Vulnerability Development mailing list archives

Re: Modern hw-killing virus feasible


From: Lucien Fransman <R.L.Fransman () ptt-post nl>
Date: Thu, 8 Mar 2001 09:34:48 +0100

-----BEGIN PGP SIGNED MESSAGE-----
Hash: MD5

Back in the old days at high-school, we had some smart kid who wrote
the first ever "IPL-virus" on the IBM S 32 which would render the
drivepacks useless by crashing the read/write heads into the disks by
altering the distance between the drum and the head :-)
He only did this once to prove a point ( the IBM engineer said it
could not be done)

About the monitor / scanrate problem :
I severely destroyed my own monitor (Comodore 1942 ) by sending wrong
info to the monitor ( changed the frequency rates around ). It gave a
terrible whine, and while I was looking for the booklet that came with
the monitor it just went dark, never to show pictures again :-(

There was a program that would generate errors in certain
intel-processors ( I think it was the P-60 ) that would cause the CPU
to overheat, but it would be noticeable because the performance of the
machine went rappidly down-hill.


On a brighter note: I think that flashing your BIOS is actually a good
place to start by "hardware-damaging viruses/trojans/ whatever
malicious code". I have a "exploit" somewhere that would wipe your
bios by sending clock-ticks to it. The fun thing was that it would
only show up after a boot. Of course it will not work with NT 4.0 and
2k, but with dos...me it would work splendidly.

Also, some mainboards have a monitoring/regulating function that would
allow you to modify the speed of the CPU fan. It would be very generic
( I could only test this with a Asus XP-something_or_other board that
went dead because of a fried CPU) and thus the "real-life tread" would
be near to nothing.

Some tape-robots could be vulnerable to hardware damaging software, as
a former employer of mine noticed when a badly written driver caused
the robot to be damaged beyond repair ( bad movement caused the arm to
crash *trough* a cardridge.

Concluding this laps into the past, I would say that as long as there
are no real standards in how to program devices or how to affect
hardware. Also, the hardware should be available to "normal" people to
have any real treat.


- --
Forgive my poor English  writing-skills as it is early in the morning
:-)

Met Vriendelijke Groet,
 Lucien

-----BEGIN PGP SIGNATURE-----
Version: 2.6

iQCVAwUAOqdEM5PZ7EDWD5HlAQH/JwP/X5R/rCp3Wru5vUG3TfiDBqFWu3z7Fzz9
mP9xqnO7yoGeaalYHPv/kjO08RjE1RG7FFogUdvpSNdLYuzCeHvo1u26toRWterK
3H+6LPCch+2kW073i23O7sCUMPKv/rHtOeYIq5ucs17Hc03opNjc2UQGrFIsdL3i
x7G0cSmM+Jo=
=P7RO
-----END PGP SIGNATURE-----


Current thread: