fv3atm update: CCPP tendencies bugfixes, global restart reproducibility, halo boundary update in dycore, transition to cap_gen.py part 1

[climbfuji]

Description of changes

• Update .gitmodules and submodule pointer for stochastic_physics, fv3atm and NEMS for code review and testing changes described in CCPP tendencies bugfixes, global restart reproducibility, halo boundary update in dycore ufs-weather-model#208 and associated PRs.
  • Note that the NEMS code is currently an open PR (update required after UFS-weather PR #213 NOAA-EMC/NEMS#82) that needs to be merged first, while the fv3atm and stochastic_physics are updates to point to the head of the original branches.
• Change -DOPENMP to -D_OPENMP to match the changes made in fv3atm

Specific notes

Are changes expected to change answers?

• expect changes for GSL physics, regional runs, and restart runs (global and regional)

Testing performed:

• somebody needs to run the s2s tests for this change

[climbfuji]

@BinLiu-NOAA @junwang-noaa could some one please run the regression tests for this PR before we merge fv3atm? Thank you ...

[junwang-noaa]

I think it is for Bin Li @binli2337.

[climbfuji]

@BinLiu-NOAA @junwang-noaa this PR is ready from my side, I don't have the regression test logs yet, though

[junwang-noaa]

Dom, this PR will be after Shan's fractional grid PR.

…

On Wed, Sep 30, 2020 at 4:58 PM Dom Heinzeller ***@***.***> wrote: @BinLiu-NOAA <https://github.com/BinLiu-NOAA> @junwang-noaa <https://github.com/junwang-noaa> this PR is ready from my side, I don't have the regression test logs yet, though — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#200 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AI7D6TN7SJRUSCYAPAFH7NDSIOLXNANCNFSM4R6J4HTQ> .

[climbfuji]

Dom, this PR will be after Shan's fractional grid PR.
…

Sure, no problem. As long as we don't have to wait with ufs-weather-model to proceed, that's fine.

[climbfuji]

@junwang-noaa I saw you asked @shansun6 to update fv3atm to my Sep 26 commit in fv3atm (hash 9e1ba7). Then this PR here will no longer be required.

[junwang-noaa]

Dom, Shan's branch is updated to current s2s fv3 version 9e1ba7 (PR#157, RRTMGP), not ab0e5ae3 in this PR. SInce we don't want to change current s2s baseline, just add a new fractional grid test to have a clean comparison of the new fractional grid test with the current non-fractional grid test. We still need this PR #200, we will create a new set of baseline for PR #200 due to results change.

…

On Mon, Oct 5, 2020 at 9:06 AM Dom Heinzeller ***@***.***> wrote: @junwang-noaa <https://github.com/junwang-noaa> I saw you asked @shansun6 <https://github.com/shansun6> to update fv3atm to my Sep 26 commit in fv3atm (hash 9e1ba7). Then this PR here will no longer be required. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#200 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AI7D6TLBXCJZP4GOFGRS35TSJHAFPANCNFSM4R6J4HTQ> .

[climbfuji]

Dom, Shan's branch is updated to current s2s fv3 version 9e1ba7 (PR#157, RRTMGP), not ab0e5ae3 in this PR. SInce we don't want to change current s2s baseline, just add a new fractional grid test to have a clean comparison of the new fractional grid test with the current non-fractional grid test. We still need this PR #200, we will create a new set of baseline for PR #200 due to results change.
…

Got it, thanks.

[ShanSunNOAA]

Hi Jun, I have added the log files from the exiting regression test cpld_fv3_ccpp_384_mom6_cice_cmeps_ww3_1d_bmark_rt which passed the regression test, and the same test using the fractional grid cpld_fv3_ccpp_384_mom6_cice_cmeps_ww3_1d_bmark_rt_frac The run dir is at /scratch2/BMC/gsd-fv3-dev/Shan.Sun/S2S_RT/rt_25890/. Please let me know if you have any questions. Thanks, Shan

…

On Mon, Oct 5, 2020 at 7:30 AM Dom Heinzeller ***@***.***> wrote: Dom, Shan's branch is updated to current s2s fv3 version 9e1ba7 (PR#157, RRTMGP), not ab0e5ae3 in this PR. SInce we don't want to change current s2s baseline, just add a new fractional grid test to have a clean comparison of the new fractional grid test with the current non-fractional grid test. We still need this PR #200 <#200>, we will create a new set of baseline for PR #200 <#200> due to results change. … <#m_-8865319997791315103_> Got it, thanks. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#200 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ALORMVTOWN5BRYD2ACPPRTTSJHC6JANCNFSM4R6J4HTQ> .

[binli2337]

@climbfuji I will run the regression tests on Hera, Orion and Dell 3 after the code is updated.

[binli2337]

@climbfuji @junwang-noaa
The thread test "cpld_fv3_ccpp_mom6_cice_cmeps_2d_2threads" failed.

Error message:
1: An error occurred in ccpp_physics_init
1: An error occured in GFS_phys_time_vary_init: Logic error: nthrds does not match size of Interstitial variable
1:
1: FATAL from PE 1: Call to CCPP physics_init step failed
1:
1: application called MPI_Abort(MPI_COMM_WORLD, 1) - process 1

[climbfuji]

@climbfuji @junwang-noaa
The thread test "cpld_fv3_ccpp_mom6_cice_cmeps_2d_2threads" failed.

Error message:
1: An error occurred in ccpp_physics_init
1: An error occured in GFS_phys_time_vary_init: Logic error: nthrds does not match size of Interstitial variable
1:
1: FATAL from PE 1: Call to CCPP physics_init step failed
1:
1: application called MPI_Abort(MPI_COMM_WORLD, 1) - process 1

@binli2337 I just heard from @SMoorthi-emc about this issue. Do you have a test case on hera (or instructions) so that I can reproduce the error? Thanks!