Subj : Re: A question about atomic_ptr To : comp.programming.threads From : Joe Seigh Date : Fri Apr 15 2005 08:37 am On Fri, 15 Apr 2005 12:30:01 +0200, Alexander Terekhov wrote: > > Joe Seigh wrote: > [...] >> Original post of algorihm was here >> http://groups.google.com/groups?hl=en&lr=&ie=UTF-8&selm=1998Jun23.155614%40bose.com >> if you've forgotten. > > I'm here since 2001. Not hard to "forget" something that occurred in > c.p.t.-1998. Anyway, intrusive approach is no-no here (from general > safety POV) and I'd keep hold on control blocks (reuse it internally > not reclaiming memory) for the sake of safe sex, so to speak. > That's all right. I forgot it also. :) http://groups.google.com/groups?selm=3EE0769C.7989FC1F@xemaps.com As far as memory safety, you can't use it to point to things like shared segments but it's as safe as some other lock-free algorithms. If you could add instructions to the architecture, you could add a "set reserve" instruction along with a load conditional and fix the side effect thing with stwarx. But if you could add instructions I'd be adding a lot more useful and cooler instructions. -- Joe Seigh .