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
|
// clang-format off
// RUN: %libomptarget-compile-generic
// RUN: env LIBOMPTARGET_INFO=16 \
// RUN: %libomptarget-run-generic 2>&1 | %fcheck-generic --check-prefix=DEFAULT
// REQUIRES: amdgpu
__attribute__((optnone)) int optnone() { return 1; }
int main() {
int N = optnone() * 4098 * 32;
// DEFAULT: [[NT:(128|256)]] (MaxFlatWorkGroupSize: [[NT]]
#pragma omp target teams distribute parallel for simd
for (int i = 0; i < N; ++i) {
optnone();
}
// DEFAULT: [[NT:(128|256)]] (MaxFlatWorkGroupSize: [[NT]]
#pragma omp target teams distribute parallel for simd
for (int i = 0; i < N; ++i) {
optnone();
}
// DEFAULT: [[NT:(128|256)]] (MaxFlatWorkGroupSize: [[NT]]
#pragma omp target teams distribute parallel for simd
for (int i = 0; i < N; ++i) {
optnone();
}
// DEFAULT: [[NT:(128|256)]] (MaxFlatWorkGroupSize: [[NT]]
#pragma omp target
#pragma omp teams distribute parallel for
for (int i = 0; i < N; ++i) {
optnone();
}
// DEFAULT: 42 (MaxFlatWorkGroupSize: 1024
#pragma omp target thread_limit(optnone() * 42)
#pragma omp teams distribute parallel for
for (int i = 0; i < N; ++i) {
optnone();
}
// DEFAULT: 42 (MaxFlatWorkGroupSize: 42
#pragma omp target thread_limit(optnone() * 42) ompx_attribute(__attribute__((amdgpu_flat_work_group_size(42, 42))))
#pragma omp teams distribute parallel for
for (int i = 0; i < N; ++i) {
optnone();
}
// DEFAULT: 42 (MaxFlatWorkGroupSize: 42
#pragma omp target ompx_attribute(__attribute__((amdgpu_flat_work_group_size(42, 42))))
#pragma omp teams distribute parallel for
for (int i = 0; i < N; ++i) {
optnone();
}
// DEFAULT: MaxFlatWorkGroupSize: 1024
#pragma omp target
#pragma omp teams distribute parallel for num_threads(optnone() * 42)
for (int i = 0; i < N; ++i) {
optnone();
}
// DEFAULT: MaxFlatWorkGroupSize: 1024
#pragma omp target teams distribute parallel for thread_limit(optnone() * 42)
for (int i = 0; i < N; ++i) {
optnone();
}
// DEFAULT: MaxFlatWorkGroupSize: 1024
#pragma omp target teams distribute parallel for num_threads(optnone() * 42)
for (int i = 0; i < N; ++i) {
optnone();
}
// DEFAULT: 9 (MaxFlatWorkGroupSize: 9
#pragma omp target
#pragma omp teams distribute parallel for num_threads(9)
for (int i = 0; i < N; ++i) {
optnone();
}
// DEFAULT: 4 (MaxFlatWorkGroupSize: 4
#pragma omp target thread_limit(4)
#pragma omp teams distribute parallel for
for (int i = 0; i < N; ++i) {
optnone();
}
// DEFAULT: 4 (MaxFlatWorkGroupSize: 4
#pragma omp target
#pragma omp teams distribute parallel for thread_limit(4)
for (int i = 0; i < N; ++i) {
optnone();
}
// DEFAULT: 9 (MaxFlatWorkGroupSize: 9
#pragma omp target teams distribute parallel for num_threads(9)
for (int i = 0; i < N; ++i) {
optnone();
}
// DEFAULT: 4 (MaxFlatWorkGroupSize: 4
#pragma omp target teams distribute parallel for simd thread_limit(4)
for (int i = 0; i < N; ++i) {
optnone();
}
}
|
