[Battlemesh] [FCC] What hardware still works?

[Paul Gardner-Stephen]

Speaking of FPGAs (which I use for mega65.org), I wonder if it would be
interesting to try to get a student project going to reverse-engineer the
FPGA bitstreams. In the first instance it would be nice to be able to work
out our memories are initialised, so that bitstreams could be modified
post-build to change the included memory contents...

Also a little related to this, we are contemplating making an FPGA-based
smart-phone, so that the end user has a whole lot of control over how their
device works.

Paul.

On Thu, Feb 25, 2016 at 1:11 AM, Juliusz Chroboczek <
jch at pps.univ-paris-diderot.fr> wrote:

> > I thought the chips themselves could be built to only cryptographically
> > accept approved firmware?
>
> Yes, they can and some chips are.  What we're expecting, however, is that
> the vendors of cheap routers won't bother -- their goal is merely to get
> FCC certification, so they'll implement the absolute minimum lockdown
> features that they believe will get them FCC certification.
>
> The other hope is that ARM boards are getting cheaper at a vertiginous
> pace, so with a little bit of luck (and a lot of hard work) the community
> should be able to produce a usable design based on an off-the-shelf board
> that is completely open and reasonably priced.  Right now, the main point
> of contention is the lack of either wifi or meiniPCI on most boards -- wifi
> chips need to be connected over USB, which sucks --, and the limitation to
> just one Ethernet port.
>
> If you're interested in learning what mechanisms can be used to lock down
> a fairly powerful SoC, have a read through chapter 32 of the Xilinx Zynq
> manual:
>
>
> http://www.xilinx.com/support/documentation/user_guides/ug585-Zynq-7000-TRM.pdf
>
> Section 32.2.4 describes the "eFuse", which confirms what David has been
> saying.
>
> Note that I'm not picking on Xilinx here (or ARM, for that matter) --
> quite the opposite, Xilinx provide comprehensive hardware documentation
> without registration, let alone an NDA.  Good luck finding similarly
> detailed information about MIPS-based Broadcom chips.
>
> (Except that Xilinx are still not documenting the bitstream format.  Grr.)
>
> -- Juliusz
> _______________________________________________
> Battlemesh mailing list
> Battlemesh at ml.ninux.org
> http://ml.ninux.org/mailman/listinfo/battlemesh
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ml.ninux.org/pipermail/battlemesh/attachments/20160225/da65a525/attachment-0001.html>