Shiyou Huang, Bowen Cai, and Jeff Huang, Texas A&M University
We present a new technique, H3, for reproducing Heisenbugs in production runs on commercial hardware. H3 integrates the hardware control flow tracing capability provided in recent Intel processors with symbolic constraint analysis. Compared to a state-of-the-art solution, CLAP, this integration allows H3 to reproduce failures with much lower runtime overhead and much more compact trace. Moreover, it allows us to develop a highly effective core-based constraint reduction technique that significantly reduces the complexity of the generated symbolic constraints. H3 has been implemented for C/C++ and evaluated on both popular benchmarks and real-world applications. It reproduces real-world Heisenbugs with overhead ranging between 1.4%- 23.4%, up to 8X more efficient than CLAP, and incurs only 4.9% runtime overhead on PARSEC benchmarks.
Open Access Media
USENIX is committed to Open Access to the research presented at our events. Papers and proceedings are freely available to everyone once the event begins. Any video, audio, and/or slides that are posted after the event are also free and open to everyone. Support USENIX and our commitment to Open Access.
author = {Shiyou Huang and Bowen Cai and Jeff Huang},
title = {Towards {Production-Run} Heisenbugs Reproduction on Commercial Hardware},
booktitle = {2017 USENIX Annual Technical Conference (USENIX ATC 17)},
year = {2017},
isbn = {978-1-931971-38-6},
address = {Santa Clara, CA},
pages = {403--415},
url = {https://www.usenix.org/conference/atc17/technical-sessions/presentation/huang},
publisher = {USENIX Association},
month = jul
}