ViC: Virtual Compiler Is All You Need For Assembly Code Search

TL;DR

This paper introduces ViC, a virtual compiler based on a pre-trained LLM, which enables assembly code generation from source code across multiple languages, significantly improving assembly code search performance.

Contribution

The paper presents ViC, a novel virtual compiler that emulates compilation for any language, facilitating large-scale dataset creation and enhancing assembly code search accuracy.

Findings

Achieved 26% improvement over baseline in assembly code search.

Successfully trained ViC on 20 billion tokens from Ubuntu packages.

Enabled cross-language virtual compilation without real compilers.

Abstract

Assembly code search is vital for reducing the burden on reverse engineers, allowing them to quickly identify specific functions using natural language within vast binary programs. Despite its significance, this critical task is impeded by the complexities involved in building high-quality datasets. This paper explores training a Large Language Model (LLM) to emulate a general compiler. By leveraging Ubuntu packages to compile a dataset of 20 billion tokens, we further continue pre-train CodeLlama as a Virtual Compiler (ViC), capable of compiling any source code of any language to assembly code. This approach allows for virtual compilation across a wide range of programming languages without the need for a real compiler, preserving semantic equivalency and expanding the possibilities for assembly code dataset construction. Furthermore, we use ViC to construct a sufficiently large dataset for assembly code search. Employing this extensive dataset, we achieve a substantial improvement in assembly code search performance, with our model surpassing the leading baseline by 26%.