MPI support on the AMBER-QUICK API interface to run parallel free energy calculation

[lizhen62017]

Hi Kurt,

I have implemented a local patch to resolve this issue. #392

Sadly, it does not compile well on the standalone version of QUICK. Thank you for taking a look at it.

[agoetz]

@lizhen62017 thanks for submitting this draft PR.

You accidentally added your builddir, can you please remove it?

Please also add a comment with details for this PR, explaining in more detail what is the objective and how you addressed it. Thanks.

Adding @ohearnk to the discussion.

[ohearnk]

Hi @lizhen62017, thanks for working on this.

I'd like to get this tested and merged before the upcoming Amber release. In order to do so, could you update this PR to only include the MPI-related changes? Specifically, please remove any local changes for other work (e.g., print statements related to PySCF / LUCJ).

Moreover, it looks like many changes included in the current HEAD commit of the master branch have been reverted in this PR (DL-FIND updates, other bugfixes, etc..). If you could update these files, that would be appreciated.

If you can do this by this week, then we can likely get this into the yearly Amber release. Let me know if you have any questions.