发明名称 Method and apparatus for synchronous processing based on multi-core system
摘要 Embodiments of the present invention relate to the field of communications network technologies and provide a method and an apparatus for synchronization processing based on a multi-core system, which can improve efficiency in system scheduling and consume fewer resources. According to the solutions provided in the present invention, an initialization setting is sent by any processing device in a first group of processing devices that synchronously process a same current task and initialization is performed; then a notification message sent by any processing device in the first group of processing devices is received and 1 is subtracted from a value of a counting semaphore; and when the value of the counting semaphore is 0, a control message is sent to a second group of processing devices through a message sending interface. The solutions provided in the present invention are applicable to processing synchronization and communication between multiple modules.
申请公布号 US9424101(B2) 申请公布日期 2016.08.23
申请号 US201314077421 申请日期 2013.11.12
申请人 Huawei Technologies Co., Ltd. 发明人 Du Xuefeng
分类号 G06F9/52;G06F9/50;G06F9/46 主分类号 G06F9/52
代理机构 Conley Rose, P.C. 代理人 Conley Rose, P.C. ;Rodolph Grant;Beaulieu Nicholas K.
主权项 1. A method for synchronous processing based on a multi-core system, comprising: receiving an initialization setting sent by any processing device in a first group of processing devices that synchronously process a same current task; performing initialization, wherein the initialization setting comprises setting a value of a counting semaphore of a current multi-core synchronous processing unit to M and setting content of a control message for the current multi-core synchronous processing unit to enable a second group of processing devices, and wherein M is the number of times that the first group of processing devices synchronously process the current task; receiving a notification message sent by any processing device in the first group of processing devices; subtracting one accordingly from the value of the counting semaphore, wherein content of the notification message is that a processing device sending the notification message completes the current task; and when the value of the counting semaphore is zero, sending the control message to the second group of processing devices through a message sending interface according to the set content of the control message for the current multi-core synchronous processing unit to enable the second group of processing devices such that the second group of processing devices processes the current task, wherein receiving the initialization setting sent by any processing device in the first group of processing devices that synchronously process the same current task comprises receiving, through any channel of an internal cache, the initialization setting sent by any processing device in the first group of processing devices that synchronously process the same current task, wherein performing initialization comprises reading the internal cache corresponding to the channel, updating content in the internal cache to the initialization setting, and saving the initialization setting, and wherein receiving the notification message sent by any processing device in the first group of processing devices and subtracting one accordingly from the value of the counting semaphore comprises: receiving, any channel of the internal cache, the notification message sent by any processing device in the first group of processing devices;reading the internal cache corresponding to the channel and acquiring a current value of the counting semaphore in the internal cache; andsubtracting one from the current value of the counting semaphore and saving the current value of the counting semaphore in the internal cache.
地址 Shenzhen CN