Coursera Spring Sale
40% Off Coursera Plus Annual!
Grab it
Explore a conference talk that introduces a novel test oracle for the BPF verifier designed to detect false negatives in fuzzing operations. Learn how current fuzzers effectively uncover deadlocks, warnings, and memory errors but struggle to identify cases where the verifier incorrectly accepts programs. Discover the proposed solution that preserves verifier expectations on registers and stack slots at pruning points, enabling runtime validation by interpreters or JIT compilers. Understand the implementation approach using BPF maps to save verifier states and expose them for debugging purposes, with initial focus on scalar values. Examine how this oracle addresses limitations in existing fuzzing methodologies and enhances the reliability of BPF program verification through runtime checking mechanisms that issue warnings when concrete values fall outside verifier expectations.
