Restore linux64 build parallelization#198
Conversation
|
@conda-forge-admin, please rerender |
|
Hi! This is the friendly automated conda-forge-linting service. I just wanted to let you know that I linted all conda-recipes in your PR ( |
…2026.05.29.18.35.1
conda-forge-admin
requested review from
bdice,
jakirkham,
kkraus14,
leofang,
m3vaz,
mmccarty,
oleksandr-pavlyk,
rparolin,
rwgk,
vyasr and
vzhurba01
as code owners
…2026.05.29.18.35.1
…utput
The CFEP-25 noarch:python recipe references only ${{ python_min }}, so
rattler-build does not include `python` in that output's variant.
Combined with the non-noarch `cuda-bindings` output (which does use
`python`), conda-smithy's _collapse_subpackage_variants takes the
intersection of used loop vars across outputs, drops `python`, and
collapses all Python variants into a single .ci_support file per
platform — serializing what used to be 6 parallel CI jobs.
Adding `build.variant.use_keys: [python]` to the noarch output forces
`python` back into its variant, restoring the per-Python matrix. The
redundant identical noarch builds dedupe on upload via hash collision.
Mirrors the same pattern used in conda-forge/root-feedstock for its
noarch:generic root_cxx_standard output.
|
@conda-forge-admin, please rerender |
…2026.05.29.18.35.1
|
@conda-forge-admin, please rerender |
|
Hi! This is the friendly automated conda-forge-webservice. I tried to rerender for you, but it looks like there was nothing to do. This message was generated by GitHub Actions workflow run https://github.com/conda-forge/conda-forge-webservices/actions/runs/26671619206. Examine the logs at this URL for more detail. |
Brings in v13.3.1 (sha 5d21c94f...) and the cuda-core/cudla test
additions. Resolves recipe/recipe.yaml `context.number` by setting
it to 1 to capture this PR's build-graph changes on top of the
13.3.1 release.
Also folds in the noarch-dedupe wiring on the cuda-python output:
* Adds explicit `build.string: py_min_${{ python_min }}_${{ number }}`
so all variant.use_keys: [python] builds produce the same artifact
filename and collide on upload (matches root-feedstock's pattern).
* Loosens the noarch test from `python ${{ python_min }}.*` to
`python >=${{ python_min }}` so the test environment can pick the
variant's Python and resolve against the locally-built
cuda-bindings ABI (matches v0 behaviour).
|
Hi! This is the friendly automated conda-forge-linting service. I just wanted to let you know that I linted all conda-recipes in your PR ( I do have some suggestions for making it better though... For recipe/recipe.yaml:
This message was generated by GitHub Actions workflow run https://github.com/conda-forge/conda-forge-webservices/actions/runs/26672111393. Examine the logs at this URL for more detail. |
|
@conda-forge-admin, please rerender |
|
Hi! This is the friendly automated conda-forge-webservice. I tried to rerender for you, but it looks like there was nothing to do. This message was generated by GitHub Actions workflow run https://github.com/conda-forge/conda-forge-webservices/actions/runs/26672151658. Examine the logs at this URL for more detail. |
|
Hi! This is the friendly conda-forge automerge bot! I considered the following status checks when analyzing this PR:
Thus the PR was passing and merged! Have a great day! |
|
It's so unfortunate that top-level requirements was what made the parallelization possible. Removing |
3 participants
Hi! This is the friendly automated conda-forge-webservice.
I've started rerendering the recipe as instructed in #197.
If I find any needed changes to the recipe, I'll push them to this PR shortly. Thank you for waiting!
Here's a checklist to do before merging.
Fixes #197