2 * Copyright (C) 2001-2011 Michael Niedermayer <michaelni@gmx.at>
4 * This file is part of FFmpeg.
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.
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.
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
21 #ifndef SWSCALE_SWSCALE_INTERNAL_H
22 #define SWSCALE_SWSCALE_INTERNAL_H
30 #include "libavutil/avutil.h"
31 #include "libavutil/log.h"
32 #include "libavutil/pixfmt.h"
33 #include "libavutil/pixdesc.h"
35 #define STR(s) AV_TOSTRING(s) // AV_STRINGIFY is too long
37 #define YUVRGB_TABLE_HEADROOM 128
39 #define FAST_BGR2YV12 // use 7-bit instead of 15-bit coefficients
41 #define MAX_FILTER_SIZE 256
46 #define ALT32_CORR (-1)
63 typedef int (*SwsFunc)(struct SwsContext *context, const uint8_t *src[],
64 int srcStride[], int srcSliceY, int srcSliceH,
65 uint8_t *dst[], int dstStride[]);
68 * Write one line of horizontally scaled data to planar output
69 * without any additional vertical scaling (or point-scaling).
71 * @param src scaled source data, 15bit for 8-10bit output,
72 * 19-bit for 16bit output (in int32_t)
73 * @param dest pointer to the output plane. For >8bit
74 * output, this is in uint16_t
75 * @param dstW width of destination in pixels
76 * @param dither ordered dither array of type int16_t and size 8
77 * @param offset Dither offset
79 typedef void (*yuv2planar1_fn)(const int16_t *src, uint8_t *dest, int dstW,
80 const uint8_t *dither, int offset);
83 * Write one line of horizontally scaled data to planar output
84 * with multi-point vertical scaling between input pixels.
86 * @param filter vertical luma/alpha scaling coefficients, 12bit [0,4096]
87 * @param src scaled luma (Y) or alpha (A) source data, 15bit for 8-10bit output,
88 * 19-bit for 16bit output (in int32_t)
89 * @param filterSize number of vertical input lines to scale
90 * @param dest pointer to output plane. For >8bit
91 * output, this is in uint16_t
92 * @param dstW width of destination pixels
93 * @param offset Dither offset
95 typedef void (*yuv2planarX_fn)(const int16_t *filter, int filterSize,
96 const int16_t **src, uint8_t *dest, int dstW,
97 const uint8_t *dither, int offset);
100 * Write one line of horizontally scaled chroma to interleaved output
101 * with multi-point vertical scaling between input pixels.
103 * @param c SWS scaling context
104 * @param chrFilter vertical chroma scaling coefficients, 12bit [0,4096]
105 * @param chrUSrc scaled chroma (U) source data, 15bit for 8-10bit output,
106 * 19-bit for 16bit output (in int32_t)
107 * @param chrVSrc scaled chroma (V) source data, 15bit for 8-10bit output,
108 * 19-bit for 16bit output (in int32_t)
109 * @param chrFilterSize number of vertical chroma input lines to scale
110 * @param dest pointer to the output plane. For >8bit
111 * output, this is in uint16_t
112 * @param dstW width of chroma planes
114 typedef void (*yuv2interleavedX_fn)(struct SwsContext *c,
115 const int16_t *chrFilter,
117 const int16_t **chrUSrc,
118 const int16_t **chrVSrc,
119 uint8_t *dest, int dstW);
122 * Write one line of horizontally scaled Y/U/V/A to packed-pixel YUV/RGB
123 * output without any additional vertical scaling (or point-scaling). Note
124 * that this function may do chroma scaling, see the "uvalpha" argument.
126 * @param c SWS scaling context
127 * @param lumSrc scaled luma (Y) source data, 15bit for 8-10bit output,
128 * 19-bit for 16bit output (in int32_t)
129 * @param chrUSrc scaled chroma (U) source data, 15bit for 8-10bit output,
130 * 19-bit for 16bit output (in int32_t)
131 * @param chrVSrc scaled chroma (V) source data, 15bit for 8-10bit output,
132 * 19-bit for 16bit output (in int32_t)
133 * @param alpSrc scaled alpha (A) source data, 15bit for 8-10bit output,
134 * 19-bit for 16bit output (in int32_t)
135 * @param dest pointer to the output plane. For 16bit output, this is
137 * @param dstW width of lumSrc and alpSrc in pixels, number of pixels
138 * to write into dest[]
139 * @param uvalpha chroma scaling coefficient for the second line of chroma
140 * pixels, either 2048 or 0. If 0, one chroma input is used
141 * for 2 output pixels (or if the SWS_FLAG_FULL_CHR_INT flag
142 * is set, it generates 1 output pixel). If 2048, two chroma
143 * input pixels should be averaged for 2 output pixels (this
144 * only happens if SWS_FLAG_FULL_CHR_INT is not set)
145 * @param y vertical line number for this output. This does not need
146 * to be used to calculate the offset in the destination,
147 * but can be used to generate comfort noise using dithering
148 * for some output formats.
150 typedef void (*yuv2packed1_fn)(struct SwsContext *c, const int16_t *lumSrc,
151 const int16_t *chrUSrc[2],
152 const int16_t *chrVSrc[2],
153 const int16_t *alpSrc, uint8_t *dest,
154 int dstW, int uvalpha, int y);
156 * Write one line of horizontally scaled Y/U/V/A to packed-pixel YUV/RGB
157 * output by doing bilinear scaling between two input lines.
159 * @param c SWS scaling context
160 * @param lumSrc scaled luma (Y) source data, 15bit for 8-10bit output,
161 * 19-bit for 16bit output (in int32_t)
162 * @param chrUSrc scaled chroma (U) source data, 15bit for 8-10bit output,
163 * 19-bit for 16bit output (in int32_t)
164 * @param chrVSrc scaled chroma (V) source data, 15bit for 8-10bit output,
165 * 19-bit for 16bit output (in int32_t)
166 * @param alpSrc scaled alpha (A) source data, 15bit for 8-10bit output,
167 * 19-bit for 16bit output (in int32_t)
168 * @param dest pointer to the output plane. For 16bit output, this is
170 * @param dstW width of lumSrc and alpSrc in pixels, number of pixels
171 * to write into dest[]
172 * @param yalpha luma/alpha scaling coefficients for the second input line.
173 * The first line's coefficients can be calculated by using
175 * @param uvalpha chroma scaling coefficient for the second input line. The
176 * first line's coefficients can be calculated by using
178 * @param y vertical line number for this output. This does not need
179 * to be used to calculate the offset in the destination,
180 * but can be used to generate comfort noise using dithering
181 * for some output formats.
183 typedef void (*yuv2packed2_fn)(struct SwsContext *c, const int16_t *lumSrc[2],
184 const int16_t *chrUSrc[2],
185 const int16_t *chrVSrc[2],
186 const int16_t *alpSrc[2],
188 int dstW, int yalpha, int uvalpha, int y);
190 * Write one line of horizontally scaled Y/U/V/A to packed-pixel YUV/RGB
191 * output by doing multi-point vertical scaling between input pixels.
193 * @param c SWS scaling context
194 * @param lumFilter vertical luma/alpha scaling coefficients, 12bit [0,4096]
195 * @param lumSrc scaled luma (Y) source data, 15bit for 8-10bit output,
196 * 19-bit for 16bit output (in int32_t)
197 * @param lumFilterSize number of vertical luma/alpha input lines to scale
198 * @param chrFilter vertical chroma scaling coefficients, 12bit [0,4096]
199 * @param chrUSrc scaled chroma (U) source data, 15bit for 8-10bit output,
200 * 19-bit for 16bit output (in int32_t)
201 * @param chrVSrc scaled chroma (V) source data, 15bit for 8-10bit output,
202 * 19-bit for 16bit output (in int32_t)
203 * @param chrFilterSize number of vertical chroma input lines to scale
204 * @param alpSrc scaled alpha (A) source data, 15bit for 8-10bit output,
205 * 19-bit for 16bit output (in int32_t)
206 * @param dest pointer to the output plane. For 16bit output, this is
208 * @param dstW width of lumSrc and alpSrc in pixels, number of pixels
209 * to write into dest[]
210 * @param y vertical line number for this output. This does not need
211 * to be used to calculate the offset in the destination,
212 * but can be used to generate comfort noise using dithering
213 * or some output formats.
215 typedef void (*yuv2packedX_fn)(struct SwsContext *c, const int16_t *lumFilter,
216 const int16_t **lumSrc, int lumFilterSize,
217 const int16_t *chrFilter,
218 const int16_t **chrUSrc,
219 const int16_t **chrVSrc, int chrFilterSize,
220 const int16_t **alpSrc, uint8_t *dest,
223 /* This struct should be aligned on at least a 32-byte boundary. */
224 typedef struct SwsContext {
226 * info on struct for av_log
228 const AVClass *av_class;
231 * Note that src, dst, srcStride, dstStride will be copied in the
232 * sws_scale() wrapper so they can be freely modified here.
235 int srcW; ///< Width of source luma/alpha planes.
236 int srcH; ///< Height of source luma/alpha planes.
237 int dstH; ///< Height of destination luma/alpha planes.
238 int chrSrcW; ///< Width of source chroma planes.
239 int chrSrcH; ///< Height of source chroma planes.
240 int chrDstW; ///< Width of destination chroma planes.
241 int chrDstH; ///< Height of destination chroma planes.
242 int lumXInc, chrXInc;
243 int lumYInc, chrYInc;
244 enum PixelFormat dstFormat; ///< Destination pixel format.
245 enum PixelFormat srcFormat; ///< Source pixel format.
246 int dstFormatBpp; ///< Number of bits per pixel of the destination pixel format.
247 int srcFormatBpp; ///< Number of bits per pixel of the source pixel format.
249 int chrSrcHSubSample; ///< Binary logarithm of horizontal subsampling factor between luma/alpha and chroma planes in source image.
250 int chrSrcVSubSample; ///< Binary logarithm of vertical subsampling factor between luma/alpha and chroma planes in source image.
251 int chrDstHSubSample; ///< Binary logarithm of horizontal subsampling factor between luma/alpha and chroma planes in destination image.
252 int chrDstVSubSample; ///< Binary logarithm of vertical subsampling factor between luma/alpha and chroma planes in destination image.
253 int vChrDrop; ///< Binary logarithm of extra vertical subsampling factor in source image chroma planes specified by user.
254 int sliceDir; ///< Direction that slices are fed to the scaler (1 = top-to-bottom, -1 = bottom-to-top).
255 double param[2]; ///< Input parameters for scaling algorithms that need them.
257 uint32_t pal_yuv[256];
258 uint32_t pal_rgb[256];
261 * @name Scaled horizontal lines ring buffer.
262 * The horizontal scaler keeps just enough scaled lines in a ring buffer
263 * so they may be passed to the vertical scaler. The pointers to the
264 * allocated buffers for each line are duplicated in sequence in the ring
265 * buffer to simplify indexing and avoid wrapping around between lines
266 * inside the vertical scaler code. The wrapping is done before the
267 * vertical scaler is called.
270 int16_t **lumPixBuf; ///< Ring buffer for scaled horizontal luma plane lines to be fed to the vertical scaler.
271 int16_t **chrUPixBuf; ///< Ring buffer for scaled horizontal chroma plane lines to be fed to the vertical scaler.
272 int16_t **chrVPixBuf; ///< Ring buffer for scaled horizontal chroma plane lines to be fed to the vertical scaler.
273 int16_t **alpPixBuf; ///< Ring buffer for scaled horizontal alpha plane lines to be fed to the vertical scaler.
274 int vLumBufSize; ///< Number of vertical luma/alpha lines allocated in the ring buffer.
275 int vChrBufSize; ///< Number of vertical chroma lines allocated in the ring buffer.
276 int lastInLumBuf; ///< Last scaled horizontal luma/alpha line from source in the ring buffer.
277 int lastInChrBuf; ///< Last scaled horizontal chroma line from source in the ring buffer.
278 int lumBufIndex; ///< Index in ring buffer of the last scaled horizontal luma/alpha line from source.
279 int chrBufIndex; ///< Index in ring buffer of the last scaled horizontal chroma line from source.
282 uint8_t *formatConvBuffer;
285 * @name Horizontal and vertical filters.
286 * To better understand the following fields, here is a pseudo-code of
287 * their usage in filtering a horizontal line:
289 * for (i = 0; i < width; i++) {
291 * for (j = 0; j < filterSize; j++)
292 * dst[i] += src[ filterPos[i] + j ] * filter[ filterSize * i + j ];
293 * dst[i] >>= FRAC_BITS; // The actual implementation is fixed-point.
298 int16_t *hLumFilter; ///< Array of horizontal filter coefficients for luma/alpha planes.
299 int16_t *hChrFilter; ///< Array of horizontal filter coefficients for chroma planes.
300 int16_t *vLumFilter; ///< Array of vertical filter coefficients for luma/alpha planes.
301 int16_t *vChrFilter; ///< Array of vertical filter coefficients for chroma planes.
302 int32_t *hLumFilterPos; ///< Array of horizontal filter starting positions for each dst[i] for luma/alpha planes.
303 int32_t *hChrFilterPos; ///< Array of horizontal filter starting positions for each dst[i] for chroma planes.
304 int32_t *vLumFilterPos; ///< Array of vertical filter starting positions for each dst[i] for luma/alpha planes.
305 int32_t *vChrFilterPos; ///< Array of vertical filter starting positions for each dst[i] for chroma planes.
306 int hLumFilterSize; ///< Horizontal filter size for luma/alpha pixels.
307 int hChrFilterSize; ///< Horizontal filter size for chroma pixels.
308 int vLumFilterSize; ///< Vertical filter size for luma/alpha pixels.
309 int vChrFilterSize; ///< Vertical filter size for chroma pixels.
312 int lumMmx2FilterCodeSize; ///< Runtime-generated MMX2 horizontal fast bilinear scaler code size for luma/alpha planes.
313 int chrMmx2FilterCodeSize; ///< Runtime-generated MMX2 horizontal fast bilinear scaler code size for chroma planes.
314 uint8_t *lumMmx2FilterCode; ///< Runtime-generated MMX2 horizontal fast bilinear scaler code for luma/alpha planes.
315 uint8_t *chrMmx2FilterCode; ///< Runtime-generated MMX2 horizontal fast bilinear scaler code for chroma planes.
319 int dstY; ///< Last destination vertical line output from last slice.
320 int flags; ///< Flags passed by the user to select scaler algorithm, optimizations, subsampling, etc...
321 void *yuvTable; // pointer to the yuv->rgb table start so it can be freed()
322 uint8_t *table_rV[256 + 2*YUVRGB_TABLE_HEADROOM];
323 uint8_t *table_gU[256 + 2*YUVRGB_TABLE_HEADROOM];
324 int table_gV[256 + 2*YUVRGB_TABLE_HEADROOM];
325 uint8_t *table_bU[256 + 2*YUVRGB_TABLE_HEADROOM];
328 int contrast, brightness, saturation; // for sws_getColorspaceDetails
329 int srcColorspaceTable[4];
330 int dstColorspaceTable[4];
331 int srcRange; ///< 0 = MPG YUV range, 1 = JPG YUV range (source image).
332 int dstRange; ///< 0 = MPG YUV range, 1 = JPG YUV range (destination image).
335 int yuv2rgb_y_offset;
337 int yuv2rgb_v2r_coeff;
338 int yuv2rgb_v2g_coeff;
339 int yuv2rgb_u2g_coeff;
340 int yuv2rgb_u2b_coeff;
342 #define RED_DITHER "0*8"
343 #define GREEN_DITHER "1*8"
344 #define BLUE_DITHER "2*8"
345 #define Y_COEFF "3*8"
346 #define VR_COEFF "4*8"
347 #define UB_COEFF "5*8"
348 #define VG_COEFF "6*8"
349 #define UG_COEFF "7*8"
350 #define Y_OFFSET "8*8"
351 #define U_OFFSET "9*8"
352 #define V_OFFSET "10*8"
353 #define LUM_MMX_FILTER_OFFSET "11*8"
354 #define CHR_MMX_FILTER_OFFSET "11*8+4*4*256"
355 #define DSTW_OFFSET "11*8+4*4*256*2" //do not change, it is hardcoded in the ASM
356 #define ESP_OFFSET "11*8+4*4*256*2+8"
357 #define VROUNDER_OFFSET "11*8+4*4*256*2+16"
358 #define U_TEMP "11*8+4*4*256*2+24"
359 #define V_TEMP "11*8+4*4*256*2+32"
360 #define Y_TEMP "11*8+4*4*256*2+40"
361 #define ALP_MMX_FILTER_OFFSET "11*8+4*4*256*2+48"
362 #define UV_OFF_PX "11*8+4*4*256*3+48"
363 #define UV_OFF_BYTE "11*8+4*4*256*3+56"
364 #define DITHER16 "11*8+4*4*256*3+64"
365 #define DITHER32 "11*8+4*4*256*3+80"
367 DECLARE_ALIGNED(8, uint64_t, redDither);
368 DECLARE_ALIGNED(8, uint64_t, greenDither);
369 DECLARE_ALIGNED(8, uint64_t, blueDither);
371 DECLARE_ALIGNED(8, uint64_t, yCoeff);
372 DECLARE_ALIGNED(8, uint64_t, vrCoeff);
373 DECLARE_ALIGNED(8, uint64_t, ubCoeff);
374 DECLARE_ALIGNED(8, uint64_t, vgCoeff);
375 DECLARE_ALIGNED(8, uint64_t, ugCoeff);
376 DECLARE_ALIGNED(8, uint64_t, yOffset);
377 DECLARE_ALIGNED(8, uint64_t, uOffset);
378 DECLARE_ALIGNED(8, uint64_t, vOffset);
379 int32_t lumMmxFilter[4 * MAX_FILTER_SIZE];
380 int32_t chrMmxFilter[4 * MAX_FILTER_SIZE];
381 int dstW; ///< Width of destination luma/alpha planes.
382 DECLARE_ALIGNED(8, uint64_t, esp);
383 DECLARE_ALIGNED(8, uint64_t, vRounder);
384 DECLARE_ALIGNED(8, uint64_t, u_temp);
385 DECLARE_ALIGNED(8, uint64_t, v_temp);
386 DECLARE_ALIGNED(8, uint64_t, y_temp);
387 int32_t alpMmxFilter[4 * MAX_FILTER_SIZE];
388 // alignment of these values is not necessary, but merely here
389 // to maintain the same offset across x8632 and x86-64. Once we
390 // use proper offset macros in the asm, they can be removed.
391 DECLARE_ALIGNED(8, ptrdiff_t, uv_off); ///< offset (in pixels) between u and v planes
392 DECLARE_ALIGNED(8, ptrdiff_t, uv_offx2); ///< offset (in bytes) between u and v planes
393 DECLARE_ALIGNED(8, uint16_t, dither16)[8];
394 DECLARE_ALIGNED(8, uint32_t, dither32)[8];
396 const uint8_t *chrDither8, *lumDither8;
399 vector signed short CY;
400 vector signed short CRV;
401 vector signed short CBU;
402 vector signed short CGU;
403 vector signed short CGV;
404 vector signed short OY;
405 vector unsigned short CSHIFT;
406 vector signed short *vYCoeffsBank, *vCCoeffsBank;
410 DECLARE_ALIGNED(4, uint32_t, oy);
411 DECLARE_ALIGNED(4, uint32_t, oc);
412 DECLARE_ALIGNED(4, uint32_t, zero);
413 DECLARE_ALIGNED(4, uint32_t, cy);
414 DECLARE_ALIGNED(4, uint32_t, crv);
415 DECLARE_ALIGNED(4, uint32_t, rmask);
416 DECLARE_ALIGNED(4, uint32_t, cbu);
417 DECLARE_ALIGNED(4, uint32_t, bmask);
418 DECLARE_ALIGNED(4, uint32_t, cgu);
419 DECLARE_ALIGNED(4, uint32_t, cgv);
420 DECLARE_ALIGNED(4, uint32_t, gmask);
424 DECLARE_ALIGNED(8, uint64_t, sparc_coeffs)[10];
428 /* function pointers for swScale() */
429 yuv2planar1_fn yuv2plane1;
430 yuv2planarX_fn yuv2planeX;
431 yuv2interleavedX_fn yuv2nv12cX;
432 yuv2packed1_fn yuv2packed1;
433 yuv2packed2_fn yuv2packed2;
434 yuv2packedX_fn yuv2packedX;
436 /// Unscaled conversion of luma plane to YV12 for horizontal scaler.
437 void (*lumToYV12)(uint8_t *dst, const uint8_t *src, const uint8_t *src2, const uint8_t *src3,
438 int width, uint32_t *pal);
439 /// Unscaled conversion of alpha plane to YV12 for horizontal scaler.
440 void (*alpToYV12)(uint8_t *dst, const uint8_t *src, const uint8_t *src2, const uint8_t *src3,
441 int width, uint32_t *pal);
442 /// Unscaled conversion of chroma planes to YV12 for horizontal scaler.
443 void (*chrToYV12)(uint8_t *dstU, uint8_t *dstV,
444 const uint8_t *src1, const uint8_t *src2, const uint8_t *src3,
445 int width, uint32_t *pal);
448 * Functions to read planar input, such as planar RGB, and convert
449 * internally to Y/UV.
452 void (*readLumPlanar)(uint8_t *dst, const uint8_t *src[4], int width);
453 void (*readChrPlanar)(uint8_t *dstU, uint8_t *dstV, const uint8_t *src[4],
458 * Scale one horizontal line of input data using a bilinear filter
459 * to produce one line of output data. Compared to SwsContext->hScale(),
460 * please take note of the following caveats when using these:
461 * - Scaling is done using only 7bit instead of 14bit coefficients.
462 * - You can use no more than 5 input pixels to produce 4 output
463 * pixels. Therefore, this filter should not be used for downscaling
464 * by more than ~20% in width (because that equals more than 5/4th
465 * downscaling and thus more than 5 pixels input per 4 pixels output).
466 * - In general, bilinear filters create artifacts during downscaling
467 * (even when <20%), because one output pixel will span more than one
468 * input pixel, and thus some pixels will need edges of both neighbor
469 * pixels to interpolate the output pixel. Since you can use at most
470 * two input pixels per output pixel in bilinear scaling, this is
471 * impossible and thus downscaling by any size will create artifacts.
472 * To enable this type of scaling, set SWS_FLAG_FAST_BILINEAR
473 * in SwsContext->flags.
476 void (*hyscale_fast)(struct SwsContext *c,
477 int16_t *dst, int dstWidth,
478 const uint8_t *src, int srcW, int xInc);
479 void (*hcscale_fast)(struct SwsContext *c,
480 int16_t *dst1, int16_t *dst2, int dstWidth,
481 const uint8_t *src1, const uint8_t *src2,
486 * Scale one horizontal line of input data using a filter over the input
487 * lines, to produce one (differently sized) line of output data.
489 * @param dst pointer to destination buffer for horizontally scaled
490 * data. If the number of bits per component of one
491 * destination pixel (SwsContext->dstBpc) is <= 10, data
492 * will be 15bpc in 16bits (int16_t) width. Else (i.e.
493 * SwsContext->dstBpc == 16), data will be 19bpc in
494 * 32bits (int32_t) width.
495 * @param dstW width of destination image
496 * @param src pointer to source data to be scaled. If the number of
497 * bits per component of a source pixel (SwsContext->srcBpc)
498 * is 8, this is 8bpc in 8bits (uint8_t) width. Else
499 * (i.e. SwsContext->dstBpc > 8), this is native depth
500 * in 16bits (uint16_t) width. In other words, for 9-bit
501 * YUV input, this is 9bpc, for 10-bit YUV input, this is
502 * 10bpc, and for 16-bit RGB or YUV, this is 16bpc.
503 * @param filter filter coefficients to be used per output pixel for
504 * scaling. This contains 14bpp filtering coefficients.
505 * Guaranteed to contain dstW * filterSize entries.
506 * @param filterPos position of the first input pixel to be used for
507 * each output pixel during scaling. Guaranteed to
508 * contain dstW entries.
509 * @param filterSize the number of input coefficients to be used (and
510 * thus the number of input pixels to be used) for
511 * creating a single output pixel. Is aligned to 4
512 * (and input coefficients thus padded with zeroes)
513 * to simplify creating SIMD code.
516 void (*hyScale)(struct SwsContext *c, int16_t *dst, int dstW,
517 const uint8_t *src, const int16_t *filter,
518 const int32_t *filterPos, int filterSize);
519 void (*hcScale)(struct SwsContext *c, int16_t *dst, int dstW,
520 const uint8_t *src, const int16_t *filter,
521 const int32_t *filterPos, int filterSize);
524 /// Color range conversion function for luma plane if needed.
525 void (*lumConvertRange)(int16_t *dst, int width);
526 /// Color range conversion function for chroma planes if needed.
527 void (*chrConvertRange)(int16_t *dst1, int16_t *dst2, int width);
529 int needs_hcscale; ///< Set if there are chroma planes to be converted.
531 //FIXME check init (where 0)
533 SwsFunc ff_yuv2rgb_get_func_ptr(SwsContext *c);
534 int ff_yuv2rgb_c_init_tables(SwsContext *c, const int inv_table[4],
535 int fullRange, int brightness,
536 int contrast, int saturation);
538 void ff_yuv2rgb_init_tables_altivec(SwsContext *c, const int inv_table[4],
539 int brightness, int contrast, int saturation);
540 void updateMMXDitherTables(SwsContext *c, int dstY, int lumBufIndex, int chrBufIndex,
541 int lastInLumBuf, int lastInChrBuf);
543 SwsFunc ff_yuv2rgb_init_mmx(SwsContext *c);
544 SwsFunc ff_yuv2rgb_init_vis(SwsContext *c);
545 SwsFunc ff_yuv2rgb_init_altivec(SwsContext *c);
546 SwsFunc ff_yuv2rgb_get_func_ptr_bfin(SwsContext *c);
547 void ff_bfin_get_unscaled_swscale(SwsContext *c);
549 #if FF_API_SWS_FORMAT_NAME
551 * @deprecated Use av_get_pix_fmt_name() instead.
554 const char *sws_format_name(enum PixelFormat format);
558 (av_pix_fmt_descriptors[x].comp[0].depth_minus1 == 15)
560 #define is9_OR_10BPS(x) \
561 (av_pix_fmt_descriptors[x].comp[0].depth_minus1 == 8 || \
562 av_pix_fmt_descriptors[x].comp[0].depth_minus1 == 9)
564 #define isNBPS(x) is9_OR_10BPS(x)
567 (av_pix_fmt_descriptors[x].flags & PIX_FMT_BE)
570 (!(av_pix_fmt_descriptors[x].flags & PIX_FMT_RGB) && \
571 av_pix_fmt_descriptors[x].nb_components >= 2)
573 #define isPlanarYUV(x) \
574 ((av_pix_fmt_descriptors[x].flags & PIX_FMT_PLANAR) && \
578 (av_pix_fmt_descriptors[x].flags & PIX_FMT_RGB)
581 (!(av_pix_fmt_descriptors[x].flags & PIX_FMT_PAL) && \
582 av_pix_fmt_descriptors[x].nb_components <= 2)
585 ((x) == PIX_FMT_GRAY8 || \
586 (x) == PIX_FMT_Y400A || \
587 (x) == PIX_FMT_GRAY16BE || \
588 (x) == PIX_FMT_GRAY16LE)
591 #define isRGBinInt(x) \
593 (x)==PIX_FMT_RGB48BE || \
594 (x)==PIX_FMT_RGB48LE || \
595 (x)==PIX_FMT_RGBA64BE || \
596 (x)==PIX_FMT_RGBA64LE || \
597 (x)==PIX_FMT_RGB32 || \
598 (x)==PIX_FMT_RGB32_1 || \
599 (x)==PIX_FMT_RGB24 || \
600 (x)==PIX_FMT_RGB565BE || \
601 (x)==PIX_FMT_RGB565LE || \
602 (x)==PIX_FMT_RGB555BE || \
603 (x)==PIX_FMT_RGB555LE || \
604 (x)==PIX_FMT_RGB444BE || \
605 (x)==PIX_FMT_RGB444LE || \
606 (x)==PIX_FMT_RGB8 || \
607 (x)==PIX_FMT_RGB4 || \
608 (x)==PIX_FMT_RGB4_BYTE || \
609 (x)==PIX_FMT_MONOBLACK || \
610 (x)==PIX_FMT_MONOWHITE \
612 #define isBGRinInt(x) \
614 (x)==PIX_FMT_BGR48BE || \
615 (x)==PIX_FMT_BGR48LE || \
616 (x)==PIX_FMT_BGRA64BE || \
617 (x)==PIX_FMT_BGRA64LE || \
618 (x)==PIX_FMT_BGR32 || \
619 (x)==PIX_FMT_BGR32_1 || \
620 (x)==PIX_FMT_BGR24 || \
621 (x)==PIX_FMT_BGR565BE || \
622 (x)==PIX_FMT_BGR565LE || \
623 (x)==PIX_FMT_BGR555BE || \
624 (x)==PIX_FMT_BGR555LE || \
625 (x)==PIX_FMT_BGR444BE || \
626 (x)==PIX_FMT_BGR444LE || \
627 (x)==PIX_FMT_BGR8 || \
628 (x)==PIX_FMT_BGR4 || \
629 (x)==PIX_FMT_BGR4_BYTE|| \
630 (x)==PIX_FMT_MONOBLACK|| \
631 (x)==PIX_FMT_MONOWHITE \
634 #define isRGBinBytes(x) ( \
635 (x)==PIX_FMT_RGB48BE \
636 || (x)==PIX_FMT_RGB48LE \
637 || (x)==PIX_FMT_RGBA64BE \
638 || (x)==PIX_FMT_RGBA64LE \
639 || (x)==PIX_FMT_RGBA \
640 || (x)==PIX_FMT_ARGB \
641 || (x)==PIX_FMT_RGB24 \
643 #define isBGRinBytes(x) ( \
644 (x)==PIX_FMT_BGR48BE \
645 || (x)==PIX_FMT_BGR48LE \
646 || (x)==PIX_FMT_BGRA64BE \
647 || (x)==PIX_FMT_BGRA64LE \
648 || (x)==PIX_FMT_BGRA \
649 || (x)==PIX_FMT_ABGR \
650 || (x)==PIX_FMT_BGR24 \
653 #define isAnyRGB(x) \
657 (x)==PIX_FMT_GBR24P \
661 (av_pix_fmt_descriptors[x].nb_components == 2 || \
662 av_pix_fmt_descriptors[x].nb_components == 4)
665 #define isPacked(x) ( \
667 || (x)==PIX_FMT_YUYV422 \
668 || (x)==PIX_FMT_UYVY422 \
669 || (x)==PIX_FMT_Y400A \
674 #define isPacked(x) \
675 ((av_pix_fmt_descriptors[x].nb_components >= 2 && \
676 !(av_pix_fmt_descriptors[x].flags & PIX_FMT_PLANAR)) || \
680 #define isPlanar(x) \
681 (av_pix_fmt_descriptors[x].nb_components >= 2 && \
682 (av_pix_fmt_descriptors[x].flags & PIX_FMT_PLANAR))
684 #define isPackedRGB(x) \
685 ((av_pix_fmt_descriptors[x].flags & \
686 (PIX_FMT_PLANAR | PIX_FMT_RGB)) == PIX_FMT_RGB)
688 #define isPlanarRGB(x) \
689 ((av_pix_fmt_descriptors[x].flags & \
690 (PIX_FMT_PLANAR | PIX_FMT_RGB)) == (PIX_FMT_PLANAR | PIX_FMT_RGB))
692 #define usePal(x) ((av_pix_fmt_descriptors[x].flags & PIX_FMT_PAL) || \
693 (av_pix_fmt_descriptors[x].flags & PIX_FMT_PSEUDOPAL) || \
694 (x) == PIX_FMT_Y400A)
696 extern const uint64_t ff_dither4[2];
697 extern const uint64_t ff_dither8[2];
698 extern const uint8_t dithers[8][8][8];
699 extern const uint16_t dither_scale[15][16];
702 extern const AVClass sws_context_class;
705 * Set c->swScale to an unscaled converter if one exists for the specific
706 * source and destination formats, bit depths, flags, etc.
708 void ff_get_unscaled_swscale(SwsContext *c);
710 void ff_swscale_get_unscaled_altivec(SwsContext *c);
713 * Return function pointer to fastest main scaler path function depending
714 * on architecture and available optimizations.
716 SwsFunc ff_getSwsFunc(SwsContext *c);
718 void ff_sws_init_input_funcs(SwsContext *c);
719 void ff_sws_init_output_funcs(SwsContext *c,
720 yuv2planar1_fn *yuv2plane1,
721 yuv2planarX_fn *yuv2planeX,
722 yuv2interleavedX_fn *yuv2nv12cX,
723 yuv2packed1_fn *yuv2packed1,
724 yuv2packed2_fn *yuv2packed2,
725 yuv2packedX_fn *yuv2packedX);
726 void ff_sws_init_swScale_altivec(SwsContext *c);
727 void ff_sws_init_swScale_mmx(SwsContext *c);
729 #endif /* SWSCALE_SWSCALE_INTERNAL_H */