Parent Directory
| 
Revision Log
Revision 434 - (view) (download)
| 1 : | chl | 434 | ;/***************************************************************************** |
| 2 : | ; * | ||
| 3 : | ; * XVID MPEG-4 VIDEO CODEC | ||
| 4 : | ; * mmx yuv planar to yv12 conversion | ||
| 5 : | ; * | ||
| 6 : | ; * Copyright (C) 2001 - Michael Militzer <isibaar@xvid.org> | ||
| 7 : | ; * | ||
| 8 : | ; * This program is an implementation of a part of one or more MPEG-4 | ||
| 9 : | ; * Video tools as specified in ISO/IEC 14496-2 standard. Those intending | ||
| 10 : | ; * to use this software module in hardware or software products are | ||
| 11 : | ; * advised that its use may infringe existing patents or copyrights, and | ||
| 12 : | ; * any such use would be at such party's own risk. The original | ||
| 13 : | ; * developer of this software module and his/her company, and subsequent | ||
| 14 : | ; * editors and their companies, will have no liability for use of this | ||
| 15 : | ; * software or modifications or derivatives thereof. | ||
| 16 : | ; * | ||
| 17 : | ; * This program is free software; you can redistribute it and/or modify | ||
| 18 : | ; * it under the terms of the GNU General Public License as published by | ||
| 19 : | ; * the Free Software Foundation; either version 2 of the License, or | ||
| 20 : | ; * (at your option) any later version. | ||
| 21 : | ; * | ||
| 22 : | ; * This program is distributed in the hope that it will be useful, | ||
| 23 : | ; * but WITHOUT ANY WARRANTY; without even the implied warranty of | ||
| 24 : | ; * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | ||
| 25 : | ; * GNU General Public License for more details. | ||
| 26 : | ; * | ||
| 27 : | ; * You should have received a copy of the GNU General Public License | ||
| 28 : | ; * along with this program; if not, write to the Free Software | ||
| 29 : | ; * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA | ||
| 30 : | ; * | ||
| 31 : | ; * $Id: yuv_to_yv12_mmx.asm,v 1.6 2002-09-06 17:48:58 chl Exp $ | ||
| 32 : | ; * | ||
| 33 : | ; ****************************************************************************/ | ||
| 34 : | edgomez | 328 | |
| 35 : | BITS 32 | ||
| 36 : | |||
| 37 : | %macro cglobal 1 | ||
| 38 : | %ifdef PREFIX | ||
| 39 : | global _%1 | ||
| 40 : | edgomez | 331 | %define %1 _%1 |
| 41 : | edgomez | 328 | %else |
| 42 : | global %1 | ||
| 43 : | %endif | ||
| 44 : | %endmacro | ||
| 45 : | |||
| 46 : | SECTION .text | ||
| 47 : | |||
| 48 : | ALIGN 64 | ||
| 49 : | |||
| 50 : | ;------------------------------------------------------------------------------ | ||
| 51 : | ; | ||
| 52 : | ; void yuv_to_yv12_xmm(uint8_t *y_out, | ||
| 53 : | ; uint8_t *u_out, | ||
| 54 : | ; uint8_t *v_out, | ||
| 55 : | ; uint8_t *src, | ||
| 56 : | ; int width, int height, int stride); | ||
| 57 : | ; | ||
| 58 : | ; This function probably also runs on PentiumII class cpu's | ||
| 59 : | ; | ||
| 60 : | ; Attention: This code assumes that width is a multiple of 16 | ||
| 61 : | ; | ||
| 62 : | ;------------------------------------------------------------------------------ | ||
| 63 : | |||
| 64 : | |||
| 65 : | cglobal yuv_to_yv12_xmm | ||
| 66 : | yuv_to_yv12_xmm: | ||
| 67 : | |||
| 68 : | push ebx | ||
| 69 : | push esi | ||
| 70 : | push edi | ||
| 71 : | push ebp | ||
| 72 : | |||
| 73 : | ; local vars allocation | ||
| 74 : | %define localsize 4 | ||
| 75 : | %define remainder esp | ||
| 76 : | sub esp, localsize | ||
| 77 : | |||
| 78 : | ; function code | ||
| 79 : | mov eax, [esp + 40 + localsize] ; height -> eax | ||
| 80 : | mov ebx, [esp + 44 + localsize] ; stride -> ebx | ||
| 81 : | mov esi, [esp + 32 + localsize] ; src -> esi | ||
| 82 : | mov edi, [esp + 20 + localsize] ; y_out -> edi | ||
| 83 : | mov ecx, [esp + 36 + localsize] ; width -> ecx | ||
| 84 : | edgomez | 331 | |
| 85 : | edgomez | 328 | sub ebx, ecx ; stride - width -> ebx |
| 86 : | |||
| 87 : | mov edx, ecx | ||
| 88 : | mov ebp, ecx | ||
| 89 : | shr edx, 6 | ||
| 90 : | mov ecx, edx ; 64 bytes copied per iteration | ||
| 91 : | shl edx, 6 | ||
| 92 : | sub ebp, edx ; remainder -> ebp | ||
| 93 : | shr ebp, 4 ; 16 bytes per iteration | ||
| 94 : | add ebp, 1 | ||
| 95 : | mov [remainder], ebp | ||
| 96 : | |||
| 97 : | mov edx, ecx | ||
| 98 : | |||
| 99 : | .y_inner_loop: | ||
| 100 : | prefetchnta [esi + 64] ; non temporal prefetch | ||
| 101 : | prefetchnta [esi + 96] | ||
| 102 : | |||
| 103 : | movq mm1, [esi] ; read from src | ||
| 104 : | movq mm2, [esi + 8] | ||
| 105 : | movq mm3, [esi + 16] | ||
| 106 : | movq mm4, [esi + 24] | ||
| 107 : | movq mm5, [esi + 32] | ||
| 108 : | movq mm6, [esi + 40] | ||
| 109 : | movq mm7, [esi + 48] | ||
| 110 : | movq mm0, [esi + 56] | ||
| 111 : | |||
| 112 : | movntq [edi], mm1 ; write to y_out | ||
| 113 : | movntq [edi + 8], mm2 | ||
| 114 : | movntq [edi + 16], mm3 | ||
| 115 : | movntq [edi + 24], mm4 | ||
| 116 : | movntq [edi + 32], mm5 | ||
| 117 : | movntq [edi + 40], mm6 | ||
| 118 : | movntq [edi + 48], mm7 | ||
| 119 : | movntq [edi + 56], mm0 | ||
| 120 : | |||
| 121 : | add esi, 64 | ||
| 122 : | add edi, 64 | ||
| 123 : | dec ecx | ||
| 124 : | jnz .y_inner_loop | ||
| 125 : | |||
| 126 : | dec ebp | ||
| 127 : | jz .y_outer_loop | ||
| 128 : | |||
| 129 : | .y_remainder_loop: | ||
| 130 : | movq mm1, [esi] ; read from src | ||
| 131 : | movq mm2, [esi + 8] | ||
| 132 : | |||
| 133 : | movntq [edi], mm1 ; write to y_out | ||
| 134 : | movntq [edi + 8], mm2 | ||
| 135 : | |||
| 136 : | add esi, 16 | ||
| 137 : | add edi, 16 | ||
| 138 : | dec ebp | ||
| 139 : | jnz .y_remainder_loop | ||
| 140 : | |||
| 141 : | .y_outer_loop: | ||
| 142 : | mov ebp, [remainder] | ||
| 143 : | mov ecx, edx | ||
| 144 : | add edi, ebx | ||
| 145 : | |||
| 146 : | dec eax | ||
| 147 : | jnz near .y_inner_loop | ||
| 148 : | |||
| 149 : | mov eax, [esp + 40 + localsize] ; height -> eax | ||
| 150 : | mov ebx, [esp + 44 + localsize] ; stride -> ebx | ||
| 151 : | mov ecx, [esp + 36 + localsize] ; width -> ecx | ||
| 152 : | mov edi, [esp + 24 + localsize] ; u_out -> edi | ||
| 153 : | |||
| 154 : | shr ecx, 1 ; width / 2 -> ecx | ||
| 155 : | shr ebx, 1 ; stride / 2 -> ebx | ||
| 156 : | shr eax, 1 ; height / 2 -> eax | ||
| 157 : | |||
| 158 : | sub ebx, ecx ; stride / 2 - width / 2 -> ebx | ||
| 159 : | |||
| 160 : | mov edx, ecx | ||
| 161 : | mov ebp, ecx | ||
| 162 : | shr edx, 6 | ||
| 163 : | mov ecx, edx ; 64 bytes copied per iteration | ||
| 164 : | shl edx, 6 | ||
| 165 : | sub ebp, edx ; remainder -> ebp | ||
| 166 : | shr ebp, 3 ; 8 bytes per iteration | ||
| 167 : | add ebp, 1 | ||
| 168 : | mov [remainder], ebp | ||
| 169 : | |||
| 170 : | mov edx, ecx | ||
| 171 : | |||
| 172 : | .u_inner_loop: | ||
| 173 : | prefetchnta [esi + 64] ; non temporal prefetch | ||
| 174 : | prefetchnta [esi + 96] | ||
| 175 : | |||
| 176 : | movq mm1, [esi] ; read from src | ||
| 177 : | movq mm2, [esi + 8] | ||
| 178 : | movq mm3, [esi + 16] | ||
| 179 : | movq mm4, [esi + 24] | ||
| 180 : | movq mm5, [esi + 32] | ||
| 181 : | movq mm6, [esi + 40] | ||
| 182 : | movq mm7, [esi + 48] | ||
| 183 : | movq mm0, [esi + 56] | ||
| 184 : | |||
| 185 : | movntq [edi], mm1 ; write to u_out | ||
| 186 : | movntq [edi + 8], mm2 | ||
| 187 : | movntq [edi + 16], mm3 | ||
| 188 : | movntq [edi + 24], mm4 | ||
| 189 : | movntq [edi + 32], mm5 | ||
| 190 : | movntq [edi + 40], mm6 | ||
| 191 : | movntq [edi + 48], mm7 | ||
| 192 : | movntq [edi + 56], mm0 | ||
| 193 : | |||
| 194 : | |||
| 195 : | add esi, 64 | ||
| 196 : | add edi, 64 | ||
| 197 : | dec ecx | ||
| 198 : | jnz .u_inner_loop | ||
| 199 : | |||
| 200 : | dec ebp | ||
| 201 : | jz .u_outer_loop | ||
| 202 : | |||
| 203 : | .u_remainder_loop: | ||
| 204 : | movq mm1, [esi] ; read from src | ||
| 205 : | movntq [edi], mm1 ; write to y_out | ||
| 206 : | |||
| 207 : | add esi, 8 | ||
| 208 : | add edi, 8 | ||
| 209 : | dec ebp | ||
| 210 : | jnz .u_remainder_loop | ||
| 211 : | |||
| 212 : | .u_outer_loop: | ||
| 213 : | mov ebp, [remainder] | ||
| 214 : | mov ecx, edx | ||
| 215 : | add edi, ebx | ||
| 216 : | |||
| 217 : | dec eax | ||
| 218 : | jnz .u_inner_loop | ||
| 219 : | |||
| 220 : | mov eax, [esp + 40 + localsize] ; height -> eax | ||
| 221 : | mov ecx, [esp + 36 + localsize] ; width -> ecx | ||
| 222 : | mov edi, [esp + 28 + localsize] ; v_out -> edi | ||
| 223 : | |||
| 224 : | shr ecx, 1 ; width / 2 -> ecx | ||
| 225 : | shr eax, 1 ; height / 2 -> eax | ||
| 226 : | |||
| 227 : | mov edx, ecx | ||
| 228 : | mov ebp, ecx | ||
| 229 : | shr edx, 6 | ||
| 230 : | mov ecx, edx ; 64 bytes copied per iteration | ||
| 231 : | shl edx, 6 | ||
| 232 : | sub ebp, edx ; remainder -> ebp | ||
| 233 : | shr ebp, 3 ; 8 bytes per iteration | ||
| 234 : | add ebp, 1 | ||
| 235 : | mov [remainder], ebp | ||
| 236 : | |||
| 237 : | mov edx, ecx | ||
| 238 : | |||
| 239 : | .v_inner_loop: | ||
| 240 : | prefetchnta [esi + 64] ; non temporal prefetch | ||
| 241 : | prefetchnta [esi + 96] | ||
| 242 : | |||
| 243 : | movq mm1, [esi] ; read from src | ||
| 244 : | movq mm2, [esi + 8] | ||
| 245 : | movq mm3, [esi + 16] | ||
| 246 : | movq mm4, [esi + 24] | ||
| 247 : | movq mm5, [esi + 32] | ||
| 248 : | movq mm6, [esi + 40] | ||
| 249 : | movq mm7, [esi + 48] | ||
| 250 : | movq mm0, [esi + 56] | ||
| 251 : | |||
| 252 : | movntq [edi], mm1 ; write to u_out | ||
| 253 : | movntq [edi + 8], mm2 | ||
| 254 : | movntq [edi + 16], mm3 | ||
| 255 : | movntq [edi + 24], mm4 | ||
| 256 : | movntq [edi + 32], mm5 | ||
| 257 : | movntq [edi + 40], mm6 | ||
| 258 : | movntq [edi + 48], mm7 | ||
| 259 : | movntq [edi + 56], mm0 | ||
| 260 : | |||
| 261 : | |||
| 262 : | add esi, 64 | ||
| 263 : | add edi, 64 | ||
| 264 : | dec ecx | ||
| 265 : | jnz .v_inner_loop | ||
| 266 : | |||
| 267 : | dec ebp | ||
| 268 : | jz .v_outer_loop | ||
| 269 : | |||
| 270 : | .v_remainder_loop: | ||
| 271 : | movq mm1, [esi] ; read from src | ||
| 272 : | movntq [edi], mm1 ; write to y_out | ||
| 273 : | |||
| 274 : | add esi, 8 | ||
| 275 : | add edi, 8 | ||
| 276 : | dec ebp | ||
| 277 : | jnz .v_remainder_loop | ||
| 278 : | |||
| 279 : | .v_outer_loop: | ||
| 280 : | mov ebp, [remainder] | ||
| 281 : | mov ecx, edx | ||
| 282 : | add edi, ebx | ||
| 283 : | |||
| 284 : | dec eax | ||
| 285 : | jnz .v_inner_loop | ||
| 286 : | |||
| 287 : | ; local vars deallocation | ||
| 288 : | add esp, localsize | ||
| 289 : | %undef localsize | ||
| 290 : | %undef remainder | ||
| 291 : | |||
| 292 : | pop ebp | ||
| 293 : | pop edi | ||
| 294 : | pop esi | ||
| 295 : | pop ebx | ||
| 296 : | |||
| 297 : | emms | ||
| 298 : | |||
| 299 : | ret | ||
| 300 : | |||
| 301 : | |||
| 302 : | |||
| 303 : | ;------------------------------------------------------------------------------ | ||
| 304 : | ; | ||
| 305 : | ; void yuv_to_yv12_mmx(uint8_t *y_out, | ||
| 306 : | ; uint8_t *u_out, | ||
| 307 : | ; uint8_t *v_out, | ||
| 308 : | ; uint8_t *src, | ||
| 309 : | ; int width, int height, int stride); | ||
| 310 : | ; | ||
| 311 : | ; Attention: This code assumes that width is a multiple of 16 | ||
| 312 : | ; | ||
| 313 : | ;------------------------------------------------------------------------------ | ||
| 314 : | |||
| 315 : | cglobal yuv_to_yv12_mmx | ||
| 316 : | yuv_to_yv12_mmx: | ||
| 317 : | |||
| 318 : | push ebx | ||
| 319 : | push esi | ||
| 320 : | push edi | ||
| 321 : | push ebp | ||
| 322 : | |||
| 323 : | ; local vars allocation | ||
| 324 : | %define localsize 4 | ||
| 325 : | %define remainder esp | ||
| 326 : | sub esp, localsize | ||
| 327 : | |||
| 328 : | |||
| 329 : | ; function code | ||
| 330 : | mov eax, [esp + 40 + localsize] ; height -> eax | ||
| 331 : | mov ebx, [esp + 44 + localsize] ; stride -> ebx | ||
| 332 : | mov esi, [esp + 32 + localsize] ; src -> esi | ||
| 333 : | mov edi, [esp + 20 + localsize] ; y_out -> edi | ||
| 334 : | edgomez | 331 | mov ecx, [esp + 36 + localsize] ; width -> ecx |
| 335 : | edgomez | 328 | |
| 336 : | sub ebx, ecx ; stride - width -> ebx | ||
| 337 : | |||
| 338 : | mov edx, ecx | ||
| 339 : | mov ebp, ecx | ||
| 340 : | shr edx, 6 | ||
| 341 : | mov ecx, edx ; 64 bytes copied per iteration | ||
| 342 : | shl edx, 6 | ||
| 343 : | sub ebp, edx ; mainder -> ebp | ||
| 344 : | shr ebp, 4 ; 16 bytes per iteration | ||
| 345 : | add ebp, 1 | ||
| 346 : | mov [remainder], ebp | ||
| 347 : | |||
| 348 : | mov edx, ecx | ||
| 349 : | |||
| 350 : | .y_inner_loop: | ||
| 351 : | movq mm1, [esi] ; read from src | ||
| 352 : | movq mm2, [esi + 8] | ||
| 353 : | movq mm3, [esi + 16] | ||
| 354 : | movq mm4, [esi + 24] | ||
| 355 : | movq mm5, [esi + 32] | ||
| 356 : | movq mm6, [esi + 40] | ||
| 357 : | movq mm7, [esi + 48] | ||
| 358 : | movq mm0, [esi + 56] | ||
| 359 : | |||
| 360 : | movq [edi], mm1 ; write to y_out | ||
| 361 : | movq [edi + 8], mm2 | ||
| 362 : | movq [edi + 16], mm3 | ||
| 363 : | movq [edi + 24], mm4 | ||
| 364 : | movq [edi + 32], mm5 | ||
| 365 : | movq [edi + 40], mm6 | ||
| 366 : | movq [edi + 48], mm7 | ||
| 367 : | movq [edi + 56], mm0 | ||
| 368 : | |||
| 369 : | add esi, 64 | ||
| 370 : | add edi, 64 | ||
| 371 : | dec ecx | ||
| 372 : | jnz .y_inner_loop | ||
| 373 : | |||
| 374 : | dec ebp | ||
| 375 : | jz .y_outer_loop | ||
| 376 : | |||
| 377 : | .y_remainder_loop: | ||
| 378 : | movq mm1, [esi] ; read from src | ||
| 379 : | movq mm2, [esi + 8] | ||
| 380 : | |||
| 381 : | movq [edi], mm1 ; write to y_out | ||
| 382 : | movq [edi + 8], mm2 | ||
| 383 : | |||
| 384 : | add esi, 16 | ||
| 385 : | add edi, 16 | ||
| 386 : | dec ebp | ||
| 387 : | jnz .y_remainder_loop | ||
| 388 : | |||
| 389 : | .y_outer_loop: | ||
| 390 : | mov ebp, [remainder] | ||
| 391 : | mov ecx, edx | ||
| 392 : | add edi, ebx | ||
| 393 : | |||
| 394 : | dec eax | ||
| 395 : | jnz near .y_inner_loop | ||
| 396 : | |||
| 397 : | mov eax, [esp + 40 + localsize] ; height -> eax | ||
| 398 : | mov ebx, [esp + 44 + localsize] ; stride -> ebx | ||
| 399 : | mov ecx, [esp + 36 + localsize] ; width -> ecx | ||
| 400 : | mov edi, [esp + 24 + localsize] ; u_out -> edi | ||
| 401 : | |||
| 402 : | shr ecx, 1 ; width / 2 -> ecx | ||
| 403 : | shr ebx, 1 ; stride / 2 -> ebx | ||
| 404 : | shr eax, 1 ; height / 2 -> eax | ||
| 405 : | |||
| 406 : | sub ebx, ecx ; stride / 2 - width / 2 -> ebx | ||
| 407 : | |||
| 408 : | mov edx, ecx | ||
| 409 : | mov ebp, ecx | ||
| 410 : | shr edx, 6 | ||
| 411 : | mov ecx, edx ; 64 bytes copied per iteration | ||
| 412 : | shl edx, 6 | ||
| 413 : | sub ebp, edx ; remainder -> ebp | ||
| 414 : | shr ebp, 3 ; 8 bytes per iteration | ||
| 415 : | add ebp, 1 | ||
| 416 : | mov [remainder], ebp | ||
| 417 : | |||
| 418 : | mov edx, ecx | ||
| 419 : | |||
| 420 : | .u_inner_loop: | ||
| 421 : | movq mm1, [esi] ; read from src | ||
| 422 : | movq mm2, [esi + 8] | ||
| 423 : | movq mm3, [esi + 16] | ||
| 424 : | movq mm4, [esi + 24] | ||
| 425 : | movq mm5, [esi + 32] | ||
| 426 : | movq mm6, [esi + 40] | ||
| 427 : | movq mm7, [esi + 48] | ||
| 428 : | movq mm0, [esi + 56] | ||
| 429 : | |||
| 430 : | movq [edi], mm1 ; write to u_out | ||
| 431 : | movq [edi + 8], mm2 | ||
| 432 : | movq [edi + 16], mm3 | ||
| 433 : | movq [edi + 24], mm4 | ||
| 434 : | movq [edi + 32], mm5 | ||
| 435 : | movq [edi + 40], mm6 | ||
| 436 : | movq [edi + 48], mm7 | ||
| 437 : | movq [edi + 56], mm0 | ||
| 438 : | |||
| 439 : | |||
| 440 : | add esi, 64 | ||
| 441 : | add edi, 64 | ||
| 442 : | dec ecx | ||
| 443 : | jnz .u_inner_loop | ||
| 444 : | |||
| 445 : | dec ebp | ||
| 446 : | jz .u_outer_loop | ||
| 447 : | |||
| 448 : | .u_remainder_loop: | ||
| 449 : | movq mm1, [esi] ; read from src | ||
| 450 : | movq [edi], mm1 ; write to y_out | ||
| 451 : | |||
| 452 : | add esi, 8 | ||
| 453 : | add edi, 8 | ||
| 454 : | dec ebp | ||
| 455 : | jnz .u_remainder_loop | ||
| 456 : | |||
| 457 : | .u_outer_loop: | ||
| 458 : | mov ebp, [remainder] | ||
| 459 : | mov ecx, edx | ||
| 460 : | add edi, ebx | ||
| 461 : | |||
| 462 : | dec eax | ||
| 463 : | jnz .u_inner_loop | ||
| 464 : | |||
| 465 : | mov eax, [esp + 40 + localsize] ; height -> eax | ||
| 466 : | mov ecx, [esp + 36 + localsize] ; width -> ecx | ||
| 467 : | mov edi, [esp + 28 + localsize] ; v_out -> edi | ||
| 468 : | |||
| 469 : | shr ecx, 1 ; width / 2 -> ecx | ||
| 470 : | shr eax, 1 ; height / 2 -> eax | ||
| 471 : | |||
| 472 : | mov edx, ecx | ||
| 473 : | mov ebp, ecx | ||
| 474 : | shr edx, 6 | ||
| 475 : | mov ecx, edx ; 64 bytes copied per iteration | ||
| 476 : | shl edx, 6 | ||
| 477 : | sub ebp, edx ; remainder -> ebp | ||
| 478 : | shr ebp, 3 ; 8 bytes per iteration | ||
| 479 : | add ebp, 1 | ||
| 480 : | mov [remainder], ebp | ||
| 481 : | |||
| 482 : | mov edx, ecx | ||
| 483 : | |||
| 484 : | .v_inner_loop: | ||
| 485 : | movq mm1, [esi] ; read from src | ||
| 486 : | movq mm2, [esi + 8] | ||
| 487 : | movq mm3, [esi + 16] | ||
| 488 : | movq mm4, [esi + 24] | ||
| 489 : | movq mm5, [esi + 32] | ||
| 490 : | movq mm6, [esi + 40] | ||
| 491 : | movq mm7, [esi + 48] | ||
| 492 : | movq mm0, [esi + 56] | ||
| 493 : | |||
| 494 : | movq [edi], mm1 ; write to u_out | ||
| 495 : | movq [edi + 8], mm2 | ||
| 496 : | movq [edi + 16], mm3 | ||
| 497 : | movq [edi + 24], mm4 | ||
| 498 : | movq [edi + 32], mm5 | ||
| 499 : | movq [edi + 40], mm6 | ||
| 500 : | movq [edi + 48], mm7 | ||
| 501 : | movq [edi + 56], mm0 | ||
| 502 : | |||
| 503 : | |||
| 504 : | add esi, 64 | ||
| 505 : | add edi, 64 | ||
| 506 : | dec ecx | ||
| 507 : | jnz .v_inner_loop | ||
| 508 : | |||
| 509 : | dec ebp | ||
| 510 : | jz .v_outer_loop | ||
| 511 : | |||
| 512 : | .v_remainder_loop: | ||
| 513 : | movq mm1, [esi] ; read from src | ||
| 514 : | movq [edi], mm1 ; write to y_out | ||
| 515 : | |||
| 516 : | add esi, 8 | ||
| 517 : | add edi, 8 | ||
| 518 : | dec ebp | ||
| 519 : | jnz .v_remainder_loop | ||
| 520 : | |||
| 521 : | .v_outer_loop: | ||
| 522 : | mov ebp, [remainder] | ||
| 523 : | mov ecx, edx | ||
| 524 : | add edi, ebx | ||
| 525 : | |||
| 526 : | dec eax | ||
| 527 : | jnz .v_inner_loop | ||
| 528 : | |||
| 529 : | ; local vars deallocation | ||
| 530 : | add esp, localsize | ||
| 531 : | %undef localsize | ||
| 532 : | %undef remainder | ||
| 533 : | |||
| 534 : | pop ebp | ||
| 535 : | pop edi | ||
| 536 : | pop esi | ||
| 537 : | pop ebx | ||
| 538 : | |||
| 539 : | emms | ||
| 540 : | |||
| 541 : | ret |
| No admin address has been configured | ViewVC Help |
| Powered by ViewVC 1.0.4 |