Skip to content

Look at changes - #2

Closed
tcclevenger wants to merge 1327 commits into
tcclevenger/aurora_working_masterfrom
master
Closed

Look at changes#2
tcclevenger wants to merge 1327 commits into
tcclevenger/aurora_working_masterfrom
master

Conversation

@tcclevenger

Copy link
Copy Markdown
Owner

No description provided.

@tcclevenger
tcclevenger marked this pull request as draft May 6, 2026 02:42
vijaysm and others added 29 commits June 4, 2026 16:53
…fix-cosp

Merge branch 'mahf708/eamxx/fix-cosp' into next (PR E3SM-Project#8411)

Makes targeted adjustments to the COSP and RRTMGP physics
modules to improve the timing logic for COSP updates and
to clean up unnecessary code in the RRTMGP process interface.

[BFB] for everything except eamxx cosp
)

Add ctest-based and python driver scripts for testing

[BFB]
The nlmaps test needs to use the driver-mct version because
the test parses the log file looking for nlmaps_verbose output
only available in driver-mct
Change mask for ne30_ne30 from gx1v6 to oEC60to30v3
The gx1v6 domain file is degenerate and we should not
be using POP grids anywhere.
Add handling of ocn_rof_two_way and lnd_rof_two_way to
prep_rof_mod.  Also fix a subtle bug in defining samegrid_ro
in prep_rof_init.
Ensure that the geo data streams are added in the same order across ranks.

[BFB]
We were allowed to call get_view<double*> on an IntType field
Prefents ctest from interpreting certain strings found
in the build output as actual error
It belongs in mach files for GNU machines
Intended to later replace AtmosphereInput, it's a very
lightweight structure with very little dependencies
(only the Field class and the scorpio interfaces).
Also provide free functions for one-off reads
Do not allow to call get_view<double*> on an IntType field.

[BFB]
Add machine specs and mach file needed to run test-all-eamxx.
Also, add a oneapi-openmp job in eamxx-sa-testing workflow.

[BFB]
Fix test problems uncovered after switch to driver-moab as default.

Change nlmaps and RCS tests to Vmct
Change mask for ne30_ne30 to oEC60to30v3
adds all 11 river nutrient fields
Fix OCN merge issue
Add proper handling of lnd_rof_two_way and ocn_rof_two_way options
Deallocate temporaries

[BFB]
Co-authored-by: Gregory Lemieux <7565064+glemieux@users.noreply.github.com>
whannah1 and others added 29 commits July 1, 2026 14:44
Use atm_pg_active to choose between the global element count
(pg2/FV cell meshes) and the global vertex count (np4 point clouds)
when determining nx_lnd from iMOAB_GetGlobalInfo for samegrid_al. Previously the
code always used the elem count, which was incorrect for vertex-based
atmosphere grids such as ne4_ne4 and ne30_ne30. The rof mesh path
is also corrected to use the
element count consistently.

Co-Authored-By: Claude <noreply@anthropic.com>
…ject#8534)

Update the default values of ATM_NCPL for ne32/ne64/ne128

[BFB] (because we don't test with these grids)
…8536)

Need ghci-snl-gnu_setup for scripts tests to work

[BFB]
E3SM-Project#8443)

Fix CICE CIME_cppdefs missing BLCK defines after buildnml -nodecomp configure

Fix missing BLCK defines in CIME_cppdefs: CICE configure script is called with
-nodecomp, so it writes CIME_cppdefs without the BLCK decomposition defines.
After buildnml appends those BLCK values to cppdefs, it now writes the complete
string back to CIME_cppdefs. Without this, a subsequent CMake reconfigure (e.g.,
after changing the pe-layout or rebuilding an existing case without cleaning) would
read an incomplete CIME_cppdefs and produce a broken build.

The fix is in the Python buildnml script and runs before any compilation. All machines
and compilers are affected equally.

This is specifically relevant when modifying and rebuilding an existing case without
cleaning: the first build succeeds because CICE_cppdefs.new triggers a full rebuild,
but a subsequent reconfigure reads the stale CIME_cppdefs and misses the BLCK defines.

[bfb]
Fixes memory leaks when accessing the string from MCT for data models

[BFB]
…ect#8546)

Accidentally added _setup to the machine name

[BFB]
…M-Project#8515)

Update global attribute including data license and contact:
Components touched: eam, elm, mosart, mpas-albany-landice, mpas-ocean, mpas-seaice, and eamxx

Closes E3SM-Project#8512
…-Project#8533)

Fix CL options for minimum and maximum land fraction. Provides explicit types for fminval and fmaxval.

[BFB]
…E3SM-Project#8335)

Validate all EAM history field names in namelist before reporting error

Validate all eam history fields in the namelist (atm_in, some via user_nl_eam),
instead of abort/report on the first invalid field, before reporting error back to the user.

[BFB]
…8079)

Make "intel" now refer to oneapi. Intel-classic compiler on chrysalis will be incompatible with Kokkos 5/C++20.
Moving the default to oneapi before updating kokkos to allow users to switch. Keep old intel compiler around for
now as intel-classic.

[non-BFB] for all "intel" tests on Chrysalis. Compiler is changing but "intel" test names aren't changing.
…ect#8219)

This adds support for the C++ GW drag schemes in EAMxx. Nothing has been 
validated yet, but we can save this effort for later when ZM and frontogenesis 
values are available. For now this PR just adds a working GW interface to EAMxx, 
but keeps it as a stealth feature.

[BFB]
Various fixes for testing issues originating from PR E3SM-Project#8421 
which was a major ZM update.
  - The default value of use_fortran_bridge was not properly 
    updated, so CIME tests with F2010xx-ZM were running the fortran
    bridge rather than the new C++ version of ZM. This does not
    explain any failure, but it was an oversight and is incorrect.
  - CIME tests with F2010xx-ZM fail on chrysalis due to a intel 
    specific problem with the allocation of the "estbl" variable. 
    The fix is certainly "hacky", but several alternatives were tested 
    that did not fix the problem. The fix does not appear to negatively 
    affect GPU builds.
  - ZM unit tests are failing due to a baseline generation problem, but 
    an additional issue was discovered where the "old_snow" value for the 
    fortran bridge was incorrectly set. The default value of this parameter 
    changed n the previous PR, so a diff was expected, but it was expected 
    to be fixed by a bless, which is not what happened. The current old_snow 
    update, plus another bless, should resolve the issue.

[non-BFB] for ZM tests and F2010xx-ZM
Bumps [DavidAnson/markdownlint-cli2-action](https://github.com/davidanson/markdownlint-cli2-action) from 23 to 24.
- [Release notes](https://github.com/davidanson/markdownlint-cli2-action/releases)
- [Commits](DavidAnson/markdownlint-cli2-action@v23...v24)

---
updated-dependencies:
- dependency-name: DavidAnson/markdownlint-cli2-action
  dependency-version: '24'
  dependency-type: direct:production
  update-type: version-update:semver-major
...

Signed-off-by: dependabot[bot] <support@github.com>
…(PR E3SM-Project#8535)

Fix null bytes in log files by confining C_NULL_CHAR to MOAB C API call sites
Use atm_pg_active to choose between the global element count (pg2/FV cell meshes) and the global vertex count (np4
point clouds) when determining nx_lnd from iMOAB_GetGlobalInfo for samegrid_al. Previously the code always used the
elem count, which was incorrect for vertex-based atmosphere grids such as ne4_ne4 and ne30_ne30. The rof mesh path is
also corrected to use the element count consistently.

[BFB]
Fixes E3SM-Project#8540
…action-24' (PR E3SM-Project#8552)

Bumps DavidAnson/markdownlint-cli2-action from 23 to 24.
    
Freshen generated package-lock.json file.
Freshen generated index.js file.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.