TaintInduce is a project which aims to automate the creation of taint propagation rules for unknown instruction sets.
One Engine To Serve 'em All: Inferring Taint Rules Without Architectural Semantics
Zheng Leong Chua, Yanhao Wang, Teodora Băluță, Prateek Saxena, Zhenkai Liang, Purui Su.
In the Network and Distributed System Security Symposium 2019, San Diego, CA, US, Feb 2019.
We are currently in the process of rewriting the prototype to better serve our goal of providing an online taint service for different architectures. For people who are interested in the implementation used in the paper, feel free to contact us.
- capstone
- keystone
- unicorn
- tqdm
- squirrel-framework
@TODO
taintinduce.py provides the inference interface and is the CLI tool to generate the rule.
Checkout the --help option on how to use the CLI tool.
python -m taintinduce.taintinduce c3 X86
taintinduce_worker.py is the compute process that SquirrelFlowDB uses.
python -m taintinduce.taintinduce_worker localhost 1234
When installing capstone and keystone-engine with pip in a virtual
environment, the shared library files are expected to be in a folder like
~/.virtualenvs/<virtual_env>/lib/python2.7/site-packages/capstone. If not, you
might run into an import error:
ImportError: ERROR: fail to load the dynamic library.
A quick solution is to find where the library is and copy it to the expected
path. For example, find - name libcapstone* inside
~/.virtualenvs/<virtual_env> and copy it to
~/.virtualenvs/<virtual_env>/lib/python2.7/site-packages/capstone.
@TODO