Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Sun, 23 Feb 2003 18:27:23 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Sun, 23 Feb 2003 18:27:23 -0500 Received: from ppp59-net2.boo.net ([216.200.67.189]:48051 "EHLO kaneda.boo.net") by vger.kernel.org with ESMTP id ; Sun, 23 Feb 2003 18:27:21 -0500 Message-Id: <3.0.6.32.20030223184931.00825e30@boo.net> X-Mailer: QUALCOMM Windows Eudora Light Version 3.0.6 (32) Date: Sun, 23 Feb 2003 18:49:31 -0500 To: linux-mm@kvack.org, linux-kernel@vger.kernel.org From: Jason Papadopoulos Subject: [PATCH] page coloring for 2.5.62 kernel, version 1 Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1182 Lines: 31 Hello again. This version of the page coloring patch implements "stealth mode", i.e. is as minimal as possible. There are many cleanups, and a few mods for the 2.5 series kernel. The biggest change is that the hot/cold per-cpu lists are individually colored now, and the patch has the effect of randomizing the cache colors of pages pumped into the per-cpu lists. The lists are still lifo and still favor pages just freed. No difference in kernel compile time on the K7 system I have; I suspect that decoupling allocation of pages from allocation of pages to processes dilutes the effectiveness of page coloring, but all of the schemes I can think of to enforce page coloring to processes are much more complicated than the one used by this patch. Patch with /proc output: www.boo.net/~jasonp/page_color-2.5.62-20030223.patch and without: www.boo.net/~jasonp/page_color-2.5.62-20030223a.patch Feedback of any sort welcome. jasonp - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/