From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S263874AbTDYL2M (ORCPT ); Fri, 25 Apr 2003 07:28:12 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S263870AbTDYL2M (ORCPT ); Fri, 25 Apr 2003 07:28:12 -0400 Received: from verein.lst.de ([212.34.181.86]:37390 "EHLO verein.lst.de") by vger.kernel.org with ESMTP id S263854AbTDYL2L (ORCPT ); Fri, 25 Apr 2003 07:28:11 -0400 Date: Fri, 25 Apr 2003 13:40:20 +0200 From: Christoph Hellwig To: linux-arch@vger.kernel.org, linux-kernel@vger.kernel.org Subject: [HEADS UP] planned change to will cause arch breakage Message-ID: <20030425134020.A5644@lst.de> Mail-Followup-To: Christoph Hellwig , linux-arch@vger.kernel.org, linux-kernel@vger.kernel.org Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org To allow including in generic code the default implemtation of the dma-mapping API in terms of the PCI API will have to go away. What does this mean for your architecture? - If your architecture doesn't use (arm, i386, parisc): Nothing. - If your architecture doesn't support PCI (cris, s390. um): Nothing, except that your arch won't break anymore if somsone includes dma-mapping.h in generic code. But it would be cool to move over your dma-capable busses to the dma-mapping API... - Everyone else: Unless you properly implement the dma-mapping in your architecture the dma_* APIs won't work anymore for you. The simplest way to get support for it on pci busses is to change your implementation of the pci_* dma functions to the dma_* prototypes and add the following to their top: struct pci_dev *pdev = to_pci_dev(dev); and maybe BUG_ON(dev->bus != &pci_bus_type); Of course it would be much nicer if you could add support for the dma-mapping API to your other dma-capable busses..