Re: [RFC v2 0/2] virtio: Support releasing lock during kick

["Michael S. Tsirkin" <mst@redhat.com>]

On Wed, Sep 07, 2011 at 02:51:39PM +0100, Stefan Hajnoczi wrote:
> This patch allows virtio-blk to release its block queue lock while kicking the
> host.  This improves scalability on SMP guests who would otherwise spin on the
> lock while another vCPU is kicking the host.
> 
> This approach can be used for other virtio devices too.  It simply splits the
> virtqueue_kick() operation into a prepare step which requires that the lock be
> held and the actual notify step which may be performed without the lock.
> Existing virtio drivers may continue to use the virtqueue_kick() interface
> which now does the prepare and notify steps internally.
> 
> I am sending this out as RFC because further performance benchmarking is
> required.  Although we have seen good results in the past, gathering number on
> a wider range of machines and verifying that there is no regression would be
> helpful.

Makes sense to me.
Acked-by: Michael S. Tsirkin <mst@redhat.com>

> Stefan Hajnoczi (2):
>   virtio_ring: split virtqueue_kick prepare/notify
>   virtio_blk: unlock vblk->lock during kick
> 
>  drivers/block/virtio_blk.c   |   10 ++++++++--
>  drivers/virtio/virtio_ring.c |   28 +++++++++++++++++++++-------
>  include/linux/virtio.h       |   13 +++++++++++++
>  3 files changed, 42 insertions(+), 9 deletions(-)
> 
> -- 
> 1.7.5.4