SKilLed Communication for Toolchains

Abstract—The creation of a program analysis toolchain in-volves design choices regarding intermediate representations (IRs). Good choices for an IR depend on the analyses performed by a toolchain. In academia, new analyses are developed fre-quently. Therefore, the best single IR of a research-oriented toolchain does not exist. Thus, we will describe our design of an IR that can be easily adapted to new requirements. I. BACKGROUND Successful research projects can outlive their dominant pro-gramming language. Cooperation with a project using another dominant programming language or usage of a library that is only available for a specific language is only feasible if a language-independent IR is used. An evaluation of serializa-tion technology revealed that any available solution would at least lack one of: independence of programming language, ex