DR.FIX: Automatically Fixing Data Races at Industry Scale
This program is tentative and subject to change.
Data races are a prevalent class of concurrency bugs in shared-memory parallel programs, posing significant challenges to software reliability and reproducibility.
While there is an extensive body of research on detecting data races and a wealth of practical detection tools across various programming languages, considerably less effort has been directed toward automatically fixing data races at an industrial scale.
In large codebases, data races are continuously introduced and exhibit myriad patterns, making automated fixing particularly challenging.
In this paper, we tackle the problem of automatically fixing data races at an industrial scale. We present \textsc{Dr.Fix}, a tool that combines large language models (LLMs) with program analysis to generate fixes for data races in real-world settings, effectively addressing a broad spectrum of racy patterns in complex code contexts.
Implemented for Go—the programming language widely used in modern microservice architectures where concurrency is pervasive and data races are common—\textsc{Dr.Fix} seamlessly integrates into existing development workflows.
We detail the design of \textsc{Dr.Fix} and examine how individual design choices influence the quality of the fixes produced.
Over the past 18 months, \textsc{Dr.Fix} has been integrated into developer workflows at Uber
demonstrating its practical utility.
During this period, \textsc{Dr.Fix} produced patches for 224 (55%) from a corpus of 404 data races spanning various categories; 193 of these patches (86%) were accepted by more than a hundred developers via code reviews and integrated into the codebase.
This program is tentative and subject to change.
Thu 19 JunDisplayed time zone: Seoul change
10:30 - 12:10 | |||
10:30 20mTalk | Fast Direct Manipulation Programming with Patch-Reconciliation Correspondence PLDI Research Papers Parker Ziegler University of California at Berkeley, Justin Lubin University of California at Berkeley, Sarah E. Chasins University of California at Berkeley DOI | ||
10:50 20mTalk | An Interactive Debugger for Rust Trait Errors PLDI Research Papers DOI Pre-print | ||
11:10 20mTalk | Spineless Traversal for Layout Invalidation PLDI Research Papers Marisa Kirisame University of Utah, Tiezhi Wang Tongji University, Pavel Panchekha University of Utah DOI | ||
11:30 20mTalk | DR.FIX: Automatically Fixing Data Races at Industry Scale PLDI Research Papers Farnaz Behrang Uber Technologies, Zhizhou (Chris) Zhang Uber Technologies, Georgian-Vlad Saioc Aarhus University, Peng Liu Uber Technologies, Milind Chabbi Uber Technologies DOI | ||
11:50 20mTalk | Program Skeletons for Automated Program Translation PLDI Research Papers Bo Wang National University of Singapore, Tianyu Li National University of Singapore, Ruishi Li National University of Singapore, Umang Mathur National University of Singapore, Prateek Saxena National University of Singapore DOI |