发明名称 Method and system for resolving dependency among the enqueued works and/or finished works and scheduling the dependency-resolved works
摘要 A scheduling apparatus and method allocate a plurality of works to a plurality of processing cores by transferring a work having no dependency on the execution completion of another work from a dependency queue to a runnable queue, transferring the work from the runnable queue to an idle one of the processing cores for execution, transferring the work executed by the one processing core to a finish queue, where the work becomes designated a finished work, and transferring a work within the dependency queue, having a dependency upon the execution completion of the finished work, to the runnable queue.
申请公布号 US8806498(B2) 申请公布日期 2014.08.12
申请号 US201113023848 申请日期 2011.02.09
申请人 Samsung Electronics Co., Ltd. 发明人 Seo Sung-Jong;Lee Sung-Hak;Im Dong-Woo;Song Hyo-Jung;Cho Seung-Mo
分类号 G06F9/46 主分类号 G06F9/46
代理机构 NSIP Law 代理人 NSIP Law
主权项 1. A scheduling apparatus to allocate works from applications to cores so that the applications share the cores, comprising: a processor; a runnable queue, using the processor, configured to enqueue therein the works in order of execution in consideration of any dependency therebetween;a finish queue, using the processor, configured to enqueue therein finished works whose execution is complete;one or more dependency queues, using the processor, configured to temporarily enqueue the works therein until the works are enqueued into the runnable queue, wherein each application corresponds to a separate dependency queue;an I/O (input/output) port configured to transmit the works in the runnable queue to the cores and receive the finished works from the cores; anda scheduler configured to search the dependency queues for one or more runnable works with no dependency and enqueue the one or more runnable works with no dependency into the runnable queue, to transmit the one or more works in the runnable queue to the cores via the I/O port so that the transmitted one or more works can be executed by any idle one of the cores, and to receive the finished works from the cores via the I/O port and enqueue the received finished works in the finish queue, whereinthe scheduler is further configured to perform dependency resolving on the works in the dependency queues in real time upon the receipt of the finished works from the cores or at regular intervals of time, andin response to there being one or more works in the dependency queues having dependency on the finished works, the scheduler is further configured to perform dependency resolving on the one or more works having dependency on the finished works and to enqueue the dependency-resolved works into the runnable queue.
地址 Suwon-si KR