microsoft / microsoft/OpenCLOn12
Improve specialization handling of the parallel compiles for the same kernel
Open
Nobody has claimed this yet.
enhancement
- Dominant language
- C++
- Stars
- 130
- Forks
- 19
- Avg merge
- 11d 23h
- Merged PRs (30d)
- 2
Description
Right now we can end up with multiple threads trying to specialize the same expensive kernel, when it'd be more efficient to just let one do it while the others wait. Need to either introduce per-kernel locks or other async/await mechanism.
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
No files, tests, or entry points are named. Trace where kernel specialization is requested and where concurrent compilation is coordinated, then determine how duplicate work can be serialized per kernel without blocking unrelated kernels; done means parallel requests no longer compile the same kernel redundantly.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- cpp
- Domain
- backend, performance
- Issue type
- Refactor
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 20/100