Attempt to get GCM 11.4 to work for ADAS

[rtodling]

This initial attempt to get ADAS to use GCM 11.4 now works for main ADAS - cycling and forecasts, but with restrictions.

Two keep components fail:

1. gcmPert

2. NNR in GAAS

This branch increments an attempt to fix (2) - but not work yet.

To test (1) please follow he following:

i) build ADAS from this branch
ii) Create an empty dir
iii) Cd to the empty dir
ln -s /gpfsm/dnb05/projects/p139/rtodling/tmp.39107/sens.20230722.000000/* .
iv) mpirun -np 216 $YOURBIN/GEOSgcmPert.x

The error the Pert gets is:

AGCM_PERTInitialize The current TIME is 2023/07/18 00:00
CFIO: Reading traj1/x50Rrt.traj.lcv.20230718_0000z.nc4 at 20230718
0
CFIO: Reading traj1/x50Rrt.traj.lcv.20230718_0000z.nc4 at 20230718
0
pe=00111 FAIL at line=00452 Shmem_implementation.F90
pe=00111 FAIL at line=00631 Shmem_implementation.F90
pe=00090 FAIL at line=00452 Shmem_implementation.F90
pe=00090 FAIL at line=00631 Shmem_implementation.F90

[github-actions]

Label error. Requires at least 1 of: 0 diff, 0 diff trivial, Non 0-diff, 0 diff structural, 0-diff trivial, Not 0-diff, 0-diff, automatic, 0-diff uncoupled. Found:

[mathomp4]

I'll ping @atrayano as well since he and Ben are the two main Shmem people I think of.

As for the GAAS NNR, do you know which one? I remember when I was helping @patricia-nasa with CMake'ing up AeroApps there were a couple of them...

[mathomp4]

The shmem error points to code like:

      if(.not.MAPL_ShmInitialized) then
         _RETURN(MAPL_NoShm)
      endif

so I'm guessing Pert needs some update to make it happy as it's like it's trying to use shmem before it's initialized.

[github-actions]

Label error. Requires at least 1 of: 0 diff, 0 diff trivial, Non 0-diff, 0 diff structural, 0-diff trivial, Not 0-diff, 0-diff, automatic, 0-diff uncoupled. Found:

[github-actions]

Label error. Requires at least 1 of: 0 diff, 0 diff trivial, Non 0-diff, 0 diff structural, 0-diff trivial, Not 0-diff, 0-diff, automatic, 0-diff uncoupled. Found: