Mercurial > libavcodec.hg

comparison arm/dsputil_arm_s.S @ 8678:6c256fc075e3 libavcodec

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
ARM: change alignment of loops in put_pixels*_arm to 32
author mru
date Tue, 27 Jan 2009 16:06:41 +0000
parents 7a463923ecd1
children 9c1ea156d893
comparison
equal deleted inserted replaced
8677:3c484b73ca73 8678:6c256fc075e3
89 add \Rd0, \Rn0, \Rd0, lsr #1 89 add \Rd0, \Rn0, \Rd0, lsr #1
90 add \Rd1, \Rn1, \Rd1, lsr #1 90 add \Rd1, \Rn1, \Rd1, lsr #1
91 .endm 91 .endm
92 92
93 @ ---------------------------------------------------------------- 93 @ ----------------------------------------------------------------
94 .align 8 94 .align 5
95 function put_pixels16_arm, export=1 95 function put_pixels16_arm, export=1
96 @ void func(uint8_t *block, const uint8_t *pixels, int line_size, int h) 96 @ void func(uint8_t *block, const uint8_t *pixels, int line_size, int h)
97 @ block = word aligned, pixles = unaligned 97 @ block = word aligned, pixles = unaligned
98 pld [r1] 98 pld [r1]
99 stmfd sp!, {r4-r11, lr} @ R14 is also called LR 99 stmfd sp!, {r4-r11, lr} @ R14 is also called LR
109 pld [r1] 109 pld [r1]
110 subs r3, r3, #1 110 subs r3, r3, #1
111 add r0, r0, r2 111 add r0, r0, r2
112 bne 1b 112 bne 1b
113 ldmfd sp!, {r4-r11, pc} 113 ldmfd sp!, {r4-r11, pc}
114 .align 8 114 .align 5
115 2: 115 2:
116 ldmia r1, {r4-r8} 116 ldmia r1, {r4-r8}
117 add r1, r1, r2 117 add r1, r1, r2
118 ADJ_ALIGN_QUADWORD_D 1, r9, r10, r11, r12, r4, r5, r6, r7, r8 118 ADJ_ALIGN_QUADWORD_D 1, r9, r10, r11, r12, r4, r5, r6, r7, r8
119 pld [r1] 119 pld [r1]
120 subs r3, r3, #1 120 subs r3, r3, #1
121 stmia r0, {r9-r12} 121 stmia r0, {r9-r12}
122 add r0, r0, r2 122 add r0, r0, r2
123 bne 2b 123 bne 2b
124 ldmfd sp!, {r4-r11, pc} 124 ldmfd sp!, {r4-r11, pc}
125 .align 8 125 .align 5
126 3: 126 3:
127 ldmia r1, {r4-r8} 127 ldmia r1, {r4-r8}
128 add r1, r1, r2 128 add r1, r1, r2
129 ADJ_ALIGN_QUADWORD_D 2, r9, r10, r11, r12, r4, r5, r6, r7, r8 129 ADJ_ALIGN_QUADWORD_D 2, r9, r10, r11, r12, r4, r5, r6, r7, r8
130 pld [r1] 130 pld [r1]
131 subs r3, r3, #1 131 subs r3, r3, #1
132 stmia r0, {r9-r12} 132 stmia r0, {r9-r12}
133 add r0, r0, r2 133 add r0, r0, r2
134 bne 3b 134 bne 3b
135 ldmfd sp!, {r4-r11, pc} 135 ldmfd sp!, {r4-r11, pc}
136 .align 8 136 .align 5
137 4: 137 4:
138 ldmia r1, {r4-r8} 138 ldmia r1, {r4-r8}
139 add r1, r1, r2 139 add r1, r1, r2
140 ADJ_ALIGN_QUADWORD_D 3, r9, r10, r11, r12, r4, r5, r6, r7, r8 140 ADJ_ALIGN_QUADWORD_D 3, r9, r10, r11, r12, r4, r5, r6, r7, r8
141 pld [r1] 141 pld [r1]
142 subs r3, r3, #1 142 subs r3, r3, #1
143 stmia r0, {r9-r12} 143 stmia r0, {r9-r12}
144 add r0, r0, r2 144 add r0, r0, r2
145 bne 4b 145 bne 4b
146 ldmfd sp!, {r4-r11,pc} 146 ldmfd sp!, {r4-r11,pc}
147 .align 8
148 5: 147 5:
149 .word 1b 148 .word 1b
150 .word 2b 149 .word 2b
151 .word 3b 150 .word 3b
152 .word 4b 151 .word 4b
153 .endfunc 152 .endfunc
154 153
155 @ ---------------------------------------------------------------- 154 @ ----------------------------------------------------------------
156 .align 8 155 .align 5
157 function put_pixels8_arm, export=1 156 function put_pixels8_arm, export=1
158 @ void func(uint8_t *block, const uint8_t *pixels, int line_size, int h) 157 @ void func(uint8_t *block, const uint8_t *pixels, int line_size, int h)
159 @ block = word aligned, pixles = unaligned 158 @ block = word aligned, pixles = unaligned
160 pld [r1] 159 pld [r1]
161 stmfd sp!, {r4-r5,lr} @ R14 is also called LR 160 stmfd sp!, {r4-r5,lr} @ R14 is also called LR
171 pld [r1] 170 pld [r1]
172 stmia r0, {r4-r5} 171 stmia r0, {r4-r5}
173 add r0, r0, r2 172 add r0, r0, r2
174 bne 1b 173 bne 1b
175 ldmfd sp!, {r4-r5,pc} 174 ldmfd sp!, {r4-r5,pc}
176 .align 8 175 .align 5
177 2: 176 2:
178 ldmia r1, {r4-r5, r12} 177 ldmia r1, {r4-r5, r12}
179 add r1, r1, r2 178 add r1, r1, r2
180 ADJ_ALIGN_DOUBLEWORD 1, r4, r5, r12 179 ADJ_ALIGN_DOUBLEWORD 1, r4, r5, r12
181 pld [r1] 180 pld [r1]
182 subs r3, r3, #1 181 subs r3, r3, #1
183 stmia r0, {r4-r5} 182 stmia r0, {r4-r5}
184 add r0, r0, r2 183 add r0, r0, r2
185 bne 2b 184 bne 2b
186 ldmfd sp!, {r4-r5,pc} 185 ldmfd sp!, {r4-r5,pc}
187 .align 8 186 .align 5
188 3: 187 3:
189 ldmia r1, {r4-r5, r12} 188 ldmia r1, {r4-r5, r12}
190 add r1, r1, r2 189 add r1, r1, r2
191 ADJ_ALIGN_DOUBLEWORD 2, r4, r5, r12 190 ADJ_ALIGN_DOUBLEWORD 2, r4, r5, r12
192 pld [r1] 191 pld [r1]
193 subs r3, r3, #1 192 subs r3, r3, #1
194 stmia r0, {r4-r5} 193 stmia r0, {r4-r5}
195 add r0, r0, r2 194 add r0, r0, r2
196 bne 3b 195 bne 3b
197 ldmfd sp!, {r4-r5,pc} 196 ldmfd sp!, {r4-r5,pc}
198 .align 8 197 .align 5
199 4: 198 4:
200 ldmia r1, {r4-r5, r12} 199 ldmia r1, {r4-r5, r12}
201 add r1, r1, r2 200 add r1, r1, r2
202 ADJ_ALIGN_DOUBLEWORD 3, r4, r5, r12 201 ADJ_ALIGN_DOUBLEWORD 3, r4, r5, r12
203 pld [r1] 202 pld [r1]
204 subs r3, r3, #1 203 subs r3, r3, #1
205 stmia r0, {r4-r5} 204 stmia r0, {r4-r5}
206 add r0, r0, r2 205 add r0, r0, r2
207 bne 4b 206 bne 4b
208 ldmfd sp!, {r4-r5,pc} 207 ldmfd sp!, {r4-r5,pc}
209 .align 8
210 5: 208 5:
211 .word 1b 209 .word 1b
212 .word 2b 210 .word 2b
213 .word 3b 211 .word 3b
214 .word 4b 212 .word 4b
215 .endfunc 213 .endfunc
216 214
217 @ ---------------------------------------------------------------- 215 @ ----------------------------------------------------------------
218 .align 8 216 .align 5
219 function put_pixels8_x2_arm, export=1 217 function put_pixels8_x2_arm, export=1
220 @ void func(uint8_t *block, const uint8_t *pixels, int line_size, int h) 218 @ void func(uint8_t *block, const uint8_t *pixels, int line_size, int h)
221 @ block = word aligned, pixles = unaligned 219 @ block = word aligned, pixles = unaligned
222 pld [r1] 220 pld [r1]
223 stmfd sp!, {r4-r10,lr} @ R14 is also called LR 221 stmfd sp!, {r4-r10,lr} @ R14 is also called LR
236 subs r3, r3, #1 234 subs r3, r3, #1
237 stmia r0, {r8-r9} 235 stmia r0, {r8-r9}
238 add r0, r0, r2 236 add r0, r0, r2
239 bne 1b 237 bne 1b
240 ldmfd sp!, {r4-r10,pc} 238 ldmfd sp!, {r4-r10,pc}
241 .align 8 239 .align 5
242 2: 240 2:
243 ldmia r1, {r4-r5, r10} 241 ldmia r1, {r4-r5, r10}
244 add r1, r1, r2 242 add r1, r1, r2
245 ADJ_ALIGN_DOUBLEWORD_D 1, r6, r7, r4, r5, r10 243 ADJ_ALIGN_DOUBLEWORD_D 1, r6, r7, r4, r5, r10
246 ADJ_ALIGN_DOUBLEWORD_D 2, r8, r9, r4, r5, r10 244 ADJ_ALIGN_DOUBLEWORD_D 2, r8, r9, r4, r5, r10
249 subs r3, r3, #1 247 subs r3, r3, #1
250 stmia r0, {r4-r5} 248 stmia r0, {r4-r5}
251 add r0, r0, r2 249 add r0, r0, r2
252 bne 2b 250 bne 2b
253 ldmfd sp!, {r4-r10,pc} 251 ldmfd sp!, {r4-r10,pc}
254 .align 8 252 .align 5
255 3: 253 3:
256 ldmia r1, {r4-r5, r10} 254 ldmia r1, {r4-r5, r10}
257 add r1, r1, r2 255 add r1, r1, r2
258 ADJ_ALIGN_DOUBLEWORD_D 2, r6, r7, r4, r5, r10 256 ADJ_ALIGN_DOUBLEWORD_D 2, r6, r7, r4, r5, r10
259 ADJ_ALIGN_DOUBLEWORD_D 3, r8, r9, r4, r5, r10 257 ADJ_ALIGN_DOUBLEWORD_D 3, r8, r9, r4, r5, r10
262 subs r3, r3, #1 260 subs r3, r3, #1
263 stmia r0, {r4-r5} 261 stmia r0, {r4-r5}
264 add r0, r0, r2 262 add r0, r0, r2
265 bne 3b 263 bne 3b
266 ldmfd sp!, {r4-r10,pc} 264 ldmfd sp!, {r4-r10,pc}
267 .align 8 265 .align 5
268 4: 266 4:
269 ldmia r1, {r4-r5, r10} 267 ldmia r1, {r4-r5, r10}
270 add r1, r1, r2 268 add r1, r1, r2
271 ADJ_ALIGN_DOUBLEWORD_D 3, r6, r7, r4, r5, r10 269 ADJ_ALIGN_DOUBLEWORD_D 3, r6, r7, r4, r5, r10
272 pld [r1] 270 pld [r1]
274 subs r3, r3, #1 272 subs r3, r3, #1
275 stmia r0, {r8-r9} 273 stmia r0, {r8-r9}
276 add r0, r0, r2 274 add r0, r0, r2
277 bne 4b 275 bne 4b
278 ldmfd sp!, {r4-r10,pc} @@ update PC with LR content. 276 ldmfd sp!, {r4-r10,pc} @@ update PC with LR content.
279 .align 8
280 5: 277 5:
281 .word 0xFEFEFEFE 278 .word 0xFEFEFEFE
282 .word 2b 279 .word 2b
283 .word 3b 280 .word 3b
284 .word 4b 281 .word 4b
285 .endfunc 282 .endfunc
286 283
287 .align 8 284 .align 5
288 function put_no_rnd_pixels8_x2_arm, export=1 285 function put_no_rnd_pixels8_x2_arm, export=1
289 @ void func(uint8_t *block, const uint8_t *pixels, int line_size, int h) 286 @ void func(uint8_t *block, const uint8_t *pixels, int line_size, int h)
290 @ block = word aligned, pixles = unaligned 287 @ block = word aligned, pixles = unaligned
291 pld [r1] 288 pld [r1]
292 stmfd sp!, {r4-r10,lr} @ R14 is also called LR 289 stmfd sp!, {r4-r10,lr} @ R14 is also called LR
305 subs r3, r3, #1 302 subs r3, r3, #1
306 stmia r0, {r8-r9} 303 stmia r0, {r8-r9}
307 add r0, r0, r2 304 add r0, r0, r2
308 bne 1b 305 bne 1b
309 ldmfd sp!, {r4-r10,pc} 306 ldmfd sp!, {r4-r10,pc}
310 .align 8 307 .align 5
311 2: 308 2:
312 ldmia r1, {r4-r5, r10} 309 ldmia r1, {r4-r5, r10}
313 add r1, r1, r2 310 add r1, r1, r2
314 ADJ_ALIGN_DOUBLEWORD_D 1, r6, r7, r4, r5, r10 311 ADJ_ALIGN_DOUBLEWORD_D 1, r6, r7, r4, r5, r10
315 ADJ_ALIGN_DOUBLEWORD_D 2, r8, r9, r4, r5, r10 312 ADJ_ALIGN_DOUBLEWORD_D 2, r8, r9, r4, r5, r10
318 subs r3, r3, #1 315 subs r3, r3, #1
319 stmia r0, {r4-r5} 316 stmia r0, {r4-r5}
320 add r0, r0, r2 317 add r0, r0, r2
321 bne 2b 318 bne 2b
322 ldmfd sp!, {r4-r10,pc} 319 ldmfd sp!, {r4-r10,pc}
323 .align 8 320 .align 5
324 3: 321 3:
325 ldmia r1, {r4-r5, r10} 322 ldmia r1, {r4-r5, r10}
326 add r1, r1, r2 323 add r1, r1, r2
327 ADJ_ALIGN_DOUBLEWORD_D 2, r6, r7, r4, r5, r10 324 ADJ_ALIGN_DOUBLEWORD_D 2, r6, r7, r4, r5, r10
328 ADJ_ALIGN_DOUBLEWORD_D 3, r8, r9, r4, r5, r10 325 ADJ_ALIGN_DOUBLEWORD_D 3, r8, r9, r4, r5, r10
331 subs r3, r3, #1 328 subs r3, r3, #1
332 stmia r0, {r4-r5} 329 stmia r0, {r4-r5}
333 add r0, r0, r2 330 add r0, r0, r2
334 bne 3b 331 bne 3b
335 ldmfd sp!, {r4-r10,pc} 332 ldmfd sp!, {r4-r10,pc}
336 .align 8 333 .align 5
337 4: 334 4:
338 ldmia r1, {r4-r5, r10} 335 ldmia r1, {r4-r5, r10}
339 add r1, r1, r2 336 add r1, r1, r2
340 ADJ_ALIGN_DOUBLEWORD_D 3, r6, r7, r4, r5, r10 337 ADJ_ALIGN_DOUBLEWORD_D 3, r6, r7, r4, r5, r10
341 pld [r1] 338 pld [r1]
343 subs r3, r3, #1 340 subs r3, r3, #1
344 stmia r0, {r8-r9} 341 stmia r0, {r8-r9}
345 add r0, r0, r2 342 add r0, r0, r2
346 bne 4b 343 bne 4b
347 ldmfd sp!, {r4-r10,pc} @@ update PC with LR content. 344 ldmfd sp!, {r4-r10,pc} @@ update PC with LR content.
348 .align 8
349 5: 345 5:
350 .word 0xFEFEFEFE 346 .word 0xFEFEFEFE
351 .word 2b 347 .word 2b
352 .word 3b 348 .word 3b
353 .word 4b 349 .word 4b
354 .endfunc 350 .endfunc
355 351
356 352
357 @ ---------------------------------------------------------------- 353 @ ----------------------------------------------------------------
358 .align 8 354 .align 5
359 function put_pixels8_y2_arm, export=1 355 function put_pixels8_y2_arm, export=1
360 @ void func(uint8_t *block, const uint8_t *pixels, int line_size, int h) 356 @ void func(uint8_t *block, const uint8_t *pixels, int line_size, int h)
361 @ block = word aligned, pixles = unaligned 357 @ block = word aligned, pixles = unaligned
362 pld [r1] 358 pld [r1]
363 stmfd sp!, {r4-r11,lr} @ R14 is also called LR 359 stmfd sp!, {r4-r11,lr} @ R14 is also called LR
384 subs r3, r3, #1 380 subs r3, r3, #1
385 stmia r0, {r8-r9} 381 stmia r0, {r8-r9}
386 add r0, r0, r2 382 add r0, r0, r2
387 bne 6b 383 bne 6b
388 ldmfd sp!, {r4-r11,pc} 384 ldmfd sp!, {r4-r11,pc}
389 .align 8 385 .align 5
390 2: 386 2:
391 ldmia r1, {r4-r6} 387 ldmia r1, {r4-r6}
392 add r1, r1, r2 388 add r1, r1, r2
393 pld [r1] 389 pld [r1]
394 ADJ_ALIGN_DOUBLEWORD 1, r4, r5, r6 390 ADJ_ALIGN_DOUBLEWORD 1, r4, r5, r6
407 RND_AVG32 r10, r11, r7, r8, r4, r5, r12 403 RND_AVG32 r10, r11, r7, r8, r4, r5, r12
408 stmia r0, {r10-r11} 404 stmia r0, {r10-r11}
409 add r0, r0, r2 405 add r0, r0, r2
410 bne 6b 406 bne 6b
411 ldmfd sp!, {r4-r11,pc} 407 ldmfd sp!, {r4-r11,pc}
412 .align 8 408 .align 5
413 3: 409 3:
414 ldmia r1, {r4-r6} 410 ldmia r1, {r4-r6}
415 add r1, r1, r2 411 add r1, r1, r2
416 pld [r1] 412 pld [r1]
417 ADJ_ALIGN_DOUBLEWORD 2, r4, r5, r6 413 ADJ_ALIGN_DOUBLEWORD 2, r4, r5, r6
430 RND_AVG32 r10, r11, r7, r8, r4, r5, r12 426 RND_AVG32 r10, r11, r7, r8, r4, r5, r12
431 stmia r0, {r10-r11} 427 stmia r0, {r10-r11}
432 add r0, r0, r2 428 add r0, r0, r2
433 bne 6b 429 bne 6b
434 ldmfd sp!, {r4-r11,pc} 430 ldmfd sp!, {r4-r11,pc}
435 .align 8 431 .align 5
436 4: 432 4:
437 ldmia r1, {r4-r6} 433 ldmia r1, {r4-r6}
438 add r1, r1, r2 434 add r1, r1, r2
439 pld [r1] 435 pld [r1]
440 ADJ_ALIGN_DOUBLEWORD 3, r4, r5, r6 436 ADJ_ALIGN_DOUBLEWORD 3, r4, r5, r6
454 stmia r0, {r10-r11} 450 stmia r0, {r10-r11}
455 add r0, r0, r2 451 add r0, r0, r2
456 bne 6b 452 bne 6b
457 ldmfd sp!, {r4-r11,pc} 453 ldmfd sp!, {r4-r11,pc}
458 454
459 .align 8
460 5: 455 5:
461 .word 0xFEFEFEFE 456 .word 0xFEFEFEFE
462 .word 2b 457 .word 2b
463 .word 3b 458 .word 3b
464 .word 4b 459 .word 4b
465 .endfunc 460 .endfunc
466 461
467 .align 8 462 .align 5
468 function put_no_rnd_pixels8_y2_arm, export=1 463 function put_no_rnd_pixels8_y2_arm, export=1
469 @ void func(uint8_t *block, const uint8_t *pixels, int line_size, int h) 464 @ void func(uint8_t *block, const uint8_t *pixels, int line_size, int h)
470 @ block = word aligned, pixles = unaligned 465 @ block = word aligned, pixles = unaligned
471 pld [r1] 466 pld [r1]
472 stmfd sp!, {r4-r11,lr} @ R14 is also called LR 467 stmfd sp!, {r4-r11,lr} @ R14 is also called LR
493 subs r3, r3, #1 488 subs r3, r3, #1
494 stmia r0, {r8-r9} 489 stmia r0, {r8-r9}
495 add r0, r0, r2 490 add r0, r0, r2
496 bne 6b 491 bne 6b
497 ldmfd sp!, {r4-r11,pc} 492 ldmfd sp!, {r4-r11,pc}
498 .align 8 493 .align 5
499 2: 494 2:
500 ldmia r1, {r4-r6} 495 ldmia r1, {r4-r6}
501 add r1, r1, r2 496 add r1, r1, r2
502 pld [r1] 497 pld [r1]
503 ADJ_ALIGN_DOUBLEWORD 1, r4, r5, r6 498 ADJ_ALIGN_DOUBLEWORD 1, r4, r5, r6
516 NO_RND_AVG32 r10, r11, r7, r8, r4, r5, r12 511 NO_RND_AVG32 r10, r11, r7, r8, r4, r5, r12
517 stmia r0, {r10-r11} 512 stmia r0, {r10-r11}
518 add r0, r0, r2 513 add r0, r0, r2
519 bne 6b 514 bne 6b
520 ldmfd sp!, {r4-r11,pc} 515 ldmfd sp!, {r4-r11,pc}
521 .align 8 516 .align 5
522 3: 517 3:
523 ldmia r1, {r4-r6} 518 ldmia r1, {r4-r6}
524 add r1, r1, r2 519 add r1, r1, r2
525 pld [r1] 520 pld [r1]
526 ADJ_ALIGN_DOUBLEWORD 2, r4, r5, r6 521 ADJ_ALIGN_DOUBLEWORD 2, r4, r5, r6
539 NO_RND_AVG32 r10, r11, r7, r8, r4, r5, r12 534 NO_RND_AVG32 r10, r11, r7, r8, r4, r5, r12
540 stmia r0, {r10-r11} 535 stmia r0, {r10-r11}
541 add r0, r0, r2 536 add r0, r0, r2
542 bne 6b 537 bne 6b
543 ldmfd sp!, {r4-r11,pc} 538 ldmfd sp!, {r4-r11,pc}
544 .align 8 539 .align 5
545 4: 540 4:
546 ldmia r1, {r4-r6} 541 ldmia r1, {r4-r6}
547 add r1, r1, r2 542 add r1, r1, r2
548 pld [r1] 543 pld [r1]
549 ADJ_ALIGN_DOUBLEWORD 3, r4, r5, r6 544 ADJ_ALIGN_DOUBLEWORD 3, r4, r5, r6
562 NO_RND_AVG32 r10, r11, r7, r8, r4, r5, r12 557 NO_RND_AVG32 r10, r11, r7, r8, r4, r5, r12
563 stmia r0, {r10-r11} 558 stmia r0, {r10-r11}
564 add r0, r0, r2 559 add r0, r0, r2
565 bne 6b 560 bne 6b
566 ldmfd sp!, {r4-r11,pc} 561 ldmfd sp!, {r4-r11,pc}
567 .align 8
568 5: 562 5:
569 .word 0xFEFEFEFE 563 .word 0xFEFEFEFE
570 .word 2b 564 .word 2b
571 .word 3b 565 .word 3b
572 .word 4b 566 .word 4b
635 add r0, r0, r2 629 add r0, r0, r2
636 bge 6b 630 bge 6b
637 ldmfd sp!, {r4-r11,pc} 631 ldmfd sp!, {r4-r11,pc}
638 .endm 632 .endm
639 633
640 .align 8 634 .align 5
641 function put_pixels8_xy2_arm, export=1 635 function put_pixels8_xy2_arm, export=1
642 @ void func(uint8_t *block, const uint8_t *pixels, int line_size, int h) 636 @ void func(uint8_t *block, const uint8_t *pixels, int line_size, int h)
643 @ block = word aligned, pixles = unaligned 637 @ block = word aligned, pixles = unaligned
644 pld [r1] 638 pld [r1]
645 stmfd sp!, {r4-r11,lr} @ R14 is also called LR 639 stmfd sp!, {r4-r11,lr} @ R14 is also called LR
649 bic r1, r1, #3 643 bic r1, r1, #3
650 ldrne pc, [r5] 644 ldrne pc, [r5]
651 1: 645 1:
652 RND_XY2_EXPAND 0 646 RND_XY2_EXPAND 0
653 647
654 .align 8 648 .align 5
655 2: 649 2:
656 RND_XY2_EXPAND 1 650 RND_XY2_EXPAND 1
657 651
658 .align 8 652 .align 5
659 3: 653 3:
660 RND_XY2_EXPAND 2 654 RND_XY2_EXPAND 2
661 655
662 .align 8 656 .align 5
663 4: 657 4:
664 RND_XY2_EXPAND 3 658 RND_XY2_EXPAND 3
665 659
666 5: 660 5:
667 .word 0x03030303 661 .word 0x03030303
671 .word 0x02020202 665 .word 0x02020202
672 .word 0xFCFCFCFC >> 2 666 .word 0xFCFCFCFC >> 2
673 .word 0x0F0F0F0F 667 .word 0x0F0F0F0F
674 .endfunc 668 .endfunc
675 669
676 .align 8 670 .align 5
677 function put_no_rnd_pixels8_xy2_arm, export=1 671 function put_no_rnd_pixels8_xy2_arm, export=1
678 @ void func(uint8_t *block, const uint8_t *pixels, int line_size, int h) 672 @ void func(uint8_t *block, const uint8_t *pixels, int line_size, int h)
679 @ block = word aligned, pixles = unaligned 673 @ block = word aligned, pixles = unaligned
680 pld [r1] 674 pld [r1]
681 stmfd sp!, {r4-r11,lr} @ R14 is also called LR 675 stmfd sp!, {r4-r11,lr} @ R14 is also called LR
685 bic r1, r1, #3 679 bic r1, r1, #3
686 ldrne pc, [r5] 680 ldrne pc, [r5]
687 1: 681 1:
688 RND_XY2_EXPAND 0 682 RND_XY2_EXPAND 0
689 683
690 .align 8 684 .align 5
691 2: 685 2:
692 RND_XY2_EXPAND 1 686 RND_XY2_EXPAND 1
693 687
694 .align 8 688 .align 5
695 3: 689 3:
696 RND_XY2_EXPAND 2 690 RND_XY2_EXPAND 2
697 691
698 .align 8 692 .align 5
699 4: 693 4:
700 RND_XY2_EXPAND 3 694 RND_XY2_EXPAND 3
701 695
702 5: 696 5:
703 .word 0x03030303 697 .word 0x03030303
707 .word 0x01010101 701 .word 0x01010101
708 .word 0xFCFCFCFC >> 2 702 .word 0xFCFCFCFC >> 2
709 .word 0x0F0F0F0F 703 .word 0x0F0F0F0F
710 .endfunc 704 .endfunc
711 705
706 .align 5
712 @ void ff_add_pixels_clamped_ARM(int16_t *block, uint8_t *dest, int stride) 707 @ void ff_add_pixels_clamped_ARM(int16_t *block, uint8_t *dest, int stride)
713 function ff_add_pixels_clamped_ARM, export=1 708 function ff_add_pixels_clamped_ARM, export=1
714 push {r4-r10} 709 push {r4-r10}
715 mov r10, #8 710 mov r10, #8
716 1: 711 1: