LCOV - code coverage report
Current view: top level - src/libswscale - utils.c (source / functions) Hit Total Coverage
Test: coverage.info Lines: 600 1183 50.7 %
Date: 2017-01-19 23:52:33 Functions: 19 47 40.4 %

          Line data    Source code
       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_SYS_MMAN_H
      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        5120 : unsigned swscale_version(void)
      76             : {
      77             :     av_assert0(LIBSWSCALE_VERSION_MICRO >= 100);
      78        5120 :     return LIBSWSCALE_VERSION_INT;
      79             : }
      80             : 
      81        5120 : const char *swscale_configuration(void)
      82             : {
      83        5120 :     return FFMPEG_CONFIGURATION;
      84             : }
      85             : 
      86           0 : const char *swscale_license(void)
      87             : {
      88             : #define LICENSE_PREFIX "libswscale license: "
      89           0 :     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_GRAY10BE]    = { 1, 1 },
     135             :     [AV_PIX_FMT_GRAY10LE]    = { 1, 1 },
     136             :     [AV_PIX_FMT_GRAY12BE]    = { 1, 1 },
     137             :     [AV_PIX_FMT_GRAY12LE]    = { 1, 1 },
     138             :     [AV_PIX_FMT_GRAY16BE]    = { 1, 1 },
     139             :     [AV_PIX_FMT_GRAY16LE]    = { 1, 1 },
     140             :     [AV_PIX_FMT_YUV440P]     = { 1, 1 },
     141             :     [AV_PIX_FMT_YUVJ440P]    = { 1, 1 },
     142             :     [AV_PIX_FMT_YUV440P10LE] = { 1, 1 },
     143             :     [AV_PIX_FMT_YUV440P10BE] = { 1, 1 },
     144             :     [AV_PIX_FMT_YUV440P12LE] = { 1, 1 },
     145             :     [AV_PIX_FMT_YUV440P12BE] = { 1, 1 },
     146             :     [AV_PIX_FMT_YUVA420P]    = { 1, 1 },
     147             :     [AV_PIX_FMT_YUVA422P]    = { 1, 1 },
     148             :     [AV_PIX_FMT_YUVA444P]    = { 1, 1 },
     149             :     [AV_PIX_FMT_YUVA420P9BE] = { 1, 1 },
     150             :     [AV_PIX_FMT_YUVA420P9LE] = { 1, 1 },
     151             :     [AV_PIX_FMT_YUVA422P9BE] = { 1, 1 },
     152             :     [AV_PIX_FMT_YUVA422P9LE] = { 1, 1 },
     153             :     [AV_PIX_FMT_YUVA444P9BE] = { 1, 1 },
     154             :     [AV_PIX_FMT_YUVA444P9LE] = { 1, 1 },
     155             :     [AV_PIX_FMT_YUVA420P10BE]= { 1, 1 },
     156             :     [AV_PIX_FMT_YUVA420P10LE]= { 1, 1 },
     157             :     [AV_PIX_FMT_YUVA422P10BE]= { 1, 1 },
     158             :     [AV_PIX_FMT_YUVA422P10LE]= { 1, 1 },
     159             :     [AV_PIX_FMT_YUVA444P10BE]= { 1, 1 },
     160             :     [AV_PIX_FMT_YUVA444P10LE]= { 1, 1 },
     161             :     [AV_PIX_FMT_YUVA420P16BE]= { 1, 1 },
     162             :     [AV_PIX_FMT_YUVA420P16LE]= { 1, 1 },
     163             :     [AV_PIX_FMT_YUVA422P16BE]= { 1, 1 },
     164             :     [AV_PIX_FMT_YUVA422P16LE]= { 1, 1 },
     165             :     [AV_PIX_FMT_YUVA444P16BE]= { 1, 1 },
     166             :     [AV_PIX_FMT_YUVA444P16LE]= { 1, 1 },
     167             :     [AV_PIX_FMT_RGB48BE]     = { 1, 1 },
     168             :     [AV_PIX_FMT_RGB48LE]     = { 1, 1 },
     169             :     [AV_PIX_FMT_RGBA64BE]    = { 1, 1, 1 },
     170             :     [AV_PIX_FMT_RGBA64LE]    = { 1, 1, 1 },
     171             :     [AV_PIX_FMT_RGB565BE]    = { 1, 1 },
     172             :     [AV_PIX_FMT_RGB565LE]    = { 1, 1 },
     173             :     [AV_PIX_FMT_RGB555BE]    = { 1, 1 },
     174             :     [AV_PIX_FMT_RGB555LE]    = { 1, 1 },
     175             :     [AV_PIX_FMT_BGR565BE]    = { 1, 1 },
     176             :     [AV_PIX_FMT_BGR565LE]    = { 1, 1 },
     177             :     [AV_PIX_FMT_BGR555BE]    = { 1, 1 },
     178             :     [AV_PIX_FMT_BGR555LE]    = { 1, 1 },
     179             :     [AV_PIX_FMT_YUV420P16LE] = { 1, 1 },
     180             :     [AV_PIX_FMT_YUV420P16BE] = { 1, 1 },
     181             :     [AV_PIX_FMT_YUV422P16LE] = { 1, 1 },
     182             :     [AV_PIX_FMT_YUV422P16BE] = { 1, 1 },
     183             :     [AV_PIX_FMT_YUV444P16LE] = { 1, 1 },
     184             :     [AV_PIX_FMT_YUV444P16BE] = { 1, 1 },
     185             :     [AV_PIX_FMT_RGB444LE]    = { 1, 1 },
     186             :     [AV_PIX_FMT_RGB444BE]    = { 1, 1 },
     187             :     [AV_PIX_FMT_BGR444LE]    = { 1, 1 },
     188             :     [AV_PIX_FMT_BGR444BE]    = { 1, 1 },
     189             :     [AV_PIX_FMT_YA8]         = { 1, 1 },
     190             :     [AV_PIX_FMT_YA16BE]      = { 1, 0 },
     191             :     [AV_PIX_FMT_YA16LE]      = { 1, 0 },
     192             :     [AV_PIX_FMT_BGR48BE]     = { 1, 1 },
     193             :     [AV_PIX_FMT_BGR48LE]     = { 1, 1 },
     194             :     [AV_PIX_FMT_BGRA64BE]    = { 1, 1, 1 },
     195             :     [AV_PIX_FMT_BGRA64LE]    = { 1, 1, 1 },
     196             :     [AV_PIX_FMT_YUV420P9BE]  = { 1, 1 },
     197             :     [AV_PIX_FMT_YUV420P9LE]  = { 1, 1 },
     198             :     [AV_PIX_FMT_YUV420P10BE] = { 1, 1 },
     199             :     [AV_PIX_FMT_YUV420P10LE] = { 1, 1 },
     200             :     [AV_PIX_FMT_YUV420P12BE] = { 1, 1 },
     201             :     [AV_PIX_FMT_YUV420P12LE] = { 1, 1 },
     202             :     [AV_PIX_FMT_YUV420P14BE] = { 1, 1 },
     203             :     [AV_PIX_FMT_YUV420P14LE] = { 1, 1 },
     204             :     [AV_PIX_FMT_YUV422P9BE]  = { 1, 1 },
     205             :     [AV_PIX_FMT_YUV422P9LE]  = { 1, 1 },
     206             :     [AV_PIX_FMT_YUV422P10BE] = { 1, 1 },
     207             :     [AV_PIX_FMT_YUV422P10LE] = { 1, 1 },
     208             :     [AV_PIX_FMT_YUV422P12BE] = { 1, 1 },
     209             :     [AV_PIX_FMT_YUV422P12LE] = { 1, 1 },
     210             :     [AV_PIX_FMT_YUV422P14BE] = { 1, 1 },
     211             :     [AV_PIX_FMT_YUV422P14LE] = { 1, 1 },
     212             :     [AV_PIX_FMT_YUV444P9BE]  = { 1, 1 },
     213             :     [AV_PIX_FMT_YUV444P9LE]  = { 1, 1 },
     214             :     [AV_PIX_FMT_YUV444P10BE] = { 1, 1 },
     215             :     [AV_PIX_FMT_YUV444P10LE] = { 1, 1 },
     216             :     [AV_PIX_FMT_YUV444P12BE] = { 1, 1 },
     217             :     [AV_PIX_FMT_YUV444P12LE] = { 1, 1 },
     218             :     [AV_PIX_FMT_YUV444P14BE] = { 1, 1 },
     219             :     [AV_PIX_FMT_YUV444P14LE] = { 1, 1 },
     220             :     [AV_PIX_FMT_GBRP]        = { 1, 1 },
     221             :     [AV_PIX_FMT_GBRP9LE]     = { 1, 1 },
     222             :     [AV_PIX_FMT_GBRP9BE]     = { 1, 1 },
     223             :     [AV_PIX_FMT_GBRP10LE]    = { 1, 1 },
     224             :     [AV_PIX_FMT_GBRP10BE]    = { 1, 1 },
     225             :     [AV_PIX_FMT_GBRAP10LE]   = { 1, 1 },
     226             :     [AV_PIX_FMT_GBRAP10BE]   = { 1, 1 },
     227             :     [AV_PIX_FMT_GBRP12LE]    = { 1, 1 },
     228             :     [AV_PIX_FMT_GBRP12BE]    = { 1, 1 },
     229             :     [AV_PIX_FMT_GBRAP12LE]   = { 1, 1 },
     230             :     [AV_PIX_FMT_GBRAP12BE]   = { 1, 1 },
     231             :     [AV_PIX_FMT_GBRP14LE]    = { 1, 1 },
     232             :     [AV_PIX_FMT_GBRP14BE]    = { 1, 1 },
     233             :     [AV_PIX_FMT_GBRP16LE]    = { 1, 1 },
     234             :     [AV_PIX_FMT_GBRP16BE]    = { 1, 1 },
     235             :     [AV_PIX_FMT_GBRAP]       = { 1, 1 },
     236             :     [AV_PIX_FMT_GBRAP16LE]   = { 1, 1 },
     237             :     [AV_PIX_FMT_GBRAP16BE]   = { 1, 1 },
     238             :     [AV_PIX_FMT_BAYER_BGGR8] = { 1, 0 },
     239             :     [AV_PIX_FMT_BAYER_RGGB8] = { 1, 0 },
     240             :     [AV_PIX_FMT_BAYER_GBRG8] = { 1, 0 },
     241             :     [AV_PIX_FMT_BAYER_GRBG8] = { 1, 0 },
     242             :     [AV_PIX_FMT_BAYER_BGGR16LE] = { 1, 0 },
     243             :     [AV_PIX_FMT_BAYER_BGGR16BE] = { 1, 0 },
     244             :     [AV_PIX_FMT_BAYER_RGGB16LE] = { 1, 0 },
     245             :     [AV_PIX_FMT_BAYER_RGGB16BE] = { 1, 0 },
     246             :     [AV_PIX_FMT_BAYER_GBRG16LE] = { 1, 0 },
     247             :     [AV_PIX_FMT_BAYER_GBRG16BE] = { 1, 0 },
     248             :     [AV_PIX_FMT_BAYER_GRBG16LE] = { 1, 0 },
     249             :     [AV_PIX_FMT_BAYER_GRBG16BE] = { 1, 0 },
     250             :     [AV_PIX_FMT_XYZ12BE]     = { 1, 1, 1 },
     251             :     [AV_PIX_FMT_XYZ12LE]     = { 1, 1, 1 },
     252             :     [AV_PIX_FMT_AYUV64LE]    = { 1, 1},
     253             :     [AV_PIX_FMT_P010LE]      = { 1, 1 },
     254             :     [AV_PIX_FMT_P010BE]      = { 1, 1 },
     255             : };
     256             : 
     257      530450 : int sws_isSupportedInput(enum AVPixelFormat pix_fmt)
     258             : {
     259      530450 :     return (unsigned)pix_fmt < AV_PIX_FMT_NB ?
     260      530450 :            format_entries[pix_fmt].is_supported_in : 0;
     261             : }
     262             : 
     263      530450 : int sws_isSupportedOutput(enum AVPixelFormat pix_fmt)
     264             : {
     265      530450 :     return (unsigned)pix_fmt < AV_PIX_FMT_NB ?
     266      530450 :            format_entries[pix_fmt].is_supported_out : 0;
     267             : }
     268             : 
     269      209891 : int sws_isSupportedEndiannessConversion(enum AVPixelFormat pix_fmt)
     270             : {
     271      209891 :     return (unsigned)pix_fmt < AV_PIX_FMT_NB ?
     272      209891 :            format_entries[pix_fmt].is_supported_endianness : 0;
     273             : }
     274             : 
     275           0 : static double getSplineCoeff(double a, double b, double c, double d,
     276             :                              double dist)
     277             : {
     278           0 :     if (dist <= 1.0)
     279           0 :         return ((d * dist + c) * dist + b) * dist + a;
     280             :     else
     281           0 :         return getSplineCoeff(0.0,
     282           0 :                                b + 2.0 * c + 3.0 * d,
     283           0 :                                c + 3.0 * d,
     284           0 :                               -b - 3.0 * c - 6.0 * d,
     285             :                               dist - 1.0);
     286             : }
     287             : 
     288       21424 : static av_cold int get_local_pos(SwsContext *s, int chr_subsample, int pos, int dir)
     289             : {
     290       21424 :     if (pos == -1 || pos <= -513) {
     291        8321 :         pos = (128 << chr_subsample) - 128;
     292             :     }
     293       21424 :     pos += 128; // relative to ideal left edge
     294       21424 :     return pos >> chr_subsample;
     295             : }
     296             : 
     297             : typedef struct {
     298             :     int flag;                   ///< flag associated to the algorithm
     299             :     const char *description;    ///< human-readable description
     300             :     int size_factor;            ///< size factor used when initing the filters
     301             : } ScaleAlgorithm;
     302             : 
     303             : static const ScaleAlgorithm scale_algorithms[] = {
     304             :     { SWS_AREA,          "area averaging",                  1 /* downscale only, for upscale it is bilinear */ },
     305             :     { SWS_BICUBIC,       "bicubic",                         4 },
     306             :     { SWS_BICUBLIN,      "luma bicubic / chroma bilinear", -1 },
     307             :     { SWS_BILINEAR,      "bilinear",                        2 },
     308             :     { SWS_FAST_BILINEAR, "fast bilinear",                  -1 },
     309             :     { SWS_GAUSS,         "Gaussian",                        8 /* infinite ;) */ },
     310             :     { SWS_LANCZOS,       "Lanczos",                        -1 /* custom */ },
     311             :     { SWS_POINT,         "nearest neighbor / point",       -1 },
     312             :     { SWS_SINC,          "sinc",                           20 /* infinite ;) */ },
     313             :     { SWS_SPLINE,        "bicubic spline",                 20 /* infinite :)*/ },
     314             :     { SWS_X,             "experimental",                    8 },
     315             : };
     316             : 
     317       10712 : static av_cold int initFilter(int16_t **outFilter, int32_t **filterPos,
     318             :                               int *outFilterSize, int xInc, int srcW,
     319             :                               int dstW, int filterAlign, int one,
     320             :                               int flags, int cpu_flags,
     321             :                               SwsVector *srcFilter, SwsVector *dstFilter,
     322             :                               double param[2], int srcPos, int dstPos)
     323             : {
     324             :     int i;
     325             :     int filterSize;
     326             :     int filter2Size;
     327             :     int minFilterSize;
     328       10712 :     int64_t *filter    = NULL;
     329       10712 :     int64_t *filter2   = NULL;
     330       10712 :     const int64_t fone = 1LL << (54 - FFMIN(av_log2(srcW/dstW), 8));
     331       10712 :     int ret            = -1;
     332             : 
     333       10712 :     emms_c(); // FIXME should not be required but IS (even for non-MMX versions)
     334             : 
     335             :     // NOTE: the +3 is for the MMX(+1) / SSE(+3) scaler which reads over the end
     336       10712 :     FF_ALLOC_ARRAY_OR_GOTO(NULL, *filterPos, (dstW + 3), sizeof(**filterPos), fail);
     337             : 
     338       17401 :     if (FFABS(xInc - 0x10000) < 10 && srcPos == dstPos) { // unscaled
     339             :         int i;
     340        6689 :         filterSize = 1;
     341        6689 :         FF_ALLOCZ_ARRAY_OR_GOTO(NULL, filter,
     342             :                                 dstW, sizeof(*filter) * filterSize, fail);
     343             : 
     344     1815554 :         for (i = 0; i < dstW; i++) {
     345     1808865 :             filter[i * filterSize] = fone;
     346     1808865 :             (*filterPos)[i]        = i;
     347             :         }
     348        4023 :     } else if (flags & SWS_POINT) { // lame looking point sampling mode
     349             :         int i;
     350             :         int64_t xDstInSrc;
     351         188 :         filterSize = 1;
     352         188 :         FF_ALLOC_ARRAY_OR_GOTO(NULL, filter,
     353             :                                dstW, sizeof(*filter) * filterSize, fail);
     354             : 
     355         188 :         xDstInSrc = ((dstPos*(int64_t)xInc)>>8) - ((srcPos*0x8000LL)>>7);
     356       43549 :         for (i = 0; i < dstW; i++) {
     357       43361 :             int xx = (xDstInSrc - ((filterSize - 1) << 15) + (1 << 15)) >> 16;
     358             : 
     359       43361 :             (*filterPos)[i] = xx;
     360       43361 :             filter[i]       = fone;
     361       43361 :             xDstInSrc      += xInc;
     362             :         }
     363        7618 :     } else if ((xInc <= (1 << 16) && (flags & SWS_AREA)) ||
     364        3835 :                (flags & SWS_FAST_BILINEAR)) { // bilinear upscale
     365             :         int i;
     366             :         int64_t xDstInSrc;
     367          52 :         filterSize = 2;
     368          52 :         FF_ALLOC_ARRAY_OR_GOTO(NULL, filter,
     369             :                                dstW, sizeof(*filter) * filterSize, fail);
     370             : 
     371          52 :         xDstInSrc = ((dstPos*(int64_t)xInc)>>8) - ((srcPos*0x8000LL)>>7);
     372       29092 :         for (i = 0; i < dstW; i++) {
     373       29040 :             int xx = (xDstInSrc - ((filterSize - 1) << 15) + (1 << 15)) >> 16;
     374             :             int j;
     375             : 
     376       29040 :             (*filterPos)[i] = xx;
     377             :             // bilinear upscale / linear interpolate / area averaging
     378       87120 :             for (j = 0; j < filterSize; j++) {
     379       58080 :                 int64_t coeff= fone - FFABS(((int64_t)xx<<16) - xDstInSrc)*(fone>>16);
     380       58080 :                 if (coeff < 0)
     381           0 :                     coeff = 0;
     382       58080 :                 filter[i * filterSize + j] = coeff;
     383       58080 :                 xx++;
     384             :             }
     385       29040 :             xDstInSrc += xInc;
     386             :         }
     387             :     } else {
     388             :         int64_t xDstInSrc;
     389        3783 :         int sizeFactor = -1;
     390             : 
     391        7474 :         for (i = 0; i < FF_ARRAY_ELEMS(scale_algorithms); i++) {
     392        7474 :             if (flags & scale_algorithms[i].flag && scale_algorithms[i].size_factor > 0) {
     393        3783 :                 sizeFactor = scale_algorithms[i].size_factor;
     394        3783 :                 break;
     395             :             }
     396             :         }
     397        3783 :         if (flags & SWS_LANCZOS)
     398           0 :             sizeFactor = param[0] != SWS_PARAM_DEFAULT ? ceil(2 * param[0]) : 6;
     399        3783 :         av_assert0(sizeFactor > 0);
     400             : 
     401        3783 :         if (xInc <= 1 << 16)
     402        2606 :             filterSize = 1 + sizeFactor;    // upscale
     403             :         else
     404        1177 :             filterSize = 1 + (sizeFactor * srcW + dstW - 1) / dstW;
     405             : 
     406        3783 :         filterSize = FFMIN(filterSize, srcW - 2);
     407        3783 :         filterSize = FFMAX(filterSize, 1);
     408             : 
     409        3783 :         FF_ALLOC_ARRAY_OR_GOTO(NULL, filter,
     410             :                                dstW, sizeof(*filter) * filterSize, fail);
     411             : 
     412        3783 :         xDstInSrc = ((dstPos*(int64_t)xInc)>>7) - ((srcPos*0x10000LL)>>7);
     413     1227930 :         for (i = 0; i < dstW; i++) {
     414     1224147 :             int xx = (xDstInSrc - (filterSize - 2) * (1LL<<16)) / (1 << 17);
     415             :             int j;
     416     1224147 :             (*filterPos)[i] = xx;
     417     8129884 :             for (j = 0; j < filterSize; j++) {
     418     6905737 :                 int64_t d = (FFABS(((int64_t)xx * (1 << 17)) - xDstInSrc)) << 13;
     419             :                 double floatd;
     420             :                 int64_t coeff;
     421             : 
     422     6905737 :                 if (xInc > 1 << 16)
     423     1696182 :                     d = d * dstW / srcW;
     424     6905737 :                 floatd = d * (1.0 / (1 << 30));
     425             : 
     426     6905737 :                 if (flags & SWS_BICUBIC) {
     427     6489910 :                     int64_t B = (param[0] != SWS_PARAM_DEFAULT ? param[0] :   0) * (1 << 24);
     428     6489910 :                     int64_t C = (param[1] != SWS_PARAM_DEFAULT ? param[1] : 0.6) * (1 << 24);
     429             : 
     430     6489910 :                     if (d >= 1LL << 31) {
     431     1230426 :                         coeff = 0.0;
     432             :                     } else {
     433     5259484 :                         int64_t dd  = (d  * d) >> 30;
     434     5259484 :                         int64_t ddd = (dd * d) >> 30;
     435             : 
     436     5259484 :                         if (d < 1LL << 30)
     437     7899828 :                             coeff =  (12 * (1 << 24) -  9 * B - 6 * C) * ddd +
     438     2633276 :                                     (-18 * (1 << 24) + 12 * B + 6 * C) *  dd +
     439     2633276 :                                       (6 * (1 << 24) -  2 * B)         * (1 << 30);
     440             :                         else
     441     7878624 :                             coeff =      (-B -  6 * C) * ddd +
     442     5252416 :                                       (6 * B + 30 * C) * dd  +
     443     2626208 :                                     (-12 * B - 48 * C) * d   +
     444     2626208 :                                       (8 * B + 24 * C) * (1 << 30);
     445             :                     }
     446     6489910 :                     coeff /= (1LL<<54)/fone;
     447             :                 }
     448             : #if 0
     449             :                 else if (flags & SWS_X) {
     450             :                     double p  = param ? param * 0.01 : 0.3;
     451             :                     coeff     = d ? sin(d * M_PI) / (d * M_PI) : 1.0;
     452             :                     coeff    *= pow(2.0, -p * d * d);
     453             :                 }
     454             : #endif
     455      415827 :                 else if (flags & SWS_X) {
     456           0 :                     double A = param[0] != SWS_PARAM_DEFAULT ? param[0] : 1.0;
     457             :                     double c;
     458             : 
     459           0 :                     if (floatd < 1.0)
     460           0 :                         c = cos(floatd * M_PI);
     461             :                     else
     462           0 :                         c = -1.0;
     463           0 :                     if (c < 0.0)
     464           0 :                         c = -pow(-c, A);
     465             :                     else
     466           0 :                         c = pow(c, A);
     467           0 :                     coeff = (c * 0.5 + 0.5) * fone;
     468      415827 :                 } else if (flags & SWS_AREA) {
     469      288927 :                     int64_t d2 = d - (1 << 29);
     470      288927 :                     if (d2 * xInc < -(1LL << (29 + 16)))
     471      218924 :                         coeff = 1.0 * (1LL << (30 + 16));
     472       70003 :                     else if (d2 * xInc < (1LL << (29 + 16)))
     473       60908 :                         coeff = -d2 * xInc + (1LL << (29 + 16));
     474             :                     else
     475        9095 :                         coeff = 0.0;
     476      288927 :                     coeff *= fone >> (30 + 16);
     477      126900 :                 } else if (flags & SWS_GAUSS) {
     478           0 :                     double p = param[0] != SWS_PARAM_DEFAULT ? param[0] : 3.0;
     479           0 :                     coeff = exp2(-p * floatd * floatd) * fone;
     480      126900 :                 } else if (flags & SWS_SINC) {
     481           0 :                     coeff = (d ? sin(floatd * M_PI) / (floatd * M_PI) : 1.0) * fone;
     482      126900 :                 } else if (flags & SWS_LANCZOS) {
     483           0 :                     double p = param[0] != SWS_PARAM_DEFAULT ? param[0] : 3.0;
     484           0 :                     coeff = (d ? sin(floatd * M_PI) * sin(floatd * M_PI / p) /
     485           0 :                              (floatd * floatd * M_PI * M_PI / p) : 1.0) * fone;
     486           0 :                     if (floatd > p)
     487           0 :                         coeff = 0;
     488      126900 :                 } else if (flags & SWS_BILINEAR) {
     489      126900 :                     coeff = (1 << 30) - d;
     490      126900 :                     if (coeff < 0)
     491       42416 :                         coeff = 0;
     492      126900 :                     coeff *= fone >> 30;
     493           0 :                 } else if (flags & SWS_SPLINE) {
     494           0 :                     double p = -2.196152422706632;
     495           0 :                     coeff = getSplineCoeff(1.0, 0.0, p, -p - 1.0, floatd) * fone;
     496             :                 } else {
     497           0 :                     av_assert0(0);
     498             :                 }
     499             : 
     500     6905737 :                 filter[i * filterSize + j] = coeff;
     501     6905737 :                 xx++;
     502             :             }
     503     1224147 :             xDstInSrc += 2 * xInc;
     504             :         }
     505             :     }
     506             : 
     507             :     /* apply src & dst Filter to filter -> filter2
     508             :      * av_free(filter);
     509             :      */
     510       10712 :     av_assert0(filterSize > 0);
     511       10712 :     filter2Size = filterSize;
     512       10712 :     if (srcFilter)
     513           0 :         filter2Size += srcFilter->length - 1;
     514       10712 :     if (dstFilter)
     515           0 :         filter2Size += dstFilter->length - 1;
     516       10712 :     av_assert0(filter2Size > 0);
     517       10712 :     FF_ALLOCZ_ARRAY_OR_GOTO(NULL, filter2, dstW, filter2Size * sizeof(*filter2), fail);
     518             : 
     519     3116125 :     for (i = 0; i < dstW; i++) {
     520             :         int j, k;
     521             : 
     522     3105413 :         if (srcFilter) {
     523           0 :             for (k = 0; k < srcFilter->length; k++) {
     524           0 :                 for (j = 0; j < filterSize; j++)
     525           0 :                     filter2[i * filter2Size + k + j] +=
     526           0 :                         srcFilter->coeff[k] * filter[i * filterSize + j];
     527             :             }
     528             :         } else {
     529    11921456 :             for (j = 0; j < filterSize; j++)
     530     8816043 :                 filter2[i * filter2Size + j] = filter[i * filterSize + j];
     531             :         }
     532             :         // FIXME dstFilter
     533             : 
     534     3105413 :         (*filterPos)[i] += (filterSize - 1) / 2 - (filter2Size - 1) / 2;
     535             :     }
     536       10712 :     av_freep(&filter);
     537             : 
     538             :     /* try to reduce the filter-size (step1 find size and shift left) */
     539             :     // Assume it is near normalized (*0.5 or *2.0 is OK but * 0.001 is not).
     540       10712 :     minFilterSize = 0;
     541     3116125 :     for (i = dstW - 1; i >= 0; i--) {
     542     3105413 :         int min = filter2Size;
     543             :         int j;
     544     3105413 :         int64_t cutOff = 0.0;
     545             : 
     546             :         /* get rid of near zero elements on the left by shifting left */
     547     3733515 :         for (j = 0; j < filter2Size; j++) {
     548             :             int k;
     549     3733515 :             cutOff += FFABS(filter2[i * filter2Size]);
     550             : 
     551     3733515 :             if (cutOff > SWS_MAX_REDUCE_CUTOFF * fone)
     552     3104312 :                 break;
     553             : 
     554             :             /* preserve monotonicity because the core can't handle the
     555             :              * filter otherwise */
     556      629203 :             if (i < dstW - 1 && (*filterPos)[i] >= (*filterPos)[i + 1])
     557        1101 :                 break;
     558             : 
     559             :             // move filter coefficients left
     560     3637170 :             for (k = 1; k < filter2Size; k++)
     561     3009068 :                 filter2[i * filter2Size + k - 1] = filter2[i * filter2Size + k];
     562      628102 :             filter2[i * filter2Size + k - 1] = 0;
     563      628102 :             (*filterPos)[i]++;
     564             :         }
     565             : 
     566     3105413 :         cutOff = 0;
     567             :         /* count near zeros on the right */
     568     4424218 :         for (j = filter2Size - 1; j > 0; j--) {
     569     2570871 :             cutOff += FFABS(filter2[i * filter2Size + j]);
     570             : 
     571     2570871 :             if (cutOff > SWS_MAX_REDUCE_CUTOFF * fone)
     572     1252066 :                 break;
     573     1318805 :             min--;
     574             :         }
     575             : 
     576     3105413 :         if (min > minFilterSize)
     577       11162 :             minFilterSize = min;
     578             :     }
     579             : 
     580             :     if (PPC_ALTIVEC(cpu_flags)) {
     581             :         // we can handle the special case 4, so we don't want to go the full 8
     582             :         if (minFilterSize < 5)
     583             :             filterAlign = 4;
     584             : 
     585             :         /* We really don't want to waste our time doing useless computation, so
     586             :          * fall back on the scalar C code for very small filters.
     587             :          * Vectorizing is worth it only if you have a decent-sized vector. */
     588             :         if (minFilterSize < 3)
     589             :             filterAlign = 1;
     590             :     }
     591             : 
     592       10712 :     if (HAVE_MMX && cpu_flags & AV_CPU_FLAG_MMX) {
     593             :         // special case for unscaled vertical filtering
     594          12 :         if (minFilterSize == 1 && filterAlign == 2)
     595           3 :             filterAlign = 1;
     596             :     }
     597             : 
     598       10712 :     av_assert0(minFilterSize > 0);
     599       10712 :     filterSize = (minFilterSize + (filterAlign - 1)) & (~(filterAlign - 1));
     600       10712 :     av_assert0(filterSize > 0);
     601       10712 :     filter = av_malloc_array(dstW, filterSize * sizeof(*filter));
     602       10712 :     if (!filter)
     603           0 :         goto fail;
     604       10712 :     if (filterSize >= MAX_FILTER_SIZE * 16 /
     605       10712 :                       ((flags & SWS_ACCURATE_RND) ? APCK_SIZE : 16)) {
     606           0 :         ret = RETCODE_USE_CASCADE;
     607           0 :         goto fail;
     608             :     }
     609       10712 :     *outFilterSize = filterSize;
     610             : 
     611       10712 :     if (flags & SWS_PRINT_INFO)
     612           0 :         av_log(NULL, AV_LOG_VERBOSE,
     613             :                "SwScaler: reducing / aligning filtersize %d -> %d\n",
     614             :                filter2Size, filterSize);
     615             :     /* try to reduce the filter-size (step2 reduce it) */
     616     3116125 :     for (i = 0; i < dstW; i++) {
     617             :         int j;
     618             : 
     619    10670293 :         for (j = 0; j < filterSize; j++) {
     620     7564880 :             if (j >= filter2Size)
     621        4320 :                 filter[i * filterSize + j] = 0;
     622             :             else
     623     7560560 :                 filter[i * filterSize + j] = filter2[i * filter2Size + j];
     624     7564880 :             if ((flags & SWS_BITEXACT) && j >= minFilterSize)
     625        4320 :                 filter[i * filterSize + j] = 0;
     626             :         }
     627             :     }
     628             : 
     629             :     // FIXME try to align filterPos if possible
     630             : 
     631             :     // fix borders
     632     3116125 :     for (i = 0; i < dstW; i++) {
     633             :         int j;
     634     3105413 :         if ((*filterPos)[i] < 0) {
     635             :             // move filter coefficients left to compensate for filterPos
     636       51977 :             for (j = 1; j < filterSize; j++) {
     637       41668 :                 int left = FFMAX(j + (*filterPos)[i], 0);
     638       41668 :                 filter[i * filterSize + left] += filter[i * filterSize + j];
     639       41668 :                 filter[i * filterSize + j]     = 0;
     640             :             }
     641       10309 :             (*filterPos)[i]= 0;
     642             :         }
     643             : 
     644     3105413 :         if ((*filterPos)[i] + filterSize > srcW) {
     645       14790 :             int shift = (*filterPos)[i] + FFMIN(filterSize - srcW, 0);
     646       14790 :             int64_t acc = 0;
     647             : 
     648       86805 :             for (j = filterSize - 1; j >= 0; j--) {
     649       72015 :                 if ((*filterPos)[i] + j >= srcW) {
     650       22263 :                     acc += filter[i * filterSize + j];
     651       22263 :                     filter[i * filterSize + j] = 0;
     652             :                 }
     653             :             }
     654       86805 :             for (j = filterSize - 1; j >= 0; j--) {
     655       72015 :                 if (j < shift) {
     656       22263 :                     filter[i * filterSize + j] = 0;
     657             :                 } else {
     658       49752 :                     filter[i * filterSize + j] = filter[i * filterSize + j - shift];
     659             :                 }
     660             :             }
     661             : 
     662       14790 :             (*filterPos)[i]-= shift;
     663       14790 :             filter[i * filterSize + srcW - 1 - (*filterPos)[i]] += acc;
     664             :         }
     665     3105413 :         av_assert0((*filterPos)[i] >= 0);
     666     3105413 :         av_assert0((*filterPos)[i] < srcW);
     667     3105413 :         if ((*filterPos)[i] + filterSize > srcW) {
     668           0 :             for (j = 0; j < filterSize; j++) {
     669           0 :                 av_assert0((*filterPos)[i] + j < srcW || !filter[i * filterSize + j]);
     670             :             }
     671             :         }
     672             :     }
     673             : 
     674             :     // Note the +1 is for the MMX scaler which reads over the end
     675             :     /* align at 16 for AltiVec (needed by hScale_altivec_real) */
     676       10712 :     FF_ALLOCZ_ARRAY_OR_GOTO(NULL, *outFilter,
     677             :                             (dstW + 3), *outFilterSize * sizeof(int16_t), fail);
     678             : 
     679             :     /* normalize & store in outFilter */
     680     3116125 :     for (i = 0; i < dstW; i++) {
     681             :         int j;
     682     3105413 :         int64_t error = 0;
     683     3105413 :         int64_t sum   = 0;
     684             : 
     685    10670293 :         for (j = 0; j < filterSize; j++) {
     686     7564880 :             sum += filter[i * filterSize + j];
     687             :         }
     688     3105413 :         sum = (sum + one / 2) / one;
     689     3105413 :         if (!sum) {
     690           0 :             av_log(NULL, AV_LOG_WARNING, "SwScaler: zero vector in scaling\n");
     691           0 :             sum = 1;
     692             :         }
     693    10670293 :         for (j = 0; j < *outFilterSize; j++) {
     694     7564880 :             int64_t v = filter[i * filterSize + j] + error;
     695     7564880 :             int intV  = ROUNDED_DIV(v, sum);
     696     7564880 :             (*outFilter)[i * (*outFilterSize) + j] = intV;
     697     7564880 :             error                                  = v - intV * sum;
     698             :         }
     699             :     }
     700             : 
     701       21424 :     (*filterPos)[dstW + 0] =
     702       21424 :     (*filterPos)[dstW + 1] =
     703       10712 :     (*filterPos)[dstW + 2] = (*filterPos)[dstW - 1]; /* the MMX/SSE scaler will
     704             :                                                       * read over the end */
     705       41057 :     for (i = 0; i < *outFilterSize; i++) {
     706       30345 :         int k = (dstW - 1) * (*outFilterSize) + i;
     707       60690 :         (*outFilter)[k + 1 * (*outFilterSize)] =
     708       60690 :         (*outFilter)[k + 2 * (*outFilterSize)] =
     709       30345 :         (*outFilter)[k + 3 * (*outFilterSize)] = (*outFilter)[k];
     710             :     }
     711             : 
     712       10712 :     ret = 0;
     713             : 
     714             : fail:
     715       10712 :     if(ret < 0)
     716           0 :         av_log(NULL, ret == RETCODE_USE_CASCADE ? AV_LOG_DEBUG : AV_LOG_ERROR, "sws: initFilter failed\n");
     717       10712 :     av_free(filter);
     718       10712 :     av_free(filter2);
     719       10712 :     return ret;
     720             : }
     721             : 
     722        1116 : static void fill_rgb2yuv_table(SwsContext *c, const int table[4], int dstRange)
     723             : {
     724             :     int64_t W, V, Z, Cy, Cu, Cv;
     725        1116 :     int64_t vr =  table[0];
     726        1116 :     int64_t ub =  table[1];
     727        1116 :     int64_t ug = -table[2];
     728        1116 :     int64_t vg = -table[3];
     729        1116 :     int64_t ONE = 65536;
     730        1116 :     int64_t cy = ONE;
     731        1116 :     uint8_t *p = (uint8_t*)c->input_rgb2yuv_table;
     732             :     int i;
     733             :     static const int8_t map[] = {
     734             :     BY_IDX, GY_IDX, -1    , BY_IDX, BY_IDX, GY_IDX, -1    , BY_IDX,
     735             :     RY_IDX, -1    , GY_IDX, RY_IDX, RY_IDX, -1    , GY_IDX, RY_IDX,
     736             :     RY_IDX, GY_IDX, -1    , RY_IDX, RY_IDX, GY_IDX, -1    , RY_IDX,
     737             :     BY_IDX, -1    , GY_IDX, BY_IDX, BY_IDX, -1    , GY_IDX, BY_IDX,
     738             :     BU_IDX, GU_IDX, -1    , BU_IDX, BU_IDX, GU_IDX, -1    , BU_IDX,
     739             :     RU_IDX, -1    , GU_IDX, RU_IDX, RU_IDX, -1    , GU_IDX, RU_IDX,
     740             :     RU_IDX, GU_IDX, -1    , RU_IDX, RU_IDX, GU_IDX, -1    , RU_IDX,
     741             :     BU_IDX, -1    , GU_IDX, BU_IDX, BU_IDX, -1    , GU_IDX, BU_IDX,
     742             :     BV_IDX, GV_IDX, -1    , BV_IDX, BV_IDX, GV_IDX, -1    , BV_IDX,
     743             :     RV_IDX, -1    , GV_IDX, RV_IDX, RV_IDX, -1    , GV_IDX, RV_IDX,
     744             :     RV_IDX, GV_IDX, -1    , RV_IDX, RV_IDX, GV_IDX, -1    , RV_IDX,
     745             :     BV_IDX, -1    , GV_IDX, BV_IDX, BV_IDX, -1    , GV_IDX, BV_IDX,
     746             :     RY_IDX, BY_IDX, RY_IDX, BY_IDX, RY_IDX, BY_IDX, RY_IDX, BY_IDX,
     747             :     BY_IDX, RY_IDX, BY_IDX, RY_IDX, BY_IDX, RY_IDX, BY_IDX, RY_IDX,
     748             :     GY_IDX, -1    , GY_IDX, -1    , GY_IDX, -1    , GY_IDX, -1    ,
     749             :     -1    , GY_IDX, -1    , GY_IDX, -1    , GY_IDX, -1    , GY_IDX,
     750             :     RU_IDX, BU_IDX, RU_IDX, BU_IDX, RU_IDX, BU_IDX, RU_IDX, BU_IDX,
     751             :     BU_IDX, RU_IDX, BU_IDX, RU_IDX, BU_IDX, RU_IDX, BU_IDX, RU_IDX,
     752             :     GU_IDX, -1    , GU_IDX, -1    , GU_IDX, -1    , GU_IDX, -1    ,
     753             :     -1    , GU_IDX, -1    , GU_IDX, -1    , GU_IDX, -1    , GU_IDX,
     754             :     RV_IDX, BV_IDX, RV_IDX, BV_IDX, RV_IDX, BV_IDX, RV_IDX, BV_IDX,
     755             :     BV_IDX, RV_IDX, BV_IDX, RV_IDX, BV_IDX, RV_IDX, BV_IDX, RV_IDX,
     756             :     GV_IDX, -1    , GV_IDX, -1    , GV_IDX, -1    , GV_IDX, -1    ,
     757             :     -1    , GV_IDX, -1    , GV_IDX, -1    , GV_IDX, -1    , GV_IDX, //23
     758             :     -1    , -1    , -1    , -1    , -1    , -1    , -1    , -1    , //24
     759             :     -1    , -1    , -1    , -1    , -1    , -1    , -1    , -1    , //25
     760             :     -1    , -1    , -1    , -1    , -1    , -1    , -1    , -1    , //26
     761             :     -1    , -1    , -1    , -1    , -1    , -1    , -1    , -1    , //27
     762             :     -1    , -1    , -1    , -1    , -1    , -1    , -1    , -1    , //28
     763             :     -1    , -1    , -1    , -1    , -1    , -1    , -1    , -1    , //29
     764             :     -1    , -1    , -1    , -1    , -1    , -1    , -1    , -1    , //30
     765             :     -1    , -1    , -1    , -1    , -1    , -1    , -1    , -1    , //31
     766             :     BY_IDX, GY_IDX, RY_IDX, -1    , -1    , -1    , -1    , -1    , //32
     767             :     BU_IDX, GU_IDX, RU_IDX, -1    , -1    , -1    , -1    , -1    , //33
     768             :     BV_IDX, GV_IDX, RV_IDX, -1    , -1    , -1    , -1    , -1    , //34
     769             :     };
     770             : 
     771        1116 :     dstRange = 0; //FIXME range = 1 is handled elsewhere
     772             : 
     773        1116 :     if (!dstRange) {
     774        1116 :         cy = cy * 255 / 219;
     775             :     } else {
     776           0 :         vr = vr * 224 / 255;
     777           0 :         ub = ub * 224 / 255;
     778           0 :         ug = ug * 224 / 255;
     779           0 :         vg = vg * 224 / 255;
     780             :     }
     781        1116 :     W = ROUNDED_DIV(ONE*ONE*ug, ub);
     782        1116 :     V = ROUNDED_DIV(ONE*ONE*vg, vr);
     783        1116 :     Z = ONE*ONE-W-V;
     784             : 
     785        1116 :     Cy = ROUNDED_DIV(cy*Z, ONE);
     786        1116 :     Cu = ROUNDED_DIV(ub*Z, ONE);
     787        1116 :     Cv = ROUNDED_DIV(vr*Z, ONE);
     788             : 
     789        1116 :     c->input_rgb2yuv_table[RY_IDX] = -ROUNDED_DIV((1 << RGB2YUV_SHIFT)*V        , Cy);
     790        1116 :     c->input_rgb2yuv_table[GY_IDX] =  ROUNDED_DIV((1 << RGB2YUV_SHIFT)*ONE*ONE  , Cy);
     791        1116 :     c->input_rgb2yuv_table[BY_IDX] = -ROUNDED_DIV((1 << RGB2YUV_SHIFT)*W        , Cy);
     792             : 
     793        1116 :     c->input_rgb2yuv_table[RU_IDX] =  ROUNDED_DIV((1 << RGB2YUV_SHIFT)*V        , Cu);
     794        1116 :     c->input_rgb2yuv_table[GU_IDX] = -ROUNDED_DIV((1 << RGB2YUV_SHIFT)*ONE*ONE  , Cu);
     795        1116 :     c->input_rgb2yuv_table[BU_IDX] =  ROUNDED_DIV((1 << RGB2YUV_SHIFT)*(Z+W)    , Cu);
     796             : 
     797        1116 :     c->input_rgb2yuv_table[RV_IDX] =  ROUNDED_DIV((1 << RGB2YUV_SHIFT)*(V+Z)    , Cv);
     798        1116 :     c->input_rgb2yuv_table[GV_IDX] = -ROUNDED_DIV((1 << RGB2YUV_SHIFT)*ONE*ONE  , Cv);
     799        1116 :     c->input_rgb2yuv_table[BV_IDX] =  ROUNDED_DIV((1 << RGB2YUV_SHIFT)*W        , Cv);
     800             : 
     801        1116 :     if(/*!dstRange && */!memcmp(table, ff_yuv2rgb_coeffs[SWS_CS_DEFAULT], sizeof(ff_yuv2rgb_coeffs[SWS_CS_DEFAULT]))) {
     802        1116 :         c->input_rgb2yuv_table[BY_IDX] =  ((int)(0.114 * 219 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));
     803        1116 :         c->input_rgb2yuv_table[BV_IDX] = (-(int)(0.081 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));
     804        1116 :         c->input_rgb2yuv_table[BU_IDX] =  ((int)(0.500 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));
     805        1116 :         c->input_rgb2yuv_table[GY_IDX] =  ((int)(0.587 * 219 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));
     806        1116 :         c->input_rgb2yuv_table[GV_IDX] = (-(int)(0.419 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));
     807        1116 :         c->input_rgb2yuv_table[GU_IDX] = (-(int)(0.331 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));
     808        1116 :         c->input_rgb2yuv_table[RY_IDX] =  ((int)(0.299 * 219 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));
     809        1116 :         c->input_rgb2yuv_table[RV_IDX] =  ((int)(0.500 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));
     810        1116 :         c->input_rgb2yuv_table[RU_IDX] = (-(int)(0.169 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));
     811             :     }
     812      313596 :     for(i=0; i<FF_ARRAY_ELEMS(map); i++)
     813      312480 :         AV_WL16(p + 16*4 + 2*i, map[i] >= 0 ? c->input_rgb2yuv_table[map[i]] : 0);
     814        1116 : }
     815             : 
     816          77 : static void fill_xyztables(struct SwsContext *c)
     817             : {
     818             :     int i;
     819          77 :     double xyzgamma = XYZ_GAMMA;
     820          77 :     double rgbgamma = 1.0 / RGB_GAMMA;
     821          77 :     double xyzgammainv = 1.0 / XYZ_GAMMA;
     822          77 :     double rgbgammainv = RGB_GAMMA;
     823             :     static const int16_t xyz2rgb_matrix[3][4] = {
     824             :         {13270, -6295, -2041},
     825             :         {-3969,  7682,   170},
     826             :         {  228,  -835,  4329} };
     827             :     static const int16_t rgb2xyz_matrix[3][4] = {
     828             :         {1689, 1464,  739},
     829             :         { 871, 2929,  296},
     830             :         {  79,  488, 3891} };
     831             :     static int16_t xyzgamma_tab[4096], rgbgamma_tab[4096], xyzgammainv_tab[4096], rgbgammainv_tab[4096];
     832             : 
     833          77 :     memcpy(c->xyz2rgb_matrix, xyz2rgb_matrix, sizeof(c->xyz2rgb_matrix));
     834          77 :     memcpy(c->rgb2xyz_matrix, rgb2xyz_matrix, sizeof(c->rgb2xyz_matrix));
     835          77 :     c->xyzgamma = xyzgamma_tab;
     836          77 :     c->rgbgamma = rgbgamma_tab;
     837          77 :     c->xyzgammainv = xyzgammainv_tab;
     838          77 :     c->rgbgammainv = rgbgammainv_tab;
     839             : 
     840          77 :     if (rgbgamma_tab[4095])
     841          56 :         return;
     842             : 
     843             :     /* set gamma vectors */
     844       86037 :     for (i = 0; i < 4096; i++) {
     845       86016 :         xyzgamma_tab[i] = lrint(pow(i / 4095.0, xyzgamma) * 4095.0);
     846       86016 :         rgbgamma_tab[i] = lrint(pow(i / 4095.0, rgbgamma) * 4095.0);
     847       86016 :         xyzgammainv_tab[i] = lrint(pow(i / 4095.0, xyzgammainv) * 4095.0);
     848       86016 :         rgbgammainv_tab[i] = lrint(pow(i / 4095.0, rgbgammainv) * 4095.0);
     849             :     }
     850             : }
     851             : 
     852       38647 : int sws_setColorspaceDetails(struct SwsContext *c, const int inv_table[4],
     853             :                              int srcRange, const int table[4], int dstRange,
     854             :                              int brightness, int contrast, int saturation)
     855             : {
     856             :     const AVPixFmtDescriptor *desc_dst;
     857             :     const AVPixFmtDescriptor *desc_src;
     858       38647 :     int need_reinit = 0;
     859             : 
     860       38647 :     handle_formats(c);
     861       38647 :     desc_dst = av_pix_fmt_desc_get(c->dstFormat);
     862       38647 :     desc_src = av_pix_fmt_desc_get(c->srcFormat);
     863             : 
     864       38647 :     if(!isYUV(c->dstFormat) && !isGray(c->dstFormat))
     865       15484 :         dstRange = 0;
     866       38647 :     if(!isYUV(c->srcFormat) && !isGray(c->srcFormat))
     867       12146 :         srcRange = 0;
     868             : 
     869       77294 :     if (c->srcRange != srcRange ||
     870       77294 :         c->dstRange != dstRange ||
     871       77294 :         c->brightness != brightness ||
     872       74616 :         c->contrast   != contrast ||
     873       71938 :         c->saturation != saturation ||
     874       71853 :         memcmp(c->srcColorspaceTable, inv_table, sizeof(int) * 4) ||
     875       35884 :         memcmp(c->dstColorspaceTable,     table, sizeof(int) * 4)
     876             :     )
     877        2763 :         need_reinit = 1;
     878             : 
     879       38647 :     memmove(c->srcColorspaceTable, inv_table, sizeof(int) * 4);
     880       38647 :     memmove(c->dstColorspaceTable, table, sizeof(int) * 4);
     881             : 
     882             : 
     883             : 
     884       38647 :     c->brightness = brightness;
     885       38647 :     c->contrast   = contrast;
     886       38647 :     c->saturation = saturation;
     887       38647 :     c->srcRange   = srcRange;
     888       38647 :     c->dstRange   = dstRange;
     889             : 
     890             :     //The srcBpc check is possibly wrong but we seem to lack a definitive reference to test this
     891             :     //and what we have in ticket 2939 looks better with this check
     892       38647 :     if (need_reinit && (c->srcBpc == 8 || !isYUV(c->srcFormat)))
     893         389 :         ff_sws_init_range_convert(c);
     894             : 
     895       38647 :     c->dstFormatBpp = av_get_bits_per_pixel(desc_dst);
     896       38647 :     c->srcFormatBpp = av_get_bits_per_pixel(desc_src);
     897             : 
     898       38647 :     if (c->cascaded_context[c->cascaded_mainindex])
     899           0 :         return sws_setColorspaceDetails(c->cascaded_context[c->cascaded_mainindex],inv_table, srcRange,table, dstRange, brightness,  contrast, saturation);
     900             : 
     901       38647 :     if (!need_reinit)
     902       35884 :         return 0;
     903             : 
     904        2763 :     if ((isYUV(c->dstFormat) || isGray(c->dstFormat)) && (isYUV(c->srcFormat) || isGray(c->srcFormat))) {
     905        3294 :         if (!c->cascaded_context[0] &&
     906        1647 :             memcmp(c->dstColorspaceTable, c->srcColorspaceTable, sizeof(int) * 4) &&
     907           0 :             c->srcW && c->srcH && c->dstW && c->dstH) {
     908             :             enum AVPixelFormat tmp_format;
     909             :             int tmp_width, tmp_height;
     910           0 :             int srcW = c->srcW;
     911           0 :             int srcH = c->srcH;
     912           0 :             int dstW = c->dstW;
     913           0 :             int dstH = c->dstH;
     914             :             int ret;
     915           0 :             av_log(c, AV_LOG_VERBOSE, "YUV color matrix differs for YUV->YUV, using intermediate RGB to convert\n");
     916             : 
     917           0 :             if (isNBPS(c->dstFormat) || is16BPS(c->dstFormat)) {
     918           0 :                 if (isALPHA(c->srcFormat) && isALPHA(c->dstFormat)) {
     919           0 :                     tmp_format = AV_PIX_FMT_BGRA64;
     920             :                 } else {
     921           0 :                     tmp_format = AV_PIX_FMT_BGR48;
     922             :                 }
     923             :             } else {
     924           0 :                 if (isALPHA(c->srcFormat) && isALPHA(c->dstFormat)) {
     925           0 :                     tmp_format = AV_PIX_FMT_BGRA;
     926             :                 } else {
     927           0 :                     tmp_format = AV_PIX_FMT_BGR24;
     928             :                 }
     929             :             }
     930             : 
     931           0 :             if (srcW*srcH > dstW*dstH) {
     932           0 :                 tmp_width  = dstW;
     933           0 :                 tmp_height = dstH;
     934             :             } else {
     935           0 :                 tmp_width  = srcW;
     936           0 :                 tmp_height = srcH;
     937             :             }
     938             : 
     939           0 :             ret = av_image_alloc(c->cascaded_tmp, c->cascaded_tmpStride,
     940             :                                 tmp_width, tmp_height, tmp_format, 64);
     941           0 :             if (ret < 0)
     942           0 :                 return ret;
     943             : 
     944           0 :             c->cascaded_context[0] = sws_alloc_set_opts(srcW, srcH, c->srcFormat,
     945             :                                                         tmp_width, tmp_height, tmp_format,
     946           0 :                                                         c->flags, c->param);
     947           0 :             if (!c->cascaded_context[0])
     948           0 :                 return -1;
     949             : 
     950           0 :             c->cascaded_context[0]->alphablend = c->alphablend;
     951           0 :             ret = sws_init_context(c->cascaded_context[0], NULL , NULL);
     952           0 :             if (ret < 0)
     953           0 :                 return ret;
     954             :             //we set both src and dst depending on that the RGB side will be ignored
     955           0 :             sws_setColorspaceDetails(c->cascaded_context[0], inv_table,
     956             :                                      srcRange, table, dstRange,
     957             :                                      brightness, contrast, saturation);
     958             : 
     959           0 :             c->cascaded_context[1] = sws_getContext(tmp_width, tmp_height, tmp_format,
     960             :                                                     dstW, dstH, c->dstFormat,
     961           0 :                                                     c->flags, NULL, NULL, c->param);
     962           0 :             if (!c->cascaded_context[1])
     963           0 :                 return -1;
     964           0 :             sws_setColorspaceDetails(c->cascaded_context[1], inv_table,
     965             :                                      srcRange, table, dstRange,
     966             :                                      0, 1 << 16, 1 << 16);
     967           0 :             return 0;
     968             :         }
     969        1647 :         return -1;
     970             :     }
     971             : 
     972        1116 :     if (!isYUV(c->dstFormat) && !isGray(c->dstFormat)) {
     973        1008 :         ff_yuv2rgb_c_init_tables(c, inv_table, srcRange, brightness,
     974             :                                  contrast, saturation);
     975             :         // FIXME factorize
     976             : 
     977             :         if (ARCH_PPC)
     978             :             ff_yuv2rgb_init_tables_ppc(c, inv_table, brightness,
     979             :                                        contrast, saturation);
     980             :     }
     981             : 
     982        1116 :     fill_rgb2yuv_table(c, table, dstRange);
     983             : 
     984        1116 :     return 0;
     985             : }
     986             : 
     987       35969 : int sws_getColorspaceDetails(struct SwsContext *c, int **inv_table,
     988             :                              int *srcRange, int **table, int *dstRange,
     989             :                              int *brightness, int *contrast, int *saturation)
     990             : {
     991       35969 :     if (!c )
     992           0 :         return -1;
     993             : 
     994       35969 :     *inv_table  = c->srcColorspaceTable;
     995       35969 :     *table      = c->dstColorspaceTable;
     996       35969 :     *srcRange   = c->srcRange;
     997       35969 :     *dstRange   = c->dstRange;
     998       35969 :     *brightness = c->brightness;
     999       35969 :     *contrast   = c->contrast;
    1000       35969 :     *saturation = c->saturation;
    1001             : 
    1002       35969 :     return 0;
    1003             : }
    1004             : 
    1005        5356 : static int handle_jpeg(enum AVPixelFormat *format)
    1006             : {
    1007        5356 :     switch (*format) {
    1008             :     case AV_PIX_FMT_YUVJ420P:
    1009          48 :         *format = AV_PIX_FMT_YUV420P;
    1010          48 :         return 1;
    1011             :     case AV_PIX_FMT_YUVJ411P:
    1012          15 :         *format = AV_PIX_FMT_YUV411P;
    1013          15 :         return 1;
    1014             :     case AV_PIX_FMT_YUVJ422P:
    1015          28 :         *format = AV_PIX_FMT_YUV422P;
    1016          28 :         return 1;
    1017             :     case AV_PIX_FMT_YUVJ444P:
    1018          36 :         *format = AV_PIX_FMT_YUV444P;
    1019          36 :         return 1;
    1020             :     case AV_PIX_FMT_YUVJ440P:
    1021          19 :         *format = AV_PIX_FMT_YUV440P;
    1022          19 :         return 1;
    1023             :     case AV_PIX_FMT_GRAY8:
    1024             :     case AV_PIX_FMT_YA8:
    1025             :     case AV_PIX_FMT_GRAY16LE:
    1026             :     case AV_PIX_FMT_GRAY16BE:
    1027             :     case AV_PIX_FMT_YA16BE:
    1028             :     case AV_PIX_FMT_YA16LE:
    1029          81 :         return 1;
    1030             :     default:
    1031        5129 :         return 0;
    1032             :     }
    1033             : }
    1034             : 
    1035       82650 : static int handle_0alpha(enum AVPixelFormat *format)
    1036             : {
    1037       82650 :     switch (*format) {
    1038          15 :     case AV_PIX_FMT_0BGR    : *format = AV_PIX_FMT_ABGR   ; return 1;
    1039          46 :     case AV_PIX_FMT_BGR0    : *format = AV_PIX_FMT_BGRA   ; return 4;
    1040          15 :     case AV_PIX_FMT_0RGB    : *format = AV_PIX_FMT_ARGB   ; return 1;
    1041          15 :     case AV_PIX_FMT_RGB0    : *format = AV_PIX_FMT_RGBA   ; return 4;
    1042       82559 :     default:                                          return 0;
    1043             :     }
    1044             : }
    1045             : 
    1046       82650 : static int handle_xyz(enum AVPixelFormat *format)
    1047             : {
    1048       82650 :     switch (*format) {
    1049          12 :     case AV_PIX_FMT_XYZ12BE : *format = AV_PIX_FMT_RGB48BE; return 1;
    1050          13 :     case AV_PIX_FMT_XYZ12LE : *format = AV_PIX_FMT_RGB48LE; return 1;
    1051       82625 :     default:                                                return 0;
    1052             :     }
    1053             : }
    1054             : 
    1055       41325 : static void handle_formats(SwsContext *c)
    1056             : {
    1057       41325 :     c->src0Alpha |= handle_0alpha(&c->srcFormat);
    1058       41325 :     c->dst0Alpha |= handle_0alpha(&c->dstFormat);
    1059       41325 :     c->srcXYZ    |= handle_xyz(&c->srcFormat);
    1060       41325 :     c->dstXYZ    |= handle_xyz(&c->dstFormat);
    1061       41325 :     if (c->srcXYZ || c->dstXYZ)
    1062          77 :         fill_xyztables(c);
    1063       41325 : }
    1064             : 
    1065        8662 : SwsContext *sws_alloc_context(void)
    1066             : {
    1067        8662 :     SwsContext *c = av_mallocz(sizeof(SwsContext));
    1068             : 
    1069             :     av_assert0(offsetof(SwsContext, redDither) + DITHER32_INT == offsetof(SwsContext, dither32));
    1070             : 
    1071        8662 :     if (c) {
    1072        8662 :         c->av_class = &ff_sws_context_class;
    1073        8662 :         av_opt_set_defaults(c);
    1074             :     }
    1075             : 
    1076        8662 :     return c;
    1077             : }
    1078             : 
    1079           0 : static uint16_t * alloc_gamma_tbl(double e)
    1080             : {
    1081           0 :     int i = 0;
    1082             :     uint16_t * tbl;
    1083           0 :     tbl = (uint16_t*)av_malloc(sizeof(uint16_t) * 1 << 16);
    1084           0 :     if (!tbl)
    1085           0 :         return NULL;
    1086             : 
    1087           0 :     for (i = 0; i < 65536; ++i) {
    1088           0 :         tbl[i] = pow(i / 65535.0, e) * 65535.0;
    1089             :     }
    1090           0 :     return tbl;
    1091             : }
    1092             : 
    1093         126 : static enum AVPixelFormat alphaless_fmt(enum AVPixelFormat fmt)
    1094             : {
    1095         126 :     switch(fmt) {
    1096           1 :     case AV_PIX_FMT_ARGB:       return AV_PIX_FMT_RGB24;
    1097           1 :     case AV_PIX_FMT_RGBA:       return AV_PIX_FMT_RGB24;
    1098           0 :     case AV_PIX_FMT_ABGR:       return AV_PIX_FMT_BGR24;
    1099          29 :     case AV_PIX_FMT_BGRA:       return AV_PIX_FMT_BGR24;
    1100           2 :     case AV_PIX_FMT_YA8:        return AV_PIX_FMT_GRAY8;
    1101             : 
    1102           0 :     case AV_PIX_FMT_YUVA420P:   return AV_PIX_FMT_YUV420P;
    1103           3 :     case AV_PIX_FMT_YUVA422P:   return AV_PIX_FMT_YUV422P;
    1104           4 :     case AV_PIX_FMT_YUVA444P:           return AV_PIX_FMT_YUV444P;
    1105             : 
    1106           1 :     case AV_PIX_FMT_GBRAP:              return AV_PIX_FMT_GBRP;
    1107             : 
    1108           0 :     case AV_PIX_FMT_GBRAP10LE:          return AV_PIX_FMT_GBRP10;
    1109           0 :     case AV_PIX_FMT_GBRAP10BE:          return AV_PIX_FMT_GBRP10;
    1110             : 
    1111           0 :     case AV_PIX_FMT_GBRAP12LE:          return AV_PIX_FMT_GBRP12;
    1112           0 :     case AV_PIX_FMT_GBRAP12BE:          return AV_PIX_FMT_GBRP12;
    1113             : 
    1114           0 :     case AV_PIX_FMT_GBRAP16LE:          return AV_PIX_FMT_GBRP16;
    1115           1 :     case AV_PIX_FMT_GBRAP16BE:          return AV_PIX_FMT_GBRP16;
    1116             : 
    1117           0 :     case AV_PIX_FMT_RGBA64LE:   return AV_PIX_FMT_RGB48;
    1118           1 :     case AV_PIX_FMT_RGBA64BE:   return AV_PIX_FMT_RGB48;
    1119           0 :     case AV_PIX_FMT_BGRA64LE:   return AV_PIX_FMT_BGR48;
    1120           0 :     case AV_PIX_FMT_BGRA64BE:   return AV_PIX_FMT_BGR48;
    1121             : 
    1122           2 :     case AV_PIX_FMT_YA16BE:             return AV_PIX_FMT_GRAY16;
    1123           0 :     case AV_PIX_FMT_YA16LE:             return AV_PIX_FMT_GRAY16;
    1124             : 
    1125           0 :     case AV_PIX_FMT_YUVA420P9BE:        return AV_PIX_FMT_YUV420P9;
    1126           0 :     case AV_PIX_FMT_YUVA422P9BE:        return AV_PIX_FMT_YUV422P9;
    1127           0 :     case AV_PIX_FMT_YUVA444P9BE:        return AV_PIX_FMT_YUV444P9;
    1128           0 :     case AV_PIX_FMT_YUVA420P9LE:        return AV_PIX_FMT_YUV420P9;
    1129           0 :     case AV_PIX_FMT_YUVA422P9LE:        return AV_PIX_FMT_YUV422P9;
    1130           0 :     case AV_PIX_FMT_YUVA444P9LE:        return AV_PIX_FMT_YUV444P9;
    1131           0 :     case AV_PIX_FMT_YUVA420P10BE:       return AV_PIX_FMT_YUV420P10;
    1132           0 :     case AV_PIX_FMT_YUVA422P10BE:       return AV_PIX_FMT_YUV422P10;
    1133           0 :     case AV_PIX_FMT_YUVA444P10BE:       return AV_PIX_FMT_YUV444P10;
    1134           0 :     case AV_PIX_FMT_YUVA420P10LE:       return AV_PIX_FMT_YUV420P10;
    1135           0 :     case AV_PIX_FMT_YUVA422P10LE:       return AV_PIX_FMT_YUV422P10;
    1136           0 :     case AV_PIX_FMT_YUVA444P10LE:       return AV_PIX_FMT_YUV444P10;
    1137           0 :     case AV_PIX_FMT_YUVA420P16BE:       return AV_PIX_FMT_YUV420P16;
    1138           0 :     case AV_PIX_FMT_YUVA422P16BE:       return AV_PIX_FMT_YUV422P16;
    1139           0 :     case AV_PIX_FMT_YUVA444P16BE:       return AV_PIX_FMT_YUV444P16;
    1140           0 :     case AV_PIX_FMT_YUVA420P16LE:       return AV_PIX_FMT_YUV420P16;
    1141           1 :     case AV_PIX_FMT_YUVA422P16LE:       return AV_PIX_FMT_YUV422P16;
    1142           0 :     case AV_PIX_FMT_YUVA444P16LE:       return AV_PIX_FMT_YUV444P16;
    1143             : 
    1144             : //     case AV_PIX_FMT_AYUV64LE:
    1145             : //     case AV_PIX_FMT_AYUV64BE:
    1146             : //     case AV_PIX_FMT_PAL8:
    1147          80 :     default: return AV_PIX_FMT_NONE;
    1148             :     }
    1149             : }
    1150             : 
    1151        2678 : av_cold int sws_init_context(SwsContext *c, SwsFilter *srcFilter,
    1152             :                              SwsFilter *dstFilter)
    1153             : {
    1154             :     int i;
    1155             :     int usesVFilter, usesHFilter;
    1156             :     int unscaled;
    1157        2678 :     SwsFilter dummyFilter = { NULL, NULL, NULL, NULL };
    1158        2678 :     int srcW              = c->srcW;
    1159        2678 :     int srcH              = c->srcH;
    1160        2678 :     int dstW              = c->dstW;
    1161        2678 :     int dstH              = c->dstH;
    1162        2678 :     int dst_stride        = FFALIGN(dstW * sizeof(int16_t) + 66, 16);
    1163             :     int flags, cpu_flags;
    1164        2678 :     enum AVPixelFormat srcFormat = c->srcFormat;
    1165        2678 :     enum AVPixelFormat dstFormat = c->dstFormat;
    1166             :     const AVPixFmtDescriptor *desc_src;
    1167             :     const AVPixFmtDescriptor *desc_dst;
    1168        2678 :     int ret = 0;
    1169             :     enum AVPixelFormat tmpFmt;
    1170             : 
    1171        2678 :     cpu_flags = av_get_cpu_flags();
    1172        2678 :     flags     = c->flags;
    1173        2678 :     emms_c();
    1174        2678 :     if (!rgb15to16)
    1175        2478 :         ff_sws_rgb2rgb_init();
    1176             : 
    1177        2678 :     unscaled = (srcW == dstW && srcH == dstH);
    1178             : 
    1179        2678 :     c->srcRange |= handle_jpeg(&c->srcFormat);
    1180        2678 :     c->dstRange |= handle_jpeg(&c->dstFormat);
    1181             : 
    1182        2678 :     if(srcFormat!=c->srcFormat || dstFormat!=c->dstFormat)
    1183         141 :         av_log(c, AV_LOG_WARNING, "deprecated pixel format used, make sure you did set range correctly\n");
    1184             : 
    1185        2678 :     if (!c->contrast && !c->saturation && !c->dstFormatBpp)
    1186        2678 :         sws_setColorspaceDetails(c, ff_yuv2rgb_coeffs[SWS_CS_DEFAULT], c->srcRange,
    1187             :                                  ff_yuv2rgb_coeffs[SWS_CS_DEFAULT],
    1188             :                                  c->dstRange, 0, 1 << 16, 1 << 16);
    1189             : 
    1190        2678 :     handle_formats(c);
    1191        2678 :     srcFormat = c->srcFormat;
    1192        2678 :     dstFormat = c->dstFormat;
    1193        2678 :     desc_src = av_pix_fmt_desc_get(srcFormat);
    1194        2678 :     desc_dst = av_pix_fmt_desc_get(dstFormat);
    1195             : 
    1196             :     // If the source has no alpha then disable alpha blendaway
    1197        2678 :     if (c->src0Alpha)
    1198          25 :         c->alphablend = SWS_ALPHA_BLEND_NONE;
    1199             : 
    1200        2679 :     if (!(unscaled && sws_isSupportedEndiannessConversion(srcFormat) &&
    1201           1 :           av_pix_fmt_swap_endianness(srcFormat) == dstFormat)) {
    1202        2678 :     if (!sws_isSupportedInput(srcFormat)) {
    1203           0 :         av_log(c, AV_LOG_ERROR, "%s is not supported as input pixel format\n",
    1204             :                av_get_pix_fmt_name(srcFormat));
    1205           0 :         return AVERROR(EINVAL);
    1206             :     }
    1207        2678 :     if (!sws_isSupportedOutput(dstFormat)) {
    1208           0 :         av_log(c, AV_LOG_ERROR, "%s is not supported as output pixel format\n",
    1209             :                av_get_pix_fmt_name(dstFormat));
    1210           0 :         return AVERROR(EINVAL);
    1211             :     }
    1212             :     }
    1213             :     av_assert2(desc_src && desc_dst);
    1214             : 
    1215        2678 :     i = flags & (SWS_POINT         |
    1216             :                  SWS_AREA          |
    1217             :                  SWS_BILINEAR      |
    1218             :                  SWS_FAST_BILINEAR |
    1219             :                  SWS_BICUBIC       |
    1220             :                  SWS_X             |
    1221             :                  SWS_GAUSS         |
    1222             :                  SWS_LANCZOS       |
    1223             :                  SWS_SINC          |
    1224             :                  SWS_SPLINE        |
    1225             :                  SWS_BICUBLIN);
    1226             : 
    1227             :     /* provide a default scaler if not set by caller */
    1228        2678 :     if (!i) {
    1229          29 :         if (dstW < srcW && dstH < srcH)
    1230           8 :             flags |= SWS_BICUBIC;
    1231          21 :         else if (dstW > srcW && dstH > srcH)
    1232           1 :             flags |= SWS_BICUBIC;
    1233             :         else
    1234          20 :             flags |= SWS_BICUBIC;
    1235          29 :         c->flags = flags;
    1236        2649 :     } else if (i & (i - 1)) {
    1237           0 :         av_log(c, AV_LOG_ERROR,
    1238             :                "Exactly one scaler algorithm must be chosen, got %X\n", i);
    1239           0 :         return AVERROR(EINVAL);
    1240             :     }
    1241             :     /* sanity check */
    1242        2678 :     if (srcW < 1 || srcH < 1 || dstW < 1 || dstH < 1) {
    1243             :         /* FIXME check if these are enough and try to lower them after
    1244             :          * fixing the relevant parts of the code */
    1245           0 :         av_log(c, AV_LOG_ERROR, "%dx%d -> %dx%d is invalid scaling dimension\n",
    1246             :                srcW, srcH, dstW, dstH);
    1247           0 :         return AVERROR(EINVAL);
    1248             :     }
    1249        2678 :     if (flags & SWS_FAST_BILINEAR) {
    1250           0 :         if (srcW < 8 || dstW < 8) {
    1251           0 :             flags ^= SWS_FAST_BILINEAR | SWS_BILINEAR;
    1252           0 :             c->flags = flags;
    1253             :         }
    1254             :     }
    1255             : 
    1256        2678 :     if (!dstFilter)
    1257        2678 :         dstFilter = &dummyFilter;
    1258        2678 :     if (!srcFilter)
    1259        2678 :         srcFilter = &dummyFilter;
    1260             : 
    1261        2678 :     c->lumXInc      = (((int64_t)srcW << 16) + (dstW >> 1)) / dstW;
    1262        2678 :     c->lumYInc      = (((int64_t)srcH << 16) + (dstH >> 1)) / dstH;
    1263        2678 :     c->dstFormatBpp = av_get_bits_per_pixel(desc_dst);
    1264        2678 :     c->srcFormatBpp = av_get_bits_per_pixel(desc_src);
    1265        2678 :     c->vRounder     = 4 * 0x0001000100010001ULL;
    1266             : 
    1267        8034 :     usesVFilter = (srcFilter->lumV && srcFilter->lumV->length > 1) ||
    1268        5356 :                   (srcFilter->chrV && srcFilter->chrV->length > 1) ||
    1269        8034 :                   (dstFilter->lumV && dstFilter->lumV->length > 1) ||
    1270        2678 :                   (dstFilter->chrV && dstFilter->chrV->length > 1);
    1271        8034 :     usesHFilter = (srcFilter->lumH && srcFilter->lumH->length > 1) ||
    1272        5356 :                   (srcFilter->chrH && srcFilter->chrH->length > 1) ||
    1273        8034 :                   (dstFilter->lumH && dstFilter->lumH->length > 1) ||
    1274        2678 :                   (dstFilter->chrH && dstFilter->chrH->length > 1);
    1275             : 
    1276        2678 :     av_pix_fmt_get_chroma_sub_sample(srcFormat, &c->chrSrcHSubSample, &c->chrSrcVSubSample);
    1277        2678 :     av_pix_fmt_get_chroma_sub_sample(dstFormat, &c->chrDstHSubSample, &c->chrDstVSubSample);
    1278             : 
    1279        2678 :     if (isAnyRGB(dstFormat) && !(flags&SWS_FULL_CHR_H_INT)) {
    1280         996 :         if (dstW&1) {
    1281          30 :             av_log(c, AV_LOG_DEBUG, "Forcing full internal H chroma due to odd output size\n");
    1282          30 :             flags |= SWS_FULL_CHR_H_INT;
    1283          30 :             c->flags = flags;
    1284             :         }
    1285             : 
    1286         996 :         if (   c->chrSrcHSubSample == 0
    1287         221 :             && c->chrSrcVSubSample == 0
    1288         221 :             && c->dither != SWS_DITHER_BAYER //SWS_FULL_CHR_H_INT is currently not supported with SWS_DITHER_BAYER
    1289         221 :             && !(c->flags & SWS_FAST_BILINEAR)
    1290             :         ) {
    1291         221 :             av_log(c, AV_LOG_DEBUG, "Forcing full internal H chroma due to input having non subsampled chroma\n");
    1292         221 :             flags |= SWS_FULL_CHR_H_INT;
    1293         221 :             c->flags = flags;
    1294             :         }
    1295             :     }
    1296             : 
    1297        2678 :     if (c->dither == SWS_DITHER_AUTO) {
    1298        2678 :         if (flags & SWS_ERROR_DIFFUSION)
    1299           0 :             c->dither = SWS_DITHER_ED;
    1300             :     }
    1301             : 
    1302        2678 :     if(dstFormat == AV_PIX_FMT_BGR4_BYTE ||
    1303        2653 :        dstFormat == AV_PIX_FMT_RGB4_BYTE ||
    1304        2632 :        dstFormat == AV_PIX_FMT_BGR8 ||
    1305             :        dstFormat == AV_PIX_FMT_RGB8) {
    1306          60 :         if (c->dither == SWS_DITHER_AUTO)
    1307          60 :             c->dither = (flags & SWS_FULL_CHR_H_INT) ? SWS_DITHER_ED : SWS_DITHER_BAYER;
    1308          60 :         if (!(flags & SWS_FULL_CHR_H_INT)) {
    1309          50 :             if (c->dither == SWS_DITHER_ED || c->dither == SWS_DITHER_A_DITHER || c->dither == SWS_DITHER_X_DITHER) {
    1310           0 :                 av_log(c, AV_LOG_DEBUG,
    1311             :                     "Desired dithering only supported in full chroma interpolation for destination format '%s'\n",
    1312             :                     av_get_pix_fmt_name(dstFormat));
    1313           0 :                 flags   |= SWS_FULL_CHR_H_INT;
    1314           0 :                 c->flags = flags;
    1315             :             }
    1316             :         }
    1317          60 :         if (flags & SWS_FULL_CHR_H_INT) {
    1318          10 :             if (c->dither == SWS_DITHER_BAYER) {
    1319           0 :                 av_log(c, AV_LOG_DEBUG,
    1320             :                     "Ordered dither is not supported in full chroma interpolation for destination format '%s'\n",
    1321             :                     av_get_pix_fmt_name(dstFormat));
    1322           0 :                 c->dither = SWS_DITHER_ED;
    1323             :             }
    1324             :         }
    1325             :     }
    1326        2678 :     if (isPlanarRGB(dstFormat)) {
    1327         222 :         if (!(flags & SWS_FULL_CHR_H_INT)) {
    1328         204 :             av_log(c, AV_LOG_DEBUG,
    1329             :                    "%s output is not supported with half chroma resolution, switching to full\n",
    1330             :                    av_get_pix_fmt_name(dstFormat));
    1331         204 :             flags   |= SWS_FULL_CHR_H_INT;
    1332         204 :             c->flags = flags;
    1333             :         }
    1334             :     }
    1335             : 
    1336             :     /* reuse chroma for 2 pixels RGB/BGR unless user wants full
    1337             :      * chroma interpolation */
    1338        2678 :     if (flags & SWS_FULL_CHR_H_INT &&
    1339         874 :         isAnyRGB(dstFormat)        &&
    1340         652 :         !isPlanarRGB(dstFormat)    &&
    1341         212 :         dstFormat != AV_PIX_FMT_RGBA64LE &&
    1342         211 :         dstFormat != AV_PIX_FMT_RGBA64BE &&
    1343         210 :         dstFormat != AV_PIX_FMT_BGRA64LE &&
    1344         209 :         dstFormat != AV_PIX_FMT_BGRA64BE &&
    1345         206 :         dstFormat != AV_PIX_FMT_RGB48LE &&
    1346         203 :         dstFormat != AV_PIX_FMT_RGB48BE &&
    1347         202 :         dstFormat != AV_PIX_FMT_BGR48LE &&
    1348         201 :         dstFormat != AV_PIX_FMT_BGR48BE &&
    1349         194 :         dstFormat != AV_PIX_FMT_RGBA  &&
    1350         191 :         dstFormat != AV_PIX_FMT_ARGB  &&
    1351         157 :         dstFormat != AV_PIX_FMT_BGRA  &&
    1352         155 :         dstFormat != AV_PIX_FMT_ABGR  &&
    1353          35 :         dstFormat != AV_PIX_FMT_RGB24 &&
    1354          24 :         dstFormat != AV_PIX_FMT_BGR24 &&
    1355          22 :         dstFormat != AV_PIX_FMT_BGR4_BYTE &&
    1356          20 :         dstFormat != AV_PIX_FMT_RGB4_BYTE &&
    1357          16 :         dstFormat != AV_PIX_FMT_BGR8 &&
    1358             :         dstFormat != AV_PIX_FMT_RGB8
    1359             :     ) {
    1360          14 :         av_log(c, AV_LOG_WARNING,
    1361             :                "full chroma interpolation for destination format '%s' not yet implemented\n",
    1362             :                av_get_pix_fmt_name(dstFormat));
    1363          14 :         flags   &= ~SWS_FULL_CHR_H_INT;
    1364          14 :         c->flags = flags;
    1365             :     }
    1366        2678 :     if (isAnyRGB(dstFormat) && !(flags & SWS_FULL_CHR_H_INT))
    1367         585 :         c->chrDstHSubSample = 1;
    1368             : 
    1369             :     // drop some chroma lines if the user wants it
    1370        2678 :     c->vChrDrop          = (flags & SWS_SRC_V_CHR_DROP_MASK) >>
    1371             :                            SWS_SRC_V_CHR_DROP_SHIFT;
    1372        2678 :     c->chrSrcVSubSample += c->vChrDrop;
    1373             : 
    1374             :     /* drop every other pixel for chroma calculation unless user
    1375             :      * wants full chroma */
    1376        2678 :     if (isAnyRGB(srcFormat) && !(flags & SWS_FULL_CHR_H_INP)   &&
    1377         227 :         srcFormat != AV_PIX_FMT_RGB8 && srcFormat != AV_PIX_FMT_BGR8 &&
    1378         226 :         srcFormat != AV_PIX_FMT_RGB4 && srcFormat != AV_PIX_FMT_BGR4 &&
    1379         225 :         srcFormat != AV_PIX_FMT_RGB4_BYTE && srcFormat != AV_PIX_FMT_BGR4_BYTE &&
    1380         223 :         srcFormat != AV_PIX_FMT_GBRP9BE   && srcFormat != AV_PIX_FMT_GBRP9LE  &&
    1381         221 :         srcFormat != AV_PIX_FMT_GBRP10BE  && srcFormat != AV_PIX_FMT_GBRP10LE &&
    1382         218 :         srcFormat != AV_PIX_FMT_GBRAP10BE && srcFormat != AV_PIX_FMT_GBRAP10LE &&
    1383         216 :         srcFormat != AV_PIX_FMT_GBRP12BE  && srcFormat != AV_PIX_FMT_GBRP12LE &&
    1384         214 :         srcFormat != AV_PIX_FMT_GBRAP12BE && srcFormat != AV_PIX_FMT_GBRAP12LE &&
    1385         212 :         srcFormat != AV_PIX_FMT_GBRP14BE  && srcFormat != AV_PIX_FMT_GBRP14LE &&
    1386         209 :         srcFormat != AV_PIX_FMT_GBRP16BE  && srcFormat != AV_PIX_FMT_GBRP16LE &&
    1387         403 :         srcFormat != AV_PIX_FMT_GBRAP16BE  && srcFormat != AV_PIX_FMT_GBRAP16LE &&
    1388         304 :         ((dstW >> c->chrDstHSubSample) <= (srcW >> 1) ||
    1389         103 :          (flags & SWS_FAST_BILINEAR)))
    1390          98 :         c->chrSrcHSubSample = 1;
    1391             : 
    1392             :     // Note the AV_CEIL_RSHIFT is so that we always round toward +inf.
    1393        2678 :     c->chrSrcW = AV_CEIL_RSHIFT(srcW, c->chrSrcHSubSample);
    1394        2678 :     c->chrSrcH = AV_CEIL_RSHIFT(srcH, c->chrSrcVSubSample);
    1395        2678 :     c->chrDstW = AV_CEIL_RSHIFT(dstW, c->chrDstHSubSample);
    1396        2678 :     c->chrDstH = AV_CEIL_RSHIFT(dstH, c->chrDstVSubSample);
    1397             : 
    1398        2678 :     FF_ALLOCZ_OR_GOTO(c, c->formatConvBuffer, FFALIGN(srcW*2+78, 16) * 2, fail);
    1399             : 
    1400        2678 :     c->srcBpc = desc_src->comp[0].depth;
    1401        2678 :     if (c->srcBpc < 8)
    1402          56 :         c->srcBpc = 8;
    1403        2678 :     c->dstBpc = desc_dst->comp[0].depth;
    1404        2678 :     if (c->dstBpc < 8)
    1405         249 :         c->dstBpc = 8;
    1406        2678 :     if (isAnyRGB(srcFormat) || srcFormat == AV_PIX_FMT_PAL8)
    1407         316 :         c->srcBpc = 16;
    1408        2678 :     if (c->dstBpc == 16)
    1409         383 :         dst_stride <<= 1;
    1410             : 
    1411        2678 :     if (INLINE_MMXEXT(cpu_flags) && c->srcBpc == 8 && c->dstBpc <= 14) {
    1412           6 :         c->canMMXEXTBeUsed = dstW >= srcW && (dstW & 31) == 0 &&
    1413           6 :                              c->chrDstW >= c->chrSrcW &&
    1414           2 :                              (srcW & 15) == 0;
    1415           2 :         if (!c->canMMXEXTBeUsed && dstW >= srcW && c->chrDstW >= c->chrSrcW && (srcW & 15) == 0
    1416             : 
    1417           0 :             && (flags & SWS_FAST_BILINEAR)) {
    1418           0 :             if (flags & SWS_PRINT_INFO)
    1419           0 :                 av_log(c, AV_LOG_INFO,
    1420             :                        "output width is not a multiple of 32 -> no MMXEXT scaler\n");
    1421             :         }
    1422           4 :         if (usesHFilter || isNBPS(c->srcFormat) || is16BPS(c->srcFormat) || isAnyRGB(c->srcFormat))
    1423           0 :             c->canMMXEXTBeUsed = 0;
    1424             :     } else
    1425        2676 :         c->canMMXEXTBeUsed = 0;
    1426             : 
    1427        2678 :     c->chrXInc = (((int64_t)c->chrSrcW << 16) + (c->chrDstW >> 1)) / c->chrDstW;
    1428        2678 :     c->chrYInc = (((int64_t)c->chrSrcH << 16) + (c->chrDstH >> 1)) / c->chrDstH;
    1429             : 
    1430             :     /* Match pixel 0 of the src to pixel 0 of dst and match pixel n-2 of src
    1431             :      * to pixel n-2 of dst, but only for the FAST_BILINEAR mode otherwise do
    1432             :      * correct scaling.
    1433             :      * n-2 is the last chrominance sample available.
    1434             :      * This is not perfect, but no one should notice the difference, the more
    1435             :      * correct variant would be like the vertical one, but that would require
    1436             :      * some special code for the first and last pixel */
    1437        2678 :     if (flags & SWS_FAST_BILINEAR) {
    1438           0 :         if (c->canMMXEXTBeUsed) {
    1439           0 :             c->lumXInc += 20;
    1440           0 :             c->chrXInc += 20;
    1441             :         }
    1442             :         // we don't use the x86 asm scaler if MMX is available
    1443           0 :         else if (INLINE_MMX(cpu_flags) && c->dstBpc <= 14) {
    1444           0 :             c->lumXInc = ((int64_t)(srcW       - 2) << 16) / (dstW       - 2) - 20;
    1445           0 :             c->chrXInc = ((int64_t)(c->chrSrcW - 2) << 16) / (c->chrDstW - 2) - 20;
    1446             :         }
    1447             :     }
    1448             : 
    1449             :     // hardcoded for now
    1450        2678 :     c->gamma_value = 2.2;
    1451        2678 :     tmpFmt = AV_PIX_FMT_RGBA64LE;
    1452             : 
    1453             : 
    1454        2678 :     if (!unscaled && c->gamma_flag && (srcFormat != tmpFmt || dstFormat != tmpFmt)) {
    1455             :         SwsContext *c2;
    1456           0 :         c->cascaded_context[0] = NULL;
    1457             : 
    1458           0 :         ret = av_image_alloc(c->cascaded_tmp, c->cascaded_tmpStride,
    1459             :                             srcW, srcH, tmpFmt, 64);
    1460           0 :         if (ret < 0)
    1461           0 :             return ret;
    1462             : 
    1463           0 :         c->cascaded_context[0] = sws_getContext(srcW, srcH, srcFormat,
    1464             :                                                 srcW, srcH, tmpFmt,
    1465           0 :                                                 flags, NULL, NULL, c->param);
    1466           0 :         if (!c->cascaded_context[0]) {
    1467           0 :             return -1;
    1468             :         }
    1469             : 
    1470           0 :         c->cascaded_context[1] = sws_getContext(srcW, srcH, tmpFmt,
    1471             :                                                 dstW, dstH, tmpFmt,
    1472           0 :                                                 flags, srcFilter, dstFilter, c->param);
    1473             : 
    1474           0 :         if (!c->cascaded_context[1])
    1475           0 :             return -1;
    1476             : 
    1477           0 :         c2 = c->cascaded_context[1];
    1478           0 :         c2->is_internal_gamma = 1;
    1479           0 :         c2->gamma     = alloc_gamma_tbl(    c->gamma_value);
    1480           0 :         c2->inv_gamma = alloc_gamma_tbl(1.f/c->gamma_value);
    1481           0 :         if (!c2->gamma || !c2->inv_gamma)
    1482           0 :             return AVERROR(ENOMEM);
    1483             : 
    1484             :         // is_internal_flag is set after creating the context
    1485             :         // to properly create the gamma convert FilterDescriptor
    1486             :         // we have to re-initialize it
    1487           0 :         ff_free_filters(c2);
    1488           0 :         if (ff_init_filters(c2) < 0) {
    1489           0 :             sws_freeContext(c2);
    1490           0 :             return -1;
    1491             :         }
    1492             : 
    1493           0 :         c->cascaded_context[2] = NULL;
    1494           0 :         if (dstFormat != tmpFmt) {
    1495           0 :             ret = av_image_alloc(c->cascaded1_tmp, c->cascaded1_tmpStride,
    1496             :                                 dstW, dstH, tmpFmt, 64);
    1497           0 :             if (ret < 0)
    1498           0 :                 return ret;
    1499             : 
    1500           0 :             c->cascaded_context[2] = sws_getContext(dstW, dstH, tmpFmt,
    1501             :                                                 dstW, dstH, dstFormat,
    1502           0 :                                                 flags, NULL, NULL, c->param);
    1503           0 :             if (!c->cascaded_context[2])
    1504           0 :                 return -1;
    1505             :         }
    1506           0 :         return 0;
    1507             :     }
    1508             : 
    1509        2678 :     if (isBayer(srcFormat)) {
    1510           0 :         if (!unscaled ||
    1511           0 :             (dstFormat != AV_PIX_FMT_RGB24 && dstFormat != AV_PIX_FMT_YUV420P)) {
    1512           0 :             enum AVPixelFormat tmpFormat = AV_PIX_FMT_RGB24;
    1513             : 
    1514           0 :             ret = av_image_alloc(c->cascaded_tmp, c->cascaded_tmpStride,
    1515             :                                 srcW, srcH, tmpFormat, 64);
    1516           0 :             if (ret < 0)
    1517           0 :                 return ret;
    1518             : 
    1519           0 :             c->cascaded_context[0] = sws_getContext(srcW, srcH, srcFormat,
    1520             :                                                     srcW, srcH, tmpFormat,
    1521           0 :                                                     flags, srcFilter, NULL, c->param);
    1522           0 :             if (!c->cascaded_context[0])
    1523           0 :                 return -1;
    1524             : 
    1525           0 :             c->cascaded_context[1] = sws_getContext(srcW, srcH, tmpFormat,
    1526             :                                                     dstW, dstH, dstFormat,
    1527           0 :                                                     flags, NULL, dstFilter, c->param);
    1528           0 :             if (!c->cascaded_context[1])
    1529           0 :                 return -1;
    1530           0 :             return 0;
    1531             :         }
    1532             :     }
    1533             : 
    1534        2678 :     if (CONFIG_SWSCALE_ALPHA && isALPHA(srcFormat) && !isALPHA(dstFormat)) {
    1535         126 :         enum AVPixelFormat tmpFormat = alphaless_fmt(srcFormat);
    1536             : 
    1537         126 :         if (tmpFormat != AV_PIX_FMT_NONE && c->alphablend != SWS_ALPHA_BLEND_NONE)
    1538           0 :         if (!unscaled ||
    1539           0 :             dstFormat != tmpFormat ||
    1540           0 :             usesHFilter || usesVFilter ||
    1541           0 :             c->srcRange != c->dstRange
    1542             :         ) {
    1543           0 :             c->cascaded_mainindex = 1;
    1544           0 :             ret = av_image_alloc(c->cascaded_tmp, c->cascaded_tmpStride,
    1545             :                                 srcW, srcH, tmpFormat, 64);
    1546           0 :             if (ret < 0)
    1547           0 :                 return ret;
    1548             : 
    1549           0 :             c->cascaded_context[0] = sws_alloc_set_opts(srcW, srcH, srcFormat,
    1550             :                                                         srcW, srcH, tmpFormat,
    1551           0 :                                                         flags, c->param);
    1552           0 :             if (!c->cascaded_context[0])
    1553           0 :                 return -1;
    1554           0 :             c->cascaded_context[0]->alphablend = c->alphablend;
    1555           0 :             ret = sws_init_context(c->cascaded_context[0], NULL , NULL);
    1556           0 :             if (ret < 0)
    1557           0 :                 return ret;
    1558             : 
    1559           0 :             c->cascaded_context[1] = sws_alloc_set_opts(srcW, srcH, tmpFormat,
    1560             :                                                         dstW, dstH, dstFormat,
    1561           0 :                                                         flags, c->param);
    1562           0 :             if (!c->cascaded_context[1])
    1563           0 :                 return -1;
    1564             : 
    1565           0 :             c->cascaded_context[1]->srcRange = c->srcRange;
    1566           0 :             c->cascaded_context[1]->dstRange = c->dstRange;
    1567           0 :             ret = sws_init_context(c->cascaded_context[1], srcFilter , dstFilter);
    1568           0 :             if (ret < 0)
    1569           0 :                 return ret;
    1570             : 
    1571           0 :             return 0;
    1572             :         }
    1573             :     }
    1574             : 
    1575             : #define USE_MMAP (HAVE_MMAP && HAVE_MPROTECT && defined MAP_ANONYMOUS)
    1576             : 
    1577             :     /* precalculate horizontal scaler filter coefficients */
    1578             :     {
    1579             : #if HAVE_MMXEXT_INLINE
    1580             : // can't downscale !!!
    1581        2678 :         if (c->canMMXEXTBeUsed && (flags & SWS_FAST_BILINEAR)) {
    1582           0 :             c->lumMmxextFilterCodeSize = ff_init_hscaler_mmxext(dstW, c->lumXInc, NULL,
    1583             :                                                              NULL, NULL, 8);
    1584           0 :             c->chrMmxextFilterCodeSize = ff_init_hscaler_mmxext(c->chrDstW, c->chrXInc,
    1585             :                                                              NULL, NULL, NULL, 4);
    1586             : 
    1587             : #if USE_MMAP
    1588           0 :             c->lumMmxextFilterCode = mmap(NULL, c->lumMmxextFilterCodeSize,
    1589             :                                           PROT_READ | PROT_WRITE,
    1590             :                                           MAP_PRIVATE | MAP_ANONYMOUS,
    1591             :                                           -1, 0);
    1592           0 :             c->chrMmxextFilterCode = mmap(NULL, c->chrMmxextFilterCodeSize,
    1593             :                                           PROT_READ | PROT_WRITE,
    1594             :                                           MAP_PRIVATE | MAP_ANONYMOUS,
    1595             :                                           -1, 0);
    1596             : #elif HAVE_VIRTUALALLOC
    1597             :             c->lumMmxextFilterCode = VirtualAlloc(NULL,
    1598             :                                                   c->lumMmxextFilterCodeSize,
    1599             :                                                   MEM_COMMIT,
    1600             :                                                   PAGE_EXECUTE_READWRITE);
    1601             :             c->chrMmxextFilterCode = VirtualAlloc(NULL,
    1602             :                                                   c->chrMmxextFilterCodeSize,
    1603             :                                                   MEM_COMMIT,
    1604             :                                                   PAGE_EXECUTE_READWRITE);
    1605             : #else
    1606             :             c->lumMmxextFilterCode = av_malloc(c->lumMmxextFilterCodeSize);
    1607             :             c->chrMmxextFilterCode = av_malloc(c->chrMmxextFilterCodeSize);
    1608             : #endif
    1609             : 
    1610             : #ifdef MAP_ANONYMOUS
    1611           0 :             if (c->lumMmxextFilterCode == MAP_FAILED || c->chrMmxextFilterCode == MAP_FAILED)
    1612             : #else
    1613             :             if (!c->lumMmxextFilterCode || !c->chrMmxextFilterCode)
    1614             : #endif
    1615             :             {
    1616           0 :                 av_log(c, AV_LOG_ERROR, "Failed to allocate MMX2FilterCode\n");
    1617           0 :                 return AVERROR(ENOMEM);
    1618             :             }
    1619             : 
    1620           0 :             FF_ALLOCZ_OR_GOTO(c, c->hLumFilter,    (dstW           / 8 + 8) * sizeof(int16_t), fail);
    1621           0 :             FF_ALLOCZ_OR_GOTO(c, c->hChrFilter,    (c->chrDstW     / 4 + 8) * sizeof(int16_t), fail);
    1622           0 :             FF_ALLOCZ_OR_GOTO(c, c->hLumFilterPos, (dstW       / 2 / 8 + 8) * sizeof(int32_t), fail);
    1623           0 :             FF_ALLOCZ_OR_GOTO(c, c->hChrFilterPos, (c->chrDstW / 2 / 4 + 8) * sizeof(int32_t), fail);
    1624             : 
    1625           0 :             ff_init_hscaler_mmxext(      dstW, c->lumXInc, c->lumMmxextFilterCode,
    1626             :                                 c->hLumFilter, (uint32_t*)c->hLumFilterPos, 8);
    1627           0 :             ff_init_hscaler_mmxext(c->chrDstW, c->chrXInc, c->chrMmxextFilterCode,
    1628             :                                 c->hChrFilter, (uint32_t*)c->hChrFilterPos, 4);
    1629             : 
    1630             : #if USE_MMAP
    1631           0 :             if (   mprotect(c->lumMmxextFilterCode, c->lumMmxextFilterCodeSize, PROT_EXEC | PROT_READ) == -1
    1632           0 :                 || mprotect(c->chrMmxextFilterCode, c->chrMmxextFilterCodeSize, PROT_EXEC | PROT_READ) == -1) {
    1633           0 :                 av_log(c, AV_LOG_ERROR, "mprotect failed, cannot use fast bilinear scaler\n");
    1634           0 :                 goto fail;
    1635             :             }
    1636             : #endif
    1637             :         } else
    1638             : #endif /* HAVE_MMXEXT_INLINE */
    1639             :         {
    1640        2678 :             const int filterAlign = X86_MMX(cpu_flags)     ? 4 :
    1641             :                                     PPC_ALTIVEC(cpu_flags) ? 8 :
    1642             :                                     have_neon(cpu_flags)   ? 8 : 1;
    1643             : 
    1644        8034 :             if ((ret = initFilter(&c->hLumFilter, &c->hLumFilterPos,
    1645             :                            &c->hLumFilterSize, c->lumXInc,
    1646             :                            srcW, dstW, filterAlign, 1 << 14,
    1647        2678 :                            (flags & SWS_BICUBLIN) ? (flags | SWS_BICUBIC) : flags,
    1648             :                            cpu_flags, srcFilter->lumH, dstFilter->lumH,
    1649        2678 :                            c->param,
    1650             :                            get_local_pos(c, 0, 0, 0),
    1651             :                            get_local_pos(c, 0, 0, 0))) < 0)
    1652           0 :                 goto fail;
    1653        8034 :             if ((ret = initFilter(&c->hChrFilter, &c->hChrFilterPos,
    1654             :                            &c->hChrFilterSize, c->chrXInc,
    1655             :                            c->chrSrcW, c->chrDstW, filterAlign, 1 << 14,
    1656        2678 :                            (flags & SWS_BICUBLIN) ? (flags | SWS_BILINEAR) : flags,
    1657             :                            cpu_flags, srcFilter->chrH, dstFilter->chrH,
    1658        2678 :                            c->param,
    1659             :                            get_local_pos(c, c->chrSrcHSubSample, c->src_h_chr_pos, 0),
    1660             :                            get_local_pos(c, c->chrDstHSubSample, c->dst_h_chr_pos, 0))) < 0)
    1661           0 :                 goto fail;
    1662             :         }
    1663             :     } // initialize horizontal stuff
    1664             : 
    1665             :     /* precalculate vertical scaler filter coefficients */
    1666             :     {
    1667        2678 :         const int filterAlign = X86_MMX(cpu_flags)     ? 2 :
    1668             :                                 PPC_ALTIVEC(cpu_flags) ? 8 :
    1669             :                                 have_neon(cpu_flags)   ? 2 : 1;
    1670             : 
    1671        8034 :         if ((ret = initFilter(&c->vLumFilter, &c->vLumFilterPos, &c->vLumFilterSize,
    1672             :                        c->lumYInc, srcH, dstH, filterAlign, (1 << 12),
    1673        2678 :                        (flags & SWS_BICUBLIN) ? (flags | SWS_BICUBIC) : flags,
    1674             :                        cpu_flags, srcFilter->lumV, dstFilter->lumV,
    1675        2678 :                        c->param,
    1676             :                        get_local_pos(c, 0, 0, 1),
    1677             :                        get_local_pos(c, 0, 0, 1))) < 0)
    1678           0 :             goto fail;
    1679        8034 :         if ((ret = initFilter(&c->vChrFilter, &c->vChrFilterPos, &c->vChrFilterSize,
    1680             :                        c->chrYInc, c->chrSrcH, c->chrDstH,
    1681             :                        filterAlign, (1 << 12),
    1682        2678 :                        (flags & SWS_BICUBLIN) ? (flags | SWS_BILINEAR) : flags,
    1683             :                        cpu_flags, srcFilter->chrV, dstFilter->chrV,
    1684        2678 :                        c->param,
    1685             :                        get_local_pos(c, c->chrSrcVSubSample, c->src_v_chr_pos, 1),
    1686             :                        get_local_pos(c, c->chrDstVSubSample, c->dst_v_chr_pos, 1))) < 0)
    1687             : 
    1688           0 :             goto fail;
    1689             : 
    1690             : #if HAVE_ALTIVEC
    1691             :         FF_ALLOC_OR_GOTO(c, c->vYCoeffsBank, sizeof(vector signed short) * c->vLumFilterSize * c->dstH,    fail);
    1692             :         FF_ALLOC_OR_GOTO(c, c->vCCoeffsBank, sizeof(vector signed short) * c->vChrFilterSize * c->chrDstH, fail);
    1693             : 
    1694             :         for (i = 0; i < c->vLumFilterSize * c->dstH; i++) {
    1695             :             int j;
    1696             :             short *p = (short *)&c->vYCoeffsBank[i];
    1697             :             for (j = 0; j < 8; j++)
    1698             :                 p[j] = c->vLumFilter[i];
    1699             :         }
    1700             : 
    1701             :         for (i = 0; i < c->vChrFilterSize * c->chrDstH; i++) {
    1702             :             int j;
    1703             :             short *p = (short *)&c->vCCoeffsBank[i];
    1704             :             for (j = 0; j < 8; j++)
    1705             :                 p[j] = c->vChrFilter[i];
    1706             :         }
    1707             : #endif
    1708             :     }
    1709             : 
    1710       13390 :     for (i = 0; i < 4; i++)
    1711       10712 :         FF_ALLOCZ_OR_GOTO(c, c->dither_error[i], (c->dstW+2) * sizeof(int), fail);
    1712             : 
    1713        2678 :     c->needAlpha = (CONFIG_SWSCALE_ALPHA && isALPHA(c->srcFormat) && isALPHA(c->dstFormat)) ? 1 : 0;
    1714             : 
    1715             :     // 64 / c->scalingBpp is the same as 16 / sizeof(scaling_intermediate)
    1716        2678 :     c->uv_off   = (dst_stride>>1) + 64 / (c->dstBpc &~ 7);
    1717        2678 :     c->uv_offx2 = dst_stride + 16;
    1718             : 
    1719        2678 :     av_assert0(c->chrDstH <= dstH);
    1720             : 
    1721        2678 :     if (flags & SWS_PRINT_INFO) {
    1722           0 :         const char *scaler = NULL, *cpucaps;
    1723             : 
    1724           0 :         for (i = 0; i < FF_ARRAY_ELEMS(scale_algorithms); i++) {
    1725           0 :             if (flags & scale_algorithms[i].flag) {
    1726           0 :                 scaler = scale_algorithms[i].description;
    1727           0 :                 break;
    1728             :             }
    1729             :         }
    1730           0 :         if (!scaler)
    1731           0 :             scaler =  "ehh flags invalid?!";
    1732           0 :         av_log(c, AV_LOG_INFO, "%s scaler, from %s to %s%s ",
    1733             :                scaler,
    1734             :                av_get_pix_fmt_name(srcFormat),
    1735             : #ifdef DITHER1XBPP
    1736           0 :                dstFormat == AV_PIX_FMT_BGR555   || dstFormat == AV_PIX_FMT_BGR565   ||
    1737           0 :                dstFormat == AV_PIX_FMT_RGB444BE || dstFormat == AV_PIX_FMT_RGB444LE ||
    1738           0 :                dstFormat == AV_PIX_FMT_BGR444BE || dstFormat == AV_PIX_FMT_BGR444LE ?
    1739             :                                                              "dithered " : "",
    1740             : #else
    1741             :                "",
    1742             : #endif
    1743             :                av_get_pix_fmt_name(dstFormat));
    1744             : 
    1745           0 :         if (INLINE_MMXEXT(cpu_flags))
    1746           0 :             cpucaps = "MMXEXT";
    1747           0 :         else if (INLINE_AMD3DNOW(cpu_flags))
    1748           0 :             cpucaps = "3DNOW";
    1749           0 :         else if (INLINE_MMX(cpu_flags))
    1750           0 :             cpucaps = "MMX";
    1751             :         else if (PPC_ALTIVEC(cpu_flags))
    1752             :             cpucaps = "AltiVec";
    1753             :         else
    1754           0 :             cpucaps = "C";
    1755             : 
    1756           0 :         av_log(c, AV_LOG_INFO, "using %s\n", cpucaps);
    1757             : 
    1758           0 :         av_log(c, AV_LOG_VERBOSE, "%dx%d -> %dx%d\n", srcW, srcH, dstW, dstH);
    1759           0 :         av_log(c, AV_LOG_DEBUG,
    1760             :                "lum srcW=%d srcH=%d dstW=%d dstH=%d xInc=%d yInc=%d\n",
    1761             :                c->srcW, c->srcH, c->dstW, c->dstH, c->lumXInc, c->lumYInc);
    1762           0 :         av_log(c, AV_LOG_DEBUG,
    1763             :                "chr srcW=%d srcH=%d dstW=%d dstH=%d xInc=%d yInc=%d\n",
    1764             :                c->chrSrcW, c->chrSrcH, c->chrDstW, c->chrDstH,
    1765             :                c->chrXInc, c->chrYInc);
    1766             :     }
    1767             : 
    1768             :     /* alpha blend special case, note this has been split via cascaded contexts if its scaled */
    1769        4993 :     if (unscaled && !usesHFilter && !usesVFilter &&
    1770        2315 :         c->alphablend != SWS_ALPHA_BLEND_NONE &&
    1771           0 :         isALPHA(srcFormat) &&
    1772           0 :         (c->srcRange == c->dstRange || isAnyRGB(dstFormat)) &&
    1773           0 :         alphaless_fmt(srcFormat) == dstFormat
    1774             :     ) {
    1775           0 :         c->swscale = ff_sws_alphablendaway;
    1776             : 
    1777           0 :         if (flags & SWS_PRINT_INFO)
    1778           0 :             av_log(c, AV_LOG_INFO,
    1779             :                     "using alpha blendaway %s -> %s special converter\n",
    1780             :                     av_get_pix_fmt_name(srcFormat), av_get_pix_fmt_name(dstFormat));
    1781           0 :         return 0;
    1782             :     }
    1783             : 
    1784             :     /* unscaled special cases */
    1785        4993 :     if (unscaled && !usesHFilter && !usesVFilter &&
    1786        2522 :         (c->srcRange == c->dstRange || isAnyRGB(dstFormat))) {
    1787        2118 :         ff_get_unscaled_swscale(c);
    1788             : 
    1789        2118 :         if (c->swscale) {
    1790         431 :             if (flags & SWS_PRINT_INFO)
    1791           0 :                 av_log(c, AV_LOG_INFO,
    1792             :                        "using unscaled %s -> %s special converter\n",
    1793             :                        av_get_pix_fmt_name(srcFormat), av_get_pix_fmt_name(dstFormat));
    1794         431 :             return 0;
    1795             :         }
    1796             :     }
    1797             : 
    1798        2247 :     c->swscale = ff_getSwsFunc(c);
    1799        2247 :     return ff_init_filters(c);
    1800             : fail: // FIXME replace things by appropriate error codes
    1801           0 :     if (ret == RETCODE_USE_CASCADE)  {
    1802           0 :         int tmpW = sqrt(srcW * (int64_t)dstW);
    1803           0 :         int tmpH = sqrt(srcH * (int64_t)dstH);
    1804           0 :         enum AVPixelFormat tmpFormat = AV_PIX_FMT_YUV420P;
    1805             : 
    1806           0 :         if (isALPHA(srcFormat))
    1807           0 :             tmpFormat = AV_PIX_FMT_YUVA420P;
    1808             : 
    1809           0 :         if (srcW*(int64_t)srcH <= 4LL*dstW*dstH)
    1810           0 :             return AVERROR(EINVAL);
    1811             : 
    1812           0 :         ret = av_image_alloc(c->cascaded_tmp, c->cascaded_tmpStride,
    1813             :                              tmpW, tmpH, tmpFormat, 64);
    1814           0 :         if (ret < 0)
    1815           0 :             return ret;
    1816             : 
    1817           0 :         c->cascaded_context[0] = sws_getContext(srcW, srcH, srcFormat,
    1818             :                                                 tmpW, tmpH, tmpFormat,
    1819           0 :                                                 flags, srcFilter, NULL, c->param);
    1820           0 :         if (!c->cascaded_context[0])
    1821           0 :             return -1;
    1822             : 
    1823           0 :         c->cascaded_context[1] = sws_getContext(tmpW, tmpH, tmpFormat,
    1824             :                                                 dstW, dstH, dstFormat,
    1825           0 :                                                 flags, NULL, dstFilter, c->param);
    1826           0 :         if (!c->cascaded_context[1])
    1827           0 :             return -1;
    1828           0 :         return 0;
    1829             :     }
    1830           0 :     return -1;
    1831             : }
    1832             : 
    1833           0 : SwsContext *sws_alloc_set_opts(int srcW, int srcH, enum AVPixelFormat srcFormat,
    1834             :                                int dstW, int dstH, enum AVPixelFormat dstFormat,
    1835             :                                int flags, const double *param)
    1836             : {
    1837             :     SwsContext *c;
    1838             : 
    1839           0 :     if (!(c = sws_alloc_context()))
    1840           0 :         return NULL;
    1841             : 
    1842           0 :     c->flags     = flags;
    1843           0 :     c->srcW      = srcW;
    1844           0 :     c->srcH      = srcH;
    1845           0 :     c->dstW      = dstW;
    1846           0 :     c->dstH      = dstH;
    1847           0 :     c->srcFormat = srcFormat;
    1848           0 :     c->dstFormat = dstFormat;
    1849             : 
    1850           0 :     if (param) {
    1851           0 :         c->param[0] = param[0];
    1852           0 :         c->param[1] = param[1];
    1853             :     }
    1854             : 
    1855           0 :     return c;
    1856             : }
    1857             : 
    1858           0 : SwsContext *sws_getContext(int srcW, int srcH, enum AVPixelFormat srcFormat,
    1859             :                            int dstW, int dstH, enum AVPixelFormat dstFormat,
    1860             :                            int flags, SwsFilter *srcFilter,
    1861             :                            SwsFilter *dstFilter, const double *param)
    1862             : {
    1863             :     SwsContext *c;
    1864             : 
    1865           0 :     c = sws_alloc_set_opts(srcW, srcH, srcFormat,
    1866             :                            dstW, dstH, dstFormat,
    1867             :                            flags, param);
    1868           0 :     if (!c)
    1869           0 :         return NULL;
    1870             : 
    1871           0 :     if (sws_init_context(c, srcFilter, dstFilter) < 0) {
    1872           0 :         sws_freeContext(c);
    1873           0 :         return NULL;
    1874             :     }
    1875             : 
    1876           0 :     return c;
    1877             : }
    1878             : 
    1879           0 : static int isnan_vec(SwsVector *a)
    1880             : {
    1881             :     int i;
    1882           0 :     for (i=0; i<a->length; i++)
    1883           0 :         if (isnan(a->coeff[i]))
    1884           0 :             return 1;
    1885           0 :     return 0;
    1886             : }
    1887             : 
    1888           0 : static void makenan_vec(SwsVector *a)
    1889             : {
    1890             :     int i;
    1891           0 :     for (i=0; i<a->length; i++)
    1892           0 :         a->coeff[i] = NAN;
    1893           0 : }
    1894             : 
    1895           0 : SwsFilter *sws_getDefaultFilter(float lumaGBlur, float chromaGBlur,
    1896             :                                 float lumaSharpen, float chromaSharpen,
    1897             :                                 float chromaHShift, float chromaVShift,
    1898             :                                 int verbose)
    1899             : {
    1900           0 :     SwsFilter *filter = av_malloc(sizeof(SwsFilter));
    1901           0 :     if (!filter)
    1902           0 :         return NULL;
    1903             : 
    1904           0 :     if (lumaGBlur != 0.0) {
    1905           0 :         filter->lumH = sws_getGaussianVec(lumaGBlur, 3.0);
    1906           0 :         filter->lumV = sws_getGaussianVec(lumaGBlur, 3.0);
    1907             :     } else {
    1908           0 :         filter->lumH = sws_getIdentityVec();
    1909           0 :         filter->lumV = sws_getIdentityVec();
    1910             :     }
    1911             : 
    1912           0 :     if (chromaGBlur != 0.0) {
    1913           0 :         filter->chrH = sws_getGaussianVec(chromaGBlur, 3.0);
    1914           0 :         filter->chrV = sws_getGaussianVec(chromaGBlur, 3.0);
    1915             :     } else {
    1916           0 :         filter->chrH = sws_getIdentityVec();
    1917           0 :         filter->chrV = sws_getIdentityVec();
    1918             :     }
    1919             : 
    1920           0 :     if (!filter->lumH || !filter->lumV || !filter->chrH || !filter->chrV)
    1921             :         goto fail;
    1922             : 
    1923           0 :     if (chromaSharpen != 0.0) {
    1924           0 :         SwsVector *id = sws_getIdentityVec();
    1925           0 :         if (!id)
    1926           0 :             goto fail;
    1927           0 :         sws_scaleVec(filter->chrH, -chromaSharpen);
    1928           0 :         sws_scaleVec(filter->chrV, -chromaSharpen);
    1929           0 :         sws_addVec(filter->chrH, id);
    1930           0 :         sws_addVec(filter->chrV, id);
    1931           0 :         sws_freeVec(id);
    1932             :     }
    1933             : 
    1934           0 :     if (lumaSharpen != 0.0) {
    1935           0 :         SwsVector *id = sws_getIdentityVec();
    1936           0 :         if (!id)
    1937           0 :             goto fail;
    1938           0 :         sws_scaleVec(filter->lumH, -lumaSharpen);
    1939           0 :         sws_scaleVec(filter->lumV, -lumaSharpen);
    1940           0 :         sws_addVec(filter->lumH, id);
    1941           0 :         sws_addVec(filter->lumV, id);
    1942           0 :         sws_freeVec(id);
    1943             :     }
    1944             : 
    1945           0 :     if (chromaHShift != 0.0)
    1946           0 :         sws_shiftVec(filter->chrH, (int)(chromaHShift + 0.5));
    1947             : 
    1948           0 :     if (chromaVShift != 0.0)
    1949           0 :         sws_shiftVec(filter->chrV, (int)(chromaVShift + 0.5));
    1950             : 
    1951           0 :     sws_normalizeVec(filter->chrH, 1.0);
    1952           0 :     sws_normalizeVec(filter->chrV, 1.0);
    1953           0 :     sws_normalizeVec(filter->lumH, 1.0);
    1954           0 :     sws_normalizeVec(filter->lumV, 1.0);
    1955             : 
    1956           0 :     if (isnan_vec(filter->chrH) ||
    1957           0 :         isnan_vec(filter->chrV) ||
    1958           0 :         isnan_vec(filter->lumH) ||
    1959           0 :         isnan_vec(filter->lumV))
    1960             :         goto fail;
    1961             : 
    1962           0 :     if (verbose)
    1963           0 :         sws_printVec2(filter->chrH, NULL, AV_LOG_DEBUG);
    1964           0 :     if (verbose)
    1965           0 :         sws_printVec2(filter->lumH, NULL, AV_LOG_DEBUG);
    1966             : 
    1967           0 :     return filter;
    1968             : 
    1969             : fail:
    1970           0 :     sws_freeVec(filter->lumH);
    1971           0 :     sws_freeVec(filter->lumV);
    1972           0 :     sws_freeVec(filter->chrH);
    1973           0 :     sws_freeVec(filter->chrV);
    1974           0 :     av_freep(&filter);
    1975           0 :     return NULL;
    1976             : }
    1977             : 
    1978           0 : SwsVector *sws_allocVec(int length)
    1979             : {
    1980             :     SwsVector *vec;
    1981             : 
    1982           0 :     if(length <= 0 || length > INT_MAX/ sizeof(double))
    1983           0 :         return NULL;
    1984             : 
    1985           0 :     vec = av_malloc(sizeof(SwsVector));
    1986           0 :     if (!vec)
    1987           0 :         return NULL;
    1988           0 :     vec->length = length;
    1989           0 :     vec->coeff  = av_malloc(sizeof(double) * length);
    1990           0 :     if (!vec->coeff)
    1991           0 :         av_freep(&vec);
    1992           0 :     return vec;
    1993             : }
    1994             : 
    1995           0 : SwsVector *sws_getGaussianVec(double variance, double quality)
    1996             : {
    1997           0 :     const int length = (int)(variance * quality + 0.5) | 1;
    1998             :     int i;
    1999           0 :     double middle  = (length - 1) * 0.5;
    2000             :     SwsVector *vec;
    2001             : 
    2002           0 :     if(variance < 0 || quality < 0)
    2003           0 :         return NULL;
    2004             : 
    2005           0 :     vec = sws_allocVec(length);
    2006             : 
    2007           0 :     if (!vec)
    2008           0 :         return NULL;
    2009             : 
    2010           0 :     for (i = 0; i < length; i++) {
    2011           0 :         double dist = i - middle;
    2012           0 :         vec->coeff[i] = exp(-dist * dist / (2 * variance * variance)) /
    2013           0 :                         sqrt(2 * variance * M_PI);
    2014             :     }
    2015             : 
    2016           0 :     sws_normalizeVec(vec, 1.0);
    2017             : 
    2018           0 :     return vec;
    2019             : }
    2020             : 
    2021             : /**
    2022             :  * Allocate and return a vector with length coefficients, all
    2023             :  * with the same value c.
    2024             :  */
    2025             : #if !FF_API_SWS_VECTOR
    2026             : static
    2027             : #endif
    2028           0 : SwsVector *sws_getConstVec(double c, int length)
    2029             : {
    2030             :     int i;
    2031           0 :     SwsVector *vec = sws_allocVec(length);
    2032             : 
    2033           0 :     if (!vec)
    2034           0 :         return NULL;
    2035             : 
    2036           0 :     for (i = 0; i < length; i++)
    2037           0 :         vec->coeff[i] = c;
    2038             : 
    2039           0 :     return vec;
    2040             : }
    2041             : 
    2042             : /**
    2043             :  * Allocate and return a vector with just one coefficient, with
    2044             :  * value 1.0.
    2045             :  */
    2046             : #if !FF_API_SWS_VECTOR
    2047             : static
    2048             : #endif
    2049           0 : SwsVector *sws_getIdentityVec(void)
    2050             : {
    2051           0 :     return sws_getConstVec(1.0, 1);
    2052             : }
    2053             : 
    2054           0 : static double sws_dcVec(SwsVector *a)
    2055             : {
    2056             :     int i;
    2057           0 :     double sum = 0;
    2058             : 
    2059           0 :     for (i = 0; i < a->length; i++)
    2060           0 :         sum += a->coeff[i];
    2061             : 
    2062           0 :     return sum;
    2063             : }
    2064             : 
    2065           0 : void sws_scaleVec(SwsVector *a, double scalar)
    2066             : {
    2067             :     int i;
    2068             : 
    2069           0 :     for (i = 0; i < a->length; i++)
    2070           0 :         a->coeff[i] *= scalar;
    2071           0 : }
    2072             : 
    2073           0 : void sws_normalizeVec(SwsVector *a, double height)
    2074             : {
    2075           0 :     sws_scaleVec(a, height / sws_dcVec(a));
    2076           0 : }
    2077             : 
    2078             : #if FF_API_SWS_VECTOR
    2079           0 : static SwsVector *sws_getConvVec(SwsVector *a, SwsVector *b)
    2080             : {
    2081           0 :     int length = a->length + b->length - 1;
    2082             :     int i, j;
    2083           0 :     SwsVector *vec = sws_getConstVec(0.0, length);
    2084             : 
    2085           0 :     if (!vec)
    2086           0 :         return NULL;
    2087             : 
    2088           0 :     for (i = 0; i < a->length; i++) {
    2089           0 :         for (j = 0; j < b->length; j++) {
    2090           0 :             vec->coeff[i + j] += a->coeff[i] * b->coeff[j];
    2091             :         }
    2092             :     }
    2093             : 
    2094           0 :     return vec;
    2095             : }
    2096             : #endif
    2097             : 
    2098           0 : static SwsVector *sws_sumVec(SwsVector *a, SwsVector *b)
    2099             : {
    2100           0 :     int length = FFMAX(a->length, b->length);
    2101             :     int i;
    2102           0 :     SwsVector *vec = sws_getConstVec(0.0, length);
    2103             : 
    2104           0 :     if (!vec)
    2105           0 :         return NULL;
    2106             : 
    2107           0 :     for (i = 0; i < a->length; i++)
    2108           0 :         vec->coeff[i + (length - 1) / 2 - (a->length - 1) / 2] += a->coeff[i];
    2109           0 :     for (i = 0; i < b->length; i++)
    2110           0 :         vec->coeff[i + (length - 1) / 2 - (b->length - 1) / 2] += b->coeff[i];
    2111             : 
    2112           0 :     return vec;
    2113             : }
    2114             : 
    2115             : #if FF_API_SWS_VECTOR
    2116           0 : static SwsVector *sws_diffVec(SwsVector *a, SwsVector *b)
    2117             : {
    2118           0 :     int length = FFMAX(a->length, b->length);
    2119             :     int i;
    2120           0 :     SwsVector *vec = sws_getConstVec(0.0, length);
    2121             : 
    2122           0 :     if (!vec)
    2123           0 :         return NULL;
    2124             : 
    2125           0 :     for (i = 0; i < a->length; i++)
    2126           0 :         vec->coeff[i + (length - 1) / 2 - (a->length - 1) / 2] += a->coeff[i];
    2127           0 :     for (i = 0; i < b->length; i++)
    2128           0 :         vec->coeff[i + (length - 1) / 2 - (b->length - 1) / 2] -= b->coeff[i];
    2129             : 
    2130           0 :     return vec;
    2131             : }
    2132             : #endif
    2133             : 
    2134             : /* shift left / or right if "shift" is negative */
    2135           0 : static SwsVector *sws_getShiftedVec(SwsVector *a, int shift)
    2136             : {
    2137           0 :     int length = a->length + FFABS(shift) * 2;
    2138             :     int i;
    2139           0 :     SwsVector *vec = sws_getConstVec(0.0, length);
    2140             : 
    2141           0 :     if (!vec)
    2142           0 :         return NULL;
    2143             : 
    2144           0 :     for (i = 0; i < a->length; i++) {
    2145           0 :         vec->coeff[i + (length    - 1) / 2 -
    2146           0 :                        (a->length - 1) / 2 - shift] = a->coeff[i];
    2147             :     }
    2148             : 
    2149           0 :     return vec;
    2150             : }
    2151             : 
    2152             : #if !FF_API_SWS_VECTOR
    2153             : static
    2154             : #endif
    2155           0 : void sws_shiftVec(SwsVector *a, int shift)
    2156             : {
    2157           0 :     SwsVector *shifted = sws_getShiftedVec(a, shift);
    2158           0 :     if (!shifted) {
    2159           0 :         makenan_vec(a);
    2160           0 :         return;
    2161             :     }
    2162           0 :     av_free(a->coeff);
    2163           0 :     a->coeff  = shifted->coeff;
    2164           0 :     a->length = shifted->length;
    2165           0 :     av_free(shifted);
    2166             : }
    2167             : 
    2168             : #if !FF_API_SWS_VECTOR
    2169             : static
    2170             : #endif
    2171           0 : void sws_addVec(SwsVector *a, SwsVector *b)
    2172             : {
    2173           0 :     SwsVector *sum = sws_sumVec(a, b);
    2174           0 :     if (!sum) {
    2175           0 :         makenan_vec(a);
    2176           0 :         return;
    2177             :     }
    2178           0 :     av_free(a->coeff);
    2179           0 :     a->coeff  = sum->coeff;
    2180           0 :     a->length = sum->length;
    2181           0 :     av_free(sum);
    2182             : }
    2183             : 
    2184             : #if FF_API_SWS_VECTOR
    2185           0 : void sws_subVec(SwsVector *a, SwsVector *b)
    2186             : {
    2187           0 :     SwsVector *diff = sws_diffVec(a, b);
    2188           0 :     if (!diff) {
    2189           0 :         makenan_vec(a);
    2190           0 :         return;
    2191             :     }
    2192           0 :     av_free(a->coeff);
    2193           0 :     a->coeff  = diff->coeff;
    2194           0 :     a->length = diff->length;
    2195           0 :     av_free(diff);
    2196             : }
    2197             : 
    2198           0 : void sws_convVec(SwsVector *a, SwsVector *b)
    2199             : {
    2200           0 :     SwsVector *conv = sws_getConvVec(a, b);
    2201           0 :     if (!conv) {
    2202           0 :         makenan_vec(a);
    2203           0 :         return;
    2204             :     }
    2205           0 :     av_free(a->coeff);
    2206           0 :     a->coeff  = conv->coeff;
    2207           0 :     a->length = conv->length;
    2208           0 :     av_free(conv);
    2209             : }
    2210             : 
    2211           0 : SwsVector *sws_cloneVec(SwsVector *a)
    2212             : {
    2213           0 :     SwsVector *vec = sws_allocVec(a->length);
    2214             : 
    2215           0 :     if (!vec)
    2216           0 :         return NULL;
    2217             : 
    2218           0 :     memcpy(vec->coeff, a->coeff, a->length * sizeof(*a->coeff));
    2219             : 
    2220           0 :     return vec;
    2221             : }
    2222             : #endif
    2223             : 
    2224             : /**
    2225             :  * Print with av_log() a textual representation of the vector a
    2226             :  * if log_level <= av_log_level.
    2227             :  */
    2228             : #if !FF_API_SWS_VECTOR
    2229             : static
    2230             : #endif
    2231           0 : void sws_printVec2(SwsVector *a, AVClass *log_ctx, int log_level)
    2232             : {
    2233             :     int i;
    2234           0 :     double max = 0;
    2235           0 :     double min = 0;
    2236             :     double range;
    2237             : 
    2238           0 :     for (i = 0; i < a->length; i++)
    2239           0 :         if (a->coeff[i] > max)
    2240           0 :             max = a->coeff[i];
    2241             : 
    2242           0 :     for (i = 0; i < a->length; i++)
    2243           0 :         if (a->coeff[i] < min)
    2244           0 :             min = a->coeff[i];
    2245             : 
    2246           0 :     range = max - min;
    2247             : 
    2248           0 :     for (i = 0; i < a->length; i++) {
    2249           0 :         int x = (int)((a->coeff[i] - min) * 60.0 / range + 0.5);
    2250           0 :         av_log(log_ctx, log_level, "%1.3f ", a->coeff[i]);
    2251           0 :         for (; x > 0; x--)
    2252           0 :             av_log(log_ctx, log_level, " ");
    2253           0 :         av_log(log_ctx, log_level, "|\n");
    2254             :     }
    2255           0 : }
    2256             : 
    2257           0 : void sws_freeVec(SwsVector *a)
    2258             : {
    2259           0 :     if (!a)
    2260           0 :         return;
    2261           0 :     av_freep(&a->coeff);
    2262           0 :     a->length = 0;
    2263           0 :     av_free(a);
    2264             : }
    2265             : 
    2266           0 : void sws_freeFilter(SwsFilter *filter)
    2267             : {
    2268           0 :     if (!filter)
    2269           0 :         return;
    2270             : 
    2271           0 :     sws_freeVec(filter->lumH);
    2272           0 :     sws_freeVec(filter->lumV);
    2273           0 :     sws_freeVec(filter->chrH);
    2274           0 :     sws_freeVec(filter->chrV);
    2275           0 :     av_free(filter);
    2276             : }
    2277             : 
    2278       40645 : void sws_freeContext(SwsContext *c)
    2279             : {
    2280             :     int i;
    2281       40645 :     if (!c)
    2282       31983 :         return;
    2283             : 
    2284       43310 :     for (i = 0; i < 4; i++)
    2285       34648 :         av_freep(&c->dither_error[i]);
    2286             : 
    2287        8662 :     av_freep(&c->vLumFilter);
    2288        8662 :     av_freep(&c->vChrFilter);
    2289        8662 :     av_freep(&c->hLumFilter);
    2290        8662 :     av_freep(&c->hChrFilter);
    2291             : #if HAVE_ALTIVEC
    2292             :     av_freep(&c->vYCoeffsBank);
    2293             :     av_freep(&c->vCCoeffsBank);
    2294             : #endif
    2295             : 
    2296        8662 :     av_freep(&c->vLumFilterPos);
    2297        8662 :     av_freep(&c->vChrFilterPos);
    2298        8662 :     av_freep(&c->hLumFilterPos);
    2299        8662 :     av_freep(&c->hChrFilterPos);
    2300             : 
    2301             : #if HAVE_MMX_INLINE
    2302             : #if USE_MMAP
    2303        8662 :     if (c->lumMmxextFilterCode)
    2304           0 :         munmap(c->lumMmxextFilterCode, c->lumMmxextFilterCodeSize);
    2305        8662 :     if (c->chrMmxextFilterCode)
    2306           0 :         munmap(c->chrMmxextFilterCode, c->chrMmxextFilterCodeSize);
    2307             : #elif HAVE_VIRTUALALLOC
    2308             :     if (c->lumMmxextFilterCode)
    2309             :         VirtualFree(c->lumMmxextFilterCode, 0, MEM_RELEASE);
    2310             :     if (c->chrMmxextFilterCode)
    2311             :         VirtualFree(c->chrMmxextFilterCode, 0, MEM_RELEASE);
    2312             : #else
    2313             :     av_free(c->lumMmxextFilterCode);
    2314             :     av_free(c->chrMmxextFilterCode);
    2315             : #endif
    2316        8662 :     c->lumMmxextFilterCode = NULL;
    2317        8662 :     c->chrMmxextFilterCode = NULL;
    2318             : #endif /* HAVE_MMX_INLINE */
    2319             : 
    2320        8662 :     av_freep(&c->yuvTable);
    2321        8662 :     av_freep(&c->formatConvBuffer);
    2322             : 
    2323        8662 :     sws_freeContext(c->cascaded_context[0]);
    2324        8662 :     sws_freeContext(c->cascaded_context[1]);
    2325        8662 :     sws_freeContext(c->cascaded_context[2]);
    2326        8662 :     memset(c->cascaded_context, 0, sizeof(c->cascaded_context));
    2327        8662 :     av_freep(&c->cascaded_tmp[0]);
    2328        8662 :     av_freep(&c->cascaded1_tmp[0]);
    2329             : 
    2330        8662 :     av_freep(&c->gamma);
    2331        8662 :     av_freep(&c->inv_gamma);
    2332             : 
    2333        8662 :     ff_free_filters(c);
    2334             : 
    2335        8662 :     av_free(c);
    2336             : }
    2337             : 
    2338           0 : struct SwsContext *sws_getCachedContext(struct SwsContext *context, int srcW,
    2339             :                                         int srcH, enum AVPixelFormat srcFormat,
    2340             :                                         int dstW, int dstH,
    2341             :                                         enum AVPixelFormat dstFormat, int flags,
    2342             :                                         SwsFilter *srcFilter,
    2343             :                                         SwsFilter *dstFilter,
    2344             :                                         const double *param)
    2345             : {
    2346             :     static const double default_param[2] = { SWS_PARAM_DEFAULT,
    2347             :                                              SWS_PARAM_DEFAULT };
    2348           0 :     int64_t src_h_chr_pos = -513, dst_h_chr_pos = -513,
    2349           0 :             src_v_chr_pos = -513, dst_v_chr_pos = -513;
    2350             : 
    2351           0 :     if (!param)
    2352           0 :         param = default_param;
    2353             : 
    2354           0 :     if (context &&
    2355           0 :         (context->srcW      != srcW      ||
    2356           0 :          context->srcH      != srcH      ||
    2357           0 :          context->srcFormat != srcFormat ||
    2358           0 :          context->dstW      != dstW      ||
    2359           0 :          context->dstH      != dstH      ||
    2360           0 :          context->dstFormat != dstFormat ||
    2361           0 :          context->flags     != flags     ||
    2362           0 :          context->param[0]  != param[0]  ||
    2363           0 :          context->param[1]  != param[1])) {
    2364             : 
    2365           0 :         av_opt_get_int(context, "src_h_chr_pos", 0, &src_h_chr_pos);
    2366           0 :         av_opt_get_int(context, "src_v_chr_pos", 0, &src_v_chr_pos);
    2367           0 :         av_opt_get_int(context, "dst_h_chr_pos", 0, &dst_h_chr_pos);
    2368           0 :         av_opt_get_int(context, "dst_v_chr_pos", 0, &dst_v_chr_pos);
    2369           0 :         sws_freeContext(context);
    2370           0 :         context = NULL;
    2371             :     }
    2372             : 
    2373           0 :     if (!context) {
    2374           0 :         if (!(context = sws_alloc_context()))
    2375           0 :             return NULL;
    2376           0 :         context->srcW      = srcW;
    2377           0 :         context->srcH      = srcH;
    2378           0 :         context->srcFormat = srcFormat;
    2379           0 :         context->dstW      = dstW;
    2380           0 :         context->dstH      = dstH;
    2381           0 :         context->dstFormat = dstFormat;
    2382           0 :         context->flags     = flags;
    2383           0 :         context->param[0]  = param[0];
    2384           0 :         context->param[1]  = param[1];
    2385             : 
    2386           0 :         av_opt_set_int(context, "src_h_chr_pos", src_h_chr_pos, 0);
    2387           0 :         av_opt_set_int(context, "src_v_chr_pos", src_v_chr_pos, 0);
    2388           0 :         av_opt_set_int(context, "dst_h_chr_pos", dst_h_chr_pos, 0);
    2389           0 :         av_opt_set_int(context, "dst_v_chr_pos", dst_v_chr_pos, 0);
    2390             : 
    2391           0 :         if (sws_init_context(context, srcFilter, dstFilter) < 0) {
    2392           0 :             sws_freeContext(context);
    2393           0 :             return NULL;
    2394             :         }
    2395             :     }
    2396           0 :     return context;
    2397             : }

Generated by: LCOV version 1.12