[https://nvbugs/6384747][fix] Add input_ids=None to MiniMaxM3MoeRoutingMethod.apply signature and forward to…

[tensorrt-cicd]

Summary

• Root cause: M3 routing-method override drops input_ids arg that parent (M2) and MoE scheduler caller both pass.
• Fix: Add input_ids=None to MiniMaxM3MoeRoutingMethod.apply signature and forward to super().apply(router_logits, input_ids); remove the matching waives.txt entry.
• Automated fix generated by repair-bot

Test plan

• Verify fix on the same GPU type as the original failure
• Check for regressions in related tests

Links

• Bug: https://nvbugs/6384747

Summary by CodeRabbit

• Bug Fixes
  • Improved routing behavior to better handle optional input IDs during model execution.
  • Removed an outdated test waiver entry, reflecting updated test expectations.

[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: Enterprise

Run ID: 5c8241ce-b0c5-43c7-be37-db91a3224355

📥 Commits

Reviewing files that changed from the base of the PR and between fde8e0a and 7a54d08.

📒 Files selected for processing (2)

• tensorrt_llm/_torch/modules/fused_moe/routing.py
• tests/integration/test_lists/waives.txt

💤 Files with no reviewable changes (1)

• tests/integration/test_lists/waives.txt

---

📝 Walkthrough

Walkthrough

MiniMaxM3MoeRoutingMethod.apply is updated to accept input_ids: Optional[torch.Tensor] = None and forward it to super().apply. The previously waived TestMiniMaxM3::test_auto_dtype[tp_size=8-ep_size=8] test on B200 is re-enabled by removing the skip entry.

MiniMaxM3 routing fix and test re-enable

Layer / File(s)	Summary
MiniMaxM3 apply signature fix and waiver removal tensorrt_llm/_torch/modules/fused_moe/routing.py, tests/integration/test_lists/waives.txt	MiniMaxM3MoeRoutingMethod.apply adds input_ids parameter and passes it to super().apply; the SKIP waiver for the corresponding B200 tp=8/ep=8 test is removed.

Estimated code review effort

🎯 1 (Trivial) | ⏱️ ~3 minutes

Possibly related PRs

• NVIDIA/TensorRT-LLM#15688: Adds the exact TestMiniMaxM3::test_auto_dtype[tp_size=8-ep_size=8] waiver entry that this PR removes.

Suggested reviewers

• crazydemo
• LarryXFly
• xinhe-nv

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly matches the bug fix and includes the correct NVBugs reference and API change.
Description check	✅ Passed	The description covers the problem, fix, test plan, and bug link, though it doesn’t use the repo’s exact template headings.
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.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

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

[BowenFu]

LGTM — waives.txt only; the code fix is already upstream and each newly-added waiver carries its own nvbug link (independent known failures, not masking this change).