excessive print statements to stdout from MAPL OpenMP_Support.F90

[junwang-noaa]

When we run with GOCART in UFS weather model, we noticed excessive print statements in the model run output file:

41: OpenMP_Support.F90 536 splitting component:
57: OpenMP_Support.F90 536 splitting component:
43: OpenMP_Support.F90 536 splitting component:
49: OpenMP_Support.F90 536 splitting component:
47: OpenMP_Support.F90 536 splitting component:
59: OpenMP_Support.F90 536 splitting component:
53: OpenMP_Support.F90 536 splitting component:
51: OpenMP_Support.F90 536 splitting component:

This becomes a problem when we run high resolution tests with large number of tasks. Further check shows that the prints are from MAPL OpenMP_Support.F90:

MAPL/generic/OpenMP_Support.F90

Lines 540 to 541 in df6786c

	if(myPET==0) print*,__FILE__,__LINE__, 'internal states labels : <',trim(comp_name), (trim(labels(i)),i=1,size(labels)), '>'
	print*,__FILE__,__LINE__, 'splitting component: <',trim(comp_name),'>'

This line seems a debug prints, could this line be commented out or only run in debug mode? Thanks

[junwang-noaa]

@tclune @weiyuan-jiang FYI.

[tclune]

@weiyuan-jiang Please either delete or convert these to use pflogger.

[weiyuan-jiang]

@junwang-noaa which tag are you using now ? I can patch your tag and also remove the extra print for the future release.

[junwang-noaa]

We are using MAPL 2.40.3.

[mathomp4]

So that'd be a new MAPL v2.40.5 as we already have v2.40.4. I can work with @weiyuan-jiang to get that released.

(Note: v2.40.4 was a minor update for find_package(MAPL).)

[mathomp4]

@junwang-noaa Thanks to @weiyuan-jiang we now have a MAPL v2.40.5. The only real change is the removal of these prints:

v2.40.4...v2.40.5

[junwang-noaa]

Thank you all!

[junwang-noaa]

@mathomp4 Another question, may I ask if you have a major release version with this change? There is request in UFS to install MAPL 2.43.0, which I believe does not contain this fix in this issue. Thanks

[mathomp4]

@junwang-noaa MAPL v2.46.2 has the fix, but if you can't move to that, we could issue MAPL v2.43.3 with the fix.

[junwang-noaa]

Thank you, @mathomp4! We will test with MAPL v2.46.2 in UFS.

[mathomp4]

@junwang-noaa Sounds good. If you can't move to it (needs ESMF 8.6.1 which is in spack-stack v1.7), then we can backport for you.

Edited to correct to ESMF 8.6.1. MAPL 8.6.1 is about 30 years in the future unless we start really cranking out API changes. 😄