On Sat, Apr 9, 2011 at 4:26 PM, Anthony Liguori <anthony@xxxxxxxxxxxxx> wrote:
> Right now, you issue synchronous I/O on the VCPU exit path. Blocking the
> VCPU to do I/O not only slows down I/O (since you're only handling one
> request at a time) but it has secondary affects on the guest since normally
> the guest can do useful work while waiting for I/O to complete.
>
> You really want to defer I/O to a thread pool to allow the VCPU to continue
> execution while the I/O is being processed.
Yup, we definitely need to do that.
Pekka
--
To unsubscribe from this list: send the line "unsubscribe kvm" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
