Lesser Bear - a light-weight process library for SMP computers

Citation
H. Oguma et Y. Nakayama, Lesser Bear - a light-weight process library for SMP computers, CONCURR COM, 13(12), 2001, pp. 1107-1120
Citations number
13
Categorie Soggetti
Computer Science & Engineering
Journal title
CONCURRENCY AND COMPUTATION-PRACTICE & EXPERIENCE
ISSN journal
15320626 → ACNP
Volume
13
Issue
12
Year of publication
2001
Pages
1107 - 1120
Database
ISI
SICI code
1532-0626(200110)13:12<1107:LB-ALP>2.0.ZU;2-X
Abstract
We have designed and implemented a light-weight process (thread) library ca lled 'Lesser Bear' for SNIP computers. Lesser Bear has high portability and thread-level parallelism. Creating UNIX processes as virtual processors an d a memory-mapped file as a huge shared-memory space enables Lesser Bear to execute threads in parallel. Lesser Bear requires exclusive operation betw een peer virtual processors, and treats a shared-memory space as a critical section for synchronization of threads. Therefore, thread functions of the previous Lesser Bear are serialized. In this paper, we present a schedulin g mechanism to execute thread functions in parallel. In the design of the p roposed mechanism, we divide the entire shared-memory space into partial sp aces for virtual processors, and prepare two queues (Protect Queue and Waiv er Queue) for each partial space. We adopt an algorithm in which lock opera tions are not necessary for enqueueing. This algorithm allows us to propose a scheduling mechanism that can reduce the scheduling overhead. The mechan ism is applied to Lesser Bear and evaluated by experimental results. Copyri ght (C) 2001 John Wiley & Sons, Ltd.