file:///svnroot/FUN3D

r43327

CCI
by thomasj at Fri Nov 06 22:30:23 EST 2009

r43326

To complete FUN3D stable release: 10.8-43309.
by kleb at Fri Nov 06 22:14:53 EST 2009

r43325

To tag FUN3D stable release: 10.8-43309.
by kleb at Fri Nov 06 22:14:36 EST 2009

r43324

To simplify triage
by thomasj at Fri Nov 06 21:31:42 EST 2009

r43323

To use i16n269 again now that it has a new power supply.
by kleb at Fri Nov 06 16:47:57 EST 2009

r43322

Initial implementation of mapping for node-based least squares.  Only
catch is it seems to cause Hiro's node-centered agglomeration to give
a different answer (but not die!), which one would think completely
unrelated.  I narrowed it down to the simple declaration, nullify, and
allocation of grid%nlsq, but do not see any direct signs of incorrect
code.  Hiro is looking into it and reports that he has seen other cases
prior to this where the coarsest grid agglomeration shows memory issues.
Redefining those golden files for now so that we can move on.  But there
seems to clearly be a memory problem with the agglomeration procedure.
by nielsen at Fri Nov 06 12:45:24 EST 2009

r43321

To complete FUN3D stable release: 10.8-43305.
by kleb at Fri Nov 06 09:29:54 EST 2009

r43320

To tag FUN3D stable release: 10.8-43305.
by kleb at Fri Nov 06 09:29:50 EST 2009

r43319

Correct minor heading value
by dana at Fri Nov 06 09:03:04 EST 2009

r43318

to fix build SIGSEGV with write-damping by adding old-timey write debug statements, valgrind still reports a cryptic 'Warning: client syscall munmap tried to modify addresses ...' but google indicates it may be openmpi bug?
by mikepark at Fri Nov 06 08:40:26 EST 2009

r43317

to make error est grid embedding valgrind-clean, only use nedgeloc edges because rad uses a level 1 grid now
by mikepark at Fri Nov 06 08:10:22 EST 2009

r43316

To add residual averaging - convergence drops from 0.17 to 0.09 per cycle
by thomasj at Fri Nov 06 06:02:40 EST 2009

r43315

To define format
by thomasj at Fri Nov 06 05:49:08 EST 2009

r43314

To make variable name treatments consistent, employ RedCloth markup where able, and split long lines into phrases to better enable future diffs.
by kleb at Thu Nov 05 22:55:43 EST 2009

r43313

To complete FUN3D stable release: 10.8-43295.
by kleb at Thu Nov 05 20:49:10 EST 2009

r43312

To tag FUN3D stable release: 10.8-43295.
by kleb at Thu Nov 05 20:49:07 EST 2009

r43311

To describe implicit line format
by thomasj at Thu Nov 05 17:58:56 EST 2009

r43309

To create a cc residiual averaging
by thomasj at Thu Nov 05 17:19:54 EST 2009

r43305

To tie http://abi3.larc.nasa.gov/projects/PHYSICS_MODULES/changeset/2192 to FUN3D.
by kleb at Thu Nov 05 13:46:46 EST 2009

r43302

Technical debt
by thomasj at Thu Nov 05 09:05:06 EST 2009

r43301

Avoiding division and stopping for MPI
by thomasj at Thu Nov 05 09:03:37 EST 2009

r43300

The out part of r43299
by thomasj at Thu Nov 05 08:51:13 EST 2009

r43299

Moving data about follow-on code (adjoint in this case) out and up
by thomasj at Thu Nov 05 08:44:20 EST 2009

r43298

To complete FUN3D stable release: 10.8-43291.
by kleb at Thu Nov 05 08:06:44 EST 2009

r43297

To tag FUN3D stable release: 10.8-43291.
by kleb at Thu Nov 05 08:06:40 EST 2009

r43296

To reduce line noise, be explicit, and update with "Grids/" directory correction.
by kleb at Thu Nov 05 07:06:09 EST 2009

r43295

To seal off from multigrid
by thomasj at Wed Nov 04 18:18:12 EST 2009

r43294

To add se_exit upon stop
by thomasj at Wed Nov 04 18:05:23 EST 2009

r43293

Fuller descrtiption
by thomasj at Wed Nov 04 17:46:21 EST 2009

r43292

To actually repair at least the partitioning ofthe SpanwiseCylinder/Generic/Laminar/OneCell testcase, put .lines_fmt in Grids directory, not the test case directory.
by kleb at Wed Nov 04 15:28:17 EST 2009

r43291

To repair SpanwiseCylinder/Generic/Laminar/OneCell case thanks to Thomas.
by kleb at Wed Nov 04 14:55:33 EST 2009

r43290

First attempt at converting from ginput.faces + ginput_translator to namelist. Known failures: three generic gas cases: OneCell due to --partition_lines and two turbulent cases (ivisc=2 not allowed).
by kleb at Wed Nov 04 07:10:45 EST 2009

r43289

To avoid dead node i16n269, load up i16n271.
by kleb at Tue Nov 03 23:43:18 EST 2009

r43288

Comment
by thomasj at Tue Nov 03 18:19:16 EST 2009

r43287

To account more fully for 2D
by thomasj at Tue Nov 03 17:54:58 EST 2009

r43286

Slight refactoring to eliminate divisions
by thomasj at Tue Nov 03 17:39:53 EST 2009

r43285

Indentation
by thomasj at Tue Nov 03 17:32:10 EST 2009

r43284

To correct an incorrect conditional
by thomasj at Tue Nov 03 17:07:23 EST 2009

r43283

Rename file extension now that convention has changed
by nielsen at Tue Nov 03 16:29:59 EST 2009

r43282

To make the move over to suggar++
by biedron at Tue Nov 03 15:55:19 EST 2009