| blob | 8add331272b83318ba86a196fa23d1d93839eaaf |
1 /*
2 * Simple IDCT
3 *
4 * Copyright (c) 2001 Michael Niedermayer <michaelni@gmx.at>
5 * Copyright (c) 2006 Mans Rullgard <mans@mansr.com>
6 *
7 * This file is part of FFmpeg.
8 *
9 * FFmpeg is free software; you can redistribute it and/or
10 * modify it under the terms of the GNU Lesser General Public
11 * License as published by the Free Software Foundation; either
12 * version 2.1 of the License, or (at your option) any later version.
13 *
14 * FFmpeg is distributed in the hope that it will be useful,
15 * but WITHOUT ANY WARRANTY; without even the implied warranty of
16 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
17 * Lesser General Public License for more details.
18 *
19 * You should have received a copy of the GNU Lesser General Public
20 * License along with FFmpeg; if not, write to the Free Software
21 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
22 */
24 #define W1 22725 /* cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 */
25 #define W2 21407 /* cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 */
26 #define W3 19266 /* cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 */
27 #define W4 16383 /* cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 */
28 #define W5 12873 /* cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 */
29 #define W6 8867 /* cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 */
30 #define W7 4520 /* cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 */
31 #define ROW_SHIFT 11
32 #define COL_SHIFT 20
34 #define W13 (W1 | (W3 << 16))
35 #define W26 (W2 | (W6 << 16))
36 #define W57 (W5 | (W7 << 16))
38 .text
39 .align
40 w13: .long W13
41 w26: .long W26
42 w57: .long W57
44 .align
45 .type idct_row_armv5te, %function
46 .func idct_row_armv5te
47 idct_row_armv5te:
48 str lr, [sp, #-4]!
50 ldrd v1, [a1, #8]
51 ldrd a3, [a1] /* a3 = row[1:0], a4 = row[3:2] */
52 orrs v1, v1, v2
53 cmpeq v1, a4
54 cmpeq v1, a3, lsr #16
55 beq row_dc_only
57 mov v1, #(1<<(ROW_SHIFT-1))
58 mov ip, #16384
59 sub ip, ip, #1 /* ip = W4 */
60 smlabb v1, ip, a3, v1 /* v1 = W4*row[0]+(1<<(RS-1)) */
61 ldr ip, [pc, #(w26-.-8)] /* ip = W2 | (W6 << 16) */
62 smultb a2, ip, a4
63 smulbb lr, ip, a4
64 add v2, v1, a2
65 sub v3, v1, a2
66 sub v4, v1, lr
67 add v1, v1, lr
69 ldr ip, [pc, #(w13-.-8)] /* ip = W1 | (W3 << 16) */
70 ldr lr, [pc, #(w57-.-8)] /* lr = W5 | (W7 << 16) */
71 smulbt v5, ip, a3
72 smultt v6, lr, a4
73 smlatt v5, ip, a4, v5
74 smultt a2, ip, a3
75 smulbt v7, lr, a3
76 sub v6, v6, a2
77 smulbt a2, ip, a4
78 smultt fp, lr, a3
79 sub v7, v7, a2
80 smulbt a2, lr, a4
81 ldrd a3, [a1, #8] /* a3=row[5:4] a4=row[7:6] */
82 sub fp, fp, a2
84 orrs a2, a3, a4
85 beq 1f
87 smlabt v5, lr, a3, v5
88 smlabt v6, ip, a3, v6
89 smlatt v5, lr, a4, v5
90 smlabt v6, lr, a4, v6
91 smlatt v7, lr, a3, v7
92 smlatt fp, ip, a3, fp
93 smulbt a2, ip, a4
94 smlatt v7, ip, a4, v7
95 sub fp, fp, a2
97 ldr ip, [pc, #(w26-.-8)] /* ip = W2 | (W6 << 16) */
98 mov a2, #16384
99 sub a2, a2, #1 /* a2 = W4 */
100 smulbb a2, a2, a3 /* a2 = W4*row[4] */
101 smultb lr, ip, a4 /* lr = W6*row[6] */
102 add v1, v1, a2 /* v1 += W4*row[4] */
103 add v1, v1, lr /* v1 += W6*row[6] */
104 add v4, v4, a2 /* v4 += W4*row[4] */
105 sub v4, v4, lr /* v4 -= W6*row[6] */
106 smulbb lr, ip, a4 /* lr = W2*row[6] */
107 sub v2, v2, a2 /* v2 -= W4*row[4] */
108 sub v2, v2, lr /* v2 -= W2*row[6] */
109 sub v3, v3, a2 /* v3 -= W4*row[4] */
110 add v3, v3, lr /* v3 += W2*row[6] */
112 1: add a2, v1, v5
113 mov a3, a2, lsr #11
114 bic a3, a3, #0x1f0000
115 sub a2, v2, v6
116 mov a2, a2, lsr #11
117 add a3, a3, a2, lsl #16
118 add a2, v3, v7
119 mov a4, a2, lsr #11
120 bic a4, a4, #0x1f0000
121 add a2, v4, fp
122 mov a2, a2, lsr #11
123 add a4, a4, a2, lsl #16
124 strd a3, [a1]
126 sub a2, v4, fp
127 mov a3, a2, lsr #11
128 bic a3, a3, #0x1f0000
129 sub a2, v3, v7
130 mov a2, a2, lsr #11
131 add a3, a3, a2, lsl #16
132 add a2, v2, v6
133 mov a4, a2, lsr #11
134 bic a4, a4, #0x1f0000
135 sub a2, v1, v5
136 mov a2, a2, lsr #11
137 add a4, a4, a2, lsl #16
138 strd a3, [a1, #8]
140 ldr pc, [sp], #4
142 row_dc_only:
143 orr a3, a3, a3, lsl #16
144 bic a3, a3, #0xe000
145 mov a3, a3, lsl #3
146 mov a4, a3
147 strd a3, [a1]
148 strd a3, [a1, #8]
150 ldr pc, [sp], #4
151 .endfunc
153 .macro idct_col
154 ldr a4, [a1] /* a4 = col[1:0] */
155 mov ip, #16384
156 sub ip, ip, #1 /* ip = W4 */
157 #if 0
158 mov v1, #(1<<(COL_SHIFT-1))
159 smlabt v2, ip, a4, v1 /* v2 = W4*col[1] + (1<<(COL_SHIFT-1)) */
160 smlabb v1, ip, a4, v1 /* v1 = W4*col[0] + (1<<(COL_SHIFT-1)) */
161 ldr a4, [a1, #(16*4)]
162 #else
163 mov v1, #((1<<(COL_SHIFT-1))/W4) /* this matches the C version */
164 add v2, v1, a4, asr #16
165 rsb v2, v2, v2, lsl #14
166 mov a4, a4, lsl #16
167 add v1, v1, a4, asr #16
168 ldr a4, [a1, #(16*4)]
169 rsb v1, v1, v1, lsl #14
170 #endif
172 smulbb lr, ip, a4
173 smulbt a3, ip, a4
174 sub v3, v1, lr
175 sub v5, v1, lr
176 add v7, v1, lr
177 add v1, v1, lr
178 sub v4, v2, a3
179 sub v6, v2, a3
180 add fp, v2, a3
181 ldr ip, [pc, #(w26-.-8)]
182 ldr a4, [a1, #(16*2)]
183 add v2, v2, a3
185 smulbb lr, ip, a4
186 smultb a3, ip, a4
187 add v1, v1, lr
188 sub v7, v7, lr
189 add v3, v3, a3
190 sub v5, v5, a3
191 smulbt lr, ip, a4
192 smultt a3, ip, a4
193 add v2, v2, lr
194 sub fp, fp, lr
195 add v4, v4, a3
196 ldr a4, [a1, #(16*6)]
197 sub v6, v6, a3
199 smultb lr, ip, a4
200 smulbb a3, ip, a4
201 add v1, v1, lr
202 sub v7, v7, lr
203 sub v3, v3, a3
204 add v5, v5, a3
205 smultt lr, ip, a4
206 smulbt a3, ip, a4
207 add v2, v2, lr
208 sub fp, fp, lr
209 sub v4, v4, a3
210 add v6, v6, a3
212 stmfd sp!, {v1, v2, v3, v4, v5, v6, v7, fp}
214 ldr ip, [pc, #(w13-.-8)]
215 ldr a4, [a1, #(16*1)]
216 ldr lr, [pc, #(w57-.-8)]
217 smulbb v1, ip, a4
218 smultb v3, ip, a4
219 smulbb v5, lr, a4
220 smultb v7, lr, a4
221 smulbt v2, ip, a4
222 smultt v4, ip, a4
223 smulbt v6, lr, a4
224 smultt fp, lr, a4
225 rsb v4, v4, #0
226 ldr a4, [a1, #(16*3)]
227 rsb v3, v3, #0
229 smlatb v1, ip, a4, v1
230 smlatb v3, lr, a4, v3
231 smulbb a3, ip, a4
232 smulbb a2, lr, a4
233 sub v5, v5, a3
234 sub v7, v7, a2
235 smlatt v2, ip, a4, v2
236 smlatt v4, lr, a4, v4
237 smulbt a3, ip, a4
238 smulbt a2, lr, a4
239 sub v6, v6, a3
240 ldr a4, [a1, #(16*5)]
241 sub fp, fp, a2
243 smlabb v1, lr, a4, v1
244 smlabb v3, ip, a4, v3
245 smlatb v5, lr, a4, v5
246 smlatb v7, ip, a4, v7
247 smlabt v2, lr, a4, v2
248 smlabt v4, ip, a4, v4
249 smlatt v6, lr, a4, v6
250 ldr a3, [a1, #(16*7)]
251 smlatt fp, ip, a4, fp
253 smlatb v1, lr, a3, v1
254 smlabb v3, lr, a3, v3
255 smlatb v5, ip, a3, v5
256 smulbb a4, ip, a3
257 smlatt v2, lr, a3, v2
258 sub v7, v7, a4
259 smlabt v4, lr, a3, v4
260 smulbt a4, ip, a3
261 smlatt v6, ip, a3, v6
262 sub fp, fp, a4
263 .endm
265 .align
266 .type idct_col_armv5te, %function
267 .func idct_col_armv5te
268 idct_col_armv5te:
269 str lr, [sp, #-4]!
271 idct_col
273 ldmfd sp!, {a3, a4}
274 adds a2, a3, v1
275 mov a2, a2, lsr #20
276 orrmi a2, a2, #0xf000
277 add ip, a4, v2
278 mov ip, ip, asr #20
279 orr a2, a2, ip, lsl #16
280 str a2, [a1]
281 subs a3, a3, v1
282 mov a2, a3, lsr #20
283 orrmi a2, a2, #0xf000
284 sub a4, a4, v2
285 mov a4, a4, asr #20
286 orr a2, a2, a4, lsl #16
287 ldmfd sp!, {a3, a4}
288 str a2, [a1, #(16*7)]
290 subs a2, a3, v3
291 mov a2, a2, lsr #20
292 orrmi a2, a2, #0xf000
293 sub ip, a4, v4
294 mov ip, ip, asr #20
295 orr a2, a2, ip, lsl #16
296 str a2, [a1, #(16*1)]
297 adds a3, a3, v3
298 mov a2, a3, lsr #20
299 orrmi a2, a2, #0xf000
300 add a4, a4, v4
301 mov a4, a4, asr #20
302 orr a2, a2, a4, lsl #16
303 ldmfd sp!, {a3, a4}
304 str a2, [a1, #(16*6)]
306 adds a2, a3, v5
307 mov a2, a2, lsr #20
308 orrmi a2, a2, #0xf000
309 add ip, a4, v6
310 mov ip, ip, asr #20
311 orr a2, a2, ip, lsl #16
312 str a2, [a1, #(16*2)]
313 subs a3, a3, v5
314 mov a2, a3, lsr #20
315 orrmi a2, a2, #0xf000
316 sub a4, a4, v6
317 mov a4, a4, asr #20
318 orr a2, a2, a4, lsl #16
319 ldmfd sp!, {a3, a4}
320 str a2, [a1, #(16*5)]
322 adds a2, a3, v7
323 mov a2, a2, lsr #20
324 orrmi a2, a2, #0xf000
325 add ip, a4, fp
326 mov ip, ip, asr #20
327 orr a2, a2, ip, lsl #16
328 str a2, [a1, #(16*3)]
329 subs a3, a3, v7
330 mov a2, a3, lsr #20
331 orrmi a2, a2, #0xf000
332 sub a4, a4, fp
333 mov a4, a4, asr #20
334 orr a2, a2, a4, lsl #16
335 str a2, [a1, #(16*4)]
337 ldr pc, [sp], #4
338 .endfunc
340 .align
341 .type idct_col_put_armv5te, %function
342 .func idct_col_put_armv5te
343 idct_col_put_armv5te:
344 str lr, [sp, #-4]!
346 idct_col
348 ldmfd sp!, {a3, a4}
349 ldr lr, [sp, #32]
350 add a2, a3, v1
351 movs a2, a2, asr #20
352 movmi a2, #0
353 cmp a2, #255
354 movgt a2, #255
355 add ip, a4, v2
356 movs ip, ip, asr #20
357 movmi ip, #0
358 cmp ip, #255
359 movgt ip, #255
360 orr a2, a2, ip, lsl #8
361 sub a3, a3, v1
362 movs a3, a3, asr #20
363 movmi a3, #0
364 cmp a3, #255
365 movgt a3, #255
366 sub a4, a4, v2
367 movs a4, a4, asr #20
368 movmi a4, #0
369 cmp a4, #255
370 ldr v1, [sp, #28]
371 movgt a4, #255
372 strh a2, [v1]
373 add a2, v1, #2
374 str a2, [sp, #28]
375 orr a2, a3, a4, lsl #8
376 rsb v2, lr, lr, lsl #3
377 ldmfd sp!, {a3, a4}
378 strh a2, [v2, v1]!
380 sub a2, a3, v3
381 movs a2, a2, asr #20
382 movmi a2, #0
383 cmp a2, #255
384 movgt a2, #255
385 sub ip, a4, v4
386 movs ip, ip, asr #20
387 movmi ip, #0
388 cmp ip, #255
389 movgt ip, #255
390 orr a2, a2, ip, lsl #8
391 strh a2, [v1, lr]!
392 add a3, a3, v3
393 movs a2, a3, asr #20
394 movmi a2, #0
395 cmp a2, #255
396 movgt a2, #255
397 add a4, a4, v4
398 movs a4, a4, asr #20
399 movmi a4, #0
400 cmp a4, #255
401 movgt a4, #255
402 orr a2, a2, a4, lsl #8
403 ldmfd sp!, {a3, a4}
404 strh a2, [v2, -lr]!
406 add a2, a3, v5
407 movs a2, a2, asr #20
408 movmi a2, #0
409 cmp a2, #255
410 movgt a2, #255
411 add ip, a4, v6
412 movs ip, ip, asr #20
413 movmi ip, #0
414 cmp ip, #255
415 movgt ip, #255
416 orr a2, a2, ip, lsl #8
417 strh a2, [v1, lr]!
418 sub a3, a3, v5
419 movs a2, a3, asr #20
420 movmi a2, #0
421 cmp a2, #255
422 movgt a2, #255
423 sub a4, a4, v6
424 movs a4, a4, asr #20
425 movmi a4, #0
426 cmp a4, #255
427 movgt a4, #255
428 orr a2, a2, a4, lsl #8
429 ldmfd sp!, {a3, a4}
430 strh a2, [v2, -lr]!
432 add a2, a3, v7
433 movs a2, a2, asr #20
434 movmi a2, #0
435 cmp a2, #255
436 movgt a2, #255
437 add ip, a4, fp
438 movs ip, ip, asr #20
439 movmi ip, #0
440 cmp ip, #255
441 movgt ip, #255
442 orr a2, a2, ip, lsl #8
443 strh a2, [v1, lr]
444 sub a3, a3, v7
445 movs a2, a3, asr #20
446 movmi a2, #0
447 cmp a2, #255
448 movgt a2, #255
449 sub a4, a4, fp
450 movs a4, a4, asr #20
451 movmi a4, #0
452 cmp a4, #255
453 movgt a4, #255
454 orr a2, a2, a4, lsl #8
455 strh a2, [v2, -lr]
457 ldr pc, [sp], #4
458 .endfunc
460 .align
461 .type idct_col_add_armv5te, %function
462 .func idct_col_add_armv5te
463 idct_col_add_armv5te:
464 str lr, [sp, #-4]!
466 idct_col
468 ldr lr, [sp, #36]
470 ldmfd sp!, {a3, a4}
471 ldrh ip, [lr]
472 add a2, a3, v1
473 mov a2, a2, asr #20
474 sub a3, a3, v1
475 and v1, ip, #255
476 adds a2, a2, v1
477 movmi a2, #0
478 cmp a2, #255
479 movgt a2, #255
480 add v1, a4, v2
481 mov v1, v1, asr #20
482 adds v1, v1, ip, lsr #8
483 movmi v1, #0
484 cmp v1, #255
485 movgt v1, #255
486 orr a2, a2, v1, lsl #8
487 ldr v1, [sp, #32]
488 sub a4, a4, v2
489 rsb v2, v1, v1, lsl #3
490 ldrh ip, [v2, lr]!
491 strh a2, [lr]
492 mov a3, a3, asr #20
493 and a2, ip, #255
494 adds a3, a3, a2
495 movmi a3, #0
496 cmp a3, #255
497 movgt a3, #255
498 mov a4, a4, asr #20
499 adds a4, a4, ip, lsr #8
500 movmi a4, #0
501 cmp a4, #255
502 movgt a4, #255
503 add a2, lr, #2
504 str a2, [sp, #28]
505 orr a2, a3, a4, lsl #8
506 strh a2, [v2]
508 ldmfd sp!, {a3, a4}
509 ldrh ip, [lr, v1]!
510 sub a2, a3, v3
511 mov a2, a2, asr #20
512 add a3, a3, v3
513 and v3, ip, #255
514 adds a2, a2, v3
515 movmi a2, #0
516 cmp a2, #255
517 movgt a2, #255
518 sub v3, a4, v4
519 mov v3, v3, asr #20
520 adds v3, v3, ip, lsr #8
521 movmi v3, #0
522 cmp v3, #255
523 movgt v3, #255
524 orr a2, a2, v3, lsl #8
525 add a4, a4, v4
526 ldrh ip, [v2, -v1]!
527 strh a2, [lr]
528 mov a3, a3, asr #20
529 and a2, ip, #255
530 adds a3, a3, a2
531 movmi a3, #0
532 cmp a3, #255
533 movgt a3, #255
534 mov a4, a4, asr #20
535 adds a4, a4, ip, lsr #8
536 movmi a4, #0
537 cmp a4, #255
538 movgt a4, #255
539 orr a2, a3, a4, lsl #8
540 strh a2, [v2]
542 ldmfd sp!, {a3, a4}
543 ldrh ip, [lr, v1]!
544 add a2, a3, v5
545 mov a2, a2, asr #20
546 sub a3, a3, v5
547 and v3, ip, #255
548 adds a2, a2, v3
549 movmi a2, #0
550 cmp a2, #255
551 movgt a2, #255
552 add v3, a4, v6
553 mov v3, v3, asr #20
554 adds v3, v3, ip, lsr #8
555 movmi v3, #0
556 cmp v3, #255
557 movgt v3, #255
558 orr a2, a2, v3, lsl #8
559 sub a4, a4, v6
560 ldrh ip, [v2, -v1]!
561 strh a2, [lr]
562 mov a3, a3, asr #20
563 and a2, ip, #255
564 adds a3, a3, a2
565 movmi a3, #0
566 cmp a3, #255
567 movgt a3, #255
568 mov a4, a4, asr #20
569 adds a4, a4, ip, lsr #8
570 movmi a4, #0
571 cmp a4, #255
572 movgt a4, #255
573 orr a2, a3, a4, lsl #8
574 strh a2, [v2]
576 ldmfd sp!, {a3, a4}
577 ldrh ip, [lr, v1]!
578 add a2, a3, v7
579 mov a2, a2, asr #20
580 sub a3, a3, v7
581 and v3, ip, #255
582 adds a2, a2, v3
583 movmi a2, #0
584 cmp a2, #255
585 movgt a2, #255
586 add v3, a4, fp
587 mov v3, v3, asr #20
588 adds v3, v3, ip, lsr #8
589 movmi v3, #0
590 cmp v3, #255
591 movgt v3, #255
592 orr a2, a2, v3, lsl #8
593 sub a4, a4, fp
594 ldrh ip, [v2, -v1]!
595 strh a2, [lr]
596 mov a3, a3, asr #20
597 and a2, ip, #255
598 adds a3, a3, a2
599 movmi a3, #0
600 cmp a3, #255
601 movgt a3, #255
602 mov a4, a4, asr #20
603 adds a4, a4, ip, lsr #8
604 movmi a4, #0
605 cmp a4, #255
606 movgt a4, #255
607 orr a2, a3, a4, lsl #8
608 strh a2, [v2]
610 ldr pc, [sp], #4
611 .endfunc
613 .align
614 .global simple_idct_armv5te
615 .type simple_idct_armv5te, %function
616 .func simple_idct_armv5te
617 simple_idct_armv5te:
618 stmfd sp!, {v1, v2, v3, v4, v5, v6, v7, fp, lr}
620 bl idct_row_armv5te
621 add a1, a1, #16
622 bl idct_row_armv5te
623 add a1, a1, #16
624 bl idct_row_armv5te
625 add a1, a1, #16
626 bl idct_row_armv5te
627 add a1, a1, #16
628 bl idct_row_armv5te
629 add a1, a1, #16
630 bl idct_row_armv5te
631 add a1, a1, #16
632 bl idct_row_armv5te
633 add a1, a1, #16
634 bl idct_row_armv5te
636 sub a1, a1, #(16*7)
638 bl idct_col_armv5te
639 add a1, a1, #4
640 bl idct_col_armv5te
641 add a1, a1, #4
642 bl idct_col_armv5te
643 add a1, a1, #4
644 bl idct_col_armv5te
646 ldmfd sp!, {v1, v2, v3, v4, v5, v6, v7, fp, pc}
647 .endfunc
649 .align
650 .global simple_idct_add_armv5te
651 .type simple_idct_add_armv5te, %function
652 .func simple_idct_add_armv5te
653 simple_idct_add_armv5te:
654 stmfd sp!, {a1, a2, v1, v2, v3, v4, v5, v6, v7, fp, lr}
656 mov a1, a3
658 bl idct_row_armv5te
659 add a1, a1, #16
660 bl idct_row_armv5te
661 add a1, a1, #16
662 bl idct_row_armv5te
663 add a1, a1, #16
664 bl idct_row_armv5te
665 add a1, a1, #16
666 bl idct_row_armv5te
667 add a1, a1, #16
668 bl idct_row_armv5te
669 add a1, a1, #16
670 bl idct_row_armv5te
671 add a1, a1, #16
672 bl idct_row_armv5te
674 sub a1, a1, #(16*7)
676 bl idct_col_add_armv5te
677 add a1, a1, #4
678 bl idct_col_add_armv5te
679 add a1, a1, #4
680 bl idct_col_add_armv5te
681 add a1, a1, #4
682 bl idct_col_add_armv5te
684 add sp, sp, #8
685 ldmfd sp!, {v1, v2, v3, v4, v5, v6, v7, fp, pc}
686 .endfunc
688 .align
689 .global simple_idct_put_armv5te
690 .type simple_idct_put_armv5te, %function
691 .func simple_idct_put_armv5te
692 simple_idct_put_armv5te:
693 stmfd sp!, {a1, a2, v1, v2, v3, v4, v5, v6, v7, fp, lr}
695 mov a1, a3
697 bl idct_row_armv5te
698 add a1, a1, #16
699 bl idct_row_armv5te
700 add a1, a1, #16
701 bl idct_row_armv5te
702 add a1, a1, #16
703 bl idct_row_armv5te
704 add a1, a1, #16
705 bl idct_row_armv5te
706 add a1, a1, #16
707 bl idct_row_armv5te
708 add a1, a1, #16
709 bl idct_row_armv5te
710 add a1, a1, #16
711 bl idct_row_armv5te
713 sub a1, a1, #(16*7)
715 bl idct_col_put_armv5te
716 add a1, a1, #4
717 bl idct_col_put_armv5te
718 add a1, a1, #4
719 bl idct_col_put_armv5te
720 add a1, a1, #4
721 bl idct_col_put_armv5te
723 add sp, sp, #8
724 ldmfd sp!, {v1, v2, v3, v4, v5, v6, v7, fp, pc}
725 .endfunc