On 23/09/2016 14:36, Alexander Graf wrote:
>>>
>>> + if ((ret == 1) && timer_ret) {
>>> + /*
>>> + * We have to exit straight away to ensure that we only
>>> + * ever notify user space once about a level change
>>> + */
>>
>> Is this really a requirement? It complicates the logic noticeably.
>
> The alternative would be to track the state on ioctl entry, which
> Christoffer disliked :).
Ok, I see. Maybe he's convinced by the ugly code. :)
It's not an architectural concept, it's a KVM API concept. Tracking the
state on entry just the equivalent of making this vmexit edge-triggered
(both rising and falling edge).
Paolo
--
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
