Concurrency data handling

[Shirley125]

Purpose

This PR addresses a race condition in the shared memory (SHM) connector that caused cannot mmap an empty file and MessagePack data is malformed errors under high concurrency. The issue occurred when the consumer attempted to read from an SHM segment before the producer had fully written the data or properly sized the segment.

Test Plan

The fix was developed to resolve concurrency issues observed with PR vllm-project#951. To validate, run the system under high concurrency and verify that the previously reported errors (cannot mmap an empty file, MessagePack data is malformed) no longer occur.

Test Result

Unit tests passed. The implemented changes are designed to prevent the reported concurrency errors by ensuring data integrity and proper synchronization during shared memory communication.

---

Essential Elements of an Effective PR Description Checklist

• The purpose of the PR, such as "Fix some issue (link existing issues this PR will resolve)".
• The test plan, such as providing test command.
• The test results, such as pasting the results comparison before and after, or e2e results
• (Optional) The necessary documentation update, such as updating supported_models.md and examples for a new model.
• (Optional) Release notes update. If your change is user facing, please update the release notes draft.

BEFORE SUBMITTING, PLEASE READ https://github.com/vllm-project/vllm-omni/blob/main/CONTRIBUTING.md

---

 

[cursor]

Cursor Agent can help with this pull request. Just @cursor in comments and I'll start working on changes in this branch.
_{Learn more about Cursor Agents}