Compiler Optimizations for Safe Insertion of Checkpoints in Intermittently Powered Systems

Abstract

A large and increasing number of Internet-of-Things devices are not equipped with batteries and harvest energy from their environment. Many of them cannot be physically accessed once they are deployed (embedded in civil engineering structures, sent in the atmosphere or deep in the oceans). When they run out of energy, they stop executing and wait until the energy level reaches a threshold. Programming such devices is challenging in terms of ensuring memory consistency and guaranteeing forward progress. Previous work has proposed to insert checkpoints in the program so that execution can resume from well-defined locations. In this work, we propose to define these checkpoint locations based on statically-computed worst-case energy consumption of code sections. We also apply classical compiler optimizations in order to decrease the required number of checkpoints at runtime. As our method is based upon worst-case energy consumption, we can guarantee memory consistency and forward progress.

Cost-optimal timed trace synthesis for scheduling of intermittent embedded systems

“…Then, an alternative direction has been explored in which the program is decomposed into subsequence, usually named tasks, executed according to a transactionnal semantics where the commit part correspond to a checkpoint. The tasks are constructed either by the programmer [4] or by the compiler [5,6], possibly with an adaptative phase during execution to activate or deactivate checkpoints. Task-based intermittent computing can be with rollbacks [4] if it runs until a power loss, or without if it stops the system after a checkpoint if it cannot safely reach the next checkpoint with the available energy [6].…”

“…The tasks are constructed either by the programmer [4] or by the compiler [5,6], possibly with an adaptative phase during execution to activate or deactivate checkpoints. Task-based intermittent computing can be with rollbacks [4] if it runs until a power loss, or without if it stops the system after a checkpoint if it cannot safely reach the next checkpoint with the available energy [6]. Task-based solution facilitates static reasoning about the behavior of the system, but may induce Cost-optimal timed trace synthesis for scheduling of intermittent embedded systems inefficient use of energy (with rollback) or raise the issue of estimating the available energy (without rollback).…”