GitLab

This change is intended to improve the behavior when multiple
CUDA contexts are created on a node with multiple CUDA GPUs,
each of which is set to process-exclusive mode.

Converted the routine for identifying molecules from a recursion to a loop.  This avoids stack overflows with very long linear molecules.

Platform specific header files get installed.  This allows plugins to be built with just an OpenMM installation, not a full source tree.

Integrators control when atom reordering gets done, instead of having it happen automatically when forces are evaluated.  This simplifies some code, and also improves unintuitive behavior with CustomIntegrator.

Began implementing new mixed precision model that does integration in double precision and force evaluation in single precision

Continuing to implement new CUDA platform: constraints, LangevinIntegrator, BrownianIntegrator, VariableLangevinIntegrator, VariableVerletIntegrator