falkTX
/
FFmpeg
mirror of https://github.com/falkTX/FFmpeg.git
1
0
Fork 0
Code Issues 0 Releases 338 Wiki Activity
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
92515 Commits
32 Branches
844MB
Tree: 86e6f0dbc7
Branches Tags
${ item.name }
Create branch ${ searchTerm }
from '86e6f0dbc7'
${ noResults }
FFmpeg/libswscale/utils.c

2429 lines
86KB
Raw Blame History 

  1. /*

  2.  * Copyright (C) 2001-2003 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.  * modify it under the terms of the GNU Lesser General Public

  8.  * License as published by the Free Software Foundation; either

  9.  * version 2.1 of the License, or (at your option) any later version.

  10.  *

  11.  * FFmpeg is distributed in the hope that it will be useful,

  12.  * but WITHOUT ANY WARRANTY; without even the implied warranty of

  13.  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU

  14.  * Lesser General Public License for more details.

  15.  *

  16.  * 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.  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA

  19.  */

  20. 

  21. #include "config.h"

  22. 

  23. #define _DEFAULT_SOURCE

  24. #define _SVID_SOURCE // needed for MAP_ANONYMOUS

  25. #define _DARWIN_C_SOURCE // needed for MAP_ANON

  26. #include <inttypes.h>

  27. #include <math.h>

  28. #include <stdio.h>

  29. #include <string.h>

  30. #if HAVE_MMAP

  31. #include <sys/mman.h>

  32. #if defined(MAP_ANON) && !defined(MAP_ANONYMOUS)

  33. #define MAP_ANONYMOUS MAP_ANON

  34. #endif

  35. #endif

  36. #if HAVE_VIRTUALALLOC

  37. #define WIN32_LEAN_AND_MEAN

  38. #include <windows.h>

  39. #endif

  40. 

  41. #include "libavutil/attributes.h"

  42. #include "libavutil/avassert.h"

  43. #include "libavutil/avutil.h"

  44. #include "libavutil/bswap.h"

  45. #include "libavutil/cpu.h"

  46. #include "libavutil/imgutils.h"

  47. #include "libavutil/intreadwrite.h"

  48. #include "libavutil/libm.h"

  49. #include "libavutil/mathematics.h"

  50. #include "libavutil/opt.h"

  51. #include "libavutil/pixdesc.h"

  52. #include "libavutil/aarch64/cpu.h"

  53. #include "libavutil/ppc/cpu.h"

  54. #include "libavutil/x86/asm.h"

  55. #include "libavutil/x86/cpu.h"

  56. 

  57. // We have to implement deprecated functions until they are removed, this is the

  58. // simplest way to prevent warnings

  59. #undef attribute_deprecated

  60. #define attribute_deprecated

  61. 

  62. #include "rgb2rgb.h"

  63. #include "swscale.h"

  64. #include "swscale_internal.h"

  65. 

  66. #if !FF_API_SWS_VECTOR

  67. static SwsVector *sws_getIdentityVec(void);

  68. static void sws_addVec(SwsVector *a, SwsVector *b);

  69. static void sws_shiftVec(SwsVector *a, int shift);

  70. static void sws_printVec2(SwsVector *a, AVClass *log_ctx, int log_level);

  71. #endif

  72. 

  73. static void handle_formats(SwsContext *c);

  74. 

  75. unsigned swscale_version(void)

  76. {

  77.     av_assert0(LIBSWSCALE_VERSION_MICRO >= 100);

  78.     return LIBSWSCALE_VERSION_INT;

  79. }

  80. 

  81. const char *swscale_configuration(void)

  82. {

  83.     return FFMPEG_CONFIGURATION;

  84. }

  85. 

  86. const char *swscale_license(void)

  87. {

  88. #define LICENSE_PREFIX "libswscale license: "

  89.     return LICENSE_PREFIX FFMPEG_LICENSE + sizeof(LICENSE_PREFIX) - 1;

  90. }

  91. 

  92. typedef struct FormatEntry {

  93.     uint8_t is_supported_in         :1;

  94.     uint8_t is_supported_out        :1;

  95.     uint8_t is_supported_endianness :1;

  96. } FormatEntry;

  97. 

  98. static const FormatEntry format_entries[AV_PIX_FMT_NB] = {

  99.     [AV_PIX_FMT_YUV420P]     = { 1, 1 },

  100.     [AV_PIX_FMT_YUYV422]     = { 1, 1 },

  101.     [AV_PIX_FMT_RGB24]       = { 1, 1 },

  102.     [AV_PIX_FMT_BGR24]       = { 1, 1 },

  103.     [AV_PIX_FMT_YUV422P]     = { 1, 1 },

  104.     [AV_PIX_FMT_YUV444P]     = { 1, 1 },

  105.     [AV_PIX_FMT_YUV410P]     = { 1, 1 },

  106.     [AV_PIX_FMT_YUV411P]     = { 1, 1 },

  107.     [AV_PIX_FMT_GRAY8]       = { 1, 1 },

  108.     [AV_PIX_FMT_MONOWHITE]   = { 1, 1 },

  109.     [AV_PIX_FMT_MONOBLACK]   = { 1, 1 },

  110.     [AV_PIX_FMT_PAL8]        = { 1, 0 },

  111.     [AV_PIX_FMT_YUVJ420P]    = { 1, 1 },

  112.     [AV_PIX_FMT_YUVJ411P]    = { 1, 1 },

  113.     [AV_PIX_FMT_YUVJ422P]    = { 1, 1 },

  114.     [AV_PIX_FMT_YUVJ444P]    = { 1, 1 },

  115.     [AV_PIX_FMT_YVYU422]     = { 1, 1 },

  116.     [AV_PIX_FMT_UYVY422]     = { 1, 1 },

  117.     [AV_PIX_FMT_UYYVYY411]   = { 0, 0 },

  118.     [AV_PIX_FMT_BGR8]        = { 1, 1 },

  119.     [AV_PIX_FMT_BGR4]        = { 0, 1 },

  120.     [AV_PIX_FMT_BGR4_BYTE]   = { 1, 1 },

  121.     [AV_PIX_FMT_RGB8]        = { 1, 1 },

  122.     [AV_PIX_FMT_RGB4]        = { 0, 1 },

  123.     [AV_PIX_FMT_RGB4_BYTE]   = { 1, 1 },

  124.     [AV_PIX_FMT_NV12]        = { 1, 1 },

  125.     [AV_PIX_FMT_NV21]        = { 1, 1 },

  126.     [AV_PIX_FMT_ARGB]        = { 1, 1 },

  127.     [AV_PIX_FMT_RGBA]        = { 1, 1 },

  128.     [AV_PIX_FMT_ABGR]        = { 1, 1 },

  129.     [AV_PIX_FMT_BGRA]        = { 1, 1 },

  130.     [AV_PIX_FMT_0RGB]        = { 1, 1 },

  131.     [AV_PIX_FMT_RGB0]        = { 1, 1 },

  132.     [AV_PIX_FMT_0BGR]        = { 1, 1 },

  133.     [AV_PIX_FMT_BGR0]        = { 1, 1 },

  134.     [AV_PIX_FMT_GRAY9BE]     = { 1, 1 },

  135.     [AV_PIX_FMT_GRAY9LE]     = { 1, 1 },

  136.     [AV_PIX_FMT_GRAY10BE]    = { 1, 1 },

  137.     [AV_PIX_FMT_GRAY10LE]    = { 1, 1 },

  138.     [AV_PIX_FMT_GRAY12BE]    = { 1, 1 },

  139.     [AV_PIX_FMT_GRAY12LE]    = { 1, 1 },

  140.     [AV_PIX_FMT_GRAY14BE]    = { 1, 1 },

  141.     [AV_PIX_FMT_GRAY14LE]    = { 1, 1 },

  142.     [AV_PIX_FMT_GRAY16BE]    = { 1, 1 },

  143.     [AV_PIX_FMT_GRAY16LE]    = { 1, 1 },

  144.     [AV_PIX_FMT_YUV440P]     = { 1, 1 },

  145.     [AV_PIX_FMT_YUVJ440P]    = { 1, 1 },

  146.     [AV_PIX_FMT_YUV440P10LE] = { 1, 1 },

  147.     [AV_PIX_FMT_YUV440P10BE] = { 1, 1 },

  148.     [AV_PIX_FMT_YUV440P12LE] = { 1, 1 },

  149.     [AV_PIX_FMT_YUV440P12BE] = { 1, 1 },

  150.     [AV_PIX_FMT_YUVA420P]    = { 1, 1 },

  151.     [AV_PIX_FMT_YUVA422P]    = { 1, 1 },

  152.     [AV_PIX_FMT_YUVA444P]    = { 1, 1 },

  153.     [AV_PIX_FMT_YUVA420P9BE] = { 1, 1 },

  154.     [AV_PIX_FMT_YUVA420P9LE] = { 1, 1 },

  155.     [AV_PIX_FMT_YUVA422P9BE] = { 1, 1 },

  156.     [AV_PIX_FMT_YUVA422P9LE] = { 1, 1 },

  157.     [AV_PIX_FMT_YUVA444P9BE] = { 1, 1 },

  158.     [AV_PIX_FMT_YUVA444P9LE] = { 1, 1 },

  159.     [AV_PIX_FMT_YUVA420P10BE]= { 1, 1 },

  160.     [AV_PIX_FMT_YUVA420P10LE]= { 1, 1 },

  161.     [AV_PIX_FMT_YUVA422P10BE]= { 1, 1 },

  162.     [AV_PIX_FMT_YUVA422P10LE]= { 1, 1 },

  163.     [AV_PIX_FMT_YUVA444P10BE]= { 1, 1 },

  164.     [AV_PIX_FMT_YUVA444P10LE]= { 1, 1 },

  165.     [AV_PIX_FMT_YUVA420P16BE]= { 1, 1 },

  166.     [AV_PIX_FMT_YUVA420P16LE]= { 1, 1 },

  167.     [AV_PIX_FMT_YUVA422P16BE]= { 1, 1 },

  168.     [AV_PIX_FMT_YUVA422P16LE]= { 1, 1 },

  169.     [AV_PIX_FMT_YUVA444P16BE]= { 1, 1 },

  170.     [AV_PIX_FMT_YUVA444P16LE]= { 1, 1 },

  171.     [AV_PIX_FMT_RGB48BE]     = { 1, 1 },

  172.     [AV_PIX_FMT_RGB48LE]     = { 1, 1 },

  173.     [AV_PIX_FMT_RGBA64BE]    = { 1, 1, 1 },

  174.     [AV_PIX_FMT_RGBA64LE]    = { 1, 1, 1 },

  175.     [AV_PIX_FMT_RGB565BE]    = { 1, 1 },

  176.     [AV_PIX_FMT_RGB565LE]    = { 1, 1 },

  177.     [AV_PIX_FMT_RGB555BE]    = { 1, 1 },

  178.     [AV_PIX_FMT_RGB555LE]    = { 1, 1 },

  179.     [AV_PIX_FMT_BGR565BE]    = { 1, 1 },

  180.     [AV_PIX_FMT_BGR565LE]    = { 1, 1 },

  181.     [AV_PIX_FMT_BGR555BE]    = { 1, 1 },

  182.     [AV_PIX_FMT_BGR555LE]    = { 1, 1 },

  183.     [AV_PIX_FMT_YUV420P16LE] = { 1, 1 },

  184.     [AV_PIX_FMT_YUV420P16BE] = { 1, 1 },

  185.     [AV_PIX_FMT_YUV422P16LE] = { 1, 1 },

  186.     [AV_PIX_FMT_YUV422P16BE] = { 1, 1 },

  187.     [AV_PIX_FMT_YUV444P16LE] = { 1, 1 },

  188.     [AV_PIX_FMT_YUV444P16BE] = { 1, 1 },

  189.     [AV_PIX_FMT_RGB444LE]    = { 1, 1 },

  190.     [AV_PIX_FMT_RGB444BE]    = { 1, 1 },

  191.     [AV_PIX_FMT_BGR444LE]    = { 1, 1 },

  192.     [AV_PIX_FMT_BGR444BE]    = { 1, 1 },

  193.     [AV_PIX_FMT_YA8]         = { 1, 1 },

  194.     [AV_PIX_FMT_YA16BE]      = { 1, 1 },

  195.     [AV_PIX_FMT_YA16LE]      = { 1, 1 },

  196.     [AV_PIX_FMT_BGR48BE]     = { 1, 1 },

  197.     [AV_PIX_FMT_BGR48LE]     = { 1, 1 },

  198.     [AV_PIX_FMT_BGRA64BE]    = { 1, 1, 1 },

  199.     [AV_PIX_FMT_BGRA64LE]    = { 1, 1, 1 },

  200.     [AV_PIX_FMT_YUV420P9BE]  = { 1, 1 },

  201.     [AV_PIX_FMT_YUV420P9LE]  = { 1, 1 },

  202.     [AV_PIX_FMT_YUV420P10BE] = { 1, 1 },

  203.     [AV_PIX_FMT_YUV420P10LE] = { 1, 1 },

  204.     [AV_PIX_FMT_YUV420P12BE] = { 1, 1 },

  205.     [AV_PIX_FMT_YUV420P12LE] = { 1, 1 },

  206.     [AV_PIX_FMT_YUV420P14BE] = { 1, 1 },

  207.     [AV_PIX_FMT_YUV420P14LE] = { 1, 1 },

  208.     [AV_PIX_FMT_YUV422P9BE]  = { 1, 1 },

  209.     [AV_PIX_FMT_YUV422P9LE]  = { 1, 1 },

  210.     [AV_PIX_FMT_YUV422P10BE] = { 1, 1 },

  211.     [AV_PIX_FMT_YUV422P10LE] = { 1, 1 },

  212.     [AV_PIX_FMT_YUV422P12BE] = { 1, 1 },

  213.     [AV_PIX_FMT_YUV422P12LE] = { 1, 1 },

  214.     [AV_PIX_FMT_YUV422P14BE] = { 1, 1 },

  215.     [AV_PIX_FMT_YUV422P14LE] = { 1, 1 },

  216.     [AV_PIX_FMT_YUV444P9BE]  = { 1, 1 },

  217.     [AV_PIX_FMT_YUV444P9LE]  = { 1, 1 },

  218.     [AV_PIX_FMT_YUV444P10BE] = { 1, 1 },

  219.     [AV_PIX_FMT_YUV444P10LE] = { 1, 1 },

  220.     [AV_PIX_FMT_YUV444P12BE] = { 1, 1 },

  221.     [AV_PIX_FMT_YUV444P12LE] = { 1, 1 },

  222.     [AV_PIX_FMT_YUV444P14BE] = { 1, 1 },

  223.     [AV_PIX_FMT_YUV444P14LE] = { 1, 1 },

  224.     [AV_PIX_FMT_GBRP]        = { 1, 1 },

  225.     [AV_PIX_FMT_GBRP9LE]     = { 1, 1 },

  226.     [AV_PIX_FMT_GBRP9BE]     = { 1, 1 },

  227.     [AV_PIX_FMT_GBRP10LE]    = { 1, 1 },

  228.     [AV_PIX_FMT_GBRP10BE]    = { 1, 1 },

  229.     [AV_PIX_FMT_GBRAP10LE]   = { 1, 1 },

  230.     [AV_PIX_FMT_GBRAP10BE]   = { 1, 1 },

  231.     [AV_PIX_FMT_GBRP12LE]    = { 1, 1 },

  232.     [AV_PIX_FMT_GBRP12BE]    = { 1, 1 },

  233.     [AV_PIX_FMT_GBRAP12LE]   = { 1, 1 },

  234.     [AV_PIX_FMT_GBRAP12BE]   = { 1, 1 },

  235.     [AV_PIX_FMT_GBRP14LE]    = { 1, 1 },

  236.     [AV_PIX_FMT_GBRP14BE]    = { 1, 1 },

  237.     [AV_PIX_FMT_GBRP16LE]    = { 1, 1 },

  238.     [AV_PIX_FMT_GBRP16BE]    = { 1, 1 },

  239.     [AV_PIX_FMT_GBRAP]       = { 1, 1 },

  240.     [AV_PIX_FMT_GBRAP16LE]   = { 1, 1 },

  241.     [AV_PIX_FMT_GBRAP16BE]   = { 1, 1 },

  242.     [AV_PIX_FMT_BAYER_BGGR8] = { 1, 0 },

  243.     [AV_PIX_FMT_BAYER_RGGB8] = { 1, 0 },

  244.     [AV_PIX_FMT_BAYER_GBRG8] = { 1, 0 },

  245.     [AV_PIX_FMT_BAYER_GRBG8] = { 1, 0 },

  246.     [AV_PIX_FMT_BAYER_BGGR16LE] = { 1, 0 },

  247.     [AV_PIX_FMT_BAYER_BGGR16BE] = { 1, 0 },

  248.     [AV_PIX_FMT_BAYER_RGGB16LE] = { 1, 0 },

  249.     [AV_PIX_FMT_BAYER_RGGB16BE] = { 1, 0 },

  250.     [AV_PIX_FMT_BAYER_GBRG16LE] = { 1, 0 },

  251.     [AV_PIX_FMT_BAYER_GBRG16BE] = { 1, 0 },

  252.     [AV_PIX_FMT_BAYER_GRBG16LE] = { 1, 0 },

  253.     [AV_PIX_FMT_BAYER_GRBG16BE] = { 1, 0 },

  254.     [AV_PIX_FMT_XYZ12BE]     = { 1, 1, 1 },

  255.     [AV_PIX_FMT_XYZ12LE]     = { 1, 1, 1 },

  256.     [AV_PIX_FMT_AYUV64LE]    = { 1, 1},

  257.     [AV_PIX_FMT_P010LE]      = { 1, 1 },

  258.     [AV_PIX_FMT_P010BE]      = { 1, 1 },

  259.     [AV_PIX_FMT_P016LE]      = { 1, 1 },

  260.     [AV_PIX_FMT_P016BE]      = { 1, 1 },

  261.     [AV_PIX_FMT_GRAYF32LE]   = { 1, 1 },

  262.     [AV_PIX_FMT_GRAYF32BE]   = { 1, 1 },

  263.     [AV_PIX_FMT_YUVA422P12BE] = { 1, 1 },

  264.     [AV_PIX_FMT_YUVA422P12LE] = { 1, 1 },

  265.     [AV_PIX_FMT_YUVA444P12BE] = { 1, 1 },

  266.     [AV_PIX_FMT_YUVA444P12LE] = { 1, 1 },

  267. };

  268. 

  269. int sws_isSupportedInput(enum AVPixelFormat pix_fmt)

  270. {

  271.     return (unsigned)pix_fmt < AV_PIX_FMT_NB ?

  272.            format_entries[pix_fmt].is_supported_in : 0;

  273. }

  274. 

  275. int sws_isSupportedOutput(enum AVPixelFormat pix_fmt)

  276. {

  277.     return (unsigned)pix_fmt < AV_PIX_FMT_NB ?

  278.            format_entries[pix_fmt].is_supported_out : 0;

  279. }

  280. 

  281. int sws_isSupportedEndiannessConversion(enum AVPixelFormat pix_fmt)

  282. {

  283.     return (unsigned)pix_fmt < AV_PIX_FMT_NB ?

  284.            format_entries[pix_fmt].is_supported_endianness : 0;

  285. }

  286. 

  287. static double getSplineCoeff(double a, double b, double c, double d,

  288.                              double dist)

  289. {

  290.     if (dist <= 1.0)

  291.         return ((d * dist + c) * dist + b) * dist + a;

  292.     else

  293.         return getSplineCoeff(0.0,

  294.                                b + 2.0 * c + 3.0 * d,

  295.                                c + 3.0 * d,

  296.                               -b - 3.0 * c - 6.0 * d,

  297.                               dist - 1.0);

  298. }

  299. 

  300. static av_cold int get_local_pos(SwsContext *s, int chr_subsample, int pos, int dir)

  301. {

  302.     if (pos == -1 || pos <= -513) {

  303.         pos = (128 << chr_subsample) - 128;

  304.     }

  305.     pos += 128; // relative to ideal left edge

  306.     return pos >> chr_subsample;

  307. }

  308. 

  309. typedef struct {

  310.     int flag;                   ///< flag associated to the algorithm

  311.     const char *description;    ///< human-readable description

  312.     int size_factor;            ///< size factor used when initing the filters

  313. } ScaleAlgorithm;

  314. 

  315. static const ScaleAlgorithm scale_algorithms[] = {

  316.     { SWS_AREA,          "area averaging",                  1 /* downscale only, for upscale it is bilinear */ },

  317.     { SWS_BICUBIC,       "bicubic",                         4 },

  318.     { SWS_BICUBLIN,      "luma bicubic / chroma bilinear", -1 },

  319.     { SWS_BILINEAR,      "bilinear",                        2 },

  320.     { SWS_FAST_BILINEAR, "fast bilinear",                  -1 },

  321.     { SWS_GAUSS,         "Gaussian",                        8 /* infinite ;) */ },

  322.     { SWS_LANCZOS,       "Lanczos",                        -1 /* custom */ },

  323.     { SWS_POINT,         "nearest neighbor / point",       -1 },

  324.     { SWS_SINC,          "sinc",                           20 /* infinite ;) */ },

  325.     { SWS_SPLINE,        "bicubic spline",                 20 /* infinite :)*/ },

  326.     { SWS_X,             "experimental",                    8 },

  327. };

  328. 

  329. static av_cold int initFilter(int16_t **outFilter, int32_t **filterPos,

  330.                               int *outFilterSize, int xInc, int srcW,

  331.                               int dstW, int filterAlign, int one,

  332.                               int flags, int cpu_flags,

  333.                               SwsVector *srcFilter, SwsVector *dstFilter,

  334.                               double param[2], int srcPos, int dstPos)

  335. {

  336.     int i;

  337.     int filterSize;

  338.     int filter2Size;

  339.     int minFilterSize;

  340.     int64_t *filter    = NULL;

  341.     int64_t *filter2   = NULL;

  342.     const int64_t fone = 1LL << (54 - FFMIN(av_log2(srcW/dstW), 8));

  343.     int ret            = -1;

  344. 

  345.     emms_c(); // FIXME should not be required but IS (even for non-MMX versions)

  346. 

  347.     // NOTE: the +3 is for the MMX(+1) / SSE(+3) scaler which reads over the end

  348.     FF_ALLOC_ARRAY_OR_GOTO(NULL, *filterPos, (dstW + 3), sizeof(**filterPos), fail);

  349. 

  350.     if (FFABS(xInc - 0x10000) < 10 && srcPos == dstPos) { // unscaled

  351.         int i;

  352.         filterSize = 1;

  353.         FF_ALLOCZ_ARRAY_OR_GOTO(NULL, filter,

  354.                                 dstW, sizeof(*filter) * filterSize, fail);

  355. 

  356.         for (i = 0; i < dstW; i++) {

  357.             filter[i * filterSize] = fone;

  358.             (*filterPos)[i]        = i;

  359.         }

  360.     } else if (flags & SWS_POINT) { // lame looking point sampling mode

  361.         int i;

  362.         int64_t xDstInSrc;

  363.         filterSize = 1;

  364.         FF_ALLOC_ARRAY_OR_GOTO(NULL, filter,

  365.                                dstW, sizeof(*filter) * filterSize, fail);

  366. 

  367.         xDstInSrc = ((dstPos*(int64_t)xInc)>>8) - ((srcPos*0x8000LL)>>7);

  368.         for (i = 0; i < dstW; i++) {

  369.             int xx = (xDstInSrc - ((filterSize - 1) << 15) + (1 << 15)) >> 16;

  370. 

  371.             (*filterPos)[i] = xx;

  372.             filter[i]       = fone;

  373.             xDstInSrc      += xInc;

  374.         }

  375.     } else if ((xInc <= (1 << 16) && (flags & SWS_AREA)) ||

  376.                (flags & SWS_FAST_BILINEAR)) { // bilinear upscale

  377.         int i;

  378.         int64_t xDstInSrc;

  379.         filterSize = 2;

  380.         FF_ALLOC_ARRAY_OR_GOTO(NULL, filter,

  381.                                dstW, sizeof(*filter) * filterSize, fail);

  382. 

  383.         xDstInSrc = ((dstPos*(int64_t)xInc)>>8) - ((srcPos*0x8000LL)>>7);

  384.         for (i = 0; i < dstW; i++) {

  385.             int xx = (xDstInSrc - ((filterSize - 1) << 15) + (1 << 15)) >> 16;

  386.             int j;

  387. 

  388.             (*filterPos)[i] = xx;

  389.             // bilinear upscale / linear interpolate / area averaging

  390.             for (j = 0; j < filterSize; j++) {

  391.                 int64_t coeff= fone - FFABS(((int64_t)xx<<16) - xDstInSrc)*(fone>>16);

  392.                 if (coeff < 0)

  393.                     coeff = 0;

  394.                 filter[i * filterSize + j] = coeff;

  395.                 xx++;

  396.             }

  397.             xDstInSrc += xInc;

  398.         }

  399.     } else {

  400.         int64_t xDstInSrc;

  401.         int sizeFactor = -1;

  402. 

  403.         for (i = 0; i < FF_ARRAY_ELEMS(scale_algorithms); i++) {

  404.             if (flags & scale_algorithms[i].flag && scale_algorithms[i].size_factor > 0) {

  405.                 sizeFactor = scale_algorithms[i].size_factor;

  406.                 break;

  407.             }

  408.         }

  409.         if (flags & SWS_LANCZOS)

  410.             sizeFactor = param[0] != SWS_PARAM_DEFAULT ? ceil(2 * param[0]) : 6;

  411.         av_assert0(sizeFactor > 0);

  412. 

  413.         if (xInc <= 1 << 16)

  414.             filterSize = 1 + sizeFactor;    // upscale

  415.         else

  416.             filterSize = 1 + (sizeFactor * srcW + dstW - 1) / dstW;

  417. 

  418.         filterSize = FFMIN(filterSize, srcW - 2);

  419.         filterSize = FFMAX(filterSize, 1);

  420. 

  421.         FF_ALLOC_ARRAY_OR_GOTO(NULL, filter,

  422.                                dstW, sizeof(*filter) * filterSize, fail);

  423. 

  424.         xDstInSrc = ((dstPos*(int64_t)xInc)>>7) - ((srcPos*0x10000LL)>>7);

  425.         for (i = 0; i < dstW; i++) {

  426.             int xx = (xDstInSrc - (filterSize - 2) * (1LL<<16)) / (1 << 17);

  427.             int j;

  428.             (*filterPos)[i] = xx;

  429.             for (j = 0; j < filterSize; j++) {

  430.                 int64_t d = (FFABS(((int64_t)xx * (1 << 17)) - xDstInSrc)) << 13;

  431.                 double floatd;

  432.                 int64_t coeff;

  433. 

  434.                 if (xInc > 1 << 16)

  435.                     d = d * dstW / srcW;

  436.                 floatd = d * (1.0 / (1 << 30));

  437. 

  438.                 if (flags & SWS_BICUBIC) {

  439.                     int64_t B = (param[0] != SWS_PARAM_DEFAULT ? param[0] :   0) * (1 << 24);

  440.                     int64_t C = (param[1] != SWS_PARAM_DEFAULT ? param[1] : 0.6) * (1 << 24);

  441. 

  442.                     if (d >= 1LL << 31) {

  443.                         coeff = 0.0;

  444.                     } else {

  445.                         int64_t dd  = (d  * d) >> 30;

  446.                         int64_t ddd = (dd * d) >> 30;

  447. 

  448.                         if (d < 1LL << 30)

  449.                             coeff =  (12 * (1 << 24) -  9 * B - 6 * C) * ddd +

  450.                                     (-18 * (1 << 24) + 12 * B + 6 * C) *  dd +

  451.                                       (6 * (1 << 24) -  2 * B)         * (1 << 30);

  452.                         else

  453.                             coeff =      (-B -  6 * C) * ddd +

  454.                                       (6 * B + 30 * C) * dd  +

  455.                                     (-12 * B - 48 * C) * d   +

  456.                                       (8 * B + 24 * C) * (1 << 30);

  457.                     }

  458.                     coeff /= (1LL<<54)/fone;

  459.                 } else if (flags & SWS_X) {

  460.                     double A = param[0] != SWS_PARAM_DEFAULT ? param[0] : 1.0;

  461.                     double c;

  462. 

  463.                     if (floatd < 1.0)

  464.                         c = cos(floatd * M_PI);

  465.                     else

  466.                         c = -1.0;

  467.                     if (c < 0.0)

  468.                         c = -pow(-c, A);

  469.                     else

  470.                         c = pow(c, A);

  471.                     coeff = (c * 0.5 + 0.5) * fone;

  472.                 } else if (flags & SWS_AREA) {

  473.                     int64_t d2 = d - (1 << 29);

  474.                     if (d2 * xInc < -(1LL << (29 + 16)))

  475.                         coeff = 1.0 * (1LL << (30 + 16));

  476.                     else if (d2 * xInc < (1LL << (29 + 16)))

  477.                         coeff = -d2 * xInc + (1LL << (29 + 16));

  478.                     else

  479.                         coeff = 0.0;

  480.                     coeff *= fone >> (30 + 16);

  481.                 } else if (flags & SWS_GAUSS) {

  482.                     double p = param[0] != SWS_PARAM_DEFAULT ? param[0] : 3.0;

  483.                     coeff = exp2(-p * floatd * floatd) * fone;

  484.                 } else if (flags & SWS_SINC) {

  485.                     coeff = (d ? sin(floatd * M_PI) / (floatd * M_PI) : 1.0) * fone;

  486.                 } else if (flags & SWS_LANCZOS) {

  487.                     double p = param[0] != SWS_PARAM_DEFAULT ? param[0] : 3.0;

  488.                     coeff = (d ? sin(floatd * M_PI) * sin(floatd * M_PI / p) /

  489.                              (floatd * floatd * M_PI * M_PI / p) : 1.0) * fone;

  490.                     if (floatd > p)

  491.                         coeff = 0;

  492.                 } else if (flags & SWS_BILINEAR) {

  493.                     coeff = (1 << 30) - d;

  494.                     if (coeff < 0)

  495.                         coeff = 0;

  496.                     coeff *= fone >> 30;

  497.                 } else if (flags & SWS_SPLINE) {

  498.                     double p = -2.196152422706632;

  499.                     coeff = getSplineCoeff(1.0, 0.0, p, -p - 1.0, floatd) * fone;

  500.                 } else {

  501.                     av_assert0(0);

  502.                 }

  503. 

  504.                 filter[i * filterSize + j] = coeff;

  505.                 xx++;

  506.             }

  507.             xDstInSrc += 2 * xInc;

  508.         }

  509.     }

  510. 

  511.     /* apply src & dst Filter to filter -> filter2

  512.      * av_free(filter);

  513.      */

  514.     av_assert0(filterSize > 0);

  515.     filter2Size = filterSize;

  516.     if (srcFilter)

  517.         filter2Size += srcFilter->length - 1;

  518.     if (dstFilter)

  519.         filter2Size += dstFilter->length - 1;

  520.     av_assert0(filter2Size > 0);

  521.     FF_ALLOCZ_ARRAY_OR_GOTO(NULL, filter2, dstW, filter2Size * sizeof(*filter2), fail);

  522. 

  523.     for (i = 0; i < dstW; i++) {

  524.         int j, k;

  525. 

  526.         if (srcFilter) {

  527.             for (k = 0; k < srcFilter->length; k++) {

  528.                 for (j = 0; j < filterSize; j++)

  529.                     filter2[i * filter2Size + k + j] +=

  530.                         srcFilter->coeff[k] * filter[i * filterSize + j];

  531.             }

  532.         } else {

  533.             for (j = 0; j < filterSize; j++)

  534.                 filter2[i * filter2Size + j] = filter[i * filterSize + j];

  535.         }

  536.         // FIXME dstFilter

  537. 

  538.         (*filterPos)[i] += (filterSize - 1) / 2 - (filter2Size - 1) / 2;

  539.     }

  540.     av_freep(&filter);

  541. 

  542.     /* try to reduce the filter-size (step1 find size and shift left) */

  543.     // Assume it is near normalized (*0.5 or *2.0 is OK but * 0.001 is not).

  544.     minFilterSize = 0;

  545.     for (i = dstW - 1; i >= 0; i--) {

  546.         int min = filter2Size;

  547.         int j;

  548.         int64_t cutOff = 0.0;

  549. 

  550.         /* get rid of near zero elements on the left by shifting left */

  551.         for (j = 0; j < filter2Size; j++) {

  552.             int k;

  553.             cutOff += FFABS(filter2[i * filter2Size]);

  554. 

  555.             if (cutOff > SWS_MAX_REDUCE_CUTOFF * fone)

  556.                 break;

  557. 

  558.             /* preserve monotonicity because the core can't handle the

  559.              * filter otherwise */

  560.             if (i < dstW - 1 && (*filterPos)[i] >= (*filterPos)[i + 1])

  561.                 break;

  562. 

  563.             // move filter coefficients left

  564.             for (k = 1; k < filter2Size; k++)

  565.                 filter2[i * filter2Size + k - 1] = filter2[i * filter2Size + k];

  566.             filter2[i * filter2Size + k - 1] = 0;

  567.             (*filterPos)[i]++;

  568.         }

  569. 

  570.         cutOff = 0;

  571.         /* count near zeros on the right */

  572.         for (j = filter2Size - 1; j > 0; j--) {

  573.             cutOff += FFABS(filter2[i * filter2Size + j]);

  574. 

  575.             if (cutOff > SWS_MAX_REDUCE_CUTOFF * fone)

  576.                 break;

  577.             min--;

  578.         }

  579. 

  580.         if (min > minFilterSize)

  581.             minFilterSize = min;

  582.     }

  583. 

  584.     if (PPC_ALTIVEC(cpu_flags)) {

  585.         // we can handle the special case 4, so we don't want to go the full 8

  586.         if (minFilterSize < 5)

  587.             filterAlign = 4;

  588. 

  589.         /* We really don't want to waste our time doing useless computation, so

  590.          * fall back on the scalar C code for very small filters.

  591.          * Vectorizing is worth it only if you have a decent-sized vector. */

  592.         if (minFilterSize < 3)

  593.             filterAlign = 1;

  594.     }

  595. 

  596.     if (HAVE_MMX && cpu_flags & AV_CPU_FLAG_MMX) {

  597.         // special case for unscaled vertical filtering

  598.         if (minFilterSize == 1 && filterAlign == 2)

  599.             filterAlign = 1;

  600.     }

  601. 

  602.     av_assert0(minFilterSize > 0);

  603.     filterSize = (minFilterSize + (filterAlign - 1)) & (~(filterAlign - 1));

  604.     av_assert0(filterSize > 0);

  605.     filter = av_malloc_array(dstW, filterSize * sizeof(*filter));

  606.     if (!filter)

  607.         goto fail;

  608.     if (filterSize >= MAX_FILTER_SIZE * 16 /

  609.                       ((flags & SWS_ACCURATE_RND) ? APCK_SIZE : 16)) {

  610.         ret = RETCODE_USE_CASCADE;

  611.         goto fail;

  612.     }

  613.     *outFilterSize = filterSize;

  614. 

  615.     if (flags & SWS_PRINT_INFO)

  616.         av_log(NULL, AV_LOG_VERBOSE,

  617.                "SwScaler: reducing / aligning filtersize %d -> %d\n",

  618.                filter2Size, filterSize);

  619.     /* try to reduce the filter-size (step2 reduce it) */

  620.     for (i = 0; i < dstW; i++) {

  621.         int j;

  622. 

  623.         for (j = 0; j < filterSize; j++) {

  624.             if (j >= filter2Size)

  625.                 filter[i * filterSize + j] = 0;

  626.             else

  627.                 filter[i * filterSize + j] = filter2[i * filter2Size + j];

  628.             if ((flags & SWS_BITEXACT) && j >= minFilterSize)

  629.                 filter[i * filterSize + j] = 0;

  630.         }

  631.     }

  632. 

  633.     // FIXME try to align filterPos if possible

  634. 

  635.     // fix borders

  636.     for (i = 0; i < dstW; i++) {

  637.         int j;

  638.         if ((*filterPos)[i] < 0) {

  639.             // move filter coefficients left to compensate for filterPos

  640.             for (j = 1; j < filterSize; j++) {

  641.                 int left = FFMAX(j + (*filterPos)[i], 0);

  642.                 filter[i * filterSize + left] += filter[i * filterSize + j];

  643.                 filter[i * filterSize + j]     = 0;

  644.             }

  645.             (*filterPos)[i]= 0;

  646.         }

  647. 

  648.         if ((*filterPos)[i] + filterSize > srcW) {

  649.             int shift = (*filterPos)[i] + FFMIN(filterSize - srcW, 0);

  650.             int64_t acc = 0;

  651. 

  652.             for (j = filterSize - 1; j >= 0; j--) {

  653.                 if ((*filterPos)[i] + j >= srcW) {

  654.                     acc += filter[i * filterSize + j];

  655.                     filter[i * filterSize + j] = 0;

  656.                 }

  657.             }

  658.             for (j = filterSize - 1; j >= 0; j--) {

  659.                 if (j < shift) {

  660.                     filter[i * filterSize + j] = 0;

  661.                 } else {

  662.                     filter[i * filterSize + j] = filter[i * filterSize + j - shift];

  663.                 }

  664.             }

  665. 

  666.             (*filterPos)[i]-= shift;

  667.             filter[i * filterSize + srcW - 1 - (*filterPos)[i]] += acc;

  668.         }

  669.         av_assert0((*filterPos)[i] >= 0);

  670.         av_assert0((*filterPos)[i] < srcW);

  671.         if ((*filterPos)[i] + filterSize > srcW) {

  672.             for (j = 0; j < filterSize; j++) {

  673.                 av_assert0((*filterPos)[i] + j < srcW || !filter[i * filterSize + j]);

  674.             }

  675.         }

  676.     }

  677. 

  678.     // Note the +1 is for the MMX scaler which reads over the end

  679.     /* align at 16 for AltiVec (needed by hScale_altivec_real) */

  680.     FF_ALLOCZ_ARRAY_OR_GOTO(NULL, *outFilter,

  681.                             (dstW + 3), *outFilterSize * sizeof(int16_t), fail);

  682. 

  683.     /* normalize & store in outFilter */

  684.     for (i = 0; i < dstW; i++) {

  685.         int j;

  686.         int64_t error = 0;

  687.         int64_t sum   = 0;

  688. 

  689.         for (j = 0; j < filterSize; j++) {

  690.             sum += filter[i * filterSize + j];

  691.         }

  692.         sum = (sum + one / 2) / one;

  693.         if (!sum) {

  694.             av_log(NULL, AV_LOG_WARNING, "SwScaler: zero vector in scaling\n");

  695.             sum = 1;

  696.         }

  697.         for (j = 0; j < *outFilterSize; j++) {

  698.             int64_t v = filter[i * filterSize + j] + error;

  699.             int intV  = ROUNDED_DIV(v, sum);

  700.             (*outFilter)[i * (*outFilterSize) + j] = intV;

  701.             error                                  = v - intV * sum;

  702.         }

  703.     }

  704. 

  705.     (*filterPos)[dstW + 0] =

  706.     (*filterPos)[dstW + 1] =

  707.     (*filterPos)[dstW + 2] = (*filterPos)[dstW - 1]; /* the MMX/SSE scaler will

  708.                                                       * read over the end */

  709.     for (i = 0; i < *outFilterSize; i++) {

  710.         int k = (dstW - 1) * (*outFilterSize) + i;

  711.         (*outFilter)[k + 1 * (*outFilterSize)] =

  712.         (*outFilter)[k + 2 * (*outFilterSize)] =

  713.         (*outFilter)[k + 3 * (*outFilterSize)] = (*outFilter)[k];

  714.     }

  715. 

  716.     ret = 0;

  717. 

  718. fail:

  719.     if(ret < 0)

  720.         av_log(NULL, ret == RETCODE_USE_CASCADE ? AV_LOG_DEBUG : AV_LOG_ERROR, "sws: initFilter failed\n");

  721.     av_free(filter);

  722.     av_free(filter2);

  723.     return ret;

  724. }

  725. 

  726. static void fill_rgb2yuv_table(SwsContext *c, const int table[4], int dstRange)

  727. {

  728.     int64_t W, V, Z, Cy, Cu, Cv;

  729.     int64_t vr =  table[0];

  730.     int64_t ub =  table[1];

  731.     int64_t ug = -table[2];

  732.     int64_t vg = -table[3];

  733.     int64_t ONE = 65536;

  734.     int64_t cy = ONE;

  735.     uint8_t *p = (uint8_t*)c->input_rgb2yuv_table;

  736.     int i;

  737.     static const int8_t map[] = {

  738.     BY_IDX, GY_IDX, -1    , BY_IDX, BY_IDX, GY_IDX, -1    , BY_IDX,

  739.     RY_IDX, -1    , GY_IDX, RY_IDX, RY_IDX, -1    , GY_IDX, RY_IDX,

  740.     RY_IDX, GY_IDX, -1    , RY_IDX, RY_IDX, GY_IDX, -1    , RY_IDX,

  741.     BY_IDX, -1    , GY_IDX, BY_IDX, BY_IDX, -1    , GY_IDX, BY_IDX,

  742.     BU_IDX, GU_IDX, -1    , BU_IDX, BU_IDX, GU_IDX, -1    , BU_IDX,

  743.     RU_IDX, -1    , GU_IDX, RU_IDX, RU_IDX, -1    , GU_IDX, RU_IDX,

  744.     RU_IDX, GU_IDX, -1    , RU_IDX, RU_IDX, GU_IDX, -1    , RU_IDX,

  745.     BU_IDX, -1    , GU_IDX, BU_IDX, BU_IDX, -1    , GU_IDX, BU_IDX,

  746.     BV_IDX, GV_IDX, -1    , BV_IDX, BV_IDX, GV_IDX, -1    , BV_IDX,

  747.     RV_IDX, -1    , GV_IDX, RV_IDX, RV_IDX, -1    , GV_IDX, RV_IDX,

  748.     RV_IDX, GV_IDX, -1    , RV_IDX, RV_IDX, GV_IDX, -1    , RV_IDX,

  749.     BV_IDX, -1    , GV_IDX, BV_IDX, BV_IDX, -1    , GV_IDX, BV_IDX,

  750.     RY_IDX, BY_IDX, RY_IDX, BY_IDX, RY_IDX, BY_IDX, RY_IDX, BY_IDX,

  751.     BY_IDX, RY_IDX, BY_IDX, RY_IDX, BY_IDX, RY_IDX, BY_IDX, RY_IDX,

  752.     GY_IDX, -1    , GY_IDX, -1    , GY_IDX, -1    , GY_IDX, -1    ,

  753.     -1    , GY_IDX, -1    , GY_IDX, -1    , GY_IDX, -1    , GY_IDX,

  754.     RU_IDX, BU_IDX, RU_IDX, BU_IDX, RU_IDX, BU_IDX, RU_IDX, BU_IDX,

  755.     BU_IDX, RU_IDX, BU_IDX, RU_IDX, BU_IDX, RU_IDX, BU_IDX, RU_IDX,

  756.     GU_IDX, -1    , GU_IDX, -1    , GU_IDX, -1    , GU_IDX, -1    ,

  757.     -1    , GU_IDX, -1    , GU_IDX, -1    , GU_IDX, -1    , GU_IDX,

  758.     RV_IDX, BV_IDX, RV_IDX, BV_IDX, RV_IDX, BV_IDX, RV_IDX, BV_IDX,

  759.     BV_IDX, RV_IDX, BV_IDX, RV_IDX, BV_IDX, RV_IDX, BV_IDX, RV_IDX,

  760.     GV_IDX, -1    , GV_IDX, -1    , GV_IDX, -1    , GV_IDX, -1    ,

  761.     -1    , GV_IDX, -1    , GV_IDX, -1    , GV_IDX, -1    , GV_IDX, //23

  762.     -1    , -1    , -1    , -1    , -1    , -1    , -1    , -1    , //24

  763.     -1    , -1    , -1    , -1    , -1    , -1    , -1    , -1    , //25

  764.     -1    , -1    , -1    , -1    , -1    , -1    , -1    , -1    , //26

  765.     -1    , -1    , -1    , -1    , -1    , -1    , -1    , -1    , //27

  766.     -1    , -1    , -1    , -1    , -1    , -1    , -1    , -1    , //28

  767.     -1    , -1    , -1    , -1    , -1    , -1    , -1    , -1    , //29

  768.     -1    , -1    , -1    , -1    , -1    , -1    , -1    , -1    , //30

  769.     -1    , -1    , -1    , -1    , -1    , -1    , -1    , -1    , //31

  770.     BY_IDX, GY_IDX, RY_IDX, -1    , -1    , -1    , -1    , -1    , //32

  771.     BU_IDX, GU_IDX, RU_IDX, -1    , -1    , -1    , -1    , -1    , //33

  772.     BV_IDX, GV_IDX, RV_IDX, -1    , -1    , -1    , -1    , -1    , //34

  773.     };

  774. 

  775.     dstRange = 0; //FIXME range = 1 is handled elsewhere

  776. 

  777.     if (!dstRange) {

  778.         cy = cy * 255 / 219;

  779.     } else {

  780.         vr = vr * 224 / 255;

  781.         ub = ub * 224 / 255;

  782.         ug = ug * 224 / 255;

  783.         vg = vg * 224 / 255;

  784.     }

  785.     W = ROUNDED_DIV(ONE*ONE*ug, ub);

  786.     V = ROUNDED_DIV(ONE*ONE*vg, vr);

  787.     Z = ONE*ONE-W-V;

  788. 

  789.     Cy = ROUNDED_DIV(cy*Z, ONE);

  790.     Cu = ROUNDED_DIV(ub*Z, ONE);

  791.     Cv = ROUNDED_DIV(vr*Z, ONE);

  792. 

  793.     c->input_rgb2yuv_table[RY_IDX] = -ROUNDED_DIV((1 << RGB2YUV_SHIFT)*V        , Cy);

  794.     c->input_rgb2yuv_table[GY_IDX] =  ROUNDED_DIV((1 << RGB2YUV_SHIFT)*ONE*ONE  , Cy);

  795.     c->input_rgb2yuv_table[BY_IDX] = -ROUNDED_DIV((1 << RGB2YUV_SHIFT)*W        , Cy);

  796. 

  797.     c->input_rgb2yuv_table[RU_IDX] =  ROUNDED_DIV((1 << RGB2YUV_SHIFT)*V        , Cu);

  798.     c->input_rgb2yuv_table[GU_IDX] = -ROUNDED_DIV((1 << RGB2YUV_SHIFT)*ONE*ONE  , Cu);

  799.     c->input_rgb2yuv_table[BU_IDX] =  ROUNDED_DIV((1 << RGB2YUV_SHIFT)*(Z+W)    , Cu);

  800. 

  801.     c->input_rgb2yuv_table[RV_IDX] =  ROUNDED_DIV((1 << RGB2YUV_SHIFT)*(V+Z)    , Cv);

  802.     c->input_rgb2yuv_table[GV_IDX] = -ROUNDED_DIV((1 << RGB2YUV_SHIFT)*ONE*ONE  , Cv);

  803.     c->input_rgb2yuv_table[BV_IDX] =  ROUNDED_DIV((1 << RGB2YUV_SHIFT)*W        , Cv);

  804. 

  805.     if(/*!dstRange && */!memcmp(table, ff_yuv2rgb_coeffs[SWS_CS_DEFAULT], sizeof(ff_yuv2rgb_coeffs[SWS_CS_DEFAULT]))) {

  806.         c->input_rgb2yuv_table[BY_IDX] =  ((int)(0.114 * 219 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));

  807.         c->input_rgb2yuv_table[BV_IDX] = (-(int)(0.081 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));

  808.         c->input_rgb2yuv_table[BU_IDX] =  ((int)(0.500 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));

  809.         c->input_rgb2yuv_table[GY_IDX] =  ((int)(0.587 * 219 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));

  810.         c->input_rgb2yuv_table[GV_IDX] = (-(int)(0.419 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));

  811.         c->input_rgb2yuv_table[GU_IDX] = (-(int)(0.331 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));

  812.         c->input_rgb2yuv_table[RY_IDX] =  ((int)(0.299 * 219 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));

  813.         c->input_rgb2yuv_table[RV_IDX] =  ((int)(0.500 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));

  814.         c->input_rgb2yuv_table[RU_IDX] = (-(int)(0.169 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));

  815.     }

  816.     for(i=0; i<FF_ARRAY_ELEMS(map); i++)

  817.         AV_WL16(p + 16*4 + 2*i, map[i] >= 0 ? c->input_rgb2yuv_table[map[i]] : 0);

  818. }

  819. 

  820. static void fill_xyztables(struct SwsContext *c)

  821. {

  822.     int i;

  823.     double xyzgamma = XYZ_GAMMA;

  824.     double rgbgamma = 1.0 / RGB_GAMMA;

  825.     double xyzgammainv = 1.0 / XYZ_GAMMA;

  826.     double rgbgammainv = RGB_GAMMA;

  827.     static const int16_t xyz2rgb_matrix[3][4] = {

  828.         {13270, -6295, -2041},

  829.         {-3969,  7682,   170},

  830.         {  228,  -835,  4329} };

  831.     static const int16_t rgb2xyz_matrix[3][4] = {

  832.         {1689, 1464,  739},

  833.         { 871, 2929,  296},

  834.         {  79,  488, 3891} };

  835.     static int16_t xyzgamma_tab[4096], rgbgamma_tab[4096], xyzgammainv_tab[4096], rgbgammainv_tab[4096];

  836. 

  837.     memcpy(c->xyz2rgb_matrix, xyz2rgb_matrix, sizeof(c->xyz2rgb_matrix));

  838.     memcpy(c->rgb2xyz_matrix, rgb2xyz_matrix, sizeof(c->rgb2xyz_matrix));

  839.     c->xyzgamma = xyzgamma_tab;

  840.     c->rgbgamma = rgbgamma_tab;

  841.     c->xyzgammainv = xyzgammainv_tab;

  842.     c->rgbgammainv = rgbgammainv_tab;

  843. 

  844.     if (rgbgamma_tab[4095])

  845.         return;

  846. 

  847.     /* set gamma vectors */

  848.     for (i = 0; i < 4096; i++) {

  849.         xyzgamma_tab[i] = lrint(pow(i / 4095.0, xyzgamma) * 4095.0);

  850.         rgbgamma_tab[i] = lrint(pow(i / 4095.0, rgbgamma) * 4095.0);

  851.         xyzgammainv_tab[i] = lrint(pow(i / 4095.0, xyzgammainv) * 4095.0);

  852.         rgbgammainv_tab[i] = lrint(pow(i / 4095.0, rgbgammainv) * 4095.0);

  853.     }

  854. }

  855. 

  856. int sws_setColorspaceDetails(struct SwsContext *c, const int inv_table[4],

  857.                              int srcRange, const int table[4], int dstRange,

  858.                              int brightness, int contrast, int saturation)

  859. {

  860.     const AVPixFmtDescriptor *desc_dst;

  861.     const AVPixFmtDescriptor *desc_src;

  862.     int need_reinit = 0;

  863. 

  864.     handle_formats(c);

  865.     desc_dst = av_pix_fmt_desc_get(c->dstFormat);

  866.     desc_src = av_pix_fmt_desc_get(c->srcFormat);

  867. 

  868.     if(!isYUV(c->dstFormat) && !isGray(c->dstFormat))

  869.         dstRange = 0;

  870.     if(!isYUV(c->srcFormat) && !isGray(c->srcFormat))

  871.         srcRange = 0;

  872. 

  873.     if (c->srcRange != srcRange ||

  874.         c->dstRange != dstRange ||

  875.         c->brightness != brightness ||

  876.         c->contrast   != contrast ||

  877.         c->saturation != saturation ||

  878.         memcmp(c->srcColorspaceTable, inv_table, sizeof(int) * 4) ||

  879.         memcmp(c->dstColorspaceTable,     table, sizeof(int) * 4)

  880.     )

  881.         need_reinit = 1;

  882. 

  883.     memmove(c->srcColorspaceTable, inv_table, sizeof(int) * 4);

  884.     memmove(c->dstColorspaceTable, table, sizeof(int) * 4);

  885. 

  886. 

  887. 

  888.     c->brightness = brightness;

  889.     c->contrast   = contrast;

  890.     c->saturation = saturation;

  891.     c->srcRange   = srcRange;

  892.     c->dstRange   = dstRange;

  893. 

  894.     //The srcBpc check is possibly wrong but we seem to lack a definitive reference to test this

  895.     //and what we have in ticket 2939 looks better with this check

  896.     if (need_reinit && (c->srcBpc == 8 || !isYUV(c->srcFormat)))

  897.         ff_sws_init_range_convert(c);

  898. 

  899.     c->dstFormatBpp = av_get_bits_per_pixel(desc_dst);

  900.     c->srcFormatBpp = av_get_bits_per_pixel(desc_src);

  901. 

  902.     if (c->cascaded_context[c->cascaded_mainindex])

  903.         return sws_setColorspaceDetails(c->cascaded_context[c->cascaded_mainindex],inv_table, srcRange,table, dstRange, brightness,  contrast, saturation);

  904. 

  905.     if (!need_reinit)

  906.         return 0;

  907. 

  908.     if ((isYUV(c->dstFormat) || isGray(c->dstFormat)) && (isYUV(c->srcFormat) || isGray(c->srcFormat))) {

  909.         if (!c->cascaded_context[0] &&

  910.             memcmp(c->dstColorspaceTable, c->srcColorspaceTable, sizeof(int) * 4) &&

  911.             c->srcW && c->srcH && c->dstW && c->dstH) {

  912.             enum AVPixelFormat tmp_format;

  913.             int tmp_width, tmp_height;

  914.             int srcW = c->srcW;

  915.             int srcH = c->srcH;

  916.             int dstW = c->dstW;

  917.             int dstH = c->dstH;

  918.             int ret;

  919.             av_log(c, AV_LOG_VERBOSE, "YUV color matrix differs for YUV->YUV, using intermediate RGB to convert\n");

  920. 

  921.             if (isNBPS(c->dstFormat) || is16BPS(c->dstFormat)) {

  922.                 if (isALPHA(c->srcFormat) && isALPHA(c->dstFormat)) {

  923.                     tmp_format = AV_PIX_FMT_BGRA64;

  924.                 } else {

  925.                     tmp_format = AV_PIX_FMT_BGR48;

  926.                 }

  927.             } else {

  928.                 if (isALPHA(c->srcFormat) && isALPHA(c->dstFormat)) {

  929.                     tmp_format = AV_PIX_FMT_BGRA;

  930.                 } else {

  931.                     tmp_format = AV_PIX_FMT_BGR24;

  932.                 }

  933.             }

  934. 

  935.             if (srcW*srcH > dstW*dstH) {

  936.                 tmp_width  = dstW;

  937.                 tmp_height = dstH;

  938.             } else {

  939.                 tmp_width  = srcW;

  940.                 tmp_height = srcH;

  941.             }

  942. 

  943.             ret = av_image_alloc(c->cascaded_tmp, c->cascaded_tmpStride,

  944.                                 tmp_width, tmp_height, tmp_format, 64);

  945.             if (ret < 0)

  946.                 return ret;

  947. 

  948.             c->cascaded_context[0] = sws_alloc_set_opts(srcW, srcH, c->srcFormat,

  949.                                                         tmp_width, tmp_height, tmp_format,

  950.                                                         c->flags, c->param);

  951.             if (!c->cascaded_context[0])

  952.                 return -1;

  953. 

  954.             c->cascaded_context[0]->alphablend = c->alphablend;

  955.             ret = sws_init_context(c->cascaded_context[0], NULL , NULL);

  956.             if (ret < 0)

  957.                 return ret;

  958.             //we set both src and dst depending on that the RGB side will be ignored

  959.             sws_setColorspaceDetails(c->cascaded_context[0], inv_table,

  960.                                      srcRange, table, dstRange,

  961.                                      brightness, contrast, saturation);

  962. 

  963.             c->cascaded_context[1] = sws_getContext(tmp_width, tmp_height, tmp_format,

  964.                                                     dstW, dstH, c->dstFormat,

  965.                                                     c->flags, NULL, NULL, c->param);

  966.             if (!c->cascaded_context[1])

  967.                 return -1;

  968.             sws_setColorspaceDetails(c->cascaded_context[1], inv_table,

  969.                                      srcRange, table, dstRange,

  970.                                      0, 1 << 16, 1 << 16);

  971.             return 0;

  972.         }

  973.         return -1;

  974.     }

  975. 

  976.     if (!isYUV(c->dstFormat) && !isGray(c->dstFormat)) {

  977.         ff_yuv2rgb_c_init_tables(c, inv_table, srcRange, brightness,

  978.                                  contrast, saturation);

  979.         // FIXME factorize

  980. 

  981.         if (ARCH_PPC)

  982.             ff_yuv2rgb_init_tables_ppc(c, inv_table, brightness,

  983.                                        contrast, saturation);

  984.     }

  985. 

  986.     fill_rgb2yuv_table(c, table, dstRange);

  987. 

  988.     return 0;

  989. }

  990. 

  991. int sws_getColorspaceDetails(struct SwsContext *c, int **inv_table,

  992.                              int *srcRange, int **table, int *dstRange,

  993.                              int *brightness, int *contrast, int *saturation)

  994. {

  995.     if (!c )

  996.         return -1;

  997. 

  998.     *inv_table  = c->srcColorspaceTable;

  999.     *table      = c->dstColorspaceTable;

  1000.     *srcRange   = c->srcRange;

  1001.     *dstRange   = c->dstRange;

  1002.     *brightness = c->brightness;

  1003.     *contrast   = c->contrast;

  1004.     *saturation = c->saturation;

  1005. 

  1006.     return 0;

  1007. }

  1008. 

  1009. static int handle_jpeg(enum AVPixelFormat *format)

  1010. {

  1011.     switch (*format) {

  1012.     case AV_PIX_FMT_YUVJ420P:

  1013.         *format = AV_PIX_FMT_YUV420P;

  1014.         return 1;

  1015.     case AV_PIX_FMT_YUVJ411P:

  1016.         *format = AV_PIX_FMT_YUV411P;

  1017.         return 1;

  1018.     case AV_PIX_FMT_YUVJ422P:

  1019.         *format = AV_PIX_FMT_YUV422P;

  1020.         return 1;

  1021.     case AV_PIX_FMT_YUVJ444P:

  1022.         *format = AV_PIX_FMT_YUV444P;

  1023.         return 1;

  1024.     case AV_PIX_FMT_YUVJ440P:

  1025.         *format = AV_PIX_FMT_YUV440P;

  1026.         return 1;

  1027.     case AV_PIX_FMT_GRAY8:

  1028.     case AV_PIX_FMT_YA8:

  1029.     case AV_PIX_FMT_GRAY9LE:

  1030.     case AV_PIX_FMT_GRAY9BE:

  1031.     case AV_PIX_FMT_GRAY10LE:

  1032.     case AV_PIX_FMT_GRAY10BE:

  1033.     case AV_PIX_FMT_GRAY12LE:

  1034.     case AV_PIX_FMT_GRAY12BE:

  1035.     case AV_PIX_FMT_GRAY14LE:

  1036.     case AV_PIX_FMT_GRAY14BE:

  1037.     case AV_PIX_FMT_GRAY16LE:

  1038.     case AV_PIX_FMT_GRAY16BE:

  1039.     case AV_PIX_FMT_YA16BE:

  1040.     case AV_PIX_FMT_YA16LE:

  1041.         return 1;

  1042.     default:

  1043.         return 0;

  1044.     }

  1045. }

  1046. 

  1047. static int handle_0alpha(enum AVPixelFormat *format)

  1048. {

  1049.     switch (*format) {

  1050.     case AV_PIX_FMT_0BGR    : *format = AV_PIX_FMT_ABGR   ; return 1;

  1051.     case AV_PIX_FMT_BGR0    : *format = AV_PIX_FMT_BGRA   ; return 4;

  1052.     case AV_PIX_FMT_0RGB    : *format = AV_PIX_FMT_ARGB   ; return 1;

  1053.     case AV_PIX_FMT_RGB0    : *format = AV_PIX_FMT_RGBA   ; return 4;

  1054.     default:                                          return 0;

  1055.     }

  1056. }

  1057. 

  1058. static int handle_xyz(enum AVPixelFormat *format)

  1059. {

  1060.     switch (*format) {

  1061.     case AV_PIX_FMT_XYZ12BE : *format = AV_PIX_FMT_RGB48BE; return 1;

  1062.     case AV_PIX_FMT_XYZ12LE : *format = AV_PIX_FMT_RGB48LE; return 1;

  1063.     default:                                                return 0;

  1064.     }

  1065. }

  1066. 

  1067. static void handle_formats(SwsContext *c)

  1068. {

  1069.     c->src0Alpha |= handle_0alpha(&c->srcFormat);

  1070.     c->dst0Alpha |= handle_0alpha(&c->dstFormat);

  1071.     c->srcXYZ    |= handle_xyz(&c->srcFormat);

  1072.     c->dstXYZ    |= handle_xyz(&c->dstFormat);

  1073.     if (c->srcXYZ || c->dstXYZ)

  1074.         fill_xyztables(c);

  1075. }

  1076. 

  1077. SwsContext *sws_alloc_context(void)

  1078. {

  1079.     SwsContext *c = av_mallocz(sizeof(SwsContext));

  1080. 

  1081.     av_assert0(offsetof(SwsContext, redDither) + DITHER32_INT == offsetof(SwsContext, dither32));

  1082. 

  1083.     if (c) {

  1084.         c->av_class = &ff_sws_context_class;

  1085.         av_opt_set_defaults(c);

  1086.     }

  1087. 

  1088.     return c;

  1089. }

  1090. 

  1091. static uint16_t * alloc_gamma_tbl(double e)

  1092. {

  1093.     int i = 0;

  1094.     uint16_t * tbl;

  1095.     tbl = (uint16_t*)av_malloc(sizeof(uint16_t) * 1 << 16);

  1096.     if (!tbl)

  1097.         return NULL;

  1098. 

  1099.     for (i = 0; i < 65536; ++i) {

  1100.         tbl[i] = pow(i / 65535.0, e) * 65535.0;

  1101.     }

  1102.     return tbl;

  1103. }

  1104. 

  1105. static enum AVPixelFormat alphaless_fmt(enum AVPixelFormat fmt)

  1106. {

  1107.     switch(fmt) {

  1108.     case AV_PIX_FMT_ARGB:       return AV_PIX_FMT_RGB24;

  1109.     case AV_PIX_FMT_RGBA:       return AV_PIX_FMT_RGB24;

  1110.     case AV_PIX_FMT_ABGR:       return AV_PIX_FMT_BGR24;

  1111.     case AV_PIX_FMT_BGRA:       return AV_PIX_FMT_BGR24;

  1112.     case AV_PIX_FMT_YA8:        return AV_PIX_FMT_GRAY8;

  1113. 

  1114.     case AV_PIX_FMT_YUVA420P:   return AV_PIX_FMT_YUV420P;

  1115.     case AV_PIX_FMT_YUVA422P:   return AV_PIX_FMT_YUV422P;

  1116.     case AV_PIX_FMT_YUVA444P:           return AV_PIX_FMT_YUV444P;

  1117. 

  1118.     case AV_PIX_FMT_GBRAP:              return AV_PIX_FMT_GBRP;

  1119. 

  1120.     case AV_PIX_FMT_GBRAP10LE:          return AV_PIX_FMT_GBRP10;

  1121.     case AV_PIX_FMT_GBRAP10BE:          return AV_PIX_FMT_GBRP10;

  1122. 

  1123.     case AV_PIX_FMT_GBRAP12LE:          return AV_PIX_FMT_GBRP12;

  1124.     case AV_PIX_FMT_GBRAP12BE:          return AV_PIX_FMT_GBRP12;

  1125. 

  1126.     case AV_PIX_FMT_GBRAP16LE:          return AV_PIX_FMT_GBRP16;

  1127.     case AV_PIX_FMT_GBRAP16BE:          return AV_PIX_FMT_GBRP16;

  1128. 

  1129.     case AV_PIX_FMT_RGBA64LE:   return AV_PIX_FMT_RGB48;

  1130.     case AV_PIX_FMT_RGBA64BE:   return AV_PIX_FMT_RGB48;

  1131.     case AV_PIX_FMT_BGRA64LE:   return AV_PIX_FMT_BGR48;

  1132.     case AV_PIX_FMT_BGRA64BE:   return AV_PIX_FMT_BGR48;

  1133. 

  1134.     case AV_PIX_FMT_YA16BE:             return AV_PIX_FMT_GRAY16;

  1135.     case AV_PIX_FMT_YA16LE:             return AV_PIX_FMT_GRAY16;

  1136. 

  1137.     case AV_PIX_FMT_YUVA420P9BE:        return AV_PIX_FMT_YUV420P9;

  1138.     case AV_PIX_FMT_YUVA422P9BE:        return AV_PIX_FMT_YUV422P9;

  1139.     case AV_PIX_FMT_YUVA444P9BE:        return AV_PIX_FMT_YUV444P9;

  1140.     case AV_PIX_FMT_YUVA420P9LE:        return AV_PIX_FMT_YUV420P9;

  1141.     case AV_PIX_FMT_YUVA422P9LE:        return AV_PIX_FMT_YUV422P9;

  1142.     case AV_PIX_FMT_YUVA444P9LE:        return AV_PIX_FMT_YUV444P9;

  1143.     case AV_PIX_FMT_YUVA420P10BE:       return AV_PIX_FMT_YUV420P10;

  1144.     case AV_PIX_FMT_YUVA422P10BE:       return AV_PIX_FMT_YUV422P10;

  1145.     case AV_PIX_FMT_YUVA444P10BE:       return AV_PIX_FMT_YUV444P10;

  1146.     case AV_PIX_FMT_YUVA420P10LE:       return AV_PIX_FMT_YUV420P10;

  1147.     case AV_PIX_FMT_YUVA422P10LE:       return AV_PIX_FMT_YUV422P10;

  1148.     case AV_PIX_FMT_YUVA444P10LE:       return AV_PIX_FMT_YUV444P10;

  1149.     case AV_PIX_FMT_YUVA420P16BE:       return AV_PIX_FMT_YUV420P16;

  1150.     case AV_PIX_FMT_YUVA422P16BE:       return AV_PIX_FMT_YUV422P16;

  1151.     case AV_PIX_FMT_YUVA444P16BE:       return AV_PIX_FMT_YUV444P16;

  1152.     case AV_PIX_FMT_YUVA420P16LE:       return AV_PIX_FMT_YUV420P16;

  1153.     case AV_PIX_FMT_YUVA422P16LE:       return AV_PIX_FMT_YUV422P16;

  1154.     case AV_PIX_FMT_YUVA444P16LE:       return AV_PIX_FMT_YUV444P16;

  1155. 

  1156. //     case AV_PIX_FMT_AYUV64LE:

  1157. //     case AV_PIX_FMT_AYUV64BE:

  1158. //     case AV_PIX_FMT_PAL8:

  1159.     default: return AV_PIX_FMT_NONE;

  1160.     }

  1161. }

  1162. 

  1163. av_cold int sws_init_context(SwsContext *c, SwsFilter *srcFilter,

  1164.                              SwsFilter *dstFilter)

  1165. {

  1166.     int i;

  1167.     int usesVFilter, usesHFilter;

  1168.     int unscaled;

  1169.     SwsFilter dummyFilter = { NULL, NULL, NULL, NULL };

  1170.     int srcW              = c->srcW;

  1171.     int srcH              = c->srcH;

  1172.     int dstW              = c->dstW;

  1173.     int dstH              = c->dstH;

  1174.     int dst_stride        = FFALIGN(dstW * sizeof(int16_t) + 66, 16);

  1175.     int flags, cpu_flags;

  1176.     enum AVPixelFormat srcFormat = c->srcFormat;

  1177.     enum AVPixelFormat dstFormat = c->dstFormat;

  1178.     const AVPixFmtDescriptor *desc_src;

  1179.     const AVPixFmtDescriptor *desc_dst;

  1180.     int ret = 0;

  1181.     enum AVPixelFormat tmpFmt;

  1182.     static const float float_mult = 1.0f / 255.0f;

  1183. 

  1184.     cpu_flags = av_get_cpu_flags();

  1185.     flags     = c->flags;

  1186.     emms_c();

  1187.     if (!rgb15to16)

  1188.         ff_sws_rgb2rgb_init();

  1189. 

  1190.     unscaled = (srcW == dstW && srcH == dstH);

  1191. 

  1192.     c->srcRange |= handle_jpeg(&c->srcFormat);

  1193.     c->dstRange |= handle_jpeg(&c->dstFormat);

  1194. 

  1195.     if(srcFormat!=c->srcFormat || dstFormat!=c->dstFormat)

  1196.         av_log(c, AV_LOG_WARNING, "deprecated pixel format used, make sure you did set range correctly\n");

  1197. 

  1198.     if (!c->contrast && !c->saturation && !c->dstFormatBpp)

  1199.         sws_setColorspaceDetails(c, ff_yuv2rgb_coeffs[SWS_CS_DEFAULT], c->srcRange,

  1200.                                  ff_yuv2rgb_coeffs[SWS_CS_DEFAULT],

  1201.                                  c->dstRange, 0, 1 << 16, 1 << 16);

  1202. 

  1203.     handle_formats(c);

  1204.     srcFormat = c->srcFormat;

  1205.     dstFormat = c->dstFormat;

  1206.     desc_src = av_pix_fmt_desc_get(srcFormat);

  1207.     desc_dst = av_pix_fmt_desc_get(dstFormat);

  1208. 

  1209.     // If the source has no alpha then disable alpha blendaway

  1210.     if (c->src0Alpha)

  1211.         c->alphablend = SWS_ALPHA_BLEND_NONE;

  1212. 

  1213.     if (!(unscaled && sws_isSupportedEndiannessConversion(srcFormat) &&

  1214.           av_pix_fmt_swap_endianness(srcFormat) == dstFormat)) {

  1215.     if (!sws_isSupportedInput(srcFormat)) {

  1216.         av_log(c, AV_LOG_ERROR, "%s is not supported as input pixel format\n",

  1217.                av_get_pix_fmt_name(srcFormat));

  1218.         return AVERROR(EINVAL);

  1219.     }

  1220.     if (!sws_isSupportedOutput(dstFormat)) {

  1221.         av_log(c, AV_LOG_ERROR, "%s is not supported as output pixel format\n",

  1222.                av_get_pix_fmt_name(dstFormat));

  1223.         return AVERROR(EINVAL);

  1224.     }

  1225.     }

  1226.     av_assert2(desc_src && desc_dst);

  1227. 

  1228.     i = flags & (SWS_POINT         |

  1229.                  SWS_AREA          |

  1230.                  SWS_BILINEAR      |

  1231.                  SWS_FAST_BILINEAR |

  1232.                  SWS_BICUBIC       |

  1233.                  SWS_X             |

  1234.                  SWS_GAUSS         |

  1235.                  SWS_LANCZOS       |

  1236.                  SWS_SINC          |

  1237.                  SWS_SPLINE        |

  1238.                  SWS_BICUBLIN);

  1239. 

  1240.     /* provide a default scaler if not set by caller */

  1241.     if (!i) {

  1242.         if (dstW < srcW && dstH < srcH)

  1243.             flags |= SWS_BICUBIC;

  1244.         else if (dstW > srcW && dstH > srcH)

  1245.             flags |= SWS_BICUBIC;

  1246.         else

  1247.             flags |= SWS_BICUBIC;

  1248.         c->flags = flags;

  1249.     } else if (i & (i - 1)) {

  1250.         av_log(c, AV_LOG_ERROR,

  1251.                "Exactly one scaler algorithm must be chosen, got %X\n", i);

  1252.         return AVERROR(EINVAL);

  1253.     }

  1254.     /* sanity check */

  1255.     if (srcW < 1 || srcH < 1 || dstW < 1 || dstH < 1) {

  1256.         /* FIXME check if these are enough and try to lower them after

  1257.          * fixing the relevant parts of the code */

  1258.         av_log(c, AV_LOG_ERROR, "%dx%d -> %dx%d is invalid scaling dimension\n",

  1259.                srcW, srcH, dstW, dstH);

  1260.         return AVERROR(EINVAL);

  1261.     }

  1262.     if (flags & SWS_FAST_BILINEAR) {

  1263.         if (srcW < 8 || dstW < 8) {

  1264.             flags ^= SWS_FAST_BILINEAR | SWS_BILINEAR;

  1265.             c->flags = flags;

  1266.         }

  1267.     }

  1268. 

  1269.     if (!dstFilter)

  1270.         dstFilter = &dummyFilter;

  1271.     if (!srcFilter)

  1272.         srcFilter = &dummyFilter;

  1273. 

  1274.     c->lumXInc      = (((int64_t)srcW << 16) + (dstW >> 1)) / dstW;

  1275.     c->lumYInc      = (((int64_t)srcH << 16) + (dstH >> 1)) / dstH;

  1276.     c->dstFormatBpp = av_get_bits_per_pixel(desc_dst);

  1277.     c->srcFormatBpp = av_get_bits_per_pixel(desc_src);

  1278.     c->vRounder     = 4 * 0x0001000100010001ULL;

  1279. 

  1280.     usesVFilter = (srcFilter->lumV && srcFilter->lumV->length > 1) ||

  1281.                   (srcFilter->chrV && srcFilter->chrV->length > 1) ||

  1282.                   (dstFilter->lumV && dstFilter->lumV->length > 1) ||

  1283.                   (dstFilter->chrV && dstFilter->chrV->length > 1);

  1284.     usesHFilter = (srcFilter->lumH && srcFilter->lumH->length > 1) ||

  1285.                   (srcFilter->chrH && srcFilter->chrH->length > 1) ||

  1286.                   (dstFilter->lumH && dstFilter->lumH->length > 1) ||

  1287.                   (dstFilter->chrH && dstFilter->chrH->length > 1);

  1288. 

  1289.     av_pix_fmt_get_chroma_sub_sample(srcFormat, &c->chrSrcHSubSample, &c->chrSrcVSubSample);

  1290.     av_pix_fmt_get_chroma_sub_sample(dstFormat, &c->chrDstHSubSample, &c->chrDstVSubSample);

  1291. 

  1292.     if (isAnyRGB(dstFormat) && !(flags&SWS_FULL_CHR_H_INT)) {

  1293.         if (dstW&1) {

  1294.             av_log(c, AV_LOG_DEBUG, "Forcing full internal H chroma due to odd output size\n");

  1295.             flags |= SWS_FULL_CHR_H_INT;

  1296.             c->flags = flags;

  1297.         }

  1298. 

  1299.         if (   c->chrSrcHSubSample == 0

  1300.             && c->chrSrcVSubSample == 0

  1301.             && c->dither != SWS_DITHER_BAYER //SWS_FULL_CHR_H_INT is currently not supported with SWS_DITHER_BAYER

  1302.             && !(c->flags & SWS_FAST_BILINEAR)

  1303.         ) {

  1304.             av_log(c, AV_LOG_DEBUG, "Forcing full internal H chroma due to input having non subsampled chroma\n");

  1305.             flags |= SWS_FULL_CHR_H_INT;

  1306.             c->flags = flags;

  1307.         }

  1308.     }

  1309. 

  1310.     if (c->dither == SWS_DITHER_AUTO) {

  1311.         if (flags & SWS_ERROR_DIFFUSION)

  1312.             c->dither = SWS_DITHER_ED;

  1313.     }

  1314. 

  1315.     if(dstFormat == AV_PIX_FMT_BGR4_BYTE ||

  1316.        dstFormat == AV_PIX_FMT_RGB4_BYTE ||

  1317.        dstFormat == AV_PIX_FMT_BGR8 ||

  1318.        dstFormat == AV_PIX_FMT_RGB8) {

  1319.         if (c->dither == SWS_DITHER_AUTO)

  1320.             c->dither = (flags & SWS_FULL_CHR_H_INT) ? SWS_DITHER_ED : SWS_DITHER_BAYER;

  1321.         if (!(flags & SWS_FULL_CHR_H_INT)) {

  1322.             if (c->dither == SWS_DITHER_ED || c->dither == SWS_DITHER_A_DITHER || c->dither == SWS_DITHER_X_DITHER) {

  1323.                 av_log(c, AV_LOG_DEBUG,

  1324.                     "Desired dithering only supported in full chroma interpolation for destination format '%s'\n",

  1325.                     av_get_pix_fmt_name(dstFormat));

  1326.                 flags   |= SWS_FULL_CHR_H_INT;

  1327.                 c->flags = flags;

  1328.             }

  1329.         }

  1330.         if (flags & SWS_FULL_CHR_H_INT) {

  1331.             if (c->dither == SWS_DITHER_BAYER) {

  1332.                 av_log(c, AV_LOG_DEBUG,

  1333.                     "Ordered dither is not supported in full chroma interpolation for destination format '%s'\n",

  1334.                     av_get_pix_fmt_name(dstFormat));

  1335.                 c->dither = SWS_DITHER_ED;

  1336.             }

  1337.         }

  1338.     }

  1339.     if (isPlanarRGB(dstFormat)) {

  1340.         if (!(flags & SWS_FULL_CHR_H_INT)) {

  1341.             av_log(c, AV_LOG_DEBUG,

  1342.                    "%s output is not supported with half chroma resolution, switching to full\n",

  1343.                    av_get_pix_fmt_name(dstFormat));

  1344.             flags   |= SWS_FULL_CHR_H_INT;

  1345.             c->flags = flags;

  1346.         }

  1347.     }

  1348. 

  1349.     /* reuse chroma for 2 pixels RGB/BGR unless user wants full

  1350.      * chroma interpolation */

  1351.     if (flags & SWS_FULL_CHR_H_INT &&

  1352.         isAnyRGB(dstFormat)        &&

  1353.         !isPlanarRGB(dstFormat)    &&

  1354.         dstFormat != AV_PIX_FMT_RGBA64LE &&

  1355.         dstFormat != AV_PIX_FMT_RGBA64BE &&

  1356.         dstFormat != AV_PIX_FMT_BGRA64LE &&

  1357.         dstFormat != AV_PIX_FMT_BGRA64BE &&

  1358.         dstFormat != AV_PIX_FMT_RGB48LE &&

  1359.         dstFormat != AV_PIX_FMT_RGB48BE &&

  1360.         dstFormat != AV_PIX_FMT_BGR48LE &&

  1361.         dstFormat != AV_PIX_FMT_BGR48BE &&

  1362.         dstFormat != AV_PIX_FMT_RGBA  &&

  1363.         dstFormat != AV_PIX_FMT_ARGB  &&

  1364.         dstFormat != AV_PIX_FMT_BGRA  &&

  1365.         dstFormat != AV_PIX_FMT_ABGR  &&

  1366.         dstFormat != AV_PIX_FMT_RGB24 &&

  1367.         dstFormat != AV_PIX_FMT_BGR24 &&

  1368.         dstFormat != AV_PIX_FMT_BGR4_BYTE &&

  1369.         dstFormat != AV_PIX_FMT_RGB4_BYTE &&

  1370.         dstFormat != AV_PIX_FMT_BGR8 &&

  1371.         dstFormat != AV_PIX_FMT_RGB8

  1372.     ) {

  1373.         av_log(c, AV_LOG_WARNING,

  1374.                "full chroma interpolation for destination format '%s' not yet implemented\n",

  1375.                av_get_pix_fmt_name(dstFormat));

  1376.         flags   &= ~SWS_FULL_CHR_H_INT;

  1377.         c->flags = flags;

  1378.     }

  1379.     if (isAnyRGB(dstFormat) && !(flags & SWS_FULL_CHR_H_INT))

  1380.         c->chrDstHSubSample = 1;

  1381. 

  1382.     // drop some chroma lines if the user wants it

  1383.     c->vChrDrop          = (flags & SWS_SRC_V_CHR_DROP_MASK) >>

  1384.                            SWS_SRC_V_CHR_DROP_SHIFT;

  1385.     c->chrSrcVSubSample += c->vChrDrop;

  1386. 

  1387.     /* drop every other pixel for chroma calculation unless user

  1388.      * wants full chroma */

  1389.     if (isAnyRGB(srcFormat) && !(flags & SWS_FULL_CHR_H_INP)   &&

  1390.         srcFormat != AV_PIX_FMT_RGB8 && srcFormat != AV_PIX_FMT_BGR8 &&

  1391.         srcFormat != AV_PIX_FMT_RGB4 && srcFormat != AV_PIX_FMT_BGR4 &&

  1392.         srcFormat != AV_PIX_FMT_RGB4_BYTE && srcFormat != AV_PIX_FMT_BGR4_BYTE &&

  1393.         srcFormat != AV_PIX_FMT_GBRP9BE   && srcFormat != AV_PIX_FMT_GBRP9LE  &&

  1394.         srcFormat != AV_PIX_FMT_GBRP10BE  && srcFormat != AV_PIX_FMT_GBRP10LE &&

  1395.         srcFormat != AV_PIX_FMT_GBRAP10BE && srcFormat != AV_PIX_FMT_GBRAP10LE &&

  1396.         srcFormat != AV_PIX_FMT_GBRP12BE  && srcFormat != AV_PIX_FMT_GBRP12LE &&

  1397.         srcFormat != AV_PIX_FMT_GBRAP12BE && srcFormat != AV_PIX_FMT_GBRAP12LE &&

  1398.         srcFormat != AV_PIX_FMT_GBRP14BE  && srcFormat != AV_PIX_FMT_GBRP14LE &&

  1399.         srcFormat != AV_PIX_FMT_GBRP16BE  && srcFormat != AV_PIX_FMT_GBRP16LE &&

  1400.         srcFormat != AV_PIX_FMT_GBRAP16BE  && srcFormat != AV_PIX_FMT_GBRAP16LE &&

  1401.         ((dstW >> c->chrDstHSubSample) <= (srcW >> 1) ||

  1402.          (flags & SWS_FAST_BILINEAR)))

  1403.         c->chrSrcHSubSample = 1;

  1404. 

  1405.     // Note the AV_CEIL_RSHIFT is so that we always round toward +inf.

  1406.     c->chrSrcW = AV_CEIL_RSHIFT(srcW, c->chrSrcHSubSample);

  1407.     c->chrSrcH = AV_CEIL_RSHIFT(srcH, c->chrSrcVSubSample);

  1408.     c->chrDstW = AV_CEIL_RSHIFT(dstW, c->chrDstHSubSample);

  1409.     c->chrDstH = AV_CEIL_RSHIFT(dstH, c->chrDstVSubSample);

  1410. 

  1411.     FF_ALLOCZ_OR_GOTO(c, c->formatConvBuffer, FFALIGN(srcW*2+78, 16) * 2, fail);

  1412. 

  1413.     c->srcBpc = desc_src->comp[0].depth;

  1414.     if (c->srcBpc < 8)

  1415.         c->srcBpc = 8;

  1416.     c->dstBpc = desc_dst->comp[0].depth;

  1417.     if (c->dstBpc < 8)

  1418.         c->dstBpc = 8;

  1419.     if (isAnyRGB(srcFormat) || srcFormat == AV_PIX_FMT_PAL8)

  1420.         c->srcBpc = 16;

  1421.     if (c->dstBpc == 16)

  1422.         dst_stride <<= 1;

  1423. 

  1424.     if (INLINE_MMXEXT(cpu_flags) && c->srcBpc == 8 && c->dstBpc <= 14) {

  1425.         c->canMMXEXTBeUsed = dstW >= srcW && (dstW & 31) == 0 &&

  1426.                              c->chrDstW >= c->chrSrcW &&

  1427.                              (srcW & 15) == 0;

  1428.         if (!c->canMMXEXTBeUsed && dstW >= srcW && c->chrDstW >= c->chrSrcW && (srcW & 15) == 0

  1429. 

  1430.             && (flags & SWS_FAST_BILINEAR)) {

  1431.             if (flags & SWS_PRINT_INFO)

  1432.                 av_log(c, AV_LOG_INFO,

  1433.                        "output width is not a multiple of 32 -> no MMXEXT scaler\n");

  1434.         }

  1435.         if (usesHFilter || isNBPS(c->srcFormat) || is16BPS(c->srcFormat) || isAnyRGB(c->srcFormat))

  1436.             c->canMMXEXTBeUsed = 0;

  1437.     } else

  1438.         c->canMMXEXTBeUsed = 0;

  1439. 

  1440.     c->chrXInc = (((int64_t)c->chrSrcW << 16) + (c->chrDstW >> 1)) / c->chrDstW;

  1441.     c->chrYInc = (((int64_t)c->chrSrcH << 16) + (c->chrDstH >> 1)) / c->chrDstH;

  1442. 

  1443.     /* Match pixel 0 of the src to pixel 0 of dst and match pixel n-2 of src

  1444.      * to pixel n-2 of dst, but only for the FAST_BILINEAR mode otherwise do

  1445.      * correct scaling.

  1446.      * n-2 is the last chrominance sample available.

  1447.      * This is not perfect, but no one should notice the difference, the more

  1448.      * correct variant would be like the vertical one, but that would require

  1449.      * some special code for the first and last pixel */

  1450.     if (flags & SWS_FAST_BILINEAR) {

  1451.         if (c->canMMXEXTBeUsed) {

  1452.             c->lumXInc += 20;

  1453.             c->chrXInc += 20;

  1454.         }

  1455.         // we don't use the x86 asm scaler if MMX is available

  1456.         else if (INLINE_MMX(cpu_flags) && c->dstBpc <= 14) {

  1457.             c->lumXInc = ((int64_t)(srcW       - 2) << 16) / (dstW       - 2) - 20;

  1458.             c->chrXInc = ((int64_t)(c->chrSrcW - 2) << 16) / (c->chrDstW - 2) - 20;

  1459.         }

  1460.     }

  1461. 

  1462.     // hardcoded for now

  1463.     c->gamma_value = 2.2;

  1464.     tmpFmt = AV_PIX_FMT_RGBA64LE;

  1465. 

  1466. 

  1467.     if (!unscaled && c->gamma_flag && (srcFormat != tmpFmt || dstFormat != tmpFmt)) {

  1468.         SwsContext *c2;

  1469.         c->cascaded_context[0] = NULL;

  1470. 

  1471.         ret = av_image_alloc(c->cascaded_tmp, c->cascaded_tmpStride,

  1472.                             srcW, srcH, tmpFmt, 64);

  1473.         if (ret < 0)

  1474.             return ret;

  1475. 

  1476.         c->cascaded_context[0] = sws_getContext(srcW, srcH, srcFormat,

  1477.                                                 srcW, srcH, tmpFmt,

  1478.                                                 flags, NULL, NULL, c->param);

  1479.         if (!c->cascaded_context[0]) {

  1480.             return -1;

  1481.         }

  1482. 

  1483.         c->cascaded_context[1] = sws_getContext(srcW, srcH, tmpFmt,

  1484.                                                 dstW, dstH, tmpFmt,

  1485.                                                 flags, srcFilter, dstFilter, c->param);

  1486. 

  1487.         if (!c->cascaded_context[1])

  1488.             return -1;

  1489. 

  1490.         c2 = c->cascaded_context[1];

  1491.         c2->is_internal_gamma = 1;

  1492.         c2->gamma     = alloc_gamma_tbl(    c->gamma_value);

  1493.         c2->inv_gamma = alloc_gamma_tbl(1.f/c->gamma_value);

  1494.         if (!c2->gamma || !c2->inv_gamma)

  1495.             return AVERROR(ENOMEM);

  1496. 

  1497.         // is_internal_flag is set after creating the context

  1498.         // to properly create the gamma convert FilterDescriptor

  1499.         // we have to re-initialize it

  1500.         ff_free_filters(c2);

  1501.         if (ff_init_filters(c2) < 0) {

  1502.             sws_freeContext(c2);

  1503.             return -1;

  1504.         }

  1505. 

  1506.         c->cascaded_context[2] = NULL;

  1507.         if (dstFormat != tmpFmt) {

  1508.             ret = av_image_alloc(c->cascaded1_tmp, c->cascaded1_tmpStride,

  1509.                                 dstW, dstH, tmpFmt, 64);

  1510.             if (ret < 0)

  1511.                 return ret;

  1512. 

  1513.             c->cascaded_context[2] = sws_getContext(dstW, dstH, tmpFmt,

  1514.                                                 dstW, dstH, dstFormat,

  1515.                                                 flags, NULL, NULL, c->param);

  1516.             if (!c->cascaded_context[2])

  1517.                 return -1;

  1518.         }

  1519.         return 0;

  1520.     }

  1521. 

  1522.     if (isBayer(srcFormat)) {

  1523.         if (!unscaled ||

  1524.             (dstFormat != AV_PIX_FMT_RGB24 && dstFormat != AV_PIX_FMT_YUV420P)) {

  1525.             enum AVPixelFormat tmpFormat = AV_PIX_FMT_RGB24;

  1526. 

  1527.             ret = av_image_alloc(c->cascaded_tmp, c->cascaded_tmpStride,

  1528.                                 srcW, srcH, tmpFormat, 64);

  1529.             if (ret < 0)

  1530.                 return ret;

  1531. 

  1532.             c->cascaded_context[0] = sws_getContext(srcW, srcH, srcFormat,

  1533.                                                     srcW, srcH, tmpFormat,

  1534.                                                     flags, srcFilter, NULL, c->param);

  1535.             if (!c->cascaded_context[0])

  1536.                 return -1;

  1537. 

  1538.             c->cascaded_context[1] = sws_getContext(srcW, srcH, tmpFormat,

  1539.                                                     dstW, dstH, dstFormat,

  1540.                                                     flags, NULL, dstFilter, c->param);

  1541.             if (!c->cascaded_context[1])

  1542.                 return -1;

  1543.             return 0;

  1544.         }

  1545.     }

  1546. 

  1547.     if (unscaled && c->srcBpc == 8 && dstFormat == AV_PIX_FMT_GRAYF32){

  1548.         for (i = 0; i < 256; ++i){

  1549.             c->uint2float_lut[i] = (float)i * float_mult;

  1550.         }

  1551.     }

  1552. 

  1553.     // float will be converted to uint16_t

  1554.     if ((srcFormat == AV_PIX_FMT_GRAYF32BE || srcFormat == AV_PIX_FMT_GRAYF32LE) &&

  1555.         (!unscaled || unscaled && dstFormat != srcFormat && (srcFormat != AV_PIX_FMT_GRAYF32 ||

  1556.         dstFormat != AV_PIX_FMT_GRAY8))){

  1557.         c->srcBpc = 16;

  1558.     }

  1559. 

  1560.     if (CONFIG_SWSCALE_ALPHA && isALPHA(srcFormat) && !isALPHA(dstFormat)) {

  1561.         enum AVPixelFormat tmpFormat = alphaless_fmt(srcFormat);

  1562. 

  1563.         if (tmpFormat != AV_PIX_FMT_NONE && c->alphablend != SWS_ALPHA_BLEND_NONE)

  1564.         if (!unscaled ||

  1565.             dstFormat != tmpFormat ||

  1566.             usesHFilter || usesVFilter ||

  1567.             c->srcRange != c->dstRange

  1568.         ) {

  1569.             c->cascaded_mainindex = 1;

  1570.             ret = av_image_alloc(c->cascaded_tmp, c->cascaded_tmpStride,

  1571.                                 srcW, srcH, tmpFormat, 64);

  1572.             if (ret < 0)

  1573.                 return ret;

  1574. 

  1575.             c->cascaded_context[0] = sws_alloc_set_opts(srcW, srcH, srcFormat,

  1576.                                                         srcW, srcH, tmpFormat,

  1577.                                                         flags, c->param);

  1578.             if (!c->cascaded_context[0])

  1579.                 return -1;

  1580.             c->cascaded_context[0]->alphablend = c->alphablend;

  1581.             ret = sws_init_context(c->cascaded_context[0], NULL , NULL);

  1582.             if (ret < 0)

  1583.                 return ret;

  1584. 

  1585.             c->cascaded_context[1] = sws_alloc_set_opts(srcW, srcH, tmpFormat,

  1586.                                                         dstW, dstH, dstFormat,

  1587.                                                         flags, c->param);

  1588.             if (!c->cascaded_context[1])

  1589.                 return -1;

  1590. 

  1591.             c->cascaded_context[1]->srcRange = c->srcRange;

  1592.             c->cascaded_context[1]->dstRange = c->dstRange;

  1593.             ret = sws_init_context(c->cascaded_context[1], srcFilter , dstFilter);

  1594.             if (ret < 0)

  1595.                 return ret;

  1596. 

  1597.             return 0;

  1598.         }

  1599.     }

  1600. 

  1601. #if HAVE_MMAP && HAVE_MPROTECT && defined(MAP_ANONYMOUS)

  1602. #define USE_MMAP 1

  1603. #else

  1604. #define USE_MMAP 0

  1605. #endif

  1606. 

  1607.     /* precalculate horizontal scaler filter coefficients */

  1608.     {

  1609. #if HAVE_MMXEXT_INLINE

  1610. // can't downscale !!!

  1611.         if (c->canMMXEXTBeUsed && (flags & SWS_FAST_BILINEAR)) {

  1612.             c->lumMmxextFilterCodeSize = ff_init_hscaler_mmxext(dstW, c->lumXInc, NULL,

  1613.                                                              NULL, NULL, 8);

  1614.             c->chrMmxextFilterCodeSize = ff_init_hscaler_mmxext(c->chrDstW, c->chrXInc,

  1615.                                                              NULL, NULL, NULL, 4);

  1616. 

  1617. #if USE_MMAP

  1618.             c->lumMmxextFilterCode = mmap(NULL, c->lumMmxextFilterCodeSize,

  1619.                                           PROT_READ | PROT_WRITE,

  1620.                                           MAP_PRIVATE | MAP_ANONYMOUS,

  1621.                                           -1, 0);

  1622.             c->chrMmxextFilterCode = mmap(NULL, c->chrMmxextFilterCodeSize,

  1623.                                           PROT_READ | PROT_WRITE,

  1624.                                           MAP_PRIVATE | MAP_ANONYMOUS,

  1625.                                           -1, 0);

  1626. #elif HAVE_VIRTUALALLOC

  1627.             c->lumMmxextFilterCode = VirtualAlloc(NULL,

  1628.                                                   c->lumMmxextFilterCodeSize,

  1629.                                                   MEM_COMMIT,

  1630.                                                   PAGE_EXECUTE_READWRITE);

  1631.             c->chrMmxextFilterCode = VirtualAlloc(NULL,

  1632.                                                   c->chrMmxextFilterCodeSize,

  1633.                                                   MEM_COMMIT,

  1634.                                                   PAGE_EXECUTE_READWRITE);

  1635. #else

  1636.             c->lumMmxextFilterCode = av_malloc(c->lumMmxextFilterCodeSize);

  1637.             c->chrMmxextFilterCode = av_malloc(c->chrMmxextFilterCodeSize);

  1638. #endif

  1639. 

  1640. #ifdef MAP_ANONYMOUS

  1641.             if (c->lumMmxextFilterCode == MAP_FAILED || c->chrMmxextFilterCode == MAP_FAILED)

  1642. #else

  1643.             if (!c->lumMmxextFilterCode || !c->chrMmxextFilterCode)

  1644. #endif

  1645.             {

  1646.                 av_log(c, AV_LOG_ERROR, "Failed to allocate MMX2FilterCode\n");

  1647.                 return AVERROR(ENOMEM);

  1648.             }

  1649. 

  1650.             FF_ALLOCZ_OR_GOTO(c, c->hLumFilter,    (dstW           / 8 + 8) * sizeof(int16_t), fail);

  1651.             FF_ALLOCZ_OR_GOTO(c, c->hChrFilter,    (c->chrDstW     / 4 + 8) * sizeof(int16_t), fail);

  1652.             FF_ALLOCZ_OR_GOTO(c, c->hLumFilterPos, (dstW       / 2 / 8 + 8) * sizeof(int32_t), fail);

  1653.             FF_ALLOCZ_OR_GOTO(c, c->hChrFilterPos, (c->chrDstW / 2 / 4 + 8) * sizeof(int32_t), fail);

  1654. 

  1655.             ff_init_hscaler_mmxext(      dstW, c->lumXInc, c->lumMmxextFilterCode,

  1656.                                 c->hLumFilter, (uint32_t*)c->hLumFilterPos, 8);

  1657.             ff_init_hscaler_mmxext(c->chrDstW, c->chrXInc, c->chrMmxextFilterCode,

  1658.                                 c->hChrFilter, (uint32_t*)c->hChrFilterPos, 4);

  1659. 

  1660. #if USE_MMAP

  1661.             if (   mprotect(c->lumMmxextFilterCode, c->lumMmxextFilterCodeSize, PROT_EXEC | PROT_READ) == -1

  1662.                 || mprotect(c->chrMmxextFilterCode, c->chrMmxextFilterCodeSize, PROT_EXEC | PROT_READ) == -1) {

  1663.                 av_log(c, AV_LOG_ERROR, "mprotect failed, cannot use fast bilinear scaler\n");

  1664.                 goto fail;

  1665.             }

  1666. #endif

  1667.         } else

  1668. #endif /* HAVE_MMXEXT_INLINE */

  1669.         {

  1670.             const int filterAlign = X86_MMX(cpu_flags)     ? 4 :

  1671.                                     PPC_ALTIVEC(cpu_flags) ? 8 :

  1672.                                     have_neon(cpu_flags)   ? 8 : 1;

  1673. 

  1674.             if ((ret = initFilter(&c->hLumFilter, &c->hLumFilterPos,

  1675.                            &c->hLumFilterSize, c->lumXInc,

  1676.                            srcW, dstW, filterAlign, 1 << 14,

  1677.                            (flags & SWS_BICUBLIN) ? (flags | SWS_BICUBIC) : flags,

  1678.                            cpu_flags, srcFilter->lumH, dstFilter->lumH,

  1679.                            c->param,

  1680.                            get_local_pos(c, 0, 0, 0),

  1681.                            get_local_pos(c, 0, 0, 0))) < 0)

  1682.                 goto fail;

  1683.             if ((ret = initFilter(&c->hChrFilter, &c->hChrFilterPos,

  1684.                            &c->hChrFilterSize, c->chrXInc,

  1685.                            c->chrSrcW, c->chrDstW, filterAlign, 1 << 14,

  1686.                            (flags & SWS_BICUBLIN) ? (flags | SWS_BILINEAR) : flags,

  1687.                            cpu_flags, srcFilter->chrH, dstFilter->chrH,

  1688.                            c->param,

  1689.                            get_local_pos(c, c->chrSrcHSubSample, c->src_h_chr_pos, 0),

  1690.                            get_local_pos(c, c->chrDstHSubSample, c->dst_h_chr_pos, 0))) < 0)

  1691.                 goto fail;

  1692.         }

  1693.     } // initialize horizontal stuff

  1694. 

  1695.     /* precalculate vertical scaler filter coefficients */

  1696.     {

  1697.         const int filterAlign = X86_MMX(cpu_flags)     ? 2 :

  1698.                                 PPC_ALTIVEC(cpu_flags) ? 8 :

  1699.                                 have_neon(cpu_flags)   ? 2 : 1;

  1700. 

  1701.         if ((ret = initFilter(&c->vLumFilter, &c->vLumFilterPos, &c->vLumFilterSize,

  1702.                        c->lumYInc, srcH, dstH, filterAlign, (1 << 12),

  1703.                        (flags & SWS_BICUBLIN) ? (flags | SWS_BICUBIC) : flags,

  1704.                        cpu_flags, srcFilter->lumV, dstFilter->lumV,

  1705.                        c->param,

  1706.                        get_local_pos(c, 0, 0, 1),

  1707.                        get_local_pos(c, 0, 0, 1))) < 0)

  1708.             goto fail;

  1709.         if ((ret = initFilter(&c->vChrFilter, &c->vChrFilterPos, &c->vChrFilterSize,

  1710.                        c->chrYInc, c->chrSrcH, c->chrDstH,

  1711.                        filterAlign, (1 << 12),

  1712.                        (flags & SWS_BICUBLIN) ? (flags | SWS_BILINEAR) : flags,

  1713.                        cpu_flags, srcFilter->chrV, dstFilter->chrV,

  1714.                        c->param,

  1715.                        get_local_pos(c, c->chrSrcVSubSample, c->src_v_chr_pos, 1),

  1716.                        get_local_pos(c, c->chrDstVSubSample, c->dst_v_chr_pos, 1))) < 0)

  1717. 

  1718.             goto fail;

  1719. 

  1720. #if HAVE_ALTIVEC

  1721.         FF_ALLOC_OR_GOTO(c, c->vYCoeffsBank, sizeof(vector signed short) * c->vLumFilterSize * c->dstH,    fail);

  1722.         FF_ALLOC_OR_GOTO(c, c->vCCoeffsBank, sizeof(vector signed short) * c->vChrFilterSize * c->chrDstH, fail);

  1723. 

  1724.         for (i = 0; i < c->vLumFilterSize * c->dstH; i++) {

  1725.             int j;

  1726.             short *p = (short *)&c->vYCoeffsBank[i];

  1727.             for (j = 0; j < 8; j++)

  1728.                 p[j] = c->vLumFilter[i];

  1729.         }

  1730. 

  1731.         for (i = 0; i < c->vChrFilterSize * c->chrDstH; i++) {

  1732.             int j;

  1733.             short *p = (short *)&c->vCCoeffsBank[i];

  1734.             for (j = 0; j < 8; j++)

  1735.                 p[j] = c->vChrFilter[i];

  1736.         }

  1737. #endif

  1738.     }

  1739. 

  1740.     for (i = 0; i < 4; i++)

  1741.         FF_ALLOCZ_OR_GOTO(c, c->dither_error[i], (c->dstW+2) * sizeof(int), fail);

  1742. 

  1743.     c->needAlpha = (CONFIG_SWSCALE_ALPHA && isALPHA(c->srcFormat) && isALPHA(c->dstFormat)) ? 1 : 0;

  1744. 

  1745.     // 64 / c->scalingBpp is the same as 16 / sizeof(scaling_intermediate)

  1746.     c->uv_off   = (dst_stride>>1) + 64 / (c->dstBpc &~ 7);

  1747.     c->uv_offx2 = dst_stride + 16;

  1748. 

  1749.     av_assert0(c->chrDstH <= dstH);

  1750. 

  1751.     if (flags & SWS_PRINT_INFO) {

  1752.         const char *scaler = NULL, *cpucaps;

  1753. 

  1754.         for (i = 0; i < FF_ARRAY_ELEMS(scale_algorithms); i++) {

  1755.             if (flags & scale_algorithms[i].flag) {

  1756.                 scaler = scale_algorithms[i].description;

  1757.                 break;

  1758.             }

  1759.         }

  1760.         if (!scaler)

  1761.             scaler =  "ehh flags invalid?!";

  1762.         av_log(c, AV_LOG_INFO, "%s scaler, from %s to %s%s ",

  1763.                scaler,

  1764.                av_get_pix_fmt_name(srcFormat),

  1765. #ifdef DITHER1XBPP

  1766.                dstFormat == AV_PIX_FMT_BGR555   || dstFormat == AV_PIX_FMT_BGR565   ||

  1767.                dstFormat == AV_PIX_FMT_RGB444BE || dstFormat == AV_PIX_FMT_RGB444LE ||

  1768.                dstFormat == AV_PIX_FMT_BGR444BE || dstFormat == AV_PIX_FMT_BGR444LE ?

  1769.                                                              "dithered " : "",

  1770. #else

  1771.                "",

  1772. #endif

  1773.                av_get_pix_fmt_name(dstFormat));

  1774. 

  1775.         if (INLINE_MMXEXT(cpu_flags))

  1776.             cpucaps = "MMXEXT";

  1777.         else if (INLINE_AMD3DNOW(cpu_flags))

  1778.             cpucaps = "3DNOW";

  1779.         else if (INLINE_MMX(cpu_flags))

  1780.             cpucaps = "MMX";

  1781.         else if (PPC_ALTIVEC(cpu_flags))

  1782.             cpucaps = "AltiVec";

  1783.         else

  1784.             cpucaps = "C";

  1785. 

  1786.         av_log(c, AV_LOG_INFO, "using %s\n", cpucaps);

  1787. 

  1788.         av_log(c, AV_LOG_VERBOSE, "%dx%d -> %dx%d\n", srcW, srcH, dstW, dstH);

  1789.         av_log(c, AV_LOG_DEBUG,

  1790.                "lum srcW=%d srcH=%d dstW=%d dstH=%d xInc=%d yInc=%d\n",

  1791.                c->srcW, c->srcH, c->dstW, c->dstH, c->lumXInc, c->lumYInc);

  1792.         av_log(c, AV_LOG_DEBUG,

  1793.                "chr srcW=%d srcH=%d dstW=%d dstH=%d xInc=%d yInc=%d\n",

  1794.                c->chrSrcW, c->chrSrcH, c->chrDstW, c->chrDstH,

  1795.                c->chrXInc, c->chrYInc);

  1796.     }

  1797. 

  1798.     /* alpha blend special case, note this has been split via cascaded contexts if its scaled */

  1799.     if (unscaled && !usesHFilter && !usesVFilter &&

  1800.         c->alphablend != SWS_ALPHA_BLEND_NONE &&

  1801.         isALPHA(srcFormat) &&

  1802.         (c->srcRange == c->dstRange || isAnyRGB(dstFormat)) &&

  1803.         alphaless_fmt(srcFormat) == dstFormat

  1804.     ) {

  1805.         c->swscale = ff_sws_alphablendaway;

  1806. 

  1807.         if (flags & SWS_PRINT_INFO)

  1808.             av_log(c, AV_LOG_INFO,

  1809.                     "using alpha blendaway %s -> %s special converter\n",

  1810.                     av_get_pix_fmt_name(srcFormat), av_get_pix_fmt_name(dstFormat));

  1811.         return 0;

  1812.     }

  1813. 

  1814.     /* unscaled special cases */

  1815.     if (unscaled && !usesHFilter && !usesVFilter &&

  1816.         (c->srcRange == c->dstRange || isAnyRGB(dstFormat) ||

  1817.          isFloat(srcFormat) || isFloat(dstFormat))){

  1818.         ff_get_unscaled_swscale(c);

  1819. 

  1820.         if (c->swscale) {

  1821.             if (flags & SWS_PRINT_INFO)

  1822.                 av_log(c, AV_LOG_INFO,

  1823.                        "using unscaled %s -> %s special converter\n",

  1824.                        av_get_pix_fmt_name(srcFormat), av_get_pix_fmt_name(dstFormat));

  1825.             return 0;

  1826.         }

  1827.     }

  1828. 

  1829.     c->swscale = ff_getSwsFunc(c);

  1830.     return ff_init_filters(c);

  1831. fail: // FIXME replace things by appropriate error codes

  1832.     if (ret == RETCODE_USE_CASCADE)  {

  1833.         int tmpW = sqrt(srcW * (int64_t)dstW);

  1834.         int tmpH = sqrt(srcH * (int64_t)dstH);

  1835.         enum AVPixelFormat tmpFormat = AV_PIX_FMT_YUV420P;

  1836. 

  1837.         if (isALPHA(srcFormat))

  1838.             tmpFormat = AV_PIX_FMT_YUVA420P;

  1839. 

  1840.         if (srcW*(int64_t)srcH <= 4LL*dstW*dstH)

  1841.             return AVERROR(EINVAL);

  1842. 

  1843.         ret = av_image_alloc(c->cascaded_tmp, c->cascaded_tmpStride,

  1844.                              tmpW, tmpH, tmpFormat, 64);

  1845.         if (ret < 0)

  1846.             return ret;

  1847. 

  1848.         c->cascaded_context[0] = sws_getContext(srcW, srcH, srcFormat,

  1849.                                                 tmpW, tmpH, tmpFormat,

  1850.                                                 flags, srcFilter, NULL, c->param);

  1851.         if (!c->cascaded_context[0])

  1852.             return -1;

  1853. 

  1854.         c->cascaded_context[1] = sws_getContext(tmpW, tmpH, tmpFormat,

  1855.                                                 dstW, dstH, dstFormat,

  1856.                                                 flags, NULL, dstFilter, c->param);

  1857.         if (!c->cascaded_context[1])

  1858.             return -1;

  1859.         return 0;

  1860.     }

  1861.     return -1;

  1862. }

  1863. 

  1864. SwsContext *sws_alloc_set_opts(int srcW, int srcH, enum AVPixelFormat srcFormat,

  1865.                                int dstW, int dstH, enum AVPixelFormat dstFormat,

  1866.                                int flags, const double *param)

  1867. {

  1868.     SwsContext *c;

  1869. 

  1870.     if (!(c = sws_alloc_context()))

  1871.         return NULL;

  1872. 

  1873.     c->flags     = flags;

  1874.     c->srcW      = srcW;

  1875.     c->srcH      = srcH;

  1876.     c->dstW      = dstW;

  1877.     c->dstH      = dstH;

  1878.     c->srcFormat = srcFormat;

  1879.     c->dstFormat = dstFormat;

  1880. 

  1881.     if (param) {

  1882.         c->param[0] = param[0];

  1883.         c->param[1] = param[1];

  1884.     }

  1885. 

  1886.     return c;

  1887. }

  1888. 

  1889. SwsContext *sws_getContext(int srcW, int srcH, enum AVPixelFormat srcFormat,

  1890.                            int dstW, int dstH, enum AVPixelFormat dstFormat,

  1891.                            int flags, SwsFilter *srcFilter,

  1892.                            SwsFilter *dstFilter, const double *param)

  1893. {

  1894.     SwsContext *c;

  1895. 

  1896.     c = sws_alloc_set_opts(srcW, srcH, srcFormat,

  1897.                            dstW, dstH, dstFormat,

  1898.                            flags, param);

  1899.     if (!c)

  1900.         return NULL;

  1901. 

  1902.     if (sws_init_context(c, srcFilter, dstFilter) < 0) {

  1903.         sws_freeContext(c);

  1904.         return NULL;

  1905.     }

  1906. 

  1907.     return c;

  1908. }

  1909. 

  1910. static int isnan_vec(SwsVector *a)

  1911. {

  1912.     int i;

  1913.     for (i=0; i<a->length; i++)

  1914.         if (isnan(a->coeff[i]))

  1915.             return 1;

  1916.     return 0;

  1917. }

  1918. 

  1919. static void makenan_vec(SwsVector *a)

  1920. {

  1921.     int i;

  1922.     for (i=0; i<a->length; i++)

  1923.         a->coeff[i] = NAN;

  1924. }

  1925. 

  1926. SwsFilter *sws_getDefaultFilter(float lumaGBlur, float chromaGBlur,

  1927.                                 float lumaSharpen, float chromaSharpen,

  1928.                                 float chromaHShift, float chromaVShift,

  1929.                                 int verbose)

  1930. {

  1931.     SwsFilter *filter = av_malloc(sizeof(SwsFilter));

  1932.     if (!filter)

  1933.         return NULL;

  1934. 

  1935.     if (lumaGBlur != 0.0) {

  1936.         filter->lumH = sws_getGaussianVec(lumaGBlur, 3.0);

  1937.         filter->lumV = sws_getGaussianVec(lumaGBlur, 3.0);

  1938.     } else {

  1939.         filter->lumH = sws_getIdentityVec();

  1940.         filter->lumV = sws_getIdentityVec();

  1941.     }

  1942. 

  1943.     if (chromaGBlur != 0.0) {

  1944.         filter->chrH = sws_getGaussianVec(chromaGBlur, 3.0);

  1945.         filter->chrV = sws_getGaussianVec(chromaGBlur, 3.0);

  1946.     } else {

  1947.         filter->chrH = sws_getIdentityVec();

  1948.         filter->chrV = sws_getIdentityVec();

  1949.     }

  1950. 

  1951.     if (!filter->lumH || !filter->lumV || !filter->chrH || !filter->chrV)

  1952.         goto fail;

  1953. 

  1954.     if (chromaSharpen != 0.0) {

  1955.         SwsVector *id = sws_getIdentityVec();

  1956.         if (!id)

  1957.             goto fail;

  1958.         sws_scaleVec(filter->chrH, -chromaSharpen);

  1959.         sws_scaleVec(filter->chrV, -chromaSharpen);

  1960.         sws_addVec(filter->chrH, id);

  1961.         sws_addVec(filter->chrV, id);

  1962.         sws_freeVec(id);

  1963.     }

  1964. 

  1965.     if (lumaSharpen != 0.0) {

  1966.         SwsVector *id = sws_getIdentityVec();

  1967.         if (!id)

  1968.             goto fail;

  1969.         sws_scaleVec(filter->lumH, -lumaSharpen);

  1970.         sws_scaleVec(filter->lumV, -lumaSharpen);

  1971.         sws_addVec(filter->lumH, id);

  1972.         sws_addVec(filter->lumV, id);

  1973.         sws_freeVec(id);

  1974.     }

  1975. 

  1976.     if (chromaHShift != 0.0)

  1977.         sws_shiftVec(filter->chrH, (int)(chromaHShift + 0.5));

  1978. 

  1979.     if (chromaVShift != 0.0)

  1980.         sws_shiftVec(filter->chrV, (int)(chromaVShift + 0.5));

  1981. 

  1982.     sws_normalizeVec(filter->chrH, 1.0);

  1983.     sws_normalizeVec(filter->chrV, 1.0);

  1984.     sws_normalizeVec(filter->lumH, 1.0);

  1985.     sws_normalizeVec(filter->lumV, 1.0);

  1986. 

  1987.     if (isnan_vec(filter->chrH) ||

  1988.         isnan_vec(filter->chrV) ||

  1989.         isnan_vec(filter->lumH) ||

  1990.         isnan_vec(filter->lumV))

  1991.         goto fail;

  1992. 

  1993.     if (verbose)

  1994.         sws_printVec2(filter->chrH, NULL, AV_LOG_DEBUG);

  1995.     if (verbose)

  1996.         sws_printVec2(filter->lumH, NULL, AV_LOG_DEBUG);

  1997. 

  1998.     return filter;

  1999. 

  2000. fail:

  2001.     sws_freeVec(filter->lumH);

  2002.     sws_freeVec(filter->lumV);

  2003.     sws_freeVec(filter->chrH);

  2004.     sws_freeVec(filter->chrV);

  2005.     av_freep(&filter);

  2006.     return NULL;

  2007. }

  2008. 

  2009. SwsVector *sws_allocVec(int length)

  2010. {

  2011.     SwsVector *vec;

  2012. 

  2013.     if(length <= 0 || length > INT_MAX/ sizeof(double))

  2014.         return NULL;

  2015. 

  2016.     vec = av_malloc(sizeof(SwsVector));

  2017.     if (!vec)

  2018.         return NULL;

  2019.     vec->length = length;

  2020.     vec->coeff  = av_malloc(sizeof(double) * length);

  2021.     if (!vec->coeff)

  2022.         av_freep(&vec);

  2023.     return vec;

  2024. }

  2025. 

  2026. SwsVector *sws_getGaussianVec(double variance, double quality)

  2027. {

  2028.     const int length = (int)(variance * quality + 0.5) | 1;

  2029.     int i;

  2030.     double middle  = (length - 1) * 0.5;

  2031.     SwsVector *vec;

  2032. 

  2033.     if(variance < 0 || quality < 0)

  2034.         return NULL;

  2035. 

  2036.     vec = sws_allocVec(length);

  2037. 

  2038.     if (!vec)

  2039.         return NULL;

  2040. 

  2041.     for (i = 0; i < length; i++) {

  2042.         double dist = i - middle;

  2043.         vec->coeff[i] = exp(-dist * dist / (2 * variance * variance)) /

  2044.                         sqrt(2 * variance * M_PI);

  2045.     }

  2046. 

  2047.     sws_normalizeVec(vec, 1.0);

  2048. 

  2049.     return vec;

  2050. }

  2051. 

  2052. /**

  2053.  * Allocate and return a vector with length coefficients, all

  2054.  * with the same value c.

  2055.  */

  2056. #if !FF_API_SWS_VECTOR

  2057. static

  2058. #endif

  2059. SwsVector *sws_getConstVec(double c, int length)

  2060. {

  2061.     int i;

  2062.     SwsVector *vec = sws_allocVec(length);

  2063. 

  2064.     if (!vec)

  2065.         return NULL;

  2066. 

  2067.     for (i = 0; i < length; i++)

  2068.         vec->coeff[i] = c;

  2069. 

  2070.     return vec;

  2071. }

  2072. 

  2073. /**

  2074.  * Allocate and return a vector with just one coefficient, with

  2075.  * value 1.0.

  2076.  */

  2077. #if !FF_API_SWS_VECTOR

  2078. static

  2079. #endif

  2080. SwsVector *sws_getIdentityVec(void)

  2081. {

  2082.     return sws_getConstVec(1.0, 1);

  2083. }

  2084. 

  2085. static double sws_dcVec(SwsVector *a)

  2086. {

  2087.     int i;

  2088.     double sum = 0;

  2089. 

  2090.     for (i = 0; i < a->length; i++)

  2091.         sum += a->coeff[i];

  2092. 

  2093.     return sum;

  2094. }

  2095. 

  2096. void sws_scaleVec(SwsVector *a, double scalar)

  2097. {

  2098.     int i;

  2099. 

  2100.     for (i = 0; i < a->length; i++)

  2101.         a->coeff[i] *= scalar;

  2102. }

  2103. 

  2104. void sws_normalizeVec(SwsVector *a, double height)

  2105. {

  2106.     sws_scaleVec(a, height / sws_dcVec(a));

  2107. }

  2108. 

  2109. #if FF_API_SWS_VECTOR

  2110. static SwsVector *sws_getConvVec(SwsVector *a, SwsVector *b)

  2111. {

  2112.     int length = a->length + b->length - 1;

  2113.     int i, j;

  2114.     SwsVector *vec = sws_getConstVec(0.0, length);

  2115. 

  2116.     if (!vec)

  2117.         return NULL;

  2118. 

  2119.     for (i = 0; i < a->length; i++) {

  2120.         for (j = 0; j < b->length; j++) {

  2121.             vec->coeff[i + j] += a->coeff[i] * b->coeff[j];

  2122.         }

  2123.     }

  2124. 

  2125.     return vec;

  2126. }

  2127. #endif

  2128. 

  2129. static SwsVector *sws_sumVec(SwsVector *a, SwsVector *b)

  2130. {

  2131.     int length = FFMAX(a->length, b->length);

  2132.     int i;

  2133.     SwsVector *vec = sws_getConstVec(0.0, length);

  2134. 

  2135.     if (!vec)

  2136.         return NULL;

  2137. 

  2138.     for (i = 0; i < a->length; i++)

  2139.         vec->coeff[i + (length - 1) / 2 - (a->length - 1) / 2] += a->coeff[i];

  2140.     for (i = 0; i < b->length; i++)

  2141.         vec->coeff[i + (length - 1) / 2 - (b->length - 1) / 2] += b->coeff[i];

  2142. 

  2143.     return vec;

  2144. }

  2145. 

  2146. #if FF_API_SWS_VECTOR

  2147. static SwsVector *sws_diffVec(SwsVector *a, SwsVector *b)

  2148. {

  2149.     int length = FFMAX(a->length, b->length);

  2150.     int i;

  2151.     SwsVector *vec = sws_getConstVec(0.0, length);

  2152. 

  2153.     if (!vec)

  2154.         return NULL;

  2155. 

  2156.     for (i = 0; i < a->length; i++)

  2157.         vec->coeff[i + (length - 1) / 2 - (a->length - 1) / 2] += a->coeff[i];

  2158.     for (i = 0; i < b->length; i++)

  2159.         vec->coeff[i + (length - 1) / 2 - (b->length - 1) / 2] -= b->coeff[i];

  2160. 

  2161.     return vec;

  2162. }

  2163. #endif

  2164. 

  2165. /* shift left / or right if "shift" is negative */

  2166. static SwsVector *sws_getShiftedVec(SwsVector *a, int shift)

  2167. {

  2168.     int length = a->length + FFABS(shift) * 2;

  2169.     int i;

  2170.     SwsVector *vec = sws_getConstVec(0.0, length);

  2171. 

  2172.     if (!vec)

  2173.         return NULL;

  2174. 

  2175.     for (i = 0; i < a->length; i++) {

  2176.         vec->coeff[i + (length    - 1) / 2 -

  2177.                        (a->length - 1) / 2 - shift] = a->coeff[i];

  2178.     }

  2179. 

  2180.     return vec;

  2181. }

  2182. 

  2183. #if !FF_API_SWS_VECTOR

  2184. static

  2185. #endif

  2186. void sws_shiftVec(SwsVector *a, int shift)

  2187. {

  2188.     SwsVector *shifted = sws_getShiftedVec(a, shift);

  2189.     if (!shifted) {

  2190.         makenan_vec(a);

  2191.         return;

  2192.     }

  2193.     av_free(a->coeff);

  2194.     a->coeff  = shifted->coeff;

  2195.     a->length = shifted->length;

  2196.     av_free(shifted);

  2197. }

  2198. 

  2199. #if !FF_API_SWS_VECTOR

  2200. static

  2201. #endif

  2202. void sws_addVec(SwsVector *a, SwsVector *b)

  2203. {

  2204.     SwsVector *sum = sws_sumVec(a, b);

  2205.     if (!sum) {

  2206.         makenan_vec(a);

  2207.         return;

  2208.     }

  2209.     av_free(a->coeff);

  2210.     a->coeff  = sum->coeff;

  2211.     a->length = sum->length;

  2212.     av_free(sum);

  2213. }

  2214. 

  2215. #if FF_API_SWS_VECTOR

  2216. void sws_subVec(SwsVector *a, SwsVector *b)

  2217. {

  2218.     SwsVector *diff = sws_diffVec(a, b);

  2219.     if (!diff) {

  2220.         makenan_vec(a);

  2221.         return;

  2222.     }

  2223.     av_free(a->coeff);

  2224.     a->coeff  = diff->coeff;

  2225.     a->length = diff->length;

  2226.     av_free(diff);

  2227. }

  2228. 

  2229. void sws_convVec(SwsVector *a, SwsVector *b)

  2230. {

  2231.     SwsVector *conv = sws_getConvVec(a, b);

  2232.     if (!conv) {

  2233.         makenan_vec(a);

  2234.         return;

  2235.     }

  2236.     av_free(a->coeff);

  2237.     a->coeff  = conv->coeff;

  2238.     a->length = conv->length;

  2239.     av_free(conv);

  2240. }

  2241. 

  2242. SwsVector *sws_cloneVec(SwsVector *a)

  2243. {

  2244.     SwsVector *vec = sws_allocVec(a->length);

  2245. 

  2246.     if (!vec)

  2247.         return NULL;

  2248. 

  2249.     memcpy(vec->coeff, a->coeff, a->length * sizeof(*a->coeff));

  2250. 

  2251.     return vec;

  2252. }

  2253. #endif

  2254. 

  2255. /**

  2256.  * Print with av_log() a textual representation of the vector a

  2257.  * if log_level <= av_log_level.

  2258.  */

  2259. #if !FF_API_SWS_VECTOR

  2260. static

  2261. #endif

  2262. void sws_printVec2(SwsVector *a, AVClass *log_ctx, int log_level)

  2263. {

  2264.     int i;

  2265.     double max = 0;

  2266.     double min = 0;

  2267.     double range;

  2268. 

  2269.     for (i = 0; i < a->length; i++)

  2270.         if (a->coeff[i] > max)

  2271.             max = a->coeff[i];

  2272. 

  2273.     for (i = 0; i < a->length; i++)

  2274.         if (a->coeff[i] < min)

  2275.             min = a->coeff[i];

  2276. 

  2277.     range = max - min;

  2278. 

  2279.     for (i = 0; i < a->length; i++) {

  2280.         int x = (int)((a->coeff[i] - min) * 60.0 / range + 0.5);

  2281.         av_log(log_ctx, log_level, "%1.3f ", a->coeff[i]);

  2282.         for (; x > 0; x--)

  2283.             av_log(log_ctx, log_level, " ");

  2284.         av_log(log_ctx, log_level, "|\n");

  2285.     }

  2286. }

  2287. 

  2288. void sws_freeVec(SwsVector *a)

  2289. {

  2290.     if (!a)

  2291.         return;

  2292.     av_freep(&a->coeff);

  2293.     a->length = 0;

  2294.     av_free(a);

  2295. }

  2296. 

  2297. void sws_freeFilter(SwsFilter *filter)

  2298. {

  2299.     if (!filter)

  2300.         return;

  2301. 

  2302.     sws_freeVec(filter->lumH);

  2303.     sws_freeVec(filter->lumV);

  2304.     sws_freeVec(filter->chrH);

  2305.     sws_freeVec(filter->chrV);

  2306.     av_free(filter);

  2307. }

  2308. 

  2309. void sws_freeContext(SwsContext *c)

  2310. {

  2311.     int i;

  2312.     if (!c)

  2313.         return;

  2314. 

  2315.     for (i = 0; i < 4; i++)

  2316.         av_freep(&c->dither_error[i]);

  2317. 

  2318.     av_freep(&c->vLumFilter);

  2319.     av_freep(&c->vChrFilter);

  2320.     av_freep(&c->hLumFilter);

  2321.     av_freep(&c->hChrFilter);

  2322. #if HAVE_ALTIVEC

  2323.     av_freep(&c->vYCoeffsBank);

  2324.     av_freep(&c->vCCoeffsBank);

  2325. #endif

  2326. 

  2327.     av_freep(&c->vLumFilterPos);

  2328.     av_freep(&c->vChrFilterPos);

  2329.     av_freep(&c->hLumFilterPos);

  2330.     av_freep(&c->hChrFilterPos);

  2331. 

  2332. #if HAVE_MMX_INLINE

  2333. #if USE_MMAP

  2334.     if (c->lumMmxextFilterCode)

  2335.         munmap(c->lumMmxextFilterCode, c->lumMmxextFilterCodeSize);

  2336.     if (c->chrMmxextFilterCode)

  2337.         munmap(c->chrMmxextFilterCode, c->chrMmxextFilterCodeSize);

  2338. #elif HAVE_VIRTUALALLOC

  2339.     if (c->lumMmxextFilterCode)

  2340.         VirtualFree(c->lumMmxextFilterCode, 0, MEM_RELEASE);

  2341.     if (c->chrMmxextFilterCode)

  2342.         VirtualFree(c->chrMmxextFilterCode, 0, MEM_RELEASE);

  2343. #else

  2344.     av_free(c->lumMmxextFilterCode);

  2345.     av_free(c->chrMmxextFilterCode);

  2346. #endif

  2347.     c->lumMmxextFilterCode = NULL;

  2348.     c->chrMmxextFilterCode = NULL;

  2349. #endif /* HAVE_MMX_INLINE */

  2350. 

  2351.     av_freep(&c->yuvTable);

  2352.     av_freep(&c->formatConvBuffer);

  2353. 

  2354.     sws_freeContext(c->cascaded_context[0]);

  2355.     sws_freeContext(c->cascaded_context[1]);

  2356.     sws_freeContext(c->cascaded_context[2]);

  2357.     memset(c->cascaded_context, 0, sizeof(c->cascaded_context));

  2358.     av_freep(&c->cascaded_tmp[0]);

  2359.     av_freep(&c->cascaded1_tmp[0]);

  2360. 

  2361.     av_freep(&c->gamma);

  2362.     av_freep(&c->inv_gamma);

  2363. 

  2364.     ff_free_filters(c);

  2365. 

  2366.     av_free(c);

  2367. }

  2368. 

  2369. struct SwsContext *sws_getCachedContext(struct SwsContext *context, int srcW,

  2370.                                         int srcH, enum AVPixelFormat srcFormat,

  2371.                                         int dstW, int dstH,

  2372.                                         enum AVPixelFormat dstFormat, int flags,

  2373.                                         SwsFilter *srcFilter,

  2374.                                         SwsFilter *dstFilter,

  2375.                                         const double *param)

  2376. {

  2377.     static const double default_param[2] = { SWS_PARAM_DEFAULT,

  2378.                                              SWS_PARAM_DEFAULT };

  2379.     int64_t src_h_chr_pos = -513, dst_h_chr_pos = -513,

  2380.             src_v_chr_pos = -513, dst_v_chr_pos = -513;

  2381. 

  2382.     if (!param)

  2383.         param = default_param;

  2384. 

  2385.     if (context &&

  2386.         (context->srcW      != srcW      ||

  2387.          context->srcH      != srcH      ||

  2388.          context->srcFormat != srcFormat ||

  2389.          context->dstW      != dstW      ||

  2390.          context->dstH      != dstH      ||

  2391.          context->dstFormat != dstFormat ||

  2392.          context->flags     != flags     ||

  2393.          context->param[0]  != param[0]  ||

  2394.          context->param[1]  != param[1])) {

  2395. 

  2396.         av_opt_get_int(context, "src_h_chr_pos", 0, &src_h_chr_pos);

  2397.         av_opt_get_int(context, "src_v_chr_pos", 0, &src_v_chr_pos);

  2398.         av_opt_get_int(context, "dst_h_chr_pos", 0, &dst_h_chr_pos);

  2399.         av_opt_get_int(context, "dst_v_chr_pos", 0, &dst_v_chr_pos);

  2400.         sws_freeContext(context);

  2401.         context = NULL;

  2402.     }

  2403. 

  2404.     if (!context) {

  2405.         if (!(context = sws_alloc_context()))

  2406.             return NULL;

  2407.         context->srcW      = srcW;

  2408.         context->srcH      = srcH;

  2409.         context->srcFormat = srcFormat;

  2410.         context->dstW      = dstW;

  2411.         context->dstH      = dstH;

  2412.         context->dstFormat = dstFormat;

  2413.         context->flags     = flags;

  2414.         context->param[0]  = param[0];

  2415.         context->param[1]  = param[1];

  2416. 

  2417.         av_opt_set_int(context, "src_h_chr_pos", src_h_chr_pos, 0);

  2418.         av_opt_set_int(context, "src_v_chr_pos", src_v_chr_pos, 0);

  2419.         av_opt_set_int(context, "dst_h_chr_pos", dst_h_chr_pos, 0);

  2420.         av_opt_set_int(context, "dst_v_chr_pos", dst_v_chr_pos, 0);

  2421. 

  2422.         if (sws_init_context(context, srcFilter, dstFilter) < 0) {

  2423.             sws_freeContext(context);

  2424.             return NULL;

  2425.         }

  2426.     }

  2427.     return context;

  2428. }