GitLab

* Make reference/CPU minimizer into a kernel

* Add per-platform support for GPU minimization

* Initial implementation of GPU minimization

* Fixes

* Increase robustness when initial gradient is huge

* Handle overflow leading to non-finite values gracefully

* Handle large forces in single precision more robustly

* Optimize kernels

* Fix kernel launch size

* Update banner years

* Don't create MinimizeKernel until first minimization requested

* Make some compile-time constants into kernel arguments

* Consolidate scale calculation kernel

* Condense alpha/beta reduction kernels using atomics

* Condense line search dot kernels with reductions

* Remove a download, and download grad norm separately

* Asynchronously check lbfgs convergence condition

* Restructure line search to avoid download waiting

* Start line search preemptively in case CPU evaluation is not needed

* In rare cases, constraint error might not decrease after one optimization round

* Better handling of unsupported 64-bit atomics, use FLT_MAX

* Pick gradient mode based on GPU vs. CPU evaluation

* Rework getDiff/getScale reduction, remove reduceBuffer

* Older CUDA might not like float hex literals

* Fix error in a comment

* Initial implementation of PythonForce

* Continuing implementation of PythonForce

* Tests for PythonForce

* Fix memory leaks

* Documentation for PythonForce

* Fixed incorrect return type

* Fix compilation error on Python older than 3.12

* Handle all dtypes

* Optimizations to PythonForce

* Optimized getPositions()

* Test all platforms

* Fix test failures

* Replace SimTK-containing file headers

* Update file headers for new Tinker reader files added

* Reference platform supports nested virtual sites

* Common platform supports nested virtual sites

* Fixed force distribution from nested virtual sites

* Fixed test failures

* Prevent child integrators from accessing deleted context

* Prevent reading property names from nullptr platform

* Don't check properties in a loop, but only when there's a Platform

* Clarified condition in "if" block
Co-authored-by: Peter Eastman <peter.eastman@gmail.com>

* DrudeLangevinIntegrator has getSystemTemperature()

* DrudeNoseHooverIntegrator has getSystemTemperature()

* StateDataReporter reports system temperature for Drude systems

* Fixed incorrect return type

* Allow querying current step count

* Fixed error building Python wrapper

Updated Context::getParameter() and Context::setParameter() error messages to also print name of incorrect parameter to aid debugging.

Added missing required kernel names (ApplyConstraintsKernel and VirtualSitesKernel) to check for platform capabilities, since these kernels are initialized in the next code block.

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