Operating Systems: Intro and Structure

Introduction: Operating Systems

• An operating system (OS) acts as an intermediary/interface between a computer user and the machine's hardware.
• OS goals include executing user programs, solving user problems easily, making the computer system user-friendly, and using computer hardware efficiently.

Computer System Structure

• A computer system has four components: hardware, the operating system, application programs, and users.
• Hardware provides basic computing resources and include CPU, memory, and I/O devices.
• The OS controls and coordinates the use of hardware by applications and users.
• Application programs define how system resources are used to solve user computing problems (e.g., word processors, compilers, web browsers, video games).
• Users can be people, machines, or other computers.

What Operating Systems Do

• Operating systems must cater to different viewpoints.
• Users prioritize convenience and ease of use, caring less about resource utilization.
• Shared computers (mainframes or minicomputers) should keep all users happy.
• Dedicated systems (workstations) are dedicated to individual users, but often use shared resources from servers.
• Handheld computers are resource-poor and optimized for usability and battery life.
• Embedded systems in devices and automobiles may have little to no user interface.

Operating System Definition

• From a hardware perspective, an OS is a resource allocator that manages all resources (CPU, memory, I/O devices).
• It decides between conflicting resource requests and aims for efficient and fair resource use.
• From a software perspective, an OS controls program execution to prevent errors and ensure proper computer use.
• There is no universally accepted definition of what exactly an OS entails.
• One view is that it's everything a vendor ships, which varies widely.
• The "kernel" is the one program running at all times, other programs are system programs (shipped with the OS) or application programs.

Computer Startup

• The bootstrap program, stored in ROM or EEPROM (firmware), is loaded at power-up or reboot.
• It initializes all system aspects and loads the OS kernel to start execution.
• Firmware is programming written to a computing device's read-only memory (ROM).
• EPROM (Erasable Programmable Read-Only Memory) retains its data when the power is switched off.

Computer System Organization

• One or more CPUs and device controllers connect through a common bus, accessing shared memory.
• Concurrent execution of CPUs and devices occurs, competing for memory cycles.

Computer System Operation

• I/O devices and the CPU can execute concurrently.
• Each device controller manages a specific device type and has a local buffer.
• The CPU moves data to/from main memory and local buffers.
• I/O is transferred from the device to the controller's local buffer.
• The device controller informs the CPU upon operation completion by causing an interrupt (system call).
• The device controller is an integral part of a system interpreting signals to/from the CPU.

Common Functions of Interrupts

• Interrupts stop the CPU's current process and transfer control to an interrupt service routine through an interrupt vector, which contains the addresses of all service routines.
• Interrupt architecture must save the address of the interrupted instruction, with the interrupt vector being a look-up table containing addresses/pointers to service routines.
• Incoming interrupts are disabled while another interrupt is being processed to prevent loss.
• A trap/exception is a software-generated interrupt caused by an error or user request, and the OS is responsible for managing them.
• An OS is interrupt-driven (multitasking).

Interrupt Handling

• The OS preserves CPU state by storing registers and the program counter.
• It determines the interrupt type and the appropriate interrupt service routine.
• Polling involves the interrupt controller querying each device to identify the requestor.
• A vectored interrupt system directly identifies the requestor.
• The OS contains separate code segments to handle various interrupt types.

I/O Structure

• After I/O starts, control returns to the user program only upon I/O completion.
• This involves waiting, which causes the CPU to idle until the next interrupt or enter a wait loop (memory contention).
• Only one I/O request can be outstanding at a time.
• After I/O starts, control returns without waiting for I/O completion via a system call to the OS.
• The OS has a device-status table with entries about I/O devices (type, address, state).
• The OS indexes into the I/O device table to get device status and modify the table entry to include interrupts.

Storage Definition and Notation Review

• The basic unit of computer storage is the bit (0 or 1).
• A Byte = 8 bits.
• A word is the architechture's native data unit
• A kilobyte (KB) is 1,024 bytes.
• A megabyte (MB) is 1,024^2 bytes.
• A gigabyte (GB) is 1,024^3 bytes.
• A terabyte (TB) is 1,024^4 bytes.
• A petabyte (PB) is 1,024^5 bytes.

Direct Memory Access Structure

• Direct memory access (DMA) is used for high-speed I/O devices, able to transmit information at memory speeds.
• The device controller transfers data blocks directly to main memory without CPU intervention.
• Only one interrupt is generated per block, not per byte.

Storage Structure

• Main memory is the only directly accessible storage medium for the CPU.
• it is typically volatile.
• Secondary storage, like hard disk or flash memory, provides large, nonvolatile storage capacity but is much slower.
• The disk surface is organized into tracks and sectors, with the disk controller managing the interaction.

Storage Hierarchy

• Storage systems are organized in a hierarchy based on speed, cost, and volatility.
• Caching involves copying information into a faster storage system, with main memory acting as a cache for secondary storage.
• Volatile memory loses its content when power is off, and RAM is volatile memory.

Caching

• Caching is an important design principle implemented at many levels (hardware, OS, software).
• Information in use is copied from slower to faster storage temporarily.
• Faster storage (cache) is checked first to see if information is there before checking slower storage.
• Cache is smaller than the storage it is being used for.
• Cache management is an important design problem including cache size and replacement policy.

Computer-System Architecture

• Most systems use a single general-purpose processor, but some have special-purpose processors.
• Multiprocessor systems are growing with various advantages.
• Including increased throughput, economic scale, and graceful degradation/fault tolerance.
• Asymmetric multiprocessing assigns specific tasks to each processor.
• Symmetric multiprocessing (SMP) allows each processor to perform all tasks within the OS.

Multiprocessing Types

• Asymmetric Multiprocessing (AMP) was the original method for handling multiple CPUs and can be less expensive.
• In AMP, all CPUs are not treated equally, potentially restricting certain CPUs to executing operating system code or I/O operations.
• Symmetric Multiprocessing (SMP) has two or more identical processors that connect to shared main memory, each has full I/O access, and the system is controlled by a single OS.
• The OS instance treats the processsors equally and reserves none for special pruporses.
• SMP is the most common architecture for multiprocessor systems today.

Modern Computing

• Modern computers features system including the cache, CPU, memory, device , DMA and more
• These follow a von Neumann architecture
• Executes through instructions, and data movements

Symmetric Multiprocessing (SMP) Architecture

• In this system architecture CPUs have registers and cache and are connected to a memory

Dual-Core Design

• Dual-core refers to a CPU that includes two complete execution cores on a single silicon chip.

Clustered Systems

• Computer clusters works together as a single stem, controlled and scheduled by software.
• Can provide high-availability service.
• Asymmetric clustering has one machine in hot-standby mode.
• Symmetric clustering has many nodes running together, monitor each other.
• Some clusters are for high-performance computing (HPC) and must use parallelization.

Operating System Structure

• Multiprogramming increases efficiencey by keeping CPU and I/O devices busy antime
• To do so the multiprogramming organizes jobs so CPU always has one to execute
• A subset of total jobs in the system is kept in memory so that processes can switch.
• Timesharing/Multitasking is a logical extension of multiprogramming in which CPU switches jobs so frequently that users can interact with each job while executed.
• If several jobs are ready to run at the same time then CPU scheduling occures
• If processses don't fit in memory, swapping occurs to move items in and out.
• Virtual memory allows execution of processes not completely in memory.

Memory layout for multiprogrammed system

• The memory is laid out so that the operating system, job 1, 2, 3 and 4 are loaded into it.

Operating-System Operations

• Interrupt driven (multitasking) by hardware in the systems.
• Software error or request creates exception or trap (e.g., division by zero or requesting an OS service.).
• Dual-mode operation allows OS to protect other systems through a mode bit.
• A mode bit indicates whether the system is running user code or kernel code.
• There are specific instructions labeled "priveleged", designed only for the kernel.
• A system call changes mode to enter kernel mode, and resets it upon return to user mode.

System transition from user to Kernal

• The Timer prevents infinite loop/process hogging resources
• This is completed by setting an interrupt after a specific period
• The operating system decrements a counter
• When the counter is zero, generate an interrupt
• The system is set up before scheduling the process to regain control or terminate the program that exceeds allotted time

Process Management

• A process is a program in execution
• A process is an active entity, whereas a program is a passive one
• A process needs resources to accomplish its task CPU, memory, I/O, files,and initialization data.
• It occurs when the process termination requires reclaim of any reusable resources
• A single-threaded process has one program counter specifying the next instruction
• Instructions are executed sequentially until completion.
• The OS allows multiple program couters per thread processes/threads.
• There is concurrency by multiplexing processes and threads.

Process Management Activities

• Create and deleting of both user and system processes
• Suspending and resuming processes
• Mechanisms for process synchronization, and comminication
• Providing mechanism for deadlock handling

Memory Management

• All data in memory occurs after processing
• The instructions occur in memory for order to execute
• Manages what's in memory
• CPU is optimised to utilize the computer and its response
• Keeps track of currently used parts by keeping track of the who
• Decides what Processes to move in and out
• Allocates and deallocates memory space

Storage Management

• The OS provides uniform, logical view of information storage.
• Storage abstracts physical properties to logical storage unit - file.
• Each medium is controlled by a device (e.g., disk drive, tape drive).
• Varying properties include access speed, capacity, data-transfer rate, access method (sequential or random).CH12
• Files usually are organized into directories.
• There is access control based on most systems
• There is uniform operation that includes: ➤ Mapping files and directories ➤ Primitives to manipulate files and dirs ➤creating and deleting files and directories

Mass-Storage Management

• Disks used when the data does not fit in memory
• Proper management is important
• Entire speed of system depends on subsystems such as disk.
• Operations ➤ Disk scheduling ➤ Storage allocation ➤ Free-space management
• Tertiary storage used for optical/magnetic
• This must have management
• Depends betweem WORM(write once and read many times) and RW(read/write)

Performance of Various Levels of Storage

• The system is designed so that is can complete explicit or implicit code, depending on the hierarchy
• All storage levels must contain the
• Register Level 1
• Cache Level 2
• Main Memory Level 3
• Disk Storage Level 4

Migration of Integer A from Disk to Register

• Multitasking must use most recent value
• A must be cached
• Multiprocessor must have Cache Coherency in hardware so CPUs have must recent value
• Distributed environment
• Has several copies for data
• Solutions will be covered in CH17

I/O Subsystem

• OS is designed to hide harware operations
• Manages operations
• Memory management ➤ Spooling Buffer cache:Stroing data for performance
• The I/O is device driven
• The system contains hardware drivers.

Protection and Security

• Protection occurs by controlling the access of processes/users in resources
• System does so by limiting what you can access or when you can access
• System defends internal and external attacks ➤ Users require identities
• System knows if there is access ➤ Identifier allows group process to operate
• Previlege Escalation - Users can escalate perms.

Distributed Computing

• Combines network and systems with the help of defined distance.
• Contains multiple types of networks
  • LAN , WAN, MAN
  • the network contains operatings System/
• Communicates via scheme
• Ilusion of single system/Faliure

Special-purpose Systems

• Real embedded (computers)
• Contain limited OS and memory and purpose of data system
• Can use small streams with limited resources

Computing Environments

• Tradional and now are over time, and includes the
• Offices
• Homes
  • and PCs connected on Network
  • Each job give a specific switch time, and has portals

Computing Environments (Cont.)

• Client-Sever computing
• A dumb terminals in PCs
• Server resonds via generated clients
• Compute contains a provided client interface to request services
• File Service store and get data
• They occur in multiple enviornments of course

Peer-to-Peer Computing

• A model that contains services and distribution
• Most include
  • All notes can be used via the server
• Requires a network
  • All register
  • Will have access via request for services and share via protocol

Web-Based Computing

• Web is everywhere in web based systems,
• Most are Network
  • Devices allow the ability to be on load balancers Access of OS will have access to systems.

Open-Source Operating Systems

• The operating system has Code available for binary source
• Provides security
• They can be copyleft in GNU
• This creates public licensing
• OpenSource code
• Code avail
• Contains multiple Operating Systems that contain linux.