Stephen Hines | 51a0ffb | 2014-02-14 00:25:07 -0800 | [diff] [blame] | 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 | #if !defined(__ARM_NEON__) && !defined(__ARM_NEON) |
| 28 | #error "NEON support not enabled" |
| 29 | #endif |
| 30 | |
| 31 | #include <stdint.h> |
| 32 | |
| 33 | typedef float float32_t; |
| 34 | typedef __fp16 float16_t; |
| 35 | #ifdef __aarch64__ |
| 36 | typedef double float64_t; |
| 37 | #endif |
| 38 | |
| 39 | #ifdef __aarch64__ |
| 40 | typedef uint8_t poly8_t; |
| 41 | typedef uint16_t poly16_t; |
| 42 | typedef uint64_t poly64_t; |
| 43 | #else |
| 44 | typedef int8_t poly8_t; |
| 45 | typedef int16_t poly16_t; |
| 46 | #endif |
| 47 | typedef __attribute__((neon_vector_type(8))) int8_t int8x8_t; |
| 48 | typedef __attribute__((neon_vector_type(16))) int8_t int8x16_t; |
| 49 | typedef __attribute__((neon_vector_type(4))) int16_t int16x4_t; |
| 50 | typedef __attribute__((neon_vector_type(8))) int16_t int16x8_t; |
| 51 | typedef __attribute__((neon_vector_type(2))) int32_t int32x2_t; |
| 52 | typedef __attribute__((neon_vector_type(4))) int32_t int32x4_t; |
| 53 | typedef __attribute__((neon_vector_type(1))) int64_t int64x1_t; |
| 54 | typedef __attribute__((neon_vector_type(2))) int64_t int64x2_t; |
| 55 | typedef __attribute__((neon_vector_type(8))) uint8_t uint8x8_t; |
| 56 | typedef __attribute__((neon_vector_type(16))) uint8_t uint8x16_t; |
| 57 | typedef __attribute__((neon_vector_type(4))) uint16_t uint16x4_t; |
| 58 | typedef __attribute__((neon_vector_type(8))) uint16_t uint16x8_t; |
| 59 | typedef __attribute__((neon_vector_type(2))) uint32_t uint32x2_t; |
| 60 | typedef __attribute__((neon_vector_type(4))) uint32_t uint32x4_t; |
| 61 | typedef __attribute__((neon_vector_type(1))) uint64_t uint64x1_t; |
| 62 | typedef __attribute__((neon_vector_type(2))) uint64_t uint64x2_t; |
| 63 | typedef __attribute__((neon_vector_type(4))) float16_t float16x4_t; |
| 64 | typedef __attribute__((neon_vector_type(8))) float16_t float16x8_t; |
| 65 | typedef __attribute__((neon_vector_type(2))) float32_t float32x2_t; |
| 66 | typedef __attribute__((neon_vector_type(4))) float32_t float32x4_t; |
| 67 | #ifdef __aarch64__ |
| 68 | typedef __attribute__((neon_vector_type(1))) float64_t float64x1_t; |
| 69 | typedef __attribute__((neon_vector_type(2))) float64_t float64x2_t; |
| 70 | #endif |
| 71 | typedef __attribute__((neon_polyvector_type(8))) poly8_t poly8x8_t; |
| 72 | typedef __attribute__((neon_polyvector_type(16))) poly8_t poly8x16_t; |
| 73 | typedef __attribute__((neon_polyvector_type(4))) poly16_t poly16x4_t; |
| 74 | typedef __attribute__((neon_polyvector_type(8))) poly16_t poly16x8_t; |
| 75 | #ifdef __aarch64__ |
| 76 | typedef __attribute__((neon_polyvector_type(1))) poly64_t poly64x1_t; |
| 77 | typedef __attribute__((neon_polyvector_type(2))) poly64_t poly64x2_t; |
| 78 | #endif |
| 79 | |
| 80 | typedef struct int8x8x2_t { |
| 81 | int8x8_t val[2]; |
| 82 | } int8x8x2_t; |
| 83 | |
| 84 | typedef struct int8x16x2_t { |
| 85 | int8x16_t val[2]; |
| 86 | } int8x16x2_t; |
| 87 | |
| 88 | typedef struct int16x4x2_t { |
| 89 | int16x4_t val[2]; |
| 90 | } int16x4x2_t; |
| 91 | |
| 92 | typedef struct int16x8x2_t { |
| 93 | int16x8_t val[2]; |
| 94 | } int16x8x2_t; |
| 95 | |
| 96 | typedef struct int32x2x2_t { |
| 97 | int32x2_t val[2]; |
| 98 | } int32x2x2_t; |
| 99 | |
| 100 | typedef struct int32x4x2_t { |
| 101 | int32x4_t val[2]; |
| 102 | } int32x4x2_t; |
| 103 | |
| 104 | typedef struct int64x1x2_t { |
| 105 | int64x1_t val[2]; |
| 106 | } int64x1x2_t; |
| 107 | |
| 108 | typedef struct int64x2x2_t { |
| 109 | int64x2_t val[2]; |
| 110 | } int64x2x2_t; |
| 111 | |
| 112 | typedef struct uint8x8x2_t { |
| 113 | uint8x8_t val[2]; |
| 114 | } uint8x8x2_t; |
| 115 | |
| 116 | typedef struct uint8x16x2_t { |
| 117 | uint8x16_t val[2]; |
| 118 | } uint8x16x2_t; |
| 119 | |
| 120 | typedef struct uint16x4x2_t { |
| 121 | uint16x4_t val[2]; |
| 122 | } uint16x4x2_t; |
| 123 | |
| 124 | typedef struct uint16x8x2_t { |
| 125 | uint16x8_t val[2]; |
| 126 | } uint16x8x2_t; |
| 127 | |
| 128 | typedef struct uint32x2x2_t { |
| 129 | uint32x2_t val[2]; |
| 130 | } uint32x2x2_t; |
| 131 | |
| 132 | typedef struct uint32x4x2_t { |
| 133 | uint32x4_t val[2]; |
| 134 | } uint32x4x2_t; |
| 135 | |
| 136 | typedef struct uint64x1x2_t { |
| 137 | uint64x1_t val[2]; |
| 138 | } uint64x1x2_t; |
| 139 | |
| 140 | typedef struct uint64x2x2_t { |
| 141 | uint64x2_t val[2]; |
| 142 | } uint64x2x2_t; |
| 143 | |
| 144 | typedef struct float16x4x2_t { |
| 145 | float16x4_t val[2]; |
| 146 | } float16x4x2_t; |
| 147 | |
| 148 | typedef struct float16x8x2_t { |
| 149 | float16x8_t val[2]; |
| 150 | } float16x8x2_t; |
| 151 | |
| 152 | typedef struct float32x2x2_t { |
| 153 | float32x2_t val[2]; |
| 154 | } float32x2x2_t; |
| 155 | |
| 156 | typedef struct float32x4x2_t { |
| 157 | float32x4_t val[2]; |
| 158 | } float32x4x2_t; |
| 159 | |
| 160 | #ifdef __aarch64__ |
| 161 | typedef struct float64x1x2_t { |
| 162 | float64x1_t val[2]; |
| 163 | } float64x1x2_t; |
| 164 | |
| 165 | typedef struct float64x2x2_t { |
| 166 | float64x2_t val[2]; |
| 167 | } float64x2x2_t; |
| 168 | |
| 169 | #endif |
| 170 | typedef struct poly8x8x2_t { |
| 171 | poly8x8_t val[2]; |
| 172 | } poly8x8x2_t; |
| 173 | |
| 174 | typedef struct poly8x16x2_t { |
| 175 | poly8x16_t val[2]; |
| 176 | } poly8x16x2_t; |
| 177 | |
| 178 | typedef struct poly16x4x2_t { |
| 179 | poly16x4_t val[2]; |
| 180 | } poly16x4x2_t; |
| 181 | |
| 182 | typedef struct poly16x8x2_t { |
| 183 | poly16x8_t val[2]; |
| 184 | } poly16x8x2_t; |
| 185 | |
| 186 | #ifdef __aarch64__ |
| 187 | typedef struct poly64x1x2_t { |
| 188 | poly64x1_t val[2]; |
| 189 | } poly64x1x2_t; |
| 190 | |
| 191 | typedef struct poly64x2x2_t { |
| 192 | poly64x2_t val[2]; |
| 193 | } poly64x2x2_t; |
| 194 | |
| 195 | #endif |
| 196 | typedef struct int8x8x3_t { |
| 197 | int8x8_t val[3]; |
| 198 | } int8x8x3_t; |
| 199 | |
| 200 | typedef struct int8x16x3_t { |
| 201 | int8x16_t val[3]; |
| 202 | } int8x16x3_t; |
| 203 | |
| 204 | typedef struct int16x4x3_t { |
| 205 | int16x4_t val[3]; |
| 206 | } int16x4x3_t; |
| 207 | |
| 208 | typedef struct int16x8x3_t { |
| 209 | int16x8_t val[3]; |
| 210 | } int16x8x3_t; |
| 211 | |
| 212 | typedef struct int32x2x3_t { |
| 213 | int32x2_t val[3]; |
| 214 | } int32x2x3_t; |
| 215 | |
| 216 | typedef struct int32x4x3_t { |
| 217 | int32x4_t val[3]; |
| 218 | } int32x4x3_t; |
| 219 | |
| 220 | typedef struct int64x1x3_t { |
| 221 | int64x1_t val[3]; |
| 222 | } int64x1x3_t; |
| 223 | |
| 224 | typedef struct int64x2x3_t { |
| 225 | int64x2_t val[3]; |
| 226 | } int64x2x3_t; |
| 227 | |
| 228 | typedef struct uint8x8x3_t { |
| 229 | uint8x8_t val[3]; |
| 230 | } uint8x8x3_t; |
| 231 | |
| 232 | typedef struct uint8x16x3_t { |
| 233 | uint8x16_t val[3]; |
| 234 | } uint8x16x3_t; |
| 235 | |
| 236 | typedef struct uint16x4x3_t { |
| 237 | uint16x4_t val[3]; |
| 238 | } uint16x4x3_t; |
| 239 | |
| 240 | typedef struct uint16x8x3_t { |
| 241 | uint16x8_t val[3]; |
| 242 | } uint16x8x3_t; |
| 243 | |
| 244 | typedef struct uint32x2x3_t { |
| 245 | uint32x2_t val[3]; |
| 246 | } uint32x2x3_t; |
| 247 | |
| 248 | typedef struct uint32x4x3_t { |
| 249 | uint32x4_t val[3]; |
| 250 | } uint32x4x3_t; |
| 251 | |
| 252 | typedef struct uint64x1x3_t { |
| 253 | uint64x1_t val[3]; |
| 254 | } uint64x1x3_t; |
| 255 | |
| 256 | typedef struct uint64x2x3_t { |
| 257 | uint64x2_t val[3]; |
| 258 | } uint64x2x3_t; |
| 259 | |
| 260 | typedef struct float16x4x3_t { |
| 261 | float16x4_t val[3]; |
| 262 | } float16x4x3_t; |
| 263 | |
| 264 | typedef struct float16x8x3_t { |
| 265 | float16x8_t val[3]; |
| 266 | } float16x8x3_t; |
| 267 | |
| 268 | typedef struct float32x2x3_t { |
| 269 | float32x2_t val[3]; |
| 270 | } float32x2x3_t; |
| 271 | |
| 272 | typedef struct float32x4x3_t { |
| 273 | float32x4_t val[3]; |
| 274 | } float32x4x3_t; |
| 275 | |
| 276 | #ifdef __aarch64__ |
| 277 | typedef struct float64x1x3_t { |
| 278 | float64x1_t val[3]; |
| 279 | } float64x1x3_t; |
| 280 | |
| 281 | typedef struct float64x2x3_t { |
| 282 | float64x2_t val[3]; |
| 283 | } float64x2x3_t; |
| 284 | |
| 285 | #endif |
| 286 | typedef struct poly8x8x3_t { |
| 287 | poly8x8_t val[3]; |
| 288 | } poly8x8x3_t; |
| 289 | |
| 290 | typedef struct poly8x16x3_t { |
| 291 | poly8x16_t val[3]; |
| 292 | } poly8x16x3_t; |
| 293 | |
| 294 | typedef struct poly16x4x3_t { |
| 295 | poly16x4_t val[3]; |
| 296 | } poly16x4x3_t; |
| 297 | |
| 298 | typedef struct poly16x8x3_t { |
| 299 | poly16x8_t val[3]; |
| 300 | } poly16x8x3_t; |
| 301 | |
| 302 | #ifdef __aarch64__ |
| 303 | typedef struct poly64x1x3_t { |
| 304 | poly64x1_t val[3]; |
| 305 | } poly64x1x3_t; |
| 306 | |
| 307 | typedef struct poly64x2x3_t { |
| 308 | poly64x2_t val[3]; |
| 309 | } poly64x2x3_t; |
| 310 | |
| 311 | #endif |
| 312 | typedef struct int8x8x4_t { |
| 313 | int8x8_t val[4]; |
| 314 | } int8x8x4_t; |
| 315 | |
| 316 | typedef struct int8x16x4_t { |
| 317 | int8x16_t val[4]; |
| 318 | } int8x16x4_t; |
| 319 | |
| 320 | typedef struct int16x4x4_t { |
| 321 | int16x4_t val[4]; |
| 322 | } int16x4x4_t; |
| 323 | |
| 324 | typedef struct int16x8x4_t { |
| 325 | int16x8_t val[4]; |
| 326 | } int16x8x4_t; |
| 327 | |
| 328 | typedef struct int32x2x4_t { |
| 329 | int32x2_t val[4]; |
| 330 | } int32x2x4_t; |
| 331 | |
| 332 | typedef struct int32x4x4_t { |
| 333 | int32x4_t val[4]; |
| 334 | } int32x4x4_t; |
| 335 | |
| 336 | typedef struct int64x1x4_t { |
| 337 | int64x1_t val[4]; |
| 338 | } int64x1x4_t; |
| 339 | |
| 340 | typedef struct int64x2x4_t { |
| 341 | int64x2_t val[4]; |
| 342 | } int64x2x4_t; |
| 343 | |
| 344 | typedef struct uint8x8x4_t { |
| 345 | uint8x8_t val[4]; |
| 346 | } uint8x8x4_t; |
| 347 | |
| 348 | typedef struct uint8x16x4_t { |
| 349 | uint8x16_t val[4]; |
| 350 | } uint8x16x4_t; |
| 351 | |
| 352 | typedef struct uint16x4x4_t { |
| 353 | uint16x4_t val[4]; |
| 354 | } uint16x4x4_t; |
| 355 | |
| 356 | typedef struct uint16x8x4_t { |
| 357 | uint16x8_t val[4]; |
| 358 | } uint16x8x4_t; |
| 359 | |
| 360 | typedef struct uint32x2x4_t { |
| 361 | uint32x2_t val[4]; |
| 362 | } uint32x2x4_t; |
| 363 | |
| 364 | typedef struct uint32x4x4_t { |
| 365 | uint32x4_t val[4]; |
| 366 | } uint32x4x4_t; |
| 367 | |
| 368 | typedef struct uint64x1x4_t { |
| 369 | uint64x1_t val[4]; |
| 370 | } uint64x1x4_t; |
| 371 | |
| 372 | typedef struct uint64x2x4_t { |
| 373 | uint64x2_t val[4]; |
| 374 | } uint64x2x4_t; |
| 375 | |
| 376 | typedef struct float16x4x4_t { |
| 377 | float16x4_t val[4]; |
| 378 | } float16x4x4_t; |
| 379 | |
| 380 | typedef struct float16x8x4_t { |
| 381 | float16x8_t val[4]; |
| 382 | } float16x8x4_t; |
| 383 | |
| 384 | typedef struct float32x2x4_t { |
| 385 | float32x2_t val[4]; |
| 386 | } float32x2x4_t; |
| 387 | |
| 388 | typedef struct float32x4x4_t { |
| 389 | float32x4_t val[4]; |
| 390 | } float32x4x4_t; |
| 391 | |
| 392 | #ifdef __aarch64__ |
| 393 | typedef struct float64x1x4_t { |
| 394 | float64x1_t val[4]; |
| 395 | } float64x1x4_t; |
| 396 | |
| 397 | typedef struct float64x2x4_t { |
| 398 | float64x2_t val[4]; |
| 399 | } float64x2x4_t; |
| 400 | |
| 401 | #endif |
| 402 | typedef struct poly8x8x4_t { |
| 403 | poly8x8_t val[4]; |
| 404 | } poly8x8x4_t; |
| 405 | |
| 406 | typedef struct poly8x16x4_t { |
| 407 | poly8x16_t val[4]; |
| 408 | } poly8x16x4_t; |
| 409 | |
| 410 | typedef struct poly16x4x4_t { |
| 411 | poly16x4_t val[4]; |
| 412 | } poly16x4x4_t; |
| 413 | |
| 414 | typedef struct poly16x8x4_t { |
| 415 | poly16x8_t val[4]; |
| 416 | } poly16x8x4_t; |
| 417 | |
| 418 | #ifdef __aarch64__ |
| 419 | typedef struct poly64x1x4_t { |
| 420 | poly64x1_t val[4]; |
| 421 | } poly64x1x4_t; |
| 422 | |
| 423 | typedef struct poly64x2x4_t { |
| 424 | poly64x2_t val[4]; |
| 425 | } poly64x2x4_t; |
| 426 | |
| 427 | #endif |
| 428 | |
| 429 | #define __ai static inline __attribute__((__always_inline__, __nodebug__)) |
| 430 | |
| 431 | __ai int16x8_t vmovl_s8(int8x8_t __a) { |
| 432 | return (int16x8_t)__builtin_neon_vmovl_v(__a, 33); } |
| 433 | __ai int32x4_t vmovl_s16(int16x4_t __a) { |
| 434 | return (int32x4_t)__builtin_neon_vmovl_v((int8x8_t)__a, 34); } |
| 435 | __ai int64x2_t vmovl_s32(int32x2_t __a) { |
| 436 | return (int64x2_t)__builtin_neon_vmovl_v((int8x8_t)__a, 35); } |
| 437 | __ai uint16x8_t vmovl_u8(uint8x8_t __a) { |
| 438 | return (uint16x8_t)__builtin_neon_vmovl_v((int8x8_t)__a, 49); } |
| 439 | __ai uint32x4_t vmovl_u16(uint16x4_t __a) { |
| 440 | return (uint32x4_t)__builtin_neon_vmovl_v((int8x8_t)__a, 50); } |
| 441 | __ai uint64x2_t vmovl_u32(uint32x2_t __a) { |
| 442 | return (uint64x2_t)__builtin_neon_vmovl_v((int8x8_t)__a, 51); } |
| 443 | |
| 444 | __ai int16x8_t vmull_s8(int8x8_t __a, int8x8_t __b) { |
| 445 | return (int16x8_t)__builtin_neon_vmull_v(__a, __b, 33); } |
| 446 | __ai int32x4_t vmull_s16(int16x4_t __a, int16x4_t __b) { |
| 447 | return (int32x4_t)__builtin_neon_vmull_v((int8x8_t)__a, (int8x8_t)__b, 34); } |
| 448 | __ai int64x2_t vmull_s32(int32x2_t __a, int32x2_t __b) { |
| 449 | return (int64x2_t)__builtin_neon_vmull_v((int8x8_t)__a, (int8x8_t)__b, 35); } |
| 450 | __ai uint16x8_t vmull_u8(uint8x8_t __a, uint8x8_t __b) { |
| 451 | return (uint16x8_t)__builtin_neon_vmull_v((int8x8_t)__a, (int8x8_t)__b, 49); } |
| 452 | __ai uint32x4_t vmull_u16(uint16x4_t __a, uint16x4_t __b) { |
| 453 | return (uint32x4_t)__builtin_neon_vmull_v((int8x8_t)__a, (int8x8_t)__b, 50); } |
| 454 | __ai uint64x2_t vmull_u32(uint32x2_t __a, uint32x2_t __b) { |
| 455 | return (uint64x2_t)__builtin_neon_vmull_v((int8x8_t)__a, (int8x8_t)__b, 51); } |
| 456 | __ai poly16x8_t vmull_p8(poly8x8_t __a, poly8x8_t __b) { |
| 457 | return (poly16x8_t)__builtin_neon_vmull_v((int8x8_t)__a, (int8x8_t)__b, 37); } |
| 458 | |
| 459 | __ai int8x8_t vabd_s8(int8x8_t __a, int8x8_t __b) { |
| 460 | return (int8x8_t)__builtin_neon_vabd_v(__a, __b, 0); } |
| 461 | __ai int16x4_t vabd_s16(int16x4_t __a, int16x4_t __b) { |
| 462 | return (int16x4_t)__builtin_neon_vabd_v((int8x8_t)__a, (int8x8_t)__b, 1); } |
| 463 | __ai int32x2_t vabd_s32(int32x2_t __a, int32x2_t __b) { |
| 464 | return (int32x2_t)__builtin_neon_vabd_v((int8x8_t)__a, (int8x8_t)__b, 2); } |
| 465 | __ai uint8x8_t vabd_u8(uint8x8_t __a, uint8x8_t __b) { |
| 466 | return (uint8x8_t)__builtin_neon_vabd_v((int8x8_t)__a, (int8x8_t)__b, 16); } |
| 467 | __ai uint16x4_t vabd_u16(uint16x4_t __a, uint16x4_t __b) { |
| 468 | return (uint16x4_t)__builtin_neon_vabd_v((int8x8_t)__a, (int8x8_t)__b, 17); } |
| 469 | __ai uint32x2_t vabd_u32(uint32x2_t __a, uint32x2_t __b) { |
| 470 | return (uint32x2_t)__builtin_neon_vabd_v((int8x8_t)__a, (int8x8_t)__b, 18); } |
| 471 | __ai float32x2_t vabd_f32(float32x2_t __a, float32x2_t __b) { |
| 472 | return (float32x2_t)__builtin_neon_vabd_v((int8x8_t)__a, (int8x8_t)__b, 8); } |
| 473 | __ai int8x16_t vabdq_s8(int8x16_t __a, int8x16_t __b) { |
| 474 | return (int8x16_t)__builtin_neon_vabdq_v(__a, __b, 32); } |
| 475 | __ai int16x8_t vabdq_s16(int16x8_t __a, int16x8_t __b) { |
| 476 | return (int16x8_t)__builtin_neon_vabdq_v((int8x16_t)__a, (int8x16_t)__b, 33); } |
| 477 | __ai int32x4_t vabdq_s32(int32x4_t __a, int32x4_t __b) { |
| 478 | return (int32x4_t)__builtin_neon_vabdq_v((int8x16_t)__a, (int8x16_t)__b, 34); } |
| 479 | __ai uint8x16_t vabdq_u8(uint8x16_t __a, uint8x16_t __b) { |
| 480 | return (uint8x16_t)__builtin_neon_vabdq_v((int8x16_t)__a, (int8x16_t)__b, 48); } |
| 481 | __ai uint16x8_t vabdq_u16(uint16x8_t __a, uint16x8_t __b) { |
| 482 | return (uint16x8_t)__builtin_neon_vabdq_v((int8x16_t)__a, (int8x16_t)__b, 49); } |
| 483 | __ai uint32x4_t vabdq_u32(uint32x4_t __a, uint32x4_t __b) { |
| 484 | return (uint32x4_t)__builtin_neon_vabdq_v((int8x16_t)__a, (int8x16_t)__b, 50); } |
| 485 | __ai float32x4_t vabdq_f32(float32x4_t __a, float32x4_t __b) { |
| 486 | return (float32x4_t)__builtin_neon_vabdq_v((int8x16_t)__a, (int8x16_t)__b, 40); } |
| 487 | |
| 488 | __ai int16x8_t vabdl_s8(int8x8_t __a, int8x8_t __b) { |
| 489 | return (int16x8_t)vmovl_u8((uint8x8_t)vabd_s8(__a, __b)); } |
| 490 | __ai int32x4_t vabdl_s16(int16x4_t __a, int16x4_t __b) { |
| 491 | return (int32x4_t)vmovl_u16((uint16x4_t)vabd_s16(__a, __b)); } |
| 492 | __ai int64x2_t vabdl_s32(int32x2_t __a, int32x2_t __b) { |
| 493 | return (int64x2_t)vmovl_u32((uint32x2_t)vabd_s32(__a, __b)); } |
| 494 | __ai uint16x8_t vabdl_u8(uint8x8_t __a, uint8x8_t __b) { |
| 495 | return vmovl_u8(vabd_u8(__a, __b)); } |
| 496 | __ai uint32x4_t vabdl_u16(uint16x4_t __a, uint16x4_t __b) { |
| 497 | return vmovl_u16(vabd_u16(__a, __b)); } |
| 498 | __ai uint64x2_t vabdl_u32(uint32x2_t __a, uint32x2_t __b) { |
| 499 | return vmovl_u32(vabd_u32(__a, __b)); } |
| 500 | |
| 501 | __ai int8x8_t vaba_s8(int8x8_t __a, int8x8_t __b, int8x8_t __c) { |
| 502 | return __a + vabd_s8(__b, __c); } |
| 503 | __ai int16x4_t vaba_s16(int16x4_t __a, int16x4_t __b, int16x4_t __c) { |
| 504 | return __a + vabd_s16(__b, __c); } |
| 505 | __ai int32x2_t vaba_s32(int32x2_t __a, int32x2_t __b, int32x2_t __c) { |
| 506 | return __a + vabd_s32(__b, __c); } |
| 507 | __ai uint8x8_t vaba_u8(uint8x8_t __a, uint8x8_t __b, uint8x8_t __c) { |
| 508 | return __a + vabd_u8(__b, __c); } |
| 509 | __ai uint16x4_t vaba_u16(uint16x4_t __a, uint16x4_t __b, uint16x4_t __c) { |
| 510 | return __a + vabd_u16(__b, __c); } |
| 511 | __ai uint32x2_t vaba_u32(uint32x2_t __a, uint32x2_t __b, uint32x2_t __c) { |
| 512 | return __a + vabd_u32(__b, __c); } |
| 513 | __ai int8x16_t vabaq_s8(int8x16_t __a, int8x16_t __b, int8x16_t __c) { |
| 514 | return __a + vabdq_s8(__b, __c); } |
| 515 | __ai int16x8_t vabaq_s16(int16x8_t __a, int16x8_t __b, int16x8_t __c) { |
| 516 | return __a + vabdq_s16(__b, __c); } |
| 517 | __ai int32x4_t vabaq_s32(int32x4_t __a, int32x4_t __b, int32x4_t __c) { |
| 518 | return __a + vabdq_s32(__b, __c); } |
| 519 | __ai uint8x16_t vabaq_u8(uint8x16_t __a, uint8x16_t __b, uint8x16_t __c) { |
| 520 | return __a + vabdq_u8(__b, __c); } |
| 521 | __ai uint16x8_t vabaq_u16(uint16x8_t __a, uint16x8_t __b, uint16x8_t __c) { |
| 522 | return __a + vabdq_u16(__b, __c); } |
| 523 | __ai uint32x4_t vabaq_u32(uint32x4_t __a, uint32x4_t __b, uint32x4_t __c) { |
| 524 | return __a + vabdq_u32(__b, __c); } |
| 525 | |
| 526 | __ai int16x8_t vabal_s8(int16x8_t __a, int8x8_t __b, int8x8_t __c) { |
| 527 | return __a + vabdl_s8(__b, __c); } |
| 528 | __ai int32x4_t vabal_s16(int32x4_t __a, int16x4_t __b, int16x4_t __c) { |
| 529 | return __a + vabdl_s16(__b, __c); } |
| 530 | __ai int64x2_t vabal_s32(int64x2_t __a, int32x2_t __b, int32x2_t __c) { |
| 531 | return __a + vabdl_s32(__b, __c); } |
| 532 | __ai uint16x8_t vabal_u8(uint16x8_t __a, uint8x8_t __b, uint8x8_t __c) { |
| 533 | return __a + vabdl_u8(__b, __c); } |
| 534 | __ai uint32x4_t vabal_u16(uint32x4_t __a, uint16x4_t __b, uint16x4_t __c) { |
| 535 | return __a + vabdl_u16(__b, __c); } |
| 536 | __ai uint64x2_t vabal_u32(uint64x2_t __a, uint32x2_t __b, uint32x2_t __c) { |
| 537 | return __a + vabdl_u32(__b, __c); } |
| 538 | |
| 539 | |
| 540 | __ai int8x8_t vabs_s8(int8x8_t __a) { |
| 541 | return (int8x8_t)__builtin_neon_vabs_v(__a, 0); } |
| 542 | __ai int16x4_t vabs_s16(int16x4_t __a) { |
| 543 | return (int16x4_t)__builtin_neon_vabs_v((int8x8_t)__a, 1); } |
| 544 | __ai int32x2_t vabs_s32(int32x2_t __a) { |
| 545 | return (int32x2_t)__builtin_neon_vabs_v((int8x8_t)__a, 2); } |
| 546 | __ai float32x2_t vabs_f32(float32x2_t __a) { |
| 547 | return (float32x2_t)__builtin_neon_vabs_v((int8x8_t)__a, 8); } |
| 548 | __ai int8x16_t vabsq_s8(int8x16_t __a) { |
| 549 | return (int8x16_t)__builtin_neon_vabsq_v(__a, 32); } |
| 550 | __ai int16x8_t vabsq_s16(int16x8_t __a) { |
| 551 | return (int16x8_t)__builtin_neon_vabsq_v((int8x16_t)__a, 33); } |
| 552 | __ai int32x4_t vabsq_s32(int32x4_t __a) { |
| 553 | return (int32x4_t)__builtin_neon_vabsq_v((int8x16_t)__a, 34); } |
| 554 | __ai float32x4_t vabsq_f32(float32x4_t __a) { |
| 555 | return (float32x4_t)__builtin_neon_vabsq_v((int8x16_t)__a, 40); } |
| 556 | |
| 557 | __ai int8x8_t vadd_s8(int8x8_t __a, int8x8_t __b) { |
| 558 | return __a + __b; } |
| 559 | __ai int16x4_t vadd_s16(int16x4_t __a, int16x4_t __b) { |
| 560 | return __a + __b; } |
| 561 | __ai int32x2_t vadd_s32(int32x2_t __a, int32x2_t __b) { |
| 562 | return __a + __b; } |
| 563 | __ai int64x1_t vadd_s64(int64x1_t __a, int64x1_t __b) { |
| 564 | return __a + __b; } |
| 565 | __ai float32x2_t vadd_f32(float32x2_t __a, float32x2_t __b) { |
| 566 | return __a + __b; } |
| 567 | __ai uint8x8_t vadd_u8(uint8x8_t __a, uint8x8_t __b) { |
| 568 | return __a + __b; } |
| 569 | __ai uint16x4_t vadd_u16(uint16x4_t __a, uint16x4_t __b) { |
| 570 | return __a + __b; } |
| 571 | __ai uint32x2_t vadd_u32(uint32x2_t __a, uint32x2_t __b) { |
| 572 | return __a + __b; } |
| 573 | __ai uint64x1_t vadd_u64(uint64x1_t __a, uint64x1_t __b) { |
| 574 | return __a + __b; } |
| 575 | __ai int8x16_t vaddq_s8(int8x16_t __a, int8x16_t __b) { |
| 576 | return __a + __b; } |
| 577 | __ai int16x8_t vaddq_s16(int16x8_t __a, int16x8_t __b) { |
| 578 | return __a + __b; } |
| 579 | __ai int32x4_t vaddq_s32(int32x4_t __a, int32x4_t __b) { |
| 580 | return __a + __b; } |
| 581 | __ai int64x2_t vaddq_s64(int64x2_t __a, int64x2_t __b) { |
| 582 | return __a + __b; } |
| 583 | __ai float32x4_t vaddq_f32(float32x4_t __a, float32x4_t __b) { |
| 584 | return __a + __b; } |
| 585 | __ai uint8x16_t vaddq_u8(uint8x16_t __a, uint8x16_t __b) { |
| 586 | return __a + __b; } |
| 587 | __ai uint16x8_t vaddq_u16(uint16x8_t __a, uint16x8_t __b) { |
| 588 | return __a + __b; } |
| 589 | __ai uint32x4_t vaddq_u32(uint32x4_t __a, uint32x4_t __b) { |
| 590 | return __a + __b; } |
| 591 | __ai uint64x2_t vaddq_u64(uint64x2_t __a, uint64x2_t __b) { |
| 592 | return __a + __b; } |
| 593 | |
| 594 | __ai int8x8_t vaddhn_s16(int16x8_t __a, int16x8_t __b) { |
| 595 | return (int8x8_t)__builtin_neon_vaddhn_v((int8x16_t)__a, (int8x16_t)__b, 0); } |
| 596 | __ai int16x4_t vaddhn_s32(int32x4_t __a, int32x4_t __b) { |
| 597 | return (int16x4_t)__builtin_neon_vaddhn_v((int8x16_t)__a, (int8x16_t)__b, 1); } |
| 598 | __ai int32x2_t vaddhn_s64(int64x2_t __a, int64x2_t __b) { |
| 599 | return (int32x2_t)__builtin_neon_vaddhn_v((int8x16_t)__a, (int8x16_t)__b, 2); } |
| 600 | __ai uint8x8_t vaddhn_u16(uint16x8_t __a, uint16x8_t __b) { |
| 601 | return (uint8x8_t)__builtin_neon_vaddhn_v((int8x16_t)__a, (int8x16_t)__b, 16); } |
| 602 | __ai uint16x4_t vaddhn_u32(uint32x4_t __a, uint32x4_t __b) { |
| 603 | return (uint16x4_t)__builtin_neon_vaddhn_v((int8x16_t)__a, (int8x16_t)__b, 17); } |
| 604 | __ai uint32x2_t vaddhn_u64(uint64x2_t __a, uint64x2_t __b) { |
| 605 | return (uint32x2_t)__builtin_neon_vaddhn_v((int8x16_t)__a, (int8x16_t)__b, 18); } |
| 606 | |
| 607 | __ai int16x8_t vaddl_s8(int8x8_t __a, int8x8_t __b) { |
| 608 | return vmovl_s8(__a) + vmovl_s8(__b); } |
| 609 | __ai int32x4_t vaddl_s16(int16x4_t __a, int16x4_t __b) { |
| 610 | return vmovl_s16(__a) + vmovl_s16(__b); } |
| 611 | __ai int64x2_t vaddl_s32(int32x2_t __a, int32x2_t __b) { |
| 612 | return vmovl_s32(__a) + vmovl_s32(__b); } |
| 613 | __ai uint16x8_t vaddl_u8(uint8x8_t __a, uint8x8_t __b) { |
| 614 | return vmovl_u8(__a) + vmovl_u8(__b); } |
| 615 | __ai uint32x4_t vaddl_u16(uint16x4_t __a, uint16x4_t __b) { |
| 616 | return vmovl_u16(__a) + vmovl_u16(__b); } |
| 617 | __ai uint64x2_t vaddl_u32(uint32x2_t __a, uint32x2_t __b) { |
| 618 | return vmovl_u32(__a) + vmovl_u32(__b); } |
| 619 | |
| 620 | __ai int16x8_t vaddw_s8(int16x8_t __a, int8x8_t __b) { |
| 621 | return __a + vmovl_s8(__b); } |
| 622 | __ai int32x4_t vaddw_s16(int32x4_t __a, int16x4_t __b) { |
| 623 | return __a + vmovl_s16(__b); } |
| 624 | __ai int64x2_t vaddw_s32(int64x2_t __a, int32x2_t __b) { |
| 625 | return __a + vmovl_s32(__b); } |
| 626 | __ai uint16x8_t vaddw_u8(uint16x8_t __a, uint8x8_t __b) { |
| 627 | return __a + vmovl_u8(__b); } |
| 628 | __ai uint32x4_t vaddw_u16(uint32x4_t __a, uint16x4_t __b) { |
| 629 | return __a + vmovl_u16(__b); } |
| 630 | __ai uint64x2_t vaddw_u32(uint64x2_t __a, uint32x2_t __b) { |
| 631 | return __a + vmovl_u32(__b); } |
| 632 | |
| 633 | __ai int8x8_t vand_s8(int8x8_t __a, int8x8_t __b) { |
| 634 | return __a & __b; } |
| 635 | __ai int16x4_t vand_s16(int16x4_t __a, int16x4_t __b) { |
| 636 | return __a & __b; } |
| 637 | __ai int32x2_t vand_s32(int32x2_t __a, int32x2_t __b) { |
| 638 | return __a & __b; } |
| 639 | __ai int64x1_t vand_s64(int64x1_t __a, int64x1_t __b) { |
| 640 | return __a & __b; } |
| 641 | __ai uint8x8_t vand_u8(uint8x8_t __a, uint8x8_t __b) { |
| 642 | return __a & __b; } |
| 643 | __ai uint16x4_t vand_u16(uint16x4_t __a, uint16x4_t __b) { |
| 644 | return __a & __b; } |
| 645 | __ai uint32x2_t vand_u32(uint32x2_t __a, uint32x2_t __b) { |
| 646 | return __a & __b; } |
| 647 | __ai uint64x1_t vand_u64(uint64x1_t __a, uint64x1_t __b) { |
| 648 | return __a & __b; } |
| 649 | __ai int8x16_t vandq_s8(int8x16_t __a, int8x16_t __b) { |
| 650 | return __a & __b; } |
| 651 | __ai int16x8_t vandq_s16(int16x8_t __a, int16x8_t __b) { |
| 652 | return __a & __b; } |
| 653 | __ai int32x4_t vandq_s32(int32x4_t __a, int32x4_t __b) { |
| 654 | return __a & __b; } |
| 655 | __ai int64x2_t vandq_s64(int64x2_t __a, int64x2_t __b) { |
| 656 | return __a & __b; } |
| 657 | __ai uint8x16_t vandq_u8(uint8x16_t __a, uint8x16_t __b) { |
| 658 | return __a & __b; } |
| 659 | __ai uint16x8_t vandq_u16(uint16x8_t __a, uint16x8_t __b) { |
| 660 | return __a & __b; } |
| 661 | __ai uint32x4_t vandq_u32(uint32x4_t __a, uint32x4_t __b) { |
| 662 | return __a & __b; } |
| 663 | __ai uint64x2_t vandq_u64(uint64x2_t __a, uint64x2_t __b) { |
| 664 | return __a & __b; } |
| 665 | |
| 666 | __ai int8x8_t vbic_s8(int8x8_t __a, int8x8_t __b) { |
| 667 | return __a & ~__b; } |
| 668 | __ai int16x4_t vbic_s16(int16x4_t __a, int16x4_t __b) { |
| 669 | return __a & ~__b; } |
| 670 | __ai int32x2_t vbic_s32(int32x2_t __a, int32x2_t __b) { |
| 671 | return __a & ~__b; } |
| 672 | __ai int64x1_t vbic_s64(int64x1_t __a, int64x1_t __b) { |
| 673 | return __a & ~__b; } |
| 674 | __ai uint8x8_t vbic_u8(uint8x8_t __a, uint8x8_t __b) { |
| 675 | return __a & ~__b; } |
| 676 | __ai uint16x4_t vbic_u16(uint16x4_t __a, uint16x4_t __b) { |
| 677 | return __a & ~__b; } |
| 678 | __ai uint32x2_t vbic_u32(uint32x2_t __a, uint32x2_t __b) { |
| 679 | return __a & ~__b; } |
| 680 | __ai uint64x1_t vbic_u64(uint64x1_t __a, uint64x1_t __b) { |
| 681 | return __a & ~__b; } |
| 682 | __ai int8x16_t vbicq_s8(int8x16_t __a, int8x16_t __b) { |
| 683 | return __a & ~__b; } |
| 684 | __ai int16x8_t vbicq_s16(int16x8_t __a, int16x8_t __b) { |
| 685 | return __a & ~__b; } |
| 686 | __ai int32x4_t vbicq_s32(int32x4_t __a, int32x4_t __b) { |
| 687 | return __a & ~__b; } |
| 688 | __ai int64x2_t vbicq_s64(int64x2_t __a, int64x2_t __b) { |
| 689 | return __a & ~__b; } |
| 690 | __ai uint8x16_t vbicq_u8(uint8x16_t __a, uint8x16_t __b) { |
| 691 | return __a & ~__b; } |
| 692 | __ai uint16x8_t vbicq_u16(uint16x8_t __a, uint16x8_t __b) { |
| 693 | return __a & ~__b; } |
| 694 | __ai uint32x4_t vbicq_u32(uint32x4_t __a, uint32x4_t __b) { |
| 695 | return __a & ~__b; } |
| 696 | __ai uint64x2_t vbicq_u64(uint64x2_t __a, uint64x2_t __b) { |
| 697 | return __a & ~__b; } |
| 698 | |
| 699 | __ai int8x8_t vbsl_s8(uint8x8_t __a, int8x8_t __b, int8x8_t __c) { |
| 700 | return (int8x8_t)__builtin_neon_vbsl_v((int8x8_t)__a, __b, __c, 0); } |
| 701 | __ai int16x4_t vbsl_s16(uint16x4_t __a, int16x4_t __b, int16x4_t __c) { |
| 702 | return (int16x4_t)__builtin_neon_vbsl_v((int8x8_t)__a, (int8x8_t)__b, (int8x8_t)__c, 1); } |
| 703 | __ai int32x2_t vbsl_s32(uint32x2_t __a, int32x2_t __b, int32x2_t __c) { |
| 704 | return (int32x2_t)__builtin_neon_vbsl_v((int8x8_t)__a, (int8x8_t)__b, (int8x8_t)__c, 2); } |
| 705 | __ai int64x1_t vbsl_s64(uint64x1_t __a, int64x1_t __b, int64x1_t __c) { |
| 706 | return (int64x1_t)__builtin_neon_vbsl_v((int8x8_t)__a, (int8x8_t)__b, (int8x8_t)__c, 3); } |
| 707 | __ai uint8x8_t vbsl_u8(uint8x8_t __a, uint8x8_t __b, uint8x8_t __c) { |
| 708 | return (uint8x8_t)__builtin_neon_vbsl_v((int8x8_t)__a, (int8x8_t)__b, (int8x8_t)__c, 16); } |
| 709 | __ai uint16x4_t vbsl_u16(uint16x4_t __a, uint16x4_t __b, uint16x4_t __c) { |
| 710 | return (uint16x4_t)__builtin_neon_vbsl_v((int8x8_t)__a, (int8x8_t)__b, (int8x8_t)__c, 17); } |
| 711 | __ai uint32x2_t vbsl_u32(uint32x2_t __a, uint32x2_t __b, uint32x2_t __c) { |
| 712 | return (uint32x2_t)__builtin_neon_vbsl_v((int8x8_t)__a, (int8x8_t)__b, (int8x8_t)__c, 18); } |
| 713 | __ai uint64x1_t vbsl_u64(uint64x1_t __a, uint64x1_t __b, uint64x1_t __c) { |
| 714 | return (uint64x1_t)__builtin_neon_vbsl_v((int8x8_t)__a, (int8x8_t)__b, (int8x8_t)__c, 19); } |
| 715 | __ai float32x2_t vbsl_f32(uint32x2_t __a, float32x2_t __b, float32x2_t __c) { |
| 716 | return (float32x2_t)__builtin_neon_vbsl_v((int8x8_t)__a, (int8x8_t)__b, (int8x8_t)__c, 8); } |
| 717 | __ai poly8x8_t vbsl_p8(uint8x8_t __a, poly8x8_t __b, poly8x8_t __c) { |
| 718 | return (poly8x8_t)__builtin_neon_vbsl_v((int8x8_t)__a, (int8x8_t)__b, (int8x8_t)__c, 4); } |
| 719 | __ai poly16x4_t vbsl_p16(uint16x4_t __a, poly16x4_t __b, poly16x4_t __c) { |
| 720 | return (poly16x4_t)__builtin_neon_vbsl_v((int8x8_t)__a, (int8x8_t)__b, (int8x8_t)__c, 5); } |
| 721 | __ai int8x16_t vbslq_s8(uint8x16_t __a, int8x16_t __b, int8x16_t __c) { |
| 722 | return (int8x16_t)__builtin_neon_vbslq_v((int8x16_t)__a, __b, __c, 32); } |
| 723 | __ai int16x8_t vbslq_s16(uint16x8_t __a, int16x8_t __b, int16x8_t __c) { |
| 724 | return (int16x8_t)__builtin_neon_vbslq_v((int8x16_t)__a, (int8x16_t)__b, (int8x16_t)__c, 33); } |
| 725 | __ai int32x4_t vbslq_s32(uint32x4_t __a, int32x4_t __b, int32x4_t __c) { |
| 726 | return (int32x4_t)__builtin_neon_vbslq_v((int8x16_t)__a, (int8x16_t)__b, (int8x16_t)__c, 34); } |
| 727 | __ai int64x2_t vbslq_s64(uint64x2_t __a, int64x2_t __b, int64x2_t __c) { |
| 728 | return (int64x2_t)__builtin_neon_vbslq_v((int8x16_t)__a, (int8x16_t)__b, (int8x16_t)__c, 35); } |
| 729 | __ai uint8x16_t vbslq_u8(uint8x16_t __a, uint8x16_t __b, uint8x16_t __c) { |
| 730 | return (uint8x16_t)__builtin_neon_vbslq_v((int8x16_t)__a, (int8x16_t)__b, (int8x16_t)__c, 48); } |
| 731 | __ai uint16x8_t vbslq_u16(uint16x8_t __a, uint16x8_t __b, uint16x8_t __c) { |
| 732 | return (uint16x8_t)__builtin_neon_vbslq_v((int8x16_t)__a, (int8x16_t)__b, (int8x16_t)__c, 49); } |
| 733 | __ai uint32x4_t vbslq_u32(uint32x4_t __a, uint32x4_t __b, uint32x4_t __c) { |
| 734 | return (uint32x4_t)__builtin_neon_vbslq_v((int8x16_t)__a, (int8x16_t)__b, (int8x16_t)__c, 50); } |
| 735 | __ai uint64x2_t vbslq_u64(uint64x2_t __a, uint64x2_t __b, uint64x2_t __c) { |
| 736 | return (uint64x2_t)__builtin_neon_vbslq_v((int8x16_t)__a, (int8x16_t)__b, (int8x16_t)__c, 51); } |
| 737 | __ai float32x4_t vbslq_f32(uint32x4_t __a, float32x4_t __b, float32x4_t __c) { |
| 738 | return (float32x4_t)__builtin_neon_vbslq_v((int8x16_t)__a, (int8x16_t)__b, (int8x16_t)__c, 40); } |
| 739 | __ai poly8x16_t vbslq_p8(uint8x16_t __a, poly8x16_t __b, poly8x16_t __c) { |
| 740 | return (poly8x16_t)__builtin_neon_vbslq_v((int8x16_t)__a, (int8x16_t)__b, (int8x16_t)__c, 36); } |
| 741 | __ai poly16x8_t vbslq_p16(uint16x8_t __a, poly16x8_t __b, poly16x8_t __c) { |
| 742 | return (poly16x8_t)__builtin_neon_vbslq_v((int8x16_t)__a, (int8x16_t)__b, (int8x16_t)__c, 37); } |
| 743 | |
| 744 | __ai uint32x2_t vcage_f32(float32x2_t __a, float32x2_t __b) { |
| 745 | return (uint32x2_t)__builtin_neon_vcage_v((int8x8_t)__a, (int8x8_t)__b, 18); } |
| 746 | __ai uint32x4_t vcageq_f32(float32x4_t __a, float32x4_t __b) { |
| 747 | return (uint32x4_t)__builtin_neon_vcageq_v((int8x16_t)__a, (int8x16_t)__b, 50); } |
| 748 | |
| 749 | __ai uint32x2_t vcagt_f32(float32x2_t __a, float32x2_t __b) { |
| 750 | return (uint32x2_t)__builtin_neon_vcagt_v((int8x8_t)__a, (int8x8_t)__b, 18); } |
| 751 | __ai uint32x4_t vcagtq_f32(float32x4_t __a, float32x4_t __b) { |
| 752 | return (uint32x4_t)__builtin_neon_vcagtq_v((int8x16_t)__a, (int8x16_t)__b, 50); } |
| 753 | |
| 754 | __ai uint32x2_t vcale_f32(float32x2_t __a, float32x2_t __b) { |
| 755 | return (uint32x2_t)__builtin_neon_vcale_v((int8x8_t)__a, (int8x8_t)__b, 18); } |
| 756 | __ai uint32x4_t vcaleq_f32(float32x4_t __a, float32x4_t __b) { |
| 757 | return (uint32x4_t)__builtin_neon_vcaleq_v((int8x16_t)__a, (int8x16_t)__b, 50); } |
| 758 | |
| 759 | __ai uint32x2_t vcalt_f32(float32x2_t __a, float32x2_t __b) { |
| 760 | return (uint32x2_t)__builtin_neon_vcalt_v((int8x8_t)__a, (int8x8_t)__b, 18); } |
| 761 | __ai uint32x4_t vcaltq_f32(float32x4_t __a, float32x4_t __b) { |
| 762 | return (uint32x4_t)__builtin_neon_vcaltq_v((int8x16_t)__a, (int8x16_t)__b, 50); } |
| 763 | |
| 764 | __ai uint8x8_t vceq_s8(int8x8_t __a, int8x8_t __b) { |
| 765 | return (uint8x8_t)(__a == __b); } |
| 766 | __ai uint16x4_t vceq_s16(int16x4_t __a, int16x4_t __b) { |
| 767 | return (uint16x4_t)(__a == __b); } |
| 768 | __ai uint32x2_t vceq_s32(int32x2_t __a, int32x2_t __b) { |
| 769 | return (uint32x2_t)(__a == __b); } |
| 770 | __ai uint32x2_t vceq_f32(float32x2_t __a, float32x2_t __b) { |
| 771 | return (uint32x2_t)(__a == __b); } |
| 772 | __ai uint8x8_t vceq_u8(uint8x8_t __a, uint8x8_t __b) { |
| 773 | return (uint8x8_t)(__a == __b); } |
| 774 | __ai uint16x4_t vceq_u16(uint16x4_t __a, uint16x4_t __b) { |
| 775 | return (uint16x4_t)(__a == __b); } |
| 776 | __ai uint32x2_t vceq_u32(uint32x2_t __a, uint32x2_t __b) { |
| 777 | return (uint32x2_t)(__a == __b); } |
| 778 | __ai uint8x8_t vceq_p8(poly8x8_t __a, poly8x8_t __b) { |
| 779 | return (uint8x8_t)(__a == __b); } |
| 780 | __ai uint8x16_t vceqq_s8(int8x16_t __a, int8x16_t __b) { |
| 781 | return (uint8x16_t)(__a == __b); } |
| 782 | __ai uint16x8_t vceqq_s16(int16x8_t __a, int16x8_t __b) { |
| 783 | return (uint16x8_t)(__a == __b); } |
| 784 | __ai uint32x4_t vceqq_s32(int32x4_t __a, int32x4_t __b) { |
| 785 | return (uint32x4_t)(__a == __b); } |
| 786 | __ai uint32x4_t vceqq_f32(float32x4_t __a, float32x4_t __b) { |
| 787 | return (uint32x4_t)(__a == __b); } |
| 788 | __ai uint8x16_t vceqq_u8(uint8x16_t __a, uint8x16_t __b) { |
| 789 | return (uint8x16_t)(__a == __b); } |
| 790 | __ai uint16x8_t vceqq_u16(uint16x8_t __a, uint16x8_t __b) { |
| 791 | return (uint16x8_t)(__a == __b); } |
| 792 | __ai uint32x4_t vceqq_u32(uint32x4_t __a, uint32x4_t __b) { |
| 793 | return (uint32x4_t)(__a == __b); } |
| 794 | __ai uint8x16_t vceqq_p8(poly8x16_t __a, poly8x16_t __b) { |
| 795 | return (uint8x16_t)(__a == __b); } |
| 796 | |
| 797 | __ai uint8x8_t vcge_s8(int8x8_t __a, int8x8_t __b) { |
| 798 | return (uint8x8_t)(__a >= __b); } |
| 799 | __ai uint16x4_t vcge_s16(int16x4_t __a, int16x4_t __b) { |
| 800 | return (uint16x4_t)(__a >= __b); } |
| 801 | __ai uint32x2_t vcge_s32(int32x2_t __a, int32x2_t __b) { |
| 802 | return (uint32x2_t)(__a >= __b); } |
| 803 | __ai uint32x2_t vcge_f32(float32x2_t __a, float32x2_t __b) { |
| 804 | return (uint32x2_t)(__a >= __b); } |
| 805 | __ai uint8x8_t vcge_u8(uint8x8_t __a, uint8x8_t __b) { |
| 806 | return (uint8x8_t)(__a >= __b); } |
| 807 | __ai uint16x4_t vcge_u16(uint16x4_t __a, uint16x4_t __b) { |
| 808 | return (uint16x4_t)(__a >= __b); } |
| 809 | __ai uint32x2_t vcge_u32(uint32x2_t __a, uint32x2_t __b) { |
| 810 | return (uint32x2_t)(__a >= __b); } |
| 811 | __ai uint8x16_t vcgeq_s8(int8x16_t __a, int8x16_t __b) { |
| 812 | return (uint8x16_t)(__a >= __b); } |
| 813 | __ai uint16x8_t vcgeq_s16(int16x8_t __a, int16x8_t __b) { |
| 814 | return (uint16x8_t)(__a >= __b); } |
| 815 | __ai uint32x4_t vcgeq_s32(int32x4_t __a, int32x4_t __b) { |
| 816 | return (uint32x4_t)(__a >= __b); } |
| 817 | __ai uint32x4_t vcgeq_f32(float32x4_t __a, float32x4_t __b) { |
| 818 | return (uint32x4_t)(__a >= __b); } |
| 819 | __ai uint8x16_t vcgeq_u8(uint8x16_t __a, uint8x16_t __b) { |
| 820 | return (uint8x16_t)(__a >= __b); } |
| 821 | __ai uint16x8_t vcgeq_u16(uint16x8_t __a, uint16x8_t __b) { |
| 822 | return (uint16x8_t)(__a >= __b); } |
| 823 | __ai uint32x4_t vcgeq_u32(uint32x4_t __a, uint32x4_t __b) { |
| 824 | return (uint32x4_t)(__a >= __b); } |
| 825 | |
| 826 | __ai uint8x8_t vcgt_s8(int8x8_t __a, int8x8_t __b) { |
| 827 | return (uint8x8_t)(__a > __b); } |
| 828 | __ai uint16x4_t vcgt_s16(int16x4_t __a, int16x4_t __b) { |
| 829 | return (uint16x4_t)(__a > __b); } |
| 830 | __ai uint32x2_t vcgt_s32(int32x2_t __a, int32x2_t __b) { |
| 831 | return (uint32x2_t)(__a > __b); } |
| 832 | __ai uint32x2_t vcgt_f32(float32x2_t __a, float32x2_t __b) { |
| 833 | return (uint32x2_t)(__a > __b); } |
| 834 | __ai uint8x8_t vcgt_u8(uint8x8_t __a, uint8x8_t __b) { |
| 835 | return (uint8x8_t)(__a > __b); } |
| 836 | __ai uint16x4_t vcgt_u16(uint16x4_t __a, uint16x4_t __b) { |
| 837 | return (uint16x4_t)(__a > __b); } |
| 838 | __ai uint32x2_t vcgt_u32(uint32x2_t __a, uint32x2_t __b) { |
| 839 | return (uint32x2_t)(__a > __b); } |
| 840 | __ai uint8x16_t vcgtq_s8(int8x16_t __a, int8x16_t __b) { |
| 841 | return (uint8x16_t)(__a > __b); } |
| 842 | __ai uint16x8_t vcgtq_s16(int16x8_t __a, int16x8_t __b) { |
| 843 | return (uint16x8_t)(__a > __b); } |
| 844 | __ai uint32x4_t vcgtq_s32(int32x4_t __a, int32x4_t __b) { |
| 845 | return (uint32x4_t)(__a > __b); } |
| 846 | __ai uint32x4_t vcgtq_f32(float32x4_t __a, float32x4_t __b) { |
| 847 | return (uint32x4_t)(__a > __b); } |
| 848 | __ai uint8x16_t vcgtq_u8(uint8x16_t __a, uint8x16_t __b) { |
| 849 | return (uint8x16_t)(__a > __b); } |
| 850 | __ai uint16x8_t vcgtq_u16(uint16x8_t __a, uint16x8_t __b) { |
| 851 | return (uint16x8_t)(__a > __b); } |
| 852 | __ai uint32x4_t vcgtq_u32(uint32x4_t __a, uint32x4_t __b) { |
| 853 | return (uint32x4_t)(__a > __b); } |
| 854 | |
| 855 | __ai uint8x8_t vcle_s8(int8x8_t __a, int8x8_t __b) { |
| 856 | return (uint8x8_t)(__a <= __b); } |
| 857 | __ai uint16x4_t vcle_s16(int16x4_t __a, int16x4_t __b) { |
| 858 | return (uint16x4_t)(__a <= __b); } |
| 859 | __ai uint32x2_t vcle_s32(int32x2_t __a, int32x2_t __b) { |
| 860 | return (uint32x2_t)(__a <= __b); } |
| 861 | __ai uint32x2_t vcle_f32(float32x2_t __a, float32x2_t __b) { |
| 862 | return (uint32x2_t)(__a <= __b); } |
| 863 | __ai uint8x8_t vcle_u8(uint8x8_t __a, uint8x8_t __b) { |
| 864 | return (uint8x8_t)(__a <= __b); } |
| 865 | __ai uint16x4_t vcle_u16(uint16x4_t __a, uint16x4_t __b) { |
| 866 | return (uint16x4_t)(__a <= __b); } |
| 867 | __ai uint32x2_t vcle_u32(uint32x2_t __a, uint32x2_t __b) { |
| 868 | return (uint32x2_t)(__a <= __b); } |
| 869 | __ai uint8x16_t vcleq_s8(int8x16_t __a, int8x16_t __b) { |
| 870 | return (uint8x16_t)(__a <= __b); } |
| 871 | __ai uint16x8_t vcleq_s16(int16x8_t __a, int16x8_t __b) { |
| 872 | return (uint16x8_t)(__a <= __b); } |
| 873 | __ai uint32x4_t vcleq_s32(int32x4_t __a, int32x4_t __b) { |
| 874 | return (uint32x4_t)(__a <= __b); } |
| 875 | __ai uint32x4_t vcleq_f32(float32x4_t __a, float32x4_t __b) { |
| 876 | return (uint32x4_t)(__a <= __b); } |
| 877 | __ai uint8x16_t vcleq_u8(uint8x16_t __a, uint8x16_t __b) { |
| 878 | return (uint8x16_t)(__a <= __b); } |
| 879 | __ai uint16x8_t vcleq_u16(uint16x8_t __a, uint16x8_t __b) { |
| 880 | return (uint16x8_t)(__a <= __b); } |
| 881 | __ai uint32x4_t vcleq_u32(uint32x4_t __a, uint32x4_t __b) { |
| 882 | return (uint32x4_t)(__a <= __b); } |
| 883 | |
| 884 | __ai int8x8_t vcls_s8(int8x8_t __a) { |
| 885 | return (int8x8_t)__builtin_neon_vcls_v(__a, 0); } |
| 886 | __ai int16x4_t vcls_s16(int16x4_t __a) { |
| 887 | return (int16x4_t)__builtin_neon_vcls_v((int8x8_t)__a, 1); } |
| 888 | __ai int32x2_t vcls_s32(int32x2_t __a) { |
| 889 | return (int32x2_t)__builtin_neon_vcls_v((int8x8_t)__a, 2); } |
| 890 | __ai int8x16_t vclsq_s8(int8x16_t __a) { |
| 891 | return (int8x16_t)__builtin_neon_vclsq_v(__a, 32); } |
| 892 | __ai int16x8_t vclsq_s16(int16x8_t __a) { |
| 893 | return (int16x8_t)__builtin_neon_vclsq_v((int8x16_t)__a, 33); } |
| 894 | __ai int32x4_t vclsq_s32(int32x4_t __a) { |
| 895 | return (int32x4_t)__builtin_neon_vclsq_v((int8x16_t)__a, 34); } |
| 896 | |
| 897 | __ai uint8x8_t vclt_s8(int8x8_t __a, int8x8_t __b) { |
| 898 | return (uint8x8_t)(__a < __b); } |
| 899 | __ai uint16x4_t vclt_s16(int16x4_t __a, int16x4_t __b) { |
| 900 | return (uint16x4_t)(__a < __b); } |
| 901 | __ai uint32x2_t vclt_s32(int32x2_t __a, int32x2_t __b) { |
| 902 | return (uint32x2_t)(__a < __b); } |
| 903 | __ai uint32x2_t vclt_f32(float32x2_t __a, float32x2_t __b) { |
| 904 | return (uint32x2_t)(__a < __b); } |
| 905 | __ai uint8x8_t vclt_u8(uint8x8_t __a, uint8x8_t __b) { |
| 906 | return (uint8x8_t)(__a < __b); } |
| 907 | __ai uint16x4_t vclt_u16(uint16x4_t __a, uint16x4_t __b) { |
| 908 | return (uint16x4_t)(__a < __b); } |
| 909 | __ai uint32x2_t vclt_u32(uint32x2_t __a, uint32x2_t __b) { |
| 910 | return (uint32x2_t)(__a < __b); } |
| 911 | __ai uint8x16_t vcltq_s8(int8x16_t __a, int8x16_t __b) { |
| 912 | return (uint8x16_t)(__a < __b); } |
| 913 | __ai uint16x8_t vcltq_s16(int16x8_t __a, int16x8_t __b) { |
| 914 | return (uint16x8_t)(__a < __b); } |
| 915 | __ai uint32x4_t vcltq_s32(int32x4_t __a, int32x4_t __b) { |
| 916 | return (uint32x4_t)(__a < __b); } |
| 917 | __ai uint32x4_t vcltq_f32(float32x4_t __a, float32x4_t __b) { |
| 918 | return (uint32x4_t)(__a < __b); } |
| 919 | __ai uint8x16_t vcltq_u8(uint8x16_t __a, uint8x16_t __b) { |
| 920 | return (uint8x16_t)(__a < __b); } |
| 921 | __ai uint16x8_t vcltq_u16(uint16x8_t __a, uint16x8_t __b) { |
| 922 | return (uint16x8_t)(__a < __b); } |
| 923 | __ai uint32x4_t vcltq_u32(uint32x4_t __a, uint32x4_t __b) { |
| 924 | return (uint32x4_t)(__a < __b); } |
| 925 | |
| 926 | __ai int8x8_t vclz_s8(int8x8_t __a) { |
| 927 | return (int8x8_t)__builtin_neon_vclz_v(__a, 0); } |
| 928 | __ai int16x4_t vclz_s16(int16x4_t __a) { |
| 929 | return (int16x4_t)__builtin_neon_vclz_v((int8x8_t)__a, 1); } |
| 930 | __ai int32x2_t vclz_s32(int32x2_t __a) { |
| 931 | return (int32x2_t)__builtin_neon_vclz_v((int8x8_t)__a, 2); } |
| 932 | __ai uint8x8_t vclz_u8(uint8x8_t __a) { |
| 933 | return (uint8x8_t)__builtin_neon_vclz_v((int8x8_t)__a, 16); } |
| 934 | __ai uint16x4_t vclz_u16(uint16x4_t __a) { |
| 935 | return (uint16x4_t)__builtin_neon_vclz_v((int8x8_t)__a, 17); } |
| 936 | __ai uint32x2_t vclz_u32(uint32x2_t __a) { |
| 937 | return (uint32x2_t)__builtin_neon_vclz_v((int8x8_t)__a, 18); } |
| 938 | __ai int8x16_t vclzq_s8(int8x16_t __a) { |
| 939 | return (int8x16_t)__builtin_neon_vclzq_v(__a, 32); } |
| 940 | __ai int16x8_t vclzq_s16(int16x8_t __a) { |
| 941 | return (int16x8_t)__builtin_neon_vclzq_v((int8x16_t)__a, 33); } |
| 942 | __ai int32x4_t vclzq_s32(int32x4_t __a) { |
| 943 | return (int32x4_t)__builtin_neon_vclzq_v((int8x16_t)__a, 34); } |
| 944 | __ai uint8x16_t vclzq_u8(uint8x16_t __a) { |
| 945 | return (uint8x16_t)__builtin_neon_vclzq_v((int8x16_t)__a, 48); } |
| 946 | __ai uint16x8_t vclzq_u16(uint16x8_t __a) { |
| 947 | return (uint16x8_t)__builtin_neon_vclzq_v((int8x16_t)__a, 49); } |
| 948 | __ai uint32x4_t vclzq_u32(uint32x4_t __a) { |
| 949 | return (uint32x4_t)__builtin_neon_vclzq_v((int8x16_t)__a, 50); } |
| 950 | |
| 951 | __ai uint8x8_t vcnt_u8(uint8x8_t __a) { |
| 952 | return (uint8x8_t)__builtin_neon_vcnt_v((int8x8_t)__a, 16); } |
| 953 | __ai int8x8_t vcnt_s8(int8x8_t __a) { |
| 954 | return (int8x8_t)__builtin_neon_vcnt_v(__a, 0); } |
| 955 | __ai poly8x8_t vcnt_p8(poly8x8_t __a) { |
| 956 | return (poly8x8_t)__builtin_neon_vcnt_v((int8x8_t)__a, 4); } |
| 957 | __ai uint8x16_t vcntq_u8(uint8x16_t __a) { |
| 958 | return (uint8x16_t)__builtin_neon_vcntq_v((int8x16_t)__a, 48); } |
| 959 | __ai int8x16_t vcntq_s8(int8x16_t __a) { |
| 960 | return (int8x16_t)__builtin_neon_vcntq_v(__a, 32); } |
| 961 | __ai poly8x16_t vcntq_p8(poly8x16_t __a) { |
| 962 | return (poly8x16_t)__builtin_neon_vcntq_v((int8x16_t)__a, 36); } |
| 963 | |
| 964 | __ai int8x16_t vcombine_s8(int8x8_t __a, int8x8_t __b) { |
| 965 | return (int8x16_t)__builtin_shufflevector((int64x1_t)__a, (int64x1_t)__b, 0, 1); } |
| 966 | __ai int16x8_t vcombine_s16(int16x4_t __a, int16x4_t __b) { |
| 967 | return (int16x8_t)__builtin_shufflevector((int64x1_t)__a, (int64x1_t)__b, 0, 1); } |
| 968 | __ai int32x4_t vcombine_s32(int32x2_t __a, int32x2_t __b) { |
| 969 | return (int32x4_t)__builtin_shufflevector((int64x1_t)__a, (int64x1_t)__b, 0, 1); } |
| 970 | __ai int64x2_t vcombine_s64(int64x1_t __a, int64x1_t __b) { |
| 971 | return (int64x2_t)__builtin_shufflevector((int64x1_t)__a, (int64x1_t)__b, 0, 1); } |
| 972 | __ai float16x8_t vcombine_f16(float16x4_t __a, float16x4_t __b) { |
| 973 | return (float16x8_t)__builtin_shufflevector((int64x1_t)__a, (int64x1_t)__b, 0, 1); } |
| 974 | __ai float32x4_t vcombine_f32(float32x2_t __a, float32x2_t __b) { |
| 975 | return (float32x4_t)__builtin_shufflevector((int64x1_t)__a, (int64x1_t)__b, 0, 1); } |
| 976 | __ai uint8x16_t vcombine_u8(uint8x8_t __a, uint8x8_t __b) { |
| 977 | return (uint8x16_t)__builtin_shufflevector((int64x1_t)__a, (int64x1_t)__b, 0, 1); } |
| 978 | __ai uint16x8_t vcombine_u16(uint16x4_t __a, uint16x4_t __b) { |
| 979 | return (uint16x8_t)__builtin_shufflevector((int64x1_t)__a, (int64x1_t)__b, 0, 1); } |
| 980 | __ai uint32x4_t vcombine_u32(uint32x2_t __a, uint32x2_t __b) { |
| 981 | return (uint32x4_t)__builtin_shufflevector((int64x1_t)__a, (int64x1_t)__b, 0, 1); } |
| 982 | __ai uint64x2_t vcombine_u64(uint64x1_t __a, uint64x1_t __b) { |
| 983 | return (uint64x2_t)__builtin_shufflevector((int64x1_t)__a, (int64x1_t)__b, 0, 1); } |
| 984 | __ai poly8x16_t vcombine_p8(poly8x8_t __a, poly8x8_t __b) { |
| 985 | return (poly8x16_t)__builtin_shufflevector((int64x1_t)__a, (int64x1_t)__b, 0, 1); } |
| 986 | __ai poly16x8_t vcombine_p16(poly16x4_t __a, poly16x4_t __b) { |
| 987 | return (poly16x8_t)__builtin_shufflevector((int64x1_t)__a, (int64x1_t)__b, 0, 1); } |
| 988 | |
| 989 | __ai int8x8_t vcreate_s8(uint64_t __a) { |
| 990 | return (int8x8_t)__a; } |
| 991 | __ai int16x4_t vcreate_s16(uint64_t __a) { |
| 992 | return (int16x4_t)__a; } |
| 993 | __ai int32x2_t vcreate_s32(uint64_t __a) { |
| 994 | return (int32x2_t)__a; } |
| 995 | __ai float16x4_t vcreate_f16(uint64_t __a) { |
| 996 | return (float16x4_t)__a; } |
| 997 | __ai float32x2_t vcreate_f32(uint64_t __a) { |
| 998 | return (float32x2_t)__a; } |
| 999 | __ai uint8x8_t vcreate_u8(uint64_t __a) { |
| 1000 | return (uint8x8_t)__a; } |
| 1001 | __ai uint16x4_t vcreate_u16(uint64_t __a) { |
| 1002 | return (uint16x4_t)__a; } |
| 1003 | __ai uint32x2_t vcreate_u32(uint64_t __a) { |
| 1004 | return (uint32x2_t)__a; } |
| 1005 | __ai uint64x1_t vcreate_u64(uint64_t __a) { |
| 1006 | return (uint64x1_t)__a; } |
| 1007 | __ai poly8x8_t vcreate_p8(uint64_t __a) { |
| 1008 | return (poly8x8_t)__a; } |
| 1009 | __ai poly16x4_t vcreate_p16(uint64_t __a) { |
| 1010 | return (poly16x4_t)__a; } |
| 1011 | __ai int64x1_t vcreate_s64(uint64_t __a) { |
| 1012 | return (int64x1_t)__a; } |
| 1013 | |
| 1014 | __ai float16x4_t vcvt_f16_f32(float32x4_t __a) { |
| 1015 | return (float16x4_t)__builtin_neon_vcvt_f16_v((int8x16_t)__a, 7); } |
| 1016 | |
| 1017 | __ai float32x2_t vcvt_f32_s32(int32x2_t __a) { |
| 1018 | return (float32x2_t)__builtin_neon_vcvt_f32_v((int8x8_t)__a, 2); } |
| 1019 | __ai float32x2_t vcvt_f32_u32(uint32x2_t __a) { |
| 1020 | return (float32x2_t)__builtin_neon_vcvt_f32_v((int8x8_t)__a, 18); } |
| 1021 | __ai float32x4_t vcvtq_f32_s32(int32x4_t __a) { |
| 1022 | return (float32x4_t)__builtin_neon_vcvtq_f32_v((int8x16_t)__a, 34); } |
| 1023 | __ai float32x4_t vcvtq_f32_u32(uint32x4_t __a) { |
| 1024 | return (float32x4_t)__builtin_neon_vcvtq_f32_v((int8x16_t)__a, 50); } |
| 1025 | |
| 1026 | __ai float32x4_t vcvt_f32_f16(float16x4_t __a) { |
| 1027 | return (float32x4_t)__builtin_neon_vcvt_f32_f16((int8x8_t)__a, 7); } |
| 1028 | |
| 1029 | #define vcvt_n_f32_s32(a, __b) __extension__ ({ \ |
| 1030 | int32x2_t __a = (a); \ |
| 1031 | (float32x2_t)__builtin_neon_vcvt_n_f32_v((int8x8_t)__a, __b, 2); }) |
| 1032 | #define vcvt_n_f32_u32(a, __b) __extension__ ({ \ |
| 1033 | uint32x2_t __a = (a); \ |
| 1034 | (float32x2_t)__builtin_neon_vcvt_n_f32_v((int8x8_t)__a, __b, 18); }) |
| 1035 | #define vcvtq_n_f32_s32(a, __b) __extension__ ({ \ |
| 1036 | int32x4_t __a = (a); \ |
| 1037 | (float32x4_t)__builtin_neon_vcvtq_n_f32_v((int8x16_t)__a, __b, 34); }) |
| 1038 | #define vcvtq_n_f32_u32(a, __b) __extension__ ({ \ |
| 1039 | uint32x4_t __a = (a); \ |
| 1040 | (float32x4_t)__builtin_neon_vcvtq_n_f32_v((int8x16_t)__a, __b, 50); }) |
| 1041 | |
| 1042 | #define vcvt_n_s32_f32(a, __b) __extension__ ({ \ |
| 1043 | float32x2_t __a = (a); \ |
| 1044 | (int32x2_t)__builtin_neon_vcvt_n_s32_v((int8x8_t)__a, __b, 2); }) |
| 1045 | #define vcvtq_n_s32_f32(a, __b) __extension__ ({ \ |
| 1046 | float32x4_t __a = (a); \ |
| 1047 | (int32x4_t)__builtin_neon_vcvtq_n_s32_v((int8x16_t)__a, __b, 34); }) |
| 1048 | |
| 1049 | #define vcvt_n_u32_f32(a, __b) __extension__ ({ \ |
| 1050 | float32x2_t __a = (a); \ |
| 1051 | (uint32x2_t)__builtin_neon_vcvt_n_u32_v((int8x8_t)__a, __b, 18); }) |
| 1052 | #define vcvtq_n_u32_f32(a, __b) __extension__ ({ \ |
| 1053 | float32x4_t __a = (a); \ |
| 1054 | (uint32x4_t)__builtin_neon_vcvtq_n_u32_v((int8x16_t)__a, __b, 50); }) |
| 1055 | |
| 1056 | __ai int32x2_t vcvt_s32_f32(float32x2_t __a) { |
| 1057 | return (int32x2_t)__builtin_neon_vcvt_s32_v((int8x8_t)__a, 2); } |
| 1058 | __ai int32x4_t vcvtq_s32_f32(float32x4_t __a) { |
| 1059 | return (int32x4_t)__builtin_neon_vcvtq_s32_v((int8x16_t)__a, 34); } |
| 1060 | |
| 1061 | __ai uint32x2_t vcvt_u32_f32(float32x2_t __a) { |
| 1062 | return (uint32x2_t)__builtin_neon_vcvt_u32_v((int8x8_t)__a, 18); } |
| 1063 | __ai uint32x4_t vcvtq_u32_f32(float32x4_t __a) { |
| 1064 | return (uint32x4_t)__builtin_neon_vcvtq_u32_v((int8x16_t)__a, 50); } |
| 1065 | |
| 1066 | #define vdup_lane_u8(a, __b) __extension__ ({ \ |
| 1067 | uint8x8_t __a = (a); \ |
| 1068 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b, __b, __b, __b, __b); }) |
| 1069 | #define vdup_lane_u16(a, __b) __extension__ ({ \ |
| 1070 | uint16x4_t __a = (a); \ |
| 1071 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b); }) |
| 1072 | #define vdup_lane_u32(a, __b) __extension__ ({ \ |
| 1073 | uint32x2_t __a = (a); \ |
| 1074 | __builtin_shufflevector(__a, __a, __b, __b); }) |
| 1075 | #define vdup_lane_s8(a, __b) __extension__ ({ \ |
| 1076 | int8x8_t __a = (a); \ |
| 1077 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b, __b, __b, __b, __b); }) |
| 1078 | #define vdup_lane_s16(a, __b) __extension__ ({ \ |
| 1079 | int16x4_t __a = (a); \ |
| 1080 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b); }) |
| 1081 | #define vdup_lane_s32(a, __b) __extension__ ({ \ |
| 1082 | int32x2_t __a = (a); \ |
| 1083 | __builtin_shufflevector(__a, __a, __b, __b); }) |
| 1084 | #define vdup_lane_p8(a, __b) __extension__ ({ \ |
| 1085 | poly8x8_t __a = (a); \ |
| 1086 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b, __b, __b, __b, __b); }) |
| 1087 | #define vdup_lane_p16(a, __b) __extension__ ({ \ |
| 1088 | poly16x4_t __a = (a); \ |
| 1089 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b); }) |
| 1090 | #define vdup_lane_f32(a, __b) __extension__ ({ \ |
| 1091 | float32x2_t __a = (a); \ |
| 1092 | __builtin_shufflevector(__a, __a, __b, __b); }) |
| 1093 | #define vdupq_lane_u8(a, __b) __extension__ ({ \ |
| 1094 | uint8x8_t __a = (a); \ |
| 1095 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b); }) |
| 1096 | #define vdupq_lane_u16(a, __b) __extension__ ({ \ |
| 1097 | uint16x4_t __a = (a); \ |
| 1098 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b, __b, __b, __b, __b); }) |
| 1099 | #define vdupq_lane_u32(a, __b) __extension__ ({ \ |
| 1100 | uint32x2_t __a = (a); \ |
| 1101 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b); }) |
| 1102 | #define vdupq_lane_s8(a, __b) __extension__ ({ \ |
| 1103 | int8x8_t __a = (a); \ |
| 1104 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b); }) |
| 1105 | #define vdupq_lane_s16(a, __b) __extension__ ({ \ |
| 1106 | int16x4_t __a = (a); \ |
| 1107 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b, __b, __b, __b, __b); }) |
| 1108 | #define vdupq_lane_s32(a, __b) __extension__ ({ \ |
| 1109 | int32x2_t __a = (a); \ |
| 1110 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b); }) |
| 1111 | #define vdupq_lane_p8(a, __b) __extension__ ({ \ |
| 1112 | poly8x8_t __a = (a); \ |
| 1113 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b); }) |
| 1114 | #define vdupq_lane_p16(a, __b) __extension__ ({ \ |
| 1115 | poly16x4_t __a = (a); \ |
| 1116 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b, __b, __b, __b, __b); }) |
| 1117 | #define vdupq_lane_f32(a, __b) __extension__ ({ \ |
| 1118 | float32x2_t __a = (a); \ |
| 1119 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b); }) |
| 1120 | #define vdup_lane_s64(a, __b) __extension__ ({ \ |
| 1121 | int64x1_t __a = (a); \ |
| 1122 | __builtin_shufflevector(__a, __a, __b); }) |
| 1123 | #define vdup_lane_u64(a, __b) __extension__ ({ \ |
| 1124 | uint64x1_t __a = (a); \ |
| 1125 | __builtin_shufflevector(__a, __a, __b); }) |
| 1126 | #define vdupq_lane_s64(a, __b) __extension__ ({ \ |
| 1127 | int64x1_t __a = (a); \ |
| 1128 | __builtin_shufflevector(__a, __a, __b, __b); }) |
| 1129 | #define vdupq_lane_u64(a, __b) __extension__ ({ \ |
| 1130 | uint64x1_t __a = (a); \ |
| 1131 | __builtin_shufflevector(__a, __a, __b, __b); }) |
| 1132 | |
| 1133 | __ai uint8x8_t vdup_n_u8(uint8_t __a) { |
| 1134 | return (uint8x8_t){ __a, __a, __a, __a, __a, __a, __a, __a }; } |
| 1135 | __ai uint16x4_t vdup_n_u16(uint16_t __a) { |
| 1136 | return (uint16x4_t){ __a, __a, __a, __a }; } |
| 1137 | __ai uint32x2_t vdup_n_u32(uint32_t __a) { |
| 1138 | return (uint32x2_t){ __a, __a }; } |
| 1139 | __ai int8x8_t vdup_n_s8(int8_t __a) { |
| 1140 | return (int8x8_t){ __a, __a, __a, __a, __a, __a, __a, __a }; } |
| 1141 | __ai int16x4_t vdup_n_s16(int16_t __a) { |
| 1142 | return (int16x4_t){ __a, __a, __a, __a }; } |
| 1143 | __ai int32x2_t vdup_n_s32(int32_t __a) { |
| 1144 | return (int32x2_t){ __a, __a }; } |
| 1145 | __ai poly8x8_t vdup_n_p8(poly8_t __a) { |
| 1146 | return (poly8x8_t){ __a, __a, __a, __a, __a, __a, __a, __a }; } |
| 1147 | __ai poly16x4_t vdup_n_p16(poly16_t __a) { |
| 1148 | return (poly16x4_t){ __a, __a, __a, __a }; } |
| 1149 | __ai float32x2_t vdup_n_f32(float32_t __a) { |
| 1150 | return (float32x2_t){ __a, __a }; } |
| 1151 | __ai uint8x16_t vdupq_n_u8(uint8_t __a) { |
| 1152 | return (uint8x16_t){ __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a }; } |
| 1153 | __ai uint16x8_t vdupq_n_u16(uint16_t __a) { |
| 1154 | return (uint16x8_t){ __a, __a, __a, __a, __a, __a, __a, __a }; } |
| 1155 | __ai uint32x4_t vdupq_n_u32(uint32_t __a) { |
| 1156 | return (uint32x4_t){ __a, __a, __a, __a }; } |
| 1157 | __ai int8x16_t vdupq_n_s8(int8_t __a) { |
| 1158 | return (int8x16_t){ __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a }; } |
| 1159 | __ai int16x8_t vdupq_n_s16(int16_t __a) { |
| 1160 | return (int16x8_t){ __a, __a, __a, __a, __a, __a, __a, __a }; } |
| 1161 | __ai int32x4_t vdupq_n_s32(int32_t __a) { |
| 1162 | return (int32x4_t){ __a, __a, __a, __a }; } |
| 1163 | __ai poly8x16_t vdupq_n_p8(poly8_t __a) { |
| 1164 | return (poly8x16_t){ __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a }; } |
| 1165 | __ai poly16x8_t vdupq_n_p16(poly16_t __a) { |
| 1166 | return (poly16x8_t){ __a, __a, __a, __a, __a, __a, __a, __a }; } |
| 1167 | __ai float32x4_t vdupq_n_f32(float32_t __a) { |
| 1168 | return (float32x4_t){ __a, __a, __a, __a }; } |
| 1169 | __ai int64x1_t vdup_n_s64(int64_t __a) { |
| 1170 | return (int64x1_t){ __a }; } |
| 1171 | __ai uint64x1_t vdup_n_u64(uint64_t __a) { |
| 1172 | return (uint64x1_t){ __a }; } |
| 1173 | __ai int64x2_t vdupq_n_s64(int64_t __a) { |
| 1174 | return (int64x2_t){ __a, __a }; } |
| 1175 | __ai uint64x2_t vdupq_n_u64(uint64_t __a) { |
| 1176 | return (uint64x2_t){ __a, __a }; } |
| 1177 | |
| 1178 | __ai int8x8_t veor_s8(int8x8_t __a, int8x8_t __b) { |
| 1179 | return __a ^ __b; } |
| 1180 | __ai int16x4_t veor_s16(int16x4_t __a, int16x4_t __b) { |
| 1181 | return __a ^ __b; } |
| 1182 | __ai int32x2_t veor_s32(int32x2_t __a, int32x2_t __b) { |
| 1183 | return __a ^ __b; } |
| 1184 | __ai int64x1_t veor_s64(int64x1_t __a, int64x1_t __b) { |
| 1185 | return __a ^ __b; } |
| 1186 | __ai uint8x8_t veor_u8(uint8x8_t __a, uint8x8_t __b) { |
| 1187 | return __a ^ __b; } |
| 1188 | __ai uint16x4_t veor_u16(uint16x4_t __a, uint16x4_t __b) { |
| 1189 | return __a ^ __b; } |
| 1190 | __ai uint32x2_t veor_u32(uint32x2_t __a, uint32x2_t __b) { |
| 1191 | return __a ^ __b; } |
| 1192 | __ai uint64x1_t veor_u64(uint64x1_t __a, uint64x1_t __b) { |
| 1193 | return __a ^ __b; } |
| 1194 | __ai int8x16_t veorq_s8(int8x16_t __a, int8x16_t __b) { |
| 1195 | return __a ^ __b; } |
| 1196 | __ai int16x8_t veorq_s16(int16x8_t __a, int16x8_t __b) { |
| 1197 | return __a ^ __b; } |
| 1198 | __ai int32x4_t veorq_s32(int32x4_t __a, int32x4_t __b) { |
| 1199 | return __a ^ __b; } |
| 1200 | __ai int64x2_t veorq_s64(int64x2_t __a, int64x2_t __b) { |
| 1201 | return __a ^ __b; } |
| 1202 | __ai uint8x16_t veorq_u8(uint8x16_t __a, uint8x16_t __b) { |
| 1203 | return __a ^ __b; } |
| 1204 | __ai uint16x8_t veorq_u16(uint16x8_t __a, uint16x8_t __b) { |
| 1205 | return __a ^ __b; } |
| 1206 | __ai uint32x4_t veorq_u32(uint32x4_t __a, uint32x4_t __b) { |
| 1207 | return __a ^ __b; } |
| 1208 | __ai uint64x2_t veorq_u64(uint64x2_t __a, uint64x2_t __b) { |
| 1209 | return __a ^ __b; } |
| 1210 | |
| 1211 | #define vext_s8(a, b, __c) __extension__ ({ \ |
| 1212 | int8x8_t __a = (a); int8x8_t __b = (b); \ |
| 1213 | (int8x8_t)__builtin_neon_vext_v(__a, __b, __c, 0); }) |
| 1214 | #define vext_u8(a, b, __c) __extension__ ({ \ |
| 1215 | uint8x8_t __a = (a); uint8x8_t __b = (b); \ |
| 1216 | (uint8x8_t)__builtin_neon_vext_v((int8x8_t)__a, (int8x8_t)__b, __c, 16); }) |
| 1217 | #define vext_p8(a, b, __c) __extension__ ({ \ |
| 1218 | poly8x8_t __a = (a); poly8x8_t __b = (b); \ |
| 1219 | (poly8x8_t)__builtin_neon_vext_v((int8x8_t)__a, (int8x8_t)__b, __c, 4); }) |
| 1220 | #define vext_s16(a, b, __c) __extension__ ({ \ |
| 1221 | int16x4_t __a = (a); int16x4_t __b = (b); \ |
| 1222 | (int16x4_t)__builtin_neon_vext_v((int8x8_t)__a, (int8x8_t)__b, __c, 1); }) |
| 1223 | #define vext_u16(a, b, __c) __extension__ ({ \ |
| 1224 | uint16x4_t __a = (a); uint16x4_t __b = (b); \ |
| 1225 | (uint16x4_t)__builtin_neon_vext_v((int8x8_t)__a, (int8x8_t)__b, __c, 17); }) |
| 1226 | #define vext_p16(a, b, __c) __extension__ ({ \ |
| 1227 | poly16x4_t __a = (a); poly16x4_t __b = (b); \ |
| 1228 | (poly16x4_t)__builtin_neon_vext_v((int8x8_t)__a, (int8x8_t)__b, __c, 5); }) |
| 1229 | #define vext_s32(a, b, __c) __extension__ ({ \ |
| 1230 | int32x2_t __a = (a); int32x2_t __b = (b); \ |
| 1231 | (int32x2_t)__builtin_neon_vext_v((int8x8_t)__a, (int8x8_t)__b, __c, 2); }) |
| 1232 | #define vext_u32(a, b, __c) __extension__ ({ \ |
| 1233 | uint32x2_t __a = (a); uint32x2_t __b = (b); \ |
| 1234 | (uint32x2_t)__builtin_neon_vext_v((int8x8_t)__a, (int8x8_t)__b, __c, 18); }) |
| 1235 | #define vext_s64(a, b, __c) __extension__ ({ \ |
| 1236 | int64x1_t __a = (a); int64x1_t __b = (b); \ |
| 1237 | (int64x1_t)__builtin_neon_vext_v((int8x8_t)__a, (int8x8_t)__b, __c, 3); }) |
| 1238 | #define vext_u64(a, b, __c) __extension__ ({ \ |
| 1239 | uint64x1_t __a = (a); uint64x1_t __b = (b); \ |
| 1240 | (uint64x1_t)__builtin_neon_vext_v((int8x8_t)__a, (int8x8_t)__b, __c, 19); }) |
| 1241 | #define vext_f32(a, b, __c) __extension__ ({ \ |
| 1242 | float32x2_t __a = (a); float32x2_t __b = (b); \ |
| 1243 | (float32x2_t)__builtin_neon_vext_v((int8x8_t)__a, (int8x8_t)__b, __c, 8); }) |
| 1244 | #define vextq_s8(a, b, __c) __extension__ ({ \ |
| 1245 | int8x16_t __a = (a); int8x16_t __b = (b); \ |
| 1246 | (int8x16_t)__builtin_neon_vextq_v(__a, __b, __c, 32); }) |
| 1247 | #define vextq_u8(a, b, __c) __extension__ ({ \ |
| 1248 | uint8x16_t __a = (a); uint8x16_t __b = (b); \ |
| 1249 | (uint8x16_t)__builtin_neon_vextq_v((int8x16_t)__a, (int8x16_t)__b, __c, 48); }) |
| 1250 | #define vextq_p8(a, b, __c) __extension__ ({ \ |
| 1251 | poly8x16_t __a = (a); poly8x16_t __b = (b); \ |
| 1252 | (poly8x16_t)__builtin_neon_vextq_v((int8x16_t)__a, (int8x16_t)__b, __c, 36); }) |
| 1253 | #define vextq_s16(a, b, __c) __extension__ ({ \ |
| 1254 | int16x8_t __a = (a); int16x8_t __b = (b); \ |
| 1255 | (int16x8_t)__builtin_neon_vextq_v((int8x16_t)__a, (int8x16_t)__b, __c, 33); }) |
| 1256 | #define vextq_u16(a, b, __c) __extension__ ({ \ |
| 1257 | uint16x8_t __a = (a); uint16x8_t __b = (b); \ |
| 1258 | (uint16x8_t)__builtin_neon_vextq_v((int8x16_t)__a, (int8x16_t)__b, __c, 49); }) |
| 1259 | #define vextq_p16(a, b, __c) __extension__ ({ \ |
| 1260 | poly16x8_t __a = (a); poly16x8_t __b = (b); \ |
| 1261 | (poly16x8_t)__builtin_neon_vextq_v((int8x16_t)__a, (int8x16_t)__b, __c, 37); }) |
| 1262 | #define vextq_s32(a, b, __c) __extension__ ({ \ |
| 1263 | int32x4_t __a = (a); int32x4_t __b = (b); \ |
| 1264 | (int32x4_t)__builtin_neon_vextq_v((int8x16_t)__a, (int8x16_t)__b, __c, 34); }) |
| 1265 | #define vextq_u32(a, b, __c) __extension__ ({ \ |
| 1266 | uint32x4_t __a = (a); uint32x4_t __b = (b); \ |
| 1267 | (uint32x4_t)__builtin_neon_vextq_v((int8x16_t)__a, (int8x16_t)__b, __c, 50); }) |
| 1268 | #define vextq_s64(a, b, __c) __extension__ ({ \ |
| 1269 | int64x2_t __a = (a); int64x2_t __b = (b); \ |
| 1270 | (int64x2_t)__builtin_neon_vextq_v((int8x16_t)__a, (int8x16_t)__b, __c, 35); }) |
| 1271 | #define vextq_u64(a, b, __c) __extension__ ({ \ |
| 1272 | uint64x2_t __a = (a); uint64x2_t __b = (b); \ |
| 1273 | (uint64x2_t)__builtin_neon_vextq_v((int8x16_t)__a, (int8x16_t)__b, __c, 51); }) |
| 1274 | #define vextq_f32(a, b, __c) __extension__ ({ \ |
| 1275 | float32x4_t __a = (a); float32x4_t __b = (b); \ |
| 1276 | (float32x4_t)__builtin_neon_vextq_v((int8x16_t)__a, (int8x16_t)__b, __c, 40); }) |
| 1277 | |
| 1278 | __ai float32x2_t vfma_f32(float32x2_t __a, float32x2_t __b, float32x2_t __c) { |
| 1279 | return (float32x2_t)__builtin_neon_vfma_v((int8x8_t)__a, (int8x8_t)__b, (int8x8_t)__c, 8); } |
| 1280 | __ai float32x4_t vfmaq_f32(float32x4_t __a, float32x4_t __b, float32x4_t __c) { |
| 1281 | return (float32x4_t)__builtin_neon_vfmaq_v((int8x16_t)__a, (int8x16_t)__b, (int8x16_t)__c, 40); } |
| 1282 | |
| 1283 | __ai int8x8_t vget_high_s8(int8x16_t __a) { |
| 1284 | return __builtin_shufflevector(__a, __a, 8, 9, 10, 11, 12, 13, 14, 15); } |
| 1285 | __ai int16x4_t vget_high_s16(int16x8_t __a) { |
| 1286 | return __builtin_shufflevector(__a, __a, 4, 5, 6, 7); } |
| 1287 | __ai int32x2_t vget_high_s32(int32x4_t __a) { |
| 1288 | return __builtin_shufflevector(__a, __a, 2, 3); } |
| 1289 | __ai int64x1_t vget_high_s64(int64x2_t __a) { |
| 1290 | return __builtin_shufflevector(__a, __a, 1); } |
| 1291 | __ai float16x4_t vget_high_f16(float16x8_t __a) { |
| 1292 | return __builtin_shufflevector(__a, __a, 4, 5, 6, 7); } |
| 1293 | __ai float32x2_t vget_high_f32(float32x4_t __a) { |
| 1294 | return __builtin_shufflevector(__a, __a, 2, 3); } |
| 1295 | __ai uint8x8_t vget_high_u8(uint8x16_t __a) { |
| 1296 | return __builtin_shufflevector(__a, __a, 8, 9, 10, 11, 12, 13, 14, 15); } |
| 1297 | __ai uint16x4_t vget_high_u16(uint16x8_t __a) { |
| 1298 | return __builtin_shufflevector(__a, __a, 4, 5, 6, 7); } |
| 1299 | __ai uint32x2_t vget_high_u32(uint32x4_t __a) { |
| 1300 | return __builtin_shufflevector(__a, __a, 2, 3); } |
| 1301 | __ai uint64x1_t vget_high_u64(uint64x2_t __a) { |
| 1302 | return __builtin_shufflevector(__a, __a, 1); } |
| 1303 | __ai poly8x8_t vget_high_p8(poly8x16_t __a) { |
| 1304 | return __builtin_shufflevector(__a, __a, 8, 9, 10, 11, 12, 13, 14, 15); } |
| 1305 | __ai poly16x4_t vget_high_p16(poly16x8_t __a) { |
| 1306 | return __builtin_shufflevector(__a, __a, 4, 5, 6, 7); } |
| 1307 | |
| 1308 | #define vget_lane_u8(a, __b) __extension__ ({ \ |
| 1309 | uint8x8_t __a = (a); \ |
| 1310 | (uint8_t)__builtin_neon_vget_lane_i8((int8x8_t)__a, __b); }) |
| 1311 | #define vget_lane_u16(a, __b) __extension__ ({ \ |
| 1312 | uint16x4_t __a = (a); \ |
| 1313 | (uint16_t)__builtin_neon_vget_lane_i16((int16x4_t)__a, __b); }) |
| 1314 | #define vget_lane_u32(a, __b) __extension__ ({ \ |
| 1315 | uint32x2_t __a = (a); \ |
| 1316 | (uint32_t)__builtin_neon_vget_lane_i32((int32x2_t)__a, __b); }) |
| 1317 | #define vget_lane_s8(a, __b) __extension__ ({ \ |
| 1318 | int8x8_t __a = (a); \ |
| 1319 | (int8_t)__builtin_neon_vget_lane_i8(__a, __b); }) |
| 1320 | #define vget_lane_s16(a, __b) __extension__ ({ \ |
| 1321 | int16x4_t __a = (a); \ |
| 1322 | (int16_t)__builtin_neon_vget_lane_i16(__a, __b); }) |
| 1323 | #define vget_lane_s32(a, __b) __extension__ ({ \ |
| 1324 | int32x2_t __a = (a); \ |
| 1325 | (int32_t)__builtin_neon_vget_lane_i32(__a, __b); }) |
| 1326 | #define vget_lane_p8(a, __b) __extension__ ({ \ |
| 1327 | poly8x8_t __a = (a); \ |
| 1328 | (poly8_t)__builtin_neon_vget_lane_i8((int8x8_t)__a, __b); }) |
| 1329 | #define vget_lane_p16(a, __b) __extension__ ({ \ |
| 1330 | poly16x4_t __a = (a); \ |
| 1331 | (poly16_t)__builtin_neon_vget_lane_i16((int16x4_t)__a, __b); }) |
| 1332 | #define vget_lane_f32(a, __b) __extension__ ({ \ |
| 1333 | float32x2_t __a = (a); \ |
| 1334 | (float32_t)__builtin_neon_vget_lane_f32(__a, __b); }) |
| 1335 | #define vgetq_lane_u8(a, __b) __extension__ ({ \ |
| 1336 | uint8x16_t __a = (a); \ |
| 1337 | (uint8_t)__builtin_neon_vgetq_lane_i8((int8x16_t)__a, __b); }) |
| 1338 | #define vgetq_lane_u16(a, __b) __extension__ ({ \ |
| 1339 | uint16x8_t __a = (a); \ |
| 1340 | (uint16_t)__builtin_neon_vgetq_lane_i16((int16x8_t)__a, __b); }) |
| 1341 | #define vgetq_lane_u32(a, __b) __extension__ ({ \ |
| 1342 | uint32x4_t __a = (a); \ |
| 1343 | (uint32_t)__builtin_neon_vgetq_lane_i32((int32x4_t)__a, __b); }) |
| 1344 | #define vgetq_lane_s8(a, __b) __extension__ ({ \ |
| 1345 | int8x16_t __a = (a); \ |
| 1346 | (int8_t)__builtin_neon_vgetq_lane_i8(__a, __b); }) |
| 1347 | #define vgetq_lane_s16(a, __b) __extension__ ({ \ |
| 1348 | int16x8_t __a = (a); \ |
| 1349 | (int16_t)__builtin_neon_vgetq_lane_i16(__a, __b); }) |
| 1350 | #define vgetq_lane_s32(a, __b) __extension__ ({ \ |
| 1351 | int32x4_t __a = (a); \ |
| 1352 | (int32_t)__builtin_neon_vgetq_lane_i32(__a, __b); }) |
| 1353 | #define vgetq_lane_p8(a, __b) __extension__ ({ \ |
| 1354 | poly8x16_t __a = (a); \ |
| 1355 | (poly8_t)__builtin_neon_vgetq_lane_i8((int8x16_t)__a, __b); }) |
| 1356 | #define vgetq_lane_p16(a, __b) __extension__ ({ \ |
| 1357 | poly16x8_t __a = (a); \ |
| 1358 | (poly16_t)__builtin_neon_vgetq_lane_i16((int16x8_t)__a, __b); }) |
| 1359 | #define vgetq_lane_f32(a, __b) __extension__ ({ \ |
| 1360 | float32x4_t __a = (a); \ |
| 1361 | (float32_t)__builtin_neon_vgetq_lane_f32(__a, __b); }) |
| 1362 | #define vget_lane_s64(a, __b) __extension__ ({ \ |
| 1363 | int64x1_t __a = (a); \ |
| 1364 | (int64_t)__builtin_neon_vget_lane_i64(__a, __b); }) |
| 1365 | #define vget_lane_u64(a, __b) __extension__ ({ \ |
| 1366 | uint64x1_t __a = (a); \ |
| 1367 | (uint64_t)__builtin_neon_vget_lane_i64((int64x1_t)__a, __b); }) |
| 1368 | #define vgetq_lane_s64(a, __b) __extension__ ({ \ |
| 1369 | int64x2_t __a = (a); \ |
| 1370 | (int64_t)__builtin_neon_vgetq_lane_i64(__a, __b); }) |
| 1371 | #define vgetq_lane_u64(a, __b) __extension__ ({ \ |
| 1372 | uint64x2_t __a = (a); \ |
| 1373 | (uint64_t)__builtin_neon_vgetq_lane_i64((int64x2_t)__a, __b); }) |
| 1374 | |
| 1375 | __ai int8x8_t vget_low_s8(int8x16_t __a) { |
| 1376 | return __builtin_shufflevector(__a, __a, 0, 1, 2, 3, 4, 5, 6, 7); } |
| 1377 | __ai int16x4_t vget_low_s16(int16x8_t __a) { |
| 1378 | return __builtin_shufflevector(__a, __a, 0, 1, 2, 3); } |
| 1379 | __ai int32x2_t vget_low_s32(int32x4_t __a) { |
| 1380 | return __builtin_shufflevector(__a, __a, 0, 1); } |
| 1381 | __ai int64x1_t vget_low_s64(int64x2_t __a) { |
| 1382 | return __builtin_shufflevector(__a, __a, 0); } |
| 1383 | __ai float16x4_t vget_low_f16(float16x8_t __a) { |
| 1384 | return __builtin_shufflevector(__a, __a, 0, 1, 2, 3); } |
| 1385 | __ai float32x2_t vget_low_f32(float32x4_t __a) { |
| 1386 | return __builtin_shufflevector(__a, __a, 0, 1); } |
| 1387 | __ai uint8x8_t vget_low_u8(uint8x16_t __a) { |
| 1388 | return __builtin_shufflevector(__a, __a, 0, 1, 2, 3, 4, 5, 6, 7); } |
| 1389 | __ai uint16x4_t vget_low_u16(uint16x8_t __a) { |
| 1390 | return __builtin_shufflevector(__a, __a, 0, 1, 2, 3); } |
| 1391 | __ai uint32x2_t vget_low_u32(uint32x4_t __a) { |
| 1392 | return __builtin_shufflevector(__a, __a, 0, 1); } |
| 1393 | __ai uint64x1_t vget_low_u64(uint64x2_t __a) { |
| 1394 | return __builtin_shufflevector(__a, __a, 0); } |
| 1395 | __ai poly8x8_t vget_low_p8(poly8x16_t __a) { |
| 1396 | return __builtin_shufflevector(__a, __a, 0, 1, 2, 3, 4, 5, 6, 7); } |
| 1397 | __ai poly16x4_t vget_low_p16(poly16x8_t __a) { |
| 1398 | return __builtin_shufflevector(__a, __a, 0, 1, 2, 3); } |
| 1399 | |
| 1400 | __ai int8x8_t vhadd_s8(int8x8_t __a, int8x8_t __b) { |
| 1401 | return (int8x8_t)__builtin_neon_vhadd_v(__a, __b, 0); } |
| 1402 | __ai int16x4_t vhadd_s16(int16x4_t __a, int16x4_t __b) { |
| 1403 | return (int16x4_t)__builtin_neon_vhadd_v((int8x8_t)__a, (int8x8_t)__b, 1); } |
| 1404 | __ai int32x2_t vhadd_s32(int32x2_t __a, int32x2_t __b) { |
| 1405 | return (int32x2_t)__builtin_neon_vhadd_v((int8x8_t)__a, (int8x8_t)__b, 2); } |
| 1406 | __ai uint8x8_t vhadd_u8(uint8x8_t __a, uint8x8_t __b) { |
| 1407 | return (uint8x8_t)__builtin_neon_vhadd_v((int8x8_t)__a, (int8x8_t)__b, 16); } |
| 1408 | __ai uint16x4_t vhadd_u16(uint16x4_t __a, uint16x4_t __b) { |
| 1409 | return (uint16x4_t)__builtin_neon_vhadd_v((int8x8_t)__a, (int8x8_t)__b, 17); } |
| 1410 | __ai uint32x2_t vhadd_u32(uint32x2_t __a, uint32x2_t __b) { |
| 1411 | return (uint32x2_t)__builtin_neon_vhadd_v((int8x8_t)__a, (int8x8_t)__b, 18); } |
| 1412 | __ai int8x16_t vhaddq_s8(int8x16_t __a, int8x16_t __b) { |
| 1413 | return (int8x16_t)__builtin_neon_vhaddq_v(__a, __b, 32); } |
| 1414 | __ai int16x8_t vhaddq_s16(int16x8_t __a, int16x8_t __b) { |
| 1415 | return (int16x8_t)__builtin_neon_vhaddq_v((int8x16_t)__a, (int8x16_t)__b, 33); } |
| 1416 | __ai int32x4_t vhaddq_s32(int32x4_t __a, int32x4_t __b) { |
| 1417 | return (int32x4_t)__builtin_neon_vhaddq_v((int8x16_t)__a, (int8x16_t)__b, 34); } |
| 1418 | __ai uint8x16_t vhaddq_u8(uint8x16_t __a, uint8x16_t __b) { |
| 1419 | return (uint8x16_t)__builtin_neon_vhaddq_v((int8x16_t)__a, (int8x16_t)__b, 48); } |
| 1420 | __ai uint16x8_t vhaddq_u16(uint16x8_t __a, uint16x8_t __b) { |
| 1421 | return (uint16x8_t)__builtin_neon_vhaddq_v((int8x16_t)__a, (int8x16_t)__b, 49); } |
| 1422 | __ai uint32x4_t vhaddq_u32(uint32x4_t __a, uint32x4_t __b) { |
| 1423 | return (uint32x4_t)__builtin_neon_vhaddq_v((int8x16_t)__a, (int8x16_t)__b, 50); } |
| 1424 | |
| 1425 | __ai int8x8_t vhsub_s8(int8x8_t __a, int8x8_t __b) { |
| 1426 | return (int8x8_t)__builtin_neon_vhsub_v(__a, __b, 0); } |
| 1427 | __ai int16x4_t vhsub_s16(int16x4_t __a, int16x4_t __b) { |
| 1428 | return (int16x4_t)__builtin_neon_vhsub_v((int8x8_t)__a, (int8x8_t)__b, 1); } |
| 1429 | __ai int32x2_t vhsub_s32(int32x2_t __a, int32x2_t __b) { |
| 1430 | return (int32x2_t)__builtin_neon_vhsub_v((int8x8_t)__a, (int8x8_t)__b, 2); } |
| 1431 | __ai uint8x8_t vhsub_u8(uint8x8_t __a, uint8x8_t __b) { |
| 1432 | return (uint8x8_t)__builtin_neon_vhsub_v((int8x8_t)__a, (int8x8_t)__b, 16); } |
| 1433 | __ai uint16x4_t vhsub_u16(uint16x4_t __a, uint16x4_t __b) { |
| 1434 | return (uint16x4_t)__builtin_neon_vhsub_v((int8x8_t)__a, (int8x8_t)__b, 17); } |
| 1435 | __ai uint32x2_t vhsub_u32(uint32x2_t __a, uint32x2_t __b) { |
| 1436 | return (uint32x2_t)__builtin_neon_vhsub_v((int8x8_t)__a, (int8x8_t)__b, 18); } |
| 1437 | __ai int8x16_t vhsubq_s8(int8x16_t __a, int8x16_t __b) { |
| 1438 | return (int8x16_t)__builtin_neon_vhsubq_v(__a, __b, 32); } |
| 1439 | __ai int16x8_t vhsubq_s16(int16x8_t __a, int16x8_t __b) { |
| 1440 | return (int16x8_t)__builtin_neon_vhsubq_v((int8x16_t)__a, (int8x16_t)__b, 33); } |
| 1441 | __ai int32x4_t vhsubq_s32(int32x4_t __a, int32x4_t __b) { |
| 1442 | return (int32x4_t)__builtin_neon_vhsubq_v((int8x16_t)__a, (int8x16_t)__b, 34); } |
| 1443 | __ai uint8x16_t vhsubq_u8(uint8x16_t __a, uint8x16_t __b) { |
| 1444 | return (uint8x16_t)__builtin_neon_vhsubq_v((int8x16_t)__a, (int8x16_t)__b, 48); } |
| 1445 | __ai uint16x8_t vhsubq_u16(uint16x8_t __a, uint16x8_t __b) { |
| 1446 | return (uint16x8_t)__builtin_neon_vhsubq_v((int8x16_t)__a, (int8x16_t)__b, 49); } |
| 1447 | __ai uint32x4_t vhsubq_u32(uint32x4_t __a, uint32x4_t __b) { |
| 1448 | return (uint32x4_t)__builtin_neon_vhsubq_v((int8x16_t)__a, (int8x16_t)__b, 50); } |
| 1449 | |
| 1450 | #define vld1q_u8(__a) __extension__ ({ \ |
| 1451 | (uint8x16_t)__builtin_neon_vld1q_v(__a, 48); }) |
| 1452 | #define vld1q_u16(__a) __extension__ ({ \ |
| 1453 | (uint16x8_t)__builtin_neon_vld1q_v(__a, 49); }) |
| 1454 | #define vld1q_u32(__a) __extension__ ({ \ |
| 1455 | (uint32x4_t)__builtin_neon_vld1q_v(__a, 50); }) |
| 1456 | #define vld1q_u64(__a) __extension__ ({ \ |
| 1457 | (uint64x2_t)__builtin_neon_vld1q_v(__a, 51); }) |
| 1458 | #define vld1q_s8(__a) __extension__ ({ \ |
| 1459 | (int8x16_t)__builtin_neon_vld1q_v(__a, 32); }) |
| 1460 | #define vld1q_s16(__a) __extension__ ({ \ |
| 1461 | (int16x8_t)__builtin_neon_vld1q_v(__a, 33); }) |
| 1462 | #define vld1q_s32(__a) __extension__ ({ \ |
| 1463 | (int32x4_t)__builtin_neon_vld1q_v(__a, 34); }) |
| 1464 | #define vld1q_s64(__a) __extension__ ({ \ |
| 1465 | (int64x2_t)__builtin_neon_vld1q_v(__a, 35); }) |
| 1466 | #define vld1q_f16(__a) __extension__ ({ \ |
| 1467 | (float16x8_t)__builtin_neon_vld1q_v(__a, 39); }) |
| 1468 | #define vld1q_f32(__a) __extension__ ({ \ |
| 1469 | (float32x4_t)__builtin_neon_vld1q_v(__a, 40); }) |
| 1470 | #define vld1q_p8(__a) __extension__ ({ \ |
| 1471 | (poly8x16_t)__builtin_neon_vld1q_v(__a, 36); }) |
| 1472 | #define vld1q_p16(__a) __extension__ ({ \ |
| 1473 | (poly16x8_t)__builtin_neon_vld1q_v(__a, 37); }) |
| 1474 | #define vld1_u8(__a) __extension__ ({ \ |
| 1475 | (uint8x8_t)__builtin_neon_vld1_v(__a, 16); }) |
| 1476 | #define vld1_u16(__a) __extension__ ({ \ |
| 1477 | (uint16x4_t)__builtin_neon_vld1_v(__a, 17); }) |
| 1478 | #define vld1_u32(__a) __extension__ ({ \ |
| 1479 | (uint32x2_t)__builtin_neon_vld1_v(__a, 18); }) |
| 1480 | #define vld1_u64(__a) __extension__ ({ \ |
| 1481 | (uint64x1_t)__builtin_neon_vld1_v(__a, 19); }) |
| 1482 | #define vld1_s8(__a) __extension__ ({ \ |
| 1483 | (int8x8_t)__builtin_neon_vld1_v(__a, 0); }) |
| 1484 | #define vld1_s16(__a) __extension__ ({ \ |
| 1485 | (int16x4_t)__builtin_neon_vld1_v(__a, 1); }) |
| 1486 | #define vld1_s32(__a) __extension__ ({ \ |
| 1487 | (int32x2_t)__builtin_neon_vld1_v(__a, 2); }) |
| 1488 | #define vld1_s64(__a) __extension__ ({ \ |
| 1489 | (int64x1_t)__builtin_neon_vld1_v(__a, 3); }) |
| 1490 | #define vld1_f16(__a) __extension__ ({ \ |
| 1491 | (float16x4_t)__builtin_neon_vld1_v(__a, 7); }) |
| 1492 | #define vld1_f32(__a) __extension__ ({ \ |
| 1493 | (float32x2_t)__builtin_neon_vld1_v(__a, 8); }) |
| 1494 | #define vld1_p8(__a) __extension__ ({ \ |
| 1495 | (poly8x8_t)__builtin_neon_vld1_v(__a, 4); }) |
| 1496 | #define vld1_p16(__a) __extension__ ({ \ |
| 1497 | (poly16x4_t)__builtin_neon_vld1_v(__a, 5); }) |
| 1498 | |
| 1499 | #define vld1q_dup_u8(__a) __extension__ ({ \ |
| 1500 | (uint8x16_t)__builtin_neon_vld1q_dup_v(__a, 48); }) |
| 1501 | #define vld1q_dup_u16(__a) __extension__ ({ \ |
| 1502 | (uint16x8_t)__builtin_neon_vld1q_dup_v(__a, 49); }) |
| 1503 | #define vld1q_dup_u32(__a) __extension__ ({ \ |
| 1504 | (uint32x4_t)__builtin_neon_vld1q_dup_v(__a, 50); }) |
| 1505 | #define vld1q_dup_u64(__a) __extension__ ({ \ |
| 1506 | (uint64x2_t)__builtin_neon_vld1q_dup_v(__a, 51); }) |
| 1507 | #define vld1q_dup_s8(__a) __extension__ ({ \ |
| 1508 | (int8x16_t)__builtin_neon_vld1q_dup_v(__a, 32); }) |
| 1509 | #define vld1q_dup_s16(__a) __extension__ ({ \ |
| 1510 | (int16x8_t)__builtin_neon_vld1q_dup_v(__a, 33); }) |
| 1511 | #define vld1q_dup_s32(__a) __extension__ ({ \ |
| 1512 | (int32x4_t)__builtin_neon_vld1q_dup_v(__a, 34); }) |
| 1513 | #define vld1q_dup_s64(__a) __extension__ ({ \ |
| 1514 | (int64x2_t)__builtin_neon_vld1q_dup_v(__a, 35); }) |
| 1515 | #define vld1q_dup_f16(__a) __extension__ ({ \ |
| 1516 | (float16x8_t)__builtin_neon_vld1q_dup_v(__a, 39); }) |
| 1517 | #define vld1q_dup_f32(__a) __extension__ ({ \ |
| 1518 | (float32x4_t)__builtin_neon_vld1q_dup_v(__a, 40); }) |
| 1519 | #define vld1q_dup_p8(__a) __extension__ ({ \ |
| 1520 | (poly8x16_t)__builtin_neon_vld1q_dup_v(__a, 36); }) |
| 1521 | #define vld1q_dup_p16(__a) __extension__ ({ \ |
| 1522 | (poly16x8_t)__builtin_neon_vld1q_dup_v(__a, 37); }) |
| 1523 | #define vld1_dup_u8(__a) __extension__ ({ \ |
| 1524 | (uint8x8_t)__builtin_neon_vld1_dup_v(__a, 16); }) |
| 1525 | #define vld1_dup_u16(__a) __extension__ ({ \ |
| 1526 | (uint16x4_t)__builtin_neon_vld1_dup_v(__a, 17); }) |
| 1527 | #define vld1_dup_u32(__a) __extension__ ({ \ |
| 1528 | (uint32x2_t)__builtin_neon_vld1_dup_v(__a, 18); }) |
| 1529 | #define vld1_dup_u64(__a) __extension__ ({ \ |
| 1530 | (uint64x1_t)__builtin_neon_vld1_dup_v(__a, 19); }) |
| 1531 | #define vld1_dup_s8(__a) __extension__ ({ \ |
| 1532 | (int8x8_t)__builtin_neon_vld1_dup_v(__a, 0); }) |
| 1533 | #define vld1_dup_s16(__a) __extension__ ({ \ |
| 1534 | (int16x4_t)__builtin_neon_vld1_dup_v(__a, 1); }) |
| 1535 | #define vld1_dup_s32(__a) __extension__ ({ \ |
| 1536 | (int32x2_t)__builtin_neon_vld1_dup_v(__a, 2); }) |
| 1537 | #define vld1_dup_s64(__a) __extension__ ({ \ |
| 1538 | (int64x1_t)__builtin_neon_vld1_dup_v(__a, 3); }) |
| 1539 | #define vld1_dup_f16(__a) __extension__ ({ \ |
| 1540 | (float16x4_t)__builtin_neon_vld1_dup_v(__a, 7); }) |
| 1541 | #define vld1_dup_f32(__a) __extension__ ({ \ |
| 1542 | (float32x2_t)__builtin_neon_vld1_dup_v(__a, 8); }) |
| 1543 | #define vld1_dup_p8(__a) __extension__ ({ \ |
| 1544 | (poly8x8_t)__builtin_neon_vld1_dup_v(__a, 4); }) |
| 1545 | #define vld1_dup_p16(__a) __extension__ ({ \ |
| 1546 | (poly16x4_t)__builtin_neon_vld1_dup_v(__a, 5); }) |
| 1547 | |
| 1548 | #define vld1q_lane_u8(__a, b, __c) __extension__ ({ \ |
| 1549 | uint8x16_t __b = (b); \ |
| 1550 | (uint8x16_t)__builtin_neon_vld1q_lane_v(__a, (int8x16_t)__b, __c, 48); }) |
| 1551 | #define vld1q_lane_u16(__a, b, __c) __extension__ ({ \ |
| 1552 | uint16x8_t __b = (b); \ |
| 1553 | (uint16x8_t)__builtin_neon_vld1q_lane_v(__a, (int8x16_t)__b, __c, 49); }) |
| 1554 | #define vld1q_lane_u32(__a, b, __c) __extension__ ({ \ |
| 1555 | uint32x4_t __b = (b); \ |
| 1556 | (uint32x4_t)__builtin_neon_vld1q_lane_v(__a, (int8x16_t)__b, __c, 50); }) |
| 1557 | #define vld1q_lane_u64(__a, b, __c) __extension__ ({ \ |
| 1558 | uint64x2_t __b = (b); \ |
| 1559 | (uint64x2_t)__builtin_neon_vld1q_lane_v(__a, (int8x16_t)__b, __c, 51); }) |
| 1560 | #define vld1q_lane_s8(__a, b, __c) __extension__ ({ \ |
| 1561 | int8x16_t __b = (b); \ |
| 1562 | (int8x16_t)__builtin_neon_vld1q_lane_v(__a, __b, __c, 32); }) |
| 1563 | #define vld1q_lane_s16(__a, b, __c) __extension__ ({ \ |
| 1564 | int16x8_t __b = (b); \ |
| 1565 | (int16x8_t)__builtin_neon_vld1q_lane_v(__a, (int8x16_t)__b, __c, 33); }) |
| 1566 | #define vld1q_lane_s32(__a, b, __c) __extension__ ({ \ |
| 1567 | int32x4_t __b = (b); \ |
| 1568 | (int32x4_t)__builtin_neon_vld1q_lane_v(__a, (int8x16_t)__b, __c, 34); }) |
| 1569 | #define vld1q_lane_s64(__a, b, __c) __extension__ ({ \ |
| 1570 | int64x2_t __b = (b); \ |
| 1571 | (int64x2_t)__builtin_neon_vld1q_lane_v(__a, (int8x16_t)__b, __c, 35); }) |
| 1572 | #define vld1q_lane_f16(__a, b, __c) __extension__ ({ \ |
| 1573 | float16x8_t __b = (b); \ |
| 1574 | (float16x8_t)__builtin_neon_vld1q_lane_v(__a, (int8x16_t)__b, __c, 39); }) |
| 1575 | #define vld1q_lane_f32(__a, b, __c) __extension__ ({ \ |
| 1576 | float32x4_t __b = (b); \ |
| 1577 | (float32x4_t)__builtin_neon_vld1q_lane_v(__a, (int8x16_t)__b, __c, 40); }) |
| 1578 | #define vld1q_lane_p8(__a, b, __c) __extension__ ({ \ |
| 1579 | poly8x16_t __b = (b); \ |
| 1580 | (poly8x16_t)__builtin_neon_vld1q_lane_v(__a, (int8x16_t)__b, __c, 36); }) |
| 1581 | #define vld1q_lane_p16(__a, b, __c) __extension__ ({ \ |
| 1582 | poly16x8_t __b = (b); \ |
| 1583 | (poly16x8_t)__builtin_neon_vld1q_lane_v(__a, (int8x16_t)__b, __c, 37); }) |
| 1584 | #define vld1_lane_u8(__a, b, __c) __extension__ ({ \ |
| 1585 | uint8x8_t __b = (b); \ |
| 1586 | (uint8x8_t)__builtin_neon_vld1_lane_v(__a, (int8x8_t)__b, __c, 16); }) |
| 1587 | #define vld1_lane_u16(__a, b, __c) __extension__ ({ \ |
| 1588 | uint16x4_t __b = (b); \ |
| 1589 | (uint16x4_t)__builtin_neon_vld1_lane_v(__a, (int8x8_t)__b, __c, 17); }) |
| 1590 | #define vld1_lane_u32(__a, b, __c) __extension__ ({ \ |
| 1591 | uint32x2_t __b = (b); \ |
| 1592 | (uint32x2_t)__builtin_neon_vld1_lane_v(__a, (int8x8_t)__b, __c, 18); }) |
| 1593 | #define vld1_lane_u64(__a, b, __c) __extension__ ({ \ |
| 1594 | uint64x1_t __b = (b); \ |
| 1595 | (uint64x1_t)__builtin_neon_vld1_lane_v(__a, (int8x8_t)__b, __c, 19); }) |
| 1596 | #define vld1_lane_s8(__a, b, __c) __extension__ ({ \ |
| 1597 | int8x8_t __b = (b); \ |
| 1598 | (int8x8_t)__builtin_neon_vld1_lane_v(__a, __b, __c, 0); }) |
| 1599 | #define vld1_lane_s16(__a, b, __c) __extension__ ({ \ |
| 1600 | int16x4_t __b = (b); \ |
| 1601 | (int16x4_t)__builtin_neon_vld1_lane_v(__a, (int8x8_t)__b, __c, 1); }) |
| 1602 | #define vld1_lane_s32(__a, b, __c) __extension__ ({ \ |
| 1603 | int32x2_t __b = (b); \ |
| 1604 | (int32x2_t)__builtin_neon_vld1_lane_v(__a, (int8x8_t)__b, __c, 2); }) |
| 1605 | #define vld1_lane_s64(__a, b, __c) __extension__ ({ \ |
| 1606 | int64x1_t __b = (b); \ |
| 1607 | (int64x1_t)__builtin_neon_vld1_lane_v(__a, (int8x8_t)__b, __c, 3); }) |
| 1608 | #define vld1_lane_f16(__a, b, __c) __extension__ ({ \ |
| 1609 | float16x4_t __b = (b); \ |
| 1610 | (float16x4_t)__builtin_neon_vld1_lane_v(__a, (int8x8_t)__b, __c, 7); }) |
| 1611 | #define vld1_lane_f32(__a, b, __c) __extension__ ({ \ |
| 1612 | float32x2_t __b = (b); \ |
| 1613 | (float32x2_t)__builtin_neon_vld1_lane_v(__a, (int8x8_t)__b, __c, 8); }) |
| 1614 | #define vld1_lane_p8(__a, b, __c) __extension__ ({ \ |
| 1615 | poly8x8_t __b = (b); \ |
| 1616 | (poly8x8_t)__builtin_neon_vld1_lane_v(__a, (int8x8_t)__b, __c, 4); }) |
| 1617 | #define vld1_lane_p16(__a, b, __c) __extension__ ({ \ |
| 1618 | poly16x4_t __b = (b); \ |
| 1619 | (poly16x4_t)__builtin_neon_vld1_lane_v(__a, (int8x8_t)__b, __c, 5); }) |
| 1620 | |
| 1621 | #define vld2q_u8(__a) __extension__ ({ \ |
| 1622 | uint8x16x2_t r; __builtin_neon_vld2q_v(&r, __a, 48); r; }) |
| 1623 | #define vld2q_u16(__a) __extension__ ({ \ |
| 1624 | uint16x8x2_t r; __builtin_neon_vld2q_v(&r, __a, 49); r; }) |
| 1625 | #define vld2q_u32(__a) __extension__ ({ \ |
| 1626 | uint32x4x2_t r; __builtin_neon_vld2q_v(&r, __a, 50); r; }) |
| 1627 | #define vld2q_s8(__a) __extension__ ({ \ |
| 1628 | int8x16x2_t r; __builtin_neon_vld2q_v(&r, __a, 32); r; }) |
| 1629 | #define vld2q_s16(__a) __extension__ ({ \ |
| 1630 | int16x8x2_t r; __builtin_neon_vld2q_v(&r, __a, 33); r; }) |
| 1631 | #define vld2q_s32(__a) __extension__ ({ \ |
| 1632 | int32x4x2_t r; __builtin_neon_vld2q_v(&r, __a, 34); r; }) |
| 1633 | #define vld2q_f16(__a) __extension__ ({ \ |
| 1634 | float16x8x2_t r; __builtin_neon_vld2q_v(&r, __a, 39); r; }) |
| 1635 | #define vld2q_f32(__a) __extension__ ({ \ |
| 1636 | float32x4x2_t r; __builtin_neon_vld2q_v(&r, __a, 40); r; }) |
| 1637 | #define vld2q_p8(__a) __extension__ ({ \ |
| 1638 | poly8x16x2_t r; __builtin_neon_vld2q_v(&r, __a, 36); r; }) |
| 1639 | #define vld2q_p16(__a) __extension__ ({ \ |
| 1640 | poly16x8x2_t r; __builtin_neon_vld2q_v(&r, __a, 37); r; }) |
| 1641 | #define vld2_u8(__a) __extension__ ({ \ |
| 1642 | uint8x8x2_t r; __builtin_neon_vld2_v(&r, __a, 16); r; }) |
| 1643 | #define vld2_u16(__a) __extension__ ({ \ |
| 1644 | uint16x4x2_t r; __builtin_neon_vld2_v(&r, __a, 17); r; }) |
| 1645 | #define vld2_u32(__a) __extension__ ({ \ |
| 1646 | uint32x2x2_t r; __builtin_neon_vld2_v(&r, __a, 18); r; }) |
| 1647 | #define vld2_u64(__a) __extension__ ({ \ |
| 1648 | uint64x1x2_t r; __builtin_neon_vld2_v(&r, __a, 19); r; }) |
| 1649 | #define vld2_s8(__a) __extension__ ({ \ |
| 1650 | int8x8x2_t r; __builtin_neon_vld2_v(&r, __a, 0); r; }) |
| 1651 | #define vld2_s16(__a) __extension__ ({ \ |
| 1652 | int16x4x2_t r; __builtin_neon_vld2_v(&r, __a, 1); r; }) |
| 1653 | #define vld2_s32(__a) __extension__ ({ \ |
| 1654 | int32x2x2_t r; __builtin_neon_vld2_v(&r, __a, 2); r; }) |
| 1655 | #define vld2_s64(__a) __extension__ ({ \ |
| 1656 | int64x1x2_t r; __builtin_neon_vld2_v(&r, __a, 3); r; }) |
| 1657 | #define vld2_f16(__a) __extension__ ({ \ |
| 1658 | float16x4x2_t r; __builtin_neon_vld2_v(&r, __a, 7); r; }) |
| 1659 | #define vld2_f32(__a) __extension__ ({ \ |
| 1660 | float32x2x2_t r; __builtin_neon_vld2_v(&r, __a, 8); r; }) |
| 1661 | #define vld2_p8(__a) __extension__ ({ \ |
| 1662 | poly8x8x2_t r; __builtin_neon_vld2_v(&r, __a, 4); r; }) |
| 1663 | #define vld2_p16(__a) __extension__ ({ \ |
| 1664 | poly16x4x2_t r; __builtin_neon_vld2_v(&r, __a, 5); r; }) |
| 1665 | |
| 1666 | #define vld2_dup_u8(__a) __extension__ ({ \ |
| 1667 | uint8x8x2_t r; __builtin_neon_vld2_dup_v(&r, __a, 16); r; }) |
| 1668 | #define vld2_dup_u16(__a) __extension__ ({ \ |
| 1669 | uint16x4x2_t r; __builtin_neon_vld2_dup_v(&r, __a, 17); r; }) |
| 1670 | #define vld2_dup_u32(__a) __extension__ ({ \ |
| 1671 | uint32x2x2_t r; __builtin_neon_vld2_dup_v(&r, __a, 18); r; }) |
| 1672 | #define vld2_dup_u64(__a) __extension__ ({ \ |
| 1673 | uint64x1x2_t r; __builtin_neon_vld2_dup_v(&r, __a, 19); r; }) |
| 1674 | #define vld2_dup_s8(__a) __extension__ ({ \ |
| 1675 | int8x8x2_t r; __builtin_neon_vld2_dup_v(&r, __a, 0); r; }) |
| 1676 | #define vld2_dup_s16(__a) __extension__ ({ \ |
| 1677 | int16x4x2_t r; __builtin_neon_vld2_dup_v(&r, __a, 1); r; }) |
| 1678 | #define vld2_dup_s32(__a) __extension__ ({ \ |
| 1679 | int32x2x2_t r; __builtin_neon_vld2_dup_v(&r, __a, 2); r; }) |
| 1680 | #define vld2_dup_s64(__a) __extension__ ({ \ |
| 1681 | int64x1x2_t r; __builtin_neon_vld2_dup_v(&r, __a, 3); r; }) |
| 1682 | #define vld2_dup_f16(__a) __extension__ ({ \ |
| 1683 | float16x4x2_t r; __builtin_neon_vld2_dup_v(&r, __a, 7); r; }) |
| 1684 | #define vld2_dup_f32(__a) __extension__ ({ \ |
| 1685 | float32x2x2_t r; __builtin_neon_vld2_dup_v(&r, __a, 8); r; }) |
| 1686 | #define vld2_dup_p8(__a) __extension__ ({ \ |
| 1687 | poly8x8x2_t r; __builtin_neon_vld2_dup_v(&r, __a, 4); r; }) |
| 1688 | #define vld2_dup_p16(__a) __extension__ ({ \ |
| 1689 | poly16x4x2_t r; __builtin_neon_vld2_dup_v(&r, __a, 5); r; }) |
| 1690 | |
| 1691 | #define vld2q_lane_u16(__a, b, __c) __extension__ ({ \ |
| 1692 | uint16x8x2_t __b = (b); \ |
| 1693 | uint16x8x2_t r; __builtin_neon_vld2q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], __c, 49); r; }) |
| 1694 | #define vld2q_lane_u32(__a, b, __c) __extension__ ({ \ |
| 1695 | uint32x4x2_t __b = (b); \ |
| 1696 | uint32x4x2_t r; __builtin_neon_vld2q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], __c, 50); r; }) |
| 1697 | #define vld2q_lane_s16(__a, b, __c) __extension__ ({ \ |
| 1698 | int16x8x2_t __b = (b); \ |
| 1699 | int16x8x2_t r; __builtin_neon_vld2q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], __c, 33); r; }) |
| 1700 | #define vld2q_lane_s32(__a, b, __c) __extension__ ({ \ |
| 1701 | int32x4x2_t __b = (b); \ |
| 1702 | int32x4x2_t r; __builtin_neon_vld2q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], __c, 34); r; }) |
| 1703 | #define vld2q_lane_f16(__a, b, __c) __extension__ ({ \ |
| 1704 | float16x8x2_t __b = (b); \ |
| 1705 | float16x8x2_t r; __builtin_neon_vld2q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], __c, 39); r; }) |
| 1706 | #define vld2q_lane_f32(__a, b, __c) __extension__ ({ \ |
| 1707 | float32x4x2_t __b = (b); \ |
| 1708 | float32x4x2_t r; __builtin_neon_vld2q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], __c, 40); r; }) |
| 1709 | #define vld2q_lane_p16(__a, b, __c) __extension__ ({ \ |
| 1710 | poly16x8x2_t __b = (b); \ |
| 1711 | poly16x8x2_t r; __builtin_neon_vld2q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], __c, 37); r; }) |
| 1712 | #define vld2_lane_u8(__a, b, __c) __extension__ ({ \ |
| 1713 | uint8x8x2_t __b = (b); \ |
| 1714 | uint8x8x2_t r; __builtin_neon_vld2_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], __c, 16); r; }) |
| 1715 | #define vld2_lane_u16(__a, b, __c) __extension__ ({ \ |
| 1716 | uint16x4x2_t __b = (b); \ |
| 1717 | uint16x4x2_t r; __builtin_neon_vld2_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], __c, 17); r; }) |
| 1718 | #define vld2_lane_u32(__a, b, __c) __extension__ ({ \ |
| 1719 | uint32x2x2_t __b = (b); \ |
| 1720 | uint32x2x2_t r; __builtin_neon_vld2_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], __c, 18); r; }) |
| 1721 | #define vld2_lane_s8(__a, b, __c) __extension__ ({ \ |
| 1722 | int8x8x2_t __b = (b); \ |
| 1723 | int8x8x2_t r; __builtin_neon_vld2_lane_v(&r, __a, __b.val[0], __b.val[1], __c, 0); r; }) |
| 1724 | #define vld2_lane_s16(__a, b, __c) __extension__ ({ \ |
| 1725 | int16x4x2_t __b = (b); \ |
| 1726 | int16x4x2_t r; __builtin_neon_vld2_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], __c, 1); r; }) |
| 1727 | #define vld2_lane_s32(__a, b, __c) __extension__ ({ \ |
| 1728 | int32x2x2_t __b = (b); \ |
| 1729 | int32x2x2_t r; __builtin_neon_vld2_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], __c, 2); r; }) |
| 1730 | #define vld2_lane_f16(__a, b, __c) __extension__ ({ \ |
| 1731 | float16x4x2_t __b = (b); \ |
| 1732 | float16x4x2_t r; __builtin_neon_vld2_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], __c, 7); r; }) |
| 1733 | #define vld2_lane_f32(__a, b, __c) __extension__ ({ \ |
| 1734 | float32x2x2_t __b = (b); \ |
| 1735 | float32x2x2_t r; __builtin_neon_vld2_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], __c, 8); r; }) |
| 1736 | #define vld2_lane_p8(__a, b, __c) __extension__ ({ \ |
| 1737 | poly8x8x2_t __b = (b); \ |
| 1738 | poly8x8x2_t r; __builtin_neon_vld2_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], __c, 4); r; }) |
| 1739 | #define vld2_lane_p16(__a, b, __c) __extension__ ({ \ |
| 1740 | poly16x4x2_t __b = (b); \ |
| 1741 | poly16x4x2_t r; __builtin_neon_vld2_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], __c, 5); r; }) |
| 1742 | |
| 1743 | #define vld3q_u8(__a) __extension__ ({ \ |
| 1744 | uint8x16x3_t r; __builtin_neon_vld3q_v(&r, __a, 48); r; }) |
| 1745 | #define vld3q_u16(__a) __extension__ ({ \ |
| 1746 | uint16x8x3_t r; __builtin_neon_vld3q_v(&r, __a, 49); r; }) |
| 1747 | #define vld3q_u32(__a) __extension__ ({ \ |
| 1748 | uint32x4x3_t r; __builtin_neon_vld3q_v(&r, __a, 50); r; }) |
| 1749 | #define vld3q_s8(__a) __extension__ ({ \ |
| 1750 | int8x16x3_t r; __builtin_neon_vld3q_v(&r, __a, 32); r; }) |
| 1751 | #define vld3q_s16(__a) __extension__ ({ \ |
| 1752 | int16x8x3_t r; __builtin_neon_vld3q_v(&r, __a, 33); r; }) |
| 1753 | #define vld3q_s32(__a) __extension__ ({ \ |
| 1754 | int32x4x3_t r; __builtin_neon_vld3q_v(&r, __a, 34); r; }) |
| 1755 | #define vld3q_f16(__a) __extension__ ({ \ |
| 1756 | float16x8x3_t r; __builtin_neon_vld3q_v(&r, __a, 39); r; }) |
| 1757 | #define vld3q_f32(__a) __extension__ ({ \ |
| 1758 | float32x4x3_t r; __builtin_neon_vld3q_v(&r, __a, 40); r; }) |
| 1759 | #define vld3q_p8(__a) __extension__ ({ \ |
| 1760 | poly8x16x3_t r; __builtin_neon_vld3q_v(&r, __a, 36); r; }) |
| 1761 | #define vld3q_p16(__a) __extension__ ({ \ |
| 1762 | poly16x8x3_t r; __builtin_neon_vld3q_v(&r, __a, 37); r; }) |
| 1763 | #define vld3_u8(__a) __extension__ ({ \ |
| 1764 | uint8x8x3_t r; __builtin_neon_vld3_v(&r, __a, 16); r; }) |
| 1765 | #define vld3_u16(__a) __extension__ ({ \ |
| 1766 | uint16x4x3_t r; __builtin_neon_vld3_v(&r, __a, 17); r; }) |
| 1767 | #define vld3_u32(__a) __extension__ ({ \ |
| 1768 | uint32x2x3_t r; __builtin_neon_vld3_v(&r, __a, 18); r; }) |
| 1769 | #define vld3_u64(__a) __extension__ ({ \ |
| 1770 | uint64x1x3_t r; __builtin_neon_vld3_v(&r, __a, 19); r; }) |
| 1771 | #define vld3_s8(__a) __extension__ ({ \ |
| 1772 | int8x8x3_t r; __builtin_neon_vld3_v(&r, __a, 0); r; }) |
| 1773 | #define vld3_s16(__a) __extension__ ({ \ |
| 1774 | int16x4x3_t r; __builtin_neon_vld3_v(&r, __a, 1); r; }) |
| 1775 | #define vld3_s32(__a) __extension__ ({ \ |
| 1776 | int32x2x3_t r; __builtin_neon_vld3_v(&r, __a, 2); r; }) |
| 1777 | #define vld3_s64(__a) __extension__ ({ \ |
| 1778 | int64x1x3_t r; __builtin_neon_vld3_v(&r, __a, 3); r; }) |
| 1779 | #define vld3_f16(__a) __extension__ ({ \ |
| 1780 | float16x4x3_t r; __builtin_neon_vld3_v(&r, __a, 7); r; }) |
| 1781 | #define vld3_f32(__a) __extension__ ({ \ |
| 1782 | float32x2x3_t r; __builtin_neon_vld3_v(&r, __a, 8); r; }) |
| 1783 | #define vld3_p8(__a) __extension__ ({ \ |
| 1784 | poly8x8x3_t r; __builtin_neon_vld3_v(&r, __a, 4); r; }) |
| 1785 | #define vld3_p16(__a) __extension__ ({ \ |
| 1786 | poly16x4x3_t r; __builtin_neon_vld3_v(&r, __a, 5); r; }) |
| 1787 | |
| 1788 | #define vld3_dup_u8(__a) __extension__ ({ \ |
| 1789 | uint8x8x3_t r; __builtin_neon_vld3_dup_v(&r, __a, 16); r; }) |
| 1790 | #define vld3_dup_u16(__a) __extension__ ({ \ |
| 1791 | uint16x4x3_t r; __builtin_neon_vld3_dup_v(&r, __a, 17); r; }) |
| 1792 | #define vld3_dup_u32(__a) __extension__ ({ \ |
| 1793 | uint32x2x3_t r; __builtin_neon_vld3_dup_v(&r, __a, 18); r; }) |
| 1794 | #define vld3_dup_u64(__a) __extension__ ({ \ |
| 1795 | uint64x1x3_t r; __builtin_neon_vld3_dup_v(&r, __a, 19); r; }) |
| 1796 | #define vld3_dup_s8(__a) __extension__ ({ \ |
| 1797 | int8x8x3_t r; __builtin_neon_vld3_dup_v(&r, __a, 0); r; }) |
| 1798 | #define vld3_dup_s16(__a) __extension__ ({ \ |
| 1799 | int16x4x3_t r; __builtin_neon_vld3_dup_v(&r, __a, 1); r; }) |
| 1800 | #define vld3_dup_s32(__a) __extension__ ({ \ |
| 1801 | int32x2x3_t r; __builtin_neon_vld3_dup_v(&r, __a, 2); r; }) |
| 1802 | #define vld3_dup_s64(__a) __extension__ ({ \ |
| 1803 | int64x1x3_t r; __builtin_neon_vld3_dup_v(&r, __a, 3); r; }) |
| 1804 | #define vld3_dup_f16(__a) __extension__ ({ \ |
| 1805 | float16x4x3_t r; __builtin_neon_vld3_dup_v(&r, __a, 7); r; }) |
| 1806 | #define vld3_dup_f32(__a) __extension__ ({ \ |
| 1807 | float32x2x3_t r; __builtin_neon_vld3_dup_v(&r, __a, 8); r; }) |
| 1808 | #define vld3_dup_p8(__a) __extension__ ({ \ |
| 1809 | poly8x8x3_t r; __builtin_neon_vld3_dup_v(&r, __a, 4); r; }) |
| 1810 | #define vld3_dup_p16(__a) __extension__ ({ \ |
| 1811 | poly16x4x3_t r; __builtin_neon_vld3_dup_v(&r, __a, 5); r; }) |
| 1812 | |
| 1813 | #define vld3q_lane_u16(__a, b, __c) __extension__ ({ \ |
| 1814 | uint16x8x3_t __b = (b); \ |
| 1815 | uint16x8x3_t r; __builtin_neon_vld3q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], __c, 49); r; }) |
| 1816 | #define vld3q_lane_u32(__a, b, __c) __extension__ ({ \ |
| 1817 | uint32x4x3_t __b = (b); \ |
| 1818 | uint32x4x3_t r; __builtin_neon_vld3q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], __c, 50); r; }) |
| 1819 | #define vld3q_lane_s16(__a, b, __c) __extension__ ({ \ |
| 1820 | int16x8x3_t __b = (b); \ |
| 1821 | int16x8x3_t r; __builtin_neon_vld3q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], __c, 33); r; }) |
| 1822 | #define vld3q_lane_s32(__a, b, __c) __extension__ ({ \ |
| 1823 | int32x4x3_t __b = (b); \ |
| 1824 | int32x4x3_t r; __builtin_neon_vld3q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], __c, 34); r; }) |
| 1825 | #define vld3q_lane_f16(__a, b, __c) __extension__ ({ \ |
| 1826 | float16x8x3_t __b = (b); \ |
| 1827 | float16x8x3_t r; __builtin_neon_vld3q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], __c, 39); r; }) |
| 1828 | #define vld3q_lane_f32(__a, b, __c) __extension__ ({ \ |
| 1829 | float32x4x3_t __b = (b); \ |
| 1830 | float32x4x3_t r; __builtin_neon_vld3q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], __c, 40); r; }) |
| 1831 | #define vld3q_lane_p16(__a, b, __c) __extension__ ({ \ |
| 1832 | poly16x8x3_t __b = (b); \ |
| 1833 | poly16x8x3_t r; __builtin_neon_vld3q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], __c, 37); r; }) |
| 1834 | #define vld3_lane_u8(__a, b, __c) __extension__ ({ \ |
| 1835 | uint8x8x3_t __b = (b); \ |
| 1836 | uint8x8x3_t r; __builtin_neon_vld3_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], __c, 16); r; }) |
| 1837 | #define vld3_lane_u16(__a, b, __c) __extension__ ({ \ |
| 1838 | uint16x4x3_t __b = (b); \ |
| 1839 | uint16x4x3_t r; __builtin_neon_vld3_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], __c, 17); r; }) |
| 1840 | #define vld3_lane_u32(__a, b, __c) __extension__ ({ \ |
| 1841 | uint32x2x3_t __b = (b); \ |
| 1842 | uint32x2x3_t r; __builtin_neon_vld3_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], __c, 18); r; }) |
| 1843 | #define vld3_lane_s8(__a, b, __c) __extension__ ({ \ |
| 1844 | int8x8x3_t __b = (b); \ |
| 1845 | int8x8x3_t r; __builtin_neon_vld3_lane_v(&r, __a, __b.val[0], __b.val[1], __b.val[2], __c, 0); r; }) |
| 1846 | #define vld3_lane_s16(__a, b, __c) __extension__ ({ \ |
| 1847 | int16x4x3_t __b = (b); \ |
| 1848 | int16x4x3_t r; __builtin_neon_vld3_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], __c, 1); r; }) |
| 1849 | #define vld3_lane_s32(__a, b, __c) __extension__ ({ \ |
| 1850 | int32x2x3_t __b = (b); \ |
| 1851 | int32x2x3_t r; __builtin_neon_vld3_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], __c, 2); r; }) |
| 1852 | #define vld3_lane_f16(__a, b, __c) __extension__ ({ \ |
| 1853 | float16x4x3_t __b = (b); \ |
| 1854 | float16x4x3_t r; __builtin_neon_vld3_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], __c, 7); r; }) |
| 1855 | #define vld3_lane_f32(__a, b, __c) __extension__ ({ \ |
| 1856 | float32x2x3_t __b = (b); \ |
| 1857 | float32x2x3_t r; __builtin_neon_vld3_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], __c, 8); r; }) |
| 1858 | #define vld3_lane_p8(__a, b, __c) __extension__ ({ \ |
| 1859 | poly8x8x3_t __b = (b); \ |
| 1860 | poly8x8x3_t r; __builtin_neon_vld3_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], __c, 4); r; }) |
| 1861 | #define vld3_lane_p16(__a, b, __c) __extension__ ({ \ |
| 1862 | poly16x4x3_t __b = (b); \ |
| 1863 | poly16x4x3_t r; __builtin_neon_vld3_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], __c, 5); r; }) |
| 1864 | |
| 1865 | #define vld4q_u8(__a) __extension__ ({ \ |
| 1866 | uint8x16x4_t r; __builtin_neon_vld4q_v(&r, __a, 48); r; }) |
| 1867 | #define vld4q_u16(__a) __extension__ ({ \ |
| 1868 | uint16x8x4_t r; __builtin_neon_vld4q_v(&r, __a, 49); r; }) |
| 1869 | #define vld4q_u32(__a) __extension__ ({ \ |
| 1870 | uint32x4x4_t r; __builtin_neon_vld4q_v(&r, __a, 50); r; }) |
| 1871 | #define vld4q_s8(__a) __extension__ ({ \ |
| 1872 | int8x16x4_t r; __builtin_neon_vld4q_v(&r, __a, 32); r; }) |
| 1873 | #define vld4q_s16(__a) __extension__ ({ \ |
| 1874 | int16x8x4_t r; __builtin_neon_vld4q_v(&r, __a, 33); r; }) |
| 1875 | #define vld4q_s32(__a) __extension__ ({ \ |
| 1876 | int32x4x4_t r; __builtin_neon_vld4q_v(&r, __a, 34); r; }) |
| 1877 | #define vld4q_f16(__a) __extension__ ({ \ |
| 1878 | float16x8x4_t r; __builtin_neon_vld4q_v(&r, __a, 39); r; }) |
| 1879 | #define vld4q_f32(__a) __extension__ ({ \ |
| 1880 | float32x4x4_t r; __builtin_neon_vld4q_v(&r, __a, 40); r; }) |
| 1881 | #define vld4q_p8(__a) __extension__ ({ \ |
| 1882 | poly8x16x4_t r; __builtin_neon_vld4q_v(&r, __a, 36); r; }) |
| 1883 | #define vld4q_p16(__a) __extension__ ({ \ |
| 1884 | poly16x8x4_t r; __builtin_neon_vld4q_v(&r, __a, 37); r; }) |
| 1885 | #define vld4_u8(__a) __extension__ ({ \ |
| 1886 | uint8x8x4_t r; __builtin_neon_vld4_v(&r, __a, 16); r; }) |
| 1887 | #define vld4_u16(__a) __extension__ ({ \ |
| 1888 | uint16x4x4_t r; __builtin_neon_vld4_v(&r, __a, 17); r; }) |
| 1889 | #define vld4_u32(__a) __extension__ ({ \ |
| 1890 | uint32x2x4_t r; __builtin_neon_vld4_v(&r, __a, 18); r; }) |
| 1891 | #define vld4_u64(__a) __extension__ ({ \ |
| 1892 | uint64x1x4_t r; __builtin_neon_vld4_v(&r, __a, 19); r; }) |
| 1893 | #define vld4_s8(__a) __extension__ ({ \ |
| 1894 | int8x8x4_t r; __builtin_neon_vld4_v(&r, __a, 0); r; }) |
| 1895 | #define vld4_s16(__a) __extension__ ({ \ |
| 1896 | int16x4x4_t r; __builtin_neon_vld4_v(&r, __a, 1); r; }) |
| 1897 | #define vld4_s32(__a) __extension__ ({ \ |
| 1898 | int32x2x4_t r; __builtin_neon_vld4_v(&r, __a, 2); r; }) |
| 1899 | #define vld4_s64(__a) __extension__ ({ \ |
| 1900 | int64x1x4_t r; __builtin_neon_vld4_v(&r, __a, 3); r; }) |
| 1901 | #define vld4_f16(__a) __extension__ ({ \ |
| 1902 | float16x4x4_t r; __builtin_neon_vld4_v(&r, __a, 7); r; }) |
| 1903 | #define vld4_f32(__a) __extension__ ({ \ |
| 1904 | float32x2x4_t r; __builtin_neon_vld4_v(&r, __a, 8); r; }) |
| 1905 | #define vld4_p8(__a) __extension__ ({ \ |
| 1906 | poly8x8x4_t r; __builtin_neon_vld4_v(&r, __a, 4); r; }) |
| 1907 | #define vld4_p16(__a) __extension__ ({ \ |
| 1908 | poly16x4x4_t r; __builtin_neon_vld4_v(&r, __a, 5); r; }) |
| 1909 | |
| 1910 | #define vld4_dup_u8(__a) __extension__ ({ \ |
| 1911 | uint8x8x4_t r; __builtin_neon_vld4_dup_v(&r, __a, 16); r; }) |
| 1912 | #define vld4_dup_u16(__a) __extension__ ({ \ |
| 1913 | uint16x4x4_t r; __builtin_neon_vld4_dup_v(&r, __a, 17); r; }) |
| 1914 | #define vld4_dup_u32(__a) __extension__ ({ \ |
| 1915 | uint32x2x4_t r; __builtin_neon_vld4_dup_v(&r, __a, 18); r; }) |
| 1916 | #define vld4_dup_u64(__a) __extension__ ({ \ |
| 1917 | uint64x1x4_t r; __builtin_neon_vld4_dup_v(&r, __a, 19); r; }) |
| 1918 | #define vld4_dup_s8(__a) __extension__ ({ \ |
| 1919 | int8x8x4_t r; __builtin_neon_vld4_dup_v(&r, __a, 0); r; }) |
| 1920 | #define vld4_dup_s16(__a) __extension__ ({ \ |
| 1921 | int16x4x4_t r; __builtin_neon_vld4_dup_v(&r, __a, 1); r; }) |
| 1922 | #define vld4_dup_s32(__a) __extension__ ({ \ |
| 1923 | int32x2x4_t r; __builtin_neon_vld4_dup_v(&r, __a, 2); r; }) |
| 1924 | #define vld4_dup_s64(__a) __extension__ ({ \ |
| 1925 | int64x1x4_t r; __builtin_neon_vld4_dup_v(&r, __a, 3); r; }) |
| 1926 | #define vld4_dup_f16(__a) __extension__ ({ \ |
| 1927 | float16x4x4_t r; __builtin_neon_vld4_dup_v(&r, __a, 7); r; }) |
| 1928 | #define vld4_dup_f32(__a) __extension__ ({ \ |
| 1929 | float32x2x4_t r; __builtin_neon_vld4_dup_v(&r, __a, 8); r; }) |
| 1930 | #define vld4_dup_p8(__a) __extension__ ({ \ |
| 1931 | poly8x8x4_t r; __builtin_neon_vld4_dup_v(&r, __a, 4); r; }) |
| 1932 | #define vld4_dup_p16(__a) __extension__ ({ \ |
| 1933 | poly16x4x4_t r; __builtin_neon_vld4_dup_v(&r, __a, 5); r; }) |
| 1934 | |
| 1935 | #define vld4q_lane_u16(__a, b, __c) __extension__ ({ \ |
| 1936 | uint16x8x4_t __b = (b); \ |
| 1937 | uint16x8x4_t r; __builtin_neon_vld4q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], __c, 49); r; }) |
| 1938 | #define vld4q_lane_u32(__a, b, __c) __extension__ ({ \ |
| 1939 | uint32x4x4_t __b = (b); \ |
| 1940 | uint32x4x4_t r; __builtin_neon_vld4q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], __c, 50); r; }) |
| 1941 | #define vld4q_lane_s16(__a, b, __c) __extension__ ({ \ |
| 1942 | int16x8x4_t __b = (b); \ |
| 1943 | int16x8x4_t r; __builtin_neon_vld4q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], __c, 33); r; }) |
| 1944 | #define vld4q_lane_s32(__a, b, __c) __extension__ ({ \ |
| 1945 | int32x4x4_t __b = (b); \ |
| 1946 | int32x4x4_t r; __builtin_neon_vld4q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], __c, 34); r; }) |
| 1947 | #define vld4q_lane_f16(__a, b, __c) __extension__ ({ \ |
| 1948 | float16x8x4_t __b = (b); \ |
| 1949 | float16x8x4_t r; __builtin_neon_vld4q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], __c, 39); r; }) |
| 1950 | #define vld4q_lane_f32(__a, b, __c) __extension__ ({ \ |
| 1951 | float32x4x4_t __b = (b); \ |
| 1952 | float32x4x4_t r; __builtin_neon_vld4q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], __c, 40); r; }) |
| 1953 | #define vld4q_lane_p16(__a, b, __c) __extension__ ({ \ |
| 1954 | poly16x8x4_t __b = (b); \ |
| 1955 | poly16x8x4_t r; __builtin_neon_vld4q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], __c, 37); r; }) |
| 1956 | #define vld4_lane_u8(__a, b, __c) __extension__ ({ \ |
| 1957 | uint8x8x4_t __b = (b); \ |
| 1958 | uint8x8x4_t r; __builtin_neon_vld4_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], __c, 16); r; }) |
| 1959 | #define vld4_lane_u16(__a, b, __c) __extension__ ({ \ |
| 1960 | uint16x4x4_t __b = (b); \ |
| 1961 | uint16x4x4_t r; __builtin_neon_vld4_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], __c, 17); r; }) |
| 1962 | #define vld4_lane_u32(__a, b, __c) __extension__ ({ \ |
| 1963 | uint32x2x4_t __b = (b); \ |
| 1964 | uint32x2x4_t r; __builtin_neon_vld4_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], __c, 18); r; }) |
| 1965 | #define vld4_lane_s8(__a, b, __c) __extension__ ({ \ |
| 1966 | int8x8x4_t __b = (b); \ |
| 1967 | int8x8x4_t r; __builtin_neon_vld4_lane_v(&r, __a, __b.val[0], __b.val[1], __b.val[2], __b.val[3], __c, 0); r; }) |
| 1968 | #define vld4_lane_s16(__a, b, __c) __extension__ ({ \ |
| 1969 | int16x4x4_t __b = (b); \ |
| 1970 | int16x4x4_t r; __builtin_neon_vld4_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], __c, 1); r; }) |
| 1971 | #define vld4_lane_s32(__a, b, __c) __extension__ ({ \ |
| 1972 | int32x2x4_t __b = (b); \ |
| 1973 | int32x2x4_t r; __builtin_neon_vld4_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], __c, 2); r; }) |
| 1974 | #define vld4_lane_f16(__a, b, __c) __extension__ ({ \ |
| 1975 | float16x4x4_t __b = (b); \ |
| 1976 | float16x4x4_t r; __builtin_neon_vld4_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], __c, 7); r; }) |
| 1977 | #define vld4_lane_f32(__a, b, __c) __extension__ ({ \ |
| 1978 | float32x2x4_t __b = (b); \ |
| 1979 | float32x2x4_t r; __builtin_neon_vld4_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], __c, 8); r; }) |
| 1980 | #define vld4_lane_p8(__a, b, __c) __extension__ ({ \ |
| 1981 | poly8x8x4_t __b = (b); \ |
| 1982 | poly8x8x4_t r; __builtin_neon_vld4_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], __c, 4); r; }) |
| 1983 | #define vld4_lane_p16(__a, b, __c) __extension__ ({ \ |
| 1984 | poly16x4x4_t __b = (b); \ |
| 1985 | poly16x4x4_t r; __builtin_neon_vld4_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], __c, 5); r; }) |
| 1986 | |
| 1987 | __ai int8x8_t vmax_s8(int8x8_t __a, int8x8_t __b) { |
| 1988 | return (int8x8_t)__builtin_neon_vmax_v(__a, __b, 0); } |
| 1989 | __ai int16x4_t vmax_s16(int16x4_t __a, int16x4_t __b) { |
| 1990 | return (int16x4_t)__builtin_neon_vmax_v((int8x8_t)__a, (int8x8_t)__b, 1); } |
| 1991 | __ai int32x2_t vmax_s32(int32x2_t __a, int32x2_t __b) { |
| 1992 | return (int32x2_t)__builtin_neon_vmax_v((int8x8_t)__a, (int8x8_t)__b, 2); } |
| 1993 | __ai uint8x8_t vmax_u8(uint8x8_t __a, uint8x8_t __b) { |
| 1994 | return (uint8x8_t)__builtin_neon_vmax_v((int8x8_t)__a, (int8x8_t)__b, 16); } |
| 1995 | __ai uint16x4_t vmax_u16(uint16x4_t __a, uint16x4_t __b) { |
| 1996 | return (uint16x4_t)__builtin_neon_vmax_v((int8x8_t)__a, (int8x8_t)__b, 17); } |
| 1997 | __ai uint32x2_t vmax_u32(uint32x2_t __a, uint32x2_t __b) { |
| 1998 | return (uint32x2_t)__builtin_neon_vmax_v((int8x8_t)__a, (int8x8_t)__b, 18); } |
| 1999 | __ai float32x2_t vmax_f32(float32x2_t __a, float32x2_t __b) { |
| 2000 | return (float32x2_t)__builtin_neon_vmax_v((int8x8_t)__a, (int8x8_t)__b, 8); } |
| 2001 | __ai int8x16_t vmaxq_s8(int8x16_t __a, int8x16_t __b) { |
| 2002 | return (int8x16_t)__builtin_neon_vmaxq_v(__a, __b, 32); } |
| 2003 | __ai int16x8_t vmaxq_s16(int16x8_t __a, int16x8_t __b) { |
| 2004 | return (int16x8_t)__builtin_neon_vmaxq_v((int8x16_t)__a, (int8x16_t)__b, 33); } |
| 2005 | __ai int32x4_t vmaxq_s32(int32x4_t __a, int32x4_t __b) { |
| 2006 | return (int32x4_t)__builtin_neon_vmaxq_v((int8x16_t)__a, (int8x16_t)__b, 34); } |
| 2007 | __ai uint8x16_t vmaxq_u8(uint8x16_t __a, uint8x16_t __b) { |
| 2008 | return (uint8x16_t)__builtin_neon_vmaxq_v((int8x16_t)__a, (int8x16_t)__b, 48); } |
| 2009 | __ai uint16x8_t vmaxq_u16(uint16x8_t __a, uint16x8_t __b) { |
| 2010 | return (uint16x8_t)__builtin_neon_vmaxq_v((int8x16_t)__a, (int8x16_t)__b, 49); } |
| 2011 | __ai uint32x4_t vmaxq_u32(uint32x4_t __a, uint32x4_t __b) { |
| 2012 | return (uint32x4_t)__builtin_neon_vmaxq_v((int8x16_t)__a, (int8x16_t)__b, 50); } |
| 2013 | __ai float32x4_t vmaxq_f32(float32x4_t __a, float32x4_t __b) { |
| 2014 | return (float32x4_t)__builtin_neon_vmaxq_v((int8x16_t)__a, (int8x16_t)__b, 40); } |
| 2015 | |
| 2016 | __ai int8x8_t vmin_s8(int8x8_t __a, int8x8_t __b) { |
| 2017 | return (int8x8_t)__builtin_neon_vmin_v(__a, __b, 0); } |
| 2018 | __ai int16x4_t vmin_s16(int16x4_t __a, int16x4_t __b) { |
| 2019 | return (int16x4_t)__builtin_neon_vmin_v((int8x8_t)__a, (int8x8_t)__b, 1); } |
| 2020 | __ai int32x2_t vmin_s32(int32x2_t __a, int32x2_t __b) { |
| 2021 | return (int32x2_t)__builtin_neon_vmin_v((int8x8_t)__a, (int8x8_t)__b, 2); } |
| 2022 | __ai uint8x8_t vmin_u8(uint8x8_t __a, uint8x8_t __b) { |
| 2023 | return (uint8x8_t)__builtin_neon_vmin_v((int8x8_t)__a, (int8x8_t)__b, 16); } |
| 2024 | __ai uint16x4_t vmin_u16(uint16x4_t __a, uint16x4_t __b) { |
| 2025 | return (uint16x4_t)__builtin_neon_vmin_v((int8x8_t)__a, (int8x8_t)__b, 17); } |
| 2026 | __ai uint32x2_t vmin_u32(uint32x2_t __a, uint32x2_t __b) { |
| 2027 | return (uint32x2_t)__builtin_neon_vmin_v((int8x8_t)__a, (int8x8_t)__b, 18); } |
| 2028 | __ai float32x2_t vmin_f32(float32x2_t __a, float32x2_t __b) { |
| 2029 | return (float32x2_t)__builtin_neon_vmin_v((int8x8_t)__a, (int8x8_t)__b, 8); } |
| 2030 | __ai int8x16_t vminq_s8(int8x16_t __a, int8x16_t __b) { |
| 2031 | return (int8x16_t)__builtin_neon_vminq_v(__a, __b, 32); } |
| 2032 | __ai int16x8_t vminq_s16(int16x8_t __a, int16x8_t __b) { |
| 2033 | return (int16x8_t)__builtin_neon_vminq_v((int8x16_t)__a, (int8x16_t)__b, 33); } |
| 2034 | __ai int32x4_t vminq_s32(int32x4_t __a, int32x4_t __b) { |
| 2035 | return (int32x4_t)__builtin_neon_vminq_v((int8x16_t)__a, (int8x16_t)__b, 34); } |
| 2036 | __ai uint8x16_t vminq_u8(uint8x16_t __a, uint8x16_t __b) { |
| 2037 | return (uint8x16_t)__builtin_neon_vminq_v((int8x16_t)__a, (int8x16_t)__b, 48); } |
| 2038 | __ai uint16x8_t vminq_u16(uint16x8_t __a, uint16x8_t __b) { |
| 2039 | return (uint16x8_t)__builtin_neon_vminq_v((int8x16_t)__a, (int8x16_t)__b, 49); } |
| 2040 | __ai uint32x4_t vminq_u32(uint32x4_t __a, uint32x4_t __b) { |
| 2041 | return (uint32x4_t)__builtin_neon_vminq_v((int8x16_t)__a, (int8x16_t)__b, 50); } |
| 2042 | __ai float32x4_t vminq_f32(float32x4_t __a, float32x4_t __b) { |
| 2043 | return (float32x4_t)__builtin_neon_vminq_v((int8x16_t)__a, (int8x16_t)__b, 40); } |
| 2044 | |
| 2045 | __ai int8x8_t vmla_s8(int8x8_t __a, int8x8_t __b, int8x8_t __c) { |
| 2046 | return __a + (__b * __c); } |
| 2047 | __ai int16x4_t vmla_s16(int16x4_t __a, int16x4_t __b, int16x4_t __c) { |
| 2048 | return __a + (__b * __c); } |
| 2049 | __ai int32x2_t vmla_s32(int32x2_t __a, int32x2_t __b, int32x2_t __c) { |
| 2050 | return __a + (__b * __c); } |
| 2051 | __ai float32x2_t vmla_f32(float32x2_t __a, float32x2_t __b, float32x2_t __c) { |
| 2052 | return __a + (__b * __c); } |
| 2053 | __ai uint8x8_t vmla_u8(uint8x8_t __a, uint8x8_t __b, uint8x8_t __c) { |
| 2054 | return __a + (__b * __c); } |
| 2055 | __ai uint16x4_t vmla_u16(uint16x4_t __a, uint16x4_t __b, uint16x4_t __c) { |
| 2056 | return __a + (__b * __c); } |
| 2057 | __ai uint32x2_t vmla_u32(uint32x2_t __a, uint32x2_t __b, uint32x2_t __c) { |
| 2058 | return __a + (__b * __c); } |
| 2059 | __ai int8x16_t vmlaq_s8(int8x16_t __a, int8x16_t __b, int8x16_t __c) { |
| 2060 | return __a + (__b * __c); } |
| 2061 | __ai int16x8_t vmlaq_s16(int16x8_t __a, int16x8_t __b, int16x8_t __c) { |
| 2062 | return __a + (__b * __c); } |
| 2063 | __ai int32x4_t vmlaq_s32(int32x4_t __a, int32x4_t __b, int32x4_t __c) { |
| 2064 | return __a + (__b * __c); } |
| 2065 | __ai float32x4_t vmlaq_f32(float32x4_t __a, float32x4_t __b, float32x4_t __c) { |
| 2066 | return __a + (__b * __c); } |
| 2067 | __ai uint8x16_t vmlaq_u8(uint8x16_t __a, uint8x16_t __b, uint8x16_t __c) { |
| 2068 | return __a + (__b * __c); } |
| 2069 | __ai uint16x8_t vmlaq_u16(uint16x8_t __a, uint16x8_t __b, uint16x8_t __c) { |
| 2070 | return __a + (__b * __c); } |
| 2071 | __ai uint32x4_t vmlaq_u32(uint32x4_t __a, uint32x4_t __b, uint32x4_t __c) { |
| 2072 | return __a + (__b * __c); } |
| 2073 | |
| 2074 | __ai int16x8_t vmlal_s8(int16x8_t __a, int8x8_t __b, int8x8_t __c) { |
| 2075 | return __a + vmull_s8(__b, __c); } |
| 2076 | __ai int32x4_t vmlal_s16(int32x4_t __a, int16x4_t __b, int16x4_t __c) { |
| 2077 | return __a + vmull_s16(__b, __c); } |
| 2078 | __ai int64x2_t vmlal_s32(int64x2_t __a, int32x2_t __b, int32x2_t __c) { |
| 2079 | return __a + vmull_s32(__b, __c); } |
| 2080 | __ai uint16x8_t vmlal_u8(uint16x8_t __a, uint8x8_t __b, uint8x8_t __c) { |
| 2081 | return __a + vmull_u8(__b, __c); } |
| 2082 | __ai uint32x4_t vmlal_u16(uint32x4_t __a, uint16x4_t __b, uint16x4_t __c) { |
| 2083 | return __a + vmull_u16(__b, __c); } |
| 2084 | __ai uint64x2_t vmlal_u32(uint64x2_t __a, uint32x2_t __b, uint32x2_t __c) { |
| 2085 | return __a + vmull_u32(__b, __c); } |
| 2086 | |
| 2087 | #define vmlal_lane_s16(a, b, c, __d) __extension__ ({ \ |
| 2088 | int32x4_t __a = (a); int16x4_t __b = (b); int16x4_t __c = (c); \ |
| 2089 | __a + vmull_s16(__b, __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 2090 | #define vmlal_lane_s32(a, b, c, __d) __extension__ ({ \ |
| 2091 | int64x2_t __a = (a); int32x2_t __b = (b); int32x2_t __c = (c); \ |
| 2092 | __a + vmull_s32(__b, __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 2093 | #define vmlal_lane_u16(a, b, c, __d) __extension__ ({ \ |
| 2094 | uint32x4_t __a = (a); uint16x4_t __b = (b); uint16x4_t __c = (c); \ |
| 2095 | __a + vmull_u16(__b, __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 2096 | #define vmlal_lane_u32(a, b, c, __d) __extension__ ({ \ |
| 2097 | uint64x2_t __a = (a); uint32x2_t __b = (b); uint32x2_t __c = (c); \ |
| 2098 | __a + vmull_u32(__b, __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 2099 | |
| 2100 | __ai int32x4_t vmlal_n_s16(int32x4_t __a, int16x4_t __b, int16_t __c) { |
| 2101 | return __a + vmull_s16(__b, (int16x4_t){ __c, __c, __c, __c }); } |
| 2102 | __ai int64x2_t vmlal_n_s32(int64x2_t __a, int32x2_t __b, int32_t __c) { |
| 2103 | return __a + vmull_s32(__b, (int32x2_t){ __c, __c }); } |
| 2104 | __ai uint32x4_t vmlal_n_u16(uint32x4_t __a, uint16x4_t __b, uint16_t __c) { |
| 2105 | return __a + vmull_u16(__b, (uint16x4_t){ __c, __c, __c, __c }); } |
| 2106 | __ai uint64x2_t vmlal_n_u32(uint64x2_t __a, uint32x2_t __b, uint32_t __c) { |
| 2107 | return __a + vmull_u32(__b, (uint32x2_t){ __c, __c }); } |
| 2108 | |
| 2109 | #define vmla_lane_s16(a, b, c, __d) __extension__ ({ \ |
| 2110 | int16x4_t __a = (a); int16x4_t __b = (b); int16x4_t __c = (c); \ |
| 2111 | __a + (__b * __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 2112 | #define vmla_lane_s32(a, b, c, __d) __extension__ ({ \ |
| 2113 | int32x2_t __a = (a); int32x2_t __b = (b); int32x2_t __c = (c); \ |
| 2114 | __a + (__b * __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 2115 | #define vmla_lane_u16(a, b, c, __d) __extension__ ({ \ |
| 2116 | uint16x4_t __a = (a); uint16x4_t __b = (b); uint16x4_t __c = (c); \ |
| 2117 | __a + (__b * __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 2118 | #define vmla_lane_u32(a, b, c, __d) __extension__ ({ \ |
| 2119 | uint32x2_t __a = (a); uint32x2_t __b = (b); uint32x2_t __c = (c); \ |
| 2120 | __a + (__b * __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 2121 | #define vmla_lane_f32(a, b, c, __d) __extension__ ({ \ |
| 2122 | float32x2_t __a = (a); float32x2_t __b = (b); float32x2_t __c = (c); \ |
| 2123 | __a + (__b * __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 2124 | #define vmlaq_lane_s16(a, b, c, __d) __extension__ ({ \ |
| 2125 | int16x8_t __a = (a); int16x8_t __b = (b); int16x4_t __c = (c); \ |
| 2126 | __a + (__b * __builtin_shufflevector(__c, __c, __d, __d, __d, __d, __d, __d, __d, __d)); }) |
| 2127 | #define vmlaq_lane_s32(a, b, c, __d) __extension__ ({ \ |
| 2128 | int32x4_t __a = (a); int32x4_t __b = (b); int32x2_t __c = (c); \ |
| 2129 | __a + (__b * __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 2130 | #define vmlaq_lane_u16(a, b, c, __d) __extension__ ({ \ |
| 2131 | uint16x8_t __a = (a); uint16x8_t __b = (b); uint16x4_t __c = (c); \ |
| 2132 | __a + (__b * __builtin_shufflevector(__c, __c, __d, __d, __d, __d, __d, __d, __d, __d)); }) |
| 2133 | #define vmlaq_lane_u32(a, b, c, __d) __extension__ ({ \ |
| 2134 | uint32x4_t __a = (a); uint32x4_t __b = (b); uint32x2_t __c = (c); \ |
| 2135 | __a + (__b * __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 2136 | #define vmlaq_lane_f32(a, b, c, __d) __extension__ ({ \ |
| 2137 | float32x4_t __a = (a); float32x4_t __b = (b); float32x2_t __c = (c); \ |
| 2138 | __a + (__b * __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 2139 | |
| 2140 | __ai int16x4_t vmla_n_s16(int16x4_t __a, int16x4_t __b, int16_t __c) { |
| 2141 | return __a + (__b * (int16x4_t){ __c, __c, __c, __c }); } |
| 2142 | __ai int32x2_t vmla_n_s32(int32x2_t __a, int32x2_t __b, int32_t __c) { |
| 2143 | return __a + (__b * (int32x2_t){ __c, __c }); } |
| 2144 | __ai uint16x4_t vmla_n_u16(uint16x4_t __a, uint16x4_t __b, uint16_t __c) { |
| 2145 | return __a + (__b * (uint16x4_t){ __c, __c, __c, __c }); } |
| 2146 | __ai uint32x2_t vmla_n_u32(uint32x2_t __a, uint32x2_t __b, uint32_t __c) { |
| 2147 | return __a + (__b * (uint32x2_t){ __c, __c }); } |
| 2148 | __ai float32x2_t vmla_n_f32(float32x2_t __a, float32x2_t __b, float32_t __c) { |
| 2149 | return __a + (__b * (float32x2_t){ __c, __c }); } |
| 2150 | __ai int16x8_t vmlaq_n_s16(int16x8_t __a, int16x8_t __b, int16_t __c) { |
| 2151 | return __a + (__b * (int16x8_t){ __c, __c, __c, __c, __c, __c, __c, __c }); } |
| 2152 | __ai int32x4_t vmlaq_n_s32(int32x4_t __a, int32x4_t __b, int32_t __c) { |
| 2153 | return __a + (__b * (int32x4_t){ __c, __c, __c, __c }); } |
| 2154 | __ai uint16x8_t vmlaq_n_u16(uint16x8_t __a, uint16x8_t __b, uint16_t __c) { |
| 2155 | return __a + (__b * (uint16x8_t){ __c, __c, __c, __c, __c, __c, __c, __c }); } |
| 2156 | __ai uint32x4_t vmlaq_n_u32(uint32x4_t __a, uint32x4_t __b, uint32_t __c) { |
| 2157 | return __a + (__b * (uint32x4_t){ __c, __c, __c, __c }); } |
| 2158 | __ai float32x4_t vmlaq_n_f32(float32x4_t __a, float32x4_t __b, float32_t __c) { |
| 2159 | return __a + (__b * (float32x4_t){ __c, __c, __c, __c }); } |
| 2160 | |
| 2161 | __ai int8x8_t vmls_s8(int8x8_t __a, int8x8_t __b, int8x8_t __c) { |
| 2162 | return __a - (__b * __c); } |
| 2163 | __ai int16x4_t vmls_s16(int16x4_t __a, int16x4_t __b, int16x4_t __c) { |
| 2164 | return __a - (__b * __c); } |
| 2165 | __ai int32x2_t vmls_s32(int32x2_t __a, int32x2_t __b, int32x2_t __c) { |
| 2166 | return __a - (__b * __c); } |
| 2167 | __ai float32x2_t vmls_f32(float32x2_t __a, float32x2_t __b, float32x2_t __c) { |
| 2168 | return __a - (__b * __c); } |
| 2169 | __ai uint8x8_t vmls_u8(uint8x8_t __a, uint8x8_t __b, uint8x8_t __c) { |
| 2170 | return __a - (__b * __c); } |
| 2171 | __ai uint16x4_t vmls_u16(uint16x4_t __a, uint16x4_t __b, uint16x4_t __c) { |
| 2172 | return __a - (__b * __c); } |
| 2173 | __ai uint32x2_t vmls_u32(uint32x2_t __a, uint32x2_t __b, uint32x2_t __c) { |
| 2174 | return __a - (__b * __c); } |
| 2175 | __ai int8x16_t vmlsq_s8(int8x16_t __a, int8x16_t __b, int8x16_t __c) { |
| 2176 | return __a - (__b * __c); } |
| 2177 | __ai int16x8_t vmlsq_s16(int16x8_t __a, int16x8_t __b, int16x8_t __c) { |
| 2178 | return __a - (__b * __c); } |
| 2179 | __ai int32x4_t vmlsq_s32(int32x4_t __a, int32x4_t __b, int32x4_t __c) { |
| 2180 | return __a - (__b * __c); } |
| 2181 | __ai float32x4_t vmlsq_f32(float32x4_t __a, float32x4_t __b, float32x4_t __c) { |
| 2182 | return __a - (__b * __c); } |
| 2183 | __ai uint8x16_t vmlsq_u8(uint8x16_t __a, uint8x16_t __b, uint8x16_t __c) { |
| 2184 | return __a - (__b * __c); } |
| 2185 | __ai uint16x8_t vmlsq_u16(uint16x8_t __a, uint16x8_t __b, uint16x8_t __c) { |
| 2186 | return __a - (__b * __c); } |
| 2187 | __ai uint32x4_t vmlsq_u32(uint32x4_t __a, uint32x4_t __b, uint32x4_t __c) { |
| 2188 | return __a - (__b * __c); } |
| 2189 | |
| 2190 | __ai int16x8_t vmlsl_s8(int16x8_t __a, int8x8_t __b, int8x8_t __c) { |
| 2191 | return __a - vmull_s8(__b, __c); } |
| 2192 | __ai int32x4_t vmlsl_s16(int32x4_t __a, int16x4_t __b, int16x4_t __c) { |
| 2193 | return __a - vmull_s16(__b, __c); } |
| 2194 | __ai int64x2_t vmlsl_s32(int64x2_t __a, int32x2_t __b, int32x2_t __c) { |
| 2195 | return __a - vmull_s32(__b, __c); } |
| 2196 | __ai uint16x8_t vmlsl_u8(uint16x8_t __a, uint8x8_t __b, uint8x8_t __c) { |
| 2197 | return __a - vmull_u8(__b, __c); } |
| 2198 | __ai uint32x4_t vmlsl_u16(uint32x4_t __a, uint16x4_t __b, uint16x4_t __c) { |
| 2199 | return __a - vmull_u16(__b, __c); } |
| 2200 | __ai uint64x2_t vmlsl_u32(uint64x2_t __a, uint32x2_t __b, uint32x2_t __c) { |
| 2201 | return __a - vmull_u32(__b, __c); } |
| 2202 | |
| 2203 | #define vmlsl_lane_s16(a, b, c, __d) __extension__ ({ \ |
| 2204 | int32x4_t __a = (a); int16x4_t __b = (b); int16x4_t __c = (c); \ |
| 2205 | __a - vmull_s16(__b, __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 2206 | #define vmlsl_lane_s32(a, b, c, __d) __extension__ ({ \ |
| 2207 | int64x2_t __a = (a); int32x2_t __b = (b); int32x2_t __c = (c); \ |
| 2208 | __a - vmull_s32(__b, __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 2209 | #define vmlsl_lane_u16(a, b, c, __d) __extension__ ({ \ |
| 2210 | uint32x4_t __a = (a); uint16x4_t __b = (b); uint16x4_t __c = (c); \ |
| 2211 | __a - vmull_u16(__b, __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 2212 | #define vmlsl_lane_u32(a, b, c, __d) __extension__ ({ \ |
| 2213 | uint64x2_t __a = (a); uint32x2_t __b = (b); uint32x2_t __c = (c); \ |
| 2214 | __a - vmull_u32(__b, __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 2215 | |
| 2216 | __ai int32x4_t vmlsl_n_s16(int32x4_t __a, int16x4_t __b, int16_t __c) { |
| 2217 | return __a - vmull_s16(__b, (int16x4_t){ __c, __c, __c, __c }); } |
| 2218 | __ai int64x2_t vmlsl_n_s32(int64x2_t __a, int32x2_t __b, int32_t __c) { |
| 2219 | return __a - vmull_s32(__b, (int32x2_t){ __c, __c }); } |
| 2220 | __ai uint32x4_t vmlsl_n_u16(uint32x4_t __a, uint16x4_t __b, uint16_t __c) { |
| 2221 | return __a - vmull_u16(__b, (uint16x4_t){ __c, __c, __c, __c }); } |
| 2222 | __ai uint64x2_t vmlsl_n_u32(uint64x2_t __a, uint32x2_t __b, uint32_t __c) { |
| 2223 | return __a - vmull_u32(__b, (uint32x2_t){ __c, __c }); } |
| 2224 | |
| 2225 | #define vmls_lane_s16(a, b, c, __d) __extension__ ({ \ |
| 2226 | int16x4_t __a = (a); int16x4_t __b = (b); int16x4_t __c = (c); \ |
| 2227 | __a - (__b * __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 2228 | #define vmls_lane_s32(a, b, c, __d) __extension__ ({ \ |
| 2229 | int32x2_t __a = (a); int32x2_t __b = (b); int32x2_t __c = (c); \ |
| 2230 | __a - (__b * __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 2231 | #define vmls_lane_u16(a, b, c, __d) __extension__ ({ \ |
| 2232 | uint16x4_t __a = (a); uint16x4_t __b = (b); uint16x4_t __c = (c); \ |
| 2233 | __a - (__b * __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 2234 | #define vmls_lane_u32(a, b, c, __d) __extension__ ({ \ |
| 2235 | uint32x2_t __a = (a); uint32x2_t __b = (b); uint32x2_t __c = (c); \ |
| 2236 | __a - (__b * __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 2237 | #define vmls_lane_f32(a, b, c, __d) __extension__ ({ \ |
| 2238 | float32x2_t __a = (a); float32x2_t __b = (b); float32x2_t __c = (c); \ |
| 2239 | __a - (__b * __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 2240 | #define vmlsq_lane_s16(a, b, c, __d) __extension__ ({ \ |
| 2241 | int16x8_t __a = (a); int16x8_t __b = (b); int16x4_t __c = (c); \ |
| 2242 | __a - (__b * __builtin_shufflevector(__c, __c, __d, __d, __d, __d, __d, __d, __d, __d)); }) |
| 2243 | #define vmlsq_lane_s32(a, b, c, __d) __extension__ ({ \ |
| 2244 | int32x4_t __a = (a); int32x4_t __b = (b); int32x2_t __c = (c); \ |
| 2245 | __a - (__b * __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 2246 | #define vmlsq_lane_u16(a, b, c, __d) __extension__ ({ \ |
| 2247 | uint16x8_t __a = (a); uint16x8_t __b = (b); uint16x4_t __c = (c); \ |
| 2248 | __a - (__b * __builtin_shufflevector(__c, __c, __d, __d, __d, __d, __d, __d, __d, __d)); }) |
| 2249 | #define vmlsq_lane_u32(a, b, c, __d) __extension__ ({ \ |
| 2250 | uint32x4_t __a = (a); uint32x4_t __b = (b); uint32x2_t __c = (c); \ |
| 2251 | __a - (__b * __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 2252 | #define vmlsq_lane_f32(a, b, c, __d) __extension__ ({ \ |
| 2253 | float32x4_t __a = (a); float32x4_t __b = (b); float32x2_t __c = (c); \ |
| 2254 | __a - (__b * __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 2255 | |
| 2256 | __ai int16x4_t vmls_n_s16(int16x4_t __a, int16x4_t __b, int16_t __c) { |
| 2257 | return __a - (__b * (int16x4_t){ __c, __c, __c, __c }); } |
| 2258 | __ai int32x2_t vmls_n_s32(int32x2_t __a, int32x2_t __b, int32_t __c) { |
| 2259 | return __a - (__b * (int32x2_t){ __c, __c }); } |
| 2260 | __ai uint16x4_t vmls_n_u16(uint16x4_t __a, uint16x4_t __b, uint16_t __c) { |
| 2261 | return __a - (__b * (uint16x4_t){ __c, __c, __c, __c }); } |
| 2262 | __ai uint32x2_t vmls_n_u32(uint32x2_t __a, uint32x2_t __b, uint32_t __c) { |
| 2263 | return __a - (__b * (uint32x2_t){ __c, __c }); } |
| 2264 | __ai float32x2_t vmls_n_f32(float32x2_t __a, float32x2_t __b, float32_t __c) { |
| 2265 | return __a - (__b * (float32x2_t){ __c, __c }); } |
| 2266 | __ai int16x8_t vmlsq_n_s16(int16x8_t __a, int16x8_t __b, int16_t __c) { |
| 2267 | return __a - (__b * (int16x8_t){ __c, __c, __c, __c, __c, __c, __c, __c }); } |
| 2268 | __ai int32x4_t vmlsq_n_s32(int32x4_t __a, int32x4_t __b, int32_t __c) { |
| 2269 | return __a - (__b * (int32x4_t){ __c, __c, __c, __c }); } |
| 2270 | __ai uint16x8_t vmlsq_n_u16(uint16x8_t __a, uint16x8_t __b, uint16_t __c) { |
| 2271 | return __a - (__b * (uint16x8_t){ __c, __c, __c, __c, __c, __c, __c, __c }); } |
| 2272 | __ai uint32x4_t vmlsq_n_u32(uint32x4_t __a, uint32x4_t __b, uint32_t __c) { |
| 2273 | return __a - (__b * (uint32x4_t){ __c, __c, __c, __c }); } |
| 2274 | __ai float32x4_t vmlsq_n_f32(float32x4_t __a, float32x4_t __b, float32_t __c) { |
| 2275 | return __a - (__b * (float32x4_t){ __c, __c, __c, __c }); } |
| 2276 | |
| 2277 | __ai int8x8_t vmovn_s16(int16x8_t __a) { |
| 2278 | return (int8x8_t)__builtin_neon_vmovn_v((int8x16_t)__a, 0); } |
| 2279 | __ai int16x4_t vmovn_s32(int32x4_t __a) { |
| 2280 | return (int16x4_t)__builtin_neon_vmovn_v((int8x16_t)__a, 1); } |
| 2281 | __ai int32x2_t vmovn_s64(int64x2_t __a) { |
| 2282 | return (int32x2_t)__builtin_neon_vmovn_v((int8x16_t)__a, 2); } |
| 2283 | __ai uint8x8_t vmovn_u16(uint16x8_t __a) { |
| 2284 | return (uint8x8_t)__builtin_neon_vmovn_v((int8x16_t)__a, 16); } |
| 2285 | __ai uint16x4_t vmovn_u32(uint32x4_t __a) { |
| 2286 | return (uint16x4_t)__builtin_neon_vmovn_v((int8x16_t)__a, 17); } |
| 2287 | __ai uint32x2_t vmovn_u64(uint64x2_t __a) { |
| 2288 | return (uint32x2_t)__builtin_neon_vmovn_v((int8x16_t)__a, 18); } |
| 2289 | |
| 2290 | __ai uint8x8_t vmov_n_u8(uint8_t __a) { |
| 2291 | return (uint8x8_t){ __a, __a, __a, __a, __a, __a, __a, __a }; } |
| 2292 | __ai uint16x4_t vmov_n_u16(uint16_t __a) { |
| 2293 | return (uint16x4_t){ __a, __a, __a, __a }; } |
| 2294 | __ai uint32x2_t vmov_n_u32(uint32_t __a) { |
| 2295 | return (uint32x2_t){ __a, __a }; } |
| 2296 | __ai int8x8_t vmov_n_s8(int8_t __a) { |
| 2297 | return (int8x8_t){ __a, __a, __a, __a, __a, __a, __a, __a }; } |
| 2298 | __ai int16x4_t vmov_n_s16(int16_t __a) { |
| 2299 | return (int16x4_t){ __a, __a, __a, __a }; } |
| 2300 | __ai int32x2_t vmov_n_s32(int32_t __a) { |
| 2301 | return (int32x2_t){ __a, __a }; } |
| 2302 | __ai poly8x8_t vmov_n_p8(poly8_t __a) { |
| 2303 | return (poly8x8_t){ __a, __a, __a, __a, __a, __a, __a, __a }; } |
| 2304 | __ai poly16x4_t vmov_n_p16(poly16_t __a) { |
| 2305 | return (poly16x4_t){ __a, __a, __a, __a }; } |
| 2306 | __ai float32x2_t vmov_n_f32(float32_t __a) { |
| 2307 | return (float32x2_t){ __a, __a }; } |
| 2308 | __ai uint8x16_t vmovq_n_u8(uint8_t __a) { |
| 2309 | return (uint8x16_t){ __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a }; } |
| 2310 | __ai uint16x8_t vmovq_n_u16(uint16_t __a) { |
| 2311 | return (uint16x8_t){ __a, __a, __a, __a, __a, __a, __a, __a }; } |
| 2312 | __ai uint32x4_t vmovq_n_u32(uint32_t __a) { |
| 2313 | return (uint32x4_t){ __a, __a, __a, __a }; } |
| 2314 | __ai int8x16_t vmovq_n_s8(int8_t __a) { |
| 2315 | return (int8x16_t){ __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a }; } |
| 2316 | __ai int16x8_t vmovq_n_s16(int16_t __a) { |
| 2317 | return (int16x8_t){ __a, __a, __a, __a, __a, __a, __a, __a }; } |
| 2318 | __ai int32x4_t vmovq_n_s32(int32_t __a) { |
| 2319 | return (int32x4_t){ __a, __a, __a, __a }; } |
| 2320 | __ai poly8x16_t vmovq_n_p8(poly8_t __a) { |
| 2321 | return (poly8x16_t){ __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a, __a }; } |
| 2322 | __ai poly16x8_t vmovq_n_p16(poly16_t __a) { |
| 2323 | return (poly16x8_t){ __a, __a, __a, __a, __a, __a, __a, __a }; } |
| 2324 | __ai float32x4_t vmovq_n_f32(float32_t __a) { |
| 2325 | return (float32x4_t){ __a, __a, __a, __a }; } |
| 2326 | __ai int64x1_t vmov_n_s64(int64_t __a) { |
| 2327 | return (int64x1_t){ __a }; } |
| 2328 | __ai uint64x1_t vmov_n_u64(uint64_t __a) { |
| 2329 | return (uint64x1_t){ __a }; } |
| 2330 | __ai int64x2_t vmovq_n_s64(int64_t __a) { |
| 2331 | return (int64x2_t){ __a, __a }; } |
| 2332 | __ai uint64x2_t vmovq_n_u64(uint64_t __a) { |
| 2333 | return (uint64x2_t){ __a, __a }; } |
| 2334 | |
| 2335 | __ai int8x8_t vmul_s8(int8x8_t __a, int8x8_t __b) { |
| 2336 | return __a * __b; } |
| 2337 | __ai int16x4_t vmul_s16(int16x4_t __a, int16x4_t __b) { |
| 2338 | return __a * __b; } |
| 2339 | __ai int32x2_t vmul_s32(int32x2_t __a, int32x2_t __b) { |
| 2340 | return __a * __b; } |
| 2341 | __ai float32x2_t vmul_f32(float32x2_t __a, float32x2_t __b) { |
| 2342 | return __a * __b; } |
| 2343 | __ai uint8x8_t vmul_u8(uint8x8_t __a, uint8x8_t __b) { |
| 2344 | return __a * __b; } |
| 2345 | __ai uint16x4_t vmul_u16(uint16x4_t __a, uint16x4_t __b) { |
| 2346 | return __a * __b; } |
| 2347 | __ai uint32x2_t vmul_u32(uint32x2_t __a, uint32x2_t __b) { |
| 2348 | return __a * __b; } |
| 2349 | __ai int8x16_t vmulq_s8(int8x16_t __a, int8x16_t __b) { |
| 2350 | return __a * __b; } |
| 2351 | __ai int16x8_t vmulq_s16(int16x8_t __a, int16x8_t __b) { |
| 2352 | return __a * __b; } |
| 2353 | __ai int32x4_t vmulq_s32(int32x4_t __a, int32x4_t __b) { |
| 2354 | return __a * __b; } |
| 2355 | __ai float32x4_t vmulq_f32(float32x4_t __a, float32x4_t __b) { |
| 2356 | return __a * __b; } |
| 2357 | __ai uint8x16_t vmulq_u8(uint8x16_t __a, uint8x16_t __b) { |
| 2358 | return __a * __b; } |
| 2359 | __ai uint16x8_t vmulq_u16(uint16x8_t __a, uint16x8_t __b) { |
| 2360 | return __a * __b; } |
| 2361 | __ai uint32x4_t vmulq_u32(uint32x4_t __a, uint32x4_t __b) { |
| 2362 | return __a * __b; } |
| 2363 | |
| 2364 | #define vmull_lane_s16(a, b, __c) __extension__ ({ \ |
| 2365 | int16x4_t __a = (a); int16x4_t __b = (b); \ |
| 2366 | vmull_s16(__a, __builtin_shufflevector(__b, __b, __c, __c, __c, __c)); }) |
| 2367 | #define vmull_lane_s32(a, b, __c) __extension__ ({ \ |
| 2368 | int32x2_t __a = (a); int32x2_t __b = (b); \ |
| 2369 | vmull_s32(__a, __builtin_shufflevector(__b, __b, __c, __c)); }) |
| 2370 | #define vmull_lane_u16(a, b, __c) __extension__ ({ \ |
| 2371 | uint16x4_t __a = (a); uint16x4_t __b = (b); \ |
| 2372 | vmull_u16(__a, __builtin_shufflevector(__b, __b, __c, __c, __c, __c)); }) |
| 2373 | #define vmull_lane_u32(a, b, __c) __extension__ ({ \ |
| 2374 | uint32x2_t __a = (a); uint32x2_t __b = (b); \ |
| 2375 | vmull_u32(__a, __builtin_shufflevector(__b, __b, __c, __c)); }) |
| 2376 | |
| 2377 | __ai int32x4_t vmull_n_s16(int16x4_t __a, int16_t __b) { |
| 2378 | return (int32x4_t)__builtin_neon_vmull_v((int8x8_t)__a, (int8x8_t)(int16x4_t){ __b, __b, __b, __b }, 34); } |
| 2379 | __ai int64x2_t vmull_n_s32(int32x2_t __a, int32_t __b) { |
| 2380 | return (int64x2_t)__builtin_neon_vmull_v((int8x8_t)__a, (int8x8_t)(int32x2_t){ __b, __b }, 35); } |
| 2381 | __ai uint32x4_t vmull_n_u16(uint16x4_t __a, uint16_t __b) { |
| 2382 | return (uint32x4_t)__builtin_neon_vmull_v((int8x8_t)__a, (int8x8_t)(uint16x4_t){ __b, __b, __b, __b }, 50); } |
| 2383 | __ai uint64x2_t vmull_n_u32(uint32x2_t __a, uint32_t __b) { |
| 2384 | return (uint64x2_t)__builtin_neon_vmull_v((int8x8_t)__a, (int8x8_t)(uint32x2_t){ __b, __b }, 51); } |
| 2385 | |
| 2386 | __ai poly8x8_t vmul_p8(poly8x8_t __a, poly8x8_t __b) { |
| 2387 | return (poly8x8_t)__builtin_neon_vmul_v((int8x8_t)__a, (int8x8_t)__b, 4); } |
| 2388 | __ai poly8x16_t vmulq_p8(poly8x16_t __a, poly8x16_t __b) { |
| 2389 | return (poly8x16_t)__builtin_neon_vmulq_v((int8x16_t)__a, (int8x16_t)__b, 36); } |
| 2390 | |
| 2391 | #define vmul_lane_s16(a, b, __c) __extension__ ({ \ |
| 2392 | int16x4_t __a = (a); int16x4_t __b = (b); \ |
| 2393 | __a * __builtin_shufflevector(__b, __b, __c, __c, __c, __c); }) |
| 2394 | #define vmul_lane_s32(a, b, __c) __extension__ ({ \ |
| 2395 | int32x2_t __a = (a); int32x2_t __b = (b); \ |
| 2396 | __a * __builtin_shufflevector(__b, __b, __c, __c); }) |
| 2397 | #define vmul_lane_f32(a, b, __c) __extension__ ({ \ |
| 2398 | float32x2_t __a = (a); float32x2_t __b = (b); \ |
| 2399 | __a * __builtin_shufflevector(__b, __b, __c, __c); }) |
| 2400 | #define vmul_lane_u16(a, b, __c) __extension__ ({ \ |
| 2401 | uint16x4_t __a = (a); uint16x4_t __b = (b); \ |
| 2402 | __a * __builtin_shufflevector(__b, __b, __c, __c, __c, __c); }) |
| 2403 | #define vmul_lane_u32(a, b, __c) __extension__ ({ \ |
| 2404 | uint32x2_t __a = (a); uint32x2_t __b = (b); \ |
| 2405 | __a * __builtin_shufflevector(__b, __b, __c, __c); }) |
| 2406 | #define vmulq_lane_s16(a, b, __c) __extension__ ({ \ |
| 2407 | int16x8_t __a = (a); int16x4_t __b = (b); \ |
| 2408 | __a * __builtin_shufflevector(__b, __b, __c, __c, __c, __c, __c, __c, __c, __c); }) |
| 2409 | #define vmulq_lane_s32(a, b, __c) __extension__ ({ \ |
| 2410 | int32x4_t __a = (a); int32x2_t __b = (b); \ |
| 2411 | __a * __builtin_shufflevector(__b, __b, __c, __c, __c, __c); }) |
| 2412 | #define vmulq_lane_f32(a, b, __c) __extension__ ({ \ |
| 2413 | float32x4_t __a = (a); float32x2_t __b = (b); \ |
| 2414 | __a * __builtin_shufflevector(__b, __b, __c, __c, __c, __c); }) |
| 2415 | #define vmulq_lane_u16(a, b, __c) __extension__ ({ \ |
| 2416 | uint16x8_t __a = (a); uint16x4_t __b = (b); \ |
| 2417 | __a * __builtin_shufflevector(__b, __b, __c, __c, __c, __c, __c, __c, __c, __c); }) |
| 2418 | #define vmulq_lane_u32(a, b, __c) __extension__ ({ \ |
| 2419 | uint32x4_t __a = (a); uint32x2_t __b = (b); \ |
| 2420 | __a * __builtin_shufflevector(__b, __b, __c, __c, __c, __c); }) |
| 2421 | |
| 2422 | __ai int16x4_t vmul_n_s16(int16x4_t __a, int16_t __b) { |
| 2423 | return __a * (int16x4_t){ __b, __b, __b, __b }; } |
| 2424 | __ai int32x2_t vmul_n_s32(int32x2_t __a, int32_t __b) { |
| 2425 | return __a * (int32x2_t){ __b, __b }; } |
| 2426 | __ai float32x2_t vmul_n_f32(float32x2_t __a, float32_t __b) { |
| 2427 | return __a * (float32x2_t){ __b, __b }; } |
| 2428 | __ai uint16x4_t vmul_n_u16(uint16x4_t __a, uint16_t __b) { |
| 2429 | return __a * (uint16x4_t){ __b, __b, __b, __b }; } |
| 2430 | __ai uint32x2_t vmul_n_u32(uint32x2_t __a, uint32_t __b) { |
| 2431 | return __a * (uint32x2_t){ __b, __b }; } |
| 2432 | __ai int16x8_t vmulq_n_s16(int16x8_t __a, int16_t __b) { |
| 2433 | return __a * (int16x8_t){ __b, __b, __b, __b, __b, __b, __b, __b }; } |
| 2434 | __ai int32x4_t vmulq_n_s32(int32x4_t __a, int32_t __b) { |
| 2435 | return __a * (int32x4_t){ __b, __b, __b, __b }; } |
| 2436 | __ai float32x4_t vmulq_n_f32(float32x4_t __a, float32_t __b) { |
| 2437 | return __a * (float32x4_t){ __b, __b, __b, __b }; } |
| 2438 | __ai uint16x8_t vmulq_n_u16(uint16x8_t __a, uint16_t __b) { |
| 2439 | return __a * (uint16x8_t){ __b, __b, __b, __b, __b, __b, __b, __b }; } |
| 2440 | __ai uint32x4_t vmulq_n_u32(uint32x4_t __a, uint32_t __b) { |
| 2441 | return __a * (uint32x4_t){ __b, __b, __b, __b }; } |
| 2442 | |
| 2443 | __ai int8x8_t vmvn_s8(int8x8_t __a) { |
| 2444 | return ~__a; } |
| 2445 | __ai int16x4_t vmvn_s16(int16x4_t __a) { |
| 2446 | return ~__a; } |
| 2447 | __ai int32x2_t vmvn_s32(int32x2_t __a) { |
| 2448 | return ~__a; } |
| 2449 | __ai uint8x8_t vmvn_u8(uint8x8_t __a) { |
| 2450 | return ~__a; } |
| 2451 | __ai uint16x4_t vmvn_u16(uint16x4_t __a) { |
| 2452 | return ~__a; } |
| 2453 | __ai uint32x2_t vmvn_u32(uint32x2_t __a) { |
| 2454 | return ~__a; } |
| 2455 | __ai poly8x8_t vmvn_p8(poly8x8_t __a) { |
| 2456 | return ~__a; } |
| 2457 | __ai int8x16_t vmvnq_s8(int8x16_t __a) { |
| 2458 | return ~__a; } |
| 2459 | __ai int16x8_t vmvnq_s16(int16x8_t __a) { |
| 2460 | return ~__a; } |
| 2461 | __ai int32x4_t vmvnq_s32(int32x4_t __a) { |
| 2462 | return ~__a; } |
| 2463 | __ai uint8x16_t vmvnq_u8(uint8x16_t __a) { |
| 2464 | return ~__a; } |
| 2465 | __ai uint16x8_t vmvnq_u16(uint16x8_t __a) { |
| 2466 | return ~__a; } |
| 2467 | __ai uint32x4_t vmvnq_u32(uint32x4_t __a) { |
| 2468 | return ~__a; } |
| 2469 | __ai poly8x16_t vmvnq_p8(poly8x16_t __a) { |
| 2470 | return ~__a; } |
| 2471 | |
| 2472 | __ai int8x8_t vneg_s8(int8x8_t __a) { |
| 2473 | return -__a; } |
| 2474 | __ai int16x4_t vneg_s16(int16x4_t __a) { |
| 2475 | return -__a; } |
| 2476 | __ai int32x2_t vneg_s32(int32x2_t __a) { |
| 2477 | return -__a; } |
| 2478 | __ai float32x2_t vneg_f32(float32x2_t __a) { |
| 2479 | return -__a; } |
| 2480 | __ai int8x16_t vnegq_s8(int8x16_t __a) { |
| 2481 | return -__a; } |
| 2482 | __ai int16x8_t vnegq_s16(int16x8_t __a) { |
| 2483 | return -__a; } |
| 2484 | __ai int32x4_t vnegq_s32(int32x4_t __a) { |
| 2485 | return -__a; } |
| 2486 | __ai float32x4_t vnegq_f32(float32x4_t __a) { |
| 2487 | return -__a; } |
| 2488 | |
| 2489 | __ai int8x8_t vorn_s8(int8x8_t __a, int8x8_t __b) { |
| 2490 | return __a | ~__b; } |
| 2491 | __ai int16x4_t vorn_s16(int16x4_t __a, int16x4_t __b) { |
| 2492 | return __a | ~__b; } |
| 2493 | __ai int32x2_t vorn_s32(int32x2_t __a, int32x2_t __b) { |
| 2494 | return __a | ~__b; } |
| 2495 | __ai int64x1_t vorn_s64(int64x1_t __a, int64x1_t __b) { |
| 2496 | return __a | ~__b; } |
| 2497 | __ai uint8x8_t vorn_u8(uint8x8_t __a, uint8x8_t __b) { |
| 2498 | return __a | ~__b; } |
| 2499 | __ai uint16x4_t vorn_u16(uint16x4_t __a, uint16x4_t __b) { |
| 2500 | return __a | ~__b; } |
| 2501 | __ai uint32x2_t vorn_u32(uint32x2_t __a, uint32x2_t __b) { |
| 2502 | return __a | ~__b; } |
| 2503 | __ai uint64x1_t vorn_u64(uint64x1_t __a, uint64x1_t __b) { |
| 2504 | return __a | ~__b; } |
| 2505 | __ai int8x16_t vornq_s8(int8x16_t __a, int8x16_t __b) { |
| 2506 | return __a | ~__b; } |
| 2507 | __ai int16x8_t vornq_s16(int16x8_t __a, int16x8_t __b) { |
| 2508 | return __a | ~__b; } |
| 2509 | __ai int32x4_t vornq_s32(int32x4_t __a, int32x4_t __b) { |
| 2510 | return __a | ~__b; } |
| 2511 | __ai int64x2_t vornq_s64(int64x2_t __a, int64x2_t __b) { |
| 2512 | return __a | ~__b; } |
| 2513 | __ai uint8x16_t vornq_u8(uint8x16_t __a, uint8x16_t __b) { |
| 2514 | return __a | ~__b; } |
| 2515 | __ai uint16x8_t vornq_u16(uint16x8_t __a, uint16x8_t __b) { |
| 2516 | return __a | ~__b; } |
| 2517 | __ai uint32x4_t vornq_u32(uint32x4_t __a, uint32x4_t __b) { |
| 2518 | return __a | ~__b; } |
| 2519 | __ai uint64x2_t vornq_u64(uint64x2_t __a, uint64x2_t __b) { |
| 2520 | return __a | ~__b; } |
| 2521 | |
| 2522 | __ai int8x8_t vorr_s8(int8x8_t __a, int8x8_t __b) { |
| 2523 | return __a | __b; } |
| 2524 | __ai int16x4_t vorr_s16(int16x4_t __a, int16x4_t __b) { |
| 2525 | return __a | __b; } |
| 2526 | __ai int32x2_t vorr_s32(int32x2_t __a, int32x2_t __b) { |
| 2527 | return __a | __b; } |
| 2528 | __ai int64x1_t vorr_s64(int64x1_t __a, int64x1_t __b) { |
| 2529 | return __a | __b; } |
| 2530 | __ai uint8x8_t vorr_u8(uint8x8_t __a, uint8x8_t __b) { |
| 2531 | return __a | __b; } |
| 2532 | __ai uint16x4_t vorr_u16(uint16x4_t __a, uint16x4_t __b) { |
| 2533 | return __a | __b; } |
| 2534 | __ai uint32x2_t vorr_u32(uint32x2_t __a, uint32x2_t __b) { |
| 2535 | return __a | __b; } |
| 2536 | __ai uint64x1_t vorr_u64(uint64x1_t __a, uint64x1_t __b) { |
| 2537 | return __a | __b; } |
| 2538 | __ai int8x16_t vorrq_s8(int8x16_t __a, int8x16_t __b) { |
| 2539 | return __a | __b; } |
| 2540 | __ai int16x8_t vorrq_s16(int16x8_t __a, int16x8_t __b) { |
| 2541 | return __a | __b; } |
| 2542 | __ai int32x4_t vorrq_s32(int32x4_t __a, int32x4_t __b) { |
| 2543 | return __a | __b; } |
| 2544 | __ai int64x2_t vorrq_s64(int64x2_t __a, int64x2_t __b) { |
| 2545 | return __a | __b; } |
| 2546 | __ai uint8x16_t vorrq_u8(uint8x16_t __a, uint8x16_t __b) { |
| 2547 | return __a | __b; } |
| 2548 | __ai uint16x8_t vorrq_u16(uint16x8_t __a, uint16x8_t __b) { |
| 2549 | return __a | __b; } |
| 2550 | __ai uint32x4_t vorrq_u32(uint32x4_t __a, uint32x4_t __b) { |
| 2551 | return __a | __b; } |
| 2552 | __ai uint64x2_t vorrq_u64(uint64x2_t __a, uint64x2_t __b) { |
| 2553 | return __a | __b; } |
| 2554 | |
| 2555 | __ai int16x4_t vpadal_s8(int16x4_t __a, int8x8_t __b) { |
| 2556 | return (int16x4_t)__builtin_neon_vpadal_v((int8x8_t)__a, __b, 1); } |
| 2557 | __ai int32x2_t vpadal_s16(int32x2_t __a, int16x4_t __b) { |
| 2558 | return (int32x2_t)__builtin_neon_vpadal_v((int8x8_t)__a, (int8x8_t)__b, 2); } |
| 2559 | __ai int64x1_t vpadal_s32(int64x1_t __a, int32x2_t __b) { |
| 2560 | return (int64x1_t)__builtin_neon_vpadal_v((int8x8_t)__a, (int8x8_t)__b, 3); } |
| 2561 | __ai uint16x4_t vpadal_u8(uint16x4_t __a, uint8x8_t __b) { |
| 2562 | return (uint16x4_t)__builtin_neon_vpadal_v((int8x8_t)__a, (int8x8_t)__b, 17); } |
| 2563 | __ai uint32x2_t vpadal_u16(uint32x2_t __a, uint16x4_t __b) { |
| 2564 | return (uint32x2_t)__builtin_neon_vpadal_v((int8x8_t)__a, (int8x8_t)__b, 18); } |
| 2565 | __ai uint64x1_t vpadal_u32(uint64x1_t __a, uint32x2_t __b) { |
| 2566 | return (uint64x1_t)__builtin_neon_vpadal_v((int8x8_t)__a, (int8x8_t)__b, 19); } |
| 2567 | __ai int16x8_t vpadalq_s8(int16x8_t __a, int8x16_t __b) { |
| 2568 | return (int16x8_t)__builtin_neon_vpadalq_v((int8x16_t)__a, __b, 33); } |
| 2569 | __ai int32x4_t vpadalq_s16(int32x4_t __a, int16x8_t __b) { |
| 2570 | return (int32x4_t)__builtin_neon_vpadalq_v((int8x16_t)__a, (int8x16_t)__b, 34); } |
| 2571 | __ai int64x2_t vpadalq_s32(int64x2_t __a, int32x4_t __b) { |
| 2572 | return (int64x2_t)__builtin_neon_vpadalq_v((int8x16_t)__a, (int8x16_t)__b, 35); } |
| 2573 | __ai uint16x8_t vpadalq_u8(uint16x8_t __a, uint8x16_t __b) { |
| 2574 | return (uint16x8_t)__builtin_neon_vpadalq_v((int8x16_t)__a, (int8x16_t)__b, 49); } |
| 2575 | __ai uint32x4_t vpadalq_u16(uint32x4_t __a, uint16x8_t __b) { |
| 2576 | return (uint32x4_t)__builtin_neon_vpadalq_v((int8x16_t)__a, (int8x16_t)__b, 50); } |
| 2577 | __ai uint64x2_t vpadalq_u32(uint64x2_t __a, uint32x4_t __b) { |
| 2578 | return (uint64x2_t)__builtin_neon_vpadalq_v((int8x16_t)__a, (int8x16_t)__b, 51); } |
| 2579 | |
| 2580 | __ai int8x8_t vpadd_s8(int8x8_t __a, int8x8_t __b) { |
| 2581 | return (int8x8_t)__builtin_neon_vpadd_v(__a, __b, 0); } |
| 2582 | __ai int16x4_t vpadd_s16(int16x4_t __a, int16x4_t __b) { |
| 2583 | return (int16x4_t)__builtin_neon_vpadd_v((int8x8_t)__a, (int8x8_t)__b, 1); } |
| 2584 | __ai int32x2_t vpadd_s32(int32x2_t __a, int32x2_t __b) { |
| 2585 | return (int32x2_t)__builtin_neon_vpadd_v((int8x8_t)__a, (int8x8_t)__b, 2); } |
| 2586 | __ai uint8x8_t vpadd_u8(uint8x8_t __a, uint8x8_t __b) { |
| 2587 | return (uint8x8_t)__builtin_neon_vpadd_v((int8x8_t)__a, (int8x8_t)__b, 16); } |
| 2588 | __ai uint16x4_t vpadd_u16(uint16x4_t __a, uint16x4_t __b) { |
| 2589 | return (uint16x4_t)__builtin_neon_vpadd_v((int8x8_t)__a, (int8x8_t)__b, 17); } |
| 2590 | __ai uint32x2_t vpadd_u32(uint32x2_t __a, uint32x2_t __b) { |
| 2591 | return (uint32x2_t)__builtin_neon_vpadd_v((int8x8_t)__a, (int8x8_t)__b, 18); } |
| 2592 | __ai float32x2_t vpadd_f32(float32x2_t __a, float32x2_t __b) { |
| 2593 | return (float32x2_t)__builtin_neon_vpadd_v((int8x8_t)__a, (int8x8_t)__b, 8); } |
| 2594 | |
| 2595 | __ai int16x4_t vpaddl_s8(int8x8_t __a) { |
| 2596 | return (int16x4_t)__builtin_neon_vpaddl_v(__a, 1); } |
| 2597 | __ai int32x2_t vpaddl_s16(int16x4_t __a) { |
| 2598 | return (int32x2_t)__builtin_neon_vpaddl_v((int8x8_t)__a, 2); } |
| 2599 | __ai int64x1_t vpaddl_s32(int32x2_t __a) { |
| 2600 | return (int64x1_t)__builtin_neon_vpaddl_v((int8x8_t)__a, 3); } |
| 2601 | __ai uint16x4_t vpaddl_u8(uint8x8_t __a) { |
| 2602 | return (uint16x4_t)__builtin_neon_vpaddl_v((int8x8_t)__a, 17); } |
| 2603 | __ai uint32x2_t vpaddl_u16(uint16x4_t __a) { |
| 2604 | return (uint32x2_t)__builtin_neon_vpaddl_v((int8x8_t)__a, 18); } |
| 2605 | __ai uint64x1_t vpaddl_u32(uint32x2_t __a) { |
| 2606 | return (uint64x1_t)__builtin_neon_vpaddl_v((int8x8_t)__a, 19); } |
| 2607 | __ai int16x8_t vpaddlq_s8(int8x16_t __a) { |
| 2608 | return (int16x8_t)__builtin_neon_vpaddlq_v(__a, 33); } |
| 2609 | __ai int32x4_t vpaddlq_s16(int16x8_t __a) { |
| 2610 | return (int32x4_t)__builtin_neon_vpaddlq_v((int8x16_t)__a, 34); } |
| 2611 | __ai int64x2_t vpaddlq_s32(int32x4_t __a) { |
| 2612 | return (int64x2_t)__builtin_neon_vpaddlq_v((int8x16_t)__a, 35); } |
| 2613 | __ai uint16x8_t vpaddlq_u8(uint8x16_t __a) { |
| 2614 | return (uint16x8_t)__builtin_neon_vpaddlq_v((int8x16_t)__a, 49); } |
| 2615 | __ai uint32x4_t vpaddlq_u16(uint16x8_t __a) { |
| 2616 | return (uint32x4_t)__builtin_neon_vpaddlq_v((int8x16_t)__a, 50); } |
| 2617 | __ai uint64x2_t vpaddlq_u32(uint32x4_t __a) { |
| 2618 | return (uint64x2_t)__builtin_neon_vpaddlq_v((int8x16_t)__a, 51); } |
| 2619 | |
| 2620 | __ai int8x8_t vpmax_s8(int8x8_t __a, int8x8_t __b) { |
| 2621 | return (int8x8_t)__builtin_neon_vpmax_v(__a, __b, 0); } |
| 2622 | __ai int16x4_t vpmax_s16(int16x4_t __a, int16x4_t __b) { |
| 2623 | return (int16x4_t)__builtin_neon_vpmax_v((int8x8_t)__a, (int8x8_t)__b, 1); } |
| 2624 | __ai int32x2_t vpmax_s32(int32x2_t __a, int32x2_t __b) { |
| 2625 | return (int32x2_t)__builtin_neon_vpmax_v((int8x8_t)__a, (int8x8_t)__b, 2); } |
| 2626 | __ai uint8x8_t vpmax_u8(uint8x8_t __a, uint8x8_t __b) { |
| 2627 | return (uint8x8_t)__builtin_neon_vpmax_v((int8x8_t)__a, (int8x8_t)__b, 16); } |
| 2628 | __ai uint16x4_t vpmax_u16(uint16x4_t __a, uint16x4_t __b) { |
| 2629 | return (uint16x4_t)__builtin_neon_vpmax_v((int8x8_t)__a, (int8x8_t)__b, 17); } |
| 2630 | __ai uint32x2_t vpmax_u32(uint32x2_t __a, uint32x2_t __b) { |
| 2631 | return (uint32x2_t)__builtin_neon_vpmax_v((int8x8_t)__a, (int8x8_t)__b, 18); } |
| 2632 | __ai float32x2_t vpmax_f32(float32x2_t __a, float32x2_t __b) { |
| 2633 | return (float32x2_t)__builtin_neon_vpmax_v((int8x8_t)__a, (int8x8_t)__b, 8); } |
| 2634 | |
| 2635 | __ai int8x8_t vpmin_s8(int8x8_t __a, int8x8_t __b) { |
| 2636 | return (int8x8_t)__builtin_neon_vpmin_v(__a, __b, 0); } |
| 2637 | __ai int16x4_t vpmin_s16(int16x4_t __a, int16x4_t __b) { |
| 2638 | return (int16x4_t)__builtin_neon_vpmin_v((int8x8_t)__a, (int8x8_t)__b, 1); } |
| 2639 | __ai int32x2_t vpmin_s32(int32x2_t __a, int32x2_t __b) { |
| 2640 | return (int32x2_t)__builtin_neon_vpmin_v((int8x8_t)__a, (int8x8_t)__b, 2); } |
| 2641 | __ai uint8x8_t vpmin_u8(uint8x8_t __a, uint8x8_t __b) { |
| 2642 | return (uint8x8_t)__builtin_neon_vpmin_v((int8x8_t)__a, (int8x8_t)__b, 16); } |
| 2643 | __ai uint16x4_t vpmin_u16(uint16x4_t __a, uint16x4_t __b) { |
| 2644 | return (uint16x4_t)__builtin_neon_vpmin_v((int8x8_t)__a, (int8x8_t)__b, 17); } |
| 2645 | __ai uint32x2_t vpmin_u32(uint32x2_t __a, uint32x2_t __b) { |
| 2646 | return (uint32x2_t)__builtin_neon_vpmin_v((int8x8_t)__a, (int8x8_t)__b, 18); } |
| 2647 | __ai float32x2_t vpmin_f32(float32x2_t __a, float32x2_t __b) { |
| 2648 | return (float32x2_t)__builtin_neon_vpmin_v((int8x8_t)__a, (int8x8_t)__b, 8); } |
| 2649 | |
| 2650 | __ai int8x8_t vqabs_s8(int8x8_t __a) { |
| 2651 | return (int8x8_t)__builtin_neon_vqabs_v(__a, 0); } |
| 2652 | __ai int16x4_t vqabs_s16(int16x4_t __a) { |
| 2653 | return (int16x4_t)__builtin_neon_vqabs_v((int8x8_t)__a, 1); } |
| 2654 | __ai int32x2_t vqabs_s32(int32x2_t __a) { |
| 2655 | return (int32x2_t)__builtin_neon_vqabs_v((int8x8_t)__a, 2); } |
| 2656 | __ai int8x16_t vqabsq_s8(int8x16_t __a) { |
| 2657 | return (int8x16_t)__builtin_neon_vqabsq_v(__a, 32); } |
| 2658 | __ai int16x8_t vqabsq_s16(int16x8_t __a) { |
| 2659 | return (int16x8_t)__builtin_neon_vqabsq_v((int8x16_t)__a, 33); } |
| 2660 | __ai int32x4_t vqabsq_s32(int32x4_t __a) { |
| 2661 | return (int32x4_t)__builtin_neon_vqabsq_v((int8x16_t)__a, 34); } |
| 2662 | |
| 2663 | __ai int8x8_t vqadd_s8(int8x8_t __a, int8x8_t __b) { |
| 2664 | return (int8x8_t)__builtin_neon_vqadd_v(__a, __b, 0); } |
| 2665 | __ai int16x4_t vqadd_s16(int16x4_t __a, int16x4_t __b) { |
| 2666 | return (int16x4_t)__builtin_neon_vqadd_v((int8x8_t)__a, (int8x8_t)__b, 1); } |
| 2667 | __ai int32x2_t vqadd_s32(int32x2_t __a, int32x2_t __b) { |
| 2668 | return (int32x2_t)__builtin_neon_vqadd_v((int8x8_t)__a, (int8x8_t)__b, 2); } |
| 2669 | __ai int64x1_t vqadd_s64(int64x1_t __a, int64x1_t __b) { |
| 2670 | return (int64x1_t)__builtin_neon_vqadd_v((int8x8_t)__a, (int8x8_t)__b, 3); } |
| 2671 | __ai uint8x8_t vqadd_u8(uint8x8_t __a, uint8x8_t __b) { |
| 2672 | return (uint8x8_t)__builtin_neon_vqadd_v((int8x8_t)__a, (int8x8_t)__b, 16); } |
| 2673 | __ai uint16x4_t vqadd_u16(uint16x4_t __a, uint16x4_t __b) { |
| 2674 | return (uint16x4_t)__builtin_neon_vqadd_v((int8x8_t)__a, (int8x8_t)__b, 17); } |
| 2675 | __ai uint32x2_t vqadd_u32(uint32x2_t __a, uint32x2_t __b) { |
| 2676 | return (uint32x2_t)__builtin_neon_vqadd_v((int8x8_t)__a, (int8x8_t)__b, 18); } |
| 2677 | __ai uint64x1_t vqadd_u64(uint64x1_t __a, uint64x1_t __b) { |
| 2678 | return (uint64x1_t)__builtin_neon_vqadd_v((int8x8_t)__a, (int8x8_t)__b, 19); } |
| 2679 | __ai int8x16_t vqaddq_s8(int8x16_t __a, int8x16_t __b) { |
| 2680 | return (int8x16_t)__builtin_neon_vqaddq_v(__a, __b, 32); } |
| 2681 | __ai int16x8_t vqaddq_s16(int16x8_t __a, int16x8_t __b) { |
| 2682 | return (int16x8_t)__builtin_neon_vqaddq_v((int8x16_t)__a, (int8x16_t)__b, 33); } |
| 2683 | __ai int32x4_t vqaddq_s32(int32x4_t __a, int32x4_t __b) { |
| 2684 | return (int32x4_t)__builtin_neon_vqaddq_v((int8x16_t)__a, (int8x16_t)__b, 34); } |
| 2685 | __ai int64x2_t vqaddq_s64(int64x2_t __a, int64x2_t __b) { |
| 2686 | return (int64x2_t)__builtin_neon_vqaddq_v((int8x16_t)__a, (int8x16_t)__b, 35); } |
| 2687 | __ai uint8x16_t vqaddq_u8(uint8x16_t __a, uint8x16_t __b) { |
| 2688 | return (uint8x16_t)__builtin_neon_vqaddq_v((int8x16_t)__a, (int8x16_t)__b, 48); } |
| 2689 | __ai uint16x8_t vqaddq_u16(uint16x8_t __a, uint16x8_t __b) { |
| 2690 | return (uint16x8_t)__builtin_neon_vqaddq_v((int8x16_t)__a, (int8x16_t)__b, 49); } |
| 2691 | __ai uint32x4_t vqaddq_u32(uint32x4_t __a, uint32x4_t __b) { |
| 2692 | return (uint32x4_t)__builtin_neon_vqaddq_v((int8x16_t)__a, (int8x16_t)__b, 50); } |
| 2693 | __ai uint64x2_t vqaddq_u64(uint64x2_t __a, uint64x2_t __b) { |
| 2694 | return (uint64x2_t)__builtin_neon_vqaddq_v((int8x16_t)__a, (int8x16_t)__b, 51); } |
| 2695 | |
| 2696 | __ai int32x4_t vqdmlal_s16(int32x4_t __a, int16x4_t __b, int16x4_t __c) { |
| 2697 | return (int32x4_t)__builtin_neon_vqdmlal_v((int8x16_t)__a, (int8x8_t)__b, (int8x8_t)__c, 34); } |
| 2698 | __ai int64x2_t vqdmlal_s32(int64x2_t __a, int32x2_t __b, int32x2_t __c) { |
| 2699 | return (int64x2_t)__builtin_neon_vqdmlal_v((int8x16_t)__a, (int8x8_t)__b, (int8x8_t)__c, 35); } |
| 2700 | |
| 2701 | #define vqdmlal_lane_s16(a, b, c, __d) __extension__ ({ \ |
| 2702 | int32x4_t __a = (a); int16x4_t __b = (b); int16x4_t __c = (c); \ |
| 2703 | vqdmlal_s16(__a, __b, __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 2704 | #define vqdmlal_lane_s32(a, b, c, __d) __extension__ ({ \ |
| 2705 | int64x2_t __a = (a); int32x2_t __b = (b); int32x2_t __c = (c); \ |
| 2706 | vqdmlal_s32(__a, __b, __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 2707 | |
| 2708 | __ai int32x4_t vqdmlal_n_s16(int32x4_t __a, int16x4_t __b, int16_t __c) { |
| 2709 | return (int32x4_t)__builtin_neon_vqdmlal_v((int8x16_t)__a, (int8x8_t)__b, (int8x8_t)(int16x4_t){ __c, __c, __c, __c }, 34); } |
| 2710 | __ai int64x2_t vqdmlal_n_s32(int64x2_t __a, int32x2_t __b, int32_t __c) { |
| 2711 | return (int64x2_t)__builtin_neon_vqdmlal_v((int8x16_t)__a, (int8x8_t)__b, (int8x8_t)(int32x2_t){ __c, __c }, 35); } |
| 2712 | |
| 2713 | __ai int32x4_t vqdmlsl_s16(int32x4_t __a, int16x4_t __b, int16x4_t __c) { |
| 2714 | return (int32x4_t)__builtin_neon_vqdmlsl_v((int8x16_t)__a, (int8x8_t)__b, (int8x8_t)__c, 34); } |
| 2715 | __ai int64x2_t vqdmlsl_s32(int64x2_t __a, int32x2_t __b, int32x2_t __c) { |
| 2716 | return (int64x2_t)__builtin_neon_vqdmlsl_v((int8x16_t)__a, (int8x8_t)__b, (int8x8_t)__c, 35); } |
| 2717 | |
| 2718 | #define vqdmlsl_lane_s16(a, b, c, __d) __extension__ ({ \ |
| 2719 | int32x4_t __a = (a); int16x4_t __b = (b); int16x4_t __c = (c); \ |
| 2720 | vqdmlsl_s16(__a, __b, __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 2721 | #define vqdmlsl_lane_s32(a, b, c, __d) __extension__ ({ \ |
| 2722 | int64x2_t __a = (a); int32x2_t __b = (b); int32x2_t __c = (c); \ |
| 2723 | vqdmlsl_s32(__a, __b, __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 2724 | |
| 2725 | __ai int32x4_t vqdmlsl_n_s16(int32x4_t __a, int16x4_t __b, int16_t __c) { |
| 2726 | return (int32x4_t)__builtin_neon_vqdmlsl_v((int8x16_t)__a, (int8x8_t)__b, (int8x8_t)(int16x4_t){ __c, __c, __c, __c }, 34); } |
| 2727 | __ai int64x2_t vqdmlsl_n_s32(int64x2_t __a, int32x2_t __b, int32_t __c) { |
| 2728 | return (int64x2_t)__builtin_neon_vqdmlsl_v((int8x16_t)__a, (int8x8_t)__b, (int8x8_t)(int32x2_t){ __c, __c }, 35); } |
| 2729 | |
| 2730 | __ai int16x4_t vqdmulh_s16(int16x4_t __a, int16x4_t __b) { |
| 2731 | return (int16x4_t)__builtin_neon_vqdmulh_v((int8x8_t)__a, (int8x8_t)__b, 1); } |
| 2732 | __ai int32x2_t vqdmulh_s32(int32x2_t __a, int32x2_t __b) { |
| 2733 | return (int32x2_t)__builtin_neon_vqdmulh_v((int8x8_t)__a, (int8x8_t)__b, 2); } |
| 2734 | __ai int16x8_t vqdmulhq_s16(int16x8_t __a, int16x8_t __b) { |
| 2735 | return (int16x8_t)__builtin_neon_vqdmulhq_v((int8x16_t)__a, (int8x16_t)__b, 33); } |
| 2736 | __ai int32x4_t vqdmulhq_s32(int32x4_t __a, int32x4_t __b) { |
| 2737 | return (int32x4_t)__builtin_neon_vqdmulhq_v((int8x16_t)__a, (int8x16_t)__b, 34); } |
| 2738 | |
| 2739 | #define vqdmulh_lane_s16(a, b, __c) __extension__ ({ \ |
| 2740 | int16x4_t __a = (a); int16x4_t __b = (b); \ |
| 2741 | vqdmulh_s16(__a, __builtin_shufflevector(__b, __b, __c, __c, __c, __c)); }) |
| 2742 | #define vqdmulh_lane_s32(a, b, __c) __extension__ ({ \ |
| 2743 | int32x2_t __a = (a); int32x2_t __b = (b); \ |
| 2744 | vqdmulh_s32(__a, __builtin_shufflevector(__b, __b, __c, __c)); }) |
| 2745 | #define vqdmulhq_lane_s16(a, b, __c) __extension__ ({ \ |
| 2746 | int16x8_t __a = (a); int16x4_t __b = (b); \ |
| 2747 | vqdmulhq_s16(__a, __builtin_shufflevector(__b, __b, __c, __c, __c, __c, __c, __c, __c, __c)); }) |
| 2748 | #define vqdmulhq_lane_s32(a, b, __c) __extension__ ({ \ |
| 2749 | int32x4_t __a = (a); int32x2_t __b = (b); \ |
| 2750 | vqdmulhq_s32(__a, __builtin_shufflevector(__b, __b, __c, __c, __c, __c)); }) |
| 2751 | |
| 2752 | __ai int16x4_t vqdmulh_n_s16(int16x4_t __a, int16_t __b) { |
| 2753 | return (int16x4_t)__builtin_neon_vqdmulh_v((int8x8_t)__a, (int8x8_t)(int16x4_t){ __b, __b, __b, __b }, 1); } |
| 2754 | __ai int32x2_t vqdmulh_n_s32(int32x2_t __a, int32_t __b) { |
| 2755 | return (int32x2_t)__builtin_neon_vqdmulh_v((int8x8_t)__a, (int8x8_t)(int32x2_t){ __b, __b }, 2); } |
| 2756 | __ai int16x8_t vqdmulhq_n_s16(int16x8_t __a, int16_t __b) { |
| 2757 | return (int16x8_t)__builtin_neon_vqdmulhq_v((int8x16_t)__a, (int8x16_t)(int16x8_t){ __b, __b, __b, __b, __b, __b, __b, __b }, 33); } |
| 2758 | __ai int32x4_t vqdmulhq_n_s32(int32x4_t __a, int32_t __b) { |
| 2759 | return (int32x4_t)__builtin_neon_vqdmulhq_v((int8x16_t)__a, (int8x16_t)(int32x4_t){ __b, __b, __b, __b }, 34); } |
| 2760 | |
| 2761 | __ai int32x4_t vqdmull_s16(int16x4_t __a, int16x4_t __b) { |
| 2762 | return (int32x4_t)__builtin_neon_vqdmull_v((int8x8_t)__a, (int8x8_t)__b, 34); } |
| 2763 | __ai int64x2_t vqdmull_s32(int32x2_t __a, int32x2_t __b) { |
| 2764 | return (int64x2_t)__builtin_neon_vqdmull_v((int8x8_t)__a, (int8x8_t)__b, 35); } |
| 2765 | |
| 2766 | #define vqdmull_lane_s16(a, b, __c) __extension__ ({ \ |
| 2767 | int16x4_t __a = (a); int16x4_t __b = (b); \ |
| 2768 | vqdmull_s16(__a, __builtin_shufflevector(__b, __b, __c, __c, __c, __c)); }) |
| 2769 | #define vqdmull_lane_s32(a, b, __c) __extension__ ({ \ |
| 2770 | int32x2_t __a = (a); int32x2_t __b = (b); \ |
| 2771 | vqdmull_s32(__a, __builtin_shufflevector(__b, __b, __c, __c)); }) |
| 2772 | |
| 2773 | __ai int32x4_t vqdmull_n_s16(int16x4_t __a, int16_t __b) { |
| 2774 | return (int32x4_t)__builtin_neon_vqdmull_v((int8x8_t)__a, (int8x8_t)(int16x4_t){ __b, __b, __b, __b }, 34); } |
| 2775 | __ai int64x2_t vqdmull_n_s32(int32x2_t __a, int32_t __b) { |
| 2776 | return (int64x2_t)__builtin_neon_vqdmull_v((int8x8_t)__a, (int8x8_t)(int32x2_t){ __b, __b }, 35); } |
| 2777 | |
| 2778 | __ai int8x8_t vqmovn_s16(int16x8_t __a) { |
| 2779 | return (int8x8_t)__builtin_neon_vqmovn_v((int8x16_t)__a, 0); } |
| 2780 | __ai int16x4_t vqmovn_s32(int32x4_t __a) { |
| 2781 | return (int16x4_t)__builtin_neon_vqmovn_v((int8x16_t)__a, 1); } |
| 2782 | __ai int32x2_t vqmovn_s64(int64x2_t __a) { |
| 2783 | return (int32x2_t)__builtin_neon_vqmovn_v((int8x16_t)__a, 2); } |
| 2784 | __ai uint8x8_t vqmovn_u16(uint16x8_t __a) { |
| 2785 | return (uint8x8_t)__builtin_neon_vqmovn_v((int8x16_t)__a, 16); } |
| 2786 | __ai uint16x4_t vqmovn_u32(uint32x4_t __a) { |
| 2787 | return (uint16x4_t)__builtin_neon_vqmovn_v((int8x16_t)__a, 17); } |
| 2788 | __ai uint32x2_t vqmovn_u64(uint64x2_t __a) { |
| 2789 | return (uint32x2_t)__builtin_neon_vqmovn_v((int8x16_t)__a, 18); } |
| 2790 | |
| 2791 | __ai uint8x8_t vqmovun_s16(int16x8_t __a) { |
| 2792 | return (uint8x8_t)__builtin_neon_vqmovun_v((int8x16_t)__a, 16); } |
| 2793 | __ai uint16x4_t vqmovun_s32(int32x4_t __a) { |
| 2794 | return (uint16x4_t)__builtin_neon_vqmovun_v((int8x16_t)__a, 17); } |
| 2795 | __ai uint32x2_t vqmovun_s64(int64x2_t __a) { |
| 2796 | return (uint32x2_t)__builtin_neon_vqmovun_v((int8x16_t)__a, 18); } |
| 2797 | |
| 2798 | __ai int8x8_t vqneg_s8(int8x8_t __a) { |
| 2799 | return (int8x8_t)__builtin_neon_vqneg_v(__a, 0); } |
| 2800 | __ai int16x4_t vqneg_s16(int16x4_t __a) { |
| 2801 | return (int16x4_t)__builtin_neon_vqneg_v((int8x8_t)__a, 1); } |
| 2802 | __ai int32x2_t vqneg_s32(int32x2_t __a) { |
| 2803 | return (int32x2_t)__builtin_neon_vqneg_v((int8x8_t)__a, 2); } |
| 2804 | __ai int8x16_t vqnegq_s8(int8x16_t __a) { |
| 2805 | return (int8x16_t)__builtin_neon_vqnegq_v(__a, 32); } |
| 2806 | __ai int16x8_t vqnegq_s16(int16x8_t __a) { |
| 2807 | return (int16x8_t)__builtin_neon_vqnegq_v((int8x16_t)__a, 33); } |
| 2808 | __ai int32x4_t vqnegq_s32(int32x4_t __a) { |
| 2809 | return (int32x4_t)__builtin_neon_vqnegq_v((int8x16_t)__a, 34); } |
| 2810 | |
| 2811 | __ai int16x4_t vqrdmulh_s16(int16x4_t __a, int16x4_t __b) { |
| 2812 | return (int16x4_t)__builtin_neon_vqrdmulh_v((int8x8_t)__a, (int8x8_t)__b, 1); } |
| 2813 | __ai int32x2_t vqrdmulh_s32(int32x2_t __a, int32x2_t __b) { |
| 2814 | return (int32x2_t)__builtin_neon_vqrdmulh_v((int8x8_t)__a, (int8x8_t)__b, 2); } |
| 2815 | __ai int16x8_t vqrdmulhq_s16(int16x8_t __a, int16x8_t __b) { |
| 2816 | return (int16x8_t)__builtin_neon_vqrdmulhq_v((int8x16_t)__a, (int8x16_t)__b, 33); } |
| 2817 | __ai int32x4_t vqrdmulhq_s32(int32x4_t __a, int32x4_t __b) { |
| 2818 | return (int32x4_t)__builtin_neon_vqrdmulhq_v((int8x16_t)__a, (int8x16_t)__b, 34); } |
| 2819 | |
| 2820 | #define vqrdmulh_lane_s16(a, b, __c) __extension__ ({ \ |
| 2821 | int16x4_t __a = (a); int16x4_t __b = (b); \ |
| 2822 | vqrdmulh_s16(__a, __builtin_shufflevector(__b, __b, __c, __c, __c, __c)); }) |
| 2823 | #define vqrdmulh_lane_s32(a, b, __c) __extension__ ({ \ |
| 2824 | int32x2_t __a = (a); int32x2_t __b = (b); \ |
| 2825 | vqrdmulh_s32(__a, __builtin_shufflevector(__b, __b, __c, __c)); }) |
| 2826 | #define vqrdmulhq_lane_s16(a, b, __c) __extension__ ({ \ |
| 2827 | int16x8_t __a = (a); int16x4_t __b = (b); \ |
| 2828 | vqrdmulhq_s16(__a, __builtin_shufflevector(__b, __b, __c, __c, __c, __c, __c, __c, __c, __c)); }) |
| 2829 | #define vqrdmulhq_lane_s32(a, b, __c) __extension__ ({ \ |
| 2830 | int32x4_t __a = (a); int32x2_t __b = (b); \ |
| 2831 | vqrdmulhq_s32(__a, __builtin_shufflevector(__b, __b, __c, __c, __c, __c)); }) |
| 2832 | |
| 2833 | __ai int16x4_t vqrdmulh_n_s16(int16x4_t __a, int16_t __b) { |
| 2834 | return (int16x4_t)__builtin_neon_vqrdmulh_v((int8x8_t)__a, (int8x8_t)(int16x4_t){ __b, __b, __b, __b }, 1); } |
| 2835 | __ai int32x2_t vqrdmulh_n_s32(int32x2_t __a, int32_t __b) { |
| 2836 | return (int32x2_t)__builtin_neon_vqrdmulh_v((int8x8_t)__a, (int8x8_t)(int32x2_t){ __b, __b }, 2); } |
| 2837 | __ai int16x8_t vqrdmulhq_n_s16(int16x8_t __a, int16_t __b) { |
| 2838 | return (int16x8_t)__builtin_neon_vqrdmulhq_v((int8x16_t)__a, (int8x16_t)(int16x8_t){ __b, __b, __b, __b, __b, __b, __b, __b }, 33); } |
| 2839 | __ai int32x4_t vqrdmulhq_n_s32(int32x4_t __a, int32_t __b) { |
| 2840 | return (int32x4_t)__builtin_neon_vqrdmulhq_v((int8x16_t)__a, (int8x16_t)(int32x4_t){ __b, __b, __b, __b }, 34); } |
| 2841 | |
| 2842 | __ai int8x8_t vqrshl_s8(int8x8_t __a, int8x8_t __b) { |
| 2843 | return (int8x8_t)__builtin_neon_vqrshl_v(__a, __b, 0); } |
| 2844 | __ai int16x4_t vqrshl_s16(int16x4_t __a, int16x4_t __b) { |
| 2845 | return (int16x4_t)__builtin_neon_vqrshl_v((int8x8_t)__a, (int8x8_t)__b, 1); } |
| 2846 | __ai int32x2_t vqrshl_s32(int32x2_t __a, int32x2_t __b) { |
| 2847 | return (int32x2_t)__builtin_neon_vqrshl_v((int8x8_t)__a, (int8x8_t)__b, 2); } |
| 2848 | __ai int64x1_t vqrshl_s64(int64x1_t __a, int64x1_t __b) { |
| 2849 | return (int64x1_t)__builtin_neon_vqrshl_v((int8x8_t)__a, (int8x8_t)__b, 3); } |
| 2850 | __ai uint8x8_t vqrshl_u8(uint8x8_t __a, int8x8_t __b) { |
| 2851 | return (uint8x8_t)__builtin_neon_vqrshl_v((int8x8_t)__a, __b, 16); } |
| 2852 | __ai uint16x4_t vqrshl_u16(uint16x4_t __a, int16x4_t __b) { |
| 2853 | return (uint16x4_t)__builtin_neon_vqrshl_v((int8x8_t)__a, (int8x8_t)__b, 17); } |
| 2854 | __ai uint32x2_t vqrshl_u32(uint32x2_t __a, int32x2_t __b) { |
| 2855 | return (uint32x2_t)__builtin_neon_vqrshl_v((int8x8_t)__a, (int8x8_t)__b, 18); } |
| 2856 | __ai uint64x1_t vqrshl_u64(uint64x1_t __a, int64x1_t __b) { |
| 2857 | return (uint64x1_t)__builtin_neon_vqrshl_v((int8x8_t)__a, (int8x8_t)__b, 19); } |
| 2858 | __ai int8x16_t vqrshlq_s8(int8x16_t __a, int8x16_t __b) { |
| 2859 | return (int8x16_t)__builtin_neon_vqrshlq_v(__a, __b, 32); } |
| 2860 | __ai int16x8_t vqrshlq_s16(int16x8_t __a, int16x8_t __b) { |
| 2861 | return (int16x8_t)__builtin_neon_vqrshlq_v((int8x16_t)__a, (int8x16_t)__b, 33); } |
| 2862 | __ai int32x4_t vqrshlq_s32(int32x4_t __a, int32x4_t __b) { |
| 2863 | return (int32x4_t)__builtin_neon_vqrshlq_v((int8x16_t)__a, (int8x16_t)__b, 34); } |
| 2864 | __ai int64x2_t vqrshlq_s64(int64x2_t __a, int64x2_t __b) { |
| 2865 | return (int64x2_t)__builtin_neon_vqrshlq_v((int8x16_t)__a, (int8x16_t)__b, 35); } |
| 2866 | __ai uint8x16_t vqrshlq_u8(uint8x16_t __a, int8x16_t __b) { |
| 2867 | return (uint8x16_t)__builtin_neon_vqrshlq_v((int8x16_t)__a, __b, 48); } |
| 2868 | __ai uint16x8_t vqrshlq_u16(uint16x8_t __a, int16x8_t __b) { |
| 2869 | return (uint16x8_t)__builtin_neon_vqrshlq_v((int8x16_t)__a, (int8x16_t)__b, 49); } |
| 2870 | __ai uint32x4_t vqrshlq_u32(uint32x4_t __a, int32x4_t __b) { |
| 2871 | return (uint32x4_t)__builtin_neon_vqrshlq_v((int8x16_t)__a, (int8x16_t)__b, 50); } |
| 2872 | __ai uint64x2_t vqrshlq_u64(uint64x2_t __a, int64x2_t __b) { |
| 2873 | return (uint64x2_t)__builtin_neon_vqrshlq_v((int8x16_t)__a, (int8x16_t)__b, 51); } |
| 2874 | |
| 2875 | #define vqrshrn_n_s16(a, __b) __extension__ ({ \ |
| 2876 | int16x8_t __a = (a); \ |
| 2877 | (int8x8_t)__builtin_neon_vqrshrn_n_v((int8x16_t)__a, __b, 0); }) |
| 2878 | #define vqrshrn_n_s32(a, __b) __extension__ ({ \ |
| 2879 | int32x4_t __a = (a); \ |
| 2880 | (int16x4_t)__builtin_neon_vqrshrn_n_v((int8x16_t)__a, __b, 1); }) |
| 2881 | #define vqrshrn_n_s64(a, __b) __extension__ ({ \ |
| 2882 | int64x2_t __a = (a); \ |
| 2883 | (int32x2_t)__builtin_neon_vqrshrn_n_v((int8x16_t)__a, __b, 2); }) |
| 2884 | #define vqrshrn_n_u16(a, __b) __extension__ ({ \ |
| 2885 | uint16x8_t __a = (a); \ |
| 2886 | (uint8x8_t)__builtin_neon_vqrshrn_n_v((int8x16_t)__a, __b, 16); }) |
| 2887 | #define vqrshrn_n_u32(a, __b) __extension__ ({ \ |
| 2888 | uint32x4_t __a = (a); \ |
| 2889 | (uint16x4_t)__builtin_neon_vqrshrn_n_v((int8x16_t)__a, __b, 17); }) |
| 2890 | #define vqrshrn_n_u64(a, __b) __extension__ ({ \ |
| 2891 | uint64x2_t __a = (a); \ |
| 2892 | (uint32x2_t)__builtin_neon_vqrshrn_n_v((int8x16_t)__a, __b, 18); }) |
| 2893 | |
| 2894 | #define vqrshrun_n_s16(a, __b) __extension__ ({ \ |
| 2895 | int16x8_t __a = (a); \ |
| 2896 | (uint8x8_t)__builtin_neon_vqrshrun_n_v((int8x16_t)__a, __b, 16); }) |
| 2897 | #define vqrshrun_n_s32(a, __b) __extension__ ({ \ |
| 2898 | int32x4_t __a = (a); \ |
| 2899 | (uint16x4_t)__builtin_neon_vqrshrun_n_v((int8x16_t)__a, __b, 17); }) |
| 2900 | #define vqrshrun_n_s64(a, __b) __extension__ ({ \ |
| 2901 | int64x2_t __a = (a); \ |
| 2902 | (uint32x2_t)__builtin_neon_vqrshrun_n_v((int8x16_t)__a, __b, 18); }) |
| 2903 | |
| 2904 | __ai int8x8_t vqshl_s8(int8x8_t __a, int8x8_t __b) { |
| 2905 | return (int8x8_t)__builtin_neon_vqshl_v(__a, __b, 0); } |
| 2906 | __ai int16x4_t vqshl_s16(int16x4_t __a, int16x4_t __b) { |
| 2907 | return (int16x4_t)__builtin_neon_vqshl_v((int8x8_t)__a, (int8x8_t)__b, 1); } |
| 2908 | __ai int32x2_t vqshl_s32(int32x2_t __a, int32x2_t __b) { |
| 2909 | return (int32x2_t)__builtin_neon_vqshl_v((int8x8_t)__a, (int8x8_t)__b, 2); } |
| 2910 | __ai int64x1_t vqshl_s64(int64x1_t __a, int64x1_t __b) { |
| 2911 | return (int64x1_t)__builtin_neon_vqshl_v((int8x8_t)__a, (int8x8_t)__b, 3); } |
| 2912 | __ai uint8x8_t vqshl_u8(uint8x8_t __a, int8x8_t __b) { |
| 2913 | return (uint8x8_t)__builtin_neon_vqshl_v((int8x8_t)__a, __b, 16); } |
| 2914 | __ai uint16x4_t vqshl_u16(uint16x4_t __a, int16x4_t __b) { |
| 2915 | return (uint16x4_t)__builtin_neon_vqshl_v((int8x8_t)__a, (int8x8_t)__b, 17); } |
| 2916 | __ai uint32x2_t vqshl_u32(uint32x2_t __a, int32x2_t __b) { |
| 2917 | return (uint32x2_t)__builtin_neon_vqshl_v((int8x8_t)__a, (int8x8_t)__b, 18); } |
| 2918 | __ai uint64x1_t vqshl_u64(uint64x1_t __a, int64x1_t __b) { |
| 2919 | return (uint64x1_t)__builtin_neon_vqshl_v((int8x8_t)__a, (int8x8_t)__b, 19); } |
| 2920 | __ai int8x16_t vqshlq_s8(int8x16_t __a, int8x16_t __b) { |
| 2921 | return (int8x16_t)__builtin_neon_vqshlq_v(__a, __b, 32); } |
| 2922 | __ai int16x8_t vqshlq_s16(int16x8_t __a, int16x8_t __b) { |
| 2923 | return (int16x8_t)__builtin_neon_vqshlq_v((int8x16_t)__a, (int8x16_t)__b, 33); } |
| 2924 | __ai int32x4_t vqshlq_s32(int32x4_t __a, int32x4_t __b) { |
| 2925 | return (int32x4_t)__builtin_neon_vqshlq_v((int8x16_t)__a, (int8x16_t)__b, 34); } |
| 2926 | __ai int64x2_t vqshlq_s64(int64x2_t __a, int64x2_t __b) { |
| 2927 | return (int64x2_t)__builtin_neon_vqshlq_v((int8x16_t)__a, (int8x16_t)__b, 35); } |
| 2928 | __ai uint8x16_t vqshlq_u8(uint8x16_t __a, int8x16_t __b) { |
| 2929 | return (uint8x16_t)__builtin_neon_vqshlq_v((int8x16_t)__a, __b, 48); } |
| 2930 | __ai uint16x8_t vqshlq_u16(uint16x8_t __a, int16x8_t __b) { |
| 2931 | return (uint16x8_t)__builtin_neon_vqshlq_v((int8x16_t)__a, (int8x16_t)__b, 49); } |
| 2932 | __ai uint32x4_t vqshlq_u32(uint32x4_t __a, int32x4_t __b) { |
| 2933 | return (uint32x4_t)__builtin_neon_vqshlq_v((int8x16_t)__a, (int8x16_t)__b, 50); } |
| 2934 | __ai uint64x2_t vqshlq_u64(uint64x2_t __a, int64x2_t __b) { |
| 2935 | return (uint64x2_t)__builtin_neon_vqshlq_v((int8x16_t)__a, (int8x16_t)__b, 51); } |
| 2936 | |
| 2937 | #define vqshlu_n_s8(a, __b) __extension__ ({ \ |
| 2938 | int8x8_t __a = (a); \ |
| 2939 | (uint8x8_t)__builtin_neon_vqshlu_n_v(__a, __b, 16); }) |
| 2940 | #define vqshlu_n_s16(a, __b) __extension__ ({ \ |
| 2941 | int16x4_t __a = (a); \ |
| 2942 | (uint16x4_t)__builtin_neon_vqshlu_n_v((int8x8_t)__a, __b, 17); }) |
| 2943 | #define vqshlu_n_s32(a, __b) __extension__ ({ \ |
| 2944 | int32x2_t __a = (a); \ |
| 2945 | (uint32x2_t)__builtin_neon_vqshlu_n_v((int8x8_t)__a, __b, 18); }) |
| 2946 | #define vqshlu_n_s64(a, __b) __extension__ ({ \ |
| 2947 | int64x1_t __a = (a); \ |
| 2948 | (uint64x1_t)__builtin_neon_vqshlu_n_v((int8x8_t)__a, __b, 19); }) |
| 2949 | #define vqshluq_n_s8(a, __b) __extension__ ({ \ |
| 2950 | int8x16_t __a = (a); \ |
| 2951 | (uint8x16_t)__builtin_neon_vqshluq_n_v(__a, __b, 48); }) |
| 2952 | #define vqshluq_n_s16(a, __b) __extension__ ({ \ |
| 2953 | int16x8_t __a = (a); \ |
| 2954 | (uint16x8_t)__builtin_neon_vqshluq_n_v((int8x16_t)__a, __b, 49); }) |
| 2955 | #define vqshluq_n_s32(a, __b) __extension__ ({ \ |
| 2956 | int32x4_t __a = (a); \ |
| 2957 | (uint32x4_t)__builtin_neon_vqshluq_n_v((int8x16_t)__a, __b, 50); }) |
| 2958 | #define vqshluq_n_s64(a, __b) __extension__ ({ \ |
| 2959 | int64x2_t __a = (a); \ |
| 2960 | (uint64x2_t)__builtin_neon_vqshluq_n_v((int8x16_t)__a, __b, 51); }) |
| 2961 | |
| 2962 | #define vqshl_n_s8(a, __b) __extension__ ({ \ |
| 2963 | int8x8_t __a = (a); \ |
| 2964 | (int8x8_t)__builtin_neon_vqshl_n_v(__a, __b, 0); }) |
| 2965 | #define vqshl_n_s16(a, __b) __extension__ ({ \ |
| 2966 | int16x4_t __a = (a); \ |
| 2967 | (int16x4_t)__builtin_neon_vqshl_n_v((int8x8_t)__a, __b, 1); }) |
| 2968 | #define vqshl_n_s32(a, __b) __extension__ ({ \ |
| 2969 | int32x2_t __a = (a); \ |
| 2970 | (int32x2_t)__builtin_neon_vqshl_n_v((int8x8_t)__a, __b, 2); }) |
| 2971 | #define vqshl_n_s64(a, __b) __extension__ ({ \ |
| 2972 | int64x1_t __a = (a); \ |
| 2973 | (int64x1_t)__builtin_neon_vqshl_n_v((int8x8_t)__a, __b, 3); }) |
| 2974 | #define vqshl_n_u8(a, __b) __extension__ ({ \ |
| 2975 | uint8x8_t __a = (a); \ |
| 2976 | (uint8x8_t)__builtin_neon_vqshl_n_v((int8x8_t)__a, __b, 16); }) |
| 2977 | #define vqshl_n_u16(a, __b) __extension__ ({ \ |
| 2978 | uint16x4_t __a = (a); \ |
| 2979 | (uint16x4_t)__builtin_neon_vqshl_n_v((int8x8_t)__a, __b, 17); }) |
| 2980 | #define vqshl_n_u32(a, __b) __extension__ ({ \ |
| 2981 | uint32x2_t __a = (a); \ |
| 2982 | (uint32x2_t)__builtin_neon_vqshl_n_v((int8x8_t)__a, __b, 18); }) |
| 2983 | #define vqshl_n_u64(a, __b) __extension__ ({ \ |
| 2984 | uint64x1_t __a = (a); \ |
| 2985 | (uint64x1_t)__builtin_neon_vqshl_n_v((int8x8_t)__a, __b, 19); }) |
| 2986 | #define vqshlq_n_s8(a, __b) __extension__ ({ \ |
| 2987 | int8x16_t __a = (a); \ |
| 2988 | (int8x16_t)__builtin_neon_vqshlq_n_v(__a, __b, 32); }) |
| 2989 | #define vqshlq_n_s16(a, __b) __extension__ ({ \ |
| 2990 | int16x8_t __a = (a); \ |
| 2991 | (int16x8_t)__builtin_neon_vqshlq_n_v((int8x16_t)__a, __b, 33); }) |
| 2992 | #define vqshlq_n_s32(a, __b) __extension__ ({ \ |
| 2993 | int32x4_t __a = (a); \ |
| 2994 | (int32x4_t)__builtin_neon_vqshlq_n_v((int8x16_t)__a, __b, 34); }) |
| 2995 | #define vqshlq_n_s64(a, __b) __extension__ ({ \ |
| 2996 | int64x2_t __a = (a); \ |
| 2997 | (int64x2_t)__builtin_neon_vqshlq_n_v((int8x16_t)__a, __b, 35); }) |
| 2998 | #define vqshlq_n_u8(a, __b) __extension__ ({ \ |
| 2999 | uint8x16_t __a = (a); \ |
| 3000 | (uint8x16_t)__builtin_neon_vqshlq_n_v((int8x16_t)__a, __b, 48); }) |
| 3001 | #define vqshlq_n_u16(a, __b) __extension__ ({ \ |
| 3002 | uint16x8_t __a = (a); \ |
| 3003 | (uint16x8_t)__builtin_neon_vqshlq_n_v((int8x16_t)__a, __b, 49); }) |
| 3004 | #define vqshlq_n_u32(a, __b) __extension__ ({ \ |
| 3005 | uint32x4_t __a = (a); \ |
| 3006 | (uint32x4_t)__builtin_neon_vqshlq_n_v((int8x16_t)__a, __b, 50); }) |
| 3007 | #define vqshlq_n_u64(a, __b) __extension__ ({ \ |
| 3008 | uint64x2_t __a = (a); \ |
| 3009 | (uint64x2_t)__builtin_neon_vqshlq_n_v((int8x16_t)__a, __b, 51); }) |
| 3010 | |
| 3011 | #define vqshrn_n_s16(a, __b) __extension__ ({ \ |
| 3012 | int16x8_t __a = (a); \ |
| 3013 | (int8x8_t)__builtin_neon_vqshrn_n_v((int8x16_t)__a, __b, 0); }) |
| 3014 | #define vqshrn_n_s32(a, __b) __extension__ ({ \ |
| 3015 | int32x4_t __a = (a); \ |
| 3016 | (int16x4_t)__builtin_neon_vqshrn_n_v((int8x16_t)__a, __b, 1); }) |
| 3017 | #define vqshrn_n_s64(a, __b) __extension__ ({ \ |
| 3018 | int64x2_t __a = (a); \ |
| 3019 | (int32x2_t)__builtin_neon_vqshrn_n_v((int8x16_t)__a, __b, 2); }) |
| 3020 | #define vqshrn_n_u16(a, __b) __extension__ ({ \ |
| 3021 | uint16x8_t __a = (a); \ |
| 3022 | (uint8x8_t)__builtin_neon_vqshrn_n_v((int8x16_t)__a, __b, 16); }) |
| 3023 | #define vqshrn_n_u32(a, __b) __extension__ ({ \ |
| 3024 | uint32x4_t __a = (a); \ |
| 3025 | (uint16x4_t)__builtin_neon_vqshrn_n_v((int8x16_t)__a, __b, 17); }) |
| 3026 | #define vqshrn_n_u64(a, __b) __extension__ ({ \ |
| 3027 | uint64x2_t __a = (a); \ |
| 3028 | (uint32x2_t)__builtin_neon_vqshrn_n_v((int8x16_t)__a, __b, 18); }) |
| 3029 | |
| 3030 | #define vqshrun_n_s16(a, __b) __extension__ ({ \ |
| 3031 | int16x8_t __a = (a); \ |
| 3032 | (uint8x8_t)__builtin_neon_vqshrun_n_v((int8x16_t)__a, __b, 16); }) |
| 3033 | #define vqshrun_n_s32(a, __b) __extension__ ({ \ |
| 3034 | int32x4_t __a = (a); \ |
| 3035 | (uint16x4_t)__builtin_neon_vqshrun_n_v((int8x16_t)__a, __b, 17); }) |
| 3036 | #define vqshrun_n_s64(a, __b) __extension__ ({ \ |
| 3037 | int64x2_t __a = (a); \ |
| 3038 | (uint32x2_t)__builtin_neon_vqshrun_n_v((int8x16_t)__a, __b, 18); }) |
| 3039 | |
| 3040 | __ai int8x8_t vqsub_s8(int8x8_t __a, int8x8_t __b) { |
| 3041 | return (int8x8_t)__builtin_neon_vqsub_v(__a, __b, 0); } |
| 3042 | __ai int16x4_t vqsub_s16(int16x4_t __a, int16x4_t __b) { |
| 3043 | return (int16x4_t)__builtin_neon_vqsub_v((int8x8_t)__a, (int8x8_t)__b, 1); } |
| 3044 | __ai int32x2_t vqsub_s32(int32x2_t __a, int32x2_t __b) { |
| 3045 | return (int32x2_t)__builtin_neon_vqsub_v((int8x8_t)__a, (int8x8_t)__b, 2); } |
| 3046 | __ai int64x1_t vqsub_s64(int64x1_t __a, int64x1_t __b) { |
| 3047 | return (int64x1_t)__builtin_neon_vqsub_v((int8x8_t)__a, (int8x8_t)__b, 3); } |
| 3048 | __ai uint8x8_t vqsub_u8(uint8x8_t __a, uint8x8_t __b) { |
| 3049 | return (uint8x8_t)__builtin_neon_vqsub_v((int8x8_t)__a, (int8x8_t)__b, 16); } |
| 3050 | __ai uint16x4_t vqsub_u16(uint16x4_t __a, uint16x4_t __b) { |
| 3051 | return (uint16x4_t)__builtin_neon_vqsub_v((int8x8_t)__a, (int8x8_t)__b, 17); } |
| 3052 | __ai uint32x2_t vqsub_u32(uint32x2_t __a, uint32x2_t __b) { |
| 3053 | return (uint32x2_t)__builtin_neon_vqsub_v((int8x8_t)__a, (int8x8_t)__b, 18); } |
| 3054 | __ai uint64x1_t vqsub_u64(uint64x1_t __a, uint64x1_t __b) { |
| 3055 | return (uint64x1_t)__builtin_neon_vqsub_v((int8x8_t)__a, (int8x8_t)__b, 19); } |
| 3056 | __ai int8x16_t vqsubq_s8(int8x16_t __a, int8x16_t __b) { |
| 3057 | return (int8x16_t)__builtin_neon_vqsubq_v(__a, __b, 32); } |
| 3058 | __ai int16x8_t vqsubq_s16(int16x8_t __a, int16x8_t __b) { |
| 3059 | return (int16x8_t)__builtin_neon_vqsubq_v((int8x16_t)__a, (int8x16_t)__b, 33); } |
| 3060 | __ai int32x4_t vqsubq_s32(int32x4_t __a, int32x4_t __b) { |
| 3061 | return (int32x4_t)__builtin_neon_vqsubq_v((int8x16_t)__a, (int8x16_t)__b, 34); } |
| 3062 | __ai int64x2_t vqsubq_s64(int64x2_t __a, int64x2_t __b) { |
| 3063 | return (int64x2_t)__builtin_neon_vqsubq_v((int8x16_t)__a, (int8x16_t)__b, 35); } |
| 3064 | __ai uint8x16_t vqsubq_u8(uint8x16_t __a, uint8x16_t __b) { |
| 3065 | return (uint8x16_t)__builtin_neon_vqsubq_v((int8x16_t)__a, (int8x16_t)__b, 48); } |
| 3066 | __ai uint16x8_t vqsubq_u16(uint16x8_t __a, uint16x8_t __b) { |
| 3067 | return (uint16x8_t)__builtin_neon_vqsubq_v((int8x16_t)__a, (int8x16_t)__b, 49); } |
| 3068 | __ai uint32x4_t vqsubq_u32(uint32x4_t __a, uint32x4_t __b) { |
| 3069 | return (uint32x4_t)__builtin_neon_vqsubq_v((int8x16_t)__a, (int8x16_t)__b, 50); } |
| 3070 | __ai uint64x2_t vqsubq_u64(uint64x2_t __a, uint64x2_t __b) { |
| 3071 | return (uint64x2_t)__builtin_neon_vqsubq_v((int8x16_t)__a, (int8x16_t)__b, 51); } |
| 3072 | |
| 3073 | __ai int8x8_t vraddhn_s16(int16x8_t __a, int16x8_t __b) { |
| 3074 | return (int8x8_t)__builtin_neon_vraddhn_v((int8x16_t)__a, (int8x16_t)__b, 0); } |
| 3075 | __ai int16x4_t vraddhn_s32(int32x4_t __a, int32x4_t __b) { |
| 3076 | return (int16x4_t)__builtin_neon_vraddhn_v((int8x16_t)__a, (int8x16_t)__b, 1); } |
| 3077 | __ai int32x2_t vraddhn_s64(int64x2_t __a, int64x2_t __b) { |
| 3078 | return (int32x2_t)__builtin_neon_vraddhn_v((int8x16_t)__a, (int8x16_t)__b, 2); } |
| 3079 | __ai uint8x8_t vraddhn_u16(uint16x8_t __a, uint16x8_t __b) { |
| 3080 | return (uint8x8_t)__builtin_neon_vraddhn_v((int8x16_t)__a, (int8x16_t)__b, 16); } |
| 3081 | __ai uint16x4_t vraddhn_u32(uint32x4_t __a, uint32x4_t __b) { |
| 3082 | return (uint16x4_t)__builtin_neon_vraddhn_v((int8x16_t)__a, (int8x16_t)__b, 17); } |
| 3083 | __ai uint32x2_t vraddhn_u64(uint64x2_t __a, uint64x2_t __b) { |
| 3084 | return (uint32x2_t)__builtin_neon_vraddhn_v((int8x16_t)__a, (int8x16_t)__b, 18); } |
| 3085 | |
| 3086 | __ai float32x2_t vrecpe_f32(float32x2_t __a) { |
| 3087 | return (float32x2_t)__builtin_neon_vrecpe_v((int8x8_t)__a, 8); } |
| 3088 | __ai uint32x2_t vrecpe_u32(uint32x2_t __a) { |
| 3089 | return (uint32x2_t)__builtin_neon_vrecpe_v((int8x8_t)__a, 18); } |
| 3090 | __ai float32x4_t vrecpeq_f32(float32x4_t __a) { |
| 3091 | return (float32x4_t)__builtin_neon_vrecpeq_v((int8x16_t)__a, 40); } |
| 3092 | __ai uint32x4_t vrecpeq_u32(uint32x4_t __a) { |
| 3093 | return (uint32x4_t)__builtin_neon_vrecpeq_v((int8x16_t)__a, 50); } |
| 3094 | |
| 3095 | __ai float32x2_t vrecps_f32(float32x2_t __a, float32x2_t __b) { |
| 3096 | return (float32x2_t)__builtin_neon_vrecps_v((int8x8_t)__a, (int8x8_t)__b, 8); } |
| 3097 | __ai float32x4_t vrecpsq_f32(float32x4_t __a, float32x4_t __b) { |
| 3098 | return (float32x4_t)__builtin_neon_vrecpsq_v((int8x16_t)__a, (int8x16_t)__b, 40); } |
| 3099 | |
| 3100 | __ai int8x8_t vreinterpret_s8_s16(int16x4_t __a) { |
| 3101 | return (int8x8_t)__a; } |
| 3102 | __ai int8x8_t vreinterpret_s8_s32(int32x2_t __a) { |
| 3103 | return (int8x8_t)__a; } |
| 3104 | __ai int8x8_t vreinterpret_s8_s64(int64x1_t __a) { |
| 3105 | return (int8x8_t)__a; } |
| 3106 | __ai int8x8_t vreinterpret_s8_u8(uint8x8_t __a) { |
| 3107 | return (int8x8_t)__a; } |
| 3108 | __ai int8x8_t vreinterpret_s8_u16(uint16x4_t __a) { |
| 3109 | return (int8x8_t)__a; } |
| 3110 | __ai int8x8_t vreinterpret_s8_u32(uint32x2_t __a) { |
| 3111 | return (int8x8_t)__a; } |
| 3112 | __ai int8x8_t vreinterpret_s8_u64(uint64x1_t __a) { |
| 3113 | return (int8x8_t)__a; } |
| 3114 | __ai int8x8_t vreinterpret_s8_f16(float16x4_t __a) { |
| 3115 | return (int8x8_t)__a; } |
| 3116 | __ai int8x8_t vreinterpret_s8_f32(float32x2_t __a) { |
| 3117 | return (int8x8_t)__a; } |
| 3118 | __ai int8x8_t vreinterpret_s8_p8(poly8x8_t __a) { |
| 3119 | return (int8x8_t)__a; } |
| 3120 | __ai int8x8_t vreinterpret_s8_p16(poly16x4_t __a) { |
| 3121 | return (int8x8_t)__a; } |
| 3122 | __ai int16x4_t vreinterpret_s16_s8(int8x8_t __a) { |
| 3123 | return (int16x4_t)__a; } |
| 3124 | __ai int16x4_t vreinterpret_s16_s32(int32x2_t __a) { |
| 3125 | return (int16x4_t)__a; } |
| 3126 | __ai int16x4_t vreinterpret_s16_s64(int64x1_t __a) { |
| 3127 | return (int16x4_t)__a; } |
| 3128 | __ai int16x4_t vreinterpret_s16_u8(uint8x8_t __a) { |
| 3129 | return (int16x4_t)__a; } |
| 3130 | __ai int16x4_t vreinterpret_s16_u16(uint16x4_t __a) { |
| 3131 | return (int16x4_t)__a; } |
| 3132 | __ai int16x4_t vreinterpret_s16_u32(uint32x2_t __a) { |
| 3133 | return (int16x4_t)__a; } |
| 3134 | __ai int16x4_t vreinterpret_s16_u64(uint64x1_t __a) { |
| 3135 | return (int16x4_t)__a; } |
| 3136 | __ai int16x4_t vreinterpret_s16_f16(float16x4_t __a) { |
| 3137 | return (int16x4_t)__a; } |
| 3138 | __ai int16x4_t vreinterpret_s16_f32(float32x2_t __a) { |
| 3139 | return (int16x4_t)__a; } |
| 3140 | __ai int16x4_t vreinterpret_s16_p8(poly8x8_t __a) { |
| 3141 | return (int16x4_t)__a; } |
| 3142 | __ai int16x4_t vreinterpret_s16_p16(poly16x4_t __a) { |
| 3143 | return (int16x4_t)__a; } |
| 3144 | __ai int32x2_t vreinterpret_s32_s8(int8x8_t __a) { |
| 3145 | return (int32x2_t)__a; } |
| 3146 | __ai int32x2_t vreinterpret_s32_s16(int16x4_t __a) { |
| 3147 | return (int32x2_t)__a; } |
| 3148 | __ai int32x2_t vreinterpret_s32_s64(int64x1_t __a) { |
| 3149 | return (int32x2_t)__a; } |
| 3150 | __ai int32x2_t vreinterpret_s32_u8(uint8x8_t __a) { |
| 3151 | return (int32x2_t)__a; } |
| 3152 | __ai int32x2_t vreinterpret_s32_u16(uint16x4_t __a) { |
| 3153 | return (int32x2_t)__a; } |
| 3154 | __ai int32x2_t vreinterpret_s32_u32(uint32x2_t __a) { |
| 3155 | return (int32x2_t)__a; } |
| 3156 | __ai int32x2_t vreinterpret_s32_u64(uint64x1_t __a) { |
| 3157 | return (int32x2_t)__a; } |
| 3158 | __ai int32x2_t vreinterpret_s32_f16(float16x4_t __a) { |
| 3159 | return (int32x2_t)__a; } |
| 3160 | __ai int32x2_t vreinterpret_s32_f32(float32x2_t __a) { |
| 3161 | return (int32x2_t)__a; } |
| 3162 | __ai int32x2_t vreinterpret_s32_p8(poly8x8_t __a) { |
| 3163 | return (int32x2_t)__a; } |
| 3164 | __ai int32x2_t vreinterpret_s32_p16(poly16x4_t __a) { |
| 3165 | return (int32x2_t)__a; } |
| 3166 | __ai int64x1_t vreinterpret_s64_s8(int8x8_t __a) { |
| 3167 | return (int64x1_t)__a; } |
| 3168 | __ai int64x1_t vreinterpret_s64_s16(int16x4_t __a) { |
| 3169 | return (int64x1_t)__a; } |
| 3170 | __ai int64x1_t vreinterpret_s64_s32(int32x2_t __a) { |
| 3171 | return (int64x1_t)__a; } |
| 3172 | __ai int64x1_t vreinterpret_s64_u8(uint8x8_t __a) { |
| 3173 | return (int64x1_t)__a; } |
| 3174 | __ai int64x1_t vreinterpret_s64_u16(uint16x4_t __a) { |
| 3175 | return (int64x1_t)__a; } |
| 3176 | __ai int64x1_t vreinterpret_s64_u32(uint32x2_t __a) { |
| 3177 | return (int64x1_t)__a; } |
| 3178 | __ai int64x1_t vreinterpret_s64_u64(uint64x1_t __a) { |
| 3179 | return (int64x1_t)__a; } |
| 3180 | __ai int64x1_t vreinterpret_s64_f16(float16x4_t __a) { |
| 3181 | return (int64x1_t)__a; } |
| 3182 | __ai int64x1_t vreinterpret_s64_f32(float32x2_t __a) { |
| 3183 | return (int64x1_t)__a; } |
| 3184 | __ai int64x1_t vreinterpret_s64_p8(poly8x8_t __a) { |
| 3185 | return (int64x1_t)__a; } |
| 3186 | __ai int64x1_t vreinterpret_s64_p16(poly16x4_t __a) { |
| 3187 | return (int64x1_t)__a; } |
| 3188 | __ai uint8x8_t vreinterpret_u8_s8(int8x8_t __a) { |
| 3189 | return (uint8x8_t)__a; } |
| 3190 | __ai uint8x8_t vreinterpret_u8_s16(int16x4_t __a) { |
| 3191 | return (uint8x8_t)__a; } |
| 3192 | __ai uint8x8_t vreinterpret_u8_s32(int32x2_t __a) { |
| 3193 | return (uint8x8_t)__a; } |
| 3194 | __ai uint8x8_t vreinterpret_u8_s64(int64x1_t __a) { |
| 3195 | return (uint8x8_t)__a; } |
| 3196 | __ai uint8x8_t vreinterpret_u8_u16(uint16x4_t __a) { |
| 3197 | return (uint8x8_t)__a; } |
| 3198 | __ai uint8x8_t vreinterpret_u8_u32(uint32x2_t __a) { |
| 3199 | return (uint8x8_t)__a; } |
| 3200 | __ai uint8x8_t vreinterpret_u8_u64(uint64x1_t __a) { |
| 3201 | return (uint8x8_t)__a; } |
| 3202 | __ai uint8x8_t vreinterpret_u8_f16(float16x4_t __a) { |
| 3203 | return (uint8x8_t)__a; } |
| 3204 | __ai uint8x8_t vreinterpret_u8_f32(float32x2_t __a) { |
| 3205 | return (uint8x8_t)__a; } |
| 3206 | __ai uint8x8_t vreinterpret_u8_p8(poly8x8_t __a) { |
| 3207 | return (uint8x8_t)__a; } |
| 3208 | __ai uint8x8_t vreinterpret_u8_p16(poly16x4_t __a) { |
| 3209 | return (uint8x8_t)__a; } |
| 3210 | __ai uint16x4_t vreinterpret_u16_s8(int8x8_t __a) { |
| 3211 | return (uint16x4_t)__a; } |
| 3212 | __ai uint16x4_t vreinterpret_u16_s16(int16x4_t __a) { |
| 3213 | return (uint16x4_t)__a; } |
| 3214 | __ai uint16x4_t vreinterpret_u16_s32(int32x2_t __a) { |
| 3215 | return (uint16x4_t)__a; } |
| 3216 | __ai uint16x4_t vreinterpret_u16_s64(int64x1_t __a) { |
| 3217 | return (uint16x4_t)__a; } |
| 3218 | __ai uint16x4_t vreinterpret_u16_u8(uint8x8_t __a) { |
| 3219 | return (uint16x4_t)__a; } |
| 3220 | __ai uint16x4_t vreinterpret_u16_u32(uint32x2_t __a) { |
| 3221 | return (uint16x4_t)__a; } |
| 3222 | __ai uint16x4_t vreinterpret_u16_u64(uint64x1_t __a) { |
| 3223 | return (uint16x4_t)__a; } |
| 3224 | __ai uint16x4_t vreinterpret_u16_f16(float16x4_t __a) { |
| 3225 | return (uint16x4_t)__a; } |
| 3226 | __ai uint16x4_t vreinterpret_u16_f32(float32x2_t __a) { |
| 3227 | return (uint16x4_t)__a; } |
| 3228 | __ai uint16x4_t vreinterpret_u16_p8(poly8x8_t __a) { |
| 3229 | return (uint16x4_t)__a; } |
| 3230 | __ai uint16x4_t vreinterpret_u16_p16(poly16x4_t __a) { |
| 3231 | return (uint16x4_t)__a; } |
| 3232 | __ai uint32x2_t vreinterpret_u32_s8(int8x8_t __a) { |
| 3233 | return (uint32x2_t)__a; } |
| 3234 | __ai uint32x2_t vreinterpret_u32_s16(int16x4_t __a) { |
| 3235 | return (uint32x2_t)__a; } |
| 3236 | __ai uint32x2_t vreinterpret_u32_s32(int32x2_t __a) { |
| 3237 | return (uint32x2_t)__a; } |
| 3238 | __ai uint32x2_t vreinterpret_u32_s64(int64x1_t __a) { |
| 3239 | return (uint32x2_t)__a; } |
| 3240 | __ai uint32x2_t vreinterpret_u32_u8(uint8x8_t __a) { |
| 3241 | return (uint32x2_t)__a; } |
| 3242 | __ai uint32x2_t vreinterpret_u32_u16(uint16x4_t __a) { |
| 3243 | return (uint32x2_t)__a; } |
| 3244 | __ai uint32x2_t vreinterpret_u32_u64(uint64x1_t __a) { |
| 3245 | return (uint32x2_t)__a; } |
| 3246 | __ai uint32x2_t vreinterpret_u32_f16(float16x4_t __a) { |
| 3247 | return (uint32x2_t)__a; } |
| 3248 | __ai uint32x2_t vreinterpret_u32_f32(float32x2_t __a) { |
| 3249 | return (uint32x2_t)__a; } |
| 3250 | __ai uint32x2_t vreinterpret_u32_p8(poly8x8_t __a) { |
| 3251 | return (uint32x2_t)__a; } |
| 3252 | __ai uint32x2_t vreinterpret_u32_p16(poly16x4_t __a) { |
| 3253 | return (uint32x2_t)__a; } |
| 3254 | __ai uint64x1_t vreinterpret_u64_s8(int8x8_t __a) { |
| 3255 | return (uint64x1_t)__a; } |
| 3256 | __ai uint64x1_t vreinterpret_u64_s16(int16x4_t __a) { |
| 3257 | return (uint64x1_t)__a; } |
| 3258 | __ai uint64x1_t vreinterpret_u64_s32(int32x2_t __a) { |
| 3259 | return (uint64x1_t)__a; } |
| 3260 | __ai uint64x1_t vreinterpret_u64_s64(int64x1_t __a) { |
| 3261 | return (uint64x1_t)__a; } |
| 3262 | __ai uint64x1_t vreinterpret_u64_u8(uint8x8_t __a) { |
| 3263 | return (uint64x1_t)__a; } |
| 3264 | __ai uint64x1_t vreinterpret_u64_u16(uint16x4_t __a) { |
| 3265 | return (uint64x1_t)__a; } |
| 3266 | __ai uint64x1_t vreinterpret_u64_u32(uint32x2_t __a) { |
| 3267 | return (uint64x1_t)__a; } |
| 3268 | __ai uint64x1_t vreinterpret_u64_f16(float16x4_t __a) { |
| 3269 | return (uint64x1_t)__a; } |
| 3270 | __ai uint64x1_t vreinterpret_u64_f32(float32x2_t __a) { |
| 3271 | return (uint64x1_t)__a; } |
| 3272 | __ai uint64x1_t vreinterpret_u64_p8(poly8x8_t __a) { |
| 3273 | return (uint64x1_t)__a; } |
| 3274 | __ai uint64x1_t vreinterpret_u64_p16(poly16x4_t __a) { |
| 3275 | return (uint64x1_t)__a; } |
| 3276 | __ai float16x4_t vreinterpret_f16_s8(int8x8_t __a) { |
| 3277 | return (float16x4_t)__a; } |
| 3278 | __ai float16x4_t vreinterpret_f16_s16(int16x4_t __a) { |
| 3279 | return (float16x4_t)__a; } |
| 3280 | __ai float16x4_t vreinterpret_f16_s32(int32x2_t __a) { |
| 3281 | return (float16x4_t)__a; } |
| 3282 | __ai float16x4_t vreinterpret_f16_s64(int64x1_t __a) { |
| 3283 | return (float16x4_t)__a; } |
| 3284 | __ai float16x4_t vreinterpret_f16_u8(uint8x8_t __a) { |
| 3285 | return (float16x4_t)__a; } |
| 3286 | __ai float16x4_t vreinterpret_f16_u16(uint16x4_t __a) { |
| 3287 | return (float16x4_t)__a; } |
| 3288 | __ai float16x4_t vreinterpret_f16_u32(uint32x2_t __a) { |
| 3289 | return (float16x4_t)__a; } |
| 3290 | __ai float16x4_t vreinterpret_f16_u64(uint64x1_t __a) { |
| 3291 | return (float16x4_t)__a; } |
| 3292 | __ai float16x4_t vreinterpret_f16_f32(float32x2_t __a) { |
| 3293 | return (float16x4_t)__a; } |
| 3294 | __ai float16x4_t vreinterpret_f16_p8(poly8x8_t __a) { |
| 3295 | return (float16x4_t)__a; } |
| 3296 | __ai float16x4_t vreinterpret_f16_p16(poly16x4_t __a) { |
| 3297 | return (float16x4_t)__a; } |
| 3298 | __ai float32x2_t vreinterpret_f32_s8(int8x8_t __a) { |
| 3299 | return (float32x2_t)__a; } |
| 3300 | __ai float32x2_t vreinterpret_f32_s16(int16x4_t __a) { |
| 3301 | return (float32x2_t)__a; } |
| 3302 | __ai float32x2_t vreinterpret_f32_s32(int32x2_t __a) { |
| 3303 | return (float32x2_t)__a; } |
| 3304 | __ai float32x2_t vreinterpret_f32_s64(int64x1_t __a) { |
| 3305 | return (float32x2_t)__a; } |
| 3306 | __ai float32x2_t vreinterpret_f32_u8(uint8x8_t __a) { |
| 3307 | return (float32x2_t)__a; } |
| 3308 | __ai float32x2_t vreinterpret_f32_u16(uint16x4_t __a) { |
| 3309 | return (float32x2_t)__a; } |
| 3310 | __ai float32x2_t vreinterpret_f32_u32(uint32x2_t __a) { |
| 3311 | return (float32x2_t)__a; } |
| 3312 | __ai float32x2_t vreinterpret_f32_u64(uint64x1_t __a) { |
| 3313 | return (float32x2_t)__a; } |
| 3314 | __ai float32x2_t vreinterpret_f32_f16(float16x4_t __a) { |
| 3315 | return (float32x2_t)__a; } |
| 3316 | __ai float32x2_t vreinterpret_f32_p8(poly8x8_t __a) { |
| 3317 | return (float32x2_t)__a; } |
| 3318 | __ai float32x2_t vreinterpret_f32_p16(poly16x4_t __a) { |
| 3319 | return (float32x2_t)__a; } |
| 3320 | __ai poly8x8_t vreinterpret_p8_s8(int8x8_t __a) { |
| 3321 | return (poly8x8_t)__a; } |
| 3322 | __ai poly8x8_t vreinterpret_p8_s16(int16x4_t __a) { |
| 3323 | return (poly8x8_t)__a; } |
| 3324 | __ai poly8x8_t vreinterpret_p8_s32(int32x2_t __a) { |
| 3325 | return (poly8x8_t)__a; } |
| 3326 | __ai poly8x8_t vreinterpret_p8_s64(int64x1_t __a) { |
| 3327 | return (poly8x8_t)__a; } |
| 3328 | __ai poly8x8_t vreinterpret_p8_u8(uint8x8_t __a) { |
| 3329 | return (poly8x8_t)__a; } |
| 3330 | __ai poly8x8_t vreinterpret_p8_u16(uint16x4_t __a) { |
| 3331 | return (poly8x8_t)__a; } |
| 3332 | __ai poly8x8_t vreinterpret_p8_u32(uint32x2_t __a) { |
| 3333 | return (poly8x8_t)__a; } |
| 3334 | __ai poly8x8_t vreinterpret_p8_u64(uint64x1_t __a) { |
| 3335 | return (poly8x8_t)__a; } |
| 3336 | __ai poly8x8_t vreinterpret_p8_f16(float16x4_t __a) { |
| 3337 | return (poly8x8_t)__a; } |
| 3338 | __ai poly8x8_t vreinterpret_p8_f32(float32x2_t __a) { |
| 3339 | return (poly8x8_t)__a; } |
| 3340 | __ai poly8x8_t vreinterpret_p8_p16(poly16x4_t __a) { |
| 3341 | return (poly8x8_t)__a; } |
| 3342 | __ai poly16x4_t vreinterpret_p16_s8(int8x8_t __a) { |
| 3343 | return (poly16x4_t)__a; } |
| 3344 | __ai poly16x4_t vreinterpret_p16_s16(int16x4_t __a) { |
| 3345 | return (poly16x4_t)__a; } |
| 3346 | __ai poly16x4_t vreinterpret_p16_s32(int32x2_t __a) { |
| 3347 | return (poly16x4_t)__a; } |
| 3348 | __ai poly16x4_t vreinterpret_p16_s64(int64x1_t __a) { |
| 3349 | return (poly16x4_t)__a; } |
| 3350 | __ai poly16x4_t vreinterpret_p16_u8(uint8x8_t __a) { |
| 3351 | return (poly16x4_t)__a; } |
| 3352 | __ai poly16x4_t vreinterpret_p16_u16(uint16x4_t __a) { |
| 3353 | return (poly16x4_t)__a; } |
| 3354 | __ai poly16x4_t vreinterpret_p16_u32(uint32x2_t __a) { |
| 3355 | return (poly16x4_t)__a; } |
| 3356 | __ai poly16x4_t vreinterpret_p16_u64(uint64x1_t __a) { |
| 3357 | return (poly16x4_t)__a; } |
| 3358 | __ai poly16x4_t vreinterpret_p16_f16(float16x4_t __a) { |
| 3359 | return (poly16x4_t)__a; } |
| 3360 | __ai poly16x4_t vreinterpret_p16_f32(float32x2_t __a) { |
| 3361 | return (poly16x4_t)__a; } |
| 3362 | __ai poly16x4_t vreinterpret_p16_p8(poly8x8_t __a) { |
| 3363 | return (poly16x4_t)__a; } |
| 3364 | __ai int8x16_t vreinterpretq_s8_s16(int16x8_t __a) { |
| 3365 | return (int8x16_t)__a; } |
| 3366 | __ai int8x16_t vreinterpretq_s8_s32(int32x4_t __a) { |
| 3367 | return (int8x16_t)__a; } |
| 3368 | __ai int8x16_t vreinterpretq_s8_s64(int64x2_t __a) { |
| 3369 | return (int8x16_t)__a; } |
| 3370 | __ai int8x16_t vreinterpretq_s8_u8(uint8x16_t __a) { |
| 3371 | return (int8x16_t)__a; } |
| 3372 | __ai int8x16_t vreinterpretq_s8_u16(uint16x8_t __a) { |
| 3373 | return (int8x16_t)__a; } |
| 3374 | __ai int8x16_t vreinterpretq_s8_u32(uint32x4_t __a) { |
| 3375 | return (int8x16_t)__a; } |
| 3376 | __ai int8x16_t vreinterpretq_s8_u64(uint64x2_t __a) { |
| 3377 | return (int8x16_t)__a; } |
| 3378 | __ai int8x16_t vreinterpretq_s8_f16(float16x8_t __a) { |
| 3379 | return (int8x16_t)__a; } |
| 3380 | __ai int8x16_t vreinterpretq_s8_f32(float32x4_t __a) { |
| 3381 | return (int8x16_t)__a; } |
| 3382 | __ai int8x16_t vreinterpretq_s8_p8(poly8x16_t __a) { |
| 3383 | return (int8x16_t)__a; } |
| 3384 | __ai int8x16_t vreinterpretq_s8_p16(poly16x8_t __a) { |
| 3385 | return (int8x16_t)__a; } |
| 3386 | __ai int16x8_t vreinterpretq_s16_s8(int8x16_t __a) { |
| 3387 | return (int16x8_t)__a; } |
| 3388 | __ai int16x8_t vreinterpretq_s16_s32(int32x4_t __a) { |
| 3389 | return (int16x8_t)__a; } |
| 3390 | __ai int16x8_t vreinterpretq_s16_s64(int64x2_t __a) { |
| 3391 | return (int16x8_t)__a; } |
| 3392 | __ai int16x8_t vreinterpretq_s16_u8(uint8x16_t __a) { |
| 3393 | return (int16x8_t)__a; } |
| 3394 | __ai int16x8_t vreinterpretq_s16_u16(uint16x8_t __a) { |
| 3395 | return (int16x8_t)__a; } |
| 3396 | __ai int16x8_t vreinterpretq_s16_u32(uint32x4_t __a) { |
| 3397 | return (int16x8_t)__a; } |
| 3398 | __ai int16x8_t vreinterpretq_s16_u64(uint64x2_t __a) { |
| 3399 | return (int16x8_t)__a; } |
| 3400 | __ai int16x8_t vreinterpretq_s16_f16(float16x8_t __a) { |
| 3401 | return (int16x8_t)__a; } |
| 3402 | __ai int16x8_t vreinterpretq_s16_f32(float32x4_t __a) { |
| 3403 | return (int16x8_t)__a; } |
| 3404 | __ai int16x8_t vreinterpretq_s16_p8(poly8x16_t __a) { |
| 3405 | return (int16x8_t)__a; } |
| 3406 | __ai int16x8_t vreinterpretq_s16_p16(poly16x8_t __a) { |
| 3407 | return (int16x8_t)__a; } |
| 3408 | __ai int32x4_t vreinterpretq_s32_s8(int8x16_t __a) { |
| 3409 | return (int32x4_t)__a; } |
| 3410 | __ai int32x4_t vreinterpretq_s32_s16(int16x8_t __a) { |
| 3411 | return (int32x4_t)__a; } |
| 3412 | __ai int32x4_t vreinterpretq_s32_s64(int64x2_t __a) { |
| 3413 | return (int32x4_t)__a; } |
| 3414 | __ai int32x4_t vreinterpretq_s32_u8(uint8x16_t __a) { |
| 3415 | return (int32x4_t)__a; } |
| 3416 | __ai int32x4_t vreinterpretq_s32_u16(uint16x8_t __a) { |
| 3417 | return (int32x4_t)__a; } |
| 3418 | __ai int32x4_t vreinterpretq_s32_u32(uint32x4_t __a) { |
| 3419 | return (int32x4_t)__a; } |
| 3420 | __ai int32x4_t vreinterpretq_s32_u64(uint64x2_t __a) { |
| 3421 | return (int32x4_t)__a; } |
| 3422 | __ai int32x4_t vreinterpretq_s32_f16(float16x8_t __a) { |
| 3423 | return (int32x4_t)__a; } |
| 3424 | __ai int32x4_t vreinterpretq_s32_f32(float32x4_t __a) { |
| 3425 | return (int32x4_t)__a; } |
| 3426 | __ai int32x4_t vreinterpretq_s32_p8(poly8x16_t __a) { |
| 3427 | return (int32x4_t)__a; } |
| 3428 | __ai int32x4_t vreinterpretq_s32_p16(poly16x8_t __a) { |
| 3429 | return (int32x4_t)__a; } |
| 3430 | __ai int64x2_t vreinterpretq_s64_s8(int8x16_t __a) { |
| 3431 | return (int64x2_t)__a; } |
| 3432 | __ai int64x2_t vreinterpretq_s64_s16(int16x8_t __a) { |
| 3433 | return (int64x2_t)__a; } |
| 3434 | __ai int64x2_t vreinterpretq_s64_s32(int32x4_t __a) { |
| 3435 | return (int64x2_t)__a; } |
| 3436 | __ai int64x2_t vreinterpretq_s64_u8(uint8x16_t __a) { |
| 3437 | return (int64x2_t)__a; } |
| 3438 | __ai int64x2_t vreinterpretq_s64_u16(uint16x8_t __a) { |
| 3439 | return (int64x2_t)__a; } |
| 3440 | __ai int64x2_t vreinterpretq_s64_u32(uint32x4_t __a) { |
| 3441 | return (int64x2_t)__a; } |
| 3442 | __ai int64x2_t vreinterpretq_s64_u64(uint64x2_t __a) { |
| 3443 | return (int64x2_t)__a; } |
| 3444 | __ai int64x2_t vreinterpretq_s64_f16(float16x8_t __a) { |
| 3445 | return (int64x2_t)__a; } |
| 3446 | __ai int64x2_t vreinterpretq_s64_f32(float32x4_t __a) { |
| 3447 | return (int64x2_t)__a; } |
| 3448 | __ai int64x2_t vreinterpretq_s64_p8(poly8x16_t __a) { |
| 3449 | return (int64x2_t)__a; } |
| 3450 | __ai int64x2_t vreinterpretq_s64_p16(poly16x8_t __a) { |
| 3451 | return (int64x2_t)__a; } |
| 3452 | __ai uint8x16_t vreinterpretq_u8_s8(int8x16_t __a) { |
| 3453 | return (uint8x16_t)__a; } |
| 3454 | __ai uint8x16_t vreinterpretq_u8_s16(int16x8_t __a) { |
| 3455 | return (uint8x16_t)__a; } |
| 3456 | __ai uint8x16_t vreinterpretq_u8_s32(int32x4_t __a) { |
| 3457 | return (uint8x16_t)__a; } |
| 3458 | __ai uint8x16_t vreinterpretq_u8_s64(int64x2_t __a) { |
| 3459 | return (uint8x16_t)__a; } |
| 3460 | __ai uint8x16_t vreinterpretq_u8_u16(uint16x8_t __a) { |
| 3461 | return (uint8x16_t)__a; } |
| 3462 | __ai uint8x16_t vreinterpretq_u8_u32(uint32x4_t __a) { |
| 3463 | return (uint8x16_t)__a; } |
| 3464 | __ai uint8x16_t vreinterpretq_u8_u64(uint64x2_t __a) { |
| 3465 | return (uint8x16_t)__a; } |
| 3466 | __ai uint8x16_t vreinterpretq_u8_f16(float16x8_t __a) { |
| 3467 | return (uint8x16_t)__a; } |
| 3468 | __ai uint8x16_t vreinterpretq_u8_f32(float32x4_t __a) { |
| 3469 | return (uint8x16_t)__a; } |
| 3470 | __ai uint8x16_t vreinterpretq_u8_p8(poly8x16_t __a) { |
| 3471 | return (uint8x16_t)__a; } |
| 3472 | __ai uint8x16_t vreinterpretq_u8_p16(poly16x8_t __a) { |
| 3473 | return (uint8x16_t)__a; } |
| 3474 | __ai uint16x8_t vreinterpretq_u16_s8(int8x16_t __a) { |
| 3475 | return (uint16x8_t)__a; } |
| 3476 | __ai uint16x8_t vreinterpretq_u16_s16(int16x8_t __a) { |
| 3477 | return (uint16x8_t)__a; } |
| 3478 | __ai uint16x8_t vreinterpretq_u16_s32(int32x4_t __a) { |
| 3479 | return (uint16x8_t)__a; } |
| 3480 | __ai uint16x8_t vreinterpretq_u16_s64(int64x2_t __a) { |
| 3481 | return (uint16x8_t)__a; } |
| 3482 | __ai uint16x8_t vreinterpretq_u16_u8(uint8x16_t __a) { |
| 3483 | return (uint16x8_t)__a; } |
| 3484 | __ai uint16x8_t vreinterpretq_u16_u32(uint32x4_t __a) { |
| 3485 | return (uint16x8_t)__a; } |
| 3486 | __ai uint16x8_t vreinterpretq_u16_u64(uint64x2_t __a) { |
| 3487 | return (uint16x8_t)__a; } |
| 3488 | __ai uint16x8_t vreinterpretq_u16_f16(float16x8_t __a) { |
| 3489 | return (uint16x8_t)__a; } |
| 3490 | __ai uint16x8_t vreinterpretq_u16_f32(float32x4_t __a) { |
| 3491 | return (uint16x8_t)__a; } |
| 3492 | __ai uint16x8_t vreinterpretq_u16_p8(poly8x16_t __a) { |
| 3493 | return (uint16x8_t)__a; } |
| 3494 | __ai uint16x8_t vreinterpretq_u16_p16(poly16x8_t __a) { |
| 3495 | return (uint16x8_t)__a; } |
| 3496 | __ai uint32x4_t vreinterpretq_u32_s8(int8x16_t __a) { |
| 3497 | return (uint32x4_t)__a; } |
| 3498 | __ai uint32x4_t vreinterpretq_u32_s16(int16x8_t __a) { |
| 3499 | return (uint32x4_t)__a; } |
| 3500 | __ai uint32x4_t vreinterpretq_u32_s32(int32x4_t __a) { |
| 3501 | return (uint32x4_t)__a; } |
| 3502 | __ai uint32x4_t vreinterpretq_u32_s64(int64x2_t __a) { |
| 3503 | return (uint32x4_t)__a; } |
| 3504 | __ai uint32x4_t vreinterpretq_u32_u8(uint8x16_t __a) { |
| 3505 | return (uint32x4_t)__a; } |
| 3506 | __ai uint32x4_t vreinterpretq_u32_u16(uint16x8_t __a) { |
| 3507 | return (uint32x4_t)__a; } |
| 3508 | __ai uint32x4_t vreinterpretq_u32_u64(uint64x2_t __a) { |
| 3509 | return (uint32x4_t)__a; } |
| 3510 | __ai uint32x4_t vreinterpretq_u32_f16(float16x8_t __a) { |
| 3511 | return (uint32x4_t)__a; } |
| 3512 | __ai uint32x4_t vreinterpretq_u32_f32(float32x4_t __a) { |
| 3513 | return (uint32x4_t)__a; } |
| 3514 | __ai uint32x4_t vreinterpretq_u32_p8(poly8x16_t __a) { |
| 3515 | return (uint32x4_t)__a; } |
| 3516 | __ai uint32x4_t vreinterpretq_u32_p16(poly16x8_t __a) { |
| 3517 | return (uint32x4_t)__a; } |
| 3518 | __ai uint64x2_t vreinterpretq_u64_s8(int8x16_t __a) { |
| 3519 | return (uint64x2_t)__a; } |
| 3520 | __ai uint64x2_t vreinterpretq_u64_s16(int16x8_t __a) { |
| 3521 | return (uint64x2_t)__a; } |
| 3522 | __ai uint64x2_t vreinterpretq_u64_s32(int32x4_t __a) { |
| 3523 | return (uint64x2_t)__a; } |
| 3524 | __ai uint64x2_t vreinterpretq_u64_s64(int64x2_t __a) { |
| 3525 | return (uint64x2_t)__a; } |
| 3526 | __ai uint64x2_t vreinterpretq_u64_u8(uint8x16_t __a) { |
| 3527 | return (uint64x2_t)__a; } |
| 3528 | __ai uint64x2_t vreinterpretq_u64_u16(uint16x8_t __a) { |
| 3529 | return (uint64x2_t)__a; } |
| 3530 | __ai uint64x2_t vreinterpretq_u64_u32(uint32x4_t __a) { |
| 3531 | return (uint64x2_t)__a; } |
| 3532 | __ai uint64x2_t vreinterpretq_u64_f16(float16x8_t __a) { |
| 3533 | return (uint64x2_t)__a; } |
| 3534 | __ai uint64x2_t vreinterpretq_u64_f32(float32x4_t __a) { |
| 3535 | return (uint64x2_t)__a; } |
| 3536 | __ai uint64x2_t vreinterpretq_u64_p8(poly8x16_t __a) { |
| 3537 | return (uint64x2_t)__a; } |
| 3538 | __ai uint64x2_t vreinterpretq_u64_p16(poly16x8_t __a) { |
| 3539 | return (uint64x2_t)__a; } |
| 3540 | __ai float16x8_t vreinterpretq_f16_s8(int8x16_t __a) { |
| 3541 | return (float16x8_t)__a; } |
| 3542 | __ai float16x8_t vreinterpretq_f16_s16(int16x8_t __a) { |
| 3543 | return (float16x8_t)__a; } |
| 3544 | __ai float16x8_t vreinterpretq_f16_s32(int32x4_t __a) { |
| 3545 | return (float16x8_t)__a; } |
| 3546 | __ai float16x8_t vreinterpretq_f16_s64(int64x2_t __a) { |
| 3547 | return (float16x8_t)__a; } |
| 3548 | __ai float16x8_t vreinterpretq_f16_u8(uint8x16_t __a) { |
| 3549 | return (float16x8_t)__a; } |
| 3550 | __ai float16x8_t vreinterpretq_f16_u16(uint16x8_t __a) { |
| 3551 | return (float16x8_t)__a; } |
| 3552 | __ai float16x8_t vreinterpretq_f16_u32(uint32x4_t __a) { |
| 3553 | return (float16x8_t)__a; } |
| 3554 | __ai float16x8_t vreinterpretq_f16_u64(uint64x2_t __a) { |
| 3555 | return (float16x8_t)__a; } |
| 3556 | __ai float16x8_t vreinterpretq_f16_f32(float32x4_t __a) { |
| 3557 | return (float16x8_t)__a; } |
| 3558 | __ai float16x8_t vreinterpretq_f16_p8(poly8x16_t __a) { |
| 3559 | return (float16x8_t)__a; } |
| 3560 | __ai float16x8_t vreinterpretq_f16_p16(poly16x8_t __a) { |
| 3561 | return (float16x8_t)__a; } |
| 3562 | __ai float32x4_t vreinterpretq_f32_s8(int8x16_t __a) { |
| 3563 | return (float32x4_t)__a; } |
| 3564 | __ai float32x4_t vreinterpretq_f32_s16(int16x8_t __a) { |
| 3565 | return (float32x4_t)__a; } |
| 3566 | __ai float32x4_t vreinterpretq_f32_s32(int32x4_t __a) { |
| 3567 | return (float32x4_t)__a; } |
| 3568 | __ai float32x4_t vreinterpretq_f32_s64(int64x2_t __a) { |
| 3569 | return (float32x4_t)__a; } |
| 3570 | __ai float32x4_t vreinterpretq_f32_u8(uint8x16_t __a) { |
| 3571 | return (float32x4_t)__a; } |
| 3572 | __ai float32x4_t vreinterpretq_f32_u16(uint16x8_t __a) { |
| 3573 | return (float32x4_t)__a; } |
| 3574 | __ai float32x4_t vreinterpretq_f32_u32(uint32x4_t __a) { |
| 3575 | return (float32x4_t)__a; } |
| 3576 | __ai float32x4_t vreinterpretq_f32_u64(uint64x2_t __a) { |
| 3577 | return (float32x4_t)__a; } |
| 3578 | __ai float32x4_t vreinterpretq_f32_f16(float16x8_t __a) { |
| 3579 | return (float32x4_t)__a; } |
| 3580 | __ai float32x4_t vreinterpretq_f32_p8(poly8x16_t __a) { |
| 3581 | return (float32x4_t)__a; } |
| 3582 | __ai float32x4_t vreinterpretq_f32_p16(poly16x8_t __a) { |
| 3583 | return (float32x4_t)__a; } |
| 3584 | __ai poly8x16_t vreinterpretq_p8_s8(int8x16_t __a) { |
| 3585 | return (poly8x16_t)__a; } |
| 3586 | __ai poly8x16_t vreinterpretq_p8_s16(int16x8_t __a) { |
| 3587 | return (poly8x16_t)__a; } |
| 3588 | __ai poly8x16_t vreinterpretq_p8_s32(int32x4_t __a) { |
| 3589 | return (poly8x16_t)__a; } |
| 3590 | __ai poly8x16_t vreinterpretq_p8_s64(int64x2_t __a) { |
| 3591 | return (poly8x16_t)__a; } |
| 3592 | __ai poly8x16_t vreinterpretq_p8_u8(uint8x16_t __a) { |
| 3593 | return (poly8x16_t)__a; } |
| 3594 | __ai poly8x16_t vreinterpretq_p8_u16(uint16x8_t __a) { |
| 3595 | return (poly8x16_t)__a; } |
| 3596 | __ai poly8x16_t vreinterpretq_p8_u32(uint32x4_t __a) { |
| 3597 | return (poly8x16_t)__a; } |
| 3598 | __ai poly8x16_t vreinterpretq_p8_u64(uint64x2_t __a) { |
| 3599 | return (poly8x16_t)__a; } |
| 3600 | __ai poly8x16_t vreinterpretq_p8_f16(float16x8_t __a) { |
| 3601 | return (poly8x16_t)__a; } |
| 3602 | __ai poly8x16_t vreinterpretq_p8_f32(float32x4_t __a) { |
| 3603 | return (poly8x16_t)__a; } |
| 3604 | __ai poly8x16_t vreinterpretq_p8_p16(poly16x8_t __a) { |
| 3605 | return (poly8x16_t)__a; } |
| 3606 | __ai poly16x8_t vreinterpretq_p16_s8(int8x16_t __a) { |
| 3607 | return (poly16x8_t)__a; } |
| 3608 | __ai poly16x8_t vreinterpretq_p16_s16(int16x8_t __a) { |
| 3609 | return (poly16x8_t)__a; } |
| 3610 | __ai poly16x8_t vreinterpretq_p16_s32(int32x4_t __a) { |
| 3611 | return (poly16x8_t)__a; } |
| 3612 | __ai poly16x8_t vreinterpretq_p16_s64(int64x2_t __a) { |
| 3613 | return (poly16x8_t)__a; } |
| 3614 | __ai poly16x8_t vreinterpretq_p16_u8(uint8x16_t __a) { |
| 3615 | return (poly16x8_t)__a; } |
| 3616 | __ai poly16x8_t vreinterpretq_p16_u16(uint16x8_t __a) { |
| 3617 | return (poly16x8_t)__a; } |
| 3618 | __ai poly16x8_t vreinterpretq_p16_u32(uint32x4_t __a) { |
| 3619 | return (poly16x8_t)__a; } |
| 3620 | __ai poly16x8_t vreinterpretq_p16_u64(uint64x2_t __a) { |
| 3621 | return (poly16x8_t)__a; } |
| 3622 | __ai poly16x8_t vreinterpretq_p16_f16(float16x8_t __a) { |
| 3623 | return (poly16x8_t)__a; } |
| 3624 | __ai poly16x8_t vreinterpretq_p16_f32(float32x4_t __a) { |
| 3625 | return (poly16x8_t)__a; } |
| 3626 | __ai poly16x8_t vreinterpretq_p16_p8(poly8x16_t __a) { |
| 3627 | return (poly16x8_t)__a; } |
| 3628 | |
| 3629 | __ai int8x8_t vrev16_s8(int8x8_t __a) { |
| 3630 | return __builtin_shufflevector(__a, __a, 1, 0, 3, 2, 5, 4, 7, 6); } |
| 3631 | __ai uint8x8_t vrev16_u8(uint8x8_t __a) { |
| 3632 | return __builtin_shufflevector(__a, __a, 1, 0, 3, 2, 5, 4, 7, 6); } |
| 3633 | __ai poly8x8_t vrev16_p8(poly8x8_t __a) { |
| 3634 | return __builtin_shufflevector(__a, __a, 1, 0, 3, 2, 5, 4, 7, 6); } |
| 3635 | __ai int8x16_t vrev16q_s8(int8x16_t __a) { |
| 3636 | return __builtin_shufflevector(__a, __a, 1, 0, 3, 2, 5, 4, 7, 6, 9, 8, 11, 10, 13, 12, 15, 14); } |
| 3637 | __ai uint8x16_t vrev16q_u8(uint8x16_t __a) { |
| 3638 | return __builtin_shufflevector(__a, __a, 1, 0, 3, 2, 5, 4, 7, 6, 9, 8, 11, 10, 13, 12, 15, 14); } |
| 3639 | __ai poly8x16_t vrev16q_p8(poly8x16_t __a) { |
| 3640 | return __builtin_shufflevector(__a, __a, 1, 0, 3, 2, 5, 4, 7, 6, 9, 8, 11, 10, 13, 12, 15, 14); } |
| 3641 | |
| 3642 | __ai int8x8_t vrev32_s8(int8x8_t __a) { |
| 3643 | return __builtin_shufflevector(__a, __a, 3, 2, 1, 0, 7, 6, 5, 4); } |
| 3644 | __ai int16x4_t vrev32_s16(int16x4_t __a) { |
| 3645 | return __builtin_shufflevector(__a, __a, 1, 0, 3, 2); } |
| 3646 | __ai uint8x8_t vrev32_u8(uint8x8_t __a) { |
| 3647 | return __builtin_shufflevector(__a, __a, 3, 2, 1, 0, 7, 6, 5, 4); } |
| 3648 | __ai uint16x4_t vrev32_u16(uint16x4_t __a) { |
| 3649 | return __builtin_shufflevector(__a, __a, 1, 0, 3, 2); } |
| 3650 | __ai poly8x8_t vrev32_p8(poly8x8_t __a) { |
| 3651 | return __builtin_shufflevector(__a, __a, 3, 2, 1, 0, 7, 6, 5, 4); } |
| 3652 | __ai poly16x4_t vrev32_p16(poly16x4_t __a) { |
| 3653 | return __builtin_shufflevector(__a, __a, 1, 0, 3, 2); } |
| 3654 | __ai int8x16_t vrev32q_s8(int8x16_t __a) { |
| 3655 | return __builtin_shufflevector(__a, __a, 3, 2, 1, 0, 7, 6, 5, 4, 11, 10, 9, 8, 15, 14, 13, 12); } |
| 3656 | __ai int16x8_t vrev32q_s16(int16x8_t __a) { |
| 3657 | return __builtin_shufflevector(__a, __a, 1, 0, 3, 2, 5, 4, 7, 6); } |
| 3658 | __ai uint8x16_t vrev32q_u8(uint8x16_t __a) { |
| 3659 | return __builtin_shufflevector(__a, __a, 3, 2, 1, 0, 7, 6, 5, 4, 11, 10, 9, 8, 15, 14, 13, 12); } |
| 3660 | __ai uint16x8_t vrev32q_u16(uint16x8_t __a) { |
| 3661 | return __builtin_shufflevector(__a, __a, 1, 0, 3, 2, 5, 4, 7, 6); } |
| 3662 | __ai poly8x16_t vrev32q_p8(poly8x16_t __a) { |
| 3663 | return __builtin_shufflevector(__a, __a, 3, 2, 1, 0, 7, 6, 5, 4, 11, 10, 9, 8, 15, 14, 13, 12); } |
| 3664 | __ai poly16x8_t vrev32q_p16(poly16x8_t __a) { |
| 3665 | return __builtin_shufflevector(__a, __a, 1, 0, 3, 2, 5, 4, 7, 6); } |
| 3666 | |
| 3667 | __ai int8x8_t vrev64_s8(int8x8_t __a) { |
| 3668 | return __builtin_shufflevector(__a, __a, 7, 6, 5, 4, 3, 2, 1, 0); } |
| 3669 | __ai int16x4_t vrev64_s16(int16x4_t __a) { |
| 3670 | return __builtin_shufflevector(__a, __a, 3, 2, 1, 0); } |
| 3671 | __ai int32x2_t vrev64_s32(int32x2_t __a) { |
| 3672 | return __builtin_shufflevector(__a, __a, 1, 0); } |
| 3673 | __ai uint8x8_t vrev64_u8(uint8x8_t __a) { |
| 3674 | return __builtin_shufflevector(__a, __a, 7, 6, 5, 4, 3, 2, 1, 0); } |
| 3675 | __ai uint16x4_t vrev64_u16(uint16x4_t __a) { |
| 3676 | return __builtin_shufflevector(__a, __a, 3, 2, 1, 0); } |
| 3677 | __ai uint32x2_t vrev64_u32(uint32x2_t __a) { |
| 3678 | return __builtin_shufflevector(__a, __a, 1, 0); } |
| 3679 | __ai poly8x8_t vrev64_p8(poly8x8_t __a) { |
| 3680 | return __builtin_shufflevector(__a, __a, 7, 6, 5, 4, 3, 2, 1, 0); } |
| 3681 | __ai poly16x4_t vrev64_p16(poly16x4_t __a) { |
| 3682 | return __builtin_shufflevector(__a, __a, 3, 2, 1, 0); } |
| 3683 | __ai float32x2_t vrev64_f32(float32x2_t __a) { |
| 3684 | return __builtin_shufflevector(__a, __a, 1, 0); } |
| 3685 | __ai int8x16_t vrev64q_s8(int8x16_t __a) { |
| 3686 | return __builtin_shufflevector(__a, __a, 7, 6, 5, 4, 3, 2, 1, 0, 15, 14, 13, 12, 11, 10, 9, 8); } |
| 3687 | __ai int16x8_t vrev64q_s16(int16x8_t __a) { |
| 3688 | return __builtin_shufflevector(__a, __a, 3, 2, 1, 0, 7, 6, 5, 4); } |
| 3689 | __ai int32x4_t vrev64q_s32(int32x4_t __a) { |
| 3690 | return __builtin_shufflevector(__a, __a, 1, 0, 3, 2); } |
| 3691 | __ai uint8x16_t vrev64q_u8(uint8x16_t __a) { |
| 3692 | return __builtin_shufflevector(__a, __a, 7, 6, 5, 4, 3, 2, 1, 0, 15, 14, 13, 12, 11, 10, 9, 8); } |
| 3693 | __ai uint16x8_t vrev64q_u16(uint16x8_t __a) { |
| 3694 | return __builtin_shufflevector(__a, __a, 3, 2, 1, 0, 7, 6, 5, 4); } |
| 3695 | __ai uint32x4_t vrev64q_u32(uint32x4_t __a) { |
| 3696 | return __builtin_shufflevector(__a, __a, 1, 0, 3, 2); } |
| 3697 | __ai poly8x16_t vrev64q_p8(poly8x16_t __a) { |
| 3698 | return __builtin_shufflevector(__a, __a, 7, 6, 5, 4, 3, 2, 1, 0, 15, 14, 13, 12, 11, 10, 9, 8); } |
| 3699 | __ai poly16x8_t vrev64q_p16(poly16x8_t __a) { |
| 3700 | return __builtin_shufflevector(__a, __a, 3, 2, 1, 0, 7, 6, 5, 4); } |
| 3701 | __ai float32x4_t vrev64q_f32(float32x4_t __a) { |
| 3702 | return __builtin_shufflevector(__a, __a, 1, 0, 3, 2); } |
| 3703 | |
| 3704 | __ai int8x8_t vrhadd_s8(int8x8_t __a, int8x8_t __b) { |
| 3705 | return (int8x8_t)__builtin_neon_vrhadd_v(__a, __b, 0); } |
| 3706 | __ai int16x4_t vrhadd_s16(int16x4_t __a, int16x4_t __b) { |
| 3707 | return (int16x4_t)__builtin_neon_vrhadd_v((int8x8_t)__a, (int8x8_t)__b, 1); } |
| 3708 | __ai int32x2_t vrhadd_s32(int32x2_t __a, int32x2_t __b) { |
| 3709 | return (int32x2_t)__builtin_neon_vrhadd_v((int8x8_t)__a, (int8x8_t)__b, 2); } |
| 3710 | __ai uint8x8_t vrhadd_u8(uint8x8_t __a, uint8x8_t __b) { |
| 3711 | return (uint8x8_t)__builtin_neon_vrhadd_v((int8x8_t)__a, (int8x8_t)__b, 16); } |
| 3712 | __ai uint16x4_t vrhadd_u16(uint16x4_t __a, uint16x4_t __b) { |
| 3713 | return (uint16x4_t)__builtin_neon_vrhadd_v((int8x8_t)__a, (int8x8_t)__b, 17); } |
| 3714 | __ai uint32x2_t vrhadd_u32(uint32x2_t __a, uint32x2_t __b) { |
| 3715 | return (uint32x2_t)__builtin_neon_vrhadd_v((int8x8_t)__a, (int8x8_t)__b, 18); } |
| 3716 | __ai int8x16_t vrhaddq_s8(int8x16_t __a, int8x16_t __b) { |
| 3717 | return (int8x16_t)__builtin_neon_vrhaddq_v(__a, __b, 32); } |
| 3718 | __ai int16x8_t vrhaddq_s16(int16x8_t __a, int16x8_t __b) { |
| 3719 | return (int16x8_t)__builtin_neon_vrhaddq_v((int8x16_t)__a, (int8x16_t)__b, 33); } |
| 3720 | __ai int32x4_t vrhaddq_s32(int32x4_t __a, int32x4_t __b) { |
| 3721 | return (int32x4_t)__builtin_neon_vrhaddq_v((int8x16_t)__a, (int8x16_t)__b, 34); } |
| 3722 | __ai uint8x16_t vrhaddq_u8(uint8x16_t __a, uint8x16_t __b) { |
| 3723 | return (uint8x16_t)__builtin_neon_vrhaddq_v((int8x16_t)__a, (int8x16_t)__b, 48); } |
| 3724 | __ai uint16x8_t vrhaddq_u16(uint16x8_t __a, uint16x8_t __b) { |
| 3725 | return (uint16x8_t)__builtin_neon_vrhaddq_v((int8x16_t)__a, (int8x16_t)__b, 49); } |
| 3726 | __ai uint32x4_t vrhaddq_u32(uint32x4_t __a, uint32x4_t __b) { |
| 3727 | return (uint32x4_t)__builtin_neon_vrhaddq_v((int8x16_t)__a, (int8x16_t)__b, 50); } |
| 3728 | |
| 3729 | __ai int8x8_t vrshl_s8(int8x8_t __a, int8x8_t __b) { |
| 3730 | return (int8x8_t)__builtin_neon_vrshl_v(__a, __b, 0); } |
| 3731 | __ai int16x4_t vrshl_s16(int16x4_t __a, int16x4_t __b) { |
| 3732 | return (int16x4_t)__builtin_neon_vrshl_v((int8x8_t)__a, (int8x8_t)__b, 1); } |
| 3733 | __ai int32x2_t vrshl_s32(int32x2_t __a, int32x2_t __b) { |
| 3734 | return (int32x2_t)__builtin_neon_vrshl_v((int8x8_t)__a, (int8x8_t)__b, 2); } |
| 3735 | __ai int64x1_t vrshl_s64(int64x1_t __a, int64x1_t __b) { |
| 3736 | return (int64x1_t)__builtin_neon_vrshl_v((int8x8_t)__a, (int8x8_t)__b, 3); } |
| 3737 | __ai uint8x8_t vrshl_u8(uint8x8_t __a, int8x8_t __b) { |
| 3738 | return (uint8x8_t)__builtin_neon_vrshl_v((int8x8_t)__a, __b, 16); } |
| 3739 | __ai uint16x4_t vrshl_u16(uint16x4_t __a, int16x4_t __b) { |
| 3740 | return (uint16x4_t)__builtin_neon_vrshl_v((int8x8_t)__a, (int8x8_t)__b, 17); } |
| 3741 | __ai uint32x2_t vrshl_u32(uint32x2_t __a, int32x2_t __b) { |
| 3742 | return (uint32x2_t)__builtin_neon_vrshl_v((int8x8_t)__a, (int8x8_t)__b, 18); } |
| 3743 | __ai uint64x1_t vrshl_u64(uint64x1_t __a, int64x1_t __b) { |
| 3744 | return (uint64x1_t)__builtin_neon_vrshl_v((int8x8_t)__a, (int8x8_t)__b, 19); } |
| 3745 | __ai int8x16_t vrshlq_s8(int8x16_t __a, int8x16_t __b) { |
| 3746 | return (int8x16_t)__builtin_neon_vrshlq_v(__a, __b, 32); } |
| 3747 | __ai int16x8_t vrshlq_s16(int16x8_t __a, int16x8_t __b) { |
| 3748 | return (int16x8_t)__builtin_neon_vrshlq_v((int8x16_t)__a, (int8x16_t)__b, 33); } |
| 3749 | __ai int32x4_t vrshlq_s32(int32x4_t __a, int32x4_t __b) { |
| 3750 | return (int32x4_t)__builtin_neon_vrshlq_v((int8x16_t)__a, (int8x16_t)__b, 34); } |
| 3751 | __ai int64x2_t vrshlq_s64(int64x2_t __a, int64x2_t __b) { |
| 3752 | return (int64x2_t)__builtin_neon_vrshlq_v((int8x16_t)__a, (int8x16_t)__b, 35); } |
| 3753 | __ai uint8x16_t vrshlq_u8(uint8x16_t __a, int8x16_t __b) { |
| 3754 | return (uint8x16_t)__builtin_neon_vrshlq_v((int8x16_t)__a, __b, 48); } |
| 3755 | __ai uint16x8_t vrshlq_u16(uint16x8_t __a, int16x8_t __b) { |
| 3756 | return (uint16x8_t)__builtin_neon_vrshlq_v((int8x16_t)__a, (int8x16_t)__b, 49); } |
| 3757 | __ai uint32x4_t vrshlq_u32(uint32x4_t __a, int32x4_t __b) { |
| 3758 | return (uint32x4_t)__builtin_neon_vrshlq_v((int8x16_t)__a, (int8x16_t)__b, 50); } |
| 3759 | __ai uint64x2_t vrshlq_u64(uint64x2_t __a, int64x2_t __b) { |
| 3760 | return (uint64x2_t)__builtin_neon_vrshlq_v((int8x16_t)__a, (int8x16_t)__b, 51); } |
| 3761 | |
| 3762 | #define vrshrn_n_s16(a, __b) __extension__ ({ \ |
| 3763 | int16x8_t __a = (a); \ |
| 3764 | (int8x8_t)__builtin_neon_vrshrn_n_v((int8x16_t)__a, __b, 0); }) |
| 3765 | #define vrshrn_n_s32(a, __b) __extension__ ({ \ |
| 3766 | int32x4_t __a = (a); \ |
| 3767 | (int16x4_t)__builtin_neon_vrshrn_n_v((int8x16_t)__a, __b, 1); }) |
| 3768 | #define vrshrn_n_s64(a, __b) __extension__ ({ \ |
| 3769 | int64x2_t __a = (a); \ |
| 3770 | (int32x2_t)__builtin_neon_vrshrn_n_v((int8x16_t)__a, __b, 2); }) |
| 3771 | #define vrshrn_n_u16(a, __b) __extension__ ({ \ |
| 3772 | uint16x8_t __a = (a); \ |
| 3773 | (uint8x8_t)__builtin_neon_vrshrn_n_v((int8x16_t)__a, __b, 16); }) |
| 3774 | #define vrshrn_n_u32(a, __b) __extension__ ({ \ |
| 3775 | uint32x4_t __a = (a); \ |
| 3776 | (uint16x4_t)__builtin_neon_vrshrn_n_v((int8x16_t)__a, __b, 17); }) |
| 3777 | #define vrshrn_n_u64(a, __b) __extension__ ({ \ |
| 3778 | uint64x2_t __a = (a); \ |
| 3779 | (uint32x2_t)__builtin_neon_vrshrn_n_v((int8x16_t)__a, __b, 18); }) |
| 3780 | |
| 3781 | #define vrshr_n_s8(a, __b) __extension__ ({ \ |
| 3782 | int8x8_t __a = (a); \ |
| 3783 | (int8x8_t)__builtin_neon_vrshr_n_v(__a, __b, 0); }) |
| 3784 | #define vrshr_n_s16(a, __b) __extension__ ({ \ |
| 3785 | int16x4_t __a = (a); \ |
| 3786 | (int16x4_t)__builtin_neon_vrshr_n_v((int8x8_t)__a, __b, 1); }) |
| 3787 | #define vrshr_n_s32(a, __b) __extension__ ({ \ |
| 3788 | int32x2_t __a = (a); \ |
| 3789 | (int32x2_t)__builtin_neon_vrshr_n_v((int8x8_t)__a, __b, 2); }) |
| 3790 | #define vrshr_n_s64(a, __b) __extension__ ({ \ |
| 3791 | int64x1_t __a = (a); \ |
| 3792 | (int64x1_t)__builtin_neon_vrshr_n_v((int8x8_t)__a, __b, 3); }) |
| 3793 | #define vrshr_n_u8(a, __b) __extension__ ({ \ |
| 3794 | uint8x8_t __a = (a); \ |
| 3795 | (uint8x8_t)__builtin_neon_vrshr_n_v((int8x8_t)__a, __b, 16); }) |
| 3796 | #define vrshr_n_u16(a, __b) __extension__ ({ \ |
| 3797 | uint16x4_t __a = (a); \ |
| 3798 | (uint16x4_t)__builtin_neon_vrshr_n_v((int8x8_t)__a, __b, 17); }) |
| 3799 | #define vrshr_n_u32(a, __b) __extension__ ({ \ |
| 3800 | uint32x2_t __a = (a); \ |
| 3801 | (uint32x2_t)__builtin_neon_vrshr_n_v((int8x8_t)__a, __b, 18); }) |
| 3802 | #define vrshr_n_u64(a, __b) __extension__ ({ \ |
| 3803 | uint64x1_t __a = (a); \ |
| 3804 | (uint64x1_t)__builtin_neon_vrshr_n_v((int8x8_t)__a, __b, 19); }) |
| 3805 | #define vrshrq_n_s8(a, __b) __extension__ ({ \ |
| 3806 | int8x16_t __a = (a); \ |
| 3807 | (int8x16_t)__builtin_neon_vrshrq_n_v(__a, __b, 32); }) |
| 3808 | #define vrshrq_n_s16(a, __b) __extension__ ({ \ |
| 3809 | int16x8_t __a = (a); \ |
| 3810 | (int16x8_t)__builtin_neon_vrshrq_n_v((int8x16_t)__a, __b, 33); }) |
| 3811 | #define vrshrq_n_s32(a, __b) __extension__ ({ \ |
| 3812 | int32x4_t __a = (a); \ |
| 3813 | (int32x4_t)__builtin_neon_vrshrq_n_v((int8x16_t)__a, __b, 34); }) |
| 3814 | #define vrshrq_n_s64(a, __b) __extension__ ({ \ |
| 3815 | int64x2_t __a = (a); \ |
| 3816 | (int64x2_t)__builtin_neon_vrshrq_n_v((int8x16_t)__a, __b, 35); }) |
| 3817 | #define vrshrq_n_u8(a, __b) __extension__ ({ \ |
| 3818 | uint8x16_t __a = (a); \ |
| 3819 | (uint8x16_t)__builtin_neon_vrshrq_n_v((int8x16_t)__a, __b, 48); }) |
| 3820 | #define vrshrq_n_u16(a, __b) __extension__ ({ \ |
| 3821 | uint16x8_t __a = (a); \ |
| 3822 | (uint16x8_t)__builtin_neon_vrshrq_n_v((int8x16_t)__a, __b, 49); }) |
| 3823 | #define vrshrq_n_u32(a, __b) __extension__ ({ \ |
| 3824 | uint32x4_t __a = (a); \ |
| 3825 | (uint32x4_t)__builtin_neon_vrshrq_n_v((int8x16_t)__a, __b, 50); }) |
| 3826 | #define vrshrq_n_u64(a, __b) __extension__ ({ \ |
| 3827 | uint64x2_t __a = (a); \ |
| 3828 | (uint64x2_t)__builtin_neon_vrshrq_n_v((int8x16_t)__a, __b, 51); }) |
| 3829 | |
| 3830 | __ai float32x2_t vrsqrte_f32(float32x2_t __a) { |
| 3831 | return (float32x2_t)__builtin_neon_vrsqrte_v((int8x8_t)__a, 8); } |
| 3832 | __ai uint32x2_t vrsqrte_u32(uint32x2_t __a) { |
| 3833 | return (uint32x2_t)__builtin_neon_vrsqrte_v((int8x8_t)__a, 18); } |
| 3834 | __ai float32x4_t vrsqrteq_f32(float32x4_t __a) { |
| 3835 | return (float32x4_t)__builtin_neon_vrsqrteq_v((int8x16_t)__a, 40); } |
| 3836 | __ai uint32x4_t vrsqrteq_u32(uint32x4_t __a) { |
| 3837 | return (uint32x4_t)__builtin_neon_vrsqrteq_v((int8x16_t)__a, 50); } |
| 3838 | |
| 3839 | __ai float32x2_t vrsqrts_f32(float32x2_t __a, float32x2_t __b) { |
| 3840 | return (float32x2_t)__builtin_neon_vrsqrts_v((int8x8_t)__a, (int8x8_t)__b, 8); } |
| 3841 | __ai float32x4_t vrsqrtsq_f32(float32x4_t __a, float32x4_t __b) { |
| 3842 | return (float32x4_t)__builtin_neon_vrsqrtsq_v((int8x16_t)__a, (int8x16_t)__b, 40); } |
| 3843 | |
| 3844 | #define vrsra_n_s8(a, b, __c) __extension__ ({ \ |
| 3845 | int8x8_t __a = (a); int8x8_t __b = (b); \ |
| 3846 | (int8x8_t)__builtin_neon_vrsra_n_v(__a, __b, __c, 0); }) |
| 3847 | #define vrsra_n_s16(a, b, __c) __extension__ ({ \ |
| 3848 | int16x4_t __a = (a); int16x4_t __b = (b); \ |
| 3849 | (int16x4_t)__builtin_neon_vrsra_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 1); }) |
| 3850 | #define vrsra_n_s32(a, b, __c) __extension__ ({ \ |
| 3851 | int32x2_t __a = (a); int32x2_t __b = (b); \ |
| 3852 | (int32x2_t)__builtin_neon_vrsra_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 2); }) |
| 3853 | #define vrsra_n_s64(a, b, __c) __extension__ ({ \ |
| 3854 | int64x1_t __a = (a); int64x1_t __b = (b); \ |
| 3855 | (int64x1_t)__builtin_neon_vrsra_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 3); }) |
| 3856 | #define vrsra_n_u8(a, b, __c) __extension__ ({ \ |
| 3857 | uint8x8_t __a = (a); uint8x8_t __b = (b); \ |
| 3858 | (uint8x8_t)__builtin_neon_vrsra_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 16); }) |
| 3859 | #define vrsra_n_u16(a, b, __c) __extension__ ({ \ |
| 3860 | uint16x4_t __a = (a); uint16x4_t __b = (b); \ |
| 3861 | (uint16x4_t)__builtin_neon_vrsra_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 17); }) |
| 3862 | #define vrsra_n_u32(a, b, __c) __extension__ ({ \ |
| 3863 | uint32x2_t __a = (a); uint32x2_t __b = (b); \ |
| 3864 | (uint32x2_t)__builtin_neon_vrsra_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 18); }) |
| 3865 | #define vrsra_n_u64(a, b, __c) __extension__ ({ \ |
| 3866 | uint64x1_t __a = (a); uint64x1_t __b = (b); \ |
| 3867 | (uint64x1_t)__builtin_neon_vrsra_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 19); }) |
| 3868 | #define vrsraq_n_s8(a, b, __c) __extension__ ({ \ |
| 3869 | int8x16_t __a = (a); int8x16_t __b = (b); \ |
| 3870 | (int8x16_t)__builtin_neon_vrsraq_n_v(__a, __b, __c, 32); }) |
| 3871 | #define vrsraq_n_s16(a, b, __c) __extension__ ({ \ |
| 3872 | int16x8_t __a = (a); int16x8_t __b = (b); \ |
| 3873 | (int16x8_t)__builtin_neon_vrsraq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 33); }) |
| 3874 | #define vrsraq_n_s32(a, b, __c) __extension__ ({ \ |
| 3875 | int32x4_t __a = (a); int32x4_t __b = (b); \ |
| 3876 | (int32x4_t)__builtin_neon_vrsraq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 34); }) |
| 3877 | #define vrsraq_n_s64(a, b, __c) __extension__ ({ \ |
| 3878 | int64x2_t __a = (a); int64x2_t __b = (b); \ |
| 3879 | (int64x2_t)__builtin_neon_vrsraq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 35); }) |
| 3880 | #define vrsraq_n_u8(a, b, __c) __extension__ ({ \ |
| 3881 | uint8x16_t __a = (a); uint8x16_t __b = (b); \ |
| 3882 | (uint8x16_t)__builtin_neon_vrsraq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 48); }) |
| 3883 | #define vrsraq_n_u16(a, b, __c) __extension__ ({ \ |
| 3884 | uint16x8_t __a = (a); uint16x8_t __b = (b); \ |
| 3885 | (uint16x8_t)__builtin_neon_vrsraq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 49); }) |
| 3886 | #define vrsraq_n_u32(a, b, __c) __extension__ ({ \ |
| 3887 | uint32x4_t __a = (a); uint32x4_t __b = (b); \ |
| 3888 | (uint32x4_t)__builtin_neon_vrsraq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 50); }) |
| 3889 | #define vrsraq_n_u64(a, b, __c) __extension__ ({ \ |
| 3890 | uint64x2_t __a = (a); uint64x2_t __b = (b); \ |
| 3891 | (uint64x2_t)__builtin_neon_vrsraq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 51); }) |
| 3892 | |
| 3893 | __ai int8x8_t vrsubhn_s16(int16x8_t __a, int16x8_t __b) { |
| 3894 | return (int8x8_t)__builtin_neon_vrsubhn_v((int8x16_t)__a, (int8x16_t)__b, 0); } |
| 3895 | __ai int16x4_t vrsubhn_s32(int32x4_t __a, int32x4_t __b) { |
| 3896 | return (int16x4_t)__builtin_neon_vrsubhn_v((int8x16_t)__a, (int8x16_t)__b, 1); } |
| 3897 | __ai int32x2_t vrsubhn_s64(int64x2_t __a, int64x2_t __b) { |
| 3898 | return (int32x2_t)__builtin_neon_vrsubhn_v((int8x16_t)__a, (int8x16_t)__b, 2); } |
| 3899 | __ai uint8x8_t vrsubhn_u16(uint16x8_t __a, uint16x8_t __b) { |
| 3900 | return (uint8x8_t)__builtin_neon_vrsubhn_v((int8x16_t)__a, (int8x16_t)__b, 16); } |
| 3901 | __ai uint16x4_t vrsubhn_u32(uint32x4_t __a, uint32x4_t __b) { |
| 3902 | return (uint16x4_t)__builtin_neon_vrsubhn_v((int8x16_t)__a, (int8x16_t)__b, 17); } |
| 3903 | __ai uint32x2_t vrsubhn_u64(uint64x2_t __a, uint64x2_t __b) { |
| 3904 | return (uint32x2_t)__builtin_neon_vrsubhn_v((int8x16_t)__a, (int8x16_t)__b, 18); } |
| 3905 | |
| 3906 | #define vset_lane_u8(a, b, __c) __extension__ ({ \ |
| 3907 | uint8_t __a = (a); uint8x8_t __b = (b); \ |
| 3908 | (uint8x8_t)__builtin_neon_vset_lane_i8(__a, (int8x8_t)__b, __c); }) |
| 3909 | #define vset_lane_u16(a, b, __c) __extension__ ({ \ |
| 3910 | uint16_t __a = (a); uint16x4_t __b = (b); \ |
| 3911 | (uint16x4_t)__builtin_neon_vset_lane_i16(__a, (int16x4_t)__b, __c); }) |
| 3912 | #define vset_lane_u32(a, b, __c) __extension__ ({ \ |
| 3913 | uint32_t __a = (a); uint32x2_t __b = (b); \ |
| 3914 | (uint32x2_t)__builtin_neon_vset_lane_i32(__a, (int32x2_t)__b, __c); }) |
| 3915 | #define vset_lane_s8(a, b, __c) __extension__ ({ \ |
| 3916 | int8_t __a = (a); int8x8_t __b = (b); \ |
| 3917 | (int8x8_t)__builtin_neon_vset_lane_i8(__a, __b, __c); }) |
| 3918 | #define vset_lane_s16(a, b, __c) __extension__ ({ \ |
| 3919 | int16_t __a = (a); int16x4_t __b = (b); \ |
| 3920 | (int16x4_t)__builtin_neon_vset_lane_i16(__a, __b, __c); }) |
| 3921 | #define vset_lane_s32(a, b, __c) __extension__ ({ \ |
| 3922 | int32_t __a = (a); int32x2_t __b = (b); \ |
| 3923 | (int32x2_t)__builtin_neon_vset_lane_i32(__a, __b, __c); }) |
| 3924 | #define vset_lane_p8(a, b, __c) __extension__ ({ \ |
| 3925 | poly8_t __a = (a); poly8x8_t __b = (b); \ |
| 3926 | (poly8x8_t)__builtin_neon_vset_lane_i8(__a, (int8x8_t)__b, __c); }) |
| 3927 | #define vset_lane_p16(a, b, __c) __extension__ ({ \ |
| 3928 | poly16_t __a = (a); poly16x4_t __b = (b); \ |
| 3929 | (poly16x4_t)__builtin_neon_vset_lane_i16(__a, (int16x4_t)__b, __c); }) |
| 3930 | #define vset_lane_f32(a, b, __c) __extension__ ({ \ |
| 3931 | float32_t __a = (a); float32x2_t __b = (b); \ |
| 3932 | (float32x2_t)__builtin_neon_vset_lane_f32(__a, __b, __c); }) |
| 3933 | #define vsetq_lane_u8(a, b, __c) __extension__ ({ \ |
| 3934 | uint8_t __a = (a); uint8x16_t __b = (b); \ |
| 3935 | (uint8x16_t)__builtin_neon_vsetq_lane_i8(__a, (int8x16_t)__b, __c); }) |
| 3936 | #define vsetq_lane_u16(a, b, __c) __extension__ ({ \ |
| 3937 | uint16_t __a = (a); uint16x8_t __b = (b); \ |
| 3938 | (uint16x8_t)__builtin_neon_vsetq_lane_i16(__a, (int16x8_t)__b, __c); }) |
| 3939 | #define vsetq_lane_u32(a, b, __c) __extension__ ({ \ |
| 3940 | uint32_t __a = (a); uint32x4_t __b = (b); \ |
| 3941 | (uint32x4_t)__builtin_neon_vsetq_lane_i32(__a, (int32x4_t)__b, __c); }) |
| 3942 | #define vsetq_lane_s8(a, b, __c) __extension__ ({ \ |
| 3943 | int8_t __a = (a); int8x16_t __b = (b); \ |
| 3944 | (int8x16_t)__builtin_neon_vsetq_lane_i8(__a, __b, __c); }) |
| 3945 | #define vsetq_lane_s16(a, b, __c) __extension__ ({ \ |
| 3946 | int16_t __a = (a); int16x8_t __b = (b); \ |
| 3947 | (int16x8_t)__builtin_neon_vsetq_lane_i16(__a, __b, __c); }) |
| 3948 | #define vsetq_lane_s32(a, b, __c) __extension__ ({ \ |
| 3949 | int32_t __a = (a); int32x4_t __b = (b); \ |
| 3950 | (int32x4_t)__builtin_neon_vsetq_lane_i32(__a, __b, __c); }) |
| 3951 | #define vsetq_lane_p8(a, b, __c) __extension__ ({ \ |
| 3952 | poly8_t __a = (a); poly8x16_t __b = (b); \ |
| 3953 | (poly8x16_t)__builtin_neon_vsetq_lane_i8(__a, (int8x16_t)__b, __c); }) |
| 3954 | #define vsetq_lane_p16(a, b, __c) __extension__ ({ \ |
| 3955 | poly16_t __a = (a); poly16x8_t __b = (b); \ |
| 3956 | (poly16x8_t)__builtin_neon_vsetq_lane_i16(__a, (int16x8_t)__b, __c); }) |
| 3957 | #define vsetq_lane_f32(a, b, __c) __extension__ ({ \ |
| 3958 | float32_t __a = (a); float32x4_t __b = (b); \ |
| 3959 | (float32x4_t)__builtin_neon_vsetq_lane_f32(__a, __b, __c); }) |
| 3960 | #define vset_lane_s64(a, b, __c) __extension__ ({ \ |
| 3961 | int64_t __a = (a); int64x1_t __b = (b); \ |
| 3962 | (int64x1_t)__builtin_neon_vset_lane_i64(__a, __b, __c); }) |
| 3963 | #define vset_lane_u64(a, b, __c) __extension__ ({ \ |
| 3964 | uint64_t __a = (a); uint64x1_t __b = (b); \ |
| 3965 | (uint64x1_t)__builtin_neon_vset_lane_i64(__a, (int64x1_t)__b, __c); }) |
| 3966 | #define vsetq_lane_s64(a, b, __c) __extension__ ({ \ |
| 3967 | int64_t __a = (a); int64x2_t __b = (b); \ |
| 3968 | (int64x2_t)__builtin_neon_vsetq_lane_i64(__a, __b, __c); }) |
| 3969 | #define vsetq_lane_u64(a, b, __c) __extension__ ({ \ |
| 3970 | uint64_t __a = (a); uint64x2_t __b = (b); \ |
| 3971 | (uint64x2_t)__builtin_neon_vsetq_lane_i64(__a, (int64x2_t)__b, __c); }) |
| 3972 | |
| 3973 | __ai int8x8_t vshl_s8(int8x8_t __a, int8x8_t __b) { |
| 3974 | return (int8x8_t)__builtin_neon_vshl_v(__a, __b, 0); } |
| 3975 | __ai int16x4_t vshl_s16(int16x4_t __a, int16x4_t __b) { |
| 3976 | return (int16x4_t)__builtin_neon_vshl_v((int8x8_t)__a, (int8x8_t)__b, 1); } |
| 3977 | __ai int32x2_t vshl_s32(int32x2_t __a, int32x2_t __b) { |
| 3978 | return (int32x2_t)__builtin_neon_vshl_v((int8x8_t)__a, (int8x8_t)__b, 2); } |
| 3979 | __ai int64x1_t vshl_s64(int64x1_t __a, int64x1_t __b) { |
| 3980 | return (int64x1_t)__builtin_neon_vshl_v((int8x8_t)__a, (int8x8_t)__b, 3); } |
| 3981 | __ai uint8x8_t vshl_u8(uint8x8_t __a, int8x8_t __b) { |
| 3982 | return (uint8x8_t)__builtin_neon_vshl_v((int8x8_t)__a, __b, 16); } |
| 3983 | __ai uint16x4_t vshl_u16(uint16x4_t __a, int16x4_t __b) { |
| 3984 | return (uint16x4_t)__builtin_neon_vshl_v((int8x8_t)__a, (int8x8_t)__b, 17); } |
| 3985 | __ai uint32x2_t vshl_u32(uint32x2_t __a, int32x2_t __b) { |
| 3986 | return (uint32x2_t)__builtin_neon_vshl_v((int8x8_t)__a, (int8x8_t)__b, 18); } |
| 3987 | __ai uint64x1_t vshl_u64(uint64x1_t __a, int64x1_t __b) { |
| 3988 | return (uint64x1_t)__builtin_neon_vshl_v((int8x8_t)__a, (int8x8_t)__b, 19); } |
| 3989 | __ai int8x16_t vshlq_s8(int8x16_t __a, int8x16_t __b) { |
| 3990 | return (int8x16_t)__builtin_neon_vshlq_v(__a, __b, 32); } |
| 3991 | __ai int16x8_t vshlq_s16(int16x8_t __a, int16x8_t __b) { |
| 3992 | return (int16x8_t)__builtin_neon_vshlq_v((int8x16_t)__a, (int8x16_t)__b, 33); } |
| 3993 | __ai int32x4_t vshlq_s32(int32x4_t __a, int32x4_t __b) { |
| 3994 | return (int32x4_t)__builtin_neon_vshlq_v((int8x16_t)__a, (int8x16_t)__b, 34); } |
| 3995 | __ai int64x2_t vshlq_s64(int64x2_t __a, int64x2_t __b) { |
| 3996 | return (int64x2_t)__builtin_neon_vshlq_v((int8x16_t)__a, (int8x16_t)__b, 35); } |
| 3997 | __ai uint8x16_t vshlq_u8(uint8x16_t __a, int8x16_t __b) { |
| 3998 | return (uint8x16_t)__builtin_neon_vshlq_v((int8x16_t)__a, __b, 48); } |
| 3999 | __ai uint16x8_t vshlq_u16(uint16x8_t __a, int16x8_t __b) { |
| 4000 | return (uint16x8_t)__builtin_neon_vshlq_v((int8x16_t)__a, (int8x16_t)__b, 49); } |
| 4001 | __ai uint32x4_t vshlq_u32(uint32x4_t __a, int32x4_t __b) { |
| 4002 | return (uint32x4_t)__builtin_neon_vshlq_v((int8x16_t)__a, (int8x16_t)__b, 50); } |
| 4003 | __ai uint64x2_t vshlq_u64(uint64x2_t __a, int64x2_t __b) { |
| 4004 | return (uint64x2_t)__builtin_neon_vshlq_v((int8x16_t)__a, (int8x16_t)__b, 51); } |
| 4005 | |
| 4006 | #define vshll_n_s8(a, __b) __extension__ ({ \ |
| 4007 | int8x8_t __a = (a); \ |
| 4008 | (int16x8_t)__builtin_neon_vshll_n_v(__a, __b, 33); }) |
| 4009 | #define vshll_n_s16(a, __b) __extension__ ({ \ |
| 4010 | int16x4_t __a = (a); \ |
| 4011 | (int32x4_t)__builtin_neon_vshll_n_v((int8x8_t)__a, __b, 34); }) |
| 4012 | #define vshll_n_s32(a, __b) __extension__ ({ \ |
| 4013 | int32x2_t __a = (a); \ |
| 4014 | (int64x2_t)__builtin_neon_vshll_n_v((int8x8_t)__a, __b, 35); }) |
| 4015 | #define vshll_n_u8(a, __b) __extension__ ({ \ |
| 4016 | uint8x8_t __a = (a); \ |
| 4017 | (uint16x8_t)__builtin_neon_vshll_n_v((int8x8_t)__a, __b, 49); }) |
| 4018 | #define vshll_n_u16(a, __b) __extension__ ({ \ |
| 4019 | uint16x4_t __a = (a); \ |
| 4020 | (uint32x4_t)__builtin_neon_vshll_n_v((int8x8_t)__a, __b, 50); }) |
| 4021 | #define vshll_n_u32(a, __b) __extension__ ({ \ |
| 4022 | uint32x2_t __a = (a); \ |
| 4023 | (uint64x2_t)__builtin_neon_vshll_n_v((int8x8_t)__a, __b, 51); }) |
| 4024 | |
| 4025 | #define vshl_n_s8(a, __b) __extension__ ({ \ |
| 4026 | int8x8_t __a = (a); \ |
| 4027 | (int8x8_t)__builtin_neon_vshl_n_v(__a, __b, 0); }) |
| 4028 | #define vshl_n_s16(a, __b) __extension__ ({ \ |
| 4029 | int16x4_t __a = (a); \ |
| 4030 | (int16x4_t)__builtin_neon_vshl_n_v((int8x8_t)__a, __b, 1); }) |
| 4031 | #define vshl_n_s32(a, __b) __extension__ ({ \ |
| 4032 | int32x2_t __a = (a); \ |
| 4033 | (int32x2_t)__builtin_neon_vshl_n_v((int8x8_t)__a, __b, 2); }) |
| 4034 | #define vshl_n_s64(a, __b) __extension__ ({ \ |
| 4035 | int64x1_t __a = (a); \ |
| 4036 | (int64x1_t)__builtin_neon_vshl_n_v((int8x8_t)__a, __b, 3); }) |
| 4037 | #define vshl_n_u8(a, __b) __extension__ ({ \ |
| 4038 | uint8x8_t __a = (a); \ |
| 4039 | (uint8x8_t)__builtin_neon_vshl_n_v((int8x8_t)__a, __b, 16); }) |
| 4040 | #define vshl_n_u16(a, __b) __extension__ ({ \ |
| 4041 | uint16x4_t __a = (a); \ |
| 4042 | (uint16x4_t)__builtin_neon_vshl_n_v((int8x8_t)__a, __b, 17); }) |
| 4043 | #define vshl_n_u32(a, __b) __extension__ ({ \ |
| 4044 | uint32x2_t __a = (a); \ |
| 4045 | (uint32x2_t)__builtin_neon_vshl_n_v((int8x8_t)__a, __b, 18); }) |
| 4046 | #define vshl_n_u64(a, __b) __extension__ ({ \ |
| 4047 | uint64x1_t __a = (a); \ |
| 4048 | (uint64x1_t)__builtin_neon_vshl_n_v((int8x8_t)__a, __b, 19); }) |
| 4049 | #define vshlq_n_s8(a, __b) __extension__ ({ \ |
| 4050 | int8x16_t __a = (a); \ |
| 4051 | (int8x16_t)__builtin_neon_vshlq_n_v(__a, __b, 32); }) |
| 4052 | #define vshlq_n_s16(a, __b) __extension__ ({ \ |
| 4053 | int16x8_t __a = (a); \ |
| 4054 | (int16x8_t)__builtin_neon_vshlq_n_v((int8x16_t)__a, __b, 33); }) |
| 4055 | #define vshlq_n_s32(a, __b) __extension__ ({ \ |
| 4056 | int32x4_t __a = (a); \ |
| 4057 | (int32x4_t)__builtin_neon_vshlq_n_v((int8x16_t)__a, __b, 34); }) |
| 4058 | #define vshlq_n_s64(a, __b) __extension__ ({ \ |
| 4059 | int64x2_t __a = (a); \ |
| 4060 | (int64x2_t)__builtin_neon_vshlq_n_v((int8x16_t)__a, __b, 35); }) |
| 4061 | #define vshlq_n_u8(a, __b) __extension__ ({ \ |
| 4062 | uint8x16_t __a = (a); \ |
| 4063 | (uint8x16_t)__builtin_neon_vshlq_n_v((int8x16_t)__a, __b, 48); }) |
| 4064 | #define vshlq_n_u16(a, __b) __extension__ ({ \ |
| 4065 | uint16x8_t __a = (a); \ |
| 4066 | (uint16x8_t)__builtin_neon_vshlq_n_v((int8x16_t)__a, __b, 49); }) |
| 4067 | #define vshlq_n_u32(a, __b) __extension__ ({ \ |
| 4068 | uint32x4_t __a = (a); \ |
| 4069 | (uint32x4_t)__builtin_neon_vshlq_n_v((int8x16_t)__a, __b, 50); }) |
| 4070 | #define vshlq_n_u64(a, __b) __extension__ ({ \ |
| 4071 | uint64x2_t __a = (a); \ |
| 4072 | (uint64x2_t)__builtin_neon_vshlq_n_v((int8x16_t)__a, __b, 51); }) |
| 4073 | |
| 4074 | #define vshrn_n_s16(a, __b) __extension__ ({ \ |
| 4075 | int16x8_t __a = (a); \ |
| 4076 | (int8x8_t)__builtin_neon_vshrn_n_v((int8x16_t)__a, __b, 0); }) |
| 4077 | #define vshrn_n_s32(a, __b) __extension__ ({ \ |
| 4078 | int32x4_t __a = (a); \ |
| 4079 | (int16x4_t)__builtin_neon_vshrn_n_v((int8x16_t)__a, __b, 1); }) |
| 4080 | #define vshrn_n_s64(a, __b) __extension__ ({ \ |
| 4081 | int64x2_t __a = (a); \ |
| 4082 | (int32x2_t)__builtin_neon_vshrn_n_v((int8x16_t)__a, __b, 2); }) |
| 4083 | #define vshrn_n_u16(a, __b) __extension__ ({ \ |
| 4084 | uint16x8_t __a = (a); \ |
| 4085 | (uint8x8_t)__builtin_neon_vshrn_n_v((int8x16_t)__a, __b, 16); }) |
| 4086 | #define vshrn_n_u32(a, __b) __extension__ ({ \ |
| 4087 | uint32x4_t __a = (a); \ |
| 4088 | (uint16x4_t)__builtin_neon_vshrn_n_v((int8x16_t)__a, __b, 17); }) |
| 4089 | #define vshrn_n_u64(a, __b) __extension__ ({ \ |
| 4090 | uint64x2_t __a = (a); \ |
| 4091 | (uint32x2_t)__builtin_neon_vshrn_n_v((int8x16_t)__a, __b, 18); }) |
| 4092 | |
| 4093 | #define vshr_n_s8(a, __b) __extension__ ({ \ |
| 4094 | int8x8_t __a = (a); \ |
| 4095 | (int8x8_t)__builtin_neon_vshr_n_v(__a, __b, 0); }) |
| 4096 | #define vshr_n_s16(a, __b) __extension__ ({ \ |
| 4097 | int16x4_t __a = (a); \ |
| 4098 | (int16x4_t)__builtin_neon_vshr_n_v((int8x8_t)__a, __b, 1); }) |
| 4099 | #define vshr_n_s32(a, __b) __extension__ ({ \ |
| 4100 | int32x2_t __a = (a); \ |
| 4101 | (int32x2_t)__builtin_neon_vshr_n_v((int8x8_t)__a, __b, 2); }) |
| 4102 | #define vshr_n_s64(a, __b) __extension__ ({ \ |
| 4103 | int64x1_t __a = (a); \ |
| 4104 | (int64x1_t)__builtin_neon_vshr_n_v((int8x8_t)__a, __b, 3); }) |
| 4105 | #define vshr_n_u8(a, __b) __extension__ ({ \ |
| 4106 | uint8x8_t __a = (a); \ |
| 4107 | (uint8x8_t)__builtin_neon_vshr_n_v((int8x8_t)__a, __b, 16); }) |
| 4108 | #define vshr_n_u16(a, __b) __extension__ ({ \ |
| 4109 | uint16x4_t __a = (a); \ |
| 4110 | (uint16x4_t)__builtin_neon_vshr_n_v((int8x8_t)__a, __b, 17); }) |
| 4111 | #define vshr_n_u32(a, __b) __extension__ ({ \ |
| 4112 | uint32x2_t __a = (a); \ |
| 4113 | (uint32x2_t)__builtin_neon_vshr_n_v((int8x8_t)__a, __b, 18); }) |
| 4114 | #define vshr_n_u64(a, __b) __extension__ ({ \ |
| 4115 | uint64x1_t __a = (a); \ |
| 4116 | (uint64x1_t)__builtin_neon_vshr_n_v((int8x8_t)__a, __b, 19); }) |
| 4117 | #define vshrq_n_s8(a, __b) __extension__ ({ \ |
| 4118 | int8x16_t __a = (a); \ |
| 4119 | (int8x16_t)__builtin_neon_vshrq_n_v(__a, __b, 32); }) |
| 4120 | #define vshrq_n_s16(a, __b) __extension__ ({ \ |
| 4121 | int16x8_t __a = (a); \ |
| 4122 | (int16x8_t)__builtin_neon_vshrq_n_v((int8x16_t)__a, __b, 33); }) |
| 4123 | #define vshrq_n_s32(a, __b) __extension__ ({ \ |
| 4124 | int32x4_t __a = (a); \ |
| 4125 | (int32x4_t)__builtin_neon_vshrq_n_v((int8x16_t)__a, __b, 34); }) |
| 4126 | #define vshrq_n_s64(a, __b) __extension__ ({ \ |
| 4127 | int64x2_t __a = (a); \ |
| 4128 | (int64x2_t)__builtin_neon_vshrq_n_v((int8x16_t)__a, __b, 35); }) |
| 4129 | #define vshrq_n_u8(a, __b) __extension__ ({ \ |
| 4130 | uint8x16_t __a = (a); \ |
| 4131 | (uint8x16_t)__builtin_neon_vshrq_n_v((int8x16_t)__a, __b, 48); }) |
| 4132 | #define vshrq_n_u16(a, __b) __extension__ ({ \ |
| 4133 | uint16x8_t __a = (a); \ |
| 4134 | (uint16x8_t)__builtin_neon_vshrq_n_v((int8x16_t)__a, __b, 49); }) |
| 4135 | #define vshrq_n_u32(a, __b) __extension__ ({ \ |
| 4136 | uint32x4_t __a = (a); \ |
| 4137 | (uint32x4_t)__builtin_neon_vshrq_n_v((int8x16_t)__a, __b, 50); }) |
| 4138 | #define vshrq_n_u64(a, __b) __extension__ ({ \ |
| 4139 | uint64x2_t __a = (a); \ |
| 4140 | (uint64x2_t)__builtin_neon_vshrq_n_v((int8x16_t)__a, __b, 51); }) |
| 4141 | |
| 4142 | #define vsli_n_s8(a, b, __c) __extension__ ({ \ |
| 4143 | int8x8_t __a = (a); int8x8_t __b = (b); \ |
| 4144 | (int8x8_t)__builtin_neon_vsli_n_v(__a, __b, __c, 0); }) |
| 4145 | #define vsli_n_s16(a, b, __c) __extension__ ({ \ |
| 4146 | int16x4_t __a = (a); int16x4_t __b = (b); \ |
| 4147 | (int16x4_t)__builtin_neon_vsli_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 1); }) |
| 4148 | #define vsli_n_s32(a, b, __c) __extension__ ({ \ |
| 4149 | int32x2_t __a = (a); int32x2_t __b = (b); \ |
| 4150 | (int32x2_t)__builtin_neon_vsli_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 2); }) |
| 4151 | #define vsli_n_s64(a, b, __c) __extension__ ({ \ |
| 4152 | int64x1_t __a = (a); int64x1_t __b = (b); \ |
| 4153 | (int64x1_t)__builtin_neon_vsli_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 3); }) |
| 4154 | #define vsli_n_u8(a, b, __c) __extension__ ({ \ |
| 4155 | uint8x8_t __a = (a); uint8x8_t __b = (b); \ |
| 4156 | (uint8x8_t)__builtin_neon_vsli_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 16); }) |
| 4157 | #define vsli_n_u16(a, b, __c) __extension__ ({ \ |
| 4158 | uint16x4_t __a = (a); uint16x4_t __b = (b); \ |
| 4159 | (uint16x4_t)__builtin_neon_vsli_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 17); }) |
| 4160 | #define vsli_n_u32(a, b, __c) __extension__ ({ \ |
| 4161 | uint32x2_t __a = (a); uint32x2_t __b = (b); \ |
| 4162 | (uint32x2_t)__builtin_neon_vsli_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 18); }) |
| 4163 | #define vsli_n_u64(a, b, __c) __extension__ ({ \ |
| 4164 | uint64x1_t __a = (a); uint64x1_t __b = (b); \ |
| 4165 | (uint64x1_t)__builtin_neon_vsli_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 19); }) |
| 4166 | #define vsli_n_p8(a, b, __c) __extension__ ({ \ |
| 4167 | poly8x8_t __a = (a); poly8x8_t __b = (b); \ |
| 4168 | (poly8x8_t)__builtin_neon_vsli_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 4); }) |
| 4169 | #define vsli_n_p16(a, b, __c) __extension__ ({ \ |
| 4170 | poly16x4_t __a = (a); poly16x4_t __b = (b); \ |
| 4171 | (poly16x4_t)__builtin_neon_vsli_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 5); }) |
| 4172 | #define vsliq_n_s8(a, b, __c) __extension__ ({ \ |
| 4173 | int8x16_t __a = (a); int8x16_t __b = (b); \ |
| 4174 | (int8x16_t)__builtin_neon_vsliq_n_v(__a, __b, __c, 32); }) |
| 4175 | #define vsliq_n_s16(a, b, __c) __extension__ ({ \ |
| 4176 | int16x8_t __a = (a); int16x8_t __b = (b); \ |
| 4177 | (int16x8_t)__builtin_neon_vsliq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 33); }) |
| 4178 | #define vsliq_n_s32(a, b, __c) __extension__ ({ \ |
| 4179 | int32x4_t __a = (a); int32x4_t __b = (b); \ |
| 4180 | (int32x4_t)__builtin_neon_vsliq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 34); }) |
| 4181 | #define vsliq_n_s64(a, b, __c) __extension__ ({ \ |
| 4182 | int64x2_t __a = (a); int64x2_t __b = (b); \ |
| 4183 | (int64x2_t)__builtin_neon_vsliq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 35); }) |
| 4184 | #define vsliq_n_u8(a, b, __c) __extension__ ({ \ |
| 4185 | uint8x16_t __a = (a); uint8x16_t __b = (b); \ |
| 4186 | (uint8x16_t)__builtin_neon_vsliq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 48); }) |
| 4187 | #define vsliq_n_u16(a, b, __c) __extension__ ({ \ |
| 4188 | uint16x8_t __a = (a); uint16x8_t __b = (b); \ |
| 4189 | (uint16x8_t)__builtin_neon_vsliq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 49); }) |
| 4190 | #define vsliq_n_u32(a, b, __c) __extension__ ({ \ |
| 4191 | uint32x4_t __a = (a); uint32x4_t __b = (b); \ |
| 4192 | (uint32x4_t)__builtin_neon_vsliq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 50); }) |
| 4193 | #define vsliq_n_u64(a, b, __c) __extension__ ({ \ |
| 4194 | uint64x2_t __a = (a); uint64x2_t __b = (b); \ |
| 4195 | (uint64x2_t)__builtin_neon_vsliq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 51); }) |
| 4196 | #define vsliq_n_p8(a, b, __c) __extension__ ({ \ |
| 4197 | poly8x16_t __a = (a); poly8x16_t __b = (b); \ |
| 4198 | (poly8x16_t)__builtin_neon_vsliq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 36); }) |
| 4199 | #define vsliq_n_p16(a, b, __c) __extension__ ({ \ |
| 4200 | poly16x8_t __a = (a); poly16x8_t __b = (b); \ |
| 4201 | (poly16x8_t)__builtin_neon_vsliq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 37); }) |
| 4202 | |
| 4203 | #define vsra_n_s8(a, b, __c) __extension__ ({ \ |
| 4204 | int8x8_t __a = (a); int8x8_t __b = (b); \ |
| 4205 | (int8x8_t)__builtin_neon_vsra_n_v(__a, __b, __c, 0); }) |
| 4206 | #define vsra_n_s16(a, b, __c) __extension__ ({ \ |
| 4207 | int16x4_t __a = (a); int16x4_t __b = (b); \ |
| 4208 | (int16x4_t)__builtin_neon_vsra_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 1); }) |
| 4209 | #define vsra_n_s32(a, b, __c) __extension__ ({ \ |
| 4210 | int32x2_t __a = (a); int32x2_t __b = (b); \ |
| 4211 | (int32x2_t)__builtin_neon_vsra_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 2); }) |
| 4212 | #define vsra_n_s64(a, b, __c) __extension__ ({ \ |
| 4213 | int64x1_t __a = (a); int64x1_t __b = (b); \ |
| 4214 | (int64x1_t)__builtin_neon_vsra_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 3); }) |
| 4215 | #define vsra_n_u8(a, b, __c) __extension__ ({ \ |
| 4216 | uint8x8_t __a = (a); uint8x8_t __b = (b); \ |
| 4217 | (uint8x8_t)__builtin_neon_vsra_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 16); }) |
| 4218 | #define vsra_n_u16(a, b, __c) __extension__ ({ \ |
| 4219 | uint16x4_t __a = (a); uint16x4_t __b = (b); \ |
| 4220 | (uint16x4_t)__builtin_neon_vsra_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 17); }) |
| 4221 | #define vsra_n_u32(a, b, __c) __extension__ ({ \ |
| 4222 | uint32x2_t __a = (a); uint32x2_t __b = (b); \ |
| 4223 | (uint32x2_t)__builtin_neon_vsra_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 18); }) |
| 4224 | #define vsra_n_u64(a, b, __c) __extension__ ({ \ |
| 4225 | uint64x1_t __a = (a); uint64x1_t __b = (b); \ |
| 4226 | (uint64x1_t)__builtin_neon_vsra_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 19); }) |
| 4227 | #define vsraq_n_s8(a, b, __c) __extension__ ({ \ |
| 4228 | int8x16_t __a = (a); int8x16_t __b = (b); \ |
| 4229 | (int8x16_t)__builtin_neon_vsraq_n_v(__a, __b, __c, 32); }) |
| 4230 | #define vsraq_n_s16(a, b, __c) __extension__ ({ \ |
| 4231 | int16x8_t __a = (a); int16x8_t __b = (b); \ |
| 4232 | (int16x8_t)__builtin_neon_vsraq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 33); }) |
| 4233 | #define vsraq_n_s32(a, b, __c) __extension__ ({ \ |
| 4234 | int32x4_t __a = (a); int32x4_t __b = (b); \ |
| 4235 | (int32x4_t)__builtin_neon_vsraq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 34); }) |
| 4236 | #define vsraq_n_s64(a, b, __c) __extension__ ({ \ |
| 4237 | int64x2_t __a = (a); int64x2_t __b = (b); \ |
| 4238 | (int64x2_t)__builtin_neon_vsraq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 35); }) |
| 4239 | #define vsraq_n_u8(a, b, __c) __extension__ ({ \ |
| 4240 | uint8x16_t __a = (a); uint8x16_t __b = (b); \ |
| 4241 | (uint8x16_t)__builtin_neon_vsraq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 48); }) |
| 4242 | #define vsraq_n_u16(a, b, __c) __extension__ ({ \ |
| 4243 | uint16x8_t __a = (a); uint16x8_t __b = (b); \ |
| 4244 | (uint16x8_t)__builtin_neon_vsraq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 49); }) |
| 4245 | #define vsraq_n_u32(a, b, __c) __extension__ ({ \ |
| 4246 | uint32x4_t __a = (a); uint32x4_t __b = (b); \ |
| 4247 | (uint32x4_t)__builtin_neon_vsraq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 50); }) |
| 4248 | #define vsraq_n_u64(a, b, __c) __extension__ ({ \ |
| 4249 | uint64x2_t __a = (a); uint64x2_t __b = (b); \ |
| 4250 | (uint64x2_t)__builtin_neon_vsraq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 51); }) |
| 4251 | |
| 4252 | #define vsri_n_s8(a, b, __c) __extension__ ({ \ |
| 4253 | int8x8_t __a = (a); int8x8_t __b = (b); \ |
| 4254 | (int8x8_t)__builtin_neon_vsri_n_v(__a, __b, __c, 0); }) |
| 4255 | #define vsri_n_s16(a, b, __c) __extension__ ({ \ |
| 4256 | int16x4_t __a = (a); int16x4_t __b = (b); \ |
| 4257 | (int16x4_t)__builtin_neon_vsri_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 1); }) |
| 4258 | #define vsri_n_s32(a, b, __c) __extension__ ({ \ |
| 4259 | int32x2_t __a = (a); int32x2_t __b = (b); \ |
| 4260 | (int32x2_t)__builtin_neon_vsri_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 2); }) |
| 4261 | #define vsri_n_s64(a, b, __c) __extension__ ({ \ |
| 4262 | int64x1_t __a = (a); int64x1_t __b = (b); \ |
| 4263 | (int64x1_t)__builtin_neon_vsri_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 3); }) |
| 4264 | #define vsri_n_u8(a, b, __c) __extension__ ({ \ |
| 4265 | uint8x8_t __a = (a); uint8x8_t __b = (b); \ |
| 4266 | (uint8x8_t)__builtin_neon_vsri_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 16); }) |
| 4267 | #define vsri_n_u16(a, b, __c) __extension__ ({ \ |
| 4268 | uint16x4_t __a = (a); uint16x4_t __b = (b); \ |
| 4269 | (uint16x4_t)__builtin_neon_vsri_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 17); }) |
| 4270 | #define vsri_n_u32(a, b, __c) __extension__ ({ \ |
| 4271 | uint32x2_t __a = (a); uint32x2_t __b = (b); \ |
| 4272 | (uint32x2_t)__builtin_neon_vsri_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 18); }) |
| 4273 | #define vsri_n_u64(a, b, __c) __extension__ ({ \ |
| 4274 | uint64x1_t __a = (a); uint64x1_t __b = (b); \ |
| 4275 | (uint64x1_t)__builtin_neon_vsri_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 19); }) |
| 4276 | #define vsri_n_p8(a, b, __c) __extension__ ({ \ |
| 4277 | poly8x8_t __a = (a); poly8x8_t __b = (b); \ |
| 4278 | (poly8x8_t)__builtin_neon_vsri_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 4); }) |
| 4279 | #define vsri_n_p16(a, b, __c) __extension__ ({ \ |
| 4280 | poly16x4_t __a = (a); poly16x4_t __b = (b); \ |
| 4281 | (poly16x4_t)__builtin_neon_vsri_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 5); }) |
| 4282 | #define vsriq_n_s8(a, b, __c) __extension__ ({ \ |
| 4283 | int8x16_t __a = (a); int8x16_t __b = (b); \ |
| 4284 | (int8x16_t)__builtin_neon_vsriq_n_v(__a, __b, __c, 32); }) |
| 4285 | #define vsriq_n_s16(a, b, __c) __extension__ ({ \ |
| 4286 | int16x8_t __a = (a); int16x8_t __b = (b); \ |
| 4287 | (int16x8_t)__builtin_neon_vsriq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 33); }) |
| 4288 | #define vsriq_n_s32(a, b, __c) __extension__ ({ \ |
| 4289 | int32x4_t __a = (a); int32x4_t __b = (b); \ |
| 4290 | (int32x4_t)__builtin_neon_vsriq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 34); }) |
| 4291 | #define vsriq_n_s64(a, b, __c) __extension__ ({ \ |
| 4292 | int64x2_t __a = (a); int64x2_t __b = (b); \ |
| 4293 | (int64x2_t)__builtin_neon_vsriq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 35); }) |
| 4294 | #define vsriq_n_u8(a, b, __c) __extension__ ({ \ |
| 4295 | uint8x16_t __a = (a); uint8x16_t __b = (b); \ |
| 4296 | (uint8x16_t)__builtin_neon_vsriq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 48); }) |
| 4297 | #define vsriq_n_u16(a, b, __c) __extension__ ({ \ |
| 4298 | uint16x8_t __a = (a); uint16x8_t __b = (b); \ |
| 4299 | (uint16x8_t)__builtin_neon_vsriq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 49); }) |
| 4300 | #define vsriq_n_u32(a, b, __c) __extension__ ({ \ |
| 4301 | uint32x4_t __a = (a); uint32x4_t __b = (b); \ |
| 4302 | (uint32x4_t)__builtin_neon_vsriq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 50); }) |
| 4303 | #define vsriq_n_u64(a, b, __c) __extension__ ({ \ |
| 4304 | uint64x2_t __a = (a); uint64x2_t __b = (b); \ |
| 4305 | (uint64x2_t)__builtin_neon_vsriq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 51); }) |
| 4306 | #define vsriq_n_p8(a, b, __c) __extension__ ({ \ |
| 4307 | poly8x16_t __a = (a); poly8x16_t __b = (b); \ |
| 4308 | (poly8x16_t)__builtin_neon_vsriq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 36); }) |
| 4309 | #define vsriq_n_p16(a, b, __c) __extension__ ({ \ |
| 4310 | poly16x8_t __a = (a); poly16x8_t __b = (b); \ |
| 4311 | (poly16x8_t)__builtin_neon_vsriq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 37); }) |
| 4312 | |
| 4313 | #define vst1q_u8(__a, b) __extension__ ({ \ |
| 4314 | uint8x16_t __b = (b); \ |
| 4315 | __builtin_neon_vst1q_v(__a, (int8x16_t)__b, 48); }) |
| 4316 | #define vst1q_u16(__a, b) __extension__ ({ \ |
| 4317 | uint16x8_t __b = (b); \ |
| 4318 | __builtin_neon_vst1q_v(__a, (int8x16_t)__b, 49); }) |
| 4319 | #define vst1q_u32(__a, b) __extension__ ({ \ |
| 4320 | uint32x4_t __b = (b); \ |
| 4321 | __builtin_neon_vst1q_v(__a, (int8x16_t)__b, 50); }) |
| 4322 | #define vst1q_u64(__a, b) __extension__ ({ \ |
| 4323 | uint64x2_t __b = (b); \ |
| 4324 | __builtin_neon_vst1q_v(__a, (int8x16_t)__b, 51); }) |
| 4325 | #define vst1q_s8(__a, b) __extension__ ({ \ |
| 4326 | int8x16_t __b = (b); \ |
| 4327 | __builtin_neon_vst1q_v(__a, __b, 32); }) |
| 4328 | #define vst1q_s16(__a, b) __extension__ ({ \ |
| 4329 | int16x8_t __b = (b); \ |
| 4330 | __builtin_neon_vst1q_v(__a, (int8x16_t)__b, 33); }) |
| 4331 | #define vst1q_s32(__a, b) __extension__ ({ \ |
| 4332 | int32x4_t __b = (b); \ |
| 4333 | __builtin_neon_vst1q_v(__a, (int8x16_t)__b, 34); }) |
| 4334 | #define vst1q_s64(__a, b) __extension__ ({ \ |
| 4335 | int64x2_t __b = (b); \ |
| 4336 | __builtin_neon_vst1q_v(__a, (int8x16_t)__b, 35); }) |
| 4337 | #define vst1q_f16(__a, b) __extension__ ({ \ |
| 4338 | float16x8_t __b = (b); \ |
| 4339 | __builtin_neon_vst1q_v(__a, (int8x16_t)__b, 39); }) |
| 4340 | #define vst1q_f32(__a, b) __extension__ ({ \ |
| 4341 | float32x4_t __b = (b); \ |
| 4342 | __builtin_neon_vst1q_v(__a, (int8x16_t)__b, 40); }) |
| 4343 | #define vst1q_p8(__a, b) __extension__ ({ \ |
| 4344 | poly8x16_t __b = (b); \ |
| 4345 | __builtin_neon_vst1q_v(__a, (int8x16_t)__b, 36); }) |
| 4346 | #define vst1q_p16(__a, b) __extension__ ({ \ |
| 4347 | poly16x8_t __b = (b); \ |
| 4348 | __builtin_neon_vst1q_v(__a, (int8x16_t)__b, 37); }) |
| 4349 | #define vst1_u8(__a, b) __extension__ ({ \ |
| 4350 | uint8x8_t __b = (b); \ |
| 4351 | __builtin_neon_vst1_v(__a, (int8x8_t)__b, 16); }) |
| 4352 | #define vst1_u16(__a, b) __extension__ ({ \ |
| 4353 | uint16x4_t __b = (b); \ |
| 4354 | __builtin_neon_vst1_v(__a, (int8x8_t)__b, 17); }) |
| 4355 | #define vst1_u32(__a, b) __extension__ ({ \ |
| 4356 | uint32x2_t __b = (b); \ |
| 4357 | __builtin_neon_vst1_v(__a, (int8x8_t)__b, 18); }) |
| 4358 | #define vst1_u64(__a, b) __extension__ ({ \ |
| 4359 | uint64x1_t __b = (b); \ |
| 4360 | __builtin_neon_vst1_v(__a, (int8x8_t)__b, 19); }) |
| 4361 | #define vst1_s8(__a, b) __extension__ ({ \ |
| 4362 | int8x8_t __b = (b); \ |
| 4363 | __builtin_neon_vst1_v(__a, __b, 0); }) |
| 4364 | #define vst1_s16(__a, b) __extension__ ({ \ |
| 4365 | int16x4_t __b = (b); \ |
| 4366 | __builtin_neon_vst1_v(__a, (int8x8_t)__b, 1); }) |
| 4367 | #define vst1_s32(__a, b) __extension__ ({ \ |
| 4368 | int32x2_t __b = (b); \ |
| 4369 | __builtin_neon_vst1_v(__a, (int8x8_t)__b, 2); }) |
| 4370 | #define vst1_s64(__a, b) __extension__ ({ \ |
| 4371 | int64x1_t __b = (b); \ |
| 4372 | __builtin_neon_vst1_v(__a, (int8x8_t)__b, 3); }) |
| 4373 | #define vst1_f16(__a, b) __extension__ ({ \ |
| 4374 | float16x4_t __b = (b); \ |
| 4375 | __builtin_neon_vst1_v(__a, (int8x8_t)__b, 7); }) |
| 4376 | #define vst1_f32(__a, b) __extension__ ({ \ |
| 4377 | float32x2_t __b = (b); \ |
| 4378 | __builtin_neon_vst1_v(__a, (int8x8_t)__b, 8); }) |
| 4379 | #define vst1_p8(__a, b) __extension__ ({ \ |
| 4380 | poly8x8_t __b = (b); \ |
| 4381 | __builtin_neon_vst1_v(__a, (int8x8_t)__b, 4); }) |
| 4382 | #define vst1_p16(__a, b) __extension__ ({ \ |
| 4383 | poly16x4_t __b = (b); \ |
| 4384 | __builtin_neon_vst1_v(__a, (int8x8_t)__b, 5); }) |
| 4385 | |
| 4386 | #define vst1q_lane_u8(__a, b, __c) __extension__ ({ \ |
| 4387 | uint8x16_t __b = (b); \ |
| 4388 | __builtin_neon_vst1q_lane_v(__a, (int8x16_t)__b, __c, 48); }) |
| 4389 | #define vst1q_lane_u16(__a, b, __c) __extension__ ({ \ |
| 4390 | uint16x8_t __b = (b); \ |
| 4391 | __builtin_neon_vst1q_lane_v(__a, (int8x16_t)__b, __c, 49); }) |
| 4392 | #define vst1q_lane_u32(__a, b, __c) __extension__ ({ \ |
| 4393 | uint32x4_t __b = (b); \ |
| 4394 | __builtin_neon_vst1q_lane_v(__a, (int8x16_t)__b, __c, 50); }) |
| 4395 | #define vst1q_lane_u64(__a, b, __c) __extension__ ({ \ |
| 4396 | uint64x2_t __b = (b); \ |
| 4397 | __builtin_neon_vst1q_lane_v(__a, (int8x16_t)__b, __c, 51); }) |
| 4398 | #define vst1q_lane_s8(__a, b, __c) __extension__ ({ \ |
| 4399 | int8x16_t __b = (b); \ |
| 4400 | __builtin_neon_vst1q_lane_v(__a, __b, __c, 32); }) |
| 4401 | #define vst1q_lane_s16(__a, b, __c) __extension__ ({ \ |
| 4402 | int16x8_t __b = (b); \ |
| 4403 | __builtin_neon_vst1q_lane_v(__a, (int8x16_t)__b, __c, 33); }) |
| 4404 | #define vst1q_lane_s32(__a, b, __c) __extension__ ({ \ |
| 4405 | int32x4_t __b = (b); \ |
| 4406 | __builtin_neon_vst1q_lane_v(__a, (int8x16_t)__b, __c, 34); }) |
| 4407 | #define vst1q_lane_s64(__a, b, __c) __extension__ ({ \ |
| 4408 | int64x2_t __b = (b); \ |
| 4409 | __builtin_neon_vst1q_lane_v(__a, (int8x16_t)__b, __c, 35); }) |
| 4410 | #define vst1q_lane_f16(__a, b, __c) __extension__ ({ \ |
| 4411 | float16x8_t __b = (b); \ |
| 4412 | __builtin_neon_vst1q_lane_v(__a, (int8x16_t)__b, __c, 39); }) |
| 4413 | #define vst1q_lane_f32(__a, b, __c) __extension__ ({ \ |
| 4414 | float32x4_t __b = (b); \ |
| 4415 | __builtin_neon_vst1q_lane_v(__a, (int8x16_t)__b, __c, 40); }) |
| 4416 | #define vst1q_lane_p8(__a, b, __c) __extension__ ({ \ |
| 4417 | poly8x16_t __b = (b); \ |
| 4418 | __builtin_neon_vst1q_lane_v(__a, (int8x16_t)__b, __c, 36); }) |
| 4419 | #define vst1q_lane_p16(__a, b, __c) __extension__ ({ \ |
| 4420 | poly16x8_t __b = (b); \ |
| 4421 | __builtin_neon_vst1q_lane_v(__a, (int8x16_t)__b, __c, 37); }) |
| 4422 | #define vst1_lane_u8(__a, b, __c) __extension__ ({ \ |
| 4423 | uint8x8_t __b = (b); \ |
| 4424 | __builtin_neon_vst1_lane_v(__a, (int8x8_t)__b, __c, 16); }) |
| 4425 | #define vst1_lane_u16(__a, b, __c) __extension__ ({ \ |
| 4426 | uint16x4_t __b = (b); \ |
| 4427 | __builtin_neon_vst1_lane_v(__a, (int8x8_t)__b, __c, 17); }) |
| 4428 | #define vst1_lane_u32(__a, b, __c) __extension__ ({ \ |
| 4429 | uint32x2_t __b = (b); \ |
| 4430 | __builtin_neon_vst1_lane_v(__a, (int8x8_t)__b, __c, 18); }) |
| 4431 | #define vst1_lane_u64(__a, b, __c) __extension__ ({ \ |
| 4432 | uint64x1_t __b = (b); \ |
| 4433 | __builtin_neon_vst1_lane_v(__a, (int8x8_t)__b, __c, 19); }) |
| 4434 | #define vst1_lane_s8(__a, b, __c) __extension__ ({ \ |
| 4435 | int8x8_t __b = (b); \ |
| 4436 | __builtin_neon_vst1_lane_v(__a, __b, __c, 0); }) |
| 4437 | #define vst1_lane_s16(__a, b, __c) __extension__ ({ \ |
| 4438 | int16x4_t __b = (b); \ |
| 4439 | __builtin_neon_vst1_lane_v(__a, (int8x8_t)__b, __c, 1); }) |
| 4440 | #define vst1_lane_s32(__a, b, __c) __extension__ ({ \ |
| 4441 | int32x2_t __b = (b); \ |
| 4442 | __builtin_neon_vst1_lane_v(__a, (int8x8_t)__b, __c, 2); }) |
| 4443 | #define vst1_lane_s64(__a, b, __c) __extension__ ({ \ |
| 4444 | int64x1_t __b = (b); \ |
| 4445 | __builtin_neon_vst1_lane_v(__a, (int8x8_t)__b, __c, 3); }) |
| 4446 | #define vst1_lane_f16(__a, b, __c) __extension__ ({ \ |
| 4447 | float16x4_t __b = (b); \ |
| 4448 | __builtin_neon_vst1_lane_v(__a, (int8x8_t)__b, __c, 7); }) |
| 4449 | #define vst1_lane_f32(__a, b, __c) __extension__ ({ \ |
| 4450 | float32x2_t __b = (b); \ |
| 4451 | __builtin_neon_vst1_lane_v(__a, (int8x8_t)__b, __c, 8); }) |
| 4452 | #define vst1_lane_p8(__a, b, __c) __extension__ ({ \ |
| 4453 | poly8x8_t __b = (b); \ |
| 4454 | __builtin_neon_vst1_lane_v(__a, (int8x8_t)__b, __c, 4); }) |
| 4455 | #define vst1_lane_p16(__a, b, __c) __extension__ ({ \ |
| 4456 | poly16x4_t __b = (b); \ |
| 4457 | __builtin_neon_vst1_lane_v(__a, (int8x8_t)__b, __c, 5); }) |
| 4458 | |
| 4459 | #define vst2q_u8(__a, b) __extension__ ({ \ |
| 4460 | uint8x16x2_t __b = (b); \ |
| 4461 | __builtin_neon_vst2q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], 48); }) |
| 4462 | #define vst2q_u16(__a, b) __extension__ ({ \ |
| 4463 | uint16x8x2_t __b = (b); \ |
| 4464 | __builtin_neon_vst2q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], 49); }) |
| 4465 | #define vst2q_u32(__a, b) __extension__ ({ \ |
| 4466 | uint32x4x2_t __b = (b); \ |
| 4467 | __builtin_neon_vst2q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], 50); }) |
| 4468 | #define vst2q_s8(__a, b) __extension__ ({ \ |
| 4469 | int8x16x2_t __b = (b); \ |
| 4470 | __builtin_neon_vst2q_v(__a, __b.val[0], __b.val[1], 32); }) |
| 4471 | #define vst2q_s16(__a, b) __extension__ ({ \ |
| 4472 | int16x8x2_t __b = (b); \ |
| 4473 | __builtin_neon_vst2q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], 33); }) |
| 4474 | #define vst2q_s32(__a, b) __extension__ ({ \ |
| 4475 | int32x4x2_t __b = (b); \ |
| 4476 | __builtin_neon_vst2q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], 34); }) |
| 4477 | #define vst2q_f16(__a, b) __extension__ ({ \ |
| 4478 | float16x8x2_t __b = (b); \ |
| 4479 | __builtin_neon_vst2q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], 39); }) |
| 4480 | #define vst2q_f32(__a, b) __extension__ ({ \ |
| 4481 | float32x4x2_t __b = (b); \ |
| 4482 | __builtin_neon_vst2q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], 40); }) |
| 4483 | #define vst2q_p8(__a, b) __extension__ ({ \ |
| 4484 | poly8x16x2_t __b = (b); \ |
| 4485 | __builtin_neon_vst2q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], 36); }) |
| 4486 | #define vst2q_p16(__a, b) __extension__ ({ \ |
| 4487 | poly16x8x2_t __b = (b); \ |
| 4488 | __builtin_neon_vst2q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], 37); }) |
| 4489 | #define vst2_u8(__a, b) __extension__ ({ \ |
| 4490 | uint8x8x2_t __b = (b); \ |
| 4491 | __builtin_neon_vst2_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], 16); }) |
| 4492 | #define vst2_u16(__a, b) __extension__ ({ \ |
| 4493 | uint16x4x2_t __b = (b); \ |
| 4494 | __builtin_neon_vst2_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], 17); }) |
| 4495 | #define vst2_u32(__a, b) __extension__ ({ \ |
| 4496 | uint32x2x2_t __b = (b); \ |
| 4497 | __builtin_neon_vst2_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], 18); }) |
| 4498 | #define vst2_u64(__a, b) __extension__ ({ \ |
| 4499 | uint64x1x2_t __b = (b); \ |
| 4500 | __builtin_neon_vst2_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], 19); }) |
| 4501 | #define vst2_s8(__a, b) __extension__ ({ \ |
| 4502 | int8x8x2_t __b = (b); \ |
| 4503 | __builtin_neon_vst2_v(__a, __b.val[0], __b.val[1], 0); }) |
| 4504 | #define vst2_s16(__a, b) __extension__ ({ \ |
| 4505 | int16x4x2_t __b = (b); \ |
| 4506 | __builtin_neon_vst2_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], 1); }) |
| 4507 | #define vst2_s32(__a, b) __extension__ ({ \ |
| 4508 | int32x2x2_t __b = (b); \ |
| 4509 | __builtin_neon_vst2_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], 2); }) |
| 4510 | #define vst2_s64(__a, b) __extension__ ({ \ |
| 4511 | int64x1x2_t __b = (b); \ |
| 4512 | __builtin_neon_vst2_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], 3); }) |
| 4513 | #define vst2_f16(__a, b) __extension__ ({ \ |
| 4514 | float16x4x2_t __b = (b); \ |
| 4515 | __builtin_neon_vst2_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], 7); }) |
| 4516 | #define vst2_f32(__a, b) __extension__ ({ \ |
| 4517 | float32x2x2_t __b = (b); \ |
| 4518 | __builtin_neon_vst2_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], 8); }) |
| 4519 | #define vst2_p8(__a, b) __extension__ ({ \ |
| 4520 | poly8x8x2_t __b = (b); \ |
| 4521 | __builtin_neon_vst2_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], 4); }) |
| 4522 | #define vst2_p16(__a, b) __extension__ ({ \ |
| 4523 | poly16x4x2_t __b = (b); \ |
| 4524 | __builtin_neon_vst2_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], 5); }) |
| 4525 | |
| 4526 | #define vst2q_lane_u16(__a, b, __c) __extension__ ({ \ |
| 4527 | uint16x8x2_t __b = (b); \ |
| 4528 | __builtin_neon_vst2q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], __c, 49); }) |
| 4529 | #define vst2q_lane_u32(__a, b, __c) __extension__ ({ \ |
| 4530 | uint32x4x2_t __b = (b); \ |
| 4531 | __builtin_neon_vst2q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], __c, 50); }) |
| 4532 | #define vst2q_lane_s16(__a, b, __c) __extension__ ({ \ |
| 4533 | int16x8x2_t __b = (b); \ |
| 4534 | __builtin_neon_vst2q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], __c, 33); }) |
| 4535 | #define vst2q_lane_s32(__a, b, __c) __extension__ ({ \ |
| 4536 | int32x4x2_t __b = (b); \ |
| 4537 | __builtin_neon_vst2q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], __c, 34); }) |
| 4538 | #define vst2q_lane_f16(__a, b, __c) __extension__ ({ \ |
| 4539 | float16x8x2_t __b = (b); \ |
| 4540 | __builtin_neon_vst2q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], __c, 39); }) |
| 4541 | #define vst2q_lane_f32(__a, b, __c) __extension__ ({ \ |
| 4542 | float32x4x2_t __b = (b); \ |
| 4543 | __builtin_neon_vst2q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], __c, 40); }) |
| 4544 | #define vst2q_lane_p16(__a, b, __c) __extension__ ({ \ |
| 4545 | poly16x8x2_t __b = (b); \ |
| 4546 | __builtin_neon_vst2q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], __c, 37); }) |
| 4547 | #define vst2_lane_u8(__a, b, __c) __extension__ ({ \ |
| 4548 | uint8x8x2_t __b = (b); \ |
| 4549 | __builtin_neon_vst2_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], __c, 16); }) |
| 4550 | #define vst2_lane_u16(__a, b, __c) __extension__ ({ \ |
| 4551 | uint16x4x2_t __b = (b); \ |
| 4552 | __builtin_neon_vst2_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], __c, 17); }) |
| 4553 | #define vst2_lane_u32(__a, b, __c) __extension__ ({ \ |
| 4554 | uint32x2x2_t __b = (b); \ |
| 4555 | __builtin_neon_vst2_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], __c, 18); }) |
| 4556 | #define vst2_lane_s8(__a, b, __c) __extension__ ({ \ |
| 4557 | int8x8x2_t __b = (b); \ |
| 4558 | __builtin_neon_vst2_lane_v(__a, __b.val[0], __b.val[1], __c, 0); }) |
| 4559 | #define vst2_lane_s16(__a, b, __c) __extension__ ({ \ |
| 4560 | int16x4x2_t __b = (b); \ |
| 4561 | __builtin_neon_vst2_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], __c, 1); }) |
| 4562 | #define vst2_lane_s32(__a, b, __c) __extension__ ({ \ |
| 4563 | int32x2x2_t __b = (b); \ |
| 4564 | __builtin_neon_vst2_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], __c, 2); }) |
| 4565 | #define vst2_lane_f16(__a, b, __c) __extension__ ({ \ |
| 4566 | float16x4x2_t __b = (b); \ |
| 4567 | __builtin_neon_vst2_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], __c, 7); }) |
| 4568 | #define vst2_lane_f32(__a, b, __c) __extension__ ({ \ |
| 4569 | float32x2x2_t __b = (b); \ |
| 4570 | __builtin_neon_vst2_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], __c, 8); }) |
| 4571 | #define vst2_lane_p8(__a, b, __c) __extension__ ({ \ |
| 4572 | poly8x8x2_t __b = (b); \ |
| 4573 | __builtin_neon_vst2_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], __c, 4); }) |
| 4574 | #define vst2_lane_p16(__a, b, __c) __extension__ ({ \ |
| 4575 | poly16x4x2_t __b = (b); \ |
| 4576 | __builtin_neon_vst2_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], __c, 5); }) |
| 4577 | |
| 4578 | #define vst3q_u8(__a, b) __extension__ ({ \ |
| 4579 | uint8x16x3_t __b = (b); \ |
| 4580 | __builtin_neon_vst3q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], 48); }) |
| 4581 | #define vst3q_u16(__a, b) __extension__ ({ \ |
| 4582 | uint16x8x3_t __b = (b); \ |
| 4583 | __builtin_neon_vst3q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], 49); }) |
| 4584 | #define vst3q_u32(__a, b) __extension__ ({ \ |
| 4585 | uint32x4x3_t __b = (b); \ |
| 4586 | __builtin_neon_vst3q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], 50); }) |
| 4587 | #define vst3q_s8(__a, b) __extension__ ({ \ |
| 4588 | int8x16x3_t __b = (b); \ |
| 4589 | __builtin_neon_vst3q_v(__a, __b.val[0], __b.val[1], __b.val[2], 32); }) |
| 4590 | #define vst3q_s16(__a, b) __extension__ ({ \ |
| 4591 | int16x8x3_t __b = (b); \ |
| 4592 | __builtin_neon_vst3q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], 33); }) |
| 4593 | #define vst3q_s32(__a, b) __extension__ ({ \ |
| 4594 | int32x4x3_t __b = (b); \ |
| 4595 | __builtin_neon_vst3q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], 34); }) |
| 4596 | #define vst3q_f16(__a, b) __extension__ ({ \ |
| 4597 | float16x8x3_t __b = (b); \ |
| 4598 | __builtin_neon_vst3q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], 39); }) |
| 4599 | #define vst3q_f32(__a, b) __extension__ ({ \ |
| 4600 | float32x4x3_t __b = (b); \ |
| 4601 | __builtin_neon_vst3q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], 40); }) |
| 4602 | #define vst3q_p8(__a, b) __extension__ ({ \ |
| 4603 | poly8x16x3_t __b = (b); \ |
| 4604 | __builtin_neon_vst3q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], 36); }) |
| 4605 | #define vst3q_p16(__a, b) __extension__ ({ \ |
| 4606 | poly16x8x3_t __b = (b); \ |
| 4607 | __builtin_neon_vst3q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], 37); }) |
| 4608 | #define vst3_u8(__a, b) __extension__ ({ \ |
| 4609 | uint8x8x3_t __b = (b); \ |
| 4610 | __builtin_neon_vst3_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], 16); }) |
| 4611 | #define vst3_u16(__a, b) __extension__ ({ \ |
| 4612 | uint16x4x3_t __b = (b); \ |
| 4613 | __builtin_neon_vst3_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], 17); }) |
| 4614 | #define vst3_u32(__a, b) __extension__ ({ \ |
| 4615 | uint32x2x3_t __b = (b); \ |
| 4616 | __builtin_neon_vst3_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], 18); }) |
| 4617 | #define vst3_u64(__a, b) __extension__ ({ \ |
| 4618 | uint64x1x3_t __b = (b); \ |
| 4619 | __builtin_neon_vst3_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], 19); }) |
| 4620 | #define vst3_s8(__a, b) __extension__ ({ \ |
| 4621 | int8x8x3_t __b = (b); \ |
| 4622 | __builtin_neon_vst3_v(__a, __b.val[0], __b.val[1], __b.val[2], 0); }) |
| 4623 | #define vst3_s16(__a, b) __extension__ ({ \ |
| 4624 | int16x4x3_t __b = (b); \ |
| 4625 | __builtin_neon_vst3_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], 1); }) |
| 4626 | #define vst3_s32(__a, b) __extension__ ({ \ |
| 4627 | int32x2x3_t __b = (b); \ |
| 4628 | __builtin_neon_vst3_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], 2); }) |
| 4629 | #define vst3_s64(__a, b) __extension__ ({ \ |
| 4630 | int64x1x3_t __b = (b); \ |
| 4631 | __builtin_neon_vst3_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], 3); }) |
| 4632 | #define vst3_f16(__a, b) __extension__ ({ \ |
| 4633 | float16x4x3_t __b = (b); \ |
| 4634 | __builtin_neon_vst3_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], 7); }) |
| 4635 | #define vst3_f32(__a, b) __extension__ ({ \ |
| 4636 | float32x2x3_t __b = (b); \ |
| 4637 | __builtin_neon_vst3_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], 8); }) |
| 4638 | #define vst3_p8(__a, b) __extension__ ({ \ |
| 4639 | poly8x8x3_t __b = (b); \ |
| 4640 | __builtin_neon_vst3_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], 4); }) |
| 4641 | #define vst3_p16(__a, b) __extension__ ({ \ |
| 4642 | poly16x4x3_t __b = (b); \ |
| 4643 | __builtin_neon_vst3_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], 5); }) |
| 4644 | |
| 4645 | #define vst3q_lane_u16(__a, b, __c) __extension__ ({ \ |
| 4646 | uint16x8x3_t __b = (b); \ |
| 4647 | __builtin_neon_vst3q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], __c, 49); }) |
| 4648 | #define vst3q_lane_u32(__a, b, __c) __extension__ ({ \ |
| 4649 | uint32x4x3_t __b = (b); \ |
| 4650 | __builtin_neon_vst3q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], __c, 50); }) |
| 4651 | #define vst3q_lane_s16(__a, b, __c) __extension__ ({ \ |
| 4652 | int16x8x3_t __b = (b); \ |
| 4653 | __builtin_neon_vst3q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], __c, 33); }) |
| 4654 | #define vst3q_lane_s32(__a, b, __c) __extension__ ({ \ |
| 4655 | int32x4x3_t __b = (b); \ |
| 4656 | __builtin_neon_vst3q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], __c, 34); }) |
| 4657 | #define vst3q_lane_f16(__a, b, __c) __extension__ ({ \ |
| 4658 | float16x8x3_t __b = (b); \ |
| 4659 | __builtin_neon_vst3q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], __c, 39); }) |
| 4660 | #define vst3q_lane_f32(__a, b, __c) __extension__ ({ \ |
| 4661 | float32x4x3_t __b = (b); \ |
| 4662 | __builtin_neon_vst3q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], __c, 40); }) |
| 4663 | #define vst3q_lane_p16(__a, b, __c) __extension__ ({ \ |
| 4664 | poly16x8x3_t __b = (b); \ |
| 4665 | __builtin_neon_vst3q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], __c, 37); }) |
| 4666 | #define vst3_lane_u8(__a, b, __c) __extension__ ({ \ |
| 4667 | uint8x8x3_t __b = (b); \ |
| 4668 | __builtin_neon_vst3_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], __c, 16); }) |
| 4669 | #define vst3_lane_u16(__a, b, __c) __extension__ ({ \ |
| 4670 | uint16x4x3_t __b = (b); \ |
| 4671 | __builtin_neon_vst3_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], __c, 17); }) |
| 4672 | #define vst3_lane_u32(__a, b, __c) __extension__ ({ \ |
| 4673 | uint32x2x3_t __b = (b); \ |
| 4674 | __builtin_neon_vst3_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], __c, 18); }) |
| 4675 | #define vst3_lane_s8(__a, b, __c) __extension__ ({ \ |
| 4676 | int8x8x3_t __b = (b); \ |
| 4677 | __builtin_neon_vst3_lane_v(__a, __b.val[0], __b.val[1], __b.val[2], __c, 0); }) |
| 4678 | #define vst3_lane_s16(__a, b, __c) __extension__ ({ \ |
| 4679 | int16x4x3_t __b = (b); \ |
| 4680 | __builtin_neon_vst3_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], __c, 1); }) |
| 4681 | #define vst3_lane_s32(__a, b, __c) __extension__ ({ \ |
| 4682 | int32x2x3_t __b = (b); \ |
| 4683 | __builtin_neon_vst3_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], __c, 2); }) |
| 4684 | #define vst3_lane_f16(__a, b, __c) __extension__ ({ \ |
| 4685 | float16x4x3_t __b = (b); \ |
| 4686 | __builtin_neon_vst3_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], __c, 7); }) |
| 4687 | #define vst3_lane_f32(__a, b, __c) __extension__ ({ \ |
| 4688 | float32x2x3_t __b = (b); \ |
| 4689 | __builtin_neon_vst3_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], __c, 8); }) |
| 4690 | #define vst3_lane_p8(__a, b, __c) __extension__ ({ \ |
| 4691 | poly8x8x3_t __b = (b); \ |
| 4692 | __builtin_neon_vst3_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], __c, 4); }) |
| 4693 | #define vst3_lane_p16(__a, b, __c) __extension__ ({ \ |
| 4694 | poly16x4x3_t __b = (b); \ |
| 4695 | __builtin_neon_vst3_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], __c, 5); }) |
| 4696 | |
| 4697 | #define vst4q_u8(__a, b) __extension__ ({ \ |
| 4698 | uint8x16x4_t __b = (b); \ |
| 4699 | __builtin_neon_vst4q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], 48); }) |
| 4700 | #define vst4q_u16(__a, b) __extension__ ({ \ |
| 4701 | uint16x8x4_t __b = (b); \ |
| 4702 | __builtin_neon_vst4q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], 49); }) |
| 4703 | #define vst4q_u32(__a, b) __extension__ ({ \ |
| 4704 | uint32x4x4_t __b = (b); \ |
| 4705 | __builtin_neon_vst4q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], 50); }) |
| 4706 | #define vst4q_s8(__a, b) __extension__ ({ \ |
| 4707 | int8x16x4_t __b = (b); \ |
| 4708 | __builtin_neon_vst4q_v(__a, __b.val[0], __b.val[1], __b.val[2], __b.val[3], 32); }) |
| 4709 | #define vst4q_s16(__a, b) __extension__ ({ \ |
| 4710 | int16x8x4_t __b = (b); \ |
| 4711 | __builtin_neon_vst4q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], 33); }) |
| 4712 | #define vst4q_s32(__a, b) __extension__ ({ \ |
| 4713 | int32x4x4_t __b = (b); \ |
| 4714 | __builtin_neon_vst4q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], 34); }) |
| 4715 | #define vst4q_f16(__a, b) __extension__ ({ \ |
| 4716 | float16x8x4_t __b = (b); \ |
| 4717 | __builtin_neon_vst4q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], 39); }) |
| 4718 | #define vst4q_f32(__a, b) __extension__ ({ \ |
| 4719 | float32x4x4_t __b = (b); \ |
| 4720 | __builtin_neon_vst4q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], 40); }) |
| 4721 | #define vst4q_p8(__a, b) __extension__ ({ \ |
| 4722 | poly8x16x4_t __b = (b); \ |
| 4723 | __builtin_neon_vst4q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], 36); }) |
| 4724 | #define vst4q_p16(__a, b) __extension__ ({ \ |
| 4725 | poly16x8x4_t __b = (b); \ |
| 4726 | __builtin_neon_vst4q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], 37); }) |
| 4727 | #define vst4_u8(__a, b) __extension__ ({ \ |
| 4728 | uint8x8x4_t __b = (b); \ |
| 4729 | __builtin_neon_vst4_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], 16); }) |
| 4730 | #define vst4_u16(__a, b) __extension__ ({ \ |
| 4731 | uint16x4x4_t __b = (b); \ |
| 4732 | __builtin_neon_vst4_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], 17); }) |
| 4733 | #define vst4_u32(__a, b) __extension__ ({ \ |
| 4734 | uint32x2x4_t __b = (b); \ |
| 4735 | __builtin_neon_vst4_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], 18); }) |
| 4736 | #define vst4_u64(__a, b) __extension__ ({ \ |
| 4737 | uint64x1x4_t __b = (b); \ |
| 4738 | __builtin_neon_vst4_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], 19); }) |
| 4739 | #define vst4_s8(__a, b) __extension__ ({ \ |
| 4740 | int8x8x4_t __b = (b); \ |
| 4741 | __builtin_neon_vst4_v(__a, __b.val[0], __b.val[1], __b.val[2], __b.val[3], 0); }) |
| 4742 | #define vst4_s16(__a, b) __extension__ ({ \ |
| 4743 | int16x4x4_t __b = (b); \ |
| 4744 | __builtin_neon_vst4_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], 1); }) |
| 4745 | #define vst4_s32(__a, b) __extension__ ({ \ |
| 4746 | int32x2x4_t __b = (b); \ |
| 4747 | __builtin_neon_vst4_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], 2); }) |
| 4748 | #define vst4_s64(__a, b) __extension__ ({ \ |
| 4749 | int64x1x4_t __b = (b); \ |
| 4750 | __builtin_neon_vst4_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], 3); }) |
| 4751 | #define vst4_f16(__a, b) __extension__ ({ \ |
| 4752 | float16x4x4_t __b = (b); \ |
| 4753 | __builtin_neon_vst4_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], 7); }) |
| 4754 | #define vst4_f32(__a, b) __extension__ ({ \ |
| 4755 | float32x2x4_t __b = (b); \ |
| 4756 | __builtin_neon_vst4_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], 8); }) |
| 4757 | #define vst4_p8(__a, b) __extension__ ({ \ |
| 4758 | poly8x8x4_t __b = (b); \ |
| 4759 | __builtin_neon_vst4_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], 4); }) |
| 4760 | #define vst4_p16(__a, b) __extension__ ({ \ |
| 4761 | poly16x4x4_t __b = (b); \ |
| 4762 | __builtin_neon_vst4_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], 5); }) |
| 4763 | |
| 4764 | #define vst4q_lane_u16(__a, b, __c) __extension__ ({ \ |
| 4765 | uint16x8x4_t __b = (b); \ |
| 4766 | __builtin_neon_vst4q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], __c, 49); }) |
| 4767 | #define vst4q_lane_u32(__a, b, __c) __extension__ ({ \ |
| 4768 | uint32x4x4_t __b = (b); \ |
| 4769 | __builtin_neon_vst4q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], __c, 50); }) |
| 4770 | #define vst4q_lane_s16(__a, b, __c) __extension__ ({ \ |
| 4771 | int16x8x4_t __b = (b); \ |
| 4772 | __builtin_neon_vst4q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], __c, 33); }) |
| 4773 | #define vst4q_lane_s32(__a, b, __c) __extension__ ({ \ |
| 4774 | int32x4x4_t __b = (b); \ |
| 4775 | __builtin_neon_vst4q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], __c, 34); }) |
| 4776 | #define vst4q_lane_f16(__a, b, __c) __extension__ ({ \ |
| 4777 | float16x8x4_t __b = (b); \ |
| 4778 | __builtin_neon_vst4q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], __c, 39); }) |
| 4779 | #define vst4q_lane_f32(__a, b, __c) __extension__ ({ \ |
| 4780 | float32x4x4_t __b = (b); \ |
| 4781 | __builtin_neon_vst4q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], __c, 40); }) |
| 4782 | #define vst4q_lane_p16(__a, b, __c) __extension__ ({ \ |
| 4783 | poly16x8x4_t __b = (b); \ |
| 4784 | __builtin_neon_vst4q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], __c, 37); }) |
| 4785 | #define vst4_lane_u8(__a, b, __c) __extension__ ({ \ |
| 4786 | uint8x8x4_t __b = (b); \ |
| 4787 | __builtin_neon_vst4_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], __c, 16); }) |
| 4788 | #define vst4_lane_u16(__a, b, __c) __extension__ ({ \ |
| 4789 | uint16x4x4_t __b = (b); \ |
| 4790 | __builtin_neon_vst4_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], __c, 17); }) |
| 4791 | #define vst4_lane_u32(__a, b, __c) __extension__ ({ \ |
| 4792 | uint32x2x4_t __b = (b); \ |
| 4793 | __builtin_neon_vst4_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], __c, 18); }) |
| 4794 | #define vst4_lane_s8(__a, b, __c) __extension__ ({ \ |
| 4795 | int8x8x4_t __b = (b); \ |
| 4796 | __builtin_neon_vst4_lane_v(__a, __b.val[0], __b.val[1], __b.val[2], __b.val[3], __c, 0); }) |
| 4797 | #define vst4_lane_s16(__a, b, __c) __extension__ ({ \ |
| 4798 | int16x4x4_t __b = (b); \ |
| 4799 | __builtin_neon_vst4_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], __c, 1); }) |
| 4800 | #define vst4_lane_s32(__a, b, __c) __extension__ ({ \ |
| 4801 | int32x2x4_t __b = (b); \ |
| 4802 | __builtin_neon_vst4_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], __c, 2); }) |
| 4803 | #define vst4_lane_f16(__a, b, __c) __extension__ ({ \ |
| 4804 | float16x4x4_t __b = (b); \ |
| 4805 | __builtin_neon_vst4_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], __c, 7); }) |
| 4806 | #define vst4_lane_f32(__a, b, __c) __extension__ ({ \ |
| 4807 | float32x2x4_t __b = (b); \ |
| 4808 | __builtin_neon_vst4_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], __c, 8); }) |
| 4809 | #define vst4_lane_p8(__a, b, __c) __extension__ ({ \ |
| 4810 | poly8x8x4_t __b = (b); \ |
| 4811 | __builtin_neon_vst4_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], __c, 4); }) |
| 4812 | #define vst4_lane_p16(__a, b, __c) __extension__ ({ \ |
| 4813 | poly16x4x4_t __b = (b); \ |
| 4814 | __builtin_neon_vst4_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], __c, 5); }) |
| 4815 | |
| 4816 | __ai int8x8_t vsub_s8(int8x8_t __a, int8x8_t __b) { |
| 4817 | return __a - __b; } |
| 4818 | __ai int16x4_t vsub_s16(int16x4_t __a, int16x4_t __b) { |
| 4819 | return __a - __b; } |
| 4820 | __ai int32x2_t vsub_s32(int32x2_t __a, int32x2_t __b) { |
| 4821 | return __a - __b; } |
| 4822 | __ai int64x1_t vsub_s64(int64x1_t __a, int64x1_t __b) { |
| 4823 | return __a - __b; } |
| 4824 | __ai float32x2_t vsub_f32(float32x2_t __a, float32x2_t __b) { |
| 4825 | return __a - __b; } |
| 4826 | __ai uint8x8_t vsub_u8(uint8x8_t __a, uint8x8_t __b) { |
| 4827 | return __a - __b; } |
| 4828 | __ai uint16x4_t vsub_u16(uint16x4_t __a, uint16x4_t __b) { |
| 4829 | return __a - __b; } |
| 4830 | __ai uint32x2_t vsub_u32(uint32x2_t __a, uint32x2_t __b) { |
| 4831 | return __a - __b; } |
| 4832 | __ai uint64x1_t vsub_u64(uint64x1_t __a, uint64x1_t __b) { |
| 4833 | return __a - __b; } |
| 4834 | __ai int8x16_t vsubq_s8(int8x16_t __a, int8x16_t __b) { |
| 4835 | return __a - __b; } |
| 4836 | __ai int16x8_t vsubq_s16(int16x8_t __a, int16x8_t __b) { |
| 4837 | return __a - __b; } |
| 4838 | __ai int32x4_t vsubq_s32(int32x4_t __a, int32x4_t __b) { |
| 4839 | return __a - __b; } |
| 4840 | __ai int64x2_t vsubq_s64(int64x2_t __a, int64x2_t __b) { |
| 4841 | return __a - __b; } |
| 4842 | __ai float32x4_t vsubq_f32(float32x4_t __a, float32x4_t __b) { |
| 4843 | return __a - __b; } |
| 4844 | __ai uint8x16_t vsubq_u8(uint8x16_t __a, uint8x16_t __b) { |
| 4845 | return __a - __b; } |
| 4846 | __ai uint16x8_t vsubq_u16(uint16x8_t __a, uint16x8_t __b) { |
| 4847 | return __a - __b; } |
| 4848 | __ai uint32x4_t vsubq_u32(uint32x4_t __a, uint32x4_t __b) { |
| 4849 | return __a - __b; } |
| 4850 | __ai uint64x2_t vsubq_u64(uint64x2_t __a, uint64x2_t __b) { |
| 4851 | return __a - __b; } |
| 4852 | |
| 4853 | __ai int8x8_t vsubhn_s16(int16x8_t __a, int16x8_t __b) { |
| 4854 | return (int8x8_t)__builtin_neon_vsubhn_v((int8x16_t)__a, (int8x16_t)__b, 0); } |
| 4855 | __ai int16x4_t vsubhn_s32(int32x4_t __a, int32x4_t __b) { |
| 4856 | return (int16x4_t)__builtin_neon_vsubhn_v((int8x16_t)__a, (int8x16_t)__b, 1); } |
| 4857 | __ai int32x2_t vsubhn_s64(int64x2_t __a, int64x2_t __b) { |
| 4858 | return (int32x2_t)__builtin_neon_vsubhn_v((int8x16_t)__a, (int8x16_t)__b, 2); } |
| 4859 | __ai uint8x8_t vsubhn_u16(uint16x8_t __a, uint16x8_t __b) { |
| 4860 | return (uint8x8_t)__builtin_neon_vsubhn_v((int8x16_t)__a, (int8x16_t)__b, 16); } |
| 4861 | __ai uint16x4_t vsubhn_u32(uint32x4_t __a, uint32x4_t __b) { |
| 4862 | return (uint16x4_t)__builtin_neon_vsubhn_v((int8x16_t)__a, (int8x16_t)__b, 17); } |
| 4863 | __ai uint32x2_t vsubhn_u64(uint64x2_t __a, uint64x2_t __b) { |
| 4864 | return (uint32x2_t)__builtin_neon_vsubhn_v((int8x16_t)__a, (int8x16_t)__b, 18); } |
| 4865 | |
| 4866 | __ai int16x8_t vsubl_s8(int8x8_t __a, int8x8_t __b) { |
| 4867 | return vmovl_s8(__a) - vmovl_s8(__b); } |
| 4868 | __ai int32x4_t vsubl_s16(int16x4_t __a, int16x4_t __b) { |
| 4869 | return vmovl_s16(__a) - vmovl_s16(__b); } |
| 4870 | __ai int64x2_t vsubl_s32(int32x2_t __a, int32x2_t __b) { |
| 4871 | return vmovl_s32(__a) - vmovl_s32(__b); } |
| 4872 | __ai uint16x8_t vsubl_u8(uint8x8_t __a, uint8x8_t __b) { |
| 4873 | return vmovl_u8(__a) - vmovl_u8(__b); } |
| 4874 | __ai uint32x4_t vsubl_u16(uint16x4_t __a, uint16x4_t __b) { |
| 4875 | return vmovl_u16(__a) - vmovl_u16(__b); } |
| 4876 | __ai uint64x2_t vsubl_u32(uint32x2_t __a, uint32x2_t __b) { |
| 4877 | return vmovl_u32(__a) - vmovl_u32(__b); } |
| 4878 | |
| 4879 | __ai int16x8_t vsubw_s8(int16x8_t __a, int8x8_t __b) { |
| 4880 | return __a - vmovl_s8(__b); } |
| 4881 | __ai int32x4_t vsubw_s16(int32x4_t __a, int16x4_t __b) { |
| 4882 | return __a - vmovl_s16(__b); } |
| 4883 | __ai int64x2_t vsubw_s32(int64x2_t __a, int32x2_t __b) { |
| 4884 | return __a - vmovl_s32(__b); } |
| 4885 | __ai uint16x8_t vsubw_u8(uint16x8_t __a, uint8x8_t __b) { |
| 4886 | return __a - vmovl_u8(__b); } |
| 4887 | __ai uint32x4_t vsubw_u16(uint32x4_t __a, uint16x4_t __b) { |
| 4888 | return __a - vmovl_u16(__b); } |
| 4889 | __ai uint64x2_t vsubw_u32(uint64x2_t __a, uint32x2_t __b) { |
| 4890 | return __a - vmovl_u32(__b); } |
| 4891 | |
| 4892 | __ai uint8x8_t vtbl1_u8(uint8x8_t __a, uint8x8_t __b) { |
| 4893 | return (uint8x8_t)__builtin_neon_vtbl1_v((int8x8_t)__a, (int8x8_t)__b, 16); } |
| 4894 | __ai int8x8_t vtbl1_s8(int8x8_t __a, int8x8_t __b) { |
| 4895 | return (int8x8_t)__builtin_neon_vtbl1_v(__a, __b, 0); } |
| 4896 | __ai poly8x8_t vtbl1_p8(poly8x8_t __a, uint8x8_t __b) { |
| 4897 | return (poly8x8_t)__builtin_neon_vtbl1_v((int8x8_t)__a, (int8x8_t)__b, 4); } |
| 4898 | |
| 4899 | __ai uint8x8_t vtbl2_u8(uint8x8x2_t __a, uint8x8_t __b) { |
| 4900 | return (uint8x8_t)__builtin_neon_vtbl2_v((int8x8_t)__a.val[0], (int8x8_t)__a.val[1], (int8x8_t)__b, 16); } |
| 4901 | __ai int8x8_t vtbl2_s8(int8x8x2_t __a, int8x8_t __b) { |
| 4902 | return (int8x8_t)__builtin_neon_vtbl2_v(__a.val[0], __a.val[1], __b, 0); } |
| 4903 | __ai poly8x8_t vtbl2_p8(poly8x8x2_t __a, uint8x8_t __b) { |
| 4904 | return (poly8x8_t)__builtin_neon_vtbl2_v((int8x8_t)__a.val[0], (int8x8_t)__a.val[1], (int8x8_t)__b, 4); } |
| 4905 | |
| 4906 | __ai uint8x8_t vtbl3_u8(uint8x8x3_t __a, uint8x8_t __b) { |
| 4907 | return (uint8x8_t)__builtin_neon_vtbl3_v((int8x8_t)__a.val[0], (int8x8_t)__a.val[1], (int8x8_t)__a.val[2], (int8x8_t)__b, 16); } |
| 4908 | __ai int8x8_t vtbl3_s8(int8x8x3_t __a, int8x8_t __b) { |
| 4909 | return (int8x8_t)__builtin_neon_vtbl3_v(__a.val[0], __a.val[1], __a.val[2], __b, 0); } |
| 4910 | __ai poly8x8_t vtbl3_p8(poly8x8x3_t __a, uint8x8_t __b) { |
| 4911 | return (poly8x8_t)__builtin_neon_vtbl3_v((int8x8_t)__a.val[0], (int8x8_t)__a.val[1], (int8x8_t)__a.val[2], (int8x8_t)__b, 4); } |
| 4912 | |
| 4913 | __ai uint8x8_t vtbl4_u8(uint8x8x4_t __a, uint8x8_t __b) { |
| 4914 | return (uint8x8_t)__builtin_neon_vtbl4_v((int8x8_t)__a.val[0], (int8x8_t)__a.val[1], (int8x8_t)__a.val[2], (int8x8_t)__a.val[3], (int8x8_t)__b, 16); } |
| 4915 | __ai int8x8_t vtbl4_s8(int8x8x4_t __a, int8x8_t __b) { |
| 4916 | return (int8x8_t)__builtin_neon_vtbl4_v(__a.val[0], __a.val[1], __a.val[2], __a.val[3], __b, 0); } |
| 4917 | __ai poly8x8_t vtbl4_p8(poly8x8x4_t __a, uint8x8_t __b) { |
| 4918 | return (poly8x8_t)__builtin_neon_vtbl4_v((int8x8_t)__a.val[0], (int8x8_t)__a.val[1], (int8x8_t)__a.val[2], (int8x8_t)__a.val[3], (int8x8_t)__b, 4); } |
| 4919 | |
| 4920 | __ai uint8x8_t vtbx1_u8(uint8x8_t __a, uint8x8_t __b, uint8x8_t __c) { |
| 4921 | return (uint8x8_t)__builtin_neon_vtbx1_v((int8x8_t)__a, (int8x8_t)__b, (int8x8_t)__c, 16); } |
| 4922 | __ai int8x8_t vtbx1_s8(int8x8_t __a, int8x8_t __b, int8x8_t __c) { |
| 4923 | return (int8x8_t)__builtin_neon_vtbx1_v(__a, __b, __c, 0); } |
| 4924 | __ai poly8x8_t vtbx1_p8(poly8x8_t __a, poly8x8_t __b, uint8x8_t __c) { |
| 4925 | return (poly8x8_t)__builtin_neon_vtbx1_v((int8x8_t)__a, (int8x8_t)__b, (int8x8_t)__c, 4); } |
| 4926 | |
| 4927 | __ai uint8x8_t vtbx2_u8(uint8x8_t __a, uint8x8x2_t __b, uint8x8_t __c) { |
| 4928 | return (uint8x8_t)__builtin_neon_vtbx2_v((int8x8_t)__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__c, 16); } |
| 4929 | __ai int8x8_t vtbx2_s8(int8x8_t __a, int8x8x2_t __b, int8x8_t __c) { |
| 4930 | return (int8x8_t)__builtin_neon_vtbx2_v(__a, __b.val[0], __b.val[1], __c, 0); } |
| 4931 | __ai poly8x8_t vtbx2_p8(poly8x8_t __a, poly8x8x2_t __b, uint8x8_t __c) { |
| 4932 | return (poly8x8_t)__builtin_neon_vtbx2_v((int8x8_t)__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__c, 4); } |
| 4933 | |
| 4934 | __ai uint8x8_t vtbx3_u8(uint8x8_t __a, uint8x8x3_t __b, uint8x8_t __c) { |
| 4935 | return (uint8x8_t)__builtin_neon_vtbx3_v((int8x8_t)__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__c, 16); } |
| 4936 | __ai int8x8_t vtbx3_s8(int8x8_t __a, int8x8x3_t __b, int8x8_t __c) { |
| 4937 | return (int8x8_t)__builtin_neon_vtbx3_v(__a, __b.val[0], __b.val[1], __b.val[2], __c, 0); } |
| 4938 | __ai poly8x8_t vtbx3_p8(poly8x8_t __a, poly8x8x3_t __b, uint8x8_t __c) { |
| 4939 | return (poly8x8_t)__builtin_neon_vtbx3_v((int8x8_t)__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__c, 4); } |
| 4940 | |
| 4941 | __ai uint8x8_t vtbx4_u8(uint8x8_t __a, uint8x8x4_t __b, uint8x8_t __c) { |
| 4942 | return (uint8x8_t)__builtin_neon_vtbx4_v((int8x8_t)__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], (int8x8_t)__c, 16); } |
| 4943 | __ai int8x8_t vtbx4_s8(int8x8_t __a, int8x8x4_t __b, int8x8_t __c) { |
| 4944 | return (int8x8_t)__builtin_neon_vtbx4_v(__a, __b.val[0], __b.val[1], __b.val[2], __b.val[3], __c, 0); } |
| 4945 | __ai poly8x8_t vtbx4_p8(poly8x8_t __a, poly8x8x4_t __b, uint8x8_t __c) { |
| 4946 | return (poly8x8_t)__builtin_neon_vtbx4_v((int8x8_t)__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], (int8x8_t)__c, 4); } |
| 4947 | |
| 4948 | __ai int8x8x2_t vtrn_s8(int8x8_t __a, int8x8_t __b) { |
| 4949 | int8x8x2_t r; __builtin_neon_vtrn_v(&r, __a, __b, 0); return r; } |
| 4950 | __ai int16x4x2_t vtrn_s16(int16x4_t __a, int16x4_t __b) { |
| 4951 | int16x4x2_t r; __builtin_neon_vtrn_v(&r, (int8x8_t)__a, (int8x8_t)__b, 1); return r; } |
| 4952 | __ai int32x2x2_t vtrn_s32(int32x2_t __a, int32x2_t __b) { |
| 4953 | int32x2x2_t r; __builtin_neon_vtrn_v(&r, (int8x8_t)__a, (int8x8_t)__b, 2); return r; } |
| 4954 | __ai uint8x8x2_t vtrn_u8(uint8x8_t __a, uint8x8_t __b) { |
| 4955 | uint8x8x2_t r; __builtin_neon_vtrn_v(&r, (int8x8_t)__a, (int8x8_t)__b, 16); return r; } |
| 4956 | __ai uint16x4x2_t vtrn_u16(uint16x4_t __a, uint16x4_t __b) { |
| 4957 | uint16x4x2_t r; __builtin_neon_vtrn_v(&r, (int8x8_t)__a, (int8x8_t)__b, 17); return r; } |
| 4958 | __ai uint32x2x2_t vtrn_u32(uint32x2_t __a, uint32x2_t __b) { |
| 4959 | uint32x2x2_t r; __builtin_neon_vtrn_v(&r, (int8x8_t)__a, (int8x8_t)__b, 18); return r; } |
| 4960 | __ai float32x2x2_t vtrn_f32(float32x2_t __a, float32x2_t __b) { |
| 4961 | float32x2x2_t r; __builtin_neon_vtrn_v(&r, (int8x8_t)__a, (int8x8_t)__b, 8); return r; } |
| 4962 | __ai poly8x8x2_t vtrn_p8(poly8x8_t __a, poly8x8_t __b) { |
| 4963 | poly8x8x2_t r; __builtin_neon_vtrn_v(&r, (int8x8_t)__a, (int8x8_t)__b, 4); return r; } |
| 4964 | __ai poly16x4x2_t vtrn_p16(poly16x4_t __a, poly16x4_t __b) { |
| 4965 | poly16x4x2_t r; __builtin_neon_vtrn_v(&r, (int8x8_t)__a, (int8x8_t)__b, 5); return r; } |
| 4966 | __ai int8x16x2_t vtrnq_s8(int8x16_t __a, int8x16_t __b) { |
| 4967 | int8x16x2_t r; __builtin_neon_vtrnq_v(&r, __a, __b, 32); return r; } |
| 4968 | __ai int16x8x2_t vtrnq_s16(int16x8_t __a, int16x8_t __b) { |
| 4969 | int16x8x2_t r; __builtin_neon_vtrnq_v(&r, (int8x16_t)__a, (int8x16_t)__b, 33); return r; } |
| 4970 | __ai int32x4x2_t vtrnq_s32(int32x4_t __a, int32x4_t __b) { |
| 4971 | int32x4x2_t r; __builtin_neon_vtrnq_v(&r, (int8x16_t)__a, (int8x16_t)__b, 34); return r; } |
| 4972 | __ai uint8x16x2_t vtrnq_u8(uint8x16_t __a, uint8x16_t __b) { |
| 4973 | uint8x16x2_t r; __builtin_neon_vtrnq_v(&r, (int8x16_t)__a, (int8x16_t)__b, 48); return r; } |
| 4974 | __ai uint16x8x2_t vtrnq_u16(uint16x8_t __a, uint16x8_t __b) { |
| 4975 | uint16x8x2_t r; __builtin_neon_vtrnq_v(&r, (int8x16_t)__a, (int8x16_t)__b, 49); return r; } |
| 4976 | __ai uint32x4x2_t vtrnq_u32(uint32x4_t __a, uint32x4_t __b) { |
| 4977 | uint32x4x2_t r; __builtin_neon_vtrnq_v(&r, (int8x16_t)__a, (int8x16_t)__b, 50); return r; } |
| 4978 | __ai float32x4x2_t vtrnq_f32(float32x4_t __a, float32x4_t __b) { |
| 4979 | float32x4x2_t r; __builtin_neon_vtrnq_v(&r, (int8x16_t)__a, (int8x16_t)__b, 40); return r; } |
| 4980 | __ai poly8x16x2_t vtrnq_p8(poly8x16_t __a, poly8x16_t __b) { |
| 4981 | poly8x16x2_t r; __builtin_neon_vtrnq_v(&r, (int8x16_t)__a, (int8x16_t)__b, 36); return r; } |
| 4982 | __ai poly16x8x2_t vtrnq_p16(poly16x8_t __a, poly16x8_t __b) { |
| 4983 | poly16x8x2_t r; __builtin_neon_vtrnq_v(&r, (int8x16_t)__a, (int8x16_t)__b, 37); return r; } |
| 4984 | |
| 4985 | __ai uint8x8_t vtst_s8(int8x8_t __a, int8x8_t __b) { |
| 4986 | return (uint8x8_t)__builtin_neon_vtst_v(__a, __b, 16); } |
| 4987 | __ai uint16x4_t vtst_s16(int16x4_t __a, int16x4_t __b) { |
| 4988 | return (uint16x4_t)__builtin_neon_vtst_v((int8x8_t)__a, (int8x8_t)__b, 17); } |
| 4989 | __ai uint32x2_t vtst_s32(int32x2_t __a, int32x2_t __b) { |
| 4990 | return (uint32x2_t)__builtin_neon_vtst_v((int8x8_t)__a, (int8x8_t)__b, 18); } |
| 4991 | __ai uint8x8_t vtst_u8(uint8x8_t __a, uint8x8_t __b) { |
| 4992 | return (uint8x8_t)__builtin_neon_vtst_v((int8x8_t)__a, (int8x8_t)__b, 16); } |
| 4993 | __ai uint16x4_t vtst_u16(uint16x4_t __a, uint16x4_t __b) { |
| 4994 | return (uint16x4_t)__builtin_neon_vtst_v((int8x8_t)__a, (int8x8_t)__b, 17); } |
| 4995 | __ai uint32x2_t vtst_u32(uint32x2_t __a, uint32x2_t __b) { |
| 4996 | return (uint32x2_t)__builtin_neon_vtst_v((int8x8_t)__a, (int8x8_t)__b, 18); } |
| 4997 | __ai uint8x8_t vtst_p8(poly8x8_t __a, poly8x8_t __b) { |
| 4998 | return (uint8x8_t)__builtin_neon_vtst_v((int8x8_t)__a, (int8x8_t)__b, 16); } |
| 4999 | __ai uint16x4_t vtst_p16(poly16x4_t __a, poly16x4_t __b) { |
| 5000 | return (uint16x4_t)__builtin_neon_vtst_v((int8x8_t)__a, (int8x8_t)__b, 17); } |
| 5001 | __ai uint8x16_t vtstq_s8(int8x16_t __a, int8x16_t __b) { |
| 5002 | return (uint8x16_t)__builtin_neon_vtstq_v(__a, __b, 48); } |
| 5003 | __ai uint16x8_t vtstq_s16(int16x8_t __a, int16x8_t __b) { |
| 5004 | return (uint16x8_t)__builtin_neon_vtstq_v((int8x16_t)__a, (int8x16_t)__b, 49); } |
| 5005 | __ai uint32x4_t vtstq_s32(int32x4_t __a, int32x4_t __b) { |
| 5006 | return (uint32x4_t)__builtin_neon_vtstq_v((int8x16_t)__a, (int8x16_t)__b, 50); } |
| 5007 | __ai uint8x16_t vtstq_u8(uint8x16_t __a, uint8x16_t __b) { |
| 5008 | return (uint8x16_t)__builtin_neon_vtstq_v((int8x16_t)__a, (int8x16_t)__b, 48); } |
| 5009 | __ai uint16x8_t vtstq_u16(uint16x8_t __a, uint16x8_t __b) { |
| 5010 | return (uint16x8_t)__builtin_neon_vtstq_v((int8x16_t)__a, (int8x16_t)__b, 49); } |
| 5011 | __ai uint32x4_t vtstq_u32(uint32x4_t __a, uint32x4_t __b) { |
| 5012 | return (uint32x4_t)__builtin_neon_vtstq_v((int8x16_t)__a, (int8x16_t)__b, 50); } |
| 5013 | __ai uint8x16_t vtstq_p8(poly8x16_t __a, poly8x16_t __b) { |
| 5014 | return (uint8x16_t)__builtin_neon_vtstq_v((int8x16_t)__a, (int8x16_t)__b, 48); } |
| 5015 | __ai uint16x8_t vtstq_p16(poly16x8_t __a, poly16x8_t __b) { |
| 5016 | return (uint16x8_t)__builtin_neon_vtstq_v((int8x16_t)__a, (int8x16_t)__b, 49); } |
| 5017 | |
| 5018 | __ai int8x8x2_t vuzp_s8(int8x8_t __a, int8x8_t __b) { |
| 5019 | int8x8x2_t r; __builtin_neon_vuzp_v(&r, __a, __b, 0); return r; } |
| 5020 | __ai int16x4x2_t vuzp_s16(int16x4_t __a, int16x4_t __b) { |
| 5021 | int16x4x2_t r; __builtin_neon_vuzp_v(&r, (int8x8_t)__a, (int8x8_t)__b, 1); return r; } |
| 5022 | __ai int32x2x2_t vuzp_s32(int32x2_t __a, int32x2_t __b) { |
| 5023 | int32x2x2_t r; __builtin_neon_vuzp_v(&r, (int8x8_t)__a, (int8x8_t)__b, 2); return r; } |
| 5024 | __ai uint8x8x2_t vuzp_u8(uint8x8_t __a, uint8x8_t __b) { |
| 5025 | uint8x8x2_t r; __builtin_neon_vuzp_v(&r, (int8x8_t)__a, (int8x8_t)__b, 16); return r; } |
| 5026 | __ai uint16x4x2_t vuzp_u16(uint16x4_t __a, uint16x4_t __b) { |
| 5027 | uint16x4x2_t r; __builtin_neon_vuzp_v(&r, (int8x8_t)__a, (int8x8_t)__b, 17); return r; } |
| 5028 | __ai uint32x2x2_t vuzp_u32(uint32x2_t __a, uint32x2_t __b) { |
| 5029 | uint32x2x2_t r; __builtin_neon_vuzp_v(&r, (int8x8_t)__a, (int8x8_t)__b, 18); return r; } |
| 5030 | __ai float32x2x2_t vuzp_f32(float32x2_t __a, float32x2_t __b) { |
| 5031 | float32x2x2_t r; __builtin_neon_vuzp_v(&r, (int8x8_t)__a, (int8x8_t)__b, 8); return r; } |
| 5032 | __ai poly8x8x2_t vuzp_p8(poly8x8_t __a, poly8x8_t __b) { |
| 5033 | poly8x8x2_t r; __builtin_neon_vuzp_v(&r, (int8x8_t)__a, (int8x8_t)__b, 4); return r; } |
| 5034 | __ai poly16x4x2_t vuzp_p16(poly16x4_t __a, poly16x4_t __b) { |
| 5035 | poly16x4x2_t r; __builtin_neon_vuzp_v(&r, (int8x8_t)__a, (int8x8_t)__b, 5); return r; } |
| 5036 | __ai int8x16x2_t vuzpq_s8(int8x16_t __a, int8x16_t __b) { |
| 5037 | int8x16x2_t r; __builtin_neon_vuzpq_v(&r, __a, __b, 32); return r; } |
| 5038 | __ai int16x8x2_t vuzpq_s16(int16x8_t __a, int16x8_t __b) { |
| 5039 | int16x8x2_t r; __builtin_neon_vuzpq_v(&r, (int8x16_t)__a, (int8x16_t)__b, 33); return r; } |
| 5040 | __ai int32x4x2_t vuzpq_s32(int32x4_t __a, int32x4_t __b) { |
| 5041 | int32x4x2_t r; __builtin_neon_vuzpq_v(&r, (int8x16_t)__a, (int8x16_t)__b, 34); return r; } |
| 5042 | __ai uint8x16x2_t vuzpq_u8(uint8x16_t __a, uint8x16_t __b) { |
| 5043 | uint8x16x2_t r; __builtin_neon_vuzpq_v(&r, (int8x16_t)__a, (int8x16_t)__b, 48); return r; } |
| 5044 | __ai uint16x8x2_t vuzpq_u16(uint16x8_t __a, uint16x8_t __b) { |
| 5045 | uint16x8x2_t r; __builtin_neon_vuzpq_v(&r, (int8x16_t)__a, (int8x16_t)__b, 49); return r; } |
| 5046 | __ai uint32x4x2_t vuzpq_u32(uint32x4_t __a, uint32x4_t __b) { |
| 5047 | uint32x4x2_t r; __builtin_neon_vuzpq_v(&r, (int8x16_t)__a, (int8x16_t)__b, 50); return r; } |
| 5048 | __ai float32x4x2_t vuzpq_f32(float32x4_t __a, float32x4_t __b) { |
| 5049 | float32x4x2_t r; __builtin_neon_vuzpq_v(&r, (int8x16_t)__a, (int8x16_t)__b, 40); return r; } |
| 5050 | __ai poly8x16x2_t vuzpq_p8(poly8x16_t __a, poly8x16_t __b) { |
| 5051 | poly8x16x2_t r; __builtin_neon_vuzpq_v(&r, (int8x16_t)__a, (int8x16_t)__b, 36); return r; } |
| 5052 | __ai poly16x8x2_t vuzpq_p16(poly16x8_t __a, poly16x8_t __b) { |
| 5053 | poly16x8x2_t r; __builtin_neon_vuzpq_v(&r, (int8x16_t)__a, (int8x16_t)__b, 37); return r; } |
| 5054 | |
| 5055 | __ai int8x8x2_t vzip_s8(int8x8_t __a, int8x8_t __b) { |
| 5056 | int8x8x2_t r; __builtin_neon_vzip_v(&r, __a, __b, 0); return r; } |
| 5057 | __ai int16x4x2_t vzip_s16(int16x4_t __a, int16x4_t __b) { |
| 5058 | int16x4x2_t r; __builtin_neon_vzip_v(&r, (int8x8_t)__a, (int8x8_t)__b, 1); return r; } |
| 5059 | __ai int32x2x2_t vzip_s32(int32x2_t __a, int32x2_t __b) { |
| 5060 | int32x2x2_t r; __builtin_neon_vzip_v(&r, (int8x8_t)__a, (int8x8_t)__b, 2); return r; } |
| 5061 | __ai uint8x8x2_t vzip_u8(uint8x8_t __a, uint8x8_t __b) { |
| 5062 | uint8x8x2_t r; __builtin_neon_vzip_v(&r, (int8x8_t)__a, (int8x8_t)__b, 16); return r; } |
| 5063 | __ai uint16x4x2_t vzip_u16(uint16x4_t __a, uint16x4_t __b) { |
| 5064 | uint16x4x2_t r; __builtin_neon_vzip_v(&r, (int8x8_t)__a, (int8x8_t)__b, 17); return r; } |
| 5065 | __ai uint32x2x2_t vzip_u32(uint32x2_t __a, uint32x2_t __b) { |
| 5066 | uint32x2x2_t r; __builtin_neon_vzip_v(&r, (int8x8_t)__a, (int8x8_t)__b, 18); return r; } |
| 5067 | __ai float32x2x2_t vzip_f32(float32x2_t __a, float32x2_t __b) { |
| 5068 | float32x2x2_t r; __builtin_neon_vzip_v(&r, (int8x8_t)__a, (int8x8_t)__b, 8); return r; } |
| 5069 | __ai poly8x8x2_t vzip_p8(poly8x8_t __a, poly8x8_t __b) { |
| 5070 | poly8x8x2_t r; __builtin_neon_vzip_v(&r, (int8x8_t)__a, (int8x8_t)__b, 4); return r; } |
| 5071 | __ai poly16x4x2_t vzip_p16(poly16x4_t __a, poly16x4_t __b) { |
| 5072 | poly16x4x2_t r; __builtin_neon_vzip_v(&r, (int8x8_t)__a, (int8x8_t)__b, 5); return r; } |
| 5073 | __ai int8x16x2_t vzipq_s8(int8x16_t __a, int8x16_t __b) { |
| 5074 | int8x16x2_t r; __builtin_neon_vzipq_v(&r, __a, __b, 32); return r; } |
| 5075 | __ai int16x8x2_t vzipq_s16(int16x8_t __a, int16x8_t __b) { |
| 5076 | int16x8x2_t r; __builtin_neon_vzipq_v(&r, (int8x16_t)__a, (int8x16_t)__b, 33); return r; } |
| 5077 | __ai int32x4x2_t vzipq_s32(int32x4_t __a, int32x4_t __b) { |
| 5078 | int32x4x2_t r; __builtin_neon_vzipq_v(&r, (int8x16_t)__a, (int8x16_t)__b, 34); return r; } |
| 5079 | __ai uint8x16x2_t vzipq_u8(uint8x16_t __a, uint8x16_t __b) { |
| 5080 | uint8x16x2_t r; __builtin_neon_vzipq_v(&r, (int8x16_t)__a, (int8x16_t)__b, 48); return r; } |
| 5081 | __ai uint16x8x2_t vzipq_u16(uint16x8_t __a, uint16x8_t __b) { |
| 5082 | uint16x8x2_t r; __builtin_neon_vzipq_v(&r, (int8x16_t)__a, (int8x16_t)__b, 49); return r; } |
| 5083 | __ai uint32x4x2_t vzipq_u32(uint32x4_t __a, uint32x4_t __b) { |
| 5084 | uint32x4x2_t r; __builtin_neon_vzipq_v(&r, (int8x16_t)__a, (int8x16_t)__b, 50); return r; } |
| 5085 | __ai float32x4x2_t vzipq_f32(float32x4_t __a, float32x4_t __b) { |
| 5086 | float32x4x2_t r; __builtin_neon_vzipq_v(&r, (int8x16_t)__a, (int8x16_t)__b, 40); return r; } |
| 5087 | __ai poly8x16x2_t vzipq_p8(poly8x16_t __a, poly8x16_t __b) { |
| 5088 | poly8x16x2_t r; __builtin_neon_vzipq_v(&r, (int8x16_t)__a, (int8x16_t)__b, 36); return r; } |
| 5089 | __ai poly16x8x2_t vzipq_p16(poly16x8_t __a, poly16x8_t __b) { |
| 5090 | poly16x8x2_t r; __builtin_neon_vzipq_v(&r, (int8x16_t)__a, (int8x16_t)__b, 37); return r; } |
| 5091 | |
| 5092 | #ifdef __aarch64__ |
| 5093 | __ai int16x8_t vmovl_high_s8(int8x16_t __a) { |
| 5094 | int8x8_t __a1 = vget_high_s8(__a); |
| 5095 | return (int16x8_t)vshll_n_s8(__a1, 0); } |
| 5096 | __ai int32x4_t vmovl_high_s16(int16x8_t __a) { |
| 5097 | int16x4_t __a1 = vget_high_s16(__a); |
| 5098 | return (int32x4_t)vshll_n_s16(__a1, 0); } |
| 5099 | __ai int64x2_t vmovl_high_s32(int32x4_t __a) { |
| 5100 | int32x2_t __a1 = vget_high_s32(__a); |
| 5101 | return (int64x2_t)vshll_n_s32(__a1, 0); } |
| 5102 | __ai uint16x8_t vmovl_high_u8(uint8x16_t __a) { |
| 5103 | uint8x8_t __a1 = vget_high_u8(__a); |
| 5104 | return (uint16x8_t)vshll_n_u8(__a1, 0); } |
| 5105 | __ai uint32x4_t vmovl_high_u16(uint16x8_t __a) { |
| 5106 | uint16x4_t __a1 = vget_high_u16(__a); |
| 5107 | return (uint32x4_t)vshll_n_u16(__a1, 0); } |
| 5108 | __ai uint64x2_t vmovl_high_u32(uint32x4_t __a) { |
| 5109 | uint32x2_t __a1 = vget_high_u32(__a); |
| 5110 | return (uint64x2_t)vshll_n_u32(__a1, 0); } |
| 5111 | |
| 5112 | __ai int16x8_t vmull_high_s8(int8x16_t __a, int8x16_t __b) { |
| 5113 | return vmull_s8(vget_high_s8(__a), vget_high_s8(__b)); } |
| 5114 | __ai int32x4_t vmull_high_s16(int16x8_t __a, int16x8_t __b) { |
| 5115 | return vmull_s16(vget_high_s16(__a), vget_high_s16(__b)); } |
| 5116 | __ai int64x2_t vmull_high_s32(int32x4_t __a, int32x4_t __b) { |
| 5117 | return vmull_s32(vget_high_s32(__a), vget_high_s32(__b)); } |
| 5118 | __ai uint16x8_t vmull_high_u8(uint8x16_t __a, uint8x16_t __b) { |
| 5119 | return vmull_u8(vget_high_u8(__a), vget_high_u8(__b)); } |
| 5120 | __ai uint32x4_t vmull_high_u16(uint16x8_t __a, uint16x8_t __b) { |
| 5121 | return vmull_u16(vget_high_u16(__a), vget_high_u16(__b)); } |
| 5122 | __ai uint64x2_t vmull_high_u32(uint32x4_t __a, uint32x4_t __b) { |
| 5123 | return vmull_u32(vget_high_u32(__a), vget_high_u32(__b)); } |
| 5124 | __ai poly16x8_t vmull_high_p8(poly8x16_t __a, poly8x16_t __b) { |
| 5125 | return vmull_p8(vget_high_p8(__a), vget_high_p8(__b)); } |
| 5126 | |
| 5127 | __ai int16x8_t vabdl_high_s8(int8x16_t __a, int8x16_t __b) { |
| 5128 | return vabdl_s8(vget_high_s8(__a), vget_high_s8(__b)); } |
| 5129 | __ai int32x4_t vabdl_high_s16(int16x8_t __a, int16x8_t __b) { |
| 5130 | return vabdl_s16(vget_high_s16(__a), vget_high_s16(__b)); } |
| 5131 | __ai int64x2_t vabdl_high_s32(int32x4_t __a, int32x4_t __b) { |
| 5132 | return vabdl_s32(vget_high_s32(__a), vget_high_s32(__b)); } |
| 5133 | __ai uint16x8_t vabdl_high_u8(uint8x16_t __a, uint8x16_t __b) { |
| 5134 | return vabdl_u8(vget_high_u8(__a), vget_high_u8(__b)); } |
| 5135 | __ai uint32x4_t vabdl_high_u16(uint16x8_t __a, uint16x8_t __b) { |
| 5136 | return vabdl_u16(vget_high_u16(__a), vget_high_u16(__b)); } |
| 5137 | __ai uint64x2_t vabdl_high_u32(uint32x4_t __a, uint32x4_t __b) { |
| 5138 | return vabdl_u32(vget_high_u32(__a), vget_high_u32(__b)); } |
| 5139 | |
| 5140 | __ai float64x1_t vabd_f64(float64x1_t __a, float64x1_t __b) { |
| 5141 | return (float64x1_t)__builtin_neon_vabd_v((int8x8_t)__a, (int8x8_t)__b, 9); } |
| 5142 | __ai float64x2_t vabdq_f64(float64x2_t __a, float64x2_t __b) { |
| 5143 | return (float64x2_t)__builtin_neon_vabdq_v((int8x16_t)__a, (int8x16_t)__b, 41); } |
| 5144 | |
| 5145 | __ai int64x1_t vabs_s64(int64x1_t __a) { |
| 5146 | return (int64x1_t)__builtin_neon_vabs_v((int8x8_t)__a, 3); } |
| 5147 | __ai float64x1_t vabs_f64(float64x1_t __a) { |
| 5148 | return (float64x1_t)__builtin_neon_vabs_v((int8x8_t)__a, 9); } |
| 5149 | __ai int64x2_t vabsq_s64(int64x2_t __a) { |
| 5150 | return (int64x2_t)__builtin_neon_vabsq_v((int8x16_t)__a, 35); } |
| 5151 | __ai float64x2_t vabsq_f64(float64x2_t __a) { |
| 5152 | return (float64x2_t)__builtin_neon_vabsq_v((int8x16_t)__a, 41); } |
| 5153 | |
| 5154 | __ai float64x1_t vadd_f64(float64x1_t __a, float64x1_t __b) { |
| 5155 | return __a + __b; } |
| 5156 | __ai float64x2_t vaddq_f64(float64x2_t __a, float64x2_t __b) { |
| 5157 | return __a + __b; } |
| 5158 | |
| 5159 | __ai int8x16_t vpaddq_s8(int8x16_t __a, int8x16_t __b) { |
| 5160 | return (int8x16_t)__builtin_neon_vpaddq_v(__a, __b, 32); } |
| 5161 | __ai int16x8_t vpaddq_s16(int16x8_t __a, int16x8_t __b) { |
| 5162 | return (int16x8_t)__builtin_neon_vpaddq_v((int8x16_t)__a, (int8x16_t)__b, 33); } |
| 5163 | __ai int32x4_t vpaddq_s32(int32x4_t __a, int32x4_t __b) { |
| 5164 | return (int32x4_t)__builtin_neon_vpaddq_v((int8x16_t)__a, (int8x16_t)__b, 34); } |
| 5165 | __ai int64x2_t vpaddq_s64(int64x2_t __a, int64x2_t __b) { |
| 5166 | return (int64x2_t)__builtin_neon_vpaddq_v((int8x16_t)__a, (int8x16_t)__b, 35); } |
| 5167 | __ai uint8x16_t vpaddq_u8(uint8x16_t __a, uint8x16_t __b) { |
| 5168 | return (uint8x16_t)__builtin_neon_vpaddq_v((int8x16_t)__a, (int8x16_t)__b, 48); } |
| 5169 | __ai uint16x8_t vpaddq_u16(uint16x8_t __a, uint16x8_t __b) { |
| 5170 | return (uint16x8_t)__builtin_neon_vpaddq_v((int8x16_t)__a, (int8x16_t)__b, 49); } |
| 5171 | __ai uint32x4_t vpaddq_u32(uint32x4_t __a, uint32x4_t __b) { |
| 5172 | return (uint32x4_t)__builtin_neon_vpaddq_v((int8x16_t)__a, (int8x16_t)__b, 50); } |
| 5173 | __ai uint64x2_t vpaddq_u64(uint64x2_t __a, uint64x2_t __b) { |
| 5174 | return (uint64x2_t)__builtin_neon_vpaddq_v((int8x16_t)__a, (int8x16_t)__b, 51); } |
| 5175 | __ai float32x4_t vpaddq_f32(float32x4_t __a, float32x4_t __b) { |
| 5176 | return (float32x4_t)__builtin_neon_vpaddq_v((int8x16_t)__a, (int8x16_t)__b, 40); } |
| 5177 | __ai float64x2_t vpaddq_f64(float64x2_t __a, float64x2_t __b) { |
| 5178 | return (float64x2_t)__builtin_neon_vpaddq_v((int8x16_t)__a, (int8x16_t)__b, 41); } |
| 5179 | |
| 5180 | __ai float64x1_t vbsl_f64(uint64x1_t __a, float64x1_t __b, float64x1_t __c) { |
| 5181 | return (float64x1_t)__builtin_neon_vbsl_v((int8x8_t)__a, (int8x8_t)__b, (int8x8_t)__c, 9); } |
| 5182 | __ai float64x2_t vbslq_f64(uint64x2_t __a, float64x2_t __b, float64x2_t __c) { |
| 5183 | return (float64x2_t)__builtin_neon_vbslq_v((int8x16_t)__a, (int8x16_t)__b, (int8x16_t)__c, 41); } |
| 5184 | __ai poly64x1_t vbsl_p64(uint64x1_t __a, poly64x1_t __b, poly64x1_t __c) { |
| 5185 | return (poly64x1_t)__builtin_neon_vbsl_v((int8x8_t)__a, (int8x8_t)__b, (int8x8_t)__c, 6); } |
| 5186 | __ai poly64x2_t vbslq_p64(uint64x2_t __a, poly64x2_t __b, poly64x2_t __c) { |
| 5187 | return (poly64x2_t)__builtin_neon_vbslq_v((int8x16_t)__a, (int8x16_t)__b, (int8x16_t)__c, 38); } |
| 5188 | |
| 5189 | __ai uint64x1_t vceq_s64(int64x1_t __a, int64x1_t __b) { |
| 5190 | return (uint64x1_t)(__a == __b); } |
| 5191 | __ai uint64x1_t vceq_u64(uint64x1_t __a, uint64x1_t __b) { |
| 5192 | return (uint64x1_t)(__a == __b); } |
| 5193 | __ai uint64x1_t vceq_f64(float64x1_t __a, float64x1_t __b) { |
| 5194 | return (uint64x1_t)(__a == __b); } |
| 5195 | __ai uint64x2_t vceqq_f64(float64x2_t __a, float64x2_t __b) { |
| 5196 | return (uint64x2_t)(__a == __b); } |
| 5197 | __ai uint64x2_t vceqq_u64(uint64x2_t __a, uint64x2_t __b) { |
| 5198 | return (uint64x2_t)(__a == __b); } |
| 5199 | __ai uint64x2_t vceqq_s64(int64x2_t __a, int64x2_t __b) { |
| 5200 | return (uint64x2_t)(__a == __b); } |
| 5201 | __ai uint64x1_t vceq_p64(poly64x1_t __a, poly64x1_t __b) { |
| 5202 | return (uint64x1_t)(__a == __b); } |
| 5203 | __ai uint64x2_t vceqq_p64(poly64x2_t __a, poly64x2_t __b) { |
| 5204 | return (uint64x2_t)(__a == __b); } |
| 5205 | |
| 5206 | __ai uint64x1_t vcge_s64(int64x1_t __a, int64x1_t __b) { |
| 5207 | return (uint64x1_t)(__a >= __b); } |
| 5208 | __ai uint64x1_t vcge_u64(uint64x1_t __a, uint64x1_t __b) { |
| 5209 | return (uint64x1_t)(__a >= __b); } |
| 5210 | __ai uint64x2_t vcgeq_s64(int64x2_t __a, int64x2_t __b) { |
| 5211 | return (uint64x2_t)(__a >= __b); } |
| 5212 | __ai uint64x2_t vcgeq_u64(uint64x2_t __a, uint64x2_t __b) { |
| 5213 | return (uint64x2_t)(__a >= __b); } |
| 5214 | __ai uint64x1_t vcge_f64(float64x1_t __a, float64x1_t __b) { |
| 5215 | return (uint64x1_t)(__a >= __b); } |
| 5216 | __ai uint64x2_t vcgeq_f64(float64x2_t __a, float64x2_t __b) { |
| 5217 | return (uint64x2_t)(__a >= __b); } |
| 5218 | |
| 5219 | __ai uint64x1_t vcgt_s64(int64x1_t __a, int64x1_t __b) { |
| 5220 | return (uint64x1_t)(__a > __b); } |
| 5221 | __ai uint64x1_t vcgt_u64(uint64x1_t __a, uint64x1_t __b) { |
| 5222 | return (uint64x1_t)(__a > __b); } |
| 5223 | __ai uint64x2_t vcgtq_s64(int64x2_t __a, int64x2_t __b) { |
| 5224 | return (uint64x2_t)(__a > __b); } |
| 5225 | __ai uint64x2_t vcgtq_u64(uint64x2_t __a, uint64x2_t __b) { |
| 5226 | return (uint64x2_t)(__a > __b); } |
| 5227 | __ai uint64x1_t vcgt_f64(float64x1_t __a, float64x1_t __b) { |
| 5228 | return (uint64x1_t)(__a > __b); } |
| 5229 | __ai uint64x2_t vcgtq_f64(float64x2_t __a, float64x2_t __b) { |
| 5230 | return (uint64x2_t)(__a > __b); } |
| 5231 | |
| 5232 | __ai uint64x1_t vcle_s64(int64x1_t __a, int64x1_t __b) { |
| 5233 | return (uint64x1_t)(__a <= __b); } |
| 5234 | __ai uint64x1_t vcle_u64(uint64x1_t __a, uint64x1_t __b) { |
| 5235 | return (uint64x1_t)(__a <= __b); } |
| 5236 | __ai uint64x2_t vcleq_s64(int64x2_t __a, int64x2_t __b) { |
| 5237 | return (uint64x2_t)(__a <= __b); } |
| 5238 | __ai uint64x2_t vcleq_u64(uint64x2_t __a, uint64x2_t __b) { |
| 5239 | return (uint64x2_t)(__a <= __b); } |
| 5240 | __ai uint64x1_t vcle_f64(float64x1_t __a, float64x1_t __b) { |
| 5241 | return (uint64x1_t)(__a <= __b); } |
| 5242 | __ai uint64x2_t vcleq_f64(float64x2_t __a, float64x2_t __b) { |
| 5243 | return (uint64x2_t)(__a <= __b); } |
| 5244 | |
| 5245 | __ai uint64x1_t vclt_s64(int64x1_t __a, int64x1_t __b) { |
| 5246 | return (uint64x1_t)(__a < __b); } |
| 5247 | __ai uint64x1_t vclt_u64(uint64x1_t __a, uint64x1_t __b) { |
| 5248 | return (uint64x1_t)(__a < __b); } |
| 5249 | __ai uint64x2_t vcltq_s64(int64x2_t __a, int64x2_t __b) { |
| 5250 | return (uint64x2_t)(__a < __b); } |
| 5251 | __ai uint64x2_t vcltq_u64(uint64x2_t __a, uint64x2_t __b) { |
| 5252 | return (uint64x2_t)(__a < __b); } |
| 5253 | __ai uint64x1_t vclt_f64(float64x1_t __a, float64x1_t __b) { |
| 5254 | return (uint64x1_t)(__a < __b); } |
| 5255 | __ai uint64x2_t vcltq_f64(float64x2_t __a, float64x2_t __b) { |
| 5256 | return (uint64x2_t)(__a < __b); } |
| 5257 | |
| 5258 | __ai uint8x8_t vceqz_s8(int8x8_t __a) { |
| 5259 | return (uint8x8_t)__builtin_neon_vceqz_v(__a, 16); } |
| 5260 | __ai uint16x4_t vceqz_s16(int16x4_t __a) { |
| 5261 | return (uint16x4_t)__builtin_neon_vceqz_v((int8x8_t)__a, 17); } |
| 5262 | __ai uint32x2_t vceqz_s32(int32x2_t __a) { |
| 5263 | return (uint32x2_t)__builtin_neon_vceqz_v((int8x8_t)__a, 18); } |
| 5264 | __ai uint64x1_t vceqz_s64(int64x1_t __a) { |
| 5265 | return (uint64x1_t)__builtin_neon_vceqz_v((int8x8_t)__a, 19); } |
| 5266 | __ai uint32x2_t vceqz_f32(float32x2_t __a) { |
| 5267 | return (uint32x2_t)__builtin_neon_vceqz_v((int8x8_t)__a, 18); } |
| 5268 | __ai uint8x8_t vceqz_u8(uint8x8_t __a) { |
| 5269 | return (uint8x8_t)__builtin_neon_vceqz_v((int8x8_t)__a, 16); } |
| 5270 | __ai uint16x4_t vceqz_u16(uint16x4_t __a) { |
| 5271 | return (uint16x4_t)__builtin_neon_vceqz_v((int8x8_t)__a, 17); } |
| 5272 | __ai uint32x2_t vceqz_u32(uint32x2_t __a) { |
| 5273 | return (uint32x2_t)__builtin_neon_vceqz_v((int8x8_t)__a, 18); } |
| 5274 | __ai uint64x1_t vceqz_u64(uint64x1_t __a) { |
| 5275 | return (uint64x1_t)__builtin_neon_vceqz_v((int8x8_t)__a, 19); } |
| 5276 | __ai uint8x8_t vceqz_p8(poly8x8_t __a) { |
| 5277 | return (uint8x8_t)__builtin_neon_vceqz_v((int8x8_t)__a, 16); } |
| 5278 | __ai uint16x4_t vceqz_p16(poly16x4_t __a) { |
| 5279 | return (uint16x4_t)__builtin_neon_vceqz_v((int8x8_t)__a, 17); } |
| 5280 | __ai uint64x1_t vceqz_p64(poly64x1_t __a) { |
| 5281 | return (uint64x1_t)__builtin_neon_vceqz_v((int8x8_t)__a, 19); } |
| 5282 | __ai uint8x16_t vceqzq_s8(int8x16_t __a) { |
| 5283 | return (uint8x16_t)__builtin_neon_vceqzq_v(__a, 48); } |
| 5284 | __ai uint16x8_t vceqzq_s16(int16x8_t __a) { |
| 5285 | return (uint16x8_t)__builtin_neon_vceqzq_v((int8x16_t)__a, 49); } |
| 5286 | __ai uint32x4_t vceqzq_s32(int32x4_t __a) { |
| 5287 | return (uint32x4_t)__builtin_neon_vceqzq_v((int8x16_t)__a, 50); } |
| 5288 | __ai uint64x2_t vceqzq_s64(int64x2_t __a) { |
| 5289 | return (uint64x2_t)__builtin_neon_vceqzq_v((int8x16_t)__a, 51); } |
| 5290 | __ai uint32x4_t vceqzq_f32(float32x4_t __a) { |
| 5291 | return (uint32x4_t)__builtin_neon_vceqzq_v((int8x16_t)__a, 50); } |
| 5292 | __ai uint8x16_t vceqzq_u8(uint8x16_t __a) { |
| 5293 | return (uint8x16_t)__builtin_neon_vceqzq_v((int8x16_t)__a, 48); } |
| 5294 | __ai uint16x8_t vceqzq_u16(uint16x8_t __a) { |
| 5295 | return (uint16x8_t)__builtin_neon_vceqzq_v((int8x16_t)__a, 49); } |
| 5296 | __ai uint32x4_t vceqzq_u32(uint32x4_t __a) { |
| 5297 | return (uint32x4_t)__builtin_neon_vceqzq_v((int8x16_t)__a, 50); } |
| 5298 | __ai uint64x2_t vceqzq_u64(uint64x2_t __a) { |
| 5299 | return (uint64x2_t)__builtin_neon_vceqzq_v((int8x16_t)__a, 51); } |
| 5300 | __ai uint8x16_t vceqzq_p8(poly8x16_t __a) { |
| 5301 | return (uint8x16_t)__builtin_neon_vceqzq_v((int8x16_t)__a, 48); } |
| 5302 | __ai uint16x8_t vceqzq_p16(poly16x8_t __a) { |
| 5303 | return (uint16x8_t)__builtin_neon_vceqzq_v((int8x16_t)__a, 49); } |
| 5304 | __ai uint64x1_t vceqz_f64(float64x1_t __a) { |
| 5305 | return (uint64x1_t)__builtin_neon_vceqz_v((int8x8_t)__a, 19); } |
| 5306 | __ai uint64x2_t vceqzq_f64(float64x2_t __a) { |
| 5307 | return (uint64x2_t)__builtin_neon_vceqzq_v((int8x16_t)__a, 51); } |
| 5308 | __ai uint64x2_t vceqzq_p64(poly64x2_t __a) { |
| 5309 | return (uint64x2_t)__builtin_neon_vceqzq_v((int8x16_t)__a, 51); } |
| 5310 | |
| 5311 | __ai uint8x8_t vcgez_s8(int8x8_t __a) { |
| 5312 | return (uint8x8_t)__builtin_neon_vcgez_v(__a, 16); } |
| 5313 | __ai uint16x4_t vcgez_s16(int16x4_t __a) { |
| 5314 | return (uint16x4_t)__builtin_neon_vcgez_v((int8x8_t)__a, 17); } |
| 5315 | __ai uint32x2_t vcgez_s32(int32x2_t __a) { |
| 5316 | return (uint32x2_t)__builtin_neon_vcgez_v((int8x8_t)__a, 18); } |
| 5317 | __ai uint64x1_t vcgez_s64(int64x1_t __a) { |
| 5318 | return (uint64x1_t)__builtin_neon_vcgez_v((int8x8_t)__a, 19); } |
| 5319 | __ai uint32x2_t vcgez_f32(float32x2_t __a) { |
| 5320 | return (uint32x2_t)__builtin_neon_vcgez_v((int8x8_t)__a, 18); } |
| 5321 | __ai uint64x1_t vcgez_f64(float64x1_t __a) { |
| 5322 | return (uint64x1_t)__builtin_neon_vcgez_v((int8x8_t)__a, 19); } |
| 5323 | __ai uint8x16_t vcgezq_s8(int8x16_t __a) { |
| 5324 | return (uint8x16_t)__builtin_neon_vcgezq_v(__a, 48); } |
| 5325 | __ai uint16x8_t vcgezq_s16(int16x8_t __a) { |
| 5326 | return (uint16x8_t)__builtin_neon_vcgezq_v((int8x16_t)__a, 49); } |
| 5327 | __ai uint32x4_t vcgezq_s32(int32x4_t __a) { |
| 5328 | return (uint32x4_t)__builtin_neon_vcgezq_v((int8x16_t)__a, 50); } |
| 5329 | __ai uint64x2_t vcgezq_s64(int64x2_t __a) { |
| 5330 | return (uint64x2_t)__builtin_neon_vcgezq_v((int8x16_t)__a, 51); } |
| 5331 | __ai uint32x4_t vcgezq_f32(float32x4_t __a) { |
| 5332 | return (uint32x4_t)__builtin_neon_vcgezq_v((int8x16_t)__a, 50); } |
| 5333 | __ai uint64x2_t vcgezq_f64(float64x2_t __a) { |
| 5334 | return (uint64x2_t)__builtin_neon_vcgezq_v((int8x16_t)__a, 51); } |
| 5335 | |
| 5336 | __ai uint8x8_t vcgtz_s8(int8x8_t __a) { |
| 5337 | return (uint8x8_t)__builtin_neon_vcgtz_v(__a, 16); } |
| 5338 | __ai uint16x4_t vcgtz_s16(int16x4_t __a) { |
| 5339 | return (uint16x4_t)__builtin_neon_vcgtz_v((int8x8_t)__a, 17); } |
| 5340 | __ai uint32x2_t vcgtz_s32(int32x2_t __a) { |
| 5341 | return (uint32x2_t)__builtin_neon_vcgtz_v((int8x8_t)__a, 18); } |
| 5342 | __ai uint64x1_t vcgtz_s64(int64x1_t __a) { |
| 5343 | return (uint64x1_t)__builtin_neon_vcgtz_v((int8x8_t)__a, 19); } |
| 5344 | __ai uint32x2_t vcgtz_f32(float32x2_t __a) { |
| 5345 | return (uint32x2_t)__builtin_neon_vcgtz_v((int8x8_t)__a, 18); } |
| 5346 | __ai uint64x1_t vcgtz_f64(float64x1_t __a) { |
| 5347 | return (uint64x1_t)__builtin_neon_vcgtz_v((int8x8_t)__a, 19); } |
| 5348 | __ai uint8x16_t vcgtzq_s8(int8x16_t __a) { |
| 5349 | return (uint8x16_t)__builtin_neon_vcgtzq_v(__a, 48); } |
| 5350 | __ai uint16x8_t vcgtzq_s16(int16x8_t __a) { |
| 5351 | return (uint16x8_t)__builtin_neon_vcgtzq_v((int8x16_t)__a, 49); } |
| 5352 | __ai uint32x4_t vcgtzq_s32(int32x4_t __a) { |
| 5353 | return (uint32x4_t)__builtin_neon_vcgtzq_v((int8x16_t)__a, 50); } |
| 5354 | __ai uint64x2_t vcgtzq_s64(int64x2_t __a) { |
| 5355 | return (uint64x2_t)__builtin_neon_vcgtzq_v((int8x16_t)__a, 51); } |
| 5356 | __ai uint32x4_t vcgtzq_f32(float32x4_t __a) { |
| 5357 | return (uint32x4_t)__builtin_neon_vcgtzq_v((int8x16_t)__a, 50); } |
| 5358 | __ai uint64x2_t vcgtzq_f64(float64x2_t __a) { |
| 5359 | return (uint64x2_t)__builtin_neon_vcgtzq_v((int8x16_t)__a, 51); } |
| 5360 | |
| 5361 | __ai uint8x8_t vclez_s8(int8x8_t __a) { |
| 5362 | return (uint8x8_t)__builtin_neon_vclez_v(__a, 16); } |
| 5363 | __ai uint16x4_t vclez_s16(int16x4_t __a) { |
| 5364 | return (uint16x4_t)__builtin_neon_vclez_v((int8x8_t)__a, 17); } |
| 5365 | __ai uint32x2_t vclez_s32(int32x2_t __a) { |
| 5366 | return (uint32x2_t)__builtin_neon_vclez_v((int8x8_t)__a, 18); } |
| 5367 | __ai uint64x1_t vclez_s64(int64x1_t __a) { |
| 5368 | return (uint64x1_t)__builtin_neon_vclez_v((int8x8_t)__a, 19); } |
| 5369 | __ai uint32x2_t vclez_f32(float32x2_t __a) { |
| 5370 | return (uint32x2_t)__builtin_neon_vclez_v((int8x8_t)__a, 18); } |
| 5371 | __ai uint64x1_t vclez_f64(float64x1_t __a) { |
| 5372 | return (uint64x1_t)__builtin_neon_vclez_v((int8x8_t)__a, 19); } |
| 5373 | __ai uint8x16_t vclezq_s8(int8x16_t __a) { |
| 5374 | return (uint8x16_t)__builtin_neon_vclezq_v(__a, 48); } |
| 5375 | __ai uint16x8_t vclezq_s16(int16x8_t __a) { |
| 5376 | return (uint16x8_t)__builtin_neon_vclezq_v((int8x16_t)__a, 49); } |
| 5377 | __ai uint32x4_t vclezq_s32(int32x4_t __a) { |
| 5378 | return (uint32x4_t)__builtin_neon_vclezq_v((int8x16_t)__a, 50); } |
| 5379 | __ai uint64x2_t vclezq_s64(int64x2_t __a) { |
| 5380 | return (uint64x2_t)__builtin_neon_vclezq_v((int8x16_t)__a, 51); } |
| 5381 | __ai uint32x4_t vclezq_f32(float32x4_t __a) { |
| 5382 | return (uint32x4_t)__builtin_neon_vclezq_v((int8x16_t)__a, 50); } |
| 5383 | __ai uint64x2_t vclezq_f64(float64x2_t __a) { |
| 5384 | return (uint64x2_t)__builtin_neon_vclezq_v((int8x16_t)__a, 51); } |
| 5385 | |
| 5386 | __ai uint8x8_t vcltz_s8(int8x8_t __a) { |
| 5387 | return (uint8x8_t)__builtin_neon_vcltz_v(__a, 16); } |
| 5388 | __ai uint16x4_t vcltz_s16(int16x4_t __a) { |
| 5389 | return (uint16x4_t)__builtin_neon_vcltz_v((int8x8_t)__a, 17); } |
| 5390 | __ai uint32x2_t vcltz_s32(int32x2_t __a) { |
| 5391 | return (uint32x2_t)__builtin_neon_vcltz_v((int8x8_t)__a, 18); } |
| 5392 | __ai uint64x1_t vcltz_s64(int64x1_t __a) { |
| 5393 | return (uint64x1_t)__builtin_neon_vcltz_v((int8x8_t)__a, 19); } |
| 5394 | __ai uint32x2_t vcltz_f32(float32x2_t __a) { |
| 5395 | return (uint32x2_t)__builtin_neon_vcltz_v((int8x8_t)__a, 18); } |
| 5396 | __ai uint64x1_t vcltz_f64(float64x1_t __a) { |
| 5397 | return (uint64x1_t)__builtin_neon_vcltz_v((int8x8_t)__a, 19); } |
| 5398 | __ai uint8x16_t vcltzq_s8(int8x16_t __a) { |
| 5399 | return (uint8x16_t)__builtin_neon_vcltzq_v(__a, 48); } |
| 5400 | __ai uint16x8_t vcltzq_s16(int16x8_t __a) { |
| 5401 | return (uint16x8_t)__builtin_neon_vcltzq_v((int8x16_t)__a, 49); } |
| 5402 | __ai uint32x4_t vcltzq_s32(int32x4_t __a) { |
| 5403 | return (uint32x4_t)__builtin_neon_vcltzq_v((int8x16_t)__a, 50); } |
| 5404 | __ai uint64x2_t vcltzq_s64(int64x2_t __a) { |
| 5405 | return (uint64x2_t)__builtin_neon_vcltzq_v((int8x16_t)__a, 51); } |
| 5406 | __ai uint32x4_t vcltzq_f32(float32x4_t __a) { |
| 5407 | return (uint32x4_t)__builtin_neon_vcltzq_v((int8x16_t)__a, 50); } |
| 5408 | __ai uint64x2_t vcltzq_f64(float64x2_t __a) { |
| 5409 | return (uint64x2_t)__builtin_neon_vcltzq_v((int8x16_t)__a, 51); } |
| 5410 | |
| 5411 | __ai uint64x1_t vtst_s64(int64x1_t __a, int64x1_t __b) { |
| 5412 | return (uint64x1_t)__builtin_neon_vtst_v((int8x8_t)__a, (int8x8_t)__b, 19); } |
| 5413 | __ai uint64x1_t vtst_u64(uint64x1_t __a, uint64x1_t __b) { |
| 5414 | return (uint64x1_t)__builtin_neon_vtst_v((int8x8_t)__a, (int8x8_t)__b, 19); } |
| 5415 | __ai uint64x2_t vtstq_s64(int64x2_t __a, int64x2_t __b) { |
| 5416 | return (uint64x2_t)__builtin_neon_vtstq_v((int8x16_t)__a, (int8x16_t)__b, 51); } |
| 5417 | __ai uint64x2_t vtstq_u64(uint64x2_t __a, uint64x2_t __b) { |
| 5418 | return (uint64x2_t)__builtin_neon_vtstq_v((int8x16_t)__a, (int8x16_t)__b, 51); } |
| 5419 | __ai uint64x1_t vtst_p64(poly64x1_t __a, poly64x1_t __b) { |
| 5420 | return (uint64x1_t)__builtin_neon_vtst_v((int8x8_t)__a, (int8x8_t)__b, 19); } |
| 5421 | __ai uint64x2_t vtstq_p64(poly64x2_t __a, poly64x2_t __b) { |
| 5422 | return (uint64x2_t)__builtin_neon_vtstq_v((int8x16_t)__a, (int8x16_t)__b, 51); } |
| 5423 | |
| 5424 | __ai float64x2_t vcombine_f64(float64x1_t __a, float64x1_t __b) { |
| 5425 | return (float64x2_t)__builtin_shufflevector((int64x1_t)__a, (int64x1_t)__b, 0, 1); } |
| 5426 | __ai poly64x2_t vcombine_p64(poly64x1_t __a, poly64x1_t __b) { |
| 5427 | return (poly64x2_t)__builtin_shufflevector((int64x1_t)__a, (int64x1_t)__b, 0, 1); } |
| 5428 | |
| 5429 | #define vcopyq_lane_s8(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5430 | int8x16_t __a1 = (a1); int8x8_t __c1 = (c1); \ |
| 5431 | int8_t __c2 = vget_lane_s8(__c1, __d1); \ |
| 5432 | vsetq_lane_s8(__c2, __a1, __b1); }) |
| 5433 | #define vcopyq_lane_s16(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5434 | int16x8_t __a1 = (a1); int16x4_t __c1 = (c1); \ |
| 5435 | int16_t __c2 = vget_lane_s16(__c1, __d1); \ |
| 5436 | vsetq_lane_s16(__c2, __a1, __b1); }) |
| 5437 | #define vcopyq_lane_s32(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5438 | int32x4_t __a1 = (a1); int32x2_t __c1 = (c1); \ |
| 5439 | int32_t __c2 = vget_lane_s32(__c1, __d1); \ |
| 5440 | vsetq_lane_s32(__c2, __a1, __b1); }) |
| 5441 | #define vcopyq_lane_s64(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5442 | int64x2_t __a1 = (a1); int64x1_t __c1 = (c1); \ |
| 5443 | int64_t __c2 = vget_lane_s64(__c1, __d1); \ |
| 5444 | vsetq_lane_s64(__c2, __a1, __b1); }) |
| 5445 | #define vcopyq_lane_u8(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5446 | uint8x16_t __a1 = (a1); uint8x8_t __c1 = (c1); \ |
| 5447 | uint8_t __c2 = vget_lane_u8(__c1, __d1); \ |
| 5448 | vsetq_lane_u8(__c2, __a1, __b1); }) |
| 5449 | #define vcopyq_lane_u16(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5450 | uint16x8_t __a1 = (a1); uint16x4_t __c1 = (c1); \ |
| 5451 | uint16_t __c2 = vget_lane_u16(__c1, __d1); \ |
| 5452 | vsetq_lane_u16(__c2, __a1, __b1); }) |
| 5453 | #define vcopyq_lane_u32(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5454 | uint32x4_t __a1 = (a1); uint32x2_t __c1 = (c1); \ |
| 5455 | uint32_t __c2 = vget_lane_u32(__c1, __d1); \ |
| 5456 | vsetq_lane_u32(__c2, __a1, __b1); }) |
| 5457 | #define vcopyq_lane_u64(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5458 | uint64x2_t __a1 = (a1); uint64x1_t __c1 = (c1); \ |
| 5459 | uint64_t __c2 = vget_lane_u64(__c1, __d1); \ |
| 5460 | vsetq_lane_u64(__c2, __a1, __b1); }) |
| 5461 | #define vcopyq_lane_p8(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5462 | poly8x16_t __a1 = (a1); poly8x8_t __c1 = (c1); \ |
| 5463 | poly8_t __c2 = vget_lane_p8(__c1, __d1); \ |
| 5464 | vsetq_lane_p8(__c2, __a1, __b1); }) |
| 5465 | #define vcopyq_lane_p16(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5466 | poly16x8_t __a1 = (a1); poly16x4_t __c1 = (c1); \ |
| 5467 | poly16_t __c2 = vget_lane_p16(__c1, __d1); \ |
| 5468 | vsetq_lane_p16(__c2, __a1, __b1); }) |
| 5469 | #define vcopyq_lane_f32(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5470 | float32x4_t __a1 = (a1); float32x2_t __c1 = (c1); \ |
| 5471 | float32_t __c2 = vget_lane_f32(__c1, __d1); \ |
| 5472 | vsetq_lane_f32(__c2, __a1, __b1); }) |
| 5473 | #define vcopyq_lane_f64(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5474 | float64x2_t __a1 = (a1); float64x1_t __c1 = (c1); \ |
| 5475 | float64_t __c2 = vget_lane_f64(__c1, __d1); \ |
| 5476 | vsetq_lane_f64(__c2, __a1, __b1); }) |
| 5477 | #define vcopyq_lane_p64(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5478 | poly64x2_t __a1 = (a1); poly64x1_t __c1 = (c1); \ |
| 5479 | poly64_t __c2 = vget_lane_p64(__c1, __d1); \ |
| 5480 | vsetq_lane_p64(__c2, __a1, __b1); }) |
| 5481 | |
| 5482 | #define vcopyq_laneq_s8(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5483 | int8x16_t __a1 = (a1); int8x16_t __c1 = (c1); \ |
| 5484 | int8_t __c2 = vgetq_lane_s8(__c1, __d1); \ |
| 5485 | vsetq_lane_s8(__c2, __a1, __b1); }) |
| 5486 | #define vcopyq_laneq_s16(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5487 | int16x8_t __a1 = (a1); int16x8_t __c1 = (c1); \ |
| 5488 | int16_t __c2 = vgetq_lane_s16(__c1, __d1); \ |
| 5489 | vsetq_lane_s16(__c2, __a1, __b1); }) |
| 5490 | #define vcopyq_laneq_s32(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5491 | int32x4_t __a1 = (a1); int32x4_t __c1 = (c1); \ |
| 5492 | int32_t __c2 = vgetq_lane_s32(__c1, __d1); \ |
| 5493 | vsetq_lane_s32(__c2, __a1, __b1); }) |
| 5494 | #define vcopyq_laneq_s64(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5495 | int64x2_t __a1 = (a1); int64x2_t __c1 = (c1); \ |
| 5496 | int64_t __c2 = vgetq_lane_s64(__c1, __d1); \ |
| 5497 | vsetq_lane_s64(__c2, __a1, __b1); }) |
| 5498 | #define vcopyq_laneq_u8(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5499 | uint8x16_t __a1 = (a1); uint8x16_t __c1 = (c1); \ |
| 5500 | uint8_t __c2 = vgetq_lane_u8(__c1, __d1); \ |
| 5501 | vsetq_lane_u8(__c2, __a1, __b1); }) |
| 5502 | #define vcopyq_laneq_u16(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5503 | uint16x8_t __a1 = (a1); uint16x8_t __c1 = (c1); \ |
| 5504 | uint16_t __c2 = vgetq_lane_u16(__c1, __d1); \ |
| 5505 | vsetq_lane_u16(__c2, __a1, __b1); }) |
| 5506 | #define vcopyq_laneq_u32(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5507 | uint32x4_t __a1 = (a1); uint32x4_t __c1 = (c1); \ |
| 5508 | uint32_t __c2 = vgetq_lane_u32(__c1, __d1); \ |
| 5509 | vsetq_lane_u32(__c2, __a1, __b1); }) |
| 5510 | #define vcopyq_laneq_u64(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5511 | uint64x2_t __a1 = (a1); uint64x2_t __c1 = (c1); \ |
| 5512 | uint64_t __c2 = vgetq_lane_u64(__c1, __d1); \ |
| 5513 | vsetq_lane_u64(__c2, __a1, __b1); }) |
| 5514 | #define vcopyq_laneq_p8(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5515 | poly8x16_t __a1 = (a1); poly8x16_t __c1 = (c1); \ |
| 5516 | poly8_t __c2 = vgetq_lane_p8(__c1, __d1); \ |
| 5517 | vsetq_lane_p8(__c2, __a1, __b1); }) |
| 5518 | #define vcopyq_laneq_p16(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5519 | poly16x8_t __a1 = (a1); poly16x8_t __c1 = (c1); \ |
| 5520 | poly16_t __c2 = vgetq_lane_p16(__c1, __d1); \ |
| 5521 | vsetq_lane_p16(__c2, __a1, __b1); }) |
| 5522 | #define vcopyq_laneq_f32(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5523 | float32x4_t __a1 = (a1); float32x4_t __c1 = (c1); \ |
| 5524 | float32_t __c2 = vgetq_lane_f32(__c1, __d1); \ |
| 5525 | vsetq_lane_f32(__c2, __a1, __b1); }) |
| 5526 | #define vcopyq_laneq_f64(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5527 | float64x2_t __a1 = (a1); float64x2_t __c1 = (c1); \ |
| 5528 | float64_t __c2 = vgetq_lane_f64(__c1, __d1); \ |
| 5529 | vsetq_lane_f64(__c2, __a1, __b1); }) |
| 5530 | #define vcopyq_laneq_p64(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5531 | poly64x2_t __a1 = (a1); poly64x2_t __c1 = (c1); \ |
| 5532 | poly64_t __c2 = vgetq_lane_p64(__c1, __d1); \ |
| 5533 | vsetq_lane_p64(__c2, __a1, __b1); }) |
| 5534 | |
| 5535 | #define vcopy_lane_s8(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5536 | int8x8_t __a1 = (a1); int8x8_t __c1 = (c1); \ |
| 5537 | int8_t __c2 = vget_lane_s8(__c1, __d1); \ |
| 5538 | vset_lane_s8(__c2, __a1, __b1); }) |
| 5539 | #define vcopy_lane_s16(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5540 | int16x4_t __a1 = (a1); int16x4_t __c1 = (c1); \ |
| 5541 | int16_t __c2 = vget_lane_s16(__c1, __d1); \ |
| 5542 | vset_lane_s16(__c2, __a1, __b1); }) |
| 5543 | #define vcopy_lane_s32(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5544 | int32x2_t __a1 = (a1); int32x2_t __c1 = (c1); \ |
| 5545 | int32_t __c2 = vget_lane_s32(__c1, __d1); \ |
| 5546 | vset_lane_s32(__c2, __a1, __b1); }) |
| 5547 | #define vcopy_lane_s64(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5548 | int64x1_t __a1 = (a1); int64x1_t __c1 = (c1); \ |
| 5549 | int64_t __c2 = vget_lane_s64(__c1, __d1); \ |
| 5550 | vset_lane_s64(__c2, __a1, __b1); }) |
| 5551 | #define vcopy_lane_p8(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5552 | poly8x8_t __a1 = (a1); poly8x8_t __c1 = (c1); \ |
| 5553 | poly8_t __c2 = vget_lane_p8(__c1, __d1); \ |
| 5554 | vset_lane_p8(__c2, __a1, __b1); }) |
| 5555 | #define vcopy_lane_p16(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5556 | poly16x4_t __a1 = (a1); poly16x4_t __c1 = (c1); \ |
| 5557 | poly16_t __c2 = vget_lane_p16(__c1, __d1); \ |
| 5558 | vset_lane_p16(__c2, __a1, __b1); }) |
| 5559 | #define vcopy_lane_u8(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5560 | uint8x8_t __a1 = (a1); uint8x8_t __c1 = (c1); \ |
| 5561 | uint8_t __c2 = vget_lane_u8(__c1, __d1); \ |
| 5562 | vset_lane_u8(__c2, __a1, __b1); }) |
| 5563 | #define vcopy_lane_u16(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5564 | uint16x4_t __a1 = (a1); uint16x4_t __c1 = (c1); \ |
| 5565 | uint16_t __c2 = vget_lane_u16(__c1, __d1); \ |
| 5566 | vset_lane_u16(__c2, __a1, __b1); }) |
| 5567 | #define vcopy_lane_u32(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5568 | uint32x2_t __a1 = (a1); uint32x2_t __c1 = (c1); \ |
| 5569 | uint32_t __c2 = vget_lane_u32(__c1, __d1); \ |
| 5570 | vset_lane_u32(__c2, __a1, __b1); }) |
| 5571 | #define vcopy_lane_u64(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5572 | uint64x1_t __a1 = (a1); uint64x1_t __c1 = (c1); \ |
| 5573 | uint64_t __c2 = vget_lane_u64(__c1, __d1); \ |
| 5574 | vset_lane_u64(__c2, __a1, __b1); }) |
| 5575 | #define vcopy_lane_p64(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5576 | poly64x1_t __a1 = (a1); poly64x1_t __c1 = (c1); \ |
| 5577 | poly64_t __c2 = vget_lane_p64(__c1, __d1); \ |
| 5578 | vset_lane_p64(__c2, __a1, __b1); }) |
| 5579 | #define vcopy_lane_f32(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5580 | float32x2_t __a1 = (a1); float32x2_t __c1 = (c1); \ |
| 5581 | float32_t __c2 = vget_lane_f32(__c1, __d1); \ |
| 5582 | vset_lane_f32(__c2, __a1, __b1); }) |
| 5583 | #define vcopy_lane_f64(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5584 | float64x1_t __a1 = (a1); float64x1_t __c1 = (c1); \ |
| 5585 | float64_t __c2 = vget_lane_f64(__c1, __d1); \ |
| 5586 | vset_lane_f64(__c2, __a1, __b1); }) |
| 5587 | |
| 5588 | #define vcopy_laneq_s8(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5589 | int8x8_t __a1 = (a1); int8x16_t __c1 = (c1); \ |
| 5590 | int8_t __c2 = vgetq_lane_s8(__c1, __d1); \ |
| 5591 | vset_lane_s8(__c2, __a1, __b1); }) |
| 5592 | #define vcopy_laneq_s16(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5593 | int16x4_t __a1 = (a1); int16x8_t __c1 = (c1); \ |
| 5594 | int16_t __c2 = vgetq_lane_s16(__c1, __d1); \ |
| 5595 | vset_lane_s16(__c2, __a1, __b1); }) |
| 5596 | #define vcopy_laneq_s32(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5597 | int32x2_t __a1 = (a1); int32x4_t __c1 = (c1); \ |
| 5598 | int32_t __c2 = vgetq_lane_s32(__c1, __d1); \ |
| 5599 | vset_lane_s32(__c2, __a1, __b1); }) |
| 5600 | #define vcopy_laneq_s64(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5601 | int64x1_t __a1 = (a1); int64x2_t __c1 = (c1); \ |
| 5602 | int64_t __c2 = vgetq_lane_s64(__c1, __d1); \ |
| 5603 | vset_lane_s64(__c2, __a1, __b1); }) |
| 5604 | #define vcopy_laneq_p8(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5605 | poly8x8_t __a1 = (a1); poly8x16_t __c1 = (c1); \ |
| 5606 | poly8_t __c2 = vgetq_lane_p8(__c1, __d1); \ |
| 5607 | vset_lane_p8(__c2, __a1, __b1); }) |
| 5608 | #define vcopy_laneq_p16(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5609 | poly16x4_t __a1 = (a1); poly16x8_t __c1 = (c1); \ |
| 5610 | poly16_t __c2 = vgetq_lane_p16(__c1, __d1); \ |
| 5611 | vset_lane_p16(__c2, __a1, __b1); }) |
| 5612 | #define vcopy_laneq_p64(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5613 | poly64x1_t __a1 = (a1); poly64x2_t __c1 = (c1); \ |
| 5614 | poly64_t __c2 = vgetq_lane_p64(__c1, __d1); \ |
| 5615 | vset_lane_p64(__c2, __a1, __b1); }) |
| 5616 | #define vcopy_laneq_u8(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5617 | uint8x8_t __a1 = (a1); uint8x16_t __c1 = (c1); \ |
| 5618 | uint8_t __c2 = vgetq_lane_u8(__c1, __d1); \ |
| 5619 | vset_lane_u8(__c2, __a1, __b1); }) |
| 5620 | #define vcopy_laneq_u16(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5621 | uint16x4_t __a1 = (a1); uint16x8_t __c1 = (c1); \ |
| 5622 | uint16_t __c2 = vgetq_lane_u16(__c1, __d1); \ |
| 5623 | vset_lane_u16(__c2, __a1, __b1); }) |
| 5624 | #define vcopy_laneq_u32(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5625 | uint32x2_t __a1 = (a1); uint32x4_t __c1 = (c1); \ |
| 5626 | uint32_t __c2 = vgetq_lane_u32(__c1, __d1); \ |
| 5627 | vset_lane_u32(__c2, __a1, __b1); }) |
| 5628 | #define vcopy_laneq_u64(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5629 | uint64x1_t __a1 = (a1); uint64x2_t __c1 = (c1); \ |
| 5630 | uint64_t __c2 = vgetq_lane_u64(__c1, __d1); \ |
| 5631 | vset_lane_u64(__c2, __a1, __b1); }) |
| 5632 | #define vcopy_laneq_f32(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5633 | float32x2_t __a1 = (a1); float32x4_t __c1 = (c1); \ |
| 5634 | float32_t __c2 = vgetq_lane_f32(__c1, __d1); \ |
| 5635 | vset_lane_f32(__c2, __a1, __b1); }) |
| 5636 | #define vcopy_laneq_f64(a1, __b1, c1, __d1) __extension__ ({ \ |
| 5637 | float64x1_t __a1 = (a1); float64x2_t __c1 = (c1); \ |
| 5638 | float64_t __c2 = vgetq_lane_f64(__c1, __d1); \ |
| 5639 | vset_lane_f64(__c2, __a1, __b1); }) |
| 5640 | |
| 5641 | __ai float64x1_t vcreate_f64(uint64_t __a) { |
| 5642 | return (float64x1_t)__a; } |
| 5643 | __ai poly64x1_t vcreate_p64(uint64_t __a) { |
| 5644 | return (poly64x1_t)__a; } |
| 5645 | |
| 5646 | #define vcvt_n_f64_s64(a, __b) __extension__ ({ \ |
| 5647 | int64x1_t __a = (a); \ |
| 5648 | (float64x1_t)__builtin_neon_vcvt_n_f64_v((int8x8_t)__a, __b, 3); }) |
| 5649 | #define vcvt_n_f64_u64(a, __b) __extension__ ({ \ |
| 5650 | uint64x1_t __a = (a); \ |
| 5651 | (float64x1_t)__builtin_neon_vcvt_n_f64_v((int8x8_t)__a, __b, 19); }) |
| 5652 | #define vcvtq_n_f64_s64(a, __b) __extension__ ({ \ |
| 5653 | int64x2_t __a = (a); \ |
| 5654 | (float64x2_t)__builtin_neon_vcvtq_n_f64_v((int8x16_t)__a, __b, 35); }) |
| 5655 | #define vcvtq_n_f64_u64(a, __b) __extension__ ({ \ |
| 5656 | uint64x2_t __a = (a); \ |
| 5657 | (float64x2_t)__builtin_neon_vcvtq_n_f64_v((int8x16_t)__a, __b, 51); }) |
| 5658 | |
| 5659 | __ai float64x1_t vdup_n_f64(float64_t __a) { |
| 5660 | return (float64x1_t){ __a }; } |
| 5661 | __ai float64x2_t vdupq_n_f64(float64_t __a) { |
| 5662 | return (float64x2_t){ __a, __a }; } |
| 5663 | __ai poly64x1_t vdup_n_p64(poly64_t __a) { |
| 5664 | return (poly64x1_t){ __a }; } |
| 5665 | __ai poly64x2_t vdupq_n_p64(poly64_t __a) { |
| 5666 | return (poly64x2_t){ __a, __a }; } |
| 5667 | |
| 5668 | __ai uint64x1_t vcage_f64(float64x1_t __a, float64x1_t __b) { |
| 5669 | return (uint64x1_t)__builtin_neon_vcage_v((int8x8_t)__a, (int8x8_t)__b, 19); } |
| 5670 | __ai uint64x2_t vcageq_f64(float64x2_t __a, float64x2_t __b) { |
| 5671 | return (uint64x2_t)__builtin_neon_vcageq_v((int8x16_t)__a, (int8x16_t)__b, 51); } |
| 5672 | |
| 5673 | __ai uint64x1_t vcagt_f64(float64x1_t __a, float64x1_t __b) { |
| 5674 | return (uint64x1_t)__builtin_neon_vcagt_v((int8x8_t)__a, (int8x8_t)__b, 19); } |
| 5675 | __ai uint64x2_t vcagtq_f64(float64x2_t __a, float64x2_t __b) { |
| 5676 | return (uint64x2_t)__builtin_neon_vcagtq_v((int8x16_t)__a, (int8x16_t)__b, 51); } |
| 5677 | |
| 5678 | __ai uint64x1_t vcale_f64(float64x1_t __a, float64x1_t __b) { |
| 5679 | return (uint64x1_t)__builtin_neon_vcale_v((int8x8_t)__a, (int8x8_t)__b, 19); } |
| 5680 | __ai uint64x2_t vcaleq_f64(float64x2_t __a, float64x2_t __b) { |
| 5681 | return (uint64x2_t)__builtin_neon_vcaleq_v((int8x16_t)__a, (int8x16_t)__b, 51); } |
| 5682 | |
| 5683 | __ai uint64x1_t vcalt_f64(float64x1_t __a, float64x1_t __b) { |
| 5684 | return (uint64x1_t)__builtin_neon_vcalt_v((int8x8_t)__a, (int8x8_t)__b, 19); } |
| 5685 | __ai uint64x2_t vcaltq_f64(float64x2_t __a, float64x2_t __b) { |
| 5686 | return (uint64x2_t)__builtin_neon_vcaltq_v((int8x16_t)__a, (int8x16_t)__b, 51); } |
| 5687 | |
| 5688 | __ai int32x2_t vcvta_s32_f32(float32x2_t __a) { |
| 5689 | return (int32x2_t)__builtin_neon_vcvta_s32_v((int8x8_t)__a, 2); } |
| 5690 | __ai int32x4_t vcvtaq_s32_f32(float32x4_t __a) { |
| 5691 | return (int32x4_t)__builtin_neon_vcvtaq_s32_v((int8x16_t)__a, 34); } |
| 5692 | |
| 5693 | __ai int64x1_t vcvta_s64_f64(float64x1_t __a) { |
| 5694 | return (int64x1_t)__builtin_neon_vcvta_s64_v((int8x8_t)__a, 3); } |
| 5695 | __ai int64x2_t vcvtaq_s64_f64(float64x2_t __a) { |
| 5696 | return (int64x2_t)__builtin_neon_vcvtaq_s64_v((int8x16_t)__a, 35); } |
| 5697 | |
| 5698 | __ai uint32x2_t vcvta_u32_f32(float32x2_t __a) { |
| 5699 | return (uint32x2_t)__builtin_neon_vcvta_u32_v((int8x8_t)__a, 18); } |
| 5700 | __ai uint32x4_t vcvtaq_u32_f32(float32x4_t __a) { |
| 5701 | return (uint32x4_t)__builtin_neon_vcvtaq_u32_v((int8x16_t)__a, 50); } |
| 5702 | |
| 5703 | __ai uint64x1_t vcvta_u64_f64(float64x1_t __a) { |
| 5704 | return (uint64x1_t)__builtin_neon_vcvta_u64_v((int8x8_t)__a, 19); } |
| 5705 | __ai uint64x2_t vcvtaq_u64_f64(float64x2_t __a) { |
| 5706 | return (uint64x2_t)__builtin_neon_vcvtaq_u64_v((int8x16_t)__a, 51); } |
| 5707 | |
| 5708 | __ai int32x2_t vcvtm_s32_f32(float32x2_t __a) { |
| 5709 | return (int32x2_t)__builtin_neon_vcvtm_s32_v((int8x8_t)__a, 2); } |
| 5710 | __ai int32x4_t vcvtmq_s32_f32(float32x4_t __a) { |
| 5711 | return (int32x4_t)__builtin_neon_vcvtmq_s32_v((int8x16_t)__a, 34); } |
| 5712 | |
| 5713 | __ai int64x1_t vcvtm_s64_f64(float64x1_t __a) { |
| 5714 | return (int64x1_t)__builtin_neon_vcvtm_s64_v((int8x8_t)__a, 3); } |
| 5715 | __ai int64x2_t vcvtmq_s64_f64(float64x2_t __a) { |
| 5716 | return (int64x2_t)__builtin_neon_vcvtmq_s64_v((int8x16_t)__a, 35); } |
| 5717 | |
| 5718 | __ai uint32x2_t vcvtm_u32_f32(float32x2_t __a) { |
| 5719 | return (uint32x2_t)__builtin_neon_vcvtm_u32_v((int8x8_t)__a, 18); } |
| 5720 | __ai uint32x4_t vcvtmq_u32_f32(float32x4_t __a) { |
| 5721 | return (uint32x4_t)__builtin_neon_vcvtmq_u32_v((int8x16_t)__a, 50); } |
| 5722 | |
| 5723 | __ai uint64x1_t vcvtm_u64_f64(float64x1_t __a) { |
| 5724 | return (uint64x1_t)__builtin_neon_vcvtm_u64_v((int8x8_t)__a, 19); } |
| 5725 | __ai uint64x2_t vcvtmq_u64_f64(float64x2_t __a) { |
| 5726 | return (uint64x2_t)__builtin_neon_vcvtmq_u64_v((int8x16_t)__a, 51); } |
| 5727 | |
| 5728 | __ai int32x2_t vcvtn_s32_f32(float32x2_t __a) { |
| 5729 | return (int32x2_t)__builtin_neon_vcvtn_s32_v((int8x8_t)__a, 2); } |
| 5730 | __ai int32x4_t vcvtnq_s32_f32(float32x4_t __a) { |
| 5731 | return (int32x4_t)__builtin_neon_vcvtnq_s32_v((int8x16_t)__a, 34); } |
| 5732 | |
| 5733 | __ai int64x1_t vcvtn_s64_f64(float64x1_t __a) { |
| 5734 | return (int64x1_t)__builtin_neon_vcvtn_s64_v((int8x8_t)__a, 3); } |
| 5735 | __ai int64x2_t vcvtnq_s64_f64(float64x2_t __a) { |
| 5736 | return (int64x2_t)__builtin_neon_vcvtnq_s64_v((int8x16_t)__a, 35); } |
| 5737 | |
| 5738 | __ai uint32x2_t vcvtn_u32_f32(float32x2_t __a) { |
| 5739 | return (uint32x2_t)__builtin_neon_vcvtn_u32_v((int8x8_t)__a, 18); } |
| 5740 | __ai uint32x4_t vcvtnq_u32_f32(float32x4_t __a) { |
| 5741 | return (uint32x4_t)__builtin_neon_vcvtnq_u32_v((int8x16_t)__a, 50); } |
| 5742 | |
| 5743 | __ai uint64x1_t vcvtn_u64_f64(float64x1_t __a) { |
| 5744 | return (uint64x1_t)__builtin_neon_vcvtn_u64_v((int8x8_t)__a, 19); } |
| 5745 | __ai uint64x2_t vcvtnq_u64_f64(float64x2_t __a) { |
| 5746 | return (uint64x2_t)__builtin_neon_vcvtnq_u64_v((int8x16_t)__a, 51); } |
| 5747 | |
| 5748 | __ai int32x2_t vcvtp_s32_f32(float32x2_t __a) { |
| 5749 | return (int32x2_t)__builtin_neon_vcvtp_s32_v((int8x8_t)__a, 2); } |
| 5750 | __ai int32x4_t vcvtpq_s32_f32(float32x4_t __a) { |
| 5751 | return (int32x4_t)__builtin_neon_vcvtpq_s32_v((int8x16_t)__a, 34); } |
| 5752 | |
| 5753 | __ai int64x1_t vcvtp_s64_f64(float64x1_t __a) { |
| 5754 | return (int64x1_t)__builtin_neon_vcvtp_s64_v((int8x8_t)__a, 3); } |
| 5755 | __ai int64x2_t vcvtpq_s64_f64(float64x2_t __a) { |
| 5756 | return (int64x2_t)__builtin_neon_vcvtpq_s64_v((int8x16_t)__a, 35); } |
| 5757 | |
| 5758 | __ai uint32x2_t vcvtp_u32_f32(float32x2_t __a) { |
| 5759 | return (uint32x2_t)__builtin_neon_vcvtp_u32_v((int8x8_t)__a, 18); } |
| 5760 | __ai uint32x4_t vcvtpq_u32_f32(float32x4_t __a) { |
| 5761 | return (uint32x4_t)__builtin_neon_vcvtpq_u32_v((int8x16_t)__a, 50); } |
| 5762 | |
| 5763 | __ai uint64x1_t vcvtp_u64_f64(float64x1_t __a) { |
| 5764 | return (uint64x1_t)__builtin_neon_vcvtp_u64_v((int8x8_t)__a, 19); } |
| 5765 | __ai uint64x2_t vcvtpq_u64_f64(float64x2_t __a) { |
| 5766 | return (uint64x2_t)__builtin_neon_vcvtpq_u64_v((int8x16_t)__a, 51); } |
| 5767 | |
| 5768 | #define vcvt_n_s64_f64(a, __b) __extension__ ({ \ |
| 5769 | float64x1_t __a = (a); \ |
| 5770 | (int64x1_t)__builtin_neon_vcvt_n_s64_v((int8x8_t)__a, __b, 3); }) |
| 5771 | #define vcvtq_n_s64_f64(a, __b) __extension__ ({ \ |
| 5772 | float64x2_t __a = (a); \ |
| 5773 | (int64x2_t)__builtin_neon_vcvtq_n_s64_v((int8x16_t)__a, __b, 35); }) |
| 5774 | |
| 5775 | #define vcvt_n_u64_f64(a, __b) __extension__ ({ \ |
| 5776 | float64x1_t __a = (a); \ |
| 5777 | (uint64x1_t)__builtin_neon_vcvt_n_u64_v((int8x8_t)__a, __b, 19); }) |
| 5778 | #define vcvtq_n_u64_f64(a, __b) __extension__ ({ \ |
| 5779 | float64x2_t __a = (a); \ |
| 5780 | (uint64x2_t)__builtin_neon_vcvtq_n_u64_v((int8x16_t)__a, __b, 51); }) |
| 5781 | |
| 5782 | __ai float32x2_t vdiv_f32(float32x2_t __a, float32x2_t __b) { |
| 5783 | return __a / __b; } |
| 5784 | __ai float64x1_t vdiv_f64(float64x1_t __a, float64x1_t __b) { |
| 5785 | return __a / __b; } |
| 5786 | __ai float32x4_t vdivq_f32(float32x4_t __a, float32x4_t __b) { |
| 5787 | return __a / __b; } |
| 5788 | __ai float64x2_t vdivq_f64(float64x2_t __a, float64x2_t __b) { |
| 5789 | return __a / __b; } |
| 5790 | |
| 5791 | __ai float32x2_t vmaxnm_f32(float32x2_t __a, float32x2_t __b) { |
| 5792 | return (float32x2_t)__builtin_neon_vmaxnm_v((int8x8_t)__a, (int8x8_t)__b, 8); } |
| 5793 | __ai float64x1_t vmaxnm_f64(float64x1_t __a, float64x1_t __b) { |
| 5794 | return (float64x1_t)__builtin_neon_vmaxnm_v((int8x8_t)__a, (int8x8_t)__b, 9); } |
| 5795 | __ai float32x4_t vmaxnmq_f32(float32x4_t __a, float32x4_t __b) { |
| 5796 | return (float32x4_t)__builtin_neon_vmaxnmq_v((int8x16_t)__a, (int8x16_t)__b, 40); } |
| 5797 | __ai float64x2_t vmaxnmq_f64(float64x2_t __a, float64x2_t __b) { |
| 5798 | return (float64x2_t)__builtin_neon_vmaxnmq_v((int8x16_t)__a, (int8x16_t)__b, 41); } |
| 5799 | |
| 5800 | __ai float32x2_t vpmaxnm_f32(float32x2_t __a, float32x2_t __b) { |
| 5801 | return (float32x2_t)__builtin_neon_vpmaxnm_v((int8x8_t)__a, (int8x8_t)__b, 8); } |
| 5802 | __ai float32x4_t vpmaxnmq_f32(float32x4_t __a, float32x4_t __b) { |
| 5803 | return (float32x4_t)__builtin_neon_vpmaxnmq_v((int8x16_t)__a, (int8x16_t)__b, 40); } |
| 5804 | __ai float64x2_t vpmaxnmq_f64(float64x2_t __a, float64x2_t __b) { |
| 5805 | return (float64x2_t)__builtin_neon_vpmaxnmq_v((int8x16_t)__a, (int8x16_t)__b, 41); } |
| 5806 | |
| 5807 | __ai float32_t vmaxnmv_f32(float32x2_t __a) { |
| 5808 | return (float32_t)__builtin_neon_vmaxnmv_f32(__a); } |
| 5809 | __ai float32_t vmaxnmvq_f32(float32x4_t __a) { |
| 5810 | return (float32_t)__builtin_neon_vmaxnmvq_f32(__a); } |
| 5811 | __ai float64_t vmaxnmvq_f64(float64x2_t __a) { |
| 5812 | return (float64_t)__builtin_neon_vmaxnmvq_f64(__a); } |
| 5813 | |
| 5814 | __ai float32x2_t vminnm_f32(float32x2_t __a, float32x2_t __b) { |
| 5815 | return (float32x2_t)__builtin_neon_vminnm_v((int8x8_t)__a, (int8x8_t)__b, 8); } |
| 5816 | __ai float64x1_t vminnm_f64(float64x1_t __a, float64x1_t __b) { |
| 5817 | return (float64x1_t)__builtin_neon_vminnm_v((int8x8_t)__a, (int8x8_t)__b, 9); } |
| 5818 | __ai float32x4_t vminnmq_f32(float32x4_t __a, float32x4_t __b) { |
| 5819 | return (float32x4_t)__builtin_neon_vminnmq_v((int8x16_t)__a, (int8x16_t)__b, 40); } |
| 5820 | __ai float64x2_t vminnmq_f64(float64x2_t __a, float64x2_t __b) { |
| 5821 | return (float64x2_t)__builtin_neon_vminnmq_v((int8x16_t)__a, (int8x16_t)__b, 41); } |
| 5822 | |
| 5823 | __ai float32x2_t vpminnm_f32(float32x2_t __a, float32x2_t __b) { |
| 5824 | return (float32x2_t)__builtin_neon_vpminnm_v((int8x8_t)__a, (int8x8_t)__b, 8); } |
| 5825 | __ai float32x4_t vpminnmq_f32(float32x4_t __a, float32x4_t __b) { |
| 5826 | return (float32x4_t)__builtin_neon_vpminnmq_v((int8x16_t)__a, (int8x16_t)__b, 40); } |
| 5827 | __ai float64x2_t vpminnmq_f64(float64x2_t __a, float64x2_t __b) { |
| 5828 | return (float64x2_t)__builtin_neon_vpminnmq_v((int8x16_t)__a, (int8x16_t)__b, 41); } |
| 5829 | |
| 5830 | __ai float32_t vminnmv_f32(float32x2_t __a) { |
| 5831 | return (float32_t)__builtin_neon_vminnmv_f32(__a); } |
| 5832 | __ai float32_t vminnmvq_f32(float32x4_t __a) { |
| 5833 | return (float32_t)__builtin_neon_vminnmvq_f32(__a); } |
| 5834 | __ai float64_t vminnmvq_f64(float64x2_t __a) { |
| 5835 | return (float64_t)__builtin_neon_vminnmvq_f64(__a); } |
| 5836 | |
| 5837 | __ai float64x1_t vfma_f64(float64x1_t __a, float64x1_t __b, float64x1_t __c) { |
| 5838 | return (float64x1_t)__builtin_neon_vfma_v((int8x8_t)__a, (int8x8_t)__b, (int8x8_t)__c, 9); } |
| 5839 | __ai float64x2_t vfmaq_f64(float64x2_t __a, float64x2_t __b, float64x2_t __c) { |
| 5840 | return (float64x2_t)__builtin_neon_vfmaq_v((int8x16_t)__a, (int8x16_t)__b, (int8x16_t)__c, 41); } |
| 5841 | |
| 5842 | __ai float32x2_t vfma_n_f32(float32x2_t __a, float32x2_t __b, float32_t __c) { |
| 5843 | return vfma_f32(__a, __b, (float32x2_t){ __c, __c }); } |
| 5844 | __ai float32x4_t vfmaq_n_f32(float32x4_t __a, float32x4_t __b, float32_t __c) { |
| 5845 | return vfmaq_f32(__a, __b, (float32x4_t){ __c, __c, __c, __c }); } |
| 5846 | |
| 5847 | __ai float32x2_t vfms_f32(float32x2_t __a, float32x2_t __b, float32x2_t __c) { |
| 5848 | return (float32x2_t)__builtin_neon_vfms_v((int8x8_t)__a, (int8x8_t)__b, (int8x8_t)__c, 8); } |
| 5849 | __ai float64x1_t vfms_f64(float64x1_t __a, float64x1_t __b, float64x1_t __c) { |
| 5850 | return (float64x1_t)__builtin_neon_vfms_v((int8x8_t)__a, (int8x8_t)__b, (int8x8_t)__c, 9); } |
| 5851 | __ai float32x4_t vfmsq_f32(float32x4_t __a, float32x4_t __b, float32x4_t __c) { |
| 5852 | return (float32x4_t)__builtin_neon_vfmsq_v((int8x16_t)__a, (int8x16_t)__b, (int8x16_t)__c, 40); } |
| 5853 | __ai float64x2_t vfmsq_f64(float64x2_t __a, float64x2_t __b, float64x2_t __c) { |
| 5854 | return (float64x2_t)__builtin_neon_vfmsq_v((int8x16_t)__a, (int8x16_t)__b, (int8x16_t)__c, 41); } |
| 5855 | |
| 5856 | __ai float32x2_t vfms_n_f32(float32x2_t __a, float32x2_t __b, float32_t __c) { |
| 5857 | return vfms_f32(__a, __b, (float32x2_t){ __c, __c }); } |
| 5858 | __ai float32x4_t vfmsq_n_f32(float32x4_t __a, float32x4_t __b, float32_t __c) { |
| 5859 | return vfmsq_f32(__a, __b, (float32x4_t){ __c, __c, __c, __c }); } |
| 5860 | |
| 5861 | __ai float64x1_t vrecpe_f64(float64x1_t __a) { |
| 5862 | return (float64x1_t)__builtin_neon_vrecpe_v((int8x8_t)__a, 9); } |
| 5863 | __ai float64x2_t vrecpeq_f64(float64x2_t __a) { |
| 5864 | return (float64x2_t)__builtin_neon_vrecpeq_v((int8x16_t)__a, 41); } |
| 5865 | |
| 5866 | __ai float64x1_t vrecps_f64(float64x1_t __a, float64x1_t __b) { |
| 5867 | return (float64x1_t)__builtin_neon_vrecps_v((int8x8_t)__a, (int8x8_t)__b, 9); } |
| 5868 | __ai float64x2_t vrecpsq_f64(float64x2_t __a, float64x2_t __b) { |
| 5869 | return (float64x2_t)__builtin_neon_vrecpsq_v((int8x16_t)__a, (int8x16_t)__b, 41); } |
| 5870 | |
| 5871 | __ai float32x2_t vrnda_f32(float32x2_t __a) { |
| 5872 | return (float32x2_t)__builtin_neon_vrnda_v((int8x8_t)__a, 8); } |
| 5873 | __ai float64x1_t vrnda_f64(float64x1_t __a) { |
| 5874 | return (float64x1_t)__builtin_neon_vrnda_v((int8x8_t)__a, 9); } |
| 5875 | __ai float32x4_t vrndaq_f32(float32x4_t __a) { |
| 5876 | return (float32x4_t)__builtin_neon_vrndaq_v((int8x16_t)__a, 40); } |
| 5877 | __ai float64x2_t vrndaq_f64(float64x2_t __a) { |
| 5878 | return (float64x2_t)__builtin_neon_vrndaq_v((int8x16_t)__a, 41); } |
| 5879 | |
| 5880 | __ai float32x2_t vrndi_f32(float32x2_t __a) { |
| 5881 | return (float32x2_t)__builtin_neon_vrndi_v((int8x8_t)__a, 8); } |
| 5882 | __ai float64x1_t vrndi_f64(float64x1_t __a) { |
| 5883 | return (float64x1_t)__builtin_neon_vrndi_v((int8x8_t)__a, 9); } |
| 5884 | __ai float32x4_t vrndiq_f32(float32x4_t __a) { |
| 5885 | return (float32x4_t)__builtin_neon_vrndiq_v((int8x16_t)__a, 40); } |
| 5886 | __ai float64x2_t vrndiq_f64(float64x2_t __a) { |
| 5887 | return (float64x2_t)__builtin_neon_vrndiq_v((int8x16_t)__a, 41); } |
| 5888 | |
| 5889 | __ai float32x2_t vrndm_f32(float32x2_t __a) { |
| 5890 | return (float32x2_t)__builtin_neon_vrndm_v((int8x8_t)__a, 8); } |
| 5891 | __ai float64x1_t vrndm_f64(float64x1_t __a) { |
| 5892 | return (float64x1_t)__builtin_neon_vrndm_v((int8x8_t)__a, 9); } |
| 5893 | __ai float32x4_t vrndmq_f32(float32x4_t __a) { |
| 5894 | return (float32x4_t)__builtin_neon_vrndmq_v((int8x16_t)__a, 40); } |
| 5895 | __ai float64x2_t vrndmq_f64(float64x2_t __a) { |
| 5896 | return (float64x2_t)__builtin_neon_vrndmq_v((int8x16_t)__a, 41); } |
| 5897 | |
| 5898 | __ai float32x2_t vrndn_f32(float32x2_t __a) { |
| 5899 | return (float32x2_t)__builtin_neon_vrndn_v((int8x8_t)__a, 8); } |
| 5900 | __ai float64x1_t vrndn_f64(float64x1_t __a) { |
| 5901 | return (float64x1_t)__builtin_neon_vrndn_v((int8x8_t)__a, 9); } |
| 5902 | __ai float32x4_t vrndnq_f32(float32x4_t __a) { |
| 5903 | return (float32x4_t)__builtin_neon_vrndnq_v((int8x16_t)__a, 40); } |
| 5904 | __ai float64x2_t vrndnq_f64(float64x2_t __a) { |
| 5905 | return (float64x2_t)__builtin_neon_vrndnq_v((int8x16_t)__a, 41); } |
| 5906 | |
| 5907 | __ai float32x2_t vrndp_f32(float32x2_t __a) { |
| 5908 | return (float32x2_t)__builtin_neon_vrndp_v((int8x8_t)__a, 8); } |
| 5909 | __ai float64x1_t vrndp_f64(float64x1_t __a) { |
| 5910 | return (float64x1_t)__builtin_neon_vrndp_v((int8x8_t)__a, 9); } |
| 5911 | __ai float32x4_t vrndpq_f32(float32x4_t __a) { |
| 5912 | return (float32x4_t)__builtin_neon_vrndpq_v((int8x16_t)__a, 40); } |
| 5913 | __ai float64x2_t vrndpq_f64(float64x2_t __a) { |
| 5914 | return (float64x2_t)__builtin_neon_vrndpq_v((int8x16_t)__a, 41); } |
| 5915 | |
| 5916 | __ai float32x2_t vrndx_f32(float32x2_t __a) { |
| 5917 | return (float32x2_t)__builtin_neon_vrndx_v((int8x8_t)__a, 8); } |
| 5918 | __ai float64x1_t vrndx_f64(float64x1_t __a) { |
| 5919 | return (float64x1_t)__builtin_neon_vrndx_v((int8x8_t)__a, 9); } |
| 5920 | __ai float32x4_t vrndxq_f32(float32x4_t __a) { |
| 5921 | return (float32x4_t)__builtin_neon_vrndxq_v((int8x16_t)__a, 40); } |
| 5922 | __ai float64x2_t vrndxq_f64(float64x2_t __a) { |
| 5923 | return (float64x2_t)__builtin_neon_vrndxq_v((int8x16_t)__a, 41); } |
| 5924 | |
| 5925 | __ai float32x2_t vrnd_f32(float32x2_t __a) { |
| 5926 | return (float32x2_t)__builtin_neon_vrnd_v((int8x8_t)__a, 8); } |
| 5927 | __ai float64x1_t vrnd_f64(float64x1_t __a) { |
| 5928 | return (float64x1_t)__builtin_neon_vrnd_v((int8x8_t)__a, 9); } |
| 5929 | __ai float32x4_t vrndq_f32(float32x4_t __a) { |
| 5930 | return (float32x4_t)__builtin_neon_vrndq_v((int8x16_t)__a, 40); } |
| 5931 | __ai float64x2_t vrndq_f64(float64x2_t __a) { |
| 5932 | return (float64x2_t)__builtin_neon_vrndq_v((int8x16_t)__a, 41); } |
| 5933 | |
| 5934 | __ai float64x1_t vrsqrte_f64(float64x1_t __a) { |
| 5935 | return (float64x1_t)__builtin_neon_vrsqrte_v((int8x8_t)__a, 9); } |
| 5936 | __ai float64x2_t vrsqrteq_f64(float64x2_t __a) { |
| 5937 | return (float64x2_t)__builtin_neon_vrsqrteq_v((int8x16_t)__a, 41); } |
| 5938 | |
| 5939 | __ai float64x1_t vrsqrts_f64(float64x1_t __a, float64x1_t __b) { |
| 5940 | return (float64x1_t)__builtin_neon_vrsqrts_v((int8x8_t)__a, (int8x8_t)__b, 9); } |
| 5941 | __ai float64x2_t vrsqrtsq_f64(float64x2_t __a, float64x2_t __b) { |
| 5942 | return (float64x2_t)__builtin_neon_vrsqrtsq_v((int8x16_t)__a, (int8x16_t)__b, 41); } |
| 5943 | |
| 5944 | __ai float32x2_t vsqrt_f32(float32x2_t __a) { |
| 5945 | return (float32x2_t)__builtin_neon_vsqrt_v((int8x8_t)__a, 8); } |
| 5946 | __ai float64x1_t vsqrt_f64(float64x1_t __a) { |
| 5947 | return (float64x1_t)__builtin_neon_vsqrt_v((int8x8_t)__a, 9); } |
| 5948 | __ai float32x4_t vsqrtq_f32(float32x4_t __a) { |
| 5949 | return (float32x4_t)__builtin_neon_vsqrtq_v((int8x16_t)__a, 40); } |
| 5950 | __ai float64x2_t vsqrtq_f64(float64x2_t __a) { |
| 5951 | return (float64x2_t)__builtin_neon_vsqrtq_v((int8x16_t)__a, 41); } |
| 5952 | |
| 5953 | #define vget_lane_f64(a, __b) __extension__ ({ \ |
| 5954 | float64x1_t __a = (a); \ |
| 5955 | (float64_t)__builtin_neon_vget_lane_f64(__a, __b); }) |
| 5956 | #define vgetq_lane_f64(a, __b) __extension__ ({ \ |
| 5957 | float64x2_t __a = (a); \ |
| 5958 | (float64_t)__builtin_neon_vgetq_lane_f64(__a, __b); }) |
| 5959 | #define vget_lane_p64(a, __b) __extension__ ({ \ |
| 5960 | poly64x1_t __a = (a); \ |
| 5961 | (poly64_t)__builtin_neon_vget_lane_i64((int64x1_t)__a, __b); }) |
| 5962 | #define vgetq_lane_p64(a, __b) __extension__ ({ \ |
| 5963 | poly64x2_t __a = (a); \ |
| 5964 | (poly64_t)__builtin_neon_vgetq_lane_i64((int64x2_t)__a, __b); }) |
| 5965 | |
| 5966 | #define vld1q_f64(__a) __extension__ ({ \ |
| 5967 | (float64x2_t)__builtin_neon_vld1q_v(__a, 41); }) |
| 5968 | #define vld1_f64(__a) __extension__ ({ \ |
| 5969 | (float64x1_t)__builtin_neon_vld1_v(__a, 9); }) |
| 5970 | #define vld1_p64(__a) __extension__ ({ \ |
| 5971 | (poly64x1_t)__builtin_neon_vld1_v(__a, 6); }) |
| 5972 | #define vld1q_p64(__a) __extension__ ({ \ |
| 5973 | (poly64x2_t)__builtin_neon_vld1q_v(__a, 38); }) |
| 5974 | |
| 5975 | #define vld1q_dup_f64(__a) __extension__ ({ \ |
| 5976 | (float64x2_t)__builtin_neon_vld1q_dup_v(__a, 41); }) |
| 5977 | #define vld1q_dup_p64(__a) __extension__ ({ \ |
| 5978 | (poly64x2_t)__builtin_neon_vld1q_dup_v(__a, 38); }) |
| 5979 | #define vld1_dup_f64(__a) __extension__ ({ \ |
| 5980 | (float64x1_t)__builtin_neon_vld1_dup_v(__a, 9); }) |
| 5981 | #define vld1_dup_p64(__a) __extension__ ({ \ |
| 5982 | (poly64x1_t)__builtin_neon_vld1_dup_v(__a, 6); }) |
| 5983 | |
| 5984 | #define vld1q_lane_f64(__a, b, __c) __extension__ ({ \ |
| 5985 | float64x2_t __b = (b); \ |
| 5986 | (float64x2_t)__builtin_neon_vld1q_lane_v(__a, (int8x16_t)__b, __c, 41); }) |
| 5987 | #define vld1q_lane_p64(__a, b, __c) __extension__ ({ \ |
| 5988 | poly64x2_t __b = (b); \ |
| 5989 | (poly64x2_t)__builtin_neon_vld1q_lane_v(__a, (int8x16_t)__b, __c, 38); }) |
| 5990 | #define vld1_lane_f64(__a, b, __c) __extension__ ({ \ |
| 5991 | float64x1_t __b = (b); \ |
| 5992 | (float64x1_t)__builtin_neon_vld1_lane_v(__a, (int8x8_t)__b, __c, 9); }) |
| 5993 | #define vld1_lane_p64(__a, b, __c) __extension__ ({ \ |
| 5994 | poly64x1_t __b = (b); \ |
| 5995 | (poly64x1_t)__builtin_neon_vld1_lane_v(__a, (int8x8_t)__b, __c, 6); }) |
| 5996 | |
| 5997 | #define vld1q_u8_x2(__a) __extension__ ({ \ |
| 5998 | uint8x16x2_t r; __builtin_neon_vld1q_x2_v(&r, __a, 48); r; }) |
| 5999 | #define vld1q_u16_x2(__a) __extension__ ({ \ |
| 6000 | uint16x8x2_t r; __builtin_neon_vld1q_x2_v(&r, __a, 49); r; }) |
| 6001 | #define vld1q_u32_x2(__a) __extension__ ({ \ |
| 6002 | uint32x4x2_t r; __builtin_neon_vld1q_x2_v(&r, __a, 50); r; }) |
| 6003 | #define vld1q_u64_x2(__a) __extension__ ({ \ |
| 6004 | uint64x2x2_t r; __builtin_neon_vld1q_x2_v(&r, __a, 51); r; }) |
| 6005 | #define vld1q_s8_x2(__a) __extension__ ({ \ |
| 6006 | int8x16x2_t r; __builtin_neon_vld1q_x2_v(&r, __a, 32); r; }) |
| 6007 | #define vld1q_s16_x2(__a) __extension__ ({ \ |
| 6008 | int16x8x2_t r; __builtin_neon_vld1q_x2_v(&r, __a, 33); r; }) |
| 6009 | #define vld1q_s32_x2(__a) __extension__ ({ \ |
| 6010 | int32x4x2_t r; __builtin_neon_vld1q_x2_v(&r, __a, 34); r; }) |
| 6011 | #define vld1q_s64_x2(__a) __extension__ ({ \ |
| 6012 | int64x2x2_t r; __builtin_neon_vld1q_x2_v(&r, __a, 35); r; }) |
| 6013 | #define vld1q_f16_x2(__a) __extension__ ({ \ |
| 6014 | float16x8x2_t r; __builtin_neon_vld1q_x2_v(&r, __a, 39); r; }) |
| 6015 | #define vld1q_f32_x2(__a) __extension__ ({ \ |
| 6016 | float32x4x2_t r; __builtin_neon_vld1q_x2_v(&r, __a, 40); r; }) |
| 6017 | #define vld1q_f64_x2(__a) __extension__ ({ \ |
| 6018 | float64x2x2_t r; __builtin_neon_vld1q_x2_v(&r, __a, 41); r; }) |
| 6019 | #define vld1q_p8_x2(__a) __extension__ ({ \ |
| 6020 | poly8x16x2_t r; __builtin_neon_vld1q_x2_v(&r, __a, 36); r; }) |
| 6021 | #define vld1q_p16_x2(__a) __extension__ ({ \ |
| 6022 | poly16x8x2_t r; __builtin_neon_vld1q_x2_v(&r, __a, 37); r; }) |
| 6023 | #define vld1q_p64_x2(__a) __extension__ ({ \ |
| 6024 | poly64x2x2_t r; __builtin_neon_vld1q_x2_v(&r, __a, 38); r; }) |
| 6025 | #define vld1_u8_x2(__a) __extension__ ({ \ |
| 6026 | uint8x8x2_t r; __builtin_neon_vld1_x2_v(&r, __a, 16); r; }) |
| 6027 | #define vld1_u16_x2(__a) __extension__ ({ \ |
| 6028 | uint16x4x2_t r; __builtin_neon_vld1_x2_v(&r, __a, 17); r; }) |
| 6029 | #define vld1_u32_x2(__a) __extension__ ({ \ |
| 6030 | uint32x2x2_t r; __builtin_neon_vld1_x2_v(&r, __a, 18); r; }) |
| 6031 | #define vld1_u64_x2(__a) __extension__ ({ \ |
| 6032 | uint64x1x2_t r; __builtin_neon_vld1_x2_v(&r, __a, 19); r; }) |
| 6033 | #define vld1_s8_x2(__a) __extension__ ({ \ |
| 6034 | int8x8x2_t r; __builtin_neon_vld1_x2_v(&r, __a, 0); r; }) |
| 6035 | #define vld1_s16_x2(__a) __extension__ ({ \ |
| 6036 | int16x4x2_t r; __builtin_neon_vld1_x2_v(&r, __a, 1); r; }) |
| 6037 | #define vld1_s32_x2(__a) __extension__ ({ \ |
| 6038 | int32x2x2_t r; __builtin_neon_vld1_x2_v(&r, __a, 2); r; }) |
| 6039 | #define vld1_s64_x2(__a) __extension__ ({ \ |
| 6040 | int64x1x2_t r; __builtin_neon_vld1_x2_v(&r, __a, 3); r; }) |
| 6041 | #define vld1_f16_x2(__a) __extension__ ({ \ |
| 6042 | float16x4x2_t r; __builtin_neon_vld1_x2_v(&r, __a, 7); r; }) |
| 6043 | #define vld1_f32_x2(__a) __extension__ ({ \ |
| 6044 | float32x2x2_t r; __builtin_neon_vld1_x2_v(&r, __a, 8); r; }) |
| 6045 | #define vld1_f64_x2(__a) __extension__ ({ \ |
| 6046 | float64x1x2_t r; __builtin_neon_vld1_x2_v(&r, __a, 9); r; }) |
| 6047 | #define vld1_p8_x2(__a) __extension__ ({ \ |
| 6048 | poly8x8x2_t r; __builtin_neon_vld1_x2_v(&r, __a, 4); r; }) |
| 6049 | #define vld1_p16_x2(__a) __extension__ ({ \ |
| 6050 | poly16x4x2_t r; __builtin_neon_vld1_x2_v(&r, __a, 5); r; }) |
| 6051 | #define vld1_p64_x2(__a) __extension__ ({ \ |
| 6052 | poly64x1x2_t r; __builtin_neon_vld1_x2_v(&r, __a, 6); r; }) |
| 6053 | |
| 6054 | #define vld2q_u64(__a) __extension__ ({ \ |
| 6055 | uint64x2x2_t r; __builtin_neon_vld2q_v(&r, __a, 51); r; }) |
| 6056 | #define vld2q_s64(__a) __extension__ ({ \ |
| 6057 | int64x2x2_t r; __builtin_neon_vld2q_v(&r, __a, 35); r; }) |
| 6058 | #define vld2q_f64(__a) __extension__ ({ \ |
| 6059 | float64x2x2_t r; __builtin_neon_vld2q_v(&r, __a, 41); r; }) |
| 6060 | #define vld2_f64(__a) __extension__ ({ \ |
| 6061 | float64x1x2_t r; __builtin_neon_vld2_v(&r, __a, 9); r; }) |
| 6062 | #define vld2_p64(__a) __extension__ ({ \ |
| 6063 | poly64x1x2_t r; __builtin_neon_vld2_v(&r, __a, 6); r; }) |
| 6064 | #define vld2q_p64(__a) __extension__ ({ \ |
| 6065 | poly64x2x2_t r; __builtin_neon_vld2q_v(&r, __a, 38); r; }) |
| 6066 | |
| 6067 | #define vld2q_dup_u8(__a) __extension__ ({ \ |
| 6068 | uint8x16x2_t r; __builtin_neon_vld2q_dup_v(&r, __a, 48); r; }) |
| 6069 | #define vld2q_dup_u16(__a) __extension__ ({ \ |
| 6070 | uint16x8x2_t r; __builtin_neon_vld2q_dup_v(&r, __a, 49); r; }) |
| 6071 | #define vld2q_dup_u32(__a) __extension__ ({ \ |
| 6072 | uint32x4x2_t r; __builtin_neon_vld2q_dup_v(&r, __a, 50); r; }) |
| 6073 | #define vld2q_dup_u64(__a) __extension__ ({ \ |
| 6074 | uint64x2x2_t r; __builtin_neon_vld2q_dup_v(&r, __a, 51); r; }) |
| 6075 | #define vld2q_dup_s8(__a) __extension__ ({ \ |
| 6076 | int8x16x2_t r; __builtin_neon_vld2q_dup_v(&r, __a, 32); r; }) |
| 6077 | #define vld2q_dup_s16(__a) __extension__ ({ \ |
| 6078 | int16x8x2_t r; __builtin_neon_vld2q_dup_v(&r, __a, 33); r; }) |
| 6079 | #define vld2q_dup_s32(__a) __extension__ ({ \ |
| 6080 | int32x4x2_t r; __builtin_neon_vld2q_dup_v(&r, __a, 34); r; }) |
| 6081 | #define vld2q_dup_s64(__a) __extension__ ({ \ |
| 6082 | int64x2x2_t r; __builtin_neon_vld2q_dup_v(&r, __a, 35); r; }) |
| 6083 | #define vld2q_dup_f16(__a) __extension__ ({ \ |
| 6084 | float16x8x2_t r; __builtin_neon_vld2q_dup_v(&r, __a, 39); r; }) |
| 6085 | #define vld2q_dup_f32(__a) __extension__ ({ \ |
| 6086 | float32x4x2_t r; __builtin_neon_vld2q_dup_v(&r, __a, 40); r; }) |
| 6087 | #define vld2q_dup_f64(__a) __extension__ ({ \ |
| 6088 | float64x2x2_t r; __builtin_neon_vld2q_dup_v(&r, __a, 41); r; }) |
| 6089 | #define vld2q_dup_p8(__a) __extension__ ({ \ |
| 6090 | poly8x16x2_t r; __builtin_neon_vld2q_dup_v(&r, __a, 36); r; }) |
| 6091 | #define vld2q_dup_p16(__a) __extension__ ({ \ |
| 6092 | poly16x8x2_t r; __builtin_neon_vld2q_dup_v(&r, __a, 37); r; }) |
| 6093 | #define vld2q_dup_p64(__a) __extension__ ({ \ |
| 6094 | poly64x2x2_t r; __builtin_neon_vld2q_dup_v(&r, __a, 38); r; }) |
| 6095 | #define vld2_dup_f64(__a) __extension__ ({ \ |
| 6096 | float64x1x2_t r; __builtin_neon_vld2_dup_v(&r, __a, 9); r; }) |
| 6097 | #define vld2_dup_p64(__a) __extension__ ({ \ |
| 6098 | poly64x1x2_t r; __builtin_neon_vld2_dup_v(&r, __a, 6); r; }) |
| 6099 | |
| 6100 | #define vld2q_lane_u8(__a, b, __c) __extension__ ({ \ |
| 6101 | uint8x16x2_t __b = (b); \ |
| 6102 | uint8x16x2_t r; __builtin_neon_vld2q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], __c, 48); r; }) |
| 6103 | #define vld2q_lane_u64(__a, b, __c) __extension__ ({ \ |
| 6104 | uint64x2x2_t __b = (b); \ |
| 6105 | uint64x2x2_t r; __builtin_neon_vld2q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], __c, 51); r; }) |
| 6106 | #define vld2q_lane_s8(__a, b, __c) __extension__ ({ \ |
| 6107 | int8x16x2_t __b = (b); \ |
| 6108 | int8x16x2_t r; __builtin_neon_vld2q_lane_v(&r, __a, __b.val[0], __b.val[1], __c, 32); r; }) |
| 6109 | #define vld2q_lane_s64(__a, b, __c) __extension__ ({ \ |
| 6110 | int64x2x2_t __b = (b); \ |
| 6111 | int64x2x2_t r; __builtin_neon_vld2q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], __c, 35); r; }) |
| 6112 | #define vld2q_lane_f64(__a, b, __c) __extension__ ({ \ |
| 6113 | float64x2x2_t __b = (b); \ |
| 6114 | float64x2x2_t r; __builtin_neon_vld2q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], __c, 41); r; }) |
| 6115 | #define vld2q_lane_p8(__a, b, __c) __extension__ ({ \ |
| 6116 | poly8x16x2_t __b = (b); \ |
| 6117 | poly8x16x2_t r; __builtin_neon_vld2q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], __c, 36); r; }) |
| 6118 | #define vld2q_lane_p64(__a, b, __c) __extension__ ({ \ |
| 6119 | poly64x2x2_t __b = (b); \ |
| 6120 | poly64x2x2_t r; __builtin_neon_vld2q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], __c, 38); r; }) |
| 6121 | #define vld2_lane_u64(__a, b, __c) __extension__ ({ \ |
| 6122 | uint64x1x2_t __b = (b); \ |
| 6123 | uint64x1x2_t r; __builtin_neon_vld2_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], __c, 19); r; }) |
| 6124 | #define vld2_lane_s64(__a, b, __c) __extension__ ({ \ |
| 6125 | int64x1x2_t __b = (b); \ |
| 6126 | int64x1x2_t r; __builtin_neon_vld2_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], __c, 3); r; }) |
| 6127 | #define vld2_lane_f64(__a, b, __c) __extension__ ({ \ |
| 6128 | float64x1x2_t __b = (b); \ |
| 6129 | float64x1x2_t r; __builtin_neon_vld2_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], __c, 9); r; }) |
| 6130 | #define vld2_lane_p64(__a, b, __c) __extension__ ({ \ |
| 6131 | poly64x1x2_t __b = (b); \ |
| 6132 | poly64x1x2_t r; __builtin_neon_vld2_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], __c, 6); r; }) |
| 6133 | |
| 6134 | #define vld3q_u64(__a) __extension__ ({ \ |
| 6135 | uint64x2x3_t r; __builtin_neon_vld3q_v(&r, __a, 51); r; }) |
| 6136 | #define vld3q_s64(__a) __extension__ ({ \ |
| 6137 | int64x2x3_t r; __builtin_neon_vld3q_v(&r, __a, 35); r; }) |
| 6138 | #define vld3q_f64(__a) __extension__ ({ \ |
| 6139 | float64x2x3_t r; __builtin_neon_vld3q_v(&r, __a, 41); r; }) |
| 6140 | #define vld3_f64(__a) __extension__ ({ \ |
| 6141 | float64x1x3_t r; __builtin_neon_vld3_v(&r, __a, 9); r; }) |
| 6142 | #define vld3_p64(__a) __extension__ ({ \ |
| 6143 | poly64x1x3_t r; __builtin_neon_vld3_v(&r, __a, 6); r; }) |
| 6144 | #define vld3q_p64(__a) __extension__ ({ \ |
| 6145 | poly64x2x3_t r; __builtin_neon_vld3q_v(&r, __a, 38); r; }) |
| 6146 | |
| 6147 | #define vld3q_dup_u8(__a) __extension__ ({ \ |
| 6148 | uint8x16x3_t r; __builtin_neon_vld3q_dup_v(&r, __a, 48); r; }) |
| 6149 | #define vld3q_dup_u16(__a) __extension__ ({ \ |
| 6150 | uint16x8x3_t r; __builtin_neon_vld3q_dup_v(&r, __a, 49); r; }) |
| 6151 | #define vld3q_dup_u32(__a) __extension__ ({ \ |
| 6152 | uint32x4x3_t r; __builtin_neon_vld3q_dup_v(&r, __a, 50); r; }) |
| 6153 | #define vld3q_dup_u64(__a) __extension__ ({ \ |
| 6154 | uint64x2x3_t r; __builtin_neon_vld3q_dup_v(&r, __a, 51); r; }) |
| 6155 | #define vld3q_dup_s8(__a) __extension__ ({ \ |
| 6156 | int8x16x3_t r; __builtin_neon_vld3q_dup_v(&r, __a, 32); r; }) |
| 6157 | #define vld3q_dup_s16(__a) __extension__ ({ \ |
| 6158 | int16x8x3_t r; __builtin_neon_vld3q_dup_v(&r, __a, 33); r; }) |
| 6159 | #define vld3q_dup_s32(__a) __extension__ ({ \ |
| 6160 | int32x4x3_t r; __builtin_neon_vld3q_dup_v(&r, __a, 34); r; }) |
| 6161 | #define vld3q_dup_s64(__a) __extension__ ({ \ |
| 6162 | int64x2x3_t r; __builtin_neon_vld3q_dup_v(&r, __a, 35); r; }) |
| 6163 | #define vld3q_dup_f16(__a) __extension__ ({ \ |
| 6164 | float16x8x3_t r; __builtin_neon_vld3q_dup_v(&r, __a, 39); r; }) |
| 6165 | #define vld3q_dup_f32(__a) __extension__ ({ \ |
| 6166 | float32x4x3_t r; __builtin_neon_vld3q_dup_v(&r, __a, 40); r; }) |
| 6167 | #define vld3q_dup_f64(__a) __extension__ ({ \ |
| 6168 | float64x2x3_t r; __builtin_neon_vld3q_dup_v(&r, __a, 41); r; }) |
| 6169 | #define vld3q_dup_p8(__a) __extension__ ({ \ |
| 6170 | poly8x16x3_t r; __builtin_neon_vld3q_dup_v(&r, __a, 36); r; }) |
| 6171 | #define vld3q_dup_p16(__a) __extension__ ({ \ |
| 6172 | poly16x8x3_t r; __builtin_neon_vld3q_dup_v(&r, __a, 37); r; }) |
| 6173 | #define vld3q_dup_p64(__a) __extension__ ({ \ |
| 6174 | poly64x2x3_t r; __builtin_neon_vld3q_dup_v(&r, __a, 38); r; }) |
| 6175 | #define vld3_dup_f64(__a) __extension__ ({ \ |
| 6176 | float64x1x3_t r; __builtin_neon_vld3_dup_v(&r, __a, 9); r; }) |
| 6177 | #define vld3_dup_p64(__a) __extension__ ({ \ |
| 6178 | poly64x1x3_t r; __builtin_neon_vld3_dup_v(&r, __a, 6); r; }) |
| 6179 | |
| 6180 | #define vld3q_lane_u8(__a, b, __c) __extension__ ({ \ |
| 6181 | uint8x16x3_t __b = (b); \ |
| 6182 | uint8x16x3_t r; __builtin_neon_vld3q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], __c, 48); r; }) |
| 6183 | #define vld3q_lane_u64(__a, b, __c) __extension__ ({ \ |
| 6184 | uint64x2x3_t __b = (b); \ |
| 6185 | uint64x2x3_t r; __builtin_neon_vld3q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], __c, 51); r; }) |
| 6186 | #define vld3q_lane_s8(__a, b, __c) __extension__ ({ \ |
| 6187 | int8x16x3_t __b = (b); \ |
| 6188 | int8x16x3_t r; __builtin_neon_vld3q_lane_v(&r, __a, __b.val[0], __b.val[1], __b.val[2], __c, 32); r; }) |
| 6189 | #define vld3q_lane_s64(__a, b, __c) __extension__ ({ \ |
| 6190 | int64x2x3_t __b = (b); \ |
| 6191 | int64x2x3_t r; __builtin_neon_vld3q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], __c, 35); r; }) |
| 6192 | #define vld3q_lane_f64(__a, b, __c) __extension__ ({ \ |
| 6193 | float64x2x3_t __b = (b); \ |
| 6194 | float64x2x3_t r; __builtin_neon_vld3q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], __c, 41); r; }) |
| 6195 | #define vld3q_lane_p8(__a, b, __c) __extension__ ({ \ |
| 6196 | poly8x16x3_t __b = (b); \ |
| 6197 | poly8x16x3_t r; __builtin_neon_vld3q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], __c, 36); r; }) |
| 6198 | #define vld3q_lane_p64(__a, b, __c) __extension__ ({ \ |
| 6199 | poly64x2x3_t __b = (b); \ |
| 6200 | poly64x2x3_t r; __builtin_neon_vld3q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], __c, 38); r; }) |
| 6201 | #define vld3_lane_u64(__a, b, __c) __extension__ ({ \ |
| 6202 | uint64x1x3_t __b = (b); \ |
| 6203 | uint64x1x3_t r; __builtin_neon_vld3_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], __c, 19); r; }) |
| 6204 | #define vld3_lane_s64(__a, b, __c) __extension__ ({ \ |
| 6205 | int64x1x3_t __b = (b); \ |
| 6206 | int64x1x3_t r; __builtin_neon_vld3_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], __c, 3); r; }) |
| 6207 | #define vld3_lane_f64(__a, b, __c) __extension__ ({ \ |
| 6208 | float64x1x3_t __b = (b); \ |
| 6209 | float64x1x3_t r; __builtin_neon_vld3_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], __c, 9); r; }) |
| 6210 | #define vld3_lane_p64(__a, b, __c) __extension__ ({ \ |
| 6211 | poly64x1x3_t __b = (b); \ |
| 6212 | poly64x1x3_t r; __builtin_neon_vld3_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], __c, 6); r; }) |
| 6213 | |
| 6214 | #define vld1q_u8_x3(__a) __extension__ ({ \ |
| 6215 | uint8x16x3_t r; __builtin_neon_vld1q_x3_v(&r, __a, 48); r; }) |
| 6216 | #define vld1q_u16_x3(__a) __extension__ ({ \ |
| 6217 | uint16x8x3_t r; __builtin_neon_vld1q_x3_v(&r, __a, 49); r; }) |
| 6218 | #define vld1q_u32_x3(__a) __extension__ ({ \ |
| 6219 | uint32x4x3_t r; __builtin_neon_vld1q_x3_v(&r, __a, 50); r; }) |
| 6220 | #define vld1q_u64_x3(__a) __extension__ ({ \ |
| 6221 | uint64x2x3_t r; __builtin_neon_vld1q_x3_v(&r, __a, 51); r; }) |
| 6222 | #define vld1q_s8_x3(__a) __extension__ ({ \ |
| 6223 | int8x16x3_t r; __builtin_neon_vld1q_x3_v(&r, __a, 32); r; }) |
| 6224 | #define vld1q_s16_x3(__a) __extension__ ({ \ |
| 6225 | int16x8x3_t r; __builtin_neon_vld1q_x3_v(&r, __a, 33); r; }) |
| 6226 | #define vld1q_s32_x3(__a) __extension__ ({ \ |
| 6227 | int32x4x3_t r; __builtin_neon_vld1q_x3_v(&r, __a, 34); r; }) |
| 6228 | #define vld1q_s64_x3(__a) __extension__ ({ \ |
| 6229 | int64x2x3_t r; __builtin_neon_vld1q_x3_v(&r, __a, 35); r; }) |
| 6230 | #define vld1q_f16_x3(__a) __extension__ ({ \ |
| 6231 | float16x8x3_t r; __builtin_neon_vld1q_x3_v(&r, __a, 39); r; }) |
| 6232 | #define vld1q_f32_x3(__a) __extension__ ({ \ |
| 6233 | float32x4x3_t r; __builtin_neon_vld1q_x3_v(&r, __a, 40); r; }) |
| 6234 | #define vld1q_f64_x3(__a) __extension__ ({ \ |
| 6235 | float64x2x3_t r; __builtin_neon_vld1q_x3_v(&r, __a, 41); r; }) |
| 6236 | #define vld1q_p8_x3(__a) __extension__ ({ \ |
| 6237 | poly8x16x3_t r; __builtin_neon_vld1q_x3_v(&r, __a, 36); r; }) |
| 6238 | #define vld1q_p16_x3(__a) __extension__ ({ \ |
| 6239 | poly16x8x3_t r; __builtin_neon_vld1q_x3_v(&r, __a, 37); r; }) |
| 6240 | #define vld1q_p64_x3(__a) __extension__ ({ \ |
| 6241 | poly64x2x3_t r; __builtin_neon_vld1q_x3_v(&r, __a, 38); r; }) |
| 6242 | #define vld1_u8_x3(__a) __extension__ ({ \ |
| 6243 | uint8x8x3_t r; __builtin_neon_vld1_x3_v(&r, __a, 16); r; }) |
| 6244 | #define vld1_u16_x3(__a) __extension__ ({ \ |
| 6245 | uint16x4x3_t r; __builtin_neon_vld1_x3_v(&r, __a, 17); r; }) |
| 6246 | #define vld1_u32_x3(__a) __extension__ ({ \ |
| 6247 | uint32x2x3_t r; __builtin_neon_vld1_x3_v(&r, __a, 18); r; }) |
| 6248 | #define vld1_u64_x3(__a) __extension__ ({ \ |
| 6249 | uint64x1x3_t r; __builtin_neon_vld1_x3_v(&r, __a, 19); r; }) |
| 6250 | #define vld1_s8_x3(__a) __extension__ ({ \ |
| 6251 | int8x8x3_t r; __builtin_neon_vld1_x3_v(&r, __a, 0); r; }) |
| 6252 | #define vld1_s16_x3(__a) __extension__ ({ \ |
| 6253 | int16x4x3_t r; __builtin_neon_vld1_x3_v(&r, __a, 1); r; }) |
| 6254 | #define vld1_s32_x3(__a) __extension__ ({ \ |
| 6255 | int32x2x3_t r; __builtin_neon_vld1_x3_v(&r, __a, 2); r; }) |
| 6256 | #define vld1_s64_x3(__a) __extension__ ({ \ |
| 6257 | int64x1x3_t r; __builtin_neon_vld1_x3_v(&r, __a, 3); r; }) |
| 6258 | #define vld1_f16_x3(__a) __extension__ ({ \ |
| 6259 | float16x4x3_t r; __builtin_neon_vld1_x3_v(&r, __a, 7); r; }) |
| 6260 | #define vld1_f32_x3(__a) __extension__ ({ \ |
| 6261 | float32x2x3_t r; __builtin_neon_vld1_x3_v(&r, __a, 8); r; }) |
| 6262 | #define vld1_f64_x3(__a) __extension__ ({ \ |
| 6263 | float64x1x3_t r; __builtin_neon_vld1_x3_v(&r, __a, 9); r; }) |
| 6264 | #define vld1_p8_x3(__a) __extension__ ({ \ |
| 6265 | poly8x8x3_t r; __builtin_neon_vld1_x3_v(&r, __a, 4); r; }) |
| 6266 | #define vld1_p16_x3(__a) __extension__ ({ \ |
| 6267 | poly16x4x3_t r; __builtin_neon_vld1_x3_v(&r, __a, 5); r; }) |
| 6268 | #define vld1_p64_x3(__a) __extension__ ({ \ |
| 6269 | poly64x1x3_t r; __builtin_neon_vld1_x3_v(&r, __a, 6); r; }) |
| 6270 | |
| 6271 | #define vld4q_u64(__a) __extension__ ({ \ |
| 6272 | uint64x2x4_t r; __builtin_neon_vld4q_v(&r, __a, 51); r; }) |
| 6273 | #define vld4q_s64(__a) __extension__ ({ \ |
| 6274 | int64x2x4_t r; __builtin_neon_vld4q_v(&r, __a, 35); r; }) |
| 6275 | #define vld4q_f64(__a) __extension__ ({ \ |
| 6276 | float64x2x4_t r; __builtin_neon_vld4q_v(&r, __a, 41); r; }) |
| 6277 | #define vld4_f64(__a) __extension__ ({ \ |
| 6278 | float64x1x4_t r; __builtin_neon_vld4_v(&r, __a, 9); r; }) |
| 6279 | #define vld4_p64(__a) __extension__ ({ \ |
| 6280 | poly64x1x4_t r; __builtin_neon_vld4_v(&r, __a, 6); r; }) |
| 6281 | #define vld4q_p64(__a) __extension__ ({ \ |
| 6282 | poly64x2x4_t r; __builtin_neon_vld4q_v(&r, __a, 38); r; }) |
| 6283 | |
| 6284 | #define vld4q_dup_u8(__a) __extension__ ({ \ |
| 6285 | uint8x16x4_t r; __builtin_neon_vld4q_dup_v(&r, __a, 48); r; }) |
| 6286 | #define vld4q_dup_u16(__a) __extension__ ({ \ |
| 6287 | uint16x8x4_t r; __builtin_neon_vld4q_dup_v(&r, __a, 49); r; }) |
| 6288 | #define vld4q_dup_u32(__a) __extension__ ({ \ |
| 6289 | uint32x4x4_t r; __builtin_neon_vld4q_dup_v(&r, __a, 50); r; }) |
| 6290 | #define vld4q_dup_u64(__a) __extension__ ({ \ |
| 6291 | uint64x2x4_t r; __builtin_neon_vld4q_dup_v(&r, __a, 51); r; }) |
| 6292 | #define vld4q_dup_s8(__a) __extension__ ({ \ |
| 6293 | int8x16x4_t r; __builtin_neon_vld4q_dup_v(&r, __a, 32); r; }) |
| 6294 | #define vld4q_dup_s16(__a) __extension__ ({ \ |
| 6295 | int16x8x4_t r; __builtin_neon_vld4q_dup_v(&r, __a, 33); r; }) |
| 6296 | #define vld4q_dup_s32(__a) __extension__ ({ \ |
| 6297 | int32x4x4_t r; __builtin_neon_vld4q_dup_v(&r, __a, 34); r; }) |
| 6298 | #define vld4q_dup_s64(__a) __extension__ ({ \ |
| 6299 | int64x2x4_t r; __builtin_neon_vld4q_dup_v(&r, __a, 35); r; }) |
| 6300 | #define vld4q_dup_f16(__a) __extension__ ({ \ |
| 6301 | float16x8x4_t r; __builtin_neon_vld4q_dup_v(&r, __a, 39); r; }) |
| 6302 | #define vld4q_dup_f32(__a) __extension__ ({ \ |
| 6303 | float32x4x4_t r; __builtin_neon_vld4q_dup_v(&r, __a, 40); r; }) |
| 6304 | #define vld4q_dup_f64(__a) __extension__ ({ \ |
| 6305 | float64x2x4_t r; __builtin_neon_vld4q_dup_v(&r, __a, 41); r; }) |
| 6306 | #define vld4q_dup_p8(__a) __extension__ ({ \ |
| 6307 | poly8x16x4_t r; __builtin_neon_vld4q_dup_v(&r, __a, 36); r; }) |
| 6308 | #define vld4q_dup_p16(__a) __extension__ ({ \ |
| 6309 | poly16x8x4_t r; __builtin_neon_vld4q_dup_v(&r, __a, 37); r; }) |
| 6310 | #define vld4q_dup_p64(__a) __extension__ ({ \ |
| 6311 | poly64x2x4_t r; __builtin_neon_vld4q_dup_v(&r, __a, 38); r; }) |
| 6312 | #define vld4_dup_f64(__a) __extension__ ({ \ |
| 6313 | float64x1x4_t r; __builtin_neon_vld4_dup_v(&r, __a, 9); r; }) |
| 6314 | #define vld4_dup_p64(__a) __extension__ ({ \ |
| 6315 | poly64x1x4_t r; __builtin_neon_vld4_dup_v(&r, __a, 6); r; }) |
| 6316 | |
| 6317 | #define vld4q_lane_u8(__a, b, __c) __extension__ ({ \ |
| 6318 | uint8x16x4_t __b = (b); \ |
| 6319 | uint8x16x4_t r; __builtin_neon_vld4q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], __c, 48); r; }) |
| 6320 | #define vld4q_lane_u64(__a, b, __c) __extension__ ({ \ |
| 6321 | uint64x2x4_t __b = (b); \ |
| 6322 | uint64x2x4_t r; __builtin_neon_vld4q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], __c, 51); r; }) |
| 6323 | #define vld4q_lane_s8(__a, b, __c) __extension__ ({ \ |
| 6324 | int8x16x4_t __b = (b); \ |
| 6325 | int8x16x4_t r; __builtin_neon_vld4q_lane_v(&r, __a, __b.val[0], __b.val[1], __b.val[2], __b.val[3], __c, 32); r; }) |
| 6326 | #define vld4q_lane_s64(__a, b, __c) __extension__ ({ \ |
| 6327 | int64x2x4_t __b = (b); \ |
| 6328 | int64x2x4_t r; __builtin_neon_vld4q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], __c, 35); r; }) |
| 6329 | #define vld4q_lane_f64(__a, b, __c) __extension__ ({ \ |
| 6330 | float64x2x4_t __b = (b); \ |
| 6331 | float64x2x4_t r; __builtin_neon_vld4q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], __c, 41); r; }) |
| 6332 | #define vld4q_lane_p8(__a, b, __c) __extension__ ({ \ |
| 6333 | poly8x16x4_t __b = (b); \ |
| 6334 | poly8x16x4_t r; __builtin_neon_vld4q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], __c, 36); r; }) |
| 6335 | #define vld4q_lane_p64(__a, b, __c) __extension__ ({ \ |
| 6336 | poly64x2x4_t __b = (b); \ |
| 6337 | poly64x2x4_t r; __builtin_neon_vld4q_lane_v(&r, __a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], __c, 38); r; }) |
| 6338 | #define vld4_lane_u64(__a, b, __c) __extension__ ({ \ |
| 6339 | uint64x1x4_t __b = (b); \ |
| 6340 | uint64x1x4_t r; __builtin_neon_vld4_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], __c, 19); r; }) |
| 6341 | #define vld4_lane_s64(__a, b, __c) __extension__ ({ \ |
| 6342 | int64x1x4_t __b = (b); \ |
| 6343 | int64x1x4_t r; __builtin_neon_vld4_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], __c, 3); r; }) |
| 6344 | #define vld4_lane_f64(__a, b, __c) __extension__ ({ \ |
| 6345 | float64x1x4_t __b = (b); \ |
| 6346 | float64x1x4_t r; __builtin_neon_vld4_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], __c, 9); r; }) |
| 6347 | #define vld4_lane_p64(__a, b, __c) __extension__ ({ \ |
| 6348 | poly64x1x4_t __b = (b); \ |
| 6349 | poly64x1x4_t r; __builtin_neon_vld4_lane_v(&r, __a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], __c, 6); r; }) |
| 6350 | |
| 6351 | #define vld1q_u8_x4(__a) __extension__ ({ \ |
| 6352 | uint8x16x4_t r; __builtin_neon_vld1q_x4_v(&r, __a, 48); r; }) |
| 6353 | #define vld1q_u16_x4(__a) __extension__ ({ \ |
| 6354 | uint16x8x4_t r; __builtin_neon_vld1q_x4_v(&r, __a, 49); r; }) |
| 6355 | #define vld1q_u32_x4(__a) __extension__ ({ \ |
| 6356 | uint32x4x4_t r; __builtin_neon_vld1q_x4_v(&r, __a, 50); r; }) |
| 6357 | #define vld1q_u64_x4(__a) __extension__ ({ \ |
| 6358 | uint64x2x4_t r; __builtin_neon_vld1q_x4_v(&r, __a, 51); r; }) |
| 6359 | #define vld1q_s8_x4(__a) __extension__ ({ \ |
| 6360 | int8x16x4_t r; __builtin_neon_vld1q_x4_v(&r, __a, 32); r; }) |
| 6361 | #define vld1q_s16_x4(__a) __extension__ ({ \ |
| 6362 | int16x8x4_t r; __builtin_neon_vld1q_x4_v(&r, __a, 33); r; }) |
| 6363 | #define vld1q_s32_x4(__a) __extension__ ({ \ |
| 6364 | int32x4x4_t r; __builtin_neon_vld1q_x4_v(&r, __a, 34); r; }) |
| 6365 | #define vld1q_s64_x4(__a) __extension__ ({ \ |
| 6366 | int64x2x4_t r; __builtin_neon_vld1q_x4_v(&r, __a, 35); r; }) |
| 6367 | #define vld1q_f16_x4(__a) __extension__ ({ \ |
| 6368 | float16x8x4_t r; __builtin_neon_vld1q_x4_v(&r, __a, 39); r; }) |
| 6369 | #define vld1q_f32_x4(__a) __extension__ ({ \ |
| 6370 | float32x4x4_t r; __builtin_neon_vld1q_x4_v(&r, __a, 40); r; }) |
| 6371 | #define vld1q_f64_x4(__a) __extension__ ({ \ |
| 6372 | float64x2x4_t r; __builtin_neon_vld1q_x4_v(&r, __a, 41); r; }) |
| 6373 | #define vld1q_p8_x4(__a) __extension__ ({ \ |
| 6374 | poly8x16x4_t r; __builtin_neon_vld1q_x4_v(&r, __a, 36); r; }) |
| 6375 | #define vld1q_p16_x4(__a) __extension__ ({ \ |
| 6376 | poly16x8x4_t r; __builtin_neon_vld1q_x4_v(&r, __a, 37); r; }) |
| 6377 | #define vld1q_p64_x4(__a) __extension__ ({ \ |
| 6378 | poly64x2x4_t r; __builtin_neon_vld1q_x4_v(&r, __a, 38); r; }) |
| 6379 | #define vld1_u8_x4(__a) __extension__ ({ \ |
| 6380 | uint8x8x4_t r; __builtin_neon_vld1_x4_v(&r, __a, 16); r; }) |
| 6381 | #define vld1_u16_x4(__a) __extension__ ({ \ |
| 6382 | uint16x4x4_t r; __builtin_neon_vld1_x4_v(&r, __a, 17); r; }) |
| 6383 | #define vld1_u32_x4(__a) __extension__ ({ \ |
| 6384 | uint32x2x4_t r; __builtin_neon_vld1_x4_v(&r, __a, 18); r; }) |
| 6385 | #define vld1_u64_x4(__a) __extension__ ({ \ |
| 6386 | uint64x1x4_t r; __builtin_neon_vld1_x4_v(&r, __a, 19); r; }) |
| 6387 | #define vld1_s8_x4(__a) __extension__ ({ \ |
| 6388 | int8x8x4_t r; __builtin_neon_vld1_x4_v(&r, __a, 0); r; }) |
| 6389 | #define vld1_s16_x4(__a) __extension__ ({ \ |
| 6390 | int16x4x4_t r; __builtin_neon_vld1_x4_v(&r, __a, 1); r; }) |
| 6391 | #define vld1_s32_x4(__a) __extension__ ({ \ |
| 6392 | int32x2x4_t r; __builtin_neon_vld1_x4_v(&r, __a, 2); r; }) |
| 6393 | #define vld1_s64_x4(__a) __extension__ ({ \ |
| 6394 | int64x1x4_t r; __builtin_neon_vld1_x4_v(&r, __a, 3); r; }) |
| 6395 | #define vld1_f16_x4(__a) __extension__ ({ \ |
| 6396 | float16x4x4_t r; __builtin_neon_vld1_x4_v(&r, __a, 7); r; }) |
| 6397 | #define vld1_f32_x4(__a) __extension__ ({ \ |
| 6398 | float32x2x4_t r; __builtin_neon_vld1_x4_v(&r, __a, 8); r; }) |
| 6399 | #define vld1_f64_x4(__a) __extension__ ({ \ |
| 6400 | float64x1x4_t r; __builtin_neon_vld1_x4_v(&r, __a, 9); r; }) |
| 6401 | #define vld1_p8_x4(__a) __extension__ ({ \ |
| 6402 | poly8x8x4_t r; __builtin_neon_vld1_x4_v(&r, __a, 4); r; }) |
| 6403 | #define vld1_p16_x4(__a) __extension__ ({ \ |
| 6404 | poly16x4x4_t r; __builtin_neon_vld1_x4_v(&r, __a, 5); r; }) |
| 6405 | #define vld1_p64_x4(__a) __extension__ ({ \ |
| 6406 | poly64x1x4_t r; __builtin_neon_vld1_x4_v(&r, __a, 6); r; }) |
| 6407 | |
| 6408 | __ai float64x1_t vmax_f64(float64x1_t __a, float64x1_t __b) { |
| 6409 | return (float64x1_t)__builtin_neon_vmax_v((int8x8_t)__a, (int8x8_t)__b, 9); } |
| 6410 | __ai float64x2_t vmaxq_f64(float64x2_t __a, float64x2_t __b) { |
| 6411 | return (float64x2_t)__builtin_neon_vmaxq_v((int8x16_t)__a, (int8x16_t)__b, 41); } |
| 6412 | |
| 6413 | __ai int8x16_t vpmaxq_s8(int8x16_t __a, int8x16_t __b) { |
| 6414 | return (int8x16_t)__builtin_neon_vpmaxq_v(__a, __b, 32); } |
| 6415 | __ai int16x8_t vpmaxq_s16(int16x8_t __a, int16x8_t __b) { |
| 6416 | return (int16x8_t)__builtin_neon_vpmaxq_v((int8x16_t)__a, (int8x16_t)__b, 33); } |
| 6417 | __ai int32x4_t vpmaxq_s32(int32x4_t __a, int32x4_t __b) { |
| 6418 | return (int32x4_t)__builtin_neon_vpmaxq_v((int8x16_t)__a, (int8x16_t)__b, 34); } |
| 6419 | __ai uint8x16_t vpmaxq_u8(uint8x16_t __a, uint8x16_t __b) { |
| 6420 | return (uint8x16_t)__builtin_neon_vpmaxq_v((int8x16_t)__a, (int8x16_t)__b, 48); } |
| 6421 | __ai uint16x8_t vpmaxq_u16(uint16x8_t __a, uint16x8_t __b) { |
| 6422 | return (uint16x8_t)__builtin_neon_vpmaxq_v((int8x16_t)__a, (int8x16_t)__b, 49); } |
| 6423 | __ai uint32x4_t vpmaxq_u32(uint32x4_t __a, uint32x4_t __b) { |
| 6424 | return (uint32x4_t)__builtin_neon_vpmaxq_v((int8x16_t)__a, (int8x16_t)__b, 50); } |
| 6425 | __ai float32x4_t vpmaxq_f32(float32x4_t __a, float32x4_t __b) { |
| 6426 | return (float32x4_t)__builtin_neon_vpmaxq_v((int8x16_t)__a, (int8x16_t)__b, 40); } |
| 6427 | __ai float64x2_t vpmaxq_f64(float64x2_t __a, float64x2_t __b) { |
| 6428 | return (float64x2_t)__builtin_neon_vpmaxq_v((int8x16_t)__a, (int8x16_t)__b, 41); } |
| 6429 | |
| 6430 | __ai float64x1_t vmin_f64(float64x1_t __a, float64x1_t __b) { |
| 6431 | return (float64x1_t)__builtin_neon_vmin_v((int8x8_t)__a, (int8x8_t)__b, 9); } |
| 6432 | __ai float64x2_t vminq_f64(float64x2_t __a, float64x2_t __b) { |
| 6433 | return (float64x2_t)__builtin_neon_vminq_v((int8x16_t)__a, (int8x16_t)__b, 41); } |
| 6434 | |
| 6435 | __ai int8x16_t vpminq_s8(int8x16_t __a, int8x16_t __b) { |
| 6436 | return (int8x16_t)__builtin_neon_vpminq_v(__a, __b, 32); } |
| 6437 | __ai int16x8_t vpminq_s16(int16x8_t __a, int16x8_t __b) { |
| 6438 | return (int16x8_t)__builtin_neon_vpminq_v((int8x16_t)__a, (int8x16_t)__b, 33); } |
| 6439 | __ai int32x4_t vpminq_s32(int32x4_t __a, int32x4_t __b) { |
| 6440 | return (int32x4_t)__builtin_neon_vpminq_v((int8x16_t)__a, (int8x16_t)__b, 34); } |
| 6441 | __ai uint8x16_t vpminq_u8(uint8x16_t __a, uint8x16_t __b) { |
| 6442 | return (uint8x16_t)__builtin_neon_vpminq_v((int8x16_t)__a, (int8x16_t)__b, 48); } |
| 6443 | __ai uint16x8_t vpminq_u16(uint16x8_t __a, uint16x8_t __b) { |
| 6444 | return (uint16x8_t)__builtin_neon_vpminq_v((int8x16_t)__a, (int8x16_t)__b, 49); } |
| 6445 | __ai uint32x4_t vpminq_u32(uint32x4_t __a, uint32x4_t __b) { |
| 6446 | return (uint32x4_t)__builtin_neon_vpminq_v((int8x16_t)__a, (int8x16_t)__b, 50); } |
| 6447 | __ai float32x4_t vpminq_f32(float32x4_t __a, float32x4_t __b) { |
| 6448 | return (float32x4_t)__builtin_neon_vpminq_v((int8x16_t)__a, (int8x16_t)__b, 40); } |
| 6449 | __ai float64x2_t vpminq_f64(float64x2_t __a, float64x2_t __b) { |
| 6450 | return (float64x2_t)__builtin_neon_vpminq_v((int8x16_t)__a, (int8x16_t)__b, 41); } |
| 6451 | |
| 6452 | __ai float64x1_t vmla_f64(float64x1_t __a, float64x1_t __b, float64x1_t __c) { |
| 6453 | return __a + (__b * __c); } |
| 6454 | __ai float64x2_t vmlaq_f64(float64x2_t __a, float64x2_t __b, float64x2_t __c) { |
| 6455 | return __a + (__b * __c); } |
| 6456 | |
| 6457 | __ai float64x1_t vmls_f64(float64x1_t __a, float64x1_t __b, float64x1_t __c) { |
| 6458 | return __a - (__b * __c); } |
| 6459 | __ai float64x2_t vmlsq_f64(float64x2_t __a, float64x2_t __b, float64x2_t __c) { |
| 6460 | return __a - (__b * __c); } |
| 6461 | |
| 6462 | __ai float64x1_t vmov_n_f64(float64_t __a) { |
| 6463 | return (float64x1_t){ __a }; } |
| 6464 | __ai float64x2_t vmovq_n_f64(float64_t __a) { |
| 6465 | return (float64x2_t){ __a, __a }; } |
| 6466 | |
| 6467 | __ai float64x1_t vmul_f64(float64x1_t __a, float64x1_t __b) { |
| 6468 | return __a * __b; } |
| 6469 | __ai float64x2_t vmulq_f64(float64x2_t __a, float64x2_t __b) { |
| 6470 | return __a * __b; } |
| 6471 | |
| 6472 | __ai float32x2_t vmulx_f32(float32x2_t __a, float32x2_t __b) { |
| 6473 | return (float32x2_t)__builtin_neon_vmulx_v((int8x8_t)__a, (int8x8_t)__b, 8); } |
| 6474 | __ai float64x1_t vmulx_f64(float64x1_t __a, float64x1_t __b) { |
| 6475 | return (float64x1_t)__builtin_neon_vmulx_v((int8x8_t)__a, (int8x8_t)__b, 9); } |
| 6476 | __ai float32x4_t vmulxq_f32(float32x4_t __a, float32x4_t __b) { |
| 6477 | return (float32x4_t)__builtin_neon_vmulxq_v((int8x16_t)__a, (int8x16_t)__b, 40); } |
| 6478 | __ai float64x2_t vmulxq_f64(float64x2_t __a, float64x2_t __b) { |
| 6479 | return (float64x2_t)__builtin_neon_vmulxq_v((int8x16_t)__a, (int8x16_t)__b, 41); } |
| 6480 | |
| 6481 | __ai int64x1_t vneg_s64(int64x1_t __a) { |
| 6482 | return -__a; } |
| 6483 | __ai float64x1_t vneg_f64(float64x1_t __a) { |
| 6484 | return -__a; } |
| 6485 | __ai float64x2_t vnegq_f64(float64x2_t __a) { |
| 6486 | return -__a; } |
| 6487 | __ai int64x2_t vnegq_s64(int64x2_t __a) { |
| 6488 | return -__a; } |
| 6489 | |
| 6490 | __ai int64x1_t vqabs_s64(int64x1_t __a) { |
| 6491 | return (int64x1_t)__builtin_neon_vqabs_v((int8x8_t)__a, 3); } |
| 6492 | __ai int64x2_t vqabsq_s64(int64x2_t __a) { |
| 6493 | return (int64x2_t)__builtin_neon_vqabsq_v((int8x16_t)__a, 35); } |
| 6494 | |
| 6495 | __ai int64x1_t vqneg_s64(int64x1_t __a) { |
| 6496 | return (int64x1_t)__builtin_neon_vqneg_v((int8x8_t)__a, 3); } |
| 6497 | __ai int64x2_t vqnegq_s64(int64x2_t __a) { |
| 6498 | return (int64x2_t)__builtin_neon_vqnegq_v((int8x16_t)__a, 35); } |
| 6499 | |
| 6500 | #define vqrshrn_high_n_s16(a, b, __c) __extension__ ({ \ |
| 6501 | int8x8_t __a = (a); int16x8_t __b = (b); \ |
| 6502 | (int8x16_t)vcombine_s16(__a, vqrshrn_n_s16(__b, __c)); }) |
| 6503 | #define vqrshrn_high_n_s32(a, b, __c) __extension__ ({ \ |
| 6504 | int16x4_t __a = (a); int32x4_t __b = (b); \ |
| 6505 | (int16x8_t)vcombine_s32(__a, vqrshrn_n_s32(__b, __c)); }) |
| 6506 | #define vqrshrn_high_n_s64(a, b, __c) __extension__ ({ \ |
| 6507 | int32x2_t __a = (a); int64x2_t __b = (b); \ |
| 6508 | (int32x4_t)vcombine_s64(__a, vqrshrn_n_s64(__b, __c)); }) |
| 6509 | #define vqrshrn_high_n_u16(a, b, __c) __extension__ ({ \ |
| 6510 | uint8x8_t __a = (a); uint16x8_t __b = (b); \ |
| 6511 | (uint8x16_t)vcombine_u16(__a, vqrshrn_n_u16(__b, __c)); }) |
| 6512 | #define vqrshrn_high_n_u32(a, b, __c) __extension__ ({ \ |
| 6513 | uint16x4_t __a = (a); uint32x4_t __b = (b); \ |
| 6514 | (uint16x8_t)vcombine_u32(__a, vqrshrn_n_u32(__b, __c)); }) |
| 6515 | #define vqrshrn_high_n_u64(a, b, __c) __extension__ ({ \ |
| 6516 | uint32x2_t __a = (a); uint64x2_t __b = (b); \ |
| 6517 | (uint32x4_t)vcombine_u64(__a, vqrshrn_n_u64(__b, __c)); }) |
| 6518 | |
| 6519 | #define vqrshrun_high_n_s16(a, b, __c) __extension__ ({ \ |
| 6520 | int8x8_t __a = (a); int16x8_t __b = (b); \ |
| 6521 | (int8x16_t)vcombine_s16(__a, vqrshrun_n_s16(__b, __c)); }) |
| 6522 | #define vqrshrun_high_n_s32(a, b, __c) __extension__ ({ \ |
| 6523 | int16x4_t __a = (a); int32x4_t __b = (b); \ |
| 6524 | (int16x8_t)vcombine_s32(__a, vqrshrun_n_s32(__b, __c)); }) |
| 6525 | #define vqrshrun_high_n_s64(a, b, __c) __extension__ ({ \ |
| 6526 | int32x2_t __a = (a); int64x2_t __b = (b); \ |
| 6527 | (int32x4_t)vcombine_s64(__a, vqrshrun_n_s64(__b, __c)); }) |
| 6528 | |
| 6529 | #define vqshrn_high_n_s16(a, b, __c) __extension__ ({ \ |
| 6530 | int8x8_t __a = (a); int16x8_t __b = (b); \ |
| 6531 | (int8x16_t)vcombine_s16(__a, vqshrn_n_s16(__b, __c)); }) |
| 6532 | #define vqshrn_high_n_s32(a, b, __c) __extension__ ({ \ |
| 6533 | int16x4_t __a = (a); int32x4_t __b = (b); \ |
| 6534 | (int16x8_t)vcombine_s32(__a, vqshrn_n_s32(__b, __c)); }) |
| 6535 | #define vqshrn_high_n_s64(a, b, __c) __extension__ ({ \ |
| 6536 | int32x2_t __a = (a); int64x2_t __b = (b); \ |
| 6537 | (int32x4_t)vcombine_s64(__a, vqshrn_n_s64(__b, __c)); }) |
| 6538 | #define vqshrn_high_n_u16(a, b, __c) __extension__ ({ \ |
| 6539 | uint8x8_t __a = (a); uint16x8_t __b = (b); \ |
| 6540 | (uint8x16_t)vcombine_u16(__a, vqshrn_n_u16(__b, __c)); }) |
| 6541 | #define vqshrn_high_n_u32(a, b, __c) __extension__ ({ \ |
| 6542 | uint16x4_t __a = (a); uint32x4_t __b = (b); \ |
| 6543 | (uint16x8_t)vcombine_u32(__a, vqshrn_n_u32(__b, __c)); }) |
| 6544 | #define vqshrn_high_n_u64(a, b, __c) __extension__ ({ \ |
| 6545 | uint32x2_t __a = (a); uint64x2_t __b = (b); \ |
| 6546 | (uint32x4_t)vcombine_u64(__a, vqshrn_n_u64(__b, __c)); }) |
| 6547 | |
| 6548 | #define vqshrun_high_n_s16(a, b, __c) __extension__ ({ \ |
| 6549 | int8x8_t __a = (a); int16x8_t __b = (b); \ |
| 6550 | (int8x16_t)vcombine_s16(__a, vqshrun_n_s16(__b, __c)); }) |
| 6551 | #define vqshrun_high_n_s32(a, b, __c) __extension__ ({ \ |
| 6552 | int16x4_t __a = (a); int32x4_t __b = (b); \ |
| 6553 | (int16x8_t)vcombine_s32(__a, vqshrun_n_s32(__b, __c)); }) |
| 6554 | #define vqshrun_high_n_s64(a, b, __c) __extension__ ({ \ |
| 6555 | int32x2_t __a = (a); int64x2_t __b = (b); \ |
| 6556 | (int32x4_t)vcombine_s64(__a, vqshrun_n_s64(__b, __c)); }) |
| 6557 | |
| 6558 | __ai int8x16_t vqmovn_high_s16(int8x8_t __a, int16x8_t __b) { |
| 6559 | int8x8_t __a1 = vqmovn_s16(__b); |
| 6560 | return __builtin_shufflevector(__a, __a1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15); } |
| 6561 | __ai int16x8_t vqmovn_high_s32(int16x4_t __a, int32x4_t __b) { |
| 6562 | int16x4_t __a1 = vqmovn_s32(__b); |
| 6563 | return __builtin_shufflevector(__a, __a1, 0, 1, 2, 3, 4, 5, 6, 7); } |
| 6564 | __ai int32x4_t vqmovn_high_s64(int32x2_t __a, int64x2_t __b) { |
| 6565 | int32x2_t __a1 = vqmovn_s64(__b); |
| 6566 | return __builtin_shufflevector(__a, __a1, 0, 1, 2, 3); } |
| 6567 | __ai uint8x16_t vqmovn_high_u16(uint8x8_t __a, uint16x8_t __b) { |
| 6568 | uint8x8_t __a1 = vqmovn_u16(__b); |
| 6569 | return __builtin_shufflevector(__a, __a1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15); } |
| 6570 | __ai uint16x8_t vqmovn_high_u32(uint16x4_t __a, uint32x4_t __b) { |
| 6571 | uint16x4_t __a1 = vqmovn_u32(__b); |
| 6572 | return __builtin_shufflevector(__a, __a1, 0, 1, 2, 3, 4, 5, 6, 7); } |
| 6573 | __ai uint32x4_t vqmovn_high_u64(uint32x2_t __a, uint64x2_t __b) { |
| 6574 | uint32x2_t __a1 = vqmovn_u64(__b); |
| 6575 | return __builtin_shufflevector(__a, __a1, 0, 1, 2, 3); } |
| 6576 | |
| 6577 | __ai int8x8_t vrbit_s8(int8x8_t __a) { |
| 6578 | return (int8x8_t)__builtin_neon_vrbit_v(__a, 0); } |
| 6579 | __ai uint8x8_t vrbit_u8(uint8x8_t __a) { |
| 6580 | return (uint8x8_t)__builtin_neon_vrbit_v((int8x8_t)__a, 16); } |
| 6581 | __ai poly8x8_t vrbit_p8(poly8x8_t __a) { |
| 6582 | return (poly8x8_t)__builtin_neon_vrbit_v((int8x8_t)__a, 4); } |
| 6583 | __ai int8x16_t vrbitq_s8(int8x16_t __a) { |
| 6584 | return (int8x16_t)__builtin_neon_vrbitq_v(__a, 32); } |
| 6585 | __ai uint8x16_t vrbitq_u8(uint8x16_t __a) { |
| 6586 | return (uint8x16_t)__builtin_neon_vrbitq_v((int8x16_t)__a, 48); } |
| 6587 | __ai poly8x16_t vrbitq_p8(poly8x16_t __a) { |
| 6588 | return (poly8x16_t)__builtin_neon_vrbitq_v((int8x16_t)__a, 36); } |
| 6589 | |
| 6590 | __ai int8x8_t vreinterpret_s8_f64(float64x1_t __a) { |
| 6591 | return (int8x8_t)__a; } |
| 6592 | __ai int8x8_t vreinterpret_s8_p64(poly64x1_t __a) { |
| 6593 | return (int8x8_t)__a; } |
| 6594 | __ai int16x4_t vreinterpret_s16_f64(float64x1_t __a) { |
| 6595 | return (int16x4_t)__a; } |
| 6596 | __ai int16x4_t vreinterpret_s16_p64(poly64x1_t __a) { |
| 6597 | return (int16x4_t)__a; } |
| 6598 | __ai int32x2_t vreinterpret_s32_f64(float64x1_t __a) { |
| 6599 | return (int32x2_t)__a; } |
| 6600 | __ai int32x2_t vreinterpret_s32_p64(poly64x1_t __a) { |
| 6601 | return (int32x2_t)__a; } |
| 6602 | __ai int64x1_t vreinterpret_s64_f64(float64x1_t __a) { |
| 6603 | return (int64x1_t)__a; } |
| 6604 | __ai int64x1_t vreinterpret_s64_p64(poly64x1_t __a) { |
| 6605 | return (int64x1_t)__a; } |
| 6606 | __ai uint8x8_t vreinterpret_u8_f64(float64x1_t __a) { |
| 6607 | return (uint8x8_t)__a; } |
| 6608 | __ai uint8x8_t vreinterpret_u8_p64(poly64x1_t __a) { |
| 6609 | return (uint8x8_t)__a; } |
| 6610 | __ai uint16x4_t vreinterpret_u16_f64(float64x1_t __a) { |
| 6611 | return (uint16x4_t)__a; } |
| 6612 | __ai uint16x4_t vreinterpret_u16_p64(poly64x1_t __a) { |
| 6613 | return (uint16x4_t)__a; } |
| 6614 | __ai uint32x2_t vreinterpret_u32_f64(float64x1_t __a) { |
| 6615 | return (uint32x2_t)__a; } |
| 6616 | __ai uint32x2_t vreinterpret_u32_p64(poly64x1_t __a) { |
| 6617 | return (uint32x2_t)__a; } |
| 6618 | __ai uint64x1_t vreinterpret_u64_f64(float64x1_t __a) { |
| 6619 | return (uint64x1_t)__a; } |
| 6620 | __ai uint64x1_t vreinterpret_u64_p64(poly64x1_t __a) { |
| 6621 | return (uint64x1_t)__a; } |
| 6622 | __ai float16x4_t vreinterpret_f16_f64(float64x1_t __a) { |
| 6623 | return (float16x4_t)__a; } |
| 6624 | __ai float16x4_t vreinterpret_f16_p64(poly64x1_t __a) { |
| 6625 | return (float16x4_t)__a; } |
| 6626 | __ai float32x2_t vreinterpret_f32_f64(float64x1_t __a) { |
| 6627 | return (float32x2_t)__a; } |
| 6628 | __ai float32x2_t vreinterpret_f32_p64(poly64x1_t __a) { |
| 6629 | return (float32x2_t)__a; } |
| 6630 | __ai float64x1_t vreinterpret_f64_s8(int8x8_t __a) { |
| 6631 | return (float64x1_t)__a; } |
| 6632 | __ai float64x1_t vreinterpret_f64_s16(int16x4_t __a) { |
| 6633 | return (float64x1_t)__a; } |
| 6634 | __ai float64x1_t vreinterpret_f64_s32(int32x2_t __a) { |
| 6635 | return (float64x1_t)__a; } |
| 6636 | __ai float64x1_t vreinterpret_f64_s64(int64x1_t __a) { |
| 6637 | return (float64x1_t)__a; } |
| 6638 | __ai float64x1_t vreinterpret_f64_u8(uint8x8_t __a) { |
| 6639 | return (float64x1_t)__a; } |
| 6640 | __ai float64x1_t vreinterpret_f64_u16(uint16x4_t __a) { |
| 6641 | return (float64x1_t)__a; } |
| 6642 | __ai float64x1_t vreinterpret_f64_u32(uint32x2_t __a) { |
| 6643 | return (float64x1_t)__a; } |
| 6644 | __ai float64x1_t vreinterpret_f64_u64(uint64x1_t __a) { |
| 6645 | return (float64x1_t)__a; } |
| 6646 | __ai float64x1_t vreinterpret_f64_f16(float16x4_t __a) { |
| 6647 | return (float64x1_t)__a; } |
| 6648 | __ai float64x1_t vreinterpret_f64_f32(float32x2_t __a) { |
| 6649 | return (float64x1_t)__a; } |
| 6650 | __ai float64x1_t vreinterpret_f64_p8(poly8x8_t __a) { |
| 6651 | return (float64x1_t)__a; } |
| 6652 | __ai float64x1_t vreinterpret_f64_p16(poly16x4_t __a) { |
| 6653 | return (float64x1_t)__a; } |
| 6654 | __ai float64x1_t vreinterpret_f64_p64(poly64x1_t __a) { |
| 6655 | return (float64x1_t)__a; } |
| 6656 | __ai poly8x8_t vreinterpret_p8_f64(float64x1_t __a) { |
| 6657 | return (poly8x8_t)__a; } |
| 6658 | __ai poly8x8_t vreinterpret_p8_p64(poly64x1_t __a) { |
| 6659 | return (poly8x8_t)__a; } |
| 6660 | __ai poly16x4_t vreinterpret_p16_f64(float64x1_t __a) { |
| 6661 | return (poly16x4_t)__a; } |
| 6662 | __ai poly16x4_t vreinterpret_p16_p64(poly64x1_t __a) { |
| 6663 | return (poly16x4_t)__a; } |
| 6664 | __ai poly64x1_t vreinterpret_p64_s8(int8x8_t __a) { |
| 6665 | return (poly64x1_t)__a; } |
| 6666 | __ai poly64x1_t vreinterpret_p64_s16(int16x4_t __a) { |
| 6667 | return (poly64x1_t)__a; } |
| 6668 | __ai poly64x1_t vreinterpret_p64_s32(int32x2_t __a) { |
| 6669 | return (poly64x1_t)__a; } |
| 6670 | __ai poly64x1_t vreinterpret_p64_s64(int64x1_t __a) { |
| 6671 | return (poly64x1_t)__a; } |
| 6672 | __ai poly64x1_t vreinterpret_p64_u8(uint8x8_t __a) { |
| 6673 | return (poly64x1_t)__a; } |
| 6674 | __ai poly64x1_t vreinterpret_p64_u16(uint16x4_t __a) { |
| 6675 | return (poly64x1_t)__a; } |
| 6676 | __ai poly64x1_t vreinterpret_p64_u32(uint32x2_t __a) { |
| 6677 | return (poly64x1_t)__a; } |
| 6678 | __ai poly64x1_t vreinterpret_p64_u64(uint64x1_t __a) { |
| 6679 | return (poly64x1_t)__a; } |
| 6680 | __ai poly64x1_t vreinterpret_p64_f16(float16x4_t __a) { |
| 6681 | return (poly64x1_t)__a; } |
| 6682 | __ai poly64x1_t vreinterpret_p64_f32(float32x2_t __a) { |
| 6683 | return (poly64x1_t)__a; } |
| 6684 | __ai poly64x1_t vreinterpret_p64_f64(float64x1_t __a) { |
| 6685 | return (poly64x1_t)__a; } |
| 6686 | __ai poly64x1_t vreinterpret_p64_p8(poly8x8_t __a) { |
| 6687 | return (poly64x1_t)__a; } |
| 6688 | __ai poly64x1_t vreinterpret_p64_p16(poly16x4_t __a) { |
| 6689 | return (poly64x1_t)__a; } |
| 6690 | __ai int8x16_t vreinterpretq_s8_f64(float64x2_t __a) { |
| 6691 | return (int8x16_t)__a; } |
| 6692 | __ai int8x16_t vreinterpretq_s8_p64(poly64x2_t __a) { |
| 6693 | return (int8x16_t)__a; } |
| 6694 | __ai int16x8_t vreinterpretq_s16_f64(float64x2_t __a) { |
| 6695 | return (int16x8_t)__a; } |
| 6696 | __ai int16x8_t vreinterpretq_s16_p64(poly64x2_t __a) { |
| 6697 | return (int16x8_t)__a; } |
| 6698 | __ai int32x4_t vreinterpretq_s32_f64(float64x2_t __a) { |
| 6699 | return (int32x4_t)__a; } |
| 6700 | __ai int32x4_t vreinterpretq_s32_p64(poly64x2_t __a) { |
| 6701 | return (int32x4_t)__a; } |
| 6702 | __ai int64x2_t vreinterpretq_s64_f64(float64x2_t __a) { |
| 6703 | return (int64x2_t)__a; } |
| 6704 | __ai int64x2_t vreinterpretq_s64_p64(poly64x2_t __a) { |
| 6705 | return (int64x2_t)__a; } |
| 6706 | __ai uint8x16_t vreinterpretq_u8_f64(float64x2_t __a) { |
| 6707 | return (uint8x16_t)__a; } |
| 6708 | __ai uint8x16_t vreinterpretq_u8_p64(poly64x2_t __a) { |
| 6709 | return (uint8x16_t)__a; } |
| 6710 | __ai uint16x8_t vreinterpretq_u16_f64(float64x2_t __a) { |
| 6711 | return (uint16x8_t)__a; } |
| 6712 | __ai uint16x8_t vreinterpretq_u16_p64(poly64x2_t __a) { |
| 6713 | return (uint16x8_t)__a; } |
| 6714 | __ai uint32x4_t vreinterpretq_u32_f64(float64x2_t __a) { |
| 6715 | return (uint32x4_t)__a; } |
| 6716 | __ai uint32x4_t vreinterpretq_u32_p64(poly64x2_t __a) { |
| 6717 | return (uint32x4_t)__a; } |
| 6718 | __ai uint64x2_t vreinterpretq_u64_f64(float64x2_t __a) { |
| 6719 | return (uint64x2_t)__a; } |
| 6720 | __ai uint64x2_t vreinterpretq_u64_p64(poly64x2_t __a) { |
| 6721 | return (uint64x2_t)__a; } |
| 6722 | __ai float16x8_t vreinterpretq_f16_f64(float64x2_t __a) { |
| 6723 | return (float16x8_t)__a; } |
| 6724 | __ai float16x8_t vreinterpretq_f16_p64(poly64x2_t __a) { |
| 6725 | return (float16x8_t)__a; } |
| 6726 | __ai float32x4_t vreinterpretq_f32_f64(float64x2_t __a) { |
| 6727 | return (float32x4_t)__a; } |
| 6728 | __ai float32x4_t vreinterpretq_f32_p64(poly64x2_t __a) { |
| 6729 | return (float32x4_t)__a; } |
| 6730 | __ai float64x2_t vreinterpretq_f64_s8(int8x16_t __a) { |
| 6731 | return (float64x2_t)__a; } |
| 6732 | __ai float64x2_t vreinterpretq_f64_s16(int16x8_t __a) { |
| 6733 | return (float64x2_t)__a; } |
| 6734 | __ai float64x2_t vreinterpretq_f64_s32(int32x4_t __a) { |
| 6735 | return (float64x2_t)__a; } |
| 6736 | __ai float64x2_t vreinterpretq_f64_s64(int64x2_t __a) { |
| 6737 | return (float64x2_t)__a; } |
| 6738 | __ai float64x2_t vreinterpretq_f64_u8(uint8x16_t __a) { |
| 6739 | return (float64x2_t)__a; } |
| 6740 | __ai float64x2_t vreinterpretq_f64_u16(uint16x8_t __a) { |
| 6741 | return (float64x2_t)__a; } |
| 6742 | __ai float64x2_t vreinterpretq_f64_u32(uint32x4_t __a) { |
| 6743 | return (float64x2_t)__a; } |
| 6744 | __ai float64x2_t vreinterpretq_f64_u64(uint64x2_t __a) { |
| 6745 | return (float64x2_t)__a; } |
| 6746 | __ai float64x2_t vreinterpretq_f64_f16(float16x8_t __a) { |
| 6747 | return (float64x2_t)__a; } |
| 6748 | __ai float64x2_t vreinterpretq_f64_f32(float32x4_t __a) { |
| 6749 | return (float64x2_t)__a; } |
| 6750 | __ai float64x2_t vreinterpretq_f64_p8(poly8x16_t __a) { |
| 6751 | return (float64x2_t)__a; } |
| 6752 | __ai float64x2_t vreinterpretq_f64_p16(poly16x8_t __a) { |
| 6753 | return (float64x2_t)__a; } |
| 6754 | __ai float64x2_t vreinterpretq_f64_p64(poly64x2_t __a) { |
| 6755 | return (float64x2_t)__a; } |
| 6756 | __ai poly8x16_t vreinterpretq_p8_f64(float64x2_t __a) { |
| 6757 | return (poly8x16_t)__a; } |
| 6758 | __ai poly8x16_t vreinterpretq_p8_p64(poly64x2_t __a) { |
| 6759 | return (poly8x16_t)__a; } |
| 6760 | __ai poly16x8_t vreinterpretq_p16_f64(float64x2_t __a) { |
| 6761 | return (poly16x8_t)__a; } |
| 6762 | __ai poly16x8_t vreinterpretq_p16_p64(poly64x2_t __a) { |
| 6763 | return (poly16x8_t)__a; } |
| 6764 | __ai poly64x2_t vreinterpretq_p64_s8(int8x16_t __a) { |
| 6765 | return (poly64x2_t)__a; } |
| 6766 | __ai poly64x2_t vreinterpretq_p64_s16(int16x8_t __a) { |
| 6767 | return (poly64x2_t)__a; } |
| 6768 | __ai poly64x2_t vreinterpretq_p64_s32(int32x4_t __a) { |
| 6769 | return (poly64x2_t)__a; } |
| 6770 | __ai poly64x2_t vreinterpretq_p64_s64(int64x2_t __a) { |
| 6771 | return (poly64x2_t)__a; } |
| 6772 | __ai poly64x2_t vreinterpretq_p64_u8(uint8x16_t __a) { |
| 6773 | return (poly64x2_t)__a; } |
| 6774 | __ai poly64x2_t vreinterpretq_p64_u16(uint16x8_t __a) { |
| 6775 | return (poly64x2_t)__a; } |
| 6776 | __ai poly64x2_t vreinterpretq_p64_u32(uint32x4_t __a) { |
| 6777 | return (poly64x2_t)__a; } |
| 6778 | __ai poly64x2_t vreinterpretq_p64_u64(uint64x2_t __a) { |
| 6779 | return (poly64x2_t)__a; } |
| 6780 | __ai poly64x2_t vreinterpretq_p64_f16(float16x8_t __a) { |
| 6781 | return (poly64x2_t)__a; } |
| 6782 | __ai poly64x2_t vreinterpretq_p64_f32(float32x4_t __a) { |
| 6783 | return (poly64x2_t)__a; } |
| 6784 | __ai poly64x2_t vreinterpretq_p64_f64(float64x2_t __a) { |
| 6785 | return (poly64x2_t)__a; } |
| 6786 | __ai poly64x2_t vreinterpretq_p64_p8(poly8x16_t __a) { |
| 6787 | return (poly64x2_t)__a; } |
| 6788 | __ai poly64x2_t vreinterpretq_p64_p16(poly16x8_t __a) { |
| 6789 | return (poly64x2_t)__a; } |
| 6790 | |
| 6791 | #define vrshrn_high_n_s16(a, b, __c) __extension__ ({ \ |
| 6792 | int8x8_t __a = (a); int16x8_t __b = (b); \ |
| 6793 | (int8x16_t)vcombine_s16(__a, vrshrn_n_s16(__b, __c)); }) |
| 6794 | #define vrshrn_high_n_s32(a, b, __c) __extension__ ({ \ |
| 6795 | int16x4_t __a = (a); int32x4_t __b = (b); \ |
| 6796 | (int16x8_t)vcombine_s32(__a, vrshrn_n_s32(__b, __c)); }) |
| 6797 | #define vrshrn_high_n_s64(a, b, __c) __extension__ ({ \ |
| 6798 | int32x2_t __a = (a); int64x2_t __b = (b); \ |
| 6799 | (int32x4_t)vcombine_s64(__a, vrshrn_n_s64(__b, __c)); }) |
| 6800 | #define vrshrn_high_n_u16(a, b, __c) __extension__ ({ \ |
| 6801 | uint8x8_t __a = (a); uint16x8_t __b = (b); \ |
| 6802 | (uint8x16_t)vcombine_u16(__a, vrshrn_n_u16(__b, __c)); }) |
| 6803 | #define vrshrn_high_n_u32(a, b, __c) __extension__ ({ \ |
| 6804 | uint16x4_t __a = (a); uint32x4_t __b = (b); \ |
| 6805 | (uint16x8_t)vcombine_u32(__a, vrshrn_n_u32(__b, __c)); }) |
| 6806 | #define vrshrn_high_n_u64(a, b, __c) __extension__ ({ \ |
| 6807 | uint32x2_t __a = (a); uint64x2_t __b = (b); \ |
| 6808 | (uint32x4_t)vcombine_u64(__a, vrshrn_n_u64(__b, __c)); }) |
| 6809 | |
| 6810 | __ai float32_t vabds_f32(float32_t __a, float32_t __b) { |
| 6811 | return (float32_t)__builtin_neon_vabds_f32(__a, __b); } |
| 6812 | __ai float64_t vabdd_f64(float64_t __a, float64_t __b) { |
| 6813 | return (float64_t)__builtin_neon_vabdd_f64(__a, __b); } |
| 6814 | |
| 6815 | __ai int64_t vabsd_s64(int64_t __a) { |
| 6816 | return (int64_t)__builtin_neon_vabsd_s64(__a); } |
| 6817 | |
| 6818 | __ai int64_t vaddd_s64(int64_t __a, int64_t __b) { |
| 6819 | return (int64_t)__builtin_neon_vaddd_s64(__a, __b); } |
| 6820 | __ai uint64_t vaddd_u64(uint64_t __a, uint64_t __b) { |
| 6821 | return (uint64_t)__builtin_neon_vaddd_u64(__a, __b); } |
| 6822 | |
| 6823 | __ai float32_t vpadds_f32(float32x2_t __a) { |
| 6824 | return (float32_t)__builtin_neon_vpadds_f32(__a); } |
| 6825 | __ai int64_t vpaddd_s64(int64x2_t __a) { |
| 6826 | return (int64_t)__builtin_neon_vpaddd_s64(__a); } |
| 6827 | __ai float64_t vpaddd_f64(float64x2_t __a) { |
| 6828 | return (float64_t)__builtin_neon_vpaddd_f64(__a); } |
| 6829 | __ai uint64_t vpaddd_u64(uint64x2_t __a) { |
| 6830 | return (uint64_t)__builtin_neon_vpaddd_u64((int64x2_t)__a); } |
| 6831 | |
| 6832 | __ai int64_t vceqd_s64(int64_t __a, int64_t __b) { |
| 6833 | return (int64_t)__builtin_neon_vceqd_s64(__a, __b); } |
| 6834 | __ai uint64_t vceqd_u64(uint64_t __a, uint64_t __b) { |
| 6835 | return (uint64_t)__builtin_neon_vceqd_u64(__a, __b); } |
| 6836 | |
| 6837 | __ai int64_t vceqzd_s64(int64_t __a) { |
| 6838 | return (int64_t)__builtin_neon_vceqzd_s64(__a); } |
| 6839 | __ai uint64_t vceqzd_u64(uint64_t __a) { |
| 6840 | return (uint64_t)__builtin_neon_vceqzd_u64(__a); } |
| 6841 | |
| 6842 | __ai int64_t vcged_s64(int64_t __a, int64_t __b) { |
| 6843 | return (int64_t)__builtin_neon_vcged_s64(__a, __b); } |
| 6844 | |
| 6845 | __ai int64_t vcgezd_s64(int64_t __a) { |
| 6846 | return (int64_t)__builtin_neon_vcgezd_s64(__a); } |
| 6847 | |
| 6848 | __ai int64_t vcgtd_s64(int64_t __a, int64_t __b) { |
| 6849 | return (int64_t)__builtin_neon_vcgtd_s64(__a, __b); } |
| 6850 | |
| 6851 | __ai int64_t vcgtzd_s64(int64_t __a) { |
| 6852 | return (int64_t)__builtin_neon_vcgtzd_s64(__a); } |
| 6853 | |
| 6854 | __ai uint64_t vcgtd_u64(uint64_t __a, uint64_t __b) { |
| 6855 | return (uint64_t)__builtin_neon_vcgtd_u64(__a, __b); } |
| 6856 | |
| 6857 | __ai uint64_t vcged_u64(uint64_t __a, uint64_t __b) { |
| 6858 | return (uint64_t)__builtin_neon_vcged_u64(__a, __b); } |
| 6859 | |
| 6860 | __ai int64_t vcled_s64(int64_t __a, int64_t __b) { |
| 6861 | return (int64_t)__builtin_neon_vcled_s64(__a, __b); } |
| 6862 | __ai uint64_t vcled_u64(uint64_t __a, uint64_t __b) { |
| 6863 | return (uint64_t)__builtin_neon_vcled_u64(__a, __b); } |
| 6864 | |
| 6865 | __ai int64_t vclezd_s64(int64_t __a) { |
| 6866 | return (int64_t)__builtin_neon_vclezd_s64(__a); } |
| 6867 | |
| 6868 | __ai int64_t vcltd_s64(int64_t __a, int64_t __b) { |
| 6869 | return (int64_t)__builtin_neon_vcltd_s64(__a, __b); } |
| 6870 | __ai uint64_t vcltd_u64(uint64_t __a, uint64_t __b) { |
| 6871 | return (uint64_t)__builtin_neon_vcltd_u64(__a, __b); } |
| 6872 | |
| 6873 | __ai int64_t vcltzd_s64(int64_t __a) { |
| 6874 | return (int64_t)__builtin_neon_vcltzd_s64(__a); } |
| 6875 | |
| 6876 | __ai int64_t vtstd_s64(int64_t __a, int64_t __b) { |
| 6877 | return (int64_t)__builtin_neon_vtstd_s64(__a, __b); } |
| 6878 | __ai uint64_t vtstd_u64(uint64_t __a, uint64_t __b) { |
| 6879 | return (uint64_t)__builtin_neon_vtstd_u64(__a, __b); } |
| 6880 | |
| 6881 | __ai uint32_t vcages_f32(float32_t __a, float32_t __b) { |
| 6882 | return (uint32_t)__builtin_neon_vcages_f32(__a, __b); } |
| 6883 | __ai uint64_t vcaged_f64(float64_t __a, float64_t __b) { |
| 6884 | return (uint64_t)__builtin_neon_vcaged_f64(__a, __b); } |
| 6885 | |
| 6886 | __ai uint32_t vcagts_f32(float32_t __a, float32_t __b) { |
| 6887 | return (uint32_t)__builtin_neon_vcagts_f32(__a, __b); } |
| 6888 | __ai uint64_t vcagtd_f64(float64_t __a, float64_t __b) { |
| 6889 | return (uint64_t)__builtin_neon_vcagtd_f64(__a, __b); } |
| 6890 | |
| 6891 | __ai uint32_t vcales_f32(float32_t __a, float32_t __b) { |
| 6892 | return (uint32_t)__builtin_neon_vcales_f32(__a, __b); } |
| 6893 | __ai uint64_t vcaled_f64(float64_t __a, float64_t __b) { |
| 6894 | return (uint64_t)__builtin_neon_vcaled_f64(__a, __b); } |
| 6895 | |
| 6896 | __ai uint32_t vcalts_f32(float32_t __a, float32_t __b) { |
| 6897 | return (uint32_t)__builtin_neon_vcalts_f32(__a, __b); } |
| 6898 | __ai uint64_t vcaltd_f64(float64_t __a, float64_t __b) { |
| 6899 | return (uint64_t)__builtin_neon_vcaltd_f64(__a, __b); } |
| 6900 | |
| 6901 | __ai uint32_t vceqs_f32(float32_t __a, float32_t __b) { |
| 6902 | return (uint32_t)__builtin_neon_vceqs_f32(__a, __b); } |
| 6903 | __ai uint64_t vceqd_f64(float64_t __a, float64_t __b) { |
| 6904 | return (uint64_t)__builtin_neon_vceqd_f64(__a, __b); } |
| 6905 | |
| 6906 | __ai uint32_t vceqzs_f32(float32_t __a) { |
| 6907 | return (uint32_t)__builtin_neon_vceqzs_f32(__a); } |
| 6908 | __ai uint64_t vceqzd_f64(float64_t __a) { |
| 6909 | return (uint64_t)__builtin_neon_vceqzd_f64(__a); } |
| 6910 | |
| 6911 | __ai uint32_t vcges_f32(float32_t __a, float32_t __b) { |
| 6912 | return (uint32_t)__builtin_neon_vcges_f32(__a, __b); } |
| 6913 | __ai uint64_t vcged_f64(float64_t __a, float64_t __b) { |
| 6914 | return (uint64_t)__builtin_neon_vcged_f64(__a, __b); } |
| 6915 | |
| 6916 | __ai uint32_t vcgezs_f32(float32_t __a) { |
| 6917 | return (uint32_t)__builtin_neon_vcgezs_f32(__a); } |
| 6918 | __ai uint64_t vcgezd_f64(float64_t __a) { |
| 6919 | return (uint64_t)__builtin_neon_vcgezd_f64(__a); } |
| 6920 | |
| 6921 | __ai uint32_t vcgts_f32(float32_t __a, float32_t __b) { |
| 6922 | return (uint32_t)__builtin_neon_vcgts_f32(__a, __b); } |
| 6923 | __ai uint64_t vcgtd_f64(float64_t __a, float64_t __b) { |
| 6924 | return (uint64_t)__builtin_neon_vcgtd_f64(__a, __b); } |
| 6925 | |
| 6926 | __ai uint32_t vcgtzs_f32(float32_t __a) { |
| 6927 | return (uint32_t)__builtin_neon_vcgtzs_f32(__a); } |
| 6928 | __ai uint64_t vcgtzd_f64(float64_t __a) { |
| 6929 | return (uint64_t)__builtin_neon_vcgtzd_f64(__a); } |
| 6930 | |
| 6931 | __ai uint32_t vcles_f32(float32_t __a, float32_t __b) { |
| 6932 | return (uint32_t)__builtin_neon_vcles_f32(__a, __b); } |
| 6933 | __ai uint64_t vcled_f64(float64_t __a, float64_t __b) { |
| 6934 | return (uint64_t)__builtin_neon_vcled_f64(__a, __b); } |
| 6935 | |
| 6936 | __ai uint32_t vclezs_f32(float32_t __a) { |
| 6937 | return (uint32_t)__builtin_neon_vclezs_f32(__a); } |
| 6938 | __ai uint64_t vclezd_f64(float64_t __a) { |
| 6939 | return (uint64_t)__builtin_neon_vclezd_f64(__a); } |
| 6940 | |
| 6941 | __ai uint32_t vclts_f32(float32_t __a, float32_t __b) { |
| 6942 | return (uint32_t)__builtin_neon_vclts_f32(__a, __b); } |
| 6943 | __ai uint64_t vcltd_f64(float64_t __a, float64_t __b) { |
| 6944 | return (uint64_t)__builtin_neon_vcltd_f64(__a, __b); } |
| 6945 | |
| 6946 | __ai uint32_t vcltzs_f32(float32_t __a) { |
| 6947 | return (uint32_t)__builtin_neon_vcltzs_f32(__a); } |
| 6948 | __ai uint64_t vcltzd_f64(float64_t __a) { |
| 6949 | return (uint64_t)__builtin_neon_vcltzd_f64(__a); } |
| 6950 | |
| 6951 | __ai int64_t vcvtad_s64_f64(float64_t __a) { |
| 6952 | return (int64_t)__builtin_neon_vcvtad_s64_f64(__a); } |
| 6953 | |
| 6954 | __ai int32_t vcvtas_s32_f32(float32_t __a) { |
| 6955 | return (int32_t)__builtin_neon_vcvtas_s32_f32(__a); } |
| 6956 | |
| 6957 | __ai uint64_t vcvtad_u64_f64(float64_t __a) { |
| 6958 | return (uint64_t)__builtin_neon_vcvtad_u64_f64(__a); } |
| 6959 | |
| 6960 | __ai uint32_t vcvtas_u32_f32(float32_t __a) { |
| 6961 | return (uint32_t)__builtin_neon_vcvtas_u32_f32(__a); } |
| 6962 | |
| 6963 | __ai int64_t vcvtmd_s64_f64(float64_t __a) { |
| 6964 | return (int64_t)__builtin_neon_vcvtmd_s64_f64(__a); } |
| 6965 | |
| 6966 | __ai int32_t vcvtms_s32_f32(float32_t __a) { |
| 6967 | return (int32_t)__builtin_neon_vcvtms_s32_f32(__a); } |
| 6968 | |
| 6969 | __ai uint64_t vcvtmd_u64_f64(float64_t __a) { |
| 6970 | return (uint64_t)__builtin_neon_vcvtmd_u64_f64(__a); } |
| 6971 | |
| 6972 | __ai uint32_t vcvtms_u32_f32(float32_t __a) { |
| 6973 | return (uint32_t)__builtin_neon_vcvtms_u32_f32(__a); } |
| 6974 | |
| 6975 | __ai int64_t vcvtnd_s64_f64(float64_t __a) { |
| 6976 | return (int64_t)__builtin_neon_vcvtnd_s64_f64(__a); } |
| 6977 | |
| 6978 | __ai int32_t vcvtns_s32_f32(float32_t __a) { |
| 6979 | return (int32_t)__builtin_neon_vcvtns_s32_f32(__a); } |
| 6980 | |
| 6981 | __ai uint64_t vcvtnd_u64_f64(float64_t __a) { |
| 6982 | return (uint64_t)__builtin_neon_vcvtnd_u64_f64(__a); } |
| 6983 | |
| 6984 | __ai uint32_t vcvtns_u32_f32(float32_t __a) { |
| 6985 | return (uint32_t)__builtin_neon_vcvtns_u32_f32(__a); } |
| 6986 | |
| 6987 | __ai int64_t vcvtpd_s64_f64(float64_t __a) { |
| 6988 | return (int64_t)__builtin_neon_vcvtpd_s64_f64(__a); } |
| 6989 | |
| 6990 | __ai int32_t vcvtps_s32_f32(float32_t __a) { |
| 6991 | return (int32_t)__builtin_neon_vcvtps_s32_f32(__a); } |
| 6992 | |
| 6993 | __ai uint64_t vcvtpd_u64_f64(float64_t __a) { |
| 6994 | return (uint64_t)__builtin_neon_vcvtpd_u64_f64(__a); } |
| 6995 | |
| 6996 | __ai uint32_t vcvtps_u32_f32(float32_t __a) { |
| 6997 | return (uint32_t)__builtin_neon_vcvtps_u32_f32(__a); } |
| 6998 | |
| 6999 | __ai float32_t vcvtxd_f32_f64(float64_t __a) { |
| 7000 | return (float32_t)__builtin_neon_vcvtxd_f32_f64(__a); } |
| 7001 | |
| 7002 | __ai int64_t vcvtd_s64_f64(float64_t __a) { |
| 7003 | return (int64_t)__builtin_neon_vcvtd_s64_f64(__a); } |
| 7004 | |
| 7005 | __ai int32_t vcvts_s32_f32(float32_t __a) { |
| 7006 | return (int32_t)__builtin_neon_vcvts_s32_f32(__a); } |
| 7007 | |
| 7008 | #define vcvts_n_s32_f32(a, __b) __extension__ ({ \ |
| 7009 | float32_t __a = (a); \ |
| 7010 | (int32_t)__builtin_neon_vcvts_n_s32_f32(__a, __b); }) |
| 7011 | |
| 7012 | #define vcvtd_n_s64_f64(a, __b) __extension__ ({ \ |
| 7013 | float64_t __a = (a); \ |
| 7014 | (int64_t)__builtin_neon_vcvtd_n_s64_f64(__a, __b); }) |
| 7015 | |
| 7016 | __ai uint64_t vcvtd_u64_f64(float64_t __a) { |
| 7017 | return (uint64_t)__builtin_neon_vcvtd_u64_f64(__a); } |
| 7018 | |
| 7019 | __ai uint32_t vcvts_u32_f32(float32_t __a) { |
| 7020 | return (uint32_t)__builtin_neon_vcvts_u32_f32(__a); } |
| 7021 | |
| 7022 | #define vcvts_n_u32_f32(a, __b) __extension__ ({ \ |
| 7023 | float32_t __a = (a); \ |
| 7024 | (uint32_t)__builtin_neon_vcvts_n_u32_f32(__a, __b); }) |
| 7025 | |
| 7026 | #define vcvtd_n_u64_f64(a, __b) __extension__ ({ \ |
| 7027 | float64_t __a = (a); \ |
| 7028 | (uint64_t)__builtin_neon_vcvtd_n_u64_f64(__a, __b); }) |
| 7029 | |
| 7030 | __ai float32_t vpmaxnms_f32(float32x2_t __a) { |
| 7031 | return (float32_t)__builtin_neon_vpmaxnms_f32(__a); } |
| 7032 | __ai float64_t vpmaxnmqd_f64(float64x2_t __a) { |
| 7033 | return (float64_t)__builtin_neon_vpmaxnmqd_f64(__a); } |
| 7034 | |
| 7035 | __ai float32_t vpmaxs_f32(float32x2_t __a) { |
| 7036 | return (float32_t)__builtin_neon_vpmaxs_f32(__a); } |
| 7037 | __ai float64_t vpmaxqd_f64(float64x2_t __a) { |
| 7038 | return (float64_t)__builtin_neon_vpmaxqd_f64(__a); } |
| 7039 | |
| 7040 | __ai float32_t vpminnms_f32(float32x2_t __a) { |
| 7041 | return (float32_t)__builtin_neon_vpminnms_f32(__a); } |
| 7042 | __ai float64_t vpminnmqd_f64(float64x2_t __a) { |
| 7043 | return (float64_t)__builtin_neon_vpminnmqd_f64(__a); } |
| 7044 | |
| 7045 | __ai float32_t vpmins_f32(float32x2_t __a) { |
| 7046 | return (float32_t)__builtin_neon_vpmins_f32(__a); } |
| 7047 | __ai float64_t vpminqd_f64(float64x2_t __a) { |
| 7048 | return (float64_t)__builtin_neon_vpminqd_f64(__a); } |
| 7049 | |
| 7050 | #define vfmas_lane_f32(a, b, c, __d) __extension__ ({ \ |
| 7051 | float32_t __a = (a); float32_t __b = (b); float32x2_t __c = (c); \ |
| 7052 | (float32_t)__builtin_neon_vfmas_lane_f32(__a, __b, __c, __d); }) |
| 7053 | #define vfmad_lane_f64(a, b, c, __d) __extension__ ({ \ |
| 7054 | float64_t __a = (a); float64_t __b = (b); float64x1_t __c = (c); \ |
| 7055 | (float64_t)__builtin_neon_vfmad_lane_f64(__a, __b, __c, __d); }) |
| 7056 | |
| 7057 | #define vfmas_laneq_f32(a, b, c, __d) __extension__ ({ \ |
| 7058 | float32_t __a = (a); float32_t __b = (b); float32x4_t __c = (c); \ |
| 7059 | (float32_t)__builtin_neon_vfmas_laneq_f32(__a, __b, __c, __d); }) |
| 7060 | #define vfmad_laneq_f64(a, b, c, __d) __extension__ ({ \ |
| 7061 | float64_t __a = (a); float64_t __b = (b); float64x2_t __c = (c); \ |
| 7062 | (float64_t)__builtin_neon_vfmad_laneq_f64(__a, __b, __c, __d); }) |
| 7063 | |
| 7064 | #define vfmss_lane_f32(a, b, c, __d) __extension__ ({ \ |
| 7065 | float32_t __a = (a); float32_t __b = (b); float32x2_t __c = (c); \ |
| 7066 | float32_t __a1 = __a; \ |
| 7067 | float32_t __b1 = __b; \ |
| 7068 | float32x2_t __c1 = __c; \ |
| 7069 | vfmas_lane_f32(__a1, __b1, -__c1, __d); }) |
| 7070 | #define vfmsd_lane_f64(a, b, c, __d) __extension__ ({ \ |
| 7071 | float64_t __a = (a); float64_t __b = (b); float64x1_t __c = (c); \ |
| 7072 | float64_t __a1 = __a; \ |
| 7073 | float64_t __b1 = __b; \ |
| 7074 | float64x1_t __c1 = __c; \ |
| 7075 | vfmad_lane_f64(__a1, __b1, -__c1, __d); }) |
| 7076 | |
| 7077 | #define vfmss_laneq_f32(a, b, c, __d) __extension__ ({ \ |
| 7078 | float32_t __a = (a); float32_t __b = (b); float32x4_t __c = (c); \ |
| 7079 | float32_t __a1 = __a; \ |
| 7080 | float32_t __b1 = __b; \ |
| 7081 | float32x4_t __c1 = __c; \ |
| 7082 | vfmas_laneq_f32(__a1, __b1, -__c1, __d); }) |
| 7083 | #define vfmsd_laneq_f64(a, b, c, __d) __extension__ ({ \ |
| 7084 | float64_t __a = (a); float64_t __b = (b); float64x2_t __c = (c); \ |
| 7085 | float64_t __a1 = __a; \ |
| 7086 | float64_t __b1 = __b; \ |
| 7087 | float64x2_t __c1 = __c; \ |
| 7088 | vfmad_laneq_f64(__a1, __b1, -__c1, __d); }) |
| 7089 | |
| 7090 | __ai float32_t vmulxs_f32(float32_t __a, float32_t __b) { |
| 7091 | return (float32_t)__builtin_neon_vmulxs_f32(__a, __b); } |
| 7092 | __ai float64_t vmulxd_f64(float64_t __a, float64_t __b) { |
| 7093 | return (float64_t)__builtin_neon_vmulxd_f64(__a, __b); } |
| 7094 | |
| 7095 | #define vmulxs_lane_f32(a, b, __c) __extension__ ({ \ |
| 7096 | float32_t __a = (a); float32x2_t __b = (b); \ |
| 7097 | float32_t __d1 = vget_lane_f32(__b, __c);\ |
| 7098 | vmulxs_f32(__a, __d1); }) |
| 7099 | #define vmulxd_lane_f64(a, b, __c) __extension__ ({ \ |
| 7100 | float64_t __a = (a); float64x1_t __b = (b); \ |
| 7101 | float64_t __d1 = vget_lane_f64(__b, __c);\ |
| 7102 | vmulxd_f64(__a, __d1); }) |
| 7103 | |
| 7104 | #define vmulxs_laneq_f32(a, b, __c) __extension__ ({ \ |
| 7105 | float32_t __a = (a); float32x4_t __b = (b); \ |
| 7106 | float32_t __d1 = vgetq_lane_f32(__b, __c);\ |
| 7107 | vmulxs_f32(__a, __d1); }) |
| 7108 | #define vmulxd_laneq_f64(a, b, __c) __extension__ ({ \ |
| 7109 | float64_t __a = (a); float64x2_t __b = (b); \ |
| 7110 | float64_t __d1 = vgetq_lane_f64(__b, __c);\ |
| 7111 | vmulxd_f64(__a, __d1); }) |
| 7112 | |
| 7113 | #define vmuls_lane_f32(a, b, __c) __extension__ ({ \ |
| 7114 | float32_t __a = (a); float32x2_t __b = (b); \ |
| 7115 | float32_t __d1 = vget_lane_f32(__b, __c);\ |
| 7116 | __a * __d1; }) |
| 7117 | #define vmuld_lane_f64(a, b, __c) __extension__ ({ \ |
| 7118 | float64_t __a = (a); float64x1_t __b = (b); \ |
| 7119 | float64_t __d1 = vget_lane_f64(__b, __c);\ |
| 7120 | __a * __d1; }) |
| 7121 | |
| 7122 | #define vmuls_laneq_f32(a, b, __c) __extension__ ({ \ |
| 7123 | float32_t __a = (a); float32x4_t __b = (b); \ |
| 7124 | float32_t __d1 = vgetq_lane_f32(__b, __c);\ |
| 7125 | __a * __d1; }) |
| 7126 | #define vmuld_laneq_f64(a, b, __c) __extension__ ({ \ |
| 7127 | float64_t __a = (a); float64x2_t __b = (b); \ |
| 7128 | float64_t __d1 = vgetq_lane_f64(__b, __c);\ |
| 7129 | __a * __d1; }) |
| 7130 | |
| 7131 | __ai float32_t vrecpes_f32(float32_t __a) { |
| 7132 | return (float32_t)__builtin_neon_vrecpes_f32(__a); } |
| 7133 | __ai float64_t vrecped_f64(float64_t __a) { |
| 7134 | return (float64_t)__builtin_neon_vrecped_f64(__a); } |
| 7135 | |
| 7136 | __ai float32_t vrecpss_f32(float32_t __a, float32_t __b) { |
| 7137 | return (float32_t)__builtin_neon_vrecpss_f32(__a, __b); } |
| 7138 | __ai float64_t vrecpsd_f64(float64_t __a, float64_t __b) { |
| 7139 | return (float64_t)__builtin_neon_vrecpsd_f64(__a, __b); } |
| 7140 | |
| 7141 | __ai float32_t vrecpxs_f32(float32_t __a) { |
| 7142 | return (float32_t)__builtin_neon_vrecpxs_f32(__a); } |
| 7143 | __ai float64_t vrecpxd_f64(float64_t __a) { |
| 7144 | return (float64_t)__builtin_neon_vrecpxd_f64(__a); } |
| 7145 | |
| 7146 | __ai float32_t vrsqrtes_f32(float32_t __a) { |
| 7147 | return (float32_t)__builtin_neon_vrsqrtes_f32(__a); } |
| 7148 | __ai float64_t vrsqrted_f64(float64_t __a) { |
| 7149 | return (float64_t)__builtin_neon_vrsqrted_f64(__a); } |
| 7150 | |
| 7151 | __ai float32_t vrsqrtss_f32(float32_t __a, float32_t __b) { |
| 7152 | return (float32_t)__builtin_neon_vrsqrtss_f32(__a, __b); } |
| 7153 | __ai float64_t vrsqrtsd_f64(float64_t __a, float64_t __b) { |
| 7154 | return (float64_t)__builtin_neon_vrsqrtsd_f64(__a, __b); } |
| 7155 | |
| 7156 | #define vget_lane_f16(a, __b) __extension__ ({ \ |
| 7157 | float16x4_t __a = (a); \ |
| 7158 | int16x4_t __a1 = vreinterpret_s16_f16(__a);\ |
| 7159 | vget_lane_s16(__a1, __b); }) |
| 7160 | #define vgetq_lane_f16(a, __b) __extension__ ({ \ |
| 7161 | float16x8_t __a = (a); \ |
| 7162 | int16x8_t __a1 = vreinterpretq_s16_f16(__a);\ |
| 7163 | vgetq_lane_s16(__a1, __b); }) |
| 7164 | |
| 7165 | __ai int64_t vnegd_s64(int64_t __a) { |
| 7166 | return (int64_t)__builtin_neon_vnegd_s64(__a); } |
| 7167 | |
| 7168 | __ai int8_t vqaddb_s8(int8_t __a, int8_t __b) { |
| 7169 | return (int8_t)__builtin_neon_vqaddb_s8(__a, __b); } |
| 7170 | __ai int16_t vqaddh_s16(int16_t __a, int16_t __b) { |
| 7171 | return (int16_t)__builtin_neon_vqaddh_s16(__a, __b); } |
| 7172 | __ai int32_t vqadds_s32(int32_t __a, int32_t __b) { |
| 7173 | return (int32_t)__builtin_neon_vqadds_s32(__a, __b); } |
| 7174 | __ai int64_t vqaddd_s64(int64_t __a, int64_t __b) { |
| 7175 | return (int64_t)__builtin_neon_vqaddd_s64(__a, __b); } |
| 7176 | __ai uint8_t vqaddb_u8(uint8_t __a, uint8_t __b) { |
| 7177 | return (uint8_t)__builtin_neon_vqaddb_u8(__a, __b); } |
| 7178 | __ai uint16_t vqaddh_u16(uint16_t __a, uint16_t __b) { |
| 7179 | return (uint16_t)__builtin_neon_vqaddh_u16(__a, __b); } |
| 7180 | __ai uint32_t vqadds_u32(uint32_t __a, uint32_t __b) { |
| 7181 | return (uint32_t)__builtin_neon_vqadds_u32(__a, __b); } |
| 7182 | __ai uint64_t vqaddd_u64(uint64_t __a, uint64_t __b) { |
| 7183 | return (uint64_t)__builtin_neon_vqaddd_u64(__a, __b); } |
| 7184 | |
| 7185 | __ai int8_t vqrshlb_s8(int8_t __a, int8_t __b) { |
| 7186 | return (int8_t)__builtin_neon_vqrshlb_s8(__a, __b); } |
| 7187 | __ai int16_t vqrshlh_s16(int16_t __a, int16_t __b) { |
| 7188 | return (int16_t)__builtin_neon_vqrshlh_s16(__a, __b); } |
| 7189 | __ai int32_t vqrshls_s32(int32_t __a, int32_t __b) { |
| 7190 | return (int32_t)__builtin_neon_vqrshls_s32(__a, __b); } |
| 7191 | __ai int64_t vqrshld_s64(int64_t __a, int64_t __b) { |
| 7192 | return (int64_t)__builtin_neon_vqrshld_s64(__a, __b); } |
| 7193 | __ai uint8_t vqrshlb_u8(uint8_t __a, uint8_t __b) { |
| 7194 | return (uint8_t)__builtin_neon_vqrshlb_u8(__a, __b); } |
| 7195 | __ai uint16_t vqrshlh_u16(uint16_t __a, uint16_t __b) { |
| 7196 | return (uint16_t)__builtin_neon_vqrshlh_u16(__a, __b); } |
| 7197 | __ai uint32_t vqrshls_u32(uint32_t __a, uint32_t __b) { |
| 7198 | return (uint32_t)__builtin_neon_vqrshls_u32(__a, __b); } |
| 7199 | __ai uint64_t vqrshld_u64(uint64_t __a, uint64_t __b) { |
| 7200 | return (uint64_t)__builtin_neon_vqrshld_u64(__a, __b); } |
| 7201 | |
| 7202 | __ai int8_t vqshlb_s8(int8_t __a, int8_t __b) { |
| 7203 | return (int8_t)__builtin_neon_vqshlb_s8(__a, __b); } |
| 7204 | __ai int16_t vqshlh_s16(int16_t __a, int16_t __b) { |
| 7205 | return (int16_t)__builtin_neon_vqshlh_s16(__a, __b); } |
| 7206 | __ai int32_t vqshls_s32(int32_t __a, int32_t __b) { |
| 7207 | return (int32_t)__builtin_neon_vqshls_s32(__a, __b); } |
| 7208 | __ai int64_t vqshld_s64(int64_t __a, int64_t __b) { |
| 7209 | return (int64_t)__builtin_neon_vqshld_s64(__a, __b); } |
| 7210 | __ai uint8_t vqshlb_u8(uint8_t __a, uint8_t __b) { |
| 7211 | return (uint8_t)__builtin_neon_vqshlb_u8(__a, __b); } |
| 7212 | __ai uint16_t vqshlh_u16(uint16_t __a, uint16_t __b) { |
| 7213 | return (uint16_t)__builtin_neon_vqshlh_u16(__a, __b); } |
| 7214 | __ai uint32_t vqshls_u32(uint32_t __a, uint32_t __b) { |
| 7215 | return (uint32_t)__builtin_neon_vqshls_u32(__a, __b); } |
| 7216 | __ai uint64_t vqshld_u64(uint64_t __a, uint64_t __b) { |
| 7217 | return (uint64_t)__builtin_neon_vqshld_u64(__a, __b); } |
| 7218 | |
| 7219 | __ai int8_t vqsubb_s8(int8_t __a, int8_t __b) { |
| 7220 | return (int8_t)__builtin_neon_vqsubb_s8(__a, __b); } |
| 7221 | __ai int16_t vqsubh_s16(int16_t __a, int16_t __b) { |
| 7222 | return (int16_t)__builtin_neon_vqsubh_s16(__a, __b); } |
| 7223 | __ai int32_t vqsubs_s32(int32_t __a, int32_t __b) { |
| 7224 | return (int32_t)__builtin_neon_vqsubs_s32(__a, __b); } |
| 7225 | __ai int64_t vqsubd_s64(int64_t __a, int64_t __b) { |
| 7226 | return (int64_t)__builtin_neon_vqsubd_s64(__a, __b); } |
| 7227 | __ai uint8_t vqsubb_u8(uint8_t __a, uint8_t __b) { |
| 7228 | return (uint8_t)__builtin_neon_vqsubb_u8(__a, __b); } |
| 7229 | __ai uint16_t vqsubh_u16(uint16_t __a, uint16_t __b) { |
| 7230 | return (uint16_t)__builtin_neon_vqsubh_u16(__a, __b); } |
| 7231 | __ai uint32_t vqsubs_u32(uint32_t __a, uint32_t __b) { |
| 7232 | return (uint32_t)__builtin_neon_vqsubs_u32(__a, __b); } |
| 7233 | __ai uint64_t vqsubd_u64(uint64_t __a, uint64_t __b) { |
| 7234 | return (uint64_t)__builtin_neon_vqsubd_u64(__a, __b); } |
| 7235 | |
| 7236 | __ai int64_t vrshld_s64(int64_t __a, int64_t __b) { |
| 7237 | return (int64_t)__builtin_neon_vrshld_s64(__a, __b); } |
| 7238 | __ai uint64_t vrshld_u64(uint64_t __a, uint64_t __b) { |
| 7239 | return (uint64_t)__builtin_neon_vrshld_u64(__a, __b); } |
| 7240 | |
| 7241 | __ai float64_t vcvtd_f64_s64(int64_t __a) { |
| 7242 | return (float64_t)__builtin_neon_vcvtd_f64_s64(__a); } |
| 7243 | |
| 7244 | __ai float32_t vcvts_f32_s32(int32_t __a) { |
| 7245 | return (float32_t)__builtin_neon_vcvts_f32_s32(__a); } |
| 7246 | |
| 7247 | #define vcvts_n_f32_s32(a, __b) __extension__ ({ \ |
| 7248 | int32_t __a = (a); \ |
| 7249 | (float32_t)__builtin_neon_vcvts_n_f32_s32(__a, __b); }) |
| 7250 | #define vcvts_n_f32_u32(a, __b) __extension__ ({ \ |
| 7251 | uint32_t __a = (a); \ |
| 7252 | (float32_t)__builtin_neon_vcvts_n_f32_u32(__a, __b); }) |
| 7253 | |
| 7254 | #define vcvtd_n_f64_s64(a, __b) __extension__ ({ \ |
| 7255 | int64_t __a = (a); \ |
| 7256 | (float64_t)__builtin_neon_vcvtd_n_f64_s64(__a, __b); }) |
| 7257 | #define vcvtd_n_f64_u64(a, __b) __extension__ ({ \ |
| 7258 | uint64_t __a = (a); \ |
| 7259 | (float64_t)__builtin_neon_vcvtd_n_f64_u64(__a, __b); }) |
| 7260 | |
| 7261 | #define vset_lane_f16(a, b, __c) __extension__ ({ \ |
| 7262 | float16_t __a = (a); float16x4_t __b = (b); \ |
| 7263 | int16_t __a1 = (int16_t)__a;\ |
| 7264 | int16x4_t __b1 = vreinterpret_s16_f16(b);\ |
| 7265 | int16x4_t __b2 = vset_lane_s16(__a1, __b1, __c);\ |
| 7266 | vreinterpret_f16_s16(__b2); }) |
| 7267 | #define vsetq_lane_f16(a, b, __c) __extension__ ({ \ |
| 7268 | float16_t __a = (a); float16x8_t __b = (b); \ |
| 7269 | int16_t __a1 = (int16_t)__a;\ |
| 7270 | int16x8_t __b1 = vreinterpretq_s16_f16(b);\ |
| 7271 | int16x8_t __b2 = vsetq_lane_s16(__a1, __b1, __c);\ |
| 7272 | vreinterpretq_f16_s16(__b2); }) |
| 7273 | |
| 7274 | __ai int64_t vshld_s64(int64_t __a, int64_t __b) { |
| 7275 | return (int64_t)__builtin_neon_vshld_s64(__a, __b); } |
| 7276 | __ai uint64_t vshld_u64(uint64_t __a, uint64_t __b) { |
| 7277 | return (uint64_t)__builtin_neon_vshld_u64(__a, __b); } |
| 7278 | |
| 7279 | #define vshld_n_s64(a, __b) __extension__ ({ \ |
| 7280 | int64_t __a = (a); \ |
| 7281 | (int64_t)__builtin_neon_vshld_n_s64(__a, __b); }) |
| 7282 | #define vshld_n_u64(a, __b) __extension__ ({ \ |
| 7283 | uint64_t __a = (a); \ |
| 7284 | (uint64_t)__builtin_neon_vshld_n_u64(__a, __b); }) |
| 7285 | |
| 7286 | #define vslid_n_s64(a, b, __c) __extension__ ({ \ |
| 7287 | int64_t __a = (a); int64_t __b = (b); \ |
| 7288 | (int64_t)__builtin_neon_vslid_n_s64(__a, __b, __c); }) |
| 7289 | #define vslid_n_u64(a, b, __c) __extension__ ({ \ |
| 7290 | uint64_t __a = (a); uint64_t __b = (b); \ |
| 7291 | (uint64_t)__builtin_neon_vslid_n_u64(__a, __b, __c); }) |
| 7292 | |
| 7293 | __ai int8_t vqabsb_s8(int8_t __a) { |
| 7294 | return (int8_t)__builtin_neon_vqabsb_s8(__a); } |
| 7295 | __ai int16_t vqabsh_s16(int16_t __a) { |
| 7296 | return (int16_t)__builtin_neon_vqabsh_s16(__a); } |
| 7297 | __ai int32_t vqabss_s32(int32_t __a) { |
| 7298 | return (int32_t)__builtin_neon_vqabss_s32(__a); } |
| 7299 | __ai int64_t vqabsd_s64(int64_t __a) { |
| 7300 | return (int64_t)__builtin_neon_vqabsd_s64(__a); } |
| 7301 | |
| 7302 | __ai int32_t vqdmlalh_s16(int32_t __a, int16_t __b, int16_t __c) { |
| 7303 | return (int32_t)__builtin_neon_vqdmlalh_s16(__a, __b, __c); } |
| 7304 | __ai int64_t vqdmlals_s32(int64_t __a, int32_t __b, int32_t __c) { |
| 7305 | return (int64_t)__builtin_neon_vqdmlals_s32(__a, __b, __c); } |
| 7306 | |
| 7307 | #define vqdmlalh_lane_s16(a, b, c, __d) __extension__ ({ \ |
| 7308 | int32_t __a = (a); int16_t __b = (b); int16x4_t __c = (c); \ |
| 7309 | (int32_t)__builtin_neon_vqdmlalh_lane_s16(__a, __b, __c, __d); }) |
| 7310 | #define vqdmlals_lane_s32(a, b, c, __d) __extension__ ({ \ |
| 7311 | int64_t __a = (a); int32_t __b = (b); int32x2_t __c = (c); \ |
| 7312 | (int64_t)__builtin_neon_vqdmlals_lane_s32(__a, __b, __c, __d); }) |
| 7313 | |
| 7314 | #define vqdmlalh_laneq_s16(a, b, c, __d) __extension__ ({ \ |
| 7315 | int32_t __a = (a); int16_t __b = (b); int16x8_t __c = (c); \ |
| 7316 | (int32_t)__builtin_neon_vqdmlalh_laneq_s16(__a, __b, __c, __d); }) |
| 7317 | #define vqdmlals_laneq_s32(a, b, c, __d) __extension__ ({ \ |
| 7318 | int64_t __a = (a); int32_t __b = (b); int32x4_t __c = (c); \ |
| 7319 | (int64_t)__builtin_neon_vqdmlals_laneq_s32(__a, __b, __c, __d); }) |
| 7320 | |
| 7321 | __ai int32_t vqdmlslh_s16(int32_t __a, int16_t __b, int16_t __c) { |
| 7322 | return (int32_t)__builtin_neon_vqdmlslh_s16(__a, __b, __c); } |
| 7323 | __ai int64_t vqdmlsls_s32(int64_t __a, int32_t __b, int32_t __c) { |
| 7324 | return (int64_t)__builtin_neon_vqdmlsls_s32(__a, __b, __c); } |
| 7325 | |
| 7326 | #define vqdmlslh_lane_s16(a, b, c, __d) __extension__ ({ \ |
| 7327 | int32_t __a = (a); int16_t __b = (b); int16x4_t __c = (c); \ |
| 7328 | (int32_t)__builtin_neon_vqdmlslh_lane_s16(__a, __b, __c, __d); }) |
| 7329 | #define vqdmlsls_lane_s32(a, b, c, __d) __extension__ ({ \ |
| 7330 | int64_t __a = (a); int32_t __b = (b); int32x2_t __c = (c); \ |
| 7331 | (int64_t)__builtin_neon_vqdmlsls_lane_s32(__a, __b, __c, __d); }) |
| 7332 | |
| 7333 | #define vqdmlslh_laneq_s16(a, b, c, __d) __extension__ ({ \ |
| 7334 | int32_t __a = (a); int16_t __b = (b); int16x8_t __c = (c); \ |
| 7335 | (int32_t)__builtin_neon_vqdmlslh_laneq_s16(__a, __b, __c, __d); }) |
| 7336 | #define vqdmlsls_laneq_s32(a, b, c, __d) __extension__ ({ \ |
| 7337 | int64_t __a = (a); int32_t __b = (b); int32x4_t __c = (c); \ |
| 7338 | (int64_t)__builtin_neon_vqdmlsls_laneq_s32(__a, __b, __c, __d); }) |
| 7339 | |
| 7340 | __ai int16_t vqdmulhh_s16(int16_t __a, int16_t __b) { |
| 7341 | return (int16_t)__builtin_neon_vqdmulhh_s16(__a, __b); } |
| 7342 | __ai int32_t vqdmulhs_s32(int32_t __a, int32_t __b) { |
| 7343 | return (int32_t)__builtin_neon_vqdmulhs_s32(__a, __b); } |
| 7344 | |
| 7345 | #define vqdmulhh_lane_s16(a, b, __c) __extension__ ({ \ |
| 7346 | int16_t __a = (a); int16x4_t __b = (b); \ |
| 7347 | vqdmulhh_s16(__a, vget_lane_s16(__b, __c)); }) |
| 7348 | #define vqdmulhs_lane_s32(a, b, __c) __extension__ ({ \ |
| 7349 | int32_t __a = (a); int32x2_t __b = (b); \ |
| 7350 | vqdmulhs_s32(__a, vget_lane_s32(__b, __c)); }) |
| 7351 | |
| 7352 | #define vqdmulhh_laneq_s16(a, b, __c) __extension__ ({ \ |
| 7353 | int16_t __a = (a); int16x8_t __b = (b); \ |
| 7354 | vqdmulhh_s16(__a, vgetq_lane_s16(__b, __c)); }) |
| 7355 | #define vqdmulhs_laneq_s32(a, b, __c) __extension__ ({ \ |
| 7356 | int32_t __a = (a); int32x4_t __b = (b); \ |
| 7357 | vqdmulhs_s32(__a, vgetq_lane_s32(__b, __c)); }) |
| 7358 | |
| 7359 | __ai int32_t vqdmullh_s16(int16_t __a, int16_t __b) { |
| 7360 | return (int32_t)__builtin_neon_vqdmullh_s16(__a, __b); } |
| 7361 | __ai int64_t vqdmulls_s32(int32_t __a, int32_t __b) { |
| 7362 | return (int64_t)__builtin_neon_vqdmulls_s32(__a, __b); } |
| 7363 | |
| 7364 | #define vqdmullh_lane_s16(a, b, __c) __extension__ ({ \ |
| 7365 | int16_t __a = (a); int16x4_t __b = (b); \ |
| 7366 | vqdmullh_s16(__a, vget_lane_s16(b, __c)); }) |
| 7367 | #define vqdmulls_lane_s32(a, b, __c) __extension__ ({ \ |
| 7368 | int32_t __a = (a); int32x2_t __b = (b); \ |
| 7369 | vqdmulls_s32(__a, vget_lane_s32(b, __c)); }) |
| 7370 | |
| 7371 | #define vqdmullh_laneq_s16(a, b, __c) __extension__ ({ \ |
| 7372 | int16_t __a = (a); int16x8_t __b = (b); \ |
| 7373 | vqdmullh_s16(__a, vgetq_lane_s16(b, __c)); }) |
| 7374 | #define vqdmulls_laneq_s32(a, b, __c) __extension__ ({ \ |
| 7375 | int32_t __a = (a); int32x4_t __b = (b); \ |
| 7376 | vqdmulls_s32(__a, vgetq_lane_s32(b, __c)); }) |
| 7377 | |
| 7378 | __ai int8_t vqnegb_s8(int8_t __a) { |
| 7379 | return (int8_t)__builtin_neon_vqnegb_s8(__a); } |
| 7380 | __ai int16_t vqnegh_s16(int16_t __a) { |
| 7381 | return (int16_t)__builtin_neon_vqnegh_s16(__a); } |
| 7382 | __ai int32_t vqnegs_s32(int32_t __a) { |
| 7383 | return (int32_t)__builtin_neon_vqnegs_s32(__a); } |
| 7384 | __ai int64_t vqnegd_s64(int64_t __a) { |
| 7385 | return (int64_t)__builtin_neon_vqnegd_s64(__a); } |
| 7386 | |
| 7387 | __ai int16_t vqrdmulhh_s16(int16_t __a, int16_t __b) { |
| 7388 | return (int16_t)__builtin_neon_vqrdmulhh_s16(__a, __b); } |
| 7389 | __ai int32_t vqrdmulhs_s32(int32_t __a, int32_t __b) { |
| 7390 | return (int32_t)__builtin_neon_vqrdmulhs_s32(__a, __b); } |
| 7391 | |
| 7392 | #define vqrdmulhh_lane_s16(a, b, __c) __extension__ ({ \ |
| 7393 | int16_t __a = (a); int16x4_t __b = (b); \ |
| 7394 | vqrdmulhh_s16(__a, vget_lane_s16(__b, __c)); }) |
| 7395 | #define vqrdmulhs_lane_s32(a, b, __c) __extension__ ({ \ |
| 7396 | int32_t __a = (a); int32x2_t __b = (b); \ |
| 7397 | vqrdmulhs_s32(__a, vget_lane_s32(__b, __c)); }) |
| 7398 | |
| 7399 | #define vqrdmulhh_laneq_s16(a, b, __c) __extension__ ({ \ |
| 7400 | int16_t __a = (a); int16x8_t __b = (b); \ |
| 7401 | vqrdmulhh_s16(__a, vgetq_lane_s16(__b, __c)); }) |
| 7402 | #define vqrdmulhs_laneq_s32(a, b, __c) __extension__ ({ \ |
| 7403 | int32_t __a = (a); int32x4_t __b = (b); \ |
| 7404 | vqrdmulhs_s32(__a, vgetq_lane_s32(__b, __c)); }) |
| 7405 | |
| 7406 | #define vqrshrnh_n_s16(a, __b) __extension__ ({ \ |
| 7407 | int16_t __a = (a); \ |
| 7408 | (int8_t)__builtin_neon_vqrshrnh_n_s16(__a, __b); }) |
| 7409 | #define vqrshrns_n_s32(a, __b) __extension__ ({ \ |
| 7410 | int32_t __a = (a); \ |
| 7411 | (int16_t)__builtin_neon_vqrshrns_n_s32(__a, __b); }) |
| 7412 | #define vqrshrnd_n_s64(a, __b) __extension__ ({ \ |
| 7413 | int64_t __a = (a); \ |
| 7414 | (int32_t)__builtin_neon_vqrshrnd_n_s64(__a, __b); }) |
| 7415 | #define vqrshrnh_n_u16(a, __b) __extension__ ({ \ |
| 7416 | uint16_t __a = (a); \ |
| 7417 | (uint8_t)__builtin_neon_vqrshrnh_n_u16(__a, __b); }) |
| 7418 | #define vqrshrns_n_u32(a, __b) __extension__ ({ \ |
| 7419 | uint32_t __a = (a); \ |
| 7420 | (uint16_t)__builtin_neon_vqrshrns_n_u32(__a, __b); }) |
| 7421 | #define vqrshrnd_n_u64(a, __b) __extension__ ({ \ |
| 7422 | uint64_t __a = (a); \ |
| 7423 | (uint32_t)__builtin_neon_vqrshrnd_n_u64(__a, __b); }) |
| 7424 | |
| 7425 | #define vqrshrunh_n_s16(a, __b) __extension__ ({ \ |
| 7426 | int16_t __a = (a); \ |
| 7427 | (int8_t)__builtin_neon_vqrshrunh_n_s16(__a, __b); }) |
| 7428 | #define vqrshruns_n_s32(a, __b) __extension__ ({ \ |
| 7429 | int32_t __a = (a); \ |
| 7430 | (int16_t)__builtin_neon_vqrshruns_n_s32(__a, __b); }) |
| 7431 | #define vqrshrund_n_s64(a, __b) __extension__ ({ \ |
| 7432 | int64_t __a = (a); \ |
| 7433 | (int32_t)__builtin_neon_vqrshrund_n_s64(__a, __b); }) |
| 7434 | |
| 7435 | #define vqshlub_n_s8(a, __b) __extension__ ({ \ |
| 7436 | int8_t __a = (a); \ |
| 7437 | (int8_t)__builtin_neon_vqshlub_n_s8(__a, __b); }) |
| 7438 | #define vqshluh_n_s16(a, __b) __extension__ ({ \ |
| 7439 | int16_t __a = (a); \ |
| 7440 | (int16_t)__builtin_neon_vqshluh_n_s16(__a, __b); }) |
| 7441 | #define vqshlus_n_s32(a, __b) __extension__ ({ \ |
| 7442 | int32_t __a = (a); \ |
| 7443 | (int32_t)__builtin_neon_vqshlus_n_s32(__a, __b); }) |
| 7444 | #define vqshlud_n_s64(a, __b) __extension__ ({ \ |
| 7445 | int64_t __a = (a); \ |
| 7446 | (int64_t)__builtin_neon_vqshlud_n_s64(__a, __b); }) |
| 7447 | |
| 7448 | #define vqshlb_n_s8(a, __b) __extension__ ({ \ |
| 7449 | int8_t __a = (a); \ |
| 7450 | (int8_t)__builtin_neon_vqshlb_n_s8(__a, __b); }) |
| 7451 | #define vqshlh_n_s16(a, __b) __extension__ ({ \ |
| 7452 | int16_t __a = (a); \ |
| 7453 | (int16_t)__builtin_neon_vqshlh_n_s16(__a, __b); }) |
| 7454 | #define vqshls_n_s32(a, __b) __extension__ ({ \ |
| 7455 | int32_t __a = (a); \ |
| 7456 | (int32_t)__builtin_neon_vqshls_n_s32(__a, __b); }) |
| 7457 | #define vqshld_n_s64(a, __b) __extension__ ({ \ |
| 7458 | int64_t __a = (a); \ |
| 7459 | (int64_t)__builtin_neon_vqshld_n_s64(__a, __b); }) |
| 7460 | #define vqshlb_n_u8(a, __b) __extension__ ({ \ |
| 7461 | uint8_t __a = (a); \ |
| 7462 | (uint8_t)__builtin_neon_vqshlb_n_u8(__a, __b); }) |
| 7463 | #define vqshlh_n_u16(a, __b) __extension__ ({ \ |
| 7464 | uint16_t __a = (a); \ |
| 7465 | (uint16_t)__builtin_neon_vqshlh_n_u16(__a, __b); }) |
| 7466 | #define vqshls_n_u32(a, __b) __extension__ ({ \ |
| 7467 | uint32_t __a = (a); \ |
| 7468 | (uint32_t)__builtin_neon_vqshls_n_u32(__a, __b); }) |
| 7469 | #define vqshld_n_u64(a, __b) __extension__ ({ \ |
| 7470 | uint64_t __a = (a); \ |
| 7471 | (uint64_t)__builtin_neon_vqshld_n_u64(__a, __b); }) |
| 7472 | |
| 7473 | #define vqshrnh_n_s16(a, __b) __extension__ ({ \ |
| 7474 | int16_t __a = (a); \ |
| 7475 | (int8_t)__builtin_neon_vqshrnh_n_s16(__a, __b); }) |
| 7476 | #define vqshrns_n_s32(a, __b) __extension__ ({ \ |
| 7477 | int32_t __a = (a); \ |
| 7478 | (int16_t)__builtin_neon_vqshrns_n_s32(__a, __b); }) |
| 7479 | #define vqshrnd_n_s64(a, __b) __extension__ ({ \ |
| 7480 | int64_t __a = (a); \ |
| 7481 | (int32_t)__builtin_neon_vqshrnd_n_s64(__a, __b); }) |
| 7482 | #define vqshrnh_n_u16(a, __b) __extension__ ({ \ |
| 7483 | uint16_t __a = (a); \ |
| 7484 | (uint8_t)__builtin_neon_vqshrnh_n_u16(__a, __b); }) |
| 7485 | #define vqshrns_n_u32(a, __b) __extension__ ({ \ |
| 7486 | uint32_t __a = (a); \ |
| 7487 | (uint16_t)__builtin_neon_vqshrns_n_u32(__a, __b); }) |
| 7488 | #define vqshrnd_n_u64(a, __b) __extension__ ({ \ |
| 7489 | uint64_t __a = (a); \ |
| 7490 | (uint32_t)__builtin_neon_vqshrnd_n_u64(__a, __b); }) |
| 7491 | |
| 7492 | #define vqshrunh_n_s16(a, __b) __extension__ ({ \ |
| 7493 | int16_t __a = (a); \ |
| 7494 | (int8_t)__builtin_neon_vqshrunh_n_s16(__a, __b); }) |
| 7495 | #define vqshruns_n_s32(a, __b) __extension__ ({ \ |
| 7496 | int32_t __a = (a); \ |
| 7497 | (int16_t)__builtin_neon_vqshruns_n_s32(__a, __b); }) |
| 7498 | #define vqshrund_n_s64(a, __b) __extension__ ({ \ |
| 7499 | int64_t __a = (a); \ |
| 7500 | (int32_t)__builtin_neon_vqshrund_n_s64(__a, __b); }) |
| 7501 | |
| 7502 | __ai int8_t vqmovnh_s16(int16_t __a) { |
| 7503 | return (int8_t)__builtin_neon_vqmovnh_s16(__a); } |
| 7504 | __ai int16_t vqmovns_s32(int32_t __a) { |
| 7505 | return (int16_t)__builtin_neon_vqmovns_s32(__a); } |
| 7506 | __ai int32_t vqmovnd_s64(int64_t __a) { |
| 7507 | return (int32_t)__builtin_neon_vqmovnd_s64(__a); } |
| 7508 | |
| 7509 | __ai int8_t vqmovunh_s16(int16_t __a) { |
| 7510 | return (int8_t)__builtin_neon_vqmovunh_s16(__a); } |
| 7511 | __ai int16_t vqmovuns_s32(int32_t __a) { |
| 7512 | return (int16_t)__builtin_neon_vqmovuns_s32(__a); } |
| 7513 | __ai int32_t vqmovund_s64(int64_t __a) { |
| 7514 | return (int32_t)__builtin_neon_vqmovund_s64(__a); } |
| 7515 | |
| 7516 | #define vsrid_n_s64(a, b, __c) __extension__ ({ \ |
| 7517 | int64_t __a = (a); int64_t __b = (b); \ |
| 7518 | (int64_t)__builtin_neon_vsrid_n_s64(__a, __b, __c); }) |
| 7519 | #define vsrid_n_u64(a, b, __c) __extension__ ({ \ |
| 7520 | uint64_t __a = (a); uint64_t __b = (b); \ |
| 7521 | (uint64_t)__builtin_neon_vsrid_n_u64(__a, __b, __c); }) |
| 7522 | |
| 7523 | #define vrshrd_n_s64(a, __b) __extension__ ({ \ |
| 7524 | int64_t __a = (a); \ |
| 7525 | (int64_t)__builtin_neon_vrshrd_n_s64(__a, __b); }) |
| 7526 | #define vrshrd_n_u64(a, __b) __extension__ ({ \ |
| 7527 | uint64_t __a = (a); \ |
| 7528 | (uint64_t)__builtin_neon_vrshrd_n_u64(__a, __b); }) |
| 7529 | |
| 7530 | #define vrsrad_n_s64(a, b, __c) __extension__ ({ \ |
| 7531 | int64_t __a = (a); int64_t __b = (b); \ |
| 7532 | (int64_t)__builtin_neon_vrsrad_n_s64(__a, __b, __c); }) |
| 7533 | #define vrsrad_n_u64(a, b, __c) __extension__ ({ \ |
| 7534 | uint64_t __a = (a); uint64_t __b = (b); \ |
| 7535 | (uint64_t)__builtin_neon_vrsrad_n_u64(__a, __b, __c); }) |
| 7536 | |
| 7537 | #define vshrd_n_s64(a, __b) __extension__ ({ \ |
| 7538 | int64_t __a = (a); \ |
| 7539 | (int64_t)__builtin_neon_vshrd_n_s64(__a, __b); }) |
| 7540 | #define vshrd_n_u64(a, __b) __extension__ ({ \ |
| 7541 | uint64_t __a = (a); \ |
| 7542 | (uint64_t)__builtin_neon_vshrd_n_u64(__a, __b); }) |
| 7543 | |
| 7544 | #define vsrad_n_s64(a, b, __c) __extension__ ({ \ |
| 7545 | int64_t __a = (a); int64_t __b = (b); \ |
| 7546 | (int64_t)__builtin_neon_vsrad_n_s64(__a, __b, __c); }) |
| 7547 | #define vsrad_n_u64(a, b, __c) __extension__ ({ \ |
| 7548 | uint64_t __a = (a); uint64_t __b = (b); \ |
| 7549 | (uint64_t)__builtin_neon_vsrad_n_u64(__a, __b, __c); }) |
| 7550 | |
| 7551 | __ai int64_t vsubd_s64(int64_t __a, int64_t __b) { |
| 7552 | return (int64_t)__builtin_neon_vsubd_s64(__a, __b); } |
| 7553 | __ai uint64_t vsubd_u64(uint64_t __a, uint64_t __b) { |
| 7554 | return (uint64_t)__builtin_neon_vsubd_u64(__a, __b); } |
| 7555 | |
| 7556 | __ai int8_t vuqaddb_s8(int8_t __a, int8_t __b) { |
| 7557 | return (int8_t)__builtin_neon_vuqaddb_s8(__a, __b); } |
| 7558 | __ai int16_t vuqaddh_s16(int16_t __a, int16_t __b) { |
| 7559 | return (int16_t)__builtin_neon_vuqaddh_s16(__a, __b); } |
| 7560 | __ai int32_t vuqadds_s32(int32_t __a, int32_t __b) { |
| 7561 | return (int32_t)__builtin_neon_vuqadds_s32(__a, __b); } |
| 7562 | __ai int64_t vuqaddd_s64(int64_t __a, int64_t __b) { |
| 7563 | return (int64_t)__builtin_neon_vuqaddd_s64(__a, __b); } |
| 7564 | |
| 7565 | __ai float64_t vcvtd_f64_u64(uint64_t __a) { |
| 7566 | return (float64_t)__builtin_neon_vcvtd_f64_u64(__a); } |
| 7567 | |
| 7568 | __ai float32_t vcvts_f32_u32(uint32_t __a) { |
| 7569 | return (float32_t)__builtin_neon_vcvts_f32_u32(__a); } |
| 7570 | |
| 7571 | __ai uint8_t vqmovnh_u16(uint16_t __a) { |
| 7572 | return (uint8_t)__builtin_neon_vqmovnh_u16(__a); } |
| 7573 | __ai uint16_t vqmovns_u32(uint32_t __a) { |
| 7574 | return (uint16_t)__builtin_neon_vqmovns_u32(__a); } |
| 7575 | __ai uint32_t vqmovnd_u64(uint64_t __a) { |
| 7576 | return (uint32_t)__builtin_neon_vqmovnd_u64(__a); } |
| 7577 | |
| 7578 | __ai uint8_t vsqaddb_u8(uint8_t __a, uint8_t __b) { |
| 7579 | return (uint8_t)__builtin_neon_vsqaddb_u8(__a, __b); } |
| 7580 | __ai uint16_t vsqaddh_u16(uint16_t __a, uint16_t __b) { |
| 7581 | return (uint16_t)__builtin_neon_vsqaddh_u16(__a, __b); } |
| 7582 | __ai uint32_t vsqadds_u32(uint32_t __a, uint32_t __b) { |
| 7583 | return (uint32_t)__builtin_neon_vsqadds_u32(__a, __b); } |
| 7584 | __ai uint64_t vsqaddd_u64(uint64_t __a, uint64_t __b) { |
| 7585 | return (uint64_t)__builtin_neon_vsqaddd_u64(__a, __b); } |
| 7586 | |
| 7587 | #define vdupb_lane_s8(a, __b) __extension__ ({ \ |
| 7588 | int8x8_t __a = (a); \ |
| 7589 | (int8_t)__builtin_neon_vdupb_lane_i8(__a, __b); }) |
| 7590 | #define vduph_lane_s16(a, __b) __extension__ ({ \ |
| 7591 | int16x4_t __a = (a); \ |
| 7592 | (int16_t)__builtin_neon_vduph_lane_i16(__a, __b); }) |
| 7593 | #define vdups_lane_s32(a, __b) __extension__ ({ \ |
| 7594 | int32x2_t __a = (a); \ |
| 7595 | (int32_t)__builtin_neon_vdups_lane_i32(__a, __b); }) |
| 7596 | #define vdupd_lane_s64(a, __b) __extension__ ({ \ |
| 7597 | int64x1_t __a = (a); \ |
| 7598 | (int64_t)__builtin_neon_vdupd_lane_i64(__a, __b); }) |
| 7599 | #define vdups_lane_f32(a, __b) __extension__ ({ \ |
| 7600 | float32x2_t __a = (a); \ |
| 7601 | (float32_t)__builtin_neon_vdups_lane_f32(__a, __b); }) |
| 7602 | #define vdupd_lane_f64(a, __b) __extension__ ({ \ |
| 7603 | float64x1_t __a = (a); \ |
| 7604 | (float64_t)__builtin_neon_vdupd_lane_f64(__a, __b); }) |
| 7605 | #define vdupb_lane_u8(a, __b) __extension__ ({ \ |
| 7606 | uint8x8_t __a = (a); \ |
| 7607 | (uint8_t)__builtin_neon_vdupb_lane_i8((int8x8_t)__a, __b); }) |
| 7608 | #define vduph_lane_u16(a, __b) __extension__ ({ \ |
| 7609 | uint16x4_t __a = (a); \ |
| 7610 | (uint16_t)__builtin_neon_vduph_lane_i16((int16x4_t)__a, __b); }) |
| 7611 | #define vdups_lane_u32(a, __b) __extension__ ({ \ |
| 7612 | uint32x2_t __a = (a); \ |
| 7613 | (uint32_t)__builtin_neon_vdups_lane_i32((int32x2_t)__a, __b); }) |
| 7614 | #define vdupd_lane_u64(a, __b) __extension__ ({ \ |
| 7615 | uint64x1_t __a = (a); \ |
| 7616 | (uint64_t)__builtin_neon_vdupd_lane_i64((int64x1_t)__a, __b); }) |
| 7617 | #define vdupb_lane_p8(a, __b) __extension__ ({ \ |
| 7618 | poly8x8_t __a = (a); \ |
| 7619 | (poly8_t)__builtin_neon_vdupb_lane_i8((int8x8_t)__a, __b); }) |
| 7620 | #define vduph_lane_p16(a, __b) __extension__ ({ \ |
| 7621 | poly16x4_t __a = (a); \ |
| 7622 | (poly16_t)__builtin_neon_vduph_lane_i16((int16x4_t)__a, __b); }) |
| 7623 | |
| 7624 | #define vdupb_laneq_s8(a, __b) __extension__ ({ \ |
| 7625 | int8x16_t __a = (a); \ |
| 7626 | (int8_t)__builtin_neon_vdupb_laneq_i8(__a, __b); }) |
| 7627 | #define vduph_laneq_s16(a, __b) __extension__ ({ \ |
| 7628 | int16x8_t __a = (a); \ |
| 7629 | (int16_t)__builtin_neon_vduph_laneq_i16(__a, __b); }) |
| 7630 | #define vdups_laneq_s32(a, __b) __extension__ ({ \ |
| 7631 | int32x4_t __a = (a); \ |
| 7632 | (int32_t)__builtin_neon_vdups_laneq_i32(__a, __b); }) |
| 7633 | #define vdupd_laneq_s64(a, __b) __extension__ ({ \ |
| 7634 | int64x2_t __a = (a); \ |
| 7635 | (int64_t)__builtin_neon_vdupd_laneq_i64(__a, __b); }) |
| 7636 | #define vdups_laneq_f32(a, __b) __extension__ ({ \ |
| 7637 | float32x4_t __a = (a); \ |
| 7638 | (float32_t)__builtin_neon_vdups_laneq_f32(__a, __b); }) |
| 7639 | #define vdupd_laneq_f64(a, __b) __extension__ ({ \ |
| 7640 | float64x2_t __a = (a); \ |
| 7641 | (float64_t)__builtin_neon_vdupd_laneq_f64(__a, __b); }) |
| 7642 | #define vdupb_laneq_u8(a, __b) __extension__ ({ \ |
| 7643 | uint8x16_t __a = (a); \ |
| 7644 | (uint8_t)__builtin_neon_vdupb_laneq_i8((int8x16_t)__a, __b); }) |
| 7645 | #define vduph_laneq_u16(a, __b) __extension__ ({ \ |
| 7646 | uint16x8_t __a = (a); \ |
| 7647 | (uint16_t)__builtin_neon_vduph_laneq_i16((int16x8_t)__a, __b); }) |
| 7648 | #define vdups_laneq_u32(a, __b) __extension__ ({ \ |
| 7649 | uint32x4_t __a = (a); \ |
| 7650 | (uint32_t)__builtin_neon_vdups_laneq_i32((int32x4_t)__a, __b); }) |
| 7651 | #define vdupd_laneq_u64(a, __b) __extension__ ({ \ |
| 7652 | uint64x2_t __a = (a); \ |
| 7653 | (uint64_t)__builtin_neon_vdupd_laneq_i64((int64x2_t)__a, __b); }) |
| 7654 | #define vdupb_laneq_p8(a, __b) __extension__ ({ \ |
| 7655 | poly8x16_t __a = (a); \ |
| 7656 | (poly8_t)__builtin_neon_vdupb_laneq_i8((int8x16_t)__a, __b); }) |
| 7657 | #define vduph_laneq_p16(a, __b) __extension__ ({ \ |
| 7658 | poly16x8_t __a = (a); \ |
| 7659 | (poly16_t)__builtin_neon_vduph_laneq_i16((int16x8_t)__a, __b); }) |
| 7660 | |
| 7661 | #define vmulx_lane_f64(a, b, __c) __extension__ ({ \ |
| 7662 | float64x1_t __a = (a); float64x1_t __b = (b); \ |
| 7663 | float64_t __d1 = vget_lane_f64(__a, 0);\ |
| 7664 | float64_t __e1 = vget_lane_f64(__b, __c);\ |
| 7665 | float64_t __f1 = vmulxd_f64(__d1, __e1);\ |
| 7666 | float64x1_t __g1;\ |
| 7667 | vset_lane_f64(__f1, __g1, __c); }) |
| 7668 | |
| 7669 | #define vmulx_laneq_f64(a, b, __c) __extension__ ({ \ |
| 7670 | float64x1_t __a = (a); float64x2_t __b = (b); \ |
| 7671 | float64_t __d1 = vget_lane_f64(__a, 0);\ |
| 7672 | float64_t __e1 = vgetq_lane_f64(__b, __c);\ |
| 7673 | float64_t __f1 = vmulxd_f64(__d1, __e1);\ |
| 7674 | float64x1_t __g1;\ |
| 7675 | vset_lane_f64(__f1, __g1, 0); }) |
| 7676 | |
| 7677 | #define vmul_lane_f64(a, b, __c) __extension__ ({ \ |
| 7678 | float64x1_t __a = (a); float64x1_t __b = (b); \ |
| 7679 | (float64x1_t)__builtin_neon_vmul_lane_v((int8x8_t)__a, (int8x8_t)__b, __c, 9); }) |
| 7680 | |
| 7681 | #define vmul_laneq_f64(a, b, __c) __extension__ ({ \ |
| 7682 | float64x1_t __a = (a); float64x2_t __b = (b); \ |
| 7683 | (float64x1_t)__builtin_neon_vmul_laneq_v((int8x8_t)__a, (int8x16_t)__b, __c, 9); }) |
| 7684 | |
| 7685 | __ai float64x1_t vmul_n_f64(float64x1_t __a, float64_t __b) { |
| 7686 | return (float64x1_t)__builtin_neon_vmul_n_f64(__a, __b); } |
| 7687 | |
| 7688 | #define vset_lane_f64(a, b, __c) __extension__ ({ \ |
| 7689 | float64_t __a = (a); float64x1_t __b = (b); \ |
| 7690 | (float64x1_t)__builtin_neon_vset_lane_f64(__a, __b, __c); }) |
| 7691 | #define vsetq_lane_f64(a, b, __c) __extension__ ({ \ |
| 7692 | float64_t __a = (a); float64x2_t __b = (b); \ |
| 7693 | (float64x2_t)__builtin_neon_vsetq_lane_f64(__a, __b, __c); }) |
| 7694 | #define vset_lane_p64(a, b, __c) __extension__ ({ \ |
| 7695 | poly64_t __a = (a); poly64x1_t __b = (b); \ |
| 7696 | (poly64x1_t)__builtin_neon_vset_lane_i64(__a, (int64x1_t)__b, __c); }) |
| 7697 | #define vsetq_lane_p64(a, b, __c) __extension__ ({ \ |
| 7698 | poly64_t __a = (a); poly64x2_t __b = (b); \ |
| 7699 | (poly64x2_t)__builtin_neon_vsetq_lane_i64(__a, (int64x2_t)__b, __c); }) |
| 7700 | |
| 7701 | #define vshll_high_n_s8(a, __b) __extension__ ({ \ |
| 7702 | int8x16_t __a = (a); \ |
| 7703 | int8x8_t __a1 = vget_high_s8(__a); \ |
| 7704 | (int16x8_t)vshll_n_s8(__a1, __b); }) |
| 7705 | #define vshll_high_n_s16(a, __b) __extension__ ({ \ |
| 7706 | int16x8_t __a = (a); \ |
| 7707 | int16x4_t __a1 = vget_high_s16(__a); \ |
| 7708 | (int32x4_t)vshll_n_s16(__a1, __b); }) |
| 7709 | #define vshll_high_n_s32(a, __b) __extension__ ({ \ |
| 7710 | int32x4_t __a = (a); \ |
| 7711 | int32x2_t __a1 = vget_high_s32(__a); \ |
| 7712 | (int64x2_t)vshll_n_s32(__a1, __b); }) |
| 7713 | #define vshll_high_n_u8(a, __b) __extension__ ({ \ |
| 7714 | uint8x16_t __a = (a); \ |
| 7715 | uint8x8_t __a1 = vget_high_u8(__a); \ |
| 7716 | (uint16x8_t)vshll_n_u8(__a1, __b); }) |
| 7717 | #define vshll_high_n_u16(a, __b) __extension__ ({ \ |
| 7718 | uint16x8_t __a = (a); \ |
| 7719 | uint16x4_t __a1 = vget_high_u16(__a); \ |
| 7720 | (uint32x4_t)vshll_n_u16(__a1, __b); }) |
| 7721 | #define vshll_high_n_u32(a, __b) __extension__ ({ \ |
| 7722 | uint32x4_t __a = (a); \ |
| 7723 | uint32x2_t __a1 = vget_high_u32(__a); \ |
| 7724 | (uint64x2_t)vshll_n_u32(__a1, __b); }) |
| 7725 | |
| 7726 | #define vshrn_high_n_s16(a, b, __c) __extension__ ({ \ |
| 7727 | int8x8_t __a = (a); int16x8_t __b = (b); \ |
| 7728 | (int8x16_t)vcombine_s16(__a, vshrn_n_s16(__b, __c)); }) |
| 7729 | #define vshrn_high_n_s32(a, b, __c) __extension__ ({ \ |
| 7730 | int16x4_t __a = (a); int32x4_t __b = (b); \ |
| 7731 | (int16x8_t)vcombine_s32(__a, vshrn_n_s32(__b, __c)); }) |
| 7732 | #define vshrn_high_n_s64(a, b, __c) __extension__ ({ \ |
| 7733 | int32x2_t __a = (a); int64x2_t __b = (b); \ |
| 7734 | (int32x4_t)vcombine_s64(__a, vshrn_n_s64(__b, __c)); }) |
| 7735 | #define vshrn_high_n_u16(a, b, __c) __extension__ ({ \ |
| 7736 | uint8x8_t __a = (a); uint16x8_t __b = (b); \ |
| 7737 | (uint8x16_t)vcombine_u16(__a, vshrn_n_u16(__b, __c)); }) |
| 7738 | #define vshrn_high_n_u32(a, b, __c) __extension__ ({ \ |
| 7739 | uint16x4_t __a = (a); uint32x4_t __b = (b); \ |
| 7740 | (uint16x8_t)vcombine_u32(__a, vshrn_n_u32(__b, __c)); }) |
| 7741 | #define vshrn_high_n_u64(a, b, __c) __extension__ ({ \ |
| 7742 | uint32x2_t __a = (a); uint64x2_t __b = (b); \ |
| 7743 | (uint32x4_t)vcombine_u64(__a, vshrn_n_u64(__b, __c)); }) |
| 7744 | |
| 7745 | #define vsli_n_p64(a, b, __c) __extension__ ({ \ |
| 7746 | poly64x1_t __a = (a); poly64x1_t __b = (b); \ |
| 7747 | (poly64x1_t)__builtin_neon_vsli_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 6); }) |
| 7748 | #define vsliq_n_p64(a, b, __c) __extension__ ({ \ |
| 7749 | poly64x2_t __a = (a); poly64x2_t __b = (b); \ |
| 7750 | (poly64x2_t)__builtin_neon_vsliq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 38); }) |
| 7751 | |
| 7752 | __ai int8x16_t vqmovun_high_s16(int8x8_t __a, int16x8_t __b) { |
| 7753 | int8x8_t __a1 = vqmovun_s16(__b); |
| 7754 | return __builtin_shufflevector(__a, __a1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15); } |
| 7755 | __ai int16x8_t vqmovun_high_s32(int16x4_t __a, int32x4_t __b) { |
| 7756 | int16x4_t __a1 = vqmovun_s32(__b); |
| 7757 | return __builtin_shufflevector(__a, __a1, 0, 1, 2, 3, 4, 5, 6, 7); } |
| 7758 | __ai int32x4_t vqmovun_high_s64(int32x2_t __a, int64x2_t __b) { |
| 7759 | int32x2_t __a1 = vqmovun_s64(__b); |
| 7760 | return __builtin_shufflevector(__a, __a1, 0, 1, 2, 3); } |
| 7761 | |
| 7762 | #define vsri_n_p64(a, b, __c) __extension__ ({ \ |
| 7763 | poly64x1_t __a = (a); poly64x1_t __b = (b); \ |
| 7764 | (poly64x1_t)__builtin_neon_vsri_n_v((int8x8_t)__a, (int8x8_t)__b, __c, 6); }) |
| 7765 | #define vsriq_n_p64(a, b, __c) __extension__ ({ \ |
| 7766 | poly64x2_t __a = (a); poly64x2_t __b = (b); \ |
| 7767 | (poly64x2_t)__builtin_neon_vsriq_n_v((int8x16_t)__a, (int8x16_t)__b, __c, 38); }) |
| 7768 | |
| 7769 | #define vst1q_f64(__a, b) __extension__ ({ \ |
| 7770 | float64x2_t __b = (b); \ |
| 7771 | __builtin_neon_vst1q_v(__a, (int8x16_t)__b, 41); }) |
| 7772 | #define vst1_f64(__a, b) __extension__ ({ \ |
| 7773 | float64x1_t __b = (b); \ |
| 7774 | __builtin_neon_vst1_v(__a, (int8x8_t)__b, 9); }) |
| 7775 | #define vst1_p64(__a, b) __extension__ ({ \ |
| 7776 | poly64x1_t __b = (b); \ |
| 7777 | __builtin_neon_vst1_v(__a, (int8x8_t)__b, 6); }) |
| 7778 | #define vst1q_p64(__a, b) __extension__ ({ \ |
| 7779 | poly64x2_t __b = (b); \ |
| 7780 | __builtin_neon_vst1q_v(__a, (int8x16_t)__b, 38); }) |
| 7781 | |
| 7782 | #define vst1q_lane_f64(__a, b, __c) __extension__ ({ \ |
| 7783 | float64x2_t __b = (b); \ |
| 7784 | __builtin_neon_vst1q_lane_v(__a, (int8x16_t)__b, __c, 41); }) |
| 7785 | #define vst1q_lane_p64(__a, b, __c) __extension__ ({ \ |
| 7786 | poly64x2_t __b = (b); \ |
| 7787 | __builtin_neon_vst1q_lane_v(__a, (int8x16_t)__b, __c, 38); }) |
| 7788 | #define vst1_lane_f64(__a, b, __c) __extension__ ({ \ |
| 7789 | float64x1_t __b = (b); \ |
| 7790 | __builtin_neon_vst1_lane_v(__a, (int8x8_t)__b, __c, 9); }) |
| 7791 | #define vst1_lane_p64(__a, b, __c) __extension__ ({ \ |
| 7792 | poly64x1_t __b = (b); \ |
| 7793 | __builtin_neon_vst1_lane_v(__a, (int8x8_t)__b, __c, 6); }) |
| 7794 | |
| 7795 | #define vst1q_u8_x2(__a, b) __extension__ ({ \ |
| 7796 | uint8x16x2_t __b = (b); \ |
| 7797 | __builtin_neon_vst1q_x2_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], 48); }) |
| 7798 | #define vst1q_u16_x2(__a, b) __extension__ ({ \ |
| 7799 | uint16x8x2_t __b = (b); \ |
| 7800 | __builtin_neon_vst1q_x2_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], 49); }) |
| 7801 | #define vst1q_u32_x2(__a, b) __extension__ ({ \ |
| 7802 | uint32x4x2_t __b = (b); \ |
| 7803 | __builtin_neon_vst1q_x2_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], 50); }) |
| 7804 | #define vst1q_u64_x2(__a, b) __extension__ ({ \ |
| 7805 | uint64x2x2_t __b = (b); \ |
| 7806 | __builtin_neon_vst1q_x2_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], 51); }) |
| 7807 | #define vst1q_s8_x2(__a, b) __extension__ ({ \ |
| 7808 | int8x16x2_t __b = (b); \ |
| 7809 | __builtin_neon_vst1q_x2_v(__a, __b.val[0], __b.val[1], 32); }) |
| 7810 | #define vst1q_s16_x2(__a, b) __extension__ ({ \ |
| 7811 | int16x8x2_t __b = (b); \ |
| 7812 | __builtin_neon_vst1q_x2_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], 33); }) |
| 7813 | #define vst1q_s32_x2(__a, b) __extension__ ({ \ |
| 7814 | int32x4x2_t __b = (b); \ |
| 7815 | __builtin_neon_vst1q_x2_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], 34); }) |
| 7816 | #define vst1q_s64_x2(__a, b) __extension__ ({ \ |
| 7817 | int64x2x2_t __b = (b); \ |
| 7818 | __builtin_neon_vst1q_x2_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], 35); }) |
| 7819 | #define vst1q_f16_x2(__a, b) __extension__ ({ \ |
| 7820 | float16x8x2_t __b = (b); \ |
| 7821 | __builtin_neon_vst1q_x2_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], 39); }) |
| 7822 | #define vst1q_f32_x2(__a, b) __extension__ ({ \ |
| 7823 | float32x4x2_t __b = (b); \ |
| 7824 | __builtin_neon_vst1q_x2_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], 40); }) |
| 7825 | #define vst1q_f64_x2(__a, b) __extension__ ({ \ |
| 7826 | float64x2x2_t __b = (b); \ |
| 7827 | __builtin_neon_vst1q_x2_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], 41); }) |
| 7828 | #define vst1q_p8_x2(__a, b) __extension__ ({ \ |
| 7829 | poly8x16x2_t __b = (b); \ |
| 7830 | __builtin_neon_vst1q_x2_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], 36); }) |
| 7831 | #define vst1q_p16_x2(__a, b) __extension__ ({ \ |
| 7832 | poly16x8x2_t __b = (b); \ |
| 7833 | __builtin_neon_vst1q_x2_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], 37); }) |
| 7834 | #define vst1q_p64_x2(__a, b) __extension__ ({ \ |
| 7835 | poly64x2x2_t __b = (b); \ |
| 7836 | __builtin_neon_vst1q_x2_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], 38); }) |
| 7837 | #define vst1_u8_x2(__a, b) __extension__ ({ \ |
| 7838 | uint8x8x2_t __b = (b); \ |
| 7839 | __builtin_neon_vst1_x2_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], 16); }) |
| 7840 | #define vst1_u16_x2(__a, b) __extension__ ({ \ |
| 7841 | uint16x4x2_t __b = (b); \ |
| 7842 | __builtin_neon_vst1_x2_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], 17); }) |
| 7843 | #define vst1_u32_x2(__a, b) __extension__ ({ \ |
| 7844 | uint32x2x2_t __b = (b); \ |
| 7845 | __builtin_neon_vst1_x2_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], 18); }) |
| 7846 | #define vst1_u64_x2(__a, b) __extension__ ({ \ |
| 7847 | uint64x1x2_t __b = (b); \ |
| 7848 | __builtin_neon_vst1_x2_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], 19); }) |
| 7849 | #define vst1_s8_x2(__a, b) __extension__ ({ \ |
| 7850 | int8x8x2_t __b = (b); \ |
| 7851 | __builtin_neon_vst1_x2_v(__a, __b.val[0], __b.val[1], 0); }) |
| 7852 | #define vst1_s16_x2(__a, b) __extension__ ({ \ |
| 7853 | int16x4x2_t __b = (b); \ |
| 7854 | __builtin_neon_vst1_x2_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], 1); }) |
| 7855 | #define vst1_s32_x2(__a, b) __extension__ ({ \ |
| 7856 | int32x2x2_t __b = (b); \ |
| 7857 | __builtin_neon_vst1_x2_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], 2); }) |
| 7858 | #define vst1_s64_x2(__a, b) __extension__ ({ \ |
| 7859 | int64x1x2_t __b = (b); \ |
| 7860 | __builtin_neon_vst1_x2_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], 3); }) |
| 7861 | #define vst1_f16_x2(__a, b) __extension__ ({ \ |
| 7862 | float16x4x2_t __b = (b); \ |
| 7863 | __builtin_neon_vst1_x2_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], 7); }) |
| 7864 | #define vst1_f32_x2(__a, b) __extension__ ({ \ |
| 7865 | float32x2x2_t __b = (b); \ |
| 7866 | __builtin_neon_vst1_x2_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], 8); }) |
| 7867 | #define vst1_f64_x2(__a, b) __extension__ ({ \ |
| 7868 | float64x1x2_t __b = (b); \ |
| 7869 | __builtin_neon_vst1_x2_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], 9); }) |
| 7870 | #define vst1_p8_x2(__a, b) __extension__ ({ \ |
| 7871 | poly8x8x2_t __b = (b); \ |
| 7872 | __builtin_neon_vst1_x2_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], 4); }) |
| 7873 | #define vst1_p16_x2(__a, b) __extension__ ({ \ |
| 7874 | poly16x4x2_t __b = (b); \ |
| 7875 | __builtin_neon_vst1_x2_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], 5); }) |
| 7876 | #define vst1_p64_x2(__a, b) __extension__ ({ \ |
| 7877 | poly64x1x2_t __b = (b); \ |
| 7878 | __builtin_neon_vst1_x2_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], 6); }) |
| 7879 | |
| 7880 | #define vst1q_u8_x3(__a, b) __extension__ ({ \ |
| 7881 | uint8x16x3_t __b = (b); \ |
| 7882 | __builtin_neon_vst1q_x3_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], 48); }) |
| 7883 | #define vst1q_u16_x3(__a, b) __extension__ ({ \ |
| 7884 | uint16x8x3_t __b = (b); \ |
| 7885 | __builtin_neon_vst1q_x3_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], 49); }) |
| 7886 | #define vst1q_u32_x3(__a, b) __extension__ ({ \ |
| 7887 | uint32x4x3_t __b = (b); \ |
| 7888 | __builtin_neon_vst1q_x3_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], 50); }) |
| 7889 | #define vst1q_u64_x3(__a, b) __extension__ ({ \ |
| 7890 | uint64x2x3_t __b = (b); \ |
| 7891 | __builtin_neon_vst1q_x3_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], 51); }) |
| 7892 | #define vst1q_s8_x3(__a, b) __extension__ ({ \ |
| 7893 | int8x16x3_t __b = (b); \ |
| 7894 | __builtin_neon_vst1q_x3_v(__a, __b.val[0], __b.val[1], __b.val[2], 32); }) |
| 7895 | #define vst1q_s16_x3(__a, b) __extension__ ({ \ |
| 7896 | int16x8x3_t __b = (b); \ |
| 7897 | __builtin_neon_vst1q_x3_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], 33); }) |
| 7898 | #define vst1q_s32_x3(__a, b) __extension__ ({ \ |
| 7899 | int32x4x3_t __b = (b); \ |
| 7900 | __builtin_neon_vst1q_x3_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], 34); }) |
| 7901 | #define vst1q_s64_x3(__a, b) __extension__ ({ \ |
| 7902 | int64x2x3_t __b = (b); \ |
| 7903 | __builtin_neon_vst1q_x3_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], 35); }) |
| 7904 | #define vst1q_f16_x3(__a, b) __extension__ ({ \ |
| 7905 | float16x8x3_t __b = (b); \ |
| 7906 | __builtin_neon_vst1q_x3_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], 39); }) |
| 7907 | #define vst1q_f32_x3(__a, b) __extension__ ({ \ |
| 7908 | float32x4x3_t __b = (b); \ |
| 7909 | __builtin_neon_vst1q_x3_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], 40); }) |
| 7910 | #define vst1q_f64_x3(__a, b) __extension__ ({ \ |
| 7911 | float64x2x3_t __b = (b); \ |
| 7912 | __builtin_neon_vst1q_x3_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], 41); }) |
| 7913 | #define vst1q_p8_x3(__a, b) __extension__ ({ \ |
| 7914 | poly8x16x3_t __b = (b); \ |
| 7915 | __builtin_neon_vst1q_x3_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], 36); }) |
| 7916 | #define vst1q_p16_x3(__a, b) __extension__ ({ \ |
| 7917 | poly16x8x3_t __b = (b); \ |
| 7918 | __builtin_neon_vst1q_x3_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], 37); }) |
| 7919 | #define vst1q_p64_x3(__a, b) __extension__ ({ \ |
| 7920 | poly64x2x3_t __b = (b); \ |
| 7921 | __builtin_neon_vst1q_x3_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], 38); }) |
| 7922 | #define vst1_u8_x3(__a, b) __extension__ ({ \ |
| 7923 | uint8x8x3_t __b = (b); \ |
| 7924 | __builtin_neon_vst1_x3_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], 16); }) |
| 7925 | #define vst1_u16_x3(__a, b) __extension__ ({ \ |
| 7926 | uint16x4x3_t __b = (b); \ |
| 7927 | __builtin_neon_vst1_x3_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], 17); }) |
| 7928 | #define vst1_u32_x3(__a, b) __extension__ ({ \ |
| 7929 | uint32x2x3_t __b = (b); \ |
| 7930 | __builtin_neon_vst1_x3_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], 18); }) |
| 7931 | #define vst1_u64_x3(__a, b) __extension__ ({ \ |
| 7932 | uint64x1x3_t __b = (b); \ |
| 7933 | __builtin_neon_vst1_x3_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], 19); }) |
| 7934 | #define vst1_s8_x3(__a, b) __extension__ ({ \ |
| 7935 | int8x8x3_t __b = (b); \ |
| 7936 | __builtin_neon_vst1_x3_v(__a, __b.val[0], __b.val[1], __b.val[2], 0); }) |
| 7937 | #define vst1_s16_x3(__a, b) __extension__ ({ \ |
| 7938 | int16x4x3_t __b = (b); \ |
| 7939 | __builtin_neon_vst1_x3_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], 1); }) |
| 7940 | #define vst1_s32_x3(__a, b) __extension__ ({ \ |
| 7941 | int32x2x3_t __b = (b); \ |
| 7942 | __builtin_neon_vst1_x3_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], 2); }) |
| 7943 | #define vst1_s64_x3(__a, b) __extension__ ({ \ |
| 7944 | int64x1x3_t __b = (b); \ |
| 7945 | __builtin_neon_vst1_x3_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], 3); }) |
| 7946 | #define vst1_f16_x3(__a, b) __extension__ ({ \ |
| 7947 | float16x4x3_t __b = (b); \ |
| 7948 | __builtin_neon_vst1_x3_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], 7); }) |
| 7949 | #define vst1_f32_x3(__a, b) __extension__ ({ \ |
| 7950 | float32x2x3_t __b = (b); \ |
| 7951 | __builtin_neon_vst1_x3_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], 8); }) |
| 7952 | #define vst1_f64_x3(__a, b) __extension__ ({ \ |
| 7953 | float64x1x3_t __b = (b); \ |
| 7954 | __builtin_neon_vst1_x3_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], 9); }) |
| 7955 | #define vst1_p8_x3(__a, b) __extension__ ({ \ |
| 7956 | poly8x8x3_t __b = (b); \ |
| 7957 | __builtin_neon_vst1_x3_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], 4); }) |
| 7958 | #define vst1_p16_x3(__a, b) __extension__ ({ \ |
| 7959 | poly16x4x3_t __b = (b); \ |
| 7960 | __builtin_neon_vst1_x3_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], 5); }) |
| 7961 | #define vst1_p64_x3(__a, b) __extension__ ({ \ |
| 7962 | poly64x1x3_t __b = (b); \ |
| 7963 | __builtin_neon_vst1_x3_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], 6); }) |
| 7964 | |
| 7965 | #define vst1q_u8_x4(__a, b) __extension__ ({ \ |
| 7966 | uint8x16x4_t __b = (b); \ |
| 7967 | __builtin_neon_vst1q_x4_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], 48); }) |
| 7968 | #define vst1q_u16_x4(__a, b) __extension__ ({ \ |
| 7969 | uint16x8x4_t __b = (b); \ |
| 7970 | __builtin_neon_vst1q_x4_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], 49); }) |
| 7971 | #define vst1q_u32_x4(__a, b) __extension__ ({ \ |
| 7972 | uint32x4x4_t __b = (b); \ |
| 7973 | __builtin_neon_vst1q_x4_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], 50); }) |
| 7974 | #define vst1q_u64_x4(__a, b) __extension__ ({ \ |
| 7975 | uint64x2x4_t __b = (b); \ |
| 7976 | __builtin_neon_vst1q_x4_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], 51); }) |
| 7977 | #define vst1q_s8_x4(__a, b) __extension__ ({ \ |
| 7978 | int8x16x4_t __b = (b); \ |
| 7979 | __builtin_neon_vst1q_x4_v(__a, __b.val[0], __b.val[1], __b.val[2], __b.val[3], 32); }) |
| 7980 | #define vst1q_s16_x4(__a, b) __extension__ ({ \ |
| 7981 | int16x8x4_t __b = (b); \ |
| 7982 | __builtin_neon_vst1q_x4_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], 33); }) |
| 7983 | #define vst1q_s32_x4(__a, b) __extension__ ({ \ |
| 7984 | int32x4x4_t __b = (b); \ |
| 7985 | __builtin_neon_vst1q_x4_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], 34); }) |
| 7986 | #define vst1q_s64_x4(__a, b) __extension__ ({ \ |
| 7987 | int64x2x4_t __b = (b); \ |
| 7988 | __builtin_neon_vst1q_x4_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], 35); }) |
| 7989 | #define vst1q_f16_x4(__a, b) __extension__ ({ \ |
| 7990 | float16x8x4_t __b = (b); \ |
| 7991 | __builtin_neon_vst1q_x4_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], 39); }) |
| 7992 | #define vst1q_f32_x4(__a, b) __extension__ ({ \ |
| 7993 | float32x4x4_t __b = (b); \ |
| 7994 | __builtin_neon_vst1q_x4_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], 40); }) |
| 7995 | #define vst1q_f64_x4(__a, b) __extension__ ({ \ |
| 7996 | float64x2x4_t __b = (b); \ |
| 7997 | __builtin_neon_vst1q_x4_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], 41); }) |
| 7998 | #define vst1q_p8_x4(__a, b) __extension__ ({ \ |
| 7999 | poly8x16x4_t __b = (b); \ |
| 8000 | __builtin_neon_vst1q_x4_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], 36); }) |
| 8001 | #define vst1q_p16_x4(__a, b) __extension__ ({ \ |
| 8002 | poly16x8x4_t __b = (b); \ |
| 8003 | __builtin_neon_vst1q_x4_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], 37); }) |
| 8004 | #define vst1q_p64_x4(__a, b) __extension__ ({ \ |
| 8005 | poly64x2x4_t __b = (b); \ |
| 8006 | __builtin_neon_vst1q_x4_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], 38); }) |
| 8007 | #define vst1_u8_x4(__a, b) __extension__ ({ \ |
| 8008 | uint8x8x4_t __b = (b); \ |
| 8009 | __builtin_neon_vst1_x4_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], 16); }) |
| 8010 | #define vst1_u16_x4(__a, b) __extension__ ({ \ |
| 8011 | uint16x4x4_t __b = (b); \ |
| 8012 | __builtin_neon_vst1_x4_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], 17); }) |
| 8013 | #define vst1_u32_x4(__a, b) __extension__ ({ \ |
| 8014 | uint32x2x4_t __b = (b); \ |
| 8015 | __builtin_neon_vst1_x4_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], 18); }) |
| 8016 | #define vst1_u64_x4(__a, b) __extension__ ({ \ |
| 8017 | uint64x1x4_t __b = (b); \ |
| 8018 | __builtin_neon_vst1_x4_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], 19); }) |
| 8019 | #define vst1_s8_x4(__a, b) __extension__ ({ \ |
| 8020 | int8x8x4_t __b = (b); \ |
| 8021 | __builtin_neon_vst1_x4_v(__a, __b.val[0], __b.val[1], __b.val[2], __b.val[3], 0); }) |
| 8022 | #define vst1_s16_x4(__a, b) __extension__ ({ \ |
| 8023 | int16x4x4_t __b = (b); \ |
| 8024 | __builtin_neon_vst1_x4_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], 1); }) |
| 8025 | #define vst1_s32_x4(__a, b) __extension__ ({ \ |
| 8026 | int32x2x4_t __b = (b); \ |
| 8027 | __builtin_neon_vst1_x4_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], 2); }) |
| 8028 | #define vst1_s64_x4(__a, b) __extension__ ({ \ |
| 8029 | int64x1x4_t __b = (b); \ |
| 8030 | __builtin_neon_vst1_x4_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], 3); }) |
| 8031 | #define vst1_f16_x4(__a, b) __extension__ ({ \ |
| 8032 | float16x4x4_t __b = (b); \ |
| 8033 | __builtin_neon_vst1_x4_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], 7); }) |
| 8034 | #define vst1_f32_x4(__a, b) __extension__ ({ \ |
| 8035 | float32x2x4_t __b = (b); \ |
| 8036 | __builtin_neon_vst1_x4_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], 8); }) |
| 8037 | #define vst1_f64_x4(__a, b) __extension__ ({ \ |
| 8038 | float64x1x4_t __b = (b); \ |
| 8039 | __builtin_neon_vst1_x4_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], 9); }) |
| 8040 | #define vst1_p8_x4(__a, b) __extension__ ({ \ |
| 8041 | poly8x8x4_t __b = (b); \ |
| 8042 | __builtin_neon_vst1_x4_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], 4); }) |
| 8043 | #define vst1_p16_x4(__a, b) __extension__ ({ \ |
| 8044 | poly16x4x4_t __b = (b); \ |
| 8045 | __builtin_neon_vst1_x4_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], 5); }) |
| 8046 | #define vst1_p64_x4(__a, b) __extension__ ({ \ |
| 8047 | poly64x1x4_t __b = (b); \ |
| 8048 | __builtin_neon_vst1_x4_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], 6); }) |
| 8049 | |
| 8050 | #define vst2q_u64(__a, b) __extension__ ({ \ |
| 8051 | uint64x2x2_t __b = (b); \ |
| 8052 | __builtin_neon_vst2q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], 51); }) |
| 8053 | #define vst2q_s64(__a, b) __extension__ ({ \ |
| 8054 | int64x2x2_t __b = (b); \ |
| 8055 | __builtin_neon_vst2q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], 35); }) |
| 8056 | #define vst2q_f64(__a, b) __extension__ ({ \ |
| 8057 | float64x2x2_t __b = (b); \ |
| 8058 | __builtin_neon_vst2q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], 41); }) |
| 8059 | #define vst2_f64(__a, b) __extension__ ({ \ |
| 8060 | float64x1x2_t __b = (b); \ |
| 8061 | __builtin_neon_vst2_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], 9); }) |
| 8062 | #define vst2_p64(__a, b) __extension__ ({ \ |
| 8063 | poly64x1x2_t __b = (b); \ |
| 8064 | __builtin_neon_vst2_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], 6); }) |
| 8065 | #define vst2q_p64(__a, b) __extension__ ({ \ |
| 8066 | poly64x2x2_t __b = (b); \ |
| 8067 | __builtin_neon_vst2q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], 38); }) |
| 8068 | |
| 8069 | #define vst2q_lane_u8(__a, b, __c) __extension__ ({ \ |
| 8070 | uint8x16x2_t __b = (b); \ |
| 8071 | __builtin_neon_vst2q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], __c, 48); }) |
| 8072 | #define vst2q_lane_u64(__a, b, __c) __extension__ ({ \ |
| 8073 | uint64x2x2_t __b = (b); \ |
| 8074 | __builtin_neon_vst2q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], __c, 51); }) |
| 8075 | #define vst2q_lane_s8(__a, b, __c) __extension__ ({ \ |
| 8076 | int8x16x2_t __b = (b); \ |
| 8077 | __builtin_neon_vst2q_lane_v(__a, __b.val[0], __b.val[1], __c, 32); }) |
| 8078 | #define vst2q_lane_s64(__a, b, __c) __extension__ ({ \ |
| 8079 | int64x2x2_t __b = (b); \ |
| 8080 | __builtin_neon_vst2q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], __c, 35); }) |
| 8081 | #define vst2q_lane_f64(__a, b, __c) __extension__ ({ \ |
| 8082 | float64x2x2_t __b = (b); \ |
| 8083 | __builtin_neon_vst2q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], __c, 41); }) |
| 8084 | #define vst2q_lane_p8(__a, b, __c) __extension__ ({ \ |
| 8085 | poly8x16x2_t __b = (b); \ |
| 8086 | __builtin_neon_vst2q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], __c, 36); }) |
| 8087 | #define vst2q_lane_p64(__a, b, __c) __extension__ ({ \ |
| 8088 | poly64x2x2_t __b = (b); \ |
| 8089 | __builtin_neon_vst2q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], __c, 38); }) |
| 8090 | #define vst2_lane_u64(__a, b, __c) __extension__ ({ \ |
| 8091 | uint64x1x2_t __b = (b); \ |
| 8092 | __builtin_neon_vst2_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], __c, 19); }) |
| 8093 | #define vst2_lane_s64(__a, b, __c) __extension__ ({ \ |
| 8094 | int64x1x2_t __b = (b); \ |
| 8095 | __builtin_neon_vst2_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], __c, 3); }) |
| 8096 | #define vst2_lane_f64(__a, b, __c) __extension__ ({ \ |
| 8097 | float64x1x2_t __b = (b); \ |
| 8098 | __builtin_neon_vst2_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], __c, 9); }) |
| 8099 | #define vst2_lane_p64(__a, b, __c) __extension__ ({ \ |
| 8100 | poly64x1x2_t __b = (b); \ |
| 8101 | __builtin_neon_vst2_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], __c, 6); }) |
| 8102 | |
| 8103 | #define vst3q_u64(__a, b) __extension__ ({ \ |
| 8104 | uint64x2x3_t __b = (b); \ |
| 8105 | __builtin_neon_vst3q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], 51); }) |
| 8106 | #define vst3q_s64(__a, b) __extension__ ({ \ |
| 8107 | int64x2x3_t __b = (b); \ |
| 8108 | __builtin_neon_vst3q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], 35); }) |
| 8109 | #define vst3q_f64(__a, b) __extension__ ({ \ |
| 8110 | float64x2x3_t __b = (b); \ |
| 8111 | __builtin_neon_vst3q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], 41); }) |
| 8112 | #define vst3_f64(__a, b) __extension__ ({ \ |
| 8113 | float64x1x3_t __b = (b); \ |
| 8114 | __builtin_neon_vst3_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], 9); }) |
| 8115 | #define vst3_p64(__a, b) __extension__ ({ \ |
| 8116 | poly64x1x3_t __b = (b); \ |
| 8117 | __builtin_neon_vst3_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], 6); }) |
| 8118 | #define vst3q_p64(__a, b) __extension__ ({ \ |
| 8119 | poly64x2x3_t __b = (b); \ |
| 8120 | __builtin_neon_vst3q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], 38); }) |
| 8121 | |
| 8122 | #define vst3q_lane_u8(__a, b, __c) __extension__ ({ \ |
| 8123 | uint8x16x3_t __b = (b); \ |
| 8124 | __builtin_neon_vst3q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], __c, 48); }) |
| 8125 | #define vst3q_lane_u64(__a, b, __c) __extension__ ({ \ |
| 8126 | uint64x2x3_t __b = (b); \ |
| 8127 | __builtin_neon_vst3q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], __c, 51); }) |
| 8128 | #define vst3q_lane_s8(__a, b, __c) __extension__ ({ \ |
| 8129 | int8x16x3_t __b = (b); \ |
| 8130 | __builtin_neon_vst3q_lane_v(__a, __b.val[0], __b.val[1], __b.val[2], __c, 32); }) |
| 8131 | #define vst3q_lane_s64(__a, b, __c) __extension__ ({ \ |
| 8132 | int64x2x3_t __b = (b); \ |
| 8133 | __builtin_neon_vst3q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], __c, 35); }) |
| 8134 | #define vst3q_lane_f64(__a, b, __c) __extension__ ({ \ |
| 8135 | float64x2x3_t __b = (b); \ |
| 8136 | __builtin_neon_vst3q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], __c, 41); }) |
| 8137 | #define vst3q_lane_p8(__a, b, __c) __extension__ ({ \ |
| 8138 | poly8x16x3_t __b = (b); \ |
| 8139 | __builtin_neon_vst3q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], __c, 36); }) |
| 8140 | #define vst3q_lane_p64(__a, b, __c) __extension__ ({ \ |
| 8141 | poly64x2x3_t __b = (b); \ |
| 8142 | __builtin_neon_vst3q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], __c, 38); }) |
| 8143 | #define vst3_lane_u64(__a, b, __c) __extension__ ({ \ |
| 8144 | uint64x1x3_t __b = (b); \ |
| 8145 | __builtin_neon_vst3_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], __c, 19); }) |
| 8146 | #define vst3_lane_s64(__a, b, __c) __extension__ ({ \ |
| 8147 | int64x1x3_t __b = (b); \ |
| 8148 | __builtin_neon_vst3_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], __c, 3); }) |
| 8149 | #define vst3_lane_f64(__a, b, __c) __extension__ ({ \ |
| 8150 | float64x1x3_t __b = (b); \ |
| 8151 | __builtin_neon_vst3_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], __c, 9); }) |
| 8152 | #define vst3_lane_p64(__a, b, __c) __extension__ ({ \ |
| 8153 | poly64x1x3_t __b = (b); \ |
| 8154 | __builtin_neon_vst3_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], __c, 6); }) |
| 8155 | |
| 8156 | #define vst4q_u64(__a, b) __extension__ ({ \ |
| 8157 | uint64x2x4_t __b = (b); \ |
| 8158 | __builtin_neon_vst4q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], 51); }) |
| 8159 | #define vst4q_s64(__a, b) __extension__ ({ \ |
| 8160 | int64x2x4_t __b = (b); \ |
| 8161 | __builtin_neon_vst4q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], 35); }) |
| 8162 | #define vst4q_f64(__a, b) __extension__ ({ \ |
| 8163 | float64x2x4_t __b = (b); \ |
| 8164 | __builtin_neon_vst4q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], 41); }) |
| 8165 | #define vst4_f64(__a, b) __extension__ ({ \ |
| 8166 | float64x1x4_t __b = (b); \ |
| 8167 | __builtin_neon_vst4_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], 9); }) |
| 8168 | #define vst4_p64(__a, b) __extension__ ({ \ |
| 8169 | poly64x1x4_t __b = (b); \ |
| 8170 | __builtin_neon_vst4_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], 6); }) |
| 8171 | #define vst4q_p64(__a, b) __extension__ ({ \ |
| 8172 | poly64x2x4_t __b = (b); \ |
| 8173 | __builtin_neon_vst4q_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], 38); }) |
| 8174 | |
| 8175 | #define vst4q_lane_u8(__a, b, __c) __extension__ ({ \ |
| 8176 | uint8x16x4_t __b = (b); \ |
| 8177 | __builtin_neon_vst4q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], __c, 48); }) |
| 8178 | #define vst4q_lane_u64(__a, b, __c) __extension__ ({ \ |
| 8179 | uint64x2x4_t __b = (b); \ |
| 8180 | __builtin_neon_vst4q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], __c, 51); }) |
| 8181 | #define vst4q_lane_s8(__a, b, __c) __extension__ ({ \ |
| 8182 | int8x16x4_t __b = (b); \ |
| 8183 | __builtin_neon_vst4q_lane_v(__a, __b.val[0], __b.val[1], __b.val[2], __b.val[3], __c, 32); }) |
| 8184 | #define vst4q_lane_s64(__a, b, __c) __extension__ ({ \ |
| 8185 | int64x2x4_t __b = (b); \ |
| 8186 | __builtin_neon_vst4q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], __c, 35); }) |
| 8187 | #define vst4q_lane_f64(__a, b, __c) __extension__ ({ \ |
| 8188 | float64x2x4_t __b = (b); \ |
| 8189 | __builtin_neon_vst4q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], __c, 41); }) |
| 8190 | #define vst4q_lane_p8(__a, b, __c) __extension__ ({ \ |
| 8191 | poly8x16x4_t __b = (b); \ |
| 8192 | __builtin_neon_vst4q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], __c, 36); }) |
| 8193 | #define vst4q_lane_p64(__a, b, __c) __extension__ ({ \ |
| 8194 | poly64x2x4_t __b = (b); \ |
| 8195 | __builtin_neon_vst4q_lane_v(__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], __c, 38); }) |
| 8196 | #define vst4_lane_u64(__a, b, __c) __extension__ ({ \ |
| 8197 | uint64x1x4_t __b = (b); \ |
| 8198 | __builtin_neon_vst4_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], __c, 19); }) |
| 8199 | #define vst4_lane_s64(__a, b, __c) __extension__ ({ \ |
| 8200 | int64x1x4_t __b = (b); \ |
| 8201 | __builtin_neon_vst4_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], __c, 3); }) |
| 8202 | #define vst4_lane_f64(__a, b, __c) __extension__ ({ \ |
| 8203 | float64x1x4_t __b = (b); \ |
| 8204 | __builtin_neon_vst4_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], __c, 9); }) |
| 8205 | #define vst4_lane_p64(__a, b, __c) __extension__ ({ \ |
| 8206 | poly64x1x4_t __b = (b); \ |
| 8207 | __builtin_neon_vst4_lane_v(__a, (int8x8_t)__b.val[0], (int8x8_t)__b.val[1], (int8x8_t)__b.val[2], (int8x8_t)__b.val[3], __c, 6); }) |
| 8208 | |
| 8209 | __ai float64x1_t vsub_f64(float64x1_t __a, float64x1_t __b) { |
| 8210 | return __a - __b; } |
| 8211 | __ai float64x2_t vsubq_f64(float64x2_t __a, float64x2_t __b) { |
| 8212 | return __a - __b; } |
| 8213 | |
| 8214 | __ai int8x8_t vuqadd_s8(int8x8_t __a, int8x8_t __b) { |
| 8215 | return (int8x8_t)__builtin_neon_vuqadd_v(__a, __b, 0); } |
| 8216 | __ai int16x4_t vuqadd_s16(int16x4_t __a, int16x4_t __b) { |
| 8217 | return (int16x4_t)__builtin_neon_vuqadd_v((int8x8_t)__a, (int8x8_t)__b, 1); } |
| 8218 | __ai int32x2_t vuqadd_s32(int32x2_t __a, int32x2_t __b) { |
| 8219 | return (int32x2_t)__builtin_neon_vuqadd_v((int8x8_t)__a, (int8x8_t)__b, 2); } |
| 8220 | __ai int64x1_t vuqadd_s64(int64x1_t __a, int64x1_t __b) { |
| 8221 | return (int64x1_t)__builtin_neon_vuqadd_v((int8x8_t)__a, (int8x8_t)__b, 3); } |
| 8222 | __ai int8x16_t vuqaddq_s8(int8x16_t __a, int8x16_t __b) { |
| 8223 | return (int8x16_t)__builtin_neon_vuqaddq_v(__a, __b, 32); } |
| 8224 | __ai int16x8_t vuqaddq_s16(int16x8_t __a, int16x8_t __b) { |
| 8225 | return (int16x8_t)__builtin_neon_vuqaddq_v((int8x16_t)__a, (int8x16_t)__b, 33); } |
| 8226 | __ai int32x4_t vuqaddq_s32(int32x4_t __a, int32x4_t __b) { |
| 8227 | return (int32x4_t)__builtin_neon_vuqaddq_v((int8x16_t)__a, (int8x16_t)__b, 34); } |
| 8228 | __ai int64x2_t vuqaddq_s64(int64x2_t __a, int64x2_t __b) { |
| 8229 | return (int64x2_t)__builtin_neon_vuqaddq_v((int8x16_t)__a, (int8x16_t)__b, 35); } |
| 8230 | |
| 8231 | __ai uint8x8_t vsqadd_u8(uint8x8_t __a, uint8x8_t __b) { |
| 8232 | return (uint8x8_t)__builtin_neon_vsqadd_v((int8x8_t)__a, (int8x8_t)__b, 16); } |
| 8233 | __ai uint16x4_t vsqadd_u16(uint16x4_t __a, uint16x4_t __b) { |
| 8234 | return (uint16x4_t)__builtin_neon_vsqadd_v((int8x8_t)__a, (int8x8_t)__b, 17); } |
| 8235 | __ai uint32x2_t vsqadd_u32(uint32x2_t __a, uint32x2_t __b) { |
| 8236 | return (uint32x2_t)__builtin_neon_vsqadd_v((int8x8_t)__a, (int8x8_t)__b, 18); } |
| 8237 | __ai uint64x1_t vsqadd_u64(uint64x1_t __a, uint64x1_t __b) { |
| 8238 | return (uint64x1_t)__builtin_neon_vsqadd_v((int8x8_t)__a, (int8x8_t)__b, 19); } |
| 8239 | __ai uint8x16_t vsqaddq_u8(uint8x16_t __a, uint8x16_t __b) { |
| 8240 | return (uint8x16_t)__builtin_neon_vsqaddq_v((int8x16_t)__a, (int8x16_t)__b, 48); } |
| 8241 | __ai uint16x8_t vsqaddq_u16(uint16x8_t __a, uint16x8_t __b) { |
| 8242 | return (uint16x8_t)__builtin_neon_vsqaddq_v((int8x16_t)__a, (int8x16_t)__b, 49); } |
| 8243 | __ai uint32x4_t vsqaddq_u32(uint32x4_t __a, uint32x4_t __b) { |
| 8244 | return (uint32x4_t)__builtin_neon_vsqaddq_v((int8x16_t)__a, (int8x16_t)__b, 50); } |
| 8245 | __ai uint64x2_t vsqaddq_u64(uint64x2_t __a, uint64x2_t __b) { |
| 8246 | return (uint64x2_t)__builtin_neon_vsqaddq_v((int8x16_t)__a, (int8x16_t)__b, 51); } |
| 8247 | |
| 8248 | __ai int16x8_t vabal_high_s8(int16x8_t __a, int8x16_t __b, int8x16_t __c) { |
| 8249 | return vabal_s8(__a, vget_high_s8(__b), vget_high_s8(__c)); } |
| 8250 | __ai int32x4_t vabal_high_s16(int32x4_t __a, int16x8_t __b, int16x8_t __c) { |
| 8251 | return vabal_s16(__a, vget_high_s16(__b), vget_high_s16(__c)); } |
| 8252 | __ai int64x2_t vabal_high_s32(int64x2_t __a, int32x4_t __b, int32x4_t __c) { |
| 8253 | return vabal_s32(__a, vget_high_s32(__b), vget_high_s32(__c)); } |
| 8254 | __ai uint16x8_t vabal_high_u8(uint16x8_t __a, uint8x16_t __b, uint8x16_t __c) { |
| 8255 | return vabal_u8(__a, vget_high_u8(__b), vget_high_u8(__c)); } |
| 8256 | __ai uint32x4_t vabal_high_u16(uint32x4_t __a, uint16x8_t __b, uint16x8_t __c) { |
| 8257 | return vabal_u16(__a, vget_high_u16(__b), vget_high_u16(__c)); } |
| 8258 | __ai uint64x2_t vabal_high_u32(uint64x2_t __a, uint32x4_t __b, uint32x4_t __c) { |
| 8259 | return vabal_u32(__a, vget_high_u32(__b), vget_high_u32(__c)); } |
| 8260 | |
| 8261 | |
| 8262 | __ai int8x16_t vaddhn_high_s16(int8x8_t __a, int16x8_t __b, int16x8_t __c) { |
| 8263 | return vcombine_s8(__a, vaddhn_s16(__b, __c)); } |
| 8264 | __ai int16x8_t vaddhn_high_s32(int16x4_t __a, int32x4_t __b, int32x4_t __c) { |
| 8265 | return vcombine_s16(__a, vaddhn_s32(__b, __c)); } |
| 8266 | __ai int32x4_t vaddhn_high_s64(int32x2_t __a, int64x2_t __b, int64x2_t __c) { |
| 8267 | return vcombine_s32(__a, vaddhn_s64(__b, __c)); } |
| 8268 | __ai uint8x16_t vaddhn_high_u16(uint8x8_t __a, uint16x8_t __b, uint16x8_t __c) { |
| 8269 | return vcombine_u8(__a, vaddhn_u16(__b, __c)); } |
| 8270 | __ai uint16x8_t vaddhn_high_u32(uint16x4_t __a, uint32x4_t __b, uint32x4_t __c) { |
| 8271 | return vcombine_u16(__a, vaddhn_u32(__b, __c)); } |
| 8272 | __ai uint32x4_t vaddhn_high_u64(uint32x2_t __a, uint64x2_t __b, uint64x2_t __c) { |
| 8273 | return vcombine_u32(__a, vaddhn_u64(__b, __c)); } |
| 8274 | |
| 8275 | __ai int16_t vaddlv_s8(int8x8_t __a) { |
| 8276 | return (int16_t)__builtin_neon_vaddlv_s8(__a); } |
| 8277 | __ai int32_t vaddlv_s16(int16x4_t __a) { |
| 8278 | return (int32_t)__builtin_neon_vaddlv_s16(__a); } |
| 8279 | __ai int64_t vaddlv_s32(int32x2_t __a) { |
| 8280 | return (int64_t)__builtin_neon_vaddlv_s32(__a); } |
| 8281 | __ai uint16_t vaddlv_u8(uint8x8_t __a) { |
| 8282 | return (uint16_t)__builtin_neon_vaddlv_u8((int8x8_t)__a); } |
| 8283 | __ai uint32_t vaddlv_u16(uint16x4_t __a) { |
| 8284 | return (uint32_t)__builtin_neon_vaddlv_u16((int16x4_t)__a); } |
| 8285 | __ai uint64_t vaddlv_u32(uint32x2_t __a) { |
| 8286 | return (uint64_t)__builtin_neon_vaddlv_u32((int32x2_t)__a); } |
| 8287 | __ai int16_t vaddlvq_s8(int8x16_t __a) { |
| 8288 | return (int16_t)__builtin_neon_vaddlvq_s8(__a); } |
| 8289 | __ai int32_t vaddlvq_s16(int16x8_t __a) { |
| 8290 | return (int32_t)__builtin_neon_vaddlvq_s16(__a); } |
| 8291 | __ai int64_t vaddlvq_s32(int32x4_t __a) { |
| 8292 | return (int64_t)__builtin_neon_vaddlvq_s32(__a); } |
| 8293 | __ai uint16_t vaddlvq_u8(uint8x16_t __a) { |
| 8294 | return (uint16_t)__builtin_neon_vaddlvq_u8((int8x16_t)__a); } |
| 8295 | __ai uint32_t vaddlvq_u16(uint16x8_t __a) { |
| 8296 | return (uint32_t)__builtin_neon_vaddlvq_u16((int16x8_t)__a); } |
| 8297 | __ai uint64_t vaddlvq_u32(uint32x4_t __a) { |
| 8298 | return (uint64_t)__builtin_neon_vaddlvq_u32((int32x4_t)__a); } |
| 8299 | |
| 8300 | __ai int16x8_t vaddl_high_s8(int8x16_t __a, int8x16_t __b) { |
| 8301 | return vmovl_high_s8(__a) + vmovl_high_s8(__b); } |
| 8302 | __ai int32x4_t vaddl_high_s16(int16x8_t __a, int16x8_t __b) { |
| 8303 | return vmovl_high_s16(__a) + vmovl_high_s16(__b); } |
| 8304 | __ai int64x2_t vaddl_high_s32(int32x4_t __a, int32x4_t __b) { |
| 8305 | return vmovl_high_s32(__a) + vmovl_high_s32(__b); } |
| 8306 | __ai uint16x8_t vaddl_high_u8(uint8x16_t __a, uint8x16_t __b) { |
| 8307 | return vmovl_high_u8(__a) + vmovl_high_u8(__b); } |
| 8308 | __ai uint32x4_t vaddl_high_u16(uint16x8_t __a, uint16x8_t __b) { |
| 8309 | return vmovl_high_u16(__a) + vmovl_high_u16(__b); } |
| 8310 | __ai uint64x2_t vaddl_high_u32(uint32x4_t __a, uint32x4_t __b) { |
| 8311 | return vmovl_high_u32(__a) + vmovl_high_u32(__b); } |
| 8312 | |
| 8313 | __ai int8_t vaddv_s8(int8x8_t __a) { |
| 8314 | return (int8_t)__builtin_neon_vaddv_s8(__a); } |
| 8315 | __ai int16_t vaddv_s16(int16x4_t __a) { |
| 8316 | return (int16_t)__builtin_neon_vaddv_s16(__a); } |
| 8317 | __ai int32_t vaddv_s32(int32x2_t __a) { |
| 8318 | return (int32_t)__builtin_neon_vaddv_s32(__a); } |
| 8319 | __ai float32_t vaddv_f32(float32x2_t __a) { |
| 8320 | return (float32_t)__builtin_neon_vaddv_f32(__a); } |
| 8321 | __ai uint8_t vaddv_u8(uint8x8_t __a) { |
| 8322 | return (uint8_t)__builtin_neon_vaddv_u8((int8x8_t)__a); } |
| 8323 | __ai uint16_t vaddv_u16(uint16x4_t __a) { |
| 8324 | return (uint16_t)__builtin_neon_vaddv_u16((int16x4_t)__a); } |
| 8325 | __ai uint32_t vaddv_u32(uint32x2_t __a) { |
| 8326 | return (uint32_t)__builtin_neon_vaddv_u32((int32x2_t)__a); } |
| 8327 | __ai int8_t vaddvq_s8(int8x16_t __a) { |
| 8328 | return (int8_t)__builtin_neon_vaddvq_s8(__a); } |
| 8329 | __ai int16_t vaddvq_s16(int16x8_t __a) { |
| 8330 | return (int16_t)__builtin_neon_vaddvq_s16(__a); } |
| 8331 | __ai int32_t vaddvq_s32(int32x4_t __a) { |
| 8332 | return (int32_t)__builtin_neon_vaddvq_s32(__a); } |
| 8333 | __ai uint8_t vaddvq_u8(uint8x16_t __a) { |
| 8334 | return (uint8_t)__builtin_neon_vaddvq_u8((int8x16_t)__a); } |
| 8335 | __ai uint16_t vaddvq_u16(uint16x8_t __a) { |
| 8336 | return (uint16_t)__builtin_neon_vaddvq_u16((int16x8_t)__a); } |
| 8337 | __ai uint32_t vaddvq_u32(uint32x4_t __a) { |
| 8338 | return (uint32_t)__builtin_neon_vaddvq_u32((int32x4_t)__a); } |
| 8339 | __ai float32_t vaddvq_f32(float32x4_t __a) { |
| 8340 | return (float32_t)__builtin_neon_vaddvq_f32(__a); } |
| 8341 | __ai float64_t vaddvq_f64(float64x2_t __a) { |
| 8342 | return (float64_t)__builtin_neon_vaddvq_f64(__a); } |
| 8343 | __ai int64_t vaddvq_s64(int64x2_t __a) { |
| 8344 | return (int64_t)__builtin_neon_vaddvq_s64(__a); } |
| 8345 | __ai uint64_t vaddvq_u64(uint64x2_t __a) { |
| 8346 | return (uint64_t)__builtin_neon_vaddvq_u64((int64x2_t)__a); } |
| 8347 | |
| 8348 | __ai int16x8_t vaddw_high_s8(int16x8_t __a, int8x16_t __b) { |
| 8349 | return __a + vmovl_high_s8(__b); } |
| 8350 | __ai int32x4_t vaddw_high_s16(int32x4_t __a, int16x8_t __b) { |
| 8351 | return __a + vmovl_high_s16(__b); } |
| 8352 | __ai int64x2_t vaddw_high_s32(int64x2_t __a, int32x4_t __b) { |
| 8353 | return __a + vmovl_high_s32(__b); } |
| 8354 | __ai uint16x8_t vaddw_high_u8(uint16x8_t __a, uint8x16_t __b) { |
| 8355 | return __a + vmovl_high_u8(__b); } |
| 8356 | __ai uint32x4_t vaddw_high_u16(uint32x4_t __a, uint16x8_t __b) { |
| 8357 | return __a + vmovl_high_u16(__b); } |
| 8358 | __ai uint64x2_t vaddw_high_u32(uint64x2_t __a, uint32x4_t __b) { |
| 8359 | return __a + vmovl_high_u32(__b); } |
| 8360 | |
| 8361 | __ai float32x2_t vcvtx_f32_f64(float64x2_t __a) { |
| 8362 | return (float32x2_t)__builtin_neon_vcvtx_f32_v((int8x16_t)__a, 41); } |
| 8363 | |
| 8364 | __ai float32x4_t vcvtx_high_f32_f64(float32x2_t __a, float64x2_t __b) { |
| 8365 | float32x2_t __a1 = vcvtx_f32_f64(__b); |
| 8366 | return __builtin_shufflevector(__a, __a1, 0, 1, 2, 3); } |
| 8367 | |
| 8368 | __ai float32x2_t vcvt_f32_f64(float64x2_t __a) { |
| 8369 | return (float32x2_t)__builtin_neon_vcvt_f32_f64((int8x16_t)__a, 41); } |
| 8370 | |
| 8371 | __ai float64x1_t vcvt_f64_s64(int64x1_t __a) { |
| 8372 | return (float64x1_t)__builtin_neon_vcvt_f64_v((int8x8_t)__a, 3); } |
| 8373 | __ai float64x1_t vcvt_f64_u64(uint64x1_t __a) { |
| 8374 | return (float64x1_t)__builtin_neon_vcvt_f64_v((int8x8_t)__a, 19); } |
| 8375 | __ai float64x2_t vcvtq_f64_s64(int64x2_t __a) { |
| 8376 | return (float64x2_t)__builtin_neon_vcvtq_f64_v((int8x16_t)__a, 35); } |
| 8377 | __ai float64x2_t vcvtq_f64_u64(uint64x2_t __a) { |
| 8378 | return (float64x2_t)__builtin_neon_vcvtq_f64_v((int8x16_t)__a, 51); } |
| 8379 | |
| 8380 | __ai float64x2_t vcvt_f64_f32(float32x2_t __a) { |
| 8381 | return (float64x2_t)__builtin_neon_vcvt_f64_f32((int8x8_t)__a, 41); } |
| 8382 | |
| 8383 | __ai float16x8_t vcvt_high_f16_f32(float16x4_t __a, float32x4_t __b) { |
| 8384 | float16x4_t __a1 = vcvt_f16_f32(__b); |
| 8385 | return __builtin_shufflevector(__a, __a1, 0, 1, 2, 3, 4, 5, 6, 7); } |
| 8386 | |
| 8387 | __ai float32x4_t vcvt_high_f32_f16(float16x8_t __a) { |
| 8388 | float16x4_t __a1 = vget_high_f16(__a); |
| 8389 | return vcvt_f32_f16(__a1); } |
| 8390 | |
| 8391 | __ai float32x4_t vcvt_high_f32_f64(float32x2_t __a, float64x2_t __b) { |
| 8392 | float32x2_t __a1 = vcvt_f32_f64(__b); |
| 8393 | return __builtin_shufflevector(__a, __a1, 0, 1, 2, 3); } |
| 8394 | |
| 8395 | __ai float64x2_t vcvt_high_f64_f32(float32x4_t __a) { |
| 8396 | float32x2_t __a1 = vget_high_f32(__a); |
| 8397 | return vcvt_f64_f32(__a1); } |
| 8398 | |
| 8399 | __ai int64x1_t vcvt_s64_f64(float64x1_t __a) { |
| 8400 | return (int64x1_t)__builtin_neon_vcvt_s64_v((int8x8_t)__a, 3); } |
| 8401 | __ai int64x2_t vcvtq_s64_f64(float64x2_t __a) { |
| 8402 | return (int64x2_t)__builtin_neon_vcvtq_s64_v((int8x16_t)__a, 35); } |
| 8403 | |
| 8404 | __ai uint64x1_t vcvt_u64_f64(float64x1_t __a) { |
| 8405 | return (uint64x1_t)__builtin_neon_vcvt_u64_v((int8x8_t)__a, 19); } |
| 8406 | __ai uint64x2_t vcvtq_u64_f64(float64x2_t __a) { |
| 8407 | return (uint64x2_t)__builtin_neon_vcvtq_u64_v((int8x16_t)__a, 51); } |
| 8408 | |
| 8409 | #define vdup_lane_f16(a, __b) __extension__ ({ \ |
| 8410 | float16x4_t __a = (a); \ |
| 8411 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b); }) |
| 8412 | #define vdup_lane_f64(a, __b) __extension__ ({ \ |
| 8413 | float64x1_t __a = (a); \ |
| 8414 | __builtin_shufflevector(__a, __a, __b); }) |
| 8415 | #define vdupq_lane_f16(a, __b) __extension__ ({ \ |
| 8416 | float16x4_t __a = (a); \ |
| 8417 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b, __b, __b, __b, __b); }) |
| 8418 | #define vdupq_lane_f64(a, __b) __extension__ ({ \ |
| 8419 | float64x1_t __a = (a); \ |
| 8420 | __builtin_shufflevector(__a, __a, __b, __b); }) |
| 8421 | #define vdup_lane_p64(a, __b) __extension__ ({ \ |
| 8422 | poly64x1_t __a = (a); \ |
| 8423 | __builtin_shufflevector(__a, __a, __b); }) |
| 8424 | #define vdupq_lane_p64(a, __b) __extension__ ({ \ |
| 8425 | poly64x1_t __a = (a); \ |
| 8426 | __builtin_shufflevector(__a, __a, __b, __b); }) |
| 8427 | |
| 8428 | #define vdup_laneq_s8(a, __b) __extension__ ({ \ |
| 8429 | int8x16_t __a = (a); \ |
| 8430 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b, __b, __b, __b, __b); }) |
| 8431 | #define vdup_laneq_s16(a, __b) __extension__ ({ \ |
| 8432 | int16x8_t __a = (a); \ |
| 8433 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b); }) |
| 8434 | #define vdup_laneq_s32(a, __b) __extension__ ({ \ |
| 8435 | int32x4_t __a = (a); \ |
| 8436 | __builtin_shufflevector(__a, __a, __b, __b); }) |
| 8437 | #define vdup_laneq_s64(a, __b) __extension__ ({ \ |
| 8438 | int64x2_t __a = (a); \ |
| 8439 | __builtin_shufflevector(__a, __a, __b); }) |
| 8440 | #define vdup_laneq_p8(a, __b) __extension__ ({ \ |
| 8441 | poly8x16_t __a = (a); \ |
| 8442 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b, __b, __b, __b, __b); }) |
| 8443 | #define vdup_laneq_p16(a, __b) __extension__ ({ \ |
| 8444 | poly16x8_t __a = (a); \ |
| 8445 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b); }) |
| 8446 | #define vdup_laneq_u8(a, __b) __extension__ ({ \ |
| 8447 | uint8x16_t __a = (a); \ |
| 8448 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b, __b, __b, __b, __b); }) |
| 8449 | #define vdup_laneq_u16(a, __b) __extension__ ({ \ |
| 8450 | uint16x8_t __a = (a); \ |
| 8451 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b); }) |
| 8452 | #define vdup_laneq_u32(a, __b) __extension__ ({ \ |
| 8453 | uint32x4_t __a = (a); \ |
| 8454 | __builtin_shufflevector(__a, __a, __b, __b); }) |
| 8455 | #define vdup_laneq_u64(a, __b) __extension__ ({ \ |
| 8456 | uint64x2_t __a = (a); \ |
| 8457 | __builtin_shufflevector(__a, __a, __b); }) |
| 8458 | #define vdup_laneq_f16(a, __b) __extension__ ({ \ |
| 8459 | float16x8_t __a = (a); \ |
| 8460 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b); }) |
| 8461 | #define vdup_laneq_f32(a, __b) __extension__ ({ \ |
| 8462 | float32x4_t __a = (a); \ |
| 8463 | __builtin_shufflevector(__a, __a, __b, __b); }) |
| 8464 | #define vdup_laneq_f64(a, __b) __extension__ ({ \ |
| 8465 | float64x2_t __a = (a); \ |
| 8466 | __builtin_shufflevector(__a, __a, __b); }) |
| 8467 | #define vdupq_laneq_s8(a, __b) __extension__ ({ \ |
| 8468 | int8x16_t __a = (a); \ |
| 8469 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b); }) |
| 8470 | #define vdupq_laneq_s16(a, __b) __extension__ ({ \ |
| 8471 | int16x8_t __a = (a); \ |
| 8472 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b, __b, __b, __b, __b); }) |
| 8473 | #define vdupq_laneq_s32(a, __b) __extension__ ({ \ |
| 8474 | int32x4_t __a = (a); \ |
| 8475 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b); }) |
| 8476 | #define vdupq_laneq_s64(a, __b) __extension__ ({ \ |
| 8477 | int64x2_t __a = (a); \ |
| 8478 | __builtin_shufflevector(__a, __a, __b, __b); }) |
| 8479 | #define vdupq_laneq_p8(a, __b) __extension__ ({ \ |
| 8480 | poly8x16_t __a = (a); \ |
| 8481 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b); }) |
| 8482 | #define vdupq_laneq_p16(a, __b) __extension__ ({ \ |
| 8483 | poly16x8_t __a = (a); \ |
| 8484 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b, __b, __b, __b, __b); }) |
| 8485 | #define vdupq_laneq_u8(a, __b) __extension__ ({ \ |
| 8486 | uint8x16_t __a = (a); \ |
| 8487 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b, __b); }) |
| 8488 | #define vdupq_laneq_u16(a, __b) __extension__ ({ \ |
| 8489 | uint16x8_t __a = (a); \ |
| 8490 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b, __b, __b, __b, __b); }) |
| 8491 | #define vdupq_laneq_u32(a, __b) __extension__ ({ \ |
| 8492 | uint32x4_t __a = (a); \ |
| 8493 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b); }) |
| 8494 | #define vdupq_laneq_u64(a, __b) __extension__ ({ \ |
| 8495 | uint64x2_t __a = (a); \ |
| 8496 | __builtin_shufflevector(__a, __a, __b, __b); }) |
| 8497 | #define vdupq_laneq_f16(a, __b) __extension__ ({ \ |
| 8498 | float16x8_t __a = (a); \ |
| 8499 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b, __b, __b, __b, __b); }) |
| 8500 | #define vdupq_laneq_f32(a, __b) __extension__ ({ \ |
| 8501 | float32x4_t __a = (a); \ |
| 8502 | __builtin_shufflevector(__a, __a, __b, __b, __b, __b); }) |
| 8503 | #define vdupq_laneq_f64(a, __b) __extension__ ({ \ |
| 8504 | float64x2_t __a = (a); \ |
| 8505 | __builtin_shufflevector(__a, __a, __b, __b); }) |
| 8506 | #define vdup_laneq_p64(a, __b) __extension__ ({ \ |
| 8507 | poly64x2_t __a = (a); \ |
| 8508 | __builtin_shufflevector(__a, __a, __b); }) |
| 8509 | #define vdupq_laneq_p64(a, __b) __extension__ ({ \ |
| 8510 | poly64x2_t __a = (a); \ |
| 8511 | __builtin_shufflevector(__a, __a, __b, __b); }) |
| 8512 | |
| 8513 | #define vext_f64(a, b, __c) __extension__ ({ \ |
| 8514 | float64x1_t __a = (a); float64x1_t __b = (b); \ |
| 8515 | (float64x1_t)__builtin_neon_vext_v((int8x8_t)__a, (int8x8_t)__b, __c, 9); }) |
| 8516 | #define vextq_f64(a, b, __c) __extension__ ({ \ |
| 8517 | float64x2_t __a = (a); float64x2_t __b = (b); \ |
| 8518 | (float64x2_t)__builtin_neon_vextq_v((int8x16_t)__a, (int8x16_t)__b, __c, 41); }) |
| 8519 | #define vext_p64(a, b, __c) __extension__ ({ \ |
| 8520 | poly64x1_t __a = (a); poly64x1_t __b = (b); \ |
| 8521 | (poly64x1_t)__builtin_neon_vext_v((int8x8_t)__a, (int8x8_t)__b, __c, 6); }) |
| 8522 | #define vextq_p64(a, b, __c) __extension__ ({ \ |
| 8523 | poly64x2_t __a = (a); poly64x2_t __b = (b); \ |
| 8524 | (poly64x2_t)__builtin_neon_vextq_v((int8x16_t)__a, (int8x16_t)__b, __c, 38); }) |
| 8525 | |
| 8526 | #define vfma_lane_f32(a, b, c, __d) __extension__ ({ \ |
| 8527 | float32x2_t __a = (a); float32x2_t __b = (b); float32x2_t __c = (c); \ |
| 8528 | (float32x2_t)__builtin_neon_vfma_lane_v((int8x8_t)__a, (int8x8_t)__b, (int8x8_t)__c, __d, 8); }) |
| 8529 | #define vfma_lane_f64(a, b, c, __d) __extension__ ({ \ |
| 8530 | float64x1_t __a = (a); float64x1_t __b = (b); float64x1_t __c = (c); \ |
| 8531 | (float64x1_t)__builtin_neon_vfma_lane_v((int8x8_t)__a, (int8x8_t)__b, (int8x8_t)__c, __d, 9); }) |
| 8532 | #define vfmaq_lane_f32(a, b, c, __d) __extension__ ({ \ |
| 8533 | float32x4_t __a = (a); float32x4_t __b = (b); float32x2_t __c = (c); \ |
| 8534 | (float32x4_t)__builtin_neon_vfmaq_lane_v((int8x16_t)__a, (int8x16_t)__b, (int8x8_t)__c, __d, 40); }) |
| 8535 | #define vfmaq_lane_f64(a, b, c, __d) __extension__ ({ \ |
| 8536 | float64x2_t __a = (a); float64x2_t __b = (b); float64x1_t __c = (c); \ |
| 8537 | (float64x2_t)__builtin_neon_vfmaq_lane_v((int8x16_t)__a, (int8x16_t)__b, (int8x8_t)__c, __d, 41); }) |
| 8538 | |
| 8539 | #define vfma_laneq_f32(a, b, c, __d) __extension__ ({ \ |
| 8540 | float32x2_t __a = (a); float32x2_t __b = (b); float32x4_t __c = (c); \ |
| 8541 | (float32x2_t)__builtin_neon_vfma_laneq_v((int8x8_t)__a, (int8x8_t)__b, (int8x16_t)__c, __d, 8); }) |
| 8542 | #define vfma_laneq_f64(a, b, c, __d) __extension__ ({ \ |
| 8543 | float64x1_t __a = (a); float64x1_t __b = (b); float64x2_t __c = (c); \ |
| 8544 | (float64x1_t)__builtin_neon_vfma_laneq_v((int8x8_t)__a, (int8x8_t)__b, (int8x16_t)__c, __d, 9); }) |
| 8545 | #define vfmaq_laneq_f32(a, b, c, __d) __extension__ ({ \ |
| 8546 | float32x4_t __a = (a); float32x4_t __b = (b); float32x4_t __c = (c); \ |
| 8547 | (float32x4_t)__builtin_neon_vfmaq_laneq_v((int8x16_t)__a, (int8x16_t)__b, (int8x16_t)__c, __d, 40); }) |
| 8548 | #define vfmaq_laneq_f64(a, b, c, __d) __extension__ ({ \ |
| 8549 | float64x2_t __a = (a); float64x2_t __b = (b); float64x2_t __c = (c); \ |
| 8550 | (float64x2_t)__builtin_neon_vfmaq_laneq_v((int8x16_t)__a, (int8x16_t)__b, (int8x16_t)__c, __d, 41); }) |
| 8551 | |
| 8552 | #define vfms_lane_f32(a, b, c, __d) __extension__ ({ \ |
| 8553 | float32x2_t __a = (a); float32x2_t __b = (b); float32x2_t __c = (c); \ |
| 8554 | float32x2_t __a1 = __a; \ |
| 8555 | float32x2_t __b1 = __b; \ |
| 8556 | float32x2_t __c1 = __c; \ |
| 8557 | vfma_lane_f32(__a1, __b1, -__c1, __d); }) |
| 8558 | #define vfms_lane_f64(a, b, c, __d) __extension__ ({ \ |
| 8559 | float64x1_t __a = (a); float64x1_t __b = (b); float64x1_t __c = (c); \ |
| 8560 | float64x1_t __a1 = __a; \ |
| 8561 | float64x1_t __b1 = __b; \ |
| 8562 | float64x1_t __c1 = __c; \ |
| 8563 | vfma_lane_f64(__a1, __b1, -__c1, __d); }) |
| 8564 | #define vfmsq_lane_f32(a, b, c, __d) __extension__ ({ \ |
| 8565 | float32x4_t __a = (a); float32x4_t __b = (b); float32x2_t __c = (c); \ |
| 8566 | float32x4_t __a1 = __a; \ |
| 8567 | float32x4_t __b1 = __b; \ |
| 8568 | float32x2_t __c1 = __c; \ |
| 8569 | vfmaq_lane_f32(__a1, __b1, -__c1, __d); }) |
| 8570 | #define vfmsq_lane_f64(a, b, c, __d) __extension__ ({ \ |
| 8571 | float64x2_t __a = (a); float64x2_t __b = (b); float64x1_t __c = (c); \ |
| 8572 | float64x2_t __a1 = __a; \ |
| 8573 | float64x2_t __b1 = __b; \ |
| 8574 | float64x1_t __c1 = __c; \ |
| 8575 | vfmaq_lane_f64(__a1, __b1, -__c1, __d); }) |
| 8576 | |
| 8577 | #define vfms_laneq_f32(a, b, c, __d) __extension__ ({ \ |
| 8578 | float32x2_t __a = (a); float32x2_t __b = (b); float32x4_t __c = (c); \ |
| 8579 | float32x2_t __a1 = __a; \ |
| 8580 | float32x2_t __b1 = __b; \ |
| 8581 | float32x4_t __c1 = __c; \ |
| 8582 | vfma_laneq_f32(__a1, __b1, -__c1, __d); }) |
| 8583 | #define vfms_laneq_f64(a, b, c, __d) __extension__ ({ \ |
| 8584 | float64x1_t __a = (a); float64x1_t __b = (b); float64x2_t __c = (c); \ |
| 8585 | float64x1_t __a1 = __a; \ |
| 8586 | float64x1_t __b1 = __b; \ |
| 8587 | float64x2_t __c1 = __c; \ |
| 8588 | vfma_laneq_f64(__a1, __b1, -__c1, __d); }) |
| 8589 | #define vfmsq_laneq_f32(a, b, c, __d) __extension__ ({ \ |
| 8590 | float32x4_t __a = (a); float32x4_t __b = (b); float32x4_t __c = (c); \ |
| 8591 | float32x4_t __a1 = __a; \ |
| 8592 | float32x4_t __b1 = __b; \ |
| 8593 | float32x4_t __c1 = __c; \ |
| 8594 | vfmaq_laneq_f32(__a1, __b1, -__c1, __d); }) |
| 8595 | #define vfmsq_laneq_f64(a, b, c, __d) __extension__ ({ \ |
| 8596 | float64x2_t __a = (a); float64x2_t __b = (b); float64x2_t __c = (c); \ |
| 8597 | float64x2_t __a1 = __a; \ |
| 8598 | float64x2_t __b1 = __b; \ |
| 8599 | float64x2_t __c1 = __c; \ |
| 8600 | vfmaq_laneq_f64(__a1, __b1, -__c1, __d); }) |
| 8601 | |
| 8602 | __ai float64x1_t vget_high_f64(float64x2_t __a) { |
| 8603 | return __builtin_shufflevector(__a, __a, 1); } |
| 8604 | __ai poly64x1_t vget_high_p64(poly64x2_t __a) { |
| 8605 | return __builtin_shufflevector(__a, __a, 1); } |
| 8606 | |
| 8607 | __ai float64x1_t vget_low_f64(float64x2_t __a) { |
| 8608 | return __builtin_shufflevector(__a, __a, 0); } |
| 8609 | __ai poly64x1_t vget_low_p64(poly64x2_t __a) { |
| 8610 | return __builtin_shufflevector(__a, __a, 0); } |
| 8611 | |
| 8612 | __ai int8_t vmaxv_s8(int8x8_t __a) { |
| 8613 | return (int8_t)__builtin_neon_vmaxv_s8(__a); } |
| 8614 | __ai int16_t vmaxv_s16(int16x4_t __a) { |
| 8615 | return (int16_t)__builtin_neon_vmaxv_s16(__a); } |
| 8616 | __ai int32_t vmaxv_s32(int32x2_t __a) { |
| 8617 | return (int32_t)__builtin_neon_vmaxv_s32(__a); } |
| 8618 | __ai float32_t vmaxv_f32(float32x2_t __a) { |
| 8619 | return (float32_t)__builtin_neon_vmaxv_f32(__a); } |
| 8620 | __ai uint8_t vmaxv_u8(uint8x8_t __a) { |
| 8621 | return (uint8_t)__builtin_neon_vmaxv_u8((int8x8_t)__a); } |
| 8622 | __ai uint16_t vmaxv_u16(uint16x4_t __a) { |
| 8623 | return (uint16_t)__builtin_neon_vmaxv_u16((int16x4_t)__a); } |
| 8624 | __ai uint32_t vmaxv_u32(uint32x2_t __a) { |
| 8625 | return (uint32_t)__builtin_neon_vmaxv_u32((int32x2_t)__a); } |
| 8626 | __ai int8_t vmaxvq_s8(int8x16_t __a) { |
| 8627 | return (int8_t)__builtin_neon_vmaxvq_s8(__a); } |
| 8628 | __ai int16_t vmaxvq_s16(int16x8_t __a) { |
| 8629 | return (int16_t)__builtin_neon_vmaxvq_s16(__a); } |
| 8630 | __ai int32_t vmaxvq_s32(int32x4_t __a) { |
| 8631 | return (int32_t)__builtin_neon_vmaxvq_s32(__a); } |
| 8632 | __ai uint8_t vmaxvq_u8(uint8x16_t __a) { |
| 8633 | return (uint8_t)__builtin_neon_vmaxvq_u8((int8x16_t)__a); } |
| 8634 | __ai uint16_t vmaxvq_u16(uint16x8_t __a) { |
| 8635 | return (uint16_t)__builtin_neon_vmaxvq_u16((int16x8_t)__a); } |
| 8636 | __ai uint32_t vmaxvq_u32(uint32x4_t __a) { |
| 8637 | return (uint32_t)__builtin_neon_vmaxvq_u32((int32x4_t)__a); } |
| 8638 | __ai float32_t vmaxvq_f32(float32x4_t __a) { |
| 8639 | return (float32_t)__builtin_neon_vmaxvq_f32(__a); } |
| 8640 | __ai float64_t vmaxvq_f64(float64x2_t __a) { |
| 8641 | return (float64_t)__builtin_neon_vmaxvq_f64(__a); } |
| 8642 | |
| 8643 | __ai int8_t vminv_s8(int8x8_t __a) { |
| 8644 | return (int8_t)__builtin_neon_vminv_s8(__a); } |
| 8645 | __ai int16_t vminv_s16(int16x4_t __a) { |
| 8646 | return (int16_t)__builtin_neon_vminv_s16(__a); } |
| 8647 | __ai int32_t vminv_s32(int32x2_t __a) { |
| 8648 | return (int32_t)__builtin_neon_vminv_s32(__a); } |
| 8649 | __ai float32_t vminv_f32(float32x2_t __a) { |
| 8650 | return (float32_t)__builtin_neon_vminv_f32(__a); } |
| 8651 | __ai uint8_t vminv_u8(uint8x8_t __a) { |
| 8652 | return (uint8_t)__builtin_neon_vminv_u8((int8x8_t)__a); } |
| 8653 | __ai uint16_t vminv_u16(uint16x4_t __a) { |
| 8654 | return (uint16_t)__builtin_neon_vminv_u16((int16x4_t)__a); } |
| 8655 | __ai uint32_t vminv_u32(uint32x2_t __a) { |
| 8656 | return (uint32_t)__builtin_neon_vminv_u32((int32x2_t)__a); } |
| 8657 | __ai int8_t vminvq_s8(int8x16_t __a) { |
| 8658 | return (int8_t)__builtin_neon_vminvq_s8(__a); } |
| 8659 | __ai int16_t vminvq_s16(int16x8_t __a) { |
| 8660 | return (int16_t)__builtin_neon_vminvq_s16(__a); } |
| 8661 | __ai int32_t vminvq_s32(int32x4_t __a) { |
| 8662 | return (int32_t)__builtin_neon_vminvq_s32(__a); } |
| 8663 | __ai uint8_t vminvq_u8(uint8x16_t __a) { |
| 8664 | return (uint8_t)__builtin_neon_vminvq_u8((int8x16_t)__a); } |
| 8665 | __ai uint16_t vminvq_u16(uint16x8_t __a) { |
| 8666 | return (uint16_t)__builtin_neon_vminvq_u16((int16x8_t)__a); } |
| 8667 | __ai uint32_t vminvq_u32(uint32x4_t __a) { |
| 8668 | return (uint32_t)__builtin_neon_vminvq_u32((int32x4_t)__a); } |
| 8669 | __ai float32_t vminvq_f32(float32x4_t __a) { |
| 8670 | return (float32_t)__builtin_neon_vminvq_f32(__a); } |
| 8671 | __ai float64_t vminvq_f64(float64x2_t __a) { |
| 8672 | return (float64_t)__builtin_neon_vminvq_f64(__a); } |
| 8673 | |
| 8674 | __ai int16x8_t vmlal_high_s8(int16x8_t __a, int8x16_t __b, int8x16_t __c) { |
| 8675 | return vmlal_s8(__a, vget_high_s8(__b), vget_high_s8(__c)); } |
| 8676 | __ai int32x4_t vmlal_high_s16(int32x4_t __a, int16x8_t __b, int16x8_t __c) { |
| 8677 | return vmlal_s16(__a, vget_high_s16(__b), vget_high_s16(__c)); } |
| 8678 | __ai int64x2_t vmlal_high_s32(int64x2_t __a, int32x4_t __b, int32x4_t __c) { |
| 8679 | return vmlal_s32(__a, vget_high_s32(__b), vget_high_s32(__c)); } |
| 8680 | __ai uint16x8_t vmlal_high_u8(uint16x8_t __a, uint8x16_t __b, uint8x16_t __c) { |
| 8681 | return vmlal_u8(__a, vget_high_u8(__b), vget_high_u8(__c)); } |
| 8682 | __ai uint32x4_t vmlal_high_u16(uint32x4_t __a, uint16x8_t __b, uint16x8_t __c) { |
| 8683 | return vmlal_u16(__a, vget_high_u16(__b), vget_high_u16(__c)); } |
| 8684 | __ai uint64x2_t vmlal_high_u32(uint64x2_t __a, uint32x4_t __b, uint32x4_t __c) { |
| 8685 | return vmlal_u32(__a, vget_high_u32(__b), vget_high_u32(__c)); } |
| 8686 | |
| 8687 | #define vmlal_high_lane_s16(a, b, c, __d) __extension__ ({ \ |
| 8688 | int32x4_t __a = (a); int16x8_t __b = (b); int16x4_t __c = (c); \ |
| 8689 | __a + vmull_s16(vget_high_s16(__b), __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 8690 | #define vmlal_high_lane_s32(a, b, c, __d) __extension__ ({ \ |
| 8691 | int64x2_t __a = (a); int32x4_t __b = (b); int32x2_t __c = (c); \ |
| 8692 | __a + vmull_s32(vget_high_s32(__b), __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 8693 | #define vmlal_high_lane_u16(a, b, c, __d) __extension__ ({ \ |
| 8694 | uint32x4_t __a = (a); uint16x8_t __b = (b); uint16x4_t __c = (c); \ |
| 8695 | __a + vmull_u16(vget_high_u16(__b), __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 8696 | #define vmlal_high_lane_u32(a, b, c, __d) __extension__ ({ \ |
| 8697 | uint64x2_t __a = (a); uint32x4_t __b = (b); uint32x2_t __c = (c); \ |
| 8698 | __a + vmull_u32(vget_high_u32(__b), __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 8699 | |
| 8700 | #define vmlal_high_laneq_s16(a, b, c, __d) __extension__ ({ \ |
| 8701 | int32x4_t __a = (a); int16x8_t __b = (b); int16x8_t __c = (c); \ |
| 8702 | __a + vmull_s16(vget_high_s16(__b), __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 8703 | #define vmlal_high_laneq_s32(a, b, c, __d) __extension__ ({ \ |
| 8704 | int64x2_t __a = (a); int32x4_t __b = (b); int32x4_t __c = (c); \ |
| 8705 | __a + vmull_s32(vget_high_s32(__b), __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 8706 | #define vmlal_high_laneq_u16(a, b, c, __d) __extension__ ({ \ |
| 8707 | uint32x4_t __a = (a); uint16x8_t __b = (b); uint16x8_t __c = (c); \ |
| 8708 | __a + vmull_u16(vget_high_u16(__b), __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 8709 | #define vmlal_high_laneq_u32(a, b, c, __d) __extension__ ({ \ |
| 8710 | uint64x2_t __a = (a); uint32x4_t __b = (b); uint32x4_t __c = (c); \ |
| 8711 | __a + vmull_u32(vget_high_u32(__b), __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 8712 | |
| 8713 | __ai int32x4_t vmlal_high_n_s16(int32x4_t __a, int16x8_t __b, int16_t __c) { |
| 8714 | return vmlal_n_s16(__a, vget_high_s16(__b), __c); } |
| 8715 | __ai int64x2_t vmlal_high_n_s32(int64x2_t __a, int32x4_t __b, int32_t __c) { |
| 8716 | return vmlal_n_s32(__a, vget_high_s32(__b), __c); } |
| 8717 | __ai uint32x4_t vmlal_high_n_u16(uint32x4_t __a, uint16x8_t __b, uint16_t __c) { |
| 8718 | return vmlal_n_u16(__a, vget_high_u16(__b), __c); } |
| 8719 | __ai uint64x2_t vmlal_high_n_u32(uint64x2_t __a, uint32x4_t __b, uint32_t __c) { |
| 8720 | return vmlal_n_u32(__a, vget_high_u32(__b), __c); } |
| 8721 | |
| 8722 | #define vmlal_laneq_s16(a, b, c, __d) __extension__ ({ \ |
| 8723 | int32x4_t __a = (a); int16x4_t __b = (b); int16x8_t __c = (c); \ |
| 8724 | __a + vmull_s16(__b, __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 8725 | #define vmlal_laneq_s32(a, b, c, __d) __extension__ ({ \ |
| 8726 | int64x2_t __a = (a); int32x2_t __b = (b); int32x4_t __c = (c); \ |
| 8727 | __a + vmull_s32(__b, __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 8728 | #define vmlal_laneq_u16(a, b, c, __d) __extension__ ({ \ |
| 8729 | uint32x4_t __a = (a); uint16x4_t __b = (b); uint16x8_t __c = (c); \ |
| 8730 | __a + vmull_u16(__b, __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 8731 | #define vmlal_laneq_u32(a, b, c, __d) __extension__ ({ \ |
| 8732 | uint64x2_t __a = (a); uint32x2_t __b = (b); uint32x4_t __c = (c); \ |
| 8733 | __a + vmull_u32(__b, __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 8734 | |
| 8735 | #define vmla_laneq_s16(a, b, c, __d) __extension__ ({ \ |
| 8736 | int16x4_t __a = (a); int16x4_t __b = (b); int16x8_t __c = (c); \ |
| 8737 | __a + (__b * __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 8738 | #define vmla_laneq_s32(a, b, c, __d) __extension__ ({ \ |
| 8739 | int32x2_t __a = (a); int32x2_t __b = (b); int32x4_t __c = (c); \ |
| 8740 | __a + (__b * __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 8741 | #define vmla_laneq_u16(a, b, c, __d) __extension__ ({ \ |
| 8742 | uint16x4_t __a = (a); uint16x4_t __b = (b); uint16x8_t __c = (c); \ |
| 8743 | __a + (__b * __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 8744 | #define vmla_laneq_u32(a, b, c, __d) __extension__ ({ \ |
| 8745 | uint32x2_t __a = (a); uint32x2_t __b = (b); uint32x4_t __c = (c); \ |
| 8746 | __a + (__b * __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 8747 | #define vmla_laneq_f32(a, b, c, __d) __extension__ ({ \ |
| 8748 | float32x2_t __a = (a); float32x2_t __b = (b); float32x4_t __c = (c); \ |
| 8749 | __a + (__b * __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 8750 | #define vmlaq_laneq_s16(a, b, c, __d) __extension__ ({ \ |
| 8751 | int16x8_t __a = (a); int16x8_t __b = (b); int16x8_t __c = (c); \ |
| 8752 | __a + (__b * __builtin_shufflevector(__c, __c, __d, __d, __d, __d, __d, __d, __d, __d)); }) |
| 8753 | #define vmlaq_laneq_s32(a, b, c, __d) __extension__ ({ \ |
| 8754 | int32x4_t __a = (a); int32x4_t __b = (b); int32x4_t __c = (c); \ |
| 8755 | __a + (__b * __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 8756 | #define vmlaq_laneq_u16(a, b, c, __d) __extension__ ({ \ |
| 8757 | uint16x8_t __a = (a); uint16x8_t __b = (b); uint16x8_t __c = (c); \ |
| 8758 | __a + (__b * __builtin_shufflevector(__c, __c, __d, __d, __d, __d, __d, __d, __d, __d)); }) |
| 8759 | #define vmlaq_laneq_u32(a, b, c, __d) __extension__ ({ \ |
| 8760 | uint32x4_t __a = (a); uint32x4_t __b = (b); uint32x4_t __c = (c); \ |
| 8761 | __a + (__b * __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 8762 | #define vmlaq_laneq_f32(a, b, c, __d) __extension__ ({ \ |
| 8763 | float32x4_t __a = (a); float32x4_t __b = (b); float32x4_t __c = (c); \ |
| 8764 | __a + (__b * __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 8765 | |
| 8766 | __ai int16x8_t vmlsl_high_s8(int16x8_t __a, int8x16_t __b, int8x16_t __c) { |
| 8767 | return vmlsl_s8(__a, vget_high_s8(__b), vget_high_s8(__c)); } |
| 8768 | __ai int32x4_t vmlsl_high_s16(int32x4_t __a, int16x8_t __b, int16x8_t __c) { |
| 8769 | return vmlsl_s16(__a, vget_high_s16(__b), vget_high_s16(__c)); } |
| 8770 | __ai int64x2_t vmlsl_high_s32(int64x2_t __a, int32x4_t __b, int32x4_t __c) { |
| 8771 | return vmlsl_s32(__a, vget_high_s32(__b), vget_high_s32(__c)); } |
| 8772 | __ai uint16x8_t vmlsl_high_u8(uint16x8_t __a, uint8x16_t __b, uint8x16_t __c) { |
| 8773 | return vmlsl_u8(__a, vget_high_u8(__b), vget_high_u8(__c)); } |
| 8774 | __ai uint32x4_t vmlsl_high_u16(uint32x4_t __a, uint16x8_t __b, uint16x8_t __c) { |
| 8775 | return vmlsl_u16(__a, vget_high_u16(__b), vget_high_u16(__c)); } |
| 8776 | __ai uint64x2_t vmlsl_high_u32(uint64x2_t __a, uint32x4_t __b, uint32x4_t __c) { |
| 8777 | return vmlsl_u32(__a, vget_high_u32(__b), vget_high_u32(__c)); } |
| 8778 | |
| 8779 | #define vmlsl_high_lane_s16(a, b, c, __d) __extension__ ({ \ |
| 8780 | int32x4_t __a = (a); int16x8_t __b = (b); int16x4_t __c = (c); \ |
| 8781 | __a - vmull_s16(vget_high_s16(__b), __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 8782 | #define vmlsl_high_lane_s32(a, b, c, __d) __extension__ ({ \ |
| 8783 | int64x2_t __a = (a); int32x4_t __b = (b); int32x2_t __c = (c); \ |
| 8784 | __a - vmull_s32(vget_high_s32(__b), __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 8785 | #define vmlsl_high_lane_u16(a, b, c, __d) __extension__ ({ \ |
| 8786 | uint32x4_t __a = (a); uint16x8_t __b = (b); uint16x4_t __c = (c); \ |
| 8787 | __a - vmull_u16(vget_high_u16(__b), __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 8788 | #define vmlsl_high_lane_u32(a, b, c, __d) __extension__ ({ \ |
| 8789 | uint64x2_t __a = (a); uint32x4_t __b = (b); uint32x2_t __c = (c); \ |
| 8790 | __a - vmull_u32(vget_high_u32(__b), __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 8791 | |
| 8792 | #define vmlsl_high_laneq_s16(a, b, c, __d) __extension__ ({ \ |
| 8793 | int32x4_t __a = (a); int16x8_t __b = (b); int16x8_t __c = (c); \ |
| 8794 | __a - vmull_s16(vget_high_s16(__b), __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 8795 | #define vmlsl_high_laneq_s32(a, b, c, __d) __extension__ ({ \ |
| 8796 | int64x2_t __a = (a); int32x4_t __b = (b); int32x4_t __c = (c); \ |
| 8797 | __a - vmull_s32(vget_high_s32(__b), __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 8798 | #define vmlsl_high_laneq_u16(a, b, c, __d) __extension__ ({ \ |
| 8799 | uint32x4_t __a = (a); uint16x8_t __b = (b); uint16x8_t __c = (c); \ |
| 8800 | __a - vmull_u16(vget_high_u16(__b), __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 8801 | #define vmlsl_high_laneq_u32(a, b, c, __d) __extension__ ({ \ |
| 8802 | uint64x2_t __a = (a); uint32x4_t __b = (b); uint32x4_t __c = (c); \ |
| 8803 | __a - vmull_u32(vget_high_u32(__b), __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 8804 | |
| 8805 | __ai int32x4_t vmlsl_high_n_s16(int32x4_t __a, int16x8_t __b, int16_t __c) { |
| 8806 | return vmlsl_n_s16(__a, vget_high_s16(__b), __c); } |
| 8807 | __ai int64x2_t vmlsl_high_n_s32(int64x2_t __a, int32x4_t __b, int32_t __c) { |
| 8808 | return vmlsl_n_s32(__a, vget_high_s32(__b), __c); } |
| 8809 | __ai uint32x4_t vmlsl_high_n_u16(uint32x4_t __a, uint16x8_t __b, uint16_t __c) { |
| 8810 | return vmlsl_n_u16(__a, vget_high_u16(__b), __c); } |
| 8811 | __ai uint64x2_t vmlsl_high_n_u32(uint64x2_t __a, uint32x4_t __b, uint32_t __c) { |
| 8812 | return vmlsl_n_u32(__a, vget_high_u32(__b), __c); } |
| 8813 | |
| 8814 | #define vmlsl_laneq_s16(a, b, c, __d) __extension__ ({ \ |
| 8815 | int32x4_t __a = (a); int16x4_t __b = (b); int16x8_t __c = (c); \ |
| 8816 | __a - vmull_s16(__b, __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 8817 | #define vmlsl_laneq_s32(a, b, c, __d) __extension__ ({ \ |
| 8818 | int64x2_t __a = (a); int32x2_t __b = (b); int32x4_t __c = (c); \ |
| 8819 | __a - vmull_s32(__b, __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 8820 | #define vmlsl_laneq_u16(a, b, c, __d) __extension__ ({ \ |
| 8821 | uint32x4_t __a = (a); uint16x4_t __b = (b); uint16x8_t __c = (c); \ |
| 8822 | __a - vmull_u16(__b, __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 8823 | #define vmlsl_laneq_u32(a, b, c, __d) __extension__ ({ \ |
| 8824 | uint64x2_t __a = (a); uint32x2_t __b = (b); uint32x4_t __c = (c); \ |
| 8825 | __a - vmull_u32(__b, __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 8826 | |
| 8827 | #define vmls_laneq_s16(a, b, c, __d) __extension__ ({ \ |
| 8828 | int16x4_t __a = (a); int16x4_t __b = (b); int16x8_t __c = (c); \ |
| 8829 | __a - (__b * __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 8830 | #define vmls_laneq_s32(a, b, c, __d) __extension__ ({ \ |
| 8831 | int32x2_t __a = (a); int32x2_t __b = (b); int32x4_t __c = (c); \ |
| 8832 | __a - (__b * __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 8833 | #define vmls_laneq_u16(a, b, c, __d) __extension__ ({ \ |
| 8834 | uint16x4_t __a = (a); uint16x4_t __b = (b); uint16x8_t __c = (c); \ |
| 8835 | __a - (__b * __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 8836 | #define vmls_laneq_u32(a, b, c, __d) __extension__ ({ \ |
| 8837 | uint32x2_t __a = (a); uint32x2_t __b = (b); uint32x4_t __c = (c); \ |
| 8838 | __a - (__b * __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 8839 | #define vmls_laneq_f32(a, b, c, __d) __extension__ ({ \ |
| 8840 | float32x2_t __a = (a); float32x2_t __b = (b); float32x4_t __c = (c); \ |
| 8841 | __a - (__b * __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 8842 | #define vmlsq_laneq_s16(a, b, c, __d) __extension__ ({ \ |
| 8843 | int16x8_t __a = (a); int16x8_t __b = (b); int16x8_t __c = (c); \ |
| 8844 | __a - (__b * __builtin_shufflevector(__c, __c, __d, __d, __d, __d, __d, __d, __d, __d)); }) |
| 8845 | #define vmlsq_laneq_s32(a, b, c, __d) __extension__ ({ \ |
| 8846 | int32x4_t __a = (a); int32x4_t __b = (b); int32x4_t __c = (c); \ |
| 8847 | __a - (__b * __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 8848 | #define vmlsq_laneq_u16(a, b, c, __d) __extension__ ({ \ |
| 8849 | uint16x8_t __a = (a); uint16x8_t __b = (b); uint16x8_t __c = (c); \ |
| 8850 | __a - (__b * __builtin_shufflevector(__c, __c, __d, __d, __d, __d, __d, __d, __d, __d)); }) |
| 8851 | #define vmlsq_laneq_u32(a, b, c, __d) __extension__ ({ \ |
| 8852 | uint32x4_t __a = (a); uint32x4_t __b = (b); uint32x4_t __c = (c); \ |
| 8853 | __a - (__b * __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 8854 | #define vmlsq_laneq_f32(a, b, c, __d) __extension__ ({ \ |
| 8855 | float32x4_t __a = (a); float32x4_t __b = (b); float32x4_t __c = (c); \ |
| 8856 | __a - (__b * __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 8857 | |
| 8858 | |
| 8859 | |
| 8860 | #define vmull_high_lane_s16(a, b, __c) __extension__ ({ \ |
| 8861 | int16x8_t __a = (a); int16x4_t __b = (b); \ |
| 8862 | vmull_s16(vget_high_s16(__a), __builtin_shufflevector(__b, __b, __c, __c, __c, __c)); }) |
| 8863 | #define vmull_high_lane_s32(a, b, __c) __extension__ ({ \ |
| 8864 | int32x4_t __a = (a); int32x2_t __b = (b); \ |
| 8865 | vmull_s32(vget_high_s32(__a), __builtin_shufflevector(__b, __b, __c, __c)); }) |
| 8866 | #define vmull_high_lane_u16(a, b, __c) __extension__ ({ \ |
| 8867 | uint16x8_t __a = (a); uint16x4_t __b = (b); \ |
| 8868 | vmull_u16(vget_high_u16(__a), __builtin_shufflevector(__b, __b, __c, __c, __c, __c)); }) |
| 8869 | #define vmull_high_lane_u32(a, b, __c) __extension__ ({ \ |
| 8870 | uint32x4_t __a = (a); uint32x2_t __b = (b); \ |
| 8871 | vmull_u32(vget_high_u32(__a), __builtin_shufflevector(__b, __b, __c, __c)); }) |
| 8872 | |
| 8873 | #define vmull_high_laneq_s16(a, b, __c) __extension__ ({ \ |
| 8874 | int16x8_t __a = (a); int16x8_t __b = (b); \ |
| 8875 | vmull_s16(vget_high_s16(__a), __builtin_shufflevector(__b, __b, __c, __c, __c, __c)); }) |
| 8876 | #define vmull_high_laneq_s32(a, b, __c) __extension__ ({ \ |
| 8877 | int32x4_t __a = (a); int32x4_t __b = (b); \ |
| 8878 | vmull_s32(vget_high_s32(__a), __builtin_shufflevector(__b, __b, __c, __c)); }) |
| 8879 | #define vmull_high_laneq_u16(a, b, __c) __extension__ ({ \ |
| 8880 | uint16x8_t __a = (a); uint16x8_t __b = (b); \ |
| 8881 | vmull_u16(vget_high_u16(__a), __builtin_shufflevector(__b, __b, __c, __c, __c, __c)); }) |
| 8882 | #define vmull_high_laneq_u32(a, b, __c) __extension__ ({ \ |
| 8883 | uint32x4_t __a = (a); uint32x4_t __b = (b); \ |
| 8884 | vmull_u32(vget_high_u32(__a), __builtin_shufflevector(__b, __b, __c, __c)); }) |
| 8885 | |
| 8886 | __ai int32x4_t vmull_high_n_s16(int16x8_t __a, int16_t __b) { |
| 8887 | return vmull_n_s16(vget_high_s16(__a), __b); } |
| 8888 | __ai int64x2_t vmull_high_n_s32(int32x4_t __a, int32_t __b) { |
| 8889 | return vmull_n_s32(vget_high_s32(__a), __b); } |
| 8890 | __ai uint32x4_t vmull_high_n_u16(uint16x8_t __a, uint16_t __b) { |
| 8891 | return vmull_n_u16(vget_high_u16(__a), __b); } |
| 8892 | __ai uint64x2_t vmull_high_n_u32(uint32x4_t __a, uint32_t __b) { |
| 8893 | return vmull_n_u32(vget_high_u32(__a), __b); } |
| 8894 | |
| 8895 | #define vmull_laneq_s16(a, b, __c) __extension__ ({ \ |
| 8896 | int16x4_t __a = (a); int16x8_t __b = (b); \ |
| 8897 | vmull_s16(__a, __builtin_shufflevector(__b, __b, __c, __c, __c, __c)); }) |
| 8898 | #define vmull_laneq_s32(a, b, __c) __extension__ ({ \ |
| 8899 | int32x2_t __a = (a); int32x4_t __b = (b); \ |
| 8900 | vmull_s32(__a, __builtin_shufflevector(__b, __b, __c, __c)); }) |
| 8901 | #define vmull_laneq_u16(a, b, __c) __extension__ ({ \ |
| 8902 | uint16x4_t __a = (a); uint16x8_t __b = (b); \ |
| 8903 | vmull_u16(__a, __builtin_shufflevector(__b, __b, __c, __c, __c, __c)); }) |
| 8904 | #define vmull_laneq_u32(a, b, __c) __extension__ ({ \ |
| 8905 | uint32x2_t __a = (a); uint32x4_t __b = (b); \ |
| 8906 | vmull_u32(__a, __builtin_shufflevector(__b, __b, __c, __c)); }) |
| 8907 | |
| 8908 | #define vmulx_lane_f32(a, b, __c) __extension__ ({ \ |
| 8909 | float32x2_t __a = (a); float32x2_t __b = (b); \ |
| 8910 | vmulx_f32(__a, __builtin_shufflevector(__b, __b, __c, __c)); }) |
| 8911 | #define vmulxq_lane_f32(a, b, __c) __extension__ ({ \ |
| 8912 | float32x4_t __a = (a); float32x2_t __b = (b); \ |
| 8913 | vmulxq_f32(__a, __builtin_shufflevector(__b, __b, __c, __c, __c, __c)); }) |
| 8914 | #define vmulxq_lane_f64(a, b, __c) __extension__ ({ \ |
| 8915 | float64x2_t __a = (a); float64x1_t __b = (b); \ |
| 8916 | vmulxq_f64(__a, __builtin_shufflevector(__b, __b, __c, __c)); }) |
| 8917 | |
| 8918 | #define vmulx_laneq_f32(a, b, __c) __extension__ ({ \ |
| 8919 | float32x2_t __a = (a); float32x4_t __b = (b); \ |
| 8920 | vmulx_f32(__a, __builtin_shufflevector(__b, __b, __c, __c)); }) |
| 8921 | #define vmulxq_laneq_f32(a, b, __c) __extension__ ({ \ |
| 8922 | float32x4_t __a = (a); float32x4_t __b = (b); \ |
| 8923 | vmulxq_f32(__a, __builtin_shufflevector(__b, __b, __c, __c, __c, __c)); }) |
| 8924 | #define vmulxq_laneq_f64(a, b, __c) __extension__ ({ \ |
| 8925 | float64x2_t __a = (a); float64x2_t __b = (b); \ |
| 8926 | vmulxq_f64(__a, __builtin_shufflevector(__b, __b, __c, __c)); }) |
| 8927 | |
| 8928 | #define vmul_laneq_s16(a, b, __c) __extension__ ({ \ |
| 8929 | int16x4_t __a = (a); int16x8_t __b = (b); \ |
| 8930 | __a * __builtin_shufflevector(__b, __b, __c, __c, __c, __c); }) |
| 8931 | #define vmul_laneq_s32(a, b, __c) __extension__ ({ \ |
| 8932 | int32x2_t __a = (a); int32x4_t __b = (b); \ |
| 8933 | __a * __builtin_shufflevector(__b, __b, __c, __c); }) |
| 8934 | #define vmul_laneq_f32(a, b, __c) __extension__ ({ \ |
| 8935 | float32x2_t __a = (a); float32x4_t __b = (b); \ |
| 8936 | __a * __builtin_shufflevector(__b, __b, __c, __c); }) |
| 8937 | #define vmul_laneq_u16(a, b, __c) __extension__ ({ \ |
| 8938 | uint16x4_t __a = (a); uint16x8_t __b = (b); \ |
| 8939 | __a * __builtin_shufflevector(__b, __b, __c, __c, __c, __c); }) |
| 8940 | #define vmul_laneq_u32(a, b, __c) __extension__ ({ \ |
| 8941 | uint32x2_t __a = (a); uint32x4_t __b = (b); \ |
| 8942 | __a * __builtin_shufflevector(__b, __b, __c, __c); }) |
| 8943 | #define vmulq_laneq_s16(a, b, __c) __extension__ ({ \ |
| 8944 | int16x8_t __a = (a); int16x8_t __b = (b); \ |
| 8945 | __a * __builtin_shufflevector(__b, __b, __c, __c, __c, __c, __c, __c, __c, __c); }) |
| 8946 | #define vmulq_laneq_s32(a, b, __c) __extension__ ({ \ |
| 8947 | int32x4_t __a = (a); int32x4_t __b = (b); \ |
| 8948 | __a * __builtin_shufflevector(__b, __b, __c, __c, __c, __c); }) |
| 8949 | #define vmulq_laneq_f32(a, b, __c) __extension__ ({ \ |
| 8950 | float32x4_t __a = (a); float32x4_t __b = (b); \ |
| 8951 | __a * __builtin_shufflevector(__b, __b, __c, __c, __c, __c); }) |
| 8952 | #define vmulq_laneq_u16(a, b, __c) __extension__ ({ \ |
| 8953 | uint16x8_t __a = (a); uint16x8_t __b = (b); \ |
| 8954 | __a * __builtin_shufflevector(__b, __b, __c, __c, __c, __c, __c, __c, __c, __c); }) |
| 8955 | #define vmulq_laneq_u32(a, b, __c) __extension__ ({ \ |
| 8956 | uint32x4_t __a = (a); uint32x4_t __b = (b); \ |
| 8957 | __a * __builtin_shufflevector(__b, __b, __c, __c, __c, __c); }) |
| 8958 | #define vmulq_laneq_f64(a, b, __c) __extension__ ({ \ |
| 8959 | float64x2_t __a = (a); float64x2_t __b = (b); \ |
| 8960 | __a * __builtin_shufflevector(__b, __b, __c, __c); }) |
| 8961 | |
| 8962 | #define vmulq_lane_f64(a, b, __c) __extension__ ({ \ |
| 8963 | float64x2_t __a = (a); float64x1_t __b = (b); \ |
| 8964 | __a * __builtin_shufflevector(__b, __b, __c, __c); }) |
| 8965 | |
| 8966 | __ai float64x2_t vmulq_n_f64(float64x2_t __a, float64_t __b) { |
| 8967 | return __a * (float64x2_t){ __b, __b }; } |
| 8968 | |
| 8969 | __ai int32x4_t vqdmlal_high_s16(int32x4_t __a, int16x8_t __b, int16x8_t __c) { |
| 8970 | return vqdmlal_s16(__a, vget_high_s16(__b), vget_high_s16(__c)); } |
| 8971 | __ai int64x2_t vqdmlal_high_s32(int64x2_t __a, int32x4_t __b, int32x4_t __c) { |
| 8972 | return vqdmlal_s32(__a, vget_high_s32(__b), vget_high_s32(__c)); } |
| 8973 | |
| 8974 | #define vqdmlal_high_lane_s16(a, b, c, __d) __extension__ ({ \ |
| 8975 | int32x4_t __a = (a); int16x8_t __b = (b); int16x4_t __c = (c); \ |
| 8976 | vqdmlal_s16(__a, vget_high_s16(__b), __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 8977 | #define vqdmlal_high_lane_s32(a, b, c, __d) __extension__ ({ \ |
| 8978 | int64x2_t __a = (a); int32x4_t __b = (b); int32x2_t __c = (c); \ |
| 8979 | vqdmlal_s32(__a, vget_high_s32(__b), __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 8980 | |
| 8981 | #define vqdmlal_high_laneq_s16(a, b, c, __d) __extension__ ({ \ |
| 8982 | int32x4_t __a = (a); int16x8_t __b = (b); int16x8_t __c = (c); \ |
| 8983 | vqdmlal_s16(__a, vget_high_s16(__b), __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 8984 | #define vqdmlal_high_laneq_s32(a, b, c, __d) __extension__ ({ \ |
| 8985 | int64x2_t __a = (a); int32x4_t __b = (b); int32x4_t __c = (c); \ |
| 8986 | vqdmlal_s32(__a, vget_high_s32(__b), __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 8987 | |
| 8988 | __ai int32x4_t vqdmlal_high_n_s16(int32x4_t __a, int16x8_t __b, int16_t __c) { |
| 8989 | return vqdmlal_n_s16(__a, vget_high_s16(__b), __c); } |
| 8990 | __ai int64x2_t vqdmlal_high_n_s32(int64x2_t __a, int32x4_t __b, int32_t __c) { |
| 8991 | return vqdmlal_n_s32(__a, vget_high_s32(__b), __c); } |
| 8992 | |
| 8993 | #define vqdmlal_laneq_s16(a, b, c, __d) __extension__ ({ \ |
| 8994 | int32x4_t __a = (a); int16x4_t __b = (b); int16x8_t __c = (c); \ |
| 8995 | vqdmlal_s16(__a, __b, __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 8996 | #define vqdmlal_laneq_s32(a, b, c, __d) __extension__ ({ \ |
| 8997 | int64x2_t __a = (a); int32x2_t __b = (b); int32x4_t __c = (c); \ |
| 8998 | vqdmlal_s32(__a, __b, __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 8999 | |
| 9000 | __ai int32x4_t vqdmlsl_high_s16(int32x4_t __a, int16x8_t __b, int16x8_t __c) { |
| 9001 | return vqdmlsl_s16(__a, vget_high_s16(__b), vget_high_s16(__c)); } |
| 9002 | __ai int64x2_t vqdmlsl_high_s32(int64x2_t __a, int32x4_t __b, int32x4_t __c) { |
| 9003 | return vqdmlsl_s32(__a, vget_high_s32(__b), vget_high_s32(__c)); } |
| 9004 | |
| 9005 | #define vqdmlsl_high_lane_s16(a, b, c, __d) __extension__ ({ \ |
| 9006 | int32x4_t __a = (a); int16x8_t __b = (b); int16x4_t __c = (c); \ |
| 9007 | vqdmlsl_s16(__a, vget_high_s16(__b), __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 9008 | #define vqdmlsl_high_lane_s32(a, b, c, __d) __extension__ ({ \ |
| 9009 | int64x2_t __a = (a); int32x4_t __b = (b); int32x2_t __c = (c); \ |
| 9010 | vqdmlsl_s32(__a, vget_high_s32(__b), __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 9011 | |
| 9012 | #define vqdmlsl_high_laneq_s16(a, b, c, __d) __extension__ ({ \ |
| 9013 | int32x4_t __a = (a); int16x8_t __b = (b); int16x8_t __c = (c); \ |
| 9014 | vqdmlsl_s16(__a, vget_high_s16(__b), __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 9015 | #define vqdmlsl_high_laneq_s32(a, b, c, __d) __extension__ ({ \ |
| 9016 | int64x2_t __a = (a); int32x4_t __b = (b); int32x4_t __c = (c); \ |
| 9017 | vqdmlsl_s32(__a, vget_high_s32(__b), __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 9018 | |
| 9019 | __ai int32x4_t vqdmlsl_high_n_s16(int32x4_t __a, int16x8_t __b, int16_t __c) { |
| 9020 | return vqdmlsl_n_s16(__a, vget_high_s16(__b), __c); } |
| 9021 | __ai int64x2_t vqdmlsl_high_n_s32(int64x2_t __a, int32x4_t __b, int32_t __c) { |
| 9022 | return vqdmlsl_n_s32(__a, vget_high_s32(__b), __c); } |
| 9023 | |
| 9024 | #define vqdmlsl_laneq_s16(a, b, c, __d) __extension__ ({ \ |
| 9025 | int32x4_t __a = (a); int16x4_t __b = (b); int16x8_t __c = (c); \ |
| 9026 | vqdmlsl_s16(__a, __b, __builtin_shufflevector(__c, __c, __d, __d, __d, __d)); }) |
| 9027 | #define vqdmlsl_laneq_s32(a, b, c, __d) __extension__ ({ \ |
| 9028 | int64x2_t __a = (a); int32x2_t __b = (b); int32x4_t __c = (c); \ |
| 9029 | vqdmlsl_s32(__a, __b, __builtin_shufflevector(__c, __c, __d, __d)); }) |
| 9030 | |
| 9031 | #define vqdmulh_laneq_s16(a, b, __c) __extension__ ({ \ |
| 9032 | int16x4_t __a = (a); int16x8_t __b = (b); \ |
| 9033 | vqdmulh_s16(__a, __builtin_shufflevector(__b, __b, __c, __c, __c, __c)); }) |
| 9034 | #define vqdmulh_laneq_s32(a, b, __c) __extension__ ({ \ |
| 9035 | int32x2_t __a = (a); int32x4_t __b = (b); \ |
| 9036 | vqdmulh_s32(__a, __builtin_shufflevector(__b, __b, __c, __c)); }) |
| 9037 | #define vqdmulhq_laneq_s16(a, b, __c) __extension__ ({ \ |
| 9038 | int16x8_t __a = (a); int16x8_t __b = (b); \ |
| 9039 | vqdmulhq_s16(__a, __builtin_shufflevector(__b, __b, __c, __c, __c, __c, __c, __c, __c, __c)); }) |
| 9040 | #define vqdmulhq_laneq_s32(a, b, __c) __extension__ ({ \ |
| 9041 | int32x4_t __a = (a); int32x4_t __b = (b); \ |
| 9042 | vqdmulhq_s32(__a, __builtin_shufflevector(__b, __b, __c, __c, __c, __c)); }) |
| 9043 | |
| 9044 | __ai int32x4_t vqdmull_high_s16(int16x8_t __a, int16x8_t __b) { |
| 9045 | return vqdmull_s16(vget_high_s16(__a), vget_high_s16(__b)); } |
| 9046 | __ai int64x2_t vqdmull_high_s32(int32x4_t __a, int32x4_t __b) { |
| 9047 | return vqdmull_s32(vget_high_s32(__a), vget_high_s32(__b)); } |
| 9048 | |
| 9049 | #define vqdmull_high_lane_s16(a, b, __c) __extension__ ({ \ |
| 9050 | int16x8_t __a = (a); int16x4_t __b = (b); \ |
| 9051 | vqdmull_s16(vget_high_s16(__a), __builtin_shufflevector(__b, __b, __c, __c, __c, __c)); }) |
| 9052 | #define vqdmull_high_lane_s32(a, b, __c) __extension__ ({ \ |
| 9053 | int32x4_t __a = (a); int32x2_t __b = (b); \ |
| 9054 | vqdmull_s32(vget_high_s32(__a), __builtin_shufflevector(__b, __b, __c, __c)); }) |
| 9055 | |
| 9056 | #define vqdmull_high_laneq_s16(a, b, __c) __extension__ ({ \ |
| 9057 | int16x8_t __a = (a); int16x8_t __b = (b); \ |
| 9058 | vqdmull_s16(vget_high_s16(__a), __builtin_shufflevector(__b, __b, __c, __c, __c, __c)); }) |
| 9059 | #define vqdmull_high_laneq_s32(a, b, __c) __extension__ ({ \ |
| 9060 | int32x4_t __a = (a); int32x4_t __b = (b); \ |
| 9061 | vqdmull_s32(vget_high_s32(__a), __builtin_shufflevector(__b, __b, __c, __c)); }) |
| 9062 | |
| 9063 | __ai int32x4_t vqdmull_high_n_s16(int16x8_t __a, int16_t __b) { |
| 9064 | return vqdmull_n_s16(vget_high_s16(__a), __b); } |
| 9065 | __ai int64x2_t vqdmull_high_n_s32(int32x4_t __a, int32_t __b) { |
| 9066 | return vqdmull_n_s32(vget_high_s32(__a), __b); } |
| 9067 | |
| 9068 | #define vqdmull_laneq_s16(a, b, __c) __extension__ ({ \ |
| 9069 | int16x4_t __a = (a); int16x8_t __b = (b); \ |
| 9070 | vqdmull_s16(__a, __builtin_shufflevector(__b, __b, __c, __c, __c, __c)); }) |
| 9071 | #define vqdmull_laneq_s32(a, b, __c) __extension__ ({ \ |
| 9072 | int32x2_t __a = (a); int32x4_t __b = (b); \ |
| 9073 | vqdmull_s32(__a, __builtin_shufflevector(__b, __b, __c, __c)); }) |
| 9074 | |
| 9075 | #define vqrdmulh_laneq_s16(a, b, __c) __extension__ ({ \ |
| 9076 | int16x4_t __a = (a); int16x8_t __b = (b); \ |
| 9077 | vqrdmulh_s16(__a, __builtin_shufflevector(__b, __b, __c, __c, __c, __c)); }) |
| 9078 | #define vqrdmulh_laneq_s32(a, b, __c) __extension__ ({ \ |
| 9079 | int32x2_t __a = (a); int32x4_t __b = (b); \ |
| 9080 | vqrdmulh_s32(__a, __builtin_shufflevector(__b, __b, __c, __c)); }) |
| 9081 | #define vqrdmulhq_laneq_s16(a, b, __c) __extension__ ({ \ |
| 9082 | int16x8_t __a = (a); int16x8_t __b = (b); \ |
| 9083 | vqrdmulhq_s16(__a, __builtin_shufflevector(__b, __b, __c, __c, __c, __c, __c, __c, __c, __c)); }) |
| 9084 | #define vqrdmulhq_laneq_s32(a, b, __c) __extension__ ({ \ |
| 9085 | int32x4_t __a = (a); int32x4_t __b = (b); \ |
| 9086 | vqrdmulhq_s32(__a, __builtin_shufflevector(__b, __b, __c, __c, __c, __c)); }) |
| 9087 | |
| 9088 | __ai uint8x8_t vqtbl1_u8(uint8x16_t __a, uint8x8_t __b) { |
| 9089 | return (uint8x8_t)__builtin_neon_vqtbl1_v((int8x16_t)__a, (int8x8_t)__b, 16); } |
| 9090 | __ai int8x8_t vqtbl1_s8(int8x16_t __a, int8x8_t __b) { |
| 9091 | return (int8x8_t)__builtin_neon_vqtbl1_v(__a, __b, 0); } |
| 9092 | __ai poly8x8_t vqtbl1_p8(poly8x16_t __a, uint8x8_t __b) { |
| 9093 | return (poly8x8_t)__builtin_neon_vqtbl1_v((int8x16_t)__a, (int8x8_t)__b, 4); } |
| 9094 | __ai uint8x16_t vqtbl1q_u8(uint8x16_t __a, uint8x16_t __b) { |
| 9095 | return (uint8x16_t)__builtin_neon_vqtbl1q_v((int8x16_t)__a, (int8x16_t)__b, 48); } |
| 9096 | __ai int8x16_t vqtbl1q_s8(int8x16_t __a, int8x16_t __b) { |
| 9097 | return (int8x16_t)__builtin_neon_vqtbl1q_v(__a, __b, 32); } |
| 9098 | __ai poly8x16_t vqtbl1q_p8(poly8x16_t __a, uint8x16_t __b) { |
| 9099 | return (poly8x16_t)__builtin_neon_vqtbl1q_v((int8x16_t)__a, (int8x16_t)__b, 36); } |
| 9100 | |
| 9101 | __ai uint8x8_t vqtbl2_u8(uint8x16x2_t __a, uint8x8_t __b) { |
| 9102 | return (uint8x8_t)__builtin_neon_vqtbl2_v((int8x16_t)__a.val[0], (int8x16_t)__a.val[1], (int8x8_t)__b, 16); } |
| 9103 | __ai int8x8_t vqtbl2_s8(int8x16x2_t __a, int8x8_t __b) { |
| 9104 | return (int8x8_t)__builtin_neon_vqtbl2_v(__a.val[0], __a.val[1], __b, 0); } |
| 9105 | __ai poly8x8_t vqtbl2_p8(poly8x16x2_t __a, uint8x8_t __b) { |
| 9106 | return (poly8x8_t)__builtin_neon_vqtbl2_v((int8x16_t)__a.val[0], (int8x16_t)__a.val[1], (int8x8_t)__b, 4); } |
| 9107 | __ai uint8x16_t vqtbl2q_u8(uint8x16x2_t __a, uint8x16_t __b) { |
| 9108 | return (uint8x16_t)__builtin_neon_vqtbl2q_v((int8x16_t)__a.val[0], (int8x16_t)__a.val[1], (int8x16_t)__b, 48); } |
| 9109 | __ai int8x16_t vqtbl2q_s8(int8x16x2_t __a, int8x16_t __b) { |
| 9110 | return (int8x16_t)__builtin_neon_vqtbl2q_v(__a.val[0], __a.val[1], __b, 32); } |
| 9111 | __ai poly8x16_t vqtbl2q_p8(poly8x16x2_t __a, uint8x16_t __b) { |
| 9112 | return (poly8x16_t)__builtin_neon_vqtbl2q_v((int8x16_t)__a.val[0], (int8x16_t)__a.val[1], (int8x16_t)__b, 36); } |
| 9113 | |
| 9114 | __ai uint8x8_t vqtbl3_u8(uint8x16x3_t __a, uint8x8_t __b) { |
| 9115 | return (uint8x8_t)__builtin_neon_vqtbl3_v((int8x16_t)__a.val[0], (int8x16_t)__a.val[1], (int8x16_t)__a.val[2], (int8x8_t)__b, 16); } |
| 9116 | __ai int8x8_t vqtbl3_s8(int8x16x3_t __a, int8x8_t __b) { |
| 9117 | return (int8x8_t)__builtin_neon_vqtbl3_v(__a.val[0], __a.val[1], __a.val[2], __b, 0); } |
| 9118 | __ai poly8x8_t vqtbl3_p8(poly8x16x3_t __a, uint8x8_t __b) { |
| 9119 | return (poly8x8_t)__builtin_neon_vqtbl3_v((int8x16_t)__a.val[0], (int8x16_t)__a.val[1], (int8x16_t)__a.val[2], (int8x8_t)__b, 4); } |
| 9120 | __ai uint8x16_t vqtbl3q_u8(uint8x16x3_t __a, uint8x16_t __b) { |
| 9121 | return (uint8x16_t)__builtin_neon_vqtbl3q_v((int8x16_t)__a.val[0], (int8x16_t)__a.val[1], (int8x16_t)__a.val[2], (int8x16_t)__b, 48); } |
| 9122 | __ai int8x16_t vqtbl3q_s8(int8x16x3_t __a, int8x16_t __b) { |
| 9123 | return (int8x16_t)__builtin_neon_vqtbl3q_v(__a.val[0], __a.val[1], __a.val[2], __b, 32); } |
| 9124 | __ai poly8x16_t vqtbl3q_p8(poly8x16x3_t __a, uint8x16_t __b) { |
| 9125 | return (poly8x16_t)__builtin_neon_vqtbl3q_v((int8x16_t)__a.val[0], (int8x16_t)__a.val[1], (int8x16_t)__a.val[2], (int8x16_t)__b, 36); } |
| 9126 | |
| 9127 | __ai uint8x8_t vqtbl4_u8(uint8x16x4_t __a, uint8x8_t __b) { |
| 9128 | return (uint8x8_t)__builtin_neon_vqtbl4_v((int8x16_t)__a.val[0], (int8x16_t)__a.val[1], (int8x16_t)__a.val[2], (int8x16_t)__a.val[3], (int8x8_t)__b, 16); } |
| 9129 | __ai int8x8_t vqtbl4_s8(int8x16x4_t __a, int8x8_t __b) { |
| 9130 | return (int8x8_t)__builtin_neon_vqtbl4_v(__a.val[0], __a.val[1], __a.val[2], __a.val[3], __b, 0); } |
| 9131 | __ai poly8x8_t vqtbl4_p8(poly8x16x4_t __a, uint8x8_t __b) { |
| 9132 | return (poly8x8_t)__builtin_neon_vqtbl4_v((int8x16_t)__a.val[0], (int8x16_t)__a.val[1], (int8x16_t)__a.val[2], (int8x16_t)__a.val[3], (int8x8_t)__b, 4); } |
| 9133 | __ai uint8x16_t vqtbl4q_u8(uint8x16x4_t __a, uint8x16_t __b) { |
| 9134 | return (uint8x16_t)__builtin_neon_vqtbl4q_v((int8x16_t)__a.val[0], (int8x16_t)__a.val[1], (int8x16_t)__a.val[2], (int8x16_t)__a.val[3], (int8x16_t)__b, 48); } |
| 9135 | __ai int8x16_t vqtbl4q_s8(int8x16x4_t __a, int8x16_t __b) { |
| 9136 | return (int8x16_t)__builtin_neon_vqtbl4q_v(__a.val[0], __a.val[1], __a.val[2], __a.val[3], __b, 32); } |
| 9137 | __ai poly8x16_t vqtbl4q_p8(poly8x16x4_t __a, uint8x16_t __b) { |
| 9138 | return (poly8x16_t)__builtin_neon_vqtbl4q_v((int8x16_t)__a.val[0], (int8x16_t)__a.val[1], (int8x16_t)__a.val[2], (int8x16_t)__a.val[3], (int8x16_t)__b, 36); } |
| 9139 | |
| 9140 | __ai uint8x8_t vqtbx1_u8(uint8x8_t __a, uint8x16_t __b, uint8x8_t __c) { |
| 9141 | return (uint8x8_t)__builtin_neon_vqtbx1_v((int8x8_t)__a, (int8x16_t)__b, (int8x8_t)__c, 16); } |
| 9142 | __ai int8x8_t vqtbx1_s8(int8x8_t __a, int8x16_t __b, int8x8_t __c) { |
| 9143 | return (int8x8_t)__builtin_neon_vqtbx1_v(__a, __b, __c, 0); } |
| 9144 | __ai poly8x8_t vqtbx1_p8(poly8x8_t __a, poly8x16_t __b, uint8x8_t __c) { |
| 9145 | return (poly8x8_t)__builtin_neon_vqtbx1_v((int8x8_t)__a, (int8x16_t)__b, (int8x8_t)__c, 4); } |
| 9146 | __ai uint8x16_t vqtbx1q_u8(uint8x16_t __a, uint8x16_t __b, uint8x16_t __c) { |
| 9147 | return (uint8x16_t)__builtin_neon_vqtbx1q_v((int8x16_t)__a, (int8x16_t)__b, (int8x16_t)__c, 48); } |
| 9148 | __ai int8x16_t vqtbx1q_s8(int8x16_t __a, int8x16_t __b, int8x16_t __c) { |
| 9149 | return (int8x16_t)__builtin_neon_vqtbx1q_v(__a, __b, __c, 32); } |
| 9150 | __ai poly8x16_t vqtbx1q_p8(poly8x16_t __a, poly8x16_t __b, uint8x16_t __c) { |
| 9151 | return (poly8x16_t)__builtin_neon_vqtbx1q_v((int8x16_t)__a, (int8x16_t)__b, (int8x16_t)__c, 36); } |
| 9152 | |
| 9153 | __ai uint8x8_t vqtbx2_u8(uint8x8_t __a, uint8x16x2_t __b, uint8x8_t __c) { |
| 9154 | return (uint8x8_t)__builtin_neon_vqtbx2_v((int8x8_t)__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x8_t)__c, 16); } |
| 9155 | __ai int8x8_t vqtbx2_s8(int8x8_t __a, int8x16x2_t __b, int8x8_t __c) { |
| 9156 | return (int8x8_t)__builtin_neon_vqtbx2_v(__a, __b.val[0], __b.val[1], __c, 0); } |
| 9157 | __ai poly8x8_t vqtbx2_p8(poly8x8_t __a, poly8x16x2_t __b, uint8x8_t __c) { |
| 9158 | return (poly8x8_t)__builtin_neon_vqtbx2_v((int8x8_t)__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x8_t)__c, 4); } |
| 9159 | __ai uint8x16_t vqtbx2q_u8(uint8x16_t __a, uint8x16x2_t __b, uint8x16_t __c) { |
| 9160 | return (uint8x16_t)__builtin_neon_vqtbx2q_v((int8x16_t)__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__c, 48); } |
| 9161 | __ai int8x16_t vqtbx2q_s8(int8x16_t __a, int8x16x2_t __b, int8x16_t __c) { |
| 9162 | return (int8x16_t)__builtin_neon_vqtbx2q_v(__a, __b.val[0], __b.val[1], __c, 32); } |
| 9163 | __ai poly8x16_t vqtbx2q_p8(poly8x16_t __a, poly8x16x2_t __b, uint8x16_t __c) { |
| 9164 | return (poly8x16_t)__builtin_neon_vqtbx2q_v((int8x16_t)__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__c, 36); } |
| 9165 | |
| 9166 | __ai uint8x8_t vqtbx3_u8(uint8x8_t __a, uint8x16x3_t __b, uint8x8_t __c) { |
| 9167 | return (uint8x8_t)__builtin_neon_vqtbx3_v((int8x8_t)__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x8_t)__c, 16); } |
| 9168 | __ai int8x8_t vqtbx3_s8(int8x8_t __a, int8x16x3_t __b, int8x8_t __c) { |
| 9169 | return (int8x8_t)__builtin_neon_vqtbx3_v(__a, __b.val[0], __b.val[1], __b.val[2], __c, 0); } |
| 9170 | __ai poly8x8_t vqtbx3_p8(poly8x8_t __a, poly8x16x3_t __b, uint8x8_t __c) { |
| 9171 | return (poly8x8_t)__builtin_neon_vqtbx3_v((int8x8_t)__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x8_t)__c, 4); } |
| 9172 | __ai uint8x16_t vqtbx3q_u8(uint8x16_t __a, uint8x16x3_t __b, uint8x16_t __c) { |
| 9173 | return (uint8x16_t)__builtin_neon_vqtbx3q_v((int8x16_t)__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__c, 48); } |
| 9174 | __ai int8x16_t vqtbx3q_s8(int8x16_t __a, int8x16x3_t __b, int8x16_t __c) { |
| 9175 | return (int8x16_t)__builtin_neon_vqtbx3q_v(__a, __b.val[0], __b.val[1], __b.val[2], __c, 32); } |
| 9176 | __ai poly8x16_t vqtbx3q_p8(poly8x16_t __a, poly8x16x3_t __b, uint8x16_t __c) { |
| 9177 | return (poly8x16_t)__builtin_neon_vqtbx3q_v((int8x16_t)__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__c, 36); } |
| 9178 | |
| 9179 | __ai uint8x8_t vqtbx4_u8(uint8x8_t __a, uint8x16x4_t __b, uint8x8_t __c) { |
| 9180 | return (uint8x8_t)__builtin_neon_vqtbx4_v((int8x8_t)__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], (int8x8_t)__c, 16); } |
| 9181 | __ai int8x8_t vqtbx4_s8(int8x8_t __a, int8x16x4_t __b, int8x8_t __c) { |
| 9182 | return (int8x8_t)__builtin_neon_vqtbx4_v(__a, __b.val[0], __b.val[1], __b.val[2], __b.val[3], __c, 0); } |
| 9183 | __ai poly8x8_t vqtbx4_p8(poly8x8_t __a, poly8x16x4_t __b, uint8x8_t __c) { |
| 9184 | return (poly8x8_t)__builtin_neon_vqtbx4_v((int8x8_t)__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], (int8x8_t)__c, 4); } |
| 9185 | __ai uint8x16_t vqtbx4q_u8(uint8x16_t __a, uint8x16x4_t __b, uint8x16_t __c) { |
| 9186 | return (uint8x16_t)__builtin_neon_vqtbx4q_v((int8x16_t)__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], (int8x16_t)__c, 48); } |
| 9187 | __ai int8x16_t vqtbx4q_s8(int8x16_t __a, int8x16x4_t __b, int8x16_t __c) { |
| 9188 | return (int8x16_t)__builtin_neon_vqtbx4q_v(__a, __b.val[0], __b.val[1], __b.val[2], __b.val[3], __c, 32); } |
| 9189 | __ai poly8x16_t vqtbx4q_p8(poly8x16_t __a, poly8x16x4_t __b, uint8x16_t __c) { |
| 9190 | return (poly8x16_t)__builtin_neon_vqtbx4q_v((int8x16_t)__a, (int8x16_t)__b.val[0], (int8x16_t)__b.val[1], (int8x16_t)__b.val[2], (int8x16_t)__b.val[3], (int8x16_t)__c, 36); } |
| 9191 | |
| 9192 | __ai int8x16_t vraddhn_high_s16(int8x8_t __a, int16x8_t __b, int16x8_t __c) { |
| 9193 | return vcombine_s8(__a, vraddhn_s16(__b, __c)); } |
| 9194 | __ai int16x8_t vraddhn_high_s32(int16x4_t __a, int32x4_t __b, int32x4_t __c) { |
| 9195 | return vcombine_s16(__a, vraddhn_s32(__b, __c)); } |
| 9196 | __ai int32x4_t vraddhn_high_s64(int32x2_t __a, int64x2_t __b, int64x2_t __c) { |
| 9197 | return vcombine_s32(__a, vraddhn_s64(__b, __c)); } |
| 9198 | __ai uint8x16_t vraddhn_high_u16(uint8x8_t __a, uint16x8_t __b, uint16x8_t __c) { |
| 9199 | return vcombine_u8(__a, vraddhn_u16(__b, __c)); } |
| 9200 | __ai uint16x8_t vraddhn_high_u32(uint16x4_t __a, uint32x4_t __b, uint32x4_t __c) { |
| 9201 | return vcombine_u16(__a, vraddhn_u32(__b, __c)); } |
| 9202 | __ai uint32x4_t vraddhn_high_u64(uint32x2_t __a, uint64x2_t __b, uint64x2_t __c) { |
| 9203 | return vcombine_u32(__a, vraddhn_u64(__b, __c)); } |
| 9204 | |
| 9205 | __ai int8x16_t vrsubhn_high_s16(int8x8_t __a, int16x8_t __b, int16x8_t __c) { |
| 9206 | return vcombine_s8(__a, vrsubhn_s16(__b, __c)); } |
| 9207 | __ai int16x8_t vrsubhn_high_s32(int16x4_t __a, int32x4_t __b, int32x4_t __c) { |
| 9208 | return vcombine_s16(__a, vrsubhn_s32(__b, __c)); } |
| 9209 | __ai int32x4_t vrsubhn_high_s64(int32x2_t __a, int64x2_t __b, int64x2_t __c) { |
| 9210 | return vcombine_s32(__a, vrsubhn_s64(__b, __c)); } |
| 9211 | __ai uint8x16_t vrsubhn_high_u16(uint8x8_t __a, uint16x8_t __b, uint16x8_t __c) { |
| 9212 | return vcombine_u8(__a, vrsubhn_u16(__b, __c)); } |
| 9213 | __ai uint16x8_t vrsubhn_high_u32(uint16x4_t __a, uint32x4_t __b, uint32x4_t __c) { |
| 9214 | return vcombine_u16(__a, vrsubhn_u32(__b, __c)); } |
| 9215 | __ai uint32x4_t vrsubhn_high_u64(uint32x2_t __a, uint64x2_t __b, uint64x2_t __c) { |
| 9216 | return vcombine_u32(__a, vrsubhn_u64(__b, __c)); } |
| 9217 | |
| 9218 | __ai int8x16_t vsubhn_high_s16(int8x8_t __a, int16x8_t __b, int16x8_t __c) { |
| 9219 | return vcombine_s8(__a, vsubhn_s16(__b, __c)); } |
| 9220 | __ai int16x8_t vsubhn_high_s32(int16x4_t __a, int32x4_t __b, int32x4_t __c) { |
| 9221 | return vcombine_s16(__a, vsubhn_s32(__b, __c)); } |
| 9222 | __ai int32x4_t vsubhn_high_s64(int32x2_t __a, int64x2_t __b, int64x2_t __c) { |
| 9223 | return vcombine_s32(__a, vsubhn_s64(__b, __c)); } |
| 9224 | __ai uint8x16_t vsubhn_high_u16(uint8x8_t __a, uint16x8_t __b, uint16x8_t __c) { |
| 9225 | return vcombine_u8(__a, vsubhn_u16(__b, __c)); } |
| 9226 | __ai uint16x8_t vsubhn_high_u32(uint16x4_t __a, uint32x4_t __b, uint32x4_t __c) { |
| 9227 | return vcombine_u16(__a, vsubhn_u32(__b, __c)); } |
| 9228 | __ai uint32x4_t vsubhn_high_u64(uint32x2_t __a, uint64x2_t __b, uint64x2_t __c) { |
| 9229 | return vcombine_u32(__a, vsubhn_u64(__b, __c)); } |
| 9230 | |
| 9231 | __ai int16x8_t vsubl_high_s8(int8x16_t __a, int8x16_t __b) { |
| 9232 | return vmovl_high_s8(__a) - vmovl_high_s8(__b); } |
| 9233 | __ai int32x4_t vsubl_high_s16(int16x8_t __a, int16x8_t __b) { |
| 9234 | return vmovl_high_s16(__a) - vmovl_high_s16(__b); } |
| 9235 | __ai int64x2_t vsubl_high_s32(int32x4_t __a, int32x4_t __b) { |
| 9236 | return vmovl_high_s32(__a) - vmovl_high_s32(__b); } |
| 9237 | __ai uint16x8_t vsubl_high_u8(uint8x16_t __a, uint8x16_t __b) { |
| 9238 | return vmovl_high_u8(__a) - vmovl_high_u8(__b); } |
| 9239 | __ai uint32x4_t vsubl_high_u16(uint16x8_t __a, uint16x8_t __b) { |
| 9240 | return vmovl_high_u16(__a) - vmovl_high_u16(__b); } |
| 9241 | __ai uint64x2_t vsubl_high_u32(uint32x4_t __a, uint32x4_t __b) { |
| 9242 | return vmovl_high_u32(__a) - vmovl_high_u32(__b); } |
| 9243 | |
| 9244 | __ai int16x8_t vsubw_high_s8(int16x8_t __a, int8x16_t __b) { |
| 9245 | return __a - vmovl_high_s8(__b); } |
| 9246 | __ai int32x4_t vsubw_high_s16(int32x4_t __a, int16x8_t __b) { |
| 9247 | return __a - vmovl_high_s16(__b); } |
| 9248 | __ai int64x2_t vsubw_high_s32(int64x2_t __a, int32x4_t __b) { |
| 9249 | return __a - vmovl_high_s32(__b); } |
| 9250 | __ai uint16x8_t vsubw_high_u8(uint16x8_t __a, uint8x16_t __b) { |
| 9251 | return __a - vmovl_high_u8(__b); } |
| 9252 | __ai uint32x4_t vsubw_high_u16(uint32x4_t __a, uint16x8_t __b) { |
| 9253 | return __a - vmovl_high_u16(__b); } |
| 9254 | __ai uint64x2_t vsubw_high_u32(uint64x2_t __a, uint32x4_t __b) { |
| 9255 | return __a - vmovl_high_u32(__b); } |
| 9256 | |
| 9257 | __ai int8x8_t vtrn1_s8(int8x8_t __a, int8x8_t __b) { |
| 9258 | return __builtin_shufflevector(__a, __b, 0, 8, 2, 10, 4, 12, 6, 14); } |
| 9259 | __ai int16x4_t vtrn1_s16(int16x4_t __a, int16x4_t __b) { |
| 9260 | return __builtin_shufflevector(__a, __b, 0, 4, 2, 6); } |
| 9261 | __ai int32x2_t vtrn1_s32(int32x2_t __a, int32x2_t __b) { |
| 9262 | return __builtin_shufflevector(__a, __b, 0, 2); } |
| 9263 | __ai uint8x8_t vtrn1_u8(uint8x8_t __a, uint8x8_t __b) { |
| 9264 | return __builtin_shufflevector(__a, __b, 0, 8, 2, 10, 4, 12, 6, 14); } |
| 9265 | __ai uint16x4_t vtrn1_u16(uint16x4_t __a, uint16x4_t __b) { |
| 9266 | return __builtin_shufflevector(__a, __b, 0, 4, 2, 6); } |
| 9267 | __ai uint32x2_t vtrn1_u32(uint32x2_t __a, uint32x2_t __b) { |
| 9268 | return __builtin_shufflevector(__a, __b, 0, 2); } |
| 9269 | __ai float32x2_t vtrn1_f32(float32x2_t __a, float32x2_t __b) { |
| 9270 | return __builtin_shufflevector(__a, __b, 0, 2); } |
| 9271 | __ai poly8x8_t vtrn1_p8(poly8x8_t __a, poly8x8_t __b) { |
| 9272 | return __builtin_shufflevector(__a, __b, 0, 8, 2, 10, 4, 12, 6, 14); } |
| 9273 | __ai poly16x4_t vtrn1_p16(poly16x4_t __a, poly16x4_t __b) { |
| 9274 | return __builtin_shufflevector(__a, __b, 0, 4, 2, 6); } |
| 9275 | __ai int8x16_t vtrn1q_s8(int8x16_t __a, int8x16_t __b) { |
| 9276 | return __builtin_shufflevector(__a, __b, 0, 16, 2, 18, 4, 20, 6, 22, 8, 24, 10, 26, 12, 28, 14, 30); } |
| 9277 | __ai int16x8_t vtrn1q_s16(int16x8_t __a, int16x8_t __b) { |
| 9278 | return __builtin_shufflevector(__a, __b, 0, 8, 2, 10, 4, 12, 6, 14); } |
| 9279 | __ai int32x4_t vtrn1q_s32(int32x4_t __a, int32x4_t __b) { |
| 9280 | return __builtin_shufflevector(__a, __b, 0, 4, 2, 6); } |
| 9281 | __ai int64x2_t vtrn1q_s64(int64x2_t __a, int64x2_t __b) { |
| 9282 | return __builtin_shufflevector(__a, __b, 0, 2); } |
| 9283 | __ai uint8x16_t vtrn1q_u8(uint8x16_t __a, uint8x16_t __b) { |
| 9284 | return __builtin_shufflevector(__a, __b, 0, 16, 2, 18, 4, 20, 6, 22, 8, 24, 10, 26, 12, 28, 14, 30); } |
| 9285 | __ai uint16x8_t vtrn1q_u16(uint16x8_t __a, uint16x8_t __b) { |
| 9286 | return __builtin_shufflevector(__a, __b, 0, 8, 2, 10, 4, 12, 6, 14); } |
| 9287 | __ai uint32x4_t vtrn1q_u32(uint32x4_t __a, uint32x4_t __b) { |
| 9288 | return __builtin_shufflevector(__a, __b, 0, 4, 2, 6); } |
| 9289 | __ai uint64x2_t vtrn1q_u64(uint64x2_t __a, uint64x2_t __b) { |
| 9290 | return __builtin_shufflevector(__a, __b, 0, 2); } |
| 9291 | __ai float32x4_t vtrn1q_f32(float32x4_t __a, float32x4_t __b) { |
| 9292 | return __builtin_shufflevector(__a, __b, 0, 4, 2, 6); } |
| 9293 | __ai float64x2_t vtrn1q_f64(float64x2_t __a, float64x2_t __b) { |
| 9294 | return __builtin_shufflevector(__a, __b, 0, 2); } |
| 9295 | __ai poly8x16_t vtrn1q_p8(poly8x16_t __a, poly8x16_t __b) { |
| 9296 | return __builtin_shufflevector(__a, __b, 0, 16, 2, 18, 4, 20, 6, 22, 8, 24, 10, 26, 12, 28, 14, 30); } |
| 9297 | __ai poly16x8_t vtrn1q_p16(poly16x8_t __a, poly16x8_t __b) { |
| 9298 | return __builtin_shufflevector(__a, __b, 0, 8, 2, 10, 4, 12, 6, 14); } |
| 9299 | __ai poly64x2_t vtrn1q_p64(poly64x2_t __a, poly64x2_t __b) { |
| 9300 | return __builtin_shufflevector(__a, __b, 0, 2); } |
| 9301 | |
| 9302 | __ai int8x8_t vtrn2_s8(int8x8_t __a, int8x8_t __b) { |
| 9303 | return __builtin_shufflevector(__a, __b, 1, 9, 3, 11, 5, 13, 7, 15); } |
| 9304 | __ai int16x4_t vtrn2_s16(int16x4_t __a, int16x4_t __b) { |
| 9305 | return __builtin_shufflevector(__a, __b, 1, 5, 3, 7); } |
| 9306 | __ai int32x2_t vtrn2_s32(int32x2_t __a, int32x2_t __b) { |
| 9307 | return __builtin_shufflevector(__a, __b, 1, 3); } |
| 9308 | __ai uint8x8_t vtrn2_u8(uint8x8_t __a, uint8x8_t __b) { |
| 9309 | return __builtin_shufflevector(__a, __b, 1, 9, 3, 11, 5, 13, 7, 15); } |
| 9310 | __ai uint16x4_t vtrn2_u16(uint16x4_t __a, uint16x4_t __b) { |
| 9311 | return __builtin_shufflevector(__a, __b, 1, 5, 3, 7); } |
| 9312 | __ai uint32x2_t vtrn2_u32(uint32x2_t __a, uint32x2_t __b) { |
| 9313 | return __builtin_shufflevector(__a, __b, 1, 3); } |
| 9314 | __ai float32x2_t vtrn2_f32(float32x2_t __a, float32x2_t __b) { |
| 9315 | return __builtin_shufflevector(__a, __b, 1, 3); } |
| 9316 | __ai poly8x8_t vtrn2_p8(poly8x8_t __a, poly8x8_t __b) { |
| 9317 | return __builtin_shufflevector(__a, __b, 1, 9, 3, 11, 5, 13, 7, 15); } |
| 9318 | __ai poly16x4_t vtrn2_p16(poly16x4_t __a, poly16x4_t __b) { |
| 9319 | return __builtin_shufflevector(__a, __b, 1, 5, 3, 7); } |
| 9320 | __ai int8x16_t vtrn2q_s8(int8x16_t __a, int8x16_t __b) { |
| 9321 | return __builtin_shufflevector(__a, __b, 1, 17, 3, 19, 5, 21, 7, 23, 9, 25, 11, 27, 13, 29, 15, 31); } |
| 9322 | __ai int16x8_t vtrn2q_s16(int16x8_t __a, int16x8_t __b) { |
| 9323 | return __builtin_shufflevector(__a, __b, 1, 9, 3, 11, 5, 13, 7, 15); } |
| 9324 | __ai int32x4_t vtrn2q_s32(int32x4_t __a, int32x4_t __b) { |
| 9325 | return __builtin_shufflevector(__a, __b, 1, 5, 3, 7); } |
| 9326 | __ai int64x2_t vtrn2q_s64(int64x2_t __a, int64x2_t __b) { |
| 9327 | return __builtin_shufflevector(__a, __b, 1, 3); } |
| 9328 | __ai uint8x16_t vtrn2q_u8(uint8x16_t __a, uint8x16_t __b) { |
| 9329 | return __builtin_shufflevector(__a, __b, 1, 17, 3, 19, 5, 21, 7, 23, 9, 25, 11, 27, 13, 29, 15, 31); } |
| 9330 | __ai uint16x8_t vtrn2q_u16(uint16x8_t __a, uint16x8_t __b) { |
| 9331 | return __builtin_shufflevector(__a, __b, 1, 9, 3, 11, 5, 13, 7, 15); } |
| 9332 | __ai uint32x4_t vtrn2q_u32(uint32x4_t __a, uint32x4_t __b) { |
| 9333 | return __builtin_shufflevector(__a, __b, 1, 5, 3, 7); } |
| 9334 | __ai uint64x2_t vtrn2q_u64(uint64x2_t __a, uint64x2_t __b) { |
| 9335 | return __builtin_shufflevector(__a, __b, 1, 3); } |
| 9336 | __ai float32x4_t vtrn2q_f32(float32x4_t __a, float32x4_t __b) { |
| 9337 | return __builtin_shufflevector(__a, __b, 1, 5, 3, 7); } |
| 9338 | __ai float64x2_t vtrn2q_f64(float64x2_t __a, float64x2_t __b) { |
| 9339 | return __builtin_shufflevector(__a, __b, 1, 3); } |
| 9340 | __ai poly8x16_t vtrn2q_p8(poly8x16_t __a, poly8x16_t __b) { |
| 9341 | return __builtin_shufflevector(__a, __b, 1, 17, 3, 19, 5, 21, 7, 23, 9, 25, 11, 27, 13, 29, 15, 31); } |
| 9342 | __ai poly16x8_t vtrn2q_p16(poly16x8_t __a, poly16x8_t __b) { |
| 9343 | return __builtin_shufflevector(__a, __b, 1, 9, 3, 11, 5, 13, 7, 15); } |
| 9344 | __ai poly64x2_t vtrn2q_p64(poly64x2_t __a, poly64x2_t __b) { |
| 9345 | return __builtin_shufflevector(__a, __b, 1, 3); } |
| 9346 | |
| 9347 | __ai int8x8_t vuzp1_s8(int8x8_t __a, int8x8_t __b) { |
| 9348 | return __builtin_shufflevector(__a, __b, 0, 2, 4, 6, 8, 10, 12, 14); } |
| 9349 | __ai int16x4_t vuzp1_s16(int16x4_t __a, int16x4_t __b) { |
| 9350 | return __builtin_shufflevector(__a, __b, 0, 2, 4, 6); } |
| 9351 | __ai int32x2_t vuzp1_s32(int32x2_t __a, int32x2_t __b) { |
| 9352 | return __builtin_shufflevector(__a, __b, 0, 2); } |
| 9353 | __ai uint8x8_t vuzp1_u8(uint8x8_t __a, uint8x8_t __b) { |
| 9354 | return __builtin_shufflevector(__a, __b, 0, 2, 4, 6, 8, 10, 12, 14); } |
| 9355 | __ai uint16x4_t vuzp1_u16(uint16x4_t __a, uint16x4_t __b) { |
| 9356 | return __builtin_shufflevector(__a, __b, 0, 2, 4, 6); } |
| 9357 | __ai uint32x2_t vuzp1_u32(uint32x2_t __a, uint32x2_t __b) { |
| 9358 | return __builtin_shufflevector(__a, __b, 0, 2); } |
| 9359 | __ai float32x2_t vuzp1_f32(float32x2_t __a, float32x2_t __b) { |
| 9360 | return __builtin_shufflevector(__a, __b, 0, 2); } |
| 9361 | __ai poly8x8_t vuzp1_p8(poly8x8_t __a, poly8x8_t __b) { |
| 9362 | return __builtin_shufflevector(__a, __b, 0, 2, 4, 6, 8, 10, 12, 14); } |
| 9363 | __ai poly16x4_t vuzp1_p16(poly16x4_t __a, poly16x4_t __b) { |
| 9364 | return __builtin_shufflevector(__a, __b, 0, 2, 4, 6); } |
| 9365 | __ai int8x16_t vuzp1q_s8(int8x16_t __a, int8x16_t __b) { |
| 9366 | return __builtin_shufflevector(__a, __b, 0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30); } |
| 9367 | __ai int16x8_t vuzp1q_s16(int16x8_t __a, int16x8_t __b) { |
| 9368 | return __builtin_shufflevector(__a, __b, 0, 2, 4, 6, 8, 10, 12, 14); } |
| 9369 | __ai int32x4_t vuzp1q_s32(int32x4_t __a, int32x4_t __b) { |
| 9370 | return __builtin_shufflevector(__a, __b, 0, 2, 4, 6); } |
| 9371 | __ai int64x2_t vuzp1q_s64(int64x2_t __a, int64x2_t __b) { |
| 9372 | return __builtin_shufflevector(__a, __b, 0, 2); } |
| 9373 | __ai uint8x16_t vuzp1q_u8(uint8x16_t __a, uint8x16_t __b) { |
| 9374 | return __builtin_shufflevector(__a, __b, 0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30); } |
| 9375 | __ai uint16x8_t vuzp1q_u16(uint16x8_t __a, uint16x8_t __b) { |
| 9376 | return __builtin_shufflevector(__a, __b, 0, 2, 4, 6, 8, 10, 12, 14); } |
| 9377 | __ai uint32x4_t vuzp1q_u32(uint32x4_t __a, uint32x4_t __b) { |
| 9378 | return __builtin_shufflevector(__a, __b, 0, 2, 4, 6); } |
| 9379 | __ai uint64x2_t vuzp1q_u64(uint64x2_t __a, uint64x2_t __b) { |
| 9380 | return __builtin_shufflevector(__a, __b, 0, 2); } |
| 9381 | __ai float32x4_t vuzp1q_f32(float32x4_t __a, float32x4_t __b) { |
| 9382 | return __builtin_shufflevector(__a, __b, 0, 2, 4, 6); } |
| 9383 | __ai float64x2_t vuzp1q_f64(float64x2_t __a, float64x2_t __b) { |
| 9384 | return __builtin_shufflevector(__a, __b, 0, 2); } |
| 9385 | __ai poly8x16_t vuzp1q_p8(poly8x16_t __a, poly8x16_t __b) { |
| 9386 | return __builtin_shufflevector(__a, __b, 0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30); } |
| 9387 | __ai poly16x8_t vuzp1q_p16(poly16x8_t __a, poly16x8_t __b) { |
| 9388 | return __builtin_shufflevector(__a, __b, 0, 2, 4, 6, 8, 10, 12, 14); } |
| 9389 | __ai poly64x2_t vuzp1q_p64(poly64x2_t __a, poly64x2_t __b) { |
| 9390 | return __builtin_shufflevector(__a, __b, 0, 2); } |
| 9391 | |
| 9392 | __ai int8x8_t vuzp2_s8(int8x8_t __a, int8x8_t __b) { |
| 9393 | return __builtin_shufflevector(__a, __b, 1, 3, 5, 7, 9, 11, 13, 15); } |
| 9394 | __ai int16x4_t vuzp2_s16(int16x4_t __a, int16x4_t __b) { |
| 9395 | return __builtin_shufflevector(__a, __b, 1, 3, 5, 7); } |
| 9396 | __ai int32x2_t vuzp2_s32(int32x2_t __a, int32x2_t __b) { |
| 9397 | return __builtin_shufflevector(__a, __b, 1, 3); } |
| 9398 | __ai uint8x8_t vuzp2_u8(uint8x8_t __a, uint8x8_t __b) { |
| 9399 | return __builtin_shufflevector(__a, __b, 1, 3, 5, 7, 9, 11, 13, 15); } |
| 9400 | __ai uint16x4_t vuzp2_u16(uint16x4_t __a, uint16x4_t __b) { |
| 9401 | return __builtin_shufflevector(__a, __b, 1, 3, 5, 7); } |
| 9402 | __ai uint32x2_t vuzp2_u32(uint32x2_t __a, uint32x2_t __b) { |
| 9403 | return __builtin_shufflevector(__a, __b, 1, 3); } |
| 9404 | __ai float32x2_t vuzp2_f32(float32x2_t __a, float32x2_t __b) { |
| 9405 | return __builtin_shufflevector(__a, __b, 1, 3); } |
| 9406 | __ai poly8x8_t vuzp2_p8(poly8x8_t __a, poly8x8_t __b) { |
| 9407 | return __builtin_shufflevector(__a, __b, 1, 3, 5, 7, 9, 11, 13, 15); } |
| 9408 | __ai poly16x4_t vuzp2_p16(poly16x4_t __a, poly16x4_t __b) { |
| 9409 | return __builtin_shufflevector(__a, __b, 1, 3, 5, 7); } |
| 9410 | __ai int8x16_t vuzp2q_s8(int8x16_t __a, int8x16_t __b) { |
| 9411 | return __builtin_shufflevector(__a, __b, 1, 3, 5, 7, 9, 11, 13, 15, 17, 19, 21, 23, 25, 27, 29, 31); } |
| 9412 | __ai int16x8_t vuzp2q_s16(int16x8_t __a, int16x8_t __b) { |
| 9413 | return __builtin_shufflevector(__a, __b, 1, 3, 5, 7, 9, 11, 13, 15); } |
| 9414 | __ai int32x4_t vuzp2q_s32(int32x4_t __a, int32x4_t __b) { |
| 9415 | return __builtin_shufflevector(__a, __b, 1, 3, 5, 7); } |
| 9416 | __ai int64x2_t vuzp2q_s64(int64x2_t __a, int64x2_t __b) { |
| 9417 | return __builtin_shufflevector(__a, __b, 1, 3); } |
| 9418 | __ai uint8x16_t vuzp2q_u8(uint8x16_t __a, uint8x16_t __b) { |
| 9419 | return __builtin_shufflevector(__a, __b, 1, 3, 5, 7, 9, 11, 13, 15, 17, 19, 21, 23, 25, 27, 29, 31); } |
| 9420 | __ai uint16x8_t vuzp2q_u16(uint16x8_t __a, uint16x8_t __b) { |
| 9421 | return __builtin_shufflevector(__a, __b, 1, 3, 5, 7, 9, 11, 13, 15); } |
| 9422 | __ai uint32x4_t vuzp2q_u32(uint32x4_t __a, uint32x4_t __b) { |
| 9423 | return __builtin_shufflevector(__a, __b, 1, 3, 5, 7); } |
| 9424 | __ai uint64x2_t vuzp2q_u64(uint64x2_t __a, uint64x2_t __b) { |
| 9425 | return __builtin_shufflevector(__a, __b, 1, 3); } |
| 9426 | __ai float32x4_t vuzp2q_f32(float32x4_t __a, float32x4_t __b) { |
| 9427 | return __builtin_shufflevector(__a, __b, 1, 3, 5, 7); } |
| 9428 | __ai float64x2_t vuzp2q_f64(float64x2_t __a, float64x2_t __b) { |
| 9429 | return __builtin_shufflevector(__a, __b, 1, 3); } |
| 9430 | __ai poly8x16_t vuzp2q_p8(poly8x16_t __a, poly8x16_t __b) { |
| 9431 | return __builtin_shufflevector(__a, __b, 1, 3, 5, 7, 9, 11, 13, 15, 17, 19, 21, 23, 25, 27, 29, 31); } |
| 9432 | __ai poly16x8_t vuzp2q_p16(poly16x8_t __a, poly16x8_t __b) { |
| 9433 | return __builtin_shufflevector(__a, __b, 1, 3, 5, 7, 9, 11, 13, 15); } |
| 9434 | __ai poly64x2_t vuzp2q_p64(poly64x2_t __a, poly64x2_t __b) { |
| 9435 | return __builtin_shufflevector(__a, __b, 1, 3); } |
| 9436 | |
| 9437 | __ai int8x8_t vzip1_s8(int8x8_t __a, int8x8_t __b) { |
| 9438 | return __builtin_shufflevector(__a, __b, 0, 8, 1, 9, 2, 10, 3, 11); } |
| 9439 | __ai int16x4_t vzip1_s16(int16x4_t __a, int16x4_t __b) { |
| 9440 | return __builtin_shufflevector(__a, __b, 0, 4, 1, 5); } |
| 9441 | __ai int32x2_t vzip1_s32(int32x2_t __a, int32x2_t __b) { |
| 9442 | return __builtin_shufflevector(__a, __b, 0, 2); } |
| 9443 | __ai uint8x8_t vzip1_u8(uint8x8_t __a, uint8x8_t __b) { |
| 9444 | return __builtin_shufflevector(__a, __b, 0, 8, 1, 9, 2, 10, 3, 11); } |
| 9445 | __ai uint16x4_t vzip1_u16(uint16x4_t __a, uint16x4_t __b) { |
| 9446 | return __builtin_shufflevector(__a, __b, 0, 4, 1, 5); } |
| 9447 | __ai uint32x2_t vzip1_u32(uint32x2_t __a, uint32x2_t __b) { |
| 9448 | return __builtin_shufflevector(__a, __b, 0, 2); } |
| 9449 | __ai float32x2_t vzip1_f32(float32x2_t __a, float32x2_t __b) { |
| 9450 | return __builtin_shufflevector(__a, __b, 0, 2); } |
| 9451 | __ai poly8x8_t vzip1_p8(poly8x8_t __a, poly8x8_t __b) { |
| 9452 | return __builtin_shufflevector(__a, __b, 0, 8, 1, 9, 2, 10, 3, 11); } |
| 9453 | __ai poly16x4_t vzip1_p16(poly16x4_t __a, poly16x4_t __b) { |
| 9454 | return __builtin_shufflevector(__a, __b, 0, 4, 1, 5); } |
| 9455 | __ai int8x16_t vzip1q_s8(int8x16_t __a, int8x16_t __b) { |
| 9456 | return __builtin_shufflevector(__a, __b, 0, 16, 1, 17, 2, 18, 3, 19, 4, 20, 5, 21, 6, 22, 7, 23); } |
| 9457 | __ai int16x8_t vzip1q_s16(int16x8_t __a, int16x8_t __b) { |
| 9458 | return __builtin_shufflevector(__a, __b, 0, 8, 1, 9, 2, 10, 3, 11); } |
| 9459 | __ai int32x4_t vzip1q_s32(int32x4_t __a, int32x4_t __b) { |
| 9460 | return __builtin_shufflevector(__a, __b, 0, 4, 1, 5); } |
| 9461 | __ai int64x2_t vzip1q_s64(int64x2_t __a, int64x2_t __b) { |
| 9462 | return __builtin_shufflevector(__a, __b, 0, 2); } |
| 9463 | __ai uint8x16_t vzip1q_u8(uint8x16_t __a, uint8x16_t __b) { |
| 9464 | return __builtin_shufflevector(__a, __b, 0, 16, 1, 17, 2, 18, 3, 19, 4, 20, 5, 21, 6, 22, 7, 23); } |
| 9465 | __ai uint16x8_t vzip1q_u16(uint16x8_t __a, uint16x8_t __b) { |
| 9466 | return __builtin_shufflevector(__a, __b, 0, 8, 1, 9, 2, 10, 3, 11); } |
| 9467 | __ai uint32x4_t vzip1q_u32(uint32x4_t __a, uint32x4_t __b) { |
| 9468 | return __builtin_shufflevector(__a, __b, 0, 4, 1, 5); } |
| 9469 | __ai uint64x2_t vzip1q_u64(uint64x2_t __a, uint64x2_t __b) { |
| 9470 | return __builtin_shufflevector(__a, __b, 0, 2); } |
| 9471 | __ai float32x4_t vzip1q_f32(float32x4_t __a, float32x4_t __b) { |
| 9472 | return __builtin_shufflevector(__a, __b, 0, 4, 1, 5); } |
| 9473 | __ai float64x2_t vzip1q_f64(float64x2_t __a, float64x2_t __b) { |
| 9474 | return __builtin_shufflevector(__a, __b, 0, 2); } |
| 9475 | __ai poly8x16_t vzip1q_p8(poly8x16_t __a, poly8x16_t __b) { |
| 9476 | return __builtin_shufflevector(__a, __b, 0, 16, 1, 17, 2, 18, 3, 19, 4, 20, 5, 21, 6, 22, 7, 23); } |
| 9477 | __ai poly16x8_t vzip1q_p16(poly16x8_t __a, poly16x8_t __b) { |
| 9478 | return __builtin_shufflevector(__a, __b, 0, 8, 1, 9, 2, 10, 3, 11); } |
| 9479 | __ai poly64x2_t vzip1q_p64(poly64x2_t __a, poly64x2_t __b) { |
| 9480 | return __builtin_shufflevector(__a, __b, 0, 2); } |
| 9481 | |
| 9482 | __ai int8x8_t vzip2_s8(int8x8_t __a, int8x8_t __b) { |
| 9483 | return __builtin_shufflevector(__a, __b, 4, 12, 5, 13, 6, 14, 7, 15); } |
| 9484 | __ai int16x4_t vzip2_s16(int16x4_t __a, int16x4_t __b) { |
| 9485 | return __builtin_shufflevector(__a, __b, 2, 6, 3, 7); } |
| 9486 | __ai int32x2_t vzip2_s32(int32x2_t __a, int32x2_t __b) { |
| 9487 | return __builtin_shufflevector(__a, __b, 1, 3); } |
| 9488 | __ai uint8x8_t vzip2_u8(uint8x8_t __a, uint8x8_t __b) { |
| 9489 | return __builtin_shufflevector(__a, __b, 4, 12, 5, 13, 6, 14, 7, 15); } |
| 9490 | __ai uint16x4_t vzip2_u16(uint16x4_t __a, uint16x4_t __b) { |
| 9491 | return __builtin_shufflevector(__a, __b, 2, 6, 3, 7); } |
| 9492 | __ai uint32x2_t vzip2_u32(uint32x2_t __a, uint32x2_t __b) { |
| 9493 | return __builtin_shufflevector(__a, __b, 1, 3); } |
| 9494 | __ai float32x2_t vzip2_f32(float32x2_t __a, float32x2_t __b) { |
| 9495 | return __builtin_shufflevector(__a, __b, 1, 3); } |
| 9496 | __ai poly8x8_t vzip2_p8(poly8x8_t __a, poly8x8_t __b) { |
| 9497 | return __builtin_shufflevector(__a, __b, 4, 12, 5, 13, 6, 14, 7, 15); } |
| 9498 | __ai poly16x4_t vzip2_p16(poly16x4_t __a, poly16x4_t __b) { |
| 9499 | return __builtin_shufflevector(__a, __b, 2, 6, 3, 7); } |
| 9500 | __ai int8x16_t vzip2q_s8(int8x16_t __a, int8x16_t __b) { |
| 9501 | return __builtin_shufflevector(__a, __b, 8, 24, 9, 25, 10, 26, 11, 27, 12, 28, 13, 29, 14, 30, 15, 31); } |
| 9502 | __ai int16x8_t vzip2q_s16(int16x8_t __a, int16x8_t __b) { |
| 9503 | return __builtin_shufflevector(__a, __b, 4, 12, 5, 13, 6, 14, 7, 15); } |
| 9504 | __ai int32x4_t vzip2q_s32(int32x4_t __a, int32x4_t __b) { |
| 9505 | return __builtin_shufflevector(__a, __b, 2, 6, 3, 7); } |
| 9506 | __ai int64x2_t vzip2q_s64(int64x2_t __a, int64x2_t __b) { |
| 9507 | return __builtin_shufflevector(__a, __b, 1, 3); } |
| 9508 | __ai uint8x16_t vzip2q_u8(uint8x16_t __a, uint8x16_t __b) { |
| 9509 | return __builtin_shufflevector(__a, __b, 8, 24, 9, 25, 10, 26, 11, 27, 12, 28, 13, 29, 14, 30, 15, 31); } |
| 9510 | __ai uint16x8_t vzip2q_u16(uint16x8_t __a, uint16x8_t __b) { |
| 9511 | return __builtin_shufflevector(__a, __b, 4, 12, 5, 13, 6, 14, 7, 15); } |
| 9512 | __ai uint32x4_t vzip2q_u32(uint32x4_t __a, uint32x4_t __b) { |
| 9513 | return __builtin_shufflevector(__a, __b, 2, 6, 3, 7); } |
| 9514 | __ai uint64x2_t vzip2q_u64(uint64x2_t __a, uint64x2_t __b) { |
| 9515 | return __builtin_shufflevector(__a, __b, 1, 3); } |
| 9516 | __ai float32x4_t vzip2q_f32(float32x4_t __a, float32x4_t __b) { |
| 9517 | return __builtin_shufflevector(__a, __b, 2, 6, 3, 7); } |
| 9518 | __ai float64x2_t vzip2q_f64(float64x2_t __a, float64x2_t __b) { |
| 9519 | return __builtin_shufflevector(__a, __b, 1, 3); } |
| 9520 | __ai poly8x16_t vzip2q_p8(poly8x16_t __a, poly8x16_t __b) { |
| 9521 | return __builtin_shufflevector(__a, __b, 8, 24, 9, 25, 10, 26, 11, 27, 12, 28, 13, 29, 14, 30, 15, 31); } |
| 9522 | __ai poly16x8_t vzip2q_p16(poly16x8_t __a, poly16x8_t __b) { |
| 9523 | return __builtin_shufflevector(__a, __b, 4, 12, 5, 13, 6, 14, 7, 15); } |
| 9524 | __ai poly64x2_t vzip2q_p64(poly64x2_t __a, poly64x2_t __b) { |
| 9525 | return __builtin_shufflevector(__a, __b, 1, 3); } |
| 9526 | |
| 9527 | __ai int8x16_t vmovn_high_s16(int8x8_t __a, int16x8_t __b) { |
| 9528 | int8x8_t __a1 = vmovn_s16(__b); |
| 9529 | return __builtin_shufflevector(__a, __a1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15); } |
| 9530 | __ai int16x8_t vmovn_high_s32(int16x4_t __a, int32x4_t __b) { |
| 9531 | int16x4_t __a1 = vmovn_s32(__b); |
| 9532 | return __builtin_shufflevector(__a, __a1, 0, 1, 2, 3, 4, 5, 6, 7); } |
| 9533 | __ai int32x4_t vmovn_high_s64(int32x2_t __a, int64x2_t __b) { |
| 9534 | int32x2_t __a1 = vmovn_s64(__b); |
| 9535 | return __builtin_shufflevector(__a, __a1, 0, 1, 2, 3); } |
| 9536 | __ai uint8x16_t vmovn_high_u16(uint8x8_t __a, uint16x8_t __b) { |
| 9537 | uint8x8_t __a1 = vmovn_u16(__b); |
| 9538 | return __builtin_shufflevector(__a, __a1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15); } |
| 9539 | __ai uint16x8_t vmovn_high_u32(uint16x4_t __a, uint32x4_t __b) { |
| 9540 | uint16x4_t __a1 = vmovn_u32(__b); |
| 9541 | return __builtin_shufflevector(__a, __a1, 0, 1, 2, 3, 4, 5, 6, 7); } |
| 9542 | __ai uint32x4_t vmovn_high_u64(uint32x2_t __a, uint64x2_t __b) { |
| 9543 | uint32x2_t __a1 = vmovn_u64(__b); |
| 9544 | return __builtin_shufflevector(__a, __a1, 0, 1, 2, 3); } |
| 9545 | |
| 9546 | #ifdef __ARM_FEATURE_CRYPTO |
| 9547 | __ai uint8x16_t vaesdq_u8(uint8x16_t __a, uint8x16_t __b) { |
| 9548 | return (uint8x16_t)__builtin_neon_vaesdq_v((int8x16_t)__a, (int8x16_t)__b, 48); } |
| 9549 | |
| 9550 | __ai uint8x16_t vaeseq_u8(uint8x16_t __a, uint8x16_t __b) { |
| 9551 | return (uint8x16_t)__builtin_neon_vaeseq_v((int8x16_t)__a, (int8x16_t)__b, 48); } |
| 9552 | |
| 9553 | __ai uint8x16_t vaesimcq_u8(uint8x16_t __a) { |
| 9554 | return (uint8x16_t)__builtin_neon_vaesimcq_v((int8x16_t)__a, 48); } |
| 9555 | |
| 9556 | __ai uint8x16_t vaesmcq_u8(uint8x16_t __a) { |
| 9557 | return (uint8x16_t)__builtin_neon_vaesmcq_v((int8x16_t)__a, 48); } |
| 9558 | |
| 9559 | __ai uint32x4_t vsha1cq_u32(uint32x4_t __a, uint32_t __b, uint32x4_t __c) { |
| 9560 | return (uint32x4_t)__builtin_neon_vsha1cq_u32((int32x4_t)__a, __b, (int32x4_t)__c); } |
| 9561 | |
| 9562 | __ai uint32_t vsha1h_u32(uint32_t __a) { |
| 9563 | return (uint32_t)__builtin_neon_vsha1h_u32(__a); } |
| 9564 | |
| 9565 | __ai uint32x4_t vsha1mq_u32(uint32x4_t __a, uint32_t __b, uint32x4_t __c) { |
| 9566 | return (uint32x4_t)__builtin_neon_vsha1mq_u32((int32x4_t)__a, __b, (int32x4_t)__c); } |
| 9567 | |
| 9568 | __ai uint32x4_t vsha1pq_u32(uint32x4_t __a, uint32_t __b, uint32x4_t __c) { |
| 9569 | return (uint32x4_t)__builtin_neon_vsha1pq_u32((int32x4_t)__a, __b, (int32x4_t)__c); } |
| 9570 | |
| 9571 | __ai uint32x4_t vsha1su0q_u32(uint32x4_t __a, uint32x4_t __b, uint32x4_t __c) { |
| 9572 | return (uint32x4_t)__builtin_neon_vsha1su0q_v((int8x16_t)__a, (int8x16_t)__b, (int8x16_t)__c, 50); } |
| 9573 | |
| 9574 | __ai uint32x4_t vsha1su1q_u32(uint32x4_t __a, uint32x4_t __b) { |
| 9575 | return (uint32x4_t)__builtin_neon_vsha1su1q_v((int8x16_t)__a, (int8x16_t)__b, 50); } |
| 9576 | |
| 9577 | __ai uint32x4_t vsha256hq_u32(uint32x4_t __a, uint32x4_t __b, uint32x4_t __c) { |
| 9578 | return (uint32x4_t)__builtin_neon_vsha256hq_v((int8x16_t)__a, (int8x16_t)__b, (int8x16_t)__c, 50); } |
| 9579 | |
| 9580 | __ai uint32x4_t vsha256h2q_u32(uint32x4_t __a, uint32x4_t __b, uint32x4_t __c) { |
| 9581 | return (uint32x4_t)__builtin_neon_vsha256h2q_v((int8x16_t)__a, (int8x16_t)__b, (int8x16_t)__c, 50); } |
| 9582 | |
| 9583 | __ai uint32x4_t vsha256su0q_u32(uint32x4_t __a, uint32x4_t __b) { |
| 9584 | return (uint32x4_t)__builtin_neon_vsha256su0q_v((int8x16_t)__a, (int8x16_t)__b, 50); } |
| 9585 | |
| 9586 | __ai uint32x4_t vsha256su1q_u32(uint32x4_t __a, uint32x4_t __b, uint32x4_t __c) { |
| 9587 | return (uint32x4_t)__builtin_neon_vsha256su1q_v((int8x16_t)__a, (int8x16_t)__b, (int8x16_t)__c, 50); } |
| 9588 | |
| 9589 | #endif |
| 9590 | |
| 9591 | #endif |
| 9592 | |
| 9593 | #undef __ai |
| 9594 | |
| 9595 | #endif /* __ARM_NEON_H */ |