Add errmsg to check_allocate abort message

[jimmielin]

Fixes #297

Following up from the tropopause_find PR I've updated check_allocate to take in errmsg and updated the tropopause_climo_read to use it and other places within SIMA. I did not update all instances of check_allocate, particularly those in dynamics and in cpl/nuopc as I was not sure if they were brought in externally or not.

If the optional errmsg argument is present a linebreak is added to the end of the abort message using the new_line intrinsic (Fortran 2003). Since errmsg in allocate is a F2003 feature I think this is safe to use? It can otherwise be removed, it's just used to make the output prettier.

This is what it looks like:

 ERROR: cam_thermo_init: Allocate of 'test(test)' failed with code 1. Memory highwater is     641.32 mb, current memory usage is      42.41 mb
Allocation failed with: Hello world this is not really a failure but just a test at /somewhere/CAM-SIMA/src/data/cam_thermo.F90:203

or an actual example made possible by a duplicate allocation...

 ERROR: cam_thermo_init: Allocate of 'kmcnd(pcols,pverp)--2test' failed with code 5014. Memory highwater is     641.35 mb, current memory usage is      43.08 mb
Allocation failed with: Attempt to allocate an allocated object at /somewhere/CAM-SIMA/src/data/cam_thermo.F90:207

[nusbaume]

Thanks for adding this feature @jimmielin! I did have one (optional) request, which is to change the beginning of the full check_allocate error message from Allocate of <varname> to Allocation of <varname>, as I realized that is probably more grammatically correct. Of course don't feel like you have to if it will hold up this PR. Thanks again!

[jimmielin]

Thanks @nusbaume! Updated to Allocation of <varname>

[peverwhee]

one small thing but it doesn't hold up my approval! thanks @jimmielin

[peverwhee]

would it make sense to change this errmsg to "shr_kind_cl" as well to match the other error messages? It's probably unlikely to be too short, but would be consistent with the other usage.

[jimmielin]

Thanks, updated!

[jimmielin]

Regression tests passed subject to existing failures

aux_sima_intel_20241219072857: 9 tests
  SMS_Ln9.mpasa480_mpasa480.FKESSLER.derecho_intel.cam-outfrq_kessler_mpas_derecho (Overall: PEND) details:
    PEND SMS_Ln9.mpasa480_mpasa480.FKESSLER.derecho_intel.cam-outfrq_kessler_mpas_derecho RUN
  SMS_Ln9.ne5pg3_ne5pg3_mg37.FCAM7.derecho_intel.cam-outfrq_se_cslam_analy_ic (Overall: PEND) details:
    PEND SMS_Ln9.ne5pg3_ne5pg3_mg37.FCAM7.derecho_intel.cam-outfrq_se_cslam_analy_ic SHAREDLIB_BUILD RERUN


aux_sima_gnu_20241219072901: 9 tests
  SMS_Ln9.mpasa480_mpasa480.FKESSLER.derecho_gnu.cam-outfrq_kessler_mpas_derecho (Overall: FAIL) details:
    FAIL SMS_Ln9.mpasa480_mpasa480.FKESSLER.derecho_gnu.cam-outfrq_kessler_mpas_derecho RUN time=20
  SMS_Ln9.ne5pg3_ne5pg3_mg37.FCAM7.derecho_gnu.cam-outfrq_se_cslam_analy_ic (Overall: PEND) details:
    PEND SMS_Ln9.ne5pg3_ne5pg3_mg37.FCAM7.derecho_gnu.cam-outfrq_se_cslam_analy_ic SHAREDLIB_BUILD RERUN