Systems Software Flashcards
What is an operating system?
It is a program or set of programs that manages the operations of the computer for the user. It provides an interface for the user, application software and hardware since the user cannot directly communicate with hardware.
How does an operating system work?
- OS is held in permanent storage eg. hard disk
- Loader (small program) is held in ROM
- Computer switched on - loader in ROM sends instructions to load OS by copying it from storage into RAM.
what are the functions of an operating system?
- memory management
- interrupt service routines
- processor scheduling
- backing store management
- management of all input and output
what is memory management?
it is when the OS must manage allocation of RAM to different programs fairly since there may not be sufficient RAM for everything to be in it at once.
what are 3 methods of memory management?
- paging
- segmentation
- virtual memory
What is paging?
- memory (RAM mainly) is divided into fixed sections called pages (4Kb)
- process in memory will be allocated enough pages to run - but these may not be contiguous physically
- page table uses mapping to store a link between physical memory address + logical address space of each process
What does contiguous mean?
next to each other/adjacent physically
What is segmentation?
- similar to paging except memory is divided into segments which can be different sizes.
- Things stored in the segment may relate to parts of a program - eg. a particular function/subroutine may occupy a segment.
- segment size relates to the program structure
what is the difference between paging and segmentation?
Paging uses chunks of memory that are the same size while segmentation uses chunks of memory which are different sizes.
What is virtual memory?
- an area of secondary storage that can be used as an extension of memory when RAM is filled up and another instruction needs processor time
- pages of the current process will be stored in virtual memory until they are needed - then they will be swapped with an unused page into RAM
- advantage: it’s extra strage and can optimise computer performance when used correctly
- disadvantage: slower than using RAM
What is disk thrashing?
It is when the OS is excessively swapping pages into and out of RAM which slows down performance significantly because so much time is spent swapping pages.
What is an interrupt?
- its a signal from a software program, hardware component or internal clock to the CPU
- software interrupt = when an application program terminates or requests certain services from the OS
- hardware interrupt = when in I/O operation is complete or an error occurs eg. printer paper out
why are interrupts needed?
- I/O process is completed/sends an interrupt signal
- an error occurs eg. printer paper out
- power failure
- scheduled interrupt from the internal clock - to enable multitasking to take place
when are interrupts detected and serviced?
The CPU checks for interrupts at the end of a clock cycle for any interrupts that need to be processed
how are interrupts dealt with?
- if the CPU detects an interrupt signal, the running program is suspended + lower priority interrupts are disabled
- all values in program counter and registers are stored in the stack
- Interrupt Service Routine (ISR) is called to deal with the interrupt
- Descending order of interrupt priorities: power fail interrupt, clock interrupt, I/O interrupt
- once interrupt has been serviced, original register values are popped off the stack and process resumes from the point it left off.
What is processor scheduling?
- 1 CPU can perform 1 instruction at a time
- processor scheduling is done by the OS scheduler where each application that needs to be run is allocated processor time in a way that makes most efficient use of the CPU
- determines the manner and order that processes and applications can use the CPU
- basically doing small parts of multiple large tasks in turn.
- Gives the illusion that the CPU is doing multiple tasks simultaneously - multitasking
What are the aims of processor scheduling?
- ensure fairness to all users of a multi-user system (eg. in a school)
- provide acceptable response time to all users
- maximise time CPU is usefully engaged
- ensure hardware resources are kept as busy as possible
- maximise throughput
What is the round robin scheduling algorithm?
- operates on a First In First Out (FIFO) basis
- every process is given an equal amount of CPU time (time slice)
- CPU cycles through every process to be completed - if a process isn’t completed in time, CPU moves onto the next process and it’s given a new time slice to be completed in after all the other processes in the cycle
- Timings are maintained by an interval timer - it generates interrupts at specific times
- this method is good because it guarantees a reasonable response time to all users
What is the first come first served scheduling algorithm?
- the first job to arrive is the first to be processed - no system of priorities
- The current process must be completed before the CPU moves onto the next one (non-preemptive)
- the completed process is removed from the queue
- not very efficient
what is the shortest remaining time scheduling algorithm?
- the process with the shortest estimated time to complete is run next
- aims to reduce waiting time for shorter jobs - therefore have fewer waiting jobs
- only works well if there is knowledge of how long a job will take and it should be a good estimate
- not great for longer jobs that get stuck behind many smaller jobs
what is the shortest job first scheduling algorithm?
- similar to shortest remaining time except the job with the shortest estimated running time is done first
- similar pros and cons as shortest remaining time
what are multi-level feedback queues?
- a scheduling algorithm designed to giver preference to short jobs and I/O bound processes and separate processes into categories based on their processor needs
- there are multiple dynamic job queues and jobs can move between queues based on how much processor time they use
- it aims to maximise processor use
- I/O devices are much slower than processor speed so it aims to keep them as busy as possible to prevent a bottleneck (eg. lots of jobs are sent to the printer)
- eg. while one job is printing another can use the CPU.
what is backing store management?
- when files/applications are loaded, they transfer from backing storage to memory
- OS must keep a directory of where files are stores so they can be quickly accessed
- also needs to know which parts of storage are free to save new files/applications.
- file management system allows user to move files/folders, delete files + protect them from unauthorised access
what is peripheral management?
- different applications need different input/output devices throughout operation
- eg. to print something: OS communicates with printer to check if it’s online, if it is a printer, and send the data to print
- data to be printed will be sent to an area of memory called the buffer so CPU can carry on with another task
what is the buffer?
an area of memory whose purpose is to compensate for the difference in speed between input/output devices and the CPU
how do you determine what type of OS to use in a particular situation?
- what functions need to be prioritised
- what will enable operations to be completed as efficiently as possible.
what is the distributed OS?
- an OS that runs across multiple computers to coordinate processing of a single job across multiple computers
- a form of parallel processing system
- one job is split up into several tasks and each is run on a seperate computer to make use of more computational power
- task distribution is managed by OS so user gets the illusion they are working on a single computer
- eg. the intranet - system is configured as a cluster of servers that share memory and tasks (better performance + power than a single large server)
- no need to write different code for it/training since OS manages task distribution - but also means programmer has no control
what is a multi-tasking OS?
- can run on a standalone computer (eg. PC/laptop)
- gives the impression that one processor can do more than one task at a time simultaneously
- achieved by scheduling processor time using scheduling algorithms or using multiple cores to run tasks simultaneously
- eg. the Windows OS
What is a multi-user OS?
- aka time-sharing system
- lots of terminals with a user are connected to a single powerful mainframe/supercomputer
- each user gets a time-slice of CPU time according to a scheduling algorithm
what is a mobile OS?
- the operating system used by smartphones, tablets, PDAs and other mobile devices
- these devices are multi-tasking computers
- they combine features of a PC OS with their own special features for mobile use - eg. managing cellular + wifi connectivity + phone access
- other special features - touch sensitivity, GPS mobile navigation, camera, speech recognition, music player etc.
- most tied to specific hardware - eg. Android and iOS
What are the parts of a mobile OS?
- main OS - controls user interface + running application software
- low-level proprietary real-time operating system - operates radio + other hardware
- low-level systems have lots of security vulnerabilities that allow others to gain control of a mobile device.
What is an embedded system?
- a specialised OS designed to perform a specific task for a device that isn’t a computer
- eg. on a washing machine or microwave
- runs the code for the specific task it’s designed for to allow the device to do its job
- allows device hardware to be accessible to software running on top
what are the features of an embedded system?
- minimal user interface - maybe a few buttons or a dial and maybe a small screen
- accept input from sensors, send output to control devices
- limited amount of RAM - complex memory management system not required
- no permanent data storage devices to manage
