A Type System for Safe Intermittent Computing

Author:

Surbatovich Milijana1ORCID,Spargo Naomi1ORCID,Jia Limin1ORCID,Lucia Brandon1ORCID

Affiliation:

1. Carnegie Mellon University, USA

Abstract

Batteryless energy-harvesting devices enable computing in inaccessible environments, at a cost to programmability and correctness. These devices operate intermittently as energy is available, using a recovery system to save and restore state. Some program tasks must execute atomically w.r.t. power failures, re-executing if power fails before completion. Any re-execution should typically be idempotent —its behavior should match the behavior of a single execution. Thus, a key aspect of correct intermittent execution is identifying and recovering state causing undesired non-idempotence. Unfortunately, past intermittent systems take an ad-hoc approach, using unsound dataflow analyses or conservatively recovering all written state. Moreover, no prior work allows the programmer to directly specify idempotence requirements (including allowable non-idempotence). We present curricle, the first type system approach to safe intermittence, for Rust. Type level reasoning allows programmers to express requirements and retains alias information crucial for sound analyses. Curricle uses information flow and type qualifiers to reject programs causing undesired non-idempotence. We implement Curricle’s type system on top of Rust’s compiler, evaluating the prototype on benchmarks from prior work. We find that Curricle benefits application programmers by allowing them to express idempotence requirements that are checked to be satisfied, and that targeting programs checked with Curricle allows intermittent system designers to write simpler recovery systems that perform better.

Funder

National Science Foundation

Publisher

Association for Computing Machinery (ACM)

Subject

Safety, Risk, Reliability and Quality,Software

Cited by 1 articles. 订阅此论文施引文献 订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献

1. Mapi-Pro: An Energy Efficient Memory Mapping Technique for Intermittent Computing;ACM Transactions on Architecture and Code Optimization;2023-12-14

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3