Re: [kvm-devel] [RFC] [VTD][patch 1/3] vt-d support for pci passthrough: kvm-vtd--kernel.patch

[Anthony L. <an...@co...>]

Kay, Allen M wrote:
>> We have to ensure we don't swap KVM guest memory while using hardware 
>> pass-through, but AFAICT, we do not need to make the memory 
>> non-reclaimable  As long as we reprogram the IOMMU with a new, valid, 
>> mapping everything should be fine.  mlock() really gives us the right 
>> semantics.
>>
>> Semantically, a PV API that supports DMA window registration simply 
>> mlock()s the DMA regions on behalf of the guest.  No special logic 
>> should be needed.
>>
>>     
>
> What should be done for unmodified guest where there is no PV driver in
> the guest?  Would a call to mlock() from
> qemu/hw/pci-passthrough.c/add_pci_passthrough_device() a reasonable
> thing to do?
>   

Yup.  The idea is to ensure that the memory is always present, without 
necessarily taking a reference to it.  This allows for memory reclaiming 
which should allow for things like NUMA page migration.  We can't swap 
of course but that doesn't mean reclaimation isn't useful.

Regards,

Anthony Liguori

> Allen
>