70 likes | 235 Views
WP 4 Efficiency Aspects of Test Case Generation. Daniel Kroening, ETHZ. WP4 / Efficient TCG − Objectives. Efficient algorithms to automatically generate test cases from models Considering models of design and architecture (Matlab/Simulink, class diagrams) Fault-based approach
E N D
WP 4Efficiency Aspects of Test Case Generation Daniel Kroening, ETHZ
WP4 / Efficient TCG − Objectives • Efficient algorithms to automatically generate test cases from models • Considering models of design and architecture (Matlab/Simulink, class diagrams) • Fault-based approach • Tackle scalability problems • Size of real-world models • Enormous number of mutations • Size of test suites • Improve accuracy of coverage metrics • Reports of >40% of faults that cannot be tested
WP4 / Efficient TCG − Progress Observation: one test casecan cover many mutants New: model checking-based method to maxmise number of covered mutants Minimises size of test suites Efficient methods to detectunobservabilityof mutants Requires proofs of unobservability k-induction Craig interpolation
WP4 / Efficient TCG − Progress New: Coverage in Interpolation-based Model Checking Hana Chockler (IBM), Daniel Kroening, Mitra Purandare Computation of mutation coverage in Model Checking To appear at DAC 2010, Anaheim, CA
WP4 / Efficient TCG − Progress Input: Model and set of mutations T Output: Test suite that covers T (to appear in the FMCO post proceedings)
WP4 / Efficient TCG − Planning Improve implementation Empirical evaluation on Ford/ReLab demonstrators Evaluation on models provided by Airbus UK Evaluation on automotive C code