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
// **************************************************************************
//
// PARALUTION www.paralution.com
//
// Copyright (C) 2015 PARALUTION Labs UG (haftungsbeschränkt) & Co. KG
// Am Hasensprung 6, 76571 Gaggenau
// Handelsregister: Amtsgericht Mannheim, HRA 706051
// Vertreten durch:
// PARALUTION Labs Verwaltungs UG (haftungsbeschränkt)
// Am Hasensprung 6, 76571 Gaggenau
// Handelsregister: Amtsgericht Mannheim, HRB 721277
// Geschäftsführer: Dimitar Lukarski, Nico Trost
//
// This program is free software: you can redistribute it and/or modify
// it under the terms of the GNU General Public License as published by
// the Free Software Foundation, either version 3 of the License, or
// (at your option) any later version.
//
// This program is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
// GNU General Public License for more details.
//
// You should have received a copy of the GNU General Public License
// along with this program. If not, see <http://www.gnu.org/licenses/>.
//
// **************************************************************************
// PARALUTION version 1.1.0
#include "../../utils/def.hpp"
#include "mic_matrix_mcsr_kernel.hpp"
#include "mic_utils.hpp"
#include "../matrix_formats_ind.hpp"
namespace paralution {
template <typename ValueType>
void spmv_mcsr(const int mic_dev,
const int *row, const int *col, const ValueType *val,
const int nrow,
const ValueType *in, ValueType *out) {
#pragma offload target(mic:mic_dev) \
in(row:length(0) MIC_REUSE MIC_RETAIN) \
in(col:length(0) MIC_REUSE MIC_RETAIN) \
in(val:length(0) MIC_REUSE MIC_RETAIN) \
in(in:length(0) MIC_REUSE MIC_RETAIN) \
in(out:length(0) MIC_REUSE MIC_RETAIN)
#pragma omp parallel for
for (int ai=0; ai<nrow; ++ai) {
out[ai] = val[ai] * in[ai];
for (int aj=row[ai]; aj<row[ai+1]; ++aj)
out[ai] += val[aj] * in[ col[aj] ];
}
}
template <typename ValueType>
void spmv_add_mcsr(const int mic_dev,
const int *row, const int *col, const ValueType *val,
const int nrow,
const ValueType scalar,
const ValueType *in, ValueType *out) {
#pragma offload target(mic:mic_dev) \
in(row:length(0) MIC_REUSE MIC_RETAIN) \
in(col:length(0) MIC_REUSE MIC_RETAIN) \
in(val:length(0) MIC_REUSE MIC_RETAIN) \
in(in:length(0) MIC_REUSE MIC_RETAIN) \
in(out:length(0) MIC_REUSE MIC_RETAIN)
#pragma omp parallel for
for (int ai=0; ai<nrow; ++ai) {
out[ai] += scalar*val[ai] * in[ai];
for (int aj=row[ai]; aj<row[ai+1]; ++aj)
out[ai] += scalar*val[aj] * in[ col[aj] ];
}
}
template void spmv_mcsr<double>(const int mic_dev,
const int *row, const int *col, const double *val,
const int nrow,
const double *in, double *out);
template void spmv_mcsr<float>(const int mic_dev,
const int *row, const int *col, const float *val,
const int nrow,
const float *in, float *out);
template void spmv_mcsr<int>(const int mic_dev,
const int *row, const int *col, const int *val,
const int nrow,
const int *in, int *out);
template void spmv_add_mcsr<double>(const int mic_dev,
const int *row, const int *col, const double *val,
const int nrow,
const double scalar,
const double *in, double *out);
template void spmv_add_mcsr<float>(const int mic_dev,
const int *row, const int *col, const float *val,
const int nrow,
const float scalar,
const float *in, float *out);
template void spmv_add_mcsr<int>(const int mic_dev,
const int *row, const int *col, const int *val,
const int nrow,
const int scalar,
const int *in, int *out);
}