Skip to content

feat: support sgl-kernel cu129#8800

Merged
zhyncs merged 1 commit intomainfrom
zhyncs/129
Aug 5, 2025
Merged

feat: support sgl-kernel cu129#8800
zhyncs merged 1 commit intomainfrom
zhyncs/129

Conversation

@zhyncs
Copy link
Copy Markdown
Collaborator

@zhyncs zhyncs commented Aug 5, 2025

Motivation

Modifications

Accuracy Test

Benchmark & Profiling

Checklist

Copy link
Copy Markdown
Contributor

@gemini-code-assist gemini-code-assist Bot left a comment

Choose a reason for hiding this comment

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

Summary of Changes

Hello @zhyncs, I'm Gemini Code Assist1! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

I've introduced support for CUDA 12.9 into the sgl-kernel build system. This change primarily updates the build.sh script to correctly configure the build environment and PyTorch installation for systems utilizing CUDA 12.9, ensuring compatibility and enabling development with the latest CUDA version.

Highlights

  • CUDA 12.9 Build Support: The build script sgl-kernel/build.sh has been updated to recognize and handle CUDA version 12.9. This involves adding a new elif condition that sets the DOCKER_IMAGE and TORCH_INSTALL variables specifically for CUDA 12.9, pointing to the cu129 PyTorch test wheel.
Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point in your pull request via creating an issue comment (i.e. comment on the pull request page) using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands.

Feature Command Description
Code Review /gemini review Performs a code review for the current pull request in its current state.
Pull Request Summary /gemini summary Provides a summary of the current pull request in its current state.
Comment @gemini-code-assist Responds in comments when explicitly tagged, both in issue comments and review comments.
Help /gemini help Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments or fill out our survey to provide feedback.

You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.

Footnotes

  1. Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution.

Copy link
Copy Markdown
Contributor

@gemini-code-assist gemini-code-assist Bot left a comment

Choose a reason for hiding this comment

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

Code Review

This pull request aims to add support for CUDA 12.9. The change in build.sh is a good first step. I've added a review comment with a suggestion to refactor the conditional logic for setting TORCH_INSTALL, which will improve code clarity and make it easier to add support for future CUDA versions.

However, there is a critical omission. The sgl-kernel/rename_wheels.sh script, which is responsible for creating CUDA-specific wheel names, has not been updated. It contains a hardcoded check for CUDA 12.8 and will not work for 12.9. This will lead to incorrectly named wheels for CUDA 12.9 builds, defeating the purpose of this PR. Please update rename_wheels.sh to correctly handle CUDA 12.9 and consider making it more robust for future versions, for example by using the CUDA_VERSION environment variable.

@zhyncs zhyncs merged commit 194561f into main Aug 5, 2025
53 of 55 checks passed
@zhyncs zhyncs deleted the zhyncs/129 branch August 5, 2025 09:33
narutolhy pushed a commit to narutolhy/sglang that referenced this pull request Aug 17, 2025
MahmoudAshraf97 pushed a commit to MahmoudAshraf97/sglang that referenced this pull request Sep 8, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant