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/intreadwrite.h"
47 #include "libavutil/mathematics.h"
48 #include "libavutil/opt.h"
49 #include "libavutil/pixdesc.h"
50 #include "libavutil/ppc/cpu.h"
51 #include "libavutil/x86/asm.h"
52 #include "libavutil/x86/cpu.h"
53 #include "rgb2rgb.h"
54 #include "swscale.h"
55 #include "swscale_internal.h"
56
57 static void handle_formats(SwsContext *c);
58
swscale_version(void)59 unsigned swscale_version(void)
60 {
61 av_assert0(LIBSWSCALE_VERSION_MICRO >= 100);
62 return LIBSWSCALE_VERSION_INT;
63 }
64
swscale_configuration(void)65 const char *swscale_configuration(void)
66 {
67 return FFMPEG_CONFIGURATION;
68 }
69
swscale_license(void)70 const char *swscale_license(void)
71 {
72 #define LICENSE_PREFIX "libswscale license: "
73 return LICENSE_PREFIX FFMPEG_LICENSE + sizeof(LICENSE_PREFIX) - 1;
74 }
75
76 typedef struct FormatEntry {
77 uint8_t is_supported_in :1;
78 uint8_t is_supported_out :1;
79 uint8_t is_supported_endianness :1;
80 } FormatEntry;
81
82 static const FormatEntry format_entries[AV_PIX_FMT_NB] = {
83 [AV_PIX_FMT_YUV420P] = { 1, 1 },
84 [AV_PIX_FMT_YUYV422] = { 1, 1 },
85 [AV_PIX_FMT_RGB24] = { 1, 1 },
86 [AV_PIX_FMT_BGR24] = { 1, 1 },
87 [AV_PIX_FMT_YUV422P] = { 1, 1 },
88 [AV_PIX_FMT_YUV444P] = { 1, 1 },
89 [AV_PIX_FMT_YUV410P] = { 1, 1 },
90 [AV_PIX_FMT_YUV411P] = { 1, 1 },
91 [AV_PIX_FMT_GRAY8] = { 1, 1 },
92 [AV_PIX_FMT_MONOWHITE] = { 1, 1 },
93 [AV_PIX_FMT_MONOBLACK] = { 1, 1 },
94 [AV_PIX_FMT_PAL8] = { 1, 0 },
95 [AV_PIX_FMT_YUVJ420P] = { 1, 1 },
96 [AV_PIX_FMT_YUVJ411P] = { 1, 1 },
97 [AV_PIX_FMT_YUVJ422P] = { 1, 1 },
98 [AV_PIX_FMT_YUVJ444P] = { 1, 1 },
99 [AV_PIX_FMT_YVYU422] = { 1, 1 },
100 [AV_PIX_FMT_UYVY422] = { 1, 1 },
101 [AV_PIX_FMT_UYYVYY411] = { 0, 0 },
102 [AV_PIX_FMT_BGR8] = { 1, 1 },
103 [AV_PIX_FMT_BGR4] = { 0, 1 },
104 [AV_PIX_FMT_BGR4_BYTE] = { 1, 1 },
105 [AV_PIX_FMT_RGB8] = { 1, 1 },
106 [AV_PIX_FMT_RGB4] = { 0, 1 },
107 [AV_PIX_FMT_RGB4_BYTE] = { 1, 1 },
108 [AV_PIX_FMT_NV12] = { 1, 1 },
109 [AV_PIX_FMT_NV21] = { 1, 1 },
110 [AV_PIX_FMT_ARGB] = { 1, 1 },
111 [AV_PIX_FMT_RGBA] = { 1, 1 },
112 [AV_PIX_FMT_ABGR] = { 1, 1 },
113 [AV_PIX_FMT_BGRA] = { 1, 1 },
114 [AV_PIX_FMT_0RGB] = { 1, 1 },
115 [AV_PIX_FMT_RGB0] = { 1, 1 },
116 [AV_PIX_FMT_0BGR] = { 1, 1 },
117 [AV_PIX_FMT_BGR0] = { 1, 1 },
118 [AV_PIX_FMT_GRAY16BE] = { 1, 1 },
119 [AV_PIX_FMT_GRAY16LE] = { 1, 1 },
120 [AV_PIX_FMT_YUV440P] = { 1, 1 },
121 [AV_PIX_FMT_YUVJ440P] = { 1, 1 },
122 [AV_PIX_FMT_YUVA420P] = { 1, 1 },
123 [AV_PIX_FMT_YUVA422P] = { 1, 1 },
124 [AV_PIX_FMT_YUVA444P] = { 1, 1 },
125 [AV_PIX_FMT_YUVA420P9BE] = { 1, 1 },
126 [AV_PIX_FMT_YUVA420P9LE] = { 1, 1 },
127 [AV_PIX_FMT_YUVA422P9BE] = { 1, 1 },
128 [AV_PIX_FMT_YUVA422P9LE] = { 1, 1 },
129 [AV_PIX_FMT_YUVA444P9BE] = { 1, 1 },
130 [AV_PIX_FMT_YUVA444P9LE] = { 1, 1 },
131 [AV_PIX_FMT_YUVA420P10BE]= { 1, 1 },
132 [AV_PIX_FMT_YUVA420P10LE]= { 1, 1 },
133 [AV_PIX_FMT_YUVA422P10BE]= { 1, 1 },
134 [AV_PIX_FMT_YUVA422P10LE]= { 1, 1 },
135 [AV_PIX_FMT_YUVA444P10BE]= { 1, 1 },
136 [AV_PIX_FMT_YUVA444P10LE]= { 1, 1 },
137 [AV_PIX_FMT_YUVA420P16BE]= { 1, 1 },
138 [AV_PIX_FMT_YUVA420P16LE]= { 1, 1 },
139 [AV_PIX_FMT_YUVA422P16BE]= { 1, 1 },
140 [AV_PIX_FMT_YUVA422P16LE]= { 1, 1 },
141 [AV_PIX_FMT_YUVA444P16BE]= { 1, 1 },
142 [AV_PIX_FMT_YUVA444P16LE]= { 1, 1 },
143 [AV_PIX_FMT_RGB48BE] = { 1, 1 },
144 [AV_PIX_FMT_RGB48LE] = { 1, 1 },
145 [AV_PIX_FMT_RGBA64BE] = { 1, 1, 1 },
146 [AV_PIX_FMT_RGBA64LE] = { 1, 1, 1 },
147 [AV_PIX_FMT_RGB565BE] = { 1, 1 },
148 [AV_PIX_FMT_RGB565LE] = { 1, 1 },
149 [AV_PIX_FMT_RGB555BE] = { 1, 1 },
150 [AV_PIX_FMT_RGB555LE] = { 1, 1 },
151 [AV_PIX_FMT_BGR565BE] = { 1, 1 },
152 [AV_PIX_FMT_BGR565LE] = { 1, 1 },
153 [AV_PIX_FMT_BGR555BE] = { 1, 1 },
154 [AV_PIX_FMT_BGR555LE] = { 1, 1 },
155 [AV_PIX_FMT_YUV420P16LE] = { 1, 1 },
156 [AV_PIX_FMT_YUV420P16BE] = { 1, 1 },
157 [AV_PIX_FMT_YUV422P16LE] = { 1, 1 },
158 [AV_PIX_FMT_YUV422P16BE] = { 1, 1 },
159 [AV_PIX_FMT_YUV444P16LE] = { 1, 1 },
160 [AV_PIX_FMT_YUV444P16BE] = { 1, 1 },
161 [AV_PIX_FMT_RGB444LE] = { 1, 1 },
162 [AV_PIX_FMT_RGB444BE] = { 1, 1 },
163 [AV_PIX_FMT_BGR444LE] = { 1, 1 },
164 [AV_PIX_FMT_BGR444BE] = { 1, 1 },
165 [AV_PIX_FMT_YA8] = { 1, 0 },
166 [AV_PIX_FMT_YA16BE] = { 1, 0 },
167 [AV_PIX_FMT_YA16LE] = { 1, 0 },
168 [AV_PIX_FMT_BGR48BE] = { 1, 1 },
169 [AV_PIX_FMT_BGR48LE] = { 1, 1 },
170 [AV_PIX_FMT_BGRA64BE] = { 1, 1, 1 },
171 [AV_PIX_FMT_BGRA64LE] = { 1, 1, 1 },
172 [AV_PIX_FMT_YUV420P9BE] = { 1, 1 },
173 [AV_PIX_FMT_YUV420P9LE] = { 1, 1 },
174 [AV_PIX_FMT_YUV420P10BE] = { 1, 1 },
175 [AV_PIX_FMT_YUV420P10LE] = { 1, 1 },
176 [AV_PIX_FMT_YUV420P12BE] = { 1, 1 },
177 [AV_PIX_FMT_YUV420P12LE] = { 1, 1 },
178 [AV_PIX_FMT_YUV420P14BE] = { 1, 1 },
179 [AV_PIX_FMT_YUV420P14LE] = { 1, 1 },
180 [AV_PIX_FMT_YUV422P9BE] = { 1, 1 },
181 [AV_PIX_FMT_YUV422P9LE] = { 1, 1 },
182 [AV_PIX_FMT_YUV422P10BE] = { 1, 1 },
183 [AV_PIX_FMT_YUV422P10LE] = { 1, 1 },
184 [AV_PIX_FMT_YUV422P12BE] = { 1, 1 },
185 [AV_PIX_FMT_YUV422P12LE] = { 1, 1 },
186 [AV_PIX_FMT_YUV422P14BE] = { 1, 1 },
187 [AV_PIX_FMT_YUV422P14LE] = { 1, 1 },
188 [AV_PIX_FMT_YUV444P9BE] = { 1, 1 },
189 [AV_PIX_FMT_YUV444P9LE] = { 1, 1 },
190 [AV_PIX_FMT_YUV444P10BE] = { 1, 1 },
191 [AV_PIX_FMT_YUV444P10LE] = { 1, 1 },
192 [AV_PIX_FMT_YUV444P12BE] = { 1, 1 },
193 [AV_PIX_FMT_YUV444P12LE] = { 1, 1 },
194 [AV_PIX_FMT_YUV444P14BE] = { 1, 1 },
195 [AV_PIX_FMT_YUV444P14LE] = { 1, 1 },
196 [AV_PIX_FMT_GBRP] = { 1, 1 },
197 [AV_PIX_FMT_GBRP9LE] = { 1, 1 },
198 [AV_PIX_FMT_GBRP9BE] = { 1, 1 },
199 [AV_PIX_FMT_GBRP10LE] = { 1, 1 },
200 [AV_PIX_FMT_GBRP10BE] = { 1, 1 },
201 [AV_PIX_FMT_GBRP12LE] = { 1, 1 },
202 [AV_PIX_FMT_GBRP12BE] = { 1, 1 },
203 [AV_PIX_FMT_GBRP14LE] = { 1, 1 },
204 [AV_PIX_FMT_GBRP14BE] = { 1, 1 },
205 [AV_PIX_FMT_GBRP16LE] = { 1, 0 },
206 [AV_PIX_FMT_GBRP16BE] = { 1, 0 },
207 [AV_PIX_FMT_XYZ12BE] = { 1, 1, 1 },
208 [AV_PIX_FMT_XYZ12LE] = { 1, 1, 1 },
209 [AV_PIX_FMT_GBRAP] = { 1, 1 },
210 [AV_PIX_FMT_GBRAP16LE] = { 1, 0 },
211 [AV_PIX_FMT_GBRAP16BE] = { 1, 0 },
212 [AV_PIX_FMT_BAYER_BGGR8] = { 1, 0 },
213 [AV_PIX_FMT_BAYER_RGGB8] = { 1, 0 },
214 [AV_PIX_FMT_BAYER_GBRG8] = { 1, 0 },
215 [AV_PIX_FMT_BAYER_GRBG8] = { 1, 0 },
216 [AV_PIX_FMT_BAYER_BGGR16LE] = { 1, 0 },
217 [AV_PIX_FMT_BAYER_BGGR16BE] = { 1, 0 },
218 [AV_PIX_FMT_BAYER_RGGB16LE] = { 1, 0 },
219 [AV_PIX_FMT_BAYER_RGGB16BE] = { 1, 0 },
220 [AV_PIX_FMT_BAYER_GBRG16LE] = { 1, 0 },
221 [AV_PIX_FMT_BAYER_GBRG16BE] = { 1, 0 },
222 [AV_PIX_FMT_BAYER_GRBG16LE] = { 1, 0 },
223 [AV_PIX_FMT_BAYER_GRBG16BE] = { 1, 0 },
224 };
225
sws_isSupportedInput(enum AVPixelFormat pix_fmt)226 int sws_isSupportedInput(enum AVPixelFormat pix_fmt)
227 {
228 return (unsigned)pix_fmt < AV_PIX_FMT_NB ?
229 format_entries[pix_fmt].is_supported_in : 0;
230 }
231
sws_isSupportedOutput(enum AVPixelFormat pix_fmt)232 int sws_isSupportedOutput(enum AVPixelFormat pix_fmt)
233 {
234 return (unsigned)pix_fmt < AV_PIX_FMT_NB ?
235 format_entries[pix_fmt].is_supported_out : 0;
236 }
237
sws_isSupportedEndiannessConversion(enum AVPixelFormat pix_fmt)238 int sws_isSupportedEndiannessConversion(enum AVPixelFormat pix_fmt)
239 {
240 return (unsigned)pix_fmt < AV_PIX_FMT_NB ?
241 format_entries[pix_fmt].is_supported_endianness : 0;
242 }
243
244 #if FF_API_SWS_FORMAT_NAME
sws_format_name(enum AVPixelFormat format)245 const char *sws_format_name(enum AVPixelFormat format)
246 {
247 const AVPixFmtDescriptor *desc = av_pix_fmt_desc_get(format);
248 if (desc)
249 return desc->name;
250 else
251 return "Unknown format";
252 }
253 #endif
254
getSplineCoeff(double a,double b,double c,double d,double dist)255 static double getSplineCoeff(double a, double b, double c, double d,
256 double dist)
257 {
258 if (dist <= 1.0)
259 return ((d * dist + c) * dist + b) * dist + a;
260 else
261 return getSplineCoeff(0.0,
262 b + 2.0 * c + 3.0 * d,
263 c + 3.0 * d,
264 -b - 3.0 * c - 6.0 * d,
265 dist - 1.0);
266 }
267
get_local_pos(SwsContext * s,int chr_subsample,int pos,int dir)268 static av_cold int get_local_pos(SwsContext *s, int chr_subsample, int pos, int dir)
269 {
270 if (pos == -1 || pos <= -513) {
271 pos = (128 << chr_subsample) - 128;
272 }
273 pos += 128; // relative to ideal left edge
274 return pos >> chr_subsample;
275 }
276
277 typedef struct {
278 int flag; ///< flag associated to the algorithm
279 const char *description; ///< human-readable description
280 int size_factor; ///< size factor used when initing the filters
281 } ScaleAlgorithm;
282
283 static const ScaleAlgorithm scale_algorithms[] = {
284 { SWS_AREA, "area averaging", 1 /* downscale only, for upscale it is bilinear */ },
285 { SWS_BICUBIC, "bicubic", 4 },
286 { SWS_BICUBLIN, "luma bicubic / chroma bilinear", -1 },
287 { SWS_BILINEAR, "bilinear", 2 },
288 { SWS_FAST_BILINEAR, "fast bilinear", -1 },
289 { SWS_GAUSS, "Gaussian", 8 /* infinite ;) */ },
290 { SWS_LANCZOS, "Lanczos", -1 /* custom */ },
291 { SWS_POINT, "nearest neighbor / point", -1 },
292 { SWS_SINC, "sinc", 20 /* infinite ;) */ },
293 { SWS_SPLINE, "bicubic spline", 20 /* infinite :)*/ },
294 { SWS_X, "experimental", 8 },
295 };
296
initFilter(int16_t ** outFilter,int32_t ** filterPos,int * outFilterSize,int xInc,int srcW,int dstW,int filterAlign,int one,int flags,int cpu_flags,SwsVector * srcFilter,SwsVector * dstFilter,double param[2],int srcPos,int dstPos)297 static av_cold int initFilter(int16_t **outFilter, int32_t **filterPos,
298 int *outFilterSize, int xInc, int srcW,
299 int dstW, int filterAlign, int one,
300 int flags, int cpu_flags,
301 SwsVector *srcFilter, SwsVector *dstFilter,
302 double param[2], int srcPos, int dstPos)
303 {
304 int i;
305 int filterSize;
306 int filter2Size;
307 int minFilterSize;
308 int64_t *filter = NULL;
309 int64_t *filter2 = NULL;
310 const int64_t fone = LLN(1) << (54 - FFMIN(av_log2(srcW/dstW), 8));
311 int ret = -1;
312
313 emms_c(); // FIXME should not be required but IS (even for non-MMX versions)
314
315 // NOTE: the +3 is for the MMX(+1) / SSE(+3) scaler which reads over the end
316 FF_ALLOC_ARRAY_OR_GOTO(NULL, *filterPos, (dstW + 3), sizeof(**filterPos), fail);
317
318 if (FFABS(xInc - 0x10000) < 10 && srcPos == dstPos) { // unscaled
319 int i;
320 filterSize = 1;
321 FF_ALLOCZ_ARRAY_OR_GOTO(NULL, filter,
322 dstW, sizeof(*filter) * filterSize, fail);
323
324 for (i = 0; i < dstW; i++) {
325 filter[i * filterSize] = fone;
326 (*filterPos)[i] = i;
327 }
328 } else if (flags & SWS_POINT) { // lame looking point sampling mode
329 int i;
330 int64_t xDstInSrc;
331 filterSize = 1;
332 FF_ALLOC_ARRAY_OR_GOTO(NULL, filter,
333 dstW, sizeof(*filter) * filterSize, fail);
334
335 xDstInSrc = ((dstPos*(int64_t)xInc)>>8) - ((srcPos*LLN(0x8000))>>7);
336 for (i = 0; i < dstW; i++) {
337 int xx = (xDstInSrc - ((filterSize - 1) << 15) + (1 << 15)) >> 16;
338
339 (*filterPos)[i] = xx;
340 filter[i] = fone;
341 xDstInSrc += xInc;
342 }
343 } else if ((xInc <= (1 << 16) && (flags & SWS_AREA)) ||
344 (flags & SWS_FAST_BILINEAR)) { // bilinear upscale
345 int i;
346 int64_t xDstInSrc;
347 filterSize = 2;
348 FF_ALLOC_ARRAY_OR_GOTO(NULL, filter,
349 dstW, sizeof(*filter) * filterSize, fail);
350
351 xDstInSrc = ((dstPos*(int64_t)xInc)>>8) - ((srcPos*LLN(0x8000))>>7);
352 for (i = 0; i < dstW; i++) {
353 int xx = (xDstInSrc - ((filterSize - 1) << 15) + (1 << 15)) >> 16;
354 int j;
355
356 (*filterPos)[i] = xx;
357 // bilinear upscale / linear interpolate / area averaging
358 for (j = 0; j < filterSize; j++) {
359 int64_t coeff= fone - FFABS(((int64_t)xx<<16) - xDstInSrc)*(fone>>16);
360 if (coeff < 0)
361 coeff = 0;
362 filter[i * filterSize + j] = coeff;
363 xx++;
364 }
365 xDstInSrc += xInc;
366 }
367 } else {
368 int64_t xDstInSrc;
369 int sizeFactor = -1;
370
371 for (i = 0; i < FF_ARRAY_ELEMS(scale_algorithms); i++) {
372 if (flags & scale_algorithms[i].flag && scale_algorithms[i].size_factor > 0) {
373 sizeFactor = scale_algorithms[i].size_factor;
374 break;
375 }
376 }
377 if (flags & SWS_LANCZOS)
378 sizeFactor = param[0] != SWS_PARAM_DEFAULT ? ceil(2 * param[0]) : 6;
379 av_assert0(sizeFactor > 0);
380
381 if (xInc <= 1 << 16)
382 filterSize = 1 + sizeFactor; // upscale
383 else
384 filterSize = 1 + (sizeFactor * srcW + dstW - 1) / dstW;
385
386 filterSize = FFMIN(filterSize, srcW - 2);
387 filterSize = FFMAX(filterSize, 1);
388
389 FF_ALLOC_ARRAY_OR_GOTO(NULL, filter,
390 dstW, sizeof(*filter) * filterSize, fail);
391
392 xDstInSrc = ((dstPos*(int64_t)xInc)>>7) - ((srcPos*LLN(0x10000))>>7);
393 for (i = 0; i < dstW; i++) {
394 int xx = (xDstInSrc - ((int64_t)(filterSize - 2) << 16)) / (1 << 17);
395 int j;
396 (*filterPos)[i] = xx;
397 for (j = 0; j < filterSize; j++) {
398 int64_t d = (FFABS(((int64_t)xx << 17) - xDstInSrc)) << 13;
399 double floatd;
400 int64_t coeff;
401
402 if (xInc > 1 << 16)
403 d = d * dstW / srcW;
404 floatd = d * (1.0 / (1 << 30));
405
406 if (flags & SWS_BICUBIC) {
407 int64_t B = (param[0] != SWS_PARAM_DEFAULT ? param[0] : 0) * (1 << 24);
408 int64_t C = (param[1] != SWS_PARAM_DEFAULT ? param[1] : 0.6) * (1 << 24);
409
410 if (d >= LLN(1) << 31) {
411 coeff = 0.0;
412 } else {
413 int64_t dd = (d * d) >> 30;
414 int64_t ddd = (dd * d) >> 30;
415
416 if (d < LLN(1) << 30)
417 coeff = (12 * (1 << 24) - 9 * B - 6 * C) * ddd +
418 (-18 * (1 << 24) + 12 * B + 6 * C) * dd +
419 (6 * (1 << 24) - 2 * B) * (1 << 30);
420 else
421 coeff = (-B - 6 * C) * ddd +
422 (6 * B + 30 * C) * dd +
423 (-12 * B - 48 * C) * d +
424 (8 * B + 24 * C) * (1 << 30);
425 }
426 coeff /= (LLN(1)<<54)/fone;
427 }
428 #if 0
429 else if (flags & SWS_X) {
430 double p = param ? param * 0.01 : 0.3;
431 coeff = d ? sin(d * M_PI) / (d * M_PI) : 1.0;
432 coeff *= pow(2.0, -p * d * d);
433 }
434 #endif
435 else if (flags & SWS_X) {
436 double A = param[0] != SWS_PARAM_DEFAULT ? param[0] : 1.0;
437 double c;
438
439 if (floatd < 1.0)
440 c = cos(floatd * M_PI);
441 else
442 c = -1.0;
443 if (c < 0.0)
444 c = -pow(-c, A);
445 else
446 c = pow(c, A);
447 coeff = (c * 0.5 + 0.5) * fone;
448 } else if (flags & SWS_AREA) {
449 int64_t d2 = d - (1 << 29);
450 if (d2 * xInc < -(LLN(1) << (29 + 16)))
451 coeff = 1.0 * (LLN(1) << (30 + 16));
452 else if (d2 * xInc < (LLN(1) << (29 + 16)))
453 coeff = -d2 * xInc + (LLN(1) << (29 + 16));
454 else
455 coeff = 0.0;
456 coeff *= fone >> (30 + 16);
457 } else if (flags & SWS_GAUSS) {
458 double p = param[0] != SWS_PARAM_DEFAULT ? param[0] : 3.0;
459 coeff = (pow(2.0, -p * floatd * floatd)) * fone;
460 } else if (flags & SWS_SINC) {
461 coeff = (d ? sin(floatd * M_PI) / (floatd * M_PI) : 1.0) * fone;
462 } else if (flags & SWS_LANCZOS) {
463 double p = param[0] != SWS_PARAM_DEFAULT ? param[0] : 3.0;
464 coeff = (d ? sin(floatd * M_PI) * sin(floatd * M_PI / p) /
465 (floatd * floatd * M_PI * M_PI / p) : 1.0) * fone;
466 if (floatd > p)
467 coeff = 0;
468 } else if (flags & SWS_BILINEAR) {
469 coeff = (1 << 30) - d;
470 if (coeff < 0)
471 coeff = 0;
472 coeff *= fone >> 30;
473 } else if (flags & SWS_SPLINE) {
474 double p = -2.196152422706632;
475 coeff = getSplineCoeff(1.0, 0.0, p, -p - 1.0, floatd) * fone;
476 } else {
477 av_assert0(0);
478 }
479
480 filter[i * filterSize + j] = coeff;
481 xx++;
482 }
483 xDstInSrc += 2 * xInc;
484 }
485 }
486
487 /* apply src & dst Filter to filter -> filter2
488 * av_free(filter);
489 */
490 av_assert0(filterSize > 0);
491 filter2Size = filterSize;
492 if (srcFilter)
493 filter2Size += srcFilter->length - 1;
494 if (dstFilter)
495 filter2Size += dstFilter->length - 1;
496 av_assert0(filter2Size > 0);
497 FF_ALLOCZ_ARRAY_OR_GOTO(NULL, filter2, dstW, filter2Size * sizeof(*filter2), fail);
498
499 for (i = 0; i < dstW; i++) {
500 int j, k;
501
502 if (srcFilter) {
503 for (k = 0; k < srcFilter->length; k++) {
504 for (j = 0; j < filterSize; j++)
505 filter2[i * filter2Size + k + j] +=
506 srcFilter->coeff[k] * filter[i * filterSize + j];
507 }
508 } else {
509 for (j = 0; j < filterSize; j++)
510 filter2[i * filter2Size + j] = filter[i * filterSize + j];
511 }
512 // FIXME dstFilter
513
514 (*filterPos)[i] += (filterSize - 1) / 2 - (filter2Size - 1) / 2;
515 }
516 av_freep(&filter);
517
518 /* try to reduce the filter-size (step1 find size and shift left) */
519 // Assume it is near normalized (*0.5 or *2.0 is OK but * 0.001 is not).
520 minFilterSize = 0;
521 for (i = dstW - 1; i >= 0; i--) {
522 int min = filter2Size;
523 int j;
524 int64_t cutOff = 0.0;
525
526 /* get rid of near zero elements on the left by shifting left */
527 for (j = 0; j < filter2Size; j++) {
528 int k;
529 cutOff += FFABS(filter2[i * filter2Size]);
530
531 if (cutOff > SWS_MAX_REDUCE_CUTOFF * fone)
532 break;
533
534 /* preserve monotonicity because the core can't handle the
535 * filter otherwise */
536 if (i < dstW - 1 && (*filterPos)[i] >= (*filterPos)[i + 1])
537 break;
538
539 // move filter coefficients left
540 for (k = 1; k < filter2Size; k++)
541 filter2[i * filter2Size + k - 1] = filter2[i * filter2Size + k];
542 filter2[i * filter2Size + k - 1] = 0;
543 (*filterPos)[i]++;
544 }
545
546 cutOff = 0;
547 /* count near zeros on the right */
548 for (j = filter2Size - 1; j > 0; j--) {
549 cutOff += FFABS(filter2[i * filter2Size + j]);
550
551 if (cutOff > SWS_MAX_REDUCE_CUTOFF * fone)
552 break;
553 min--;
554 }
555
556 if (min > minFilterSize)
557 minFilterSize = min;
558 }
559
560 #if HAVE_ALTIVEC
561 if (PPC_ALTIVEC(cpu_flags)) {
562 // we can handle the special case 4, so we don't want to go the full 8
563 if (minFilterSize < 5)
564 filterAlign = 4;
565
566 /* We really don't want to waste our time doing useless computation, so
567 * fall back on the scalar C code for very small filters.
568 * Vectorizing is worth it only if you have a decent-sized vector. */
569 if (minFilterSize < 3)
570 filterAlign = 1;
571 }
572 #endif
573
574 if (HAVE_MMX && cpu_flags & AV_CPU_FLAG_MMX) {
575 // special case for unscaled vertical filtering
576 if (minFilterSize == 1 && filterAlign == 2)
577 filterAlign = 1;
578 }
579
580 av_assert0(minFilterSize > 0);
581 filterSize = (minFilterSize + (filterAlign - 1)) & (~(filterAlign - 1));
582 av_assert0(filterSize > 0);
583 filter = av_malloc_array(dstW, filterSize * sizeof(*filter));
584 if (!filter)
585 goto fail;
586 if (filterSize >= MAX_FILTER_SIZE * 16 /
587 ((flags & SWS_ACCURATE_RND) ? APCK_SIZE : 16)) {
588 av_log(NULL, AV_LOG_ERROR, "sws: filterSize %d is too large, try less extreme scaling or set --sws-max-filter-size and recompile\n",
589 FF_CEIL_RSHIFT((filterSize+1) * ((flags & SWS_ACCURATE_RND) ? APCK_SIZE : 16), 4));
590 goto fail;
591 }
592 *outFilterSize = filterSize;
593
594 if (flags & SWS_PRINT_INFO)
595 av_log(NULL, AV_LOG_VERBOSE,
596 "SwScaler: reducing / aligning filtersize %d -> %d\n",
597 filter2Size, filterSize);
598 /* try to reduce the filter-size (step2 reduce it) */
599 for (i = 0; i < dstW; i++) {
600 int j;
601
602 for (j = 0; j < filterSize; j++) {
603 if (j >= filter2Size)
604 filter[i * filterSize + j] = 0;
605 else
606 filter[i * filterSize + j] = filter2[i * filter2Size + j];
607 if ((flags & SWS_BITEXACT) && j >= minFilterSize)
608 filter[i * filterSize + j] = 0;
609 }
610 }
611
612 // FIXME try to align filterPos if possible
613
614 // fix borders
615 for (i = 0; i < dstW; i++) {
616 int j;
617 if ((*filterPos)[i] < 0) {
618 // move filter coefficients left to compensate for filterPos
619 for (j = 1; j < filterSize; j++) {
620 int left = FFMAX(j + (*filterPos)[i], 0);
621 filter[i * filterSize + left] += filter[i * filterSize + j];
622 filter[i * filterSize + j] = 0;
623 }
624 (*filterPos)[i]= 0;
625 }
626
627 if ((*filterPos)[i] + filterSize > srcW) {
628 int shift = (*filterPos)[i] + filterSize - srcW;
629 // move filter coefficients right to compensate for filterPos
630 for (j = filterSize - 2; j >= 0; j--) {
631 int right = FFMIN(j + shift, filterSize - 1);
632 filter[i * filterSize + right] += filter[i * filterSize + j];
633 filter[i * filterSize + j] = 0;
634 }
635 (*filterPos)[i]= srcW - filterSize;
636 }
637 }
638
639 // Note the +1 is for the MMX scaler which reads over the end
640 /* align at 16 for AltiVec (needed by hScale_altivec_real) */
641 FF_ALLOCZ_ARRAY_OR_GOTO(NULL, *outFilter,
642 (dstW + 3), *outFilterSize * sizeof(int16_t), fail);
643
644 /* normalize & store in outFilter */
645 for (i = 0; i < dstW; i++) {
646 int j;
647 int64_t error = 0;
648 int64_t sum = 0;
649
650 for (j = 0; j < filterSize; j++) {
651 sum += filter[i * filterSize + j];
652 }
653 sum = (sum + one / 2) / one;
654 if (!sum) {
655 av_log(NULL, AV_LOG_WARNING, "SwScaler: zero vector in scaling\n");
656 sum = 1;
657 }
658 for (j = 0; j < *outFilterSize; j++) {
659 int64_t v = filter[i * filterSize + j] + error;
660 int intV = ROUNDED_DIV(v, sum);
661 (*outFilter)[i * (*outFilterSize) + j] = intV;
662 error = v - intV * sum;
663 }
664 }
665
666 (*filterPos)[dstW + 0] =
667 (*filterPos)[dstW + 1] =
668 (*filterPos)[dstW + 2] = (*filterPos)[dstW - 1]; /* the MMX/SSE scaler will
669 * read over the end */
670 for (i = 0; i < *outFilterSize; i++) {
671 int k = (dstW - 1) * (*outFilterSize) + i;
672 (*outFilter)[k + 1 * (*outFilterSize)] =
673 (*outFilter)[k + 2 * (*outFilterSize)] =
674 (*outFilter)[k + 3 * (*outFilterSize)] = (*outFilter)[k];
675 }
676
677 ret = 0;
678
679 fail:
680 if(ret < 0)
681 av_log(NULL, AV_LOG_ERROR, "sws: initFilter failed\n");
682 av_free(filter);
683 av_free(filter2);
684 return ret;
685 }
686
fill_rgb2yuv_table(SwsContext * c,const int table[4],int dstRange)687 static void fill_rgb2yuv_table(SwsContext *c, const int table[4], int dstRange)
688 {
689 int64_t W, V, Z, Cy, Cu, Cv;
690 int64_t vr = table[0];
691 int64_t ub = table[1];
692 int64_t ug = -table[2];
693 int64_t vg = -table[3];
694 int64_t ONE = 65536;
695 int64_t cy = ONE;
696 uint8_t *p = (uint8_t*)c->input_rgb2yuv_table;
697 int i;
698 static const int8_t map[] = {
699 BY_IDX, GY_IDX, -1 , BY_IDX, BY_IDX, GY_IDX, -1 , BY_IDX,
700 RY_IDX, -1 , GY_IDX, RY_IDX, RY_IDX, -1 , GY_IDX, RY_IDX,
701 RY_IDX, GY_IDX, -1 , RY_IDX, RY_IDX, GY_IDX, -1 , RY_IDX,
702 BY_IDX, -1 , GY_IDX, BY_IDX, BY_IDX, -1 , GY_IDX, BY_IDX,
703 BU_IDX, GU_IDX, -1 , BU_IDX, BU_IDX, GU_IDX, -1 , BU_IDX,
704 RU_IDX, -1 , GU_IDX, RU_IDX, RU_IDX, -1 , GU_IDX, RU_IDX,
705 RU_IDX, GU_IDX, -1 , RU_IDX, RU_IDX, GU_IDX, -1 , RU_IDX,
706 BU_IDX, -1 , GU_IDX, BU_IDX, BU_IDX, -1 , GU_IDX, BU_IDX,
707 BV_IDX, GV_IDX, -1 , BV_IDX, BV_IDX, GV_IDX, -1 , BV_IDX,
708 RV_IDX, -1 , GV_IDX, RV_IDX, RV_IDX, -1 , GV_IDX, RV_IDX,
709 RV_IDX, GV_IDX, -1 , RV_IDX, RV_IDX, GV_IDX, -1 , RV_IDX,
710 BV_IDX, -1 , GV_IDX, BV_IDX, BV_IDX, -1 , GV_IDX, BV_IDX,
711 RY_IDX, BY_IDX, RY_IDX, BY_IDX, RY_IDX, BY_IDX, RY_IDX, BY_IDX,
712 BY_IDX, RY_IDX, BY_IDX, RY_IDX, BY_IDX, RY_IDX, BY_IDX, RY_IDX,
713 GY_IDX, -1 , GY_IDX, -1 , GY_IDX, -1 , GY_IDX, -1 ,
714 -1 , GY_IDX, -1 , GY_IDX, -1 , GY_IDX, -1 , GY_IDX,
715 RU_IDX, BU_IDX, RU_IDX, BU_IDX, RU_IDX, BU_IDX, RU_IDX, BU_IDX,
716 BU_IDX, RU_IDX, BU_IDX, RU_IDX, BU_IDX, RU_IDX, BU_IDX, RU_IDX,
717 GU_IDX, -1 , GU_IDX, -1 , GU_IDX, -1 , GU_IDX, -1 ,
718 -1 , GU_IDX, -1 , GU_IDX, -1 , GU_IDX, -1 , GU_IDX,
719 RV_IDX, BV_IDX, RV_IDX, BV_IDX, RV_IDX, BV_IDX, RV_IDX, BV_IDX,
720 BV_IDX, RV_IDX, BV_IDX, RV_IDX, BV_IDX, RV_IDX, BV_IDX, RV_IDX,
721 GV_IDX, -1 , GV_IDX, -1 , GV_IDX, -1 , GV_IDX, -1 ,
722 -1 , GV_IDX, -1 , GV_IDX, -1 , GV_IDX, -1 , GV_IDX, //23
723 -1 , -1 , -1 , -1 , -1 , -1 , -1 , -1 , //24
724 -1 , -1 , -1 , -1 , -1 , -1 , -1 , -1 , //25
725 -1 , -1 , -1 , -1 , -1 , -1 , -1 , -1 , //26
726 -1 , -1 , -1 , -1 , -1 , -1 , -1 , -1 , //27
727 -1 , -1 , -1 , -1 , -1 , -1 , -1 , -1 , //28
728 -1 , -1 , -1 , -1 , -1 , -1 , -1 , -1 , //29
729 -1 , -1 , -1 , -1 , -1 , -1 , -1 , -1 , //30
730 -1 , -1 , -1 , -1 , -1 , -1 , -1 , -1 , //31
731 BY_IDX, GY_IDX, RY_IDX, -1 , -1 , -1 , -1 , -1 , //32
732 BU_IDX, GU_IDX, RU_IDX, -1 , -1 , -1 , -1 , -1 , //33
733 BV_IDX, GV_IDX, RV_IDX, -1 , -1 , -1 , -1 , -1 , //34
734 };
735
736 dstRange = 0; //FIXME range = 1 is handled elsewhere
737
738 if (!dstRange) {
739 cy = cy * 255 / 219;
740 } else {
741 vr = vr * 224 / 255;
742 ub = ub * 224 / 255;
743 ug = ug * 224 / 255;
744 vg = vg * 224 / 255;
745 }
746 W = ROUNDED_DIV(ONE*ONE*ug, ub);
747 V = ROUNDED_DIV(ONE*ONE*vg, vr);
748 Z = ONE*ONE-W-V;
749
750 Cy = ROUNDED_DIV(cy*Z, ONE);
751 Cu = ROUNDED_DIV(ub*Z, ONE);
752 Cv = ROUNDED_DIV(vr*Z, ONE);
753
754 c->input_rgb2yuv_table[RY_IDX] = -ROUNDED_DIV((1 << RGB2YUV_SHIFT)*V , Cy);
755 c->input_rgb2yuv_table[GY_IDX] = ROUNDED_DIV((1 << RGB2YUV_SHIFT)*ONE*ONE , Cy);
756 c->input_rgb2yuv_table[BY_IDX] = -ROUNDED_DIV((1 << RGB2YUV_SHIFT)*W , Cy);
757
758 c->input_rgb2yuv_table[RU_IDX] = ROUNDED_DIV((1 << RGB2YUV_SHIFT)*V , Cu);
759 c->input_rgb2yuv_table[GU_IDX] = -ROUNDED_DIV((1 << RGB2YUV_SHIFT)*ONE*ONE , Cu);
760 c->input_rgb2yuv_table[BU_IDX] = ROUNDED_DIV((1 << RGB2YUV_SHIFT)*(Z+W) , Cu);
761
762 c->input_rgb2yuv_table[RV_IDX] = ROUNDED_DIV((1 << RGB2YUV_SHIFT)*(V+Z) , Cv);
763 c->input_rgb2yuv_table[GV_IDX] = -ROUNDED_DIV((1 << RGB2YUV_SHIFT)*ONE*ONE , Cv);
764 c->input_rgb2yuv_table[BV_IDX] = ROUNDED_DIV((1 << RGB2YUV_SHIFT)*W , Cv);
765
766 if(/*!dstRange && */!memcmp(table, ff_yuv2rgb_coeffs[SWS_CS_DEFAULT], sizeof(ff_yuv2rgb_coeffs[SWS_CS_DEFAULT]))) {
767 c->input_rgb2yuv_table[BY_IDX] = ((int)(0.114 * 219 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));
768 c->input_rgb2yuv_table[BV_IDX] = (-(int)(0.081 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));
769 c->input_rgb2yuv_table[BU_IDX] = ((int)(0.500 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));
770 c->input_rgb2yuv_table[GY_IDX] = ((int)(0.587 * 219 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));
771 c->input_rgb2yuv_table[GV_IDX] = (-(int)(0.419 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));
772 c->input_rgb2yuv_table[GU_IDX] = (-(int)(0.331 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));
773 c->input_rgb2yuv_table[RY_IDX] = ((int)(0.299 * 219 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));
774 c->input_rgb2yuv_table[RV_IDX] = ((int)(0.500 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));
775 c->input_rgb2yuv_table[RU_IDX] = (-(int)(0.169 * 224 / 255 * (1 << RGB2YUV_SHIFT) + 0.5));
776 }
777 for(i=0; i<FF_ARRAY_ELEMS(map); i++)
778 AV_WL16(p + 16*4 + 2*i, map[i] >= 0 ? c->input_rgb2yuv_table[map[i]] : 0);
779 }
780
fill_xyztables(struct SwsContext * c)781 static void fill_xyztables(struct SwsContext *c)
782 {
783 int i;
784 double xyzgamma = XYZ_GAMMA;
785 double rgbgamma = 1.0 / RGB_GAMMA;
786 double xyzgammainv = 1.0 / XYZ_GAMMA;
787 double rgbgammainv = RGB_GAMMA;
788 static const int16_t xyz2rgb_matrix[3][4] = {
789 {13270, -6295, -2041},
790 {-3969, 7682, 170},
791 { 228, -835, 4329} };
792 static const int16_t rgb2xyz_matrix[3][4] = {
793 {1689, 1464, 739},
794 { 871, 2929, 296},
795 { 79, 488, 3891} };
796 static int16_t xyzgamma_tab[4096], rgbgamma_tab[4096], xyzgammainv_tab[4096], rgbgammainv_tab[4096];
797
798 memcpy(c->xyz2rgb_matrix, xyz2rgb_matrix, sizeof(c->xyz2rgb_matrix));
799 memcpy(c->rgb2xyz_matrix, rgb2xyz_matrix, sizeof(c->rgb2xyz_matrix));
800 c->xyzgamma = xyzgamma_tab;
801 c->rgbgamma = rgbgamma_tab;
802 c->xyzgammainv = xyzgammainv_tab;
803 c->rgbgammainv = rgbgammainv_tab;
804
805 if (rgbgamma_tab[4095])
806 return;
807
808 /* set gamma vectors */
809 for (i = 0; i < 4096; i++) {
810 xyzgamma_tab[i] = lrint(pow(i / 4095.0, xyzgamma) * 4095.0);
811 rgbgamma_tab[i] = lrint(pow(i / 4095.0, rgbgamma) * 4095.0);
812 xyzgammainv_tab[i] = lrint(pow(i / 4095.0, xyzgammainv) * 4095.0);
813 rgbgammainv_tab[i] = lrint(pow(i / 4095.0, rgbgammainv) * 4095.0);
814 }
815 }
816
sws_setColorspaceDetails(struct SwsContext * c,const int inv_table[4],int srcRange,const int table[4],int dstRange,int brightness,int contrast,int saturation)817 int sws_setColorspaceDetails(struct SwsContext *c, const int inv_table[4],
818 int srcRange, const int table[4], int dstRange,
819 int brightness, int contrast, int saturation)
820 {
821 const AVPixFmtDescriptor *desc_dst;
822 const AVPixFmtDescriptor *desc_src;
823 int need_reinit = 0;
824 memmove(c->srcColorspaceTable, inv_table, sizeof(int) * 4);
825 memmove(c->dstColorspaceTable, table, sizeof(int) * 4);
826
827 handle_formats(c);
828 desc_dst = av_pix_fmt_desc_get(c->dstFormat);
829 desc_src = av_pix_fmt_desc_get(c->srcFormat);
830
831 if(!isYUV(c->dstFormat) && !isGray(c->dstFormat))
832 dstRange = 0;
833 if(!isYUV(c->srcFormat) && !isGray(c->srcFormat))
834 srcRange = 0;
835
836 c->brightness = brightness;
837 c->contrast = contrast;
838 c->saturation = saturation;
839 if (c->srcRange != srcRange || c->dstRange != dstRange)
840 need_reinit = 1;
841 c->srcRange = srcRange;
842 c->dstRange = dstRange;
843
844 //The srcBpc check is possibly wrong but we seem to lack a definitive reference to test this
845 //and what we have in ticket 2939 looks better with this check
846 if (need_reinit && (c->srcBpc == 8 || !isYUV(c->srcFormat)))
847 ff_sws_init_range_convert(c);
848
849 if ((isYUV(c->dstFormat) || isGray(c->dstFormat)) && (isYUV(c->srcFormat) || isGray(c->srcFormat)))
850 return -1;
851
852 c->dstFormatBpp = av_get_bits_per_pixel(desc_dst);
853 c->srcFormatBpp = av_get_bits_per_pixel(desc_src);
854
855 if (!isYUV(c->dstFormat) && !isGray(c->dstFormat)) {
856 ff_yuv2rgb_c_init_tables(c, inv_table, srcRange, brightness,
857 contrast, saturation);
858 // FIXME factorize
859
860 #if (ARCH_PPC == 1)
861 if (ARCH_PPC)
862 ff_yuv2rgb_init_tables_ppc(c, inv_table, brightness,
863 contrast, saturation);
864 #endif
865 }
866
867 fill_rgb2yuv_table(c, table, dstRange);
868
869 return 0;
870 }
871
sws_getColorspaceDetails(struct SwsContext * c,int ** inv_table,int * srcRange,int ** table,int * dstRange,int * brightness,int * contrast,int * saturation)872 int sws_getColorspaceDetails(struct SwsContext *c, int **inv_table,
873 int *srcRange, int **table, int *dstRange,
874 int *brightness, int *contrast, int *saturation)
875 {
876 if (!c )
877 return -1;
878
879 *inv_table = c->srcColorspaceTable;
880 *table = c->dstColorspaceTable;
881 *srcRange = c->srcRange;
882 *dstRange = c->dstRange;
883 *brightness = c->brightness;
884 *contrast = c->contrast;
885 *saturation = c->saturation;
886
887 return 0;
888 }
889
handle_jpeg(enum AVPixelFormat * format)890 static int handle_jpeg(enum AVPixelFormat *format)
891 {
892 switch (*format) {
893 case AV_PIX_FMT_YUVJ420P:
894 *format = AV_PIX_FMT_YUV420P;
895 return 1;
896 case AV_PIX_FMT_YUVJ411P:
897 *format = AV_PIX_FMT_YUV411P;
898 return 1;
899 case AV_PIX_FMT_YUVJ422P:
900 *format = AV_PIX_FMT_YUV422P;
901 return 1;
902 case AV_PIX_FMT_YUVJ444P:
903 *format = AV_PIX_FMT_YUV444P;
904 return 1;
905 case AV_PIX_FMT_YUVJ440P:
906 *format = AV_PIX_FMT_YUV440P;
907 return 1;
908 case AV_PIX_FMT_GRAY8:
909 case AV_PIX_FMT_GRAY16LE:
910 case AV_PIX_FMT_GRAY16BE:
911 return 1;
912 default:
913 return 0;
914 }
915 }
916
handle_0alpha(enum AVPixelFormat * format)917 static int handle_0alpha(enum AVPixelFormat *format)
918 {
919 switch (*format) {
920 case AV_PIX_FMT_0BGR : *format = AV_PIX_FMT_ABGR ; return 1;
921 case AV_PIX_FMT_BGR0 : *format = AV_PIX_FMT_BGRA ; return 4;
922 case AV_PIX_FMT_0RGB : *format = AV_PIX_FMT_ARGB ; return 1;
923 case AV_PIX_FMT_RGB0 : *format = AV_PIX_FMT_RGBA ; return 4;
924 default: return 0;
925 }
926 }
927
handle_xyz(enum AVPixelFormat * format)928 static int handle_xyz(enum AVPixelFormat *format)
929 {
930 switch (*format) {
931 case AV_PIX_FMT_XYZ12BE : *format = AV_PIX_FMT_RGB48BE; return 1;
932 case AV_PIX_FMT_XYZ12LE : *format = AV_PIX_FMT_RGB48LE; return 1;
933 default: return 0;
934 }
935 }
936
handle_formats(SwsContext * c)937 static void handle_formats(SwsContext *c)
938 {
939 c->src0Alpha |= handle_0alpha(&c->srcFormat);
940 c->dst0Alpha |= handle_0alpha(&c->dstFormat);
941 c->srcXYZ |= handle_xyz(&c->srcFormat);
942 c->dstXYZ |= handle_xyz(&c->dstFormat);
943 if (c->srcXYZ || c->dstXYZ)
944 fill_xyztables(c);
945 }
946
sws_alloc_context(void)947 SwsContext *sws_alloc_context(void)
948 {
949 SwsContext *c = av_mallocz(sizeof(SwsContext));
950
951 av_assert0(offsetof(SwsContext, redDither) + DITHER32_INT == offsetof(SwsContext, dither32));
952
953 if (c) {
954 c->av_class = &sws_context_class;
955 av_opt_set_defaults(c);
956 }
957
958 return c;
959 }
960
sws_init_context(SwsContext * c,SwsFilter * srcFilter,SwsFilter * dstFilter)961 av_cold int sws_init_context(SwsContext *c, SwsFilter *srcFilter,
962 SwsFilter *dstFilter)
963 {
964 int i, j;
965 int usesVFilter, usesHFilter;
966 int unscaled;
967 SwsFilter dummyFilter = { NULL, NULL, NULL, NULL };
968 int srcW = c->srcW;
969 int srcH = c->srcH;
970 int dstW = c->dstW;
971 int dstH = c->dstH;
972 int dst_stride = FFALIGN(dstW * sizeof(int16_t) + 66, 16);
973 int flags, cpu_flags;
974 enum AVPixelFormat srcFormat = c->srcFormat;
975 enum AVPixelFormat dstFormat = c->dstFormat;
976 const AVPixFmtDescriptor *desc_src;
977 const AVPixFmtDescriptor *desc_dst;
978
979 cpu_flags = av_get_cpu_flags();
980 flags = c->flags;
981 emms_c();
982 if (!rgb15to16)
983 sws_rgb2rgb_init();
984
985 unscaled = (srcW == dstW && srcH == dstH);
986
987 c->srcRange |= handle_jpeg(&c->srcFormat);
988 c->dstRange |= handle_jpeg(&c->dstFormat);
989
990 if(srcFormat!=c->srcFormat || dstFormat!=c->dstFormat)
991 av_log(c, AV_LOG_WARNING, "deprecated pixel format used, make sure you did set range correctly\n");
992
993 if (!c->contrast && !c->saturation && !c->dstFormatBpp)
994 sws_setColorspaceDetails(c, ff_yuv2rgb_coeffs[SWS_CS_DEFAULT], c->srcRange,
995 ff_yuv2rgb_coeffs[SWS_CS_DEFAULT],
996 c->dstRange, 0, 1 << 16, 1 << 16);
997
998 handle_formats(c);
999 srcFormat = c->srcFormat;
1000 dstFormat = c->dstFormat;
1001 desc_src = av_pix_fmt_desc_get(srcFormat);
1002 desc_dst = av_pix_fmt_desc_get(dstFormat);
1003
1004 if (!(unscaled && sws_isSupportedEndiannessConversion(srcFormat) &&
1005 av_pix_fmt_swap_endianness(srcFormat) == dstFormat)) {
1006 if (!sws_isSupportedInput(srcFormat)) {
1007 av_log(c, AV_LOG_ERROR, "%s is not supported as input pixel format\n",
1008 av_get_pix_fmt_name(srcFormat));
1009 return AVERROR(EINVAL);
1010 }
1011 if (!sws_isSupportedOutput(dstFormat)) {
1012 av_log(c, AV_LOG_ERROR, "%s is not supported as output pixel format\n",
1013 av_get_pix_fmt_name(dstFormat));
1014 return AVERROR(EINVAL);
1015 }
1016 }
1017
1018 i = flags & (SWS_POINT |
1019 SWS_AREA |
1020 SWS_BILINEAR |
1021 SWS_FAST_BILINEAR |
1022 SWS_BICUBIC |
1023 SWS_X |
1024 SWS_GAUSS |
1025 SWS_LANCZOS |
1026 SWS_SINC |
1027 SWS_SPLINE |
1028 SWS_BICUBLIN);
1029
1030 /* provide a default scaler if not set by caller */
1031 if (!i) {
1032 if (dstW < srcW && dstH < srcH)
1033 flags |= SWS_BICUBIC;
1034 else if (dstW > srcW && dstH > srcH)
1035 flags |= SWS_BICUBIC;
1036 else
1037 flags |= SWS_BICUBIC;
1038 c->flags = flags;
1039 } else if (i & (i - 1)) {
1040 av_log(c, AV_LOG_ERROR,
1041 "Exactly one scaler algorithm must be chosen, got %X\n", i);
1042 return AVERROR(EINVAL);
1043 }
1044 /* sanity check */
1045 if (srcW < 1 || srcH < 1 || dstW < 1 || dstH < 1) {
1046 /* FIXME check if these are enough and try to lower them after
1047 * fixing the relevant parts of the code */
1048 av_log(c, AV_LOG_ERROR, "%dx%d -> %dx%d is invalid scaling dimension\n",
1049 srcW, srcH, dstW, dstH);
1050 return AVERROR(EINVAL);
1051 }
1052
1053 if (!dstFilter)
1054 dstFilter = &dummyFilter;
1055 if (!srcFilter)
1056 srcFilter = &dummyFilter;
1057
1058 c->lumXInc = (((int64_t)srcW << 16) + (dstW >> 1)) / dstW;
1059 c->lumYInc = (((int64_t)srcH << 16) + (dstH >> 1)) / dstH;
1060 c->dstFormatBpp = av_get_bits_per_pixel(desc_dst);
1061 c->srcFormatBpp = av_get_bits_per_pixel(desc_src);
1062 c->vRounder = 4 * ULLN(0x0001000100010001);
1063
1064 usesVFilter = (srcFilter->lumV && srcFilter->lumV->length > 1) ||
1065 (srcFilter->chrV && srcFilter->chrV->length > 1) ||
1066 (dstFilter->lumV && dstFilter->lumV->length > 1) ||
1067 (dstFilter->chrV && dstFilter->chrV->length > 1);
1068 usesHFilter = (srcFilter->lumH && srcFilter->lumH->length > 1) ||
1069 (srcFilter->chrH && srcFilter->chrH->length > 1) ||
1070 (dstFilter->lumH && dstFilter->lumH->length > 1) ||
1071 (dstFilter->chrH && dstFilter->chrH->length > 1);
1072
1073 av_pix_fmt_get_chroma_sub_sample(srcFormat, &c->chrSrcHSubSample, &c->chrSrcVSubSample);
1074 av_pix_fmt_get_chroma_sub_sample(dstFormat, &c->chrDstHSubSample, &c->chrDstVSubSample);
1075
1076 if (isAnyRGB(dstFormat) && !(flags&SWS_FULL_CHR_H_INT)) {
1077 if (dstW&1) {
1078 av_log(c, AV_LOG_DEBUG, "Forcing full internal H chroma due to odd output size\n");
1079 flags |= SWS_FULL_CHR_H_INT;
1080 c->flags = flags;
1081 }
1082
1083 if ( c->chrSrcHSubSample == 0
1084 && c->chrSrcVSubSample == 0
1085 && c->dither != SWS_DITHER_BAYER //SWS_FULL_CHR_H_INT is currently not supported with SWS_DITHER_BAYER
1086 && !(c->flags & SWS_FAST_BILINEAR)
1087 ) {
1088 av_log(c, AV_LOG_DEBUG, "Forcing full internal H chroma due to input having non subsampled chroma\n");
1089 flags |= SWS_FULL_CHR_H_INT;
1090 c->flags = flags;
1091 }
1092 }
1093
1094 if (c->dither == SWS_DITHER_AUTO) {
1095 if (flags & SWS_ERROR_DIFFUSION)
1096 c->dither = SWS_DITHER_ED;
1097 }
1098
1099 if(dstFormat == AV_PIX_FMT_BGR4_BYTE ||
1100 dstFormat == AV_PIX_FMT_RGB4_BYTE ||
1101 dstFormat == AV_PIX_FMT_BGR8 ||
1102 dstFormat == AV_PIX_FMT_RGB8) {
1103 if (c->dither == SWS_DITHER_AUTO)
1104 c->dither = (flags & SWS_FULL_CHR_H_INT) ? SWS_DITHER_ED : SWS_DITHER_BAYER;
1105 if (!(flags & SWS_FULL_CHR_H_INT)) {
1106 if (c->dither == SWS_DITHER_ED || c->dither == SWS_DITHER_A_DITHER || c->dither == SWS_DITHER_X_DITHER) {
1107 av_log(c, AV_LOG_DEBUG,
1108 "Desired dithering only supported in full chroma interpolation for destination format '%s'\n",
1109 av_get_pix_fmt_name(dstFormat));
1110 flags |= SWS_FULL_CHR_H_INT;
1111 c->flags = flags;
1112 }
1113 }
1114 if (flags & SWS_FULL_CHR_H_INT) {
1115 if (c->dither == SWS_DITHER_BAYER) {
1116 av_log(c, AV_LOG_DEBUG,
1117 "Ordered dither is not supported in full chroma interpolation for destination format '%s'\n",
1118 av_get_pix_fmt_name(dstFormat));
1119 c->dither = SWS_DITHER_ED;
1120 }
1121 }
1122 }
1123 if (isPlanarRGB(dstFormat)) {
1124 if (!(flags & SWS_FULL_CHR_H_INT)) {
1125 av_log(c, AV_LOG_DEBUG,
1126 "%s output is not supported with half chroma resolution, switching to full\n",
1127 av_get_pix_fmt_name(dstFormat));
1128 flags |= SWS_FULL_CHR_H_INT;
1129 c->flags = flags;
1130 }
1131 }
1132
1133 /* reuse chroma for 2 pixels RGB/BGR unless user wants full
1134 * chroma interpolation */
1135 if (flags & SWS_FULL_CHR_H_INT &&
1136 isAnyRGB(dstFormat) &&
1137 !isPlanarRGB(dstFormat) &&
1138 dstFormat != AV_PIX_FMT_RGBA &&
1139 dstFormat != AV_PIX_FMT_ARGB &&
1140 dstFormat != AV_PIX_FMT_BGRA &&
1141 dstFormat != AV_PIX_FMT_ABGR &&
1142 dstFormat != AV_PIX_FMT_RGB24 &&
1143 dstFormat != AV_PIX_FMT_BGR24 &&
1144 dstFormat != AV_PIX_FMT_BGR4_BYTE &&
1145 dstFormat != AV_PIX_FMT_RGB4_BYTE &&
1146 dstFormat != AV_PIX_FMT_BGR8 &&
1147 dstFormat != AV_PIX_FMT_RGB8
1148 ) {
1149 av_log(c, AV_LOG_WARNING,
1150 "full chroma interpolation for destination format '%s' not yet implemented\n",
1151 av_get_pix_fmt_name(dstFormat));
1152 flags &= ~SWS_FULL_CHR_H_INT;
1153 c->flags = flags;
1154 }
1155 if (isAnyRGB(dstFormat) && !(flags & SWS_FULL_CHR_H_INT))
1156 c->chrDstHSubSample = 1;
1157
1158 // drop some chroma lines if the user wants it
1159 c->vChrDrop = (flags & SWS_SRC_V_CHR_DROP_MASK) >>
1160 SWS_SRC_V_CHR_DROP_SHIFT;
1161 c->chrSrcVSubSample += c->vChrDrop;
1162
1163 /* drop every other pixel for chroma calculation unless user
1164 * wants full chroma */
1165 if (isAnyRGB(srcFormat) && !(flags & SWS_FULL_CHR_H_INP) &&
1166 srcFormat != AV_PIX_FMT_RGB8 && srcFormat != AV_PIX_FMT_BGR8 &&
1167 srcFormat != AV_PIX_FMT_RGB4 && srcFormat != AV_PIX_FMT_BGR4 &&
1168 srcFormat != AV_PIX_FMT_RGB4_BYTE && srcFormat != AV_PIX_FMT_BGR4_BYTE &&
1169 srcFormat != AV_PIX_FMT_GBRP9BE && srcFormat != AV_PIX_FMT_GBRP9LE &&
1170 srcFormat != AV_PIX_FMT_GBRP10BE && srcFormat != AV_PIX_FMT_GBRP10LE &&
1171 srcFormat != AV_PIX_FMT_GBRP12BE && srcFormat != AV_PIX_FMT_GBRP12LE &&
1172 srcFormat != AV_PIX_FMT_GBRP14BE && srcFormat != AV_PIX_FMT_GBRP14LE &&
1173 srcFormat != AV_PIX_FMT_GBRP16BE && srcFormat != AV_PIX_FMT_GBRP16LE &&
1174 ((dstW >> c->chrDstHSubSample) <= (srcW >> 1) ||
1175 (flags & SWS_FAST_BILINEAR)))
1176 c->chrSrcHSubSample = 1;
1177
1178 // Note the FF_CEIL_RSHIFT is so that we always round toward +inf.
1179 c->chrSrcW = FF_CEIL_RSHIFT(srcW, c->chrSrcHSubSample);
1180 c->chrSrcH = FF_CEIL_RSHIFT(srcH, c->chrSrcVSubSample);
1181 c->chrDstW = FF_CEIL_RSHIFT(dstW, c->chrDstHSubSample);
1182 c->chrDstH = FF_CEIL_RSHIFT(dstH, c->chrDstVSubSample);
1183
1184 FF_ALLOC_OR_GOTO(c, c->formatConvBuffer, FFALIGN(srcW*2+78, 16) * 2, fail);
1185
1186 c->srcBpc = 1 + desc_src->comp[0].depth_minus1;
1187 if (c->srcBpc < 8)
1188 c->srcBpc = 8;
1189 c->dstBpc = 1 + desc_dst->comp[0].depth_minus1;
1190 if (c->dstBpc < 8)
1191 c->dstBpc = 8;
1192 if (isAnyRGB(srcFormat) || srcFormat == AV_PIX_FMT_PAL8)
1193 c->srcBpc = 16;
1194 if (c->dstBpc == 16)
1195 dst_stride <<= 1;
1196
1197 if (INLINE_MMXEXT(cpu_flags) && c->srcBpc == 8 && c->dstBpc <= 14) {
1198 c->canMMXEXTBeUsed = dstW >= srcW && (dstW & 31) == 0 &&
1199 c->chrDstW >= c->chrSrcW &&
1200 (srcW & 15) == 0;
1201 if (!c->canMMXEXTBeUsed && dstW >= srcW && c->chrDstW >= c->chrSrcW && (srcW & 15) == 0
1202
1203 && (flags & SWS_FAST_BILINEAR)) {
1204 if (flags & SWS_PRINT_INFO)
1205 av_log(c, AV_LOG_INFO,
1206 "output width is not a multiple of 32 -> no MMXEXT scaler\n");
1207 }
1208 if (usesHFilter || isNBPS(c->srcFormat) || is16BPS(c->srcFormat) || isAnyRGB(c->srcFormat))
1209 c->canMMXEXTBeUsed = 0;
1210 } else
1211 c->canMMXEXTBeUsed = 0;
1212
1213 c->chrXInc = (((int64_t)c->chrSrcW << 16) + (c->chrDstW >> 1)) / c->chrDstW;
1214 c->chrYInc = (((int64_t)c->chrSrcH << 16) + (c->chrDstH >> 1)) / c->chrDstH;
1215
1216 /* Match pixel 0 of the src to pixel 0 of dst and match pixel n-2 of src
1217 * to pixel n-2 of dst, but only for the FAST_BILINEAR mode otherwise do
1218 * correct scaling.
1219 * n-2 is the last chrominance sample available.
1220 * This is not perfect, but no one should notice the difference, the more
1221 * correct variant would be like the vertical one, but that would require
1222 * some special code for the first and last pixel */
1223 if (flags & SWS_FAST_BILINEAR) {
1224 if (c->canMMXEXTBeUsed) {
1225 c->lumXInc += 20;
1226 c->chrXInc += 20;
1227 }
1228 // we don't use the x86 asm scaler if MMX is available
1229 else if (INLINE_MMX(cpu_flags) && c->dstBpc <= 14) {
1230 c->lumXInc = ((int64_t)(srcW - 2) << 16) / (dstW - 2) - 20;
1231 c->chrXInc = ((int64_t)(c->chrSrcW - 2) << 16) / (c->chrDstW - 2) - 20;
1232 }
1233 }
1234
1235 #define USE_MMAP (HAVE_MMAP && HAVE_MPROTECT && defined MAP_ANONYMOUS)
1236
1237 /* precalculate horizontal scaler filter coefficients */
1238 {
1239 #if HAVE_MMXEXT_INLINE
1240 // can't downscale !!!
1241 if (c->canMMXEXTBeUsed && (flags & SWS_FAST_BILINEAR)) {
1242 c->lumMmxextFilterCodeSize = ff_init_hscaler_mmxext(dstW, c->lumXInc, NULL,
1243 NULL, NULL, 8);
1244 c->chrMmxextFilterCodeSize = ff_init_hscaler_mmxext(c->chrDstW, c->chrXInc,
1245 NULL, NULL, NULL, 4);
1246
1247 #if USE_MMAP
1248 c->lumMmxextFilterCode = mmap(NULL, c->lumMmxextFilterCodeSize,
1249 PROT_READ | PROT_WRITE,
1250 MAP_PRIVATE | MAP_ANONYMOUS,
1251 -1, 0);
1252 c->chrMmxextFilterCode = mmap(NULL, c->chrMmxextFilterCodeSize,
1253 PROT_READ | PROT_WRITE,
1254 MAP_PRIVATE | MAP_ANONYMOUS,
1255 -1, 0);
1256 #elif HAVE_VIRTUALALLOC
1257 c->lumMmxextFilterCode = VirtualAlloc(NULL,
1258 c->lumMmxextFilterCodeSize,
1259 MEM_COMMIT,
1260 PAGE_EXECUTE_READWRITE);
1261 c->chrMmxextFilterCode = VirtualAlloc(NULL,
1262 c->chrMmxextFilterCodeSize,
1263 MEM_COMMIT,
1264 PAGE_EXECUTE_READWRITE);
1265 #else
1266 c->lumMmxextFilterCode = av_malloc(c->lumMmxextFilterCodeSize);
1267 c->chrMmxextFilterCode = av_malloc(c->chrMmxextFilterCodeSize);
1268 #endif
1269
1270 #ifdef MAP_ANONYMOUS
1271 if (c->lumMmxextFilterCode == MAP_FAILED || c->chrMmxextFilterCode == MAP_FAILED)
1272 #else
1273 if (!c->lumMmxextFilterCode || !c->chrMmxextFilterCode)
1274 #endif
1275 {
1276 av_log(c, AV_LOG_ERROR, "Failed to allocate MMX2FilterCode\n");
1277 return AVERROR(ENOMEM);
1278 }
1279
1280 FF_ALLOCZ_OR_GOTO(c, c->hLumFilter, (dstW / 8 + 8) * sizeof(int16_t), fail);
1281 FF_ALLOCZ_OR_GOTO(c, c->hChrFilter, (c->chrDstW / 4 + 8) * sizeof(int16_t), fail);
1282 FF_ALLOCZ_OR_GOTO(c, c->hLumFilterPos, (dstW / 2 / 8 + 8) * sizeof(int32_t), fail);
1283 FF_ALLOCZ_OR_GOTO(c, c->hChrFilterPos, (c->chrDstW / 2 / 4 + 8) * sizeof(int32_t), fail);
1284
1285 ff_init_hscaler_mmxext( dstW, c->lumXInc, c->lumMmxextFilterCode,
1286 c->hLumFilter, (uint32_t*)c->hLumFilterPos, 8);
1287 ff_init_hscaler_mmxext(c->chrDstW, c->chrXInc, c->chrMmxextFilterCode,
1288 c->hChrFilter, (uint32_t*)c->hChrFilterPos, 4);
1289
1290 #if USE_MMAP
1291 if ( mprotect(c->lumMmxextFilterCode, c->lumMmxextFilterCodeSize, PROT_EXEC | PROT_READ) == -1
1292 || mprotect(c->chrMmxextFilterCode, c->chrMmxextFilterCodeSize, PROT_EXEC | PROT_READ) == -1) {
1293 av_log(c, AV_LOG_ERROR, "mprotect failed, cannot use fast bilinear scaler\n");
1294 goto fail;
1295 }
1296 #endif
1297 } else
1298 #endif /* HAVE_MMXEXT_INLINE */
1299 {
1300 const int filterAlign = X86_MMX(cpu_flags) ? 4 :
1301 PPC_ALTIVEC(cpu_flags) ? 8 : 1;
1302 if (initFilter(&c->hLumFilter, &c->hLumFilterPos,
1303 &c->hLumFilterSize, c->lumXInc,
1304 srcW, dstW, filterAlign, 1 << 14,
1305 (flags & SWS_BICUBLIN) ? (flags | SWS_BICUBIC) : flags,
1306 cpu_flags, srcFilter->lumH, dstFilter->lumH,
1307 c->param,
1308 get_local_pos(c, 0, 0, 0),
1309 get_local_pos(c, 0, 0, 0)) < 0)
1310 goto fail;
1311 if (initFilter(&c->hChrFilter, &c->hChrFilterPos,
1312 &c->hChrFilterSize, c->chrXInc,
1313 c->chrSrcW, c->chrDstW, filterAlign, 1 << 14,
1314 (flags & SWS_BICUBLIN) ? (flags | SWS_BILINEAR) : flags,
1315 cpu_flags, srcFilter->chrH, dstFilter->chrH,
1316 c->param,
1317 get_local_pos(c, c->chrSrcHSubSample, c->src_h_chr_pos, 0),
1318 get_local_pos(c, c->chrDstHSubSample, c->dst_h_chr_pos, 0)) < 0)
1319 goto fail;
1320 }
1321 } // initialize horizontal stuff
1322
1323 /* precalculate vertical scaler filter coefficients */
1324 {
1325 const int filterAlign = X86_MMX(cpu_flags) ? 2 :
1326 PPC_ALTIVEC(cpu_flags) ? 8 : 1;
1327
1328 if (initFilter(&c->vLumFilter, &c->vLumFilterPos, &c->vLumFilterSize,
1329 c->lumYInc, srcH, dstH, filterAlign, (1 << 12),
1330 (flags & SWS_BICUBLIN) ? (flags | SWS_BICUBIC) : flags,
1331 cpu_flags, srcFilter->lumV, dstFilter->lumV,
1332 c->param,
1333 get_local_pos(c, 0, 0, 1),
1334 get_local_pos(c, 0, 0, 1)) < 0)
1335 goto fail;
1336 if (initFilter(&c->vChrFilter, &c->vChrFilterPos, &c->vChrFilterSize,
1337 c->chrYInc, c->chrSrcH, c->chrDstH,
1338 filterAlign, (1 << 12),
1339 (flags & SWS_BICUBLIN) ? (flags | SWS_BILINEAR) : flags,
1340 cpu_flags, srcFilter->chrV, dstFilter->chrV,
1341 c->param,
1342 get_local_pos(c, c->chrSrcVSubSample, c->src_v_chr_pos, 1),
1343 get_local_pos(c, c->chrDstVSubSample, c->dst_v_chr_pos, 1)) < 0)
1344
1345 goto fail;
1346
1347 #if HAVE_ALTIVEC
1348 FF_ALLOC_OR_GOTO(c, c->vYCoeffsBank, sizeof(vector signed short) * c->vLumFilterSize * c->dstH, fail);
1349 FF_ALLOC_OR_GOTO(c, c->vCCoeffsBank, sizeof(vector signed short) * c->vChrFilterSize * c->chrDstH, fail);
1350
1351 for (i = 0; i < c->vLumFilterSize * c->dstH; i++) {
1352 int j;
1353 short *p = (short *)&c->vYCoeffsBank[i];
1354 for (j = 0; j < 8; j++)
1355 p[j] = c->vLumFilter[i];
1356 }
1357
1358 for (i = 0; i < c->vChrFilterSize * c->chrDstH; i++) {
1359 int j;
1360 short *p = (short *)&c->vCCoeffsBank[i];
1361 for (j = 0; j < 8; j++)
1362 p[j] = c->vChrFilter[i];
1363 }
1364 #endif
1365 }
1366
1367 // calculate buffer sizes so that they won't run out while handling these damn slices
1368 c->vLumBufSize = c->vLumFilterSize;
1369 c->vChrBufSize = c->vChrFilterSize;
1370 for (i = 0; i < dstH; i++) {
1371 int chrI = (int64_t)i * c->chrDstH / dstH;
1372 int nextSlice = FFMAX(c->vLumFilterPos[i] + c->vLumFilterSize - 1,
1373 ((c->vChrFilterPos[chrI] + c->vChrFilterSize - 1)
1374 << c->chrSrcVSubSample));
1375
1376 nextSlice >>= c->chrSrcVSubSample;
1377 nextSlice <<= c->chrSrcVSubSample;
1378 if (c->vLumFilterPos[i] + c->vLumBufSize < nextSlice)
1379 c->vLumBufSize = nextSlice - c->vLumFilterPos[i];
1380 if (c->vChrFilterPos[chrI] + c->vChrBufSize <
1381 (nextSlice >> c->chrSrcVSubSample))
1382 c->vChrBufSize = (nextSlice >> c->chrSrcVSubSample) -
1383 c->vChrFilterPos[chrI];
1384 }
1385
1386 for (i = 0; i < 4; i++)
1387 FF_ALLOCZ_OR_GOTO(c, c->dither_error[i], (c->dstW+2) * sizeof(int), fail);
1388
1389 /* Allocate pixbufs (we use dynamic allocation because otherwise we would
1390 * need to allocate several megabytes to handle all possible cases) */
1391 FF_ALLOC_OR_GOTO(c, c->lumPixBuf, c->vLumBufSize * 3 * sizeof(int16_t *), fail);
1392 FF_ALLOC_OR_GOTO(c, c->chrUPixBuf, c->vChrBufSize * 3 * sizeof(int16_t *), fail);
1393 FF_ALLOC_OR_GOTO(c, c->chrVPixBuf, c->vChrBufSize * 3 * sizeof(int16_t *), fail);
1394 if (CONFIG_SWSCALE_ALPHA && isALPHA(c->srcFormat) && isALPHA(c->dstFormat))
1395 FF_ALLOCZ_OR_GOTO(c, c->alpPixBuf, c->vLumBufSize * 3 * sizeof(int16_t *), fail);
1396 /* Note we need at least one pixel more at the end because of the MMX code
1397 * (just in case someone wants to replace the 4000/8000). */
1398 /* align at 16 bytes for AltiVec */
1399 for (i = 0; i < c->vLumBufSize; i++) {
1400 FF_ALLOCZ_OR_GOTO(c, c->lumPixBuf[i + c->vLumBufSize],
1401 dst_stride + 16, fail);
1402 c->lumPixBuf[i] = c->lumPixBuf[i + c->vLumBufSize];
1403 }
1404 // 64 / c->scalingBpp is the same as 16 / sizeof(scaling_intermediate)
1405 c->uv_off = (dst_stride>>1) + 64 / (c->dstBpc &~ 7);
1406 c->uv_offx2 = dst_stride + 16;
1407 for (i = 0; i < c->vChrBufSize; i++) {
1408 FF_ALLOC_OR_GOTO(c, c->chrUPixBuf[i + c->vChrBufSize],
1409 dst_stride * 2 + 32, fail);
1410 c->chrUPixBuf[i] = c->chrUPixBuf[i + c->vChrBufSize];
1411 c->chrVPixBuf[i] = c->chrVPixBuf[i + c->vChrBufSize]
1412 = c->chrUPixBuf[i] + (dst_stride >> 1) + 8;
1413 }
1414 if (CONFIG_SWSCALE_ALPHA && c->alpPixBuf)
1415 for (i = 0; i < c->vLumBufSize; i++) {
1416 FF_ALLOCZ_OR_GOTO(c, c->alpPixBuf[i + c->vLumBufSize],
1417 dst_stride + 16, fail);
1418 c->alpPixBuf[i] = c->alpPixBuf[i + c->vLumBufSize];
1419 }
1420
1421 // try to avoid drawing green stuff between the right end and the stride end
1422 for (i = 0; i < c->vChrBufSize; i++)
1423 if(desc_dst->comp[0].depth_minus1 == 15){
1424 av_assert0(c->dstBpc > 14);
1425 for(j=0; j<dst_stride/2+1; j++)
1426 ((int32_t*)(c->chrUPixBuf[i]))[j] = 1<<18;
1427 } else
1428 for(j=0; j<dst_stride+1; j++)
1429 ((int16_t*)(c->chrUPixBuf[i]))[j] = 1<<14;
1430
1431 av_assert0(c->chrDstH <= dstH);
1432
1433 if (flags & SWS_PRINT_INFO) {
1434 const char *scaler = NULL, *cpucaps;
1435
1436 for (i = 0; i < FF_ARRAY_ELEMS(scale_algorithms); i++) {
1437 if (flags & scale_algorithms[i].flag) {
1438 scaler = scale_algorithms[i].description;
1439 break;
1440 }
1441 }
1442 if (!scaler)
1443 scaler = "ehh flags invalid?!";
1444 av_log(c, AV_LOG_INFO, "%s scaler, from %s to %s%s ",
1445 scaler,
1446 av_get_pix_fmt_name(srcFormat),
1447 #ifdef DITHER1XBPP
1448 dstFormat == AV_PIX_FMT_BGR555 || dstFormat == AV_PIX_FMT_BGR565 ||
1449 dstFormat == AV_PIX_FMT_RGB444BE || dstFormat == AV_PIX_FMT_RGB444LE ||
1450 dstFormat == AV_PIX_FMT_BGR444BE || dstFormat == AV_PIX_FMT_BGR444LE ?
1451 "dithered " : "",
1452 #else
1453 "",
1454 #endif
1455 av_get_pix_fmt_name(dstFormat));
1456
1457 if (INLINE_MMXEXT(cpu_flags))
1458 cpucaps = "MMXEXT";
1459 else if (INLINE_AMD3DNOW(cpu_flags))
1460 cpucaps = "3DNOW";
1461 else if (INLINE_MMX(cpu_flags))
1462 cpucaps = "MMX";
1463 else if (PPC_ALTIVEC(cpu_flags))
1464 cpucaps = "AltiVec";
1465 else
1466 cpucaps = "C";
1467
1468 av_log(c, AV_LOG_INFO, "using %s\n", cpucaps);
1469
1470 av_log(c, AV_LOG_VERBOSE, "%dx%d -> %dx%d\n", srcW, srcH, dstW, dstH);
1471 av_log(c, AV_LOG_DEBUG,
1472 "lum srcW=%d srcH=%d dstW=%d dstH=%d xInc=%d yInc=%d\n",
1473 c->srcW, c->srcH, c->dstW, c->dstH, c->lumXInc, c->lumYInc);
1474 av_log(c, AV_LOG_DEBUG,
1475 "chr srcW=%d srcH=%d dstW=%d dstH=%d xInc=%d yInc=%d\n",
1476 c->chrSrcW, c->chrSrcH, c->chrDstW, c->chrDstH,
1477 c->chrXInc, c->chrYInc);
1478 }
1479
1480 /* unscaled special cases */
1481 if (unscaled && !usesHFilter && !usesVFilter &&
1482 (c->srcRange == c->dstRange || isAnyRGB(dstFormat))) {
1483 ff_get_unscaled_swscale(c);
1484
1485 if (c->swscale) {
1486 if (flags & SWS_PRINT_INFO)
1487 av_log(c, AV_LOG_INFO,
1488 "using unscaled %s -> %s special converter\n",
1489 av_get_pix_fmt_name(srcFormat), av_get_pix_fmt_name(dstFormat));
1490 return 0;
1491 }
1492 }
1493
1494 c->swscale = ff_getSwsFunc(c);
1495 return 0;
1496 fail: // FIXME replace things by appropriate error codes
1497 return -1;
1498 }
1499
sws_getContext(int srcW,int srcH,enum AVPixelFormat srcFormat,int dstW,int dstH,enum AVPixelFormat dstFormat,int flags,SwsFilter * srcFilter,SwsFilter * dstFilter,const double * param)1500 SwsContext *sws_getContext(int srcW, int srcH, enum AVPixelFormat srcFormat,
1501 int dstW, int dstH, enum AVPixelFormat dstFormat,
1502 int flags, SwsFilter *srcFilter,
1503 SwsFilter *dstFilter, const double *param)
1504 {
1505 SwsContext *c;
1506
1507 if (!(c = sws_alloc_context()))
1508 return NULL;
1509
1510 c->flags = flags;
1511 c->srcW = srcW;
1512 c->srcH = srcH;
1513 c->dstW = dstW;
1514 c->dstH = dstH;
1515 c->srcFormat = srcFormat;
1516 c->dstFormat = dstFormat;
1517
1518 if (param) {
1519 c->param[0] = param[0];
1520 c->param[1] = param[1];
1521 }
1522
1523 if (sws_init_context(c, srcFilter, dstFilter) < 0) {
1524 sws_freeContext(c);
1525 return NULL;
1526 }
1527
1528 return c;
1529 }
1530
sws_getDefaultFilter(float lumaGBlur,float chromaGBlur,float lumaSharpen,float chromaSharpen,float chromaHShift,float chromaVShift,int verbose)1531 SwsFilter *sws_getDefaultFilter(float lumaGBlur, float chromaGBlur,
1532 float lumaSharpen, float chromaSharpen,
1533 float chromaHShift, float chromaVShift,
1534 int verbose)
1535 {
1536 SwsFilter *filter = av_malloc(sizeof(SwsFilter));
1537 if (!filter)
1538 return NULL;
1539
1540 if (lumaGBlur != 0.0) {
1541 filter->lumH = sws_getGaussianVec(lumaGBlur, 3.0);
1542 filter->lumV = sws_getGaussianVec(lumaGBlur, 3.0);
1543 } else {
1544 filter->lumH = sws_getIdentityVec();
1545 filter->lumV = sws_getIdentityVec();
1546 }
1547
1548 if (chromaGBlur != 0.0) {
1549 filter->chrH = sws_getGaussianVec(chromaGBlur, 3.0);
1550 filter->chrV = sws_getGaussianVec(chromaGBlur, 3.0);
1551 } else {
1552 filter->chrH = sws_getIdentityVec();
1553 filter->chrV = sws_getIdentityVec();
1554 }
1555
1556 if (chromaSharpen != 0.0) {
1557 SwsVector *id = sws_getIdentityVec();
1558 sws_scaleVec(filter->chrH, -chromaSharpen);
1559 sws_scaleVec(filter->chrV, -chromaSharpen);
1560 sws_addVec(filter->chrH, id);
1561 sws_addVec(filter->chrV, id);
1562 sws_freeVec(id);
1563 }
1564
1565 if (lumaSharpen != 0.0) {
1566 SwsVector *id = sws_getIdentityVec();
1567 sws_scaleVec(filter->lumH, -lumaSharpen);
1568 sws_scaleVec(filter->lumV, -lumaSharpen);
1569 sws_addVec(filter->lumH, id);
1570 sws_addVec(filter->lumV, id);
1571 sws_freeVec(id);
1572 }
1573
1574 if (chromaHShift != 0.0)
1575 sws_shiftVec(filter->chrH, (int)(chromaHShift + 0.5));
1576
1577 if (chromaVShift != 0.0)
1578 sws_shiftVec(filter->chrV, (int)(chromaVShift + 0.5));
1579
1580 sws_normalizeVec(filter->chrH, 1.0);
1581 sws_normalizeVec(filter->chrV, 1.0);
1582 sws_normalizeVec(filter->lumH, 1.0);
1583 sws_normalizeVec(filter->lumV, 1.0);
1584
1585 if (verbose)
1586 sws_printVec2(filter->chrH, NULL, AV_LOG_DEBUG);
1587 if (verbose)
1588 sws_printVec2(filter->lumH, NULL, AV_LOG_DEBUG);
1589
1590 return filter;
1591 }
1592
sws_allocVec(int length)1593 SwsVector *sws_allocVec(int length)
1594 {
1595 SwsVector *vec;
1596
1597 if(length <= 0 || length > INT_MAX/ sizeof(double))
1598 return NULL;
1599
1600 vec = av_malloc(sizeof(SwsVector));
1601 if (!vec)
1602 return NULL;
1603 vec->length = length;
1604 vec->coeff = av_malloc(sizeof(double) * length);
1605 if (!vec->coeff)
1606 av_freep(&vec);
1607 return vec;
1608 }
1609
sws_getGaussianVec(double variance,double quality)1610 SwsVector *sws_getGaussianVec(double variance, double quality)
1611 {
1612 const int length = (int)(variance * quality + 0.5) | 1;
1613 int i;
1614 double middle = (length - 1) * 0.5;
1615 SwsVector *vec;
1616
1617 if(variance < 0 || quality < 0)
1618 return NULL;
1619
1620 vec = sws_allocVec(length);
1621
1622 if (!vec)
1623 return NULL;
1624
1625 for (i = 0; i < length; i++) {
1626 double dist = i - middle;
1627 vec->coeff[i] = exp(-dist * dist / (2 * variance * variance)) /
1628 sqrt(2 * variance * M_PI);
1629 }
1630
1631 sws_normalizeVec(vec, 1.0);
1632
1633 return vec;
1634 }
1635
sws_getConstVec(double c,int length)1636 SwsVector *sws_getConstVec(double c, int length)
1637 {
1638 int i;
1639 SwsVector *vec = sws_allocVec(length);
1640
1641 if (!vec)
1642 return NULL;
1643
1644 for (i = 0; i < length; i++)
1645 vec->coeff[i] = c;
1646
1647 return vec;
1648 }
1649
sws_getIdentityVec(void)1650 SwsVector *sws_getIdentityVec(void)
1651 {
1652 return sws_getConstVec(1.0, 1);
1653 }
1654
sws_dcVec(SwsVector * a)1655 static double sws_dcVec(SwsVector *a)
1656 {
1657 int i;
1658 double sum = 0;
1659
1660 for (i = 0; i < a->length; i++)
1661 sum += a->coeff[i];
1662
1663 return sum;
1664 }
1665
sws_scaleVec(SwsVector * a,double scalar)1666 void sws_scaleVec(SwsVector *a, double scalar)
1667 {
1668 int i;
1669
1670 for (i = 0; i < a->length; i++)
1671 a->coeff[i] *= scalar;
1672 }
1673
sws_normalizeVec(SwsVector * a,double height)1674 void sws_normalizeVec(SwsVector *a, double height)
1675 {
1676 sws_scaleVec(a, height / sws_dcVec(a));
1677 }
1678
sws_getConvVec(SwsVector * a,SwsVector * b)1679 static SwsVector *sws_getConvVec(SwsVector *a, SwsVector *b)
1680 {
1681 int length = a->length + b->length - 1;
1682 int i, j;
1683 SwsVector *vec = sws_getConstVec(0.0, length);
1684
1685 if (!vec)
1686 return NULL;
1687
1688 for (i = 0; i < a->length; i++) {
1689 for (j = 0; j < b->length; j++) {
1690 vec->coeff[i + j] += a->coeff[i] * b->coeff[j];
1691 }
1692 }
1693
1694 return vec;
1695 }
1696
sws_sumVec(SwsVector * a,SwsVector * b)1697 static SwsVector *sws_sumVec(SwsVector *a, SwsVector *b)
1698 {
1699 int length = FFMAX(a->length, b->length);
1700 int i;
1701 SwsVector *vec = sws_getConstVec(0.0, length);
1702
1703 if (!vec)
1704 return NULL;
1705
1706 for (i = 0; i < a->length; i++)
1707 vec->coeff[i + (length - 1) / 2 - (a->length - 1) / 2] += a->coeff[i];
1708 for (i = 0; i < b->length; i++)
1709 vec->coeff[i + (length - 1) / 2 - (b->length - 1) / 2] += b->coeff[i];
1710
1711 return vec;
1712 }
1713
sws_diffVec(SwsVector * a,SwsVector * b)1714 static SwsVector *sws_diffVec(SwsVector *a, SwsVector *b)
1715 {
1716 int length = FFMAX(a->length, b->length);
1717 int i;
1718 SwsVector *vec = sws_getConstVec(0.0, length);
1719
1720 if (!vec)
1721 return NULL;
1722
1723 for (i = 0; i < a->length; i++)
1724 vec->coeff[i + (length - 1) / 2 - (a->length - 1) / 2] += a->coeff[i];
1725 for (i = 0; i < b->length; i++)
1726 vec->coeff[i + (length - 1) / 2 - (b->length - 1) / 2] -= b->coeff[i];
1727
1728 return vec;
1729 }
1730
1731 /* shift left / or right if "shift" is negative */
sws_getShiftedVec(SwsVector * a,int shift)1732 static SwsVector *sws_getShiftedVec(SwsVector *a, int shift)
1733 {
1734 int length = a->length + FFABS(shift) * 2;
1735 int i;
1736 SwsVector *vec = sws_getConstVec(0.0, length);
1737
1738 if (!vec)
1739 return NULL;
1740
1741 for (i = 0; i < a->length; i++) {
1742 vec->coeff[i + (length - 1) / 2 -
1743 (a->length - 1) / 2 - shift] = a->coeff[i];
1744 }
1745
1746 return vec;
1747 }
1748
sws_shiftVec(SwsVector * a,int shift)1749 void sws_shiftVec(SwsVector *a, int shift)
1750 {
1751 SwsVector *shifted = sws_getShiftedVec(a, shift);
1752 av_free(a->coeff);
1753 a->coeff = shifted->coeff;
1754 a->length = shifted->length;
1755 av_free(shifted);
1756 }
1757
sws_addVec(SwsVector * a,SwsVector * b)1758 void sws_addVec(SwsVector *a, SwsVector *b)
1759 {
1760 SwsVector *sum = sws_sumVec(a, b);
1761 av_free(a->coeff);
1762 a->coeff = sum->coeff;
1763 a->length = sum->length;
1764 av_free(sum);
1765 }
1766
sws_subVec(SwsVector * a,SwsVector * b)1767 void sws_subVec(SwsVector *a, SwsVector *b)
1768 {
1769 SwsVector *diff = sws_diffVec(a, b);
1770 av_free(a->coeff);
1771 a->coeff = diff->coeff;
1772 a->length = diff->length;
1773 av_free(diff);
1774 }
1775
sws_convVec(SwsVector * a,SwsVector * b)1776 void sws_convVec(SwsVector *a, SwsVector *b)
1777 {
1778 SwsVector *conv = sws_getConvVec(a, b);
1779 av_free(a->coeff);
1780 a->coeff = conv->coeff;
1781 a->length = conv->length;
1782 av_free(conv);
1783 }
1784
sws_cloneVec(SwsVector * a)1785 SwsVector *sws_cloneVec(SwsVector *a)
1786 {
1787 SwsVector *vec = sws_allocVec(a->length);
1788
1789 if (!vec)
1790 return NULL;
1791
1792 memcpy(vec->coeff, a->coeff, a->length * sizeof(*a->coeff));
1793
1794 return vec;
1795 }
1796
sws_printVec2(SwsVector * a,AVClass * log_ctx,int log_level)1797 void sws_printVec2(SwsVector *a, AVClass *log_ctx, int log_level)
1798 {
1799 int i;
1800 double max = 0;
1801 double min = 0;
1802 double range;
1803
1804 for (i = 0; i < a->length; i++)
1805 if (a->coeff[i] > max)
1806 max = a->coeff[i];
1807
1808 for (i = 0; i < a->length; i++)
1809 if (a->coeff[i] < min)
1810 min = a->coeff[i];
1811
1812 range = max - min;
1813
1814 for (i = 0; i < a->length; i++) {
1815 int x = (int)((a->coeff[i] - min) * 60.0 / range + 0.5);
1816 av_log(log_ctx, log_level, "%1.3f ", a->coeff[i]);
1817 for (; x > 0; x--)
1818 av_log(log_ctx, log_level, " ");
1819 av_log(log_ctx, log_level, "|\n");
1820 }
1821 }
1822
sws_freeVec(SwsVector * a)1823 void sws_freeVec(SwsVector *a)
1824 {
1825 if (!a)
1826 return;
1827 av_freep(&a->coeff);
1828 a->length = 0;
1829 av_free(a);
1830 }
1831
sws_freeFilter(SwsFilter * filter)1832 void sws_freeFilter(SwsFilter *filter)
1833 {
1834 if (!filter)
1835 return;
1836
1837 sws_freeVec(filter->lumH);
1838 sws_freeVec(filter->lumV);
1839 sws_freeVec(filter->chrH);
1840 sws_freeVec(filter->chrV);
1841 av_free(filter);
1842 }
1843
sws_freeContext(SwsContext * c)1844 void sws_freeContext(SwsContext *c)
1845 {
1846 int i;
1847 if (!c)
1848 return;
1849
1850 if (c->lumPixBuf) {
1851 for (i = 0; i < c->vLumBufSize; i++)
1852 av_freep(&c->lumPixBuf[i]);
1853 av_freep(&c->lumPixBuf);
1854 }
1855
1856 if (c->chrUPixBuf) {
1857 for (i = 0; i < c->vChrBufSize; i++)
1858 av_freep(&c->chrUPixBuf[i]);
1859 av_freep(&c->chrUPixBuf);
1860 av_freep(&c->chrVPixBuf);
1861 }
1862
1863 if (CONFIG_SWSCALE_ALPHA && c->alpPixBuf) {
1864 for (i = 0; i < c->vLumBufSize; i++)
1865 av_freep(&c->alpPixBuf[i]);
1866 av_freep(&c->alpPixBuf);
1867 }
1868
1869 for (i = 0; i < 4; i++)
1870 av_freep(&c->dither_error[i]);
1871
1872 av_freep(&c->vLumFilter);
1873 av_freep(&c->vChrFilter);
1874 av_freep(&c->hLumFilter);
1875 av_freep(&c->hChrFilter);
1876 #if HAVE_ALTIVEC
1877 av_freep(&c->vYCoeffsBank);
1878 av_freep(&c->vCCoeffsBank);
1879 #endif
1880
1881 av_freep(&c->vLumFilterPos);
1882 av_freep(&c->vChrFilterPos);
1883 av_freep(&c->hLumFilterPos);
1884 av_freep(&c->hChrFilterPos);
1885
1886 #if HAVE_MMX_INLINE
1887 #if USE_MMAP
1888 if (c->lumMmxextFilterCode)
1889 munmap(c->lumMmxextFilterCode, c->lumMmxextFilterCodeSize);
1890 if (c->chrMmxextFilterCode)
1891 munmap(c->chrMmxextFilterCode, c->chrMmxextFilterCodeSize);
1892 #elif HAVE_VIRTUALALLOC
1893 if (c->lumMmxextFilterCode)
1894 VirtualFree(c->lumMmxextFilterCode, 0, MEM_RELEASE);
1895 if (c->chrMmxextFilterCode)
1896 VirtualFree(c->chrMmxextFilterCode, 0, MEM_RELEASE);
1897 #else
1898 av_free(c->lumMmxextFilterCode);
1899 av_free(c->chrMmxextFilterCode);
1900 #endif
1901 c->lumMmxextFilterCode = NULL;
1902 c->chrMmxextFilterCode = NULL;
1903 #endif /* HAVE_MMX_INLINE */
1904
1905 av_freep(&c->yuvTable);
1906 av_freep(&c->formatConvBuffer);
1907
1908 av_free(c);
1909 }
1910
sws_getCachedContext(struct SwsContext * context,int srcW,int srcH,enum AVPixelFormat srcFormat,int dstW,int dstH,enum AVPixelFormat dstFormat,int flags,SwsFilter * srcFilter,SwsFilter * dstFilter,const double * param)1911 struct SwsContext *sws_getCachedContext(struct SwsContext *context, int srcW,
1912 int srcH, enum AVPixelFormat srcFormat,
1913 int dstW, int dstH,
1914 enum AVPixelFormat dstFormat, int flags,
1915 SwsFilter *srcFilter,
1916 SwsFilter *dstFilter,
1917 const double *param)
1918 {
1919 static const double default_param[2] = { SWS_PARAM_DEFAULT,
1920 SWS_PARAM_DEFAULT };
1921
1922 if (!param)
1923 param = default_param;
1924
1925 if (context &&
1926 (context->srcW != srcW ||
1927 context->srcH != srcH ||
1928 context->srcFormat != srcFormat ||
1929 context->dstW != dstW ||
1930 context->dstH != dstH ||
1931 context->dstFormat != dstFormat ||
1932 context->flags != flags ||
1933 context->param[0] != param[0] ||
1934 context->param[1] != param[1])) {
1935 sws_freeContext(context);
1936 context = NULL;
1937 }
1938
1939 if (!context) {
1940 if (!(context = sws_alloc_context()))
1941 return NULL;
1942 context->srcW = srcW;
1943 context->srcH = srcH;
1944 context->srcFormat = srcFormat;
1945 context->dstW = dstW;
1946 context->dstH = dstH;
1947 context->dstFormat = dstFormat;
1948 context->flags = flags;
1949 context->param[0] = param[0];
1950 context->param[1] = param[1];
1951 if (sws_init_context(context, srcFilter, dstFilter) < 0) {
1952 sws_freeContext(context);
1953 return NULL;
1954 }
1955 }
1956 return context;
1957 }
1958