Creating Memory Efficient and High Performance IP Stack for Zephyr OS
summary
Native IP stack for Zephyr replaces the old lwIP-based code with a fully integrated dual IPv4/IPv6 stack that supports multiple simultaneous interfaces, per-interface TX/RX fibers, and kconfig-driven options. It uses fragmented small network buffers (e.g. ~128 B fragments for 802.15.4) chained to allow >1280 B transfers, plus a minimal copy data path and preallocated link‑header space for zero-copy where possible. TCP and RPL code are ported and heavily adapted, unit/component tests run on each commit, and ad‑hoc tests show higher speed and much lower memory footprint.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 1:00Why a Native IP Stack
- 6:50Native IP Stack
- 9:00Key Features
- 14:00Conformance Testing
- 18:30LowLevel Details
- 20:10Merge
- 22:30Duplicate
related talks
45mImplementing Network Protocols in Zephyr Project
20mZephyr Network Subsystem Status and Overview
24mNew Implementation Proposal for Zephyr SD Protocol Stack
37mCreating Bluetooth-based IoT Solutions with Zephyr Project
25mEmpowering High Performance User Space Network Stack on Arm
27mZephyr Porting Efforts in High Performance SoCs
40mOptimizing Zephyr for Peak Performance
30mZephyr ❤️ Internet: How Zephyr Speeds Implementation for New IoT Devices