On Wed, Aug 29, 2018 at 03:08:38PM +0100, Qais Yousef wrote: > It was hard to track the history of this and *why* it's needed.
Yeah, my bad..
So at some point I wanted to do dynamic capacity and dynamic smt gain by using the x86 APERF/MPERF stuff. But it never quite worked and we got stuck with the remnants.