App Genome: Callback Sequencing in Android

摘要

Recent analysis shows that the callback sequences are of great importance in the analysis of Android applications (apps for short), due to the app's event-driven nature. However, existing works only extract a part of the callback sequences, depending on the need for their specific properties. We propose App Genome sequencing, an automatic fine-grained callback extraction, covering lifecycle and non-lifecycle, inner-and inter-component callback relations, as well as related attributes, including global objects and operations, along the callback sequences. The extracted App Genome facilitates more complete analysis of Android apps, since it contains more callback sequences and data information, than existing works. We use a process algebra called CSP# to represent the App Genome. We implement our method as a tool, which takes an app as input, automatically generates the CSP# model of the App Genome and automatically invokes the model checker to verify a given property.