-
Notifications
You must be signed in to change notification settings - Fork 3
/
ChangeLog
5572 lines (3484 loc) · 160 KB
/
ChangeLog
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
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
2006-04-05 03:42 leif
* [r2670] Added a "make check" target which runs several of the
examples. Deleted the redundant 'pyre' directory because it was
confusing 'python'.
2006-04-01 01:54 leif
* [r2660] Allow building without MPI -- i.e., with a PETSc
configured --with-mpi=0 and a Pythia configured --without-mpi.
This means not asking for Pythia's 'mpi' module (trivial), and
checking for PETSc includes and libs instead of checking for MPI
directly (this involved writing new PETSc Autoconf macros).
Also, factored-out the Fortran I/O Autoconf tests into the CIG
Autoconf archive, to make 'configure.ac' pretty. Prevent Libtool
from looking for another F77 compiler by AC_PROVIDE-ing
AC_PROG_F77.
2006-03-29 04:10 leif
* [r2654] Fix for crash on 64-bit systems with ifort: use 'size_t'
instead of 'int' for the implicit string length parameter of
Fortran functions.
2006-03-28 19:26 walter
* [r2653] Changed externals from svn:// to http://
2006-03-28 02:39 leif
* [r2643] Added new file 'binio.h' from previous checkin to
Makefile.am.
2006-03-25 04:22 leif
* [r2642] I/O improvements: automatically configure Fortran
"append" mode and binary I/O, even under 'gfortran'... this
means deferring some configuration decisions until runtime; see
'open_append.F' and 'try_binio.F'. Also, validate I/O pathnames
in Python: if _any_ input/output files cannot be opened, dump a
concise error report detailing _all_ the files that couldn't be
opened, and what the errors were. As an added bonus, I
macro-ized the pathnames in the inventory, and tightened up
exceptionhandler() a little.
2006-02-25 03:26 leif
* [r2618] Fixed a bunch of bugs in the bindings.
2006-02-23 21:12 leif
* [r2617] Added the capability to embed Python instead of
extending it; this creates one, big monolithic executable,
'pylithomop3d', which contains all Lithomop code and an embedded
Python interpreter. Embedding works-around MPI portability
issues. Embed by default; '--with-embedding=no" to disable.
2006-02-22 01:21 leif
* [r2615] Avoid PETSC_*_BASIC variables; they are apparently for
internal use only. Also, 'sed' PYTHON into shebang line of
top-level Python script. (Lithomop now installs successfully on
the Pangu cluster.)
2006-02-17 00:13 leif
* [r2611] Bake 'PYTHONPATH' into lithomop3dapp.py, so that users
don't have to set it. Also, install Python scripts into
'pyexecdir' instead of 'pythondir', since this package contains
a native module... on systems where 'pyexecdir' and 'pythondir'
are distinct, this is critical.
2005-12-01 16:43 willic3
* [r2508] Added $(FCLIBS) to list for lithomop3dmodule_la_LIBADD.
Without this, I was getting missing g95 symbols on Mac OS X.
This will need to be tested on other platforms.
2005-11-21 20:59 willic3
* [r2503] This file has been moved to 'makefile.test' to avoid
conflicts with the new build procedure.
2005-11-16 22:02 willic3
* [r2499] Changed the version number of LithoMop from 1.0.0 to
0.7.1, since I think we're still not ready for a 1.0 release.
2005-10-21 14:02 willic3
* [r2465] New directory to hold code change summary information.
2005-10-21 14:00 willic3
* [r2464] Moved all files describing code history to info
directory. It might make more sense to rename 'info' to
'history'.
2005-10-21 13:59 willic3
* [r2463] Moved all files describing code history to info
directory. It might make more sense to rename 'info' to
'history'.
2005-10-11 07:10 willic3
* [r2410] Moved makefile for compilation tests to makefile.test,
so it does not conflict with Makefile produced by the new build
procedure.
2005-10-09 03:12 leif
* [r2405] Ignore files generated by Autoconf/Automake/Libtool.
2005-10-09 01:10 willic3
* [r2404] All files that require specific fortran features have
been modified to use information provided by the new autoconf
build procedure.
2005-10-09 01:08 willic3
* [r2403] Removed lithomop3d/future files from python make
targets, since they are not being used.
2005-10-09 01:05 willic3
* [r2402] Modified configure.ac to perform tests based on fortran
capabilities, rather than type/name. Note that it still may be
necessary to use AC_RUN_IFELSE rather than AC_COMPILE_IFELSE to
truly check some capabilities, although this will then mess up
cross-compiling.
2005-09-30 21:07 leif
* [r2386] Set the language to C++ for MPI tests, which are now
sensitive to that sort of thing.
2005-09-30 18:04 leif
* [r2384] Added files for the GNU Build System.
2005-08-06 03:17 willic3
* [r1448] Modified *.F files to treat i/o differently for
different compilers. Modified Make.mm to replace previous *.f
files with *.F files. Modified local.def to add
FORTRAN_COMPILER_TYPE (either FORTRAN_TYPE_F77 or
FORTRAN_TYPE_F95) and FORTRAN_COMPILER_NAME (presently only need
to test for FORTRAN_NAME_GFORTRAN) to PROJ_F77_DEFINES. These
must presently be set as environmental variables in the user's
environment. There is probably already a config value that
defines this information, but I haven't found it yet.
2005-08-23 02:34 unknown_user
* [r1447] [vcp] delete of edited revision with dead state
2005-08-06 03:04 willic3
* [r1446] These obsolete files have been moved to 'unused'.
2005-08-06 02:58 willic3
* [r1445] All of the *.f routines above have been replaced by *.F
routines that now need preprocessing. All of the *.F routines
have been modified to account for changes in the handling of i/o
between f77 and f95. There are 2 main factors that need to be
considered: 1. For f95, the access="append" argument is no
longer accepted in the open statement. This has been replaced by
the position="append" argument. 2. The previous method of doing
direct access, unformatted (binary) file writing is no longer
available in f95. However, it turns out that g95 supports stream
i/o, which provides very similar functionality. The gfortran
compiler does not support this, and at present it is assumed
that any f95 compiler other than gfortran will support it. Those
using gfortran receive a warning message when compiling, and
binary UCD output is turned off.
2005-07-09 04:32 willic3
* [r1444] Added PROJ_DLL stuff, in hopes of making things work for
cygwin.
2005-06-30 22:26 willic3
* [r1443] Quick and dirty routine to add extra nodes on either
side of a fault. I'm not sure if this has been tested or not.
2005-06-30 22:24 willic3
* [r1442] Dummies as placeholders for future mesh output routines.
2005-06-25 03:26 willic3
* [r1441] Removed obsolete petsc_solver argument from command
line. Explicitly specify binary UCD output using the
lm3dscan.ucdOutput variable (even though binary is now the
default).
2005-06-25 03:22 willic3
* [r1440] Changed ucdOutput inventory variable from a boolean to a
choice with possible values of ("none", "ascii", "binary") with
a default value of "binary". Changed usage and definition of the
integer equivalent in Lithomop3d_setup.py, and added the integer
equivalent to the argument list for write_ucd_mesh_f.
2005-06-25 03:17 willic3
* [r1439] Updated bindings to pass iucd parameter to
write_ucd_mesh_f.
2005-06-25 03:13 willic3
* [r1438] Added new section for writing binary UCD nodal values
and computing nodal min/max values.
2005-06-25 03:12 willic3
* [r1437] Added entire new section for writing binary UCD mesh
info.
2005-06-25 03:11 willic3
* [r1436] Added new sections to write header info for binary UCD
files.
2005-06-25 03:10 willic3
* [r1435] Updated argument lists to computation routines and added
new sections for writing binary UCD file info. Fixed previous
bug where UCD files were not being closed after use.
2005-06-25 03:09 willic3
* [r1434] Added new sections for writing binary UCD files and
computing min/max values for state variables.
2005-06-25 03:08 willic3
* [r1433] Added new option for opening a binary file using direct
access.
2005-06-25 03:06 willic3
* [r1432] Fixed calls to write_ucd_node_vals for updated usage of
iucd parameter.
2005-06-20 23:59 willic3
* [r1431] Initial version of a list of important changes from
June, 2004 to June, 2005.
2005-06-18 03:39 willic3
* [r1430] The numbered logs are raw information using cvs log with
a date filter. The changes log is an overview of the changes
compiled into a single file. This information needs to be
further distilled into a summary log that summarizes important
changes over date ranges. These logs are for my own information
as well as references when summarizing lithomop progress.
2005-06-09 04:49 willic3
* [r1429] Fixed ascii output to reflect the fact that displayed
rotation angles are in radians rather than degrees.
2005-06-09 04:48 willic3
* [r1428] Fixed stupid logic error when deciding to perform a
local coordinate rotation (used an and when I should have used
an or).
2005-06-08 02:39 willic3
* [r1427] Altered all bindings that use MPI in any way (including
those that use PETSc). There is a conflict between names defined
by mpi.h and those defined in stdio.h. One way to resolve the
conflict is to make sure that mpi.h is always included before
stdio.h. All of the C++ codes that initially referenced petsc
include files have been altered so that the petsc files are the
first ones included. Also, petsc includes have been removed from
lithomop3d_externs.h, since this resulted in additional
problems. Instead, any files that were getting their petsc
definitions from this file now include the petsc definitions
individually.
2005-06-08 02:32 willic3
* [r1426] Removed 'include petsc/default.def' and put this in
local.def. In local.def, remove $(MATH_LIBS) from
PROJ_LCXX_FLAGS and replaced it with $(EXTERNAL_LIBS). The lack
of EXTERNAL_LIBS was what was causing all the Mac build
problems, since it turns out that EXTERNAL_LIBS is not used by
default when building a library (although EXTERNAL_LIBPATH is).
By including this, now all PETSc/MPI stuff is linked when
building the library. If this link does not occur, the common
blocks defined by PETSc and mpich2 are flagged as 'defined'
rather than 'undefined', which will results in a
multiply-defined symbol error when creating the module (at least
on the Mac).
2005-05-13 22:16 willic3
* [r1425] Fixed point numbering when writing state variables to
UCD file. The number should correspond to the Gauss point number
rather than the element number.
2005-05-13 20:45 willic3
* [r1424] Fixed UCD output so that filenames for different cycles
don't conflict with each other. This is done by using the total
number of steps as the name basis rather than the number of
steps within the curren cycle.
2005-05-12 04:07 willic3
* [r1423] Moved element type logic out of do loop. Fixed header
information to be consistent with nodal output values.
2005-05-06 04:31 willic3
* [r1422] Problem to test usage of time-varying Winkler elements
for locking and unlocking a fault.
2005-05-06 04:17 willic3
* [r1421] Fixed problem where stiffness matrix was not being
zeroed before being reformed. Also, winkler coefficients were
not being added to the diagonals, so this was also fixed.
2005-05-06 04:15 willic3
* [r1420] Fixed problem where stiffness matrix was not being
zeroed before being reformed (using MatZeroEntries).
2005-05-06 03:42 willic3
* [r1419] Replaced incorrect usage of dcur with deld when
computing nodal loads equivalent to kinematic BC.
2005-05-06 03:37 willic3
* [r1418] Removed usage of petscdeveloper.h.
2005-08-23 02:31 unknown_user
* [r1417] [vcp] delete of edited revision with dead state
2005-05-06 03:36 willic3
* [r1416] This include file is no longer necessary, as the release
of a new PETSc version has made it obsolete. Also, PETSc now
includes a tag that duplicates this functionality.
2005-05-06 03:00 willic3
* [r1415] Simplified logic for determining which version of PETSc
is being used. It is now assumed that anyone using a developer
version will be using version 2.3 or later, so it is not
necessary to use the PETSC_VERSION_RELEASE tag.
2005-05-05 01:44 willic3
* [r1414] Revised fix for multiple PETSc versions. I have
simplified the logic. The supported versions are: 2.2.x
(developer or non-developer) 2.3.x (any) Any other version will
bomb. I have only tested this with PETSc 2.3.0.
2005-05-04 01:48 willic3
* [r1413] Updated bindings for usage of Bwink and Bwinkx.
2005-05-04 01:47 willic3
* [r1412] Added usage of Bwink and Bwinkx, as well as adding
corresponding entries to force flags array.
2005-05-04 01:44 willic3
* [r1411] Created new routine to zero winkler forces when a fault
is unlocked. This routine may still be unnecessary, as there is
still a problem with Winkler forces.
2005-05-04 01:42 willic3
* [r1410] Added nwinkflag and nwinkxflag.
* [r1409] Fixed header output for ascii files.
2005-05-04 01:41 willic3
* [r1408] Altered winklf to increment bwink (or bwinkx) while
updating bintern.
2005-05-04 01:40 willic3
* [r1407] Added usage of bwink and bwinkx. Changed calls to winklf.
2005-05-04 01:39 willic3
* [r1406] Added usage of bwink and bwinkx. Altered methods for
updating internal force vector in viscos.F. There are still
problems with Winkler forces that need to be resolved.
2005-05-04 01:38 willic3
* [r1405] Added unlock.f to the build list.
2005-05-03 03:45 willic3
* [r1404] This is an attempt to fix the problems introduced by the
presence of a new PETSc version (2.3.0). In the near future, I
plan to end support for all previous versions of PETSc, as the
nested list of ifs and ifdefs is getting too long to keep track
of. This is a tentative fix, and may need to be changed after I
have a guinea pig try it out (I don't have version 2.3.0
installed yet on my machine).
2005-04-22 06:22 willic3
* [r1403] Got rid of input_misc.h and replaced it with
write_modelinfo.h. Fixed commas that should have been semicolons.
2005-04-22 06:21 willic3
* [r1402] Moved local, localf, and localx into this section.
2005-04-22 06:20 willic3
* [r1401] Fixed problem where I had left off the _nodes portion of
the routine names for sort_slip and sort_split.
2005-04-22 06:19 willic3
* [r1400] Fixed problem where I had not finished updating array
names for new read_wink routine.
2005-04-22 06:17 willic3
* [r1399] Added assign_wink and removed local, localf, and localx,
which now belong in sparse.
2005-04-22 06:16 willic3
* [r1398] Added sorting.h to list of includes.
2005-04-22 06:15 willic3
* [r1397] Fixed problem where time step variables were being
initialized in the wrong place. Deleted stray line that didn't
belong. Fixed typo for calling write_split.
2005-04-22 06:13 willic3
* [r1396] Fixed typo in build list.
2005-04-21 08:34 willic3
* [r1395] Added in missing usage of dprev array.
2005-04-21 08:33 willic3
* [r1394] Commented out debugging statements.
2005-04-21 08:09 willic3
* [r1393] Changed all necessary references to point to sorted
element node array rather than original array. Removed
unnecessary arguments from write_slip and write_split.
2005-04-21 08:07 willic3
* [r1392] Removed arguments that weren't needed.
2005-04-21 08:05 willic3
* [r1391] Added new sorting routines.
2005-04-21 08:00 willic3
* [r1390] Added new output routines.
2005-04-21 07:59 willic3
* [r1389] Fixed argument types to correspond to updated argument
lists.
2005-08-23 02:30 unknown_user
* [r1388] [vcp] delete of edited revision with dead state
2005-04-21 07:32 willic3
* [r1387] All of the routines that were previously in these files
are now in write_meshinfo.cc and .h, along with the new output
routines.
2005-04-21 07:09 willic3
* [r1386] Altered bindings to reflect usage of sorted element
nodes array.
2005-04-21 07:07 willic3
* [r1385] Removed extra close brace.
2005-04-21 07:04 willic3
* [r1384] Added new function calls.
2005-04-21 06:46 willic3
* [r1383] Fixed code so we are now using sorted element node array
(Iens rather than Ien). Removed usage of Mhist material history
array.
2005-04-21 06:08 willic3
* [r1382] Added new sets of bindings (sorting and
write_modelinfo), and removed input_misc.
2005-04-21 06:06 willic3
* [r1381] Modified bindings to no longer use Mhist (material
history) array.
2005-04-21 06:03 willic3
* [r1380] Created new bindings for all output routines. This
includes routines that existed previously plus new routines.
2005-04-21 03:40 willic3
* [r1379] Moved in equation localization routines from
numbering.cc and modified bindings for new argument lists.
2005-04-21 02:00 willic3
* [r1378] These routines are no longer needed.
2005-04-21 01:59 willic3
* [r1377] Removed obsolete routines from build list and added all
the new routines.
2005-04-20 07:52 willic3
* [r1376] Revised bindings for new argument lists. Also, routines
that localize equation numbers have been removed. They will be
moved to sparse.cc.
2005-04-20 07:29 willic3
* [r1375] Revised all bindings for new routine calls and removed
bindings for read_mathist and read_winkx, which are no longer
needed.
2005-08-23 02:30 unknown_user
* [r1374] [vcp] delete of edited revision with dead state
2005-04-20 07:25 willic3
* [r1373] Moved this routine to 'unused'. Since formats are the
same for regular Winkler and slippery node Winkler forces, the
same routine may be used to read them both.
2005-08-23 02:30 unknown_user
* [r1372] [vcp] delete of edited revision with dead state
2005-04-20 07:11 willic3
* [r1371] I have removed material histories for now.
2005-08-23 02:30 unknown_user
* [r1370] [vcp] delete of edited revision with dead state
2005-04-20 07:10 willic3
* [r1369] I have eliminated material histories for now. They will
become obsolete once we are using a spatial database.
2005-04-16 07:48 willic3
* [r1368] Cosmetic changes.
2005-04-16 07:47 willic3
* [r1367] Fixed some improper closing of files.
* [r1366] Updated output text and fixed some improper closing of
files.
2005-04-16 07:46 willic3
* [r1365] Added some new parameters to be output, corresponding to
the way elements are presently handled (no mixing of element
types). This may be revised later.
2005-04-16 07:45 willic3
* [r1364] Changed how Winkler forces are handled. Definitions are
now read into temporary arrays, and the actual required arrays
(which require global equation numbers) are determined later in
subroutine assign_wink.
2005-04-16 07:43 willic3
* [r1363] Cosmetic changes.
2005-04-16 07:40 willic3
* [r1362] All of these routines have been modified to no longer
use material histories.
2005-04-16 07:39 willic3
* [r1361] Cosmetic changes.
2005-04-16 07:38 willic3
* [r1360] All of these routines have been modified to no longer
use material histories. Once the spatial database is being used,
material histories will become obsolete.
2005-04-16 07:35 willic3
* [r1359] New routine to take winkler definitions and replace them
with values that are needed by the code. It was necessary to
break out this functionality from the original read_wink routine
so that equation numbers were not required in an input section.
2005-04-16 07:33 willic3
* [r1358] Cosmetic changes.
* [r1357] New routine to replace element numbers with reordered
numbers.
2005-04-16 07:31 willic3
* [r1356] Added all of these new output routines. Previously,
input and output were performed in the same routine in many
cases.
2005-04-16 07:28 willic3
* [r1355] Added and deleted arguments to several functions to
conform to revised interface.
2005-04-15 07:26 willic3
* [r1354] Deallocated some additional temporary arrays.
2005-04-15 07:18 willic3
* [r1353] Altered main function to call different functions for
the revised setup section. Note that both of these codes need
testing, some new f77 routines still need to be written, and the
bindings need to be updated.
2005-04-15 07:16 willic3
* [r1352] Initial revised version of setup section. Things are now
divided into several functions: initialize: Get necessary values
from scanner read: Read all information and populate
corresponding arrays numberequations: Set up global equation
numbers based on BC and slippery nodes sortmesh: Sort elements
into element families. Additional sorting is done for things
that were specified in terms of the original element ordering.
sparsesetup: Set up local to global element id arrays and set up
sparse matrix structure. allocateremaining:Allocate all
additional arrays that are needed for computations. meshwrite:
Outputs parameters and mesh description to ascii, plot, and/or
UCD files.
2005-04-14 08:01 willic3
* [r1351] The write_ucd_node_vals routine has been modified to
output slippery and split node displacements, and the calls in
autoprestr, elastc, and viscos have been modified accordingly.
2005-04-14 07:59 willic3
* [r1350] All of these routines have been modified so that they
only perform input, and all sections in which they perform
output, equation reordering, etc., have been removed. These
functions will be performed by separate routines.
2005-04-14 07:57 willic3
* [r1349] Cosmetic changes.
2005-04-14 07:56 willic3
* [r1348] Removed all output sections from this routine.
2005-04-14 07:55 willic3
* [r1347] Removed input unit and file information that was no
longer necessary.
2005-04-14 07:48 willic3
* [r1346] Initial version of a simple routine that sorts split
node entries according to the new element ordering.
2005-04-14 05:04 willic3
* [r1345] Added this file as a kludgy fix to let the code know
whether the user is using the developer's version of PETSc or
not. If so, the tag PETSC_DEVELOPER_VERSION needs to be set to 1.
2005-04-14 05:02 willic3
* [r1344] Messed around some more with fixing things to work with
developer's and non-developer's versions of PETSc. The current
fix requires the user to set the value PETSC_DEVELOPER_VERSION
to 1 if the developer's version is being used. This is done in
petscdeveloper.h. If this ends up being a long-term solution, it
will need to be documented. It is hoped that with the new
release of PETSc, this fix will be unnecessary. We will simply
support only the most current version.
2005-04-13 07:37 willic3
* [r1343] Left out usage of the infin array for now, as infinite
elements are presently broken anyway.
2005-04-13 07:23 willic3
* [r1342] Initial version of a routine to sort elements into
families. The routine is somewhat inefficient with memory, as it
requires an additional element node array to hold the sorted
results (rather than attempting an in-place sort). The original
array may be deleted after use, however. The routine also
determines several global dimensions and creates an element
family info array as well as an index array containing the
original element numbering.
2005-04-13 06:17 willic3
* [r1341] Stripped out all output, element renumbering, and
indexing. This will all be handled by separate routines.
2005-04-13 05:33 willic3
* [r1340] Stripped out everything but input. Output of BC info (if
requested) will be handled by a separate routine.
2005-04-11 01:10 willic3
* [r1339] Fixed problem where slippery node output always produced
nonzero slip weights.
2005-04-09 02:21 willic3
* [r1336] Updated example files to current version of the code.
2005-04-09 02:10 willic3
* [r1334] Additional revisions after testing example.
2005-04-09 01:29 willic3
* [r1332] Updated these input files for current code version.
2005-04-09 01:20 willic3
* [r1331] Simple test problem using slippery nodes.
2005-04-09 00:38 willic3
* [r1330] Moved the definition of pointerToIpslp out of an if
block. This array is passed to several driver routines even when
it is not used. Having it defined inside the if block cause an
error message at the end of code execution.
2005-04-08 07:47 willic3
* [r1329] Added '-ksp_rtol 1e-9' to the command line arguments.
This dramatically improves the accuracy for slippery node tests,
and likely will help for the other test problems. I need to
change all of the Make.mm files for example problems to be
similar to these, and also need to modify all input files for
current formats.
2005-04-08 07:41 willic3
* [r1328] Cosmetic changes.
2005-04-08 07:40 willic3
* [r1327] Changed routine so all indexing is done with local
arrays that are altered to work with PETSc. Fixed problem where
absolute value of lmx array (suitably altered) was needed for
MatSetValues rather than the actual array. Fixed problem where
row and column index arrays appeared to be swapped in
MatSetValues in cases involving slippery nodes.
2005-04-08 07:37 willic3
* [r1326] Fixed stupid bug where I was not indexing the ien array
when sending it to addsn.
2005-04-08 07:33 willic3
* [r1325] Retained commented-out debugging statements.
* [r1324] Retained commented-out debugging statement that uses
PETSc MatView routine.
2005-04-08 07:31 willic3
* [r1323] Cosmetic changes and retained commented-out debugging
statements.
2005-04-08 07:30 willic3
* [r1322] Changed definition of alpha parameter from a parameter
statement to a data statement since it is being passed as an
argument.
2005-04-06 06:14 willic3
* [r1321] Added input files to run SCEC BM5. This serves as a test
of split nodes.
2005-04-06 06:07 willic3
* [r1320] Modified calls to read_connect, read_split, and
read_slip to create and use element indexing array. It is then
destroyed after use.
2005-04-06 06:04 willic3
* [r1319] Fixed major bug where I wasn't indexing state variable
and other arrays before passing them to computational routines.
2005-04-06 06:03 willic3
* [r1318] Cosmetic changes.
2005-04-06 06:02 willic3
* [r1317] Removed usage of ielg element counter, which isn't
needed.
2005-04-06 06:01 willic3
* [r1316] Cosmetic changes.
2005-04-06 06:00 willic3
* [r1315] Fixed improper indexing of itmp array.
* [r1314] Cosmetic changes.
2005-04-06 05:56 willic3
* [r1313] Updated call to addstf to pass stemp array.
2005-04-06 05:54 willic3
* [r1312] Updated call to addstf to pass stemp array.
2005-04-06 05:52 willic3
* [r1311] Modified this routine for cases where there are slippery
nodes: Added new index array that is the absolute value of the
slippery node index array. According to the MatSetValues
documentation, negative indices are skipped, which would mean
entries from the negative side of the fault would never get
added. This should fix that problem. Changed setup of routine so
that portion related to slippery nodes is never executed if
element contains no slippery nodes. Changed routine for
efficiency. Rather than switching stiffness matrix signs back
and forth, I use two arrays, and never change the entries of the
local stiffness array. Instead, the properly-signed values are
transferred into the stemp array, and this is used instead. To
do this, I now pass in the stemp array from all calling
routines. This routine needs to be tested.
2005-04-06 05:39 willic3
* [r1310] Modified these routines to account for the fact that
split and slippery nodes are specified using the original
element ordering rather than the revised ordering that occurs
after they are sorted into families. Routine read_connect
creates the index array to do this, and both read_slip and
read_split have been modified to use it. After use, the index
array may be destroyed.
2005-04-06 05:36 willic3
* [r1309] Updated bindings for read_connect, read_split, and
read_slip to use element index array. This accounts for the fact
that split and slippery nodes are specified using the original
element numbering rather than the numbering that occurs after
elements are sorted into families.
2005-04-05 22:47 willic3
* [r1308] Fixed conditional compilation to key on
PETSC_VERSION_PATCH. Only version 2.2.1 is supported, and it is
assumed that all versions less than version patch 42 do not
include MatSetSizes. This needs to be tested with different
patch versions.
2005-04-03 11:56 willic3
* [r1307] Added conditional compilation depending on whether
MatSetSizes is defined.
2005-04-03 06:11 willic3
* [r1306] Cosmetic change.
2005-04-02 07:59 willic3
* [r1305] Updated these to account for new and deleted parameters,
and new method of specifying state variable output.
2005-04-02 07:58 willic3
* [r1304] Added commented-out line for PETSc debugging when
desired. This is presently not working.
2005-04-02 07:56 willic3
* [r1303] Fixed bindings for createPETScMat to account for new
PETSc calling conventions. Matrix size is now set separately
using MatSetSizes.
2005-04-02 07:54 willic3
* [r1302] Fixed bindings for scan_connect.
* [r1301] Updated bindings to use Dprev vector. Also fixed
definitions in lithomop3d_externs.h to account for changed call
to scan_connect.
2005-04-02 07:50 willic3
* [r1300] Added ElementTypeDef.py to EXPORT_PYTHON_MODULES.
2005-04-02 07:49 willic3
* [r1299] Added autoprestrStage. Put back in the Dprev vector.
Added back in connectivitySize which is now a function of
numberElementNodes and numberVolumeElements. Put in definition
of reformFlagInt. Fixed usage of ElementTypeDef class. Fixed
output of subiteration parameters. Fixed some typos. Removed
unused lists.
2005-04-02 07:45 willic3
* [r1298] Added or re-added some parameters that were needed.
Increased dimension of VolumeElementFamilyList to allow tracking
of material model, and then made the corresponding changes in
the call to scan_connect.
2005-04-02 07:42 willic3
* [r1297] Added autoprestrStage to PETSc stages, and added usage
of Dprev vector when a nonzero initial guess is desired.
2005-04-02 07:40 willic3
* [r1296] Modified all parameter names to explicitly include
'Volume' name, to indicate element type.
2005-04-02 07:37 willic3
* [r1295] Put this routine back in, since it is needed to compute
element stiffness matrix.
2005-04-02 07:36 willic3
* [r1294] Commented out debugging statements.
2005-04-02 07:35 willic3
* [r1293] Build list was modified to include getmat.f, which is
needed to compute the local stiffness matrix.
2005-04-02 07:34 willic3
* [r1292] All of these routines were modified to pass along the
iddmat index array, which is needed by routine stiff_ss.
2005-04-02 07:31 willic3
* [r1291] Fixed incredibly stupid problem where I had substituted
a BLAS matrix-vector operation for a matrix-matrix operation.
Since it appears that BLAS does not have a routine to multiply a
packed symmetric matrix by another matrix, it was also necessary
to use the iddmat index array, which I had previously deleted.
This also made it necessary to use the getmat routine, which
transfers the packed array to a full-size array. Note that these
problems can all be avoided once I get rid of the global dmat
array. Once this is no longer stored globally there will be no