NVIDIA / NVIDIA/cuCollections

[ENHANCEMENT]: Remove fast_int storage from dynamic valid extents

Open
#838 1 comment 1 reaction 0 assignees View on GitHub

Nobody has claimed this yet.

type: improvement
Dominant language
Cuda
Stars
667
Forks
120
Avg merge
7d 5h
Merged PRs (30d)
4

Description

Is your feature request related to a problem?

Blocked by #837.

Dynamic valid_extent currently inherits from cuco::utility::fast_int, storing the extent value plus precomputed division metadata.

After #837 removes modulo from probe advancement, the extent is no longer used as a runtime divisor. The additional state becomes unnecessary but is still constructed and copied through storage references and probing iterators.

Describe the solution you'd like

Store only the underlying SizeType in dynamic valid_extent, preserving its existing value access, conversions, and required arithmetic operations.

The standalone public cuco::utility::fast_int utility must remain available and unchanged to avoid breaking its API.

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start with the dynamic valid_extent implementation and trace its storage references and probing iterators, using the changes in #837 as context. Done means dynamic valid_extent stores only SizeType while preserving its value access, conversions, and required arithmetic, and the public cuco::utility::fast_int API remains unchanged.

Written by the indexing model from the issue text.

Assessment

Tech stack
cpp
Domain
data
Issue type
Refactor
Difficulty
3/5
Estimated time
1-2 days
Activity status
Active
Clarity
Mostly clear
Newbie friendliness
64/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.