This project implements a high-performance ringbuffer utilizing virtual memory and supports the use of hugepages for efficient memory management. The ringbuffer is designed to work with large-scale systems where memory performance is critical, and it allows for fast, lock-free read/write operations where memory is abundant.
- Ringbuffer Design: The core structure is a circular buffer that efficiently handles data in a continuous stream.
- Hugepages Support: The buffer is designed to take advantage of hugepages, reducing overhead associated with page table management and increasing performance for large memory allocations.
- Virtual Memory: The buffer uses virtual memory for flexibility, enabling large buffers that may not fit entirely in physical memory.
- Efficient Memory Usage: By leveraging hugepages, the system minimizes memory fragmentation and optimizes cache utilization.
- Linux-based OS
- Kernel with support for hugepages (typically available by default)
- C compiler (GCC/Clang)
- Standard C library
- Clone the repository:
git clone https://github.com/Irreq/ringbuffer.git cd ringbuffer
In order to use hugepages, you must have twice the number of hugepages available due to the ringbuffer using dual-pages. See the number of available hugepages on the system:
cat /proc/meminfo | grep HugePagesSet hugepages (example for 2 pages of 2MB)
echo 2 > /proc/sys/vm/nr_hugepages