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
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
|
/*
* Simd Library (http://ermig1979.github.io/Simd).
*
* Copyright (c) 2011-2022 Yermalayeu Ihar.
*
* Permission is hereby granted, free of charge, to any person obtaining a copy
* of this software and associated documentation files (the "Software"), to deal
* in the Software without restriction, including without limitation the rights
* to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
* copies of the Software, and to permit persons to whom the Software is
* furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included in
* all copies or substantial portions of the Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
* SOFTWARE.
*/
#include "Simd/SimdMemory.h"
#include "Simd/SimdStore.h"
#include "Simd/SimdResizer.h"
#include "Simd/SimdResizerCommon.h"
namespace Simd
{
#ifdef SIMD_SSE41_ENABLE
namespace Sse41
{
ResizerByteBilinear::ResizerByteBilinear(const ResParam& param)
: Base::ResizerByteBilinear(param)
, _blocks(0)
{
}
size_t ResizerByteBilinear::BlockCountMax(size_t align)
{
return (size_t)Simd::Max(::ceil(float(_param.srcW) / (align - 1)), ::ceil(float(_param.dstW) * 2.0f / align));
}
void ResizerByteBilinear::EstimateParams()
{
if (_ax.data)
return;
if (_param.channels == 1 && _param.srcW < 4 * _param.dstW)
_blocks = BlockCountMax(A);
float scale = (float)_param.srcW / _param.dstW;
_ax.Resize(AlignHi(_param.dstW, A) * _param.channels * 2, false, _param.align);
uint8_t* alphas = _ax.data;
if (_blocks)
{
_ixg.Resize(_blocks);
int block = 0;
_ixg[0].src = 0;
_ixg[0].dst = 0;
for (int dstIndex = 0; dstIndex < (int)_param.dstW; ++dstIndex)
{
float alpha = (float)((dstIndex + 0.5) * scale - 0.5);
int srcIndex = (int)::floor(alpha);
alpha -= srcIndex;
if (srcIndex < 0)
{
srcIndex = 0;
alpha = 0;
}
if (srcIndex > (int)_param.srcW - 2)
{
srcIndex = (int)_param.srcW - 2;
alpha = 1;
}
int dst = 2 * dstIndex - _ixg[block].dst;
int src = srcIndex - _ixg[block].src;
if (src >= A - 1 || dst >= A)
{
block++;
_ixg[block].src = Simd::Min(srcIndex, int(_param.srcW - A));
_ixg[block].dst = 2 * dstIndex;
dst = 0;
src = srcIndex - _ixg[block].src;
}
_ixg[block].shuffle[dst] = src;
_ixg[block].shuffle[dst + 1] = src + 1;
alphas[1] = (uint8_t)(alpha * Base::FRACTION_RANGE + 0.5);
alphas[0] = (uint8_t)(Base::FRACTION_RANGE - alphas[1]);
alphas += 2;
}
_blocks = block + 1;
}
else
{
_ix.Resize(_param.dstW);
for (size_t i = 0; i < _param.dstW; ++i)
{
float alpha = (float)((i + 0.5) * scale - 0.5);
ptrdiff_t index = (ptrdiff_t)::floor(alpha);
alpha -= index;
if (index < 0)
{
index = 0;
alpha = 0;
}
if (index > (ptrdiff_t)_param.srcW - 2)
{
index = _param.srcW - 2;
alpha = 1;
}
_ix[i] = (int)index;
alphas[1] = (uint8_t)(alpha * Base::FRACTION_RANGE + 0.5);
alphas[0] = (uint8_t)(Base::FRACTION_RANGE - alphas[1]);
for (size_t channel = 1; channel < _param.channels; channel++)
((uint16_t*)alphas)[channel] = *(uint16_t*)alphas;
alphas += 2 * _param.channels;
}
}
size_t size = AlignHi(_param.dstW, _param.align) * _param.channels * 2 + SIMD_ALIGN;
_bx[0].Resize(size, false, _param.align);
_bx[1].Resize(size, false, _param.align);
}
template <size_t N> void ResizerByteBilinearInterpolateX(const __m128i* alpha, __m128i* buffer);
template <> SIMD_INLINE void ResizerByteBilinearInterpolateX<1>(const __m128i* alpha, __m128i* buffer)
{
_mm_store_si128(buffer, _mm_maddubs_epi16(_mm_load_si128(buffer), _mm_load_si128(alpha)));
}
const __m128i K8_SHUFFLE_X2 = SIMD_MM_SETR_EPI8(0x0, 0x2, 0x1, 0x3, 0x4, 0x6, 0x5, 0x7, 0x8, 0xA, 0x9, 0xB, 0xC, 0xE, 0xD, 0xF);
SIMD_INLINE void ResizerByteBilinearInterpolateX2(const __m128i* alpha, __m128i* buffer)
{
__m128i src = _mm_shuffle_epi8(_mm_load_si128(buffer), K8_SHUFFLE_X2);
_mm_store_si128(buffer, _mm_maddubs_epi16(src, _mm_load_si128(alpha)));
}
template <> SIMD_INLINE void ResizerByteBilinearInterpolateX<2>(const __m128i* alpha, __m128i* buffer)
{
ResizerByteBilinearInterpolateX2(alpha + 0, buffer + 0);
ResizerByteBilinearInterpolateX2(alpha + 1, buffer + 1);
}
const __m128i K8_SHUFFLE_X3_00 = SIMD_MM_SETR_EPI8(0x0, 0x3, 0x1, 0x4, 0x2, 0x5, 0x6, 0x9, 0x7, 0xA, 0x8, 0xB, 0xC, 0xF, 0xD, -1);
const __m128i K8_SHUFFLE_X3_01 = SIMD_MM_SETR_EPI8(-1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, 0x0);
const __m128i K8_SHUFFLE_X3_10 = SIMD_MM_SETR_EPI8(0xE, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1);
const __m128i K8_SHUFFLE_X3_11 = SIMD_MM_SETR_EPI8(-1, 0x1, 0x2, 0x5, 0x3, 0x6, 0x4, 0x7, 0x8, 0xB, 0x9, 0xC, 0xA, 0xD, 0xE, -1);
const __m128i K8_SHUFFLE_X3_12 = SIMD_MM_SETR_EPI8(-1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, 0x1);
const __m128i K8_SHUFFLE_X3_21 = SIMD_MM_SETR_EPI8(0xF, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1);
const __m128i K8_SHUFFLE_X3_22 = SIMD_MM_SETR_EPI8(-1, 0x2, 0x0, 0x3, 0x4, 0x7, 0x5, 0x8, 0x6, 0x9, 0xA, 0xD, 0xB, 0xE, 0xC, 0xF);
template <> SIMD_INLINE void ResizerByteBilinearInterpolateX<3>(const __m128i* alpha, __m128i* buffer)
{
__m128i src[3], shuffled[3];
src[0] = _mm_load_si128(buffer + 0);
src[1] = _mm_load_si128(buffer + 1);
src[2] = _mm_load_si128(buffer + 2);
shuffled[0] = _mm_shuffle_epi8(src[0], K8_SHUFFLE_X3_00);
shuffled[0] = _mm_or_si128(shuffled[0], _mm_shuffle_epi8(src[1], K8_SHUFFLE_X3_01));
_mm_store_si128(buffer + 0, _mm_maddubs_epi16(shuffled[0], _mm_load_si128(alpha + 0)));
shuffled[1] = _mm_shuffle_epi8(src[0], K8_SHUFFLE_X3_10);
shuffled[1] = _mm_or_si128(shuffled[1], _mm_shuffle_epi8(src[1], K8_SHUFFLE_X3_11));
shuffled[1] = _mm_or_si128(shuffled[1], _mm_shuffle_epi8(src[2], K8_SHUFFLE_X3_12));
_mm_store_si128(buffer + 1, _mm_maddubs_epi16(shuffled[1], _mm_load_si128(alpha + 1)));
shuffled[2] = _mm_shuffle_epi8(src[1], K8_SHUFFLE_X3_21);
shuffled[2] = _mm_or_si128(shuffled[2], _mm_shuffle_epi8(src[2], K8_SHUFFLE_X3_22));
_mm_store_si128(buffer + 2, _mm_maddubs_epi16(shuffled[2], _mm_load_si128(alpha + 2)));
}
const __m128i K8_SHUFFLE_X4 = SIMD_MM_SETR_EPI8(0x0, 0x4, 0x1, 0x5, 0x2, 0x6, 0x3, 0x7, 0x8, 0xC, 0x9, 0xD, 0xA, 0xE, 0xB, 0xF);
SIMD_INLINE void ResizerByteBilinearInterpolateX4(const __m128i* alpha, __m128i* buffer)
{
__m128i src = _mm_shuffle_epi8(_mm_load_si128(buffer), K8_SHUFFLE_X4);
_mm_store_si128(buffer, _mm_maddubs_epi16(src, _mm_load_si128(alpha)));
}
template <> SIMD_INLINE void ResizerByteBilinearInterpolateX<4>(const __m128i* alpha, __m128i* buffer)
{
ResizerByteBilinearInterpolateX4(alpha + 0, buffer + 0);
ResizerByteBilinearInterpolateX4(alpha + 1, buffer + 1);
ResizerByteBilinearInterpolateX4(alpha + 2, buffer + 2);
ResizerByteBilinearInterpolateX4(alpha + 3, buffer + 3);
}
const __m128i K16_FRACTION_ROUND_TERM = SIMD_MM_SET1_EPI16(Base::BILINEAR_ROUND_TERM);
template<bool align> SIMD_INLINE __m128i ResizerByteBilinearInterpolateY(const __m128i* pbx0, const __m128i* pbx1, __m128i alpha[2])
{
__m128i sum = _mm_add_epi16(_mm_mullo_epi16(Load<align>(pbx0), alpha[0]), _mm_mullo_epi16(Load<align>(pbx1), alpha[1]));
return _mm_srli_epi16(_mm_add_epi16(sum, K16_FRACTION_ROUND_TERM), Base::BILINEAR_SHIFT);
}
template<bool align> SIMD_INLINE void ResizerByteBilinearInterpolateY(const uint8_t* bx0, const uint8_t* bx1, __m128i alpha[2], uint8_t* dst)
{
__m128i lo = ResizerByteBilinearInterpolateY<align>((__m128i*)bx0 + 0, (__m128i*)bx1 + 0, alpha);
__m128i hi = ResizerByteBilinearInterpolateY<align>((__m128i*)bx0 + 1, (__m128i*)bx1 + 1, alpha);
Store<false>((__m128i*)dst, _mm_packus_epi16(lo, hi));
}
template<size_t N> void ResizerByteBilinear::Run(const uint8_t* src, size_t srcStride, uint8_t* dst, size_t dstStride)
{
struct One { uint8_t val[N * 1]; };
struct Two { uint8_t val[N * 2]; };
size_t size = 2 * _param.dstW * N;
size_t aligned = AlignHi(size, DA) - DA;
const size_t step = A * N;
ptrdiff_t previous = -2;
__m128i a[2];
uint8_t* bx[2] = { _bx[0].data, _bx[1].data };
const uint8_t* ax = _ax.data;
const int32_t* ix = _ix.data;
size_t dstW = _param.dstW;
for (size_t yDst = 0; yDst < _param.dstH; yDst++, dst += dstStride)
{
a[0] = _mm_set1_epi16(int16_t(Base::FRACTION_RANGE - _ay[yDst]));
a[1] = _mm_set1_epi16(int16_t(_ay[yDst]));
ptrdiff_t sy = _iy[yDst];
int k = 0;
if (sy == previous)
k = 2;
else if (sy == previous + 1)
{
Swap(bx[0], bx[1]);
k = 1;
}
previous = sy;
for (; k < 2; k++)
{
Two* pb = (Two*)bx[k];
const One* psrc = (const One*)(src + (sy + k) * srcStride);
for (size_t x = 0; x < dstW; x++)
pb[x] = *(Two*)(psrc + ix[x]);
uint8_t* pbx = bx[k];
for (size_t i = 0; i < size; i += step)
ResizerByteBilinearInterpolateX<N>((__m128i*)(ax + i), (__m128i*)(pbx + i));
}
for (size_t ib = 0, id = 0; ib < aligned; ib += DA, id += A)
ResizerByteBilinearInterpolateY<true>(bx[0] + ib, bx[1] + ib, a, dst + id);
size_t i = size - DA;
ResizerByteBilinearInterpolateY<false>(bx[0] + i, bx[1] + i, a, dst + i / 2);
}
}
template <class Idx> SIMD_INLINE void ResizerByteBilinearLoadGrayInterpolated(const uint8_t* src, const Idx& index, const uint8_t* alpha, uint8_t* dst)
{
__m128i _src = _mm_loadu_si128((__m128i*)(src + index.src));
__m128i _shuffle = _mm_loadu_si128((__m128i*) & index.shuffle);
__m128i _alpha = _mm_loadu_si128((__m128i*)(alpha + index.dst));
_mm_storeu_si128((__m128i*)(dst + index.dst), _mm_maddubs_epi16(_mm_shuffle_epi8(_src, _shuffle), _alpha));
}
void ResizerByteBilinear::RunG(const uint8_t* src, size_t srcStride, uint8_t* dst, size_t dstStride)
{
size_t bufW = AlignHi(_param.dstW, A) * 2;
size_t size = 2 * _param.dstW;
size_t aligned = AlignHi(size, DA) - DA;
size_t blocks = _blocks;
ptrdiff_t previous = -2;
__m128i a[2];
uint8_t* bx[2] = { _bx[0].data, _bx[1].data };
const uint8_t* ax = _ax.data;
const Idx* ixg = _ixg.data;
for (size_t yDst = 0; yDst < _param.dstH; yDst++, dst += dstStride)
{
a[0] = _mm_set1_epi16(int16_t(Base::FRACTION_RANGE - _ay[yDst]));
a[1] = _mm_set1_epi16(int16_t(_ay[yDst]));
ptrdiff_t sy = _iy[yDst];
int k = 0;
if (sy == previous)
k = 2;
else if (sy == previous + 1)
{
Swap(bx[0], bx[1]);
k = 1;
}
previous = sy;
for (; k < 2; k++)
{
const uint8_t* psrc = src + (sy + k) * srcStride;
uint8_t* pdst = bx[k];
for (size_t i = 0; i < blocks; ++i)
ResizerByteBilinearLoadGrayInterpolated(psrc, ixg[i], ax, pdst);
}
for (size_t ib = 0, id = 0; ib < aligned; ib += DA, id += A)
ResizerByteBilinearInterpolateY<true>(bx[0] + ib, bx[1] + ib, a, dst + id);
size_t i = size - DA;
ResizerByteBilinearInterpolateY<false>(bx[0] + i, bx[1] + i, a, dst + i / 2);
}
}
void ResizerByteBilinear::Run(const uint8_t* src, size_t srcStride, uint8_t* dst, size_t dstStride)
{
assert(_param.dstW >= A);
EstimateParams();
switch (_param.channels)
{
case 1:
if (_blocks)
RunG(src, srcStride, dst, dstStride);
else
Run<1>(src, srcStride, dst, dstStride);
break;
case 2: Run<2>(src, srcStride, dst, dstStride); break;
case 3: Run<3>(src, srcStride, dst, dstStride); break;
case 4: Run<4>(src, srcStride, dst, dstStride); break;
default:
assert(0);
}
}
//-----------------------------------------------------------------------------------------
ResizerShortBilinear::ResizerShortBilinear(const ResParam& param)
: Base::ResizerShortBilinear(param)
{
}
template<size_t N> void ResizerShortBilinear::RunB(const uint16_t* src, size_t srcStride, uint16_t* dst, size_t dstStride)
{
size_t rs = _param.dstW * N;
float* pbx[2] = { _bx[0].data, _bx[1].data };
int32_t prev = -2;
size_t rs3 = AlignLoAny(rs - 1, 3);
size_t rs4 = AlignLo(rs, 4);
size_t rs8 = AlignLo(rs, 8);
__m128 _1 = _mm_set1_ps(1.0f);
for (size_t dy = 0; dy < _param.dstH; dy++, dst += dstStride)
{
float fy1 = _ay[dy];
float fy0 = 1.0f - fy1;
int32_t sy = _iy[dy];
int32_t k = 0;
if (sy == prev)
k = 2;
else if (sy == prev + 1)
{
Swap(pbx[0], pbx[1]);
k = 1;
}
prev = sy;
for (; k < 2; k++)
{
float* pb = pbx[k];
const uint16_t* ps = src + (sy + k) * srcStride;
size_t dx = 0;
if (N == 1)
{
for (; dx < rs4; dx += 4)
{
__m128 fx1 = _mm_loadu_ps(_ax.data + dx);
__m128 fx0 = _mm_sub_ps(_1, fx1);
_mm_storeu_ps(pb + dx, BilColS1(ps, _ix.data + dx, fx0, fx1));
}
}
if (N == 2)
{
for (; dx < rs4; dx += 4)
{
__m128 fx1 = _mm_loadu_ps(_ax.data + dx);
__m128 fx0 = _mm_sub_ps(_1, fx1);
_mm_storeu_ps(pb + dx, BilColS2(ps, _ix.data + dx, fx0, fx1));
}
}
if (N == 3)
{
for (; dx < rs3; dx += 3)
{
__m128 fx1 = _mm_loadu_ps(_ax.data + dx);
__m128 fx0 = _mm_sub_ps(_1, fx1);
_mm_storeu_ps(pb + dx, BilColS3(ps + _ix[dx], fx0, fx1));
}
}
if (N == 4)
{
for (; dx < rs4; dx += 4)
{
__m128 fx1 = _mm_loadu_ps(_ax.data + dx);
__m128 fx0 = _mm_sub_ps(_1, fx1);
_mm_storeu_ps(pb + dx, BilColS4(ps + _ix[dx], fx0, fx1));
}
}
for (; dx < rs; dx++)
{
int32_t sx = _ix[dx];
float fx = _ax[dx];
pb[dx] = ps[sx] * (1.0f - fx) + ps[sx + N] * fx;
}
}
size_t dx = 0;
__m128 _fy0 = _mm_set1_ps(fy0);
__m128 _fy1 = _mm_set1_ps(fy1);
for (; dx < rs8; dx += 8)
{
__m128 m00 = _mm_mul_ps(_mm_loadu_ps(pbx[0] + dx + 0), _fy0);
__m128 m01 = _mm_mul_ps(_mm_loadu_ps(pbx[1] + dx + 0), _fy1);
__m128i i0 = _mm_cvttps_epi32(_mm_add_ps(m00, m01));
__m128 m10 = _mm_mul_ps(_mm_loadu_ps(pbx[0] + dx + 4), _fy0);
__m128 m11 = _mm_mul_ps(_mm_loadu_ps(pbx[1] + dx + 4), _fy1);
__m128i i1 = _mm_cvttps_epi32(_mm_add_ps(m10, m11));
_mm_storeu_si128((__m128i*)(dst + dx), _mm_packus_epi32(i0, i1));
}
for (; dx < rs4; dx += 4)
{
__m128 m0 = _mm_mul_ps(_mm_loadu_ps(pbx[0] + dx), _fy0);
__m128 m1 = _mm_mul_ps(_mm_loadu_ps(pbx[1] + dx), _fy1);
__m128i i0 = _mm_cvttps_epi32(_mm_add_ps(m0, m1));
_mm_storel_epi64((__m128i*)(dst + dx), _mm_packus_epi32(i0, K_ZERO));
}
for (; dx < rs; dx++)
dst[dx] = Round(pbx[0][dx] * fy0 + pbx[1][dx] * fy1);
}
}
template<size_t N> void ResizerShortBilinear::RunS(const uint16_t* src, size_t srcStride, uint16_t* dst, size_t dstStride)
{
size_t rs = _param.dstW * N;
size_t rs3 = AlignLoAny(rs - 1, 3);
size_t rs6 = AlignLoAny(rs - 1, 6);
size_t rs4 = AlignLo(rs, 4);
size_t rs8 = AlignLo(rs, 8);
__m128 _1 = _mm_set1_ps(1.0f);
for (size_t dy = 0; dy < _param.dstH; dy++, dst += dstStride)
{
float fy1 = _ay[dy];
float fy0 = 1.0f - fy1;
int32_t sy = _iy[dy];
const uint16_t* ps0 = src + (sy + 0) * srcStride;
const uint16_t* ps1 = src + (sy + 1) * srcStride;
size_t dx = 0;
__m128 _fy0 = _mm_set1_ps(fy0);
__m128 _fy1 = _mm_set1_ps(fy1);
if (N == 1)
{
for (; dx < rs8; dx += 8)
{
__m128 fx01 = _mm_loadu_ps(_ax.data + dx + 0);
__m128 fx00 = _mm_sub_ps(_1, fx01);
__m128 m00 = _mm_mul_ps(BilColS1(ps0, _ix.data + dx + 0, fx00, fx01), _fy0);
__m128 m01 = _mm_mul_ps(BilColS1(ps1, _ix.data + dx + 0, fx00, fx01), _fy1);
__m128i i0 = _mm_cvttps_epi32(_mm_add_ps(m00, m01));
__m128 fx11 = _mm_loadu_ps(_ax.data + dx + 4);
__m128 fx10 = _mm_sub_ps(_1, fx11);
__m128 m10 = _mm_mul_ps(BilColS1(ps0, _ix.data + dx + 4, fx10, fx11), _fy0);
__m128 m11 = _mm_mul_ps(BilColS1(ps1, _ix.data + dx + 4, fx10, fx11), _fy1);
__m128i i1 = _mm_cvttps_epi32(_mm_add_ps(m10, m11));
_mm_storeu_si128((__m128i*)(dst + dx), _mm_packus_epi32(i0, i1));
}
for (; dx < rs4; dx += 4)
{
__m128 fx1 = _mm_loadu_ps(_ax.data + dx);
__m128 fx0 = _mm_sub_ps(_1, fx1);
__m128 m0 = _mm_mul_ps(BilColS1(ps0, _ix.data + dx, fx0, fx1), _fy0);
__m128 m1 = _mm_mul_ps(BilColS1(ps1, _ix.data + dx, fx0, fx1), _fy1);
__m128i i0 = _mm_cvttps_epi32(_mm_add_ps(m0, m1));
_mm_storel_epi64((__m128i*)(dst + dx), _mm_packus_epi32(i0, K_ZERO));
}
}
if (N == 2)
{
for (; dx < rs8; dx += 8)
{
__m128 fx01 = _mm_loadu_ps(_ax.data + dx + 0);
__m128 fx00 = _mm_sub_ps(_1, fx01);
__m128 m00 = _mm_mul_ps(BilColS2(ps0, _ix.data + dx + 0, fx00, fx01), _fy0);
__m128 m01 = _mm_mul_ps(BilColS2(ps1, _ix.data + dx + 0, fx00, fx01), _fy1);
__m128i i0 = _mm_cvttps_epi32(_mm_add_ps(m00, m01));
__m128 fx11 = _mm_loadu_ps(_ax.data + dx + 4);
__m128 fx10 = _mm_sub_ps(_1, fx11);
__m128 m10 = _mm_mul_ps(BilColS2(ps0, _ix.data + dx + 4, fx10, fx11), _fy0);
__m128 m11 = _mm_mul_ps(BilColS2(ps1, _ix.data + dx + 4, fx10, fx11), _fy1);
__m128i i1 = _mm_cvttps_epi32(_mm_add_ps(m10, m11));
_mm_storeu_si128((__m128i*)(dst + dx), _mm_packus_epi32(i0, i1));
}
for (; dx < rs4; dx += 4)
{
__m128 fx1 = _mm_loadu_ps(_ax.data + dx);
__m128 fx0 = _mm_sub_ps(_1, fx1);
__m128 m0 = _mm_mul_ps(BilColS2(ps0, _ix.data + dx, fx0, fx1), _fy0);
__m128 m1 = _mm_mul_ps(BilColS2(ps1, _ix.data + dx, fx0, fx1), _fy1);
__m128i i0 = _mm_cvttps_epi32(_mm_add_ps(m0, m1));
_mm_storel_epi64((__m128i*)(dst + dx), _mm_packus_epi32(i0, K_ZERO));
}
}
if (N == 3)
{
for (; dx < rs6; dx += 6)
{
__m128 fx01 = _mm_loadu_ps(_ax.data + dx + 0);
__m128 fx00 = _mm_sub_ps(_1, fx01);
__m128 m00 = _mm_mul_ps(BilColS3(ps0 + _ix[dx + 0], fx00, fx01), _fy0);
__m128 m01 = _mm_mul_ps(BilColS3(ps1 + _ix[dx + 0], fx00, fx01), _fy1);
__m128i i0 = _mm_cvttps_epi32(_mm_add_ps(m00, m01));
__m128 fx11 = _mm_loadu_ps(_ax.data + dx + 3);
__m128 fx10 = _mm_sub_ps(_1, fx11);
__m128 m10 = _mm_mul_ps(BilColS3(ps0 + _ix[dx + 3], fx10, fx11), _fy0);
__m128 m11 = _mm_mul_ps(BilColS3(ps1 + _ix[dx + 3], fx10, fx11), _fy1);
__m128i i1 = _mm_cvttps_epi32(_mm_add_ps(m10, m11));
_mm_storeu_si128((__m128i*)(dst + dx), _mm_shuffle_epi8(_mm_packus_epi32(i0, i1), RSB_3_P));
}
for (; dx < rs3; dx += 3)
{
__m128 fx1 = _mm_loadu_ps(_ax.data + dx);
__m128 fx0 = _mm_sub_ps(_1, fx1);
__m128 m0 = _mm_mul_ps(BilColS3(ps0 + _ix[dx], fx0, fx1), _fy0);
__m128 m1 = _mm_mul_ps(BilColS3(ps1 + _ix[dx], fx0, fx1), _fy1);
__m128i i0 = _mm_cvttps_epi32(_mm_add_ps(m0, m1));
_mm_storel_epi64((__m128i*)(dst + dx), _mm_packus_epi32(i0, K_ZERO));
}
}
if (N == 4)
{
for (; dx < rs8; dx += 8)
{
__m128 fx01 = _mm_loadu_ps(_ax.data + dx + 0);
__m128 fx00 = _mm_sub_ps(_1, fx01);
__m128 m00 = _mm_mul_ps(BilColS4(ps0 + _ix[dx + 0], fx00, fx01), _fy0);
__m128 m01 = _mm_mul_ps(BilColS4(ps1 + _ix[dx + 0], fx00, fx01), _fy1);
__m128i i0 = _mm_cvttps_epi32(_mm_add_ps(m00, m01));
__m128 fx11 = _mm_loadu_ps(_ax.data + dx + 4);
__m128 fx10 = _mm_sub_ps(_1, fx11);
__m128 m10 = _mm_mul_ps(BilColS4(ps0 + _ix[dx + 4], fx10, fx11), _fy0);
__m128 m11 = _mm_mul_ps(BilColS4(ps1 + _ix[dx + 4], fx10, fx11), _fy1);
__m128i i1 = _mm_cvttps_epi32(_mm_add_ps(m10, m11));
_mm_storeu_si128((__m128i*)(dst + dx), _mm_packus_epi32(i0, i1));
}
for (; dx < rs4; dx += 4)
{
__m128 fx1 = _mm_loadu_ps(_ax.data + dx);
__m128 fx0 = _mm_sub_ps(_1, fx1);
__m128 m0 = _mm_mul_ps(BilColS4(ps0 + _ix[dx], fx0, fx1), _fy0);
__m128 m1 = _mm_mul_ps(BilColS4(ps1 + _ix[dx], fx0, fx1), _fy1);
__m128i i0 = _mm_cvttps_epi32(_mm_add_ps(m0, m1));
_mm_storel_epi64((__m128i*)(dst + dx), _mm_packus_epi32(i0, K_ZERO));
}
}
for (; dx < rs; dx++)
{
int32_t sx = _ix[dx];
float fx1 = _ax[dx];
float fx0 = 1.0f - fx1;
float r0 = ps0[sx] * fx0 + ps0[sx + N] * fx1;
float r1 = ps1[sx] * fx0 + ps1[sx + N] * fx1;
dst[dx] = Round(r0 * fy0 + r1 * fy1);
}
}
}
void ResizerShortBilinear::Run(const uint16_t* src, size_t srcStride, uint16_t* dst, size_t dstStride)
{
bool sparse = _param.dstH * 2.0 <= _param.srcH;
switch (_param.channels)
{
case 1: sparse ? RunS<1>(src, srcStride, dst, dstStride) : RunB<1>(src, srcStride, dst, dstStride); return;
case 2: sparse ? RunS<2>(src, srcStride, dst, dstStride) : RunB<2>(src, srcStride, dst, dstStride); return;
case 3: sparse ? RunS<3>(src, srcStride, dst, dstStride) : RunB<3>(src, srcStride, dst, dstStride); return;
case 4: sparse ? RunS<4>(src, srcStride, dst, dstStride) : RunB<4>(src, srcStride, dst, dstStride); return;
default:
assert(0);
}
}
}
#endif
}
|
