CIDFuzz: Fuzz testing for continuous integration

As agile software development and extreme programing have become increasingly popular, continuous integration (CI) has a widely used collaborative work method. However, it is common to make changes frequently project during CI. If existing testing methods are applied CI directly, will be difficult resources focus on generated by CI, which results in insufficient for changes. To solve this problem, we propose fuzz method First, differential analysis performed determine the change points added taint source set, static conducted calculate distances between each basic block sources. Then, under test instrumented according distances. During testing, allocated based seed coverage effectively. Using proposed methods, implement CIDFuzz as prototype tool, experiments four open-source projects that use Experimental show that, compared with AFL AFLGo, can reduce time costs of covering up 39.59% 41.64%, respectively. Also, reproducing vulnerabilities 34.78% 25.55%.