From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mailman by lists.gnu.org with tmda-scanned (Exim 4.43) id 1NK877-000513-52 for qemu-devel@nongnu.org; Mon, 14 Dec 2009 05:25:13 -0500 Received: from exim by lists.gnu.org with spam-scanned (Exim 4.43) id 1NK872-0004wx-2F for qemu-devel@nongnu.org; Mon, 14 Dec 2009 05:25:12 -0500 Received: from [199.232.76.173] (port=44728 helo=monty-python.gnu.org) by lists.gnu.org with esmtp (Exim 4.43) id 1NK871-0004wi-Mj for qemu-devel@nongnu.org; Mon, 14 Dec 2009 05:25:07 -0500 Received: from mx1.redhat.com ([209.132.183.28]:10217) by monty-python.gnu.org with esmtp (Exim 4.60) (envelope-from ) id 1NK871-0004l4-EY for qemu-devel@nongnu.org; Mon, 14 Dec 2009 05:25:07 -0500 Message-ID: <4B261269.7080801@redhat.com> Date: Mon, 14 Dec 2009 11:24:41 +0100 From: Gerd Hoffmann MIME-Version: 1.0 References: <20091213204341.GA25823@redhat.com> <4B260846.9020503@redhat.com> <20091214094255.GA32140@redhat.com> In-Reply-To: <20091214094255.GA32140@redhat.com> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Subject: [Qemu-devel] Re: [PATCH RFC] virtio: add features qdev property List-Id: qemu-devel.nongnu.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: "Michael S. Tsirkin" Cc: qemu-devel@nongnu.org On 12/14/09 10:42, Michael S. Tsirkin wrote: > On Mon, Dec 14, 2009 at 10:41:26AM +0100, Gerd Hoffmann wrote: >> On 12/13/09 21:43, Michael S. Tsirkin wrote: >>> Add features property to virtio. This makes it >>> possible to e.g. define machine without indirect >>> buffer support, which is required for 0.10 >>> compatibility. or without hardware checksum >>> support, which is required for 0.11 compatibility. >> >> I'd suggest to add flags for the individual features to the drivers >> which actually use it instead, so you'll have >> >> -device virtio-net-pci,hw-checksum=0 >> >> and >> >> -device virtio-blk-pci,indirect-buffers=0 >> >> cheers, >> Gerd > > Hmm. I hoped to avoid it, there are lots of features so it's a lot of > work and in practice, this will most likely be set by machine > description ... MSI-X aka vectors property is already done this way, so I'd tend to continue this way. It is also more user friendly. Sure, these are most likely not used on a daily base by users, but being able to turn off -- say -- indirect buffers for testing and/or bug hunting reasons without having to construct magic hex numbers from virtio header files would be nice. Can you give a list of features? The patch description sounded like it is just the two listed above ... cheers, Gerd