October 2010 Commit Log

Number of Commits:
104
Number of Active Developers:
12
tsaad 2010-10-30 16:21 Rev.: 46413

Major cleanup for UpwindInterpolant.
- Moved all counter arithmetic to constructor and added required private variables.
- In the process, removed some of un-necessary member variables (dim_, hasPlusFace_).

57 lines of code changed in 1 file:

harman 2010-10-29 16:39 Rev.: 46412

cleaned out so silly make clean commands

added command/macro to run the local regression tests

make runLocalRT

It's current hardwired to run ICE local tests.

12 lines of code changed in 1 file:

harman 2010-10-29 16:00 Rev.: 46411

runSusTests.py
- the "mode" can be either dbg, opt or local.

startTester
- alphabetized environmental vars
- now exporting LOCAL_OR_NIGHTLY_TEST

Remaining files.
- formatting
- added local tests

318 lines of code changed in 14 files:

harman 2010-10-29 14:31 Rev.: 46410

initial commit of the local regression tester. It does 80% of what I want it to do.

139 lines of code changed in 1 file:

harman 2010-10-29 09:40 Rev.: 46409

hard coded the name of the summary log.

1 lines of code changed in 1 file:

dav 2010-10-28 19:22 Rev.: 46408

Apparently some people use NDEBUG for optimized builds... who knew? :)

10 lines of code changed in 2 files:

guilkey 2010-10-28 18:56 Rev.: 46407

Remove the hardwiring of q in the cohesive zones. The effect is subtle, but
this keeps it more general according to the equations.

0 lines of code changed in 2 files:

dav 2010-10-28 17:35 Rev.: 46406

Per James, turn off tests for exprlib and spatialops wasatch libs (speeds up build time).

2 lines of code changed in 1 file:

jsutherland 2010-10-28 15:32 Rev.: 46405

quite compiler warning on Linux

3 lines of code changed in 1 file:

harman 2010-10-28 14:22 Rev.: 46404

create the file summary_log before you cat to it.

1 lines of code changed in 1 file:

dav 2010-10-28 14:08 Rev.: 46403

Added define of NDEBUG if in debug mode, even if user/configurer gives other flags too.

2 lines of code changed in 2 files:

dav 2010-10-28 14:05 Rev.: 46402

Added define of NDEBUG if in debug mode

2 lines of code changed in 2 files:

harman 2010-10-28 13:53 Rev.: 46401

build_and_run
- if the code didn't compile OR if "NOTESTS" has been choosen then exit script.
- eliminated 2 extra conditional statements.
- moved the loop over all componentTests to this script.

run
- removed loop over all componentTests. This script is only designed to run
a single component tests.
- added bulletproofing on required environmental variables.

119 lines of code changed in 2 files:

jsutherland 2010-10-28 12:42 Rev.: 46400

remove some comments

1 lines of code changed in 1 file:

jsutherland 2010-10-28 12:41 Rev.: 46399

First hack at getting memory window support from SpatialOps to propagate into Wasatch.
Still need to get the configure system up to date to correctly deal with requirements from SpatialOps

67 lines of code changed in 8 files:

harman 2010-10-28 09:18 Rev.: 46398

find all the component tests, which are python scripts,
and loop over thoses tests.

6 lines of code changed in 1 file:

guilkey 2010-10-27 21:01 Rev.: 46396

Pulled the code that does rotation into the CZMaterial code as a function.

76 lines of code changed in 4 files:

guilkey 2010-10-27 20:33 Rev.: 46395

Make execution of the code to do rotations of the CZ directions optional.
Default to false for now.

15 lines of code changed in 3 files:

guilkey 2010-10-27 19:50 Rev.: 46394

Remove tabs, improve formatting, improve scoping.

71 lines of code changed in 1 file:

harman 2010-10-27 17:16 Rev.: 46393

moved the creation of the symbolic links and testing for the GS from run to
build_and_run.

run:
changed test -> componentTest

75 lines of code changed in 2 files:

guilkey 2010-10-27 15:11 Rev.: 46392

Use M_PI in place of various approximations of Pi.

2 lines of code changed in 1 file:

jpeterson 2010-10-27 15:02 Rev.: 46391

CPDI is now split in two components. The Fast component has an issue when particle corners span more than one cell (have over an entire cell between them). Upon detecting that scenario an error is printed and the program exits without a segfault.

846 lines of code changed in 8 files:

harman 2010-10-27 12:39 Rev.: 46390

removed ability to run weekly RTs

11 lines of code changed in 7 files:

harman 2010-10-27 12:23 Rev.: 46389

removed the ability for individual users to have their own RTs.

30 lines of code changed in 4 files:

harman 2010-10-26 13:19 Rev.: 46388

removed inferno pbs script

0 lines of code changed in 1 file:

harman 2010-10-26 13:17 Rev.: 46387

removed Steve's todo list

0 lines of code changed in 1 file:

guilkey 2010-10-22 13:02 Rev.: 46386

Change the way the rigid reaction force is computed. The new way gives
smoother results. This change was suggested and implemented by Jonah Lee
of U Alaska Fairbanks.

0 lines of code changed in 2 files:

qymeng 2010-10-21 19:58 Rev.: 46385

Fixed a problem when doing AMR particle subset gather

0 lines of code changed in 2 files:

guilkey 2010-10-21 17:53 Rev.: 46384

Adding an axisymmetric version of the cpdi interpolator. Housecleaning on the
original cpdiInterpolator.

202 lines of code changed in 6 files:

harman 2010-10-21 17:01 Rev.: 46383

removed Jim's version of
- interpolateParticlesToGrid()
- computeZoneOfInfluence()

0 lines of code changed in 1 file:

harman 2010-10-21 16:57 Rev.: 46382

Global:
- cleaned up doMPMOnLevel conditional statements in all the scheduling tasks.


scheduleComputeZoneOfInfluence()
- only compute zone of influence on the fine levels

scheduleCoarsenNodalData_CFI()
- removed weeds

interpolateParticlesToGrid()
- added section of debugging code.

InterpolateParticlesToGrid_CFI()
- Size of S and ni vectors are now dynamic.

-nBoundaryCells is now set to max element of refinement ratio

-added section of debugging code.

CoarsenNodalData_CFI()
-added debugging code.


computeZoneOfInfluence()
Added code to set the zoi to the coarse level dx. Currently not used while I figure out what the right
thing to do is.

177 lines of code changed in 1 file:

guilkey 2010-10-21 16:11 Rev.: 46381

Reverting cpdi back to pre-Joseph so that we can use it until he's able to fix
the bugs in his version.

239 lines of code changed in 2 files:

harman 2010-10-21 15:03 Rev.: 46380

Completed more descriptive variable name change. These changes should have been
in the last commit.

2 lines of code changed in 1 file:

harman 2010-10-21 11:19 Rev.: 46379

-Improved comments.
-more descriptive variables names ni -> CFI_ni, lo -> finePatch_lo ...etc
-Changed test for detecting if a node is on a CFI. Now using a inside a box test.
-The number of elements in a shape function is variable. Using push_back.
-Added gobs of debugging output code, currently turned off

66 lines of code changed in 1 file:

qymeng 2010-10-20 20:00 Rev.: 46378

fixed a problem when copying amr psets.
added some missing read locks.

10 lines of code changed in 1 file:

luitjens 2010-10-20 14:50 Rev.: 46377

Potentially fixed the checkGetAccess code when requesting psets from different levels.

48 lines of code changed in 5 files:

harman 2010-10-20 14:37 Rev.: 46376

not used

0 lines of code changed in 2 files:

harman 2010-10-19 15:06 Rev.: 46375

cosmetics changes. Formatted some of the function parameter lists.

19 lines of code changed in 1 file:

luitjens 2010-10-19 13:48 Rev.: 46374

Added the parsing for reduce_uda.

3 lines of code changed in 1 file:

harman 2010-10-18 12:44 Rev.: 46372

MPMLabel.cc .h
- added gSumWeightsLabel : Sum over all particles of the interpolation weights.

AMRMPM.cc .h
- added new task coarsenNodalData_CFI()
- copy the fine level nodal data to the underlying coarse nodes only at
the coarse fine interface.

Works for a simple 3L, RR:2, 1D problem with linear interpolation.

159 lines of code changed in 4 files:

jthornoc 2010-10-15 16:01 Rev.: 46370

Fixed a bug where iv's weren't always being matched properly with
TabProp's ordering of ivs.

19 lines of code changed in 1 file:

harman 2010-10-15 08:51 Rev.: 46369

AMR.cc
getCoarseLevelRangeNodes()
- added nBoundaryCells to the computeVariableExtents argument list.
- using mapCellToCoarser() instead of mapNodeToCoarser()
- this commit has debugging couts

AMRMPM
Global: removed "\t" from printTask and printSchedule calls
For debugging pColor is initialized to the level the particle is on.

InterpolateParticlesToGrid_CFI()
- using zoi on the fine level instead of the coarse level
- using AMR version of findCellAndWeights
- The size of the node index vector is dynamic.
- lots of debugging cout statements.

ComputeZoneOfInfluence()
- At the CFI faces (fine level) the stencil element modified is the same
as the CFI face. Previously I had set it to be the opposite element.

In this commit InterpolateParticlesToGrid_CFI is doing the right thing for a simple
3L, RR: 2, 1-D problem. More testing is needed!

100 lines of code changed in 3 files:

harman 2010-10-14 17:06 Rev.: 46368

findCellAndWeight AMRMPM Version
- complete change in the way the nodes are found. Now returning only the
nodes along the coarse fine interface.

60 lines of code changed in 1 file:

guilkey 2010-10-14 15:21 Rev.: 46367

Add code for 3D cohesive zones written by Ravi Bhide of EGI.
Fix compiler warning in NodalSVFContact.


92 lines of code changed in 3 files:

harman 2010-10-14 14:33 Rev.: 46366

added comment to mapNodeToFiner

0 lines of code changed in 2 files:

harman 2010-10-14 13:21 Rev.: 46365

All: added comment

# Note: the "folder name" must be the same as input file without the extension.

Wasatch:
changed folder name from BasicScalarTransport -> BasicScalarTransportEquation.

37 lines of code changed in 12 files:

harman 2010-10-14 10:55 Rev.: 46364

mapNodeToCoarser()
- removed warning message and added ascii art to show what the
function returns.

10 lines of code changed in 1 file:

harman 2010-10-14 09:15 Rev.: 46363

mapNodeToCoarser()
- added warning to not use this function. It is being used by examples/poisson3.cc

4 lines of code changed in 1 file:

jas 2010-10-13 16:16 Rev.: 46362

Rename diamm.f to diamm.F so things will build on Debian Squeeze.

3506 lines of code changed in 4 files:

jas 2010-10-12 17:47 Rev.: 46359

Change the name of the Wasatch directory.

0 lines of code changed in 2 files:

jsutherland 2010-10-12 15:37 Rev.: 46358

Add the ability to shove source terms on transport equations. This will help with MMS in particular.

8 lines of code changed in 1 file:

luitjens 2010-10-12 13:25 Rev.: 46357

Don't delete the taskgraph doc. For some reason this causes some memory corruption. Since this is used for testing only i'm just going to let this leak.

3 lines of code changed in 1 file:

harman 2010-10-12 13:11 Rev.: 46356

more descriptive ostream message.

2 lines of code changed in 1 file:

luitjens 2010-10-12 13:04 Rev.: 46355

Also emit regular tasks. For some reason only reduction tasks were being emitted.

0 lines of code changed in 2 files:

luitjens 2010-10-12 13:02 Rev.: 46354

Removed some more dust from the task graph emitting code.

2 lines of code changed in 1 file:

harman 2010-10-12 12:33 Rev.: 46353

Initial commit of:

findCellAndWeights()

which is used by AMRMPM

105 lines of code changed in 4 files:

jas 2010-10-12 12:05 Rev.: 46352

Added wasatch test infrastructure.

14 lines of code changed in 1 file:

luitjens 2010-10-12 11:52 Rev.: 46351

Dusted off some of the task graph output code. -emit_taskgraph still does not work because the emit_taskgraph variable is not being set to true for some reason. Though you can set it manually to true in the code to get the xml taskgraph output.

1 lines of code changed in 1 file:

kunkel 2010-10-11 16:23 Rev.: 46350

Fixed the section that sets material indices alpha and beta.

5 lines of code changed in 3 files:

jpedel 2010-10-11 15:26 Rev.: 46349

Convection term was set to zero by mistake

0 lines of code changed in 2 files:

harman 2010-10-11 13:55 Rev.: 46348

getCoarseFineFaceRange()
variable name change
interpolationOrder -> nCells
more general.

6 lines of code changed in 2 files:

harman 2010-10-11 13:40 Rev.: 46347

getCoarseFineFaceRange()
Made the function more general. The faceIteratorType is now passed in
to support node & SFC(x,y,z) based variables.

25 lines of code changed in 7 files:

harman 2010-10-11 12:13 Rev.: 46346

removed \t from all printTask and printSchedule. Formatting is handled
by the functions.

153 lines of code changed in 9 files:

guilkey 2010-10-08 16:29 Rev.: 46345

Results are improving. Getting something that looks like a detonation, but the C-J pressure isn't right yet.

Added the delta_L term (based on original particle volume).
Added pressure stabilization (this helps reduce the noise in the pressure a lot)
Added a pressure mixer based on the progress variable and the pressure from the
reacted and unreacted EOS's.

80 lines of code changed in 1 file:

tsaad 2010-10-08 16:25 Rev.: 46344

Fixed typo in UpwindInterplant.h - Now convective flux work in all 3 directions.

4 lines of code changed in 1 file:

qymeng 2010-10-08 16:00 Rev.: 46342

Check if task is accesssing a right level when Task::CoarseLevel or Task::FineLevel is specified.

4 lines of code changed in 1 file:

harman 2010-10-08 15:21 Rev.: 46341

added interpolateParticlesToGrid_CFI tasks

11 lines of code changed in 1 file:

harman 2010-10-08 15:11 Rev.: 46340

actuallyInitialize()
- initialize the particle color by level index.


problemSetup()
- fixed typo dataArchive->dataArchiver

scheduleTimeAdvance()
- added scheduleInterpolateParticlesToGrid_CFI

Added:
scheduleInterpolateParticlesToGrid_CFI and interpolateParticlesToGrid_CFI
Still thrashing and testing these tasks.


180 lines of code changed in 1 file:

guilkey 2010-10-08 15:10 Rev.: 46339

Add pressure stabilization. Makes riemannMPM results much worse, probably since
it is a two material problem.

0 lines of code changed in 2 files:

harman 2010-10-08 14:19 Rev.: 46338

tabs

1 lines of code changed in 1 file:

harman 2010-10-08 14:16 Rev.: 46337

turned off memory tests

1 lines of code changed in 1 file:

qymeng 2010-10-08 12:27 Rev.: 46336

Fixed a bug in isInDepMap() when patches_dom is Task::CoarseLevel or Task::FineLevel
(assuming it may access a patch in any level)

3 lines of code changed in 1 file:

harman 2010-10-08 11:43 Rev.: 46335

-turned off some input argument output.
-Added flags to detect if replacement was successful
-Only output to the log if something was replaced if the flags indicate so.
-Added do_uda = true to the replace uda conditional statement.

42 lines of code changed in 1 file:

harman 2010-10-08 10:35 Rev.: 46334

formatted the function argument lists so it's easier to find which version of the
function you're using

ostream << operator:
-If the patches_domain has been specified output that value.

isInDepMap()
-added {} to conditionals

226 lines of code changed in 1 file:

tsaad 2010-10-07 22:08 Rev.: 46333

Fixed bug in assigning proper directions for calculating the coordinates

8 lines of code changed in 1 file:

harman 2010-10-07 16:47 Rev.: 46332

removed prototype without associated code in the cc file.

2 lines of code changed in 1 file:

jsutherland 2010-10-07 15:17 Rev.: 46331

Quiet a few more warnings on Linux.

11 lines of code changed in 1 file:

jsutherland 2010-10-07 14:46 Rev.: 46330

Use inheritance for the upwind convective flux expression builder.

17 lines of code changed in 1 file:

jsutherland 2010-10-07 13:54 Rev.: 46329

Move some things out of the ParseEquation.h header and into the implementation file.

35 lines of code changed in 2 files:

jsutherland 2010-10-07 13:20 Rev.: 46328

quiet some warnings on Linux

4 lines of code changed in 1 file:

jsutherland 2010-10-07 13:08 Rev.: 46327

more git pains - fixing a problem from my last commit

0 lines of code changed in 1 file:

jsutherland 2010-10-07 12:58 Rev.: 46326

Fix some memory leaks in Wasatch.

9 lines of code changed in 3 files:

jsutherland 2010-10-07 11:03 Rev.: 46325

Fix some memory leaks in Wasatch.

41 lines of code changed in 4 files:

harman 2010-10-07 10:39 Rev.: 46324

added some formatting of particle variable section

92 lines of code changed in 1 file:

jsutherland 2010-10-07 10:36 Rev.: 46323

- fix some more things that broke in my recent failed merge.

68 lines of code changed in 2 files:

jsutherland 2010-10-07 10:14 Rev.: 46322

Fix a mistake in an earlier commit that resulted from a bad merge

122 lines of code changed in 1 file:

jsutherland 2010-10-07 08:31 Rev.: 46321

Merge branch 'cleanup'

112 lines of code changed in 3 files:

harman 2010-10-06 17:04 Rev.: 46319

changed name of sci_debug flags.

TaskGraph0 -> TaskGraph
TaskGraph -> TaskGraphDetailed.

4 lines of code changed in 3 files:

tsaad 2010-10-06 15:23 Rev.: 46318

In this commit:
1. Debugged UpwindInterpolant.
2. Added support for Central differencing for convection.
3. Added convection-test.ups file. This is NOT a comprehensive test yet.
4. Modified ConvectiveFlux.h by adding inhertience for upwind type interpolants.

717 lines of code changed in 4 files:

tsaad 2010-10-06 12:58 Rev.: 46317

Fixed error in timestep calculation. Must use old data warehouse in the task requirement.

1 lines of code changed in 1 file:

harman 2010-10-06 11:47 Rev.: 46316

added level index to task ostream

Cleanup formatting for ostream for computes/requires, see below:

0 requires: [p.temperature , No patches, MI: none, OldDW (mapped to dw index 0), Ghost::AroundNodes:1]
0 requires: [p.size , No patches, MI: none, OldDW (mapped to dw index 0), Ghost::AroundNodes:1]
0 requires: [p.deformationMeasure, No patches, MI: none, OldDW (mapped to dw index 0), Ghost::AroundNodes:1]
0 computes: [g.mass , No patches, MI: none, NewDW (mapped to dw index 1), Ghost::None]
0 computes: [g.volume , No patches, MI: none, NewDW (mapped to dw index 1), Ghost::None]
0 computes: [g.velocity , No patches, MI: none, NewDW (mapped to dw index 1), Ghost::None]


7 lines of code changed in 1 file:

harman 2010-10-06 10:12 Rev.: 46315

added level information to exception message

0 lines of code changed in 2 files:

guilkey 2010-10-05 17:48 Rev.: 46314

Starting to look like a detonation, at least in 1-D.

19 lines of code changed in 3 files:

jsutherland 2010-10-05 16:41 Rev.: 46313

Fix the timestepper - previously we were getting deltat from the wrong data warehouse (go figure).

2 lines of code changed in 1 file:

dav 2010-10-05 15:30 Rev.: 46312

fix shell script typo

2 lines of code changed in 2 files:

harman 2010-10-04 17:59 Rev.: 46311

Now using Dav's build script to compile Wasatch 3P.

1 lines of code changed in 1 file:

dav 2010-10-04 16:49 Rev.: 46310

M configure
M configure.ac

- Added --enable-wasatch_3p which will cause configure to checkout,
build, and use the SpatialOps, ExprLib, and TabProps libs. (They will
be placed in <bin>/Wasatch3P/.)

If you already have the Wasatch 3p built somewhere else, then just use
the --with-spatialops, --with-tabprops, and --with-exprlib flags to point
at your version.

- Found a problem with our static build... need to add -static to the link
line. Placed a comment in configure.ac, but have not updated this yet
as it will break things until all libraries on our test machines are updated
to include .a files.

AM build_scripts/build_wasatch_3p.sh

- Script to checkout (git) the Wasatch 3p libs and build them.

243 lines of code changed in 3 files:

guilkey 2010-10-04 11:38 Rev.: 46309

Make the readability slightly less obnoxious.

11 lines of code changed in 1 file:

harman 2010-10-04 09:11 Rev.: 46308

turned off restart tests.

1 lines of code changed in 1 file:

harman 2010-10-01 14:37 Rev.: 46302

latest settings

0 lines of code changed in 2 files:

harman 2010-10-01 12:39 Rev.: 46296

removed extra "

1 lines of code changed in 1 file:

luitjens 2010-10-01 11:37 Rev.: 46295

Disabled allgather hack by default.

1 lines of code changed in 1 file:

harman 2010-10-01 11:17 Rev.: 46294

Added gui interface to replace_gold_standard

Added a master replace_GS script so all the GS can be replaced via one
script in either opt or dbg.

290 lines of code changed in 4 files:

harman 2010-10-01 10:46 Rev.: 46292

added petsc and hypre to the configure line. I'm not sure why this would change the
answers in ARCHES.

2 lines of code changed in 1 file:

September 2010 »

Generated by StatSVN 0.4.0