hccl v2 interface

[luanyundu]

add compiler macro to control whether use v2 interface

[gemini-code-assist]

Summary of Changes

Hello @luanyundu, 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 focuses on upgrading and refining the underlying communication and data handling mechanisms for Mixture of Experts (MoE) operations. It introduces a new version of the HCCL interface, enabling future enhancements and compatibility. Concurrently, it updates batch size related attributes for finer control over token processing and streamlines the internal management of tiling data, contributing to a more robust and maintainable codebase.

Highlights

• HCCL V2 Interface Integration: Introduced a compiler macro USE_V2_INTERFACE to conditionally enable a new version (v2) of the HCCL (Huawei Collective Communication Library) interface across multiple MoE (Mixture of Experts) related operations. This allows for flexible switching between HCCL versions.
• Batch Size Attribute Update: The CamMoeCombineNormal operation's global_bs attribute has been replaced with real_max_bs, round, and per_round_tokens. This change provides more granular control over batch size and token processing, likely for optimization or improved resource management.
• Tiling Data Handling Refactoring: Refactored the way tiling data is accessed and initialized in several kernel files. This includes using GET_TILING_DATA_WITH_STRUCT for consistent data retrieval and inlining InitTilingData logic directly into Init functions in some cases, simplifying the code structure.
• Code Simplification: Many Init function signatures for MoE operations have been simplified by removing explicit tilingData, mc2InitTiling, and mc2CcTiling parameters, now passing a single tilingGM address, which is then parsed internally.

🧠 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 to enable Memory in your admin console.

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 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.

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 introduces a new v2 interface for HCCL, controlled by the USE_V2_INTERFACE compiler macro. The changes are extensive, touching many files to conditionally use hccl_.InitV2 and hccl_.SetCcTilingV2. A significant part of this PR is also a commendable refactoring effort, where GET_TILING_DATA_WITH_STRUCT calls are moved inside Init methods, simplifying call sites and reducing code duplication. Additionally, the attributes for cam_moe_combine_normal have been updated, replacing global_bs with more descriptive attributes like real_max_bs. The changes are consistent and well-executed across the codebase. I have one minor suggestion to improve an error message.

[gemini-code-assist]

There's a minor formatting issue in the error message. The two string literals will be concatenated, resulting in ...ranks.but.... Adding a space and capitalizing 'But' will improve readability.

                "realMaxBs is invalid, only support 0 or value greater than or equal to the BS of all ranks. "
                "But got realMaxBs=%ld, bs=%ld",