🖲️Operating Systems Unit 3 – Memory Management

What's Memory Management?

• Involves managing computer memory resources to optimize system performance and ensure efficient utilization
• Includes allocating memory to processes, deallocating memory when no longer needed, and managing memory hierarchy
• Handles memory requests from processes and decides how to allocate available memory among them
• Tracks which parts of memory are currently in use and which are free to be allocated
• Manages virtual memory, allowing processes to use more memory than physically available by swapping data between RAM and disk
• Ensures memory protection by preventing unauthorized access to memory and isolating process memory spaces
• Aims to minimize memory fragmentation, which occurs when there are many small, non-contiguous free memory blocks
• Plays a crucial role in the overall performance and stability of an operating system

Types of Memory

• RAM (Random Access Memory) provides fast, temporary storage for active processes and data
  • Volatile memory loses its contents when power is turned off
  • Divided into physical frames, each typically 4KB in size
• ROM (Read-Only Memory) stores permanent, non-volatile data such as firmware and boot instructions
• Cache memory is a high-speed memory located close to the CPU, used to store frequently accessed data and instructions
  • Levels of cache: L1 (fastest, smallest), L2, and L3 (largest, slowest)
• Registers are small, high-speed memory units within the CPU used for temporary storage during instruction execution
• Virtual memory is a technique that allows processes to use more memory than physically available by swapping data between RAM and disk
• Shared memory is a memory region that can be accessed by multiple processes for inter-process communication
• Non-volatile memory retains data even when power is turned off (ROM, flash memory, hard disks)

Memory Allocation Techniques

• Static memory allocation assigns memory to processes at compile-time, before the program execution begins
  • Memory size is fixed and cannot be changed during runtime
  • Suitable for systems with predictable memory requirements
• Dynamic memory allocation assigns memory to processes during runtime, as and when requested
  • Memory is allocated from a heap, a large pool of free memory
  • Allows flexible memory usage and efficient utilization of available memory
• Contiguous memory allocation assigns a single, contiguous block of memory to each process
  • Simplifies memory management but can lead to fragmentation
• Non-contiguous memory allocation allows a process to be allocated memory in non-contiguous blocks
  • Reduces fragmentation but requires more complex memory management
• Buddy system is a memory allocation technique that divides memory into fixed-size blocks (powers of 2) and allocates memory by splitting larger blocks into smaller ones
• Slab allocation is a technique used in kernel memory allocation to efficiently manage memory for frequently allocated objects
• Stack allocation is used for local variables and function call frames, with memory allocated and deallocated in a last-in-first-out (LIFO) order

Virtual Memory Explained

• Virtual memory is a memory management technique that allows processes to use more memory than physically available in RAM
• Provides a virtual address space to each process, which is mapped to physical memory by the operating system
• Enables the execution of processes that require more memory than available RAM by swapping data between RAM and disk
• Divided into fixed-size pages (typically 4KB) that are mapped to physical memory frames
• Address translation hardware (MMU) translates virtual addresses to physical addresses using page tables
• Page tables store the mapping between virtual pages and physical frames
• When a process accesses a virtual address not present in RAM, a page fault occurs, and the operating system loads the required page from disk into memory
• Demand paging loads pages into memory only when they are accessed, reducing memory usage and startup time
• Swapping involves moving entire processes between RAM and disk when memory is scarce
• Thrashing occurs when a system spends more time swapping pages than executing useful work due to insufficient RAM

Paging and Segmentation

• Paging is a memory management scheme that divides virtual memory into fixed-size pages and physical memory into frames
  • Each process has its own page table, which maps virtual pages to physical frames
  • Simplifies memory allocation and enables efficient utilization of memory
• Segmentation is a memory management scheme that divides a process's virtual address space into variable-size segments
  • Each segment represents a logical unit of the program, such as code, data, or stack
  • Provides a more flexible and logical view of memory compared to paging
• Segmentation with paging combines the advantages of both techniques
  • Virtual address space is divided into segments, which are further divided into pages
  • Provides both logical separation and efficient memory management
• Hierarchical paging uses multiple levels of page tables to reduce the size of page tables and improve memory efficiency
• Inverted page tables store the mapping of physical frames to virtual pages, reducing the memory overhead of traditional page tables
• Translation Lookaside Buffer (TLB) is a hardware cache that stores recently used page table entries to speed up address translation

Memory Protection Strategies

• Memory protection is essential to ensure the integrity and security of a system by preventing unauthorized access to memory
• Process isolation ensures that each process has its own private memory space and cannot access memory belonging to other processes
  • Achieved through virtual memory and address space separation
• Memory access control restricts access to memory based on the type of access (read, write, execute) and the privilege level of the accessing entity
• Base and limit registers define the valid range of memory addresses a process can access, preventing out-of-bounds memory accesses
• Virtual memory provides a separate virtual address space for each process, preventing direct access to physical memory
• Memory segmentation can enforce access control by assigning different protection levels to each segment (e.g., read-only code segment)
• Memory protection units (MPUs) are hardware components that enforce memory access control rules set by the operating system
• Kernel memory is protected from user-mode processes to prevent unauthorized modification of critical system data and code
• Secure memory management techniques, such as address space layout randomization (ASLR), help mitigate memory-based attacks by randomizing the location of memory regions

Common Memory Issues

• Memory leaks occur when allocated memory is not properly deallocated, leading to a gradual depletion of available memory
  • Can cause performance degradation and eventual system failure
  • Often caused by programming errors, such as forgetting to free dynamically allocated memory
• Memory fragmentation occurs when there are many small, non-contiguous free memory blocks, making it difficult to allocate large contiguous blocks
  • External fragmentation happens when there is enough total free memory but no single contiguous block large enough to satisfy an allocation request
  • Internal fragmentation occurs when allocated memory is larger than the actual requested size, wasting memory within the allocated block
• Insufficient memory can cause thrashing, where the system spends more time swapping pages than executing useful work
  • Occurs when the total memory demand of running processes exceeds the available physical memory
• Memory corruption happens when the contents of memory are unintentionally modified, leading to program crashes or undefined behavior
  • Can be caused by buffer overflows, invalid pointer operations, or accessing uninitialized memory
• Memory-related security vulnerabilities, such as buffer overflow attacks, can allow attackers to execute arbitrary code or gain unauthorized access to the system
• Out-of-memory (OOM) errors occur when the system is unable to allocate memory due to insufficient free memory, often leading to process termination or system instability
• Shared memory synchronization issues can arise when multiple processes access shared memory concurrently without proper synchronization, leading to data races and inconsistencies

Advanced Memory Management Concepts

• Copy-on-write (COW) is a technique used to optimize memory usage by allowing multiple processes to share the same memory pages initially
  • When a process attempts to modify a shared page, a private copy of the page is created for that process
  • Reduces memory usage and improves performance by deferring memory copying until necessary
• Memory-mapped files allow files to be accessed through memory, enabling efficient file I/O and sharing of memory between processes
  • File contents are mapped to a region of the process's virtual address space
  • Modifications to the mapped memory are reflected in the underlying file
• Kernel same-page merging (KSM) is a memory deduplication technique used in virtualized environments to reduce memory usage
  • Identifies identical memory pages across virtual machines and merges them into a single copy
  • Reduces memory footprint and improves memory utilization in virtualized systems
• Non-uniform memory access (NUMA) is a memory architecture where memory access times depend on the memory location relative to the processor
  • Optimizing memory allocation and thread scheduling based on NUMA topology can improve performance
• Transparent huge pages (THP) is a technique that automatically promotes memory pages to larger sizes (e.g., 2MB or 1GB) to reduce the overhead of page table management
  • Improves performance by reducing the number of page faults and TLB misses
• Memory compression is a technique used to reduce memory usage by compressing infrequently accessed or inactive memory pages
  • Compressed pages are stored in memory and decompressed when accessed
  • Allows more data to be stored in memory, reducing the need for swapping
• Non-volatile memory (NVM) technologies, such as Intel Optane DC persistent memory, blur the line between memory and storage
  • Provide high-capacity, persistent memory that can be accessed using load/store instructions
  • Enable new memory management techniques and data persistence models