Refactoring of USER-TALLY computes to handle sparse/hybrid system for many processors plus bugfixes

Refactor and bugfix for some TIP4P pair styles

Updated NetCDF dump style (dump netcdf)

Kokkos library update to v2.03.05

Some extensions/cleanup in USER-MANIFOLD

Add `reinit` keyword to rigid body fixes

Implement check whether commands or styles try to change cached properties in rigid body integrators

Remove unnecessary thread fences in Kokkos package

Integrate compute cnp/atom contributed by Paulo Branicio (USC)

add "atomonly" optimized neighbor list build styles to USER-OMP

Using correct ndegree instead of nnn

Add USER-OMP compatible OpenMP support to USER-REAXC

added several features to the NEB

with sparse and hybrid systems, Pair::ev_tally() may not be called on
every processor and thus the computes in USER-TALLY may hang during
reverse communication because of the error->all() call after checking
whether callback from Pair::ev_tally() has been called at least once.
To address this cleanly, a second callback function needs to be added,
which is run during Pair::ev_setup() and will now handle all memory
re-allocation and clearing of accumulators, just like it is done for
regular tallied data.

- make heatj a pointer instead of a static array
- fix memory leaks for eatom, stress
- simplify and streamline computation

MAINT: Turned 'global' options into a 'thermo yes'/'thermo no' option that enables dumping of thermo data to the netcdf file (for parallel NetCDF/MPIIO variant).

MAINT: Turned 'global' options into a 'thermo yes'/'thermo no' option that enables dumping of thermo data to the netcdf file.

make pair styles lj/cut/tip4p/long/omp, lj/long/tip4p/long and lj/long/tip4p/long/omp consistent with the reset of tip4p styles

MAINT: Switched NetCDF from 64BIT_OFFSET to 64BIT_DATA which can handle frames (of unlimited dimension) > 2 GB. This becomes important for system sizes 100 Mio atoms and upwards.