[https://nvbugs/6018647][test] Add unit test for Lifecycle Race Condition error in disagg sever

[yingguo-trt]

Add regression test StoreBlocksForReuseWithPinDoesNotCreateGhostFreeBlocks to verify that storeBlocksForReuse(pin=true) correctly calls claimBlock() on zero-ref blocks before incRefCount(), preventing duplicate free queue entries that caused hangs and crashes in disaggregated serving with block reuse enabled.

Summary by CodeRabbit

• Tests
  • Added regression test for KV-cache block lifecycle management to validate correct behavior when enabling block reuse with pinned blocks, ensuring free-block counts remain accurate and preventing resource leaks.

Description

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.

[yingguo-trt]

/bot run

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: c37dcf4a-8a59-4882-afef-c68b3bba36af

📥 Commits

Reviewing files that changed from the base of the PR and between 63cb1f9 and 9c96699.

📒 Files selected for processing (1)

• cpp/tests/unit_tests/batch_manager/kvCacheManagerTest.cpp

---

📝 Walkthrough

Walkthrough

Added a regression test case that validates KV-cache block lifecycle behavior when block reuse and pinning are enabled together, ensuring no "ghost free" blocks inflate free-block counts during storeBlocksForReuse operations.

Changes

Cohort / File(s)	Summary
KV-Cache Block Management Test cpp/tests/unit_tests/batch_manager/kvCacheManagerTest.cpp	Added regression test StoreBlocksForReuseWithPinDoesNotCreateGhostFreeBlocks that validates block lifecycle during reuse with pinning, checking that free-block counts remain consistent.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~10 minutes

🚥 Pre-merge checks | ✅ 1 | ❌ 2

❌ Failed checks (2 warnings)

Check name	Status	Explanation	Resolution
Description check	⚠️ Warning	The description explains the test's purpose and what it validates, but required template sections (Description, Test Coverage, and PR Checklist details) are largely unfilled or left as comments.	Fill in the Description section with details about the lifecycle race condition bug, expand Test Coverage to describe which test safeguards the changes, and complete the PR Checklist items appropriately.
Docstring Coverage	⚠️ Warning	Docstring coverage is 33.33% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (1 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title accurately describes the main change: adding a regression unit test for a lifecycle race condition in disaggregated serving with block reuse.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

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

[tensorrt-cicd]

PR_Github #42122 [ run ] triggered by Bot. Commit: 9c96699 Link to invocation

[yingguo-trt]

/bot run

[tensorrt-cicd]

PR_Github #42134 [ run ] triggered by Bot. Commit: 9c96699 Link to invocation

[tensorrt-cicd]

PR_Github #42134 [ run ] completed with state SUCCESS. Commit: 9c96699
/LLM/main/L0_MergeRequest_PR pipeline #32969 completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

• Please check the failed tests and fix your PR
• If you cannot view the failures, ask the CI triggerer to share details
• Once fixed, request an NVIDIA team member to trigger CI again

Link to invocation

[yingguo-trt]

/bot run --skip-test

[tensorrt-cicd]

PR_Github #42226 [ run ] triggered by Bot. Commit: 9c96699 Link to invocation

[tensorrt-cicd]

PR_Github #42226 [ run ] completed with state SUCCESS. Commit: 9c96699
/LLM/main/L0_MergeRequest_PR pipeline #33039 (Partly Tested) completed with status: 'SUCCESS'

CI Report

Link to invocation

[yingguo-trt]

/bot reuse-pipeline

[tensorrt-cicd]

PR_Github #42244 [ reuse-pipeline ] triggered by Bot. Commit: 9c96699 Link to invocation

[tensorrt-cicd]

PR_Github #42244 [ reuse-pipeline ] completed with state SUCCESS. Commit: 9c96699
Reusing PR_Github #42226 (Partly Tested) for commit 9c96699

Link to invocation