fix issue with using compset longname in cesm testing by jedwards4b · Pull Request #1 · ESCOMP/FMS

jedwards4b · 2024-07-30T19:19:24Z

Description
This PR adjusts string lengths to be longer than 128 characters and
modifies the get_time_string function to allow for the use of % signs in
cesm test names. This allows us to run tests with MOM using the cesm
testing mechanism with long compset names.

Fixes # (issue)

How Has This Been Tested?
I ran test ERS.ne30pg3_t232.1850_CAM60_CLM50%BGC-CROP_CICE_MOM6_MOSART_DGLC%NOEVOLVE_SWAV.derecho_intel.allactive-defaultio

This test will fail in mom - first because the filename string length was hardcoded to 128 characters and second because fms assumes that the % sign will not be used in filenames.

Checklist:

My code follows the style guidelines of this project
I have performed a self-review of my own code
I have commented my code, particularly in hard-to-understand areas
I have made corresponding changes to the documentation
My changes generate no new warnings
Any dependent changes have been merged and published in downstream modules
New check tests, if applicable, are included
make distcheck passes

jedwards4b · 2024-07-30T19:21:13Z

    CALL get_date(current_time, yr1, mo1, dy1, hr1, mi1, sc1)
    len = LEN_TRIM(filename)
-    first_percent = INDEX(filename, '%')
+    do i=1,9


In this loop instead of just looking for a % sign we look for %\d where \d is a digit 1-9. This allows other uses of the % sign in the filename.

mnlevy1981 · 2024-07-30T19:50:22Z

+       write(first_fms_percent,"('%',i1)") i 
+       first_percent = INDEX(filename, first_fms_percent)
+       if (first_percent > 0) then
+          exit
+       endif
+    enddo


I don't think this will find the index of first percent of the string xyz%8f%2g%1300%400 correctly; could you do something like

first_percent = INDEX(filename, '%1') do i=2,9 write(first_fms_percent,"('%',i1)") i first_percent_loc = INDEX(filename, first_fms_percent) if (first_percent_loc > 0) then if (first_percent == 0) then first_percent = first_percent_loc else first_percent = min(first_percent, first_percent_loc) endif enddo

Also, are you sure you don't need to capture %0?

Oh, you pushed changes while I was typing that :) I think this is what you want in find_first_fms_percent()

I don't understand your loop - the one I wrote appears to work correctly now.

I see now, you are correct - I will update the PR. Thanks

mnlevy1981

This looks good to me; I'd like to see test suite output before approving it (I assume it'll be bit-for-bit with aux_mom / pr_mom, but it would be nice to verify). I should have time to run those tests in the next day or two

mnlevy1981

I ran aux_mom and the baseline comparisons all passed (except we still have a B1850MOM compset in the test list and for some reason create_newcase failed on the ERS.TL319_t061_wt061.GMOM_JRA_WD.derecho_intel). The test cases are all in /glade/derecho/scratch/mlevy/tests/aux_mom/cime6.1.0-cesm_cice6_5_0_12-mi_240705_newfms.intel; I'm seeing a bunch of NLCOMP failures that point to diag_table, which makes me wonder if there's a problem with NLCOMP because nothing changed in the generation of diag_table:

14	FAIL
	7	NLCOMP (FAIL)
      1 DIMCS_Ld1.TL319_t232.GMOM_JRA.derecho_intel.mom-cfc_mods--mom-debug
      1 ERI.TL319_t232.GMOM_JRA.derecho_intel.mom-debug
      1 ERS.T62_t061.GMOM_JRA.derecho_intel.mom-debug
      1 PET.TL319_t232.CMOM_JRA.derecho_intel
      1 SMS_D.TL319_t232.GMOM_JRA_RYF.derecho_intel
      1 SMS_Ld40.TL319_t232.CMOM_JRA.derecho_intel
      1 SMS.T62_t025.CMOM.derecho_intel
	2	COMPARE (FAIL)
      1 ERI.TL319_t232.GMOM_JRA.derecho_intel.mom-debug
      1 PET.TL319_t232.CMOM_JRA.derecho_intel
	1	MEMCOMP (FAIL)
      1 ERI.TL319_t232.GMOM_JRA.derecho_intel.mom-debug
	1	RUN (FAIL)
      1 DIMCS_Ld1.TL319_t232.GMOM_JRA.derecho_intel.mom-cfc_mods--mom-debug
	1	BASELINE (FAIL)
      1 PET.TL319_t232.CMOM_JRA.derecho_intel
	2	CREATE_NEWCASE (FAIL)
      1 ERS.TL319_t061_wt061.GMOM_JRA_WD.derecho_intel
      1 SMS_Ld2.f09_t232.B1850MOM.derecho_intel.mom-bcompset

(the DIMCS_Ld1 test ran out of walltime, but something like 10 or 11 of the scaling tests finished and all looked good)

From the SMS_Ld40.TL319_t232.CMOM_JRA.derecho_intel TestStatus.log file:

2024-07-31 11:34:04: NLCOMP
Comparison failed between '/glade/derecho/scratch/mlevy/tests/aux_mom/cime6.1.0-cesm_cice6_5_0_12-mi_240705_newfms.intel/SMS_Ld40.TL319_t232.CMOM_JRA.derecho_intel.GC.20240731_113353_yhxrk0/CaseDocs/diag_table' with '/glade/derecho/scratch/mlevy/baselines/aux_mom/cime6.1.0-cesm_cice6_5_0_12-mi_240705/SMS_Ld40.TL319_t232.CMOM_JRA.derecho_intel/CaseDocs/diag_table'
Inequivalent lines 'MOM6 diagnostic fields table for CESM case: SMS_Ld40.TL319_t232.CMOM_JRA.derecho_intel.GC.20240731_070336_6vi8sd' != 'MOM6 diagnostic fields table for CESM case: SMS_Ld40.TL319_t232.CMOM_JRA.derecho_intel.GC.20240731_113353_yhxrk0'
  NORMALIZED: 'MOM6 diagnostic fields table for CESM case: SMS_Ld40.TL319_t232.CMOM_JRA.derecho_intel.GC.20240731_070336_6vi8sd' != 'MOM6 diagnostic fields table for CESM case: SMS_Ld40.TL319_t232.CMOM_JRA.derecho_intel.GC.20240731_113353_yhxrk0'
Inequivalent lines 'SMS_Ld40.TL319_t232.CMOM_JRA.derecho_intel.GC.20240731_070336_6vi8sd.mom6.h.rho2%4yr-%2mo',                        1,  'days',   1, 'days',   'time', 1, 'days' != 'SMS_Ld40.TL319_t232.CMOM_JRA.derecho_intel.GC.20240731_113353_yhxrk0.mom6.h.rho2%4yr-%2mo',                        1,  'days',   1, 'days',   'time', 1, 'days'
  NORMALIZED: 'SMS_Ld40.TL319_t232.CMOM_JRA.derecho_intel.GC.20240731_070336_6vi8sd.mom6.h.rho2%4yr-%2mo',                        1,  'days',   1, 'days',   'time', 1, 'days' != 'SMS_Ld40.TL319_t232.CMOM_JRA.derecho_intel.GC.20240731_113353_yhxrk0.mom6.h.rho2%4yr-%2mo',                        1,  'days',   1, 'days',   'time', 1, 'days'
Inequivalent lines 'SMS_Ld40.TL319_t232.CMOM_JRA.derecho_intel.GC.20240731_070336_6vi8sd.mom6.h.native%4yr-%2mo',                      1,  'days',   1, 'days',   'time', 1, 'days' != 'SMS_Ld40.TL319_t232.CMOM_JRA.derecho_intel.GC.20240731_113353_yhxrk0.mom6.h.native%4yr-%2mo',                      1,  'days',   1, 'days',   'time', 1, 'days'
  NORMALIZED: 'SMS_Ld40.TL319_t232.CMOM_JRA.derecho_intel.GC.20240731_070336_6vi8sd.mom6.h.native%4yr-%2mo',                      1,  'days',   1, 'days',   'time', 1, 'days' != 'SMS_Ld40.TL319_t232.CMOM_JRA.derecho_intel.GC.20240731_113353_yhxrk0.mom6.h.native%4yr-%2mo',                      1,  'days',   1, 'days',   'time', 1, 'days'
Inequivalent lines 'SMS_Ld40.TL319_t232.CMOM_JRA.derecho_intel.GC.20240731_070336_6vi8sd.mom6.h.z%4yr-%2mo',                           1,  'days',   1, 'days',   'time', 1, 'days' != 'SMS_Ld40.TL319_t232.CMOM_JRA.derecho_intel.GC.20240731_113353_yhxrk0.mom6.h.z%4yr-%2mo',                           1,  'days',   1, 'days',   'time', 1, 'days'
  NORMALIZED: 'SMS_Ld40.TL319_t232.CMOM_JRA.derecho_intel.GC.20240731_070336_6vi8sd.mom6.h.z%4yr-%2mo',                           1,  'days',   1, 'days',   'time', 1, 'days' != 'SMS_Ld40.TL319_t232.CMOM_JRA.derecho_intel.GC.20240731_113353_yhxrk0.mom6.h.z%4yr-%2mo',                           1,  'days',   1, 'days',   'time', 1, 'days'
Inequivalent lines 'SMS_Ld40.TL319_t232.CMOM_JRA.derecho_intel.GC.20240731_070336_6vi8sd.mom6.h.static',                               -1, 'days',   1, 'days',   'time' != 'SMS_Ld40.TL319_t232.CMOM_JRA.derecho_intel.GC.20240731_113353_yhxrk0.mom6.h.static',                               -1, 'days',   1, 'days',   'time'
  NORMALIZED: 'SMS_Ld40.TL319_t232.CMOM_JRA.derecho_intel.GC.20240731_070336_6vi8sd.mom6.h.static',                               -1, 'days',   1, 'days',   'time' != 'SMS_Ld40.TL319_t232.CMOM_JRA.derecho_intel.GC.20240731_113353_yhxrk0.mom6.h.static',                               -1, 'days',   1, 'days',   'time'
Inequivalent lines 'ocean_model_rho2', 'volcello','volcello','SMS_Ld40.TL319_t232.CMOM_JRA.derecho_intel.GC.20240731_070336_6vi8sd.mom6.h.rho2%4yr-%2mo', 'all', 'mean', 'none', 1 != 'ocean_model_rho2', 'volcello','volcello','SMS_Ld40.TL319_t232.CMOM_JRA.derecho_intel.GC.20240731_113353_yhxrk0.mom6.h.rho2%4yr-%2mo', 'all', 'mean', 'none', 1
  NORMALIZED: 'ocean_model_rho2', 'volcello','volcello','SMS_Ld40.TL319_t232.CMOM_JRA.derecho_intel.GC.20240731_070336_6vi8sd.mom6.h.rho2%4yr-%2mo', 'all', 'mean', 'none', 1 != 'ocean_model_rho2', 'volcello','volcello','SMS_Ld40.TL319_t232.CMOM_JRA.derecho_intel.GC.20240731_113353_yhxrk0.mom6.h.rho2%4yr-%2mo', 'all', 'mean', 'none', 1

So I think we have some work to do before aux_mom is useful again, but all in all this PR looks fine. Should someone from the CAM side run some tests as well?

jedwards4b · 2024-08-01T14:18:13Z

It looks like the normalize mechanism should strip the GC.20240731_070336_6vi8sd part of the file name but isn't doing it.

fix issue with using compset longname in cesm testing

c9b902d

jedwards4b commented Jul 30, 2024

View reviewed changes

jedwards4b requested a review from alperaltuntas July 30, 2024 19:21

fix issue with openfile function

b64355d

mnlevy1981 reviewed Jul 30, 2024

View reviewed changes

correct algorythm thanks to Mike Levy

85ef657

mnlevy1981 reviewed Jul 30, 2024

View reviewed changes

mnlevy1981 approved these changes Aug 1, 2024

View reviewed changes

alperaltuntas approved these changes Aug 6, 2024

View reviewed changes

alperaltuntas merged commit 5a030fc into ESCOMP:dev/ncar Aug 6, 2024

jedwards4b mentioned this pull request Aug 13, 2024

intel-oneapi/2024.0.2 ICE #3

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

fix issue with using compset longname in cesm testing#1

fix issue with using compset longname in cesm testing#1
alperaltuntas merged 3 commits into
ESCOMP:dev/ncarfrom
jedwards4b:fix/string_lengths

jedwards4b commented Jul 30, 2024

Uh oh!

jedwards4b Jul 30, 2024

Uh oh!

mnlevy1981 Jul 30, 2024

Uh oh!

mnlevy1981 Jul 30, 2024

Uh oh!

jedwards4b Jul 30, 2024

Uh oh!

jedwards4b Jul 30, 2024

Uh oh!

mnlevy1981 left a comment

Uh oh!

mnlevy1981 left a comment

Uh oh!

jedwards4b commented Aug 1, 2024

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Conversation

jedwards4b commented Jul 30, 2024

Uh oh!

jedwards4b Jul 30, 2024

Choose a reason for hiding this comment

Uh oh!

mnlevy1981 Jul 30, 2024

Choose a reason for hiding this comment

Uh oh!

mnlevy1981 Jul 30, 2024

Choose a reason for hiding this comment

Uh oh!

jedwards4b Jul 30, 2024

Choose a reason for hiding this comment

Uh oh!

jedwards4b Jul 30, 2024

Choose a reason for hiding this comment

Uh oh!

mnlevy1981 left a comment

Choose a reason for hiding this comment

Uh oh!

mnlevy1981 left a comment

Choose a reason for hiding this comment

Uh oh!

jedwards4b commented Aug 1, 2024

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants