Memory Management in OS: Techniques, and Functions

What is Memory Management in OS?

The main memory, primarily called the Random Access Memory (RAM), meets the memory requirement for all the tasks performed on the operating system. RAM is temporarily stored and emptied once the computer is shut down. This memory is categorized into addressable cells, each capable of containing data in bytes.

Memory management refers to organizing memory units. It is achieved through data transfer from primary to secondary memory and vice versa. The transfer, as needed, enhances system performance, improves memory utilization, and increases efficiency during concurrent operations. Memory management also deals with tracking allocated and unallocated memory and available memory.

Advance your career as a MERN Stack Developer. Master top skills such as MongoDB, Express.js, React, and Node.js (MERN), plus GIT, HTML, CSS, and JavaScript to build and deploy interactive applications and services. 🎯

Why is Memory Management Required?

Operating systems come with limited space or memory for various tasks. Memory management in OS allows easy access, whenever required, for storing diverse data types and executing processing. Hence, effective management aims to maximize memory utilization as per the relevance and thus enhances the device’s performance, stability, and speed. Diving more into necessity, here are the detailed insights:

• It allows the execution of heavy programs even with small memory availability, all with the help of proper memory allocation. Memory swapping plays a key role in this task
• It offers abstraction to the primary memory, allowing running programs to assume the allocation of large amounts of memory to themselves
• Memory manager prevents memory corruption, which can lead to malfunction and inability to carry out processes
• It lowers memory fragmentation and ensures data integrity during process execution
• The access permission and memory isolation also contribute to enhanced security by offering only limited access to specific memory spaces

Memory Management Techniques in OS

There are two memory management techniques in OS. These are static and dynamic loading.

Technique 1: Static Loading

Here, memory is allocated to a fixed address during the program launch. Static loading comprises the loading of both executable data and code and may not result in utilizing all the loaded components. Hence, static loading is an inefficient process that requires more space. However, this technique simplifies the program distribution.

Technique 2: Dynamic Loading 

Here, loading the program’s components takes place only on requirement during execution. It leads to the loading of essential required elements only. Hence, the dynamic loading technique is based on a flexible approach that ensures efficient memory usage. The method is better suited for programs comprising extensive functionalities or libraries not utilized in every session. This memory management technique is used in modern operating systems of both mobiles and desktops.

Memory Allocation Strategies

The strategies of memory management in operating systems can be categorized as follows:

1. Uniprogramming

2. Multiprogramming

a. Contiguous memory allocation

• Single contiguous memory allocation
• Partitioned allocation
• Dynamic memory allocation

b. Non-contagious memory allocation

• Paged memory management
• Segmented memory management

Let’s dive into the details of each strategy of memory management in OS. The contagious and non-contagious memory allocation makes the key approaches to memory management. The uniprogramming involves a single contiguous memory for running programs and operating systems. Information on multiprogrammed memory allocation is covered below.

2.a.i. Single Contiguous Memory Allocation

• It is the most straightforward and earliest memory management technique. RAM is divided into two parts, one for the operating system and the other for the user process.
• The operating system partition is used during computer startup, and the user process allocation stores user-related data. The approach is used in MS-DOS. However, it possesses challenges due to a lack of process isolation and increased memory waste.

2.a.ii. Partitioned Allocation

• It involves creating predetermined equal-sized partitions in the RAM. The fixed-size allocation leads to the use of excess memory even for small running programs.
• Further, the extensive program may lack the required space for proper functionality. The partitioned allocation also results in internal and external fragmentation.

2.a.iii. Dynamic Memory Allocation

Also known as buddy memory allocation, it involves dividing RAM into multiple blocks in the range of powers of 2, such as 2kB, 4kB, 8kB, and so on. As per the request, memory management in OS begins with allocating available sizes that fit the purpose of the request. It also splits and merges the blocks depending on availability.

The binary tree structure handles the allocation and deallocation of multiple-sized blocks. This method is effective as it reduces fragmentation while improving memory allocation efficiency. The Linux Kernel uses dynamic memory allocation. 

2. b. Non-contiguous Memory Allocation 

This technique is based on non-linear memory allocation, where processes receive memory according to availability. It is a complex technique but effective in dealing with fragmentation issues. It is commonly used in modern operating systems.

2.b.i. Paging Memory Management

Let’s first look at what paging is in the OS. Paging is the memory management technique that allows access to more than the available physical memory. It achieves so by dividing pages and frames into blocks of the same size. Pages are virtual memory blocks with logical addresses, while page frames are RAM blocks with physical addresses.

It involves swapping the data between primary and secondary blocks, while the page table allows tracking the address mapping between the pages and frames. The approach is faster, reduces fragmentation, exhibits efficiency, and enhances memory utilization.

2.b.ii. Segmentation Memory Management 

The blocks are divided into logical segments, each responsible for distinct process tasks and functions. Each segment possesses segment descriptors comprising information like the base address limit and permissions. The address of each segment is calculated using the offset and base address. Segmentation in operating systems is an efficient but complex approach, limiting its use to specific cases.

Did You Know? 🔍

MongoDB's cloud offering, Atlas, serves over 60 billion requests per day? It’s a go-to solution for MERN developers seeking easy database management.
(Source: MongoDB)

Common Memory Management Issues

Some of the issues encountered during memory management in OS include:

1. Fragmentation 

There can be external and internal fragmentation. External fragmentation involves the absence of free space for memory allocation due to a lack of contiguous availability. It leads to splitting the file into pieces, resulting in storage on different disk parts. Internal fragmentation refers to allocating more than the required amount of memory. It occurs due to fixed-size blocks and results in space wastage.

2. Memory Leak 

This issue of memory management in OS occurs when allocation occurs continuously without freeing up memory space. It results in a memory shortage. 

3. Race Condition 

System crashes, data corruption, and other issues result when two or more processes compete for the same memory location. Operating Systems have semaphores to help deal with such problems.

4. Premature Frees 

The situation arises when memory is freed up, yet cannot be used. It either crashes or exhibits random behavior. A surviving reference to the memory, commonly referred to as a dangling pointer, also remains. The problem occurs in manual memory management.

5. Interface Complexity 

Memory management in OS is essential in interface design for efficient memory allocation if objects are passed between the modules.

6. Inflexible Design 

Memory managers are designed based on assumptions like block size, object lifetime, and reference patterns. Processing time will increase if assumptions are distinct or unaligned with the actual requirements.

7. Memory Reference Location 

Memory managers exhibit better speed and performance in the presence of closely placed memory locations. If the ones used together are placed far apart, it leads to a poor locality of reference.

Become a Software Developer in 6 Months

Memory management in OS plays a significant role in the function performance and speed of delivery. The concept involves various techniques, strategies, and challenges contributing to efficient memory usage. For any individual aiming for a career in software development, detailed insights into memory management are critical for completing routine tasks.

Keeping this in mind, find here the Full Stack Developer—MERN Stack, which covers the essential tools, skills, and techniques in collaboration with IBM. The Master's program is conducted online in an interactive format and offers hands-on project experience. It helps candidates get noticed by top hiring companies.

FAQs

1. What is the difference between stack and heap memory allocation?

Stack allocation is the fast and efficient automatic memory assignment for local variables. The process occurs when a function is called and involves instant memory freeing at the end of the function. Heap memory allocation involves dynamic allocation during execution. It does not include freeing up space after the function ends.

2. What is thrashing in the context of memory management?

It refers to a situation where the OS uses more time for data swapping than process execution, compromising the system's performance.

3. ​How does virtual memory work in modern operating systems?

Virtual memory offers more than the physically available memory. It does so by temporarily transferring data from RAM to disk storage.

4. ​How does an operating system allocate memory to processes?

Memory management in OS allocates space through tracking the allotted space, paging, segmentation, deallocation, swapping, and other methods.