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
|
/*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public License
* as published by the Free Software Foundation; either version 2
* of the License, or (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software Foundation,
* Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
*
* Copyright 2011, Blender Foundation.
*/
#include "COM_OpenCLDevice.h"
#include "COM_ExecutionGroup.h"
#include "COM_ReadBufferOperation.h"
namespace blender::compositor {
enum COM_VendorID { NVIDIA = 0x10DE, AMD = 0x1002 };
const cl_image_format IMAGE_FORMAT_COLOR = {
CL_RGBA,
CL_FLOAT,
};
const cl_image_format IMAGE_FORMAT_VECTOR = {
CL_RGB,
CL_FLOAT,
};
const cl_image_format IMAGE_FORMAT_VALUE = {
CL_R,
CL_FLOAT,
};
OpenCLDevice::OpenCLDevice(cl_context context,
cl_device_id device,
cl_program program,
cl_int vendorId)
{
this->m_device = device;
this->m_context = context;
this->m_program = program;
this->m_queue = nullptr;
this->m_vendorID = vendorId;
cl_int error;
this->m_queue = clCreateCommandQueue(this->m_context, this->m_device, 0, &error);
}
OpenCLDevice::OpenCLDevice(OpenCLDevice &&other) noexcept
: m_context(other.m_context),
m_device(other.m_device),
m_program(other.m_program),
m_queue(other.m_queue),
m_vendorID(other.m_vendorID)
{
other.m_queue = nullptr;
}
OpenCLDevice::~OpenCLDevice()
{
if (this->m_queue) {
clReleaseCommandQueue(this->m_queue);
}
}
void OpenCLDevice::execute(WorkPackage *work_package)
{
const unsigned int chunkNumber = work_package->chunk_number;
ExecutionGroup *executionGroup = work_package->execution_group;
MemoryBuffer **inputBuffers = executionGroup->getInputBuffersOpenCL(chunkNumber);
MemoryBuffer *outputBuffer = executionGroup->allocateOutputBuffer(work_package->rect);
executionGroup->getOutputOperation()->executeOpenCLRegion(
this, &work_package->rect, chunkNumber, inputBuffers, outputBuffer);
delete outputBuffer;
executionGroup->finalizeChunkExecution(chunkNumber, inputBuffers);
}
cl_mem OpenCLDevice::COM_clAttachMemoryBufferToKernelParameter(cl_kernel kernel,
int parameterIndex,
int offsetIndex,
std::list<cl_mem> *cleanup,
MemoryBuffer **inputMemoryBuffers,
SocketReader *reader)
{
return COM_clAttachMemoryBufferToKernelParameter(kernel,
parameterIndex,
offsetIndex,
cleanup,
inputMemoryBuffers,
(ReadBufferOperation *)reader);
}
const cl_image_format *OpenCLDevice::determineImageFormat(MemoryBuffer *memoryBuffer)
{
switch (memoryBuffer->get_num_channels()) {
case 1:
return &IMAGE_FORMAT_VALUE;
break;
case 3:
return &IMAGE_FORMAT_VECTOR;
break;
case 4:
return &IMAGE_FORMAT_COLOR;
break;
default:
BLI_assert_msg(0, "Unsupported num_channels.");
}
return &IMAGE_FORMAT_COLOR;
}
cl_mem OpenCLDevice::COM_clAttachMemoryBufferToKernelParameter(cl_kernel kernel,
int parameterIndex,
int offsetIndex,
std::list<cl_mem> *cleanup,
MemoryBuffer **inputMemoryBuffers,
ReadBufferOperation *reader)
{
cl_int error;
MemoryBuffer *result = reader->getInputMemoryBuffer(inputMemoryBuffers);
const cl_image_format *imageFormat = determineImageFormat(result);
cl_mem clBuffer = clCreateImage2D(this->m_context,
CL_MEM_READ_ONLY | CL_MEM_COPY_HOST_PTR,
imageFormat,
result->getWidth(),
result->getHeight(),
0,
result->getBuffer(),
&error);
if (error != CL_SUCCESS) {
printf("CLERROR[%d]: %s\n", error, clewErrorString(error));
}
if (error == CL_SUCCESS) {
cleanup->push_back(clBuffer);
}
error = clSetKernelArg(kernel, parameterIndex, sizeof(cl_mem), &clBuffer);
if (error != CL_SUCCESS) {
printf("CLERROR[%d]: %s\n", error, clewErrorString(error));
}
COM_clAttachMemoryBufferOffsetToKernelParameter(kernel, offsetIndex, result);
return clBuffer;
}
void OpenCLDevice::COM_clAttachMemoryBufferOffsetToKernelParameter(cl_kernel kernel,
int offsetIndex,
MemoryBuffer *memoryBuffer)
{
if (offsetIndex != -1) {
cl_int error;
const rcti &rect = memoryBuffer->get_rect();
cl_int2 offset = {{rect.xmin, rect.ymin}};
error = clSetKernelArg(kernel, offsetIndex, sizeof(cl_int2), &offset);
if (error != CL_SUCCESS) {
printf("CLERROR[%d]: %s\n", error, clewErrorString(error));
}
}
}
void OpenCLDevice::COM_clAttachSizeToKernelParameter(cl_kernel kernel,
int offsetIndex,
NodeOperation *operation)
{
if (offsetIndex != -1) {
cl_int error;
cl_int2 offset = {{(cl_int)operation->getWidth(), (cl_int)operation->getHeight()}};
error = clSetKernelArg(kernel, offsetIndex, sizeof(cl_int2), &offset);
if (error != CL_SUCCESS) {
printf("CLERROR[%d]: %s\n", error, clewErrorString(error));
}
}
}
void OpenCLDevice::COM_clAttachOutputMemoryBufferToKernelParameter(cl_kernel kernel,
int parameterIndex,
cl_mem clOutputMemoryBuffer)
{
cl_int error;
error = clSetKernelArg(kernel, parameterIndex, sizeof(cl_mem), &clOutputMemoryBuffer);
if (error != CL_SUCCESS) {
printf("CLERROR[%d]: %s\n", error, clewErrorString(error));
}
}
void OpenCLDevice::COM_clEnqueueRange(cl_kernel kernel, MemoryBuffer *outputMemoryBuffer)
{
cl_int error;
const size_t size[] = {
(size_t)outputMemoryBuffer->getWidth(),
(size_t)outputMemoryBuffer->getHeight(),
};
error = clEnqueueNDRangeKernel(
this->m_queue, kernel, 2, nullptr, size, nullptr, 0, nullptr, nullptr);
if (error != CL_SUCCESS) {
printf("CLERROR[%d]: %s\n", error, clewErrorString(error));
}
}
void OpenCLDevice::COM_clEnqueueRange(cl_kernel kernel,
MemoryBuffer *outputMemoryBuffer,
int offsetIndex,
NodeOperation *operation)
{
cl_int error;
const int width = outputMemoryBuffer->getWidth();
const int height = outputMemoryBuffer->getHeight();
int offsetx;
int offsety;
int localSize = 1024;
size_t size[2];
cl_int2 offset;
if (this->m_vendorID == NVIDIA) {
localSize = 32;
}
bool breaked = false;
for (offsety = 0; offsety < height && (!breaked); offsety += localSize) {
offset.s[1] = offsety;
if (offsety + localSize < height) {
size[1] = localSize;
}
else {
size[1] = height - offsety;
}
for (offsetx = 0; offsetx < width && (!breaked); offsetx += localSize) {
if (offsetx + localSize < width) {
size[0] = localSize;
}
else {
size[0] = width - offsetx;
}
offset.s[0] = offsetx;
error = clSetKernelArg(kernel, offsetIndex, sizeof(cl_int2), &offset);
if (error != CL_SUCCESS) {
printf("CLERROR[%d]: %s\n", error, clewErrorString(error));
}
error = clEnqueueNDRangeKernel(
this->m_queue, kernel, 2, nullptr, size, nullptr, 0, nullptr, nullptr);
if (error != CL_SUCCESS) {
printf("CLERROR[%d]: %s\n", error, clewErrorString(error));
}
clFlush(this->m_queue);
if (operation->isBraked()) {
breaked = false;
}
}
}
}
cl_kernel OpenCLDevice::COM_clCreateKernel(const char *kernelname,
std::list<cl_kernel> *clKernelsToCleanUp)
{
cl_int error;
cl_kernel kernel = clCreateKernel(this->m_program, kernelname, &error);
if (error != CL_SUCCESS) {
printf("CLERROR[%d]: %s\n", error, clewErrorString(error));
}
else {
if (clKernelsToCleanUp) {
clKernelsToCleanUp->push_back(kernel);
}
}
return kernel;
}
} // namespace blender::compositor
|