Questions about checking free hardware

Paul van der Vlis paul at vandervlis.nl
Wed Sep 16 10:51:39 UTC 2015


Hello Alessandro,

Op 16-09-15 om 12:20 schreef Alessandro Rubini:
>> When somebody would make a free processor,
> 
> There are free processors, today. 

You are right.

> They are written in HDL.
>
>> would it be possible to do
>> that in such a way that it's possible to check that the hardware is the
>> working-out of the drawings?
> 
> There are no drawings involved, as far as I know. Any drawing is an
> intermediate object file, thus not important in the overall design
> flow.

What I mean are the "sources". I see often the term "Verilog".
Maybe HDL is the correct name.

>> In software you can build your own binary and check the shasum, but
>> in hardware this is not possible.
> 
> Also in software it is not possible, most of the time.  The build
> often contains a timestamp, so your build won't match what you
> received.  And if it is programmed in some media, likely you can't
> even checksum the binary you receive (think embedded systems).

You are right, but there are solutions, see e.g.:
https://wiki.debian.org/ReproducibleBuilds

> Anyways, I don't see the point of your question: the same applies to
> physical object (3d-printed or otherwise manufactured): you get a free
> design and an object that is said to be the same. You just don't know.

My question is in other words, maybe still not perfect: are there ways
to check if a physical object is the working-out of a HDL?

> Actually, I'm interested in discussion about this case; the CPU one is
> too theoretical to be interesting at this point in time.

Why?

> Anyways, I take this chance to advertise the Open Risc Conference,
> where a number of free processor are presented. 

Interesting. Can you give links or other information?

> It's held at CERN,
> Switzerland, on Oct 10th and 11th.  See orconf.org for details.

With regards,
Paul van der Vlis.



-- 
Paul van der Vlis Linux systeembeheer Groningen
https://www.vandervlis.nl/



More information about the Discussion mailing list