Izumi nag tests fail with debug off

[samsrabin]

Brief summary of bug

Izumi nag tests fail in SHAREDLIB_BUILD with debug off.

Previously reported as part of these issues:

• Issue Two FATES tests that fail #1129
• Issue Meier roughness changes answers on processor count #2219
• Issue Test build failure: SMS_P128x1_Lm25.f10_f10_mg37.IHistClm60BgcCrop.izumi_nag.clm-RxCropCalsAdaptGGCMI #2659
• Issue FatesColdAllVars tests with Clm60Fates compsets: Izumi nag build errors #3330

General bug information

CTSM version you are using: ctsm5.3.064

Does this bug cause significantly incorrect results in the model's science? No

Configurations affected: Izumi nag tests with debug off

Important details of your setup / configuration so we can reproduce the bug

SMS_Ld5.f10_f10_mg37.I2000Clm60FatesRs.izumi_nag.clm-FatesColdAllVars

Important output or errors that show the problem

In the lnd.bldlog file, you'll see something like:

Panic: /home/samrabin/ctsm_fates_ts_20250711/src/fates/parteh/PRTAllometricCNPMod.F90: Unexpected expr node type 434

But notice also the many lines like this:

Warning: /home/samrabin/ctsm_fates_ts_20250711/src/fates/parteh/PRTAllometricCNPMod.F90, line 14: File /home/jedwards/src/ESMF/src/Infrastructure/Util/src/ESMF_UtilTypes.F90 containing module ESMF_UTILTYPESMOD needs recompilation
         detected at :@PRT_GLOBAL_TYPE

[samsrabin]

I've submitted a PR to CIME that will at least give a decent error message in such tests: ESMCI/cime#4820.

[ekluzek]

We talked about this at the standup this morning. The problem with the cime fix, is that it does look like there are other CESM tests that run izumi_nag without DEBUG, that do work.

I was thinking about it and I think the way to solve it is to make some changes to buildlib in CTSM to die with an error if izumi_nag is trying to be built without DEBUG TRUE. I'll start a branch for that.