Summary
Well done on completing this long and useful topic on tracing within the Linux kernel. We began this chapter with an overview of the many available tracing mechanisms on the Linux kernel – the first couple of figures nicely summarized this. A large portion of this chapter dealt with how you can leverage the powerful ftrace kernel infrastructure. It is high-performance and minimally invasive, with pretty much no dependencies, making it ideal for even constrained embedded systems!
To make it easier though, several useful frontends for ftrace exist. We covered using trace-cmd, the KernelShark GUI, and the perf-tools project. We finished the chapter with an introduction to using LTTng for kernel tracing and the Trace Compass GUI for visualizing the trace.
You'll tend to find that one tracing/visualization tool may be superior to another in some respects but is inferior in others. This, of course, is very typical (of pretty much everything – trade-offs,...
