1 /*
2 * Copyright (c) 2009, Michael Lehn
3 *
4 * All rights reserved.
5 *
6 * Redistribution and use in source and binary forms, with or without
7 * modification, are permitted provided that the following conditions
8 * are met:
9 *
10 * 1) Redistributions of source code must retain the above copyright
11 * notice, this list of conditions and the following disclaimer.
12 * 2) Redistributions in binary form must reproduce the above copyright
13 * notice, this list of conditions and the following disclaimer in
14 * the documentation and/or other materials provided with the
15 * distribution.
16 * 3) Neither the name of the FLENS development group nor the names of
17 * its contributors may be used to endorse or promote products derived
18 * from this software without specific prior written permission.
19 *
20 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
21 * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
22 * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
23 * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
24 * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
25 * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
26 * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
27 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
28 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
29 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
30 * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
31 */
32
33 #ifndef CXXBLAS_LEVEL2_HBMV_TCC
34 #define CXXBLAS_LEVEL2_HBMV_TCC 1
35
36 #include <complex>
37 #include <cxxblas/level1/level1.h>
38
39 namespace cxxblas {
40
41 template <typename IndexType, typename ALPHA, typename MA, typename VX,
42 typename BETA, typename VY>
43 void
44 hbmv_generic(StorageOrder order, StorageUpLo upLo, Transpose conjugateA,
45 IndexType n, IndexType k,
46 const ALPHA &alpha,
47 const MA *A, IndexType ldA,
48 const VX *x, IndexType incX,
49 const BETA &beta,
50 VY *y, IndexType incY)
51 {
52 using std::max;
53 using std::min;
54 // TODO: using cxxblas::real <- causes an error
55
56 if (order==ColMajor) {
57 upLo = (upLo==Upper) ? Lower : Upper;
58 conjugateA = Transpose(conjugateA^Conj);
59 }
60 scal_generic(n, beta, y, incY);
61 if (upLo==Upper) {
62 if (conjugateA==Conj) {
63 for (IndexType i=0, iX=0, iY=0; i<n; ++i, iX+=incX, iY+=incY) {
64 IndexType len = min(k+1, n-i);
65
66 y[iY] += alpha*cxxblas::real(A[ldA*i])*x[iX];
67
68 VY _y;
69 dot_generic(len-1, A+ldA*i+1, IndexType(1),
70 x+iX+incX, IndexType(incX),
71 _y);
72 y[iY] += alpha*_y;
73
74 axpy_generic(len-1, x[iX] * alpha,
75 A+ldA*i+1, IndexType(1),
76 y+iY+incY, incY);
77 }
78 } else {
79 for (IndexType i=0, iX=0, iY=0; i<n; ++i, iX+=incX, iY+=incY) {
80 IndexType len = min(k+1, n-i);
81
82 y[iY] += alpha*cxxblas::real(A[ldA*i])*x[iX];
83
84 VY _y;
85 dotu_generic(len-1, A+ldA*i+1, IndexType(1),
86 x+iX+incX, IndexType(incX),
87 _y);
88 y[iY] += alpha*_y;
89
90 acxpy_generic(len-1, x[iX] * alpha,
91 A+ldA*i+1, IndexType(1),
92 y+iY+incY, incY);
93 }
94 }
95 } else { /* upLo==Lower */
96 if (conjugateA==Conj) {
97 for (IndexType i=0, iY=0; i<n; ++i, iY+=incY) {
98 IndexType iA = max(k-i, IndexType(0));
99 IndexType len = min(k, i) + 1;
100 IndexType _i = max(i-k, IndexType(0));
101
102 y[iY] += alpha*cxxblas::real(A[ldA*i+iA+ len-1])*x[i*incX];
103
104 VY _y;
105 dot_generic(len-1, A+ldA*i+iA, IndexType(1),
106 x+_i*incX, IndexType(incX),
107 _y);
108 y[iY] += alpha*_y;
109
110 axpy_generic(len-1, x[i*incX] * alpha,
111 A+ldA*i+iA, IndexType(1),
112 y+_i*incY, incY);
113 }
114 } else {
115 for (IndexType i=0, iY=0; i<n; ++i, iY+=incY) {
116 IndexType iA = max(k-i, IndexType(0));
117 IndexType len = min(k, i) + 1;
118 IndexType _i = max(i-k, IndexType(0));
119
120 y[iY] += alpha*cxxblas::real(A[ldA*i+iA+ len-1])*x[i*incX];
121
122 VY _y;
123 dotu_generic(len-1, A+ldA*i+iA, IndexType(1),
124 x+_i*incX, IndexType(incX),
125 _y);
126 y[iY] += alpha*_y;
127
128 acxpy_generic(len-1, x[i*incX] * alpha,
129 A+ldA*i+iA, IndexType(1),
130 y+_i*incY, incY);
131 }
132 }
133 }
134 }
135
136 //------------------------------------------------------------------------------
137
138 template <typename IndexType, typename ALPHA, typename MA, typename VX,
139 typename BETA, typename VY>
140 void
141 hbmv(StorageOrder order, StorageUpLo upLo,
142 IndexType n, IndexType k,
143 const ALPHA &alpha,
144 const MA *A, IndexType ldA,
145 const VX *x, IndexType incX,
146 const BETA &beta,
147 VY *y, IndexType incY)
148 {
149 CXXBLAS_DEBUG_OUT("hbmv_generic");
150
151 if (n==0) {
152 return;
153 }
154 if (incX<0) {
155 x -= incX*(n-1);
156 }
157 if (incY<0) {
158 y -= incY*(n-1);
159 }
160 hbmv_generic(order, upLo, NoTrans,
161 n, k,
162 alpha, A, ldA,
163 x, incX,
164 beta,
165 y, incY);
166 }
167
168 } // namespace cxxblas
169
170 #endif // CXXBLAS_LEVEL2_HBMV_TCC
2 * Copyright (c) 2009, Michael Lehn
3 *
4 * All rights reserved.
5 *
6 * Redistribution and use in source and binary forms, with or without
7 * modification, are permitted provided that the following conditions
8 * are met:
9 *
10 * 1) Redistributions of source code must retain the above copyright
11 * notice, this list of conditions and the following disclaimer.
12 * 2) Redistributions in binary form must reproduce the above copyright
13 * notice, this list of conditions and the following disclaimer in
14 * the documentation and/or other materials provided with the
15 * distribution.
16 * 3) Neither the name of the FLENS development group nor the names of
17 * its contributors may be used to endorse or promote products derived
18 * from this software without specific prior written permission.
19 *
20 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
21 * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
22 * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
23 * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
24 * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
25 * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
26 * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
27 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
28 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
29 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
30 * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
31 */
32
33 #ifndef CXXBLAS_LEVEL2_HBMV_TCC
34 #define CXXBLAS_LEVEL2_HBMV_TCC 1
35
36 #include <complex>
37 #include <cxxblas/level1/level1.h>
38
39 namespace cxxblas {
40
41 template <typename IndexType, typename ALPHA, typename MA, typename VX,
42 typename BETA, typename VY>
43 void
44 hbmv_generic(StorageOrder order, StorageUpLo upLo, Transpose conjugateA,
45 IndexType n, IndexType k,
46 const ALPHA &alpha,
47 const MA *A, IndexType ldA,
48 const VX *x, IndexType incX,
49 const BETA &beta,
50 VY *y, IndexType incY)
51 {
52 using std::max;
53 using std::min;
54 // TODO: using cxxblas::real <- causes an error
55
56 if (order==ColMajor) {
57 upLo = (upLo==Upper) ? Lower : Upper;
58 conjugateA = Transpose(conjugateA^Conj);
59 }
60 scal_generic(n, beta, y, incY);
61 if (upLo==Upper) {
62 if (conjugateA==Conj) {
63 for (IndexType i=0, iX=0, iY=0; i<n; ++i, iX+=incX, iY+=incY) {
64 IndexType len = min(k+1, n-i);
65
66 y[iY] += alpha*cxxblas::real(A[ldA*i])*x[iX];
67
68 VY _y;
69 dot_generic(len-1, A+ldA*i+1, IndexType(1),
70 x+iX+incX, IndexType(incX),
71 _y);
72 y[iY] += alpha*_y;
73
74 axpy_generic(len-1, x[iX] * alpha,
75 A+ldA*i+1, IndexType(1),
76 y+iY+incY, incY);
77 }
78 } else {
79 for (IndexType i=0, iX=0, iY=0; i<n; ++i, iX+=incX, iY+=incY) {
80 IndexType len = min(k+1, n-i);
81
82 y[iY] += alpha*cxxblas::real(A[ldA*i])*x[iX];
83
84 VY _y;
85 dotu_generic(len-1, A+ldA*i+1, IndexType(1),
86 x+iX+incX, IndexType(incX),
87 _y);
88 y[iY] += alpha*_y;
89
90 acxpy_generic(len-1, x[iX] * alpha,
91 A+ldA*i+1, IndexType(1),
92 y+iY+incY, incY);
93 }
94 }
95 } else { /* upLo==Lower */
96 if (conjugateA==Conj) {
97 for (IndexType i=0, iY=0; i<n; ++i, iY+=incY) {
98 IndexType iA = max(k-i, IndexType(0));
99 IndexType len = min(k, i) + 1;
100 IndexType _i = max(i-k, IndexType(0));
101
102 y[iY] += alpha*cxxblas::real(A[ldA*i+iA+ len-1])*x[i*incX];
103
104 VY _y;
105 dot_generic(len-1, A+ldA*i+iA, IndexType(1),
106 x+_i*incX, IndexType(incX),
107 _y);
108 y[iY] += alpha*_y;
109
110 axpy_generic(len-1, x[i*incX] * alpha,
111 A+ldA*i+iA, IndexType(1),
112 y+_i*incY, incY);
113 }
114 } else {
115 for (IndexType i=0, iY=0; i<n; ++i, iY+=incY) {
116 IndexType iA = max(k-i, IndexType(0));
117 IndexType len = min(k, i) + 1;
118 IndexType _i = max(i-k, IndexType(0));
119
120 y[iY] += alpha*cxxblas::real(A[ldA*i+iA+ len-1])*x[i*incX];
121
122 VY _y;
123 dotu_generic(len-1, A+ldA*i+iA, IndexType(1),
124 x+_i*incX, IndexType(incX),
125 _y);
126 y[iY] += alpha*_y;
127
128 acxpy_generic(len-1, x[i*incX] * alpha,
129 A+ldA*i+iA, IndexType(1),
130 y+_i*incY, incY);
131 }
132 }
133 }
134 }
135
136 //------------------------------------------------------------------------------
137
138 template <typename IndexType, typename ALPHA, typename MA, typename VX,
139 typename BETA, typename VY>
140 void
141 hbmv(StorageOrder order, StorageUpLo upLo,
142 IndexType n, IndexType k,
143 const ALPHA &alpha,
144 const MA *A, IndexType ldA,
145 const VX *x, IndexType incX,
146 const BETA &beta,
147 VY *y, IndexType incY)
148 {
149 CXXBLAS_DEBUG_OUT("hbmv_generic");
150
151 if (n==0) {
152 return;
153 }
154 if (incX<0) {
155 x -= incX*(n-1);
156 }
157 if (incY<0) {
158 y -= incY*(n-1);
159 }
160 hbmv_generic(order, upLo, NoTrans,
161 n, k,
162 alpha, A, ldA,
163 x, incX,
164 beta,
165 y, incY);
166 }
167
168 } // namespace cxxblas
169
170 #endif // CXXBLAS_LEVEL2_HBMV_TCC