Revise Memory management and virtual memory for IB Computer Science (6.2)

The operating system is the ultimate resource manager, and memory is one of its most precious resources. Its primary memory management duties include: keeping track of which parts of memory are currently being used and by which process, deciding which processes (or parts of processes) to load into memory when space becomes available, allocating and deallocating memory space as needed by processes, and ensuring that one process cannot interfere with the memory space of another. Without this careful management, a single faulty program could crash the entire system.

To manage memory effectively, the OS needs a way to divide it up. Two primary methods are paging and segmentation. Paging involves dividing a process's logical address space into fixed-size blocks called pages. Physical memory (RAM) is similarly divided into fixed-size frames. The OS maintains a page table for each process, which maps the pages to their corresponding frames in RAM. This allows a process's memory to be non-contiguous, meaning its pages can be scattered across different physical memory locations, which is highly flexible.

Segmentation takes a different approach. Instead of fixed-size blocks, it divides a program into logical units of varying sizes, called segments. A typical program might be segmented into a code segment, a data segment, and a stack segment. Each segment is a contiguous block of memory. This model is more intuitive from a programmer's perspective, as it mirrors the logical structure of the program and allows for granular protection (e.g., making the code segment read-only).

Virtual memory is a powerful technique that allows a system to run programs that are larger than the physical RAM available. It achieves this by using a portion of the secondary storage (like an SSD or HDD) as an extension of RAM. The OS moves data between RAM and this secondary storage area (often called the 'swap file' or 'paging file') as needed. This process is typically implemented using demand paging, where pages are only loaded into RAM when they are actually requested by the CPU.

When a program tries to access a memory address on a page that isn't currently in RAM, a hardware trap called a page fault occurs. The OS's page fault handler then takes over. It finds the required page on the disk, finds a free frame in RAM (or swaps out an existing page if RAM is full), loads the required page into the frame, updates the page table, and resumes the instruction that caused the fault. To the program, this entire process is transparent, apart from a slight delay.

While virtual memory is incredibly useful, it has a significant drawback if overused. If a system has too many processes competing for too little physical memory, it can enter a state known as thrashing. In this state, the OS is constantly swapping pages in and out of memory because processes are frequently faulting for pages they need. The system spends the vast majority of its time servicing page faults rather than executing program code, leading to a catastrophic drop in performance. The CPU utilisation plummets because processes are always waiting for the disk.