Emr
Effective Memory Reclamation
Install / Use
/learn @mpoeter/EmrREADME
I have started a new project based on this work: https://github.com/mpoeter/xenium
All further development is done there!
Effective Memory Reclamation
This repository contains the C++ implementations of various lock-free reclamation schemes that I have implemented in the course of my Master's Thesis.
The implemented schemes are:
- Lock-Free Reference Counting
- Hazard Pointers
- Quiescent State Based Reclamation
- Epoch Based Reclamation
- New Epoch Based Reclamation
- DEBRA
- Stamp-it
The last one (Stamp-it) is a new reclamation scheme that I have developed as part my thesis. All schemes and their implementations are described in detail in the thesis.
The implementations are based on an adapted version of the interfaces proposed by Robison in Policy-Based Design for Safe Destruction in Concurrent Containers. A detailed list of how my version differs from the original proposal can be found in the Thesis.
This repository also contains implementations of Michael and Scott's lock-free queue, as well as Michael and Harris' list based set and hash-map. These data structures are implemented using said interface and can therefore be used with the different reclamation schemes.
The benchmark folder contains a set of benchmarks that were used to analyze the performance impact of the various reclamation schemes in different scenarios and on different architectures; the results can be found in https://github.com/mpoeter/emr-benchmarks
Related Skills
node-connect
344.1kDiagnose OpenClaw node connection and pairing failures for Android, iOS, and macOS companion apps
frontend-design
96.8kCreate distinctive, production-grade frontend interfaces with high design quality. Use this skill when the user asks to build web components, pages, or applications. Generates creative, polished code that avoids generic AI aesthetics.
openai-whisper-api
344.1kTranscribe audio via OpenAI Audio Transcriptions API (Whisper).
qqbot-media
344.1kQQBot 富媒体收发能力。使用 <qqmedia> 标签,系统根据文件扩展名自动识别类型(图片/语音/视频/文件)。
