From owner-freebsd-current@FreeBSD.ORG Fri Aug 3 10:01:41 2007 Return-Path: Delivered-To: current@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 5DCD816A419 for ; Fri, 3 Aug 2007 10:01:41 +0000 (UTC) (envelope-from jroberson@chesapeake.net) Received: from webaccess-cl.virtdom.com (webaccess-cl.virtdom.com [216.240.101.25]) by mx1.freebsd.org (Postfix) with ESMTP id 21F5813C46C for ; Fri, 3 Aug 2007 10:01:41 +0000 (UTC) (envelope-from jroberson@chesapeake.net) Received: from [192.168.1.101] (c-71-231-138-78.hsd1.or.comcast.net [71.231.138.78]) (authenticated bits=0) by webaccess-cl.virtdom.com (8.13.6/8.13.6) with ESMTP id l73A1YMo091567 (version=TLSv1/SSLv3 cipher=DHE-DSS-AES256-SHA bits=256 verify=NO); Fri, 3 Aug 2007 06:01:35 -0400 (EDT) (envelope-from jroberson@chesapeake.net) Date: Fri, 3 Aug 2007 03:04:08 -0700 (PDT) From: Jeff Roberson X-X-Sender: jroberson@10.0.0.1 To: current@freebsd.org In-Reply-To: <20070730181414.G561@10.0.0.1> Message-ID: <20070803030226.Y561@10.0.0.1> References: <20070730181414.G561@10.0.0.1> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed Cc: Michael Plass Subject: Re: SCHED_ULE wacky load averages on -current, fixed. 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: Fri, 03 Aug 2007 10:01:41 -0000 On Mon, 30 Jul 2007, Jeff Roberson wrote: > On Mon, 30 Jul 2007, Michael Plass wrote: >> >> Jeff, were you ever able to reproduce the wacky load averages referenced >> above? I updated from -current and built with SCHED_ULE, and I'm seeing >> this on an idle dual-P4: > > Yes, I have a patch to fix it in the works. Should have some time to get > back to this soon. Please test http://people.freebsd.org/~jeff/ulehtt.diff. This should fix HTT machines. It also has a small performance improvement in it for interrupt heavy workloads that may cause a lot of preemption. Thanks, Jeff