Search papers, labs, and topics across Lattice.
This paper introduces LEAF, a dynamic analysis framework tailored for Rust, addressing the current lack of general-purpose tools in this rapidly evolving programming ecosystem. By leveraging Rust's rich semantic information, LEAF captures and augments MIR-level data with runtime facts, providing an event-driven interface for various analyses. The framework's effectiveness is demonstrated through three dynamic analyses, revealing its practicality despite manageable compile-time and runtime overheads.
LEAF transforms Rust program analysis by providing a dynamic framework that captures rich semantic information at runtime, enabling advanced analysis tasks.
This paper presents LEAF, an instrumentation-based dynamic analysis framework for Rust. Although Rust has grown rapidly in recent years, the landscape of program analysis tools for Rust is still in relatively early stages. One notable gap is the lack of a general-purpose dynamic analysis framework that can support different analysis tasks. LEAF aims to fill this gap by providing a Rust-native framework for analyzing Rust programs at runtime. Rust provides rich semantic information through its ownership model, type system, memory model, and compiler-level representation. Therefore, LEAF focuses on how to make this information available to dynamic analyses. In particular, LEAF captures MIR-level semantic information, augments it with runtime facts, and delivers it to analyses as an event stream through Dynamic MIR (DMIR), an event-driven programming interface. Through three substantial dynamic analyses -- a concolic executor, a Rust-specific sanitizer, and a control-flow tracer -- we demonstrate the practicality and expressiveness of LEAF. Our evaluation further shows that LEAF's compile-time and runtime overhead is meaningful but manageable.
