output.c 97.6 KB
Newer Older
1
/*
2
 * Copyright (C) 2001-2012 Michael Niedermayer <michaelni@gmx.at>
3
 *
4
 * This file is part of FFmpeg.
5
 *
6
 * FFmpeg is free software; you can redistribute it and/or
7 8 9 10
 * modify it under the terms of the GNU Lesser General Public
 * License as published by the Free Software Foundation; either
 * version 2.1 of the License, or (at your option) any later version.
 *
11
 * FFmpeg is distributed in the hope that it will be useful,
12 13 14 15 16
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * Lesser General Public License for more details.
 *
 * You should have received a copy of the GNU Lesser General Public
17
 * License along with FFmpeg; if not, write to the Free Software
18 19 20 21 22 23 24 25
 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
 */

#include <math.h>
#include <stdint.h>
#include <stdio.h>
#include <string.h>

26
#include "libavutil/attributes.h"
27
#include "libavutil/avutil.h"
28
#include "libavutil/avassert.h"
29 30 31 32 33 34 35 36 37 38
#include "libavutil/bswap.h"
#include "libavutil/cpu.h"
#include "libavutil/intreadwrite.h"
#include "libavutil/mathematics.h"
#include "libavutil/pixdesc.h"
#include "config.h"
#include "rgb2rgb.h"
#include "swscale.h"
#include "swscale_internal.h"

39
DECLARE_ALIGNED(8, const uint8_t, ff_dither_2x2_4)[][8] = {
40 41
{  1,   3,   1,   3,   1,   3,   1,   3, },
{  2,   0,   2,   0,   2,   0,   2,   0, },
42
{  1,   3,   1,   3,   1,   3,   1,   3, },
43 44
};

45
DECLARE_ALIGNED(8, const uint8_t, ff_dither_2x2_8)[][8] = {
46 47
{  6,   2,   6,   2,   6,   2,   6,   2, },
{  0,   4,   0,   4,   0,   4,   0,   4, },
48
{  6,   2,   6,   2,   6,   2,   6,   2, },
49 50
};

51
DECLARE_ALIGNED(8, const uint8_t, ff_dither_4x4_16)[][8] = {
52 53 54 55
{  8,   4,  11,   7,   8,   4,  11,   7, },
{  2,  14,   1,  13,   2,  14,   1,  13, },
{ 10,   6,   9,   5,  10,   6,   9,   5, },
{  0,  12,   3,  15,   0,  12,   3,  15, },
56
{  8,   4,  11,   7,   8,   4,  11,   7, },
57 58
};

59
DECLARE_ALIGNED(8, const uint8_t, ff_dither_8x8_32)[][8] = {
60 61 62 63 64 65 66 67
{ 17,   9,  23,  15,  16,   8,  22,  14, },
{  5,  29,   3,  27,   4,  28,   2,  26, },
{ 21,  13,  19,  11,  20,  12,  18,  10, },
{  0,  24,   6,  30,   1,  25,   7,  31, },
{ 16,   8,  22,  14,  17,   9,  23,  15, },
{  4,  28,   2,  26,   5,  29,   3,  27, },
{ 20,  12,  18,  10,  21,  13,  19,  11, },
{  1,  25,   7,  31,   0,  24,   6,  30, },
68
{ 17,   9,  23,  15,  16,   8,  22,  14, },
69 70
};

71
DECLARE_ALIGNED(8, const uint8_t, ff_dither_8x8_73)[][8] = {
72 73 74 75 76 77 78 79
{  0,  55,  14,  68,   3,  58,  17,  72, },
{ 37,  18,  50,  32,  40,  22,  54,  35, },
{  9,  64,   5,  59,  13,  67,   8,  63, },
{ 46,  27,  41,  23,  49,  31,  44,  26, },
{  2,  57,  16,  71,   1,  56,  15,  70, },
{ 39,  21,  52,  34,  38,  19,  51,  33, },
{ 11,  66,   7,  62,  10,  65,   6,  60, },
{ 48,  30,  43,  25,  47,  29,  42,  24, },
80
{  0,  55,  14,  68,   3,  58,  17,  72, },
81 82 83
};

#if 1
84
DECLARE_ALIGNED(8, const uint8_t, ff_dither_8x8_220)[][8] = {
85 86 87 88 89 90 91 92
{117,  62, 158, 103, 113,  58, 155, 100, },
{ 34, 199,  21, 186,  31, 196,  17, 182, },
{144,  89, 131,  76, 141,  86, 127,  72, },
{  0, 165,  41, 206,  10, 175,  52, 217, },
{110,  55, 151,  96, 120,  65, 162, 107, },
{ 28, 193,  14, 179,  38, 203,  24, 189, },
{138,  83, 124,  69, 148,  93, 134,  79, },
{  7, 172,  48, 213,   3, 168,  45, 210, },
93
{117,  62, 158, 103, 113,  58, 155, 100, },
94 95 96
};
#elif 1
// tries to correct a gamma of 1.5
97
DECLARE_ALIGNED(8, const uint8_t, ff_dither_8x8_220)[][8] = {
98 99 100 101 102 103 104 105
{  0, 143,  18, 200,   2, 156,  25, 215, },
{ 78,  28, 125,  64,  89,  36, 138,  74, },
{ 10, 180,   3, 161,  16, 195,   8, 175, },
{109,  51,  93,  38, 121,  60, 105,  47, },
{  1, 152,  23, 210,   0, 147,  20, 205, },
{ 85,  33, 134,  71,  81,  30, 130,  67, },
{ 14, 190,   6, 171,  12, 185,   5, 166, },
{117,  57, 101,  44, 113,  54,  97,  41, },
106
{  0, 143,  18, 200,   2, 156,  25, 215, },
107 108 109
};
#elif 1
// tries to correct a gamma of 2.0
110
DECLARE_ALIGNED(8, const uint8_t, ff_dither_8x8_220)[][8] = {
111 112 113 114 115 116 117 118
{  0, 124,   8, 193,   0, 140,  12, 213, },
{ 55,  14, 104,  42,  66,  19, 119,  52, },
{  3, 168,   1, 145,   6, 187,   3, 162, },
{ 86,  31,  70,  21,  99,  39,  82,  28, },
{  0, 134,  11, 206,   0, 129,   9, 200, },
{ 62,  17, 114,  48,  58,  16, 109,  45, },
{  5, 181,   2, 157,   4, 175,   1, 151, },
{ 95,  36,  78,  26,  90,  34,  74,  24, },
119
{  0, 124,   8, 193,   0, 140,  12, 213, },
120 121 122
};
#else
// tries to correct a gamma of 2.5
123
DECLARE_ALIGNED(8, const uint8_t, ff_dither_8x8_220)[][8] = {
124 125 126 127 128 129 130 131
{  0, 107,   3, 187,   0, 125,   6, 212, },
{ 39,   7,  86,  28,  49,  11, 102,  36, },
{  1, 158,   0, 131,   3, 180,   1, 151, },
{ 68,  19,  52,  12,  81,  25,  64,  17, },
{  0, 119,   5, 203,   0, 113,   4, 195, },
{ 45,   9,  96,  33,  42,   8,  91,  30, },
{  2, 172,   1, 144,   2, 165,   0, 137, },
{ 77,  23,  60,  15,  72,  21,  56,  14, },
132
{  0, 107,   3, 187,   0, 125,   6, 212, },
133 134 135 136 137 138 139 140 141 142 143 144 145 146 147
};
#endif

#define output_pixel(pos, val, bias, signedness) \
    if (big_endian) { \
        AV_WB16(pos, bias + av_clip_ ## signedness ## 16(val >> shift)); \
    } else { \
        AV_WL16(pos, bias + av_clip_ ## signedness ## 16(val >> shift)); \
    }

static av_always_inline void
yuv2plane1_16_c_template(const int32_t *src, uint16_t *dest, int dstW,
                         int big_endian, int output_bits)
{
    int i;
148 149
    int shift = 3;
    av_assert0(output_bits == 16);
150 151 152 153 154 155 156 157 158 159 160 161 162

    for (i = 0; i < dstW; i++) {
        int val = src[i] + (1 << (shift - 1));
        output_pixel(&dest[i], val, 0, uint);
    }
}

static av_always_inline void
yuv2planeX_16_c_template(const int16_t *filter, int filterSize,
                         const int32_t **src, uint16_t *dest, int dstW,
                         int big_endian, int output_bits)
{
    int i;
163 164
    int shift = 15;
    av_assert0(output_bits == 16);
165 166

    for (i = 0; i < dstW; i++) {
167
        int val = 1 << (shift - 1);
168 169 170 171 172 173 174 175 176
        int j;

        /* range of val is [0,0x7FFFFFFF], so 31 bits, but with lanczos/spline
         * filters (or anything with negative coeffs, the range can be slightly
         * wider in both directions. To account for this overflow, we subtract
         * a constant so it always fits in the signed range (assuming a
         * reasonable filterSize), and re-add that at the end. */
        val -= 0x40000000;
        for (j = 0; j < filterSize; j++)
177
            val += src[j][i] * (unsigned)filter[j];
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

        output_pixel(&dest[i], val, 0x8000, int);
    }
}

#undef output_pixel

#define output_pixel(pos, val) \
    if (big_endian) { \
        AV_WB16(pos, av_clip_uintp2(val >> shift, output_bits)); \
    } else { \
        AV_WL16(pos, av_clip_uintp2(val >> shift, output_bits)); \
    }

static av_always_inline void
yuv2plane1_10_c_template(const int16_t *src, uint16_t *dest, int dstW,
                         int big_endian, int output_bits)
{
    int i;
    int shift = 15 - output_bits;

    for (i = 0; i < dstW; i++) {
        int val = src[i] + (1 << (shift - 1));
        output_pixel(&dest[i], val);
    }
}

static av_always_inline void
yuv2planeX_10_c_template(const int16_t *filter, int filterSize,
                         const int16_t **src, uint16_t *dest, int dstW,
                         int big_endian, int output_bits)
{
    int i;
    int shift = 11 + 16 - output_bits;

    for (i = 0; i < dstW; i++) {
214
        int val = 1 << (shift - 1);
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
        int j;

        for (j = 0; j < filterSize; j++)
            val += src[j][i] * filter[j];

        output_pixel(&dest[i], val);
    }
}

#undef output_pixel

#define yuv2NBPS(bits, BE_LE, is_be, template_size, typeX_t) \
static void yuv2plane1_ ## bits ## BE_LE ## _c(const int16_t *src, \
                              uint8_t *dest, int dstW, \
                              const uint8_t *dither, int offset)\
{ \
    yuv2plane1_ ## template_size ## _c_template((const typeX_t *) src, \
                         (uint16_t *) dest, dstW, is_be, bits); \
}\
static void yuv2planeX_ ## bits ## BE_LE ## _c(const int16_t *filter, int filterSize, \
                              const int16_t **src, uint8_t *dest, int dstW, \
                              const uint8_t *dither, int offset)\
{ \
    yuv2planeX_## template_size ## _c_template(filter, \
                         filterSize, (const typeX_t **) src, \
                         (uint16_t *) dest, dstW, is_be, bits); \
}
yuv2NBPS( 9, BE, 1, 10, int16_t)
yuv2NBPS( 9, LE, 0, 10, int16_t)
yuv2NBPS(10, BE, 1, 10, int16_t)
yuv2NBPS(10, LE, 0, 10, int16_t)
246 247 248 249
yuv2NBPS(12, BE, 1, 10, int16_t)
yuv2NBPS(12, LE, 0, 10, int16_t)
yuv2NBPS(14, BE, 1, 10, int16_t)
yuv2NBPS(14, LE, 0, 10, int16_t)
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
yuv2NBPS(16, BE, 1, 16, int32_t)
yuv2NBPS(16, LE, 0, 16, int32_t)

static void yuv2planeX_8_c(const int16_t *filter, int filterSize,
                           const int16_t **src, uint8_t *dest, int dstW,
                           const uint8_t *dither, int offset)
{
    int i;
    for (i=0; i<dstW; i++) {
        int val = dither[(i + offset) & 7] << 12;
        int j;
        for (j=0; j<filterSize; j++)
            val += src[j][i] * filter[j];

        dest[i]= av_clip_uint8(val>>19);
    }
}

static void yuv2plane1_8_c(const int16_t *src, uint8_t *dest, int dstW,
                           const uint8_t *dither, int offset)
{
    int i;
    for (i=0; i<dstW; i++) {
        int val = (src[i] + dither[(i + offset) & 7]) >> 7;
        dest[i]= av_clip_uint8(val);
    }
}

static void yuv2nv12cX_c(SwsContext *c, const int16_t *chrFilter, int chrFilterSize,
                        const int16_t **chrUSrc, const int16_t **chrVSrc,
                        uint8_t *dest, int chrDstW)
{
282
    enum AVPixelFormat dstFormat = c->dstFormat;
283 284 285
    const uint8_t *chrDither = c->chrDither8;
    int i;

286
    if (dstFormat == AV_PIX_FMT_NV12)
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
        for (i=0; i<chrDstW; i++) {
            int u = chrDither[i & 7] << 12;
            int v = chrDither[(i + 3) & 7] << 12;
            int j;
            for (j=0; j<chrFilterSize; j++) {
                u += chrUSrc[j][i] * chrFilter[j];
                v += chrVSrc[j][i] * chrFilter[j];
            }

            dest[2*i]= av_clip_uint8(u>>19);
            dest[2*i+1]= av_clip_uint8(v>>19);
        }
    else
        for (i=0; i<chrDstW; i++) {
            int u = chrDither[i & 7] << 12;
            int v = chrDither[(i + 3) & 7] << 12;
            int j;
            for (j=0; j<chrFilterSize; j++) {
                u += chrUSrc[j][i] * chrFilter[j];
                v += chrVSrc[j][i] * chrFilter[j];
            }

            dest[2*i]= av_clip_uint8(v>>19);
            dest[2*i+1]= av_clip_uint8(u>>19);
        }
}

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

#define output_pixel(pos, val) \
    if (big_endian) { \
        AV_WB16(pos, av_clip_uintp2(val >> shift, 10) << 6); \
    } else { \
        AV_WL16(pos, av_clip_uintp2(val >> shift, 10) << 6); \
    }

static void yuv2p010l1_c(const int16_t *src,
                         uint16_t *dest, int dstW,
                         int big_endian)
{
    int i;
    int shift = 5;

    for (i = 0; i < dstW; i++) {
        int val = src[i] + (1 << (shift - 1));
        output_pixel(&dest[i], val);
    }
}

static void yuv2p010lX_c(const int16_t *filter, int filterSize,
                         const int16_t **src, uint16_t *dest, int dstW,
                         int big_endian)
{
    int i, j;
    int shift = 17;

    for (i = 0; i < dstW; i++) {
        int val = 1 << (shift - 1);

        for (j = 0; j < filterSize; j++)
            val += src[j][i] * filter[j];

        output_pixel(&dest[i], val);
    }
}

static void yuv2p010cX_c(SwsContext *c, const int16_t *chrFilter, int chrFilterSize,
                         const int16_t **chrUSrc, const int16_t **chrVSrc,
                         uint8_t *dest8, int chrDstW)
{
    uint16_t *dest = (uint16_t*)dest8;
    int shift = 17;
    int big_endian = c->dstFormat == AV_PIX_FMT_P010BE;
    int i, j;

    for (i = 0; i < chrDstW; i++) {
        int u = 1 << (shift - 1);
        int v = 1 << (shift - 1);

        for (j = 0; j < chrFilterSize; j++) {
            u += chrUSrc[j][i] * chrFilter[j];
            v += chrVSrc[j][i] * chrFilter[j];
        }

        output_pixel(&dest[2*i]  , u);
        output_pixel(&dest[2*i+1], v);
    }
}

static void yuv2p010l1_LE_c(const int16_t *src,
                            uint8_t *dest, int dstW,
                            const uint8_t *dither, int offset)
{
    yuv2p010l1_c(src, (uint16_t*)dest, dstW, 0);
}

static void yuv2p010l1_BE_c(const int16_t *src,
                            uint8_t *dest, int dstW,
                            const uint8_t *dither, int offset)
{
    yuv2p010l1_c(src, (uint16_t*)dest, dstW, 1);
}

static void yuv2p010lX_LE_c(const int16_t *filter, int filterSize,
                            const int16_t **src, uint8_t *dest, int dstW,
                            const uint8_t *dither, int offset)
{
    yuv2p010lX_c(filter, filterSize, src, (uint16_t*)dest, dstW, 0);
}

static void yuv2p010lX_BE_c(const int16_t *filter, int filterSize,
                            const int16_t **src, uint8_t *dest, int dstW,
                            const uint8_t *dither, int offset)
{
    yuv2p010lX_c(filter, filterSize, src, (uint16_t*)dest, dstW, 1);
}

#undef output_pixel


406 407
#define accumulate_bit(acc, val) \
    acc <<= 1; \
408
    acc |= (val) >= 234
409
#define output_pixel(pos, acc) \
410
    if (target == AV_PIX_FMT_MONOBLACK) { \
411 412 413 414 415 416 417 418 419 420 421
        pos = acc; \
    } else { \
        pos = ~acc; \
    }

static av_always_inline void
yuv2mono_X_c_template(SwsContext *c, const int16_t *lumFilter,
                      const int16_t **lumSrc, int lumFilterSize,
                      const int16_t *chrFilter, const int16_t **chrUSrc,
                      const int16_t **chrVSrc, int chrFilterSize,
                      const int16_t **alpSrc, uint8_t *dest, int dstW,
422
                      int y, enum AVPixelFormat target)
423
{
424
    const uint8_t * const d128 = ff_dither_8x8_220[y&7];
425 426
    int i;
    unsigned acc = 0;
427
    int err = 0;
428

429
    for (i = 0; i < dstW; i += 2) {
430 431 432 433 434 435 436 437 438 439 440 441 442 443
        int j;
        int Y1 = 1 << 18;
        int Y2 = 1 << 18;

        for (j = 0; j < lumFilterSize; j++) {
            Y1 += lumSrc[j][i]   * lumFilter[j];
            Y2 += lumSrc[j][i+1] * lumFilter[j];
        }
        Y1 >>= 19;
        Y2 >>= 19;
        if ((Y1 | Y2) & 0x100) {
            Y1 = av_clip_uint8(Y1);
            Y2 = av_clip_uint8(Y2);
        }
444
        if (c->dither == SWS_DITHER_ED) {
445
            Y1 += (7*err + 1*c->dither_error[0][i] + 5*c->dither_error[0][i+1] + 3*c->dither_error[0][i+2] + 8 - 256)>>4;
446 447 448 449
            c->dither_error[0][i] = err;
            acc = 2*acc + (Y1 >= 128);
            Y1 -= 220*(acc&1);

450
            err = Y2 + ((7*Y1 + 1*c->dither_error[0][i+1] + 5*c->dither_error[0][i+2] + 3*c->dither_error[0][i+3] + 8 - 256)>>4);
451 452 453 454 455 456 457
            c->dither_error[0][i+1] = Y1;
            acc = 2*acc + (err >= 128);
            err -= 220*(acc&1);
        } else {
            accumulate_bit(acc, Y1 + d128[(i + 0) & 7]);
            accumulate_bit(acc, Y2 + d128[(i + 1) & 7]);
        }
458 459 460 461
        if ((i & 7) == 6) {
            output_pixel(*dest++, acc);
        }
    }
462
    c->dither_error[0][i] = err;
463 464 465 466

    if (i & 6) {
        output_pixel(*dest, acc);
    }
467 468 469 470 471 472 473
}

static av_always_inline void
yuv2mono_2_c_template(SwsContext *c, const int16_t *buf[2],
                      const int16_t *ubuf[2], const int16_t *vbuf[2],
                      const int16_t *abuf[2], uint8_t *dest, int dstW,
                      int yalpha, int uvalpha, int y,
474
                      enum AVPixelFormat target)
475 476
{
    const int16_t *buf0  = buf[0],  *buf1  = buf[1];
477
    const uint8_t * const d128 = ff_dither_8x8_220[y & 7];
478
    int  yalpha1 = 4096 - yalpha;
479
    int i;
480
    av_assert2(yalpha  <= 4096U);
481

482
    if (c->dither == SWS_DITHER_ED) {
483 484 485 486 487 488
        int err = 0;
        int acc = 0;
        for (i = 0; i < dstW; i +=2) {
            int Y;

            Y = (buf0[i + 0] * yalpha1 + buf1[i + 0] * yalpha) >> 19;
489
            Y += (7*err + 1*c->dither_error[0][i] + 5*c->dither_error[0][i+1] + 3*c->dither_error[0][i+2] + 8 - 256)>>4;
490 491 492 493 494
            c->dither_error[0][i] = err;
            acc = 2*acc + (Y >= 128);
            Y -= 220*(acc&1);

            err = (buf0[i + 1] * yalpha1 + buf1[i + 1] * yalpha) >> 19;
495
            err += (7*Y + 1*c->dither_error[0][i+1] + 5*c->dither_error[0][i+2] + 3*c->dither_error[0][i+3] + 8 - 256)>>4;
496 497 498 499 500 501 502 503 504
            c->dither_error[0][i+1] = Y;
            acc = 2*acc + (err >= 128);
            err -= 220*(acc&1);

            if ((i & 7) == 6)
                output_pixel(*dest++, acc);
        }
        c->dither_error[0][i] = err;
    } else {
505
    for (i = 0; i < dstW; i += 8) {
506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524
        int Y, acc = 0;

        Y = (buf0[i + 0] * yalpha1 + buf1[i + 0] * yalpha) >> 19;
        accumulate_bit(acc, Y + d128[0]);
        Y = (buf0[i + 1] * yalpha1 + buf1[i + 1] * yalpha) >> 19;
        accumulate_bit(acc, Y + d128[1]);
        Y = (buf0[i + 2] * yalpha1 + buf1[i + 2] * yalpha) >> 19;
        accumulate_bit(acc, Y + d128[2]);
        Y = (buf0[i + 3] * yalpha1 + buf1[i + 3] * yalpha) >> 19;
        accumulate_bit(acc, Y + d128[3]);
        Y = (buf0[i + 4] * yalpha1 + buf1[i + 4] * yalpha) >> 19;
        accumulate_bit(acc, Y + d128[4]);
        Y = (buf0[i + 5] * yalpha1 + buf1[i + 5] * yalpha) >> 19;
        accumulate_bit(acc, Y + d128[5]);
        Y = (buf0[i + 6] * yalpha1 + buf1[i + 6] * yalpha) >> 19;
        accumulate_bit(acc, Y + d128[6]);
        Y = (buf0[i + 7] * yalpha1 + buf1[i + 7] * yalpha) >> 19;
        accumulate_bit(acc, Y + d128[7]);

525 526
        output_pixel(*dest++, acc);
    }
527
    }
528 529 530 531 532 533
}

static av_always_inline void
yuv2mono_1_c_template(SwsContext *c, const int16_t *buf0,
                      const int16_t *ubuf[2], const int16_t *vbuf[2],
                      const int16_t *abuf0, uint8_t *dest, int dstW,
534
                      int uvalpha, int y, enum AVPixelFormat target)
535
{
536
    const uint8_t * const d128 = ff_dither_8x8_220[y & 7];
537 538
    int i;

539
    if (c->dither == SWS_DITHER_ED) {
540 541 542 543 544 545
        int err = 0;
        int acc = 0;
        for (i = 0; i < dstW; i +=2) {
            int Y;

            Y = ((buf0[i + 0] + 64) >> 7);
546
            Y += (7*err + 1*c->dither_error[0][i] + 5*c->dither_error[0][i+1] + 3*c->dither_error[0][i+2] + 8 - 256)>>4;
547 548 549 550 551
            c->dither_error[0][i] = err;
            acc = 2*acc + (Y >= 128);
            Y -= 220*(acc&1);

            err = ((buf0[i + 1] + 64) >> 7);
552
            err += (7*Y + 1*c->dither_error[0][i+1] + 5*c->dither_error[0][i+2] + 3*c->dither_error[0][i+3] + 8 - 256)>>4;
553 554 555 556 557 558 559 560 561
            c->dither_error[0][i+1] = Y;
            acc = 2*acc + (err >= 128);
            err -= 220*(acc&1);

            if ((i & 7) == 6)
                output_pixel(*dest++, acc);
        }
        c->dither_error[0][i] = err;
    } else {
562
    for (i = 0; i < dstW; i += 8) {
563
        int acc = 0;
564 565 566 567 568 569 570 571
        accumulate_bit(acc, ((buf0[i + 0] + 64) >> 7) + d128[0]);
        accumulate_bit(acc, ((buf0[i + 1] + 64) >> 7) + d128[1]);
        accumulate_bit(acc, ((buf0[i + 2] + 64) >> 7) + d128[2]);
        accumulate_bit(acc, ((buf0[i + 3] + 64) >> 7) + d128[3]);
        accumulate_bit(acc, ((buf0[i + 4] + 64) >> 7) + d128[4]);
        accumulate_bit(acc, ((buf0[i + 5] + 64) >> 7) + d128[5]);
        accumulate_bit(acc, ((buf0[i + 6] + 64) >> 7) + d128[6]);
        accumulate_bit(acc, ((buf0[i + 7] + 64) >> 7) + d128[7]);
572

573 574
        output_pixel(*dest++, acc);
    }
575
    }
576 577 578
}

#undef output_pixel
579
#undef accumulate_bit
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

#define YUV2PACKEDWRAPPER(name, base, ext, fmt) \
static void name ## ext ## _X_c(SwsContext *c, const int16_t *lumFilter, \
                                const int16_t **lumSrc, int lumFilterSize, \
                                const int16_t *chrFilter, const int16_t **chrUSrc, \
                                const int16_t **chrVSrc, int chrFilterSize, \
                                const int16_t **alpSrc, uint8_t *dest, int dstW, \
                                int y) \
{ \
    name ## base ## _X_c_template(c, lumFilter, lumSrc, lumFilterSize, \
                                  chrFilter, chrUSrc, chrVSrc, chrFilterSize, \
                                  alpSrc, dest, dstW, y, fmt); \
} \
 \
static void name ## ext ## _2_c(SwsContext *c, const int16_t *buf[2], \
                                const int16_t *ubuf[2], const int16_t *vbuf[2], \
                                const int16_t *abuf[2], uint8_t *dest, int dstW, \
                                int yalpha, int uvalpha, int y) \
{ \
    name ## base ## _2_c_template(c, buf, ubuf, vbuf, abuf, \
                                  dest, dstW, yalpha, uvalpha, y, fmt); \
} \
 \
static void name ## ext ## _1_c(SwsContext *c, const int16_t *buf0, \
                                const int16_t *ubuf[2], const int16_t *vbuf[2], \
                                const int16_t *abuf0, uint8_t *dest, int dstW, \
                                int uvalpha, int y) \
{ \
    name ## base ## _1_c_template(c, buf0, ubuf, vbuf, \
                                  abuf0, dest, dstW, uvalpha, \
                                  y, fmt); \
}

613 614
YUV2PACKEDWRAPPER(yuv2mono,, white, AV_PIX_FMT_MONOWHITE)
YUV2PACKEDWRAPPER(yuv2mono,, black, AV_PIX_FMT_MONOBLACK)
615 616

#define output_pixels(pos, Y1, U, Y2, V) \
617
    if (target == AV_PIX_FMT_YUYV422) { \
618 619 620 621
        dest[pos + 0] = Y1; \
        dest[pos + 1] = U;  \
        dest[pos + 2] = Y2; \
        dest[pos + 3] = V;  \
622 623 624 625 626 627
    } else if (target == AV_PIX_FMT_YVYU422) { \
        dest[pos + 0] = Y1; \
        dest[pos + 1] = V;  \
        dest[pos + 2] = Y2; \
        dest[pos + 3] = U;  \
    } else { /* AV_PIX_FMT_UYVY422 */ \
628 629 630 631 632 633 634 635 636 637 638 639
        dest[pos + 0] = U;  \
        dest[pos + 1] = Y1; \
        dest[pos + 2] = V;  \
        dest[pos + 3] = Y2; \
    }

static av_always_inline void
yuv2422_X_c_template(SwsContext *c, const int16_t *lumFilter,
                     const int16_t **lumSrc, int lumFilterSize,
                     const int16_t *chrFilter, const int16_t **chrUSrc,
                     const int16_t **chrVSrc, int chrFilterSize,
                     const int16_t **alpSrc, uint8_t *dest, int dstW,
640
                     int y, enum AVPixelFormat target)
641 642 643
{
    int i;

644
    for (i = 0; i < ((dstW + 1) >> 1); i++) {
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
        int j;
        int Y1 = 1 << 18;
        int Y2 = 1 << 18;
        int U  = 1 << 18;
        int V  = 1 << 18;

        for (j = 0; j < lumFilterSize; j++) {
            Y1 += lumSrc[j][i * 2]     * lumFilter[j];
            Y2 += lumSrc[j][i * 2 + 1] * lumFilter[j];
        }
        for (j = 0; j < chrFilterSize; j++) {
            U += chrUSrc[j][i] * chrFilter[j];
            V += chrVSrc[j][i] * chrFilter[j];
        }
        Y1 >>= 19;
        Y2 >>= 19;
        U  >>= 19;
        V  >>= 19;
        if ((Y1 | Y2 | U | V) & 0x100) {
            Y1 = av_clip_uint8(Y1);
            Y2 = av_clip_uint8(Y2);
            U  = av_clip_uint8(U);
            V  = av_clip_uint8(V);
        }
        output_pixels(4*i, Y1, U, Y2, V);
    }
}

static av_always_inline void
yuv2422_2_c_template(SwsContext *c, const int16_t *buf[2],
                     const int16_t *ubuf[2], const int16_t *vbuf[2],
                     const int16_t *abuf[2], uint8_t *dest, int dstW,
                     int yalpha, int uvalpha, int y,
678
                     enum AVPixelFormat target)
679 680 681 682
{
    const int16_t *buf0  = buf[0],  *buf1  = buf[1],
                  *ubuf0 = ubuf[0], *ubuf1 = ubuf[1],
                  *vbuf0 = vbuf[0], *vbuf1 = vbuf[1];
683 684
    int  yalpha1 = 4096 - yalpha;
    int uvalpha1 = 4096 - uvalpha;
685
    int i;
686 687
    av_assert2(yalpha  <= 4096U);
    av_assert2(uvalpha <= 4096U);
688

689
    for (i = 0; i < ((dstW + 1) >> 1); i++) {
690 691 692 693 694
        int Y1 = (buf0[i * 2]     * yalpha1  + buf1[i * 2]     * yalpha)  >> 19;
        int Y2 = (buf0[i * 2 + 1] * yalpha1  + buf1[i * 2 + 1] * yalpha)  >> 19;
        int U  = (ubuf0[i]        * uvalpha1 + ubuf1[i]        * uvalpha) >> 19;
        int V  = (vbuf0[i]        * uvalpha1 + vbuf1[i]        * uvalpha) >> 19;

695 696 697 698 699 700 701
        if ((Y1 | Y2 | U | V) & 0x100) {
            Y1 = av_clip_uint8(Y1);
            Y2 = av_clip_uint8(Y2);
            U  = av_clip_uint8(U);
            V  = av_clip_uint8(V);
        }

702 703 704 705 706 707 708 709
        output_pixels(i * 4, Y1, U, Y2, V);
    }
}

static av_always_inline void
yuv2422_1_c_template(SwsContext *c, const int16_t *buf0,
                     const int16_t *ubuf[2], const int16_t *vbuf[2],
                     const int16_t *abuf0, uint8_t *dest, int dstW,
710
                     int uvalpha, int y, enum AVPixelFormat target)
711
{
712
    const int16_t *ubuf0 = ubuf[0], *vbuf0 = vbuf[0];
713 714 715
    int i;

    if (uvalpha < 2048) {
716
        for (i = 0; i < ((dstW + 1) >> 1); i++) {
717 718 719 720
            int Y1 = (buf0[i * 2    ]+64) >> 7;
            int Y2 = (buf0[i * 2 + 1]+64) >> 7;
            int U  = (ubuf0[i]       +64) >> 7;
            int V  = (vbuf0[i]       +64) >> 7;
721

722 723 724 725 726 727 728
            if ((Y1 | Y2 | U | V) & 0x100) {
                Y1 = av_clip_uint8(Y1);
                Y2 = av_clip_uint8(Y2);
                U  = av_clip_uint8(U);
                V  = av_clip_uint8(V);
            }

729 730 731 732 733
            Y1 = av_clip_uint8(Y1);
            Y2 = av_clip_uint8(Y2);
            U  = av_clip_uint8(U);
            V  = av_clip_uint8(V);

734 735 736
            output_pixels(i * 4, Y1, U, Y2, V);
        }
    } else {
737
        const int16_t *ubuf1 = ubuf[1], *vbuf1 = vbuf[1];
738
        for (i = 0; i < ((dstW + 1) >> 1); i++) {
739 740 741 742
            int Y1 = (buf0[i * 2    ]    + 64) >> 7;
            int Y2 = (buf0[i * 2 + 1]    + 64) >> 7;
            int U  = (ubuf0[i] + ubuf1[i]+128) >> 8;
            int V  = (vbuf0[i] + vbuf1[i]+128) >> 8;
743

744 745 746 747 748 749 750
            if ((Y1 | Y2 | U | V) & 0x100) {
                Y1 = av_clip_uint8(Y1);
                Y2 = av_clip_uint8(Y2);
                U  = av_clip_uint8(U);
                V  = av_clip_uint8(V);
            }

751 752 753 754 755
            Y1 = av_clip_uint8(Y1);
            Y2 = av_clip_uint8(Y2);
            U  = av_clip_uint8(U);
            V  = av_clip_uint8(V);

756 757 758 759 760 761 762
            output_pixels(i * 4, Y1, U, Y2, V);
        }
    }
}

#undef output_pixels

763
YUV2PACKEDWRAPPER(yuv2, 422, yuyv422, AV_PIX_FMT_YUYV422)
764
YUV2PACKEDWRAPPER(yuv2, 422, yvyu422, AV_PIX_FMT_YVYU422)
765
YUV2PACKEDWRAPPER(yuv2, 422, uyvy422, AV_PIX_FMT_UYVY422)
766

767 768
#define R_B ((target == AV_PIX_FMT_RGB48LE || target == AV_PIX_FMT_RGB48BE || target == AV_PIX_FMT_RGBA64LE || target == AV_PIX_FMT_RGBA64BE) ? R : B)
#define B_R ((target == AV_PIX_FMT_RGB48LE || target == AV_PIX_FMT_RGB48BE || target == AV_PIX_FMT_RGBA64LE || target == AV_PIX_FMT_RGBA64BE) ? B : R)
769 770 771 772 773 774 775
#define output_pixel(pos, val) \
    if (isBE(target)) { \
        AV_WB16(pos, val); \
    } else { \
        AV_WL16(pos, val); \
    }

Paul B Mahol's avatar
Paul B Mahol committed
776 777 778 779 780 781
static av_always_inline void
yuv2rgba64_X_c_template(SwsContext *c, const int16_t *lumFilter,
                       const int32_t **lumSrc, int lumFilterSize,
                       const int16_t *chrFilter, const int32_t **chrUSrc,
                       const int32_t **chrVSrc, int chrFilterSize,
                       const int32_t **alpSrc, uint16_t *dest, int dstW,
782
                       int y, enum AVPixelFormat target, int hasAlpha, int eightbytes)
Paul B Mahol's avatar
Paul B Mahol committed
783 784
{
    int i;
785
    int A1 = 0xffff<<14, A2 = 0xffff<<14;
Paul B Mahol's avatar
Paul B Mahol committed
786 787

    for (i = 0; i < ((dstW + 1) >> 1); i++) {
788
        int j;
Paul B Mahol's avatar
Paul B Mahol committed
789 790
        int Y1 = -0x40000000;
        int Y2 = -0x40000000;
791 792
        int U  = -(128 << 23); // 19
        int V  = -(128 << 23);
Paul B Mahol's avatar
Paul B Mahol committed
793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810
        int R, G, B;

        for (j = 0; j < lumFilterSize; j++) {
            Y1 += lumSrc[j][i * 2]     * (unsigned)lumFilter[j];
            Y2 += lumSrc[j][i * 2 + 1] * (unsigned)lumFilter[j];
        }
        for (j = 0; j < chrFilterSize; j++) {;
            U += chrUSrc[j][i] * (unsigned)chrFilter[j];
            V += chrVSrc[j][i] * (unsigned)chrFilter[j];
        }

        if (hasAlpha) {
            A1 = -0x40000000;
            A2 = -0x40000000;
            for (j = 0; j < lumFilterSize; j++) {
                A1 += alpSrc[j][i * 2]     * (unsigned)lumFilter[j];
                A2 += alpSrc[j][i * 2 + 1] * (unsigned)lumFilter[j];
            }
811 812 813 814
            A1 >>= 1;
            A1 += 0x20002000;
            A2 >>= 1;
            A2 += 0x20002000;
Paul B Mahol's avatar
Paul B Mahol committed
815 816
        }

817
        // 8 bits: 12+15=27; 16 bits: 12+19=31
Paul B Mahol's avatar
Paul B Mahol committed
818 819 820 821 822 823 824
        Y1 >>= 14; // 10
        Y1 += 0x10000;
        Y2 >>= 14;
        Y2 += 0x10000;
        U  >>= 14;
        V  >>= 14;

825
        // 8 bits: 27 -> 17 bits, 16 bits: 31 - 14 = 17 bits
Paul B Mahol's avatar
Paul B Mahol committed
826 827 828 829 830 831
        Y1 -= c->yuv2rgb_y_offset;
        Y2 -= c->yuv2rgb_y_offset;
        Y1 *= c->yuv2rgb_y_coeff;
        Y2 *= c->yuv2rgb_y_coeff;
        Y1 += 1 << 13; // 21
        Y2 += 1 << 13;
832
        // 8 bits: 17 + 13 bits = 30 bits, 16 bits: 17 + 13 bits = 30 bits
Paul B Mahol's avatar
Paul B Mahol committed
833 834 835 836 837

        R = V * c->yuv2rgb_v2r_coeff;
        G = V * c->yuv2rgb_v2g_coeff + U * c->yuv2rgb_u2g_coeff;
        B =                            U * c->yuv2rgb_u2b_coeff;

838
        // 8 bits: 30 - 22 = 8 bits, 16 bits: 30 bits - 14 = 16 bits
839
        output_pixel(&dest[0], av_clip_uintp2(R_B + Y1, 30) >> 14);
Paul B Mahol's avatar
Paul B Mahol committed
840
        output_pixel(&dest[1], av_clip_uintp2(  G + Y1, 30) >> 14);
841
        output_pixel(&dest[2], av_clip_uintp2(B_R + Y1, 30) >> 14);
842 843 844 845 846 847 848 849 850 851 852 853 854
        if (eightbytes) {
            output_pixel(&dest[3], av_clip_uintp2(A1      , 30) >> 14);
            output_pixel(&dest[4], av_clip_uintp2(R_B + Y2, 30) >> 14);
            output_pixel(&dest[5], av_clip_uintp2(  G + Y2, 30) >> 14);
            output_pixel(&dest[6], av_clip_uintp2(B_R + Y2, 30) >> 14);
            output_pixel(&dest[7], av_clip_uintp2(A2      , 30) >> 14);
            dest += 8;
        } else {
            output_pixel(&dest[3], av_clip_uintp2(R_B + Y2, 30) >> 14);
            output_pixel(&dest[4], av_clip_uintp2(  G + Y2, 30) >> 14);
            output_pixel(&dest[5], av_clip_uintp2(B_R + Y2, 30) >> 14);
            dest += 6;
        }
Paul B Mahol's avatar
Paul B Mahol committed
855 856 857 858 859 860 861 862
    }
}

static av_always_inline void
yuv2rgba64_2_c_template(SwsContext *c, const int32_t *buf[2],
                       const int32_t *ubuf[2], const int32_t *vbuf[2],
                       const int32_t *abuf[2], uint16_t *dest, int dstW,
                       int yalpha, int uvalpha, int y,
863
                       enum AVPixelFormat target, int hasAlpha, int eightbytes)
Paul B Mahol's avatar
Paul B Mahol committed
864 865 866 867 868 869 870 871 872
{
    const int32_t *buf0  = buf[0],  *buf1  = buf[1],
                  *ubuf0 = ubuf[0], *ubuf1 = ubuf[1],
                  *vbuf0 = vbuf[0], *vbuf1 = vbuf[1],
                  *abuf0 = hasAlpha ? abuf[0] : NULL,
                  *abuf1 = hasAlpha ? abuf[1] : NULL;
    int  yalpha1 = 4096 - yalpha;
    int uvalpha1 = 4096 - uvalpha;
    int i;
873
    int A1 = 0xffff<<14, A2 = 0xffff<<14;
Paul B Mahol's avatar
Paul B Mahol committed
874

875 876 877
    av_assert2(yalpha  <= 4096U);
    av_assert2(uvalpha <= 4096U);

Paul B Mahol's avatar
Paul B Mahol committed
878 879 880
    for (i = 0; i < ((dstW + 1) >> 1); i++) {
        int Y1 = (buf0[i * 2]     * yalpha1  + buf1[i * 2]     * yalpha) >> 14;
        int Y2 = (buf0[i * 2 + 1] * yalpha1  + buf1[i * 2 + 1] * yalpha) >> 14;
881 882
        int U  = (ubuf0[i]        * uvalpha1 + ubuf1[i]        * uvalpha - (128 << 23)) >> 14;
        int V  = (vbuf0[i]        * uvalpha1 + vbuf1[i]        * uvalpha - (128 << 23)) >> 14;
Paul B Mahol's avatar
Paul B Mahol committed
883 884 885 886 887 888 889 890 891 892 893 894 895 896
        int R, G, B;

        Y1 -= c->yuv2rgb_y_offset;
        Y2 -= c->yuv2rgb_y_offset;
        Y1 *= c->yuv2rgb_y_coeff;
        Y2 *= c->yuv2rgb_y_coeff;
        Y1 += 1 << 13;
        Y2 += 1 << 13;

        R = V * c->yuv2rgb_v2r_coeff;
        G = V * c->yuv2rgb_v2g_coeff + U * c->yuv2rgb_u2g_coeff;
        B =                            U * c->yuv2rgb_u2b_coeff;

        if (hasAlpha) {
897 898
            A1 = (abuf0[i * 2    ] * yalpha1 + abuf1[i * 2    ] * yalpha) >> 1;
            A2 = (abuf0[i * 2 + 1] * yalpha1 + abuf1[i * 2 + 1] * yalpha) >> 1;
Paul B Mahol's avatar
Paul B Mahol committed
899 900 901 902 903

            A1 += 1 << 13;
            A2 += 1 << 13;
        }

904
        output_pixel(&dest[0], av_clip_uintp2(R_B + Y1, 30) >> 14);
Paul B Mahol's avatar
Paul B Mahol committed
905
        output_pixel(&dest[1], av_clip_uintp2(  G + Y1, 30) >> 14);
906
        output_pixel(&dest[2], av_clip_uintp2(B_R + Y1, 30) >> 14);
907 908 909 910 911 912 913 914 915 916 917 918 919
        if (eightbytes) {
            output_pixel(&dest[3], av_clip_uintp2(A1      , 30) >> 14);
            output_pixel(&dest[4], av_clip_uintp2(R_B + Y2, 30) >> 14);
            output_pixel(&dest[5], av_clip_uintp2(  G + Y2, 30) >> 14);
            output_pixel(&dest[6], av_clip_uintp2(B_R + Y2, 30) >> 14);
            output_pixel(&dest[7], av_clip_uintp2(A2      , 30) >> 14);
            dest += 8;
        } else {
            output_pixel(&dest[3], av_clip_uintp2(R_B + Y2, 30) >> 14);
            output_pixel(&dest[4], av_clip_uintp2(  G + Y2, 30) >> 14);
            output_pixel(&dest[5], av_clip_uintp2(B_R + Y2, 30) >> 14);
            dest += 6;
        }
Paul B Mahol's avatar
Paul B Mahol committed
920 921 922 923 924 925 926
    }
}

static av_always_inline void
yuv2rgba64_1_c_template(SwsContext *c, const int32_t *buf0,
                       const int32_t *ubuf[2], const int32_t *vbuf[2],
                       const int32_t *abuf0, uint16_t *dest, int dstW,
927
                       int uvalpha, int y, enum AVPixelFormat target, int hasAlpha, int eightbytes)
Paul B Mahol's avatar
Paul B Mahol committed
928 929 930
{
    const int32_t *ubuf0 = ubuf[0], *vbuf0 = vbuf[0];
    int i;
931
    int A1 = 0xffff<<14, A2= 0xffff<<14;
Paul B Mahol's avatar
Paul B Mahol committed
932 933 934 935 936

    if (uvalpha < 2048) {
        for (i = 0; i < ((dstW + 1) >> 1); i++) {
            int Y1 = (buf0[i * 2]    ) >> 2;
            int Y2 = (buf0[i * 2 + 1]) >> 2;
937 938
            int U  = (ubuf0[i] - (128 << 11)) >> 2;
            int V  = (vbuf0[i] - (128 << 11)) >> 2;
Paul B Mahol's avatar
Paul B Mahol committed
939 940 941 942 943 944 945 946 947 948
            int R, G, B;

            Y1 -= c->yuv2rgb_y_offset;
            Y2 -= c->yuv2rgb_y_offset;
            Y1 *= c->yuv2rgb_y_coeff;
            Y2 *= c->yuv2rgb_y_coeff;
            Y1 += 1 << 13;
            Y2 += 1 << 13;

            if (hasAlpha) {
949 950
                A1 = abuf0[i * 2    ] << 11;
                A2 = abuf0[i * 2 + 1] << 11;
Paul B Mahol's avatar
Paul B Mahol committed
951 952 953 954 955 956 957 958 959

                A1 += 1 << 13;
                A2 += 1 << 13;
            }

            R = V * c->yuv2rgb_v2r_coeff;
            G = V * c->yuv2rgb_v2g_coeff + U * c->yuv2rgb_u2g_coeff;
            B =                            U * c->yuv2rgb_u2b_coeff;

960
            output_pixel(&dest[0], av_clip_uintp2(R_B + Y1, 30) >> 14);
Paul B Mahol's avatar
Paul B Mahol committed
961
            output_pixel(&dest[1], av_clip_uintp2(  G + Y1, 30) >> 14);
962
            output_pixel(&dest[2], av_clip_uintp2(B_R + Y1, 30) >> 14);
963 964 965 966 967 968 969 970 971 972 973 974 975
            if (eightbytes) {
                output_pixel(&dest[3], av_clip_uintp2(A1      , 30) >> 14);
                output_pixel(&dest[4], av_clip_uintp2(R_B + Y2, 30) >> 14);
                output_pixel(&dest[5], av_clip_uintp2(  G + Y2, 30) >> 14);
                output_pixel(&dest[6], av_clip_uintp2(B_R + Y2, 30) >> 14);
                output_pixel(&dest[7], av_clip_uintp2(A2      , 30) >> 14);
                dest += 8;
            } else {
                output_pixel(&dest[3], av_clip_uintp2(R_B + Y2, 30) >> 14);
                output_pixel(&dest[4], av_clip_uintp2(  G + Y2, 30) >> 14);
                output_pixel(&dest[5], av_clip_uintp2(B_R + Y2, 30) >> 14);
                dest += 6;
            }
Paul B Mahol's avatar
Paul B Mahol committed
976 977 978
        }
    } else {
        const int32_t *ubuf1 = ubuf[1], *vbuf1 = vbuf[1];
979
        int A1 = 0xffff<<14, A2 = 0xffff<<14;
Paul B Mahol's avatar
Paul B Mahol committed
980 981 982
        for (i = 0; i < ((dstW + 1) >> 1); i++) {
            int Y1 = (buf0[i * 2]    ) >> 2;
            int Y2 = (buf0[i * 2 + 1]) >> 2;
983 984
            int U  = (ubuf0[i] + ubuf1[i] - (128 << 12)) >> 3;
            int V  = (vbuf0[i] + vbuf1[i] - (128 << 12)) >> 3;
Paul B Mahol's avatar
Paul B Mahol committed
985 986 987 988 989 990 991 992 993 994
            int R, G, B;

            Y1 -= c->yuv2rgb_y_offset;
            Y2 -= c->yuv2rgb_y_offset;
            Y1 *= c->yuv2rgb_y_coeff;
            Y2 *= c->yuv2rgb_y_coeff;
            Y1 += 1 << 13;
            Y2 += 1 << 13;

            if (hasAlpha) {
995 996
                A1 = abuf0[i * 2    ] << 11;
                A2 = abuf0[i * 2 + 1] << 11;
Paul B Mahol's avatar
Paul B Mahol committed
997 998 999 1000 1001 1002 1003 1004 1005

                A1 += 1 << 13;
                A2 += 1 << 13;
            }

            R = V * c->yuv2rgb_v2r_coeff;
            G = V * c->yuv2rgb_v2g_coeff + U * c->yuv2rgb_u2g_coeff;
            B =                            U * c->yuv2rgb_u2b_coeff;

1006
            output_pixel(&dest[0], av_clip_uintp2(R_B + Y1, 30) >> 14);
Paul B Mahol's avatar
Paul B Mahol committed
1007
            output_pixel(&dest[1], av_clip_uintp2(  G + Y1, 30) >> 14);
1008
            output_pixel(&dest[2], av_clip_uintp2(B_R + Y1, 30) >> 14);
1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021
            if (eightbytes) {
                output_pixel(&dest[3], av_clip_uintp2(A1      , 30) >> 14);
                output_pixel(&dest[4], av_clip_uintp2(R_B + Y2, 30) >> 14);
                output_pixel(&dest[5], av_clip_uintp2(  G + Y2, 30) >> 14);
                output_pixel(&dest[6], av_clip_uintp2(B_R + Y2, 30) >> 14);
                output_pixel(&dest[7], av_clip_uintp2(A2      , 30) >> 14);
                dest += 8;
            } else {
                output_pixel(&dest[3], av_clip_uintp2(R_B + Y2, 30) >> 14);
                output_pixel(&dest[4], av_clip_uintp2(  G + Y2, 30) >> 14);
                output_pixel(&dest[5], av_clip_uintp2(B_R + Y2, 30) >> 14);
                dest += 6;
            }
1022 1023 1024 1025
        }
    }
}

1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039
static av_always_inline void
yuv2rgba64_full_X_c_template(SwsContext *c, const int16_t *lumFilter,
                       const int32_t **lumSrc, int lumFilterSize,
                       const int16_t *chrFilter, const int32_t **chrUSrc,
                       const int32_t **chrVSrc, int chrFilterSize,
                       const int32_t **alpSrc, uint16_t *dest, int dstW,
                       int y, enum AVPixelFormat target, int hasAlpha, int eightbytes)
{
    int i;
    int A = 0xffff<<14;

    for (i = 0; i < dstW; i++) {
        int j;
        int Y  = -0x40000000;
1040 1041
        int U  = -(128 << 23); // 19
        int V  = -(128 << 23);
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 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106
        int R, G, B;

        for (j = 0; j < lumFilterSize; j++) {
            Y += lumSrc[j][i]  * (unsigned)lumFilter[j];
        }
        for (j = 0; j < chrFilterSize; j++) {;
            U += chrUSrc[j][i] * (unsigned)chrFilter[j];
            V += chrVSrc[j][i] * (unsigned)chrFilter[j];
        }

        if (hasAlpha) {
            A = -0x40000000;
            for (j = 0; j < lumFilterSize; j++) {
                A += alpSrc[j][i] * (unsigned)lumFilter[j];
            }
            A >>= 1;
            A += 0x20002000;
        }

        // 8bit: 12+15=27; 16-bit: 12+19=31
        Y  >>= 14; // 10
        Y += 0x10000;
        U  >>= 14;
        V  >>= 14;

        // 8bit: 27 -> 17bit, 16bit: 31 - 14 = 17bit
        Y -= c->yuv2rgb_y_offset;
        Y *= c->yuv2rgb_y_coeff;
        Y += 1 << 13; // 21
        // 8bit: 17 + 13bit = 30bit, 16bit: 17 + 13bit = 30bit

        R = V * c->yuv2rgb_v2r_coeff;
        G = V * c->yuv2rgb_v2g_coeff + U * c->yuv2rgb_u2g_coeff;
        B =                            U * c->yuv2rgb_u2b_coeff;

        // 8bit: 30 - 22 = 8bit, 16bit: 30bit - 14 = 16bit
        output_pixel(&dest[0], av_clip_uintp2(R_B + Y, 30) >> 14);
        output_pixel(&dest[1], av_clip_uintp2(  G + Y, 30) >> 14);
        output_pixel(&dest[2], av_clip_uintp2(B_R + Y, 30) >> 14);
        if (eightbytes) {
            output_pixel(&dest[3], av_clip_uintp2(A, 30) >> 14);
            dest += 4;
        } else {
            dest += 3;
        }
    }
}

static av_always_inline void
yuv2rgba64_full_2_c_template(SwsContext *c, const int32_t *buf[2],
                       const int32_t *ubuf[2], const int32_t *vbuf[2],
                       const int32_t *abuf[2], uint16_t *dest, int dstW,
                       int yalpha, int uvalpha, int y,
                       enum AVPixelFormat target, int hasAlpha, int eightbytes)
{
    const int32_t *buf0  = buf[0],  *buf1  = buf[1],
                  *ubuf0 = ubuf[0], *ubuf1 = ubuf[1],
                  *vbuf0 = vbuf[0], *vbuf1 = vbuf[1],
                  *abuf0 = hasAlpha ? abuf[0] : NULL,
                  *abuf1 = hasAlpha ? abuf[1] : NULL;
    int  yalpha1 = 4096 - yalpha;
    int uvalpha1 = 4096 - uvalpha;
    int i;
    int A = 0xffff<<14;

1107 1108 1109
    av_assert2(yalpha  <= 4096U);
    av_assert2(uvalpha <= 4096U);

1110 1111
    for (i = 0; i < dstW; i++) {
        int Y  = (buf0[i]     * yalpha1  + buf1[i]     * yalpha) >> 14;
1112 1113
        int U  = (ubuf0[i]   * uvalpha1 + ubuf1[i]     * uvalpha - (128 << 23)) >> 14;
        int V  = (vbuf0[i]   * uvalpha1 + vbuf1[i]     * uvalpha - (128 << 23)) >> 14;
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 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154
        int R, G, B;

        Y -= c->yuv2rgb_y_offset;
        Y *= c->yuv2rgb_y_coeff;
        Y += 1 << 13;

        R = V * c->yuv2rgb_v2r_coeff;
        G = V * c->yuv2rgb_v2g_coeff + U * c->yuv2rgb_u2g_coeff;
        B =                            U * c->yuv2rgb_u2b_coeff;

        if (hasAlpha) {
            A = (abuf0[i] * yalpha1 + abuf1[i] * yalpha) >> 1;

            A += 1 << 13;
        }

        output_pixel(&dest[0], av_clip_uintp2(R_B + Y, 30) >> 14);
        output_pixel(&dest[1], av_clip_uintp2(  G + Y, 30) >> 14);
        output_pixel(&dest[2], av_clip_uintp2(B_R + Y, 30) >> 14);
        if (eightbytes) {
            output_pixel(&dest[3], av_clip_uintp2(A, 30) >> 14);
            dest += 4;
        } else {
            dest += 3;
        }
    }
}

static av_always_inline void
yuv2rgba64_full_1_c_template(SwsContext *c, const int32_t *buf0,
                       const int32_t *ubuf[2], const int32_t *vbuf[2],
                       const int32_t *abuf0, uint16_t *dest, int dstW,
                       int uvalpha, int y, enum AVPixelFormat target, int hasAlpha, int eightbytes)
{
    const int32_t *ubuf0 = ubuf[0], *vbuf0 = vbuf[0];
    int i;
    int A = 0xffff<<14;

    if (uvalpha < 2048) {
        for (i = 0; i < dstW; i++) {
            int Y  = (buf0[i]) >> 2;
1155 1156
            int U  = (ubuf0[i] - (128 << 11)) >> 2;
            int V  = (vbuf0[i] - (128 << 11)) >> 2;
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
            int R, G, B;

            Y -= c->yuv2rgb_y_offset;
            Y *= c->yuv2rgb_y_coeff;
            Y += 1 << 13;

            if (hasAlpha) {
                A = abuf0[i] << 11;

                A += 1 << 13;
            }

            R = V * c->yuv2rgb_v2r_coeff;
            G = V * c->yuv2rgb_v2g_coeff + U * c->yuv2rgb_u2g_coeff;
            B =                            U * c->yuv2rgb_u2b_coeff;

            output_pixel(&dest[0], av_clip_uintp2(R_B + Y, 30) >> 14);
            output_pixel(&dest[1], av_clip_uintp2(  G + Y, 30) >> 14);
            output_pixel(&dest[2], av_clip_uintp2(B_R + Y, 30) >> 14);
            if (eightbytes) {
                output_pixel(&dest[3], av_clip_uintp2(A, 30) >> 14);
                dest += 4;
            } else {
                dest += 3;
            }
        }
    } else {
        const int32_t *ubuf1 = ubuf[1], *vbuf1 = vbuf[1];
        int A = 0xffff<<14;
        for (i = 0; i < dstW; i++) {
            int Y  = (buf0[i]    ) >> 2;
1188 1189
            int U  = (ubuf0[i] + ubuf1[i] - (128 << 12)) >> 3;
            int V  = (vbuf0[i] + vbuf1[i] - (128 << 12)) >> 3;
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
            int R, G, B;

            Y -= c->yuv2rgb_y_offset;
            Y *= c->yuv2rgb_y_coeff;
            Y += 1 << 13;

            if (hasAlpha) {
                A = abuf0[i] << 11;

                A += 1 << 13;
            }

            R = V * c->yuv2rgb_v2r_coeff;
            G = V * c->yuv2rgb_v2g_coeff + U * c->yuv2rgb_u2g_coeff;
            B =                            U * c->yuv2rgb_u2b_coeff;

            output_pixel(&dest[0], av_clip_uintp2(R_B + Y, 30) >> 14);
            output_pixel(&dest[1], av_clip_uintp2(  G + Y, 30) >> 14);
            output_pixel(&dest[2], av_clip_uintp2(B_R + Y, 30) >> 14);
            if (eightbytes) {
                output_pixel(&dest[3], av_clip_uintp2(A, 30) >> 14);
                dest += 4;
            } else {
                dest += 3;
            }
        }
    }
}

1219 1220 1221 1222
#undef output_pixel
#undef r_b
#undef b_r

1223
#define YUV2PACKED16WRAPPER(name, base, ext, fmt, hasAlpha, eightbytes) \
1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237
static void name ## ext ## _X_c(SwsContext *c, const int16_t *lumFilter, \
                        const int16_t **_lumSrc, int lumFilterSize, \
                        const int16_t *chrFilter, const int16_t **_chrUSrc, \
                        const int16_t **_chrVSrc, int chrFilterSize, \
                        const int16_t **_alpSrc, uint8_t *_dest, int dstW, \
                        int y) \
{ \
    const int32_t **lumSrc  = (const int32_t **) _lumSrc, \
                  **chrUSrc = (const int32_t **) _chrUSrc, \
                  **chrVSrc = (const int32_t **) _chrVSrc, \
                  **alpSrc  = (const int32_t **) _alpSrc; \
    uint16_t *dest = (uint16_t *) _dest; \
    name ## base ## _X_c_template(c, lumFilter, lumSrc, lumFilterSize, \
                          chrFilter, chrUSrc, chrVSrc, chrFilterSize, \
1238
                          alpSrc, dest, dstW, y, fmt, hasAlpha, eightbytes); \
1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251
} \
 \
static void name ## ext ## _2_c(SwsContext *c, const int16_t *_buf[2], \
                        const int16_t *_ubuf[2], const int16_t *_vbuf[2], \
                        const int16_t *_abuf[2], uint8_t *_dest, int dstW, \
                        int yalpha, int uvalpha, int y) \
{ \
    const int32_t **buf  = (const int32_t **) _buf, \
                  **ubuf = (const int32_t **) _ubuf, \
                  **vbuf = (const int32_t **) _vbuf, \
                  **abuf = (const int32_t **) _abuf; \
    uint16_t *dest = (uint16_t *) _dest; \
    name ## base ## _2_c_template(c, buf, ubuf, vbuf, abuf, \
1252
                          dest, dstW, yalpha, uvalpha, y, fmt, hasAlpha, eightbytes); \
1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265
} \
 \
static void name ## ext ## _1_c(SwsContext *c, const int16_t *_buf0, \
                        const int16_t *_ubuf[2], const int16_t *_vbuf[2], \
                        const int16_t *_abuf0, uint8_t *_dest, int dstW, \
                        int uvalpha, int y) \
{ \
    const int32_t *buf0  = (const int32_t *)  _buf0, \
                 **ubuf  = (const int32_t **) _ubuf, \
                 **vbuf  = (const int32_t **) _vbuf, \
                  *abuf0 = (const int32_t *)  _abuf0; \
    uint16_t *dest = (uint16_t *) _dest; \
    name ## base ## _1_c_template(c, buf0, ubuf, vbuf, abuf0, dest, \
1266
                                  dstW, uvalpha, y, fmt, hasAlpha, eightbytes); \
1267 1268
}

1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280
YUV2PACKED16WRAPPER(yuv2, rgba64, rgb48be, AV_PIX_FMT_RGB48BE, 0, 0)
YUV2PACKED16WRAPPER(yuv2, rgba64, rgb48le, AV_PIX_FMT_RGB48LE, 0, 0)
YUV2PACKED16WRAPPER(yuv2, rgba64, bgr48be, AV_PIX_FMT_BGR48BE, 0, 0)
YUV2PACKED16WRAPPER(yuv2, rgba64, bgr48le, AV_PIX_FMT_BGR48LE, 0, 0)
YUV2PACKED16WRAPPER(yuv2, rgba64, rgba64be, AV_PIX_FMT_RGBA64BE, 1, 1)
YUV2PACKED16WRAPPER(yuv2, rgba64, rgba64le, AV_PIX_FMT_RGBA64LE, 1, 1)
YUV2PACKED16WRAPPER(yuv2, rgba64, rgbx64be, AV_PIX_FMT_RGBA64BE, 0, 1)
YUV2PACKED16WRAPPER(yuv2, rgba64, rgbx64le, AV_PIX_FMT_RGBA64LE, 0, 1)
YUV2PACKED16WRAPPER(yuv2, rgba64, bgra64be, AV_PIX_FMT_BGRA64BE, 1, 1)
YUV2PACKED16WRAPPER(yuv2, rgba64, bgra64le, AV_PIX_FMT_BGRA64LE, 1, 1)
YUV2PACKED16WRAPPER(yuv2, rgba64, bgrx64be, AV_PIX_FMT_BGRA64BE, 0, 1)
YUV2PACKED16WRAPPER(yuv2, rgba64, bgrx64le, AV_PIX_FMT_BGRA64LE, 0, 1)
1281

1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294
YUV2PACKED16WRAPPER(yuv2, rgba64_full, rgb48be_full, AV_PIX_FMT_RGB48BE, 0, 0)
YUV2PACKED16WRAPPER(yuv2, rgba64_full, rgb48le_full, AV_PIX_FMT_RGB48LE, 0, 0)
YUV2PACKED16WRAPPER(yuv2, rgba64_full, bgr48be_full, AV_PIX_FMT_BGR48BE, 0, 0)
YUV2PACKED16WRAPPER(yuv2, rgba64_full, bgr48le_full, AV_PIX_FMT_BGR48LE, 0, 0)
YUV2PACKED16WRAPPER(yuv2, rgba64_full, rgba64be_full, AV_PIX_FMT_RGBA64BE, 1, 1)
YUV2PACKED16WRAPPER(yuv2, rgba64_full, rgba64le_full, AV_PIX_FMT_RGBA64LE, 1, 1)
YUV2PACKED16WRAPPER(yuv2, rgba64_full, rgbx64be_full, AV_PIX_FMT_RGBA64BE, 0, 1)
YUV2PACKED16WRAPPER(yuv2, rgba64_full, rgbx64le_full, AV_PIX_FMT_RGBA64LE, 0, 1)
YUV2PACKED16WRAPPER(yuv2, rgba64_full, bgra64be_full, AV_PIX_FMT_BGRA64BE, 1, 1)
YUV2PACKED16WRAPPER(yuv2, rgba64_full, bgra64le_full, AV_PIX_FMT_BGRA64LE, 1, 1)
YUV2PACKED16WRAPPER(yuv2, rgba64_full, bgrx64be_full, AV_PIX_FMT_BGRA64BE, 0, 1)
YUV2PACKED16WRAPPER(yuv2, rgba64_full, bgrx64le_full, AV_PIX_FMT_BGRA64LE, 0, 1)

1295 1296 1297 1298 1299 1300 1301 1302 1303
/*
 * Write out 2 RGB pixels in the target pixel format. This function takes a
 * R/G/B LUT as generated by ff_yuv2rgb_c_init_tables(), which takes care of
 * things like endianness conversion and shifting. The caller takes care of
 * setting the correct offset in these tables from the chroma (U/V) values.
 * This function then uses the luminance (Y1/Y2) values to write out the
 * correct RGB values into the destination buffer.
 */
static av_always_inline void
1304
yuv2rgb_write(uint8_t *_dest, int i, int Y1, int Y2,
1305 1306
              unsigned A1, unsigned A2,
              const void *_r, const void *_g, const void *_b, int y,
1307
              enum AVPixelFormat target, int hasAlpha)
1308
{
1309 1310
    if (target == AV_PIX_FMT_ARGB || target == AV_PIX_FMT_RGBA ||
        target == AV_PIX_FMT_ABGR || target == AV_PIX_FMT_BGRA) {
1311 1312 1313 1314 1315 1316
        uint32_t *dest = (uint32_t *) _dest;
        const uint32_t *r = (const uint32_t *) _r;
        const uint32_t *g = (const uint32_t *) _g;
        const uint32_t *b = (const uint32_t *) _b;

#if CONFIG_SMALL
1317
        int sh = hasAlpha ? ((target == AV_PIX_FMT_RGB32_1 || target == AV_PIX_FMT_BGR32_1) ? 0 : 24) : 0;
1318 1319 1320 1321 1322

        dest[i * 2 + 0] = r[Y1] + g[Y1] + b[Y1] + (hasAlpha ? A1 << sh : 0);
        dest[i * 2 + 1] = r[Y2] + g[Y2] + b[Y2] + (hasAlpha ? A2 << sh : 0);
#else
        if (hasAlpha) {
1323
            int sh = (target == AV_PIX_FMT_RGB32_1 || target == AV_PIX_FMT_BGR32_1) ? 0 : 24;
1324

1325
            av_assert2((((r[Y1] + g[Y1] + b[Y1]) >> sh) & 0xFF) == 0);
1326 1327 1328
            dest[i * 2 + 0] = r[Y1] + g[Y1] + b[Y1] + (A1 << sh);
            dest[i * 2 + 1] = r[Y2] + g[Y2] + b[Y2] + (A2 << sh);
        } else {
1329
#if defined(ASSERT_LEVEL) && ASSERT_LEVEL > 1
1330 1331 1332 1333
            int sh = (target == AV_PIX_FMT_RGB32_1 || target == AV_PIX_FMT_BGR32_1) ? 0 : 24;

            av_assert2((((r[Y1] + g[Y1] + b[Y1]) >> sh) & 0xFF) == 0xFF);
#endif
1334 1335 1336 1337
            dest[i * 2 + 0] = r[Y1] + g[Y1] + b[Y1];
            dest[i * 2 + 1] = r[Y2] + g[Y2] + b[Y2];
        }
#endif
1338
    } else if (target == AV_PIX_FMT_RGB24 || target == AV_PIX_FMT_BGR24) {
1339 1340 1341 1342 1343
        uint8_t *dest = (uint8_t *) _dest;
        const uint8_t *r = (const uint8_t *) _r;
        const uint8_t *g = (const uint8_t *) _g;
        const uint8_t *b = (const uint8_t *) _b;

1344 1345
#define r_b ((target == AV_PIX_FMT_RGB24) ? r : b)
#define b_r ((target == AV_PIX_FMT_RGB24) ? b : r)
1346

1347 1348 1349 1350 1351 1352 1353 1354
        dest[i * 6 + 0] = r_b[Y1];
        dest[i * 6 + 1] =   g[Y1];
        dest[i * 6 + 2] = b_r[Y1];
        dest[i * 6 + 3] = r_b[Y2];
        dest[i * 6 + 4] =   g[Y2];
        dest[i * 6 + 5] = b_r[Y2];
#undef r_b
#undef b_r
1355 1356 1357
    } else if (target == AV_PIX_FMT_RGB565 || target == AV_PIX_FMT_BGR565 ||
               target == AV_PIX_FMT_RGB555 || target == AV_PIX_FMT_BGR555 ||
               target == AV_PIX_FMT_RGB444 || target == AV_PIX_FMT_BGR444) {
1358 1359 1360 1361 1362 1363
        uint16_t *dest = (uint16_t *) _dest;
        const uint16_t *r = (const uint16_t *) _r;
        const uint16_t *g = (const uint16_t *) _g;
        const uint16_t *b = (const uint16_t *) _b;
        int dr1, dg1, db1, dr2, dg2, db2;

1364
        if (target == AV_PIX_FMT_RGB565 || target == AV_PIX_FMT_BGR565) {
1365 1366 1367 1368 1369 1370
            dr1 = ff_dither_2x2_8[ y & 1     ][0];
            dg1 = ff_dither_2x2_4[ y & 1     ][0];
            db1 = ff_dither_2x2_8[(y & 1) ^ 1][0];
            dr2 = ff_dither_2x2_8[ y & 1     ][1];
            dg2 = ff_dither_2x2_4[ y & 1     ][1];
            db2 = ff_dither_2x2_8[(y & 1) ^ 1][1];
1371
        } else if (target == AV_PIX_FMT_RGB555 || target == AV_PIX_FMT_BGR555) {
1372 1373 1374 1375 1376 1377
            dr1 = ff_dither_2x2_8[ y & 1     ][0];
            dg1 = ff_dither_2x2_8[ y & 1     ][1];
            db1 = ff_dither_2x2_8[(y & 1) ^ 1][0];
            dr2 = ff_dither_2x2_8[ y & 1     ][1];
            dg2 = ff_dither_2x2_8[ y & 1     ][0];
            db2 = ff_dither_2x2_8[(y & 1) ^ 1][1];
1378
        } else {
1379 1380 1381 1382 1383 1384
            dr1 = ff_dither_4x4_16[ y & 3     ][0];
            dg1 = ff_dither_4x4_16[ y & 3     ][1];
            db1 = ff_dither_4x4_16[(y & 3) ^ 3][0];
            dr2 = ff_dither_4x4_16[ y & 3     ][1];
            dg2 = ff_dither_4x4_16[ y & 3     ][0];
            db2 = ff_dither_4x4_16[(y & 3) ^ 3][1];
1385 1386 1387 1388
        }

        dest[i * 2 + 0] = r[Y1 + dr1] + g[Y1 + dg1] + b[Y1 + db1];
        dest[i * 2 + 1] = r[Y2 + dr2] + g[Y2 + dg2] + b[Y2 + db2];
1389
    } else /* 8/4 bits */ {
1390 1391 1392 1393 1394 1395
        uint8_t *dest = (uint8_t *) _dest;
        const uint8_t *r = (const uint8_t *) _r;
        const uint8_t *g = (const uint8_t *) _g;
        const uint8_t *b = (const uint8_t *) _b;
        int dr1, dg1, db1, dr2, dg2, db2;

1396
        if (target == AV_PIX_FMT_RGB8 || target == AV_PIX_FMT_BGR8) {
1397 1398
            const uint8_t * const d64 = ff_dither_8x8_73[y & 7];
            const uint8_t * const d32 = ff_dither_8x8_32[y & 7];
1399 1400 1401 1402 1403
            dr1 = dg1 = d32[(i * 2 + 0) & 7];
            db1 =       d64[(i * 2 + 0) & 7];
            dr2 = dg2 = d32[(i * 2 + 1) & 7];
            db2 =       d64[(i * 2 + 1) & 7];
        } else {
1404 1405
            const uint8_t * const d64  = ff_dither_8x8_73 [y & 7];
            const uint8_t * const d128 = ff_dither_8x8_220[y & 7];
1406 1407 1408 1409 1410 1411
            dr1 = db1 = d128[(i * 2 + 0) & 7];
            dg1 =        d64[(i * 2 + 0) & 7];
            dr2 = db2 = d128[(i * 2 + 1) & 7];
            dg2 =        d64[(i * 2 + 1) & 7];
        }

1412
        if (target == AV_PIX_FMT_RGB4 || target == AV_PIX_FMT_BGR4) {
1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427
            dest[i] = r[Y1 + dr1] + g[Y1 + dg1] + b[Y1 + db1] +
                    ((r[Y2 + dr2] + g[Y2 + dg2] + b[Y2 + db2]) << 4);
        } else {
            dest[i * 2 + 0] = r[Y1 + dr1] + g[Y1 + dg1] + b[Y1 + db1];
            dest[i * 2 + 1] = r[Y2 + dr2] + g[Y2 + dg2] + b[Y2 + db2];
        }
    }
}

static av_always_inline void
yuv2rgb_X_c_template(SwsContext *c, const int16_t *lumFilter,
                     const int16_t **lumSrc, int lumFilterSize,
                     const int16_t *chrFilter, const int16_t **chrUSrc,
                     const int16_t **chrVSrc, int chrFilterSize,
                     const int16_t **alpSrc, uint8_t *dest, int dstW,
1428
                     int y, enum AVPixelFormat target, int hasAlpha)
1429 1430 1431
{
    int i;

1432
    for (i = 0; i < ((dstW + 1) >> 1); i++) {
1433
        int j, A1, A2;
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
        int Y1 = 1 << 18;
        int Y2 = 1 << 18;
        int U  = 1 << 18;
        int V  = 1 << 18;
        const void *r, *g, *b;

        for (j = 0; j < lumFilterSize; j++) {
            Y1 += lumSrc[j][i * 2]     * lumFilter[j];
            Y2 += lumSrc[j][i * 2 + 1] * lumFilter[j];
        }
        for (j = 0; j < chrFilterSize; j++) {
            U += chrUSrc[j][i] * chrFilter[j];
            V += chrVSrc[j][i] * chrFilter[j];
        }
        Y1 >>= 19;
        Y2 >>= 19;
        U  >>= 19;
        V  >>= 19;
        if (hasAlpha) {
            A1 = 1 << 18;
            A2 = 1 << 18;
            for (j = 0; j < lumFilterSize; j++) {
                A1 += alpSrc[j][i * 2    ] * lumFilter[j];
                A2 += alpSrc[j][i * 2 + 1] * lumFilter[j];
            }
            A1 >>= 19;
            A2 >>= 19;
            if ((A1 | A2) & 0x100) {
                A1 = av_clip_uint8(A1);
                A2 = av_clip_uint8(A2);
            }
        }

1467 1468 1469
        r =  c->table_rV[V + YUVRGB_TABLE_HEADROOM];
        g = (c->table_gU[U + YUVRGB_TABLE_HEADROOM] + c->table_gV[V + YUVRGB_TABLE_HEADROOM]);
        b =  c->table_bU[U + YUVRGB_TABLE_HEADROOM];
1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480

        yuv2rgb_write(dest, i, Y1, Y2, hasAlpha ? A1 : 0, hasAlpha ? A2 : 0,
                      r, g, b, y, target, hasAlpha);
    }
}

static av_always_inline void
yuv2rgb_2_c_template(SwsContext *c, const int16_t *buf[2],
                     const int16_t *ubuf[2], const int16_t *vbuf[2],
                     const int16_t *abuf[2], uint8_t *dest, int dstW,
                     int yalpha, int uvalpha, int y,
1481
                     enum AVPixelFormat target, int hasAlpha)
1482 1483 1484 1485 1486 1487
{
    const int16_t *buf0  = buf[0],  *buf1  = buf[1],
                  *ubuf0 = ubuf[0], *ubuf1 = ubuf[1],
                  *vbuf0 = vbuf[0], *vbuf1 = vbuf[1],
                  *abuf0 = hasAlpha ? abuf[0] : NULL,
                  *abuf1 = hasAlpha ? abuf[1] : NULL;
1488 1489
    int  yalpha1 = 4096 - yalpha;
    int uvalpha1 = 4096 - uvalpha;
1490
    int i;
1491 1492
    av_assert2(yalpha  <= 4096U);
    av_assert2(uvalpha <= 4096U);
1493

1494
    for (i = 0; i < ((dstW + 1) >> 1); i++) {
1495 1496 1497 1498 1499
        int Y1 = (buf0[i * 2]     * yalpha1  + buf1[i * 2]     * yalpha)  >> 19;
        int Y2 = (buf0[i * 2 + 1] * yalpha1  + buf1[i * 2 + 1] * yalpha)  >> 19;
        int U  = (ubuf0[i]        * uvalpha1 + ubuf1[i]        * uvalpha) >> 19;
        int V  = (vbuf0[i]        * uvalpha1 + vbuf1[i]        * uvalpha) >> 19;
        int A1, A2;
1500 1501 1502
        const void *r =  c->table_rV[V + YUVRGB_TABLE_HEADROOM],
                   *g = (c->table_gU[U + YUVRGB_TABLE_HEADROOM] + c->table_gV[V + YUVRGB_TABLE_HEADROOM]),
                   *b =  c->table_bU[U + YUVRGB_TABLE_HEADROOM];
1503 1504 1505 1506

        if (hasAlpha) {
            A1 = (abuf0[i * 2    ] * yalpha1 + abuf1[i * 2    ] * yalpha) >> 19;
            A2 = (abuf0[i * 2 + 1] * yalpha1 + abuf1[i * 2 + 1] * yalpha) >> 19;
1507 1508
            A1 = av_clip_uint8(A1);
            A2 = av_clip_uint8(A2);
1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519
        }

        yuv2rgb_write(dest, i, Y1, Y2, hasAlpha ? A1 : 0, hasAlpha ? A2 : 0,
                      r, g, b, y, target, hasAlpha);
    }
}

static av_always_inline void
yuv2rgb_1_c_template(SwsContext *c, const int16_t *buf0,
                     const int16_t *ubuf[2], const int16_t *vbuf[2],
                     const int16_t *abuf0, uint8_t *dest, int dstW,
1520
                     int uvalpha, int y, enum AVPixelFormat target,
1521 1522
                     int hasAlpha)
{
1523
    const int16_t *ubuf0 = ubuf[0], *vbuf0 = vbuf[0];
1524 1525 1526
    int i;

    if (uvalpha < 2048) {
1527
        for (i = 0; i < ((dstW + 1) >> 1); i++) {
1528 1529 1530 1531
            int Y1 = (buf0[i * 2    ] + 64) >> 7;
            int Y2 = (buf0[i * 2 + 1] + 64) >> 7;
            int U  = (ubuf0[i]        + 64) >> 7;
            int V  = (vbuf0[i]        + 64) >> 7;
1532
            int A1, A2;
1533 1534 1535
            const void *r =  c->table_rV[V + YUVRGB_TABLE_HEADROOM],
                       *g = (c->table_gU[U + YUVRGB_TABLE_HEADROOM] + c->table_gV[V + YUVRGB_TABLE_HEADROOM]),
                       *b =  c->table_bU[U + YUVRGB_TABLE_HEADROOM];
1536 1537

            if (hasAlpha) {
1538 1539
                A1 = abuf0[i * 2    ] * 255 + 16384 >> 15;
                A2 = abuf0[i * 2 + 1] * 255 + 16384 >> 15;
1540 1541
                A1 = av_clip_uint8(A1);
                A2 = av_clip_uint8(A2);
1542 1543 1544 1545 1546 1547
            }

            yuv2rgb_write(dest, i, Y1, Y2, hasAlpha ? A1 : 0, hasAlpha ? A2 : 0,
                          r, g, b, y, target, hasAlpha);
        }
    } else {
1548
        const int16_t *ubuf1 = ubuf[1], *vbuf1 = vbuf[1];
1549
        for (i = 0; i < ((dstW + 1) >> 1); i++) {
1550 1551 1552 1553
            int Y1 = (buf0[i * 2    ]     +  64) >> 7;
            int Y2 = (buf0[i * 2 + 1]     +  64) >> 7;
            int U  = (ubuf0[i] + ubuf1[i] + 128) >> 8;
            int V  = (vbuf0[i] + vbuf1[i] + 128) >> 8;
1554
            int A1, A2;
1555 1556 1557
            const void *r =  c->table_rV[V + YUVRGB_TABLE_HEADROOM],
                       *g = (c->table_gU[U + YUVRGB_TABLE_HEADROOM] + c->table_gV[V + YUVRGB_TABLE_HEADROOM]),
                       *b =  c->table_bU[U + YUVRGB_TABLE_HEADROOM];
1558 1559

            if (hasAlpha) {
1560 1561
                A1 = (abuf0[i * 2    ] + 64) >> 7;
                A2 = (abuf0[i * 2 + 1] + 64) >> 7;
1562 1563
                A1 = av_clip_uint8(A1);
                A2 = av_clip_uint8(A2);
1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583
            }

            yuv2rgb_write(dest, i, Y1, Y2, hasAlpha ? A1 : 0, hasAlpha ? A2 : 0,
                          r, g, b, y, target, hasAlpha);
        }
    }
}

#define YUV2RGBWRAPPERX(name, base, ext, fmt, hasAlpha) \
static void name ## ext ## _X_c(SwsContext *c, const int16_t *lumFilter, \
                                const int16_t **lumSrc, int lumFilterSize, \
                                const int16_t *chrFilter, const int16_t **chrUSrc, \
                                const int16_t **chrVSrc, int chrFilterSize, \
                                const int16_t **alpSrc, uint8_t *dest, int dstW, \
                                int y) \
{ \
    name ## base ## _X_c_template(c, lumFilter, lumSrc, lumFilterSize, \
                                  chrFilter, chrUSrc, chrVSrc, chrFilterSize, \
                                  alpSrc, dest, dstW, y, fmt, hasAlpha); \
}
1584 1585

#define YUV2RGBWRAPPERX2(name, base, ext, fmt, hasAlpha) \
1586 1587 1588 1589 1590 1591 1592 1593
YUV2RGBWRAPPERX(name, base, ext, fmt, hasAlpha) \
static void name ## ext ## _2_c(SwsContext *c, const int16_t *buf[2], \
                                const int16_t *ubuf[2], const int16_t *vbuf[2], \
                                const int16_t *abuf[2], uint8_t *dest, int dstW, \
                                int yalpha, int uvalpha, int y) \
{ \
    name ## base ## _2_c_template(c, buf, ubuf, vbuf, abuf, \
                                  dest, dstW, yalpha, uvalpha, y, fmt, hasAlpha); \
1594 1595 1596 1597
}

#define YUV2RGBWRAPPER(name, base, ext, fmt, hasAlpha) \
YUV2RGBWRAPPERX2(name, base, ext, fmt, hasAlpha) \
1598 1599 1600 1601 1602 1603 1604 1605 1606 1607
static void name ## ext ## _1_c(SwsContext *c, const int16_t *buf0, \
                                const int16_t *ubuf[2], const int16_t *vbuf[2], \
                                const int16_t *abuf0, uint8_t *dest, int dstW, \
                                int uvalpha, int y) \
{ \
    name ## base ## _1_c_template(c, buf0, ubuf, vbuf, abuf0, dest, \
                                  dstW, uvalpha, y, fmt, hasAlpha); \
}

#if CONFIG_SMALL
1608 1609
YUV2RGBWRAPPER(yuv2rgb,,  32_1,  AV_PIX_FMT_RGB32_1,   CONFIG_SWSCALE_ALPHA && c->needAlpha)
YUV2RGBWRAPPER(yuv2rgb,,  32,    AV_PIX_FMT_RGB32,     CONFIG_SWSCALE_ALPHA && c->needAlpha)
1610 1611
#else
#if CONFIG_SWSCALE_ALPHA
1612 1613
YUV2RGBWRAPPER(yuv2rgb,, a32_1,  AV_PIX_FMT_RGB32_1,   1)
YUV2RGBWRAPPER(yuv2rgb,, a32,    AV_PIX_FMT_RGB32,     1)
1614
#endif
1615 1616
YUV2RGBWRAPPER(yuv2rgb,, x32_1,  AV_PIX_FMT_RGB32_1,   0)
YUV2RGBWRAPPER(yuv2rgb,, x32,    AV_PIX_FMT_RGB32,     0)
1617
#endif
1618 1619 1620 1621 1622 1623 1624 1625
YUV2RGBWRAPPER(yuv2, rgb, rgb24, AV_PIX_FMT_RGB24,   0)
YUV2RGBWRAPPER(yuv2, rgb, bgr24, AV_PIX_FMT_BGR24,   0)
YUV2RGBWRAPPER(yuv2rgb,,  16,    AV_PIX_FMT_RGB565,    0)
YUV2RGBWRAPPER(yuv2rgb,,  15,    AV_PIX_FMT_RGB555,    0)
YUV2RGBWRAPPER(yuv2rgb,,  12,    AV_PIX_FMT_RGB444,    0)
YUV2RGBWRAPPER(yuv2rgb,,   8,    AV_PIX_FMT_RGB8,      0)
YUV2RGBWRAPPER(yuv2rgb,,   4,    AV_PIX_FMT_RGB4,      0)
YUV2RGBWRAPPER(yuv2rgb,,   4b,   AV_PIX_FMT_RGB4_BYTE, 0)
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
static av_always_inline void yuv2rgb_write_full(SwsContext *c,
    uint8_t *dest, int i, int Y, int A, int U, int V,
    int y, enum AVPixelFormat target, int hasAlpha, int err[4])
{
    int R, G, B;
    int isrgb8 = target == AV_PIX_FMT_BGR8 || target == AV_PIX_FMT_RGB8;

    Y -= c->yuv2rgb_y_offset;
    Y *= c->yuv2rgb_y_coeff;
    Y += 1 << 21;
    R = Y + V*c->yuv2rgb_v2r_coeff;
    G = Y + V*c->yuv2rgb_v2g_coeff + U*c->yuv2rgb_u2g_coeff;
    B = Y +                          U*c->yuv2rgb_u2b_coeff;
    if ((R | G | B) & 0xC0000000) {
        R = av_clip_uintp2(R, 30);
        G = av_clip_uintp2(G, 30);
        B = av_clip_uintp2(B, 30);
    }

    switch(target) {
    case AV_PIX_FMT_ARGB:
        dest[0] = hasAlpha ? A : 255;
        dest[1] = R >> 22;
        dest[2] = G >> 22;
        dest[3] = B >> 22;
        break;
    case AV_PIX_FMT_RGB24:
        dest[0] = R >> 22;
        dest[1] = G >> 22;
        dest[2] = B >> 22;
        break;
    case AV_PIX_FMT_RGBA:
        dest[0] = R >> 22;
        dest[1] = G >> 22;
        dest[2] = B >> 22;
        dest[3] = hasAlpha ? A : 255;
        break;
    case AV_PIX_FMT_ABGR:
        dest[0] = hasAlpha ? A : 255;
        dest[1] = B >> 22;
        dest[2] = G >> 22;
        dest[3] = R >> 22;
        break;
    case AV_PIX_FMT_BGR24:
        dest[0] = B >> 22;
        dest[1] = G >> 22;
        dest[2] = R >> 22;
        break;
    case AV_PIX_FMT_BGRA:
        dest[0] = B >> 22;
        dest[1] = G >> 22;
        dest[2] = R >> 22;
        dest[3] = hasAlpha ? A : 255;
        break;
    case AV_PIX_FMT_BGR4_BYTE:
    case AV_PIX_FMT_RGB4_BYTE:
    case AV_PIX_FMT_BGR8:
    case AV_PIX_FMT_RGB8:
    {
        int r,g,b;
1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717

        switch (c->dither) {
        default:
        case SWS_DITHER_AUTO:
        case SWS_DITHER_ED:
            R >>= 22;
            G >>= 22;
            B >>= 22;
            R += (7*err[0] + 1*c->dither_error[0][i] + 5*c->dither_error[0][i+1] + 3*c->dither_error[0][i+2])>>4;
            G += (7*err[1] + 1*c->dither_error[1][i] + 5*c->dither_error[1][i+1] + 3*c->dither_error[1][i+2])>>4;
            B += (7*err[2] + 1*c->dither_error[2][i] + 5*c->dither_error[2][i+1] + 3*c->dither_error[2][i+2])>>4;
            c->dither_error[0][i] = err[0];
            c->dither_error[1][i] = err[1];
            c->dither_error[2][i] = err[2];
            r = R >> (isrgb8 ? 5 : 7);
            g = G >> (isrgb8 ? 5 : 6);
            b = B >> (isrgb8 ? 6 : 7);
            r = av_clip(r, 0, isrgb8 ? 7 : 1);
            g = av_clip(g, 0, isrgb8 ? 7 : 3);
            b = av_clip(b, 0, isrgb8 ? 3 : 1);
            err[0] = R - r*(isrgb8 ? 36 : 255);
            err[1] = G - g*(isrgb8 ? 36 : 85);
            err[2] = B - b*(isrgb8 ? 85 : 255);
            break;
        case SWS_DITHER_A_DITHER:
            if (isrgb8) {
  /* see http://pippin.gimp.org/a_dither/ for details/origin */
#define A_DITHER(u,v)   (((((u)+((v)*236))*119)&0xff))
                r = (((R >> 19) + A_DITHER(i,y)  -96)>>8);
                g = (((G >> 19) + A_DITHER(i + 17,y) - 96)>>8);
                b = (((B >> 20) + A_DITHER(i + 17*2,y) -96)>>8);
1718 1719 1720
                r = av_clip_uintp2(r, 3);
                g = av_clip_uintp2(g, 3);
                b = av_clip_uintp2(b, 2);
1721 1722 1723 1724
            } else {
                r = (((R >> 21) + A_DITHER(i,y)-256)>>8);
                g = (((G >> 19) + A_DITHER(i + 17,y)-256)>>8);
                b = (((B >> 21) + A_DITHER(i + 17*2,y)-256)>>8);
1725 1726 1727
                r = av_clip_uintp2(r, 1);
                g = av_clip_uintp2(g, 2);
                b = av_clip_uintp2(b, 1);
1728 1729 1730 1731 1732 1733 1734 1735 1736
            }
            break;
        case SWS_DITHER_X_DITHER:
            if (isrgb8) {
  /* see http://pippin.gimp.org/a_dither/ for details/origin */
#define X_DITHER(u,v)   (((((u)^((v)*237))*181)&0x1ff)/2)
                r = (((R >> 19) + X_DITHER(i,y) - 96)>>8);
                g = (((G >> 19) + X_DITHER(i + 17,y) - 96)>>8);
                b = (((B >> 20) + X_DITHER(i + 17*2,y) - 96)>>8);
1737 1738 1739
                r = av_clip_uintp2(r, 3);
                g = av_clip_uintp2(g, 3);
                b = av_clip_uintp2(b, 2);
1740 1741 1742 1743
            } else {
                r = (((R >> 21) + X_DITHER(i,y)-256)>>8);
                g = (((G >> 19) + X_DITHER(i + 17,y)-256)>>8);
                b = (((B >> 21) + X_DITHER(i + 17*2,y)-256)>>8);
1744 1745 1746
                r = av_clip_uintp2(r, 1);
                g = av_clip_uintp2(g, 2);
                b = av_clip_uintp2(b, 1);
1747 1748 1749 1750 1751
            }

            break;
        }

1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765
        if(target == AV_PIX_FMT_BGR4_BYTE) {
            dest[0] = r + 2*g + 8*b;
        } else if(target == AV_PIX_FMT_RGB4_BYTE) {
            dest[0] = b + 2*g + 8*r;
        } else if(target == AV_PIX_FMT_BGR8) {
            dest[0] = r + 8*g + 64*b;
        } else if(target == AV_PIX_FMT_RGB8) {
            dest[0] = b + 4*g + 32*r;
        } else
            av_assert2(0);
        break;}
    }
}

1766 1767 1768 1769 1770 1771
static av_always_inline void
yuv2rgb_full_X_c_template(SwsContext *c, const int16_t *lumFilter,
                          const int16_t **lumSrc, int lumFilterSize,
                          const int16_t *chrFilter, const int16_t **chrUSrc,
                          const int16_t **chrVSrc, int chrFilterSize,
                          const int16_t **alpSrc, uint8_t *dest,
1772
                          int dstW, int y, enum AVPixelFormat target, int hasAlpha)
1773 1774
{
    int i;
1775
    int step = (target == AV_PIX_FMT_RGB24 || target == AV_PIX_FMT_BGR24) ? 3 : 4;
1776
    int err[4] = {0};
1777
    int A = 0; //init to silence warning
1778 1779 1780 1781

    if(   target == AV_PIX_FMT_BGR4_BYTE || target == AV_PIX_FMT_RGB4_BYTE
       || target == AV_PIX_FMT_BGR8      || target == AV_PIX_FMT_RGB8)
        step = 1;
1782 1783 1784

    for (i = 0; i < dstW; i++) {
        int j;
1785 1786 1787
        int Y = 1<<9;
        int U = (1<<9)-(128 << 19);
        int V = (1<<9)-(128 << 19);
1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799

        for (j = 0; j < lumFilterSize; j++) {
            Y += lumSrc[j][i] * lumFilter[j];
        }
        for (j = 0; j < chrFilterSize; j++) {
            U += chrUSrc[j][i] * chrFilter[j];
            V += chrVSrc[j][i] * chrFilter[j];
        }
        Y >>= 10;
        U >>= 10;
        V >>= 10;
        if (hasAlpha) {
1800
            A = 1 << 18;
1801 1802 1803 1804 1805 1806 1807
            for (j = 0; j < lumFilterSize; j++) {
                A += alpSrc[j][i] * lumFilter[j];
            }
            A >>= 19;
            if (A & 0x100)
                A = av_clip_uint8(A);
        }
1808
        yuv2rgb_write_full(c, dest, i, Y, A, U, V, y, target, hasAlpha, err);
1809 1810
        dest += step;
    }
1811 1812 1813
    c->dither_error[0][i] = err[0];
    c->dither_error[1][i] = err[1];
    c->dither_error[2][i] = err[2];
1814 1815
}

1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832
static av_always_inline void
yuv2rgb_full_2_c_template(SwsContext *c, const int16_t *buf[2],
                     const int16_t *ubuf[2], const int16_t *vbuf[2],
                     const int16_t *abuf[2], uint8_t *dest, int dstW,
                     int yalpha, int uvalpha, int y,
                     enum AVPixelFormat target, int hasAlpha)
{
    const int16_t *buf0  = buf[0],  *buf1  = buf[1],
                  *ubuf0 = ubuf[0], *ubuf1 = ubuf[1],
                  *vbuf0 = vbuf[0], *vbuf1 = vbuf[1],
                  *abuf0 = hasAlpha ? abuf[0] : NULL,
                  *abuf1 = hasAlpha ? abuf[1] : NULL;
    int  yalpha1 = 4096 - yalpha;
    int uvalpha1 = 4096 - uvalpha;
    int i;
    int step = (target == AV_PIX_FMT_RGB24 || target == AV_PIX_FMT_BGR24) ? 3 : 4;
    int err[4] = {0};
1833
    int A = 0; // init to silcene warning
1834

1835 1836 1837
    av_assert2(yalpha  <= 4096U);
    av_assert2(uvalpha <= 4096U);

1838 1839 1840 1841 1842 1843 1844 1845 1846 1847
    if(   target == AV_PIX_FMT_BGR4_BYTE || target == AV_PIX_FMT_RGB4_BYTE
       || target == AV_PIX_FMT_BGR8      || target == AV_PIX_FMT_RGB8)
        step = 1;

    for (i = 0; i < dstW; i++) {
        int Y = ( buf0[i] * yalpha1  +  buf1[i] * yalpha             ) >> 10; //FIXME rounding
        int U = (ubuf0[i] * uvalpha1 + ubuf1[i] * uvalpha-(128 << 19)) >> 10;
        int V = (vbuf0[i] * uvalpha1 + vbuf1[i] * uvalpha-(128 << 19)) >> 10;

        if (hasAlpha) {
1848 1849 1850
            A = (abuf0[i] * yalpha1 + abuf1[i] * yalpha + (1<<18)) >> 19;
            if (A & 0x100)
                A = av_clip_uint8(A);
1851 1852 1853 1854 1855 1856 1857 1858 1859 1860
        }

        yuv2rgb_write_full(c, dest, i, Y, A, U, V, y, target, hasAlpha, err);
        dest += step;
    }
    c->dither_error[0][i] = err[0];
    c->dither_error[1][i] = err[1];
    c->dither_error[2][i] = err[2];
}

1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877
static av_always_inline void
yuv2rgb_full_1_c_template(SwsContext *c, const int16_t *buf0,
                     const int16_t *ubuf[2], const int16_t *vbuf[2],
                     const int16_t *abuf0, uint8_t *dest, int dstW,
                     int uvalpha, int y, enum AVPixelFormat target,
                     int hasAlpha)
{
    const int16_t *ubuf0 = ubuf[0], *vbuf0 = vbuf[0];
    int i;
    int step = (target == AV_PIX_FMT_RGB24 || target == AV_PIX_FMT_BGR24) ? 3 : 4;
    int err[4] = {0};

    if(   target == AV_PIX_FMT_BGR4_BYTE || target == AV_PIX_FMT_RGB4_BYTE
       || target == AV_PIX_FMT_BGR8      || target == AV_PIX_FMT_RGB8)
        step = 1;

    if (uvalpha < 2048) {
1878
        int A = 0; //init to silence warning
1879 1880
        for (i = 0; i < dstW; i++) {
            int Y = buf0[i] << 2;
1881 1882
            int U = (ubuf0[i] - (128<<7)) * 4;
            int V = (vbuf0[i] - (128<<7)) * 4;
1883 1884

            if (hasAlpha) {
1885 1886 1887
                A = (abuf0[i] + 64) >> 7;
                if (A & 0x100)
                    A = av_clip_uint8(A);
1888 1889 1890 1891 1892 1893 1894
            }

            yuv2rgb_write_full(c, dest, i, Y, A, U, V, y, target, hasAlpha, err);
            dest += step;
        }
    } else {
        const int16_t *ubuf1 = ubuf[1], *vbuf1 = vbuf[1];
1895
        int A = 0; //init to silence warning
1896 1897 1898 1899 1900 1901
        for (i = 0; i < dstW; i++) {
            int Y = buf0[i] << 2;
            int U = (ubuf0[i] + ubuf1[i] - (128<<8)) << 1;
            int V = (vbuf0[i] + vbuf1[i] - (128<<8)) << 1;

            if (hasAlpha) {
1902 1903 1904
                A = (abuf0[i] + 64) >> 7;
                if (A & 0x100)
                    A = av_clip_uint8(A);
1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916
            }

            yuv2rgb_write_full(c, dest, i, Y, A, U, V, y, target, hasAlpha, err);
            dest += step;
        }
    }

    c->dither_error[0][i] = err[0];
    c->dither_error[1][i] = err[1];
    c->dither_error[2][i] = err[2];
}

1917
#if CONFIG_SMALL
1918 1919 1920 1921
YUV2RGBWRAPPER(yuv2, rgb_full, bgra32_full, AV_PIX_FMT_BGRA,  CONFIG_SWSCALE_ALPHA && c->needAlpha)
YUV2RGBWRAPPER(yuv2, rgb_full, abgr32_full, AV_PIX_FMT_ABGR,  CONFIG_SWSCALE_ALPHA && c->needAlpha)
YUV2RGBWRAPPER(yuv2, rgb_full, rgba32_full, AV_PIX_FMT_RGBA,  CONFIG_SWSCALE_ALPHA && c->needAlpha)
YUV2RGBWRAPPER(yuv2, rgb_full, argb32_full, AV_PIX_FMT_ARGB,  CONFIG_SWSCALE_ALPHA && c->needAlpha)
1922 1923
#else
#if CONFIG_SWSCALE_ALPHA
1924 1925 1926 1927
YUV2RGBWRAPPER(yuv2, rgb_full, bgra32_full, AV_PIX_FMT_BGRA,  1)
YUV2RGBWRAPPER(yuv2, rgb_full, abgr32_full, AV_PIX_FMT_ABGR,  1)
YUV2RGBWRAPPER(yuv2, rgb_full, rgba32_full, AV_PIX_FMT_RGBA,  1)
YUV2RGBWRAPPER(yuv2, rgb_full, argb32_full, AV_PIX_FMT_ARGB,  1)
1928
#endif
1929 1930 1931 1932
YUV2RGBWRAPPER(yuv2, rgb_full, bgrx32_full, AV_PIX_FMT_BGRA,  0)
YUV2RGBWRAPPER(yuv2, rgb_full, xbgr32_full, AV_PIX_FMT_ABGR,  0)
YUV2RGBWRAPPER(yuv2, rgb_full, rgbx32_full, AV_PIX_FMT_RGBA,  0)
YUV2RGBWRAPPER(yuv2, rgb_full, xrgb32_full, AV_PIX_FMT_ARGB,  0)
1933
#endif
1934 1935
YUV2RGBWRAPPER(yuv2, rgb_full, bgr24_full,  AV_PIX_FMT_BGR24, 0)
YUV2RGBWRAPPER(yuv2, rgb_full, rgb24_full,  AV_PIX_FMT_RGB24, 0)
1936

1937 1938 1939 1940
YUV2RGBWRAPPER(yuv2, rgb_full, bgr4_byte_full,  AV_PIX_FMT_BGR4_BYTE, 0)
YUV2RGBWRAPPER(yuv2, rgb_full, rgb4_byte_full,  AV_PIX_FMT_RGB4_BYTE, 0)
YUV2RGBWRAPPER(yuv2, rgb_full, bgr8_full,   AV_PIX_FMT_BGR8,  0)
YUV2RGBWRAPPER(yuv2, rgb_full, rgb8_full,   AV_PIX_FMT_RGB8,  0)
1941

1942 1943
static void
yuv2gbrp_full_X_c(SwsContext *c, const int16_t *lumFilter,
1944 1945 1946 1947 1948
                  const int16_t **lumSrc, int lumFilterSize,
                  const int16_t *chrFilter, const int16_t **chrUSrc,
                  const int16_t **chrVSrc, int chrFilterSize,
                  const int16_t **alpSrc, uint8_t **dest,
                  int dstW, int y)
1949
{
1950
    const AVPixFmtDescriptor *desc = av_pix_fmt_desc_get(c->dstFormat);
1951
    int i;
1952
    int hasAlpha = (desc->flags & AV_PIX_FMT_FLAG_ALPHA) && alpSrc;
1953
    uint16_t **dest16 = (uint16_t**)dest;
1954
    int SH = 22 + 8 - desc->comp[0].depth;
1955
    int A = 0; // init to silence warning
1956 1957 1958

    for (i = 0; i < dstW; i++) {
        int j;
1959 1960 1961
        int Y = 1 << 9;
        int U = (1 << 9) - (128 << 19);
        int V = (1 << 9) - (128 << 19);
1962
        int R, G, B;
1963

1964
        for (j = 0; j < lumFilterSize; j++)
1965
            Y += lumSrc[j][i] * lumFilter[j];
1966

1967 1968 1969 1970
        for (j = 0; j < chrFilterSize; j++) {
            U += chrUSrc[j][i] * chrFilter[j];
            V += chrVSrc[j][i] * chrFilter[j];
        }
1971

1972 1973 1974
        Y >>= 10;
        U >>= 10;
        V >>= 10;
1975

1976 1977
        if (hasAlpha) {
            A = 1 << 18;
1978 1979

            for (j = 0; j < lumFilterSize; j++)
1980
                A += alpSrc[j][i] * lumFilter[j];
1981

1982 1983
            if (A & 0xF8000000)
                A =  av_clip_uintp2(A, 27);
1984
        }
1985

1986 1987
        Y -= c->yuv2rgb_y_offset;
        Y *= c->yuv2rgb_y_coeff;
1988
        Y += 1 << (SH-1);
1989 1990 1991 1992
        R = Y + V * c->yuv2rgb_v2r_coeff;
        G = Y + V * c->yuv2rgb_v2g_coeff + U * c->yuv2rgb_u2g_coeff;
        B = Y +                            U * c->yuv2rgb_u2b_coeff;

1993 1994 1995 1996 1997 1998
        if ((R | G | B) & 0xC0000000) {
            R = av_clip_uintp2(R, 30);
            G = av_clip_uintp2(G, 30);
            B = av_clip_uintp2(B, 30);
        }

1999 2000 2001 2002
        if (SH != 22) {
            dest16[0][i] = G >> SH;
            dest16[1][i] = B >> SH;
            dest16[2][i] = R >> SH;
2003
            if (hasAlpha)
2004
                dest16[3][i] = A >> (SH - 3);
2005 2006 2007 2008
        } else {
            dest[0][i] = G >> 22;
            dest[1][i] = B >> 22;
            dest[2][i] = R >> 22;
2009
            if (hasAlpha)
2010
                dest[3][i] = A >> 19;
2011 2012 2013 2014 2015 2016 2017
        }
    }
    if (SH != 22 && (!isBE(c->dstFormat)) != (!HAVE_BIGENDIAN)) {
        for (i = 0; i < dstW; i++) {
            dest16[0][i] = av_bswap16(dest16[0][i]);
            dest16[1][i] = av_bswap16(dest16[1][i]);
            dest16[2][i] = av_bswap16(dest16[2][i]);
2018 2019
            if (hasAlpha)
                dest16[3][i] = av_bswap16(dest16[3][i]);
2020
        }
2021 2022 2023
    }
}

2024 2025 2026 2027 2028
static void
yuv2gbrp16_full_X_c(SwsContext *c, const int16_t *lumFilter,
                    const int16_t **lumSrcx, int lumFilterSize,
                    const int16_t *chrFilter, const int16_t **chrUSrcx,
                    const int16_t **chrVSrcx, int chrFilterSize,
2029
                    const int16_t **alpSrcx, uint8_t **dest,
2030 2031 2032 2033
                    int dstW, int y)
{
    const AVPixFmtDescriptor *desc = av_pix_fmt_desc_get(c->dstFormat);
    int i;
2034
    int hasAlpha = (desc->flags & AV_PIX_FMT_FLAG_ALPHA) && alpSrcx;
2035 2036 2037 2038
    uint16_t **dest16 = (uint16_t**)dest;
    const int32_t **lumSrc  = (const int32_t**)lumSrcx;
    const int32_t **chrUSrc = (const int32_t**)chrUSrcx;
    const int32_t **chrVSrc = (const int32_t**)chrVSrcx;
2039
    const int32_t **alpSrc  = (const int32_t**)alpSrcx;
2040 2041 2042 2043 2044 2045

    for (i = 0; i < dstW; i++) {
        int j;
        int Y = -0x40000000;
        int U = -(128 << 23);
        int V = -(128 << 23);
2046
        int R, G, B, A;
2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061

        for (j = 0; j < lumFilterSize; j++)
            Y += lumSrc[j][i] * (unsigned)lumFilter[j];

        for (j = 0; j < chrFilterSize; j++) {
            U += chrUSrc[j][i] * (unsigned)chrFilter[j];
            V += chrVSrc[j][i] * (unsigned)chrFilter[j];
        }

        Y >>= 14;
        Y += 0x10000;
        U >>= 14;
        V >>= 14;

        if (hasAlpha) {
2062
            A = -0x40000000;
2063 2064 2065 2066

            for (j = 0; j < lumFilterSize; j++)
                A += alpSrc[j][i] * lumFilter[j];

2067 2068
            A >>= 1;
            A += 0x20002000;
2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085
        }

        Y -= c->yuv2rgb_y_offset;
        Y *= c->yuv2rgb_y_coeff;
        Y += 1 << 13;
        R = V * c->yuv2rgb_v2r_coeff;
        G = V * c->yuv2rgb_v2g_coeff + U * c->yuv2rgb_u2g_coeff;
        B =                            U * c->yuv2rgb_u2b_coeff;

        R = av_clip_uintp2(Y + R, 30);
        G = av_clip_uintp2(Y + G, 30);
        B = av_clip_uintp2(Y + B, 30);

        dest16[0][i] = G >> 14;
        dest16[1][i] = B >> 14;
        dest16[2][i] = R >> 14;
        if (hasAlpha)
2086
            dest16[3][i] = av_clip_uintp2(A, 30) >> 14;
2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098
    }
    if ((!isBE(c->dstFormat)) != (!HAVE_BIGENDIAN)) {
        for (i = 0; i < dstW; i++) {
            dest16[0][i] = av_bswap16(dest16[0][i]);
            dest16[1][i] = av_bswap16(dest16[1][i]);
            dest16[2][i] = av_bswap16(dest16[2][i]);
            if (hasAlpha)
                dest16[3][i] = av_bswap16(dest16[3][i]);
        }
    }
}

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
static void
yuv2ya8_1_c(SwsContext *c, const int16_t *buf0,
            const int16_t *ubuf[2], const int16_t *vbuf[2],
            const int16_t *abuf0, uint8_t *dest, int dstW,
            int uvalpha, int y)
{
    int hasAlpha = !!abuf0;
    int i;

    for (i = 0; i < dstW; i++) {
        int Y = (buf0[i] + 64) >> 7;
        int A;

        Y = av_clip_uint8(Y);

        if (hasAlpha) {
            A = (abuf0[i] + 64) >> 7;
            if (A & 0x100)
                A = av_clip_uint8(A);
        }

        dest[i * 2    ] = Y;
        dest[i * 2 + 1] = hasAlpha ? A : 255;
    }
}

static void
yuv2ya8_2_c(SwsContext *c, const int16_t *buf[2],
            const int16_t *ubuf[2], const int16_t *vbuf[2],
            const int16_t *abuf[2], uint8_t *dest, int dstW,
            int yalpha, int uvalpha, int y)
{
2131
    int hasAlpha = abuf && abuf[0] && abuf[1];
2132 2133 2134 2135 2136 2137
    const int16_t *buf0  = buf[0],  *buf1  = buf[1],
                  *abuf0 = hasAlpha ? abuf[0] : NULL,
                  *abuf1 = hasAlpha ? abuf[1] : NULL;
    int  yalpha1 = 4096 - yalpha;
    int i;

2138 2139
    av_assert2(yalpha  <= 4096U);

2140
    for (i = 0; i < dstW; i++) {
2141
        int Y = (buf0[i] * yalpha1 + buf1[i] * yalpha) >> 19;
2142 2143 2144 2145 2146
        int A;

        Y = av_clip_uint8(Y);

        if (hasAlpha) {
2147
            A = (abuf0[i] * yalpha1 + abuf1[i] * yalpha) >> 19;
2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191
            A = av_clip_uint8(A);
        }

        dest[i * 2    ] = Y;
        dest[i * 2 + 1] = hasAlpha ? A : 255;
    }
}

static void
yuv2ya8_X_c(SwsContext *c, const int16_t *lumFilter,
            const int16_t **lumSrc, int lumFilterSize,
            const int16_t *chrFilter, const int16_t **chrUSrc,
            const int16_t **chrVSrc, int chrFilterSize,
            const int16_t **alpSrc, uint8_t *dest, int dstW, int y)
{
    int hasAlpha = !!alpSrc;
    int i;

    for (i = 0; i < dstW; i++) {
        int j;
        int Y = 1 << 18, A = 1 << 18;

        for (j = 0; j < lumFilterSize; j++)
            Y += lumSrc[j][i] * lumFilter[j];

        Y >>= 19;
        if (Y  & 0x100)
            Y = av_clip_uint8(Y);

        if (hasAlpha) {
            for (j = 0; j < lumFilterSize; j++)
                A += alpSrc[j][i] * lumFilter[j];

            A >>= 19;

            if (A & 0x100)
                A = av_clip_uint8(A);
        }

        dest[2 * i    ] = Y;
        dest[2 * i + 1] = hasAlpha ? A : 255;
    }
}

2192 2193
static void
yuv2ayuv64le_X_c(SwsContext *c, const int16_t *lumFilter,
2194 2195 2196 2197
                 const int16_t **_lumSrc, int lumFilterSize,
                 const int16_t *chrFilter, const int16_t **_chrUSrc,
                 const int16_t **_chrVSrc, int chrFilterSize,
                 const int16_t **_alpSrc, uint8_t *dest, int dstW, int y)
2198
{
2199 2200 2201 2202
    const int32_t **lumSrc  = (const int32_t **) _lumSrc,
                  **chrUSrc = (const int32_t **) _chrUSrc,
                  **chrVSrc = (const int32_t **) _chrVSrc,
                  **alpSrc  = (const int32_t **) _alpSrc;
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 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240
    int hasAlpha = !!alpSrc;
    int i;

    for (i = 0; i < dstW; i++) {
        int Y = 1 << 14, U = 1 << 14;
        int V = 1 << 14, A = 1 << 14;
        int j;

        Y -= 0x40000000;
        U -= 0x40000000;
        V -= 0x40000000;
        A -= 0x40000000;

        for (j = 0; j < lumFilterSize; j++)
            Y += lumSrc[j][i] * (unsigned)lumFilter[j];

        for (j = 0; j < chrFilterSize; j++)
            U += chrUSrc[j][i] * (unsigned)chrFilter[j];

        for (j = 0; j < chrFilterSize; j++)
            V += chrVSrc[j][i] * (unsigned)chrFilter[j];

        if (hasAlpha)
            for (j = 0; j < lumFilterSize; j++)
                A += alpSrc[j][i] * (unsigned)lumFilter[j];

        Y = 0x8000 + av_clip_int16(Y >> 15);
        U = 0x8000 + av_clip_int16(U >> 15);
        V = 0x8000 + av_clip_int16(V >> 15);
        A = 0x8000 + av_clip_int16(A >> 15);

        AV_WL16(dest + 8 * i, hasAlpha ? A : 65535);
        AV_WL16(dest + 8 * i + 2, Y);
        AV_WL16(dest + 8 * i + 4, U);
        AV_WL16(dest + 8 * i + 6, V);
    }
}

2241 2242 2243 2244 2245 2246
av_cold void ff_sws_init_output_funcs(SwsContext *c,
                                      yuv2planar1_fn *yuv2plane1,
                                      yuv2planarX_fn *yuv2planeX,
                                      yuv2interleavedX_fn *yuv2nv12cX,
                                      yuv2packed1_fn *yuv2packed1,
                                      yuv2packed2_fn *yuv2packed2,
2247 2248
                                      yuv2packedX_fn *yuv2packedX,
                                      yuv2anyX_fn *yuv2anyX)
2249
{
2250
    enum AVPixelFormat dstFormat = c->dstFormat;
2251
    const AVPixFmtDescriptor *desc = av_pix_fmt_desc_get(dstFormat);
2252

2253 2254 2255 2256 2257
    if (dstFormat == AV_PIX_FMT_P010LE || dstFormat == AV_PIX_FMT_P010BE) {
        *yuv2plane1 = isBE(dstFormat) ? yuv2p010l1_BE_c : yuv2p010l1_LE_c;
        *yuv2planeX = isBE(dstFormat) ? yuv2p010lX_BE_c : yuv2p010lX_LE_c;
        *yuv2nv12cX = yuv2p010cX_c;
    } else if (is16BPS(dstFormat)) {
2258 2259
        *yuv2planeX = isBE(dstFormat) ? yuv2planeX_16BE_c  : yuv2planeX_16LE_c;
        *yuv2plane1 = isBE(dstFormat) ? yuv2plane1_16BE_c  : yuv2plane1_16LE_c;
2260
    } else if (isNBPS(dstFormat)) {
2261
        if (desc->comp[0].depth == 9) {
2262 2263
            *yuv2planeX = isBE(dstFormat) ? yuv2planeX_9BE_c  : yuv2planeX_9LE_c;
            *yuv2plane1 = isBE(dstFormat) ? yuv2plane1_9BE_c  : yuv2plane1_9LE_c;
2264
        } else if (desc->comp[0].depth == 10) {
2265 2266
            *yuv2planeX = isBE(dstFormat) ? yuv2planeX_10BE_c  : yuv2planeX_10LE_c;
            *yuv2plane1 = isBE(dstFormat) ? yuv2plane1_10BE_c  : yuv2plane1_10LE_c;
2267
        } else if (desc->comp[0].depth == 12) {
2268 2269
            *yuv2planeX = isBE(dstFormat) ? yuv2planeX_12BE_c  : yuv2planeX_12LE_c;
            *yuv2plane1 = isBE(dstFormat) ? yuv2plane1_12BE_c  : yuv2plane1_12LE_c;
2270
        } else if (desc->comp[0].depth == 14) {
2271 2272 2273 2274
            *yuv2planeX = isBE(dstFormat) ? yuv2planeX_14BE_c  : yuv2planeX_14LE_c;
            *yuv2plane1 = isBE(dstFormat) ? yuv2plane1_14BE_c  : yuv2plane1_14LE_c;
        } else
            av_assert0(0);
2275 2276 2277
    } else {
        *yuv2plane1 = yuv2plane1_8_c;
        *yuv2planeX = yuv2planeX_8_c;
2278
        if (dstFormat == AV_PIX_FMT_NV12 || dstFormat == AV_PIX_FMT_NV21)
2279 2280 2281 2282 2283
            *yuv2nv12cX = yuv2nv12cX_c;
    }

    if(c->flags & SWS_FULL_CHR_H_INT) {
        switch (dstFormat) {
2284
            case AV_PIX_FMT_RGBA:
2285 2286
#if CONFIG_SMALL
                *yuv2packedX = yuv2rgba32_full_X_c;
2287
                *yuv2packed2 = yuv2rgba32_full_2_c;
2288
                *yuv2packed1 = yuv2rgba32_full_1_c;
2289 2290
#else
#if CONFIG_SWSCALE_ALPHA
2291
                if (c->needAlpha) {
2292
                    *yuv2packedX = yuv2rgba32_full_X_c;
2293
                    *yuv2packed2 = yuv2rgba32_full_2_c;
2294
                    *yuv2packed1 = yuv2rgba32_full_1_c;
2295 2296 2297 2298
                } else
#endif /* CONFIG_SWSCALE_ALPHA */
                {
                    *yuv2packedX = yuv2rgbx32_full_X_c;
2299
                    *yuv2packed2 = yuv2rgbx32_full_2_c;
2300
                    *yuv2packed1 = yuv2rgbx32_full_1_c;
2301 2302 2303
                }
#endif /* !CONFIG_SMALL */
                break;
2304
            case AV_PIX_FMT_ARGB:
2305 2306
#if CONFIG_SMALL
                *yuv2packedX = yuv2argb32_full_X_c;
2307
                *yuv2packed2 = yuv2argb32_full_2_c;
2308
                *yuv2packed1 = yuv2argb32_full_1_c;
2309 2310
#else
#if CONFIG_SWSCALE_ALPHA
2311
                if (c->needAlpha) {
2312
                    *yuv2packedX = yuv2argb32_full_X_c;
2313
                    *yuv2packed2 = yuv2argb32_full_2_c;
2314
                    *yuv2packed1 = yuv2argb32_full_1_c;
2315 2316 2317 2318
                } else
#endif /* CONFIG_SWSCALE_ALPHA */
                {
                    *yuv2packedX = yuv2xrgb32_full_X_c;
2319
                    *yuv2packed2 = yuv2xrgb32_full_2_c;
2320
                    *yuv2packed1 = yuv2xrgb32_full_1_c;
2321 2322 2323
                }
#endif /* !CONFIG_SMALL */
                break;
2324
            case AV_PIX_FMT_BGRA:
2325 2326
#if CONFIG_SMALL
                *yuv2packedX = yuv2bgra32_full_X_c;
2327
                *yuv2packed2 = yuv2bgra32_full_2_c;
2328
                *yuv2packed1 = yuv2bgra32_full_1_c;
2329 2330
#else
#if CONFIG_SWSCALE_ALPHA
2331
                if (c->needAlpha) {
2332
                    *yuv2packedX = yuv2bgra32_full_X_c;
2333
                    *yuv2packed2 = yuv2bgra32_full_2_c;
2334
                    *yuv2packed1 = yuv2bgra32_full_1_c;
2335 2336 2337 2338
                } else
#endif /* CONFIG_SWSCALE_ALPHA */
                {
                    *yuv2packedX = yuv2bgrx32_full_X_c;
2339
                    *yuv2packed2 = yuv2bgrx32_full_2_c;
2340
                    *yuv2packed1 = yuv2bgrx32_full_1_c;
2341 2342 2343
                }
#endif /* !CONFIG_SMALL */
                break;
2344
            case AV_PIX_FMT_ABGR:
2345 2346
#if CONFIG_SMALL
                *yuv2packedX = yuv2abgr32_full_X_c;
2347
                *yuv2packed2 = yuv2abgr32_full_2_c;
2348
                *yuv2packed1 = yuv2abgr32_full_1_c;
2349 2350
#else
#if CONFIG_SWSCALE_ALPHA
2351
                if (c->needAlpha) {
2352
                    *yuv2packedX = yuv2abgr32_full_X_c;
2353
                    *yuv2packed2 = yuv2abgr32_full_2_c;
2354
                    *yuv2packed1 = yuv2abgr32_full_1_c;
2355 2356 2357 2358
                } else
#endif /* CONFIG_SWSCALE_ALPHA */
                {
                    *yuv2packedX = yuv2xbgr32_full_X_c;
2359
                    *yuv2packed2 = yuv2xbgr32_full_2_c;
2360
                    *yuv2packed1 = yuv2xbgr32_full_1_c;
2361 2362 2363
                }
#endif /* !CONFIG_SMALL */
                break;
2364 2365
        case AV_PIX_FMT_RGBA64LE:
#if CONFIG_SWSCALE_ALPHA
2366
            if (c->needAlpha) {
2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379
                *yuv2packedX = yuv2rgba64le_full_X_c;
                *yuv2packed2 = yuv2rgba64le_full_2_c;
                *yuv2packed1 = yuv2rgba64le_full_1_c;
            } else
#endif /* CONFIG_SWSCALE_ALPHA */
            {
                *yuv2packedX = yuv2rgbx64le_full_X_c;
                *yuv2packed2 = yuv2rgbx64le_full_2_c;
                *yuv2packed1 = yuv2rgbx64le_full_1_c;
            }
            break;
        case AV_PIX_FMT_RGBA64BE:
#if CONFIG_SWSCALE_ALPHA
2380
            if (c->needAlpha) {
2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393
                *yuv2packedX = yuv2rgba64be_full_X_c;
                *yuv2packed2 = yuv2rgba64be_full_2_c;
                *yuv2packed1 = yuv2rgba64be_full_1_c;
            } else
#endif /* CONFIG_SWSCALE_ALPHA */
            {
                *yuv2packedX = yuv2rgbx64be_full_X_c;
                *yuv2packed2 = yuv2rgbx64be_full_2_c;
                *yuv2packed1 = yuv2rgbx64be_full_1_c;
            }
            break;
        case AV_PIX_FMT_BGRA64LE:
#if CONFIG_SWSCALE_ALPHA
2394
            if (c->needAlpha) {
2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407
                *yuv2packedX = yuv2bgra64le_full_X_c;
                *yuv2packed2 = yuv2bgra64le_full_2_c;
                *yuv2packed1 = yuv2bgra64le_full_1_c;
            } else
#endif /* CONFIG_SWSCALE_ALPHA */
            {
                *yuv2packedX = yuv2bgrx64le_full_X_c;
                *yuv2packed2 = yuv2bgrx64le_full_2_c;
                *yuv2packed1 = yuv2bgrx64le_full_1_c;
            }
            break;
        case AV_PIX_FMT_BGRA64BE:
#if CONFIG_SWSCALE_ALPHA
2408
            if (c->needAlpha) {
2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421
                *yuv2packedX = yuv2bgra64be_full_X_c;
                *yuv2packed2 = yuv2bgra64be_full_2_c;
                *yuv2packed1 = yuv2bgra64be_full_1_c;
            } else
#endif /* CONFIG_SWSCALE_ALPHA */
            {
                *yuv2packedX = yuv2bgrx64be_full_X_c;
                *yuv2packed2 = yuv2bgrx64be_full_2_c;
                *yuv2packed1 = yuv2bgrx64be_full_1_c;
            }
            break;

        case AV_PIX_FMT_RGB24:
2422
            *yuv2packedX = yuv2rgb24_full_X_c;
2423
            *yuv2packed2 = yuv2rgb24_full_2_c;
2424
            *yuv2packed1 = yuv2rgb24_full_1_c;
2425
            break;
2426
        case AV_PIX_FMT_BGR24:
2427
            *yuv2packedX = yuv2bgr24_full_X_c;
2428
            *yuv2packed2 = yuv2bgr24_full_2_c;
2429
            *yuv2packed1 = yuv2bgr24_full_1_c;
2430
            break;
2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450
        case AV_PIX_FMT_RGB48LE:
            *yuv2packedX = yuv2rgb48le_full_X_c;
            *yuv2packed2 = yuv2rgb48le_full_2_c;
            *yuv2packed1 = yuv2rgb48le_full_1_c;
            break;
        case AV_PIX_FMT_BGR48LE:
            *yuv2packedX = yuv2bgr48le_full_X_c;
            *yuv2packed2 = yuv2bgr48le_full_2_c;
            *yuv2packed1 = yuv2bgr48le_full_1_c;
            break;
        case AV_PIX_FMT_RGB48BE:
            *yuv2packedX = yuv2rgb48be_full_X_c;
            *yuv2packed2 = yuv2rgb48be_full_2_c;
            *yuv2packed1 = yuv2rgb48be_full_1_c;
            break;
        case AV_PIX_FMT_BGR48BE:
            *yuv2packedX = yuv2bgr48be_full_X_c;
            *yuv2packed2 = yuv2bgr48be_full_2_c;
            *yuv2packed1 = yuv2bgr48be_full_1_c;
            break;
2451 2452
        case AV_PIX_FMT_BGR4_BYTE:
            *yuv2packedX = yuv2bgr4_byte_full_X_c;
2453
            *yuv2packed2 = yuv2bgr4_byte_full_2_c;
2454
            *yuv2packed1 = yuv2bgr4_byte_full_1_c;
2455 2456 2457
            break;
        case AV_PIX_FMT_RGB4_BYTE:
            *yuv2packedX = yuv2rgb4_byte_full_X_c;
2458
            *yuv2packed2 = yuv2rgb4_byte_full_2_c;
2459
            *yuv2packed1 = yuv2rgb4_byte_full_1_c;
2460 2461 2462
            break;
        case AV_PIX_FMT_BGR8:
            *yuv2packedX = yuv2bgr8_full_X_c;
2463
            *yuv2packed2 = yuv2bgr8_full_2_c;
2464
            *yuv2packed1 = yuv2bgr8_full_1_c;
2465 2466 2467
            break;
        case AV_PIX_FMT_RGB8:
            *yuv2packedX = yuv2rgb8_full_X_c;
2468
            *yuv2packed2 = yuv2rgb8_full_2_c;
2469
            *yuv2packed1 = yuv2rgb8_full_1_c;
2470
            break;
2471
        case AV_PIX_FMT_GBRP:
2472 2473 2474 2475 2476 2477 2478 2479
        case AV_PIX_FMT_GBRP9BE:
        case AV_PIX_FMT_GBRP9LE:
        case AV_PIX_FMT_GBRP10BE:
        case AV_PIX_FMT_GBRP10LE:
        case AV_PIX_FMT_GBRP12BE:
        case AV_PIX_FMT_GBRP12LE:
        case AV_PIX_FMT_GBRP14BE:
        case AV_PIX_FMT_GBRP14LE:
2480
        case AV_PIX_FMT_GBRAP:
2481 2482
        case AV_PIX_FMT_GBRAP10BE:
        case AV_PIX_FMT_GBRAP10LE:
2483 2484
        case AV_PIX_FMT_GBRAP12BE:
        case AV_PIX_FMT_GBRAP12LE:
2485 2486
            *yuv2anyX = yuv2gbrp_full_X_c;
            break;
2487 2488 2489 2490 2491 2492
        case AV_PIX_FMT_GBRP16BE:
        case AV_PIX_FMT_GBRP16LE:
        case AV_PIX_FMT_GBRAP16BE:
        case AV_PIX_FMT_GBRAP16LE:
            *yuv2anyX = yuv2gbrp16_full_X_c;
            break;
2493
        }
2494
        if (!*yuv2packedX && !*yuv2anyX)
2495
            goto YUV_PACKED;
2496
    } else {
2497
        YUV_PACKED:
2498
        switch (dstFormat) {
Paul B Mahol's avatar
Paul B Mahol committed
2499 2500
        case AV_PIX_FMT_RGBA64LE:
#if CONFIG_SWSCALE_ALPHA
2501
            if (c->needAlpha) {
Paul B Mahol's avatar
Paul B Mahol committed
2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514
                *yuv2packed1 = yuv2rgba64le_1_c;
                *yuv2packed2 = yuv2rgba64le_2_c;
                *yuv2packedX = yuv2rgba64le_X_c;
            } else
#endif /* CONFIG_SWSCALE_ALPHA */
            {
                *yuv2packed1 = yuv2rgbx64le_1_c;
                *yuv2packed2 = yuv2rgbx64le_2_c;
                *yuv2packedX = yuv2rgbx64le_X_c;
            }
            break;
        case AV_PIX_FMT_RGBA64BE:
#if CONFIG_SWSCALE_ALPHA
2515
            if (c->needAlpha) {
Paul B Mahol's avatar
Paul B Mahol committed
2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526
                *yuv2packed1 = yuv2rgba64be_1_c;
                *yuv2packed2 = yuv2rgba64be_2_c;
                *yuv2packedX = yuv2rgba64be_X_c;
            } else
#endif /* CONFIG_SWSCALE_ALPHA */
            {
                *yuv2packed1 = yuv2rgbx64be_1_c;
                *yuv2packed2 = yuv2rgbx64be_2_c;
                *yuv2packedX = yuv2rgbx64be_X_c;
            }
            break;
2527 2528
        case AV_PIX_FMT_BGRA64LE:
#if CONFIG_SWSCALE_ALPHA
2529
            if (c->needAlpha) {
2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 2542
                *yuv2packed1 = yuv2bgra64le_1_c;
                *yuv2packed2 = yuv2bgra64le_2_c;
                *yuv2packedX = yuv2bgra64le_X_c;
            } else
#endif /* CONFIG_SWSCALE_ALPHA */
            {
                *yuv2packed1 = yuv2bgrx64le_1_c;
                *yuv2packed2 = yuv2bgrx64le_2_c;
                *yuv2packedX = yuv2bgrx64le_X_c;
            }
            break;
        case AV_PIX_FMT_BGRA64BE:
#if CONFIG_SWSCALE_ALPHA
2543
            if (c->needAlpha) {
2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554
                *yuv2packed1 = yuv2bgra64be_1_c;
                *yuv2packed2 = yuv2bgra64be_2_c;
                *yuv2packedX = yuv2bgra64be_X_c;
            } else
#endif /* CONFIG_SWSCALE_ALPHA */
            {
                *yuv2packed1 = yuv2bgrx64be_1_c;
                *yuv2packed2 = yuv2bgrx64be_2_c;
                *yuv2packedX = yuv2bgrx64be_X_c;
            }
            break;
2555
        case AV_PIX_FMT_RGB48LE:
2556 2557 2558 2559
            *yuv2packed1 = yuv2rgb48le_1_c;
            *yuv2packed2 = yuv2rgb48le_2_c;
            *yuv2packedX = yuv2rgb48le_X_c;
            break;
2560
        case AV_PIX_FMT_RGB48BE:
2561 2562 2563 2564
            *yuv2packed1 = yuv2rgb48be_1_c;
            *yuv2packed2 = yuv2rgb48be_2_c;
            *yuv2packedX = yuv2rgb48be_X_c;
            break;
2565
        case AV_PIX_FMT_BGR48LE:
2566 2567 2568 2569
            *yuv2packed1 = yuv2bgr48le_1_c;
            *yuv2packed2 = yuv2bgr48le_2_c;
            *yuv2packedX = yuv2bgr48le_X_c;
            break;
2570
        case AV_PIX_FMT_BGR48BE:
2571 2572 2573 2574
            *yuv2packed1 = yuv2bgr48be_1_c;
            *yuv2packed2 = yuv2bgr48be_2_c;
            *yuv2packedX = yuv2bgr48be_X_c;
            break;
2575 2576
        case AV_PIX_FMT_RGB32:
        case AV_PIX_FMT_BGR32:
2577 2578 2579 2580 2581 2582
#if CONFIG_SMALL
            *yuv2packed1 = yuv2rgb32_1_c;
            *yuv2packed2 = yuv2rgb32_2_c;
            *yuv2packedX = yuv2rgb32_X_c;
#else
#if CONFIG_SWSCALE_ALPHA
2583
                if (c->needAlpha) {
2584 2585 2586 2587 2588 2589 2590 2591 2592 2593 2594 2595
                    *yuv2packed1 = yuv2rgba32_1_c;
                    *yuv2packed2 = yuv2rgba32_2_c;
                    *yuv2packedX = yuv2rgba32_X_c;
                } else
#endif /* CONFIG_SWSCALE_ALPHA */
                {
                    *yuv2packed1 = yuv2rgbx32_1_c;
                    *yuv2packed2 = yuv2rgbx32_2_c;
                    *yuv2packedX = yuv2rgbx32_X_c;
                }
#endif /* !CONFIG_SMALL */
            break;
2596 2597
        case AV_PIX_FMT_RGB32_1:
        case AV_PIX_FMT_BGR32_1:
2598 2599 2600 2601 2602 2603
#if CONFIG_SMALL
                *yuv2packed1 = yuv2rgb32_1_1_c;
                *yuv2packed2 = yuv2rgb32_1_2_c;
                *yuv2packedX = yuv2rgb32_1_X_c;
#else
#if CONFIG_SWSCALE_ALPHA
2604
                if (c->needAlpha) {
2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616
                    *yuv2packed1 = yuv2rgba32_1_1_c;
                    *yuv2packed2 = yuv2rgba32_1_2_c;
                    *yuv2packedX = yuv2rgba32_1_X_c;
                } else
#endif /* CONFIG_SWSCALE_ALPHA */
                {
                    *yuv2packed1 = yuv2rgbx32_1_1_c;
                    *yuv2packed2 = yuv2rgbx32_1_2_c;
                    *yuv2packedX = yuv2rgbx32_1_X_c;
                }
#endif /* !CONFIG_SMALL */
                break;
2617
        case AV_PIX_FMT_RGB24:
2618 2619 2620 2621
            *yuv2packed1 = yuv2rgb24_1_c;
            *yuv2packed2 = yuv2rgb24_2_c;
            *yuv2packedX = yuv2rgb24_X_c;
            break;
2622
        case AV_PIX_FMT_BGR24:
2623 2624 2625 2626
            *yuv2packed1 = yuv2bgr24_1_c;
            *yuv2packed2 = yuv2bgr24_2_c;
            *yuv2packedX = yuv2bgr24_X_c;
            break;
2627 2628 2629 2630
        case AV_PIX_FMT_RGB565LE:
        case AV_PIX_FMT_RGB565BE:
        case AV_PIX_FMT_BGR565LE:
        case AV_PIX_FMT_BGR565BE:
2631 2632 2633 2634
            *yuv2packed1 = yuv2rgb16_1_c;
            *yuv2packed2 = yuv2rgb16_2_c;
            *yuv2packedX = yuv2rgb16_X_c;
            break;
2635 2636 2637 2638
        case AV_PIX_FMT_RGB555LE:
        case AV_PIX_FMT_RGB555BE:
        case AV_PIX_FMT_BGR555LE:
        case AV_PIX_FMT_BGR555BE:
2639 2640 2641 2642
            *yuv2packed1 = yuv2rgb15_1_c;
            *yuv2packed2 = yuv2rgb15_2_c;
            *yuv2packedX = yuv2rgb15_X_c;
            break;
2643 2644 2645 2646
        case AV_PIX_FMT_RGB444LE:
        case AV_PIX_FMT_RGB444BE:
        case AV_PIX_FMT_BGR444LE:
        case AV_PIX_FMT_BGR444BE:
2647 2648 2649 2650
            *yuv2packed1 = yuv2rgb12_1_c;
            *yuv2packed2 = yuv2rgb12_2_c;
            *yuv2packedX = yuv2rgb12_X_c;
            break;
2651 2652
        case AV_PIX_FMT_RGB8:
        case AV_PIX_FMT_BGR8:
2653 2654 2655 2656
            *yuv2packed1 = yuv2rgb8_1_c;
            *yuv2packed2 = yuv2rgb8_2_c;
            *yuv2packedX = yuv2rgb8_X_c;
            break;
2657 2658
        case AV_PIX_FMT_RGB4:
        case AV_PIX_FMT_BGR4:
2659 2660 2661 2662
            *yuv2packed1 = yuv2rgb4_1_c;
            *yuv2packed2 = yuv2rgb4_2_c;
            *yuv2packedX = yuv2rgb4_X_c;
            break;
2663 2664
        case AV_PIX_FMT_RGB4_BYTE:
        case AV_PIX_FMT_BGR4_BYTE:
2665 2666 2667 2668 2669 2670 2671
            *yuv2packed1 = yuv2rgb4b_1_c;
            *yuv2packed2 = yuv2rgb4b_2_c;
            *yuv2packedX = yuv2rgb4b_X_c;
            break;
        }
    }
    switch (dstFormat) {
2672
    case AV_PIX_FMT_MONOWHITE:
2673 2674 2675 2676
        *yuv2packed1 = yuv2monowhite_1_c;
        *yuv2packed2 = yuv2monowhite_2_c;
        *yuv2packedX = yuv2monowhite_X_c;
        break;
2677
    case AV_PIX_FMT_MONOBLACK:
2678 2679 2680 2681
        *yuv2packed1 = yuv2monoblack_1_c;
        *yuv2packed2 = yuv2monoblack_2_c;
        *yuv2packedX = yuv2monoblack_X_c;
        break;
2682
    case AV_PIX_FMT_YUYV422:
2683 2684 2685 2686
        *yuv2packed1 = yuv2yuyv422_1_c;
        *yuv2packed2 = yuv2yuyv422_2_c;
        *yuv2packedX = yuv2yuyv422_X_c;
        break;
2687 2688 2689 2690 2691
    case AV_PIX_FMT_YVYU422:
        *yuv2packed1 = yuv2yvyu422_1_c;
        *yuv2packed2 = yuv2yvyu422_2_c;
        *yuv2packedX = yuv2yvyu422_X_c;
        break;
2692
    case AV_PIX_FMT_UYVY422:
2693 2694 2695 2696
        *yuv2packed1 = yuv2uyvy422_1_c;
        *yuv2packed2 = yuv2uyvy422_2_c;
        *yuv2packedX = yuv2uyvy422_X_c;
        break;
2697 2698 2699 2700 2701
    case AV_PIX_FMT_YA8:
        *yuv2packed1 = yuv2ya8_1_c;
        *yuv2packed2 = yuv2ya8_2_c;
        *yuv2packedX = yuv2ya8_X_c;
        break;
2702 2703 2704
    case AV_PIX_FMT_AYUV64LE:
        *yuv2packedX = yuv2ayuv64le_X_c;
        break;
2705 2706
    }
}