From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1764033AbYEOUOS (ORCPT ); Thu, 15 May 2008 16:14:18 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1754965AbYEOUOB (ORCPT ); Thu, 15 May 2008 16:14:01 -0400 Received: from smtp-out003.kontent.com ([81.88.40.217]:47877 "EHLO smtp-out003.kontent.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753126AbYEOUOA (ORCPT ); Thu, 15 May 2008 16:14:00 -0400 From: Oliver Neukum Organization: NOvell To: Greg KH Subject: Re: [PATCH] USB: add Sensoray 2255 v4l driver Date: Thu, 15 May 2008 22:13:59 +0200 User-Agent: KMail/1.9.9 Cc: mchehab@infradead.org, v4l-dvb-maintainer@linuxtv.org, linux-usb@vger.kernel.org, linux-kernel@vger.kernel.org, video4linux-list@redhat.com References: <20080514205927.GA13134@kroah.com> <200805152154.13630.oliver@neukum.org> <20080515201049.GA24999@kroah.com> In-Reply-To: <20080515201049.GA24999@kroah.com> MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Content-Disposition: inline Message-Id: <200805152214.00400.oliver@neukum.org> Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Am Donnerstag 15 Mai 2008 22:10:49 schrieb Greg KH: > On Thu, May 15, 2008 at 09:54:12PM +0200, Oliver Neukum wrote: > > Am Donnerstag 15 Mai 2008 20:44:24 schrieb Greg KH: > > > On Thu, May 15, 2008 at 02:03:18PM +0200, Oliver Neukum wrote: > > > > Am Donnerstag 15 Mai 2008 13:38:37 schrieb Oliver Neukum: > > > > > 3. The firmware stuff. That's an interesting solution. However: > > > > > > > > Actually, on second thought, I take that back. It's a bad solution. > > > > If you don't want to do it in probe(), the only other sensible place > > > > is in open(). That way you can avoid the whole trouble if nobody > > > > opens the device. And you need to handle the case of unloaded > > > > firmware anyway, so you can trigger firmware load there. > > > > > > No, we want to do firmware load on probe, I'll change it to be async so > > > > Could you state your reasons for that preference? > > I don't want to create the device nodes, and have userspace think the > device really is working, only to have everything fall down and die if > open() is called and the firmware isn't present. I'd rather not create > the video devices if we know this isn't going to work at all. Yes, that makes sense. However, then you might want to look into spawning a thread per bus rather than a delayed failure mechanism. Regards Oliver