Data Race Detection in Large-Scale Systems
Contributor Active
Detecting concurrency issues in major open-source systems.
Applied advanced data race detection and concurrency analysis techniques, including ThreadSanitizer, RaceHunter, and fuzzing-based approaches, to large-scale software systems.
Target systems included:
- JSVM,
- PostgreSQL,
- QEMU,
- the Linux kernel, and
- OpenMP runtimes.
This work led to the discovery of multiple concurrency-related defects, including subtle data races and synchronization issues, followed by submission of bug reports and patches.