From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S268522AbTGISbb (ORCPT ); Wed, 9 Jul 2003 14:31:31 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S266084AbTGISba (ORCPT ); Wed, 9 Jul 2003 14:31:30 -0400 Received: from fw-az.mvista.com ([65.200.49.158]:26102 "EHLO zipcode.az.mvista.com") by vger.kernel.org with ESMTP id S268504AbTGISb1 (ORCPT ); Wed, 9 Jul 2003 14:31:27 -0400 Message-ID: <3F0C6288.4030502@mvista.com> Date: Wed, 09 Jul 2003 11:44:24 -0700 From: Steven Dake User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.4) Gecko/20030624 X-Accept-Language: en-us, en MIME-Version: 1.0 To: linux-kernel@vger.kernel.org Subject: Re: interesting problem with priorities and /sbin/hotplug References: <42050DF556283A4D977B111EB70632080DD9B5@orsmsx407.jf.intel.com> <3F0C5E3D.1090402@mvista.com> <20030709183435.GA17969@kroah.com> In-Reply-To: <20030709183435.GA17969@kroah.com> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Greg KH wrote: >On Wed, Jul 09, 2003 at 11:26:05AM -0700, Steven Dake wrote: > > >>Greg, >> >>We found a small problem with the scheduler with our Linux 2.4 O1 >>backport in that a certain kernel process, keventd, would create a >>priority inversion. If a process was run at realtime priority level 99, >>keventd would never run. The keventd process is required for the >>console, creating the appearance that the system had stopped responding. >> >>This problem started me thinking about priority inversions with >>/sbin/hotplug. I wanted to get your thoughts about priority inversion >>impact on /sbin/hotplug. Should /sbin/hotplug run at priority level 99 >>realtime? >> >> > >I don't know, why would that help anything out? How would there be a >priority inversion in the existing code? > > I'm not sure if there is a problem, but one example to consider is the following: If a process is running with realtime priority 99 totally consuming the system, and an /sbin/hotplug process is started by the kernel because of a device insert or remove, /sbin/hotplug will never be scheduled if it runs with standard priority. If that service depends on those hotplug events, it will never see them (even though they have occured). Since I'm no scheduler expert, I don't know if this is actually how the system would behave.... I do know that it would be desireable to have /sbin/hotplug always run even if other processes in the system are running in realtime priority....