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
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
|
PROGRAM CTIMEE
*
* -- LAPACK timing routine (version 3.0) --
* Univ. of Tennessee, Univ. of California Berkeley, NAG Ltd.,
* Courant Institute, Argonne National Lab, and Rice University
* June 30, 1999
*
* Purpose
* =======
*
* CTIMEE is the main timing program for the COMPLEX matrix
* eigenvalue routines in LAPACK.
*
* There are four sets of routines that can be timed:
*
* NEP (Nonsymmetric Eigenvalue Problem):
* Includes CGEHRD, CHSEQR, CTREVC, and CHSEIN
*
* SEP (Hermitian Eigenvalue Problem):
* Includes CHETRD, CSTEQR, and SSTERF
*
* SVD (Singular Value Decomposition):
* Includes CGEBRD, CUNGBR, CBDSQR, and CGESDD
*
* GEP (Generalized nonsymmetric Eigenvalue Problem):
* Includes CGGHRD, CHGEQZ, and CTGEVC
*
* Each test path has a different input file. The first line of the
* input file should contain the characters NEP, SEP, SVD, or GEP in
* columns 1-3. The number of remaining lines depends on what is found
* on the first line.
*
*-----------------------------------------------------------------------
*
* NEP input file:
*
* line 2: NN, INTEGER
* Number of values of N.
*
* line 3: NVAL, INTEGER array, dimension (NN)
* The values for the matrix dimension N.
*
* line 4: NPARM, INTEGER
* Number of values of the parameters NB, NS, MAXB, and LDA.
*
* line 5: NBVAL, INTEGER array, dimension (NPARM)
* The values for the blocksize NB.
*
* line 6: NSVAL, INTEGER array, dimension (NPARM)
* The values for the number of shifts.
*
* line 7: MXBVAL, INTEGER array, dimension (NPARM)
* The values for MAXB, used in determining whether multishift
* will be used.
*
* line 8: LDAVAL, INTEGER array, dimension (NPARM)
* The values for the leading dimension LDA.
*
* line 9: TIMMIN, REAL
* The minimum time (in seconds) that a subroutine will be
* timed. If TIMMIN is zero, each routine should be timed only
* once.
*
* line 10: NTYPES, INTEGER
* The number of matrix types to be used in the timing run.
* If NTYPES >= MAXTYP, all the types are used.
*
* If 0 < NTYPES < MAXTYP, then line 11 specifies NTYPES integer
* values, which are the numbers of the matrix types to be used.
*
* The remaining lines specify a path name and the specific routines to
* be timed. For the nonsymmetric eigenvalue problem, the path name is
* 'CHS'. A line to request all the routines in this path has the form
* CHS T T T T T T T T T T T T
* where the first 3 characters specify the path name, and up to MAXTYP
* nonblank characters may appear in columns 4-80. If the k-th such
* character is 'T' or 't', the k-th routine will be timed. If at least
* one but fewer than 12 nonblank characters are specified, the
* remaining routines will not be timed. If columns 4-80 are blank, all
* the routines will be timed, so the input line
* CHS
* is equivalent to the line above.
*
*-----------------------------------------------------------------------
*
* SEP input file:
*
* line 2: NN, INTEGER
* Number of values of N.
*
* line 3: NVAL, INTEGER array, dimension (NN)
* The values for the matrix dimension N.
*
* line 4: NPARM, INTEGER
* Number of values of the parameters NB and LDA.
*
* line 5: NBVAL, INTEGER array, dimension (NPARM)
* The values for the blocksize NB.
*
* line 6: LDAVAL, INTEGER array, dimension (NPARM)
* The values for the leading dimension LDA.
*
* line 7: TIMMIN, REAL
* The minimum time (in seconds) that a subroutine will be
* timed. If TIMMIN is zero, each routine should be timed only
* once.
*
* line 8: NTYPES, INTEGER
* The number of matrix types to be used in the timing run.
* If NTYPES >= MAXTYP, all the types are used.
*
* If 0 < NTYPES < MAXTYP, then line 9 specifies NTYPES integer
* values, which are the numbers of the matrix types to be used.
*
* The remaining lines specify a path name and the specific routines to
* be timed as for the NEP input file. For the symmetric eigenvalue
* problem, the path name is 'CST' and up to 8 routines may be timed.
*
*-----------------------------------------------------------------------
*
* SVD input file:
*
* line 2: NN, INTEGER
* Number of values of M and N.
*
* line 3: MVAL, INTEGER array, dimension (NN)
* The values for the matrix dimension M.
*
* line 4: NVAL, INTEGER array, dimension (NN)
* The values for the matrix dimension N.
*
* line 5: NPARM, INTEGER
* Number of values of the parameters NB and LDA.
*
* line 6: NBVAL, INTEGER array, dimension (NPARM)
* The values for the blocksize NB.
*
* line 7: LDAVAL, INTEGER array, dimension (NPARM)
* The values for the leading dimension LDA.
*
* line 8: TIMMIN, REAL
* The minimum time (in seconds) that a subroutine will be
* timed. If TIMMIN is zero, each routine should be timed only
* once.
*
* line 9: NTYPES, INTEGER
* The number of matrix types to be used in the timing run.
* If NTYPES >= MAXTYP, all the types are used.
*
* If 0 < NTYPES < MAXTYP, then line 10 specifies NTYPES integer
* values, which are the numbers of the matrix types to be used.
*
* The remaining lines specify a path name and the specific routines to
* be timed as for the NEP input file. For the singular value
* decomposition the path name is 'CBD' and up to 16 routines may be
* timed.
*
*-----------------------------------------------------------------------
*
* GEP input file:
*
* line 2: NN, INTEGER
* Number of values of N.
*
* line 3: NVAL, INTEGER array, dimension (NN)
* The values for the matrix dimension N.
*
* line 4: NPARM, INTEGER
* Number of values of the parameters NB, NS, MAXB, and LDA.
*
* line 5: NBVAL, INTEGER array, dimension (NPARM)
* The values for the blocksize NB.
*
* line 6: NSVAL, INTEGER array, dimension (NPARM)
* The values for the number of shifts.
*
* line 7: NEIVAL, INTEGER array, dimension (NPARM)
* The values for NEISP, used in determining whether multishift
* will be used.
*
* line 8: NBMVAL, INTEGER array, dimension (NPARM)
* The values for MINNB, used in determining minimum blocksize.
*
* line 9: NBKVAL, INTEGER array, dimension (NPARM)
* The values for MINBLK, also used in determining minimum
* blocksize.
*
* line 10: LDAVAL, INTEGER array, dimension (NPARM)
* The values for the leading dimension LDA.
*
* line 11: TIMMIN, REAL
* The minimum time (in seconds) that a subroutine will be
* timed. If TIMMIN is zero, each routine should be timed only
* once.
*
* line 12: NTYPES, INTEGER
* The number of matrix types to be used in the timing run.
* If NTYPES >= MAXTYP, all the types are used.
*
* If 0 < NTYPES < MAXTYP, then line 13 specifies NTYPES integer
* values, which are the numbers of the matrix types to be used.
*
* The remaining lines specify a path name and the specific routines to
* be timed. For the nonsymmetric eigenvalue problem, the path name is
* 'CHG'. A line to request all the routines in this path has the form
* CHG T T T T T T T T T T T T T T T T T T
* where the first 3 characters specify the path name, and up to MAXTYP
* nonblank characters may appear in columns 4-80. If the k-th such
* character is 'T' or 't', the k-th routine will be timed. If at least
* one but fewer than 18 nonblank characters are specified, the
* remaining routines will not be timed. If columns 4-80 are blank, all
* the routines will be timed, so the input line
* CHG
* is equivalent to the line above.
*
*=======================================================================
*
* The workspace requirements in terms of square matrices for the
* different test paths are as follows:
*
* NEP: 3 N**2 + N*(3*NB+2)
* SEP: 2 N**2 + N*(2*N) + N
* SVD: 4 N**2 + MAX( 6*N, MAXIN*MAXPRM*MAXT )
* GEP: 6 N**2 + 3*N
*
* MAXN is currently set to 400,
* LG2MXN = ceiling of log-base-2 of MAXN = 9, and LDAMAX = 420.
* The real work space needed is LWORK = MAX( MAXN*(4*MAXN+2),
* 2*LDAMAX+1+3*MAXN+2*MAXN*LG2MXN+3*MAXN**2 ), and the integer
* workspace needed is LIWRK2 = 6 + 6*MAXN + 5*MAXN*LG2MXN.
* For SVD, we assume NRHS may be as big
* as N. The parameter NEED is set to 6 to allow for 4 NxN matrices
* for GEP.
*
* The EISPACK routines tested use two real arrays to represent complex
* data, whereas the LAPACK routines use complex arrays. The LAPACK
* arrays are called A, D, and WORK and the corresponding EISPACK arrays
* are called AR, DR, and WORKR. To conserve space, we have
* EQUIVALENCEd the real arrays to their complex analogs.
* !!!*** Note ***!!!
* This EQUIVALENCE is a violation of the FORTRAN-77 standard because
* the equivalenced arrays are both passed to a subroutine and both
* modified there. Most compilers will permit this, but if not, users
* are advised to comment out these EQUIVALENCE statements in the code
* below.
*
* The work arrays RWORK and RWORK1 are also equivalenced to get the
* effect of one array with length = max( MAXN, MAXIN*MAXT*MAXPRM ).
* Generally, RWORK1 will be the larger, and even if MAXN is so large as
* to make RWORK larger, most machines will not complain. If this
* becomes a problem, though, pass RWORK instead of RWORK1 to CTIM21.
*
*
* .. Parameters ..
INTEGER MAXN, LDAMAX, LG2MXN
PARAMETER ( MAXN = 400, LDAMAX = MAXN+20, LG2MXN = 9 )
INTEGER NEED
PARAMETER ( NEED = 6 )
INTEGER LIWRK2
PARAMETER ( LIWRK2 = 6+6*MAXN+5*MAXN*LG2MXN )
INTEGER LWORK
PARAMETER ( LWORK = 2*LDAMAX+1+3*MAXN+2*MAXN*LG2MXN+
$ 4*MAXN**2 )
INTEGER MAXIN, MAXPRM, MAXT, MAXSUB
PARAMETER ( MAXIN = 12, MAXPRM = 10, MAXT = 10,
$ MAXSUB = 20 )
INTEGER LRWORK
PARAMETER ( LRWORK = MAXIN*MAXT*MAXPRM+
$ 1+3*MAXN+2*MAXN*LG2MXN+3*MAXN**2 )
INTEGER NIN, NOUT
PARAMETER ( NIN = 5, NOUT = 6 )
* ..
* .. Local Scalars ..
LOGICAL FATAL, GEP, NEP, SEP, SVD
CHARACTER*3 C3, PATH
CHARACTER*6 VNAME
CHARACTER*80 LINE
INTEGER I, INFO, MAXTYP, NN, NPARMS, NTYPES
REAL S1, S2, TIMMIN
* ..
* .. Local Arrays ..
LOGICAL DOTYPE( MAXT ), LOGWRK( MAXN )
INTEGER ISEED( 4 ), IWORK( MAXT ), IWORK2( LIWRK2 ),
$ LDAVAL( MAXPRM ), MVAL( MAXIN ),
$ MXBVAL( MAXPRM ), MXTYPE( 4 ),
$ NBKVAL( MAXPRM ), NBMVAL( MAXPRM ),
$ NBVAL( MAXPRM ), NSVAL( MAXPRM ), NVAL( MAXIN )
REAL AR( LDAMAX*MAXN, 2*NEED ), DR( MAXN, 2*NEED ),
$ E2( MAXN ), OPCNTS( MAXPRM, MAXT, MAXIN,
$ MAXSUB ), RESULT( MAXPRM, MAXT, MAXIN,
$ MAXSUB ), RWORK( 6*MAXN ), RWORK1( LRWORK ),
$ WORKR( LWORK, 2 )
COMPLEX A( LDAMAX*MAXN, NEED ), D( MAXN, NEED ),
$ WORK( LWORK )
* ..
* .. External Functions ..
LOGICAL LSAMEN
REAL SECOND
EXTERNAL LSAMEN, SECOND
* ..
* .. External Subroutines ..
EXTERNAL CTIM21, CTIM22, CTIM26, CTIM51
* ..
* .. Scalars in Common ..
REAL ITCNT, OPS
* ..
* .. Arrays in Common ..
INTEGER IPARMS( 100 )
* ..
* .. Common blocks ..
COMMON / CLAENV / IPARMS
COMMON / LATIME / OPS, ITCNT
* ..
* .. Save statement ..
SAVE / CLAENV /
* ..
* .. Equivalences ..
EQUIVALENCE ( A, AR ), ( D, DR ), ( WORK, WORKR )
EQUIVALENCE ( RWORK, RWORK1 )
* ..
* .. Intrinsic Functions ..
INTRINSIC MAX
* ..
* .. Data statements ..
DATA ISEED / 0, 0, 0, 1 /
DATA MXTYPE / 8, 4, 5, 4 /
* ..
* .. Executable Statements ..
*
S1 = SECOND( )
FATAL = .FALSE.
NEP = .FALSE.
SEP = .FALSE.
SVD = .FALSE.
GEP = .FALSE.
*
* Read the 3-character test path
*
READ( NIN, FMT = '(A3)', END = 160 )PATH
NEP = LSAMEN( 3, PATH, 'NEP' ) .OR. LSAMEN( 3, PATH, 'CHS' )
SEP = LSAMEN( 3, PATH, 'SEP' ) .OR. LSAMEN( 3, PATH, 'CST' )
SVD = LSAMEN( 3, PATH, 'SVD' ) .OR. LSAMEN( 3, PATH, 'CBD' )
GEP = LSAMEN( 3, PATH, 'GEP' ) .OR. LSAMEN( 3, PATH, 'CHG' )
*
* Report values of parameters as they are read.
*
IF( NEP ) THEN
WRITE( NOUT, FMT = 9993 )
ELSE IF( SEP ) THEN
WRITE( NOUT, FMT = 9992 )
ELSE IF( SVD ) THEN
WRITE( NOUT, FMT = 9991 )
ELSE IF( GEP ) THEN
WRITE( NOUT, FMT = 9990 )
ELSE
WRITE( NOUT, FMT = 9996 )PATH
STOP
END IF
WRITE( NOUT, FMT = 9985 )
WRITE( NOUT, FMT = 9989 )
*
* Read the number of values of M and N.
*
READ( NIN, FMT = * )NN
IF( NN.LT.1 ) THEN
WRITE( NOUT, FMT = 9995 )'NN ', NN, 1
NN = 0
FATAL = .TRUE.
ELSE IF( NN.GT.MAXIN ) THEN
WRITE( NOUT, FMT = 9994 )'NN ', NN, MAXIN
NN = 0
FATAL = .TRUE.
END IF
*
* Read the values of M
*
READ( NIN, FMT = * )( MVAL( I ), I = 1, NN )
IF( SVD ) THEN
VNAME = ' M'
ELSE
VNAME = ' N'
END IF
DO 10 I = 1, NN
IF( MVAL( I ).LT.0 ) THEN
WRITE( NOUT, FMT = 9995 )VNAME, MVAL( I ), 0
FATAL = .TRUE.
ELSE IF( MVAL( I ).GT.MAXN ) THEN
WRITE( NOUT, FMT = 9994 )VNAME, MVAL( I ), MAXN
FATAL = .TRUE.
END IF
10 CONTINUE
*
* Read the values of N
*
IF( SVD ) THEN
WRITE( NOUT, FMT = 9988 )'M ', ( MVAL( I ), I = 1, NN )
READ( NIN, FMT = * )( NVAL( I ), I = 1, NN )
DO 20 I = 1, NN
IF( NVAL( I ).LT.0 ) THEN
WRITE( NOUT, FMT = 9995 )'N ', NVAL( I ), 0
FATAL = .TRUE.
ELSE IF( NVAL( I ).GT.MAXN ) THEN
WRITE( NOUT, FMT = 9994 )'N ', NVAL( I ), MAXN
FATAL = .TRUE.
END IF
20 CONTINUE
ELSE
DO 30 I = 1, NN
NVAL( I ) = MVAL( I )
30 CONTINUE
END IF
WRITE( NOUT, FMT = 9988 )'N ', ( NVAL( I ), I = 1, NN )
*
* Read the number of parameter values.
*
READ( NIN, FMT = * )NPARMS
IF( NPARMS.LT.1 ) THEN
WRITE( NOUT, FMT = 9995 )'NPARMS', NPARMS, 1
NPARMS = 0
FATAL = .TRUE.
ELSE IF( NPARMS.GT.MAXIN ) THEN
WRITE( NOUT, FMT = 9994 )'NPARMS', NPARMS, MAXIN
NPARMS = 0
FATAL = .TRUE.
END IF
*
* Read the values of NB
*
READ( NIN, FMT = * )( NBVAL( I ), I = 1, NPARMS )
DO 40 I = 1, NPARMS
IF( NBVAL( I ).LT.0 ) THEN
WRITE( NOUT, FMT = 9995 )'NB ', NBVAL( I ), 0
FATAL = .TRUE.
END IF
40 CONTINUE
WRITE( NOUT, FMT = 9988 )'NB ', ( NBVAL( I ), I = 1, NPARMS )
*
IF( NEP .OR. GEP ) THEN
*
* Read the values of NSHIFT
*
READ( NIN, FMT = * )( NSVAL( I ), I = 1, NPARMS )
DO 50 I = 1, NPARMS
IF( NSVAL( I ).LT.0 ) THEN
WRITE( NOUT, FMT = 9995 )'NS ', NSVAL( I ), 0
FATAL = .TRUE.
END IF
50 CONTINUE
WRITE( NOUT, FMT = 9988 )'NS ', ( NSVAL( I ), I = 1, NPARMS )
*
* Read the values of MAXB
*
READ( NIN, FMT = * )( MXBVAL( I ), I = 1, NPARMS )
DO 60 I = 1, NPARMS
IF( MXBVAL( I ).LT.0 ) THEN
WRITE( NOUT, FMT = 9995 )'MAXB', MXBVAL( I ), 0
FATAL = .TRUE.
END IF
60 CONTINUE
WRITE( NOUT, FMT = 9988 )'MAXB',
$ ( MXBVAL( I ), I = 1, NPARMS )
ELSE
DO 70 I = 1, NPARMS
NSVAL( I ) = 1
MXBVAL( I ) = 1
70 CONTINUE
END IF
*
IF( GEP ) THEN
*
* Read the values of NBMIN
*
READ( NIN, FMT = * )( NBMVAL( I ), I = 1, NPARMS )
DO 80 I = 1, NPARMS
IF( NBMVAL( I ).LT.0 ) THEN
WRITE( NOUT, FMT = 9995 )'NBMIN', NBMVAL( I ), 0
FATAL = .TRUE.
END IF
80 CONTINUE
WRITE( NOUT, FMT = 9988 )'NBMIN',
$ ( NBMVAL( I ), I = 1, NPARMS )
*
* Read the values of MINBLK
*
READ( NIN, FMT = * )( NBKVAL( I ), I = 1, NPARMS )
DO 90 I = 1, NPARMS
IF( NBKVAL( I ).LT.0 ) THEN
WRITE( NOUT, FMT = 9995 )'MINBLK', NBKVAL( I ), 0
FATAL = .TRUE.
END IF
90 CONTINUE
WRITE( NOUT, FMT = 9988 )'MINBLK',
$ ( NBKVAL( I ), I = 1, NPARMS )
ELSE
DO 100 I = 1, NPARMS
NBMVAL( I ) = MAXN + 1
NBKVAL( I ) = MAXN + 1
100 CONTINUE
END IF
*
* Read the values of LDA
*
READ( NIN, FMT = * )( LDAVAL( I ), I = 1, NPARMS )
DO 110 I = 1, NPARMS
IF( LDAVAL( I ).LT.0 ) THEN
WRITE( NOUT, FMT = 9995 )'LDA ', LDAVAL( I ), 0
FATAL = .TRUE.
ELSE IF( LDAVAL( I ).GT.LDAMAX ) THEN
WRITE( NOUT, FMT = 9994 )'LDA ', LDAVAL( I ), LDAMAX
FATAL = .TRUE.
END IF
110 CONTINUE
WRITE( NOUT, FMT = 9988 )'LDA ', ( LDAVAL( I ), I = 1, NPARMS )
*
* Read the minimum time a subroutine will be timed.
*
READ( NIN, FMT = * )TIMMIN
WRITE( NOUT, FMT = 9987 )TIMMIN
*
* Read the number of matrix types to use in timing.
*
READ( NIN, FMT = * )NTYPES
IF( NTYPES.LT.0 ) THEN
WRITE( NOUT, FMT = 9995 )'NTYPES', NTYPES, 0
FATAL = .TRUE.
NTYPES = 0
END IF
*
* Read the matrix types.
*
IF( NEP ) THEN
MAXTYP = MXTYPE( 1 )
ELSE IF( SEP ) THEN
MAXTYP = MXTYPE( 2 )
ELSE IF( SVD ) THEN
MAXTYP = MXTYPE( 3 )
ELSE
MAXTYP = MXTYPE( 4 )
END IF
IF( NTYPES.LT.MAXTYP ) THEN
READ( NIN, FMT = * )( IWORK( I ), I = 1, NTYPES )
DO 120 I = 1, MAXTYP
DOTYPE( I ) = .FALSE.
120 CONTINUE
DO 130 I = 1, NTYPES
IF( IWORK( I ).LT.0 ) THEN
WRITE( NOUT, FMT = 9995 )'TYPE', IWORK( I ), 0
FATAL = .TRUE.
ELSE IF( IWORK( I ).GT.MAXTYP ) THEN
WRITE( NOUT, FMT = 9994 )'TYPE', IWORK( I ), MAXTYP
FATAL = .TRUE.
ELSE
DOTYPE( IWORK( I ) ) = .TRUE.
END IF
130 CONTINUE
ELSE
NTYPES = MAXTYP
DO 140 I = 1, MAXT
DOTYPE( I ) = .TRUE.
140 CONTINUE
END IF
*
IF( FATAL ) THEN
WRITE( NOUT, FMT = 9999 )
9999 FORMAT( / ' Execution not attempted due to input errors' )
STOP
END IF
*
* Read the input lines indicating the test path and the routines
* to be timed. The first three characters indicate the test path.
*
150 CONTINUE
READ( NIN, FMT = '(A80)', END = 160 )LINE
C3 = LINE( 1: 3 )
*
* -------------------------------------
* NEP: Nonsymmetric Eigenvalue Problem
* -------------------------------------
*
IF( LSAMEN( 3, C3, 'CHS' ) .OR. LSAMEN( 3, C3, 'NEP' ) ) THEN
CALL CTIM21( LINE, NN, NVAL, MAXTYP, DOTYPE, NPARMS, NBVAL,
$ NSVAL, MXBVAL, LDAVAL, TIMMIN, NOUT, ISEED,
$ A( 1, 1 ), AR( 1, 1 ), AR( 1, 2 ), A( 1, 2 ),
$ AR( 1, 3 ), AR( 1, 4 ), A( 1, 3 ), AR( 1, 5 ),
$ AR( 1, 6 ), D( 1, 1 ), DR( 1, 1 ), DR( 1, 3 ),
$ WORK, WORKR( 1, 1 ), WORKR( 1, 2 ), LWORK, RWORK1,
$ LOGWRK, IWORK2, RESULT, MAXPRM, MAXT, MAXIN,
$ OPCNTS, MAXPRM, MAXT, MAXIN, INFO )
IF( INFO.NE.0 )
$ WRITE( NOUT, FMT = 9986 )'CTIM21', INFO
*
* ----------------------------------
* SEP: Symmetric Eigenvalue Problem
* ----------------------------------
*
ELSE IF( LSAMEN( 3, C3, 'CST' ) .OR. LSAMEN( 3, C3, 'SEP' ) ) THEN
CALL CTIM22( LINE, NN, NVAL, MAXTYP, DOTYPE, NPARMS, NBVAL,
$ LDAVAL, TIMMIN, NOUT, ISEED, A( 1, 1 ),
$ DR( 1, 3 ), DR( 1, 4 ), E2, A( 1, 2 ), AR( 1, 3 ),
$ AR( 1, 4 ), D( 1, 1 ), DR( 1, 1 ), A( 1, 3 ),
$ AR( 1, 5 ), AR( 1, 6 ), WORK, LWORK, RWORK1,
$ LOGWRK, IWORK2, RESULT, MAXPRM, MAXT, MAXIN,
$ OPCNTS, MAXPRM, MAXT, MAXIN, INFO )
IF( INFO.NE.0 )
$ WRITE( NOUT, FMT = 9986 )'CTIM22', INFO
*
* ----------------------------------
* SVD: Singular Value Decomposition
* ----------------------------------
*
ELSE IF( LSAMEN( 3, C3, 'CBD' ) .OR. LSAMEN( 3, C3, 'SVD' ) ) THEN
CALL CTIM26( LINE, NN, NVAL, MVAL, MAXTYP, DOTYPE, NPARMS,
$ NBVAL, LDAVAL, TIMMIN, NOUT, ISEED, A( 1, 1 ),
$ A( 1, 2 ), A( 1, 3 ), A( 1, 4 ), DR( 1, 1 ),
$ D( 1, 1 ), DR( 1, 3 ), D( 1, 2 ), D( 1, 3 ),
$ D( 1, 4 ), WORK, LWORK, WORKR, IWORK, LOGWRK,
$ RESULT, MAXPRM, MAXT, MAXIN, OPCNTS, MAXPRM, MAXT,
$ MAXIN, INFO )
IF( INFO.NE.0 )
$ WRITE( NOUT, FMT = 9986 )'CTIM26', INFO
*
* -------------------------------------
* GEP: Nonsymmetric Eigenvalue Problem
* -------------------------------------
*
ELSE IF( LSAMEN( 3, C3, 'CHG' ) .OR. LSAMEN( 3, C3, 'GEP' ) ) THEN
CALL CTIM51( LINE, NN, NVAL, MAXTYP, DOTYPE, NPARMS, NBVAL,
$ NSVAL, MXBVAL, NBMVAL, NBKVAL, LDAVAL, TIMMIN,
$ NOUT, ISEED, A( 1, 1 ), AR( 1, 1 ), AR( 1, 2 ),
$ A( 1, 2 ), AR( 1, 3 ), AR( 1, 4 ), A( 1, 3 ),
$ AR( 1, 5 ), AR( 1, 6 ), A( 1, 4 ), AR( 1, 7 ),
$ AR( 1, 8 ), A( 1, 5 ), AR( 1, 9 ), AR( 1, 10 ),
$ A( 1, 6 ), AR( 1, 11 ), AR( 1, 12 ), D, DR, WORK,
$ LWORK, RWORK1, LOGWRK, RESULT, MAXPRM, MAXT,
$ MAXIN, OPCNTS, MAXPRM, MAXT, MAXIN, INFO )
IF( INFO.NE.0 )
$ WRITE( NOUT, FMT = 9986 )'CTIM51', INFO
ELSE
WRITE( NOUT, FMT = * )
WRITE( NOUT, FMT = * )
WRITE( NOUT, FMT = 9996 )C3
END IF
GO TO 150
160 CONTINUE
WRITE( NOUT, FMT = 9998 )
9998 FORMAT( / / ' End of timing run' )
S2 = SECOND( )
WRITE( NOUT, FMT = 9997 )S2 - S1
*
9997 FORMAT( ' Total time used = ', F12.2, ' seconds', / )
9996 FORMAT( 1X, A3, ': Unrecognized path name' )
9995 FORMAT( ' *** Invalid input value: ', A6, '=', I6, '; must be >=',
$ I6 )
9994 FORMAT( ' *** Invalid input value: ', A6, '=', I6, '; must be <=',
$ I6 )
9993 FORMAT( ' Timing the Nonsymmetric Eigenvalue Problem routines',
$ / ' CGEHRD, CHSEQR, CTREVC, and CHSEIN' )
9992 FORMAT( ' Timing the Symmetric Eigenvalue Problem routines',
$ / ' CHETRD, CSTEQR, and SSTERF' )
9991 FORMAT( ' Timing the Singular Value Decomposition routines',
$ / ' CGEBRD, CBDSQR, and CUNGBR ' )
9990 FORMAT( ' Timing the Generalized Eigenvalue Problem routines',
$ / ' CGGHRD, CHGEQZ, and CTGEVC ' )
9989 FORMAT( / ' The following parameter values will be used:' )
9988 FORMAT( ' Values of ', A5, ': ', 10I6, / 19X, 10I6 )
9987 FORMAT( / ' Minimum time a subroutine will be timed = ', F8.2,
$ ' seconds', / )
9986 FORMAT( ' *** Error code from ', A6, ' = ', I4 )
9985 FORMAT( / ' LAPACK VERSION 3.0, released June 30, 1999 ' )
*
* End of CTIMEE
*
END
|
