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
|
//
// Copyright 2014 The ANGLE 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.
//
#include "common/mathutil.h"
#include <string.h>
namespace angle
{
namespace priv
{
template <typename T>
inline T *OffsetDataPointer(uint8_t *data, size_t y, size_t z, size_t rowPitch, size_t depthPitch)
{
return reinterpret_cast<T*>(data + (y * rowPitch) + (z * depthPitch));
}
template <typename T>
inline const T *OffsetDataPointer(const uint8_t *data, size_t y, size_t z, size_t rowPitch, size_t depthPitch)
{
return reinterpret_cast<const T*>(data + (y * rowPitch) + (z * depthPitch));
}
} // namespace priv
template <typename type, size_t componentCount>
inline void LoadToNative(size_t width, size_t height, size_t depth,
const uint8_t *input, size_t inputRowPitch, size_t inputDepthPitch,
uint8_t *output, size_t outputRowPitch, size_t outputDepthPitch)
{
const size_t rowSize = width * sizeof(type) * componentCount;
const size_t layerSize = rowSize * height;
const size_t imageSize = layerSize * depth;
if (layerSize == inputDepthPitch && layerSize == outputDepthPitch)
{
ASSERT(rowSize == inputRowPitch && rowSize == outputRowPitch);
memcpy(output, input, imageSize);
}
else if (rowSize == inputRowPitch && rowSize == outputRowPitch)
{
for (size_t z = 0; z < depth; z++)
{
const type *source = priv::OffsetDataPointer<type>(input, 0, z, inputRowPitch, inputDepthPitch);
type *dest = priv::OffsetDataPointer<type>(output, 0, z, outputRowPitch, outputDepthPitch);
memcpy(dest, source, layerSize);
}
}
else
{
for (size_t z = 0; z < depth; z++)
{
for (size_t y = 0; y < height; y++)
{
const type *source = priv::OffsetDataPointer<type>(input, y, z, inputRowPitch, inputDepthPitch);
type *dest = priv::OffsetDataPointer<type>(output, y, z, outputRowPitch, outputDepthPitch);
memcpy(dest, source, width * sizeof(type) * componentCount);
}
}
}
}
template <typename type, uint32_t fourthComponentBits>
inline void LoadToNative3To4(size_t width, size_t height, size_t depth,
const uint8_t *input, size_t inputRowPitch, size_t inputDepthPitch,
uint8_t *output, size_t outputRowPitch, size_t outputDepthPitch)
{
const type fourthValue = gl::bitCast<type>(fourthComponentBits);
for (size_t z = 0; z < depth; z++)
{
for (size_t y = 0; y < height; y++)
{
const type *source = priv::OffsetDataPointer<type>(input, y, z, inputRowPitch, inputDepthPitch);
type *dest = priv::OffsetDataPointer<type>(output, y, z, outputRowPitch, outputDepthPitch);
for (size_t x = 0; x < width; x++)
{
memcpy(&dest[x * 4], &source[x * 3], sizeof(type) * 3);
dest[x * 4 + 3] = fourthValue;
}
}
}
}
template <size_t componentCount>
inline void Load32FTo16F(size_t width, size_t height, size_t depth,
const uint8_t *input, size_t inputRowPitch, size_t inputDepthPitch,
uint8_t *output, size_t outputRowPitch, size_t outputDepthPitch)
{
const size_t elementWidth = componentCount * width;
for (size_t z = 0; z < depth; z++)
{
for (size_t y = 0; y < height; y++)
{
const float *source = priv::OffsetDataPointer<float>(input, y, z, inputRowPitch, inputDepthPitch);
uint16_t *dest = priv::OffsetDataPointer<uint16_t>(output, y, z, outputRowPitch, outputDepthPitch);
for (size_t x = 0; x < elementWidth; x++)
{
dest[x] = gl::float32ToFloat16(source[x]);
}
}
}
}
template <size_t blockWidth, size_t blockHeight, size_t blockDepth, size_t blockSize>
inline void LoadCompressedToNative(size_t width, size_t height, size_t depth,
const uint8_t *input, size_t inputRowPitch, size_t inputDepthPitch,
uint8_t *output, size_t outputRowPitch, size_t outputDepthPitch)
{
const size_t columns = (width + (blockWidth - 1)) / blockWidth;
const size_t rows = (height + (blockHeight - 1)) / blockHeight;
const size_t layers = (depth + (blockDepth - 1)) / blockDepth;
for (size_t z = 0; z < layers; ++z)
{
for (size_t y = 0; y < rows; ++y)
{
const uint8_t *source = priv::OffsetDataPointer<uint8_t>(input, y, z, inputRowPitch, inputDepthPitch);
uint8_t *dest = priv::OffsetDataPointer<uint8_t>(output, y, z, outputRowPitch, outputDepthPitch);
memcpy(dest, source, columns * blockSize);
}
}
}
template <typename type, uint32_t firstBits, uint32_t secondBits, uint32_t thirdBits, uint32_t fourthBits>
inline void Initialize4ComponentData(size_t width, size_t height, size_t depth,
uint8_t *output, size_t outputRowPitch, size_t outputDepthPitch)
{
type writeValues[4] =
{
gl::bitCast<type>(firstBits),
gl::bitCast<type>(secondBits),
gl::bitCast<type>(thirdBits),
gl::bitCast<type>(fourthBits),
};
for (size_t z = 0; z < depth; z++)
{
for (size_t y = 0; y < height; y++)
{
type *destRow = priv::OffsetDataPointer<type>(output, y, z, outputRowPitch, outputDepthPitch);
for (size_t x = 0; x < width; x++)
{
type* destPixel = destRow + x * 4;
// This could potentially be optimized by generating an entire row of initialization
// data and copying row by row instead of pixel by pixel.
memcpy(destPixel, writeValues, sizeof(type) * 4);
}
}
}
}
template <size_t blockWidth, size_t blockHeight>
inline void LoadASTCToRGBA8(size_t width,
size_t height,
size_t depth,
const uint8_t *input,
size_t inputRowPitch,
size_t inputDepthPitch,
uint8_t *output,
size_t outputRowPitch,
size_t outputDepthPitch)
{
LoadASTCToRGBA8Inner(width, height, depth, blockWidth, blockHeight, input, inputRowPitch,
inputDepthPitch, output, outputRowPitch, outputDepthPitch);
}
template <uint32_t indexBits, uint32_t redBlueBits, uint32_t greenBits, uint32_t alphaBits>
inline void LoadPalettedToRGBA8(size_t width,
size_t height,
size_t depth,
const uint8_t *input,
size_t inputRowPitch,
size_t inputDepthPitch,
uint8_t *output,
size_t outputRowPitch,
size_t outputDepthPitch)
{
static_assert(indexBits == 4 || indexBits == 8);
static_assert(redBlueBits == 4 || redBlueBits == 5 || redBlueBits == 8);
static_assert(greenBits == 4 || greenBits == 5 || greenBits == 6 || greenBits == 8);
static_assert(alphaBits == 0 || alphaBits == 1 || alphaBits == 4 || alphaBits == 8);
constexpr uint32_t colorBits = 2 * redBlueBits + greenBits + alphaBits;
static_assert(colorBits == 16 || colorBits == 24 || colorBits == 32);
LoadPalettedToRGBA8Impl(width, height, depth,
indexBits, redBlueBits, greenBits, alphaBits,
input, inputRowPitch, inputDepthPitch,
output, outputRowPitch, outputDepthPitch);
}
} // namespace angle
|
