From owner-freebsd-current@FreeBSD.ORG Wed Jul 6 15:28:50 2011 Return-Path: Delivered-To: freebsd-current@FreeBSD.org Received: by hub.freebsd.org (Postfix, from userid 1233) id 8BB741065670; Wed, 6 Jul 2011 15:28:50 +0000 (UTC) Date: Wed, 6 Jul 2011 15:28:50 +0000 From: Alexander Best To: Andriy Gapon Message-ID: <20110706152850.GA4139@freebsd.org> References: <4E1421D9.7080808@zedat.fu-berlin.de> <20110706133337.GA89910@freebsd.org> <4E14783C.9030909@FreeBSD.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <4E14783C.9030909@FreeBSD.org> Cc: FreeBSD Current , arrowdodger <6yearold@gmail.com> Subject: Re: Heavy I/O blocks FreeBSD box for several seconds X-BeenThere: freebsd-current@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Discussions about the use of FreeBSD-current List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 06 Jul 2011 15:28:50 -0000 On Wed Jul 6 11, Andriy Gapon wrote: > on 06/07/2011 16:33 Alexander Best said the following: > > you might also want to try enabling options IPI_PREEMPTION. no idea, if this > > improves your situation, though. > > Just in case, this option has effect for 4BSD scheduler only. thanks. i did not know that. maybe we could add a small note to NOTES or even mention in sched_ule(4) and sched_4bsd(4), which kernel options affect the according scheduler. sched_ule(4) e.g. doesn't mention kern.sched.preemption, so one can assume that defining PREEMPTION in the kernel or not doesn't make a difference. however it *does* make a huge difference. i believe sched_ule(4) in general needs a lot more details about the various sysctl vars available and their semantics. cheers. alex > > -- > Andriy Gapon