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
|
/*
* Copyright (c) 2014 The WebM project authors. All Rights Reserved.
*
* Use of this source code is governed by a BSD-style license
* that can be found in the LICENSE file in the root of the source
* tree. An additional intellectual property rights grant can be found
* in the file PATENTS. All contributing project authors may
* be found in the AUTHORS file in the root of the source tree.
*/
#include "./vpx_config.h"
#include "vp9/common/vp9_common.h"
#include "vp9/encoder/vp9_variance.h"
#include "vpx_ports/mem.h"
#define DECL(w, opt) \
int vp9_highbd_sub_pixel_variance##w##xh_##opt(const uint16_t *src, \
ptrdiff_t src_stride, \
int x_offset, int y_offset, \
const uint16_t *dst, \
ptrdiff_t dst_stride, \
int height, unsigned int *sse);
#define DECLS(opt1, opt2) \
DECL(8, opt1); \
DECL(16, opt1)
DECLS(sse2, sse);
// DECLS(ssse3, ssse3);
#undef DECLS
#undef DECL
#define FN(w, h, wf, wlog2, hlog2, opt, cast) \
uint32_t vp9_highbd_sub_pixel_variance##w##x##h##_##opt(const uint8_t *src8, \
int src_stride, \
int x_offset, \
int y_offset, \
const uint8_t *dst8, \
int dst_stride, \
uint32_t *sse_ptr) { \
uint32_t sse; \
uint16_t *src = CONVERT_TO_SHORTPTR(src8); \
uint16_t *dst = CONVERT_TO_SHORTPTR(dst8); \
int se = vp9_highbd_sub_pixel_variance##wf##xh_##opt(src, src_stride, \
x_offset, y_offset, \
dst, dst_stride, h, \
&sse); \
if (w > wf) { \
unsigned int sse2; \
int se2 = vp9_highbd_sub_pixel_variance##wf##xh_##opt(src + 16, \
src_stride, \
x_offset, y_offset, \
dst + 16, \
dst_stride, \
h, &sse2); \
se += se2; \
sse += sse2; \
if (w > wf * 2) { \
se2 = vp9_highbd_sub_pixel_variance##wf##xh_##opt(src + 32, src_stride, \
x_offset, y_offset, \
dst + 32, dst_stride, \
h, &sse2); \
se += se2; \
sse += sse2; \
se2 = vp9_highbd_sub_pixel_variance##wf##xh_##opt( \
src + 48, src_stride, x_offset, y_offset, \
dst + 48, dst_stride, h, &sse2); \
se += se2; \
sse += sse2; \
} \
} \
*sse_ptr = sse; \
return sse - ((cast se * se) >> (wlog2 + hlog2)); \
} \
\
uint32_t vp9_highbd_10_sub_pixel_variance##w##x##h##_##opt( \
const uint8_t *src8, int src_stride, int x_offset, int y_offset, \
const uint8_t *dst8, int dst_stride, uint32_t *sse_ptr) { \
uint32_t sse; \
uint16_t *src = CONVERT_TO_SHORTPTR(src8); \
uint16_t *dst = CONVERT_TO_SHORTPTR(dst8); \
int se = vp9_highbd_sub_pixel_variance##wf##xh_##opt(src, src_stride, \
x_offset, y_offset, \
dst, dst_stride, \
h, &sse); \
if (w > wf) { \
uint32_t sse2; \
int se2 = vp9_highbd_sub_pixel_variance##wf##xh_##opt(src + 16, \
src_stride, \
x_offset, y_offset, \
dst + 16, \
dst_stride, \
h, &sse2); \
se += se2; \
sse += sse2; \
if (w > wf * 2) { \
se2 = vp9_highbd_sub_pixel_variance##wf##xh_##opt(src + 32, src_stride, \
x_offset, y_offset, \
dst + 32, dst_stride, \
h, &sse2); \
se += se2; \
sse += sse2; \
se2 = vp9_highbd_sub_pixel_variance##wf##xh_##opt(src + 48, src_stride, \
x_offset, y_offset, \
dst + 48, dst_stride, \
h, &sse2); \
se += se2; \
sse += sse2; \
} \
} \
se = ROUND_POWER_OF_TWO(se, 2); \
sse = ROUND_POWER_OF_TWO(sse, 4); \
*sse_ptr = sse; \
return sse - ((cast se * se) >> (wlog2 + hlog2)); \
} \
\
uint32_t vp9_highbd_12_sub_pixel_variance##w##x##h##_##opt( \
const uint8_t *src8, int src_stride, int x_offset, int y_offset, \
const uint8_t *dst8, int dst_stride, uint32_t *sse_ptr) { \
int start_row; \
uint32_t sse; \
int se = 0; \
uint64_t long_sse = 0; \
uint16_t *src = CONVERT_TO_SHORTPTR(src8); \
uint16_t *dst = CONVERT_TO_SHORTPTR(dst8); \
for (start_row = 0; start_row < h; start_row +=16) { \
uint32_t sse2; \
int height = h - start_row < 16 ? h - start_row : 16; \
int se2 = vp9_highbd_sub_pixel_variance##wf##xh_##opt( \
src + (start_row * src_stride), src_stride, \
x_offset, y_offset, dst + (start_row * dst_stride), \
dst_stride, height, &sse2); \
se += se2; \
long_sse += sse2; \
if (w > wf) { \
se2 = vp9_highbd_sub_pixel_variance##wf##xh_##opt( \
src + 16 + (start_row * src_stride), src_stride, \
x_offset, y_offset, dst + 16 + (start_row * dst_stride), \
dst_stride, height, &sse2); \
se += se2; \
long_sse += sse2; \
if (w > wf * 2) { \
se2 = vp9_highbd_sub_pixel_variance##wf##xh_##opt( \
src + 32 + (start_row * src_stride), src_stride, \
x_offset, y_offset, dst + 32 + (start_row * dst_stride), \
dst_stride, height, &sse2); \
se += se2; \
long_sse += sse2; \
se2 = vp9_highbd_sub_pixel_variance##wf##xh_##opt( \
src + 48 + (start_row * src_stride), src_stride, \
x_offset, y_offset, dst + 48 + (start_row * dst_stride), \
dst_stride, height, &sse2); \
se += se2; \
long_sse += sse2; \
}\
} \
} \
se = ROUND_POWER_OF_TWO(se, 4); \
sse = ROUND_POWER_OF_TWO(long_sse, 8); \
*sse_ptr = sse; \
return sse - ((cast se * se) >> (wlog2 + hlog2)); \
}
#define FNS(opt1, opt2) \
FN(64, 64, 16, 6, 6, opt1, (int64_t)); \
FN(64, 32, 16, 6, 5, opt1, (int64_t)); \
FN(32, 64, 16, 5, 6, opt1, (int64_t)); \
FN(32, 32, 16, 5, 5, opt1, (int64_t)); \
FN(32, 16, 16, 5, 4, opt1, (int64_t)); \
FN(16, 32, 16, 4, 5, opt1, (int64_t)); \
FN(16, 16, 16, 4, 4, opt1, (int64_t)); \
FN(16, 8, 16, 4, 3, opt1, (int64_t)); \
FN(8, 16, 8, 3, 4, opt1, (int64_t)); \
FN(8, 8, 8, 3, 3, opt1, (int64_t)); \
FN(8, 4, 8, 3, 2, opt1, (int64_t));
FNS(sse2, sse);
#undef FNS
#undef FN
#define DECL(w, opt) \
int vp9_highbd_sub_pixel_avg_variance##w##xh_##opt(const uint16_t *src, \
ptrdiff_t src_stride, \
int x_offset, int y_offset, \
const uint16_t *dst, \
ptrdiff_t dst_stride, \
const uint16_t *sec, \
ptrdiff_t sec_stride, \
int height, \
unsigned int *sse);
#define DECLS(opt1) \
DECL(16, opt1) \
DECL(8, opt1)
DECLS(sse2);
#undef DECL
#undef DECLS
#define FN(w, h, wf, wlog2, hlog2, opt, cast) \
uint32_t vp9_highbd_sub_pixel_avg_variance##w##x##h##_##opt( \
const uint8_t *src8, int src_stride, int x_offset, int y_offset, \
const uint8_t *dst8, int dst_stride, uint32_t *sse_ptr, \
const uint8_t *sec8) { \
uint32_t sse; \
uint16_t *src = CONVERT_TO_SHORTPTR(src8); \
uint16_t *dst = CONVERT_TO_SHORTPTR(dst8); \
uint16_t *sec = CONVERT_TO_SHORTPTR(sec8); \
int se = vp9_highbd_sub_pixel_avg_variance##wf##xh_##opt( \
src, src_stride, x_offset, \
y_offset, dst, dst_stride, sec, w, h, &sse); \
if (w > wf) { \
uint32_t sse2; \
int se2 = vp9_highbd_sub_pixel_avg_variance##wf##xh_##opt( \
src + 16, src_stride, x_offset, y_offset, \
dst + 16, dst_stride, sec + 16, w, h, &sse2); \
se += se2; \
sse += sse2; \
if (w > wf * 2) { \
se2 = vp9_highbd_sub_pixel_avg_variance##wf##xh_##opt( \
src + 32, src_stride, x_offset, y_offset, \
dst + 32, dst_stride, sec + 32, w, h, &sse2); \
se += se2; \
sse += sse2; \
se2 = vp9_highbd_sub_pixel_avg_variance##wf##xh_##opt( \
src + 48, src_stride, x_offset, y_offset, \
dst + 48, dst_stride, sec + 48, w, h, &sse2); \
se += se2; \
sse += sse2; \
} \
} \
*sse_ptr = sse; \
return sse - ((cast se * se) >> (wlog2 + hlog2)); \
} \
\
uint32_t vp9_highbd_10_sub_pixel_avg_variance##w##x##h##_##opt( \
const uint8_t *src8, int src_stride, int x_offset, int y_offset, \
const uint8_t *dst8, int dst_stride, uint32_t *sse_ptr, \
const uint8_t *sec8) { \
uint32_t sse; \
uint16_t *src = CONVERT_TO_SHORTPTR(src8); \
uint16_t *dst = CONVERT_TO_SHORTPTR(dst8); \
uint16_t *sec = CONVERT_TO_SHORTPTR(sec8); \
int se = vp9_highbd_sub_pixel_avg_variance##wf##xh_##opt( \
src, src_stride, x_offset, \
y_offset, dst, dst_stride, \
sec, w, h, &sse); \
if (w > wf) { \
uint32_t sse2; \
int se2 = vp9_highbd_sub_pixel_avg_variance##wf##xh_##opt( \
src + 16, src_stride, \
x_offset, y_offset, \
dst + 16, dst_stride, \
sec + 16, w, h, &sse2); \
se += se2; \
sse += sse2; \
if (w > wf * 2) { \
se2 = vp9_highbd_sub_pixel_avg_variance##wf##xh_##opt( \
src + 32, src_stride, \
x_offset, y_offset, \
dst + 32, dst_stride, \
sec + 32, w, h, &sse2); \
se += se2; \
sse += sse2; \
se2 = vp9_highbd_sub_pixel_avg_variance##wf##xh_##opt( \
src + 48, src_stride, \
x_offset, y_offset, \
dst + 48, dst_stride, \
sec + 48, w, h, &sse2); \
se += se2; \
sse += sse2; \
} \
} \
se = ROUND_POWER_OF_TWO(se, 2); \
sse = ROUND_POWER_OF_TWO(sse, 4); \
*sse_ptr = sse; \
return sse - ((cast se * se) >> (wlog2 + hlog2)); \
} \
\
uint32_t vp9_highbd_12_sub_pixel_avg_variance##w##x##h##_##opt( \
const uint8_t *src8, int src_stride, int x_offset, int y_offset, \
const uint8_t *dst8, int dst_stride, uint32_t *sse_ptr, \
const uint8_t *sec8) { \
int start_row; \
uint32_t sse; \
int se = 0; \
uint64_t long_sse = 0; \
uint16_t *src = CONVERT_TO_SHORTPTR(src8); \
uint16_t *dst = CONVERT_TO_SHORTPTR(dst8); \
uint16_t *sec = CONVERT_TO_SHORTPTR(sec8); \
for (start_row = 0; start_row < h; start_row +=16) { \
uint32_t sse2; \
int height = h - start_row < 16 ? h - start_row : 16; \
int se2 = vp9_highbd_sub_pixel_avg_variance##wf##xh_##opt( \
src + (start_row * src_stride), src_stride, x_offset, \
y_offset, dst + (start_row * dst_stride), dst_stride, \
sec + (start_row * w), w, height, &sse2); \
se += se2; \
long_sse += sse2; \
if (w > wf) { \
se2 = vp9_highbd_sub_pixel_avg_variance##wf##xh_##opt( \
src + 16 + (start_row * src_stride), src_stride, \
x_offset, y_offset, \
dst + 16 + (start_row * dst_stride), dst_stride, \
sec + 16 + (start_row * w), w, height, &sse2); \
se += se2; \
long_sse += sse2; \
if (w > wf * 2) { \
se2 = vp9_highbd_sub_pixel_avg_variance##wf##xh_##opt( \
src + 32 + (start_row * src_stride), src_stride, \
x_offset, y_offset, \
dst + 32 + (start_row * dst_stride), dst_stride, \
sec + 32 + (start_row * w), w, height, &sse2); \
se += se2; \
long_sse += sse2; \
se2 = vp9_highbd_sub_pixel_avg_variance##wf##xh_##opt( \
src + 48 + (start_row * src_stride), src_stride, \
x_offset, y_offset, \
dst + 48 + (start_row * dst_stride), dst_stride, \
sec + 48 + (start_row * w), w, height, &sse2); \
se += se2; \
long_sse += sse2; \
} \
} \
} \
se = ROUND_POWER_OF_TWO(se, 4); \
sse = ROUND_POWER_OF_TWO(long_sse, 8); \
*sse_ptr = sse; \
return sse - ((cast se * se) >> (wlog2 + hlog2)); \
}
#define FNS(opt1) \
FN(64, 64, 16, 6, 6, opt1, (int64_t)); \
FN(64, 32, 16, 6, 5, opt1, (int64_t)); \
FN(32, 64, 16, 5, 6, opt1, (int64_t)); \
FN(32, 32, 16, 5, 5, opt1, (int64_t)); \
FN(32, 16, 16, 5, 4, opt1, (int64_t)); \
FN(16, 32, 16, 4, 5, opt1, (int64_t)); \
FN(16, 16, 16, 4, 4, opt1, (int64_t)); \
FN(16, 8, 16, 4, 3, opt1, (int64_t)); \
FN(8, 16, 8, 4, 3, opt1, (int64_t)); \
FN(8, 8, 8, 3, 3, opt1, (int64_t)); \
FN(8, 4, 8, 3, 2, opt1, (int64_t));
FNS(sse2);
#undef FNS
#undef FN
|
