/trunk/src Developers: bpeterson

Activity by Clock Time

Activity in Directories

Most Recent Commits

bpeterson 2016-11-19 14:14 Rev.: 55998

Seems the trunk GPU RMCRT code had the fixed random number flag turned on. Fixing that by turning it back off

0 lines of code changed in 2 files:

• CCA/Components/Models/Radiation/RMCRT: RayGPUKernel.cu (changed)

bpeterson 2016-11-19 14:12 Rev.: 55997

Tweaking the code so float support for GPU RMCRT works again on data onion

10 lines of code changed in 1 file:

• CCA/Components/Models/Radiation/RMCRT: RayGPUKernel.cu (+10 -2)

bpeterson 2016-11-13 09:05 Rev.: 55971

This is a fix for Todd. Derek has been running this fix for the past two weeks and it's worked for him too. It just allows for extra cells to be involved in the coarsening process.

0 lines of code changed in 2 files:

• CCA/Components/Schedulers: TaskGraph.cc (changed)

bpeterson 2016-10-24 02:47 Rev.: 55907

Added a method to help us work in L-shaped domains with less confusion

0 lines of code changed in 4 files:

• Core/Grid: Level.cc (changed), Level.h (new)

bpeterson 2016-10-22 02:31 Rev.: 55904

Forgot to remove a cout statement

0 lines of code changed in 1 file:

• Core/Grid: Patch.cc (-1)

bpeterson 2016-10-22 02:17 Rev.: 55903

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:

• Core/Grid: Patch.cc (+20 -20)

bpeterson 2016-10-21 02:17 Rev.: 55902

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:

• CCA/Components/Schedulers: MPIScheduler.cc (+1 -1)
• Core/Parallel: BufferInfo.cc (+11 -5)

bpeterson 2016-10-20 17:48 Rev.: 55900

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:

• CCA/Components/Schedulers: MPIScheduler.cc (+6 -2)

bpeterson 2016-10-20 17:23 Rev.: 55899

A defensive error check to catch errors I see if packing is turned off.

4 lines of code changed in 1 file:

• CCA/Components/Schedulers: MPIScheduler.cc (+4 -1)

bpeterson 2016-10-16 21:20 Rev.: 55892

Committed the wrong things last time. This is the correct set of stuff to commit

4 lines of code changed in 1 file:

• CCA/Components/Models/Radiation/RMCRT: Ray.cc (+4 -23)

bpeterson 2016-10-16 01:17 Rev.: 55891

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:

• CCA/Components/Models/Radiation/RMCRT: Ray.cc (+8 -1)

bpeterson 2016-10-14 20:33 Rev.: 55890

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:

• CCA/Components/Schedulers: OnDemandDataWarehouse.cc (new)

bpeterson 2016-10-14 20:12 Rev.: 55889

Just some minor cleanup on this computeVariableExtentsWithBoundaryCheck()

10 lines of code changed in 1 file:

• Core/Grid: Patch.cc (+10 -5)

bpeterson 2016-10-08 05:17 Rev.: 55850

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:

• CCA/Components/Models/Radiation/RMCRT: Ray.cc (+19 -17)

bpeterson 2016-10-06 16:33 Rev.: 55846

Just adding a clearer error message

1 lines of code changed in 1 file:

• CCA/Components/Schedulers: OnDemandDataWarehouse.cc (+1 -1)

bpeterson 2016-09-30 10:03 Rev.: 55818

A bug fix to make sure the random ray ID # is never more than the amount of total rays

14 lines of code changed in 1 file:

• CCA/Components/Models/Radiation/RMCRT: RayGPUKernel.cu (+14 -7)

bpeterson 2016-08-04 10:42 Rev.: 55571

A fix for an RMCRT/Arches scenario in which a single dependency source was needed for multiple patches at the destination.

169 lines of code changed in 1 file:

• CCA/Components/Schedulers: UnifiedScheduler.cc (+169 -161)

bpeterson 2016-08-02 17:15 Rev.: 55565

This patch has two parts. The first part changes the PerPatch variable so that uses C++11 shared_ptr so that it never makes a copy of the value, it instead preserves a pointer to a single value. This helps Wasatch who needs pointers to direct data. I've run this file against the full gold standard tests without issue. The others are some sizing fixes to the UnifiedScheduler knows how many PerPatch bytes to copy, before it was copying 0 bytes.

27 lines of code changed in 6 files:

• CCA/Components/Schedulers: GPUGridVariableInfo.cc (+5 -4), GPUGridVariableInfo.h (new), UnifiedScheduler.cc (+8 -1)
• Core/Grid/Variables: PerPatch.h (new)

bpeterson 2016-08-02 16:15 Rev.: 55564

Some fixes to ensure PerPatch and Reduction variables go into the GPU properly and with concurrency checks

100 lines of code changed in 1 file:

• CCA/Components/Schedulers: UnifiedScheduler.cc (+100 -94)

bpeterson 2016-07-26 16:35 Rev.: 55542

I'm starting work on a Unified Data Warehouse. This file is just here to get it in the repository for insurance purposes. It won't be showing up in any sub.mk anytime soon.

520 lines of code changed in 1 file:

• CCA/Components/Schedulers: UnifiedDataWarehouse.h (new 520)

(30 more)