From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S263963AbUBDTc5 (ORCPT ); Wed, 4 Feb 2004 14:32:57 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S263996AbUBDTc4 (ORCPT ); Wed, 4 Feb 2004 14:32:56 -0500 Received: from kinesis.swishmail.com ([209.10.110.86]:13581 "EHLO kinesis.swishmail.com") by vger.kernel.org with ESMTP id S263963AbUBDTcx (ORCPT ); Wed, 4 Feb 2004 14:32:53 -0500 Message-ID: <40214A11.3060007@techsource.com> Date: Wed, 04 Feb 2004 14:37:53 -0500 From: Timothy Miller MIME-Version: 1.0 To: root@chaos.analogic.com CC: Alok Mooley , Dave Hansen , linux-kernel , linux-mm Subject: Re: Active Memory Defragmentation: Our implementation & problems References: <20040204185446.91810.qmail@web9705.mail.yahoo.com> In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Richard B. Johnson wrote: > If this is an Intel x86 machine, it is impossible for pages > to get fragmented in the first place. The hardware allows any > page, from anywhere in memory, to be concatenated into linear > virtual address space. Even the kernel address space is virtual. > The only time you need physically-adjacent pages is if you > are doing DMA that is more than a page-length at a time. The > kernel keeps a bunch of those pages around for just that > purpose. > > So, if you are making a "memory defragmenter", it is a CPU time-sink. > That's all. Would memory fragmentation have any appreciable impact on L2 cache line collisions? Would defragmenting it help? In the case of the Opteron, there is a 1M cache that is (I forget) N-way set associative, and it's physically indexed. If a bunch of pages were located such that there were a disproportionately large number of lines which hit the same tag, you could be thrashing the cache. There are two ways to deal with this: (1) intelligently locates pages in physical memory; (2) hope that natural entropy keeps things random enough that it doesn't matter.