Re: [RFC PATCH 10/30] vfio/pci: Export vfio dma-buf specific info for importers

[Ackerley Tng]

Xu Yilun <yilun.xu@xxxxxxxxxxxxxxx> writes:

>> > If VFIO can create one
>
> It is an interesting idea to let VFIO create a guest_memfd instance, if
> DMABUF doesn't work. But to achieve this goal, I think we still need to
> refactor guest_memfd to allow for different memory/MMIO resource
> exporters/creators, and different resource importers/users like KVM MMU
> & IOMMUFD.
>

Yup I think both ways we'd need some level of integration from both
sides :)

>>
>> At guest_memfd creation time, guest_memfd is always created for a
>> KVM. This is the first place where it is bound to a kvm instance. This
>> is one place where it helps Yilun with Confidential VMs, so KVM can be
>> sure that the memory was meant for a specific CoCo kvm and not any VM.
>>
>> I think it might be weird to have VFIO take a kvm fd to create a
>> guest_memfd?
>
> I think having VFIO create a guest_memfd is better than having KVM
> create a guest_memfd on behalf of VFIO.
>
> The lifecycle of the MMIO resources should be managed by VFIO, cause
> the availability of these MMIO resouces are impacted by device
> operations, e.g., device reset, PCI Memory Space Enable (MSE) ...
> You can't export MMIO resources to KVM MMU without VFIO's awareness.
>

Thanks for explaining the above!

Do you mean that as long as VFIO knows that some addresses are not
usable, those addressses have to be removed from the stage 2 page tables
that KVM manages?

> And VFIO doesn't have to take a kvm fd paramter to create a guest_memfd,
> VFIO has got the kvm instance via KVM_DEV_VFIO_FILE_ADD.
>

I'm new to this - looks to me KVM_DEV_VFIO_FILE_ADD is for linking VFIO
to KVM but some mmap-ing is also required (pre-CoCo) to set up memslots?

With CoCo, setting up memslots would use some fd (guest_memfd wrapping
VFIO or the VFIO-created guest_memfd like you were suggesting) and I'd
think there will still be some notifiers between KVM and VFIO for the
mapping.

Btw is it right to say that KVM_DEV_VFIO_FILE_ADD is only in the case of
setting up irqchip? Is irqchip a requirement for Confidential IO?


Hope you also had a chance to look at [1] that David Woodhouse is
working on :)

[1] https://lore.kernel.org/all/1dc299af6b795a40c8887b3d84f915024f6446a9.camel@xxxxxxxxxxxxx/

>>
>> I was thinking that the userspace VMM could do something like:
>>
>>   int vfio_fd = open("/dev/vfio/devices/vfio0");
>>   int gmem_fd = ioctl(KVM_CREATE_GUEST_MEMFD, { .fd = vfio_fd });