1 |
|
/** |
2 |
|
* HMLP (High-Performance Machine Learning Primitives) |
3 |
|
* |
4 |
|
* Copyright (C) 2014-2017, The University of Texas at Austin |
5 |
|
* |
6 |
|
* This program is free software: you can redistribute it and/or modify |
7 |
|
* it under the terms of the GNU General Public License as published by |
8 |
|
* the Free Software Foundation, either version 3 of the License, or |
9 |
|
* (at your option) any later version. |
10 |
|
* |
11 |
|
* This program is distributed in the hope that it will be useful, |
12 |
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of |
13 |
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
14 |
|
* GNU General Public License for more details. |
15 |
|
* |
16 |
|
* You should have received a copy of the GNU General Public License |
17 |
|
* along with this program. If not, see the LICENSE file. |
18 |
|
* |
19 |
|
**/ |
20 |
|
|
21 |
|
|
22 |
|
|
23 |
|
|
24 |
|
/** STRASSEN templates */ |
25 |
|
#include <primitives/strassen.hpp> |
26 |
|
|
27 |
|
/** Haswell micro-kernel */ |
28 |
|
#include <rank_k_d8x6.hpp> |
29 |
|
|
30 |
|
using namespace hmlp; |
31 |
|
|
32 |
|
void strassen |
33 |
|
( |
34 |
|
hmlpOperation_t transA, hmlpOperation_t transB, |
35 |
|
int m, int n, int k, |
36 |
|
float *A, int lda, |
37 |
|
float *B, int ldb, |
38 |
|
float *C, int ldc |
39 |
|
) |
40 |
|
{ |
41 |
|
printf( "no implementation\n" ); |
42 |
|
exit( 1 ); |
43 |
|
}; |
44 |
|
|
45 |
|
void strassen |
46 |
|
( |
47 |
|
hmlpOperation_t transA, hmlpOperation_t transB, |
48 |
|
int m, int n, int k, |
49 |
|
double *A, int lda, |
50 |
|
double *B, int ldb, |
51 |
|
double *C, int ldc |
52 |
|
) |
53 |
|
{ |
54 |
|
rank_k_asm_d8x6 stra_semiringkernel; |
55 |
|
rank_k_asm_d8x6 stra_microkernel; |
56 |
|
|
57 |
|
strassen::strassen< |
58 |
|
72, 960, 256, 8, 6, |
59 |
|
72, 960, 8, 6, 32, |
60 |
|
false, |
61 |
|
rank_k_asm_d8x6, |
62 |
|
rank_k_asm_d8x6, |
63 |
|
double, double, double, double> |
64 |
|
( |
65 |
|
transA, transB, |
66 |
|
m, n, k, |
67 |
|
A, lda, |
68 |
|
B, ldb, |
69 |
|
C, ldc, |
70 |
|
stra_semiringkernel, |
71 |
|
stra_microkernel |
72 |
|
); |
73 |
|
|
74 |
|
}; |
75 |
|
|
76 |
|
|