Skip to content

RickdeJager/Dagro-Hacks

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

39 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Dagro-Hacks

This repo contains a PoC that can be used to gain a shell on several Dagro IP Cameras.

TLDR

Our tool can be used to gain a root shell on a number of different IP Cameras. The only thing the user needs to enter is the IP of the camera. Hacking in progress

Installing

The exploit was developed on linux, and will require some minor changes to work on Windows. Our installation instructions assume you are running a fresh install of Ubuntu 18.

We also provide a Dockerfile. Keep in mind that your docker host needs to support the "host" networking driver which is, at time of writing, only available on linux. Skip to Docker for instructions.

You need the following tools to run this exploit:

GCC toolchain

The camera we conducted our tests on was running armv5le. We used the arm-linux-gnueabihf toolchain to build our shellcode. It can be installed on Ubuntu using:

sudo apt install binutils-arm-linux-gnueabihf

Python

We used python 3.7.5+ to build this tool. All requirements are contained in requirements.txt.

The fastest way to setup your environment is to use venv:

  1. python3 -m pip install --user virtualenv
  2. python3 -m venv env
  3. source env/bin/activate
  4. pip install -r requirements.txt

When you want to leave this env, run deactivate.

Running the script

To run the script, navigate to the click2pwn directory, activate the venv and run:

python main.py [target ip]

Typing help will show a full list of supported commands. The most interesting one being shell, which opens a reverse shell on the target host.

Docker

Every modern repository comes with a Dockerfile and we did not want to fall behind. To run our tool using docker:

  1. docker build . -t dagro-hacks
  2. docker run --network=host -it dagro-hacks

FAQ / Troubleshooting

The camera crashes before the port scan completes

Sadly, the model we tested with has a randomized port scan in the 20000-62000 range. This requires a large port scan which can overwhelm the camera. The nmap arguments can be changed in network/port_scan.py.

The port scan is too slow

The port scan might take up to 5 minutes. In our experience, the scan is slightly slower in the docker container. We tried to balance speed against reliability. The port scan can be sped up by editing the nmap arguments.

The webserver port of my camera is not in the scan range

You can manually set the port of the camera by typing:

set tport xxxx

Background information

More information on how our exploit was developed can be found here.

Acknowledgments

This exploit was developed by a team of 3 students, namely:

It was developed as a part of the "Offensive Security" course, taught at Eindhoven University of Technology, by Luca Allodi. We'd like to thank Dr. Allodi and all TA's involved in the course for teaching us about various exploits, attack vectors, and for helping us trouble-shoot issues.

About

Hacking cheap IP cameras for fun and ECTS!

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published