Skip to content

[FEA] Add from_handle() methods for Kernel. - #1413

Merged
rparolin merged 22 commits into
NVIDIA:mainfrom
rparolin:rparolin/192_issue_fix
Jan 14, 2026
Merged

[FEA] Add from_handle() methods for Kernel. #1413
rparolin merged 22 commits into
NVIDIA:mainfrom
rparolin:rparolin/192_issue_fix

Conversation

@rparolin

@rparolinrparolin commented Dec 19, 2025•
edited
Loading

Copy link
Copy Markdown
Collaborator

Fixes#192

This PR implements from_handle() static methods for Kernel classes enabling interoperability with foreign CUDA handles.

✨ Key Changes

  1. Kernel.from_handle()
  2. Thread Safety Improvements
    2.1. Added thread-safe lazy initialization using double-checked locking pattern
    2.2. Refactored global state access through helper functions that ensure initialization

@copy-pr-bot

copy-pr-botBot commented Dec 19, 2025

Copy link
Copy Markdown
Contributor

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

@rparolinrparolin changed the title Rparolin/192 issue fix[core] Add from_handle() methods for Program, ObjectCode, and Kernel. Dec 19, 2025
@rparolinrparolin changed the title [core] Add from_handle() methods for Program, ObjectCode, and Kernel. [FEA] Add from_handle() methods for Program, ObjectCode, and Kernel. Dec 19, 2025
@rparolin

rparolin commented Dec 19, 2025

Copy link
Copy Markdown
CollaboratorAuthor

/ok to test 30dfcdc

@rparolin
rparolin marked this pull request as ready for review December 19, 2025 22:27
@rparolin

rparolin commented Dec 19, 2025

Copy link
Copy Markdown
CollaboratorAuthor

/ok to test 47ad63e

@github-actions

This comment has been minimized.

Sign in to view
leofang
leofang requested changes Dec 20, 2025
View reviewed changes
Comment threadcuda_core/cuda/core/_module.py
@rparolin

rparolin commented Jan 12, 2026

Copy link
Copy Markdown
CollaboratorAuthor

/ok to test de17dc8

@rparolinrparolin changed the title [FEA] Add from_handle() methods for Program, ObjectCode, and Kernel. [FEA] Add from_handle() methods for Kernel. Jan 12, 2026
leofang
leofang requested changes Jan 13, 2026
View reviewed changes
Comment threadcuda_core/cuda/core/_module.py Outdated
Comment threadcuda_core/cuda/core/_module.py Outdated
Comment threadcuda_core/tests/test_module.py Outdated
@rparolin

rparolin commented Jan 13, 2026

Copy link
Copy Markdown
CollaboratorAuthor

/ok to test 5f01231

@rparolin
rparolin requested a review from leofangJanuary 13, 2026 19:41
@rparolin
rparolin enabled auto-merge (squash) January 13, 2026 20:43
@rparolin

rparolin commented Jan 13, 2026

Copy link
Copy Markdown
CollaboratorAuthor

/ok to test c9cca32

@rparolin
rparolin merged commit 39e3293 into NVIDIA:mainJan 14, 2026
80 checks passed
@leofangleofang added this to the cuda.core beta 12 milestone Jan 14, 2026
@github-actions

github-actionsBot commented Jan 14, 2026

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

Sign up for freeto join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@leofangleofangleofang approved these changes
@cpcloudcpcloudAwaiting requested review from cpcloud

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

cuda.core v0.6.0

Development

Successfully merging this pull request may close these issues.

Implement from_handle() for all cuda.core objects

2 participants

@rparolin@leofang