Skip to content

Conversation

PointKernel
Copy link
Member

@PointKernel PointKernel commented Oct 14, 2025

Description

This PR updates cuco-related components in cudf to ensure compatibility with the new cuco stream-ordered allocator.

Checklist

  • I am familiar with the Contributing Guidelines.
  • New or existing tests cover these changes.
  • The documentation is up to date with these changes.

@PointKernel PointKernel added the DO NOT MERGE Hold off on merging; see PR for details label Oct 14, 2025
Copy link

copy-pr-bot bot commented Oct 14, 2025

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

Contributors can view more details about this message here.

@PointKernel PointKernel changed the title Test cuco stream allocator1 Test cuco stream allocator Oct 14, 2025
@github-actions github-actions bot added libcudf Affects libcudf (C++/CUDA) code. CMake CMake build issue labels Oct 14, 2025
@PointKernel
Copy link
Member Author

/ok to test bc5ac51

@PointKernel
Copy link
Member Author

/ok to test 02f3696

@PointKernel PointKernel added non-breaking Non-breaking change improvement Improvement / enhancement to an existing function labels Oct 16, 2025
@PointKernel PointKernel changed the title Test cuco stream allocator Update the code to be compatible with the new cuco stream-ordered allocator Oct 16, 2025
@PointKernel PointKernel marked this pull request as ready for review October 16, 2025 01:45
@PointKernel PointKernel requested review from a team as code owners October 16, 2025 01:45
@PointKernel
Copy link
Member Author

This PR is ready for review. Please note that the upstream rapids-cmake PR depends on this one and cannot be merged until it’s approved. The temporary CMake changes here will be reverted once rapidsai/rapids-cmake#929 is merged.

@PointKernel PointKernel added the 3 - Ready for Review Ready for review by team label Oct 16, 2025
Copy link
Contributor

@ttnghia ttnghia left a comment

Choose a reason for hiding this comment

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

Is there any performance impact due to such upgrade?

@PointKernel
Copy link
Member Author

Is there any performance impact due to such upgrade?

Probably not for libcudf, since the upstream change only updated the cuco allocator from cudaMalloc/cudaFree to cudaMallocAsync/cudaFreeAsync. As libcudf doesn’t use the cuco allocator, this change shouldn’t have any impact on it.

@PointKernel PointKernel added 5 - Ready to Merge Testing and reviews complete, ready to merge and removed 3 - Ready for Review Ready for review by team labels Oct 16, 2025
@GregoryKimball GregoryKimball moved this to Burndown in libcudf Oct 17, 2025
Copy link
Member

@mhaseeb123 mhaseeb123 left a comment

Choose a reason for hiding this comment

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

Just adding a request changes to avoid merging this PR without accommodating for new code in #20201 to use the same rmm::polymorphic_allocator. See #20201 (review)

Update: #20201 has been merged and we can now make the required changes here.

@PointKernel PointKernel removed DO NOT MERGE Hold off on merging; see PR for details CMake CMake build issue labels Oct 22, 2025
@PointKernel
Copy link
Member Author

/merge

@rapids-bot rapids-bot bot merged commit 764c7e2 into rapidsai:main Oct 22, 2025
136 checks passed
@PointKernel PointKernel deleted the test-cuco-stream-alloc branch October 22, 2025 04:28
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

5 - Ready to Merge Testing and reviews complete, ready to merge improvement Improvement / enhancement to an existing function libcudf Affects libcudf (C++/CUDA) code. non-breaking Non-breaking change

Projects

Status: Burndown

Development

Successfully merging this pull request may close these issues.

3 participants