Return-Path: Date: Wed, 24 Oct 2007 10:43:52 +0200 From: Cornelia Huck To: "Alon Bar-Lev" Cc: "Marcel Holtmann" , "Jiri Kosina" , linux-kernel@vger.kernel.org, bluez-devel@lists.sourceforge.net Subject: Re: [BUG] rfcomm] Message-ID: <20071024104352.294d3b03@gondolin.boeblingen.de.ibm.com> In-Reply-To: <9e0cf0bf0710231155n71cf38fey4d1edae6401db173@mail.gmail.com> References: <200710231950.31313.alon.barlev@gmail.com> <1193162106.6184.208.camel@violet> <1193163905.6184.214.camel@violet> <9e0cf0bf0710231127o485a52e1sc7a06df7c8bc0189@mail.gmail.com> <1193164210.6184.221.camel@violet> <9e0cf0bf0710231155n71cf38fey4d1edae6401db173@mail.gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII List-ID: On Tue, 23 Oct 2007 20:55:54 +0200, "Alon Bar-Lev" wrote: > On 10/23/07, Marcel Holtmann wrote: > > the code in net/bluetooth/rfcomm/tty.c contains two calls to > > device_move(). Please comment these out and try again. At least then we > > know that it is in issue on how we use the driver model. > > No bugs... > Looks OK, although I don't know what removed :) Hm, this might be similar to http://marc.info/?l=linux-kernel&m=119124209009812&w=2. To confirm, could you try with CONFIG_DEBUG_DRIVER and CONFIG_DEBUG_KOBJECT set (and the device_move() calls in again)?