Re: [RFC -v4 PATCH 3/3] kvm: use yield_to instead of sleep in kvm_vcpu_on_spin

[Rik van Riel <riel@redhat.com>]

On 01/13/2011 08:16 AM, Avi Kivity wrote:

>> + for (pass = 0; pass< 2&& !yielded; pass++) {
>> + kvm_for_each_vcpu(i, vcpu, kvm) {
>> + struct task_struct *task = vcpu->task;
>> + if (!pass&& i< last_boosted_vcpu) {
>> + i = last_boosted_vcpu;
>> + continue;
>> + } else if (pass&& i> last_boosted_vcpu)
>> + break;
>> + if (vcpu == me)
>> + continue;
>> + if (!task)
>> + continue;
>> + if (waitqueue_active(&vcpu->wq))
>> + continue;
>
> Suppose the vcpu exits at this point, and its task terminates.

Arghh, good point.

> I think the first patch needs some reference counting... I'd move it to
> the outermost KVM_RUN loop to reduce the performance impact.

I don't see how refcounting from that other thread could
possibly help, and I now see that the task_struct_cachep
does not have SLAB_DESTROY_BY_LRU, either :(

What do you have in mind here that would both work and
be acceptable to you as KVM maintainer?

-- 
All rights reversed