[None][chore] Skip failing import of mxfp4_moe

[brb-nv]

Description

This MR skips a failing import to unblock pre-merge CI. Impacted test:
tests/unittest/_torch/auto_deploy/unit/multigpu/custom_ops/test_mxfp4_moe_ep.py::test_mxfp4_mlp_ep_dtypes

Test Coverage

N/A

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

• 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

/bot [-h] ['run', 'kill', 'skip', 'reuse-pipeline'] ...

Provide a user friendly way for developers to interact with a Jenkins server.

Run /bot [-h|--help] to print this help message.

See details below for each supported subcommand.

run [--reuse-test (optional)pipeline-id --disable-fail-fast --skip-test --stage-list "A10-PyTorch-1, xxx" --gpu-type "A30, H100_PCIe" --test-backend "pytorch, cpp" --add-multi-gpu-test --only-multi-gpu-test --disable-multi-gpu-test --post-merge --extra-stage "H100_PCIe-TensorRT-Post-Merge-1, xxx" --detailed-log --debug(experimental)]

Launch build/test pipelines. All previously running jobs will be killed.

--reuse-test (optional)pipeline-id (OPTIONAL) : Allow the new pipeline to reuse build artifacts and skip successful test stages from a specified pipeline or the last pipeline if no pipeline-id is indicated. If the Git commit ID has changed, this option will be always ignored. The DEFAULT behavior of the bot is to reuse build artifacts and successful test results from the last pipeline.

--disable-reuse-test (OPTIONAL) : Explicitly prevent the pipeline from reusing build artifacts and skipping successful test stages from a previous pipeline. Ensure that all builds and tests are run regardless of previous successes.

--disable-fail-fast (OPTIONAL) : Disable fail fast on build/tests/infra failures.

--skip-test (OPTIONAL) : Skip all test stages, but still run build stages, package stages and sanity check stages. Note: Does NOT update GitHub check status.

--stage-list "A10-PyTorch-1, xxx" (OPTIONAL) : Only run the specified test stages. Examples: "A10-PyTorch-1, xxx". Note: Does NOT update GitHub check status.

--gpu-type "A30, H100_PCIe" (OPTIONAL) : Only run the test stages on the specified GPU types. Examples: "A30, H100_PCIe". Note: Does NOT update GitHub check status.

--test-backend "pytorch, cpp" (OPTIONAL) : Skip test stages which don't match the specified backends. Only support [pytorch, cpp, tensorrt, triton]. Examples: "pytorch, cpp" (does not run test stages with tensorrt or triton backend). Note: Does NOT update GitHub pipeline status.

--only-multi-gpu-test (OPTIONAL) : Only run the multi-GPU tests. Note: Does NOT update GitHub check status.

--disable-multi-gpu-test (OPTIONAL) : Disable the multi-GPU tests. Note: Does NOT update GitHub check status.

--add-multi-gpu-test (OPTIONAL) : Force run the multi-GPU tests in addition to running L0 pre-merge pipeline.

--post-merge (OPTIONAL) : Run the L0 post-merge pipeline instead of the ordinary L0 pre-merge pipeline.

--extra-stage "H100_PCIe-TensorRT-Post-Merge-1, xxx" (OPTIONAL) : Run the ordinary L0 pre-merge pipeline and specified test stages. Examples: --extra-stage "H100_PCIe-TensorRT-Post-Merge-1, xxx".

--detailed-log (OPTIONAL) : Enable flushing out all logs to the Jenkins console. This will significantly increase the log volume and may slow down the job.

--debug (OPTIONAL) : Experimental feature. Enable access to the CI container for debugging purpose. Note: Specify exactly one stage in the stage-list parameter to access the appropriate container environment. Note: Does NOT update GitHub check status.

For guidance on mapping tests to stage names, see docs/source/reference/ci-overview.md
and the scripts/test_to_stage_mapping.py helper.

kill

kill

Kill all running builds associated with pull request.

skip

skip --comment COMMENT

Skip testing for latest commit on pull request. --comment "Reason for skipping build/test" is required. IMPORTANT NOTE: This is dangerous since lack of user care and validation can cause top of tree to break.

reuse-pipeline

reuse-pipeline

Reuse a previous pipeline to validate current commit. This action will also kill all currently running builds associated with the pull request. IMPORTANT NOTE: This is dangerous since lack of user care and validation can cause top of tree to break.

Summary by CodeRabbit

Release Notes

No user-visible changes

This release contains internal test infrastructure updates with no impact on end-user functionality.

[coderabbitai]

📝 Walkthrough

Walkthrough

A test file import of IS_TRITON_KERNELS_AVAILABLE was replaced with a locally defined constant set to False, with a FIXME note. This disables Triton kernel-related test execution paths.

Changes

Cohort / File(s)	Summary
Triton kernel test flag tests/unittest/_torch/auto_deploy/unit/multigpu/custom_ops/test_mxfp4_moe_ep.py	Replaced external import of IS_TRITON_KERNELS_AVAILABLE with local constant assignment to False (marked with FIXME), causing Triton kernel test paths to be skipped.

Estimated code review effort

🎯 1 (Trivial) | ⏱️ ~3 minutes

Pre-merge checks and finishing touches

✅ Passed checks (3 passed)

Check name	Status	Explanation
Title Check	✅ Passed	The pull request title "[None][chore] Skip failing import of mxfp4_moe" follows the required template format and directly corresponds to the main change in the changeset. According to the raw summary, the change replaces an external import with a locally defined constant set to False, effectively disabling Triton-kernel related test paths. The title succinctly captures this action of skipping the failing import without unnecessary noise or vague terminology, making it clear to teammates reviewing the history what the primary change accomplishes.
Description Check	✅ Passed	The pull request description includes all key template sections: Description, Test Coverage, and PR Checklist. The Description section explains the purpose (skipping a failing import to unblock pre-merge CI) and identifies the impacted test. The Test Coverage section appropriately marks N/A given this is a chore to disable problematic import paths rather than new functionality. The author has completed the PR checklist with the final checkbox checked. While the description is brief, it provides the essential information needed for this type of hotfix PR and follows the template structure appropriately.
Docstring Coverage	✅ Passed	Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%.

✨ Finishing touches

• 📝 Generate docstrings

🧪 Generate unit tests (beta)

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

---

📜 Recent review details

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between df689f8 and ae26efd.

📒 Files selected for processing (1)

• tests/unittest/_torch/auto_deploy/unit/multigpu/custom_ops/test_mxfp4_moe_ep.py (1 hunks)

🧰 Additional context used

📓 Path-based instructions (3)

**/*.{h,hpp,hh,hxx,cpp,cxx,cc,cu,cuh,py}

📄 CodeRabbit inference engine (CODING_GUIDELINES.md)

Use only spaces, no tabs; indent with 4 spaces.

Files:

• tests/unittest/_torch/auto_deploy/unit/multigpu/custom_ops/test_mxfp4_moe_ep.py

**/*.py

📄 CodeRabbit inference engine (CODING_GUIDELINES.md)

**/*.py: Python code must target Python 3.8+.
Indent Python code with 4 spaces; do not use tabs.
Maintain module namespace when importing; prefer 'from package.subpackage import foo' then 'foo.SomeClass()' instead of importing the class directly.
Python filenames should be snake_case (e.g., some_file.py).
Python classes use PascalCase names.
Functions and methods use snake_case names.
Local variables use snake_case; prefix 'k' for variables that start with a number (e.g., k_99th_percentile).
Global variables use upper SNAKE_CASE prefixed with 'G' (e.g., G_MY_GLOBAL).
Constants use upper SNAKE_CASE (e.g., MY_CONSTANT).
Avoid shadowing variables from an outer scope.
Initialize all externally visible members of a class in the constructor.
Prefer docstrings for interfaces that may be used outside a file; comments for in-function or file-local interfaces.
Use Google-style docstrings for classes and functions (Sphinx-parsable).
Document attributes and variables inline so they render under the class/function docstring.
Avoid reflection when a simpler, explicit approach suffices (e.g., avoid dict(**locals()) patterns).
In try/except, catch the most specific exceptions possible.
For duck-typing try/except, keep the try body minimal and use else for the main logic.

Files:

• tests/unittest/_torch/auto_deploy/unit/multigpu/custom_ops/test_mxfp4_moe_ep.py

**/*.{cpp,cxx,cc,h,hpp,hh,hxx,cu,cuh,py}

📄 CodeRabbit inference engine (CODING_GUIDELINES.md)

Prepend the NVIDIA Apache-2.0 copyright header with current year to the top of all source files (e.g., .cpp, .h, .cu, .py).

Files:

• tests/unittest/_torch/auto_deploy/unit/multigpu/custom_ops/test_mxfp4_moe_ep.py

🧬 Code graph analysis (1)

tests/unittest/_torch/auto_deploy/unit/multigpu/custom_ops/test_mxfp4_moe_ep.py (1)

tensorrt_llm/_torch/auto_deploy/distributed/common.py (1)

• spawn_multiprocess_job (248-252)

⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (1)

• GitHub Check: Pre-commit Check

🔇 Additional comments (1)

tests/unittest/_torch/auto_deploy/unit/multigpu/custom_ops/test_mxfp4_moe_ep.py (1)

10-12: Now I need to examine the actual test file to understand the complete context of the workaround:

Based on my verification through web searches and code analysis:

1. NVIDIA Internal Bug Tracking: NVIDIA uses nvbugs internal bug tracker, as evidenced in TensorRT-LLM release notes
2. Feature Context: TensorRT-LLM implements multiple MoE backends including TritonFusedMoE, and MXFP4 support was added for Llama and Mixtral models
3. Expert Parallelism: TensorRT-LLM supports Expert Parallel parallelism patterns for MoE structures
4. Triton Kernel MoE Support: OpenAI Triton kernel MoE support is used in both TensorRT-LLM and vLLM

The workaround is appropriate and correctly unblocks CI by skipping a failing import. The bug is being tracked via an internal NVIDIA tracker following their established patterns.

---

Skip failing import by hardcoding IS_TRITON_KERNELS_AVAILABLE = False to unblock CI.

The temporary fix appropriately handles the import failure and disables the associated test until the underlying MXFP4 MoE expert-parallel Triton kernel import issue (tracked in bug 5604136) is resolved. The FIXME comment with bug reference provides adequate traceability.

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

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

[coderabbitai]

📝 Walkthrough

Walkthrough

A test file was modified to replace an import of IS_TRITON_KERNELS_AVAILABLE from an external module with a local constant set to False, along with a FIXME comment. This ensures the test skips unconditionally without relying on external module state.

Changes

Cohort / File(s)	Summary
Test Skip Condition Update tests/unittest/_torch/auto_deploy/unit/multigpu/custom_ops/test_mxfp4_moe_ep.py	Replaced import of IS_TRITON_KERNELS_AVAILABLE with local constant definition set to False; added FIXME comment. Unconditionalizes the test skip behavior.

Estimated code review effort

🎯 1 (Trivial) | ⏱️ ~3 minutes

Pre-merge checks and finishing touches

✅ Passed checks (3 passed)

Check name	Status	Explanation
Title Check	✅ Passed	The title "[None][chore] Skip failing import of mxfp4_moe" follows the required template format and accurately relates to the core change in the pull request. The phrase "skip failing import" refers to the strategy of working around a failing import by replacing it with a local constant, which effectively prevents the import issue from blocking tests. While the title could be more technically precise about the mechanism (replacing an import with a local False constant), it clearly conveys the main objective and the affected component (mxfp4_moe).
Description Check	✅ Passed	The pull request description addresses all required template sections: a Description section explaining that a failing import is being skipped to unblock pre-merge CI with the specific impacted test identified, a Test Coverage section marked as "N/A" (appropriate for a chore/workaround PR), and the PR Checklist with the acknowledgment checkbox marked. While the Description could provide more technical detail about the implementation approach, it sufficiently explains the what (skipping a failing import) and why (unblocking CI), meeting the "mostly complete" standard for this type of PR.
Docstring Coverage	✅ Passed	No functions found in the changes. Docstring coverage check skipped.

✨ Finishing touches

• 📝 Generate docstrings

🧪 Generate unit tests (beta)

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

---

📜 Recent review details

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between df689f8 and 5accb4c.

📒 Files selected for processing (1)

• tests/unittest/_torch/auto_deploy/unit/multigpu/custom_ops/test_mxfp4_moe_ep.py (1 hunks)

🧰 Additional context used

📓 Path-based instructions (3)

**/*.{h,hpp,hh,hxx,cpp,cxx,cc,cu,cuh,py}

📄 CodeRabbit inference engine (CODING_GUIDELINES.md)

Use only spaces, no tabs; indent with 4 spaces.

Files:

• tests/unittest/_torch/auto_deploy/unit/multigpu/custom_ops/test_mxfp4_moe_ep.py

**/*.py

📄 CodeRabbit inference engine (CODING_GUIDELINES.md)

**/*.py: Python code must target Python 3.8+.
Indent Python code with 4 spaces; do not use tabs.
Maintain module namespace when importing; prefer 'from package.subpackage import foo' then 'foo.SomeClass()' instead of importing the class directly.
Python filenames should be snake_case (e.g., some_file.py).
Python classes use PascalCase names.
Functions and methods use snake_case names.
Local variables use snake_case; prefix 'k' for variables that start with a number (e.g., k_99th_percentile).
Global variables use upper SNAKE_CASE prefixed with 'G' (e.g., G_MY_GLOBAL).
Constants use upper SNAKE_CASE (e.g., MY_CONSTANT).
Avoid shadowing variables from an outer scope.
Initialize all externally visible members of a class in the constructor.
Prefer docstrings for interfaces that may be used outside a file; comments for in-function or file-local interfaces.
Use Google-style docstrings for classes and functions (Sphinx-parsable).
Document attributes and variables inline so they render under the class/function docstring.
Avoid reflection when a simpler, explicit approach suffices (e.g., avoid dict(**locals()) patterns).
In try/except, catch the most specific exceptions possible.
For duck-typing try/except, keep the try body minimal and use else for the main logic.

Files:

• tests/unittest/_torch/auto_deploy/unit/multigpu/custom_ops/test_mxfp4_moe_ep.py

**/*.{cpp,cxx,cc,h,hpp,hh,hxx,cu,cuh,py}

📄 CodeRabbit inference engine (CODING_GUIDELINES.md)

Prepend the NVIDIA Apache-2.0 copyright header with current year to the top of all source files (e.g., .cpp, .h, .cu, .py).

Files:

• tests/unittest/_torch/auto_deploy/unit/multigpu/custom_ops/test_mxfp4_moe_ep.py

🔇 Additional comments (1)

tests/unittest/_torch/auto_deploy/unit/multigpu/custom_ops/test_mxfp4_moe_ep.py (1)

8-10: Verify bug tracking and timeline for blocked import issue.

This change hardcodes IS_TRITON_KERNELS_AVAILABLE = False, causing test_mxfp4_mlp_ep_dtypes to skip unconditionally. The workaround is intentional and documented with a FIXME comment linking to bug 5604136, but it removes test coverage for the triton_mxfp4_moe_ep functionality.

Verify that:

1. Bug 5604136 is actively tracked with an assigned owner and defined timeline for resolution
2. The import issue will be addressed before this workaround becomes stale

Optional: Consider removing the commented-out import (line 9) if it serves no reference value.

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

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

[brb-nv]

/bot run

[tensorrt-cicd]

PR_Github #22205 [ run ] triggered by Bot. Commit: c3b6d5d

[brb-nv]

/bot skip --comment "skipping module import to unblock pre-merge"

[tensorrt-cicd]

PR_Github #22206 [ skip ] triggered by Bot. Commit: ae26efd

[tensorrt-cicd]

PR_Github #22205 [ run ] completed with state ABORTED. Commit: c3b6d5d
LLM/main/L0_MergeRequest_PR #16744 (Blue Ocean) completed with status: ABORTED

[lucaslie]

Here is the actual bug fix: #8593

[tensorrt-cicd]

PR_Github #22206 [ skip ] completed with state SUCCESS. Commit: ae26efd
Skipping testing for commit ae26efd