cbfba74 | Rajesh Kommu | 08 September 2014, 18:08:34 UTC | changed default value of use_petsc flag to off | 08 September 2014, 18:08:34 UTC |
27720cb | Rajesh Kommu | 08 September 2014, 18:06:14 UTC | Added PETSc configuration and usage sections to the CitcomS manual | 08 September 2014, 18:06:14 UTC |
774ebeb | Rajesh Kommu | 08 September 2014, 17:01:39 UTC | Setting petsc_schur=on will fail with a "Schur complement reduction approach is not implemented yet" messgage | 08 September 2014, 17:01:39 UTC |
8969fe9 | Rajesh Kommu | 08 September 2014, 16:48:58 UTC | Added general_stokes_solver_teardown() method for PETSc related cleanup | 08 September 2014, 16:48:58 UTC |
23c90a7 | Rajesh Kommu | 23 August 2014, 19:16:05 UTC | correctly generate CPPFLAGS, LDFLAGS, LIBS for building with PETSc | 23 August 2014, 19:16:05 UTC |
52c9d9d | Rajesh Kommu | 21 August 2014, 20:45:53 UTC | fixed the signature of parallel_process_finalize | 21 August 2014, 20:45:53 UTC |
5faeb98 | Rajesh Kommu | 21 August 2014, 20:43:18 UTC | added some missing USE_PETSC guards | 21 August 2014, 20:43:18 UTC |
0c79a2e | Rajesh Kommu | 21 August 2014, 20:30:40 UTC | configure.ac now checks for the existence of PETSc libraries if --with-petsc=yes is specified. All PETSc related code is enclosed by USE_PETSC conditional compilation guards | 21 August 2014, 20:30:40 UTC |
2d80412 | Rajesh Kommu | 20 August 2014, 22:12:17 UTC | removed the empty definition of solve_Ahat_p_fhat_petsc | 20 August 2014, 22:12:17 UTC |
c0bff4a | Rajesh Kommu | 20 August 2014, 22:09:40 UTC | removed commented out v_from_vector_petsc, v_from_vector_pseudo_surf_petsc, p_to_nodes_petsc | 20 August 2014, 22:09:40 UTC |
c09bc23 | Rajesh Kommu | 20 August 2014, 21:59:40 UTC | cleaned up strip_bcs_from_residual_PETSc function | 20 August 2014, 21:59:40 UTC |
15e05e4 | Rajesh Kommu | 20 August 2014, 19:37:00 UTC | completed solve_Ahat_p_fhat_PETSc_Schur implementation | 20 August 2014, 19:37:00 UTC |
9f568ea | Rajesh Kommu | 20 August 2014, 19:36:33 UTC | Added velocity and pressure vector initialization to solve_Ahat_p_fhat_PETSc_Schur | 20 August 2014, 19:36:33 UTC |
f90ce24 | Rajesh Kommu | 20 August 2014, 19:36:02 UTC | Started on the solve_Ahat_p_fhat_PETSc_Schur implementation | 20 August 2014, 19:36:02 UTC |
6890fb0 | Rajesh Kommu | 20 August 2014, 19:35:36 UTC | completed replacing VecGetOwnershipRange with VecGetArray | 20 August 2014, 19:35:36 UTC |
6c6619f | Rajesh Kommu | 20 August 2014, 19:35:06 UTC | fixed some bugs; started replacing VecGetOwnershipRange with VecGetArray | 20 August 2014, 19:35:06 UTC |
748ea48 | Rajesh Kommu | 20 August 2014, 19:34:21 UTC | Added calls to various PETSc versions of solve_Ahat_p_fhat functions | 20 August 2014, 19:34:21 UTC |
145990d | Rajesh Kommu | 20 August 2014, 19:33:42 UTC | cleaned up the signatures of solve_Ahat_p_fhat_CG_PETSc and solve_Ahat_p_fhat_BiCG_PETSc and added a petsc_uzawa_tol parameter | 20 August 2014, 19:33:42 UTC |
9a1bf03 | Rajesh Kommu | 20 August 2014, 19:33:17 UTC | added solve_Ahat_p_fhat_CG_PETSc implementation | 20 August 2014, 19:33:17 UTC |
0071683 | Rajesh Kommu | 20 August 2014, 19:32:19 UTC | added solve_Ahat_p_fhat_BiCG_PETSc implementation | 20 August 2014, 19:32:19 UTC |
df559c0 | Rajesh Kommu | 20 August 2014, 19:30:48 UTC | Added PETSc related initialization to general_stokes_solver_setup Moved MATSHELL and PCSHELL contexts to global_defs.h from petsc_citcoms.h | 20 August 2014, 19:30:48 UTC |
4bfc600 | Rajesh Kommu | 20 August 2014, 19:29:51 UTC | commented out p_to_nodes_petsc, v_from_vector_petsc, v_from_vector_pseudo_surf_petsc, solve_constrained_flow_iterative_petsc | 20 August 2014, 19:29:51 UTC |
53fbc9c | Rajesh Kommu | 20 August 2014, 19:29:28 UTC | completed the implementation of p_to_nodes_petsc | 20 August 2014, 19:29:28 UTC |
b8f1712 | Rajesh Kommu | 20 August 2014, 19:28:46 UTC | started working on the v_from_vector_petsc, v_from_vector_pseudo_surf_petsc, p_to_nodes_petsc implementations | 20 August 2014, 19:28:46 UTC |
f2ced72 | Rajesh Kommu | 20 August 2014, 19:28:26 UTC | started working on solve_constrained_flow_iterative_petsc and solve_ahat_p_fhat_petsc methods | 20 August 2014, 19:28:26 UTC |
e0055e4 | Rajesh Kommu | 20 August 2014, 19:28:05 UTC | Added the petsc_citcoms.h and Petsc_citcoms.c files, modified lib/Makefile.am, fixed some minor bugs | 20 August 2014, 19:28:05 UTC |
7d785cb | Rajesh Kommu | 20 August 2014, 19:27:39 UTC | Added PETSc related flags, use_petsc, petsc_linear, petsc_nonlinear and the initialization of these flags | 20 August 2014, 19:27:39 UTC |
3958782 | Rajesh Kommu | 20 August 2014, 19:27:22 UTC | Added PETSc header files, initialization and finalization code | 20 August 2014, 19:27:22 UTC |
fb1839e | Eric Heien | 14 February 2014, 22:32:46 UTC | Merge pull request #1 from karredondo/master Add ability to combine "heating" output | 14 February 2014, 22:32:46 UTC |
7ca37a6 | karredondo | 14 February 2014, 22:30:31 UTC | Add ability to combine "heating" output Katrina Arredondo Heating output is in terms of elements instead of nodes. The data is placed in a third combine file "ele" aside from "cap" and "opt." | 14 February 2014, 22:30:31 UTC |
c770f8e | Rajesh Kommu | 13 February 2014, 19:46:45 UTC | Further cleanup of the cover page of the manual | 13 February 2014, 19:46:45 UTC |
3ebabdc | Eric Heien | 13 February 2014, 18:02:15 UTC | Improved manual cover layout, removed old cover file | 13 February 2014, 18:02:15 UTC |
90cb771 | Rajesh Kommu | 11 February 2014, 00:19:39 UTC | Checked in the LaTeX source file for the CitcomS manual, along with the images needed to generate the PDF version of the manual. | 11 February 2014, 00:19:39 UTC |
8fde56f | Eh Tan | 18 October 2013, 03:58:15 UTC | update ignore list | 18 October 2013, 03:58:15 UTC |
fd2b0d6 | CIG Backend | 11 October 2013, 16:35:38 UTC | Update gitignore | 11 October 2013, 16:35:38 UTC |
754dc95 | CIG Backend | 11 October 2013, 16:24:31 UTC | Readd archimedes | 11 October 2013, 16:24:31 UTC |
7c6b5ad | CIG Backend | 11 October 2013, 05:10:13 UTC | m4 submodule | 11 October 2013, 05:10:13 UTC |
8c294e9 | CIG Backend | 10 October 2013, 22:03:31 UTC | Readd anisotropic code | 10 October 2013, 22:03:31 UTC |
f886a03 | Thorsten Becker | 20 September 2013, 14:58:59 UTC | | 20 September 2013, 14:58:59 UTC |
3cd9e30 | Eric Heien | 24 May 2013, 17:42:50 UTC | Added reference viscosity fix from Shijie Zhong | 24 May 2013, 17:42:50 UTC |
ed21917 | Thorsten Becker | 25 March 2013, 21:56:47 UTC | Updating to newest version for testing purposes. | 25 March 2013, 21:56:47 UTC |
c5526ec | Thorsten Becker | 01 September 2012, 23:51:37 UTC | Fixed ggrd velocity boundary condition weirdness, I hope. This may go back to 16065 vs. 16067, and bug may have stuck through 20621. | 01 September 2012, 23:51:37 UTC |
2b8ed1f | Thorsten Becker | 22 August 2012, 21:21:52 UTC | Implemented improved tracer fix by Rene. "a new method of finding the element of a tracer that is too close to an element boundary. Up to now such a tracer was shifted by a constant epsilon theta and phi. If the element boundary is parallel to this theta/phi direction it is not guaranteed to work well (thus the number_of_tries check), and additionally I got the problem that sometimes all elements refuse this tracer. Eh checked in a workaround for this (r15742), deleting orphan tracers in Tracer_setup.c. Because I did not know this, I created my own bugfix moving the tracers an epsilon amount orthogonal to all boundaries that are too close. In order to save computing time I use the already computed vectors for the element boundaries (this assumes that the element boundaries are nearly orthogonal to each other, but unless somebody tries to change CitcomS elements that should work fine). The shift happens now in cartesian coordinates since the boundary-vectors are cartesian and the radius-coordinate of the tracer is normalized prior to this check anyway, so I just need to re-normalize it after the shift. For now I did not touch all the now (hopefully) useless security checks but as far as I can see they do no harm either, so we can remove them later." | 22 August 2012, 21:21:52 UTC |
11854a5 | Thorsten Becker | 21 August 2012, 16:30:28 UTC | Implemented Output_vtk.c fixes from Rene. | 21 August 2012, 16:30:28 UTC |
74c56ef | Thorsten Becker | 28 June 2012, 09:30:30 UTC | Minor changes to instructions | 28 June 2012, 09:30:30 UTC |
d5fbdb0 | Eric Heien | 02 February 2012, 18:23:14 UTC | Changing to version 3.2.0 since there are many significant changes | 02 February 2012, 18:23:14 UTC |
8d8016f | Eric Heien | 02 February 2012, 00:41:19 UTC | Added missing file to distribution | 02 February 2012, 00:41:19 UTC |
b0d8a75 | Eric Heien | 02 February 2012, 00:29:10 UTC | Updated NEWS for past 2.5 years | 02 February 2012, 00:29:10 UTC |
162e668 | Eric Heien | 01 February 2012, 21:11:35 UTC | Updated ChangeLog, configure.ac for 3.1.2 release Set default configuration to not setup Pyre Added note in examples directory regarding Pyre and Cookbooks | 01 February 2012, 21:11:35 UTC |
eb74d9f | Thorsten Becker | 24 January 2012, 21:54:19 UTC | Fixed minor ggrd assignment bug. | 24 January 2012, 21:54:19 UTC |
199d1d8 | Thorsten Becker | 24 January 2012, 18:44:23 UTC | Debugging GGrd | 24 January 2012, 18:44:23 UTC |
74ce5d8 | Thorsten Becker | 24 January 2012, 18:03:52 UTC | Updating to latest version for sync. | 24 January 2012, 18:03:52 UTC |
552bc26 | Eh Tan | 19 January 2012, 19:24:59 UTC | Stop configure if pyre is enabled and python 2.7 is used. See Issue622. | 19 January 2012, 19:24:59 UTC |
f4cf4db | Eh Tan | 17 January 2012, 04:57:06 UTC | updated installation instruction | 17 January 2012, 04:57:06 UTC |
3a23d00 | Eh Tan | 17 January 2012, 02:41:15 UTC | When python 2.7 is used, ask users to download pythia manually. See issue622 and issue606. | 17 January 2012, 02:41:15 UTC |
15a09f1 | Thorsten Becker | 14 January 2012, 08:00:41 UTC | Made sure that compositional viscosity without compositional buoyancy can be assigned without setting ibuoy=1 and the buoyancy number to zero. | 14 January 2012, 08:00:41 UTC |
c9704b3 | Thorsten Becker | 20 December 2011, 18:04:26 UTC | Merged with recent check in, minor changes. | 20 December 2011, 18:04:26 UTC |
51e1984 | Eh Tan | 14 November 2011, 05:06:30 UTC | Fixed a bug in reflective TBC where y-index should be used, not z-index. The bug is harmless though. I checked the code history. This bug exists since CitcomS 1.0, but not in CitcomCU nor CitcomT. | 14 November 2011, 05:06:30 UTC |
d8f08c4 | Eh Tan | 14 October 2011, 18:27:24 UTC | fixed a bug that h5close_field() didn't have a return statement, reported by Nicolas Flament in issue611. | 14 October 2011, 18:27:24 UTC |
0f1fbbf | Eh Tan | 18 September 2011, 03:54:13 UTC | fixed a bug that when the perturblayer is on the internal boundary, the temperature perturbation wasn't added to the lower processor. | 18 September 2011, 03:54:13 UTC |
cebdfcb | Thorsten Becker | 31 July 2011, 20:38:35 UTC | Updating to newest version, don't think that there should be any significant changes on my side. | 31 July 2011, 20:38:35 UTC |
1e7ff88 | Eh Tan | 12 July 2011, 16:27:58 UTC | enhancing VTK output, mostly contributed by Rene Gassmoeller at GFZ. - A new input parameter 'vtk_format', which can be either 'ascii' (default) or 'binary'. When 'binary', gzip and base64 encoding are used. - The node ordering is changed to CitcomS native ordering. The consequence is the axis are rotated as z in CitcomS is mapped to X in vtk, and x to Y, y to Z. - More data can be outputted in vtk format - Using .pvts file for regional model; .vtm and .visit file for global model (.vtm format provided by Tobias Hoeink at Rice, .visit format provided by Kat at visitusers.org/forum) | 12 July 2011, 16:27:58 UTC |
56f6015 | Thorsten Becker | 21 March 2011, 00:13:58 UTC | Fixed typo for anisotropic stress computation. | 21 March 2011, 00:13:58 UTC |
a0153f3 | Thorsten Becker | 22 February 2011, 04:13:21 UTC | Added a flag force_iteration which if on (off by default), will enforce an iterative solution to the velocity fields even if linear rheologies were chosen. This is useful for debugging/robustness tests. | 22 February 2011, 04:13:21 UTC |
966fcd8 | Thorsten Becker | 18 February 2011, 19:45:49 UTC | Modified storage of anisotropic viscosity files. Those are now | 18 February 2011, 19:45:49 UTC |
7ea7b1d | Thorsten Becker | 14 February 2011, 05:45:26 UTC | Improved vtk output, still for Dan to test. | 14 February 2011, 05:45:26 UTC |
446b149 | Thorsten Becker | 12 February 2011, 05:30:30 UTC | Aedded the actual files because I didn't know how to link from CitcomCU. Those are didentical | 12 February 2011, 05:30:30 UTC |
7a6fa87 | Thorsten Becker | 12 February 2011, 05:20:32 UTC | | 12 February 2011, 05:20:32 UTC |
1cfc861 | Thorsten Becker | 08 February 2011, 04:06:28 UTC | Removed the srand() call that seemed like a good idea (rev. 17175) but screwed up tracer-based runs with large number of processors (12x3x3x2). Weird, but oh well. Someone advise why this didn't work? | 08 February 2011, 04:06:28 UTC |
9086505 | Thorsten Becker | 07 February 2011, 19:00:57 UTC | Added one missing line for anisotropic viscosity init. | 07 February 2011, 19:00:57 UTC |
8507b40 | Thorsten Becker | 07 February 2011, 00:32:59 UTC | Reversed, temporarily, the use of nearneighbor because it makes it easier to compare model computations with previous versions. | 07 February 2011, 00:32:59 UTC |
12ac22c | Thorsten Becker | 06 February 2011, 23:57:18 UTC | Fixed typo that was introduced in 17555 and made temperature field assignment apparently screwed up. | 06 February 2011, 23:57:18 UTC |
4fb41eb | Thorsten Becker | 05 February 2011, 23:45:04 UTC | Backward compatibility of neglect of both divergence and pressure for convergence checks. | 05 February 2011, 23:45:04 UTC |
c21beb1 | Thorsten Becker | 04 February 2011, 05:03:01 UTC | Temporarily moved velocity gradient computation to ensure compilation. | 04 February 2011, 05:03:01 UTC |
9c045e8 | Thorsten Becker | 03 February 2011, 21:06:35 UTC | Uptdated prototypes file. Fixed a few compiler complaints. Made sure that atemp is always set in Viscosity_structures, making mat_control > 0 feasible, not just 0 or 1 | 03 February 2011, 21:06:35 UTC |
8dcea68 | Thorsten Becker | 03 February 2011, 17:11:42 UTC | Forgot some functions, not sure why the linker didn't complain previously. | 03 February 2011, 17:11:42 UTC |
93f3131 | Thorsten Becker | 02 February 2011, 01:09:51 UTC | Futher synchronization with CitcomCU anisotropic viscosity. | 02 February 2011, 01:09:51 UTC |
4d1c25b | Thorsten Becker | 02 February 2011, 00:30:25 UTC | Trial synchronization between CitcomCU and CitcomS anisotropic viscosity. | 02 February 2011, 00:30:25 UTC |
2f2115e | Thorsten Becker | 26 January 2011, 00:14:53 UTC | Added parentheses for logic statements as per compiler suggestion | 26 January 2011, 00:14:53 UTC |
2598655 | Thorsten Becker | 19 January 2011, 21:00:20 UTC | Checking in Shijie Zhong's revised Advection_diffusion routine. | 19 January 2011, 21:00:20 UTC |
452a6d7 | Thorsten Becker | 29 December 2010, 20:47:43 UTC | Added support for non-interpolated grd input. | 29 December 2010, 20:47:43 UTC |
b6fe6e9 | Thorsten Becker | 22 December 2010, 21:53:36 UTC | Minor fixes for compile, still hunting for convergence problem. | 22 December 2010, 21:53:36 UTC |
dc5cd22 | Thorsten Becker | 08 December 2010, 21:06:36 UTC | Reverted to separate horizontal_bc routines for regional and full. the former has a check for ROW==E->lmesh.NOZ[level] && E->parallel.me_loc[3]==E->parallel.nprocz-1 while the full version has ROW==E->mesh.NOZ[level]) && (E->parallel.me_loc[3]==E->parallel.nprocz-1) ) | 08 December 2010, 21:06:36 UTC |
4ea70e2 | Thorsten Becker | 08 December 2010, 20:14:32 UTC | Reverted the rowl adjustment since it didn't fix the problem reported by Dan. | 08 December 2010, 20:14:32 UTC |
df821c0 | Thorsten Becker | 08 December 2010, 18:43:58 UTC | Added definition of TRUE (1) and FALSE (0) in case undefined (was defined for GMT/ggrd compile before). | 08 December 2010, 18:43:58 UTC |
cf7495e | Thorsten Becker | 08 December 2010, 00:54:48 UTC | Addded a test for top or bottom nodal row back into horizontal_bc (earlier moved from Regional_boundary_conditions and Full_boundary_conditions to BC_util.c) to see if this helps fix a bug reported by Dan that I could not reproduce in the non-pyre version. | 08 December 2010, 00:54:48 UTC |
9bc8fb6 | Thorsten Becker | 12 October 2010, 06:36:01 UTC | More accurate computation of element coordinates in anisotropic part. | 12 October 2010, 06:36:01 UTC |
010b460 | Thorsten Becker | 24 September 2010, 03:35:36 UTC | Augmented G matrix with anisotropic computation, may or may not be a good idea. | 24 September 2010, 03:35:36 UTC |
cbf9149 | Thorsten Becker | 15 September 2010, 10:40:50 UTC | Missed one routine. | 15 September 2010, 10:40:50 UTC |
1e59dd6 | Thorsten Becker | 15 September 2010, 10:36:58 UTC | Suggested use of general prototypes Retired N>3 matrix determinant functions (because they caused compile problems) | 15 September 2010, 10:36:58 UTC |
0c84e99 | Thorsten Becker | 14 September 2010, 01:08:34 UTC | Read in sdepv_misfit regardless of rheology. | 14 September 2010, 01:08:34 UTC |
bf59b6b | Thorsten Becker | 13 September 2010, 14:07:10 UTC | Added option to start with isotropic viscosity during first anisotropic iteration. | 13 September 2010, 14:07:10 UTC |
a112ea8 | Thorsten Becker | 13 September 2010, 12:01:43 UTC | Forgot part of the rotation matrix, minor fixes else, still experimental. | 13 September 2010, 12:01:43 UTC |
ab35881 | Thorsten Becker | 13 September 2010, 07:13:07 UTC | Made sure current version compiled nicely without additional flags. Anisotropic viscosity is set up more flexibly. | 13 September 2010, 07:13:07 UTC |
b5d6c30 | Thorsten Becker | 07 September 2010, 07:05:09 UTC | Made sure that viscosity prefactors get assigned for material dependent viscosity only (TDEPV = off) | 07 September 2010, 07:05:09 UTC |
a345eef | Thorsten Becker | 07 September 2010, 02:22:48 UTC | Added random number generator initialization to ensure exact reproducibility of all runs, including those with tracers. Earlier reported solver issues might be related to our cluster, at least all revisions seems to give the same answer now. | 07 September 2010, 02:22:48 UTC |
1cf026d | Thorsten Becker | 07 September 2010, 01:40:02 UTC | Checking in current version because of need for debugging. Revision 17172, compared to (presumably) 17153 give strange multigrid convergence behavior. Checking why that is now. | 07 September 2010, 01:40:02 UTC |
1efa65c | Thorsten Becker | 06 September 2010, 05:20:44 UTC | Trial implementation of anisotropic viscosity. | 06 September 2010, 05:20:44 UTC |
4ecc25c | Eh Tan | 30 August 2010, 20:41:40 UTC | Added support of multigrid solver in Exchanger. Lifted the solver assertion in Coupler. Cookbook9 can be converted to multigrid solver once appropriate parameters (*solver.mesher.levels=2 and *solver.vsolver.Solver=multigrid) are added. However, the embedded solver will converge incorrectly. Setting esolver.vsolver.accuracy=6e-3 seems to fix the problem. | 30 August 2010, 20:41:40 UTC |
bd28b8f | Thorsten Becker | 25 May 2010, 02:47:46 UTC | And a typo fix. | 25 May 2010, 02:47:46 UTC |