Skip to content
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.

Sign up for GitHub

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

Jump to bottom

Merge plan for PR #101: CUDA 12.0 conda support #112

Closed
13 of 17 tasks
bdice opened this issue Jul 11, 2023 · 5 comments
Closed
13 of 17 tasks

Merge plan for PR #101: CUDA 12.0 conda support #112

bdice opened this issue Jul 11, 2023 · 5 comments

Comments

@bdice
Copy link
Contributor

bdice commented Jul 11, 2023 •
edited
Loading

We're planning to merge PR #101 on Tuesday, July 18. Any repos using workflows that has not yet moved to the branch cuda-120 will need to make PRs that filter out CUDA 12 jobs from their GitHub Actions workflows like this:

matrix_filter: map(select(.CUDA_VER | startswith("11")))

See this cusignal PR demonstrating this matrix_filter.

Currently this list of repos is using branch-23.08. These need to migrate to use the shared-action-workflows branch cuda-120 before July 18 or introduce matrix filters that skip CUDA 12 builds.

Tasks

(Note that xgboost and other packages aren't using the same shared-action-workflows.)
@robertmaynard robertmaynard mentioned this issue Jul 12, 2023
Merged
3 tasks
rapids-bot bot pushed a commit to rapidsai/rapids-cmake that referenced this issue Jul 13, 2023
@robertmaynard
Update rapids-cmake ci to support conda-forge CUDA 12 (#437)
05be912 
Needed due to the changes coming to the rapids shared-action-workflow: rapidsai/shared-workflows#112

Authors:
  - Robert Maynard (https://github.com/robertmaynard)
  - Bradley Dice (https://github.com/bdice)

Approvers:
  - Bradley Dice (https://github.com/bdice)
  - Ray Douglass (https://github.com/raydouglass)

URL: #437
@jakirkham
Copy link
Member

Let's drop xgboost-conda as that repo is going away. It is being replaced with rapidsai/xgboost-feedstock

Do we still need the integration repo?

Am ok either skipping cuCIM or handling it with the others here. We are pretty much down to the CUDA 12 PR there and it is getting close

@bdice
Copy link
Contributor Author

bdice commented Jul 17, 2023

I don't think we can make a decision to "drop" any repos unless broken CI after merging PR #101 is a viable outcome for those repos.

@jakirkham
Copy link
Member

Not following

The xgboost-conda repo is being archived. Let's not spend time on that

The cuCIM repo is down to the CUDA 12 PR that is nearly done. Not sure it is worth changing the GHA branch to only change it back again

Idk the state of the integration repo. Maybe @raydouglass can chime in

@bdice
Copy link
Contributor Author

bdice commented Jul 17, 2023

We still rely on integration for the rapids metapackage, which is our primary conda installation entrypoint. See rapidsai/integration#662

@bdice bdice mentioned this issue Jul 17, 2023
Merged
raydouglass pushed a commit to rapidsai/cugraph that referenced this issue Jul 18, 2023
@bdice
Only run cugraph conda CI for CUDA 11. (#3713)
c33d4bb 
We are planning to enable CUDA 12 CI jobs on July 18, per rapidsai/shared-workflows#112. This PR must be merged to disable CUDA 12 CI jobs for cugraph until #3271 / #3456 are merged.

Authors:
   - Bradley Dice (https://github.com/bdice)

Approvers:
   - Ray Douglass (https://github.com/raydouglass)
@bdice
Copy link
Contributor Author

bdice commented Jul 18, 2023

This has been merged -- feel free to comment here if we've missed any repos for posterity.

@bdice bdice closed this as completed Jul 18, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
2 participants
@jakirkham @bdice