From: "Robert Comer" 

> That's what I was talking about too. Whatever layers there are outside the
> VM are still just as exposed from inside the VM as from the network. I
> think
> there's been enough worms that it should be clear there are still plenty
> of
> opportunities. Any place you communicate outside the VM (disk, com ports,
> network, video, IR, keyboard, cd, bluetooth, fan control, whatever, you
> are
> still exposed to driver hacks and there are enough of these interfaces
> available from inside the VM that aren't available from the network that
> the
> risk should still be greater than being hacked from the network.

I don't think you understand just how isolated those drivers in the VM are
from the actual hardware. (they don't actually touch the hardware at all,
they only touch an emulated device.)  The CPU is a possible attack point as
that's only virtualized, but even it is enough so I haven't seen it crash
the host.  Sure, an errant driver for say the sound card can be crashed by
a VM running, but I don't see an actual attack vector from the VM.

--
Bob Comer


"Geo"  wrote in message news:4557b001{at}w3.nls.net...
> That's what I was talking about too. Whatever layers there are outside the
> VM are still just as exposed from inside the VM as from the network. I
> think
> there's been enough worms that it should be clear there are still plenty
> of
> opportunities. Any place you communicate outside the VM (disk, com ports,
> network, video, IR, keyboard, cd, bluetooth, fan control, whatever, you
> are
> still exposed to driver hacks and there are enough of these interfaces
> available from inside the VM that aren't available from the network that
> the
> risk should still be greater than being hacked from the network.
>
> A VM really only sandboxes application hacks to some degree, but those
> hacks
> then yeild access to these additional interfaces. And I'm not aware of any
> way to firewall these interfaces.
>
> Geo.
>
>
> "Robert Comer"  wrote in message
> news:45572284{at}w3.nls.net...
>> > Why would it be any more difficult than breaking into a PC from the
>> > network?
>>
>> It's identical from that type of attack -- what he's talking about is
>> getting from the VM to the host in some easier way than a LAN attack.
>
>
>

--- BBBS/NT v4.01 Flag-5