release/public-v1: partial fix to threading issue

[climbfuji]

This PR

• removes the unused file num_parthds_stochy.f, and
• modifies the OpenMP threaded regions to use the number of OpenMP threads provided from the calling model at most

This PR is required for NOAA-EMC/fv3atm#62 and ufs-community/ufs-weather-model#55

[jswhit2]

Shouldn't this be merged into develop also? Seems like a bug.

[pjpegion]

I'm running the regression test, if it passes I will merge into ufs-relase and develop -Phil

…

On Mon, Feb 10, 2020 at 3:23 PM Jeffrey Whitaker ***@***.***> wrote: Shouldn't this be merged into develop also? Seems like a bug. — You are receiving this because your review was requested. Reply to this email directly, view it on GitHub <#13>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AJIRVJGGM5OUVWXJGW37U3DRCHHVTANCNFSM4KSWGWOQ> .

-- Philip Pegion NOAA/ESRL/PSD-CIRES (303) 497-7897 philip.pegion@noaa.gov

[climbfuji]

Yes, something similar should be done for the develop, but:

• removing file num_parthds_stochy.f also requires changes in fv3atm develop (which you need to coordinate with EMC)
• this is not the end of the story, the current PR just allows you to set the number of threads to 1 when calling the stochastic_physics run routine - it does not solve the threading deadlock that is somewhere in the code

I would only merge this into the UFS public release and create an issue to fix the code entirely in develop.

[pjpegion]

Changes look good, since the regression tests also pass (stochastic physics w/ threading is not tested) I will also merge into psd/develop, but not master.