Search papers, labs, and topics across Lattice.
This paper introduces WasmMend, a novel system designed to automatically repair functional discrepancies between WebAssembly (Wasm) and native binaries by employing a two-stage approach. The first stage utilizes differential trace analysis to pinpoint where the executions diverge, and the second stage leverages LLM agents to reason about the discrepancies and generate corrective patches. Experimental results indicate that WasmMend achieves a 70.0% fix rate, significantly outperforming existing methods, thereby highlighting the effectiveness of divergence-guided reasoning in cross-platform repair tasks.
WasmMend achieves a remarkable 70% fix rate for discrepancies between WebAssembly and native binaries, showcasing the power of divergence-guided reasoning in automated repair.
WebAssembly (Wasm) promises seamless reuse of C/C++ codebases as portable, fast, sandboxed binaries. In practice, however, this promise often falls short: recent studies show that cross-compiling the same C/C++ source to Wasm and native binaries frequently leads to runtime discrepancies, owing to library implementation differences or compiler bugs. Since the root causes lie in the platform-level runtime and are hidden beneath the source code, even state-of-the-art LLM-based repair agents often fail to fix these discrepancies. In this paper, we present WasmMend, the first system to automatically repair Native-Wasm functional discrepancies. WasmMend converts the undirected exploration to a focused reasoning task in two stages: First, a novel differential trace analysis approach localizes the function where Wasm and native executions initially diverge; guided by this localization, LLM agents then reason about the root causes and generate patches that eliminate the divergent behavior. Experiments on real-world C/C++ projects show that WasmMend achieves a fix rate of 70.0%, compared to 50.2% for the agentic baseline and 54.5\% for the approach augmented with repair-time LLM-based instrumentation, demonstrating the value of divergence-guided reasoning for cross-platform repair.
