Por favor, use este identificador para citar o enlazar este ítem:
https://doi.org/10.1145/3470496.3527385
Twittear
Título: | Free Atomics: Hardware Atomic Operations without Fences |
Fecha de publicación: | 11-jun-2022 |
Editorial: | Association for Computing Machinery |
Cita bibliográfica: | ISCA '22: The 49th Annual International Symposium on Computer Architecture New York New York Pages 14–26 |
ISSN: | 1063-6897 |
ISBN: | 978-1-4503-8610-4 |
Palabras clave: | Multi-core architectures Microarchitecture Atomic Read-Modify-Write instructions Total-Store-Order (TSO) Store-to-load forwarding |
Resumen: | Atomic Read-Modify-Write (RMW) instructions are primitive synchronization operations implemented in hardware that provide the building blocks for higher-abstraction synchronization mechanisms to programmers. According to publicly available documentation, current x86 implementations serialize atomic RMW operations, i.e., the store buffer is drained before issuing atomic RMWs and subsequent memory operations are stalled until the atomic RMW commits. This serialization, carried out by memory fences, incurs a performance cost which is expected to increase with deeper pipelines. This work proposes Free atomics, a lightweight, speculative, deadlock-free implementation of atomic operations that removes the need for memory fences, thus improving performance, while preserving atomicity and consistency. Free atomics is, to the best of our knowledge, the first proposal to enable store-to-load forwarding for atomic RMWs. Free atomics only requires simple modifications and incurs a small area overhead (15 bytes). Our evaluation using gem5-20 shows that, for a 32-core configuration, Free atomics improves performance by 12.5%, on average, for a large range of parallel workloads and 25.2%, on average, for atomic-intensive parallel workloads over a fenced atomic RMW implementation. |
Autor/es principal/es: | Asgharzadeh, Ashkan Cebrian, Juan M. Perais, Arthur Kaxiras, Stefanos Ros, Alberto |
Facultad/Departamentos/Servicios: | Facultades, Departamentos, Servicios y Escuelas::Departamentos de la UMU::Ingeniería de la Información y las Comunicaciones |
Forma parte de: | 49th International Symposium on Computer Architecture (ISCA) |
Versión del editor: | https://dl.acm.org/doi/10.1145/3470496.3527385 |
URI: | http://hdl.handle.net/10201/124764 |
DOI: | https://doi.org/10.1145/3470496.3527385 |
Tipo de documento: | info:eu-repo/semantics/article |
Número páginas / Extensión: | 13 |
Derechos: | info:eu-repo/semantics/openAccess Attribution-NonCommercial-NoDerivatives 4.0 Internacional |
Descripción: | © 2022. The authors. This document is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0 This document is the accepted version of a published work that appeared in final form in ISCA '22: The 49th Annual International Symposium on Computer Architecture New York New York. To access the final work, see DOI: https://doi.org/10.1145/3470496.3527385 |
Aparece en las colecciones: | Artículos: Ingeniería y Tecnología de Computadores |
Ficheros en este ítem:
Fichero | Descripción | Tamaño | Formato | |
---|---|---|---|---|
aasgharzadeh-isca22.pdf | 2,26 MB | Adobe PDF | Visualizar/Abrir |
Este ítem está sujeto a una licencia Creative Commons Licencia Creative Commons