Re: Making C compiler generate obfuscated code

Martin Ward <martin@gkc.org.uk> wrote:
+---------------
| A major theoretical result in this area is the paper "On the
| (Im)possibility of Obfuscating Programs" by Boaz Barak et al,
| published in: CRYPTO '01 Proceedings of the 21st Annual International
| Cryptology Conference on Advances in Cryptology. Boaz Barak gives a
| non-technical description of the results and their meaning here:
|
| http://www.math.ias.edu/~boaz/Papers/obf_informal.html
+---------------


See also more recent results mentioned here:


        http://www.wisdom.weizmann.ac.il/~oded/p_obfuscate.html


In particular, there's a 2010 revised version of the CRYPTO'01
full paper here (54 pages, 0.5 MB PDF):


        http://www.wisdom.weizmann.ac.il/~oded/PS/obf4.pdf


Some snippets from the abstract:


        Our main result is that, even under very weak formalizations
        of the above intuition, obfuscation is impossible.
        ...
        We extend our impossibility result in a number of ways,
        including even obfuscators that (a) are not necessarily
        computable in polynomial time, (b) only approximately
        preserve the functionality, and (c) only need to work
        for very restricted models of computation (TC0). We also
        rule out several potential applications of obfuscators,
        by constructing "unobfuscatable" signature schemes,
        encryption schemes, and pseudorandom function families.


-Rob


-----
Rob Warnock <rpw3@rpw3.org>
627 26th Avenue <URL:http://rpw3.org/>
San Mateo, CA 94403 (650)572-2607

Return to the comp.compilers page.
Search the comp.compilers archives again.