Automated fuzz testing of closed source applications based on the code coverage
summary
Automated mutation fuzzing uses corpus distillation to reduce hundreds of thousands of PDF samples to a few hundred that produce the same code coverage then fuzzes those to find bugs the pipeline records basic‑block coverage (using IDA Pro to extract blocks), patches the binary to speed coverage logging and runs mutations so 366000 PDFs shrink to 726 while preserving coverage the run finds thousands of crashes but only tens of unique vulnerabilities across Adobe Apple Microsoft document readers and the tooling proves practical and fast.
AI-written summary · about
chapters (4)
related talks
01h 15mMake your programs more reliable with Fuzzing
25mGoing Beyond Coverage-Guided Fuzzing with Structured Fuzzing
52mStructure aware fuzzing
39mMaking Fuzzing Part of Your Software Development Lifecycle
25mHow fuzzing helps to find bugs
24mWhy you should be using automated fuzzing to test for security issues in your code
47mFuzzing Your Code
31m“Fuzz or lose..."