Jgfouca/branch for acme split 2018 03 27 #2406
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
[BFB] - Bit-For-Bit
Fix domain files setting for the CONUS RRM configuration. Specification of the domain files for the CONUS setup were missing from config_grids.xml, so hgrid was not being set, causing set_horiz_grid to fail during case setup. Adding specification of which domain files to use fixes this issue. Fixes #2147.
Add a small test suite to test basic functionality of supported RRM configurations. These tests use large grids, so more nodes may be required to provide sufficient memory. This can be specified when running the test suite using the --force-procs option to create_test. I.e., create_test e3sm_rrm --force-procs 64.
Add a default PE layout for all RRM grids (any atmosphere grid that begins with "ne0np4"). These grids are all large, and require a larger amount of processors to have access to sufficient memory.
Add RRM-specific tests from the e3sm_rrm test suite to the e3sm_integration test suite to test basic RRM functionality before merging new code changes.
This fixes the CTest script for machines with nersc_slurm queue types. An alternative (and potentially preferred) fix would be to figure out how to create a subtype of the slurm queue in the xml and just use that.
Fix CONUS RRM configuration Fix two issues with the CONUS RRM configuration that prevented configuring a case with CONUS. First, the domain files were not specified for the CONUS grid in config_grids.xml, so hgrid ended up remaining UNSET, causing set_horiz_grid to fail with all components having grid values UNSET. This PR puts the domain file specification back into config_grids.xml for CONUS, and fixes #2147. Second, an older land initial condition (finidat) file was being used for sim year 2000 runs, which was incompatible with recent changes, causing runs to fail. This PR replaces finidat for sim year 2000 with the updated version. Fixes #1899 This PR also adds a test suite for RRM grids. [BFB]
Move the RRM test suite to e3sm_extra_coverage and keep only the CONUS RRM test in e3sm_integration to speed up the integration testing.
Move the allactive usermods_dir to the same level as the testmods_dir
Remove the CONUS test from the e3sm_extra_coverage suite since it is already in e3sm_integration.
Add a testmod to turn on most of the options of the v1 CMIP6 DECK runs including: extra atm output files, cosp, coupler budgets.
Add CMIP6 wcycl ne30 test cases to production (2 days) and integration (1 day with debugging).
Remove SMS.ne30_oECv3_ICG.A_WCYCL1850S. This is replaced by the SMS_D_Ld1 test with A_WCYCL1850S_CMIP6
So the test work completely fail if git describe fails
Add a ne30_oECv3_ICG.A_WCYCL1850S_CMIP6 test to e3sm_prod and include test mods that turn on production-level output. Also add this case to e3sm_integration but with debugging on. Remove the SMS.ne30_oECv3_ICG.A_WCYCL1850S, aka "CMIP5", test. Move the allactive "usermods_dirs" subdirectory up a level to the same as testmods_dirs. [BFB]
Adding configuration file sections for Summit using PGI, IBM compilers. Several settings are adopted from Summitdev, a testbed which was a precursor to Summit. * Added specific compiler sections to config_machines.xml * Added new machine section in config_machines.xml * Updated LSF section in config_batch.xml,removed obsolete lsf_old sec. * mpirun.summit script is installed locally in project-wide accessible directory. Work in progress to tune it for the environment. Testing: Ran FC5AV1C with ne4_ne4 and ne30_ne30. [BFB] - Bit-For-Bit
Adding configuration file sections for Summit using PGI, IBM compilers. Several settings are adopted from Summitdev, a testbed which was a precursor to Summit. - Added specific compiler sections to config_machines.xml - Added new machine section in config_machines.xml - Updated LSF section in config_batch.xml,removed obsolete lsf_old sec. Note: mpirun.summit script is installed locally in project-wide accessible directory. Work in progress to tune it for the environment. Testing: Ran FC5AV1C with ne4_ne4 and ne30_ne30. [BFB] - Bit-For-Bit
Fixes NAG identified errors in atm, homme and clm models Most of changes are related to line length longer than 132 characters, which NAG doesn't allow. Other changes are related where the whole F90 files were enclosed in #if defines. NAG complains about these files as it sees no code in those files when those #if defines are not true. Previously Gautam identified that these kind of files may force recompiliation of the some codes even if no change is made to any of the F90 files. [BFB] - Bit-For-Bit
Add pet testmods for Theta so that multi-submit test jobs go through production queue. [BFB]
Minor changes to be compatible with the recent changes in CIME. [BFB]
Update cime to ESMCI 5.4.0-alpha.25 Feature list: * Improve handling of "non-core" jobs (st.archive) allow for better queue selection for this job * Reinstate ability to select queue by num tasks * Add support for environment and module to depend on job * Mpirun can now depend on queue * Implement improvements to st_archive regular expressions * enabling archiving of dates year >9999 * Add ability to xmlquery derived attributes (e.g. NUM_TASKS) * Provide infrastructure to allow SCM to run with SE dycore * Allow custom batch directive prefixes Bug fixes: * Report lock file info as debug output, not info. * Remove aliasing of element blocks in env_batch * Remove most usage of scan_children and replace with root discovery * Fix mistake in scripts-regression-tests help dump * Improve create_test error message when user provides non-existent test suite [BFB]
This test is typically, by far, the slowest of our integration tests to complete. Since we are only doing a Ln9 run, it makes no sense to use a large peylayout. A much better approach is to use the PS test option to force a smaller layout so we can get in the queue faster. This PR also defined a small grid layout for sandiatoss3. This PR allows both phases of the above test to fit into the sandiatoss3 "short" queue. [BFB]
* esmci_remote_for_split/master: (111 commits) remove debug print statement fixes a py3 issue Allow case_diff to work on cases that are not tests Cannot combine soft operations. remove refs to ccsm remove refs to ccsm Add coment Permit IOErrors as well as OSErrors when trying to copy crpnc log use __init__.py to move entire case object to subdirectory add a README add __init__ file more rearrangement fix pylint issues update templates put case objects in subdirectory take it to the limit cleanup acme -> e3sm references in rst documentation Add an A and X compset test to aux_cime_baselines. fix comment fix comment ...
jgfouca
commented
Mar 27, 2018
| @@ -384,7 +384,7 @@ def _archive_restarts_date_comp(case, archive, archive_entry, | |||
| for i in range(ninst): | |||
| restfiles = "" | |||
| if compname.find("mpas") == 0: | |||
| pattern = compname + suffix + '_'.join(datename_str.rsplit('-', 1)) | |||
| pattern = compname + r'\.' + suffix + r'\.' + '_'.join(datename_str.rsplit('-', 1)) | |||
There was a problem hiding this comment.
Important fix that was needed on the E3SM side.
|
Tests passed. I don't think there is much here of interest to @jedwards4b , so I think it's OK to merge and let him review at a later date. |
Closed
jgfouca
pushed a commit
that referenced
this pull request
Aug 8, 2018
Vectorize loops in coupler attr-vector multiplications [BFB]
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Test suite: scripts_regression_tests
Test baseline:
Test namelist changes:
Test status: bit for bit
Fixes [CIME Github issue #]
User interface changes?: N
Update gh-pages html (Y/N)?: N
Code review: @jedwards4b