From owner-freebsd-current Sun Jun 11 3:18:14 2000 Delivered-To: freebsd-current@freebsd.org Received: from pub.dn.ua (pub.dn.ua [193.124.70.70]) by hub.freebsd.org (Postfix) with ESMTP id 12A2837BF58 for ; Sun, 11 Jun 2000 03:18:09 -0700 (PDT) (envelope-from arcade@pub.dn.ua) Received: (from arcade@localhost) by pub.dn.ua (Aist/Vlad) id NAA05774; Sun, 11 Jun 2000 13:16:59 +0300 (EEST) Message-ID: <20000611131659.A3540@dn.ua> Date: Sun, 11 Jun 2000 13:16:59 +0300 From: Vladimir Kostyrko To: "Jacob A. Hart" Cc: freebsd-current@freebsd.org Subject: Re: Scheduler changes? References: <20000528135331.A241@carcass.au.hartware.com> <20000611200322.A236@carcass.au.hartware.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.93.2i In-Reply-To: <20000611200322.A236@carcass.au.hartware.com>; from Jacob A. Hart on Sun, Jun 11, 2000 at 08:03:22PM +1000 Sender: owner-freebsd-current@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG On Sun, Jun 11, 2000 at 08:03:22PM +1000, Jacob A. Hart wrote: > Using idprio as Volodymyr suggested seems to be a viable workaround. You > mentioned in another message that idprio could potentially deadlock my > machine, though. Under what conditions could this happen (and how likely > is it to occur)? That things were beautifully described in [id|rt]prio man page. idprio makes process idle, so it gets only time slices left _after_ ditributing them among normal processes. rtprio makes process priority realtime so each time slice is distributed to this process and only leftover went to system. If you make rtprio dnetc, the system near hangs and understands only three buttons but one minute after pressing them.. ;) So using idprio would not hang the machine. -- [WBR], Arcade Zardos. [AIST] [SAT Astronomy//Think to survive!] [mp3.aist.net] To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message