Skip to content

[FEA]: Add public class methods for retrieving the legacy default stream as well as the per-thread default stream #1445

Closed

Description

@XSShawnZeng

Is this a duplicate?

Area

cuda.core

Is your feature request related to a problem? Please describe.

Currently I am using stream = Stream.from_handle(0) for retrieving the legacy default stream in CUDA python samples.

Describe the solution you'd like

The Stream class already has _legacy_default() and _per_thread_default() class methods. It'd be better that they are made public to provide a clean, documented way to access default streams, similar to the existing Stream.from_handle() pattern.

Describe alternatives you've considered

No response

Additional context

No response

Activity

  1. added
    enhancementAny code-related improvements
    cuda.coreEverything related to the cuda.core module
    and removed
    triageNeeds the team's attention
    on Jan 9, 2026
  2. Monishver11 commented on Jan 11, 2026

    @Monishver11
    Contributor

    Hello @XSShawnZeng @kkraus14, may I work on this? If so, could you please assign it to me?

  3. leofang commented on Jan 14, 2026

    @leofang
    Member

    Feel free to send a PR! We don't use the assignment system for people outside of the team.

  4. Monishver11 commented on Jan 14, 2026

    @Monishver11
    Contributor

    Thanks for the reply, @leofang. I'll raise the PR shortly.

  5. added 2 commits that reference this issue on Jan 14, 2026
    61b5de2
    8c9a466
  6. Monishver11 commented on Jan 14, 2026

    @Monishver11
    Contributor

    Hello @leofang, I've made the changes and raised the PR. Could you review it and let me know if any modifications are needed? - Feature/public default stream methods #1483, Thanks.

  7. added a commit that references this issue on Jan 14, 2026
    264524b
  8. added a commit that references this issue on Jan 16, 2026
    04586be
  9. added a commit that references this issue on Feb 18, 2026
    73725e6
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    cuda.coreEverything related to the cuda.core moduleenhancementAny code-related improvementsgood first issueGood for newcomers

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions