Note: you need to clone this repo using the --recursive flag since this repo has submodules, e.g.,
git clone git@github.com:nevillegrech/MadMax.git --recursive
Madmax consists of a series of analyses and queries that find gas-focussed vulnerabilities in Ethereum smart contracts. The analyses are performed on the Gigahose IR, which is lifted from Ethereum bytecode. The first version of MadMax used Vandal.
First follow the instructions in gigahorse-toolchain for instructions on installation of Gigahorse. In a nutshell, this requires the installation of the Souffle Datalog engine, custom functors and Boost.
In order to run MadMax using Gigahorse, you can use the following incantation:
gigahorse-toolchain/gigahorse.py -C madmax.dl <contract.hex>
Where <contract.hex> is a compiled Ethereum contract, or a directory of contracts. If you're running this for the first time it will take longer due to compilation of Datalog files. The output of the analysis results can be found under .temp/**/out/*.csv and results.json. A summary is also printed to the screen.
To see whether an individual contract is flagged or not if, check whether there are any entries inside the WalletGriefing, UnboundedMassOp and OverflowLoopIterator relations.
MadMax is now deployed as a client for the Gigahorse framework. One can see the latest version in action on contract-library.com. If you would like to test your own contract please deploy it on an Ethereum test network (e.g. Ropsten) and then view the results of the analysis at contract-library.
For a list of contracts flagged by MadMax on the entire Ethereum chain (updated in realtime to reflect all deployed contracts), please visit the following pages, for each vulnerability type, respectively:
MadMax: surviving out-of-gas conditions in Ethereum smart contracts Neville Grech, Michael Kong, Anton Jurisevic, Lexi Brent, Bernhard Scholz, and Yannis Smaragdakis Proceedings of the ACM in Programming Languages (OOPSLA) 2018 PDF
🏆 Distinguished Paper 🏆
MadMax: Analyzing the Out-of-Gas World of Smart Contracts Neville Grech, Michael Kong, Anton Jurisevic, Lexi Brent, Bernhard Scholz, and Yannis Smaragdakis Communications of the ACM 2020 PDF
🏆 CACM research highlight 🏆
