Jinn: Synthesizing a Dynamic Bug Detector for Foreign Language Interfaces

Abstract

Programming language specifications mandate static and dynamic analyses to preclude syntactic and semantic errors. Although individual languages are usually well-specified, composing languages in multilingual programs is not. Because multilingual programs are prevalent, poor specification is a source of many errors. For example, virtually all Java programs compose Java and C with the Java Native Interface (JNI). Unfortunately, JNI is informally specified and thus, Java compilers and virtual machines (VMs) check only a small subset of JNI constraints. Worse, Java compiler and VM implementations inconsistently check constraints. This paper’s most significant contribution is to show how to synthesize dynamic analyses from state machines to detect foreign function interface (FFI) violations. To demonstrate the generality of our approach, we build FFI state machines that encode specifications for JNI and Python/C. Although we identify over a thousand FFI correctness constraints, we show that they fall into three classes and a modest number of state machines encode them. From these state machines, we generate context-specific FFI dynamic analysis. For Java, we insert this analysis in a library that interposes on all language transitions and thus is compiler and VM independent. We call the resulting dynamic bug detection tool Jinn. We show Jinn detects and diagnoses a wide variety of FFI bugs that other tools do not. This paper lays the foundation for better specification and enforcement of FFIs and a more principled approach to developing correct multilingual software. 1.