From owner-freebsd-smp@FreeBSD.ORG Thu Jan 12 10:28:50 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 7962216A41F for ; Thu, 12 Jan 2006 10:28:50 +0000 (GMT) (envelope-from david@landgren.net) Received: from sferics.mongueurs.net (sferics.mongueurs.net [81.80.147.197]) by mx1.FreeBSD.org (Postfix) with ESMTP id 1DBD943D46 for ; Thu, 12 Jan 2006 10:28:47 +0000 (GMT) (envelope-from david@landgren.net) Received: from [127.0.0.1] (exo.bpinet.com [81.80.147.206]) by sferics.mongueurs.net (Postfix) with ESMTP id E5132AC6F for ; Thu, 12 Jan 2006 11:28:45 +0100 (CET) Message-ID: <43C62F16.8030401@landgren.net> Date: Thu, 12 Jan 2006 11:27:34 +0100 From: David Landgren Organization: The Lusty Decadent Delights of Imperial Pompeii User-Agent: Thunderbird 1.5 (Windows/20051201) MIME-Version: 1.0 To: freebsd-smp@freebsd.org Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Subject: Benchmarks of 5.4 and 6.0 on a 6-CPU host (HP Netserver LT 6000r) 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: Thu, 12 Jan 2006 10:28:50 -0000 List, A while back, I mentioned that I was going to bring a six-processor box (an HP Netserver LT 6000r) from 5.x to 6.0-STABLE, and someone asked for some before and after benchmarks. With the recent spate of advisories, I figured it was time to recompile the world, which gave me the chance to perform the other side of the benchmark, to see how 6.0 performs. The benchmark was of course to buildworld and buildkernel. Back in november I was running a reasonably recent 5.4-STABLE. I first ran the build with -j12 to give the system a workout, and then afterwards without, for the real thing. Each time I moved /usr/obj to /usr/obj-old, in order to have a fresh /usr/obj directory tree created each time. (In fact, I didn't have much choice in the matter: cruft from 5.4 builds or something or other cause the very first make to fail. Zapping /usr/obj fixed that). In all cases the machine was very lightly loaded. 5.4-STABLE compilation times ---------------------------- # Build world, 12 processes: time env -i make -j12 -DALWAYS_CHECK_MAKE buildworld real 43m28.093s user 152m18.214s sys 41m6.976s # Build word, normal: time env -i make -DALWAYS_CHECK_MAKE buildworld real 157m46.084s user 143m20.122s sys 17m53.311s # Build kernel, 12 processes: time env -i make -j12 -DALWAYS_CHECK_MAKE buildkernel KERNCONF=PROFANE real 18m30.613s user 30m46.221s sys 6m0.858s # Build kernel, normal: time env -i make -DALWAYS_CHECK_MAKE buildkernel KERNCONF=PROFANE real 32m51.860s user 29m53.228s sys 3m30.556s 6.0-STABLE compilation times ---------------------------- # Build world, 12 processes: time env -i make -j12 -DALWAYS_CHECK_MAKE buildworld real 43m40.423s user 155m43.345s sys 30m15.898s (Of note, system time has declined by 25%) # Build word, normal: time env -i make -DALWAYS_CHECK_MAKE buildworld real 160m28.020s user 146m26.300s sys 16m4.736s (No difference) # Build kernel, 12 processes: time env -i make -j12 -DALWAYS_CHECK_MAKE buildkernel KERNCONF=PROFANE real 18m12.984s user 31m31.825s sys 4m26.606s (again, a clear reduction of 25%) # Build kernel, normal: time env -i make -DALWAYS_CHECK_MAKE buildkernel KERNCONF=PROFANE real 33m30.404s user 30m42.769s sys 2m52.994s (no difference) So my naive reading of the above is yes, the kernel 6.0 is quite a bit more efficient on multiprocessor systems, the more the machine is loaded, the more the difference. David Landgren -- "It's overkill of course, but you can never have too much overkill." From owner-freebsd-smp@FreeBSD.ORG Thu Jan 12 16:45:31 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 8E68E16A41F for ; Thu, 12 Jan 2006 16:45:31 +0000 (GMT) (envelope-from delphij@gmail.com) Received: from wproxy.gmail.com (wproxy.gmail.com [64.233.184.202]) by mx1.FreeBSD.org (Postfix) with ESMTP id 1856F43D45 for ; Thu, 12 Jan 2006 16:45:30 +0000 (GMT) (envelope-from delphij@gmail.com) Received: by wproxy.gmail.com with SMTP id 71so431128wra for ; Thu, 12 Jan 2006 08:45:30 -0800 (PST) DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:message-id:date:from:reply-to:to:subject:cc:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; b=IchF6Q9nvEnWEbeCAHqhq/u+paF9X791g7P6pg50NM1Qlu0Elv0l0oVGvvZItnZ/TtfrJZZxPtThYOxjv10UbMYWkm35R4FvhaZ797qGYxb07lABnef/BVUUxjATXGLwu351YuppIeecgGAp9CfIAHDr2Z9eMFnmVJ/spnUNtFM= Received: by 10.65.233.8 with SMTP id k8mr878645qbr; Thu, 12 Jan 2006 08:45:29 -0800 (PST) Received: by 10.65.72.5 with HTTP; Thu, 12 Jan 2006 08:45:29 -0800 (PST) Message-ID: Date: Fri, 13 Jan 2006 00:45:29 +0800 From: Xin LI To: David Landgren In-Reply-To: <43C62F16.8030401@landgren.net> MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Content-Disposition: inline References: <43C62F16.8030401@landgren.net> Cc: freebsd-smp@freebsd.org Subject: Re: Benchmarks of 5.4 and 6.0 on a 6-CPU host (HP Netserver LT 6000r) X-BeenThere: freebsd-smp@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list Reply-To: delphij@delphij.net List-Id: FreeBSD SMP implementation group List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 12 Jan 2006 16:45:31 -0000 Hi, David, On 1/12/06, David Landgren wrote: > The benchmark was of course to buildworld and buildkernel. A quick question: are you sure that you are comparing exactly same workload between the two? I think the build process is different when doing cross-build, so it is not quite comparable if you are going to build same codeset on different FreeBSD releases (the compiler is slightly different between 5.4 and 6.0, too)... Cheers, -- Xin LI http://www.delphij.net From owner-freebsd-smp@FreeBSD.ORG Thu Jan 12 19:06:13 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 654D316A41F for ; Thu, 12 Jan 2006 19:06:13 +0000 (GMT) (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 20BAA43D5C for ; Thu, 12 Jan 2006 19:06:12 +0000 (GMT) (envelope-from julian@elischer.org) Received: from unknown (HELO [10.251.23.146]) ([10.251.23.146]) by a50.ironport.com with ESMTP; 12 Jan 2006 11:06:09 -0800 X-IronPort-Anti-Spam-Filtered: true Message-ID: <43C6A8A0.3080704@elischer.org> Date: Thu, 12 Jan 2006 11:06:08 -0800 From: Julian Elischer User-Agent: Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.7.11) Gecko/20050727 X-Accept-Language: en-us, en MIME-Version: 1.0 To: David Landgren References: <43C62F16.8030401@landgren.net> In-Reply-To: <43C62F16.8030401@landgren.net> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: freebsd-smp@freebsd.org Subject: Re: Benchmarks of 5.4 and 6.0 on a 6-CPU host (HP Netserver LT 6000r) 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: Thu, 12 Jan 2006 19:06:13 -0000 David Landgren wrote: > List, > > A while back, I mentioned that I was going to bring a six-processor > box (an HP Netserver LT 6000r) from 5.x to 6.0-STABLE, and someone > asked for some before and after benchmarks. With the recent spate of > advisories, I figured it was time to recompile the world, which gave > me the chance to perform the other side of the benchmark, to see how > 6.0 performs. > > The benchmark was of course to buildworld and buildkernel. > > Back in november I was running a reasonably recent 5.4-STABLE. I first > ran the build with -j12 to give the system a workout, and then > afterwards without, for the real thing. Each time I moved /usr/obj to > /usr/obj-old, in order to have a fresh /usr/obj directory tree created > each time. (In fact, I didn't have much choice in the matter: cruft > from 5.4 builds or something or other cause the very first make to > fail. Zapping /usr/obj fixed that). > > In all cases the machine was very lightly loaded. > > 5.4-STABLE compilation times > ---------------------------- > > # Build world, 12 processes: > time env -i make -j12 -DALWAYS_CHECK_MAKE buildworld > real 43m28.093s > user 152m18.214s > sys 41m6.976s > > # Build word, normal: > time env -i make -DALWAYS_CHECK_MAKE buildworld > real 157m46.084s > user 143m20.122s > sys 17m53.311s > > # Build kernel, 12 processes: > time env -i make -j12 -DALWAYS_CHECK_MAKE buildkernel KERNCONF=PROFANE > real 18m30.613s > user 30m46.221s > sys 6m0.858s > > # Build kernel, normal: > time env -i make -DALWAYS_CHECK_MAKE buildkernel KERNCONF=PROFANE > real 32m51.860s > user 29m53.228s > sys 3m30.556s > > 6.0-STABLE compilation times > ---------------------------- > > # Build world, 12 processes: > time env -i make -j12 -DALWAYS_CHECK_MAKE buildworld > real 43m40.423s > user 155m43.345s > sys 30m15.898s > > (Of note, system time has declined by 25%) > > # Build word, normal: > time env -i make -DALWAYS_CHECK_MAKE buildworld > real 160m28.020s > user 146m26.300s > sys 16m4.736s > > (No difference) > > # Build kernel, 12 processes: > time env -i make -j12 -DALWAYS_CHECK_MAKE buildkernel KERNCONF=PROFANE > real 18m12.984s > user 31m31.825s > sys 4m26.606s > > (again, a clear reduction of 25%) > > # Build kernel, normal: > time env -i make -DALWAYS_CHECK_MAKE buildkernel KERNCONF=PROFANE > real 33m30.404s > user 30m42.769s > sys 2m52.994s > > (no difference) > > So my naive reading of the above is yes, the kernel 6.0 is quite a bit > more efficient on multiprocessor systems, the more the machine is > loaded, the more the difference. but the real times didn't change.. (in fact got a bit worse in some cases) which I guess is something that could be looked at.. did you build in a chroot so athat the same tools were being used both times :-) > > David Landgren From owner-freebsd-smp@FreeBSD.ORG Fri Jan 13 07:43:31 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 4EF6816A41F for ; Fri, 13 Jan 2006 07:43:31 +0000 (GMT) (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 EC73243D48 for ; Fri, 13 Jan 2006 07:43:30 +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 CB0D71A3C26; Thu, 12 Jan 2006 23:43:30 -0800 (PST) Received: by obsecurity.dyndns.org (Postfix, from userid 1000) id C196851488; Fri, 13 Jan 2006 02:43:29 -0500 (EST) Date: Fri, 13 Jan 2006 02:43:29 -0500 From: Kris Kennaway To: David Landgren Message-ID: <20060113074328.GA53487@xor.obsecurity.org> References: <43C62F16.8030401@landgren.net> Mime-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-sha1; protocol="application/pgp-signature"; boundary="C7zPtVaVf+AK4Oqc" Content-Disposition: inline In-Reply-To: <43C62F16.8030401@landgren.net> User-Agent: Mutt/1.4.2.1i Cc: freebsd-smp@freebsd.org Subject: Re: Benchmarks of 5.4 and 6.0 on a 6-CPU host (HP Netserver LT 6000r) 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, 13 Jan 2006 07:43:31 -0000 --C7zPtVaVf+AK4Oqc Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On Thu, Jan 12, 2006 at 11:27:34AM +0100, David Landgren wrote: > List, >=20 > A while back, I mentioned that I was going to bring a six-processor box= =20 > (an HP Netserver LT 6000r) from 5.x to 6.0-STABLE, and someone asked for= =20 > some before and after benchmarks. With the recent spate of advisories, I= =20 > figured it was time to recompile the world, which gave me the chance to= =20 > perform the other side of the benchmark, to see how 6.0 performs. >=20 > The benchmark was of course to buildworld and buildkernel. Unfortunately this isn't valid because you're not compiling the same code using the same tools. A valid benchmark would=20 * Use the same compiler (e.g. 5.4 gcc toolchain, on 5.4 and 6.0 kernels) * Compile the same code, e.g. a 5.4 world (you can't use a 6.0 world because there is extra bootstrapping compilation enabled when building 6.0 on a 5.4 kernel). The easiest way to do this is to extract a 5.4 world + source tree into a subdirectory, boot your chosen kernel, and chroot to the subdirectory, then time the builds in both cases. Kris --C7zPtVaVf+AK4Oqc Content-Type: application/pgp-signature Content-Disposition: inline -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.2 (FreeBSD) iD8DBQFDx1ogWry0BWjoQKURAl6MAKCuZVXG5jSROy2knG1KpdoY4VKNWQCgio2R eZ48OBe8pW6qb7bSiaYptdM= =wqAf -----END PGP SIGNATURE----- --C7zPtVaVf+AK4Oqc--