dsputil_armv6.S 20.4 KB
Newer Older
1 2 3
/*
 * Copyright (c) 2009 Mans Rullgard <mans@mansr.com>
 *
4
 * This file is part of Libav.
5
 *
6
 * Libav is free software; you can redistribute it and/or
7 8 9 10
 * modify it under the terms of the GNU Lesser General Public
 * License as published by the Free Software Foundation; either
 * version 2.1 of the License, or (at your option) any later version.
 *
11
 * Libav is distributed in the hope that it will be useful,
12 13 14 15 16
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * Lesser General Public License for more details.
 *
 * You should have received a copy of the GNU Lesser General Public
17
 * License along with Libav; if not, write to the Free Software
18 19 20
 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
 */

21
#include "libavutil/arm/asm.S"
22

23 24
        preserve8

25 26 27 28 29 30 31 32
.macro  call_2x_pixels  type, subp
function ff_\type\()_pixels16\subp\()_armv6, export=1
        push            {r0-r3, lr}
        bl              ff_\type\()_pixels8\subp\()_armv6
        pop             {r0-r3, lr}
        add             r0,  r0,  #8
        add             r1,  r1,  #8
        b               ff_\type\()_pixels8\subp\()_armv6
33
endfunc
34 35 36 37 38 39 40 41 42 43 44 45 46 47
.endm

call_2x_pixels          avg
call_2x_pixels          put, _x2
call_2x_pixels          put, _y2
call_2x_pixels          put, _x2_no_rnd
call_2x_pixels          put, _y2_no_rnd

function ff_put_pixels16_armv6, export=1
        push            {r4-r11}
1:
        ldr             r5,  [r1, #4]
        ldr             r6,  [r1, #8]
        ldr             r7,  [r1, #12]
48
        ldr_post        r4,  r1,  r2
49 50
        strd            r6,  r7,  [r0, #8]
        ldr             r9,  [r1, #4]
51
        strd_post       r4,  r5,  r0,  r2
52 53
        ldr             r10, [r1, #8]
        ldr             r11, [r1, #12]
54
        ldr_post        r8,  r1,  r2
55 56
        strd            r10, r11, [r0, #8]
        subs            r3,  r3,  #2
57
        strd_post       r8,  r9,  r0,  r2
58 59 60 61
        bne             1b

        pop             {r4-r11}
        bx              lr
62
endfunc
63 64 65 66 67

function ff_put_pixels8_armv6, export=1
        push            {r4-r7}
1:
        ldr             r5,  [r1, #4]
68
        ldr_post        r4,  r1,  r2
69
        ldr             r7,  [r1, #4]
70 71
        strd_post       r4,  r5,  r0,  r2
        ldr_post        r6,  r1,  r2
72
        subs            r3,  r3,  #2
73
        strd_post       r6,  r7,  r0,  r2
74 75 76 77
        bne             1b

        pop             {r4-r7}
        bx              lr
78
endfunc
79 80 81 82 83 84 85 86 87 88 89 90

function ff_put_pixels8_x2_armv6, export=1
        push            {r4-r11, lr}
        mov             r12, #1
        orr             r12, r12, r12, lsl #8
        orr             r12, r12, r12, lsl #16
1:
        ldr             r4,  [r1]
        subs            r3,  r3,  #2
        ldr             r5,  [r1, #4]
        ldr             r7,  [r1, #5]
        lsr             r6,  r4,  #8
91
        ldr_pre         r8,  r1,  r2
92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112
        orr             r6,  r6,  r5,  lsl #24
        ldr             r9,  [r1, #4]
        ldr             r11, [r1, #5]
        lsr             r10, r8,  #8
        add             r1,  r1,  r2
        orr             r10, r10, r9,  lsl #24
        eor             r14, r4,  r6
        uhadd8          r4,  r4,  r6
        eor             r6,  r5,  r7
        uhadd8          r5,  r5,  r7
        and             r14, r14, r12
        and             r6,  r6,  r12
        uadd8           r4,  r4,  r14
        eor             r14, r8,  r10
        uadd8           r5,  r5,  r6
        eor             r6,  r9,  r11
        uhadd8          r8,  r8,  r10
        and             r14, r14, r12
        uhadd8          r9,  r9,  r11
        and             r6,  r6,  r12
        uadd8           r8,  r8,  r14
113
        strd_post       r4,  r5,  r0,  r2
114
        uadd8           r9,  r9,  r6
115
        strd_post       r8,  r9,  r0,  r2
116 117 118
        bne             1b

        pop             {r4-r11, pc}
119
endfunc
120 121 122 123 124 125 126 127

function ff_put_pixels8_y2_armv6, export=1
        push            {r4-r11}
        mov             r12, #1
        orr             r12, r12, r12, lsl #8
        orr             r12, r12, r12, lsl #16
        ldr             r4,  [r1]
        ldr             r5,  [r1, #4]
128
        ldr_pre         r6,  r1,  r2
129 130 131 132 133 134 135 136
        ldr             r7,  [r1, #4]
1:
        subs            r3,  r3,  #2
        uhadd8          r8,  r4,  r6
        eor             r10, r4,  r6
        uhadd8          r9,  r5,  r7
        eor             r11, r5,  r7
        and             r10, r10, r12
137
        ldr_pre         r4,  r1,  r2
138 139 140 141 142 143 144 145 146 147 148
        uadd8           r8,  r8,  r10
        and             r11, r11, r12
        uadd8           r9,  r9,  r11
        ldr             r5,  [r1, #4]
        uhadd8          r10, r4,  r6
        eor             r6,  r4,  r6
        uhadd8          r11, r5,  r7
        and             r6,  r6,  r12
        eor             r7,  r5,  r7
        uadd8           r10, r10, r6
        and             r7,  r7,  r12
149
        ldr_pre         r6,  r1,  r2
150
        uadd8           r11, r11, r7
151
        strd_post       r8,  r9,  r0,  r2
152
        ldr             r7,  [r1, #4]
153
        strd_post       r10, r11, r0,  r2
154 155 156 157
        bne             1b

        pop             {r4-r11}
        bx              lr
158
endfunc
159 160 161 162 163 164 165 166

function ff_put_pixels8_x2_no_rnd_armv6, export=1
        push            {r4-r9, lr}
1:
        subs            r3,  r3,  #2
        ldr             r4,  [r1]
        ldr             r5,  [r1, #4]
        ldr             r7,  [r1, #5]
167
        ldr_pre         r8,  r1,  r2
168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185
        ldr             r9,  [r1, #4]
        ldr             r14, [r1, #5]
        add             r1,  r1,  r2
        lsr             r6,  r4,  #8
        orr             r6,  r6,  r5,  lsl #24
        lsr             r12, r8,  #8
        orr             r12, r12, r9,  lsl #24
        uhadd8          r4,  r4,  r6
        uhadd8          r5,  r5,  r7
        uhadd8          r8,  r8,  r12
        uhadd8          r9,  r9,  r14
        stm             r0,  {r4,r5}
        add             r0,  r0,  r2
        stm             r0,  {r8,r9}
        add             r0,  r0,  r2
        bne             1b

        pop             {r4-r9, pc}
186
endfunc
187 188 189 190 191

function ff_put_pixels8_y2_no_rnd_armv6, export=1
        push            {r4-r9, lr}
        ldr             r4,  [r1]
        ldr             r5,  [r1, #4]
192
        ldr_pre         r6,  r1,  r2
193 194 195 196
        ldr             r7,  [r1, #4]
1:
        subs            r3,  r3,  #2
        uhadd8          r8,  r4,  r6
197
        ldr_pre         r4,  r1,  r2
198 199 200
        uhadd8          r9,  r5,  r7
        ldr             r5,  [r1, #4]
        uhadd8          r12, r4,  r6
201
        ldr_pre         r6,  r1,  r2
202 203 204 205 206 207 208 209 210
        uhadd8          r14, r5,  r7
        ldr             r7,  [r1, #4]
        stm             r0,  {r8,r9}
        add             r0,  r0,  r2
        stm             r0,  {r12,r14}
        add             r0,  r0,  r2
        bne             1b

        pop             {r4-r9, pc}
211
endfunc
212 213 214 215 216 217 218 219 220

function ff_avg_pixels8_armv6, export=1
        pld             [r1, r2]
        push            {r4-r10, lr}
        mov             lr,  #1
        orr             lr,  lr,  lr,  lsl #8
        orr             lr,  lr,  lr,  lsl #16
        ldrd            r4,  r5,  [r0]
        ldr             r10, [r1, #4]
221
        ldr_post        r9,  r1,  r2
222 223 224 225 226 227
        subs            r3,  r3,  #2
1:
        pld             [r1, r2]
        eor             r8,  r4,  r9
        uhadd8          r4,  r4,  r9
        eor             r12, r5,  r10
228
        ldrd_reg        r6,  r7,  r0,  r2
229 230 231 232 233
        uhadd8          r5,  r5,  r10
        and             r8,  r8,  lr
        ldr             r10, [r1, #4]
        and             r12, r12, lr
        uadd8           r4,  r4,  r8
234
        ldr_post        r9,  r1,  r2
235 236 237 238 239
        eor             r8,  r6,  r9
        uadd8           r5,  r5,  r12
        pld             [r1, r2,  lsl #1]
        eor             r12, r7,  r10
        uhadd8          r6,  r6,  r9
240
        strd_post       r4,  r5,  r0,  r2
241 242 243
        uhadd8          r7,  r7,  r10
        beq             2f
        and             r8,  r8,  lr
244
        ldrd_reg        r4,  r5,  r0,  r2
245 246 247 248 249
        uadd8           r6,  r6,  r8
        ldr             r10, [r1, #4]
        and             r12, r12, lr
        subs            r3,  r3,  #2
        uadd8           r7,  r7,  r12
250 251
        ldr_post        r9,  r1,  r2
        strd_post       r6,  r7,  r0,  r2
252 253 254 255 256 257
        b               1b
2:
        and             r8,  r8,  lr
        and             r12, r12, lr
        uadd8           r6,  r6,  r8
        uadd8           r7,  r7,  r12
258
        strd_post       r6,  r7,  r0,  r2
259 260

        pop             {r4-r10, pc}
261
endfunc
262

263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284
function ff_add_pixels_clamped_armv6, export=1
        push            {r4-r8,lr}
        mov             r3,  #8
1:
        ldm             r0!, {r4,r5,r12,lr}
        ldrd            r6,  r7,  [r1]
        pkhbt           r8,  r4,  r5,  lsl #16
        pkhtb           r5,  r5,  r4,  asr #16
        pkhbt           r4,  r12, lr,  lsl #16
        pkhtb           lr,  lr,  r12, asr #16
        pld             [r1, r2]
        uxtab16         r8,  r8,  r6
        uxtab16         r5,  r5,  r6,  ror #8
        uxtab16         r4,  r4,  r7
        uxtab16         lr,  lr,  r7,  ror #8
        usat16          r8,  #8,  r8
        usat16          r5,  #8,  r5
        usat16          r4,  #8,  r4
        usat16          lr,  #8,  lr
        orr             r6,  r8,  r5,  lsl #8
        orr             r7,  r4,  lr,  lsl #8
        subs            r3,  r3,  #1
285
        strd_post       r6,  r7,  r1,  r2
286 287
        bgt             1b
        pop             {r4-r8,pc}
288
endfunc
289

290 291 292 293 294
function ff_get_pixels_armv6, export=1
        pld             [r1, r2]
        push            {r4-r8, lr}
        mov             lr,  #8
1:
295
        ldrd_post       r4,  r5,  r1,  r2
296 297 298 299 300 301 302 303 304 305 306 307 308 309
        subs            lr,  lr,  #1
        uxtb16          r6,  r4
        uxtb16          r4,  r4,  ror #8
        uxtb16          r12, r5
        uxtb16          r8,  r5,  ror #8
        pld             [r1, r2]
        pkhbt           r5,  r6,  r4,  lsl #16
        pkhtb           r6,  r4,  r6,  asr #16
        pkhbt           r7,  r12, r8,  lsl #16
        pkhtb           r12, r8,  r12, asr #16
        stm             r0!, {r5,r6,r7,r12}
        bgt             1b

        pop             {r4-r8, pc}
310
endfunc
311

312 313 314 315 316 317
function ff_diff_pixels_armv6, export=1
        pld             [r1, r3]
        pld             [r2, r3]
        push            {r4-r9, lr}
        mov             lr,  #8
1:
318 319
        ldrd_post       r4,  r5,  r1,  r3
        ldrd_post       r6,  r7,  r2,  r3
320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342
        uxtb16          r8,  r4
        uxtb16          r4,  r4,  ror #8
        uxtb16          r9,  r6
        uxtb16          r6,  r6,  ror #8
        pld             [r1, r3]
        ssub16          r9,  r8,  r9
        ssub16          r6,  r4,  r6
        uxtb16          r8,  r5
        uxtb16          r5,  r5,  ror #8
        pld             [r2, r3]
        pkhbt           r4,  r9,  r6,  lsl #16
        pkhtb           r6,  r6,  r9,  asr #16
        uxtb16          r9,  r7
        uxtb16          r7,  r7,  ror #8
        ssub16          r9,  r8,  r9
        ssub16          r5,  r5,  r7
        subs            lr,  lr,  #1
        pkhbt           r8,  r9,  r5,  lsl #16
        pkhtb           r9,  r5,  r9,  asr #16
        stm             r0!, {r4,r6,r8,r9}
        bgt             1b

        pop             {r4-r9, pc}
343
endfunc
344

345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371
function ff_pix_abs16_armv6, export=1
        ldr             r0,  [sp]
        push            {r4-r9, lr}
        mov             r12, #0
        mov             lr,  #0
        ldm             r1,  {r4-r7}
        ldr             r8,  [r2]
1:
        ldr             r9,  [r2, #4]
        pld             [r1, r3]
        usada8          r12, r4,  r8,  r12
        ldr             r8,  [r2, #8]
        pld             [r2, r3]
        usada8          lr,  r5,  r9,  lr
        ldr             r9,  [r2, #12]
        usada8          r12, r6,  r8,  r12
        subs            r0,  r0,  #1
        usada8          lr,  r7,  r9,  lr
        beq             2f
        add             r1,  r1,  r3
        ldm             r1,  {r4-r7}
        add             r2,  r2,  r3
        ldr             r8,  [r2]
        b               1b
2:
        add             r0,  r12, lr
        pop             {r4-r9, pc}
372
endfunc
373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426

function ff_pix_abs16_x2_armv6, export=1
        ldr             r12, [sp]
        push            {r4-r11, lr}
        mov             r0,  #0
        mov             lr,  #1
        orr             lr,  lr,  lr,  lsl #8
        orr             lr,  lr,  lr,  lsl #16
1:
        ldr             r8,  [r2]
        ldr             r9,  [r2, #4]
        lsr             r10, r8,  #8
        ldr             r4,  [r1]
        lsr             r6,  r9,  #8
        orr             r10, r10, r9,  lsl #24
        ldr             r5,  [r2, #8]
        eor             r11, r8,  r10
        uhadd8          r7,  r8,  r10
        orr             r6,  r6,  r5,  lsl #24
        and             r11, r11, lr
        uadd8           r7,  r7,  r11
        ldr             r8,  [r1, #4]
        usada8          r0,  r4,  r7,  r0
        eor             r7,  r9,  r6
        lsr             r10, r5,  #8
        and             r7,  r7,  lr
        uhadd8          r4,  r9,  r6
        ldr             r6,  [r2, #12]
        uadd8           r4,  r4,  r7
        pld             [r1, r3]
        orr             r10, r10, r6,  lsl #24
        usada8          r0,  r8,  r4,  r0
        ldr             r4,  [r1, #8]
        eor             r11, r5,  r10
        ldrb            r7,  [r2, #16]
        and             r11, r11, lr
        uhadd8          r8,  r5,  r10
        ldr             r5,  [r1, #12]
        uadd8           r8,  r8,  r11
        pld             [r2, r3]
        lsr             r10, r6,  #8
        usada8          r0,  r4,  r8,  r0
        orr             r10, r10, r7,  lsl #24
        subs            r12,  r12,  #1
        eor             r11, r6,  r10
        add             r1,  r1,  r3
        uhadd8          r9,  r6,  r10
        and             r11, r11, lr
        uadd8           r9,  r9,  r11
        add             r2,  r2,  r3
        usada8          r0,  r5,  r9,  r0
        bgt             1b

        pop             {r4-r11, pc}
427
endfunc
428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484

.macro  usad_y2         p0,  p1,  p2,  p3,  n0,  n1,  n2,  n3
        ldr             \n0, [r2]
        eor             \n1, \p0, \n0
        uhadd8          \p0, \p0, \n0
        and             \n1, \n1, lr
        ldr             \n2, [r1]
        uadd8           \p0, \p0, \n1
        ldr             \n1, [r2, #4]
        usada8          r0,  \p0, \n2, r0
        pld             [r1,  r3]
        eor             \n3, \p1, \n1
        uhadd8          \p1, \p1, \n1
        and             \n3, \n3, lr
        ldr             \p0, [r1, #4]
        uadd8           \p1, \p1, \n3
        ldr             \n2, [r2, #8]
        usada8          r0,  \p1, \p0, r0
        pld             [r2,  r3]
        eor             \p0, \p2, \n2
        uhadd8          \p2, \p2, \n2
        and             \p0, \p0, lr
        ldr             \p1, [r1, #8]
        uadd8           \p2, \p2, \p0
        ldr             \n3, [r2, #12]
        usada8          r0,  \p2, \p1, r0
        eor             \p1, \p3, \n3
        uhadd8          \p3, \p3, \n3
        and             \p1, \p1, lr
        ldr             \p0,  [r1, #12]
        uadd8           \p3, \p3, \p1
        add             r1,  r1,  r3
        usada8          r0,  \p3, \p0,  r0
        add             r2,  r2,  r3
.endm

function ff_pix_abs16_y2_armv6, export=1
        pld             [r1]
        pld             [r2]
        ldr             r12, [sp]
        push            {r4-r11, lr}
        mov             r0,  #0
        mov             lr,  #1
        orr             lr,  lr,  lr,  lsl #8
        orr             lr,  lr,  lr,  lsl #16
        ldr             r4,  [r2]
        ldr             r5,  [r2, #4]
        ldr             r6,  [r2, #8]
        ldr             r7,  [r2, #12]
        add             r2,  r2,  r3
1:
        usad_y2         r4,  r5,  r6,  r7,  r8,  r9,  r10, r11
        subs            r12, r12, #2
        usad_y2         r8,  r9,  r10, r11, r4,  r5,  r6,  r7
        bgt             1b

        pop             {r4-r11, pc}
485
endfunc
486 487 488 489 490 491 492

function ff_pix_abs8_armv6, export=1
        pld             [r2, r3]
        ldr             r12, [sp]
        push            {r4-r9, lr}
        mov             r0,  #0
        mov             lr,  #0
493
        ldrd_post       r4,  r5,  r1,  r3
494 495 496
1:
        subs            r12, r12, #2
        ldr             r7,  [r2, #4]
497 498
        ldr_post        r6,  r2,  r3
        ldrd_post       r8,  r9,  r1,  r3
499 500 501 502
        usada8          r0,  r4,  r6,  r0
        pld             [r2, r3]
        usada8          lr,  r5,  r7,  lr
        ldr             r7,  [r2, #4]
503
        ldr_post        r6,  r2,  r3
504
        beq             2f
505
        ldrd_post       r4,  r5,  r1,  r3
506 507 508 509 510 511 512 513 514
        usada8          r0,  r8,  r6,  r0
        pld             [r2, r3]
        usada8          lr,  r9,  r7,  lr
        b               1b
2:
        usada8          r0,  r8,  r6,  r0
        usada8          lr,  r9,  r7,  lr
        add             r0,  r0,  lr
        pop             {r4-r9, pc}
515
endfunc
Måns Rullgård's avatar
Måns Rullgård committed
516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565

function ff_sse16_armv6, export=1
        ldr             r12, [sp]
        push            {r4-r9, lr}
        mov             r0,  #0
1:
        ldrd            r4,  r5,  [r1]
        ldr             r8,  [r2]
        uxtb16          lr,  r4
        uxtb16          r4,  r4,  ror #8
        uxtb16          r9,  r8
        uxtb16          r8,  r8,  ror #8
        ldr             r7,  [r2, #4]
        usub16          lr,  lr,  r9
        usub16          r4,  r4,  r8
        smlad           r0,  lr,  lr,  r0
        uxtb16          r6,  r5
        uxtb16          lr,  r5,  ror #8
        uxtb16          r8,  r7
        uxtb16          r9,  r7,  ror #8
        smlad           r0,  r4,  r4,  r0
        ldrd            r4,  r5,  [r1, #8]
        usub16          r6,  r6,  r8
        usub16          r8,  lr,  r9
        ldr             r7,  [r2, #8]
        smlad           r0,  r6,  r6,  r0
        uxtb16          lr,  r4
        uxtb16          r4,  r4,  ror #8
        uxtb16          r9,  r7
        uxtb16          r7,  r7, ror #8
        smlad           r0,  r8,  r8,  r0
        ldr             r8,  [r2, #12]
        usub16          lr,  lr,  r9
        usub16          r4,  r4,  r7
        smlad           r0,  lr,  lr,  r0
        uxtb16          r6,  r5
        uxtb16          r5,  r5,  ror #8
        uxtb16          r9,  r8
        uxtb16          r8,  r8,  ror #8
        smlad           r0,  r4,  r4,  r0
        usub16          r6,  r6,  r9
        usub16          r5,  r5,  r8
        smlad           r0,  r6,  r6,  r0
        add             r1,  r1,  r3
        add             r2,  r2,  r3
        subs            r12, r12, #1
        smlad           r0,  r5,  r5,  r0
        bgt             1b

        pop             {r4-r9, pc}
566
endfunc
567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595

function ff_pix_norm1_armv6, export=1
        push            {r4-r6, lr}
        mov             r12, #16
        mov             lr,  #0
1:
        ldm             r0,  {r2-r5}
        uxtb16          r6,  r2
        uxtb16          r2,  r2,  ror #8
        smlad           lr,  r6,  r6,  lr
        uxtb16          r6,  r3
        smlad           lr,  r2,  r2,  lr
        uxtb16          r3,  r3,  ror #8
        smlad           lr,  r6,  r6,  lr
        uxtb16          r6,  r4
        smlad           lr,  r3,  r3,  lr
        uxtb16          r4,  r4,  ror #8
        smlad           lr,  r6,  r6,  lr
        uxtb16          r6,  r5
        smlad           lr,  r4,  r4,  lr
        uxtb16          r5,  r5,  ror #8
        smlad           lr,  r6,  r6,  lr
        subs            r12, r12, #1
        add             r0,  r0,  r1
        smlad           lr,  r5,  r5,  lr
        bgt             1b

        mov             r0,  lr
        pop             {r4-r6, pc}
596
endfunc
597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613

function ff_pix_sum_armv6, export=1
        push            {r4-r7, lr}
        mov             r12, #16
        mov             r2,  #0
        mov             r3,  #0
        mov             lr,  #0
        ldr             r4,  [r0]
1:
        subs            r12, r12, #1
        ldr             r5,  [r0, #4]
        usada8          r2,  r4,  lr,  r2
        ldr             r6,  [r0, #8]
        usada8          r3,  r5,  lr,  r3
        ldr             r7,  [r0, #12]
        usada8          r2,  r6,  lr,  r2
        beq             2f
614
        ldr_pre         r4,  r0,  r1
615 616 617 618 619 620
        usada8          r3,  r7,  lr,  r3
        bgt             1b
2:
        usada8          r3,  r7,  lr,  r3
        add             r0,  r2,  r3
        pop             {r4-r7, pc}
621
endfunc