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
|
/*------------------------------------------------------------------------------------------------*/
/* Copyright (C) by the DBCSR developers group - All rights reserved */
/* This file is part of the DBCSR library. */
/* */
/* For information on the license, see the LICENSE file. */
/* For further information please visit https://dbcsr.cp2k.org */
/* SPDX-License-Identifier: GPL-2.0+ */
/*------------------------------------------------------------------------------------------------*/
#include <algorithm>
#include <cstdint>
#include <cstdio>
#include <cstdlib>
#include <iostream>
#include <random>
#include <vector>
#include <mpi.h>
#include <dbcsr.h>
// Random distribution by using round-robin assignment
// of blocks to processors
std::vector<int> random_dist(int dist_size, int nbins) {
std::vector<int> dist(dist_size);
for (int i = 0; i < dist_size; i++) dist[i] = i % nbins;
return dist;
}
// DBCSR example 3
// This example shows how to multiply two DBCSR matrices
int main(int argc, char* argv[]) {
// initialize MPI
MPI_Init(&argc, &argv);
// setup the mpi environment
int mpi_size, mpi_rank;
MPI_Comm_size(MPI_COMM_WORLD, &mpi_size);
MPI_Comm_rank(MPI_COMM_WORLD, &mpi_rank);
// make 2D grid
int dims[2] = {0};
MPI_Dims_create(mpi_size, 2, dims);
int periods[2] = {1};
int reorder = 0;
MPI_Comm group;
MPI_Cart_create(MPI_COMM_WORLD, 2, dims, periods, reorder, &group);
int coord[2];
MPI_Cart_coords(group, mpi_rank, 2, coord);
for (int i = 0; i != mpi_size; ++i) {
if (mpi_rank == i) {
std::cout << "I'm processor " << mpi_rank << " over " << mpi_size << " proc"
<< ", (" << coord[0] << ", " << coord[1] << ") in the 2D grid" << std::endl;
}
MPI_Barrier(MPI_COMM_WORLD);
}
// initialize the DBCSR library
c_dbcsr_init_lib(MPI_COMM_WORLD, nullptr);
// Total number of blocks
int nrows_1 = 4;
int ncols_1 = 5;
int nrows_2 = 5;
int ncols_2 = 4;
// Block sizes
std::vector<int> row_blk_sizes_1 = {2, 3, 5, 2};
std::vector<int> col_blk_sizes_1 = {3, 3, 4, 6, 2};
std::vector<int> row_blk_sizes_2 = col_blk_sizes_1;
std::vector<int> col_blk_sizes_2 = {5, 2, 5, 3};
auto row_dist_1 = random_dist(nrows_1, dims[0]);
auto col_dist_1 = random_dist(ncols_1, dims[1]);
auto row_dist_2 = random_dist(nrows_2, dims[0]);
auto col_dist_2 = random_dist(ncols_2, dims[1]);
dbcsr_distribution dist1 = nullptr;
dbcsr_distribution dist2 = nullptr;
dbcsr_distribution dist3 = nullptr;
//create distributions
c_dbcsr_distribution_new(&dist1, group, row_dist_1.data(), row_dist_1.size(), col_dist_1.data(), col_dist_1.size());
c_dbcsr_distribution_new(&dist2, group, row_dist_2.data(), row_dist_2.size(), col_dist_2.data(), col_dist_2.size());
c_dbcsr_distribution_new(&dist3, group, row_dist_1.data(), row_dist_1.size(), col_dist_2.data(), col_dist_2.size());
// Fill all blocks, i.e. dense matrices
auto fill_matrix = [&](void* matrix, std::vector<int>& irblks, std::vector<int>& icblks) {
std::vector<double> block;
std::vector<int> loc_irblks, loc_icblks;
for (int i = 0; i != (int)irblks.size(); ++i) {
int blk_proc = -1;
int ix = irblks[i];
int jx = icblks[i];
c_dbcsr_get_stored_coordinates(matrix, ix, jx, &blk_proc);
if (mpi_rank == blk_proc) {
loc_irblks.push_back(ix);
loc_icblks.push_back(jx);
}
}
c_dbcsr_reserve_blocks(matrix, loc_irblks.data(), loc_icblks.data(), loc_irblks.size());
void* iter = nullptr;
c_dbcsr_iterator_start(&iter, matrix, nullptr, nullptr, nullptr, nullptr, nullptr);
while (c_dbcsr_iterator_blocks_left(iter)) {
int i = -1;
int j = -1;
int nblk = -1;
int rsize = -1;
int csize = -1;
bool tr = false;
double* blk = nullptr;
c_dbcsr_iterator_next_2d_block_d(iter, &i, &j, &blk, &tr, &nblk, &rsize, &csize, nullptr, nullptr);
std::generate(blk, blk + rsize * csize, [&]() { return static_cast<double>(std::rand()) / RAND_MAX; });
}
c_dbcsr_iterator_stop(&iter);
};
dbcsr_matrix matrix_a = nullptr;
dbcsr_matrix matrix_b = nullptr;
dbcsr_matrix matrix_c = nullptr;
c_dbcsr_create_new(&matrix_a, "matrix a", dist1, dbcsr_type_no_symmetry, row_blk_sizes_1.data(), row_blk_sizes_1.size(),
col_blk_sizes_1.data(), col_blk_sizes_1.size(), nullptr, nullptr, nullptr, nullptr, nullptr, nullptr);
c_dbcsr_create_new(&matrix_b, "matrix b", dist2, dbcsr_type_no_symmetry, row_blk_sizes_2.data(), row_blk_sizes_2.size(),
col_blk_sizes_2.data(), col_blk_sizes_2.size(), nullptr, nullptr, nullptr, nullptr, nullptr, nullptr);
c_dbcsr_create_new(&matrix_c, "matrix c", dist3, dbcsr_type_no_symmetry, row_blk_sizes_1.data(), row_blk_sizes_1.size(),
col_blk_sizes_2.data(), col_blk_sizes_2.size(), nullptr, nullptr, nullptr, nullptr, nullptr, nullptr);
// indices of non-zero blocks
std::vector<int> irblks_1 = {0, 0, 0, 0, 1, 1, 1, 2, 2, 2, 3, 3, 3};
std::vector<int> icblks_1 = {0, 1, 2, 4, 0, 2, 3, 1, 3, 4, 0, 1, 2};
std::vector<int> irblks_2 = {0, 0, 0, 1, 1, 1, 1, 2, 2, 2, 3, 3, 3, 4, 4, 4, 4};
std::vector<int> icblks_2 = {0, 2, 3, 0, 1, 2, 3, 0, 2, 3, 1, 2, 3, 0, 1, 2, 3};
std::vector<int> irblks_3 = {0, 0, 0, 0, 1, 1, 1, 2, 2, 2, 3, 3, 3, 3};
std::vector<int> icblks_3 = {0, 1, 2, 3, 0, 2, 3, 1, 2, 3, 0, 1, 2, 3};
fill_matrix(matrix_a, irblks_1, icblks_1);
c_dbcsr_finalize(matrix_a);
fill_matrix(matrix_b, irblks_2, icblks_2);
c_dbcsr_finalize(matrix_b);
fill_matrix(matrix_c, irblks_3, icblks_3);
c_dbcsr_finalize(matrix_c);
// Compute C = 3.0 * A * B + 2.0 * C
c_dbcsr_multiply_d('N', 'N', 3.0, matrix_a, matrix_b, 2.0, matrix_c, nullptr, nullptr, nullptr, nullptr, nullptr, nullptr,
nullptr, nullptr, nullptr);
c_dbcsr_print(matrix_c);
// release the matrices
c_dbcsr_release(&matrix_a);
c_dbcsr_release(&matrix_b);
c_dbcsr_release(&matrix_c);
c_dbcsr_distribution_release(&dist1);
c_dbcsr_distribution_release(&dist2);
c_dbcsr_distribution_release(&dist3);
MPI_Comm_free(&group);
// finalize the DBCSR library
c_dbcsr_finalize_lib();
// finalize MPI
MPI_Finalize();
return 0;
}
|
