gemini-code-assist[bot] commented on PR #18823: URL: https://github.com/apache/tvm/pull/18823#issuecomment-3961251330
## Summary of Changes Hello @ksgr5566, I'm Gemini Code Assist[^1]! 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 enhances WebGPU support by introducing a flexible mechanism for utilizing subgroup shuffle primitives. Recognizing that not all WebGPU devices support subgroups, the changes ensure that shared memory reductions are used as a default, universally compatible approach. For devices that do support subgroups, a new target parameter (`supports_subgroups`) and an associated CLI flag (`--enable-subgroups` in MLC-LLM) allow for their optional activation, leading to potential performance improvements. This gating logic is applied at the reduction path selection level, preventing the generation of unsupported subgroup operations. ### Highlights * **Conditional Subgroup Shuffles**: Implemented gating logic to enable WebGPU subgroup shuffle primitives only when explicitly supported by the target device. * **Default to Shared Memory**: Ensured that shared memory reductions are used by default for universal compatibility across WebGPU devices. * **Target Parameter and CLI Flag**: Introduced a `supports_subgroups` target parameter and a corresponding `thread_warp_size` adjustment, along with a `--enable-subgroups` CLI flag in MLC-LLM, to allow users to opt-in to subgroup optimizations. 🧠**New Feature in Public Preview:** You can now enable **Memory** to help **Gemini Code Assist** learn from your team's feedback. This makes future code reviews more consistent and personalized to your project's style. **Click [here](https://codeassist.google/code-review/login) to enable Memory in your admin console.** <details> <summary><b>Changelog</b></summary> * **src/target/source/codegen_webgpu.cc** * Enabled the `subgroups` WGSL extension in the generated header. * Initialized the `enable_subgroups_` flag based on target attributes. * **src/target/source/codegen_webgpu.h** * Added a boolean member `enable_subgroups_` to track subgroup support. * **src/target/source/intrin_rule_webgpu.cc** * Defined WebGPU-specific intrinsic rules to map generic `tvm_warp_shuffle` operations to `tir.webgpu.subgroup_shuffle` built-ins. * **src/target/target_kind.cc** * Implemented `UpdateWebGPUAttrs` to dynamically set `thread_warp_size` to 32 when `supports_subgroups` is true. * Registered new target attributes for WebGPU, including `supports_subgroups` and `thread_warp_size`. * **src/tir/transforms/lower_thread_allreduce.cc** * Modified the `WarpShuffle` function to conditionally cast the delta argument to `uint32` for WebGPU. * Extended `IsWarpReduction` to include WebGPU as a target for warp reductions, while disabling the warp shuffle mask for it. * **web/package-lock.json** * Updated various JavaScript package dependencies, including `@babel` components and `semver`. * **web/src/webgpu.ts** * Added logic to detect and request the `subgroups` feature when initializing a WebGPU adapter. </details> <details> <summary><b>Activity</b></summary> * No specific activity was provided in the context for this pull request. </details> <details> <summary><b>Using Gemini Code Assist</b></summary> <br> The full guide for Gemini Code Assist can be found on our [documentation page](https://developers.google.com/gemini-code-assist/docs/review-github-code), here are some quick tips. <b>Invoking Gemini</b> You can request assistance from Gemini at any point by creating a comment using either `/gemini <command>` or `@gemini-code-assist <command>`. Below is a summary of the supported commands on the current page. 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 pull request comments and review comments. Help | `/gemini help` | Displays a list of available commands. <b>Customization</b> 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](https://developers.google.com/gemini-code-assist/docs/customize-gemini-behavior-github). <b>Limitations & Feedback</b> Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with :thumbsup: and :thumbsdown: 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](https://google.qualtrics.com/jfe/form/SV_2cyuGuTWsEw84yG). <b>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](https://cloud.google.com/products/gemini/code-assist).</b> </details> [^1]: Review the [Privacy Notices](https://policies.google.com/privacy), [Generative AI Prohibited Use Policy](https://policies.google.com/terms/generative-ai/use-policy), [Terms of Service](https://policies.google.com/terms), and learn how to configure Gemini Code Assist in GitHub [here](https://developers.google.com/gemini-code-assist/docs/customize-gemini-behavior-github). Gemini can make mistakes, so double check it and [use code with caution](https://support.google.com/legal/answer/13505487). -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
