Persistent HyTM via Fast Path Fine-Grained Locking

2025-01-04

arXiv.org

AI Summary

This paper addresses a critical issue in hardware transactional memory (HTM)–non-volatile memory (NVM) co-persistence: NVM writes frequently abort HTM transactions, undermining performance and correctness. To resolve this, we propose a novel paradigm that leverages HTM’s fast path not for direct data updates, but for fine-grained lock acquisition—thereby protecting NVM-resident data while ensuring atomicity and persistence. This is the first approach to repurpose HTM exclusively for synchronization (lock management) rather than data mutation, enabling persistent linearizability and supporting configurable weak or strong progress guarantees. Evaluated under read-intensive workloads, our design significantly outperforms state-of-the-art persistent software transactional memory (STM) and hybrid TM (HyTM) systems. Microbenchmarks demonstrate substantial throughput improvements, while rigorously preserving persistence semantics and consistency guarantees.

Technology Category

• Machine Learning: Hardware-aware ML
• Data Mining & Knowledge Management: Scalability, Parallel & Distributed Systems
• Multiagent Systems: Mechanism Design

Application Category

• Security and Privacy: Data transparency and provenance
• Systems and Infrastructure for Web, Mobile and WoT: Data management and stream processing for Web, mobile and wireless applications
• Responsible Web: Data and user privacy-enhancing technologies for the Web

Abstract

Utilizing hardware transactional memory (HTM) in conjunction with non-volatile memory (NVM) to achieve persistence is quite difficult and somewhat awkward due to the fact that the primitives utilized to write data to NVM will abort HTM transactions. We present several persistent hybrid transactional memory (HyTM) that, perhaps counterintuitively, utilize an HTM fast path primarily to read or acquire fine-grained locks which protect data items. Our implementations guarantee durable linearizable transactions and the STM path satisfies either weak progressiveness or strong progressiveness. We discuss the design choices related to the differing progress guarantees and we examine how these design choices impact performance. We evaluate our persistent HyTM implementations using various microbenchmarks. Our implementations achieve improved performance especially for read dominant workloads compared to state of the art persistent STMs and persistent HyTMs despite the challenges and apparent awkwardness of using current implementation HTM to achieve persistence.

Problem

• Achieving persistence with HTM and NVM is challenging
• Designing persistent HyTM with durable linearizable transactions
• Improving performance for read-dominant workloads in HyTM

Innovation

• Hybrid TM with HTM fast path
• Fine-grained locking for NVM persistence
• Durable linearizable transaction guarantees