-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathBranchs_Old_Gitlog
3864 lines (2744 loc) · 149 KB
/
Branchs_Old_Gitlog
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
commit a812af8cf6daf1adb07d7934fea2b61f769e4c6e
Author: Ben O'Neill <beon9273@colorado.edu>
Date: Tue Jul 25 11:40:01 2017 -0400
Makefile changes
commit 8a4bd3b361fe502546b5187f8e17a0125655a6c6
Author: Ben O'Neill <beon9273@colorado.edu>
Date: Tue Jul 25 10:49:13 2017 -0400
Clean up pass mostly fixing the naming of the variables in the load balancing code.
commit beb1dbea1ffd8b6a84a6d425ea3cdcbb57a1689d
Author: Ben O'Neill <beon9273@colorado.edu>
Date: Thu Jul 6 16:22:30 2017 -0400
Bug fixes
commit 49aca755586ab0c01367232c31d0383d4def78c1
Merge: 5b395dd 5cc085a
Author: Ben O'Neill <beon9273@colorado.edu>
Date: Thu Jul 6 16:06:48 2017 -0400
Merged in load balancing
commit 5cc085add7b2b458d16a4b8008d80bd46cead0bf
Author: Ben O'Neill <beon9273@colorado.edu>
Date: Wed Jul 5 10:50:54 2017 -0400
Added the calling function to all status structures. When doing adaptation it is nice to know what function is calling the step status, so that we can complete the error estimation process only one time per iteration.
commit 92625e4cf4233d18bdf2f3ec96d7d2858d212f35
Author: Ben O'Neill <beon9273@colorado.edu>
Date: Fri Jun 23 14:50:37 2017 -0400
Picked up a memory error with Valgrind. Recv_buffer is of length size.
commit 510df329bd3aaad3c5cf9a071f5b1d4d61546c7b
Author: Ben O'Neill <beon9273@colorado.edu>
Date: Fri Jun 23 14:44:08 2017 -0400
Bug fixes for the Error estimation Functions
commit 5b395dd835bd19bb1e93208b23e84a3a29d76f0b
Author: Ben O'Neill <beon9273@colorado.edu>
Date: Fri May 5 14:31:53 2017 -0400
Added the example, and changed WFactors so that they are set \n directly in the core by the user
commit 71b5fbefe75bd6de346803372ff5bf96f406c0fa
Author: Ben O'Neill <beon9273@colorado.edu>
Date: Fri May 5 09:43:28 2017 -0400
Reinstating ex-01 from master branch
commit e183f51847a8c0d91f594f6d027600b1b1c78af6
Author: Ben O'Neill <beon9273@colorado.edu>
Date: Mon Apr 17 15:14:04 2017 -0400
Merged in a commit adding the global access feature. The user can optinally set a global access function that allows full access to the global solution, residuals, rfactors, etc, prior to FRefine. This function can then be used to decide where to refine, or to quit all together
commit dc7432ac5787bd11f29df0558f66b4c42c419241
Author: Ben O'Neill <beon9273@colorado.edu>
Date: Thu May 4 15:57:52 2017 -0400
First commit with error estimation and richardson extrapolation added.
commit 866ec7384cc992a108aee0dec2b3f6c016014763
Author: Ben O'Neill <beon9273@colorado.edu>
Date: Mon Apr 17 12:17:06 2017 -0400
Added temporal load balancing
commit 094e58e6d9a94679257775d52789a79398e219e3
Author: Ben O'Neill <beon9273@colorado.edu>
Date: Fri Apr 14 14:35:07 2017 -0400
Added in load balancing. Need to test
commit 714b7f7f31fc876ef2280ea400f57932086c1977
Author: Ben O'Neill <beon9273@colorado.edu>
Date: Fri Apr 14 11:27:19 2017 -0400
Added the communication minimzation feature. When this is set true,
the first point on any processor must be a C-point. The idea is that
splitting coarse grid intervals up across time processors does not
give us anything.
commit 7a35d4398f2822bc544bd843779b62a52ccd22de
Author: Matthieu B Lecouvez <lecouvez1@llnl.gov>
Date: Thu Apr 13 16:50:10 2017 -0700
Just forgotten a few lines for the access to NLevels
commit 0a43b86645883ddb5d347ce4ffc9a9426dfa3910
Author: Matthieu B Lecouvez <lecouvez1@llnl.gov>
Date: Thu Apr 13 16:42:38 2017 -0700
Gives access to the total number of levels from Step, Coarsen/Refine
and Access user functions through the *status structures.
commit 5fd5c53f9938a003c4f188fdff3a132268f892b6
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Tue Apr 4 11:14:51 2017 -0700
- Fixed some discrepancies between ex-03, ex-03-serial and driver-diffusion-2D.c
regarding how CFL numbers were treated
- Fixed a small bug in drive-diffusion-2D regarding how PFMG iterations
were tracked and how max-pfmg-iter was enforced (basically, A_idx and level
were not equal when skip was used).
- Removed the "RDF Hack" that put F-refinement in ex-03. We now use
ex-01-refinement.c for our regression testing of F-refinement
commit 7d3da43e79abebb744bdaa280842429fca54a82f
Merge: 63a657c afe974f
Author: Rob Falgout <rfalgout@llnl.gov>
Date: Fri Mar 31 16:14:12 2017 -0700
Merge branch 'master' of ssh://mystash.llnl.gov:7999/xbraid/xbraid
commit 63a657c88ca1bed92c3070221d86c366a2c1e30c
Author: Rob Falgout <rfalgout@llnl.gov>
Date: Fri Mar 31 16:13:11 2017 -0700
Fixed some issues with the CFL computation
commit afe974fca8c7000650c7e6fe643b9f433a7acb69
Author: Matthieu B Lecouvez <lecouvez1@llnl.gov>
Date: Wed Jan 25 16:10:17 2017 -0800
Prevent the use of the option skip and seqsoln simultaneously. Changes associated test.
commit c04422feb32e9b8b2e5fab646531d3a3c5ce01c1
Merge: fb8389f 42d33f6
Author: Schroder, Jacob B <schroder2@llnl.gov>
Date: Wed Jan 18 13:15:51 2017 -0800
Merge pull request #8 in XBRAID/xbraid from fullrefinement-dev to master
* commit '42d33f6a72085b70ee38120d5af1e11067142024':
Modifies the user manual to reflect modification regarding the removal of the refinement constrain. Also adds the new example in the doc.
Updated FRefine() comments
Adds regression tests for the ode1D problem with refinement
Adds regression test for ex-03 with refinement factor larger that coarsening factor.
Modifies ex-03 and associated tests to accept a maximum refinement factor. This way, all test results remain unchanged.
Adds option to refine in time the example ex-01 based on LTE.
Adds option to the refinement example to limit the refinement factor.
Removes the limitation on the refinement factor. It is not limited to the coarsening factor anymore. Adds an example for refinement.
commit 42d33f6a72085b70ee38120d5af1e11067142024
Merge: 2eeac8c a6133b2
Author: Matthieu B Lecouvez <lecouvez1@llnl.gov>
Date: Wed Jan 18 13:08:11 2017 -0800
Merge branch 'fullrefinement-dev' of ssh://mystash.llnl.gov:7999/xbraid/xbraid into fullrefinement-dev
commit 2eeac8ccba48c3f2b49ba25e621883a57f806b07
Author: Matthieu B Lecouvez <lecouvez1@llnl.gov>
Date: Wed Jan 18 13:06:52 2017 -0800
Modifies the user manual to reflect modification regarding the removal of
the refinement constrain. Also adds the new example in the doc.
commit a6133b2b6f09ec04e07f6119f215492d00b3d9d8
Author: Rob Falgout <rfalgout@llnl.gov>
Date: Wed Jan 18 10:49:42 2017 -0800
Updated FRefine() comments
commit 6cad1eec264b18dc1d4c27e360540ef52a1efcaf
Author: Matthieu B Lecouvez <lecouvez1@llnl.gov>
Date: Tue Jan 17 10:14:36 2017 -0800
Adds regression tests for the ode1D problem with refinement
commit f102b1d9dbe7441e452993e64bb27177c550d0b8
Author: Matthieu B Lecouvez <lecouvez1@llnl.gov>
Date: Tue Jan 17 09:38:55 2017 -0800
Adds regression test for ex-03 with refinement factor larger that
coarsening factor.
commit f99762002b1eadd33f573663b76a7b9b6bd12569
Author: Matthieu B Lecouvez <lecouvez1@llnl.gov>
Date: Tue Jan 17 08:59:51 2017 -0800
Modifies ex-03 and associated tests to accept a maximum refinement factor.
This way, all test results remain unchanged.
commit f050856da4d1daab36aae79188c52a86fbef0c45
Author: Matthieu B Lecouvez <lecouvez1@llnl.gov>
Date: Fri Jan 13 15:28:06 2017 -0800
Adds option to refine in time the example ex-01 based on LTE.
commit 660ae1909fecfa8bb546835e9358963fa0aa8997
Author: Matthieu B Lecouvez <lecouvez1@llnl.gov>
Date: Fri Jan 13 14:28:50 2017 -0800
Adds option to the refinement example to limit the refinement factor.
commit a51f460b638e10e71b58f7637dd0be5ec8914403
Author: Matthieu B Lecouvez <lecouvez1@llnl.gov>
Date: Fri Jan 13 14:11:41 2017 -0800
Removes the limitation on the refinement factor. It is not limited
to the coarsening factor anymore. Adds an example for refinement.
commit fb8389fe3301308f21b85c4d5a7be6c5e04a7021
Author: Matthieu B Lecouvez <lecouvez1@llnl.gov>
Date: Tue Dec 13 10:05:57 2016 -0800
Adds an option to set the level of output-to-file. For now, only one
file is created (braid.out.cycle). It can be set to 0 (no output)
or to 1 (the cycle is written to the file, default).
commit bd7eee90acf9ec9176b80631a40e29de10738c57
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Mon Dec 5 09:50:32 2016 -0800
Updated the debugging section in the documentation again, based
on some impressions from the tutorial.
commit 2d9f37b2ae03549894e58237a6e7a1ed4a4f9ece
Author: Jacob Schrider <schroder2@vpnb-user-128-15-245-208.llnl.gov>
Date: Sun Dec 4 17:12:35 2016 -0800
Adding one more bit to the canonical debugging advice section in
Introduction.md.
commit 3d53af9f04bec1a9f4348b5baedbb4938aba91ff
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Fri Nov 18 09:58:38 2016 -0800
One more change for v2.1.0
commit 90d932b0f5eb12a1db665b72136ae41ee78ba0d1
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Fri Nov 18 09:48:44 2016 -0800
Version 2.1.0 Commit
Adding a CHANGELOG and a few more documentation tweaks
commit 7924f985b462ec6f69cc96b3196e76b20c76d5a6
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Thu Nov 17 18:04:22 2016 -0800
Re-wrote some of the users manual for a new release
commit ffe045db3c0dc9e6a9b4699cf1c8aefcc44bc9ec
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Thu Oct 6 16:23:45 2016 -0700
Updating the new regression tests for the 1D heat equation example.
commit 257e821bcbd2a554918ec82060a058507742617d
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Tue Oct 4 17:06:07 2016 -0700
Adding Simple 1D diffusion example, plus accompanying regression tests.
The regression tests need to be run on my tux-box to get the right output
to store in the .saved files. That will come in the next commit.
commit 80352609a35161261e14ec11605bd70d74c54c5c
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Mon Oct 3 20:20:27 2016 -0700
Changed make system so that the sequential option is now enabled with
make sequential=yes
similar to the other two options
make debug=yes|no
make valgrind=yes|no
commit b784b865c157bb63cc685dfb965e9a4065445767
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Mon Oct 3 17:07:16 2016 -0700
Implemented most of ex-02.c, a simple 1D heat equation example. This
will be the main "tutorial" example for showing people things like
spatial coarsening and the residual option.
commit 3ed4eca99923e202cf760d61b68c598943c06ec9
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Fri Sep 30 16:59:02 2016 -0700
During the restructuring of the examples and drivers, these
viz.py files got left out.
commit dfdf869c6ad7327ed8d8aefda9e19e0eb33d1177
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Fri Sep 30 16:54:48 2016 -0700
This commit does a major re-working of the structure of the examples and
drivers directories. The idea moving forward is to
Only do development with drivers--not examples
I've been guilty of mis-using examples, myself. But, it was leading to
things becoming a bit of a mess.
commit a5e041d7cfc83bbeb2036fdeb45f78fbb893e87d
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Thu Sep 29 22:20:46 2016 -0700
Modified ex-01-expanded-f.f90 so that it's style and structure
resemble (in so far as possbile) the C version, ex-01-expanded.c
commit 82dc00bcdb14b7004feb799df1132b0ca058e343
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Thu Sep 29 17:04:36 2016 -0700
Changed some coding style and structure in ex-01-expanded-bdf2 so that
it mirrors ex-01 and ex-01-expanded
commit 9bc0d688b6578e776ab01a752d95ab6c04e5c7e6
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Thu Sep 29 15:35:48 2016 -0700
Issue 6: Added a braid test to test the compatibility of residual
and test, that
f - Residual( Step(u_{i-1}, f_i), u_{i-1}) \approx 0
commit 0a44d6d75ae4bc52b0c31c9ead0790b9fd4f8a59
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Thu Sep 29 10:15:22 2016 -0700
Upon realizing that we have yet another ex-01 type implementation
of the scalar ODE, I've now re-organized things into four ex-01 drivers
ex-01.c Simplest example (scalar ODE)
ex-01-expanded.c Same as ex-01.c, but with more Braid options
ex-01-expanded-f.f90 Same as ex-01-expanded.c, only in F90
ex-01-expanded-bdf2.f90 Same as ex-01-expanded.c, only uses BDF2
and the shell vector
commit def2e1347e0c90089e89813d0b1842f43fbf13cc
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Wed Sep 28 12:50:47 2016 -0700
Updated examples/ex-01**
We now have
ex-01.c Simplest example (scalar ODE)
ex-01b.c Same as ex-01.c, but with more Braid options
ex-01b-f.f90 Same as ex-01b.c, only in F90
commit 590836608fcc99a42e39f638681934ef1baf4cdf
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Tue Sep 27 11:37:18 2016 -0700
update docs.sh regression test, to be more intelligent with its grepping
for "error" in the output.
commit 032d1834966b94547118ff27d0849e81299ff32f
Author: Veselin Dobrev <dobrev@llnl.gov>
Date: Mon Sep 26 12:11:34 2016 -0700
Add drive-08.cpp -- a time-dependent elasticity problem using mfem.
commit 464dddc31632e3f73d744e5e87d1177bef58213e
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Sat Sep 24 19:39:44 2016 -0700
A number of small changes regarding the recent updates to the status structure
- Changed comments (and docs/*.conf) files so that the doxygen PDF better presents
the new status structures
- Fixed a couple small bugs, like setting tstart in the C++ interface
- Removed comments in _braid_status.c, that were duplicate from braid_status.h
- Moved a few things from braid_status.h to _braid_status.h that weren't
user interface related
commit 4d3980437ddad09882e368987ab4b56bfbcac4f6
Merge: 111c213 e582086
Author: Schroder, Jacob B <schroder2@llnl.gov>
Date: Thu Sep 22 17:06:19 2016 -0700
Merge pull request #7 in XBRAID/xbraid from status-facto to master
* commit 'e582086445bfc2ed25c9afd7bee4edae05556287':
Modifies the SetRSpace. Since now core == status, we have to implement this differently (i.e. directly in the SetRSpace)
Remove the _braid_DeriveStatusElt, which is replaced by _braid_StatusElt. Also, move a typedef to the appropriate header.
Moves all the properties of _braid_Status_struct to the core. Now _braid_Status_struct inherits from _braid_Core_struct.
Change the name of the BufferStatus property "size" to "size_buffer"
Temporary state with the new header file _braid_status.h, that contains the actual implementation of the _braid_*Status_struct. This step is required for the next step, which will be to merge the _braid_Status_struct into _braid_Core_struct.
Removes the intermediate structure GlobalStatus. Now, all *Status derivate from the main structure _braid_Status so that they can be easily cast between each other. In braid, we only have one braid_Status, the other astatus, bstatus, cstatus and sstatus are the same structure cast into different types.
Change the headers and functions of the accessors for the derived *Status to macros. This way, when adding a new parameter in the Status, only only need to implement the global accessor and add one macro call per derived accessor.
Modification of the status implementation. We now have one main structure _braid_Status_struct wich is a merge of all the previous *Status. All the other status structures now contain only a pointer to this main status structure. A new structure called _braid_GlobalStatus_struct appears (and also contains only a pointer to the main status structure). This new structure allows the user to access properties that he would normally not be able to access, by casting the *Status given to him by XBraid into this GlobalStatus. All accessors have been implemented, with very few API changes: - for StepStatus, GetTstart becomes GetT for uniformization purposes - for *Status, GetIStop becomes GetTIndex, and always return the index in ta such that ta[GetTIndex()] == GetT(). This was not consistent so far (in StepStatus, we had tstop, in AccessStatus, we had t)
Change the names of some funciton argument to avoid collision with free()...
commit e582086445bfc2ed25c9afd7bee4edae05556287
Author: Matthieu B Lecouvez <lecouvez1@llnl.gov>
Date: Thu Sep 22 15:58:07 2016 -0700
Modifies the SetRSpace. Since now core == status, we have to implement
this differently (i.e. directly in the SetRSpace)
commit 5cd4eaacf7c87ede58e1f0eda4ba9fb3774b599b
Author: Matthieu B Lecouvez <lecouvez1@llnl.gov>
Date: Thu Sep 22 15:53:40 2016 -0700
Remove the _braid_DeriveStatusElt, which is replaced by _braid_StatusElt.
Also, move a typedef to the appropriate header.
commit 31b185f143892f10a264c7d86ce30839dba33eb2
Author: Matthieu B Lecouvez <lecouvez1@llnl.gov>
Date: Thu Sep 22 11:35:31 2016 -0700
Moves all the properties of _braid_Status_struct to the core. Now
_braid_Status_struct inherits from _braid_Core_struct.
commit 0d9e67e5a417514f3877f0e7802b204616837e0d
Author: Matthieu B Lecouvez <lecouvez1@llnl.gov>
Date: Thu Sep 22 11:03:38 2016 -0700
Change the name of the BufferStatus property "size" to "size_buffer"
commit b4e0e3c61e715d7aa6ac4b21e80632f938693ced
Author: Matthieu B Lecouvez <lecouvez1@llnl.gov>
Date: Thu Sep 22 10:56:50 2016 -0700
Temporary state with the new header file _braid_status.h, that contains
the actual implementation of the _braid_*Status_struct. This step is
required for the next step, which will be to merge the _braid_Status_struct
into _braid_Core_struct.
commit fb0d7eba23a24a01d86dcbbae18aa69ad9a3e173
Author: Matthieu B Lecouvez <lecouvez1@llnl.gov>
Date: Mon Sep 19 11:40:17 2016 -0700
Removes the intermediate structure GlobalStatus. Now, all *Status
derivate from the main structure _braid_Status so that they can be
easily cast between each other. In braid, we only have one braid_Status,
the other astatus, bstatus, cstatus and sstatus are the same structure
cast into different types.
commit 3f8752e039ee1e7798200cc53b0aef27aff52dc7
Author: Matthieu B Lecouvez <lecouvez1@llnl.gov>
Date: Fri Sep 16 10:57:50 2016 -0700
Change the headers and functions of the accessors for the derived
*Status to macros. This way, when adding a new parameter in the Status,
only only need to implement the global accessor and add one macro
call per derived accessor.
commit e1d55931b8ebff6a42316cc10ccd0e46f21ac0c3
Author: Matthieu B Lecouvez <lecouvez1@llnl.gov>
Date: Thu Sep 15 16:52:31 2016 -0700
Modification of the status implementation. We now have one main structure
_braid_Status_struct wich is a merge of all the previous *Status.
All the other status structures now contain only a pointer to this
main status structure. A new structure called _braid_GlobalStatus_struct
appears (and also contains only a pointer to the main status structure).
This new structure allows the user to access properties that he would
normally not be able to access, by casting the *Status given to him by
XBraid into this GlobalStatus.
All accessors have been implemented, with very few API changes:
- for StepStatus, GetTstart becomes GetT for uniformization purposes
- for *Status, GetIStop becomes GetTIndex, and always return the index
in ta such that ta[GetTIndex()] == GetT(). This was not consistent
so far (in StepStatus, we had tstop, in AccessStatus, we had t)
XBraid compiles and pass the tests. But these modifications have not
been heavily tested. Note that the c++ wrappers are probably out-to-date
and miss some accessors, but should still work.
commit 37b69fbb26125b495cfc34b1e19eddb3981b11df
Author: Matthieu B Lecouvez <lecouvez1@llnl.gov>
Date: Thu Sep 15 15:51:00 2016 -0700
Change the names of some funciton argument to avoid collision with free()...
commit 111c213cb91a67af21e740e57ce3e66f7ec6b8eb
Author: Matthieu B Lecouvez <lecouvez1@llnl.gov>
Date: Wed Sep 14 17:26:44 2016 -0700
Removal of extra comma in some functions argument lists, that prevented
from compiling the fortran interface of spatial coarsening and
refinement. There are still warnings to be addressed.
commit 8f9184f0dd8d25d835fe6453f7e4da4e0fec3c86
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Tue Aug 23 08:36:34 2016 -0700
Had to change how ex-01f is compiled in the regression test so that
the Fortran warnings about the unused app parameter are ignored. The
app has to be a part of the interface (obviously), but for such a simple
example, it isn't always used.
commit 31414967d792b9ee6c3e60563bdffb6d4c58b571
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Mon Aug 22 08:56:36 2016 -0700
Two final changes required for Andrea's regression tests to run
from the immediately preceding pull request
- add ode1.sh to machine-tux.sh
- enable the user defined time grid in the fortran interface
commit 5ced904aa4865e713bc7993492828c5c86882834
Merge: 191d166 8cb58b6
Author: Schroder, Jacob B <schroder2@llnl.gov>
Date: Mon Aug 22 08:47:14 2016 -0700
Merge pull request #6 in XBRAID/xbraid from hessenthaler to master
* commit '8cb58b636f33b9e0232dccf2cce20aa46af9a2a7':
Forgot to turn off this part of the Fortran interface
Last edits before pulling this branch into master
Edits related to making a pull request -- cleaned up a few things
Added: step_status_get_istop in Fortran interface.
Removed commented line.
Added: Regression test for ex-01/ex-01f testing the user-defined fine time grid feature. Added: braid_access_status_get_istop in Fortran interface. Bugfix: ex-01f now queries access status to determine filename index as we don't have dt = const, in general.
Set preprocessor directive braid_Fortran_TimeGrid to 0 by default
Added: Fortran interface to enable user supplied fine time grid. Modified: Pass values by reference so we have interoperability for C/Fortran. Modified: ex-01f now demonstrates how to define a fine time grid. Bugfix: Set initial tgrid function pointer to NULL. Todo: Use macros for passing data to Fortran in braid_TimeGrid_F90_Iface.
Introduced an option for the user to supply an initial fine grid. The use of this option is demonstrated in examples/ex-01.c I added an index in AccessStatus/StepStatus to simplify output to files based on time point index value. Since we know the index in braid, this makes it easier for the user to assign a correct file name.
Exposed more functions in Fortran interface Added parameters in PrintStats, destroy bstatus
commit 8cb58b636f33b9e0232dccf2cce20aa46af9a2a7
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Fri Aug 19 14:49:00 2016 -0700
Forgot to turn off this part of the Fortran interface
commit b84ab26c1321478cc92bcceb820cd96bbc785b9a
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Fri Aug 19 14:45:43 2016 -0700
Last edits before pulling this branch into master
commit 9bb85c51403193e9780879e75584a5295eb1d524
Merge: 90d95df 191d166
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Fri Aug 19 14:11:25 2016 -0700
Merge branch 'master' into hessenthaler
Conflicts:
_braid.c
braid_status.c
braid_status.h
commit 90d95dfe4ee9dd521d44de474c6fc67c6b5d5e96
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Fri Aug 19 13:43:54 2016 -0700
Edits related to making a pull request -- cleaned up a few things
commit 191d1664e4d62d227eea86c2f6ab7911d5a24154
Author: Ben O'Neill <oneill10@llnl.gov>
Date: Tue Aug 9 09:48:59 2016 -0700
Bug-Fix: Recv buffer was being used uninitialized on processors where iupper > gupper ( i.e processors with no time points. ), as no message was recieved there.
commit ade6c45b0f6856e9fa974bdc0064f002d050e1c5
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Thu Aug 4 15:06:59 2016 -0700
Updating the access level regression tests for the
new access levels 2 and 3.
commit 310b4757ce747791a779180eabad1b44b7f767d8
Merge: c33b639 843245d
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Thu Aug 4 01:00:04 2016 -0700
Merge branch 'master' of ssh://mystash.llnl.gov:7999/xbraid/xbraid
commit 843245d6ce2ab3697b8894dc1fa45f167c8549e8
Merge: 5b05af7 462a55c
Author: Schroder, Jacob B <schroder2@llnl.gov>
Date: Wed Aug 3 13:42:48 2016 -0700
Merge pull request #2 in XBRAID/xbraid from shellvector to master
* commit '462a55c45500c17b9993eded035acdb48073bfd2':
Adds a comment to describe the problem solved in ex-bdf2
Change interface of _braid_UGetIndex. The function now returns two integers. One is the local index if something (shell or vector) is stored or -1 if nothing is stored. The second is a flag indicating the storage status: 0 if it's a full vector, -1 if it's a shell, and -2 if nothing is stored.
Following Jacob's suggestion, adds some more explicit comments, rename a variable and correct indentation
Adds a verification on the number of my_Access calls for different access level parameter (0 to 3).
Adds an example using the shellvector feature. It's a simple 1D linear problem using a BDF2 method. I have added the associated test script and expected results.
Adds and modifies comments to reflect the new shell feature and the new access levels
Removes tabs and correction of indentation
Re-add the access_level variable in braid_Drive that has been removed during merge, but that I still use.
Modification of the access_level. Now, we have: - access_level == 0 : do nothing - access_level == 1 : call at the end of the simulation (fine grid) - access_level == 2 : call at the end of each iteration and end of simulation (fine grid only) - access_level == 3 : same as other, plus call in FRestrict, FInterp and FRefine (on all levels).
Adds a missing initialization of sfree to NULL when not using shell vectors. Normally this initialization is useless.
Correction of two bugs. The first is in FRefine, the array r_ta now as a size of r_npoints+2, to have r_npoints[-1] and r_npoints[r_npoints]. We need these values when doind spatial refinement. The second is in _braid_InitHierarchy, if very few points so that only one level is built, then core.max_levels is not reset to 1 anymore.
Bug of missing shell after refining. Now all points from coarse grid are passed to th new coarse point on refined grid
Modifications to use shell vectors when possible. When the shell feature is activated, U Vectors are saved on all points in time ( C- and F-points ). On C-points, vectors are full, and on F-points only the shell structure is saved. It is up to the used to define InitShell, CloneShell and FreeShell. I think only full vectors are communicated between processors, so no modification of the *Buffer functions are required. Apparently work with my code...
First modifications to take into account the shell structure of braid vectors. Does not modify anything in the algorithm yet.
commit 462a55c45500c17b9993eded035acdb48073bfd2
Author: Matthieu B Lecouvez <lecouvez1@llnl.gov>
Date: Wed Aug 3 10:45:04 2016 -0700
Adds a comment to describe the problem solved in ex-bdf2
commit e84cee4b742ba38d385188534bae7e6f319b29ca
Author: Matthieu B Lecouvez <lecouvez1@llnl.gov>
Date: Mon Aug 1 11:04:57 2016 -0700
Change interface of _braid_UGetIndex. The function now returns two integers.
One is the local index if something (shell or vector) is stored or -1 if
nothing is stored. The second is a flag indicating the storage status:
0 if it's a full vector, -1 if it's a shell, and -2 if nothing is stored.
commit 7e7f980635cd84b232f3961d3c2b2b78daeebfb3
Author: Matthieu B Lecouvez <lecouvez1@llnl.gov>
Date: Mon Aug 1 10:16:53 2016 -0700
Following Jacob's suggestion, adds some more explicit comments,
rename a variable and correct indentation
commit d4255bddb8c60fb77c4e13bcfcbfa7f87e3041a6
Author: Andreas Hessenthaler <hessenthaler1@llnl.gov>
Date: Fri Jul 29 11:47:33 2016 -0700
Added: step_status_get_istop in Fortran interface.
commit fa9518c67245b49ad8a0a8a1c3ad0b7a9c63c338
Author: Andreas Hessenthaler <hessenthaler1@llnl.gov>
Date: Thu Jul 28 17:18:08 2016 -0700
Removed commented line.
commit 1dd8a58e167405ebb2fa3d135d0edbd3832f89f1
Author: Andreas Hessenthaler <hessenthaler1@llnl.gov>
Date: Thu Jul 28 17:13:38 2016 -0700
Added: Regression test for ex-01/ex-01f testing the user-defined fine time grid feature.
Added: braid_access_status_get_istop in Fortran interface.
Bugfix: ex-01f now queries access status to determine filename index as we don't have dt = const, in general.
commit 1862b9094cb8d16cdea81b8c2f6d5ed0f7320333
Author: Andreas Hessenthaler <hessenthaler1@llnl.gov>
Date: Thu Jul 28 09:47:16 2016 -0700
Set preprocessor directive braid_Fortran_TimeGrid to 0 by default
commit 70d2ae77987cd310ee39044e15004c0225fd1888
Author: Andreas Hessenthaler <hessenthaler1@llnl.gov>
Date: Thu Jul 28 09:37:55 2016 -0700
Added: Fortran interface to enable user supplied fine time grid.
Modified: Pass values by reference so we have interoperability for C/Fortran.
Modified: ex-01f now demonstrates how to define a fine time grid.
Bugfix: Set initial tgrid function pointer to NULL.
Todo: Use macros for passing data to Fortran in braid_TimeGrid_F90_Iface.
commit d13528e590cbc6ee68a55a50d307065f3388f742
Author: Andreas Hessenthaler <hessenthaler1@llnl.gov>
Date: Wed Jul 27 10:53:37 2016 -0700
Introduced an option for the user to supply an initial fine grid.
The use of this option is demonstrated in examples/ex-01.c
I added an index in AccessStatus/StepStatus to simplify output to files
based on time point index value. Since we know the index in braid,
this makes it easier for the user to assign a correct file name.
I will work on the Fortran interface for this option next.
commit c33b6391b1f6dadf738591b66e4b0802243ad797
Merge: df5b1b4 5b05af7
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Wed Jul 27 01:00:05 2016 -0700
Merge branch 'master' of ssh://mystash.llnl.gov:7999/xbraid/xbraid
commit c68c00c11dc3300f1055167fdf6673d64256b09c
Merge: 651136e 5b05af7
Author: Matthieu B Lecouvez <lecouvez1@llnl.gov>
Date: Tue Jul 26 08:33:51 2016 -0700
Merge branch 'master' into shellvector
commit 5b05af7bed3a26a1a2bcbd40e94adc6329b9e069
Author: Matthieu B Lecouvez <lecouvez1@llnl.gov>
Date: Tue Jul 26 08:31:13 2016 -0700
Bug Fix -- Deallocate the BufferStatus object in braid_Destroy.
commit 651136e591dbacc33adf748171fae779ff34d753
Author: Matthieu B Lecouvez <lecouvez1@llnl.gov>
Date: Tue Jul 26 08:21:25 2016 -0700
Adds a verification on the number of my_Access calls for different
access level parameter (0 to 3).
commit 8dbfccbfa00ef793711fccf7e34d3e7e485cee7e
Author: Matthieu B Lecouvez <lecouvez1@llnl.gov>
Date: Mon Jul 25 15:50:28 2016 -0700
Adds an example using the shellvector feature. It's a simple 1D linear
problem using a BDF2 method. I have added the associated test script
and expected results.
commit f661b95f0fbeefb4ff4560664887332455dea2e8
Merge: 82cd542 d8875db
Author: Andreas Hessenthaler <hessenthaler1@llnl.gov>
Date: Mon Jul 25 08:47:10 2016 -0700
Merge branch 'shellvector' into hessenthaler
commit 82cd5422c262aa18260c96049e54a77be588adf1
Author: Andreas Hessenthaler <hessenthaler1@llnl.gov>
Date: Mon Jul 25 08:27:52 2016 -0700
Exposed more functions in Fortran interface
Added parameters in PrintStats, destroy bstatus
commit d8875dbe2540c73b37ce71ce873a523c0145b82b
Merge: a58b718 01a8374
Author: Matthieu B Lecouvez <lecouvez1@llnl.gov>
Date: Wed Jul 13 16:37:30 2016 -0700
Merge branch 'master' into shellvector
commit df5b1b47458728b1a8c3a83921d45c97df96f9d8
Merge: 9f66ae1 01a8374
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Tue Jul 12 01:00:07 2016 -0700
Merge branch 'master' of ssh://mystash.llnl.gov:7999/xbraid/xbraid
commit 01a8374bce7bfd3c367e0b2686411ad7049d6153
Merge: 751b368 74e8a8e
Author: Falgout, Rob <falgout2@llnl.gov>
Date: Mon Jul 11 12:35:09 2016 -0700
Merge pull request #5 in XBRAID/xbraid from oneill to master
* commit '74e8a8e6721cdcb29e58804a04daca4d4bd4a693':
For some reason, the commit I pushed for the buffer status change didnt have this line. It is needed to initialize the buffer status in the core.
Bug fix: FRefine stalls if there are more processors than there are new time steps on the new fine grid. Tracked it down to these f_next and f_first values. If a processor has zero time points on the new fine grid then f_next doesnt get initialized. This causes recv_msg and send_msg to be set on that processor even though it have no points
Rather specific bug fix. Make it explicit that if nlevels = 1 we should not complete a v cycle.
commit 9f66ae15e06fb85b60abb99f71f42e055292c804
Merge: 49027fa 751b368
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Fri Jul 1 01:00:03 2016 -0700
Merge branch 'master' of ssh://mystash.llnl.gov:7999/xbraid/xbraid
commit 751b368f6666835bf3c3fa09a780009d4055d81a
Author: Ben O'neill <oneill10@sierra330.llnl.gov>
Date: Thu Jun 30 17:25:42 2016 -0700
Bug Fix -- Initialize the BufferStatus object in the core
commit 49027fa42762eae1f860a22ed34e9ae865d2ec5d
Merge: d96bddc 8918cb4
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Thu Jun 30 01:00:06 2016 -0700
Merge branch 'master' of ssh://mystash.llnl.gov:7999/xbraid/xbraid
commit 8918cb40d0956208955dfec8bbdf3d446d847d91
Author: Tzanio <tzanio@llnl.gov>
Date: Wed Jun 29 10:32:22 2016 -0700
Fixed compiler errors.
commit 74e8a8e6721cdcb29e58804a04daca4d4bd4a693
Author: O'neill <oneill10@mollow.llnl.gov>
Date: Fri Jun 24 12:23:03 2016 -0700
For some reason, the commit I pushed for the buffer status change didnt have this line. It is needed to initialize the buffer status in the core.
commit 26b8fc1b6bca88204c3d9a6c847078a3bd73c58f
Author: O'neill <oneill10@mollow.llnl.gov>
Date: Thu Jun 23 09:14:07 2016 -0700
Bug fix: FRefine stalls if there are more processors than there are new time steps on the new fine grid. Tracked it down to these f_next and f_first values. If a processor has zero time points on the new fine grid then f_next doesnt get initialized. This causes recv_msg and send_msg to be set on that processor even though it have no points
commit 22207da346b25a6945bfb70608c57fac62963c9f
Author: O'neill <oneill10@mollow.llnl.gov>
Date: Wed Jun 8 09:56:11 2016 -0700
Rather specific bug fix. Make it explicit that if nlevels = 1 we should not complete a v cycle.
I had an example where FRefinment was completed, but the resulting grid still only had one time level. This caused a seg fault as a V cycle was initialised with only one grid level in time
commit d96bddce25f6821bb4064b7794047f926817d1d8
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Fri Jun 3 15:57:49 2016 -0700
Fixed two bugs regarding the new BufStatus structure. The regression
tests now pass cleanly
commit e8384bb3b12383e1a2e5f04f47ecab8809db5b79
Merge: a35a804 d287977
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Fri Jun 3 14:59:35 2016 -0700
Merge branch 'oneill' of ssh://mystash.llnl.gov:7999/xbraid/xbraid
Conflicts:
docs/Examples.md
commit d287977464d282671c4083b37a7b131d98770b48
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Fri Jun 3 14:53:35 2016 -0700
Cleaned up a few small things, mostly comments and then
changed docs/Examples.md to reflect the new BufStatus structure
commit a35a804ae1cdba9a713c4b6c3eb1d17f0f2e2c10
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Fri Jun 3 14:36:44 2016 -0700
Added a few simple regression tests for the MFEM drivers. These had been
missing, but now there's a few checks on the MFEM interface.
commit da05aa7717d838cabc11d629136df2a6cf261efe
Author: O'neill <oneill10@mollow.llnl.gov>
Date: Thu Jun 2 08:04:22 2016 -0700
frefine --> messagetype
commit fbed16345e020685f58df8f843b1df54d215fa79
Author: Rob Falgout <rfalgout@llnl.gov>
Date: Wed Jun 1 16:28:34 2016 -0700
Minor cosmetic changes.
commit 0f64819c35c6bb878f7a319d7cbb6a1a7d28b247
Author: O'neill <oneill10@mollow.llnl.gov>
Date: Wed Jun 1 11:44:28 2016 -0700
small bug fixes and comment updates
commit 163a171e4178978541d71ab4e11edd823a434032
Author: O'neill <oneill10@mollow.llnl.gov>
Date: Wed Jun 1 09:03:45 2016 -0700
Update the examples to agree with the new interface for buf**
commit f7c9fd1f2e9da08dc44d48895376d53a3cce121d
Author: O'neill <oneill10@mollow.llnl.gov>
Date: Tue May 31 16:52:58 2016 -0700
BufPack now accepts a buffer and status structure with the option on setting the buf_size through a status structure get routine
commit 7840333e8fb25c22d018de7d7c7842bc5f213836
Author: O'neill <oneill10@mollow.llnl.gov>
Date: Tue May 31 15:48:44 2016 -0700
Fixes based on comments regarding r_space variable. r_space is now set either 0 or 1 based on the user set value in the step status structure. Also changed the logic for when to call space refinment inside the F refine function
commit 010a73773aa29a65485a2f9529f09c725351092c
Author: O'neill <oneill10@mollow.llnl.gov>
Date: Fri May 27 15:34:46 2016 -0700
Add a buffer status structure -- braid.hpp is untested
commit fc40b639ca149f92586e7a9d3d57ac86053817a4
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Wed May 25 08:52:33 2016 -0700
Braid version 2.0.0, This commit just updates the docs a little
commit 199b3ba27f9f08cf3fa8272c9bb6dd973ef73b81
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Tue May 24 22:36:07 2016 -0700
Updated some documentation. It still needs work, but it's good enough for the next
release.
commit 3f6386d06c692e10840456e51a00e1dace3b4abb
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Mon May 23 11:37:14 2016 -0700
Woops...this line got accidentally deleted in the last commit
commit 297de4b41975dd5d426f2f878f68ea08484940c5
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Mon May 23 11:32:54 2016 -0700
Updated version numbers to 2.0.0
commit 3d85b8f4a04de7cda54fbffd067153f5761f2ec0
Merge: ad368e2 da9a4df
Author: Rob Falgout <rfalgout@llnl.gov>
Date: Fri May 20 17:31:29 2016 -0700
Merge branch 'master' of ssh://mystash.llnl.gov:7999/xbraid/xbraid
commit ad368e247d2a2987b08c4b8eb62ba6b5b34dd32c
Author: Rob Falgout <rfalgout@llnl.gov>
Date: Fri May 20 17:28:09 2016 -0700
First pass at modifying the documentation for the next release
This changes the function Phi to Step and adds more description (mostly to
Example 1) on the new residual, adaptive time refinement, and storage options.
commit da9a4dfd6ecdc25a8febdc0f5ae5354e03a85a8c
Author: Jacob Schroder <schroder2@llnl.gov>
Date: Fri May 20 13:00:55 2016 -0700
mall updates to create_release.sh
- Changing release number to 2.0.0
- Changed it to point to stash, and not /usr/casc/hypre/braid/git/braid
This was way out-of-date
commit 135c448131f6a70d57ca21a40837196cef2e0de9
Merge: f392ef5 4856a48
Author: Schroder, Jacob B <schroder2@llnl.gov>
Date: Fri May 20 12:48:50 2016 -0700
Merge pull request #3 in XBRAID/xbraid from rebalance-dev-fix to master
* commit '4856a48eafd6ec2b11de0d95972634ce61927226':
In class MFEMBraidApp, replace the local restriction operators with the local prolongation operators - this is what mfem constructs now.
Changed MFEMBraidApp::InitMultilevelApp so it compiles with current MFEM version where the Mesh TWO_LEVEL interface was removed. This code should hopefully be equivalent.
commit 4856a48eafd6ec2b11de0d95972634ce61927226
Author: Veselin Dobrev <dobrev@llnl.gov>
Date: Thu May 19 17:01:17 2016 -0700
In class MFEMBraidApp, replace the local restriction operators with the local
prolongation operators - this is what mfem constructs now.
commit 5ac403a7f84a9006c687553b3c43a27328186a17
Author: Jakub Cerveny <cerveny1@llnl.gov>
Date: Fri May 13 16:54:17 2016 -0700
Changed MFEMBraidApp::InitMultilevelApp so it compiles with current MFEM
version where the Mesh TWO_LEVEL interface was removed. This code should
hopefully be equivalent.
commit 400228fb7ed367deaea684c3f94805c47e33f8eb
Author: O'neill <oneill10@mollow.llnl.gov>
Date: Thu May 12 10:21:03 2016 -0700
Set size_ptr prior to calling bufpack such that: size_ptr = 0 @ entry
to bufpack indicates the message will be used for completing a time
step on the next processor. @entry size_ptr = 1 indicates the message
will be used for load balenceing in FRefine.
Reason: Far less information is needed when completing a time step
than when moving an entire vector to a new processor. In my
case each vector saves its entire spatial heirachy of meshes.
To complete a time step on the next processor the message only
needs to contain the current function and the current mesh. To
move an entire vector to a new processor the message must contain
the fine grid solution and mesh + all other coarse grid messages.
commit 2820b12b2137d36c80316fa61fcef4321577ff1f
Author: O'neill <oneill10@mollow.llnl.gov>
Date: Thu May 12 09:38:46 2016 -0700
Added Functionality that allows for spatial refinement following a
Vcycle. Currently the user can only complete spatial refinement following
a Vcycle if temporal refinement is also completed. Using the StepStatus
struct the user can set the flag r_space. Follwing the Vcycle, if refine is
turned on, no temporal refinment was completed (globally), and r_space > 0 for
some local time step, then the users spatial refinment routine is called at all
local time points. An MPI_AllReduce checks to see if spatial refinement was completed
on any processor. If yes then the *refined_ptr returns 2 and the Vcycle is repeated, if
not, then *refined_ptr returns 0 and braid moves onto the next Vcycle.
commit f392ef5ccd1bbc5efbdeabaf3b136342741ec5fb
Author: Veselin Dobrev <dobrev@llnl.gov>
Date: Mon May 9 15:05:36 2016 -0700
Moved the additional "hypre" functions from drive-04.cpp to hypre_extra.hpp, so
they can be used in other drivers.
Added a number of options to drive-05: implicit time stepping; different options
for the implicit solver; a switch between second (diffusion) and fourth order
dissipation terms; random initialization for the initial vectors, etc.
Some formatting edits.
commit f5860c1cc7582bacea279436d005de6f15235f89
Author: Veselin Dobrev <dobrev@llnl.gov>
Date: Mon May 9 14:58:03 2016 -0700
In the build system, added support for compiling the mfem examples using any
external packages that mfem was configured with, e.g. SuiteSparse.
Also, iclude makefile.user (if present) from makefile.inc -- this is just
another way for users or developers to configure braid without modifying
makefile.inc.
commit a58b718b3a9815aa160b90e23a372ef149e4f497
Author: Matthieu B Lecouvez <lecouvez1@llnl.gov>
Date: Fri Apr 22 17:02:59 2016 -0700
Adds and modifies comments to reflect the new shell feature and the
new access levels
commit f5f5e227af9ca04365d8e89de43a33dbc344cbfb
Author: Matthieu B Lecouvez <lecouvez1@llnl.gov>
Date: Fri Apr 22 14:22:56 2016 -0700
Removes tabs and correction of indentation
commit abbf7b643f5d14fdb7ad510ab58a9130e3bb2180
Author: Matthieu B Lecouvez <lecouvez1@llnl.gov>
Date: Thu Apr 21 14:49:57 2016 -0700