From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754674Ab1JUR5d (ORCPT ); Fri, 21 Oct 2011 13:57:33 -0400 Received: from service87.mimecast.com ([91.220.42.44]:56408 "EHLO service87.mimecast.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753333Ab1JUR5c convert rfc822-to-8bit (ORCPT ); Fri, 21 Oct 2011 13:57:32 -0400 Subject: Re: [PATCH v3] virtio: Add platform bus driver for memory mapped virtio device From: Pawel Moll To: Rusty Russell Cc: "linux-kernel@vger.kernel.org" , "linux-arm-kernel@lists.infradead.org" , "virtualization@lists.linux-foundation.org" , Anthony Liguori , "Michael S.Tsirkin" In-Reply-To: <87ty75adtb.fsf@rustcorp.com.au> References: <1317841030-17758-1-git-send-email-pawel.moll@arm.com> <1317919843-8343-1-git-send-email-pawel.moll@arm.com> <1317921230.28696.102.camel@hornet.cambridge.arm.com> <87lisjq6tu.fsf@rustcorp.com.au> <1318931075.3146.285.camel@hornet.cambridge.arm.com> <87ty75adtb.fsf@rustcorp.com.au> Date: Fri, 21 Oct 2011 18:57:24 +0100 Message-ID: <1319219844.5094.16.camel@hornet.cambridge.arm.com> Mime-Version: 1.0 X-Mailer: Evolution 2.28.1 X-OriginalArrivalTime: 21 Oct 2011 17:57:24.0651 (UTC) FILETIME=[E3746FB0:01CC901A] X-MC-Unique: 111102118572800201 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8BIT Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Wed, 2011-10-19 at 03:57 +0100, Rusty Russell wrote: > > Oh, really? My host-side implementation is just doing that: > > > > addr += align - 1; > > addr &= ~(align - 1); > > OK, so you're assuming power of 2. Make sure you kill the guest or at > least the device if it's not though. Yep, I have assertions all around such places :-) (it's a non-production code yet so I can do that) > > \item The dynamic configuration changes, as described in p. 2.4.3 > > ``Dealing With Configuration Changes'' are not permitted. > > This means some devices simply won't work, at least in theory. Why > don't you support this? Uh. I simply forgot about it - my Host block device doesn't do that, so I ignored that feature initially and then it slipped through cracks. And till now I didn't realize that most of the drivers actually use this :-O My fault. Simple to fix anyway - I'll just add InterruptStatus register and use second bit (same with InterruptACK) to get this through. Will be done on Monday. Any other final complaints regarding the interface while I'm on it? ;-) Cheers! Paweł