Commits · 558b3636d2187b6ff5eda706ddcdb54de478a020 · tsoc / openmm

24 Mar, 2015 1 commit
- Fixed a race condition in OpenCL · 558b3636
  Peter Eastman authored Mar 24, 2015
  
  558b3636
19 Mar, 2015 1 commit
- Simplification to reference platform tests · f66c9b61
  peastman authored Mar 19, 2015
  
  f66c9b61
17 Mar, 2015 1 commit
- CUDA and OpenCL implementations of CMAPTorsionForce::updateParametersInContext() · 6dc2f9a8
  peastman authored Mar 16, 2015
  
  6dc2f9a8
16 Mar, 2015 1 commit
- Added updateParametersInContext() to CMAPTorsionForce, and implemented in reference platform · 474bbf7a
  peastman authored Mar 16, 2015
  
  474bbf7a
10 Mar, 2015 2 commits
- Fixed a performance regression in multi-GPU on CUDA · 9d3a655b
  peastman authored Mar 10, 2015
  
  9d3a655b
- More bug fixes to multi-GPU · f7ef2dd0
  peastman authored Mar 10, 2015
  
  f7ef2dd0
06 Mar, 2015 2 commits
- For multi-GPU, exceeding the size of the neighbor list causes the forces to be marked as invalid · bd666c27
  peastman authored Mar 06, 2015
  
  bd666c27
- Load balancing between GPUs forces the neighbor list to be rebuilt · 62f7c44c
  peastman authored Mar 06, 2015
  
  62f7c44c
24 Feb, 2015 1 commit
- Fixed problems running in PNaCl · aa7bd1cf
  peastman authored Feb 24, 2015
  
  aa7bd1cf
21 Feb, 2015 2 commits
- Deleted obsolete debugging code · f72062af
  peastman authored Feb 20, 2015
  
  f72062af
- Combined subfolders in reference source, and renamed a couple of badly named classes · 3d69185a
  peastman authored Feb 20, 2015
  
  3d69185a
20 Feb, 2015 4 commits
- Elimininated obsolete SimTKOpenMMCommon files · 0671c7b0
  peastman authored Feb 20, 2015
  
  0671c7b0
- Cleaned up lots of formatting to be more consistent with the rest of OpenMM · 7fb10336
  peastman authored Feb 20, 2015
  
  7fb10336
- Deleted obsolete logging class · 1d3ffd7b
  peastman authored Feb 20, 2015
  
  1d3ffd7b
- Moved reference platform classes into OpenMM namespace · 9111cca0
  peastman authored Feb 20, 2015
  
  9111cca0
19 Feb, 2015 2 commits
- Fixed some test cases that were failing · 1ce17590
  peastman authored Feb 19, 2015
  
  1ce17590
- Minor changes to speed up some test cases · f2958043
  peastman authored Feb 18, 2015
  
  f2958043
17 Feb, 2015 1 commit
- Workaround for Nvidia bug on Maxwell in CUDA 6 · bfe7570e
  peastman authored Feb 17, 2015
  
  bfe7570e
13 Feb, 2015 3 commits
- Custom expressions support floor() and ceil(). · dce6a437
  peastman authored Feb 13, 2015
  
  dce6a437
- Cleanup to CUDA compiler plugin · 8fa08870
  Peter Eastman authored Feb 13, 2015
  
  8fa08870
- Created plugin for CUDA runtime compiler · 7e29671a
  Peter Eastman authored Feb 13, 2015
  
  7e29671a
06 Feb, 2015 1 commit
- Bug fixes to triclinic boxes · 050e1262
  Peter Eastman authored Feb 05, 2015
  
  050e1262
20 Jan, 2015 1 commit
- Check for nan positions was being done incorrectly · f75bf1bd
  peastman authored Jan 20, 2015
  
  f75bf1bd
17 Jan, 2015 1 commit
- Variable step size integrators were scaling the step size incorrectly · bd01a379
  peastman authored Jan 16, 2015
  
  bd01a379
15 Jan, 2015 1 commit
- Catch nan coordinates before they can produce a segfault · 7dc2ff9e
  peastman authored Jan 15, 2015
  
  7dc2ff9e
14 Jan, 2015 1 commit
- Removed some unneeded code · 3b2579a5
  Peter Eastman authored Jan 14, 2015
  
  3b2579a5
12 Jan, 2015 2 commits
- Updated test cases to catch a bug that was just fixed · 5ae844b2
  peastman authored Jan 12, 2015
  
  5ae844b2
- Fixed bug in reference CustomCompoundBondForce · 29856d18
  peastman authored Jan 12, 2015
  
  29856d18
10 Jan, 2015 1 commit
- Change comments so that it does not reference generating seeds from the clock. · d422c306
  Jason Swails authored Jan 10, 2015
  
  d422c306
09 Jan, 2015 2 commits

Fix an include. · 118aab89
Jason Swails authored Jan 09, 2015

118aab89

Convert a random seed set to 0 into one based on the wall clock in more places. · 601a696f

Jason Swails authored Jan 09, 2015

This commit hits the CPU platform (I had previously thought it just used the
Reference generator) as well as the CustomIntegrator stream that is set
separately from the standard "integrator" stream attached to each platform for
CUDA and OpenCL platforms (it seemed to me that the Reference platform's
CustomIntegrator just used the integrator random stream).

601a696f

08 Jan, 2015 2 commits
- Adjusted the tolerance on a test that was failing on some computers · 38acdd6e
  Peter Eastman authored Jan 08, 2015
  
  38acdd6e
- OpenCL supports PME with triclinic boxes · 295ae001
  Peter Eastman authored Jan 07, 2015
  
  295ae001
07 Jan, 2015 3 commits
- Continuing OpenCL support for triclinic boxes · 9bcba51e
  Peter Eastman authored Jan 07, 2015
  
  9bcba51e
- Began implementing OpenCL support for triclinic boxes · 195bcecf
  Peter Eastman authored Jan 07, 2015
  
  195bcecf
- Fixed some methods that should have been marked const · 17c06daa
  Peter Eastman authored Jan 07, 2015
  
  17c06daa
06 Jan, 2015 2 commits
- CUDA support for PME with triclinic boxes · e7eb7852
  Peter Eastman authored Jan 06, 2015
  
  e7eb7852
- Continuing to implement CUDA support for triclinic boxes · 6eb302fb
  Peter Eastman authored Jan 05, 2015
  
  6eb302fb
05 Jan, 2015 2 commits
- Beginning of CUDA support for triclinic boxes · 61458a5f
  Peter Eastman authored Jan 05, 2015
  
  61458a5f
- Attempt to move the implementation of separate osrngseed() calls for each · e04a7368
  Jason Swails authored Jan 05, 2015
```
Context into the platform rather than hacked into getRandomNumberSeed...

Someone please check this.
```
  e04a7368