Re: [PATCH v3 2/8] KVM: add PCI MSI hypercall for VFIO passthrough

[Jose Ignacio Tornos Martinez <jtornosm@redhat.com>]

> This is not quite what I suggested, this still relies on the host
> setting up the msi using the trapped MSI emulation, which is the
> primary hack..
>
> Was it too hard to make this an 'allocate msi' operation?

I tried this approach as a simpler mechanism to just read the necessary
info from the host side. Ok, I will explore a true allocate operation
for the next version.

> Yet here it is called alloc?

> And this isn't alloc, it is some query..

You are right, the naming was misleading. The implementation was
a query, not an allocation.

> I feel like VFIO shouldn't be involved here, for the purpose of MSI
> all we need is to inform KVM what struct pci_device * is at a vBDF and
> then KVM can do the interrupt bits. Maybe this is already happening?

> Nothing should use vendor/device ID either, it should all be drive on
> vBDF.

I appreciate the hint. I will explore the vBDF-based approach
without VFIO in the MSI path.

> I probably would not have selected this, it is why I suggested
> virtio-msi or something that could be optional and seperate from the
> core KVM guest ABI.
>
> But maybe there is interest in doing it this way?

Since there was no feedback from KVM maintainers, I will try to
follow your guidance more closely for the next version.

Thank you for your help

Best regards
Jose Ignacio