SELF-INSTRUCTIONAL STUDY MATERIAL FOR JGND PSOU, ALL COPYRIGHTS WITH JGND PSOU, PATIALA

JAGAT GURU NANAK DEV

PUNJAB STATE OPEN UNIVERSITY, PATIALA
(Established by Act No. 19 of 2019 of the Legislature of State of Punjab)

The Motto of the University

(SEWA)
SKILL ENHANCEMENT EMPLOYABILITY WISDOM ACCESSIBILITY

M.Sc. (Computer Science)

Course Name: Operating Systems
Course Code: MSCS-2-01T
Programme Coordinator: Dr. Karan Sukhija
ADDRESS: C/28, THE LOWER MALL, PATIALA-147001
WEBSITE: www.psou.ac.in
1
 JAGAT GURU NANAK DEV
PUNJAB STATE OPEN UNIVERSITY PATIALA
(Established by Act No.19 of 2019 of Legislature of the State of Punjab)

M.Sc. (Computer Science) Programme Coordinator:

Dr. Karan Sukhija (Assistant Professor)
School of Sciences and Emerging Technologies
JGND PSOU, Patiala

Faculty of School of Science and Emerging Technologies:

Dr. Baljit Singh Khera (Head)

Professor, School of Sciences and Emerging Technologies
Jagat Guru Nanak Dev Punjab State Open University, Patiala

Dr. Kanwalvir Singh Dhindsa

Professor, School of Sciences and Emerging Technologies
Jagat Guru Nanak Dev Punjab State Open University, Patiala

Dr. Amitoj Singh

Associate Professor, School of Sciences and Emerging Technologies
Jagat Guru Nanak Dev Punjab State Open University, Patiala

Dr. Monika Pathak

Assistant Professor, School of Sciences and Emerging Technologies
Jagat Guru Nanak Dev Punjab State Open University, Patiala

Faculty of School of Business Management & Commerce:

Dr. Pooja Aggarwal
Assistant Professor, School of Business & Commerce
Jagat Guru Nanak Dev Punjab State Open University, Patiala

Faculty of School of Social Sciences and Liberal Arts:

Dr. Pinky Sra
Assistant Professor, School of Social Sciences and Liberal Arts
Jagat Guru Nanak Dev Punjab State Open University, Patiala

2
 JAGAT GURU NANAK DEV
PUNJAB STATE OPEN UNIVERSITY PATIALA
(Established by Act No.19 of 2019 of Legislature of the State of Punjab)

PROGRAMME COORDINATOR
Dr. Karan Sukhija (Assistant Professor)
School of Sciences and Emerging Technologies
JGND PSOU, Patiala

COURSE COORDINATOR :
Dr. Monika Pathak
Assistant Professor, School of Sciences and Emerging Technologies
Jagat Guru Nanak Dev Punjab State Open University, Patiala

COURSE OUTCOMES(COs)
After Completion of the course, learners will be able to

1. Understand the structure of computing systems, from the hardware level through the
operating system level and onto the applications level.
2. Understand basics of operating system viz. system programs, system calls, user mode and
kernel mode.
3. Learn the working with CPU scheduling algorithms for specific situation, and analyze the
environment leading to deadlock and its rectification.
4. Explore the memory management techniques viz. caching, paging, segmentation, virtual
memory, and thrashing.

3
 JAGAT GURU NANAK DEV
PUNJAB STATE OPEN UNIVERSITY PATIALA
(Established by Act No.19 of 2019 of Legislature of the State of Punjab)

PREFACE

Jagat Guru Nanak Dev Punjab State Open University, Patiala was established in Decembas
2019 by Act 19 of the Legislature of State of Punjab. It is the first and only Open Universit of the
State, entrusted with the responsibility of making higher education accessible to all especially to
those sections of society who do not have the means, time or opportunity to pursue regular
education.

In keeping with the nature of an Open University, this University provides a flexible
education system to suit every need. The time given to complete a programme is double the duration
of a regular mode programme. Well-designed study material has been prepared in consultation with
experts in their respective fields.

The University offers programmes which have been designed to provide relevant, skill-
based and employability-enhancing education. The study material provided in this booklet is self
instructional, with self-assessment exercises, and recommendations for further readings. The
syllabus has been divided in sections, and provided as units for simplification.

The Learner Support Centres/Study Centres are located in the Government and Government
aided colleges of Punjab, to enable students to make use of reading facilities, and for curriculum-
based counselling and practicals. We, at the University, welcome you to be a part of this institution
of knowledge.

Prof. G. S. Batra,
Dean Academic Affairs

4
 M.Sc. (Computer Science)
Semester II
MSCS-2-01T: Operating Systems
Total Marks: 100
External Marks: 70
Internal Marks: 30
Credits: 4
Pass Percentage: 40%

SECTION A

UNIT- I: Introduction and System Structures: Computer-System Organization, Computer-

System Architecture, Operating-System Structure, Operating-System Operations, Process
Management, Memory Management, Storage Management, Protection and Security, Computing
Environments, Operating-System Services, User and Operating-System Interface, System Calls,
Types of System Calls, System Programs.

UNIT II: Process Management: Process Concept, Process Scheduling, Operations on Processes,
Multi-threaded programming: Multithreading Models, Process Scheduling: Basic Concepts,
Scheduling Criteria, and Scheduling Algorithms.

Unit III: Deadlock: System Model, Deadlock Characterization, Methods for Handling Deadlocks,
Deadlock Prevention, Deadlock Avoidance, Deadlock Detection, Recovery from Deadlock.

UNIT IV: Memory Management: Basic Hardware, Address Binding, Logical and Physical
Address, Dynamic linking and loading, Swapping, Contiguous Memory Allocation, Segmentation,
Paging, Demand Paging, Page Replacement algorithms

SECTION B

UNIT V: File Systems: File Concept, Access Methods, Directory and Disk Structure, File-System
Structure, File-System Implementation, Directory Implementation, Allocation Methods, Free-Space
Management.

UNIT VI: Introduction to Linux: Linux's shell, Kernel, Features of Linux, Using file system:
Filenames, Introduction to different types of directories: Parent, Subdirectory, Home directory; rules
to name a directory, Important directories in Linux File System,

UNIT VII: Linux Commands: cal, date, echo, bc, who, cd, mkdir, rmdir, ls, cat cp, rm, mv, more,
5
gzip, tar, File ownership, file permissions, chmod, Directory permission, change file ownership,

UNIT VIII: Shell Scripting: Creating and Executing Shell Programs, Using variables: Assigning a
value to a variable, Accessing the value of a variable, Positional Parameters and other Built-In Shell
Variables; Special Characters, Conditional Statements : if Statement, case Statement; Iteration
Statements : for Statement, while Statement, until Statement

Suggested Readings
1. A Silberschatz, P.B. Galvin, G. Gagne, Operating Systems Concepts, 8th Edition, John Wiley
Publications, 2009
2. A.S. Tanenbaum, Modern Operating Systems, 3rd Edition, Pearson Education, 2014
3. G. Nutt, Operating Systems: A Modern Perspective, 2nd Edition Pearson Education, 2000
4. S. Das, Unix Concepts and Applications, 4th edition, McGraw Hill Education, 2017

6
 M.Sc. (Computer Science)
SEMESTER-II
OPERATING SYSTEM

UNIT 1: INTRODUCTION AND SYSTEM STRUCTURES

STRUCTURE
1. Objective

1.1. Operating System

1.2. Views of Operating System

1.3. Types of Operating System

1.4. Functions of Operation System

1.5. Networking

1.6. Protection or Security

1.7. System Calls

1.7.1. Process Control

1.7.2. File Manipulation

1.7.3. Device Management

1.7.4. Information Maintenance

1.7.5. Communication

1.8. System Programs

1.9. System Structure

1.10. Operating System Services

1.11. Practice Exercise

7
1.OBJECTIVE

To understand the following

Computer-System Architecture
Operating-System Structure, Operating-System Operations
Operating-System Services, User and Operating-System Interface
System Calls, Types of System Calls, System Programs.
1.1.Operating System

An operating system is a program which manages all the computer hardware.

It provides the base for application program and acts as an intermediary between a user
and thecomputer hardware.
The operating system has two objectives such as:

 Firstly, an operating system controls the computer‘s hardware.

 The second objective is to provide an interactive interface to the user and

interpretcommands so that it can communicate with the hardware.
The operating system is very important part of almost every computer system.

Managing Hardware

The prime objective of operating system is to manage & control the various hardware
resourcesof a computer system.
These hardware resources include processer, memory, and disk space and so on.

The output result was display in monitor. In addition to communicating with the
hardware the operating system provides on error handling procedure and display an
8
 error notification.
If a device not functioning properly, the operating system cannot be communicate with
thedevice.
Providing an Interface
It provides a stable and consistent way for applications to deal with the hardware without
theuser having known details of the hardware.

If the program is not functioning properly, the operating system again takes control, stops
theapplication and displays the appropriate error message.
Computer system components are divided into 5 parts

 Computer hardware

 Operating system

 Utilities

 Application programs

 End user

The operating system controls and coordinate a user of hardware and various
applicationprograms for various users.
It is a program that directly interacts with the hardware.

The operating system is the first encoded with the Computer and it remains on the
memory alltime thereafter.

System Goals
The purpose of an operating system is to be provided an environment in which an
user canexecute programs.
Its primary goals are to make the computer system convenience for the user.

Its secondary goals are to use the computer hardware in efficient manner.

1.2.VIEW OF OPERATING SYSTEM

User view:The user view of the computer varies by the interface being used. The

9
 examplesare -windows XP, vista, windows 7 etc. Most computer user sit in the in front
of personal computer (pc) in this case the operating system is designed mostly for easy
use with some attention paid to resource utilization. Some user sit at a terminal
connected to amainframe/minicomputer. In this case other users are accessing the same
computer through the other terminals. There user are share resources and may
exchange the information. The operating system in this case is designed to maximize
resources utilization to assume that all available CPU time, memory and I/O are used
efficiently and no individual user takes more than his/her fair and share.The other
users sit at workstations connected to network of other workstations and servers.
These users have dedicated resources but they share resources such as networking
and servers like file, compute and print server. Here the operating system is
designed to compromise between individual usability and resource utilization.
System view: From the computer point of view the operating system is the program
which is most intermediate with the hardware. An operating system has resources as
hardware and software which may be required to solve a problem like CPU time,
memory space, file storage space and I/O devices and so on. That‘s why the operating

system acts as manager of these resources. Another view of the operating system is it is
a control program. A control program manages the execution of user programs to
present the errors in proper use of the computer. It is especially concerned of the user
the operation and controls the I/O devices.

1.3. TYPES OF OPERATING SYSTEM

1. Mainframe System: It is the system where the first computer used to handle many
commercialscientific applications. The growth of mainframe systems traced from simple
batch system where the computer runs one and only one application to time shared
systems which allowed for user interaction with the computer system
a. Batch /Early System: Early computers were physically large machine. The common
input devices were card readers, tape drivers. The common output devices were
line printers, tape drivers and card punches. In these systems the user did not
interact directlywith the computer system. Instead the user preparing a job which
consists of programming data and some control information and then submitted it
to the computer operator after some time the output is appeared.
10
 The output in these early computer was fairly simple is main task was to transfer
control automatically from one job to next. Theoperating system always resides in
the memory. To speed up processing operators batched the jobs with similar needs
and ran then together as a group. The disadvantages of batch system are that in this
execution environment the CPU is often idle because thespeed up of I/O devices is
much slower than the CPU.

b. Multiprogrammed System: Multiprogramming concept increases CPU utilization by

organization jobs so that the CPU always has one job to execute the idea behind
multiprogramming concept. This set of job is subset of the jobs kept in the job pool.
The operating system picks and beginning to execute one of the jobs in the memory.
In this environment the operating system simply switches and executes another job.
When a job needs to wait the CPU is simply switched to another job and so on. The
multiprogramming operating system is sophisticated because the operating system
makes decisions for the user. This is known asscheduling. If several jobs are ready
torun at the same time the system choose one among them. This is known as CPU
scheduling. The disadvantages of the multiprogrammedsystem are
 It does not provide user interaction with the computer system during the
program execution.
 The introduction of disk technology solved these problems rather than
reading the cards from card reader into disk. This form of processing is
known as spooling.
SPOOL stands for simultaneous peripheral operations online. It uses the disk as a
huge buffer for reading from input devices and for storing output data until the

11
 output devices accept them. It is also use for processing data at remote sides. The
remote processing is done and its own speed with no CPU intervention. Spooling
overlaps the input, output one job with computation of other jobs. Spooling has a
beneficial effect onthe performance of the systems by keeping both CPU and I/O
devices working at much higher time.
c. Time Sharing System: The time sharing system is also known as multi user systems.
The CPU executes multiple jobs by switching among them but the switches occurs
so frequently that the user can interact with each program while it is running. An
interactive computer system provides direct communication between a user and
system. The user gives instruction to the operating systems or to a program directly
using keyboard or mouse and wait for immediate results. So the response time will
be short. The time sharing system allows many users to share the computer
simultaneously. Since each action in this system is short, only a little CPU time is
needed for each user. The system switchesrapidly from one user to the next so each
user feels as if the entire computer system is dedicated to his use, even though it is
being shared by many users. The disadvantages oftime sharing system are:
 It is more complex than multiprogrammed operating system

 The system must have memory management & protection, since several jobs are
keptin memory at the same time.
 Time sharing system must also provide a file system, so disk management is required.

 It provides mechanism for concurrent execution which requires complex CPU

scheduling schemes.
2. Personal Computer System/Desktop System: Personal computer appeared in 1970‘s.
They are microcomputers that are smaller & less expensive than mainframe systems.
Instead of maximizing CPU & peripheral utilization, the systems opt for maximizing user
convenience & responsiveness. At first file protection was not necessary on a personal
machine. But when other computers 2nd other users can access the files on a pc file
protection becomes necessary.The lack of protection made if easy for malicious programs
to destroy data on such systems. These programs may be self replicating& they spread via

12
 worm or virus mechanisms. They can disrupt entire companies or even world wide
networks. E.g : windows 98, windows 2000, Linux.
3. Microprocessor Systems/ Parallel Systems/ Tightly coupled Systems: These
Systems have more than one processor in close communications which share the computer
bus, clock, memory & peripheral devices. Ex: UNIX, LINUX. Multiprocessor Systems
have 3 main advantages.
a. Increased throughput: No. of processes computed per unit time. By increasing the
no. of processors move work can be done in less time. The speed up ratio with N
processors is not N, but it is less than N. Because a certain amount of overhead is
incurred in keeping all the parts working correctly.
b. Increased Reliability: If functions can be properly distributed among several
processors, then the failure of one processor will not halt the system, but slow it
down. This ability to continue to operate in spite of failure makes the system fault
tolerant.
c. Economic scale: Multiprocessor systems can save money as they can share
peripherals,storage & power supplies.
The various types of multiprocessing systems are:

 Symmetric Multiprocessing (SMP): Each processor runs an identical copy of the

operating system & these copies communicate with one another as required. Ex:
Encore‘s version of UNIX for multi max computer. Virtually, all modern operating
systemincluding Windows NT, Solaris, Digital UNIX, OS/2 & LINUX now provide
support forSMP.

 Asymmetric Multiprocessing (Master – Slave Processors): Each processor is

designed for a specific task. A master processor controls the system & schedules &
allocates the work to the slave processors. Ex- Sun‘s Operating system SUNOS
version 4provides asymmetric multiprocessing.
4. Distributed System/Loosely Coupled Systems: In contrast to tightly coupled systems,
the processors do not share memory or a clock. Instead, each processor has its own local
memory. The processors communicate with each other by various communication lines
such as high speed buses or telephone lines. Distributed systems depend on networking
for their functionalities. By being able to communicate distributed systems are able to
share computational tasks and provide a rich set of features to the users. Networks vary by
13
 the protocols used, the distances between the nodes and transport media. TCP/IP is the
most common network protocol. The processor is a distributed system varies in size and
function. Itmay microprocessors, work stations, minicomputer, and large general purpose
computers. Network types are based on the distance between the nodes such as LAN
(within a room, flooror building) and WAN (between buildings, cities or countries). The
advantages of distributed system are resource sharing, computation speed up, reliability,
communication.
5. Real time Systems: Real time system is used when there are rigid time requirements on

the operation of a processor or flow of data. Sensors bring data to the computers. The
computer analyzes data and adjusts controls to modify the sensors inputs. System that
controls scientific experiments, medical imaging systems and some display systems are
real time systems. The disadvantages of real time system are:
a. A real time system is considered to function correctly only if it returns the correct
result within the time constraints.
b. Secondary storage is limited or missing instead data is usually stored in short term
memory or ROM.
c. Advanced OS features are absent.Real time system is of two types such as:

 Hard real time systems: It guarantees that the critical task has been completed on
time.The sudden task is takes place at a sudden instant of time.
 Soft real time systems: It is a less restrictive type of real time system where a critical
task gets priority over other tasks and retains that priority until it computes. These
have more limited utility than hard real time systems. Missing an occasional
deadline is acceptable
e.g. QNX, VX works. Digital audio or multimedia is included in this category.

It is a special purpose OS in which there are rigid time requirements on the operation of a
processor. A real time OS has well defined fixed time constraints. Processing must be
done within the time constraint or the system will fail. A real time system is said to function
correctlyonly if it returns the correct result within the time constraint. These systems are
characterized by having time as a key parameter.

1.4. Functions of Operation System

The various functions of operating system are as follows:
14
1. Process Management
 A program does nothing unless their instructions are executed by a CPU.A process is a
programin execution. A time shared user program such as a complier is a process. A
word processing program being run by an individual user on a pc is a process.
 A system task such as sending output to a printer is also a process. A process needs
certain resources including CPU time, memory files & I/O devices to accomplish its
task.
 These resources are either given to the process when it is created or allocated to it while
it is running. The OS is responsible for the following activities of process management.
 Creating & deleting both user & system processes.

 Suspending & resuming processes.

 Providing mechanism for process synchronization.

 Providing mechanism for process communication.

 Providing mechanism for deadlock handling.

2. Main Memory Management

The main memory is central to the operation of a modern computer system. Main memory
is a large array of words or bytes ranging in size from hundreds of thousand to billions.
Main memory stores the quickly accessible data shared by the CPU & I/O device. The
central processor reads instruction from main memory during instruction fetch cycle & it
both reads & writes data from main memory during the data fetch cycle. The main memory
is generally theonly large storage device that the CPU is able to address & access directly.
For example, for theCPU to process data from disk. Those data must first be transferred to
main memory by CPU generated E/O calls. Instruction must be in memory for the CPU to
execute them. The OS is responsible for the following activities in connection with
memory management.
 Keeping track of which parts of memory are currently being used & by whom.

 Deciding which processes are to be loaded into memory when memory space
becomesavailable.
 Allocating &deallocating memory space as needed.

15
3. File Management

File management is one of the most important components of an OS computer can store
information on several different types of physical media magnetic tape, magnetic disk &
opticaldisk are the most common media. Each medium is controlled by a device such as
disk drive or tape drive those has unique characteristics. These characteristics include
access speed, capacity, data transfer rate & access method (sequential or random).For
convenient use of computer system the OS provides a uniform logical view of information
storage. The OS abstracts from the physical properties of its storage devices to define a
logical storage unit the file. A file is collection of related information defined by its
creator. The OS is responsible for the followingactivities of file management.
 Creating & deleting files.

 Creating & deleting directories.

 Supporting primitives for manipulating files & directories.

 Mapping files into secondary storage.

 Backing up files on non-volatile media.

4. I/O System Management

One of the purposes of an OS is to hide the peculiarities of specific hardware devices from
the user. For example, in UNIX the peculiarities of I/O devices are hidden from the bulk of
the OSitself by the I/O subsystem. The I/O subsystem consists of:
 A memory management component that includes buffering, catching & spooling.

 A general device- driver interfaces drivers for specific hardware devices. Only the
device driver knows the peculiarities of the specific device to which it is assigned.

5. Secondary Storage Management

The main purpose of computer system is to execute programs. These programs with the
data they access must be in main memory during execution. As the main memory is too
small to accommodate all data & programs & because the data that it holds are lost when
power is lost. The computer system must provide secondary storage to back-up main
memory. Most modern computer systems are disks as the storage medium to store data &

16
 program. The operating system is responsible for the following activities of disk
management.
 Free space management.

 Storage allocation.

 Disk scheduling

Because secondary storage is used frequently it must be used efficiently.

1.5. NETWORKING
A distributed system is a collection of processors that don‘t share memory peripheral devices
or a clock. Each processor has its own local memory & clock and the processor communicate
with one another through various communication lines such as high speed buses or networks.
The processors in the system are connected through communication networks which are
configured in a number ofdifferent ways. The communication network design must consider
message routing & connection strategies are the problems of connection & security.

1.6. PROTECTION OR SECURITY

If a computer system has multi users & allow the concurrent execution of multiple processes
then the various processes must be protected from one another‘s activities. For that purpose,
mechanisms ensure that files, memory segments, CPU & other resources can be operated on by
onlythose processes that have gained proper authorization from the OS.

1.7. SYSTEM CALLS

System calls provide the interface between a process & the OS. These are usually available in
the form of assembly language instruction. Some systems allow system calls to be made
directly from ahigh level language program like C, BCPL and PERL etc. systems calls occur
in different ways depending on the computer in use. System calls can be roughly grouped into 5
major categories.

1.7.1. Process Control

End, abort: A running program needs to be able to has its execution either normally
17
 (end) orabnormally (abort).
Load, execute: A process or job executing one program may want to load and
executes another program.
Create Process, terminate process: There is a system call specifying for the purpose

of creating a new process or job (create process or submit job). We may want to
terminate a job or process that we created (terminates process, if we find that it is
incorrect or no longer needed).
Get process attributes, set process attributes: If we create a new job or process we
should able to control its execution. This control requires the ability to determine &
reset the attributes of a job or processes (get process attributes, set process attributes).
Wait time: After creating new jobs or processes, we may need to wait for them to
finish theirexecution (wait time).
Wait event, signal event: We may wait for a specific event to occur (wait event). The
jobs orprocesses then signal when that event has occurred (signal event).

1.7.2. File Manipulation

Create file, delete file: We first need to be able to create & delete files. Both the
system callsrequire the name of the file & some of its attributes.
Open file, close file: Once the file is created, we need to open it & use it. We close
the file when we are no longer using it.
Read, write, reposition file: After opening, we may also read, write or reposition the
file (rewind or skip to the end of the file).
Get file attributes, set file attributes: For either files or directories, we need to be
able to determine the values of various attributes & reset them if necessary. Two
system calls get fileattribute & set file attributes are required for their purpose.

1.7.3. Device Management

Request device, release device: If there are multiple users of the system, we first
request the device. After we finished with the device, we must release it.
Read, write, reposition: Once the device has been requested & allocated to us, we can
read,write & reposition the device.
18
1.7.4. Information Maintenance
Get system data, set system data: Other system calls may return
information about thesystem like number of current users, version number of OS,
amount of free memory etc.
Get process attributes, set process attributes: The OS keeps information
aboutall itsprocesses & there are system calls to access this information.

1.7.5. Communication
There are two modes of communication such as:

a. Message passing model: Information is exchanged through an inter process

communication facility provided by operating system. Each computer in a network
has a name by which it is known. Similarly, each process has a process name which
is translated toan equivalent identifier by which the OS can refer to it. The get hostid

and get processed systems calls to do this translation. These identifiers are then
passed to the general purpose open & close calls provided by the file system or to
specific open connection system call. Therecipient process must give its permission
for communication to take place with an accept connection call. The source of the
communication known as client & receiver known as server exchange messages by
read message & write message system calls. The closeconnection call terminates the
connection.

b. Shared memory model: processes use map memory system calls to access regions
of memory owned by other processes. They exchange information by reading &
writing data inthe shared areas. The processes ensure that they are not writing to the
same locationsimultaneously.

1.8. SYSTEM PROGRAMS

System programs provide a convenient environment for program development & execution.
Theyare divided into the following categories.
a. File manipulation: These programs create, delete, copy, rename, print & manipulate
filesand directories.
19
 b. Status information: Some programs ask the system for date, time & amount of
availablememory or disk space, no. of users or similar status information.

i. File modification:Several text editors are available to create and modify the
contents of filestored on disk.
ii. Programming language support: compliers, assemblers & interpreters are
provided to the user with the OS.
iii. Programming loading and execution: Once a program is assembled or
compiled, it mustbe loaded into memory to be executed.
iv. Communications: These programs provide the mechanism for creating virtual
connectionsamong processes users 2nd different computer systems.

v. Application programs: Most OS are supplied with programs that are useful to
solve common problems or perform common operations. Ex: web browsers,
word processors & text formatters etc.

1.9. SYSTEM STRUCTURE

1. Simple structure: There are several commercial system that don‘t have a well-
defined structure such operating systems begins as small, simple & limited systems and
then grow beyond their original scope. MS-DOS is an example of such system. It was

not divided intomodules carefully. Another example of limited structuring is the UNIX
operating system.

2. Layered approach: In the layered approach, the OS is broken into a number of layers
(levels) each built on top of lower layers. The bottom layer (layer o ) is the hardware &
20
 top
most layer (layer N) is the user interface.
The main advantage of the layered approach
is modularity.
The layers are selected such that each users
functions (or operations) & services of only
lower layer.

This approach simplifies debugging & system verification, i.e. the first layer can be
debugged without concerning the rest of the system. Once the first layer is debugged, its
correct functioning is assumed while the 2nd layer is debugged & so on.
If an error is found during the debugging of a particular layer, the error must be on that
layer because the layers below it are already debugged. Thus the design & implementation
of the system are simplified when the system is broken down into layers.

Layers Functions
5 User Program
4 I/O Management
3 Operator Process Communication
2 Memory Management
1 CPU Scheduling
0 Hardware

Each layer is implemented using only operations provided by lower layers. A layer doesn‘t
need to know how these operations are implemented; it only needs to know what these
operations do.
The layer approach was first used in the operating system. It was defined in six layers.

The main disadvantage of the layered approach is:

 The main difficulty with this approach involves the careful definition of the layers,
because a layer can use only those layers below it. For example, the device driver for
thedisk space used by virtual memory algorithm must be at a level lower than that

21
 of the memory management routines, because memory management requires the
ability to usethe disk space.
 It is less efficient than a non layered system (Each layer adds overhead to the system
call & the net result is a system call that take longer time than on a non layered
system).
1.10. Operating System Services

An operating system provides an environment for the execution of the program. It provides
some services to the programs. The various services provided by an operating system are as
follows:
 Program Execution: The system must be able to load a program into memory and to
run that program. The program must be able to terminate this execution either normally
or abnormally.
 I/O Operation: A running program may require I/O. This I/O may involve a file or a
I/O device for specific device. Some special function can be desired. Therefore the
operating system must provide a means to do I/O.
 File System Manipulation: The programs need to create and delete files by name and
read and write files. Therefore the operating system must maintain each and every files
correctly.
 Communication: The communication is implemented via shared memory or by the
technique of message passing in which packets of information are moved between the
processes by the operating system.
 Error detection: The operating system should take the appropriate actions for the
occurrences of any type like arithmetic overflow, access to the illegal memory location
andtoo large user CPU time.
 Research Allocation: When multiple users are logged on to the system the resources
must be allocated to each of them. For current distribution of the resource among the
various processes the operating system uses the CPU scheduling run times which
determine which process will be allocated with the resource.
 Accounting: The operating system keep track of which users use how many and which
kindof computer resources.
 Protection: The operating system is responsible for both hardware as well as software
protection. The operating system protects the information stored in a multiuser
22
 computer system.
1.11. PRACTICE EXERCISE

a. Why is the operating system important?

b. What's the main purpose of an OS? What are the different types of OS?

c. What are the benefits of a multiprocessor system?

d. What is a bootstrap program in OS?

e. What is the main purpose of an operating system?

f. What are the different operating systems?

g. What is kernel?

h. What is monolithic kernel?

i. What do you mean by a process?

23
 M.Sc. (Computer Science)
OPERATING SYSTEM

UNIT 1: INTRODUCTION AND SYSTEM STRUCTURES

STRUCTURE
Objective
Introduction to Process
Process Basics
Process Description
Process Control Block
Role of PCB
Process Schedulers
Long-term Scheduler
Medium-term Scheduler
Short-term Scheduler
Operation on Processes
Multi-Threaded Programming
Multi-Threaded Models
CPU Scheduling
CPU Scheduling Criteria
Practice Exercise

24
OBJECTIVES
To understand the following
 Different types of Process Scheduling
 Operations on Processes
 Multi-threaded programming and Model
 CPU Scheduling
 Scheduling Concepts
 Scheduling Criteria
 Scheduling Algorithms

INTRODUCTION TO PROCESS
Process Basics

The fundamental activity of an operating system is the creation, management, and

termination of processes. Now the question comes to mind, what is a process?
A process is a program under execution or the ―animated‖ existence of a program or
an identifiable entity executed on a processor by the operating system.
A process may be defined as an instance of a program in execution. It is also
known as a task. An operating system manages each hardware resource attached to the
computer by representing it as an abstraction. Abstraction hides unwanted details from
the users and programmers allowing them to have a view of the resources in the form,
which is convenient to them. A process is an abstract model of a sequential program in
execution. The operating system can schedule a process as a unit of work. A process can
be identifying in an operating system by its following components:

The object program (or code) to be executed.

The data on which the program will execute.
The status of the process execution.
The stack is associated with a process to store parameters and calling addresses.

A process may be represented schematically as in the figure.

25
 Process Abstraction
Process Description
To control and manage the processes, the Operating system records the information about
them in the primary process table. The primary process table is used to keep one entry per
process in the operating system. Each entry contains at least one pointer to a process
image. The Process Image contains User Data: It contains program data that can be
modified etc., Code: The sequence of instructions (program) to be executed, Stack: Each
process has one or more stacks associated with it. A stack is used to saved parameters of
the process and calling addresses for process, system calls, and Process Control Block
(PCB) of process in which data needed by the operating system to control the process
(attributes and information about the process) is stored.

Process Control Block

The data structure that stores information about a process is called Process Control
Block. When a process is initialized, the corresponding process control block of the
process is created. Information in a process control block is updated during the transition
of process states.
A process control block is a location in the main memory, where various
information of a process regarding memory, process, and I/O management is stored. Each
process has a single process control block. When a process is completed the process
control block is unloaded from the memory. The information stored in a process control
block is:

Pointer Process
State
Process Number
26
 Program Counter
Registers
Memory Limits
List of open files
.
.
.
.

Figure: Process Control Board (PCB)

Each process has a priority, implemented in terms of numbers. The higher priority
processes have precedence over lower priority processes. The priority field is used
for storing the priority of a process.
A new process can be created from the existing process. The existing process is
called the parent of the newly created process. The field, link (pointer) to the parent
process stores the address of the process control block of the parent process in the
main memory. The field, link (pointer) to the child process stores the address of the
process control block of the child process in the main memory.

Process State field stores the information about the recent state of the process. The
state of the process may be new, waiting, ready, running, and so on.
Process Number filled to store the numeric value which is an identifier of the
process.
The PCB stores information regarding the programming environment of a process.
The programming environment information includes the value of the registers,
stack, and program counter.
A program counter is a special register that saves the address of the next
instruction to be executed for this process.
A PCB also stores information regarding memory management, such as the number
of memory units allocated to the process and the addresses of the memory chunks
allocated.
Registers: It includes a general-purpose register, stack pointers, index register, and

27
 accumulators, etc. Many register and accumulators etc. Some register and type of
register depend upon the computer architecture.
The PCB control block stores file management information. An example of the file
management information stored in the process control block is:

 The number of files opens.

 List of the open files.
 The access right of the files opens, such as read-only or read-write.

Role of PCB
The process control block is the most important data structure in an OS. Each
Process Control Block contains all of the data about a process that is required by the
Operating System. The blocks are read as well as modified by virtually every unit in the
Operating System, including those which involved scheduling, interrupt processing,
resource allocation, and performance monitoring and analysis of the process.
One can say that the set of process control blocks defines the state of the OS. This brings
up an important design issue. Many routines within the OS will need access to
information in process control blocks. The provision of direct access to these tables is not
difficult. Each process is equipped with a unique ID, and this can be used as an index into
a table of pointers to the process control blocks.
The difficulty is not access but rather protection. There are two problems :
1. A bug in a single routine, such as an interrupt handler, could damage process
control blocks, which could destroy the system‘s ability to manage the affected.
2. A design change in the structure or semantics of the process control block could
affect many modules in the OS.

PROCESS SCHEDULERS
Several types of schedulers can be used in an OS. Schedulers are classified according to
the type and duration of processes. Schedulers are classified as:
Long-term Scheduler/High-Level Scheduler
Medium-term Scheduler/ Intermediator Scheduler
Short-term Scheduler/Low-Level Schedular

28
Long-term Scheduler

Long-term scheduling identifying which programs is to be admitted to the system as new

processes. Once a new process is accepted, it may enter the scheduling queues in one of
two places:
If all resources are initially fully available to the new process, they may be
admitted to the tail of the ready queue.
If all resources are not immediately available, the new process may be entered in
the blocked-suspended queue until those resources are provided.

Long-term Scheduler plans the CPU scheduling for batch jobs. Processes, which are
resource-intensive and have a low priority, are called batch jobs are executed in a group
or bunch. An example of a batch job is a user request for printing a bunch of files.

Medium-term Scheduler
Medium-term scheduling is part of the swapping role of an operating system. Medium-
term Scheduler plans CPU scheduling for the processes that have been waiting for the
completion of another process or an I/O task that requires a long time. A process is
suspended or blocked marked as waiting if it is waiting for the completion of a long time
I/O task. These processes are removed from the main memory and stored in the swapped-
out queue in the secondary memory to create space in the main memory. The swapped-
out queue is implemented in the secondary memory for storing the waiting processes that
have been swapped out of the main memory. After completion of the I/O operation, the
suspended or blocked processes are resumed and placed in the ready queue in the main
memory. If the process is waiting for the completion of a short-term I/O task, the process
is not swapped out of the main memory and is not handled by the medium-term
scheduler. The success of the medium-term schedules is based on the degree of
multiprogramming that it can maintain, by keeping as many processes ―runnable‖ as
possible.

29
 Figure: Working of medium-term scheduler
Short-term Scheduler
Short-term Scheduler plans the scheduling of the processes that are in a ready state.
Short-term schedulers retrieve a process from the ready queue and allocate CPU time to
it. The process state is changed from ready to run. If an interrupt or time-out occurs the
scheduler places the running process back into the ready queue and marks the running
process as ready. The figure below shows how short, medium, long term schedulers
work.

30
 Figure: Role of various type schedulers

If a new process is a batch job, it is placed in the batch queue; otherwise, it is ready
into the ready queue. The long-term scheduler selects a batch job on a first-come-first-
served basis and sends it to the ready queue for execution.
The short-term scheduler selects a process from the ready queue for execution. If a
process waits for a short time for completion of an I/O task it is placed in the waiting
queue, otherwise, the waiting process is swapped out of the main memory and placed in
the swapped-out queue, which is implemented in the secondary memory.

Sr. Long Term Short Term Medium Term

No.
1. It is a job scheduler. It is a CPU scheduler. It is swapping.
2. Speed is less than short term Speed is very fast. Speed is in between
schedular both.
3. It controls the degree of Less control over the Reduce the degree of
Multiprogramming. degree of Multiprogramming.
Multiprogramming.
4. Absent or minimal in time Minimal in a time-sharing The Time-sharing
sharing system system system uses a
medium-term
scheduler.
5. It selects processes from It selects from among the The process can be
pool and load them into processes that are ready to reintroduced into
memory for execution. execute memory and its
execution can be
continued.
6. Process state is Process state is Process state is
(new to Ready.) (Ready to Running) (waiting)
7. Select a good process, a mix Select a new process for Select suspended
of I/O bound and CPU a CPU quite frequently. Process
bound

31
OPERATION ON PROCESSES
Process State

The principal function of a processor is to execute machine instructions residing in the

main memory. We can view the process from two points of view: Program View and
Processor View.

Process View

Program View Processor View

Different Process View

Program View: Its execution involves a sequence of instructions within that

program. The behavior of individual process can be characterized by a list of the
sequence of instructions – a trace of the process

Processor View: It executes instructions from the main memory, as dictated by

changing values in the program counter register. The behavior of the processor can be
characterized by showing how the traces of various processes are interleaved.

The Simple Two-State Process Model

The operating system‘s principal responsibility is controlling the execution of processes.
This includes determining the interleaving pattern for execution and allocating resources
to processes. The first step in designing an OS to control processes is to describe the
behavior that we would like the processes to exhibit.

We can construct the simplest possible model by observing that, at any time, a process is
either being executed by a processor or not. In this model, a process may be in one of two
states: Running or Not Running, as shown in the figure Two-State Process Model State
transition Diagram. When the OS creates a new process, it creates a Process Control
Block for the process and enters that process into the system in the Not Running state.
The process exists, is known to the OS, and is waiting for an opportunity to execute.
32
From time to time, the currently running process will be interrupted and the dispatcher
portion of the OS will select some other process to run. The former process moves from
the Running state to the Not Running state, and one of the other processes moves to the
Running state.

From this simple model, we can already begin to appreciate some of the design
elements of the OS. Each process must be represented in some way so that the OS can
keep track of it. That is, there must be some information relating to each process,
including the current state and location in memory; this is the Process Control Block.
Processes that are not running must be kept in some sort of queue, waiting their turn to
execute.

Two-State Process Model State transition Diagram

Two-State Process Model Queuing Diagram

33
Process Creation

When a new process is to be added to those currently being managed, the OS builds the
data structures that are used to manage the process and allocates address space in the
main memory to the process. These actions constitute the creation of a new process. Four
common events leading to the creation of a process. In a batch environment, a process is
created in response to the submission of a job. In an interactive environment, a process is
created when a new user attempts to log on. In both cases, the OS is responsible for the
creation of the new process. An OS may also create a process on behalf of an application.
For example, if a user requests that a file be printed, the OS can create a process that will
manage the printing. The requesting process can thus proceed independently of the time
required to complete the printing task.

Traditionally, the OS created all processes in a way that was transparent to the user
or application program, and this is still commonly found with many contemporary
operating systems. However, it can be useful to allow one process to cause the creation of
another. For example, an application process may generate another process to receive
data that the application is generating and to organize those data into a form suitable for
later analysis. The new process runs in parallel to the original process and is activated
from time to time when new data are available. This arrangement can be very useful in
structuring the application. As another example, a server process (which may be a print
server or file server) may create a new process for each request that it handles. When the
OS creates a new process at the explicit request by another process, this work is known as
process spawning.

Parent Process

When one process spawns another, the former is referred to as the parent process.

Child Process

The spawned process is referred to as the child process.

Typically, the ―related‖ processes need to communicate and cooperate. Achieving

this cooperation is a difficult task for the programmer.

Process Termination

34
 Any computer system must provide a means for a process to indicate its completion.
A batch job should include a Halt instruction or an explicit OS service call for
termination. In the former case, the Halt instruction will generate an interrupt to alert the
OS that a process has been completed. For an interactive application, the action of the
user will indicate when the process is completed. When the parent process terminates, the
OS may automatically terminate all its children. For example, in a time-sharing system,
the process for a particular user is to be terminated when the user logs off or turns off his
or her terminal. On a personal computer or workstation, a user may quit an application
(e.g., word processing or spreadsheet). All of these actions ultimately result in a service
request to the OS to terminate the requesting process.

Five State Process Model

A process is a program in execution. In the five-state process model, there are five states
of a process. A process may be in any one of the states during its lifetime. First of all, a
process arrives into the system for its execution, then it becomes ready for execution and,
then it gets the attention of the processor, and at last, the process is terminated after the
completion of its execution. When the parent process terminates, the operating system
may automatically terminate all its children.

The queue is a first-in-first-out list and the processor operates in a round-robin

fashion on the available processes (each process in the queue is given a certain amount of
time, in turn, to execute and then returned to the queue, unless blocked). Some processes
in the Not Running state are ready to execute, while others are blocked, waiting for an
I/O operation to complete.

Thus, using a single queue, the dispatcher could not just select the process at the
oldest end of the queue. Rather, the dispatcher would have to scan the list looking for the
process that is not blocked and that has been in the queue the longest.
A more natural way to handle this situation is to split the Not Running state into two
states: Ready and Blocked. We have added two additional states that will prove useful.
The five states in this new diagram are as follows:

35
 (i) Five-State Process Model State Transition Diagram

(ii) Five-State Process Model Queuing Diagram

So, the process either be in any of one of the below-mentioned states -:

1. NEW: The process is said to be in the NEW state when it arrives into the system
for its execution. In other words, the process enters the JOB POOL of the system. A
process that has been newly created but has not admitted yet into the pool of executable
processes by the OS. Typically, a new process has not yet been loaded into the main
memory, although its process control block has been created.

2. READY: The process is supposed to be in the READY state when it is ready to

get the attention of the CPU In other words, the process enters the READY QUEUE of
the system. A READY process is that, which is prepared to execute when given the
opportunity.
36
 3. RUNNING: A process is supposed to be running, when it gets the attention of
the CPU for its execution. In other words, the process is being executed by the CPU in
this stage.

4. WAITING: A process is said to be waiting when the process is blocked for some
time due to some reason. In other words, the process switches (jumps) from the
RUNNING stage to the BLOCKED stage for some time. A process that cannot execute
until some event happens, the event may be the completion of an Input or Output
operation. Waiting is a frequently used alternative term for Blocked as a process state.
There are various reasons for the blocking of a process like
I/O Required, Time Slice Elapses, Higher Priority Job Arrives, etc.

5. TERMINATED: A process is supposed to be in the TERMINATED state when

the process completes its execution successfully. A process which released from the pool
of executable processes by the Operating System, either because it paused or because it
was abandoned for some reason.

There are two queues now: ready queue and blocked queue
When the process is admitted in the system, it is placed in the ready queue and when
a process is removed from the processor, it is either placed in the ready queue or a
blocked queue (depending on circumstances). If event time out occurs, then it moves to
ready queue, and if event wait occurs then it moves to blocked queue. When an event
occurs, all the processes waiting on that event are moved from the blocked queue onto the
ready queue.

The New and Exit states are useful constructs for process management. The New
state refers to a process when it has been just defined. For example, if a new user starts to
log onto a system or a new batch job is submitted to the OS for execution, the OS can
define a new process in two stages. First, the OS performs the necessary housekeeping
chores. An identifier is associated with the process. Any tables that will be needed to
manage the process are allocated and built. At this point, the process is in a New state.
This means that the OS has performed the necessary actions to create the process but has
not committed itself to the execution of the process. For example, the OS may limit the
number of processes that may be in the system for reasons of performance or main
memory limitation. While a process is in the new state, information concerning the
37
 process that is needed by the OS is maintained in control tables in the main memory.
However, the process itself is not in the main memory. That is, the code of the program to

be executed is not in the main memory, and no space has been allocated for the data
associated with that program.
While the process is in the New state, the program remains in secondary storage,
typically disk storage Similarly, a process exits a system in two stages. First, a process is
terminated once it reaches its usual completion point, when it aborts due to an
unrecoverable error, or when another process with the appropriate authority causes the
process to abort. Termination moves the process to the exit state. At this point, the
process is no longer eligible for execution. The tables and other information associated
with the job are temporarily preserved by the OS, which provides time for auxiliary or
support programs to extract any needed information. For example, an accounting
program may need to record the processor time and other resources utilized by the
process for billing purposes. A utility program may need to extract information about the
history of the process for purposes related to performance or utilization analysis. Once
these programs have extracted the needed information, the OS no longer needs to
maintain any data relating to the process and the process is deleted from the system.
Memory committed to existing processes. This limit assures that there are not so many
active processes as to degrade performance.

MULTI-THREADED PROGRAMMING

A thread is a single sequential flow of control within a program. A process is defined

sometimes as a heavyweight process and a thread is defined as a lightweight process. A
thread belongs only to one process. It is a unit of computation associated with a
particular heavyweight process, using many of the associated process‘s resources. It has a
minimum internal state and a minimum of allocated resources. Threads can share the
same resources (files, memory space, etc) which are in that process. Thread can be
created faster as compare to the process. Threads are widely used in real-time operating
systems and modern operating systems. Each thread has its control block, with a state
(Running/Blocked/etc.), saved registers, instruction pointer. Threads improve application
performance through parallelism. This concept is useful in a server-client environment. A
process can have a single thread or multiple threads. In multiple threads, each thread is
38
associated precisely to one process and it always inside of the process, which means no
thread can exist outside a process.

Figure: Representation of Single Thread Process

Figure: Representation of Multi-Thread Process

Advantages of Threads
Thread minimizes context switching time.
Threads help in the parallel execution of an application on shared-memory
multiprocessors.
39
 The benefits of multi-threading can be greatly increased in a multiprocessor
architecture.
It is more cost-effective to create and context switch threads.
Threads are dependent on each other so it has efficient communication.
Threads improve application performance through parallelism.
Threads can be created at the user level and the kernel level so it is implemented by
two ways:
(i) User Level (ii) Kernel Level

(i) User Level

In the user-level thread, all of the work of thread management is done by the
application and the Kernel does not aware of its existence. In user space, the thread
library contains code for creating and destroying threads, message passing, data,
scheduling of thread execution, etc.

Characteristic of User Level Threads

It is generally fast to create and easy to manage.
User-type threads can run on any OS.
A multithreaded application cannot take advantage of multiprocessing.
Scheduling can be application-specific.

(ii) Kernel Level

In Kernel-level thread, all of the work of thread management is done by the Kernel.
Kernel threads are directly supported by the operating system. The kernel maintains
context information for the process as a whole and individual threads within the process.
In Kernel space, thread library contains code for creating and destroying threads, message
passing, data, scheduling of thread execution, etc.

Characteristic of Kernel Level threads

 In the kernel-level thread, scheduling is on a thread basis.

 It is generally slow to create.
 The kernel can schedule multiple threads from the same process on multiple
processes.
40
  Kernel routines themselves can be multithreaded.
 Context switching between threads is time-consuming.

Comparison between process and thread

Process Thread
The process is termed the heavyweight Thread is termed lightweight process
process

Processes are independent of one another Threads are not independent of one another
In multiple process implementations, each All threads can share the same set of open
the process executes the same code but has files, child processes.
its memory and file resources

It takes more time to create a new process. It takes less time to create a new thread.

It takes more time to terminate a process It takes less time to terminate a thread

While context switching needs the interface In the thread, switching does not need to
with the operating system. call an OS and cause an interrupt to the
kernel.
It takes more time to switch between two It takes less time to switch between two
processes threads

Difference between User-level thread and Kernel level thread

User Level Thread Kernel level thread

It is maintained at the user level. It is maintained at the Kernel level.

User-level threads are quicker to create Kernel level threads are not quickly created

It is easy and speedily managed It is slowly managed.

It runs on any Operating System These are specific to Operating System.

Implemented by a thread library at the user OS support directly to Kernel threads.
level

Daemon

A Daemon is a system process. It is created at boot time and keeps executing in the

41
background. A daemon is created to perform specific tasks. It gets activated
automatically when a request is received for performing a particular task. After
completion of the task, it again goes back to the background.

MULTI-THREADED MODELS

Some operating system allows the facility of both level threads user-level thread and
Kernel level thread. An example of this combined approach is Solaris. Multithreading
permits the accomplishment of multiple parts of a program run in parallel. These parts of
a program are called threads. It was also known as the lightweight process. By using
multitasking and multithreading, CPU utilization increased.

Multithreading models are 3 models namely:

one-one model, many to one model, and many to many models which are described
below:

One to One Model

There is 1 to1 association between both the thread's user-level thread and kernel-level
thread. This model offers more concurrency than the M:1 model. It also permits another
thread to complete its work when a thread makes a blocking system call. A major
disadvantage of this is when a user-level thread is created then the requirement of
corresponding kernel thread. Therefore a lot of kernel-level threads are required which is
a burden on the system, but there is a limit on the number of threads in the system.

A diagram that reveals the one-to-one model is given below –

42
 Many to One Model

The many to one model, in this model many of the user threads maps to a single kernel
thread. This model is relatively efficient as compare to the 1:1 model as in this the user
space was managed by the thread management.

A disadvantage of this model is that when a thread blocking system calls then it blocks
the whole process. Another disadvantage is that multiple threads cannot run
simultaneously because one thread can access the kernel at a single time.

A given below diagram that shows the many to one model:-

Many to Many Model

In the many-to-many model, any number of user threads correspond with any number of
kernel-level threads. There is no limit. So there is no disadvantage as compared to other
models. These threads can execute simultaneously on a multiprocessor.
CPU SCHEDULING
Scheduling refers to the set of policies and mechanisms that an OS supports for
determining the order of execution of pending jobs and processes. A scheduler is an OS
Module that determines the next pending job to be admitted into the system for execution
or the next ready process to be dispatched to RUN state.

43
Introduction

In the multiprogramming OS, the method or procedure for switching the CPU among
multiple processes is called CPU scheduling. The CPU scheduler is a part of an OS,
which is responsible for CPU scheduling. When a process performs an I/O-related task, it
does not use certain resources, such as CPU and these resources remain idle.
CPU scheduling enables processes to utilize idle resources by assigning them to
other processes. Whenever CPU becomes idle the CPU scheduler chooses a process
among the processes which are in the ready queue and sends the process to the CPU for
execution.
For example, process A is running and needs to perform an I/O-related task. Process
A does not need the CPU while performing the I/O-related task. Process scheduler
changes the state of process A from running to waiting and enables process B to use the
CPU.

44
 Figure: CPU Scheduling

For example, in the process of CPU scheduling, two processes, A and B, share the
CPU times. When process A uses the CPU, process B performing I/O operations, and
when the CPU is accessed by process B, Process A performs I/O operations.

Goals of Scheduling
(i) To optimize the utilization of system resources.
(ii) To ensure that more critical processes get priority over others processes.
(iii) To provide as fair a deal as possible to all jobs and processes which are pending.

When a process, in the ready state, is allocated the CPU in the place of a partially
done process, the value of the various field of the partially done process, such as process
state and I/O status is updated in the Process Control Block of the partially done process.
The OS marks the partially done process as ready and sends it back to the ready queue.
The OS reads the data from the process control block of the ready process and allocates
CPU to it and marks the process state running. This is called context switching.
In other words, ―Context switching‖ refer to the process of transferring control of
the CPU from the currently running process to another process from the ready queue.
For example, a time-out has occurred for process A. Process A will be released.
Process B is to be allocated the CPU. The steps performed by the dispatcher in context

45
 switching are:
It retrieves and analyzes information about process A regarding program
counter,memory, and registers.
It updates the PCB of process A by writing the new values of the various
fields of PCB. This includes changing the state of process A from running to
ready or waiting.
It moves process A to the appropriate queue. If the state of process A is changed
into a waiting state, it is inserted in the waiting queue. If process A is
transformed into the ready state, it is inserted into the ready queue.
It retrieves and analyzes information about process B from the PCB of process B
regarding program counter, memory, and registers.
It allocates the CPU time to process B
It restores the environment values, such as program counter, memory, and
registers of process B.

The context switching is pure overhead. The extent of this overhead depends on the size
of the process content. The larger the process content, the higher will be the context-
switching overhead.

CPU SCHEDULING CRITERIA

Factors for measuring the performance of an operating system

There are certain factors on which the performance of an operating system depends. The
efficiency and overall performance of the operating system can be measured in terms of
the following factors:

(a) CPU utilization: CPU utilization refers to the usage of the processor during the
execution of a process.CPU utilization may vary from 40% to 90%.CPU should

46
 remain as busy as possible. So, CPU utilization should be the maximum for the
better efficiency of a system.

(b) Throughput: Throughput may be defined as the total number of processes

completed per unit of time. It is the measure of work done by the CPU. It is
expressed in terms of the number of jobs done in a given unit of time. It is
important to know that the value of throughput does not depend only on the
capability of the system but also on the nature of jobs, so, it may vary
accordingly. If the jobs are CPU bound then the throughput will be less and if
the jobs are I/O bound, the throughput will be high.
If the jobs are long and heavy, the throughput may be one or two processes per
hour. On the other hand, if the jobs are short and light, the throughput may be
100 processes per hour. However, it should be as maximum as possible.

(c) Turnaround time: Turnaround time is the time which is the difference between
the time of submission or entered and the time of ending of the job. It is a metric
for batch systems. Turnaround time is the sum of the following components:

The time spends in waiting for entry into the system.

Total time spent in the ready queue.
Total time spent in the device queue.
Time spent in the execution of the process.
Time spend in doing the Input/Output operation.
However, turnaround time should be as minimum as possible.

(d) Response time: Response time is defined as the difference between the
time of submission of the job for processing and the time when it gets the first
response of the system. It is considered the best metric for interactive systems.

(e) Waiting time: Waiting time may be defined as the sum of intervals for which a
process has to wait in the ready queue. It should be minimum for the better
efficiency of the system. The scheduling strategies try to minimize the waiting
time for the processes.
Average Turnaround Time: The turnaround time of a process is the total time elapsed
from the time the process is submitted to the time the process is completed. It is

47
calculated as

Turnaround time (TAT) = Process finish time (T1) – Process Arrwal time (T0)
Average Turnaround Time =
The lower the average turnaround time, the better it is.

Average Waiting Time: Waiting time of a process is defined as the total time spent
by the process while waiting in a ready state or suspended state.

Waiting time (WT) = Turnaround time (TAT) – Actual Execution time (

Average Waiting Time =
The lower the average waiting time, the better it is.
For the better performance and efficiency of a system the CPU utilization should be maximum,
Response time should be minimum, Waiting time should be minimum, Throughput should be
maximum and the Turnaround time should be minimum.
DISPATCHER

The dispatcher is a module of an OS that gives control of the CPU to the process selected
by the scheduler. The dispatcher is responsible for the context switching. The time taken
by the dispatcher for halting the running process and starting the process selected by the
scheduler for executions is called dispatch latency.

SCHEDULING STRATEGIES

Scheduling strategies refer to the various algorithms used to choose a particular process
from among the various processes in the ready queue.
There may be several processes waiting for the attention of the CPU in the ready
queue. The operating system chooses a particular process from the ready queue and
allocates the processor to it. So, the operating system requires a mechanism to decide,
which process should be chosen next for execution.

There are two types of jobs:

CPU Bound Jobs: The jobs, that spend more time performing CPU operations
and less I/O operations are called CPU-bound jobs/processes.

I/O Bound Jobs: The jobs that spend more time performing I/O operations and
48
 less time doing CPU operations are called I/O bound jobs/processes.

CPU Scheduling can be preemptive or non-preemptive.

In preemptive scheduling, the scheduler removes the running process from the
CPU before its completion so that another process can run. In other words, in preemptive
scheduling, the running process only gives up the control of the CPU voluntarily.

In non-preemptive scheduling, nothing can remove a process from utilizing CPU

time until it completes or a time-out occurs.
There are several strategies used for CPU Scheduling, called Scheduling Strategies
like:
1. FIRST COMES FIRST SERVE (FCFS)
2. SHORTEST JOB FIRST (SJF) (Preemptive and Non-preemptive)
3. PRIORITY SCHEDULING (Preemptive and Non-preemptive)
4. ROUND ROBIN SCHEDULING (RR SCHEDULING)

The detailed description of the scheduling strategies is as follows:

First Come First Serve (FCFS)

It is the simplest of all the scheduling algorithms. It is purely non-preemptive
scheduling. The key concept of this algorithm is:

“allocate the CPU to the processes in the order in which they arrive”.
According to this algorithm, the process which arrives first will get the CPU before
any other process. Same way, the process which arrives as a second, will get the CPU
after the first process and so on. It assumes the Ready Queue as the FIFO QUEUE. When
a process completes its execution, the CPU is allocated to that process which is the first
process in queue i.e the FRONT of the queue. If a new process arrived then it enters in
REAR of the queue.

When a process starts running then it is removed from the queue. This algorithm is
NON-PREEMPTIVE by default. This means once the CPU is assigned to a process, that
process keeps the CPU till the end of its execution.

49
Advantages:
The code for FCFS scheduling is simple to write and understand.
It is suitable for Batch systems.
It is considered to be a fair policy, as the job which arrives first will get the CPU
first.

Disadvantages:
If shorter jobs arrive after the longer jobs, then the waiting time will be large.
It is not suitable for Time-sharing systems.
Low CPU utilization, because all the other processes wait for one long job to get
off the CPU.
This algorithm is never recommended whenever performance is a major issue.

Example:
Consider the following snapshot of processes that arrive at a different time with the CPU-
burst in milliseconds.

Note: Ignoring the arrival time in non-preemptive scheduling, the lowest number has the
highest priority and time slice 2ms.

Process ID Arrival Time (T0) Priority Next CPU Burst

ms Time (Δt) ms

P1 0 2 16

P2 1 3 8

P3 2 1 2

The Gantt chart of execution of the processes according to FCFS is the following:

P1 P2 P3
0 16 24 26
16 8
2

50
 Waiting time for P1 = 0 ms
Waiting time for P2 = 16 ms

Waiting time for P3 = 24 ms

Table depicting performance of FCFS

Process ID Arrival Time Next CPU Finish Time Turnaround Waiting

(T0) ms Burst Time (T1) ms Time Time=
(Δt) ms TAT=T1-T0 TAT-Δt

(ms) (ms)

P1 0 16 16 16 0

P2 0 8 24 24 16

P3 0 2 26 26 24

66 40

As mentioned in question ignore the arrival time in non-preemptive scheduling so,

here arrival time mentioned 0

Total waiting time = 0+16+24 = 40 milliseconds

Average waiting time = 40/3 = 13.33 milliseconds.
Total turnaround time = 16+24+26 = 66 milliseconds
Average turnaround time = 66/3 = 22 milliseconds.
However, if the jobs arrive in a different order, then the waiting time and turnaround
time can be reduced. So, the waiting time and turnaround time depend upon the order of
the jobs in which they arrive.

Shortest Job First (SJF)

The key concept of this algorithm is

“allocate the processor to the job which has the least CPU burst time”.
The ready queue has all the processes which require the processor for their
execution. According to this algorithm, the processor is allocated to that job that has the
51
smallest CPU burst time amongst all the processes in the ready queue.
If two processes have the same CPU burst, then the processor is allocated to the
process which arrives first. This algorithm can either be PREEMPTIVE or NON-
PREEMPTIVE.

Advantages
This is considered to be an optimal algorithm as it helps to achieve the
minimum waiting time.
The shorter jobs have to wait for less time as compared to the longer jobs.

Disadvantages
There is a need for the mechanism to know about the CPU burst of all the
processes in advance.
If the shorter jobs arrive again and again, the longer jobs may wait for a longperiod.

Example

NON-PREEMPTIVE SJF
Consider the following snapshot of processes that arrive at a different time with the CPU-
burst in milliseconds.

Note: Ignoring the arrival time in non-preemptive scheduling, the lowest number has the
highest priority and time slice 2ms.

Process ID Arrival Time (T0) Priority Next CPU Burst

ms Time (Δt) ms

P1 0 2 16

P2 1 3 8

P3 2 1 2
The Gantt chart of execution of the processes according to NON-PREEMPTIVE SJF is
the following:

P3 P2 P1
0 2 10 26
2 16
8
52
 Waiting time for P1 = 10 ms
Waiting time for P2 = 2 ms
Waiting time for P3 = 0 ms

Table depicting performance of non-preemptive SJF

Process ID Arrival Time Next CPU Finish Time Turnaround Waiting

(T0) ms Burst Time (T1) ms Time Time=
(Δt) ms TAT=T1-T0 TAT-Δt

(ms) (ms)

P1 0 16 26 26 10

P2 0 8 10 10 2

P3 0 2 2 2 0

38 12

Total waiting time = 10+2+0 = 12 milliseconds

Average waiting time = 12/3 =4 milliseconds.

Total turnaround time = 26+10+2 = 38 milliseconds

Average turnaround time =38/3 = 12.6 milliseconds.

PREEMPTIVE SJF: It is also termed as Shortest Remaining Time Next (SRTN) and
Shortest Remaining Time First(SRTF).

Consider the set of the following processes with the CPU-burst in milliseconds.

Process ID Arrival Time (T0) Priority Next CPU Burst

ms Time (Δt) ms

P1 0 2 16

P2 1 3 8

P3 2 1 2
53
Here, first of all, P1 arrives and the processor is allocated to it. After one second P2
arrives, since P2 has a smaller CPU burst as compare to P1, therefore the processor is
preempted from P1 and is allocated to P2. Same way, after one second, P3 arrives whose
CPU burst is less than the P2, so the processor is preempted from P2 and allocated to the
P3. After the execution of P3, the processor is again allocated to P2 and so on.
The Gantt chart of execution of the processes according to PRE EMPTIVE SJF is
the following:

P1 P2 P3 P2 P1

0 1 2 4 11 26
1 2 7 15
1

Table depicting performance of preemptive SJF

Process ID Arrival Time Next CPU Finish Time Turnaround Waiting

(T0) ms Burst Time (T1) ms Time Time=
(Δt) ms TAT=T1-T0 TAT-Δt

(ms) (ms)

P1 0 16 26 26 10

P2 1 8 11 10 2

P3 2 2 4 2 0

38 12

Total waiting time = 10 + 2 + 0 = 12 milliseconds

Average waiting time =12/3 = 4 milliseconds.
Total turnaround time = 26 + 10 + 2 = 38 milliseconds
Average turnaround time =38/3 = 12.6 milliseconds.

54
 In the case where it is not possible to know the CPU time for each process, this is
estimated using predictors:

– Pn = aOn-1 + (1-a)Pn-1 where

• On-1 = previous service time

• Pn-1 = previous predictor

• a is within [0,1] range

– If a = 1 then Pn-1 is ignored

– Pn is dependent upon the history of the process evolution

Priority Scheduling
Priority scheduling is the scheduling mechanism in which each process in the system is
assigned a priority. The processor is allocated to the processes according to their priority.
The key concept of this algorithm is

―Allocate the processor to the process which has higher priority”.

The ready queue is assumed to be a priority queue in which each process is assigned
a priority. First of all, the processor is allocated to the process having higher priority and
then to the process having lower priority and so on. Priority scheduling can be of two
types:

PREEMPTIVE or NON-PREEMPTIVE

Example:
NON-PREEMPTIVE PRIORITY SCHEDULING

Consider the set of the following processes with the CPU-burst in milliseconds.

Note: Ignoring the arrival time in non-preemptive scheduling, the lowest number has the
highest priority and time slice 2ms.

Process ID Arrival Time (T0) Priority Next CPU Burst

ms Time (Δt) ms

55
 P1 0 2 16

P2 1 3 8

P3 2 1 2

The Gantt chart for the execution of the programs is as follows:

P3 P1 P2
0 2 18 26
2 16 8

Table depicting performance of non-preemptive priority scheduling

Process ID Arrival Time Next CPU Finish Turnaround Waiting

(T0) ms Burst Time Time (T1) Time Time= TAT-
(Δt) ms ms TAT=T1-T0 (ms) Δt (ms)

P1 0 16 18 18 2

P2 0 8 26 26 18

P3 0 2 2 2 0

46 20
Waiting time for P1 = 2 ms
Waiting time for P2 = 18 ms
Waiting time for P3 = 0 ms

Total waiting time = 2 +18 + 0 = 20 milliseconds

Average waiting time =20/3 = 6.6 milliseconds.
Total turnaround time = 18 + 26 + 2 = 46 milliseconds
Average turnaround time =46/3 = 15.3 milliseconds.

PREEMPTIVE PRIORITY SCHEDULING

Consider the set of the following processes with the CPU-burst in milliseconds.

Process ID Arrival Time (T0) Priority Next CPU Burst

56
 ms Time (Δt) ms

P1 0 2 16

P2 1 3 8

P3 2 1 2

The Gantt chart for the execution of the programs is as follows:

P1 P3 P1 P2

0 2 2 4 14 18 8 26
2

Table depicting performance of preemptive priority scheduling

Process ID Arrival Time Next CPU Finish Time Turnaround Waiting

(T0) ms Burst Time (T1) ms Time Time=
(Δt) ms TAT=T1-T0 TAT-Δt

(ms) (ms)

P1 0 16 18 18 2

P2 1 8 26 25 17

P3 2 2 4 2 0

46 19

Total waiting time = 2 + 17 + 0 = 19 milliseconds

Average waiting time =19/3 = 6.3 milliseconds.
Total turnaround time = 18 + 25 + 2 = 45 milliseconds
Average turnaround time =45/3 = 15 milliseconds.

The problem in Preemptive Priority Scheduling (Starvation or Blocking)

The main problem with priority scheduling is that low-priority processes can wait
for a long time due to high-priority process arrivals. If an Operating system has not any
precautions and just chooses the process with the highest priority, low priority processes
57
wouldn‘t get CPU, as long as high priority processes are runnable. This problem is
known as Starvation.

Solution of Starvation (AGING)

Aging is a method of slowly increasing the priority of processes that are waiting in
the system for a long time. The simplest solution is dynamic priorities. On one hand, the
operating system can reduce the priority of a running process for each time quantum it
used the CPU and on the other hand, it could increase the priority of other processes
(which finally leads to the first situation because the boost should only be temporary)
which didn‘t get the CPU for a certain amount of time. Whether the operating system
uses one or another solution, processes have a base priority that remains unchanged, and
a real priority that is used for scheduling. Often the real priority is limited to a specific
range so that important process still gets the processor when they need it.
Another way to avoid this problem is static priorities. The OS has to keep a record
of how long a process has used the CPU. If it reaches a certain limit, the next highest
priority process is allowed to run. This is not practically good.

Round Robin Algorithm (RR Algo)

The Round Robin scheduling algorithm is designed especially for the Time Sharing
systems. It can be considered as FCFS scheduling along with the preemption. The
processor is allocated to a process for a fixed amount of time called, TIME SLOT or
TIME QUANTUM, or, TIME SLICE. It is a purely preemptive algorithm. Likewise
FCFS, the processor is allocated to the processes in the order in which they arrive, but the
ed from the process after the time slice is over, and the processor is allocated to the next
process in the ready queue. A time quantum generally varies from 10 milliseconds to 100
milliseconds. The ready queue is assumed to be a circular queue. Theshort-term scheduler
goes on allocating the processor to the processes in the ready queue for a fixed amount of
time.

However, if a process has its CPU burst less than the time quantum, then the process
releases the CPU voluntarily (itself). Otherwise, the processor is allocated to the process
in the ready queue for a fixed amount of time (time quantum) and after that CONTEXT
SWITCHING take place, and the processor is allocated to the next process in the ready
58
 queue, keeping the former process at the end of the ready queue. The RR algorithm is
PREEMPTIVE by default.

The performance of the Round Robin algorithm depends on the value of time
quantum or slice
If the value of time quantum is large, then this algorithm becomes the same as
FCFS.
If the value of the time quantum is small, then the number of the context
switching will be increased considerably, this is not desirable at all. It affects the
system throughput adversely.
Thus, the size of the time quantum should neither be very large nor too small
forbetter efficiency.

Example:
Consider the set of the following processes with the CPU-burst in milliseconds.
Note: Ignoring the arrival time in non-preemptive scheduling, the lowest number has the
highest priority and time slice 2ms.

Process ID Arrival Time (T0) Priority Next CPU Burst

ms Time (Δt) ms

P1 0 2 16

P2 1 3 8

P3 2 1 2
The Gantt chart of execution of the processes according to the RR scheduling algorithm
is the following. Let the time quantum be 2 milliseconds-:

P1 P2 P3 P1 P2 P1 P2 P1 P2 P1

0 2 4 6 8 10 12 14 16 18
26 2 2 2 2 8
2 2 2 2 2

59
 Table depicting performance of Round Robin scheduling

Process ID Arrival Time Next CPU Finish Time Turnaround Waiting

(T0) ms Burst Time (T1) ms Time Time=
(Δt) ms TAT=T1-T0 TAT-Δt

(ms) (ms)

P1 0 16 26 26 10

P2 1 8 18 17 9

P3 2 2 6 4 2

47 21

Waiting time for P1 = 10 ms

Waiting time for P2 = 9 ms
Waiting time for P3 = 2 ms

Total waiting time = 10 + 9 + 2 = 21 milliseconds

Average waiting time = 21/3 = 7 milliseconds.
Total turnaround time = 26 + 17 + 4 = 47 milliseconds
Average turnaround time = =47/3 = 15.6 milliseconds.

Comparison between FCFS and Round Robin Scheduling Algorithm

FCFS Round Robin

FCFS is purely non-preemptive Scheduling It is purely preemptive Scheduling

It has minimum overhead It has a higher overhead as compare to

FCFS as context switching occurs more.
Response time depends upon the size of the It offers a better response time
process

It is not designed for a time-sharing system It is designed for a time-sharing system

The workload is simply processed in the It is similar to FCFS but it uses a time slice
order process arrived which means one process maximum uses
60
 CPU for time slice at once.

Multilevel Queue Scheduling

A multilevel queue scheduling algorithm is used when the processes are to be divided
into different groups. Sometimes such a situation arises where two or more processes
have different response time requirements, so there is a need for a different kind of
schedule for both processes.
A multilevel queue scheduling division of the ready queue into numerous separate
queues. The processes are permanently allocated to a particular queue, based on the
properties of the process.
Sometimes there are two or more processes, among which some processes have
higher priority and other having low priority. So, there is a need for a separate queue for
the processes with higher priority and the processes with lower priority.

Consider an example of multilevel queue scheduling for different kinds of processes

in the ready queue. There are four different queues for different processes which are
made based on their priority in the system.

Highest priority queue SYSTEM PROCESSES

INTERACTIVE PROCESSES

BATCH PROCESSES

61
 Lowest priority Queue STUDENT PROCESSES

Each queue may have its scheduling criteria, based upon the nature and properties of the
jobs.

Advantages

Low scheduling overhead can be achieved by using a multilevel scheduling

algorithm.

Disadvantages
It is not considered as efficient in some of the cases, like if higher priority
queues don‘t become empty for a long time, then the lower priority jobs may
starve.

Multilevel Feedback Queue Scheduling

It is an enhancement of multilevel queue scheduling.
Actually, in multilevel queue scheduling, the processes are assigned a particular
queue on entering the system. The processes do not move from a queue to any other
queue. So, it is an inflexible approach.
Multilevel feedback queue scheduling permits a process to move between the
queues. For example, if a process CPU-bound process which means it uses too much
CPU then the process is moved (shifted) to a lower priority queue. Correspondingly, a
lower priority job can also be shifted to a higher priority queue if it is waiting for a long
time, etc.

Advantages
o It allows a process to move to any other queue.
o It is more flexible as compared to any other scheduling algorithm.

62
 o A lower priority job, which is waiting for a long time, can be shifted to a higher
priority queue.
o A higher priority job, affecting negatively the efficiency, can be moved to a
lower priority queue.

Multilevel Queue Scheduling

Disadvantages
o However, it is considered to be a complex scheduling algorithm.
o While moving the processes from a queue to any other queue increases the CPU
overhead.

Illustration

With reference to the following set of processes

Process ID Arrival Time (T0) Next CPU Burst Priority

ms Time (Δt) ms

P1 0 29 1

P2 6 14 5

63
 P3 8 10 3

P4 10 8 1

P5 13 6 2

Determine Average Waiting Time and Turn Around Time, for the following scheduling.

(a) First Come First Serve (FCFS)

(b) Shortest Job First (SJF) (Non-preemptive)
(c) Shortest Job First (Preemptive) or Shortest Remaining Time Next (SRTN)
(d) Priority Scheduling (Preemptive)
(e) Priority Scheduling (Non-preemptive)

(f) Round Robin (RR)

Note: Time Slice – 5 ms, the highest number has the highest priority and ignores the
arrival time in non-preemptive scheduling.

(a) First Come First Serve (FCFS)

In this scheduling, jobs or processes are scheduled to run in the same order as those
that have arrived in the system. It's a pure non-preemptive algorithm. Process P1. will
execute first because its ID is 1, then P2, P3, P4, and P5. So its Gantt Chart is given
below:

P1 P2 P3 P4 P5

0 29 43 53 61
29 14 10 8
67 6

Calculating Waiting Time and Turnaround Time

In this problem, in a non-preemptive algorithm arrival time is ignored so, waiting

time for all processes is when they start to run and turnaround time is when they complete
64
the process.

Note Arrival time 0 in this problem because in non-preemptive algorithm arrival time is
ignored.

Table depicting performance of FCFS

Process ID Arrival Time Next CPU Finish Time Turnaround Waiting

(T0) ms Burst Time (T1) ms Time Time=
(Δt) ms TAT=T1-T0 TAT-Δt

(ms) (ms)

P1 0 29 29 29 0

P2 0 14 14 43 29

P3 0 10 10 53 43

P4 0 8 8 61 53

P5 0 6 6 67 61

253 186

Average Waiting Time = (0+29+43+53+61)/5=37.2 ms

Average Turnaround Time = (29+43+53+61+67)/5=50.6ms

(b) Shortest Job First (SJF)

In shortest job first, job to be dispatch will be the one, which happens to be the
shortest amongst the pending jobs. This is non-preemptive. So a job, once scheduled, is
permitted to complete its next burst. In this problem, Arrival time is ignored for non-
preemptive scheduling so, P5 has the shortest next burst time hence it will run first then
P4, P3, P2 & P1 will run. So its Gantt chart is

P5 P4 P3 P2 P1

0 6 14 24 38 67
8 10 14 29
6
65
Calculation of Waiting Time & Turnaround Time
In this problem, in a non-preemptive algorithm arrival time is ignored so, waiting
time for all processes is when they start to run and turnaround time is when they complete
the process.

Table depicting performance of non-preemptive SJF scheduling

Process ID Arrival Time Next CPU Finish Time Turnaround Waiting

(T0) ms Burst Time (T1) ms Time Time=
(Δt) ms TAT=T1-T0 TAT-Δt

(ms) (ms)

P1 0 29 67 67 38

P2 0 14 38 38 24

P3 0 10 24 24 14

P4 0 8 14 14 6

P5 0 6 6 6 0

149 82

Note: Arrival time is 0 in this problem because non-preemptive algorithm arrival time is
ignored.

Average Waiting Time = (38+24+14+6+0)/5=16.4 ms

Average Turnaround Time = (67+38+24+14+6)/5=29.8 ms

(c) Shortest Remaining Time Next (SRTN)

This is a preemptive algorithm, where the next job/process to be dispatched will be

one that happens to be shortest amongst the pending jobs, at the time of making the
decision. However, if a process/job arrives later, whose next burst time to be less than the
remaining burst time of the currently running process, the currently running process will
be preempted by the new process. The preempted process will be later re-dispatched
when its remaining burst happens to be shortest amongst the pending processes.

66
 In this problem, at 0 arrival time, only one process is there i.e. P1. So, P1 will
execute first. After 6ms, a new process P2 arrives which has a 14ms burst time and P1
has 23ms (29-6) remaining burst time. So, P1 will be preempted by the P2 because P2 has
the shortest burst time at 6ms. After 2ms i.e. 8ms a new process P3 arrives which has
10ms burst time and P2 has 12ms (14-2) remains burst time P3 has the shortest burst time
at 8ms. So, P2 will be preemptive by P3. After 10ms a new process P4 arrives which has
an 8ms burst time and P3 has 8ms (10-2) remains burst time. So there is a tie between P3
and P4. To break tie FCFS algorithm is adopted so, P3 will continue to execute. At time
13ms a new process, P5 arrived with 6ms burst time. At that time P3 has (10-(13-8)) i.e.
5ms remaining burst time which shortest than among all-time so P3 will continue. After
completion of P3 at 18ms time. There is no new process arrived. Now remains burst time
for a process are P1 is 23ms, P2 is 12ms, P4 is 8ms and P5 is 6ms. So amongst P5 is
shortest than P4, P2 and P1 will execute. Gantt chart of this is given below:

P1 P2 P3 P5 P4 P2 P1

6 2 10 6 8 12 23

0 6 8 18 24 32 44 67

Table Depicting Performance of SRTN Algorithm

Process ID Arrival Time Next CPU Finish Time Turnaround Waiting

(T0) ms Burst Time (T1) ms Time Time=
(Δt) ms TAT=T1-T0 TAT-Δt

(ms) (ms)

P1 0 29 67 67 38

P2 6 14 44 38 24

P3 8 10 18 10 0

P4 10 8 32 22 14

P5 13 6 24 11 5
67
 148 81

Average Waiting Time =(38+24+0+14+5)/5=16.2 ms

Average Turnaround Time = (67+38+10+22+11)/5= 29.6 ms

(d) Priority Scheduling (Preemptive)

At the time of schedules, a dispatcher dispatched that process which has the highest
priority amongst the processes which are waiting in the ready queue. When a process Pi
is executing, if another process Pj arrived and has higher priority, then Pi will be
preempted by Pj.

In this problem, at time 0ms only one process P1 is there so it is executed till a new
process has arrived i.e. 6ms. After 6ms a new process P2 comes which has priority 5
which is greater than P1 priority i.e. 2 so, P1 will be preempted by P2. Now P2 will
execute at 8ms a new process P3 arrive which have priority 3 but P2 have the highest
priority at that time so, P2 will continue, after 10ms P4 arrive which have priority value 1
which lowest so, P2 will continue. After that at 13ms P5 comes whose priority is lower
than P2. So P2 will continue till 20ms i.e. (6 + 14).
Now there are 4 processes at 20ms P1, P3, P4, P5 are ready and priorities are 1, 3, 1,
2 respectively. Among them, 3 is the highest priority. So, P3 will execute then P5. After
that P1, P4 have the same priority so, there is a tie. To break the tie FCFS algorithm was
adopted so P1 executes than P4. Gantt chart of this algorithm is:

P1 P2 P3 P5 P1 P4

0 6 20 30 36 59
6 67 14 10 6 23 8

Table depicting performance of a priority-based preemptive algorithm.

68
 Process ID Arrival Time Next CPU Finish Time Turnaround Waiting
(T0) ms Burst Time (T1) ms Time Time=
(Δt) ms TAT=T1-T0 TAT-Δt

(ms) (ms)

P1 0 29 59 59 30

P2 6 14 20 14 0

P3 8 10 30 22 12

P4 10 8 67 57 49

P5 13 6 36 23 17

175 108

Average Waiting Time = (30+0+12+49+17)/5=21.6 ms

Average Turnaround Time = (59+14+22+57+23)/5=35 ms

(e) Priority Scheduling (Non-preemptive)

At the time of scheduling, a process that has the highest priority amongst all the
processes which are waiting in the ready queue. Once dispatched, a process Pi is allowed
to complete its burst time, even if Pj another process of having higher priority becomes
ready while running Pi.
In this scheduling, arrival time is ignored so amongst all processes P2 has the
highest priority which will execute first. Priority of above processes is:

P2 > P3 > P5 > P1 > P4

Gantt chart is:

P2 P3 P5 P1 P4

29
14 5 10 6
8

0 14 24 30 59 67

69
 Table depicting performance of a priority-based non-preemptive algorithm.

Process ID Arrival Time Next CPU Finish Time Turnaround Waiting

(T0) ms Burst Time (T1) ms Time Time=
(Δt) ms TAT=T1-T0 TAT-Δt

(ms) (ms)

P1 0 29 59 59 30

P2 0 14 14 14 0

P3 0 10 24 24 14

P4 0 8 67 67 59

P5 0 6 30 30 24

194 127
Note: Arrival time is 0 in this problem because non-preemptive algorithm arrival time
is ignored.
Average Waiting Time = (30+0+14+59+24)/5=25.4 ms
Average Turnaround Time = (59+14+24+67+30)/5=38.8 ms
(f) Round Robin (RR)
It is purely a preemptive scheduling algorithm. A small unit of time is called a time
slice or time quantum. It is the maximum time for which a process can execute at a time.
New processes are added at the end of the ready queue and the head of the ready queue
process Pi is dispatched for the time slice. If the process Pi has not finished its execution,
then it is linked to the tail of the ready queue, and the next process in the ready queue is
dispatched for the next time slice and so on.
In this problem, the time slice is 5ms so P1 will execute for 5ms but there is no more
process at 5ms so, P1 again executes for the next 5ms. After 10ms P2 execute for 5ms
and P3, P4, P5 so on. This process will continue till the process complete their execution.
Gantt Chart is:

P1 P1 P2 P3 P4 P5 P1 P2 P3 P4 P5 P1 P2 P1
70
 49 54 58
5 5 5 5 5 5 3 5 4 9
5 5 5
0 5 10 15 20 25 30 35 40 45 48 67
1

Table depicting performance of round robin algorithm

Process ID Arrival Time Next CPU Finish Time Turnaround Waiting

(T0) ms Burst Time (T1) ms Time Time=
(Δt) ms TAT=T1-T0 TAT-Δt

(ms) (ms)

P1 0 29 67 67 38

P2 6 14 58 52 38

P3 8 10 45 37 27

P4 10 8 48 38 30

P5 13 6 49 36 30

230 163

Average Waiting Time = (38+38+27+30+30)/5 =32.6 ms

Average Turnaround Time = (67+52+37+38+36)/5=46 ms

CPU Scheduling Algorithms analysis

Scheduling Turnaround Waiting Average Average
Algorithm Time Time Turnaround Waiting
Time Time
FCFS 253 186 50.6 37.2
SJF 149 82 29.8 16.4
SRTN 148 81 29.6 16.2
Priority(Preemptive) 175 108 35 21.6
Priority(non- 194 127 38.8 25.4

71
 Preemptive)

Round Robin 230 163 46 32.6

The above table depicts that SRTN scheduling gives better performance among the
other scheduling.

Note: Some time context switching time is given in problem then that time is switching
time and it is pure overhead. It will be increased the waiting time and turnaround time. If
the context switch time is 1ms then in the above example the Gantt chart of FCFS as
below:

P1 P2 P3 P4 P5

0 29 29 30 14 44 45 10 558 56 64 65
71 6

Performance of all CPU Scheduling Algorithms

Scheduling CPU Response Average Average Throughput
Algorithm Utilization Time Waiting Turnaround
Time Time
FCFS Low Low High High Low
SJF Medium Medium Medium Medium High
Priority Medium High High High Low
Round Robin High High Medium Medium Medium
Multi-Level High Medium Medium Medium High

Points to Remember

 A program is a static entity.

 A process is an instance of a program under execution.
 The process is a dynamic entity.

72
 Process state is defined as the current activity of the process.
 In two states process model process has two states: Running or not running.
 In five states process model process has five states: New, Ready, Waiting
Running, and Terminating.
 Each process has its Process Control Board (PCB).
 In-Process Control Board (PCB) attributes and information about the process is
stored which is needed by OS to control the process.
 Schedulers are of three types: Long term Scheduler, Medium Term Scheduler,
Short Term Scheduler.
 Long term scheduler is also known as a job scheduler. It selects the processes
and loads them into memory. It changes process states from new to ready state.
 A short-term scheduler is also known as a CPU scheduler or dispatcher. It
changes process states from ready to running state.
 The process is called the heavyweight process.
 A process can have a single thread as well as multiple threads.
 Thread is called a lightweight process.
 Threads are not independent of one another.
 Threads can be created at a user level and the kernel level.
 A context switch is switching the CPU to another process which requires saving
the state of the old process and loaded the saved state for the new process.
 A Daemon is a system process that is created at boot time and keeps executing in
the background.
 CPU scheduling is used to increase CPU utilization.
 In non-preemptive scheduling, once the process has been assigned to the CPU,
the CPU cannot be taken away from that process until it terminates or is blocked.
 In preemptive scheduling, the CPU can be taken away from the process during
execution.
 Context switching is required in preemptive scheduling.

 Throughput means how many processes the system can execute in a unit of time.

73
 Higher the number, the better it is.
 Response Time is the time from the entry of request until the first response is
produced.
 Finish time means when the process finishes its execution (T1).
 Arrival time is the time when the process arrived (T0).
 Burst Time is the estimated time the process needed CPU for execution (Δt).
 Turnaround time is computed by subtracting the time the process entered the
system from the time it terminated (TAT= T1-T0).
 Waiting Time is time spent by the process for CPU (WT=TAT- Δt).
 CPU-bound process spends most of its time in CPU.
 I/O bound process spends most of its time in I/O operations.
 First Come First Serve (FCFS) is a purely non-preemptive algorithm.
 In FCFS, the CPU is allocated to the process in order of arrival.
 Shortest Job First (SJF) scheduling algorithm can be either a non-preemptive or
preemptive algorithm.
 In SJF, the CPU is allocated to the process which has the smallest burst time.
 Preemptive SJF is known as Shortest Remaining Time Next (SRTN).
 In Shortest Job First (SJF) scheduling (non-preemptive and preemptive) the next
CPU burst time must be known in advance.
 A priority scheduling algorithm can be either a non-preemptive or preemptive
algorithm.
 In priority scheduling, the CPU is allocated to the process which has the highest
priority.
 In preemptive priority scheduling, there is a big problem known as starvation.
 In starvation, blocking of low priority processes due to high priority jobs keep
arriving one after another.
 Aging is a solution to starvation, after some time makes the priority of a process
go up the longer it stays run-able but isn‘t run.
 Round Robin is a purely preemptive algorithm.
 CPU is allocated to all processes in a queue for small-time, which is known as
Time Slice, Time Quantum.
74
  Multilevel queue algorithms allow different algorithms to be used for various
classes of processes.
 Multilevel feedback queues allow processes to move from one queue to another.

EXERCISES

1. What is a Process? Explain different states of a process.

2. Explain the Process State diagram in detail.
3. What do you mean by Process Creation and Termination?
4. What is Process Control Block?
5. Explain the following terms:
(i) Throughput
(ii) Waiting Time

(iii) Turn Around Time

(iv) Response Time
(v) CPU utilization
6. What are the various factors for measuring the performance of an operating system?
7. Explain the Operations on Processes.
8. What do you mean by Thread and explain different types of threads?
9. Difference between Thread and Process.
10. Difference between User Level Thread and Kernel Level Thread.
11. Explain how threads can improve system performance.
12. Discuss the following terms:
Process Spawning
Parent Process
Child Process
Halt
Dispatcher
Ready Queue
Blocked Queue
Daemon

75
13. What do you mean by CPU scheduling and why it is required?
14. What do you mean by I/O bound and CPU bound process?
15. Define the differences between preemptive and non-preemptive scheduling?
16. What are the scheduling criteria? Explain it.
17. Explain the different types of schedulers.
18. Consider the following set of processes

Process Arrival Time Priority Burst Time

P0 0 4 20
P1 1 2 8
P2 3 1 10
P3 5 3 4
P4 8 1 2
P5 9 5 7

Calculate Average Turnaround Time, Average waiting time for the following
algorithms. The time slice is 3 ms and the lowest number has the highest priority for
the following scheduling.
(a) FCFS (b) SJF (Preemptive and non-
preemptive)
(c) Priority (Preemptive and non-preemptive) (d) Round Robin

19. What is Multilevel queue scheduling. Why we use it?

20. Explain between long-term and short-term scheduler.
21. Explain with example the following:
(a) FCFS (b) SJF

22. Explain with example the following:

(a) Priority Scheduling (b) Round Robin Scheduling

76
23. Difference between FCFS and RR scheduling.
24. Consider the following set of processes

Process Arrival Time Priority Burst Time

P1 0 3 5
P2 1 1 2
P3 1 2 8
P4 3 1 5
P5 7 4 7

Calculate Average Turnaround Time, Average waiting time for the following
algorithms. The time slice is 2 ms and the highest number has the highest priority
(ignore the arrival time in non-preemptive scheduling) for the following scheduling.
(a) FCFS (b) SJF (Preemptive and non-
preemptive)
(c) Priority (Preemptive and non-preemptive) (d) Round Robin

25. Consider the following set of processes

Process Arrival Time Priority Burst Time

P1 0 1 20
P2 0 5 12
P3 3 3 3
P4 4 4 18
P5 8 2 8

Calculate Average Turnaround Time, Average waiting time for the following
algorithms. The time slice is 5 ms and the lowest number has the highest priority (1
ms is context switching time and ignore the arrival time in non-preemptive
scheduling) for the following scheduling.
(a) FCFS (b) SJF (Preemptive and non-
preemptive)
(c) Priority (Preemptive and non-preemptive) (d) Round Robin

77
 M.Sc. (Computer Science)
SEMESTER-1
COURSE: DBMS
UNIT 1:

STRUCTURE
Objective
Introduction to Deadlock
System Model
Deadlock Characterization
Necessary Conditions For Deadlock
Deadlock Detection
Deadlocks Management
Deadlock Prevention
Deadlock Avoidance
Deadlock Dectection & Recovery
Deadlock Ignorance
Practice Excercises

78
OBJECTIVES

 To understand the Deadlock Problem

 System Model
 Deadlock Characterization
 Methods for Handling Deadlocks
 Deadlock Prevention
 Deadlock Avoidance
 Deadlock Detection
 Recovery from Deadlock
INTRODUCTION TO DEADLOCK
There is a situation in which two processes wait for each other. The resources of a system
may be limited or less as compared to the number of processes. In multiprogramming,
numerous processes may require a predetermined number of resources. When a process
requests for a resource, the Operating System checks whether the resource is available or not.
If the resource is available, then it is allocated to the process. On the other hand, if the
resource is not free (available), the process enters the waiting state. Sometimes it happens that
the process remains to wait, for a long time because the requested resource is holding by
other waiting processes. This situation leads to DEADLOCK. It is a condition, wherein a set
of processes are waiting forever for the resources, held by each other. None of them can
proceed with its execution.

In routine life, a traffic jam on the narrow bridge is an example of Deadlock, when cars came
from both side, but only a single car crossed from the bridge at a time and both car drivers
refuge to back the cars as shown in the figure below:

79
 Example: Consider a scenario where there are three processes P1, P2, P3, and three
resources R1, R2, R3. P1 requires two resources R1 and R2 for its execution. Also, P2
requires two resources R1 & R2 for its execution .P3 requires only R3. Initially, R1 is
allocated to P1, and R2 is allocated to P2.

Here, (i) R1 is allocated to P1 and P1 is requesting R2.

(ii) R2 is allocated to P2 and P2 is requesting for R1.

(iii) R3 is allocated to P3.

Allocated
P1 . R1

R2
P2
Allocated
.

P3 Allocated
R3

Allocated

Requested

R2 R1

P1 P2
. . P1

SYSTEM MODEL

In the system, there are n number of processes namely p1,p2,p3, …… pn-1, pn, and m number of
resource types namely R1, R2,…..Rm-1, Rm. In the System model, R resources are to be
circulated among some processes P. The resources are then divided into many types, each

80
consisting of some definite quantity of identical instances which is represented as w. The
main example of resource types is CPU cycles, memory space, Input-Output devices such as
keyboards, printers, and CD-DVD drives, directories, and files. Each resource type has
instances. Instances mean the number of resource types that represent as w. Each resource
type Ri has Wi instances When a system has 4 printers, then the resource type printer got four
instances. Each process uses a resource as the following sequence:

• Request: Process request for the resources. If resources are available then the
system allocates to process otherwise process waits for resources.

• Use: Then process used the resources when the system allocates them.

• Release: Process used the resources and then released them.

DEADLOCK CHARACTERIZATION

Necessary Conditions For Deadlock

The following four conditions must hold simultaneously, for a deadlock to occur:

1. Mutual exclusion

Mutual exclusion implies that a resource can be utilized exclusively by only one process
at a time in a non-shareable mode. If any other process wants to use that resource, then it
must have to wait, until it is released by the former process. For example, P1 holds R1 and P2
holds R2, both in mutually-exclusive mode.

2. Hold and wait

It implies that there exists a process that is holding at least one resource which is waiting
for another resource or resources, which is being held by other processes. It implies that there
exists a process that must be holding some resources (at least one) in a non-sharable mode
and at the same time must be waiting for the other resources, which are held by other
processes in a non-sharable mode. Like in the above example: P1 is holding R1 in a non-
shareable mode and at the same time trying to acquire R2, which is currently held by P2 in a
non-sharable mode. Similarly, P2 is holding R2 in a non-shareable mode and at the same time
trying to acquire R1, which is currently held by P1 in a non-sharable mode. This situation is
called hold and wait for conditions.

3. No pre-emption

81
 It implies that the resources cannot be pre-empted forcefully from a process. After the
completion of the process, a resource is released voluntarily by the process.

4. Circular wait

Circular wait implies that the processes are waiting for resources that are circularly held
by another process. In other words, Pi is waiting for a resource held by Pi+1 and so on Pn-1 is
waiting for a resource which is held by Pn, and at last, Pn is waiting for a resource held by Pi.
For example, Let a set of processes, say (P1, P2, P3, ---, Pn). They must wait in a circular
way for the resources held by each other i.e.

P1 is waiting for a resource which is held by P2

P2 is waiting for a resource which is held by P3

Pn-1 is waiting for a resource which is held by Pn

Pn is waiting for a resource which is held by P1

P4 Pn-1 Pn
P1 P2 P3

82
 Deadlock Detection

When all the four essential condition holds at the same time then deadlock occurs. Another
way to detect deadlock is Resource Allocation Graph (RAG)

Resource Allocation Graph

The Resource Allocation Graph is a directed graph which is a set of vertices which is denoted
by V and a set of edges which is denoted by E. Vertices (V) is divided into two types.
Processes (P) and Resources (R). P = {P1, P2, - - -, Pn-1, Pn} the set consisting of all the
processes in the system. A process is represented by a circle, with the process name indicated
with a label inside the circle. R = {R1, R2, - - - -, Rn-1, Rn} the set containing of all
resources types in the system. A resource is represented by a square, with dots inside the
square representing different instances of that resource.

There are two types of edges: Request Edge & Assignment edge.
An edge from process P1 to Resource Rj represents a request edge.

Rj

Pi .

Request Edge

An edge from resource Rj to process Pi represents an assignment edge.

Rj
P2

Assignment Edge

Note: The assignment edges originate from the instance within the resource symbol, which
indicates which instance is assigned to the process.
83
 If there is no cycle in RAG, it indicates that there is no Deadlock occurred.

If there is a cycle detected in the graph and the resources involved in the cycle have only
one instance per resource, then it indicates that a deadlock exists.

If there is a cycle detected in the graph and some of the resources involved in the cycle
have multiple instances per resource, there may be a deadlock exists.

For example

Let there are 3 processes P1, P2, P3, and 4 resources R1, R2, R3, R4. R1 has one
instance, R2 has 2 instances and R3 has one instance, R4 has one instance. R1, R2 is

allocated to P1 and requested for R3. R2 is allocated to P2 and requested for R3. R3, R4 are
allocated to P3 and requested for R2.

P1 P2 P3

R3
R1 R2 R4

Resource Allocation Graph

Example of RAG with a cycle Deadlock

84
 R2 R3

P1 P2 P3

R R4
1

Resource Allocation Graph with Deadlock

Example of RAG with a cycle but no deadlock

R1

P1 P2

R2 P3

Resource Allocation Graph with Cycle but no Deadlock

85
Process Wait for Graph (PWFG)
PWFG can be obtained by collapsing the resources symbol in RAG. An edge from
process P1 to P2 indicates that P1 waiting for a resource that is currently held by
P2.
R3
Consider the following RAG

P2 P3
PP11

R1 R2 R3

Example of Resource Allocation Graph

86
Equivalent PWFG for above RAG

P2

P3
P1

Process Wait for Graph (PWFG)

The PWFG contains a cycle, thus indicated that deadlock exists.

Deadlocks Management

Different OS adopts different strategies to handle deadlocks. The policies to handle deadlockscan
be classified into four major groups:

 Deadlock Prevention
 Deadlock Avoidance
 Deadlock Detection and Recovery
 Deadlock Ignorance

The deadlock prevention techniques prevent the occurrence of at least one of the four
conditions that cause deadlocks. By preventing are of the necessary four conditions deadlock can be
prevented. The deadlock avoidance techniques acquire information in advance about which
resource a process will claim at what stage of execution. The OS allocates resources in such a
manner that no deadlock occurs. Deadlocks can be avoided by maintaining the system always in a
safe state. The system is said to be in a safe state if all pending processes can be successfully
executed in some sequence. That sequence is called a safe sequence and processes are in a safe
state. The third category of deadlock management strategies is known as detection and recovery.
These types of techniques allow deadlocks to occur detect deadlock when it occurs and then apply

87
certain methods to recover from deadlock.

Finally, an OS can assume that deadlock will never happen or rarely occur and fully ignore it.
This approach might be referred to as the no-policy approach. The positive point of this approach is
it saves CPU time and memory space for deadlock management, unlike those required for detection,
prevention, or avoidance methods. This strategy has been adopted in UNIX.

Deadlock Prevention

This technique allocates resources in such a manner that at least one of the four necessary
conditions of deadlock cannot occur. It denies at least one of the four conditions required for the
occurrence of a deadlock. These techniques do not utilize a resource properly. For example, an OS
adopting a deadlock prevention technique allows recourses to be preempted from a blocked process.
No preemption condition is violated and deadlock does not occur.

We have already discussed that there are four conditions to be satisfied for the occurrence of a
deadlock. Below is a brief discussion about how the conditions for a deadlockcan be prevented.

1. Mutual exclusion condition

The mutual exclusion condition indicates the locking of resources in exclusive, also known as
the non-shareable mode that leads to blocking of resources. An OS can avoid blocking resources by
locking the resources in shareable mode, if feasible. For example, two processes need to read data
from a file.

The OS can allow locking of that file in shareable mode by two processes simultaneously.
Locking resources in shareable mode prevents waiting for a resource. When a process request to
look at a resource in shareable mode, the OS allows the locking instantly and no time is wasted
waiting to lock the resource exclusively.
Some resources cannot be shared by multiple processes. For example, two processes cannot
simultaneously share a scanner. Again when a process writes data to a file, it needs to lock the file
in exclusive mode.

2. Hold and Wait

An OS can avoid this by adopting a strategy that a process must request for all resources it
requires at the same time. This strategy can be implemented in two ways.
Firstly, a process p requires all the resources it needs when the execution begins. On the other
88
hand, a process can start execution with a minimum set of resources required and request the other
resources when required at the time of execution. In the second approach, a process must release all
the resources it presently holds before requesting any other required resources.
For example, a process p requires resources R1, R2, R3, and R4 in exclusive mode in the
sequence: R3 and R1 simultaneously at the beginning of execution, then R1, R2, and R3
simultaneously, and lastly R1 and R4 simultaneously. When applying the first strategy, process P
acquires all four simultaneously before it starts execution.
When applying the second strategy, the process p needs not to acquire all the four resources at
the beginning. The process p1 acquires the resources R1 and R3 at the time of starting execution.
Soon the process p requires R2. The process p releases all the acquired resources R1 and R3 and
sends a request to the OS for acquiring the resources R1, R2, and R3. Later, the resources R2 and
R3 are not required and R4 is needed. The process p releases all the acquired resources R1, R2, and
R3 and sends a request to acquire R1 and R4.

3. No Preemption

An OS can adopt a policy to avoid this condition. If a process requests an unavailable resource,
the process must release all the resources it has presently acquired and wait for the required
resources. In other words, when a process, P requests a busy resource this policy allows other
processes to preempt the resources currently held by P.
Certain resources are good candidates to be preempted. For example, CPU, and memory need a
mention. While revoking the control of the resources from a process the OS updates the process
control block of the process. Not all the resources can be preempted. For example, a partly updated
file cannot be preempted since data will be lost.

4. Circular Wait

An OS can impose an ordering of resources for avoiding the circular wait. Resources in a
system are grouped into certain categories. For example, all magnetic disks form a group and all
scanners from another group. Each category is assigned a number. When a process acquires a
resource belonging to a specific category, no other process belonging to a category having a lower
number can be claimed.
Formally, if there are n categories of resources in a system, ranging from 0 to n-1, and if a
process acquires a resource belonging to category c, then the process can only request a resource
belonging to category c+1.
89
 This helps to prevent a formation circle because a process holding the control of a resource of
the category n-1 cannot claim a resource of the type 0. The disadvantage of this technique is each
process needs to acquire all the required resources in a predetermined and specific order depending
upon the arrangement of numbers assigned to various categories of resources

Deadlock Avoidance

In deadlock avoidance, Operating System requires additional information about which process
requires which resource in advance to avoid a deadlock. The fundamental concept of deadlock
avoidance is that OS only entertains those requests for resources by processes that will not lead to a
deadlock. The deadlock-avoidance algorithm regularly inspects the resource-allocation state of the
processes to confirm that there can never occur a circular-wait condition. Resource-allocation state
is explained by the total number of available resources and number of resources allocated to the
processes, and the maximum demands of resources by the processes.

Safe State

When processes request resources and it is available in the system and allocated immediately
then the system in a safe state. The system is in a safe state if there exists an order of processes
<P1, P2, …, Pn> which satisfied their resources in that sequence. For each Pi, the resources that Pi
can still request can be satisfied by currently available resources + resources held by all the Pj, with
j < i. That is: If Pi resources wanted are not instantly available, then Pi can wait till all Pj have
finished. When Pj is finished, Pi can get needed resources, execute, return allocated resources, and
terminate. When Pi terminates, Pi +1 can obtain its needed resources, and so on.

If a system is in a safe state then there is no Deadlock. If a system is in an unsafe state then there is
the possibility of deadlock. To avoid deadlock, it is must confirm that a system never entered an
unsafe state and it should be in a safe state.

90
 Deadlock Unsafe

Safe

Relationship between safe, unsafe, and deadlock state

Avoidance algorithms

There are two algorithms to avoid deadlock

For the Single instance of a resource type, in this algorithm, we use a Resource- Allocation
Graph (RAG) and the second algorithm which is deals with multiple instances of a resource type,
then we use the Banker‟s algorithm.

Resource-Allocation Graph (RAG) Scheme

In Resource allocation Graph Scheme Claim edge Pi → Rj showed that process Pi may
demand resource Rj; it is denoted by a dashed line. Claim edge changes to request edge when a
process needs a resource. Request edge changed to an assignment edge when the resource is
assigned to the process. When a resource is free by a process after execution, the assignment edge
reconverts into a claim edge. Resources must be demanded in advance in the system.

Suppose that process Pi requests a resource Rj. The request can be approved if and only if
changing the request edge to an assignment edge does not result in the creation of a cycle in the
Resource Allocation Graph(RAG).

91
 R2

P3
P1

R1

(i) Unsafe state in RAG

92
 R2

P3
P1

R1

Figure showing (i) safe and (ii) unsafe state in RAG

BANKER’S ALGORITHM

The banker’s algorithm adopts the deadlock avoidance strategy. The name, banker‟s algorithm,
depicts the similarity of the concept in the field of banking. A banker never allots the cash available
in such a manner that the banker cannot fulfill the needs of its clients.

A scheduling algorithm that can avoid deadlocks is BANKER‟S ALGORITHM. It is modeled

on the way that a single banker might deal with a group of customers to whom he has granted
lines of credit. The banker‟s algorithm can be implemented for a single resource or multiple
resources. The banker algorithm keeps track of all the resources currently available to the operating
system and all the resources which are allocated to the processes. It maintains the various data
structures to store the information about the number (instances) of resources. Whenever a process
requests a resource, the banker‟s algorithm checks the availability of the resources and acts
accordingly. Finally, it helps to check that the system is in a SAFE STATE or not.

93
Data Structures for the Banker’s Algorithm

Suppose n be the number of processes, and m be the number of resource types.

Available: Array of length m. If available [j] = k, it means that k instances are available of
resource type Rj.

Max: n x m matrix. where n is the number of rows and m is the number of columns. Max
[i,j] = k, it means that k instances maximum needed by process Pi of the resource typeRj.

Allocation: n x m matrix. where n is the number of rows and m is the number of columns
Allocation[i,j] = k, it means that k instances are currently allocated to process Pi of the resources Rj.

Need: n x m matrix. where n is the number of rows and m is the number of columns Need[i,j] =
k, it means that Pi may need k more instances of Rj to complete its task. It can be calculated by
matrix subtraction:

Need [i,j] = Max[i,j] – Allocation [i,j].

Safe State Algorithm

This algorithm is applied by an Operating System to determine whether a system is in a safe state or
not. A system is safe if the system can allocate resources to every process in some order and still is
in the safe state which means that its avoiding deadlock. In other words, a system is in a safe state
only if there occurs, a safe sequence.

Let „m‟ be the total number of resources and „n‟ be the total number of processes in thesystem.

Let „Work‟ and „Finish‟ be the two arrays of length „m‟ and „n‟.

(1) Set Work = Available and also Set value FALSE for all the members of array Finish

FINISH[i] = FALSE, for i = 0 to n - 1

In other words, it is assumed that all the processes are to be executed and neither of the
processes is completed yet.

(2) Find a process ―i‖ such that following both conditions should be satisfied

Finish[i] = FALSE and Need[i] < = Work If no

94
 such ―i‖ exists ,then go to step (4).

In other words, find a process that is not completed yet and whose need is less than
available.

(3) Set Work = Work + Allocationi (Allocation of process i)Set

Finish[i] = TRUE.
Go to step (2).

In other words, after the completion of the execution of a process, the process will returnall the
resources to the system. So, the available resources of the system will increase.

(4) Check if Finish[i] = TRUE for all ―i‖ at that time

System is in safe state
else

System is not in safe state

Consider an example

Let there are 3 resources and 5 processes such that

Processes Allocation Max Available

R1 R2 R3 R1 R2 R3 R1 R2 R3
P0 0 1 0 6 5 3 2 3 2
P1 2 0 0 3 3 2
P2 3 0 2 9 0 2
P3 2 1 1 2 2 2
P4 0 0 2 5 3 2

Find the Need matrix by using the formula:Need

= Max – Allocation
So, the need matrix is:

Processes Need
R1 R2 R3

95
 P0 6 4 3
P1 1 3 2
P2 6 0 0
P3 0 1 1
P4 5 3 0

Hereby considering the need of all the processes, we have to find such a process whose need is
less than available. It is visibly that the need of the process P1 is less than Available, so the process
P1 will be executed first. (As 1 3 2 < 2 3 2)

After the execution of process P1, the P1will return all the resources to the system. So, the
available resources will increase as the following :

Available = Available + Allocationi

In the same way, after the completion of a process, the Available will be updated, as the
process returns the resources to the system after being executed.

So, the processes will be executed in the following sequence

< P1, P3, P4, P0, P2>

Explanation of Safety Algorithm

First, we set all processes in the false state which means that no processes are executed or
completed yet, and set work is equal to available. Then we find repeatedly a process that is false
state and whose need is less than equal to work. If we find such a process then we change its state
to true and its resources are deallocated which is added in work. In this way, the number of
available resources is increased. Then we check all the processes are true or not. If all the processes
are true then the system is in a safe state. If even one process is in a false state then the system is in
an unsafe state.

Resource Request Algorithm

The second part of the banker‟s algorithm, known as the resource-request algorithm determines
whether granting a resource according to any claim by a process leads to an unsafe state. This
96
algorithm is applied when a process sends a request to grant one or a set of resources.

Requesti = request vector for processPi =

If requesti [j] = K,

it means that Pi requested the K instances of resource type Rj.

(1) If Request[i] <= Need [i] (Check request is genuine or not)then

Goto step (2)

Else

―Raise an error condition‖

(Because the process has exceeded its maximum claim)

(2) If Request[i] <= Available (Check weather system has available resources or not)

Then

Goto step (3)Else

The process Pi has to wait because the resource is not available yet.

(3) Let the system pretend that the resources have been allocated by changing the following:
Available = Availablei – Requesti ;
Allocationi = Allocationi + Requesti ;Needi =
Needi – Requesti;
Then check with the safety algorithm, if the result of the safety algorithm is safe, then thetransaction
is completed and the process is allocated the requested resources.

However, if the system is in an unsafe state, then the process must wait and the old allocation
state is restored.

Explanation of Resource Request Algorithm

In this algorithm, first, we check whether the request of the process is less than its need or
not. If the request is more than need then it is an error because the process claims more resources
from its maximum claim. Else we check current request is less than available or not. If it is less
than or equal to available then we allocate the request and update data structure such as allocation
of the process by adding a request in allocation, the request is subtracted from need as well as
97
available. Then we check the system is in a safe state or not by applying a safe state algorithm. If
available is less than need then the request can not be granted and the process will wait because the
resource is not available yet.

Advantages

Advantages of banker‟s algorithm, it allows mutual exclusion, hold and wait, no preemption
conditions. The system assures that a process will be allocated resources without deadlock.

Disadvantages
The disadvantage of a banker‟s algorithm is the overhead of simulation and calculation before
each time a process requests for a resource. Deadlock avoidance techniques cannot be applied until
an OS knows the resource requirement of a process in advance. This approach may lead a process
to starvation while for a long time to get the requested resources allocated by the OS.

Illustration

Consider the following table with 5 processes P0 to P4 and 3 resource types namely X Yand
Z; X has 9 instances, Y has 5 instances, Z has 8 instances.

ALLOCATION MAX
XY Z XYZ

P0 2, 2, 3 6, 2, 3

P1 2, 0, 1 6, 1, 2

P2 0, 1, 1 3, 1, 2

P3 1, 0, 0 1, 0, 1

P4 1, 1, 2 6, 2, 3

Using Banker‟s algorithm check whether the system is in a safe state or not. If it is a safe state
then if P1 requests (1, 0, 1) then it is a safe state or not. If it is not in a safe state then, what changes
are required in available resources to make the system safe?

Total instance of resources is

98
 X = 9, Y = 5, Z = 8

Total allocated resources to P0 to P4 areX = 6,

Y = 4, Z = 7

Now available is
X = 3, Y = 1, Z = 1

Now calculate Need: i.e. Needi = MAXi – Allocationi

Process Allocation MAX NEED Available

XYZ XYZ XYZ XYZ

P0 2, 2, 3 6, 2, 3 4,0,0 3,1,1

P1 2, 0, 1 6, 1, 2 4,1,1

P2 0, 1, 1 3, 1, 2 3,0,1

P3 1, 0, 0 1, 0, 1 0,0,1

P4 1, 1, 2 6, 2, 3 5,1,1

Now apply safe state algorithm to check whether the system is safe state or not.First
step is set all process equal to false i.e. P0, P1, P2, P3, P4 are FALSE. Now find such
process which false and whose needi < Available i.e. P2.

Now update data structure and status of processes i.e. finishi

Available = Available + Allocation2

Finish2 = True

Now new Available = 3, 1, 1 + 0,1, 1

= 3, 2, 2

All processes are not true so again go to step - 2 to find a process that is yet false andneeds
is less than equal to available.

99
 Need3 < = Available i.e. P3

Now update data structure and status of processes i.e. finishiNow

Available = Available + Allocation3
New Available = 3, 2, 2 + 1, 0, 0

=> 4, 2, 2

Finish3 = True

Now 3 processes are false i.e. P0, P1 & P4 & Available = 4, 2, 2

All processes are not true so again go to step - 2 to find a process that is yet false andneeds
is less than equal to available.
Need0 < = Available i.e. P0

Now update data structure and status of processes i.e. finishiNow

Available = Available + Allocation0
=> 4, 2, 2 + 2,2,3

= 6, 4, 5

Finish0 = True

Now 2 process are false i.e. P1 & P4 and Available = 6, 4, 5

All processes are not true so again go to step - 2 to find a process that is yet false andneeds
is less than equal to available.
Need1 < = Available i.e. P1

Now update data structure and status of processes i.e. finishiNew

Available = Available + Allocation1
= 6,4,5 + 2, 0, 1

=> 8, 4, 6

Finish1 = True

100
 Now only one process i.e. P4 is false if its need is greater than Available then system isnot safe
state but need4 < Available

Update Available & Finish

New Available = Available + Allocation4

= 8, 4, 6 + 1, 1, 2

=> 9, 5, 8

Now finish is true so there is no false in finish and all resources are available now i.e. 7,
3, 6.

< P2, P3, P0, P1, P4>

The system is in a safe state if the process runs in the above sequence.

Now check if P1 requests (1, 0, 1) then if it is allocated to the process then whether it is in
safe-state or not? To check that the resource request algorithm follows.

In this algorithm first request is to check whether it is less than Needi or not. If it is greater
than Needi its mean Pi request more resources as it claims earlier so it can‟t be given & request
directly terminated. If Requesti is less than Needi then we check Requesti < = Available if it is true
then the request is granted & we check that now system is in a safe state or not by using the safe
state algorithm used earlier. If it is false then at that time Requesti can not be fulfilled and the
process Pi made to wait.

In this case Request P1 is (1, 0, 1) which less than Need1 and Available so we run resource
request algorithm.

Set Available = Available - Requesti

= 3, 3, 1 - 1, 0, 1

101
 Now Available = 2, 3, 0

Set Allocation1 = Allocation1 + Request1

= 2, 0, 1 + 1, 0, 1

Now Allocation1 = 3, 0, 2

Set Need1 = Need1 - Request1

= 4, 1, 1 - 1, 0, 1

Now Need1 = 3, 1, 0

Now Request is granted now we check after allocating the request whether the system is in
a safe state or not. To check that we used a safe state algorithm. Now the situation of processes is
given below:

Process Allocation MAX NEED Available

XYZ XYZ XYZ XYZ

P0 2, 2, 3 6, 2, 3 4,0,0 2,3,0

P1 3, 0, 2 6, 1, 2 3,1,0

P2 0, 1, 1 3, 1, 2 3,0,1

P3 1, 0, 0 1, 0, 1 0,0,1

P4 1, 1, 2 6, 2, 3 5,1,1

Now by applying a safety algorithm there is no such process whose needi < = available. So if
the request of process1 is granted then the system comes in an unsafe state. So, we revoke the
request of Process1 so that system may run safely. To overcome this problem we must need at least
one A-type and at least two C-type Resource. If it will available in the available list then the system
may come in a safe state.

Deadlock Detection and Recovery

Another approach to deadlock management is detection and recovery. In this approach, the
system is not prevented from occurring a deadlock. The operating system periodically searches or
102
analyzes that whether a deadlock has occurred. If the system is in deadlock, OS recovers from
deadlock.

This approach of deadlock management has two phases, detection, and recovery of deadlocks.
The algorithm to detect deadlocks in an OS can use a data structure similar to that of the data
structures used in the context of deadlock avoidance.
A variation of the safety algorithm discussed in deadlock avoidance has been discussed below.
In this algorithm, the data structures ALLOCATION, CLAIMS (Request), and AVAILABLE are
the same as the data structure used in the context of deadlock avoidance. Another data structure
used in this algorithm is FLAG.
The FLAG is a vector of the length equal to the number of processes in a system. FLAG is
used to mark and unmark a process. For example, FLAG[X]=TRUE means the process Px is
marked, and FLAG[Y]=FALSE means that the process Py is unmarked.
An OS verifies after a certain interval whether a deadlock has occurred. This algorithm finds a
sequence of execution of processes that do not lead to an unsafe state. For example, there are 3
processes P1, P2, not lead the
system to deadlock. The steps in this algorithm are:
1. Initialize FLAG[i] to FALSE, where 0 <= i<=s-1, and s denotes the number of
processes.

2. Find a process Px such that FLAG[X]=FALSE and CLAIMSX<=AVAILABLE. Goto

step 6, if no such process is found.

3. Update the AVAILABLE vector because, after the termination of the process, Px, OS
revokes the control of all the resources held by Px. Perform AVAILABLE =
AVAILABLE + ALLOCATIONX.

4. Mark the process Px as FLAG[x]=TRUE.

5. Go to step 2.

6. If FLAG[i]=TRUE for all i, where 0<=i<=s-1, then all processes can be completed
properly and there exists no deadlock, else the system is deadlocked with the processes
which are unmarked that is FLAG[i] is set to FALSE.

103
 The algorithm needs an order of O(m × n2) processes to detect that the system is in a deadlock
state or not.

The frequency of verifyin5r4ag the system whether a deadlock has occurred is an important
issue. An OS can adopt any of the two policies. First, an OS can verify the system state whenever a
process sends a request for resources. Alternatively, an OS can verify the system state after a certain
interval of time.

An OS can adopt two strategies to recover from deadlock. Firstly, all the processes involved in
the deadlock can be terminated forcefully. Alternatively, the OS can apply a preemption
technique from a deadlock period.

Aborting deadlocked processes can be done in two ways.

An OS can abort all the deadlocked processes when a deadlock is detected.

On the other hand, an OS can apply the trial and error method. This means, one of the
deadlocked processes is aborted and it is verified that whether the deadlock is over. This procedure
is continued until the system reaches a safe state.

Which process is to be aborted first depends on the cost of terminating the process. If a process
is 95% done, OS tries to avoid aborting that process. The number of processes to be aborted is also
important in this context. For example, four processes are deadlocked. The OS can recover itself
from deadlock by aborting either all of the processes A, B, and C or a single process D. In this case,
an OS will sacrifice a single process to recover from deadlock.

The second approach of recovery from deadlock is preempting a resource from a process. In
this case, a victim is select, the process from which a resource is revoked by an OS is rolled back
to a safe previous state and continued later on. In this approach, the starvation problem may occur
by selecting the same process as a victim again and again.

Deadlock Ignorance

In deadlock ignorance, we assume that our system never enters in the deadlock condition. It is
only a theoretical concept. Practically it is not possible because if the system is not in a safe state
then the system may be in a deadlock state or we can say that if the system satisfied all the
necessary four conditions simultaneously then the system must be deadlock state which can be
104
ignored.
Comparison Between Various Deadlock Management

Deadlock Prevention Deadlock Avoidance Deadlock Detection and

Recovery

It is easy method It is complex method It is also a complex method It

adversely affects Utilization Utilization improved Depends on the frequency of
at a cost detection

Involves violating at least one RAG Algorithm Maintain a wait-for graph of the
necessary conditions for Bankers Algorithm and periodically invoke
deadlock (Multiple instances) a ―cycle-search‖ algorithm
Used in critical systems Used in applications that Used in Database

have deterministic needs Applications

Points to Remember

 Deadlock: Process is said to be in a deadlock state if the process is waiting for a

specific event that will not occur.
 The reusable resource can be safely used by only one process at a time.
 Four necessary conditions for deadlock are:
 (i) Mutual exclusion (ii) Hold and wait
 (iii) Circular waiting (iv) No preemption
 Resource Allocation Graph is used to detect the deadlock.
 Resource Allocation Graph is also known as RAG.
 If RAG contains a cycle then it may or may not deadlock the situation.
 In RAG, the process is represented by a circle and the resource is represented by a
square.

105
  Process Wait For Graph is obtained from Resource Allocation Graph.
 Process Wait For Graph is known as PWFG.
 The policies to handle deadlock can be classified into four major groups
 Deadlock Prevention
o Deadlock Avoidance
o Deadlock Detection and Recovery
o Deadlock Ignorance
 Deadlock prevention is a set of methods for ensuring that at least one of the necessary
conditions cannot hold.
 Deadlock avoidance needs extra information about how resources are to be
demanded.
 Safe order is an order of the process in which there exists one order in which all the
processes can be executed without resulting in a deadlock.
 The deadlock avoidance algorithm enthusiastically examines the resource-allocation
state to confirm that a circular wait condition can never exist.
 A safe state is not a deadlock state.
 Not all unsafe states are deadlocks.

 Deadlocks occur only when some process makes a request that cannot be granted
immediately.
 Banker‟s algorithm is a deadlock avoidance algorithm.
 A safety algorithm is used to find the safe state of the system.

PRACTICCE EXERCISES

1. What is deadlock?

2. What are the necessary conditions for the occurrence of a deadlock?

3. How deadlock can be avoided.

4. What do you mean by RAG?

5. Explain methods for deadlock detection.

106
6. Explain safe state.

7. What are the different methods of handling deadlock.

8. Explain the Banker‟s algorithm with an example.

9. Explain the safe state algorithm with an example.

10. Explain the resource algorithm with an example.

11. Consider the following snapshot of a system.

Process Allocation MAX Available

MNOP MNOP MNOP

P0 2, 2, 1, 3 7, 5, 3, 3 1, 1, 2, 1

P1 2, 2, 1, 1 6, 2, 2, 4

P2 1, 2, 1, 0 4, 2, 4, 0

P3 1, 1, 0, 1 2, 2, 2, 3

P4 2, 1, 2, 0 2, 4, 3, 3

Whether the system is a safe state or not. If process P3 request 0, 0, 1, 2 can this request
granted immediately or not and also check after granting above request system is insate state or
not. If it is not in a safe state then what changes are required in available resources to make the
system safe.

107
 B.Sc.(DATA SCIENCE)
SEMESTER-II
OPERATING SYSTEM

UNIT IV: MEMORY MANAGEMENT

STRUCTURE

Objectives

Introduction

Address Binding

Dynamic Loading And Linking

Logical vs. Physical Address Space

Swapping

Memory Allocation

Practice Exercises

108
OBJECTIVES
 Understanding Memory, Address Binding, Logical Vs Physical Memory
 Understanding Concept of Dynamic Loading and Linking, Swapping, Contiguous
Memory Allocation, Segmentation, Paging, Demand Paging, Page Replacement
algorithms

INTRODUCTION

The important task of allocating memory to processes, and efficiently, ensuring that
processes have their instructions and data in main memory when needed, is termed as
Memory Management.

Memory is one of the important resources of the computer system. The main
memory is usually divided into two partitions, one for the resident operating system and
the other for the user processes.

The operating system is responsible for memory management. It keeps track of the
status of the memory, makes policy, allocates the memory, and then deallocates the
memory from the process.

Figure: Main Memory

The functions of Memory Management

An operating system performs various activities for memory management:

1. It keeps track of the status of the memory, that whether it is free, allocated,
available, or not available (full).
2. If the memory is free for the execution of the process, the Operating System
109
 chooses a policy to allocate the memory to the process.
3. According to the chosen policy (algorithm), it allocates the memory to the
process.

4. After the execution of the process, it deallocates the memory.

ADDRESS BINDING
Address binding means map from one address space to another i.e. the logical addresses
to real physical addresses in memory. To execute a program, it must be loaded into the
main memory at a particular location. The instructions that use addresses in a program
must be bound to proper address space in the main memory. Many instructions use
―fixed‖ addresses these must be bound to ―fixed‖ locations in the memory.

This binding of addresses of instruction and data to actual physical addresses can
take place at compile-time, load time, and run time during the execution of a process.
To run a program there is a sequence of steps to execute a particular program which
is as follow:

 A program is known as a source program and it is loaded into the main memory.

 A compiler or assembler or interpreter is required for compilation or assembly.

 After compilation or assembly, a code is generated by them which, is known

as Object code.
 When a program is executed then, the program links with other object
modules or system library which are loaded into linkage editors with the help of
linker and loader.

At last, the process is stored in memory as a binary image, and system library files
are dynamically loaded and dynamically linked.

DYNAMIC LOADING AND LINKING

In any program execution linking and loading plays a key role. Linking means creating
an executable module of a program by linking the object codes which is created by the
assembler or compiler. Other hand loaders, load these runnable modules to the main
memory for execution.
110
Linking: It is the procedure of linking all the modules or the sub-program or all the
functions in a program for whole program execution. It takes more than one object
module and combines it into a single object file. The linker is also called a link editor. It
takes object modules from the compiler or assembler and forms a runnable file for the
loader.

Based on time Linking is classified into two categories – static linking and dynamic
linking:

Linking of object modules is done at compile time and at load time. Compile-time linking
is completed when the source code is converted into machine code is called Static
linking. The load-time linking is done while the program is loaded into memory by the
loader.

Loading

Loading is the procedure of loading the program from auxiliary memory to the main
memory for execution.

Dynamic loading

It involves loading routines into memory only when required. This is done during
execution. Dynamic loading reduces the memory requirements of large programs. This is
especially the case if there is a large set of infrequently used routines.

Dynamic linking

It is often used for libraries. Only a ―stub‖ of the library is kept in the image of the
program. When a program calls one of these routines, the routine is loaded and linked
into memory. All programs share one copy of the same library routine.

111
 Figure Processing of a User Program

Address binding of instructions and data to memory addresses can happen at three
different stages i.e. compile time, load time, execution time (as shown in figure
Processing of User Program)
Compile-time: If the memory location is known a priori, absolute code can be
generated. It must recompile code if starting location changes, after compiling the
process.
Load time: It must generate relocatable code if memory location is not known at
compile time. Final binding is delayed until load time.

Execution time: If the process can be moved during its execution from one memory
segment to another, binding may be delayed. Then hardware support is needed for
address maps. For example base and limit registers. Most general-purpose operating
systems use the execution time-binding method.

LOGICAL VS. PHYSICAL ADDRESS SPACE

When a process is executing, the CPU would generate addresses, called Logical Ad
112
dresses. There are two registers used to define logical address space. These are the Base
register and Limit register. Base register stores the starting memory address of the
process and the Limit register stores the size of a process. Let a process P whose size is
5000 i.e. value of the Limit register and it has the start address is 1000. i.e. value of Base
Register. A pair of base and limit registers define the logical address space as the shown
figure below

0 Operating System

500

1000
1000

Base

6000
5000

Limit

Figure Base and Limit Register

113
 Executing processes that occupy corresponding physical addresses in the physical
memory is known as Physical Addresses. Logical and physical addresses are the same
in compile-time and load-time address-binding schemes but logical (virtual) and physical
addresses differ in the execution-time address-binding scheme. A physical address is the
effective memory address of instruction or data.
SWAPPING
Swapping is a mechanism in which a process can be swapped temporarily out of memory
to a storage device (backing store) and another process brings in memory for execution.
If a swapped-out process requires again then it is brought back into memory for
continued execution.

Figure Logical and Physical Address Space

The backing store is a fast disk large enough to accommodate copies of all memory
images for all users. It must provide direct access to these memory images. When the
process swapped out then it is called Roll out and when it brings in then it is called Roll
in. The swapping variant is used for priority-based scheduling algorithms where the
lower-priority process is swapped out due to the arrival of a higher-priority process. It
can be loaded and executed. A major part of swap time is transfer time. The system
maintains a ready queue of ready-to-run processes which have memory images on disk.

Example: Let there are n processes in the storage device and P4 in main memory which
is not required. Now P2 is required so P4 is roll out and P2 is rolled in which is known as
swapping as shown in the figure.
114
 Figure: Swapping

MEMORY ALLOCATION

Memory management includes the various methods of allocating the memory to the
different processes. The memory management techniques are basically of two types:

(i) Contiguous memory allocation.

(ii) Non-contiguous memory allocation.

Contiguous Memory Allocation

In contiguous memory allocation, the data and instructions of a program are sure to reside
in a single contiguous memory area. It is a simple technique of allocating memory to the
processes. There is no possibility of sharing data and instructions among the processes,
which are loaded into entirely different portions of memory. The entire process is
allocated a single contiguous memory area for its execution. Addressing is very simple as
compare to the non-contiguous allocation methods.

Fragmentation

In contiguous memory allocation, one problem occurs i.e. Fragmentation. It can be

classified into two categories:
1. Internal Fragmentation
2. External Fragmentation

1. Internal Fragmentation: In fixed-size memory management, allocated memory

115
to a process may be slightly larger than the requested memory by the process. The size
difference between occupied memory and partition size, which is not used by another
program is known as Internal Fragmentation.

Example of Internal Fragmentation: Let, there are three processes P1, P2, P3
whose size is 48, 45, 50 respectively. The total memory size is 256 and OS occupies 56K
and the rest of the memory is divided into fix partition of 50k.

56

Process 1
48
106 2 Free
Process 3 Internal
Fragmentation
251 156
50

Process 2
206
50

256 45
5

Example of Internal Fragmentation

2. External Fragmentation: In variable size memory management, some small

chunks are available in memory that cannot be effectively used. Total free memory is
larger than the required memory by the requested process but it is not contiguous. It is
known as External Fragmentation.

116
 Example of External Fragmentation: Let, there are three processes P1, P2, P3
whose size is 40,25,50 respectively. The total memory size is 256 and OS occupies 56K
and the rest of the memory is divided into fix partition of 50k.

56

Process 1
40
106 10
Process 3

231 156
50

50
Process 2
206

25
256
25

Free External
Fragmentation

117
 Example of External Fragmentation

Non-contiguous Memory Allocation

In non-contiguous memory allocation, the data and instructions of a program may reside
in non-contiguous memory locations. It is possible to allocate different memory locations
to a process for its execution. In other words, the process can be divided into parts and
can be allocated to different memory areas. It offers various advantages over, contiguous
memory allocation like it permits sharing of code and data amongst processes. There is
no external fragmentation of physical memory. It supports the vertical memory concept.

Contiguous Memory Allocation Methods

The memory is partitioned into different blocks of different sizes for accommodating the
programs. The partitioning is of three types:

1. Single partitioning memory management

2. Multiple Fixed Partitioning

3. Multiple Variable Partitioning

Single Partition Memory Management

The main memory is divided into two partitions. One partition is for the resident
operating system and, the other for the user processes. In the single partition memory

management technique, the user area has a single partition. It is the simplest memory
management technique.

Consider an example, where the total main memory is 256 MB. 56 MB is reserved
for the operating system and 200 MB is for the user processes.

118
 Figure: Single Partitioned Main Memory

Here, the user area has a single partition. At a time, only one program can be
executed. The maximum length of the program, which the user can execute, is equal to
the size of the partition of the user area. Here, the maximum size of a program can be 200
MB.

Internal fragmentation is much more. Since there is a single partition of memory, we

can execute only one program at a time. Consider an example, where the size of a
program is 50 MB. In this case, there will be an internal fragmentation of about 150.

Here, external fragmentation is also more. However, if the size of the program
increases from the size of the partition, then it is not possible to execute the program.

Advantages

 It is a very simple memory management technique.

 Any program of size less than the size of the user area can be executed easily.

Disadvantages

 Multiprogramming is not possible.

119
  Internal fragmentation is more.

 It is not possible to execute a program of size greater than the size of the user area.

 The maximum size of a program to be executed is fixed.

Multiple Partitions with Equal Size

In this, whole memory is divided into a fixed number of partitions of different sizes,
which may suit the program sizes. Each partition accommodates exactly one process.
When a program needs, it is loaded into a partition that one big enough to accommodate
the program. Sometimes, some space may be left unoccupied in a partition after loading a
program, which is not used by another program that space is wasted and it is known as
Internal Fragmentation. The degree of multiprogramming is fixed since the number of
partitions is fixed is the drawback of this partitioning.

Figure: Multiple Fixed Partitioned Main Memory

In this technique, internal fragmentation is less as compared to the single partition

technique. Consider an example, where a program of size 50 MB arrives. In this case,
there will be no internal fragmentation. However, if a program of size 40 MB or 30MB
arrives, then there will be an internal fragmentation of 10 MB or 20 MB respectively.

In this technique, external fragmentation may be more. Consider an example, where

120
a process of size 55 MB arrives. Even if four of the partitions(200 MB) are free, but it is
not possible to execute a program of 55 MB.

Advantages

 It is possible to execute more than one program at a time. In other words,

multiprogramming is possible.
 Internal fragmentation is less as compared to a single partition.

 It is simple to maintain.

 The degree of multiprogramming depends upon the number of partitions.

Disadvantages

oIt limits the size of the program, the user can execute.
oExternal fragmentation is high.
oIt is not possible to execute a program having its size greater than the size of the
partition.

Multiple Partitions with Variable Size

This scheme is free from the limitation encountered in the case of fixed partitioning. In
this, the entire available memory is treated as a single partition. All programs, requesting
memory are store in a waiting queue and loaded only when a free partition available
which is big enough to occupy that program. When a program is allocated a space exactly
equal to its size, the balance unoccupied space is treated as another free partition. When a
free partition is too small to accommodate any program, it is called External
Fragmentation.

121
 Figure: Multiple Variable Size Partitioned Main Memory

Three strategies can be used to allocate the memory to a process.

1. First Fit: According to this strategy, the O/S chooses the very first partition
available, whose size is equal to or greater than the size of the process.
2. Best Fit: According to this strategy, the O/S chooses a partition that is smallest
and whose size is greater than or equal to the size of the process. It may lead to the
formation of small-sized unusable holes of memory.
3. Worst Fit: According to this strategy, the O/S chooses the largest partition and
allocates it to the process. It may result in bigger-sized unusable holes of memory.
From the point of view of the size of the unusable holes, the first fit and best fit are
better than the worst fit. From the point of view of time taken for searching the partition,
the first fit is better than the other two i.e. Best fit and Worst fit

Example of Best Fit

Consider the requests from processes in given order 180K, 40K, 100K,400k, and
190K. Let there be four blocks of memory available of size 100K, 200k, 180K, 400k
followed by a block size 100K.

122
 100 Memory used by

200 Free Memory

180

400

100

Let there are three process P1, P2, P3, P4, and P5 of sizes 180, 40, 100, 400, and 190
respectively
By using an algorithm of First Fit, Best Fit & Worst Fit is shown by the figure
below:

123
Total free 260 k Total free 660 k but
Total free 70 k all process P4- 400 k and
but process 190k
process allocated P5- 190k not loaded as
not loaded as
in memory memory is not in
memory is not in
contiguous contiguous

Compaction

Compaction is a technique to deal with the problem of external fragmentation.

Compaction means to move the processes in such a way that scattered pieces of unused
(free) memory blocks can be placed together so that they can be used by any other
process.

124
 Figure: Before Compaction and After Compaction

Compaction involves the dynamic relocation of a program. This is done by using the
relocation register.
However, the compaction algorithm is costly. The problem with the compaction
algorithm is that it needs the dynamic relocation of addresses at the execution time of the
process. So, the compaction is not possible if the relocation is done statically.
The cost of the compaction increases with the increase in the number of processes to
be moved. Moreover, there is a need for a strategy to decide the direction (whether to
move upwards or downwards) to move the processes.

Advantages

 Internal fragmentation is very less.

 The degree of multiprogramming can vary.

Disadvantages

 The O/S has to decide about the partition size, almost whenever a program arrives
for execution.
 External fragmentation exists.

Memory Protection Contiguous Memory Allocation

125
When a logical address is generated by the CPU, then it is compared with the content of
Limit Register (M). If the logical address is less than equal to M, then it is a valid address
else it is an invalid address error, and the process is terminated. If the logical address is
valid then the corresponding physical address is computed by adding the content of
Relocation Register (B) i.e. base address to the logical address.
Physical Address = Logical Address L + Relocation value R

Figure: Memory Protection in Contiguous Memory Allocation

Non-Contiguous Memory Allocation Methods

Non-contiguous memory allocation involves a complex implementation and involves

additional costs in terms of memory and processing.
The non-contiguous memory allocation method includes the various strategies like:

1. Paging
2. Segmentation
3. Paging with Segmentation

Paging

Many solutions have been suggested to handle the external fragmentation problem of the
memory during the process execution. One scheme permits the logical address space of a
process to be noncontiguous, thus allowing a process to be allocating physical memory

126
wherever the latter is available. Memory paging implements this scheme. In the memory
paging technique, the logical address space is divided into fixed-sized blocks known as
pages. The physical address space is also divided into fixed-size blocks known as
frames. A page is mapped into a frame. Individual pages and frames are recognized by a
unique number known as page number and frame number respectively. The size of a page
is equal to the size of a frame. A page number forms a part of the logical address and a
frame number forms a part of a physical address.
Both page and frame size is usually a power of 2 and depends on the hardware. If a

logical address consists of P number of digits, this means there are 2p addresses in the
logical address space. Similarly, if physical addresses consist of q number of digits, this

means there are 2q addresses in the physical address space. It is always power by 2
because this is to divide logical address into a page number and page offset, easily.

Figure: Paging Hardware

Mapping of Pages to Frame

A page table is maintained for mapping. It divides into two parts: Page Number & Page
offset.

127
 p d

Page No Offset

A logical address consists of two parts: Page number and Offset. If a logical address is P

digits long and page size is 2n digits long, then the higher or leftmost p-q digits in a
logical address denote the page number, and the rest of the logical address denotes offset.

Memory Table for Mapping a Logical Address

For example, the logical address space can hold 213 addresses. This means a
logical address is 13 bits long. In other words, the logical address space is 8 KB. There

are 8 pages in the logical address space. This means the size of each page is 2 10 bits=1

KB. The size of the physical address space is 212 bits= 4 KB. This means a physical
address is 12 bit long. In the logical address of 13 bits the higher 13 bits (13-10) =3 bits
denote the page number and the rest 10 bits denote the offsets. The logical address 011
1010101010 denotes page number 011 and the offset 1010101010.
A physical address consists of two parts: frame number and offset.

128
 f d

Frame No Offset

The offset is the same as it is in the corresponding logical address. For example, in
the previous example, the offset 1010101010 of the logical address 011 1010101010 is 10
bits long. The physical address corresponding to this logical address can be 01
1010101010, where 01 is the frame number and 1010101010 is the offset.

The operating system maintains a table to convert a logical address into its
corresponding physical address. This table is called a memory page table. The method
of converting a logical address into its corresponding physical address is known as
address translation.

There are two fields in the page table. The first field stores the corresponding frame

number of a page. For example, page 3 is stored in the 13th frame. In the first field of the

page table value of the 4th cell will be 13. The second field stores a binary value either 0
or 1 and is known as the presence bit. If there exists a frame corresponding to a page, the
value of the cell is marked as 1 otherwise the value is set to 0.

129
 Figure: Mapping

Address Translation in a paging Scheme

The logical address 011 1010101010 is mapped into memory. From the page table,
the operating system verifies the presence bit of cell 011, which is 1. This indicates that
page 011 has been stored in a frame in the main memory.

Another example Using a page size of 2 bytes and a physical memory of 16 bytes (8
pages), we show how the user‗s view of memory can be mapped into physical memory.

130
so physical address calculated (frame no x size)+ offset

The physical address of ‗f‗ calculated: ‗f‗ stores on page no 2 and offset is 1. page no 2 is
loaded in frame 6.
So address is (6x2)+1 i.e. 13.

In a paging scheme, an operating system maintains a list of free frames. This is also
known as free frame pool, or chain or list. Any frame that is not allocated to any
process is known as a free frame. Whenever a page is released from memory after a
process is terminated or aborted the page is added to the free list. When an operating
system allocated pages to a process for storing instruction or data, it assigns a free page
from the free frames list.

Structure of Page Table

Hierarchical Page Tables
With the increasing need for larger applications, there is also a need for larger
memory space in a computer. As a result, it becomes difficult to work with long memory
addresses. In paged memory system the page table itself becomes very lengthy. The
solution is hierarchical paging.
In this, the page table itself is paged. There are two-page tables: outer and inner.

The inner page table is similar to that of a single-page table scenario. The outer page
table stores the links for the inner page table. The logical address consists of three parts:

131
outer page table entry, inner page table entry, and offset.

Figure: Hierarchical Page Table

In the process of hierarchical paging, a logical address has been divided into three
parts: X, Y, and Z. the symbol X represents the entry in the outer page table, and Y
represents the entry in the inner page table, Z represents the offset. The Xth position in
the outer page table store the address of the inner page table associated with X. The Y
position is searched in that inner page table and Y contains the address of the frame
containing the searched page.

Hashed Page Table

This page table is created of length M. Whenever, logical address of generated, a

hashing function is applied to the page Number p, to generate an index value i.

i = p % M;

The index value i is used to indexing into the page table. Each entry in the page
132
table is a pointer to a link list. It will provide a mapping between page number p and the
corresponding frame number f. It accessed through the index value i will be traversed, till
a match forms for page number p and the corresponding frame number f is obtained.

Figure Hashed Page Table

Inverted Page Table (IPT)

An inverted page table is similar to a simple page table along with another entry in
the table i.e. process id along with page number. IPT is not process-specific and it does
not need switching during content switching. A logical address generated by CPU
contains:
Process id (P id),
Page number (p), and
Page offset (d).
Process id & page number is found and corresponding offset (d) which gives the
frame number (f), where the desired page is residing. The frame number (f) combined
with the offset d, gives the intended physical address.

133
 Figure: Inverted Page Table

Translation Look Aside Buffer (TLB)

The page table is implemented in the hardware can enhance the performance
substantially. However, it does increase the cost proportionally. Though several schemes
of a hardware implementation of the page table exist, in the simplest case, the page table
is implemented as a set of dedicated registers. Care must be taken to employ very high-
speed logic for these register otherwise the performance will not be up to the mark. The
CPU dispatcher reloads these registers, just as it reloads the other registers. Instruction to
load is modified the page-table register are, of course, privileged, so that only the
operating system can change the memory map. This architecture is used in the DEC PDP-
11 computers. It has a 16-bit address while the page size is 8K. As a result, the page
consists of 8 entries that are kept in fast registers.

Having a dedicated set of registers for the implementation of the page table has a
serious limitation. The register can be used for the purpose only when the page table is
reasonably small. Modern computers are capable of having page tables containing as
many as 1 million entries. Register implementation in such a system is grossly infeasible.
The modern practice is to implement the page into the main memory rather than in the
register. The page table is maintained in the main memory and is accessed by a pointer
stored in the base register PTBR (Page Table Base Register). One advantage of this
practice is that changing the page table requires only changing the register value thus
avoiding time-consuming context switching.
134
 This scheme is not without its downside. It suffers from memory access delay. With
this approach, a memory location is accessed indirectly. First, the logical memory is
indexed into the page table. The page table itself is accessed using the content of the
PTBR register which provides the frame number, which is combined with the page offset
to produce the actual address.

This approach requires two accesses to memory – one for the page table and one for
the index into the page table thereby reducing the CPU time it may become unacceptable.
The process can be accelerated using high-speed associative memory or a set of
translation look-aside buffer (TLBs). In associative memory, each register memory each
register contains two entries – one for key and another for a value. During a search, the
keys are matched with the search value and the corresponding value field is the result if
the key was found. The use of associative memory enhances the search speed though at a
price for the additional hardware support.

Even associative memory is not unlimited. Therefore, only a few entries of the page
table are stored in the associative memory. When the CPU generates a logical address
while executing a process the page number of the generated address is searched into the
current associative memory for where the frame number is acquired. However, when the
page number being searched does not exist in the associative memory the same is
searched into the memory and the associative memory is updated with this page number
so that it may be found in the next reference top the associative memory.

In this arrangement, a page being searched may not always be found in the
associative registers. The extent to which a reference page is found in the associative
memory is expressed in terms of hit ratio. Hit ratio is the percentage of times that a page
number being searched is found in the memory. For instance, a hit ratio of 60% indicates
that out of 100 times a page being searched it is found in the associative memory 60
times. Therefore, one of the performance goals of an operating system is to increase this
hit ratio.

The time delay in memory access can be easily estimated in the following manner.
Assuming that it takes 10 nanoseconds to search the associative registers, and 50
nanoseconds to access memory, the mapped memory access would take 10 + 50 = 60 ns
provided the page number is found in the associative registers.
135
 However, if the search ends in a miss in the associative register after spending 10 ns,
then memory access is made for the page table and frame number taking additional 50 ns.
Following this, the desired byte is accessed in the memory costing another 50 ns. The
total time elapsed comes out to be 50 + 50 + 10=110 ns.

Example: Assuming a hit ratio of 60%, the effective or expected memory-access time
can be calculated as shown here under:

Effective access time = (Probability of hit) * Access Time on hit +

(Probability of miss) * Access Time on miss
= 60% of 60 + 40% of 110

= 80 nanosecond

One way of increasing the hit ratio is to use more associative registers. However, as
mentioned earlier it can be very costly. Therefore, usually, a trade-off is considered in
most cases. One estimate indicates that a hit ratio of 80% to 90% can be achieved with 16
to 12 associative registers.

Figure: Paging with TLB

Protection

136
 The issue of memory protection in a paging system is handled by attaching a few
bits to each of the frames in the page table. The bit may indicate read-only or read-write
or execute-only behaviors of the frame. At the time of accessing the frame number from
the page table, the corresponding bits are also examined.

Here if a process attempts to write into a read-only frame a system interrupt is

generated by the hardware. This interrupt is caught by the operating system, which in turn
takes the memory protection violation action.

Notes

In addition to the protection bits, one more bit attached to each frame. This bit
determines whether the frame/page is valid or invalid.

The ―valid‖ status of the bit indicates that the associated page in the process‗s
logical address space, and is thus a legal page. Otherwise, the page is not in the process‗s
logical address space. Illegal addresses are trapped by the operating system using the
valid-invalid bit. The operating system controls the accessibility of a frame by setting this
bit.

1 E
Page A
2
Page Page No. Valid/
Page B
Invalid Bit 3 B
Page C 4
A I
Page D 5
B 3 V
Page E 6
C 7 V
Page F 7 C
D I
Main Memory
Logical
Memory E 1 V

F I

137
 Page Table

Figure: Paging Protection

Memory Segmentation

Despite all the obvious merits of a memory-paging scheme of memory management, it

does not reflect the way a programmer would like to view the memory. In a paging
memory system, there is a clear-cut distinction between the users and system views of
memory. The user‗s view of memory is not the same as the actual physical memory. The
user‗s view is translated onto physical memory by the operating system with the use of
special hardware. It is due to this translation that the logical view is different from the
physical view of the memory.

A programmer would not like to think that the memory is simply a larger array of
bytes wherein at someplace data is stored and at another executable code is stored. The
most preferred view of the majority of programmers is that a program is divided into
different sections or segments. If you have programmed in COBOL you would appreciate
that a COBOL program has different divisions each having one or more optional sections.

Similarly, in a C program, the code is divided between many functions. Your

program may have any number of functions, subroutines, data structures, and several
modules. Each of these modules or data elements is assigned a unique name by the
programmer. Where and how these elements are loaded in the memory is not the concern
of the programmer. The programmer does care as to whether the segment containing
symbols table is stored in the memory before the main function or after it. Each of these
segments must be of variable length. The element of the program should be addressable
by the respective names assigned to them.

138
 Figure: User’s view of a program

An operating system may adopt a memory management scheme to simulate this

view of memory.
Segmentation is a technique of non-contiguous memory management technique that
employs this scheme. In this scheme, the logical memory space is divided into certain
unequal size chunks known as segments. According to segmentation, an application is
loaded into memory as a collection of modules. In other words, logically related
instruction and data items are grouped and loaded into a segment.
For example, an application can consist of five modules: the main program, function
A, function B, the stack for storing data when a function is called, and shared data for all
modules such as a global variable. Each module is loaded in a distinct segment in the
main memory.

Accessing Addresses in Segmentation:

In a segmented memory, a logical address consists of two parts: segment number

and offset.

139
 s d

Segment No Offset

Figure Address of Segment

An offset is a location within a segment, where any particular data is stored. For
example, the logical address 3: 10000 indicates offset 10000 and segment number 3. A
segment table is used for the validation of an address. A segment table has two fields:
base address of the segment, and size of the segment. The base address of a segment is
the location of a particular segment memory and its corresponding segment table:
0
10000
Segment Base Address of Size of the
Segment 1
No the Segment Segment 20000

0 80000 6000
Segment 2
1 10000 10000 35000
2 20000 15000 40000

3 40000 35000
Segment 3
4 90000 5000
75000
80000
Segment 0
86000
90000

Segment 4
95000
Free Space

Figure: Segment Table

Segment and Segment Table

The above figure segment table elaborates various memory segments in the main memory
and their corresponding entries in the segment table. Segments need not be loaded
sequentially or contiguously. For example, segment 2 has been loaded in a lower position
140
in the memory than segment 0.

The starting address and size of a segment are stored in the segment table. For
example, the base or starting address of segment 2 is 20000, and is the size of segment 2
is 15000 bytes.

When accessing an address in a segmented memory, the segment table is used for
address validation. The logical address 2: 12000 indicates the location 12000 starting
from the base of segment 2. Segment 2 starts from location 20000. The size of segment 2
is 15000 bytes. In other words, range of the segment 2 is 20000 to 74FFF. The physical
address corresponding to the logical address 2: 12000 is 20000 + 12000=32000. the
address 32000 is within the range of segment 2 (i.e. 20000 to 74FFF as 32000<74FFF).
Hence, the logical address 2: 12000 is a valid address.

Implementation of Segment Tables

As in the case of the paging system, segmentation can be implemented in a variety

of ways. The segment table can reside either in the register or in fast memory. The
advantage of register implementation is that both the limits checking and referencing can
be done extremely rapidly and simultaneously.
The feasibility of register implementation depends on the number of segments a
program has. The larger the number the more register are required making the proposition
infeasible. In smaller segmentation register implementation may be recommended.
However, modern programs tend to become very large causing a large number of
segments. Therefore, in such a system it is more appropriate to maintain the segment
table in the memory rather than in the registers. The data structure used for
implementation is often a Segment-Table Base Register or STBR. This pointer points
to the segment table located in the memory. Note that the number of segments used by a
program varies widely. Therefore, a Segment-Table Length Register or STLR is used.
The address translation for a given logical address (s, d) proceeds as follows:

1. Check that the segment number is legal (i.e. S< STLR)

2. Add the segment number to the STBR(i.e., STBR+S)

3. Read entry from memory at STBR + S

4. Check the offset against the segment length (d< limit)

141
 (a) if true then compute the physical address of the desired bytes as S+d

(b) else raise trap to the OS

Figure: Segmentation Hardware

Both the schemes- paging and segmentation requires two memory references during
address translation. This slows down the OS effectively by a factor of 2. As in the case of
paging, a set of an associative registers can be used to speed up the operation. It has been
found that a small set of associative registers generally reduces the time required for
memory accesses.

Protection and Sharing

Segmentation enforces protection intuitively. Since each segment is a semantically

cohesive unit all the corresponding entries are likely used similarly. Therefore, one needs
to protect at the segment level using a few protection bits. For instance, since codes do
not rewrite themselves the code segment can be assigned a read-only status whereas the
data segment can be allowed to perform read and write operations. The memory mapping
hardware will check the protection bit associated with each segment table entry to
prevent illegal access to memory, such as attempts to write into a read-only segment or to
use an execute-only segment as data. For example, by placing an array in its segment, the
memory-management hardware will automatically check that array indexes are legal and
142
do not stray outside the array boundaries. This enables the system to take corrective
measures early on if a memory protection error does occur.
The main advantage of segmentation is that it can be shared between processes. This
is true, particularly with code segments. For each process, a segment table is associated,
which the dispatcher uses to define the hardware segment table when this process can
share the same code segment as shown in the figure.

Segment Table P1
0
4000

Segment 1
4200
Segment Base Address of Size of the
No the Segment Segment
0 6000 500 5000

1 4000 200
Segment 3
2 8000 600 5400
3 5000 400
6000
Segment Table P2
Segment 0
Segment Base Address of Size of the
6500
No the Segment Segment
1 4000 200 8000

4 8600 200 Segment 2

8600
5 8800 200
Segment 4
8800
143
Segment 5
9000
Free Space

Figure: Segmentation Sharing

Here Segment 1 is shared between P1 & P2

Sharing of the segment in a segment memory system

Here in this example, segment (1) is being shared between the two processes p1 and
p2. Simply by managing entries in the segment table of processes a segment can be
shared among them. The sharing occurs at the segment level. Thus, any information can
be shared if it is defined to be segmented. Several segments can be shared in this way.

Segmentation with paging

Segmentation of memory can be implemented with a paging scheme. Unequal size

segments can be represented by equal size pages. For example, in a logical address space
of 64 KB, there are 32 pages of 2KB each. A program is divided into 3 segments of the
size 15 KB, 18 KB, and 10 KB. These three segments require 8, 9, and 5 pages to be
loaded into memory. The first segment occupies the memory space of 8 pages of 2 KB
each or 16 KB, whereas its requirement is 15 KB memory space. Thus internal
fragmentation of 1 KB occurs.
In segmentation with a paging scheme two tables are used: a segment table and a
page table. The segment table has two fields: base address and size. The base address of a
segment is the address of the first page in the segment. The size of a segment is the
number of pages a segment occupies. The page table stores the addresses of the frame

where a page is loaded into the main memory. For example, the value of the 15th cell of
the page table is 38000. This means page 15 has been loaded into the frame located at the
address 38000 in the main memory.

144
 A logical address has three parts: segment number, page number, and offset. The

logical address 4:3: 9E indicates the address 9E on page number 3 of segment number 4.

Figure: Segmentation with paging

The main memory is divided into certain equal size frames of 256 bytes (decimal
256 = 100 in hexadecimal) each. The address 6A of page number 2 of segment number 1

is being located. In the segment table, we found that the address of the first page in the
page table of segment 1 is 4A. Segment 1 is consists of 4 pages.
From the page table we see that address of the frame corresponding to page 2 has
been stored in the location 4C in the page table (because 4a + 2 = 4C). The cell number
4C of the page table is loaded in the frame located at the address 00200 in the main
memory.

Shared Pages

In multiprogramming, more than one program can be loaded. It is common for many
users to be executing the same program. If individual copies of these programs were
given to each user, much of the main memory would be wasted. Its solution is to share
those pages that can be shared. One copy of read-only code shared among processes (i.e.,
text editors, compilers, window systems). Shared code must appear in the same location
in the logical address space of all processes. Each process keeps a separate copy of the
code and data. The pages for the private code and data can appear anywhere in the logical
address space.

145
Example

Let, there are four processes P1, P2, P3, and P4. P1 has four pages lib1, lib2, lib3,
and data1. P2 has four pages lib1, lib4, lib5, and data2. P3 has four pages lib1, lib2, lib5,
and data3. P4 has four pages lib1, lib2, lib4, and data4. So, to load all processes 16
frames are required. In all these processes some pages are shared. So, if we share those
pages and loaded all shared pages once by giving the same frame number, then it needs
only 8 pages as shown in the figure below. 0 lib 5
1
2 lib1
lib1 2 lib1 2 3
lib 2 6 lib 4 12 4 data 1
5
lib 3 9 lib 5 0
6 lib 2
data 1 4 data 2 7 7 data 2
Process Page Process Page 8
9 lib 3
P1 Table P1 P2 Table P2
10 data 3
11
12 lib 4
13
14 data 4
lib1 2 lib1 2
15
lib 2 6 lib 2 6
lib 5 0 lib 4 12
data 3 10 data 4 14
Process Page Process Page
P3 Table P3 P4 Table P4

Shared Pages among Processes

Figure: Shared Pages

Demand Paging

The basic concept of virtual memory is storing instructions and data of a program in the
secondary memory, and then they are loaded in the main memory. In other words, the
part of a program that is required at any instant is loaded in the main memory while the

146
rest of the program is in the secondary memory.

Virtual Memory

Virtual memory is useful especially in a multitasking operating system, where memory

available in the main memory for user application is divided among several user
programs and every program has to compete for its memory requirement.

Virtual memory is a technique that permits the execution of processes, with their code
only partially loaded into physical memory. Virtual memory is used for the separation of
logical address space available to the user and the actual physical memory. CPU-
generated addresses are known as logical addresses or virtual addresses.

Programmers use virtual addresses in applications. The MMU converts the virtual
addresses into the corresponding physical memory address.

Programmers are notified that they can fully utilize the logical address space. Since
the logical address space, programmers have the illusion that they have a larger memory
space at their disposal.

Virtual memory is implemented using a non-contiguous memory allocation technique

known as demand paging. This technique is similar to paging except that in-demand
paging, pages are swapped in and out of main memory.
A program is initially stored in the secondary memory. When a page in the program
is required, it is swapped into the main memory. This is called demand paging because
until a page is not required, it is not loaded.
In demand paging, MMU also maintains an indexed table for address translation,
known as a page table, similar to that of a simple paging scheme. The page table stores
the information about, which page of a program is stored in which frame of the physical
memory.
When a page is required, the operating system scans the page table to find out the
physical address in the main memory that is in which frame the page is loaded. If the
page is found, the operating system continues its processing. In case the page is not
present in the main memory, a page fault occurs. A page fault is not a fault or an error;
rather it indicates a situation that the page is requested by a program that is not currently
147
loaded in the main memory.
When a page fault occurs, MMU swaps in the requested page from the secondary
memory into the main memory. In case there is no free space in the main memory, MMU
finds free space in the secondary memory, which is also known as the backing store in

the context of virtual memory. MMU swaps out an unused page from the main memory
and stores the unused page in the free space in the secondary memory for creating space
in the main memory. The requested page is loaded into the main memory. Finally, the
page table is updated.
Another way is a page table with a valid or invalid bit. The page table includes a
valid or invalid bit for each entry. When a page is loaded in memory its frame number is
entered and the page validity bit is set to valid. Thus if the bit is set to valid, it indicates
that the page is in memory. If the page presence bit is invalid, it indicates that either page
does not belong to the logical address space of the process, or it is still not loaded into
memory.

CPU

Request Page 1 which is not in Page Table, So Page Fault occurs. Frame nowhere the
pages stored in the memory

Page No Frame No Presence Bit

Frame
0 2 1
No
0 Page 3
1 0 Page 6
0
1 Page 5
2 0 Page 2
2 Page 0
3 1
3 3 Page 4
1 1
4 Physical Memory
0 1
5 0 Page 7

6 Page 1
7
Secondary
Page Table
Memory
148
 0 Page 0
1 Page 1
Page 2
2
Page 3
3 Page 4
Page 5
4
Page 6
5 Page 7
6
7
Logical Memory
Figure: Before Page Replacement

Let, CPU request page 1. This page is not present in the page table because the
presence bit of page1 in the page table is 0 that indicates the cell is empty. Thus page
fault occurs.
When a page fault occurs, the operating system searches and loads that page in the
memory replacing an existing page.
Page 1 caused a page fault. Page 0 is swapped out of the main memory and stored in
the disk to create room for page 1 in the main memory. Then page 1 is swapped in the
main memory.
Page No Frame No Presence Bit
Frame
0 0
No
2 1
1 0 Page 6
0
1 Page 5
2 0
2 Page 1
3 1
3 3 Page 4 Page 3
1 1
4 Physical
0 1
Memory Page 2
5 0

6 Page 0

7 Page 7

Secondary
149 Memory
 Page Table

0 Page 0
Page 1
1
Page 2
2 Page 3
3 Page 4
Page 5
4 Page 6
5 Page 7

6
7
Logical Memory

Figure: After Page Replacement

In the above figure, an abstract and basic idea of the functionality of virtual memory has
been shown. Actual implementation varies in different operating systems.

Advantages

 It uses memory more efficiently.

 There is no limit on the degree of multiprogramming.

Steps in Handling a Page Fault

Initially, memory is empty, if there is a reference of a page, the first reference to that page
will trap to an operating system which is page faults. To handle the page fault there are
the following steps:

1. Operating system looks at another table to decide whether the demand is valid
or not in memory. If it is invalid demand then it is aborted, else it is not in
memory.
2. After that it searches the empty frame if it is not available then the victim finds
and swaps out.
3. Demanded page is swapped into that frame.
4. Page tables are reset which assign the frame number to the corresponding page
number.
150
 5. Set the valid-invalid bit to valid bit.
6. Restart the instruction that caused the page fault.

Figure: Steps in Handling Page Fault

Performance of Demand Paging

Demand paging can have a significant effect on the performance of a computer system.
Let us calculate effective access time for a demand paged memory. Let P be the
probability of a page fault 0 < P < 1.0 if P = 0 no page faults and if P = 1 every reference
is a fault.

Effective Access time =

(1 – p) × ma + pft
p = page fault

ma = memory access time

pft = p × (page fault overhead + swap page out + swap page in + restart overhead)

EAT is directly proportional to the page fault rate. If the page fault rate is low then
EAT is decreased otherwise.

151
Example: Let memory access time = 50 nanoseconds
Average page fault service time = 4 ms
Then EAT = (1 – p) × ma + pft

= 1 – p × 50 + p × (4 ms)

= 50 - 50p + 4000000p

= 50 + 3999950p (nanosecond)

Page Replacement

As the number of processes and the number of pages in the main memory for each
process increase, at some point in time, all the page frames become occupied. At this
time, if a new page is to be brought in, the OS has to overwrite some existing pages in the
memory. The page to be chosen is selected by the page replacement policy.

A page is a fixed-length block of memory that is used as a unit of transfer between

physical memory and external storage like a disk, and a page fault is an interrupt (or
exception) to the software raised by the hardware when a program accesses a page that is
mapped in address space but not loaded in physical memory. The hardware that detects
this situation is the memory management unit in a processor. The exception handling
software that handles the page fault is generally part of an operating system. The
operating system tries to handle the page fault by making the required page accessible at
a location in physical memory or kills the program in case it is illegal to access.

The hardware generates a page fault for page accesses where:

The page corresponding to the requested address is not loaded in memory.

The page corresponding to the memory address accessed is loaded, but its
present status is not updated in hardware.

Example

Let program A have four pages i.e. P1, P2, P3, P4. Currently, P1, P2, P3 are in
memory and it requests page 4 i.e. P4. Program B has four pages i.e. P8, P9, P10, P11.

152
Currently, P8, P10, P11 are in memory. Only six pages can be stored in memory and
currently, P1, P2, P3, P8, P10, P11 are in memory so there is no memory space. To load
P4, one page must be swapped out. Let it be P2, P2 store in frame 3 which is swapped
out, and P4 is loaded in frame 3 which is currently free. The figure shows memory
position before swapping page 4 i.e. P4 and the next figure shows the memory position
after swapping page 4.

Page No Frame No Valid/

Invalid Bit Operating
0 Page 1 Page 1 4 v System
Page 2 3 v Page 11 1
1 Page 2
Page 3 6 v Page 8 2
2 Page 4 Page 3 Page 2 3
i
3 Page 1 4
Page 4
Page 10 5
Logical Address Page Table of Program A Page 3 6
Space of Program A Secondary
Memory

Page No Frame No Valid/

0 Page 8 Invalid Bit
Page 9 2 v
1 Page 8
Page 10 i
2 Page 11 Page 9 5 v
3 1 v
Page 10
Logical Address Page 11
Space of Program B Page Table of Program B

153
Figure showing the page table, logical address, and main memory before swapping

Page No Frame No Valid/

Invalid Bit
0 Page 1
1 Page 2 Page 1 4 v
2 Page 3 i
Page 2
3 Page 4 6 v
Logical Address Page 3 3 v
Space of Program A
Page 4
Page Table of Program A

Operating
System Page No Frame No Valid/
0 Page 8 Page 11 1 Invalid Bit
Page 9 Page 8 2
1
Page 10 Page 8 2 v
Page 4 3
2 Page 11 i
Page 1 4 Page 9
5 v
3 Page 10 5
Page 10 1 v
Logical Address Page 3 6
Figure
Secondary Page 11
Space of Program B showing Memory
the page table, logical address, and Page Table of Program

main memory after swapping

Page Replacement Algorithms

154
When a page fault occurs, the operating system has to choose a page to remove from
memory to brought a demanded page in memory.
A page replacement algorithm is logic or policy regarding how to select a page to be
swapped out from main memory to create space for the page, known as the requested
page, which has caused a page fault. These are several page replacement algorithms such
as :

 First In First Out (FIFO)

 Optimal Page Replacement Algorithm
 Least Recently Used (LRU)
 Clock Algorithm (Second Chance Algorithm)
 Counting Based Algorithms
 Least Frequently Used (LFU)
 Most Frequently Used (MFU)

First In First Out (FIFO) Algorithm

The FIFO algorithm is the simplest of all the page replacement algorithms. It conveys a
basic idea that when a page fault occurs, the oldest page in the main memory is to be
swapped out of the main memory to create a room or the required page that needs to be
executed. It replaces the page that has been in the memory longest.
The oldest page in the main memory is one that should be selected for
replacement first. ( if the number of frames is 3 then page repeated 3 times
will replace)
For example, let there are four pages i.e. 14, 21, 18, 36. Pages number 14, 21, and 18
are present in the memory and page number 21 was loaded first, followed by 14 and then
18. If page number 36 is required to be accessed by the CPU, it results in a page fault
because page number 36 is not present in the main memory. The page loaded first, that is

155
page number 21 is swapped out of the main memory and stored in the secondary memory
to create room for page number 36. Then page 36 is loaded in the memory. One possible
implementation is a FIFO queue of existing pages in memory. The oldest page will be at
the FRONT of the queue. Whenever a page fault occurs, the page at the FRONT of the
queue is mode victim and the new page is put at the REAR of the queue.

Advantages

It is very simple and easy to implement.

Programmers can easily code this algorithm.

Disadvantages

This algorithm has a severe drawback. If the oldest page is accessed frequently, the
performance of this algorithm declines, since whenever the page swapped out
will be required, it will result in another page fault, which degrades the
performance.

Reference String

A reference string refers to the sequence of page numbers referenced by a program during
its execution.
Assume a reference string:
1, 2, 3, 2,1,4, 2,1, 6, 1, 4, 2,5, 6, 2,1, 6, 3, 2
Let frames 3 are available for allocation page frame. Find number of page faults.

1 2 3 2 1 4 2 1 6 1 4 2 5 6 2 1 6 3 2
1 1 1 4 4 4 2 2 2 6 6 6
2 2 2 1 1 1 5 5 5 3 3
3 3 3 6 6 6 1 1 1 2

Initially

156
empty Frames No Page Fault

Initially, all 3 slots are empty, so when 1, 2,3 came they are
allocated to the empty slots —> 3 Page Faults.

Next page 2 refer, it is already in memory so —> 0 Page Faults.

next page 1 refers, it is already in memory so —> 0 Page Faults.

next page 4 refer, it is not available in memory and no free frame so it replaces the oldest
page slot i.e 1. —>4th Page Fault

Next page 2 refer, it is already in memory so —> 0 Page Faults.

Next page 1 refers, it is not available in memory and no free frame so it replaces the
oldest page slot i.e 2. —>5th Page Fault

next page 6 refer, it is not available in memory and no free frame so it replaces the oldest
page slot i.e 3. —>6th Page Fault

Next page 1 refers, it is already in memory so —> 0 Page Faults.

Next page 4 refer, it is already in memory so —> 0 Page Faults.

next page 2 refer, it is not available in memory and no free frame so it replaces the oldest
page slot i.e. 4 —>7th Page Fault

next page 5 refer, it is not available in memory and no free frame so it replaces the oldest
page slot i.e. 1 —>8th Page Fault

Next page 6 refer, it is already in memory so —> 0 Page Faults.

Next page 2 refer, it is already in memory so —> 0 Page Faults.

next page 1 refers, it is not available in memory and no free frame so it replaces the
oldest page slot i.e. 6 —>9th Page Fault

next page 6 refer, it is not available in memory and no free frame so it replaces the oldest
page slot i.e. 2 —>10th Page Fault
157
next page 3 refer, it is not available in memory and no free frame so it replaces the oldest
page slot i.e. 5 —>11th Page Fault

next page 2 refer, it is not available in memory and no free frame so it replaces the oldest
page slot i.e. 1 —>12th Page Fault

The number of page faults in FIFO is 12

Let frames 4 are available for allocation:

1 2 3 2 1 4 2 1 6 1 4 2 5 6 2 1 6 3 2
1 1 1 1 6 6 6 6 3
2 2 2 2 1 1 1 1
3 3 3 3 2 2 2
4 4 4 4 5 5

Page fault in FIFO with 4 frames is 9.

By observing these two examples it states that if the number of frames will increase
then the number of page fault must be decrease

158
 Graph of Page Faults Versus The Number of Frames

The above graph reveals that if the number of frames increases then the number of
page fault decreases but Belady‗s anomaly proves that if the number of frames increases
then it may be increased in page fault also while using the First in First Out (FIFO) page
replacement algorithm.
Let string is 5, 6, 7, 8, 5, 6, 9, 5, 6, 7, 8, 9, 5, and frameset is 3.

5 6 7 8 5 6 9 5 6 7 8 9 5
5 5 5 8 8 8 9 9 9 5
6 6 6 5 5 5 7 7 7
7 7 7 6 6 6 8 8

There are 10-page faults

Now with 4 frames set
5 6 7 8 5 6 9 5 6 7 8 9 5
5 5 5 5 9 9 9 9 8 8 8
6 6 6 6 5 5 5 5 9 9
7 7 7 7 6 6 6 6 5
8 8 8 8 7 7 7 7

there are 11-page faults.

So by the increasing number of frames, it may be the increase of the number of page
faults like in the above example where with 3 frames set there are 10-page faults and with
4 framesets there are 11-page faults. This is the Belady‗s Anomaly.

159
 FIFO Illustrating Belady’s Anomaly

The Optimal (OPT) Page Replacement Algorithm

Optimal page replacement algorithm is considered the best possible page replacement
policy in a virtual memory theoretically but it is difficult to implement. According to the
optimal page replacement policy, the page in the main memory, which will not be
referred to for the longest time is swapped out from the main memory to create room for
the requested page. A page should be replaced, which is to be referenced in the most
distant future. Since, it requires knowledge of the future reference string, which is not
practical. In the optional page replacement algorithm, the number of page faults is
minimum as compared to another algorithm.
Practically, implementation of the optimal page replacement algorithm is done as
follows:
Usually, all pages are labeled with the number of instructions that will be executed
before this page will be used again in the future. When a page fault occurs the page with
the highest number is replaced with the requested page that has caused a page fault.
For example, let pages 14, 21, and 18 are present in the main memory. Page 82 is
required to be loaded resulting in a page fault since page 82 is not there in the memory at
160
 present. Page 14 is not required till the next 2000 instruction. Page 21 is not required till
the next 1500 instructions. Page 18 is not required till the next 1700 instructions. This

means page 14 is the one that will not be accessed by the CPU for the longest time. Page
14 is swapped out of the main memory to create room for page 82.

Advantages

It has the lowest rate of occurrence of page faults.

It improves the system performance by reducing overhead for several page faults and
swapping pages in and out when a page fault occurs.

Disadvantages

It is very difficult to implement.

The situation is very similar to that of implementing the SJF algorithm in process
management. It becomes very difficult for an operating system to calculate after
what interval a page is to be referred to.

Example:

Let a reference string as follow:

1, 2, 3, 2,1,4, 2,1, 6, 1, 4, 2,5, 6, 2,1, 6, 3, 2
Let frames 3 are available for allocation page frame. Find number of page faults.

1 2 3 2 1 4 2 1 6 1 4 2 5 6 2 1 6 3 2
1 1 1 1 1 1 5 1 3
2 2 2 6 6 6 6 6
3 4 4 2 2 2 2

Initially
empty Frames No Page Fault

There is a 9-page fault in the optional page replacement

161
algorithm but in FIFO there are 12-page faults.

Least Recently Used (LRU)

The LRU algorithm uses information about the pages accessed in the recent past to
predict the near future. The LRU algorithm is when a page fault occurs, the page that has
not been referred to for the longest time is swapped out of the main memory to create
space for the requested page that has caused the page fault. It replaces the page which has
been used least recently.

Implementation of LRU can be done in various ways. One of the common methods
to apply the LRU in a scheme for virtual management is using an array. The array stores
the information about the page present in the main memory. The front end of the array

stores the page accessed recently. The rear end of the array stores the page that has not
been accessed for the longest time.

Whenever a page, that is present in the main memory, is accessed, the information
about the page in the array is shifted to the front end of the array. If a page fault occurs
the page indicated by the rear end of the array is swapped out of the main memory and
the requested page is swapped in the main memory. Information about the page swapped
in is stored in the front end of the array.

For example, the array stores four-page numbers {12, 56, 27, 61}. Page 12 is at the
front end of the array. This indicates that page 12 has been accessed recently. Page 61 is
at the rear end of the array. This indicates that the page that has not been referenced for
the longest time is page 61. Page 27 is accessed. No page fault occurs because page 27 is
present as the main memory. Information about page 27 is shifted at the front end of the
array. The array becomes { 27, 12, 56, 61}. Page 43 is required to be accessed. A page
fault occurs since page 43 is not in the main memory. Page 61 is at the rear end of the
array, which indicates that it has not been accessed for the longest time. Page 61 is
swapped out of the main memory and information about page 61 is removed from the
array. Page 43 is swapped in the main memory and information about page 43 is inserted
at the front end of the array. The array becomes { 43, 27, 12, 56}.

Advantages

162
 It is very feasible to implement.
This is not as simple as the FIFO algorithm but not as complicated to implement as
the optimal page replacement algorithm.

Disadvantage

The LRU algorithm requires additional data structure and hardware support for
its implementation.

Let a reference string as follow:

1, 2, 3, 2,1,4, 2,1, 6, 1, 4, 2,5, 6, 2,1, 6, 3, 2

Let frames 3 are available for allocation page frame. Find number of page faults.

1 2 3 2 1 4 2 1 6 1 4 2 5 6 2 1 6 3 2
1 1 1 1 1 1 1 5 5 1 1 1
2 2 2 2 4 4 4 6 6 6 6
3 4 6 6 2 2 2 2 3 2

Initially
empty Frames
No Page Fault

Page fault in LRU is 12. In this case, it is higher than optimal because a page that swap
out is needed after one page so, it increases the page fault but in normal case page fault in
LRU is less than FIFO but greater than the optional page replacement algorithm.

Clock algorithm

The clock algorithm also known as the second chance algorithm is a variation of the
FIFO algorithm. In this, the page present for the longest time in the memory is given a
second chance to remain loaded in the main memory. When that page is encountered for
the second time, is swapped out to create room for the page that has caused a page fault.
It is also referred to as Not Recently Used (NRU). It replaces a resident page, which has
163
not been accessed in the near past.

Working

This is implemented using the concept of a circular queue. Each cell in the circular
queue contains two values: a page number and its corresponding reference bit. The value
of the reference bit can be either 0 or 1. If the value of a reference bit of a page is 1 it
means that the page was encountered as the oldest page and followed a second chance. If
the value of the reference bit is 0 this indicates that this page has not been encountered as
the oldest page yet. When a page is found the oldest page present in the memory for the
first time, its reference bit is set from 0 to 1. The next time when that page is found, it is
swapped out of the main memory for creating free space in the main memory.

Page Number Pointer

18 10 12 61 39

0 0 1 0 1

Reference Bit

Page Number Pointer

18 10 12 61 39

0 0 1 0 1

164
Pointer

18 10 12 61 43

0 0 1 0 0

Page 43 swapped in from the secondary memory replacing page 39

A circular chain or pull is organized to implement the clock algorithm for selecting a
page to be swapped out in a virtual memory scheme. The pointer is directing presently at
page number 61. a page fault occurs. Page 61 has not been given a chance to remain
loaded in the memory since its references bit is 0. The reference bit of page 61 is made 1
and the pointer is moved to the next page in the pull that is page 39.
The reference bit on page 39 is 1. This indicates that previously page 39 has been
given a chance to be remain loaded in the memory. Thus this page is swapped out of the
memory to create space for the new to swapped on page 43.

Counting Based Algorithms

Some page replacement algorithms apply the logic based on how many times a page
has been accessed. Usually, these types of algorithms are implemented using an array that
stores information about a page number and the number of times it has been accessed.
The type of counting algorithms is:

Least Frequently Used (LFU) page replacement algorithm

Most Frequently Used (MFU) page replacement algorithm
When a page fault occurs, the operating system verifies how many times it has been
accessed and proceeds according to the logic provided by the counting algorithm adopted
for page replacement in the virtual memory by the operating system.

Least Frequently Used (LFU)

165
 In the LFU algorithm, the page that has been accessed for the lowest/fewest number
of times from the time when the page is loaded in the memory is replaced when a page
fault occurs. The logic behind applying this algorithm is that some pages are accessed
more frequently than others. Counting how many times a page has been accessed is used
as an estimate of the probability of a page being referenced. Whenever a replacement is
necessary, a page with the least count is replaced. The main drawback of this algorithm is
that some pages may have a high usage initially and may build a high count but they have
low usage subsequently, would remain in memory due to high count.
Let a reference string as follow:

1, 2, 3, 2,1,4, 2,1, 6, 1, 4, 2,5, 6, 2,1, 6, 3, 2

Let frames 3 are available for allocation page frame. Find number of page faults.

1 2 3 2 1 4 2 1 6 1 4 2 5 6 2 1 6 3 2
1 1 1 1 1 1 1 1 1
2 2 2 2 2 2 2 2
3 4 6 4 6 6 3

Initially
empty Frames
No Page Fault

There are 9-page faults is LFU

Note: If the counter of all the pages is the same then we adopt FIFO.

Most Frequently Used (MFU)

This algorithm replaces the page with the largest usage count. It is based on the
assumption that the pages, with a smaller count, have been brought in recently and would
need to be resident.
1 2 3 2 1 4 2 1 6 1 4 2 5 6 2 1 6 3 2

166
 1 1 1 4 4 4 4 4 4 2 1 3 3
2 2 2 1 6 6 6 6 6 6 6 2
3 3 3 3 1 2 5 5 5 5 5

Initially
empty Frames
No Page Fault

There are 13-page fault is MFU

Note: If the counter of all the pages is the same then we adopt FIFO.

Thrashing

In certain situations, a system spends more time processing page fault by swapping
in and out pages than executing an instruction of processes. In other words, sometimes
handling the pages faults becomes a huge overhead. This situation is called thrashing.
Thrashing degrades the performance of a system. When multiprogramming increase more
which degrades the performance is known as thrashing.

In a situation, where too many pages are active and required frequently, a page is
required that is not present in the memory resulting in a page fault. One of the pages
loaded in the main memory, which is also accessed frequently, is swapped out to create
room for the required page. The required page is swapped in; the page swapped out is
required for execution causing another page fault. Thus, a series of pages fault occurs and
swapping becomes a large overhead.
In case a large number of processes are running in a multitasking operating system
simultaneously, the memory becomes over-committed and thrashing occurs. In another
word, the degree of multiprogramming is directly related to thrashing. The term, degree
of multiprogramming, indicates many processes that are being simultaneous. This figure
shows how system performance declines because of thrashing:

167
 Figure: Effect of Thrashing on System Performance

This figure shows that system performance in terms of throughput reaches an

optimal level to an extent if the degree of multiprogramming is increased. If the degree of
multiprogramming is further extended thrashing occurs and system performance
degrades. The peek of the graph represents the optimal performance of the system.
In certain cases, thrashing can be avoided. In most cases, an operating system
attempt to recover from thrashing by suspending the execution of current processes and
preventing the execution of new processes to start.
Selecting an appropriate page replacement policy to operate virtual memory plays an
important role in the paging of thrashing. A page replacement strategy based on the local
mode prevents the occurrence of thrashing to an extent. In local mode, while running a
process, When a page fault occurs, swapping out of a page, which belongs to another
process, is not allowed. A page belonging to the same running process is selected and

168
swapped out to create room for the required page. In such cases, the thrashing of a
particular process does not affect other processes. If other processes have occupied
enough frames in the main memory, they can continue execution properly.
A page replacement policy based on the global model can lead the system to thrash.
In global mode, when a page fault occurs, any page, regardless of which process it
belongs to, is swapped out to create room for the required process that needs immediate
execution. In global mode, a process may occupy a huge part of the main memory while
other processes are competing for room in the main memory and hence resulting in page
fault frequently.

Points to Remember

The logical address is generated by the CPU.

The logical address is also known as the virtual address.
Logical and Physical addresses are different for execution time address binding.
The base register is also called the relocation register.
The user program deals with the logical address.
Swapping is a technique to temporarily removing the inactive program from the
memory of a system.
Memory allocation methods are: first fit, Best fit, Worst fit.
Paging is a memory management method.
Logical memory divided it into similar size called pages.
Physical memory divided it into similar size called frames.
Paging eliminates fragmentation.
In segmentation, the program is divided into variable size segments.
Paging and segmentation can be shared.
Virtual memory allows the execution of partially loaded processes.
Virtual memory is the separation of user logical memory from physical memory.
Demand paging is similar to a paging system with swapping.
With the use of virtual memory, CPU utilization can be increased.
In demand paging, pages are swapped in and out of the main memory.
In demand paging, MMU also maintains an indexed table for address translation,
169
 known as a page table.
This table is similar to that of a simple paging scheme.
When the page is not present in the main memory a page fault occurs.
Demand paging uses memory more efficiently.
In demand paging, there is no limit on the degree of multiprogramming.
Effective Access time calculated as
(1 – p) × ma + pft
where p = page fault

ma = memory access time

pft = p × (page fault overhead + swap page out + swap page in + restart overhead)

EAT is directly proportional to the page fault rate.

Demand paging can have a significant effect on the performance of a computer
system.
Page replacement algorithms are:
First In First Out (FIFO)
Optimal Page Replacement Algorithm
Least Recently Used (LRU)
Clock Algorithm (Second Chance Algorithm)
Counting Based Algorithms
Least Frequently Used (LFU)
Most Frequently Used (MFU)
FIFO replaces the page at FRONT of Queue.
Optimal it replaces the page that will not be used for the longest time.
LRU replaces the page that has not been used for the longest time.
LFU replaces the page, which has the smallest count.
MFU replaces the page, which has the highest count.
Belady anomaly state that page fault rate may increase as the number of allocated
frames increases,
FIFO replacement algorithm suffers from Belady anomaly.
The degree of multiprogramming increases more which degrade the performance of
the system is known as thrashing.
170
 Thrashing degrades the performance of a system.

PRACTICE EXERCISES

1. What is memory management?

2. How Operating system helps in memory management?
3. What are the different types of memory management techniques?

4. Explain the following terms:

(i) Internal fragmentation

(ii) External fragmentation.
(iii) Compaction

5. Explain the strategies of the first fit, best fit, and worst fit in the multiple partition
memory management techniques?
6. What is the basic difference between Single partition and Multiple partition memory
management techniques?
7. What is Paging?
8. Why is page size always power by 2?
9. Explain the technique of Segmentation in memory management?

10. Why paging is used?

11. Describe the following:
— Swapping
— Structure of Page Table
— First Fit
— Best Fit
— Worst Fit
— Address Binding
— Production and Sharing
12 Write a note on virtual memory.
13. Explain the demand paging.
14. What do you mean by page fault?
171
15. Write steps handling the page fault.
16. Explain the performance of demand paging.
17. Why we need the page replacement algorithm.
18. Describe the working of various page replacement algorithms.
19. What do you mean by reference string?
20. What is the Belady anomaly?
21. Write a short note on Thrashing.
22. Consider the following page reference string:
1, 2, 3, 4, 7, 5, 1, 2, 1, 7, 5, 3, 2, 4, 2, 3, 1, 6, 4, 2, 1, 2, 3, 6
How many page faults will occur for the following page replacement algorithms?
Assume a set of four page-frames (initially all empty).
FIFO
LRU
Optimal

172
 M.Sc. (Computer Science)
OPERATING SYSTEM

UNIT IV : FILE SYSTEM

STRUCTURE

Objective
File Concept
File System
Introduction
File Attributes and Naming
File Attributes
User-Defined Attributes
System-Defined Attributes
Naming
File Operations
File Operations
Directory and Disk Structure
Single- Level Structure
Two- Level Structure
Hierarchical Structure or Tree-Structured Directories
Acyclic Graph Directories
General Graph Directories
File-System Structure
File-System Implementation
Directory Implementation
Allocation Methods
Free-Space Management
Practice Exercises

173
 OBJECTIVE

Understanding Concept of File, different methods to access files.

Using File system structure and their implementation,
Concept of Directory and its allocation methods

FILE CONCEPT
A file is a group of similar records which is stored in memory. The file is treated as one unit
by users and applications. It may be mentioned by name. The filename should be sole which
means in the same location file's name should be unique. It may be created, deleted,
appended, truncated. There should be file manager which provides a protection mechanism to
allow machine user to administrator how processes executing on behalf of different users can
access the information in a file.

File System
The file system in the operating system is assigned with the work of storing, controlling,
and managing data that is stored on disks or secondary storage in the form of files. File
system management is responsible for maintaining consistency in data when multiple users
access files concurrently. It also provides measures for file protection at times when the
system crashes.

Introduction
Files are stored permanently on secondary storage devices, such as hard disks. A file system
is a part of the OS that is responsible for controlling secondary storage space. It hides device-
specific complexities and provides a uniform logical view of users. A function of file system
includes:

It allows users to provide a facility to give the name of the file as user-defined
names, to create, append, truncate, and delete files.

It maps user-defined names with low-level spotters, names in a machine-

understandable format so that the machine finds a file uniquely.
It provides a uniform logical view of data to users rather than a physical view i.e.
internal structure of files in which they are stored on the disk, by giving user-
friendly interface.
It controls transferring of data blocks between secondary storage and main memory
and also between different files.
174
 It offers semantics or the file-sharing rules and regulations between numerous
processes and users.
It also allocates and manages space for files on secondary storage devices, such as
disks or magnetic tapes. Space management is an important part of the file system.
It protects the files from system failures and applies measures for recovery and
backup.
It provides security measures for confidential data such as electronic funds or
criminal records.

It also provides encryption and decryption facilities to users. Encryption is a

mechanism of converting data into some code form that is unreadable to everyone
except the recipient. Decryption is the reverse process of encryption.
A file system is implemented as a layer of OS and is placed between the kernel and the
memory manager. It consists of utility programs that run as constrained applications that are
used to control access to files. Users may access files with the help of the file system only.

File Attributes and Naming

The file is the named collection i.e. each unit is having a name, that is unique and the end-
user identifies a file by its name, usually considered as a linear array i.e. sequentially
arranged, of date and records. Almost all input/output operations are performed through files.
All inputs are done via files and all outputs are recorded in files. Data or information stored in
files is of many types. A file has a certain defined structure according to its types such as
executable programs, numeric and textual data, pictures, images, and sound recordings. Each
file is saved in a directory which is acts as a container for that file. Access to file is done
through these directories. A user assigns a name with each file through which it can refer.
Moreover, the file system requires more information about a file that is attached with the file
in the form of attributes to maintain consistency and reliability of data and control multiple
accesses.

File Attributes
File attributes are required by any file system to manage or maintain a file. It may differ from
one operating system to another operating system. File attributes are the information about a
file that is associated with every file. Some attributes are unique for each file in disks such as
file locations, name and creation, date, and time. Few attributes are accessible for users, such
as access privileges, name, or size of a file, whereas some of them are specifically assigned to
a file for file system usage.
175
 File attributes vary from one OS to another, but few of them are needed by every OS.
The major types of attributes are:-

(1) User-Defined attributes

(2) System-Defined Attributes

User-Defined attributes

File name: An identifier chosen by the user to address the file. Usually a string of
alphanumeric characters, some OS allows the use of special characters, such as
#, *, or $, in file names, and must be unique in its file directory.

File type: Type of information stored i.e. binary file, text file, picture file, or
program file.
Owner: Name of the creator of files that controls and provides access privileges,
such as read-only or read-write, to other users.
Permitted privileges: It contains information that determines read privileges, write
privileges and execute privileges.

System-Defined Attributes

Low-level identifier: Machine understandable names, usually in binary digits that

are used by hardware to identify a file by mapping it with a user-defined name. This
attribute also consists of numbers.

Location: Address of sector or area where the file is stored on the disk. Usually,
this information is used as a pointer; a value used by programs to find the location
of a certain sector or file, to the location and consists of numbers.

File size: Current file size in bytes, or words, or in blocks.

Creation date: This contains the date when the file was created.

Allocated size: Total allocated size of a file by the file system.

Volume: Indicates which type of device is used for storing files.

Date of last modification: Contains date and time of last update, insertion, and
deletion in a file.

All the attributes of a file are stored in a directory where the file resides. Their storage
takes more than 1 Kb space per file. Some attributes are stored in the header record associates
176
 with each file.

Naming

The most significant attribute of a file is its name that is given by the user who created it.
Users should associate a name with every file to uniquely identify it and access files through
these names. The file‘s name should be unique in its directory. In a shared system, it is
recommended that the user must assign a unique name to a file.

Files are accessed by giving a complete path or address i.e., you have to specify the
names of directory and subdirectories with the filename, all when combined makes a
complete address for a file. It is possible, that a directory is also stored in another directory
that is contained in the root directory, a directory at the top level. Therefore, a file in a system
can be located by a giving-by-giving complete path from the root directory to the files,
specifying all the intermediates directories, to the file system. All the directories, other than
the root directory, are the subdirectories of the root directory. The root directory is denoted
by ‗\‘.

An example of the pathname for a file:

C: \ Rohit \ rsts \ os.doc

The pathname of a file is also called the complete address for a file. Slash is used to
restricting names in a sequence and also indicate subdirectories of a directory. Two or more
files are allowed to have similar names provided they have different pathnames i.e. their
parent directory should be different.

The process that calls a file is associated with it a current directory also called the
working directory. In this case, a complete pathname is not required rather; a user can access
files giving pathname starting from this working directory.

File Operations

Basic operations on files are:

Create a File: For creating a file, address space in the file system is required. After
creating a file, the entry of that file must be entered in the directory. Then directory entry
contains the file name and location of that file which is stored in the file system.

Writing a File: A system call is used for writing into a file. It requires two parameters
one file name and the second information which is to be written.

177
 Reading a File: A system call is called to read a file. It requires two parameters: one the
file name and its memory address.

Delete a File: System will search the directory with the file to be deleted. It releases all
file space that can be reused by another file.

Truncating File: The user may need to remove the contents of the file but want to keep
its attributes. It recreates a file.
ACCESS METHODS
In the file, information can be accessed in various ways. Different types of file access
methods are:

1. Sequential access

2. Direct access

3. Indexed Access

1. Sequential Access

It is a very simple method among the other methods. In the file, information is serially
accessed which means one record after other records. For example: in a tape recorder, the
tape plays in sequential mode. A read process reads the file‘s next portion and automatically
advances a file pointer, that keeps tracks of Input/Output
location. A write operation Current Head appends the end of the file.

Position

Beginning of File End of File

Rewind Read or Write

178
2. Direct Access

It allows random access to any block. This model is based on a disk model of a file. It
allows programs to read and write records in any order.

3. Indexed Access

The records in a logical sequence according to a key contained in each record. The
system maintains an index containing the physical address of certain records. By using the
key attributes, the records can be indexed directly.

Sequential and direct access method both are not supported by all operating system.
Some operating system uses sequential access method, some operating system uses direct
access method and some operating system used both access methods.

DIRECTORY AND DISK STRUCTURE

Directories are considered as symbolic tables of files that are store all the related information
about the file it holds, with the content. This information includes file attributes, location,
type, and access privileges. Directories are also known as containers of files.

The directory is itself a file that is owned by the OS. Millions of files present in the
system need to be managed. Directories provide means to organize files in a structure. Each
entry in a directory contains information about a file. Similar to files, operations such as
insertion, deletion, and searching can be performed on it. Operations performed on these
entries are:

Searching a file: Whenever a file is referenced, the directory must search for the
related entry.

Create a file: An entry of every newly created file needs to be added to the
directory.
Delete a file: Whenever a file is deleted, the related entry should be removed from
the directory.
List directory: A list of files in a directory should be shown whenever a user
requests it.
Rename a file: The name should be changeable when the use of the file changed or
its location changes.
Update directory: Whenever a file attributes changes, its corresponding entry

179
 needs to be updated.
Based on these entries and their operations, the structure for directories can be organized
in different ways. The three most common structures for organizing directory are, single-
level, two-level, and hierarchical structures.

Single-Level Structure

It is the simplest form of directory structure having only one level of directories. The entire
files are contained in the same directory. It appears like the list of files or sequential files
having file names serving as the key. The logical structure of a single directory is given in the
figure below:

Root D1 D2 D3 D4 D5
Director
y
F1

File
s
Single Level Directory Structure

The directory structure is implemented in an earlier single-user system. It becomes

180
 outdated and inadequate in the multiple-user system. Even for a single user, it is difficult to
keep track of the files if the number of files increases. Moreover, files are of different types,
such as graphic files, text files, and executable files, and if the user wants to arrange these
files in an organized manner such as group files by type, this structure becomes inconvenient.

Files in the single-level directory should have unique names because they are contained
in one single directory. In a shared system, unique naming becomes a serious problem. These
drawbacks lead us to design another structure of directories named a two-level structure.

Two-Level Structure

As the name suggests, this structure is divided into two levels of directories i.e. a master
directory and user, and all these directories are contained and indexed in the master directory.
The user directory represents a simple list of files of that user.

The two-level structure looks like an inverted tree of height 2. The root of this tree is the
master directory having user directories as its branches. Files are the leaves of these branches.
The logical structure of two-level directories is shown in the figure below:

D1
Root
Director
y

User D11 D12 D13

Director
y
F1 F2 F3 F5

Files
F4

Two Level Directory Structure

A user name and file name are the pathnames for a file. This structure solves the
181
 problem of unique names up to a certain extent i.e. user can assign duplicate names to files
provided files are present in different directories. Names need to be unique in the user‘s

directory because two file names cannot be the same in one single directory. User searches a
file in his directory only, which allows different users to have files with the same names.

To create or delete a file, OS searches only the user‘s directory that initiates the
command. OS uses a special system program or system calls, to create or delete a user
directory. This system program creates or deletes a user directory entry from the master
directory.

This structure provides no help in grouping files of different types. In a shared system,
one user wants to access another user‘s file because files are shared in a shared system
network, which again creates the problem of uniqueness in file names. The user has to give a
complete pathname to name a file in another user's directory.

Hierarchical Structure or Tree-Structured Directories

This is the most powerful and flexible structure and is implemented in almost every OS. The
two-level structure is extended into a more advanced hierarchical structure of arbitrary levels.
It uses the same concept of a two-level structure of master directory having user-directories
as subdirectories. In addition, each user directory, in turn, has subdirectories and files as its
branches and leaves. A typical hierarchical structure of directories and files is shown in the
figure below.

182
 Directory

File

Tree Level Directory Structure

Users can create their subdirectories to organize files of different types, such as separate
subdirectories for graphic files, or separate subdirectories for text files. Special system calls

are used to create or delete directories. Internal formats, i.e. the internal structure in which the
details of the directory are stored, of each directory ha

s an entry that stores special bits representing a subdirectory or a file i.e. 0 bit represents a file
and 1 bit represents a directory.

The user always works on the files in the current directory. The current directory holds
all the files, which a user currently requires. OS searches the current directory for reference to
a file. In a hierarchical structure, the user can access a file, which is not in the current
directory, giving pathname. Users can change the current directory also through system calls.

In a hierarchical structure, a file can be referenced in two ways, absolute pathname, and
relative pathname.
183
 Absolute pathname starts from the root and ends at the required file following a path of
directories and subdirectories.

Relative pathname starts from the current directory to the file. Users can access another
user file by giving its pathname. In a hierarchical structure, the pathname to a file can be
longer than a two-level directory. This increases the search time for a file that resides in other
user directories.

Acyclic Graph Directories

It always directories to shared subdirectories and files. Some files or directories may be in
two different directories. Shared files and directories can be implemented by using links.
Link is implemented as an absolute path or relative pathname. It is more flexible than a
simple tree structure but sometimes it is more complain.

The acyclic graph directory structure

184
 Characteristics

1. Files and sub-directories can be shared.

2. It is more flexible than three structures.

3. Deletion of a shared file is difficult.

4. Consistency should be taken care of.

General Graph Directories

In this structure, cycles are also permissible within a directory structure are from more than
one parent directory, multiple directories can be made. The main disadvantage of the general
graph directory structure is to compute the total size or memory required which has been
taken by the files and directories.

Characteristics

It permits more cycles.

It is more flexible as compared to the other directories structure.

It is more expensive than others directory structures.
It requires garbage collection.

A General Graph Directory structure

FILE-SYSTEM STRUCTURE
185
The file system resembles a complete directory structure that includes a root directory
following subdirectories and files under it. The file system is required to enhance the function
of retrieval and storage of files. We can mount more than one file system that creates a
directory structure showing an image of a single file system. Commands used for file system
operations are:

Chfs: Change the file system characteristics

Crfs: Add a new file system

Rmfs: Remove a file system

Mount: Make files system available for user access.

In the real world, different Operating systems use different file systems. Each file system
has a different directory structure and different ways to represent files.

1. CD ROM

CD ROM is an acronym for Compact Disc-Read Only Memory. Data is stored on one
side of it that is coated with the highly reflected material. Data is stored in the form of series
of microscopic pits, i.e. data is stored in binary digits of 0 and 1 and a pit represents a binary
1, which is read by a low-powered laser in a drive unit. To read data, the laser rotates the disk
at various speeds at a Constant Linear Velocity (CLV).

CD ROM is used for storing audio in digital format and is of size 6547.4 MB. CD ROM
can store data equivalent to 500 floppy disks. Data on CD ROM can be accessed much faster
than any other tape. Accessing speed of CD ROM is 75-150 sectors per second. To read a CD
ROM, we need a CD ROM reader and an appropriate software driver that controls the
functionality of CD ROM.

Data is stored in the form of sectors with error-correction codes i.e. codes that are used
to debug any error if occurred in the stored data. Data on CD ROM can be stored and
retrieved, as files for that we need a file system to manage the stored data. The file system
allows you to access the contents of CD ROM using the interface of a normal file system i.e.
the file system stored on the disk.

2. FAT16

To understand the FAT 16 file system, a discussion on File Allocation Table (FAT) is

186
necessary. FAT is a table that has entries in the form of pointers. These pointers point to the
files that are present in a partition. Partition is a logical division of the table. Each entry in
this table consists of a block number, a 0 valued block number represents as unused block
any new entry is stored on it. Partition of the disk depends on the file system used. The two
most common file systems are FAT 16 and FAT 32.

FAT 16 is the old MS-DOS file system created by Microsoft in 1977 that used 16 bits
table. FAT 16 support filenames of 8.3 standard i.e. 8 characters for the name and 3
characters for file extension.

The value stored in the FAT table is 16 bits. A 16-bit file allocation table means that the
size of the address stored in the entry of FAT is limited to 16 bits. This implies that you can‘t
have storage allocation units i.e. the entries in the FAT 16 table are restricted to 65536 units.
This is the main drawback of FAT 16.

Another drawback of FAT 16 is that its partition is of fixed size i.e. of 32k. That wastes
a lot of disk space in storing files of small size such as pictures, HTML documents, etc.

3. FAT32

It is an advancement of FAT 16 that supports 32 bits addresses. Entries in FAT 32 tables

can hold a value up to 32 bits. It divides the hard disk into smaller partitions of a small disk
area than FAT 16 so that disk space is used more efficiently with increased storage capacity.

FAT 32 allows you to create a partition up to 2048 GB size. The number of clusters in
FAT 32 ranges around 4 billion i.e. 4,294,967,296 inexact. FAT 32 supports long file names
of a maximum of up to 255 characters. File names are not case sensitive i.e. no matter
whether the names are in capital letters or small letters; it saves the case of filenames once
created. Minimum disk space needed for FAT 32 is 8 GB. Os that supports FAT 32 are:
OSR2, Windows 98, Windows 2000, and Linux.

4. Unix

Unix is a multi-user and multitasking OS that allows multiple users to access multiple
files concurrently. It is a data structure that resides on disks. Unix file system has a tree-like
structure having a root directory (/) at the top node. This hierarchical structure appears to
users as a single file system having all types of files in one tree, the concept of drives such as
A, C: etc is not found here. Each node, a vertex of the UNIX file system tree, represents a
separate file system.

187
 Thus, the separate device is mounted in one single file system. A file in UNIX is a
collection of randomly addressable bytes, not sequentially; each byte has a different address,
not in the continuity of a previously-stored byte. Files are allocated on the block randomly on
the disk. The index method is used to keep track of each file. A filename can extent up to 255
characters excluding forward-slash (/) & Null and are case sensitive. File in Unix may
contain holes i.e. no data blocks. The directory is also a file except the user can‘t write on it.
Unix associates its file with three access permissions: read, write, and executable. Users can
access files giving absolute and relative pathname.

UNIX file system is made up of four parts:

Boot block: Contains boot program and load kernel into memory when a user boots
the system or switch on the system.

Superblock: Contains all information about the attributes of a file.

I-list: Contain a list of I-nodes of the file system. I-node is a data structure that
holds information about an individual file, such as file size, access privileges, or file
name. Each file is related to I-node and is identified by I-nodes number.

Data block: Contains actual file contents.

FILE-SYSTEM IMPLEMENTATION
File system structured in the layers which are: Application Programs, Logical File System,

File Organization Module, Basic File System, Input/Output Control, Devices

File
Application Logical File Basic File Input/Output
FILE SYSTEM Organization Devices
Programs System System Control
Module

Input/Output Control level –

Device drivers perform as an interface between devices and Operating Systems,
these drivers help to transfer data from disk to main memory and vice versa. It
picks block numbers as input and as output, it returns low-level hardware-
specific instruction.

Basic file system –

It refers to normal commands to the particular device driver to do operations:
read and write on particular physical blocks on the disk. A block is part of
188
 memory in the buffer that can contain the contents of the disk block and cache
stores frequently used file system metadata. It also handles the memory buffers
and caches.

File organization Module –

It contains information about the files such as their location, and their logical
blocks, and physical blocks. Physical blocks were entirely different so they do
not match with logical numbers of the logical block which are numbered from 0
to N. It contains some free space that is tracked by unallocated blocks.
Logical file system –
It manages metadata information about a file i.e includes all details about a file
except the actual contents of the file. It also keeps with the help of File Control
Blocks. File Control Block (FCB) contains all the information about a file such
as the owner of the file, size of the file, permissions granted, file location, and its
contents.
Advantages :
1. Replication of code is lessened.
2. Each file system can have its independent logical file system.
Disadvantages :
If someone accesses many files simultaneously then its outcomes will reduce the
performance of the system.
We can implement a file system by using data structures which are:
1. On-disk Structures –
Generally, it contains the following information:
a. total number of disk blocks available
b. free disk blocks available
c. location of them.
Below given are different on-disk structures :

189
 1. Boot Control Block –
It is generally the 1st block of volume and it has information that is
needed to boot any operating system. In UNIX, this block is known as a
boot block and in NTFS it is known as a partition boot sector.
2. Volume Control Block –
It contains information about a particular partition for example:- count of
free blocks, size of blocks, and its pointers also. In UNIX it is known as
superblock and in NTFS it is stored in the master file table.
3. Directory Structure –
They kept file names and linked inode numbers. In UNIX, it contains file
names and associated file names and in NTFS, it is stored in the master
file table.
4. Per-File FCB –
It holds details about files and it has an exclusive identifier number to
allow association with the directory entry. In NTFS it is kept in the
master file table.

Structure of File Control Block (FCB)

In-Memory Structure :
They are maintained in the main memory and these are helpful for file system

190
 management for caching. Several in-memory structures given below:
1. Mount Table: It is a table in which information about every mounted
volume is stored.
2. Directory-Structure cache – It contains the directory information about
which directory is freshly accessed.
3. System-wide open-file table –It holds the copy of the File Control Block
(FCB) of every open file.

4. Per-process open-file table – It holds information on which process

opened the file and it also maps with appropriate system-wide open-file.

DIRECTORY IMPLEMENTATION

A directory can be implemented in two ways:

1. Linear List

2. Hash Table

Linear List

The linear list is the simplest method. In this linear list of filenames used. For searching
purposes, a linear search technique is used to find a particular entry. It is very simple but
time-consuming to execute. Directory information is used often. Users also notice the slow
implementation of access to it.

Hash Table

With the hash table, it decreases the directory search time. Insertion and deletion are
very simple. Hash table takes the value which is compiled from the file name and then it
returns a pointer to a file name in the form of a linear list. It always uses a fixed size.

ALLOCATION METHODS
The space allocation strategy is often closely related to the efficiency of file accessing and of
logical to physical mapping of disk addresses. Three major methods of allocating disk space
are widely used which are:

Contiguous Allocation Method

The contiguous allocation method always requires every file to occupy a set of adjoining

191
addresses or blocks on the disk. Disk addresses describe in a linear order on the disk. The
important thing to be noted is that, in this ordering, the block of memory represents as B,
accessing block where is B+1 after block B normally requires no head movement. When head
movement is needed, it is only one track. Thus, the number of disks seeks required for
accessing contiguous allocated files is minimal.

Contiguous allocation of a file is defined by the disk address or blocks and the length of
the first block. Suppose, the file has a size of N blocks, and its starting location is location B,
then it must occupy consecutive blocks namely B, B+1, B+2,….., B+n-1. The directory entry
must contain for each file indicates the address of the starting block of the file and the total
length of the file in terms of blocks. This is the simplest method of allocation. Performance is
good because the entire file can be read from the disk in a single operation.

The difficulty with a contiguous allocation is finding space for a new file. If the file to be
created is n blocks long, then the OS must search for n free contiguous blocks. The most
common approaches used to select a free hole among the set of available holes are First-fit,
best-fit, and worst fit. Both first-it and best-fit are a better option as compare to worst fit in

terms of both criteria which is time and storage utilization. Neither first-fit nor best fit is best
in terms of storage utilization, but first-fit is generally faster.

These algorithms also have a disadvantage which is external fragmentation. As files are
allocated and also deleted, the free disk space is broken into tiny chunks. External
fragmentation exists when enough total disk space exists to satisfy a request, but this space
not contiguous; storage is fragmented into a larger number of small holes.

Another problem with a contiguous allocation is determining how much disk space is
needed for a file. At the file creation time, how much space it requires must be known and
allocated. One question arises here, how does the creator know about the size of the file
which is to be created? In some cases, this determination may be fairly simple, but generally,
the estimation of the size of an output file is a little difficult. A directory containing the file
information is maintained. Starting block address and a total number of blocks in a file are
read from this directory to read/write a file.

192
 Directory

File Start Length

Tushar
Tushar 3 5
1 2 3 4 5 6
Saatvik 25 3

7 8 9 10 11
Noor 20 4
12
Sheetal 17 2
Sheetal Rohit 32 4
13 14 15 16 17 18

Noor
19 20 21 22 23 24

Saatvik USED
25 26 27 28 29 30

31 32 33 34 35 36

FREE

Contiguous File Allocation

Characteristics

1. It supports flexible size portions.

2. It requires Pre-allocation.

3. It needs only a single entry of a file.

4. Allocation frequency is only once.

193
Linked List Allocation method

The problems in contiguous allocation can be traced directly to the requirement that the
spaces be allocated contiguously and that the files that need these spaces are of different
sizes. These requirements can be avoided by using linked allocation.
Directory
File Start End
Sachu 17 27

1 2 3 4 5 6

7 8 9 10 11 12

13 14 15 16 17
Sachu 18

20 21 22 23 24

19

25 26 -1 28 29 30

31 32 33 34 35 36

Link List Allocation

In linked allocation, every file is fragmented into Blocks and makes a linked list of these
disk blocks. The directory stores a pointer to the first Block of the file. For example, a file of

194
7 blocks starts at block 17, they might be next block 6, then block 21, block 32, block 23,
block 34, and finally last block 27. Each block holds a pointer to the next block and the last

block contains a NULL pointer which shows that end of the link list. The value -1 may be
used for NULL to differentiate it from block 0.

In the linked allocation method, every directory entry contains a pointer to the file‘s first
disk block. Initially, the pointer is initialized with a null value which shows that it is an empty
file. A write to a file removes the first free block from the free block list and writes to that
block. Then the address of this new block is associated with the last block of the file and then
linked to this block at the end of the file. To read a file, the pointers are used by a followed
pointer from block to block. The problem of the Contiguous File allocation method i.e.,
external fragmentation is resolved in the linked allocation. Any free block can be utilized to
fulfill a request. Notice also that there is no need to declare the size of a file when that file is
created. A file will still grow as possible as there are free blocks.

The linked allocation has some disadvantages. The foremost problem is that it is
inefficient to support direct access; it is effective only for sequential-access files. To locate

the ith block of a file, it should begin at the start of that file and follow the pointers till the ith
block is reached. Note that every access to a pointer needs a disk read.

Another severe problem is reliability. A bug in OS or disk hardware failure might result
in the pointer being lost and damaged. The effect of which could be picking up a wrong
pointer and linking it to a free block or into another file.

Characteristics

1. For a File there is one entry in the File allocation table size.

2. It‘s supports fixed size portions.

3. In this allocation Pre-allocation is feasible.

4. Allocation occurrence is Low to High.

Indexed Allocation Method

The problem of both contiguous and linked allocation is resolved in the indexed allocation
method. In this allocation, one block is used as an Index block in which all the pointers bring
together into one location.

195
 Of course, the index block will occupy some space and thus could be considered as an
overhead of the method. In indexed allocation, every file has its induvial index block, which

is nothing an array of disk sectors of addresses. The jth sector of the file is pointed by the jth
entry in the index block. The directory keeps the address of the index block of a file only. To

read the jth sector of the file, from the index block, the pointer in the jth entry is read to locate
the desires sector. Indexed allocation supports direct access, it resolves the external
fragmentation. Any free block anyplace on the disk might satisfy a request for extra space.

196
 Directory

File Index Block

Bob 24

1 2 3 4 5

7 8 9 10 B1
ob1

12

13 14 17
15 16 17 18
6

19 34
20 21 22 23 24
32

21
25 26 27 28 29 30

Inde
31 32 33 34 35 x
36
Blo
ck

Indexed Allocation Method

Characteristics

1. It supports both sequential and direct access.

2. There is no external fragmentation.

3. It does suffer wasted space.

4. Manage the Pointer is overhead.

197
 5. It is faster than the other two approaches.

Free-Space Management

Since there is only a restricted amount of disk space, it is necessary to utilize the space from
deleted files for new files. To keep track of free disk space, a free-space list is maintained by
the system. This list keeps records of all disk blocks which are free. Whenever a file is
created, the free space list has to be searched for the required amount of space and allocate
space to a newly created file. After that from the free-space list that space is removed.
Whenever a file is deleted then the free space list is updated by deleted file space added to the
free-space list.

Bitmaps or Bit – Vector

In the bitmap method a table, called bit map is maintained for keeping track of the
information about which part of memory is allocated to which process and which portion of
memory is free. The bitmap is dynamic. This means the table is updated each time when a
new process is allocated memory space, an existing process is swapped out of the main
memory, a process is swapped in the main memory, or a process is completed and released
from memory.

In the bitmap method, the memory is divided into certain equal size units or blocks.
While allocating memory space to a program an OS can only allocate a complete unit of
memory space and not a part of the unit. There is a cell in the bitmap for each unit of
memory. If a unit is allocated to any process the value of the cell corresponding to the unit is

198
 1 2 3 4 5 6

7 8 9 10 11 12

13 14 15 16 17 18

19 20 21 22 23 24

25 26 27 28 29 30

31 32 33 34 35 36

set to 1 otherwise if the memory unit is free, the value is set to 0.

Figure Bit Vector

1, 2, 8, 9, 10, 11, 12,13,14,15,16,19,24,28,29,30,31,36 are flocks free so bit map of this

example is 0,0,1,1,1,1,1,0,0,0,0,0,0,0,0,0,1,1,0,1,1,1,1,0,1,1,1,0,0,0,0,1,1,1,1,0

There is a cell in the bit map corresponding to every memory unit. The value of the 3
cell in the bitmap is 1 because its corresponding memory unit is in use by the process. Value
of the first and second cell is 0, because it‘s free.

An advantage of the bitmap method is that it is very simple to implement. On the other
hand, the disadvantage of this method is it is very slow. OS needs much time to search for a
hole, when allocating memory space to a new process or an existing process swapped in.

Linked List

In this method, a pool or chain of nodes is created to keep track of memory that is free
and maintain a linked list i.e. the first block contains the address of the next free block and so
199
on. The address of the very first free block is stored in a special node called as START node.
The last block of the linked list contains NULL, which implies that there is no other free
block.

Free Space List Header

1 2 3 4 5 6

7 8 9 10 11 12

13 14 15 16 17 18

19 20 21 22 23 24

28 29 30

34 35 36

25 26 27

31 32 33

Link Free Space List on Disk

For accessing the information in the linked list, it is loaded into physical memory.
Whenever any allocation is to be done, blocks are removed from the head of the list and
allocated. It occupies a large space. Traversing is also time-consuming. The following list
shows a free block list.

Grouping

A modification of the free-list approach, in the first free block, keeps the addresses of n free
200
blocks. There are only the first n-1 blocks that are free. The last one is the disk address of
another block containing addresses of another n free block. The importance of this
implementation is that a block contains the addresses of a larger number of free blocks so it
can be found quickly if we need the number of free blocks.

Example: Block 1 Contains the address of Block 2. Block 2 contains the address of the first
block of the next group block i.e., 8, Block 8 stores the address of 9 to 16 and the last block is
16 so it contains the address of the first block of the next group 19. As 19 is a single free
block so, it contains the next group address i.e. 28. Block 28 keeps the address of 29 and 30.
Block 30 contains the address of 31 and again Block 31 is the only single block in Grouping
so it contains the address of 36. 36 is the last block os it stores nothing

Counting

Another method is to take benefit of the fact that, generally, many contiguous blocks may be
allocated or freed contiguously, mainly when we used the contiguous allocation method.
Thus, rather than keeping the entire list of free disk addresses, we noted the address of the
first free block along with it the number n of free contiguous blocks which follow the first
block. Each entry in the free-space list then contains the disk address along with the count.
Although each entry needs more space as compared to a simple disk address, the overall list
will be shorter, as long as the count is generally greater than 1.

First Free Block Count

1 2
8 9
19 1
28 3
31 1
36 1

Points to Remember

The file system must identify and locate the selected file before operation on a file.
The file control block contains information about the file.
A disk can be divided into multiple partitions.
The file can be accessed by
201
— Sequential Access

— Direct Access

— Indexed Access

Three major methods of allocating disk space are widely used which are
Contiguous Allocation Method
Linked Allocation Method

— Indexed Allocation Method

The directory can be implemented in two ways:

— Linear List

— Hash Table

Free space management techniques are:

— Bit Vector

— Linked List

— Grouping

— Counting

PRACTICE EXERCISES

1. What is a File system?

2. Write a note on file attributes and naming?

3. What is File Organization?

4. Discuss the various methods of file allocation?

5. What are the disadvantages of the contiguous file allocation method?

6. What is the difference between linked allocation and indexed allocation methods?

7. What is a directory?

8. What are the different levels of the directory structure?

202
9. Discuss the various file systems?

10. Discuss the concept of free space management?

11. What are the various methods of Free Space Management?

203
 M.Sc. (Computer Science)
OPERATING SYSTEM

UNIT V1: INTRODUCTION TO LINUX

STRUCTURE
Objectives
Introduction to Linux
Linux Distributions
Characteristics of Linux Operating System
History of LINUX Operating System
Architecture of Linux
Windows Vs Linux
Unix Vs Linux
Types of Files
Linux Directory Structure
Parent, Subdirectory, Home directory
Naming Rules for Directory and Files
Practice Questions

OBJECTIVES

To Understand Linux's shell, Kernel, and file system of Linux.

Introduction to different types of directories: Parent, Subdirectory, Home directory;
rules to name a directory, Important directories in Linux File System

INTRODUCTION TO LINUX

Linux which is pronounced as Lin-nucks is an operating system similar to UNIX. Some

people also called it as descended form of UNIX. Linux is first released in 1991 by Linus
Torvalds at the University of Helsinki. Since then it has gain huge popularity among the
programmers. The Linux is an open-source operating system. It means it comes with the
source code, so that one can change and customize the operating system according to the
requirements. By the term open source it is meant that

The independence to run the program in operating system for any function.

The freedom to study any program that how it works and to alter it to make it work
according to your need.

The liberty to redistribute copies of the program to facilitate your neighbours.

204
 The free will to share copies of your modified versions to others for use.

Initially, Linux was created for personal computers and gradually it was used in other
machines like servers, mainframe computers, supercomputers, etc. Nowadays, Linux is also
used in embedded systems like routers, automation controls, televisions, digital video
recorders, video game consoles, smart watches, etc. The biggest success of Linux is Android
(operating system) it is based on the Linux kernel that is running on smart phones and tablets.
Due to android Linux has the largest installed base of all general-purpose operating systems.
Linux is generally packaged in a Linux distribution.

LINUX DISTRIBUTIONS

Linux has a numeral of dissimilar versions to ensemble any kind of user. Roughly all the
editions of the Linux can be downloaded free of charge. These editions are labelled
allocations (or, in the small form, ―distros‖). There are additional 600 distributions of Linux.
Some of the general distributions of the Linux are:

LINUX MINT

MY LINUX

MANJARO

DEBIAN

UBUNTU

DEEPIN

ANTERGOS

ARCH LINUX

GENTOO LINUX

205
 KALI LINUX

SOLUS

FEDORA

ELEMENTARY OS

OPENSUSE

Every distribution has a dissimilar take on the desktop. A small amount of the Server
distributions of the Linux comprises:

Red Hat Enterprise Linux

Ubuntu Server

Centos

SUSE Enterprise Linux

A few of the above server distributions are at no cost (such as Ubuntu Server and CentOS)
and a little have an associated cost (such as Red Hat Enterprise Linux and SUSE Enterprise
Linux). Those with an associated price also comprise support.
CHARACTERISTICS OF LINUX OPERATING SYSTEM
Following are a few of the significant characteristics of Linux Operating System.

Portable – Portability means software be able to work on dissimilar types of

hardware in similar way. Linux kernel and application programs sustain their setting
up on any kind of hardware stage.

Open Source – Linux source code is liberally accessible and it is community based
growth project. Multiple teams carry out task in collaboration to augment the potential
of Linux operating system and it is constantly evolving.

Multi-User – Linux is a multiuser scheme means numerous users can access network
206
 resources like memory/ ram/ application programs at similar time.

Multiprogramming – Linux is a multiprogramming arrangement means numerous

functions can run at similar time.

Hierarchical File System – Linux offers a standard file arrangement in which

organization files/ user files are ordered.

Shell – Linux provides a particular interpreter program that can be employed to carry
out commands of the operating system. It can be employed to do a variety of kinds of
processes, call application programs etc.

Security – Linux offers user security by means of authentication characteristics like

password protection/ controlled admittance to particular files/ encryption of
information.

Linux is fast, free and easy to use, power laptops and servers around the world. Linux
has numerous more characteristics to amaze its user such as:

Live CD/USB: Approximately all Linux distributions have Live CD/USB

characteristic by which user can run/try the OS even with no mounting it on the
scheme.

Graphical user interface (X Window System): People think that Linux is a

command line OS, wherever its precise also but not essentially, Linux have packages
that can be mounted to construct the whole OS graphics based as Windows.

Support’s most national or customized keyboards: Linux is engaged worldwide and

consequently obtainable in many languages, and supports on the entire of their custom
national keyboards.

Application Support: Linux has its own software ordnance through which users can
download and set up thousands of desires just by concerning a command in Linux
Terminal or Shell. Linux can3 besides run Windows demand if essential.

HISTORY OF LINUX OPERATING SYSTEM

The History of Linux commenced in 1991 with the beginning of a personal project by a
Finland student Linus Torvalds to create a new open operating scheme kernel. From then,
the resulting Linux kernel has been perceptible by fixed expansion during history.
207
 In the year 1991, Linux was started by a Finland student Linus Torvalds.
Hewlett Packard UNIX(HP-UX) 8.0 was enlightened.
In the year 1992, Hewlett Packard 9.0 was liberated.
In the year 1993, NetBSD 0.8 and FreeBSD 1.0 liberated.
In the year 1994, Red Hat Linux was initiated. Caldera was originated by Bryan
Sparks and Ransom Love and NetBSD1.0 liberated.
In the year 1995, FreeBSD 2.0 and HP UX 10.0 were liberated.
In the year 1996, K Desktop Environment was initiated by Matthias Ettrich.
In the year 1997, HP-UX 11.0 was liberated.
In the year 1998, the fifth invention of SGI Unix i.e IRIX 6.5, Sun Solaris 7
operating system, and Free BSD 3.0 was liberated.
In the year 2000, the conformity of Caldera Systems with the SCO server software
division and the professional services division was declared.
In the year 2001, Linus Torvalds liberated the Linux 2.4 edition source code.

In the year 2001, Microsoft ordered a trademark suit besides Lindows.com

In the year 2004, Lindows name was altered to Linspire.
In the year 2004, the first liberate of Ubuntu was discharged.
In the year 2005, The project, open SUSE begin a complimentary allocation from
Novell‘s community.
In the year 2006, Oracle liberated its own allocation of Red Hat.
Dell initiated allocated laptops with Ubuntu pre-installed in it, in the year 2007.
The Linux kernel 3.0 version was liberated in the year 2011.
Google Linux-based Android declared 75% of the smart phone market share, in
form of the number of phones dispatched, In the year 2013
Ubuntu asserted 22,000,000 users in the year 2014.

ARCHITECTURE OF LINUX

Linux architecture has the subsequent components:

208
 Figure 6.1: Architecture of Linux Operating System

I. Kernel: Kernel is the heart of the Linux supported operating system. The kernel is the
mainly significant part or mind of every operating system as it handles the conversation
among a machine‘s hardware and its software. In easy words, the kernel acts as a spirit
of the network and handles the memory, peripheral apparatus, and CPU. The kernel
places at the ―lowest‖ point of the OS. This creates the method seem as if it is the only
procedure running on the device. The kernel too well is accountable for avoiding and
justifying clashes among diverse processes. Particular kernel types are as

 Monolithic Kernel,

 Hybrid Kernels,

 Micro Kernels and

 Kernels from Exo.

A few of the key design ethics executed by Linus kernel are:

 In Linux, the whole thing is a file philosophy

 Multi-users ability
209
  Multitasking capability
 Portability among GPU architecture
 Modularity
 Security
 Configurability
II. Shell: It is an interface to the kernel which covers the difficulty of the kernel‘s tasks
from the client. It takes instructions from the user and carry out the kernel‘s functions.
Linux include two kinds of command keys: text mode comparable to those originated
in mainly UNIX systems (example the bourne shell, the bourne over shell, the c shell,
the turbo C shell and the korn shell) and graphical user interface (GUI) such as the
KDE (K Desktop Environment) and GNOME (GNU Network Object Model
Environment). Terminal is the purpose that brings it all collectively, in the sense that
it offers a visual illustration of the shell for the client to penetrate commands. In other
words, in a GUI (graphical user interface), where requests and other descriptions are
visually symbolized by images that the client can manoeuvre by ticking on them with
a cursor, a terminal request release a window where the client can type in instructions
for the shell to infer into binary communication for the kernel.

III. System Library: It is the exceptional kind of tasks that are used to execute the
functionality of the operating system. These pieces carry out the background facility
of the Linux operating system similar to scheduling, printing, sound, etc that each
start-up during boot of the operating scheme, or after you register into your system.

IV. Hardware Layer: This layer comprises of all the peripheral apparatus like RAM/
HDD/ CPU etc. Some of the substantial hardware that is incorporated in any tool
comprises of mouse, keyboard, the graphics chipset, display, and, but you do have
one, your system interface card

V. System Utility: It offers the functionalities of an operating system to the client. A

Linux-based scheme will typically appear with such a set of established Unix-like
utility services; these were all normally basic tasks used in day-to-day operating
system procedures, and also precise devices and requests. This is frequently software
that was available and liberated under the open-source license by the GNU Project,
consequently the software can be explicitly downloaded, restructured, and reallocated
to all.

210
Advantages of Linux

The main benefit of Linux, is it is an open-source operating system. This means the source
code is simply obtainable for everybody and you are authorized to give, alter and allocate the
code to anybody without any consent.

 In terms of security, Linux is further secure than any other operating scheme. It does
not signify that Linux is 100 percent safe it has a little malware for it but is fewer
susceptible than any other operating scheme. So, it does not necessitate any anti-virus
software.
 The software renewed in Linux are simple and recurrent.
 A variety of Linux distributions are accessible so that you be able to utilize them
according to your desires or according to your flavour.
 Linux is liberally available to employ on the internet.
 It has huge community support.
 It offers high constancy. It seldom slows down or freezes and there is no necessitating
rebooting it after a small time.
 It preserves the privacy of the client.
 The presentation of the Linux scheme is much elevated than other operating scheme.
It permits a large quantity of people to employee at the similar time and it grips them
resourcefully.
 It is network pleasant.
 The flexibility of Linux is elevated. There is no necessitate to establish a complete
Linux suit; you are authorized to establish only required apparatus.
 Linux is companionable with a huge number of file set-ups.
 It is quick and simple to fit from the web. It can also mount on any hardware even on
your previous computer structure.
 It carry out all tasks correctly even if it has partial liberty on the hard disk.
Disadvantages of Linux

 It is not extremely user-friendly. So, it might be puzzling for beginners.

 It has tiny marginal hardware drivers as contrast to windows.

211
 Widows Vs Linux

Few Differences between Linux and Windows

S.NO Linux Windows

1. Linux is a open source operating scheme. While windows are the not the open
source operating scheme.

2. Linux is at no cost. While it is expensive.

S.NO Linux Windows

3. It‘s file name case-sensitive. While it‘s file name is case-insensitive.

4. In linux, monolithic kernel is employed. While in this, micro kernel is employed.

5. Linux is more competent in comparison While windows are fewer competent.

of windows.

6. There is forward slash is employed for While there is back slash is employed for
separating the directories. isolating the directories.

7. Linux offers more protection than While it offers less safety than linux.
windows.

8. Linux is extensively employed in hacking While windows does not offers much
point based systems. competence in hacking.

UNIX VS LINUX
Now, we will observe what is the distinction between linux and unix:

Key
Linux Unix
Differences

212
 Linux is liberally dispersed, downloaded
Diverse flavors of Unix have
during magazines, Books, website, etc.
Cost dissimilar pricing depending
There are waged versions also
upon the kind of vendor.
accessible for Linux.
Unix systems have dissimilar
Linux is Open Source, and thousands of
versions. These editions are
Development programmers work together online and
primarily created by AT&T as
supply to its expansion.
well as other business vendors.
Everybody. From home customer to The UNIX can be employed in
User developers and computer aficionado internet servers, workstations,
alike. and PCs.
BASH is the Linux evasion shell. It Initially prepared to job in
Text made
presents support for numerous command Bourne Shell. Though, it is now

Key
Linux Unix
Differences
interface interpreters. well-suited with numerous others
software.
Linux offers two GUIs,viz., KDE and
Gnome. Although there are numerous General Desktop atmosphere and
GUI
substitutes such as Mate, LXDE, Xfce, also has Gnome.
etc.
Linux has had concerning 60-100 virus There are among 80 to 120
Viruses listed today which are at present not viruses informed till date in
scattering. Unix.
Threat finding and way out is extremely
quick since Linux is chiefly community
Unix client necessitate longer
Threat driven. So, if any Linux clients post any
wait instance, to acquire the
detection type of threat, a team of competent
appropriate bug fixing patch.
developers commence working to
overcome this risk.

213
 Originally urbanized for Intel's x86
hardware processors. It is obtainable for It is obtainable on PA-RISC and
Architectures
over twenty dissimilar kinds of CPU Itanium machines.
which also comprises an ARM.
Linux OS can be fitted on a variety of The UNIX operating scheme is
Usage kinds of plans like mobile, tablet employed for internet servers,
computers. workstations & PCs.
Feta ZFS - next age group file
Best feature Kernel renewed with no reboot system DTrace - dynamic Kernel
Tracing
Dissimilar editions of Linux are Dissimilar editions of Unix are
Versions
OpenSuse ,Redhat, Ubuntu, etc. BSD ,HP-UX, AIS, etc.
The File arrangement maintained by file The File arrangement sustained
Supported file
kind like devpts, xfs, nfs, cramfsm ext 1 by file kinds are vxfs, zfs, hfx,
type
to 4, ufs, NTFS. GPS, xfs.
Linux is moveable and is booted as of a
Portability Unix is not moveable
USB Stick
Key
Linux Unix
Differences
The source is obtainable to the common The source code is not accessible
Source Code
public to everyone.

TYPES OF FILES
In Linux and UNIX, the whole thing is a file. Directories are records; files are files, and tools
like keyboard, Printer, mouse, etc. are files.

Let's look into the File types in more detail.

General Files

General Files also called as Ordinary files. They can include image, video, program or simply
text. They can be in ASCII or a Binary format. These are the most commonly used files by
Linux Users.
214
 Directory Files

These files are a warehouse for other file types. You can have a directory file within a
directory (sub-directory).You can take them as 'Folders' found in Windows operating system.

Device Files

In MS Windows, devices like Printers, CD-ROM, and hard drives are represented as drive
letters like G: or H:. In Linux, these are represented as files. For example, if the first SATA
hard drive had three primary partitions, they would be named and numbered as /dev/sda1,
/dev/sda2 and /dev/sda3.

Note: All device files reside in the directory /dev/

All the above file types (including devices) have permissions, which allow a user to read, edit
or execute (run) them. This is a powerful Linux/Unix feature. Access restrictions can be
applied for different kinds of users, by changing permissions.

Linux Directory Structure

In Linux, everything is a file. A directory is nothing but a special file that contains details
about all the files and subdirectories housed within it. The Linux contains following
directories for different purpose:

i. Root Directory ( / )
The root directory also represented as ― / ― is the topmost level of the system drive.
Every other directory in your Linux system is located under the root directory. You
may imagine the / directory similar to the C:\ directory or drive on your Windows
system. This is not at all true but you may understand it as imaginary vision because

Linux doesn‘t have drive letters. It is commonly used to represent a filesystem. A

filesystem is the hierarchy of directories that is used to organize directories and files
on a computer.
ii. Essential User binaries ( /bin)
This folder contains the essential user programs which are also called as binaries
whichare present when the system is mounted in single-user mode. It also contain the

215
 files for common commands likes ls, cat, cd, cp etc. The '/bin' directory also contains
executable files
iii. Static Boot Files (/boot)
It contains the files which are required to boot the system such Linux kernel files,
GRUB configuration files.
iv. Historical mount point for CD-ROMs (/cdrom)
/mnt/cdrom and /cdrom are the mount point directories for the CD-ROM drive. The
/cdrom directory isn‘t part of the Filesystem Hierarchy Standard (FHS), but you‘ll
still find it on Ubuntu and other operating systems.
v. Device Files (/dev)
In Linux, the /dev directory contains device and other special files. To see a list of
these items, run the following command in a terminal session.
ls –l /dev
These include terminal devices, usb, or any device attached to the system.
vi. Configuration Files (/etc)
The /etc/ directory contains system-wide configuration files — user-specific
configuration files are located in each user‘s home directory.
vii. Home Folder ( /home )
The home directory for each user takes the form /home/username (where username is
the name of the user account). For example, if your user name is john, the system will
have a home folder located at /home/john
viii. Essential Shared Libraries ( /lib )
The lib folder is a library files directory which contains libraries needed by the
essential binaries to be used by the system. These are supportive files which are used
by an application or a process for their proper execution. The commands in /bin or
/sbin dynamic library files are located just in this directory.
ix. Recovered Files ( /lost+found )
Whenever a there is crashed in the file system, the system boots again and check is
performed. Any file which got corrupted by crash will be found in lost+found
folder.Every disk partition has a lost+found directory.
x. Removable Media ( /media )
This Directory is used for mounting files systems on removable media like Zip drives,
floppy drives, CD-ROM drives.

216
 xi. Temporary Mount Points ( /mnt )
This directory is used for temporarily mounted filesystem. For example, if you‘re
mounting a FAT partition for some operations, it might be mounted at /mnt/fat.
xii. Optional Packages ( /opt )

It contains subdirectories for optional software packages which are written as a result
of copy/install operations.
xiii. Kernel and Process Files ( /proc )
It is a special directory for virtual filesystem. Instead of standard files it contains
special files that represent system and process information.
xiv. Root Home Directory ( /root )
This directory is distinct from / directory which is used for the system root directory.
It is the home directory for root user. Instead of using the /home/root, it uses the
location at /root.
xv. Application State Files ( /run )
This folder contains data which describe the system since it has been booted. It gives
applications a standard place to store temporary/transient files containing data related
to sockets and process Ids.
xvi. System Administration Binaries ( /sbin )
It contains files related to administrative commands. This directory includes essential
binaries that might be needed to run by the root user for system administrator of the
system.
xvii. SELinux Virtual File System ( /selinux )
Ubuntu doesn‘t use SELinux. The /selinux directory contains special files used by
SELinux for security (example Fedora and Red Hat). It is similar to /proc folder.
xviii. Service Data ( /srv )
This holds the data for system services like Apache HTTP, FTP, etc.
xix. Temporary Files ( / tmp )
When the system or user executes the applications, the temporary files are stored in
this folder. These temporary files are usually deleted automatically when the system
gets started again.
xx. User Binaries and Read Only Data ( /usr )
This directory is meant to contain the applications and files which belong to users,
dissimilar to applications and files used by the system.

217
 xxi. Variable data Files ( /var )
This folder holds numerous system files which may be related to log, mail directories,
print spool, etc. that may change over time in numbers and size.

Parent, Subdirectory, Home Directory

Root Directory ( / ) holds the top position as the main directory. All the directories mentioned
in above section from sr. No. 2 to sr. No. 21 are the subdirectory of the root directory. /Home
and /root are the home directories for the normal user and root user respectively.

Naming Rules for Directory and Files

1. The file names in the Linux are case sensitive. For example a filenamesarea.txt,
Area.txt and AREA.txt all are considered as different files.

2. The filename can be created either by using uppercase letters, lowercase letters,
numeric numbers, ―_‖ (underscore) and ―.‖ (dot) symbols.

3. In Linux, other special characters such as blank space can also be used, but the rules
for using other symbols are hard and it is heal their not to use them.

4. As mentioned in above point filenames may contain any character but one cannot use
slash sign―/‖ (root directory), because it is reserved character and can only be used as
the separator between files and directories while mentioning the pathname. One
cannot use the null character.

5. .(dot) sign should not be used in the filename. Some time dot sign increases the
readability of filenames but this may add confusion in understanding the extension of
files. Dot sign should be used to identify the extension of the file. For example:

o .tar.gz = zipped or Compressed files

o .sh = used to identify the shell file

6. Older versions of the Unix operating system uses filenames upto 14 characters long.
But in today‘s era latest version of Linux and UNIX can use filename length upto 255
characters (255 bytes).

7. A filename should be unique in a particular directory. For example, in directory

/home/john, one cannot give two file name as try.txt and try.txt. However, different
directory can have same filenames inside them. For example, you can create try.txt in
218
 /tmp directory and try.txt inside /home/john directory.

PRACTICE QUESTIONS

Q1.What are the basic components of Linux?

Q2.Write down the name of some Linux variants.
Q3.Which popular office suite is available free for both Microsoft and Linux?
Q4.Suppose your company is recently switched from Microsoft to Linux and you have some
MS Word document to save and work in Linux, what will you do?
Q5.What is the difference between Linux and Unix?
Q6.What is Linux Kernel? Is it legal to edit Linux Kernel?
Q7.What is the difference between BASH and DOS?
Q8.What are the process states in Linux?
Q9.What is the advantage of open source?

219
 M.Sc. (Computer Science)
OPERATING SYSTEM

UNIT VII : LINUX COMMANDS

STRUCTURE

Objective
Linux Commands
File Permissions in Linux
Change in Ownership and the Associated Group
Advice before using Commands
Summary
Practice Exercise

OBJECTIVE

Understanding Linux commands

OVERVIEW OF LINUX COMMANDS

The command line is one of the most powerful features of Linux. There are a countless
number of commands in Linux. In this chapter, we will introduce few of most commonly and
regularly used Linux commands for easy learning.

cal / ncal command

The cal command is used to print the calendar for a specified month and/or year on
the monitor i.e. standard output device.

$cal: This will print the calendar for the current

$cal 2020: This command will print the twelve month calendar for the year 2020.

$cal 7 2020: This command will print the calendar July 2020.

mkdir command

This command is used to make the subdirectory under the current directory

220
 $mkdir student: This command will create the directory student under the
current directory

cd command

The cd command is used to change the current directory to the specified directory

$cd student: This command will change the current directory to student directory.

mv command

This command is used move the files from one folder to another folder.

$mv /home/john/try.txt /home/merry/. This command will move the try.txt file
from /home/john directory to home/merry folder

$mv /home/merry/*.* /home/john/. This command will copy all the files in
/home/merry folder to /home/john folder.

cp command

This command is used to copy the files from one folder to another folder.

$cp /home/john/try.txt /home/merry/. This command will copy the try.txt file
from /home/john directory to home/merry folder

$cp /home/merry/*.cpp /home/john/. This command will copy all the

files with extension .cpp from /home/merry folder to /home/john folder.

date command

The date command displays the current day, date, time, and year.

$ date: It will display current system date like ―Mon May 24 17:14:57 IST 2021‖
echo command
The echo command prints whatever you will write except quote. This command is
221
 used to insert the text on a file as shown below.

$echo "Welcome to The Linux Class": This command will print the line
Welcome to The Linux Class.

rm command:

One can use ―rm‖ command to delete any file or directory. To delete directories
recursively we can use rm command along with -r , -R

$rm –r fname: This command will remove the folder with name fname.

$rm rea*.* : This command will remove all files starting with name prefix
rea in current folder

rmdir command:

one can use rmdir to delete a directory. This command can only be used to delete an
empty directory.

$rmdir dir_name: This command will remove the folder with name dir_name.

cat command:

cat is used to display the contents of the file whose name is mntioned along with the
command. Example to see the text inside the file area.txt we may write the command
as follows.

$cat area.txt: It will display the contents inside the filename area.txt

It is also used to merge the contents of the two files. The output is printed on the
standard output device.

$cat filename1 filename2: It will display the contents of both the files together
after merging.

222
pwd command

pwd command is used to display the name of present working directory. With the help
of pwd command we are able to know about the directory in which we are working
with. It gives us the absolute path, which means the path that starts from the root.

$pwd : It will the complete path of current directory. Example if you are
working under user john under home directory then we will get output as /home/john.

who command

who command used to find out last system boot time, the system‟s current run level,
List of the users who logged in the system and more. The different variations of who
command are as follows:

Description Example
The who command displays the $who
following information for each user
currently logged in to the system if no
option is provided
To display host name and user $who -m –H
associated with standard input such as
keyboard
To show all active processes which $who -p –H
are spawned by INIT process
To show status of the users message $who -T –H
as +, – or ?
To show list of users logged in to $who –u
system
To show time of the system when it $who -b –H
booted last time
To show details of all dead processes $who -d –H

To show system login process details $who -l -H

223
 To count number of users logged on $who -q –H
to system
To display current run level of the $who -r
system

To display all details of current $who –a

logged in user
To display system‟s username $whoami

To display list of users and their $w

activities
To display user identification $ id
information

pwd command
pwd command is used to display the name of present working directory. With the help
of pwd command we are able to know about the directory in which we are working
with. It gives us the absolute path, which means the path that starts from the root.

$pwd : It will the complete path of current directory. Example if you are
working under user john under home directory then we will get output as /home/john.

ls command
This is called list command. It is used to display the files stored in a particular
directory.

ls: To view the brief, multi-column list of the files in the current directory.

ls –a: To also see files starting with symbol "dot" (configuration files that begin with
a period, such as .login).

ls –la: To see the permissions, owners and size of the files along with their names.

ls -la | less: If the listing is enough long that it is not fit into one screen view then
reading becomes difficult. This command combines ls with the less utility for the
224
 same.

bc Command

This command is used for arbitrary precision CLI calculator which can be used like
this: $ echo 30.07 + 16.00 | bc

more Command
This command enables us to view the large text data one screen display at a time.
gzip Command
gzip command is used to compress a file, replaces it with a file with .gz extension.
The example of usage of gzip command is as shown below:

$ gzip passwds.txt
$ cat filename1 filename2 | gzip > try.gz
tar Command
This powerful utility command is a used for archiving files in Linux as follows.
$ tar -czf demo.tar.gz
FILE PERMISSIONS IN LINUX
Linux is a multi-user operating system which means it can be accessed by many users
simultaneously. The multi-user concern of Linux imposes security concerns as an unsolicited
or malign user can corrupt, change or remove crucial data. Linux categorizes
authorization into 2 levels.
1. Ownership
2. Permission
Ownership in Linux
Every file and directory in Linux operating system can be assigned 3 different types of
owner, given below.

User
The person who creates the file is by default the owner of the file. Such user is the owner of
the file. Therefore, a user is also sometimes called an owner.

Group
A group or user- group consists of multiple users. All the users which belong to a same group
will have the same access permissions to the file. For example there is a project where the
225
number of people/users needs access to a file. You could add all users to a single group and
assign the permissions to the group instead of assigning the permissions to each user
individually. No one other than the group member can read or modify the files.

Other
In this category, any general user who can access to a file is included in other category. This
person has neither produced the file, nor do he / she belong to a particular usergroup who
could own the file. Practically, it means everybody else. Therefore, this category is also
called as the set permissions for the world.

It is matter of the question that how Linux distinguishes between the three category of the
users. One user 'A' cannot affect a file that contains the information of some other user 'B'. It
means that you do not want another colleague in your organization to access your data. This
is where the role of setting permissions comes in, and this also defines user behavior.

Let us understand the role of setting permissions in the Linux OS.

Permissions
Every file and directory in your Linux operating system has three types of permissions
defined for all the 3 types of owners.

 Read: The read permission gives the authority to read as well as open the file. This
permission on a directory gives you the capability to view its contents.

 Write: This permission gives a person the authority to modify the contents inside the
file. You can rename, remove and add files present in the directory. For example if
you don‟t have write permission on a folder but have the write permission on one of
its file then you will be able to change the contents of the file but you will not be able
to remove, move or rename the file placed inside the directory.

 Execute: Similar to Windows where files with ―.exe‖ extension are executable. But in
Linux OS, one cannot execute any file without having permission on it. If the execute
permission is not granted on file, then you will not be able to execute the file.
However you can see/modify the contents inside the file.

226
 Categorization of Permissions that can be assigned to different types of Users

Let us see view this with the help of commands:

Issue command ls - l on terminal gives you the filenames with file types and access
permissions.

ls – l
home$ ls –l
-rw-rw-r-- 1 home home – 2020-04-25 16:27 MyFilename

The string '-rw-rw-r--' in the output describes about the permissions given to the owner, user
group and the world.

Here, the first '-' represents that this line indicates information about file

If it could have been a directory then d would have been shown as below:
227
drwxr-xr-x 2 ubuntu ubuntu 80 Apr 5 08:24 Downloads

In the above output, the first character „d‟ represent the directory.

The characters shown in the above example are easy to remember.

- = have no permission
x = the permission to execute
w = the permission to write
r = the permission to read
Let us explore the meaning of the output as follows:

The first part in the output line is 'rw-'. This indicates that the owner 'home' can have two
permissions i.e. read and write the file but the owner is not having the permission to execute
that is why sign „-„ has been used.

However, many distributions of the Linux like Ubuntu, CentOs and Fedora etc. can add users
to the same group name which is the name of the user. Therefore a username 'john' may be
added to a group having name as 'john'.

The second part of the output is 'rw-' which indicates that group having named as „home‟
have the permission to read and write the file but don‟t have the permission to edit the file.

Third part is used to represent the permissions for the other person from the outside world. It
means there could be any other user. The string 'r--' indicates that other users have the

228
permission to only read the file but these persons may not be able to write and execute the
file.

Using „chmod‟ command to change the directory/file permissions

Permissions help you to restrict the others in performing operations on files/directories in

form of read, write and execute. This can be accomplished by changing permissions on
files/directories.

We can use the 'chmod' command which stands for 'change mode'. Using the command, we
can set permissions (read, write, execute) on a file/directory for the owner, group and the
world. The Syntax of command is as follows:

chmod permissions filename

There are 2 types of modes for using chmod command -

1. Absolute mode
2. Symbolic mode
Absolute Mode / Numeric Mode

In this technique/mode, the permissions on files are represented as three-digit octal number
instead as characters. The following table represents the different numbers for each type of
permissions.

Number Permission Type Symbol

0 No Permission ---

1 Execute --x

2 Write -w-

3 Execute + Write -wx

4 Read r--

5 Read + Execute r-x

229
 6 Read +Write rw-

7 Read + Write +Execute Rwx

Let us look at the examples of using chmod command.

Command “ls –l MyFilename” shows the current permissions on the MyFilename

If we issue the command “chmod 764 MyFilename” then we will see the following results

In the above-given terminal window, we have changed the permissions of the file 'sample to
'764'.

'764' absolute coding can be interoperated as follows:

 Owner can read, write and execute

 User group can read and write

 World can only read

This can be seen in the output as '-rwxrw-r-

In this way; you can change the permissions on file by assigning an absolute number.

230
Symbolic Mode

In the Absolute mode, you change permissions for all 3 owners. In the symbolic mode, you
can modify permissions of a specific owner. It makes use of mathematical symbols to modify
the file permissions.

Operator Action of applying the operator

+ Adds the particular permission to the file / directory

- Removes the particular permission from file / directory

= Sets the new permission by overriding the previous permissions.

The various types of owners of file and directories are given as follows:

User Denotations

U user/owner

G Group

O Other

A All

In symbolic mode we shall be setting permissions by making use of characters like rwx
unlike numbers in absolute mode eg 764. Let us apply example for symbolic mode using the
previous file permissions on MyFilename:

Command “ls –l MyFilename” shows the current permissions on the MyFilename

231
After applying the command “Chmod u = rwx MyFilename”, we shall get the following
result:

In above command ―u‖ stands for user. It means new permissions will be set for the user. The
command will assign the user permissions of read, write and execute to the MyFilename.

Let us apply another command as “chmod o+x MyFilename”. Then the permissions will be
changed as follows.

String “o+x” in above command will add the execute permission for the other user.

Let us again apply command as “chmod o-x MyFilename”. Then the execute permission
which was given just now will be removed. The output of “ls –l MyFilename” command
will look as follows after applying the command.

CHANGE IN OWNERSHIP AND THE ASSOCIATED GROUP

To change the ownership of particular file/directory, we may use the “chown” command as
follows:
chown new_user_name

If there is need to change the group along with the user for a file or directory, then we may
use the following command

chown new_user_name:new_group_name filename

Let us look at the output of “ls – l MyFilename” command

232
Now apply the ―chown john MyFilename” command and then apply “ls – l MyFilename”
command. It you are not logged in as root user then you may use “sudo chown john
MyFilename” command to act as super user. The system may ask you the password for the
super user in such case.

Let us change the user and group both to root by applying “sudo chown root:root
MyFilename” command. The output of the file may look as follows after applying the
command.

If you want to change only the group of the user then may also apply “chgrp” command.
Command ―chgrp” stands for change group.

Let us apply the command “sudo chgrp john MyFilename” to change the group from root
to john. Then we will get output as follows:

ADVICE BEFORE USING COMMANDS

The file stored in the /etc/group contains the name of all the groups defined in the
system at any moment of time.

We may use the command "groups" on terminal to find the name of all the groups
where we are a members.

We may use the command “newgrp” to work as a member a group other than your
default group

No file/directory can own to two groups simultaneously.

Linux do not have nested groups.

x- eXecuting a directory means Being allowed to "enter" a dir and gain possible
access to sub-directories.
233
SUMMARY

Linux being a multi-user system uses permissions and ownership for security.

There are three user types on a Linux system viz. User, Group and Other

Linux divides the file permissions into read, write and execute denoted by r,w, and x

The permissions on a file can be changed by 'chmod' command which can be further
divided into Absolute and Symbolic mode

The 'chown' command can change the ownership of a file/directory. Use the following
commands: chown user file or chown user:group file

The 'chgrp' command can change the group ownership chrgrp group filename

What does x - executing a directory mean? A: Being allowed to "enter" a dir and gain
possible access to sub-dirs.

PRACTICE QUESTIONS

Q1.What are the basic commands for user management?

Q2.Which command is used to uncompressed grip files?
Q3.What are the modes used in VI editor?
Q4.What are the file permissions in Linux?
Q5.Which are the Linux Directory Commands?
Q6.What are inode and process id?
Q7.Explain Process Management System Calls in Linux
Q8.Explain the redirection operator.
Q9.How to copy a file in Linux?

Q10.How to terminate a running process in Linux?

Q11.How to create a new file or modify an existing file in vi?

Q12.Explain Regular Expressions and Grep

234
 M.Sc. (Computer Science)
OPERATING SYSTEM

UNIT VIII: SHELL SCRIPTING

STRUCTURE

Objectives
Introduction to Shell
Shell Scripting
Shell Prompt
Variable Names
Positional Parameter / Command Line Arguments
Practice Questions

OBJECTIVE
To learn the usefulness of shell scripting
INTRODUCTION TO SHELL
• The shell is a user program or it is an environment provided for user interaction.
• It is a command language interpreter that executes commands read from the standard
input device such as keyboard or from a file.
• The shell gets started when you log in or open a console (terminal).
• Quick and dirty way to execute utilities.
• The shell is not part of system kernel, but uses the system kernel to execute programs,
create files etc.
Different shells are available in Linux which are as under:
• BASH ( Bourne-Again SHell ) - Most common shell in Linux. It's Open Source.
• CSH (C SHell) - The C shell's syntax and usage are very similar to the C
programming language.
• KSH (Korn SHell) - Created by David Korn at AT & T Bell Labs. The Korn Shell
also was the base for the POSIX Shell standard specifications.
• TCSH - It is an enhanced but completely compatible version of the Berkeley UNIX C
shell (CSH).
SHELL SCRIPTING
A shell script is just a normal Linux file which contains Linux and shell commands. The
simplest shell scripts simply group together commonly used sequences of commands.

235
 More complex scripts use the shell's programming syntax to perform more advanced tasks.
Shell scripts are not compiled but interpreted. This means that each time they are run a shell
is executed to read the file and run the commands it contains. Languages which are compiled,
such as C, will produce a compiled code which will run faster than an equivalent shell script.

Shell scripts usually begin with a #! and a shell name.

For example: #!/bin/sh
If they do not, the user's current shell will be used

SHELL PROMPT
There are various ways to get shell access:
• Terminal - Linux desktop provide a GUI based login system. Once logged in you can
gain access to a shell by running X Terminal (XTerm), Gnome Terminal (GTerm), or
KDE Terminal (KTerm) application.
• Connect via secure shell (SSH) - You will get a shell prompt as soon as you log in
into remote server or workstation.
• Use the console - A few Linux system also provides a text-based login system.
Generally you get a shell prompt as soon as you log in to the system.
You may type the following into the Terminal to find out the shells available in your system.

cat /etc/shells
Why write shell scripts?
– To avoid repetition: If you do a sequence of steps with standard Linux commands over and
over, why not do it all with just one command?
– To automate difficult tasks: Many commands have subtle and difficult options that you
don‘t want to figure out or remember every time.
– Creating your own power tools/utilities.
– Automating command input or entry.
– Customizing administrative tasks.
– Creating simple applications.
– Since scripts are well tested, the chances of errors are reduced while configuring services or
system administration tasks such as adding new users.

Advantages
236
 • Easy to use.
• Quick start, and interactive debugging.
• Time Saving.
• Sys Admin task automation.
• Shell scripts can execute without any additional effort on nearly any modern UNIX /
Linux / BSD / Mac OS X operating system as they are written an interpreted
language.

Disadvantages
• Compatibility problems between different platforms.
• Slow execution speed.
• A new process launched for almost every shell command executed.

VARIABLE NAMES

Variables names may comprise upper and lower case alphabetic characters, digits and
underscores. A user defined variable name cannot start with a digit. As with most things in
Unix variable names are case sensitive, e.g. FOO and foo are two distinct variables.

In Linux, there are two types of variable

1) System variables - Created and maintained by Linux itself. This type of variable defined
in CAPITAL LETTERS.
2) User defined variables (UDV) - Created and maintained by user. This type of variable
defined normally by using lower case LETTERS.
Some System variables
You can see system variables by giving command like $ set, Some of the important System
variables are

237
System Variable Meaning
BASH=/bin/bash Our shell name
BASH_VERSION=1.14.7(1) Our shell version name
COLUMNS=80 No. of columns for our screen
HOME=/home/vivek Our home directory
LINES=25 Number of columns for our screen
LOGNAME=students Our logging name
OSTYPE=Linux Our o/s type : -)
PATH=/usr/bin:/sbin:/bin:/usr/sbin Our path settings
PS1=[\u@\h \W]\$ Our prompt settings
PWD=/home/students/Common Our current working directory
SHELL=/bin/bash Our shell name
USERNAME=vivek User name who is currently login to this PC

Special Shell Variables

Some variables have special significance to the shell. For example, the PATH variable is used
by the shell to contain the list of directories to search for commands. Other variables have
special meaning to other Linux utilities, for example the TERM variable contains the current
terminal type.

bash$ variable1=23
Use echo command to display variable value.
bash$ echo variable1
variable1
bash$ echo $variable1
23
To display the program search path, type:
bash$ echo "$PATH"
To display your prompt setting, type:
bash$ echo "$PS1"
All variable names must be prefixed with $ symbol, and the entire construct should be
enclosed in quotes. Try the following example to display the value of a variable without using
$ prefix:
238
bash$ echo "HOME"
To display the value of a variable with echo $HOME:
bash$ echo "$HOME"
You must use $ followed by variable name to print a variable's contents.
The variable name may also be enclosed in braces:
bash$ echo "${HOME}"
This is useful when the variable name is followed by a character that could be part of a
variable name:
bash$ echo "${HOME} work"
To define User Defined Variable (UDV) use following syntax
Syntax: variablename=value

NOTE: Here 'value' is assigned to given 'variablename' and Value must be on right side =
sign For e.g.
$ no=10 # this is ok
$ 10=no # Error, NOT Ok, Value must be on right side of = sign.
To define variable called 'vech' having value Bus
$ vech=Bus
To define variable called n having value 10
$ n=10
How to Define variable x with value 10 and print it on screen
$ x=10
$ echo $x
Linux Shell Script Tutorial
How to Define variable xn with value Rani and print it on screen
$ xn=Rani
$ echo $xn
How to print sum of two numbers, let's say 6 and 3
$ echo 6 + 3
This will print 6 + 3, not the sum 9, To do sum or math operations in shell use expr, syntax is
as
follows Syntax: expr op1 operator op2

239
 Where, op1 and op2 are any Integer Number (Number without decimal point) and operator
can be
+ Addition
- Subtraction
/ Division
% Modular, to find remainder For e.g. 20 / 3 = 6 , to find remainder 20 % 3 = 2, (Remember
its
integer calculation)
\* Multiplication
$ expr 6 + 3
Now It will print sum as 9 , But
$ expr 6+3
will not work because space is required between number and operator (See Shell Arithmetic)
Q.4.How to define two variable x=20, y=5 and then to print division of x and y (i.e. x/y)
$x=20
$ y=5
$ expr x / y
Q.5.Modify above and store division of x and y to variable called z
$ x=20
$ y=5
$ z=`expr x / y`
$ echo $z
Note : For third statement, read Shell Arithmetic.

POSITIONAL PARAMETER / COMMAND LINE ARGUMENTS

Command line arguments are important part of writing scripts. Command line arguments
define the expected input into a shell script. For example, we may want to pass a file name or
folder name or some other type of argument to a shell script.

A positional parameter is a variable within a shell program; its value is set from
an argument specified on the command line that invokes the program. Positional
parameters are numbered and are referred to with a preceding ``$'': $1, $2, $3, and so on.
A shell program may reference up to nine positional parameters.

240
Consider the following bash command. Let the command name is mycommand. The
command line has three parameters: one, two, and three four.

$ mycommand one two ―three four‖

Variable name Value

$0 Mycommand
$1 One
$2 Two
$3 three four
$# 3
$@ one two three four
$* one two three four
$! This parameter represents the process number of the
background that was executed last.
$? This parameter represents exit status of the last command that
was executed. Here 0 represents success and 1 represents
failure.

$_ This parameter represents the command which is being

executed previously.
$- This parameter will print the current options flags where the set
command can be used to modify the options flags.

241
$ cat program.sh
echo "The File Name is: $0"
echo "The First argument is: $1"
echo "The Second argument is: $2"

$ sh program.sh ab cd
The File Name: program.sh
The First argument is: ab
The Second argument is: cd

If any arguments are supplied, they become the positional parameters when filename is
executed. Otherwise, the positional parameters remain unchanged.

if-then-else
Like most programming languages, shell script supports the if statement, with or without an
else. The general form is below:

if (condition)
simple commands

if (condition)
then
commands
…
fi

if (condition)
then

242
 … Else

…fi
commands

Commands

You can have nested if-else-fi

The condition used as the predicate can be any program or expression. The results areevaluated
with a 0 return being true and a non-0 return being false.

If ever there is the need for an empty if-block, the null command, a :, can be used in place ofa command to
keep the syntax legal.

The following is a nice, quick example of an if-else:

if [ "$LOGNAME"="guna" ]then
elsefi printf "%s is logged in" $LOGNAMEprintf "Intruder! Intruder!"

243
The elif construct
Shell scripting also has another construct that is very helpful in reducing deep nesting. It is unfamilar to
those of us who come from languages like C and Perl. It is the elif, the "else if". This probably made its way
into shell scripting because it drastically reduces the nesting that would otherwise result from the many
special cases that real-world situatins present --without functions to hide complexity (shell does have
functions, but not parameters -- and they are more frequently used by csh shell scripters than
traniditonalists).

if conditionthen
commandcommand
...
commandelif condition then
commandcommand
...
commandelif condition then
commandcommand
...
command
fi

The switch statement

Much like C, C++, or Java, shell has a case/switch case statement. The form is as follows:
case ―$variable‖ inpattern1)
command

command
...
command
;; # Two ;;'s serve as the breakPattern2)
commandcommand
...
command
;; # Two ;;'s serve as the breakPattern3)

244
commandcommand
...
command
;; # Two ;;'s serve as the breakesac
Here's a quick example:

#!/bin/sh case "$char"in

"+")
ans=`expr $1 + $3`
printf "%d %s %d = %d\n" $1 $2 $3 $ans
;;
"-")
ans=`expr $1 - $3`
printf "%d %s %d = %d\n" $1 $2 $3 $ans
;;
"\*")
ans=`expr "$1 * $3"`
printf "%d %s %d = %d\n" $1 $2 $3 $ans
;;
"/")
ans=`expr $1 / $3`
printf "%d %s %d = %d\n" $1 $2 $3 $ans
;;
# Notice this: the default case is a simple *
*)

esac
printf "Don't know how
to do that.\n"
;

245
The for Loop
The for loop provides a tool for processing a list of input. The input to for loop is a list of values. Every
iteration through the loop extracts one value into a variable and then enters the body of the loop. The loop
stops when the extract fails because there are no more values in the list.
for loop operates on a list and repeats commands in the block for each element on the list.

for x in [ list ]do

commands
done
Let's consider the following example which prints each of the command line arguments, oneat a time.
We'll extract them from "$@" into $arg:

for var in "$@"do

printf "%s\n" $var
done
Much like C or Java, shell has a break command, also. As you might guess, it can be used to get out of a
loop. Consider this example which stops printing command line arguments, whenit gets to one whose value
is "quit":

for var in "$@"do

if [ "$var" = "quit" ]break
thenfi printf "%s\n" $var
done
Similarly, shell has a continue that works just like it does in C or Java.

for var in "$@"do

if [ "$var" = "me" ]
then
continue
elif [ "$var" = "mine" ]then
continue
elif [ "$var" = "myself" ]

246
then
continue
fi
if [ "$var" = "quit" ]then
break
fi
printf "%s\n" $vardone

The while and until Loops

Shell has a while loop similar to that seen in C or Java. It continues until the predicate is false. And, like
the other loops within shell, break and continue can be used. Here's an example of a simple while loop:

# This lists the files in a directory in alphabetical order

# It continues until the read fails because it has reached the end of inputls | sort |
while read filedo
echo $file
done
In the above code, | called a ―pipe‖ directs output from one process to another process. For
example, ls | sort takes the output from ls command, and sort the output stream received. Pipeis a form of
inter process communication which we will discuss later.
There is a similar loop, the until loop that continues until the condition is successful -- inother words,
while the command failes. This will pound the user for input until it gets it:
printf "ANSWER ME! "until read $answer
do
printf "ANSWER ME! "
done

PRACTICE QUESTIONS

Program1: Write a shell script program to display “HELLO WORLD”.Program2: Write a

shell script to find the factorial of given integer
Program3: Write a shell Script program to check whether the given number is even
or odd.

247