Date: Wed, 27 Oct 2004 16:31:16 -0400 From: John Baldwin <jhb@FreeBSD.org> To: Julian Elischer <julian@elischer.org> Cc: Perforce Change Reviews <perforce@FreeBSD.org> Subject: Re: PERFORCE change 63836 for review Message-ID: <200410271631.16160.jhb@FreeBSD.org> In-Reply-To: <417FFC9D.1080500@elischer.org> References: <200410271747.i9RHlpjg005855@repoman.freebsd.org> <417FFC9D.1080500@elischer.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Wednesday 27 October 2004 03:53 pm, Julian Elischer wrote: > John Baldwin wrote: > > on needresched? If so, it goes into sched_userret(). > >+- Need lots of thought on realtime and idle time priorities and handling > >+ tsleep() priority "boosts" for such processes. Might need to make > >+ the priority boost be something passsed to sched_sleep() rather than > >+ a sched_prio() call. > > I'd like to see the boost kept separate and added during the calcualtions > that way various algorythms could be used on the boost.. > for example I'd like to see the boost exponentially reduced each > clocktick encounterred in userspace > or maybe the boost could include a suggested lifetime.. > e.g. "boost by 10% for 20mSec" This would let the scheduler do that. Right now we just change the priority via sched_prio(). This change would instead tell the scheduler at sched_sleep() what the boost is and it is free to store it and use it however it wants. -- John Baldwin <jhb@FreeBSD.org> <>< http://www.FreeBSD.org/~jhb/ "Power Users Use the Power to Serve" = http://www.FreeBSD.org
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200410271631.16160.jhb>