Skip to content

Conversation

@fbusato
Copy link
Contributor

@fbusato fbusato commented Dec 19, 2025

Description

The PR implements conversion utilities that take a (host/device/managed) mdspan and produce a DLTensor view of the same underlying memory.

The main issue is how to handle the memory associated with the shape and stride array of DLPack.
The implementation provides a small wrapper that owns the shape and strides arrays and stores a DLTensor whose shape/strides pointers refer into those arrays.

Todo:

  • documentation

@fbusato fbusato self-assigned this Dec 19, 2025
@fbusato fbusato requested a review from a team as a code owner December 19, 2025 00:55
@fbusato fbusato requested a review from wmaxey December 19, 2025 00:55
@fbusato fbusato added the 3.2.0 Targeted for 3.2.0 release label Dec 19, 2025
@fbusato fbusato added this to CCCL Dec 19, 2025
@github-project-automation github-project-automation bot moved this to Todo in CCCL Dec 19, 2025
@cccl-authenticator-app cccl-authenticator-app bot moved this from Todo to In Review in CCCL Dec 19, 2025
@github-actions

This comment has been minimized.

@github-project-automation github-project-automation bot moved this from In Review to In Progress in CCCL Dec 19, 2025
@fbusato fbusato requested a review from a team as a code owner December 19, 2025 18:21
@fbusato fbusato requested a review from gonidelis December 19, 2025 18:21
@github-actions

This comment has been minimized.

@davebayer
Copy link
Contributor

I was also wondering whether we don't want to introduce a special namespace (something like cuda::interop) for interoperability with other APIs apart from CTK

@github-actions

This comment has been minimized.

@fbusato fbusato moved this from In Progress to In Review in CCCL Dec 20, 2025
@github-actions
Copy link
Contributor

😬 CI Workflow Results

🟥 Finished in 1h 10m: Pass: 82%/91 | Total: 17h 13m | Max: 1h 10m | Hits: 97%/213965

See results here.

@fbusato fbusato mentioned this pull request Dec 23, 2025
1 task
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

3.2.0 Targeted for 3.2.0 release

Projects

Status: In Review

Development

Successfully merging this pull request may close these issues.

2 participants