From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S265702AbTFNSzG (ORCPT ); Sat, 14 Jun 2003 14:55:06 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S265703AbTFNSzG (ORCPT ); Sat, 14 Jun 2003 14:55:06 -0400 Received: from dsl081-085-006.lax1.dsl.speakeasy.net ([64.81.85.6]:33216 "EHLO jyro.mirai.cx") by vger.kernel.org with ESMTP id S265702AbTFNSy7 (ORCPT ); Sat, 14 Jun 2003 14:54:59 -0400 Message-ID: <3EEB72C0.80702@tmsusa.com> Date: Sat, 14 Jun 2003 12:08:48 -0700 From: Joe User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.4b) Gecko/20030612 X-Accept-Language: en-us, en MIME-Version: 1.0 To: linux-kernel Subject: Re: Xeon processors &&Hyper-Threading References: <3EE9FDFA.6020803@mindspring.com> <3EEB6A4B.7040106@mindspring.com> In-Reply-To: <3EEB6A4B.7040106@mindspring.com> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Don't linux vendors ship a variety of precompiled kernels to match the various hardware scenarios? I'm trying to imagine why a customer would ever want or need to roll their own... Joe Joe wrote: > Thanks, I was asked this by one of our clients, and told them that > they would probably have to recompile their kernel. > > Joe > > > Richard B. Johnson wrote: > >> On Fri, 13 Jun 2003, Joe wrote: >> >> >>> Does Linux support the Xeon (p4) processor and its capabilities? >>> >>> The company I work for recently ported its application to Linux and one >>> of our current HP clients asked this and I figure it would be just a >>> recompile the kernel as a P4, but not sure if this would do it. >>> >>> I'm not asking if Linux can RUN the Xeon processor. >>> >>> I'm asking if Linux processor takes any advantage of the >>> Hyper-Threading >>> built into this processor? >>> >>> below is a link to more info on this. >>> >>> http://www.intel.com/design/xeon/prodbref/ >>> >>> Joe >>> >> >> >> You recompile the kernel for SMP as well as P4. If the motherboard >> hasn't disabled HT capabilities, you will take full advantage of >> the processor under Linux. Whatever "full advantage" means, is >> not absolute, but whatever it is, will be used to its fullest. >> Basically, if the code is I/O bound, you'll not see any difference. >> If the code is compute-intensive, you will. >> >> Cheers, >> Dick Johnson >> Penguin : Linux version 2.4.20 on an i686 machine (797.90 BogoMips). >> Why is the government concerned about the lunatic fringe? Think about >> it. >>