From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:55435) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1aWjPf-0003Yn-Dp for qemu-devel@nongnu.org; Fri, 19 Feb 2016 06:39:56 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1aWjPa-0006vq-BL for qemu-devel@nongnu.org; Fri, 19 Feb 2016 06:39:55 -0500 Received: from mx1.redhat.com ([209.132.183.28]:60363) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1aWjPa-0006vd-5g for qemu-devel@nongnu.org; Fri, 19 Feb 2016 06:39:50 -0500 Date: Fri, 19 Feb 2016 19:39:34 +0800 From: Peter Xu Message-ID: <20160219113934.GB11961@pxdev.xzpeter.org> References: <1455852618-5224-1-git-send-email-peterx@redhat.com> <56C6BA42.1050302@web.de> <20160219074300.GC17229@pxdev.xzpeter.org> <56C6D3A0.9070302@web.de> <20160219092931.GD17229@pxdev.xzpeter.org> <56C6E734.90204@redhat.com> <56C6EB2A.1010503@web.de> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline In-Reply-To: <56C6EB2A.1010503@web.de> Subject: Re: [Qemu-devel] [PATCH 00/13] IOMMU: Enable interrupt remapping for Intel IOMMU List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Jan Kiszka Cc: Rita Sinha , ehabkost@redhat.com, Radim =?utf-8?B?S3LEjW3DocWZ?= , jasowang@redhat.com, mst@redhat.com, qemu-devel@nongnu.org, imammedo@redhat.com, Paolo Bonzini , rth@twiddle.net On Fri, Feb 19, 2016 at 11:15:06AM +0100, Jan Kiszka wrote: > On 2016-02-19 10:58, Paolo Bonzini wrote: > > I think you're not verifying the SVT, SID and SQ fields in the IRTE. > > Exactly. > > > > > The source ID can be passed to the IOMMU using the MemTxAttrs mechanism. > > Ah, that's a nice new channel, resolving the need for using/passing > source-specific target memory regions for this. At least for PCI > devices, it should already be populated with the required information, > others (IOAPIC, HPET) probably require additional work to pass what I > defined as Q35_PSEUDO_BUS_PLATFORM, Q35_PSEUDO_DEVFN_IOAPIC/HPET. Jan, Rita, Paolo, One more thing to ask: do we have any plan to move on the work to support vhost/pass-through devices? Do you think we need these too? Thanks. Peter