KS2011: Tracing for large-scale data centers

"The solution, according to Ingo, is to move the tools closer to the source of the events; in this case, that means putting the tools into the kernel source tree. The tooling, he said, is even more important than the kernel side of things. [...] But that is how it goes, responded Ingo; if the tool is useful and the code is open we will not break it."

Aren't those outcomes opposites of each other? Even if one accepts a fuzzy pseudo-ABI treatment for some parts of the kernel-userspace interface, the moment some "useful and open" external tool uses it, it'd have to be preserved. At that point, there's no point moving more tool sources into the kernel tree, just to enable changing that pseudo-ABI more easily.