CSCI315

Operating Systems- CPU Scheduling

Dr. Reem Ibrahim

Dr. Reem Ibrahim

Operating System Concepts, Avi Silberschatz, Peter Baer Galvin, Greg Gagne,
John Wiley & Sons, ISBN 978-1-118-06333-0 , 9th Ed., 2012.
CPU Scheduling
 Basic Concepts
 Scheduling Criteria
 Scheduling Algorithms
 Multiple-Processor Scheduling
 Real-Time CPU Scheduling
 Operating Systems Examples

2 Operating System Dr. Reem Ibrahim

CPU Scheduling
 Basic Concepts
 CPU Scheduling is a process of determining which
process will own CPU for execution while another
process is on hold. The main task of CPU scheduling is to
make sure that whenever the CPU remains idle, the OS
at least select one of the processes available in the ready
queue for execution. The selection process will be
carried out by the CPU scheduler.
 CPU–I/O Burst Cycle – Process execution consists of a
cycle of CPU execution and I/O wait
 CPU burst followed by I/O burst
 CPU burst distribution is of main concern
 CPU burst: the amount of time the process uses the processor
before it is no longer ready waiting for some input or
interrupted by some other process
 I/O Burst :It is the amount of time, a process waits for input-
output before needing CPU time.

3 Operating System Dr. Reem Ibrahim

CPU Scheduling
 CPU Scheduler
 Short-term scheduler selects from among the processes
in ready queue, and allocates the CPU to one of them
 Queue may be ordered in various ways
 CPU scheduling decisions may take place when a process:
1. Switches from running to waiting state
2. Switches from running to ready state
3. Switches from waiting to ready
4. Terminates
Preemptive nonPreemptive
Scheduling Scheduling

4 Operating System Dr. Reem Ibrahim

Non-preemptive & Preemptive
Scheduling

5 Operating System Dr. Reem Ibrahim

CPU Scheduling
 Dispatcher
 Dispatcher module gives control of the CPU to the process selected by
the short-term scheduler; this involves:
 Switching context
 Switching to user mode
 Jumping to the proper location in the user program to restart that program
 Dispatch latency – time it takes for the dispatcher to stop one process
and start another running.

6 Operating System Dr. Reem Ibrahim

CPU Scheduling
 Scheduling Criteria
 CPU utilization – keep the CPU as busy as possible  MAX
 Throughput – # of processes that complete their execution per time
unit MAX
 Turnaround time – amount of time to execute a particular
process MIN
 Waiting time – amount of time a process has been waiting in the
ready queue MIN
 Response time – amount of time it takes from when a request was
submitted until the first response is produced, not output (for time-
sharing environment)  MIN

7 Operating System Dr. Reem Ibrahim

Scheduling Algorithms

8 Operating System Dr. Reem Ibrahim

Scheduling Algorithms
 First-Come, First-Served Scheduling
Process Burst Time
P1 24
P2 3
P3 3
 Suppose that the processes arrive in the order: P1 , P2 , P3
 The Gantt Chart for the schedule is:

P1 P2 P3
0 24 27 30

9 Operating System Dr. Reem Ibrahim

Scheduling Algorithms
 First-Come, First-Served Scheduling
Suppose that the processes arrive in the order:
P2 , P3 , P1
 The Gantt chart for the schedule is:

P2 P3 P1
0 3 6 30

 One of the major drawbacks of this scheme is that the waiting time and
the average turnaround time is often quite long.
10 Operating System Dr. Reem Ibrahim
Scheduling Algorithms
 Shortest-Job-First (SJF) Scheduling
 Associate with each process the length of its next CPU burst
 Use these lengths to schedule the process with the shortest time
 SJF is optimal – gives minimum average waiting time for a given set of
processes
 The difficulty is knowing the length of the next CPU request
 Could ask the user
 Can be done by using the length of previous CPU bursts, using exponential
averaging
1. t n  actual length of n CPU burst
th

2.  n 1  predicted value for the next CPU burst

3.  , 0    1
4. Define :  n 1   t n  1    n .

 Commonly, α set to ½

11 Operating System Dr. Reem Ibrahim

Scheduling Algorithms
 Shortest-Job-First (SJF) Scheduling

 n 1   t n  1    n .

 =0
n+1 = n
Recent history does not
count
 =1
n+1 =  tn
Only the actual last CPU
burst counts

12 Operating System Dr. Reem Ibrahim

Scheduling Algorithms
 Shortest-Job-First (SJF) Scheduling (Non-Preemptive)
ProcessArriva l Time Burst Time
P1 0.0 6
P2 2.0 8
P3 4.0 7
P4 5.0 3
 SJF scheduling chart
P4 P1 P3 P2
0 3 9 16 24

13 Operating System Dr. Reem Ibrahim

Scheduling Algorithms
 Shortest-Job-First (SJF) Scheduling
 Preemptive version called shortest-remaining-time-first
 Example of Shortest-remaining-time-first
 Adding the concepts of varying arrival times and preemption to the analysis
ProcessAarri Arrival TimeT Burst Time
P1 0 8
P2 1 4
P3 2 9
P4 3 5
 Preemptive SJF Gantt Chart

P1 P2 P4 P1 P3
0 1 5 10 17 26

14 Operating System Dr. Reem Ibrahim

Scheduling Algorithms
 Shortest-Job-First (SJF) Scheduling(Preemptive)
P1 P2 P4 P1 P3
0 1 5 10 17 26

15 Operating System Dr. Reem Ibrahim

Scheduling Algorithms
 Priority Scheduling
 A priority number (integer) is associated with each process

 The CPU is allocated to the process with the highest priority (smallest
integer  highest priority)
 Preemptive
 Nonpreemptive

 SJF is priority scheduling where priority is the inverse of predicted next CPU
burst time

 Problem  Starvation – low priority processes may never execute

 Solution  Aging – as time progresses increase the priority of the process

16 Operating System Dr. Reem Ibrahim

Scheduling Algorithms
 Priority Scheduling
 Example of Priority Scheduling
 Priority scheduling Gantt Chart

17 Operating System Dr. Reem Ibrahim

Scheduling Algorithms
 Round Robin (RR) Scheduling
 Each process gets a small unit of CPU time (time quantum q),
usually 10-100 milliseconds. After this time has elapsed, the
process is preempted and added to the end of the ready queue.
 If there are 𝑛 processes in the ready queue and the time
quantum is 𝑞, then each process gets 1/𝑛 of the CPU time in
chunks of at most 𝑞 time units at once.
 No process waits more than (𝑛−1)𝑞time units.
 Timer interrupts every quantum to schedule next process
 Performance
 q large  FIFO
 q small  q must be large with respect to context switch, otherwise
overhead is too high
18 Operating System Dr. Reem Ibrahim
Scheduling Algorithms
 Round Robin (RR) Scheduling
The Gantt chart is:

P1 P2 P3 P1 P1 P1 P1 P1
0 4 7 10 14 18 22 26 30

19 Operating System Dr. Reem Ibrahim

Scheduling Algorithms
 Round Robin (RR) Scheduling
Typically, higher average turnaround than SJF, but better response
 q should be large compared to context switch time
 q usually 10ms to 100ms, context switch < 10 usec

Time Quantum and Context Switch Time

20 Operating System Dr. Reem Ibrahim

Scheduling Algorithms
 Multilevel Queue
 Ready queue is partitioned into separate queues, eg:
 foreground (interactive)
 background (batch)
 Process permanently in a given queue
 Each queue has its own scheduling algorithm:
 foreground – RR
 background – FCFS
 Scheduling must be done between the queues:
 Fixed priority scheduling; (i.e., serve all from foreground then from background).
Possibility of starvation.
 Time slice – each queue gets a certain amount of CPU time which it can schedule
amongst its processes; i.e., 80% to foreground in RR, 20% to background in FCFS.

21 Operating System Dr. Reem Ibrahim

Scheduling Algorithms
 Multilevel Feedback Queue
 A process can move between the various queues; aging can
be implemented this way
 Multilevel-feedback-queue scheduler defined by the
following parameters:
 number of queues
 scheduling algorithms for each queue
 method used to determine when to upgrade a process
 method used to determine when to demote a process
 method used to determine which queue a process will enter
when that process needs service

22 Operating System Dr. Reem Ibrahim

Scheduling Algorithms
 Multilevel Queue
 Example of Multilevel Feedback Queue
 Three queues:
 Q0 – RR with time quantum 8 milliseconds
 Q1 – RR time quantum 16 milliseconds
 Q2 – FCFS

 Scheduling
 A new job enters queue Q0 which is served FCFS
 When it gains CPU, job receives 8 milliseconds
 If it does not finish in 8 milliseconds, job is moved to queue Q1
 At Q1 job is again served FCFS and receives 16 additional milliseconds
 If it still does not complete, it is preempted and moved to queue Q2

23 Operating System Dr. Reem Ibrahim

Scheduling Algorithms
 Example of Multilevel Feedback Queue

24 Operating System Dr. Reem Ibrahim

Multiple-Processor Scheduling
 CPU scheduling more complex when
multiple CPUs are available
 Homogeneous processors within a
multiprocessor
 Asymmetric multiprocessing – only
one processor accesses the system data
structures, alleviating the need for data Asymmetric multiprocessing
sharing (Master-Slave)
 Symmetric multiprocessing (SMP) –
each processor is self-scheduling, all
processes in common ready queue, or
each has its own private queue of ready
processes
Symmetric multiprocessing
 Currently, most common

25 Operating System Dr. Reem Ibrahim

Multiple-Processor Scheduling
 Processor affinity – process has affinity for
processor on which it is currently running.
 Try to avoid migration of processes from one
processor to another and try to keep a
process running on the same processor. This is
known as Processor affinity.
 Soft affinity : refers to an operating
system's policy of keeping a process
running on the same processor, and it will
not ensure that it will do so.
 Hard affinity : It enables a process to
NUMA and CPU Scheduling
define a subset of processors on which it
may execute.
 The main-memory architecture of a system
can affect processor affinity issues.
NUMA : non-uniform memory access.

26 Operating System Dr. Reem Ibrahim

Multiple-Processor Scheduling
 If SMP, need to keep all CPUs loaded for efficiency.
 Load Balancing is the phenomena which keeps
the workload evenly distributed across all processors in an SMP system.
 Push migration – periodic task checks load on each processor, and if found
pushes task from overloaded CPU to other CPUs
 Pull migration – idle processors pulls waiting task from busy processor

27 Operating System Dr. Reem Ibrahim

Multicore Processors
 Recent trend to place multiple
processor cores on same physical chip
 Faster and consumes less power
 Multiple threads per core also
growing
 Takes advantage of memory stall to make
progress on another thread while memory
retrieve happens.
 When processor accesses memory then it
spends a significant amount of time waiting
for the data to become available. This
situation is called MEMORY STALL

Multithreaded Multicore System

28 Operating System Dr. Reem Ibrahim
Real-Time CPU Scheduling
 Real-time systems are systems that
carry real-time tasks. These tasks need
to be performed immediately with a
certain degree of urgency.
 Soft real-time systems – no
guarantee as to when critical real-time
process will be scheduled
 Hard real-time systems – task
must be serviced by its deadline
 Two types of latencies affect
performance
1. Interrupt latency – time from arrival of
interrupt to start of routine that services
interrupt
2. Dispatch latency – time for schedule to take
current process off CPU and switch to another

29 Operating System Dr. Reem Ibrahim

Real-Time CPU Scheduling
 Priority-based Scheduling
 For real-time scheduling, scheduler must support preemptive, priority-
based scheduling
 But only guarantees soft real-time
 For hard real-time must also provide ability to meet deadlines
 Processes have new characteristics: periodic ones require CPU at
constant intervals
 Has processing time t, deadline d, period p
 0≤t≤d≤p
 Rate of periodic task is 1/p

30 Operating System Dr. Reem Ibrahim

Real-Time CPU Scheduling
 Rate-Monotonic Scheduling
 A priority is assigned based on the inverse of its period
 Shorter periods = higher priority;
 Longer periods = lower priority
 P1 is assigned a higher priority than P2.

 Example: We have two processes, P1 and P2. The periods for P1 and P2
are 50 and 100, respectively—that is, p1 = 50 and p2 = 100. The processing
times are t1 = 20 for P1 and t2 = 35 for P2. The deadline for each process
requires that it complete its CPU burst by the start of its next period

31 Operating System Dr. Reem Ibrahim

Real-Time CPU Scheduling
 Rate-Monotonic Scheduling
 Another Example: Assume that process P1 has a period of p1
= 50 and a CPU burst of t1 = 25. For P2, the corresponding
values are p2 = 80 and t2 = 35
 Missed the deadline for P2 although the total CPU utilization
of the two processes is (25/50)+(35/80) = 0.94

32 Operating System Dr. Reem Ibrahim

Real-Time CPU Scheduling
 Earliest Deadline First Scheduling (EDF)
 Priorities are assigned according to deadlines:
 the earlier the deadline, the higher the priority;
 the later the deadline, the lower the priority

 Process P1 has the earliest deadline, so its initial priority is

higher than that of process P2.
 Whereas rate-monotonic scheduling allows P1 to preempt P2
at the beginning of its next period at time 50, EDF scheduling
allows process P2 to continue running.
33 Operating System Dr. Reem Ibrahim
Operating System Examples
 Windows scheduling
 Windows uses priority-based preemptive scheduling
 Highest-priority thread runs next
 Dispatcher is scheduler
 Thread runs until (1) calls a blocking system call, such as for I/O, (2) uses time slice, (3)
preempted by higher-priority thread
 Real-time threads can preempt non-real-time
 32-level priority scheme
 Variable class is 1-15, real-time class is 16-31
 Priority 0 is memory-management thread
 Queue for each priority
 If no run-able thread, runs idle thread
 Windows API identifies several priority classes to which a process can belong
 REALTIME_PRIORITY_CLASS, HIGH_PRIORITY_CLASS,
ABOVE_NORMAL_PRIORITY_CLASS,NORMAL_PRIORITY_CLASS,
BELOW_NORMAL_PRIORITY_CLASS, IDLE_PRIORITY_CLASS
 All are variable except REALTIME
34 Operating System Dr. Reem Ibrahim
Operating System Examples
 Windows scheduling
 A thread within a given priority class has a relative priority
 TIME_CRITICAL, HIGHEST, ABOVE_NORMAL, NORMAL, BELOW_NORMAL, LOWEST,
IDLE
 Priority class and relative priority combine to give numeric priority
 Base priority is NORMAL within the class
 If quantum expires, priority lowered, but never below base
 If wait occurs, priority boosted depending on what was waited for
 Foreground window given 3x priority boost
 Windows 7 added User-Mode Scheduling (UMS)
 Applications create and manage threads independent of kernel
 For large number of threads, much more efficient
 UMS schedulers come from programming language libraries like
C++ Concurrent Runtime (ConcRT) framework

35 Operating System Dr. Reem Ibrahim

Operating System Examples
 Windows scheduling
 The priority of each thread is based on both the priority class it belongs to and its relative
priority within that class. The values of the priority classes appear in the top row. The left
column contains the values for the relative priorities. For example, if the relative priority
of a thread in the ABOVE NORMAL PRIORITY CLASS is NORMAL, the numeric priority of that
thread is 10.
Priority classes
 NORMAL relative priority
 for each priority class are as follows:
Values for the relative priorities
 REALTIME PRIORITY CLASS—24
 HIGH PRIORITY CLASS—13
 ABOVE NORMAL PRIORITY CLASS—10
 NORMAL PRIORITY CLASS—8
 BELOW NORMAL PRIORITY CLASS—6
 IDLE PRIORITY CLASS—4

Windows thread priorities

36 Operating System Dr. Reem Ibrahim