Skip to main content

Linux Kernel Mainline Real-Time History, Support and Experience Based on Robotic and Automotive Projects

UD2.208 (Decroly) | Day 2 | 09:00 - 09:30 | Speakers: Pavel Pisa

Linux Kernel Mainline Real-Time History, Support and Experience Based on Robotic and Automotive Projects
A picture of a devroom at FOSDEM 2024
Open in browser

Notes

Abstract

In September 2024, the last modification required to enable PREEMP_RT for the x86, AArch64 and RISC-V architectures was submitted to the main development tree by Thomas Gleixner (Linutronix). Even the previous real-time-related enhancements brought so many valuable fixes and features for kernel deployment outside the RT area that Linus Torvalds recognized the community around RT deployment of GNU/Linux as valuable and worthy of gradually incorporating its work over the years. The first meeting of the community interested in using GNU/Linux in RT applications took place in Vienna exactly 25 years ago (Real Time Linux Workshop). Another round anniversary is 20 years since kernel list debate following by possible solutions for real-time Linux deployment debate at the Real Time Linux Workshop conference at Lanzhou University in China. The first minor modifications (runtime locking correctness validator) were included in the main development tree by that time as well as mutex priority inheritance, the fundamental requirement for each RTOS. To date, about 10,000 changes have been submitted to the main kernel tree directly by the RT development team and 5,000 changes in the related drivers and subsystems added by their maintainers in response to patch requests and other needs.

In this talk, I will present the development history, highlight and gaps learned, testing for automotive communication etc. because I have followed the development from the beginning to allow our applications and projects realization.

Attachments

Speakers

Pavel Pisa

Notice: The placeholder video image is licensed under CC BY-SA 4.0. The original image can be found hereChanges made to the image are: Cropped the image to a new ratio, part of the image was cut off.