From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S270455AbTGMXj4 (ORCPT ); Sun, 13 Jul 2003 19:39:56 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S270454AbTGMXj4 (ORCPT ); Sun, 13 Jul 2003 19:39:56 -0400 Received: from smtp.bitmover.com ([192.132.92.12]:54161 "EHLO smtp.bitmover.com") by vger.kernel.org with ESMTP id S270451AbTGMXjw (ORCPT ); Sun, 13 Jul 2003 19:39:52 -0400 Date: Sun, 13 Jul 2003 16:54:24 -0700 From: Larry McVoy To: "David S. Miller" Cc: Larry McVoy , roland@topspin.com, alan@storlinksemi.com, linux-kernel@vger.kernel.org, linux-net@vger.kernel.org, netdev@oss.sgi.com Subject: Re: TCP IP Offloading Interface Message-ID: <20030713235424.GB31793@work.bitmover.com> Mail-Followup-To: Larry McVoy , "David S. Miller" , Larry McVoy , roland@topspin.com, alan@storlinksemi.com, linux-kernel@vger.kernel.org, linux-net@vger.kernel.org, netdev@oss.sgi.com References: <20030713004818.4f1895be.davem@redhat.com> <52u19qwg53.fsf@topspin.com> <20030713160200.571716cf.davem@redhat.com> <20030713233503.GA31793@work.bitmover.com> <20030713164003.21839eb4.davem@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20030713164003.21839eb4.davem@redhat.com> User-Agent: Mutt/1.4i X-MailScanner-Information: Please contact the ISP for more information X-MailScanner: Found to be clean X-MailScanner-SpamCheck: not spam (whitelisted), SpamAssassin (score=0.5, required 7, AWL, DATE_IN_PAST_06_12) Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org > The whole point is to not touch any of this data. > > The idea is to push the pages directly into the page cache > of the filesystem. It doesn't work. Measure the cost of the VM operations before you go down this path. Just set up a system call that swaps a page with a kernel allocated buffer and then see how many of those you can do a second. Maybe Linux is so blindingly fast this makes sense but IRIX certainly wasn't, the VM overhead hurt like crazy. Every time I tried to push the page flip idea or offloading or any of that crap, Andy Bechtolsheim would tell "the CPUs will get faster faster than you can make that work". He was right. -- --- Larry McVoy lm at bitmover.com http://www.bitmover.com/lm