Summary

In this, our second chapter on CPU scheduling on the Linux OS, you have learned several key things. Among them, you learned how to visualize kernel flow with powerful tools such as LTTng and the Trace Compass GUI, as well as with the trace-cmd(1) utility, a convenient frontend to the kernel's powerful Ftrace framework. You then saw how to programatically query and set any thread's CPU affinity mask. This naturally led to a discussion on how you can programmatically query and set any thread's scheduling policy and priority. The whole notion of being "completely fair" (via the CFS implementation) was brought into question, and some light was shed on the elegant solution called cgroups. You even learned how to leverage the cgroups v2 CPU controller to allocate CPU bandwidth as desired to processes in a sub-group. We then understood that though Linux is a GPOS, an RTL patchset very much exists, which, once applied and the kernel is configured and...