incorporated suggestion for a header
29 lines of code changed in 3 files:
removed the ui callback
0 lines of code changed in 1 file:
Peg Uintah to specific versions of W3P.
This will allow SVN to roll-back and pick up the appropriate W3P libs consistent with that version of Uintah.
Note: this will only work from this point in time forward, as I will maintain the git hashes as needed. For backward compatibility, you will need to manually enter the git hashes as appropriate for W3P.
21 lines of code changed in 1 file:
- Fix bounding box issue for EllipsoidGeometryPiece
12 lines of code changed in 3 files:
Fixing the soot radiation property model so that it doesn't reset the gas absorbtion coefficient. Oops!!
0 lines of code changed in 2 files:
added command callback for the custom UI
7 lines of code changed in 2 files:
Rename reg to regridder. White space.
21 lines of code changed in 1 file:
Update the BC logic so that the BCs do the right thing when using fast_cpdi.
Update bulletproofing in SerialMPM.
26 lines of code changed in 3 files:
Cleaned up some whitespace
3 lines of code changed in 1 file:
Uncommented ASSERT
1 lines of code changed in 1 file:
Fixed linking error
334 lines of code changed in 20 files:
Added capability to have material specific models for computing conductivity values based on concentration
655 lines of code changed in 16 files:
changed the sus format output cleanup the wall times calculations
395 lines of code changed in 3 files:
white space clean up
1 lines of code changed in 1 file:
removed errant subtraction of the bytes/sec measure
0 lines of code changed in 2 files:
changed var name for clarity
0 lines of code changed in 6 files:
changed the wall time and delta T to a table
332 lines of code changed in 11 files:
removed errant subtraction of the bytes/sec measure
15 lines of code changed in 1 file:
Added a method to help us work in L-shaped domains with less confusion
0 lines of code changed in 4 files:
Clamp down on synchronization in IntrusionBC::setCellType. This fixes a race condition occuring with the 8-corner boiler case at high patch/core counts during multi-threaded init timestep.
0 lines of code changed in 2 files:
Replaced the geometry based EllipsoidGeometryPiece instantiation and boundary check with a linear algebra based one.
Ellipsoids with arbitrary axes vector directions now work properly.
84 lines of code changed in 2 files:
Forgot to remove a cout statement
0 lines of code changed in 1 file:
This fixes a bug determining fine level patches from coarse level patches. The bug seemed to only manifest itself in an L-shaped domain. This fix is explicit where the prior code had a fudge factor.
20 lines of code changed in 1 file:
This allows for non-packed MPI messages to work correctly. Specifically if multiple items are being sent in a single message.
12 lines of code changed in 3 files:
Punting, assuming for now that the input is specifying 2D ellipses that are in
the Z-plane.
14 lines of code changed in 1 file:
Cleaned up the error message and got the error check on both the send and the receive
6 lines of code changed in 1 file:
A defensive error check to catch errors I see if packing is turned off.
4 lines of code changed in 1 file:
Mercy! Fixing (I hope) round off issues that I didn't see before.
15 lines of code changed in 1 file:
Converging on getting the EGP to work properly.
12 lines of code changed in 3 files:
Additional fix to this code.
3 lines of code changed in 1 file:
Changed abs to fabs, atan to atan2, but most importantly, added some fabs in
the bounding box code.
30 lines of code changed in 1 file:
Use d_ naming convention. Default <PIDX> tag vars to avoid segfaults. White space.
0 lines of code changed in 4 files:
add check for null return for the map find operator
0 lines of code changed in 2 files:
Committed the wrong things last time. This is the correct set of stuff to commit
4 lines of code changed in 1 file:
Fixes an ROI extents issue. This was tested to help get the BSF_rmcrt.ups Titan run to proceed past the errors on timestep 1.
8 lines of code changed in 1 file:
Bulletproofing now works regardless of whether it is cubic or non-cubic. We want these error messages being thrown whenever they pop up
0 lines of code changed in 2 files:
Just some minor cleanup on this computeVariableExtentsWithBoundaryCheck()
10 lines of code changed in 1 file:
added bulletproofing to catch non-physical values of abskg, sigmaT4 and cellType.
71 lines of code changed in 5 files:
Added new method for computing VariableExtents. It defaults to the standard computeExtents unless the computationnal domain is non-cubic and
the the number of ghostcells requested is > 5.
Brad & Todd
81 lines of code changed in 3 files:
fixed boolean chaeck and added modified var mapper for saving vars in the timestep.xml file
147 lines of code changed in 24 files:
Added new ESConductivityModel to compute the conductivity values at cell face centers based on concentration values of particles in neighboring cells. ESConductivityModel will act as a base class that can be extended to new conductivity models.
276 lines of code changed in 5 files:
added physical location where a NaN or inf is detected.
1 lines of code changed in 1 file:
added the ability of the user to init the simulation but wait until VisIt is connected to run
108 lines of code changed in 7 files:
Added adjustment factor for fluxes. This is needed for the new quadrature set, Legenre-Chebyshev. This will change the Fluxes for all radiation tests. In most cases the Diff should be isolated to the fluxes.
0 lines of code changed in 4 files:
removed debugging cout
1 lines of code changed in 1 file:
more descriptive exception message(s) when multiple level boxes and multiple levels don't align.
0 lines of code changed in 2 files:
moved the increment and time setting to the top of the loop
10 lines of code changed in 1 file:
Make fast_cpdi an option for interpolators.
8 lines of code changed in 1 file:
Now compiling fastCpdiInterpolator. Seems to work...
353 lines of code changed in 4 files:
moved first check to before adjustDelT
5 lines of code changed in 1 file:
Add new test cases to nightlyRT.
4 lines of code changed in 1 file:
Progress towards bringing the fast implementation up to date.
39 lines of code changed in 2 files:
clean up the current time incrementing, delT calculation, and sim controller reporting
210 lines of code changed in 7 files:
These may not work yet, but at least they'll compile now if added to sub.mk.
12 lines of code changed in 2 files:
In previous commit, forgot to comment out corresonding computes and requires.
2 lines of code changed in 1 file:
Adding commented out code to project particle color to grid. This has
occasional utility, but not enough to tolerate the additional work.
10 lines of code changed in 1 file:
insideBoxes_ups()
- removed flawed logic.
1 lines of code changed in 1 file:
These are changes needed to couple the current soot model and the radiation coefficients.
220 lines of code changed in 3 files:
Update configure to remove Wasatch .o files when W3P changes (need to rebuild configure but wanted to get tihs in).
11 lines of code changed in 2 files:
Store the level boxes specified in the ups file in a vector<BBox>. This vector is searched
if an RMCRT ray is going out of bounds.
Added <Grid>
<nonCubic_domain> bool </nonCubic_domain>
</Grid>
Enabled bulletproofing in RMCRT in the bounding ray approach.
104 lines of code changed in 4 files:
Logic was backwards for the intrusion deletion when using multiple
patches/mpi-rank.
19 lines of code changed in 1 file:
Fix needed so that concentration reduction variables only are computed for specified materials.
30 lines of code changed in 6 files:
cleanup:
- added dividers
- horizontal and vertical spacing
- more descriptive exception messages
116 lines of code changed in 1 file:
Add a new constraint option for Friction Contact. Currently commented out, not
ready to deal with breaking the RT just yet.
54 lines of code changed in 4 files:
Added ability to build NSCBC library.
33 lines of code changed in 4 files:
Add in the Wasatch 3P NSCBC library build/check in configure system. If you are using Wasatch and pointing
to old Watach 3Ps, this will cause your build to break... you will need to update your W3P to include
the NSCBC library.
M build_scripts/build_wasatch_3p.sh
- Added in Tony Saad's build commands for NSCBC library. Added hooks to allow configure to
specify whether to build it or not.
M configVars.mk.in
- Added in HAVE_NSCBC and NSCBC_INCLUDE flags.
M configure.ac
M configure
- Order Wasatch 3P lines to always be SpatialOps, ExprLib, TabProps, RadProps, NSCBC, POKITT.
- Add in verification of NSCBC files and NSCBC flags.
- Re-ordered the Pokitt test to be after the NSCBC test (just to keep using the same order when looking W3P libs).
M CCA/Components/SimulationController/AMRSimulationController.cc
- Few layout cleanups for easier reading.
M CCA/Components/SimulationController/SimulationController.cc
- Rename 't' to 'time'. Please don't use one letter variable names.
- Commented out "stdDeviation()" function as it isn't used and causes compiler warnings.
M Core/DataArchive/DataArchive.cc
- Fix error message when looking for <Data> to accurately report the name of the file that was searched.
75 lines of code changed in 6 files:
initalize d_prev_delt and made adjustDelT more readable
36 lines of code changed in 2 files:
Added linearInterpolation for stencils. Not tested at this point.
0 lines of code changed in 2 files:
Tweaking Todd's bulletproofing so that it only looks within cells within patches and not cells in the entire level
19 lines of code changed in 1 file:
added bulletproofing in rayTrace and updateSumI to catch Nans or infs floating around.
50 lines of code changed in 2 files:
Update to conform to new ExprLib API.
29 lines of code changed in 45 files:
updateSumI()
- if disMin is non-physical ~ -1e-15 then add some numerical fuzz 1e-12.
Removed exception about calc frequency and gpus.
7 lines of code changed in 2 files:
Just adding a clearer error message
1 lines of code changed in 1 file:
Cleaned up memory leak issue associated with two new VarLabels not being destroyed.
2 lines of code changed in 2 files:
Updated AutoFluxBC so that user can define setpoints based on min and max values or average.
73 lines of code changed in 6 files:
Fix the W3P build script for GPU builds of RadProps.
1 lines of code changed in 1 file:
Enable GPU calculation with RadProps (in Wasatch).
Changes to RadProps consistent with API changes made in conjunction with this commit.
49 lines of code changed in 15 files:
M CCA/Components/DataArchiver/DataArchiver.cc
- Comment end of function.
M CCA/Components/SimulationController/AMRSimulationController.cc
- Set a var ('checkpointing') to remember if this is a checkpoint so we don't have to do the check twice.
The previous code was problematic as it used an individual rank's currentSeconds() and we
were doing the check (for the same timestep) twice with different wall times.
4 lines of code changed in 2 files:
Fix a nasty synchronization bug in checkpointing based on wall clock time.
This, I believe, mainly only shows up with very quick time steps, but it
could, conceivably, show up 'randomly' any time. The problem was that
each rank was determining "what the wall clock time is"... And each processors
clock could be off, or the check could be just a micro second delayed,
causing one core to not think that it is time for a checkpoint, and all
others to try to checkpoint, thus causing a hang (or segfault).
It would appear that this issue has been in the code for a long time.
In almost (?) all cases where real time is used, the time is found by
rank 0 and broadcast to everyone else. However, (at least) in this case
all ranks were calling currentSeconds(). I believe that I have found
the two locations this was happening and fixed it, and want to get this
fix committed for Sid to try and so that Ben's runs can be attempted.
However, it would behoove us to actually go through this code in detail
to make sure that we haven't missed anything and that redundant work
is not being performed. Also, I don't think that the broadcast will
be a major problem, but I'd like to get Alan's (et al) thoughts on this.
M CCA/Components/DataArchiver/DataArchiver.cc
- Removed tabs, Minor white space clean ups. Use > 80 columns of code where it makes sense.
Be consistent with placement of '{' in for loops, and use of else on a new line.
- d_checkpointWalltimeInterval is an integer, though there were (are?) a
few places it was being treated like a float. Probably didn't make a
difference, but we should be consistent.
- Have rank 0 determine the time and bcast it if we are doing wall clock
check pointing.
M CCA/Components/SimulationController/AMRSimulationController.cc
- Have rank 0 determine the time and bcast it if we are doing wall clock
check pointing.
52 lines of code changed in 2 files:
Maintain list of spawned threads. This way we can do RAII style management of thread join/detach.
0 lines of code changed in 2 files:
added debug streams
8 lines of code changed in 2 files:
reset index
1 lines of code changed in 1 file:
changed to Port
1 lines of code changed in 1 file:
This commit allows for dumping the grid.xml file in binary. It is turned off by default at this point.
Additionally, the <Data> section is split out of the grid.xml file and put into its own file (XML/ASCII) -
but only when using the XML_TEXTWRITER or the binary output.
If the grid is output in binary, you can use the "puda/grid_reader.cc" stand alone program to
display it (in XML) so you can visually see what is in the file. It is important to note that
when outputing in XML, we can (and do) ignore (not write) tags that are meaningless (in certain situations)
or have a default when they do not exist in the file. However, in the binary version, we have to write them all
out. I have tested against our local regression tests, but it is possible that there are some
seldomly used flags that are not yet written to the binary grid.xml file that we will need to take
care of adding in if/when they come up.
M CCA/Components/DataArchiver/DataArchiver.h
M CCA/Components/DataArchiver/DataArchiver.cc
- Finished implementing the writeGridBinary() function.
- Drop out of writeto_xml_files() immediately if not an I/O timestep. Previously it didn't
actually do anything, but it was not clear that it was not doing anything.
- Currently, to turn on binary output of grid.xml, you need to edit the code and
uncomment the writeGridBinary() call (and comment out the writeGridOriginal()).
Hopefully soon this will just be the default so we can get rid of the multiple
cases and extra code.
- Create a shared writeDataTextWriter() so that both the XML_TEXTWRITER and binary version
can share the code to dump the data.xml file.
- procOnLevel is a boolean array, so treat it as such.
- Swat some endls that take a lot more typing then just a \n.
- White space for better readability, update some formatting for consistency.
M CCA/Components/SimulationController/AMRSimulationController.cc
- Indentation is your friend.
- Remove tabs, line up code, white space.
M CCA/Components/SimulationController/SimulationController.cc
- Move local stdDeviation() routine to just above where it is used. Make static.
- Swat endls.
- White space for readability.
- Remove end of lines from exception messages where they are not needed.
M Core/DataArchive/DataArchive.h
M Core/DataArchive/DataArchive.cc
- Alphabetize #includes.
- Add a constant magic number to designate binary grid.xml files. Used by the code to verify
which version (xml vs binary) is being read. GRID_MAGIC_NUMBER
- Remove the "ref_cnt" and "lock" class variables as it appears they aren't used, and they don't
do anything. I believe they are left over vestiges of days long past.
- Remove "d_grid_path_and_filename" as all we need is the path and that is already stored.
- White space and consistent indentation.
- Added code to read in the new binary grid.xml files.
- Removed a huge "#if 0" section. Most likely left over from a previous version of UDA reading.
It would be good to converge on a single method and only have one set of code...
M Core/Grid/Grid.h
M Core/Grid/Grid.cc
- Add in readLevelsFromFileBinary() code to support DataArchive reading of new binary grid.xml.
As you can see from the (short) size of this routine, reading the binary version is much
simpler (especially compared to the streaming (non-DOM based) xml parsing that we also
currently do).
- Make use of longer lines as we non-longer are limites to 80 columns - where it makes sense.
- Add "{" to single line if statements as per coding standard.
- Remove some debug statements.
- Line up code for easier reading.
M Core/Grid/Level.cc
- White space / line up code for easier reading.
M Core/Grid/SimulationTime.cc
- White space / line up code for easier reading.
- Removed the deprecated "<max_iterations>" check. It hasn't been used in a long time.
M StandAlone/tools/puda/grid_reader.cc
- New, stand alone program to read a binary grid.xml file and display a human readable XML version.
- If you update this code, make sure it matches the writing in DataArchiver and the reading in
DataArchive.
M StandAlone/tools/puda/sub.mk
- Build grid_reader.
665 lines of code changed in 16 files:
ESMPM now couples amrmpm to the finite volume electrostatic solver.
238 lines of code changed in 8 files: