On Tue, Jul 07, 2015 at 04:09:07PM +0200, Paolo Bonzini wrote:
> The guest should not take ages to boot, which is typical of messed-up
> MTRR/PAT. And if you add some actual usage of the attached device, it
> actually is sufficient. Even a simple ping test, or typing through a
> USB keyboard on a
On 07/07/2015 16:06, Joerg Roedel wrote:
> > I do not have any AMD machines that support an IOMMU, so I would like
> > some help testing these patches. Thanks,
>
> What kind of testing do you want? Booting a guest with a device attached
> is probably not sufficient, right?
The guest should not
Hi Paolo,
On Tue, Jul 07, 2015 at 03:45:35PM +0200, Paolo Bonzini wrote:
> I do not have any AMD machines that support an IOMMU, so I would like
> some help testing these patches. Thanks,
What kind of testing do you want? Booting a guest with a device attached
is probably not sufficient, right?
This part of the MTRR patches was dropped by Xiao. Bring SVM on feature
parity with VMX, and then do guest MTRR virtualization for both VMX and SVM.
The IPAT bit of VMX extended page tables is emulated by mangling the guest
PAT value.
I do not have any AMD machines that support an IOMMU, so I wo
