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
|
module xxx
contains
subroutine writelog(msg, val)
character(len=*), intent(in) :: msg
integer, optional, intent(in) :: val
integer, external :: omp_get_thread_num
if ( present(val) ) then
write(*,*) "---", omp_get_thread_num(), ": ", msg, val
else
write(*,*) "---", omp_get_thread_num(), ": ", msg
endif
return
end subroutine writelog
end module xxx
program trans
use xxx
integer, parameter :: iterations = 4
integer :: i, k = 0
integer :: lck
logical, external :: omp_test_lock
!$omp inst init
call writelog("sequential", 0);
! ---- plain parallel region
!$omp parallel
call writelog("parallel", 0);
!$omp end parallel
call writelog("sequential", 1)
! ---- parallel region
!$omp parallel
call writelog("parallel", 1)
!$omp inst begin(worksharing)
! ---- worksharing for loop without synchronisation
!$omp do
do i=1,iterations
call writelog("for nowait iteration", i)
enddo
!$omp end do nowait
! ---- user specified barrier
!$omp barrier
! ---- worksharing for loop with implicit synchronisation
!$omp do
do i=1,iterations
call writelog("for iteration", i)
enddo
!$omp end do
! ---- worksharing tasks without synchronisation
!$omp sections
!$omp section
call writelog("section nowait", 1)
!$omp section
call writelog("section nowait", 2)
!$omp end sections nowait
! ---- worksharing tasks with implicit synchronisation
!$omp sections
!$omp section
call writelog("section", 1)
!$omp section
call writelog("section", 2)
!$omp end sections
!$omp inst end(worksharing)
!$omp inst begin(synchronisation)
! ---- critical section
!$omp critical
call writelog("critical\n")
k = k+1
!$omp end critical
! ---- named critical section
!$omp critical(kincr)
call writelog("critical\n")
k = k+1
!$omp end critical(kincr)
! ---- atomic expression
!$omp atomic
k = k+1
! ---- update k just once without synchronisation
!$omp single
call writelog("single nowait\n")
k = k+1
!$omp end single nowait
! ---- update k just once with implicit synchronisation
!$omp single
call writelog("single\n")
k = k+1
!$omp end single
!$omp master
call writelog("master\n")
write(*,*) "k = ", k
k = 0
!$omp end master
!$omp inst end(synchronisation)
!$omp end parallel
call writelog("sequential", 2)
!$omp inst begin(parallelworksharing)
! ---- combined parallel worksharing for loop
!$omp parallel do &
!$omp reduction(+:k) &
!$omp private(i) &
!$omp schedule(dynamic)
do i=1,iterations
call writelog("pfor", i)
enddo
!$omp end parallel do
call writelog("sequential", 3)
! ---- combined parallel worksharing tasks
!$omp parallel sections
!$omp section
call writelog("psection", 1)
!$omp section
call writelog("psection", 2)
!$omp end parallel sections
call writelog("sequential", 4)
!$omp inst end(parallelworksharing)
! ---- OpenMP locking API
!$omp inst begin(locking)
call omp_init_lock(lck)
!$omp parallel shared(lck)
call omp_set_lock(lck)
call writelog("got lock")
call omp_unset_lock(lck)
do while (.not. omp_test_lock(lck))
call writelog("skipping")
enddo
call writelog("working")
call omp_unset_lock(lck)
!$omp end parallel
call omp_destroy_lock(lck)
!$omp flush(k)
!$omp inst end(locking)
end program trans
|
