From: alberto on

I'm not that sure either, but accessing MSRs can be done with
processor instructions, which hopefully should be handled by the
Virtualizer if not directly by Intel VT. It's worth a try! When I have
a few cycles available, I'll see if I can get it to work.

Alberto.


On Jan 27, 5:11 pm, Kalle Olavi Niemitalo <k...(a)iki.fi> wrote:
> alberto <amore...(a)ieee.org> writes:
> > Or, even simpler, maybe write a tiny Linux driver and run it under
> > Hyper-V, VMWare or VirtualBox if you really want it to work in a
> > Windows environment.
>
> Linux 2.6 has such a driver already; see arch/x86/kernel/msr.c.
> It provides both read and write access.
> I don't know though whether those virtualizers pass the MSR
> accesses through to the actual CPU.

From: Scott Noone on
I guess this is going to rely heavily on which MSR you're reading. For
example, returning the host's value of SYSENTER_EIP_MSR/ SYSENTER_ESP_MSR
would probably be a really bad thing.

Why does the OP need this in the first place?

-scott

--
Scott Noone
Consulting Associate
OSR Open Systems Resources, Inc.
http://www.osronline.com


"alberto" <amoreira(a)ieee.org> wrote in message
news:6e620c64-5d3c-4646-bdfb-aba6eed4d3a3(a)o3g2000vbo.googlegroups.com...
>
> I'm not that sure either, but accessing MSRs can be done with
> processor instructions, which hopefully should be handled by the
> Virtualizer if not directly by Intel VT. It's worth a try! When I have
> a few cycles available, I'll see if I can get it to work.
>
> Alberto.
>
>
> On Jan 27, 5:11 pm, Kalle Olavi Niemitalo <k...(a)iki.fi> wrote:
>> alberto <amore...(a)ieee.org> writes:
>> > Or, even simpler, maybe write a tiny Linux driver and run it under
>> > Hyper-V, VMWare or VirtualBox if you really want it to work in a
>> > Windows environment.
>>
>> Linux 2.6 has such a driver already; see arch/x86/kernel/msr.c.
>> It provides both read and write access.
>> I don't know though whether those virtualizers pass the MSR
>> accesses through to the actual CPU.
>