[None][fix] AutoDeploy: Unwaive llmc standalone tests

[bmarimuthu-nv]

Summary by CodeRabbit

• Improvements
  • Enhanced standalone package generation with expanded test file exclusions and runtime dependency validation to prevent accidental imports
  • Strengthened test isolation for standalone mode through stricter subprocess environment controls and upfront verification
  • Improved graph processing with topological order restoration before canonicalization
  • Made operation resolution more robust for dynamic dispatch

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)

• If PR introduces API changes, an appropriate PR label is added - either api-compatible or api-breaking. For api-breaking, include BREAKING in the PR title.

• 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.

[bmarimuthu-nv]

@CodeRabbit summary

[coderabbitai]

✅ Actions performed

Summary regeneration triggered.

[coderabbitai]

📝 Walkthrough

Walkthrough

Hardens standalone auto-deploy: add dependency guards and local constants, restore FX node topological order before canonicalization, safely resolve optional auto-deploy ops, expand standalone packaging exclusions and conftest import guard, and run standalone tests in an isolated subprocess environment with an import probe.

Changes

Standalone auto-deploy environment isolation

Layer / File(s)	Summary
Dependency conditionals for standalone mode tensorrt_llm/_torch/auto_deploy/models/quant_config_reader.py, tensorrt_llm/_torch/auto_deploy/models/custom/modeling_minimax_m2.py, tensorrt_llm/_torch/auto_deploy/transform/library/fuse_rope_mla.py	Modules now check TensorRT-LLM availability and provide fallback stubs for optional ModelOPT helpers, switch an ActivationType import source, and define the MLA RoPE info key locally to allow loading without the full dependency.
Graph ordering and op resolution utilities tensorrt_llm/_torch/auto_deploy/utils/_graph.py, tensorrt_llm/_torch/auto_deploy/utils/node_utils.py	Add _restore_topological_order to reorder FX nodes before canonicalization and _auto_deploy_op to safely resolve optional torch.ops.auto_deploy overloads; canonicalization and MOE detection are updated to use them.
Standalone packaging and conftest enforcement examples/auto_deploy/llmc/create_standalone_package.py	Expand EXCLUDE_TEST_FILES for the standalone package and generate a tests/conftest.py that checks importlib.util.find_spec("tensorrt_llm") and raises RuntimeError if the package is importable.
Standalone test environment isolation and validation tests/unittest/auto_deploy/standalone/test_standalone_package.py, tests/integration/test_lists/waives.txt	Construct a filtered isolated subprocess environment (strip host Python path vars, set PYTHONNOUSERSITE/PYTHONSAFEPATH), run an isolation probe using python -I to ensure tensorrt_llm is not importable, invoke pytest with the venv Python in isolated mode, and remove waivers so the tests run.

Estimated code review effort

🎯 4 (Complex) | ⏱️ ~45 minutes

Possibly related PRs

• NVIDIA/TensorRT-LLM#13790: Modifies tests/integration/test_lists/waives.txt to change which auto-deploy integration tests are SKIP-waived.
• NVIDIA/TensorRT-LLM#13504: Also updates the CI waiver list affecting which auto-deploy tests are waived.

Suggested reviewers

• govind-ramnarayan
• galagam
• jieli-matrix
• xinhe-nv
• StanleySun639

🚥 Pre-merge checks | ✅ 3 | ❌ 2

❌ Failed checks (2 warnings)

Check name	Status	Explanation	Resolution
Description check	⚠️ Warning	The PR description uses the provided template but leaves the critical Description and Test Coverage sections entirely empty, containing only comments placeholders and a completed checklist.	Fill in the Description section explaining why waives were removed and how test isolation improvements ensure tests can run safely, and populate Test Coverage with relevant test identifiers.
Docstring Coverage	⚠️ Warning	Docstring coverage is 44.44% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (3 passed)

Check name	Status	Explanation
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Title check	✅ Passed	The title clearly summarizes the main change: removing test waivers from LLMC standalone tests, which aligns with the file changes removing skip entries and strengthening the standalone test suite.

_{✏️ 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.}

[bmarimuthu-nv]

/bot run --disable-fail-fast

[coderabbitai]

Actionable comments posted: 0

[tensorrt-cicd]

PR_Github #50841 [ run ] triggered by Bot. Commit: 9a39702 Link to invocation

[tensorrt-cicd]

PR_Github #50841 [ run ] completed with state FAILURE. Commit: 9a39702
/LLM/main/L0_MergeRequest_PR pipeline #40308 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

CI Agent Failure Analysis

Link to invocation

[bmarimuthu-nv]

/bot run --disable-fail-fast

[tensorrt-cicd]

PR_Github #50871 [ run ] triggered by Bot. Commit: 9a39702 Link to invocation

[tensorrt-cicd]

PR_Github #50871 [ run ] completed with state FAILURE. Commit: 9a39702
/LLM/main/L0_MergeRequest_PR pipeline #40338 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

CI Agent Failure Analysis

Link to invocation

[bmarimuthu-nv]

/bot run --disable-fail-fast

[tensorrt-cicd]

PR_Github #50891 [ run ] triggered by Bot. Commit: 9a39702 Link to invocation

[tensorrt-cicd]

PR_Github #50891 [ run ] completed with state FAILURE. Commit: 9a39702
/LLM/main/L0_MergeRequest_PR pipeline #40355 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

CI Agent Failure Analysis

Link to invocation

[bmarimuthu-nv]

/bot run --disable-fail-fast

[tensorrt-cicd]

PR_Github #50955 [ run ] triggered by Bot. Commit: fd318c7 Link to invocation

[tensorrt-cicd]

PR_Github #50955 [ run ] completed with state FAILURE. Commit: fd318c7
/LLM/main/L0_MergeRequest_PR pipeline #40412 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

CI Agent Failure Analysis

Link to invocation

[bmarimuthu-nv]

/bot run --disable-fail-fast

[tensorrt-cicd]

PR_Github #51138 [ run ] triggered by Bot. Commit: b94a94c Link to invocation

[tensorrt-cicd]

PR_Github #51138 [ run ] completed with state SUCCESS. Commit: b94a94c
/LLM/main/L0_MergeRequest_PR pipeline #40574 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

CI Agent Failure Analysis

Link to invocation

[bmarimuthu-nv]

/bot run --disable-fail-fast

[tensorrt-cicd]

PR_Github #51414 [ run ] triggered by Bot. Commit: 1380b43 Link to invocation

[tensorrt-cicd]

PR_Github #51414 [ run ] completed with state FAILURE. Commit: 1380b43
/LLM/main/L0_MergeRequest_PR pipeline #40824 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

[bmarimuthu-nv]

/bot run --disable-fail-fast

[tensorrt-cicd]

PR_Github #51429 [ run ] triggered by Bot. Commit: 1380b43 Link to invocation

[tensorrt-cicd]

PR_Github #51429 [ run ] completed with state SUCCESS. Commit: 1380b43
/LLM/main/L0_MergeRequest_PR pipeline #40838 completed with status: 'SUCCESS'
Pipeline passed with automatic retried tests. Check the rerun report for details.

CI Report

Link to invocation