On Thu, Dec 29, 2011 at 10:31 PM, Avi Kivity <avi@redhat.com> wrote:
> On 12/29/2011 04:03 PM, Liu ping fan wrote:
>> > Why do we want an independent grace period, is hotunplugging a vcpu that
>> > much different from hotunplugging memory?
>> >
>> I thought that if less readers on the same srcu lock, then
>> synchronize_srcu_expedited() may success to return more quickly.
>
> It would be good to measure it, otherwise it's premature optimization.
>
Yes, after using kprobetrace to measure it, I found it was  premature
optimization. So I will resort to the kvm->srcu, instead of creating a
new one in next version.

Thanks and regards
ping fan
> --
> error compiling committee.c: too many arguments to function
>