-
Notifications
You must be signed in to change notification settings - Fork 16
/
declare_copyin.c
250 lines (224 loc) · 5.24 KB
/
declare_copyin.c
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
#define DECLARE_TEST
#define DECLARE_COPYIN
int mult_copyin = 2;
#include "acc_testsuite_declare.h"
#include "acc_testsuite.h"
#pragma acc declare copyin(fixed_size_array)
#pragma acc declare copyin(scalar)
#pragma acc declare copyin(datapointer)
#pragma acc declare copyin(n)
#pragma acc routine vector
void multiplyData(real_t *a){
#pragma acc loop vector
for (int x = 0; x < n; ++x){
a[x] = a[x] * 2;
}
}
#ifndef T1
//T1:declare,construct-independent,V:1.0-2.7
int test1(){
int err = 0;
srand(SEED);
real_t * a = (real_t *)malloc(n * sizeof(real_t));
real_t * b = (real_t *)malloc(n * sizeof(real_t));
int mult = 2;
for (int x = 0; x < n; ++x){
a[x] = rand() / (real_t)(RAND_MAX / 10);
b[x] = 0;
}
#pragma acc data copyin(a[0:n]) copyout(b[0:n]) present(fixed_size_array)
{
#pragma acc parallel
{
#pragma acc loop
for (int x = 0; x < n; ++x){
b[x] = a[x] + fixed_size_array[x%10];
}
}
}
for (int x = 0; x < n; ++x){
if (fabs(b[x] - (a[x] + fixed_size_array[x%10])) > PRECISION){
err += 1;
break;
}
}
return err;
}
#endif
#ifndef T2
//T2:declare,construct-independent,V:1.0-2.7
int test2(){
int err = 0;
srand(SEED);
real_t * a = (real_t *)malloc(n * sizeof(real_t));
real_t * b = (real_t *)malloc(n * sizeof(real_t));
int mult = 2;
for (int x = 0; x < n; ++x){
a[x] = rand() / (real_t)(RAND_MAX / 10);
b[x] = 0;
}
#pragma acc data copyin(a[0:n]) copyout(b[0:n]) present(scalar)
{
#pragma acc parallel
{
#pragma acc loop
for (int x = 0; x < n; ++x){
b[x] = a[x] + scalar;
}
}
}
for (int x = 0; x < n; ++x){
if (fabs(b[x] - (a[x] + scalar)) > PRECISION){
err += 1;
break;
}
}
return err;
}
#endif
#ifndef T3
//T3:declare,construct-independent,V:2.0-2.7
int test3(){
int err = 0;
srand(SEED);
real_t * a = (real_t *)malloc(n * sizeof(real_t));
real_t * b = (real_t *)malloc(n * sizeof(real_t));
int mult = 2;
for (int x = 0; x < n; ++x){
a[x] = rand() / (real_t)(RAND_MAX / 10);
b[x] = a[x];
}
#pragma acc data copy(a[0:n])
{
#pragma acc parallel
{
#pragma acc loop
for (int x = 0; x < 1; ++x){
extern_multiplyData_copyin(a, n);
}
}
}
for (int x = 0; x < n; ++x){
if (fabs(a[x] - (b[x] * 2)) > PRECISION){
err += 1;
break;
}
}
return err;
}
#endif
#ifndef T4
//T4:declare,construct-independent,V:2.0-2.7
int test4(){
int err = 0;
srand(SEED);
real_t * a = (real_t *)malloc(n * sizeof(real_t));
real_t * b = (real_t *)malloc(n * sizeof(real_t));
int mult = 2;
for (int x = 0; x < n; ++x){
a[x] = rand() / (real_t)(RAND_MAX / 10);
b[x] = a[x];
}
#pragma acc data copy(a[0:n])
{
#pragma acc parallel
{
#pragma acc loop
for (int x = 0; x < 1; ++x){
multiplyData(a);
}
}
}
for (int x = 0; x < n; ++x){
if (fabs(a[x] - (b[x] * 2)) > PRECISION){
err += 1;
break;
}
}
return err;
}
#endif
#ifndef T5
//T5:declare,construct-independent,attach,V:2.6-2.7
int test5(){
int err = 0;
srand(SEED);
real_t * a = (real_t *)malloc(n * sizeof(real_t));
real_t * b = (real_t *)malloc(n * sizeof(real_t));
int mult = 2;
for (int x = 0; x < n; ++x){
a[x] = rand() / (real_t)(RAND_MAX / 10);
b[x] = a[x];
}
datapointer = a;
#pragma acc enter data copyin(a[0:n]) attach(datapointer)
#pragma acc data present(datapointer[0:n])
{
#pragma acc parallel
{
#pragma acc loop
for (int x = 0; x < n; ++x){
datapointer[x] = datapointer[x] * 2;
}
}
}
#pragma acc exit data copyout(a[0:n])
for (int x = 0; x < n; ++x){
if (fabs(a[x] - (b[x] * 2)) > PRECISION){
err += 1;
break;
}
}
return err;
}
#endif
int main(){
int failcode = 0;
int failed;
#ifndef T1
failed = 0;
for (int x = 0; x < NUM_TEST_CALLS; ++x){
failed = failed + test1();
}
if (failed != 0){
failcode = failcode + (1 << 0);
}
#endif
#ifndef T2
failed = 0;
for (int x = 0; x < NUM_TEST_CALLS; ++x){
failed = failed + test2();
}
if (failed != 0){
failcode = failcode + (1 << 1);
}
#endif
#ifndef T3
failed = 0;
for (int x = 0; x < NUM_TEST_CALLS; ++x){
failed = failed + test3();
}
if (failed != 0){
failcode = failcode + (1 << 2);
}
#endif
#ifndef T4
failed = 0;
for (int x = 0; x < NUM_TEST_CALLS; ++x){
failed = failed + test4();
}
if (failed != 0){
failcode = failcode + (1 << 3);
}
#endif
#ifndef T5
failed = 0;
for (int x = 0; x < NUM_TEST_CALLS; ++x){
failed = failed + test5();
}
if (failed != 0){
failcode = failcode + (1 << 4);
}
#endif
return failcode;
}