On Thu, 12 Oct 2017 16:48:26 +0200 Pierre Morel <pmo...@linux.vnet.ibm.com> wrote:
> On 11/10/2017 14:20, Cornelia Huck wrote: > > On Tue, 10 Oct 2017 18:01:53 +0200 > > Pierre Morel <pmo...@linux.vnet.ibm.com> wrote: > > > >> Since in the case we have no AIS in KVM only PCI is impacted, do you > >> think we can fence IRQFD for MSI with > >> > >> kvm_msi_via_irqfd_allowed = false; > >> > >> in kvm_arch_init_irq_routing() ? > > > > Hm, it seems we never set that for s390x... does that imply that we > > never support irqfd for pci anyway? > > ... yes, we never supported irqfd for virtio-pci anyway. > This explains a lot of things to me. > :) :D > > > > >> > >> In fact, if we also verify the AIS availability in the same routine we > >> can avoid the ugly include of vfio/pci.h > > > > If we can get the serialization right (and setting that value to false > > does indeed have the desired effect), that's probably the way to go. > > > > So now, in v2, we will be able to **enable** irqfd if we detect that we > handle AIS in kernel. :) Let's see how that works out :)
