Introduction to Operating Systems

• Introduction to operating systems covering topics such as what an operating system is, system architecture, system components, and memory layout

What is an Operating System?

• Acts as an intermediary between a computer user and its hardware
• Goal is to execute user programs, solve user problems and make the system convenient and efficient.

Computer System Components

• Hardware provides the basic computing resources like CPU, memory, and I/O devices
• The operating system controls and coordinates hardware use among different application programs and various users
• System and application programs define how system resources are used to solve computing problems
• Users can be people, machines, or other computers

Operating System Definitions

• Resource Allocator manages and allocates resources
• Control Program controls the execution of user programs and I/O device operations
• Kernel is the one program running at all times, with all other programs being system or application programs

Multiprogramming Batched Systems

• Multiprogramming involves keeping several jobs in main memory simultaneously
• CPU is multiplexed among them, requiring memory management and protection
• The OS picks and begins executing one job from memory
• While a job requires an I/O operation the OS switches to another job, keeping the CPU busy
• The number of jobs in memory is typically less than the number of jobs in the job pool on disk
• If there isn't enough room in memory, the system chooses which jobs get loaded in, also known as job scheduling
• The system must also decide which jobs get run at the same time, or CPU scheduling
• Having multiple programs in memory at the same time requires effectively managing the memory
• The CPU sits idle in a non-multiprogrammed system, but is always in use in a multiprogramming system

OS Features Needed for Multiprogramming

• I/O routines are supplied by the system
• Memory management allocates memory to several jobs
• Job scheduling chooses jobs ready to run from disk
• CPU scheduling chooses jobs ready to run in memory
• Includes allocation of devices and general protection
• Multiprogrammed batched systems cannot be used to interact with jobs as users are executing
• Programmers also cannot modify a program as it executes to study its behavior

Time-Sharing Systems - Interactive Computing

• Time-sharing systems utilize CPU scheduling and multiprogramming to provide a fraction of a time-shared computer to each user
• The CPU is multiplexed among jobs in memory and on disk, and only allocated to jobs in memory
• When a job requires an I/O operation, the CPU switches to another job, which keeps the CPU busy
• Jobs get swapped between memory and disk, with the disk to act as a backup in main memory
• On-line communication is provided between the user and the system
• The OS gets the next control statement from the user's keyboard and not a card reader
• It must be available for users to access data and code

Time-Sharing Systems

• Features include an on-line file system, that manages disk collections
• A mechanism for concurrent execution, involving CPU scheduling schemes
• There is also job synchronization and communication to ensure orderly execution
• A mechanism prevents deadlock, in case a job is waiting for another job forever

Personal-Computer Systems

• Personal computers are computer systems dedicated to a single user
• I/O devices include keyboards, mice, display screens and small printers
• PC operating systems were not initially multi-user or multi-tasking
• The main goal of PC Operating systems is to maximize user convenience and responsiveness instead of maximizing CPU and I/O utilization
• Examples include Microsoft Windows and Apple Macintosh

Parallel Systems

• Multiprocessor systems consists of multiple CPUs in close communication
• In tightly coupled systems, processors share memory and a clock, where communication happens through shared memory
• Advantages include increased throughput(completed processes per unit of time)
• More economical, removing the need for multiple data copies
• Faster because work is divided across processors
• They also feature increased reliability and fault tolerance by picking up work from failed processors

Parallel Systems (Symmetric Multiprocessing)

• Symmetric multiprocessing (SMP) model involves each processor running an identical copy of the operating system
• Most modern operating systems support SMP, like UNIX for Multimax computers

Parallel Systems (Asymmetric Multiprocessing)

• An asymmetric multiprocessing model involves each processor being assigned a specific task
• The master processor schedules and allocates work to slave processors
• The master handles I/O and executes only the operating system while the slave runs user programs only
• If the master fails, the entire system cannot perform I/O and the same is true if there are lost slave computations
• This model is more common in extremely large systems

Distributed Systems

• Involves distributing computation among multiple physical processors
• A loosely coupled structure describes two or more computer systems connected via a communication link
• Each processor has its own operating system and local memory
• Processors communicate using message passing over communication lines like high-speed buses or telephone lines
• Advantages of distributed systems include sharing of resources(files and printers)
• Computation can be partitioned for concurrent processing, which improves the system reliability and fault tolerance
• This is also used for communications with electronic mail, ftp etc.

Distributed Systems - Networks

• Local Area Network (LAN)
• Wide Area Network (WAN)
• Metropolitan Area Network (MAN)

Distributed Systems - Client-Server Systems

• The compute-server provides an interface for clients to request services like accessing a database
• The file-server provides an interface for clients to store and retrieve files

Distributed Systems - Peer-to-Peer (P2P)

• P2P systems do not distinguish between clients and servers
• All nodes are considered equals and can act as both client and server
• Nodes can join the network by registering with a central lookup service or broadcasting requests via a discovery protocol
• Napster and Gnutella represent examples of this

Clustered Systems

• Consists of multiple systems working together like multiprocessor systems
• Storage is usually shared via a storage-area network (SAN)
• Provides high-availability through asymmetric(one machine is hot-standby) or symmetric clustering(multiple nodes running applications, monitoring each other)
• Some clusters are used for high-performance computing (HPC), requiring applications to be written with parallelism in mind

Real-Time Systems

• Real-time systems are characterized by the need to supply immediate responses, particularly when sensors bring data
• They are often used as a control device in scientific experiments, medical imaging, industrial control systems, etc.
• Such systems typically have well-defined, fixed-time constraints
• Hard real-time systems have limited or no secondary storage
• Data is stored in short-term memory or read-only memory (ROM)
• These systems conflict with time-sharing systems and are not supported by general-purpose operating systems
• Soft real-time systems have limited utility in industrial control or robotics
• They can be used in multimedia applications, virtual reality requiring advanced operating-system features

Computer-System Architecture

• I/O devices and the CPU can execute concurrently
• Each device controller manages a specific device type and has a local buffer
• The CPU transfers data to and from main memory and local buffers
• Memory controller ensures orderly access to shared memory, synchronizing all shared memory accesses
• Device controllers notify the CPU upon operation completion by creating an interrupt

Storage Structure

• Main memory is a large storage medium that the CPU accesses directly for program execution
• Secondary storage is an extension of main memory, with large, non-volatile storage capacity
• Magnetic disks have rigid metal or glass platters covered with magnetic recording material
• Disk surfaces are logically divided into tracks which are then further divided into sectors
• Disk controller handles the interface between the device and computer

Magnetic Disk Mechanism

• Platters have two magnetic material-covered surfaces
• Platter diameters may range from 1.8 to 5.25 inches
• Disk drives may have thousands of cylinders organized into concentric cylinders, and each track may contain hundreds of sectors
• The transfer rate is the rate at which data flows between the drive and the computer
• Positioning time (or random access-time) consists of both the seek time and rotational latency
• Seek time is the time required to move the disk arm to the desired cylinder
• Rotational latency is the time required for the desired sector to rotate to the disk head

Memory & Storage Structure

• Floppy disks are composed of one platter and a head that sits directly on the surface
• Consisting of less storage at 1.4MB, it features slower speeds compared to hard disks, and are generally removable
• The load moves a word from main memory to an internal register for CPU execution
• The store instruction moves the content of a register to main memory
• Programs/data are generally not stored in the main memory permanently because the main memory's maximum storage is limited
• Main memory (RAM) is volatile, losing it's contents when power is turned off or lost
• Therefore, data is generally backed up in secondary storage such as disks, CDs, drives etc.

Storage Hierarchy

• Storage systems are organized based on levels
• Attributes include Speed, Cost, Size, and Volatility
• Caching involves copying information into faster storage system
• Main memory acts as a last cache for secondary storage
• CPU registers, caches, and main memory are volatile
• All storage beyond main memory is non-volatile
• Caching first checks cache memory, upon data absence it goes to main memory and assumes that there is a high probability that it will be needed again
• Data must be moved from secondary storage into main memory before being used
• Data transfer from cache to CPU and registers is a hardware function without OS control
• Data transfer from disk to memory is controlled by the operating system
• In multi-processor systems, when any shared data is held by 2+ caches, Cache Coherency and Consistency ensures that no 2 caches hold different values of the shared data simultaneously

Hardware Protection

• Dual-Mode Operation
• I/O Protection
• Memory Protection

Dual-Mode Operation

• Addresses shared resource concerns
• Ensures that an incorrect program does not impact the correct running of other programs
• Hardware supports two different modes of operation, User Mode and Monitor Mode
• User mode is execution done on behalf of a user
• Monitor mode is execution done by the operating system
• At system boot time, the hardware starts in Monitor mode; the Operating system is loaded and starts user processes in User mode
• Whenever an interrupt occurs, the hardware switches from User mode to Monitor mode
• Whenever the O.S. gains control of the computer, it is in Monitor mode
• MS-DOS for 8088 architecture does not have a dual mode and can be wiped over or written down
• MS-Widows NT and IBM OS/2 take advantage of dual mode feature and provide greater protection for the O.S
• A mode bit is added to the computer hardware to indicate the current mode: monitor (0) or user (1)
• When either an interrupt or fault occurs, the hardware switches to monitor mode
• Privileged instructions can only be issued in monitor mode

I/O Protection

• All I/O instructions are privileged instructions of monitor mode
• This ensures that a user program cannot gain control of the computer in monitor

System Call to Perform I/O

• Only the operating system can process I/O instructions and operations
• A user program executes the system call to request the operating system perform I/O

Memory Protection

• The operating system and user programs must have memory protected from each other
• In order to have memory protection, two registers can be added to define a range of legal addresses
• The base register holds the smallest legal physical memory address of code
• The limit register contains the size of the range
• Any memory access outside of what range is protected
• This protection is accomplished by CPU hardware comparing every address generated in User mode with the registers
• The O.S. can load the base and limit registers