Re: [PATCH v2 2/3] KVM: x86: introduce KVM_MEM_PCI_HOLE memory

[Sean Christopherson <sean.j.christopherson@intel.com>]

On Fri, Aug 14, 2020 at 10:30:14AM -0400, Michael S. Tsirkin wrote:
> On Thu, Aug 13, 2020 at 07:31:39PM -0700, Sean Christopherson wrote:
> > > @@ -2318,6 +2338,11 @@ static int __kvm_read_guest_page(struct kvm_memory_slot *slot, gfn_t gfn,
> > >  	int r;
> > >  	unsigned long addr;
> > >  
> > > +	if (unlikely(slot && (slot->flags & KVM_MEM_PCI_HOLE))) {
> > > +		memset(data, 0xff, len);
> > > +		return 0;
> > > +	}
> > 
> > This feels wrong, shouldn't we be treating PCI_HOLE as MMIO?  Given that
> > this is performance oriented, I would think we'd want to leverage the
> > GPA from the VMCS instead of doing a full translation.
> > 
> > That brings up a potential alternative to adding a memslot flag.  What if
> > we instead add a KVM_MMIO_BUS device similar to coalesced MMIO?  I think
> > it'd be about the same amount of KVM code, and it would provide userspace
> > with more flexibility, e.g. I assume it would allow handling even writes
> > wholly within the kernel for certain ranges and/or use cases, and it'd
> > allow stuffing a value other than 0xff (though I have no idea if there is
> > a use case for this).
> 
> I still think down the road the way to go is to map
> valid RO page full of 0xff to avoid exit on read.
> I don't think a KVM_MMIO_BUS device will allow this, will it?

No, it would not, but adding KVM_MEM_PCI_HOLE doesn't get us any closer to
solving that problem either.

What if we add a flag to allow routing all GFNs in a memslot to a single
HVA?  At a glance, it doesn't seem to heinous.  It would have several of the
same touchpoints as this series, e.g. __kvm_set_memory_region() and
kvm_alloc_memslot_metadata().

The functional changes (for x86) would be a few lines in
__gfn_to_hva_memslot() and some new logic in kvm_handle_hva_range().  The
biggest concern is probably the fragility of such an implementation, as KVM
has a habit of open coding operations on memslots.

The new flags could then be paired with KVM_MEM_READONLY to yield the desired
behavior of reading out 0xff for an arbitrary range without requiring copious
memslots and/or host pages.

diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h
index 852fc8274bdd..875243a0ab36 100644
--- a/include/linux/kvm_host.h
+++ b/include/linux/kvm_host.h
@@ -1103,6 +1103,9 @@ __gfn_to_memslot(struct kvm_memslots *slots, gfn_t gfn)
 static inline unsigned long
 __gfn_to_hva_memslot(struct kvm_memory_slot *slot, gfn_t gfn)
 {
+       if (unlikely(slot->flags & KVM_MEM_SINGLE_HVA))
+               return slot->userspace_addr;
+
        return slot->userspace_addr + (gfn - slot->base_gfn) * PAGE_SIZE;
 }