Computation of total forces should be skipped after the system's topology was redefined #778
Description
> How about skipping `calc_force_invgrads` after the first step of reinitatoms? Is that possible?
That would change an already rather delicate logic inside the library. If you propose to go that way, I would rather just set to zero all elements of colvarproxy::atoms_total_forces and call it a day.
Originally posted by @giacomofiorin in #764 (comment)