Monday, 25 July 2022

[re:educate] Basic in reverse engineering executable files

Basic in reverse engineering executable files

This article is a part of our program, #re:educate where we empowering
cybersecurity students and beginners to share their understanding
about anything related to offensive security. For more info, refer to
this link RE:HACK - #re:educate

Author: Wan Muhammad Khairuddin
University: Universiti Teknikal Malaysia Melaka

Hi readers! This is a step by step, in basic, on how to reverse engineer an executable files that is written in C language. Reverse engineering is simply a process of understanding how a program works. When we got the end product, we disassemble and poke it around to see the behaviour of the product and then get some idea how the program works in the background. But in our context, we disassemble a binary executable files, run it and do analysis to understand the behaviour of the executable files.

In this walkthrough, we will use some program that will help us to reverse engineer the executable files. Reverse engineering can be overwhelming for beginners but we will try our best to explain and share tips on how you can easily understand the flow of the program just by reading the assembly code (static analysis).

What is Reverse Engineering ?

Reverse engineering is a process or method through which one attempts to understand how an application works when we don’t have the source codes. We understand it by disassembling the application and look at each function’s implementation.

Prerequisite

The most important things that need to be understood before we can start reverse engineer an executable file are:

  • Assembly language
  • Registers

We would not be able to get the source code of an application written in C or C++ because it will be converted into assembly language once compiled. You don’t have to be a master or can write a program by using assembly language to do reverse engineering. The code can be understood simply by knowing a few basic instructions and keywords.

Register is one of a small set of data holding places that are part of the computer processor. A register may hold an instruction, a storage address, or any kind of data (such as a bit sequence or individual characters). Some instructions specify registers as part of the instruction. Register is like a variable that stores data. It is also used to perform mathematical calculations. You can learn about common registers here. We would not explain much about register in this article.

Reverse Engineer a Simple Program

Now we will start the main purpose of this article. We will reverse engineer this file from Crackmes. Let’s get started.

Tools needed (for Windows):

First we need to know what file we are dealing with. I will use file command in git bash to find out the type of the file.

enter image description here

From the command, it says the file is a 32-bit PE executable file. Thus, we know that it is for Windows environment. We would not need a Linux machine in this case.

Next, we can try run the file through our terminal (cmd.exe).

1

What the program does above is, it asks the user to enter a password and then display a message if the entered password is wrong or correct. From this behaviour, we can conclude that our goal is to find out the correct password. So let’s disassemble this file and figure out how the password checking functionality works.

Open our IDA > Choose new > Drag our file into IDA.

2

3

Complete the importing process and then it will look like this.

4

It does look overwhelming, but rest assured, we will walk you through each process to reverse engineer this file and collect the correct password.

First let us look at the top section.

5

You’ll see a lot of instructions here. It is easy to understand what is happening in it by focusing at the most significant things which is the function.

First we look at the call instruction. This instruction calls a function ; printf, gets and strcmp.

printf is when the program is printing something on the screen. In the beginning, we noted that the program displays a welcome message. So now we know our position or which part of the program we are looking at.

6

Move on to the next function is printf again. But this time, it displays the password: before the program asks for the user input.

7

Next is gets function. As we know gets function takes user input so what this assembly instructions do is taking the user input.

8

The instruction above says that it takes the address of szPassword variable and store it inside $eax register. Then after gets function finishes, the user input is stored inside whatever address the $eax holds. In this case $eax holds address of szPassword. So the user input is stored inside szPassword variable.

Last function is strcmp. This function compares two strings and check if these two strings are equal or not.

8

As you see above, before strcmp was called, the instructions push something onto the stack. The first one is szPassword, which stores the user input and then the instructions push str1onto the stack. Then strcmp is called. These two strings, szPassword and str1 are compared by the strcmp function. So basically our input is being compared to str1.

Next, let’s take a look at what happen after our input being compared.

9

There are two paths here; the red path and the green path. If the strcmp return 0 (which means if our input is equal to str1), we will be following the red path. Otherwise, we will be following the green path. Since we are looking for a correct password, we must follow the red path (look at the final message shown in the printf)

10

Further digging the file, we can see there’s a value set at the str1 which is LiL2281337.

11

Let’s try if LiL2281337 is the correct password.

12

Success! It seems we have found the correct password!

Additional notes

Some may ask, how do I know that szPassword and str1 are the strings that are being compared by strcmp? Let me explain in-depth how function calls in assembly.

strcmp is a function that takes two parameters.

int strcmp (const char* str1, const char* str2);

In 32-bit assembly, before a function with parameters being called, the value of the parameters will be pushed onto the stack first. Then when the function is called, it will take whatever value from the top of the stack as its parameter value. Let’s take a look at this instructions again.

13

As you can see there are two push instructions, push ecx pushes our input onto the stack and push offset str1 pushes the correct password onto the stack. Now our input and the correct password are placed at the top of the stack.

14

Then strcmp is called, it will take two items from the top of the stack to fulfill its two empty parameters. That’s is how function with parameters being called in 32-bit assembly.

Similarly to the printf. printf takes string to be displayed as its parameter. So the string must be pushed onto the stack first before printf is called.

15

Conclusion

Reverse engineering is quite complex for a beginner but with the right mindset and methodology, anyone can understand the flow of the program even though they have only a little knowledge about assembly language. The most important thing is to find a suitable methodology and know what to look for.

To anyone who wants to explore reverse engineering, I suggest reading as many reverse engineering writeups, articles, and research papers as possible. We will be able to learn more about reverse engineering as a result of this.

Share:

Wednesday, 20 July 2022

[Tips & Tricks] Configuring BurpSuite to capture requests from a proxy unaware thick client

Configuring BurpSuite to capture requests from a proxy unaware thick client

What is a thick client application

Thick client applications generally installed on a user’s local desktop/laptop/workstation. They sometimes called as Desktop Application.

These applications can run own its own (independently) without need to be connected to the internet. Best examples are desktop chat applications such as Teams, Zoom, Slack, etc.

Types of thick client based on proxy configuration

There are two:

  1. Proxy aware: The type that have an option to configure the proxy settings in it so a user could monitor the outgoing and incoming communications through the proxy server/tool.

  2. Proxy unaware: The type that have no option to configure the proxy settings. To monitor the requests, the user needs to make changes on their own machine’s host file.

How to configure

We had a situation where we required to perform a security assessment against the proxy unaware type thick client. To ensure we were able to collect all the incoming and outgoing requests, we configured our machine as the following:

Initial process

After configuring

Steps

  1. Edit the /etc/hosts file as the following (you may need to be a superuser):
##
# Host Database
#
# localhost is used to configure the loopback interface
# when the system is booting.  Do not change this entry.
##
127.0.0.1       localhost
127.0.0.1       targetapi.io
  1. In BurpSuite, navigate to your Proxy setting and add configure the Proxy Listeners to bind to the actual targetapi.io port (in this case is 443 and set it as “Loopback only”


  1. In the Request Handling, add the actual IP address of the targetapi.io and port 443. Tick on “Force use of TLS” (if it communicates over TLS) and enable invisible proxying. You may need to play around with the invisible proxying sometimes.

That’s all. Now you should be able to capture the thick client’s upcoming and outgoing requests in your BurpSuite.
Worth noting that if you are using port below 1024 in Burp’s proxy setting, you may need to run it as a superuser.

Share:

Wednesday, 6 July 2022

[Disclosure] Bypassing the fixed reflected XSS on👨‍⚕️ website

Bypassing the fixed reflected XSS on👨‍⚕️ website

After we noted about a vulnerability disclosure program from BadApe NFT website (see http://blog.rehack.xyz/2022/06/disclosure-dom-based-xss-on-badape-nft.html), our team decided to poke around and see if there is any other Malaysian-based company that has the similar program on it.

We found one :). While they have a vulnerability disclosure program, further a discussion with them, they agreed for us to publish this article but without mentioning their name. Fair enough.

By simply browsing their website, we noted an obvious Reflected XSS and one DOM-Based XSS.

DOM-Based XSS

The DOM-Based XSS was interesting. The injection point was only detected when the URL included with a page parameter. Without this parameter, the value will not be reflected in the DOM. We used Untrusted-Types Devtools plugin by @filedescriptor to detect this.


This was too straightforward. By inserting the XSS payload, the XSS was successfully executed.

Reflected XSS

The reflected XSS was also found on an obvious location. The page parameter on a different endpoint from the above was insufficiently sanitise the user input, thus it was possible for us to insert our XSS payload such as '"><img src=x onerror=alert(document.domain)>

We notified their security team and they immediately mitigated the issue. Once our team received their reply stating that the fix was done, we casually verified the fix and noted that it can be bypassed.

The initial fix was to allow page parameter to only accepted up to 4 characters. More than that will redirect us to a 404 error page. We were able to bypass that via Parameter Pollution technique, where the second page parameter appended on the request was still accepted and reflected in the response body. Thus, this allowed us to include the XSS payload again and successfully bypassed the initial fix.


Again, our team was happy with the quick response and their commitment to security. They immediately reviewed and corrected the problem. We reviewed and confirmed the fix was correctly implemented this time.

Thank you to their security team, and we hope more companies in Malaysia will accept vulnerability disclosures from outsiders.

Share: