Palladium and classic computing

From: Patrick Finnegan <pat_at_purdueriots.com>
Date: Thu Jun 27 19:15:24 2002

On 27 Jun 2002, Frank McConnell wrote:

> "Bill Sudbrink" <wh.sudbrink_at_verizon.net> wrote:
> > Basically, the ID in the PIII is disabled by all standard BIOSs... AWARD,
> > etc. (although it can be enabled by software) and, in any Pentium 1.5GHz or
> > faster, there is no ID or supporting circuits in the silicon.
>
> It's this "it can be enabled by software" bit that I was getting at.
> As I remember it, there was a great hue and cry from privacy advocates,
> Intel said "look, we'll make it disabled by default, and give people
> software they can use to turn it on", and the privacy advocates said
> "well, then, that's OK" and got quiet. Hello? If I can run some
> software to enable ID reporting, why wouldn't any closed-source software
> provider who really wants to know just have their software enable ID
> reporting, get the ID, then restore/disable ID reporting to cover its
> tracks?

Actually, the processor starts with the ID enabled on reset, and then has
to be disabled in software. Once disabled, you have to reset the
processor to re-enable it. This means that to enable the bit, you have to
be writing software that runs in ring 0 if you're using protected mode,
somthing easy for the OS, but hard for applications to do, unless it's
running under Winderz 3.1/95/98/ME.

Still, I don't like what's possible. It made the decision to get an AMD
Athlon all that much easier.

-- Pat
Received on Thu Jun 27 2002 - 19:15:24 BST