Re: [PATCH v3] x86/hvm: Add per-vcpu evtchn upcalls

[David Vrabel <david.vrabel@citrix.com>]

On 10/11/14 11:13, Paul Durrant wrote:
> HVM guests have always been confined to using the domain callback
> via (see HVM_PARAM_CALLBACK_IRQ) to receive event notifications.
> This is usually an IOAPIC vector and is only used if the event
> channel is bound to vcpu 0.
> 
> PV-on-HVM Linux uses a local APIC vector for event channel upcall,
> set using HVM_PARAM_CALLBACK_IRQ by ORing in a special bit (bit 57)
> into the value (see params.h), but this mechanism is not suitable
> in the general case since the vector must be the same on all vcpus
> which cannot be guaranteed on some OS (e.g. Windows).

"local APIC vector" doesn't make much sense to me here since PVHVM Linux
bypasses the APIC entirely for event channels.

Reword to:

"PVHVM Linux uses a pre-defined interrupt vector for the event channel
upcall, set using HVM_PARAM_CALLBACK_IRQ by ORing in a special bit
(bit 57) into the value (see params.h), This mechanism does not assert
the interrupt via the local APIC.

This mechanism is not suitable in the general case since Windows (and
potentially other OSes):

- cannot guarantee the same vector for all VCPUs
- require the upcall to be asserted via the local APIC."

or similar.

David