Openwall GNU/*/Linux - a small security-enhanced Linux distro for servers
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Fri, 29 Sep 2017 01:03:31 +0200
From: Guido Vranken <guidovranken@...il.com>
To: oss-security@...ts.openwall.com
Subject: Re: The Internet Bug Bounty: Data Processing (hackerone.com)

I found a buffer overflow in one of the projects within 30 minutes,
and there are probably many more issues to be found (as in virtually
any large, unaudited project). What makes this project special
compared to other bug bounties for C libraries (such as the regular
Internet Big Bounty programs) is that they require a full, reliable
exploit.

If they would be willing to be lenient in their qualification of what
constitutes a working exploit, such as exploitation of a binary
without advanced anti-exploit protections such ASLR, I might bother,
otherwise I won't. Enhancing open source projects is a honourable
pursuit indeed and I've done it many times for free, but if I'm going
to hack for money I might as well choose something that is easier or
more profitable or both at the same time. You can fetch $500 for any
old XSS on a web page or a buffer overflow in the clusterfucks that
are the PHP and Python code
(https://hackerone.com/directory?query=ibb%3Ayes&sort=published_at%3Adescending&page=1
-- see the sheer number of submissions to both those programs).

Right after the program was announced, I sent an email to the IBB
asking if exploitation of a non-ASLR configuration of the binary at
hand would be sufficient. Unfortunately, I have not yet received a
reply. The reason they want full exploits is, I think, to cut the
chaff from the grain and solicit bugs that at least have real
potential. A nice middle ground would be paying a percentage (25%?) of
their current bounty offering for raw submissions of bugs that are
generally assumed to constitute a security risk. It will attract a
larger body of researchers for sure, and in the end this will be more
beneficial to the overall security of the internet than under their
current approach.

A Heartbleed-like vulnerability in an image parsing or conversion
library, where an attacker can send a crafted image file resulting in
exposure of unrelated memory, would not be eligible under this
program. Case in point: see Chris Evans' Yahoobleed:
https://scarybeastsecurity.blogspot.nl/2017/05/bleed-more-powerful-dumping-yahoo.html

All in all I think they should reconsider their current program
stipulations, if only to increase their own return-on-investment
(making the internet safer with a limited funding).

Guido

Powered by blists - more mailing lists

Your e-mail address:

Please check out the Open Source Software Security Wiki, which is counterpart to this mailing list.

Confused about mailing lists and their use? Read about mailing lists on Wikipedia and check out these guidelines on proper formatting of your messages.