Re: [PATCH v6 4/5] [FUTURE] xen/arm: enable vPCI for domUs

[Jan Beulich <jbeulich@xxxxxxxx>]

On 02.12.2023 03:56, Stefano Stabellini wrote:
> On Fri, 1 Dec 2023, Roger Pau Monné wrote:
>> On Mon, Nov 13, 2023 at 05:21:13PM -0500, Stewart Hildebrand wrote:
>>> @@ -1618,6 +1630,14 @@ int iommu_do_pci_domctl(
>>>          bus = PCI_BUS(machine_sbdf);
>>>          devfn = PCI_DEVFN(machine_sbdf);
>>>  
>>> +        if ( needs_vpci(d) && !has_vpci(d) )
>>> +        {
>>> +            printk(XENLOG_G_WARNING "Cannot assign %pp to %pd: vPCI 
>>> support not enabled\n",
>>> +                   &PCI_SBDF(seg, bus, devfn), d);
>>> +            ret = -EPERM;
>>> +            break;
>>
>> I think this is likely too restrictive going forward.  The current
>> approach is indeed to enable vPCI on a per-domain basis because that's
>> how PVH dom0 uses it, due to being unable to use ioreq servers.
>>
>> If we start to expose vPCI suport to guests the interface should be on
>> a per-device basis, so that vPCI could be enabled for some devices,
>> while others could still be handled by ioreq servers.
>>
>> We might want to add a new flag to xen_domctl_assign_device (used by
>> XEN_DOMCTL_assign_device) in order to signal whether the device will
>> use vPCI.
> 
> Actually I don't think this is a good idea. I am all for flexibility but
> supporting multiple different configurations comes at an extra cost for
> both maintainers and contributors. I think we should try to reduce the
> amount of configurations we support rather than increasing them
> (especially on x86 where we have PV, PVH, HVM).
> 
> I don't think we should enable IOREQ servers to handle PCI passthrough
> for PVH guests and/or guests with vPCI. If the domain has vPCI, PCI
> Passthrough can be handled by vPCI just fine. I think this should be a
> good anti-feature to have (a goal to explicitly not add this feature) to
> reduce complexity. Unless you see a specific usecase to add support for
> it?

I could see very special devices to want accompanying by a special ioreq
server. I could also see tandem pass-through/emulated device pairs to
potentially be of use and require coordination in the same ioreq server.

That said, I wouldn't insist on allowing a mix of vPCI and ioreq servers
to be used until an actual use case (with code supporting it) actually
appears.

Jan