Consider restoring don't-preempt-SYS_write-until-k-ms optimization
Nobody has claimed this yet.
- Dominant language
- C++
- Stars
- 10.7k
- Forks
- 662
- Avg merge
- 2d 3h
- Merged PRs (30d)
- 2
Description
As part of #7, I'm ripping out the code that previously disabled preemption for write but up to a "grace period". write will soon be a buffered syscall and we'll usually not even get a trace event for it; this is optimal. And even if it is preempted, we'll be able to near-optimally resume it.
So the previous optimization would only apply to the (hopefully!) rare case when the syscall buffer is full.
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
Start by reviewing the changes associated with #7 and locating the code for the previous don't-preempt-SYS_write-until-k-ms optimization. Determine how buffered write syscalls and full syscall buffers affect preemption, then establish whether restoring the optimization is worthwhile for that rare full-buffer case.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- cpp, linux
- 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