Date: Tue, 29 Oct 2002 13:53:42 -0800 From: Terry Lambert <tlambert2@mindspring.com> To: Matthew Dillon <dillon@apollo.backplane.com> Cc: Raymond Kohler <raymond.j.kohler@lmco.com>, current@FreeBSD.ORG Subject: Re: questions about the state of current Message-ID: <3DBF0366.10FCC8E@mindspring.com> References: <2570443.1035916854787.JavaMail.wshttp@emss03g01.ems.lmco.com> <3DBEF55E.A0F9ED1B@mindspring.com> <200210292106.g9TL6aoc010659@apollo.backplane.com> <3DBEFE24.1E9DDB89@mindspring.com> <200210292146.g9TLkvWi010975@apollo.backplane.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Matthew Dillon wrote: > Interrupt threads have 'grown' on me. I like them. > But I come from an embedded world where switching threads > costs no more then a procedure call. The way I figure it, > we will eventually be able to make -current's scheduler > efficient enough such that the overhead of switching to > an interrupt thread becomes a non-issue, and they take care > of the big problem we've always had with interrupts under > SMP... managing interrupts in an SMP environment. Don't get me wrong... 15% is heavy overhead, but I expect that, over time, that performance gap will at least narrow, if not disappear, if hyperthreading becomes domething other than a parketing buzzword. > I am somewhat partial to the interrupt context stealing > idea too, though I'm not sure if the added complexity is > worth it (time may be better spent improving the scheduler). I like context stealing, too. I've liked it ever since I first saw it in Windows 95 back in 1996; it's been common practice in the Windows world for a long time. -- Terry To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3DBF0366.10FCC8E>