GitLab

Thread count for kReduceObcGbsaBornForces_kernel/kReduceGbviBornForces_kernel were too low for some systems/boards

Simplified constraint code.  In the process, this fixed a bug that caused bubbles to appear in water when running on Fermi GPUs.

Refactored ForceImpl, lots of KernelImpl subclasses, and other related classes to avoid redundant calculations when requesting a State with both forces and energies

Fixed bug where MonteCarloBarostat would leave molecules seeming to be outside the first periodic box

Fixed same pressure unit conversion error in MonteCarloBarostatImpl.cpp that occurs in tests.  Tests now pass.

Optimization to nonbonded forces: replaced divides by multiplies when applying periodic boundary conditions