OPERATING SYSTEMS - WordPress.com System Evolution Networks and Communication Department Time Sharing Systems: to use computer system resources efficiently, multiprogramming was introduced

OPERATING SYSTEMS

Lecture 3: we will explore the role of the operating system in a computer

Networks and

Communication

Department

1

Outline

Networks and Communication Department

Define the operating system

Discuss the process of bootstrapping

Operating system evolution

Identify the components of an operating system

Memory manager

Process manager

Device manager

File manager

Computer System


7.4 Figure 7.1 A computer system

Computer System


A computer is a system composed of 2 major

components:

hardware and software.

Computer hardware: is the physical equipment.

Software is the collection of programs that

allows the hardware to do its job.

Computer System


Computer software is divided into 2 broad

categories:

The operating system and application

programs

Application programs: use the computer

hardware to solve users’ problems.

The operating system: controls the access to

hardware by users.

Operating System Introduction


Operating System


An operating system: is an interface between the

hardware of a computer and the user (programs or

humans) that facilitates the execution of other

programs and the access to hardware and software

resources

Two major design goals of an operating system are:

Efficient use of hardware.

Ease of use of resources.

Bootstrap Process


Bootstrap Process


The operating system is responsible for loading

other programs into memory for execution.

However, the operating system itself is a program

that needs to be loaded into the memory and be

run. How is this dilemma solved?

What is the Bootstrap process?


So a very small section of memory is made of ROM

and holds a small program called the bootstrap

program (Firmware).

The program is only responsible for loading the

operating system itself, or that part of it required to

start up the computer, into RAM memory.

Bootstrap Process


1. When the computer is turned on, the CPU counter

(a type of the CPU registers) is set to the first

instruction of this bootstrap program and

executes the instructions in this program.

2. When loading is done, the program counter is set

to the first instruction of the operating system in

RAM.

7.13 Figure 7.2 The bootstrap process

Operating System Evolution


Operating systems have gone through a long

history of evolution



Batch operating systems were designed in the

1950s to control mainframe computers.

At that time, computers were large machines that

used :

punched cards for input

line printers for output

tape drives for secondary storage media



Punched Cards



Tape

Drives





In Batch operating systems each program to be executed was called a job.

A programmer who wished to execute a job sends a request to the operating room along with punched cards.

The punched cards were fed to the computer by an operator.

If the program was successful, a print out of the result was sent to the programmer.

If not, a print out of the error was sent.

Batch System Example


The idea behind it was to collect a tray full of jobs in the

input room and then read them onto a magnetic tape

using a small (relatively) inexpensive computer, such as

the IBM 1401, which was very good at reading cards,

copying tapes, and printing output, but not at all good at

numerical calculations. Other, much more expensive

machines, such as the IBM 7094, were used for the real

computing.

Batch System Example

(a) Programmers bring cards to 1401.

(b) 1401 reads batch of jobs onto tape.

(c) Operator carries input tape to 7094.

(d) 7094 does computing.

(e) Operator carries output

tape to 1401.

(f) 1401 prints output.



Time Sharing Systems: to use computer system resources efficiently, multiprogramming was introduced. The idea is to hold several jobs in memory at a time, and only assign a resource to a job that needs it on the condition that the resource is available.

Multiprogramming brought the idea of :

time sharing: resources could be shared between different jobs, with each job being allocated a portion of time to use a resource.

Because a computer is much faster than a human, time sharing is hidden from the user—each user has the impression that the whole system is serving them exclusively.



Time Sharing Systems required a more complex operating

system.

They had to do scheduling which is : allocating resources to

different programs and deciding which program should use

which resource and when.

During this era, a new term was coined which is a process.

A job: is a program to be run.

But a process: is a program that is in memory and waiting

for resources.



Personal Systems: When personal computers were

introduced, there was a need for an operating system for

this new type of computer.

During this era, single-user operating systems such as DOS

(Disk Operating System) were introduced.



The need for more speed and efficiency led to the design

of:

parallel systems: multiple CPUs on the same machine.

Each CPU can be used to serve one program or a part of

a program, which means that many tasks can be

accomplished in parallel instead of serially.

The operating systems required for this are more

complex than those that support single CPUs.



Distributed Systems: Networking and internetworking

have created a new dimension in operating systems.

A job that was previously done on one computer can

now be shared between computers that may be

thousands of miles apart.

A program can be run partially on one computer and

partially on another if they are connected through a

network.

Resources can be distributed too.

Distributed systems combine features of the previous

generation with new duties such as controlling security.



Real-Time Systems: A real-time system is expected to do

a task within a specific time constraint. They are used

with real-time applications, which monitor, respond to or

control external processes or environments.

Such as : traffic control, patient monitoring and military

control systems.

The typical components of an operating

system


Operating System Components


Today’s operating systems are very complex.

An operating system needs to manage different

resources in a computer system.

It resembles an organization with several managers

at the top level.

Operating System Components


Each manager is responsible for managing

their department, but also needs to cooperate

with others and coordinate activities.

A modern operating system has at least four

duties: memory manager, process manager,

device manager and file manager.

OS Components- User interface


Each operating system has a :

user interface: a program that accepts requests

from users (processes) and interprets them for the

rest of the operating system.

A user interface in some operating systems, such

as UNIX, is called a shell. In others, it is called a

window to denote that it is menu driven and has

a GUI (graphical user interface) component.

User Operating System Interface - CLI


Command Line Interface (CLI) or command

interpreter allows direct command entry

Which is an operating system shell that uses

alphanumeric characters typed on a keyboard to

provide instructions and data to the operating system,

interactively.

User Operating System Interface - GUI


User-friendly desktop metaphor interface

Usually mouse, keyboard, and monitor

Icons represent files, programs, actions, etc

Various mouse buttons over objects in the interface cause various actions (provide information, options, execute function, open directory (known as a folder)

Many systems now include both CLI and GUI interfaces

Microsoft Windows is GUI with CLI “command” shell

Apple Mac OS X as “Aqua” GUI interface with UNIX kernel underneath and shells available

Solaris is CLI with optional GUI interfaces (Java Desktop, KDE)

OS Components- Memory manager


Although the memory size of computers has

increased tremendously in recent years, so has the

size of the programs and data to be processed.

Memory allocation must be managed to prevent

applications from running out of memory.

Operating systems can be divided into two broad

categories of memory management:

monoprogramming and multiprogramming.

Monoprogramming


In monoprogramming, most of the memory capacity is dedicated to a single program.

only a small part is needed to hold the operating system.

In this configuration, the whole program is in memory for execution.

When the program finishes running, the program area is occupied by another program.

Monoprogramming


Figure shows memory in

Monoprogramming a environment

Monoprogramming (Cont.)


There are several problems with this technique:

If the size of memory is less than the size of the program,

can the program run?

the program cannot be run

Inefficient use of memory and CPU time

Multiprogramming


In multiprogramming:

more than one program is in memory at the

same time and they are executed concurrently

with the CPU switching rapidly between the

programs

Multiprogramming


Figure shows memory in a multiprogramming

environment

Categories of multiprogramming


Multiprogramming (Cont.)




Nonswapping category:

This means that the program remains in

memory for the duration of execution

Two techniques belong to the this category:

Partitioning

paging



Swapping category:

This means that, during execution, the

program can be swapped between memory

and disk one or more times.

Two techniques belong to the this category:

Demand paging

Demand segmentation

Partitioning


The first technique used in multiprogramming.

Memory is divided into variable length sections

Each section or partition holds one program

The CPU switches between programs

With this technique, each program is entirely in

memory and occupying contiguous locations

Partitioning- how its work ?!


1. The CPU starts with one program, executing some instructions until it either encounters an input/output operation or the time allocated for that program has expired

2. Then, it saves the address of the memory location where the last instruction was executed and moves to the next program.

3. The same procedure is repeated with the second program

4. After all the programs have been served, the CPU moves back to the first program

Partitioning- how its work ?! (Cont.)


Priority levels can be used to control the amount of

CPU time allocated to each program

Its improves the efficiency of the CPU, but there are

still some issues

Partitioning- how its work ?! (Cont.)


Figure shows memory with

partitioning technique

Partitioning issues


The size of the partitions has to be determined

beforehand by the memory manager

what will happen if partition sizes are small or large?!

The memory manager can compact the partitions to

remove the holes and creates new partitions, but this

creates extra overhead on the system

Paging


Paging improves the efficiency of partitioning

Memory is divided into equally sized sections called

frames

Programs are also divided, into equally sized

sections called pages

The size of a page and frame is usually the same

and equal to the size of the block used by the

system to retrieve information from a storage

device

Paging - how its work!!


A page is loaded into a frame in memory

The program does not have to be contiguous in

memory

There is no need for the new program to wait

until set of contiguous frames are free before

being loaded into memory

Paging - how its work!!


Figure shows memory with paging technique

Paging issues


The whole program still needs to be in memory

before being executed

Ex: A program has six pages, cannot be loaded into

memory if there are only four unoccupied frames

Demand paging


In this technique the program is divided into

pages

but the pages can be loaded into memory one

by one, executed, and replaced by another

page.

In addition, a page can be loaded into any

free frame

Demand paging (Cont.)


An example of demand paging is shown in figure

below. Two pages from program A, one page from

program B, and one page from program C are in

the memory

Demand segmentation


A technique similar to paging is segmentation

A programmer thinks in terms of modules

A program is usually made up of a main program and subprograms

In demand segmentation, the program is divided into segments that match the programmer’s view

These are loaded into memory, executed, and replaced by another module from the same or a different program

since segments in memory are of equal size, part of a segment may remain empty

Demand segmentation


An example of demand

segmentation is shown in figure

Demand paging and segmentation


Demand paging and segmentation can be

combined to further improve the efficiency of the

system

A segment may be too large to fit any available

free space in memory

Memory can be divided into frames, and module

can be divided into pages

The pages of a module can then be loaded into

memory one by one and executed

Virtual Memory


Virtual memory


Demand paging and demand segmentation mean that, when a program is being executed, part of the program is in memory and part is on disk.

This means that, for example, a memory size of 10 MB can execute 10 programs, each of size 3 MB, for a total of 30 MB.

At any moment, 10 MB of the 10 programs are in memory and 20 MB are on disk.

There is therefore an actual memory size of 10 MB, but a virtual memory size of 30 MB

Virtual memory (Cont.)


Virtual memory, which implies demand paging,

demand segmentation or both, is used in almost all

operating systems today.

Figure below shows the concept.

OS Components- Process manager


A second function of an operating system is process management, but before discussing this concept, we need to define some terms.

A program is a non-active set of instructions stored on disk. It may or may not become a job

A program becomes a job from the moment it is selected for execution until it has finished running and becomes a program again.

A process is a program in execution. It is a program that has started but has not finished.

State diagrams


The relationship between a program, a job and

a process becomes clearer if we consider how

a program becomes a job and how a job

becomes a process.

This can be illustrated with a state diagram

that shows the different states of each of these

entities.

State diagrams (Cont.)


Figure shows a State diagram with boundaries between program, job and process

State diagrams (Cont.)


A program becomes a job when selected by the operating system and brought to the hold state

It remains in this state until it can be loaded into memory

When there is memory space available to load the program totally or partially, the job moves to the ready state. It now becomes a process

It remains in memory and in this state until the CPU can execute it, moving it to the running state at this time

When in the running state, one of three things can happen:

The process executes until it needs I/O resources, it goes into the waiting state

The process exhausts its allocated time slot, it goes into the ready state

The process terminates, it goes into the terminated state

Schedulers


To move a job or process from one state to

another,

The process manager uses two schedulers

The job scheduler

The process scheduler

Schedulers (Cont.)


Job scheduler:

Moves a job from the hold state to the ready state or

from the running state to the terminated state

(Figure shows job scheduler)

Schedulers (Cont.)


Process scheduler:

Moves a process between the running, waiting, and

ready states many times before it goes to the

terminated state and is no longer a process

As shown in the figure below

Queuing


In reality, there are many jobs and many

processes competing with each other for

computer resources.

Queuing


To handle multiple processes and jobs, the process manager uses queues (waiting lists).

A job control block or process control block is associated with each job or process.

This is a block of memory that stores information about that job or process.

The process manager stores the job or process control block in the queues instead of the job or process itself.

Queuing (Cont.)


Figure shows the circulation of jobs and processes

through three queues ( the job queue, the ready

queue , and the I/O queue)

Queuing (Cont.)


The job queue holds the jobs that are waiting for

memory

The ready queue holds the processes that are in

memory, ready to be run and waiting for the CPU

The I/O queue holds the processes that are waiting

for an I/O device (one queue for each I/O device)

Selecting next job or process


The process manager can have different policies for

selecting the next job or process from a queue:

It could be first in first out (FIFO), shortest length

first, highest priority first, and so on

Process synchronization


The whole idea behind process management is

to synchronize different processes with

different resources.

Whenever resources can be used by more than

one user (or process, in this case), we can have

two problematic situations: deadlock and

starvation.

Deadlock on a narrow bridge


Deadlock


There are four necessary conditions for deadlock to occur:

Mutual exclusion only one process can hold a resource

Resource holding A process holds a resource even though it cannot use it until other resources are available

No preemption The OS cannot temporarily reallocate a resource

Circular waiting All processes and resources involved form a loop

Deadlock occurs when the operating system does not

put resource restrictions on processes.

Deadlock example


Assume that there are two processes, A and B

Process A is holding File 1, and it has requested File

2 and cannot release File 1 until it acquires File 2

Process B is holding File 2, and it has requested File

1 and cannot release File 2 until it acquires File 1

Files in most system are not sharable

If there is no provision in this situation to force a

process to release a file, deadlock is created (as

shown in figure below)

Deadlock example (Cont.)


How the system preventing deadlock?


One solution is not to allow a process to start running

until the required resources are free, but this creates

another problem (Starvation)

The second solution is: do not allow one of deadlock

conditions to happen (i.e. limit the time a process can

hold a resource)

The dining philosophers problem


Starvation


Starvation is the opposite of deadlock. It can happen

when the operating system puts too many resource

restrictions on a process.

Starvation example


For example, imagine an operating system that specifies that a process must have possession of its required resources before it can be run

In Figure below,

Imagine that process A needs two files, File1 and File2

File1 is being used by process B and File2 is being used by process E

Process B terminates and releases File1

Process C, which needs only File1, is allowed to run

Process E terminates and releases File2

Process D, which needs only File2, is allowed to run

Starvation example (Cont.)


OS Components- Device manager


The device manager, or input/output manager,

is responsible for access to input/ output

devices.

There are limitations on the number and speed

of input/output devices in a computer system.

The device manager is responsible for the

efficient use of input/output devices

Device manager responsibilities


The device manager monitors every input/output device constantly to ensure that the device is functioning properly.

The device manager maintains a queue for each input/output device or one or more queues for similar input/output devices.

The device manager controls the different policies for accessing input/output devices. i.e. it may use FIFO for one device and shortest length first for another

OS Components- File manager


Operating systems today use a file manager to control access to files. The responsibilities of a file manager: Controls access to files. The type of access can vary

(read, write, execute)

Supervises the creation, deletion, and modification of files.

Controls the naming of files.

Supervises the storage of files. ( how, where,.. etc)

is responsible for archiving and backups.


86

Any Q

uest

ions

?

References


Behrouz Forouzan and Firouz Mosharraf,

“Foundations of computer science”, Second edition,

chapter7, pp. 188-203

Documents

OPERATING SYSTEMS - WordPress.com System Evolution Networks and Communication Department Time Sharing Systems: to use computer system resources efficiently, multiprogramming was introduced