From owner-freebsd-current@FreeBSD.ORG Wed Aug 18 21:12:33 2010 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 287FA106564A for ; Wed, 18 Aug 2010 21:12:33 +0000 (UTC) (envelope-from dimitry@andric.com) Received: from tensor.andric.com (cl-327.ede-01.nl.sixxs.net [IPv6:2001:7b8:2ff:146::2]) by mx1.freebsd.org (Postfix) with ESMTP id D91F78FC17 for ; Wed, 18 Aug 2010 21:12:32 +0000 (UTC) Received: from [IPv6:2001:7b8:3a7:0:f4aa:bf1e:4470:d911] (unknown [IPv6:2001:7b8:3a7:0:f4aa:bf1e:4470:d911]) (using TLSv1 with cipher DHE-RSA-CAMELLIA256-SHA (256/256 bits)) (No client certificate requested) by tensor.andric.com (Postfix) with ESMTPSA id F41845C59; Wed, 18 Aug 2010 23:12:30 +0200 (CEST) Message-ID: <4C6C4CBF.1040200@andric.com> Date: Wed, 18 Aug 2010 23:12:31 +0200 From: Dimitry Andric User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.2; en-US; rv:1.9.2.9pre) Gecko/20100814 Lanikai/3.1.3pre MIME-Version: 1.0 To: Peter Jeremy References: <4C673F02.8000805@FreeBSD.org> <20100815013438.GA8958@troutmask.apl.washington.edu> <4C67492C.5020206@FreeBSD.org> <8639ufd78w.fsf@ds4.des.no> <4C6844D8.5070602@andric.com> <86sk2faqdl.fsf@ds4.des.no> <4C6AAA88.5080606@andric.com> <4C6AF13A.1080606@andric.com> <20100818205206.GA55620@server.vk2pj.dyndns.org> In-Reply-To: <20100818205206.GA55620@server.vk2pj.dyndns.org> Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Cc: Alan Cox , current@freebsd.org Subject: Re: Official request: Please make GNU grep the default 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: Wed, 18 Aug 2010 21:12:33 -0000 On 2010-08-18 22:52, Peter Jeremy wrote: >> grep with normal mmap() 1396s >> grep with prefault mmap() 1354s >> grep with regular read() 1354s > > Is this with uncached (ie remount the filesystem on each test) or cached > data? This is all on the same filesystem, and the test file is ~370MB, so eventually all data will be in RAM, most likely. E.g. normal mmap() seems to add a bit of overhead that explains the slower result. > Which filesystem (and does it change for different filesystems > (particularly between UFS and ZFS))? I only checked on UFS2. > And one trial is not statistically valid - especially given the small > differences. How about multiple multiple trials with ministat. The result were averages of three trials; they were fairly close to each other, but I didn't calculate the standard deviation. I was not aware of ministat, which looks like a real handy program. :)