From mboxrd@z Thu Jan 1 00:00:00 1970 From: Carsten Otte Subject: Re: [PATCH/RFC 7/9] Virtual network guest device driver Date: Wed, 23 May 2007 10:15:03 +0200 Message-ID: <4653F807.2010209@de.ibm.com> References: <8FFF7E42E93CC646B632AB40643802A8032793AC@scsmsx412.amr.corp.intel.com> Reply-To: carsteno-tA70FqPdS9bQT0dZR+AlfA@public.gmane.org Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Cc: Jimi Xenidis , jmk-zzFmDc4TPjtKvsKVC3L/VUEOCMrvLtNR@public.gmane.org, Christian Borntraeger , mschwid2-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org, Christoph Hellwig , Suzanne McIntosh , kvm-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org To: "Nakajima, Jun" Return-path: In-Reply-To: <8FFF7E42E93CC646B632AB40643802A8032793AC-1a9uaKK1+wJcIJlls4ac1rfspsVTdybXVpNB7YpNyf8@public.gmane.org> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: kvm-devel-bounces-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org Errors-To: kvm-devel-bounces-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org List-Id: kvm.vger.kernel.org I have been closely following thisvery interresting discussion. Here's my summary: - PV capabilities is something we'll want - being able to surface virtual devices to the guest as PCI is preferable to Windows - we need an additional way to surface virtual devices to the guest. We don't have PCI on s390, and Ron doesn't want PCI in his guests. - complex interfaces are a mess to implement and maintain in different hypervisors and guest operating systems, we need a simple and clear structure like plan9 has today To me, it looks like we need a virtual device abstraction both in the guest kernel and in the kvm/qemu. This abstraction needs to be simple and fast, and needs to be representable as PCI device and in a simpler way. PCI obstacles are supposed to be transparent to the virutal device. For me, plan9 does provide answers to a lot of above requirements. However, it does not provide capabilities for shared memory and it adds extra complexity. It's been designed to solve a different problem. I think the virtual device abstraction should provide the following functionality: - hypercall guest to host with parameters and return value - interrupt from host to guest with parameters - thin interrupt from host to guest, no parameters - shared memory between guest and host - dma access to guest memory, possibly via kmap on the host - copy from/to guest memory so long, Carsten ------------------------------------------------------------------------- This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/