1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
|
/*
* Copyright (c) 2016, Alliance for Open Media. All rights reserved
*
* This source code is subject to the terms of the BSD 2 Clause License and
* the Alliance for Open Media Patent License 1.0. If the BSD 2 Clause License
* was not distributed with this source code in the LICENSE file, you can
* obtain it at www.aomedia.org/license/software. If the Alliance for Open
* Media Patent License 1.0 was not distributed with this source code in the
* PATENTS file, you can obtain it at www.aomedia.org/license/patent.
*/
#include <stdlib.h>
#include "config/aom_config.h"
#include "config/aom_dsp_rtcd.h"
#include "aom/aom_integer.h"
#include "aom_ports/mem.h"
#include "aom_dsp/blend.h"
/* Sum the difference between every corresponding element of the buffers. */
static INLINE unsigned int sad(const uint8_t *a, int a_stride, const uint8_t *b,
int b_stride, int width, int height) {
int y, x;
unsigned int sad = 0;
for (y = 0; y < height; y++) {
for (x = 0; x < width; x++) {
sad += abs(a[x] - b[x]);
}
a += a_stride;
b += b_stride;
}
return sad;
}
#define SADMXN(m, n) \
unsigned int aom_sad##m##x##n##_c(const uint8_t *src, int src_stride, \
const uint8_t *ref, int ref_stride) { \
return sad(src, src_stride, ref, ref_stride, m, n); \
} \
unsigned int aom_sad##m##x##n##_avg_c(const uint8_t *src, int src_stride, \
const uint8_t *ref, int ref_stride, \
const uint8_t *second_pred) { \
uint8_t comp_pred[m * n]; \
aom_comp_avg_pred(comp_pred, second_pred, m, n, ref, ref_stride); \
return sad(src, src_stride, comp_pred, m, m, n); \
} \
unsigned int aom_dist_wtd_sad##m##x##n##_avg_c( \
const uint8_t *src, int src_stride, const uint8_t *ref, int ref_stride, \
const uint8_t *second_pred, const DIST_WTD_COMP_PARAMS *jcp_param) { \
uint8_t comp_pred[m * n]; \
aom_dist_wtd_comp_avg_pred_c(comp_pred, second_pred, m, n, ref, \
ref_stride, jcp_param); \
return sad(src, src_stride, comp_pred, m, m, n); \
} \
unsigned int aom_sad_skip_##m##x##n##_c(const uint8_t *src, int src_stride, \
const uint8_t *ref, \
int ref_stride) { \
return 2 * sad(src, 2 * src_stride, ref, 2 * ref_stride, (m), (n / 2)); \
}
// Calculate sad against 4 reference locations and store each in sad_array
#define SAD_MXNX4D(m, n) \
void aom_sad##m##x##n##x4d_c(const uint8_t *src, int src_stride, \
const uint8_t *const ref_array[4], \
int ref_stride, uint32_t sad_array[4]) { \
int i; \
for (i = 0; i < 4; ++i) { \
sad_array[i] = \
aom_sad##m##x##n##_c(src, src_stride, ref_array[i], ref_stride); \
} \
} \
void aom_sad_skip_##m##x##n##x4d_c(const uint8_t *src, int src_stride, \
const uint8_t *const ref_array[4], \
int ref_stride, uint32_t sad_array[4]) { \
int i; \
for (i = 0; i < 4; ++i) { \
sad_array[i] = 2 * sad(src, 2 * src_stride, ref_array[i], \
2 * ref_stride, (m), (n / 2)); \
} \
}
// Call SIMD version of aom_sad_mxnx4d if the 3d version is unavailable.
#define SAD_MXNX3D(m, n) \
void aom_sad##m##x##n##x3d_c(const uint8_t *src, int src_stride, \
const uint8_t *const ref_array[4], \
int ref_stride, uint32_t sad_array[4]) { \
aom_sad##m##x##n##x4d(src, src_stride, ref_array, ref_stride, sad_array); \
}
// 128x128
SADMXN(128, 128)
SAD_MXNX4D(128, 128)
SAD_MXNX3D(128, 128)
// 128x64
SADMXN(128, 64)
SAD_MXNX4D(128, 64)
SAD_MXNX3D(128, 64)
// 64x128
SADMXN(64, 128)
SAD_MXNX4D(64, 128)
SAD_MXNX3D(64, 128)
// 64x64
SADMXN(64, 64)
SAD_MXNX4D(64, 64)
SAD_MXNX3D(64, 64)
// 64x32
SADMXN(64, 32)
SAD_MXNX4D(64, 32)
SAD_MXNX3D(64, 32)
// 32x64
SADMXN(32, 64)
SAD_MXNX4D(32, 64)
SAD_MXNX3D(32, 64)
// 32x32
SADMXN(32, 32)
SAD_MXNX4D(32, 32)
SAD_MXNX3D(32, 32)
// 32x16
SADMXN(32, 16)
SAD_MXNX4D(32, 16)
SAD_MXNX3D(32, 16)
// 16x32
SADMXN(16, 32)
SAD_MXNX4D(16, 32)
SAD_MXNX3D(16, 32)
// 16x16
SADMXN(16, 16)
SAD_MXNX4D(16, 16)
SAD_MXNX3D(16, 16)
// 16x8
SADMXN(16, 8)
SAD_MXNX4D(16, 8)
SAD_MXNX3D(16, 8)
// 8x16
SADMXN(8, 16)
SAD_MXNX4D(8, 16)
SAD_MXNX3D(8, 16)
// 8x8
SADMXN(8, 8)
SAD_MXNX4D(8, 8)
SAD_MXNX3D(8, 8)
// 8x4
SADMXN(8, 4)
SAD_MXNX4D(8, 4)
SAD_MXNX3D(8, 4)
// 4x8
SADMXN(4, 8)
SAD_MXNX4D(4, 8)
SAD_MXNX3D(4, 8)
// 4x4
SADMXN(4, 4)
SAD_MXNX4D(4, 4)
SAD_MXNX3D(4, 4)
#if !CONFIG_REALTIME_ONLY
SADMXN(4, 16)
SAD_MXNX4D(4, 16)
SADMXN(16, 4)
SAD_MXNX4D(16, 4)
SADMXN(8, 32)
SAD_MXNX4D(8, 32)
SADMXN(32, 8)
SAD_MXNX4D(32, 8)
SADMXN(16, 64)
SAD_MXNX4D(16, 64)
SADMXN(64, 16)
SAD_MXNX4D(64, 16)
SAD_MXNX3D(4, 16)
SAD_MXNX3D(16, 4)
SAD_MXNX3D(8, 32)
SAD_MXNX3D(32, 8)
SAD_MXNX3D(16, 64)
SAD_MXNX3D(64, 16)
#endif // !CONFIG_REALTIME_ONLY
#if CONFIG_AV1_HIGHBITDEPTH
static INLINE unsigned int highbd_sad(const uint8_t *a8, int a_stride,
const uint8_t *b8, int b_stride,
int width, int height) {
int y, x;
unsigned int sad = 0;
const uint16_t *a = CONVERT_TO_SHORTPTR(a8);
const uint16_t *b = CONVERT_TO_SHORTPTR(b8);
for (y = 0; y < height; y++) {
for (x = 0; x < width; x++) {
sad += abs(a[x] - b[x]);
}
a += a_stride;
b += b_stride;
}
return sad;
}
static INLINE unsigned int highbd_sadb(const uint8_t *a8, int a_stride,
const uint8_t *b8, int b_stride,
int width, int height) {
int y, x;
unsigned int sad = 0;
const uint16_t *a = CONVERT_TO_SHORTPTR(a8);
const uint16_t *b = CONVERT_TO_SHORTPTR(b8);
for (y = 0; y < height; y++) {
for (x = 0; x < width; x++) {
sad += abs(a[x] - b[x]);
}
a += a_stride;
b += b_stride;
}
return sad;
}
#define HIGHBD_SADMXN(m, n) \
unsigned int aom_highbd_sad##m##x##n##_c(const uint8_t *src, int src_stride, \
const uint8_t *ref, \
int ref_stride) { \
return highbd_sad(src, src_stride, ref, ref_stride, m, n); \
} \
unsigned int aom_highbd_sad##m##x##n##_avg_c( \
const uint8_t *src, int src_stride, const uint8_t *ref, int ref_stride, \
const uint8_t *second_pred) { \
uint16_t comp_pred[m * n]; \
uint8_t *const comp_pred8 = CONVERT_TO_BYTEPTR(comp_pred); \
aom_highbd_comp_avg_pred(comp_pred8, second_pred, m, n, ref, ref_stride); \
return highbd_sadb(src, src_stride, comp_pred8, m, m, n); \
} \
unsigned int aom_highbd_dist_wtd_sad##m##x##n##_avg_c( \
const uint8_t *src, int src_stride, const uint8_t *ref, int ref_stride, \
const uint8_t *second_pred, const DIST_WTD_COMP_PARAMS *jcp_param) { \
uint16_t comp_pred[m * n]; \
uint8_t *const comp_pred8 = CONVERT_TO_BYTEPTR(comp_pred); \
aom_highbd_dist_wtd_comp_avg_pred(comp_pred8, second_pred, m, n, ref, \
ref_stride, jcp_param); \
return highbd_sadb(src, src_stride, comp_pred8, m, m, n); \
} \
unsigned int aom_highbd_sad_skip_##m##x##n##_c( \
const uint8_t *src, int src_stride, const uint8_t *ref, \
int ref_stride) { \
return 2 * \
highbd_sad(src, 2 * src_stride, ref, 2 * ref_stride, (m), (n / 2)); \
}
#define HIGHBD_SAD_MXNX4D(m, n) \
void aom_highbd_sad##m##x##n##x4d_c(const uint8_t *src, int src_stride, \
const uint8_t *const ref_array[4], \
int ref_stride, uint32_t sad_array[4]) { \
int i; \
for (i = 0; i < 4; ++i) { \
sad_array[i] = aom_highbd_sad##m##x##n##_c(src, src_stride, \
ref_array[i], ref_stride); \
} \
} \
void aom_highbd_sad_skip_##m##x##n##x4d_c( \
const uint8_t *src, int src_stride, const uint8_t *const ref_array[4], \
int ref_stride, uint32_t sad_array[4]) { \
int i; \
for (i = 0; i < 4; ++i) { \
sad_array[i] = 2 * highbd_sad(src, 2 * src_stride, ref_array[i], \
2 * ref_stride, (m), (n / 2)); \
} \
}
// Call SIMD version of aom_highbd_sad_mxnx4d if the 3d version is unavailable.
#define HIGHBD_SAD_MXNX3D(m, n) \
void aom_highbd_sad##m##x##n##x3d_c(const uint8_t *src, int src_stride, \
const uint8_t *const ref_array[4], \
int ref_stride, uint32_t sad_array[4]) { \
aom_highbd_sad##m##x##n##x4d(src, src_stride, ref_array, ref_stride, \
sad_array); \
}
// 128x128
HIGHBD_SADMXN(128, 128)
HIGHBD_SAD_MXNX4D(128, 128)
HIGHBD_SAD_MXNX3D(128, 128)
// 128x64
HIGHBD_SADMXN(128, 64)
HIGHBD_SAD_MXNX4D(128, 64)
HIGHBD_SAD_MXNX3D(128, 64)
// 64x128
HIGHBD_SADMXN(64, 128)
HIGHBD_SAD_MXNX4D(64, 128)
HIGHBD_SAD_MXNX3D(64, 128)
// 64x64
HIGHBD_SADMXN(64, 64)
HIGHBD_SAD_MXNX4D(64, 64)
HIGHBD_SAD_MXNX3D(64, 64)
// 64x32
HIGHBD_SADMXN(64, 32)
HIGHBD_SAD_MXNX4D(64, 32)
HIGHBD_SAD_MXNX3D(64, 32)
// 32x64
HIGHBD_SADMXN(32, 64)
HIGHBD_SAD_MXNX4D(32, 64)
HIGHBD_SAD_MXNX3D(32, 64)
// 32x32
HIGHBD_SADMXN(32, 32)
HIGHBD_SAD_MXNX4D(32, 32)
HIGHBD_SAD_MXNX3D(32, 32)
// 32x16
HIGHBD_SADMXN(32, 16)
HIGHBD_SAD_MXNX4D(32, 16)
HIGHBD_SAD_MXNX3D(32, 16)
// 16x32
HIGHBD_SADMXN(16, 32)
HIGHBD_SAD_MXNX4D(16, 32)
HIGHBD_SAD_MXNX3D(16, 32)
// 16x16
HIGHBD_SADMXN(16, 16)
HIGHBD_SAD_MXNX4D(16, 16)
HIGHBD_SAD_MXNX3D(16, 16)
// 16x8
HIGHBD_SADMXN(16, 8)
HIGHBD_SAD_MXNX4D(16, 8)
HIGHBD_SAD_MXNX3D(16, 8)
// 8x16
HIGHBD_SADMXN(8, 16)
HIGHBD_SAD_MXNX4D(8, 16)
HIGHBD_SAD_MXNX3D(8, 16)
// 8x8
HIGHBD_SADMXN(8, 8)
HIGHBD_SAD_MXNX4D(8, 8)
HIGHBD_SAD_MXNX3D(8, 8)
// 8x4
HIGHBD_SADMXN(8, 4)
HIGHBD_SAD_MXNX4D(8, 4)
HIGHBD_SAD_MXNX3D(8, 4)
// 4x8
HIGHBD_SADMXN(4, 8)
HIGHBD_SAD_MXNX4D(4, 8)
HIGHBD_SAD_MXNX3D(4, 8)
// 4x4
HIGHBD_SADMXN(4, 4)
HIGHBD_SAD_MXNX4D(4, 4)
HIGHBD_SAD_MXNX3D(4, 4)
HIGHBD_SADMXN(4, 16)
HIGHBD_SAD_MXNX4D(4, 16)
HIGHBD_SADMXN(16, 4)
HIGHBD_SAD_MXNX4D(16, 4)
HIGHBD_SADMXN(8, 32)
HIGHBD_SAD_MXNX4D(8, 32)
HIGHBD_SADMXN(32, 8)
HIGHBD_SAD_MXNX4D(32, 8)
HIGHBD_SADMXN(16, 64)
HIGHBD_SAD_MXNX4D(16, 64)
HIGHBD_SADMXN(64, 16)
HIGHBD_SAD_MXNX4D(64, 16)
#if !CONFIG_REALTIME_ONLY
HIGHBD_SAD_MXNX3D(4, 16)
HIGHBD_SAD_MXNX3D(16, 4)
HIGHBD_SAD_MXNX3D(8, 32)
HIGHBD_SAD_MXNX3D(32, 8)
HIGHBD_SAD_MXNX3D(16, 64)
HIGHBD_SAD_MXNX3D(64, 16)
#endif // !CONFIG_REALTIME_ONLY
#endif // CONFIG_AV1_HIGHBITDEPTH
|
