Skip to content

Latest commit

 

History

History
184 lines (133 loc) · 6.71 KB

06-1-spoc-discussion.md

File metadata and controls

184 lines (133 loc) · 6.71 KB

调度算法概念(lec 15) spoc 思考题

  • 有"spoc"标记的题是要求拿清华学分的同学要在实体课上完成，并按时提交到学生对应的ucore_code和os_exercises的git repo上。

个人思考题

处理机调度概念

  • 如何判断操作系统是否是可抢先的？
  • 当操作系统的处理机调度导致线程切换时，暂停进程的当前指令指针可能在什么位置？用户态代码或内核代码？给出理由和实例。

调度准则

先来先服务、短进程优先和最高响应比优先调度算法

  • FIFO、SPN、HRRN算法的思路？
  • 如何调试你的调度算法？

时间片轮转、多级反馈队列、公平共享调度算法和ucore调度框架

  • RR、MLFQ和FSS算法的思路？
  • RR算法选择时间片长度的依据有哪些？
  • 定义调度算法接口需要考虑哪些因素？
  • 尝试跟踪就绪进程选择、进程切换和调度算法的选择依据。

实时调度和多处理器调度

优先级反置

  • 什么是优先级继承(Priority Inheritance)和优先级天花板协议(priority ceiling protocol)？它们的区别是什么？

小组练习与思考题

(1)(spoc) 理解并完善调度算法

实现３种调度算法（SJF，FIFO，RR），可基于python, ruby, C, C++，LISP等）模拟实现，并给出测试。请参考scheduler-homework.py代码或独自实现。 最后统计采用不同调度算法的每个任务的相关时间和总体的平均时间： 　- turnaround time　周转时间 　- response time 响应时间 　- wait time　等待时间

对模拟环境的抽象

  • 任务/进程，及其执行时间 Job 0 (length = 1) Job 1 (length = 4) Job 2 (length = 7)

  • 何时切换？

  • 如何统计？

执行结果

采用FIFO调度算法

 ./scheduler-homework.py  -p FIFO
ARG policy FIFO
ARG jobs 3
ARG maxlen 10
ARG seed 0

Here is the job list, with the run time of each job: 
  Job 0 ( length = 9 )
  Job 1 ( length = 8 )
  Job 2 ( length = 5 )


** Solutions **

Execution trace:
  [ time   0 ] Run job 0 for 9.00 secs ( DONE at 9.00 )
  [ time   9 ] Run job 1 for 8.00 secs ( DONE at 17.00 )
  [ time  17 ] Run job 2 for 5.00 secs ( DONE at 22.00 )

Final statistics:
  Job   0 -- Response: 0.00  Turnaround 9.00  Wait 0.00
  Job   1 -- Response: 9.00  Turnaround 17.00  Wait 9.00
  Job   2 -- Response: 17.00  Turnaround 22.00  Wait 17.00

  Average -- Response: 8.67  Turnaround 16.00  Wait 8.67

采用SJF调度算法

 ./scheduler-homework.py  -p SJF
ARG policy SJF
ARG jobs 3
ARG maxlen 10
ARG seed 0

Here is the job list, with the run time of each job: 
  Job 0 ( length = 9 )
  Job 1 ( length = 8 )
  Job 2 ( length = 5 )


** Solutions **

Execution trace:
  [ time   0 ] Run job 2 for 5.00 secs ( DONE at 5.00 )
  [ time   5 ] Run job 1 for 8.00 secs ( DONE at 13.00 )
  [ time  13 ] Run job 0 for 9.00 secs ( DONE at 22.00 )

Final statistics:
  Job   2 -- Response: 0.00  Turnaround 5.00  Wait 0.00
  Job   1 -- Response: 5.00  Turnaround 13.00  Wait 5.00
  Job   0 -- Response: 13.00  Turnaround 22.00  Wait 13.00

  Average -- Response: 6.00  Turnaround 13.33  Wait 6.00

采用RR调度算法

 ./scheduler-homework.py  -p RR
ARG policy RR
ARG jobs 3
ARG maxlen 10
ARG seed 0

Here is the job list, with the run time of each job: 
  Job 0 ( length = 9 )
  Job 1 ( length = 8 )
  Job 2 ( length = 5 )


** Solutions **

Execution trace:
  [ time   0 ] Run job   0 for 1.00 secs
  [ time   1 ] Run job   1 for 1.00 secs
  [ time   2 ] Run job   2 for 1.00 secs
  [ time   3 ] Run job   0 for 1.00 secs
  [ time   4 ] Run job   1 for 1.00 secs
  [ time   5 ] Run job   2 for 1.00 secs
  [ time   6 ] Run job   0 for 1.00 secs
  [ time   7 ] Run job   1 for 1.00 secs
  [ time   8 ] Run job   2 for 1.00 secs
  [ time   9 ] Run job   0 for 1.00 secs
  [ time  10 ] Run job   1 for 1.00 secs
  [ time  11 ] Run job   2 for 1.00 secs
  [ time  12 ] Run job   0 for 1.00 secs
  [ time  13 ] Run job   1 for 1.00 secs
  [ time  14 ] Run job   2 for 1.00 secs ( DONE at 15.00 )
  [ time  15 ] Run job   0 for 1.00 secs
  [ time  16 ] Run job   1 for 1.00 secs
  [ time  17 ] Run job   0 for 1.00 secs
  [ time  18 ] Run job   1 for 1.00 secs
  [ time  19 ] Run job   0 for 1.00 secs
  [ time  20 ] Run job   1 for 1.00 secs ( DONE at 21.00 )
  [ time  21 ] Run job   0 for 1.00 secs ( DONE at 22.00 )

Final statistics:
  Job   0 -- Response: 0.00  Turnaround 22.00  Wait 13.00
  Job   1 -- Response: 1.00  Turnaround 21.00  Wait 13.00
  Job   2 -- Response: 2.00  Turnaround 15.00  Wait 10.00

  Average -- Response: 1.00  Turnaround 19.33  Wait 12.00

(２)扩展练习1:理解并实现MLFQ调度算法　可基于python, ruby, C, C++，LISP等）模拟实现，并给出测试，在试验报告写出设计思路和测试结果分析。

(3)扩展练习2:理解并实现stride调度算法　可基于python, ruby, C, C++，LISP等）模拟实现，并给出测试，在试验报告写出设计思路和测试结果分析。

(4)扩展练习３：理解并实现EDF, RM实时调度算法和优先级反置方法　可基于python, ruby, C, C++，LISP等）模拟实现，并给出测试，在试验报告写出设计思路和测试结果分析。