[Feature]: Support hipIpcEventHandler
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 202
- Forks
- 47
- Avg merge
- 6d 11h
- Merged PRs (30d)
- 4
Description
Suggestion Description
According to examples, it uses torch.distributed.barrier (previously torch.cuda.synchronize and mpi's world_barrier) to handle synchronization.
Very noob question. Can we add hipIpcEventHandler for fine-grained control? I'm not sure whether torch APIs use hipIpcEventHandler under the hood.
Operating System
No response
GPU
No response
ROCm Component
No response
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
The issue names torch.distributed.barrier, torch.cuda.synchronize, and MPI's world_barrier, but no repository files, tests, or entry points. Start by locating the synchronization examples and existing HIP integration, then determine whether hipIpcEventHandler can be exposed safely; done requires an agreed API and validation of fine-grained synchronization behavior.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- pytorch
- Domain
- distributed-systems
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 20/100