post-5.4.n01.alpha-ctsm5.4.CMIP7.21.ctsm5.3.085: Update post-5.4 branch to latest alpha5.4 tag#3649
Conversation
I used WhatsNewInCTSM5.3.md as the starting point for WhatsNewInCTSM5.4.md and will start modifying in subsequent commits
b4b-dev merge 2025-10-16
Merge ctsm5.3.080 into b4b-dev
…tream docs-build-and-deploy: Only run on upstream
…e for CTSM, this compiles with intel on Derecho
… in the local object
…into carbon_isotopes_in_streams
… CTSM stream base type
…ve memory when appropriate
…e method a bit, so that varnames and name aren't sent in, also have to make sdat public, so that the dshr_fldbun_getFldPtr method can use it's internal components
…s for C13/C14 timeseries handling
…ilename was incorrect as well as a typo for C13 in the Advance and Interp routines
…n be used easier in the code, this increases memory, but matches the rest of CTSM better and isolates the handling of the grid for the C14 data which makes it easier for the grid to change, this is a prestep to moving to streams as this is how streams works
…ase_notes alpha-ctsm5.4.CMIP7.20.ctsm5.3.085: WhatsNewInCTSM5.4.md
Introduce WhatsNewInCTSM5.4.md and update other documentation: PR ESCOMP#3364
alpha-ctsm5.4.CMIP7.21.ctsm5.3.085: Three user_nl_clm and two .cfg files with ctsm5.4 updates
….85-to-post-5.4 Update 3 user_nl_clm + 2 .cfg files used in testing plus clean-up PR ESCOMP#3637 resolves Issue ESCOMP#3495 An izumi test failure Issue ESCOMP#3642 Add H2OSNO to fincl1 in AD and SASU modes I did some python testing in this PR. I'm saving the full ctsm5.4 testing for PR ESCOMP#3638, which comes next and has no planned code changes.
|
Actually, I've been leaving off advancing the branch tag number when it's merely an update to latest master. So I think this should be post-5.4.n01.alpha-ctsm5.4.CMIP7.21.ctsm5.3.05. And actually you might as well merge this to ctsm5.3.000 to make this post-5.4.n01.ctsm5.3.000. The merge to ctsm5.3.000 will just bring in ChangeLog updates. It would be good for us to be consistent in how we manage the "nZZ" for branches. We could discuss this in a meeting if it would help. |
ekluzek
left a comment
There was a problem hiding this comment.
I think this needs to have the base changed from master to post-5.4. I'll let you make sure that's right and do it. Then I can review it.
I also suggest you update to ctsm5.4.000 as it's now tagged and even just to shorten the tag name.
|
Oh, I see you ran testing on this. And the description clearly states this was meant to be to post-5.4 so I'll go ahead and rebase here. |
|
The answer changing parts of this are for FATES due to going to ctsm5.3.083 (084 and 085 are b4b so don't matter). And then there are the changes from the ctsm5.4 alpha branch from alpha-ctsm5.4.CMIP7.18 to alpha-ctsm5.4.CMIP7.21. Looks like "19" fixed some LII tests which would show up as a diff, we must have skipped "20" and then "21" will have a few tests with differences. @samsrabin by the way thanks for updating the tagnames to include the entire base version, it took me a bit to figure out what base version post-5.4.n01 corresponded to. |
|
OK, 200 tests are different, but most of those are only due to fieldlist differences or missing baselines. The "21" ctsm5.4 update changed smallville and PLUMBER2 tests. So filtering out those and the Fates tests there are 27 with differences: Looking at one of the short tests above I see that the finidat file is different (which came in ctsm5.4 "19"). So that explains those. So then looking through the above list to see if the namelists only differ by "finidat" -- I don't see any other differences. So it looks like ALL of the differences are explained. Which seemed surprising at the initial list of 200 tests showing differences. |
ekluzek
left a comment
There was a problem hiding this comment.
OK, cool. With the rebase we see the proper differences as expected. And going through the answer changes I was able to justify all 200 of the tests that showed differences (on derecho anyway). I think we don't need to do the same on Izumi. So this looks good to go to me! Thanks for working on this @samsrabin
Oh, I still recommend bringing in the documentation update to ctsm5.3.000 to simplify the branch tag name. And using n01 in the name. But, I'll let you do that part and it's fine to disagree and move forward and talk later...
|
Description of changes
Merges tag
alpha-ctsm5.4.CMIP7.21.ctsm5.3.085into thepost-5.4branch at tagpost-5.4.n01.Specific notes
Contributors other than yourself, if any: None
CTSM Issues Fixed: None
Are answers expected to change (and if so in what way)? Yes; lots of ways.
Any User Interface Changes (namelist or namelist defaults changes)? Yes; lots.
Does this create a need to change or add documentation? Did you do so? No
Testing performed, if any:
aux_clmhas a bunch of diffs, but I'm assuming these are to be expected?/glade/derecho/scratch/samrabin/tests_1204-163718de//scratch/cluster/samrabin/tests_1204-163723iz/