page_pool, nvme_driver: decide if dma allocations need to be numa aware
Open
Nobody has claimed this yet.
bug
storage
- Dominant language
- Rust
- Stars
- 1.9k
- Forks
- 238
- Avg merge
- 1d 15h
- Merged PRs (30d)
- 100
Description
Today, allocations for queues are not numa aware. We probably want to make them numa aware via change to the dma apis and allocators, and figure out what performance delta there is. This also means when we use the private page pool, figure out how much of each numa node to carve out, etc.
Contributor guide
No contributing guide indexed for this repository
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
Start with the page_pool and nvme_driver allocation paths, then trace the DMA APIs and allocators used for queue allocations. Determine whether NUMA-aware allocation is needed, measure the performance delta, and define how the private page pool should be divided across NUMA nodes.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- rust
- Domain
- operating-systems, performance
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100