Alter fillpatch and boundary plane calls for overset simulations

[mbkuhn]

Summary

Because exawind-driver calls exchange_solution (and thereby, fillpatch) more than the plain AMR-Wind algorithm, and at slightly different instances in the time step, the Tioga Interface calls to fillpatch need to be made smarter to use the correct time in the algorithm. This hinges on the advance_states call that occurs at the beginning of the timestep. On top of that, the treatment of reading boundary plane data needs to be modified because of when pre_advance and exchange_solution calls happen from the overset solver.

Pull request type

Please check the type of change introduced:

• Bugfix
• Feature
• Code style update (formatting, renaming)
• Refactoring (no functional changes, no api changes)
• Build related changes
• Documentation content changes
• Other (please describe):

Checklist

The following is included:

• new unit-test(s)
• new regression test(s)
• documentation for new capability

This PR was tested by running:

• the unit tests
  • on GPU
  • on CPU
• the regression tests
  • on GPU
  • on CPU

Additional background

Without this fix, the overset solver fails an assert when reading boundary planes in AMR-Wind.

[marchdf]

I like where this is going. I just had a minor thought on clarity of naming functions.

[mbkuhn]

The way I set it up should still be correct with multiple iterations in AMR-Wind, too

[mbkuhn]

it would be good to accompany this with an exawind-driver reg test

[marchdf]

Good work! 100% tests passed, 0 tests failed out of 136