feat(core): Implement parallel FC for read only tools.#18791
feat(core): Implement parallel FC for read only tools.#18791joshualitt merged 2 commits intomainfrom
Conversation
Summary of ChangesHello @joshualitt, 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 introduces a fundamental improvement to the tool execution pipeline by enabling parallel processing of read-only operations. This change enhances efficiency by allowing multiple non-mutating tasks to run concurrently, while maintaining sequential execution for tools that modify the system state. The core mechanism involves a new Highlights
Changelog
Activity
Using Gemini Code AssistThe 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 by creating a comment using either
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 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
|
There was a problem hiding this comment.
Code Review
This pull request introduces parallel execution for read-only tools, which is a great feature for improving performance. The implementation correctly identifies read-only tools and batches them for concurrent execution. However, I've identified a critical race condition in the CoreToolScheduler that could lead to incorrect state management when parallel tool calls complete. My review includes a detailed explanation of the issue. Addressing this is crucial for the stability of the new parallel execution logic.
d94b569 to
03d5ae3
Compare
|
Size Change: -9.53 kB (-0.04%) Total Size: 24.5 MB
ℹ️ View Unchanged
|
c8af2a9 to
54f6675
Compare
gundermanc
left a comment
There was a problem hiding this comment.
LGTM though I think we should reconsider whether the 'isReadonly' check is necessary. We'll need the ability to have subagents that write in parallel. I'd rather us figure out what's needed to make that work and skip the intermediate step.
a56fdf7 to
e821574
Compare
e821574 to
2789df5
Compare
1fca1a2 to
0aeeb02
Compare
0aeeb02 to
c66eed5
Compare
Fixes #17549