Inteligentní testování a analýza paralelních programů
Project Period: 1. 4. 2013 - 31. 3. 2015
Project Type: grant
Agency: Ministry of Education, Youth and Sports Czech Republic
Program: KONTAKT II (2011-2017)
Multi-threaded programs, Concurrency, Testing, Verification, Static and dynamic analysis, Noise injection, Meta-heuristics, Genetic algorithms, Genetic programming, Data mining.
The goal of the project is to conduct a joint research with the Israeli partner that will significantly improve the efficiency, generality, and degree of automation of testing and analysis of multi-threaded programs. Improved efficiency should be achieved by developing new heuristics for injection of noise into thread scheduling which will increase the probability of discovering synchronization-related errors when testing concurrent programs. Increased generality should be achieved by developing new dynamic error detection algorithms, targeting even new and/or so far less studied synchronization mechanisms as well as currently less supported classes of programs (e.g., noise-based dynamic analysis of concurrent binary code). Higher automation should be achieved by employing advanced optimization algorithms (meta-heuristics) and data mining on test data for an automated optimization of the process of testing of concurrent code wrt. various criteria without a need of a manual intervention of testers.