Bugfix: some typos in Persistent kernel

[Edenzzzz]

📌 Description

tests pass

🔍 Related Issues

🚀 Pull Request Checklist

Thank you for contributing to FlashInfer! Before we review your pull request, please make sure the following items are complete.

✅ Pre-commit Checks

• I have installed pre-commit by running pip install pre-commit (or used your preferred method).
• I have installed the hooks with pre-commit install.
• I have run the hooks manually with pre-commit run --all-files and fixed any reported issues.

If you are unsure about how to set up pre-commit, see the pre-commit documentation.

🧪 Tests

• Tests have been added or updated as needed.
• All tests are passing (unittest, etc.).

Reviewer Notes

[gemini-code-assist]

Summary of Changes

Hello @Edenzzzz, 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!

This pull request addresses several issues and introduces optimizations within the Persistent attention kernel, primarily focusing on memory management, indexing, and numerical stability. It refactors data structures for better handling of len_kv_chunk and adjusts memory allocations for attention outputs, ensuring correct behavior and improved performance, especially for larger sequence lengths and multi-head configurations.

Highlights

• Persistent Kernel Refinements: Corrected len_kv_chunk access, introduced a normalize_d function for online softmax, and optimized kv_chunk_idx calculation.
• Memory Allocation Adjustments: Increased the float_workspace_buffer size and adjusted partial_o_offset and partial_lse_offset calculations to prevent under-allocation, particularly for multi-head attention.
• Data Structure Refactoring: Modified the HolisticPlanInfo struct to centralize len_kv_chunk_offset, streamlining data management for attention planning.
• Numerical Stability & Edge Case Handling: Ensured lse tensor has float32 dtype and added a max(..., 0) check for packed_causal_kv_end to prevent negative values.
• Test Coverage Expansion: Added new test cases for large sequence lengths and masked tokens, and included LSE output assertion to verify numerical correctness.

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. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here.

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. ↩

[gemini-code-assist]

Code Review

This pull request provides several important bug fixes and refactorings for the persistent attention kernel, which go beyond simple typo corrections. Key improvements include correcting memory allocation for intermediate buffers, fixing a potential negative index bug, and refactoring the len_kv_chunk calculation for better efficiency. The Python tests have also been enhanced with more comprehensive checks. I've identified one potential issue with a performance heuristic that I'd like you to look at.