Closed elopez closed 8 months ago
I'll dig in a bit more in depth later, but my top-level comment is that you don't necessarily need to keep to the TLV format for fuzzer test cases if it doesn't make sense. Does https://github.com/google/fuzzing/blob/master/docs/split-inputs.md#fuzzed-data-provider make sense for this fuzzer?
This PR refactors the existing TLV handling code to make it a bit more reusable. It then adds a new standalone harness that fuzzes the bufq buffer handling.