Remove use of "master" in CTSM

[ekluzek]

We currently have about a thousand instances of the word "master" in CTSM (and about 2k in cime). Most of this is the use of the variable "masterproc". It would be good to get away from this terminology and use a different term in its place. Doing this should be bit-for-bit and relatively straight forward to do. Although because it's used a lot it won't be a trivial thing to do.

[billsacks]

Let's keep #1048 for the documentation-related pieces of this, which relate to "master" as in "master branch". This issue can be used for pieces that involve Fortran code changes, such as the use of "master" in the MPI sense.

[ekluzek]

Note, that CAM is working on this same thing. They are suggesting to move to root_proc for masterproc. Other terms I'm seeing suggested in general to replace master are: primary, director, initiator, controller, leader, conductor, source, or main.

https://github.com/ESCOMP/CAM/issues/327

[DavidBergeron-eng]

The vault is classified documents not allowed entry by anyone but me or designer...ME!!!

[DavidBergeron-eng]

What are you wanTing to pull from vault

[ekluzek]

CAM closed their issue as a "not planned". #1048 will deal with the part of this for github branches.

But, the big outstanding piece would be to change masterproc to something else in our MPI code. Which is a big change.

There is also a use of master_doc in one place lilac/docs/conf.py.

Uses of "master" I see in github are: master_doc, masterproc and mastertask in the code. And the following phrases in documentation or comments:

• master proc
• master task
• master restart
• master clock
• documentation master file
• master input namelist file
• Repository URL paths that include "master" as the branch for it