blob: 7369045c0b7847f61cdffc0ca35c142e3aec03c7 (
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
|
#include <vector>
void vecaddgpu(float* r, float* a, float* b, std::size_t n)
{
#pragma acc kernels loop present(r, a, b)
for (std::size_t i = 0; i < n; ++i)
r[i] = a[i] + b[i];
}
int main(int, char* [])
{
const std::size_t n = 100000; /* vector length */
std::vector<float> a(n); /* input vector 1 */
std::vector<float> b(n); /* input vector 2 */
std::vector<float> r(n); /* output vector */
std::vector<float> e(n); /* expected output values */
for (std::size_t i = 0; i < n; ++i) {
a[i] = static_cast<float>(i + 1);
b[i] = static_cast<float>(1000 * i);
}
/* compute on the GPU */
auto a_ptr = a.data();
auto b_ptr = b.data();
auto r_ptr = r.data();
#pragma acc data copyin(a_ptr [0:n], b_ptr [0:n]) copyout(r_ptr [0:n])
{
vecaddgpu(r_ptr, a_ptr, b_ptr, n);
}
/* compute on the host to compare */
for (std::size_t i = 0; i < n; ++i)
e[i] = a[i] + b[i];
/* compare results */
int errs = 0;
for (std::size_t i = 0; i < n; ++i) {
if (r[i] != e[i]) {
++errs;
}
}
return errs;
}
|
