CSC422 Operating Systems - Exam 1

For Dr. Bennet's CSC422 class at Mississippi College

Memory hierarchy

A system where smaller, faster storage (like cache) is used to keep recent data, working with larger, cheaper storage (like main memory) to create a seemingly large and cheap storage system

Address Mapping

The process where the CPU translates virtual addresses used by programs into real addresses before sending them to the memory unit

Real addresses

The actual memory locations used by the hardware
Virtual addresses

Kernel

The core component of an operating system

Applications

Programs designed for specific user tasks, distinct from the operating system kernel and utilities

Utilities

Software tools that provide services between the operating system kernel and applications

Extended machine

The concept that an operating system adds functionality to the machine, making it seem more capable (also sometimes called a virtual machine in this context)

Resource manager

The role of an operating system in allocating and enforcing the use of resources such as memory, disk, and exclusive devices

First Generation (OS History)

The earliest era (1940s-1950s) of computing, characterized by hardware like relays or vacuum tubes, direct programmer interaction, and no operating systems

Second Generation (Batch systems)

An era (mid-1960s to mid-1970s) using transistors, where jobs were collected into batches on magnetic tape and run sequentially by the computer

Third Generation

An era (1965-1980) marked by integrated circuits, multiprogramming, spooling, and time-sharing systems like the IBM 360

Multiprogramming

A technique allowing multiple programs to reside in memory at the same time, combining compute- and I/O-bound jobs to improve CPU utilization

Spooling

A method allowing the CPU and peripheral devices to operate in parallel on a single machine

Time-sharing

An application of multiprogramming that allows multiple users to interact with a computer simultaneously, often via text-based terminals (TTYs)

Multics

A large, ambitious joint project (MIT and ATT) to create a utility-like computing service, which was a commercial failure but advanced technology

Unix

An influential operating system developed by ex-Multics programmers at ATT, known for its "small-is-beautiful" approach and having many descendants

Minicomputers

Smaller and cheaper computers than mainframes, such as the DEC PDP series, on which Unix was first developed

Fourth Generation (Personal computing)

The era (1980-present) defined by integrated CPUs, personal computers (e.g., Apple II, IBM PC), and graphical user interfaces (GUIs)

GUI (Graphical User Interface)

A user interface paradigm, pioneered at Xerox PARC and popularized by Apple Mac, which uses visual elements like windows, icons, and menus

Fifth Generation (Mobile computing)

The current generation of computing, characterized by devices with limited power, computing power, and peripherals, such as smartphones running iOS or Android

Registers

Small, high-speed storage locations within the CPU, including general-purpose registers, the program counter, stack pointer, and program status register

Program counter (PC)

A CPU register that stores the memory address of the next instruction to be executed

Context switch

The process of saving the current CPU state (registers) of one program and loading the saved state of another program to allow the CPU to switch between them

Multi-threaded processor

A CPU designed to run multiple threads concurrently, often by having multiple sets of registers (e.g., Intel's Hyperthreaded technology)

Multi-core

A CPU architecture where multiple processing units (cores) are integrated onto a single chip, essentially providing multiple CPUs

Solid-State storage (flash)

A type of storage (e.g., SSDs) that offers much faster reads than mechanical disks, without concern for operation order, but where writes cause wear and must be distributed

Memory-mapped I/O

A technique where device control registers are assigned memory addresses, allowing the CPU to communicate with devices by reading from and writing to these addresses

Port address space I/O

A technique where device control registers have their own dedicated address space, separate from memory, for CPU communication

Polling

A method of device control where the CPU continuously checks the status of an I/O device until it is ready

Interrupts

Hardware signals that cause the CPU to suspend its current task and transfer control to an interrupt handler in the operating system, used for I/O completion, errors, or periodic clock ticks

DMA (Direct Memory Access)

A hardware feature that allows I/O devices to transfer data directly to and from main memory without involving the CPU, improving efficiency

Traps

Software-initiated events, such as system calls, that cause the CPU to switch from user mode to supervisor (kernel) mode, transferring control to the operating system

CPU modes (User and Supervisor/Kernel)

Distinct operating modes of the CPU; user mode restricts access to privileged instructions (like direct I/O), while supervisor (kernel) mode grants full access to hardware

System Call (Syscall)

An interface through which user programs request services from the operating system kernel, typically involving a trap to switch to kernel mode

Memory Protection

Hardware mechanisms that enforce memory allocations, preventing programs from accessing memory regions they are not authorized to use, and generating an interrupt on violations

Clock interrupt

A periodic interrupt generated by a hardware timer, used by the operating system to regain control, enforce time-sharing, and perform scheduling

Process

An operating system abstraction that represents a running program, encompassing its identity, ownership, communication, control, and its own address space

Address space

The range of memory addresses that a process can refer to

Files

An abstraction used by the operating system to store and organize data on storage devices, characterized by hierarchy, permissions, and handles

Protection

The mechanisms, both hardware and software, used by the operating system to control access to system resources like memory and files

I/O (Input/Output)

Operations that involve data transfer between the computer and the outside world, which are exclusively performed by the OS at the request of user programs

Shell

A program that acts as a command interpreter, allowing users to interact with the operating system, either through a text-based interface or a graphical one

Monolithic structure

An operating system architecture where the entire OS, including all services, runs as a single, large program in kernel mode

Layered system

An operating system architecture where functions are organized into hierarchical layers, with each layer only communicating with the layer directly below it

Microkernels

An operating system architecture that minimizes the kernel by moving most OS services (like file systems and device drivers) into user-space server processes, which communicate via message passing

Virtual machines

Software that simulates a complete hardware environment, allowing one or more guest operating systems to run on top of a host system

Hypervisor (Type 1)

A virtual machine monitor that runs directly on the host hardware without a separate host operating system

Hypervisor (Type 2)

A virtual machine monitor that runs as an application on a host operating system, which then hosts guest operating systems

Paravirtualization

A virtualization technique where the guest operating system is modified to cooperate with the hypervisor, improving performance and efficiency, though sacrificing some generality

Multiprogramming

The rapid switching of the CPU among multiple processes, giving the illusion that they are running concurrently (pseudoparallelism)

Pseudoparallelism

The appearance of multiple processes running at the same time, achieved by rapid CPU switching in a single-processor system

Program in execution

Synonymous with a process; it includes the program code and its complete execution context (CPU state, registers, memory, etc.)

Process switching

The operating system's action of saving the current process's CPU state, updating system records, and loading the state of another process to give it control of the CPU

Process Creation

The act of initiating a new process, typically done by the kernel either during system boot, at the request of an existing process, or by a user command

Process Termination

The ending of a process, which can be voluntary (normal exit, error) or involuntary (killed by another process)

Process Hierarchies (Unix)

A system where processes are organized in a parent-child relationship, forming a tree structure, with process 1 (init) as the ultimate ancestor

Process table (Process Control Block / PCB)

A data structure maintained by the OS for each running process, containing essential information like process ID, state, scheduling parameters, resources, and saved register values

CreateProcess (Windows)

A Windows API function used to create a new process and execute a specified program, returning a handle for future management

Fork (Unix)

A Unix system call that creates an exact copy of the calling process, with both parent and child processes returning from the call

Exec (Unix)

A Unix system call that replaces the current process's memory space with a new program, effectively running a new program within the existing process context and not returning to the caller

CPU Utilization

The percentage of time the CPU is actively performing work, which can be estimated by the formula 1 - p^n, where p is the fraction of time a process waits for I/O and n is the degree of multiprogramming

Degree of multiprogramming

The number of active programs currently loaded into memory at the same time

Thread

A light-weight unit of execution within a process that has its own program counter, registers, and stack, but shares the memory space and other resources (like open files) with other threads in the same process

Traditional process

A process that contains only a single thread of execution

Threaded process (Modern process)

A process that can contain multiple threads of execution

Race condition

A situation that occurs when multiple threads access and modify shared data concurrently, and the final outcome depends on the non-deterministic order in which the threads' operations are interleaved

User space threads

Threads managed entirely by a user-level library without kernel awareness, meaning the kernel only schedules the containing process

Kernel threads

Threads that are directly managed and scheduled by the operating system kernel

Hybrid threads

A threading model that combines aspects of both user-level and kernel-level threads, often by mapping multiple user-level threads to a smaller number of kernel threads

Scheduler Activations

A model where the kernel provides a set of virtual processors (kernel threads) to a user-level thread library, and the kernel notifies the library (via an "up-call") of events like a thread blocking

Up-call

A mechanism where the operating system (kernel) invokes a function in a user-space scheduler, essentially a "system call in reverse"

Pop-up thread

A thread that is automatically created by the system in response to an incoming event, such as a network message, often used for quickly handling requests

Thread-local storage

A mechanism that allows each thread to have its own private copy of data that would otherwise be shared global or static variables, preventing race conditions without explicit synchronization