Fuzzing 101(github.com)
github.com
Fuzzing 101
https://github.com/antonio-morales/Fuzzing101
17 comments
Google turns up a CNET article from 2007 (probably because eEye was "pumping press releases left and right"):[1]
> Researchers at eEye used a standard process of code auditing in discovering the vulnerabilities, [eEye CEO Ross] Brown added. He noted that Microsoft either did not do a 'good job' with its code auditing, or it may not have had enough people working on such a task.
I don't really get this culture of racing to find a bug in another company's product, then strutting about finding one (in Microsoft Publisher of all things) and throwing shade. I guess we should all be so lucky to have a company whose "standard process" is to pull a week of all nighters testing our product.
[1] https://www.cnet.com/news/privacy/flaw-found-in-office-2007/
> Researchers at eEye used a standard process of code auditing in discovering the vulnerabilities, [eEye CEO Ross] Brown added. He noted that Microsoft either did not do a 'good job' with its code auditing, or it may not have had enough people working on such a task.
I don't really get this culture of racing to find a bug in another company's product, then strutting about finding one (in Microsoft Publisher of all things) and throwing shade. I guess we should all be so lucky to have a company whose "standard process" is to pull a week of all nighters testing our product.
[1] https://www.cnet.com/news/privacy/flaw-found-in-office-2007/
The style of writing certainly added a lot to it.
Edit: I just checked the author, I might actually know him from IRC. The "Mantis" and "infosec" checks out.
Edit: I just checked the author, I might actually know him from IRC. The "Mantis" and "infosec" checks out.
small world we live in...
Fantastic read. Funny, relatable, all the technical details, and so much heart. Thank you for sharing it!
The interesting thing to me is the stark difference between this and golang's approach.
With golang, you can run fuzzing as simply as you run tests, which means that it's trivial to target specific parts of your application or library. It obsoletes so much of those techniques.
I'm quite curious of techniques to guide more the fuzzing. It seems like the best you can do is provide a seed corpus and hope for the best.
With golang, you can run fuzzing as simply as you run tests, which means that it's trivial to target specific parts of your application or library. It obsoletes so much of those techniques.
I'm quite curious of techniques to guide more the fuzzing. It seems like the best you can do is provide a seed corpus and hope for the best.
some fuzzing tools (libFuzzer for example) leverage LLVM's intermediate representation to provide code-coverage metrics that they feed back into their fuzzing algorithms, increasing test coverage
Golang does that natively ;-)
LibFuzzer is packaged with clang, so there is no additional installation [0]. You just have to provide an entry function and link it with a command-line flag. However, since C and C++ lack reflection you have to work with raw bytes as input.
LibFuzzer has the option to provide callbacks that customize mutation, which can help with obtaining coverage.
[0] https://llvm.org/docs/LibFuzzer.html
LibFuzzer has the option to provide callbacks that customize mutation, which can help with obtaining coverage.
[0] https://llvm.org/docs/LibFuzzer.html
I proposed using reinforcement learning to guide coverage as a potential phd topic, but didn't really go down that path, no idea if it could work
Did you try making small changes to your phd proposal to see if it opened up new paths?
</fuzzingjoke>
</fuzzingjoke>
I think it would go the other way where you use coverage to guide reinforcement. Crank the temperature up to increase variation and you would probably produce a model that could approximate the file format you were targeting.
Please tell us more!
Fuzzing is often a special case of genetic algorithms, so there is already a tiny connection to RL. I'm curious to hear what your proposal was.
Fuzzing is often a special case of genetic algorithms, so there is already a tiny connection to RL. I'm curious to hear what your proposal was.
> Fuzzing is often a special case of genetic algorithms
Yes, that was sort of why I thought RL guided fuzzing could work, and possibly better. Also, for things like XSS fuzzing (which I have a little experience in), it is possible for an experienced attacker to intelligibly guide the fuzzing to a payload, which theoretically could be mimicked through RL.
There wasn't really anything novel in the proposal, it was just for a graduate cyber-security course, and one of the deliverables was a project proposal for something related. There were already some existing works that time (around 2-3 years ago) where people tried combining RL with fuzzing, and I just mish-mashed some ideas together so I could hand in something.
My main concern at the time however was that with fuzzing the positive signal would be so rare compared to the negative signal, since most randomly fuzzed inputs would just return the same negative feedback. I wasn't sure that would be enough signal to train an RL system. I'm not quite sure what new progress has been made in the field since then.
Yes, that was sort of why I thought RL guided fuzzing could work, and possibly better. Also, for things like XSS fuzzing (which I have a little experience in), it is possible for an experienced attacker to intelligibly guide the fuzzing to a payload, which theoretically could be mimicked through RL.
There wasn't really anything novel in the proposal, it was just for a graduate cyber-security course, and one of the deliverables was a project proposal for something related. There were already some existing works that time (around 2-3 years ago) where people tried combining RL with fuzzing, and I just mish-mashed some ideas together so I could hand in something.
My main concern at the time however was that with fuzzing the positive signal would be so rare compared to the negative signal, since most randomly fuzzed inputs would just return the same negative feedback. I wasn't sure that would be enough signal to train an RL system. I'm not quite sure what new progress has been made in the field since then.
I am surprised Heartbleed is not on the list, it’s very easy to duplicate.
The tutorials here all use local files or local data rather than network-based fuzzing, so I think that's why it doesn't appear.
Thank you for sharing this - very useful!!!
great read if you wanna waste 11 minutes