2 * Copyright 2011 The LibYuv Project Authors. All rights reserved.
4 * Use of this source code is governed by a BSD-style license
5 * that can be found in the LICENSE file in the root of the source
6 * tree. An additional intellectual property rights grant can be found
7 * in the file PATENTS. All contributing project authors may
8 * be found in the AUTHORS file in the root of the source tree.
11 #ifndef INCLUDE_LIBYUV_ROW_H_ // NOLINT
12 #define INCLUDE_LIBYUV_ROW_H_
14 #include <stdlib.h> // For malloc.
16 #include "libyuv/basic_types.h"
23 #define IS_ALIGNED(p, a) (!((uintptr_t)(p) & ((a) - 1)))
26 #define align_buffer_64(var, size) \
27 uint8* var##_mem = reinterpret_cast<uint8*>(malloc((size) + 63)); \
28 uint8* var = reinterpret_cast<uint8*> \
29 ((reinterpret_cast<intptr_t>(var##_mem) + 63) & ~63)
31 #define align_buffer_64(var, size) \
32 uint8* var##_mem = (uint8*)(malloc((size) + 63)); /* NOLINT */ \
33 uint8* var = (uint8*)(((intptr_t)(var##_mem) + 63) & ~63) /* NOLINT */
36 #define free_aligned_buffer_64(var) \
40 #if defined(__pnacl__) || defined(__CLR_VER) || defined(COVERAGE_ENABLED) || \
41 defined(TARGET_IPHONE_SIMULATOR)
42 #define LIBYUV_DISABLE_X86
44 // True if compiling for SSSE3 as a requirement.
45 #if defined(__SSSE3__) || (defined(_M_IX86_FP) && (_M_IX86_FP >= 3))
46 #define LIBYUV_SSSE3_ONLY
49 // Enable for NaCL pepper 33 for bundle and AVX2 support.
50 // #define NEW_BINUTILS
52 // The following are available on all x86 platforms:
53 #if !defined(LIBYUV_DISABLE_X86) && \
54 (defined(_M_IX86) || defined(__x86_64__) || defined(__i386__))
56 #define HAS_ARGBADDROW_SSE2
57 #define HAS_ARGBAFFINEROW_SSE2
58 #define HAS_ARGBATTENUATEROW_SSSE3
59 #define HAS_ARGBBLENDROW_SSSE3
60 #define HAS_ARGBCOLORMATRIXROW_SSSE3
61 #define HAS_ARGBCOLORTABLEROW_X86
62 #define HAS_ARGBCOPYALPHAROW_SSE2
63 #define HAS_ARGBCOPYYTOALPHAROW_SSE2
64 #define HAS_ARGBGRAYROW_SSSE3
65 #define HAS_ARGBLUMACOLORTABLEROW_SSSE3
66 #define HAS_ARGBMIRRORROW_SSSE3
67 #define HAS_ARGBMULTIPLYROW_SSE2
68 #define HAS_ARGBPOLYNOMIALROW_SSE2
69 #define HAS_ARGBQUANTIZEROW_SSE2
70 #define HAS_ARGBSEPIAROW_SSSE3
71 #define HAS_ARGBSHADEROW_SSE2
72 #define HAS_ARGBSUBTRACTROW_SSE2
73 #define HAS_ARGBTOUVROW_SSSE3
74 #define HAS_ARGBUNATTENUATEROW_SSE2
75 #define HAS_COMPUTECUMULATIVESUMROW_SSE2
76 #define HAS_CUMULATIVESUMTOAVERAGEROW_SSE2
77 #define HAS_INTERPOLATEROW_SSE2
78 #define HAS_INTERPOLATEROW_SSSE3
79 #define HAS_RGBCOLORTABLEROW_X86
80 #define HAS_SOBELROW_SSE2
81 #define HAS_SOBELTOPLANEROW_SSE2
82 #define HAS_SOBELXROW_SSE2
83 #define HAS_SOBELXYROW_SSE2
84 #define HAS_SOBELYROW_SSE2
87 #define HAS_ABGRTOUVROW_SSSE3
88 #define HAS_ABGRTOYROW_SSSE3
89 #define HAS_ARGB1555TOARGBROW_SSE2
90 #define HAS_ARGB4444TOARGBROW_SSE2
91 #define HAS_ARGBSHUFFLEROW_SSE2
92 #define HAS_ARGBSHUFFLEROW_SSSE3
93 #define HAS_ARGBTOARGB1555ROW_SSE2
94 #define HAS_ARGBTOARGB4444ROW_SSE2
95 #define HAS_ARGBTOBAYERGGROW_SSE2
96 #define HAS_ARGBTOBAYERROW_SSSE3
97 #define HAS_ARGBTORAWROW_SSSE3
98 #define HAS_ARGBTORGB24ROW_SSSE3
99 #define HAS_ARGBTORGB565ROW_SSE2
100 #define HAS_ARGBTOUV422ROW_SSSE3
101 #define HAS_ARGBTOUV444ROW_SSSE3
102 #define HAS_ARGBTOUVJROW_SSSE3
103 #define HAS_ARGBTOYJROW_SSSE3
104 #define HAS_ARGBTOYROW_SSSE3
105 #define HAS_BGRATOUVROW_SSSE3
106 #define HAS_BGRATOYROW_SSSE3
107 #define HAS_COPYROW_ERMS
108 #define HAS_COPYROW_SSE2
109 #define HAS_COPYROW_X86
110 #define HAS_HALFROW_SSE2
111 #define HAS_I400TOARGBROW_SSE2
112 #define HAS_I411TOARGBROW_SSSE3
113 #define HAS_I422TOARGB1555ROW_SSSE3
114 #define HAS_I422TOABGRROW_SSSE3
115 #define HAS_I422TOARGB1555ROW_SSSE3
116 #define HAS_I422TOARGB4444ROW_SSSE3
117 #define HAS_I422TOARGBROW_SSSE3
118 #define HAS_I422TOBGRAROW_SSSE3
119 #define HAS_I422TORAWROW_SSSE3
120 #define HAS_I422TORGB24ROW_SSSE3
121 #define HAS_I422TORGB565ROW_SSSE3
122 #define HAS_I422TORGBAROW_SSSE3
123 #define HAS_I422TOUYVYROW_SSE2
124 #define HAS_I422TOYUY2ROW_SSE2
125 #define HAS_I444TOARGBROW_SSSE3
126 #define HAS_MERGEUVROW_SSE2
127 #define HAS_MIRRORROW_SSE2
128 #define HAS_MIRRORROW_SSSE3
129 #define HAS_MIRRORROW_UV_SSSE3
130 #define HAS_MIRRORUVROW_SSSE3
131 #define HAS_NV12TOARGBROW_SSSE3
132 #define HAS_NV12TORGB565ROW_SSSE3
133 #define HAS_NV21TOARGBROW_SSSE3
134 #define HAS_NV21TORGB565ROW_SSSE3
135 #define HAS_RAWTOARGBROW_SSSE3
136 #define HAS_RAWTOYROW_SSSE3
137 #define HAS_RGB24TOARGBROW_SSSE3
138 #define HAS_RGB24TOYROW_SSSE3
139 #define HAS_RGB565TOARGBROW_SSE2
140 #define HAS_RGBATOUVROW_SSSE3
141 #define HAS_RGBATOYROW_SSSE3
142 #define HAS_SETROW_X86
143 #define HAS_SPLITUVROW_SSE2
144 #define HAS_UYVYTOARGBROW_SSSE3
145 #define HAS_UYVYTOUV422ROW_SSE2
146 #define HAS_UYVYTOUVROW_SSE2
147 #define HAS_UYVYTOYROW_SSE2
148 #define HAS_YTOARGBROW_SSE2
149 #define HAS_YUY2TOARGBROW_SSSE3
150 #define HAS_YUY2TOUV422ROW_SSE2
151 #define HAS_YUY2TOUVROW_SSE2
152 #define HAS_YUY2TOYROW_SSE2
155 // GCC >= 4.7.0 required for AVX2.
156 #if defined(__GNUC__) && (defined(__x86_64__) || defined(__i386__))
157 #if (__GNUC__ > 4) || (__GNUC__ == 4 && (__GNUC_MINOR__ >= 7))
158 #define GCC_HAS_AVX2 1
159 #endif // GNUC >= 4.7
162 // clang >= 3.4.0 required for AVX2.
163 #if defined(__clang__) && (defined(__x86_64__) || defined(__i386__))
164 #if (__clang_major__ > 3) || (__clang_major__ == 3 && (__clang_minor__ >= 4))
165 #define CLANG_HAS_AVX2 1
166 #endif // clang >= 3.4
169 // Visual C 2012 required for AVX2.
170 #if defined(_M_IX86) && defined(_MSC_VER) && _MSC_VER >= 1700
171 #define VISUALC_HAS_AVX2 1
172 #endif // VisualStudio >= 2012
174 // The following are available on all x86 platforms, but
175 // require VS2012, clang 3.4 or gcc 4.7.
176 // The code supports NaCL but requires a new compiler and validator.
177 #if !defined(LIBYUV_DISABLE_X86) && (defined(VISUALC_HAS_AVX2) || \
178 defined(CLANG_HAS_AVX2) || defined(GCC_HAS_AVX2))
180 #define HAS_ARGBPOLYNOMIALROW_AVX2
181 #define HAS_ARGBSHUFFLEROW_AVX2
182 #define HAS_ARGBCOPYALPHAROW_AVX2
183 #define HAS_ARGBCOPYYTOALPHAROW_AVX2
186 // The following are require VS2012.
187 // TODO(fbarchard): Port to gcc.
188 #if !defined(LIBYUV_DISABLE_X86) && defined(VISUALC_HAS_AVX2)
189 #define HAS_ARGBTOUVROW_AVX2
190 #define HAS_ARGBTOYJROW_AVX2
191 #define HAS_ARGBTOYROW_AVX2
192 #define HAS_HALFROW_AVX2
193 #define HAS_I422TOARGBROW_AVX2
194 #define HAS_INTERPOLATEROW_AVX2
195 #define HAS_MERGEUVROW_AVX2
196 #define HAS_MIRRORROW_AVX2
197 #define HAS_SPLITUVROW_AVX2
198 #define HAS_UYVYTOUV422ROW_AVX2
199 #define HAS_UYVYTOUVROW_AVX2
200 #define HAS_UYVYTOYROW_AVX2
201 #define HAS_YUY2TOUV422ROW_AVX2
202 #define HAS_YUY2TOUVROW_AVX2
203 #define HAS_YUY2TOYROW_AVX2
206 #define HAS_ARGBADDROW_AVX2
207 #define HAS_ARGBATTENUATEROW_AVX2
208 #define HAS_ARGBMIRRORROW_AVX2
209 #define HAS_ARGBMULTIPLYROW_AVX2
210 #define HAS_ARGBSUBTRACTROW_AVX2
211 #define HAS_ARGBUNATTENUATEROW_AVX2
212 #endif // defined(VISUALC_HAS_AVX2)
214 // The following are Yasm x86 only:
215 // TODO(fbarchard): Port AVX2 to inline.
216 #if !defined(LIBYUV_DISABLE_X86) && defined(HAVE_YASM)
217 (defined(_M_IX86) || defined(_M_X64) || \
218 defined(__x86_64__) || defined(__i386__))
219 #define HAS_MERGEUVROW_AVX2
220 #define HAS_MERGEUVROW_MMX
221 #define HAS_SPLITUVROW_AVX2
222 #define HAS_SPLITUVROW_MMX
223 #define HAS_UYVYTOYROW_AVX2
224 #define HAS_UYVYTOYROW_MMX
225 #define HAS_YUY2TOYROW_AVX2
226 #define HAS_YUY2TOYROW_MMX
229 // The following are disabled when SSSE3 is available:
230 #if !defined(LIBYUV_DISABLE_X86) && \
231 (defined(_M_IX86) || defined(__x86_64__) || defined(__i386__)) && \
232 !defined(LIBYUV_SSSE3_ONLY)
233 #define HAS_ARGBBLENDROW_SSE2
234 #define HAS_ARGBATTENUATEROW_SSE2
235 #define HAS_MIRRORROW_SSE2
238 // The following are available on Neon platforms:
239 #if !defined(LIBYUV_DISABLE_NEON) && \
240 (defined(__ARM_NEON__) || defined(LIBYUV_NEON)) && \
241 !defined(__native_client__)
242 #define HAS_ABGRTOUVROW_NEON
243 #define HAS_ABGRTOYROW_NEON
244 #define HAS_ARGB1555TOARGBROW_NEON
245 #define HAS_ARGB1555TOUVROW_NEON
246 #define HAS_ARGB1555TOYROW_NEON
247 #define HAS_ARGB4444TOARGBROW_NEON
248 #define HAS_ARGB4444TOUVROW_NEON
249 #define HAS_ARGB4444TOYROW_NEON
250 #define HAS_ARGBTOARGB1555ROW_NEON
251 #define HAS_ARGBTOARGB4444ROW_NEON
252 #define HAS_ARGBTOBAYERROW_NEON
253 #define HAS_ARGBTOBAYERGGROW_NEON
254 #define HAS_ARGBTORAWROW_NEON
255 #define HAS_ARGBTORGB24ROW_NEON
256 #define HAS_ARGBTORGB565ROW_NEON
257 #define HAS_ARGBTOUV411ROW_NEON
258 #define HAS_ARGBTOUV422ROW_NEON
259 #define HAS_ARGBTOUV444ROW_NEON
260 #define HAS_ARGBTOUVROW_NEON
261 #define HAS_ARGBTOUVJROW_NEON
262 #define HAS_ARGBTOYROW_NEON
263 #define HAS_ARGBTOYJROW_NEON
264 #define HAS_BGRATOUVROW_NEON
265 #define HAS_BGRATOYROW_NEON
266 #define HAS_COPYROW_NEON
267 #define HAS_HALFROW_NEON
268 #define HAS_I400TOARGBROW_NEON
269 #define HAS_I411TOARGBROW_NEON
270 #define HAS_I422TOABGRROW_NEON
271 #define HAS_I422TOARGB1555ROW_NEON
272 #define HAS_I422TOARGB4444ROW_NEON
273 #define HAS_I422TOARGBROW_NEON
274 #define HAS_I422TOBGRAROW_NEON
275 #define HAS_I422TORAWROW_NEON
276 #define HAS_I422TORGB24ROW_NEON
277 #define HAS_I422TORGB565ROW_NEON
278 #define HAS_I422TORGBAROW_NEON
279 #define HAS_I422TOUYVYROW_NEON
280 #define HAS_I422TOYUY2ROW_NEON
281 #define HAS_I444TOARGBROW_NEON
282 #define HAS_MERGEUVROW_NEON
283 #define HAS_MIRRORROW_NEON
284 #define HAS_MIRRORUVROW_NEON
285 #define HAS_NV12TOARGBROW_NEON
286 #define HAS_NV12TORGB565ROW_NEON
287 #define HAS_NV21TOARGBROW_NEON
288 #define HAS_NV21TORGB565ROW_NEON
289 #define HAS_RAWTOARGBROW_NEON
290 #define HAS_RAWTOUVROW_NEON
291 #define HAS_RAWTOYROW_NEON
292 #define HAS_RGB24TOARGBROW_NEON
293 #define HAS_RGB24TOUVROW_NEON
294 #define HAS_RGB24TOYROW_NEON
295 #define HAS_RGB565TOARGBROW_NEON
296 #define HAS_RGB565TOUVROW_NEON
297 #define HAS_RGB565TOYROW_NEON
298 #define HAS_RGBATOUVROW_NEON
299 #define HAS_RGBATOYROW_NEON
300 #define HAS_SETROW_NEON
301 #define HAS_SPLITUVROW_NEON
302 #define HAS_UYVYTOARGBROW_NEON
303 #define HAS_UYVYTOUV422ROW_NEON
304 #define HAS_UYVYTOUVROW_NEON
305 #define HAS_UYVYTOYROW_NEON
306 #define HAS_YTOARGBROW_NEON
307 #define HAS_YUY2TOARGBROW_NEON
308 #define HAS_YUY2TOUV422ROW_NEON
309 #define HAS_YUY2TOUVROW_NEON
310 #define HAS_YUY2TOYROW_NEON
313 #define HAS_ARGBADDROW_NEON
314 #define HAS_ARGBATTENUATEROW_NEON
315 #define HAS_ARGBBLENDROW_NEON
316 #define HAS_ARGBGRAYROW_NEON
317 #define HAS_ARGBMIRRORROW_NEON
318 #define HAS_ARGBMULTIPLYROW_NEON
319 #define HAS_ARGBQUANTIZEROW_NEON
320 #define HAS_ARGBSEPIAROW_NEON
321 #define HAS_ARGBSHADEROW_NEON
322 #define HAS_ARGBSUBTRACTROW_NEON
323 #define HAS_SOBELROW_NEON
324 #define HAS_SOBELTOPLANEROW_NEON
325 #define HAS_SOBELXYROW_NEON
326 #define HAS_SOBELXROW_NEON
327 #define HAS_SOBELYROW_NEON
328 #define HAS_INTERPOLATEROW_NEON
329 // TODO(fbarchard): Investigate neon unittest failure.
330 // #define HAS_ARGBCOLORMATRIXROW_NEON
333 // The following are available on Mips platforms:
334 #if !defined(LIBYUV_DISABLE_MIPS) && defined(__mips__)
335 #define HAS_COPYROW_MIPS
336 #if defined(__mips_dsp) && (__mips_dsp_rev >= 2)
337 #define HAS_I422TOABGRROW_MIPS_DSPR2
338 #define HAS_I422TOARGBROW_MIPS_DSPR2
339 #define HAS_I422TOBGRAROW_MIPS_DSPR2
340 #define HAS_INTERPOLATEROWS_MIPS_DSPR2
341 #define HAS_MIRRORROW_MIPS_DSPR2
342 #define HAS_MIRRORUVROW_MIPS_DSPR2
343 #define HAS_SPLITUVROW_MIPS_DSPR2
347 #if defined(_MSC_VER) && !defined(__CLR_VER)
348 #define SIMD_ALIGNED(var) __declspec(align(16)) var
349 typedef __declspec(align(16)) int16 vec16[8];
350 typedef __declspec(align(16)) int32 vec32[4];
351 typedef __declspec(align(16)) int8 vec8[16];
352 typedef __declspec(align(16)) uint16 uvec16[8];
353 typedef __declspec(align(16)) uint32 uvec32[4];
354 typedef __declspec(align(16)) uint8 uvec8[16];
355 typedef __declspec(align(32)) int16 lvec16[16];
356 typedef __declspec(align(32)) int32 lvec32[8];
357 typedef __declspec(align(32)) int8 lvec8[32];
358 typedef __declspec(align(32)) uint16 ulvec16[16];
359 typedef __declspec(align(32)) uint32 ulvec32[8];
360 typedef __declspec(align(32)) uint8 ulvec8[32];
362 #elif defined(__GNUC__)
363 // Caveat GCC 4.2 to 4.7 have a known issue using vectors with const.
364 #define SIMD_ALIGNED(var) var __attribute__((aligned(16)))
365 typedef int16 __attribute__((vector_size(16))) vec16;
366 typedef int32 __attribute__((vector_size(16))) vec32;
367 typedef int8 __attribute__((vector_size(16))) vec8;
368 typedef uint16 __attribute__((vector_size(16))) uvec16;
369 typedef uint32 __attribute__((vector_size(16))) uvec32;
370 typedef uint8 __attribute__((vector_size(16))) uvec8;
372 #define SIMD_ALIGNED(var) var
373 typedef int16 vec16[8];
374 typedef int32 vec32[4];
375 typedef int8 vec8[16];
376 typedef uint16 uvec16[8];
377 typedef uint32 uvec32[4];
378 typedef uint8 uvec8[16];
381 #if defined(__APPLE__) || defined(__x86_64__) || defined(__llvm__)
384 #define OMITFP __attribute__((optimize("omit-frame-pointer")))
387 // NaCL macros for GCC x86 and x64.
389 // TODO(nfullagar): When pepper_33 toolchain is distributed, default to
390 // NEW_BINUTILS and remove all BUNDLEALIGN occurances.
391 #if defined(__native_client__)
392 #define LABELALIGN ".p2align 5\n"
394 #define LABELALIGN ".p2align 2\n"
396 #if defined(__native_client__) && defined(__x86_64__)
397 #if defined(NEW_BINUTILS)
398 #define BUNDLELOCK ".bundle_lock\n"
399 #define BUNDLEUNLOCK ".bundle_unlock\n"
400 #define BUNDLEALIGN "\n"
402 #define BUNDLELOCK "\n"
403 #define BUNDLEUNLOCK "\n"
404 #define BUNDLEALIGN ".p2align 5\n"
406 #define MEMACCESS(base) "%%nacl:(%%r15,%q" #base ")"
407 #define MEMACCESS2(offset, base) "%%nacl:" #offset "(%%r15,%q" #base ")"
408 #define MEMLEA(offset, base) #offset "(%q" #base ")"
409 #define MEMLEA3(offset, index, scale) \
410 #offset "(,%q" #index "," #scale ")"
411 #define MEMLEA4(offset, base, index, scale) \
412 #offset "(%q" #base ",%q" #index "," #scale ")"
413 #define MEMMOVESTRING(s, d) "%%nacl:(%q" #s "),%%nacl:(%q" #d "), %%r15"
414 #define MEMSTORESTRING(reg, d) "%%" #reg ",%%nacl:(%q" #d "), %%r15"
415 #define MEMOPREG(opcode, offset, base, index, scale, reg) \
417 "lea " #offset "(%q" #base ",%q" #index "," #scale "),%%r14d\n" \
418 #opcode " (%%r15,%%r14),%%" #reg "\n" \
420 #define MEMOPMEM(opcode, reg, offset, base, index, scale) \
422 "lea " #offset "(%q" #base ",%q" #index "," #scale "),%%r14d\n" \
423 #opcode " %%" #reg ",(%%r15,%%r14)\n" \
425 #define MEMOPARG(opcode, offset, base, index, scale, arg) \
427 "lea " #offset "(%q" #base ",%q" #index "," #scale "),%%r14d\n" \
428 #opcode " (%%r15,%%r14),%" #arg "\n" \
431 #define BUNDLEALIGN "\n"
432 #define MEMACCESS(base) "(%" #base ")"
433 #define MEMACCESS2(offset, base) #offset "(%" #base ")"
434 #define MEMLEA(offset, base) #offset "(%" #base ")"
435 #define MEMLEA3(offset, index, scale) \
436 #offset "(,%" #index "," #scale ")"
437 #define MEMLEA4(offset, base, index, scale) \
438 #offset "(%" #base ",%" #index "," #scale ")"
439 #define MEMMOVESTRING(s, d)
440 #define MEMSTORESTRING(reg, d)
441 #define MEMOPREG(opcode, offset, base, index, scale, reg) \
442 #opcode " " #offset "(%" #base ",%" #index "," #scale "),%%" #reg "\n"
443 #define MEMOPMEM(opcode, reg, offset, base, index, scale) \
444 #opcode " %%" #reg ","#offset "(%" #base ",%" #index "," #scale ")\n"
445 #define MEMOPARG(opcode, offset, base, index, scale, arg) \
446 #opcode " " #offset "(%" #base ",%" #index "," #scale "),%" #arg "\n"
449 void I444ToARGBRow_NEON(const uint8* src_y,
454 void I422ToARGBRow_NEON(const uint8* src_y,
459 void I411ToARGBRow_NEON(const uint8* src_y,
464 void I422ToBGRARow_NEON(const uint8* src_y,
469 void I422ToABGRRow_NEON(const uint8* src_y,
474 void I422ToRGBARow_NEON(const uint8* src_y,
479 void I422ToRGB24Row_NEON(const uint8* src_y,
484 void I422ToRAWRow_NEON(const uint8* src_y,
489 void I422ToRGB565Row_NEON(const uint8* src_y,
494 void I422ToARGB1555Row_NEON(const uint8* src_y,
499 void I422ToARGB4444Row_NEON(const uint8* src_y,
504 void NV12ToARGBRow_NEON(const uint8* src_y,
508 void NV21ToARGBRow_NEON(const uint8* src_y,
512 void NV12ToRGB565Row_NEON(const uint8* src_y,
516 void NV21ToRGB565Row_NEON(const uint8* src_y,
520 void YUY2ToARGBRow_NEON(const uint8* src_yuy2,
523 void UYVYToARGBRow_NEON(const uint8* src_uyvy,
527 void ARGBToYRow_AVX2(const uint8* src_argb, uint8* dst_y, int pix);
528 void ARGBToYRow_Any_AVX2(const uint8* src_argb, uint8* dst_y, int pix);
529 void ARGBToYRow_SSSE3(const uint8* src_argb, uint8* dst_y, int pix);
530 void ARGBToYJRow_AVX2(const uint8* src_argb, uint8* dst_y, int pix);
531 void ARGBToYJRow_Any_AVX2(const uint8* src_argb, uint8* dst_y, int pix);
532 void ARGBToYJRow_SSSE3(const uint8* src_argb, uint8* dst_y, int pix);
533 void BGRAToYRow_SSSE3(const uint8* src_bgra, uint8* dst_y, int pix);
534 void ABGRToYRow_SSSE3(const uint8* src_abgr, uint8* dst_y, int pix);
535 void RGBAToYRow_SSSE3(const uint8* src_rgba, uint8* dst_y, int pix);
536 void RGB24ToYRow_SSSE3(const uint8* src_rgb24, uint8* dst_y, int pix);
537 void RAWToYRow_SSSE3(const uint8* src_raw, uint8* dst_y, int pix);
538 void ARGBToYRow_Unaligned_SSSE3(const uint8* src_argb, uint8* dst_y, int pix);
539 void ARGBToYJRow_Unaligned_SSSE3(const uint8* src_argb, uint8* dst_y, int pix);
540 void BGRAToYRow_Unaligned_SSSE3(const uint8* src_bgra, uint8* dst_y, int pix);
541 void ABGRToYRow_Unaligned_SSSE3(const uint8* src_abgr, uint8* dst_y, int pix);
542 void RGBAToYRow_Unaligned_SSSE3(const uint8* src_rgba, uint8* dst_y, int pix);
543 void RGB24ToYRow_Unaligned_SSSE3(const uint8* src_rgb24, uint8* dst_y, int pix);
544 void RAWToYRow_Unaligned_SSSE3(const uint8* src_raw, uint8* dst_y, int pix);
545 void ARGBToYRow_NEON(const uint8* src_argb, uint8* dst_y, int pix);
546 void ARGBToYJRow_NEON(const uint8* src_argb, uint8* dst_y, int pix);
547 void ARGBToUV444Row_NEON(const uint8* src_argb, uint8* dst_u, uint8* dst_v,
549 void ARGBToUV422Row_NEON(const uint8* src_argb, uint8* dst_u, uint8* dst_v,
551 void ARGBToUV411Row_NEON(const uint8* src_argb, uint8* dst_u, uint8* dst_v,
553 void ARGBToUVRow_NEON(const uint8* src_argb, int src_stride_argb,
554 uint8* dst_u, uint8* dst_v, int pix);
555 void ARGBToUVJRow_NEON(const uint8* src_argb, int src_stride_argb,
556 uint8* dst_u, uint8* dst_v, int pix);
557 void BGRAToUVRow_NEON(const uint8* src_bgra, int src_stride_bgra,
558 uint8* dst_u, uint8* dst_v, int pix);
559 void ABGRToUVRow_NEON(const uint8* src_abgr, int src_stride_abgr,
560 uint8* dst_u, uint8* dst_v, int pix);
561 void RGBAToUVRow_NEON(const uint8* src_rgba, int src_stride_rgba,
562 uint8* dst_u, uint8* dst_v, int pix);
563 void RGB24ToUVRow_NEON(const uint8* src_rgb24, int src_stride_rgb24,
564 uint8* dst_u, uint8* dst_v, int pix);
565 void RAWToUVRow_NEON(const uint8* src_raw, int src_stride_raw,
566 uint8* dst_u, uint8* dst_v, int pix);
567 void RGB565ToUVRow_NEON(const uint8* src_rgb565, int src_stride_rgb565,
568 uint8* dst_u, uint8* dst_v, int pix);
569 void ARGB1555ToUVRow_NEON(const uint8* src_argb1555, int src_stride_argb1555,
570 uint8* dst_u, uint8* dst_v, int pix);
571 void ARGB4444ToUVRow_NEON(const uint8* src_argb4444, int src_stride_argb4444,
572 uint8* dst_u, uint8* dst_v, int pix);
573 void BGRAToYRow_NEON(const uint8* src_bgra, uint8* dst_y, int pix);
574 void ABGRToYRow_NEON(const uint8* src_abgr, uint8* dst_y, int pix);
575 void RGBAToYRow_NEON(const uint8* src_rgba, uint8* dst_y, int pix);
576 void RGB24ToYRow_NEON(const uint8* src_rgb24, uint8* dst_y, int pix);
577 void RAWToYRow_NEON(const uint8* src_raw, uint8* dst_y, int pix);
578 void RGB565ToYRow_NEON(const uint8* src_rgb565, uint8* dst_y, int pix);
579 void ARGB1555ToYRow_NEON(const uint8* src_argb1555, uint8* dst_y, int pix);
580 void ARGB4444ToYRow_NEON(const uint8* src_argb4444, uint8* dst_y, int pix);
581 void ARGBToYRow_C(const uint8* src_argb, uint8* dst_y, int pix);
582 void ARGBToYJRow_C(const uint8* src_argb, uint8* dst_y, int pix);
583 void BGRAToYRow_C(const uint8* src_bgra, uint8* dst_y, int pix);
584 void ABGRToYRow_C(const uint8* src_abgr, uint8* dst_y, int pix);
585 void RGBAToYRow_C(const uint8* src_rgba, uint8* dst_y, int pix);
586 void RGB24ToYRow_C(const uint8* src_rgb24, uint8* dst_y, int pix);
587 void RAWToYRow_C(const uint8* src_raw, uint8* dst_y, int pix);
588 void RGB565ToYRow_C(const uint8* src_rgb565, uint8* dst_y, int pix);
589 void ARGB1555ToYRow_C(const uint8* src_argb1555, uint8* dst_y, int pix);
590 void ARGB4444ToYRow_C(const uint8* src_argb4444, uint8* dst_y, int pix);
591 void ARGBToYRow_Any_SSSE3(const uint8* src_argb, uint8* dst_y, int pix);
592 void ARGBToYJRow_Any_SSSE3(const uint8* src_argb, uint8* dst_y, int pix);
593 void BGRAToYRow_Any_SSSE3(const uint8* src_bgra, uint8* dst_y, int pix);
594 void ABGRToYRow_Any_SSSE3(const uint8* src_abgr, uint8* dst_y, int pix);
595 void RGBAToYRow_Any_SSSE3(const uint8* src_rgba, uint8* dst_y, int pix);
596 void RGB24ToYRow_Any_SSSE3(const uint8* src_rgb24, uint8* dst_y, int pix);
597 void RAWToYRow_Any_SSSE3(const uint8* src_raw, uint8* dst_y, int pix);
598 void ARGBToYRow_Any_NEON(const uint8* src_argb, uint8* dst_y, int pix);
599 void ARGBToYJRow_Any_NEON(const uint8* src_argb, uint8* dst_y, int pix);
600 void BGRAToYRow_Any_NEON(const uint8* src_bgra, uint8* dst_y, int pix);
601 void ABGRToYRow_Any_NEON(const uint8* src_abgr, uint8* dst_y, int pix);
602 void RGBAToYRow_Any_NEON(const uint8* src_rgba, uint8* dst_y, int pix);
603 void RGB24ToYRow_Any_NEON(const uint8* src_rgb24, uint8* dst_y, int pix);
604 void RAWToYRow_Any_NEON(const uint8* src_raw, uint8* dst_y, int pix);
605 void RGB565ToYRow_Any_NEON(const uint8* src_rgb565, uint8* dst_y, int pix);
606 void ARGB1555ToYRow_Any_NEON(const uint8* src_argb1555, uint8* dst_y, int pix);
607 void ARGB4444ToYRow_Any_NEON(const uint8* src_argb4444, uint8* dst_y, int pix);
609 void ARGBToUVRow_AVX2(const uint8* src_argb, int src_stride_argb,
610 uint8* dst_u, uint8* dst_v, int width);
611 void ARGBToUVRow_Any_AVX2(const uint8* src_argb, int src_stride_argb,
612 uint8* dst_u, uint8* dst_v, int width);
613 void ARGBToUVRow_SSSE3(const uint8* src_argb, int src_stride_argb,
614 uint8* dst_u, uint8* dst_v, int width);
615 void ARGBToUVJRow_SSSE3(const uint8* src_argb, int src_stride_argb,
616 uint8* dst_u, uint8* dst_v, int width);
617 void BGRAToUVRow_SSSE3(const uint8* src_bgra, int src_stride_bgra,
618 uint8* dst_u, uint8* dst_v, int width);
619 void ABGRToUVRow_SSSE3(const uint8* src_abgr, int src_stride_abgr,
620 uint8* dst_u, uint8* dst_v, int width);
621 void RGBAToUVRow_SSSE3(const uint8* src_rgba, int src_stride_rgba,
622 uint8* dst_u, uint8* dst_v, int width);
623 void ARGBToUVRow_Unaligned_SSSE3(const uint8* src_argb, int src_stride_argb,
624 uint8* dst_u, uint8* dst_v, int width);
625 void ARGBToUVJRow_Unaligned_SSSE3(const uint8* src_argb, int src_stride_argb,
626 uint8* dst_u, uint8* dst_v, int width);
627 void BGRAToUVRow_Unaligned_SSSE3(const uint8* src_bgra, int src_stride_bgra,
628 uint8* dst_u, uint8* dst_v, int width);
629 void ABGRToUVRow_Unaligned_SSSE3(const uint8* src_abgr, int src_stride_abgr,
630 uint8* dst_u, uint8* dst_v, int width);
631 void RGBAToUVRow_Unaligned_SSSE3(const uint8* src_rgba, int src_stride_rgba,
632 uint8* dst_u, uint8* dst_v, int width);
633 void ARGBToUVRow_Any_SSSE3(const uint8* src_argb, int src_stride_argb,
634 uint8* dst_u, uint8* dst_v, int width);
635 void ARGBToUVJRow_Any_SSSE3(const uint8* src_argb, int src_stride_argb,
636 uint8* dst_u, uint8* dst_v, int width);
637 void BGRAToUVRow_Any_SSSE3(const uint8* src_bgra, int src_stride_bgra,
638 uint8* dst_u, uint8* dst_v, int width);
639 void ABGRToUVRow_Any_SSSE3(const uint8* src_abgr, int src_stride_abgr,
640 uint8* dst_u, uint8* dst_v, int width);
641 void RGBAToUVRow_Any_SSSE3(const uint8* src_rgba, int src_stride_rgba,
642 uint8* dst_u, uint8* dst_v, int width);
643 void ARGBToUV444Row_Any_NEON(const uint8* src_argb, uint8* dst_u, uint8* dst_v,
645 void ARGBToUV422Row_Any_NEON(const uint8* src_argb, uint8* dst_u, uint8* dst_v,
647 void ARGBToUV411Row_Any_NEON(const uint8* src_argb, uint8* dst_u, uint8* dst_v,
649 void ARGBToUVRow_Any_NEON(const uint8* src_argb, int src_stride_argb,
650 uint8* dst_u, uint8* dst_v, int pix);
651 void ARGBToUVJRow_Any_NEON(const uint8* src_argb, int src_stride_argb,
652 uint8* dst_u, uint8* dst_v, int pix);
653 void BGRAToUVRow_Any_NEON(const uint8* src_bgra, int src_stride_bgra,
654 uint8* dst_u, uint8* dst_v, int pix);
655 void ABGRToUVRow_Any_NEON(const uint8* src_abgr, int src_stride_abgr,
656 uint8* dst_u, uint8* dst_v, int pix);
657 void RGBAToUVRow_Any_NEON(const uint8* src_rgba, int src_stride_rgba,
658 uint8* dst_u, uint8* dst_v, int pix);
659 void RGB24ToUVRow_Any_NEON(const uint8* src_rgb24, int src_stride_rgb24,
660 uint8* dst_u, uint8* dst_v, int pix);
661 void RAWToUVRow_Any_NEON(const uint8* src_raw, int src_stride_raw,
662 uint8* dst_u, uint8* dst_v, int pix);
663 void RGB565ToUVRow_Any_NEON(const uint8* src_rgb565, int src_stride_rgb565,
664 uint8* dst_u, uint8* dst_v, int pix);
665 void ARGB1555ToUVRow_Any_NEON(const uint8* src_argb1555,
666 int src_stride_argb1555,
667 uint8* dst_u, uint8* dst_v, int pix);
668 void ARGB4444ToUVRow_Any_NEON(const uint8* src_argb4444,
669 int src_stride_argb4444,
670 uint8* dst_u, uint8* dst_v, int pix);
671 void ARGBToUVRow_C(const uint8* src_argb, int src_stride_argb,
672 uint8* dst_u, uint8* dst_v, int width);
673 void ARGBToUVJRow_C(const uint8* src_argb, int src_stride_argb,
674 uint8* dst_u, uint8* dst_v, int width);
675 void BGRAToUVRow_C(const uint8* src_bgra, int src_stride_bgra,
676 uint8* dst_u, uint8* dst_v, int width);
677 void ABGRToUVRow_C(const uint8* src_abgr, int src_stride_abgr,
678 uint8* dst_u, uint8* dst_v, int width);
679 void RGBAToUVRow_C(const uint8* src_rgba, int src_stride_rgba,
680 uint8* dst_u, uint8* dst_v, int width);
681 void RGB24ToUVRow_C(const uint8* src_rgb24, int src_stride_rgb24,
682 uint8* dst_u, uint8* dst_v, int width);
683 void RAWToUVRow_C(const uint8* src_raw, int src_stride_raw,
684 uint8* dst_u, uint8* dst_v, int width);
685 void RGB565ToUVRow_C(const uint8* src_rgb565, int src_stride_rgb565,
686 uint8* dst_u, uint8* dst_v, int width);
687 void ARGB1555ToUVRow_C(const uint8* src_argb1555, int src_stride_argb1555,
688 uint8* dst_u, uint8* dst_v, int width);
689 void ARGB4444ToUVRow_C(const uint8* src_argb4444, int src_stride_argb4444,
690 uint8* dst_u, uint8* dst_v, int width);
692 void ARGBToUV444Row_SSSE3(const uint8* src_argb,
693 uint8* dst_u, uint8* dst_v, int width);
694 void ARGBToUV444Row_Unaligned_SSSE3(const uint8* src_argb,
695 uint8* dst_u, uint8* dst_v, int width);
696 void ARGBToUV444Row_Any_SSSE3(const uint8* src_argb,
697 uint8* dst_u, uint8* dst_v, int width);
699 void ARGBToUV422Row_SSSE3(const uint8* src_argb,
700 uint8* dst_u, uint8* dst_v, int width);
701 void ARGBToUV422Row_Unaligned_SSSE3(const uint8* src_argb,
702 uint8* dst_u, uint8* dst_v, int width);
703 void ARGBToUV422Row_Any_SSSE3(const uint8* src_argb,
704 uint8* dst_u, uint8* dst_v, int width);
706 void ARGBToUV444Row_C(const uint8* src_argb,
707 uint8* dst_u, uint8* dst_v, int width);
708 void ARGBToUV422Row_C(const uint8* src_argb,
709 uint8* dst_u, uint8* dst_v, int width);
710 void ARGBToUV411Row_C(const uint8* src_argb,
711 uint8* dst_u, uint8* dst_v, int width);
713 void MirrorRow_AVX2(const uint8* src, uint8* dst, int width);
714 void MirrorRow_SSSE3(const uint8* src, uint8* dst, int width);
715 void MirrorRow_SSE2(const uint8* src, uint8* dst, int width);
716 void MirrorRow_NEON(const uint8* src, uint8* dst, int width);
717 void MirrorRow_MIPS_DSPR2(const uint8* src, uint8* dst, int width);
718 void MirrorRow_C(const uint8* src, uint8* dst, int width);
720 void MirrorUVRow_SSSE3(const uint8* src_uv, uint8* dst_u, uint8* dst_v,
722 void MirrorUVRow_NEON(const uint8* src_uv, uint8* dst_u, uint8* dst_v,
724 void MirrorUVRow_MIPS_DSPR2(const uint8* src_uv, uint8* dst_u, uint8* dst_v,
726 void MirrorUVRow_C(const uint8* src_uv, uint8* dst_u, uint8* dst_v,
729 void ARGBMirrorRow_AVX2(const uint8* src, uint8* dst, int width);
730 void ARGBMirrorRow_SSSE3(const uint8* src, uint8* dst, int width);
731 void ARGBMirrorRow_NEON(const uint8* src, uint8* dst, int width);
732 void ARGBMirrorRow_C(const uint8* src, uint8* dst, int width);
734 void SplitUVRow_C(const uint8* src_uv, uint8* dst_u, uint8* dst_v, int pix);
735 void SplitUVRow_SSE2(const uint8* src_uv, uint8* dst_u, uint8* dst_v, int pix);
736 void SplitUVRow_AVX2(const uint8* src_uv, uint8* dst_u, uint8* dst_v, int pix);
737 void SplitUVRow_NEON(const uint8* src_uv, uint8* dst_u, uint8* dst_v, int pix);
738 void SplitUVRow_MIPS_DSPR2(const uint8* src_uv, uint8* dst_u, uint8* dst_v,
740 void SplitUVRow_Unaligned_SSE2(const uint8* src_uv, uint8* dst_u, uint8* dst_v,
742 void SplitUVRow_Unaligned_MIPS_DSPR2(const uint8* src_uv, uint8* dst_u,
743 uint8* dst_v, int pix);
744 void SplitUVRow_Any_SSE2(const uint8* src_uv, uint8* dst_u, uint8* dst_v,
746 void SplitUVRow_Any_AVX2(const uint8* src_uv, uint8* dst_u, uint8* dst_v,
748 void SplitUVRow_Any_NEON(const uint8* src_uv, uint8* dst_u, uint8* dst_v,
750 void SplitUVRow_Any_MIPS_DSPR2(const uint8* src_uv, uint8* dst_u, uint8* dst_v,
753 void MergeUVRow_C(const uint8* src_u, const uint8* src_v, uint8* dst_uv,
755 void MergeUVRow_SSE2(const uint8* src_u, const uint8* src_v, uint8* dst_uv,
757 void MergeUVRow_AVX2(const uint8* src_u, const uint8* src_v, uint8* dst_uv,
759 void MergeUVRow_NEON(const uint8* src_u, const uint8* src_v, uint8* dst_uv,
761 void MergeUVRow_Unaligned_SSE2(const uint8* src_u, const uint8* src_v,
762 uint8* dst_uv, int width);
763 void MergeUVRow_Any_SSE2(const uint8* src_u, const uint8* src_v, uint8* dst_uv,
765 void MergeUVRow_Any_AVX2(const uint8* src_u, const uint8* src_v, uint8* dst_uv,
767 void MergeUVRow_Any_NEON(const uint8* src_u, const uint8* src_v, uint8* dst_uv,
770 void CopyRow_SSE2(const uint8* src, uint8* dst, int count);
771 void CopyRow_ERMS(const uint8* src, uint8* dst, int count);
772 void CopyRow_X86(const uint8* src, uint8* dst, int count);
773 void CopyRow_NEON(const uint8* src, uint8* dst, int count);
774 void CopyRow_MIPS(const uint8* src, uint8* dst, int count);
775 void CopyRow_C(const uint8* src, uint8* dst, int count);
777 void CopyRow_16_C(const uint16* src, uint16* dst, int count);
779 void ARGBCopyAlphaRow_C(const uint8* src_argb, uint8* dst_argb, int width);
780 void ARGBCopyAlphaRow_SSE2(const uint8* src_argb, uint8* dst_argb, int width);
781 void ARGBCopyAlphaRow_AVX2(const uint8* src_argb, uint8* dst_argb, int width);
783 void ARGBCopyYToAlphaRow_C(const uint8* src_y, uint8* dst_argb, int width);
784 void ARGBCopyYToAlphaRow_SSE2(const uint8* src_y, uint8* dst_argb, int width);
785 void ARGBCopyYToAlphaRow_AVX2(const uint8* src_y, uint8* dst_argb, int width);
787 void SetRow_X86(uint8* dst, uint32 v32, int count);
788 void ARGBSetRows_X86(uint8* dst, uint32 v32, int width,
789 int dst_stride, int height);
790 void SetRow_NEON(uint8* dst, uint32 v32, int count);
791 void ARGBSetRows_NEON(uint8* dst, uint32 v32, int width,
792 int dst_stride, int height);
793 void SetRow_C(uint8* dst, uint32 v32, int count);
794 void ARGBSetRows_C(uint8* dst, uint32 v32, int width, int dst_stride,
797 // ARGBShufflers for BGRAToARGB etc.
798 void ARGBShuffleRow_C(const uint8* src_argb, uint8* dst_argb,
799 const uint8* shuffler, int pix);
800 void ARGBShuffleRow_SSE2(const uint8* src_argb, uint8* dst_argb,
801 const uint8* shuffler, int pix);
802 void ARGBShuffleRow_SSSE3(const uint8* src_argb, uint8* dst_argb,
803 const uint8* shuffler, int pix);
804 void ARGBShuffleRow_AVX2(const uint8* src_argb, uint8* dst_argb,
805 const uint8* shuffler, int pix);
806 void ARGBShuffleRow_NEON(const uint8* src_argb, uint8* dst_argb,
807 const uint8* shuffler, int pix);
808 void ARGBShuffleRow_Unaligned_SSSE3(const uint8* src_argb, uint8* dst_argb,
809 const uint8* shuffler, int pix);
810 void ARGBShuffleRow_Any_SSE2(const uint8* src_argb, uint8* dst_argb,
811 const uint8* shuffler, int pix);
812 void ARGBShuffleRow_Any_SSSE3(const uint8* src_argb, uint8* dst_argb,
813 const uint8* shuffler, int pix);
814 void ARGBShuffleRow_Any_AVX2(const uint8* src_argb, uint8* dst_argb,
815 const uint8* shuffler, int pix);
816 void ARGBShuffleRow_Any_NEON(const uint8* src_argb, uint8* dst_argb,
817 const uint8* shuffler, int pix);
819 void RGB24ToARGBRow_SSSE3(const uint8* src_rgb24, uint8* dst_argb, int pix);
820 void RAWToARGBRow_SSSE3(const uint8* src_raw, uint8* dst_argb, int pix);
821 void RGB565ToARGBRow_SSE2(const uint8* src_rgb565, uint8* dst_argb, int pix);
822 void ARGB1555ToARGBRow_SSE2(const uint8* src_argb1555, uint8* dst_argb,
824 void ARGB4444ToARGBRow_SSE2(const uint8* src_argb4444, uint8* dst_argb,
827 void RGB24ToARGBRow_NEON(const uint8* src_rgb24, uint8* dst_argb, int pix);
828 void RAWToARGBRow_NEON(const uint8* src_raw, uint8* dst_argb, int pix);
829 void RGB565ToARGBRow_NEON(const uint8* src_rgb565, uint8* dst_argb, int pix);
830 void ARGB1555ToARGBRow_NEON(const uint8* src_argb1555, uint8* dst_argb,
832 void ARGB4444ToARGBRow_NEON(const uint8* src_argb4444, uint8* dst_argb,
834 void RGB24ToARGBRow_C(const uint8* src_rgb24, uint8* dst_argb, int pix);
835 void RAWToARGBRow_C(const uint8* src_raw, uint8* dst_argb, int pix);
836 void RGB565ToARGBRow_C(const uint8* src_rgb, uint8* dst_argb, int pix);
837 void ARGB1555ToARGBRow_C(const uint8* src_argb, uint8* dst_argb, int pix);
838 void ARGB4444ToARGBRow_C(const uint8* src_argb, uint8* dst_argb, int pix);
839 void RGB24ToARGBRow_Any_SSSE3(const uint8* src_rgb24, uint8* dst_argb, int pix);
840 void RAWToARGBRow_Any_SSSE3(const uint8* src_raw, uint8* dst_argb, int pix);
841 void RGB565ToARGBRow_Any_SSE2(const uint8* src_rgb565, uint8* dst_argb,
843 void ARGB1555ToARGBRow_Any_SSE2(const uint8* src_argb1555, uint8* dst_argb,
845 void ARGB4444ToARGBRow_Any_SSE2(const uint8* src_argb4444, uint8* dst_argb,
847 void RGB24ToARGBRow_Any_NEON(const uint8* src_rgb24, uint8* dst_argb, int pix);
848 void RAWToARGBRow_Any_NEON(const uint8* src_raw, uint8* dst_argb, int pix);
849 void RGB565ToARGBRow_Any_NEON(const uint8* src_rgb565, uint8* dst_argb,
851 void ARGB1555ToARGBRow_Any_NEON(const uint8* src_argb1555, uint8* dst_argb,
853 void ARGB4444ToARGBRow_Any_NEON(const uint8* src_argb4444, uint8* dst_argb,
856 void ARGBToRGB24Row_SSSE3(const uint8* src_argb, uint8* dst_rgb, int pix);
857 void ARGBToRAWRow_SSSE3(const uint8* src_argb, uint8* dst_rgb, int pix);
858 void ARGBToRGB565Row_SSE2(const uint8* src_argb, uint8* dst_rgb, int pix);
859 void ARGBToARGB1555Row_SSE2(const uint8* src_argb, uint8* dst_rgb, int pix);
860 void ARGBToARGB4444Row_SSE2(const uint8* src_argb, uint8* dst_rgb, int pix);
862 void ARGBToRGB24Row_NEON(const uint8* src_argb, uint8* dst_rgb, int pix);
863 void ARGBToRAWRow_NEON(const uint8* src_argb, uint8* dst_rgb, int pix);
864 void ARGBToRGB565Row_NEON(const uint8* src_argb, uint8* dst_rgb, int pix);
865 void ARGBToARGB1555Row_NEON(const uint8* src_argb, uint8* dst_rgb, int pix);
866 void ARGBToARGB4444Row_NEON(const uint8* src_argb, uint8* dst_rgb, int pix);
868 void ARGBToRGBARow_C(const uint8* src_argb, uint8* dst_rgb, int pix);
869 void ARGBToRGB24Row_C(const uint8* src_argb, uint8* dst_rgb, int pix);
870 void ARGBToRAWRow_C(const uint8* src_argb, uint8* dst_rgb, int pix);
871 void ARGBToRGB565Row_C(const uint8* src_argb, uint8* dst_rgb, int pix);
872 void ARGBToARGB1555Row_C(const uint8* src_argb, uint8* dst_rgb, int pix);
873 void ARGBToARGB4444Row_C(const uint8* src_argb, uint8* dst_rgb, int pix);
875 void I400ToARGBRow_SSE2(const uint8* src_y, uint8* dst_argb, int pix);
876 void I400ToARGBRow_Unaligned_SSE2(const uint8* src_y, uint8* dst_argb, int pix);
877 void I400ToARGBRow_NEON(const uint8* src_y, uint8* dst_argb, int pix);
878 void I400ToARGBRow_C(const uint8* src_y, uint8* dst_argb, int pix);
879 void I400ToARGBRow_Any_SSE2(const uint8* src_y, uint8* dst_argb, int pix);
880 void I400ToARGBRow_Any_NEON(const uint8* src_y, uint8* dst_argb, int pix);
882 void I444ToARGBRow_C(const uint8* src_y,
887 void I422ToARGBRow_C(const uint8* src_y,
892 void I411ToARGBRow_C(const uint8* src_y,
897 void NV12ToARGBRow_C(const uint8* src_y,
901 void NV21ToRGB565Row_C(const uint8* src_y,
905 void NV12ToRGB565Row_C(const uint8* src_y,
909 void NV21ToARGBRow_C(const uint8* src_y,
913 void YUY2ToARGBRow_C(const uint8* src_yuy2,
916 void UYVYToARGBRow_C(const uint8* src_uyvy,
919 void I422ToBGRARow_C(const uint8* src_y,
924 void I422ToABGRRow_C(const uint8* src_y,
929 void I422ToRGBARow_C(const uint8* src_y,
934 void I422ToRGB24Row_C(const uint8* src_y,
939 void I422ToRAWRow_C(const uint8* src_y,
944 void I422ToARGB4444Row_C(const uint8* src_y,
949 void I422ToARGB1555Row_C(const uint8* src_y,
954 void I422ToRGB565Row_C(const uint8* src_y,
959 void YToARGBRow_C(const uint8* src_y,
962 void I422ToARGBRow_AVX2(const uint8* src_y,
967 void I444ToARGBRow_SSSE3(const uint8* src_y,
972 void I422ToARGBRow_SSSE3(const uint8* src_y,
977 void I411ToARGBRow_SSSE3(const uint8* src_y,
982 void NV12ToARGBRow_SSSE3(const uint8* src_y,
986 void NV21ToARGBRow_SSSE3(const uint8* src_y,
990 void NV12ToRGB565Row_SSSE3(const uint8* src_y,
994 void NV21ToRGB565Row_SSSE3(const uint8* src_y,
998 void YUY2ToARGBRow_SSSE3(const uint8* src_yuy2,
1001 void UYVYToARGBRow_SSSE3(const uint8* src_uyvy,
1004 void I422ToBGRARow_SSSE3(const uint8* src_y,
1009 void I422ToABGRRow_SSSE3(const uint8* src_y,
1014 void I422ToRGBARow_SSSE3(const uint8* src_y,
1019 void I422ToARGB4444Row_SSSE3(const uint8* src_y,
1024 void I422ToARGB1555Row_SSSE3(const uint8* src_y,
1029 void I422ToRGB565Row_SSSE3(const uint8* src_y,
1034 // RGB24/RAW are unaligned.
1035 void I422ToRGB24Row_SSSE3(const uint8* src_y,
1040 void I422ToRAWRow_SSSE3(const uint8* src_y,
1046 void I444ToARGBRow_Unaligned_SSSE3(const uint8* src_y,
1051 void I422ToARGBRow_Unaligned_SSSE3(const uint8* src_y,
1056 void I411ToARGBRow_Unaligned_SSSE3(const uint8* src_y,
1061 void NV12ToARGBRow_Unaligned_SSSE3(const uint8* src_y,
1062 const uint8* src_uv,
1065 void NV21ToARGBRow_Unaligned_SSSE3(const uint8* src_y,
1066 const uint8* src_vu,
1069 void YUY2ToARGBRow_Unaligned_SSSE3(const uint8* src_yuy2,
1072 void UYVYToARGBRow_Unaligned_SSSE3(const uint8* src_uyvy,
1075 void I422ToBGRARow_Unaligned_SSSE3(const uint8* src_y,
1080 void I422ToABGRRow_Unaligned_SSSE3(const uint8* src_y,
1085 void I422ToRGBARow_Unaligned_SSSE3(const uint8* src_y,
1090 void I422ToARGBRow_Any_AVX2(const uint8* src_y,
1095 void I444ToARGBRow_Any_SSSE3(const uint8* src_y,
1100 void I422ToARGBRow_Any_SSSE3(const uint8* src_y,
1105 void I411ToARGBRow_Any_SSSE3(const uint8* src_y,
1110 void NV12ToARGBRow_Any_SSSE3(const uint8* src_y,
1111 const uint8* src_uv,
1114 void NV21ToARGBRow_Any_SSSE3(const uint8* src_y,
1115 const uint8* src_vu,
1118 void NV12ToRGB565Row_Any_SSSE3(const uint8* src_y,
1119 const uint8* src_uv,
1122 void NV21ToRGB565Row_Any_SSSE3(const uint8* src_y,
1123 const uint8* src_vu,
1126 void YUY2ToARGBRow_Any_SSSE3(const uint8* src_yuy2,
1129 void UYVYToARGBRow_Any_SSSE3(const uint8* src_uyvy,
1132 void I422ToBGRARow_Any_SSSE3(const uint8* src_y,
1137 void I422ToABGRRow_Any_SSSE3(const uint8* src_y,
1142 void I422ToRGBARow_Any_SSSE3(const uint8* src_y,
1147 void I422ToARGB4444Row_Any_SSSE3(const uint8* src_y,
1152 void I422ToARGB1555Row_Any_SSSE3(const uint8* src_y,
1157 void I422ToRGB565Row_Any_SSSE3(const uint8* src_y,
1162 // RGB24/RAW are unaligned.
1163 void I422ToRGB24Row_Any_SSSE3(const uint8* src_y,
1168 void I422ToRAWRow_Any_SSSE3(const uint8* src_y,
1173 void YToARGBRow_SSE2(const uint8* src_y,
1176 void YToARGBRow_NEON(const uint8* src_y,
1179 void YToARGBRow_Any_SSE2(const uint8* src_y,
1182 void YToARGBRow_Any_NEON(const uint8* src_y,
1186 // ARGB preattenuated alpha blend.
1187 void ARGBBlendRow_SSSE3(const uint8* src_argb, const uint8* src_argb1,
1188 uint8* dst_argb, int width);
1189 void ARGBBlendRow_SSE2(const uint8* src_argb, const uint8* src_argb1,
1190 uint8* dst_argb, int width);
1191 void ARGBBlendRow_NEON(const uint8* src_argb, const uint8* src_argb1,
1192 uint8* dst_argb, int width);
1193 void ARGBBlendRow_C(const uint8* src_argb, const uint8* src_argb1,
1194 uint8* dst_argb, int width);
1196 // ARGB multiply images. Same API as Blend, but these require
1197 // pointer and width alignment for SSE2.
1198 void ARGBMultiplyRow_C(const uint8* src_argb, const uint8* src_argb1,
1199 uint8* dst_argb, int width);
1200 void ARGBMultiplyRow_SSE2(const uint8* src_argb, const uint8* src_argb1,
1201 uint8* dst_argb, int width);
1202 void ARGBMultiplyRow_Any_SSE2(const uint8* src_argb, const uint8* src_argb1,
1203 uint8* dst_argb, int width);
1204 void ARGBMultiplyRow_AVX2(const uint8* src_argb, const uint8* src_argb1,
1205 uint8* dst_argb, int width);
1206 void ARGBMultiplyRow_Any_AVX2(const uint8* src_argb, const uint8* src_argb1,
1207 uint8* dst_argb, int width);
1208 void ARGBMultiplyRow_NEON(const uint8* src_argb, const uint8* src_argb1,
1209 uint8* dst_argb, int width);
1210 void ARGBMultiplyRow_Any_NEON(const uint8* src_argb, const uint8* src_argb1,
1211 uint8* dst_argb, int width);
1214 void ARGBAddRow_C(const uint8* src_argb, const uint8* src_argb1,
1215 uint8* dst_argb, int width);
1216 void ARGBAddRow_SSE2(const uint8* src_argb, const uint8* src_argb1,
1217 uint8* dst_argb, int width);
1218 void ARGBAddRow_Any_SSE2(const uint8* src_argb, const uint8* src_argb1,
1219 uint8* dst_argb, int width);
1220 void ARGBAddRow_AVX2(const uint8* src_argb, const uint8* src_argb1,
1221 uint8* dst_argb, int width);
1222 void ARGBAddRow_Any_AVX2(const uint8* src_argb, const uint8* src_argb1,
1223 uint8* dst_argb, int width);
1224 void ARGBAddRow_NEON(const uint8* src_argb, const uint8* src_argb1,
1225 uint8* dst_argb, int width);
1226 void ARGBAddRow_Any_NEON(const uint8* src_argb, const uint8* src_argb1,
1227 uint8* dst_argb, int width);
1229 // ARGB subtract images. Same API as Blend, but these require
1230 // pointer and width alignment for SSE2.
1231 void ARGBSubtractRow_C(const uint8* src_argb, const uint8* src_argb1,
1232 uint8* dst_argb, int width);
1233 void ARGBSubtractRow_SSE2(const uint8* src_argb, const uint8* src_argb1,
1234 uint8* dst_argb, int width);
1235 void ARGBSubtractRow_Any_SSE2(const uint8* src_argb, const uint8* src_argb1,
1236 uint8* dst_argb, int width);
1237 void ARGBSubtractRow_AVX2(const uint8* src_argb, const uint8* src_argb1,
1238 uint8* dst_argb, int width);
1239 void ARGBSubtractRow_Any_AVX2(const uint8* src_argb, const uint8* src_argb1,
1240 uint8* dst_argb, int width);
1241 void ARGBSubtractRow_NEON(const uint8* src_argb, const uint8* src_argb1,
1242 uint8* dst_argb, int width);
1243 void ARGBSubtractRow_Any_NEON(const uint8* src_argb, const uint8* src_argb1,
1244 uint8* dst_argb, int width);
1246 void ARGBToRGB24Row_Any_SSSE3(const uint8* src_argb, uint8* dst_rgb, int pix);
1247 void ARGBToRAWRow_Any_SSSE3(const uint8* src_argb, uint8* dst_rgb, int pix);
1248 void ARGBToRGB565Row_Any_SSE2(const uint8* src_argb, uint8* dst_rgb, int pix);
1249 void ARGBToARGB1555Row_Any_SSE2(const uint8* src_argb, uint8* dst_rgb, int pix);
1250 void ARGBToARGB4444Row_Any_SSE2(const uint8* src_argb, uint8* dst_rgb, int pix);
1252 void ARGBToRGB24Row_Any_NEON(const uint8* src_argb, uint8* dst_rgb, int pix);
1253 void ARGBToRAWRow_Any_NEON(const uint8* src_argb, uint8* dst_rgb, int pix);
1254 void ARGBToRGB565Row_Any_NEON(const uint8* src_argb, uint8* dst_rgb, int pix);
1255 void ARGBToARGB1555Row_Any_NEON(const uint8* src_argb, uint8* dst_rgb, int pix);
1256 void ARGBToARGB4444Row_Any_NEON(const uint8* src_argb, uint8* dst_rgb, int pix);
1258 void I444ToARGBRow_Any_NEON(const uint8* src_y,
1263 void I422ToARGBRow_Any_NEON(const uint8* src_y,
1268 void I411ToARGBRow_Any_NEON(const uint8* src_y,
1273 void I422ToBGRARow_Any_NEON(const uint8* src_y,
1278 void I422ToABGRRow_Any_NEON(const uint8* src_y,
1283 void I422ToRGBARow_Any_NEON(const uint8* src_y,
1288 void I422ToRGB24Row_Any_NEON(const uint8* src_y,
1293 void I422ToRAWRow_Any_NEON(const uint8* src_y,
1298 void I422ToARGB4444Row_Any_NEON(const uint8* src_y,
1303 void I422ToARGB1555Row_Any_NEON(const uint8* src_y,
1308 void I422ToRGB565Row_Any_NEON(const uint8* src_y,
1313 void NV12ToARGBRow_Any_NEON(const uint8* src_y,
1314 const uint8* src_uv,
1317 void NV21ToARGBRow_Any_NEON(const uint8* src_y,
1318 const uint8* src_uv,
1321 void NV12ToRGB565Row_Any_NEON(const uint8* src_y,
1322 const uint8* src_uv,
1325 void NV21ToRGB565Row_Any_NEON(const uint8* src_y,
1326 const uint8* src_uv,
1329 void YUY2ToARGBRow_Any_NEON(const uint8* src_yuy2,
1332 void UYVYToARGBRow_Any_NEON(const uint8* src_uyvy,
1335 void I422ToARGBRow_MIPS_DSPR2(const uint8* src_y,
1340 void I422ToBGRARow_MIPS_DSPR2(const uint8* src_y,
1345 void I422ToABGRRow_MIPS_DSPR2(const uint8* src_y,
1350 void I422ToARGBRow_MIPS_DSPR2(const uint8* src_y,
1355 void I422ToBGRARow_MIPS_DSPR2(const uint8* src_y,
1360 void I422ToABGRRow_MIPS_DSPR2(const uint8* src_y,
1366 void YUY2ToYRow_AVX2(const uint8* src_yuy2, uint8* dst_y, int pix);
1367 void YUY2ToUVRow_AVX2(const uint8* src_yuy2, int stride_yuy2,
1368 uint8* dst_u, uint8* dst_v, int pix);
1369 void YUY2ToUV422Row_AVX2(const uint8* src_yuy2,
1370 uint8* dst_u, uint8* dst_v, int pix);
1371 void YUY2ToYRow_SSE2(const uint8* src_yuy2, uint8* dst_y, int pix);
1372 void YUY2ToUVRow_SSE2(const uint8* src_yuy2, int stride_yuy2,
1373 uint8* dst_u, uint8* dst_v, int pix);
1374 void YUY2ToUV422Row_SSE2(const uint8* src_yuy2,
1375 uint8* dst_u, uint8* dst_v, int pix);
1376 void YUY2ToYRow_Unaligned_SSE2(const uint8* src_yuy2,
1377 uint8* dst_y, int pix);
1378 void YUY2ToUVRow_Unaligned_SSE2(const uint8* src_yuy2, int stride_yuy2,
1379 uint8* dst_u, uint8* dst_v, int pix);
1380 void YUY2ToUV422Row_Unaligned_SSE2(const uint8* src_yuy2,
1381 uint8* dst_u, uint8* dst_v, int pix);
1382 void YUY2ToYRow_NEON(const uint8* src_yuy2, uint8* dst_y, int pix);
1383 void YUY2ToUVRow_NEON(const uint8* src_yuy2, int stride_yuy2,
1384 uint8* dst_u, uint8* dst_v, int pix);
1385 void YUY2ToUV422Row_NEON(const uint8* src_yuy2,
1386 uint8* dst_u, uint8* dst_v, int pix);
1387 void YUY2ToYRow_C(const uint8* src_yuy2, uint8* dst_y, int pix);
1388 void YUY2ToUVRow_C(const uint8* src_yuy2, int stride_yuy2,
1389 uint8* dst_u, uint8* dst_v, int pix);
1390 void YUY2ToUV422Row_C(const uint8* src_yuy2,
1391 uint8* dst_u, uint8* dst_v, int pix);
1392 void YUY2ToYRow_Any_AVX2(const uint8* src_yuy2, uint8* dst_y, int pix);
1393 void YUY2ToUVRow_Any_AVX2(const uint8* src_yuy2, int stride_yuy2,
1394 uint8* dst_u, uint8* dst_v, int pix);
1395 void YUY2ToUV422Row_Any_AVX2(const uint8* src_yuy2,
1396 uint8* dst_u, uint8* dst_v, int pix);
1397 void YUY2ToYRow_Any_SSE2(const uint8* src_yuy2, uint8* dst_y, int pix);
1398 void YUY2ToUVRow_Any_SSE2(const uint8* src_yuy2, int stride_yuy2,
1399 uint8* dst_u, uint8* dst_v, int pix);
1400 void YUY2ToUV422Row_Any_SSE2(const uint8* src_yuy2,
1401 uint8* dst_u, uint8* dst_v, int pix);
1402 void YUY2ToYRow_Any_NEON(const uint8* src_yuy2, uint8* dst_y, int pix);
1403 void YUY2ToUVRow_Any_NEON(const uint8* src_yuy2, int stride_yuy2,
1404 uint8* dst_u, uint8* dst_v, int pix);
1405 void YUY2ToUV422Row_Any_NEON(const uint8* src_yuy2,
1406 uint8* dst_u, uint8* dst_v, int pix);
1407 void UYVYToYRow_AVX2(const uint8* src_uyvy, uint8* dst_y, int pix);
1408 void UYVYToUVRow_AVX2(const uint8* src_uyvy, int stride_uyvy,
1409 uint8* dst_u, uint8* dst_v, int pix);
1410 void UYVYToUV422Row_AVX2(const uint8* src_uyvy,
1411 uint8* dst_u, uint8* dst_v, int pix);
1412 void UYVYToYRow_SSE2(const uint8* src_uyvy, uint8* dst_y, int pix);
1413 void UYVYToUVRow_SSE2(const uint8* src_uyvy, int stride_uyvy,
1414 uint8* dst_u, uint8* dst_v, int pix);
1415 void UYVYToUV422Row_SSE2(const uint8* src_uyvy,
1416 uint8* dst_u, uint8* dst_v, int pix);
1417 void UYVYToYRow_Unaligned_SSE2(const uint8* src_uyvy,
1418 uint8* dst_y, int pix);
1419 void UYVYToUVRow_Unaligned_SSE2(const uint8* src_uyvy, int stride_uyvy,
1420 uint8* dst_u, uint8* dst_v, int pix);
1421 void UYVYToUV422Row_Unaligned_SSE2(const uint8* src_uyvy,
1422 uint8* dst_u, uint8* dst_v, int pix);
1423 void UYVYToYRow_AVX2(const uint8* src_uyvy, uint8* dst_y, int pix);
1424 void UYVYToUVRow_AVX2(const uint8* src_uyvy, int stride_uyvy,
1425 uint8* dst_u, uint8* dst_v, int pix);
1426 void UYVYToUV422Row_AVX2(const uint8* src_uyvy,
1427 uint8* dst_u, uint8* dst_v, int pix);
1428 void UYVYToYRow_NEON(const uint8* src_uyvy, uint8* dst_y, int pix);
1429 void UYVYToUVRow_NEON(const uint8* src_uyvy, int stride_uyvy,
1430 uint8* dst_u, uint8* dst_v, int pix);
1431 void UYVYToUV422Row_NEON(const uint8* src_uyvy,
1432 uint8* dst_u, uint8* dst_v, int pix);
1434 void UYVYToYRow_C(const uint8* src_uyvy, uint8* dst_y, int pix);
1435 void UYVYToUVRow_C(const uint8* src_uyvy, int stride_uyvy,
1436 uint8* dst_u, uint8* dst_v, int pix);
1437 void UYVYToUV422Row_C(const uint8* src_uyvy,
1438 uint8* dst_u, uint8* dst_v, int pix);
1439 void UYVYToYRow_Any_AVX2(const uint8* src_uyvy, uint8* dst_y, int pix);
1440 void UYVYToUVRow_Any_AVX2(const uint8* src_uyvy, int stride_uyvy,
1441 uint8* dst_u, uint8* dst_v, int pix);
1442 void UYVYToUV422Row_Any_AVX2(const uint8* src_uyvy,
1443 uint8* dst_u, uint8* dst_v, int pix);
1444 void UYVYToYRow_Any_SSE2(const uint8* src_uyvy, uint8* dst_y, int pix);
1445 void UYVYToUVRow_Any_SSE2(const uint8* src_uyvy, int stride_uyvy,
1446 uint8* dst_u, uint8* dst_v, int pix);
1447 void UYVYToUV422Row_Any_SSE2(const uint8* src_uyvy,
1448 uint8* dst_u, uint8* dst_v, int pix);
1449 void UYVYToYRow_Any_NEON(const uint8* src_uyvy, uint8* dst_y, int pix);
1450 void UYVYToUVRow_Any_NEON(const uint8* src_uyvy, int stride_uyvy,
1451 uint8* dst_u, uint8* dst_v, int pix);
1452 void UYVYToUV422Row_Any_NEON(const uint8* src_uyvy,
1453 uint8* dst_u, uint8* dst_v, int pix);
1455 void HalfRow_C(const uint8* src_uv, int src_uv_stride,
1456 uint8* dst_uv, int pix);
1457 void HalfRow_SSE2(const uint8* src_uv, int src_uv_stride,
1458 uint8* dst_uv, int pix);
1459 void HalfRow_AVX2(const uint8* src_uv, int src_uv_stride,
1460 uint8* dst_uv, int pix);
1461 void HalfRow_NEON(const uint8* src_uv, int src_uv_stride,
1462 uint8* dst_uv, int pix);
1464 void HalfRow_16_C(const uint16* src_uv, int src_uv_stride,
1465 uint16* dst_uv, int pix);
1467 void ARGBToBayerRow_C(const uint8* src_argb, uint8* dst_bayer,
1468 uint32 selector, int pix);
1469 void ARGBToBayerRow_SSSE3(const uint8* src_argb, uint8* dst_bayer,
1470 uint32 selector, int pix);
1471 void ARGBToBayerRow_NEON(const uint8* src_argb, uint8* dst_bayer,
1472 uint32 selector, int pix);
1473 void ARGBToBayerRow_Any_SSSE3(const uint8* src_argb, uint8* dst_bayer,
1474 uint32 selector, int pix);
1475 void ARGBToBayerRow_Any_NEON(const uint8* src_argb, uint8* dst_bayer,
1476 uint32 selector, int pix);
1477 void ARGBToBayerGGRow_C(const uint8* src_argb, uint8* dst_bayer,
1478 uint32 /* selector */, int pix);
1479 void ARGBToBayerGGRow_SSE2(const uint8* src_argb, uint8* dst_bayer,
1480 uint32 /* selector */, int pix);
1481 void ARGBToBayerGGRow_NEON(const uint8* src_argb, uint8* dst_bayer,
1482 uint32 /* selector */, int pix);
1483 void ARGBToBayerGGRow_Any_SSE2(const uint8* src_argb, uint8* dst_bayer,
1484 uint32 /* selector */, int pix);
1485 void ARGBToBayerGGRow_Any_NEON(const uint8* src_argb, uint8* dst_bayer,
1486 uint32 /* selector */, int pix);
1488 void I422ToYUY2Row_C(const uint8* src_y,
1491 uint8* dst_yuy2, int width);
1492 void I422ToUYVYRow_C(const uint8* src_y,
1495 uint8* dst_uyvy, int width);
1496 void I422ToYUY2Row_SSE2(const uint8* src_y,
1499 uint8* dst_yuy2, int width);
1500 void I422ToUYVYRow_SSE2(const uint8* src_y,
1503 uint8* dst_uyvy, int width);
1504 void I422ToYUY2Row_Any_SSE2(const uint8* src_y,
1507 uint8* dst_yuy2, int width);
1508 void I422ToUYVYRow_Any_SSE2(const uint8* src_y,
1511 uint8* dst_uyvy, int width);
1512 void I422ToYUY2Row_NEON(const uint8* src_y,
1515 uint8* dst_yuy2, int width);
1516 void I422ToUYVYRow_NEON(const uint8* src_y,
1519 uint8* dst_uyvy, int width);
1520 void I422ToYUY2Row_Any_NEON(const uint8* src_y,
1523 uint8* dst_yuy2, int width);
1524 void I422ToUYVYRow_Any_NEON(const uint8* src_y,
1527 uint8* dst_uyvy, int width);
1529 // Effects related row functions.
1530 void ARGBAttenuateRow_C(const uint8* src_argb, uint8* dst_argb, int width);
1531 void ARGBAttenuateRow_SSE2(const uint8* src_argb, uint8* dst_argb, int width);
1532 void ARGBAttenuateRow_SSSE3(const uint8* src_argb, uint8* dst_argb, int width);
1533 void ARGBAttenuateRow_AVX2(const uint8* src_argb, uint8* dst_argb, int width);
1534 void ARGBAttenuateRow_NEON(const uint8* src_argb, uint8* dst_argb, int width);
1535 void ARGBAttenuateRow_Any_SSE2(const uint8* src_argb, uint8* dst_argb,
1537 void ARGBAttenuateRow_Any_SSSE3(const uint8* src_argb, uint8* dst_argb,
1539 void ARGBAttenuateRow_Any_AVX2(const uint8* src_argb, uint8* dst_argb,
1541 void ARGBAttenuateRow_Any_NEON(const uint8* src_argb, uint8* dst_argb,
1544 // Inverse table for unattenuate, shared by C and SSE2.
1545 extern const uint32 fixed_invtbl8[256];
1546 void ARGBUnattenuateRow_C(const uint8* src_argb, uint8* dst_argb, int width);
1547 void ARGBUnattenuateRow_SSE2(const uint8* src_argb, uint8* dst_argb, int width);
1548 void ARGBUnattenuateRow_AVX2(const uint8* src_argb, uint8* dst_argb, int width);
1549 void ARGBUnattenuateRow_Any_SSE2(const uint8* src_argb, uint8* dst_argb,
1551 void ARGBUnattenuateRow_Any_AVX2(const uint8* src_argb, uint8* dst_argb,
1554 void ARGBGrayRow_C(const uint8* src_argb, uint8* dst_argb, int width);
1555 void ARGBGrayRow_SSSE3(const uint8* src_argb, uint8* dst_argb, int width);
1556 void ARGBGrayRow_NEON(const uint8* src_argb, uint8* dst_argb, int width);
1558 void ARGBSepiaRow_C(uint8* dst_argb, int width);
1559 void ARGBSepiaRow_SSSE3(uint8* dst_argb, int width);
1560 void ARGBSepiaRow_NEON(uint8* dst_argb, int width);
1562 void ARGBColorMatrixRow_C(const uint8* src_argb, uint8* dst_argb,
1563 const int8* matrix_argb, int width);
1564 void ARGBColorMatrixRow_SSSE3(const uint8* src_argb, uint8* dst_argb,
1565 const int8* matrix_argb, int width);
1566 void ARGBColorMatrixRow_NEON(const uint8* src_argb, uint8* dst_argb,
1567 const int8* matrix_argb, int width);
1569 void ARGBColorTableRow_C(uint8* dst_argb, const uint8* table_argb, int width);
1570 void ARGBColorTableRow_X86(uint8* dst_argb, const uint8* table_argb, int width);
1572 void RGBColorTableRow_C(uint8* dst_argb, const uint8* table_argb, int width);
1573 void RGBColorTableRow_X86(uint8* dst_argb, const uint8* table_argb, int width);
1575 void ARGBQuantizeRow_C(uint8* dst_argb, int scale, int interval_size,
1576 int interval_offset, int width);
1577 void ARGBQuantizeRow_SSE2(uint8* dst_argb, int scale, int interval_size,
1578 int interval_offset, int width);
1579 void ARGBQuantizeRow_NEON(uint8* dst_argb, int scale, int interval_size,
1580 int interval_offset, int width);
1582 void ARGBShadeRow_C(const uint8* src_argb, uint8* dst_argb, int width,
1584 void ARGBShadeRow_SSE2(const uint8* src_argb, uint8* dst_argb, int width,
1586 void ARGBShadeRow_NEON(const uint8* src_argb, uint8* dst_argb, int width,
1590 void CumulativeSumToAverageRow_SSE2(const int32* topleft, const int32* botleft,
1591 int width, int area, uint8* dst, int count);
1592 void ComputeCumulativeSumRow_SSE2(const uint8* row, int32* cumsum,
1593 const int32* previous_cumsum, int width);
1595 void CumulativeSumToAverageRow_C(const int32* topleft, const int32* botleft,
1596 int width, int area, uint8* dst, int count);
1597 void ComputeCumulativeSumRow_C(const uint8* row, int32* cumsum,
1598 const int32* previous_cumsum, int width);
1601 void ARGBAffineRow_C(const uint8* src_argb, int src_argb_stride,
1602 uint8* dst_argb, const float* uv_dudv, int width);
1604 void ARGBAffineRow_SSE2(const uint8* src_argb, int src_argb_stride,
1605 uint8* dst_argb, const float* uv_dudv, int width);
1607 // Used for I420Scale, ARGBScale, and ARGBInterpolate.
1608 void InterpolateRow_C(uint8* dst_ptr, const uint8* src_ptr,
1609 ptrdiff_t src_stride_ptr,
1610 int width, int source_y_fraction);
1611 void InterpolateRow_SSE2(uint8* dst_ptr, const uint8* src_ptr,
1612 ptrdiff_t src_stride_ptr, int width,
1613 int source_y_fraction);
1614 void InterpolateRow_SSSE3(uint8* dst_ptr, const uint8* src_ptr,
1615 ptrdiff_t src_stride_ptr, int width,
1616 int source_y_fraction);
1617 void InterpolateRow_AVX2(uint8* dst_ptr, const uint8* src_ptr,
1618 ptrdiff_t src_stride_ptr, int width,
1619 int source_y_fraction);
1620 void InterpolateRow_NEON(uint8* dst_ptr, const uint8* src_ptr,
1621 ptrdiff_t src_stride_ptr, int width,
1622 int source_y_fraction);
1623 void InterpolateRows_MIPS_DSPR2(uint8* dst_ptr, const uint8* src_ptr,
1624 ptrdiff_t src_stride_ptr, int width,
1625 int source_y_fraction);
1626 void InterpolateRow_Unaligned_SSE2(uint8* dst_ptr, const uint8* src_ptr,
1627 ptrdiff_t src_stride_ptr, int width,
1628 int source_y_fraction);
1629 void InterpolateRow_Unaligned_SSSE3(uint8* dst_ptr, const uint8* src_ptr,
1630 ptrdiff_t src_stride_ptr, int width,
1631 int source_y_fraction);
1632 void InterpolateRow_Any_NEON(uint8* dst_ptr, const uint8* src_ptr,
1633 ptrdiff_t src_stride_ptr, int width,
1634 int source_y_fraction);
1635 void InterpolateRow_Any_SSE2(uint8* dst_ptr, const uint8* src_ptr,
1636 ptrdiff_t src_stride_ptr, int width,
1637 int source_y_fraction);
1638 void InterpolateRow_Any_SSSE3(uint8* dst_ptr, const uint8* src_ptr,
1639 ptrdiff_t src_stride_ptr, int width,
1640 int source_y_fraction);
1641 void InterpolateRow_Any_AVX2(uint8* dst_ptr, const uint8* src_ptr,
1642 ptrdiff_t src_stride_ptr, int width,
1643 int source_y_fraction);
1644 void InterpolateRows_Any_MIPS_DSPR2(uint8* dst_ptr, const uint8* src_ptr,
1645 ptrdiff_t src_stride_ptr, int width,
1646 int source_y_fraction);
1648 void InterpolateRow_16_C(uint16* dst_ptr, const uint16* src_ptr,
1649 ptrdiff_t src_stride_ptr,
1650 int width, int source_y_fraction);
1653 void SobelXRow_C(const uint8* src_y0, const uint8* src_y1, const uint8* src_y2,
1654 uint8* dst_sobelx, int width);
1655 void SobelXRow_SSE2(const uint8* src_y0, const uint8* src_y1,
1656 const uint8* src_y2, uint8* dst_sobelx, int width);
1657 void SobelXRow_NEON(const uint8* src_y0, const uint8* src_y1,
1658 const uint8* src_y2, uint8* dst_sobelx, int width);
1659 void SobelYRow_C(const uint8* src_y0, const uint8* src_y1,
1660 uint8* dst_sobely, int width);
1661 void SobelYRow_SSE2(const uint8* src_y0, const uint8* src_y1,
1662 uint8* dst_sobely, int width);
1663 void SobelYRow_NEON(const uint8* src_y0, const uint8* src_y1,
1664 uint8* dst_sobely, int width);
1665 void SobelRow_C(const uint8* src_sobelx, const uint8* src_sobely,
1666 uint8* dst_argb, int width);
1667 void SobelRow_SSE2(const uint8* src_sobelx, const uint8* src_sobely,
1668 uint8* dst_argb, int width);
1669 void SobelRow_NEON(const uint8* src_sobelx, const uint8* src_sobely,
1670 uint8* dst_argb, int width);
1671 void SobelToPlaneRow_C(const uint8* src_sobelx, const uint8* src_sobely,
1672 uint8* dst_y, int width);
1673 void SobelToPlaneRow_SSE2(const uint8* src_sobelx, const uint8* src_sobely,
1674 uint8* dst_y, int width);
1675 void SobelToPlaneRow_NEON(const uint8* src_sobelx, const uint8* src_sobely,
1676 uint8* dst_y, int width);
1677 void SobelXYRow_C(const uint8* src_sobelx, const uint8* src_sobely,
1678 uint8* dst_argb, int width);
1679 void SobelXYRow_SSE2(const uint8* src_sobelx, const uint8* src_sobely,
1680 uint8* dst_argb, int width);
1681 void SobelXYRow_NEON(const uint8* src_sobelx, const uint8* src_sobely,
1682 uint8* dst_argb, int width);
1684 void ARGBPolynomialRow_C(const uint8* src_argb,
1685 uint8* dst_argb, const float* poly,
1687 void ARGBPolynomialRow_SSE2(const uint8* src_argb,
1688 uint8* dst_argb, const float* poly,
1690 void ARGBPolynomialRow_AVX2(const uint8* src_argb,
1691 uint8* dst_argb, const float* poly,
1694 void ARGBLumaColorTableRow_C(const uint8* src_argb, uint8* dst_argb, int width,
1695 const uint8* luma, uint32 lumacoeff);
1696 void ARGBLumaColorTableRow_SSSE3(const uint8* src_argb, uint8* dst_argb,
1698 const uint8* luma, uint32 lumacoeff);
1702 } // namespace libyuv
1705 #endif // INCLUDE_LIBYUV_ROW_H_ NOLINT