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
|
/* { dg-do run } */
#include <stdlib.h>
#define EPS 0.0001
#define N 1024*1024
void init (float B[], float C[], int n)
{
int i;
for (i = 0; i < n; i++)
{
B[i] = 0.1 * i;
C[i] = 0.01 * i * i;
}
}
float dotprod_ref (float B[], float C[], int n)
{
int i;
float sum = 0.0;
for (i = 0; i < n; i++)
sum += B[i] * C[i];
return sum;
}
float dotprod (float B[], float C[], int n, int block_size,
int num_teams, int block_threads)
{
int i, i0;
float sum = 0;
#pragma omp target map(to: B[0:n], C[0:n])
#pragma omp teams num_teams(num_teams) thread_limit(block_threads) \
reduction(+:sum)
#pragma omp distribute
for (i0 = 0; i0 < n; i0 += block_size)
#pragma omp parallel for reduction(+:sum)
for (i = i0; i < ((i0 + block_size > n) ? n : i0 + block_size); i++)
sum += B[i] * C[i];
return sum;
}
void check (float a, float b)
{
float err = (b == 0.0) ? a : (a - b) / b;
if (((err > 0) ? err : -err) > EPS)
abort ();
}
int main ()
{
float *v1 = (float *) malloc (N * sizeof (float));
float *v2 = (float *) malloc (N * sizeof (float));
float p1, p2;
init (v1, v2, N);
p1 = dotprod_ref (v1, v2, N);
p2 = dotprod (v1, v2, N, N / 8, 2, 8);
check (p1, p2);
free (v1);
free (v2);
return 0;
}
|
