Skip to content

Conversation

yewentao256
Copy link
Member

@yewentao256 yewentao256 commented Jul 23, 2025

Purpose

A follow up PR for #21208

Fix Compile Warning 1444-D: type "cub::CUB_200802_SM_1000::Sum" was declared deprecated ("use cuda::std::plus instead")

Now tested on AMD MI300, this version should work.

Test

Using /mnt/nvme5n1p1/wentao/.venv/lib/python3.12/site-packages
Searching for mdurl==0.1.2
Best match: mdurl 0.1.2
Adding mdurl 0.1.2 to easy-install.pth file

Using /mnt/nvme5n1p1/wentao/.venv/lib/python3.12/site-packages
Finished processing dependencies for vllm==0.10.0rc2.dev74+g7c734ee09.d20250723.rocm624
(wentao) wentao@ENC1-CLS01-SVR08:~/vllm$ pip show vllm
Name: vllm
Version: 0.10.0rc2.dev74+g7c734ee09.d20250723.rocm624
Summary: A high-throughput and memory-efficient inference and serving engine for LLMs
Home-page: https://github.com/vllm-project/vllm

Signed-off-by: yewentao256 <[email protected]>
Copy link

👋 Hi! Thank you for contributing to the vLLM project.

💬 Join our developer Slack at https://slack.vllm.ai to discuss your PR in #pr-reviews, coordinate on features in #feat- channels, or join special interest groups in #sig- channels.

Just a reminder: PRs would not trigger full CI run by default. Instead, it would only run fastcheck CI which starts running only a small and essential subset of CI tests to quickly catch errors. You can run other CI tests on top of those by going to your fastcheck build on Buildkite UI (linked in the PR checks section) and unblock them. If you do not have permission to unblock, ping simon-mo or khluu to add you in our Buildkite org.

Once the PR is approved and ready to go, your PR reviewer(s) can run CI to test the changes comprehensively before merging.

To run CI, PR reviewers can either: Add ready label to the PR or enable auto-merge.

🚀

Copy link
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 addresses a compile-time deprecation warning for cub::Sum on CUDA by replacing it with the recommended cuda::std::plus<float>. The change is well-implemented using a type alias AddOp to handle the platform-specific code for CUDA and ROCm, ensuring that ROCm builds continue to use cub::Sum which is not deprecated on that platform. The changes are correct, targeted, and improve the maintainability of the codebase by resolving the warning. The code modifications are sound, and I have no further feedback.

@yewentao256
Copy link
Member Author

@houseroad CC, thanks~

@mgoin mgoin added bug Something isn't working ready ONLY add when PR is ready to merge/full CI is needed labels Jul 31, 2025
@mgoin
Copy link
Member

mgoin commented Jul 31, 2025

This seems safe, thanks

@mgoin mgoin enabled auto-merge (squash) July 31, 2025 21:43
@vllm-bot vllm-bot merged commit a59cd9d into vllm-project:main Aug 1, 2025
105 of 108 checks passed
npanpaliya pushed a commit to odh-on-pz/vllm-upstream that referenced this pull request Aug 6, 2025
jinzhen-lin pushed a commit to jinzhen-lin/vllm that referenced this pull request Aug 9, 2025
noamgat pushed a commit to noamgat/vllm that referenced this pull request Aug 9, 2025
paulpak58 pushed a commit to paulpak58/vllm that referenced this pull request Aug 13, 2025
diegocastanibm pushed a commit to diegocastanibm/vllm that referenced this pull request Aug 15, 2025
epwalsh pushed a commit to epwalsh/vllm that referenced this pull request Aug 28, 2025
zhewenl pushed a commit to zhewenl/vllm that referenced this pull request Aug 28, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
bug Something isn't working ready ONLY add when PR is ready to merge/full CI is needed
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants