From owner-freebsd-stable@FreeBSD.ORG Fri Sep 20 14:36:20 2013 Return-Path: Delivered-To: freebsd-stable@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:1900:2254:206a::19:1]) (using TLSv1 with cipher ADH-AES256-SHA (256/256 bits)) (No client certificate requested) by hub.freebsd.org (Postfix) with ESMTP id A3BF2633 for ; Fri, 20 Sep 2013 14:36:20 +0000 (UTC) (envelope-from gpalmer@freebsd.org) Received: from mail.in-addr.com (unknown [IPv6:2001:470:8:162::1]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mx1.freebsd.org (Postfix) with ESMTPS id 758A02B68 for ; Fri, 20 Sep 2013 14:36:20 +0000 (UTC) Received: from gjp by mail.in-addr.com with local (Exim 4.80.1 (FreeBSD)) (envelope-from ) id 1VN1od-000ELL-SQ; Fri, 20 Sep 2013 10:36:15 -0400 Date: Fri, 20 Sep 2013 10:36:15 -0400 From: Gary Palmer To: Thomas Laus Subject: Re: FreeBSD 9-Stable + Atom D510 Freeze Message-ID: <20130920143615.GA40029@in-addr.com> References: <523AFF46.29244.4657A1@lausts.acm.org> <523B91F2.8070004@heuristicsystems.com.au> <523C49A9.5719.289C1E@lausts.acm.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <523C49A9.5719.289C1E@lausts.acm.org> X-SA-Exim-Connect-IP: X-SA-Exim-Mail-From: gpalmer@freebsd.org X-SA-Exim-Scanned: No (on mail.in-addr.com); SAEximRunCond expanded to false Cc: freebsd-stable@freebsd.org, Dewayne Geraghty X-BeenThere: freebsd-stable@freebsd.org X-Mailman-Version: 2.1.14 Precedence: list List-Id: Production branch of FreeBSD source code List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 20 Sep 2013 14:36:20 -0000 On Fri, Sep 20, 2013 at 09:12:09AM -0400, Thomas Laus wrote: > > Tom, > > I have had multiple D510's and now D525's that are part of my test > > systems, all are 4GB machines and all run the latest (ie 2 days old) 9.X > > Stable. They're faultless. I have a D510 in production serving 30 > > users - yes its a 1G system running, sendmail, squid, samba as PDC. > > It's been in place for at least 7 months and runs without any hiccups. > > > > Though I would point out that the Atom processor does NOT do out of > > order processing, so a VIA motherboard that is of lower GHz builds > > worlds/ports in less time that a supposedly faster Atom. > > > > Your question re HT, yes HT introduces some additional latency, but is > > unlikely to be the problem. > > > Thanks for the information about the HT CPU's. I asked the question to the > group because I did not know if they were functionally any different than a > traditional CPU. I successfully built my problem port, Tshark, yesterday > while monitoring 'top' on another console. I observed that all 4 cpu's were > in service for the build and at times were running at 100 percent each. The > State column on all 4 occasionally showed a 'pfault' on all 4 but recovered > and the build continued to successful completion. > > > When I experience something like spurious reboots and it is definately > > not hardware, then I delete /usr/src and /usr/ports and perform a > > complete rebuild. (Yes seriously, and on the Atom's we're talking days, > > aren't we :) ) > > > I have been using this Atom D510 since it was released about 3 years ago. It > ran on FreeBSD 8-Stable until about a month ago. I installed an Intel 520 > SSD and loaded a fresh copy of a FreeBSD 9 Snapshot. After getting the > source and ports tarballs, I used svnup to bring both up to date. I built > and installed world and the kernel to bring me up to Stable. I rebuilt all > of my ports using Portmaster. > > The spurious reboot issue existed for the last 3 years when running FreeBSD-8 > Stable. I never had the problem building world or kernel. It only occurred > when building some ports. Subversion and Tshark more often than others. > FreeBSD 9-Stable was frozen when I tried to build tshark, but I was able to > build it OK yesterday. Everything hardware related other than the Atom > microprocessor and the Intel motherboard itself is new. The OS is now a > different version and all of the source was rebuilt monthly. The ports have > been been built many times in the last 3 years. When building kernel & world do you use the '-j' argument to do parallel builds? AFAIK thats not done by default, but it is for some ports. Gary