Einar Krogh

An Introduction to Windows
Operating System

2
Download free eBooks at bookboon.com
An Introduction to Windows Operating System
1st edition
© 2015 Einar Krogh & bookboon.com
ISBN 978-87-403-1169-3

3
Download free eBooks at bookboon.com
 An Introduction to
Windows Operating System Contents

Contents

Introduction 11

1 About Windows history 12

1.1 MS-DOS 12
1.2 The first versions of Windows 12
1.3 Windows NT 13
1.4 Windows versions based on Windows NT 14
1.5 Windows Server 16
1.6 Control Questions 18

2 The tasks of an operating system

360° 19
2.1
2.2
2.3
About the construction of computers
Central tasks for an operating system
Control Questions thinking . 19
21
22

360°
thinking . 360°
thinking .
Discover the truth at www.deloitte.ca/careers Dis

© Deloitte & Touche LLP and affiliated entities.

Discover the truth at www.deloitte.ca/careers © Deloitte & Touche LLP and affiliated entities.

Deloitte & Touche LLP and affiliated entities.

Discover the truth

4 at www.deloitte.ca/careers
Click on the ad to read more
Download free eBooks at bookboon.com

© Deloitte & Touche LLP and affiliated entities.

 An Introduction to
Windows Operating System Contents

3 Some components and concepts of the Windows operating system 23

3.1 Windows API 23
3.2 Objects and Handles 25
3.3 Windows runs in two modes 26
3.4 Processes and threads 27
3.5 Services, functions and routines 27
3.6 Unicode 29
3.7 Control Questions 29

4 The architecture of Windows 30

4.1 Some aims for the Windows NT operating system 30
4.2 About the structure of Windows 30
4.3 User mode
TMP PRODUCTION NY026057B 4
31
12/13/2013
4.4 Kernel Mode 32
6x4 PSTANKIE ACCCTR00
4.5 Windows Executive Services 33
gl/rv/rv/baf Bookboon Ad Creative
4.6 The kernel 33
4.7 Hardware Abstraction Layer 34
4.8 Device Drivers 35

All rights reserved.

© 2013 Accenture.

Bring your talent and passion to a

global organization at the forefront of
business, technology and innovation.
Discover how great you can be.
Visit accenture.com/bookboon

5 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Contents

4.9 The window and graphics system 36

4.10 System Processes 36
4.11 Control Questions 37

5 System Mechanisms 38
5.1 Interrupts 38
5.2 Exceptions 39
5.3 Object Manager 40
5.4 32-bit applications on a 64-bit operating system 41
5.5 Control Questions 41

6 Mechanisms for data management 42

6.1 The Registry 42
6.2 Configuration Manager 43
6.3 Control Questions 43

The Wake
the only emission we want to leave behind

.QYURGGF'PIKPGU/GFKWOURGGF'PIKPGU6WTDQEJCTIGTU2TQRGNNGTU2TQRWNUKQP2CEMCIGU2TKOG5GTX

6JGFGUKIPQHGEQHTKGPFN[OCTKPGRQYGTCPFRTQRWNUKQPUQNWVKQPUKUETWEKCNHQT/#0&KGUGN6WTDQ
2QYGTEQORGVGPEKGUCTGQHHGTGFYKVJVJGYQTNFoUNCTIGUVGPIKPGRTQITCOOGsJCXKPIQWVRWVUURCPPKPI
HTQOVQM9RGTGPIKPG)GVWRHTQPV
(KPFQWVOQTGCVYYYOCPFKGUGNVWTDQEQO

6 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Contents

7 Processes, Threads and Jobs 44

7.1 About processes 44
7.2 About threads 47
7.3 About fibers 49
7.4 About jobs 50
7.5 The Process Manager 51
7.6 Multitasking 51
7.7 Communication between processes 52
7.8 Control Questions 53

8 Synchronization of threads 54
8.1 Threads cooperating 54
8.2 Threads competing 54
8.3 Critical region 55
8.4 Mutual Exclusion 55
8.5 Semaphore 56
8.6 Mutex 57
8.7 Some synchronization mechanisms in Windows 58
8.8 Control Questions 60

30 FR
da EE
SMS from your computer ys
...Sync'd with your Android phone & number tria
l!

Go to

BrowserTexting.com

and start texting from

your computer!

... BrowserTexting

7 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Contents

9 Process planning on one processor 61

9.1 Voluntary and involuntary sharing of the processor 62
9.2 Non-preemptive process planning 63
9.3 Preemptive process planning 63
9.4 Scheduling for various operating systems 64
9.5 Context Switching 66
9.6 About the length of a Quantum 66
9.7 About scheduling in Windows 67
9.8 Control Questions 68

10 Process planning on multiple processors 69

10.1 Multi-core processors 70
10.2 The organization of multiple processors 70
10.3 Symmetric Multiprocessing 71
10.4 Multiple processors and synchronization 72
10.5 Process scheduling on multiple processors 73
10.6 Multiprocessing in Windows 74
10.7 Control Questions 75

11 Deadlocks 76
11.1 Examples of deadlocks 78
11.2 Conditions for a deadlock to occur 80
11.3 How the operating system can manage deadlocks 80
11.4 Tools to find deadlocks 82
11.5 Control Questions 83

12 Memory Management 84
12.1 Swapping 84
12.2 Some memory models 84
12.3 Virtual Memory 85
12.4 Memory management in Windows 87
12.5 The Memory Manager 87
12.6 The Heap Manager 88
12.7 Stacks 88
12.8 Control Questions 88

8
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Contents

13 Caching in Windows 89
13.1 The Cache Manager 89
13.2 Control Questions 89

14 Input and output 90

14.1 Problems with input and output 90
14.2 Block devices and character devices 91
14.3 Data Streams 92
14.4 Treatment of errors in input and output 92
14.5 Synchronous and asynchronous input/output 93
14.6 Use of a buffer 93
14.7 Polling 94
14.8 Interrupts 94
14.9 Direct access to memory 95
14.10 About drivers for input/output devices 96
14.11 Input and output in Windows 96
14.12 The Input/output Manager 97
14.13 The Plug and Play Manager 98
14.14 The Power Manager 98
14.15 Control Questions 99

15 File systems in Windows 100

15.1 FAT 100
15.2 NT File System 100
15.3 Support for SSD disks 101
15.4 File systems for CD-ROM and DVDs 102
15.5 Control Questions 102

16 Storage Management 103

16.1 About storage on hard drives 103
16.2 Organization of hard disks 104
16.3 Storing data on servers 107
16.4 Control Questions 108

17 Networking Features in Windows 109

17.1 Network Services 109
17.2 Some network APIs 110
17.3 About network drivers in Windows 111
17.4 Control Questions 112

9
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Contents

18 Security in Windows 113

18.1 About security in computer systems 113
18.2 Security in Windows NT 113
18.3 Security Mechanisms in Windows 114
18.4 Defense against Spyware 115
18.5 Virus Protection 116
18.6 The firewall in Windows 117
18.7 Windows Update 117
18.8 Control Questions 117

19 When Windows crashes 118

19.1 The blue screen 118
19.2 Some reasons that Windows can crash 119
19.3 File dumping by system crash 120
19.4 Control Questions 120

Wordlist 121

Some references 126

10
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Introduction

Introduction
This book is about Windows operating system written for courses in operating systems.

Students of operating systems need some knowledge of Windows operating system. This book presents
the key components of Windows operating system in a short and straightforward way and introduces
how a computer system using Windows operating system works.

The book also discuss general theory of operating systems. Central subjects in the book are how
applications are running on a computer and how applications communicate with hardware.

The book is therefore both suitable for introductory courses on how Windows operating system functions
and for beginning courses in operating systems in general.

11
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System About Windows history

1 About Windows history

A Personal Computer (PC) is a computer designed for use by private individuals. The typical use of a
PC has been for word processing, spreadsheets, databases, Web browsing, e-mail and computer games.

The history of personal computers began in earnest in 1977, with the introduction of the microcomputer.
When the microprocessor was developed, it was possible to create computers so affordable to purchase
that private individuals could buy them.

When IBM (International Business Computers) began producing a Personal Computer in 1980 it created a
need for an operating system for this new computer. IBM approached Microsoft with a request. Microsoft
was not developing operating systems at this time, but since the mid-1970s had been a leader in the
development of tools for the programming language, BASIC. Microsoft acquired an operating system
by purchasing it from Seattle Computer Products, and then made changes to this system so that it was
suited for PCs. They called the operating system MS-DOS.

1.1 MS-DOS
Microsoft called the new operating system MS-DOS 1.0

DOS = Disk Operating System

MS-DOS was an operating system based on a simple communication between user and computer. The
user entered commands on the screen using the keyboard, as a mouse was used to a limited extent at
this point.

MS-DOS was the most common operating system for PCs before Windows, and was also the base
operating system in the first versions of Windows. MS-DOS therefore continued to live long after the
introduction of Windows.

Microsoft has made many versions of MS-DOS, with MS-DOS 8.0 being the latest version, which was
launched in the year 2000.

1.2 The first versions of Windows

The development of operating systems with a graphical user interface started in the 1980s, which was
when Microsoft decided to give MS-DOS a graphical user interface. Microsoft created a graphical program
for this purpose called Windows. At first, Windows was not a separate operating system, but instead a
graphical application that used MS-DOS as operating system.

12
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System About Windows history

Windows 1.0 was the first version and came in 1983, though the first versions of Windows (1.0 and 2.0)
were not a success. The reason for this was that there was little software for Windows, while the existing
software was unstable and simple. It was not until Windows 3.0 and then Windows 3.1 that Windows
became widely taken into use.

Release Version About the version

1990 Windows 3.0 Ten million copies sold.

1991 Windows 3.1 Windows becomes widespread.

1995 Windows 95 New and improved version.

1998 Windows 98 Integrating the Internet.

2000 Windows Millennium Focus on multimedia.

Figure 1.1: The figure shows the first versions of Windows

The operating systems Windows 95, Windows 98 and Windows Millennium were quite similar, so they
were therefore called Windows 9x.

Windows 95

Windows 95 was a major improvement from earlier versions of Windows, central to this being an
enhanced usability and better networking capabilities. Windows 95 was not just a GUI for MS-DOS as
earlier versions of Windows had been. MS-DOS was still at the bottom, but Windows 95 also used its
own drivers, such as for a mouse, CD-ROM and an audio card. The new features in Windows 95 were
multitasking and the automatic detection and configuration of equipment (Plug and Play).

Windows 98

Windows 98 was an upgrade and improvement of Windows 95, as Microsoft wanted to implement
the Internet in Windows at this time. As a result, Windows 98 contained Internet Explorer and other
programs for the Internet.

Windows Millennium

Windows Millennium was the last release in the series based on the Windows 9x platform, and was
aimed at the domestic market with a focus on multimedia.

1.3 Windows NT
Computers acquired more and more memory, a higher processor speed and more disk space. MS-DOS
could not handle a lot of memory, and could not run multiple applications simultaneously, so there was
a need for a new operating system. This led to the developing of Windows NT, in which NT stands for
New Technology.

13
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System About Windows history

Microsoft released the first version of Windows NT in 1993 and called it Windows NT 3.1, because it
came at the time when Windows 3.1 was in use. Windows NT 3.1 had the same appearance and user
interface as Windows 3.1, but they were completely different operating systems.

Windows NT 4.0 came in 1996, with the two basic versions of Windows NT 4.0 being:

• Windows NT Server
• Windows NT Workstation

Windows NT 4.0 Server was a network operating system designed to be used on servers in local networks.
Windows NT 4.0 Workstation was designed for use on both home computers and workstations, but was
primarily used by companies as workstations in local networks.

Windows NT 4.0 was an advanced 32-bit operating system designed to be secure, stable and flexible, thus
making it possible to use multiple processors on the same computer. Windows NT 4.0 could utilize a
large memory and large hard drives, which meant that the operating system was well scalable. However,
Windows NT 4.0 did not have all the features of Windows 95, such as automatic detection and the
configuration of equipment (Plug and Play).

Windows NT introduced users as a part of the system’s security model, so in order to use a workstation in
a modern Windows system you must be a registered user. One can define multiple users on a workstation,
and each user has certain rights to files and to access system resources in general.

1.4 Windows versions based on Windows NT

Windows NT operating system is the basis for the current versions of Windows, though with the exception
of Windows 2000, the latest versions of Windows are used on both home computers and workstations
in a network.

Release Version About the version

2000 Windows 2000 Not intended for home use.

2001 Windows XP Both for home use and in local networks.

2006 Windows Vista New user interface with Windows Aero.

2009 Windows 7 Emphasis on functionality and performance.

2012 Windows 8 New graphical interface Microsoft Metro.

Figure 1.2: The figure shows the versions of Windows based on Windows NT

Windows NT has become a widespread operating system in the computer world, as today you can find
variations of the original Windows NT on laptops, desktops, servers and Xbox consoles worldwide.

14
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System About Windows history

Windows 2000

Windows 2000 (Windows NT 5.0) was built on Windows NT 4.0. In addition, Windows 2000 had most
of the useful qualities of Windows 98, such as support for Plug and Play. Windows 2000’s operating
system was available in several versions. One version was for workstations, and there were several
versions for servers.

Windows XP

Launched in August 2001, Windows XP has been the most popular version of Windows, based on
the number of copies sold. Windows XP first came as Windows NT 5.1, and was later upgraded to
Windows NT 5.2.

There have been several versions of Windows XP, with the two most important being:

ƒƒ Windows XP Home (for home use)

ƒƒ Windows XP Professional (for companies)

These versions were the same, but XP Professional had some additional features compared to
Windows XP Home.

Windows Vista

Launched in November 2006, Windows Vista (Windows NT 6.0) contained hundreds of new and revised
features. One of the new features in Windows Vista was an updated graphical user interface called
Windows Aero, and Windows Vista also introduced Bitlocker, a tool to encrypt hard drives. Windows
Vista had improved safety, because Windows XP and earlier versions of Windows had been criticized
for a lack of security.

Windows 7

Windows 7 (Windows NT 6.1) came in October 2009. Unlike previous versions of Windows, Windows 7
did not contain many new features, basically being more of an upgrade of Windows Vista. The objective
of Windows 7 was an operating system that had increased functionality and performance over
previous versions.

Windows 8

Windows 8 (Windows NT 6.2) came in October 2012, and contained a new graphical interface known
as Metro, which is used for tablets, laptops, desktops and Windows Phone. Optimized for touch screens,
Metro can also be controlled with a mouse and keyboard. Additionally, Windows 8 has a faster and more
stable startup than previous versions of Windows.

15
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System About Windows history

In April 2014, Microsoft released an update called Windows 8.1. The main new features were in the user
interface, security and hardware compatibility and functionality.

1.5 Windows Server

Microsoft has made several versions of the Windows operating system adapted to act as a server in
local networks.

Microsoft’s first attempt at creating an operating system with network features was Windows 3.11,
which came at the same time as Windows 3.1 (1992). Another name for Windows 3.11 is Windows for
Workgroups, and this version had some additional features that gave network support:

ƒƒ Network cards and cables

ƒƒ Sharing directories, disks and printers
ƒƒ E-mail and instant messaging functionality

However, it was with Windows NT that Microsoft began developing a network operating system intended
to act as a server.

Brain power By 2020, wind could provide one-tenth of our planet’s

electricity needs. Already today, SKF’s innovative know-
how is crucial to running a large proportion of the
world’s wind turbines.
Up to 25 % of the generating costs relate to mainte-
nance. These can be reduced dramatically thanks to our
systems for on-line condition monitoring and automatic
lubrication. We help make it more economical to create
cleaner, cheaper energy out of thin air.
By sharing our experience, expertise, and creativity,
industries can boost performance beyond expectations.
Therefore we need the best employees who can
meet this challenge!

The Power of Knowledge Engineering

Plug into The Power of Knowledge Engineering.

Visit us at www.skf.com/knowledge

16 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System About Windows history

Year Version

1996 Windows NT 4.0 Server

2000 Windows Server 2000

2003 Windows Server 2003

2008 Windows Server 2008

2012 Windows Server 2012

Figure 1.3: The figure shows the server versions of Windows

Windows NT 4.0 Server

It was first with Windows NT 4.0 that Microsoft got a proper network operating system adapted to the task
as a server in a local area network, as the Windows NT 4.0 Server could serve many users in a network.

Windows Server 2000

The sequel to Windows NT 4.0 Server was Windows Server 2000, which had many of the features
of Windows 98. However, the main novelty in Windows Server 2000 was Active Directory, which is
Microsoft’s directory service for managing users and resources in a local area network.

Other news in Windows Server 2000:

1. Protection of system files. Created automatically by the operating system, if deleted.

2. Microsoft Installer. For easy installation and uninstallation of applications.
3. Better stability and security. Ability to encrypt files.
4. Better support for hardware devices. Among other high-speed USB pens.

Windows Server 2003

Built on Windows XP, Windows Server 2003 had several improvements over Windows Server 2000, the
most important of which were the Active Directory, disk management and security. New in Windows
Server 2003 was the Internet Information Services (IIS) used to create web pages for servers.

Windows Server 2008

Launched in February 2008, Windows Server 2008 was an upgrade of Windows Server 2003. Windows
Server 2008 had enhancements to Active Directory, Group Policy, disk management and security.

Windows Server 2008 was built on the same code as Windows Vista. When Windows 7 came out in
2009, there was a need for an upgrade. The name of this upgrade was Windows Server 2008 R2, which
was the first Microsoft operating system that just came as 64-bit.

17
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System About Windows history

Windows Server 2012

Windows Server 2012, also called Windows Server 8, debuted in September 2012. New with Windows
Server 2012 was better support for server virtualization and the use of the cloud (Cloud Computing).
Windows Server 2012 includes a new file system (ReFS), but this file system is only used by file servers.

An upgrade to Windows Server 2012, called Windows Server 2012 R2, was released along with
Windows 8.1 in October 2013.

1.6 Control Questions

1. At what point in time was Windows taken into use as an operating system?
2. Which versions of Windows are called Windows 9x?
3. What was the task of MS-DOS in the first versions of Windows?
4. Mention some reasons that Microsoft developed Windows NT.
5. Mention some differences between Windows NT and Windows 9x.
6. Mention some differences between Windows Server and Windows for PCs.

> Apply now

redefine your future

AxA globAl grAduAte
progrAm 2015
- © Photononstop

axa_ad_grad_prog_170x115.indd 1 19/12/13 16:36

18 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System The tasks of an operating system

2 The tasks of an operating system

There are two types of software running on a computer: software applications and system software. A
software application is a program designed for end users, while system software consists of low-level
programs belonging to the operating system, compilers and utilities for managing resources.

An operating system is software that enables services for software applications to run on a computer.
An important task of an operating system is taking care of the communication between the software
applications and hardware devices attached to your computer. For example, a word processor
communicates with devices such as a keyboard and mouse.

Operating systems are large programs consisting of thousands of functions, which provide services of
various kinds. Often called by events in the system, the functions perform a service when needed. To
make operating systems fast and most effective, the functions are often written in C or C++, but also
low-level programming languages such as Assembly has occasionally been used.

2.1 About the construction of computers

To understand the tasks of an operating system, it is necessary to have some knowledge of computer
systems. In this text, there will be no detailed description of the constructions of computers, but we will
look at some key components of the computer system and show why these need to be administered by
an operating system.

Item Description

Processor A processor is the part of the computer system that is able to execute a program code.

Memory Computer memory is a temporary storage location for data and the program code. Data is lost
in the memory when the power of the computer is turned off.

Hard drive The task of hard drives is to permanently store data on a computer. The hard drives store
program files, documents and data files.

Input/output Equipment used for communication between the computer and a user. The most common
devices input/output devices are the keyboard, mouse and screen.

Figure 2.1: The figure shows some key devices in a computer system

A processor executes the instructions that a computer program is comprised of. The processor is therefore
the central unit of program execution.

Several applications are often simultaneously executed on a computer. It is therefore a task for the
operating system to launch programs and allocate time for programs in one or more processors.

19
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System The tasks of an operating system

The applications running are loaded into a computer’s memory for storage. Consequently, it is therefore
necessary to keep track of where in the memory different data resides. In addition, there is a need to
keep track of which parts of the memory are available and which parts are in use. Managing computer
memory is a task for the operating system.

Files are stored on a hard disk. The operating system keeps track of where files on the hard disk are
situated, as well as what parts of the hard disk are in use and what parts are not in use.

Input/output means communication between software applications and screen, keyboard, mouse and
similar devices. The operating system takes care of the input/output to make it easier for applications to
communicate with different drivers for input/output devices.

In addition to creating services for the computer system, it is a task for an operating system to provide
a user interface for the computer. Windows has a graphical user interface in which the user uses the
keyboard, mouse and touchscreen to interact with applications.

LIGS University
based in Hawaii, USA
is currently enrolling in the
Interactive Online BBA, MBA, MSc,
DBA and PhD programs:

▶▶ enroll by October 31st, 2014 and

▶▶ save up to 11% on the tuition!
▶▶ pay in 10 installments / 2 years
▶▶ Interactive Online education
▶▶ visit www.ligsuniversity.com to
find out more!

Note: LIGS University is not accredited by any

nationally recognized accrediting agency listed
by the US Secretary of Education.
More info here.

20 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System The tasks of an operating system

2.2 Central tasks for an operating system

We want fast and reliable computers; hence an important task for an operating system is to utilize
computer resources in the best possible way.

There are two main goals of an operating system:

• Managing applications and giving applications access to hardware.

• Manage data system resources optimally.

Important tasks for an operating system are to:

1. Manage applications running on the computer.

2. Take care of input and output.
3. Manage communication between software applications and hardware.
4. Manage computer memory.
5. Manage the file system.
6. Manage networking.
7. Take care of the security of the computer system.
8. Provide a user interface for the computer.

The operating system manages applications running on the computer. The operating system starts and
stops applications, and provides time for applications in the processor. An important task for an operating
system is to help facilitate communication between applications and hardware, while user applications
access the hardware through the operating system.

An operating system is a layer between applications and hardware. A user interacts with an application, the
application interacts with the operating system and the operating system communicates with hardware.

8VHU

$SSOLFDWLRQ

2SHUDWLQJV\VWHP

+DUGZDUH

Figure 2.2: The interaction in a computer system

21
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System The tasks of an operating system

Hardware equipment consists of various types and uses different drivers, whereas the operating system
takes care of the communication with the hardware. This makes it much easier to create computer
programs since programmers do not need to consider communication with different drivers.

Modern operating systems use multitasking, which allows multiple applications to run simultaneously,
thereby making it is possible to perform more than one action on the computer at approximately the
same time. To help achieve this, programs take turns running in the processor. An application executes
in a processor for a short time before it must be out of the processor, and another application is loaded
into the processor for execution.

Data security is important when using computer technology, particularly for computers attached to
a network. The objective of computer security is to therefore provide information and data against
unauthorized access, theft, destruction and natural disasters.

2.3 Control Questions

1. Describe the tasks of the processor, memory, hard drive and input/output devices.
2. Explain why it is beneficial that an operating system manages the processor, memory, hard
drive and input/output devices.
3. What makes it difficult for applications to communicate directly with hardware?
4. What are the two main goals for an operating system?
5. Mention the most important tasks for an operating system.

22 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to Some components and concepts of the
Windows Operating System Windows operating system

3 Some components and

concepts of the Windows
operating system
In this chapter, we get to know some basic components and concepts of the Windows operating system.

3.1 Windows API

Windows API is a library of functions that form an interface between the Windows operating system
and software applications running on a computer.

API = Application Program Interface

The main task of the Windows API is to take care of the communication between user applications
and the operating system, as any software application designed for Windows uses Windows API. For
example, events in Windows such as keystrokes, movement of the mouse and the insertion of a CD call
on functions in Windows API that execute the event.

WriteFile Writes data to a file.

CreateProcess Starts a process.

SetFocus Gives focus to an object.

EndDialog Closes a dialog box and removes it from the memory.

CloseWindow Removes a window from the screen.

Figure 3.1: The figure shows some examples of the functions in Windows API

Programmers who create applications for Windows do not use the functions in Windows API directly.
When a programmer who develops a program for Windows wants to communicate with the operating
system, this is done through calls to library functions in the development tool. The library functions will
then send further calls to functions in Windows API. A function whose task is to call another function
is called a Wrapper Function.

Windows API is an old library used by all versions of Windows from Windows 95 and onward. Windows
operating systems have changed with each version of Windows, as has Windows API. Windows API was
originally 16-bit, but Windows API has gradually become 32-bit and 64-bit. The 32-bit API has been
called Win32 API, and today 32-bit and 64-bit live side-by-side. Windows 8 is available in both 32-bit
and 64-bit versions, so Windows API will therefore refer to both 32-bit and 64-bit versions.

23
Download free eBooks at bookboon.com
An Introduction to Some components and concepts of the
Windows Operating System Windows operating system

About 32-bit and 64-bit architecture

32-bit and 64-bit architecture refers to the length of the memory address used by the processor. The 64-bit version
of Windows, which handles large amounts of RAM (Random Access Memory), is more effective than a 32-bit system.
However, much of today’s software was made during the 32-bit time. To fully benefit from a 64-bit operating system,
you must have software optimized for 64-bit processing.

Windows API has services used by all Windows applications, which consists of thousands of features,
divided into the following categories:

• Administration and management

• Diagnosis
• Graphics and multimedia
• Network
• Security
• System services
• Windows user interface

The administration and management part of Windows API has many services for applications and
systems. One of them is the installation and configuration of software applications.

The diagnosis part of the Windows API handles problems that occur during program execution, including
functions for logging, processing errors, troubleshooting and more. The diagnostic part of the Windows
API is attempting to fix bugs and system errors. What can happen if a program has a problem is that
the Windows API manages to remedy the defect or offers to restart the program.

The Graphics and Multimedia section of the Windows API provides applications for the functionality
to print graphics to the screen, printers and similar equipment. The graphics and multimedia section
of the Windows API allows applications to incorporate formatted text, graphics, audio and video, and
there are functions to draw lines, curves and shapes, as well as use pens, images and colors. Microsoft
DirectX is a collection of APIs that takes care of tasks related to multimedia on Microsoft platforms.
DirectX is also a tool used in game programming.

The Network part of the Windows API has facilities that make it possible to achieve communication
between applications on different computers within a network. There are several standards such as
Windows Sockets, Remote Procedure Call (RPC) and Simple Network Management Protocol (SNMP).

Windows API has many services for security, and some of them include passwords at logon, the protection
of shared system objects, access control, rights, monitoring, encryption and more.

24
Download free eBooks at bookboon.com
An Introduction to Some components and concepts of the
Windows Operating System Windows operating system

System Services in Windows API provides applications access to resources on the computer such as
memory, file systems, input/output devices, hardware devices, processes, threads and more.

Windows API has functions that deal with the user interface of Windows. For example, there are functions
in the Windows API that create and manage windows, in addition to functions that manage controls
such as menus, cursors, icons and combo boxes. Windows API also contains a number of features to
manage windows, geometric shapes, text, fonts, scrollbars, dialog boxes, menus and similar GUI features.

3.2 Objects and Handles

An object in the Windows operating system is a structure used to store data. An object consists of
system-defined data types called attributes and functions that treat the attributes.

Objects are widely used in Windows operating system, e.g. files, windows, pictures, semaphores, processes
and threads. Nonetheless, not all data in Windows operating system are objects, as only data that needs
sharing, protection or to be visible to user programs is located in objects.

About Windows Objects

Objects in the Windows operating system are similar to objects used in object-oriented programming. However,
objects in Windows lack some of the characteristics in object-oriented programming systems, such as inheritance
and polymorphism.

25 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to Some components and concepts of the
Windows Operating System Windows operating system

The Windows operating system distinguishes between three types of objects. There are user objects,
graphical objects (GDI Objects) and kernel objects. The task of user objects is to support the management
of windows, while the task of graphical objects is to support graphics. Kernel objects support memory
management, process execution and communication between processes.

Objects have the following important tasks in the operating system:

• To provide the names of system resources that are readable for humans.
• To share resources and data between processes.
• To protect resources from unauthorized access.
• To keep track of data that is no longer in use, and is ready for deletion.

To increase security, each object has a Security Descriptor attached. A Security Descriptor contains
information about which operations are allowed to perform on the object.

In Windows NT, API functions create objects in the kernel mode. Every time a new object is created,
a handle to the object is returned.

A handle is a type of pointer used in Windows. A user application does not have direct access to the
data objects. Therefore, the applications use handles in order to reach and change the data in objects.

Handles are used to perform operations on objects, just as pointers and references do in object-oriented
programming. To get ahold of data and functions in an object by means of a handle is faster than looking
it up using the name of the object.

3.3 Windows runs in two modes

The Windows operating system runs programs in two modes:

ƒƒ User Mode
ƒƒ Kernel Mode

Software applications run in user mode, whereas processes associated with the operating system runs
in kernel mode.

The purpose of two modes in Windows is to increase the security of the operating system, as using two
modes protects the operating system from applications that can damage or alter critical data belonging
to the operating system.

26
Download free eBooks at bookboon.com
An Introduction to Some components and concepts of the
Windows Operating System Windows operating system

The difference between kernel mode and user mode is the access that processes have to system resources.
Processes running in kernel mode have access to all hardware and can perform as many tasks as possible
at the computer. Processes belonging to applications running in user mode only have limited access to
system resources.

3.4 Processes and threads

The terms program, process and thread are closely related in the theory of operating systems. Programs
give rise to processes and threads, although processes and threads are quite different from programs.

Term Definition

Program A program is program code and corresponding data that resides at one or more files on a hard drive.

Process When a program starts and is loaded into the computer memory, it creates a process.

Thread A thread is a portion of a process code to be executed in the processor, and is the dynamic part of
a process.

Figure 3.2: The figure provides a definition of a program, process and thread.

The management of processes and threads is a central task for an operating system. The operating system
must start and terminate processes, allocate resources to processes, take care of the synchronization of
processes, organize communication between processes and protect processes from each other.

Processes and threads are central concepts in the theory of operating systems, and we shall consider
them in more detail later in the chapter on processes, threads and jobs.

3.5 Services, functions and routines

In connection with operating systems, we have the concepts of function, routine, subroutine and service:

1. In the C programming language, a function is a piece of code in a program called by other

parts of the program.
2. A routine is defined as a piece of code in a program that is used repeatedly by calls when the
program is run.
3. A subroutine can be defined as a unit with program instructions for performing a specific task.
4. A service is usually one or more routines that execute when you call them.

The meaning of function, routine and service is a matter of definition. In the following, we shall look at
how these terms are used in connection with different parts of the Windows operating system.

27
Download free eBooks at bookboon.com
An Introduction to Some components and concepts of the
Windows Operating System Windows operating system

In conjunction with the Windows operating system, the following terminology is often used:

• Windows API functions

• Native System Services
• Kernel functions
• Windows services
• DLLs

Windows API functions are documented subroutines called from user applications. You therefore use
the word functions in connection with library routines in the Windows API.

Native System Services in Windows are undocumented services running in kernel mode. Drivers can call
these services directly, whereas applications in user mode can call them via Windows API. For example,
when a user application creates a file, the function CreateFile in Windows API will call an internal system
service with the name NtCreateFile. Many system services have a name that starts with Nt, in which Nt
refers to the Native System Services.

Kernel Support Functions or Routines are subroutines in the operating system only called from the
kernel mode.

Windows Services are processes that run in the background, and perform some function. The Windows
operating system has approximately 50 default services, e.g. DNS is used to find an IP address for a web
address. Another example of a Windows service is a Web server that runs in the background.

It is also possible for users to add their own services. You can configure Windows services to start
automatically when you restart, but you can also start Windows services by hand.

Many Windows Services can be seen in the Windows Task Manager, where they often have names such
as System, Local Service or Network Service.

Dynamic Link Libraries (DLLs) are subroutines used by applications in user mode, and Windows
programs often use DLLs. The advantage of DLLs is that multiple processes can share them, which leads
to a better utilization of the memory when the operating system ensures that only one copy of a DLL
is in the memory at a time.

The opposite of DLLs are Static Libraries, in which code is copied into each application that uses them.

28
Download free eBooks at bookboon.com
An Introduction to Some components and concepts of the
Windows Operating System Windows operating system

3.6 Unicode
Unicode is a standard for representing characters written on a computer, for example characters used in a
word processor, and are also an international character set that can represent all the characters in the world.

The most widely used standard for characters has previously been ASCII (American Standard Code for
Information Interchange). This character set is suited for the English language, though it only has 256
characters, which is not enough in an international world. For example, Chinese, Korean and Japanese
have 10,000 commonly used characters. Consequently, Unicode was developed. Unicode has 65,536
characters, which is sufficient for all the languages in the world.

The advantage of Unicode is that it allows multilingual communication, including the representation
of multiple languages in a single document. For many languages, it is difficult or impossible to do this
without using Unicode.

3.7 Control Questions

1. What is API an abbreviation for?
2. What is the task of the Windows API?
3. How do programmers use Windows API?
4. What is an object in Windows and why is the Windows operating system using objects?
5. What is the task of a handle in the Windows operating system?
6. Windows operating system runs in two modes. What are the names of these two modes?
7. Give a definition of the concepts of program, process and thread.
DO YOU WANT TO KNOW:
8. What is Unicode, and what is the reason that Windows uses Unicode?

What your staff really want?

The top issues troubling them?

How to make staff assessments

work for you & them, painlessly?

How to retain your

top staff Get your free trial
FIND OUT NOW FOR FREE
Because happy staff get more done

29 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System The architecture of Windows

4 The architecture of Windows

In this chapter we look at the inner structure of Windows. Windows NT is the basis for today’s Windows,
and we shall first look at some aims for Windows NT.

4.1 Some aims for the Windows NT operating system

When planning Windows NT in the early 1990s, the following objectives for the operating system were set.

1. Extensibility
2. Transferability
3. Reliability and robustness
4. Compatibility
5. Performance

The operating system needed to be expandable, meaning that the code to the operating system had to
be able to grow and change with the times and development.

The operating system also needed to be transferable, which means that the operating system had to work
on different types of hardware architecture. The operating system should also be easily adapted to new
types of hardware architecture that would come in the future.

A reliable and robust operating system has to cope with bugs in their own code. An error in an application
running on the system should not be able to harm the operating system or user applications running.

When Windows NT was launched, it was compatible with earlier versions of Windows and MS-DOS.
This means that applications for earlier versions of Windows and MS-DOS had to be able to run on
Windows NT.

The operating system has to perform well, which means that the operating system will work as quickly
as possible on different types of platforms.

4.2 About the structure of Windows

There are different ways to organize an operating system. One way is to create the operating system as
a big application that runs on a computer such as a user application. However, this is not the way that
modern operating systems were created, instead consisting of several layers, in which each layer has a
specific task.

30
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System The architecture of Windows

The architecture of the Windows operating system is a layered system with the two main components
of user mode and kernel mode. User applications run in user mode, while processes belonging to the
operating system run in kernel mode.

User Mode User Applications

Kernel mode The operating system processes

consists of executive services,
device drivers, the kernel and
HAL.

Figure 4.1: Programs are running in two modes

Central components in kernel mode are the kernel, a hardware layer (HAL), drivers and a variety of
services called Executive Services. Another important component of the kernel mode is the device drivers,
which manage physical devices attached to your computer, the file system, network protocols and more.

4.3 User mode

To protect the operating system from user applications, which can damage or critical alter data belonging
to Windows operating system, applications runs in user mode. User mode also protects applications
against each other.

Applications in user mode communicate with the kernel through the Windows API and access the
hardware through the operating system. Applications in user mode cannot access data in other
applications memory unless through Windows API.

Processes from user applications have a lower priority than the processes belonging to the operating
system.

31
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System The architecture of Windows

4.4 Kernel Mode

Processes in kernel mode have full access to the hardware and system resources. To get good security,
processes in kernel mode runs in a protected area of memory.

The key parts of kernel mode are the kernel, the executive services, the hardware layer and device drivers.

The kernel

Executive services

Drivers

Hardware Layer (HAL)

Graphical User Interface (GUI)

Figure 4.2: The organization of the kernel mode

The purpose of the kernel mode is to prevent applications in the user mode from being able to access
critical areas of the operating system. Applications in the user mode must get help from the kernel mode
to perform these kinds of operations.

Challenge the way we run

EXPERIENCE THE POWER OF

FULL ENGAGEMENT…

RUN FASTER.
RUN LONGER.. READ MORE & PRE-ORDER TODAY
RUN EASIER… WWW.GAITEYE.COM

1349906_A6_4+0.indd 1 22-08-2014 12:56:57

32 Click on the ad to read more
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System The architecture of Windows

4.5 Windows Executive Services

Executive Services are services in the Windows operating system that help applications in the user
mode to run on the computer system. Executive services include many basic services in the Windows
operating system.

The operating system file Ntoskrnl.exe

The upper layer in the OS file Ntoskrnl.exe is the executive part of Windows, while the lower layer in Ntoskrnl.exe is the
kernel. The executive part of Windows has the name Windows Executive.

The executive services in Windows are in kernel mode. The executive services do central services for
the operating system, including the management of processes and threads, memory management and
processing of input/output.

Some of the main components of the Windows Executive Services:

• Object Manager
• Configuration Manager
• Process and thread Manager
• Input/output Manager
• Installation and configuration Manager
• Memory Manager
• Cache Manager
• Power Manager

Executive services also include a variety of other features that are important to the infrastructure of the
Windows operating system.

4.6 The kernel

The kernel refers to the most central part of the operating system, and controls the use of the processors.
The kernel manages the scheduling of threads, context switching, interrupt signals, exception signals
and multiprocessor synchronization.

The kernel sits between the Hardware Layer (HAL) and the executive services. The kernel forms a bridge
between user applications and data processing at the hardware level.

The most important task for the kernel is to allow applications to run and give applications access to
hardware. To run a program, the kernel sets up an address space for the program, loads the program
code into the memory and provides the program with a stack.

33
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System The architecture of Windows

Other important tasks for the kernel are to manage your computer resources and to allow applications
to use these resources. These resources are:

1. The processor – Processes get access to the processor by the kernel.

2. The computer’s memory – The kernel has full access to the computer’s memory and allocates
memory to the processor as needed.
3. Input/output devices – The kernel receives requests from processes to perform input/output,
and provides access to relevant equipment.

When Windows is starting, the kernel is also started. The kernel starts the drivers needed for the
Windows operating system to start and run. The kernel runs in kernel mode, and sets up paging and
virtual memory. It then starts some system processes and lets them run in user mode.

The kernel keeps track of hardware devices connected to the computer by having a list of this equipment.
The kernel provides processes with access to hardware devices when needed.

4.7 Hardware Abstraction Layer

Hardware Abstractions are a set of routines that give applications access to hardware devices regardless
of which manufacturer the equipment belongs to.

This e-book
is made with SETASIGN
SetaPDF

PDF components for PHP developers

www.setasign.com

34 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System The architecture of Windows

The abbreviation for the Hardware Abstraction Layer is HAL, which is a layer between the physical
hardware and the software that runs on the operating system. On a PC, HAL can be seen as a driver for
the motherboard that allows instructions from high-level programming language to communicate with
low-level components like hardware.

A task for HAL is to get the Windows operating system to be able to work with different hardware platforms.
HAL is therefore an interface to the hardware equipment in the computer operating system. Programs in
Windows will not communicate directly with hardware, but communicate with hardware through HAL.

4.8 Device Drivers

Device Drivers are small programs designed to help programs at a higher level to communicate
with hardware.

Drivers are control programs that enable communication between hardware devices and the operating
system, and are necessary for Windows to communicate with hardware.Device Drivers running in kernel
mode, and provide an interface between the input/output manager and hardware.

There are several types of device drivers. The following provides an overview of some of them:

ƒƒ Drivers for hardware devices. Using the hardware layer, these drivers take care of input/
output for physical equipment.
ƒƒ File system drivers are Windows drivers that take care of input/output to files.
ƒƒ File system filter drivers perform tasks such as encryption or writing data to more than
one disk. File system drivers for the network transfer the file system input/output to other
computers at a network.
ƒƒ Protocol drivers implement a networking protocol such as TCP/IP or NetBUI.

Microsoft introduced the Windows Driver Model (WDM) to allow driver developers to write device
drivers that are source-code compatible across all versions of Windows. Kernel-mode drivers that follow
WDM rules are called WDM drivers, and there are three kinds of WDM drivers: bus drivers, function
drivers and filter drivers.

ƒƒ A bus driver drives an individual I/O bus device, and provides a functionality that is device-
independent. Bus drivers also detect and report other devices that are connected to the bus.
ƒƒ A function driver drives an individual device.
ƒƒ A filter driver filters I/O requests for a device or a bus.

Windows includes installation drivers that support a huge number of different hardware devices.
Thousands of drivers are available through a Windows Update, and there are hundreds of new drivers
every month.

35
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System The architecture of Windows

4.9 The window and graphics system

A graphical user interface (GUI) is a type of interface that allows users to communicate with a computer
in ways other than using the keyboard.

The window and graphics system implements the graphical user interfaces for Windows.

This system takes care of processing windows, drawing in windows, processing controls such as menus
and toolbars, receiving mouse and keyboard input and more.

4.10 System Processes

There are many system processes that always run in any Windows system, and we will look at some of these:

• Idle Process
• Session Manager
• Winlogon
• Windows Subsystem

Idle Process consists of one or more threads in the kernel running in the processor when there are no
other threads running on the computer. In a computer system with multiple processors, there is an Idle
Process for each processor.

The original reason for an Idle Process was to avoid the special situation for Process Manager of when
no programs were running on the system.

In computer science, a session is an active exchange of information, also called a dialogue between two
devices. Session Manager is the first process created in user mode, and is responsible for many important
steps in the Windows startup. Session Manager also creates system variables, starts the kernel and user
modes, creates virtual memory and starts winlogon.

The Windows Logon Process manages the login and logout of the computer system. Winlogon can be
activated on the keyboard by using Ctrl + Alt + Delete.

The Windows Subsystem consists of several components that have key responsibilities in the operating
system, including the management of the keyboard, mouse and screen.

The Windows Subsystem is required for Windows to run.

36
Download free eBooks at bookboon.com
 An Introduction to
Windows Operating System System Mechanisms

4.11 Control Questions

1. Why is the Windows operating system divided into user mode and kernel mode?
2. What is the difference between user mode and kernel mode?
3. Which layers are in the kernel mode?
4. What are the tasks for Executive Services in the Windows operating system?
5. What are the tasks for the kernel in the Windows operating system?
6. What are the tasks for the hardware layer (HAL) in the Windows operating system?
7. What are the tasks for the device drivers? Mention some types of device drivers.
8. Mention the tasks for the following system processes: Idle Process, Session Manager,
Winlogon and the Windows Subsystem.

360°
thinking .

360°
thinking . 360°
thinking .
Discover the truth at www.deloitte.ca/careers Dis

© Deloitte & Touche LLP and affiliated entities.

Discover the truth at www.deloitte.ca/careers © Deloitte & Touche LLP and affiliated entities.

Deloitte & Touche LLP and affiliated entities.

Discover the truth

37 at www.deloitte.ca/careers
Click on the ad to read more
Download free eBooks at bookboon.com

© Deloitte & Touche LLP and affiliated entities.

An Introduction to
Windows Operating System System Mechanisms

5 System Mechanisms
The Windows operating system has many mechanisms that are used by components in kernel mode.
These mechanisms are used by the Executive services, the kernel and the device drivers.

Some of these system mechanisms are interrupts, exceptions, object manager, synchronization, global
flags and wow64.

In the following, we will be looking at some of the most important system mechanisms.

5.1 Interrupts
Windows is an interactive operating system that waits for events to occur. Examples of such events are
key pressure, the movement of a mouse, and insertion of a CD or USB pen.

When such events occur, the processor must execute them. Moreover, this must happen immediately,
so the processor must interrupt what it is doing to handle the event.

Windows is using interrupts to carry out events as soon as they happen. This means that an event
interrupts the thread running in the processor, so that the processor can execute the event instead.

The advantage of using interrupts is that the processor is able to work undisturbed most of the time. The
processor only gets an interrupt when it is necessary to perform an event or a service.

An interrupt is an asynchronous event, which means that the event can occur at any time, and is
independent of what happens in the processor.

All hardware devices can use interrupts to receive services from the operating system. Not only hardware,
but also the software system can send interrupts.

Interrupt signals from the hardware mostly come from input/output devices that need to make the
processor aware that it must perform a service. However, almost all types of peripheral devices may use
interrupts as the primary method to get the operating system to perform actions.

An interrupt service routine is a routine that hardware calls on in response to an interrupt signal.
Windows has a table for interrupt signals, which locates the interrupt service, and is designed to handle
a specific interrupt signal.

38
Download free eBooks at bookboon.com
 An Introduction to
Windows Operating System System Mechanisms

5.2 Exceptions
Exceptions are mechanisms that allow applications to handle errors or unexpected events during
program execution.

Exceptions are similar to interrupts. However, unlike interrupts, which can take place at any time,
exceptions can only arise from program execution.

Exceptions are sent when an error occurs during program execution. Some examples of conditions that
can cause exceptions are division by zero, insufficient memory to perform an operation, stack faults,
page faults in memory and floating point errors.

Programmers can use exceptions in the applications code to handle errors or situations, where something
TMP not
does PRODUCTION NY026057B
go according to the plan. For example, what 4
happens if an application wants to12/13/2013
open a file and
6 xthe
4 file is not on the hard drive? If opening the file causes an exception, the program can switch to a
PSTANKIE ACCCTR00

code written
gl/rv/rv/baf to process the error that occurred. Bookboon Ad Creative

All rights reserved.

© 2013 Accenture.

Bring your talent and passion to a

global organization at the forefront of
business, technology and innovation.
Discover how great you can be.
Visit accenture.com/bookboon

39 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System System Mechanisms

try
{
StreamReader sr = new StreamReader(filnavn);
String line;

// Reads and displays lines from the file.

while ((line = sr.ReadLine()) != null)
Console.WriteLine(" " + line);
}
catch (Exception e)
{
// Gives a message about what went wrong.
Console.WriteLine("Couldn’t read the file: ");
Console.WriteLine(e.Message);
}

Figure 5.1: The program code above shows the use of an exception in a C# program.

If the program cannot open or read the file, an exception will occur and in this case a message will be
printed out on the screen.

Windows uses a system called structured exception handling, which gives applications the opportunity
to gain control of the situation when an error occurs. The program can then try to fix the error itself
with a code written in case an error should occur.

When an exception happens, the system will generally take care of (save) the current code to be executed,
and program execution switches over to a subroutine called an Exception Handler. Depending on the
situation, the interrupted code can start again when the Exception Handler is finished.

5.3 Object Manager

Object Manager is a part of the executive services in Windows, and its task is to manage Windows
resources. All resources are treated as objects, with some examples of these being processes, threads,
files, semaphores, drivers and events.

One can say that an object is a data structure in the kernel mode available for virtual memory. Each
object has a head and a body. The head contains information (Properties) used by the object manager,
while the body contains code and data.

Objects can be either kernel objects or executive objects. Moreover, applications running in user mode
can use executive objects through executive services.

Only the kernel use kernel objects. The kernel objects represent resources such as devices or services
such as synchronization.

40
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System System Mechanisms

The Object Manager takes care of the following tasks:

1. Create and return handles to objects.

2. Making multiple handles to an object if necessary.
3. To check if a process has the right to use an object.
4. Keeping track of which quotas belong to which resources.
5. Delete objects and close the handles to objects.

In total, the Object Manager takes care of the lifecycle of objects from creation to deletion. The
Object Manager is very important because it takes care of the management of system resources and
data structures.

The operating system uses objects to regulate access to the system for two reasons.

1. The use of objects ensures that Microsoft can update the system’s functionality. When new
versions of the operating system are made, the properties of objects can be changed without
causing major consequences for the system.
2. The use of objects increases the security of the Windows operating system. Each object has a
list that specifies what actions are possible on the object.

5.4 32-bit applications on a 64-bit operating system

The WoW64 subsystem is a layer in the Windows operating system that allows 32-bit applications to
run on a 64-bit system without having to make changes to the software:

• WoW64 stands for a Windows 32-bit on a Windows 64-bit.

WoW64 consists of DLLs that allows 32-bit applications to run on 64-bit versions of Windows. WoW64
is included with all 64-bit versions of Windows.

5.5 Control Questions

1. What is the advantage of using interrupts?
2. How does the Windows operating system use interrupts?
3. In what contexts does the Windows operating system use exceptions?
4. How can programmers use exceptions?
5. What is the role of the Object Manager?
6. What is the function of WoW64?

41
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Mechanisms for data managemen

6 Mechanisms for
data management
Windows has many system settings that must be handled and taken care of, and are stored in a database
called the registry. The Configuration Manager takes care of- and manages the registry.

6.1 The Registry

The registry is a hierarchical database that contains information about the various settings used by the
operating system. The registry contains system settings used at the Windows startup. An example of this
is user settings such as the colors and background on a desktop.

The registry plays an important role in setting up and controlling the Windows operating system. It
keeps track of system settings, user settings and what hardware and applications are on the computer.

The registry contains information about the applications installed on your PC, in addition to information
about the processor, memory, drivers, network and similar aspects.

The Wake
the only emission we want to leave behind

.QYURGGF'PIKPGU/GFKWOURGGF'PIKPGU6WTDQEJCTIGTU2TQRGNNGTU2TQRWNUKQP2CEMCIGU2TKOG5GTX

6JGFGUKIPQHGEQHTKGPFN[OCTKPGRQYGTCPFRTQRWNUKQPUQNWVKQPUKUETWEKCNHQT/#0&KGUGN6WTDQ
2QYGTEQORGVGPEKGUCTGQHHGTGFYKVJVJGYQTNFoUNCTIGUVGPIKPGRTQITCOOGsJCXKPIQWVRWVUURCPPKPI
HTQOVQM9RGTGPIKPG)GVWRHTQPV
(KPFQWVOQTGCVYYYOCPFKGUGNVWTDQEQO

42 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Mechanisms for data managemen

Windows has quite a bit of important information in the registry, including:

ƒƒ Settings that determine the layout of Windows.

ƒƒ What drivers that belong to different equipment.
ƒƒ What software to start at boot time.

Windows retrieves data in the registry under the following conditions:

1. When starting the computer (Boot Process).

2. When loading the kernel (Kernel Boot Process).
3. When a user logs onto the system.
4. At the startup of applications.

The registry is organized into separate volumes called Hives, with each Hive set in its own file in
the director:

C: \ Windows \ system32 \ config

The Configuration Manager takes care of- and manages the registry.

6.2 Configuration Manager

The Configuration Manager is an important component of the Windows executive services, which take
care of- and manage the registry.

We have seen that data is stored in registry files called hives, and the Configuration Manager reads data
from these files.

The configuration to a computer will change over time. The Configuration Manager updates the registry
as programs and other components of the operating system change settings​​​​.

At startup, the Configuration Manager provides the ability for computer to receive this new configuration.

6.3 Control Questions

1. What is the role of the registry in the Windows operating system?
2. What kind of data is stored in the registry?
3. In what situations do Windows retrieve data from the registry?
4. What is the task of the Configuration Manager?

43
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Processes, Threads and Jobs

7 Processes, Threads and Jobs

A process, thread and job are key concepts associated with programs running on a computer. We will
therefore begin with a definition of these terms.

Process

When we start a program and it is loaded into the computer memory we have a process, which is a
program that runs on the computer.

Thread

A thread is a code to execute in the processor. A thread belongs to a process and is called the dynamic
part of a process.

Fiber

A fiber is a kind of thread, and is often called a lightweight tread. A fiber is scheduled in a different way
than ordinary threads.

Job

Windows can group processes that cooperate, with such a group of processes called a job.

7.1 About processes

A computer program is one or more files on a hard drive containing code and program resources. When
a program is started and is loaded into a computer’s memory, we have a process.

A process is thus a program running on a computer, which can give rise to multiple processes on a
computer if someone starts the program several times at the same time.

A process is work to do on a computer. A processor runs a process by executing the process’s code. On
a computer, there are often many processes, but only one or a few processors. Consequently, if there are
more processes than processors, they must execute in the processor or the processors in turn.

44
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Processes, Threads and Jobs

A process consists of two parts:

• The static part

• The dynamic part

The static part of a process is resources on your computer that the process uses. The static part of a
process can consist of a piece of memory that belongs to the process, a directory, an open file and other
related resources.

The dynamic part of a process consists of one or several threads. The dynamic part of a process is a
program code that runs on your computer, i.e. as a set of instructions executed one after another in
the processor.

30 FR
da EE
SMS from your computer ys
...Sync'd with your Android phone & number tria
l!

Go to

BrowserTexting.com

and start texting from

your computer!

... BrowserTexting

45 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Processes, Threads and Jobs

Figure 7.1: Task Manager in Windows shows the processes running on your computer

In the Windows operating system, a process is defined in the following way:

• A process is a container in the memory with the resources required to run a program on
the computer.

A Windows process consists of:

1. An address space consisting of a set of virtual addresses.

2. An executable program.
3. At least one line of code to execute in the processor.
4. A list of handles to various system resources.
5. A unique process ID used to identify the process.
6. A security structure that regulates access to the process.

46
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Processes, Threads and Jobs

For each process that is running, the operating system has a structure called a Process Descriptor, which
includes the following information:

1. The name of the process.

2. The start time for the process.
3. The status of the process.
4. A list of threads related to the process.
5. A description of addresses in the memory.
6. A list of resources that belong to the process.

The Process Manager takes care of processes, threads and the resources that belong to the processes. The
Process Manager therefore controls the activities of the process and the resources used by the process.

7.2 About threads

A thread comprises some code to execute in the processor. A process can create one or more threads, and
if a process has multiple threads, the threads will run independently and compete for time in the processor.

A process that has only one thread is a single-threaded process, while a process with multiple threads
is called a multithreaded process. The advantage of multithreaded processes is that they can perform
different tasks simultaneously.

Why use multiple threads?

One reason to use multiple threads in an application is that several activities are going on simultaneously. One of these
activities can sometimes stop, for example, waiting for input from the user. It may then be fine if other activities can
continue to work independently, which they can do, if they are separate threads.

Using multiple threads is particularly beneficial on computers with multiple processors, as the threads
run in parallel for increased speed.

Since threads are competing for processor time, each thread has a status. The status is either active or
inactive, meaning that a thread is either active in the processor or on a standby status.

State Description

Blocked The thread is in a wait state.

Ready The thread is ready to be loaded into the processor.

Active The thread is executed in the processor.

Figure 7.2: The three possible states for a thread

47
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Processes, Threads and Jobs

In Windows, a thread is a process object to which the operating system is allocating processor time.
The process itself is a primary thread, which can launch other threads. The process has a list with the
threads that belongs to the process.

A thread has a Thread Descriptor in the computer’s memory with information about the thread. A
Thread Descriptor is a data structure that the operating system uses to obtain information that it needs
to process a thread, and has the following information about the thread:

1. A description of the thread.

2. The thread status.
3. The start time for the thread.
4. A reference to the Process Descriptor.
5. A list of threads associated with the thread.
6. A reference to the thread’s resources.

All threads in a process have common access to the resources belonging to the process. Most of the resources
used by a thread are allocated to the process, though resources can be allocated directly to a thread.

Brain power By 2020, wind could provide one-tenth of our planet’s

electricity needs. Already today, SKF’s innovative know-
how is crucial to running a large proportion of the
world’s wind turbines.
Up to 25 % of the generating costs relate to mainte-
nance. These can be reduced dramatically thanks to our
systems for on-line condition monitoring and automatic
lubrication. We help make it more economical to create
cleaner, cheaper energy out of thin air.
By sharing our experience, expertise, and creativity,
industries can boost performance beyond expectations.
Therefore we need the best employees who can
meet this challenge!

The Power of Knowledge Engineering

Plug into The Power of Knowledge Engineering.

Visit us at www.skf.com/knowledge

48 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Processes, Threads and Jobs

There are usually many threads running simultaneously on a computer, and the operating system must
therefore allocate time for them in the processors. A processor is running thread after thread, and thus
shifts constantly between running threads.

Context Switching is the procedure to remove a thread from the processor and to load in a new thread.
When Context Switching has removed a thread from the processor, the thread is on standby status. All
information about the thread is then stored so that the thread can continue to run in exactly the same
way the next time it comes into the processor.

In general, several threads compete to get into the processor. Scheduling determines the order in which
the threads will enter the processor, so the main task of a scheduler is to allocate time in the processor
for threads in the best possible way.

One way to organize scheduling is to give threads different priorities, as threads with a high priority
have easier access to the processor than those with a low priority.

7.3 About fibers

Fibers are quite new in the Windows operating system, and were introduced in Windows Vista. A fiber
is code to execute in the processor like a thread. A fiber is a sub-thread that belongs to a conventional
thread, and fibers are often called lightweight threads.

The operating system does not control process planning for fibers as it does for threads. A fiber belongs
to a thread, which takes care of the process planning for the fiber. A thread can schedule multiple fibers,
and can also make a switch between fibers in the processor.

Fibers require less support from the operating system than threads, and in contrast to threads, fibers
may be using cooperative multitasking. Cooperative multitasking means that a thread is in the processor
until it voluntarily relinquishes the place.

Process

Thread Thread

Fiber Fiber Fiber Fiber

Figure 7.3: The relationship between processes, threads and fibers

49
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Processes, Threads and Jobs

An advantage of fibers is that it is easy to change from one fiber to another in a processor. To switch
between fibers only requires storing and the reading of some registers, while to switch between threads
requires them to go out of- or into the kernel.

One drawback with fibers is that they cannot utilize multiple processors without using the operating
system’s controlled threads.

7.4 About jobs

It is possible to group processes so that they can work together. In Windows, such a group of processes
is called a job. A process can only be in one job at a time.

To group processes in a job is only applicable when the processes need to cooperate.

In the Windows operating system a job is an object. The function of a job object is to handle groups
of processes as a unit. A job object is a kernel object with a name, and which possesses a high security.

> Apply now

redefine your future

AxA globAl grAduAte
progrAm 2015
- © Photononstop

axa_ad_grad_prog_170x115.indd 1 19/12/13 16:36

50 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Processes, Threads and Jobs

7.5 The Process Manager

A Process Manager manages processes and threads, and takes care of the following tasks:

ƒƒ Starting and ending processes and threads.

ƒƒ Provide parameters and characteristics of processes and threads.
ƒƒ Allocation of resources to processes and threads.
ƒƒ Implementation of input and output.
ƒƒ Implementation of addresses in the memory.

The Process Manager does not do operations on processes and threads such as scheduling and
synchronization, instead being taken care of by the kernel.

7.6 Multitasking
Previously, a PC just had one processor, which could only run one thread at a time. This is sufficient if
there is just one program running on your computer, as it was for the first PCs. However, when graphical
operating systems were developed, it became possible to run multiple programs in different windows.
It was then necessary for computers to be able to run multiple programs simultaneously. As a result, an
operating system like Windows thus has a need for multitasking.

Windows supports multitasking, which means that two or more threads can run simultaneously on one
processor while taking turns in the processor.

Multitasking allows a user to run multiple applications simultaneously and to switch between applications.
For instance, while simultaneously working with a word processor and a spreadsheet, you may transfer
the data from the spreadsheet to the editor by cutting and pasting.

A multitasking operating system gives processor time to the threads that need it. Multitasking is done
by Context Switching. In using Context Switching, a thread runs in the processor until the operating
system interrupts it or the thread voluntarily leaves the processor.

We distinguish between two types of multitasking:

ƒƒ Cooperative multitasking.
ƒƒ Preemptive multitasking.

In cooperative multitasking, programs share CPU time between them. A thread holds the processor
until it voluntarily gives it away.

Preemptive multitasking ensures that the various processes only run for a short time in the processor. A
thread can only be in the processor for a time interval called a quantum. When a quantum is finished,
the operating will automatically do a context switch.

51
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Processes, Threads and Jobs

7.7 Communication between processes

Inter-process communication is when a thread in a process interacts with a thread in another process.
This applies when:

1. The processes are on the same computer.

2. The processes are on different computers.

Communication between processes on the same computer is quite straightforward. The operating system
can copy data from the memory to the sending thread to the memory of the receiving thread.

Communication between processes in different computers is as follows. The operating system on the
computer that sends will copy the data from its memory to a communication unit. The communication
unit will then send the data to the receiving computer.

Windows operating system have several mechanisms to support communication between processes:

Clipboard

A Clipboard is widely used to transfer data from one application to another. When a user copies or cuts
data from an application, the data can be transferred to another application by pasting it.

File Mapping

File Mapping is a way for a process to manage the content of a file as if it were in the memory of the
process. The process can use a pointer to retrieve data from the file and modify the contents of the
file. File Mapping is a simple way for two or more processes to share data, but the method requires
synchronization.

Pipes

Pipes are a simple model for process communication. One thread writes data to a buffer, so that another
thread can read the data from the buffer. We distinguish between two types of pipes: Anonymous Pipes
and Named Pipes.

Moreover, Anonymous Pipes can be used to transfer data between threads in the same process.

Named Pipes can transfer data from one application to another application on a computer, as well as
transfer data from one computer to another computer over a network. Named Pipes has a name that
will allow other processes to identify them. In Windows, named pipes are based on a client-server
communication similar to that used by Sockets.

52
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Processes, Threads and Jobs

7.8 Control Questions

1. What is the static part, and what is the dynamic part of a process?
2. What is the definition of a process in the Windows operating system?
3. Give a definition of a thread.
4. Under what conditions can a program benefit from using multiple threads?
5. What is Context Switching?
6. What is the difference between a thread and a fiber?
7. How is scheduling done for threads?
8. How is scheduling done for fibers?
9. When may it be appropriate to group processes in a job?
10. What is multitasking?
11. What is the difference between cooperative and preemptive multitasking?
12. What is the difference between Anonymous Pipes and Named Pipes?

LIGS University
based in Hawaii, USA
is currently enrolling in the
Interactive Online BBA, MBA, MSc,
DBA and PhD programs:

▶▶ enroll by October 31st, 2014 and

▶▶ save up to 11% on the tuition!
▶▶ pay in 10 installments / 2 years
▶▶ Interactive Online education
▶▶ visit www.ligsuniversity.com to
find out more!

Note: LIGS University is not accredited by any

nationally recognized accrediting agency listed
by the US Secretary of Education.
More info here.

53 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Synchronization of threads

8 Synchronization of threads
There are often multiple threads simultaneously running on a computer. It may then be necessary to
synchronize the threads to avoid collisions.

There is a need for the synchronization of threads in two situations:

1. Threads that cooperate.

2. Threads competing for a resource.

8.1 Threads cooperating

Threads can cooperate in various situations. For example, a thread can be responsible to perform a task
for another thread.

Example of thread cooperation

Two threads working towards a common data structure, such as a list, is an example of cooperation between threads.
One thread can have the task to create and hang items on the list, while the other thread can remove items from the
list and use them. The second thread is then dependent on the first thread.

If all the threads in a computer system were running independently without having anything to do with
other threads, it would not be difficult to allocate time to them in the processor. What makes processing
difficult is that the threads can cooperate and rely on each other. The operating system must therefore
be able to manage cooperation between threads.

8.2 Threads competing

A competition between threads occurs when two threads want to use the same resource at the same
time. An example of this is two threads that write to the screen simultaneously, which could lead to
collisions and messy prints.

Two threads writing to the screen at the same time may provide collisions. For example:

ƒƒ Thread 1 writes Hello.

ƒƒ Thread 2 writes Hi.

A result on the screen could be HeHillo. However, in practice, the printing will be a lot messier. Collisions
can cause the positions of the letters to crash, and the letters will then be written out at random locations
on the screen.

54
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Synchronization of threads

Many resources can only use one thread at a time. If two threads try to use such a resource at the same
time there will be a collision. Hence, to avoid collisions, an operating system must be able to manage
competition between threads.

8.3 Critical region

The term critical region is important in connection to the synchronization of threads. A critical region
is a place where collisions may occur. In traffic, an intersection can be a critical region because cars can
easily collide there. Therefore, traffic lights, roundabouts and traffic rules control intersections.

A critical region in conjunction with an operating system is a piece of a program code that attempts to
obtain a resource that cannot be shared. If different threads use the same resource at the same time, this
can lead to collisions with the result that an application does not function normally.

A file is an example of a resource that cannot be shared. If two threads write to the same file at the same
time, the contents of the file will become an unreadable mix of characters.

Another example of a resource that cannot be shared is a printer. If two processes are printing their
documents at the same printer at the same time, the printout will probably not be readable.

In addition, variables are resources that cannot be shared. If two threads attempt to change the value
of a shared variable at the same time, context switching may lead to the loss of one of the updates. The
variable will then not have the correct value afterwards.

Example of resources simultaneously used by multiple threads are hardware devices, memory blocks,
and files and shared variables. A thread that will use such a resource should have exclusive rights to the
resource when it is used.

8.4 Mutual Exclusion

Mutual Exclusion is an important concept in connection to synchronization. That threads can get mutual
exclusion to a resource is necessary for a computer system to function properly.

We have seen that if two threads access a resource at the same time, it may lead to collisions. It is therefore
important that only one thread at a time can access a resource that cannot be shared. If only one thread
at a point of time has access to a resource so that other threads cannot access, the thread has Mutual
Exclusion to the resource.

55
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Synchronization of threads

Having Mutual Exclusion to a resource is required when the resource cannot be shared, because sharing
the resource will lead to unpredictable results. Consequently, operating systems must use different
mechanisms to ensure that threads have a mutual exclusion.

In the following, we shall look at some commonly used mechanisms used for synchronizing threads.

8.5 Semaphore
One way to control access to a resource is to use a semaphore. A semaphore may be described as an
integer, and can have the value 0, 1, 2 …, n, where n is a number that indicates how many threads are
allowed to access the resource at the same time.

For instance, two threads that are simultaneously writing to the screen can use a semaphore to avoid
collisions. The semaphore can either have the value zero or the value one.

Value semaphore State Consequence

0 The screen is in use. No more threads can access.

1 The screen is not in use. One thread can access.

Figure 8.1: A semaphore can provide mutual exclusion to a resource

Before a thread writes to the screen, it checks the semaphore. If the value of the semaphore is one, the
value is set to zero and the thread has mutual exclusion to the screen. When the thread is finished using
the screen the value of the semaphore is set to one again, so that other threads can then access the screen:

MySemaphore.WaitOne();
// Now, only this thread has access to the screen.

Console.Write(“Message from the thread: {0}”, message);

MySemaphore.Release();
// Releases the semaphore for other treads to use the screen.

Figure 8.2: The above code shows the use of a semaphore in a C# program. Printing to the screen using Write() is a
critical region. Therefore, a semaphore is used to ensure that only one thread is allowed to use the screen at any given
point in time.

A semaphore can have more values ​​than only zero and one. It may sometimes be possible that multiple
threads can simultaneously access a resource. You can then use a semaphore with several values.

For example, a semaphore with a numeric value of ​​0, 1, and 2 will be able to allow two threads
simultaneously. The semaphore can start with a value of 2, and each thread that then enters the semaphore
decreases the value by 1. When the value is 0, then no more threads can access the resource.

56
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Synchronization of threads

Semaphores can regulate access to resources, but there are some disadvantages to using semaphores that
you must bear in mind:

1. Threads are not forced to use semaphores. Some threads may not use a semaphore when it
is necessary.
2. If a thread with a low priority that is starved gains access to a semaphore, this will lock the
system. This is because other threads with a higher priority are unable to access the semaphore.
3. The incorrect use of a semaphore can lead to deadlock.

Many programming languages ​​support semaphores. For this reason, semaphores are still widely used
by programmers.

8.6 Mutex
Mutex is an abbreviation for Mutual Exclusion. A Mutex is a synchronization mechanism made for cases
when you want the mutual exclusion of a resource, in addition to being a kind of semaphore that only
can have the values zero and one.

Semaphores that have more values than zero and one do not provide the mutual exclusion of a resource
because multiple threads can simultaneously access the resource. For a semaphore to provide mutual
exclusion, it must only allow the values zero and one.

57 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Synchronization of threads

A difference between a Mutex and a semaphore is that semaphores are dependent on the operating
system to function. A Mutex may be implemented by more specialized and faster procedures. It is
therefore advisable for programmers to use a Mutex instead of a semaphore when the goal is to obtain
mutual exclusion.

8.7 Some synchronization mechanisms in Windows

There are many synchronization mechanisms other than semaphore and Mutex. In the following, we
will review some synchronization mechanisms used in the Windows operating system:

ƒƒ Keyed Events
ƒƒ Spinlocks
ƒƒ Fast Mutex and Guarded Mutex
ƒƒ Pushlocks

Keyed Events

Event objects are a type of synchronization mechanism used in cases in which threads cooperate. An
event object is used when a thread must wait until an event has occurred before it can perform some
task. This event may be that another thread has finished a job.

An example of the use of Events

Given two threads and a file, one thread writes to the file while another thread reads the file. This gives an example of
when the two threads must wait for each other. Thread 2 cannot read the file before Thread 1 has written to it. Moreover,
Thread 1 cannot write to the file again before Thread 2 has read it.

An event object can be used to synchronize reading and writing to the file.When Thread 1 has written to the file, it
updates the event object, and before reading the file, Thread 2 then can check the event object to be certain whether
it is written to the file or not. When Thread 2 has read the file, it updates the event object.

Thread 1 and Thread 2 can constantly keep abreast of the read/write status of the file using the event object.

Windows uses a synchronization mechanism called Keyed Events, which is an object that allows a thread
to specify a key for what event it is waiting for. The thread will wait until another thread signals that an
event with the same key has occurred.

Spinlocks

Spinlock is a locking mechanism used by the kernel in a Windows operating system to achieve the
mutual exclusion of a resource. Before a thread in the kernel can access a critical resource, it must ask
a spinlock for access. If the spinlock is not free, the thread will keep trying to access the spinlock until
it becomes available.

58
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Synchronization of threads

The thread is waiting in a loop (spinning) until the spinlock becomes available. The thread will then be in
a wait state (not active), while waiting for the spinlock to become available. Once a thread has access to
a spinlock, it will have the mutual exclusion of the resource until it lets go of it again. To use a spinlock
is therefore only effective if threads only have the spinlock for a short period.

Queued Spinlocks are also used. A Queued Spinlock has a queue, and when a thread attempts to access
a queued spinlock that is busy, the thread is placed in a queue belonging to the spinlock. When the
spinlock becomes available, the thread that is first the in queue is the next to access the critical resource.

Fast Mutex and Guarded Mutex

The Windows operating system uses two types of Mutex, which are called Fast Mutex and Guarded Mutex.

Fast Mutex provides better performance than a common Mutex. The reason for this is that it can
bypass normal process planning, which provides Fast Mutex with a particularly good performance in
environments with multiple processors. Fast Mutex is widely used in communication with drivers.

Guarded Mutex is used by Windows Server 2003 and later versions of Windows, and is used for the
same tasks as Fast Mutex, but is even faster.

59 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Synchronization of threads

Pushlocks

A pushlock is a synchronization mechanism that was first introduced in Windows Server 2003. Pushlocks
are used instead of spinlocks to protect data structures in the kernel, with one advantage of pushlocks
being that they use little space.

Access to a pushlock can be both exclusive and shared, and can be shared if data is to be read. If a thread
shall be able to change the content of the data, access to the data must be exclusive. If a thread has
exclusive access to a pushlock, other threads trying to access the pushlock will be put on a waiting list.

Pushlocks is only used by Windows operating systems, but there are also some drivers who use pushlocks.

8.8 Control Questions

1. Explain why there is a need for the synchronizing of threads.
2. What is a critical region in conjunction to the theory of synchronization?
3. Provide some examples of resources that only can be accessed by one thread at a time.
4. What is mutual exclusion in connection with the synchronization of threads?
5. Explain the behavior of a semaphore.
6. What is the difference between a semaphore and a Mutex?
7. In what situations is the synchronization mechanism Keyed Events used?
8. Explain the behavior of a spinlock.

DO YOU WANT TO KNOW:

What your staff really want?

The top issues troubling them?

How to make staff assessments

work for you & them, painlessly?

How to retain your

top staff Get your free trial
FIND OUT NOW FOR FREE
Because happy staff get more done

60 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Process planning on one processo

9 Process planning on
one processor
There are usually more threads running on a computer than there are processors. It is therefore necessary
to organize how threads shall use the processor or the processors.

Planning how threads should use one or several processors does not differ from other planning, such as
planning the use of classrooms at a school or planning a traffic flow using traffic rules.

In this section, we will look at process planning for multiple threads on one processor, and we will return
to process planning on multiple processors in the next chapter.

Processor scheduling is the task of managing the sharing of a processor between different groups of
threads. The scheduler that is a part of the operating system is taking care of process planning.

There are usually several threads on a computer waiting to execute in the processor. When a thread has
finished processing and a new thread is to be loaded, the process manager must decide which thread
is to be next.

A policy for process planning determines which thread is the next to be loaded into the processor, and
must determine two things:

ƒƒ When the time has come for a thread to leave the processor.
ƒƒ Which thread is the next to be loaded into the processor?

There are four reasons why a thread leaves the processor:

1. The thread is finished, for example at the program’s end.

2. The thread needs a resource to continue, but the resource is currently not available.
3. The thread decides to voluntarily leave the processor.
4. The thread must leave the processor, because the time interval it can have in the processor
is finished.

Process planning can have a huge impact on computer performance since it determines when a thread
is to be loaded into the processor. Ignoring a thread and never loading it into the processor will lead to
the starvation of the tread.

61
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Process planning on one processo

There are many ways to do scheduling for threads, as how to organize scheduling is a classical research
problem for operating systems. Modern algorithms for process planning tend to use priority, i.e. the
different threads have different priorities, and threads of higher priority enter the processor more easily.

The importance of process planning

Process planning is much more important on a server in a network than for a single PC, as there are often many processes
on a server competing to get into the processor.

Process planning is not so important on individual PCs. A person using a PC often works with only one application at a
time, such as a word processor or a web browser. Today’s PCs are so fast that there is no problem to run a few processes.

9.1 Voluntary and involuntary sharing of the processor

Threads can share the use of a processor in two ways, through:

ƒƒ Voluntary sharing.
ƒƒ Involuntary sharing.

Voluntary sharing of the processor means that threads leave the processor by themselves, as the voluntary
sharing of the processor is the same as cooperative multitasking. One problem with using the method
of voluntarily sharing the processor is that there may be threads that are not programmed to cooperate.

Involuntary sharing of the processor means that there is an interruption system that can interrupt a
thread running in the processor. This system employs a timer that interrupts the thread when a certain
time interval has passed.

The Processor Manager determines which threads get to enter the processor, as well determining
how long the threads will be in the processor. How the Processor Manager is organized has a major
impact on the computer’s speed and performance, as threads that readily enter the processor will have
a high performance. Conversely, a thread will have a low performance if neglected, and rarely enters
the processor.

Process scheduling algorithms designed for modern operating systems use priority to determine which
processes get to access the processor. Each thread has a priority that indicates how important the thread is.

One can distinguish between two types of algorithms for process planning:

ƒƒ Non-preemptive algorithms.
ƒƒ Preemptive algorithms.

62
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Process planning on one processo

Non-preemptive algorithms allow a process that has come into the processor to run until it is finished
or until it leaves the processor by itself.

Preemptive algorithms allow processes to run for only a specific time interval, and threads are allowed
to be executed in the processor for just a specific time interval.

9.2 Non-preemptive process planning

Some theoretical ways an operating system using a non-preemptive process planning may work:

1. The job that comes first is first performed.

2. The smallest job first.
3. Priority process scheduling.
4. Deadline process planning.

The job that comes first is first performed strategy allows threads to come into the processor as they are
started. One way to organize this is to have a list with the threads that are ready to go into the processor.
The threads are then placed back on the list as they come. Whoever is first on the list then comes first
into the processor, but since this strategy does not work well in practice it is rarely used.

The smallest job first strategy is that the thread that is fastest to the finish comes first into the processor.
This strategy reduces the waiting time because the small threads are running away fast. The disadvantage
of this strategy is that it can lead to the starvation of large threads if there are too many small threads.

Priority process scheduling means that threads are chosen to run in the processor by priority, as a process
with a high priority will more easily enter the processor than a process with a low priority. The disadvantage
of this method is that it can lead to the starvation of threads with low priority.

In real-time systems, it is often necessary that some threads have to run in the processor in order to keep
a deadline. System performance depends on such deadlines, which is called deadline process planning.

9.3 Preemptive process planning

Some ways that an operating system using a preemptive process planning can work:

1. Round Robin.
2. Queues with different priorities.
3. Foreground and background processes.

63
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Process planning on one processo

Of all the scheduling algorithms, Round Robin is the most widely used. The threads are running in the
processor in turn, with each thread processing a specific time interval called a Quantum. All threads
are treated equally without priority, as the idea behind Round Robin is that all threads will get equal
time in the processor.

Queues with different priorities comprise a system in which there is a queue for each priority. Threads
with the same priority are hung on the same queue, so queues with a different priority are thus an
extension of the priority system. There are many different strategies for giving threads on all queues
time in the processor.

Foreground and background processes are one way to organize threads. Threads in the foreground are
running on the processor to perform tasks for a user, whereas threads in the background get into the
processor when there is an opportunity, i.e. when there are no active threads in the foreground.

9.4 Scheduling for various operating systems

How to organize process planning depends of the kind of operating system being used. For example,
real-time processes in a real-time operating system must respond immediately. Real-time processes must
therefore have a much higher priority than usual processes.

Challenge the way we run

EXPERIENCE THE POWER OF

FULL ENGAGEMENT…

RUN FASTER.
RUN LONGER.. READ MORE & PRE-ORDER TODAY
RUN EASIER… WWW.GAITEYE.COM

1349906_A6_4+0.indd 1 22-08-2014 12:56:57

64 Click on the ad to read more
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Process planning on one processo

What scheduling algorithm is best depends on the type of operating system in use, and we will look at
three different systems:

ƒƒ Batch System
ƒƒ Interactive System
ƒƒ Real-time System

Batch systems perform a task described in a file. The first computers used batch systems, and batch
systems are still used today in the business world.

In batch systems, no user is waiting for a response, so there is no need for preemptive algorithms. A
batch process is running in the processor until it is finished.

In an interactive system, a user interacts with a computer’s operating system. This gives a need for
preemptive algorithms, which use context switching. This replacement of threads makes it possible to
work with several applications at once, and ensures that your computer does not halt if an error occurs
in a process.

Real-time systems often have sophisticated algorithms for process planning, involving processes that
must respond immediately to have a high priority. Moreover, these processes must always be ready to
respond to external events within the system.

The following gives the aim of process planning for these three types of operating systems:

1. The aim for the process scheduling of algorithms for batch systems is to perform as many jobs
as possible in the shortest possible time, i.e. to keep the processor active at all times.
2. The aim for the process scheduling of algorithms for interactive systems is to be able to react
to events as quickly as possible. It must be able to satisfy user’s expectations of the system.
3. The aim for the process scheduling of algorithms for real-time systems is to be able to react to
events immediately. The system must be predictable, and things have to happen at the right time.

The aim for the process scheduling of algorithms for all systems is to provide all the processes access to
the processor, i.e. to exploit the system in the best possible way.

65
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Process planning on one processo

9.5 Context Switching

Context Switching is the process to remove a thread from a processor and to load in a new thread. The
most common causes of a context switching is:

ƒƒ The time for a thread to execute in the processor is finished.

ƒƒ A thread with a higher priority has become ready to run.
ƒƒ A thread running has to wait for a resource. When a thread has to wait, it is taken out of
the processor.

A thread in the processor uses multiple registers with data and information about the thread. When
context switching happens, data in these registers are saved, so that they can be loaded into the processor
the next time the thread execution continues.

The Process Manager has a queue for each priority with threads that are next to execute in the processor.

The following occurs when a thread leaves the processor:

1. The data for the thread leaving is stored.

2. The thread leaving is put at the end of the queue in its priority class.
3. A new thread in the queue with the highest priority is looked up.
4. The new thread is removed from the queue and is loaded into the processor.

9.6 About the length of a Quantum

Preemptive algorithms ensure that each thread gets a moment in a processor called a quantum. Once a
quantum is finished, the system interrupts the thread running and another thread with the same priority
is loaded into the processor.

However, the fact that a thread must leave the processor when a quantum is over is not always
unconditional. If there is not another thread with the same priority, Windows will allow a thread to
run a quantum more.

How long a quantum is varies with the type of operating system.

In Windows 8, a thread is about two clock intervals in processor, whereas in Windows Server, a thread
is about 12 clock intervals in the processor.

The length of a clock interval varies with the type of hardware platform. For an x86 single processor,
a clock interval is approximately 10 milliseconds. For x86 and x64 multi-processors, a clock interval is
approximately 15 milliseconds.

66
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Process planning on one processo

9.7 About scheduling in Windows

MS-DOS and early versions of Windows did not have multitasking, and therefore did not have a process
manager. Windows 95, 98 and Millennium used non-preemptive process planning; hence, processes
would then run in the processor until they were finished.

Today, Windows is a preemptive multitasking operating system, meaning that the operating system can
carry out a context switching to satisfy a process scheduling policy. Preemptive process planning makes it
possible that a process with a high priority takes a processor from a process that is active in the processor.

Operating systems using preemptive multitasking guarantees that each thread gets some time in the
processor, thereby making the preemptive multitasking reliable.

The Process Manager manages multitasking by determining which of the competing threads will be the
next to execute in the processor, and the Process Manager in Windows uses priority to determine this.

The priority of each thread is determined by the following:

ƒƒ The priority level of the thread.

ƒƒ The priority class of the process.

This e-book
is made with SETASIGN
SetaPDF

PDF components for PHP developers

www.setasign.com

67 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Process planning on one processo

Threads are scheduled by priority, with each thread having a priority that varies between 0 and 31. The
priorities from 0 to 15 are normal priorities, while priorities from 16 to 31 are real-time priorities with
respect to threads that must react quickly.

Processes also have priority, and a process belongs to one of the following priority classes:

IDLE_PRIORITY_CLASS
BELOW_NORMAL_PRIORITY_CLASS
NORMAL_PRIORITY_CLASS
ABOVE_NORMAL_PRIORITY_CLASS
HIGH_PRIORITY_CLASS
REALTIME_PRIORITY_CLASS

A process will normally belong to NORMAL_PRIORITY_CLASS. Screensavers are an example of threads

that can have an IDLE_PRIORITY_CLASS.

A HIGH_PRIORITY_CLASS and REALTIME_PRIORITY_CLASS should be used with caution, as using

them can cause other threads to not get enough time in the processor.

9.8 Control Questions

1. What is processor scheduling?
2. What is the difference between non-preemptive algorithms and preemptive algorithms for
process planning?
3. How does Round Robin work?
4. What is the difference between the batch system, the interactive system and the
real-time system?
5. Give some reasons why the operating system performs context switching.
6. How long is a Quantum?
7. How is process planning done in the Windows operating system?

68
Download free eBooks at bookboon.com
 An Introduction to
Windows Operating System Process planning on multiple processor

10 Process planning on
multiple processors
Previously, all PCs had just one processor, called the Central Processing Unit (CPU). In recent years, it
has been common to have multiple processors in a personal computer, which run in parallel and increase
the speed of the system.

About computer speed

In the history of the PC, the speed of computers has steadily increased every year. The usual way to speed up has been
to increase the computer’s clock speed. However, there is a limit to how fast signals can be transmitted in a computer
system. Moreover, the faster the signal transmit, the more heat it develops, thus leading to overheating. The new trend
for increasing a computer’s speed is to therefore use multiple processors.

360°
.
Multiple processors provide particularly good performance when the number of threads is the same as

thinking
the number of processors. Performance also improves when there are more threads than processors, but
not to the same extent as when each thread has a processor to them self.

360°
thinking . 360°
thinking .
Discover the truth at www.deloitte.ca/careers Dis

© Deloitte & Touche LLP and affiliated entities.

Discover the truth at www.deloitte.ca/careers © Deloitte & Touche LLP and affiliated entities.

Deloitte & Touche LLP and affiliated entities.

Discover the truth

69 at www.deloitte.ca/careers
Click on the ad to read more
Download free eBooks at bookboon.com

© Deloitte & Touche LLP and affiliated entities.

An Introduction to
Windows Operating System Process planning on multiple processor

10.1 Multi-core processors

The core is the part of a processor that reads and executes instructions. Previously, processors only had
one core, while today’s new PCs have one processor with multiple cores.

A multi-core processor is a processing system with two or more cores that has an integrated circuit with
two or more individual processors, which in this context are called cores. A dual-core processor has two
cores, while a quad core processor has four cores. A multi-core processor implements multiprocessing
in a single physical package.

Cores connect to a multi-core device either tightly or loosely, e.g. cores can share or not share cache
and memory.

Tightly coupled multiprocessors also have multiple processors that connect at the bus level.

Loosely coupled multiprocessors form a cluster that consists of one or two multiple individual processors
connected by a high-speed communication system.

It is also possible to make clusters of computers to acquire more processor power. A computer cluster
consists of loosely interconnected computers that work together, so that they can be considered a
single system.

10.2 The organization of multiple processors

Multiple processors in computers make new demands of operating systems. Multiprocessor operating
systems are mostly like single processor operating systems, but must have enhanced features in terms
of process scheduling, synchronization and the management of resources.

We will now look at some ways to organize the use of multiple processors.

Each processor has its own operating system

One way to organize the use of multiple processors is that each processor has its own operating system,
which helps to avoid excessive collisions. Each processor does not have its own entire operating system,
but does have a copy of the data structures to the operating system, while sharing the operating system
code. The collaboration of different processors via a shared memory, is hence an advantage.

This model, in which each processor has its own operating system, was used in the past, though it is
rare today because it is a poor use of system resources.

70
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Process planning on multiple processor

Master-Slave model

Another way to organize operating systems with multiple processors is the Master-Slave model. In this
model, the operating system is running in one of the processors (Master), whereas the user applications
are running in the other processors (Slaves). This method provides an easy way to organize the use of
multiple processors in a computer.

Master Slave Slave Slave

Figure 10.1: The Master-Slave model with four processors

The downside of the Master-Slave model is that all slave processors must communicate with the operating
system in the master processor. The master processor therefore becomes a bottleneck, and all processors
are dependent on the master processor. The consequence of this is that processes in the slave processors
must often wait for the master processor.

Equivalent processors

Another way to organize multiprocessor operating systems is to have the operating system into memory
and allow any processor to execute threads. All processors are equal and can run both operating system
threads and threads that belongs to applications. Threads are just loaded into the processor that is
first available.

This model is called symmetric multiprocessing or SMP shortened. Using SMP, there are no overload
of one processor like the Master-Slave model.

10.3 Symmetric Multiprocessing

SMP (Symmetric Multi-Processing) is the computer architecture that modern operating systems use.
The SMP processors all share one operating system and a shared memory. SMP allows each processor
to run any thread. SMP can easily move tasks between processors so that tasks evenly balance between
the processors.

Processor Processor Processor Processor

Figure 10.2: With SMP, each processor run threads from both user applications and threads belonging to the
operating system

71
Download free eBooks at bookboon.com
 An Introduction to
Windows Operating System Process planning on multiple processor

A problem with SMP is that there may be collisions if two processors try to access the same resource
at the same time. One way to get out of this problem is to use a Mutex so that two processors cannot
use the same resource at the same time. In this way, just one processor can use the operating system at
a time. This works, but the problem is an inefficient system because the processors must wait for each
other, and then of course some of the intent with multiple processors falls away.

However, a Mutex can improve a system that regulates access to the operating system since many parts of
the operating system are independent of each other. For example, one processor can process scheduling,
while another processor processes the file system. A Mutex controls each critical region in the operating
system so that only one processor can use it at a time.

Most modern operating systems use SMP, but this needs to be done with care to avoid deadlocks.
TMP PRODUCTION NY026057B 4 12/13/2013
10.4 Multiple processors and synchronization
6x4 PSTANKIE ACCCTR00
Multiple processors working simultaneously increase the need for synchronization. TheBookboon
gl/rv/rv/baf processors
Aduse a
Creative
shared memory, and with threads working in parallel this can easily lead to collisions. There is therefore
a need for Mutexes to synchronize critical regions; however, the disadvantage of using a Mutex is a delay
in time.

All rights reserved.

© 2013 Accenture.

Bring your talent and passion to a

global organization at the forefront of
business, technology and innovation.
Discover how great you can be.
Visit accenture.com/bookboon

72 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Process planning on multiple processor

It is possible to treat a thread that must wait in two ways: either the thread is spinning or the system
performs a context switch.

Spin Spinning means that a thread is waiting for a Mutex until it becomes free. With spinning, the thread
is repeatedly trying to gain access to a Mutex, as long as the Mutex is locked.

Switch A thread that is waiting (spinning) for ​​a Mutex, can be taken out of the processor to release room to
another thread.

Figure 10.3: Definition of spin and switch

Spinning is a waste of time. However, context switching takes time to perform. What is best will depend
on how long a time the thread is spinning. Spinning will pay off if the thread just has to wait for a short
time, whereas context switching will only pay off if the thread must wait for a while.

In practice, it may be beneficial to use both methods. The system keeps track of the last spin time, and
uses this to decide whether to use spinning or context switching.

10.5 Process scheduling on multiple processors

When there is only one processor, it is only necessary to do a scheduling of the threads. If there are
several processors, the system must also schedule the use of processors. This gives scheduling in two
dimensions, which is much more complicated than scheduling for only one processor.

Sometimes threads are cooperating, which makes process scheduling more difficult when there are
multiple processors. This means that some of the threads running on the system are independent of the
others, while other threads cooperate with each other. In that case, it would be beneficial to coordinate
threads that are cooperating.

We will look at two ways to organize scheduling on multiple processors: time-sharing and space-sharing.

Time-sharing This organizes process planning as for threads working independently. Depending on priority, the
threads hang on one or more lists, and the threads get into a processor when it becomes available.

This organizing is simple and effective.

Space-sharing Here, process planning is organized as for threads that cooperate. The system waits until each
thread can have one processor each. Then the threads start simultaneously and can cooperate.

An advantage of space-sharing is that context switching is not necessary for the threads to cooperate.

Figure 10.4: Definition of time-sharing and space-sharing

A disadvantage of space-sharing is that sometimes a processor is blocked and nothing happens. A system
that combines time-sharing and space-sharing may then be desirable, with Gang scheduling being one
way to do this.

73
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Process planning on multiple processor

The Gang scheduling process takes place as follows:

1. Groups of threads that cooperate are treated as a unit called a gang.

2. All members of the gang run simultaneously on different processors as with timesharing.
3. All gang members start and end their time slices together.

Gang scheduling combines time-sharing and space-sharing to ensure a high overall system throughput
and short response times for interactive tasks.

The advantage of threads running simultaneously is that cooperation takes place immediately. A thread
will easily be able to communicate with another thread if both run at the same time.

10.6 Multiprocessing in Windows

Multitasking is a technique in which multiple threads share a single processor. However, it is an advantage
that a computer has multiple processors, because then the threads can run parallel.

It has been an objective for the Windows operating system to work well on multiprocessor computer
systems. Originally, Windows supported up to 32 processors, but some versions support even more
processors. 64-bit versions of Windows can support up to 64 processors.

The Wake
the only emission we want to leave behind

.QYURGGF'PIKPGU/GFKWOURGGF'PIKPGU6WTDQEJCTIGTU2TQRGNNGTU2TQRWNUKQP2CEMCIGU2TKOG5GTX

6JGFGUKIPQHGEQHTKGPFN[OCTKPGRQYGTCPFRTQRWNUKQPUQNWVKQPUKUETWEKCNHQT/#0&KGUGN6WTDQ
2QYGTEQORGVGPEKGUCTGQHHGTGFYKVJVJGYQTNFoUNCTIGUVGPIKPGRTQITCOOGsJCXKPIQWVRWVUURCPPKPI
HTQOVQM9RGTGPIKPG)GVWRHTQPV
(KPFQWVOQTGCVYYYOCPFKGUGNVWTDQEQO

74 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Process planning on multiple processor

Windows is a symmetric multiprocessing system (SMP). SMP run threads on any available processor,
which ensures that processors and resources are efficiently utilized.

Windows supports multi-core processors, as well as two modern types of multiprocessor systems:

ƒƒ Hyper threading
ƒƒ Non-uniform Memory Access (NUMA)

Hyper threading

Hyper threading is a technology introduced by Intel that supports multiple logical processors on one
physical processor. For each physical processor core, the operating system creates two virtual processors
and distributes tasks between them when possible.

Hyper threading works by duplicating parts of the processor without duplicating the device that performs
the instructions, and is used to improve parallel actions performed by processors on PCs. This system
requires that the operating system support multiple processors.

NUMA systems

In Non-uniform Memory Access (NUMA) systems, processors are grouped into smaller units called nodes.
Each node has its own processor and memory, and connects through a cache-coherent coordinated bus.

By the use of NUMA, a processor has faster access to its own memory than a shared memory. The
advantage of a separate memory for each processor is to avoid that several processors attempt to work
on the same memory, which causes delays.

NUMA makes memory accesses faster, which gives modern computers better performance.

10.7 Control Questions

1. Why have multiple processors become common in PCs?
2. What are the disadvantages of the Master-Slave model?
3. Describe how symmetrical multi-processing (SMP) works.
4. Why is scheduling more difficult when a computer has multiple processors?
5. Describe process scheduling by time-sharing and room-sharing.
6. Describe how Gang Scheduling works.
7. How does Hyper threading work?
8. What are the benefits of NUMA systems?

75
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Deadlocks

11 Deadlocks
When multiple threads are competing for the same resources it can cause something called a deadlock,
which refers to a lock that cannot be opened.

A thread in a deadlock stops running because it is waiting to access a resource. As long as the thread
cannot access the resource, it will continue to wait. The thread can therefore wait forever if it does not
obtain access to the resource.

For a deadlock to occur, there must be two or more threads involved. Two threads with a status deadlock
are in a situation where each thread is waiting for a resource that the other thread occupies. Both threads
will be in a blocked state.

30 FR
da EE
SMS from your computer ys
...Sync'd with your Android phone & number tria
l!

Go to

BrowserTexting.com

and start texting from

your computer!

... BrowserTexting

76 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Deadlocks

Thread 1 Resource 1

Resource 2 Thread 2

Figure 11.1: A deadlock with two threads and two resources.

Thread 1 has access to Resource 1, and tries to access Resource 2.
Thread 2 has access to Resource 2, and tries to access Resource 1.

An example of a deadlock is when two threads have each opened a file for writing, and both threads
want to open the other thread’s file. The figure below illustrates the situation:

File 1 File 2

Thread A Has File 1 open. Trying to open File 2.

Thread B Trying to open File 1. Has File 2 open.

Figure 11.2: Example of deadlock with two threads and two files

Both threads attempts to open the other thread’s file, but a thread cannot access a file that is open for
writing. Both threads will therefore wait forever.

Sometimes there are more than two threads and two resources involved in the same deadlock. For
example, thread 1 is waiting for a resource that thread 2 has, thread 2 is waiting for a resource that
thread 3 has and thread 3 is waiting for a resource that thread 1 has.

Thread 1 Resource 2

Resource 1 Thread 2

Thread 3 Resource 3

Figure 11.3: Deadlock with three threads and three resources.

The figure illustrates why deadlocks are called circular holding.

77
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Deadlocks

The following gives a definition of deadlock:

ƒƒ Some processes are in a deadlock if each process is waiting for an event that only one of the
other processes can cause.

Deadlock is a potential problem in any operating system. Deadlock can occur in a variety of ways, and
there will always be two or more threads involved in a deadlock.

11.1 Examples of deadlocks

In a computer system, many resources can only be used by one thread at a time. These are resources
such as printers, CD players, files and similar items. A deadlock can occur when two threads are both
trying to get hold of a resource that the other thread controls.

An example of a deadlock

The following is a practical example of a deadlock. Two processes both want to scan an image and to
write to a DVD. Both processes therefore have use of the same scanner and DVD.

Process 1 first accesses the scanner and Process 2 first accesses the DVD. Process 1 then attempts to
access the DVD, and Process 2 tries to access the scanner at the same time. This results in the situation
in the figure below:

Scanner DVD

Process 1 Has access to the scanner. Attempting to access the DVD.

Process 2 Attempting to access the scanner. Has access to the DVD.

Figure 11.4: Example of deadlock with a scanner and a DVD

Both processes will stop running and wait for each other, so there is a deadlock.

Deadlock in a Database

A deadlock may occur in many different situations. The following gives an example, in which two
processes are using a database:

1. Process A closes some data in the database.

2. Process B closes some other data in the database.

Now, if both processes try to use the data that the other process has closed at the same time, a
deadlock occurs.

78
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Deadlocks

Deadlocks in a network
Deadlocks can occur in networks with several computers involved. Often, printers, scanners and similar
devices are resources in the network. A deadlock can then occur if multiple computers are simultaneously
attempting to get hold of shared resources in the network.

Communication deadlock
Communication deadlock is a type of deadlock that does not have to do with resources. Suppose that
Process A sends a request to Process B, and then Process A blocks until it gets a reply. If the response
from Process B becomes lost, this will lead to a situation similar to a deadlock.

Starvation is similar to a deadlock

Starvation is an issue that is closely related to a deadlock. Threads, which do not enter the processor, are
in a locked situation similar to a deadlock.

Brain power By 2020, wind could provide one-tenth of our planet’s

electricity needs. Already today, SKF’s innovative know-
how is crucial to running a large proportion of the
world’s wind turbines.
Up to 25 % of the generating costs relate to mainte-
nance. These can be reduced dramatically thanks to our
systems for on-line condition monitoring and automatic
lubrication. We help make it more economical to create
cleaner, cheaper energy out of thin air.
By sharing our experience, expertise, and creativity,
industries can boost performance beyond expectations.
Therefore we need the best employees who can
meet this challenge!

The Power of Knowledge Engineering

Plug into The Power of Knowledge Engineering.

Visit us at www.skf.com/knowledge

79 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Deadlocks

11.2 Conditions for a deadlock to occur

For a deadlock to occur, we must have the following conditions:

1. The mutual exclusion condition

2. The hold and wait condition
3. Non-preemptive condition
4. The circular wait condition

The mutual exclusion condition means that if a thread has access to a resource, no other threads can
access the resource.

The hold and wait condition describes the situation of a deadlock. The threads are holding a resource,
and are waiting for access to another resource.

The non-preemptive condition means that the operating system will not intervene and take a resource
from a thread. The resource will therefore not be released unless the thread willingly lets go of it.

A process will keep a resource until it lets go of it. The circular wait condition is when process P1 holds
resource R1 and process P2 holds resource R2, while process P2 is trying to obtain resource R1 and
process P1 is trying to obtain resource R2. This situation may also involve more than two threads.

11.3 How the operating system can manage deadlocks

The operating system can deal with deadlocks in four ways:

1. Ignore deadlocks.
2. Try to find deadlocks and correct it.
3. Try to avoid that deadlocks occur when allocating resources.
4. Try to avoid deadlocks by avoiding the conditions for them.

1 Ignore deadlocks

One way to deal with deadlocks is to ignore them, which may be acceptable if deadlocks are a rare
phenomenon. If a computer stops working once a year because of a deadlock, it will have little impact
if the computer can easily restart.

2 Finding deadlocks and correct it

The operating system can try to find deadlocks and correct it, as there are techniques developed to detect
deadlocks. Operating systems that do this are usually not trying to prevent deadlocks.

80
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Deadlocks

If the operating system has detected a deadlock, there are different strategies to remove it:

1. Restore through that the operating system intervenes.

2. Restore through reset.
3. Recovery by ending processes.

Restore through that the operating system intervenes means that the operating system takes a resource
from a process and gives it to another process. However, it is often difficult or impossible to correct a
deadlock in this way.

Restore through reset means that a process in a deadlock is set back in time so that it can begin the
allocation of resources again. To achieve the reset of processes, something called checkpoints are used.
The checkpoints of a process mean that the status of the process is stored in a file at specified times,
which means that the process can start again from a checkpoint.

Recovery by ending processes means that the operating system terminates one or more processes to
eliminate a deadlock, which may give other processes an opportunity to proceed.

> Apply now

redefine your future

AxA globAl grAduAte
progrAm 2015
- © Photononstop

axa_ad_grad_prog_170x115.indd 1 19/12/13 16:36

81 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Deadlocks

3 Try to avoid that deadlocks occur when resources are allocated

Instead of trying to find and correct deadlocks, operating systems may attempt to avoid deadlocks. The
operating system can try to avoid a deadlock from occurring when allocating resources.

4 Try to avoid deadlocks by avoiding the conditions for them

The operating system can also try to avoid deadlocks by avoiding the four conditions for deadlocks:

1. The mutual exclusion condition

2. The hold and wait condition
3. The non-preemptive condition
4. The circular wait condition

The mutual exclusion condition implies that only one process at a time can access a resource. However,
this condition is difficult to use to avoid deadlock, because mutual exclusion is necessary for many
resources such as printers, files and similar items.

Hold and wait condition means that for a deadlock to occur, processes will hold a resource while it
attempts to access another resource. One way to avoid this is to require that a process give away the
resources it has, before it gains access to a new resource.

A non-preemptive condition means that the operating system will not take a resource from a process. To
take a resource from a process can have adverse consequences for the process, e.g. to take a printer from
a process that is in the middle of a printing would be drastic. However, it is possible to avoid deadlocks
in connection with the printers if the process writes the document to a file on a hard drive. The process
can then continue and the printer can print the file later.

A circular wait condition means that for a deadlock to occur, a process must attempt to simultaneously
access multiple resources. This means that if a process only tries to access one resource at a time,
deadlocks will not occur.

11.4 Tools to find deadlocks

Driver Verifier is a tool in Window, designed to find bugs in drivers. Hardware manufacturers can use
Driver Verifier to test whether drivers are working correctly under Windows, and are not making illegal
function calls or creating system damage.

Driver Verifier can also check for possible deadlocks by choosing the option known as Deadlock
Detection, which can explore the use of Spinlocks and Mutexes, as well as looking for patterns that might
indicate deadlock. If Driver Verifier detects a deadlock, it stops the system and gives an indication of
which driver is causing the deadlock.

82
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Deadlocks

11.5 Control Questions

1. Give an example of a deadlock.
2. What conditions must be present for a deadlock to occur?
3. What is circular holding in connection to deadlocks?
4. Why do deadlocks resemble starvation?
5. Give some ways an operating system can deal with a deadlock.
6. How can the operating system attempt to avoid deadlocks from occurring?

LIGS University
based in Hawaii, USA
is currently enrolling in the
Interactive Online BBA, MBA, MSc,
DBA and PhD programs:

▶▶ enroll by October 31st, 2014 and

▶▶ save up to 11% on the tuition!
▶▶ pay in 10 installments / 2 years
▶▶ Interactive Online education
▶▶ visit www.ligsuniversity.com to
find out more!

Note: LIGS University is not accredited by any

nationally recognized accrediting agency listed
by the US Secretary of Education.
More info here.

83 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Memory Management

12 Memory Management
Managing memory concerns in terms of how to allocate memory for programs and how to free memory
that is no longer needed. How this is done will affect computer performance, which makes memory
management an important task in a computer system.

The following provides an overview of some of the problems with memory management:

1. There must be space for several programs in the memory at the same time.
2. There may not be room for all programs in the memory at the same time.
3. Programs will have different addresses in the memory at different runs.

Models for treating memory must consider these issues. We will now be looking at some ways to organize
the management of memory.

12.1 Swapping
If there is not space for several programs in the memory at once, it is possible to use swapping. Swapping
means that the processes take turns in disposing the memory.

A process is loaded into the memory and runs for a while. The process is then removed from the memory
and another process is loaded into the memory and runs for a while. In this way, all processes take turns
in disposing the memory.

Swapping is an old-fashioned method that has previously been used. Modern operating systems use
virtual memory, which allows applications to run without loading whole programs into the memory.

12.2 Some memory models

Most modern memory models fall into three categories:

1. A Flat Memory Model

2. A Paged Memory Model
3. A Segmented Memory Model

A Flat Memory Model uses linear indexing for memory addressing, as the indexes start at 0 and go up to
a maximum. The processor uses indexes to directly access data stored in the memory, with a flat model
providing a simple organization of the memory. It needs little resources for memory management, and
the system acquires fast access to data.

84
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Memory Management

A Paged Memory Model is more complicated than the Flat model, and gives a little slower access to
data. By paging memory, the contents are moved to a hard disk so that the memory is released to other
processes or the operating system itself. Data is divided into equally sized blocks called pages. The
advantage of paging is that it is not necessary to keep all programs in the memory at the same time, and
that the system always work with units of the same size.

The paged memory model is suitable for multitasking, in which threads are constantly swapped in the
processor. Paging is an important part of the technique used for virtual memory.

The Segmented Memory Model is similar to the Paged Memory Model, but the sides differ in size. This
makes the segmented memory model more flexible and efficient than paging.

By segmenting, programs are divided into independent segments addresses. Each segment has its own
address space that goes from zero to a maximum. The segment length can be anything between zero
and the maximum. When a program is loaded into the memory, the operating system makes a segment
table that contains the starting address of each segment.

The disadvantage with the Segmented Memory Model is that it is a complex system, which is difficult
to program and can easily result in failure.

12.3 Virtual Memory

Although there is a lot of memory on a computer, it is a possible situation that there is not room for
all processes to run simultaneously in the memory. The operating system put some processes or parts
of processes on a hard drive, so it is then necessary to manage processes scattered on various storage
media. This is usually done by a technique called virtual memory.

Virtual memory is a technique which gives applications the impression that your computer has a unified
and coherent memory, even if the memory can be physically located on different media. The advantage
of the virtual memory architecture is that a part of a process can reside on your hard drive if there is
not enough memory space.

85
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Memory Management

9LUWXDOPHPRU\  
   0HPRU\5$0
   
   
   
   
   
   
   ,ĂƌĚĚƌŝǀĞ
   
   
   
   
   
Figure 12.1: Virtual Memory points to physical memory that can be on various storage media

Almost all the implementation of virtual memory shares the virtual address space into pages. When
a process is running in the processor, the processor reads the code page-by-page in the same way as
humans read a book.

86 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Memory Management

Operating systems use the technique of virtual memory to make it easier to use the physical memory,
which makes it easier for large applications to run on computers. All modern operating systems use the
techniques of virtual memory.

12.4 Memory management in Windows

Memory management in Windows uses the virtual memory architecture. In Windows, each process has
a virtual address space, which is divided into pages.

Windows has a virtual Memory Manager that uses paging, so that only the active parts of programs are
loaded into the memory when they run.

Windows supports Demand-Paged Virtual Memory. This means that the operating system only loads
pages into memory that a running process requests.

A Cache Manager keeps track of virtual addresses used to map files into the memory. This improves the
input/output performance for many applications because read operations can be done without using
the hard drive.

In today’s Windows, Win32 and Win64 coexist. Win32 has a 32-bit memory structure that uses pointers
with 32-bit memory addresses, whereas Win64 has the opportunity for a much larger virtual address
space because Win64 uses 64-bit pointers.

12.5 The Memory Manager

The Memory Manager in Windows takes care of the memory management. It allocates space for programs
in memory when needed, and it removes programs from memory again when they finish.

Memory Manager implements virtual memory, perform tasks on the memory and gives support for
large memories. Memory Manager also creates services such as keeping track of files that are part of the
memory and gives support to the Cache Manager.

The Memory Manager has system services that allocate and free virtual memory, shares memory
between processes and manages pages. Memory Manager also has many services that allocate and free
physical memory.

Programmers can use most of the services in the Memory Manager through the Windows API.

87
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Memory Management

12.6 The Heap Manager

Dynamic memory allocation is a method used to allocate memory to processes running on a computer.
The heap is a portion of the memory used for dynamic memory allocation. Programmers can use the
heap to store variables and objects. Variables and objects created in the heap will be there until deleted
by the process or by the Garbage Collector.

Windows divides the memory into heaps. When each thread has a free heap, there will be no competition
for the memory between threads, which increases system performance.

Each process has at least one free heap. Once the process starts, it creates a heap called the Default
Process Heap, which exists until the process is completed. Processes can also create additional private
free heaps using a Windows API function called HeapCreate.

The Heap Manager optimizes the performance of memory usage for allocations in the heap. Heaps are
Windows objects and processes, which thus have access to the free heap by means of a handle.

12.7 Stacks
When a thread is running, it must have access to a temporary storage where it can store local variables,
have function parameters and return addresses for function calls. A stack is a portion of the memory
used for this.

A stack is a last-in, first-out data structure. A new element is always added at the top of a stack, and
when removing an item from a stack, the top item is always removed. Items at the bottom of a stack will
therefore be there until all the items over it are removed.

The Memory Manager in Windows uses two stacks for each thread, a user stack and a kernel stack. When
a thread is created, it will automatically get a user stack and a kernel stack. User stacks generally has a
size of 1MB, while kernel stacks are often much less, with a common size being 12 KB.

12.8 Control Questions

1. What makes memory management difficult?
2. What is swapping?
3. Describe how the Paged Memory Model works.
4. What is the difference between the Paged Memory Model and the Segmented Memory Model?
5. Describe the technique of virtual memory.
6. How is memory management done in Windows?
7. What is the use of the free heap?
8. What is the task of the Heap Manager?
9. What is a stack?

88
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Caching in Windows

13 Caching in Windows
The purpose of caching is to make input/output run more efficiently, which is done by storing data in
a cache. A cache is a piece of memory for the temporary storage of data that is expected to be used
several times.

Processors use a cache to speed up the computer system. It is faster to retrieve data from a small cache
near the processor than to retrieve data from a large memory. When a processor will retrieve data from
the memory, it first checks whether the data is in the cache. If there is no data in the cache, it is retrieved
from the memory and then added to the cache for future use.

The cache is often used in conjunction with reading hard disks. To retrieve a file from a cache is much
faster than reading the file from a hard disk, so files are therefore often stored in a cache. Files that are
read into a memory from a hard drive are taken care of in a cache so that the files can be retrieved
quickly from the cache instead of reading them from the hard drive.

Internet browsers use caching to work faster, as it saves time to have web pages that are much read in a
cache at an Internet network provider. Web pages can then be retrieved from a cache instead of being
loaded down via the Internet from a distant part of the world.

13.1 The Cache Manager

The Cache Manager in Windows takes care of the organization of caching. The Cache Manager consists
of a set of functions in kernel mode, as well as threads that cooperate with the Memory Manager.

An important task for the Cache Manager is to ensure that processes that are trying to obtain data get
the latest version of the data, and not older versions.

The Cache Manager provides a high-speed intelligent mechanism to reduce input/output for disk drives
and improve computer system performance.

13.2 Control Questions

1. What is the purpose of caching?
2. Describe how caching takes place.
3. In what connections is caching used?

89
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Input and output

14 Input and output

A computer is connected to devices such as a screen, keyboard, mouse, hard disk, printer, CD/DVD, etc.
An important task for an operating system is to communicate with the input/output devices connected
to the computer.

14.1 Problems with input and output

Input and output have been a problem in operating systems. One reason for this is the large difference
in speeds within a computer system.

In the processor and the memory, data processes very quickly, though by comparison communication
with peripheral devices is very slow. To write to the screen and a file is a slow process. A user typing on
a keyboard can wait several seconds before he types a character, while for the processor, it is not a good
use of time to wait for input/output devices.

90 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Input and output

Type of equipment Transfer speed

Keyboard 10 bytes/second

Mouse 100 bytes/second

72x CD-ROM 11 Mb/second

SAS disk 129 Mb/second

USB 3.0 400 Mb/second

PCI bus (64-bit) 533 MB/second

Figure 14.1: The table provides an overview of the speed of some input/output devices

Another challenge with input/output is that hardware connecting to a computer is from different
manufacturers. For example, there are many types of screens, and these require separate drivers. It would
hence be cumbersome if programmers had to create code for each type of screen driver that might be
using the program.

The operating system takes care of communication with hardware

It makes programming much easier when programmers can communicate with the operating system, rather than directly
with the hardware. For example, in his code a programmer can write an output to the screen as Write (“A Message”).

The operating system takes care of the printing to the screen, so the programmer does not have to think about
communication with the various drivers that may be for screens from different manufacturers.

The operating system takes care of the communication with the screen, mouse, printers and
similar equipment. Applications can therefore work on different computers with different hardware
devices connected.

14.2 Block devices and character devices

There are two categories for input/output devices:

ƒƒ Block Devices
ƒƒ Character Devices

Block Devices store data in blocks, each of which has an address. The typical size of a block has generally
been between 512 bytes and 32768 bytes. Hard drives, CD-ROM and USB drives are examples of
block devices.

Character Devices are another type of input/output devices than Block Devices. Character devices
transmit and receive data as characters. Examples of Character Devices are the keyboard, the mouse,
printers and network connections. Devices that are not discs are usually Character Devices.

91
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Input and output

14.3 Data Streams

An application must open a connection to an input/output device before using it, with a stream being
the term used for such a connection.

When a new process starts, it automatically creates three Data Streams:

1. Standard input from the keyboard

2. Standard output to the screen
3. Standard error to the screen

When a stream opens, it can stay open as long as needed. However, the operating system must keep track
of which Data Streams are open, and thus has a data structure to keep track of streams.

14.4 Treatment of errors in input and output

The operating system takes care of errors that can occur when communicating with input/output devices.
The operating system should be able to:

1. Detect errors.
2. Correct mistakes.

DO YOU WANT TO KNOW:

What your staff really want?

The top issues troubling them?

How to make staff assessments

work for you & them, painlessly?

How to retain your

top staff Get your free trial
FIND OUT NOW FOR FREE
Because happy staff get more done

92 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Input and output

Many errors can occur when using input/output devices. The operating system should be able to detect
whether a process tries to write to a file opened for reading, and also recognize corrupt data.

Sometimes the operating system successfully corrects errors that occur. If the operating system is able
to correct a mistake, it will not normally notify the program where the error occurred.

14.5 Synchronous and asynchronous input/output

Input and output are slow processes, which means that the system must wait for input/output operations.
How long the system must wait varies, from a few milliseconds to save a file to a hard drive to up to
several minutes of waiting for a character from the keyboard.

Instead of waiting for input/output operations, it may be better for a thread to do something else. However,
a thread often has to perform an input/output operation before it can continue. For instance, if a thread
uses ReadLine( ) to read a line of data from the screen, the thread cannot proceed until the line is read
if the thread requires the data from the line in the further program execution.

Output goes somewhat better. If we have written Write (“A Message”) in a thread, the thread can proceed
without waiting for the print to the screen to happen.

There are two ways for a thread to deal with input and output:

1. Synchronous I/O
2. Asynchronous I/O

Through the use of synchronous I/O, a process will wait to continue until an input/output operation
is complete, as the thread goes to sleep while waiting. Another name for synchronous I/O is wait I/O.

Through the use of asynchronous I/O, the process continues to run without waiting until an input/output
operation is performed. Another term for asynchronous I/O is no wait I/O.

Asynchronous I/O requires synchronization in the code of the thread. It is possible to use a semaphore,
which causes the thread to wait until it gets a signal from the driver. Using a semaphore with asynchronous
I/O provides the same effect as synchronous I/O. However, an advantage of asynchronous I/O is that
the thread can execute some other code before the thread starts to wait for the input/output operation
to finish.

14.6 Use of a buffer

When doing an output, it may be helpful to use a buffer. For example, if a thread wants to write to a
printer and the printer is busy, the thread can write the output to a buffer instead. The thread can then
continue, and the printer can print the contents of the buffer later.

93
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Input and output

The advantage of using a buffer is to save time. Once a thread has written to a buffer, it can continue
executing other codes. The disadvantage of using a buffer is that it requires space on a hard drive.

When a thread is writing to a file, it may be necessary to close the access to other processes. Locks are
mechanisms used to close files.

14.7 Polling
We will now take a look at how the operating system communicates with hardware. What happens is
that signals from peripheral devices such as keyboards transfer to the computer’s memory.

One way to treat the data from the keyboard is to put it in a register. The processor can then regularly
check the register and update applications with this data. This method is called polling.

Using a register to take care of keystrokes has the disadvantage that it does not synchronize reading and
writing to the register, as writing to the register can go faster than reading or vice versa. Two presses of the
keyboard before the processor has read the register can overwrite a keystroke. Alternatively, the processor
may read the same keystroke twice because there have not been more keystrokes since the last check.

To help solve this problem, two registers are used:

1. A register that takes care of the keystroke.

2. A register that indicates whether the processor has read the register or not.

In the polling method, registers are regularly controlled. However, this is cumbersome since there are
many registers, and each peripheral has registers.

The polling method is not good because it lacks synchronization, and the processor may possibly check
at the wrong time. Either there is nothing in the registers when the processor checks or they are overfull.
It is therefore better to use interrupt signals.

14.8 Interrupts
The processor should not have to waste time checking registers such as when using the polling method.
Interrupts are a better method than polling since interrupt signals allow peripheral devices to notify the
processor every time there is something new in a register.

With an interrupt system, the processor can keep doing what it does without worrying about
peripheral devices.

The peripheral device notifies the processor when there are data to be processed. When the processor
gets an interrupt signal, it ends what it is doing and reads the register.

94
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Input and output

14.9 Direct access to memory

Interrupt signals with data registers work for key presses and similar, although this system is not so
good for reading from a file when the amount of data read into a memory is large. It is a burden for the
processor to handle at lot of data like a file transfer, so a system called Direct Memory Access (DMA)
is used instead.

Like polling, the DMA method uses a status register and an interrupt system. However, DMA does not
use any data register, as the data transfers directly from the file into the memory.

DMA allows certain hardware subsystems to directly access memory independently of the CPU. A usual
input/output transfer makes the processor busy all the time during the input/output, and the processor
will not be able to perform other operations. Using DMA, the processor will initiate input/output transfer
and do other tasks while the transfer takes place, in addition to receiving an interrupt signal when the
transfer is complete.

DMA is an important feature of modern computer systems because it makes it possible for devices to
transfer data without too much burdening of the processor.

Challenge the way we run

EXPERIENCE THE POWER OF

FULL ENGAGEMENT…

RUN FASTER.
RUN LONGER.. READ MORE & PRE-ORDER TODAY
RUN EASIER… WWW.GAITEYE.COM

1349906_A6_4+0.indd 1 22-08-2014 12:56:57

95 Click on the ad to read more
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Input and output

14.10 About drivers for input/output devices

Each physical device attached to a computer needs software specifically designed for it and its tasks. A driver
is software that makes it possible for the operating system to communicate with hardware, and is either
a part of the operating system or a small program designed to communicate with the operating system.

When an application needs to use a driver, it calls a function that belongs to the operating system’s
library. The operating system then calls that driver’s code. When the peripheral device is ready, it sends
an interrupt signal to the processor.

The operating system calls a driver using a Device Switch, which is an array. Each driver has an index
in this array that contains a collection of pointers that point to methods used to use the driver.

Different operating systems manage drivers in two ways:

1. The driver is always in the memory, so that it is sleeping when not in use.
2. The driver is loaded into the memory only when necessary.

A data structure called an IORB is important for a driver.

IORB = Input Output Request Block

This data structure contains parameters for input/output driver, and the information in an IORB can be:

ƒƒ A flag indicating whether to read, print or similar.

ƒƒ Where data is located in memory.
ƒƒ The Number of bytes to transfer.
ƒƒ The peripheral unit number.

To make sure that the drivers for hardware work well with the rest of the Windows operating system,
Microsoft has defined a model called a Windows Driver Model (WMD) that drivers are expected to
work with. Microsoft also has a Driver Verifier that examines many of the functions of drivers to make
sure that they follow the requirements of WDM.

There are hundreds of thousands of hardware drivers for a Windows operating system, which provides
a lot of code. Application errors in the driver code means that the system stops, and That Windows will
then show the Blue Screen of Death. The normal program execution then stops.

14.11 Input and output in Windows

A Windows operating system provides an interface to handle a variety of input/output services and
drivers. A Windows input/output system consists of many components that together treat hardware
devices and provide a user interface to hardware devices for applications.

96
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Input and output

Some key components of a Windows input/output system are the Input/output Manager, device drivers,
the Installation and Configuration Manager, the Power Manager and the registry.

A device driver provides an interface between the operating system and a hardware device, and receives
commands via the Input/output Manager.

The Installation and Configuration Manager detects changes in the layout of the hardware, and load and
removes drivers. The Power Manager tries to reduce the power consumption of hardware equipment
not in use. The Input/output Manager also provides support for manipulating the input/output objects
in the kernel.

The registry serves as a database that stores application settings and descriptions of hardware devices
connected to the computer. The registry also has information at the startup and configuration of drivers.

14.12 The Input/output Manager

The Input/output Manager is the central component of the input/output system. It forms a connection
between software and hardware that provides an infrastructure that supports hardware drivers.

The Input/output Manager takes care of system hardware drivers, and does a variety of services related
to the equipment attached.

This e-book
is made with SETASIGN
SetaPDF

PDF components for PHP developers

www.setasign.com

97 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Input and output

The Input/output Manager takes care of the following tasks:

ƒƒ The implementation of drivers.

ƒƒ The configuration of drivers.
ƒƒ Access to hardware devices.
ƒƒ Operations on hardware equipment.

14.13 The Plug and Play Manager

The Plug and Play Manager takes care of connecting new devices to a computer, as well as the removal
of equipment from the computer. If a new hardware device does not have a driver, the Plug and Play
Manager acquires a driver for it.

The Input/output Manager includes the installation and configuration of the new hardware connected
(Plug and Play). The Plug and Play Manager starts when a new hardware component is detected on the
computer, and then attempts to find a driver for this hardware component.

The mission of the Plug and Play Manager is to make sure that Windows can recognize and adapt to
different hardware devices, with the advantage being that users do not have to think about installing
the drivers themselves. When new equipment connects to a computer, the Plug and Play Manager
automatically makes a configuration and adds drivers.

Some features and functions of the Plug and Play Manager:

1. It automatically detects the installed hardware devices.

2. It allocates resources to the hardware equipment.
3. It finds the correct drivers for the hardware devices.
4. It also supports network devices such as printers.

14.14 The Power Manager

The Power Manager manages the system’s power usage. Previously, reduced electricity use would turn off
the monitor and the hard disk. Now, newer methods reduce the power consumption of the computer’s
components when not in use.

Windows supports a method for shutdown called sleep mode (hibernation), which copies all physical
memory to the hard disk. Power consumption is then reduced to a minimum.

An alternative to Hibernation is Standby Mode, which also reduces power consumption to a minimum
for the entire system.

98
Download free eBooks at bookboon.com
 An Introduction to
Windows Operating System Input and output

There are so many computers in the world that a reduced power consumption of computers has great
economic importance. The power consumption of computers has been so large that a reduction could
reduce the number of nuclear power plants in the world.

14.15 Control Questions

1. What makes input/output to a problem in a computer system?
2. What is the difference between block devices and character devices?
3. What is a stream? Give an example of a stream.
4. What is the difference between synchronous I/O and asynchronous I/O?
5. What are the advantages and disadvantages of using a buffer?
6. Describe the polling method.
7. Why are interrupts better than the polling method?

360°
8. Describe the DMA (Direct Memory Access) method.
9. What is a Device Switch?

.
10. What are the tasks for the Input/output Manager in Windows?
11. What are the tasks for the Plug and Play Manager in Windows?
12. Describe some ways the Power Manager can save power on a computer. thinking

360°
thinking . 360°
thinking .
Discover the truth at www.deloitte.ca/careers Dis

© Deloitte & Touche LLP and affiliated entities.

Discover the truth at www.deloitte.ca/careers © Deloitte & Touche LLP and affiliated entities.

Deloitte & Touche LLP and affiliated entities.

Discover the truth

99 at www.deloitte.ca/careers
Click on the ad to read more
Download free eBooks at bookboon.com

© Deloitte & Touche LLP and affiliated entities.

An Introduction to
Windows Operating System File systems in Windows

15 File systems in Windows

Before you can start using a new hard drive, you must format it to use a file system.

A file system stores and organizes the files on a hard disk, so it is desirable that it should be easy to find
and retrieve files. File systems are therefore made for this purpose.

To take care of files and file management, the Windows operating system uses the File Manager. The
mission of the File Manager is to organize the files so that users can obtain them quickly and easily.

In the Windows operating system, there are three file systems used on hard drives: there are the NTFS
file system, the older FAT and FAT32. Windows also supports file systems for CD-ROM and DVD.

15.1 FAT
FAT is an abbreviation for File Allocation Table, and is the file system used in MS-DOS and early versions
of Windows. There have been several versions of FAT, including FAT12, FAT16, FAT32 and exFAT.

FAT12 was a 12-bit address system designed for floppy disks, while FAT16 was developed when PCs
with hard disks were taken in use. The first versions of MS-DOS and the very first versions of Windows
used FAT12 and FAT16, which are no longer in use.

The early versions of the Windows operating system were Windows 95, Windows 98 and Windows
Millennium, all of which used FAT32, and which is no longer in general use.

exFAT, which is also called FAT64, is a newer version of FAT. exFAT is designed for smaller storage
devices such as USB pens, so is therefore a version of FAT still in use.

15.2 NT File System

The NT File System (NTFS) is the file system developed for Windows NT, and is the file system that
current versions of Windows use.

NTFS supports long file names, security, fault tolerance, encryption, disk compression and very large
files and volumes.

NTFS has several advantages over FAT32 when it comes to safety, reliability, extensibility and efficiency.

100
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System File systems in Windows

Safety

Security is enhanced by the fact that users are given access to just the directories and files they need. In
FAT32, all users could access all the files on a hard disk.

Reliability

NTFS keeps track of changes in the file system by keeping a journal. NTFS uses log files to keep track
of all disk activity, which allows an NTFS volume to recover quickly after a disk crash.

Extensibility

Using NTFS formatted volumes can expand the storage capacity to existing volumes without having to
take backup, to repartition, to reformat or to restore anything.

Efficiency

NTFS volumes will manage partitions larger than 8 GB of memory more efficiently than the old FAT32
file system.

Compression

NTFS supports the compression of files and directories. You can create a file as compressed, and then
NTFS automatically compresses the contents of the file.

File names

A file name in NTFS can be up to 255 characters long. File names are in Unicode, which means that you
can use file names in character sets other than Latin, e.g. Greek, Chinese, Russian, etc.

15.3 Support for SSD disks

Solid State Drive (SSD) is a storage device that uses flash memory instead of mechanical components such
as hard drives. This makes SSDs faster and less noisy than the usual hard drives, though the disadvantage
of SSDs is that they have been more expensive than hard drives.

Versions of Microsoft Windows prior to 7 did not take any special measures to support solid state drives.
Windows 7 has support for SSDs, with the support in Windows 8 and 8.1 being similar. The operating
system detects the presence of an SSD and optimizes its operation accordingly.

101
Download free eBooks at bookboon.com
 An Introduction to
Windows Operating System File systems in Windows

15.4 File systems for CD-ROM and DVDs

Windows supports a file system for CD-ROM called a Compact Disc File System (CDFS), which is the
standard for optical disc media. CDFS supports various file systems such as Windows, Mac OS and
UNIX systems, so that data can be exchanged between operating systems.

Windows also supports a file system called a Universal Disk Format (UDF). A UDF is a file system
standard for data storage on all optical media, but is mostly used for DVDs and newer optical media.
For example, Blu-ray and DVD video use different versions of UDF.

15.5 Control Questions

1. Which versions of Windows used FAT32?
2. What is the use of exFAT?
TMP PRODUCTION NY026057B 4 12/13/2013
3. What file system is used by current versions of Windows?
6x4 PSTANKIE ACCCTR00
4. What advantages does NTFS have compared to FAT32?
gl/rv/rv/baf Bookboon Ad Creative
5. What is the difference between SSD disks and traditional hard drives?
6. What versions of Windows have support for SSD disks?
7. Which file systems does Windows support for CD-ROM and DVDs?

All rights reserved.

© 2013 Accenture.

Bring your talent and passion to a

global organization at the forefront of
business, technology and innovation.
Discover how great you can be.
Visit accenture.com/bookboon

102 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Storage Management

16 Storage Management
Storage Management defines how operating systems interact with disks and storage media. Windows
provides support for many types of storage media, including hard drives, USB drives, tape drives and network
storage such as SAN (Storage Area Networks) and iSCSI (Internet Small Computer System Interface).

In the following, we shall look at storage in hard drives, since hard drives have long been the primary
storage media for computers.

16.1 About storage on hard drives

A disk consists of sectors, which are a block of a specific size that has an address. A hard drive consists
of a series of blocks. The size of the blocks has been from 2 to 8 Kbytes, and the blocks have numbers
from zero and upwards.

One task for an operating system is to manage hard drives. When a new file is to be stored, the operating
system must allocate space on the hard disk. Some ways to do this are:

1. Interconnecting allocation
2. Linked list
3. File Map
4. Indexed allocation

Interconnecting allocation means that the file is stored in blocks that are in consecutive order, e.g. 1, 2,
3, 4, 5… 285. The operating system then only needs to keep track of the starting block and the number
of blocks that the file occupies.

The problem with the connecting allocation is that:

1. It is not always that contiguous blocks are free on the hard disk when creating a file.
2. When expanding a file, free blocks may not be available so that the expansion will not
be continuous.

Another way to keep track of a file is to use a linked list. Each block has a pointer that points to the next
block in the file, hence the blocks do not need to n consecutive order. A disadvantage of a linked list is
that if a pointer is not positioned correctly to a block, all blocks that follow will be lost.

A file map is an enhancement of the linked list. All the pointers are located in one place, so in this way,
we get a map of the file. If one pointer is wrong, it does not influence the others.

103
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Storage Management

Indexed allocation is also an improvement of the linked list, as all the pointers are in one place. The
pointers are accessed using indexes instead of using a pointer to the next block in the way of the linked list.

16.2 Organization of hard disks

There are different ways to organize the data storing in a computer system. A PC typically only has one
volume on one hard drive. However, it may be beneficial to organize a disk by dividing it into several volumes.

Volumes used by servers will often extend across multiple hard drives, the purpose of this being to
increase the reading speed and security.

Partitions

A hard drive consists of one or more partitions, which is a section of a hard drive. Each partition can
then function as a separate disk.

A partition is a collection of contiguous sectors on a disk. A partition table or other databases for disk
management stores the starting sector and the size of the partitions.

The task of the Partition Manager is to create, delete and manage partitions, thereby ensuring that all
partitions have a unique ID.

The Wake
the only emission we want to leave behind

.QYURGGF'PIKPGU/GFKWOURGGF'PIKPGU6WTDQEJCTIGTU2TQRGNNGTU2TQRWNUKQP2CEMCIGU2TKOG5GTX

6JGFGUKIPQHGEQHTKGPFN[OCTKPGRQYGTCPFRTQRWNUKQPUQNWVKQPUKUETWEKCNHQT/#0&KGUGN6WTDQ
2QYGTEQORGVGPEKGUCTGQHHGTGFYKVJVJGYQTNFoUNCTIGUVGPIKPGRTQITCOOGsJCXKPIQWVRWVUURCPPKPI
HTQOVQM9RGTGPIKPG)GVWRHTQPV
(KPFQWVOQTGCVYYYOCPFKGUGNVWTDQEQO

104 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Storage Management

Volumes
The storage on a computer is divided into volumes designated by a letter such as C, D or E.

A simple volume is only one partition, but you can organize volumes so that they consist of multiple
partitions on one or more hard drives.

A simple volume uses only one hard drive, which means that if the hard disk crashes, the volume is out
of use. Using multiple volumes and multiple disks avoids this, hence you can still access data even if
one hard drive is not functioning.

There are two types of volumes:

1. Simple volumes representing sectors on a single partition

2. Multi-partition volumes representing sectors from multiple partitions

Advantages of multi-partition volumes are performance, reliability and the size of volumes. Some types
of multi-partition volumes are spanned volume, striped volume, RAID 1 volume and RAID 5 volume.

Spanned Volume
A spanned volume is a volume that consists of multiple partitions.

The advantage of using a spanned volume is that one can expand the volume without having to replace
the drive with a new larger drive, whereas the disadvantage of a spanned volume is a greater risk of disk
failure when the partitions are located on different drives.

Spanned volumes have no fault tolerance such as RAID, so if a disk fails, the entire volume will be lost.

Striped volume
A striped volume consists of multiple partitions on multiple hard drives. When you write to a striped
volume, the data is distributed across all hard drives.

The advantage of a striped volume is that the reading speed increases because the computer can read the
data from disks in parallel. The disadvantage of a striped volume is that it can easily cause drive failure.

RAID 1 volume
A RAID 1 volume consists of two hard drives. The same data is on both drives, which means that one
disk is a copy of the other.

The advantage of RAID 1 volume is security. If one of the hard drives fails, you still access the data from
the other hard drive. The disadvantage of RAID 1 is that it uses twice as much storage space.

105
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Storage Management

RAID 5 volumes

RAID 5 volumes have been widely used in servers because it increases both the reading speed and
security. To use RAID 5, at least three disks are necessary, as RAID 5 writes the data to multiple disks.
Security improves when additional information about the data (parity data) is on multiple disks.

RAID 5 results in an increased reading speed because you can read the data from disks in parallel.
However, RAID 5 entails a little slower write speed, as parity data is also written.

RAID 5 can tolerate that one of the disks fails since the other disks will continue to work so that you
can retrieve data there. Even so, if a disc is out of use the reading speed will be slower, as you must use
the parity data.

RAID 6 is an improvement of RAID 5, and is quite similar to RAID 5. RAID 6 is an alternative to RAID
5 if the storage system contains four disks or more because RAID 6 tolerates that two disks fails.

Standard disks and dynamic disks

Windows distinguishes between two types of disks:

1. Basic Disks
2. Dynamic Disks

30 FR
da EE
SMS from your computer ys
...Sync'd with your Android phone & number tria
l!

Go to

BrowserTexting.com

and start texting from

your computer!

... BrowserTexting

106 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Storage Management

A Basic Disk has a fixed size and is only on one physical hard drive. A basic disk can contain several
partitions, and there may be several volumes of a standard disk.

Dynamic disks are more flexible than basic disks, but a disk must be dynamic to share volume with
other disks. It is therefore necessary to use dynamic disks if you use multi-partition volumes such as
spanned volume, striped volume or RAID 5 volume. One advantage of dynamic disks is that you can
resize a disk without restarting Windows.

16.3 Storing data on servers

Networked computers offer new possibilities for data storage. The organizing of data storage on servers
is usually done in another way than for a single PC.

SAN

A SAN (Storage Area Network) is a storage medium that is available to servers via a network so that it
looks as if the storage media is located locally on the server.

An advantage of using a SAN is that multiple servers can use a single storage medium, which is
called storage sharing. Storage sharing is economical, simplifies management and is more effective if a
server crashes.

iSCSI

iSCSI (Internet Small Computer System Interface) is a network standard based on IP (Internet Protocol)
to access a storage device via a network. The purpose of iSCSI is to help facilitate data transmission over
the network and manage the storage of data over long distances. iSCSI can be used to transmit data over
LANs (Local Area Networks), WANs (Wide Area Networks) and the Internet.

iSCSI has become a popular way to store data on a network. This is because it is possible to store data
on a remote server in the same manner as a local computer.

Data Storage in a Cloud

Data can be stored in a Cloud. You can store your data on servers that are located at remote server parks
connected to the Internet. Instead of having your own storage system in a LAN, you can rent storage
space on a server via the Internet.

It is expected that a Cloud will greatly influence the IT industry and the use of computers in the future
since the cloud can provide a combination of cost effectiveness and flexibility. Instead of having to buy
servers themselves, companies can lease capacity when needed.

107
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Storage Management

16.4 Control Questions

1. What are the advantages and disadvantages of using a connecting allocation when files are
stored in a hard drive?
2. Define the following terms associated with hard disks: partition, volume, simple volumes,
multiple partition volumes, spanned volume and striped volume.
3. What is the benefit of using RAID 1 volume?
4. What is the reason that RAID 5 volume has been widely used?
5. What is the difference between a basic disk and a dynamic disk?
6. What are the advantages of using SAN as a storage medium?
7. What is iSCSI and why has iSCSI become a popular way to store data?

Brain power By 2020, wind could provide one-tenth of our planet’s

electricity needs. Already today, SKF’s innovative know-
how is crucial to running a large proportion of the
world’s wind turbines.
Up to 25 % of the generating costs relate to mainte-
nance. These can be reduced dramatically thanks to our
systems for on-line condition monitoring and automatic
lubrication. We help make it more economical to create
cleaner, cheaper energy out of thin air.
By sharing our experience, expertise, and creativity,
industries can boost performance beyond expectations.
Therefore we need the best employees who can
meet this challenge!

The Power of Knowledge Engineering

Plug into The Power of Knowledge Engineering.

Visit us at www.skf.com/knowledge

108 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Networking Features in Window

17 Networking Features
in Windows
The first of Microsoft’s operating systems had little network support, although Windows NT was
developed to function in a network. In today’s Windows, there is therefore broad support for network
tasks in the input/output system and the Windows API.

The common task for a network application is to take a request from an application on one computer and
send it to another computer. The remote computer then performs the request and sends back the result.
For this to occur, the request must often be transformed several times. For example, a request sent over
the Internet will be divided into several packages.

An operating system hence needs services that enable communication between computers. A Windows
operating system has networking software just for this purpose.

There are four types of networking software in Windows: network services, network APIs, protocols
and drivers for network equipment.

17.1 Network Services

Windows has several network services based on the API components.
Two of these are remote access and Active Directory.

Remote Access

Remote Access allows clients to associate a connection to servers so that they can obtain resources via
a network connection. This may be resources on a server such as files, printers and network services.
Windows allows two types of remote access, which are a dial-up connection and a Virtual Private
Network (VPN).

A dial-up connection allows clients to connect to a server via telephone lines or a similar infrastructure.
Dial-up is a temporary physical or virtual connection between a client and a server.

Remote Access with VPN establishes a connection to a server over an IP network such as the Internet.
You can log on with VPN to a server in local network from anywhere in the world via the Internet.

109
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Networking Features in Window

Active Directory

Active Directory is a tool used to manage a local network using a Windows Server operating system,
and is a directory that provides an overview of all users and all devices in the network.

With Active Directory, a network administrator organizes users and computers in groups. In large
networks, it is necessary to group users and computers to maintain an overview of the network. Active
Directory makes it possible to manage very large networks with up to millions of users.

17.2 Some network APIs

Network APIs are network functions in Windows API. Windows has several network APIs that provide
support for software, and applications can use these to communicate with programs on other computers.

Some network APIs include Windows Sockets, Remote Procedure Call, Named Pipes and Mailslots.

Windows Sockets

Windows Sockets API, also called Winsock, is a technical specification that defines how Windows network
software should cooperate with network services such as TCP/IP. Windows Sockets provides an interface
between a Windows TCP/IP client program and the underlying TCP/IP protocol.

> Apply now

redefine your future

AxA globAl grAduAte
progrAm 2015
- © Photononstop

axa_ad_grad_prog_170x115.indd 1 19/12/13 16:36

110 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Networking Features in Window

Windows Sockets makes it possible for developers to create advanced network of applications for both
the Internet and intranet using Microsoft Windows networking functions, regardless of the network
protocol used.

Remote Procedure Call

Remote Procedure Call (RPC) is a type of process communication that allows a computer to execute
code on another computer over a network, without the programmer having to code the details of the
communication. In other words, the programmer writes about the same code, whether the code is executed
on the local computer or on a remotely located computer. RPC therefore makes the communication
process as simple as a function call, operating between processes at different computers in a network.

Named Pipes

Named Pipes is a programming API for communication between processes, which takes place between a
named pipe server and a named pipe client. A named pipe server is a program that creates a named pipe
that clients can use, as data is transferred via a buffer. A process writes data to a buffer so that another
process can read the data from the buffer.

In Windows, Named Pipes is a client/server communication that works quite similarly to Sockets.

Mailslot

Mailslot is a broadcast mechanism for one-way communication that allows communication between
processes, both locally and over a network. The messages are usually sent via a local network or an Internet
network. A process that creates a mailslot is called a mailslot server. Other processes (clients) can send
messages to a mailslot server that has a name. Mailslots provides a simple way of sending short messages.

17.3 About network drivers in Windows

Network API drivers take API requests and translate them into network protocol requests, so that it is
possible to send them via a network. API drivers use transport protocol drivers to do this translation.

In 1989, 3Com and Microsoft developed NDIS (Network Driver Interface Specification), which allows
protocol drivers to communicate with network adapter drivers. NDIS is independent of the type of
equipment used by a computer. Network adapter drivers that use NDIS are called NDIS drivers or NDIS
miniport drivers.

Transport Driver Interface (TDI) is an interface developed by Microsoft to make it easier for drivers to
communicate with various network transport protocols. The advantage of using TDI is that services are
independent of different protocols for transport in networks.

111
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Networking Features in Window

TDI transports, also called transports and NDIS protocol drivers, are drivers in kernel mode. They receive
packets from TDI and send them further.

NDIS miniport drivers are drivers in kernel mode, which provides an interface for TDI transports to
network adapters.

About adapters
An adapter is a transition that allows the transfer of data between devices that are not compatible. Adapters are common
as transitions between different connections in computer equipment, in which plug-in formats in periods have been
non-standardized.

17.4 Control Questions

1. What kind of network software is there in standard Windows?
2. How does remote access by VPN work?
3. What is the task for an Active Directory in a local network?
4. What are network APIs?
5. Mention some network APIs in the Windows operating system.
6. What is the role of network API drivers?
7. What is the task of NDIS (Network Driver Interface Specification)?
8. What is the role of TDI (Transport Driver Interface)?

LIGS University
based in Hawaii, USA
is currently enrolling in the
Interactive Online BBA, MBA, MSc,
DBA and PhD programs:

▶▶ enroll by October 31st, 2014 and

▶▶ save up to 11% on the tuition!
▶▶ pay in 10 installments / 2 years
▶▶ Interactive Online education
▶▶ visit www.ligsuniversity.com to
find out more!

Note: LIGS University is not accredited by any

nationally recognized accrediting agency listed
by the US Secretary of Education.
More info here.

112 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Security in Windows

18 Security in Windows
In a computer system, there is often sensitive data, to which it is not desirable that there should be public
access. The operating system must be able to protect files, memory and setup data so that unauthorized
persons cannot read or modify the data.

18.1 About security in computer systems

There are four groups of security for computer systems.

1. Data Confidentiality
2. Data Integrity
3. Access to the system
4. Attacks from outside

Data Confidentiality concerns how to prevent unauthorized access to data on computers. This is important
for protecting documents to prevent unauthorized persons from gaining access to- or reading them. This
applies not only to secret documents, but also personal data that could be sensitive.

Data Integrity is about preventing unauthorized changes to data in files. This applies not only to a change
of data, but also to remove or to add false data. An example of this is a student who will attempt to get
into the school’s computer system to change his/her grade.

Security in terms of access to the system revolves around nobody being allowed to disturb the system
or to put it out of operation.

The fourth type of security concerns the preventing of attacks via the Internet, as hackers may attempt
to gain control of computers via the Internet. One way to do this is by the use of a virus. By gaining
control of a computer, hackers can use the computer for illegal activities or to send e-mails (spam).

18.2 Security in Windows NT

With Windows NT came the following security in the Windows operating system:

ƒƒ Secure login
ƒƒ Access control for files
ƒƒ Privileged access control
ƒƒ Address space defense for each process
ƒƒ Clearance of pages in memory
ƒƒ Auditing of computer systems

113
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Security in Windows

Secure login means that all users use a password to login. Ctrl + Alt + Delete has been used to login,
the purpose of the Ctrl + Alt + Delete being that no one should be able to add fake login windows to
capture users’ passwords.

Access control to files allows a user who owns a file to decide who else can access the file. Privileged
access control means that the administrator has the right to determine access to files if needed, i.e. the
administrator can change users’ rights.

Address space defense for each process means that each process has protected addresses that unauthorized
processes cannot access.

The clearance of pages in a memory means that new pages loaded into the memory will not be able to find
information left behind by the previous pages. This makes it difficult for spyware to snoop into memory.

Auditing means that the system writes events in a network that may affect safety to a log file. The
administrator will then be able to get information about what is happening on the system by reading
the log files. The administrator can decide which events to monitor, and such events can be:

ƒƒ Someone tries to do something on the system that is not allowed.

ƒƒ Someone attempts to log on several times because the login fails.

18.3 Security Mechanisms in Windows

Microsoft has put a lot of resources into making the Windows operating system more secure. The reason
is that in recent years there have been more and more attacks against computer systems around the
world, and some of these attacks have been successful. Such attacks have managed to put the computer
systems of entire countries or large businesses, which can cost society billions of dollars.

The Windows operating system therefore has a highly developed security system, which is based on
access control and integrity levels. We will now look at how security system protects Windows processes
and data.

Security ID

There is a need to identify devices such as threads, which can perform operations on the system. Instead
of using names to identify such devices, the Windows operating system uses a SID (Security ID). A SID
is a number, and each SID is unique in the world.

A SID can be assigned to either a user or group of users in a network. When a process starts, the process
and the threads run under the user’s SID. Other threads will not be able to access the process unless
they have a SID with special authorization to do so.

114
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Security in Windows

Security Descriptor

Each process has information about its reliability that tells what privileges the user and the process have.
Each process has a Security Descriptor attached that a Security Descriptor points to for controlling lists.
These checklists contain access information that can deny access for users or groups of users.

Access to Objects

Central to the security of the Windows operating system is the protection of objects. Windows has a
comprehensive security model that prevents unauthorized access to objects, which requires that before
a thread can have access to an object, it must first specify what actions it will perform on the object.

Objects protected in the Windows operating system include files, hardware devices, mailslots, pipes,
processes, threads, events Mutexes, semaphores, shared memory, input/output ports, timers, volumes,
network shares, services, printers, etc.

18.4 Defense against Spyware

Spyware is a type of software that infiltrates a computer system without the owner being aware of it.
Spyware can pick up personal information about users on a PC and send it over the Internet back to
the originator of the spyware.

115 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Security in Windows

Windows Defender, also known as Microsoft Anti Spyware, is a program from Microsoft that has the
function to prevent, remove and isolate spyware in Microsoft Windows.

Windows Defender uses two mechanisms to detect spyware:

1. Scanning
2. Real Time protection

Windows Defender scans your computer and control programs against a database of information
about spyware.

Real Time Protection is a process that runs in the background and is looking for spyware that tries to
install itself or run on your computer.

Windows Defender can also remove ActiveX applications and block programs that start automatically
at Windows startup.

Windows Defender is included in Windows Vista, Windows 7 and Windows 8.

In Windows 8, Windows Defender is upgraded to an antivirus program that also looks for common
viruses, and not just spyware.

18.5 Virus Protection

A virus is a program that makes a copy of itself by attaching itself to another object. Viruses can attach
themselves to program files, documents and file system structures, and can run when the program
linked to it is executed. Viruses can also be in the memory and spread to the documents that a user
opens or saves.

A worm is an independent program that spreads by copying itself from one computer to another. The
usual way is via a network or attached to an e-mail. Worms are similar to viruses, though the difference
between viruses and worms is not so important in practice.

A computer can do without an antivirus program through careful use. One reason for this is that an
antivirus program is only one of many layers of security in a protected network. For example, routers
prevent viruses from entering via Web browsers through the Internet. E-mail servers also block viruses.

Earlier versions of the Windows operating system have not had their own antivirus program. However,
it has been possible to download it for free over the Internet. Microsoft Security Essentials (MSE) is
an antivirus program that fights viruses, Trojan horses and spyware mm. MSE is an antivirus program
designed for Windows XP, Windows Vista and Windows 7.

Today, Windows 8 has Windows Defender included, which is an upgrade to take over the role of MSE.

116
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Security in Windows

18.6 The firewall in Windows

A firewall is a part of a computer system or network designed to block unauthorized access and to allow
authorized access. Both hardware and software can implement firewalls.

The purpose of a firewall is to prevent unauthorized Internet users from accessing local network connected
to the Internet, especially intranets. The firewall will investigate all messages entering or leaving the
intranet through the firewall, and will also block messages if the safety criteria do not hold.

The firewall in Windows filters both incoming and outgoing packets. All incoming packets to your
computer are blocked unless they are a response to a request from your computer, while all outgoing
packets from your computer are permitted unless they violate a set rule.

Windows Firewall was first introduced as part of Windows XP Service Pack 2, and later versions of
Windows have improved the Firewall.

18.7 Windows Update

An important part of keeping a computer system safe is to obtain the latest upgrades to the operating
system. Microsoft is constantly working on new upgrades; this may be updated drivers or improvements
of code that have contained faults.

Many upgrades just give improvements in performance and functionality, but some are also security
updates to the system.

Windows include Windows Update, which is a program that updates the Windows operating system
for computers all over the world once a month. Using automatic updates, a Windows operating system
upgrades itself over the Internet without having to use a browser. The upgrade is usually the second
Tuesday of the month. However, critical upgrades can take place more often if necessary.

18.8 Control Questions

1. Mention some reasons why there is a need for security in a computer system.
2. What is the task of a Security ID (SID)?
3. What is the task of the Security Descriptor?
4. What possibilities does a user have to scan the computer for viruses in Windows?
5. What mechanisms does Microsoft Defender use to detect spyware?
6. What is new in Microsoft Defender in Windows 8?
7. Explain briefly how the firewall in Windows filters the packets.
8. What is the task for Windows Update and how does Windows Update work?

117
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System When Windows crashes

19 When Windows crashes

Many conditions can cause Windows to crash. If there is a system crash, Windows stops and displays
the blue screen.

19.1 The blue screen

The blue screen (The Blue Screen of Death) will appear in Windows when it has encountered a critical
system failure. The system then goes down to prevent the occurrence of serious faults that could damage
the system.

According to Microsoft, the reason Windows shows the blue screen is poorly programmed drivers or
hardware that is not working properly. The blue screen can also be a result of memory error, power
failure, overheated components or hardware used in an improper way.

Often, the blue screen appears when you install new software or new hardware. For example, if you have
installed a new driver and tried to reboot, you can get the blue screen. You will then have the opportunity
to undo the installation and get back the old configuration.

118 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System When Windows crashes

A question is why errors cause Windows to crash. Why does the system not ignore the error and just
continue? The reason is that the error that causes windows to crash is often part of a larger problem. To
let Windows continue will lead to more and more serious errors.

The blue screen has been in all versions of Windows since Windows 3.11.

19.2 Some reasons that Windows can crash

Windows may crash for various reasons. A detailed description of all circumstances that may cause
Windows to crash will be very extensive. In the following, we shall therefore only look at a few reasons:

Storage failure

The most common cause of Windows crashes is storage failure (Pool Corruption). Pool corruption can
occur when a driver runs into problems because a buffer cannot accommodate all the data it receives.
Pool corruption can also occur if a driver writes to a storage place that it had before, but has given up.

Page Fault in memory

An error on a page in memory (Page Fault) can lead to system crash.

Poor or no power supply

A driver or a function associated with the operating system will not work without power. Poor power
or a lack of power can thus cause Windows to crash.

Access Violation

An access violation occurs if there is an attempt to write to a page into a memory that is only allowed
to read, or if the system tries to read an address that does not exist.

Error in memory

If the Memory Manager detects that a data structure in a memory is corrupt, this can cause a system crash.

Hardware

Errors in hardware can cause system crashes. This includes errors on a disk when the memory manager
attempts to read data.

119
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System When Windows crashes

USB pen

If an error occurs while performing an operation on a USB pen, this can lead to a system crash.

Error in the file system

A fatal error in the file system can lead to system crash.

19.3 File dumping by system crash

Windows is set up to try to provide information of the system’s status when the system crashes. The
system writes memory contents to a file (Memory Dump).

Information about a system crash may occur at three levels.

Complete Memory Dump A Complete Memory Dump writes all the contents that were in the memory
when the crash occurred.

Kernel Memory Dump A Kernel Memory Dump only provides to read/write pages in kernel mode that
was in the memory when the crash occurred.

Small Memory Dump A Small Memory Dump provides an overview of drivers and running processes
and threads that were in the memory when the system crashed.

You can see what recommendations Windows has for dumping memory on your computer. In Windows,
you can open the Control Panel and select the following:

System and Security – System – Advanced System Settings

This brings up the System Properties. Choose Settings under Startup and Recovery.

19.4 Control Questions

1. Under what circumstances will Windows show the blue screen?
2. Mention some conditions that can make Windows crash.
3. Why does Windows ignore an error and just continue, instead of showing the blue screen?
4. How can Windows give information about what caused a system crash?

120
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Wordlist

Wordlist
The wordlist provides an explanation of some terms of the operating systems used in this presentation.

Asynchronous I/O

Asynchronous input/output lets a program continue to perform other tasks without waiting for an input/
output operation to be performed.

Cache

The purpose of caching is to make the input/output more efficiently.

Data is stored in a cache for fast access.

Cache Manager

A Cache Manager is a component in Windows that makes caching services for NTFS and other
file systems.

DO YOU WANT TO KNOW:

What your staff really want?

The top issues troubling them?

How to make staff assessments

work for you & them, painlessly?

How to retain your

top staff Get your free trial
FIND OUT NOW FOR FREE
Because happy staff get more done

121 Click on the ad to read more

Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Wordlist

Context switching

The purpose of context switching is to remove a thread from the processor and to load it in a new thread
for execution.

Critical Section

A Critical Section is a piece of program code that tries to use a resource that can only be used by one
thread at a time.

Deadlock

Deadlock occurs when two threads stop running because both threads are waiting for a resource that
the other threads control. There may be more than two threads involved in a deadlock.

Device Driver

A driver is a control program that enables communication between a hardware device and the operating
system. For the Windows operating system, a driver is necessary to use and communicate with
hardware equipment.

Executive Services

Executive Services in the Windows operating system helps applications in user mode to run on a
computer’s system. Executive Services include many basic services in the Windows operating system.

Fiber

A fiber is a lightweight thread. The operating system does not schedule fibers like threads, as fibers
belong to a thread that schedules them.

Handle

A handle is a type of pointer used in the Windows operating system. User applications do not have direct
access to data in objects. Therefore, applications use a handle in order to reach and change data in objects.

Hardware Abstraction Layer

A hardware layer (HAL) is a layer between the physical hardware and software that runs on a computer. A
task for HAL is to get the Windows operating system to be able to work with different hardware platforms.

Heap

The heap is a portion of memory allocated to a process. The process can use the heap to store variables
and objects associated with the process.

122
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Wordlist

Inter Process Communication

When a thread in one process interacts with a thread in another process it is called an Inter Process
Communication.

Interrupts

Events in a computer system can happen at any time, and then the processor must process them. Input/
output devices, timers, etc. send interrupts to the processor when they have tasks for the processor.

Job

Windows can group processes that cooperate. Such a group of processes is called a job.

Kernel

The kernel is an essential component that takes care of the operating system’s use of the processors.

The kernel controls the use of the processors, in addition to managing process planning, context switching,
interrupt signals, exception signals and multiprocessor synchronization.

Kernel Mode

The Kernel Mode is a privileged mode for processes belonging to the operating system.
All system resources in the operating system available for processes are running in the Kernel Mode.

Mutual Exclusion

Mutual Exclusion is when one thread has received exclusive control to a resource, so that other threads
cannot access the resource.

Multitasking

Multitasking is a technique that the operating system uses to share one processor for multiple threads.
The threads take turns in the processor, and a thread is only in the processor a short time before being
replaced by another thread.

Object

An operating system object is a data structure that represents a system resource such as a file, a thread,
a semaphore, a driver, an event, etc.

123
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Wordlist

Paging

Through paging, the contents of the memory are moved to a hard disk, so that the memory is released to
other processes or to the operating system itself. Data is divided into a set of equally sized blocks called
pages. The advantage of paging is that you do not have to keep all programs in the memory at the same
time, and that you always work with units of the same size.

Process

When a program starts and is loaded into a computer’s memory, we have a process.
In Windows, a process is a container in the memory of a program running on the computer.

Processor Scheduling

The task of Processor Scheduling is to handle the sharing of one or more processors among a group
of threads.

Quantum

A quantum is a time interval that a thread is allowed to run in on the processor before it must exit and
give the processor to another thread.

Challenge the way we run

EXPERIENCE THE POWER OF

FULL ENGAGEMENT…

RUN FASTER.
RUN LONGER.. READ MORE & PRE-ORDER TODAY
RUN EASIER… WWW.GAITEYE.COM

1349906_A6_4+0.indd 1 22-08-2014 12:56:57

124 Click on the ad to read more
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Wordlist

Registry

The registry is a system database that contains information to start Windows and to configure the system.
The registry keeps track of system settings and user preferences, as well as information about hardware
and applications on your computer.

Scheduler

Scheduling or process planning is taken care of by the Process Scheduler, which is part of the
operating system.

Synchronizing

Synchronization is a coordination of threads to avoid collisions. Synchronization ensures that only one
thread at a time can access a resource that cannot be shared.

Synchronous I/O

Synchronous input/output will stop a thread to execute when it performs an input/output. The thread
will not continue before the input/output operation is finished.

Timer

A timer controls events in an application code. This occurs with function calls at a determinable time
interval, for example every second. The code in the function will then be performed.

Thread

A piece of program code to be executed in the processor is called a thread, which is the dynamic part
of a process.

User Mode

User applications are running in user mode. For security reasons, applications running in user mode
have restricted access to data that belongs to the operating system.

Virtual Memory

Virtual Memory is a technique that provides applications with the impression that your computer has a
contiguous memory, even if some of the memory can be physically located on different storage media
such as the memory and a hard disk.

Volume

The hard drive(s) on a computer can be divided into volumes that are designated by letters C, D, E… A
single volume is only one partition, but a volume can span multiple partitions on one or more hard drives.

125
Download free eBooks at bookboon.com
An Introduction to
Windows Operating System Some references

Some references
Windows Internals. Mark E. Russinovich, David A. Solomon

Modern Operating Systems. Andrew S. Tanenbaum

Operating Systems. William Stallings

Windows Inside Out. Ed Bott, Carl Siechert, Craig Stinson

Windows System Programming. Johnson M. Hart

Wikipedia http://en.wikipedia.org/wiki/Main_Page

MSDN Library http://msdn.microsoft.com/en-us/library/ms123401.aspx

This e-book
is made with SETASIGN
SetaPDF

PDF components for PHP developers

www.setasign.com

126 Click on the ad to read more

Download free eBooks at bookboon.com