From mboxrd@z Thu Jan 1 00:00:00 1970 From: Bartlomiej Zolnierkiewicz Subject: Re: [PATCH 1/1] build modular usb isd200 with modular ide Date: Sun, 24 Oct 2004 14:45:53 +0200 Sender: linux-ide-owner@vger.kernel.org Message-ID: <58cb370e041024054575c09679@mail.gmail.com> References: <200410240055.i9O0taCf006206@falcon10.austin.ibm.com> <20041024100319.GA17183@lst.de> Reply-To: Bartlomiej Zolnierkiewicz Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Return-path: Received: from rproxy.gmail.com ([64.233.170.199]:56818 "EHLO rproxy.gmail.com") by vger.kernel.org with ESMTP id S261464AbUJXMpy (ORCPT ); Sun, 24 Oct 2004 08:45:54 -0400 Received: by rproxy.gmail.com with SMTP id 77so338129rnk for ; Sun, 24 Oct 2004 05:45:53 -0700 (PDT) In-Reply-To: <20041024100319.GA17183@lst.de> List-Id: linux-ide@vger.kernel.org To: Christoph Hellwig Cc: Doug Maxey , Paul Mackerras , Greg Kroah-Hartman , Linux SCSI Mailing List , Linux IDE Mailing List , linuxppc64-dev@ozlabs.org On Sun, 24 Oct 2004 12:03:19 +0200, Christoph Hellwig wrote: > On Sat, Oct 23, 2004 at 07:55:36PM -0500, Doug Maxey wrote: > > > > Name: inline ide_fix_driveid() > > > > Rationale: > > This is a fix for bugme.osdl 3819. > > bugme.osdl.org doesn't know of a bug #3819. > > > With any of the 2.6.9 release flavors (vanilla, mm1, ac3), one > > cannot build the usb isd200 module due to the dependency on > > ide_fix_driveid() being exported from ide-iops. > > > > Description: > > When building IDE modular, the current ide_fix_driveid() is > > exported from ide-iops.c. This patch makes the function an inline. > > Still doesn't make any sense. ide_fix_driveid is properly exported from > ide-iops.c, so you use it from other modules. The only case that > doesn't work is modular ide and builtin usb-storage, and the BLK_DEV_IDE > depency should fix that one. The new ide_fix_driveid function seems buggy, ie. it byte-swaps id->max_multsect with id->vendor3. > If you think that depency is ugly (I do) just copy the routine to > isd200.c, it's a) too large to inline but b) just a trivial byteswap > that should need much changes over time. The dependency is a bug, is for IDE driver only. Doug, if you kill debugging code in isd200.c then only: id->command_set_1 id->model id->fw_rev id->capability id->lba_capacity id->heads id->cyls id->sectors id->command_set_2 need to be byte-swapped.