rediscovering_bugs

Comparison to RTL Fuzzer Cascade (Section 5.4) & Sequence Generation Evaluation (Section 5.5)

This directory provides the scripts used to compare our sequence generation to Cascade and to evaluate the weighted sequence generation. To start the evaluation of all workloads (GhostWrite, C906 Halt and Cascade) run:

./gen_all_results.sh

This will take 3-10 hours and produces two json files results_sequence_gen_eval.json and results_sequence_gen_eval_weighted.json. As the results are deterministic we recommend to use the provided files.

From these Table 5 can be generated:

make pdf
xdg-open out/result_table.pdf