Re: [Xen-devel] [PATCH v3] xen: introduce VCPUOP_register_runstate_phys_memory_area hypercall

[Julien Grall <julien.grall@xxxxxxx>]

On 11/06/2019 13:26, Andrii Anisov wrote:
> On 11.06.19 15:12, Julien Grall wrote:
> > > Julien, I'm not sure I understand how work on (p?)CPU could be delayed. We 
> > > are here with interrupts enabled, so here guest would just spend his own vcpu 
> > > time budget. 
> > Xen only supports only voluntary preemption.
> Oh, really? Let me look into it a bit closer.
Did you expect Xen to be fully preemptible?

The function to do the scheduling is schedule(). This is either call from a 
softirq or directly in a few places (e.g wait()).
The only place in this path where do_softirq() will be called is on return to 
the guest (see leave_hypervisor_tail).
Cheers,

--
Julien Grall

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel