From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S262105AbULQSgU (ORCPT ); Fri, 17 Dec 2004 13:36:20 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S262107AbULQSgU (ORCPT ); Fri, 17 Dec 2004 13:36:20 -0500 Received: from almesberger.net ([63.105.73.238]:33287 "EHLO host.almesberger.net") by vger.kernel.org with ESMTP id S262105AbULQSgP (ORCPT ); Fri, 17 Dec 2004 13:36:15 -0500 Date: Fri, 17 Dec 2004 15:36:02 -0300 From: Werner Almesberger To: linux-kernel@vger.kernel.org Cc: Rajesh Venkatasubramanian , Andrew Morton Subject: [PATCH 0/3] prio_tree generalization Message-ID: <20041217153602.D31842@almesberger.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org This patch set for 2.6.10-rc3-bk10 (*) generalizes the prio_tree code such that other subsystems than just VMA can use it. (*) I'm not sure which tree is the most useful base for this. Should I use 2.6.10-rc3-mm* instead of -bk* ? (Or is it already too late for 2.6.10 anyway ?) While there are currently no other in-tree users than VMA, the prio_tree algorithm will be useful for detecting overlapping disk IO requests, which is a prerequisite for efficient handling of barriers, which - besides being a good idea in general - in turn is a prerequisite for useful disk IO priorities. The three patches that follow are incremental and must be applied in the order specified. - Werner -- _________________________________________________________________________ / Werner Almesberger, Buenos Aires, Argentina werner@almesberger.net / /_http://www.almesberger.net/____________________________________________/