Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Reusing Legacy Code in WebAssembly: Key Challenges of Cross-Compilation and Code Semantics Preservation

Sara Baradaran, Liyan Huang, Mukund Raghothaman, Weihang Wang

TL;DR

This work tackles the challenge of reusing legacy C/C++ code by cross-compiling to WebAssembly and preserving code semantics. It conducts a two-phase study: first identifying practical build-time challenges when porting code to Wasm, then evaluating semantic fidelity using a differential testing framework. The authors introduce WasmChecker, which leverages CodeQL-based static analysis, automated build transformations, and VFS file preloading to align WebAssembly builds with native binaries, and they validate semantic differences across 135 projects with 34,480 tests, uncovering 226 discrepancies (with 220 true and 6 false positives) and 11 new Emscripten bugs. The work provides a valuable dataset and open-source tooling to improve cross-compilation reliability and informs future improvements in Wasm compiler and runtime ecosystems. The findings highlight that WebAssembly trades some code reuse ease for portability and performance, and that careful configuration and code adjustments are often required to achieve faithful semantics across native and Wasm binaries.

Abstract

WebAssembly (Wasm) has emerged as a powerful technology for executing high-performance code and reusing legacy code in web browsers. With its increasing adoption, ensuring the reliability of WebAssembly code becomes paramount. In this paper, we investigate how well WebAssembly compilers fulfill code reusability. Specifically, we inquire (1) what challenges arise when cross-compiling a high-level language codebase into WebAssembly and (2) how faithfully WebAssembly compilers preserve code semantics in this new binary. Through a study on 115 open-source codebases, we identify the key challenges in cross-compiling legacy C/C++ code into WebAssembly, highlighting the risks of silent miscompilation and compile-time errors. We categorize these challenges based on their root causes and propose corresponding solutions. We then introduce a differential testing approach, implemented in a framework named WasmChecker, to investigate the semantics equivalency of code between native x86-64 and WebAssembly binaries. Using WasmChecker, we provide a witness that WebAssembly compilers do not necessarily preserve code semantics when cross-compiling high-level language code into WebAssembly due to different implementations of standard libraries, unsupported system calls/APIs, WebAssembly's unique features, and compiler bugs. Furthermore, we have identified 11 new bugs in the Emscripten compiler toolchain, all confirmed by Emscripten developers. As proof of concept, we make our framework and the collected dataset of open-source codebases publicly available.

Reusing Legacy Code in WebAssembly: Key Challenges of Cross-Compilation and Code Semantics Preservation

TL;DR

This work tackles the challenge of reusing legacy C/C++ code by cross-compiling to WebAssembly and preserving code semantics. It conducts a two-phase study: first identifying practical build-time challenges when porting code to Wasm, then evaluating semantic fidelity using a differential testing framework. The authors introduce WasmChecker, which leverages CodeQL-based static analysis, automated build transformations, and VFS file preloading to align WebAssembly builds with native binaries, and they validate semantic differences across 135 projects with 34,480 tests, uncovering 226 discrepancies (with 220 true and 6 false positives) and 11 new Emscripten bugs. The work provides a valuable dataset and open-source tooling to improve cross-compilation reliability and informs future improvements in Wasm compiler and runtime ecosystems. The findings highlight that WebAssembly trades some code reuse ease for portability and performance, and that careful configuration and code adjustments are often required to achieve faithful semantics across native and Wasm binaries.

Abstract

WebAssembly (Wasm) has emerged as a powerful technology for executing high-performance code and reusing legacy code in web browsers. With its increasing adoption, ensuring the reliability of WebAssembly code becomes paramount. In this paper, we investigate how well WebAssembly compilers fulfill code reusability. Specifically, we inquire (1) what challenges arise when cross-compiling a high-level language codebase into WebAssembly and (2) how faithfully WebAssembly compilers preserve code semantics in this new binary. Through a study on 115 open-source codebases, we identify the key challenges in cross-compiling legacy C/C++ code into WebAssembly, highlighting the risks of silent miscompilation and compile-time errors. We categorize these challenges based on their root causes and propose corresponding solutions. We then introduce a differential testing approach, implemented in a framework named WasmChecker, to investigate the semantics equivalency of code between native x86-64 and WebAssembly binaries. Using WasmChecker, we provide a witness that WebAssembly compilers do not necessarily preserve code semantics when cross-compiling high-level language code into WebAssembly due to different implementations of standard libraries, unsupported system calls/APIs, WebAssembly's unique features, and compiler bugs. Furthermore, we have identified 11 new bugs in the Emscripten compiler toolchain, all confirmed by Emscripten developers. As proof of concept, we make our framework and the collected dataset of open-source codebases publicly available.
Paper Structure (13 sections, 1 equation, 12 figures, 5 tables)

This paper contains 13 sections, 1 equation, 12 figures, 5 tables.

Table of Contents

  1. Introduction
  2. Methodology
  3. Data Collection
  4. Building the Codebases
  5. RQ1: Challenges in Migrating Codebases to WebAssembly
  6. RQ2: Reliability of Cross-Compiled Code in WebAssembly
  7. Analysis of Fixable WebAssembly Test Failures
  8. WasmChecker: A Differential Testing Framework
  9. Effectiveness of WasmChecker
  10. Analysis of Semantics Divergence Among Binaries
  11. Discussion
  12. Related Work
  13. Conclusion

Figures (12)

  • Figure 1: Emscripten compiler architecture.
  • Figure 2: Commands for building a sample codebase in x86-64 and their equivalent for building in WebAssembly.
  • Figure 3: Stack smashing protection in C/C++ compilers.
  • Figure 4: Sample code snippets to show how the default settings of Emscripten produces high-performance WebAssembly code by sacrificing runtime reliability and disabling support for features that bring overhead.
  • Figure 5: Architecture of WasmChecker.
  • ...and 7 more figures