-
Notifications
You must be signed in to change notification settings - Fork 47
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Update spack submodule from spack/develop as of 2024/01/26 #977
Update spack submodule from spack/develop as of 2024/01/26 #977
Conversation
…c/source/NewSiteConfigs.rst
…eature/update_spack_from_dev_20240126
…ommented-out version before merging, after extensive testing
…ck-stack/stack/stack_env.py
…eature/update_spack_from_dev_20240126
…eature/update_spack_from_dev_20240126
…eature/update_spack_from_dev_20240126
…eature/update_spack_from_dev_20240126
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Self-review
@@ -271,7 +271,8 @@ def write(self): | |||
os.makedirs(env_pkgs_path, exist_ok=False) | |||
with open(os.path.join(env_repo_path, "repo.yaml"), "w") as f: | |||
f.write("repo:\n namespace: envrepo") | |||
repo_paths = spack.config.get("repos", scope=spack.config.default_list_scope()) | |||
# DH* ??? | |||
repo_paths = spack.config.get("repos", scope=None) # scope=spack.config.default_list_scope()) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I think this works - need to test chained environments (use spack stack create env --upstream ...
).
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This looks okay (remember to test with --modify-pkg
option). Just using repo_paths = spack.config.get("repos")
with no scope argument seems to give the right behavior when I tested it just now.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Then let me change it to that, thanks!
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Done in 05d44b4 - CI does test chained environments, so we will see.
@@ -285,5 +285,3 @@ modules: | |||
- PKG_CONFIG_PATH | |||
lib64/pkgconfig: | |||
- PKG_CONFIG_PATH | |||
'': | |||
- CMAKE_PREFIX_PATH |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This was a duplicate, spack sets this variable automatically
@@ -284,5 +284,3 @@ modules: | |||
- PKG_CONFIG_PATH | |||
lib64/pkgconfig: | |||
- PKG_CONFIG_PATH | |||
'': |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
ditto
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I am building unified-env on my arm64 based Mac, and when I ran the list of spack find
commands, I got this in my site/packages.yaml file:
python:
externals:
- spec: [email protected]+bz2+crypt+ctypes+dbm+lzma+nis+pyexpat~pythoncmd+readline+sqlite3+ssl+t ix+tkinter+uuid+zlib
prefix: /usr
It didn't impact the python spec/requirement in that after concretizing spack decided to build [email protected] with no duplicates of the python spec. However, when I commented out the above section from the site/packages.yaml file, the spec for fms changed from 32-bit to 64-bit. Very strange, but I thought I would bring it to your attention in case it is of significance.
@@ -249,7 +249,7 @@ Remember to activate the ``lua`` module environment and have MacTeX in your sear | |||
# Check your clang version then add it to your site compiler config. | |||
clang --version | |||
spack config add "packages:all:compiler:[apple-clang@YOUR-VERSION]" | |||
spack config add "packages:all:providers:mpi:[openmpi@4.1.6]" | |||
spack config add "packages:all:providers:mpi:[openmpi@5.0.1]" |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This looks like a good update too. Perhaps this will help with the MacOS build issue #971?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
maybe
We shouldn't be finding any external Python. If that's the case, then we need to exclude that from the Regarding |
And as discussed yesterday, I've tested JEDI-Skylab with [email protected] extensively. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks for the quick responses. All looks good!
Thanks for your review. I am going to add the |
…Github CI scripts
Done in 0ade34d |
…os)' in spack-ext/lib/jcsda-emc/spack-stack/stack/stack_env.py
CI passed. I am going to merge this and then work on removing mysql from the default requirements on the HPCs and non-JCSDA user systems next. |
Summary
Required updates in spack-stack for JCSDA/spack#401 (merge
spack/develop
as of 2024/01/26 intojcsda_emc_spack_stack
). Unpin some package versions that we don't really care about (i.e. take the latest as long as it works for us).Reviewers: Some core spack functionality changed that we use in our spack extension, therefore please take a close look at
spack-ext/lib/*
.Testing
[email protected]
and[email protected]
. Build and run a c12 GEOS-GCM test case![email protected]
and[email protected]
and run small testsskylab-trace-gas
experiment unrelated to this PR (reported with spack-stack-1.6.0 since last week), other Skylab experiments tested ran fineApplications affected
All
Systems affected
Potentially all (although the changes are small) - but check in particular chained environments when using our
spack stack create env --upstream
functionality.Dependencies
Issue(s) addressed
This is the last part of compiling JEDI+GEOS with spack-stack (i.e. not using Baselibs) on more than just Discover. This PR resolves https://github.com/JCSDA-internal/AOP23/issues/33.
It also closes #918, because the update of
eccodes
to version2.33.0
is done in this PR (and the update from spack develop brings the new version to our fork).Checklist