[https://nvbugs/5477730][fix] Fix the alltoall case when tp_size larger than ep_size #7331
Conversation
…er than ep_size Signed-off-by: Fred Wei <20514172+WeiHaocheng@users.noreply.github.com>
📝 WalkthroughWalkthroughShifts MOE-related logic from tensor-parallel (tp_) fields to new MOE endpoint (moe_ep_) and MOE topology (moe_tp_) fields. Updates communicator color/rank derivation, workspace sizing per rank, and workspace initialization to use moe_ep_ and moe_tp_* in tensorrt_llm/_mnnvl_utils.py. Adds/renames Mapping attributes accordingly. Changes
Sequence Diagram(s)sequenceDiagram
autonumber
participant Caller
participant Mapping
participant Communicator
participant Workspace
Note over Mapping,Workspace: New MOE-oriented fields (moe_ep_*, moe_tp_*)
Caller->>Mapping: Read moe_ep_size, moe_ep_rank
Caller->>Mapping: Read moe_tp_size, moe_tp_rank
Mapping-->>Caller: Values
Caller->>Communicator: compose_color(moe_tp_size, moe_tp_rank)
Communicator-->>Caller: color, rank_in_color
Caller->>Workspace: calc_size_per_rank(moe_ep_size)
Workspace-->>Caller: workspace_size_per_rank
Caller->>Workspace: init(moe_ep_rank, moe_ep_size)
Workspace-->>Caller: init_ok
Note right of Workspace: Sizing/initialization aligned to MOE endpoints
Estimated code review effort🎯 4 (Complex) | ⏱️ ~60 minutes Suggested labels
Suggested reviewers
Tip 🔌 Remote MCP (Model Context Protocol) integration is now available!Pro plan users can now connect to remote MCP servers from the Integrations page. Connect with popular remote MCPs such as Notion and Linear to add more context to your reviews and chats. ✨ Finishing Touches
🧪 Generate unit tests
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. 🪧 TipsChatThere are 3 ways to chat with CodeRabbit:
SupportNeed help? Create a ticket on our support page for assistance with any issues or questions. CodeRabbit Commands (Invoked using PR/Issue comments)Type Other keywords and placeholders
Status, Documentation and Community
|
|
/bot run |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 0
🧹 Nitpick comments (1)
tensorrt_llm/_mnnvl_utils.py (1)
382-384: LGTM: prepare-workspace size per rank uses moe_ep_sizeConsistent with the main workspace path; this should prevent over-allocation when tp_size > ep_size.
Optionally mirror a post-allocation barrier here (as done for the main workspace) if upstream users can race on first-touch of the prepare buffer.
📜 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.
📒 Files selected for processing (1)
tensorrt_llm/_mnnvl_utils.py(3 hunks)
🧰 Additional context used
🧠 Learnings (1)
📓 Common learnings
Learnt from: djns99
PR: NVIDIA/TensorRT-LLM#6915
File: cpp/tensorrt_llm/kernels/cutlass_kernels/moe_gemm/moe_kernels.cu:4616-4626
Timestamp: 2025-08-19T03:35:20.866Z
Learning: In the MOE profiler TMA workspace preparation (cpp/tensorrt_llm/kernels/cutlass_kernels/moe_gemm/moe_kernels.cu), the overlapping of TMA WS regions for NONE and FINALIZE variants is deliberate design to save memory space, as confirmed by djns99. The comment "reuse the same pointers to save space" reflects this intentional behavior.
🧬 Code graph analysis (1)
tensorrt_llm/_mnnvl_utils.py (2)
tensorrt_llm/_torch/distributed/communicator.py (1)
cp_size(38-39)tensorrt_llm/mapping.py (2)
moe_tp_rank(352-353)moe_ep_rank(360-361)
⏰ 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 (2)
tensorrt_llm/_mnnvl_utils.py (2)
366-373: LGTM: workspace sizing/initialization now correctly keyed to moe_ep_{size,rank}Switching from tp_* to moe_ep_* aligns the workspace with the EP dimension and fixes the tp_size > ep_size case.
Consider adding an E2E test with tp_size=8, moe_ep_size=4 validating that moe_commworkspace_size_per_rank and moe_initialize_workspace don’t depend on tp_size.
107-110: Use mapping.moe_ep_rank as Split key to align communicator ranks with ep_rank
In tensorrt_llm/_mnnvl_utils.py, replacecomm = mpi_comm().Split( (mapping.pp_rank * mapping.cp_size + mapping.cp_rank) * mapping.moe_tp_size + mapping.moe_tp_rank, mapping.tp_rank, )with
color = (mapping.pp_rank * mapping.cp_size + mapping.cp_rank) * mapping.moe_tp_size \ + mapping.moe_tp_rank key = mapping.moe_ep_rank comm = mpi_comm().Split(color, key) assert comm.Get_size() == mapping.moe_ep_size, \ f"Unexpected comm size: {comm.Get_size()} vs moe_ep_size={mapping.moe_ep_size}"Then verify at runtime:
assert MnnvlMemory.get_comm(mapping).Get_rank() == mapping.moe_ep_rank[tag: verify_review_comment]
dongxuy04
changed the title
|
PR_Github #16821 [ run ] triggered by Bot |
dongxuy04
changed the title
dongxuy04
changed the title
|
PR_Github #16821 [ run ] completed with state |
|
/bot run --disable-fail-fast |
|
PR_Github #16848 [ run ] triggered by Bot |
|
PR_Github #16848 [ run ] completed with state |
|
/bot run --disable-fail-fast |
1 similar comment
|
/bot run --disable-fail-fast |
|
PR_Github #17275 [ run ] triggered by Bot |
|
PR_Github #17275 [ run ] completed with state |
|
/bot run --post-merge |
|
PR_Github #17329 [ run ] triggered by Bot |
|
PR_Github #17329 [ run ] completed with state |
|
/bot run --post-merge |
|
PR_Github #17525 [ run ] triggered by Bot |
|
PR_Github #17525 [ run ] completed with state |
|
/bot run --post-merge --disable-fail-fast |
|
PR_Github #17594 [ run ] triggered by Bot |
|
PR_Github #17594 [ run ] completed with state |
…er than ep_size (NVIDIA#7331) Signed-off-by: Fred Wei <20514172+WeiHaocheng@users.noreply.github.com>
3 participants
…er than ep_size
Summary by CodeRabbit
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)
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 thestage-listparameter 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.mdand the
scripts/test_to_stage_mapping.pyhelper.kill
killKill all running builds associated with pull request.
skip
skip --comment COMMENTSkip 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-pipelineReuse 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.