From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S261338AbTIKQRK (ORCPT ); Thu, 11 Sep 2003 12:17:10 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S261352AbTIKQRJ (ORCPT ); Thu, 11 Sep 2003 12:17:09 -0400 Received: from ns.suse.de ([195.135.220.2]:202 "EHLO Cantor.suse.de") by vger.kernel.org with ESMTP id S261338AbTIKQRH (ORCPT ); Thu, 11 Sep 2003 12:17:07 -0400 To: Matthew Wilcox Cc: linux-kernel@vger.kernel.org Subject: Re: Memory mapped IO vs Port IO References: <20030911160116.GI21596@parcelfarce.linux.theplanet.co.uk.suse.lists.linux.kernel> From: Andi Kleen Date: 11 Sep 2003 18:17:02 +0200 In-Reply-To: <20030911160116.GI21596@parcelfarce.linux.theplanet.co.uk.suse.lists.linux.kernel> Message-ID: User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2 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 Matthew Wilcox writes: > There's a lot of drivers in the tree that allow you to access the device > either via IO port space or IO mem space. Here's some examples: My gut feeling is to just fix the drivers to make this runtime switchable and get rid of the compile time options. This would help distributions (who normally want to build conservative by default, but still allow the users easy tuning without recompilation) For that it would be nice if a standard module parameter or maybe sysfs option existed. The overhead of checking for PIO vs mmio at runtime in the drivers should be completely in the noise on any non ancient CPU (both MMIO and PIO typically take hundreds or thousands of CPU cycles for the bus access, having an dynamic function call or an if before that is makes no difference at all) -Andi