RT Linux in Safety-Critical Systems: The Potential and the Challenges
summary
Shows how Real Time Linux (mainline + Preempt RT) behaves under real workloads by running cyclic and deadline tests (one-hour runs, sometimes with six concurrent kernel builds) and comparing Preempt‑RT vs vanilla kernels. Measures focus on missed-deadline counts, latency and scheduler/VM stats (allocation stalls, compaction and page‑migration failures, nr_mlocked pages). Uses RT‑tests, recommends mlock'ing working sets, examines CPU‑affinity/CPU‑sets and RT throttling, and enables OS and hardware noise tracer to separate firmware/hardware noise from OS noise.
AI-written summary · about
chapters (5)
- 0:00Start
- 5:29Methodology
- 11:04Resource management
- 21:43Understand your workload
- 31:42Deadline scheduling server - in progress
related talks
01h 08mRT Linux in Safety Critical Systems: the potential and the challenges
26mReal-time Linux: what is next?
47mReal-Time Linux: What Is Next?
50mInside the RT Patch
42mBoF: Realtime Linux
44mReal-Time Linux on Embedded Multicore Processors
35mPreempt-RT Raspberry Pi Linux
51mReal Time is Coming to Linux; What Does that Mean to You?
