h264_intrapred.asm 70.5 KB
Newer Older
1 2 3
;******************************************************************************
;* H.264 intra prediction asm optimizations
;* Copyright (c) 2010 Jason Garrett-Glaser
4 5 6
;* Copyright (c) 2010 Holger Lubitz
;* Copyright (c) 2010 Loren Merritt
;* Copyright (c) 2010 Ronald S. Bultje
7
;*
8
;* This file is part of Libav.
9
;*
10
;* Libav is free software; you can redistribute it and/or
11 12 13 14
;* 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.
;*
15
;* Libav is distributed in the hope that it will be useful,
16 17 18 19 20
;* 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
21
;* License along with Libav; if not, write to the Free Software
22 23 24 25
;* 51, Inc., Foundation Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
;******************************************************************************

%include "x86inc.asm"
26
%include "x86util.asm"
27 28 29 30

SECTION_RODATA

tm_shuf: times 8 db 0x03, 0x80
31
pw_ff00: times 8 dw 0xff00
32 33 34 35 36 37 38 39
plane_shuf:  db -8, -7, -6, -5, -4, -3, -2, -1
             db  1,  2,  3,  4,  5,  6,  7,  8
plane8_shuf: db -4, -3, -2, -1,  0,  0,  0,  0
             db  1,  2,  3,  4,  0,  0,  0,  0
pw_0to7:     dw  0,  1,  2,  3,  4,  5,  6,  7
pw_1to8:     dw  1,  2,  3,  4,  5,  6,  7,  8
pw_m8tom1:   dw -8, -7, -6, -5, -4, -3, -2, -1
pw_m4to4:    dw -4, -3, -2, -1,  1,  2,  3,  4
40 41 42

SECTION .text

43
cextern pb_1
44
cextern pb_3
45
cextern pw_4
46
cextern pw_5
47
cextern pw_8
48 49 50
cextern pw_16
cextern pw_17
cextern pw_32
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

;-----------------------------------------------------------------------------
; void pred16x16_vertical(uint8_t *src, int stride)
;-----------------------------------------------------------------------------

cglobal pred16x16_vertical_mmx, 2,3
    sub   r0, r1
    mov   r2, 8
    movq mm0, [r0+0]
    movq mm1, [r0+8]
.loop:
    movq [r0+r1*1+0], mm0
    movq [r0+r1*1+8], mm1
    movq [r0+r1*2+0], mm0
    movq [r0+r1*2+8], mm1
    lea   r0, [r0+r1*2]
    dec   r2
    jg .loop
    REP_RET

cglobal pred16x16_vertical_sse, 2,3
    sub   r0, r1
    mov   r2, 4
    movaps xmm0, [r0]
.loop:
    movaps [r0+r1*1], xmm0
    movaps [r0+r1*2], xmm0
    lea   r0, [r0+r1*2]
    movaps [r0+r1*1], xmm0
    movaps [r0+r1*2], xmm0
    lea   r0, [r0+r1*2]
    dec   r2
    jg .loop
    REP_RET

;-----------------------------------------------------------------------------
; void pred16x16_horizontal(uint8_t *src, int stride)
;-----------------------------------------------------------------------------

%macro PRED16x16_H 1
cglobal pred16x16_horizontal_%1, 2,3
    mov       r2, 8
%ifidn %1, ssse3
    mova      m2, [pb_3]
%endif
.loop:
    movd      m0, [r0+r1*0-4]
    movd      m1, [r0+r1*1-4]

%ifidn %1, ssse3
    pshufb    m0, m2
    pshufb    m1, m2
%else
    punpcklbw m0, m0
    punpcklbw m1, m1
%ifidn %1, mmxext
    pshufw    m0, m0, 0xff
    pshufw    m1, m1, 0xff
%else
    punpckhwd m0, m0
    punpckhwd m1, m1
    punpckhdq m0, m0
    punpckhdq m1, m1
%endif
    mova [r0+r1*0+8], m0
    mova [r0+r1*1+8], m1
%endif

    mova [r0+r1*0], m0
    mova [r0+r1*1], m1
    lea       r0, [r0+r1*2]
    dec       r2
    jg .loop
    REP_RET
%endmacro

INIT_MMX
PRED16x16_H mmx
PRED16x16_H mmxext
INIT_XMM
PRED16x16_H ssse3

;-----------------------------------------------------------------------------
; void pred16x16_dc(uint8_t *src, int stride)
;-----------------------------------------------------------------------------

137
%macro PRED16x16_DC 1
138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160
cglobal pred16x16_dc_%1, 2,7
    mov       r4, r0
    sub       r0, r1
    pxor      mm0, mm0
    pxor      mm1, mm1
    psadbw    mm0, [r0+0]
    psadbw    mm1, [r0+8]
    dec        r0
    movzx     r5d, byte [r0+r1*1]
    paddw     mm0, mm1
    movd      r6d, mm0
    lea        r0, [r0+r1*2]
%rep 7
    movzx     r2d, byte [r0+r1*0]
    movzx     r3d, byte [r0+r1*1]
    add       r5d, r2d
    add       r6d, r3d
    lea        r0, [r0+r1*2]
%endrep
    movzx     r2d, byte [r0+r1*0]
    add       r5d, r6d
    lea       r2d, [r2+r5+16]
    shr       r2d, 5
161
%ifidn %1, mmxext
162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178
    movd       m0, r2d
    punpcklbw  m0, m0
    pshufw     m0, m0, 0
%elifidn %1, sse2
    movd       m0, r2d
    punpcklbw  m0, m0
    pshuflw    m0, m0, 0
    punpcklqdq m0, m0
%elifidn %1, ssse3
    pxor       m1, m1
    movd       m0, r2d
    pshufb     m0, m1
%endif

%if mmsize==8
    mov       r3d, 8
.loop:
179 180 181 182
    mova [r4+r1*0+0], m0
    mova [r4+r1*0+8], m0
    mova [r4+r1*1+0], m0
    mova [r4+r1*1+8], m0
183 184 185
%else
    mov       r3d, 4
.loop:
186 187
    mova [r4+r1*0], m0
    mova [r4+r1*1], m0
188
    lea   r4, [r4+r1*2]
189 190
    mova [r4+r1*0], m0
    mova [r4+r1*1], m0
191 192 193 194 195 196 197 198
%endif
    lea   r4, [r4+r1*2]
    dec   r3d
    jg .loop
    REP_RET
%endmacro

INIT_MMX
199
PRED16x16_DC mmxext
200
INIT_XMM
201 202
PRED16x16_DC   sse2
PRED16x16_DC  ssse3
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

;-----------------------------------------------------------------------------
; void pred16x16_tm_vp8(uint8_t *src, int stride)
;-----------------------------------------------------------------------------

%macro PRED16x16_TM_MMX 1
cglobal pred16x16_tm_vp8_%1, 2,5
    sub        r0, r1
    pxor      mm7, mm7
    movq      mm0, [r0+0]
    movq      mm2, [r0+8]
    movq      mm1, mm0
    movq      mm3, mm2
    punpcklbw mm0, mm7
    punpckhbw mm1, mm7
    punpcklbw mm2, mm7
    punpckhbw mm3, mm7
    movzx     r3d, byte [r0-1]
    mov       r4d, 16
.loop:
    movzx     r2d, byte [r0+r1-1]
    sub       r2d, r3d
    movd      mm4, r2d
%ifidn %1, mmx
    punpcklwd mm4, mm4
    punpckldq mm4, mm4
%else
    pshufw    mm4, mm4, 0
%endif
    movq      mm5, mm4
    movq      mm6, mm4
    movq      mm7, mm4
    paddw     mm4, mm0
    paddw     mm5, mm1
    paddw     mm6, mm2
    paddw     mm7, mm3
    packuswb  mm4, mm5
    packuswb  mm6, mm7
    movq [r0+r1+0], mm4
    movq [r0+r1+8], mm6
    add        r0, r1
    dec       r4d
    jg .loop
    REP_RET
%endmacro

PRED16x16_TM_MMX mmx
PRED16x16_TM_MMX mmxext

cglobal pred16x16_tm_vp8_sse2, 2,6,6
    sub          r0, r1
    pxor       xmm2, xmm2
    movdqa     xmm0, [r0]
    movdqa     xmm1, xmm0
    punpcklbw  xmm0, xmm2
    punpckhbw  xmm1, xmm2
    movzx       r4d, byte [r0-1]
    mov         r5d, 8
.loop:
    movzx       r2d, byte [r0+r1*1-1]
    movzx       r3d, byte [r0+r1*2-1]
    sub         r2d, r4d
    sub         r3d, r4d
    movd       xmm2, r2d
    movd       xmm4, r3d
    pshuflw    xmm2, xmm2, 0
    pshuflw    xmm4, xmm4, 0
    punpcklqdq xmm2, xmm2
    punpcklqdq xmm4, xmm4
    movdqa     xmm3, xmm2
    movdqa     xmm5, xmm4
    paddw      xmm2, xmm0
    paddw      xmm3, xmm1
    paddw      xmm4, xmm0
    paddw      xmm5, xmm1
    packuswb   xmm2, xmm3
    packuswb   xmm4, xmm5
    movdqa [r0+r1*1], xmm2
    movdqa [r0+r1*2], xmm4
    lea          r0, [r0+r1*2]
    dec         r5d
    jg .loop
    REP_RET

287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448
;-----------------------------------------------------------------------------
; void pred16x16_plane(uint8_t *src, int stride)
;-----------------------------------------------------------------------------

%macro H264_PRED16x16_PLANE 3
cglobal pred16x16_plane_%3_%1, 2, 7, %2
    mov          r2, r1           ; +stride
    neg          r1               ; -stride

    movh         m0, [r0+r1  -1]
%if mmsize == 8
    pxor         m4, m4
    movh         m1, [r0+r1  +3 ]
    movh         m2, [r0+r1  +8 ]
    movh         m3, [r0+r1  +12]
    punpcklbw    m0, m4
    punpcklbw    m1, m4
    punpcklbw    m2, m4
    punpcklbw    m3, m4
    pmullw       m0, [pw_m8tom1  ]
    pmullw       m1, [pw_m8tom1+8]
    pmullw       m2, [pw_1to8    ]
    pmullw       m3, [pw_1to8  +8]
    paddw        m0, m2
    paddw        m1, m3
%else ; mmsize == 16
%ifidn %1, sse2
    pxor         m2, m2
    movh         m1, [r0+r1  +8]
    punpcklbw    m0, m2
    punpcklbw    m1, m2
    pmullw       m0, [pw_m8tom1]
    pmullw       m1, [pw_1to8]
    paddw        m0, m1
%else ; ssse3
    movhps       m0, [r0+r1  +8]
    pmaddubsw    m0, [plane_shuf] ; H coefficients
%endif
    movhlps      m1, m0
%endif
    paddw        m0, m1
%ifidn %1, mmx
    mova         m1, m0
    psrlq        m1, 32
%elifidn %1, mmx2
    pshufw       m1, m0, 0xE
%else ; mmsize == 16
    pshuflw      m1, m0, 0xE
%endif
    paddw        m0, m1
%ifidn %1, mmx
    mova         m1, m0
    psrlq        m1, 16
%elifidn %1, mmx2
    pshufw       m1, m0, 0x1
%else
    pshuflw      m1, m0, 0x1
%endif
    paddw        m0, m1           ; sum of H coefficients

    lea          r4, [r0+r2*8-1]
    lea          r3, [r0+r2*4-1]
    add          r4, r2

%ifdef ARCH_X86_64
%define e_reg r11
%else
%define e_reg r0
%endif

    movzx     e_reg, byte [r3+r2*2   ]
    movzx        r5, byte [r4+r1     ]
    sub          r5, e_reg

    movzx     e_reg, byte [r3+r2     ]
    movzx        r6, byte [r4        ]
    sub          r6, e_reg
    lea          r5, [r5+r6*2]

    movzx     e_reg, byte [r3+r1     ]
    movzx        r6, byte [r4+r2*2   ]
    sub          r6, e_reg
    lea          r5, [r5+r6*4]

    movzx     e_reg, byte [r3        ]
%ifdef ARCH_X86_64
    movzx       r10, byte [r4+r2     ]
    sub         r10, e_reg
%else
    movzx        r6, byte [r4+r2     ]
    sub          r6, e_reg
    lea          r5, [r5+r6*4]
    sub          r5, r6
%endif

    lea       e_reg, [r3+r1*4]
    lea          r3, [r4+r2*4]

    movzx        r4, byte [e_reg+r2  ]
    movzx        r6, byte [r3        ]
    sub          r6, r4
%ifdef ARCH_X86_64
    lea          r6, [r10+r6*2]
    lea          r5, [r5+r6*2]
    add          r5, r6
%else
    lea          r5, [r5+r6*4]
    lea          r5, [r5+r6*2]
%endif

    movzx        r4, byte [e_reg     ]
%ifdef ARCH_X86_64
    movzx       r10, byte [r3   +r2  ]
    sub         r10, r4
    sub          r5, r10
%else
    movzx        r6, byte [r3   +r2  ]
    sub          r6, r4
    lea          r5, [r5+r6*8]
    sub          r5, r6
%endif

    movzx        r4, byte [e_reg+r1  ]
    movzx        r6, byte [r3   +r2*2]
    sub          r6, r4
%ifdef ARCH_X86_64
    add          r6, r10
%endif
    lea          r5, [r5+r6*8]

    movzx        r4, byte [e_reg+r2*2]
    movzx        r6, byte [r3   +r1  ]
    sub          r6, r4
    lea          r5, [r5+r6*4]
    add          r5, r6           ; sum of V coefficients

%ifndef ARCH_X86_64
    mov          r0, r0m
%endif

%ifidn %3, h264
    lea          r5, [r5*5+32]
    sar          r5, 6
%elifidn %3, rv40
    lea          r5, [r5*5]
    sar          r5, 6
%elifidn %3, svq3
    test         r5, r5
    lea          r6, [r5+3]
    cmovs        r5, r6
    sar          r5, 2            ; V/4
    lea          r5, [r5*5]       ; 5*(V/4)
    test         r5, r5
    lea          r6, [r5+15]
    cmovs        r5, r6
    sar          r5, 4            ; (5*(V/4))/16
%endif

    movzx        r4, byte [r0+r1  +15]
    movzx        r3, byte [r3+r2*2   ]
    lea          r3, [r3+r4+1]
    shl          r3, 4
449

450 451
    movd        r1d, m0
    movsx       r1d, r1w
452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471
%ifnidn %3, svq3
%ifidn %3, h264
    lea         r1d, [r1d*5+32]
%else ; rv40
    lea         r1d, [r1d*5]
%endif
    sar         r1d, 6
%else ; svq3
    test        r1d, r1d
    lea         r4d, [r1d+3]
    cmovs       r1d, r4d
    sar         r1d, 2           ; H/4
    lea         r1d, [r1d*5]     ; 5*(H/4)
    test        r1d, r1d
    lea         r4d, [r1d+15]
    cmovs       r1d, r4d
    sar         r1d, 4           ; (5*(H/4))/16
%endif
    movd         m0, r1d

472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692
    add         r1d, r5d
    add         r3d, r1d
    shl         r1d, 3
    sub         r3d, r1d          ; a

    movd         m1, r5d
    movd         m3, r3d
%ifidn %1, mmx
    punpcklwd    m0, m0
    punpcklwd    m1, m1
    punpcklwd    m3, m3
    punpckldq    m0, m0
    punpckldq    m1, m1
    punpckldq    m3, m3
%elifidn %1, mmx2
    pshufw       m0, m0, 0x0
    pshufw       m1, m1, 0x0
    pshufw       m3, m3, 0x0
%else
    pshuflw      m0, m0, 0x0
    pshuflw      m1, m1, 0x0
    pshuflw      m3, m3, 0x0
    punpcklqdq   m0, m0           ; splat H (words)
    punpcklqdq   m1, m1           ; splat V (words)
    punpcklqdq   m3, m3           ; splat a (words)
%endif
%ifidn %3, svq3
    SWAP          0, 1
%endif
    mova         m2, m0
%if mmsize == 8
    mova         m5, m0
%endif
    pmullw       m0, [pw_0to7]    ; 0*H, 1*H, ..., 7*H  (words)
%if mmsize == 16
    psllw        m2, 3
%else
    psllw        m5, 3
    psllw        m2, 2
    mova         m6, m5
    paddw        m6, m2
%endif
    paddw        m0, m3           ; a + {0,1,2,3,4,5,6,7}*H
    paddw        m2, m0           ; a + {8,9,10,11,12,13,14,15}*H
%if mmsize == 8
    paddw        m5, m0           ; a + {8,9,10,11}*H
    paddw        m6, m0           ; a + {12,13,14,15}*H
%endif

    mov          r4, 8
.loop
    mova         m3, m0           ; b[0..7]
    mova         m4, m2           ; b[8..15]
    psraw        m3, 5
    psraw        m4, 5
    packuswb     m3, m4
    mova       [r0], m3
%if mmsize == 8
    mova         m3, m5           ; b[8..11]
    mova         m4, m6           ; b[12..15]
    psraw        m3, 5
    psraw        m4, 5
    packuswb     m3, m4
    mova     [r0+8], m3
%endif
    paddw        m0, m1
    paddw        m2, m1
%if mmsize == 8
    paddw        m5, m1
    paddw        m6, m1
%endif

    mova         m3, m0           ; b[0..7]
    mova         m4, m2           ; b[8..15]
    psraw        m3, 5
    psraw        m4, 5
    packuswb     m3, m4
    mova    [r0+r2], m3
%if mmsize == 8
    mova         m3, m5           ; b[8..11]
    mova         m4, m6           ; b[12..15]
    psraw        m3, 5
    psraw        m4, 5
    packuswb     m3, m4
    mova  [r0+r2+8], m3
%endif
    paddw        m0, m1
    paddw        m2, m1
%if mmsize == 8
    paddw        m5, m1
    paddw        m6, m1
%endif

    lea          r0, [r0+r2*2]
    dec          r4
    jg .loop
    REP_RET
%endmacro

INIT_MMX
H264_PRED16x16_PLANE mmx,   0, h264
H264_PRED16x16_PLANE mmx,   0, rv40
H264_PRED16x16_PLANE mmx,   0, svq3
H264_PRED16x16_PLANE mmx2,  0, h264
H264_PRED16x16_PLANE mmx2,  0, rv40
H264_PRED16x16_PLANE mmx2,  0, svq3
INIT_XMM
H264_PRED16x16_PLANE sse2,  8, h264
H264_PRED16x16_PLANE sse2,  8, rv40
H264_PRED16x16_PLANE sse2,  8, svq3
H264_PRED16x16_PLANE ssse3, 8, h264
H264_PRED16x16_PLANE ssse3, 8, rv40
H264_PRED16x16_PLANE ssse3, 8, svq3

;-----------------------------------------------------------------------------
; void pred8x8_plane(uint8_t *src, int stride)
;-----------------------------------------------------------------------------

%macro H264_PRED8x8_PLANE 2
cglobal pred8x8_plane_%1, 2, 7, %2
    mov          r2, r1           ; +stride
    neg          r1               ; -stride

    movd         m0, [r0+r1  -1]
%if mmsize == 8
    pxor         m2, m2
    movh         m1, [r0+r1  +4 ]
    punpcklbw    m0, m2
    punpcklbw    m1, m2
    pmullw       m0, [pw_m4to4]
    pmullw       m1, [pw_m4to4+8]
%else ; mmsize == 16
%ifidn %1, sse2
    pxor         m2, m2
    movd         m1, [r0+r1  +4]
    punpckldq    m0, m1
    punpcklbw    m0, m2
    pmullw       m0, [pw_m4to4]
%else ; ssse3
    movhps       m0, [r0+r1  +4]   ; this reads 4 bytes more than necessary
    pmaddubsw    m0, [plane8_shuf] ; H coefficients
%endif
    movhlps      m1, m0
%endif
    paddw        m0, m1

%ifnidn %1, ssse3
%ifidn %1, mmx
    mova         m1, m0
    psrlq        m1, 32
%elifidn %1, mmx2
    pshufw       m1, m0, 0xE
%else ; mmsize == 16
    pshuflw      m1, m0, 0xE
%endif
    paddw        m0, m1
%endif ; !ssse3

%ifidn %1, mmx
    mova         m1, m0
    psrlq        m1, 16
%elifidn %1, mmx2
    pshufw       m1, m0, 0x1
%else
    pshuflw      m1, m0, 0x1
%endif
    paddw        m0, m1           ; sum of H coefficients

    lea          r4, [r0+r2*4-1]
    lea          r3, [r0     -1]
    add          r4, r2

%ifdef ARCH_X86_64
%define e_reg r11
%else
%define e_reg r0
%endif

    movzx     e_reg, byte [r3+r2*2   ]
    movzx        r5, byte [r4+r1     ]
    sub          r5, e_reg

    movzx     e_reg, byte [r3        ]
%ifdef ARCH_X86_64
    movzx       r10, byte [r4+r2     ]
    sub         r10, e_reg
    sub          r5, r10
%else
    movzx        r6, byte [r4+r2     ]
    sub          r6, e_reg
    lea          r5, [r5+r6*4]
    sub          r5, r6
%endif

    movzx     e_reg, byte [r3+r1     ]
    movzx        r6, byte [r4+r2*2   ]
    sub          r6, e_reg
%ifdef ARCH_X86_64
    add          r6, r10
%endif
    lea          r5, [r5+r6*4]

    movzx     e_reg, byte [r3+r2     ]
    movzx        r6, byte [r4        ]
    sub          r6, e_reg
    lea          r6, [r5+r6*2]

    lea          r5, [r6*9+16]
    lea          r5, [r5+r6*8]
    sar          r5, 5

%ifndef ARCH_X86_64
    mov          r0, r0m
%endif

    movzx        r3, byte [r4+r2*2  ]
    movzx        r4, byte [r0+r1  +7]
    lea          r3, [r3+r4+1]
    shl          r3, 4
    movd        r1d, m0
    movsx       r1d, r1w
693 694 695 696
    imul        r1d, 17
    add         r1d, 16
    sar         r1d, 5
    movd         m0, r1d
697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777
    add         r1d, r5d
    sub         r3d, r1d
    add         r1d, r1d
    sub         r3d, r1d          ; a

    movd         m1, r5d
    movd         m3, r3d
%ifidn %1, mmx
    punpcklwd    m0, m0
    punpcklwd    m1, m1
    punpcklwd    m3, m3
    punpckldq    m0, m0
    punpckldq    m1, m1
    punpckldq    m3, m3
%elifidn %1, mmx2
    pshufw       m0, m0, 0x0
    pshufw       m1, m1, 0x0
    pshufw       m3, m3, 0x0
%else
    pshuflw      m0, m0, 0x0
    pshuflw      m1, m1, 0x0
    pshuflw      m3, m3, 0x0
    punpcklqdq   m0, m0           ; splat H (words)
    punpcklqdq   m1, m1           ; splat V (words)
    punpcklqdq   m3, m3           ; splat a (words)
%endif
%if mmsize == 8
    mova         m2, m0
%endif
    pmullw       m0, [pw_0to7]    ; 0*H, 1*H, ..., 7*H  (words)
    paddw        m0, m3           ; a + {0,1,2,3,4,5,6,7}*H
%if mmsize == 8
    psllw        m2, 2
    paddw        m2, m0           ; a + {4,5,6,7}*H
%endif

    mov          r4, 4
ALIGN 16
.loop
%if mmsize == 16
    mova         m3, m0           ; b[0..7]
    paddw        m0, m1
    psraw        m3, 5
    mova         m4, m0           ; V+b[0..7]
    paddw        m0, m1
    psraw        m4, 5
    packuswb     m3, m4
    movh       [r0], m3
    movhps  [r0+r2], m3
%else ; mmsize == 8
    mova         m3, m0           ; b[0..3]
    mova         m4, m2           ; b[4..7]
    paddw        m0, m1
    paddw        m2, m1
    psraw        m3, 5
    psraw        m4, 5
    mova         m5, m0           ; V+b[0..3]
    mova         m6, m2           ; V+b[4..7]
    paddw        m0, m1
    paddw        m2, m1
    psraw        m5, 5
    psraw        m6, 5
    packuswb     m3, m4
    packuswb     m5, m6
    mova       [r0], m3
    mova    [r0+r2], m5
%endif

    lea          r0, [r0+r2*2]
    dec          r4
    jg .loop
    REP_RET
%endmacro

INIT_MMX
H264_PRED8x8_PLANE mmx,   0
H264_PRED8x8_PLANE mmx2,  0
INIT_XMM
H264_PRED8x8_PLANE sse2,  8
H264_PRED8x8_PLANE ssse3, 8

778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835
;-----------------------------------------------------------------------------
; void pred8x8_vertical(uint8_t *src, int stride)
;-----------------------------------------------------------------------------

cglobal pred8x8_vertical_mmx, 2,2
    sub    r0, r1
    movq  mm0, [r0]
%rep 3
    movq [r0+r1*1], mm0
    movq [r0+r1*2], mm0
    lea    r0, [r0+r1*2]
%endrep
    movq [r0+r1*1], mm0
    movq [r0+r1*2], mm0
    RET

;-----------------------------------------------------------------------------
; void pred8x8_horizontal(uint8_t *src, int stride)
;-----------------------------------------------------------------------------

%macro PRED8x8_H 1
cglobal pred8x8_horizontal_%1, 2,3
    mov       r2, 4
%ifidn %1, ssse3
    mova      m2, [pb_3]
%endif
.loop:
    movd      m0, [r0+r1*0-4]
    movd      m1, [r0+r1*1-4]
%ifidn %1, ssse3
    pshufb    m0, m2
    pshufb    m1, m2
%else
    punpcklbw m0, m0
    punpcklbw m1, m1
%ifidn %1, mmxext
    pshufw    m0, m0, 0xff
    pshufw    m1, m1, 0xff
%else
    punpckhwd m0, m0
    punpckhwd m1, m1
    punpckhdq m0, m0
    punpckhdq m1, m1
%endif
%endif
    mova [r0+r1*0], m0
    mova [r0+r1*1], m1
    lea       r0, [r0+r1*2]
    dec       r2
    jg .loop
    REP_RET
%endmacro

INIT_MMX
PRED8x8_H mmx
PRED8x8_H mmxext
PRED8x8_H ssse3

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
;-----------------------------------------------------------------------------
; void pred8x8_top_dc_mmxext(uint8_t *src, int stride)
;-----------------------------------------------------------------------------
%ifdef CONFIG_GPL
cglobal pred8x8_top_dc_mmxext, 2,5
    sub         r0, r1
    movq       mm0, [r0]
    pxor       mm1, mm1
    pxor       mm2, mm2
    lea         r2, [r0+r1*2]
    punpckhbw  mm1, mm0
    punpcklbw  mm0, mm2
    psadbw     mm1, mm2        ; s1
    lea         r3, [r2+r1*2]
    psadbw     mm0, mm2        ; s0
    psrlw      mm1, 1
    psrlw      mm0, 1
    pavgw      mm1, mm2
    lea         r4, [r3+r1*2]
    pavgw      mm0, mm2
    pshufw     mm1, mm1, 0
    pshufw     mm0, mm0, 0     ; dc0 (w)
    packuswb   mm0, mm1        ; dc0,dc1 (b)
    movq [r0+r1*1], mm0
    movq [r0+r1*2], mm0
    lea         r0, [r3+r1*2]
    movq [r2+r1*1], mm0
    movq [r2+r1*2], mm0
    movq [r3+r1*1], mm0
    movq [r3+r1*2], mm0
    movq [r0+r1*1], mm0
    movq [r0+r1*2], mm0
    RET

870 871 872
;-----------------------------------------------------------------------------
; void pred8x8_dc_mmxext(uint8_t *src, int stride)
;-----------------------------------------------------------------------------
873

874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931
INIT_MMX
cglobal pred8x8_dc_mmxext, 2,5
    sub       r0, r1
    pxor      m7, m7
    movd      m0, [r0+0]
    movd      m1, [r0+4]
    psadbw    m0, m7            ; s0
    mov       r4, r0
    psadbw    m1, m7            ; s1

    movzx    r2d, byte [r0+r1*1-1]
    movzx    r3d, byte [r0+r1*2-1]
    lea       r0, [r0+r1*2]
    add      r2d, r3d
    movzx    r3d, byte [r0+r1*1-1]
    add      r2d, r3d
    movzx    r3d, byte [r0+r1*2-1]
    add      r2d, r3d
    lea       r0, [r0+r1*2]
    movd      m2, r2d            ; s2
    movzx    r2d, byte [r0+r1*1-1]
    movzx    r3d, byte [r0+r1*2-1]
    lea       r0, [r0+r1*2]
    add      r2d, r3d
    movzx    r3d, byte [r0+r1*1-1]
    add      r2d, r3d
    movzx    r3d, byte [r0+r1*2-1]
    add      r2d, r3d
    movd      m3, r2d            ; s3

    punpcklwd m0, m1
    mov       r0, r4
    punpcklwd m2, m3
    punpckldq m0, m2            ; s0, s1, s2, s3
    pshufw    m3, m0, 11110110b ; s2, s1, s3, s3
    lea       r2, [r0+r1*2]
    pshufw    m0, m0, 01110100b ; s0, s1, s3, s1
    paddw     m0, m3
    lea       r3, [r2+r1*2]
    psrlw     m0, 2
    pavgw     m0, m7            ; s0+s2, s1, s3, s1+s3
    lea       r4, [r3+r1*2]
    packuswb  m0, m0
    punpcklbw m0, m0
    movq      m1, m0
    punpcklbw m0, m0
    punpckhbw m1, m1
    movq [r0+r1*1], m0
    movq [r0+r1*2], m0
    movq [r2+r1*1], m0
    movq [r2+r1*2], m0
    movq [r3+r1*1], m1
    movq [r3+r1*2], m1
    movq [r4+r1*1], m1
    movq [r4+r1*2], m1
    RET
%endif

932 933 934 935
;-----------------------------------------------------------------------------
; void pred8x8_dc_rv40(uint8_t *src, int stride)
;-----------------------------------------------------------------------------

936
cglobal pred8x8_dc_rv40_mmxext, 2,7
937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069
    mov       r4, r0
    sub       r0, r1
    pxor      mm0, mm0
    psadbw    mm0, [r0]
    dec        r0
    movzx     r5d, byte [r0+r1*1]
    movd      r6d, mm0
    lea        r0, [r0+r1*2]
%rep 3
    movzx     r2d, byte [r0+r1*0]
    movzx     r3d, byte [r0+r1*1]
    add       r5d, r2d
    add       r6d, r3d
    lea        r0, [r0+r1*2]
%endrep
    movzx     r2d, byte [r0+r1*0]
    add       r5d, r6d
    lea       r2d, [r2+r5+8]
    shr       r2d, 4
    movd      mm0, r2d
    punpcklbw mm0, mm0
    pshufw    mm0, mm0, 0
    mov       r3d, 4
.loop:
    movq [r4+r1*0], mm0
    movq [r4+r1*1], mm0
    lea   r4, [r4+r1*2]
    dec   r3d
    jg .loop
    REP_RET

;-----------------------------------------------------------------------------
; void pred8x8_tm_vp8(uint8_t *src, int stride)
;-----------------------------------------------------------------------------

%macro PRED8x8_TM_MMX 1
cglobal pred8x8_tm_vp8_%1, 2,6
    sub        r0, r1
    pxor      mm7, mm7
    movq      mm0, [r0]
    movq      mm1, mm0
    punpcklbw mm0, mm7
    punpckhbw mm1, mm7
    movzx     r4d, byte [r0-1]
    mov       r5d, 4
.loop:
    movzx     r2d, byte [r0+r1*1-1]
    movzx     r3d, byte [r0+r1*2-1]
    sub       r2d, r4d
    sub       r3d, r4d
    movd      mm2, r2d
    movd      mm4, r3d
%ifidn %1, mmx
    punpcklwd mm2, mm2
    punpcklwd mm4, mm4
    punpckldq mm2, mm2
    punpckldq mm4, mm4
%else
    pshufw    mm2, mm2, 0
    pshufw    mm4, mm4, 0
%endif
    movq      mm3, mm2
    movq      mm5, mm4
    paddw     mm2, mm0
    paddw     mm3, mm1
    paddw     mm4, mm0
    paddw     mm5, mm1
    packuswb  mm2, mm3
    packuswb  mm4, mm5
    movq [r0+r1*1], mm2
    movq [r0+r1*2], mm4
    lea        r0, [r0+r1*2]
    dec       r5d
    jg .loop
    REP_RET
%endmacro

PRED8x8_TM_MMX mmx
PRED8x8_TM_MMX mmxext

cglobal pred8x8_tm_vp8_sse2, 2,6,4
    sub          r0, r1
    pxor       xmm1, xmm1
    movq       xmm0, [r0]
    punpcklbw  xmm0, xmm1
    movzx       r4d, byte [r0-1]
    mov         r5d, 4
.loop:
    movzx       r2d, byte [r0+r1*1-1]
    movzx       r3d, byte [r0+r1*2-1]
    sub         r2d, r4d
    sub         r3d, r4d
    movd       xmm2, r2d
    movd       xmm3, r3d
    pshuflw    xmm2, xmm2, 0
    pshuflw    xmm3, xmm3, 0
    punpcklqdq xmm2, xmm2
    punpcklqdq xmm3, xmm3
    paddw      xmm2, xmm0
    paddw      xmm3, xmm0
    packuswb   xmm2, xmm3
    movq   [r0+r1*1], xmm2
    movhps [r0+r1*2], xmm2
    lea          r0, [r0+r1*2]
    dec         r5d
    jg .loop
    REP_RET

cglobal pred8x8_tm_vp8_ssse3, 2,3,6
    sub          r0, r1
    movdqa     xmm4, [tm_shuf]
    pxor       xmm1, xmm1
    movq       xmm0, [r0]
    punpcklbw  xmm0, xmm1
    movd       xmm5, [r0-4]
    pshufb     xmm5, xmm4
    mov         r2d, 4
.loop:
    movd       xmm2, [r0+r1*1-4]
    movd       xmm3, [r0+r1*2-4]
    pshufb     xmm2, xmm4
    pshufb     xmm3, xmm4
    psubw      xmm2, xmm5
    psubw      xmm3, xmm5
    paddw      xmm2, xmm0
    paddw      xmm3, xmm0
    packuswb   xmm2, xmm3
    movq   [r0+r1*1], xmm2
    movhps [r0+r1*2], xmm2
    lea          r0, [r0+r1*2]
    dec         r2d
    jg .loop
    REP_RET
1070

1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082
; dest, left, right, src, tmp
; output: %1 = (t[n-1] + t[n]*2 + t[n+1] + 2) >> 2
%macro PRED4x4_LOWPASS 5
    mova    %5, %2
    pavgb   %2, %3
    pxor    %3, %5
    mova    %1, %4
    pand    %3, [pb_1]
    psubusb %2, %3
    pavgb   %1, %2
%endmacro

1083
;-----------------------------------------------------------------------------
1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140
; void pred8x8l_top_dc(uint8_t *src, int has_topleft, int has_topright, int stride)
;-----------------------------------------------------------------------------
%ifdef CONFIG_GPL
%macro PRED8x8L_TOP_DC 1
cglobal pred8x8l_top_dc_%1, 4,4
    sub          r0, r3
    pxor        mm7, mm7
    movq        mm0, [r0-8]
    movq        mm3, [r0]
    movq        mm1, [r0+8]
    movq        mm2, mm3
    movq        mm4, mm3
    PALIGNR     mm2, mm0, 7, mm0
    PALIGNR     mm1, mm4, 1, mm4
    test         r1, r1 ; top_left
    jz .fix_lt_2
    test         r2, r2 ; top_right
    jz .fix_tr_1
    jmp .body
.fix_lt_2:
    movq        mm5, mm3
    pxor        mm5, mm2
    psllq       mm5, 56
    psrlq       mm5, 56
    pxor        mm2, mm5
    test         r2, r2 ; top_right
    jnz .body
.fix_tr_1:
    movq        mm5, mm3
    pxor        mm5, mm1
    psrlq       mm5, 56
    psllq       mm5, 56
    pxor        mm1, mm5
.body
    PRED4x4_LOWPASS mm0, mm2, mm1, mm3, mm5
    psadbw   mm7, mm0
    paddw    mm7, [pw_4]
    psrlw    mm7, 3
    pshufw   mm7, mm7, 0
    packuswb mm7, mm7
%rep 3
    movq [r0+r3*1], mm7
    movq [r0+r3*2], mm7
    lea    r0, [r0+r3*2]
%endrep
    movq [r0+r3*1], mm7
    movq [r0+r3*2], mm7
    RET
%endmacro

INIT_MMX
%define PALIGNR PALIGNR_MMX
PRED8x8L_TOP_DC mmxext
%define PALIGNR PALIGNR_SSSE3
PRED8x8L_TOP_DC ssse3

;-----------------------------------------------------------------------------
1141 1142
;void pred8x8l_dc(uint8_t *src, int has_topleft, int has_topright, int stride)
;-----------------------------------------------------------------------------
1143

1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243
%macro PRED8x8L_DC 1
cglobal pred8x8l_dc_%1, 4,5
    sub          r0, r3
    lea          r4, [r0+r3*2]
    movq        mm0, [r0+r3*1-8]
    punpckhbw   mm0, [r0+r3*0-8]
    movq        mm1, [r4+r3*1-8]
    punpckhbw   mm1, [r0+r3*2-8]
    mov          r4, r0
    punpckhwd   mm1, mm0
    lea          r0, [r0+r3*4]
    movq        mm2, [r0+r3*1-8]
    punpckhbw   mm2, [r0+r3*0-8]
    lea          r0, [r0+r3*2]
    movq        mm3, [r0+r3*1-8]
    punpckhbw   mm3, [r0+r3*0-8]
    punpckhwd   mm3, mm2
    punpckhdq   mm3, mm1
    lea          r0, [r0+r3*2]
    movq        mm0, [r0+r3*0-8]
    movq        mm1, [r4]
    mov          r0, r4
    movq        mm4, mm3
    movq        mm2, mm3
    PALIGNR     mm4, mm0, 7, mm0
    PALIGNR     mm1, mm2, 1, mm2
    test        r1, r1
    jnz .do_left
.fix_lt_1:
    movq        mm5, mm3
    pxor        mm5, mm4
    psrlq       mm5, 56
    psllq       mm5, 48
    pxor        mm1, mm5
    jmp .do_left
.fix_lt_2:
    movq        mm5, mm3
    pxor        mm5, mm2
    psllq       mm5, 56
    psrlq       mm5, 56
    pxor        mm2, mm5
    test         r2, r2
    jnz .body
.fix_tr_1:
    movq        mm5, mm3
    pxor        mm5, mm1
    psrlq       mm5, 56
    psllq       mm5, 56
    pxor        mm1, mm5
    jmp .body
.do_left:
    movq        mm0, mm4
    PRED4x4_LOWPASS mm2, mm1, mm4, mm3, mm5
    movq        mm4, mm0
    movq        mm7, mm2
    PRED4x4_LOWPASS mm1, mm3, mm0, mm4, mm5
    psllq       mm1, 56
    PALIGNR     mm7, mm1, 7, mm3
    movq        mm0, [r0-8]
    movq        mm3, [r0]
    movq        mm1, [r0+8]
    movq        mm2, mm3
    movq        mm4, mm3
    PALIGNR     mm2, mm0, 7, mm0
    PALIGNR     mm1, mm4, 1, mm4
    test         r1, r1
    jz .fix_lt_2
    test         r2, r2
    jz .fix_tr_1
.body
    lea          r1, [r0+r3*2]
    PRED4x4_LOWPASS mm6, mm2, mm1, mm3, mm5
    pxor        mm0, mm0
    pxor        mm1, mm1
    lea          r2, [r1+r3*2]
    psadbw      mm0, mm7
    psadbw      mm1, mm6
    paddw       mm0, [pw_8]
    paddw       mm0, mm1
    lea          r4, [r2+r3*2]
    psrlw       mm0, 4
    pshufw      mm0, mm0, 0
    packuswb    mm0, mm0
    movq [r0+r3*1], mm0
    movq [r0+r3*2], mm0
    movq [r1+r3*1], mm0
    movq [r1+r3*2], mm0
    movq [r2+r3*1], mm0
    movq [r2+r3*2], mm0
    movq [r4+r3*1], mm0
    movq [r4+r3*2], mm0
    RET
%endmacro
INIT_MMX
%define PALIGNR PALIGNR_MMX
PRED8x8L_DC mmxext
%define PALIGNR PALIGNR_SSSE3
PRED8x8L_DC ssse3

;-----------------------------------------------------------------------------
1244 1245
; void pred8x8l_horizontal(uint8_t *src, int has_topleft, int has_topright, int stride)
;-----------------------------------------------------------------------------
1246

1247 1248 1249 1250 1251
%macro PRED8x8L_HORIZONTAL 1
cglobal pred8x8l_horizontal_%1, 4,4
    sub          r0, r3
    lea          r2, [r0+r3*2]
    movq        mm0, [r0+r3*1-8]
1252 1253 1254 1255
    test         r1, r1
    lea          r1, [r0+r3]
    cmovnz       r1, r0
    punpckhbw   mm0, [r1+r3*0-8]
1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269
    movq        mm1, [r2+r3*1-8]
    punpckhbw   mm1, [r0+r3*2-8]
    mov          r2, r0
    punpckhwd   mm1, mm0
    lea          r0, [r0+r3*4]
    movq        mm2, [r0+r3*1-8]
    punpckhbw   mm2, [r0+r3*0-8]
    lea          r0, [r0+r3*2]
    movq        mm3, [r0+r3*1-8]
    punpckhbw   mm3, [r0+r3*0-8]
    punpckhwd   mm3, mm2
    punpckhdq   mm3, mm1
    lea          r0, [r0+r3*2]
    movq        mm0, [r0+r3*0-8]
1270
    movq        mm1, [r1+r3*0-8]
1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315
    mov          r0, r2
    movq        mm4, mm3
    movq        mm2, mm3
    PALIGNR     mm4, mm0, 7, mm0
    PALIGNR     mm1, mm2, 1, mm2
    movq        mm0, mm4
    PRED4x4_LOWPASS mm2, mm1, mm4, mm3, mm5
    movq        mm4, mm0
    movq        mm7, mm2
    PRED4x4_LOWPASS mm1, mm3, mm0, mm4, mm5
    psllq       mm1, 56
    PALIGNR     mm7, mm1, 7, mm3
    movq        mm3, mm7
    lea         r1, [r0+r3*2]
    movq       mm7, mm3
    punpckhbw  mm3, mm3
    punpcklbw  mm7, mm7
    pshufw     mm0, mm3, 0xff
    pshufw     mm1, mm3, 0xaa
    lea         r2, [r1+r3*2]
    pshufw     mm2, mm3, 0x55
    pshufw     mm3, mm3, 0x00
    pshufw     mm4, mm7, 0xff
    pshufw     mm5, mm7, 0xaa
    pshufw     mm6, mm7, 0x55
    pshufw     mm7, mm7, 0x00
    movq [r0+r3*1], mm0
    movq [r0+r3*2], mm1
    movq [r1+r3*1], mm2
    movq [r1+r3*2], mm3
    movq [r2+r3*1], mm4
    movq [r2+r3*2], mm5
    lea         r0, [r2+r3*2]
    movq [r0+r3*1], mm6
    movq [r0+r3*2], mm7
    RET
%endmacro

INIT_MMX
%define PALIGNR PALIGNR_MMX
PRED8x8L_HORIZONTAL mmxext
%define PALIGNR PALIGNR_SSSE3
PRED8x8L_HORIZONTAL ssse3

;-----------------------------------------------------------------------------
1316 1317
; void pred8x8l_vertical(uint8_t *src, int has_topleft, int has_topright, int stride)
;-----------------------------------------------------------------------------
1318

1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366
%macro PRED8x8L_VERTICAL 1
cglobal pred8x8l_vertical_%1, 4,4
    sub          r0, r3
    movq        mm0, [r0-8]
    movq        mm3, [r0]
    movq        mm1, [r0+8]
    movq        mm2, mm3
    movq        mm4, mm3
    PALIGNR     mm2, mm0, 7, mm0
    PALIGNR     mm1, mm4, 1, mm4
    test         r1, r1 ; top_left
    jz .fix_lt_2
    test         r2, r2 ; top_right
    jz .fix_tr_1
    jmp .body
.fix_lt_2:
    movq        mm5, mm3
    pxor        mm5, mm2
    psllq       mm5, 56
    psrlq       mm5, 56
    pxor        mm2, mm5
    test         r2, r2 ; top_right
    jnz .body
.fix_tr_1:
    movq        mm5, mm3
    pxor        mm5, mm1
    psrlq       mm5, 56
    psllq       mm5, 56
    pxor        mm1, mm5
.body
    PRED4x4_LOWPASS mm0, mm2, mm1, mm3, mm5
%rep 3
    movq [r0+r3*1], mm0
    movq [r0+r3*2], mm0
    lea    r0, [r0+r3*2]
%endrep
    movq [r0+r3*1], mm0
    movq [r0+r3*2], mm0
    RET
%endmacro

INIT_MMX
%define PALIGNR PALIGNR_MMX
PRED8x8L_VERTICAL mmxext
%define PALIGNR PALIGNR_SSSE3
PRED8x8L_VERTICAL ssse3

;-----------------------------------------------------------------------------
1367 1368
;void pred8x8l_down_left(uint8_t *src, int has_topleft, int has_topright, int stride)
;-----------------------------------------------------------------------------
1369

1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478
INIT_MMX
%define PALIGNR PALIGNR_MMX
cglobal pred8x8l_down_left_mmxext, 4,5
    sub          r0, r3
    movq        mm0, [r0-8]
    movq        mm3, [r0]
    movq        mm1, [r0+8]
    movq        mm2, mm3
    movq        mm4, mm3
    PALIGNR     mm2, mm0, 7, mm0
    PALIGNR     mm1, mm4, 1, mm4
    test         r1, r1
    jz .fix_lt_2
    test         r2, r2
    jz .fix_tr_1
    jmp .do_top
.fix_lt_2:
    movq        mm5, mm3
    pxor        mm5, mm2
    psllq       mm5, 56
    psrlq       mm5, 56
    pxor        mm2, mm5
    test         r2, r2
    jnz .do_top
.fix_tr_1:
    movq        mm5, mm3
    pxor        mm5, mm1
    psrlq       mm5, 56
    psllq       mm5, 56
    pxor        mm1, mm5
    jmp .do_top
.fix_tr_2:
    punpckhbw   mm3, mm3
    pshufw      mm1, mm3, 0xFF
    jmp .do_topright
.do_top:
    PRED4x4_LOWPASS mm4, mm2, mm1, mm3, mm5
    movq        mm7, mm4
    test         r2, r2
    jz .fix_tr_2
    movq        mm0, [r0+8]
    movq        mm5, mm0
    movq        mm2, mm0
    movq        mm4, mm0
    psrlq       mm5, 56
    PALIGNR     mm2, mm3, 7, mm3
    PALIGNR     mm5, mm4, 1, mm4
    PRED4x4_LOWPASS mm1, mm2, mm5, mm0, mm4
.do_topright:
    lea          r1, [r0+r3*2]
    movq        mm6, mm1
    psrlq       mm1, 56
    movq        mm4, mm1
    lea          r2, [r1+r3*2]
    movq        mm2, mm6
    PALIGNR     mm2, mm7, 1, mm0
    movq        mm3, mm6
    PALIGNR     mm3, mm7, 7, mm0
    PALIGNR     mm4, mm6, 1, mm0
    movq        mm5, mm7
    movq        mm1, mm7
    movq        mm7, mm6
    lea          r4, [r2+r3*2]
    psllq       mm1, 8
    PRED4x4_LOWPASS mm0, mm1, mm2, mm5, mm6
    PRED4x4_LOWPASS mm1, mm3, mm4, mm7, mm6
    movq  [r4+r3*2], mm1
    movq        mm2, mm0
    psllq       mm1, 8
    psrlq       mm2, 56
    psllq       mm0, 8
    por         mm1, mm2
    movq  [r4+r3*1], mm1
    movq        mm2, mm0
    psllq       mm1, 8
    psrlq       mm2, 56
    psllq       mm0, 8
    por         mm1, mm2
    movq  [r2+r3*2], mm1
    movq        mm2, mm0
    psllq       mm1, 8
    psrlq       mm2, 56
    psllq       mm0, 8
    por         mm1, mm2
    movq  [r2+r3*1], mm1
    movq        mm2, mm0
    psllq       mm1, 8
    psrlq       mm2, 56
    psllq       mm0, 8
    por         mm1, mm2
    movq  [r1+r3*2], mm1
    movq        mm2, mm0
    psllq       mm1, 8
    psrlq       mm2, 56
    psllq       mm0, 8
    por         mm1, mm2
    movq  [r1+r3*1], mm1
    movq        mm2, mm0
    psllq       mm1, 8
    psrlq       mm2, 56
    psllq       mm0, 8
    por         mm1, mm2
    movq  [r0+r3*2], mm1
    psllq       mm1, 8
    psrlq       mm0, 56
    por         mm1, mm0
    movq  [r0+r3*1], mm1
    RET

1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569
%macro PRED8x8L_DOWN_LEFT 1
cglobal pred8x8l_down_left_%1, 4,4
    sub          r0, r3
    movq        mm0, [r0-8]
    movq        mm3, [r0]
    movq        mm1, [r0+8]
    movq        mm2, mm3
    movq        mm4, mm3
    PALIGNR     mm2, mm0, 7, mm0
    PALIGNR     mm1, mm4, 1, mm4
    test         r1, r1 ; top_left
    jz .fix_lt_2
    test         r2, r2 ; top_right
    jz .fix_tr_1
    jmp .do_top
.fix_lt_2:
    movq        mm5, mm3
    pxor        mm5, mm2
    psllq       mm5, 56
    psrlq       mm5, 56
    pxor        mm2, mm5
    test         r2, r2 ; top_right
    jnz .do_top
.fix_tr_1:
    movq        mm5, mm3
    pxor        mm5, mm1
    psrlq       mm5, 56
    psllq       mm5, 56
    pxor        mm1, mm5
    jmp .do_top
.fix_tr_2:
    punpckhbw   mm3, mm3
    pshufw      mm1, mm3, 0xFF
    jmp .do_topright
.do_top:
    PRED4x4_LOWPASS mm4, mm2, mm1, mm3, mm5
    movq2dq    xmm3, mm4
    test         r2, r2 ; top_right
    jz .fix_tr_2
    movq        mm0, [r0+8]
    movq        mm5, mm0
    movq        mm2, mm0
    movq        mm4, mm0
    psrlq       mm5, 56
    PALIGNR     mm2, mm3, 7, mm3
    PALIGNR     mm5, mm4, 1, mm4
    PRED4x4_LOWPASS mm1, mm2, mm5, mm0, mm4
.do_topright:
    movq2dq    xmm4, mm1
    psrlq       mm1, 56
    movq2dq    xmm5, mm1
    lea         r1, [r0+r3*2]
    pslldq    xmm4, 8
    por       xmm3, xmm4
    movdqa    xmm2, xmm3
    psrldq    xmm2, 1
    pslldq    xmm5, 15
    por       xmm2, xmm5
    lea         r2, [r1+r3*2]
    movdqa    xmm1, xmm3
    pslldq    xmm1, 1
INIT_XMM
    PRED4x4_LOWPASS xmm0, xmm1, xmm2, xmm3, xmm4
    psrldq    xmm0, 1
    movq [r0+r3*1], xmm0
    psrldq    xmm0, 1
    movq [r0+r3*2], xmm0
    psrldq    xmm0, 1
    lea         r0, [r2+r3*2]
    movq [r1+r3*1], xmm0
    psrldq    xmm0, 1
    movq [r1+r3*2], xmm0
    psrldq    xmm0, 1
    movq [r2+r3*1], xmm0
    psrldq    xmm0, 1
    movq [r2+r3*2], xmm0
    psrldq    xmm0, 1
    movq [r0+r3*1], xmm0
    psrldq    xmm0, 1
    movq [r0+r3*2], xmm0
    RET
%endmacro

INIT_MMX
%define PALIGNR PALIGNR_MMX
PRED8x8L_DOWN_LEFT sse2
INIT_MMX
%define PALIGNR PALIGNR_SSSE3
PRED8x8L_DOWN_LEFT ssse3

;-----------------------------------------------------------------------------
1570 1571
;void pred8x8l_down_right_mmxext(uint8_t *src, int has_topleft, int has_topright, int stride)
;-----------------------------------------------------------------------------
1572

1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704
INIT_MMX
%define PALIGNR PALIGNR_MMX
cglobal pred8x8l_down_right_mmxext, 4,5
    sub          r0, r3
    lea          r4, [r0+r3*2]
    movq        mm0, [r0+r3*1-8]
    punpckhbw   mm0, [r0+r3*0-8]
    movq        mm1, [r4+r3*1-8]
    punpckhbw   mm1, [r0+r3*2-8]
    mov          r4, r0
    punpckhwd   mm1, mm0
    lea          r0, [r0+r3*4]
    movq        mm2, [r0+r3*1-8]
    punpckhbw   mm2, [r0+r3*0-8]
    lea          r0, [r0+r3*2]
    movq        mm3, [r0+r3*1-8]
    punpckhbw   mm3, [r0+r3*0-8]
    punpckhwd   mm3, mm2
    punpckhdq   mm3, mm1
    lea          r0, [r0+r3*2]
    movq        mm0, [r0+r3*0-8]
    movq        mm1, [r4]
    mov          r0, r4
    movq        mm4, mm3
    movq        mm2, mm3
    PALIGNR     mm4, mm0, 7, mm0
    PALIGNR     mm1, mm2, 1, mm2
    test        r1, r1 ; top_left
    jz .fix_lt_1
.do_left:
    movq        mm0, mm4
    PRED4x4_LOWPASS mm2, mm1, mm4, mm3, mm5
    movq        mm4, mm0
    movq        mm7, mm2
    movq        mm6, mm2
    PRED4x4_LOWPASS mm1, mm3, mm0, mm4, mm5
    psllq       mm1, 56
    PALIGNR     mm7, mm1, 7, mm3
    movq        mm0, [r0-8]
    movq        mm3, [r0]
    movq        mm1, [r0+8]
    movq        mm2, mm3
    movq        mm4, mm3
    PALIGNR     mm2, mm0, 7, mm0
    PALIGNR     mm1, mm4, 1, mm4
    test         r1, r1 ; top_left
    jz .fix_lt_2
    test         r2, r2 ; top_right
    jz .fix_tr_1
.do_top:
    PRED4x4_LOWPASS mm4, mm2, mm1, mm3, mm5
    movq        mm5, mm4
    jmp .body
.fix_lt_1:
    movq        mm5, mm3
    pxor        mm5, mm4
    psrlq       mm5, 56
    psllq       mm5, 48
    pxor        mm1, mm5
    jmp .do_left
.fix_lt_2:
    movq        mm5, mm3
    pxor        mm5, mm2
    psllq       mm5, 56
    psrlq       mm5, 56
    pxor        mm2, mm5
    test         r2, r2 ; top_right
    jnz .do_top
.fix_tr_1:
    movq        mm5, mm3
    pxor        mm5, mm1
    psrlq       mm5, 56
    psllq       mm5, 56
    pxor        mm1, mm5
    jmp .do_top
.body
    lea         r1, [r0+r3*2]
    movq       mm1, mm7
    movq       mm7, mm5
    movq       mm5, mm6
    movq       mm2, mm7
    lea         r2, [r1+r3*2]
    PALIGNR    mm2, mm6, 1, mm0
    movq       mm3, mm7
    PALIGNR    mm3, mm6, 7, mm0
    movq       mm4, mm7
    lea         r4, [r2+r3*2]
    psrlq      mm4, 8
    PRED4x4_LOWPASS mm0, mm1, mm2, mm5, mm6
    PRED4x4_LOWPASS mm1, mm3, mm4, mm7, mm6
    movq [r4+r3*2], mm0
    movq       mm2, mm1
    psrlq      mm0, 8
    psllq      mm2, 56
    psrlq      mm1, 8
    por        mm0, mm2
    movq [r4+r3*1], mm0
    movq       mm2, mm1
    psrlq      mm0, 8
    psllq      mm2, 56
    psrlq      mm1, 8
    por        mm0, mm2
    movq [r2+r3*2], mm0
    movq       mm2, mm1
    psrlq      mm0, 8
    psllq      mm2, 56
    psrlq      mm1, 8
    por        mm0, mm2
    movq [r2+r3*1], mm0
    movq       mm2, mm1
    psrlq      mm0, 8
    psllq      mm2, 56
    psrlq      mm1, 8
    por        mm0, mm2
    movq [r1+r3*2], mm0
    movq       mm2, mm1
    psrlq      mm0, 8
    psllq      mm2, 56
    psrlq      mm1, 8
    por        mm0, mm2
    movq [r1+r3*1], mm0
    movq       mm2, mm1
    psrlq      mm0, 8
    psllq      mm2, 56
    psrlq      mm1, 8
    por        mm0, mm2
    movq [r0+r3*2], mm0
    psrlq      mm0, 8
    psllq      mm1, 56
    por        mm0, mm1
    movq [r0+r3*1], mm0
    RET
1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821

%macro PRED8x8L_DOWN_RIGHT 1
cglobal pred8x8l_down_right_%1, 4,5
    sub          r0, r3
    lea          r4, [r0+r3*2]
    movq        mm0, [r0+r3*1-8]
    punpckhbw   mm0, [r0+r3*0-8]
    movq        mm1, [r4+r3*1-8]
    punpckhbw   mm1, [r0+r3*2-8]
    mov          r4, r0
    punpckhwd   mm1, mm0
    lea          r0, [r0+r3*4]
    movq        mm2, [r0+r3*1-8]
    punpckhbw   mm2, [r0+r3*0-8]
    lea          r0, [r0+r3*2]
    movq        mm3, [r0+r3*1-8]
    punpckhbw   mm3, [r0+r3*0-8]
    punpckhwd   mm3, mm2
    punpckhdq   mm3, mm1
    lea          r0, [r0+r3*2]
    movq        mm0, [r0+r3*0-8]
    movq        mm1, [r4]
    mov          r0, r4
    movq        mm4, mm3
    movq        mm2, mm3
    PALIGNR     mm4, mm0, 7, mm0
    PALIGNR     mm1, mm2, 1, mm2
    test        r1, r1
    jz .fix_lt_1
    jmp .do_left
.fix_lt_1:
    movq        mm5, mm3
    pxor        mm5, mm4
    psrlq       mm5, 56
    psllq       mm5, 48
    pxor        mm1, mm5
    jmp .do_left
.fix_lt_2:
    movq        mm5, mm3
    pxor        mm5, mm2
    psllq       mm5, 56
    psrlq       mm5, 56
    pxor        mm2, mm5
    test         r2, r2
    jnz .do_top
.fix_tr_1:
    movq        mm5, mm3
    pxor        mm5, mm1
    psrlq       mm5, 56
    psllq       mm5, 56
    pxor        mm1, mm5
    jmp .do_top
.do_left:
    movq        mm0, mm4
    PRED4x4_LOWPASS mm2, mm1, mm4, mm3, mm5
    movq        mm4, mm0
    movq        mm7, mm2
    movq2dq    xmm3, mm2
    PRED4x4_LOWPASS mm1, mm3, mm0, mm4, mm5
    psllq       mm1, 56
    PALIGNR     mm7, mm1, 7, mm3
    movq2dq    xmm1, mm7
    movq        mm0, [r0-8]
    movq        mm3, [r0]
    movq        mm1, [r0+8]
    movq        mm2, mm3
    movq        mm4, mm3
    PALIGNR     mm2, mm0, 7, mm0
    PALIGNR     mm1, mm4, 1, mm4
    test         r1, r1
    jz .fix_lt_2
    test         r2, r2
    jz .fix_tr_1
.do_top:
    PRED4x4_LOWPASS mm4, mm2, mm1, mm3, mm5
    movq2dq   xmm4, mm4
    lea         r1, [r0+r3*2]
    movdqa    xmm0, xmm3
    pslldq    xmm4, 8
    por       xmm3, xmm4
    lea         r2, [r1+r3*2]
    pslldq    xmm4, 1
    por       xmm1, xmm4
    psrldq    xmm0, 7
    pslldq    xmm0, 15
    psrldq    xmm0, 7
    por       xmm1, xmm0
    lea         r0, [r2+r3*2]
    movdqa    xmm2, xmm3
    psrldq    xmm2, 1
INIT_XMM
    PRED4x4_LOWPASS xmm0, xmm1, xmm2, xmm3, xmm4
    movdqa    xmm1, xmm0
    psrldq    xmm1, 1
    movq [r0+r3*2], xmm0
    movq [r0+r3*1], xmm1
    psrldq    xmm0, 2
    psrldq    xmm1, 2
    movq [r2+r3*2], xmm0
    movq [r2+r3*1], xmm1
    psrldq    xmm0, 2
    psrldq    xmm1, 2
    movq [r1+r3*2], xmm0
    movq [r1+r3*1], xmm1
    psrldq    xmm0, 2
    psrldq    xmm1, 2
    movq [r4+r3*2], xmm0
    movq [r4+r3*1], xmm1
    RET
%endmacro

INIT_MMX
%define PALIGNR PALIGNR_MMX
PRED8x8L_DOWN_RIGHT sse2
INIT_MMX
%define PALIGNR PALIGNR_SSSE3
PRED8x8L_DOWN_RIGHT ssse3
1822 1823

;-----------------------------------------------------------------------------
1824 1825
; void pred8x8l_vertical_right(uint8_t *src, int has_topleft, int has_topright, int stride)
;-----------------------------------------------------------------------------
1826

1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933
INIT_MMX
%define PALIGNR PALIGNR_MMX
cglobal pred8x8l_vertical_right_mmxext, 4,5
    sub          r0, r3
    lea          r4, [r0+r3*2]
    movq        mm0, [r0+r3*1-8]
    punpckhbw   mm0, [r0+r3*0-8]
    movq        mm1, [r4+r3*1-8]
    punpckhbw   mm1, [r0+r3*2-8]
    mov          r4, r0
    punpckhwd   mm1, mm0
    lea          r0, [r0+r3*4]
    movq        mm2, [r0+r3*1-8]
    punpckhbw   mm2, [r0+r3*0-8]
    lea          r0, [r0+r3*2]
    movq        mm3, [r0+r3*1-8]
    punpckhbw   mm3, [r0+r3*0-8]
    punpckhwd   mm3, mm2
    punpckhdq   mm3, mm1
    lea          r0, [r0+r3*2]
    movq        mm0, [r0+r3*0-8]
    movq        mm1, [r4]
    mov          r0, r4
    movq        mm4, mm3
    movq        mm2, mm3
    PALIGNR     mm4, mm0, 7, mm0
    PALIGNR     mm1, mm2, 1, mm2
    test        r1, r1
    jz .fix_lt_1
    jmp .do_left
.fix_lt_1:
    movq        mm5, mm3
    pxor        mm5, mm4
    psrlq       mm5, 56
    psllq       mm5, 48
    pxor        mm1, mm5
    jmp .do_left
.fix_lt_2:
    movq        mm5, mm3
    pxor        mm5, mm2
    psllq       mm5, 56
    psrlq       mm5, 56
    pxor        mm2, mm5
    test         r2, r2
    jnz .do_top
.fix_tr_1:
    movq        mm5, mm3
    pxor        mm5, mm1
    psrlq       mm5, 56
    psllq       mm5, 56
    pxor        mm1, mm5
    jmp .do_top
.do_left:
    movq        mm0, mm4
    PRED4x4_LOWPASS mm2, mm1, mm4, mm3, mm5
    movq        mm7, mm2
    movq        mm0, [r0-8]
    movq        mm3, [r0]
    movq        mm1, [r0+8]
    movq        mm2, mm3
    movq        mm4, mm3
    PALIGNR     mm2, mm0, 7, mm0
    PALIGNR     mm1, mm4, 1, mm4
    test         r1, r1
    jz .fix_lt_2
    test         r2, r2
    jz .fix_tr_1
.do_top
    PRED4x4_LOWPASS mm6, mm2, mm1, mm3, mm5
    lea         r1, [r0+r3*2]
    movq       mm2, mm6
    movq       mm3, mm6
    PALIGNR    mm3, mm7, 7, mm0
    PALIGNR    mm6, mm7, 6, mm1
    movq       mm4, mm3
    pavgb      mm3, mm2
    lea         r2, [r1+r3*2]
    PRED4x4_LOWPASS mm0, mm6, mm2, mm4, mm5
    movq [r0+r3*1], mm3
    movq [r0+r3*2], mm0
    movq       mm5, mm0
    movq       mm6, mm3
    movq       mm1, mm7
    movq       mm2, mm1
    psllq      mm2, 8
    movq       mm3, mm1
    psllq      mm3, 16
    lea         r4, [r2+r3*2]
    PRED4x4_LOWPASS mm0, mm1, mm3, mm2, mm4
    PALIGNR    mm6, mm0, 7, mm2
    movq [r1+r3*1], mm6
    psllq      mm0, 8
    PALIGNR    mm5, mm0, 7, mm1
    movq [r1+r3*2], mm5
    psllq      mm0, 8
    PALIGNR    mm6, mm0, 7, mm2
    movq [r2+r3*1], mm6
    psllq      mm0, 8
    PALIGNR    mm5, mm0, 7, mm1
    movq [r2+r3*2], mm5
    psllq      mm0, 8
    PALIGNR    mm6, mm0, 7, mm2
    movq [r4+r3*1], mm6
    psllq      mm0, 8
    PALIGNR    mm5, mm0, 7, mm1
    movq [r4+r3*2], mm5
    RET
1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048

%macro PRED8x8L_VERTICAL_RIGHT 1
cglobal pred8x8l_vertical_right_%1, 4,5,7
    sub          r0, r3
    lea          r4, [r0+r3*2]
    movq        mm0, [r0+r3*1-8]
    punpckhbw   mm0, [r0+r3*0-8]
    movq        mm1, [r4+r3*1-8]
    punpckhbw   mm1, [r0+r3*2-8]
    mov          r4, r0
    punpckhwd   mm1, mm0
    lea          r0, [r0+r3*4]
    movq        mm2, [r0+r3*1-8]
    punpckhbw   mm2, [r0+r3*0-8]
    lea          r0, [r0+r3*2]
    movq        mm3, [r0+r3*1-8]
    punpckhbw   mm3, [r0+r3*0-8]
    punpckhwd   mm3, mm2
    punpckhdq   mm3, mm1
    lea          r0, [r0+r3*2]
    movq        mm0, [r0+r3*0-8]
    movq        mm1, [r4]
    mov          r0, r4
    movq        mm4, mm3
    movq        mm2, mm3
    PALIGNR     mm4, mm0, 7, mm0
    PALIGNR     mm1, mm2, 1, mm2
    test        r1, r1
    jnz .do_left
.fix_lt_1:
    movq        mm5, mm3
    pxor        mm5, mm4
    psrlq       mm5, 56
    psllq       mm5, 48
    pxor        mm1, mm5
    jmp .do_left
.fix_lt_2:
    movq        mm5, mm3
    pxor        mm5, mm2
    psllq       mm5, 56
    psrlq       mm5, 56
    pxor        mm2, mm5
    test         r2, r2
    jnz .do_top
.fix_tr_1:
    movq        mm5, mm3
    pxor        mm5, mm1
    psrlq       mm5, 56
    psllq       mm5, 56
    pxor        mm1, mm5
    jmp .do_top
.do_left:
    movq        mm0, mm4
    PRED4x4_LOWPASS mm2, mm1, mm4, mm3, mm5
    movq2dq    xmm0, mm2
    movq        mm0, [r0-8]
    movq        mm3, [r0]
    movq        mm1, [r0+8]
    movq        mm2, mm3
    movq        mm4, mm3
    PALIGNR     mm2, mm0, 7, mm0
    PALIGNR     mm1, mm4, 1, mm4
    test         r1, r1
    jz .fix_lt_2
    test         r2, r2
    jz .fix_tr_1
.do_top
    PRED4x4_LOWPASS mm6, mm2, mm1, mm3, mm5
    lea           r1, [r0+r3*2]
    movq2dq     xmm4, mm6
    pslldq      xmm4, 8
    por         xmm0, xmm4
    movdqa      xmm6, [pw_ff00]
    movdqa      xmm1, xmm0
    lea           r2, [r1+r3*2]
    movdqa      xmm2, xmm0
    movdqa      xmm3, xmm0
    pslldq      xmm0, 1
    pslldq      xmm1, 2
    pavgb       xmm2, xmm0
INIT_XMM
    PRED4x4_LOWPASS xmm4, xmm3, xmm1, xmm0, xmm5
    pandn       xmm6, xmm4
    movdqa      xmm5, xmm4
    psrlw       xmm4, 8
    packuswb    xmm6, xmm4
    movhlps     xmm4, xmm6
    movhps [r0+r3*2], xmm5
    movhps [r0+r3*1], xmm2
    psrldq      xmm5, 4
    movss       xmm5, xmm6
    psrldq      xmm2, 4
    movss       xmm2, xmm4
    lea           r0, [r2+r3*2]
    psrldq      xmm5, 1
    psrldq      xmm2, 1
    movq        [r0+r3*2], xmm5
    movq        [r0+r3*1], xmm2
    psrldq      xmm5, 1
    psrldq      xmm2, 1
    movq        [r2+r3*2], xmm5
    movq        [r2+r3*1], xmm2
    psrldq      xmm5, 1
    psrldq      xmm2, 1
    movq        [r1+r3*2], xmm5
    movq        [r1+r3*1], xmm2
    RET
%endmacro

INIT_MMX
%define PALIGNR PALIGNR_MMX
PRED8x8L_VERTICAL_RIGHT sse2
INIT_MMX
%define PALIGNR PALIGNR_SSSE3
PRED8x8L_VERTICAL_RIGHT ssse3
2049 2050

;-----------------------------------------------------------------------------
2051 2052
;void pred8x8l_vertical_left(uint8_t *src, int has_topleft, int has_topright, int stride)
;-----------------------------------------------------------------------------
2053

2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141
%macro PRED8x8L_VERTICAL_LEFT 1
cglobal pred8x8l_vertical_left_%1, 4,4
    sub          r0, r3
    movq        mm0, [r0-8]
    movq        mm3, [r0]
    movq        mm1, [r0+8]
    movq        mm2, mm3
    movq        mm4, mm3
    PALIGNR     mm2, mm0, 7, mm0
    PALIGNR     mm1, mm4, 1, mm4
    test         r1, r1
    jz .fix_lt_2
    test         r2, r2
    jz .fix_tr_1
    jmp .do_top
.fix_lt_2:
    movq        mm5, mm3
    pxor        mm5, mm2
    psllq       mm5, 56
    psrlq       mm5, 56
    pxor        mm2, mm5
    test         r2, r2
    jnz .do_top
.fix_tr_1:
    movq        mm5, mm3
    pxor        mm5, mm1
    psrlq       mm5, 56
    psllq       mm5, 56
    pxor        mm1, mm5
    jmp .do_top
.fix_tr_2:
    punpckhbw   mm3, mm3
    pshufw      mm1, mm3, 0xFF
    jmp .do_topright
.do_top:
    PRED4x4_LOWPASS mm4, mm2, mm1, mm3, mm5
    movq2dq    xmm4, mm4
    test         r2, r2
    jz .fix_tr_2
    movq        mm0, [r0+8]
    movq        mm5, mm0
    movq        mm2, mm0
    movq        mm4, mm0
    psrlq       mm5, 56
    PALIGNR     mm2, mm3, 7, mm3
    PALIGNR     mm5, mm4, 1, mm4
    PRED4x4_LOWPASS mm1, mm2, mm5, mm0, mm4
.do_topright:
    movq2dq   xmm3, mm1
    lea         r1, [r0+r3*2]
    pslldq    xmm3, 8
    por       xmm4, xmm3
    movdqa    xmm2, xmm4
    movdqa    xmm1, xmm4
    movdqa    xmm3, xmm4
    psrldq    xmm2, 1
    pslldq    xmm1, 1
    pavgb     xmm3, xmm2
    lea         r2, [r1+r3*2]
INIT_XMM
    PRED4x4_LOWPASS xmm0, xmm1, xmm2, xmm4, xmm5
    psrldq    xmm0, 1
    movq [r0+r3*1], xmm3
    movq [r0+r3*2], xmm0
    lea         r0, [r2+r3*2]
    psrldq    xmm3, 1
    psrldq    xmm0, 1
    movq [r1+r3*1], xmm3
    movq [r1+r3*2], xmm0
    psrldq    xmm3, 1
    psrldq    xmm0, 1
    movq [r2+r3*1], xmm3
    movq [r2+r3*2], xmm0
    psrldq    xmm3, 1
    psrldq    xmm0, 1
    movq [r0+r3*1], xmm3
    movq [r0+r3*2], xmm0
    RET
%endmacro

INIT_MMX
%define PALIGNR PALIGNR_MMX
PRED8x8L_VERTICAL_LEFT sse2
%define PALIGNR PALIGNR_SSSE3
INIT_MMX
PRED8x8L_VERTICAL_LEFT ssse3

;-----------------------------------------------------------------------------
2142 2143
; void pred8x8l_horizontal_up(uint8_t *src, int has_topleft, int has_topright, int stride)
;-----------------------------------------------------------------------------
2144

2145 2146 2147 2148 2149
%macro PRED8x8L_HORIZONTAL_UP 1
cglobal pred8x8l_horizontal_up_%1, 4,4
    sub          r0, r3
    lea          r2, [r0+r3*2]
    movq        mm0, [r0+r3*1-8]
2150 2151 2152 2153
    test         r1, r1
    lea          r1, [r0+r3]
    cmovnz       r1, r0
    punpckhbw   mm0, [r1+r3*0-8]
2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167
    movq        mm1, [r2+r3*1-8]
    punpckhbw   mm1, [r0+r3*2-8]
    mov          r2, r0
    punpckhwd   mm1, mm0
    lea          r0, [r0+r3*4]
    movq        mm2, [r0+r3*1-8]
    punpckhbw   mm2, [r0+r3*0-8]
    lea          r0, [r0+r3*2]
    movq        mm3, [r0+r3*1-8]
    punpckhbw   mm3, [r0+r3*0-8]
    punpckhwd   mm3, mm2
    punpckhdq   mm3, mm1
    lea          r0, [r0+r3*2]
    movq        mm0, [r0+r3*0-8]
2168
    movq        mm1, [r1+r3*0-8]
2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229
    mov          r0, r2
    movq        mm4, mm3
    movq        mm2, mm3
    PALIGNR     mm4, mm0, 7, mm0
    PALIGNR     mm1, mm2, 1, mm2
    movq       mm0, mm4
    PRED4x4_LOWPASS mm2, mm1, mm4, mm3, mm5
    movq       mm4, mm0
    movq       mm7, mm2
    PRED4x4_LOWPASS mm1, mm3, mm0, mm4, mm5
    psllq      mm1, 56
    PALIGNR    mm7, mm1, 7, mm3
    lea         r1, [r0+r3*2]
    pshufw     mm0, mm7, 00011011b ; l6 l7 l4 l5 l2 l3 l0 l1
    psllq      mm7, 56             ; l7 .. .. .. .. .. .. ..
    movq       mm2, mm0
    psllw      mm0, 8
    psrlw      mm2, 8
    por        mm2, mm0            ; l7 l6 l5 l4 l3 l2 l1 l0
    movq       mm3, mm2
    movq       mm4, mm2
    movq       mm5, mm2
    psrlq      mm2, 8
    psrlq      mm3, 16
    lea         r2, [r1+r3*2]
    por        mm2, mm7            ; l7 l7 l6 l5 l4 l3 l2 l1
    punpckhbw  mm7, mm7
    por        mm3, mm7            ; l7 l7 l7 l6 l5 l4 l3 l2
    pavgb      mm4, mm2
    PRED4x4_LOWPASS mm1, mm3, mm5, mm2, mm6
    movq       mm5, mm4
    punpcklbw  mm4, mm1            ; p4 p3 p2 p1
    punpckhbw  mm5, mm1            ; p8 p7 p6 p5
    movq       mm6, mm5
    movq       mm7, mm5
    movq       mm0, mm5
    PALIGNR    mm5, mm4, 2, mm1
    pshufw     mm1, mm6, 11111001b
    PALIGNR    mm6, mm4, 4, mm2
    pshufw     mm2, mm7, 11111110b
    PALIGNR    mm7, mm4, 6, mm3
    pshufw     mm3, mm0, 11111111b
    movq [r0+r3*1], mm4
    movq [r0+r3*2], mm5
    lea         r0, [r2+r3*2]
    movq [r1+r3*1], mm6
    movq [r1+r3*2], mm7
    movq [r2+r3*1], mm0
    movq [r2+r3*2], mm1
    movq [r0+r3*1], mm2
    movq [r0+r3*2], mm3
    RET
%endmacro

INIT_MMX
%define PALIGNR PALIGNR_MMX
PRED8x8L_HORIZONTAL_UP mmxext
%define PALIGNR PALIGNR_SSSE3
PRED8x8L_HORIZONTAL_UP ssse3

;-----------------------------------------------------------------------------
2230 2231
;void pred8x8l_horizontal_down(uint8_t *src, int has_topleft, int has_topright, int stride)
;-----------------------------------------------------------------------------
2232

2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347
INIT_MMX
%define PALIGNR PALIGNR_MMX
cglobal pred8x8l_horizontal_down_mmxext, 4,5
    sub          r0, r3
    lea          r4, [r0+r3*2]
    movq        mm0, [r0+r3*1-8]
    punpckhbw   mm0, [r0+r3*0-8]
    movq        mm1, [r4+r3*1-8]
    punpckhbw   mm1, [r0+r3*2-8]
    mov          r4, r0
    punpckhwd   mm1, mm0
    lea          r0, [r0+r3*4]
    movq        mm2, [r0+r3*1-8]
    punpckhbw   mm2, [r0+r3*0-8]
    lea          r0, [r0+r3*2]
    movq        mm3, [r0+r3*1-8]
    punpckhbw   mm3, [r0+r3*0-8]
    punpckhwd   mm3, mm2
    punpckhdq   mm3, mm1
    lea          r0, [r0+r3*2]
    movq        mm0, [r0+r3*0-8]
    movq        mm1, [r4]
    mov          r0, r4
    movq        mm4, mm3
    movq        mm2, mm3
    PALIGNR     mm4, mm0, 7, mm0
    PALIGNR     mm1, mm2, 1, mm2
    test        r1, r1
    jnz .do_left
.fix_lt_1:
    movq        mm5, mm3
    pxor        mm5, mm4
    psrlq       mm5, 56
    psllq       mm5, 48
    pxor        mm1, mm5
    jmp .do_left
.fix_lt_2:
    movq        mm5, mm3
    pxor        mm5, mm2
    psllq       mm5, 56
    psrlq       mm5, 56
    pxor        mm2, mm5
    test         r2, r2
    jnz .do_top
.fix_tr_1:
    movq        mm5, mm3
    pxor        mm5, mm1
    psrlq       mm5, 56
    psllq       mm5, 56
    pxor        mm1, mm5
    jmp .do_top
.do_left:
    movq        mm0, mm4
    PRED4x4_LOWPASS mm2, mm1, mm4, mm3, mm5
    movq        mm4, mm0
    movq        mm7, mm2
    movq        mm6, mm2
    PRED4x4_LOWPASS mm1, mm3, mm0, mm4, mm5
    psllq       mm1, 56
    PALIGNR     mm7, mm1, 7, mm3
    movq        mm0, [r0-8]
    movq        mm3, [r0]
    movq        mm1, [r0+8]
    movq        mm2, mm3
    movq        mm4, mm3
    PALIGNR     mm2, mm0, 7, mm0
    PALIGNR     mm1, mm4, 1, mm4
    test         r1, r1
    jz .fix_lt_2
    test         r2, r2
    jz .fix_tr_1
.do_top:
    PRED4x4_LOWPASS mm4, mm2, mm1, mm3, mm5
    movq       mm5, mm4
    lea         r1, [r0+r3*2]
    psllq      mm7, 56
    movq       mm2, mm5
    movq       mm3, mm6
    movq       mm4, mm2
    PALIGNR    mm2, mm6, 7, mm5
    PALIGNR    mm6, mm7, 7, mm0
    lea         r2, [r1+r3*2]
    PALIGNR    mm4, mm3, 1, mm7
    movq       mm5, mm3
    pavgb      mm3, mm6
    PRED4x4_LOWPASS mm0, mm4, mm6, mm5, mm7
    movq       mm4, mm2
    movq       mm1, mm2
    lea         r4, [r2+r3*2]
    psrlq      mm4, 16
    psrlq      mm1, 8
    PRED4x4_LOWPASS mm6, mm4, mm2, mm1, mm5
    movq       mm7, mm3
    punpcklbw  mm3, mm0
    punpckhbw  mm7, mm0
    movq       mm1, mm7
    movq       mm0, mm7
    movq       mm4, mm7
    movq [r4+r3*2], mm3
    PALIGNR    mm7, mm3, 2, mm5
    movq [r4+r3*1], mm7
    PALIGNR    mm1, mm3, 4, mm5
    movq [r2+r3*2], mm1
    PALIGNR    mm0, mm3, 6, mm3
    movq [r2+r3*1], mm0
    movq       mm2, mm6
    movq       mm3, mm6
    movq [r1+r3*2], mm4
    PALIGNR    mm6, mm4, 2, mm5
    movq [r1+r3*1], mm6
    PALIGNR    mm2, mm4, 4, mm5
    movq [r0+r3*2], mm2
    PALIGNR    mm3, mm4, 6, mm4
    movq [r0+r3*1], mm3
    RET
2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478

%macro PRED8x8L_HORIZONTAL_DOWN 1
cglobal pred8x8l_horizontal_down_%1, 4,5
    sub          r0, r3
    lea          r4, [r0+r3*2]
    movq        mm0, [r0+r3*1-8]
    punpckhbw   mm0, [r0+r3*0-8]
    movq        mm1, [r4+r3*1-8]
    punpckhbw   mm1, [r0+r3*2-8]
    mov          r4, r0
    punpckhwd   mm1, mm0
    lea          r0, [r0+r3*4]
    movq        mm2, [r0+r3*1-8]
    punpckhbw   mm2, [r0+r3*0-8]
    lea          r0, [r0+r3*2]
    movq        mm3, [r0+r3*1-8]
    punpckhbw   mm3, [r0+r3*0-8]
    punpckhwd   mm3, mm2
    punpckhdq   mm3, mm1
    lea          r0, [r0+r3*2]
    movq        mm0, [r0+r3*0-8]
    movq        mm1, [r4]
    mov          r0, r4
    movq        mm4, mm3
    movq        mm2, mm3
    PALIGNR     mm4, mm0, 7, mm0
    PALIGNR     mm1, mm2, 1, mm2
    test        r1, r1
    jnz .do_left
.fix_lt_1:
    movq        mm5, mm3
    pxor        mm5, mm4
    psrlq       mm5, 56
    psllq       mm5, 48
    pxor        mm1, mm5
    jmp .do_left
.fix_lt_2:
    movq        mm5, mm3
    pxor        mm5, mm2
    psllq       mm5, 56
    psrlq       mm5, 56
    pxor        mm2, mm5
    test         r2, r2
    jnz .do_top
.fix_tr_1:
    movq        mm5, mm3
    pxor        mm5, mm1
    psrlq       mm5, 56
    psllq       mm5, 56
    pxor        mm1, mm5
    jmp .do_top
.fix_tr_2:
    punpckhbw   mm3, mm3
    pshufw      mm1, mm3, 0xFF
    jmp .do_topright
.do_left:
    movq        mm0, mm4
    PRED4x4_LOWPASS mm2, mm1, mm4, mm3, mm5
    movq2dq    xmm0, mm2
    pslldq     xmm0, 8
    movq        mm4, mm0
    PRED4x4_LOWPASS mm1, mm3, mm0, mm4, mm5
    movq2dq    xmm2, mm1
    pslldq     xmm2, 15
    psrldq     xmm2, 8
    por        xmm0, xmm2
    movq        mm0, [r0-8]
    movq        mm3, [r0]
    movq        mm1, [r0+8]
    movq        mm2, mm3
    movq        mm4, mm3
    PALIGNR     mm2, mm0, 7, mm0
    PALIGNR     mm1, mm4, 1, mm4
    test         r1, r1
    jz .fix_lt_2
    test         r2, r2
    jz .fix_tr_1
.do_top:
    PRED4x4_LOWPASS mm4, mm2, mm1, mm3, mm5
    movq2dq    xmm1, mm4
    test         r2, r2
    jz .fix_tr_2
    movq        mm0, [r0+8]
    movq        mm5, mm0
    movq        mm2, mm0
    movq        mm4, mm0
    psrlq       mm5, 56
    PALIGNR     mm2, mm3, 7, mm3
    PALIGNR     mm5, mm4, 1, mm4
    PRED4x4_LOWPASS mm1, mm2, mm5, mm0, mm4
.do_topright:
    movq2dq    xmm5, mm1
    pslldq     xmm5, 8
    por        xmm1, xmm5
INIT_XMM
    lea         r2, [r4+r3*2]
    movdqa    xmm2, xmm1
    movdqa    xmm3, xmm1
    PALIGNR   xmm1, xmm0, 7, xmm4
    PALIGNR   xmm2, xmm0, 9, xmm5
    lea         r1, [r2+r3*2]
    PALIGNR   xmm3, xmm0, 8, xmm0
    movdqa    xmm4, xmm1
    pavgb     xmm4, xmm3
    lea         r0, [r1+r3*2]
    PRED4x4_LOWPASS xmm0, xmm1, xmm2, xmm3, xmm5
    punpcklbw xmm4, xmm0
    movhlps   xmm0, xmm4
    movq   [r0+r3*2], xmm4
    movq   [r2+r3*2], xmm0
    psrldq xmm4, 2
    psrldq xmm0, 2
    movq   [r0+r3*1], xmm4
    movq   [r2+r3*1], xmm0
    psrldq xmm4, 2
    psrldq xmm0, 2
    movq   [r1+r3*2], xmm4
    movq   [r4+r3*2], xmm0
    psrldq xmm4, 2
    psrldq xmm0, 2
    movq   [r1+r3*1], xmm4
    movq   [r4+r3*1], xmm0
    RET
%endmacro

INIT_MMX
%define PALIGNR PALIGNR_MMX
PRED8x8L_HORIZONTAL_DOWN sse2
INIT_MMX
%define PALIGNR PALIGNR_SSSE3
PRED8x8L_HORIZONTAL_DOWN ssse3
2479 2480 2481
%endif

;-----------------------------------------------------------------------------
2482 2483 2484
; void pred4x4_dc_mmxext(uint8_t *src, const uint8_t *topright, int stride)
;-----------------------------------------------------------------------------

2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508
cglobal pred4x4_dc_mmxext, 3,5
    pxor   mm7, mm7
    mov     r4, r0
    sub     r0, r2
    movd   mm0, [r0]
    psadbw mm0, mm7
    movzx  r1d, byte [r0+r2*1-1]
    movd   r3d, mm0
    add    r3d, r1d
    movzx  r1d, byte [r0+r2*2-1]
    lea     r0, [r0+r2*2]
    add    r3d, r1d
    movzx  r1d, byte [r0+r2*1-1]
    add    r3d, r1d
    movzx  r1d, byte [r0+r2*2-1]
    add    r3d, r1d
    add    r3d, 4
    shr    r3d, 3
    imul   r3d, 0x01010101
    mov   [r4+r2*0], r3d
    mov   [r0+r2*0], r3d
    mov   [r0+r2*1], r3d
    mov   [r0+r2*2], r3d
    RET
2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 2586

;-----------------------------------------------------------------------------
; void pred4x4_tm_vp8_mmxext(uint8_t *src, const uint8_t *topright, int stride)
;-----------------------------------------------------------------------------

%macro PRED4x4_TM_MMX 1
cglobal pred4x4_tm_vp8_%1, 3,6
    sub        r0, r2
    pxor      mm7, mm7
    movd      mm0, [r0]
    punpcklbw mm0, mm7
    movzx     r4d, byte [r0-1]
    mov       r5d, 2
.loop:
    movzx     r1d, byte [r0+r2*1-1]
    movzx     r3d, byte [r0+r2*2-1]
    sub       r1d, r4d
    sub       r3d, r4d
    movd      mm2, r1d
    movd      mm4, r3d
%ifidn %1, mmx
    punpcklwd mm2, mm2
    punpcklwd mm4, mm4
    punpckldq mm2, mm2
    punpckldq mm4, mm4
%else
    pshufw    mm2, mm2, 0
    pshufw    mm4, mm4, 0
%endif
    paddw     mm2, mm0
    paddw     mm4, mm0
    packuswb  mm2, mm2
    packuswb  mm4, mm4
    movd [r0+r2*1], mm2
    movd [r0+r2*2], mm4
    lea        r0, [r0+r2*2]
    dec       r5d
    jg .loop
    REP_RET
%endmacro

PRED4x4_TM_MMX mmx
PRED4x4_TM_MMX mmxext

cglobal pred4x4_tm_vp8_ssse3, 3,3
    sub         r0, r2
    movq       mm6, [tm_shuf]
    pxor       mm1, mm1
    movd       mm0, [r0]
    punpcklbw  mm0, mm1
    movd       mm7, [r0-4]
    pshufb     mm7, mm6
    lea         r1, [r0+r2*2]
    movd       mm2, [r0+r2*1-4]
    movd       mm3, [r0+r2*2-4]
    movd       mm4, [r1+r2*1-4]
    movd       mm5, [r1+r2*2-4]
    pshufb     mm2, mm6
    pshufb     mm3, mm6
    pshufb     mm4, mm6
    pshufb     mm5, mm6
    psubw      mm2, mm7
    psubw      mm3, mm7
    psubw      mm4, mm7
    psubw      mm5, mm7
    paddw      mm2, mm0
    paddw      mm3, mm0
    paddw      mm4, mm0
    paddw      mm5, mm0
    packuswb   mm2, mm2
    packuswb   mm3, mm3
    packuswb   mm4, mm4
    packuswb   mm5, mm5
    movd [r0+r2*1], mm2
    movd [r0+r2*2], mm3
    movd [r1+r2*1], mm4
    movd [r1+r2*2], mm5
    RET
2587 2588 2589 2590 2591 2592 2593 2594 2595 2596 2597 2598 2599 2600 2601 2602 2603 2604 2605 2606

;-----------------------------------------------------------------------------
; void pred4x4_vertical_vp8_mmxext(uint8_t *src, const uint8_t *topright, int stride)
;-----------------------------------------------------------------------------

INIT_MMX
cglobal pred4x4_vertical_vp8_mmxext, 3,3
    sub       r0, r2
    movd      m1, [r0-1]
    movd      m0, [r0]
    mova      m2, m0   ;t0 t1 t2 t3
    punpckldq m0, [r1] ;t0 t1 t2 t3 t4 t5 t6 t7
    lea       r1, [r0+r2*2]
    psrlq     m0, 8    ;t1 t2 t3 t4
    PRED4x4_LOWPASS m3, m1, m0, m2, m4
    movd [r0+r2*1], m3
    movd [r0+r2*2], m3
    movd [r1+r2*1], m3
    movd [r1+r2*2], m3
    RET
2607 2608 2609 2610

;-----------------------------------------------------------------------------
; void pred4x4_down_left_mmxext(uint8_t *src, const uint8_t *topright, int stride)
;-----------------------------------------------------------------------------
2611
%ifdef CONFIG_GPL
2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625 2626 2627 2628 2629 2630 2631 2632 2633 2634
INIT_MMX
cglobal pred4x4_down_left_mmxext, 3,3
    sub       r0, r2
    movq      m1, [r0]
    punpckldq m1, [r1]
    movq      m2, m1
    movq      m3, m1
    movq      m4, m1
    psllq     m1, 8
    pxor      m2, m1
    psrlq     m2, 8
    pxor      m3, m2
    PRED4x4_LOWPASS m0, m1, m3, m4, m5
    lea       r1, [r0+r2*2]
    psrlq     m0, 8
    movd      [r0+r2*1], m0
    psrlq     m0, 8
    movd      [r0+r2*2], m0
    psrlq     m0, 8
    movd      [r1+r2*1], m0
    psrlq     m0, 8
    movd      [r1+r2*2], m0
    RET
2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659

;-----------------------------------------------------------------------------
; void pred4x4_vertical_left_mmxext(uint8_t *src, const uint8_t *topright, int stride)
;-----------------------------------------------------------------------------

INIT_MMX
cglobal pred4x4_vertical_left_mmxext, 3,3
    sub       r0, r2
    movq      m1, [r0]
    punpckldq m1, [r1]
    movq      m3, m1
    movq      m2, m1
    psrlq     m3, 8
    psrlq     m2, 16
    movq      m4, m3
    pavgb     m4, m1
    PRED4x4_LOWPASS m0, m1, m2, m3, m5
    lea       r1, [r0+r2*2]
    movh      [r0+r2*1], m4
    movh      [r0+r2*2], m0
    psrlq     m4, 8
    psrlq     m0, 8
    movh      [r1+r2*1], m4
    movh      [r1+r2*2], m0
    RET
2660 2661 2662 2663 2664 2665 2666 2667 2668

;-----------------------------------------------------------------------------
; void pred4x4_horizontal_up_mmxext(uint8_t *src, const uint8_t *topright, int stride)
;-----------------------------------------------------------------------------

INIT_MMX
cglobal pred4x4_horizontal_up_mmxext, 3,3
    sub       r0, r2
    lea       r1, [r0+r2*2]
2669 2670 2671 2672
    movd      m0, [r0+r2*1-4]
    punpcklbw m0, [r0+r2*2-4]
    movd      m1, [r1+r2*1-4]
    punpcklbw m1, [r1+r2*2-4]
2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692
    punpckhwd m0, m1
    movq      m1, m0
    punpckhbw m1, m1
    pshufw    m1, m1, 0xFF
    punpckhdq m0, m1
    movq      m2, m0
    movq      m3, m0
    movq      m7, m0
    psrlq     m2, 16
    psrlq     m3, 8
    pavgb     m7, m3
    PRED4x4_LOWPASS m4, m0, m2, m3, m5
    punpcklbw m7, m4
    movd    [r0+r2*1], m7
    psrlq    m7, 16
    movd    [r0+r2*2], m7
    psrlq    m7, 16
    movd    [r1+r2*1], m7
    movd    [r1+r2*2], m1
    RET
2693 2694 2695 2696 2697 2698 2699 2700 2701 2702 2703 2704 2705

;-----------------------------------------------------------------------------
; void pred4x4_horizontal_down_mmxext(uint8_t *src, const uint8_t *topright, int stride)
;-----------------------------------------------------------------------------

INIT_MMX
%define PALIGNR PALIGNR_MMX
cglobal pred4x4_horizontal_down_mmxext, 3,3
    sub       r0, r2
    lea       r1, [r0+r2*2]
    movh      m0, [r0-4]      ; lt ..
    punpckldq m0, [r0]        ; t3 t2 t1 t0 lt .. .. ..
    psllq     m0, 8           ; t2 t1 t0 lt .. .. .. ..
2706 2707 2708 2709
    movd      m1, [r1+r2*2-4] ; l3
    punpcklbw m1, [r1+r2*1-4] ; l2 l3
    movd      m2, [r0+r2*2-4] ; l1
    punpcklbw m2, [r0+r2*1-4] ; l0 l1
2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728
    punpckhwd m1, m2          ; l0 l1 l2 l3
    punpckhdq m1, m0          ; t2 t1 t0 lt l0 l1 l2 l3
    movq      m0, m1
    movq      m2, m1
    movq      m5, m1
    psrlq     m0, 16          ; .. .. t2 t1 t0 lt l0 l1
    psrlq     m2, 8           ; .. t2 t1 t0 lt l0 l1 l2
    pavgb     m5, m2
    PRED4x4_LOWPASS m3, m1, m0, m2, m4
    punpcklbw m5, m3
    psrlq     m3, 32
    PALIGNR   m3, m5, 6, m4
    movh      [r1+r2*2], m5
    psrlq     m5, 16
    movh      [r1+r2*1], m5
    psrlq     m5, 16
    movh      [r0+r2*2], m5
    movh      [r0+r2*1], m3
    RET
2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751 2752 2753 2754 2755 2756 2757 2758 2759

;-----------------------------------------------------------------------------
; void pred4x4_vertical_right_mmxext(uint8_t *src, const uint8_t *topright, int stride)
;-----------------------------------------------------------------------------

INIT_MMX
%define PALIGNR PALIGNR_MMX
cglobal pred4x4_vertical_right_mmxext, 3,3
    sub     r0, r2
    lea     r1, [r0+r2*2]
    movh    m0, [r0]                    ; ........t3t2t1t0
    movq    m5, m0
    PALIGNR m0, [r0-8], 7, m1           ; ......t3t2t1t0lt
    pavgb   m5, m0
    PALIGNR m0, [r0+r2*1-8], 7, m1      ; ....t3t2t1t0ltl0
    movq    m1, m0
    PALIGNR m0, [r0+r2*2-8], 7, m2      ; ..t3t2t1t0ltl0l1
    movq    m2, m0
    PALIGNR m0, [r1+r2*1-8], 7, m3      ; t3t2t1t0ltl0l1l2
    PRED4x4_LOWPASS m3, m1, m0, m2, m4
    movq    m1, m3
    psrlq   m3, 16
    psllq   m1, 48
    movh    [r0+r2*1], m5
    movh    [r0+r2*2], m3
    PALIGNR m5, m1, 7, m2
    psllq   m1, 8
    movh    [r1+r2*1], m5
    PALIGNR m3, m1, 7, m1
    movh    [r1+r2*2], m3
    RET
2760 2761 2762 2763 2764 2765 2766 2767 2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787 2788

;-----------------------------------------------------------------------------
; void pred4x4_down_right_mmxext(uint8_t *src, const uint8_t *topright, int stride)
;-----------------------------------------------------------------------------

INIT_MMX
%define PALIGNR PALIGNR_MMX
cglobal pred4x4_down_right_mmxext, 3,3
    sub       r0, r2
    lea       r1, [r0+r2*2]
    movq      m1, [r1-8]
    movq      m2, [r0+r2*1-8]
    punpckhbw m2, [r0-8]
    movh      m3, [r0]
    punpckhwd m1, m2
    PALIGNR   m3, m1, 5, m1
    movq      m1, m3
    PALIGNR   m3, [r1+r2*1-8], 7, m4
    movq      m2, m3
    PALIGNR   m3, [r1+r2*2-8], 7, m4
    PRED4x4_LOWPASS m0, m3, m1, m2, m4
    movh      [r1+r2*2], m0
    psrlq     m0, 8
    movh      [r1+r2*1], m0
    psrlq     m0, 8
    movh      [r0+r2*2], m0
    psrlq     m0, 8
    movh      [r0+r2*1], m0
    RET
2789
%endif