FPGA reverse-engineering challenge @ Hackito Ergo Sum


Hackito Ergo Sum – Paris, April 8-10 2010
http://www.hackitoergosum.org

SRAM-based FPGAs are often touted as being “secure”, in the sense that a design for which only the programming file (“bitstream”) is available cannot be analyzed [1]. Many security features are built on this assumption: anti-cloning protections for hardware, evaluation versions of FPGA designs (IP cores) that stop working after a few hours, …

This security relies on the fact that the bitstream format, despite being unencrypted, is largely undocumented and proprietary and even assuming perfect understanding of the format, recovering a design’s netlist from the information contained in the bitstream solely is a daunting task.

Many security researchers believe that security through obscurity does not work. Could FPGAs prove them wrong? Could obscurity be, for once, an effective means to achieve security? In this case, what would prevent an attacker with knowledge gained from the FPGA manufacturer from injecting malicious code into your sensitive bitstreams without fear of being discovered?

In a world where concerns about hardware security are growing (two years ago, the DARPA launched a challenge where people were supposed to discover malicious insertions made into chips [2]), we will challenge these questions at Hackito Ergo Sum.

You are given bitstreams implementing security features, and your role is to break them. Even though they are significantly simpler than the security systems commonly found in the industry, the basic technique should be essentially the same especially for the last levels of difficulty. The very last one is close to what a real system would look like!

In the first levels, you have more information than the bitstream only, in order to ease your job and give you an insight about what needs to be done in the next levels. Furthermore, there will be a workshop at the beginning of the conference to get you acquainted with some FPGA tools.

THE WORKSHOP
A hands-on workshop will be given at the beginning of HES and shortly after the challenge begins. It will revolve around three subjects:
1/ logic design crash course.
2/ how an FPGA works internally.
3/ how to build a FPGA system-on-chip and how to connect a new peripheral (such as the security device) to an on-chip bus.
This workshop is merely a digest of the previous workshops at /tmp/lab [4] [5].

LEVELS
There are 6 levels of increasing difficulty. They will be announced at the same time as the files for the challenge are posted.

RULES
* Complete files for the challenge will be posted to my blog [9] on April 8th, 10:00 local time.
* Mail the answer(s) to the level(s) you have solved to seb AT tmplab DOT org before April 10th, 18:00 local time.
* You can work alone or in teams.
* You have to explain how you found the answer(s).
* Breaking into my computer is cheating and gets you disqualified. Don’t do it.
* If you do not want to waste time fixing libraries, installing virtual machines, downloading multi-GB files, etc., install Xilinx ISE Webpack before coming and make sure that it works (it rarely does the first time), including the fpga_editor command.
* All bitstreams can be tested on the cheap Avnet Spartan 3A Evaluation Kit [3], making the challenge accessible to everybody. A comprehensive list of commands to perform various operations on this kit is given in [7].
* Some kits are available for lending during the conference, in exchange for a deposit of 70 euros which will be given back to you when you return the kit in good working condition.

REFERENCES
[1] http://www.cl.cam.ac.uk/~sd410/papers/fpga_security.pdf
[2] http://spectrum.ieee.org/semiconductors/design/the-hunt-for-the-kill-switch
[3] http://www.xilinx.com/products/devkits/aes_sp3a_eval400_avnet.htm
[4] http://lekernel.net/blog/?p=668
[5] http://lekernel.net/blog/?p=429
[6] http://www.milkymist.org
[7] http://www.milkymist.org/wiki/index.php?title=Installing_the_Spartan_3A_evaluation_kit_mini-port
[8] http://www.ulogic.org
[9] http://lekernel.net/blog

  1. #1 by DanL on April 15, 2011 - 5:06 am

    What ever happened to this challenge? Any attempts, winners?

Comments are closed.