computeTau (X,Y,Z)
Interpolate vol_frac_CC to the face center in the same manner as
the scalarDiffusion routine.
This will change the answers
58 lines of code changed in 5 files:
added bulletproofing
- your can't have a periodic boundary condition on a face with extraCells>0
0 lines of code changed in 2 files:
updated heat conduction term in the multimaterial energy equation
0 lines of code changed in 4 files:
The face-centered diffusion coefficient is now calculated with:
d_c_L = (diff_coeff[L]*vol_frac_CC[L]);
d_c_R = (diff_coeff[R]*vol_frac_CC[R]);
diff_coeff_FC = (2.0 * d_c_L * d_c_R )/( d_c_L + d_c_R + SMALL_NUM);
This fixed a bug in diffusion when there are multiple materials.
30 lines of code changed in 13 files:
Fixes to allow VisIt plugin to be automatically built on the binary
side of the tree (if so specified to configure).
M configure
M configure.ac
- Add --with-visit=DIR flag.
- Add warning about glui.h
M configVars.mk.in
- Indent.
- Add BUILD_VISIT vars.
M Packages/Uintah/Core/DataArchive/DataArchive.h
- Indent. Use white space.
M Packages/Uintah/StandAlone/tools/sub.mk
- Build visit stuff if configured to do so.
M Packages/Uintah/StandAlone/tools/radiusMaker/radius_maker.cc
- Trying to get radius to come out right... this needs to be looked into more.
M Packages/Uintah/StandAlone/tools/uda2vis/sub.mk
Remove #if.
D Packages/Uintah/VisIt/udaReaderMTMD/avtudaReaderMTMDFileFormat.C
A Packages/Uintah/VisIt/udaReaderMTMD/testavtudaReaderMTMDFileFormat.C.in
- The avtudaReaderMTMDFileFormat.C has been replaced with the test*.in file which
allows configure to insert the path to the library directly into the file, thus
avoiding the user having to set an environment var.
M Packages/Uintah/VisIt/udaReaderMTMD/udaReaderMTMDCommonPluginInfo.C
- Use the full path #include: <Packages/Uintah/...>
M Packages/Uintah/VisIt/udaReaderMTMD/avtudaReaderMTMDFileFormat.h
- Fix #includes to use full path. Fix indentation.
A Packages/Uintah/VisIt/udaReaderMTMD/sub.mk
- Somewhat complicated sub.mk file to build visit library plugin. It
creates symbolic links on the binary side to the visit source files on
the src side of the tree (as the Makefile.visit that will be generated
by VisIt must have all the files in the same location). Currently it
runs the creation of the Makefile one time, but this probably needs to
be fixed and run as necessary. This commit is just to get a somewhat
working system in to the repository. I think the testavt...C.in file
is correctly updated to avt...C if the .in file changes, but need to
verify. This VisIt stuff should only happen when VisIt is configured
via configure.
A Packages/Uintah/VisIt/udaReaderMTMD/README
- Just a place to put notes.
M Packages/Uintah/sub.mk
- Build VisIt if configure'd.
9373 lines of code changed in 19 files:
fixed an array indexing bug
1 lines of code changed in 1 file:
Use the new name for the mpmarches pipe test, aka mpmpipe_test.ups.
3 lines of code changed in 1 file:
Add mpmarches tests.
21 lines of code changed in 1 file:
Rename the pipe test to mpmpipe_test.
0 lines of code changed in 2 files:
Clean ups for newer version of gcc. Perhaps this is a last checkin as
people move to Manta...
M rtrt/Core/NormalMapMaterial.cc
M rtrt/Core/Heightfield.cc
M rtrt/Core/BumpMaterial.cc
M rtrt/Core/Stealth.cc
M rtrt/Core/HTVolumeBrick.cc
M rtrt/Core/HVolumeBrickColor.cc
M rtrt/Core/DpyBase.cc
M rtrt/Core/ASETokens.h
M rtrt/Core/ObjReader.cc
M rtrt/Core/ASEReader.cc
M rtrt/Core/PathTracer/PathTraceEngine.cc
M rtrt/Core/HVolume.h
M rtrt/Core/RServer.cc
M rtrt/Core/ColorMapDpy.cc
M rtrt/Core/VideoMap.cc
M rtrt/Core/PNGImage.cc
M rtrt/StandAlone/scenes/min-museum.cc
M rtrt/StandAlone/scenes/3ds.cc
M rtrt/StandAlone/gl.cc
M rtrt/StandAlone/tex-utils/genpttex.cc
M rtrt/visinfo/visinfo.h
M rtrt/visinfo/visinfo.c
- Using #include <cstring/cstdlib/etc> to fix needs of newer compiler.
- Group #includes into related groups.
- Indent.
- Newer compiler wants "const char *", instead of "char *".
M rtrt/Core/bps.cc
Use {}.
M rtrt/Core/Grid.cc
- Use Thread::exitAll() (instead of exit) as this will help with rtrt
exiting fully (and not hanging.)
- Indent.
M rtrt/Core/HierarchicalGrid.cc
- Some indentation. Untabify.
M rtrt/Core/GridSpheres.cc
- Thread::exitAll()
- Better error messages.
M rtrt/StandAlone/scenes/t0.cc
- Parse arg as a 'string' as it is cleaner/easier.
0 lines of code changed in 54 files:
Add more Arches tests.
52 lines of code changed in 1 file:
Add Arches tests.
79 lines of code changed in 1 file:
Fix restarting so that it returns the proper return code for success or
failure.
11 lines of code changed in 1 file:
Change the location of the TestData.
1 lines of code changed in 1 file:
Save the timestep's time in the NRRD for future reference.
19 lines of code changed in 8 files:
Fix memory leak.
0 lines of code changed in 2 files:
Remove the TemperatureBoundCond include file.
0 lines of code changed in 1 file:
Remove the specific boundary condition classes.
0 lines of code changed in 14 files:
Use the new boundary condition interface with uses strings instead of
specific boundary condition classes.
31 lines of code changed in 6 files:
Generalize the boundary condition specification to use strings instead of
various classes. Eliminate the various boundary condition classes, i.e.
TemperatureBoundCond, VelocityBoundCond, etc.
146 lines of code changed in 20 files:
Volume data with/ without boundary
0 lines of code changed in 4 files:
Changes with the AMR support
161 lines of code changed in 2 files:
Modified LB times to narrow in on some inefficiencies.
21 lines of code changed in 1 file:
interpInitialConditionToStaggeredGrid()
- conversion to new patch.
- replaced triple loops with cell iterator
This commit should not change the answers
14 lines of code changed in 1 file:
ignore restart tests.
-restart tests fail because mmComputePropsFirst() is not called on restarts.
1 lines of code changed in 1 file:
Ignore the first 2 timesteps of profile data because it is not representative of future timesteps.
4 lines of code changed in 1 file:
Added an mpmArches case that has reaction/radiation (unlike the pipe test which doesn't have reaction/radiation).
2 lines of code changed in 1 file:
Fixed a bug in the fine level task profiling code.
1 lines of code changed in 1 file:
buildLinearMatrix()
-further cleanup.
BoundaryCondition.cc
-removed unused #define
This commit should not change the answers
20 lines of code changed in 3 files:
Fixed a bug in the new profiling code.
2 lines of code changed in 1 file:
removed mmspressureBC() -- not used.
removed commented out code.
1 lines of code changed in 3 files:
Final step in converting pressCoeff to CCVariable<Stencil7>
- rewrote pressureBC
- removed fortran code.
This commit should not change the answers
93 lines of code changed in 7 files:
Tasks that require data from a fine level are now profiled sperately from normal tasks. This will hopefully allow the profiling to be more accurate.
321 lines of code changed in 18 files:
get method for the profile type.
1 lines of code changed in 1 file:
Added the profile type to DetailedTasks. The profile type can be either Normal or Fine.
Normal means that that the task executes proportional to the number of cells/particles in the patch.
Fine means that the task executes proportional to the number of cells/particles on the fine level below it (for example coarsening).
10 lines of code changed in 2 files:
Pass the task into the profiler instead of the patches that the task ran on. This will be used to profile the coarse interfaces seperately.
8 lines of code changed in 5 files:
Beginning of new advanced profiling methods.
Seperated the profiling driver from the profiling interface so that it can be reused easily.
421 lines of code changed in 6 files:
Added a SCI_DEBUG called MPIStats. This debug will printout the number of messages that the task graph sends and the total message volume.
51 lines of code changed in 6 files:
Removed unused code.
24 lines of code changed in 2 files:
Add a quick debug build to check for compile errors.
Add a fully optimized build that has the following tests turned on:
ice, mpm, mpmice, impm
Add a nightly debug and optimized build that has the following tests turned on:
ice, mpm, mpmice, impm
211 lines of code changed in 1 file:
Report the maximum time taken with LBTimes in addition the the average.
12 lines of code changed in 1 file:
Added a sci debug stream called LBTimes.
51 lines of code changed in 1 file:
Modifications to the barrier debug stream.
41 lines of code changed in 1 file:
Reworked the dependency combining code. I'm unsure the current code will work for all cases. I have now changed the dependency combing code to hopefully always consistently combine and be quicker.
44 lines of code changed in 2 files:
pruned some unused code.
0 lines of code changed in 1 file:
Turn on heating due to dissipation resulting from artificial
viscosity. This will change the results for those tests that
use these models.
3 lines of code changed in 3 files:
Fixed a bug where sendstates and dependencies could become inconsistent causing all kinds of bad and semi random things to happen
27 lines of code changed in 1 file:
Rolling back some changes until I can confirm that the differences from the current gold standard are fuzz.
This should fix the failing Arches and MPMArches tests.
39 lines of code changed in 1 file:
Exit out a little bit later in the function.
9 lines of code changed in 1 file:
If the same dependency already exists don't delete it and readd it. Instead just keep the old dependency.
9 lines of code changed in 1 file:
Remove test_corrug_plate test temporarily.
2 lines of code changed in 1 file:
Create new function run_case that eliminates the need for all of the
*_test functions.
Track when a set of tests actually fails and exit with a non-zero quantity.
Remove poisson2 from the example set of tests.
Always run the model tests with MPI.
67 lines of code changed in 1 file:
Undid part of a patch from earlier this week that is likely to affect performance.
14 lines of code changed in 1 file:
Remove -nsigs it does not work with openmpi
0 lines of code changed in 2 files:
Fixed a bug in the cost profiler that would cause 1 processor to be assigned all of the profiling work.
1 lines of code changed in 1 file:
Don't double allocate the boundary conditions.
0 lines of code changed in 2 files:
Instantiate boundary conditions only if the processor owns the patch.
6 lines of code changed in 5 files:
Missed a header file in commit 42202
4 lines of code changed in 1 file:
Fixed a small bug in the profiler related to new regions.
34 lines of code changed in 1 file:
Simplified the profiling mechinism when dealing with initialization.
43 lines of code changed in 1 file:
Turned off exception that is thrown if an even number of cells isn't on a patch.
0 lines of code changed in 2 files:
Add environment variables for open mpi so we do not have to use the -mpi option
for sus.
3 lines of code changed in 1 file:
Major change:
- added struct that contains upper/lower loop limits for the CC and FC variables
- face centered variables are now stored on the x- faces instead of the x+
- fixed multiple problems with the advection operator that showed up in both
symmetric converging and diverging velocity fields.
To do: Something is not quite right with the 2nd order version of the
advection operator. The shock tube results look like smooth first order results.
0 lines of code changed in 16 files:
Don't grab ghost particles when counting particles on your own patch.
0 lines of code changed in 2 files:
Fixed multi-patch bug.
115 lines of code changed in 3 files:
Minor error in debugging output
0 lines of code changed in 2 files:
Debugging code
8 lines of code changed in 1 file:
Beginings of an overhaul on the scalar equation.
* Replaced old "StencilMatrix" type matrix coefficients with CCVariable<Stencil7>
* Removed some fortran calls and replaced them with their c++ equivalent
* Added an explicit integrator as a template
* Removed some redundancy and extra loops
Note: I left in all the old code but just commented it out for now.
624 lines of code changed in 18 files:
Fixed a fairly major bug in the load balancer. This bug caused incorrect accounting of the mean amount of work per unassigned processors and would cause high rank processors to recieve no work. This lead to large load imbalances. This was primarily noticiable on large processor runs.
10 lines of code changed in 1 file: