cmake cleanup in CCPP; cmake in FV3dycore

[aerorahul]

Description

This PR:

• removes the CMake build of the dycore from fv3atm
• places the CMake build of the dycore in the dycore repo. See GFDL_atmos_cubed_sphere PR 141
• Uses installed packages of CCPP from Cleanup cmake build NCAR/ccpp-framework#382 and Cleanup cmake build NCAR/ccpp-physics#705

Testing

How were these changes tested? RT's
What compilers / HPCs was it tested with? RT's
Are the changes covered by regression tests? Yes.
Have the ufs-weather-model regression test been run? On what platform? In progress

• Will the code updates change regression test baseline? No
• Please commit the regression test log files in your ufs-weather-model branch

Dependencies

If testing this branch requires non-default branches in other repositories, list them.
Those branches should have matching names (ideally)

Do PRs in upstream repositories need to be merged first?
If so add the "waiting for other repos" label and list the upstream PRs

• waiting on build the dycore with cmake NOAA-GFDL/GFDL_atmos_cubed_sphere#141
• waiting on Cleanup cmake build NCAR/ccpp-framework#382
• waiting on Cleanup cmake build NCAR/ccpp-physics#705

[junwang-noaa]

Do we have CCPP PRs for this fv3 PR? can we list them in the Dependencies section?

[aerorahul]

Do we have CCPP PRs for this fv3 PR? can we list them in the Dependencies section?

As soon as @climbfuji opens the relevant PR's, I will update the description here.

[climbfuji]

Do we have CCPP PRs for this fv3 PR? can we list them in the Dependencies section?

As soon as @climbfuji opens the relevant PR's, I will update the description here.

These PRs were opened many weeks ago ...

Associated PRs (see also ufs-weather-model PR):

NOAA-GFDL/GFDL_atmos_cubed_sphere#141
NCAR/ccpp-framework#382
NCAR/ccpp-physics#705
#385
ufs-community/ufs-weather-model#791

[aerorahul]

Do we have CCPP PRs for this fv3 PR? can we list them in the Dependencies section?

done.

[climbfuji]

This PR, together with NOAA-GFDL/GFDL_atmos_cubed_sphere#141, changed the regression test results on AWS (CI) and Jet with Intel for yet unknown reasons.

[aerorahul]

The atmosphere dycore build is treated differently than the other components of the UFS.
Isolating and fixing this issue will likely result in non-reproducible baseline due to this special handling.