GitLab

Fixes #5001.
Signed-off-by: Yannick Lemke <yannick.lemke@posteo.de>

* Optimize CustomNonbondedForce.updateParametersInContext()

* Optimized uploading changed values to GPU

* Optimized updateParametersInContext() for lots of bonded forces

* Optimized updateParametersInContext() for CustomExternalForce

* Optimized updateParametersInContext() for NonbondedForce

* Code changes for HIP platform

* Reference implementation of computed values for CustomNonbondedForce

* CPU implementation of computed values for CustomNonbondedForce

* Common implementation of computed values for CustomNonbondedForce

* Serialization of computed values

* ForceField supports computed values

- Don't make a Copy method on FunctionInfo
- Have the Copy() methods on the TabulatedFunction subclasses return a pointer

necessitated creating a Copy method to do that on TabulatedFunction classes.

segfaults when you use the long-range dispersion correction with the
CustomNonbondedForce if the force uses TabulatedFunction objects. Basically, it
prevents the copy from taking ownership (and deleting) the functions in its
destructor.

The use cases for CustomNonbondedForce copying currently is to recompute the
long-range prefactor if the parameters change (so the CustomNonbondedForceImpl
kernels need a copy of the Force object). Doing a full (deep) copy would be a
bit slower (although would only need to be paid once for each
CustomNonbondedForce object being used).

Renamed methods for dealing with TabulatedFunctions to avoid problems in wrapper APIs.  Also fixed a few bugs.

Created API and reference implementation of restricting CustomNonbondedForce to selected interaction groups

CustomNonbondedForce.cpp, CudaNonbondedUtilities.cpp and CudaKernels.cpp are now fully reverted to 3478.

LPW attempted bugfixes: (1) Changed unit multipliers in AmoebaVdwForce::getParticleParameters (2) Sorted vectors in exclusionLists so it agrees with atomExclusions

Redesigned CustomNonbondedForce to eliminate combining rules and allow the expression to directly use parameters for individual particles.