input.c 36.1 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38
/*
 * Copyright (C) 2001-2003 Michael Niedermayer <michaelni@gmx.at>
 *
 * This file is part of Libav.
 *
 * Libav is free software; you can redistribute it and/or
 * 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.
 *
 * Libav is distributed in the hope that it will be useful,
 * 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
 * License along with Libav; if not, write to the Free Software
 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
 */

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

#include "libavutil/avutil.h"
#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"

#define RGB2YUV_SHIFT 15
39 40 41 42 43 44 45 46 47
#define BY  ((int)(0.114 * 219 / 255 * (1 << RGB2YUV_SHIFT) + 0.5))
#define BV (-(int)(0.081 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5))
#define BU  ((int)(0.500 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5))
#define GY  ((int)(0.587 * 219 / 255 * (1 << RGB2YUV_SHIFT) + 0.5))
#define GV (-(int)(0.419 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5))
#define GU (-(int)(0.331 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5))
#define RY  ((int)(0.299 * 219 / 255 * (1 << RGB2YUV_SHIFT) + 0.5))
#define RV  ((int)(0.500 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5))
#define RU (-(int)(0.169 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5))
48 49 50 51 52 53

#define input_pixel(pos) (isBE(origin) ? AV_RB16(pos) : AV_RL16(pos))

#define r ((origin == PIX_FMT_BGR48BE || origin == PIX_FMT_BGR48LE) ? b_r : r_b)
#define b ((origin == PIX_FMT_BGR48BE || origin == PIX_FMT_BGR48LE) ? r_b : b_r)

54 55 56
static av_always_inline void rgb48ToY_c_template(uint16_t *dst,
                                                 const uint16_t *src, int width,
                                                 enum PixelFormat origin)
57 58 59
{
    int i;
    for (i = 0; i < width; i++) {
60 61 62
        unsigned int r_b = input_pixel(&src[i * 3 + 0]);
        unsigned int g   = input_pixel(&src[i * 3 + 1]);
        unsigned int b_r = input_pixel(&src[i * 3 + 2]);
63

64
        dst[i] = (RY * r + GY * g + BY * b + (0x2001 << (RGB2YUV_SHIFT - 1))) >> RGB2YUV_SHIFT;
65 66 67
    }
}

68 69 70 71 72 73
static av_always_inline void rgb48ToUV_c_template(uint16_t *dstU,
                                                  uint16_t *dstV,
                                                  const uint16_t *src1,
                                                  const uint16_t *src2,
                                                  int width,
                                                  enum PixelFormat origin)
74 75
{
    int i;
76
    assert(src1 == src2);
77
    for (i = 0; i < width; i++) {
78 79 80
        int r_b = input_pixel(&src1[i * 3 + 0]);
        int g   = input_pixel(&src1[i * 3 + 1]);
        int b_r = input_pixel(&src1[i * 3 + 2]);
81

82 83
        dstU[i] = (RU * r + GU * g + BU * b + (0x10001 << (RGB2YUV_SHIFT - 1))) >> RGB2YUV_SHIFT;
        dstV[i] = (RV * r + GV * g + BV * b + (0x10001 << (RGB2YUV_SHIFT - 1))) >> RGB2YUV_SHIFT;
84 85 86
    }
}

87 88 89 90 91 92
static av_always_inline void rgb48ToUV_half_c_template(uint16_t *dstU,
                                                       uint16_t *dstV,
                                                       const uint16_t *src1,
                                                       const uint16_t *src2,
                                                       int width,
                                                       enum PixelFormat origin)
93 94
{
    int i;
95
    assert(src1 == src2);
96
    for (i = 0; i < width; i++) {
97 98 99 100 101 102
        int r_b = (input_pixel(&src1[6 * i + 0]) +
                   input_pixel(&src1[6 * i + 3]) + 1) >> 1;
        int g   = (input_pixel(&src1[6 * i + 1]) +
                   input_pixel(&src1[6 * i + 4]) + 1) >> 1;
        int b_r = (input_pixel(&src1[6 * i + 2]) +
                   input_pixel(&src1[6 * i + 5]) + 1) >> 1;
103

104 105
        dstU[i] = (RU * r + GU * g + BU * b + (0x10001 << (RGB2YUV_SHIFT - 1))) >> RGB2YUV_SHIFT;
        dstV[i] = (RV * r + GV * g + BV * b + (0x10001 << (RGB2YUV_SHIFT - 1))) >> RGB2YUV_SHIFT;
106 107 108 109 110 111 112
    }
}

#undef r
#undef b
#undef input_pixel

113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149
#define rgb48funcs(pattern, BE_LE, origin)                              \
static void pattern ## 48 ## BE_LE ## ToY_c(uint8_t *_dst,              \
                                            const uint8_t *_src,        \
                                            int width,                  \
                                            uint32_t *unused)           \
{                                                                       \
    const uint16_t *src = (const uint16_t *)_src;                       \
    uint16_t *dst       = (uint16_t *)_dst;                             \
    rgb48ToY_c_template(dst, src, width, origin);                       \
}                                                                       \
                                                                        \
static void pattern ## 48 ## BE_LE ## ToUV_c(uint8_t *_dstU,            \
                                             uint8_t *_dstV,            \
                                             const uint8_t *_src1,      \
                                             const uint8_t *_src2,      \
                                             int width,                 \
                                             uint32_t *unused)          \
{                                                                       \
    const uint16_t *src1 = (const uint16_t *)_src1,                     \
                   *src2 = (const uint16_t *)_src2;                     \
    uint16_t *dstU = (uint16_t *)_dstU,                                 \
             *dstV = (uint16_t *)_dstV;                                 \
    rgb48ToUV_c_template(dstU, dstV, src1, src2, width, origin);        \
}                                                                       \
                                                                        \
static void pattern ## 48 ## BE_LE ## ToUV_half_c(uint8_t *_dstU,       \
                                                  uint8_t *_dstV,       \
                                                  const uint8_t *_src1, \
                                                  const uint8_t *_src2, \
                                                  int width,            \
                                                  uint32_t *unused)     \
{                                                                       \
    const uint16_t *src1 = (const uint16_t *)_src1,                     \
                   *src2 = (const uint16_t *)_src2;                     \
    uint16_t *dstU = (uint16_t *)_dstU,                                 \
             *dstV = (uint16_t *)_dstV;                                 \
    rgb48ToUV_half_c_template(dstU, dstV, src1, src2, width, origin);   \
150 151 152 153 154 155 156
}

rgb48funcs(rgb, LE, PIX_FMT_RGB48LE)
rgb48funcs(rgb, BE, PIX_FMT_RGB48BE)
rgb48funcs(bgr, LE, PIX_FMT_BGR48LE)
rgb48funcs(bgr, BE, PIX_FMT_BGR48BE)

157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175
#define input_pixel(i) ((origin == PIX_FMT_RGBA ||                      \
                         origin == PIX_FMT_BGRA ||                      \
                         origin == PIX_FMT_ARGB ||                      \
                         origin == PIX_FMT_ABGR)                        \
                        ? AV_RN32A(&src[(i) * 4])                       \
                        : (isBE(origin) ? AV_RB16(&src[(i) * 2])        \
                                        : AV_RL16(&src[(i) * 2])))

static av_always_inline void rgb16_32ToY_c_template(uint8_t *dst,
                                                    const uint8_t *src,
                                                    int width,
                                                    enum PixelFormat origin,
                                                    int shr, int shg,
                                                    int shb, int shp,
                                                    int maskr, int maskg,
                                                    int maskb, int rsh,
                                                    int gsh, int bsh, int S)
{
    const int ry       = RY << rsh, gy = GY << gsh, by = BY << bsh;
176 177 178 179 180
    const unsigned rnd = 33u << (S - 1);
    int i;

    for (i = 0; i < width; i++) {
        int px = input_pixel(i) >> shp;
181 182 183
        int b  = (px & maskb) >> shb;
        int g  = (px & maskg) >> shg;
        int r  = (px & maskr) >> shr;
184 185 186 187 188

        dst[i] = (ry * r + gy * g + by * b + rnd) >> S;
    }
}

189 190 191 192 193 194 195 196 197 198
static av_always_inline void rgb16_32ToUV_c_template(uint8_t *dstU,
                                                     uint8_t *dstV,
                                                     const uint8_t *src,
                                                     int width,
                                                     enum PixelFormat origin,
                                                     int shr, int shg,
                                                     int shb, int shp,
                                                     int maskr, int maskg,
                                                     int maskb, int rsh,
                                                     int gsh, int bsh, int S)
199
{
200 201
    const int ru       = RU << rsh, gu = GU << gsh, bu = BU << bsh,
              rv       = RV << rsh, gv = GV << gsh, bv = BV << bsh;
202 203 204 205 206
    const unsigned rnd = 257u << (S - 1);
    int i;

    for (i = 0; i < width; i++) {
        int px = input_pixel(i) >> shp;
207 208 209
        int b  = (px & maskb)   >> shb;
        int g  = (px & maskg)   >> shg;
        int r  = (px & maskr)   >> shr;
210 211 212 213 214 215

        dstU[i] = (ru * r + gu * g + bu * b + rnd) >> S;
        dstV[i] = (rv * r + gv * g + bv * b + rnd) >> S;
    }
}

216 217 218 219 220 221 222 223 224 225
static av_always_inline void rgb16_32ToUV_half_c_template(uint8_t *dstU,
                                                          uint8_t *dstV,
                                                          const uint8_t *src,
                                                          int width,
                                                          enum PixelFormat origin,
                                                          int shr, int shg,
                                                          int shb, int shp,
                                                          int maskr, int maskg,
                                                          int maskb, int rsh,
                                                          int gsh, int bsh, int S)
226
{
227 228 229
    const int ru       = RU << rsh, gu = GU << gsh, bu = BU << bsh,
              rv       = RV << rsh, gv = GV << gsh, bv = BV << bsh,
              maskgx   = ~(maskr | maskb);
230 231 232
    const unsigned rnd = 257u << S;
    int i;

233 234 235
    maskr |= maskr << 1;
    maskb |= maskb << 1;
    maskg |= maskg << 1;
236 237 238 239 240 241 242
    for (i = 0; i < width; i++) {
        int px0 = input_pixel(2 * i + 0) >> shp;
        int px1 = input_pixel(2 * i + 1) >> shp;
        int b, r, g = (px0 & maskgx) + (px1 & maskgx);
        int rb = px0 + px1 - g;

        b = (rb & maskb) >> shb;
243 244
        if (shp ||
            origin == PIX_FMT_BGR565LE || origin == PIX_FMT_BGR565BE ||
245 246 247
            origin == PIX_FMT_RGB565LE || origin == PIX_FMT_RGB565BE) {
            g >>= shg;
        } else {
248
            g = (g & maskg) >> shg;
249 250 251 252 253 254 255 256 257 258
        }
        r = (rb & maskr) >> shr;

        dstU[i] = (ru * r + gu * g + bu * b + rnd) >> (S + 1);
        dstV[i] = (rv * r + gv * g + bv * b + rnd) >> (S + 1);
    }
}

#undef input_pixel

259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306
#define rgb16_32_wrapper(fmt, name, shr, shg, shb, shp, maskr,          \
                         maskg, maskb, rsh, gsh, bsh, S)                \
static void name ## ToY_c(uint8_t *dst, const uint8_t *src,             \
                          int width, uint32_t *unused)                  \
{                                                                       \
    rgb16_32ToY_c_template(dst, src, width, fmt, shr, shg, shb, shp,    \
                           maskr, maskg, maskb, rsh, gsh, bsh, S);      \
}                                                                       \
                                                                        \
static void name ## ToUV_c(uint8_t *dstU, uint8_t *dstV,                \
                           const uint8_t *src, const uint8_t *dummy,    \
                           int width, uint32_t *unused)                 \
{                                                                       \
    rgb16_32ToUV_c_template(dstU, dstV, src, width, fmt,                \
                            shr, shg, shb, shp,                         \
                            maskr, maskg, maskb, rsh, gsh, bsh, S);     \
}                                                                       \
                                                                        \
static void name ## ToUV_half_c(uint8_t *dstU, uint8_t *dstV,           \
                                const uint8_t *src,                     \
                                const uint8_t *dummy,                   \
                                int width, uint32_t *unused)            \
{                                                                       \
    rgb16_32ToUV_half_c_template(dstU, dstV, src, width, fmt,           \
                                 shr, shg, shb, shp,                    \
                                 maskr, maskg, maskb,                   \
                                 rsh, gsh, bsh, S);                     \
}

rgb16_32_wrapper(PIX_FMT_BGR32,    bgr32,  16, 0,  0, 0, 0xFF0000, 0xFF00,   0x00FF,  8, 0,  8, RGB2YUV_SHIFT + 8)
rgb16_32_wrapper(PIX_FMT_BGR32_1,  bgr321, 16, 0,  0, 8, 0xFF0000, 0xFF00,   0x00FF,  8, 0,  8, RGB2YUV_SHIFT + 8)
rgb16_32_wrapper(PIX_FMT_RGB32,    rgb32,   0, 0, 16, 0,   0x00FF, 0xFF00, 0xFF0000,  8, 0,  8, RGB2YUV_SHIFT + 8)
rgb16_32_wrapper(PIX_FMT_RGB32_1,  rgb321,  0, 0, 16, 8,   0x00FF, 0xFF00, 0xFF0000,  8, 0,  8, RGB2YUV_SHIFT + 8)
rgb16_32_wrapper(PIX_FMT_BGR565LE, bgr16le, 0, 0,  0, 0,   0x001F, 0x07E0,   0xF800, 11, 5,  0, RGB2YUV_SHIFT + 8)
rgb16_32_wrapper(PIX_FMT_BGR555LE, bgr15le, 0, 0,  0, 0,   0x001F, 0x03E0,   0x7C00, 10, 5,  0, RGB2YUV_SHIFT + 7)
rgb16_32_wrapper(PIX_FMT_BGR444LE, bgr12le, 0, 0,  0, 0,   0x000F, 0x00F0,   0x0F00,  8, 4,  0, RGB2YUV_SHIFT + 4)
rgb16_32_wrapper(PIX_FMT_RGB565LE, rgb16le, 0, 0,  0, 0,   0xF800, 0x07E0,   0x001F,  0, 5, 11, RGB2YUV_SHIFT + 8)
rgb16_32_wrapper(PIX_FMT_RGB555LE, rgb15le, 0, 0,  0, 0,   0x7C00, 0x03E0,   0x001F,  0, 5, 10, RGB2YUV_SHIFT + 7)
rgb16_32_wrapper(PIX_FMT_RGB444LE, rgb12le, 0, 0,  0, 0,   0x0F00, 0x00F0,   0x000F,  0, 4,  8, RGB2YUV_SHIFT + 4)
rgb16_32_wrapper(PIX_FMT_BGR565BE, bgr16be, 0, 0,  0, 0,   0x001F, 0x07E0,   0xF800, 11, 5,  0, RGB2YUV_SHIFT + 8)
rgb16_32_wrapper(PIX_FMT_BGR555BE, bgr15be, 0, 0,  0, 0,   0x001F, 0x03E0,   0x7C00, 10, 5,  0, RGB2YUV_SHIFT + 7)
rgb16_32_wrapper(PIX_FMT_BGR444BE, bgr12be, 0, 0,  0, 0,   0x000F, 0x00F0,   0x0F00,  8, 4,  0, RGB2YUV_SHIFT + 4)
rgb16_32_wrapper(PIX_FMT_RGB565BE, rgb16be, 0, 0,  0, 0,   0xF800, 0x07E0,   0x001F,  0, 5, 11, RGB2YUV_SHIFT + 8)
rgb16_32_wrapper(PIX_FMT_RGB555BE, rgb15be, 0, 0,  0, 0,   0x7C00, 0x03E0,   0x001F,  0, 5, 10, RGB2YUV_SHIFT + 7)
rgb16_32_wrapper(PIX_FMT_RGB444BE, rgb12be, 0, 0,  0, 0,   0x0F00, 0x00F0,   0x000F,  0, 4,  8, RGB2YUV_SHIFT + 4)

static void abgrToA_c(uint8_t *dst, const uint8_t *src, int width,
                      uint32_t *unused)
307 308
{
    int i;
309 310
    for (i = 0; i < width; i++)
        dst[i] = src[4 * i];
311 312
}

313 314
static void rgbaToA_c(uint8_t *dst, const uint8_t *src, int width,
                      uint32_t *unused)
315 316
{
    int i;
317 318
    for (i = 0; i < width; i++)
        dst[i] = src[4 * i + 3];
319 320 321 322 323
}

static void palToY_c(uint8_t *dst, const uint8_t *src, int width, uint32_t *pal)
{
    int i;
324 325
    for (i = 0; i < width; i++) {
        int d = src[i];
326

327
        dst[i] = pal[d] & 0xFF;
328 329 330 331 332 333 334 335 336
    }
}

static void palToUV_c(uint8_t *dstU, uint8_t *dstV,
                      const uint8_t *src1, const uint8_t *src2,
                      int width, uint32_t *pal)
{
    int i;
    assert(src1 == src2);
337 338
    for (i = 0; i < width; i++) {
        int p = pal[src1[i]];
339

340 341
        dstU[i] = p >> 8;
        dstV[i] = p >> 16;
342 343 344 345 346 347 348
    }
}

static void monowhite2Y_c(uint8_t *dst, const uint8_t *src,
                          int width, uint32_t *unused)
{
    int i, j;
349 350
    width = (width + 7) >> 3;
    for (i = 0; i < width; i++) {
351 352 353
        int d = ~src[i];
        for (j = 0; j < 8; j++)
            dst[8 * i + j] = ((d >> (7 - j)) & 1) * 255;
354 355 356 357 358 359 360
    }
}

static void monoblack2Y_c(uint8_t *dst, const uint8_t *src,
                          int width, uint32_t *unused)
{
    int i, j;
361 362
    width = (width + 7) >> 3;
    for (i = 0; i < width; i++) {
363 364 365
        int d = src[i];
        for (j = 0; j < 8; j++)
            dst[8 * i + j] = ((d >> (7 - j)) & 1) * 255;
366 367 368 369 370 371 372
    }
}

static void yuy2ToY_c(uint8_t *dst, const uint8_t *src, int width,
                      uint32_t *unused)
{
    int i;
373 374
    for (i = 0; i < width; i++)
        dst[i] = src[2 * i];
375 376 377 378 379 380
}

static void yuy2ToUV_c(uint8_t *dstU, uint8_t *dstV, const uint8_t *src1,
                       const uint8_t *src2, int width, uint32_t *unused)
{
    int i;
381 382 383
    for (i = 0; i < width; i++) {
        dstU[i] = src1[4 * i + 1];
        dstV[i] = src1[4 * i + 3];
384 385 386 387
    }
    assert(src1 == src2);
}

388 389
static void bswap16Y_c(uint8_t *_dst, const uint8_t *_src, int width,
                       uint32_t *unused)
390 391
{
    int i;
392 393 394
    const uint16_t *src = (const uint16_t *)_src;
    uint16_t *dst       = (uint16_t *)_dst;
    for (i = 0; i < width; i++)
395 396 397 398 399 400 401
        dst[i] = av_bswap16(src[i]);
}

static void bswap16UV_c(uint8_t *_dstU, uint8_t *_dstV, const uint8_t *_src1,
                        const uint8_t *_src2, int width, uint32_t *unused)
{
    int i;
402 403 404 405
    const uint16_t *src1 = (const uint16_t *)_src1,
    *src2                = (const uint16_t *)_src2;
    uint16_t *dstU       = (uint16_t *)_dstU, *dstV = (uint16_t *)_dstV;
    for (i = 0; i < width; i++) {
406 407 408 409 410 411
        dstU[i] = av_bswap16(src1[i]);
        dstV[i] = av_bswap16(src2[i]);
    }
}

/* This is almost identical to the previous, end exists only because
412
 * yuy2ToY/UV)(dst, src + 1, ...) would have 100% unaligned accesses. */
413 414 415 416
static void uyvyToY_c(uint8_t *dst, const uint8_t *src, int width,
                      uint32_t *unused)
{
    int i;
417 418
    for (i = 0; i < width; i++)
        dst[i] = src[2 * i + 1];
419 420 421 422 423 424
}

static void uyvyToUV_c(uint8_t *dstU, uint8_t *dstV, const uint8_t *src1,
                       const uint8_t *src2, int width, uint32_t *unused)
{
    int i;
425 426 427
    for (i = 0; i < width; i++) {
        dstU[i] = src1[4 * i + 0];
        dstV[i] = src1[4 * i + 2];
428 429 430 431 432 433 434 435 436
    }
    assert(src1 == src2);
}

static av_always_inline void nvXXtoUV_c(uint8_t *dst1, uint8_t *dst2,
                                        const uint8_t *src, int width)
{
    int i;
    for (i = 0; i < width; i++) {
437 438
        dst1[i] = src[2 * i + 0];
        dst2[i] = src[2 * i + 1];
439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461
    }
}

static void nv12ToUV_c(uint8_t *dstU, uint8_t *dstV,
                       const uint8_t *src1, const uint8_t *src2,
                       int width, uint32_t *unused)
{
    nvXXtoUV_c(dstU, dstV, src1, width);
}

static void nv21ToUV_c(uint8_t *dstU, uint8_t *dstV,
                       const uint8_t *src1, const uint8_t *src2,
                       int width, uint32_t *unused)
{
    nvXXtoUV_c(dstV, dstU, src1, width);
}

#define input_pixel(pos) (isBE(origin) ? AV_RB16(pos) : AV_RL16(pos))

static void bgr24ToY_c(uint8_t *dst, const uint8_t *src,
                       int width, uint32_t *unused)
{
    int i;
462 463 464 465
    for (i = 0; i < width; i++) {
        int b = src[i * 3 + 0];
        int g = src[i * 3 + 1];
        int r = src[i * 3 + 2];
466

467
        dst[i] = ((RY * r + GY * g + BY * b + (33 << (RGB2YUV_SHIFT - 1))) >> RGB2YUV_SHIFT);
468 469 470 471 472 473 474
    }
}

static void bgr24ToUV_c(uint8_t *dstU, uint8_t *dstV, const uint8_t *src1,
                        const uint8_t *src2, int width, uint32_t *unused)
{
    int i;
475 476 477 478
    for (i = 0; i < width; i++) {
        int b = src1[3 * i + 0];
        int g = src1[3 * i + 1];
        int r = src1[3 * i + 2];
479

480 481
        dstU[i] = (RU * r + GU * g + BU * b + (257 << (RGB2YUV_SHIFT - 1))) >> RGB2YUV_SHIFT;
        dstV[i] = (RV * r + GV * g + BV * b + (257 << (RGB2YUV_SHIFT - 1))) >> RGB2YUV_SHIFT;
482 483 484 485 486 487 488 489
    }
    assert(src1 == src2);
}

static void bgr24ToUV_half_c(uint8_t *dstU, uint8_t *dstV, const uint8_t *src1,
                             const uint8_t *src2, int width, uint32_t *unused)
{
    int i;
490 491 492 493
    for (i = 0; i < width; i++) {
        int b = src1[6 * i + 0] + src1[6 * i + 3];
        int g = src1[6 * i + 1] + src1[6 * i + 4];
        int r = src1[6 * i + 2] + src1[6 * i + 5];
494

495 496
        dstU[i] = (RU * r + GU * g + BU * b + (257 << RGB2YUV_SHIFT)) >> (RGB2YUV_SHIFT + 1);
        dstV[i] = (RV * r + GV * g + BV * b + (257 << RGB2YUV_SHIFT)) >> (RGB2YUV_SHIFT + 1);
497 498 499 500 501 502 503 504
    }
    assert(src1 == src2);
}

static void rgb24ToY_c(uint8_t *dst, const uint8_t *src, int width,
                       uint32_t *unused)
{
    int i;
505 506 507 508
    for (i = 0; i < width; i++) {
        int r = src[i * 3 + 0];
        int g = src[i * 3 + 1];
        int b = src[i * 3 + 2];
509

510
        dst[i] = ((RY * r + GY * g + BY * b + (33 << (RGB2YUV_SHIFT - 1))) >> RGB2YUV_SHIFT);
511 512 513 514 515 516 517
    }
}

static void rgb24ToUV_c(uint8_t *dstU, uint8_t *dstV, const uint8_t *src1,
                        const uint8_t *src2, int width, uint32_t *unused)
{
    int i;
518 519 520 521 522 523 524 525
    assert(src1 == src2);
    for (i = 0; i < width; i++) {
        int r = src1[3 * i + 0];
        int g = src1[3 * i + 1];
        int b = src1[3 * i + 2];

        dstU[i] = (RU * r + GU * g + BU * b + (257 << (RGB2YUV_SHIFT - 1))) >> RGB2YUV_SHIFT;
        dstV[i] = (RV * r + GV * g + BV * b + (257 << (RGB2YUV_SHIFT - 1))) >> RGB2YUV_SHIFT;
526 527 528 529 530 531 532
    }
}

static void rgb24ToUV_half_c(uint8_t *dstU, uint8_t *dstV, const uint8_t *src1,
                             const uint8_t *src2, int width, uint32_t *unused)
{
    int i;
533 534 535 536 537 538 539 540
    assert(src1 == src2);
    for (i = 0; i < width; i++) {
        int r = src1[6 * i + 0] + src1[6 * i + 3];
        int g = src1[6 * i + 1] + src1[6 * i + 4];
        int b = src1[6 * i + 2] + src1[6 * i + 5];

        dstU[i] = (RU * r + GU * g + BU * b + (257 << RGB2YUV_SHIFT)) >> (RGB2YUV_SHIFT + 1);
        dstV[i] = (RV * r + GV * g + BV * b + (257 << RGB2YUV_SHIFT)) >> (RGB2YUV_SHIFT + 1);
541 542 543 544 545 546 547 548 549 550 551 552 553 554 555
    }
}

static void planar_rgb_to_y(uint8_t *dst, const uint8_t *src[4], int width)
{
    int i;
    for (i = 0; i < width; i++) {
        int g = src[0][i];
        int b = src[1][i];
        int r = src[2][i];

        dst[i] = ((RY * r + GY * g + BY * b + (33 << (RGB2YUV_SHIFT - 1))) >> RGB2YUV_SHIFT);
    }
}

556
static void planar_rgb_to_uv(uint8_t *dstU, uint8_t *dstV, const uint8_t *src[4], int width)
557 558 559
{
    int i;
    for (i = 0; i < width; i++) {
560 561 562
        int g = src[0][i];
        int b = src[1][i];
        int r = src[2][i];
563

564 565
        dstU[i] = (RU * r + GU * g + BU * b + (257 << (RGB2YUV_SHIFT - 1))) >> RGB2YUV_SHIFT;
        dstV[i] = (RV * r + GV * g + BV * b + (257 << (RGB2YUV_SHIFT - 1))) >> RGB2YUV_SHIFT;
566 567 568
    }
}

569 570 571 572
#define rdpx(src) \
    is_be ? AV_RB16(src) : AV_RL16(src)
static av_always_inline void planar_rgb16_to_y(uint8_t *_dst, const uint8_t *_src[4],
                                               int width, int bpc, int is_be)
573 574
{
    int i;
575 576
    const uint16_t **src = (const uint16_t **)_src;
    uint16_t *dst        = (uint16_t *)_dst;
577
    for (i = 0; i < width; i++) {
578 579 580
        int g = rdpx(src[0] + i);
        int b = rdpx(src[1] + i);
        int r = rdpx(src[2] + i);
581

582
        dst[i] = ((RY * r + GY * g + BY * b + (33 << (RGB2YUV_SHIFT + bpc - 9))) >> RGB2YUV_SHIFT);
583 584 585
    }
}

586
static void planar_rgb9le_to_y(uint8_t *dst, const uint8_t *src[4], int w)
587
{
588 589
    planar_rgb16_to_y(dst, src, w, 9, 0);
}
590

591 592 593
static void planar_rgb9be_to_y(uint8_t *dst, const uint8_t *src[4], int w)
{
    planar_rgb16_to_y(dst, src, w, 9, 1);
594 595
}

596
static void planar_rgb10le_to_y(uint8_t *dst, const uint8_t *src[4], int w)
597
{
598 599
    planar_rgb16_to_y(dst, src, w, 10, 0);
}
600

601 602 603
static void planar_rgb10be_to_y(uint8_t *dst, const uint8_t *src[4], int w)
{
    planar_rgb16_to_y(dst, src, w, 10, 1);
604 605
}

606 607 608 609 610 611 612 613 614 615 616 617 618
static void planar_rgb16le_to_y(uint8_t *dst, const uint8_t *src[4], int w)
{
    planar_rgb16_to_y(dst, src, w, 16, 0);
}

static void planar_rgb16be_to_y(uint8_t *dst, const uint8_t *src[4], int w)
{
    planar_rgb16_to_y(dst, src, w, 16, 1);
}

static av_always_inline void planar_rgb16_to_uv(uint8_t *_dstU, uint8_t *_dstV,
                                                const uint8_t *_src[4], int width,
                                                int bpc, int is_be)
619 620
{
    int i;
621 622 623
    const uint16_t **src = (const uint16_t **)_src;
    uint16_t *dstU       = (uint16_t *)_dstU;
    uint16_t *dstV       = (uint16_t *)_dstV;
624
    for (i = 0; i < width; i++) {
625 626 627
        int g = rdpx(src[0] + i);
        int b = rdpx(src[1] + i);
        int r = rdpx(src[2] + i);
628

629 630
        dstU[i] = (RU * r + GU * g + BU * b + (257 << (RGB2YUV_SHIFT + bpc - 9))) >> RGB2YUV_SHIFT;
        dstV[i] = (RV * r + GV * g + BV * b + (257 << (RGB2YUV_SHIFT + bpc - 9))) >> RGB2YUV_SHIFT;
631 632
    }
}
633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669
#undef rdpx

static void planar_rgb9le_to_uv(uint8_t *dstU, uint8_t *dstV,
                                const uint8_t *src[4], int w)
{
    planar_rgb16_to_uv(dstU, dstV, src, w, 9, 0);
}

static void planar_rgb9be_to_uv(uint8_t *dstU, uint8_t *dstV,
                                const uint8_t *src[4], int w)
{
    planar_rgb16_to_uv(dstU, dstV, src, w, 9, 1);
}

static void planar_rgb10le_to_uv(uint8_t *dstU, uint8_t *dstV,
                                 const uint8_t *src[4], int w)
{
    planar_rgb16_to_uv(dstU, dstV, src, w, 10, 0);
}

static void planar_rgb10be_to_uv(uint8_t *dstU, uint8_t *dstV,
                                 const uint8_t *src[4], int w)
{
    planar_rgb16_to_uv(dstU, dstV, src, w, 10, 1);
}

static void planar_rgb16le_to_uv(uint8_t *dstU, uint8_t *dstV,
                                 const uint8_t *src[4], int w)
{
    planar_rgb16_to_uv(dstU, dstV, src, w, 16, 0);
}

static void planar_rgb16be_to_uv(uint8_t *dstU, uint8_t *dstV,
                                 const uint8_t *src[4], int w)
{
    planar_rgb16_to_uv(dstU, dstV, src, w, 16, 1);
}
670 671 672 673 674 675

av_cold void ff_sws_init_input_funcs(SwsContext *c)
{
    enum PixelFormat srcFormat = c->srcFormat;

    c->chrToYV12 = NULL;
676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696
    switch (srcFormat) {
    case PIX_FMT_YUYV422:
        c->chrToYV12 = yuy2ToUV_c;
        break;
    case PIX_FMT_UYVY422:
        c->chrToYV12 = uyvyToUV_c;
        break;
    case PIX_FMT_NV12:
        c->chrToYV12 = nv12ToUV_c;
        break;
    case PIX_FMT_NV21:
        c->chrToYV12 = nv21ToUV_c;
        break;
    case PIX_FMT_RGB8:
    case PIX_FMT_BGR8:
    case PIX_FMT_PAL8:
    case PIX_FMT_BGR4_BYTE:
    case PIX_FMT_RGB4_BYTE:
        c->chrToYV12 = palToUV_c;
        break;
    case PIX_FMT_GBRP9LE:
697 698
        c->readChrPlanar = planar_rgb9le_to_uv;
        break;
699
    case PIX_FMT_GBRP10LE:
700 701
        c->readChrPlanar = planar_rgb10le_to_uv;
        break;
702 703 704 705
    case PIX_FMT_GBRP16LE:
        c->readChrPlanar = planar_rgb16le_to_uv;
        break;
    case PIX_FMT_GBRP9BE:
706 707
        c->readChrPlanar = planar_rgb9be_to_uv;
        break;
708
    case PIX_FMT_GBRP10BE:
709 710
        c->readChrPlanar = planar_rgb10be_to_uv;
        break;
711 712 713 714 715 716
    case PIX_FMT_GBRP16BE:
        c->readChrPlanar = planar_rgb16be_to_uv;
        break;
    case PIX_FMT_GBRP:
        c->readChrPlanar = planar_rgb_to_uv;
        break;
717
#if HAVE_BIGENDIAN
718 719 720 721 722 723 724 725 726 727 728
    case PIX_FMT_YUV444P9LE:
    case PIX_FMT_YUV422P9LE:
    case PIX_FMT_YUV420P9LE:
    case PIX_FMT_YUV422P10LE:
    case PIX_FMT_YUV444P10LE:
    case PIX_FMT_YUV420P10LE:
    case PIX_FMT_YUV420P16LE:
    case PIX_FMT_YUV422P16LE:
    case PIX_FMT_YUV444P16LE:
        c->chrToYV12 = bswap16UV_c;
        break;
729
#else
730 731 732 733 734 735 736 737 738 739 740
    case PIX_FMT_YUV444P9BE:
    case PIX_FMT_YUV422P9BE:
    case PIX_FMT_YUV420P9BE:
    case PIX_FMT_YUV444P10BE:
    case PIX_FMT_YUV422P10BE:
    case PIX_FMT_YUV420P10BE:
    case PIX_FMT_YUV420P16BE:
    case PIX_FMT_YUV422P16BE:
    case PIX_FMT_YUV444P16BE:
        c->chrToYV12 = bswap16UV_c;
        break;
741 742 743
#endif
    }
    if (c->chrSrcHSubSample) {
744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810
        switch (srcFormat) {
        case PIX_FMT_RGB48BE:
            c->chrToYV12 = rgb48BEToUV_half_c;
            break;
        case PIX_FMT_RGB48LE:
            c->chrToYV12 = rgb48LEToUV_half_c;
            break;
        case PIX_FMT_BGR48BE:
            c->chrToYV12 = bgr48BEToUV_half_c;
            break;
        case PIX_FMT_BGR48LE:
            c->chrToYV12 = bgr48LEToUV_half_c;
            break;
        case PIX_FMT_RGB32:
            c->chrToYV12 = bgr32ToUV_half_c;
            break;
        case PIX_FMT_RGB32_1:
            c->chrToYV12 = bgr321ToUV_half_c;
            break;
        case PIX_FMT_BGR24:
            c->chrToYV12 = bgr24ToUV_half_c;
            break;
        case PIX_FMT_BGR565LE:
            c->chrToYV12 = bgr16leToUV_half_c;
            break;
        case PIX_FMT_BGR565BE:
            c->chrToYV12 = bgr16beToUV_half_c;
            break;
        case PIX_FMT_BGR555LE:
            c->chrToYV12 = bgr15leToUV_half_c;
            break;
        case PIX_FMT_BGR555BE:
            c->chrToYV12 = bgr15beToUV_half_c;
            break;
        case PIX_FMT_BGR444LE:
            c->chrToYV12 = bgr12leToUV_half_c;
            break;
        case PIX_FMT_BGR444BE:
            c->chrToYV12 = bgr12beToUV_half_c;
            break;
        case PIX_FMT_BGR32:
            c->chrToYV12 = rgb32ToUV_half_c;
            break;
        case PIX_FMT_BGR32_1:
            c->chrToYV12 = rgb321ToUV_half_c;
            break;
        case PIX_FMT_RGB24:
            c->chrToYV12 = rgb24ToUV_half_c;
            break;
        case PIX_FMT_RGB565LE:
            c->chrToYV12 = rgb16leToUV_half_c;
            break;
        case PIX_FMT_RGB565BE:
            c->chrToYV12 = rgb16beToUV_half_c;
            break;
        case PIX_FMT_RGB555LE:
            c->chrToYV12 = rgb15leToUV_half_c;
            break;
        case PIX_FMT_RGB555BE:
            c->chrToYV12 = rgb15beToUV_half_c;
            break;
        case PIX_FMT_RGB444LE:
            c->chrToYV12 = rgb12leToUV_half_c;
            break;
        case PIX_FMT_RGB444BE:
            c->chrToYV12 = rgb12beToUV_half_c;
            break;
811 812
        }
    } else {
813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879
        switch (srcFormat) {
        case PIX_FMT_RGB48BE:
            c->chrToYV12 = rgb48BEToUV_c;
            break;
        case PIX_FMT_RGB48LE:
            c->chrToYV12 = rgb48LEToUV_c;
            break;
        case PIX_FMT_BGR48BE:
            c->chrToYV12 = bgr48BEToUV_c;
            break;
        case PIX_FMT_BGR48LE:
            c->chrToYV12 = bgr48LEToUV_c;
            break;
        case PIX_FMT_RGB32:
            c->chrToYV12 = bgr32ToUV_c;
            break;
        case PIX_FMT_RGB32_1:
            c->chrToYV12 = bgr321ToUV_c;
            break;
        case PIX_FMT_BGR24:
            c->chrToYV12 = bgr24ToUV_c;
            break;
        case PIX_FMT_BGR565LE:
            c->chrToYV12 = bgr16leToUV_c;
            break;
        case PIX_FMT_BGR565BE:
            c->chrToYV12 = bgr16beToUV_c;
            break;
        case PIX_FMT_BGR555LE:
            c->chrToYV12 = bgr15leToUV_c;
            break;
        case PIX_FMT_BGR555BE:
            c->chrToYV12 = bgr15beToUV_c;
            break;
        case PIX_FMT_BGR444LE:
            c->chrToYV12 = bgr12leToUV_c;
            break;
        case PIX_FMT_BGR444BE:
            c->chrToYV12 = bgr12beToUV_c;
            break;
        case PIX_FMT_BGR32:
            c->chrToYV12 = rgb32ToUV_c;
            break;
        case PIX_FMT_BGR32_1:
            c->chrToYV12 = rgb321ToUV_c;
            break;
        case PIX_FMT_RGB24:
            c->chrToYV12 = rgb24ToUV_c;
            break;
        case PIX_FMT_RGB565LE:
            c->chrToYV12 = rgb16leToUV_c;
            break;
        case PIX_FMT_RGB565BE:
            c->chrToYV12 = rgb16beToUV_c;
            break;
        case PIX_FMT_RGB555LE:
            c->chrToYV12 = rgb15leToUV_c;
            break;
        case PIX_FMT_RGB555BE:
            c->chrToYV12 = rgb15beToUV_c;
            break;
        case PIX_FMT_RGB444LE:
            c->chrToYV12 = rgb12leToUV_c;
            break;
        case PIX_FMT_RGB444BE:
            c->chrToYV12 = rgb12beToUV_c;
            break;
880 881 882 883 884 885 886
        }
    }

    c->lumToYV12 = NULL;
    c->alpToYV12 = NULL;
    switch (srcFormat) {
    case PIX_FMT_GBRP9LE:
887 888
        c->readLumPlanar = planar_rgb9le_to_y;
        break;
889
    case PIX_FMT_GBRP10LE:
890 891
        c->readLumPlanar = planar_rgb10le_to_y;
        break;
892 893 894
    case PIX_FMT_GBRP16LE:
        c->readLumPlanar = planar_rgb16le_to_y;
        break;
895
    case PIX_FMT_GBRP9BE:
896 897
        c->readLumPlanar = planar_rgb9be_to_y;
        break;
898
    case PIX_FMT_GBRP10BE:
899 900
        c->readLumPlanar = planar_rgb10be_to_y;
        break;
901 902 903 904 905 906
    case PIX_FMT_GBRP16BE:
        c->readLumPlanar = planar_rgb16be_to_y;
        break;
    case PIX_FMT_GBRP:
        c->readLumPlanar = planar_rgb_to_y;
        break;
907 908 909 910 911 912 913 914 915 916
#if HAVE_BIGENDIAN
    case PIX_FMT_YUV444P9LE:
    case PIX_FMT_YUV422P9LE:
    case PIX_FMT_YUV420P9LE:
    case PIX_FMT_YUV444P10LE:
    case PIX_FMT_YUV422P10LE:
    case PIX_FMT_YUV420P10LE:
    case PIX_FMT_YUV420P16LE:
    case PIX_FMT_YUV422P16LE:
    case PIX_FMT_YUV444P16LE:
917 918 919
    case PIX_FMT_GRAY16LE:
        c->lumToYV12 = bswap16Y_c;
        break;
920 921 922 923 924 925 926 927 928 929
#else
    case PIX_FMT_YUV444P9BE:
    case PIX_FMT_YUV422P9BE:
    case PIX_FMT_YUV420P9BE:
    case PIX_FMT_YUV444P10BE:
    case PIX_FMT_YUV422P10BE:
    case PIX_FMT_YUV420P10BE:
    case PIX_FMT_YUV420P16BE:
    case PIX_FMT_YUV422P16BE:
    case PIX_FMT_YUV444P16BE:
930 931 932
    case PIX_FMT_GRAY16BE:
        c->lumToYV12 = bswap16Y_c;
        break;
933
#endif
934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985
    case PIX_FMT_YUYV422:
    case PIX_FMT_Y400A:
        c->lumToYV12 = yuy2ToY_c;
        break;
    case PIX_FMT_UYVY422:
        c->lumToYV12 = uyvyToY_c;
        break;
    case PIX_FMT_BGR24:
        c->lumToYV12 = bgr24ToY_c;
        break;
    case PIX_FMT_BGR565LE:
        c->lumToYV12 = bgr16leToY_c;
        break;
    case PIX_FMT_BGR565BE:
        c->lumToYV12 = bgr16beToY_c;
        break;
    case PIX_FMT_BGR555LE:
        c->lumToYV12 = bgr15leToY_c;
        break;
    case PIX_FMT_BGR555BE:
        c->lumToYV12 = bgr15beToY_c;
        break;
    case PIX_FMT_BGR444LE:
        c->lumToYV12 = bgr12leToY_c;
        break;
    case PIX_FMT_BGR444BE:
        c->lumToYV12 = bgr12beToY_c;
        break;
    case PIX_FMT_RGB24:
        c->lumToYV12 = rgb24ToY_c;
        break;
    case PIX_FMT_RGB565LE:
        c->lumToYV12 = rgb16leToY_c;
        break;
    case PIX_FMT_RGB565BE:
        c->lumToYV12 = rgb16beToY_c;
        break;
    case PIX_FMT_RGB555LE:
        c->lumToYV12 = rgb15leToY_c;
        break;
    case PIX_FMT_RGB555BE:
        c->lumToYV12 = rgb15beToY_c;
        break;
    case PIX_FMT_RGB444LE:
        c->lumToYV12 = rgb12leToY_c;
        break;
    case PIX_FMT_RGB444BE:
        c->lumToYV12 = rgb12beToY_c;
        break;
    case PIX_FMT_RGB8:
    case PIX_FMT_BGR8:
    case PIX_FMT_PAL8:
986
    case PIX_FMT_BGR4_BYTE:
987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019
    case PIX_FMT_RGB4_BYTE:
        c->lumToYV12 = palToY_c;
        break;
    case PIX_FMT_MONOBLACK:
        c->lumToYV12 = monoblack2Y_c;
        break;
    case PIX_FMT_MONOWHITE:
        c->lumToYV12 = monowhite2Y_c;
        break;
    case PIX_FMT_RGB32:
        c->lumToYV12 = bgr32ToY_c;
        break;
    case PIX_FMT_RGB32_1:
        c->lumToYV12 = bgr321ToY_c;
        break;
    case PIX_FMT_BGR32:
        c->lumToYV12 = rgb32ToY_c;
        break;
    case PIX_FMT_BGR32_1:
        c->lumToYV12 = rgb321ToY_c;
        break;
    case PIX_FMT_RGB48BE:
        c->lumToYV12 = rgb48BEToY_c;
        break;
    case PIX_FMT_RGB48LE:
        c->lumToYV12 = rgb48LEToY_c;
        break;
    case PIX_FMT_BGR48BE:
        c->lumToYV12 = bgr48BEToY_c;
        break;
    case PIX_FMT_BGR48LE:
        c->lumToYV12 = bgr48LEToY_c;
        break;
1020 1021 1022 1023
    }
    if (c->alpPixBuf) {
        switch (srcFormat) {
        case PIX_FMT_BGRA:
1024 1025 1026
        case PIX_FMT_RGBA:
            c->alpToYV12 = rgbaToA_c;
            break;
1027
        case PIX_FMT_ABGR:
1028 1029 1030 1031 1032 1033
        case PIX_FMT_ARGB:
            c->alpToYV12 = abgrToA_c;
            break;
        case PIX_FMT_Y400A:
            c->alpToYV12 = uyvyToY_c;
            break;
1034 1035 1036
        }
    }
}