OT: Re: Security question (sort of)/goodbye - test-drb@ccmp.vtda.org

20 Mar 1999

A few posts back I remember this young man relying on MSD to figure out what
he had.  I personally have always found MSD to be worse than useless.  If
he's confused, that's why.

As for the serial ports, the UART or maybe ART (also a valid acronymn, I
guess, since they're hardly universal any more) these are normally
recognized by means of the interrupt they return.  If, let's assume, the
plug-N-play was altered by enabling the control parameter in the CMOS ram,
which allows this, then it might physically assign a different interrupt
than previously.  Possibly a new hardware device was installed.  It then
causes the firmware to reallocate resources, but perhaps not modify the
tables, hence fail to recognize devices because it was not told to update
the table.  Now it doesn't get the correct interrupt from the serial ports,
hence concludes the port is absent.

Does that sound possible?

Dick
-----Original Message-----
From: Tony Duell &lt;ard(a)p850ug1.demon.co.uk&gt;
To: Discussion re-collecting of classic computers
&lt;classiccmp(a)u.washington.edu&gt;
Date: Saturday, March 20, 1999 6:29 PM
Subject: Re: OT: Re: Security question (sort of)/goodbye

...
 > As for those of you addressing a BIOS problem, you
are wrong.  First of all,
...
   I could no
longer enter the CMOS setup- the computer would freeze. 
But isn't this one heck of a clue? Damage to serial ports can't affect
configuring the CMOS RAM AFAIK. But damage to the bios ROM itself could.
As could (in theory) misprogramming an address decoder....

> Secondly, I bought a new AMIBOIS chip at the computer show today, and
> installed it in the motherboard.  Booted, reconfigured, re-ran Windows
> setup. No serial ports detected (yes, I am smart enough to enable them in
> the CMOS setup).  Windows says that the secondary controller of the dual IDE
...
 > controller is "not present or not working
properly" - yes, it is enabled in
...
   the CMOS.

Again, this looks like an address conflict or similar.

Are the UARTs on the motherboard or an expansion card? Are they part of
some large ASIC, or do you have a chance to probe the chip select pin?

That's what I would do if possible, btw. Trigger a logic analyser off the
CS/ pin and see what's on the address bus at the time.

-tony