Data races complicate programming language semantics, and a data race is often a bug. Existing techniques detect data races and define their semantics by detecting conflicts between synchronization-free regions (SFRs). However, such techniques either modify hardware or slow programs dramatically, preventing always-on use today. This paper describes Valor, a sound, precise, software-only region conflict detection analysis that achieves high performance by eliminating the costly analysis on each read operation that prior approaches require. Valor instead logs a region's reads and lazily detects conflicts for logged reads when the region ends. As a comparison, we have also developed FastRCD, a conflict detector that leverages the epoch optimization strategy of the FastTrack data race detector. We evaluate Valor, FastRCD, and FastTrack, showing that Valor dramatically outperforms FastRCD and FastTrack. Valor is the first region conflict detector to provide strong semantic guarantees for racy program executions with under 2X slowdown. Overall, Valor advances the state of the art in always-on support for strong behavioral guarantees for data races.
Wed 28 Oct Times are displayed in time zone: Eastern Time (US & Canada) change
15:30 - 17:00: 4. ConcurrencyOOPSLA at Grand Station 1 Chair(s): Wolfgang De MeuterVrije Universiteit Brussel | |||
15:30 - 15:52 Talk | AutoMO: Automatic Inference of Memory Order Parameters for C/C++11 OOPSLA Peizhao OuUniversity of California at Irvine, USA, Brian DemskyUniversity of California at Irvine, USA DOI | ||
15:52 - 16:15 Talk | Valor: Efficient, Software-Only Region Conflict Exceptions OOPSLA Swarnendu BiswasOhio State University, USA, Minjia ZhangOhio State University, USA, Michael D. BondOhio State University, USA, Brandon LuciaCarnegie Mellon University, USA DOI Pre-print | ||
16:15 - 16:37 Talk | Automatic Memory Reclamation for Lock-Free Data Structures OOPSLA DOI | ||
16:37 - 17:00 Talk | Protocol-Based Verification of Message-Passing Parallel Programs OOPSLA Hugo LópezTechnical University of Denmark, Eduardo MarquesUniversity of Lisbon, Portugal, Francisco MartinsUniversity of Lisbon, Portugal, Nicholas NgImperial College London, UK, César SantosUniversity of Lisbon, Portugal, Vasco VasconcelosUniversity of Lisbon, Portugal, Nobuko YoshidaImperial College London, UK Link to publication |