Operating Systems Concepts

Contents Introduction 3

Virtual Memory 3

First In First Out 4

Least Recently Used 5

References 6

Introduction
As we gain a better understanding of how computers and their operating systems work, we are able to better grasp how they handle the execution of programs. As we first learn to use computers we really do not think about the things that programs need in order to run, such as memory, as long as everything works. Although as we advance our knowledge of how such things operate, we begin to understand the mechanics of such things as physical and virtual memory.
Virtual Memory Normally when we think of memory, in terms of computers, we generally tend to think about things like the hard drive or the RAM. These items are generally used in the storage of programs, documents and multimedia, however when it comes to running a program their functions are slightly altered. Programs that are executed to run on a computer require a certain amount of memory in order to run. As modern operating systems have become capable of running multiple processes, or programs, with limited resources of memory, operating systems need a way to manage these resources. This is referred to as memory management, which allocates the memory resources as they are needed. When a memory manager has the capability to allocate the memory resources to multiple programs this is known as virtual memory. Virtual memory is a technique that is used to give the illusion of addition memory when a system is low on physical memory, the manager would divide the available space into pages and then stores its contents in mass storage. When paging is used, the processor divides the linear address space into fixed-size pages that can be mapped into physical memory and/or disk storage. When a program or task references a logical address in memory, the processor translates the address into a linear address and then uses its paging mechanism to translate the linear address into a corresponding physical address. Windows is known to use paging where Linux, Mac OS X, Solaris and Amiga OS 4 uses file swap. The process is, swapping which is the moving of memory pages between RAM and disk. By default Windows OS configuration, 2 GB of this virtual address space are designated for each process’ private use and the other 2 GB are shared between all processes and the operating system.
As an example of this I have found a program that emulates the process that memory management goes through when it allocates memory to a program. Once you press start, there are roughly 8 processes that are taking turns in requesting the use of the systems memory resources. All of the processes start out as light blue, once the programs start the ones that begin to request resources turn dark blue. There are also 16 pages of virtual memory along with 8 pages of primary memory.
The virtual memory is yellow while the primary memory starts out as white, as they are currently empty. Once a process request resources it will randomly chooses a page, as the virtual memory is brought into primary memory. If a process request a page that is already in use the process goes from green to yellow, indicating that its contents have been changed and that the original contents have been saved before the page is used by another program. This is an excellent example of what the memory manager does to maintain computers resources, for multiple programs.
First In First Out
First in first out, or FIFO, it allows for a process keeps running on a CPU until it is blocked or terminated. The main advantage of the FIFO policy is simplicity, as the oldest pages are always replaced. The FIFO also provides a certain degree of fairness, since requests are served in order. However, the main disadvantage is that short jobs can get stuck behind long running jobs. As well as the fact that the pages which are being deleted does not mean that it is not in use, as pages that are accessed more frequently may be replaced. Constantly doing this could be considered bad judgment.
Least Recently Used
Least recently used, or LRU, replaces the page which has been unused for the longest time, as well as the primary memory that has gone the longest time used. LRU maintains a linked list of all pages ordered from the most recently used to the least recently used. Maintaining a list on every instruction execution is very expensive and time consuming.
In relation to the last two methods discussed, I have found a simulation that best emulates these methods. It simulates 16 pages of virtual memory along with 8 pages of primary memory, I then inputted a sample set of numbers and then set it to sample string. Once I pressed start the LRU and FIFO went through the string of numbers, as they did they proceeded to go through the number of page faults, number of page references and percentage of page faults. The number of page faults went from one to eight, as the numbers went up the page references and fault numbers increased as well. As a clicked on fast and slow the number of reference pages steadily increased, as the percentage of page faults fluctuated wildly between numbers. This is caused by the fact that as the number of page faults and references increases the percentage of fault or dropped pages changes as some go unused or was in service for long periods of time.

References
William Stallings 6th Edition, Operating Systems internals and Design Principles, Chapter 7 and 8
Page replacement algorithm information retrieved from http://en.wikipedia.org/wiki/Page_replacement_algorithm