In this repository, we deal with the implementation of the "Weighted Buddy Method for Dynamic Storage Allocation" (proposed by Shen and Peterson) and "OneBin Dynamic Storage Allocation".
-
Updated
Apr 5, 2021 - C
In this repository, we deal with the implementation of the "Weighted Buddy Method for Dynamic Storage Allocation" (proposed by Shen and Peterson) and "OneBin Dynamic Storage Allocation".
An optimized, thread-aware memory allocator in C
Memory allocator in C11: malloc/free with 3 placement policies, debug layer, benchmarks
H100 LLM KV eval: cuMem 2MiB leaf pack (42% VRAM @70%) + vLLM 0.21 TurboQuant 4bit KV (-58% bytes, ~0.86x decode). Eval-only repro.
An arena/linear memory allocator built in C
A malloc replacement in C11: segregated free lists, boundary-tag coalescing, dedicated mmap for large blocks. LD_PRELOAD-able.
A zero-dependency, thread-safe-ready custom memory allocator built from scratch in C using Linux mmap system calls. Features include block splitting and memory coalescing.
A collection of C++ systems programming projects — a DFA-based regex engine (followpos algorithm), a grep utility with ERE support and ANSI highlighting, and custom memory allocators (general-purpose & pool-based).
A bottom-up systems programming lab: memory foundations, allocator primitives, and a Linux custom allocator.
Memory pool for media-sized buffers with a hard budget and return-to-OS policies
A custom global allocator to track memory operations.
A custom implementation of the C standard library's dynamic memory allocators (malloc, free, realloc, calloc). This project explores memory management, data structures for free lists, and systems-level programming in C.
header-only C++20 arena allocator + pool. zero deps, 31 unit tests.
C++ Workshop about C++17th pma and how to do it in C++14. Therefore this is workshop code not high performance code
A malloc replacement in C: one mmap arena, first-fit free list with splitting and coalescing, thread-safe, with a live heap visualizer.
Thread-caching general-purpose malloc written from scratch in C: per-thread heaps with sharded free lists, macOS malloc-zone and LD_PRELOAD replacement, guard mode, benchmarked against the system allocator, mimalloc and jemalloc
To associate your repository with the memory-allocator topic, visit your repo's landing page and select "manage topics."