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
|
/*
* 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/SimdResizer.h"
#include "Simd/SimdResizerCommon.h"
#include "Simd/SimdStore.h"
#include "Simd/SimdSet.h"
#include "Simd/SimdUpdate.h"
namespace Simd
{
#ifdef SIMD_AVX2_ENABLE
namespace Avx2
{
ResizerNearest::ResizerNearest(const ResParam& param)
: Sse41::ResizerNearest(param)
{
}
void ResizerNearest::EstimateParams()
{
if (_pixelSize)
return;
size_t pixelSize = _param.PixelSize();
if (pixelSize == 4 || pixelSize == 8 ||
(pixelSize == 3 && _param.dstW <= _param.srcW) ||
(pixelSize == 2 && _param.dstW * 2 <= _param.srcW))
Base::ResizerNearest::EstimateParams();
}
const __m256i K8_SHUFFLE_UVXX_TO_UV = SIMD_MM256_SETR_EPI8(
0x0, 0x1, 0x4, 0x5, 0x8, 0x9, 0xC, 0xD, -1, -1, -1, -1, -1, -1, -1, -1,
0x0, 0x1, 0x4, 0x5, 0x8, 0x9, 0xC, 0xD, -1, -1, -1, -1, -1, -1, -1, -1);
const __m256i K32_PERMUTE_UVXX_TO_UV = SIMD_MM256_SETR_EPI32(0x0, 0x1, 0x4, 0x5, -1, -1, -1, -1);
SIMD_INLINE void Gather8x2(const uint8_t* src, const int32_t* idx, uint8_t* dst)
{
__m256i _idx = _mm256_loadu_si256((__m256i*)idx);
__m256i uvxx = _mm256_i32gather_epi32((int32_t*)src, _idx, 1);
__m256i uv = _mm256_permutevar8x32_epi32(_mm256_shuffle_epi8(uvxx, K8_SHUFFLE_UVXX_TO_UV), K32_PERMUTE_UVXX_TO_UV);
_mm_storeu_si128((__m128i*)dst, _mm256_castsi256_si128(uv));
}
void ResizerNearest::Gather2(const uint8_t* src, size_t srcStride, uint8_t* dst, size_t dstStride)
{
size_t body = AlignLo(_param.dstW, 8);
size_t tail = _param.dstW - 8;
for (size_t dy = 0; dy < _param.dstH; dy++)
{
const uint8_t* srcRow = src + _iy[dy] * srcStride;
for (size_t dx = 0, offs = 0; dx < body; dx += 8, offs += 16)
Avx2::Gather8x2(srcRow, _ix.data + dx, dst + offs);
Avx2::Gather8x2(srcRow, _ix.data + tail, dst + tail * 2);
dst += dstStride;
}
}
SIMD_INLINE __m256i Gather8x3(const uint8_t* src, const int32_t* idx)
{
__m256i _idx = _mm256_loadu_si256((__m256i*)idx);
__m256i bgrx = _mm256_i32gather_epi32((int32_t*)src, _idx, 1);
return _mm256_permutevar8x32_epi32(_mm256_shuffle_epi8(bgrx, K8_SHUFFLE_BGRA_TO_BGR), K32_PERMUTE_BGRA_TO_BGR);
}
void ResizerNearest::Gather3(const uint8_t* src, size_t srcStride, uint8_t* dst, size_t dstStride)
{
size_t body = AlignLo(_param.dstW - 1, 8);
size_t tail = _param.dstW - 8;
for (size_t dy = 0; dy < _param.dstH; dy++)
{
const uint8_t* srcRow = src + _iy[dy] * srcStride;
for (size_t dx = 0, offs = 0; dx < body; dx +=8, offs += 24)
_mm256_storeu_si256((__m256i*)(dst + offs), Gather8x3(srcRow, _ix.data + dx));
Store24<false>(dst + tail * 3, Gather8x3(srcRow, _ix.data + tail));
dst += dstStride;
}
}
SIMD_INLINE void Gather8x4(const int32_t * src, const int32_t* idx, int32_t* dst)
{
__m256i _idx = _mm256_loadu_si256((__m256i*)idx);
__m256i val = _mm256_i32gather_epi32(src, _idx, 1);
_mm256_storeu_si256((__m256i*)dst, val);
}
void ResizerNearest::Gather4(const uint8_t* src, size_t srcStride, uint8_t* dst, size_t dstStride)
{
size_t body = AlignLo(_param.dstW, 8);
size_t tail = _param.dstW - 8;
for (size_t dy = 0; dy < _param.dstH; dy++)
{
const int32_t* srcRow = (int32_t*)(src + _iy[dy] * srcStride);
for (size_t dx = 0; dx < body; dx += 8)
Avx2::Gather8x4(srcRow, _ix.data + dx, (int32_t*)dst + dx);
Avx2::Gather8x4(srcRow, _ix.data + tail, (int32_t*)dst + tail);
dst += dstStride;
}
}
SIMD_INLINE void Gather4x8(const int64_t* src, const int32_t* idx, int64_t* dst)
{
__m128i _idx = _mm_loadu_si128((__m128i*)idx);
__m256i val = _mm256_i32gather_epi64((long long*)src, _idx, 1);
_mm256_storeu_si256((__m256i*)dst, val);
}
void ResizerNearest::Gather8(const uint8_t* src, size_t srcStride, uint8_t* dst, size_t dstStride)
{
size_t body = AlignLo(_param.dstW, 4);
size_t tail = _param.dstW - 4;
for (size_t dy = 0; dy < _param.dstH; dy++)
{
const int64_t* srcRow = (int64_t*)(src + _iy[dy] * srcStride);
for (size_t dx = 0; dx < body; dx += 4)
Avx2::Gather4x8(srcRow, _ix.data + dx, (int64_t*)dst + dx);
Avx2::Gather4x8(srcRow, _ix.data + tail, (int64_t*)dst + tail);
dst += dstStride;
}
}
void ResizerNearest::Run(const uint8_t* src, size_t srcStride, uint8_t* dst, size_t dstStride)
{
EstimateParams();
if (_pixelSize == 2 && _param.dstW * 2 <= _param.srcW)
Gather2(src, srcStride, dst, dstStride);
else if (_pixelSize == 3 && _param.dstW <= _param.srcW)
Gather3(src, srcStride, dst, dstStride);
else if (_pixelSize == 4)
Gather4(src, srcStride, dst, dstStride);
else if (_pixelSize == 8)
Gather8(src, srcStride, dst, dstStride);
else
Sse41::ResizerNearest::Run(src, srcStride, dst, dstStride);
}
}
#endif //SIMD_AVX2_ENABLE
}
|
