rv40dsp_neon.S 33.4 KB
Newer Older
1 2
/*
 * Copyright (c) 2011 Janne Grunau <janne-libav@jannau.net>
3
 * Copyright (c) 2011 Mans Rullgard <mans@mansr.com>
4
 *
5
 * This file is part of FFmpeg.
6
 *
7
 * FFmpeg is free software; you can redistribute it and/or
8 9 10 11
 * 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.
 *
12
 * FFmpeg is distributed in the hope that it will be useful,
13 14 15 16 17
 * 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
18
 * License along with FFmpeg; if not, write to the Free Software
19 20 21
 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
 */

22
#include "libavutil/arm/asm.S"
23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374
#include "neon.S"

.macro  qpel_lowpass    r0,  r1,  rc1, rc2, shift
        vext.8          d25, \r0, \r1, #1       @ src[-1]
        vext.8          d26, \r0, \r1, #4       @ src[ 2]
        vext.8          d24, \r0, \r1, #5       @ src[ 3]
        vaddl.u8        q9,  d25, d26
        vaddl.u8        q8,  \r0, d24
        vext.8          d27, \r0, \r1, #2       @ src[ 0]
        vshl.s16        q12, q9,  #2
        vsub.s16        q8,  q8,  q9
        vext.8          d28, \r0, \r1, #3       @ src[ 1]
        vsub.s16        q8,  q8,  q12
        vmlal.u8        q8,  d27, \rc1
        vmlal.u8        q8,  d28, \rc2
        vqrshrun.s16    \r0, q8,  #\shift
.endm

.macro  qpel_lowpass_x2 r0,  r1,  r2,  r3,  rc1, rc2, shift
        vext.8          d25, \r0, \r1, #1       @ src[-1]
        vext.8          d26, \r0, \r1, #4       @ src[ 2]
        vext.8          d24, \r0, \r1, #5       @ src[ 3]
        vaddl.u8        q9,  d25, d26
        vaddl.u8        q8,  \r0, d24
        vext.8          d29, \r0, \r1, #2       @ src[ 0]
        vext.8          d28, \r0, \r1, #3       @ src[ 1]
        vshl.s16        q10, q9,  #2
        vext.8          \r1, \r2, \r3, #1       @ src[-1]
        vsub.s16        q8,  q8,  q9
        vext.8          d22, \r2, \r3, #4       @ src[ 2]
        vext.8          \r0, \r2, \r3, #5       @ src[ 3]
        vaddl.u8        q13, \r1, d22
        vaddl.u8        q12, \r2, \r0
        vsub.s16        q8,  q8,  q10
        vshl.s16        q9,  q13, #2
        vsub.s16        q12, q12, q13
        vmlal.u8        q8,  d29, \rc1
        vmlal.u8        q8,  d28, \rc2
        vsub.s16        q12, q12, q9
        vext.8          d26, \r2, \r3, #2       @ src[ 0]
        vext.8          d27, \r2, \r3, #3       @ src[ 1]
        vmlal.u8        q12, d26, \rc1
        vmlal.u8        q12, d27, \rc2
        vqrshrun.s16    \r0, q8,  #\shift
        vqrshrun.s16    \r2, q12, #\shift
.endm

.macro  rv40_qpel8_h    shift
function put_rv40_qpel8_h_lp_packed_s\shift\()_neon
1:
        vld1.8          {q2},     [r1], r2
        vld1.8          {q3},     [r1], r2
        qpel_lowpass_x2 d4,  d5,  d6,  d7,  d0,  d1,  \shift
        vst1.8          {d4},     [r12,:64]!
        vst1.8          {d6},     [r12,:64]!
        subs            r3,  r3,  #2
        bgt             1b
        vld1.8          {q2},     [r1]
        qpel_lowpass    d4,  d5,  d0,  d1,  \shift
        vst1.8          {d4},     [r12,:64]!
        bx              lr
endfunc
.endm

.macro  rv40_qpel8_v    shift, type
function \type\()_rv40_qpel8_v_lp_packed_s\shift\()_neon
        vld1.64         {d2},     [r1,:64]!
        vld1.64         {d3},     [r1,:64]!
        vld1.64         {d4},     [r1,:64]!
        vld1.64         {d5},     [r1,:64]!
        vld1.64         {d6},     [r1,:64]!
        vld1.64         {d7},     [r1,:64]!
        vld1.64         {d8},     [r1,:64]!
        vld1.64         {d9},     [r1,:64]!
        vld1.64         {d10},    [r1,:64]!
        vld1.64         {d11},    [r1,:64]!
        vld1.64         {d12},    [r1,:64]!
        vld1.64         {d13},    [r1,:64]!
        vld1.64         {d14},    [r1,:64]!
        transpose_8x8   d2,  d3,  d4,  d5,  d6,  d7,  d8,  d9
        transpose_8x8   d10, d11, d12, d13, d14, d15, d30, d31
        qpel_lowpass_x2 d2,  d10, d3,  d11, d0,  d1,  \shift
        qpel_lowpass_x2 d4,  d12, d5,  d13, d0,  d1,  \shift
        qpel_lowpass_x2 d6,  d14, d7,  d15, d0,  d1,  \shift
        qpel_lowpass_x2 d8,  d30, d9,  d31, d0,  d1,  \shift
        transpose_8x8   d2,  d3,  d4,  d5,  d6,  d7,  d8,  d9
  .ifc \type,avg
        vld1.64         d12,      [r0,:64], r2
        vld1.64         d13,      [r0,:64], r2
        vld1.64         d14,      [r0,:64], r2
        vld1.64         d15,      [r0,:64], r2
        vld1.64         d16,      [r0,:64], r2
        vld1.64         d17,      [r0,:64], r2
        vld1.64         d18,      [r0,:64], r2
        vld1.64         d19,      [r0,:64], r2
        sub             r0,  r0,  r2,  lsl #3
        vrhadd.u8       q1,  q1,  q6
        vrhadd.u8       q2,  q2,  q7
        vrhadd.u8       q3,  q3,  q8
        vrhadd.u8       q4,  q4,  q9
  .endif
        vst1.64         d2,       [r0,:64], r2
        vst1.64         d3,       [r0,:64], r2
        vst1.64         d4,       [r0,:64], r2
        vst1.64         d5,       [r0,:64], r2
        vst1.64         d6,       [r0,:64], r2
        vst1.64         d7,       [r0,:64], r2
        vst1.64         d8,       [r0,:64], r2
        vst1.64         d9,       [r0,:64], r2
        bx              lr
endfunc
.endm

        rv40_qpel8_h    5
        rv40_qpel8_h    6

.macro  rv40_qpel       type
function \type\()_rv40_qpel8_h_lowpass_neon
  .ifc \type,avg
        mov             r12, r0
  .endif
1:
        vld1.8          {q2},     [r1], r2
        vld1.8          {q3},     [r1], r2
        qpel_lowpass_x2 d4,  d5,  d6,  d7,  d0,  d1,  6
  .ifc \type,avg
        vld1.8          {d3},     [r12,:64], r2
        vld1.8          {d16},    [r12,:64], r2
        vrhadd.u8       d4,  d4,  d3
        vrhadd.u8       d6,  d6,  d16
  .endif
        vst1.8          {d4},     [r0,:64], r2
        vst1.8          {d6},     [r0,:64], r2
        subs            r3,  r3,  #2
        bgt             1b
        bx              lr
endfunc

function \type\()_rv40_qpel8_v_lowpass_neon
        vld1.64         {d2},     [r1], r2
        vld1.64         {d3},     [r1], r2
        vld1.64         {d4},     [r1], r2
        vld1.64         {d5},     [r1], r2
        vld1.64         {d6},     [r1], r2
        vld1.64         {d7},     [r1], r2
        vld1.64         {d8},     [r1], r2
        vld1.64         {d9},     [r1], r2
        vld1.64         {d10},    [r1], r2
        vld1.64         {d11},    [r1], r2
        vld1.64         {d12},    [r1], r2
        vld1.64         {d13},    [r1], r2
        vld1.64         {d14},    [r1]
        transpose_8x8   d2,  d3,  d4,  d5,  d6,  d7,  d8,  d9
        transpose_8x8   d10, d11, d12, d13, d14, d15, d30, d31
        qpel_lowpass_x2 d2,  d10, d3,  d11, d0,  d1,  6
        qpel_lowpass_x2 d4,  d12, d5,  d13, d0,  d1,  6
        qpel_lowpass_x2 d6,  d14, d7,  d15, d0,  d1,  6
        qpel_lowpass_x2 d8,  d30, d9,  d31, d0,  d1,  6
        transpose_8x8   d2,  d3,  d4,  d5,  d6,  d7,  d8,  d9
  .ifc \type,avg
        vld1.64         d12,      [r0,:64], r2
        vld1.64         d13,      [r0,:64], r2
        vld1.64         d14,      [r0,:64], r2
        vld1.64         d15,      [r0,:64], r2
        vld1.64         d16,      [r0,:64], r2
        vld1.64         d17,      [r0,:64], r2
        vld1.64         d18,      [r0,:64], r2
        vld1.64         d19,      [r0,:64], r2
        sub             r0,  r0,  r2,  lsl #3
        vrhadd.u8       q1,  q1,  q6
        vrhadd.u8       q2,  q2,  q7
        vrhadd.u8       q3,  q3,  q8
        vrhadd.u8       q4,  q4,  q9
  .endif
        vst1.64         d2,       [r0,:64], r2
        vst1.64         d3,       [r0,:64], r2
        vst1.64         d4,       [r0,:64], r2
        vst1.64         d5,       [r0,:64], r2
        vst1.64         d6,       [r0,:64], r2
        vst1.64         d7,       [r0,:64], r2
        vst1.64         d8,       [r0,:64], r2
        vst1.64         d9,       [r0,:64], r2
        bx              lr
endfunc

        rv40_qpel8_v    5, \type
        rv40_qpel8_v    6, \type

function ff_\type\()_rv40_qpel8_mc10_neon, export=1
        sub             r1,  r1,  #2
        mov             r3,  #8
        vmov.i8         d0,  #52
        vmov.i8         d1,  #20
        b               \type\()_rv40_qpel8_h_lowpass_neon
endfunc

function ff_\type\()_rv40_qpel8_mc30_neon, export=1
        sub             r1,  r1,  #2
        mov             r3,  #8
        vmov.i8         d0,  #20
        vmov.i8         d1,  #52
        b               \type\()_rv40_qpel8_h_lowpass_neon
endfunc

function ff_\type\()_rv40_qpel8_mc01_neon, export=1
        push            {r4, lr}
        vpush           {d8-d15}
        sub             r1,  r1,  r2,  lsl #1
        vmov.i8         d0,  #52
        vmov.i8         d1,  #20
        bl              \type\()_rv40_qpel8_v_lowpass_neon
        vpop            {d8-d15}
        pop             {r4, pc}
endfunc

function ff_\type\()_rv40_qpel8_mc11_neon, export=1
        push            {r4, lr}
        vpush           {d8-d15}
        sub             sp,  sp,  #14*8
        add             r12, sp,  #7
        bic             r12, r12, #7
        sub             r1,  r1,  r2,  lsl #1
        sub             r1,  r1,  #2
        mov             r3,  #12
        vmov.i8         d0,  #52
        vmov.i8         d1,  #20
        bl              put_rv40_qpel8_h_lp_packed_s6_neon
        add             r1,  sp,  #7
        bic             r1,  r1,  #7
        bl              \type\()_rv40_qpel8_v_lp_packed_s6_neon
        add             sp,  sp,  #14*8
        vpop            {d8-d15}
        pop             {r4, pc}
endfunc

function ff_\type\()_rv40_qpel8_mc21_neon, export=1
        push            {r4, lr}
        vpush           {d8-d15}
        sub             sp,  sp,  #14*8
        add             r12, sp,  #7
        bic             r12, r12, #7
        sub             r1,  r1,  r2,  lsl #1
        sub             r1,  r1,  #2
        mov             r3,  #12
        vmov.i8         d0,  #20
        vmov.i8         d1,  #20
        bl              put_rv40_qpel8_h_lp_packed_s5_neon
        add             r1,  sp,  #7
        bic             r1,  r1,  #7
        vmov.i8         d0,  #52
        bl              \type\()_rv40_qpel8_v_lp_packed_s6_neon
        add             sp,  sp,  #14*8
        vpop            {d8-d15}
        pop             {r4, pc}
endfunc

function ff_\type\()_rv40_qpel8_mc31_neon, export=1
        push            {r4, lr}
        vpush           {d8-d15}
        sub             sp,  sp,  #14*8
        add             r12, sp,  #7
        bic             r12, r12, #7
        sub             r1,  r1,  r2,  lsl #1
        sub             r1,  r1,  #2
        mov             r3,  #12
        vmov.i8         d0,  #20
        vmov.i8         d1,  #52
        bl              put_rv40_qpel8_h_lp_packed_s6_neon
        add             r1,  sp,  #7
        bic             r1,  r1,  #7
        vswp            d0,  d1
        bl              \type\()_rv40_qpel8_v_lp_packed_s6_neon
        add             sp,  sp,  #14*8
        vpop            {d8-d15}
        pop             {r4, pc}
endfunc

function ff_\type\()_rv40_qpel8_mc12_neon, export=1
        push            {r4, lr}
        vpush           {d8-d15}
        sub             sp,  sp,  #14*8
        add             r12, sp,  #7
        bic             r12, r12, #7
        sub             r1,  r1,  r2,  lsl #1
        sub             r1,  r1,  #2
        mov             r3,  #12
        vmov.i8         d0,  #52
        vmov.i8         d1,  #20
        bl              put_rv40_qpel8_h_lp_packed_s6_neon
        add             r1,  sp,  #7
        bic             r1,  r1,  #7
        vmov.i8         d0,  #20
        bl              \type\()_rv40_qpel8_v_lp_packed_s5_neon
        add             sp,  sp,  #14*8
        vpop            {d8-d15}
        pop             {r4, pc}
endfunc

function ff_\type\()_rv40_qpel8_mc22_neon, export=1
        push            {r4, lr}
        vpush           {d8-d15}
        sub             sp,  sp,  #14*8
        add             r12, sp,  #7
        bic             r12, r12, #7
        sub             r1,  r1,  r2,  lsl #1
        sub             r1,  r1,  #2
        mov             r3,  #12
        vmov.i8         d0,  #20
        vmov.i8         d1,  #20
        bl              put_rv40_qpel8_h_lp_packed_s5_neon
        add             r1,  sp,  #7
        bic             r1,  r1,  #7
        bl              \type\()_rv40_qpel8_v_lp_packed_s5_neon
        add             sp,  sp,  #14*8
        vpop            {d8-d15}
        pop             {r4, pc}
endfunc

function ff_\type\()_rv40_qpel8_mc32_neon, export=1
        push            {r4, lr}
        vpush           {d8-d15}
        sub             sp,  sp,  #14*8
        add             r12, sp,  #7
        bic             r12, r12, #7
        sub             r1,  r1,  r2,  lsl #1
        sub             r1,  r1,  #2
        mov             r3,  #12
        vmov.i8         d0,  #20
        vmov.i8         d1,  #52
        bl              put_rv40_qpel8_h_lp_packed_s6_neon
        add             r1,  sp,  #7
        bic             r1,  r1,  #7
        vmov.i8         d1,  #20
        bl              \type\()_rv40_qpel8_v_lp_packed_s5_neon
        add             sp,  sp,  #14*8
        vpop            {d8-d15}
        pop             {r4, pc}
endfunc

function ff_\type\()_rv40_qpel8_mc03_neon, export=1
        push            {r4, lr}
        vpush           {d8-d15}
        sub             r1,  r1,  r2,  lsl #1
        vmov.i8         d0,  #20
        vmov.i8         d1,  #52
        bl              \type\()_rv40_qpel8_v_lowpass_neon
        vpop            {d8-d15}
        pop             {r4, pc}
endfunc

function ff_\type\()_rv40_qpel8_mc33_neon, export=1
        mov             r3,  #8
375
        b               X(ff_\type\()_pixels8_xy2_neon)
376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654
endfunc

function ff_\type\()_rv40_qpel8_mc13_neon, export=1
        push            {r4, lr}
        vpush           {d8-d15}
        sub             sp,  sp,  #14*8
        add             r12, sp,  #7
        bic             r12, r12, #7
        sub             r1,  r1,  r2,  lsl #1
        sub             r1,  r1,  #2
        mov             r3,  #12
        vmov.i8         d0,  #52
        vmov.i8         d1,  #20
        bl              put_rv40_qpel8_h_lp_packed_s6_neon
        add             r1,  sp,  #7
        bic             r1,  r1,  #7
        vswp            d0,  d1
        bl              \type\()_rv40_qpel8_v_lp_packed_s6_neon
        add             sp,  sp,  #14*8
        vpop            {d8-d15}
        pop             {r4, pc}
endfunc

function ff_\type\()_rv40_qpel8_mc23_neon, export=1
        push            {r4, lr}
        vpush           {d8-d15}
        sub             sp,  sp,  #14*8
        add             r12, sp,  #7
        bic             r12, r12, #7
        sub             r1,  r1,  r2,  lsl #1
        sub             r1,  r1,  #2
        mov             r3,  #12
        vmov.i8         d0,  #20
        vmov.i8         d1,  #20
        bl              put_rv40_qpel8_h_lp_packed_s5_neon
        add             r1,  sp,  #7
        bic             r1,  r1,  #7
        vmov.i8         d1,  #52
        bl              \type\()_rv40_qpel8_v_lp_packed_s6_neon
        add             sp,  sp,  #14*8
        vpop            {d8-d15}
        pop             {r4, pc}
endfunc

function ff_\type\()_rv40_qpel16_mc10_neon, export=1
        vmov.i8         d0,  #52
        vmov.i8         d1,  #20
.L\type\()_rv40_qpel16_h:
        push            {r1, lr}
        sub             r1,  r1,  #2
        mov             r3,  #16
        bl              \type\()_rv40_qpel8_h_lowpass_neon
        pop             {r1, lr}
        sub             r0,  r0,  r2,  lsl #4
        add             r0,  r0,  #8
        add             r1,  r1,  #6
        mov             r3,  #16
        b               \type\()_rv40_qpel8_h_lowpass_neon
endfunc

function ff_\type\()_rv40_qpel16_mc30_neon, export=1
        vmov.i8         d0,  #20
        vmov.i8         d1,  #52
        b               .L\type\()_rv40_qpel16_h
endfunc

function ff_\type\()_rv40_qpel16_mc01_neon, export=1
        vmov.i8         d0,  #52
        vmov.i8         d1,  #20
.L\type\()_rv40_qpel16_v:
        sub             r1,  r1,  r2,  lsl #1
        push            {r1, lr}
        vpush           {d8-d15}
        bl              \type\()_rv40_qpel8_v_lowpass_neon
        sub             r1,  r1,  r2,  lsl #2
        bl              \type\()_rv40_qpel8_v_lowpass_neon
        ldr             r1,  [sp, #64]
        sub             r0,  r0,  r2,  lsl #4
        add             r0,  r0,  #8
        add             r1,  r1,  #8
        bl              \type\()_rv40_qpel8_v_lowpass_neon
        sub             r1,  r1,  r2,  lsl #2
        bl              \type\()_rv40_qpel8_v_lowpass_neon
        vpop            {d8-d15}
        pop             {r1, pc}
endfunc

function ff_\type\()_rv40_qpel16_mc11_neon, export=1
        sub             r1,  r1,  r2,  lsl #1
        sub             r1,  r1,  #2
        push            {r1, lr}
        vpush           {d8-d15}
        sub             sp,  sp,  #44*8
        add             r12, sp,  #7
        bic             r12, r12, #7
        mov             r3,  #20
        vmov.i8         d0,  #52
        vmov.i8         d1,  #20
        bl              put_rv40_qpel8_h_lp_packed_s6_neon
        ldr             r1,  [sp, #416]
        add             r1,  r1,  #8
        mov             r3,  #20
        bl              put_rv40_qpel8_h_lp_packed_s6_neon
.L\type\()_rv40_qpel16_v_s6:
        add             r1,  sp,  #7
        bic             r1,  r1,  #7
        bl              \type\()_rv40_qpel8_v_lp_packed_s6_neon
        sub             r1,  r1,  #40
        bl              \type\()_rv40_qpel8_v_lp_packed_s6_neon
        sub             r0,  r0,  r2,  lsl #4
        add             r0,  r0,  #8
        bl              \type\()_rv40_qpel8_v_lp_packed_s6_neon
        sub             r1,  r1,  #40
        bl              \type\()_rv40_qpel8_v_lp_packed_s6_neon
        add             sp,  sp,  #44*8
        vpop            {d8-d15}
        pop             {r1, pc}
endfunc

function ff_\type\()_rv40_qpel16_mc21_neon, export=1
        sub             r1,  r1,  r2,  lsl #1
        sub             r1,  r1,  #2
        push            {r1, lr}
        vpush           {d8-d15}
        sub             sp,  sp,  #44*8
        add             r12, sp,  #7
        bic             r12, r12, #7
        mov             r3,  #20
        vmov.i8         d0,  #20
        vmov.i8         d1,  #20
        bl              put_rv40_qpel8_h_lp_packed_s5_neon
        ldr             r1,  [sp, #416]
        add             r1,  r1,  #8
        mov             r3,  #20
        bl              put_rv40_qpel8_h_lp_packed_s5_neon
        vmov.i8         d0,  #52
        b               .L\type\()_rv40_qpel16_v_s6
endfunc

function ff_\type\()_rv40_qpel16_mc31_neon, export=1
        sub             r1,  r1,  r2,  lsl #1
        sub             r1,  r1,  #2
        push            {r1, lr}
        vpush           {d8-d15}
        sub             sp,  sp,  #44*8
        add             r12, sp,  #7
        bic             r12, r12, #7
        mov             r3,  #20
        vmov.i8         d0,  #20
        vmov.i8         d1,  #52
        bl              put_rv40_qpel8_h_lp_packed_s6_neon
        ldr             r1,  [sp, #416]
        add             r1,  r1,  #8
        mov             r3,  #20
        bl              put_rv40_qpel8_h_lp_packed_s6_neon
        vswp            d0,  d1
        b               .L\type\()_rv40_qpel16_v_s6
endfunc

function ff_\type\()_rv40_qpel16_mc12_neon, export=1
        sub             r1,  r1,  r2,  lsl #1
        sub             r1,  r1,  #2
        push            {r1, lr}
        vpush           {d8-d15}
        sub             sp,  sp,  #44*8
        add             r12, sp,  #7
        bic             r12, r12, #7
        mov             r3,  #20
        vmov.i8         d0,  #52
        vmov.i8         d1,  #20
        bl              put_rv40_qpel8_h_lp_packed_s6_neon
        ldr             r1,  [sp, #416]
        add             r1,  r1,  #8
        mov             r3,  #20
        bl              put_rv40_qpel8_h_lp_packed_s6_neon
        vmov.i8         d0,  #20
.L\type\()_rv40_qpel16_v_s5:
        add             r1,  sp,  #7
        bic             r1,  r1,  #7
        bl              \type\()_rv40_qpel8_v_lp_packed_s5_neon
        sub             r1,  r1,  #40
        bl              \type\()_rv40_qpel8_v_lp_packed_s5_neon
        sub             r0,  r0,  r2,  lsl #4
        add             r0,  r0,  #8
        bl              \type\()_rv40_qpel8_v_lp_packed_s5_neon
        sub             r1,  r1,  #40
        bl              \type\()_rv40_qpel8_v_lp_packed_s5_neon
        add             sp,  sp,  #44*8
        vpop            {d8-d15}
        pop             {r1, pc}
endfunc

function ff_\type\()_rv40_qpel16_mc22_neon, export=1
        sub             r1,  r1,  r2,  lsl #1
        sub             r1,  r1,  #2
        push            {r1, lr}
        vpush           {d8-d15}
        sub             sp,  sp,  #44*8
        add             r12, sp,  #7
        bic             r12, r12, #7
        mov             r3,  #20
        vmov.i8         d0,  #20
        vmov.i8         d1,  #20
        bl              put_rv40_qpel8_h_lp_packed_s5_neon
        ldr             r1,  [sp, #416]
        add             r1,  r1,  #8
        mov             r3,  #20
        bl              put_rv40_qpel8_h_lp_packed_s5_neon
        b               .L\type\()_rv40_qpel16_v_s5
endfunc

function ff_\type\()_rv40_qpel16_mc32_neon, export=1
        sub             r1,  r1,  r2,  lsl #1
        sub             r1,  r1,  #2
        push            {r1, lr}
        vpush           {d8-d15}
        sub             sp,  sp,  #44*8
        add             r12, sp,  #7
        bic             r12, r12, #7
        mov             r3,  #20
        vmov.i8         d0,  #20
        vmov.i8         d1,  #52
        bl              put_rv40_qpel8_h_lp_packed_s6_neon
        ldr             r1,  [sp, #416]
        add             r1,  r1,  #8
        mov             r3,  #20
        bl              put_rv40_qpel8_h_lp_packed_s6_neon
        vmov.i8         d1,  #20
        b               .L\type\()_rv40_qpel16_v_s5
endfunc

function ff_\type\()_rv40_qpel16_mc03_neon, export=1
        vmov.i8         d0,  #20
        vmov.i8         d1,  #52
        b               .L\type\()_rv40_qpel16_v
endfunc

function ff_\type\()_rv40_qpel16_mc13_neon, export=1
        sub             r1,  r1,  r2,  lsl #1
        sub             r1,  r1,  #2
        push            {r1, lr}
        vpush           {d8-d15}
        sub             sp,  sp,  #44*8
        add             r12, sp,  #7
        bic             r12, r12, #7
        mov             r3,  #20
        vmov.i8         d0,  #52
        vmov.i8         d1,  #20
        bl              put_rv40_qpel8_h_lp_packed_s6_neon
        ldr             r1,  [sp, #416]
        add             r1,  r1,  #8
        mov             r3,  #20
        bl              put_rv40_qpel8_h_lp_packed_s6_neon
        vswp            d0,  d1
        b               .L\type\()_rv40_qpel16_v_s6
endfunc

function ff_\type\()_rv40_qpel16_mc23_neon, export=1
        sub             r1,  r1,  r2,  lsl #1
        sub             r1,  r1,  #2
        push            {r1, lr}
        vpush           {d8-d15}
        sub             sp,  sp,  #44*8
        add             r12, sp,  #7
        bic             r12, r12, #7
        mov             r3,  #20
        vmov.i8         d0,  #20
        vmov.i8         d1,  #20
        bl              put_rv40_qpel8_h_lp_packed_s5_neon
        ldr             r1,  [sp, #416]
        add             r1,  r1,  #8
        mov             r3,  #20
        bl              put_rv40_qpel8_h_lp_packed_s5_neon
        vmov.i8         d1,  #52
        b               .L\type\()_rv40_qpel16_v_s6
endfunc

function ff_\type\()_rv40_qpel16_mc33_neon, export=1
        mov             r3,  #16
655
        b               X(ff_\type\()_pixels16_xy2_neon)
656 657 658 659 660
endfunc
.endm

        rv40_qpel       put
        rv40_qpel       avg
661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724

.macro  rv40_weight
        vmovl.u8        q8,  d2
        vmovl.u8        q9,  d3
        vmovl.u8        q10, d4
        vmovl.u8        q11, d5
        vmull.u16       q2,  d16, d0[2]
        vmull.u16       q3,  d17, d0[2]
        vmull.u16       q8,  d18, d0[2]
        vmull.u16       q9,  d19, d0[2]
        vmull.u16       q12, d20, d0[0]
        vmull.u16       q13, d21, d0[0]
        vmull.u16       q14, d22, d0[0]
        vmull.u16       q15, d23, d0[0]
        vshrn.i32       d4,  q2,  #9
        vshrn.i32       d5,  q3,  #9
        vshrn.i32       d6,  q8,  #9
        vshrn.i32       d7,  q9,  #9
        vshrn.i32       d16, q12, #9
        vshrn.i32       d17, q13, #9
        vshrn.i32       d18, q14, #9
        vshrn.i32       d19, q15, #9
        vadd.u16        q2,  q2,  q8
        vadd.u16        q3,  q3,  q9
        vrshrn.i16      d2,  q2,  #5
        vrshrn.i16      d3,  q3,  #5
.endm

/* void ff_rv40_weight_func_16_neon(uint8_t *dst, uint8_t *src1, uint8_t *src2,
                                    int w1, int w2, int stride) */
function ff_rv40_weight_func_16_neon, export=1
        ldr             r12, [sp]
        vmov            d0,  r3,  r12
        ldr             r12, [sp, #4]
        mov             r3,  #16
1:
        vld1.8          {q1},     [r1,:128], r12
        vld1.8          {q2},     [r2,:128], r12
        rv40_weight
        vst1.8          {q1},     [r0,:128], r12
        subs            r3,  r3,  #1
        bne             1b
        bx              lr
endfunc

/* void ff_rv40_weight_func_8_neon(uint8_t *dst, uint8_t *src1, uint8_t *src2,
                                   int w1, int w2, int stride) */
function ff_rv40_weight_func_8_neon, export=1
        ldr             r12, [sp]
        vmov            d0,  r3,  r12
        ldr             r12, [sp, #4]
        mov             r3,  #8
1:
        vld1.8          {d2},     [r1,:64], r12
        vld1.8          {d3},     [r1,:64], r12
        vld1.8          {d4},     [r2,:64], r12
        vld1.8          {d5},     [r2,:64], r12
        rv40_weight
        vst1.8          {d2},     [r0,:64], r12
        vst1.8          {d3},     [r0,:64], r12
        subs            r3,  r3,  #2
        bne             1b
        bx              lr
endfunc
725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810

function ff_rv40_h_loop_filter_strength_neon, export=1
        pkhbt           r2,  r3,  r2,  lsl #18

        ldr             r3,  [r0]
        ldr_dpre        r12, r0,  r1
        teq             r3,  r12
        beq             1f

        sub             r0,  r0,  r1,  lsl #1

        vld1.32         {d4[]},   [r0,:32], r1  @ -3
        vld1.32         {d0[]},   [r0,:32], r1  @ -2
        vld1.32         {d4[1]},  [r0,:32], r1  @ -1
        vld1.32         {d5[]},   [r0,:32], r1  @  0
        vld1.32         {d1[]},   [r0,:32], r1  @  1
        vld1.32         {d5[0]},  [r0,:32], r1  @  2

        vpaddl.u8       q8,  q0                 @ -2, -2, -2, -2,  1,  1,  1,  1
        vpaddl.u8       q9,  q2                 @ -3, -3, -1, -1,  2,  2,  0,  0
        vdup.32         d30, r2                 @ beta2, beta << 2
        vpadd.u16       d16, d16, d17           @ -2, -2,  1,  1
        vpadd.u16       d18, d18, d19           @ -3, -1,  2,  0
        vabd.u16        d16, d18, d16
        vclt.u16        d16, d16, d30

        ldrd            r2,  r3,  [sp, #4]
        vmovl.u16       q12, d16
        vtrn.16         d16, d17
        vshr.u32        q12, q12, #15
        ldr             r0,  [sp]
        vst1.32         {d24[1]}, [r2,:32]
        vst1.32         {d25[1]}, [r3,:32]

        cmp             r0,  #0
        it              eq
        bxeq            lr

        vand            d18, d16, d17
        vtrn.32         d18, d19
        vand            d18, d18, d19
        vmov.u16        r0,  d18[0]
        bx              lr
1:
        ldrd            r2,  r3,  [sp, #4]
        mov             r0,  #0
        str             r0,  [r2]
        str             r0,  [r3]
        bx              lr
endfunc

function ff_rv40_v_loop_filter_strength_neon, export=1
        sub             r0,  r0,  #3
        pkhbt           r2,  r3,  r2,  lsl #18

        vld1.8          {d0},     [r0], r1
        vld1.8          {d1},     [r0], r1
        vld1.8          {d2},     [r0], r1
        vld1.8          {d3},     [r0], r1

        vaddl.u8        q0,  d0,  d1
        vaddl.u8        q1,  d2,  d3
        vdup.32         q15, r2
        vadd.u16        q0,  q0,  q1            @ -3, -2, -1,  0,  1,  2
        vext.16         q1,  q0,  q0,  #1       @ -2, -1,  0,  1,  2
        vabd.u16        q0,  q1,  q0
        vclt.u16        q0,  q0,  q15

        ldrd            r2,  r3,  [sp, #4]
        vmovl.u16       q1,  d0
        vext.16         d1,  d0,  d1,  #3
        vshr.u32        q1,  q1,  #15
        ldr             r0,  [sp]
        vst1.32         {d2[1]},  [r2,:32]
        vst1.32         {d3[1]},  [r3,:32]

        cmp             r0,  #0
        it              eq
        bxeq            lr

        vand            d0,  d0,  d1
        vtrn.16         d0,  d1
        vand            d0,  d0,  d1
        vmov.u16        r0,  d0[0]
        bx              lr
endfunc
811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920

.macro  rv40_weak_loop_filter
        vdup.16         d30, r2                 @ filter_p1
        vdup.16         d31, r3                 @ filter_q1
        ldrd            r2,  r3,  [sp]
        vdup.16         d28, r2                 @ alpha
        vdup.16         d29, r3                 @ beta
        ldr             r12, [sp, #8]
        vdup.16         d25, r12                @ lim_p0q0
        ldrd            r2,  r3,  [sp, #12]
        vsubl.u8        q9,  d5,  d4            @ x, t
        vabdl.u8        q8,  d5,  d4            @ x, abs(t)
        vneg.s16        q15, q15
        vceq.i16        d16, d19, #0            @ !t
        vshl.s16        d19, d19, #2            @ t << 2
        vmul.u16        d18, d17, d28           @ alpha * abs(t)
        vand            d24, d30, d31           @ filter_p1 & filter_q1
        vsubl.u8        q1,  d0,  d4            @ p1p2, p1p0
        vsubl.u8        q3,  d1,  d5            @ q1q2, q1q0
        vmov.i16        d22, #3
        vshr.u16        d18, d18, #7
        vadd.i16        d22, d22, d24           @ 3 - (filter_p1 & filter_q1)
        vsubl.u8        q10, d0,  d1            @ src[-2] - src[1]
        vcle.u16        d18, d18, d22
        vand            d20, d20, d24
        vneg.s16        d23, d25                @ -lim_p0q0
        vadd.s16        d19, d19, d20
        vbic            d16, d18, d16           @ t && u <= 3 - (fp1 & fq1)
        vtrn.32         d4,  d5                 @ -3,  2, -1,  0
        vrshr.s16       d19, d19, #3
        vmov            d28, d29                @ beta
        vswp            d3,  d6                 @ q1q2, p1p0
        vmin.s16        d19, d19, d25
        vand            d30, d30, d16
        vand            d31, d31, d16
        vadd.s16        q10, q1,  q3            @ p1p2 + p1p0, q1q2 + q1q0
        vmax.s16        d19, d19, d23           @ diff
        vabs.s16        q1,  q1                 @ abs(p1p2), abs(q1q2)
        vand            d18, d19, d16           @ diff
        vcle.u16        q1,  q1,  q14
        vneg.s16        d19, d18                @ -diff
        vdup.16         d26, r3                 @ lim_p1
        vaddw.u8        q2,  q9,  d5            @ src[-1]+diff, src[0]-diff
        vhsub.s16       q11, q10, q9
        vand            q1,  q1,  q15
        vqmovun.s16     d4,  q2                 @ -1,  0
        vand            q9,  q11, q1
        vdup.16         d27, r2                 @ lim_q1
        vneg.s16        q9,  q9
        vneg.s16        q14, q13
        vmin.s16        q9,  q9,  q13
        vtrn.32         d0,  d1                 @ -2,  1,  -2,  1
        vmax.s16        q9,  q9,  q14
        vaddw.u8        q3,  q9,  d0
        vqmovun.s16     d5,  q3                 @ -2,  1
.endm

function ff_rv40_h_weak_loop_filter_neon, export=1
        sub             r0,  r0,  r1,  lsl #1
        sub             r0,  r0,  r1

        vld1.32         {d4[]},   [r0,:32], r1
        vld1.32         {d0[]},   [r0,:32], r1
        vld1.32         {d4[1]},  [r0,:32], r1
        vld1.32         {d5[]},   [r0,:32], r1
        vld1.32         {d1[]},   [r0,:32], r1
        vld1.32         {d5[0]},  [r0,:32]

        sub             r0,  r0,  r1,  lsl #2

        rv40_weak_loop_filter

        vst1.32         {d5[0]},  [r0,:32], r1
        vst1.32         {d4[0]},  [r0,:32], r1
        vst1.32         {d4[1]},  [r0,:32], r1
        vst1.32         {d5[1]},  [r0,:32], r1

        bx              lr
endfunc

function ff_rv40_v_weak_loop_filter_neon, export=1
        sub             r12, r0,  #3
        sub             r0,  r0,  #2

        vld1.8          {d4},     [r12], r1
        vld1.8          {d5},     [r12], r1
        vld1.8          {d2},     [r12], r1
        vld1.8          {d3},     [r12], r1

        vtrn.16         q2,  q1
        vtrn.8          d4,  d5
        vtrn.8          d2,  d3

        vrev64.32       d5,  d5
        vtrn.32         q2,  q1
        vdup.32         d0,  d3[0]
        vdup.32         d1,  d2[0]

        rv40_weak_loop_filter

        vtrn.32         q2,  q3
        vswp            d4,  d5

        vst4.8          {d4[0],d5[0],d6[0],d7[0]}, [r0], r1
        vst4.8          {d4[1],d5[1],d6[1],d7[1]}, [r0], r1
        vst4.8          {d4[2],d5[2],d6[2],d7[2]}, [r0], r1
        vst4.8          {d4[3],d5[3],d6[3],d7[3]}, [r0], r1

        bx              lr
endfunc