Skip to content

Use CCCL memory resource aliases directly - #5685

Draft
bdice wants to merge 3 commits into
rapidsai:mainfrom
bdice:cugraph-resource-ref-aliases
Draft

bdice wants to merge 3 commits into
rapidsai:mainfrom
bdice:cugraph-resource-ref-aliases

Conversation

@bdice

@bdice bdice commented Sep 29, 2026 •

Copy link
Copy Markdown
Contributor

Description

Use cuda::mr::device_resource_ref directly instead of rmm::device_async_resource_ref, and cuda::mr::any_device_resource instead of its concrete any_resource<device_accessible> spelling. Include <cuda/memory_resource> rather than introducing a cuGraph alias/header.

Keep direct <rmm/mr/per_device_resource.hpp> includes wherever the current-device resource accessor is used. The underlying types, ownership, and allocation behavior are unchanged.

Requires NVIDIA/cccl#11751, included by rapidsai/rapids-cmake#1097. Part of rapidsai/build-planning#337.

@bdice bdice added improvement Improvement / enhancement to an existing function non-breaking Non-breaking change labels Sep 29, 2026
@copy-pr-bot

copy-pr-bot Bot commented Sep 29, 2026

Copy link
Copy Markdown

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.

@bdice
bdice force-pushed the cugraph-resource-ref-aliases branch from 5d96486 to 1ecdeb4 Compare October 2, 2026 14:25
@bdice bdice changed the title Define cugraph::device_resource_ref over CCCL resource_ref type Use CCCL memory resource aliases directly Oct 2, 2026
@bdice
bdice force-pushed the cugraph-resource-ref-aliases branch from f22962f to 71faa38 Compare October 9, 2026 22:22
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

improvement Improvement / enhancement to an existing function non-breaking Non-breaking change

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant