Sean Christopherson <sean.j.christopherson@xxxxxxxxx> writes: > > Why bother preserving backwards compatibility? AIUI, both KVM and guest > will support async #PF iff interrupt delivery is enabled. Why not make > the interrupt delivery approach KVM_ASYNC_PF_V2 and completely redefine the > ABI? E.g. to make it compatible with reflecting !PRESENT faults without a > VM-Exit via Intel's EPT Violation #VE? > That's the plan, actually. 'page not present' becomes #VE and 'page present' becomes an interrupt (we don't need to inject them synchronously). These two parts are fairly independent but can be merged to reuse the same new capability/CPUID. -- Vitaly
