Skip to content

CI: Always fetch the compute-sanitizer from the CTK 12.8.0 #593

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

Merged
merged 3 commits into from
May 1, 2025

Conversation

carterbox
Copy link
Contributor

@carterbox carterbox commented Apr 30, 2025

Description

We want to always use the latest compute-sanitizer because there could be bug fixes for the compute-sanitizer, and it is OK to use the newest compute-sanitizer on programs running with old CTK versions.

closes #594

Checklist

  • New or existing tests cover these changes.
  • The documentation is up to date with these changes.

Copy link
Contributor

copy-pr-bot bot commented Apr 30, 2025

Auto-sync is disabled for ready for review pull requests in this repository. Workflows must be run manually.

Contributors can view more details about this message here.

@carterbox
Copy link
Contributor Author

/ok to test b86f71c

@carterbox
Copy link
Contributor Author

Probably need to change the modules list so that we have a cache miss.

@carterbox
Copy link
Contributor Author

/ok to test b22711c

@carterbox
Copy link
Contributor Author

Closing in favor of different approach #594

@carterbox carterbox closed this Apr 30, 2025

This comment has been minimized.

@leofang
Copy link
Member

leofang commented May 1, 2025

Let me reopen this and look into it further

@leofang leofang reopened this May 1, 2025
Copy link
Member

@leofang leofang left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Let me purge the cache again and try this.

@leofang
Copy link
Member

leofang commented May 1, 2025

/ok to test 28b585b

@leofang leofang added enhancement Any code-related improvements P0 High priority - Must do! CI/CD CI/CD infrastructure labels May 1, 2025
@leofang leofang requested a review from rwgk May 1, 2025 17:08
Copy link
Collaborator

@rwgk rwgk left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM — I will try this in combination with #578 right now, if you want to wait for the outcome.

rwgk added a commit to rwgk/cuda-python that referenced this pull request May 1, 2025
@leofang
Copy link
Member

leofang commented May 1, 2025

Let's merge for now.

@leofang leofang merged commit bd770e1 into NVIDIA:main May 1, 2025
78 checks passed
rwgk added a commit to rwgk/cuda-python that referenced this pull request May 1, 2025
@leofang leofang changed the title CI: Always fetch the compute-sanitizer from the CTK 12.8.1 CI: Always fetch the compute-sanitizer from the CTK 12.8.0 May 1, 2025
@leofang leofang added this to the cuda-python 12.9.0 & 11.8.7 milestone May 1, 2025
Copy link

github-actions bot commented May 1, 2025

Doc Preview CI
Preview removed because the pull request was closed or merged.

leofang added a commit to carterbox/cuda-python that referenced this pull request May 2, 2025
…nitizer"

This reverts commit bd770e1, reversing
changes made to 19df0d9.
leofang added a commit that referenced this pull request May 3, 2025
* CI: Install latest compute-sanitizer separately from CTK

* Use mkdir -p to reuse CUDA_PATH

* BUG: Always start with clean CUDA_PATH when building cache

* REF: Setup CUDA component cache in TMP directory

Prevent collision with components already installed to the CUDA_PATH

* DEV: Skip non python-3.12 for debugging

* fix

* Revert "Merge pull request #593 from carterbox/dching/latest-sanitizer"

This reverts commit bd770e1, reversing
changes made to 19df0d9.

* use guess_latest.sh to get the latest CTK ver

* fix cache restoration

* fix cache restoration

* guess_latest needs wget

* dunno what I was thinking

* not sure if this would work on git for windows...

* fix

* give rsync a shot

* debug

* trailing slash, it's always the trailing slash

* for some reason --strip-components=1 does not work...

* the test runners do not have rsync

* restore full matrix

* move

* avoid rsync...

* fix

* add inline comments

---------

Co-authored-by: Leo Fang <[email protected]>
Co-authored-by: Leo Fang <[email protected]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
CI/CD CI/CD infrastructure enhancement Any code-related improvements P0 High priority - Must do!
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants