Cornucopia: A Framework for Feedback Guided Generation of Binaries
Binary analysis is an important capability required for many security and software engineering applications. Consequently, there are many binary analysis techniques and tools with varied capabilities. However, testing these tools requires a large, varied binary dataset with corresponding source-level information. In this paper, we present Cornucopia, an architecture agnostic automated framework that can generate a plethora of binaries from corresponding program source by exploiting compiler optimizations and feedback-guided learning. Our evaluation shows that Cornucopia was able to generate 309K binaries across four architectures (x86, x64, ARM, MIPS) with an average of 403 binaries for each program and outperforms Bintuner, a similar technique. Our experiments revealed issues with the LLVM optimization scheduler resulting in compiler crashes ($\sim$300). Our evaluation of four popular binary analysis tools Angr, Ghidra, Idapro, and Radare, using Cornucopia generated binaries, revealed various issues with these tools. Specifically, we found 263 crashes in Angr and one memory corruption issue in Idapro. Our differential testing on the analysis results revealed various semantic bugs in these tools. We also tested machine learning tools, Asmvec, Safe, and Debin, that claim to capture binary semantics and show that they perform poorly (For instance, Debin F1 score dropped to 12.9% from reported 63.1%) on Cornucopia generated binaries. In summary, our exhaustive evaluation shows that Cornucopia is an effective mechanism to generate binaries for testing binary analysis techniques effectively.