PowerBI Data Analyst - Create visualizations and dashboards from scratch
Save 40% on 3 months of Coursera Plus
Overview
Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Learn about KFuzzTest, a novel framework for targeted fuzzing of internal Linux kernel functions, in this 20-minute conference talk from the Linux Plumbers Conference. Discover how traditional system-call fuzzers struggle to reach deeply nested kernel functions, leaving complex data parsers and other critical components under-tested and potentially vulnerable. Explore the innovative KFuzzTest architecture that bridges this gap by directly exposing stateless or low-state internal kernel functions to userspace fuzzers through a developer-friendly macro-based API. Understand how kernel developers can define fuzz test targets alongside their functions using input domain constraints and type annotations, which are compiled into dedicated ELF sections for automatic discovery. Examine the communication mechanism between the fuzzer and in-kernel test harness via debugfs entries, and see how the framework integrates with syzkaller for coverage-guided fuzzing. Gain insights into the design principles, implementation details, and the roadmap for upstreaming this work to the Linux community, empowering developers to write more effective and targeted tests for their kernel code.
Syllabus
KFuzzTest: Targeted Fuzzing of Internal Kernel Functions - Ethan Graham (ETH Zurich)
Taught by
Linux Plumbers Conference