CISC3595
CPU Scheduling
Basic Concepts
 Multiprogramming
Load multiple program into memory, and
run them simultaneously
 A process (or kernel-level thread) is
executed until it must wait (for I/O
completion, child exiting…), OS then
switches to run another process (or
thread)

 Goal: improve
2
CPU utilization
Process Scheduling
 Short-term

scheduling (CPU scheduling)
Select one process from ready queue to be
executed
 Medium-term
scheduling, Long-term
scheduling
 What is ready queue ?
Contains PCB for all processes that lines up
waiting for CPU
 Not necessarily FIFO queue

3
Process Control Block (PCB): review
Information associated with each process


Process state
Program counter:


CPU registers:


Amount of CPU time used …
I/O status information

4
4
Value of base and limit registers, page/segmentation tables (to be
studied later)
Accounting information


Process priority, …
Memory-management information


Contents of CPU registers
CPU scheduling information


address of next instruction to be executed
I/O device allocated, list of open files…
Process Control Block (PCB): review
5
5
Context Switch: review
6
6
P0, P1 are running concurrently, i.e.,
their executions are interleaved.
Process Scheduling Queues: review

OS maintain multiple process queues for scheduling
purpose:




Job queue – set of all processes in the system
Ready queue – set of all processes residing in main
memory, ready and waiting to execute
Device queues – set of processes waiting for an I/O
device
Processes migrate among the various queues
7
7
Ready Queue And I/O Device Queues: review
8
8
CPU and I/O Bursts
• Process execution
consists of a cycle of CPU
execution and I/O wait
• CPU burst:
• Data processing,
calculation
• E.g. sorting, searching
• I/O burst:
• Read from I/O device
• File system access
•What about the myCopy
program?
9
Histogram of CPU-burst Times
A large number
short CPU burst
Small number of
long CPU burst
10
CPU Scheduler


Selects one process from ready queue, and
allocates CPU to execute it
When to make CPU scheduling :
1. a process switches from running to waiting state
2. a process switches from running to ready state
3. a Process switches from waiting to ready
4. a process terminates

Nonpreemptive scheme: only run CPU scheduling
under 1 and 4


i.e., current running process gives up CPU
Otherwise, preemptive scheduling scheme
11
Dispatcher module


Scheduler => Dispatcher
Dispatcher module: gives control of CPU to the process
selected by short-term scheduler




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

12
Including context switching latency
CPU Scheduling Algorithms
First-come, First-Served Scheduling
 Shortest-Job-First Scheduling
 Priority Scheduling
 Round-Robin Scheduling
 Multilevel Queue Scheduling
 Multilevel Feedback Queue Scheduling

13
CPU Scheduling Criteria



CPU utilization – percentage of time CPU is busy (not idle)
Throughput – # of processes that complete their execution
per time unit
Turnaround time – amount of time to execute a particular
process



From process submission to time of completion
Waiting time – amount of time a process has been waiting
in ready queue
Response time – amount of time it takes from when a
request was submitted until first response is produced (for
time-sharing environment)
14
Optimization Criteria






Maximize CPU utilization
Maximize throughput
Minimize turnaround time
Minimize waiting time
Minimize response time
Achieved performance measure is random variable
depending on the statistics of workload



15
Usually, optimize average measure (average case performance)
Sometimes, optimize minimum or maximum value (worst case
performance)
Or sometimes, variance of the measure
First-Come First-Served Scheduling
Serve processes in the order of arrivals, nonpreemptive
Process Burst Time
P1
24
P2
3
P3
3

Suppose that processes arrive in the order: P1 , P2 , P3
Gantt Chart for FCFS schedule is:
P1
0


16
P2
24
P3
27
Waiting time for P1 = 0; P2 = 24; P3 = 27
Average waiting time: (0 + 24 + 27)/3 = 17
30
FCFS Scheduling (Cont.)
Process Burst Time
P1
24
P2
3
P3
3
Suppose that processes arrive in the order P2 , P3 , P1
 The Gantt chart for FCFS schedule is:
P2
0



3
P1
6
Waiting time for P1 = 6;P2 = 0; P3 = 3
Average waiting time: (6 + 0 + 3)/3 = 3
Much better than previous case

17
P3
Short process behind long process
30
Exercise

Draw Gantt chart for FCFS schedule for the following set
of processes arrived in the order of P1, P2, P3, P4.
Process
P1
P2
P3
P4
Burst Time
10
4
6
5
Average waiting time
What order of arrival of these processes will give smallest
average waiting time ?
18
FCFS Scheduling

What will happen if the system has



A Convoy effect happens when




one CPU-bound process
many I/O-bound processes
a set of processes need to use a resource for a short time
one process holds the resource for a long time, blocking all
of the other processes
Result: poor utilization of the other resources in the system.
Other problems of FCFS ?

19
Ill-suited for time-sharing systems (interactive environment)
Shortest-Job-First (SJF) Scheduling


Schedule process with shortest-next-CPU-burst (i.e., the
length/duration of next CPU burst)
Two schemes:


20
nonpreemptive – once CPU given to the process it cannot be
preempted until completes its CPU burst
preemptive – if a new process arrives with CPU burst length
less than remaining time of current executing process, preempt.
Also called Shortest-Remaining-Time-First (SRTF)
Shortest-Job-First (SJF) Scheduling

Among non-preemptive scheduling algorithms, SJF
achieves minimum average waiting time for a given set of
processes that arrives at the same time.
Suppose the duration of CPU burst of n processes is
l1,l2,…ln , li<lj for i<j, (it’s sorted in ascending
order
Assume the schedule that achieves minimum average waiting time is
not SJF, it means that there exists at least a pair of processes, j and
k, such that lj<lk, but process j is scheduled to run after
process k
The schedule: … lk … lj …
We can switch lk and lj to decrease avg. waiting time.
Assumption is wrong, and SJF achieves minimum avg. waiting time.

21
Example of Non-Preemptive SJF

Process Arrival Time
P1
0.0
P2
2.0
P3
4.0
P4
5.0
SJF (non-preemptive)
P1
0

3
Burst Time
7
4
1
4
P3
7
P2
8
P4
12
Average waiting time = (0 + 6 + 3 + 7)/4 = 4
22
16
Example of Preemptive SJF

Process Arrival Time Burst Time
P1
0.0
7
P2
2.0
4
P3
4.0
1
P4
5.0
4
SJF (preemptive)
P1
0

P2
2
P3
4
P2
5
P4
7
P1
11
16
Average waiting time = (9 + 1 + 0 +2)/4 = 3
23
Length of Next CPU Burst ?

Predict next CPU burst length

Using length of previous CPU bursts, using exponential moving
averaging
1. use subscript n to denote the n th CPU burst
2. tn  actual length of n th CPU burst
3.  n 1  predicted (n  1) th CPU burst length
4.  , 0    1
 n 1   t n  1    n .
4. Define :
24
Exponential Average Prediction
25
Examples of Exponential Averaging

 =0


n+1 =  tn, only previous CPU
burst counts
Expand formula:


n+1 = n,, recent history does
not count
 =1


 n 1   t n  1    n .
n+1 =  tn+(1 - ) tn -1 + …
+(1 -  )j  tn -j + …
+(1 -  )n +1 0
each successive term has less
weight than its predecessor

26
Both  and (1 - ) are less than or
equal to 1
Priority Scheduling

CPU is allocated to process with the highest priority



Can be Preemptive or Nonpreemptive


A priority number (integer) is associated with each process
smallest integer  highest priority
Preemptive: newly arrived higher priority process preempt
current running process
Priority can be internal or external

Internal: decided based on some quantities of process


27
resource requirement, ratio of I/O burst and avg burst, next CPU
bursts length
External: importance, funds paid,…
Priority Scheduling

Problem of priority scheduling


Indefinite blocking, starvation: low priority processes may never
execute
Solution


28
Aging – as time progresses increase the priority of the waiting
process
E.g., increase the priority of a waiting process by 1 every 15
minutes…
Round Robin (RR)

So far, we see FCFS, SJF, Priority Scheduling



How they perform for time-sharing system ?
Each process gets a small unit of CPU time (time quantum),
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 ready queue and time quantum
is q, then each process gets 1/n of CPU time in chunks of at
most q time units at once. No process waits more than (n1)q time units.
29
Example of RR with Time Quantum = 20

Process Burst Time
P1
53
P2
17
P3
68
P4
24
The Gantt chart is:
P1
0

30
P2
20
37
P3
P4
57
P1
77
P3
97
117
P4
P1
P3
121 134
P3
154 162
Typically, higher average turnaround than SJF, but
better response
Round Robin (RR)

Choosing q, time quantum



q large  FIFO
q small  processor sharing
Downside of having very small q ?

31
q must be large with respect to context switch, otherwise
overhead is too high
Quantum and Context Switch
32
Turnaround Time Varies vs Time Quantum
33
Multilevel Queue

Ready queue is partitioned into separate queues




Each queue has its own scheduling algorithm




Realtime process
foreground (interactive) process
background (batch) process
Realtime – Earliest Deadline First
foreground – RR
background – FCFS
Scheduling must be done between queues

Fixed priority scheduling; (i.e., serve all from realtime, then
foreground, then from background).


Time slice – each queue gets a certain amount of CPU time

34
Low priority process might experience starvation
i.e., 80% to foreground in RR, 20% to background in FCFS
Multilevel Queue Scheduling
35
How do we assign process to appropriate queue ?
Multilevel Feedback Queue Scheduling

Compared to multilevel queue scheduling


A process can move between various queues; aging can be
implemented this way
To favor shorter jobs, or process with short CPU burst

Penalize process that have been run longer by moving it to
lower priority queue …

36
Feed the process back to a different queue
Example of Multilevel Feedback Queue
Process becomes
ready
I/O, terminate
Timer goes off
I/O, terminate
I/O, terminate
Q0 – RR with quantum 8 milliseconds
Q1 – RR quantum 16 milliseconds
Q2 – FCFS
Strict priority among queues Q1>Q2>Q3
37
Multilevel Feedback Queue

Effect of multilevel feedback queue



I/O bound and interactive processes stay in highest priority queue
CPU bound process stays in low priority queue
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
38
Multiple-Processor Scheduling



CPU scheduling more complex when multiple CPUs are
available
Homogeneous processors within a multiprocessor
Two appraches

Symmetric multiprocessing: each processor is self-scheduling


Asymmetric multiprocessing: one processor (master server) handles
scheduling decision, I/O Processing, and other system activities

39
Share a common queue, or private queue
– only one processor accesses the system data structures, alleviating
the need for data sharing
Multiple-Processor Scheduling:
considerations

Recall L1 cache ?

To avoid cache invalidation and repopulation, scheduling
taking into account processor affinity: a process has an
affinity to the processor that it is currently running on


Avoid migrating a process to other processor
Load balancing: keep workload balanced among
processors


40
Needed if each processor has its private queue
Push migration and pull migration
Real-Time Scheduling


Hard real-time systems – required to complete a critical
task within a guaranteed amount of time
Soft real-time computing – requires that critical
processes receive priority over less fortunate ones
41
Thread Scheduling

Local Scheduling – How the threads library decides which
thread to put onto an available LWP

Global Scheduling – How the kernel decides which kernel
thread to run next
42
Case studies: Windows XP and Linux

Homework: comparing CPU scheduling used by
Windows XP and Linux


Details to come
Evaluation of different CPU scheduling algorithms




43
By hand
Through mathematical model
Through simulation studies
Actual Implementation and measurement