Virtual Memory (week 9)

Question 1

where is virtual memory of the computer

Answer 1

it is on the fast secondary memory (hard disk) of your computer

Question 2

why is sole use of primary memory not enough to run programs

Answer 2

–> Because large portions of most programs are not executed most of the time.

–> Therefore, large portions of programs need not be in memory.

–> Need to maximize the use of primary memory by loading as many programs as possible so as to increase CPU utilization.

Question 3

what is virtual memory and what does it do ?

Answer 3

–> Virtual memory allows the execution of
processes that are not completely in memory

–> Each process’s address space is partitioned into parts that can be loaded into primary memory

–> Partitions which are not needed are written to secondary memory

–> Allows program that are larger than physical memory to be executed

Question 4

Every process has code and data ______

Answer 4

locality

Question 5

when is there a change in locality

Answer 5

As computation moves to a different
phase there is a change in locality

Question 6

how is the address space partitioned

Answer 6

Address space is logically partitioned

–> Text, data, stack
–> Initialization, main, error handle

Question 7

how is virtual memory organized

Answer 7

for example if there is 2 parts of a program (Pi) which is part A and part B first the part A will be loaded into primary memory and after part A has been executed it will be sent back to virtual memory and part B will loaded into primary memory.

Question 8

why is there a need to implement an additional layer of abstraction for virtual memory

Answer 8

An additional layer of abstraction is implemented to map the physical addresses of data stored in RAM to virtual memory addresses that can be used by the OS

Question 9

.

Answer 9

.

Question 10

To support virtual memory, the computer
system needs to manage memory in _______

Answer 10

blocks

Question 11

Virtual memory system transfers “blocks” of
the address space to/from _______

Answer 11

primary memory

Question 12

what are the 2 types of blocks

Answer 12

Fixed size blocks (paging)
Variable size blocks (segmentation)

Question 13

what is the commercially dominant form of virtual memory today ?

Answer 13

demand paging

Question 14

what is fixed size block s

Answer 14

System-defined pages are moved back and forth between primary and secondary memory

Question 15

what is variable size blocks

Answer 15

Programmer-defined segments - corresponding to logical fragments – are the unit of movement

Question 16

what is a page

Answer 16

A page is a fixed size, 2^h, block of virtual
addresses, where h is an integer

Question 17

what is a page frame

Answer 17

A page frame is a fixed size, 2^h, block of
physical memory (the same size as a page)

Question 18

why is the size of a page exponent of 2

Answer 18

to minimize translation time between virtual and physical addresses

Question 19

what happens in demand paging

Answer 19

–> In demand paging, the pager brings the necessary pages into memory.

–> Avoids reading in memory pages that will not be used, decreasing the swap time and the amount of physical memory needed.

–> In a pure demand paging scheme, the system never bring a page into the memory until it is needed.

Question 20

what are the benefits of using virtual memory

Answer 20

–> Programs no longer constrained by the
amount of physical memory

–> Each program takes less physical memory, allowing more to run at the same time, increasing CPU utilization/throughput

Question 21

What happens when a page is required
but that page is not in memory but in the
virtual memory?

Answer 21

page fault occurs

Question 22

what is a page fault

Answer 22

A page fault is an event when a page is
required but is not found in the primary
memory, but is in virtual memory

Question 23

what is a page table

Answer 23

Page table is a hardware mechanism implemented to map the logical page of the program to the physical page frame where the page is loaded.

Question 24

what is the valid-invalid scheme

Answer 24

The valid-invalid scheme is used to
distinguished between those pages that are in memory and those that are on the disk.

Question 25

what does it mean if the table has bit 0/1

Answer 25

–> Each page table entry is associated with a bit (1 - in memory, 0 - not in memory).

–> If the bit is set to “valid”, the page is both legal and in memory. Otherwise, the page is either not valid or is valid but is on disk.

–> Access to page marked invalid causes a pagefault trap.

Question 26

what is the page fault rate

Answer 26

Page Fault Rate 0 ≤ p ≤ 1.0

if p = 0 no page faults
if p = 1, every reference is a fault

Question 27

what is Effective Access Time (EAT)

Answer 27

shows performance of demand paging

EAT = (1 – p) x memory access time
+ p (page fault overhead
+ [swap page out ]
+ swap page in
+ restart overhead)

avg page service fault time = page fault overhead + [swap page out ] + swap page in + restart overhead

Question 28

EAT is directly proportional to ____________

Answer 28

page-fault rate

Question 29

what is page replacement

Answer 29

When a computer needs to access a page that is not currently in RAM, it must first find a page to remove from RAM to make space for the new page.

This process is known as page replacement. The computer uses an algorithm to determine which page to remove, with the goal of minimizing the number of page faults.

Question 30

when is the page fault very high

Answer 30

If a process does not have “enough” pages

Question 31

when is a process thrashing

Answer 31

A process is thrashing if it is spending more time paging than execution.

–> The CPU utilization is low.
–> The operating system thinks that it needs to increase the degree of multiprogramming.

Question 32

how can the effects of thrashing be limited

Answer 32

by using locality replacement algorithm

Question 33

why does trahsming occur

Answer 33

Σ size of locality > total memory size (running too many programs at the same time)

Question 34

why does paging work in limiting thrashing occurrences

Answer 34

because of the Locality model
–> Process migrates from one locality to another.
–> Localities may overlap (more efficient use of its available RAM.)