From mboxrd@z Thu Jan 1 00:00:00 1970 From: Paolo Bonzini Subject: Re: [RFC PATCH] virtio_console: link vq to port with a private pointer in struct virtqueue Date: Tue, 08 May 2012 08:43:01 +0200 Message-ID: <4FA8C075.3030101@redhat.com> References: <1334756013-11752-1-git-send-email-pbonzini@redhat.com> <87397be80k.fsf@rustcorp.com.au> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <87397be80k.fsf@rustcorp.com.au> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: virtualization-bounces@lists.linux-foundation.org Errors-To: virtualization-bounces@lists.linux-foundation.org To: Rusty Russell Cc: Amit Shah , "Michael S. Tsirkin" , linux-kernel@vger.kernel.org, virtualization@lists.linux-foundation.org List-Id: virtualization@lists.linuxfoundation.org Il 08/05/2012 04:11, Rusty Russell ha scritto: >> > For virtio-scsi multiqueue support I would like to have an easy and >> > fast way to go from a virtqueue to the internal struct for that >> > queue. >> > >> > It turns out that virtio-serial has the same need, but it gets >> > by with a simple list walk. >> > >> > This patch adds a pointer to struct virtqueue that is reserved for >> > the virtio device, and uses it in virtio-serial. > I ike the concept, but share Michael's concern with naming confusion. > > How bad would be it to get rid of the current ->priv and use > container_of() instead? ie. have virtio_pci, virtio_mmio, lguest_bus > and s390's kvm_virtio embed the struct virtqueue? How bad is not the question, it would actually be pretty good... the question is more how hard! :) lguest and s390 are a bit different from the others because ->priv points into a memory-mapped descriptor provided by the host; PCI and MMIO have lots of similarities. Looks doable... Paolo