Re: [PATCH 1/9] vfio: Make vfio_(un)register_notifier accept a vfio_device

["Wang, Zhi A" <zhi.a.wang@intel.com>]

On 4/13/22 5:37 PM, Jason Gunthorpe wrote:
> On Wed, Apr 13, 2022 at 06:29:46PM +0200, Christoph Hellwig wrote:
>> On Wed, Apr 13, 2022 at 01:18:14PM -0300, Jason Gunthorpe wrote:
>>> Yeah, I was thinking about that too, but on the other hand I think it
>>> is completely wrong that gvt requires kvm at all. A vfio_device is not
>>> supposed to be tightly linked to KVM - the only exception possibly
>>> being s390..
>>
>> So i915/gvt uses it for:
>>
>>  - poking into the KVM GFN translations
>>  - using the KVM page track notifier
>>
>> No idea how these could be solved in a more generic way.
> 
> TBH I'm not sure how any of this works fully correctly..
> 
> I see this code getting something it calls a GFN and then passing
> them to vfio - which makes no sense. Either a value is a GFN - the
> physical memory address of the VM, or it is an IOVA. VFIO only takes
> in IOVA and kvm only takes in GFN. So these are probably IOVAs really..
> 
Can you let me know the place? So that I can take a look.

> But then, I see this code taking GFNs (which are probably IOVAs?) and
> passing them to the kvm page track notifier? That can't be right, VFIO
> needs to translate the IOVA to a GFN, not assume 1:1...
> 
GFNs are from the guest page table. It takes the GFN from a entry belongs
to a guest page table and request the kvm_page_track to track it, so that
the shadow page table can be updated accordingly.

> It seems the purpose is to shadow a page table, and it is capturing
> user space CPU writes to this page table memory I guess?
> 
Yes.The shadow page will be built according to the guest GPU page table.
When a guest workload is executed in the GPU, the root pointer of the
shadow page table in the shadow GPU context is used. If the host enables
the IOMMU, the pages used by the shadow page table needs to be mapped as
IOVA, and the PFNs in the shadow entries are IOVAs.

> GFN's seems to come from gen8_gtt_get_pfn which seems to be parsing
> some guest page table?
> 
Yes. It's to extract the PFNs from a page table entry.

> Jason
>