Scheduling

What is Scheduling?

 Deciding which tasks to process, for how long, and in what order is achieved
through scheduling algorithms
 A CPU is responsible for processing tasks as fast as possible
 Different algorithms are used to prioritise and process tasks that need CPU time
 The algorithms have different uses, benefits and drawbacks.

Scheduling categories

 Preemptive: allocates the CPU for time-limited slots

 Non-preemptive: allocates the CPU to tasks for unlimited time slots

Preemptive Scheduling

 Allocates the CPU for a specific time quantum to a process

 Allows interruption of processes currently being handled
 It can result in low-priority processes being neglected if high-priority processes
arrive frequently
 Example algorithms include Round Robin and Shortest Remaining Time First

Non-Preemptive Scheduling

 Once the CPU is allocated to a process, the process holds it until it completes
its burst time or switches to a 'waiting' state
 A process cannot be interrupted unless it completes or its burst time is reached
 If a process with a long burst time is running, shorter processes will be neglected
 Example algorithms include First Come First Serve and Shortest Job First

Scheduling Algorithms

Round robin (RR)

 RR is a pre-emptive algorithm, equally distributing processor time amongst all

processes
 Each process is given a time quantum to execute
 Processes that are ready to be worked on get queued
 If a process hasn’t been completed by the end of its time quantum, it will be
moved to the back of the queue
 Round robin scheduling algorithm

First-Come-First-Served (FCFS)

 FCFS is non-preemptive, prioritising processes that arrive at the queue first

 The process currently being worked on will block all other processes until it is
complete
 All new tasks join the back of the queue

First-Come-First-Served scheduling algorithm

Shortest job first (SJF)

 SJF is non-preemptive, where all processes are continuously sorted by burst

time from shortest to longest
  When new processes arrive on the queue, they are prioritised based on their
burst time in the next cycle
 Shorter jobs are placed at the front of the priority queue
 Longer jobs have lower priority, so they are placed at the back

Shortest job first scheduling algorithm

Shortest remaining time first (SRTF)

 SRTF is a preemptive version of SJF, where processes with the shortest

remaining time are higher priority
 Time quantum is set, and if a task doesn’t complete in time, it will be re-queued
for further processing
 Before the next cycle starts, all processes are inspected and ordered by the
shortest remaining time to complete
 Memory Management
What is Memory Management?

 Memory management is a fundamental role of the operating system, dealing with

the allocation and deallocation of the computer's primary memory
 When a user opens an application, its data is loaded from storage into active memory so
that it can run smoothly
 When a user opens a file from the file system, e.g. word document, the CPU loads
this file data, as well as application data, into the primary memory
 Primary memory is a limited resource in the system, so it needs careful management
 Efficient allocation of memory enables a system to continue multitasking
 Memory management is made more efficient through 3 techniques:
o Paging
o Segmentation
o Virtual Memory
 Links between different types of memory

Exam Tip

 The main benefit of memory management is to make computer systems run smoothly.
Most users don’t realise that as they effortlessly move between applications, it’s made
possible because the OS is rapidly reallocating memory depending on the task that the
user is completing.
 Make sure you can name one benefit and one drawback for each memory management
method in this revision note.

Paging
What is Paging?

 Data stored in memory will lead to the smooth running of applications

 Paging is a method of chunking the primary memory into equal-sized blocks
 When an application is launched, data will be moved from the hard disk into Pages for
faster access
 As users move between applications, memory is dynamically allocated
 Pages will be taken away from applications not in active use and granted to applications
that are in active use
 Paging can lead to internal fragmentation
 If a 200KB file is divided into four 64KB Pages, the last Page would have 8KB of unused
space
 Unused space in a Page is wasteful because other unrelated data cannot be stored in this
Page
 Over time, more pockets of wasted space will exist across the memory; this process is
called internal fragmentation
 The image below shows a single 64KB Page with 4KB of unoccupied space
 The box below this shows many Pages, each with varying sizes of internal fragments
 Internal fragmentation

Segmentation
What is Segmentation?

 Segmentation is a method of chunking memory into blocks that correspond to

different types of data needed by an application
 A video editing application may have a Segment for video data, audio data and special
effects
 Segments are not all the same size; they are sized depending on their allocated data
 Segmentation is space-efficient due to only allocating space depending on the amount an
application needs
 Segmentation can lead to external fragmentation
 As Segments fill up the memory, physical gaps reduce the maximum size of new
Segments that can be allocated
 Below (left) shows different application data assigned to a Segment
 The arrangement of data in the segment becomes more fragmented over time because
as blocks are taken away it's not possible to guarantee a new block will occupy the same
amount of space
 Below (right) shows a defragmented version of the Segment to highlight the total unused
space
 External fragmentation

Virtual Memory
Virtual Memory

 If a computer is running low on primary memory, it can make secondary storage act as
an 'extension' of the main memory
 The operating system can offload data from the primary memory into virtual memory
 Virtual memory creates an illusion of a larger memory and enables applications to
continue to multitask
 However, accessing data in virtual memory is considerably slower compared to RAM
 Solid-state drives are faster than traditional hard-disk drives, but neither are as fast as
RAM
 Over-reliance on virtual memory can lead to performance issues

Memory
Management Description Example Benefits Drawbacks
Technique

A process needing Facilitates efficient

Divides memory 200KB of memory is memory This can lead to
Paging into fixed-sized divided into four 64KB management and internal
blocks (pages) pages, leaving 8KB enables the use of fragmentation
unused in the last page virtual memory
 In a video editing
Divides memory
application, different
into variable-sized Allows for intuitive This can result in
segments may be
Segmentation segments based on and efficient external
created for video data,
logical parts of a memory access fragmentation
audio data, effects, and
process
UI elements

When memory-intensive Slower to access

Allows more
Uses hard drive applications exceed the than physical
extensive programs
space as an available RAM, the OS memory, which
Virtual Memory to be run and
'extension' of moves less frequently degrades
facilitates effective
RAM accessed pages to the performance if
multitasking
hard disk overused
Exam Tip

 Virtual memory is a method of memory management.

o Not to be confused with Virtual Storage, which is storing files in the cloud.

Worked example

Describe how the operating system would use virtual memory to load program C when there's
not enough space in physical memory.

[3]

Answer that gets full marks:

The operating system can use virtual memory to act as an extension of the computer's primary
memory. This means less-critical data can be offloaded from the primary memory into virtual
storage, useful when a higher-priority set of processes require immediate attention. If the OS
offloaded data from the RAM into virtual memory, this would free up space for program C to be
loaded into RAM.

Acceptable answers you could have given instead:

Program C can be loaded into the RAM if the operating system moves files and data into virtual
memory. Virtual memory acts as an extension of the RAM.
Worked example

Imogen buys a desktop computer. It comes with an operating system installed.

Describe two ways that an operating system could manage physical memory.

[4]
How to answer this question

 Recall two methods of memory management (Paging, segmentation, virtual memory)

 Explain how each of these works and how they assist in memory management

Example answer that gets full marks:

Imogen's operating system could manage physical memory in two ways. Paging, which is a
method of dividing memory into fixed-size chunks known as pages. Application data can be
allocated and deallocated to pages making it a flexible system depending on the activity of the
user.

The system could also use virtual memory, a method of extending the available physical memory
by using a portion of the hard drive. This allows more programs to run simultaneously and
enhances the overall system performance.

Acceptable answers you could have given instead:

Imogen's computer uses paging to manage memory. This means it breaks down the memory into
fixed-size pieces and swaps them in and out as needed.

The computer also uses something called virtual memory. It uses part of the hard drive to act like
extra memory, so more programs can run at the same time, and everything works faster.

Shortest remaining time first scheduling algorithm

Comparison and Summary of Scheduling Algorithms

Algorithm Benefits Drawbacks

All processes get a fair share of the
CPU Choosing the right time quantum can
be difficult
Round Robin Good for time-sharing systems
This can lead to a high turnaround time
Predictable, as every process gets equal and waiting time for long processes
time
This can lead to poor performance if a
Simple and easy to understand long process arrives before shorter
First Come, processes
First Served Fair in the sense that processes are
served in the order they arrive High-priority tasks wait for their turn
in the queue
Requires knowing the burst time of
Minimises waiting time processes in advance
Shortest Job
First
Efficient and fast for short processes Long processes can starve if short
processes keep arriving
Ideal for jobs that have shorter burst
Like SJF, it requires knowing the burst
times
time of processes in advance
Shortest Time
Remaining It is preemptive, so it can be aligned
High context switching overhead due
with CPU for best performance (time
to preemption
quantum)

 The suitability of a scheduling algorithm largely depends on the specific scenario

and the system requirements
 A drawback in one scenario may not be a drawback in another

Worked example

A company makes anti-virus software. When running anti-virus software, an operating

system uses a scheduling algorithm to allocate CPU time to the anti-virus software.

Explain why a First Come First Served scheduling algorithm would not be suitable in
this situation.

[2]

How to answer this question:

 Think of the conditions that anti-virus software runs optimally

  Recall the way the FCFS algorithm works and its benefits and drawbacks
 Link how the optimal running of anti-virus is incompatible with FCFS scheduling

Example answer that gets full marks:

Anti-virus software is high-priority because it scans the operating system constantly,

looking for threats. When a threat is detected, anti-virus will quarantine or eliminate
them. To work effectively, anti-virus software needs high-priority access to CPU time.

Using FCFS could delay these critical tasks if many other processes are in the queue
ahead of the anti-virus software. Other less crucial tasks could get CPU time before the
anti-virus process, leading to potential security risks.

Acceptable answers you could have given instead:

The FCFS algorithm is unsuitable because essential antivirus processing would be

placed at the back of the queue and wait for its turn. Lower-priority tasks would use
valuable CPU time, meaning the system could be at risk.

System Interrupts
What are interrupts?

 An interrupt is a signal to the processor that stops its current task and performs
a different task temporarily
 Interrupts can be hardware events or time-sensitive tasks
 When an interrupt occurs, the processor suspends the current
program execution and transfers control to an interrupt service routine

Purpose and Role of Interrupts

 Real-time Event Handling: hardware errors and signals from input devices e.g.
hard disk failure
 Device Communication: alerts from external devices e.g. printer jams and
network errors
 Multitasking: suspending processing in one application so that the user can
switch to another

Types of Interrupts

Type Definition Example

Generated by Keyboard input, mouse
Hardware
external movements, disk I/O
Interrupts
devices requests
 Triggered by
Application requests to
Software software or the
open a file, division by
Interrupts operating
zero errors
system
Intentionally Software debugging,
Trap
triggered by a handling unexpected error
Interrupts
program cases

The Interrupt Process

1. Interrupt Request (IRQ)

o An external device or software generates an interrupt, signalling the
processor to stop its current task
o The interrupt controller passes this to the interrupt handler for assessment
2. Interrupt Acknowledge
o The interrupt handler decides if the interrupt needs to be dealt with now or
later
o If yes, the current contents of the processor registers are saved in
memory
3. Interrupt Service Routine (ISR) Lookup
o The processor fetches the corresponding Interrupt Service Routine (ISR)
associated with the interrupt type
4. ISR Execution
o The processor transfers control to the ISR and executes the routine to
handle the specific interrupt
5. Interrupt Exit
o After the ISR completes, the processor restores the content of the
registers from step 2
o The fetch-decode-execute cycle is resumed

The interrupt process

What is an ISR?
  An ISR is a special function that handles a particular interrupt type
 Each type of interrupt has a corresponding routine, e.g. printer jam, hard disk
failure, file download error, network connection error all have routines to be
followed when they happen
 ISRs should be concise, efficient, and carefully designed to minimise the time
taken to execute, as they often need to handle time-sensitive events

Interrupt priority and nesting

 Interrupt prioritisation means the processor can acknowledge and switch to

resolving a higher-priority interrupt
 Prioritising interrupts is vital because many things can go wrong at the same time
 Lower-priority ISRs may be temporarily suspended until the higher-priority
ISR completes the execution
 Nesting of interrupts refers to the ability of the processor to handle interrupts
within interrupts
 Proper management of nested interrupts avoids potential conflicts and ensures
system stability

Nesting of interrupts

Interrupt priority handling

 System handling of interrupt priority

What is a Device Driver?

 A device driver is a piece of software that enables communication between an

operating system and specific hardware devices such as:
o Printers
o Graphics cards
o Network cards
 Device drivers allow the OS to control and interact with those devices
 Because many external devices have embedded system software, a
driver bridges the gap between a major operating system and a tiny hardware
OS
 Device drivers make it possible to perform specific operations on the hardware
e.g. a printer driver enables the OS to send print commands and manage print
jobs
 Most hardware manufacturers write their own device driver software, meaning a
single operating system may have several printer drivers installed

Worked example

Adding a DVD drive to a computer often requires the installation of a piece of software
called a device driver.

State the purpose of a device driver.

 [1]

How to answer this question:

 Pick one example of a hardware device and describe what happens between it
and the operating system
 Explain that this is made possible through device driver software

Example answer that gets full marks:

The purpose of a device driver is to enable the operating system to interact with and
control a hardware device. For example, a printer would have an associated device
driver so the OS could send or receive data from a printer system.

Acceptable answers you could have given instead:

The purpose of a device driver is to enable communication between an operating

system and external hardware.

Virtual Machines
What are virtual machines?

 Virtual machines (VMs) are entire operating systems running inside another
operating system
 A user running Windows 11 could run a virtual machine of MacOS
 This would allow them to navigate the GUI of MacOS and install software on it
 Running a virtual machine helps access software that is only designed to run on
specific operating systems
 VM management software includes a Hypervisor that monitors all activity
happening inside the VM

Structure of several virtual machines running on a single piece of hardware

Cross-platform and forwards compatibility

 Not all software is designed to run on all operating systems

 Apple commonly makes software that only runs on MacOS for performance
reasons
 A Windows user could run a virtual machine of MacOS and install the software
they need
 Most software needs to be updated to work on the latest versions of operating
systems
 A user running the latest release of Windows may need to run a virtual machine
of a previous release of Windows to use an application that hasn't received
a forwards-compatibility update

In software testing

 VMs are a way to create isolated test environments, that leave the host
operating system unaffected
 Isolated environments allow a developer to:
o Monitor the way their software affects system performance
o Test on a clean-slate system, while no other applications are running
 VM management software can create virtual machines that act like they
have older hardware
 This allows developers to build software that can be run on older hardware so
that more users can use the software
 A developer can test against various operating systems, such as MacOS,
Linux and Windows, for greater compatibility
  This is achieved through compiling into intermediate code and running against
different operating systems

Consequences

 VMs share the same system hardware as the host OS

 Over-use of VMs can exhaust the host OS of the system of CPU, hard disk and
memory
o VM software such as VirtualBox can set maximum limits on system
resources
o A low-specification machine could be configured to allocate only 1GB of
memory and 20% of CPU
o A high-specification machine could afford up to 8GB of memory and
50% of the CPU
 Operating systems are commonly free to download, but require an activation
payment to access all features

Worked example

Explain why the programmers of anti-virus software may make use of virtual machines
when developing the updates.

[3]

How to answer this question:

 Recall two benefits of using virtual machines in developing software

 Link these benefits to some considerations needed when developing anti-virus
software

Example answer that gets full marks:

Virtual machines (VMs) are essential in developing anti-virus software updates for
several reasons. VMs create an isolated environment that is separate from the host
operating system. This enables developers to safely work with virus test code that could
risk the integrity of their own operating system. VMs can also be configured to emulate
various types of hardware. This allows developers to understand how their anti-virus
software will perform on different devices and under different conditions, ensuring a
wider range of compatibility.

Acceptable answers you could have given instead:

Virtual machines create an isolated test environment from the host operating system.
This means working with harmful test code carries less risk. Virtual machine
management software can monitor the VM through the Hypervisor. This will show how
the software affects VM system performance, allowing the programmer to make
changes where needed.
Stages of Compilation
What is Compilation?

 Compilation is a process that translates a program written in a high-level

programming language into machine code
 Only machine code can be executed by a computer
 There are four stages involved in this process:
o Lexical Analysis
o Syntax Analysis
o Code Generation
o Optimisation

Lexical Analysis

 Lexical analysis means studying the words or vocabulary of a language

 This stage involves identifying lexical 'tokens' in the code
 Tokens represent small meaningful units in the programming language, such as:

o Keywords
 var, const, function, for, while, if
o Identifiers
 Variable names, function names
o Operators
 '+', '++', '-', '*'
o Separators
 ',' ';', '{', '}', '(', ')'

 During this stage, unnecessary elements like comments and whitespace are
ignored
 For example, if the following code is being compiled:

var x = function(x,y) {
if(x>2) {
return x*y;
}
return x+y;
}

 The result of lexical analysis is a token table

 Token Type

1 var Keyword

2 x Identifier

3 = Operator

4 function Keyword

5 ( Separator

6 x Identifier

7 , Separators

8 y Identifier

9 ) Separator

10 { Separator

11 return Keyword

12 x Identifier

13 * Operator

14 y Identifier

15 ; Separator

16 } Separator

Syntax Analysis

 Now that tokens have been identified, syntax analysis makes sure they
all adhere to the syntax rules of the programming language
 A symbol, e.g. '$' could be a valid token but not a valid character according to
particular programming languages
 The dollar symbol would be flagged as breaking the syntax rules
 Other syntax errors programmers commonly make include mismatched
parentheses or missing semicolons
 If the code passes the syntax analysis, the compiler can create an Abstract
Syntax Tree (AST)
 An AST is a graph-based representation of the code being compiled
 An AST is an efficient way to represent the code for the next step
Example abstract syntax tree

 For the same code as above, the following abstract syntax tree can be created

Code Generation

 This step takes the AST and traverses it to generate object code that can be
executed by the computer

Optimisation

 This step modifies the code to make it more efficient without changing its
functionality
 This is important to attempt because it reduces the memory required to run the
code, which leads to faster execution
 A common optimisation action is removing duplicate code
  If an 'add' function is written twice in the source code, a sophisticated compiler
will notice this and include it only once in the object code

Worked example

Imogen is writing application software for her company and is ready to compile it.

const celsius = (fahrenheit) => {

return (5/9) * (fahrenheit-32);
}
Referring to the example above, explain what happens during Lexical Analysis.

[2]

How to answer this question:

 Recall the purpose of lexical analysis and what it aims to produce

 Recall the different types of tokens that can be identified in the code
 Use examples from the code block to write your answer

Answer that gets full marks:

'Lexical analysis' breaks the code into tokens, ignoring whitespace and comments.
Tokens are identified by their type:

 keyword: 'return'
 operator: '*', '-'
 identifier: celsius, fahrenheit
 delimiter: ';', '(', ')' etc

When all tokens have been identified, a token table is produced for the next step in the
compilation.

Acceptable answers you could have given instead:

The compiler will look at all the lexical tokens in the code e.g. 'fahrenheit' is an identifier,
'return' is a keyword. All the tokens are placed into a tokens table for the next step