OT: topic...

[Don Y]

On 11/23/2021 5:11 PM, John Robertson wrote:

For example, one problem I am looking for a solution is trying to read old
single CHIP CPUs such as the Signetics/Philips MAB84XX series of chips. I know
the ROM is embedded in the chip and can\'t normally be gotten at, I\'m thinking
thought that there is a rare document floating around that would show how to
verify the ROM/RAM code by sending some signal in the T (est) pin-Input pin,

Are you sure that\'s not the T(imer) I/O pin?

testable using the JT1 or JNT1 instructions, whatever those are! Not covered in
any documents I\'ve found so far though. I\'ve found these so far which aren\'t bad:

Some old devices allow \"program verify\" to be performed from the pins
(e.g., the venerable 8048, et al.). These will gladly \"dump\" their
contents if asked to do so. (you have to be careful about what bit of
kit you *expect* to be able to do this; some will gladly toast the part
for you!)

Other devices were (are!) designed with the intent to make this \"impossible\".
(AFAICT, all have failed -- but in different ways!).

There are firms that will do this for you -- for a fee, of course.
Or, a precocious university student. Aside from devices that were
designed with the INTENDED ability to dump their contents, most
others require a fair bit of \"hacking\" (i.e., thus, a financial
incentive). Folks who do it \"professionally\", develop a \"book of
tricks\" for different devices, attack vectors, etc.

[I was actively involved in such Red/Blue team efforts years ago.
It\'s an interesting mindset. And, worthwhile skills developed to
help *protect* designs. But, on the whole, trying to copy other
designs is *so* uninteresting when contrasted with creating new...]

It\'s usually a lot easier (for small processors/applications) to just
reimplement the functionality in whatever device you choose. But,
that\'s not cheap, either (the advantage being you now have a design
that YOU own and can augment -- so \"yours\" is better than the original!)