Using dynamic binary analysis for tracking pointer data

The examination and monitoring of binaries during runtime, referred to as dynamic binary analysis, is a widely adopted approach, especially in the field of security and software vulnerabilities. Fundamentally, it provides one with a means to understand and reason about binary executions. There are various applications of dynamic binary analysis, including vulnerability analysis, malware analysis, and Web security. One technique typically employed to perform dynamic analysis is taint analysis, which revolves around inspecting interesting information flows [3]. In this approach, taint marks are associated with values that are (1) introduced via defined sources and (2) propagated to other values to keep track of information flow. Marks may also be removed (untainted) once a defined sink has been reached. In addition, taint checking is also carried out in order to determine whether or not certain runtime behaviours of the program occur. The properties describing how taint analysis is performed, i.e taint introduction, propagation and checking, are specified by a set of rules referred to as a taint policy. One convenient way to define taint rules is in the form of operational semantics rules, as it avoids ambiguity issues. Rule 1 specifies the general form of a taint rule used in this paper. Given the current machine context of the program 4 and a statement, the rule specifies the end result, after the computation has been carried out.peer-reviewe