To update dycore to the GFDL dev/emc (the FV3 202101 version )

[XiaqiongZhou-NOAA]

This PR is to use the GFDL dycore repository dev/emc. The GFDL dev/emc branch of dycore was synchronized with the GFDL master, which included the FV3 202101 upgrade except inline GFDL_MP:
https://github.com/NOAA-GFDL/GFDL_atmos_cubed_sphere/releases/tag/FV3-202101-public

New baseline should be created.

• hera.intel
• hera.gnu
• orion.intel
• cheyenne.intel
• cheyenne.gnu
• gaea.intel
• jet.intel
• wcoss_cray
• wcoss_dell_p3

Dependencies

NOAA-EMC/ufsatm#262
NCAR/ccpp-physics#597

[climbfuji]

Thanks. We will need to check out the two dycore branches, our current dev/emc in NOAA-EMC and the new branch in NOAA-GFDL, to see the differences. What I am a little worried about is the added public statements in the ccpp-physics PR. Are these variables used (i.e. imported) somewhere in the dycore? It goes somewhat against the CCPP rules to do that. We can certainly do the merge, and later on think about how to do this in a more CCPP-compliant way.

[climbfuji]

Approved - .gitmodules and submodule pointer for fv3atm will need to be updated once NOAA-EMC/ufsatm#262 is merged.

[climbfuji]

fv3atm submodule pointer is correct - do we need for the CI tests to finish or has this been done already?

[DusanJovic-NOAA]

I do not see any green check marks, so probably not. I forgot, how do we start CI?

[MinsukJi-NOAA]

I do not see any green check marks, so probably not. I forgot, how do we start CI?

write "run-ci" in a commit message

[DusanJovic-NOAA]

write "run-ci" in a commit message

Thanks. @XiaqiongZhou-NOAA please make another commit with "run-ci" in a commit message.

[XiaqiongZhou-NOAA]

Should I commit something? or just type" git commit -m "run-ci" "

…

On Tue, Mar 30, 2021 at 9:31 AM Dusan Jovic ***@***.***> wrote: write "run-ci" in a commit message Thanks. @XiaqiongZhou-NOAA <https://github.com/XiaqiongZhou-NOAA> please make another commit with "run-ci" in a commit message. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#481 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ALQE7UUIWJOP4PTE4SPAK2TTGHHCZANCNFSM4ZNCUOCQ> .

[climbfuji]

Should I commit something? or just type" git commit -m "run-ci" "
…

If you do that, you probably need to add something like --allow-empty, but git will complain and tell you what to do.

[MinsukJi-NOAA]

Some CI tests are failing due to issues in auto management of the ec2 instances: More recent commits caused the instances to shut down in the middle of test runs. This will have to be fixed in a future PR. For this PR, I suggest we move forward since some tests managed to complete.

[climbfuji]

Some CI tests are failing due to issues in auto management of the ec2 instances: More recent commits caused the instances to shut down in the middle of test runs. This will have to be fixed in a future PR. For this PR, I suggest we move forward since some tests managed to complete.

Ok with me, please merge.

[DusanJovic-NOAA]

Are there any tests in CI that are not in rt_gnu tests? If there are maybe we should add them to rt_gnu.

[MinsukJi-NOAA]

Are there any tests in CI that are not in rt_gnu tests? If there are maybe we should add them to rt_gnu.

I see fv3_control_debug in rt_gnu. That's about it.