October 2013 Commit Log

Number of Commits:
159
Number of Active Developers:
12
jsutherland 2013-10-31 20:04 Rev.: 51200

Fix a name clash with "lock" global variable and a definition in the standard namespace.

4 lines of code changed in 1 file:

jsutherland 2013-10-31 19:46

another one in an #ifdef

0 lines of code changed in 4 files:

jsutherland 2013-10-31 19:44 Rev.: 51197

Missed this one because it was in an #ifdef that wasn't active on my build.
There will probably be a few more...

3 lines of code changed in 1 file:

jsutherland 2013-10-31 19:36 Rev.: 51196

This one has been a long time coming.

Rip out the standard namespace opening that has been polluting header files throughout the framework. This crept into all sorts of places.

DO NOT open namespaces in header files. Ever.

1975 lines of code changed in 833 files:

harman 2013-10-31 13:46 Rev.: 51195

Increased
MAX_NAME from 10 -> 20

get() & getModifiable()
- output detailed information when a GPUDataWarehouse item cannot be found.

39 lines of code changed in 2 files:

harman 2013-10-31 13:41 Rev.: 51194


added thread0_Blk0()
useful in conditional statements for limiting output.

74 lines of code changed in 4 files:

jsutherland 2013-10-31 13:09 Rev.: 51193

More namespace cleanup. Grrr!

53 lines of code changed in 34 files:

jsutherland 2013-10-31 12:38 Rev.: 51192

Apparently "rank" is defined in std namespace as well. Tighten this up on mpi_test.cc to get it compiling on new systems.

4 lines of code changed in 1 file:

jthornoc 2013-10-31 12:37 Rev.: 51191

Fixing a timestepping option that was accidently turned off (and thus
left unitialized variables) for the computeStableTimestep option.

2 lines of code changed in 1 file:

jsutherland 2013-10-31 12:34 Rev.: 51190

Fix type deduction that was failing on Mavericks (and new Xcode).

4 lines of code changed in 1 file:

jsutherland 2013-10-31 11:47 Rev.: 51189

to_string is in both the SCIRun and std namespaces. Due to some abuse in opening up namespaces in Uintah, some clashes were occurring on the new mac platforms. This should eliminate those clashes.

0 lines of code changed in 8 files:

jthornoc 2013-10-31 09:26 Rev.: 51187

Get rid of a memory error with the constant property.

0 lines of code changed in 2 files:

ahumphrey 2013-10-30 18:23 Rev.: 51186

Fix linking order of NVCC_LIBS.
Generate new configure script.

2 lines of code changed in 3 files:

tsaad 2013-10-30 13:12 Rev.: 51184

forgot to add parsing of Rayleigh-Taylor instability initial condition

19 lines of code changed in 1 file:

jthornoc 2013-10-29 15:43 Rev.: 51181

Cleaning up some warnings.

0 lines of code changed in 6 files:

jthornoc 2013-10-29 11:20 Rev.: 51175

Adding back in the option of setting an initial dt. This can actually
come in handy.

5 lines of code changed in 2 files:

jthornoc 2013-10-28 15:57 Rev.: 51173

No .h

0 lines of code changed in 2 files:

jthornoc 2013-10-28 15:17 Rev.: 51172

Get tabprops interface compiling.

22 lines of code changed in 2 files:

jthornoc 2013-10-28 15:17 Rev.: 51171

Cleanup of input files and removing unused options.

5 lines of code changed in 2 files:

jthornoc 2013-10-28 15:17 Rev.: 51170

More label cleanup.

The normalized scalar variance model was simplified to only compute the
normalized value of scalar variance. The absolute value of variance was
removed.

Now requiring soot for DO radiation model (for properties).

21 lines of code changed in 11 files:

jthornoc 2013-10-28 15:17 Rev.: 51169

Even more labels gone.

0 lines of code changed in 6 files:

jthornoc 2013-10-28 15:17 Rev.: 51168

Cleaning up more labels.

2 lines of code changed in 4 files:

jthornoc 2013-10-28 15:17 Rev.: 51167

Removing several old/unused labels.

24 lines of code changed in 13 files:

jthornoc 2013-10-28 15:17 Rev.: 51166

Getting rid of tempIN label.

319 lines of code changed in 31 files:

jthornoc 2013-10-28 15:17 Rev.: 51164

**MORE BIG CHANGES**
Continued cleanup of unused/old functionality. All RT tests will fail
due to changes in input files or output in uda. I have verified at this
point that the answer still haven't changed.

71 lines of code changed in 98 files:

jthornoc 2013-10-28 15:16 Rev.: 51163

Removing old RT tests.

0 lines of code changed in 1 file:

jthornoc 2013-10-28 15:16 Rev.: 51162

***BIG CHANGE***
This commit removes the old boundary condition mechanism in ARCHES along
with its dangling tentacles. As such, it will invalidate all the old
regression tests, which will need to be removed. A previous commit has
established new, equivalent RT tests.

Note that the old functions/cc code has been moved to the attic.

More work is required to rid Arches of the some of the old transport modules.

The new RT tests should still provide the same answers.

277 lines of code changed in 35 files:

jsutherland 2013-10-28 13:47 Rev.: 51161

Fix a subtle bug that was introduced in my large set of commits that showed up on optimized but not debug builds.

2 lines of code changed in 1 file:

jsutherland 2013-10-26 16:41 Rev.: 51160

Tony's suggested changes to MomentumTransportEquation.

20 lines of code changed in 1 file:

jsutherland 2013-10-26 16:41 Rev.: 51158

Fix a potential problem in parsing ordering by getting the density stuff parsed early on.

32 lines of code changed in 1 file:

jsutherland 2013-10-26 16:41 Rev.: 51157

Update comments in ScalarRHS to provide more clear documentation.

0 lines of code changed in 2 files:

jsutherland 2013-10-26 16:41 Rev.: 51156

Remove some obsolete comments.

0 lines of code changed in 1 file:

jsutherland 2013-10-26 16:40 Rev.: 51154

Update Arches side of Warches to reflect density changes. (+1 squashed commit)
Squashed commits:
[5981256] fix warches input files (+1 squashed commit)
Squashed commits:
[1e445e8] more input file updates (+1 squashed commit)
Squashed commits:
[03891f6] Update wasatch & warches input files to reflect changes to density specification.

11 lines of code changed in 1 file:

jsutherland 2013-10-26 16:40 Rev.: 51153

Modify Density specification in Wasatch to be a bit more concise in the case where constant density is used.

25 lines of code changed in 1 file:

jsutherland 2013-10-26 16:39 Rev.: 51152

Formatting changes, remove an output statement that would spam in parallel on cases with embedded boundaries.

6 lines of code changed in 2 files:

jsutherland 2013-10-26 16:39 Rev.: 51151

Fix some formulation inconsistencies and documentation for weak form and constant density cases.

71 lines of code changed in 5 files:

harman 2013-10-25 12:29 Rev.: 51150

Now using system boost on the RT machine.

2 lines of code changed in 1 file:

harman 2013-10-25 08:11 Rev.: 51149

fixed compiler warnings

Using "d_" for global variables.

35 lines of code changed in 2 files:

harman 2013-10-24 16:29 Rev.: 51148

More descriptive error messages.
Fixed typo in variable name.

23 lines of code changed in 1 file:

qymeng 2013-10-24 15:31 Rev.: 51147

schedule ray CPU task if -gpu argument not present, instead of depends on HAVE_CUDA macro.

17 lines of code changed in 5 files:

harman 2013-10-24 15:04 Rev.: 51146

Global: changed _pi -> M_PI which is defined in math.h

11 lines of code changed in 2 files:

jsutherland 2013-10-24 14:07 Rev.: 51145

Major step toward getting weak form of scalar transport equations implemented.

- Move embedded boundary area/volume fraction tags into FieldSelector/FieldTagInfo

- for weak form RHS, scale convective term by density so that the entire RHS can later be divided by density.

- for weak form RHS, add the drho/dt term to the RHS

- Add the primitive variable to the FieldTagInfo for use in RHS when weak form is used.

Note that the QMOM equations and the Scalability test equations are not currently set up to use weak form.

220 lines of code changed in 10 files:

jsutherland 2013-10-24 14:07 Rev.: 51144

Simplify some things in setting up terms in the scalar transport equation in preparation for implementation of weak form.

The only change when using weak form will be in the RHS. No changes in convective or diffusive terms are needed.

19 lines of code changed in 3 files:

qymeng 2013-10-24 13:18 Rev.: 51143

Add support for post GPU kernel processing callback.
A GPU task will be called multipule times and each time an event argument will be passed.
The task function can then use the event to decide what to run when
1) Task::CPU task is scheduled to run on CPU
2) Task::GPU task is scheduled to run on GPU
3) Task::postGPU task is finished execution on GPU, DW has been copied back.

CPU only task interface is not unchanged.

282 lines of code changed in 11 files:

jsutherland 2013-10-24 10:32 Rev.: 51142

Put in a missing \rho in the documentation.

1 lines of code changed in 1 file:

jsutherland 2013-10-24 10:23 Rev.: 51141

- Update some documentation on the scalar transport equation.
- Update doxygen to use MathJax for nicer displayed equations.
- misc. formatting changes.

89 lines of code changed in 10 files:

awabboud 2013-10-24 09:49 Rev.: 51140

Changed 2 cout to proc0cout to reduce spam in cases with large number of processors.

3 lines of code changed in 2 files:

tsaad 2013-10-24 09:27 Rev.: 51139

add some documentation to BCHelper

13 lines of code changed in 2 files:

tsaad 2013-10-24 09:17 Rev.: 51138

document the verify_boundary_conditions member function for transport equations.

7 lines of code changed in 1 file:

tsaad 2013-10-24 09:05 Rev.: 51137

remove an obsolete expression

1 lines of code changed in 4 files:

tsaad 2013-10-24 09:05 Rev.: 51136

This commit introduces a dynamic coefficient to the Wasatch variable density formulation. This coefficient is based on scaling arguments for the continuity equation as well as the degenerate case of constant density. The model is not turned on at the moment as we go through V&V. The pressure source now computes multiple quantities related to this dynamic coefficient. This commit will break all varden RTs due to a change in algebraic operations and the use of new fields to store intermediate variables.

79 lines of code changed in 8 files:

harman 2013-10-23 16:43 Rev.: 51135

added ideal scaling line

0 lines of code changed in 2 files:

dav 2013-10-23 16:04 Rev.: 51134


M Point.h
M CompGeom.cc

Removed the Point's vector() operator that did a hard cast of the
Point into a Vector. While the Point and Vector have exactly the same
data members and this cast has (in theory) worked forever, at least on
newer Macs, casting one to the other does not work. It appears that this
was only used in one or two places, and they should just use the other
(constructor, etc) available functions for type safety reasons anyway.

Note, 'vector()' was also used in: src/CCA/Components/Wasatch/CoordHelper.cc
but James fixed this a commit or two ago. This error was what clued us
into the problem.

14 lines of code changed in 3 files:

harman 2013-10-23 12:29 Rev.: 51133

Corrected the replace_all_GS instructions.

1 lines of code changed in 1 file:

jsutherland 2013-10-23 10:11 Rev.: 51132

Fix a compiler warning coming out of a switch statement, improve error trapping, and minor reformatting.

17 lines of code changed in 1 file:

jsutherland 2013-10-23 07:34 Rev.: 51131

Update wasatch3p build script for new boost finding stuff.

3 lines of code changed in 1 file:

qymeng 2013-10-22 14:11 Rev.: 51130

fix build on blue waters due to new mpi libary changes.

0 lines of code changed in 2 files:

dav 2013-10-22 13:47 Rev.: 51129


Few cleanups based on icc compiler.

M configure
M configure.ac

- Need to clear out the list of test_libraries so that the next test only looks at those specified,
and not libs from a previous test.

M CCA/Components/LoadBalancers/ProfileDriver.h
M CCA/Components/LoadBalancers/ProfileDriver.cc

- Use d_ for class variables.
- Add some whitespace to make code readable.
- Line up variable names to make them much easier to read.

M CCA/Components/Arches/EfficiencyCalculator.h
M CCA/Components/Arches/NonlinearSolver.h
M Core/Grid/BoundaryConditions/BoundCondBase.h

- Returning a 'const' basic type does not make sense and by removing the 'const', we quiet the compiler warning.
- Line up the variables so they can be easily read.
- Remove unnecessary ; from the end of class member functions.


108 lines of code changed in 9 files:

harman 2013-10-22 11:18 Rev.: 51128

Check for dialog and awk and warn user if they aren't found. Removed hard coded path.

0 lines of code changed in 2 files:

amir 2013-10-22 01:54 Rev.: 51127

Adding rho0 and rho1 to the VarDensMMS BCExpressions which depend on them so we can specify them from the input file

113 lines of code changed in 7 files:

jsutherland 2013-10-21 20:43 Rev.: 51126

Minor changes to DynamicSmagorisnkyCoefficient:
- Use inline Nebo implementation of BoxFilter.
- Make GPU runnable.
- inline in a few places where possible
- remove a cone use in favor of an if statement for better efficiency.
- Release a few temporary fields to reduce the memory footprint slightly.

99 lines of code changed in 3 files:

jsutherland 2013-10-21 20:42 Rev.: 51125

Fix an initialization order to silence a compiler warning.

0 lines of code changed in 2 files:

tsaad 2013-10-21 15:46 Rev.: 51124

Add a StepFunction with perturbation at the interface. This can be used as an initial condition for a Rayleigh-Taylor instability simulation.

161 lines of code changed in 1 file:

jthornoc 2013-10-21 11:15 Rev.: 51122

Getting rid of some warnings.

Martin: Please make sure I didn't screw up your Newton solve in EnthalpyShaddix by initializing the old temperature guess. Previously, this value was uninitialized.

7 lines of code changed in 11 files:

jthornoc 2013-10-21 09:01 Rev.: 51121

Fix the error checking on IVs for boundaries.
Make sure that the StABL BC gets a celltype.

7 lines of code changed in 3 files:

tsaad 2013-10-21 08:34 Rev.: 51120

reduce the number of operators used in the PressureSource expression.

24 lines of code changed in 2 files:

tsaad 2013-10-19 13:20 Rev.: 51119

1. move drhodtNP1 from the PostProcessing directory to the Expression directory.
2. Rename drhodtNP1 to Drhodt in anticipation of using it to compute the pressure source term
3. Rename "pressure-source-term" to "pressure_src" and add that to the TagNames for ease of access. This will break the varden regression tests due to this rename.

322 lines of code changed in 11 files:

dav 2013-10-18 16:51 Rev.: 51118

New MKL lib list for newest Ember OS configuration

26 lines of code changed in 2 files:

jthornoc 2013-10-18 15:07 Rev.: 51117

Fix the parsing of the freestream vel for the StABL BC.

33 lines of code changed in 1 file:

tsaad 2013-10-18 12:08 Rev.: 51115

do not report warning when no values are specified for symmetry and zeroNeumann BCs. To the best of my knowledge, those two bcs are specific to ICE.

3 lines of code changed in 1 file:

jthornoc 2013-10-17 18:24 Rev.: 51113

Cleanup of an earlier commit. Made OPL required for the hottel charts.

28 lines of code changed in 2 files:

jthornoc 2013-10-17 18:24 Rev.: 51112

Turning on new tests which will replace the old ones.

14 lines of code changed in 1 file:

jthornoc 2013-10-17 18:24 Rev.: 51111

Removing the _has_been_computed boolean usage.

15 lines of code changed in 2 files:

qymeng 2013-10-17 11:20 Rev.: 51110

fix a bug in find_sendset from last commit.

0 lines of code changed in 2 files:

ahumphrey 2013-10-17 11:02 Rev.: 51109

Add support for arbitrary Uintah constVariables to return their data sizes (templated) as well as copy their underlying data to a destination buffer for automatic host<-->device copies.

Also make UnifiedScheduler a friend of OnDemandDataWarehouse.

29 lines of code changed in 7 files:

jthornoc 2013-10-16 15:49 Rev.: 51106

1) Better error checking on boundaries -- checking for complete spec of table IVs
2) Cleaning up some dummy solves and an "oldTableHack".
3) Update mpmarches input file to have full IV spec. This should get this methaneFireWContainer passing.

40 lines of code changed in 8 files:

ahumphrey 2013-10-16 14:22 Rev.: 51105

Add support for single GPU usage on multi-GPU nodes. This should ease the pain of debugging in multi-GPU environments.

8 lines of code changed in 1 file:

jsutherland 2013-10-16 14:18 Rev.: 51104

Use value_type rather than now obsolete AtomicT.

1 lines of code changed in 1 file:

jsutherland 2013-10-16 12:08 Rev.: 51102

Remove references to Stencil2.h, s2detail, etc.

0 lines of code changed in 8 files:

jsutherland 2013-10-16 10:03 Rev.: 51100

Reorder a few things in ConvectiveFlux to improve efficiency for central difference situations.

0 lines of code changed in 2 files:

jsutherland 2013-10-16 10:03 Rev.: 51099

Rework the way we are building the windowed "src" fields to be a bit more robust.

0 lines of code changed in 4 files:

ahumphrey 2013-10-16 09:41 Rev.: 51098

Add support for arbitrary Uintah Variables to return their data sizes (templated) as well as copy underlying data to a destination buffer.

This functionality will be essential for the Unified Scheduler to genericaly handle automatic copying of any Uintah Variable (of type<T>). Note: most of this functionality analogous to the preparing MPI buffers for Uintah variables.

144 lines of code changed in 21 files:

jthornoc 2013-10-16 09:14 Rev.: 51097

Updating the StABL BC to use a vector rather than a single value.
If the elevation goes below zero, set a constant Neumann BC.
Note that if the h < -dx, then this BC won't work because of the log.

79 lines of code changed in 2 files:

harman 2013-10-15 17:23 Rev.: 51093

Added bulletproofing.

The user ,ust be consistent if they're adjusting the output/checkpoint intervals. You can't
specify a "timestep interval" and a "interval."

0 lines of code changed in 2 files:

jsutherland 2013-10-15 16:34 Rev.: 51092

minor cleanup

5 lines of code changed in 1 file:

jsutherland 2013-10-15 16:34 Rev.: 51091

Fix coordinate retrieval that was breaking Wasatch on my mac. For some reason, with the new compilers I installed exposed a problem converting SCIRun::Point to SCIRun::Vector. This commit fixes that issue.

4 lines of code changed in 1 file:

ihunsaker 2013-10-15 16:33 Rev.: 51090

Completion of interface that allows RMCRT to call the hottel
property models.
Also contains some minor code cleanup.

22 lines of code changed in 6 files:

tsaad 2013-10-15 13:48 Rev.: 51088

rip out an obsolete function that was replace with function_NEW(). Enforce the type of boundary on difference and union boundaries.

0 lines of code changed in 4 files:

qymeng 2013-10-15 08:55 Rev.: 51087

fix a particle sending state thread safety issue.

16 lines of code changed in 3 files:

jsutherland 2013-10-14 18:07 Rev.: 51086

Changes to conform to Abhishek's recent commit on ExprLib.

2 lines of code changed in 1 file:

jthornoc 2013-10-14 16:36 Rev.: 51084

Updated the methane/container problem to use new BCs. **This will
change the answers**.

Note that this change basically disables the coupled heat transfer
between arches and mpm. MPMArches needs to be updated to allow for new
enthalpy naming convention.

Removed intrusion test from mpmArches because it is now covered under
Arches tests.

0 lines of code changed in 1 file:

jthornoc 2013-10-14 16:21 Rev.: 51083

Added a new line to the boundary warning.

Changed mpmpipe_test to use new BCs. **This will break RT**

6 lines of code changed in 2 files:

amir 2013-10-14 16:05 Rev.: 51082

Adding ability to handle the variable density cases and the MMS test case for pressure projection when the continuity residual is being computed

473 lines of code changed in 8 files:

tsaad 2013-10-14 15:49 Rev.: 51081

use proc0cout for a warning that I recently put in BoundCondFactory

2 lines of code changed in 1 file:

jsutherland 2013-10-14 15:49 Rev.: 51080

Silence a few compiler warnings by removing a few unused variables.

0 lines of code changed in 2 files:

guilkey 2013-10-14 15:47 Rev.: 51079

Add the Thomsen-Hartka water model files.

584 lines of code changed in 2 files:

guilkey 2013-10-14 15:15 Rev.: 51078

Adding the Thomsen-Hartka model for water to the MPM constitutive models.

14 lines of code changed in 9 files:

harman 2013-10-14 14:30 Rev.: 51076

-Heavily instrumented code.
-random number generators not being used.
It compiles and that's about it.


212 lines of code changed in 4 files:

ahumphrey 2013-10-14 14:04 Rev.: 51075

Fix typo in previously committed CUDA error handling wrapper.

1 lines of code changed in 1 file:

jthornoc 2013-10-14 13:53 Rev.: 51074

Changing the default behavior of arches to not suppress corner
velocity cells. This behavior is restored by adding the
<suppress_corner_recirculation/> tag to <Arches><BoundaryConditions>...

Updated input files to suppress the recirculation to keep RT passing.

207 lines of code changed in 2 files:

ahumphrey 2013-10-14 12:39 Rev.: 51072

Add error handling wrapper for CUDA Rand API calls.

9 lines of code changed in 1 file:

qymeng 2013-10-14 10:01 Rev.: 51071

default constructor for GPU vector

8 lines of code changed in 1 file:

jsutherland 2013-10-11 19:17 Rev.: 51070

Fix the part I missed that broke Warches.

1 lines of code changed in 1 file:

qymeng 2013-10-11 16:11 Rev.: 51069

Use derived class instead of union to support dimensional index.
This has been verfied to produce the same assumbly code as the orignal struct when using dimensional index.

40 lines of code changed in 1 file:

jsutherland 2013-10-11 15:46 Rev.: 51068

Major changes to how we handle Uintah::PerPatch<double> in Wasatch.

These fields are now operable with all SpatialOps functionality including GPUs. This is the first major change in what will likely be a few iterations to get this fully functional on GPUs, however.

The major areas of change relate to Reduction variables and any expressions that use the simulation time or time step value.

462 lines of code changed in 60 files:

jsutherland 2013-10-11 15:46 Rev.: 51067

Silence a compiler warning about order of initialization in a constructor.

1 lines of code changed in 1 file:

awabboud 2013-10-11 15:01 Rev.: 51066

Fix documentation for environment flag.

1 lines of code changed in 1 file:

awabboud 2013-10-11 14:57 Rev.: 51065

Move around the QMOM debug stream so that it only occurs if a problem arises, which avoids writing to the stream every single computation.

0 lines of code changed in 2 files:

jthornoc 2013-10-11 10:36 Rev.: 51063

Cleanup of input files.

7 lines of code changed in 1 file:

awabboud 2013-10-11 09:54 Rev.: 51062

Add in a minimum S for cylindrical diffusion and kinetic growth rates.

42 lines of code changed in 5 files:

jthornoc 2013-10-10 15:47 Rev.: 51061

Whoa....don't know how that got there...

0 lines of code changed in 1 file:

jthornoc 2013-10-10 15:26 Rev.: 51060

Removing from debug print statements.
Adding "new" tests to the RT.

21 lines of code changed in 3 files:

ihunsaker 2013-10-10 15:13 Rev.: 51059

Put in the actual calculations for hottel.F,
most of which was borrowed from radcoef.F.
Added necessary pieces to call the hottel
function in RadPropertyCalculator.

229 lines of code changed in 5 files:

tsaad 2013-10-10 14:47 Rev.: 51058

tidy up some input files and turn on the Velocity boundary. This will further reduce BC specification clutter.

14 lines of code changed in 1 file:

awabboud 2013-10-10 13:36 Rev.: 51057

Change the qmom debugging from ifdef output to a sci_debug stream. Change the warning for unrealizable momenets to a debug stream.

43 lines of code changed in 2 files:

tsaad 2013-10-09 15:52 Rev.: 51055

remove an obsolète cout statement

0 lines of code changed in 1 file:

jthornoc 2013-10-09 14:55 Rev.: 51054

Missing "void".

1 lines of code changed in 1 file:

jthornoc 2013-10-09 14:45 Rev.: 51053

Missing headers.

5 lines of code changed in 1 file:

jthornoc 2013-10-09 14:32 Rev.: 51052

RadPropertyCalculator:
Moving the derived class defs. into a *.cc file.
Adding the hottel/sarofim interface. Isaac to fill in the details.
Adding corresponding (empty for now) fortran code for Isaac.

499 lines of code changed in 8 files:

tsaad 2013-10-09 14:07 Rev.: 51051

1. remove trailing spaces when modifying bc values using the python script
2. convert all Wasatch input files to use the new bc value attribute
3. fix a Wasatch bug when parsing functors from attributes
4. specify which materials are allowed for DefaultMaterial

5 lines of code changed in 3 files:

tsaad 2013-10-09 13:11 Rev.: 51049

1. Remove the material id attribute from BCs in all Wasatch input files.
2. Add some error checking to the python scripts that make the above possible.

32 lines of code changed in 3 files:

ahumphrey 2013-10-09 13:07 Rev.: 51048

Add build system support for Cray PETSC libraries. This is for DOD HPCMP Open Research Systems (garnet and copper).
Generate new configure script.

53 lines of code changed in 2 files:

tsaad 2013-10-09 12:30 Rev.: 51047

1. Add a python script that converts ups <value> spec to value attribute
2. Add a <DefaultMaterial> option in the <BoundaryConditions> specification. This will allow us to get rid of the material id attribute in BCType for components that have a single material only. The behavior here is fully backwards compatible. If neither DefaulMaterial nor id are specified, then an exception is thrown.
3. Add a python script that will remove the material id and add a <DefaultMaterial> node to your ups file.
Let me know if you need help using these scripts.

294 lines of code changed in 5 files:

harman 2013-10-09 11:49 Rev.: 51046

Now using consistent naming conventions for local and global variables.


392 lines of code changed in 17 files:

dav 2013-10-08 16:51 Rev.: 51044

See previous commit: A few more libs need the fix for mac linking of zlib.

0 lines of code changed in 6 files:

dav 2013-10-08 16:43 Rev.: 51043

This commit will (hopefully) fix the Mac OSX regression tester.

M Makefile.in

- Don't run the verify_zlib if it is a static build.

M CCA/Components/ProblemSpecification/sub.mk
M Core/ProblemSpec/sub.mk

- On some Mac installs, the MPI library brings in a library load path
that has alternative system libraries in it. The most usual case is
with libz.dylib. If we don't add the MPI_LIBRARY line to this link
line, then these libs will link vs the system (usually /usr/lib/libz.dylib)
and thus we will have two different compression libraries linked
into the same executable... which will eventually cause a bad UDA to
be created.

M CCA/Components/ProblemSpecification/ProblemSpecReader.cc

- Add a better error message.

M CCA/Components/Regridder/TiledRegridder.cc

- Alphabetize the includes.

30 lines of code changed in 9 files:

dav 2013-10-08 15:50 Rev.: 51042

Fix compilation problem... I can't locate a "SCIRun::Iterator"
type/class/etc... I assume this is supposed to be a
"Uintah::Iterator". It at least compiles successfully this way... If
the RT finds a problem, I will revert this fix.

7 lines of code changed in 1 file:

jthornoc 2013-10-08 13:50 Rev.: 51041

[Dav de St. Germain commit via Jeremy's account/git:]

The following commit fixes a problem with the way Fortran files are
added into the Uintah framework. As you may know, the perl script
'fspec.pl.in' reads in a fortran spec file (eg: bcuvel.fspec) and
converts it into a .h file (bcuvel_fort.h), converting things such as
IntVectors into arrays of integers (etc) in order to pass Uintah
variables back and forth with the fortran code. To use the fortran
code, the generated .h file is included into a Uintah .cc file and the
function is called. Compilation problems arose when the fortran .h
file was included before a blanket "using namespace Uintah" appeared.
This was because the generated fortran .h file did not use the correct
"Uintah::" namespace qualifiers, and instead relied on some other file
to open up the namespace. I have fixed the script to now use
"Uintah::" correctly. Additionally, I have removed a lot of namespace
polluting "using namespaces" from around the code.

CCA/Components/Arches/Arches.cc
CCA/Components/Arches/CompDynamicProcedure.h
CCA/Components/Arches/IncDynamicProcedure.h

- Clean up due to "using namespace std" being removed from included .h file.

CCA/Components/Arches/Radiation/DORadiationModel.cc
CCA/Components/Arches/Source.cc

- Clean up due to "using namespace std" being removed from included .h file.
- Some better indentation.
- Put all _fort.h includes next to each other.

CCA/Components/Arches/water/watertp.h
CCA/Components/Models/FluidsBased/ArchesTable.cc
CCA/Ports/SFC.h
Core/Exceptions/InvalidCompressionMode.h
Core/Exceptions/InvalidValue.h
Core/Exceptions/ProblemSetupException.h
Core/Exceptions/UintahPetscError.h
Core/IO/SpecializedRunLengthEncoder.h
Core/Math/Short27.h

- Remove polluting "using namespace" from .h file.

CCA/Components/LoadBalancers/CostModelForecaster.cc
CCA/Components/LoadBalancers/CostModelForecaster.h

- Indentation, spacing.
- Use 'd_' for a class variable to be consistent with rest of class.

Core/Grid/Variables/Variable.cc
StandAlone/Benchmarks/SimpleMath.cc
StandAlone/tools/pfs/rawToUniqueGrains.cc

- Move "using namespace" into the .cc file as it was removed from the .h files.

tools/fspec.pl.in

- Make the perl script include the "Uintah::" for Uintah
variables so that a "using namespace Uintah" statement is not
required.

106 lines of code changed in 28 files:

jthornoc 2013-10-08 13:49 Rev.: 51040

Stub in advanced clipping for Ben.

82 lines of code changed in 3 files:

tsaad 2013-10-08 12:22 Rev.: 51038

Allow users to specify BC values via a value attribute instead of an xml child. This will help reduce the number of lines in the BC specification by up to 66%. Based on conversation with Todd and Jeremy, this commit is fully backward compatible to smooth the transition. You don't have to change any of your input files at the moment. if you want to adopt the new specification, see my next commit for examples. If you are a user, please consult with your component developer for possible implications of using this new interface. A script is being developed to help you convert your input files into using the new format.

99 lines of code changed in 5 files:

beckvermit 2013-10-08 11:44 Rev.: 51037

changes to puda script

0 lines of code changed in 2 files:

harman 2013-10-07 16:38 Rev.: 51034

More work on adding variance to the inlet velocity profile.

175 lines of code changed in 9 files:

harman 2013-10-07 16:37 Rev.: 51033

added printTask() and printSchedule() calls to each task

83 lines of code changed in 2 files:

jthornoc 2013-10-07 13:24 Rev.: 51032

Cleaning out the last of the dummies in Arches.

34 lines of code changed in 67 files:

jthornoc 2013-10-07 13:24 Rev.: 51031

Check for clipping before performing the grid loop to clip.

13 lines of code changed in 1 file:

qymeng 2013-10-04 15:45 Rev.: 51029

add new union type Int3 uInt3 Double3 Float3, so that it can be accessed by dimension index.

95 lines of code changed in 2 files:

awabboud 2013-10-04 14:02 Rev.: 51027

Fix complier warning.

1 lines of code changed in 1 file:

awabboud 2013-10-04 13:25 Rev.: 51026

Make a change to the kinetic growth calculation. Add in a parameter to force growth coefficents to zero based on supersaturation values.

46 lines of code changed in 4 files:

harman 2013-10-04 12:41 Rev.: 51025

Now adding a 'kick' or variance to the inlet velocity profile. The kick is based off of previous
researchers turbulent kinetic energy profiles. This is only applied to vel_CC in this version.

82 lines of code changed in 2 files:

qymeng 2013-10-04 11:11 Rev.: 51023

add const GPU Grid Variable support

70 lines of code changed in 8 files:

jsutherland 2013-10-03 18:25 Rev.: 51022

- silence some compiler warnings by removing some inconsequential "const" usage
- formatting

8 lines of code changed in 2 files:

harman 2013-10-03 16:49 Rev.: 51021

Moved MersenneTwister.h to Core/Math so others can use it.

449 lines of code changed in 3 files:

jsutherland 2013-10-03 12:40 Rev.: 51020

Runtime ghosts in Wasatch.
As of now, this still results in hard-coded ghost cells per field - no variation is allowed. It is the first step toward variable ghosts per task though.

83 lines of code changed in 5 files:

jsutherland 2013-10-03 12:40 Rev.: 51019

Silence a type comparison compiler warning.

2 lines of code changed in 1 file:

jsutherland 2013-10-03 11:21 Rev.: 51017

Cleaning up some namespace issues.

4 lines of code changed in 2 files:

tsaad 2013-10-03 11:04 Rev.: 51016

allow field clipping on transported variables to be applied at the end of the timestep. This will be useful for visualization. Here's an example of how to use this:
<FieldClipping tasklist="post_processing">
<FieldExpression type ="SVOL" min="0.0" max="1.0">
<NameTag name="rhof" state="STATE_NP1"/>
</FieldExpression>
</FieldClipping>

20 lines of code changed in 2 files:

jsutherland 2013-10-03 10:48 Rev.: 51015

More warning fixes (hopefully this one works better).

4 lines of code changed in 1 file:

jsutherland 2013-10-03 10:44 Rev.: 51014

Minor formatting changes.

38 lines of code changed in 1 file:

jsutherland 2013-10-03 10:43 Rev.: 51013

Clean up a compiler warning.

4 lines of code changed in 1 file:

awabboud 2013-10-03 10:18 Rev.: 51012

Add in a kernel for dissolution death of a precipitate, and modify the reaction extent source term to refelct this kernel and remove an small chunk of obsolete code.

297 lines of code changed in 3 files:

dav 2013-10-02 18:00 Rev.: 51011

Apparently inside a makefile, the regular expression for a . has to be in [].

5 lines of code changed in 1 file:

dav 2013-10-02 15:56 Rev.: 51010

Include list of duplicate libs in error message.

13 lines of code changed in 1 file:

dav 2013-10-02 15:43 Rev.: 51009

libz.so is too specific... perhaps this way...

2 lines of code changed in 1 file:

dav 2013-10-02 15:23 Rev.: 51005

Need to be more specific then 'libz'... trying 'libz.so'. Also, some shells do not like negative exit numbers so use 1.

4 lines of code changed in 1 file:

dav 2013-10-02 14:42 Rev.: 51003

This is an update to try and catch the problem when multiple
(different) versions of zlib are linked into the same sus... This
mostly has occurred on the Mac, though it seems like it could also
happen under linux. Usually, sus, one of the Uintah .so's (or .dylibs
on a Mac), or a thirdparty lib that sus links against, somehow picks
up a different libz. Then, depending on which library, etc, calls for
compression, potentially a different version of libz is used during
the same execution. This occurs (only?) when saving UDAs with gzip
turned on, and unfortunatley is not detected until someone tries to
read the UDA back in... at which point the program crashes.

I have added two fixes that will hopefully help solve this problem.

1) Makefile.in

The Makefile will now attempt to verify that sus and the Uintah
libs only link vs a single libz. I have tested this on 'fake'
examples, but don't have a version of the code where the real problem
manifests, so am not 100% sure that it is correct.

2) Core/Grid/Variables/Variable.cc

The error manifests itself when the size of the compressed data is
read in and an invalid number is seen (usually something HUGE). The
committed fix is a hack as we can't be sure exactly how big the
invalid number will be, but this should be ok. Currently I am using
'1,000,000,000'. I don't believe a single variable for a single patch
would ever be this big. Note, the 'fix' only reports a potential
error when this occurs... though in theory if it is an error, then a
segfault will happen directly after the user notification.

46 lines of code changed in 2 files:

harman 2013-10-02 10:21 Rev.: 51001

More descriptive printf statements when there's an error.

Formatting
- consistent indentation, added horizontal/vertical spaces.

149 lines of code changed in 2 files:

harman 2013-10-02 08:49 Rev.: 50997

Removed old code that was part of SCIRun and is not being used.

5 lines of code changed in 9 files:

awabboud 2013-10-01 15:14 Rev.: 50996

Add an if statement to check for moment closure expression registration. This will allow multiple growth terms to be used for a PBE.

9 lines of code changed in 1 file:

harman 2013-10-01 12:17 Rev.: 50995

Global:
- cleanup
- formatting
- added more scratch labels to aid in debugging (L_ij, M_ij, S_ij...etc)
These are controlled by #ifdef conditionals.

DynamicModel:
- fixed several multipatch bugs.
- added new method for setting boundary conditions. With some of the filtered
quantities the array size is bigger than the patch and you must set the BCs
in the ghost cells.

multi-patch impVortex == 1 patch impVortex

298 lines of code changed in 15 files:

September 2013 »

Generated by StatSVN 0.7.0