Memory Management: Early Systems
Virtual Memory Methods
Memory Manager
• Access Memory
• Random Access Memory
• Core Memory
• Primary Storage
Early Memory Allocation Schemes
• Single-User Systems
• Fixed Partitions
• Dynamic Partitions
• Relocatable Dynamic Partitions
Single-User Contiguous Scheme
• Each program to be processed was loaded in its
entirety into memory and
• allocated as much contiguous space in memory as it
needed
• It doesn’t support multiprogramming or networking
Fixed Partitions
• static partitions
• one partition for each job
• protection of the job’s memory space
• internal fragmentation
Dynamic Partitions
• Contiguous blocks
• Jobs are given only as much memory
Best-Fit Versus First-Fit Allocation
• First-fit memory allocation
- first partition fitting the requirements
• Best-fit memory allocation
- least wasted space, the smallest partition
fitting the requirements
• Space versus speed
Deallocation
• Fixed partition system
- Resets the status of the memory block to “free”
• Dynamic Partition
Case 1: Joining Two Free Blocks
Case 2: Joining Three Free Blocks
Case 3: Deallocating an Isolated Block
Relocatable Dynamic Partitions
• Memory Manager relocates programs to gather
together all of the empty blocks and compact them to
make one block of memory large enough to
accommodate some or all of the jobs waiting to get in
• The compaction of memory, sometimes referred to as
garbage collection or defragmentation, is performed by
the operating system to reclaim fragmented sections of
the memory space.
Memory Management: Virtual Memory
• Paged Memory Allocation
• Demand Paging
• Segmented Memory Allocation
• Segmented/Demand Paged Memory Allocation
Page Memory Allocation
• Page Memory Allocation based on the concept of dividing
each incoming job into pages of equal size.
• The sections of a disk are called sectors (or sometimes
blocks), and the sections of main memory are called page
frames.
• The pages do not have to be loaded in adjacent memory
blocks. In fact, each page can be stored in any available page
frame anywhere in main memory.
• Job Table (JT) contains two values for each active job: the
size of the job (shown on the left) and the memory location
where its Page Map Table is stored (on the right)
• Page Map Table (PMT), which contains the vital
information for each page—the page number and its
corresponding page frame memory address. • The Memory
Map Table (MMT) has one entry for each page frame listing
its location and free/busy status.
• The displacement, or offset, of a byte (that is, how far
away a byte is from the beginning of its page) is the factor
used to locate that byte within its page frame.
Demand Paging
• Demand paging introduced the concept of loading only a
part of the program into memory for processing.
Page Replacement Policies and Concepts
• The first-in first-out (FIFO) page replacement policy will
remove the pages that have been in memory the longest.
• The least recently used (LRU) page replacement policy
swaps out the pages that show the least amount of recent
activity, figuring that these pages are the least likely to be
used again in the immediate future.
 Introducing Operating Systems File Management
• The File Manager keeps track of every file in the
Operating System system, including data files, program files, compilers,
• Operating system software is the chief piece of and applications.
software • By using predetermined access policies, it enforces
• Portion of computing system that manages all of the restrictions on who has access to which files.
hardware and all of the other software. • For example, a user might have read-only access,
• Controls who can use the system and how. read-and-write access, or the authority to create and
• It’s the boss! delete files. • Managing access control is a key part of
file management.
Essential Managers of OS
• Memory Manager Network Management
• Processor Manager • Operating systems with Internet or networking
• Device Manager capability have a fifth essential manager called the
• File Manager Network Manager • It provides a convenient way for
• Network Manager users to share resources while controlling users’ access
to them.
Manager Tasks • These resources include hardware and software
• Monitor its resources continuously
• Enforce the policies that determine who gets what, User Interface
when, and how much • The user interface is the portion of the operating
• Allocate the resource when appropriate system that users interact with directly.
• Deallocate the resource when appropriate • In the old days, the user interface consisted of
commands typed on a keyboard and displayed on a
Main Memory Management monitor.
•The Memory Manager is in charge of main memory • Now most systems allow users to choose a menu
•It checks the validity of each request for memory option from a list.
space
•It allocates a portion of memory that isn’t already in History of Machine Hardware
use •It deallocates memory.
•Protect the space in main memory occupied by the OS History of Machine Hardware
itself. • Mainframe
• Minicomputer
Processor Management • Supercomputer
• The Processor Manager decides how to allocate the • Microcomputer
central processing unit • Servers
• Keep track of the status of each process
• Monitors whether the CPU is executing a process or Mainframe
waiting for a READ or WRITE command to finish • a large machine - in size and in internal memory
execution capacity • The IBM 360 (1964) is a classic example of
• Can be compared to a traffic controller an early mainframe.

Processor Management Level of Responsibility IBM 360 model 30

• required an air-conditioned room about 18 feet
• To handle jobs as they enter the system square
• Job Scheduler, the high-level portion of the Processor • The CPU was 5 feet high and 6 feet wide
Manager, which accepts or rejects the incoming jobs. • internal memory of 64K
• To manage each process within those jobs • Process • $200,000 in 1964 dollars
Scheduler, the low-level portion of the Processor
Manager, which is responsible for deciding which Minicomputer
process gets the CPU and for how long. • It was developed to meet the needs of smaller
institutions • Digital Equipment Corporation (early
Device Management 1970s)
• PDP-8 (less than $18,000)
• The Device Manager monitors every device, channel,
and control unit. Minicomputer vs Mainframe
• Its job is to choose the most efficient way to allocate • Smaller in size and memory capacity
all of the system’s devices • Cheaper
• Midrange computers
Supercomputer
• It was developed primarily for government
applications
• Business and industry became interested in the
technology when the massive computers became faster
and less expensive.
Cray supercomputer
•six to thousands of processors
•performing up to 2.4 trillion floating point operations
per second (2.4 teraflops).
Supercomputer Usage
• scientific research to customer support and product
development
• often used to perform the intricate calculations
required to create animated motion pictures
• help oil companies in their search for oil by analyzing
massive amounts of data
Microcomputer
• It was developed to offer inexpensive computation
capability to individual users in the late 1970s.
• amount of memory: 64K
• physical size was smaller than the minicomputers
• Eventually grew to accommodate software with larger
capacity and greater speed.
Workstation
•Powerful microcomputers developed for use by
commercial, educational, and government enterprises
•networked together
•used to support engineering and technical users
•perform massive mathematical computations or
computer-aided design (CAD)
Servers
• Servers are powerful computers that provide
specialized services to other computers on
client/server networks.
• Examples can include print servers, Internet servers,
e-mail servers, etc.
• Each performs critical network tasks.
Types of Operating System
• Operating systems for computers large and small fall
into five categories distinguished by response time and
how data is entered into the system: batch, interactive,
real-time, hybrid, and embedded systems.
Types of Operating System
• Batch systems
• Interactive systems
• Real-time systems
• Hybrid systems
• Embedded systems
Batch Systems
• Batch systems date from the earliest computers,
when they relied on stacks of punched cards or reels of
magnetic tape for input.
• Jobs were entered by assembling the cards into a
deck and running the entire deck of cards through a
card reader as a group—a batch.
• The efficiency of a batch system is measured in
throughput—the number of jobs completed in a given
amount of time (for example, 550 jobs per hour)
Interactive Systems
•Give a faster turnaround than batch systems but are
slower than the real-time systems
•They were introduced to satisfy the demands of users
who needed fast turnaround when debugging their
programs. •It required the development of time-
sharing software
•It provides immediate feedback to the user and
response time can be measured in fractions of a
second.
Real-time Systems
• Used in time-critical environments where
reliability is key and data must be processed
within a strict time limit.
•The time limit need not be ultra-fast (though it often
is), but system response time must meet the deadline
or risk significant consequences.
•These systems also need to provide contingencies to
fail gracefully—that is, preserve as much of the
system’s capabilities and data as possible to facilitate
recovery.
Two Types of Real-time Systems
• Hard real-time systems risk total system failure if
the predicted time deadline is missed.
• Soft real-time systems suffer performance
degradation, but not total system failure, as a
consequence of a missed deadline.
Embedded Systems
• Computers placed inside other products to add
features and capabilities.
• For example, you find embedded computers in
household appliances, automobiles, digital music
players, elevators, and pacemakers.
Embedded Systems
• Each one is designed to perform a set of specific
programs, which are not interchangeable among
systems.
• This permits the designers to make the operating
system more efficient and take advantage of the
computer’s limited resources, such as memory, to their
maximum.
Hybrid Systems
•Combination of batch and interactive
• Users can access the system and get fast responses
•Accepts and runs batch programs in the background
when the interactive load is light
•It takes advantage of the free time between high
demand usage of the system and low-demand times.
• Many large computer systems are hybrids.
Operating System Development
1940
• time of vacuum tube technology
• computers are the size of classrooms
• little need for standard operating system software
familiar with the idiosyncrasies of their hardware
• compilers and assemblers
1950
• developed to meet the needs of new markets—
government and business researchers
• cost effectiveness of the system
• still very expensive
• Two improvements were widely adopted:
- Computer operators were hired to facilitate
each machine’s operation
- job scheduling was instituted
• Job scheduling introduced the need for control cards,
which defined the exact nature of each program and its
requirements.
• This was one of the first uses of a job control language
• Speed of I/O devices increased
• Blocking
• Buffering
• Spooling
1960
• Passive multiprogramming
- the operating system didn’t control the
interrupts but waited for each job to end an
execution sequence.
• Active multiprogramming
- allowed each program to use only a preset slice
of CPU time.
1970
• Programmers soon became more removed from the
intricacies of the computer
• application programs started using English-like
words, modular structures, and standard operations.
1980
• Improved the cost/performance ratio of computer
components
• Hardware was more flexible, with logical functions
built on easily replaceable circuit boards
• firmware, a word used to indicate that a program is
permanently held in read-only memory (ROM)
• multiprocessing (having more than one processor),
• more complex languages were designed to coordinate
the activities of the multiple processors servicing a
single job
• With network operating systems, users generally
became aware of the existence of many networked
resources,
• could log in to remote locations, and
• could manipulate files on networked computers
distributed over a wide geographical area
1990
• The overwhelming demand for Internet capability
sparked the proliferation of networking capability
• The World Wide Web by Tim Berners-Lee made the
Internet accessible by computer users worldwide
• introduced a proliferation of multimedia applications
demanding additional power, flexibility, and device
compatibility for most operating systems
2000
• virtual machines
• multiple operating systems running at the same time
and sharing resources
• Virtualization is the creation of partitions on a single
server, with each partition supporting a different
operating system. • A thread (or lightweight process)
can be defined as a unit smaller than a process, which
can be scheduled and executed.
Object-oriented design
• An important area of research that resulted in
substantial efficiencies was that of the system
architecture of operating systems.
• the way their components are programmed and
organized, specifically the use of object-oriented
design and the reorganization of the operating
system’s nucleus, the kernel.
• The kernel is the part of the operating system that
resides in memory at all times, performs the most
essential operating system tasks, and is protected by
hardware from user tampering.