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
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
|
/*
-- MAGMA (version 2.9.0) --
Univ. of Tennessee, Knoxville
Univ. of California, Berkeley
Univ. of Colorado, Denver
@date January 2025
@generated from magmablas/blas_zbatched.cpp, normal z -> c, Wed Jan 22 14:42:10 2025
@author Ahmad Abdelfattah
Implementation of batch BLAS on the host ( CPU ) using OpenMP
*/
#include "magma_internal.h"
#include "commonblas_c.h"
#if defined(_OPENMP)
#include <omp.h>
#include "magma_threadsetting.h"
#endif
/*******************************************************************************/
extern "C" void
blas_clacpy_batched(
magma_uplo_t uplo, magma_int_t m, magma_int_t n,
magmaFloatComplex const * const * hA_array, magma_int_t lda,
magmaFloatComplex **hB_array, magma_int_t ldb,
magma_int_t batchCount )
{
#if defined(_OPENMP)
magma_int_t nthreads = magma_get_lapack_numthreads();
magma_set_lapack_numthreads(1);
magma_set_omp_numthreads(nthreads);
#pragma omp parallel for schedule(dynamic)
#endif
for (int i=0; i < batchCount; i++) {
lapackf77_clacpy( lapack_uplo_const(uplo),
&m, &n,
hA_array[i], &lda,
hB_array[i], &ldb );
}
#if defined(_OPENMP)
magma_set_lapack_numthreads(nthreads);
#endif
}
/*******************************************************************************/
extern "C" void
blas_cgemm_batched(
magma_trans_t transA, magma_trans_t transB,
magma_int_t m, magma_int_t n, magma_int_t k,
magmaFloatComplex alpha,
magmaFloatComplex const * const * hA_array, magma_int_t lda,
magmaFloatComplex const * const * hB_array, magma_int_t ldb,
magmaFloatComplex beta,
magmaFloatComplex **hC_array, magma_int_t ldc,
magma_int_t batchCount )
{
#if defined(_OPENMP)
magma_int_t nthreads = magma_get_lapack_numthreads();
magma_set_lapack_numthreads(1);
magma_set_omp_numthreads(nthreads);
#pragma omp parallel for schedule(dynamic)
#endif
for (int i=0; i < batchCount; i++) {
blasf77_cgemm( lapack_trans_const(transA),
lapack_trans_const(transB),
&m, &n, &k,
&alpha, hA_array[i], &lda,
hB_array[i], &ldb,
&beta, hC_array[i], &ldc );
}
#if defined(_OPENMP)
magma_set_lapack_numthreads(nthreads);
#endif
}
/*******************************************************************************/
extern "C" void
blas_ctrsm_batched(
magma_side_t side, magma_uplo_t uplo, magma_trans_t transA, magma_diag_t diag,
magma_int_t m, magma_int_t n,
magmaFloatComplex alpha,
magmaFloatComplex **hA_array, magma_int_t lda,
magmaFloatComplex **hB_array, magma_int_t ldb,
magma_int_t batchCount )
{
#if defined(_OPENMP)
magma_int_t nthreads = magma_get_lapack_numthreads();
magma_set_lapack_numthreads(1);
magma_set_omp_numthreads(nthreads);
#pragma omp parallel for schedule(dynamic)
#endif
for (int s=0; s < batchCount; s++) {
blasf77_ctrsm(
lapack_side_const(side), lapack_uplo_const(uplo),
lapack_trans_const(transA), lapack_diag_const(diag),
&m, &n, &alpha,
hA_array[s], &lda,
hB_array[s], &ldb );
}
#if defined(_OPENMP)
magma_set_lapack_numthreads(nthreads);
#endif
}
/*******************************************************************************/
extern "C" void
blas_ctrmm_batched(
magma_side_t side, magma_uplo_t uplo, magma_trans_t transA, magma_diag_t diag,
magma_int_t m, magma_int_t n,
magmaFloatComplex alpha,
magmaFloatComplex **hA_array, magma_int_t lda,
magmaFloatComplex **hB_array, magma_int_t ldb,
magma_int_t batchCount )
{
#if defined(_OPENMP)
magma_int_t nthreads = magma_get_lapack_numthreads();
magma_set_lapack_numthreads(1);
magma_set_omp_numthreads(nthreads);
#pragma omp parallel for schedule(dynamic)
#endif
for (int s=0; s < batchCount; s++) {
blasf77_ctrmm(
lapack_side_const(side), lapack_uplo_const(uplo),
lapack_trans_const(transA), lapack_diag_const(diag),
&m, &n, &alpha,
hA_array[s], &lda,
hB_array[s], &ldb );
}
#if defined(_OPENMP)
magma_set_lapack_numthreads(nthreads);
#endif
}
/*******************************************************************************/
extern "C" void
blas_chemm_batched(
magma_side_t side, magma_uplo_t uplo,
magma_int_t m, magma_int_t n,
magmaFloatComplex alpha,
magmaFloatComplex **hA_array, magma_int_t lda,
magmaFloatComplex **hB_array, magma_int_t ldb,
magmaFloatComplex beta,
magmaFloatComplex **hC_array, magma_int_t ldc,
magma_int_t batchCount )
{
#if defined(_OPENMP)
magma_int_t nthreads = magma_get_lapack_numthreads();
magma_set_lapack_numthreads(1);
magma_set_omp_numthreads(nthreads);
#pragma omp parallel for schedule(dynamic)
#endif
for (int i=0; i < batchCount; i++) {
blasf77_chemm( lapack_side_const(side),
lapack_uplo_const(uplo),
&m, &n,
&alpha, hA_array[i], &lda,
hB_array[i], &ldb,
&beta, hC_array[i], &ldc );
}
#if defined(_OPENMP)
magma_set_lapack_numthreads(nthreads);
#endif
}
/*******************************************************************************/
extern "C" void
blas_cherk_batched(
magma_uplo_t uplo, magma_trans_t trans,
magma_int_t n, magma_int_t k,
float alpha, magmaFloatComplex const * const * hA_array, magma_int_t lda,
float beta, magmaFloatComplex **hC_array, magma_int_t ldc,
magma_int_t batchCount )
{
#if defined(_OPENMP)
magma_int_t nthreads = magma_get_lapack_numthreads();
magma_set_lapack_numthreads(1);
magma_set_omp_numthreads(nthreads);
#pragma omp parallel for schedule(dynamic)
#endif
for (int s=0; s < batchCount; s++) {
blasf77_cherk( lapack_uplo_const(uplo),
lapack_trans_const(trans),
&n, &k,
&alpha, hA_array[s], &lda,
&beta, hC_array[s], &ldc );
}
#if defined(_OPENMP)
magma_set_lapack_numthreads(nthreads);
#endif
}
/*******************************************************************************/
extern "C" void
blas_cher2k_batched(
magma_uplo_t uplo, magma_trans_t trans,
magma_int_t n, magma_int_t k,
magmaFloatComplex alpha, magmaFloatComplex const * const * hA_array, magma_int_t lda,
magmaFloatComplex const * const * hB_array, magma_int_t ldb,
float beta, magmaFloatComplex **hC_array, magma_int_t ldc,
magma_int_t batchCount )
{
#if defined(_OPENMP)
magma_int_t nthreads = magma_get_lapack_numthreads();
magma_set_lapack_numthreads(1);
magma_set_omp_numthreads(nthreads);
#pragma omp parallel for schedule(dynamic)
#endif
for (int i=0; i < batchCount; i++) {
blasf77_cher2k( lapack_uplo_const(uplo),
lapack_trans_const(trans),
&n, &k,
&alpha, hA_array[i], &lda,
hB_array[i], &ldb,
&beta, hC_array[i], &ldc );
}
#if defined(_OPENMP)
magma_set_lapack_numthreads(nthreads);
#endif
}
|
