System and method for optimizing dispatch latency of tasks in a data
processing system
    1.
    发明授权
    System and method for optimizing dispatch latency of tasks in a data processing system 失效
    优化数据处理系统任务调度延迟的系统和方法

    公开(公告)号:US6021425A

    公开(公告)日:2000-02-01

    申请号:US862888

    申请日:1992-04-03

    IPC分类号: G06F9/48 G06F9/44

    CPC分类号: G06F9/4881

    摘要: The invention provides a system and method of enhancing efficiency in a data processing system having a processor, a memory, and a multitasking operating system for managing the processor and the memory. A normal and an expedited scheduling path are provided for scheduling tasks on the processor. The tasks are each assigned a priority for execution on the processor. A queue is provided for the placement of tasks ready for execution. Upon entry into the ready-to-run queue, the execution priority of the new task is compared to the execution priority of the executing task. Responsive to the new task holding a higher execution priority or to absence of an executing task, the expedited scheduling path is invoked. Otherwise the normal scheduling path is invoked.

    摘要翻译: 本发明提供一种提高具有处理器,存储器和用于管理处理器和存储器的多任务操作系统的数据处理系统的效率的系统和方法。 提供了正常和加速的调度路径,用于在处理器上调度任务。 这些任务各自被分配在处理器上执行的优先级。 提供了一个队列来放置可以执行的任务。 进入准备运行队列后,将新任务的执行优先级与执行任务的执行优先级进行比较。 响应于具有较高执行优先级或不执行任务的新任务,调用加速调度路径。 否则调用正常的调度路径。

    Method of integrated system load control through dynamic time-slicing in
a virtual storage environment
    2.
    发明授权
    Method of integrated system load control through dynamic time-slicing in a virtual storage environment 失效
    在虚拟存储环境中通过动态时间分片进行集成系统负载控制的方法

    公开(公告)号:US5386561A

    公开(公告)日:1995-01-31

    申请号:US861221

    申请日:1992-03-31

    IPC分类号: G06F9/48 G06F9/38

    CPC分类号: G06F9/4881

    摘要: A priority preemptive, time distribution operating system operating in accordance with the method of the present invention provides enhanced efficiency in operation of a data processing system having a central processing unit, a computer memory and an auxiliary memory. The operating system manages the central processing unit, the computer memory and the auxiliary memory where a plurality of processes are presented for execution on the central processing unit by: ordering the processes for execution; allocating a predetermined time slice for execution of a process from the ordering; executing a first current process in the ordering on the central processing unit for the predetermined time slice; responsive to occurrence of a storage access operation during execution of the first current process, allocating a supplemental time slice to the current process for execution; and continuing execution of the current process for the supplemental time slice.

    摘要翻译: 根据本发明的方法操作的优先抢占式时间分配操作系统提供具有中央处理单元,计算机存储器和辅助存储器的数据处理系统的操作的增强的效率。 操作系统管理中央处理单元,计算机存储器和辅助存储器,其中多个处理被呈现以在中央处理单元上执行:排序执行过程; 从所述排序中分配用于执行处理的预定时间片; 在所述中央处理单元的订购中执行所述预定时间片的第一当前处理; 响应于执行第一当前进程期间的存储访问操作的发生,向当前进程分配补充时间片以供执行; 并继续执行补充时间片的当前进程。

    Method and system for utilizing benign fault occurrence to measure
interrupt-blocking times
    3.
    发明授权
    Method and system for utilizing benign fault occurrence to measure interrupt-blocking times 失效
    利用良性故障发生测量中断时间的方法和系统

    公开(公告)号:US5301312A

    公开(公告)日:1994-04-05

    申请号:US748149

    申请日:1991-08-21

    CPC分类号: G06F11/3616 G06F2201/865

    摘要: A method in a computer system for monitoring time intervals during which external interrupts are inhibited within the computer system in a selected program being run on the computer system. The method and apparatus of the present invention includes identifying a first plurality of instructions, capable of blocking external interrupts and identifying a second plurality of instructions, capable of unblocking external interrupts. After identifying these instructions, a unique benign fault is inserted proximate to selected ones of the first plurality of instructions and selected ones of the second plurality of instructions to produce a special version program. The special version program is then run in the computer system. Faults which occur during the running of the special version program and associated external interrupt-blocking times are monitored.

    摘要翻译: 一种计算机系统中的方法,用于监视在所述计算机系统中在计算机系统上运行的所选程序中的外部中断被禁止的时间间隔。 本发明的方法和装置包括识别能够阻止外部中断并识别能够解除外部中断的第二多个指令的第一多个指令。 在识别这些指令之后,将独特的良性故障插入到第一组多个指令中的选定的指令和第二组指令中的选定的指令以产生特殊的版本程序。 然后在计算机系统中运行特殊版本程序。 监视在特殊版本程序运行期间发生的故障和相关的外部中断阻塞时间。

    System and method for handling a segmented program in a memory for a
multitasking data processing system utilizing paged virtual storage
    4.
    发明授权
    System and method for handling a segmented program in a memory for a multitasking data processing system utilizing paged virtual storage 失效
    用于处理利用分页虚拟存储的多任务数据处理系统的存储器中的分段程序的系统和方法

    公开(公告)号:US5539899A

    公开(公告)日:1996-07-23

    申请号:US862911

    申请日:1992-04-03

    IPC分类号: G06F12/10 G06F12/02

    CPC分类号: G06F12/109

    摘要: A system and method for loading a segmented program into a memory for a multitasking data processing system utilizing paged virtual storage for processes is disclosed. A virtual address space having a plurality of pages for the segmented program is established. Segments of the segmented program suitable for packing end to end across page boundaries are identified. A virtual storage object over a contiguous portion of the virtual address space is defined and the segments identified for packing into the virtual storage object are mapped into the contiguous portion. Any segments not suitable for packing across page boundaries start on page boundaries in the virtual address space. Then segments corresponding to selected pages of the virtual address space are demand loaded into frames of the memory in the same manner in which they are addressed into the virtual address space.

    摘要翻译: 公开了一种用于将分段程序加载到多任务数据处理系统的存储器中的系统和方法,所述多任务数据处理系统利用用于进程的分页虚拟存储。 建立具有分段程序的多页的虚拟地址空间。 识别分段程序的分段,适用于跨页边界的端到端打包。 定义虚拟地址空间的连续部分上的虚拟存储对象,并将识别用于打包到虚拟存储对象中的段映射到连续部分。 任何不适合跨页边界打包的片段都从虚拟地址空间的页边界开始。 然后,与虚拟地址空间的选定页面相对应的段需要以与将它们寻址到虚拟地址空间相同的方式加载到存储器的帧中。