Remove some obsolete Warches-related comments in FieldAdaptor. Thanks James.
0 lines of code changed in 1 file:
Change CalculateKE xml spec to ComputeKE for nomenclature consistency and being elegance.
2 lines of code changed in 1 file:
Rip out all leftovers of dmomdt calculations. the varden tests will fail checkpoint comparison because we now have one less variable to save at checkpoint. no differences in answers whatsoever are expected on all the Wasatch RT tests.
0 lines of code changed in 2 files:
Fix bug in computation of the boundary condition edge cells and their translation to SpatialOps-Friendly indices. Cleanup some nomenclature for edge cells.
30 lines of code changed in 5 files:
Fix numerous variable names to conform to Wasatch's coding conventions.
61 lines of code changed in 1 file:
Add boundary edge masks to the BCHelper. Allow boundary conditions to set special values at those edges. Not all BCs will require to set this, but some will need to.
100 lines of code changed in 6 files:
rip out the the remains of dmomdt terms from the pressure_rhs caculation. new formulation shows that we don't need those.
10 lines of code changed in 6 files:
Add a POSTPROCESSING graph that allows us to schedule tasks post-time-advance for quantities that require updated transported variables. Add an option to compute the continuity residual error.
334 lines of code changed in 13 files:
Add in an extra parameter for homogeneous nucleation. Add a check to keep aggregation efficiencies greater than 0. Eliminate numerical error in aggregation for 3rd moment by making it 0.
61 lines of code changed in 7 files:
Change the order of blas and lapack libraries for static linking.
24 lines of code changed in 30 files:
Change Atmosphere boundary type to Open (thanks J. C. Sutherland for the suggestion). add a test to exercise the new bcs.
8 lines of code changed in 4 files:
Add outflow and atmosphere boundary conditions. Improve the logic when applying user-friendly boundary types. Move pressure-related bcs (matrix coefs, pressure) to the BCHelper. This commit could potentially break two tests with diffs @ machine precision.
629 lines of code changed in 17 files:
Quiet run time warnings about affineTransformation entries.
0 lines of code changed in 2 files:
Display start/end time when running local RT.
8 lines of code changed in 1 file:
removed extra spaces.
removed old version of updateSumIDevice()
rayTraceKernel()
- added get() calls for abskg, sigmaT4 and celltype for
the entire computational domain.
22 lines of code changed in 1 file:
More heavy construction in GPU:RMCRT
global variable name change:
size->nRaySteps
It will compile.
163 lines of code changed in 4 files:
added
- findRayDirectionDevice
- rayLocationDevice
- updateSumIDevice -- still under construction.
a bunch of operators for double3 and uint3 variables.
Still under heavy construction.
361 lines of code changed in 3 files:
Make some changes to the QMOM subgrid mixing model to maintain stability, and fix some index typos from the refactor.
42 lines of code changed in 5 files:
Construction zone.
-created RMCRT_flags struct. Now passing that around.
-created core conditional blocks inside of the Kernel.
It compiles.
114 lines of code changed in 3 files:
created several structs that are being passed into the kernel.
170 lines of code changed in 5 files:
Rename all _rhs_full momentum quantities to _rhs for consistency with our transport equation model. This will break some RT tests due to the new variable names.
11 lines of code changed in 3 files:
cleanup and simplify some the some IC and BC code for momentum.
49 lines of code changed in 2 files:
Repair a nasty bug that made it to the bc modifiers.
1 lines of code changed in 1 file:
Add support for vector bcs. this should fix Warches failures.
7 lines of code changed in 2 files:
Add more functionality to the BCHelper class. This commit paves the way for logical reasoning on boundary types (Wall, outlet...): one can check the validity of user-specified BCs, or automate the application of BCs for certain equations. This greatly simplifies the input file and takes a removes most of the guess-work that a user may be faced with. The neat thing is that we also allow a USER boundary type for freedom in applying any desired BC.
322 lines of code changed in 14 files:
extraSetup()
- passing grid into the method.
- define archesLevel here.
- added bulletproofing.
For readability removed extra_local_labels vector and explicitly state the
varLabels.
69 lines of code changed in 4 files:
Convert line endings in some recently committed Wasatch files from CRLF to LF.
1156 lines of code changed in 4 files:
_archesLevelIndex is now a global variable that never changes.
cleaned up sched_initialize() and initialized()
- cellType is initialized on the non-arches level
- divQ is only computed on the arches level.
36 lines of code changed in 2 files:
ARCHES.py
- added multi-level tests to the RMCRTTESTS
rmcrt_bm1_ML.ups
- only save divQ on the Arches Level.
3 lines of code changed in 1 file:
Revert previous change. It doesn't work for all task.
0 lines of code changed in 1 file:
Updates due to an API change in SpatialOps.
1159 lines of code changed in 3 files:
More descriptive multiple computes message.
9 lines of code changed in 3 files:
Report a moving mean for the hypre solver instead of a total mean. This moving average is based on an a simple exponential moving average formula and is the same as that used to report the mean timestep size, for consistency.
13 lines of code changed in 1 file:
1. Fix a typo in the function faceTypeToString that I added last week.
2. Allow one to access a boundary face iterator when no boundary conditions have been specified at that boundary face.
17 lines of code changed in 3 files:
RMCRT_test.cc: initializeWithUda()
- Simplify how you load data from a uda and put it in the datawarehouse.
This will fix the memory leak.
RMCRT_udaInit.ups
- dump 2 timesteps
- dump 1 checkpoint so the regression tester has something to compare with.
0 lines of code changed in 2 files:
Fix RMCRT_Test compiler error.
25 lines of code changed in 3 files:
Some cleanup on the UnifiedScheduler test component.
17 lines of code changed in 3 files:
First wave of refactoring for RMCRT to use the new Unified Scheduler GPU-DataWarehouse interface.
This is essentially all the "external plumbing". Internal work will be to next re-write the rayTrace kernel to fit with the recent CPU-RMCRT changes.
170 lines of code changed in 7 files:
Clean up old UnifiedScheduler GPU interface usage in RMCRT test component.
6 lines of code changed in 2 files:
Cleanup misleading typo in Task class code:
"void* device" -> "void* stream"
37 lines of code changed in 4 files:
restore svn:executable property for Wasatch.py. This prevented the nightly RT from running Wasatch.
1 lines of code changed in 1 file:
Remove GPUScheduler test component. Eliminates confusion with UnifiedScheduler.
0 lines of code changed in 3 files:
Further interface updates to the new BC treatment in Wasatch. Today's version is based on Boundaries instead of Boundary-Conditions. As a result, a new Boundary Type was added to the Face spec in Uintah. This allows us to reason on boundary types and select appropriate boundary conditions.
470 lines of code changed in 11 files:
Added restartUda_RT used to in RMCRT regression testing.
Added RMCRT RT test that restarts a previously run Arches simulation
and runs for 1 timestep. Scattering and reflections are turned on.
10 lines of code changed in 1 file:
Ray.cc: updateSumI() in Ray_scatter section
- let the user control if the random number generator seed is random.
RMCRT_test.cc
- now using the correct timeIndex in all dataArchive queries. Previously,
we were using the timestep number.
31 lines of code changed in 2 files:
should have been added to last commit
2 lines of code changed in 1 file:
RayTrace()
- added intensity clamp if reflections are not allowed.
reflect()
- adjust the sign on the ray direction
17 lines of code changed in 1 file:
use d_allmpi to check if running under MPI
0 lines of code changed in 2 files:
Last chance for this RT test. If it continues to fail, it will be
removed.
Isaac: Please see me if you need help debugging the input file.
2 lines of code changed in 1 file:
Clean up a few warnings coming out of BCHelper.cc
0 lines of code changed in 1 file:
Commit on Abhishek's behalf. Minor changes getting some things ready for Wasatch to use GPUs.
128 lines of code changed in 10 files:
RMCRT_Test::initializeWithUda ()
- added machinery to take into account if the user saved the temperature and absk as floats,
instead of doubles.
added more descriptive output.
74 lines of code changed in 2 files:
deleted local variable
2 lines of code changed in 1 file:
Do not call MPI_Comm_dup if not running under MPI/threads.
1 lines of code changed in 1 file:
Allocate MPI communicators based on taskgraph.
This fixed a hung when there are large number of reduction tasks in the taskgraph.
18 lines of code changed in 4 files:
Minor cleanup on Tony's awesome commit.
- Fix some const correctness issues
- cosmetic changes
- Use a typedef for bcFunctorMap_ throughout to reduce a lot of boilerplate template code when handling this map.
218 lines of code changed in 7 files:
Source terms were calling problemSetup multiple times. This has been
fixed.
Also updating the spec on the methane_RCCE input file.
23 lines of code changed in 1 file:
minor formatting changes
6 lines of code changed in 1 file:
Divided the global struct d_customBC_var_basket into two structs global & local.
Previously the var_basket was a "container" that was used to pass around both local and
global variables. In the threaded world that doesn't work. Therefore there's now
customBC_globalVars and customBC_localVars
The local struct contains all variables and structs that are local to the task/function.
The global struct contains only global variables.
This commit should not change the answers but the RT tests do not test all of the custom
boundary conditions.
295 lines of code changed in 12 files:
Cleanup the varden boundary conditions. rip out an un-necessary enum that was added there as well. Restrict the options that the user can specify through the input file for those varden BCs.
53 lines of code changed in 5 files:
This is a major (and beautiful) commit for Wasatch. It entails a complete refactoring of the boundary conditions for ease of use, efficiency (?), and the new Nebo syntax for handling masks (e.g. BC iterators) in general. Some of the things were changed in this commit include:
1. ripping out the functor_name from the uintah boundary condition suite. those are now set in the <value> tag of the boundary condition.
2. converting all boundary conditions to use ijk indexing instead of flat. Nebo will eventually do its magic behind the scenes.
3. A new BCHelper class that manages all boundary conditions, their iterators, names etc...
4. A pre-processing step on input files to guarantee that <Face> specification has a unique name. This pre-processing step will automatically rename duplicates or put unique names if the user has not specified any.
A key assumption for this new approach is centered around the <Face> specification as being a single boundary condition - that is, you MUST specify ALL relevant boundary conditions within the same <Face> spec.
1693 lines of code changed in 37 files:
Add a var label to the normalized scalar variance model, so that the raw scalar variance value can be used elsewhere.
26 lines of code changed in 4 files:
Turning the test back on.
2 lines of code changed in 1 file:
Remove const_props test until memory leak is fixed.
3 lines of code changed in 1 file:
Change UnifiedScheduler numthreads range message to, [2,64]
1 lines of code changed in 1 file:
Forgot to uncomment the radiation petsc test in one of my pervious commits.
1 lines of code changed in 1 file:
Second attempt at removing the auxiliary boundary condition. This time, I put in the fixes - to the best of my knowledge - for the other components. The majority of the problems were fixed by a simple fix to the BCUtils.h getIteratorBCValueBCKind function. The other change was in MPMBoundCond.cc which I hope the MPM developers will agree with. Local regression passed on all components that provide localtests. I leave it to the buildbot and the nightly RT to check for the rest. 'If it ain't broke, then it probably violates the second law' (Tony Saad).
30 lines of code changed in 5 files:
Fix errors in ARCHES.py that were preventing the regression tests from running, including the buildbot and the nightly RT.
3 lines of code changed in 1 file:
add new operator to GPUGridVariable.
GPUGridVaraible can be accessed as var(x,y,z)
28 lines of code changed in 3 files:
Added array boundary checking for GPU varaible.
Fixed multi-GPU issue.
82 lines of code changed in 6 files:
detect and report GPU task kernel error
56 lines of code changed in 5 files:
Added a methane rmcrt multi-processor run.
Decreased number of rays for rmcrt_bm1_1L.ups.
3 lines of code changed in 1 file:
better multi-GPU debug output
19 lines of code changed in 3 files:
Cleanup old GPU interface.
0 lines of code changed in 4 files:
Remove unused variable to silence complier warning.
0 lines of code changed in 1 file:
Implemented a lightweight version of GPUDataWarehouse, GPUGridVariable and GPUArray3D inside GPU.
This commit included a new set of interfaces for better supporting GPU task.
See example in Examples/UnifiedSchedulerTest
691 lines of code changed in 27 files:
Add a regression test for the 2-equation scalar variance model.
2 lines of code changed in 1 file:
Refactor ClassicTableInterface member variables for consistency.
13 lines of code changed in 2 files:
Add in the source term and property model required for the 2-equation scalar variance model.
632 lines of code changed in 9 files:
Add methods to remove and replace xml attributes in ProblemSpec. This is part of a major refactor of Wasatch BC treatment - those specific methods will allow us to manipulate <Face> properties (e.g. name) to guarantee and/or generate unique names, for example.
31 lines of code changed in 2 files: