Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Reverse Engineering an Integrated Circuit for Pwn2Win 2017 CTF (blog.dragonsector.pl)
67 points by q3k on Oct 26, 2017 | hide | past | favorite | 9 comments


Reading blog posts like this this is kind of like listening to Usain Bolt talk about how to run fast. Step 1: be better than anyone else. Step 2: work crazy hard.


As the author of this blog post, I'd like to stress that it's much, much more of Step 2 than 1. At least for me, can't really speak for Usain Bolt.

Also, don't forget Step 0: have fun doing what you're doing.


For things like this, "have broad-but-shallow knowledge of a field" and "be lazy like Larry Wall" also work.


It might seem that way if you're new to hardware stuff, but it's really not all that complicated. I'm not super knowledgeable myself (took all of 3 undergrad courses in digital logic design a few years ago and haven't worked with it since) but found the post very readable and interesting.


> Step 1: work crazy hard. Step 2: be better than anyone else.

FTFY.


If by fix you mean completely change what was being expressed to the opposite, sure.


Alright then:

> Step 1: work crazy hard. Step 2: be better than anyone else. Step 3: work crazy hard.

FTFY.


I've actually written the intermediate section of that in production code. The DEF netlist format contains a "wire" as a set of rectangles linking an input A to outputs B,C,D...etc, and we wanted to show only those rectanges involved in e.g. the A-D part of a route. The solution was to make a graph and then use standard graph traversal algorithms to select only the needed rectangles.

This reverse engineering challenge is made a lot easier by all the preparatory work of decapping and photographing the layers not being necessary.

If you like this sort of thing, you may like the Visual 6502: http://www.visual6502.org/


This is awesome, nice work!




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: