret = (uint16x4_t) __builtin_neon_vceqz_f16((int8x8_t)__p0, 17);
34038 return __ret;
34039 }
34040 #else
vceqz_f16(float16x4_t __p0)34041 __ai __attribute__((target("fullfp16"))) uint16x4_t vceqz_f16(float16x4_t __p0) {
34042 uint16x4_t __ret;
34043 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34044 __ret = (uint16x4_t) __builtin_neon_vceqz_f16((int8x8_t)__rev0, 17);
34045 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
34046 return __ret;
34047 }
34048 #endif
34049
34050 #ifdef __LITTLE_ENDIAN__
vcgeq_f16(float16x8_t __p0,float16x8_t __p1)34051 __ai __attribute__((target("fullfp16"))) uint16x8_t vcgeq_f16(float16x8_t __p0, float16x8_t __p1) {
34052 uint16x8_t __ret;
34053 __ret = (uint16x8_t)(__p0 >= __p1);
34054 return __ret;
34055 }
34056 #else
vcgeq_f16(float16x8_t __p0,float16x8_t __p1)34057 __ai __attribute__((target("fullfp16"))) uint16x8_t vcgeq_f16(float16x8_t __p0, float16x8_t __p1) {
34058 uint16x8_t __ret;
34059 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34060 float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
34061 __ret = (uint16x8_t)(__rev0 >= __rev1);
34062 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
34063 return __ret;
34064 }
34065 #endif
34066
34067 #ifdef __LITTLE_ENDIAN__
vcge_f16(float16x4_t __p0,float16x4_t __p1)34068 __ai __attribute__((target("fullfp16"))) uint16x4_t vcge_f16(float16x4_t __p0, float16x4_t __p1) {
34069 uint16x4_t __ret;
34070 __ret = (uint16x4_t)(__p0 >= __p1);
34071 return __ret;
34072 }
34073 #else
vcge_f16(float16x4_t __p0,float16x4_t __p1)34074 __ai __attribute__((target("fullfp16"))) uint16x4_t vcge_f16(float16x4_t __p0, float16x4_t __p1) {
34075 uint16x4_t __ret;
34076 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34077 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
34078 __ret = (uint16x4_t)(__rev0 >= __rev1);
34079 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
34080 return __ret;
34081 }
34082 #endif
34083
34084 #ifdef __LITTLE_ENDIAN__
vcgezq_f16(float16x8_t __p0)34085 __ai __attribute__((target("fullfp16"))) uint16x8_t vcgezq_f16(float16x8_t __p0) {
34086 uint16x8_t __ret;
34087 __ret = (uint16x8_t) __builtin_neon_vcgezq_f16((int8x16_t)__p0, 49);
34088 return __ret;
34089 }
34090 #else
vcgezq_f16(float16x8_t __p0)34091 __ai __attribute__((target("fullfp16"))) uint16x8_t vcgezq_f16(float16x8_t __p0) {
34092 uint16x8_t __ret;
34093 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34094 __ret = (uint16x8_t) __builtin_neon_vcgezq_f16((int8x16_t)__rev0, 49);
34095 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
34096 return __ret;
34097 }
34098 #endif
34099
34100 #ifdef __LITTLE_ENDIAN__
vcgez_f16(float16x4_t __p0)34101 __ai __attribute__((target("fullfp16"))) uint16x4_t vcgez_f16(float16x4_t __p0) {
34102 uint16x4_t __ret;
34103 __ret = (uint16x4_t) __builtin_neon_vcgez_f16((int8x8_t)__p0, 17);
34104 return __ret;
34105 }
34106 #else
vcgez_f16(float16x4_t __p0)34107 __ai __attribute__((target("fullfp16"))) uint16x4_t vcgez_f16(float16x4_t __p0) {
34108 uint16x4_t __ret;
34109 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34110 __ret = (uint16x4_t) __builtin_neon_vcgez_f16((int8x8_t)__rev0, 17);
34111 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
34112 return __ret;
34113 }
34114 #endif
34115
34116 #ifdef __LITTLE_ENDIAN__
vcgtq_f16(float16x8_t __p0,float16x8_t __p1)34117 __ai __attribute__((target("fullfp16"))) uint16x8_t vcgtq_f16(float16x8_t __p0, float16x8_t __p1) {
34118 uint16x8_t __ret;
34119 __ret = (uint16x8_t)(__p0 > __p1);
34120 return __ret;
34121 }
34122 #else
vcgtq_f16(float16x8_t __p0,float16x8_t __p1)34123 __ai __attribute__((target("fullfp16"))) uint16x8_t vcgtq_f16(float16x8_t __p0, float16x8_t __p1) {
34124 uint16x8_t __ret;
34125 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34126 float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
34127 __ret = (uint16x8_t)(__rev0 > __rev1);
34128 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
34129 return __ret;
34130 }
34131 #endif
34132
34133 #ifdef __LITTLE_ENDIAN__
vcgt_f16(float16x4_t __p0,float16x4_t __p1)34134 __ai __attribute__((target("fullfp16"))) uint16x4_t vcgt_f16(float16x4_t __p0, float16x4_t __p1) {
34135 uint16x4_t __ret;
34136 __ret = (uint16x4_t)(__p0 > __p1);
34137 return __ret;
34138 }
34139 #else
vcgt_f16(float16x4_t __p0,float16x4_t __p1)34140 __ai __attribute__((target("fullfp16"))) uint16x4_t vcgt_f16(float16x4_t __p0, float16x4_t __p1) {
34141 uint16x4_t __ret;
34142 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34143 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
34144 __ret = (uint16x4_t)(__rev0 > __rev1);
34145 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
34146 return __ret;
34147 }
34148 #endif
34149
34150 #ifdef __LITTLE_ENDIAN__
vcgtzq_f16(float16x8_t __p0)34151 __ai __attribute__((target("fullfp16"))) uint16x8_t vcgtzq_f16(float16x8_t __p0) {
34152 uint16x8_t __ret;
34153 __ret = (uint16x8_t) __builtin_neon_vcgtzq_f16((int8x16_t)__p0, 49);
34154 return __ret;
34155 }
34156 #else
vcgtzq_f16(float16x8_t __p0)34157 __ai __attribute__((target("fullfp16"))) uint16x8_t vcgtzq_f16(float16x8_t __p0) {
34158 uint16x8_t __ret;
34159 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34160 __ret = (uint16x8_t) __builtin_neon_vcgtzq_f16((int8x16_t)__rev0, 49);
34161 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
34162 return __ret;
34163 }
34164 #endif
34165
34166 #ifdef __LITTLE_ENDIAN__
vcgtz_f16(float16x4_t __p0)34167 __ai __attribute__((target("fullfp16"))) uint16x4_t vcgtz_f16(float16x4_t __p0) {
34168 uint16x4_t __ret;
34169 __ret = (uint16x4_t) __builtin_neon_vcgtz_f16((int8x8_t)__p0, 17);
34170 return __ret;
34171 }
34172 #else
vcgtz_f16(float16x4_t __p0)34173 __ai __attribute__((target("fullfp16"))) uint16x4_t vcgtz_f16(float16x4_t __p0) {
34174 uint16x4_t __ret;
34175 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34176 __ret = (uint16x4_t) __builtin_neon_vcgtz_f16((int8x8_t)__rev0, 17);
34177 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
34178 return __ret;
34179 }
34180 #endif
34181
34182 #ifdef __LITTLE_ENDIAN__
vcleq_f16(float16x8_t __p0,float16x8_t __p1)34183 __ai __attribute__((target("fullfp16"))) uint16x8_t vcleq_f16(float16x8_t __p0, float16x8_t __p1) {
34184 uint16x8_t __ret;
34185 __ret = (uint16x8_t)(__p0 <= __p1);
34186 return __ret;
34187 }
34188 #else
vcleq_f16(float16x8_t __p0,float16x8_t __p1)34189 __ai __attribute__((target("fullfp16"))) uint16x8_t vcleq_f16(float16x8_t __p0, float16x8_t __p1) {
34190 uint16x8_t __ret;
34191 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34192 float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
34193 __ret = (uint16x8_t)(__rev0 <= __rev1);
34194 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
34195 return __ret;
34196 }
34197 #endif
34198
34199 #ifdef __LITTLE_ENDIAN__
vcle_f16(float16x4_t __p0,float16x4_t __p1)34200 __ai __attribute__((target("fullfp16"))) uint16x4_t vcle_f16(float16x4_t __p0, float16x4_t __p1) {
34201 uint16x4_t __ret;
34202 __ret = (uint16x4_t)(__p0 <= __p1);
34203 return __ret;
34204 }
34205 #else
vcle_f16(float16x4_t __p0,float16x4_t __p1)34206 __ai __attribute__((target("fullfp16"))) uint16x4_t vcle_f16(float16x4_t __p0, float16x4_t __p1) {
34207 uint16x4_t __ret;
34208 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34209 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
34210 __ret = (uint16x4_t)(__rev0 <= __rev1);
34211 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
34212 return __ret;
34213 }
34214 #endif
34215
34216 #ifdef __LITTLE_ENDIAN__
vclezq_f16(float16x8_t __p0)34217 __ai __attribute__((target("fullfp16"))) uint16x8_t vclezq_f16(float16x8_t __p0) {
34218 uint16x8_t __ret;
34219 __ret = (uint16x8_t) __builtin_neon_vclezq_f16((int8x16_t)__p0, 49);
34220 return __ret;
34221 }
34222 #else
vclezq_f16(float16x8_t __p0)34223 __ai __attribute__((target("fullfp16"))) uint16x8_t vclezq_f16(float16x8_t __p0) {
34224 uint16x8_t __ret;
34225 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34226 __ret = (uint16x8_t) __builtin_neon_vclezq_f16((int8x16_t)__rev0, 49);
34227 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
34228 return __ret;
34229 }
34230 #endif
34231
34232 #ifdef __LITTLE_ENDIAN__
vclez_f16(float16x4_t __p0)34233 __ai __attribute__((target("fullfp16"))) uint16x4_t vclez_f16(float16x4_t __p0) {
34234 uint16x4_t __ret;
34235 __ret = (uint16x4_t) __builtin_neon_vclez_f16((int8x8_t)__p0, 17);
34236 return __ret;
34237 }
34238 #else
vclez_f16(float16x4_t __p0)34239 __ai __attribute__((target("fullfp16"))) uint16x4_t vclez_f16(float16x4_t __p0) {
34240 uint16x4_t __ret;
34241 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34242 __ret = (uint16x4_t) __builtin_neon_vclez_f16((int8x8_t)__rev0, 17);
34243 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
34244 return __ret;
34245 }
34246 #endif
34247
34248 #ifdef __LITTLE_ENDIAN__
vcltq_f16(float16x8_t __p0,float16x8_t __p1)34249 __ai __attribute__((target("fullfp16"))) uint16x8_t vcltq_f16(float16x8_t __p0, float16x8_t __p1) {
34250 uint16x8_t __ret;
34251 __ret = (uint16x8_t)(__p0 < __p1);
34252 return __ret;
34253 }
34254 #else
vcltq_f16(float16x8_t __p0,float16x8_t __p1)34255 __ai __attribute__((target("fullfp16"))) uint16x8_t vcltq_f16(float16x8_t __p0, float16x8_t __p1) {
34256 uint16x8_t __ret;
34257 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34258 float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
34259 __ret = (uint16x8_t)(__rev0 < __rev1);
34260 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
34261 return __ret;
34262 }
34263 #endif
34264
34265 #ifdef __LITTLE_ENDIAN__
vclt_f16(float16x4_t __p0,float16x4_t __p1)34266 __ai __attribute__((target("fullfp16"))) uint16x4_t vclt_f16(float16x4_t __p0, float16x4_t __p1) {
34267 uint16x4_t __ret;
34268 __ret = (uint16x4_t)(__p0 < __p1);
34269 return __ret;
34270 }
34271 #else
vclt_f16(float16x4_t __p0,float16x4_t __p1)34272 __ai __attribute__((target("fullfp16"))) uint16x4_t vclt_f16(float16x4_t __p0, float16x4_t __p1) {
34273 uint16x4_t __ret;
34274 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34275 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
34276 __ret = (uint16x4_t)(__rev0 < __rev1);
34277 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
34278 return __ret;
34279 }
34280 #endif
34281
34282 #ifdef __LITTLE_ENDIAN__
vcltzq_f16(float16x8_t __p0)34283 __ai __attribute__((target("fullfp16"))) uint16x8_t vcltzq_f16(float16x8_t __p0) {
34284 uint16x8_t __ret;
34285 __ret = (uint16x8_t) __builtin_neon_vcltzq_f16((int8x16_t)__p0, 49);
34286 return __ret;
34287 }
34288 #else
vcltzq_f16(float16x8_t __p0)34289 __ai __attribute__((target("fullfp16"))) uint16x8_t vcltzq_f16(float16x8_t __p0) {
34290 uint16x8_t __ret;
34291 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34292 __ret = (uint16x8_t) __builtin_neon_vcltzq_f16((int8x16_t)__rev0, 49);
34293 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
34294 return __ret;
34295 }
34296 #endif
34297
34298 #ifdef __LITTLE_ENDIAN__
vcltz_f16(float16x4_t __p0)34299 __ai __attribute__((target("fullfp16"))) uint16x4_t vcltz_f16(float16x4_t __p0) {
34300 uint16x4_t __ret;
34301 __ret = (uint16x4_t) __builtin_neon_vcltz_f16((int8x8_t)__p0, 17);
34302 return __ret;
34303 }
34304 #else
vcltz_f16(float16x4_t __p0)34305 __ai __attribute__((target("fullfp16"))) uint16x4_t vcltz_f16(float16x4_t __p0) {
34306 uint16x4_t __ret;
34307 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34308 __ret = (uint16x4_t) __builtin_neon_vcltz_f16((int8x8_t)__rev0, 17);
34309 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
34310 return __ret;
34311 }
34312 #endif
34313
34314 #ifdef __LITTLE_ENDIAN__
vcvtq_f16_u16(uint16x8_t __p0)34315 __ai __attribute__((target("fullfp16"))) float16x8_t vcvtq_f16_u16(uint16x8_t __p0) {
34316 float16x8_t __ret;
34317 __ret = (float16x8_t) __builtin_neon_vcvtq_f16_u16((int8x16_t)__p0, 49);
34318 return __ret;
34319 }
34320 #else
vcvtq_f16_u16(uint16x8_t __p0)34321 __ai __attribute__((target("fullfp16"))) float16x8_t vcvtq_f16_u16(uint16x8_t __p0) {
34322 float16x8_t __ret;
34323 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34324 __ret = (float16x8_t) __builtin_neon_vcvtq_f16_u16((int8x16_t)__rev0, 49);
34325 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
34326 return __ret;
34327 }
34328 #endif
34329
34330 #ifdef __LITTLE_ENDIAN__
vcvtq_f16_s16(int16x8_t __p0)34331 __ai __attribute__((target("fullfp16"))) float16x8_t vcvtq_f16_s16(int16x8_t __p0) {
34332 float16x8_t __ret;
34333 __ret = (float16x8_t) __builtin_neon_vcvtq_f16_s16((int8x16_t)__p0, 33);
34334 return __ret;
34335 }
34336 #else
vcvtq_f16_s16(int16x8_t __p0)34337 __ai __attribute__((target("fullfp16"))) float16x8_t vcvtq_f16_s16(int16x8_t __p0) {
34338 float16x8_t __ret;
34339 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34340 __ret = (float16x8_t) __builtin_neon_vcvtq_f16_s16((int8x16_t)__rev0, 33);
34341 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
34342 return __ret;
34343 }
34344 #endif
34345
34346 #ifdef __LITTLE_ENDIAN__
vcvt_f16_u16(uint16x4_t __p0)34347 __ai __attribute__((target("fullfp16"))) float16x4_t vcvt_f16_u16(uint16x4_t __p0) {
34348 float16x4_t __ret;
34349 __ret = (float16x4_t) __builtin_neon_vcvt_f16_u16((int8x8_t)__p0, 17);
34350 return __ret;
34351 }
34352 #else
vcvt_f16_u16(uint16x4_t __p0)34353 __ai __attribute__((target("fullfp16"))) float16x4_t vcvt_f16_u16(uint16x4_t __p0) {
34354 float16x4_t __ret;
34355 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34356 __ret = (float16x4_t) __builtin_neon_vcvt_f16_u16((int8x8_t)__rev0, 17);
34357 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
34358 return __ret;
34359 }
34360 #endif
34361
34362 #ifdef __LITTLE_ENDIAN__
vcvt_f16_s16(int16x4_t __p0)34363 __ai __attribute__((target("fullfp16"))) float16x4_t vcvt_f16_s16(int16x4_t __p0) {
34364 float16x4_t __ret;
34365 __ret = (float16x4_t) __builtin_neon_vcvt_f16_s16((int8x8_t)__p0, 1);
34366 return __ret;
34367 }
34368 #else
vcvt_f16_s16(int16x4_t __p0)34369 __ai __attribute__((target("fullfp16"))) float16x4_t vcvt_f16_s16(int16x4_t __p0) {
34370 float16x4_t __ret;
34371 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34372 __ret = (float16x4_t) __builtin_neon_vcvt_f16_s16((int8x8_t)__rev0, 1);
34373 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
34374 return __ret;
34375 }
34376 #endif
34377
34378 #ifdef __LITTLE_ENDIAN__
34379 #define vcvtq_n_f16_u16(__p0, __p1) __extension__ ({ \
34380 float16x8_t __ret; \
34381 uint16x8_t __s0 = __p0; \
34382 __ret = (float16x8_t) __builtin_neon_vcvtq_n_f16_u16((int8x16_t)__s0, __p1, 49); \
34383 __ret; \
34384 })
34385 #else
34386 #define vcvtq_n_f16_u16(__p0, __p1) __extension__ ({ \
34387 float16x8_t __ret; \
34388 uint16x8_t __s0 = __p0; \
34389 uint16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
34390 __ret = (float16x8_t) __builtin_neon_vcvtq_n_f16_u16((int8x16_t)__rev0, __p1, 49); \
34391 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
34392 __ret; \
34393 })
34394 #endif
34395
34396 #ifdef __LITTLE_ENDIAN__
34397 #define vcvtq_n_f16_s16(__p0, __p1) __extension__ ({ \
34398 float16x8_t __ret; \
34399 int16x8_t __s0 = __p0; \
34400 __ret = (float16x8_t) __builtin_neon_vcvtq_n_f16_s16((int8x16_t)__s0, __p1, 33); \
34401 __ret; \
34402 })
34403 #else
34404 #define vcvtq_n_f16_s16(__p0, __p1) __extension__ ({ \
34405 float16x8_t __ret; \
34406 int16x8_t __s0 = __p0; \
34407 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
34408 __ret = (float16x8_t) __builtin_neon_vcvtq_n_f16_s16((int8x16_t)__rev0, __p1, 33); \
34409 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
34410 __ret; \
34411 })
34412 #endif
34413
34414 #ifdef __LITTLE_ENDIAN__
34415 #define vcvt_n_f16_u16(__p0, __p1) __extension__ ({ \
34416 float16x4_t __ret; \
34417 uint16x4_t __s0 = __p0; \
34418 __ret = (float16x4_t) __builtin_neon_vcvt_n_f16_u16((int8x8_t)__s0, __p1, 17); \
34419 __ret; \
34420 })
34421 #else
34422 #define vcvt_n_f16_u16(__p0, __p1) __extension__ ({ \
34423 float16x4_t __ret; \
34424 uint16x4_t __s0 = __p0; \
34425 uint16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
34426 __ret = (float16x4_t) __builtin_neon_vcvt_n_f16_u16((int8x8_t)__rev0, __p1, 17); \
34427 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
34428 __ret; \
34429 })
34430 #endif
34431
34432 #ifdef __LITTLE_ENDIAN__
34433 #define vcvt_n_f16_s16(__p0, __p1) __extension__ ({ \
34434 float16x4_t __ret; \
34435 int16x4_t __s0 = __p0; \
34436 __ret = (float16x4_t) __builtin_neon_vcvt_n_f16_s16((int8x8_t)__s0, __p1, 1); \
34437 __ret; \
34438 })
34439 #else
34440 #define vcvt_n_f16_s16(__p0, __p1) __extension__ ({ \
34441 float16x4_t __ret; \
34442 int16x4_t __s0 = __p0; \
34443 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
34444 __ret = (float16x4_t) __builtin_neon_vcvt_n_f16_s16((int8x8_t)__rev0, __p1, 1); \
34445 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
34446 __ret; \
34447 })
34448 #endif
34449
34450 #ifdef __LITTLE_ENDIAN__
34451 #define vcvtq_n_s16_f16(__p0, __p1) __extension__ ({ \
34452 int16x8_t __ret; \
34453 float16x8_t __s0 = __p0; \
34454 __ret = (int16x8_t) __builtin_neon_vcvtq_n_s16_f16((int8x16_t)__s0, __p1, 33); \
34455 __ret; \
34456 })
34457 #else
34458 #define vcvtq_n_s16_f16(__p0, __p1) __extension__ ({ \
34459 int16x8_t __ret; \
34460 float16x8_t __s0 = __p0; \
34461 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
34462 __ret = (int16x8_t) __builtin_neon_vcvtq_n_s16_f16((int8x16_t)__rev0, __p1, 33); \
34463 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
34464 __ret; \
34465 })
34466 #endif
34467
34468 #ifdef __LITTLE_ENDIAN__
34469 #define vcvt_n_s16_f16(__p0, __p1) __extension__ ({ \
34470 int16x4_t __ret; \
34471 float16x4_t __s0 = __p0; \
34472 __ret = (int16x4_t) __builtin_neon_vcvt_n_s16_f16((int8x8_t)__s0, __p1, 1); \
34473 __ret; \
34474 })
34475 #else
34476 #define vcvt_n_s16_f16(__p0, __p1) __extension__ ({ \
34477 int16x4_t __ret; \
34478 float16x4_t __s0 = __p0; \
34479 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
34480 __ret = (int16x4_t) __builtin_neon_vcvt_n_s16_f16((int8x8_t)__rev0, __p1, 1); \
34481 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
34482 __ret; \
34483 })
34484 #endif
34485
34486 #ifdef __LITTLE_ENDIAN__
34487 #define vcvtq_n_u16_f16(__p0, __p1) __extension__ ({ \
34488 uint16x8_t __ret; \
34489 float16x8_t __s0 = __p0; \
34490 __ret = (uint16x8_t) __builtin_neon_vcvtq_n_u16_f16((int8x16_t)__s0, __p1, 49); \
34491 __ret; \
34492 })
34493 #else
34494 #define vcvtq_n_u16_f16(__p0, __p1) __extension__ ({ \
34495 uint16x8_t __ret; \
34496 float16x8_t __s0 = __p0; \
34497 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
34498 __ret = (uint16x8_t) __builtin_neon_vcvtq_n_u16_f16((int8x16_t)__rev0, __p1, 49); \
34499 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
34500 __ret; \
34501 })
34502 #endif
34503
34504 #ifdef __LITTLE_ENDIAN__
34505 #define vcvt_n_u16_f16(__p0, __p1) __extension__ ({ \
34506 uint16x4_t __ret; \
34507 float16x4_t __s0 = __p0; \
34508 __ret = (uint16x4_t) __builtin_neon_vcvt_n_u16_f16((int8x8_t)__s0, __p1, 17); \
34509 __ret; \
34510 })
34511 #else
34512 #define vcvt_n_u16_f16(__p0, __p1) __extension__ ({ \
34513 uint16x4_t __ret; \
34514 float16x4_t __s0 = __p0; \
34515 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
34516 __ret = (uint16x4_t) __builtin_neon_vcvt_n_u16_f16((int8x8_t)__rev0, __p1, 17); \
34517 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
34518 __ret; \
34519 })
34520 #endif
34521
34522 #ifdef __LITTLE_ENDIAN__
vcvtq_s16_f16(float16x8_t __p0)34523 __ai __attribute__((target("fullfp16"))) int16x8_t vcvtq_s16_f16(float16x8_t __p0) {
34524 int16x8_t __ret;
34525 __ret = (int16x8_t) __builtin_neon_vcvtq_s16_f16((int8x16_t)__p0, 33);
34526 return __ret;
34527 }
34528 #else
vcvtq_s16_f16(float16x8_t __p0)34529 __ai __attribute__((target("fullfp16"))) int16x8_t vcvtq_s16_f16(float16x8_t __p0) {
34530 int16x8_t __ret;
34531 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34532 __ret = (int16x8_t) __builtin_neon_vcvtq_s16_f16((int8x16_t)__rev0, 33);
34533 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
34534 return __ret;
34535 }
34536 #endif
34537
34538 #ifdef __LITTLE_ENDIAN__
vcvt_s16_f16(float16x4_t __p0)34539 __ai __attribute__((target("fullfp16"))) int16x4_t vcvt_s16_f16(float16x4_t __p0) {
34540 int16x4_t __ret;
34541 __ret = (int16x4_t) __builtin_neon_vcvt_s16_f16((int8x8_t)__p0, 1);
34542 return __ret;
34543 }
34544 #else
vcvt_s16_f16(float16x4_t __p0)34545 __ai __attribute__((target("fullfp16"))) int16x4_t vcvt_s16_f16(float16x4_t __p0) {
34546 int16x4_t __ret;
34547 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34548 __ret = (int16x4_t) __builtin_neon_vcvt_s16_f16((int8x8_t)__rev0, 1);
34549 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
34550 return __ret;
34551 }
34552 #endif
34553
34554 #ifdef __LITTLE_ENDIAN__
vcvtq_u16_f16(float16x8_t __p0)34555 __ai __attribute__((target("fullfp16"))) uint16x8_t vcvtq_u16_f16(float16x8_t __p0) {
34556 uint16x8_t __ret;
34557 __ret = (uint16x8_t) __builtin_neon_vcvtq_u16_f16((int8x16_t)__p0, 49);
34558 return __ret;
34559 }
34560 #else
vcvtq_u16_f16(float16x8_t __p0)34561 __ai __attribute__((target("fullfp16"))) uint16x8_t vcvtq_u16_f16(float16x8_t __p0) {
34562 uint16x8_t __ret;
34563 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34564 __ret = (uint16x8_t) __builtin_neon_vcvtq_u16_f16((int8x16_t)__rev0, 49);
34565 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
34566 return __ret;
34567 }
34568 #endif
34569
34570 #ifdef __LITTLE_ENDIAN__
vcvt_u16_f16(float16x4_t __p0)34571 __ai __attribute__((target("fullfp16"))) uint16x4_t vcvt_u16_f16(float16x4_t __p0) {
34572 uint16x4_t __ret;
34573 __ret = (uint16x4_t) __builtin_neon_vcvt_u16_f16((int8x8_t)__p0, 17);
34574 return __ret;
34575 }
34576 #else
vcvt_u16_f16(float16x4_t __p0)34577 __ai __attribute__((target("fullfp16"))) uint16x4_t vcvt_u16_f16(float16x4_t __p0) {
34578 uint16x4_t __ret;
34579 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34580 __ret = (uint16x4_t) __builtin_neon_vcvt_u16_f16((int8x8_t)__rev0, 17);
34581 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
34582 return __ret;
34583 }
34584 #endif
34585
34586 #ifdef __LITTLE_ENDIAN__
vcvtaq_s16_f16(float16x8_t __p0)34587 __ai __attribute__((target("fullfp16"))) int16x8_t vcvtaq_s16_f16(float16x8_t __p0) {
34588 int16x8_t __ret;
34589 __ret = (int16x8_t) __builtin_neon_vcvtaq_s16_f16((int8x16_t)__p0, 33);
34590 return __ret;
34591 }
34592 #else
vcvtaq_s16_f16(float16x8_t __p0)34593 __ai __attribute__((target("fullfp16"))) int16x8_t vcvtaq_s16_f16(float16x8_t __p0) {
34594 int16x8_t __ret;
34595 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34596 __ret = (int16x8_t) __builtin_neon_vcvtaq_s16_f16((int8x16_t)__rev0, 33);
34597 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
34598 return __ret;
34599 }
34600 #endif
34601
34602 #ifdef __LITTLE_ENDIAN__
vcvta_s16_f16(float16x4_t __p0)34603 __ai __attribute__((target("fullfp16"))) int16x4_t vcvta_s16_f16(float16x4_t __p0) {
34604 int16x4_t __ret;
34605 __ret = (int16x4_t) __builtin_neon_vcvta_s16_f16((int8x8_t)__p0, 1);
34606 return __ret;
34607 }
34608 #else
vcvta_s16_f16(float16x4_t __p0)34609 __ai __attribute__((target("fullfp16"))) int16x4_t vcvta_s16_f16(float16x4_t __p0) {
34610 int16x4_t __ret;
34611 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34612 __ret = (int16x4_t) __builtin_neon_vcvta_s16_f16((int8x8_t)__rev0, 1);
34613 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
34614 return __ret;
34615 }
34616 #endif
34617
34618 #ifdef __LITTLE_ENDIAN__
vcvtaq_u16_f16(float16x8_t __p0)34619 __ai __attribute__((target("fullfp16"))) uint16x8_t vcvtaq_u16_f16(float16x8_t __p0) {
34620 uint16x8_t __ret;
34621 __ret = (uint16x8_t) __builtin_neon_vcvtaq_u16_f16((int8x16_t)__p0, 49);
34622 return __ret;
34623 }
34624 #else
vcvtaq_u16_f16(float16x8_t __p0)34625 __ai __attribute__((target("fullfp16"))) uint16x8_t vcvtaq_u16_f16(float16x8_t __p0) {
34626 uint16x8_t __ret;
34627 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34628 __ret = (uint16x8_t) __builtin_neon_vcvtaq_u16_f16((int8x16_t)__rev0, 49);
34629 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
34630 return __ret;
34631 }
34632 #endif
34633
34634 #ifdef __LITTLE_ENDIAN__
vcvta_u16_f16(float16x4_t __p0)34635 __ai __attribute__((target("fullfp16"))) uint16x4_t vcvta_u16_f16(float16x4_t __p0) {
34636 uint16x4_t __ret;
34637 __ret = (uint16x4_t) __builtin_neon_vcvta_u16_f16((int8x8_t)__p0, 17);
34638 return __ret;
34639 }
34640 #else
vcvta_u16_f16(float16x4_t __p0)34641 __ai __attribute__((target("fullfp16"))) uint16x4_t vcvta_u16_f16(float16x4_t __p0) {
34642 uint16x4_t __ret;
34643 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34644 __ret = (uint16x4_t) __builtin_neon_vcvta_u16_f16((int8x8_t)__rev0, 17);
34645 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
34646 return __ret;
34647 }
34648 #endif
34649
34650 #ifdef __LITTLE_ENDIAN__
vcvtmq_s16_f16(float16x8_t __p0)34651 __ai __attribute__((target("fullfp16"))) int16x8_t vcvtmq_s16_f16(float16x8_t __p0) {
34652 int16x8_t __ret;
34653 __ret = (int16x8_t) __builtin_neon_vcvtmq_s16_f16((int8x16_t)__p0, 33);
34654 return __ret;
34655 }
34656 #else
vcvtmq_s16_f16(float16x8_t __p0)34657 __ai __attribute__((target("fullfp16"))) int16x8_t vcvtmq_s16_f16(float16x8_t __p0) {
34658 int16x8_t __ret;
34659 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34660 __ret = (int16x8_t) __builtin_neon_vcvtmq_s16_f16((int8x16_t)__rev0, 33);
34661 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
34662 return __ret;
34663 }
34664 #endif
34665
34666 #ifdef __LITTLE_ENDIAN__
vcvtm_s16_f16(float16x4_t __p0)34667 __ai __attribute__((target("fullfp16"))) int16x4_t vcvtm_s16_f16(float16x4_t __p0) {
34668 int16x4_t __ret;
34669 __ret = (int16x4_t) __builtin_neon_vcvtm_s16_f16((int8x8_t)__p0, 1);
34670 return __ret;
34671 }
34672 #else
vcvtm_s16_f16(float16x4_t __p0)34673 __ai __attribute__((target("fullfp16"))) int16x4_t vcvtm_s16_f16(float16x4_t __p0) {
34674 int16x4_t __ret;
34675 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34676 __ret = (int16x4_t) __builtin_neon_vcvtm_s16_f16((int8x8_t)__rev0, 1);
34677 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
34678 return __ret;
34679 }
34680 #endif
34681
34682 #ifdef __LITTLE_ENDIAN__
vcvtmq_u16_f16(float16x8_t __p0)34683 __ai __attribute__((target("fullfp16"))) uint16x8_t vcvtmq_u16_f16(float16x8_t __p0) {
34684 uint16x8_t __ret;
34685 __ret = (uint16x8_t) __builtin_neon_vcvtmq_u16_f16((int8x16_t)__p0, 49);
34686 return __ret;
34687 }
34688 #else
vcvtmq_u16_f16(float16x8_t __p0)34689 __ai __attribute__((target("fullfp16"))) uint16x8_t vcvtmq_u16_f16(float16x8_t __p0) {
34690 uint16x8_t __ret;
34691 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34692 __ret = (uint16x8_t) __builtin_neon_vcvtmq_u16_f16((int8x16_t)__rev0, 49);
34693 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
34694 return __ret;
34695 }
34696 #endif
34697
34698 #ifdef __LITTLE_ENDIAN__
vcvtm_u16_f16(float16x4_t __p0)34699 __ai __attribute__((target("fullfp16"))) uint16x4_t vcvtm_u16_f16(float16x4_t __p0) {
34700 uint16x4_t __ret;
34701 __ret = (uint16x4_t) __builtin_neon_vcvtm_u16_f16((int8x8_t)__p0, 17);
34702 return __ret;
34703 }
34704 #else
vcvtm_u16_f16(float16x4_t __p0)34705 __ai __attribute__((target("fullfp16"))) uint16x4_t vcvtm_u16_f16(float16x4_t __p0) {
34706 uint16x4_t __ret;
34707 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34708 __ret = (uint16x4_t) __builtin_neon_vcvtm_u16_f16((int8x8_t)__rev0, 17);
34709 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
34710 return __ret;
34711 }
34712 #endif
34713
34714 #ifdef __LITTLE_ENDIAN__
vcvtnq_s16_f16(float16x8_t __p0)34715 __ai __attribute__((target("fullfp16"))) int16x8_t vcvtnq_s16_f16(float16x8_t __p0) {
34716 int16x8_t __ret;
34717 __ret = (int16x8_t) __builtin_neon_vcvtnq_s16_f16((int8x16_t)__p0, 33);
34718 return __ret;
34719 }
34720 #else
vcvtnq_s16_f16(float16x8_t __p0)34721 __ai __attribute__((target("fullfp16"))) int16x8_t vcvtnq_s16_f16(float16x8_t __p0) {
34722 int16x8_t __ret;
34723 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34724 __ret = (int16x8_t) __builtin_neon_vcvtnq_s16_f16((int8x16_t)__rev0, 33);
34725 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
34726 return __ret;
34727 }
34728 #endif
34729
34730 #ifdef __LITTLE_ENDIAN__
vcvtn_s16_f16(float16x4_t __p0)34731 __ai __attribute__((target("fullfp16"))) int16x4_t vcvtn_s16_f16(float16x4_t __p0) {
34732 int16x4_t __ret;
34733 __ret = (int16x4_t) __builtin_neon_vcvtn_s16_f16((int8x8_t)__p0, 1);
34734 return __ret;
34735 }
34736 #else
vcvtn_s16_f16(float16x4_t __p0)34737 __ai __attribute__((target("fullfp16"))) int16x4_t vcvtn_s16_f16(float16x4_t __p0) {
34738 int16x4_t __ret;
34739 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34740 __ret = (int16x4_t) __builtin_neon_vcvtn_s16_f16((int8x8_t)__rev0, 1);
34741 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
34742 return __ret;
34743 }
34744 #endif
34745
34746 #ifdef __LITTLE_ENDIAN__
vcvtnq_u16_f16(float16x8_t __p0)34747 __ai __attribute__((target("fullfp16"))) uint16x8_t vcvtnq_u16_f16(float16x8_t __p0) {
34748 uint16x8_t __ret;
34749 __ret = (uint16x8_t) __builtin_neon_vcvtnq_u16_f16((int8x16_t)__p0, 49);
34750 return __ret;
34751 }
34752 #else
vcvtnq_u16_f16(float16x8_t __p0)34753 __ai __attribute__((target("fullfp16"))) uint16x8_t vcvtnq_u16_f16(float16x8_t __p0) {
34754 uint16x8_t __ret;
34755 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34756 __ret = (uint16x8_t) __builtin_neon_vcvtnq_u16_f16((int8x16_t)__rev0, 49);
34757 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
34758 return __ret;
34759 }
34760 #endif
34761
34762 #ifdef __LITTLE_ENDIAN__
vcvtn_u16_f16(float16x4_t __p0)34763 __ai __attribute__((target("fullfp16"))) uint16x4_t vcvtn_u16_f16(float16x4_t __p0) {
34764 uint16x4_t __ret;
34765 __ret = (uint16x4_t) __builtin_neon_vcvtn_u16_f16((int8x8_t)__p0, 17);
34766 return __ret;
34767 }
34768 #else
vcvtn_u16_f16(float16x4_t __p0)34769 __ai __attribute__((target("fullfp16"))) uint16x4_t vcvtn_u16_f16(float16x4_t __p0) {
34770 uint16x4_t __ret;
34771 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34772 __ret = (uint16x4_t) __builtin_neon_vcvtn_u16_f16((int8x8_t)__rev0, 17);
34773 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
34774 return __ret;
34775 }
34776 #endif
34777
34778 #ifdef __LITTLE_ENDIAN__
vcvtpq_s16_f16(float16x8_t __p0)34779 __ai __attribute__((target("fullfp16"))) int16x8_t vcvtpq_s16_f16(float16x8_t __p0) {
34780 int16x8_t __ret;
34781 __ret = (int16x8_t) __builtin_neon_vcvtpq_s16_f16((int8x16_t)__p0, 33);
34782 return __ret;
34783 }
34784 #else
vcvtpq_s16_f16(float16x8_t __p0)34785 __ai __attribute__((target("fullfp16"))) int16x8_t vcvtpq_s16_f16(float16x8_t __p0) {
34786 int16x8_t __ret;
34787 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34788 __ret = (int16x8_t) __builtin_neon_vcvtpq_s16_f16((int8x16_t)__rev0, 33);
34789 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
34790 return __ret;
34791 }
34792 #endif
34793
34794 #ifdef __LITTLE_ENDIAN__
vcvtp_s16_f16(float16x4_t __p0)34795 __ai __attribute__((target("fullfp16"))) int16x4_t vcvtp_s16_f16(float16x4_t __p0) {
34796 int16x4_t __ret;
34797 __ret = (int16x4_t) __builtin_neon_vcvtp_s16_f16((int8x8_t)__p0, 1);
34798 return __ret;
34799 }
34800 #else
vcvtp_s16_f16(float16x4_t __p0)34801 __ai __attribute__((target("fullfp16"))) int16x4_t vcvtp_s16_f16(float16x4_t __p0) {
34802 int16x4_t __ret;
34803 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34804 __ret = (int16x4_t) __builtin_neon_vcvtp_s16_f16((int8x8_t)__rev0, 1);
34805 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
34806 return __ret;
34807 }
34808 #endif
34809
34810 #ifdef __LITTLE_ENDIAN__
vcvtpq_u16_f16(float16x8_t __p0)34811 __ai __attribute__((target("fullfp16"))) uint16x8_t vcvtpq_u16_f16(float16x8_t __p0) {
34812 uint16x8_t __ret;
34813 __ret = (uint16x8_t) __builtin_neon_vcvtpq_u16_f16((int8x16_t)__p0, 49);
34814 return __ret;
34815 }
34816 #else
vcvtpq_u16_f16(float16x8_t __p0)34817 __ai __attribute__((target("fullfp16"))) uint16x8_t vcvtpq_u16_f16(float16x8_t __p0) {
34818 uint16x8_t __ret;
34819 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34820 __ret = (uint16x8_t) __builtin_neon_vcvtpq_u16_f16((int8x16_t)__rev0, 49);
34821 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
34822 return __ret;
34823 }
34824 #endif
34825
34826 #ifdef __LITTLE_ENDIAN__
vcvtp_u16_f16(float16x4_t __p0)34827 __ai __attribute__((target("fullfp16"))) uint16x4_t vcvtp_u16_f16(float16x4_t __p0) {
34828 uint16x4_t __ret;
34829 __ret = (uint16x4_t) __builtin_neon_vcvtp_u16_f16((int8x8_t)__p0, 17);
34830 return __ret;
34831 }
34832 #else
vcvtp_u16_f16(float16x4_t __p0)34833 __ai __attribute__((target("fullfp16"))) uint16x4_t vcvtp_u16_f16(float16x4_t __p0) {
34834 uint16x4_t __ret;
34835 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34836 __ret = (uint16x4_t) __builtin_neon_vcvtp_u16_f16((int8x8_t)__rev0, 17);
34837 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
34838 return __ret;
34839 }
34840 #endif
34841
34842 #ifdef __LITTLE_ENDIAN__
34843 #define vextq_f16(__p0, __p1, __p2) __extension__ ({ \
34844 float16x8_t __ret; \
34845 float16x8_t __s0 = __p0; \
34846 float16x8_t __s1 = __p1; \
34847 __ret = (float16x8_t) __builtin_neon_vextq_f16((int8x16_t)__s0, (int8x16_t)__s1, __p2, 40); \
34848 __ret; \
34849 })
34850 #else
34851 #define vextq_f16(__p0, __p1, __p2) __extension__ ({ \
34852 float16x8_t __ret; \
34853 float16x8_t __s0 = __p0; \
34854 float16x8_t __s1 = __p1; \
34855 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
34856 float16x8_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 7, 6, 5, 4, 3, 2, 1, 0); \
34857 __ret = (float16x8_t) __builtin_neon_vextq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, __p2, 40); \
34858 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
34859 __ret; \
34860 })
34861 #endif
34862
34863 #ifdef __LITTLE_ENDIAN__
34864 #define vext_f16(__p0, __p1, __p2) __extension__ ({ \
34865 float16x4_t __ret; \
34866 float16x4_t __s0 = __p0; \
34867 float16x4_t __s1 = __p1; \
34868 __ret = (float16x4_t) __builtin_neon_vext_f16((int8x8_t)__s0, (int8x8_t)__s1, __p2, 8); \
34869 __ret; \
34870 })
34871 #else
34872 #define vext_f16(__p0, __p1, __p2) __extension__ ({ \
34873 float16x4_t __ret; \
34874 float16x4_t __s0 = __p0; \
34875 float16x4_t __s1 = __p1; \
34876 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
34877 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__s1, __s1, 3, 2, 1, 0); \
34878 __ret = (float16x4_t) __builtin_neon_vext_f16((int8x8_t)__rev0, (int8x8_t)__rev1, __p2, 8); \
34879 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
34880 __ret; \
34881 })
34882 #endif
34883
34884 #ifdef __LITTLE_ENDIAN__
vfmaq_f16(float16x8_t __p0,float16x8_t __p1,float16x8_t __p2)34885 __ai __attribute__((target("fullfp16"))) float16x8_t vfmaq_f16(float16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
34886 float16x8_t __ret;
34887 __ret = (float16x8_t) __builtin_neon_vfmaq_f16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 40);
34888 return __ret;
34889 }
34890 #else
vfmaq_f16(float16x8_t __p0,float16x8_t __p1,float16x8_t __p2)34891 __ai __attribute__((target("fullfp16"))) float16x8_t vfmaq_f16(float16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
34892 float16x8_t __ret;
34893 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34894 float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
34895 float16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
34896 __ret = (float16x8_t) __builtin_neon_vfmaq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 40);
34897 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
34898 return __ret;
34899 }
__noswap_vfmaq_f16(float16x8_t __p0,float16x8_t __p1,float16x8_t __p2)34900 __ai __attribute__((target("fullfp16"))) float16x8_t __noswap_vfmaq_f16(float16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
34901 float16x8_t __ret;
34902 __ret = (float16x8_t) __builtin_neon_vfmaq_f16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 40);
34903 return __ret;
34904 }
34905 #endif
34906
34907 #ifdef __LITTLE_ENDIAN__
vfma_f16(float16x4_t __p0,float16x4_t __p1,float16x4_t __p2)34908 __ai __attribute__((target("fullfp16"))) float16x4_t vfma_f16(float16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
34909 float16x4_t __ret;
34910 __ret = (float16x4_t) __builtin_neon_vfma_f16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 8);
34911 return __ret;
34912 }
34913 #else
vfma_f16(float16x4_t __p0,float16x4_t __p1,float16x4_t __p2)34914 __ai __attribute__((target("fullfp16"))) float16x4_t vfma_f16(float16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
34915 float16x4_t __ret;
34916 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34917 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
34918 float16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
34919 __ret = (float16x4_t) __builtin_neon_vfma_f16((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 8);
34920 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
34921 return __ret;
34922 }
__noswap_vfma_f16(float16x4_t __p0,float16x4_t __p1,float16x4_t __p2)34923 __ai __attribute__((target("fullfp16"))) float16x4_t __noswap_vfma_f16(float16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
34924 float16x4_t __ret;
34925 __ret = (float16x4_t) __builtin_neon_vfma_f16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 8);
34926 return __ret;
34927 }
34928 #endif
34929
34930 #ifdef __LITTLE_ENDIAN__
vfmsq_f16(float16x8_t __p0,float16x8_t __p1,float16x8_t __p2)34931 __ai __attribute__((target("fullfp16"))) float16x8_t vfmsq_f16(float16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
34932 float16x8_t __ret;
34933 __ret = vfmaq_f16(__p0, -__p1, __p2);
34934 return __ret;
34935 }
34936 #else
vfmsq_f16(float16x8_t __p0,float16x8_t __p1,float16x8_t __p2)34937 __ai __attribute__((target("fullfp16"))) float16x8_t vfmsq_f16(float16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
34938 float16x8_t __ret;
34939 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34940 float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
34941 float16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
34942 __ret = __noswap_vfmaq_f16(__rev0, -__rev1, __rev2);
34943 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
34944 return __ret;
34945 }
34946 #endif
34947
34948 #ifdef __LITTLE_ENDIAN__
vfms_f16(float16x4_t __p0,float16x4_t __p1,float16x4_t __p2)34949 __ai __attribute__((target("fullfp16"))) float16x4_t vfms_f16(float16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
34950 float16x4_t __ret;
34951 __ret = vfma_f16(__p0, -__p1, __p2);
34952 return __ret;
34953 }
34954 #else
vfms_f16(float16x4_t __p0,float16x4_t __p1,float16x4_t __p2)34955 __ai __attribute__((target("fullfp16"))) float16x4_t vfms_f16(float16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
34956 float16x4_t __ret;
34957 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34958 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
34959 float16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
34960 __ret = __noswap_vfma_f16(__rev0, -__rev1, __rev2);
34961 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
34962 return __ret;
34963 }
34964 #endif
34965
34966 #ifdef __LITTLE_ENDIAN__
vmaxq_f16(float16x8_t __p0,float16x8_t __p1)34967 __ai __attribute__((target("fullfp16"))) float16x8_t vmaxq_f16(float16x8_t __p0, float16x8_t __p1) {
34968 float16x8_t __ret;
34969 __ret = (float16x8_t) __builtin_neon_vmaxq_f16((int8x16_t)__p0, (int8x16_t)__p1, 40);
34970 return __ret;
34971 }
34972 #else
vmaxq_f16(float16x8_t __p0,float16x8_t __p1)34973 __ai __attribute__((target("fullfp16"))) float16x8_t vmaxq_f16(float16x8_t __p0, float16x8_t __p1) {
34974 float16x8_t __ret;
34975 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
34976 float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
34977 __ret = (float16x8_t) __builtin_neon_vmaxq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, 40);
34978 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
34979 return __ret;
34980 }
34981 #endif
34982
34983 #ifdef __LITTLE_ENDIAN__
vmax_f16(float16x4_t __p0,float16x4_t __p1)34984 __ai __attribute__((target("fullfp16"))) float16x4_t vmax_f16(float16x4_t __p0, float16x4_t __p1) {
34985 float16x4_t __ret;
34986 __ret = (float16x4_t) __builtin_neon_vmax_f16((int8x8_t)__p0, (int8x8_t)__p1, 8);
34987 return __ret;
34988 }
34989 #else
vmax_f16(float16x4_t __p0,float16x4_t __p1)34990 __ai __attribute__((target("fullfp16"))) float16x4_t vmax_f16(float16x4_t __p0, float16x4_t __p1) {
34991 float16x4_t __ret;
34992 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
34993 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
34994 __ret = (float16x4_t) __builtin_neon_vmax_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 8);
34995 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
34996 return __ret;
34997 }
34998 #endif
34999
35000 #ifdef __LITTLE_ENDIAN__
vminq_f16(float16x8_t __p0,float16x8_t __p1)35001 __ai __attribute__((target("fullfp16"))) float16x8_t vminq_f16(float16x8_t __p0, float16x8_t __p1) {
35002 float16x8_t __ret;
35003 __ret = (float16x8_t) __builtin_neon_vminq_f16((int8x16_t)__p0, (int8x16_t)__p1, 40);
35004 return __ret;
35005 }
35006 #else
vminq_f16(float16x8_t __p0,float16x8_t __p1)35007 __ai __attribute__((target("fullfp16"))) float16x8_t vminq_f16(float16x8_t __p0, float16x8_t __p1) {
35008 float16x8_t __ret;
35009 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
35010 float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
35011 __ret = (float16x8_t) __builtin_neon_vminq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, 40);
35012 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
35013 return __ret;
35014 }
35015 #endif
35016
35017 #ifdef __LITTLE_ENDIAN__
vmin_f16(float16x4_t __p0,float16x4_t __p1)35018 __ai __attribute__((target("fullfp16"))) float16x4_t vmin_f16(float16x4_t __p0, float16x4_t __p1) {
35019 float16x4_t __ret;
35020 __ret = (float16x4_t) __builtin_neon_vmin_f16((int8x8_t)__p0, (int8x8_t)__p1, 8);
35021 return __ret;
35022 }
35023 #else
vmin_f16(float16x4_t __p0,float16x4_t __p1)35024 __ai __attribute__((target("fullfp16"))) float16x4_t vmin_f16(float16x4_t __p0, float16x4_t __p1) {
35025 float16x4_t __ret;
35026 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35027 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
35028 __ret = (float16x4_t) __builtin_neon_vmin_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 8);
35029 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
35030 return __ret;
35031 }
35032 #endif
35033
35034 #ifdef __LITTLE_ENDIAN__
vmulq_f16(float16x8_t __p0,float16x8_t __p1)35035 __ai __attribute__((target("fullfp16"))) float16x8_t vmulq_f16(float16x8_t __p0, float16x8_t __p1) {
35036 float16x8_t __ret;
35037 __ret = __p0 * __p1;
35038 return __ret;
35039 }
35040 #else
vmulq_f16(float16x8_t __p0,float16x8_t __p1)35041 __ai __attribute__((target("fullfp16"))) float16x8_t vmulq_f16(float16x8_t __p0, float16x8_t __p1) {
35042 float16x8_t __ret;
35043 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
35044 float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
35045 __ret = __rev0 * __rev1;
35046 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
35047 return __ret;
35048 }
35049 #endif
35050
35051 #ifdef __LITTLE_ENDIAN__
vmul_f16(float16x4_t __p0,float16x4_t __p1)35052 __ai __attribute__((target("fullfp16"))) float16x4_t vmul_f16(float16x4_t __p0, float16x4_t __p1) {
35053 float16x4_t __ret;
35054 __ret = __p0 * __p1;
35055 return __ret;
35056 }
35057 #else
vmul_f16(float16x4_t __p0,float16x4_t __p1)35058 __ai __attribute__((target("fullfp16"))) float16x4_t vmul_f16(float16x4_t __p0, float16x4_t __p1) {
35059 float16x4_t __ret;
35060 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35061 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
35062 __ret = __rev0 * __rev1;
35063 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
35064 return __ret;
35065 }
35066 #endif
35067
35068 #ifdef __LITTLE_ENDIAN__
35069 #define vmulq_lane_f16(__p0_153, __p1_153, __p2_153) __extension__ ({ \
35070 float16x8_t __ret_153; \
35071 float16x8_t __s0_153 = __p0_153; \
35072 float16x4_t __s1_153 = __p1_153; \
35073 __ret_153 = __s0_153 * splatq_lane_f16(__s1_153, __p2_153); \
35074 __ret_153; \
35075 })
35076 #else
35077 #define vmulq_lane_f16(__p0_154, __p1_154, __p2_154) __extension__ ({ \
35078 float16x8_t __ret_154; \
35079 float16x8_t __s0_154 = __p0_154; \
35080 float16x4_t __s1_154 = __p1_154; \
35081 float16x8_t __rev0_154; __rev0_154 = __builtin_shufflevector(__s0_154, __s0_154, 7, 6, 5, 4, 3, 2, 1, 0); \
35082 float16x4_t __rev1_154; __rev1_154 = __builtin_shufflevector(__s1_154, __s1_154, 3, 2, 1, 0); \
35083 __ret_154 = __rev0_154 * __noswap_splatq_lane_f16(__rev1_154, __p2_154); \
35084 __ret_154 = __builtin_shufflevector(__ret_154, __ret_154, 7, 6, 5, 4, 3, 2, 1, 0); \
35085 __ret_154; \
35086 })
35087 #endif
35088
35089 #ifdef __LITTLE_ENDIAN__
35090 #define vmul_lane_f16(__p0_155, __p1_155, __p2_155) __extension__ ({ \
35091 float16x4_t __ret_155; \
35092 float16x4_t __s0_155 = __p0_155; \
35093 float16x4_t __s1_155 = __p1_155; \
35094 __ret_155 = __s0_155 * splat_lane_f16(__s1_155, __p2_155); \
35095 __ret_155; \
35096 })
35097 #else
35098 #define vmul_lane_f16(__p0_156, __p1_156, __p2_156) __extension__ ({ \
35099 float16x4_t __ret_156; \
35100 float16x4_t __s0_156 = __p0_156; \
35101 float16x4_t __s1_156 = __p1_156; \
35102 float16x4_t __rev0_156; __rev0_156 = __builtin_shufflevector(__s0_156, __s0_156, 3, 2, 1, 0); \
35103 float16x4_t __rev1_156; __rev1_156 = __builtin_shufflevector(__s1_156, __s1_156, 3, 2, 1, 0); \
35104 __ret_156 = __rev0_156 * __noswap_splat_lane_f16(__rev1_156, __p2_156); \
35105 __ret_156 = __builtin_shufflevector(__ret_156, __ret_156, 3, 2, 1, 0); \
35106 __ret_156; \
35107 })
35108 #endif
35109
35110 #ifdef __LITTLE_ENDIAN__
35111 #define vmulq_n_f16(__p0, __p1) __extension__ ({ \
35112 float16x8_t __ret; \
35113 float16x8_t __s0 = __p0; \
35114 float16_t __s1 = __p1; \
35115 __ret = __s0 * (float16x8_t) {__s1, __s1, __s1, __s1, __s1, __s1, __s1, __s1}; \
35116 __ret; \
35117 })
35118 #else
35119 #define vmulq_n_f16(__p0, __p1) __extension__ ({ \
35120 float16x8_t __ret; \
35121 float16x8_t __s0 = __p0; \
35122 float16_t __s1 = __p1; \
35123 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 7, 6, 5, 4, 3, 2, 1, 0); \
35124 __ret = __rev0 * (float16x8_t) {__s1, __s1, __s1, __s1, __s1, __s1, __s1, __s1}; \
35125 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0); \
35126 __ret; \
35127 })
35128 #endif
35129
35130 #ifdef __LITTLE_ENDIAN__
35131 #define vmul_n_f16(__p0, __p1) __extension__ ({ \
35132 float16x4_t __ret; \
35133 float16x4_t __s0 = __p0; \
35134 float16_t __s1 = __p1; \
35135 __ret = __s0 * (float16x4_t) {__s1, __s1, __s1, __s1}; \
35136 __ret; \
35137 })
35138 #else
35139 #define vmul_n_f16(__p0, __p1) __extension__ ({ \
35140 float16x4_t __ret; \
35141 float16x4_t __s0 = __p0; \
35142 float16_t __s1 = __p1; \
35143 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__s0, __s0, 3, 2, 1, 0); \
35144 __ret = __rev0 * (float16x4_t) {__s1, __s1, __s1, __s1}; \
35145 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0); \
35146 __ret; \
35147 })
35148 #endif
35149
35150 #ifdef __LITTLE_ENDIAN__
vnegq_f16(float16x8_t __p0)35151 __ai __attribute__((target("fullfp16"))) float16x8_t vnegq_f16(float16x8_t __p0) {
35152 float16x8_t __ret;
35153 __ret = -__p0;
35154 return __ret;
35155 }
35156 #else
vnegq_f16(float16x8_t __p0)35157 __ai __attribute__((target("fullfp16"))) float16x8_t vnegq_f16(float16x8_t __p0) {
35158 float16x8_t __ret;
35159 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
35160 __ret = -__rev0;
35161 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
35162 return __ret;
35163 }
35164 #endif
35165
35166 #ifdef __LITTLE_ENDIAN__
vneg_f16(float16x4_t __p0)35167 __ai __attribute__((target("fullfp16"))) float16x4_t vneg_f16(float16x4_t __p0) {
35168 float16x4_t __ret;
35169 __ret = -__p0;
35170 return __ret;
35171 }
35172 #else
vneg_f16(float16x4_t __p0)35173 __ai __attribute__((target("fullfp16"))) float16x4_t vneg_f16(float16x4_t __p0) {
35174 float16x4_t __ret;
35175 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35176 __ret = -__rev0;
35177 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
35178 return __ret;
35179 }
35180 #endif
35181
35182 #ifdef __LITTLE_ENDIAN__
vpadd_f16(float16x4_t __p0,float16x4_t __p1)35183 __ai __attribute__((target("fullfp16"))) float16x4_t vpadd_f16(float16x4_t __p0, float16x4_t __p1) {
35184 float16x4_t __ret;
35185 __ret = (float16x4_t) __builtin_neon_vpadd_f16((int8x8_t)__p0, (int8x8_t)__p1, 8);
35186 return __ret;
35187 }
35188 #else
vpadd_f16(float16x4_t __p0,float16x4_t __p1)35189 __ai __attribute__((target("fullfp16"))) float16x4_t vpadd_f16(float16x4_t __p0, float16x4_t __p1) {
35190 float16x4_t __ret;
35191 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35192 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
35193 __ret = (float16x4_t) __builtin_neon_vpadd_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 8);
35194 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
35195 return __ret;
35196 }
35197 #endif
35198
35199 #ifdef __LITTLE_ENDIAN__
vpmax_f16(float16x4_t __p0,float16x4_t __p1)35200 __ai __attribute__((target("fullfp16"))) float16x4_t vpmax_f16(float16x4_t __p0, float16x4_t __p1) {
35201 float16x4_t __ret;
35202 __ret = (float16x4_t) __builtin_neon_vpmax_f16((int8x8_t)__p0, (int8x8_t)__p1, 8);
35203 return __ret;
35204 }
35205 #else
vpmax_f16(float16x4_t __p0,float16x4_t __p1)35206 __ai __attribute__((target("fullfp16"))) float16x4_t vpmax_f16(float16x4_t __p0, float16x4_t __p1) {
35207 float16x4_t __ret;
35208 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35209 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
35210 __ret = (float16x4_t) __builtin_neon_vpmax_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 8);
35211 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
35212 return __ret;
35213 }
35214 #endif
35215
35216 #ifdef __LITTLE_ENDIAN__
vpmin_f16(float16x4_t __p0,float16x4_t __p1)35217 __ai __attribute__((target("fullfp16"))) float16x4_t vpmin_f16(float16x4_t __p0, float16x4_t __p1) {
35218 float16x4_t __ret;
35219 __ret = (float16x4_t) __builtin_neon_vpmin_f16((int8x8_t)__p0, (int8x8_t)__p1, 8);
35220 return __ret;
35221 }
35222 #else
vpmin_f16(float16x4_t __p0,float16x4_t __p1)35223 __ai __attribute__((target("fullfp16"))) float16x4_t vpmin_f16(float16x4_t __p0, float16x4_t __p1) {
35224 float16x4_t __ret;
35225 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35226 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
35227 __ret = (float16x4_t) __builtin_neon_vpmin_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 8);
35228 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
35229 return __ret;
35230 }
35231 #endif
35232
35233 #ifdef __LITTLE_ENDIAN__
vrecpeq_f16(float16x8_t __p0)35234 __ai __attribute__((target("fullfp16"))) float16x8_t vrecpeq_f16(float16x8_t __p0) {
35235 float16x8_t __ret;
35236 __ret = (float16x8_t) __builtin_neon_vrecpeq_f16((int8x16_t)__p0, 40);
35237 return __ret;
35238 }
35239 #else
vrecpeq_f16(float16x8_t __p0)35240 __ai __attribute__((target("fullfp16"))) float16x8_t vrecpeq_f16(float16x8_t __p0) {
35241 float16x8_t __ret;
35242 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
35243 __ret = (float16x8_t) __builtin_neon_vrecpeq_f16((int8x16_t)__rev0, 40);
35244 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
35245 return __ret;
35246 }
35247 #endif
35248
35249 #ifdef __LITTLE_ENDIAN__
vrecpe_f16(float16x4_t __p0)35250 __ai __attribute__((target("fullfp16"))) float16x4_t vrecpe_f16(float16x4_t __p0) {
35251 float16x4_t __ret;
35252 __ret = (float16x4_t) __builtin_neon_vrecpe_f16((int8x8_t)__p0, 8);
35253 return __ret;
35254 }
35255 #else
vrecpe_f16(float16x4_t __p0)35256 __ai __attribute__((target("fullfp16"))) float16x4_t vrecpe_f16(float16x4_t __p0) {
35257 float16x4_t __ret;
35258 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35259 __ret = (float16x4_t) __builtin_neon_vrecpe_f16((int8x8_t)__rev0, 8);
35260 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
35261 return __ret;
35262 }
35263 #endif
35264
35265 #ifdef __LITTLE_ENDIAN__
vrecpsq_f16(float16x8_t __p0,float16x8_t __p1)35266 __ai __attribute__((target("fullfp16"))) float16x8_t vrecpsq_f16(float16x8_t __p0, float16x8_t __p1) {
35267 float16x8_t __ret;
35268 __ret = (float16x8_t) __builtin_neon_vrecpsq_f16((int8x16_t)__p0, (int8x16_t)__p1, 40);
35269 return __ret;
35270 }
35271 #else
vrecpsq_f16(float16x8_t __p0,float16x8_t __p1)35272 __ai __attribute__((target("fullfp16"))) float16x8_t vrecpsq_f16(float16x8_t __p0, float16x8_t __p1) {
35273 float16x8_t __ret;
35274 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
35275 float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
35276 __ret = (float16x8_t) __builtin_neon_vrecpsq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, 40);
35277 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
35278 return __ret;
35279 }
35280 #endif
35281
35282 #ifdef __LITTLE_ENDIAN__
vrecps_f16(float16x4_t __p0,float16x4_t __p1)35283 __ai __attribute__((target("fullfp16"))) float16x4_t vrecps_f16(float16x4_t __p0, float16x4_t __p1) {
35284 float16x4_t __ret;
35285 __ret = (float16x4_t) __builtin_neon_vrecps_f16((int8x8_t)__p0, (int8x8_t)__p1, 8);
35286 return __ret;
35287 }
35288 #else
vrecps_f16(float16x4_t __p0,float16x4_t __p1)35289 __ai __attribute__((target("fullfp16"))) float16x4_t vrecps_f16(float16x4_t __p0, float16x4_t __p1) {
35290 float16x4_t __ret;
35291 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35292 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
35293 __ret = (float16x4_t) __builtin_neon_vrecps_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 8);
35294 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
35295 return __ret;
35296 }
35297 #endif
35298
35299 #ifdef __LITTLE_ENDIAN__
vrev64q_f16(float16x8_t __p0)35300 __ai __attribute__((target("fullfp16"))) float16x8_t vrev64q_f16(float16x8_t __p0) {
35301 float16x8_t __ret;
35302 __ret = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0, 7, 6, 5, 4);
35303 return __ret;
35304 }
35305 #else
vrev64q_f16(float16x8_t __p0)35306 __ai __attribute__((target("fullfp16"))) float16x8_t vrev64q_f16(float16x8_t __p0) {
35307 float16x8_t __ret;
35308 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
35309 __ret = __builtin_shufflevector(__rev0, __rev0, 3, 2, 1, 0, 7, 6, 5, 4);
35310 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
35311 return __ret;
35312 }
35313 #endif
35314
35315 #ifdef __LITTLE_ENDIAN__
vrev64_f16(float16x4_t __p0)35316 __ai __attribute__((target("fullfp16"))) float16x4_t vrev64_f16(float16x4_t __p0) {
35317 float16x4_t __ret;
35318 __ret = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35319 return __ret;
35320 }
35321 #else
vrev64_f16(float16x4_t __p0)35322 __ai __attribute__((target("fullfp16"))) float16x4_t vrev64_f16(float16x4_t __p0) {
35323 float16x4_t __ret;
35324 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35325 __ret = __builtin_shufflevector(__rev0, __rev0, 3, 2, 1, 0);
35326 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
35327 return __ret;
35328 }
35329 #endif
35330
35331 #ifdef __LITTLE_ENDIAN__
vrsqrteq_f16(float16x8_t __p0)35332 __ai __attribute__((target("fullfp16"))) float16x8_t vrsqrteq_f16(float16x8_t __p0) {
35333 float16x8_t __ret;
35334 __ret = (float16x8_t) __builtin_neon_vrsqrteq_f16((int8x16_t)__p0, 40);
35335 return __ret;
35336 }
35337 #else
vrsqrteq_f16(float16x8_t __p0)35338 __ai __attribute__((target("fullfp16"))) float16x8_t vrsqrteq_f16(float16x8_t __p0) {
35339 float16x8_t __ret;
35340 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
35341 __ret = (float16x8_t) __builtin_neon_vrsqrteq_f16((int8x16_t)__rev0, 40);
35342 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
35343 return __ret;
35344 }
35345 #endif
35346
35347 #ifdef __LITTLE_ENDIAN__
vrsqrte_f16(float16x4_t __p0)35348 __ai __attribute__((target("fullfp16"))) float16x4_t vrsqrte_f16(float16x4_t __p0) {
35349 float16x4_t __ret;
35350 __ret = (float16x4_t) __builtin_neon_vrsqrte_f16((int8x8_t)__p0, 8);
35351 return __ret;
35352 }
35353 #else
vrsqrte_f16(float16x4_t __p0)35354 __ai __attribute__((target("fullfp16"))) float16x4_t vrsqrte_f16(float16x4_t __p0) {
35355 float16x4_t __ret;
35356 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35357 __ret = (float16x4_t) __builtin_neon_vrsqrte_f16((int8x8_t)__rev0, 8);
35358 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
35359 return __ret;
35360 }
35361 #endif
35362
35363 #ifdef __LITTLE_ENDIAN__
vrsqrtsq_f16(float16x8_t __p0,float16x8_t __p1)35364 __ai __attribute__((target("fullfp16"))) float16x8_t vrsqrtsq_f16(float16x8_t __p0, float16x8_t __p1) {
35365 float16x8_t __ret;
35366 __ret = (float16x8_t) __builtin_neon_vrsqrtsq_f16((int8x16_t)__p0, (int8x16_t)__p1, 40);
35367 return __ret;
35368 }
35369 #else
vrsqrtsq_f16(float16x8_t __p0,float16x8_t __p1)35370 __ai __attribute__((target("fullfp16"))) float16x8_t vrsqrtsq_f16(float16x8_t __p0, float16x8_t __p1) {
35371 float16x8_t __ret;
35372 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
35373 float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
35374 __ret = (float16x8_t) __builtin_neon_vrsqrtsq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, 40);
35375 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
35376 return __ret;
35377 }
35378 #endif
35379
35380 #ifdef __LITTLE_ENDIAN__
vrsqrts_f16(float16x4_t __p0,float16x4_t __p1)35381 __ai __attribute__((target("fullfp16"))) float16x4_t vrsqrts_f16(float16x4_t __p0, float16x4_t __p1) {
35382 float16x4_t __ret;
35383 __ret = (float16x4_t) __builtin_neon_vrsqrts_f16((int8x8_t)__p0, (int8x8_t)__p1, 8);
35384 return __ret;
35385 }
35386 #else
vrsqrts_f16(float16x4_t __p0,float16x4_t __p1)35387 __ai __attribute__((target("fullfp16"))) float16x4_t vrsqrts_f16(float16x4_t __p0, float16x4_t __p1) {
35388 float16x4_t __ret;
35389 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35390 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
35391 __ret = (float16x4_t) __builtin_neon_vrsqrts_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 8);
35392 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
35393 return __ret;
35394 }
35395 #endif
35396
35397 #ifdef __LITTLE_ENDIAN__
vsubq_f16(float16x8_t __p0,float16x8_t __p1)35398 __ai __attribute__((target("fullfp16"))) float16x8_t vsubq_f16(float16x8_t __p0, float16x8_t __p1) {
35399 float16x8_t __ret;
35400 __ret = __p0 - __p1;
35401 return __ret;
35402 }
35403 #else
vsubq_f16(float16x8_t __p0,float16x8_t __p1)35404 __ai __attribute__((target("fullfp16"))) float16x8_t vsubq_f16(float16x8_t __p0, float16x8_t __p1) {
35405 float16x8_t __ret;
35406 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
35407 float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
35408 __ret = __rev0 - __rev1;
35409 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
35410 return __ret;
35411 }
35412 #endif
35413
35414 #ifdef __LITTLE_ENDIAN__
vsub_f16(float16x4_t __p0,float16x4_t __p1)35415 __ai __attribute__((target("fullfp16"))) float16x4_t vsub_f16(float16x4_t __p0, float16x4_t __p1) {
35416 float16x4_t __ret;
35417 __ret = __p0 - __p1;
35418 return __ret;
35419 }
35420 #else
vsub_f16(float16x4_t __p0,float16x4_t __p1)35421 __ai __attribute__((target("fullfp16"))) float16x4_t vsub_f16(float16x4_t __p0, float16x4_t __p1) {
35422 float16x4_t __ret;
35423 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35424 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
35425 __ret = __rev0 - __rev1;
35426 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
35427 return __ret;
35428 }
35429 #endif
35430
35431 #ifdef __LITTLE_ENDIAN__
vtrnq_f16(float16x8_t __p0,float16x8_t __p1)35432 __ai __attribute__((target("fullfp16"))) float16x8x2_t vtrnq_f16(float16x8_t __p0, float16x8_t __p1) {
35433 float16x8x2_t __ret;
35434 __builtin_neon_vtrnq_f16(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 40);
35435 return __ret;
35436 }
35437 #else
vtrnq_f16(float16x8_t __p0,float16x8_t __p1)35438 __ai __attribute__((target("fullfp16"))) float16x8x2_t vtrnq_f16(float16x8_t __p0, float16x8_t __p1) {
35439 float16x8x2_t __ret;
35440 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
35441 float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
35442 __builtin_neon_vtrnq_f16(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 40);
35443
35444 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
35445 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
35446 return __ret;
35447 }
35448 #endif
35449
35450 #ifdef __LITTLE_ENDIAN__
vtrn_f16(float16x4_t __p0,float16x4_t __p1)35451 __ai __attribute__((target("fullfp16"))) float16x4x2_t vtrn_f16(float16x4_t __p0, float16x4_t __p1) {
35452 float16x4x2_t __ret;
35453 __builtin_neon_vtrn_f16(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 8);
35454 return __ret;
35455 }
35456 #else
vtrn_f16(float16x4_t __p0,float16x4_t __p1)35457 __ai __attribute__((target("fullfp16"))) float16x4x2_t vtrn_f16(float16x4_t __p0, float16x4_t __p1) {
35458 float16x4x2_t __ret;
35459 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35460 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
35461 __builtin_neon_vtrn_f16(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 8);
35462
35463 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
35464 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
35465 return __ret;
35466 }
35467 #endif
35468
35469 #ifdef __LITTLE_ENDIAN__
vuzpq_f16(float16x8_t __p0,float16x8_t __p1)35470 __ai __attribute__((target("fullfp16"))) float16x8x2_t vuzpq_f16(float16x8_t __p0, float16x8_t __p1) {
35471 float16x8x2_t __ret;
35472 __builtin_neon_vuzpq_f16(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 40);
35473 return __ret;
35474 }
35475 #else
vuzpq_f16(float16x8_t __p0,float16x8_t __p1)35476 __ai __attribute__((target("fullfp16"))) float16x8x2_t vuzpq_f16(float16x8_t __p0, float16x8_t __p1) {
35477 float16x8x2_t __ret;
35478 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
35479 float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
35480 __builtin_neon_vuzpq_f16(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 40);
35481
35482 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
35483 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
35484 return __ret;
35485 }
35486 #endif
35487
35488 #ifdef __LITTLE_ENDIAN__
vuzp_f16(float16x4_t __p0,float16x4_t __p1)35489 __ai __attribute__((target("fullfp16"))) float16x4x2_t vuzp_f16(float16x4_t __p0, float16x4_t __p1) {
35490 float16x4x2_t __ret;
35491 __builtin_neon_vuzp_f16(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 8);
35492 return __ret;
35493 }
35494 #else
vuzp_f16(float16x4_t __p0,float16x4_t __p1)35495 __ai __attribute__((target("fullfp16"))) float16x4x2_t vuzp_f16(float16x4_t __p0, float16x4_t __p1) {
35496 float16x4x2_t __ret;
35497 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35498 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
35499 __builtin_neon_vuzp_f16(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 8);
35500
35501 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
35502 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
35503 return __ret;
35504 }
35505 #endif
35506
35507 #ifdef __LITTLE_ENDIAN__
vzipq_f16(float16x8_t __p0,float16x8_t __p1)35508 __ai __attribute__((target("fullfp16"))) float16x8x2_t vzipq_f16(float16x8_t __p0, float16x8_t __p1) {
35509 float16x8x2_t __ret;
35510 __builtin_neon_vzipq_f16(&__ret, (int8x16_t)__p0, (int8x16_t)__p1, 40);
35511 return __ret;
35512 }
35513 #else
vzipq_f16(float16x8_t __p0,float16x8_t __p1)35514 __ai __attribute__((target("fullfp16"))) float16x8x2_t vzipq_f16(float16x8_t __p0, float16x8_t __p1) {
35515 float16x8x2_t __ret;
35516 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
35517 float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
35518 __builtin_neon_vzipq_f16(&__ret, (int8x16_t)__rev0, (int8x16_t)__rev1, 40);
35519
35520 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 7, 6, 5, 4, 3, 2, 1, 0);
35521 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 7, 6, 5, 4, 3, 2, 1, 0);
35522 return __ret;
35523 }
35524 #endif
35525
35526 #ifdef __LITTLE_ENDIAN__
vzip_f16(float16x4_t __p0,float16x4_t __p1)35527 __ai __attribute__((target("fullfp16"))) float16x4x2_t vzip_f16(float16x4_t __p0, float16x4_t __p1) {
35528 float16x4x2_t __ret;
35529 __builtin_neon_vzip_f16(&__ret, (int8x8_t)__p0, (int8x8_t)__p1, 8);
35530 return __ret;
35531 }
35532 #else
vzip_f16(float16x4_t __p0,float16x4_t __p1)35533 __ai __attribute__((target("fullfp16"))) float16x4x2_t vzip_f16(float16x4_t __p0, float16x4_t __p1) {
35534 float16x4x2_t __ret;
35535 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35536 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
35537 __builtin_neon_vzip_f16(&__ret, (int8x8_t)__rev0, (int8x8_t)__rev1, 8);
35538
35539 __ret.val[0] = __builtin_shufflevector(__ret.val[0], __ret.val[0], 3, 2, 1, 0);
35540 __ret.val[1] = __builtin_shufflevector(__ret.val[1], __ret.val[1], 3, 2, 1, 0);
35541 return __ret;
35542 }
35543 #endif
35544
35545 #ifdef __LITTLE_ENDIAN__
vmmlaq_u32(uint32x4_t __p0,uint8x16_t __p1,uint8x16_t __p2)35546 __ai __attribute__((target("i8mm"))) uint32x4_t vmmlaq_u32(uint32x4_t __p0, uint8x16_t __p1, uint8x16_t __p2) {
35547 uint32x4_t __ret;
35548 __ret = (uint32x4_t) __builtin_neon_vmmlaq_u32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 50);
35549 return __ret;
35550 }
35551 #else
vmmlaq_u32(uint32x4_t __p0,uint8x16_t __p1,uint8x16_t __p2)35552 __ai __attribute__((target("i8mm"))) uint32x4_t vmmlaq_u32(uint32x4_t __p0, uint8x16_t __p1, uint8x16_t __p2) {
35553 uint32x4_t __ret;
35554 uint32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35555 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
35556 uint8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
35557 __ret = (uint32x4_t) __builtin_neon_vmmlaq_u32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 50);
35558 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
35559 return __ret;
35560 }
35561 #endif
35562
35563 #ifdef __LITTLE_ENDIAN__
vmmlaq_s32(int32x4_t __p0,int8x16_t __p1,int8x16_t __p2)35564 __ai __attribute__((target("i8mm"))) int32x4_t vmmlaq_s32(int32x4_t __p0, int8x16_t __p1, int8x16_t __p2) {
35565 int32x4_t __ret;
35566 __ret = (int32x4_t) __builtin_neon_vmmlaq_s32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 34);
35567 return __ret;
35568 }
35569 #else
vmmlaq_s32(int32x4_t __p0,int8x16_t __p1,int8x16_t __p2)35570 __ai __attribute__((target("i8mm"))) int32x4_t vmmlaq_s32(int32x4_t __p0, int8x16_t __p1, int8x16_t __p2) {
35571 int32x4_t __ret;
35572 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35573 int8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
35574 int8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
35575 __ret = (int32x4_t) __builtin_neon_vmmlaq_s32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 34);
35576 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
35577 return __ret;
35578 }
35579 #endif
35580
35581 #ifdef __LITTLE_ENDIAN__
vusdotq_s32(int32x4_t __p0,uint8x16_t __p1,int8x16_t __p2)35582 __ai __attribute__((target("i8mm"))) int32x4_t vusdotq_s32(int32x4_t __p0, uint8x16_t __p1, int8x16_t __p2) {
35583 int32x4_t __ret;
35584 __ret = (int32x4_t) __builtin_neon_vusdotq_s32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 34);
35585 return __ret;
35586 }
35587 #else
vusdotq_s32(int32x4_t __p0,uint8x16_t __p1,int8x16_t __p2)35588 __ai __attribute__((target("i8mm"))) int32x4_t vusdotq_s32(int32x4_t __p0, uint8x16_t __p1, int8x16_t __p2) {
35589 int32x4_t __ret;
35590 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35591 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
35592 int8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
35593 __ret = (int32x4_t) __builtin_neon_vusdotq_s32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 34);
35594 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
35595 return __ret;
35596 }
__noswap_vusdotq_s32(int32x4_t __p0,uint8x16_t __p1,int8x16_t __p2)35597 __ai __attribute__((target("i8mm"))) int32x4_t __noswap_vusdotq_s32(int32x4_t __p0, uint8x16_t __p1, int8x16_t __p2) {
35598 int32x4_t __ret;
35599 __ret = (int32x4_t) __builtin_neon_vusdotq_s32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 34);
35600 return __ret;
35601 }
35602 #endif
35603
35604 #ifdef __LITTLE_ENDIAN__
vusdot_s32(int32x2_t __p0,uint8x8_t __p1,int8x8_t __p2)35605 __ai __attribute__((target("i8mm"))) int32x2_t vusdot_s32(int32x2_t __p0, uint8x8_t __p1, int8x8_t __p2) {
35606 int32x2_t __ret;
35607 __ret = (int32x2_t) __builtin_neon_vusdot_s32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 2);
35608 return __ret;
35609 }
35610 #else
vusdot_s32(int32x2_t __p0,uint8x8_t __p1,int8x8_t __p2)35611 __ai __attribute__((target("i8mm"))) int32x2_t vusdot_s32(int32x2_t __p0, uint8x8_t __p1, int8x8_t __p2) {
35612 int32x2_t __ret;
35613 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
35614 uint8x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
35615 int8x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
35616 __ret = (int32x2_t) __builtin_neon_vusdot_s32((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 2);
35617 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
35618 return __ret;
35619 }
__noswap_vusdot_s32(int32x2_t __p0,uint8x8_t __p1,int8x8_t __p2)35620 __ai __attribute__((target("i8mm"))) int32x2_t __noswap_vusdot_s32(int32x2_t __p0, uint8x8_t __p1, int8x8_t __p2) {
35621 int32x2_t __ret;
35622 __ret = (int32x2_t) __builtin_neon_vusdot_s32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 2);
35623 return __ret;
35624 }
35625 #endif
35626
35627 #ifdef __LITTLE_ENDIAN__
35628 #define vusdotq_lane_s32(__p0_157, __p1_157, __p2_157, __p3_157) __extension__ ({ \
35629 int32x4_t __ret_157; \
35630 int32x4_t __s0_157 = __p0_157; \
35631 uint8x16_t __s1_157 = __p1_157; \
35632 int8x8_t __s2_157 = __p2_157; \
35633 int8x8_t __reint_157 = __s2_157; \
35634 __ret_157 = vusdotq_s32(__s0_157, __s1_157, (int8x16_t)(splatq_lane_s32(*(int32x2_t *) &__reint_157, __p3_157))); \
35635 __ret_157; \
35636 })
35637 #else
35638 #define vusdotq_lane_s32(__p0_158, __p1_158, __p2_158, __p3_158) __extension__ ({ \
35639 int32x4_t __ret_158; \
35640 int32x4_t __s0_158 = __p0_158; \
35641 uint8x16_t __s1_158 = __p1_158; \
35642 int8x8_t __s2_158 = __p2_158; \
35643 int32x4_t __rev0_158; __rev0_158 = __builtin_shufflevector(__s0_158, __s0_158, 3, 2, 1, 0); \
35644 uint8x16_t __rev1_158; __rev1_158 = __builtin_shufflevector(__s1_158, __s1_158, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); \
35645 int8x8_t __rev2_158; __rev2_158 = __builtin_shufflevector(__s2_158, __s2_158, 7, 6, 5, 4, 3, 2, 1, 0); \
35646 int8x8_t __reint_158 = __rev2_158; \
35647 __ret_158 = __noswap_vusdotq_s32(__rev0_158, __rev1_158, (int8x16_t)(__noswap_splatq_lane_s32(*(int32x2_t *) &__reint_158, __p3_158))); \
35648 __ret_158 = __builtin_shufflevector(__ret_158, __ret_158, 3, 2, 1, 0); \
35649 __ret_158; \
35650 })
35651 #endif
35652
35653 #ifdef __LITTLE_ENDIAN__
35654 #define vusdot_lane_s32(__p0_159, __p1_159, __p2_159, __p3_159) __extension__ ({ \
35655 int32x2_t __ret_159; \
35656 int32x2_t __s0_159 = __p0_159; \
35657 uint8x8_t __s1_159 = __p1_159; \
35658 int8x8_t __s2_159 = __p2_159; \
35659 int8x8_t __reint_159 = __s2_159; \
35660 __ret_159 = vusdot_s32(__s0_159, __s1_159, (int8x8_t)(splat_lane_s32(*(int32x2_t *) &__reint_159, __p3_159))); \
35661 __ret_159; \
35662 })
35663 #else
35664 #define vusdot_lane_s32(__p0_160, __p1_160, __p2_160, __p3_160) __extension__ ({ \
35665 int32x2_t __ret_160; \
35666 int32x2_t __s0_160 = __p0_160; \
35667 uint8x8_t __s1_160 = __p1_160; \
35668 int8x8_t __s2_160 = __p2_160; \
35669 int32x2_t __rev0_160; __rev0_160 = __builtin_shufflevector(__s0_160, __s0_160, 1, 0); \
35670 uint8x8_t __rev1_160; __rev1_160 = __builtin_shufflevector(__s1_160, __s1_160, 7, 6, 5, 4, 3, 2, 1, 0); \
35671 int8x8_t __rev2_160; __rev2_160 = __builtin_shufflevector(__s2_160, __s2_160, 7, 6, 5, 4, 3, 2, 1, 0); \
35672 int8x8_t __reint_160 = __rev2_160; \
35673 __ret_160 = __noswap_vusdot_s32(__rev0_160, __rev1_160, (int8x8_t)(__noswap_splat_lane_s32(*(int32x2_t *) &__reint_160, __p3_160))); \
35674 __ret_160 = __builtin_shufflevector(__ret_160, __ret_160, 1, 0); \
35675 __ret_160; \
35676 })
35677 #endif
35678
35679 #ifdef __LITTLE_ENDIAN__
vusmmlaq_s32(int32x4_t __p0,uint8x16_t __p1,int8x16_t __p2)35680 __ai __attribute__((target("i8mm"))) int32x4_t vusmmlaq_s32(int32x4_t __p0, uint8x16_t __p1, int8x16_t __p2) {
35681 int32x4_t __ret;
35682 __ret = (int32x4_t) __builtin_neon_vusmmlaq_s32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 34);
35683 return __ret;
35684 }
35685 #else
vusmmlaq_s32(int32x4_t __p0,uint8x16_t __p1,int8x16_t __p2)35686 __ai __attribute__((target("i8mm"))) int32x4_t vusmmlaq_s32(int32x4_t __p0, uint8x16_t __p1, int8x16_t __p2) {
35687 int32x4_t __ret;
35688 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35689 uint8x16_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
35690 int8x16_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
35691 __ret = (int32x4_t) __builtin_neon_vusmmlaq_s32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 34);
35692 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
35693 return __ret;
35694 }
35695 #endif
35696
35697 #ifdef __LITTLE_ENDIAN__
vqrdmlahq_s32(int32x4_t __p0,int32x4_t __p1,int32x4_t __p2)35698 __ai __attribute__((target("v8.1a"))) int32x4_t vqrdmlahq_s32(int32x4_t __p0, int32x4_t __p1, int32x4_t __p2) {
35699 int32x4_t __ret;
35700 __ret = (int32x4_t) __builtin_neon_vqrdmlahq_s32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 34);
35701 return __ret;
35702 }
35703 #else
vqrdmlahq_s32(int32x4_t __p0,int32x4_t __p1,int32x4_t __p2)35704 __ai __attribute__((target("v8.1a"))) int32x4_t vqrdmlahq_s32(int32x4_t __p0, int32x4_t __p1, int32x4_t __p2) {
35705 int32x4_t __ret;
35706 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35707 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
35708 int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
35709 __ret = (int32x4_t) __builtin_neon_vqrdmlahq_s32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 34);
35710 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
35711 return __ret;
35712 }
__noswap_vqrdmlahq_s32(int32x4_t __p0,int32x4_t __p1,int32x4_t __p2)35713 __ai __attribute__((target("v8.1a"))) int32x4_t __noswap_vqrdmlahq_s32(int32x4_t __p0, int32x4_t __p1, int32x4_t __p2) {
35714 int32x4_t __ret;
35715 __ret = (int32x4_t) __builtin_neon_vqrdmlahq_s32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 34);
35716 return __ret;
35717 }
35718 #endif
35719
35720 #ifdef __LITTLE_ENDIAN__
vqrdmlahq_s16(int16x8_t __p0,int16x8_t __p1,int16x8_t __p2)35721 __ai __attribute__((target("v8.1a"))) int16x8_t vqrdmlahq_s16(int16x8_t __p0, int16x8_t __p1, int16x8_t __p2) {
35722 int16x8_t __ret;
35723 __ret = (int16x8_t) __builtin_neon_vqrdmlahq_s16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 33);
35724 return __ret;
35725 }
35726 #else
vqrdmlahq_s16(int16x8_t __p0,int16x8_t __p1,int16x8_t __p2)35727 __ai __attribute__((target("v8.1a"))) int16x8_t vqrdmlahq_s16(int16x8_t __p0, int16x8_t __p1, int16x8_t __p2) {
35728 int16x8_t __ret;
35729 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
35730 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
35731 int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
35732 __ret = (int16x8_t) __builtin_neon_vqrdmlahq_s16((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 33);
35733 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
35734 return __ret;
35735 }
__noswap_vqrdmlahq_s16(int16x8_t __p0,int16x8_t __p1,int16x8_t __p2)35736 __ai __attribute__((target("v8.1a"))) int16x8_t __noswap_vqrdmlahq_s16(int16x8_t __p0, int16x8_t __p1, int16x8_t __p2) {
35737 int16x8_t __ret;
35738 __ret = (int16x8_t) __builtin_neon_vqrdmlahq_s16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 33);
35739 return __ret;
35740 }
35741 #endif
35742
35743 #ifdef __LITTLE_ENDIAN__
vqrdmlah_s32(int32x2_t __p0,int32x2_t __p1,int32x2_t __p2)35744 __ai __attribute__((target("v8.1a"))) int32x2_t vqrdmlah_s32(int32x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
35745 int32x2_t __ret;
35746 __ret = (int32x2_t) __builtin_neon_vqrdmlah_s32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 2);
35747 return __ret;
35748 }
35749 #else
vqrdmlah_s32(int32x2_t __p0,int32x2_t __p1,int32x2_t __p2)35750 __ai __attribute__((target("v8.1a"))) int32x2_t vqrdmlah_s32(int32x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
35751 int32x2_t __ret;
35752 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
35753 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
35754 int32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
35755 __ret = (int32x2_t) __builtin_neon_vqrdmlah_s32((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 2);
35756 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
35757 return __ret;
35758 }
__noswap_vqrdmlah_s32(int32x2_t __p0,int32x2_t __p1,int32x2_t __p2)35759 __ai __attribute__((target("v8.1a"))) int32x2_t __noswap_vqrdmlah_s32(int32x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
35760 int32x2_t __ret;
35761 __ret = (int32x2_t) __builtin_neon_vqrdmlah_s32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 2);
35762 return __ret;
35763 }
35764 #endif
35765
35766 #ifdef __LITTLE_ENDIAN__
vqrdmlah_s16(int16x4_t __p0,int16x4_t __p1,int16x4_t __p2)35767 __ai __attribute__((target("v8.1a"))) int16x4_t vqrdmlah_s16(int16x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
35768 int16x4_t __ret;
35769 __ret = (int16x4_t) __builtin_neon_vqrdmlah_s16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 1);
35770 return __ret;
35771 }
35772 #else
vqrdmlah_s16(int16x4_t __p0,int16x4_t __p1,int16x4_t __p2)35773 __ai __attribute__((target("v8.1a"))) int16x4_t vqrdmlah_s16(int16x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
35774 int16x4_t __ret;
35775 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35776 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
35777 int16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
35778 __ret = (int16x4_t) __builtin_neon_vqrdmlah_s16((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 1);
35779 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
35780 return __ret;
35781 }
__noswap_vqrdmlah_s16(int16x4_t __p0,int16x4_t __p1,int16x4_t __p2)35782 __ai __attribute__((target("v8.1a"))) int16x4_t __noswap_vqrdmlah_s16(int16x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
35783 int16x4_t __ret;
35784 __ret = (int16x4_t) __builtin_neon_vqrdmlah_s16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 1);
35785 return __ret;
35786 }
35787 #endif
35788
35789 #ifdef __LITTLE_ENDIAN__
35790 #define vqrdmlahq_lane_s32(__p0_161, __p1_161, __p2_161, __p3_161) __extension__ ({ \
35791 int32x4_t __ret_161; \
35792 int32x4_t __s0_161 = __p0_161; \
35793 int32x4_t __s1_161 = __p1_161; \
35794 int32x2_t __s2_161 = __p2_161; \
35795 __ret_161 = vqrdmlahq_s32(__s0_161, __s1_161, splatq_lane_s32(__s2_161, __p3_161)); \
35796 __ret_161; \
35797 })
35798 #else
35799 #define vqrdmlahq_lane_s32(__p0_162, __p1_162, __p2_162, __p3_162) __extension__ ({ \
35800 int32x4_t __ret_162; \
35801 int32x4_t __s0_162 = __p0_162; \
35802 int32x4_t __s1_162 = __p1_162; \
35803 int32x2_t __s2_162 = __p2_162; \
35804 int32x4_t __rev0_162; __rev0_162 = __builtin_shufflevector(__s0_162, __s0_162, 3, 2, 1, 0); \
35805 int32x4_t __rev1_162; __rev1_162 = __builtin_shufflevector(__s1_162, __s1_162, 3, 2, 1, 0); \
35806 int32x2_t __rev2_162; __rev2_162 = __builtin_shufflevector(__s2_162, __s2_162, 1, 0); \
35807 __ret_162 = __noswap_vqrdmlahq_s32(__rev0_162, __rev1_162, __noswap_splatq_lane_s32(__rev2_162, __p3_162)); \
35808 __ret_162 = __builtin_shufflevector(__ret_162, __ret_162, 3, 2, 1, 0); \
35809 __ret_162; \
35810 })
35811 #endif
35812
35813 #ifdef __LITTLE_ENDIAN__
35814 #define vqrdmlahq_lane_s16(__p0_163, __p1_163, __p2_163, __p3_163) __extension__ ({ \
35815 int16x8_t __ret_163; \
35816 int16x8_t __s0_163 = __p0_163; \
35817 int16x8_t __s1_163 = __p1_163; \
35818 int16x4_t __s2_163 = __p2_163; \
35819 __ret_163 = vqrdmlahq_s16(__s0_163, __s1_163, splatq_lane_s16(__s2_163, __p3_163)); \
35820 __ret_163; \
35821 })
35822 #else
35823 #define vqrdmlahq_lane_s16(__p0_164, __p1_164, __p2_164, __p3_164) __extension__ ({ \
35824 int16x8_t __ret_164; \
35825 int16x8_t __s0_164 = __p0_164; \
35826 int16x8_t __s1_164 = __p1_164; \
35827 int16x4_t __s2_164 = __p2_164; \
35828 int16x8_t __rev0_164; __rev0_164 = __builtin_shufflevector(__s0_164, __s0_164, 7, 6, 5, 4, 3, 2, 1, 0); \
35829 int16x8_t __rev1_164; __rev1_164 = __builtin_shufflevector(__s1_164, __s1_164, 7, 6, 5, 4, 3, 2, 1, 0); \
35830 int16x4_t __rev2_164; __rev2_164 = __builtin_shufflevector(__s2_164, __s2_164, 3, 2, 1, 0); \
35831 __ret_164 = __noswap_vqrdmlahq_s16(__rev0_164, __rev1_164, __noswap_splatq_lane_s16(__rev2_164, __p3_164)); \
35832 __ret_164 = __builtin_shufflevector(__ret_164, __ret_164, 7, 6, 5, 4, 3, 2, 1, 0); \
35833 __ret_164; \
35834 })
35835 #endif
35836
35837 #ifdef __LITTLE_ENDIAN__
35838 #define vqrdmlah_lane_s32(__p0_165, __p1_165, __p2_165, __p3_165) __extension__ ({ \
35839 int32x2_t __ret_165; \
35840 int32x2_t __s0_165 = __p0_165; \
35841 int32x2_t __s1_165 = __p1_165; \
35842 int32x2_t __s2_165 = __p2_165; \
35843 __ret_165 = vqrdmlah_s32(__s0_165, __s1_165, splat_lane_s32(__s2_165, __p3_165)); \
35844 __ret_165; \
35845 })
35846 #else
35847 #define vqrdmlah_lane_s32(__p0_166, __p1_166, __p2_166, __p3_166) __extension__ ({ \
35848 int32x2_t __ret_166; \
35849 int32x2_t __s0_166 = __p0_166; \
35850 int32x2_t __s1_166 = __p1_166; \
35851 int32x2_t __s2_166 = __p2_166; \
35852 int32x2_t __rev0_166; __rev0_166 = __builtin_shufflevector(__s0_166, __s0_166, 1, 0); \
35853 int32x2_t __rev1_166; __rev1_166 = __builtin_shufflevector(__s1_166, __s1_166, 1, 0); \
35854 int32x2_t __rev2_166; __rev2_166 = __builtin_shufflevector(__s2_166, __s2_166, 1, 0); \
35855 __ret_166 = __noswap_vqrdmlah_s32(__rev0_166, __rev1_166, __noswap_splat_lane_s32(__rev2_166, __p3_166)); \
35856 __ret_166 = __builtin_shufflevector(__ret_166, __ret_166, 1, 0); \
35857 __ret_166; \
35858 })
35859 #endif
35860
35861 #ifdef __LITTLE_ENDIAN__
35862 #define vqrdmlah_lane_s16(__p0_167, __p1_167, __p2_167, __p3_167) __extension__ ({ \
35863 int16x4_t __ret_167; \
35864 int16x4_t __s0_167 = __p0_167; \
35865 int16x4_t __s1_167 = __p1_167; \
35866 int16x4_t __s2_167 = __p2_167; \
35867 __ret_167 = vqrdmlah_s16(__s0_167, __s1_167, splat_lane_s16(__s2_167, __p3_167)); \
35868 __ret_167; \
35869 })
35870 #else
35871 #define vqrdmlah_lane_s16(__p0_168, __p1_168, __p2_168, __p3_168) __extension__ ({ \
35872 int16x4_t __ret_168; \
35873 int16x4_t __s0_168 = __p0_168; \
35874 int16x4_t __s1_168 = __p1_168; \
35875 int16x4_t __s2_168 = __p2_168; \
35876 int16x4_t __rev0_168; __rev0_168 = __builtin_shufflevector(__s0_168, __s0_168, 3, 2, 1, 0); \
35877 int16x4_t __rev1_168; __rev1_168 = __builtin_shufflevector(__s1_168, __s1_168, 3, 2, 1, 0); \
35878 int16x4_t __rev2_168; __rev2_168 = __builtin_shufflevector(__s2_168, __s2_168, 3, 2, 1, 0); \
35879 __ret_168 = __noswap_vqrdmlah_s16(__rev0_168, __rev1_168, __noswap_splat_lane_s16(__rev2_168, __p3_168)); \
35880 __ret_168 = __builtin_shufflevector(__ret_168, __ret_168, 3, 2, 1, 0); \
35881 __ret_168; \
35882 })
35883 #endif
35884
35885 #ifdef __LITTLE_ENDIAN__
vqrdmlshq_s32(int32x4_t __p0,int32x4_t __p1,int32x4_t __p2)35886 __ai __attribute__((target("v8.1a"))) int32x4_t vqrdmlshq_s32(int32x4_t __p0, int32x4_t __p1, int32x4_t __p2) {
35887 int32x4_t __ret;
35888 __ret = (int32x4_t) __builtin_neon_vqrdmlshq_s32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 34);
35889 return __ret;
35890 }
35891 #else
vqrdmlshq_s32(int32x4_t __p0,int32x4_t __p1,int32x4_t __p2)35892 __ai __attribute__((target("v8.1a"))) int32x4_t vqrdmlshq_s32(int32x4_t __p0, int32x4_t __p1, int32x4_t __p2) {
35893 int32x4_t __ret;
35894 int32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35895 int32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
35896 int32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
35897 __ret = (int32x4_t) __builtin_neon_vqrdmlshq_s32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 34);
35898 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
35899 return __ret;
35900 }
__noswap_vqrdmlshq_s32(int32x4_t __p0,int32x4_t __p1,int32x4_t __p2)35901 __ai __attribute__((target("v8.1a"))) int32x4_t __noswap_vqrdmlshq_s32(int32x4_t __p0, int32x4_t __p1, int32x4_t __p2) {
35902 int32x4_t __ret;
35903 __ret = (int32x4_t) __builtin_neon_vqrdmlshq_s32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 34);
35904 return __ret;
35905 }
35906 #endif
35907
35908 #ifdef __LITTLE_ENDIAN__
vqrdmlshq_s16(int16x8_t __p0,int16x8_t __p1,int16x8_t __p2)35909 __ai __attribute__((target("v8.1a"))) int16x8_t vqrdmlshq_s16(int16x8_t __p0, int16x8_t __p1, int16x8_t __p2) {
35910 int16x8_t __ret;
35911 __ret = (int16x8_t) __builtin_neon_vqrdmlshq_s16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 33);
35912 return __ret;
35913 }
35914 #else
vqrdmlshq_s16(int16x8_t __p0,int16x8_t __p1,int16x8_t __p2)35915 __ai __attribute__((target("v8.1a"))) int16x8_t vqrdmlshq_s16(int16x8_t __p0, int16x8_t __p1, int16x8_t __p2) {
35916 int16x8_t __ret;
35917 int16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
35918 int16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
35919 int16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
35920 __ret = (int16x8_t) __builtin_neon_vqrdmlshq_s16((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 33);
35921 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
35922 return __ret;
35923 }
__noswap_vqrdmlshq_s16(int16x8_t __p0,int16x8_t __p1,int16x8_t __p2)35924 __ai __attribute__((target("v8.1a"))) int16x8_t __noswap_vqrdmlshq_s16(int16x8_t __p0, int16x8_t __p1, int16x8_t __p2) {
35925 int16x8_t __ret;
35926 __ret = (int16x8_t) __builtin_neon_vqrdmlshq_s16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 33);
35927 return __ret;
35928 }
35929 #endif
35930
35931 #ifdef __LITTLE_ENDIAN__
vqrdmlsh_s32(int32x2_t __p0,int32x2_t __p1,int32x2_t __p2)35932 __ai __attribute__((target("v8.1a"))) int32x2_t vqrdmlsh_s32(int32x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
35933 int32x2_t __ret;
35934 __ret = (int32x2_t) __builtin_neon_vqrdmlsh_s32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 2);
35935 return __ret;
35936 }
35937 #else
vqrdmlsh_s32(int32x2_t __p0,int32x2_t __p1,int32x2_t __p2)35938 __ai __attribute__((target("v8.1a"))) int32x2_t vqrdmlsh_s32(int32x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
35939 int32x2_t __ret;
35940 int32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
35941 int32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
35942 int32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
35943 __ret = (int32x2_t) __builtin_neon_vqrdmlsh_s32((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 2);
35944 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
35945 return __ret;
35946 }
__noswap_vqrdmlsh_s32(int32x2_t __p0,int32x2_t __p1,int32x2_t __p2)35947 __ai __attribute__((target("v8.1a"))) int32x2_t __noswap_vqrdmlsh_s32(int32x2_t __p0, int32x2_t __p1, int32x2_t __p2) {
35948 int32x2_t __ret;
35949 __ret = (int32x2_t) __builtin_neon_vqrdmlsh_s32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 2);
35950 return __ret;
35951 }
35952 #endif
35953
35954 #ifdef __LITTLE_ENDIAN__
vqrdmlsh_s16(int16x4_t __p0,int16x4_t __p1,int16x4_t __p2)35955 __ai __attribute__((target("v8.1a"))) int16x4_t vqrdmlsh_s16(int16x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
35956 int16x4_t __ret;
35957 __ret = (int16x4_t) __builtin_neon_vqrdmlsh_s16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 1);
35958 return __ret;
35959 }
35960 #else
vqrdmlsh_s16(int16x4_t __p0,int16x4_t __p1,int16x4_t __p2)35961 __ai __attribute__((target("v8.1a"))) int16x4_t vqrdmlsh_s16(int16x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
35962 int16x4_t __ret;
35963 int16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
35964 int16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
35965 int16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
35966 __ret = (int16x4_t) __builtin_neon_vqrdmlsh_s16((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 1);
35967 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
35968 return __ret;
35969 }
__noswap_vqrdmlsh_s16(int16x4_t __p0,int16x4_t __p1,int16x4_t __p2)35970 __ai __attribute__((target("v8.1a"))) int16x4_t __noswap_vqrdmlsh_s16(int16x4_t __p0, int16x4_t __p1, int16x4_t __p2) {
35971 int16x4_t __ret;
35972 __ret = (int16x4_t) __builtin_neon_vqrdmlsh_s16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 1);
35973 return __ret;
35974 }
35975 #endif
35976
35977 #ifdef __LITTLE_ENDIAN__
35978 #define vqrdmlshq_lane_s32(__p0_169, __p1_169, __p2_169, __p3_169) __extension__ ({ \
35979 int32x4_t __ret_169; \
35980 int32x4_t __s0_169 = __p0_169; \
35981 int32x4_t __s1_169 = __p1_169; \
35982 int32x2_t __s2_169 = __p2_169; \
35983 __ret_169 = vqrdmlshq_s32(__s0_169, __s1_169, splatq_lane_s32(__s2_169, __p3_169)); \
35984 __ret_169; \
35985 })
35986 #else
35987 #define vqrdmlshq_lane_s32(__p0_170, __p1_170, __p2_170, __p3_170) __extension__ ({ \
35988 int32x4_t __ret_170; \
35989 int32x4_t __s0_170 = __p0_170; \
35990 int32x4_t __s1_170 = __p1_170; \
35991 int32x2_t __s2_170 = __p2_170; \
35992 int32x4_t __rev0_170; __rev0_170 = __builtin_shufflevector(__s0_170, __s0_170, 3, 2, 1, 0); \
35993 int32x4_t __rev1_170; __rev1_170 = __builtin_shufflevector(__s1_170, __s1_170, 3, 2, 1, 0); \
35994 int32x2_t __rev2_170; __rev2_170 = __builtin_shufflevector(__s2_170, __s2_170, 1, 0); \
35995 __ret_170 = __noswap_vqrdmlshq_s32(__rev0_170, __rev1_170, __noswap_splatq_lane_s32(__rev2_170, __p3_170)); \
35996 __ret_170 = __builtin_shufflevector(__ret_170, __ret_170, 3, 2, 1, 0); \
35997 __ret_170; \
35998 })
35999 #endif
36000
36001 #ifdef __LITTLE_ENDIAN__
36002 #define vqrdmlshq_lane_s16(__p0_171, __p1_171, __p2_171, __p3_171) __extension__ ({ \
36003 int16x8_t __ret_171; \
36004 int16x8_t __s0_171 = __p0_171; \
36005 int16x8_t __s1_171 = __p1_171; \
36006 int16x4_t __s2_171 = __p2_171; \
36007 __ret_171 = vqrdmlshq_s16(__s0_171, __s1_171, splatq_lane_s16(__s2_171, __p3_171)); \
36008 __ret_171; \
36009 })
36010 #else
36011 #define vqrdmlshq_lane_s16(__p0_172, __p1_172, __p2_172, __p3_172) __extension__ ({ \
36012 int16x8_t __ret_172; \
36013 int16x8_t __s0_172 = __p0_172; \
36014 int16x8_t __s1_172 = __p1_172; \
36015 int16x4_t __s2_172 = __p2_172; \
36016 int16x8_t __rev0_172; __rev0_172 = __builtin_shufflevector(__s0_172, __s0_172, 7, 6, 5, 4, 3, 2, 1, 0); \
36017 int16x8_t __rev1_172; __rev1_172 = __builtin_shufflevector(__s1_172, __s1_172, 7, 6, 5, 4, 3, 2, 1, 0); \
36018 int16x4_t __rev2_172; __rev2_172 = __builtin_shufflevector(__s2_172, __s2_172, 3, 2, 1, 0); \
36019 __ret_172 = __noswap_vqrdmlshq_s16(__rev0_172, __rev1_172, __noswap_splatq_lane_s16(__rev2_172, __p3_172)); \
36020 __ret_172 = __builtin_shufflevector(__ret_172, __ret_172, 7, 6, 5, 4, 3, 2, 1, 0); \
36021 __ret_172; \
36022 })
36023 #endif
36024
36025 #ifdef __LITTLE_ENDIAN__
36026 #define vqrdmlsh_lane_s32(__p0_173, __p1_173, __p2_173, __p3_173) __extension__ ({ \
36027 int32x2_t __ret_173; \
36028 int32x2_t __s0_173 = __p0_173; \
36029 int32x2_t __s1_173 = __p1_173; \
36030 int32x2_t __s2_173 = __p2_173; \
36031 __ret_173 = vqrdmlsh_s32(__s0_173, __s1_173, splat_lane_s32(__s2_173, __p3_173)); \
36032 __ret_173; \
36033 })
36034 #else
36035 #define vqrdmlsh_lane_s32(__p0_174, __p1_174, __p2_174, __p3_174) __extension__ ({ \
36036 int32x2_t __ret_174; \
36037 int32x2_t __s0_174 = __p0_174; \
36038 int32x2_t __s1_174 = __p1_174; \
36039 int32x2_t __s2_174 = __p2_174; \
36040 int32x2_t __rev0_174; __rev0_174 = __builtin_shufflevector(__s0_174, __s0_174, 1, 0); \
36041 int32x2_t __rev1_174; __rev1_174 = __builtin_shufflevector(__s1_174, __s1_174, 1, 0); \
36042 int32x2_t __rev2_174; __rev2_174 = __builtin_shufflevector(__s2_174, __s2_174, 1, 0); \
36043 __ret_174 = __noswap_vqrdmlsh_s32(__rev0_174, __rev1_174, __noswap_splat_lane_s32(__rev2_174, __p3_174)); \
36044 __ret_174 = __builtin_shufflevector(__ret_174, __ret_174, 1, 0); \
36045 __ret_174; \
36046 })
36047 #endif
36048
36049 #ifdef __LITTLE_ENDIAN__
36050 #define vqrdmlsh_lane_s16(__p0_175, __p1_175, __p2_175, __p3_175) __extension__ ({ \
36051 int16x4_t __ret_175; \
36052 int16x4_t __s0_175 = __p0_175; \
36053 int16x4_t __s1_175 = __p1_175; \
36054 int16x4_t __s2_175 = __p2_175; \
36055 __ret_175 = vqrdmlsh_s16(__s0_175, __s1_175, splat_lane_s16(__s2_175, __p3_175)); \
36056 __ret_175; \
36057 })
36058 #else
36059 #define vqrdmlsh_lane_s16(__p0_176, __p1_176, __p2_176, __p3_176) __extension__ ({ \
36060 int16x4_t __ret_176; \
36061 int16x4_t __s0_176 = __p0_176; \
36062 int16x4_t __s1_176 = __p1_176; \
36063 int16x4_t __s2_176 = __p2_176; \
36064 int16x4_t __rev0_176; __rev0_176 = __builtin_shufflevector(__s0_176, __s0_176, 3, 2, 1, 0); \
36065 int16x4_t __rev1_176; __rev1_176 = __builtin_shufflevector(__s1_176, __s1_176, 3, 2, 1, 0); \
36066 int16x4_t __rev2_176; __rev2_176 = __builtin_shufflevector(__s2_176, __s2_176, 3, 2, 1, 0); \
36067 __ret_176 = __noswap_vqrdmlsh_s16(__rev0_176, __rev1_176, __noswap_splat_lane_s16(__rev2_176, __p3_176)); \
36068 __ret_176 = __builtin_shufflevector(__ret_176, __ret_176, 3, 2, 1, 0); \
36069 __ret_176; \
36070 })
36071 #endif
36072
36073 #ifdef __LITTLE_ENDIAN__
vcadd_rot270_f32(float32x2_t __p0,float32x2_t __p1)36074 __ai __attribute__((target("v8.3a"))) float32x2_t vcadd_rot270_f32(float32x2_t __p0, float32x2_t __p1) {
36075 float32x2_t __ret;
36076 __ret = (float32x2_t) __builtin_neon_vcadd_rot270_f32((int8x8_t)__p0, (int8x8_t)__p1, 9);
36077 return __ret;
36078 }
36079 #else
vcadd_rot270_f32(float32x2_t __p0,float32x2_t __p1)36080 __ai __attribute__((target("v8.3a"))) float32x2_t vcadd_rot270_f32(float32x2_t __p0, float32x2_t __p1) {
36081 float32x2_t __ret;
36082 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
36083 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
36084 __ret = (float32x2_t) __builtin_neon_vcadd_rot270_f32((int8x8_t)__rev0, (int8x8_t)__rev1, 9);
36085 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
36086 return __ret;
36087 }
36088 #endif
36089
36090 #ifdef __LITTLE_ENDIAN__
vcadd_rot90_f32(float32x2_t __p0,float32x2_t __p1)36091 __ai __attribute__((target("v8.3a"))) float32x2_t vcadd_rot90_f32(float32x2_t __p0, float32x2_t __p1) {
36092 float32x2_t __ret;
36093 __ret = (float32x2_t) __builtin_neon_vcadd_rot90_f32((int8x8_t)__p0, (int8x8_t)__p1, 9);
36094 return __ret;
36095 }
36096 #else
vcadd_rot90_f32(float32x2_t __p0,float32x2_t __p1)36097 __ai __attribute__((target("v8.3a"))) float32x2_t vcadd_rot90_f32(float32x2_t __p0, float32x2_t __p1) {
36098 float32x2_t __ret;
36099 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
36100 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
36101 __ret = (float32x2_t) __builtin_neon_vcadd_rot90_f32((int8x8_t)__rev0, (int8x8_t)__rev1, 9);
36102 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
36103 return __ret;
36104 }
36105 #endif
36106
36107 #ifdef __LITTLE_ENDIAN__
vcaddq_rot270_f32(float32x4_t __p0,float32x4_t __p1)36108 __ai __attribute__((target("v8.3a"))) float32x4_t vcaddq_rot270_f32(float32x4_t __p0, float32x4_t __p1) {
36109 float32x4_t __ret;
36110 __ret = (float32x4_t) __builtin_neon_vcaddq_rot270_f32((int8x16_t)__p0, (int8x16_t)__p1, 41);
36111 return __ret;
36112 }
36113 #else
vcaddq_rot270_f32(float32x4_t __p0,float32x4_t __p1)36114 __ai __attribute__((target("v8.3a"))) float32x4_t vcaddq_rot270_f32(float32x4_t __p0, float32x4_t __p1) {
36115 float32x4_t __ret;
36116 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
36117 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
36118 __ret = (float32x4_t) __builtin_neon_vcaddq_rot270_f32((int8x16_t)__rev0, (int8x16_t)__rev1, 41);
36119 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
36120 return __ret;
36121 }
36122 #endif
36123
36124 #ifdef __LITTLE_ENDIAN__
vcaddq_rot90_f32(float32x4_t __p0,float32x4_t __p1)36125 __ai __attribute__((target("v8.3a"))) float32x4_t vcaddq_rot90_f32(float32x4_t __p0, float32x4_t __p1) {
36126 float32x4_t __ret;
36127 __ret = (float32x4_t) __builtin_neon_vcaddq_rot90_f32((int8x16_t)__p0, (int8x16_t)__p1, 41);
36128 return __ret;
36129 }
36130 #else
vcaddq_rot90_f32(float32x4_t __p0,float32x4_t __p1)36131 __ai __attribute__((target("v8.3a"))) float32x4_t vcaddq_rot90_f32(float32x4_t __p0, float32x4_t __p1) {
36132 float32x4_t __ret;
36133 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
36134 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
36135 __ret = (float32x4_t) __builtin_neon_vcaddq_rot90_f32((int8x16_t)__rev0, (int8x16_t)__rev1, 41);
36136 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
36137 return __ret;
36138 }
36139 #endif
36140
36141 #ifdef __LITTLE_ENDIAN__
vcmlaq_f32(float32x4_t __p0,float32x4_t __p1,float32x4_t __p2)36142 __ai __attribute__((target("v8.3a"))) float32x4_t vcmlaq_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
36143 float32x4_t __ret;
36144 __ret = (float32x4_t) __builtin_neon_vcmlaq_f32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
36145 return __ret;
36146 }
36147 #else
vcmlaq_f32(float32x4_t __p0,float32x4_t __p1,float32x4_t __p2)36148 __ai __attribute__((target("v8.3a"))) float32x4_t vcmlaq_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
36149 float32x4_t __ret;
36150 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
36151 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
36152 float32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
36153 __ret = (float32x4_t) __builtin_neon_vcmlaq_f32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 41);
36154 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
36155 return __ret;
36156 }
__noswap_vcmlaq_f32(float32x4_t __p0,float32x4_t __p1,float32x4_t __p2)36157 __ai __attribute__((target("v8.3a"))) float32x4_t __noswap_vcmlaq_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
36158 float32x4_t __ret;
36159 __ret = (float32x4_t) __builtin_neon_vcmlaq_f32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
36160 return __ret;
36161 }
36162 #endif
36163
36164 #ifdef __LITTLE_ENDIAN__
vcmla_f32(float32x2_t __p0,float32x2_t __p1,float32x2_t __p2)36165 __ai __attribute__((target("v8.3a"))) float32x2_t vcmla_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
36166 float32x2_t __ret;
36167 __ret = (float32x2_t) __builtin_neon_vcmla_f32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 9);
36168 return __ret;
36169 }
36170 #else
vcmla_f32(float32x2_t __p0,float32x2_t __p1,float32x2_t __p2)36171 __ai __attribute__((target("v8.3a"))) float32x2_t vcmla_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
36172 float32x2_t __ret;
36173 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
36174 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
36175 float32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
36176 __ret = (float32x2_t) __builtin_neon_vcmla_f32((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 9);
36177 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
36178 return __ret;
36179 }
__noswap_vcmla_f32(float32x2_t __p0,float32x2_t __p1,float32x2_t __p2)36180 __ai __attribute__((target("v8.3a"))) float32x2_t __noswap_vcmla_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
36181 float32x2_t __ret;
36182 __ret = (float32x2_t) __builtin_neon_vcmla_f32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 9);
36183 return __ret;
36184 }
36185 #endif
36186
36187 #ifdef __LITTLE_ENDIAN__
36188 #define vcmla_lane_f32(__p0_177, __p1_177, __p2_177, __p3_177) __extension__ ({ \
36189 float32x2_t __ret_177; \
36190 float32x2_t __s0_177 = __p0_177; \
36191 float32x2_t __s1_177 = __p1_177; \
36192 float32x2_t __s2_177 = __p2_177; \
36193 float32x2_t __reint_177 = __s2_177; \
36194 uint64x1_t __reint1_177 = (uint64x1_t) {vget_lane_u64(*(uint64x1_t *) &__reint_177, __p3_177)}; \
36195 __ret_177 = vcmla_f32(__s0_177, __s1_177, *(float32x2_t *) &__reint1_177); \
36196 __ret_177; \
36197 })
36198 #else
36199 #define vcmla_lane_f32(__p0_178, __p1_178, __p2_178, __p3_178) __extension__ ({ \
36200 float32x2_t __ret_178; \
36201 float32x2_t __s0_178 = __p0_178; \
36202 float32x2_t __s1_178 = __p1_178; \
36203 float32x2_t __s2_178 = __p2_178; \
36204 float32x2_t __rev0_178; __rev0_178 = __builtin_shufflevector(__s0_178, __s0_178, 1, 0); \
36205 float32x2_t __rev1_178; __rev1_178 = __builtin_shufflevector(__s1_178, __s1_178, 1, 0); \
36206 float32x2_t __rev2_178; __rev2_178 = __builtin_shufflevector(__s2_178, __s2_178, 1, 0); \
36207 float32x2_t __reint_178 = __rev2_178; \
36208 uint64x1_t __reint1_178 = (uint64x1_t) {vget_lane_u64(*(uint64x1_t *) &__reint_178, __p3_178)}; \
36209 __ret_178 = __noswap_vcmla_f32(__rev0_178, __rev1_178, *(float32x2_t *) &__reint1_178); \
36210 __ret_178 = __builtin_shufflevector(__ret_178, __ret_178, 1, 0); \
36211 __ret_178; \
36212 })
36213 #endif
36214
36215 #ifdef __LITTLE_ENDIAN__
36216 #define vcmlaq_lane_f32(__p0_179, __p1_179, __p2_179, __p3_179) __extension__ ({ \
36217 float32x4_t __ret_179; \
36218 float32x4_t __s0_179 = __p0_179; \
36219 float32x4_t __s1_179 = __p1_179; \
36220 float32x2_t __s2_179 = __p2_179; \
36221 float32x2_t __reint_179 = __s2_179; \
36222 uint64x2_t __reint1_179 = (uint64x2_t) {vget_lane_u64(*(uint64x1_t *) &__reint_179, __p3_179), vget_lane_u64(*(uint64x1_t *) &__reint_179, __p3_179)}; \
36223 __ret_179 = vcmlaq_f32(__s0_179, __s1_179, *(float32x4_t *) &__reint1_179); \
36224 __ret_179; \
36225 })
36226 #else
36227 #define vcmlaq_lane_f32(__p0_180, __p1_180, __p2_180, __p3_180) __extension__ ({ \
36228 float32x4_t __ret_180; \
36229 float32x4_t __s0_180 = __p0_180; \
36230 float32x4_t __s1_180 = __p1_180; \
36231 float32x2_t __s2_180 = __p2_180; \
36232 float32x4_t __rev0_180; __rev0_180 = __builtin_shufflevector(__s0_180, __s0_180, 3, 2, 1, 0); \
36233 float32x4_t __rev1_180; __rev1_180 = __builtin_shufflevector(__s1_180, __s1_180, 3, 2, 1, 0); \
36234 float32x2_t __rev2_180; __rev2_180 = __builtin_shufflevector(__s2_180, __s2_180, 1, 0); \
36235 float32x2_t __reint_180 = __rev2_180; \
36236 uint64x2_t __reint1_180 = (uint64x2_t) {vget_lane_u64(*(uint64x1_t *) &__reint_180, __p3_180), vget_lane_u64(*(uint64x1_t *) &__reint_180, __p3_180)}; \
36237 __ret_180 = __noswap_vcmlaq_f32(__rev0_180, __rev1_180, *(float32x4_t *) &__reint1_180); \
36238 __ret_180 = __builtin_shufflevector(__ret_180, __ret_180, 3, 2, 1, 0); \
36239 __ret_180; \
36240 })
36241 #endif
36242
36243 #ifdef __LITTLE_ENDIAN__
36244 #define vcmla_laneq_f32(__p0_181, __p1_181, __p2_181, __p3_181) __extension__ ({ \
36245 float32x2_t __ret_181; \
36246 float32x2_t __s0_181 = __p0_181; \
36247 float32x2_t __s1_181 = __p1_181; \
36248 float32x4_t __s2_181 = __p2_181; \
36249 float32x4_t __reint_181 = __s2_181; \
36250 uint64x1_t __reint1_181 = (uint64x1_t) {vgetq_lane_u64(*(uint64x2_t *) &__reint_181, __p3_181)}; \
36251 __ret_181 = vcmla_f32(__s0_181, __s1_181, *(float32x2_t *) &__reint1_181); \
36252 __ret_181; \
36253 })
36254 #else
36255 #define vcmla_laneq_f32(__p0_182, __p1_182, __p2_182, __p3_182) __extension__ ({ \
36256 float32x2_t __ret_182; \
36257 float32x2_t __s0_182 = __p0_182; \
36258 float32x2_t __s1_182 = __p1_182; \
36259 float32x4_t __s2_182 = __p2_182; \
36260 float32x2_t __rev0_182; __rev0_182 = __builtin_shufflevector(__s0_182, __s0_182, 1, 0); \
36261 float32x2_t __rev1_182; __rev1_182 = __builtin_shufflevector(__s1_182, __s1_182, 1, 0); \
36262 float32x4_t __rev2_182; __rev2_182 = __builtin_shufflevector(__s2_182, __s2_182, 3, 2, 1, 0); \
36263 float32x4_t __reint_182 = __rev2_182; \
36264 uint64x1_t __reint1_182 = (uint64x1_t) {__noswap_vgetq_lane_u64(*(uint64x2_t *) &__reint_182, __p3_182)}; \
36265 __ret_182 = __noswap_vcmla_f32(__rev0_182, __rev1_182, *(float32x2_t *) &__reint1_182); \
36266 __ret_182 = __builtin_shufflevector(__ret_182, __ret_182, 1, 0); \
36267 __ret_182; \
36268 })
36269 #endif
36270
36271 #ifdef __LITTLE_ENDIAN__
36272 #define vcmlaq_laneq_f32(__p0_183, __p1_183, __p2_183, __p3_183) __extension__ ({ \
36273 float32x4_t __ret_183; \
36274 float32x4_t __s0_183 = __p0_183; \
36275 float32x4_t __s1_183 = __p1_183; \
36276 float32x4_t __s2_183 = __p2_183; \
36277 float32x4_t __reint_183 = __s2_183; \
36278 uint64x2_t __reint1_183 = (uint64x2_t) {vgetq_lane_u64(*(uint64x2_t *) &__reint_183, __p3_183), vgetq_lane_u64(*(uint64x2_t *) &__reint_183, __p3_183)}; \
36279 __ret_183 = vcmlaq_f32(__s0_183, __s1_183, *(float32x4_t *) &__reint1_183); \
36280 __ret_183; \
36281 })
36282 #else
36283 #define vcmlaq_laneq_f32(__p0_184, __p1_184, __p2_184, __p3_184) __extension__ ({ \
36284 float32x4_t __ret_184; \
36285 float32x4_t __s0_184 = __p0_184; \
36286 float32x4_t __s1_184 = __p1_184; \
36287 float32x4_t __s2_184 = __p2_184; \
36288 float32x4_t __rev0_184; __rev0_184 = __builtin_shufflevector(__s0_184, __s0_184, 3, 2, 1, 0); \
36289 float32x4_t __rev1_184; __rev1_184 = __builtin_shufflevector(__s1_184, __s1_184, 3, 2, 1, 0); \
36290 float32x4_t __rev2_184; __rev2_184 = __builtin_shufflevector(__s2_184, __s2_184, 3, 2, 1, 0); \
36291 float32x4_t __reint_184 = __rev2_184; \
36292 uint64x2_t __reint1_184 = (uint64x2_t) {__noswap_vgetq_lane_u64(*(uint64x2_t *) &__reint_184, __p3_184), __noswap_vgetq_lane_u64(*(uint64x2_t *) &__reint_184, __p3_184)}; \
36293 __ret_184 = __noswap_vcmlaq_f32(__rev0_184, __rev1_184, *(float32x4_t *) &__reint1_184); \
36294 __ret_184 = __builtin_shufflevector(__ret_184, __ret_184, 3, 2, 1, 0); \
36295 __ret_184; \
36296 })
36297 #endif
36298
36299 #ifdef __LITTLE_ENDIAN__
vcmlaq_rot180_f32(float32x4_t __p0,float32x4_t __p1,float32x4_t __p2)36300 __ai __attribute__((target("v8.3a"))) float32x4_t vcmlaq_rot180_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
36301 float32x4_t __ret;
36302 __ret = (float32x4_t) __builtin_neon_vcmlaq_rot180_f32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
36303 return __ret;
36304 }
36305 #else
vcmlaq_rot180_f32(float32x4_t __p0,float32x4_t __p1,float32x4_t __p2)36306 __ai __attribute__((target("v8.3a"))) float32x4_t vcmlaq_rot180_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
36307 float32x4_t __ret;
36308 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
36309 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
36310 float32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
36311 __ret = (float32x4_t) __builtin_neon_vcmlaq_rot180_f32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 41);
36312 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
36313 return __ret;
36314 }
__noswap_vcmlaq_rot180_f32(float32x4_t __p0,float32x4_t __p1,float32x4_t __p2)36315 __ai __attribute__((target("v8.3a"))) float32x4_t __noswap_vcmlaq_rot180_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
36316 float32x4_t __ret;
36317 __ret = (float32x4_t) __builtin_neon_vcmlaq_rot180_f32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
36318 return __ret;
36319 }
36320 #endif
36321
36322 #ifdef __LITTLE_ENDIAN__
vcmla_rot180_f32(float32x2_t __p0,float32x2_t __p1,float32x2_t __p2)36323 __ai __attribute__((target("v8.3a"))) float32x2_t vcmla_rot180_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
36324 float32x2_t __ret;
36325 __ret = (float32x2_t) __builtin_neon_vcmla_rot180_f32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 9);
36326 return __ret;
36327 }
36328 #else
vcmla_rot180_f32(float32x2_t __p0,float32x2_t __p1,float32x2_t __p2)36329 __ai __attribute__((target("v8.3a"))) float32x2_t vcmla_rot180_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
36330 float32x2_t __ret;
36331 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
36332 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
36333 float32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
36334 __ret = (float32x2_t) __builtin_neon_vcmla_rot180_f32((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 9);
36335 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
36336 return __ret;
36337 }
__noswap_vcmla_rot180_f32(float32x2_t __p0,float32x2_t __p1,float32x2_t __p2)36338 __ai __attribute__((target("v8.3a"))) float32x2_t __noswap_vcmla_rot180_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
36339 float32x2_t __ret;
36340 __ret = (float32x2_t) __builtin_neon_vcmla_rot180_f32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 9);
36341 return __ret;
36342 }
36343 #endif
36344
36345 #ifdef __LITTLE_ENDIAN__
36346 #define vcmla_rot180_lane_f32(__p0_185, __p1_185, __p2_185, __p3_185) __extension__ ({ \
36347 float32x2_t __ret_185; \
36348 float32x2_t __s0_185 = __p0_185; \
36349 float32x2_t __s1_185 = __p1_185; \
36350 float32x2_t __s2_185 = __p2_185; \
36351 float32x2_t __reint_185 = __s2_185; \
36352 uint64x1_t __reint1_185 = (uint64x1_t) {vget_lane_u64(*(uint64x1_t *) &__reint_185, __p3_185)}; \
36353 __ret_185 = vcmla_rot180_f32(__s0_185, __s1_185, *(float32x2_t *) &__reint1_185); \
36354 __ret_185; \
36355 })
36356 #else
36357 #define vcmla_rot180_lane_f32(__p0_186, __p1_186, __p2_186, __p3_186) __extension__ ({ \
36358 float32x2_t __ret_186; \
36359 float32x2_t __s0_186 = __p0_186; \
36360 float32x2_t __s1_186 = __p1_186; \
36361 float32x2_t __s2_186 = __p2_186; \
36362 float32x2_t __rev0_186; __rev0_186 = __builtin_shufflevector(__s0_186, __s0_186, 1, 0); \
36363 float32x2_t __rev1_186; __rev1_186 = __builtin_shufflevector(__s1_186, __s1_186, 1, 0); \
36364 float32x2_t __rev2_186; __rev2_186 = __builtin_shufflevector(__s2_186, __s2_186, 1, 0); \
36365 float32x2_t __reint_186 = __rev2_186; \
36366 uint64x1_t __reint1_186 = (uint64x1_t) {vget_lane_u64(*(uint64x1_t *) &__reint_186, __p3_186)}; \
36367 __ret_186 = __noswap_vcmla_rot180_f32(__rev0_186, __rev1_186, *(float32x2_t *) &__reint1_186); \
36368 __ret_186 = __builtin_shufflevector(__ret_186, __ret_186, 1, 0); \
36369 __ret_186; \
36370 })
36371 #endif
36372
36373 #ifdef __LITTLE_ENDIAN__
36374 #define vcmlaq_rot180_lane_f32(__p0_187, __p1_187, __p2_187, __p3_187) __extension__ ({ \
36375 float32x4_t __ret_187; \
36376 float32x4_t __s0_187 = __p0_187; \
36377 float32x4_t __s1_187 = __p1_187; \
36378 float32x2_t __s2_187 = __p2_187; \
36379 float32x2_t __reint_187 = __s2_187; \
36380 uint64x2_t __reint1_187 = (uint64x2_t) {vget_lane_u64(*(uint64x1_t *) &__reint_187, __p3_187), vget_lane_u64(*(uint64x1_t *) &__reint_187, __p3_187)}; \
36381 __ret_187 = vcmlaq_rot180_f32(__s0_187, __s1_187, *(float32x4_t *) &__reint1_187); \
36382 __ret_187; \
36383 })
36384 #else
36385 #define vcmlaq_rot180_lane_f32(__p0_188, __p1_188, __p2_188, __p3_188) __extension__ ({ \
36386 float32x4_t __ret_188; \
36387 float32x4_t __s0_188 = __p0_188; \
36388 float32x4_t __s1_188 = __p1_188; \
36389 float32x2_t __s2_188 = __p2_188; \
36390 float32x4_t __rev0_188; __rev0_188 = __builtin_shufflevector(__s0_188, __s0_188, 3, 2, 1, 0); \
36391 float32x4_t __rev1_188; __rev1_188 = __builtin_shufflevector(__s1_188, __s1_188, 3, 2, 1, 0); \
36392 float32x2_t __rev2_188; __rev2_188 = __builtin_shufflevector(__s2_188, __s2_188, 1, 0); \
36393 float32x2_t __reint_188 = __rev2_188; \
36394 uint64x2_t __reint1_188 = (uint64x2_t) {vget_lane_u64(*(uint64x1_t *) &__reint_188, __p3_188), vget_lane_u64(*(uint64x1_t *) &__reint_188, __p3_188)}; \
36395 __ret_188 = __noswap_vcmlaq_rot180_f32(__rev0_188, __rev1_188, *(float32x4_t *) &__reint1_188); \
36396 __ret_188 = __builtin_shufflevector(__ret_188, __ret_188, 3, 2, 1, 0); \
36397 __ret_188; \
36398 })
36399 #endif
36400
36401 #ifdef __LITTLE_ENDIAN__
36402 #define vcmla_rot180_laneq_f32(__p0_189, __p1_189, __p2_189, __p3_189) __extension__ ({ \
36403 float32x2_t __ret_189; \
36404 float32x2_t __s0_189 = __p0_189; \
36405 float32x2_t __s1_189 = __p1_189; \
36406 float32x4_t __s2_189 = __p2_189; \
36407 float32x4_t __reint_189 = __s2_189; \
36408 uint64x1_t __reint1_189 = (uint64x1_t) {vgetq_lane_u64(*(uint64x2_t *) &__reint_189, __p3_189)}; \
36409 __ret_189 = vcmla_rot180_f32(__s0_189, __s1_189, *(float32x2_t *) &__reint1_189); \
36410 __ret_189; \
36411 })
36412 #else
36413 #define vcmla_rot180_laneq_f32(__p0_190, __p1_190, __p2_190, __p3_190) __extension__ ({ \
36414 float32x2_t __ret_190; \
36415 float32x2_t __s0_190 = __p0_190; \
36416 float32x2_t __s1_190 = __p1_190; \
36417 float32x4_t __s2_190 = __p2_190; \
36418 float32x2_t __rev0_190; __rev0_190 = __builtin_shufflevector(__s0_190, __s0_190, 1, 0); \
36419 float32x2_t __rev1_190; __rev1_190 = __builtin_shufflevector(__s1_190, __s1_190, 1, 0); \
36420 float32x4_t __rev2_190; __rev2_190 = __builtin_shufflevector(__s2_190, __s2_190, 3, 2, 1, 0); \
36421 float32x4_t __reint_190 = __rev2_190; \
36422 uint64x1_t __reint1_190 = (uint64x1_t) {__noswap_vgetq_lane_u64(*(uint64x2_t *) &__reint_190, __p3_190)}; \
36423 __ret_190 = __noswap_vcmla_rot180_f32(__rev0_190, __rev1_190, *(float32x2_t *) &__reint1_190); \
36424 __ret_190 = __builtin_shufflevector(__ret_190, __ret_190, 1, 0); \
36425 __ret_190; \
36426 })
36427 #endif
36428
36429 #ifdef __LITTLE_ENDIAN__
36430 #define vcmlaq_rot180_laneq_f32(__p0_191, __p1_191, __p2_191, __p3_191) __extension__ ({ \
36431 float32x4_t __ret_191; \
36432 float32x4_t __s0_191 = __p0_191; \
36433 float32x4_t __s1_191 = __p1_191; \
36434 float32x4_t __s2_191 = __p2_191; \
36435 float32x4_t __reint_191 = __s2_191; \
36436 uint64x2_t __reint1_191 = (uint64x2_t) {vgetq_lane_u64(*(uint64x2_t *) &__reint_191, __p3_191), vgetq_lane_u64(*(uint64x2_t *) &__reint_191, __p3_191)}; \
36437 __ret_191 = vcmlaq_rot180_f32(__s0_191, __s1_191, *(float32x4_t *) &__reint1_191); \
36438 __ret_191; \
36439 })
36440 #else
36441 #define vcmlaq_rot180_laneq_f32(__p0_192, __p1_192, __p2_192, __p3_192) __extension__ ({ \
36442 float32x4_t __ret_192; \
36443 float32x4_t __s0_192 = __p0_192; \
36444 float32x4_t __s1_192 = __p1_192; \
36445 float32x4_t __s2_192 = __p2_192; \
36446 float32x4_t __rev0_192; __rev0_192 = __builtin_shufflevector(__s0_192, __s0_192, 3, 2, 1, 0); \
36447 float32x4_t __rev1_192; __rev1_192 = __builtin_shufflevector(__s1_192, __s1_192, 3, 2, 1, 0); \
36448 float32x4_t __rev2_192; __rev2_192 = __builtin_shufflevector(__s2_192, __s2_192, 3, 2, 1, 0); \
36449 float32x4_t __reint_192 = __rev2_192; \
36450 uint64x2_t __reint1_192 = (uint64x2_t) {__noswap_vgetq_lane_u64(*(uint64x2_t *) &__reint_192, __p3_192), __noswap_vgetq_lane_u64(*(uint64x2_t *) &__reint_192, __p3_192)}; \
36451 __ret_192 = __noswap_vcmlaq_rot180_f32(__rev0_192, __rev1_192, *(float32x4_t *) &__reint1_192); \
36452 __ret_192 = __builtin_shufflevector(__ret_192, __ret_192, 3, 2, 1, 0); \
36453 __ret_192; \
36454 })
36455 #endif
36456
36457 #ifdef __LITTLE_ENDIAN__
vcmlaq_rot270_f32(float32x4_t __p0,float32x4_t __p1,float32x4_t __p2)36458 __ai __attribute__((target("v8.3a"))) float32x4_t vcmlaq_rot270_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
36459 float32x4_t __ret;
36460 __ret = (float32x4_t) __builtin_neon_vcmlaq_rot270_f32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
36461 return __ret;
36462 }
36463 #else
vcmlaq_rot270_f32(float32x4_t __p0,float32x4_t __p1,float32x4_t __p2)36464 __ai __attribute__((target("v8.3a"))) float32x4_t vcmlaq_rot270_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
36465 float32x4_t __ret;
36466 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
36467 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
36468 float32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
36469 __ret = (float32x4_t) __builtin_neon_vcmlaq_rot270_f32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 41);
36470 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
36471 return __ret;
36472 }
__noswap_vcmlaq_rot270_f32(float32x4_t __p0,float32x4_t __p1,float32x4_t __p2)36473 __ai __attribute__((target("v8.3a"))) float32x4_t __noswap_vcmlaq_rot270_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
36474 float32x4_t __ret;
36475 __ret = (float32x4_t) __builtin_neon_vcmlaq_rot270_f32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
36476 return __ret;
36477 }
36478 #endif
36479
36480 #ifdef __LITTLE_ENDIAN__
vcmla_rot270_f32(float32x2_t __p0,float32x2_t __p1,float32x2_t __p2)36481 __ai __attribute__((target("v8.3a"))) float32x2_t vcmla_rot270_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
36482 float32x2_t __ret;
36483 __ret = (float32x2_t) __builtin_neon_vcmla_rot270_f32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 9);
36484 return __ret;
36485 }
36486 #else
vcmla_rot270_f32(float32x2_t __p0,float32x2_t __p1,float32x2_t __p2)36487 __ai __attribute__((target("v8.3a"))) float32x2_t vcmla_rot270_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
36488 float32x2_t __ret;
36489 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
36490 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
36491 float32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
36492 __ret = (float32x2_t) __builtin_neon_vcmla_rot270_f32((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 9);
36493 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
36494 return __ret;
36495 }
__noswap_vcmla_rot270_f32(float32x2_t __p0,float32x2_t __p1,float32x2_t __p2)36496 __ai __attribute__((target("v8.3a"))) float32x2_t __noswap_vcmla_rot270_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
36497 float32x2_t __ret;
36498 __ret = (float32x2_t) __builtin_neon_vcmla_rot270_f32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 9);
36499 return __ret;
36500 }
36501 #endif
36502
36503 #ifdef __LITTLE_ENDIAN__
36504 #define vcmla_rot270_lane_f32(__p0_193, __p1_193, __p2_193, __p3_193) __extension__ ({ \
36505 float32x2_t __ret_193; \
36506 float32x2_t __s0_193 = __p0_193; \
36507 float32x2_t __s1_193 = __p1_193; \
36508 float32x2_t __s2_193 = __p2_193; \
36509 float32x2_t __reint_193 = __s2_193; \
36510 uint64x1_t __reint1_193 = (uint64x1_t) {vget_lane_u64(*(uint64x1_t *) &__reint_193, __p3_193)}; \
36511 __ret_193 = vcmla_rot270_f32(__s0_193, __s1_193, *(float32x2_t *) &__reint1_193); \
36512 __ret_193; \
36513 })
36514 #else
36515 #define vcmla_rot270_lane_f32(__p0_194, __p1_194, __p2_194, __p3_194) __extension__ ({ \
36516 float32x2_t __ret_194; \
36517 float32x2_t __s0_194 = __p0_194; \
36518 float32x2_t __s1_194 = __p1_194; \
36519 float32x2_t __s2_194 = __p2_194; \
36520 float32x2_t __rev0_194; __rev0_194 = __builtin_shufflevector(__s0_194, __s0_194, 1, 0); \
36521 float32x2_t __rev1_194; __rev1_194 = __builtin_shufflevector(__s1_194, __s1_194, 1, 0); \
36522 float32x2_t __rev2_194; __rev2_194 = __builtin_shufflevector(__s2_194, __s2_194, 1, 0); \
36523 float32x2_t __reint_194 = __rev2_194; \
36524 uint64x1_t __reint1_194 = (uint64x1_t) {vget_lane_u64(*(uint64x1_t *) &__reint_194, __p3_194)}; \
36525 __ret_194 = __noswap_vcmla_rot270_f32(__rev0_194, __rev1_194, *(float32x2_t *) &__reint1_194); \
36526 __ret_194 = __builtin_shufflevector(__ret_194, __ret_194, 1, 0); \
36527 __ret_194; \
36528 })
36529 #endif
36530
36531 #ifdef __LITTLE_ENDIAN__
36532 #define vcmlaq_rot270_lane_f32(__p0_195, __p1_195, __p2_195, __p3_195) __extension__ ({ \
36533 float32x4_t __ret_195; \
36534 float32x4_t __s0_195 = __p0_195; \
36535 float32x4_t __s1_195 = __p1_195; \
36536 float32x2_t __s2_195 = __p2_195; \
36537 float32x2_t __reint_195 = __s2_195; \
36538 uint64x2_t __reint1_195 = (uint64x2_t) {vget_lane_u64(*(uint64x1_t *) &__reint_195, __p3_195), vget_lane_u64(*(uint64x1_t *) &__reint_195, __p3_195)}; \
36539 __ret_195 = vcmlaq_rot270_f32(__s0_195, __s1_195, *(float32x4_t *) &__reint1_195); \
36540 __ret_195; \
36541 })
36542 #else
36543 #define vcmlaq_rot270_lane_f32(__p0_196, __p1_196, __p2_196, __p3_196) __extension__ ({ \
36544 float32x4_t __ret_196; \
36545 float32x4_t __s0_196 = __p0_196; \
36546 float32x4_t __s1_196 = __p1_196; \
36547 float32x2_t __s2_196 = __p2_196; \
36548 float32x4_t __rev0_196; __rev0_196 = __builtin_shufflevector(__s0_196, __s0_196, 3, 2, 1, 0); \
36549 float32x4_t __rev1_196; __rev1_196 = __builtin_shufflevector(__s1_196, __s1_196, 3, 2, 1, 0); \
36550 float32x2_t __rev2_196; __rev2_196 = __builtin_shufflevector(__s2_196, __s2_196, 1, 0); \
36551 float32x2_t __reint_196 = __rev2_196; \
36552 uint64x2_t __reint1_196 = (uint64x2_t) {vget_lane_u64(*(uint64x1_t *) &__reint_196, __p3_196), vget_lane_u64(*(uint64x1_t *) &__reint_196, __p3_196)}; \
36553 __ret_196 = __noswap_vcmlaq_rot270_f32(__rev0_196, __rev1_196, *(float32x4_t *) &__reint1_196); \
36554 __ret_196 = __builtin_shufflevector(__ret_196, __ret_196, 3, 2, 1, 0); \
36555 __ret_196; \
36556 })
36557 #endif
36558
36559 #ifdef __LITTLE_ENDIAN__
36560 #define vcmla_rot270_laneq_f32(__p0_197, __p1_197, __p2_197, __p3_197) __extension__ ({ \
36561 float32x2_t __ret_197; \
36562 float32x2_t __s0_197 = __p0_197; \
36563 float32x2_t __s1_197 = __p1_197; \
36564 float32x4_t __s2_197 = __p2_197; \
36565 float32x4_t __reint_197 = __s2_197; \
36566 uint64x1_t __reint1_197 = (uint64x1_t) {vgetq_lane_u64(*(uint64x2_t *) &__reint_197, __p3_197)}; \
36567 __ret_197 = vcmla_rot270_f32(__s0_197, __s1_197, *(float32x2_t *) &__reint1_197); \
36568 __ret_197; \
36569 })
36570 #else
36571 #define vcmla_rot270_laneq_f32(__p0_198, __p1_198, __p2_198, __p3_198) __extension__ ({ \
36572 float32x2_t __ret_198; \
36573 float32x2_t __s0_198 = __p0_198; \
36574 float32x2_t __s1_198 = __p1_198; \
36575 float32x4_t __s2_198 = __p2_198; \
36576 float32x2_t __rev0_198; __rev0_198 = __builtin_shufflevector(__s0_198, __s0_198, 1, 0); \
36577 float32x2_t __rev1_198; __rev1_198 = __builtin_shufflevector(__s1_198, __s1_198, 1, 0); \
36578 float32x4_t __rev2_198; __rev2_198 = __builtin_shufflevector(__s2_198, __s2_198, 3, 2, 1, 0); \
36579 float32x4_t __reint_198 = __rev2_198; \
36580 uint64x1_t __reint1_198 = (uint64x1_t) {__noswap_vgetq_lane_u64(*(uint64x2_t *) &__reint_198, __p3_198)}; \
36581 __ret_198 = __noswap_vcmla_rot270_f32(__rev0_198, __rev1_198, *(float32x2_t *) &__reint1_198); \
36582 __ret_198 = __builtin_shufflevector(__ret_198, __ret_198, 1, 0); \
36583 __ret_198; \
36584 })
36585 #endif
36586
36587 #ifdef __LITTLE_ENDIAN__
36588 #define vcmlaq_rot270_laneq_f32(__p0_199, __p1_199, __p2_199, __p3_199) __extension__ ({ \
36589 float32x4_t __ret_199; \
36590 float32x4_t __s0_199 = __p0_199; \
36591 float32x4_t __s1_199 = __p1_199; \
36592 float32x4_t __s2_199 = __p2_199; \
36593 float32x4_t __reint_199 = __s2_199; \
36594 uint64x2_t __reint1_199 = (uint64x2_t) {vgetq_lane_u64(*(uint64x2_t *) &__reint_199, __p3_199), vgetq_lane_u64(*(uint64x2_t *) &__reint_199, __p3_199)}; \
36595 __ret_199 = vcmlaq_rot270_f32(__s0_199, __s1_199, *(float32x4_t *) &__reint1_199); \
36596 __ret_199; \
36597 })
36598 #else
36599 #define vcmlaq_rot270_laneq_f32(__p0_200, __p1_200, __p2_200, __p3_200) __extension__ ({ \
36600 float32x4_t __ret_200; \
36601 float32x4_t __s0_200 = __p0_200; \
36602 float32x4_t __s1_200 = __p1_200; \
36603 float32x4_t __s2_200 = __p2_200; \
36604 float32x4_t __rev0_200; __rev0_200 = __builtin_shufflevector(__s0_200, __s0_200, 3, 2, 1, 0); \
36605 float32x4_t __rev1_200; __rev1_200 = __builtin_shufflevector(__s1_200, __s1_200, 3, 2, 1, 0); \
36606 float32x4_t __rev2_200; __rev2_200 = __builtin_shufflevector(__s2_200, __s2_200, 3, 2, 1, 0); \
36607 float32x4_t __reint_200 = __rev2_200; \
36608 uint64x2_t __reint1_200 = (uint64x2_t) {__noswap_vgetq_lane_u64(*(uint64x2_t *) &__reint_200, __p3_200), __noswap_vgetq_lane_u64(*(uint64x2_t *) &__reint_200, __p3_200)}; \
36609 __ret_200 = __noswap_vcmlaq_rot270_f32(__rev0_200, __rev1_200, *(float32x4_t *) &__reint1_200); \
36610 __ret_200 = __builtin_shufflevector(__ret_200, __ret_200, 3, 2, 1, 0); \
36611 __ret_200; \
36612 })
36613 #endif
36614
36615 #ifdef __LITTLE_ENDIAN__
vcmlaq_rot90_f32(float32x4_t __p0,float32x4_t __p1,float32x4_t __p2)36616 __ai __attribute__((target("v8.3a"))) float32x4_t vcmlaq_rot90_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
36617 float32x4_t __ret;
36618 __ret = (float32x4_t) __builtin_neon_vcmlaq_rot90_f32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
36619 return __ret;
36620 }
36621 #else
vcmlaq_rot90_f32(float32x4_t __p0,float32x4_t __p1,float32x4_t __p2)36622 __ai __attribute__((target("v8.3a"))) float32x4_t vcmlaq_rot90_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
36623 float32x4_t __ret;
36624 float32x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
36625 float32x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
36626 float32x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
36627 __ret = (float32x4_t) __builtin_neon_vcmlaq_rot90_f32((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 41);
36628 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
36629 return __ret;
36630 }
__noswap_vcmlaq_rot90_f32(float32x4_t __p0,float32x4_t __p1,float32x4_t __p2)36631 __ai __attribute__((target("v8.3a"))) float32x4_t __noswap_vcmlaq_rot90_f32(float32x4_t __p0, float32x4_t __p1, float32x4_t __p2) {
36632 float32x4_t __ret;
36633 __ret = (float32x4_t) __builtin_neon_vcmlaq_rot90_f32((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 41);
36634 return __ret;
36635 }
36636 #endif
36637
36638 #ifdef __LITTLE_ENDIAN__
vcmla_rot90_f32(float32x2_t __p0,float32x2_t __p1,float32x2_t __p2)36639 __ai __attribute__((target("v8.3a"))) float32x2_t vcmla_rot90_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
36640 float32x2_t __ret;
36641 __ret = (float32x2_t) __builtin_neon_vcmla_rot90_f32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 9);
36642 return __ret;
36643 }
36644 #else
vcmla_rot90_f32(float32x2_t __p0,float32x2_t __p1,float32x2_t __p2)36645 __ai __attribute__((target("v8.3a"))) float32x2_t vcmla_rot90_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
36646 float32x2_t __ret;
36647 float32x2_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 1, 0);
36648 float32x2_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 1, 0);
36649 float32x2_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 1, 0);
36650 __ret = (float32x2_t) __builtin_neon_vcmla_rot90_f32((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 9);
36651 __ret = __builtin_shufflevector(__ret, __ret, 1, 0);
36652 return __ret;
36653 }
__noswap_vcmla_rot90_f32(float32x2_t __p0,float32x2_t __p1,float32x2_t __p2)36654 __ai __attribute__((target("v8.3a"))) float32x2_t __noswap_vcmla_rot90_f32(float32x2_t __p0, float32x2_t __p1, float32x2_t __p2) {
36655 float32x2_t __ret;
36656 __ret = (float32x2_t) __builtin_neon_vcmla_rot90_f32((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 9);
36657 return __ret;
36658 }
36659 #endif
36660
36661 #ifdef __LITTLE_ENDIAN__
36662 #define vcmla_rot90_lane_f32(__p0_201, __p1_201, __p2_201, __p3_201) __extension__ ({ \
36663 float32x2_t __ret_201; \
36664 float32x2_t __s0_201 = __p0_201; \
36665 float32x2_t __s1_201 = __p1_201; \
36666 float32x2_t __s2_201 = __p2_201; \
36667 float32x2_t __reint_201 = __s2_201; \
36668 uint64x1_t __reint1_201 = (uint64x1_t) {vget_lane_u64(*(uint64x1_t *) &__reint_201, __p3_201)}; \
36669 __ret_201 = vcmla_rot90_f32(__s0_201, __s1_201, *(float32x2_t *) &__reint1_201); \
36670 __ret_201; \
36671 })
36672 #else
36673 #define vcmla_rot90_lane_f32(__p0_202, __p1_202, __p2_202, __p3_202) __extension__ ({ \
36674 float32x2_t __ret_202; \
36675 float32x2_t __s0_202 = __p0_202; \
36676 float32x2_t __s1_202 = __p1_202; \
36677 float32x2_t __s2_202 = __p2_202; \
36678 float32x2_t __rev0_202; __rev0_202 = __builtin_shufflevector(__s0_202, __s0_202, 1, 0); \
36679 float32x2_t __rev1_202; __rev1_202 = __builtin_shufflevector(__s1_202, __s1_202, 1, 0); \
36680 float32x2_t __rev2_202; __rev2_202 = __builtin_shufflevector(__s2_202, __s2_202, 1, 0); \
36681 float32x2_t __reint_202 = __rev2_202; \
36682 uint64x1_t __reint1_202 = (uint64x1_t) {vget_lane_u64(*(uint64x1_t *) &__reint_202, __p3_202)}; \
36683 __ret_202 = __noswap_vcmla_rot90_f32(__rev0_202, __rev1_202, *(float32x2_t *) &__reint1_202); \
36684 __ret_202 = __builtin_shufflevector(__ret_202, __ret_202, 1, 0); \
36685 __ret_202; \
36686 })
36687 #endif
36688
36689 #ifdef __LITTLE_ENDIAN__
36690 #define vcmlaq_rot90_lane_f32(__p0_203, __p1_203, __p2_203, __p3_203) __extension__ ({ \
36691 float32x4_t __ret_203; \
36692 float32x4_t __s0_203 = __p0_203; \
36693 float32x4_t __s1_203 = __p1_203; \
36694 float32x2_t __s2_203 = __p2_203; \
36695 float32x2_t __reint_203 = __s2_203; \
36696 uint64x2_t __reint1_203 = (uint64x2_t) {vget_lane_u64(*(uint64x1_t *) &__reint_203, __p3_203), vget_lane_u64(*(uint64x1_t *) &__reint_203, __p3_203)}; \
36697 __ret_203 = vcmlaq_rot90_f32(__s0_203, __s1_203, *(float32x4_t *) &__reint1_203); \
36698 __ret_203; \
36699 })
36700 #else
36701 #define vcmlaq_rot90_lane_f32(__p0_204, __p1_204, __p2_204, __p3_204) __extension__ ({ \
36702 float32x4_t __ret_204; \
36703 float32x4_t __s0_204 = __p0_204; \
36704 float32x4_t __s1_204 = __p1_204; \
36705 float32x2_t __s2_204 = __p2_204; \
36706 float32x4_t __rev0_204; __rev0_204 = __builtin_shufflevector(__s0_204, __s0_204, 3, 2, 1, 0); \
36707 float32x4_t __rev1_204; __rev1_204 = __builtin_shufflevector(__s1_204, __s1_204, 3, 2, 1, 0); \
36708 float32x2_t __rev2_204; __rev2_204 = __builtin_shufflevector(__s2_204, __s2_204, 1, 0); \
36709 float32x2_t __reint_204 = __rev2_204; \
36710 uint64x2_t __reint1_204 = (uint64x2_t) {vget_lane_u64(*(uint64x1_t *) &__reint_204, __p3_204), vget_lane_u64(*(uint64x1_t *) &__reint_204, __p3_204)}; \
36711 __ret_204 = __noswap_vcmlaq_rot90_f32(__rev0_204, __rev1_204, *(float32x4_t *) &__reint1_204); \
36712 __ret_204 = __builtin_shufflevector(__ret_204, __ret_204, 3, 2, 1, 0); \
36713 __ret_204; \
36714 })
36715 #endif
36716
36717 #ifdef __LITTLE_ENDIAN__
36718 #define vcmla_rot90_laneq_f32(__p0_205, __p1_205, __p2_205, __p3_205) __extension__ ({ \
36719 float32x2_t __ret_205; \
36720 float32x2_t __s0_205 = __p0_205; \
36721 float32x2_t __s1_205 = __p1_205; \
36722 float32x4_t __s2_205 = __p2_205; \
36723 float32x4_t __reint_205 = __s2_205; \
36724 uint64x1_t __reint1_205 = (uint64x1_t) {vgetq_lane_u64(*(uint64x2_t *) &__reint_205, __p3_205)}; \
36725 __ret_205 = vcmla_rot90_f32(__s0_205, __s1_205, *(float32x2_t *) &__reint1_205); \
36726 __ret_205; \
36727 })
36728 #else
36729 #define vcmla_rot90_laneq_f32(__p0_206, __p1_206, __p2_206, __p3_206) __extension__ ({ \
36730 float32x2_t __ret_206; \
36731 float32x2_t __s0_206 = __p0_206; \
36732 float32x2_t __s1_206 = __p1_206; \
36733 float32x4_t __s2_206 = __p2_206; \
36734 float32x2_t __rev0_206; __rev0_206 = __builtin_shufflevector(__s0_206, __s0_206, 1, 0); \
36735 float32x2_t __rev1_206; __rev1_206 = __builtin_shufflevector(__s1_206, __s1_206, 1, 0); \
36736 float32x4_t __rev2_206; __rev2_206 = __builtin_shufflevector(__s2_206, __s2_206, 3, 2, 1, 0); \
36737 float32x4_t __reint_206 = __rev2_206; \
36738 uint64x1_t __reint1_206 = (uint64x1_t) {__noswap_vgetq_lane_u64(*(uint64x2_t *) &__reint_206, __p3_206)}; \
36739 __ret_206 = __noswap_vcmla_rot90_f32(__rev0_206, __rev1_206, *(float32x2_t *) &__reint1_206); \
36740 __ret_206 = __builtin_shufflevector(__ret_206, __ret_206, 1, 0); \
36741 __ret_206; \
36742 })
36743 #endif
36744
36745 #ifdef __LITTLE_ENDIAN__
36746 #define vcmlaq_rot90_laneq_f32(__p0_207, __p1_207, __p2_207, __p3_207) __extension__ ({ \
36747 float32x4_t __ret_207; \
36748 float32x4_t __s0_207 = __p0_207; \
36749 float32x4_t __s1_207 = __p1_207; \
36750 float32x4_t __s2_207 = __p2_207; \
36751 float32x4_t __reint_207 = __s2_207; \
36752 uint64x2_t __reint1_207 = (uint64x2_t) {vgetq_lane_u64(*(uint64x2_t *) &__reint_207, __p3_207), vgetq_lane_u64(*(uint64x2_t *) &__reint_207, __p3_207)}; \
36753 __ret_207 = vcmlaq_rot90_f32(__s0_207, __s1_207, *(float32x4_t *) &__reint1_207); \
36754 __ret_207; \
36755 })
36756 #else
36757 #define vcmlaq_rot90_laneq_f32(__p0_208, __p1_208, __p2_208, __p3_208) __extension__ ({ \
36758 float32x4_t __ret_208; \
36759 float32x4_t __s0_208 = __p0_208; \
36760 float32x4_t __s1_208 = __p1_208; \
36761 float32x4_t __s2_208 = __p2_208; \
36762 float32x4_t __rev0_208; __rev0_208 = __builtin_shufflevector(__s0_208, __s0_208, 3, 2, 1, 0); \
36763 float32x4_t __rev1_208; __rev1_208 = __builtin_shufflevector(__s1_208, __s1_208, 3, 2, 1, 0); \
36764 float32x4_t __rev2_208; __rev2_208 = __builtin_shufflevector(__s2_208, __s2_208, 3, 2, 1, 0); \
36765 float32x4_t __reint_208 = __rev2_208; \
36766 uint64x2_t __reint1_208 = (uint64x2_t) {__noswap_vgetq_lane_u64(*(uint64x2_t *) &__reint_208, __p3_208), __noswap_vgetq_lane_u64(*(uint64x2_t *) &__reint_208, __p3_208)}; \
36767 __ret_208 = __noswap_vcmlaq_rot90_f32(__rev0_208, __rev1_208, *(float32x4_t *) &__reint1_208); \
36768 __ret_208 = __builtin_shufflevector(__ret_208, __ret_208, 3, 2, 1, 0); \
36769 __ret_208; \
36770 })
36771 #endif
36772
36773 #ifdef __LITTLE_ENDIAN__
vcadd_rot270_f16(float16x4_t __p0,float16x4_t __p1)36774 __ai __attribute__((target("v8.3a,fullfp16"))) float16x4_t vcadd_rot270_f16(float16x4_t __p0, float16x4_t __p1) {
36775 float16x4_t __ret;
36776 __ret = (float16x4_t) __builtin_neon_vcadd_rot270_f16((int8x8_t)__p0, (int8x8_t)__p1, 8);
36777 return __ret;
36778 }
36779 #else
vcadd_rot270_f16(float16x4_t __p0,float16x4_t __p1)36780 __ai __attribute__((target("v8.3a,fullfp16"))) float16x4_t vcadd_rot270_f16(float16x4_t __p0, float16x4_t __p1) {
36781 float16x4_t __ret;
36782 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
36783 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
36784 __ret = (float16x4_t) __builtin_neon_vcadd_rot270_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 8);
36785 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
36786 return __ret;
36787 }
36788 #endif
36789
36790 #ifdef __LITTLE_ENDIAN__
vcadd_rot90_f16(float16x4_t __p0,float16x4_t __p1)36791 __ai __attribute__((target("v8.3a,fullfp16"))) float16x4_t vcadd_rot90_f16(float16x4_t __p0, float16x4_t __p1) {
36792 float16x4_t __ret;
36793 __ret = (float16x4_t) __builtin_neon_vcadd_rot90_f16((int8x8_t)__p0, (int8x8_t)__p1, 8);
36794 return __ret;
36795 }
36796 #else
vcadd_rot90_f16(float16x4_t __p0,float16x4_t __p1)36797 __ai __attribute__((target("v8.3a,fullfp16"))) float16x4_t vcadd_rot90_f16(float16x4_t __p0, float16x4_t __p1) {
36798 float16x4_t __ret;
36799 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
36800 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
36801 __ret = (float16x4_t) __builtin_neon_vcadd_rot90_f16((int8x8_t)__rev0, (int8x8_t)__rev1, 8);
36802 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
36803 return __ret;
36804 }
36805 #endif
36806
36807 #ifdef __LITTLE_ENDIAN__
vcaddq_rot270_f16(float16x8_t __p0,float16x8_t __p1)36808 __ai __attribute__((target("v8.3a,fullfp16"))) float16x8_t vcaddq_rot270_f16(float16x8_t __p0, float16x8_t __p1) {
36809 float16x8_t __ret;
36810 __ret = (float16x8_t) __builtin_neon_vcaddq_rot270_f16((int8x16_t)__p0, (int8x16_t)__p1, 40);
36811 return __ret;
36812 }
36813 #else
vcaddq_rot270_f16(float16x8_t __p0,float16x8_t __p1)36814 __ai __attribute__((target("v8.3a,fullfp16"))) float16x8_t vcaddq_rot270_f16(float16x8_t __p0, float16x8_t __p1) {
36815 float16x8_t __ret;
36816 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
36817 float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
36818 __ret = (float16x8_t) __builtin_neon_vcaddq_rot270_f16((int8x16_t)__rev0, (int8x16_t)__rev1, 40);
36819 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
36820 return __ret;
36821 }
36822 #endif
36823
36824 #ifdef __LITTLE_ENDIAN__
vcaddq_rot90_f16(float16x8_t __p0,float16x8_t __p1)36825 __ai __attribute__((target("v8.3a,fullfp16"))) float16x8_t vcaddq_rot90_f16(float16x8_t __p0, float16x8_t __p1) {
36826 float16x8_t __ret;
36827 __ret = (float16x8_t) __builtin_neon_vcaddq_rot90_f16((int8x16_t)__p0, (int8x16_t)__p1, 40);
36828 return __ret;
36829 }
36830 #else
vcaddq_rot90_f16(float16x8_t __p0,float16x8_t __p1)36831 __ai __attribute__((target("v8.3a,fullfp16"))) float16x8_t vcaddq_rot90_f16(float16x8_t __p0, float16x8_t __p1) {
36832 float16x8_t __ret;
36833 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
36834 float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
36835 __ret = (float16x8_t) __builtin_neon_vcaddq_rot90_f16((int8x16_t)__rev0, (int8x16_t)__rev1, 40);
36836 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
36837 return __ret;
36838 }
36839 #endif
36840
36841 #ifdef __LITTLE_ENDIAN__
vcmlaq_f16(float16x8_t __p0,float16x8_t __p1,float16x8_t __p2)36842 __ai __attribute__((target("v8.3a,fullfp16"))) float16x8_t vcmlaq_f16(float16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
36843 float16x8_t __ret;
36844 __ret = (float16x8_t) __builtin_neon_vcmlaq_f16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 40);
36845 return __ret;
36846 }
36847 #else
vcmlaq_f16(float16x8_t __p0,float16x8_t __p1,float16x8_t __p2)36848 __ai __attribute__((target("v8.3a,fullfp16"))) float16x8_t vcmlaq_f16(float16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
36849 float16x8_t __ret;
36850 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
36851 float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
36852 float16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
36853 __ret = (float16x8_t) __builtin_neon_vcmlaq_f16((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 40);
36854 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
36855 return __ret;
36856 }
__noswap_vcmlaq_f16(float16x8_t __p0,float16x8_t __p1,float16x8_t __p2)36857 __ai __attribute__((target("v8.3a,fullfp16"))) float16x8_t __noswap_vcmlaq_f16(float16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
36858 float16x8_t __ret;
36859 __ret = (float16x8_t) __builtin_neon_vcmlaq_f16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 40);
36860 return __ret;
36861 }
36862 #endif
36863
36864 #ifdef __LITTLE_ENDIAN__
vcmla_f16(float16x4_t __p0,float16x4_t __p1,float16x4_t __p2)36865 __ai __attribute__((target("v8.3a,fullfp16"))) float16x4_t vcmla_f16(float16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
36866 float16x4_t __ret;
36867 __ret = (float16x4_t) __builtin_neon_vcmla_f16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 8);
36868 return __ret;
36869 }
36870 #else
vcmla_f16(float16x4_t __p0,float16x4_t __p1,float16x4_t __p2)36871 __ai __attribute__((target("v8.3a,fullfp16"))) float16x4_t vcmla_f16(float16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
36872 float16x4_t __ret;
36873 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
36874 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
36875 float16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
36876 __ret = (float16x4_t) __builtin_neon_vcmla_f16((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 8);
36877 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
36878 return __ret;
36879 }
__noswap_vcmla_f16(float16x4_t __p0,float16x4_t __p1,float16x4_t __p2)36880 __ai __attribute__((target("v8.3a,fullfp16"))) float16x4_t __noswap_vcmla_f16(float16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
36881 float16x4_t __ret;
36882 __ret = (float16x4_t) __builtin_neon_vcmla_f16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 8);
36883 return __ret;
36884 }
36885 #endif
36886
36887 #ifdef __LITTLE_ENDIAN__
36888 #define vcmla_lane_f16(__p0_209, __p1_209, __p2_209, __p3_209) __extension__ ({ \
36889 float16x4_t __ret_209; \
36890 float16x4_t __s0_209 = __p0_209; \
36891 float16x4_t __s1_209 = __p1_209; \
36892 float16x4_t __s2_209 = __p2_209; \
36893 float16x4_t __reint_209 = __s2_209; \
36894 uint32x2_t __reint1_209 = (uint32x2_t) {vget_lane_u32(*(uint32x2_t *) &__reint_209, __p3_209), vget_lane_u32(*(uint32x2_t *) &__reint_209, __p3_209)}; \
36895 __ret_209 = vcmla_f16(__s0_209, __s1_209, *(float16x4_t *) &__reint1_209); \
36896 __ret_209; \
36897 })
36898 #else
36899 #define vcmla_lane_f16(__p0_210, __p1_210, __p2_210, __p3_210) __extension__ ({ \
36900 float16x4_t __ret_210; \
36901 float16x4_t __s0_210 = __p0_210; \
36902 float16x4_t __s1_210 = __p1_210; \
36903 float16x4_t __s2_210 = __p2_210; \
36904 float16x4_t __rev0_210; __rev0_210 = __builtin_shufflevector(__s0_210, __s0_210, 3, 2, 1, 0); \
36905 float16x4_t __rev1_210; __rev1_210 = __builtin_shufflevector(__s1_210, __s1_210, 3, 2, 1, 0); \
36906 float16x4_t __rev2_210; __rev2_210 = __builtin_shufflevector(__s2_210, __s2_210, 3, 2, 1, 0); \
36907 float16x4_t __reint_210 = __rev2_210; \
36908 uint32x2_t __reint1_210 = (uint32x2_t) {__noswap_vget_lane_u32(*(uint32x2_t *) &__reint_210, __p3_210), __noswap_vget_lane_u32(*(uint32x2_t *) &__reint_210, __p3_210)}; \
36909 __ret_210 = __noswap_vcmla_f16(__rev0_210, __rev1_210, *(float16x4_t *) &__reint1_210); \
36910 __ret_210 = __builtin_shufflevector(__ret_210, __ret_210, 3, 2, 1, 0); \
36911 __ret_210; \
36912 })
36913 #endif
36914
36915 #ifdef __LITTLE_ENDIAN__
36916 #define vcmlaq_lane_f16(__p0_211, __p1_211, __p2_211, __p3_211) __extension__ ({ \
36917 float16x8_t __ret_211; \
36918 float16x8_t __s0_211 = __p0_211; \
36919 float16x8_t __s1_211 = __p1_211; \
36920 float16x4_t __s2_211 = __p2_211; \
36921 float16x4_t __reint_211 = __s2_211; \
36922 uint32x4_t __reint1_211 = (uint32x4_t) {vget_lane_u32(*(uint32x2_t *) &__reint_211, __p3_211), vget_lane_u32(*(uint32x2_t *) &__reint_211, __p3_211), vget_lane_u32(*(uint32x2_t *) &__reint_211, __p3_211), vget_lane_u32(*(uint32x2_t *) &__reint_211, __p3_211)}; \
36923 __ret_211 = vcmlaq_f16(__s0_211, __s1_211, *(float16x8_t *) &__reint1_211); \
36924 __ret_211; \
36925 })
36926 #else
36927 #define vcmlaq_lane_f16(__p0_212, __p1_212, __p2_212, __p3_212) __extension__ ({ \
36928 float16x8_t __ret_212; \
36929 float16x8_t __s0_212 = __p0_212; \
36930 float16x8_t __s1_212 = __p1_212; \
36931 float16x4_t __s2_212 = __p2_212; \
36932 float16x8_t __rev0_212; __rev0_212 = __builtin_shufflevector(__s0_212, __s0_212, 7, 6, 5, 4, 3, 2, 1, 0); \
36933 float16x8_t __rev1_212; __rev1_212 = __builtin_shufflevector(__s1_212, __s1_212, 7, 6, 5, 4, 3, 2, 1, 0); \
36934 float16x4_t __rev2_212; __rev2_212 = __builtin_shufflevector(__s2_212, __s2_212, 3, 2, 1, 0); \
36935 float16x4_t __reint_212 = __rev2_212; \
36936 uint32x4_t __reint1_212 = (uint32x4_t) {__noswap_vget_lane_u32(*(uint32x2_t *) &__reint_212, __p3_212), __noswap_vget_lane_u32(*(uint32x2_t *) &__reint_212, __p3_212), __noswap_vget_lane_u32(*(uint32x2_t *) &__reint_212, __p3_212), __noswap_vget_lane_u32(*(uint32x2_t *) &__reint_212, __p3_212)}; \
36937 __ret_212 = __noswap_vcmlaq_f16(__rev0_212, __rev1_212, *(float16x8_t *) &__reint1_212); \
36938 __ret_212 = __builtin_shufflevector(__ret_212, __ret_212, 7, 6, 5, 4, 3, 2, 1, 0); \
36939 __ret_212; \
36940 })
36941 #endif
36942
36943 #ifdef __LITTLE_ENDIAN__
36944 #define vcmla_laneq_f16(__p0_213, __p1_213, __p2_213, __p3_213) __extension__ ({ \
36945 float16x4_t __ret_213; \
36946 float16x4_t __s0_213 = __p0_213; \
36947 float16x4_t __s1_213 = __p1_213; \
36948 float16x8_t __s2_213 = __p2_213; \
36949 float16x8_t __reint_213 = __s2_213; \
36950 uint32x2_t __reint1_213 = (uint32x2_t) {vgetq_lane_u32(*(uint32x4_t *) &__reint_213, __p3_213), vgetq_lane_u32(*(uint32x4_t *) &__reint_213, __p3_213)}; \
36951 __ret_213 = vcmla_f16(__s0_213, __s1_213, *(float16x4_t *) &__reint1_213); \
36952 __ret_213; \
36953 })
36954 #else
36955 #define vcmla_laneq_f16(__p0_214, __p1_214, __p2_214, __p3_214) __extension__ ({ \
36956 float16x4_t __ret_214; \
36957 float16x4_t __s0_214 = __p0_214; \
36958 float16x4_t __s1_214 = __p1_214; \
36959 float16x8_t __s2_214 = __p2_214; \
36960 float16x4_t __rev0_214; __rev0_214 = __builtin_shufflevector(__s0_214, __s0_214, 3, 2, 1, 0); \
36961 float16x4_t __rev1_214; __rev1_214 = __builtin_shufflevector(__s1_214, __s1_214, 3, 2, 1, 0); \
36962 float16x8_t __rev2_214; __rev2_214 = __builtin_shufflevector(__s2_214, __s2_214, 7, 6, 5, 4, 3, 2, 1, 0); \
36963 float16x8_t __reint_214 = __rev2_214; \
36964 uint32x2_t __reint1_214 = (uint32x2_t) {__noswap_vgetq_lane_u32(*(uint32x4_t *) &__reint_214, __p3_214), __noswap_vgetq_lane_u32(*(uint32x4_t *) &__reint_214, __p3_214)}; \
36965 __ret_214 = __noswap_vcmla_f16(__rev0_214, __rev1_214, *(float16x4_t *) &__reint1_214); \
36966 __ret_214 = __builtin_shufflevector(__ret_214, __ret_214, 3, 2, 1, 0); \
36967 __ret_214; \
36968 })
36969 #endif
36970
36971 #ifdef __LITTLE_ENDIAN__
36972 #define vcmlaq_laneq_f16(__p0_215, __p1_215, __p2_215, __p3_215) __extension__ ({ \
36973 float16x8_t __ret_215; \
36974 float16x8_t __s0_215 = __p0_215; \
36975 float16x8_t __s1_215 = __p1_215; \
36976 float16x8_t __s2_215 = __p2_215; \
36977 float16x8_t __reint_215 = __s2_215; \
36978 uint32x4_t __reint1_215 = (uint32x4_t) {vgetq_lane_u32(*(uint32x4_t *) &__reint_215, __p3_215), vgetq_lane_u32(*(uint32x4_t *) &__reint_215, __p3_215), vgetq_lane_u32(*(uint32x4_t *) &__reint_215, __p3_215), vgetq_lane_u32(*(uint32x4_t *) &__reint_215, __p3_215)}; \
36979 __ret_215 = vcmlaq_f16(__s0_215, __s1_215, *(float16x8_t *) &__reint1_215); \
36980 __ret_215; \
36981 })
36982 #else
36983 #define vcmlaq_laneq_f16(__p0_216, __p1_216, __p2_216, __p3_216) __extension__ ({ \
36984 float16x8_t __ret_216; \
36985 float16x8_t __s0_216 = __p0_216; \
36986 float16x8_t __s1_216 = __p1_216; \
36987 float16x8_t __s2_216 = __p2_216; \
36988 float16x8_t __rev0_216; __rev0_216 = __builtin_shufflevector(__s0_216, __s0_216, 7, 6, 5, 4, 3, 2, 1, 0); \
36989 float16x8_t __rev1_216; __rev1_216 = __builtin_shufflevector(__s1_216, __s1_216, 7, 6, 5, 4, 3, 2, 1, 0); \
36990 float16x8_t __rev2_216; __rev2_216 = __builtin_shufflevector(__s2_216, __s2_216, 7, 6, 5, 4, 3, 2, 1, 0); \
36991 float16x8_t __reint_216 = __rev2_216; \
36992 uint32x4_t __reint1_216 = (uint32x4_t) {__noswap_vgetq_lane_u32(*(uint32x4_t *) &__reint_216, __p3_216), __noswap_vgetq_lane_u32(*(uint32x4_t *) &__reint_216, __p3_216), __noswap_vgetq_lane_u32(*(uint32x4_t *) &__reint_216, __p3_216), __noswap_vgetq_lane_u32(*(uint32x4_t *) &__reint_216, __p3_216)}; \
36993 __ret_216 = __noswap_vcmlaq_f16(__rev0_216, __rev1_216, *(float16x8_t *) &__reint1_216); \
36994 __ret_216 = __builtin_shufflevector(__ret_216, __ret_216, 7, 6, 5, 4, 3, 2, 1, 0); \
36995 __ret_216; \
36996 })
36997 #endif
36998
36999 #ifdef __LITTLE_ENDIAN__
vcmlaq_rot180_f16(float16x8_t __p0,float16x8_t __p1,float16x8_t __p2)37000 __ai __attribute__((target("v8.3a,fullfp16"))) float16x8_t vcmlaq_rot180_f16(float16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
37001 float16x8_t __ret;
37002 __ret = (float16x8_t) __builtin_neon_vcmlaq_rot180_f16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 40);
37003 return __ret;
37004 }
37005 #else
vcmlaq_rot180_f16(float16x8_t __p0,float16x8_t __p1,float16x8_t __p2)37006 __ai __attribute__((target("v8.3a,fullfp16"))) float16x8_t vcmlaq_rot180_f16(float16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
37007 float16x8_t __ret;
37008 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
37009 float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
37010 float16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
37011 __ret = (float16x8_t) __builtin_neon_vcmlaq_rot180_f16((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 40);
37012 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
37013 return __ret;
37014 }
__noswap_vcmlaq_rot180_f16(float16x8_t __p0,float16x8_t __p1,float16x8_t __p2)37015 __ai __attribute__((target("v8.3a,fullfp16"))) float16x8_t __noswap_vcmlaq_rot180_f16(float16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
37016 float16x8_t __ret;
37017 __ret = (float16x8_t) __builtin_neon_vcmlaq_rot180_f16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 40);
37018 return __ret;
37019 }
37020 #endif
37021
37022 #ifdef __LITTLE_ENDIAN__
vcmla_rot180_f16(float16x4_t __p0,float16x4_t __p1,float16x4_t __p2)37023 __ai __attribute__((target("v8.3a,fullfp16"))) float16x4_t vcmla_rot180_f16(float16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
37024 float16x4_t __ret;
37025 __ret = (float16x4_t) __builtin_neon_vcmla_rot180_f16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 8);
37026 return __ret;
37027 }
37028 #else
vcmla_rot180_f16(float16x4_t __p0,float16x4_t __p1,float16x4_t __p2)37029 __ai __attribute__((target("v8.3a,fullfp16"))) float16x4_t vcmla_rot180_f16(float16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
37030 float16x4_t __ret;
37031 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
37032 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
37033 float16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
37034 __ret = (float16x4_t) __builtin_neon_vcmla_rot180_f16((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 8);
37035 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
37036 return __ret;
37037 }
__noswap_vcmla_rot180_f16(float16x4_t __p0,float16x4_t __p1,float16x4_t __p2)37038 __ai __attribute__((target("v8.3a,fullfp16"))) float16x4_t __noswap_vcmla_rot180_f16(float16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
37039 float16x4_t __ret;
37040 __ret = (float16x4_t) __builtin_neon_vcmla_rot180_f16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 8);
37041 return __ret;
37042 }
37043 #endif
37044
37045 #ifdef __LITTLE_ENDIAN__
37046 #define vcmla_rot180_lane_f16(__p0_217, __p1_217, __p2_217, __p3_217) __extension__ ({ \
37047 float16x4_t __ret_217; \
37048 float16x4_t __s0_217 = __p0_217; \
37049 float16x4_t __s1_217 = __p1_217; \
37050 float16x4_t __s2_217 = __p2_217; \
37051 float16x4_t __reint_217 = __s2_217; \
37052 uint32x2_t __reint1_217 = (uint32x2_t) {vget_lane_u32(*(uint32x2_t *) &__reint_217, __p3_217), vget_lane_u32(*(uint32x2_t *) &__reint_217, __p3_217)}; \
37053 __ret_217 = vcmla_rot180_f16(__s0_217, __s1_217, *(float16x4_t *) &__reint1_217); \
37054 __ret_217; \
37055 })
37056 #else
37057 #define vcmla_rot180_lane_f16(__p0_218, __p1_218, __p2_218, __p3_218) __extension__ ({ \
37058 float16x4_t __ret_218; \
37059 float16x4_t __s0_218 = __p0_218; \
37060 float16x4_t __s1_218 = __p1_218; \
37061 float16x4_t __s2_218 = __p2_218; \
37062 float16x4_t __rev0_218; __rev0_218 = __builtin_shufflevector(__s0_218, __s0_218, 3, 2, 1, 0); \
37063 float16x4_t __rev1_218; __rev1_218 = __builtin_shufflevector(__s1_218, __s1_218, 3, 2, 1, 0); \
37064 float16x4_t __rev2_218; __rev2_218 = __builtin_shufflevector(__s2_218, __s2_218, 3, 2, 1, 0); \
37065 float16x4_t __reint_218 = __rev2_218; \
37066 uint32x2_t __reint1_218 = (uint32x2_t) {__noswap_vget_lane_u32(*(uint32x2_t *) &__reint_218, __p3_218), __noswap_vget_lane_u32(*(uint32x2_t *) &__reint_218, __p3_218)}; \
37067 __ret_218 = __noswap_vcmla_rot180_f16(__rev0_218, __rev1_218, *(float16x4_t *) &__reint1_218); \
37068 __ret_218 = __builtin_shufflevector(__ret_218, __ret_218, 3, 2, 1, 0); \
37069 __ret_218; \
37070 })
37071 #endif
37072
37073 #ifdef __LITTLE_ENDIAN__
37074 #define vcmlaq_rot180_lane_f16(__p0_219, __p1_219, __p2_219, __p3_219) __extension__ ({ \
37075 float16x8_t __ret_219; \
37076 float16x8_t __s0_219 = __p0_219; \
37077 float16x8_t __s1_219 = __p1_219; \
37078 float16x4_t __s2_219 = __p2_219; \
37079 float16x4_t __reint_219 = __s2_219; \
37080 uint32x4_t __reint1_219 = (uint32x4_t) {vget_lane_u32(*(uint32x2_t *) &__reint_219, __p3_219), vget_lane_u32(*(uint32x2_t *) &__reint_219, __p3_219), vget_lane_u32(*(uint32x2_t *) &__reint_219, __p3_219), vget_lane_u32(*(uint32x2_t *) &__reint_219, __p3_219)}; \
37081 __ret_219 = vcmlaq_rot180_f16(__s0_219, __s1_219, *(float16x8_t *) &__reint1_219); \
37082 __ret_219; \
37083 })
37084 #else
37085 #define vcmlaq_rot180_lane_f16(__p0_220, __p1_220, __p2_220, __p3_220) __extension__ ({ \
37086 float16x8_t __ret_220; \
37087 float16x8_t __s0_220 = __p0_220; \
37088 float16x8_t __s1_220 = __p1_220; \
37089 float16x4_t __s2_220 = __p2_220; \
37090 float16x8_t __rev0_220; __rev0_220 = __builtin_shufflevector(__s0_220, __s0_220, 7, 6, 5, 4, 3, 2, 1, 0); \
37091 float16x8_t __rev1_220; __rev1_220 = __builtin_shufflevector(__s1_220, __s1_220, 7, 6, 5, 4, 3, 2, 1, 0); \
37092 float16x4_t __rev2_220; __rev2_220 = __builtin_shufflevector(__s2_220, __s2_220, 3, 2, 1, 0); \
37093 float16x4_t __reint_220 = __rev2_220; \
37094 uint32x4_t __reint1_220 = (uint32x4_t) {__noswap_vget_lane_u32(*(uint32x2_t *) &__reint_220, __p3_220), __noswap_vget_lane_u32(*(uint32x2_t *) &__reint_220, __p3_220), __noswap_vget_lane_u32(*(uint32x2_t *) &__reint_220, __p3_220), __noswap_vget_lane_u32(*(uint32x2_t *) &__reint_220, __p3_220)}; \
37095 __ret_220 = __noswap_vcmlaq_rot180_f16(__rev0_220, __rev1_220, *(float16x8_t *) &__reint1_220); \
37096 __ret_220 = __builtin_shufflevector(__ret_220, __ret_220, 7, 6, 5, 4, 3, 2, 1, 0); \
37097 __ret_220; \
37098 })
37099 #endif
37100
37101 #ifdef __LITTLE_ENDIAN__
37102 #define vcmla_rot180_laneq_f16(__p0_221, __p1_221, __p2_221, __p3_221) __extension__ ({ \
37103 float16x4_t __ret_221; \
37104 float16x4_t __s0_221 = __p0_221; \
37105 float16x4_t __s1_221 = __p1_221; \
37106 float16x8_t __s2_221 = __p2_221; \
37107 float16x8_t __reint_221 = __s2_221; \
37108 uint32x2_t __reint1_221 = (uint32x2_t) {vgetq_lane_u32(*(uint32x4_t *) &__reint_221, __p3_221), vgetq_lane_u32(*(uint32x4_t *) &__reint_221, __p3_221)}; \
37109 __ret_221 = vcmla_rot180_f16(__s0_221, __s1_221, *(float16x4_t *) &__reint1_221); \
37110 __ret_221; \
37111 })
37112 #else
37113 #define vcmla_rot180_laneq_f16(__p0_222, __p1_222, __p2_222, __p3_222) __extension__ ({ \
37114 float16x4_t __ret_222; \
37115 float16x4_t __s0_222 = __p0_222; \
37116 float16x4_t __s1_222 = __p1_222; \
37117 float16x8_t __s2_222 = __p2_222; \
37118 float16x4_t __rev0_222; __rev0_222 = __builtin_shufflevector(__s0_222, __s0_222, 3, 2, 1, 0); \
37119 float16x4_t __rev1_222; __rev1_222 = __builtin_shufflevector(__s1_222, __s1_222, 3, 2, 1, 0); \
37120 float16x8_t __rev2_222; __rev2_222 = __builtin_shufflevector(__s2_222, __s2_222, 7, 6, 5, 4, 3, 2, 1, 0); \
37121 float16x8_t __reint_222 = __rev2_222; \
37122 uint32x2_t __reint1_222 = (uint32x2_t) {__noswap_vgetq_lane_u32(*(uint32x4_t *) &__reint_222, __p3_222), __noswap_vgetq_lane_u32(*(uint32x4_t *) &__reint_222, __p3_222)}; \
37123 __ret_222 = __noswap_vcmla_rot180_f16(__rev0_222, __rev1_222, *(float16x4_t *) &__reint1_222); \
37124 __ret_222 = __builtin_shufflevector(__ret_222, __ret_222, 3, 2, 1, 0); \
37125 __ret_222; \
37126 })
37127 #endif
37128
37129 #ifdef __LITTLE_ENDIAN__
37130 #define vcmlaq_rot180_laneq_f16(__p0_223, __p1_223, __p2_223, __p3_223) __extension__ ({ \
37131 float16x8_t __ret_223; \
37132 float16x8_t __s0_223 = __p0_223; \
37133 float16x8_t __s1_223 = __p1_223; \
37134 float16x8_t __s2_223 = __p2_223; \
37135 float16x8_t __reint_223 = __s2_223; \
37136 uint32x4_t __reint1_223 = (uint32x4_t) {vgetq_lane_u32(*(uint32x4_t *) &__reint_223, __p3_223), vgetq_lane_u32(*(uint32x4_t *) &__reint_223, __p3_223), vgetq_lane_u32(*(uint32x4_t *) &__reint_223, __p3_223), vgetq_lane_u32(*(uint32x4_t *) &__reint_223, __p3_223)}; \
37137 __ret_223 = vcmlaq_rot180_f16(__s0_223, __s1_223, *(float16x8_t *) &__reint1_223); \
37138 __ret_223; \
37139 })
37140 #else
37141 #define vcmlaq_rot180_laneq_f16(__p0_224, __p1_224, __p2_224, __p3_224) __extension__ ({ \
37142 float16x8_t __ret_224; \
37143 float16x8_t __s0_224 = __p0_224; \
37144 float16x8_t __s1_224 = __p1_224; \
37145 float16x8_t __s2_224 = __p2_224; \
37146 float16x8_t __rev0_224; __rev0_224 = __builtin_shufflevector(__s0_224, __s0_224, 7, 6, 5, 4, 3, 2, 1, 0); \
37147 float16x8_t __rev1_224; __rev1_224 = __builtin_shufflevector(__s1_224, __s1_224, 7, 6, 5, 4, 3, 2, 1, 0); \
37148 float16x8_t __rev2_224; __rev2_224 = __builtin_shufflevector(__s2_224, __s2_224, 7, 6, 5, 4, 3, 2, 1, 0); \
37149 float16x8_t __reint_224 = __rev2_224; \
37150 uint32x4_t __reint1_224 = (uint32x4_t) {__noswap_vgetq_lane_u32(*(uint32x4_t *) &__reint_224, __p3_224), __noswap_vgetq_lane_u32(*(uint32x4_t *) &__reint_224, __p3_224), __noswap_vgetq_lane_u32(*(uint32x4_t *) &__reint_224, __p3_224), __noswap_vgetq_lane_u32(*(uint32x4_t *) &__reint_224, __p3_224)}; \
37151 __ret_224 = __noswap_vcmlaq_rot180_f16(__rev0_224, __rev1_224, *(float16x8_t *) &__reint1_224); \
37152 __ret_224 = __builtin_shufflevector(__ret_224, __ret_224, 7, 6, 5, 4, 3, 2, 1, 0); \
37153 __ret_224; \
37154 })
37155 #endif
37156
37157 #ifdef __LITTLE_ENDIAN__
vcmlaq_rot270_f16(float16x8_t __p0,float16x8_t __p1,float16x8_t __p2)37158 __ai __attribute__((target("v8.3a,fullfp16"))) float16x8_t vcmlaq_rot270_f16(float16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
37159 float16x8_t __ret;
37160 __ret = (float16x8_t) __builtin_neon_vcmlaq_rot270_f16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 40);
37161 return __ret;
37162 }
37163 #else
vcmlaq_rot270_f16(float16x8_t __p0,float16x8_t __p1,float16x8_t __p2)37164 __ai __attribute__((target("v8.3a,fullfp16"))) float16x8_t vcmlaq_rot270_f16(float16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
37165 float16x8_t __ret;
37166 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
37167 float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
37168 float16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
37169 __ret = (float16x8_t) __builtin_neon_vcmlaq_rot270_f16((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 40);
37170 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
37171 return __ret;
37172 }
__noswap_vcmlaq_rot270_f16(float16x8_t __p0,float16x8_t __p1,float16x8_t __p2)37173 __ai __attribute__((target("v8.3a,fullfp16"))) float16x8_t __noswap_vcmlaq_rot270_f16(float16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
37174 float16x8_t __ret;
37175 __ret = (float16x8_t) __builtin_neon_vcmlaq_rot270_f16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 40);
37176 return __ret;
37177 }
37178 #endif
37179
37180 #ifdef __LITTLE_ENDIAN__
vcmla_rot270_f16(float16x4_t __p0,float16x4_t __p1,float16x4_t __p2)37181 __ai __attribute__((target("v8.3a,fullfp16"))) float16x4_t vcmla_rot270_f16(float16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
37182 float16x4_t __ret;
37183 __ret = (float16x4_t) __builtin_neon_vcmla_rot270_f16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 8);
37184 return __ret;
37185 }
37186 #else
vcmla_rot270_f16(float16x4_t __p0,float16x4_t __p1,float16x4_t __p2)37187 __ai __attribute__((target("v8.3a,fullfp16"))) float16x4_t vcmla_rot270_f16(float16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
37188 float16x4_t __ret;
37189 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
37190 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
37191 float16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
37192 __ret = (float16x4_t) __builtin_neon_vcmla_rot270_f16((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 8);
37193 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
37194 return __ret;
37195 }
__noswap_vcmla_rot270_f16(float16x4_t __p0,float16x4_t __p1,float16x4_t __p2)37196 __ai __attribute__((target("v8.3a,fullfp16"))) float16x4_t __noswap_vcmla_rot270_f16(float16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
37197 float16x4_t __ret;
37198 __ret = (float16x4_t) __builtin_neon_vcmla_rot270_f16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 8);
37199 return __ret;
37200 }
37201 #endif
37202
37203 #ifdef __LITTLE_ENDIAN__
37204 #define vcmla_rot270_lane_f16(__p0_225, __p1_225, __p2_225, __p3_225) __extension__ ({ \
37205 float16x4_t __ret_225; \
37206 float16x4_t __s0_225 = __p0_225; \
37207 float16x4_t __s1_225 = __p1_225; \
37208 float16x4_t __s2_225 = __p2_225; \
37209 float16x4_t __reint_225 = __s2_225; \
37210 uint32x2_t __reint1_225 = (uint32x2_t) {vget_lane_u32(*(uint32x2_t *) &__reint_225, __p3_225), vget_lane_u32(*(uint32x2_t *) &__reint_225, __p3_225)}; \
37211 __ret_225 = vcmla_rot270_f16(__s0_225, __s1_225, *(float16x4_t *) &__reint1_225); \
37212 __ret_225; \
37213 })
37214 #else
37215 #define vcmla_rot270_lane_f16(__p0_226, __p1_226, __p2_226, __p3_226) __extension__ ({ \
37216 float16x4_t __ret_226; \
37217 float16x4_t __s0_226 = __p0_226; \
37218 float16x4_t __s1_226 = __p1_226; \
37219 float16x4_t __s2_226 = __p2_226; \
37220 float16x4_t __rev0_226; __rev0_226 = __builtin_shufflevector(__s0_226, __s0_226, 3, 2, 1, 0); \
37221 float16x4_t __rev1_226; __rev1_226 = __builtin_shufflevector(__s1_226, __s1_226, 3, 2, 1, 0); \
37222 float16x4_t __rev2_226; __rev2_226 = __builtin_shufflevector(__s2_226, __s2_226, 3, 2, 1, 0); \
37223 float16x4_t __reint_226 = __rev2_226; \
37224 uint32x2_t __reint1_226 = (uint32x2_t) {__noswap_vget_lane_u32(*(uint32x2_t *) &__reint_226, __p3_226), __noswap_vget_lane_u32(*(uint32x2_t *) &__reint_226, __p3_226)}; \
37225 __ret_226 = __noswap_vcmla_rot270_f16(__rev0_226, __rev1_226, *(float16x4_t *) &__reint1_226); \
37226 __ret_226 = __builtin_shufflevector(__ret_226, __ret_226, 3, 2, 1, 0); \
37227 __ret_226; \
37228 })
37229 #endif
37230
37231 #ifdef __LITTLE_ENDIAN__
37232 #define vcmlaq_rot270_lane_f16(__p0_227, __p1_227, __p2_227, __p3_227) __extension__ ({ \
37233 float16x8_t __ret_227; \
37234 float16x8_t __s0_227 = __p0_227; \
37235 float16x8_t __s1_227 = __p1_227; \
37236 float16x4_t __s2_227 = __p2_227; \
37237 float16x4_t __reint_227 = __s2_227; \
37238 uint32x4_t __reint1_227 = (uint32x4_t) {vget_lane_u32(*(uint32x2_t *) &__reint_227, __p3_227), vget_lane_u32(*(uint32x2_t *) &__reint_227, __p3_227), vget_lane_u32(*(uint32x2_t *) &__reint_227, __p3_227), vget_lane_u32(*(uint32x2_t *) &__reint_227, __p3_227)}; \
37239 __ret_227 = vcmlaq_rot270_f16(__s0_227, __s1_227, *(float16x8_t *) &__reint1_227); \
37240 __ret_227; \
37241 })
37242 #else
37243 #define vcmlaq_rot270_lane_f16(__p0_228, __p1_228, __p2_228, __p3_228) __extension__ ({ \
37244 float16x8_t __ret_228; \
37245 float16x8_t __s0_228 = __p0_228; \
37246 float16x8_t __s1_228 = __p1_228; \
37247 float16x4_t __s2_228 = __p2_228; \
37248 float16x8_t __rev0_228; __rev0_228 = __builtin_shufflevector(__s0_228, __s0_228, 7, 6, 5, 4, 3, 2, 1, 0); \
37249 float16x8_t __rev1_228; __rev1_228 = __builtin_shufflevector(__s1_228, __s1_228, 7, 6, 5, 4, 3, 2, 1, 0); \
37250 float16x4_t __rev2_228; __rev2_228 = __builtin_shufflevector(__s2_228, __s2_228, 3, 2, 1, 0); \
37251 float16x4_t __reint_228 = __rev2_228; \
37252 uint32x4_t __reint1_228 = (uint32x4_t) {__noswap_vget_lane_u32(*(uint32x2_t *) &__reint_228, __p3_228), __noswap_vget_lane_u32(*(uint32x2_t *) &__reint_228, __p3_228), __noswap_vget_lane_u32(*(uint32x2_t *) &__reint_228, __p3_228), __noswap_vget_lane_u32(*(uint32x2_t *) &__reint_228, __p3_228)}; \
37253 __ret_228 = __noswap_vcmlaq_rot270_f16(__rev0_228, __rev1_228, *(float16x8_t *) &__reint1_228); \
37254 __ret_228 = __builtin_shufflevector(__ret_228, __ret_228, 7, 6, 5, 4, 3, 2, 1, 0); \
37255 __ret_228; \
37256 })
37257 #endif
37258
37259 #ifdef __LITTLE_ENDIAN__
37260 #define vcmla_rot270_laneq_f16(__p0_229, __p1_229, __p2_229, __p3_229) __extension__ ({ \
37261 float16x4_t __ret_229; \
37262 float16x4_t __s0_229 = __p0_229; \
37263 float16x4_t __s1_229 = __p1_229; \
37264 float16x8_t __s2_229 = __p2_229; \
37265 float16x8_t __reint_229 = __s2_229; \
37266 uint32x2_t __reint1_229 = (uint32x2_t) {vgetq_lane_u32(*(uint32x4_t *) &__reint_229, __p3_229), vgetq_lane_u32(*(uint32x4_t *) &__reint_229, __p3_229)}; \
37267 __ret_229 = vcmla_rot270_f16(__s0_229, __s1_229, *(float16x4_t *) &__reint1_229); \
37268 __ret_229; \
37269 })
37270 #else
37271 #define vcmla_rot270_laneq_f16(__p0_230, __p1_230, __p2_230, __p3_230) __extension__ ({ \
37272 float16x4_t __ret_230; \
37273 float16x4_t __s0_230 = __p0_230; \
37274 float16x4_t __s1_230 = __p1_230; \
37275 float16x8_t __s2_230 = __p2_230; \
37276 float16x4_t __rev0_230; __rev0_230 = __builtin_shufflevector(__s0_230, __s0_230, 3, 2, 1, 0); \
37277 float16x4_t __rev1_230; __rev1_230 = __builtin_shufflevector(__s1_230, __s1_230, 3, 2, 1, 0); \
37278 float16x8_t __rev2_230; __rev2_230 = __builtin_shufflevector(__s2_230, __s2_230, 7, 6, 5, 4, 3, 2, 1, 0); \
37279 float16x8_t __reint_230 = __rev2_230; \
37280 uint32x2_t __reint1_230 = (uint32x2_t) {__noswap_vgetq_lane_u32(*(uint32x4_t *) &__reint_230, __p3_230), __noswap_vgetq_lane_u32(*(uint32x4_t *) &__reint_230, __p3_230)}; \
37281 __ret_230 = __noswap_vcmla_rot270_f16(__rev0_230, __rev1_230, *(float16x4_t *) &__reint1_230); \
37282 __ret_230 = __builtin_shufflevector(__ret_230, __ret_230, 3, 2, 1, 0); \
37283 __ret_230; \
37284 })
37285 #endif
37286
37287 #ifdef __LITTLE_ENDIAN__
37288 #define vcmlaq_rot270_laneq_f16(__p0_231, __p1_231, __p2_231, __p3_231) __extension__ ({ \
37289 float16x8_t __ret_231; \
37290 float16x8_t __s0_231 = __p0_231; \
37291 float16x8_t __s1_231 = __p1_231; \
37292 float16x8_t __s2_231 = __p2_231; \
37293 float16x8_t __reint_231 = __s2_231; \
37294 uint32x4_t __reint1_231 = (uint32x4_t) {vgetq_lane_u32(*(uint32x4_t *) &__reint_231, __p3_231), vgetq_lane_u32(*(uint32x4_t *) &__reint_231, __p3_231), vgetq_lane_u32(*(uint32x4_t *) &__reint_231, __p3_231), vgetq_lane_u32(*(uint32x4_t *) &__reint_231, __p3_231)}; \
37295 __ret_231 = vcmlaq_rot270_f16(__s0_231, __s1_231, *(float16x8_t *) &__reint1_231); \
37296 __ret_231; \
37297 })
37298 #else
37299 #define vcmlaq_rot270_laneq_f16(__p0_232, __p1_232, __p2_232, __p3_232) __extension__ ({ \
37300 float16x8_t __ret_232; \
37301 float16x8_t __s0_232 = __p0_232; \
37302 float16x8_t __s1_232 = __p1_232; \
37303 float16x8_t __s2_232 = __p2_232; \
37304 float16x8_t __rev0_232; __rev0_232 = __builtin_shufflevector(__s0_232, __s0_232, 7, 6, 5, 4, 3, 2, 1, 0); \
37305 float16x8_t __rev1_232; __rev1_232 = __builtin_shufflevector(__s1_232, __s1_232, 7, 6, 5, 4, 3, 2, 1, 0); \
37306 float16x8_t __rev2_232; __rev2_232 = __builtin_shufflevector(__s2_232, __s2_232, 7, 6, 5, 4, 3, 2, 1, 0); \
37307 float16x8_t __reint_232 = __rev2_232; \
37308 uint32x4_t __reint1_232 = (uint32x4_t) {__noswap_vgetq_lane_u32(*(uint32x4_t *) &__reint_232, __p3_232), __noswap_vgetq_lane_u32(*(uint32x4_t *) &__reint_232, __p3_232), __noswap_vgetq_lane_u32(*(uint32x4_t *) &__reint_232, __p3_232), __noswap_vgetq_lane_u32(*(uint32x4_t *) &__reint_232, __p3_232)}; \
37309 __ret_232 = __noswap_vcmlaq_rot270_f16(__rev0_232, __rev1_232, *(float16x8_t *) &__reint1_232); \
37310 __ret_232 = __builtin_shufflevector(__ret_232, __ret_232, 7, 6, 5, 4, 3, 2, 1, 0); \
37311 __ret_232; \
37312 })
37313 #endif
37314
37315 #ifdef __LITTLE_ENDIAN__
vcmlaq_rot90_f16(float16x8_t __p0,float16x8_t __p1,float16x8_t __p2)37316 __ai __attribute__((target("v8.3a,fullfp16"))) float16x8_t vcmlaq_rot90_f16(float16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
37317 float16x8_t __ret;
37318 __ret = (float16x8_t) __builtin_neon_vcmlaq_rot90_f16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 40);
37319 return __ret;
37320 }
37321 #else
vcmlaq_rot90_f16(float16x8_t __p0,float16x8_t __p1,float16x8_t __p2)37322 __ai __attribute__((target("v8.3a,fullfp16"))) float16x8_t vcmlaq_rot90_f16(float16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
37323 float16x8_t __ret;
37324 float16x8_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 7, 6, 5, 4, 3, 2, 1, 0);
37325 float16x8_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 7, 6, 5, 4, 3, 2, 1, 0);
37326 float16x8_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 7, 6, 5, 4, 3, 2, 1, 0);
37327 __ret = (float16x8_t) __builtin_neon_vcmlaq_rot90_f16((int8x16_t)__rev0, (int8x16_t)__rev1, (int8x16_t)__rev2, 40);
37328 __ret = __builtin_shufflevector(__ret, __ret, 7, 6, 5, 4, 3, 2, 1, 0);
37329 return __ret;
37330 }
__noswap_vcmlaq_rot90_f16(float16x8_t __p0,float16x8_t __p1,float16x8_t __p2)37331 __ai __attribute__((target("v8.3a,fullfp16"))) float16x8_t __noswap_vcmlaq_rot90_f16(float16x8_t __p0, float16x8_t __p1, float16x8_t __p2) {
37332 float16x8_t __ret;
37333 __ret = (float16x8_t) __builtin_neon_vcmlaq_rot90_f16((int8x16_t)__p0, (int8x16_t)__p1, (int8x16_t)__p2, 40);
37334 return __ret;
37335 }
37336 #endif
37337
37338 #ifdef __LITTLE_ENDIAN__
vcmla_rot90_f16(float16x4_t __p0,float16x4_t __p1,float16x4_t __p2)37339 __ai __attribute__((target("v8.3a,fullfp16"))) float16x4_t vcmla_rot90_f16(float16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
37340 float16x4_t __ret;
37341 __ret = (float16x4_t) __builtin_neon_vcmla_rot90_f16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 8);
37342 return __ret;
37343 }
37344 #else
vcmla_rot90_f16(float16x4_t __p0,float16x4_t __p1,float16x4_t __p2)37345 __ai __attribute__((target("v8.3a,fullfp16"))) float16x4_t vcmla_rot90_f16(float16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
37346 float16x4_t __ret;
37347 float16x4_t __rev0; __rev0 = __builtin_shufflevector(__p0, __p0, 3, 2, 1, 0);
37348 float16x4_t __rev1; __rev1 = __builtin_shufflevector(__p1, __p1, 3, 2, 1, 0);
37349 float16x4_t __rev2; __rev2 = __builtin_shufflevector(__p2, __p2, 3, 2, 1, 0);
37350 __ret = (float16x4_t) __builtin_neon_vcmla_rot90_f16((int8x8_t)__rev0, (int8x8_t)__rev1, (int8x8_t)__rev2, 8);
37351 __ret = __builtin_shufflevector(__ret, __ret, 3, 2, 1, 0);
37352 return __ret;
37353 }
__noswap_vcmla_rot90_f16(float16x4_t __p0,float16x4_t __p1,float16x4_t __p2)37354 __ai __attribute__((target("v8.3a,fullfp16"))) float16x4_t __noswap_vcmla_rot90_f16(float16x4_t __p0, float16x4_t __p1, float16x4_t __p2) {
37355 float16x4_t __ret;
37356 __ret = (float16x4_t) __builtin_neon_vcmla_rot90_f16((int8x8_t)__p0, (int8x8_t)__p1, (int8x8_t)__p2, 8);
37357 return __ret;
37358 }
37359 #endif
37360
37361 #ifdef __LITTLE_ENDIAN__
37362 #define vcmla_rot90_lane_f16(__p0_233, __p1_233, __p2_233, __p3_233) __extension__ ({ \
37363 float16x4_t __ret_233; \
37364 float16x4_t __s0_233 = __p0_233; \
37365 float16x4_t __s1_233 = __p1_233; \
37366 float16x4_t __s2_233 = __p2_233; \
37367 float16x4_t __reint_233 = __s2_233; \
37368 uint32x2_t __reint1_233 = (uint32x2_t) {vget_lane_u32(*(uint32x2_t *) &__reint_233, __p3_233), vget_lane_u32(*(uint32x2_t *) &__reint_233, __p3_233)}; \
37369 __ret_233 = vcmla_rot90_f16(__s0_233, __s1_233, *(float16x4_t *) &__reint1_233); \
37370 __ret_233; \
37371 })
37372 #else
37373 #define vcmla_rot90_lane_f16(__p0_234, __p1_234, __p2_234, __p3_234) __extension__ ({ \
37374 float16x4_t __ret_234; \
37375 float16x4_t __s0_234 = __p0_234; \
37376 float16x4_t __s1_234 = __p1_234; \
37377 float16x4_t __s2_234 = __p2_234; \
37378 float16x4_t __rev0_234; __rev0_234 = __builtin_shufflevector(__s0_234, __s0_234, 3, 2, 1, 0); \
37379 float16x4_t __rev1_234; __rev1_234 = __builtin_shufflevector(__s1_234, __s1_234, 3, 2, 1, 0); \
37380 float16x4_t __rev2_234; __rev2_234 = __builtin_shufflevector(__s2_234, __s2_234, 3, 2, 1, 0); \
37381 float16x4_t __reint_234 = __rev2_234; \
37382 uint32x2_t __reint1_234 = (uint32x2_t) {__noswap_vget_lane_u32(*(uint32x2_t *) &__reint_234, __p3_234), __noswap_vget_lane_u32(*(uint32x2_t *) &__reint_234, __p3_234)}; \
37383 __ret_234 = __noswap_vcmla_rot90_f16(__rev0_234, __rev1_234, *(float16x4_t *) &__reint1_234); \
37384 __ret_234 = __builtin_shufflevector(__ret_234, __ret_234, 3, 2, 1, 0); \
37385 __ret_234; \
37386 })
37387 #endif
37388
37389 #ifdef __LITTLE_ENDIAN__
37390 #define vcmlaq_rot90_lane_f16(__p0_235, __p1_235, __p2_235, __p3_235) __extension__ ({ \
37391 float16x8_t __ret_235; \
37392 float16x8_t __s0_235 = __p0_235; \
37393 float16x8_t __s1_235 = __p1_235; \
37394 float16x4_t __s2_235 = __p2_235; \
37395 float16x4_t __reint_235 = __s2_235; \
37396 uint32x4_t __reint1_235 = (uint32x4_t) {vget_lane_u32(*(uint32x2_t *) &__reint_235, __p3_235), vget_lane_u32(*(uint32x2_t *) &__reint_235, __p3_235), vget_lane_u32(*(uint32x2_t *) &__reint_235, __p3_235), vget_lane_u32(*(uint32x2_t *) &__reint_235, __p3_235)}; \
37397 __ret_235 = vcmlaq_rot90_f16(__s0_235, __s1_235, *(float16x8_t *) &__reint1_235); \
37398 __ret_235; \
37399 })
37400 #else
37401 #define vcmlaq_rot90_lane_f16(__p0_236, __p1_236, __p2_236, __p3_236) __extension__ ({ \
37402 float16x8_t __ret_236; \
37403 float16x8_t __s0_236 = __p0_236; \
37404 float16x8_t __s1_236 = __p1_236; \
37405 float16x4_t __s2_236 = __p2_236; \
37406 float16x8_t __rev0_236; __rev0_236 = __builtin_shufflevector(__s0_236, __s0_236, 7, 6, 5, 4, 3, 2, 1, 0); \
37407 float16x8_t __rev1_236; __rev1_236 = __builtin_shufflevector(__s1_236, __s1_236, 7, 6, 5, 4, 3, 2, 1, 0); \
37408 float16x4_t __rev2_236; __rev2_236 = __builtin_shufflevector(__s2_236, __s2_236, 3, 2, 1, 0); \
37409 float16x4_t __reint_236 = __rev2_236; \
37410 uint32x4_t __reint1_236 = (uint32x4_t) {__noswap_vget_lane_u32(*(uint32x2_t *) &__reint_236, __p3_236), __noswap_vget_lane_u32(*(uint32x2_t *) &__reint_236, __p3_236), __noswap_vget_lane_u32(*(uint32x2_t *) &__reint_236, __p3_236), __noswap_vget_lane_u32(*(uint32x2_t *) &__reint_236, __p3_236)}; \
37411 __ret_236 = __noswap_vcmlaq_rot90_f16(__rev0_236, __rev1_236, *(float16x8_t *) &__reint1_236); \
37412 __ret_236 = __builtin_shufflevector(__ret_236, __ret_236, 7, 6, 5, 4, 3, 2, 1, 0); \
37413 __ret_236; \
37414 })
37415 #endif
37416
37417 #ifdef __LITTLE_ENDIAN__
37418 #define vcmla_rot90_laneq_f16(__p0_237, __p1_237, __p2_237, __p3_237) __extension__ ({ \
37419 float16x4_t __ret_237; \
37420 float16x4_t __s0_237 = __p0_237; \
37421 float16x4_t __s1_237 = __p1_237; \
37422 float16x8_t __s2_237 = __p2_237; \
37423 float16x8_t __reint_237 = __s2_237; \
37424 uint32x2_t __reint1_237 = (uint32x2_t) {vgetq_lane_u32(*(uint32x4_t *) &__reint_237, __p3_237), vgetq_lane_u32(*(uint32x4_t *) &__reint_237, __p3_237)}; \
37425 __ret_237 = vcmla_rot90_f16(__s0_237, __s1_237, *(float16x4_t *) &__reint1_237); \
37426 __ret_237; \
37427 })
37428 #else
37429 #define vcmla_rot90_laneq_f16(__p0_238, __p1_238, __p2_238, __p3_238) __extension__ ({ \
37430 float16x4_t __ret_238; \
37431 float16x4_t __s0_238 = __p0_238; \
37432 float16x4_t __s1_238 = __p1_238; \
37433 float16x8_t __s2_238 = __p2_238; \
37434 float16x4_t __rev0_238; __rev0_238 = __builtin_shufflevector(__s0_238, __s0_238, 3, 2, 1, 0); \
37435 float16x4_t __rev1_238; __rev1_238 = __builtin_shufflevector(__s1_238, __s1_238, 3, 2, 1, 0); \
37436 float16x8_t __rev2_238; __rev2_238 = __builtin_shufflevector(__s2_238, __s2_238, 7, 6, 5, 4, 3, 2, 1, 0); \
37437 float16x8_t __reint_238 = __rev2_238; \
37438 uint32x2_t __reint1_238 = (uint32x2_t) {__noswap_vgetq_lane_u32(*(uint32x4_t *) &__reint_238, __p3_238), __noswap_vgetq_lane_u32(*(uint32x4_t *) &__reint_238, __p3_238)}; \
37439 __ret_238 = __noswap_vcmla_rot90_f16(__rev0_238, __rev1_238, *(float16x4_t *) &__reint1_238); \
37440 __ret_238 = __builtin_shufflevector(__ret_238, __ret_238, 3, 2, 1, 0); \
37441 __ret_238; \
37442 })
37443 #endif
37444
37445 #ifdef __LITTLE_ENDIAN__
37446 #define vcmlaq_rot90_laneq_f16(__p0_239, __p1_239, __p2_239, __p3_239) __extension__ ({ \
37447 float16x8_t __ret_239; \
37448 float16x8_t __s0_239 = __p0_239; \
37449 float16x8_t __s1_239 = __p1_239; \
37450 float16x8_t __s2_239 = __p2_239; \
37451 float16x8_t __reint_239 = __s2_239; \
37452 uint32x4_t __reint1_239 = (uint32x4_t) {vgetq_lane_u32(*(uint32x4_t *) &__reint_239, __p3_239), vgetq_lane_u32(*(uint32x4_t *) &__reint_239, __p3_239), vgetq_lane_u32(*(uint32x4_t *) &__reint_239, __p3_239), vgetq_lane_u32(*(uint32x4_t *) &__reint_239, __p3_239)}; \
37453 __ret_239 = vcmlaq_rot90_f16(__s0_239, __s1_239, *(float16x8_t *) &__reint1_239); \
37454 __ret_239; \
37455 })
37456 #else
37457 #define vcmlaq_rot90_laneq_f16(__p0_240, __p1_240, __p2_240, __p3_240) __extension__ ({ \
37458 float16x8_t __ret_240; \
37459 float16x8_t __s0_240 = __p0_240; \
37460 float16x8_t __s1_240 = __p1_240; \
37461 float16x8_t __s2_240 = __p2_240; \
37462 float16x8_t __rev0_240; __rev0_240 = __builtin_shufflevector(__s0_240, __s0_240, 7, 6, 5, 4, 3, 2, 1, 0); \
37463 float16x8_t __rev1_240; __rev1_240 = __builtin_shufflevector(__s1_240, __s1_240, 7, 6, 5, 4, 3, 2, 1, 0); \
37464 float16x8_t __rev2_240; __rev2_240 = __builtin_shufflevector(__s2_240, __s2_240, 7, 6, 5, 4, 3, 2, 1, 0); \
37465 float16x8_t __reint_240 = __rev2_240; \
37466 uint32x4_t __reint1_240 = (uint32x4_t) {__noswap_vgetq_lane_u32(*(uint32x4_t *) &__reint_240, __p3_240), __noswap_vgetq_lane_u32(*(uint32x4_t *) &__reint_240, __p3_240), __noswap_vgetq_lane_u32(*(uint32x4_t *) &__reint_240, __p3_240), __noswap_vgetq_lane_u32(*(uint32x4_t *) &__reint_240, __p3_240)}; \
37467 __ret_240 = __noswap_vcmlaq_rot90_f16(__rev0_240, __rev1_240, *(float16x8_t *) &__reint1_240); \
37468 __ret_240 = __builtin_shufflevector(__ret_240, __ret_240, 7, 6, 5, 4, 3, 2, 1, 0); \
37469 __ret_240; \
37470 })
37471 #endif
37472
37473 #if !defined(__aarch64__)
37474 #ifdef __LITTLE_ENDIAN__
37475 #define vqdmulhq_lane_s32(__p0_241, __p1_241, __p2_241) __extension__ ({ \
37476 int32x4_t __ret_241; \
37477 int32x4_t __s0_241 = __p0_241; \
37478 int32x2_t __s1_241 = __p1_241; \
37479 __ret_241 = vqdmulhq_s32(__s0_241, splatq_lane_s32(__s1_241, __p2_241)); \
37480 __ret_241; \
37481 })
37482 #else
37483 #define vqdmulhq_lane_s32(__p0_242, __p1_242, __p2_242) __extension__ ({ \
37484 int32x4_t __ret_242; \
37485 int32x4_t __s0_242 = __p0_242; \
37486 int32x2_t __s1_242 = __p1_242; \
37487 int32x4_t __rev0_242; __rev0_242 = __builtin_shufflevector(__s0_242, __s0_242, 3, 2, 1, 0); \
37488 int32x2_t __rev1_242; __rev1_242 = __builtin_shufflevector(__s1_242, __s1_242, 1, 0); \
37489 __ret_242 = __noswap_vqdmulhq_s32(__rev0_242, __noswap_splatq_lane_s32(__rev1_242, __p2_242)); \
37490 __ret_242 = __builtin_shufflevector(__ret_242, __ret_242, 3, 2, 1, 0); \
37491 __ret_242; \
37492 })
37493 #endif
37494
37495 #ifdef __LITTLE_ENDIAN__
37496 #define vqdmulhq_lane_s16(__p0_243, __p1_243, __p2_243) __extension__ ({ \
37497 int16x8_t __ret_243; \
37498 int16x8_t __s0_243 = __p0_243; \
37499 int16x4_t __s1_243 = __p1_243; \
37500 __ret_243 = vqdmulhq_s16(__s0_243, splatq_lane_s16(__s1_243, __p2_243)); \
37501 __ret_243; \
37502 })
37503 #else
37504 #define vqdmulhq_lane_s16(__p0_244, __p1_244, __p2_244) __extension__ ({ \
37505 int16x8_t __ret_244; \
37506 int16x8_t __s0_244 = __p0_244; \
37507 int16x4_t __s1_244 = __p1_244; \
37508 int16x8_t __rev0_244; __rev0_244 = __builtin_shufflevector(__s0_244, __s0_244, 7, 6, 5, 4, 3, 2, 1, 0); \
37509 int16x4_t __rev1_244; __rev1_244 = __builtin_shufflevector(__s1_244, __s1_244, 3, 2, 1, 0); \
37510 __ret_244 = __noswap_vqdmulhq_s16(__rev0_244, __noswap_splatq_lane_s16(__rev1_244, __p2_244)); \
37511 __ret_244 = __builtin_shufflevector(__ret_244, __ret_244, 7, 6, 5, 4, 3, 2, 1, 0); \
37512 __ret_244; \
37513 })
37514 #endif
37515
37516 #ifdef __LITTLE_ENDIAN__
37517 #define vqdmulh_lane_s32(__p0_245, __p1_245, __p2_245) __extension__ ({ \
37518 int32x2_t __ret_245; \
37519 int32x2_t __s0_245 = __p0_245; \
37520 int32x2_t __s1_245 = __p1_245; \
37521 __ret_245 = vqdmulh_s32(__s0_245, splat_lane_s32(__s1_245, __p2_245)); \
37522 __ret_245; \
37523 })
37524 #else
37525 #define vqdmulh_lane_s32(__p0_246, __p1_246, __p2_246) __extension__ ({ \
37526 int32x2_t __ret_246; \
37527 int32x2_t __s0_246 = __p0_246; \
37528 int32x2_t __s1_246 = __p1_246; \
37529 int32x2_t __rev0_246; __rev0_246 = __builtin_shufflevector(__s0_246, __s0_246, 1, 0); \
37530 int32x2_t __rev1_246; __rev1_246 = __builtin_shufflevector(__s1_246, __s1_246, 1, 0); \
37531 __ret_246 = __noswap_vqdmulh_s32(__rev0_246, __noswap_splat_lane_s32(__rev1_246, __p2_246)); \
37532 __ret_246 = __builtin_shufflevector(__ret_246, __ret_246, 1, 0); \
37533 __ret_246; \
37534 })
37535 #endif
37536
37537 #ifdef __LITTLE_ENDIAN__
37538 #define vqdmulh_lane_s16(__p0_247, __p1_247, __p2_247) __extension__ ({ \
37539 int16x4_t __ret_247; \
37540 int16x4_t __s0_247 = __p0_247; \
37541 int16x4_t __s1_247 = __p1_247; \
37542 __ret_247 = vqdmulh_s16(__s0_247, splat_lane_s16(__s1_247, __p2_247)); \
37543 __ret_247; \
37544 })
37545 #else
37546 #define vqdmulh_lane_s16(__p0_248, __p1_248, __p2_248) __extension__ ({ \
37547 int16x4_t __ret_248; \
37548 int16x4_t __s0_248 = __p0_248; \
37549 int16x4_t __s1_248 = __p1_248; \
37550 int16x4_t __rev0_248; __rev0_248 = __builtin_shufflevector(__s0_248, __s0_248, 3, 2, 1, 0); \
37551 int16x4_t __rev1_248; __rev1_248 = __builtin_shufflevector(__s1_248, __s1_248, 3, 2, 1, 0); \
37552 __ret_248 = __noswap_vqdmulh_s16(__rev0_248, __noswap_splat_lane_s16(__rev1_248, __p2_248)); \
37553 __ret_248 = __builtin_shufflevector(__ret_248, __ret_248, 3, 2, 1, 0); \
37554 __ret_248; \
37555 })
37556 #endif
37557
37558 #ifdef __LITTLE_ENDIAN__
37559 #define vqrdmulhq_lane_s32(__p0_249, __p1_249, __p2_249) __extension__ ({ \
37560 int32x4_t __ret_249; \
37561 int32x4_t __s0_249 = __p0_249; \
37562 int32x2_t __s1_249 = __p1_249; \
37563 __ret_249 = vqrdmulhq_s32(__s0_249, splatq_lane_s32(__s1_249, __p2_249)); \
37564 __ret_249; \
37565 })
37566 #else
37567 #define vqrdmulhq_lane_s32(__p0_250, __p1_250, __p2_250) __extension__ ({ \
37568 int32x4_t __ret_250; \
37569 int32x4_t __s0_250 = __p0_250; \
37570 int32x2_t __s1_250 = __p1_250; \
37571 int32x4_t __rev0_250; __rev0_250 = __builtin_shufflevector(__s0_250, __s0_250, 3, 2, 1, 0); \
37572 int32x2_t __rev1_250; __rev1_250 = __builtin_shufflevector(__s1_250, __s1_250, 1, 0); \
37573 __ret_250 = __noswap_vqrdmulhq_s32(__rev0_250, __noswap_splatq_lane_s32(__rev1_250, __p2_250)); \
37574 __ret_250 = __builtin_shufflevector(__ret_250, __ret_250, 3, 2, 1, 0); \
37575 __ret_250; \
37576 })
37577 #endif
37578
37579 #ifdef __LITTLE_ENDIAN__
37580 #define vqrdmulhq_lane_s16(__p0_251, __p1_251, __p2_251) __extension__ ({ \
37581 int16x8_t __ret_251; \
37582 int16x8_t __s0_251 = __p0_251; \
37583 int16x4_t __s1_251 = __p1_251; \
37584 __ret_251 = vqrdmulhq_s16(__s0_251, splatq_lane_s16(__s1_251, __p2_251)); \
37585 __ret_251; \
37586 })
37587 #else
37588 #define vqrdmulhq_lane_s16(__p0_252, __p1_252, __p2_252) __extension__ ({ \
37589 int16x8_t __ret_252; \
37590 int16x8_t __s0_252 = __p0_252; \
37591 int16x4_t __s1_252 = __p1_252; \
37592 int16x8_t __rev0_252; __rev0_252 = __builtin_shufflevector(__s0_252, __s0_252, 7, 6, 5, 4, 3, 2, 1, 0); \
37593 int16x4_t __rev1_252; __rev1_252 = __builtin_shufflevector(__s1_252, __s1_252, 3, 2, 1, 0); \
37594 __ret_252 = __noswap_vqrdmulhq_s16(__rev0_252, __noswap_splatq_lane_s16(__rev1_252, __p2_252)); \
37595 __ret_252 = __builtin_shufflevector(__ret_252, __ret_252, 7, 6, 5, 4, 3, 2, 1, 0); \
37596 __ret_252; \
37597 })
37598 #endif
37599
37600 #ifdef __LITTLE_ENDIAN__
37601 #define vqrdmulh_lane_s32(__p0_253, __p1_253, __p2_253) __extension__ ({ \
37602 int32x2_t __ret_253; \
37603 int32x2_t __s0_253 = __p0_253; \
37604 int32x2_t __s1_253 = __p1_253; \
37605 __ret_253 = vqrdmulh_s32(__s0_253, splat_lane_s32(__s1_253, __p2_253)); \
37606 __ret_253; \
37607 })
37608 #else
37609 #define vqrdmulh_lane_s32(__p0_254, __p1_254, __p2_254) __extension__ ({ \
37610 int32x2_t __ret_254; \
37611 int32x2_t __s0_254 = __p0_254; \
37612 int32x2_t __s1_254 = __p1_254; \
37613 int32x2_t __rev0_254; __rev0_254 = __builtin_shufflevector(__s0_254, __s0_254, 1, 0); \
37614 int32x2_t __rev1_254; __rev1_254 = __builtin_shufflevector(__s1_254, __s1_254, 1, 0); \
37615 __ret_254 = __noswap_vqrdmulh_s32(__rev0_254, __noswap_splat_lane_s32(__rev1_254, __p2_254)); \
37616 __ret_254 = __builtin_shufflevector(__ret_254, __ret_254, 1, 0); \
37617 __ret_254; \
37618 })
37619 #endif
37620
37621 #ifdef __LITTLE_ENDIAN__
37622 #define vqrdmulh_lane_s16(__p0_255, __p1_255, __p2_255) __extension__ ({ \
37623 int16x4_t __ret_255; \
37624 int16x4_t __s0_255 = __p0_255; \
37625 int16x4_t __s1_255 = __p1_255; \
37626 __ret_255 = vqrdmulh_s16(__s0_255, splat_lane_s16(__s1_255, __p2_255)); \
37627 __ret_255; \
37628 })
37629 #else
37630 #define vqrdmulh_lane_s16(__p0_256, __p1_256, __p2_256) __extension__ ({ \
37631 int16x4_t __ret_256; \
37632 int16x4_t __s0_256 = __p0_256; \
37633 int16x4_t __s1_256 = __p1_256; \
37634 int16x4_t __rev0_256; __rev0_256 = __builtin_shufflevector(__s0_256, __s0_256, 3, 2, 1, 0); \
37635 int16x4_t __rev1_256; __rev1_256 = __builtin_shufflevector(__s1_256, __s1_256, 3, 2, 1, 0); \
37636 __ret_256 = __noswap_vqrdmulh_s16(__rev0_256, __noswap_splat_lane_s16(__rev1_256, __p2_256)); \
37637 __ret_256 = __builtin_shufflevector(__ret_256, __ret_256, 3, 2, 1, 0); \
37638 __ret_256; \
37639 })
37640 #endif
37641
vreinterpret_p8_p16(poly16x4_t __p0)37642 __ai poly8x8_t vreinterpret_p8_p16(poly16x4_t __p0) {
37643 poly8x8_t __ret;
37644 __ret = (poly8x8_t)(__p0);
37645 return __ret;
37646 }
vreinterpret_p8_u8(uint8x8_t __p0)37647 __ai poly8x8_t vreinterpret_p8_u8(uint8x8_t __p0) {
37648 poly8x8_t __ret;
37649 __ret = (poly8x8_t)(__p0);
37650 return __ret;
37651 }
vreinterpret_p8_u32(uint32x2_t __p0)37652 __ai poly8x8_t vreinterpret_p8_u32(uint32x2_t __p0) {
37653 poly8x8_t __ret;
37654 __ret = (poly8x8_t)(__p0);
37655 return __ret;
37656 }
vreinterpret_p8_u64(uint64x1_t __p0)37657 __ai poly8x8_t vreinterpret_p8_u64(uint64x1_t __p0) {
37658 poly8x8_t __ret;
37659 __ret = (poly8x8_t)(__p0);
37660 return __ret;
37661 }
vreinterpret_p8_u16(uint16x4_t __p0)37662 __ai poly8x8_t vreinterpret_p8_u16(uint16x4_t __p0) {
37663 poly8x8_t __ret;
37664 __ret = (poly8x8_t)(__p0);
37665 return __ret;
37666 }
vreinterpret_p8_s8(int8x8_t __p0)37667 __ai poly8x8_t vreinterpret_p8_s8(int8x8_t __p0) {
37668 poly8x8_t __ret;
37669 __ret = (poly8x8_t)(__p0);
37670 return __ret;
37671 }
vreinterpret_p8_f32(float32x2_t __p0)37672 __ai poly8x8_t vreinterpret_p8_f32(float32x2_t __p0) {
37673 poly8x8_t __ret;
37674 __ret = (poly8x8_t)(__p0);
37675 return __ret;
37676 }
vreinterpret_p8_f16(float16x4_t __p0)37677 __ai poly8x8_t vreinterpret_p8_f16(float16x4_t __p0) {
37678 poly8x8_t __ret;
37679 __ret = (poly8x8_t)(__p0);
37680 return __ret;
37681 }
vreinterpret_p8_s32(int32x2_t __p0)37682 __ai poly8x8_t vreinterpret_p8_s32(int32x2_t __p0) {
37683 poly8x8_t __ret;
37684 __ret = (poly8x8_t)(__p0);
37685 return __ret;
37686 }
vreinterpret_p8_s64(int64x1_t __p0)37687 __ai poly8x8_t vreinterpret_p8_s64(int64x1_t __p0) {
37688 poly8x8_t __ret;
37689 __ret = (poly8x8_t)(__p0);
37690 return __ret;
37691 }
vreinterpret_p8_s16(int16x4_t __p0)37692 __ai poly8x8_t vreinterpret_p8_s16(int16x4_t __p0) {
37693 poly8x8_t __ret;
37694 __ret = (poly8x8_t)(__p0);
37695 return __ret;
37696 }
vreinterpret_p16_p8(poly8x8_t __p0)37697 __ai poly16x4_t vreinterpret_p16_p8(poly8x8_t __p0) {
37698 poly16x4_t __ret;
37699 __ret = (poly16x4_t)(__p0);
37700 return __ret;
37701 }
vreinterpret_p16_u8(uint8x8_t __p0)37702 __ai poly16x4_t vreinterpret_p16_u8(uint8x8_t __p0) {
37703 poly16x4_t __ret;
37704 __ret = (poly16x4_t)(__p0);
37705 return __ret;
37706 }
vreinterpret_p16_u32(uint32x2_t __p0)37707 __ai poly16x4_t vreinterpret_p16_u32(uint32x2_t __p0) {
37708 poly16x4_t __ret;
37709 __ret = (poly16x4_t)(__p0);
37710 return __ret;
37711 }
vreinterpret_p16_u64(uint64x1_t __p0)37712 __ai poly16x4_t vreinterpret_p16_u64(uint64x1_t __p0) {
37713 poly16x4_t __ret;
37714 __ret = (poly16x4_t)(__p0);
37715 return __ret;
37716 }
vreinterpret_p16_u16(uint16x4_t __p0)37717 __ai poly16x4_t vreinterpret_p16_u16(uint16x4_t __p0) {
37718 poly16x4_t __ret;
37719 __ret = (poly16x4_t)(__p0);
37720 return __ret;
37721 }
vreinterpret_p16_s8(int8x8_t __p0)37722 __ai poly16x4_t vreinterpret_p16_s8(int8x8_t __p0) {
37723 poly16x4_t __ret;
37724 __ret = (poly16x4_t)(__p0);
37725 return __ret;
37726 }
vreinterpret_p16_f32(float32x2_t __p0)37727 __ai poly16x4_t vreinterpret_p16_f32(float32x2_t __p0) {
37728 poly16x4_t __ret;
37729 __ret = (poly16x4_t)(__p0);
37730 return __ret;
37731 }
vreinterpret_p16_f16(float16x4_t __p0)37732 __ai poly16x4_t vreinterpret_p16_f16(float16x4_t __p0) {
37733 poly16x4_t __ret;
37734 __ret = (poly16x4_t)(__p0);
37735 return __ret;
37736 }
vreinterpret_p16_s32(int32x2_t __p0)37737 __ai poly16x4_t vreinterpret_p16_s32(int32x2_t __p0) {
37738 poly16x4_t __ret;
37739 __ret = (poly16x4_t)(__p0);
37740 return __ret;
37741 }
vreinterpret_p16_s64(int64x1_t __p0)37742 __ai poly16x4_t vreinterpret_p16_s64(int64x1_t __p0) {
37743 poly16x4_t __ret;
37744 __ret = (poly16x4_t)(__p0);
37745 return __ret;
37746 }
vreinterpret_p16_s16(int16x4_t __p0)37747 __ai poly16x4_t vreinterpret_p16_s16(int16x4_t __p0) {
37748 poly16x4_t __ret;
37749 __ret = (poly16x4_t)(__p0);
37750 return __ret;
37751 }
vreinterpretq_p8_p16(poly16x8_t __p0)37752 __ai poly8x16_t vreinterpretq_p8_p16(poly16x8_t __p0) {
37753 poly8x16_t __ret;
37754 __ret = (poly8x16_t)(__p0);
37755 return __ret;
37756 }
vreinterpretq_p8_u8(uint8x16_t __p0)37757 __ai poly8x16_t vreinterpretq_p8_u8(uint8x16_t __p0) {
37758 poly8x16_t __ret;
37759 __ret = (poly8x16_t)(__p0);
37760 return __ret;
37761 }
vreinterpretq_p8_u32(uint32x4_t __p0)37762 __ai poly8x16_t vreinterpretq_p8_u32(uint32x4_t __p0) {
37763 poly8x16_t __ret;
37764 __ret = (poly8x16_t)(__p0);
37765 return __ret;
37766 }
vreinterpretq_p8_u64(uint64x2_t __p0)37767 __ai poly8x16_t vreinterpretq_p8_u64(uint64x2_t __p0) {
37768 poly8x16_t __ret;
37769 __ret = (poly8x16_t)(__p0);
37770 return __ret;
37771 }
vreinterpretq_p8_u16(uint16x8_t __p0)37772 __ai poly8x16_t vreinterpretq_p8_u16(uint16x8_t __p0) {
37773 poly8x16_t __ret;
37774 __ret = (poly8x16_t)(__p0);
37775 return __ret;
37776 }
vreinterpretq_p8_s8(int8x16_t __p0)37777 __ai poly8x16_t vreinterpretq_p8_s8(int8x16_t __p0) {
37778 poly8x16_t __ret;
37779 __ret = (poly8x16_t)(__p0);
37780 return __ret;
37781 }
vreinterpretq_p8_f32(float32x4_t __p0)37782 __ai poly8x16_t vreinterpretq_p8_f32(float32x4_t __p0) {
37783 poly8x16_t __ret;
37784 __ret = (poly8x16_t)(__p0);
37785 return __ret;
37786 }
vreinterpretq_p8_f16(float16x8_t __p0)37787 __ai poly8x16_t vreinterpretq_p8_f16(float16x8_t __p0) {
37788 poly8x16_t __ret;
37789 __ret = (poly8x16_t)(__p0);
37790 return __ret;
37791 }
vreinterpretq_p8_s32(int32x4_t __p0)37792 __ai poly8x16_t vreinterpretq_p8_s32(int32x4_t __p0) {
37793 poly8x16_t __ret;
37794 __ret = (poly8x16_t)(__p0);
37795 return __ret;
37796 }
vreinterpretq_p8_s64(int64x2_t __p0)37797 __ai poly8x16_t vreinterpretq_p8_s64(int64x2_t __p0) {
37798 poly8x16_t __ret;
37799 __ret = (poly8x16_t)(__p0);
37800 return __ret;
37801 }
vreinterpretq_p8_s16(int16x8_t __p0)37802 __ai poly8x16_t vreinterpretq_p8_s16(int16x8_t __p0) {
37803 poly8x16_t __ret;
37804 __ret = (poly8x16_t)(__p0);
37805 return __ret;
37806 }
vreinterpretq_p16_p8(poly8x16_t __p0)37807 __ai poly16x8_t vreinterpretq_p16_p8(poly8x16_t __p0) {
37808 poly16x8_t __ret;
37809 __ret = (poly16x8_t)(__p0);
37810 return __ret;
37811 }
vreinterpretq_p16_u8(uint8x16_t __p0)37812 __ai poly16x8_t vreinterpretq_p16_u8(uint8x16_t __p0) {
37813 poly16x8_t __ret;
37814 __ret = (poly16x8_t)(__p0);
37815 return __ret;
37816 }
vreinterpretq_p16_u32(uint32x4_t __p0)37817 __ai poly16x8_t vreinterpretq_p16_u32(uint32x4_t __p0) {
37818 poly16x8_t __ret;
37819 __ret = (poly16x8_t)(__p0);
37820 return __ret;
37821 }
vreinterpretq_p16_u64(uint64x2_t __p0)37822 __ai poly16x8_t vreinterpretq_p16_u64(uint64x2_t __p0) {
37823 poly16x8_t __ret;
37824 __ret = (poly16x8_t)(__p0);
37825 return __ret;
37826 }
vreinterpretq_p16_u16(uint16x8_t __p0)37827 __ai poly16x8_t vreinterpretq_p16_u16(uint16x8_t __p0) {
37828 poly16x8_t __ret;
37829 __ret = (poly16x8_t)(__p0);
37830 return __ret;
37831 }
vreinterpretq_p16_s8(int8x16_t __p0)37832 __ai poly16x8_t vreinterpretq_p16_s8(int8x16_t __p0) {
37833 poly16x8_t __ret;
37834 __ret = (poly16x8_t)(__p0);
37835 return __ret;
37836 }
vreinterpretq_p16_f32(float32x4_t __p0)37837 __ai poly16x8_t vreinterpretq_p16_f32(float32x4_t __p0) {
37838 poly16x8_t __ret;
37839 __ret = (poly16x8_t)(__p0);
37840 return __ret;
37841 }
vreinterpretq_p16_f16(float16x8_t __p0)37842 __ai poly16x8_t vreinterpretq_p16_f16(float16x8_t __p0) {
37843 poly16x8_t __ret;
37844 __ret = (poly16x8_t)(__p0);
37845 return __ret;
37846 }
vreinterpretq_p16_s32(int32x4_t __p0)37847 __ai poly16x8_t vreinterpretq_p16_s32(int32x4_t __p0) {
37848 poly16x8_t __ret;
37849 __ret = (poly16x8_t)(__p0);
37850 return __ret;
37851 }
vreinterpretq_p16_s64(int64x2_t __p0)37852 __ai poly16x8_t vreinterpretq_p16_s64(int64x2_t __p0) {
37853 poly16x8_t __ret;
37854 __ret = (poly16x8_t)(__p0);
37855 return __ret;
37856 }
vreinterpretq_p16_s16(int16x8_t __p0)37857 __ai poly16x8_t vreinterpretq_p16_s16(int16x8_t __p0) {
37858 poly16x8_t __ret;
37859 __ret = (poly16x8_t)(__p0);
37860 return __ret;
37861 }
vreinterpretq_u8_p8(poly8x16_t __p0)37862 __ai uint8x16_t vreinterpretq_u8_p8(poly8x16_t __p0) {
37863 uint8x16_t __ret;
37864 __ret = (uint8x16_t)(__p0);
37865 return __ret;
37866 }
vreinterpretq_u8_p16(poly16x8_t __p0)37867 __ai uint8x16_t vreinterpretq_u8_p16(poly16x8_t __p0) {
37868 uint8x16_t __ret;
37869 __ret = (uint8x16_t)(__p0);
37870 return __ret;
37871 }
vreinterpretq_u8_u32(uint32x4_t __p0)37872 __ai uint8x16_t vreinterpretq_u8_u32(uint32x4_t __p0) {
37873 uint8x16_t __ret;
37874 __ret = (uint8x16_t)(__p0);
37875 return __ret;
37876 }
vreinterpretq_u8_u64(uint64x2_t __p0)37877 __ai uint8x16_t vreinterpretq_u8_u64(uint64x2_t __p0) {
37878 uint8x16_t __ret;
37879 __ret = (uint8x16_t)(__p0);
37880 return __ret;
37881 }
vreinterpretq_u8_u16(uint16x8_t __p0)37882 __ai uint8x16_t vreinterpretq_u8_u16(uint16x8_t __p0) {
37883 uint8x16_t __ret;
37884 __ret = (uint8x16_t)(__p0);
37885 return __ret;
37886 }
vreinterpretq_u8_s8(int8x16_t __p0)37887 __ai uint8x16_t vreinterpretq_u8_s8(int8x16_t __p0) {
37888 uint8x16_t __ret;
37889 __ret = (uint8x16_t)(__p0);
37890 return __ret;
37891 }
vreinterpretq_u8_f32(float32x4_t __p0)37892 __ai uint8x16_t vreinterpretq_u8_f32(float32x4_t __p0) {
37893 uint8x16_t __ret;
37894 __ret = (uint8x16_t)(__p0);
37895 return __ret;
37896 }
vreinterpretq_u8_f16(float16x8_t __p0)37897 __ai uint8x16_t vreinterpretq_u8_f16(float16x8_t __p0) {
37898 uint8x16_t __ret;
37899 __ret = (uint8x16_t)(__p0);
37900 return __ret;
37901 }
vreinterpretq_u8_s32(int32x4_t __p0)37902 __ai uint8x16_t vreinterpretq_u8_s32(int32x4_t __p0) {
37903 uint8x16_t __ret;
37904 __ret = (uint8x16_t)(__p0);
37905 return __ret;
37906 }
vreinterpretq_u8_s64(int64x2_t __p0)37907 __ai uint8x16_t vreinterpretq_u8_s64(int64x2_t __p0) {
37908 uint8x16_t __ret;
37909 __ret = (uint8x16_t)(__p0);
37910 return __ret;
37911 }
vreinterpretq_u8_s16(int16x8_t __p0)37912 __ai uint8x16_t vreinterpretq_u8_s16(int16x8_t __p0) {
37913 uint8x16_t __ret;
37914 __ret = (uint8x16_t)(__p0);
37915 return __ret;
37916 }
vreinterpretq_u32_p8(poly8x16_t __p0)37917 __ai uint32x4_t vreinterpretq_u32_p8(poly8x16_t __p0) {
37918 uint32x4_t __ret;
37919 __ret = (uint32x4_t)(__p0);
37920 return __ret;
37921 }
vreinterpretq_u32_p16(poly16x8_t __p0)37922 __ai uint32x4_t vreinterpretq_u32_p16(poly16x8_t __p0) {
37923 uint32x4_t __ret;
37924 __ret = (uint32x4_t)(__p0);
37925 return __ret;
37926 }
vreinterpretq_u32_u8(uint8x16_t __p0)37927 __ai uint32x4_t vreinterpretq_u32_u8(uint8x16_t __p0) {
37928 uint32x4_t __ret;
37929 __ret = (uint32x4_t)(__p0);
37930 return __ret;
37931 }
vreinterpretq_u32_u64(uint64x2_t __p0)37932 __ai uint32x4_t vreinterpretq_u32_u64(uint64x2_t __p0) {
37933 uint32x4_t __ret;
37934 __ret = (uint32x4_t)(__p0);
37935 return __ret;
37936 }
vreinterpretq_u32_u16(uint16x8_t __p0)37937 __ai uint32x4_t vreinterpretq_u32_u16(uint16x8_t __p0) {
37938 uint32x4_t __ret;
37939 __ret = (uint32x4_t)(__p0);
37940 return __ret;
37941 }
vreinterpretq_u32_s8(int8x16_t __p0)37942 __ai uint32x4_t vreinterpretq_u32_s8(int8x16_t __p0) {
37943 uint32x4_t __ret;
37944 __ret = (uint32x4_t)(__p0);
37945 return __ret;
37946 }
vreinterpretq_u32_f32(float32x4_t __p0)37947 __ai uint32x4_t vreinterpretq_u32_f32(float32x4_t __p0) {
37948 uint32x4_t __ret;
37949 __ret = (uint32x4_t)(__p0);
37950 return __ret;
37951 }
vreinterpretq_u32_f16(float16x8_t __p0)37952 __ai uint32x4_t vreinterpretq_u32_f16(float16x8_t __p0) {
37953 uint32x4_t __ret;
37954 __ret = (uint32x4_t)(__p0);
37955 return __ret;
37956 }
vreinterpretq_u32_s32(int32x4_t __p0)37957 __ai uint32x4_t vreinterpretq_u32_s32(int32x4_t __p0) {
37958 uint32x4_t __ret;
37959 __ret = (uint32x4_t)(__p0);
37960 return __ret;
37961 }
vreinterpretq_u32_s64(int64x2_t __p0)37962 __ai uint32x4_t vreinterpretq_u32_s64(int64x2_t __p0) {
37963 uint32x4_t __ret;
37964 __ret = (uint32x4_t)(__p0);
37965 return __ret;
37966 }
vreinterpretq_u32_s16(int16x8_t __p0)37967 __ai uint32x4_t vreinterpretq_u32_s16(int16x8_t __p0) {
37968 uint32x4_t __ret;
37969 __ret = (uint32x4_t)(__p0);
37970 return __ret;
37971 }
vreinterpretq_u64_p8(poly8x16_t __p0)37972 __ai uint64x2_t vreinterpretq_u64_p8(poly8x16_t __p0) {
37973 uint64x2_t __ret;
37974 __ret = (uint64x2_t)(__p0);
37975 return __ret;
37976 }
vreinterpretq_u64_p16(poly16x8_t __p0)37977 __ai uint64x2_t vreinterpretq_u64_p16(poly16x8_t __p0) {
37978 uint64x2_t __ret;
37979 __ret = (uint64x2_t)(__p0);
37980 return __ret;
37981 }
vreinterpretq_u64_u8(uint8x16_t __p0)37982 __ai uint64x2_t vreinterpretq_u64_u8(uint8x16_t __p0) {
37983 uint64x2_t __ret;
37984 __ret = (uint64x2_t)(__p0);
37985 return __ret;
37986 }
vreinterpretq_u64_u32(uint32x4_t __p0)37987 __ai uint64x2_t vreinterpretq_u64_u32(uint32x4_t __p0) {
37988 uint64x2_t __ret;
37989 __ret = (uint64x2_t)(__p0);
37990 return __ret;
37991 }
vreinterpretq_u64_u16(uint16x8_t __p0)37992 __ai uint64x2_t vreinterpretq_u64_u16(uint16x8_t __p0) {
37993 uint64x2_t __ret;
37994 __ret = (uint64x2_t)(__p0);
37995 return __ret;
37996 }
vreinterpretq_u64_s8(int8x16_t __p0)37997 __ai uint64x2_t vreinterpretq_u64_s8(int8x16_t __p0) {
37998 uint64x2_t __ret;
37999 __ret = (uint64x2_t)(__p0);
38000 return __ret;
38001 }
vreinterpretq_u64_f32(float32x4_t __p0)38002 __ai uint64x2_t vreinterpretq_u64_f32(float32x4_t __p0) {
38003 uint64x2_t __ret;
38004 __ret = (uint64x2_t)(__p0);
38005 return __ret;
38006 }
vreinterpretq_u64_f16(float16x8_t __p0)38007 __ai uint64x2_t vreinterpretq_u64_f16(float16x8_t __p0) {
38008 uint64x2_t __ret;
38009 __ret = (uint64x2_t)(__p0);
38010 return __ret;
38011 }
vreinterpretq_u64_s32(int32x4_t __p0)38012 __ai uint64x2_t vreinterpretq_u64_s32(int32x4_t __p0) {
38013 uint64x2_t __ret;
38014 __ret = (uint64x2_t)(__p0);
38015 return __ret;
38016 }
vreinterpretq_u64_s64(int64x2_t __p0)38017 __ai uint64x2_t vreinterpretq_u64_s64(int64x2_t __p0) {
38018 uint64x2_t __ret;
38019 __ret = (uint64x2_t)(__p0);
38020 return __ret;
38021 }
vreinterpretq_u64_s16(int16x8_t __p0)38022 __ai uint64x2_t vreinterpretq_u64_s16(int16x8_t __p0) {
38023 uint64x2_t __ret;
38024 __ret = (uint64x2_t)(__p0);
38025 return __ret;
38026 }
vreinterpretq_u16_p8(poly8x16_t __p0)38027 __ai uint16x8_t vreinterpretq_u16_p8(poly8x16_t __p0) {
38028 uint16x8_t __ret;
38029 __ret = (uint16x8_t)(__p0);
38030 return __ret;
38031 }
vreinterpretq_u16_p16(poly16x8_t __p0)38032 __ai uint16x8_t vreinterpretq_u16_p16(poly16x8_t __p0) {
38033 uint16x8_t __ret;
38034 __ret = (uint16x8_t)(__p0);
38035 return __ret;
38036 }
vreinterpretq_u16_u8(uint8x16_t __p0)38037 __ai uint16x8_t vreinterpretq_u16_u8(uint8x16_t __p0) {
38038 uint16x8_t __ret;
38039 __ret = (uint16x8_t)(__p0);
38040 return __ret;
38041 }
vreinterpretq_u16_u32(uint32x4_t __p0)38042 __ai uint16x8_t vreinterpretq_u16_u32(uint32x4_t __p0) {
38043 uint16x8_t __ret;
38044 __ret = (uint16x8_t)(__p0);
38045 return __ret;
38046 }
vreinterpretq_u16_u64(uint64x2_t __p0)38047 __ai uint16x8_t vreinterpretq_u16_u64(uint64x2_t __p0) {
38048 uint16x8_t __ret;
38049 __ret = (uint16x8_t)(__p0);
38050 return __ret;
38051 }
vreinterpretq_u16_s8(int8x16_t __p0)38052 __ai uint16x8_t vreinterpretq_u16_s8(int8x16_t __p0) {
38053 uint16x8_t __ret;
38054 __ret = (uint16x8_t)(__p0);
38055 return __ret;
38056 }
vreinterpretq_u16_f32(float32x4_t __p0)38057 __ai uint16x8_t vreinterpretq_u16_f32(float32x4_t __p0) {
38058 uint16x8_t __ret;
38059 __ret = (uint16x8_t)(__p0);
38060 return __ret;
38061 }
vreinterpretq_u16_f16(float16x8_t __p0)38062 __ai uint16x8_t vreinterpretq_u16_f16(float16x8_t __p0) {
38063 uint16x8_t __ret;
38064 __ret = (uint16x8_t)(__p0);
38065 return __ret;
38066 }
vreinterpretq_u16_s32(int32x4_t __p0)38067 __ai uint16x8_t vreinterpretq_u16_s32(int32x4_t __p0) {
38068 uint16x8_t __ret;
38069 __ret = (uint16x8_t)(__p0);
38070 return __ret;
38071 }
vreinterpretq_u16_s64(int64x2_t __p0)38072 __ai uint16x8_t vreinterpretq_u16_s64(int64x2_t __p0) {
38073 uint16x8_t __ret;
38074 __ret = (uint16x8_t)(__p0);
38075 return __ret;
38076 }
vreinterpretq_u16_s16(int16x8_t __p0)38077 __ai uint16x8_t vreinterpretq_u16_s16(int16x8_t __p0) {
38078 uint16x8_t __ret;
38079 __ret = (uint16x8_t)(__p0);
38080 return __ret;
38081 }
vreinterpretq_s8_p8(poly8x16_t __p0)38082 __ai int8x16_t vreinterpretq_s8_p8(poly8x16_t __p0) {
38083 int8x16_t __ret;
38084 __ret = (int8x16_t)(__p0);
38085 return __ret;
38086 }
vreinterpretq_s8_p16(poly16x8_t __p0)38087 __ai int8x16_t vreinterpretq_s8_p16(poly16x8_t __p0) {
38088 int8x16_t __ret;
38089 __ret = (int8x16_t)(__p0);
38090 return __ret;
38091 }
vreinterpretq_s8_u8(uint8x16_t __p0)38092 __ai int8x16_t vreinterpretq_s8_u8(uint8x16_t __p0) {
38093 int8x16_t __ret;
38094 __ret = (int8x16_t)(__p0);
38095 return __ret;
38096 }
vreinterpretq_s8_u32(uint32x4_t __p0)38097 __ai int8x16_t vreinterpretq_s8_u32(uint32x4_t __p0) {
38098 int8x16_t __ret;
38099 __ret = (int8x16_t)(__p0);
38100 return __ret;
38101 }
vreinterpretq_s8_u64(uint64x2_t __p0)38102 __ai int8x16_t vreinterpretq_s8_u64(uint64x2_t __p0) {
38103 int8x16_t __ret;
38104 __ret = (int8x16_t)(__p0);
38105 return __ret;
38106 }
vreinterpretq_s8_u16(uint16x8_t __p0)38107 __ai int8x16_t vreinterpretq_s8_u16(uint16x8_t __p0) {
38108 int8x16_t __ret;
38109 __ret = (int8x16_t)(__p0);
38110 return __ret;
38111 }
vreinterpretq_s8_f32(float32x4_t __p0)38112 __ai int8x16_t vreinterpretq_s8_f32(float32x4_t __p0) {
38113 int8x16_t __ret;
38114 __ret = (int8x16_t)(__p0);
38115 return __ret;
38116 }
vreinterpretq_s8_f16(float16x8_t __p0)38117 __ai int8x16_t vreinterpretq_s8_f16(float16x8_t __p0) {
38118 int8x16_t __ret;
38119 __ret = (int8x16_t)(__p0);
38120 return __ret;
38121 }
vreinterpretq_s8_s32(int32x4_t __p0)38122 __ai int8x16_t vreinterpretq_s8_s32(int32x4_t __p0) {
38123 int8x16_t __ret;
38124 __ret = (int8x16_t)(__p0);
38125 return __ret;
38126 }
vreinterpretq_s8_s64(int64x2_t __p0)38127 __ai int8x16_t vreinterpretq_s8_s64(int64x2_t __p0) {
38128 int8x16_t __ret;
38129 __ret = (int8x16_t)(__p0);
38130 return __ret;
38131 }
vreinterpretq_s8_s16(int16x8_t __p0)38132 __ai int8x16_t vreinterpretq_s8_s16(int16x8_t __p0) {
38133 int8x16_t __ret;
38134 __ret = (int8x16_t)(__p0);
38135 return __ret;
38136 }
vreinterpretq_f32_p8(poly8x16_t __p0)38137 __ai float32x4_t vreinterpretq_f32_p8(poly8x16_t __p0) {
38138 float32x4_t __ret;
38139 __ret = (float32x4_t)(__p0);
38140 return __ret;
38141 }
vreinterpretq_f32_p16(poly16x8_t __p0)38142 __ai float32x4_t vreinterpretq_f32_p16(poly16x8_t __p0) {
38143 float32x4_t __ret;
38144 __ret = (float32x4_t)(__p0);
38145 return __ret;
38146 }
vreinterpretq_f32_u8(uint8x16_t __p0)38147 __ai float32x4_t vreinterpretq_f32_u8(uint8x16_t __p0) {
38148 float32x4_t __ret;
38149 __ret = (float32x4_t)(__p0);
38150 return __ret;
38151 }
vreinterpretq_f32_u32(uint32x4_t __p0)38152