From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mailman by lists.gnu.org with tmda-scanned (Exim 4.43) id 1NkJMZ-0007P8-HO for qemu-devel@nongnu.org; Wed, 24 Feb 2010 10:41:23 -0500 Received: from [199.232.76.173] (port=38212 helo=monty-python.gnu.org) by lists.gnu.org with esmtp (Exim 4.43) id 1NkJMZ-0007On-3N for qemu-devel@nongnu.org; Wed, 24 Feb 2010 10:41:23 -0500 Received: from Debian-exim by monty-python.gnu.org with spam-scanned (Exim 4.60) (envelope-from ) id 1NkJMX-0002DS-Gv for qemu-devel@nongnu.org; Wed, 24 Feb 2010 10:41:22 -0500 Received: from mail-qy0-f197.google.com ([209.85.221.197]:34593) by monty-python.gnu.org with esmtp (Exim 4.60) (envelope-from ) id 1NkJMX-0002DM-8j for qemu-devel@nongnu.org; Wed, 24 Feb 2010 10:41:21 -0500 Received: by qyk35 with SMTP id 35so464208qyk.18 for ; Wed, 24 Feb 2010 07:41:20 -0800 (PST) Message-ID: <4B85489D.60607@codemonkey.ws> Date: Wed, 24 Feb 2010 09:41:17 -0600 From: Anthony Liguori MIME-Version: 1.0 Subject: Re: [Qemu-devel] [PATCH] VirtIO Support for Memory Regions References: <20100223205225.14860.50990.stgit@st-brides.cs.ualberta.ca> <4B84432C.6090104@codemonkey.ws> <8286e4ee1002232220l79bf50e4t70cf7591d798cebe@mail.gmail.com> <4B85420A.6090903@codemonkey.ws> In-Reply-To: <4B85420A.6090903@codemonkey.ws> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit List-Id: qemu-devel.nongnu.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Cam Macdonell Cc: Christian Borntraeger , qemu-devel@nongnu.org, Alexander Graf On 02/24/2010 09:13 AM, Anthony Liguori wrote: >> to work on that if we can agree on what the underlying device looks >> like whether it should be VirtIO or just PCI. > Agreed, you had suggested uio_pci for my PCI driver and I'd be happy > > I feel pretty strongly that we shouldn't add shared memory to virtio. > I'm pretty sure Christian would too because it eliminates the > possibility of implementing a copy-based transport for virtio > (virtio-over-ethernet). Actually, ignore my crazy justification based on virtio-over-ethernet :-) Certain hypervisors (like Xen and PHYP) have very limited/non-existent support for shared memory. In the case of Xen, there's a very small pool that can be used for DMA buffers. In the case of PHYP, you have hardware accelerated copying for all I/O. Adding shared memory to virtio means that virtio could not be supported (or at least, all virtio devices) on these hypervisors. Regards, Anthony Liguori