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
|
// Copyright (c) 2017-2023, University of Tennessee. All rights reserved.
// SPDX-License-Identifier: BSD-3-Clause
// This program is free software: you can redistribute it and/or modify it under
// the terms of the BSD 3-Clause license. See the accompanying LICENSE file.
#include "test.hh"
#include "lapack.hh"
#include "lapack/flops.hh"
#include "print_matrix.hh"
#include "error.hh"
#include "lapacke_wrappers.hh"
#include <vector>
// -----------------------------------------------------------------------------
template< typename scalar_t >
void test_larfg_work( Params& params, bool run )
{
using blas::real;
using blas::imag;
using real_t = blas::real_type< scalar_t >;
// get & mark input values
int64_t n = params.dim.n();
int64_t incx = params.incx();
scalar_t alpha_tst = params.alpha.get<scalar_t>();
scalar_t alpha_ref = alpha_tst;
int64_t verbose = params.verbose();
// mark non-standard output values
params.ref_time();
params.ref_gflops();
params.gflops();
if (! run)
return;
// ---------- setup
scalar_t tau_tst;
scalar_t tau_ref;
size_t size_X = (size_t) (1+(n-2)*std::abs(incx));
std::vector< scalar_t > X_tst( size_X );
std::vector< scalar_t > X_ref( size_X );
int64_t idist = 1;
int64_t iseed[4] = { 0, 1, 2, 3 };
lapack::larnv( idist, iseed, X_tst.size(), &X_tst[0] );
X_ref = X_tst;
if (verbose >= 1) {
printf( "x incx %lld, size %lld\n", llong( incx ), llong( size_X ) );
}
if (verbose >= 2) {
printf( "alpha = %.4e + %.4ei\n", real(alpha_tst), imag(alpha_tst) );
printf( "x = " ); print_vector( n-1, &X_tst[0], incx );
printf( "xref = " ); print_vector( n-1, &X_ref[0], incx );
}
// ---------- run test
testsweeper::flush_cache( params.cache() );
double time = testsweeper::get_wtime();
lapack::larfg( n, &alpha_tst, &X_tst[0], incx, &tau_tst );
time = testsweeper::get_wtime() - time;
params.time() = time;
double gflop = lapack::Gflop< scalar_t >::larfg( n );
params.gflops() = gflop / time;
if (verbose >= 2) {
printf( "alpha2 = %.4e\n", real(alpha_tst) );
printf( "x2 = " ); print_vector( n-1, &X_tst[0], incx );
printf( "tau = %.4e\n", real(tau_tst) );
}
if (params.ref() == 'y' || params.check() == 'y') {
// ---------- run reference
testsweeper::flush_cache( params.cache() );
time = testsweeper::get_wtime();
int64_t info_ref = LAPACKE_larfg( n, &alpha_ref, &X_ref[0], incx, &tau_ref );
time = testsweeper::get_wtime() - time;
if (info_ref != 0) {
fprintf( stderr, "LAPACKE_larfg returned error %lld\n", llong( info_ref ) );
}
params.ref_time() = time;
params.ref_gflops() = gflop / time;
if (verbose >= 2) {
printf( "alpha2ref = %.4e\n", real(alpha_ref) );
printf( "x2ref = " ); print_vector( n-1, &X_ref[0], incx );
printf( "tau_ref = %.4e\n", real(tau_ref) );
}
// ---------- check error compared to reference
real_t error = 0;
error += std::abs( alpha_tst - alpha_ref );
error += abs_error( X_tst, X_ref );
error += std::abs( tau_tst - tau_ref );
params.error() = error;
params.okay() = (error == 0); // expect lapackpp == lapacke
}
}
// -----------------------------------------------------------------------------
void test_larfg( Params& params, bool run )
{
switch (params.datatype()) {
case testsweeper::DataType::Single:
test_larfg_work< float >( params, run );
break;
case testsweeper::DataType::Double:
test_larfg_work< double >( params, run );
break;
case testsweeper::DataType::SingleComplex:
test_larfg_work< std::complex<float> >( params, run );
break;
case testsweeper::DataType::DoubleComplex:
test_larfg_work< std::complex<double> >( params, run );
break;
default:
throw std::runtime_error( "unknown datatype" );
break;
}
}
|
