Mining Input Structures

Dear students,

Reminder that we will hold our final presentations next Tuesday, 27.08.

The presentations will happen in person from 10:00 to 16:00 at room 3.21 (third floor) of the CISPA C0 Building (at university campus building E9 1).

We're planning to order food from a local place (La Döneria) so we can have lunch together during the meeting (paid by myself).

We look forwards to seeing you all in person and watching your nice presentations.

Cheers,

Rafael and Andreas.

Dear students,

The topics for the final presentations are assigned as follows:

Autogram: Christoph and Arina
Mimid: Tim Schneider and Abdullah
SYGRAM: Julian
Arvada: Tim Scheckenbach
TreeVada: Jaideep
REINAM: Ali
WEIZZ: Kevin
FormatFuzzer: Zheyuan
AIFORE: Florian
ByteRI: Niklas

We were able to give everyone their 1st, 2nd or 3rd choice :)

And the preferred date for the presentations is on Tuesday, 27.08.
10 students have indicated at least some time available on this date. We look forwards to meeting you all in person for the presentations.
For the students who cannot make it on this date, we can schedule a separate presentation slot individually.

You should now prepare a 15-20 minute presentation on your assigned topic, including experiments you designed and conducted.
If you want to discuss how to best perform experiments for your topic, feel free to contact me by email on rafael.dutra@cispa.de

Finally, as you prepare your presentation, you may find it helpful to revisit our talk on "How to give a good research talk" here: https://youtu.be/2YS3YboZ7nw

Cheers,
Rafael and Andreas.

For the next meeting on 04.07, Kevin will present WEIZZ and I'll present FormatFuzzer. You can choose one of the two papers to read and write about in your abstract.

WEIZZ: automatic grey-box fuzzing for structured binary formats
https://arxiv.org/pdf/1911.00621

FormatFuzzer: Effective Fuzzing of Binary File Formats
https://arxiv.org/pdf/2109.11277

 

For the meeting on 11.07, we have Jaideep and Florian scheduled to present AIFORE.

AIFORE: smart fuzzing based on automatic input format reverse engineering
https://hexhive.epfl.ch/publications/files/23SEC2.pdf

 

For the meeting on 18.07, we have Niklas and Zheyuan scheduled to present ByteRI.

Recovering Structure of Input of a Binary Program
https://dl.acm.org/doi/pdf/10.1145/3508398.3511508

Dear students,

The paper for next meeting will be TreeVada, recent work building up on Arvada:

Fast Deterministic Black-box Context-free Grammar Inference
https://arxiv.org/pdf/2308.06163v2.pdf

The programming assignment will be on the evaluation of SYGRAM and Mimid on Tiny-C.

Link to the assignment is here: https://dl.cispa.de/s/wMn44zHMqo3BGJN

Cheers,

Rafael and Andreas.

Dear students,

I've added some feedback comments on your previous abstract and assignment submissions on the CMS page.

The programming assignment for next class can be found here: https://dl.cispa.de/s/XCNsFqWFRaPrSJo

In this assignment, you will use the GrammarMiner and Mimid to mine grammars for the same program and then compare the obtained grammars.

Please follow the instructions on sheet.ipynb

Cheers,

Rafael and Andreas.

Dear students,

Tomorrow there will be no meeting (holiday). For the next meeting on 06.06, we'll discuss a paper on SYGRAM, a technique for mining grammars based on symbolic execution.

Link to the paper: https://dl.cispa.de/s/d5574HwFX57drG3

I'll also post a new programming assignment soon.

Cheers,

Rafael and Andreas.

Dear students,

The next meeting will exceptionally take place on Wednesday 22 of May at 12:00 pm.

Before this meeting, you should read the Mimid paper:

Mining Input Grammars from Dynamic Control Flow
https://publications.cispa.saarland/3101/11/fse2020-mimid.pdf

and submit a small abstract for the paper.

You should also complete the Programming Assignment 1, which will start exploring the GrammarMiner chapter of The Fuzzing Book.

Download the assignment from here: https://dl.cispa.de/s/4YANEZA2ffiAs4j

and follow the exercises on sheet.ipynb

Cheers,

Rafael and Andreas.

Dear students,

Remember to submit your abstract on Autogram and the programming assignment 0 on CMS before our next meeting on Thursday.

The link to the assignment is on the last post on CMS.

And for the 2 students presenting, prepare your presentations.

See you on Thursday!

Cheers,

Rafael and Andreas.

Dear students,

In our seminar today we'll have the talk on "How to give a great research talk", by Andreas Zeller.

Since next Thursday is a holiday, our next seminar meeting will be on the 16th of May.
Before the next meeting, we ask you to complete the following tasks:

1) Read the first paper for the course, describing the Autogram grammar miner:

Mining Input Grammars from Dynamic Taints (Autogram)
https://www.st.cs.uni-saarland.de/models/autogram/ase2016-autogram.pdf

More information on Autogram: https://www.st.cs.uni-saarland.de/models/autogram/
Corresponding chapter from The Fuzzing Book: https://www.fuzzingbook.org/html/GrammarMiner.html

2) Write and submit an abstract on the paper you just read.
The abstract should focus on describing how the technique works and the key points of the paper.
A critical analysis of the papers is always welcome. What are the assumptions behind the proposed technique?
What are its limitations? Strengths? How complex is it to implement and how effective is it?

3) If you are one of the students selected to present the paper, prepare a 5-minutes presentation on the paper to begin the discussion on the next meeting.
Just like with the abstract, your presentation should focus on explaining how the approach works and the key points of the paper.

4) Do the programming assignment number 0. You can download the assignment from this link: https://dl.cispa.de/s/kJgypYokMJQ7spN
The goal of assignment 0 is to get all the infrastructure setup for your future assignments, so you should install Python and Jupyter and be able to import and use Jupyter notebooks from The Fuzzing Book. Submit your assignment as a ZIP file on CMS.

Students attending the course as a proseminar are not required to do programming assignments.

Cheers,

Rafael and Andreas.

Dear all,

Welcome to "Mining Input Structures"! The (pro)seminar is now all set – we have a date (Thursday noon) and a set of participants.

Important: Please register in LSF within the next three weeks (e.g. by May 15) for the (pro)seminar. If you do not register, you cannot get credit points.

If you do not find the seminar in LSF, it should be set up by the exam office in the next few days (Update: on Thursday at the latest).

Looking forward to see you next week on Thursday,

Andreas + Rafael