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
|
//
// Copyright (c) 2013-2014 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.
//
// blocklayout.cpp:
// Implementation for block layout classes and methods.
//
#include "common/blocklayout.h"
#include "common/shadervars.h"
#include "common/mathutil.h"
#include "common/utilities.h"
namespace sh
{
BlockLayoutEncoder::BlockLayoutEncoder()
: mCurrentOffset(0)
{
}
void BlockLayoutEncoder::encodeInterfaceBlockFields(const std::vector<InterfaceBlockField> &fields)
{
for (unsigned int fieldIndex = 0; fieldIndex < fields.size(); fieldIndex++)
{
const InterfaceBlockField &variable = fields[fieldIndex];
if (variable.fields.size() > 0)
{
const unsigned int elementCount = std::max(1u, variable.arraySize);
for (unsigned int elementIndex = 0; elementIndex < elementCount; elementIndex++)
{
enterAggregateType();
encodeInterfaceBlockFields(variable.fields);
exitAggregateType();
}
}
else
{
encodeInterfaceBlockField(variable);
}
}
}
BlockMemberInfo BlockLayoutEncoder::encodeInterfaceBlockField(const InterfaceBlockField &field)
{
int arrayStride;
int matrixStride;
ASSERT(field.fields.empty());
getBlockLayoutInfo(field.type, field.arraySize, field.isRowMajorMatrix, &arrayStride, &matrixStride);
const BlockMemberInfo memberInfo(mCurrentOffset * BytesPerComponent, arrayStride * BytesPerComponent, matrixStride * BytesPerComponent, field.isRowMajorMatrix);
advanceOffset(field.type, field.arraySize, field.isRowMajorMatrix, arrayStride, matrixStride);
return memberInfo;
}
void BlockLayoutEncoder::encodeType(GLenum type, unsigned int arraySize, bool isRowMajorMatrix)
{
int arrayStride;
int matrixStride;
getBlockLayoutInfo(type, arraySize, isRowMajorMatrix, &arrayStride, &matrixStride);
const BlockMemberInfo memberInfo(mCurrentOffset * BytesPerComponent, arrayStride * BytesPerComponent, matrixStride * BytesPerComponent, isRowMajorMatrix);
advanceOffset(type, arraySize, isRowMajorMatrix, arrayStride, matrixStride);
}
void BlockLayoutEncoder::nextRegister()
{
mCurrentOffset = rx::roundUp<size_t>(mCurrentOffset, ComponentsPerRegister);
}
Std140BlockEncoder::Std140BlockEncoder()
{
}
void Std140BlockEncoder::enterAggregateType()
{
nextRegister();
}
void Std140BlockEncoder::exitAggregateType()
{
nextRegister();
}
void Std140BlockEncoder::getBlockLayoutInfo(GLenum type, unsigned int arraySize, bool isRowMajorMatrix, int *arrayStrideOut, int *matrixStrideOut)
{
// We assume we are only dealing with 4 byte components (no doubles or half-words currently)
ASSERT(gl::VariableComponentSize(gl::VariableComponentType(type)) == BytesPerComponent);
size_t baseAlignment = 0;
int matrixStride = 0;
int arrayStride = 0;
if (gl::IsMatrixType(type))
{
baseAlignment = ComponentsPerRegister;
matrixStride = ComponentsPerRegister;
if (arraySize > 0)
{
const int numRegisters = gl::MatrixRegisterCount(type, isRowMajorMatrix);
arrayStride = ComponentsPerRegister * numRegisters;
}
}
else if (arraySize > 0)
{
baseAlignment = ComponentsPerRegister;
arrayStride = ComponentsPerRegister;
}
else
{
const int numComponents = gl::VariableComponentCount(type);
baseAlignment = (numComponents == 3 ? 4u : static_cast<size_t>(numComponents));
}
mCurrentOffset = rx::roundUp(mCurrentOffset, baseAlignment);
*matrixStrideOut = matrixStride;
*arrayStrideOut = arrayStride;
}
void Std140BlockEncoder::advanceOffset(GLenum type, unsigned int arraySize, bool isRowMajorMatrix, int arrayStride, int matrixStride)
{
if (arraySize > 0)
{
mCurrentOffset += arrayStride * arraySize;
}
else if (gl::IsMatrixType(type))
{
ASSERT(matrixStride == ComponentsPerRegister);
const int numRegisters = gl::MatrixRegisterCount(type, isRowMajorMatrix);
mCurrentOffset += ComponentsPerRegister * numRegisters;
}
else
{
mCurrentOffset += gl::VariableComponentCount(type);
}
}
HLSLBlockEncoder::HLSLBlockEncoder(HLSLBlockEncoderStrategy strategy)
: mEncoderStrategy(strategy)
{
}
void HLSLBlockEncoder::enterAggregateType()
{
nextRegister();
}
void HLSLBlockEncoder::exitAggregateType()
{
}
void HLSLBlockEncoder::getBlockLayoutInfo(GLenum type, unsigned int arraySize, bool isRowMajorMatrix, int *arrayStrideOut, int *matrixStrideOut)
{
// We assume we are only dealing with 4 byte components (no doubles or half-words currently)
ASSERT(gl::VariableComponentSize(gl::VariableComponentType(type)) == BytesPerComponent);
int matrixStride = 0;
int arrayStride = 0;
// if variables are not to be packed, or we're about to
// pack a matrix or array, skip to the start of the next
// register
if (!isPacked() ||
gl::IsMatrixType(type) ||
arraySize > 0)
{
nextRegister();
}
if (gl::IsMatrixType(type))
{
matrixStride = ComponentsPerRegister;
if (arraySize > 0)
{
const int numRegisters = gl::MatrixRegisterCount(type, isRowMajorMatrix);
arrayStride = ComponentsPerRegister * numRegisters;
}
}
else if (arraySize > 0)
{
arrayStride = ComponentsPerRegister;
}
else if (isPacked())
{
int numComponents = gl::VariableComponentCount(type);
if ((numComponents + (mCurrentOffset % ComponentsPerRegister)) > ComponentsPerRegister)
{
nextRegister();
}
}
*matrixStrideOut = matrixStride;
*arrayStrideOut = arrayStride;
}
void HLSLBlockEncoder::advanceOffset(GLenum type, unsigned int arraySize, bool isRowMajorMatrix, int arrayStride, int matrixStride)
{
if (arraySize > 0)
{
mCurrentOffset += arrayStride * (arraySize - 1);
}
if (gl::IsMatrixType(type))
{
ASSERT(matrixStride == ComponentsPerRegister);
const int numRegisters = gl::MatrixRegisterCount(type, isRowMajorMatrix);
const int numComponents = gl::MatrixComponentCount(type, isRowMajorMatrix);
mCurrentOffset += ComponentsPerRegister * (numRegisters - 1);
mCurrentOffset += numComponents;
}
else if (isPacked())
{
mCurrentOffset += gl::VariableComponentCount(type);
}
else
{
mCurrentOffset += ComponentsPerRegister;
}
}
void HLSLBlockEncoder::skipRegisters(unsigned int numRegisters)
{
mCurrentOffset += (numRegisters * ComponentsPerRegister);
}
HLSLBlockEncoder::HLSLBlockEncoderStrategy HLSLBlockEncoder::GetStrategyFor(ShShaderOutput outputType)
{
switch (outputType)
{
case SH_HLSL9_OUTPUT: return ENCODE_LOOSE;
case SH_HLSL11_OUTPUT: return ENCODE_PACKED;
default: UNREACHABLE(); return ENCODE_PACKED;
}
}
template <class ShaderVarType>
void HLSLVariableRegisterCount(const ShaderVarType &variable, HLSLBlockEncoder *encoder)
{
if (variable.isStruct())
{
for (size_t arrayElement = 0; arrayElement < variable.elementCount(); arrayElement++)
{
encoder->enterAggregateType();
for (size_t fieldIndex = 0; fieldIndex < variable.fields.size(); fieldIndex++)
{
HLSLVariableRegisterCount(variable.fields[fieldIndex], encoder);
}
encoder->exitAggregateType();
}
}
else
{
// We operate only on varyings and uniforms, which do not have matrix layout qualifiers
encoder->encodeType(variable.type, variable.arraySize, false);
}
}
unsigned int HLSLVariableRegisterCount(const Varying &variable)
{
HLSLBlockEncoder encoder(HLSLBlockEncoder::ENCODE_PACKED);
HLSLVariableRegisterCount(variable, &encoder);
const size_t registerBytes = (encoder.BytesPerComponent * encoder.ComponentsPerRegister);
return static_cast<unsigned int>(rx::roundUp<size_t>(encoder.getBlockSize(), registerBytes) / registerBytes);
}
unsigned int HLSLVariableRegisterCount(const Uniform &variable, ShShaderOutput outputType)
{
HLSLBlockEncoder encoder(HLSLBlockEncoder::GetStrategyFor(outputType));
HLSLVariableRegisterCount(variable, &encoder);
const size_t registerBytes = (encoder.BytesPerComponent * encoder.ComponentsPerRegister);
return static_cast<unsigned int>(rx::roundUp<size_t>(encoder.getBlockSize(), registerBytes) / registerBytes);
}
}
|
