Re: [PATCH 1/4] KVM: count number of assigned devices

2015-07-08 Thread Alex Williamson
On Wed, 2015-07-08 at 17:18 +0200, Paolo Bonzini wrote: > If there are no assigned devices, the guest PAT are not providing > any useful information and can be overridden to writeback; VMX > always does this because it has the "IPAT" bit in its extended > page table entries, but SVM does not have a

[PATCH 1/4] KVM: count number of assigned devices

2015-07-08 Thread Paolo Bonzini
If there are no assigned devices, the guest PAT are not providing any useful information and can be overridden to writeback; VMX always does this because it has the "IPAT" bit in its extended page table entries, but SVM does not have anything similar. Hook into VFIO and legacy device assignment so

Re: [PATCH 1/4] KVM: count number of assigned devices

2015-07-07 Thread Paolo Bonzini
On 07/07/2015 17:22, Alex Williamson wrote: >> > -#define __KVM_HAVE_ARCH_NONCOHERENT_DMA >> > +#define __KVM_HAVE_ARCH_VFIO_HOOKS > Do we really want to tie these two things together under something > that's not strictly a "vfio" option? Legacy assignment also makes use > of these, as shown in

Re: [PATCH 1/4] KVM: count number of assigned devices

2015-07-07 Thread Alex Williamson
On Tue, 2015-07-07 at 15:45 +0200, Paolo Bonzini wrote: > If there are no assigned devices, the guest PAT are not providing > any useful information and can be overridden to writeback; VMX > always does this because it has the "IPAT" bit in its extended > page table entries, but SVM does not have a

[PATCH 1/4] KVM: count number of assigned devices

2015-07-07 Thread Paolo Bonzini
If there are no assigned devices, the guest PAT are not providing any useful information and can be overridden to writeback; VMX always does this because it has the "IPAT" bit in its extended page table entries, but SVM does not have anything similar. Hook into VFIO and legacy device assignment so