User/z1l/group pass #33
Conversation
…group_pass and complete_group_pass
|
Bob, Note that this includes the commits in pull request #32 + one other commit efd6249. I'm not sure but I think that handling this pull request will automatically handle #32 because github pull requests are based on refs. To save time, try handling #33 first. Request #32 is based on the ref c152896 which is second to last in this request. If handling #33 doesn't close #32 then we'll just close the other by hand. This will be an interesting experiment with github and we are watching your handling of the pull request with great interest - no pressure! :^) |
|
I tried doing the checkout and merge commands, but it appears that a new version of the shared code is required for these changes to compile. I am using the tikal_201403 shared code. What version of the shared code do these updates require? |
Oceanographer and Ocean Group Head NOAA GFDL || Phone: (609) 452-6508 |
|
Popping popcorn. Whit Anderson, Ph.D., Oceanographer Carpe diem (cum consensus uxori). On Jul 14, 2014, at 7:25 PM, Robert Hallberg wrote:
|
|
Hi Bob, Sorry I forgot to mention the shared codes update. The following shared These fixes is to make non-blocking halo pass to work. I only tested the Thanks. Zhi On Mon, Jul 14, 2014 at 7:25 PM, Robert Hallberg [email protected]
|
Oceanographer and Ocean Group Head NOAA GFDL || Phone: (609) 452-6508 #33 (comment) |
|
Hi Bob, I think this is a good idea. I will check in updates in shared codes in Zhi On Tue, Jul 15, 2014 at 9:57 AM, Robert Hallberg [email protected]
|
|
Did you mean gitlab? -A. Dr Alistair Adcroft ([email protected]) On Tue, Jul 15, 2014 at 10:21 AM, Zhi Liang [email protected]
|
|
Hi Alistair, I thought github and gitlab are synchronized. I will check in either Zhi On Tue, Jul 15, 2014 at 10:30 AM, Alistair Adcroft (GFDL) <
|
|
We'll have to check with Seth/Balaji - I think committing directly to the Dr Alistair Adcroft ([email protected]) On Tue, Jul 15, 2014 at 10:36 AM, Zhi Liang [email protected]
|
|
The github version of shared will be a mirror of the gitlab version. So Currently we don't have an official gitlab version of shared. I will Doing both CVS and git is not ideal, but as long as it isn't something that Seth Underwood Modeling Services Group (609) 452-5847 Office On Tue, Jul 15, 2014 at 10:48 AM, Alistair Adcroft - NOAA Affiliate <
|
|
The frequency will be higher than the released patches but hopefully longer -A. Dr Alistair Adcroft ([email protected]) On Tue, Jul 15, 2014 at 11:01 AM, Seth Underwood - NOAA Affiliate <
|
…t do halo pass and cause the model to crash
|
Bob, Seth HAS pushed the necessary FMS updates to the github so external users can now get the code. I'll post the necessary steps on the github wiki and send out an email to all later. In the meantime, you can process this PR. -A. |
…d together for halo pass
|
Zhi, Bob found answer changes with this PR for the following experiments: solo_ocean/flow_downslope/layer I checked also and concur that the above have different results. Any idea why? -A. |
|
Hi Alistair, I will find out the reason. Zhi On Mon, Jul 21, 2014 at 12:18 PM, Alistair Adcroft (GFDL) <
|
|
Hi Alistair, I figured out the reason. This is a change I made in the branch Zhi On Mon, Jul 21, 2014 at 12:18 PM, Alistair Adcroft (GFDL) <
|
|
Those three were fixed but now solo_ocean/DOME is failing (unless I missed it last time)? |
|
Hi Alistair, I will find out the reason. Thanks, Zhi On Mon, Jul 21, 2014 at 4:59 PM, Alistair Adcroft (GFDL) <
|
|
Hi Alistair, I made a fix in MOM_barotropic.F90 (OBC related) and all the solo Zhi On Mon, Jul 21, 2014 at 5:16 PM, Zhi Liang - NOAA Affiliate <
|
- Updated SIS_parameter_doc files. - No answer changes. - NOAA-GFDL/SIS2@66b25e2 Merge branch 'user/nnz/fix_k_loop' into dev/master - NOAA-GFDL/SIS2@41e9b4c Merge branch 'dev/master' into user/nnz/fix_k_loop - NOAA-GFDL/SIS2@53991cd This fixes issue mom-ocean#33 - NOAA-GFDL/SIS2@eff04b1 Merge branch 'ashao-dev/master' into dev/master - NOAA-GFDL/SIS2@5abfb11 Merge branch 'dev/master' of https://github.com/ashao/SIS2 into ashao-dev/master - NOAA-GFDL/SIS2@947eca6 Explicitly initialize all of domore_u and domore_v
* use flag to control lag startup * additional log msg cleanup * clarify restart_mode control
- The loop for k was missing
The old length (128) of standard_name is now too short. Increasing it to 256 for CMIP6 runs. Resloves mom-ocean#33
Use group halo pass interface: group_pass_type, create_group_pass, do_group_pass, start_group_pass, complete_group_pass.