Record compile-time parameters to restart files

[gnwong]

During my tracer particle run (for the Ma-0.94 384x192x192), I noticed that the evolution seemed to differ from the original library run. It's possible that this is due to different versions of the code having been used to run the tracer particles and the original library run.

I started from restart_00000240.h5 and dumped for 1k M, but the different should be noticeable even in images before ~20M.

[c-prather]

This is impossible to track down at this point, and likely not that useful, since a number of (beneficial) changes have been made to how iharm3d applies floors since the library files were run originally.

However, the feature here would be recording the floor values and other compile-time parameters to restart files and checking them on load -- that would give an indication of when evolution is expected to be different or not.