On Mon, Jan 31, 2011 at 03:07:49PM -0700, Alex Williamson wrote:
> On Tue, 2011-02-01 at 00:02 +0200, Michael S. Tsirkin wrote:
> > On Mon, Jan 31, 2011 at 02:47:34PM -0700, Alex Williamson wrote:
> > > On Mon, 2011-01-31 at 23:19 +0200, Michael S. Tsirkin wrote:
> > > > When MSI is off, each inter
On Tue, 2011-02-01 at 00:02 +0200, Michael S. Tsirkin wrote:
> On Mon, Jan 31, 2011 at 02:47:34PM -0700, Alex Williamson wrote:
> > On Mon, 2011-01-31 at 23:19 +0200, Michael S. Tsirkin wrote:
> > > When MSI is off, each interrupt needs to be bounced through the io
> > > thread when it's set/cleare
On Mon, 2011-01-31 at 23:19 +0200, Michael S. Tsirkin wrote:
> When MSI is off, each interrupt needs to be bounced through the io
> thread when it's set/cleared, so vhost-net causes more context switches and
> higher CPU utilization than userspace virtio which handles networking in
> the same threa
On Mon, Jan 31, 2011 at 02:47:34PM -0700, Alex Williamson wrote:
> On Mon, 2011-01-31 at 23:19 +0200, Michael S. Tsirkin wrote:
> > When MSI is off, each interrupt needs to be bounced through the io
> > thread when it's set/cleared, so vhost-net causes more context switches and
> > higher CPU utili
On 01/31/2011 03:19 PM, Michael S. Tsirkin wrote:
When MSI is off, each interrupt needs to be bounced through the io
thread when it's set/cleared, so vhost-net causes more context switches and
higher CPU utilization than userspace virtio which handles networking in
the same thread.
We'll need to
