mlcommons / mlcommons/chakra

trace_linker: [WARNING]: No CUDA runtime operator found for correlation ID

Open
#194 3 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

question
Dominant language
Python
Stars
196
Forks
84
PR merge metrics
No merged PRs in 30d

Description

Please provide a detailed description of your question or the information you seek.

I encountered the following warning while using chakra link:

[2025-06-04 04:00:01,109] trace_linker.py:679 [WARNING]: No CUDA runtime operator found for correlation ID 17297502. This is not a common case, and there should be a corresponding CUDA runtime operator for a given GPU kernel operator. It can be a case where CUDA runtime operators are not properly identified and added to the map, kineto_correlation_cuda_runtime_map. Please manually check if the corresponding CUDA runtime operator with the correlation is dropped by mistake. It is likely that it is because of incomplete map, cuda_launch_operations, in is_kernel_launch_op. Please update the map properly to cover all CUDA runtime launch operators.
[2025-06-04 04:00:01,109] trace_linker.py:625 [WARNING]: Missing parent CPU operator for GPU op 'void at::native::(anonymous namespace)::multi_tensor_apply_kernel<at::native::(anonymous namespace)::TensorListScalarListMetadata<float, 3>, at::native::(anonymous namespace)::PointwiseOpScalarListFunctor<float, 3, 3, 0>, std::divides<float> >(at::native::(anonymous namespace)::TensorListScalarListMetadata<float, 3>, at::native::(anonymous namespace)::PointwiseOpScalarListFunctor<float, 3, 3, 0>, std::divides<float>)'. Orphaned GPU operator.

I run distributed training with dp=4 on 4XA6000 machines. Below is the repository where I train repo, I'm not sure if this warning is due to me doing something wrong, I didn't run into this problem while collecting traces while training on megatron.

I also raised this issue in the pytorch community issue
The trace of rank0 will not encounter this problem, but some kernels of other ranks will encounter this problem, I am not sure if this is normal.

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start in trace_linker.py around the warnings at lines 625 and 679, then inspect kineto_correlation_cuda_runtime_map, cuda_launch_operations, and is_kernel_launch_op. Reproduce with the distributed traces described in the issue and compare ranks against the linked PyTorch report. Done means determining whether the warnings indicate dropped or unrecognized operators and documenting or correcting the relevant behavior.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
tooling
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.