blob: c6885ff73213de77a3b2ca9e53c3e5ae8237e2d6 (
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
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
|
// PR libgomp/69555
// { dg-do run }
#include <omp.h>
__attribute__((noinline, noclone)) void
f1 (int y)
{
int a[y - 2];
int (&c)[y - 2] = a;
c[0] = 111;
int e = 0;
#pragma omp parallel private (c) num_threads (4) reduction (+:e)
{
int v = omp_get_thread_num ();
for (int i = 0; i < y - 2; i++)
c[i] = i + v;
#pragma omp barrier
for (int i = 0; i < y - 2; i++)
if (c[i] != i + v)
e++;
}
if (c[0] != 111 || e)
__builtin_abort ();
}
__attribute__((noinline, noclone)) void
f2 (int y)
{
int a[y - 2];
int (&c)[y - 2] = a;
c[0] = 111;
#pragma omp task private (c)
{
int v = omp_get_thread_num ();
for (int i = 0; i < y - 2; i++)
c[i] = i + v;
asm volatile ("" : : "r" (&c[0]) : "memory");
for (int i = 0; i < y - 2; i++)
if (c[i] != i + v)
__builtin_abort ();
}
if (c[0] != 111)
__builtin_abort ();
}
__attribute__((noinline, noclone)) void
f3 (int y)
{
int a[y - 2];
int (&c)[y - 2] = a;
for (int i = 0; i < y - 2; i++)
c[i] = i + 4;
#pragma omp parallel firstprivate (c) num_threads (4)
{
int v = omp_get_thread_num ();
for (int i = 0; i < y - 2; i++)
{
if (c[i] != i + 4)
__builtin_abort ();
c[i] = i + v;
}
#pragma omp barrier
for (int i = 0; i < y - 2; i++)
if (c[i] != i + v)
__builtin_abort ();
}
for (int i = 0; i < y - 2; i++)
if (c[i] != i + 4)
__builtin_abort ();
}
__attribute__((noinline, noclone)) void
f4 (int y)
{
int a[y - 2];
int (&c)[y - 2] = a;
for (int i = 0; i < y - 2; i++)
c[i] = i + 4;
#pragma omp task firstprivate (c)
{
int v = omp_get_thread_num ();
for (int i = 0; i < y - 2; i++)
{
if (c[i] != i + 4)
__builtin_abort ();
c[i] = i + v;
}
asm volatile ("" : : "r" (&c[0]) : "memory");
for (int i = 0; i < y - 2; i++)
if (c[i] != i + v)
__builtin_abort ();
}
for (int i = 0; i < y - 2; i++)
if (c[i] != i + 4)
__builtin_abort ();
}
int
main ()
{
f1 (6);
f3 (6);
#pragma omp parallel num_threads (4)
{
f2 (6);
f4 (6);
}
return 0;
}
|
