cc: tech-kern since it also touches kerneland
On Mon, Apr 15, 2013 at 12:17 AM, zhixiang yu <[email protected]> wrote: > Hi List, > > I'm a graduate student from China. I'm interested in the NetBSD's "Light > weight precision user level time reading" project. I'd like to take it as > my GSoC 2013 project. > > But before sending the detailed proposal, I'd like listen to your > advices/comments since my solution is a bit different: Allocate one special > page on which the kernel and user-space process can share data such as > bintime,etc. The kernel will update the shared data at some event point, > the kernel will also need to tell which timecounter it's using. The > userspace just reads out the bintime and read the timecounter to adjust the > final result. Sure the concurrency between kernel and userspace should be > resolved properly. > > What do you think about this solution? Is it acceptable? > > Thanks in advance >
