RTLA Requests and TODOs
summary
RTLA is a set of user-space binaries that measure scheduling latency and root‑cause it: a timerlot sampler flags latencies over ~30 µs and breaks them into kernel‑side events; an OS noise tracer emits per‑CPU interrupt/dispatch counts and execution times; a hardware noise tool reports device glitches. Immediate requests: ftrace histogram support, ability to enable two tracers in one instance, trace-cmd.dat recording, parse per_cpu/cpu/trace_pipe, user-space access to per‑CPU data (via per-CPU mmap or similar), execution-time tracer and perf counter integration, and KVM vCPU correlation.
AI-written summary · about
chapters (11)
- 0:00Introduction and motivation
- 1:04What is RTLA and its tools
- 3:11Demo of timerlat tool
- 4:56User space and kernel space integration
- 7:01OS noise tracer and future work
- 10:15Sharing per-CPU data between spaces
- 12:26BPF and memory mapping options
- 14:13Auto analysis for other tools
- 17:52Pseudo-random workload generator
- 19:42Future tools and requests
- 20:46Running multiple tracers together
related talks
09mrtla: what is next?
51mRTLA Timerlat - the next step in scheduling latency analysis
50mrtla timerlat: Debugging Real-Time Linux Scheduling Latency
28mrtla osnoise: what is missing?
59mProposing a new tracer to monitor RT task behavior
47mReal-Time Linux: What Is Next?
26mReal-time Linux: what is next?
40mRT Linux in Safety-Critical Systems: The Potential and the Challenges