From mboxrd@z Thu Jan 1 00:00:00 1970 From: Stefan Richter Subject: Re: [PATCH] Fix scsi/scsi_transport.h compile error Date: Mon, 31 Jul 2006 17:26:18 +0200 Message-ID: <44CE211A.2030205@s5r6.in-berlin.de> References: <200607310943.25384.eike-kernel@sf-tec.de> <20060731131446.GB6087@athena.road.mcmartin.ca> <44CE07AA.9030202@s5r6.in-berlin.de> <20060731143839.GC6087@athena.road.mcmartin.ca> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Return-path: Received: from hp3.statik.TU-Cottbus.De ([141.43.120.68]:33205 "EHLO hp3.statik.tu-cottbus.de") by vger.kernel.org with ESMTP id S1030191AbWGaP3q (ORCPT ); Mon, 31 Jul 2006 11:29:46 -0400 In-Reply-To: <20060731143839.GC6087@athena.road.mcmartin.ca> Sender: linux-scsi-owner@vger.kernel.org List-Id: linux-scsi@vger.kernel.org To: Kyle McMartin Cc: Rolf Eike Beer , linux-scsi@vger.kernel.org, James Bottomley Kyle McMartin wrote: > On Mon, Jul 31, 2006 at 03:37:46PM +0200, Stefan Richter wrote: >> That function accesses members of the struct, i.e. the compiler needs to >> know offsets into struct scsi_device at this point. Therefore it needs >> the definition, not a declaration. > > Odd, ISTR that it only needed to be defined at the point the inline was > used. Oh well, my bad. One would think so. Maybe it depends on the compiler version. (Actually I haven't checked scsi_transport.h myself but I recall fixing another header with the same issue. Although I don't remember anymore whether it lead to a warning or to an error and which gcc version I used then.) -- Stefan Richter -=====-=-==- -=== ==--- http://arcgraph.de/sr/