[None] [test] Add MNNVL AlltoAll tests to pre-merge #7465

kaiyux · 2025-09-02T06:05:35Z

Summary by CodeRabbit

Tests
- Added a new MoE all-to-all test variant to multiple GPU test suites.
- Re-enabled an FP4 MoE all-to-all test with targeted runtime skips for unstable methods.
- Aligned test tensor allocations to GPU to improve reliability.

Description

Cherry-picking #7071 from @yuantailing .

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
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.

kaiyux · 2025-09-02T06:06:43Z

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

coderabbitai · 2025-09-02T06:11:39Z

📝 Walkthrough

Walkthrough

Updates test lists to add/switch the MoE alltoall MNNVL variant and adjust FP4 coverage. In unit tests, moves certain weight tensors to CUDA and replaces a decorator skip with a runtime conditional skip for DeepEP/DeepEPLowLatency in FP4 tests.

Changes

Cohort / File(s)	Summary
Integration test lists: MoE alltoall variants `tests/integration/test_lists/test-db/l0_dgx_b200.yml`, `tests/integration/test_lists/test-db/l0_dgx_h100.yml`	l0_dgx_b200: Switch pre_merge 4xGPU PyTorch FP4 entry from DeepEPLowLatency to MNNVL. l0_dgx_h100: Add `test_fused_moe_alltoall[MNNVL]` in two pre_merge blocks (auto_trigger: others, deepseek), alongside existing DeepEP and DeepEPLowLatency.
Unit tests: fused MoE alltoall `tests/unittest/_torch/modules/test_fused_moe.py`	Allocate `w1_weight`, `w2_weight`, `w3_weight` on CUDA. Remove decorator skip for `test_fused_moe_alltoall_fp4`; add inline runtime skip when `alltoall_method_type` is DeepEPLowLatency or DeepEP (nvbugs/5467531).

Sequence Diagram(s)

sequenceDiagram
  autonumber
  actor Dev as PyTest Runner
  participant T as test_fused_moe_alltoall_fp4
  participant C as per_rank_test_fused_moe_alltoall
  Note over T: New: runtime skip for DeepEP/DeepEPLowLatency
  Dev->>T: execute()
  T->>T: determine alltoall_method_type
  alt method in {DeepEP, DeepEPLowLatency}
    T-->>Dev: skip(reason="nvbugs/5467531")
  else other method (e.g., MNNVL)
    T->>C: setup & run test
    Note right of C: Weights allocated on device="cuda"
    C-->>T: results
    T-->>Dev: assert pass/fail
  end

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~10 minutes

Possibly related PRs

[None][fix] Add FP4 all2all unitest and fix a bug for module WideEPMoE #6784 — Also modifies test_fused_moe_alltoall_fp4 and related test-list entries around FP4 all-to-all variants.
[None][feat] DeepEP LL combine FP4 #6822 — Adjusts FP4 DeepEP low-latency paths relevant to fused MoE alltoall behavior and test expectations.
[None][fix] Fix possible hang issue in WideEP and move some tests to pre-merge #7262 — Touches MoE alltoall test variants, including introducing or moving MNNVL entries.

Suggested reviewers

chzblych
QiJune
StanleySun639

✨ Finishing Touches

📝 Generate Docstrings

🧪 Generate unit tests

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

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

🪧 Tips

Chat

There are 3 ways to chat with CodeRabbit:

Review comments: Directly reply to a review comment made by CodeRabbit. Example:
- I pushed a fix in commit <commit_id>, please review it.
- Open a follow-up GitHub issue for this discussion.
Files and specific lines of code (under the "Files changed" tab): Tag @coderabbitai in a new review comment at the desired location with your query.
PR comments: Tag @coderabbitai in a new PR comment to ask questions about the PR branch. For the best results, please provide a very specific query, as very limited context is provided in this mode. Examples:
- @coderabbitai gather interesting stats about this repository and render them as a table. Additionally, render a pie chart showing the language distribution in the codebase.
- @coderabbitai read the files in the src/scheduler package and generate a class diagram using mermaid and a README in the markdown format.

Support

Need help? Create a ticket on our support page for assistance with any issues or questions.

CodeRabbit Commands (Invoked using PR/Issue comments)

Type @coderabbitai help to get the list of available commands.

Other keywords and placeholders

Add @coderabbitai ignore or @coderabbit ignore anywhere in the PR description to prevent this PR from being reviewed.
Add @coderabbitai summary to generate the high-level summary at a specific location in the PR description.
Add @coderabbitai or @coderabbitai title anywhere in the PR title to generate the title automatically.

Status, Documentation and Community

Visit our Status Page to check the current availability of CodeRabbit.
Visit our Documentation for detailed information on how to use CodeRabbit.
Join our Discord Community to get help, request features, and share feedback.
Follow us on X/Twitter for updates and announcements.

coderabbitai

Actionable comments posted: 0

🧹 Nitpick comments (1)

tests/unittest/_torch/modules/test_fused_moe.py (1)

304-306: Compact the skip condition for readability.

Same behavior, clearer expression.

-    if alltoall_method_type == AlltoallMethodType.DeepEPLowLatency \
-            or alltoall_method_type == AlltoallMethodType.DeepEP:
-        pytest.skip("Skipped due to https://nvbugs/5467531")
+    if alltoall_method_type in {AlltoallMethodType.DeepEPLowLatency,
+                                AlltoallMethodType.DeepEP}:
+        pytest.skip("Skipped due to https://nvbugs/5467531")

📜 Review details

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

💡 Knowledge Base configuration:

MCP integration is disabled by default for public repositories
Jira integration is disabled by default for public repositories
Linear integration is disabled by default for public repositories

You can enable these sources in your CodeRabbit configuration.

📥 Commits

Reviewing files that changed from the base of the PR and between 935c2c1 and 09478d0.

📒 Files selected for processing (3)

tests/integration/test_lists/test-db/l0_dgx_b200.yml (1 hunks)
tests/integration/test_lists/test-db/l0_dgx_h100.yml (1 hunks)
tests/unittest/_torch/modules/test_fused_moe.py (2 hunks)

🧰 Additional context used

📓 Path-based instructions (4)

**/*

📄 CodeRabbit inference engine (CODING_GUIDELINES.md)

Filenames compiled into a target must be case-insensitively unique

Files:

tests/unittest/_torch/modules/test_fused_moe.py
tests/integration/test_lists/test-db/l0_dgx_h100.yml
tests/integration/test_lists/test-db/l0_dgx_b200.yml

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

📄 CodeRabbit inference engine (CODING_GUIDELINES.md)

Use spaces, not tabs; indent 4 spaces

Files:

tests/unittest/_torch/modules/test_fused_moe.py

**/*.py

📄 CodeRabbit inference engine (CODING_GUIDELINES.md)

**/*.py: Code must target Python 3.8+
Indent with 4 spaces; do not use tabs (Python)
Maintain module namespace on import: prefer from package.subpackage import foo; use foo.Symbol()
Python filenames use snake_case
Python class names use PascalCase
Python functions and methods use snake_case
Python local variables use snake_case; if starting with a number concept, prefix with k (e.g., k_99th_percentile)
Python global variables use G_ prefix with UPPER_SNAKE_CASE
Python constants use UPPER_SNAKE_CASE
Avoid shadowing variables from outer scopes
Initialize all externally visible class members in init
For public interfaces, prefer docstrings over comments; comments should be for in-function or file-local interfaces
Use Google-style docstrings for classes and functions (Sphinx-parsable)
Document attributes and variables inline with docstrings immediately after assignment
Avoid reflection when a non-reflective approach suffices
Limit except clauses to specific exceptions where possible
When using try/except for duck-typing, keep try body minimal and move logic to else

Files:

tests/unittest/_torch/modules/test_fused_moe.py

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

📄 CodeRabbit inference engine (CODING_GUIDELINES.md)

Prepend NVIDIA copyright header (current year) to all source files

Files:

tests/unittest/_torch/modules/test_fused_moe.py

🧠 Learnings (1)

📚 Learning: 2025-07-28T17:06:08.621Z

Learnt from: moraxu
PR: NVIDIA/TensorRT-LLM#6303
File: tests/integration/test_lists/qa/examples_test_list.txt:494-494
Timestamp: 2025-07-28T17:06:08.621Z
Learning: In TensorRT-LLM testing, it's common to have both CLI flow tests (test_cli_flow.py) and PyTorch API tests (test_llm_api_pytorch.py) for the same model. These serve different purposes: CLI flow tests validate the traditional command-line workflow, while PyTorch API tests validate the newer LLM API backend. Both are legitimate and should coexist.

Applied to files:

tests/integration/test_lists/test-db/l0_dgx_b200.yml

🧬 Code graph analysis (1)

tests/unittest/_torch/modules/test_fused_moe.py (1)

tensorrt_llm/_torch/modules/fused_moe/fused_moe_wide_ep.py (1)

AlltoallMethodType (27-35)

⏰ 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 (3)

tests/integration/test_lists/test-db/l0_dgx_b200.yml (1)

19-19: Switch FP4 A2A variant to MNNVL aligns with test runtime skip logic.

test_fused_moe_alltoall_fp4 now skips DeepEP/DeepEPLowLatency at runtime; moving the pre-merge entry to [MNNVL] guarantees the job actually runs instead of being auto-skipped. LGTM.

tests/integration/test_lists/test-db/l0_dgx_h100.yml (1)

83-83: Adding MNNVL coverage on H100 (non-FP4) — verify CI support.

Good addition for parity across A2A methods. Please confirm the H100 pre-merge runners have MNNVL enabled to avoid unexpected infra failures; if not, consider gating to B200-only or adding a runtime capability check.

tests/unittest/_torch/modules/test_fused_moe.py (1)

215-222: Allocate expert weights directly on CUDA to avoid extra H2D copies.

This reduces overhead and matches the device set via torch.cuda.set_device(rank). Looks good.

tensorrt-cicd · 2025-09-02T06:28:55Z

PR_Github #17311 [ run ] triggered by Bot

tensorrt-cicd · 2025-09-02T15:36:03Z

PR_Github #17311 [ run ] completed with state SUCCESS
/LLM/release-1.1.0rc2/L0_MergeRequest_PR pipeline #15 completed with status: 'FAILURE'

kaiyux · 2025-09-02T16:17:35Z

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

tensorrt-cicd · 2025-09-02T16:22:58Z

PR_Github #17384 [ run ] triggered by Bot

tensorrt-cicd · 2025-09-02T18:11:34Z

PR_Github #17384 [ run ] completed with state SUCCESS
/LLM/release-1.1.0rc2/L0_MergeRequest_PR pipeline #22 completed with status: 'FAILURE'

kaiyux · 2025-09-02T21:09:51Z

/bot run --disable-fail-fast

tensorrt-cicd · 2025-09-02T21:15:12Z

PR_Github #17409 [ run ] triggered by Bot

tensorrt-cicd · 2025-09-03T14:19:25Z

PR_Github #17409 [ run ] completed with state SUCCESS
/LLM/release-1.1.0rc2/L0_MergeRequest_PR pipeline #27 completed with status: 'FAILURE'

Signed-off-by: Kaiyu Xie <26294424+kaiyux@users.noreply.github.com>

Signed-off-by: Zongfei Jing <20381269+zongfeijing@users.noreply.github.com>

kaiyux · 2025-09-04T17:03:08Z

/bot run --disable-fail-fast

tensorrt-cicd · 2025-09-04T17:08:14Z

PR_Github #17699 [ run ] triggered by Bot

tensorrt-cicd · 2025-09-04T22:04:41Z

PR_Github #17699 [ run ] completed with state SUCCESS
/LLM/release-1.1.0rc2/L0_MergeRequest_PR pipeline #69 completed with status: 'FAILURE'

kaiyux · 2025-09-04T22:21:01Z

/bot run --disable-fail-fast

tensorrt-cicd · 2025-09-04T22:26:57Z

PR_Github #17714 [ run ] triggered by Bot

tensorrt-cicd · 2025-09-05T01:40:56Z

PR_Github #17714 [ run ] completed with state SUCCESS
/LLM/release-1.1.0rc2/L0_MergeRequest_PR pipeline #71 completed with status: 'FAILURE'

kaiyux · 2025-09-05T03:28:27Z

/bot run --disable-fail-fast

tensorrt-cicd · 2025-09-05T03:33:39Z

PR_Github #17740 [ run ] triggered by Bot

tensorrt-cicd · 2025-09-05T06:16:49Z

PR_Github #17740 [ run ] completed with state SUCCESS
/LLM/release-1.1.0rc2/L0_MergeRequest_PR pipeline #75 completed with status: 'FAILURE'

kaiyux · 2025-09-05T07:58:54Z

/bot run --disable-fail-fast

tensorrt-cicd · 2025-09-05T08:04:30Z

PR_Github #17763 [ run ] triggered by Bot

tensorrt-cicd · 2025-09-05T13:29:53Z

PR_Github #17763 [ run ] completed with state SUCCESS
/LLM/release-1.1.0rc2/L0_MergeRequest_PR pipeline #77 completed with status: 'SUCCESS'

Signed-off-by: Kaiyu Xie <26294424+kaiyux@users.noreply.github.com> Signed-off-by: Zongfei Jing <20381269+zongfeijing@users.noreply.github.com> Co-authored-by: Zongfei Jing <20381269+zongfeijing@users.noreply.github.com>

Signed-off-by: Kaiyu Xie <26294424+kaiyux@users.noreply.github.com> Signed-off-by: Zongfei Jing <20381269+zongfeijing@users.noreply.github.com> Co-authored-by: Zongfei Jing <20381269+zongfeijing@users.noreply.github.com> Signed-off-by: Kaiyu Xie <26294424+kaiyux@users.noreply.github.com>

kaiyux requested review from dongxuy04 and yuantailing September 2, 2025 06:05

yuantailing approved these changes Sep 2, 2025

View reviewed changes

coderabbitai bot reviewed Sep 2, 2025

View reviewed changes

zongfeijing requested review from a team as code owners September 4, 2025 13:00

zongfeijing requested review from Wanli-Jiang and hlu1 September 4, 2025 13:00

zongfeijing force-pushed the user/kaiyu/mnnvl_a2a_test branch from 28b08a6 to 256f642 Compare September 4, 2025 13:36

kaiyux and others added 3 commits September 4, 2025 10:02

Add MNNVL AlltoAll tests to pre-merge

e0cef6b

Signed-off-by: Kaiyu Xie <26294424+kaiyux@users.noreply.github.com>

Update

9bfea2c

Signed-off-by: Kaiyu Xie <26294424+kaiyux@users.noreply.github.com>

Fix moe UT

9b53779

Signed-off-by: Zongfei Jing <20381269+zongfeijing@users.noreply.github.com>

kaiyux force-pushed the user/kaiyu/mnnvl_a2a_test branch from 256f642 to 9b53779 Compare September 4, 2025 17:02

kaiyux mentioned this pull request Sep 4, 2025

[None] [test] Add MNNVL AlltoAll tests to pre-merge #7466

Merged

1 task

kaiyux enabled auto-merge (squash) September 4, 2025 22:21

hlu1 approved these changes Sep 5, 2025

View reviewed changes

kaiyux merged commit 1455074 into NVIDIA:release/1.1.0rc2 Sep 5, 2025
5 checks passed

kaiyux deleted the user/kaiyu/mnnvl_a2a_test branch September 5, 2025 17:41

[None] [test] Add MNNVL AlltoAll tests to pre-merge #7465

[None] [test] Add MNNVL AlltoAll tests to pre-merge #7465

Conversation

kaiyux commented Sep 2, 2025 • edited by coderabbitai bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary by CodeRabbit

Description

Test Coverage

PR Checklist

GitHub Bot Help

kill

skip

reuse-pipeline

Uh oh!

kaiyux commented Sep 2, 2025

Uh oh!

coderabbitai bot commented Sep 2, 2025

Walkthrough

Changes

Sequence Diagram(s)

Estimated code review effort

Possibly related PRs

Suggested reviewers

Chat

Support

CodeRabbit Commands (Invoked using PR/Issue comments)

Other keywords and placeholders

Status, Documentation and Community

Uh oh!

coderabbitai bot left a comment

Choose a reason for hiding this comment

Uh oh!

tensorrt-cicd commented Sep 2, 2025

Uh oh!

tensorrt-cicd commented Sep 2, 2025

Uh oh!

kaiyux commented Sep 2, 2025

Uh oh!

tensorrt-cicd commented Sep 2, 2025

Uh oh!

tensorrt-cicd commented Sep 2, 2025

Uh oh!

kaiyux commented Sep 2, 2025

Uh oh!

tensorrt-cicd commented Sep 2, 2025

Uh oh!

tensorrt-cicd commented Sep 3, 2025

Uh oh!

kaiyux commented Sep 4, 2025

Uh oh!

tensorrt-cicd commented Sep 4, 2025

Uh oh!

tensorrt-cicd commented Sep 4, 2025

Uh oh!

kaiyux commented Sep 4, 2025

Uh oh!

tensorrt-cicd commented Sep 4, 2025

Uh oh!

tensorrt-cicd commented Sep 5, 2025

Uh oh!

kaiyux commented Sep 5, 2025

Uh oh!

tensorrt-cicd commented Sep 5, 2025

Uh oh!

tensorrt-cicd commented Sep 5, 2025

Uh oh!

kaiyux commented Sep 5, 2025

Uh oh!

tensorrt-cicd commented Sep 5, 2025

Uh oh!

tensorrt-cicd commented Sep 5, 2025

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

kaiyux commented Sep 2, 2025 •

edited by coderabbitai bot

Loading