Skip to content

Lecture No. 04

Dated: 26-04-2025

Layered Approach

The benefit of this approach is modularity.
The system is made up of different layers.
The bottom layer is the hardware layer and the upper layer is called the user interface layer.
The middle (layer - M) is the operating system1 layer which consists of data structures and routines which can be invoked by higher layers.
It also simplifies debugging and system verification.
Though it is less efficient as it creates overhead with each system call.

Example

  • The operating system1 by Dijkstra
  • IBM's OS/2

Micro Kernels

This method structures the operating system1 by removing all non essential components of the kernel and implement system and user level programs.
This results into smaller (micro) kernels which provide communication facilities between client program and various services also running in the user space.
It also provides minimum process and memory management.
This provides easy of extending the operating system1 or making it run on different hardware designs because most of the time, the kernel doesn't need to be modified as the new services are added to user space and even if kernel does need to be modified then those changes are very small.

Examples

  • Mach
  • MacOS X Server
  • QNX
  • OS/2
  • Windows NT

Virtual Machines

Conceptually, a computer is made up of different layers.

  1. Hardware
  2. Kernel
  3. System programs
  4. Application programs

Each layer has access to layers below it.

cs604_i_4_1.png

Non virtual machines vs virtual machines

There are 2 advantages of virtual machines despite the fact that they are hard to implement.

  • They provides a robust level of security.
  • System development to be done without disrupting normal system operation.

Java Virtual Machine

The JVM loads, verifies and executes the java bytecode.

cs604_e_4_1.svg

Examples

Following are the examples of virtual machines.1

  • VMWare
  • Virtual PC
  • Virtual Box

System Design and Implementation

Design Goals

At the highest level, the design of the system will be affected by the choice of hardware and type of system.
Beyond this, requirements might be much harder to specify and can be divided into 2 groups.

  • User goals
  • System goals

An important design goal is separation of mechanism and policies.

Mechanism

They determine how to do something and a general mechanism is more desirable.

Example

CPU Protection2

Policy

Determine what will be done.

Example

Initial value in the counter used for CPU protection.2

This separation is important because policies can change over place or time.

Implementation

Back then, operating systems1 were written in assembly language but now a days, we use C or C++ for that as the code is written faster, is more compact, easier to understand and easier to port.

Unix / Linux Directory Structure

History

Dennis Ritchie and Ken Thomsom wrote UNIX at the Bell Labs in 1969.
It was initially written in assembly language and a higher level language called Bit and was later converted to C.
Linus Torvalds, an undergraduate student at the University of Helsinki, Finland, wrote Linux in 1991.

UNIX uses a hierarchical file system structure consisting of a root directory denoted as / with other directories and files hanging under it.

A path is separated by / and last component of it can be either a file or a directory.

When you login, the system places you in home directory (also called the login directory) and it is referred to, by

  • ~ or $PATH in bash, sh, tsh
  • $path in csh, tcsh

The shells understand relative and absolute paths.

Absolute path starts with / meanwhile relative path can start with either

  • Home Directory (~)
  • Current Directory(.)
  • Parent Directory(..)

Commonly Used Directory Structure in Linux

/

The root directory is similar to C:\, D:\ drive in windows etc. But the difference here is that unlike windows, everything is under one, singular root node.

/bin

These are binaries or executables, available to all users.

/boot

This directory contains essential files for system boot, including kernel image.

/dev

This directory contains all the devices which are available to Linux.

/etc

Contains the configuration files (which are text files and are easy to edit).

Example

  • /etc/inittab contains what processes will be started alongside system boot.
  • /etc/fstab identifies file systems and their mount points.
  • /etc/passwd is where users are defined.

/home

Every available user is present here.

Example

If there are multiple users, say chris and tom then you will find

  • /home/chris
  • /home/tom

/lib

Shared libraries and kernel modules are stored in this directory.
These can be dynamically linked and are very similar to DLLs in windows.

/lost+found

Here, linux keeps files which need to be restored after a crash or when partition has not been properly unmounted before a shutdown.

/mnt

File systems can be mounted anywhere but this is the convenient place to mount temporary file systems.

/opt

Often used for storage of large applications packages.

/proc

This is a virtual directory and doesn't exist physically.
Here, the system processes are stored.

/root

The home directory for the super user (root).

/sbin

Utilities used for system administration such as

  • halt
  • ifconfig
  • fdisk

The /sbin/init.d are scripts which are used by /sbin/init to start the system.

/tmp

Used to store temporary files.

/usr

A shareable, read-only directory containing user applications and their support files. Key subdirectories include:

  • /usr/X11R6: X Window System files.
  • /usr/bin: user-accessible commands.
  • /usr/doc: documentation for /usr applications.
  • /usr/include: C compiler header files.
  • /usr/include/g++: C++ compiler header files.
  • /usr/lib: libraries, binaries, and object files not directly executed by users.
  • /usr/local: locally installed software, safe from system updates.
  • /usr/man: manual pages.
  • /usr/share: architecture-independent, read-only data files.
  • /usr/src: source code for installed applications and the kernel.

/var

Contains variable data files such as

  • Logs in /var/log
  • Mails in /var/mail
  • Spools in /var/spool

References

  1. Read more about operating systems. ↩↩↩↩↩↩

  2. Read more about CPU protection. ↩↩