Model-checking software library API usage rules

摘要

Modern software increasingly relies on using third-party libraries which are accessed via application programming interfaces (APIs). Libraries usually impose constraints on how API functions can be used (API usage rules) and programmers have to obey these API usage rules. However, API usage rules often are not well documented or documented informally. In this work, we show how to use the SCTPL and SLTPL logics to precisely and formally specify API usage rules in libraries, where SCTPL/SLTPL can be seen as an extension of the branching/linear temporal logic CTL/LTL with variables, quantifiers and predicates over the stack. This allows library providers to formally describe API usage rules without knowing how their libraries will be used by programmers. We propose an automated approach to check whether programs using libraries violate API usage rules or not. Our approach consists in modeling programs as pushdown systems (PDSs) and checking API usage rules by SCTPL/SLTPL model-checking for PDSs. To make the model-checking procedure more efficient and precise, we propose an abstraction that reduces drastically the size of the program model and integrate may-alias analysis into our approach to reduce false alarms. Moreover, we characterize two sublogics rSCTPL and rSLTPL of SCTPL and SLTPL that are preserved by the abstraction. We implement our techniques in a tool and apply the tool to check several open-source programs. Our tool finds several previously unknown bugs in several programs. The may-alias analysis avoids most of the false alarms that occur using SCTPL or SLTPL model-checking techniques without may-alias analysis.