[https://nvbugs/5955173][fix] Add abort method for GenerationResultBa…

[JunyiXu-nv]

…se (#11970)

Summary by CodeRabbit

Release Notes

• New Features

  • Implemented an abort mechanism for generation operations, enabling users to cancel ongoing tasks and verify abort status.

• Tests

  • Added unit tests validating abort functionality and state transitions.

Description

Cherry-pick #11970

Test Coverage

PR Checklist

Please review the following before submitting your PR:

• PR description clearly explains what and why. If using CodeRabbit's summary, please make sure it makes sense.

• PR Follows TRT-LLM CODING GUIDELINES to the best of your knowledge.

• Test cases are provided for new code paths (see test instructions)

• Any new dependencies have been scanned for license and vulnerabilities

• CODEOWNERS updated if ownership changes

• Documentation updated as needed

• Update tava architecture diagram if there is a significant design change in PR.

• The reviewers assigned automatically/manually are appropriate for the PR.

• Please check this after reviewing the above items as appropriate for this PR.

GitHub Bot Help

To see a list of available CI bot commands, please comment /bot help.

[JunyiXu-nv]

/bot run

[coderabbitai]

Caution

Review failed

The pull request is closed.

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: aa4a8f06-8d3f-4322-b639-41e59a97f06a

📥 Commits

Reviewing files that changed from the base of the PR and between 8de01ac and e4279cb.

📒 Files selected for processing (3)

• tensorrt_llm/executor/postproc_worker.py
• tensorrt_llm/executor/result.py
• tests/unittest/llmapi/test_executor.py

---

📝 Walkthrough

Walkthrough

This change introduces a centralized abortion mechanism for generation results. A should_abort boolean flag is added to PostprocWorker.Output. When processing responses, the system checks this flag and triggers abort() on GenerationResultBase if set. GenerationResult now delegates abortion to its base class instead of managing abort state locally.

Changes

Cohort / File(s)	Summary
Abortion State Propagation tensorrt_llm/executor/postproc_worker.py, tensorrt_llm/executor/result.py	Introduces should_abort: bool field to PostprocWorker.Output and adds centralized abort() / aborted() methods to GenerationResultBase. Coordinates abortion via _handle_response checking the should_abort flag. GenerationResult now delegates to base class for abortion handling.
Unit Tests tests/unittest/llmapi/test_executor.py	Adds test coverage for GenerationResultBase.abort() / aborted() behavior, PostprocWorker.Output.should_abort flag defaults and construction, and cross-component abortion propagation through _handle_response.

Sequence Diagram(s)

sequenceDiagram
    participant PW as PostprocWorker
    participant HR as Response Handler<br/>(_handle_response)
    participant GRB as GenerationResultBase
    
    PW->>PW: Compute should_abort from record
    PW->>PW: Create Output(should_abort=True)
    PW->>HR: Emit Output with should_abort flag
    
    HR->>HR: Check response.should_abort
    
    alt should_abort == True and not already aborted
        HR->>GRB: abort()
        GRB->>GRB: Set _aborted = True
    end
    
    HR->>GRB: Query aborted()
    GRB-->>HR: Return _aborted state

Loading

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~20 minutes

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment

---

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[tensorrt-cicd]

PR_Github #38674 [ ] completed with state FAILURE. Commit: e4279cb
Not allowed on merged PR

Link to invocation