An introduction to real-time Linux
Vložit
- čas přidán 28. 08. 2024
- In 22.04, Canonical announced a beta version of the Ubuntu kernel with the PREEMPT_RT patchset integrated. The new real-time kernel serves extreme latency-dependent use cases and provides deterministic response times to service events. By meeting stringent preemption specifications, real-time is suitable across a broad range of verticals, from telco applications to dedicated devices in industrial automation and robotics.
But does a real-time kernel necessarily result in optimized performance? And what would be the consequences of missing a deadline? If you are a Linux kernel enthusiast or just getting started with real-time systems and want to learn more about PREEMPT_RT, this webinar is for you.
This webinar will provide a broad overview of the target use cases and real-time applications. Starting with the building blocks, we will learn:
What is a real-time system?
What are the use cases of real-time across market segments?
What is the role of the PREEMPT_RT patchset?
How would one know if a low-latency kernel is acceptable over PREEMPT_RT?
And much more. Having set the stage and the necessary context in mind, we will then deep-dive into the Linux kernel build with the out-of-tree PREEMPT_RT patch applied.
Find out more about real time and get in touch at ubuntu.com/rea...
#realtime #linux #iot #ubuntu
Find out more about real time and get in touch at ubuntu.com/real-time
Thank you for sharing your knowledge
What does "not fully upstream" mean in plain talk?
Custom, non integrated to upstream project, patches.
Mainline kernel is a moving target. Realtime is not a requirement for new functionalities in kernel. It is the PREEMPT_RT people that are interested in making subsystems realtime. The subsystem maintainers don’t necessarily agree in that way, especially if it adds extra complexity for subsystem maintenance. So there are always gaps between PREEMPT_RT and mainline, and that’s what makes it is so hard to mainline.