RON: One-Way Circular Shortest Routing to Achieve Efficient and Bounded-waiting Spinlocks
summary
RON uses a routing table and precomputes a shortest circular path to hand locks in a one-way circular routing among cores, turning lock handover into an online TSP minimization. It runs on an AMD 2990wx (32 cores, 64 SMT) and reduces handover latency and variability, improving Google LevelDB user-space throughput by 22.1% and 24.2% versus ShflLock and C-BO-MCS, 1.8× faster in-kernel, and RON-plock achieves bounded waiting while giving 3.7× and 18.9× gains under oversubscription.
AI-written summary · about
chapters (6)
- 0:00Research motivation
- 1:10Performance and fairness
- 3:12The Run algorithm
- 5:08Addressing oversubscription
- 8:25Performance evaluation
- 13:07Conclusion
related talks
37mTicket Spinlocks in Zephyr RTOS: What, Where & Why
01h 38mLock-free Programming is Dead - Long Live Lock-free Programming!
37mTicket Spinlocks in Zephyr RTOS: What, Where & Why
20mRRR-SMR: Reduce, Reuse, Recycle: Better Methods for Practical Lock-Free Data Structures
17m60 Adaptive userspace spinlocks with rseq
42mReader-Writer Lock versus Mutex - Understanding a Lost Bet
40mKernel Locking Engineering
28mRevisit Multi-Reader Synchronizations for Scalable Applications
