Yeditepe University
Department of Computer Engineering
CSE 232
Systems Programming
LECTURE NOTES #8
INTRODUCTION TO OPERATING SYSTEMS
1. Introduction
A computer system can be divided into 4 components:
-
the hardware (CPU, memory, input/output devices, etc.),
the operating system,
the system programs (word processors, spread sheets, accounting softwares,
compilers,…)
the application programs.
Application Programs
System Programs
Operating
System
Hardware
Figure 1. Hierarchy of computer components
1
2. Computer System Structure
printer
disk
CPU
disk
disk
controller
printer
controller
tape-drive
controller
system bus
memory
controller
memory
Figure 2. Hardware of a computer system
2.1. Run-time Environment
-
Contains a set of service routines that are available for use during program execution (eg. I/O
functions:
Program invokes a service routine, OS takes care of status testing, counting of bytes
transferred, error checking)
-
Provides facilities for managing the resources
(eg. CPU is switched among user jobs according to some predefined policy)
2.2. Input and Output (I/O)
I/O devices:
 block devices
 character devices
I/O types:
 Programmed I/O
 Interrupt driven I/O (Interrupt service routine)
 Direct memory access (DMA)
2
Example 1: Keyboard
Keyboard interface (controller)
Bus
Port 0
(Status)
CPU
Keyboard
electronics
Port 1
(Data)
Figure 3. Keyboard controller
Port: an addressable register
- input only ports: CPU only reads from it
- output only ports: CPU only writes to it
- ports for both input and output
Every port has a number. A port can be referred to in an assembly language program by its
number, such as:
IN 1
Status port: checked by the CPU to see if a key has been pressed and data is ready to be sent to
CPU
Data port: when a key is pressed, its ASCII code is found in the data port
Polling: CPU continually examines the status port until a key is pressed (Programmed I/O)
Ex: Reading a character from keyboard
KEYIN: IN
ANA
JZ
IN
RET
0
A
KEYIN
1
; read status port into A
; test A (10H – key is pressed, 00H – otherwise)
; jump if 00H
; take data into A
3
Example 2: Printer
Printer interface (controller)
Bus
Port 2
(Control)
CPU
Printer
interface
electronics
Port 3
(Data)
Figure 4. Printer controller
Data port: character to be printed is put in this port
Control port: when the character is ready in the data port, strobe signal (that tells the printer that
it can take the character in data port) is sent.
A timing loop must be used in the CPU so that the characters are never passed to the printer at a
rate faster than it can accept.

Special characters:
- carriage return: causes the printer head to move to the left side of the paper
- line feed: causes the printer to advance the paper by one line
Ex: Writing a character to printer
CHAROUT:
OUT
CALL
MVI
OUT
MVI
OUT
RET
3
DELAY
A, 10H
2
A, 00H
2
; put char. in accumulator into output port
; make sure printer is ready
; set up strobe word (10H – char. in data port is available for printing)
; put it into control port to start strobe signal to printer
; reset strobe (00H – char. in data port is not available for printing)
; put it into control port to finish strobe signal
4
Interrupt Processing
An interrupt is a signal that causes the computer to alter its normal flow of instruction execution.
eg. divide by zero, time out, I/O operation
The interrupt automatically transfers control to an interrupt processing routine which takes some
action in response to the condition that caused the interrupt, and then control is returned to the
program at which its execution was interrupted.
program
interrupt routine
interrupt occurs
Some interrupt types:
- I/O interrupt: Generated by an I/O channel or device.
- Program interrupt: Generated by some condition that occurs during program execution (eg.
divide by zero, illegal machine instruction, etc.)
- Timer interrupt: Generated by an internal timer within the CPU
- Supervisor call interrupt (SVC): A SVC instruction generates an interrupt that transfers control
to an OS service routine. The CPU switches from user mode to supervisor mode.
When an interrupt occurs, the status of the CPU is saved (registers, program counter etc.) and the
interrupt service routine is executed. After the interrupt is processed, they are restored by the
interrupt routine and control is transferred to the program. Saving and restoring the processor
status is called context switching.
Interrupt priority: During the processing of an interrupt, all interrupts of equal or lower priority
are inhibited, higher priority interrupts are allowed. At the end of an interrupt, if there are more
than one interrupt pending, one with the highest priority is executed.
Interrupt masking, is used to prevent certain interrupts from occurring while the first one is being
processed.
5
Example 3: Disk
Bus request line
CPU
Disk controller
RAM
R/W line
DMA
Device
other
electronics
interface
registers
Address bus
Data bus
Control bus
Figure 7. Disk controller
Interface registers
Status port
Control port
DMA address
registers
DMA (Direct Memory Access): is a procedure for transferring data directly between controller
and memory without the help of CPU.
DMA device: is like a separate CPU with functions related to data transfer, such as:
- shuts the CPU down for short periods
- seizes control of the system bus
- transfers data between memory and external device without the intermediation of CPU
Disk controller: accepts commands from CPU such as:
- select disk drive
- set up DMA registers
- return current sector address
- return current track address
- move head in/out one track
- move head to track 0
- read/write current sector
Status port: contains the result of a query from CPU or status
Control port: commands for the controller are placed
DMA address registers: they receive addresses from CPU
6
3. Operating Systems
An operating system is a program that acts as an intermediary between a user of a computer and
the computer hardware.
Main functions of an operating system are:
1. The operating system is an interface between the user and the computer
 provides access to the to the utility programs such as editors, debuggers, etc.
 starts the execution of the programs
 provides low level I/O routines to the user
 controls access to the files
 provides access functions for the protection of the resources and data from unauthorized
users
 detects errors such as memory error, device failure, arithmetic overflow, etc. and provides
error handling routines
 collects performance statistics
2. The operating system manages the resources of the system
 controls and allocates resources
System resources are memory, CPU, I/O devices, etc.
3.1. Historical Review of Operating Systems
Monitors
-
Monitors are the simplest operating systems.
Single user systems
Allow user interaction
Simple Batch Systems
The user prepares a job (which consists of program, data and some control information), submits
it to the computer operator and receives the output after the program is executed. The job is
usually punched on cards and the output is usually printed. To speed up processing, jobs with
similar needs can be batched together and run as a group. Thus, the operator sorts the programs
with similar requirements into batches, runs each batch and sends the output to the programmer.
Therefore in a batch operating system there is a lack of interaction between the user and the job
while that job is executing.
The operating system is always in memory and its job is to transfer control automatically from
one job to the next.
As the I/O devices are slower than the speed of the CPU (the difference may be three orders of
magnitude or more), the CPU is often idle. For using the system more efficiently, cards can be
7
read from the card reader onto the disk. The operating system records their location on disk in a
table. When a job requests the printer, the output is copied into a system buffer and is written to
the disk. When the job is completed, the output is printed. This form of processing is called
spooling (simultaneous peripheral operation on line). Spooling is also used for processing data at
remote sites. The CPU send the data via communication paths to a remote site and the processing
is done with no CPU intervention. The CPU just needs to be notified when the processing is
completed, so that it can spool the next batch of data.
Spooling overlaps the I/O of one job with the computation of other jobs. During the execution of
one job, the spooler may be reading the input of another job while printing the output of a
different job.
disk
card reader
I/O
line printer
CPU
Figure 8. Spooling
Multiprogrammed Batch Systems
Spooling provides a job pool on disk. In order to increase CPU utilization, jobs in the pool may
be scheduled in different ways, such as, first-come first-served, shortest job first, priority basis,
etc.
The most important aspect of job scheduling is the ability to multiprogram. The operating system
keeps several jobs (a subset of jobs in the job pool) in memory at a time. During the execution of
one job, if it waits for an I/O operation to complete, the operating system switches to and
executes another job. In this way, as long as there is always some job to execute, the CPU will
never be idle.
Time-Sharing Systems
In multiprogrammed batch systems, the user cannot interact with the program during its
execution. In time-sharing or multitasking systems, multiple jobs are executed by the CPU
switching between them, but the switching occurs so frequently that the user may interact with
each program while it is running. Time-sharing operating systems:
- uses CPU scheduling and multiprogramming,
- uses time-slice mechanism,
- allows interactive I/O,
- allows many users to share the computer simultaneously.
8
2.5. Personal Computer (PC) Systems
A computer system dedicated to a single user is referred to as a PC. In the first PCs, the operating
system was neither multiuser nor multitasking (eg. MS-DOS). The operating system concepts
used in mainframes and minicomputers, today, are also used in PCs (eg. UNIX, Microsoft
Windows NT, Macintosh OS).
2.6. Parallel Systems
Parallel systems have more than one processor. In multiprocessor systems (tightly coupled),
processors are in close communication, such as they share computer bus, clock, memory or
peripherals.
2.7. Distributed Systems
Distributed systems also have more than one processor. Each processor has its local memory.
Processors communicate through communication lines (eg. Telephone lines, high-speed bus,
etc.). Processors are referred to as sites, nodes, computers depending on the context in which they
are mentioned. Multicomputer systems are loosely coupled. Example applications are e-mail,
web server, etc.
2.8. Real-time Systems
Real-time systems are special purpose operating systems. They are used when there are rigid time
requirements on the operation of a processor or the flow of data, and thus it is often used as a
control device in a dedicated application (eg. fuel injection systems, weapon systems, industrial
control systems, …). It has well defined, fixed time constraints. The processing must be done
within the defined constraints, or the system fails.
Two types:
- Hard real-time systems guarantee that critical tasks complete on time.
- In Soft real-time systems, a critical real-time task gets priority over other tasks, and the
task retains that priority until it completes.
3.2. User Interfaces for Operating Systems
Include:
- simple command language
- a menu
- graphical representation of programs and data
- I/O routines (service routines)
9
Operating System Features




Bootstrap program: initializes CPU, loads OS kernel into memory
OS waits for event, signaled by an interrupt either from a hardware or software (system
call, monitor call), interrupt vector
Dual mode of operation: user mode, kernel mode (supervisor mode)
Hardware Protection (base, limit registers)
user program 1
User programs
run in user mode
user program 2
Kernel call
4
3
1
service
procedure
dispatch table
Operating
system
runs in
kernel mode
2
Figure 9. How a system call can be made
BIOS (Basic Input Output System)
It is an interface between the operating system and hardware. Changes from one computer to
another.
Ex:
- transfers a single character to printer
- reads a sector from disk
BIOS routines of CP/M:
BOOT
WBOOT
CONST
CONIN
CONOUT
LIST
PUNCH
HOME
READER
SELDSK
SELTRK
SETSEC
SETDMA
READ
WRITE
10
To call a BIOS routine:
Jump Table: (found at the beginning of BIOS routines)
1 byte
BBASE
JMP
JMP
JMP
JMP
…
…
To call WBOOT:
2 bytes
BOOT
WBOOT
CONST
CONIN
CALL BBASE+3
11
4. Some Operating System Functions
4.1. Process Scheduling
CPU is assigned to processes (tasks) by the operating system. In a multiprogramming system,
process scheduling is the management of the CPU by switching control among the various
processes according to some scheduling policy.
blocked
waiting for
some event
awaited event
has occurred
time slice expired
running
ready
dispatch
3.2. Memory Management
Real Memory
Real memory is divided into partitions, each process is assigned to a different partition:
- Fixed partitions
- Variable partitions
- Relocatable partitions
Virtual Memory
In a virtual memory system, user programs are allowed to use a large continuous virtual address
space (virtual memory), which may be even larger than the total amount of the real memory
available. Portions of the virtual memory are mapped to the real memory as they are needed by
the program. Demand paging is one of the mapping methods. This mapping process is
completely invisible to the user.
12
Virtual memory
Real memory
mapping
disk
virtual address
mapping
real address
3.3. File Management
File management function of an operating system is an intermediate stage between the user
program and the I/O supervision.
user prog.
record
file
information
logical request
file
manager
block
physical request
I/O
supervisor
I/O interrupt
supervisor I/O
I/O
channel
13
3.4. Job Scheduling
Job scheduling is the task of selecting the next user job to begin execution. In a
multiprogramming environment, job scheduler specifies the order of job execution.
input queue
job
scheduler
active
jobs
dispatcher
CPU
3.5. Resource Allocation
The resources such as CPU, memory, I/O channels are needed by all user jobs, and their
allocation is handled automatically by the operating system. Some problems to be handled by the
operating system are:
- mutual exclusion
- process synchronization,
- deadlock avoidance
- security and protection
14