Modular data-race-freedom guarantees in the promising semantics

11Citations
Citations of this article
6Readers
Mendeley users who have this article in their library.
Get full text

Abstract

Local data-race-freedom guarantees, ensuring strong semantics for locations accessed by non-racy instructions, provide a fruitful methodology for modular reasoning in relaxed memory concurrency. We observe that standard compiler optimizations are in inherent conflict with such guarantees in general fully-relaxed memory models. Nevertheless, for a certain strengthening of the promising model by Lee et al. that only excludes relaxed RMW-store reorderings, we establish multiple useful local data-racefreedom guarantees that enhance the programmability aspect of the model.We also demonstrate that the performance price of forbidding these reorderings is insignificant. To the best of our knowledge, these results are the first to identify a model that includes the standard concurrency constructs, supports the efficient mapping of relaxed reads and writes to plain hardware loads and stores, and yet validates several local data-race-freedom guarantees. To gain confidence, our results are fully mechanized in Coq.

Cite

CITATION STYLE

APA

Cho, M., Lee, S. H., Hur, C. K., & Lahav, O. (2021). Modular data-race-freedom guarantees in the promising semantics. In Proceedings of the ACM SIGPLAN Conference on Programming Language Design and Implementation (PLDI) (pp. 867–882). Association for Computing Machinery. https://doi.org/10.1145/3453483.3454082

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Save time finding and organizing research with Mendeley

Sign up for free