/trunk/src Developers: ahumphrey

Login name:
ahumphrey
Total Commits:
392 (3.2%)
Lines of Code:
11,261 (8.2%)
Most Recent Commit:
2015-10-19 16:38
Tweet this:
Tweet this

Activity by Clock Time

Activity by Hour of Day for ahumphrey

Activity by Day of Week for ahumphrey

Activity in Directories

Directory Changes Lines of Code Lines per Change
Totals 392 (100.0%) 11261 (100.0%) 28.7
CCA/Components/Schedulers/ 197 (50.3%) 8981 (79.8%) 45.5
CCA/Components/Models/Radiation/RMCRT/ 29 (7.4%) 1240 (11.0%) 42.7
Core/Containers/ 4 (1.0%) 371 (3.3%) 92.7
Core/Parallel/ 19 (4.8%) 135 (1.2%) 7.1
/ 18 (4.6%) 134 (1.2%) 7.4
CCA/Ports/ 5 (1.3%) 99 (0.9%) 19.8
Core/Grid/Variables/ 4 (1.0%) 93 (0.8%) 23.2
CCA/Components/LoadBalancers/ 8 (2.0%) 45 (0.4%) 5.6
Core/Geometry/ 2 (0.5%) 39 (0.3%) 19.5
Core/Malloc/ 2 (0.5%) 38 (0.3%) 19.0
Core/Grid/ 15 (3.8%) 23 (0.2%) 1.5
CCA/Components/SimulationController/ 4 (1.0%) 17 (0.2%) 4.2
CCA/Components/Regridder/ 3 (0.8%) 16 (0.1%) 5.3
build_scripts/ 3 (0.8%) 7 (0.1%) 2.3
R_Tester/ 1 (0.3%) 7 (0.1%) 7.0
Core/Thread/ 5 (1.3%) 4 (0.0%) 0.8
CCA/Components/Examples/ 1 (0.3%) 4 (0.0%) 4.0
StandAlone/ 4 (1.0%) 3 (0.0%) 0.7
Core/Util/ 6 (1.5%) 2 (0.0%) 0.3
CCA/Components/Arches/ 7 (1.8%) 2 (0.0%) 0.2
CCA/Components/MPM/ConstitutiveModel/UnusedCM/ 2 (0.5%) 1 (0.0%) 0.5
include/sci_defs/ 2 (0.5%) 0 (0.0%) 0.0
include/ 3 (0.8%) 0 (0.0%) 0.0
Core/Basis/ 42 (10.7%) 0 (0.0%) 0.0
CCA/Components/Parent/ 4 (1.0%) 0 (0.0%) 0.0
CCA/Components/MPM/ 1 (0.3%) 0 (0.0%) 0.0
CCA/Components/DataArchiver/ 1 (0.3%) 0 (0.0%) 0.0

Activity of ahumphrey

Most Recent Commits

ahumphrey 2015-10-19 16:38 Rev.: 54196

Make sure and distinguish between ExecTimes, WaitTimes and MPIScheduler_TimingsOut SCI_DEBUG flags. This was missed when this code, common to all schedulers was extracted into a method (outputTimingStats) and put into MPIScheduler.

0 lines of code changed in 2 files:

  • CCA/Components/Schedulers: MPIScheduler.cc (new)
ahumphrey 2015-09-25 16:27 Rev.: 54056

Some cleanup and work toward refining and better understanding MPI timings and stats.
Pull ancient VAMPIR weeds.

134 lines of code changed in 18 files:

  • CCA/Components/Schedulers: DynamicMPIScheduler.cc (+1 -1), MPIScheduler.cc (+14 -26), SchedulerCommon.cc (+2 -2), UnifiedScheduler.cc (+107 -99), sub.mk (+1 -1)
  • Core/Grid: Level.h (new), Patch.h (+1 -1), fixedvector.h (new)
  • Core/Parallel: Vampir.cc (del), Vampir.h (del), sub.mk (new)
  • StandAlone: sus.cc (-8)
ahumphrey 2015-09-17 15:48 Rev.: 53977

Unified scheduler can run multi-threaded on intial timestep. This speeds up the 0th timestep considerably, especially with large problems and particle creation, etc.

4 lines of code changed in 1 file:

  • CCA/Components/Schedulers: UnifiedScheduler.cc (+4 -4)
ahumphrey 2015-09-17 15:39 Rev.: 53976

Consolidate ExecTimes, WaitTimes, and MPI Timing code into parent (MPI) scheduler. This eliminates miles of redundnat and error-prone code. Now it only has to be correct in one place.

162 lines of code changed in 7 files:

  • CCA/Components/Schedulers: DynamicMPIScheduler.cc (+16 -175), MPIScheduler.cc (+132 -130), MPIScheduler.h (new), ThreadedMPIScheduler.cc (new), UnifiedScheduler.cc (+10 -175)
ahumphrey 2015-09-14 17:43 Rev.: 53946

Make the SCI_DEBUG flag "DynamicMPI_TimingsOut" actually output timings files. This is needed for the Sandia frameworks tech report.
Also do some various formatting, cleanup and add some comments.

51 lines of code changed in 3 files:

  • CCA/Components/Schedulers: DynamicMPIScheduler.cc (+34 -24), MPIScheduler.cc (+5 -6), MPIScheduler.h (+12 -4)
ahumphrey 2015-09-11 16:14 Rev.: 53932

Pull some leftover MSC_VER weeds

0 lines of code changed in 10 files:

  • Core/Thread: Thread.cc (new)
  • Core/Util: DynamicLoader.cc (new), Environment.cc (changed)
  • include: sci_hash_map.h (changed)
  • include/sci_defs: compile_testdefs.h.in (changed)
ahumphrey 2015-09-11 14:44 Rev.: 53930

Remove antiquated TAU support.

124 lines of code changed in 57 files:

  • configVars.mk.in (+33 -87), configure (+7 -105), configure.ac (-39), disjointPackageMakefile.in (new)
  • CCA/Components/Arches: CompDynamicProcedure.cc (new), IncDynamicProcedure.cc (new), MomentumSolver.cc (-1), Properties.cc (new)
  • CCA/Components/DataArchiver: DataArchiver.cc (-2)
  • CCA/Components/LoadBalancers: DynamicLoadBalancer.cc (new), LoadBalancerCommon.cc (-3), ParticleLoadBalancer.cc (-3)
  • CCA/Components/MPM: PetscSolver.cc (-12)
  • CCA/Components/MPM/ConstitutiveModel/UnusedCM: CompNeoHookImplicit.cc (+1 -1)
  • CCA/Components/Regridder: RegridderCommon.cc (+1 -4), TiledRegridder.cc (new)
  • CCA/Components/Schedulers: DetailedTasks.cc (new), DynamicMPIScheduler.cc (+2 -63), MPIScheduler.cc (+3 -75), OnDemandDataWarehouse.cc (+4 -8), SchedulerCommon.cc (-15), TaskGraph.cc (+4 -43), ThreadPool.cc (-71), ThreadedMPIScheduler.cc (-17), UnifiedScheduler.cc (+1 -19), sub.mk (+1 -1)
  • CCA/Components/SimulationController: AMRSimulationController.cc (+4 -20)
  • Core/Grid: Grid.cc (+3 -5), Level.cc (+9 -16), Patch.cc (-3), PatchRangeTree.cc (new)
  • Core/Grid/Variables: ParticleVariable.h (new)
  • Core/Parallel: MPI_Communicator.cc (+1 -1)
  • Core/Thread: Thread_pthreads.cc (-4)
  • StandAlone: sus.cc (-21)
  • build_scripts: program.mk (new), so_epilogue.mk (+5 -4)
  • include: TauProfilerForSCIRun.h (del)
ahumphrey 2015-08-13 17:20 Rev.: 53819

Cleanup in Switcher component - for merge with and usage in arl_dev branch.

0 lines of code changed in 4 files:

  • CCA/Components/Parent: Switcher.cc (new), Switcher.h (new)
ahumphrey 2015-07-22 18:49 Rev.: 53752

We can now use the static version of the CUDA runtime lib on Titan.

12 lines of code changed in 2 files:

  • configure (+6 -13), configure.ac (+6 -13)
ahumphrey 2015-07-20 21:59 Rev.: 53743

Following r53735, varLabel name change boundFlux -> RMCRTboundFlux for GPU code.

9 lines of code changed in 1 file:

  • CCA/Components/Models/Radiation/RMCRT: RayGPUKernel.cu (+9 -9)
ahumphrey 2015-06-10 12:19 Rev.: 53618

Eliminate linking with cudart_static with static builds. This is not working on Titan due to the current CUDA Toolkit version, 5.5. Once Titan is upgraded to CUDA 6.5, we can then use cudart_static in static builds.

* generated new configure script

24 lines of code changed in 2 files:

  • configure (+12 -5), configure.ac (+12 -5)
ahumphrey 2015-06-10 10:14 Rev.: 53617

Accurately report specs on all available on-node GPUs.

5 lines of code changed in 1 file:

  • CCA/Components/Schedulers: UnifiedScheduler.cc (+5 -6)
ahumphrey 2015-06-09 16:55 Rev.: 53616

Comment out multi-GPU code that's not working on local SCI cluster. Will fix this soon.

7 lines of code changed in 1 file:

  • CCA/Components/Schedulers: UnifiedScheduler.cc (+7 -7)
ahumphrey 2015-06-09 14:51 Rev.: 53615

Remove ProblemSetupException thrown when trying to use > 1 available GPU per node. We correctly support multiple GPUs.

5 lines of code changed in 1 file:

  • CCA/Components/Schedulers: UnifiedScheduler.cc (+5 -16)
ahumphrey 2015-06-05 08:38 Rev.: 53599

Eliminate compiler warning about comparison between signed and unsigned integer expressions.

1 lines of code changed in 1 file:

  • CCA/Ports: SolverInterface.h (+1 -1)
ahumphrey 2015-06-05 08:36 Rev.: 53598

* Don't pass c++11 flag to NVCC. NVCC from CUDA <=6.0 doesn't support it. We don't use C++11 features in any NVCC compiled code anyway (yet).
* this allows us to run using C++11 features in Uintah code and CUDA 5.5 on Titan
* generate new configure script

* Fix C++11 compiler error about inability to find string literal in Allocator.cc (SCIRun::PrintTag(void*))


46 lines of code changed in 5 files:

  • aclocal.m4 (+4 -1), configure (+4 -1)
  • Core/Malloc: Allocator.cc (new)
ahumphrey 2015-05-19 10:46 Rev.: 53515

Do not throw exception when fine-level halo used in GPU::RMCRT_DO. This is all working properly now.

0 lines of code changed in 1 file:

  • CCA/Components/Models/Radiation/RMCRT: Ray.cc (-8)
ahumphrey 2015-05-18 17:32 Rev.: 53513

Get rid of operator new calls on device. Use compile time constant to size arrays for now.

4 lines of code changed in 1 file:

  • CCA/Components/Models/Radiation/RMCRT: RayGPUKernel.cu (+4 -13)
ahumphrey 2015-05-05 12:34 Rev.: 53426

Cleanup TaskDBG output in threaded schedulers, reporting in ThreadedMPI and Unified schedulers has not been consistent. Also restrict thread-specific debugstream output.

20 lines of code changed in 4 files:

  • CCA/Components/Schedulers: ThreadedMPIScheduler.cc (+12 -11), UnifiedScheduler.cc (+4 -4)
  • Core/Grid: Task.h (new)
ahumphrey 2015-04-24 14:01 Rev.: 53406

Some minor refactoring and debug stream improvements while hunting down the multi-threaded Hypre bug.

13 lines of code changed in 6 files:

  • CCA/Components/Schedulers: DetailedTasks.h (new), DynamicMPIScheduler.cc (+2 -2), MPIScheduler.cc (+1 -1), ThreadedMPIScheduler.cc (+6 -11), UnifiedScheduler.cc (+3 -3)

(65 more)

Generated by StatSVN 0.7.0