FuzzBtor2: A Random Generator of Word-Level Model Checking Problems in Btor2 Format

Abstract

Abstract We present , a fuzzer to generate random word-level model checking problems in Btor2 format. Btor2 is one of the mainstream input formats for word-level hardware model checking and was used in the most recent hardware model checking competition. Compared to bit-level one, word-level model checking is a more complex research field at an earlier stage of development. Therefore, it is necessary to develop a tool that can produce a large number of test cases in Btor2 format to test either existing or under-developed word-level model checkers. To evaluate the practicality of , we tested the state-of-the-art word-level model checkers and with the generated benchmarks. Experimental results show that both tools are buggy and not mature enough, which reflects the practical value of .