The Research And Implementation Of A Mutation-Based C Program Fault Injector

Fault injection plays a critical role in the field of software testing. It puts the faults in a software system on purpose to see how the system reacts. As a part of fault injection, code fault injection is popular as its low cost, convenience and portability. However, most of the injection work is done by hand at present. It is inefficient, time-consuming and the output sometimes is useless. These problems can be solved by the automatic way and that is why it is important to do research into the automatic fault injection.The main points are how to inject fault and what kinds of fault are injected. This paper presents a mutation-based fault.injection method for the C program. To achieve the goal, the method gets the information of the program by static analysis and locates the injection point by using the abstract syntax tree. Then the semantic environment would be checked to narrow the scale of mutations. At last, a mutation would be picked from the candidates and the work is done. Moreover, a strategy of using semantic-based mutators is proposed to improve the efficiency of mutation testing. To make the faults more real, the edit distance of string is adopted for choosing mutations.Based on the Code Test System (CTS), the author developed a fault injector containing all of these research findings. It can inject faults into C programs automatically and the outputs are adaptive for mutation testing.