sPuReMD: merge changes from branch 'new_setup'. Rework interface for geometry initialization without file I/O. Other code clean-up.

sPuReMD: fix BGF parser to correctly remap atom coordinates to internal simulation box boundaries. Fix reax_list to handle case when total interactions is zero instead of terminating program.

sPuReMD: rework Fortran F90 interfaces for Amber specifics and declare this interface for such usage.

sPuReMD: fix Coulomb energy calculations for QM/MM mode. Indicate QEq model not supported with QM/MM mode. Add Fortran interface for QM/MM mode.

sPuReMD: add QMMM library interface for external codes (e.g., Amber). Rework file I/O parsing to jive with multiple consecutive simulation code changes. Improve file parser error checking (numeric types). Other code clean-up.

sPuReMD: fixes for multiple consecutive simulation codepath. Reorganization to this codepath (some data structure now carry a flag to designation their allocation state). Fix bug in force field parser (atom type not truncated correctly). Charge sparse_matrix structs to be non-pointers to jive with new allocation state flag. Other code clean-up.

sPuReMD: finalize sPuReMD multiple consecutive simulation code (for integration with other codebases).

PG-PuReMD: fix BGF parser for >1 process (correct atom counts within local simulation box boundaries).

PG-PuReMD: fixes to NVT with Berendsen thermostat. Use updated constants from sPuReMD. Add BGF geometry file parser. Other code clean-up.

sPuReMD: fix memory leak with ICHOLT preconditioner (droptol for OpenMP only) for charge solver. Default to Jacobi preconditioner. Resolve several compiler warnings around C strings (snprintf, etc.). Other code clean-up.

sPuReMD: add functionality to run multiple MD simulations (for usage via integration with external codebases).

PG-PuReMD: in-progress implementation of CUDA-aware MPI for charge solver communications using OpenMPI (default off).

sPuReMD: resolve OpenMP build errors with GCC v10 (behavior changes for function arguments). Suppress possibly unitialized variable warnings for BiCGStab.

PG-PuReMD: fix van der Waals and Coulomb force computations for 1 warp per atom implementation, and enable this implementation by default.

PG-PuReMD: use registers for local arithmetic in CUDA nonbonded calculations and fix shared memory usage in 1 warp per atom implementation.

PG-PuReMD: switch timers to use CUDA events and add additional timers for kernels to match MPI code. Other code cleanup.

This reverts commit 6a1385e9.

PG-PuReMD: align CSR sparse matrix data structure to 128-bit/256-bit segments for memory coalescing (for 1 warp per row SpMV implementation). Change default interaction list and sparse matrix allocations to be multiples of 32 for aforementioned memory coalescing.

PG-PuReMD: switch from using strtok_r to strtok to avoid reliance on GNU extensions with building for the C11/C++11 standard.

PG-PuReMD: only copy host generated grid to device on reneighboring steps (after rebinning atoms and regenerating the grid). Fix small NVE integration bug (reallocation checks before copying new atom info to device).

PG-PuReMD: change CUDA kernel error checking to only make calls to cudaDeviceSynchronize when in debugging mode (non-debug builds thus have better performance with kernel asynchronous launch behavior in single default stream). Change file I/O to only flush when in debugging mode (non-debug builds allow the I/O operations to be grouped for better performance).

PG-PuReMD: add support for symmetric, half stored format (SYM_HALF_MATRIX) of the sparse matrix for the charge model (add initialization routines and fix-up solver).

PG-PuReMD: rework custom reduction functions to use less shared memory and to have the correct thread and block counts. Fix bug in charge matrix initialization (under-allocated space previously). Split charge matrix and bonds/hydrogen bonds memory management routines to mirror the similar splitting of the initialization routines. Code clean-up related to half vs. full list and sparse matrix formats.

PG-PuReMD: improve performance by conditionally re-running initialization kernels within Cuda_Init_Forces based on out-of-memory conditions (for charge matrix, bonds list, hydrogen bonds list).