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

RE: [Xen-devel] Windows SMP

> On 29/12/2008 08:47, "James Harper" <james.harper@xxxxxxxxxxxxxxxx>
> >> The WHQL tests are oblivious to it. It's just a patching of mmio
> >> writes to the APIC TPR register.
> >
> > Looking at the way KVM does it, it appears to detect writes to the
> > register when they are trapped, and then give the DomU (or whatever
> > calls it) the address of the instruction so that the DomU driver can
> > then patch it. Is that what Citrix is doing? Does the current
> > tree have such a mechanism in it?
> The result is the same, but there's no hypervisor component, so none
of it
> is open sourced.


> You could get similar results by putting static
> fixup tables in your drivers,

As in 'write bytes to offset x into kernel function y', with x depending
on the exact kernel build? Wouldn't the rootkit detectors complain about

> or I'd be open to having a KVM type of
> interaction between Xen and your GPLPV drivers. Putting the payload in
> generic virtual BIOS seemed kind of gross to me.

Is it possible for a virtualised DomU to trap the MMIO write itself, or
can it only be trapped by the hypervisor?

Btw, is it the vmexit that is slow about these TPR writes, or is it the
writes themselves?



Xen-devel mailing list



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