Skip to content

Remove _CCCL_GRID_CONSTANT from DeviceSelectSweepKernel parameters - #9795

Merged
nanan-nvidia merged 1 commit into
NVIDIA:mainfrom
nanan-nvidia:select-if-nogc
Jul 10, 2026
Merged

nanan-nvidia merged 1 commit into
NVIDIA:mainfrom
nanan-nvidia:select-if-nogc

Conversation

@nanan-nvidia

Copy link
Copy Markdown
Contributor

Fixes #9763
Fixes #9765

For benchmarks: #9765 (comment)

@copy-pr-bot

copy-pr-bot Bot commented Jul 10, 2026

Copy link
Copy Markdown
Contributor

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

@nanan-nvidia

Copy link
Copy Markdown
Contributor Author

/ok to test 4c703f2

@nanan-nvidia
nanan-nvidia marked this pull request as ready for review July 10, 2026 20:42
@nanan-nvidia
nanan-nvidia requested a review from a team as a code owner July 10, 2026 20:42
@nanan-nvidia
nanan-nvidia requested a review from elstehle July 10, 2026 20:42
@cccl-authenticator-app cccl-authenticator-app Bot moved this from Todo to In Progress in CCCL Jul 10, 2026
@cccl-authenticator-app cccl-authenticator-app Bot moved this from In Progress to In Review in CCCL Jul 10, 2026
@nanan-nvidia nanan-nvidia changed the title Remove _CCCL_GRID_CONSTANT from DeviceSelectSweepKernel parametes Remove _CCCL_GRID_CONSTANT from DeviceSelectSweepKernel parameters Jul 10, 2026
@coderabbitai

coderabbitai Bot commented Jul 10, 2026

Copy link
Copy Markdown
Contributor

Review Change Stack

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info
⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Enterprise

Run ID: cd37f1c1-ac65-4978-961a-c0a2d34ea258

📥 Commits

Reviewing files that changed from the base of the PR and between f03362e and 4c703f2.

📒 Files selected for processing (1)
  • cub/cub/device/dispatch/dispatch_select_if.cuh

📝 Walkthrough

Summary by CodeRabbit

  • Bug Fixes
    • Improved compatibility and reliability for the device selection sweep operation without changing its public interface or behavior.

Walkthrough

The select sweep kernel entry point removes _CCCL_GRID_CONSTANT qualifiers from seven parameters while preserving their types, ordering, and remaining qualifiers.

Changes

Select dispatch kernel parameters

Layer / File(s) Summary
Update select sweep kernel parameters
cub/cub/device/dispatch/dispatch_select_if.cuh
DeviceSelectSweepKernel removes _CCCL_GRID_CONSTANT from its input, output, count, tile, and streaming-context parameters without changing their types or ordering.

Assessment against linked issues

Objective Addressed Explanation
Restore select_flagged performance in CCCL 3.4 on H100 [#9763] The qualifier removal may target the reported regression, but no benchmark or validation evidence is provided.
Restore select_if performance in CCCL 3.4 [#9765] The kernel declaration is changed, but the supplied summary does not confirm performance results across the affected GPUs.

Comment @coderabbitai help to get the list of available commands.

@github-actions

Copy link
Copy Markdown
Contributor

🥳 CI Workflow Results

🟩 Finished in 2h 34m: Pass: 100%/287 | Total: 12d 21h | Max: 2h 32m | Hits: 15%/1210034

See results here.

@nanan-nvidia
nanan-nvidia merged commit aeaa386 into NVIDIA:main Jul 10, 2026
310 of 311 checks passed
@github-project-automation github-project-automation Bot moved this from In Review to Done in CCCL Jul 10, 2026
@github-actions

Copy link
Copy Markdown
Contributor

Successfully created backport PR for branch/3.4.x:

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

Archived in project

Development

Successfully merging this pull request may close these issues.

select_if regresses in CCCL 3.4 select_flagged regresses in CCCL 3.4 on H100

2 participants