More work on refactoring the RMCRT class.
- moved common methods,functions and variables from Ray.cc -> RMCRTCommon
*This should not change the answers
727 lines of code changed in 8 files:
Remove obsolete schedulers.
NOTE: This is going to break tests, due the SVN diffs themselves containing the very line /R_Tester/helpers/runSusTests.py looks for to bail: "Caught exception" (see below).
# was an exception thrown
exception = system("grep -q 'Caught exception' sus.log.txt");
Changing this string results in the script catching its own SVN diffs during RT.
Todd - How do we want to handle this?
26 lines of code changed in 11 files:
fix warches: part 2.
225 lines of code changed in 16 files:
fix warches
4 lines of code changed in 2 files:
updated variable name change abskg -> abskg_BC
turn on rmcrt_bm1_ML
2 lines of code changed in 1 file:
fix the buildbot
4 lines of code changed in 3 files:
Add particle transport to Wasatch. Currently supporting:
1. Particle position
2. Particle momentum (drag, buoyancy, weight)
3. Particle mass (no sources yet)
4. Particle size (no sources yet)
This commit provides a working and easy to use framework for adding particle transport equations and related physics.
3640 lines of code changed in 40 files:
remove a cout statement from the relocation algorithm
0 lines of code changed in 1 file:
Add particle relocation version that doesn't require pre-relocation variables. Please use this with caution as it requires additional testing. A warning has be provided in the documentation of the associated functions to remind users of that caution.
572 lines of code changed in 6 files:
change the inheritance structure of transport equations. Now derive from an EquationBase class in anticipation of adding particle transport equations.
233 lines of code changed in 6 files:
Initial commit for extracting the radiometer code from RMCRT so it can called
in conjunction with the DO radiation model.
Heavy construction zone, phase 0.0.1, ridiculous amounts of code duplication.
1075 lines of code changed in 3 files:
rayTrace()
- clean out IO code.
0 lines of code changed in 2 files:
Clean up more warnings.
18 lines of code changed in 4 files:
Clean up a few warnings.
14 lines of code changed in 2 files:
updated gnuplot plotting scripts and added a few more tests.
0 lines of code changed in 12 files:
added mutex lock to << operator. Striving for warning free thread sanitizer output.
3 lines of code changed in 1 file:
rip out an obsolete flag for TabProps
0 lines of code changed in 1 file:
fixed previous sloppy commit.
6 lines of code changed in 1 file:
fix pow() call for floats.
2 lines of code changed in 1 file:
accidently turned on float:RMCRT in last commit. Turned it off.
1 lines of code changed in 1 file:
- converted abskg and color from CCVariable<double> to CCVariable<float>.
- added a HACK around the setBC calls. The BC infrastructure
wasn't designed to work with floats and it will need a face lift
if we decide to continue down this path. For now
copy floats -> doubles
setBC (doubles)
copy doubles -> floats
The float version runs.
170 lines of code changed in 4 files:
Added threaded RMCRT tests back into the regression testing. This was necessary due to an unintentional removal on the previous commit.
4 lines of code changed in 1 file:
Changed how Arches.cc interfaces with radproperty_Calculator.
Changed default name of abskp to abskp_HS.
Removed redundant and unneeded line in DORadiation.cc.
9 lines of code changed in 4 files:
Added float instantiations of coarsenDriver(), fineToCoarseOperator(), compute_mag_gradient()
Cleaned up the alignment of parameter lists.
0 lines of code changed in 6 files:
Commit on behalf of Abhishek for further GPU related issues in the Wasatch-ExprLib interface.
14 lines of code changed in 1 file:
Comment out all SCI_DEBUG.setActive() calls. This
will reduce the amount of output then using the thread sanitizer.
10 lines of code changed in 5 files:
Calculate weights and abscissas for moments durign problem setup. This will break the regression tester for CQMOM tests.
1 lines of code changed in 1 file:
Add in the CQMOM specific convection term for when particle velocity is an internal coordinate of the system.
922 lines of code changed in 7 files:
Add a CQMOM specific boundary condition, add debug couts, changes to xml
18 lines of code changed in 3 files:
#if 0 out all cuda code. GPU code should compile.
0 lines of code changed in 2 files:
Added a "sandbox" for testing using floats instead of doubles in RMCRT. This is just a copy
of RMCRT with no changes.
4377 lines of code changed in 9 files:
formatting.
Added new restart test, regridTestArches. Restart a single patch methane uda and regrid once. The new patch layout has 8 patches.
63 lines of code changed in 1 file:
Use a faster nebo assignment for boundary conditions.
1 lines of code changed in 1 file:
* Add GPUPerPatch support.
* Some cleanup and formatting in GPUDataWarehouse and other GPU Variable source files.
592 lines of code changed in 26 files:
Of course, if you make changes to configure.ac or aclocal.m4, you have
to generate a new configure script to be checked in...
M configure
- Add the -O2 to the nvcc compilation to avoid the missing symbols problem
that occurs with dlink of a debug build.
9 lines of code changed in 1 file:
This should fix the problem with the Uintah code in a CUDA enabled
build not linking if it is in debug mode. This is somewhat a hack, in
that all I do is replace the -O0 in the NVCC_CXXFLAGS with -O2 for
debug builds... which means that they will be use both -g and -O2 and
be an optimized debug. This should still allow for debugging as
necessary. However, if not, we will have to look at an alternative
approach.
M Makefile.in
- Add in target "cleancuda" which just removes the .o files associated
with .cu files. Used for debugging.
M aclocal.m4
- Change -O0 to -O2 to fix bug in nvcc linking.
13 lines of code changed in 2 files:
Use shorter list of libs
2 lines of code changed in 1 file:
Fixes for the CUDA linking for non-static builds. For CUDA builds, we
need to link in all libraries for executables (eg: sus, test programs,
etc) because Core_Util has CUDA in it and thus the extra dlink object
file must be linked against anything that uses Core/Util, but the
dlink file also has code from many other libraries (they all have to link
together to get the full dlink file). It might be possible to
separate all the CUDA linking out, but at least for now this will work.
Created ALL_STATIC_PSE_LIBS and ALL_PSE_LIBS variables in Makefile.in
that can be used to link executables.
Placed a list of all cuda files (.o version) in the variable
DLINK_FILES (it is added to in each sub.mk). Then, once all libraries
are built, this list of files is linked with -dlink (previously was
just linking all the archive (.a) files together, but nvcc won't let
you do this with shared (.so) libs.
Updated a number of sub.mk files to specify the complete list of libs
necessary to link and added in the CUDA_LIBRARY.
Alphabetized and removed duplicates in a lot of sub.mk files.
M CCA/Components/Models/ModelFactory.cc
- Don't use the RadiationDriver if radation was not turned on.
M Makefile.in
M CCA/Components/Models/Radiation/RMCRT/sub.mk
M CCA/Components/Models/Radiation/sub.mk
M CCA/Components/MPM/sub.mk
M CCA/Components/Wasatch/Transport/sub.mk
M CCA/Components/Wasatch/sub.mk
M CCA/Components/Wasatch/Expressions/sub.mk
M CCA/Components/Wasatch/Expressions/Turbulence/sub.mk
M CCA/Components/Wasatch/Operators/sub.mk
M CCA/Components/sub.mk
M CCA/Components/Schedulers/sub.mk
M CCA/Components/Arches/sub.mk
M CCA/Components/Examples/sub.mk
M Core/Util/sub.mk
M StandAlone/sub.mk
M StandAlone/tools/mpi_test/sub.mk
M StandAlone/tools/pfs/sub.mk
M StandAlone/tools/fsspeed/sub.mk
M StandAlone/tools/puda/sub.mk
M StandAlone/tools/tracker/sub.mk
M StandAlone/tools/dumpfields/sub.mk
M StandAlone/tools/graphview/sub.mk
M StandAlone/tools/extractors/sub.mk
M StandAlone/tools/sub.mk
M StandAlone/tools/compare_mms/sub.mk
M StandAlone/Benchmarks/sub.mk
M testprograms/CubeRootTest/sub.mk
M testprograms/TestFastMatrix/sub.mk
M testprograms/RegionTest/sub.mk
M testprograms/TestMatrix3/sub.mk
M testprograms/IteratorTest/sub.mk
M testprograms/Regridders/sub.mk
M testprograms/sub.mk
M testprograms/PatchBVH/sub.mk
M testprograms/Malloc/sub.mk
M testprograms/Thread/sub.mk
M testprograms/BNRRegridder/sub.mk
560 lines of code changed in 60 files:
Missed two files in previous commit...
M CCA/Components/Schedulers/GPUDataWarehouse.cu
- Remove the implementation of isThread0_Blk0 from this file as it is in Core/Util/GPU.cu.
- #include GPU.h to get isThread0_Blk0.
- Fix indention.
M CCA/Components/Examples/UnifiedSchedulerTestKernel.cu
- Don't #include the .cu file, only need GPUDataWarehouse.h now.
- Added whitespace to make function variable list readable.
- Fix some indentation.
45 lines of code changed in 3 files:
Fix for static build when GPU (NVCC) is turned on. If you have
external device specific functions in .cu files, then you need to link
them with the -dc (in addition to the -c) flag. Then you have to link
all of these files together in a separate step using the nvcc
compiler's -dlink flag. This update removes the compilation of .cu
files in multiple other files (which thus created multiple definitions
of the same symbols).
I've created a new faux library lib/libgpu_extra_link.a that holds the
output from the -dlink command, and this is linked into sus.
M configVars.mk.in
- Added "-dc" to the nvcc link line.
M Makefile.in
- Build lib/libgpu_extra_link.a if Static and CUDA.
M StandAlone/sub.mk
- Sus now depends on lib/libgpu_extra_link.a (if Static/CUDA).
M include/sci_defs/cuda_testdefs.h.in
- Cosmetic fixes: cleaned up indentation and a few comments.
M CCA/Components/Models/Radiation/RMCRT/RayGPUKernel.cu
- No longer include the .cu files, just include the .h file.
M CCA/Components/Solvers/sub.mk
- Cosmetic fixes: Alphabetize, remove duplicate libs, consolidate if's.
M CCA/Components/Schedulers/GPUDataWarehouse.h
- Cosmetic fixes, clean up spacing, make easier to read.
- Remove unnecessary ;'s after constructor/destructor.
M CCA/Components/Schedulers/UnifiedScheduler.cc
- Indent.
- Group common variables.
M CCA/Components/Arches/sub.mk
- Cosmetic fixes: Cleanup, Indentation, consolidate if's.
M Core/Util/GPU.h
- Fix namespace: uintah -> Uintah
- isThread0_Blk0() is an external function, declare it as such.
M Core/Util/GPU.cu
- Fix namespace: uintah -> Uintah
- Fix function specification layout.
136 lines of code changed in 15 files:
Alphabetize list of libs (and thereby remove duplicates)
5 lines of code changed in 1 file:
Removed bug in Arches.cc relating to variable abskp of the radiation property calculator.
4 lines of code changed in 2 files:
Removed cout statement in problem initilization.
0 lines of code changed in 1 file:
Fixed syntax error in python script.
1 lines of code changed in 1 file:
--Changed the default name of abskp so regression testing passes.
--Added Coal case to the nightly RT.
2 lines of code changed in 2 files:
This patch changes the interface used to invoke radiation property models: hottel-sarofim, Burns-christian, and constant.
///---IMPORTANT!! OLD INPUT FILES WILL NO LONGER PARSE CORRECTLY AND MUST INVOKE THE DESIRED PROPERTY MODEL USING "property_calculator" WITHIN THE RADIATION MODEL --//
The property models are now called using property_calculator. The regression test input files show the correct syntax to invoke the property models.
This was done to improve the modularity of the code.
A coal regression test has also been added.
Other notes:
--The OPL for hottel-sarofim can no longer be passed through the radiation model and must also be passed through the property_calculator.
--The label for the absorption coefficient must be set through the property_calculator.
--For coal cases, the label of abskp must be set to "Julien_abskp" in both the PropertyModel and property_calculator. The user will be given more freedom in future patches. This was done to prevent a multiple-computes error.
--For coal cases, computation of the abskp values is less than ideal. Currently, abskp is computed for every time step, however it is only needed when then the radiation solve is executed. This will be removed in future patches.
--For coal cases, a bug was removed where abskp was summed to be 3x its correct value when using second order Runge-Kutta. Fortunately, first order Runge-Kutta computed the correct value for abskp.
468 lines of code changed in 17 files:
* Prepare build system CUDA support for separate CUDA compilation and device code linking (Titan static builds)
* need NVCC flag -arch=sm_* to work with NVCC flags: -dc and -dlink
* Generate new configure script
* Cleanup Examples sub.mk - remove old lines refering to non-existent GPU example components
* Add temporary fix for CPU/GPU RMCRT carry forward issue
10 lines of code changed in 4 files:
Fix case of transport in sub.mk
1 lines of code changed in 1 file:
Fix case in #includes
14 lines of code changed in 12 files:
Copied the r51974 version of 'transport' into Transport
5516 lines of code changed in 15 files:
Removing 'Transport' in preparation for re-adding it with file history.
0 lines of code changed in 15 files:
Clean up a few warnings and modify some formatting.
22 lines of code changed in 1 file:
Adding a gaussian initialization function to the property model.
To do: This really could be generalized to transport eqns, sources, etc...
61 lines of code changed in 3 files:
Commit on behalf of Abhishek: Adding Prism to comply with make runLocalRT for GPU tests.
1 lines of code changed in 1 file:
This commit is on behalf of Abhishek and contains GPU related changes.
198 lines of code changed in 5 files:
Don't know how double slashes got in there... remove them.
15 lines of code changed in 13 files:
Fix the transition to the Transport (capital T) directory from the old
lowercase 'transport' in Wasatch. For some reason, in a previous
commit today, both the Transport and transport were added to the tree
and coexisted. I have removed 'transport' and made sure Uintah compiles
on both Macs and other systems with case sensitive and insensitive file
systems.
Note, unfortunately I think the SVN history on the files in the Transport
directory have been lost to this point, though you probably could check
out a tree at the previous revision and find them.
5439 lines of code changed in 20 files:
There are currently two transport directorys: 'Transport' and 'transport' in the SVN repository. I am removing the lowercase one. I will then fix everything to compile correctly.
0 lines of code changed in 15 files:
AMRICE.cc & DDT1.cc
problemSetup()
- allow the single level regridder to pass through the bulletproofing
SingleLevelRegridder.cc
problemSetup()
- compute the refinement ratio for each level. Do not use the refinement ratio from the oldGrid.
The regridder's definition of the RR is different than what is stored in the Level.
AMRSimulationController.cc
run()
-reversed the order of the calls to doRegridding()
26 lines of code changed in 5 files:
ScheduleAndDoDataCopy() & copyDataToNewGrid
- Added {}
- Added comments
- Changed loop over levels index variable i-> L
- More descriptive variables names copySet -> copyPatchSet & refineSet -> refinePatchSet
100 lines of code changed in 1 file:
Add PIDX output context file.
117 lines of code changed in 2 files:
Add pidx to sub.mk files.
25 lines of code changed in 8 files:
Oops - hopefully this one does the job. Apparently my previous commit removed quite a few files from the build.
0 lines of code changed in 1 file:
Disable building of the missing PIDX file until John can commit it.
1 lines of code changed in 1 file:
Reverse changes to a sub.mk
1 lines of code changed in 1 file:
Revert "Capitalize the transport directory (e.g. Transport) to be consistent with the rest of Wasatch naming convention."
This reverts commit 90963897ecbf8a340efcce9305eddfc3eb8d3fb5.
5481 lines of code changed in 30 files:
Revert "more fixes for changing the transport directory name"
This reverts commit 08baebe605d97042840fa229267e0a6fb1f335af.
20 lines of code changed in 6 files:
Revert "sorry for the spam..."
This reverts commit 8ff5568bafaaa8fcd0e62d3456659e7a38c5ccec.
14 lines of code changed in 10 files:
Revert "fix warches"
This reverts commit 3c5513938f0dd4f980bbf4985906e21752357323.
6 lines of code changed in 3 files:
Add include files for pidx/visus.
73 lines of code changed in 2 files:
Squashed commit of the following:
commit be6e7652561b8d5cc85fb4cc6fe6bf3a178b9330
Author: John Schmidt <john.schmidt@utah.edu>
Date: Wed Jun 11 15:51:31 2014 -0600
Initial incorporation of PIDX.
1278 lines of code changed in 9 files:
computeMinMax()
- removed duplicate "put" for lastCompTimeLabel. The reduction variable is updated in MinMax::doAnalysis() task.
0 lines of code changed in 2 files:
fix warches
6 lines of code changed in 3 files:
sorry for the spam...
10 lines of code changed in 9 files:
more fixes for changing the transport directory name
8 lines of code changed in 6 files:
Forgot to commit this with the previous commit of Arenisca.cc
4 lines of code changed in 1 file:
Untested implementation of the ability to specify a variable amount of
Initial Disaggregation Volumetric Strain within a given instance of a material.
User needs to specify <volumeFraction>double</volumeFraction> in the geom_object
associated with the material for which an IDVS > 0 is desired.
66 lines of code changed in 1 file:
document a few header functions and fix the build system
1 lines of code changed in 2 files:
Capitalize the transport directory (e.g. Transport) to be consistent with the rest of Wasatch naming convention.
5662 lines of code changed in 28 files:
if (newGrid != oldGrid) {
sharedState->setRegridTimestep(true) On **ALL** mpi ranks not just rank 0
}
6 lines of code changed in 1 file:
add nebo-based boundary conditions to Wasatch. This change will allow us to execute boundary conditions on GPUs. This is a transitional commit and you will notice an increase in memory usage due to supporting both types of applying boundary conditions. Not all boundary conditions are currently FULLY GPU-ready. As we go through the transition, more boundary conditions will be GPU ready. With this commit however, all constant-value boundary conditions are GPU ready.
511 lines of code changed in 7 files:
Make "ReductionTasks" (SCI_DEBUG flag) output readable when reported by multiple threads.
8 lines of code changed in 1 file:
Clean up tabs and excessive spacing
152 lines of code changed in 1 file:
Removing the conditional that the traction be nearly zero when applying contact.
It still must be in compression or approaching. This causes a few regression
tests to fail, but inspection of the results indicate very tiny changes, and
this change improves results in at least one case I studied carefully.
Also, changed a BC in the Charpy test so that the striker doesn't tend to stick
to the top boundary.
0 lines of code changed in 2 files:
Added option to initialize the velocity field to a power law profile with variance.
0 lines of code changed in 4 files:
scheduleSolve()
Fixed the bulletproofing when periodic BCs are used.
The non-periodic directions can have any grid resolution
0 lines of code changed in 2 files:
Making sure that the intrusion temperature is initialized at t=0.
Modifying a test to set the intrusion T to something other than the
default.
2 lines of code changed in 1 file:
Removed code that has been #if 0 out for years.
0 lines of code changed in 1 file:
A bit more cleanup and formatting in OnDemandDataWarehouse.
270 lines of code changed in 1 file:
Actually fix the potential bug this time...
2 lines of code changed in 1 file:
* Fix issue related to adding uninitialized ReductionVariables to lock-free DataWarehouse.
* Add plentiful, consistent formatting, whitespace, and "{ }" to help code readability.
* Add Mutex for std::cout (for readable output when using multiple threads).
* Add #ifdef for GPU-related DebugStream.
1265 lines of code changed in 1 file:
Adding a special outflow boundary condition for variable density cases. Adding a new dynamic model for alpha in the pressure projection algorithm.
712 lines of code changed in 15 files:
Fix a potential bug.
1 lines of code changed in 1 file:
Removed tabs.
1 lines of code changed in 1 file:
Adding a simple du/dx = 0 boundary condition @ the outlet.
255 lines of code changed in 3 files:
Fix indentation so code is readable.
3 lines of code changed in 3 files:
Remove configure's search for a now obsolete header.
2 lines of code changed in 2 files:
Eliminate obsolete header file.
0 lines of code changed in 12 files:
clean up a few warnings.
10 lines of code changed in 2 files:
minor formatting changes
7 lines of code changed in 1 file:
clean up some formatting on TaskInterface and use a BOOST_FOREACH rather than an iterator loop.
34 lines of code changed in 1 file:
Change CQMOM to not require density.
1 lines of code changed in 1 file:
Global formatting:
-added {}
-added vertical and horizontal spaces.
90 lines of code changed in 1 file:
Changed the DOM16 regression test to use the new interface for radiation. Also fixed a bug associated with the new interface.
3 lines of code changed in 1 file:
Add in error check to CQMOM. Remove left over headers from when this was stand-alone. Silence complier warnings.
24 lines of code changed in 2 files:
Add multi-threaded NIGHTLY, LOCAL and RMCRT tests to ARCHES component.
21 lines of code changed in 1 file:
updated xml variable
2 lines of code changed in 1 file:
Re-enable the CQMOM regression tests, and add a CQMOM-only option for the tests.
15 lines of code changed in 1 file:
Add multi-threaded nightly tests to Examples components.
5 lines of code changed in 1 file:
Comment out CQMOM sources until these are developed. This should fix a memory leak.
12 lines of code changed in 1 file:
Fix CQMOM indexes for non-square moment specification. Remove some comments.
4 lines of code changed in 2 files:
scheduleTimeAdvance()
- set the d_doingRestart = true when performing a regrid on a restart.
When you regrid on a restart scheduleTimeAdvance() is called twice.
6 lines of code changed in 1 file:
Added Disaggregation Algorithm to Arenisca(2)
159 lines of code changed in 2 files:
May 2014 »