[https://nvbugs/6644644][fix] Replay both verified orphan payloads onto current origin/main as a single… by trtllm-agent · Pull Request #18142 · NVIDIA/TensorRT-LLM

trtllm-agent · 2026-08-24T11:46:28Z

Summary

Root cause: Four Sm100BlockScaledContiguous*Runner names are imported at module scope but defined only inside cute_dsl_custom_ops' else-less if IS_CUTLASS_DSL_AVAILABLE: block, so on a cutlass-less host the ImportError propagates via create_moe → _torch/models → create_input_processor and kills every LLM() construction; fixing that unmasks four tests calling DSL-only ops with no availability mark.
Fix: Replay both verified orphan payloads onto current origin/main as a single commit — function-local import at the sole dominated use site, plus the standard IS_CUTLASS_DSL_AVAILABLE skipif the sibling cute-DSL test modules already carry.
Original test: pytest "tests/unittest/_torch/multimodal/test_mm_encoder_standalone.py::test_kv_event_mm_keys_with_uuid[False-hex]" -v
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/6644644

Dev Engineer Review

Guarded Sm100BlockScaledContiguous*Runner imports with IS_CUTLASS_DSL_AVAILABLE.
Prevented import failures on hosts without Cutlass DSL.
Kept _TILE_SIZE_CHECKED_RUNNERS empty when Cutlass DSL is unavailable.
Updated tactic compatibility checks to use the guarded runner tuple.
Preserved behavior when Cutlass DSL is installed.
Introduced no public API, configuration, or test-list changes.

QA Engineer Review

Modified tests/unittest/_torch/thop/parallel/test_cute_dsl_moe.py.
Added Cutlass DSL availability skips to four Blackwell NVFP4/CUTLASS DSL tests.
Preserved the existing GPU architecture checks.
Added no, modified no, and removed no test functions.
The affected tests are not covered in tests/integration/test_lists/.
Verdict: needs follow-up.

…used Two layers of one optional-dependency defect, landed together because they are the same bug at two sites and the failing frame must stay visible in this commit's own scope. Layer 1 -- tensorrt_llm/_torch/modules/fused_moe/fused_moe_cute_dsl.py cute_dsl_custom_ops' `if IS_CUTLASS_DSL_AVAILABLE:` block runs to end of file with no else-branch, so everything it defines is simply absent when the optional cutlass DSL is not installed. This module imported four Sm100BlockScaledContiguous*Runner names from it at module scope. Only GroupedGemmInputsHelper, also in that import, is top-level; the other four are guard-only. Because create_moe imports this module eagerly and it sits under _torch.models, the resulting ImportError propagated through fused_moe/__init__.py, modeling_utils.py, _torch/models/__init__.py and registry.create_input_processor to llm.py's _build_model -- so on a cutlass-less Blackwell host every LLM() construction died at import, well past the CuTe backends. That is what failed this test: it builds a multimodal encoder LLM and never reached its own assertions. The four names are needed at exactly one place, the isinstance check in runner_tactic_comb_checker, and they are flat TunableRunner leaves with no subclasses, no __all__ and no other reference in the tree. Reaching that line means a CuteDslFusedMoENvfp4Runner is already under autotune, so the DSL is necessarily installed. Import them there instead of at module scope, matching how cute_dsl_mla.py and dsa/metadata.py already reach into this guard from modules that must stay importable. This adds no module-scope binding, no fallback branch and no new runtime state, so there is no reachable path where the feature is quietly off: with the DSL present the runners resolve and the isinstance tuple is unchanged; without it those symbols never existed, and the crash is what is removed, not a capability. The provider side was ruled out mechanically rather than by preference -- cute_dsl_custom_ops.py fails confidentiality-scan on unmodified origin/main via pre-existing upstream prose, so no commit touching it can ship. Layer 2 -- tests/unittest/_torch/thop/parallel/test_cute_dsl_moe.py Fixing layer 1 makes this module collectable for the first time, which exposes a second missing guard: four nvfp4 grouped-GEMM tests call torch.ops.trtllm.cute_dsl_nvfp4_* ops registered only inside the same availability block, but carried only an SM-version skip. On an SM100 host without the DSL that skip passes, so they ran to the op call and died with AttributeError on the trtllm op namespace. Add the same IS_CUTLASS_DSL_AVAILABLE mark the other cute-DSL test modules already use (test_fp4_linear.py, test_indexer_topk.py). Coverage strictly expands, so this is not a waiver: with the DSL absent the file goes from 0 tests collected (collection error) to 420 run and 360 skipped; with the DSL present it goes from 0 to all 780 run. Nothing that previously executed stops executing. Verified on B200: target test 1 passed, and the full module now exits 0 with 420 passed / 360 skipped, the skips printing as [72] [72] [72] [144]. Signed-off-by: trtllm-agent <296075020+trtllm-agent@users.noreply.github.com>

coderabbitai · 2026-08-24T11:52:17Z

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: f353c49a-bc50-4a08-a4a9-34542f346865

📥 Commits

Reviewing files that changed from the base of the PR and between 3715cc6 and 983462a.

📒 Files selected for processing (1)

tensorrt_llm/_torch/modules/fused_moe/fused_moe_cute_dsl.py

💤 Files with no reviewable changes (1)

tensorrt_llm/_torch/modules/fused_moe/fused_moe_cute_dsl.py

Included review availability: Your plan provides up to 12 included reviews per hour; 10 remain after this review.

Walkthrough

The fused MoE module conditionally imports CuteDSL runners and checks tile-size tactics only when CuteDSL is available. Related Blackwell NVFP4 tests skip when the dependency is unavailable.

Changes

CuteDSL availability handling

Layer / File(s)	Summary
Conditional CuteDSL runner support `tensorrt_llm/_torch/modules/fused_moe/fused_moe_cute_dsl.py`	CuteDSL runner imports and tile-size runner checks are conditional on `IS_CUTLASS_DSL_AVAILABLE`.
Availability-gated Blackwell tests `tests/unittest/_torch/thop/parallel/test_cute_dsl_moe.py`	Four Blackwell NVFP4 and grouped GEMM tests skip when CuteDSL is unavailable.

Estimated code review effort: 2 (Simple) | ~10 minutes

Merge Risk: ⚪ Minimal · up to 98346

The localized import and test-skip changes leave no actionable merge-blocking risk at the current head; the PR is merge-ready after normal checks and review.

Suggested reviewers: xxi-nv, leslie-fang25

🚥 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%. Docstring coverage is scoped to functions touched by this diff. Analyzed 1 functions across 2 files.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Description check	✅ Passed	The description explains the root cause, fix, affected tests, test plan, and linked bug. It omits the template headings and checklist, but it contains the critical information needed to understand the…
Title check	✅ Passed	The title uses the required NVBugs and fix format and refers to the commit replay described in the objectives. It does not clearly state the primary code fix, but it remains related to the pull reques…
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.

Full details: Description check

Explanation

The description explains the root cause, fix, affected tests, test plan, and linked bug. It omits the template headings and checklist, but it contains the critical information needed to understand the change.

Full details: Title check

Explanation

The title uses the required NVBugs and fix format and refers to the commit replay described in the objectives. It does not clearly state the primary code fix, but it remains related to the pull request.

Fix all pre-merge checks with AI

✨ Finishing Touches

🧪 Generate unit tests (beta)

Create PR with unit tests

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

xxi-nv · 2026-08-25T00:41:09Z

It looks it was duplicating with #15986

sunnyqgg

LGTM

2ez4bz · 2026-08-26T06:02:21Z

/bot run

tensorrt-cicd · 2026-08-26T06:07:53Z

PR_Github #69337 [ run ] triggered by Bot. Commit: b60e874 Link to invocation

tensorrt-cicd · 2026-08-26T08:18:19Z

PR_Github #69337 [ run ] completed with state FAILURE. Commit: b60e874
/LLM/main/L0_MergeRequest_PR pipeline #56683 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

2ez4bz · 2026-08-26T16:02:21Z

/bot run

tensorrt-cicd · 2026-08-26T16:28:15Z

PR_Github #69466 [ run ] triggered by Bot. Commit: b60e874 Link to invocation

tensorrt-cicd · 2026-08-26T18:40:46Z

PR_Github #69466 [ run ] completed with state SUCCESS. Commit: b60e874
/LLM/main/L0_MergeRequest_PR pipeline #56797 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

2ez4bz · 2026-08-26T19:47:28Z

/bot run

tensorrt-cicd · 2026-08-26T19:52:50Z

PR_Github #69516 [ run ] triggered by Bot. Commit: b60e874 Link to invocation

tensorrt-cicd · 2026-08-26T23:24:25Z

PR_Github #69516 [ run ] completed with state SUCCESS. Commit: b60e874
/LLM/main/L0_MergeRequest_PR pipeline #56842 completed with status: 'UNSTABLE'

CI Report

⚠️ Multi-GPU Label Required:
Multi-GPU tests require the ci: full pre-merge approved label on this PR. Ask a member of NVIDIA/trt-llm-ci-approvers to add the label, then re-trigger CI with the same bot command (no rebase needed).

⚠️ 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

Signed-off-by: William Zhang <133824995+2ez4bz@users.noreply.github.com>

coderabbitai · 2026-08-28T04:05:31Z

Note

GitHub couldn't provide a complete incremental comparison for this pull request, so CodeRabbit is performing a full review instead. This review may take a little longer.

2ez4bz · 2026-08-28T04:06:00Z

/bot help

github-actions · 2026-08-28T04:06:10Z

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.

Details

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) --high-priority]

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. Supports wildcard * for pattern matching (e.g., "*PerfSanity*" matches all stages containing PerfSanity). Examples: "A10-PyTorch-1, xxx", "PerfSanity". The patterns "*", "*Post-Merge*", and "*PerfSanity*", including equivalent escaped or repeated-star forms and their use in comma-separated lists, require the ci: post-merge approved PR label. 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. Requires the ci: full pre-merge approved label on the PR (ask a member of NVIDIA/trt-llm-ci-approvers). 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. Requires the ci: full pre-merge approved label on the PR (ask a member of NVIDIA/trt-llm-ci-approvers).

--post-merge (OPTIONAL) : Run the L0 post-merge pipeline instead of the ordinary L0 pre-merge pipeline. Requires the ci: post-merge approved PR label applied by an active member of NVIDIA/trt-llm-ci-approvers. The approval label remains in place when new commits are pushed.

--extra-stage "H100_PCIe-TensorRT-Post-Merge-1, xxx" (OPTIONAL) : Run the ordinary L0 pre-merge pipeline and specified test stages. Supports wildcard * for pattern matching. Examples: --extra-stage "H100_PCIe-TensorRT-Post-Merge-1, xxx", --extra-stage "Post-Merge". The patterns "*", "*Post-Merge*", and "*PerfSanity*", including equivalent escaped or repeated-star forms and their use in comma-separated lists, require the ci: post-merge approved PR label.

--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.

--high-priority (OPTIONAL) : Run the pipeline with high priority. This option is restricted to authorized users only and will route the job to a high-priority queue.

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.

2ez4bz · 2026-08-28T04:06:33Z

/bot run --add-multi-gpu-test

tensorrt-cicd · 2026-08-28T04:12:59Z

PR_Github #69875 [ run ] triggered by Bot. Commit: 3715cc6 Link to invocation

tensorrt-cicd · 2026-08-28T04:51:31Z

PR_Github #69875 [ run ] completed with state FAILURE. Commit: 3715cc6
/LLM/main/L0_MergeRequest_PR pipeline #57166 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

Signed-off-by: William Zhang <133824995+2ez4bz@users.noreply.github.com>

2ez4bz · 2026-08-28T06:58:02Z

/bot run

2ez4bz · 2026-08-28T06:58:11Z

/bot run --add-multi-gpu-test

tensorrt-cicd · 2026-08-28T07:04:10Z

PR_Github #69916 [ run ] triggered by Bot. Commit: 983462a Link to invocation

tensorrt-cicd · 2026-08-28T07:05:38Z

PR_Github #69918 [ run ] triggered by Bot. Commit: 983462a Link to invocation

tensorrt-cicd · 2026-08-28T07:05:43Z

PR_Github #69916 [ run ] completed with state ABORTED. Commit: 983462a

Link to invocation

tensorrt-cicd · 2026-08-28T11:37:18Z

PR_Github #69918 [ run ] completed with state FAILURE. Commit: 983462a
/LLM/main/L0_MergeRequest_PR pipeline #57204 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

2ez4bz · 2026-08-28T15:44:17Z

/bot run --add-multi-gpu-test

tensorrt-cicd · 2026-08-28T15:51:08Z

PR_Github #69994 [ run ] triggered by Bot. Commit: 983462a Link to invocation

tensorrt-cicd · 2026-08-28T17:27:23Z

PR_Github #69994 [ run ] completed with state SUCCESS. Commit: 983462a
/LLM/main/L0_MergeRequest_PR pipeline #57277 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

2ez4bz · 2026-08-28T18:00:30Z

/bot run --add-multi-gpu-test

tensorrt-cicd · 2026-08-28T18:05:59Z

PR_Github #70037 [ run ] triggered by Bot. Commit: 983462a Link to invocation

tensorrt-cicd · 2026-08-28T19:14:46Z

PR_Github #70037 [ run ] completed with state SUCCESS. Commit: 983462a
/LLM/main/L0_MergeRequest_PR pipeline #57312 completed with status: 'SUCCESS'

CI Report

Link to invocation

trtllm-agent requested a review from a team as a code owner August 24, 2026 11:46

trtllm-agent assigned 2ez4bz Aug 24, 2026

trtllm-agent requested review from leslie-fang25 and xxi-nv August 24, 2026 11:46

github-actions Bot assigned trtllm-agent Aug 24, 2026

xinhe-nv approved these changes Aug 26, 2026

View reviewed changes

sunnyqgg approved these changes Aug 26, 2026

View reviewed changes

2ez4bz enabled auto-merge (squash) August 26, 2026 06:02

mikeiovine added the ci: full pre-merge approved label Aug 27, 2026

Merge branch 'main' into repair-bot-bug6644644

3715cc6

Signed-off-by: William Zhang <133824995+2ez4bz@users.noreply.github.com>

Fix autoflake issue

983462a

Signed-off-by: William Zhang <133824995+2ez4bz@users.noreply.github.com>

2ez4bz merged commit 61aa99a into NVIDIA:main Aug 28, 2026
10 checks passed

Uh oh!

Conversation

trtllm-agent commented Aug 24, 2026 • edited by coderabbitai Bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Test plan

Links

Dev Engineer Review

QA Engineer Review

Uh oh!

coderabbitai Bot commented Aug 24, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Walkthrough

Changes

❌ Failed checks (1 warning)

Uh oh!

xxi-nv commented Aug 25, 2026

Uh oh!

sunnyqgg left a comment

Choose a reason for hiding this comment

Uh oh!

2ez4bz commented Aug 26, 2026

Uh oh!

tensorrt-cicd commented Aug 26, 2026

Uh oh!

tensorrt-cicd commented Aug 26, 2026

Uh oh!

2ez4bz commented Aug 26, 2026

Uh oh!

tensorrt-cicd commented Aug 26, 2026

Uh oh!

tensorrt-cicd commented Aug 26, 2026

Uh oh!

2ez4bz commented Aug 26, 2026

Uh oh!

tensorrt-cicd commented Aug 26, 2026

Uh oh!

tensorrt-cicd commented Aug 26, 2026

Uh oh!

coderabbitai Bot commented Aug 28, 2026

Uh oh!

2ez4bz commented Aug 28, 2026

Uh oh!

github-actions Bot commented Aug 28, 2026

GitHub Bot Help

kill

skip

reuse-pipeline

Uh oh!

2ez4bz commented Aug 28, 2026

Uh oh!

tensorrt-cicd commented Aug 28, 2026

Uh oh!

tensorrt-cicd commented Aug 28, 2026

Uh oh!

2ez4bz commented Aug 28, 2026

Uh oh!

2ez4bz commented Aug 28, 2026

Uh oh!

tensorrt-cicd commented Aug 28, 2026

Uh oh!

tensorrt-cicd commented Aug 28, 2026

Uh oh!

tensorrt-cicd commented Aug 28, 2026

Uh oh!

tensorrt-cicd commented Aug 28, 2026

Uh oh!

2ez4bz commented Aug 28, 2026

Uh oh!

tensorrt-cicd commented Aug 28, 2026

Uh oh!

tensorrt-cicd commented Aug 28, 2026

Uh oh!

2ez4bz commented Aug 28, 2026

Uh oh!

tensorrt-cicd commented Aug 28, 2026

Uh oh!

tensorrt-cicd commented Aug 28, 2026

Uh oh!

Uh oh!

trtllm-agent commented Aug 24, 2026 •

edited by coderabbitai Bot

Loading

coderabbitai Bot commented Aug 24, 2026 •

edited

Loading