微线程的构建基于软中断,用于允许动态生成可延迟函数。
Tasklets are built on top of softirqs to allow dynamic creation of deferrable functions.
这时需要调用函数tasklet_enable来实现微线程可被调度。
This requires that the tasklet_enable function be invoked to make the tasklet schedulable.
来自高优先级矢量的微线程先得到服务,随后是来自正常优先级矢量的微线程。
Tasklets from the high-priority vector are serviced first, followed by those on the normal vector.
工作队列可以有比微线程更高的时延,并为任务延迟提供功能更丰富的API。
Work queues can have higher latency than tasklets but include a richer API for work deferral.
甚至所谓的“轻量级”线程库中的线程也比这里提出的微线程“重”好几个数量级。
Even so-called "lightweight" threading libraries have threads that are orders of magnitude "heavier" than those presented here.
当该模块被卸载,函数tasklet _ kill将被调用来保证微线程不处于可调度状态。
When the module is unloaded, the tasklet_kill function is called to ensure that the tasklet is not in a schedulable state.
最后,微线程生成之后,就可以通过函数tasklet _ kill来停止微线程(参见清单4)。
Finally, after a tasklet has been created, it's possible to stop a tasklet through the tasklet_kill functions (see Listing 4).
微线程提供低延迟机制,该方式简单而直接,而工作队列提供复杂的API来允许对多个任务项目进行排队。
Tasklets provide a low-latency mechanism that is simple and straightforward, while work queues provide a flexible API that permits queuing of multiple work items.
有了微线程,我们就可能在目前中等性能的PC 机上运行数以万计的并行进程,还可以每秒钟几十万次地在上下文之间切换。
With microthreads, it is possible to run tens of thousands of parallel processes on a moderately capable modern PC, and to switch between contexts hundreds of thousands of times every second.
可以分别利用函数tasklet_enable和函数tasklet _disable实现允许和禁止一个微线程(从调度的角度)。
You can enable and disable a tasklet (from a scheduling perspective) using the tasklet_enable and tasklet_disable functions, respectively.
函数tasklet _kill保证微线程不会再运行,并且,如果按进度该微线程应该运行,将会等到它运行完,然后再kill该线程。
The tasklet_kill function ensures that the tasklet will not run again and, if the tasklet is currently scheduled to run, will wait for its completion, and then kill it.
通过软中断机制来调度微线程,当机器处于严重软件中断负荷之下时, 可通过ksoftirqd(一种每CPU内核线程)软中断来调度。
Tasklets are scheduled through the softirq mechanism, sometimes through ksoftirqd (a per-CPU kernel thread), when the machine is under heavy soft-interrupt load.
因此新浪微博通过线程技术允许用户将发表的任何一个单一的信息作为论坛主题进行讨论。
It built in threading features that allow any single message to act as a forum for discussion.
它遵循微核的基本原理,因为它只在特权微核中实现地址空间、线程管理和IPC,以及一些虚拟化功能。
It follows the fundamental principles of microkernels in that it implements address Spaces, thread management, and IPC only in the privileged microkernel along with virtualization capabilities.
Codezero的调度器包含针对来宾线程和微核线程的内核优先(以及用于优先的时间片)。
Codezero's scheduler includes kernel preemption for both guest threads and microkernel threads (in addition to time slices for preemption).
本文介绍了一个专门面向网络协议处理的硬件多线程包处理微引擎NRS05的设计。
This paper proposed a packet processing engine architecture called NRS05, that promotes the efficient dynamic thread scheduling for hiding long latency operations and coping with pipeline stalls.
本文介绍了一个专门面向网络协议处理的硬件多线程包处理微引擎NRS05的设计。
This paper proposed a packet processing engine architecture called NRS05, that promotes the efficient dynamic thread scheduling for hiding long latency operations and coping with pipeline stalls.
应用推荐