78
79
80
81
82
83
84 INTEGER KNT, LMAX, NINFO
85 DOUBLE PRECISION RMAX
86
87
88
89
90
91 DOUBLE PRECISION ZERO, ONE
92 parameter( zero = 0.0d0, one = 1.0d0 )
93 DOUBLE PRECISION TWO, FOUR
94 parameter( two = 2.0d0, four = 4.0d0 )
95
96
97 CHARACTER TRANA, TRANB
98 INTEGER I, IMA, IMB, IMLDA1, IMLDA2, IMLDB1, IMLOFF,
99 $ INFO, ISGN, ITRANA, ITRANB, J, M, N
100 DOUBLE PRECISION BIGNUM, CNRM, EPS, RES, RES1, RMUL, SCALE,
101 $ SMLNUM, TNRM, XNRM
102
103
104 INTEGER IDIM( 8 ), IVAL( 6, 6, 8 )
105 DOUBLE PRECISION A( 6, 6 ), B( 6, 6 ), C( 6, 6 ), CC( 6, 6 ),
106 $ DUM( 1 ), VM1( 3 ), VM2( 3 )
107
108
109 DOUBLE PRECISION DLAMCH, DLANGE
111
112
114
115
116 INTRINSIC abs, dble, max, sin, sqrt
117
118
119 DATA idim / 1, 2, 3, 4, 3, 3, 6, 4 /
120 DATA ival / 1, 35*0, 1, 2, 4*0, -2, 0, 28*0, 1, 5*0,
121 $ 5, 1, 2, 3*0, -8, -2, 1, 21*0, 3, 4, 4*0, -5,
122 $ 3, 4*0, 1, 2, 1, 4, 2*0, -3, -9, -1, 1, 14*0,
123 $ 1, 5*0, 2, 3, 4*0, 5, 6, 7, 21*0, 1, 5*0, 1, 3,
124 $ -4, 3*0, 2, 5, 2, 21*0, 1, 2, 4*0, -2, 0, 4*0,
125 $ 5, 6, 3, 4, 2*0, -1, -9, -5, 2, 2*0, 4*8, 5, 6,
126 $ 4*9, -7, 5, 1, 5*0, 1, 5, 2, 3*0, 2, -21, 5,
127 $ 3*0, 1, 2, 3, 4, 14*0 /
128
129
130
131
132
134 smlnum =
dlamch(
'S' )*four / eps
135 bignum = one / smlnum
136
137
138
139 vm1( 1 ) = sqrt( smlnum )
140 vm1( 2 ) = one
141 vm1( 3 ) = sqrt( bignum )
142 vm2( 1 ) = one
143 vm2( 2 ) = one + two*eps
144 vm2( 3 ) = two
145
146 knt = 0
147 ninfo = 0
148 lmax = 0
149 rmax = zero
150
151
152
153 DO 150 itrana = 1, 2
154 DO 140 itranb = 1, 2
155 DO 130 isgn = -1, 1, 2
156 DO 120 ima = 1, 8
157 DO 110 imlda1 = 1, 3
158 DO 100 imlda2 = 1, 3
159 DO 90 imloff = 1, 2
160 DO 80 imb = 1, 8
161 DO 70 imldb1 = 1, 3
162 IF( itrana.EQ.1 )
163 $ trana = 'N'
164 IF( itrana.EQ.2 )
165 $ trana = 'T'
166 IF( itranb.EQ.1 )
167 $ tranb = 'N'
168 IF( itranb.EQ.2 )
169 $ tranb = 'T'
170 m = idim( ima )
171 n = idim( imb )
172 tnrm = zero
173 DO 20 i = 1, m
174 DO 10 j = 1, m
175 a( i, j ) = ival( i, j, ima )
176 IF( abs( i-j ).LE.1 ) THEN
177 a( i, j ) = a( i, j )*
178 $ vm1( imlda1 )
179 a( i, j ) = a( i, j )*
180 $ vm2( imlda2 )
181 ELSE
182 a( i, j ) = a( i, j )*
183 $ vm1( imloff )
184 END IF
185 tnrm = max( tnrm,
186 $ abs( a( i, j ) ) )
187 10 CONTINUE
188 20 CONTINUE
189 DO 40 i = 1, n
190 DO 30 j = 1, n
191 b( i, j ) = ival( i, j, imb )
192 IF( abs( i-j ).LE.1 ) THEN
193 b( i, j ) = b( i, j )*
194 $ vm1( imldb1 )
195 ELSE
196 b( i, j ) = b( i, j )*
197 $ vm1( imloff )
198 END IF
199 tnrm = max( tnrm,
200 $ abs( b( i, j ) ) )
201 30 CONTINUE
202 40 CONTINUE
203 cnrm = zero
204 DO 60 i = 1, m
205 DO 50 j = 1, n
206 c( i, j ) = sin( dble( i*j ) )
207 cnrm = max( cnrm, c( i, j ) )
208 cc( i, j ) = c( i, j )
209 50 CONTINUE
210 60 CONTINUE
211 knt = knt + 1
212 CALL dtrsyl( trana, tranb, isgn, m, n,
213 $ a, 6, b, 6, c, 6, scale,
214 $ info )
215 IF( info.NE.0 )
216 $ ninfo = ninfo + 1
217 xnrm =
dlange(
'M', m, n, c, 6, dum )
218 rmul = one
219 IF( xnrm.GT.one .AND. tnrm.GT.one )
220 $ THEN
221 IF( xnrm.GT.bignum / tnrm ) THEN
222 rmul = one / max( xnrm, tnrm )
223 END IF
224 END IF
225 CALL dgemm( trana,
'N', m, n, m, rmul,
226 $ a, 6, c, 6, -scale*rmul,
227 $ cc, 6 )
228 CALL dgemm(
'N', tranb, m, n, n,
229 $ dble( isgn )*rmul, c, 6, b,
230 $ 6, one, cc, 6 )
231 res1 =
dlange(
'M', m, n, cc, 6, dum )
232 res = res1 / max( smlnum, smlnum*xnrm,
233 $ ( ( rmul*tnrm )*eps )*xnrm )
234 IF( res.GT.rmax ) THEN
235 lmax = knt
236 rmax = res
237 END IF
238 70 CONTINUE
239 80 CONTINUE
240 90 CONTINUE
241 100 CONTINUE
242 110 CONTINUE
243 120 CONTINUE
244 130 CONTINUE
245 140 CONTINUE
246 150 CONTINUE
247
248 RETURN
249
250
251
subroutine dgemm(transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc)
DGEMM
double precision function dlamch(cmach)
DLAMCH
double precision function dlange(norm, m, n, a, lda, work)
DLANGE returns the value of the 1-norm, Frobenius norm, infinity-norm, or the largest absolute value ...
subroutine dtrsyl(trana, tranb, isgn, m, n, a, lda, b, ldb, c, ldc, scale, info)
DTRSYL