Fixes for ifx

[billsacks]

Description of changes

(1) ifx apparently doesn't do short-circuit evaluation, so SMS_D_Ld1.ne30pg3_t232.I1850Clm50BgcSpinup.derecho_intel.clm-cplhist was crashing with floating invalid when using the ifx compiler (via ccs_config_cesm1.0.75).

(2) Some input files need NaNs converted to non-NaN fill values

Specific notes

Contributors other than yourself, if any: @jedwards4b

CTSM Issues Fixed (include github issue #):

• Resolves Fix issues uncovered with ifx compiler #3762
• Resolves Floating invalid with ifx compiler in UrbanTimeVarType.F90 #3763
• Resolves NaN values in some input files cause problems with ifx compiler #3764

Are answers expected to change (and if so in what way)? No

Any User Interface Changes (namelist or namelist defaults changes)? No

Does this create a need to change or add documentation? Did you do so? No No

Testing performed, if any: will do regular

[billsacks]

With the changes here together with the change in ESCOMP/MOSART#126, the failing derecho-intel I compset tests are now passing with the recent ccs_config update:

DAE_C2_D_Lh12.f10_f10_mg37.I2000Clm50BgcCrop.derecho_intel.clm-DA_multidrv
ERP_D_Ld3_PS.f09_g17.I2000Clm50Sp.derecho_intel.clm-prescribed
ERP_P256x2_D_Ld5.f19_g17_gris4.I1850Clm50BgcCropG.derecho_intel.clm-glcMEC_increase
ERP_P64x2_D.f10_f10_mg37.I2000Clm50SpRtmFl.derecho_intel.clm-default--clm-nofireemis
SMS_D_Ld1_Mmpi-serial.ne3_ne3_mg37.I2000Clm50SpRs.derecho_intel.clm-ptsRLA
SMS_D_Ld1.ne30pg3_t232.I1850Clm50BgcSpinup.derecho_intel.clm-cplhist--clm-matrixcnOn
SMS_D_Ld1.ne30pg3_t232.I1850Clm50BgcSpinup.derecho_intel.clm-cplhist
SMS_D_Ld3.f10_f10_mg37.I1850Clm50BgcCrop.derecho_intel.clm-default--clm-matrixcnOn
SMS_D_Ld3.f10_f10_mg37.I1850Clm50BgcCrop.derecho_intel.clm-default

My next steps:

• Move the new Prigent file into inputdata (needs fixed directory permissions - currently I tested by temporarily pointing to a file in my scratch space)
• Rebase this onto the bit-for-bit branch
• Run aux_clm derecho-intel tests

I'll mark this as ready for review once these things are done.

[jedwards4b]

Where is your new file?

ls /glade/campaign/cesm/cesmdata/inputdata/lnd/clm2/dustemisdata/Prigent_2005_roughness_0.25x0.25*
/glade/campaign/cesm/cesmdata/inputdata/lnd/clm2/dustemisdata/Prigent_2005_roughness_0.25x0.25_cdf5_c240127.nc

I don't see it in inputdata.

[ekluzek]

Things I will do:

• Merge this to b4b-dev
• Make the b4b-dev tag
• Make MOSART tag
• Update .gitmodules to MOSART and ccs_config tag in a different PR Derecho Intel compiler update #3758
• Run all test suites

[ekluzek]

Where is your new file?

@jedwards4b @billsacks mentions this above. There was a permission problem, that's now fixed, so he couldn't put it in inputdata. But, he'll do that now that the permission problem is resolved. That was my bad...

[ekluzek]

Since, this is b4b we shouldn't include the .gitmodules update here. I'll make a different PR for doing that.

[billsacks]

I just moved the Prigent file into the inputdata space (thanks, @ekluzek !).

@ekluzek my understanding of your recent post is that we'll run full testing on this without updating ccs_config in order to ensure bit-for-bit; I think that's the right approach.