DR.FIX: Automatically Fixing Data Races at Industry Scale
Farnaz Behrang, Zhizhou Zhang, Georgian-Vlad Saioc, Peng Liu, Milind Chabbi · 2025
Read DR.FIX: Automatically Fixing Data Races at Industry Scale by Farnaz Behrang, Zhizhou Zhang, Georgian-Vlad Saioc, Peng Liu, Milind Chabbi (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-15637.txt.
Introduction
Farnaz Behrang and four co-authors present a tool that repairs data races in this two thousand twenty-five paper. Data races strike when parallel code touches shared memory without order. They are a common and slippery class of bugs. Finding them has good tools; fixing them at company scale does not. In huge codebases, races appear in endless patterns.
Their tool, Dr.Fix, pairs a large language model with program analysis. The analysis grounds the model's suggested patches in the real code. The system targets the Go language, widely used for concurrent servers. It has been applied to real industrial code, not toy examples. The paper charts which race patterns the tool handles. Software teams fighting flaky parallel bugs are the audience.
Catalog record
| CREATOR | Farnaz Behrang, Zhizhou Zhang, Georgian-Vlad Saioc, Peng Liu, Milind Chabbi |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 117 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2504-15637