[Full-disclosure] gnupg diff available

Type securityvulns
Reporter Securityvulns
Modified 2007-01-15T00:00:00



I did a gnupg audit recently. I was, frankly, appalled by the code quality. It is a desert of pointer manipulation, string copying, memcpy and strcpy are used all over the place, and sprintf, too.

You can find my diff at


Please note that

a) I might have missed something b) I don't claim all fixed parts were exploitable. My attention span is quite short. If I couldn't figure out where data is coming from in 30 seconds, I assumed it was user settable and put a fix in. c) I focused on write accesses. I probably missed a few out of bounds read accesses. Crashing seems to be an acceptable error handling mechanism in gnupg (there are already lots of assert() and BUG() calls) so I also used assert() in most cases.

People are always saying how free and open source software is more secure because so many eyes look at it... well, I didn't see much evidence of that in gnupg. Please do use some of your time to actually look at source code and find bugs.

I tried to give Werner Koch (the author) advance warning, but he was neither helpful nor did he appear interested. So please don't make 0-days out of this.

Thank you,


Full-Disclosure - We believe in it. Charter: http://lists.grok.org.uk/full-disclosure-charter.html Hosted and sponsored by Secunia - http://secunia.com/