[https://nvbugs/5527655][feat] Cherry-pick https://github.com/NVIDIA/TensorRT-LLM/pull/8805 #9112
Conversation
📝 WalkthroughWalkthroughCPU affinity configuration for workers is refactored from implementation in Changes
Sequence DiagramsequenceDiagram
participant Worker as BaseWorker.init()
participant Rank as _get_comm_ranks_device_id()
participant Affinity as _configure_affinity()
participant Util as get_numa_aware_cpu_affinity()
participant NVML as NVML (GPU Topology)
participant OS as psutil (Current Affinity)
Worker->>Rank: determine device_id
Rank->>Affinity: _configure_affinity(device_id)
Affinity->>OS: check current affinity
rect rgb(220, 240, 255)
Note over OS: Log warning if constrained
end
Affinity->>Affinity: check TLLM_NUMA_* env vars
alt auto-configure enabled or explicit request
Affinity->>Util: get_numa_aware_cpu_affinity(device_id)
Util->>NVML: query GPU-CPU topology
rect rgb(240, 220, 255)
Note over NVML: Return CPU mask for device
end
Util-->>Affinity: return CPU ID list
rect rgb(220, 255, 220)
Note over Affinity: Apply affinity to process
end
else constrained and no override
Note over Affinity: Leave affinity unchanged
end
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~25 minutes
Pre-merge checks and finishing touches❌ Failed checks (2 warnings)
✅ Passed checks (1 passed)
✨ Finishing touches
🧪 Generate unit tests (beta)
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. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 1
📜 Review details
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (3)
tensorrt_llm/executor/base_worker.py(4 hunks)tensorrt_llm/executor/worker.py(1 hunks)tensorrt_llm/llmapi/utils.py(2 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:
tensorrt_llm/executor/worker.pytensorrt_llm/llmapi/utils.pytensorrt_llm/executor/base_worker.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:
tensorrt_llm/executor/worker.pytensorrt_llm/llmapi/utils.pytensorrt_llm/executor/base_worker.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:
tensorrt_llm/executor/worker.pytensorrt_llm/llmapi/utils.pytensorrt_llm/executor/base_worker.py
🧠 Learnings (1)
📚 Learning: 2025-07-17T09:01:27.402Z
Learnt from: amitz-nv
Repo: NVIDIA/TensorRT-LLM PR: 5616
File: tensorrt_llm/executor/worker.py:375-384
Timestamp: 2025-07-17T09:01:27.402Z
Learning: In tensorrt_llm/executor/worker.py, the LoRA adapter cache optimization logic that checks `is_adapter_in_cpu_cache()` and conditionally passes None for weights/config has a known race condition issue that cannot be solved with simple error handling or verification checks. This is a known limitation that requires a more comprehensive solution.
Applied to files:
tensorrt_llm/executor/worker.pytensorrt_llm/executor/base_worker.py
🧬 Code graph analysis (2)
tensorrt_llm/executor/worker.py (1)
tensorrt_llm/llmapi/utils.py (2)
print_colored_debug(66-70)print_traceback_on_error(33-44)
tensorrt_llm/executor/base_worker.py (2)
tensorrt_llm/llmapi/utils.py (5)
_SyncQueue(389-456)get_numa_aware_cpu_affinity(459-509)print_colored_debug(66-70)get(356-386)get(439-456)tensorrt_llm/logger.py (1)
warning(132-133)
🪛 Ruff (0.14.4)
tensorrt_llm/llmapi/utils.py
506-506: Do not use bare except
(E722)
506-507: try-except-pass detected, consider logging the exception
(S110)
⏰ 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
| try: | ||
| # initialize NVML | ||
| import pynvml | ||
| pynvml.nvmlInit() | ||
|
|
||
| # Get the number of bits per ulong | ||
| c_ulong_bits = ctypes.sizeof(ctypes.c_ulong) * 8 | ||
|
|
||
| # Determine how large our cpu set array from NVML needs to be | ||
| cpu_set_size = math.ceil(cpu_count / c_ulong_bits) | ||
|
|
||
| # Get the optimal CPU affinity for this device according to the NUMA | ||
| # topology | ||
| handle = pynvml.nvmlDeviceGetHandleByIndex(device_id) | ||
| affinity_masks = pynvml.nvmlDeviceGetCpuAffinity(handle, cpu_set_size) | ||
|
|
||
| # Convert CPU masks to python list | ||
| cpu_affinity = [] | ||
| for cpu_id in range(cpu_count): | ||
| mask_array_index = cpu_id // c_ulong_bits | ||
| mask_bit_index = cpu_id % c_ulong_bits | ||
| if affinity_masks[mask_array_index] & (1 << mask_bit_index): | ||
| cpu_affinity.append(cpu_id) | ||
| finally: | ||
| try: | ||
| pynvml.nvmlShutdown() | ||
| except: | ||
| pass # Ignore shutdown errors |
There was a problem hiding this comment.
Guard NUMA affinity when NVML isn’t available
If pynvml is missing or NVML init fails, the finally block still tries to call pynvml.nvmlShutdown(), but pynvml is undefined in that scenario. The resulting UnboundLocalError bubbles up into BaseWorker._configure_affinity and prevents workers from starting on any host without NVML installed—exactly the environment where we need to fall back to the unconstrained affinity declared at the top of this helper. Please short-circuit when the import or NVML calls fail so we return the default affinity safely.
Apply this diff:
- try:
- # initialize NVML
- import pynvml
- pynvml.nvmlInit()
+ try:
+ import pynvml
+ except ModuleNotFoundError:
+ return cpu_affinity
+
+ try:
+ pynvml.nvmlInit()
@@
- finally:
- try:
- pynvml.nvmlShutdown()
- except:
- pass # Ignore shutdown errors
+ except pynvml.NVMLError:
+ return cpu_affinity
+ finally:
+ try:
+ pynvml.nvmlShutdown()
+ except pynvml.NVMLError:
+ pass # Ignore shutdown errors🧰 Tools
🪛 Ruff (0.14.4)
506-506: Do not use bare except
(E722)
506-507: try-except-pass detected, consider logging the exception
(S110)
Signed-off-by: Dan Hansen <[email protected]> Co-authored-by: Dan Hansen <[email protected]> Signed-off-by: Dan Hansen <[email protected]>
dhansen-nvidia
force-pushed
the
numa_aware_affinity_1.1
branch
from
6ed9572 to
99bd4b9
Compare
dhansen-nvidia
changed the title
|
/bot run --disable-fail-fast |
|
PR_Github #24357 [ run ] triggered by Bot. Commit: |
|
PR_Github #24357 [ run ] completed with state |
|
/bot run --disable-fail-fast |
1 similar comment
|
/bot run --disable-fail-fast |
|
PR_Github #24506 [ run ] triggered by Bot. Commit: |
|
PR_Github #24507 [ run ] triggered by Bot. Commit: |
|
PR_Github #24506 [ run ] completed with state |
|
PR_Github #24507 [ run ] completed with state |
|
/bot run --disable-fail-fast |
|
PR_Github #24605 [ run ] triggered by Bot. Commit: |
|
PR_Github #24605 [ run ] completed with state |
|
/bot run --disable-fail-fast |
|
PR_Github #25228 [ run ] triggered by Bot. Commit: |
|
PR_Github #25228 [ run ] completed with state |
|
/bot run --disable-fail-fast |
|
/bot kill |
|
PR_Github #25367 [ run ] triggered by Bot. Commit: |
|
PR_Github #25368 [ kill ] triggered by Bot. Commit: |
|
PR_Github #25367 [ run ] completed with state |
|
PR_Github #25368 [ kill ] completed with state |
|
/bot run --disable-fail-fast |
|
PR_Github #25371 [ run ] triggered by Bot. Commit: |
|
PR_Github #25371 [ run ] completed with state |
|
/bot run --disable-fail-fast |
|
PR_Github #25768 [ run ] triggered by Bot. Commit: |
|
PR_Github #25768 [ run ] completed with state |
|
After calibration with Xiaowei/Tao, this PR can affect default CPU affinity logics which can bring perf regression, so should not be merged to 1.1 release. |
4 participants
… (#8805)
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
Update tava architecture diagram if there is a significant design change in PR.
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.
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)]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.