From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S933427Ab2FHP5l (ORCPT ); Fri, 8 Jun 2012 11:57:41 -0400 Received: from www.hansjkoch.de ([178.63.77.200]:49349 "EHLO www.hansjkoch.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932388Ab2FHP5j (ORCPT ); Fri, 8 Jun 2012 11:57:39 -0400 Date: Fri, 8 Jun 2012 17:57:29 +0200 From: "Hans J. Koch" To: Dominic Eschweiler Cc: "Hans J. Koch" , "Michael S. Tsirkin" , Greg Kroah-Hartman , kvm@vger.kernel.org, linux-kernel@vger.kernel.org Subject: Re: [PATCH] uio_pci_generic does not export memory resources Message-ID: <20120608155728.GB9705@local> References: <1339156616.3870.9.camel@blech> <20120608130351.GB1964@redhat.com> <1339165738.3870.18.camel@blech> <20120608151843.GA9705@local> <1339170312.26423.6.camel@tesla> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1339170312.26423.6.camel@tesla> User-Agent: Mutt/1.5.21 (2010-09-15) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Fri, Jun 08, 2012 at 05:45:12PM +0200, Dominic Eschweiler wrote: > Am Freitag, den 08.06.2012, 17:18 +0200 schrieb Hans J. Koch: > > Then there's something fundamentally wrong in your driver. Check the > > return value of uio_register_device(). > > > I'm talking about the uio_pci_generic module, not about UIO in general. > > > > I was very confused when I tried UIO the first time > > > and it did not behave like it is described in the documentation. > > > > UIO is the mainline since 2007, and I can assure you it works like > > described. Lots of people use it. > > > I know and I really don't have a problem with UIO itself. The > uio_pci_generic module does not export correctly, maybe you take a look > on its code? My guess is, that the uio_pci_generic module isn't commonly > used and therefore the problem wasn't reported before. As Michael already said, the intention was to map BAR ressources using /sys/bus/pci/devices/... OK, I agree, that's the standard PCI way of doing this, not the standard UIO way. What problem do you have with this approach? Thanks, Hans