Real-time scheduling theory aimed at multicore systems has become increasingly sophisticated and diverse as multicore computing hardware becomes more prevalent. Real-time operating systems (RTOSs) are ill-suited for this kind of rapid change, and the slow-moving RTOS ecosystem is falling behind advances in real-time scheduling theory.
In a paper presented at the 2013 IEEE 19th Real-Time and Embedded Technology and Applications Symposium (RTAS 2013), researchers from the University of North Carolina at Chapel Hill describe a middleware solution running in userspace — that is, outside the RTOS kernel. The solution supports preemptive, dynamic-priority, migrating real-time tasks on multicore hardware. Empirical latency and overhead measurements on an eight-core Intel Xeon platform are in the range of ones to tens of microseconds under most-tested configurations. They see this approach as potentially superior to a kernel-based approach for a subset of future real-world real-time applications.
“Bringing Theory into Practice: A Userspace Library for Multicore Real-Time Scheduling” and other papers from RTAS 2013 are available to IEEE Computer Society members and paid subscribers via the Computer Society Digital Library.