From mboxrd@z Thu Jan 1 00:00:00 1970 From: Dominic Eschweiler Subject: Re: UIO: missing resource mapping Date: Fri, 13 Jul 2012 16:44:07 +0200 Message-ID: <1342190647.6607.37.camel@blech> References: <4FFE7C1F.7080702@gmx.net> <20120712194432.GA2592@local> <20120712231632.GC9317@redhat.com> <1342166955.6607.5.camel@blech> <20120713132223.GA10959@redhat.com> <20120713141816.GB2554@local> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: "Michael S. Tsirkin" , Andreas Schallenberg , "linux-kernel@vger.kernel.org" , Greg Kroah-Hartman , kvm@vger.kernel.org To: "Hans J. Koch" Return-path: In-Reply-To: <20120713141816.GB2554@local> Sender: linux-kernel-owner@vger.kernel.org List-Id: kvm.vger.kernel.org Am Freitag, den 13.07.2012, 16:18 +0200 schrieb Hans J. Koch: > If somebody maps the card's memory through the UIO driver and somebod= y > else also maps it using sysfs, that is possible. What happens if both > write to the same hardware registers is a different topic. Writing a > driver implies some responsibility, even if it's done in userspace.=20 That's true and I would not complain. I'm just saying that there is the possibility to disturb a kernel driver by mapping memory etc. I think i= t is worth to considerate moving the BAR mapping code to UIO. Especially, when I look at the discussions about what effects mappable DMA memory can have on the system security.=20 --=20 Gru=C3=9F Dominic