This post discusses the use of mutation-based fuzzing for LLVM optimization passes. It introduces a custom mutator for LLVM IR and describes the results of fuzzing with this mutator. Bugs found using alive-mutate are also highlighted.
Sort: