A Look Inside CERT Fuzzing Tools
Hi, this is Allen Householder of the CERT Vulnerability Analysis team. If you've been following this blog for a while, you are probably familiar with our fuzzing tools: Dranzer, the CERT Basic Fuzzing Framework (BFF), and the CERT Failure Observation Engine (FOE). While creating tools that can find and analyze vulnerabilities makes up a significant portion of our work in the CERT Vulnerability Analysis team, our focus is on developing and communicating the knowledge we've built into those systems.
To that end, we recently published a pair of reports that describe a few of the heuristics and algorithms implemented in the BFF and FOE fuzzing tools. We briefly mentioned these techniques in the release announcements for the tools, but did not describe how they work in detail. Abstracts and links to the reports can be found below.
Probability-Based Parameter Selection for Black-Box Fuzz Testing
Dynamic, randomized-input functional testing, or black-box fuzz testing, is an effective technique for finding security vulnerabilities in software applications. Parameters for an invocation of black-box fuzz testing generally include known-good input to use as a basis for randomization (i.e., a seed file) and a specification of how much of the seed file to randomize (i.e., the range). This report describes an algorithm that applies basic statistical theory to the parameter selection problem and automates selection of seed files and ranges. This algorithm was implemented in an open-source, file-interface testing tool and was used to find and mitigate vulnerabilities in several software applications. This report generalizes the parameter selection problem, explains the algorithm, and analyzes empirical data collected from the implementation. Results of using the algorithm show a marked improvement in the efficiency of discovering unique application errors over basic parameter selection techniques.
Well There's Your Problem: Isolating the Crash-Inducing Bits in a Fuzzed File
Mutational input testing (fuzzing, and in particular dumb fuzzing) is an effective technique for discovering vulnerabilities in software. However, many of the bitwise changes in fuzzed input files are not relevant to the actual software crashes found. This report describes an algorithm that efficiently reverts bits from the fuzzed file to those found in the original seed file, keeping only the minimal bits required to recreate the crash under investigation. This technique reduces the complexity of analyzing a crashing test case by eliminating the changes to the seed file that are not essential to the crash being evaluated.
By the way, we're hiring. We've got a few different vulnerability analysis openings along with a variety of other technical security positions posted at www.cert.org/jobs/. If improving how vulnerabilities are discovered, analyzed, and remediated is your thing, we'd like to talk to you.
More By The Author
Prioritizing Vulnerability Response with a Stakeholder-Specific Vulnerability Categorization
Comments on Voluntary Voting System Guidelines 2.0 Principles and Guidelines
More In CERT/CC Vulnerabilities
The Latest Work from the SEI: Coordinated Vulnerability Disclosure, Cybersecurity Research, Cyber Risk and Resilience, and the Importance of Fostering Diversity in Software Engineering
This post has been shared 0 times.
Get updates on our latest work.
Sign up to have the latest post sent to your inbox weekly.