Custom Memory Allocator in C++

This project is a custom memory allocator written in C++ that demonstrates efficient, low-level memory management. It serves as a replacement for the standard malloc/free or new/delete by managing a pre-allocated pool of memory.

Core Features

This allocator was built to be efficient and reduce memory waste using two key techniques:

1. Block Splitting: When a memory request is made, the allocator finds a free block. If the block is larger than necessary, it is split into two: one part is allocated to the user, and the smaller, leftover part is returned to the free list. This minimizes internal fragmentation.

2. Block Coalescing: When a block of memory is deallocated, the allocator checks its immediate physical neighbors. If an adjacent block is also free, they are merged (coalesced) into a single, larger free block. This fights external fragmentation and ensures large contiguous blocks remain available.

Data Structure Used

The core of this allocator is a doubly linked list that keeps track of all the free memory blocks. The nodes of this list are cleverly stored inside the free blocks themselves (in the BlockHeader), meaning no extra memory is wasted on managing the list.

How to Build and Run

This project is self-contained in main.cpp. You can compile and run it using a C++17 compliant compiler (like g++ or Clang).

# Build the project using the Makefile
make

# Run the executable to see the test cases
./allocator

# (Optional) Clean up the build artifacts
make clean