[production/RRFS.v1] Fix 'check all' using 0-sized arrays

[grantfirl]

Commit Queue Requirements:

• Fill out all sections of this template.
• All sub component pull requests have been reviewed by their code managers.
• Run the full Intel+GNU RT suite (compared to current baselines) on either Hera/Derecho/Hercules
• [N/A] Commit 'test_changes.list' from previous step

---

Description:

This PR at least partially addresses #2023 (at least any runtime errors associated with the CCPP). The strategy is to simply make sure that every variable is allocated, although for variables that had been conditionally-allocated to save memory, they are now being allocated with zero size. This has a ripple effect in other repos where one needs to change from checking whether a variable is allocated to whether the size is > 0.

The change to Intel.cmake was used for testing only and I can remove this if we're not ready to permanently add it.

Commit Message:

* UFSWM - Add zero-size variable allocations to pass the `-check all` compiler option for CCPP-related code.
  * AQM - 
  * CDEPS - 
  * CICE - 
  * CMEPS - 
  * CMakeModules - 
  * FV3 - Add zero-size variable allocations to pass the `-check all` compiler option for CCPP-related code.
    * ccpp-physics - Add zero-size variable allocations to pass the `-check all` compiler option for CCPP-related code.
    * atmos_cubed_sphere - Add zero-size variable allocations to pass the `-check all` compiler option for CCPP-related code.
  * GOCART - 
  * HYCOM - 
  * MOM6 - 
  * NOAHMP - 
  * WW3 - 
  * stochastic_physics - 

Priority:

• High: Needed for RRFS.v1 release

Git Tracking

UFSWM:

• Partially addresses Add -check all -check noarg_temp_created -ftrapuv to all DEBUG build flags of UFSWM and subcomponents #2023

Sub component Pull Requests:

• FV3: [production/RRFS.v1] Fix 'check all' using 0-sized arrays NOAA-EMC/fv3atm#789
  • ccpp-physics: [production/RRFS.v1] Fix 'check all' with 0-size allocations ccpp-physics#178
  • ccpp-framework: Dummy-allocate inactive local variables NCAR/ccpp-framework#536
  • atmos_cubed_sphere: [production/RRFS.v1] Fix 'check all' using 0-sized allocations NOAA-EMC/GFDL_atmos_cubed_sphere#86

UFSWM Blocking Dependencies:

• None

---

Changes

Regression Test Changes (Please commit test_changes.list):

• No Baseline Changes.

Input data Changes:

• None.

Library Changes/Upgrades:

• No Updates

---

Testing Log:

• RDHPCS
  • Hera
  • Orion
  • Hercules
  • Jet
  • Gaea
  • Derecho
• WCOSS2
  • Dogwood/Cactus
  • Acorn
• CI
• opnReqTest (complete task if unnecessary)

[grantfirl]

Replaces #2102 and this is targeted at the RRFS release branch.

[grantfirl]

Retesting on Hera. Will post results when finished.

[grantfirl]

@jkbk2004 I kicked off the tests thinking that the baseline from where the RRFS branch was cut in the history would still be valid, but I see from #2147 that you're creating a new one. I'll retest using that baseline when it's ready.

[grantfirl]

@jkbk2004 My RTs were successful (no BL changes), but that was comparing against the saved baselines from the develop branch.

[grantfirl]

@jkbk2004 I'm re-running RTs pointing at the new RRFS baselines for 20240227. Just to confirm, those new baselines do not include the changes in #2147, right?

[grantfirl]

Throughput on Hera is slow for our account right now. It looks like it'll be a while.

[grantfirl]

@jkbk2004 Hi re-running RTs again because the last time that I did, there were unexpected failures, probably because the RRFS baselines already contained #2147.

[jkbk2004]

@grantfirl I checked code base of production/rrfs.v1 branch. All looks good. On orion, baseline was reproduced ok. Somehow baseline files look screwed up on hera. I restored baselines and cleaned up to go only with rt.conf_rrfs. @MatthewPyle-NOAA hera/orion/hercules tests are done. This PR will be ready once you confirm on WCOSS2 test

[MatthewPyle-NOAA]

I had a lot of tests fail on WCOSS, which really didn't make sense to me. I'm recreating the baselines somewhat differently than I did the first time (using a -b rt.conf_rrfs option this time)

[BrianCurtis-NOAA]

I had a lot of tests fail on WCOSS, which really didn't make sense to me. I'm recreating the baselines somewhat differently than I did the first time (using a -b rt.conf_rrfs option this time)

You would want ./rt.sh -e -c -l rt.conf_rrfs

[jkbk2004]

@MatthewPyle-NOAA I checked production/RRFS.v1 code base are ok during the merge process of #2147. So we are ok if you need to re-create baseline on wcoss. Sounds like similar things happened on hera.

[grantfirl]

@jkbk2004 @MatthewPyle-NOAA @BrianCurtis-NOAA I'm glad that tests are passing on Hera. This saves me from a "wild goose chase" of trying to figure out the failures that I was getting. My next course of action was to produce my own baseline from the top of production/RRFS.v1 and use that, but it sounds like everything is working OK now.

[MatthewPyle-NOAA]

I'm good with it now on WCOSS.

[MatthewPyle-NOAA]

@jkbk2004 Let me know if you need more from me at this stage. I do have a WCOSS log to add eventually.

[jkbk2004]

@jkbk2004 Let me know if you need more from me at this stage. I do have a WCOSS log to add eventually.

@MatthewPyle-NOAA sure! go ahead to push wcoss log. then we can move on to merge/

[MatthewPyle-NOAA]

@jkbk2004 Let me know if you need more from me at this stage. I do have a WCOSS log to add eventually.

@MatthewPyle-NOAA sure! go ahead to push wcoss log. then we can move on to merge/

@jkbk2004 Where/how do I push it? I tested with Grant's feature branch, but don't have the permission to push to his space.

[jkbk2004]

@jkbk2004 Let me know if you need more from me at this stage. I do have a WCOSS log to add eventually.

@MatthewPyle-NOAA sure! go ahead to push wcoss log. then we can move on to merge/

@jkbk2004 Where/how do I push it? I tested with Grant's feature branch, but don't have the permission to push to his space.

@grantfirl can you add Mathew for the permission to your repo?

[grantfirl]

@jkbk2004 Let me know if you need more from me at this stage. I do have a WCOSS log to add eventually.

@MatthewPyle-NOAA sure! go ahead to push wcoss log. then we can move on to merge/

@jkbk2004 Where/how do I push it? I tested with Grant's feature branch, but don't have the permission to push to his space.

@grantfirl can you add Mathew for the permission to your repo?

I've added @MatthewPyle-NOAA as a collaborator on my fork, so he should have permission to push now.

[MatthewPyle-NOAA]

Thanks @grantfirl WCOSS log file has been pushed now.

[grantfirl]

@jkbk2004 @MatthewPyle-NOAA Are we ready to start merging? I can do ccpp-physics and ccpp-framework, but nobody has reviewed NOAA-EMC/GFDL_atmos_cubed_sphere#86. NOAA-GFDL/GFDL_atmos_cubed_sphere#315 was approved, but I opened a new PR when I changed the target and rebased.

They both contain the same simple change, so I imagine that the new PR should have implicit approval.

[grantfirl]

OK, NOAA-EMC/GFDL_atmos_cubed_sphere#86 was approved. I think that we're ready to merge.

[jkbk2004]

@MatthewPyle-NOAA It will be good idea to clean up commit messages when squash/merge. Do you want me to merge this pr?

[MatthewPyle-NOAA]

I can merge it after cleaning up the merge message.