From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from moutng.kundenserver.de (moutng.kundenserver.de [212.227.126.179]) by ozlabs.org (Postfix) with ESMTP id 05094DDE42 for ; Wed, 1 Aug 2007 08:07:27 +1000 (EST) From: Arnd Bergmann To: Dave Jiang Subject: Re: [PATCH 2/2] powerpc: MPC85xx EDAC device driver Date: Wed, 1 Aug 2007 00:07:12 +0200 References: <20070726222225.GB10427@blade.az.mvista.com> <200707302158.16530.arnd@arndb.de> <46AFA040.7070303@mvista.com> In-Reply-To: <46AFA040.7070303@mvista.com> MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Message-Id: <200708010007.12922.arnd@arndb.de> Cc: linuxppc-dev@ozlabs.org, bluesmoke-devel@lists.sourceforge.net, norsk5@yahoo.com List-Id: Linux on PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , On Tuesday 31 July 2007, Dave Jiang wrote: > Actually it seems that for me to grab the interrupt number I have to do the > platform device creation in fsl_soc.c and call arch_init() instead of doing it > from fsl_add_bridge(). fsl_add_bridge() is called way too early and the mpic > interrupt mapping has not been setup yet for me to acquire the interrupt number > from of_interrupt_to_resource() call. I think in general, it would be a better solution to have the add_bridge code called much later, like we do with the 64 bit PCI code when using an of_device for the PCI bridge. I don't think it's easy to do though, so that shouldn't stop you from doing it the other way. Please make the platform_device a child of the PCI host bridge though, so it doesn't need to change if/when the PCI initialization gets changed to allow doing it from fsl_pci. Arnd <><