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
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
|
/*
-- MAGMA (version 2.9.0) --
Univ. of Tennessee, Knoxville
Univ. of California, Berkeley
Univ. of Colorado, Denver
@date January 2025
@precisions normal z -> c d s
@author Mark Gates
*/
// includes, system
#include <stdlib.h>
#include <stdio.h>
#include <string.h>
#include <math.h>
// includes, project
#include "flops.h"
#include "magma_v2.h"
#include "magma_lapack.h"
#include "magma_operators.h"
#include "testings.h"
/* ////////////////////////////////////////////////////////////////////////////
-- Testing zgemm
*/
int main( int argc, char** argv)
{
#ifdef MAGMA_HAVE_OPENCL
#define dA(i_, j_) dA, ((i_) + (j_)*ldda)
#define dB(i_, j_) dB, ((i_) + (j_)*lddb)
#define dC(i_, j_) dC, ((i_) + (j_)*lddc)
#else
#define dA(i_, j_) (dA + (i_) + (j_)*ldda)
#define dB(i_, j_) (dB + (i_) + (j_)*lddb)
#define dC(i_, j_) (dC + (i_) + (j_)*lddc)
#endif
TESTING_CHECK( magma_init() );
magma_print_environment();
real_Double_t gflops, magma_perf, magma_time, dev_perf, dev_time, cpu_perf, cpu_time;
double magma_error, dev_error, work[1];
magma_int_t M, N, K;
magma_int_t Am, An, Bm, Bn;
magma_int_t sizeA, sizeB, sizeC;
magma_int_t lda, ldb, ldc, ldda, lddb, lddc;
magma_int_t ione = 1;
magma_int_t ISEED[4] = {0,0,0,1};
int status = 0;
magmaDoubleComplex *hA, *hB, *hC, *hCmagma, *hCdev;
magmaDoubleComplex_ptr dA, dB, dC;
magmaDoubleComplex c_neg_one = MAGMA_Z_NEG_ONE;
magmaDoubleComplex alpha = MAGMA_Z_MAKE( 0.29, -0.86 );
magmaDoubleComplex beta = MAGMA_Z_MAKE( -0.48, 0.38 );
// used only with CUDA
MAGMA_UNUSED( magma_perf );
MAGMA_UNUSED( magma_time );
MAGMA_UNUSED( magma_error );
magma_opts opts;
opts.parse_opts( argc, argv );
// Allow 3*eps; complex needs 2*sqrt(2) factor; see Higham, 2002, sec. 3.6.
double eps = lapackf77_dlamch("E");
double tol = 3*eps;
#if defined(MAGMA_HAVE_CUDA) || defined(MAGMA_HAVE_HIP)
// for CUDA/HIP, we can check MAGMA vs. CUBLAS/hipBLAS, without running LAPACK
printf("%% If running lapack (option --lapack), MAGMA and %s error are both computed\n"
"%% relative to CPU BLAS result. Else, MAGMA error is computed relative to %s result.\n\n",
g_platform_str, g_platform_str );
printf("%% transA = %s, transB = %s\n",
lapack_trans_const(opts.transA),
lapack_trans_const(opts.transB) );
printf("%% M N K MAGMA Gflop/s (ms) %s Gflop/s (ms) CPU Gflop/s (ms) MAGMA error %s error\n",
g_platform_str, g_platform_str );
#else
// for others, we need LAPACK for check
opts.lapack |= opts.check; // check (-c) implies lapack (-l)
printf("%% transA = %s, transB = %s\n",
lapack_trans_const(opts.transA),
lapack_trans_const(opts.transB) );
printf("%% M N K %s Gflop/s (ms) CPU Gflop/s (ms) %s error\n",
g_platform_str, g_platform_str );
#endif
printf("%%========================================================================================================\n");
for( int itest = 0; itest < opts.ntest; ++itest ) {
for( int iter = 0; iter < opts.niter; ++iter ) {
M = opts.msize[itest];
N = opts.nsize[itest];
K = opts.ksize[itest];
gflops = FLOPS_ZGEMM( M, N, K ) / 1e9;
if ( opts.transA == MagmaNoTrans ) {
lda = Am = M;
An = K;
} else {
lda = Am = K;
An = M;
}
if ( opts.transB == MagmaNoTrans ) {
ldb = Bm = K;
Bn = N;
} else {
ldb = Bm = N;
Bn = K;
}
ldc = M;
ldda = magma_roundup( lda, opts.align ); // multiple of 32 by default
lddb = magma_roundup( ldb, opts.align ); // multiple of 32 by default
lddc = magma_roundup( ldc, opts.align ); // multiple of 32 by default
sizeA = lda*An;
sizeB = ldb*Bn;
sizeC = ldc*N;
TESTING_CHECK( magma_zmalloc_cpu( &hA, lda*An ));
TESTING_CHECK( magma_zmalloc_cpu( &hB, ldb*Bn ));
TESTING_CHECK( magma_zmalloc_cpu( &hC, ldc*N ));
TESTING_CHECK( magma_zmalloc_cpu( &hCmagma, ldc*N ));
TESTING_CHECK( magma_zmalloc_cpu( &hCdev, ldc*N ));
TESTING_CHECK( magma_zmalloc( &dA, ldda*An ));
TESTING_CHECK( magma_zmalloc( &dB, lddb*Bn ));
TESTING_CHECK( magma_zmalloc( &dC, lddc*N ));
/* Initialize the matrices */
lapackf77_zlarnv( &ione, ISEED, &sizeA, hA );
lapackf77_zlarnv( &ione, ISEED, &sizeB, hB );
lapackf77_zlarnv( &ione, ISEED, &sizeC, hC );
magma_zsetmatrix( Am, An, hA, lda, dA(0,0), ldda, opts.queue );
magma_zsetmatrix( Bm, Bn, hB, ldb, dB(0,0), lddb, opts.queue );
// for error checks
double Anorm = lapackf77_zlange( "F", &Am, &An, hA, &lda, work );
double Bnorm = lapackf77_zlange( "F", &Bm, &Bn, hB, &ldb, work );
double Cnorm = lapackf77_zlange( "F", &M, &N, hC, &ldc, work );
/* =====================================================================
Performs operation using MAGMABLAS (currently only with CUDA)
=================================================================== */
#if defined(MAGMA_HAVE_CUDA) || defined(MAGMA_HAVE_HIP)
magma_zsetmatrix( M, N, hC, ldc, dC, lddc, opts.queue );
magma_flush_cache( opts.cache );
magma_time = magma_sync_wtime( opts.queue );
magmablas_zgemm( opts.transA, opts.transB, M, N, K,
alpha, dA, ldda,
dB, lddb,
beta, dC, lddc,
opts.queue );
magma_time = magma_sync_wtime( opts.queue ) - magma_time;
magma_perf = gflops / magma_time;
magma_zgetmatrix( M, N, dC, lddc, hCmagma, ldc, opts.queue );
#endif
/* =====================================================================
Performs operation using CUBLAS / hipBLAS
=================================================================== */
magma_zsetmatrix( M, N, hC, ldc, dC(0,0), lddc, opts.queue );
magma_flush_cache( opts.cache );
dev_time = magma_sync_wtime( opts.queue );
magma_zgemm( opts.transA, opts.transB, M, N, K,
alpha, dA(0,0), ldda,
dB(0,0), lddb,
beta, dC(0,0), lddc, opts.queue );
dev_time = magma_sync_wtime( opts.queue ) - dev_time;
dev_perf = gflops / dev_time;
magma_zgetmatrix( M, N, dC(0,0), lddc, hCdev, ldc, opts.queue );
/* =====================================================================
Performs operation using CPU BLAS
=================================================================== */
if ( opts.lapack ) {
magma_flush_cache( opts.cache );
cpu_time = magma_wtime();
blasf77_zgemm( lapack_trans_const(opts.transA), lapack_trans_const(opts.transB), &M, &N, &K,
&alpha, hA, &lda,
hB, &ldb,
&beta, hC, &ldc );
cpu_time = magma_wtime() - cpu_time;
cpu_perf = gflops / cpu_time;
}
/* =====================================================================
Check the result
=================================================================== */
if ( opts.lapack ) {
// Compute forward error bound (see Higham, 2002, sec. 3.5),
// modified to include alpha, beta, and input C.
// ||R_magma - R_ref||_p / (gamma_{K+2} |alpha| ||A||_p ||B||_p + 2 |beta| ||C||_p ) < eps/2.
// This should work with p = 1, inf, fro, but numerical tests
// show p = 1, inf are very spiky and sometimes exceed eps.
// We use gamma_n = sqrt(n)*u instead of n*u/(1-n*u), since the
// former accurately represents statistical average rounding.
// We allow a slightly looser tolerance.
// use LAPACK for R_ref
blasf77_zaxpy( &sizeC, &c_neg_one, hC, &ione, hCdev, &ione );
dev_error = lapackf77_zlange( "F", &M, &N, hCdev, &ldc, work )
/ (sqrt(double(K+2))*fabs(alpha)*Anorm*Bnorm + 2*fabs(beta)*Cnorm);
#if defined(MAGMA_HAVE_CUDA) || defined(MAGMA_HAVE_HIP)
blasf77_zaxpy( &sizeC, &c_neg_one, hC, &ione, hCmagma, &ione );
magma_error = lapackf77_zlange( "F", &M, &N, hCmagma, &ldc, work )
/ (sqrt(double(K+2))*fabs(alpha)*Anorm*Bnorm + 2*fabs(beta)*Cnorm);
bool okay = (magma_error < tol && dev_error < tol);
status += ! okay;
printf("%5lld %5lld %5lld %7.2f (%7.2f) %7.2f (%7.2f) %7.2f (%7.2f) %8.2e %8.2e %s\n",
(long long) M, (long long) N, (long long) K,
magma_perf, 1000.*magma_time,
dev_perf, 1000.*dev_time,
cpu_perf, 1000.*cpu_time,
magma_error, dev_error,
(okay ? "ok" : "failed"));
#else
bool okay = (dev_error < tol);
status += ! okay;
printf("%5lld %5lld %5lld %7.2f (%7.2f) %7.2f (%7.2f) %8.2e %s\n",
(long long) M, (long long) N, (long long) K,
dev_perf, 1000.*dev_time,
cpu_perf, 1000.*cpu_time,
dev_error,
(okay ? "ok" : "failed"));
#endif
}
else {
#if defined(MAGMA_HAVE_CUDA) || defined(MAGMA_HAVE_HIP)
// use cuBLAS for R_ref (currently only with CUDA)
blasf77_zaxpy( &sizeC, &c_neg_one, hCdev, &ione, hCmagma, &ione );
magma_error = lapackf77_zlange( "F", &M, &N, hCmagma, &ldc, work )
/ (sqrt(double(K+2))*fabs(alpha)*Anorm*Bnorm + 2*fabs(beta)*Cnorm);
bool okay = (magma_error < tol);
status += ! okay;
printf("%5lld %5lld %5lld %7.2f (%7.2f) %7.2f (%7.2f) --- ( --- ) %8.2e --- %s\n",
(long long) M, (long long) N, (long long) K,
magma_perf, 1000.*magma_time,
dev_perf, 1000.*dev_time,
magma_error,
(okay ? "ok" : "failed"));
#else
printf("%5lld %5lld %5lld %7.2f (%7.2f) --- ( --- ) ---\n",
(long long) M, (long long) N, (long long) K,
dev_perf, 1000.*dev_time );
#endif
}
magma_free_cpu( hA );
magma_free_cpu( hB );
magma_free_cpu( hC );
magma_free_cpu( hCmagma );
magma_free_cpu( hCdev );
magma_free( dA );
magma_free( dB );
magma_free( dC );
fflush( stdout );
}
if ( opts.niter > 1 ) {
printf( "\n" );
}
}
opts.cleanup();
TESTING_CHECK( magma_finalize() );
return status;
}
|
