Linux 5.16's New Cluster Scheduling Is Causing Regression, Further Hurting Alder Lake

I fired up some other Intel systems and there weren't these regressions with Linux 5.16. That included some Ice Lake Xeon, Tiger Lake, and Rocket Lake systems. In those cases, no regressions on 5.16 were spotted in these same workloads. So what the heck is going on for Alder Lake with Linux 5.16? It was another weekend of kernel bisecting of an upstream performance regression... At least with using the Core i9 12900K for bisecting, the 24 threads made for quick kernel builds.

It quickly became clear that the same commit was responsible for this assortment of performance regressions...

The problematic commit was found and in retrospect makes sense though unfortunate it wasn't spotted more quickly by Intel engineers.