From owner-freebsd-smp@FreeBSD.ORG Mon Jun 5 09:58:10 2006 Return-Path: X-Original-To: freebsd-smp@freebsd.org Delivered-To: freebsd-smp@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id AC75A16A473 for ; Mon, 5 Jun 2006 09:58:10 +0000 (UTC) (envelope-from paul.lipps@gmail.com) Received: from wx-out-0102.google.com (wx-out-0102.google.com [66.249.82.197]) by mx1.FreeBSD.org (Postfix) with ESMTP id 13A4643D45 for ; Mon, 5 Jun 2006 09:58:09 +0000 (GMT) (envelope-from paul.lipps@gmail.com) Received: by wx-out-0102.google.com with SMTP id i31so709032wxd for ; Mon, 05 Jun 2006 02:58:09 -0700 (PDT) DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:mime-version:to:message-id:content-type:from:subject:date:x-mailer; b=YE+dNE6EgUP8n8RYSEZTwl0K8GuTbiYvwsI5huUrs7yPKXpkwVC4Llia2ztDWVEKIn5vBLfuXyWybOyXYrEPEchzrr0dTgiG5OPt1h5cVndFT4BKyrE+dirISaQrzOCVNkk9xlxT+zR3V9HZ5LFo8yFftQ61eadNUYBHWfrO7z4= Received: by 10.70.23.6 with SMTP id 6mr2387427wxw; Mon, 05 Jun 2006 02:58:09 -0700 (PDT) Received: from ?10.0.1.4? ( [72.64.93.129]) by mx.gmail.com with ESMTP id h20sm4873301wxd.2006.06.05.02.58.08; Mon, 05 Jun 2006 02:58:09 -0700 (PDT) Mime-Version: 1.0 (Apple Message framework v750) To: freebsd-smp@freebsd.org Message-Id: <988CE3F0-B4FB-40B3-AE6E-C079E54BD290@gmail.com> From: Paul Lipps Date: Mon, 5 Jun 2006 04:58:06 -0500 X-Mailer: Apple Mail (2.750) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit X-Content-Filtered-By: Mailman/MimeDel 2.1.5 Subject: High Pitched Whine X-BeenThere: freebsd-smp@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: FreeBSD SMP implementation group List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 05 Jun 2006 09:58:10 -0000 Hello, I am fairly new to FreeBSD (5.4 was my first version) and I have ran into an issue that I think is related to the SMP capabilities of FreeBSD. I apologize in advance if this topic is inappropriate for this mailing list. When I am running FreeBSD, my computer emits a high pitched whining noise. It's not a 'physical' noise like a fan or hard disk grinding, it's more of a electrical whine. It's driving me nuts though, it's like a dog whistle. This does not happen when I run any versions of Windows or Linux. This has been the case for all versions of FreeBSD I have had installed, 5.4, 6.0, and 6.1. I recently discovered when I do: sysctl machdep.cpu_idle_hlt=0 the high pitch whine stops. I have a P4 with HyperThreading, and FreeBSD 6.1 installed a SMP kernel by default. I built a non-SMP kernel, and the high pitch whine is still there. So I did a: sysctl machdep.cpu_idle_hlt=0 after installing the non-SMP kernel and again the whine stopped. I tried disabling HTT in the BIOS under both the SMP and the non-SMP kernel and the whine remained. So regardless if HTT is enabled in the BIOS or not, and regardless if I'm running an SMP kernel or not, the only thing that gets rid of the high pitch whine is: sysctl machdep.cpu_idle_hlt=0 Now there is another problem. When I do the command above, the CPU temperature rises rapidly. I can tell as my cooling system has a auto- throttling fan (Shuttle XPC) and I hear the fan increase in speed quickly and steadily after I 'unhalt' the 2nd logical CPU. So the 2nd logical CPU must be getting stuck in a loop or something, I'm not sure really how to see what is happening. If anyone has any suggestions as to how this issue could be resolved, or could point me in the right direction, I would appreciate it. I would be happy to provide dmesg or any other information if it would be helpful. Thank You for your time, Paul Lipps paul.lipps@gmail.com From owner-freebsd-smp@FreeBSD.ORG Mon Jun 5 23:52:54 2006 Return-Path: X-Original-To: freebsd-smp@freebsd.org Delivered-To: freebsd-smp@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 1A3FD16D649 for ; Mon, 5 Jun 2006 22:58:06 +0000 (UTC) (envelope-from mime@traveller.cz) Received: from ss.eunet.cz (ss.eunet.cz [193.85.228.13]) by mx1.FreeBSD.org (Postfix) with ESMTP id 9691643D55 for ; Mon, 5 Jun 2006 22:58:01 +0000 (GMT) (envelope-from mime@traveller.cz) Received: from localhost.i.cz (ss.eunet.cz [193.85.228.13]) by ss.eunet.cz (8.13.6/8.13.6) with ESMTP id k55MvxMI024009 (version=TLSv1/SSLv3 cipher=RC4-MD5 bits=128 verify=NO); Tue, 6 Jun 2006 00:57:59 +0200 (CEST) (envelope-from mime@traveller.cz) From: Michal Mertl To: Paul Lipps In-Reply-To: <988CE3F0-B4FB-40B3-AE6E-C079E54BD290@gmail.com> References: <988CE3F0-B4FB-40B3-AE6E-C079E54BD290@gmail.com> Content-Type: text/plain Date: Tue, 06 Jun 2006 00:57:57 +0200 Message-Id: <1149548277.1024.25.camel@genius.i.cz> Mime-Version: 1.0 X-Mailer: Evolution 2.6.1 FreeBSD GNOME Team Port Content-Transfer-Encoding: 7bit Cc: freebsd-smp@freebsd.org Subject: Re: High Pitched Whine X-BeenThere: freebsd-smp@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: FreeBSD SMP implementation group List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 05 Jun 2006 23:53:00 -0000 Paul Lipps wrote: > Hello, > > I am fairly new to FreeBSD (5.4 was my first version) and I have ran > into an issue that I think is related to the SMP capabilities of > FreeBSD. I apologize in advance if this topic is inappropriate for > this mailing list. I believe it isn't really appropriate. > When I am running FreeBSD, my computer emits a high pitched whining > noise. It's not a 'physical' noise like a fan or hard disk grinding, > it's more of a electrical whine. It's driving me nuts though, it's > like a dog whistle. This does not happen when I run any versions of > Windows or Linux. This has been the case for all versions of FreeBSD > I have had installed, 5.4, 6.0, and 6.1. I think that someone reported it before and I think the cure was to change the hz setting - try to enter kern.hz=100 in the loader prompt (or add it to /boot/loader.conf). > I recently discovered when I do: > > sysctl machdep.cpu_idle_hlt=0 > > the high pitch whine stops. > > I have a P4 with HyperThreading, and FreeBSD 6.1 installed a SMP > kernel by default. I built a non-SMP kernel, and the high pitch whine > is still there. > > So I did a: > > sysctl machdep.cpu_idle_hlt=0 > > after installing the non-SMP kernel and again the whine stopped. > > I tried disabling HTT in the BIOS under both the SMP and the non-SMP > kernel and the whine remained. So regardless if HTT is enabled in the > BIOS or not, and regardless if I'm running an SMP kernel or not, the > only thing that gets rid of the high pitch whine is: > > sysctl machdep.cpu_idle_hlt=0 > > Now there is another problem. When I do the command above, the CPU > temperature rises rapidly. I can tell as my cooling system has a auto- > throttling fan (Shuttle XPC) and I hear the fan increase in speed > quickly and steadily after I 'unhalt' the 2nd logical CPU. So the 2nd > logical CPU must be getting stuck in a loop or something, I'm not > sure really how to see what is happening. You are probably misunderstanding what machdep.cpu_idle_hlt does - it tells the idle thread(s) that when there's no work for a processor it issues the hlt instuction. Processors enter some light sleeping mode, reduce clockspeed or whatever and therefore get cooler. There are better/more modern methods to reduce the power and get the temperature down - ACPI Cx states and others methods but they aren't generally much available/effective on desktop machines AFAIK. It should normally work out of the box. See cpufreq manual page, sysctl dev.cpu, dev.acpi_throttle, dev.cpufreq, dev.est, dev.p4tcc and hw.acpi.cpu.cx_supported. I haven't seen much of it documented anywhere though and I don't really understand all of it either. With modern FreeBSD, HTT cores aren't actually used for anything by default. Performance improvements of HTT are usually minimal at best anyway and there is known security vulnerability in HTT. To enable HTT in FreeBSD it needs to be enabled in the HW (normally via a BIOS setting) and in FreeBSD - as all the computers with HTT CPU's I have access to are configured to disable it in the BIOS I am not sure how to enable it in FreeBSD - there is probably tunable/sysctl machdep.hyperthreading_allowed and hlt_logical_cpus which can probably be used to turn it on. To see whether HTT really works you should see in top command column 'C' and some processes (other than idle) should be running on CPU 1. > If anyone has any suggestions as to how this issue could be resolved, > or could point me in the right direction, I would appreciate it. I > would be happy to provide dmesg or any other information if it would > be helpful. > > Thank You for your time, > > Paul Lipps > paul.lipps@gmail.com > > _______________________________________________ > freebsd-smp@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-smp > To unsubscribe, send any mail to "freebsd-smp-unsubscribe@freebsd.org" > From owner-freebsd-smp@FreeBSD.ORG Tue Jun 6 00:21:48 2006 Return-Path: X-Original-To: freebsd-smp@freebsd.org Delivered-To: freebsd-smp@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id E95F716B70B for ; Mon, 5 Jun 2006 23:21:57 +0000 (UTC) (envelope-from paul.lipps@gmail.com) Received: from wr-out-0506.google.com (wr-out-0506.google.com [64.233.184.230]) by mx1.FreeBSD.org (Postfix) with ESMTP id CE97243D76 for ; Mon, 5 Jun 2006 23:21:45 +0000 (GMT) (envelope-from paul.lipps@gmail.com) Received: by wr-out-0506.google.com with SMTP id i20so1122855wra for ; Mon, 05 Jun 2006 16:21:44 -0700 (PDT) DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:in-reply-to:references:mime-version:content-type:message-id:cc:content-transfer-encoding:from:subject:date:to:x-mailer; b=XhtaT+3dk4KPpacqOQXBetlg2tnbJPiKdD6INoyCO8N2P7v4yEu6PjIFUOgv3q2ZKLNluKNh2l0dNtjf/mo6n5E/KFMS5kVFRPgpS2fx75EOtfW1MfUEzsrnRIebq8s8Q6WdWv8BwHvFbXJ3qZ9myz5QX6DALGQSAmLSQvhzR3A= Received: by 10.54.91.15 with SMTP id o15mr63288wrb; Mon, 05 Jun 2006 16:21:44 -0700 (PDT) Received: from ?10.0.1.4? ( [72.64.93.129]) by mx.gmail.com with ESMTP id 26sm58471wra.2006.06.05.16.21.43; Mon, 05 Jun 2006 16:21:44 -0700 (PDT) In-Reply-To: <1149548277.1024.25.camel@genius.i.cz> References: <988CE3F0-B4FB-40B3-AE6E-C079E54BD290@gmail.com> <1149548277.1024.25.camel@genius.i.cz> Mime-Version: 1.0 (Apple Message framework v750) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Message-Id: <626767E9-DD5A-4460-A9D0-4BF2774BE2D4@gmail.com> Content-Transfer-Encoding: 7bit From: Paul Lipps Date: Mon, 5 Jun 2006 18:21:43 -0500 To: Michal Mertl X-Mailer: Apple Mail (2.750) Cc: freebsd-smp@freebsd.org Subject: Re: High Pitched Whine X-BeenThere: freebsd-smp@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: FreeBSD SMP implementation group List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 06 Jun 2006 00:21:56 -0000 While it might not have been appropriate, you have been most helpful. :) Could you please let me know what list would have been most appropriate? I will try the cure you mentioned regarding the Hz change. It has also been mentioned that hot glue on the motherboard coils might help. It just seems to me something is not quite right in FreeBSD's hardware detection as I don't have this issue with any version of Linux or Windows I have installed. I understand HTT is not utilized in FreeBSD, which is why it puzzles me that starting with FreeBSD 6.1 an SMP kernel is installed on my machine by default now. Disabling HTT in the BIOS is not an option unfortunately as my other installed operating systems take advantage of HTT. Thanks again, Paul Lipps paul.lipps@gmail.com On Jun 5, 2006, at 5:57 PM, Michal Mertl wrote: > Paul Lipps wrote: >> Hello, >> >> I am fairly new to FreeBSD (5.4 was my first version) and I have ran >> into an issue that I think is related to the SMP capabilities of >> FreeBSD. I apologize in advance if this topic is inappropriate for >> this mailing list. > > I believe it isn't really appropriate. > >> When I am running FreeBSD, my computer emits a high pitched whining >> noise. It's not a 'physical' noise like a fan or hard disk grinding, >> it's more of a electrical whine. It's driving me nuts though, it's >> like a dog whistle. This does not happen when I run any versions of >> Windows or Linux. This has been the case for all versions of FreeBSD >> I have had installed, 5.4, 6.0, and 6.1. > > I think that someone reported it before and I think the cure was to > change the hz setting - try to enter kern.hz=100 in the loader prompt > (or add it to /boot/loader.conf). > >> I recently discovered when I do: >> >> sysctl machdep.cpu_idle_hlt=0 >> >> the high pitch whine stops. >> >> I have a P4 with HyperThreading, and FreeBSD 6.1 installed a SMP >> kernel by default. I built a non-SMP kernel, and the high pitch whine >> is still there. >> >> So I did a: >> >> sysctl machdep.cpu_idle_hlt=0 >> >> after installing the non-SMP kernel and again the whine stopped. >> >> I tried disabling HTT in the BIOS under both the SMP and the non-SMP >> kernel and the whine remained. So regardless if HTT is enabled in the >> BIOS or not, and regardless if I'm running an SMP kernel or not, the >> only thing that gets rid of the high pitch whine is: >> >> sysctl machdep.cpu_idle_hlt=0 >> >> Now there is another problem. When I do the command above, the CPU >> temperature rises rapidly. I can tell as my cooling system has a >> auto- >> throttling fan (Shuttle XPC) and I hear the fan increase in speed >> quickly and steadily after I 'unhalt' the 2nd logical CPU. So the 2nd >> logical CPU must be getting stuck in a loop or something, I'm not >> sure really how to see what is happening. > > You are probably misunderstanding what machdep.cpu_idle_hlt does - it > tells the idle thread(s) that when there's no work for a processor it > issues the hlt instuction. Processors enter some light sleeping mode, > reduce clockspeed or whatever and therefore get cooler. > > There are better/more modern methods to reduce the power and get the > temperature down - ACPI Cx states and others methods but they aren't > generally much available/effective on desktop machines AFAIK. It > should > normally work out of the box. See cpufreq manual page, sysctl dev.cpu, > dev.acpi_throttle, dev.cpufreq, dev.est, dev.p4tcc and > hw.acpi.cpu.cx_supported. I haven't seen much of it documented > anywhere > though and I don't really understand all of it either. > > With modern FreeBSD, HTT cores aren't actually used for anything by > default. Performance improvements of HTT are usually minimal at best > anyway and there is known security vulnerability in HTT. To enable HTT > in FreeBSD it needs to be enabled in the HW (normally via a BIOS > setting) and in FreeBSD - as all the computers with HTT CPU's I have > access to are configured to disable it in the BIOS I am not sure > how to > enable it in FreeBSD - there is probably tunable/sysctl > machdep.hyperthreading_allowed and hlt_logical_cpus which can probably > be used to turn it on. > > To see whether HTT really works you should see in top command > column 'C' > and some processes (other than idle) should be running on CPU 1. > > >> If anyone has any suggestions as to how this issue could be resolved, >> or could point me in the right direction, I would appreciate it. I >> would be happy to provide dmesg or any other information if it would >> be helpful. >> >> Thank You for your time, >> >> Paul Lipps >> paul.lipps@gmail.com >> >> _______________________________________________ >> freebsd-smp@freebsd.org mailing list >> http://lists.freebsd.org/mailman/listinfo/freebsd-smp >> To unsubscribe, send any mail to "freebsd-smp- >> unsubscribe@freebsd.org" >> > From owner-freebsd-smp@FreeBSD.ORG Tue Jun 6 01:35:00 2006 Return-Path: X-Original-To: freebsd-smp@freebsd.org Delivered-To: freebsd-smp@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id EF1E816B055 for ; Tue, 6 Jun 2006 01:03:36 +0000 (UTC) (envelope-from stephen@math.missouri.edu) Received: from sccmmhc92.asp.att.net (sccmmhc92.asp.att.net [204.127.203.212]) by mx1.FreeBSD.org (Postfix) with ESMTP id 279D343D72 for ; Tue, 6 Jun 2006 01:03:29 +0000 (GMT) (envelope-from stephen@math.missouri.edu) Received: from [10.0.0.4] (12-216-248-146.client.mchsi.com[12.216.248.146]) by sccmmhc92.asp.att.net (sccmmhc92) with ESMTP id <20060606010327m9200r50o6e>; Tue, 6 Jun 2006 01:03:28 +0000 Message-ID: <4484D456.10901@math.missouri.edu> Date: Mon, 05 Jun 2006 20:03:18 -0500 From: Stephen Montgomery-Smith User-Agent: Mozilla/5.0 (X11; U; FreeBSD i386; en-US; rv:1.7.13) Gecko/20060531 X-Accept-Language: en-us, en MIME-Version: 1.0 To: Paul Lipps References: <988CE3F0-B4FB-40B3-AE6E-C079E54BD290@gmail.com> In-Reply-To: <988CE3F0-B4FB-40B3-AE6E-C079E54BD290@gmail.com> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Cc: freebsd-smp@freebsd.org Subject: Re: High Pitched Whine X-BeenThere: freebsd-smp@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: FreeBSD SMP implementation group List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 06 Jun 2006 01:35:06 -0000 This reply probably is unrelated to your problem as well as smp, but I had a dual AMD processor which I gave a very cpu intensive numerical computation, taking about one second to do one loop, and several days to do the whole thing. You could literally hear it do the computations, with a high pitched whine just as you described, but with a one second beat to it. Stephen From owner-freebsd-smp@FreeBSD.ORG Tue Jun 6 06:26:43 2006 Return-Path: X-Original-To: freebsd-smp@freebsd.org Delivered-To: freebsd-smp@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 68AC816AD61 for ; Tue, 6 Jun 2006 06:16:53 +0000 (UTC) (envelope-from joseph.koshy@gmail.com) Received: from ug-out-1314.google.com (ug-out-1314.google.com [66.249.92.170]) by mx1.FreeBSD.org (Postfix) with ESMTP id C412043D49 for ; Tue, 6 Jun 2006 06:16:52 +0000 (GMT) (envelope-from joseph.koshy@gmail.com) Received: by ug-out-1314.google.com with SMTP id m3so1734617uge for ; Mon, 05 Jun 2006 23:16:51 -0700 (PDT) DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:message-id:date:from:to:subject:cc:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; b=Q2WtM1Ovp/Uu/b6+ebkYwEZxXvAwf0XExx6Q0aiWY8gWq/tbTaC9aiXEzGP3pC4VND/vJxRtaqUIUSxt9vHIm60xOMC1oR59Zt9KIfWgDPriMMAsfaD4tPlDsfq7tJYbxtKVAf0KSP07vBJ4NTVq3E4YVHenuShQrfxsSohV4W0= Received: by 10.78.51.9 with SMTP id y9mr1057956huy; Mon, 05 Jun 2006 23:10:31 -0700 (PDT) Received: by 10.78.71.19 with HTTP; Mon, 5 Jun 2006 23:10:31 -0700 (PDT) Message-ID: <84dead720606052310s62ad2a5dt884fce4dd192b05d@mail.gmail.com> Date: Tue, 6 Jun 2006 11:40:31 +0530 From: "Joseph Koshy" To: "Paul Lipps" In-Reply-To: <988CE3F0-B4FB-40B3-AE6E-C079E54BD290@gmail.com> MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Content-Disposition: inline References: <988CE3F0-B4FB-40B3-AE6E-C079E54BD290@gmail.com> Cc: freebsd-smp@freebsd.org Subject: Re: High Pitched Whine X-BeenThere: freebsd-smp@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: FreeBSD SMP implementation group List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 06 Jun 2006 06:26:43 -0000 > When I am running FreeBSD, my computer emits a high pitched > whining noise. It's not a 'physical' noise like a fan or hard > disk grinding, it's more of a electrical whine. The only thing that I can think of that could think of that could whine is a flaky magnetic component. Does the frequency of the whine change if you change HZ? Does the whine reduce if the processor is fully compute bound? On an idle system, the processor should quickly go back to sleep after handling the clock interrupt. Since this should take a few uSecs at most, the processor's power requirement surge should be handled by the reserves in the power supply's output capacitors. My guess is that FreeBSD's clock handling is taking long enough that your SMPS needs to (briefly) increase its current output. This change in current output could cause a whine if a magnetic coil in the SMPS was loose. Someone with a spare 'scope could check this hypothesis out ... if there's a correlation between the ripple on the 5V line and the clock rate on an otherwise idle system, then we're spending too many cycles in our clock handler. -- FreeBSD Volunteer, http://people.freebsd.org/~jkoshy From owner-freebsd-smp@FreeBSD.ORG Tue Jun 6 09:06:22 2006 Return-Path: X-Original-To: freebsd-smp@freebsd.org Delivered-To: freebsd-smp@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 2315816AE2C for ; Tue, 6 Jun 2006 09:01:52 +0000 (UTC) (envelope-from paul.lipps@gmail.com) Received: from wr-out-0506.google.com (wr-out-0506.google.com [64.233.184.225]) by mx1.FreeBSD.org (Postfix) with ESMTP id 9813043D4C for ; Tue, 6 Jun 2006 09:01:51 +0000 (GMT) (envelope-from paul.lipps@gmail.com) Received: by wr-out-0506.google.com with SMTP id i20so1206434wra for ; Tue, 06 Jun 2006 02:01:51 -0700 (PDT) DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:in-reply-to:references:mime-version:content-type:message-id:cc:content-transfer-encoding:from:subject:date:to:x-mailer; b=KNgLIyuXo/byhs8D6TrHJeApIaNiT+g+piuFS46pp1kgEZmr5+8rcHQxZEQDOMPEkZ2CTp/NGNco8RG4mQopsUHALwyvE7z83cPtTxdZ0MXyuHZqes9iHOrvus2J1xlcePiSSMDlKqTe/WrbpVTfhtrUSUhrLbBuRxpPNKYsdXw= Received: by 10.54.95.10 with SMTP id s10mr515575wrb; Tue, 06 Jun 2006 02:01:50 -0700 (PDT) Received: from ?10.0.1.4? ( [72.64.93.129]) by mx.gmail.com with ESMTP id 5sm4805187wrh.2006.06.06.02.01.50; Tue, 06 Jun 2006 02:01:50 -0700 (PDT) In-Reply-To: <84dead720606052310s62ad2a5dt884fce4dd192b05d@mail.gmail.com> References: <988CE3F0-B4FB-40B3-AE6E-C079E54BD290@gmail.com> <84dead720606052310s62ad2a5dt884fce4dd192b05d@mail.gmail.com> Mime-Version: 1.0 (Apple Message framework v750) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Message-Id: <4BA3465C-BEAD-4DAE-8975-804BB8EB25E6@gmail.com> Content-Transfer-Encoding: 7bit From: Paul Lipps Date: Tue, 6 Jun 2006 04:01:48 -0500 To: Joseph Koshy X-Mailer: Apple Mail (2.750) Cc: freebsd-smp@freebsd.org Subject: Re: High Pitched Whine X-BeenThere: freebsd-smp@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: FreeBSD SMP implementation group List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 06 Jun 2006 09:06:26 -0000 On Jun 6, 2006, at 1:10 AM, Joseph Koshy wrote: > The only thing that I can think of that could think of that > could whine is a flaky magnetic component. > > Does the frequency of the whine change if you change HZ? > Does the whine reduce if the processor is fully compute > bound? Yes, the whine actually goes away under a full load. It is intermittent when the CPU is being stressed intermittently, and of course constant when the CPU is idle. I am going to try compiling a new kernel with a reduced HZ setting as suggested. Is the output of this command: sysctl kern.clockrate displaying the current HZ setting? If so it's 1000 at the moment. I will compile a new kernel using the GENERIC configuration file with only the HZ option changed to 100 rather then 1000 and report my findings. From owner-freebsd-smp@FreeBSD.ORG Tue Jun 6 09:23:17 2006 Return-Path: X-Original-To: freebsd-smp@freebsd.org Delivered-To: freebsd-smp@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 1766116A5DA for ; Tue, 6 Jun 2006 09:16:32 +0000 (UTC) (envelope-from paul.lipps@gmail.com) Received: from wx-out-0102.google.com (wx-out-0102.google.com [66.249.82.205]) by mx1.FreeBSD.org (Postfix) with ESMTP id 7E02A43D46 for ; Tue, 6 Jun 2006 09:16:31 +0000 (GMT) (envelope-from paul.lipps@gmail.com) Received: by wx-out-0102.google.com with SMTP id i31so891340wxd for ; Tue, 06 Jun 2006 02:16:31 -0700 (PDT) DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:mime-version:to:message-id:content-type:references:from:subject:date:x-mailer; b=MY51hRpr7VLDO4lG4zLBvJsq9A7TwjekQ8aEMbHYbqe8RPKENAqwvD7JMXNUaHrZKA0bkpW2QO2dJm+jynuyoyBh7K6WLGLJbKGi39uuAyN+xZarc5qicdUuQS/XvYMwPKUFE97psY+dAmg3uxit+lrfRG4GmBXqyNG+K1tubAY= Received: by 10.70.73.13 with SMTP id v13mr7241556wxa; Tue, 06 Jun 2006 02:16:30 -0700 (PDT) Received: from ?10.0.1.4? ( [72.64.93.129]) by mx.gmail.com with ESMTP id h20sm5975146wxd.2006.06.06.02.16.30; Tue, 06 Jun 2006 02:16:30 -0700 (PDT) Mime-Version: 1.0 (Apple Message framework v750) To: freebsd-smp@freebsd.org Message-Id: References: <4BA3465C-BEAD-4DAE-8975-804BB8EB25E6@gmail.com> From: Paul Lipps Date: Tue, 6 Jun 2006 04:16:28 -0500 X-Mailer: Apple Mail (2.750) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit X-Content-Filtered-By: Mailman/MimeDel 2.1.5 Subject: Fwd: High Pitched Whine X-BeenThere: freebsd-smp@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: FreeBSD SMP implementation group List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 06 Jun 2006 09:23:18 -0000 Sorry, for some reason I thought I was going to have to build a new kernel. I set my HZ to 100 in the loader.conf, and it did lower the pitch of the whine to where I can hardly hear it now. I can still hear a difference when I toggle the CPU halt off and on, but the difference between 1000 and 100 HZ is like night and day. Thank you very much, my ears appreciate it. :) Paul Lipps paul.lipps@gmail.com Begin forwarded message: > From: Paul Lipps > Date: June 6, 2006 4:01:48 AM CDT > To: Joseph Koshy > Cc: freebsd-smp@freebsd.org > Subject: Re: High Pitched Whine > > On Jun 6, 2006, at 1:10 AM, Joseph Koshy wrote: >> The only thing that I can think of that could think of that >> could whine is a flaky magnetic component. >> >> Does the frequency of the whine change if you change HZ? >> Does the whine reduce if the processor is fully compute >> bound? > > Yes, the whine actually goes away under a full load. It is > intermittent when the CPU is being stressed intermittently, and of > course constant when the CPU is idle. > > I am going to try compiling a new kernel with a reduced HZ setting > as suggested. Is the output of this command: > > sysctl kern.clockrate > > displaying the current HZ setting? If so it's 1000 at the moment. I > will compile a new kernel using the GENERIC configuration file with > only the HZ option changed to 100 rather then 1000 and report my > findings. > > > From owner-freebsd-smp@FreeBSD.ORG Tue Jun 6 14:18:05 2006 Return-Path: X-Original-To: freebsd-smp@freebsd.org Delivered-To: freebsd-smp@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 0F66A16C651 for ; Tue, 6 Jun 2006 14:18:05 +0000 (UTC) (envelope-from julian@elischer.org) Received: from a50.ironport.com (a50.ironport.com [63.251.108.112]) by mx1.FreeBSD.org (Postfix) with ESMTP id 9F3B943D49 for ; Tue, 6 Jun 2006 14:18:01 +0000 (GMT) (envelope-from julian@elischer.org) Received: from unknown (HELO [192.168.3.4]) ([10.251.60.69]) by a50.ironport.com with ESMTP; 06 Jun 2006 07:18:00 -0700 Message-ID: <44858E96.1000500@elischer.org> Date: Tue, 06 Jun 2006 22:17:58 +0800 From: Julian Elischer User-Agent: Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.7.13) Gecko/20060414 X-Accept-Language: en-us, en MIME-Version: 1.0 To: Paul Lipps References: <988CE3F0-B4FB-40B3-AE6E-C079E54BD290@gmail.com> <84dead720606052310s62ad2a5dt884fce4dd192b05d@mail.gmail.com> <4BA3465C-BEAD-4DAE-8975-804BB8EB25E6@gmail.com> In-Reply-To: <4BA3465C-BEAD-4DAE-8975-804BB8EB25E6@gmail.com> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: freebsd-smp@freebsd.org, Joseph Koshy Subject: Re: High Pitched Whine X-BeenThere: freebsd-smp@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: FreeBSD SMP implementation group List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 06 Jun 2006 14:18:08 -0000 Paul Lipps wrote: > On Jun 6, 2006, at 1:10 AM, Joseph Koshy wrote: > >> The only thing that I can think of that could think of that >> could whine is a flaky magnetic component. >> >> Does the frequency of the whine change if you change HZ? >> Does the whine reduce if the processor is fully compute >> bound? > > > Yes, the whine actually goes away under a full load. It is > intermittent when the CPU is being stressed intermittently, and of > course constant when the CPU is idle. > > I am going to try compiling a new kernel with a reduced HZ setting as > suggested. Is the output of this command: > > sysctl kern.clockrate > > displaying the current HZ setting? If so it's 1000 at the moment. I > will compile a new kernel using the GENERIC configuration file with > only the HZ option changed to 100 rather then 1000 and report my > findings. > > > > _______________________________________________ > freebsd-smp@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-smp > To unsubscribe, send any mail to "freebsd-smp-unsubscribe@freebsd.org" The whine is probably coming from an element in the power supply. The hlt systctl that you mention allows the processor to halt when not in use and only wake up when busy or at interrupts (minimum of 'hz' times per second) A halt state uses a lot less power so the current load goes up and down with a 1kHz base and some random variance. When loaded this can not be heard because the current is more ccontinuous. From owner-freebsd-smp@FreeBSD.ORG Tue Jun 6 20:02:13 2006 Return-Path: X-Original-To: smp@freeBSD.org Delivered-To: freebsd-smp@FreeBSD.ORG Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 7DCC916CC33 for ; Tue, 6 Jun 2006 19:59:41 +0000 (UTC) (envelope-from kris@obsecurity.org) Received: from elvis.mu.org (elvis.mu.org [192.203.228.196]) by mx1.FreeBSD.org (Postfix) with ESMTP id 6821D43D46 for ; Tue, 6 Jun 2006 19:59:41 +0000 (GMT) (envelope-from kris@obsecurity.org) Received: from obsecurity.dyndns.org (elvis.mu.org [192.203.228.196]) by elvis.mu.org (Postfix) with ESMTP id 49ABC1A3C24 for ; Tue, 6 Jun 2006 12:59:39 -0700 (PDT) Received: by obsecurity.dyndns.org (Postfix, from userid 1000) id 8EEE651616; Tue, 6 Jun 2006 15:59:38 -0400 (EDT) Date: Tue, 6 Jun 2006 15:59:38 -0400 From: Kris Kennaway To: smp@freeBSD.org Message-ID: <20060606195938.GA6581@xor.obsecurity.org> Mime-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-sha1; protocol="application/pgp-signature"; boundary="/04w6evG8XlLl3ft" Content-Disposition: inline User-Agent: Mutt/1.4.2.1i Cc: Subject: Concluding the SMPng project X-BeenThere: freebsd-smp@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: FreeBSD SMP implementation group List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 06 Jun 2006 20:02:19 -0000 --/04w6evG8XlLl3ft Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable Several of us have been discussing recently whether it is time to bring the SMPng project to a formal conclusion. According to the SMPng project webpage, "The end goal of the SMPng Project is to decompose the Giant lock into a number of smaller locks, resulting in reduced contention (and improved SMP performance)." Thanks to the hard work of many developers over the past ~6 years, this goal is now complete. While Giant has not been completely eliminated from the kernel and several subsystems are still giant-locked (notably ipv6, tty, and CAM, although work is in progress on all of these fronts), kernel profiling traces show that for many real-world application loads the Giant lock is simply no longer a factor in the performance of the SMP kernel. See e.g.=20 http://www.bsdcan.org/2006/papers/FilesystemPerformance.pdf for one such measurement of the extent of Giant locking in FreeBSD 6.x; other real-world application workloads are similar. Some of the benefits of formally concluding the SMPng project are: * The focus of SMP development work has largely changed from "break up Giant everywhere" to "carefully measure the effects of the locking decisions that were made, and optimize for greater performance and scalability". This is a major milestone and should be announced to the world, perhaps under the banner of a new "FreeBSD Scalability Project". * For example, a number of us are looking very closely at the nascent FreeBSD port to the Sun Ultrasparc T1, which provides 32 virtual CPUs (4 threads on 8 CPU cores) on a single chip. Optimizing for the new generation of SMP hardware is going to be a major effort over the coming year. * We get to draw a line under the significant architectural changes of the past 6 years and move forward. Throughout the 5.x branch I think there was still a lot of developer mentality that it represented a work in progress, but I think there's a broad consensus that 6.x is the natural conclusion of that process, and we have a new baseline upon which to build for the future. * There were many people who were unsure whether the goals of the SMPng project could be realised or whether the tasks that the FreeBSD project set for itself were just too ambitious. While SMP work remains ongoing, I think it's fair to say that the project has addressed this criticism admirably, and we should not be shy about advertising this achievement. Kris --/04w6evG8XlLl3ft Content-Type: application/pgp-signature Content-Disposition: inline -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.3 (FreeBSD) iD8DBQFEhd6pWry0BWjoQKURAm1ZAKClsUtg0oRQH2girKdm4fcsbBuICgCfbvPo kiiSKJHd1X5tL6/kIKlDEOc= =Tv+4 -----END PGP SIGNATURE----- --/04w6evG8XlLl3ft-- From owner-freebsd-smp@FreeBSD.ORG Tue Jun 6 20:29:24 2006 Return-Path: X-Original-To: smp@freeBSD.org Delivered-To: freebsd-smp@FreeBSD.ORG Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 58BD216C6B0 for ; Tue, 6 Jun 2006 20:24:20 +0000 (UTC) (envelope-from phk@critter.freebsd.dk) Received: from phk.freebsd.dk (phk.freebsd.dk [130.225.244.222]) by mx1.FreeBSD.org (Postfix) with ESMTP id CF17C43D49 for ; Tue, 6 Jun 2006 20:24:19 +0000 (GMT) (envelope-from phk@critter.freebsd.dk) Received: from critter.freebsd.dk (critter.freebsd.dk [192.168.48.2]) by phk.freebsd.dk (Postfix) with ESMTP id 4D38E170DF; Tue, 6 Jun 2006 20:24:18 +0000 (UTC) To: Kris Kennaway From: "Poul-Henning Kamp" In-Reply-To: Your message of "Tue, 06 Jun 2006 15:59:38 -0400." <20060606195938.GA6581@xor.obsecurity.org> Date: Tue, 06 Jun 2006 20:24:18 +0000 Message-ID: <1915.1149625458@critter.freebsd.dk> Sender: phk@critter.freebsd.dk Cc: smp@freeBSD.org Subject: Re: Concluding the SMPng project X-BeenThere: freebsd-smp@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: FreeBSD SMP implementation group List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 06 Jun 2006 20:29:28 -0000 In message <20060606195938.GA6581@xor.obsecurity.org>, Kris Kennaway writes: >Several of us have been discussing recently whether it is time to >bring the SMPng project to a formal conclusion. As the instigator of this particular notion, I am obviously all for it. -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 phk@FreeBSD.ORG | TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. From owner-freebsd-smp@FreeBSD.ORG Fri Jun 9 00:15:13 2006 Return-Path: X-Original-To: freebsd-smp@freebsd.org Delivered-To: freebsd-smp@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 9D39716A41F for ; Fri, 9 Jun 2006 00:15:04 +0000 (UTC) (envelope-from joseph.koshy@gmail.com) Received: from ug-out-1314.google.com (ug-out-1314.google.com [66.249.92.174]) by mx1.FreeBSD.org (Postfix) with ESMTP id CEE6343D5A for ; Fri, 9 Jun 2006 00:15:03 +0000 (GMT) (envelope-from joseph.koshy@gmail.com) Received: by ug-out-1314.google.com with SMTP id e2so936196ugf for ; Thu, 08 Jun 2006 17:15:00 -0700 (PDT) DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:message-id:date:from:to:subject:cc:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; b=q6cQes9jhGgWZvp24h/5wILrb1oEBoSsTgC8i/MA6d1dXUklTaRYoRLER1w+kX0OIDX34MZ9DPuZA+YD2hdatHRxZJZcYvV3MLj7gg6YuUFt0pDx6bJ70Xd88QuMEcN1Rx1+8rqVMza+jjYDrBls2FfYFevUv3MYFb9ZAimcRYU= Received: by 10.78.57.11 with SMTP id f11mr1119088hua; Tue, 06 Jun 2006 05:28:23 -0700 (PDT) Received: by 10.78.71.19 with HTTP; Tue, 6 Jun 2006 05:28:20 -0700 (PDT) Message-ID: <84dead720606060528i1d776few85eb396b84bb7d5a@mail.gmail.com> Date: Tue, 6 Jun 2006 17:58:20 +0530 From: "Joseph Koshy" To: "Paul Lipps" In-Reply-To: MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Content-Disposition: inline References: <4BA3465C-BEAD-4DAE-8975-804BB8EB25E6@gmail.com> Cc: freebsd-smp@freebsd.org Subject: Re: High Pitched Whine X-BeenThere: freebsd-smp@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: FreeBSD SMP implementation group List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 09 Jun 2006 00:15:13 -0000 pl> Thank you very much, my ears appreciate it. :) The 'whine' indicates mechanical stress somewhere -- something is vibrating at HZ. It might be prudent to investigate and fix the affected part. -- FreeBSD Volunteer, http://people.freebsd.org/~jkoshy