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
| #define CL_USE_DEPRECATED_OPENCL_1_2_APIS
#include <CL/cl.h>
#include <stdio.h>
#define CL_CHECK(_expr) \
do { cl_int _err = _expr; \
if (_err == CL_SUCCESS) break; \
fprintf(stderr, "OpenCL Error: '%s'returned %d!\n", #_expr, (int)_err); \
abort(); \
} while (0)
const char* programSource[] = {
"__kernel void simpleMultiply(__global float* outputC, ",
" __global float* inputA, ",
" __global float* inputB, ",
" int width){ ",
" int row = get_global_id(1); ",
" int col = get_global_id(0); ",
" float sum = 0.0f; ",
" for (int i = 0; i < width; i++) ",
" sum += inputA[row*width+i] * inputB[col*width+i]; ",
" outputC[row*width+col] = sum; ",
"} "
};
int main() {
/* Generation des matrices a multiplier*/
const int tailleMatrices = 4;
int HA = tailleMatrices;
int HB = tailleMatrices;
int WB = tailleMatrices;
int WA = tailleMatrices;
int datasize = tailleMatrices*tailleMatrices*sizeof(float);
float *MatriceA =(float*) malloc(tailleMatrices*tailleMatrices*sizeof(float));
for(int i=0;i<tailleMatrices;i++)
{
for(int j=0;j<tailleMatrices;j++)
{
MatriceA[i*tailleMatrices+j]=i+j;
}
}
float *MatriceB =(float*) malloc(tailleMatrices*tailleMatrices*sizeof(float));
for(int i=0;i<tailleMatrices;i++)
{
for(int j=0;j<tailleMatrices;j++)
{
MatriceB[i*tailleMatrices+j]=i+j;
}
}
float *MatriceResult = (float*)malloc(tailleMatrices*tailleMatrices*sizeof(float));
for(int i=0;i<tailleMatrices;i++)
{
for(int j=0;j<tailleMatrices;j++)
{
MatriceResult[i*tailleMatrices+j]=0;
}
}
cl_ulong time_start=0.0f, time_end=0.0f;
long total_time=0.0f;
cl_event event;
// Use this to check the output of each API call
cl_int status;
////////////////////////// STEP 1 //////////////////////////////////////////
cl_uint numPlatforms = 0;
cl_platform_id *platforms = NULL;
// Use clGetPlatformIDs() to retrieve the number of platforms
status = clGetPlatformIDs(0, NULL,&numPlatforms);
// Allocate enough space for each platform
platforms = (cl_platform_id*)malloc(numPlatforms * sizeof(cl_platform_id));
// Fill in platforms with clGetPlatformIDs()
status = clGetPlatformIDs(numPlatforms, platforms, NULL);
///////////////////////// STEP 2 //////////////////////////////////////////
cl_uint numDevices = 0;
cl_device_id *devices = NULL;
// Use clGetDeviceIDs() to retrieve the number of devices present
status = clGetDeviceIDs(platforms[0],CL_DEVICE_TYPE_ALL,0, NULL, &numDevices);
// Allocate enough space for each device
devices = (cl_device_id*)malloc(numDevices *sizeof(cl_device_id));
// Fill in devices with clGetDeviceIDs()
status = clGetDeviceIDs(platforms[0],CL_DEVICE_TYPE_ALL,numDevices, devices,NULL);
////////////////////////// STEP 3 //////////////////////////////////////////
cl_context context = NULL;
// Create a context using clCreateContext() and associate it with the devices
context = clCreateContext(NULL, numDevices,devices, NULL, NULL, &status);
////////////////////////// STEP 4 //////////////////////////////////////////
cl_command_queue cmdQueue;
// Create a command queue using
// clCreateCommandQueue(), and associate it with
// the device you want to execute on
cmdQueue = clCreateCommandQueue(context, devices[0], CL_QUEUE_PROFILING_ENABLE, &status); // Add CL_QUEUE_PROFILING_ENABLE
////////////////////////// STEP 5 //////////////////////////////////////////
// Use clCreateBuffer() to create buffer objects
// that will contain the data from the host arrays
cl_mem bufferA = clCreateBuffer(context,CL_MEM_READ_ONLY, WA*HA*sizeof(float), NULL, &status);
cl_mem bufferB = clCreateBuffer(context,CL_MEM_READ_ONLY, WB*HB*sizeof(float), NULL, &status);
cl_mem bufferResult= clCreateBuffer(context, CL_MEM_WRITE_ONLY, datasize, NULL, &status);
////////////////////////// STEP 6 //////////////////////////////////////////
// Use clEnqueueWriteBuffer() to write input array A
status = clEnqueueWriteBuffer(cmdQueue,bufferA, CL_TRUE, 0, WA*HA*sizeof(float), (void *)MatriceA, 0, NULL, NULL);
status = clEnqueueWriteBuffer(cmdQueue,bufferB, CL_TRUE, 0, WB*HB*sizeof(float), (void *)MatriceB, 0, NULL, NULL);
////////////////////////// STEP 7 //////////////////////////////////////////
// Create a program using
// clCreateProgramWithSource()
cl_program program = clCreateProgramWithSource(context, sizeof(programSource)/sizeof(*programSource), programSource, NULL, &status);
// Build (compile) the program for the devices with clBuildProgram()
status = clBuildProgram(program, numDevices, devices, NULL, NULL, NULL);
/*
2. Ajouter des appels à clGetProgramBuildInfo et tester la compilation dynamique en introduisant des erreurs dans le kernel
*/
for(unsigned int i=0;i<numDevices;i++)
{
size_t len;
clGetProgramBuildInfo(program, devices[i], CL_PROGRAM_BUILD_LOG, 0, NULL, &len);
if(len>2)
{
char *buffer = (char *)malloc(len);
clGetProgramBuildInfo(program, devices[i], CL_PROGRAM_BUILD_LOG, len, buffer, NULL);
printf("[*] Error kernel %d: %s\n",i,buffer);
}
}
// Fin question 2
////////////////////////// STEP 8 //////////////////////////////////////////
cl_kernel kernel = NULL;
// Use clCreateKernel() to create a kernel from the vector addition function (named "vecadd")
kernel = clCreateKernel(program, "simpleMultiply", &status);
////////////////////////// STEP 9 //////////////////////////////////////////
// Associate the input and output buffers with the kernel using clSetKernelArg()
status = clSetKernelArg(kernel, 0, sizeof(cl_mem), &bufferResult);
status |= clSetKernelArg(kernel, 1, sizeof(cl_mem), &bufferA);
status |= clSetKernelArg(kernel, 2, sizeof(cl_mem), &bufferB);
status |= clSetKernelArg(kernel, 3, sizeof(tailleMatrices), &tailleMatrices);
////////////////////////// STEP 10 //////////////////////////////////////////
// Define an index space (global work size) of work
// items for execution. A workgroup size (local work
// size) is not required, but can be used.
size_t globalWorkSize[2];
// There are 'elements' work-items
globalWorkSize[0] = WB;
globalWorkSize[1] = HA;
////////////////////////// STEP 11 //////////////////////////////////////////
// Execute the kernel by using clEnqueueNDRangeKernel().
// 'globalWorkSize' is the 1D dimension of the work-items
/* question 3 */
status = clEnqueueNDRangeKernel(cmdQueue,kernel, 2, NULL, globalWorkSize,NULL, 0, NULL, &event); // add &event
clGetEventProfilingInfo(event, CL_PROFILING_COMMAND_START, sizeof(time_start), &time_start, NULL);
clGetEventProfilingInfo(event, CL_PROFILING_COMMAND_END, sizeof(time_end), &time_end, NULL);
total_time = time_end - time_start;
printf("\nExecution time in milliseconds = %0.3f ms\n", (total_time / 1000000.0) );
/* fin question 3 */
clEnqueueReadBuffer(cmdQueue, bufferResult, CL_TRUE, 0, datasize, MatriceResult, 0, NULL, NULL);
for(int i=0;i<tailleMatrices;i++)
{
for(int j=0;j<tailleMatrices;j++)
{
printf("%f -",MatriceResult[i*tailleMatrices+j]);
}
printf("\n");
}
////////////////////////// STEP 13 //////////////////////////////////////////
// Free OpenCL ressources
clReleaseKernel(kernel);
clReleaseProgram(program);
clReleaseCommandQueue(cmdQueue);
clReleaseMemObject(bufferA);
clReleaseMemObject(bufferB);
clReleaseContext(context);
// Free host ressources
free(MatriceA);
free(MatriceB);
free(platforms);
free(devices);
} |
