"PNutts" <> wrote in message
news:0EE7624C-7E98-45CD-96C5-...
> Agreed. From the eWeek article dated 8/4/2006:
>
> "We've already fixed that path [of attack] … It's beta software that will
> have bugs. That [attack scenario] has already been fixed in later builds,"
> Fathi said.
>
Actually out of the links that one is by far the scariest. The exploit has
wide ranging implications for all OS' not just Vista. It uses a flaw in
hardware virtualization that could allow a program to have complete control
of a computer in ways that would be almost impossible to detect. This
doesn't just affect Vista but any OS that runs on the newer AMD and Intel
CPUs with hardware virtualization. Any exploit for any OS could be used to
install a program that takes advantage of this hardware exploit. Once
installed the only way to get rid of it would be to remove the hard drive
and do it from another computer.
The actual flaw or exploit for Vista had two parts. The main one about
unsigned drivers was patched long before the RTM and only applies to Vista
x64. That vector of attack has always existed if unsigned drivers are
allowed so in a sense Vista x86 is vulnerable It's not an exploit per se but
a design decision to allow unsigned drivers. The exploit got around the
unsigned driver blocking in Vista X64. The other part of her exploit was
that it was easy to bypass UAC (which stopped the exploit by the way) by
just clicking on OK. That's a specious argument as that is the way UAC is
supposed to work. It gives you a choice to allow or disallow something from
happening. If social engineering can be used to persuade you click to OK
then that's a human exploit not a Vista exploit. The good news about the
whole thing is that UAC worked. If you didn't allow the program to continue
at the first UAC prompt then it was stopped dead.
--
Kerry Brown
Microsoft MVP - Shell/User
http://www.vistahelp.ca