From owner-freebsd-ppc@FreeBSD.ORG Mon Sep 6 22:10:22 2010 Return-Path: Delivered-To: freebsd-ppc@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 0DBB91065713; Mon, 6 Sep 2010 22:10:22 +0000 (UTC) (envelope-from andreast@gcc.gnu.org) Received: from smtp.fgznet.ch (mail.fgznet.ch [81.92.96.47]) by mx1.freebsd.org (Postfix) with ESMTP id 845FE8FC22; Mon, 6 Sep 2010 22:10:20 +0000 (UTC) Received: from deuterium.andreas.nets (dhclient-91-190-8-131.flashcable.ch [91.190.8.131]) by smtp.fgznet.ch (8.13.8/8.13.8/Submit_SMTPAUTH) with ESMTP id o86LlfcO047944; Mon, 6 Sep 2010 23:47:42 +0200 (CEST) (envelope-from andreast@gcc.gnu.org) Message-ID: <4C85617D.4080207@gcc.gnu.org> Date: Mon, 06 Sep 2010 23:47:41 +0200 From: Andreas Toler User-Agent: Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10.5; en-US; rv:1.9.2.8) Gecko/20100802 Lightning/1.0b2 Thunderbird/3.1.2 MIME-Version: 1.0 To: Alexander Motin References: <4C83BB60.6050400@FreeBSD.org> In-Reply-To: <4C83BB60.6050400@FreeBSD.org> Content-Type: text/plain; charset=KOI8-R; format=flowed Content-Transfer-Encoding: 7bit X-Scanned-By: MIMEDefang 2.64 on 81.92.96.47 Cc: freebsd-ppc@freebsd.org Subject: Re: [RFC] Event timers on PowerPC X-BeenThere: freebsd-ppc@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Porting FreeBSD to the PowerPC List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 06 Sep 2010 22:10:22 -0000 On 05.09.10 17:46, Alexander Motin wrote: > Hi. > > I have refactored PowerPC event timers code to use new event timers > infrastructure. I have successfully tested it on PowerMac G4 and > compiled for MPC85XX. > > Patch for HEAD can be found here: > http://people.freebsd.org/~mav/timers_powerpc.patch > > With few additional lines: > http://people.freebsd.org/~mav/timers_powerpc2.patch > this patch successfully utilizes dynamic ticks mode of my latest > one-shot timers patch: > http://people.freebsd.org/~mav/timers_oneshot8.patch > , generating only about 50 interrupts per second in idle. With my old > Motorola PowerPC 7400 CPU I haven't noticed any thermal difference, > though it would be interesting to try it on faster/consumptive CPUs. > > Comments and reviews welcome. I built and booted the kernel on my G5 SMP. Below the figures before and after the patch. Both runs were taken after a buildkernel. Do you want me to look for some specific figures? Andreas G5 before: ---------- [andreast@bohrium] /usr/src/> vmstat -i interrupt total rate irq42: iichb0 1963 0 irq312: htpic0 1964 0 irq316: aoagpio0 1 0 irq282: iichb1 30 0 irq257: pcm0 1 0 irq283: ohci0 1 0 irq284: ohci1 1 0 irq319: ohci2 ohci+ 5597 0 irq295: ata0 39 0 irq296: fwohci0 1 0 irq297: gem0 1148434 156 irq256: atapci0 65198 8 irq376: IPI 669014 91 Total 1892244 257 G5 after: --------- [bohrium:~] andreast% vmstat -i interrupt total rate cpu0:decrementer 1957106 783 irq42: iichb0 994 0 irq312: htpic0 995 0 irq316: aoagpio0 1 0 irq282: iichb1 30 0 irq257: pcm0 1 0 irq283: ohci0 1 0 irq284: ohci1 1 0 irq319: ohci2 ohci+ 1980 0 irq295: ata0 39 0 irq296: fwohci0 1 0 irq297: gem0 1115680 446 irq256: atapci0 55529 22 irq376: IPI 445211 178 cpu1:decrementer 176153140 70517 Total 179730709 71949