[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] [PATCH v4 3/4] x86/pt: enable binding of GSIs to a PVH Dom0



On 06/02/2017 04:41 AM, Roger Pau Monne wrote:
> On Thu, Jun 01, 2017 at 06:13:22PM -0400, Boris Ostrovsky wrote:
>>> @@ -696,7 +777,7 @@ int hvm_do_IRQ_dpci(struct domain *d, struct pirq *pirq)
>>>      struct hvm_irq_dpci *dpci = domain_get_irq_dpci(d);
>>>      struct hvm_pirq_dpci *pirq_dpci = pirq_dpci(pirq);
>>>  
>>> -    if ( !iommu_enabled || !dpci || !pirq_dpci ||
>>> +    if ( !iommu_enabled || (!is_hardware_domain(d) && !dpci) || !pirq_dpci 
>>> ||
>>>           !(pirq_dpci->flags & HVM_IRQ_DPCI_MAPPED) )
>>>          return 0;
>> We also need to return for !is_hvm_domain(d).
> Hm, I see. I guess this affects a PV Dom0 because it's a hardware
> domain and doesn't have dpci. I will send a new version of this patch
> alone.
>
> In any case, pirq_dpci->flags shouldn't have the HVM_IRQ_DPCI_MAPPED
> flag set for PV guests?

These flags are part of arch_pirq.hvm and may have garbage on PV. At
least that's what I think I was seeing.

-boris

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
https://lists.xen.org/xen-devel

 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.