Open Source Summit + Embedded Linux Conference North America 2026

This talk offers a comprehensive look at what's changed in the embedded Linux world over the past year. Walt will walk through the latest kernel developments most relevant to embedded developers, survey key userspace projects shaping modern embedded designs, and cover the broader community, industry, and legal landscape — from the status of major processor architectures to initiatives at the Linux Foundation and beyond.

Whether you're tracking changes to subsystems you already rely on or looking for new tools and techniques to improve your workflow, this session will help you stay current in a fast-moving ecosystem. Come find out what's new, what's shifting, and what it means for your embedded Linux work.

There are several different RISC-V based single board computers out in market and coming in future. Yocto project, is a leading embedded linux framework, and RISCV is first tier architecture supported in project, core supports RISCV64 QEMU and runs all tests. This talk will discuss using meta-riscv layers to add the support for new RISCV SBCs. meta-riscv has best practices and pre-existing support for known SBCs which can be used as template to bring-up the new board quickly. The talk will cover the content of meta-riscv in detail and the project setups using Kas and the SBC specific documentation using markdown files, detailing the flashing and build instructions, sharing common details but clearly differentiating board specific intructions.
This talk will also cover the challanges and future roadmap for meta-riscv and RISCV architecture support in Yocto Project.

Transitioning from bare-metal firmware development to Linux kernel development presents unique challenges, particularly in debugging methodologies. Traditional approaches such as halting execution via JTAG alone may not straightforwardly work for embedded Linux.

In this presentation we talk about challenges I faced and techniques I came across to debug Linux MMC block device drivers interactively using tools such as T32/GDB debuggers on embedded systems. This talk briefly covers MMC driver and block layer interactions and key golden breakpoints to use for MMC bus driver debugging. I will also discuss tools and techniques to take full control of eMMC block drivers, generating block IO requests as needed, setting up triggers and probing signals on an oscilloscope for detailed waveform level debugging.

key topics: Embedded Linux setup for interactive debug (single CPU, KASLR, WDT, ramfs, RCU, softlocks), strategic SDHCI breakpoints, GPIO-triggered oscilloscope capture signals, handling filesystem mounts; leveraging mmc_test module for generating controlled transactions for debug.

This talk aims to analyze the performance of most common Linux synchronization primitives under PREEMPT_RT, comparing their behavior across a range of workloads. We’ll explore how each primitive scales with thread count, handles contention, and impacts determinism. Expect graphs, latency histograms, and a few surprises. If you’re building real-time systems or tuning performance, this session will help you make smarter, faster, and safer decisions.

As automotive E/E architectures evolve toward centralized high-performance computing (HPC) real-time operating systems (RTOS) need better visibility into system timing behavior. Additionally software defined vehicle (SDV) trends where open, flexible, non-proprietary software stacks are the goal. In this landscape, legacy concepts like dedicated microcontrollers in-service to HPC safety are being challenged. ADAS workloads are balancing HPC compute power with strict time constraint requirements in real-time behavior. Combining mixed-critical workloads on a single HPC platform is a viable solution for OEMs and Tier 1s building perception, control, and safety-relevant HPC domain controller functions. However, the struggle to ensure real-time performance and deterministic timing is an ongoing challenge.
This presentation will demonstrate how Linux can provide measurable, stable, and predictable real-time performance, enabling ADAS teams to run time-critical functions on a modern automotive-grade Linux stack. The audience will gain confidence in a concept which accelerates development for SDV without proprietary RTOS islands, and charters a path to ASIL B/SIL 3 certification