Real-time process synchronization for systems with accelerators

Kuan Chung Huang; Yu Chen Wu; Che Wei Chang; Tei Wei Kuo; Chi Sheng Shih; Qingxu Deng

doi:10.1145/2663761.2664220

Real-time process synchronization for systems with accelerators

Kuan Chung Huang^*, Yu Chen Wu, Che Wei Chang, Tei Wei Kuo, Chi Sheng Shih, Qingxu Deng

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

This work is motivated by the needs to manage the priority inversion problem without sacrificing the utilization of increasingly popular hardware accelerators. A new mechanism is developed to dedicate accelerators to selected higher-priority tasks. The floor and ceiling priorities of accelerators are thus proposed as an extension of the concept of semaphore priority ceiling to guarantee at most two priority inversions for any real-time task in a uniprocessor system with multiple accelerators. The properties of the proposed concept are explored with respect to blocking behaviors over the CPU and accelerators and are verified by a series of experiments, for which the insight of the simple but effective idea is evaluated and presented.

Original language	English
Title of host publication	Proceedings of the 2014 Research in Adaptive and Convergent Systems, RACS 2014
Publisher	Association for Computing Machinery
Pages	350-355
Number of pages	6
ISBN (Electronic)	9781450330602
DOIs	https://doi.org/10.1145/2663761.2664220
State	Published - 05 10 2014
Externally published	Yes
Event	2014 Conference on Research in Adaptive and Convergent Systems, RACS 2014 - Towson, United States Duration: 05 10 2014 → 08 10 2014

Publication series

Name	Proceedings of the 2014 Research in Adaptive and Convergent Systems, RACS 2014

Conference

Conference	2014 Conference on Research in Adaptive and Convergent Systems, RACS 2014
Country/Territory	United States
City	Towson
Period	05/10/14 → 08/10/14

Bibliographical note

Publisher Copyright:
© 2014 ACM.

Keywords

Blocking time analysis
Dedicated accelerators
Synchronization protocols

Access to Document

10.1145/2663761.2664220

Cite this

Huang, K. C., Wu, Y. C., Chang, C. W., Kuo, T. W., Shih, C. S., & Deng, Q. (2014). Real-time process synchronization for systems with accelerators. In Proceedings of the 2014 Research in Adaptive and Convergent Systems, RACS 2014 (pp. 350-355). (Proceedings of the 2014 Research in Adaptive and Convergent Systems, RACS 2014). Association for Computing Machinery. https://doi.org/10.1145/2663761.2664220

@inproceedings{ea63dcc76d634692b00cd6914088024c,

title = "Real-time process synchronization for systems with accelerators",

abstract = "This work is motivated by the needs to manage the priority inversion problem without sacrificing the utilization of increasingly popular hardware accelerators. A new mechanism is developed to dedicate accelerators to selected higher-priority tasks. The floor and ceiling priorities of accelerators are thus proposed as an extension of the concept of semaphore priority ceiling to guarantee at most two priority inversions for any real-time task in a uniprocessor system with multiple accelerators. The properties of the proposed concept are explored with respect to blocking behaviors over the CPU and accelerators and are verified by a series of experiments, for which the insight of the simple but effective idea is evaluated and presented.",

keywords = "Blocking time analysis, Dedicated accelerators, Synchronization protocols",

author = "Huang, {Kuan Chung} and Wu, {Yu Chen} and Chang, {Che Wei} and Kuo, {Tei Wei} and Shih, {Chi Sheng} and Qingxu Deng",

note = "Publisher Copyright: {\textcopyright} 2014 ACM.; 2014 Conference on Research in Adaptive and Convergent Systems, RACS 2014 ; Conference date: 05-10-2014 Through 08-10-2014",

year = "2014",

month = oct,

day = "5",

doi = "10.1145/2663761.2664220",

language = "英语",

series = "Proceedings of the 2014 Research in Adaptive and Convergent Systems, RACS 2014",

publisher = "Association for Computing Machinery",

pages = "350--355",

booktitle = "Proceedings of the 2014 Research in Adaptive and Convergent Systems, RACS 2014",

address = "美国",

}

Huang, KC, Wu, YC, Chang, CW, Kuo, TW, Shih, CS & Deng, Q 2014, Real-time process synchronization for systems with accelerators. in Proceedings of the 2014 Research in Adaptive and Convergent Systems, RACS 2014. Proceedings of the 2014 Research in Adaptive and Convergent Systems, RACS 2014, Association for Computing Machinery, pp. 350-355, 2014 Conference on Research in Adaptive and Convergent Systems, RACS 2014, Towson, United States, 05/10/14. https://doi.org/10.1145/2663761.2664220

Real-time process synchronization for systems with accelerators. / Huang, Kuan Chung; Wu, Yu Chen; Chang, Che Wei et al.
Proceedings of the 2014 Research in Adaptive and Convergent Systems, RACS 2014. Association for Computing Machinery, 2014. p. 350-355 (Proceedings of the 2014 Research in Adaptive and Convergent Systems, RACS 2014).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Real-time process synchronization for systems with accelerators

AU - Huang, Kuan Chung

AU - Wu, Yu Chen

AU - Chang, Che Wei

AU - Kuo, Tei Wei

AU - Shih, Chi Sheng

AU - Deng, Qingxu

PY - 2014/10/5

Y1 - 2014/10/5

N2 - This work is motivated by the needs to manage the priority inversion problem without sacrificing the utilization of increasingly popular hardware accelerators. A new mechanism is developed to dedicate accelerators to selected higher-priority tasks. The floor and ceiling priorities of accelerators are thus proposed as an extension of the concept of semaphore priority ceiling to guarantee at most two priority inversions for any real-time task in a uniprocessor system with multiple accelerators. The properties of the proposed concept are explored with respect to blocking behaviors over the CPU and accelerators and are verified by a series of experiments, for which the insight of the simple but effective idea is evaluated and presented.

AB - This work is motivated by the needs to manage the priority inversion problem without sacrificing the utilization of increasingly popular hardware accelerators. A new mechanism is developed to dedicate accelerators to selected higher-priority tasks. The floor and ceiling priorities of accelerators are thus proposed as an extension of the concept of semaphore priority ceiling to guarantee at most two priority inversions for any real-time task in a uniprocessor system with multiple accelerators. The properties of the proposed concept are explored with respect to blocking behaviors over the CPU and accelerators and are verified by a series of experiments, for which the insight of the simple but effective idea is evaluated and presented.

KW - Blocking time analysis

KW - Dedicated accelerators

KW - Synchronization protocols

UR - http://www.scopus.com/inward/record.url?scp=84910013631&partnerID=8YFLogxK

U2 - 10.1145/2663761.2664220

DO - 10.1145/2663761.2664220

M3 - 会议稿件

AN - SCOPUS:84910013631

T3 - Proceedings of the 2014 Research in Adaptive and Convergent Systems, RACS 2014

SP - 350

EP - 355

BT - Proceedings of the 2014 Research in Adaptive and Convergent Systems, RACS 2014

PB - Association for Computing Machinery

T2 - 2014 Conference on Research in Adaptive and Convergent Systems, RACS 2014

Y2 - 5 October 2014 through 8 October 2014

ER -

Real-time process synchronization for systems with accelerators

Abstract

Publication series

Conference

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this