neu-se / CONFETTI

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

CONFETTI: CONcolic Fuzzer Employing Taint Tracking Information

CONFETTI is a fuzzer that hybridizes the coverage-guided fuzzing guidance algorithm, Zest, with customized mutators based on dynamic taint tracking and concolic execution information.

Overview

Fuzz testing (fuzzing) allows developers to detect bugs and vulnerabilities in code by automatically generating defect-revealing inputs. Most fuzzers operate by generating inputs for applications and mutating the bytes of those inputs, guiding the fuzzing process with branch coverage feedback via instrumentation. Whitebox guidance (e.g., taint tracking or concolic execution) is sometimes integrated with coverage-guided fuzzing to help cover tricky-to-reach branches that are guarded by complex conditions (so-called "magic values"). This integration typically takes the form of a targeted input mutation, for example placing particular byte values at a specific offset of some input in order to cover a branch. However, these dynamic analysis techniques are not perfect in practice, which can result in the loss of important relationships between input bytes and branch predicates, thus reducing the effective power of the technique.

CONFETTI introduces a new, surprisingly simple, but effective technique, global hinting, which allows the fuzzer to insert these interesting bytes not only at a targeted position, but in any position of any input. We implemented this idea in Java, creating CONFETTI, which uses both targeted and global hints for fuzzing. In an empirical comparison with two baseline approaches, a state-of-the-art greybox Java fuzzer and a version of CONFETTI without global hinting, we found that CONFETTI covers more branches and finds 15 previously unreported bugs, including 9 that neither baseline could find.

CONFETTI is a research prototype, but nonetheless, we have had success applying it to fuzz the open-source projects Apache Ant, BCEL and Maven, Google's Closure Compiler, and Mozilla's Rhino engine.

You can read more about CONFETTI's design and evaluation in our ICSE 2022 paper.

Architecture

CONFETTI consists of three key components that run in coordination: 1) the parametric fuzzer, which is responsible for input generation and execution of the target program, 2) the whitebox analysis process, which is responsible for performing dynamic taint tracking and constraint collection on interesting inputs, and 3) the CONFETTI coordinator process, which transmits inputs between the fuzzer and the analyzer and interacts with a constraint solver.

CONFETTI 3-component architecture diagram

Building

CONFETTI is currently only compatible with Java 8. To build CONFETTI, be sure that you are using a Java 8 JDK, and then, in the root directory of this repository, simply invoke mvn install.

All of CONFETTI's dependencies are managed using Maven - there are no additional dependencies that you will need to install to compile CONFETTI. For development purposes, it might be useful to also take a look at the following external libraries that are tightly-integrated with CONFETTI:

  • gmu-swe/knarr - Concolic tracing engine, executes the system under test with an input and records the path constraints on each input
  • gmu-swe/phosphor - Dynamic taint tracking runtime for Java. Knarr is built on top of Phosphor.
  • gmu-swe/green-solver - SMT library interface for Java. This is our fork of GreenSolver/green, to which we have added additional Z3 support
  • rohanpadhye/JQF - We use the JQF fuzzing platform to implement CONFETTI; this repository is a fork of JQF, diverging after e7f667bcb22c455f8c06a8f67b0ca0f9afd8a211.

Running

CONFETTI runs on MacOS X and Linux, using Java 8. Please be sure that your JAVA_HOME environment variable is set, and pointing to a Java 8 JVM. CONFETTI requires at least 16GB of RAM. We expect that it will be possible to use CONFETTI on Windows (particularly using WSL), but have not tested this configuration.

CONFETTI can be invoked in almost the same way as JQF. To run CONFETTI using one of the example programs, use the bin/confetti script. For example, to run the Closure compiler test, run the command:

./bin/confetti -c $(scripts/examples_classpath.sh) edu.berkeley.cs.jqf.examples.closure.CompilerTest testWithGenerator fuzz-results

The default behavior is for the fuzzer to run until you terminate it. To run CONFETTI using a timeout, pass the -Dtime=duration property to CONFETTI, which is most easily passed using the environmental variable JVM_OPTS (example: export JVM_OPTS='-Dtime=24h').

Contact

Please feel free to open an issue on GitHub if you run into any issues with CONFETTI. For other matters, please direct your emails to Jonathan Bell.

License

CONFETTI is released under the BSD 2-clause license.

Acknowledgements

CONFETTI builds atop JQF, which is distributed under the BSD 2-clause license.

About

License:BSD 2-Clause "Simplified" License


Languages

Language:Java 92.3%Language:Shell 4.4%Language:Python 2.4%Language:C 0.9%Language:Makefile 0.0%