blob: 40dbe5d7e2d235e7474c6152edbdb7a0a6e83f91 (
plain)
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
|
#include <iostream>
#include "file1.h"
#include "file2.h"
int file4_launch_kernel(int x);
int file5_launch_kernel(int x);
int choose_cuda_device()
{
int nDevices = 0;
cudaError_t err = cudaGetDeviceCount(&nDevices);
if (err != cudaSuccess) {
std::cerr << "Failed to retrieve the number of CUDA enabled devices"
<< std::endl;
return 1;
}
for (int i = 0; i < nDevices; ++i) {
cudaDeviceProp prop;
cudaError_t err = cudaGetDeviceProperties(&prop, i);
if (err != cudaSuccess) {
std::cerr << "Could not retrieve properties from CUDA device " << i
<< std::endl;
return 1;
}
if (prop.major >= 3) {
err = cudaSetDevice(i);
if (err != cudaSuccess) {
std::cout << "Could not select CUDA device " << i << std::endl;
} else {
return 0;
}
}
}
std::cout << "Could not find a CUDA enabled card supporting compute >=3.0"
<< std::endl;
return 1;
}
int main(int argc, char** argv)
{
int ret = choose_cuda_device();
if (ret) {
return 0;
}
cudaError_t err;
file4_launch_kernel(42);
err = cudaGetLastError();
if (err != cudaSuccess) {
std::cerr << "file4_launch_kernel: kernel launch failed: "
<< cudaGetErrorString(err) << std::endl;
return 1;
}
file5_launch_kernel(42);
err = cudaGetLastError();
if (err != cudaSuccess) {
std::cerr << "file5_launch_kernel: kernel launch failed: "
<< cudaGetErrorString(err) << std::endl;
return 1;
}
return 0;
}
|