On the previously mentioned HP server today I was able to get closer to reproducing the situation by testing with bionic (4.15.0-47-generic) instead of xenial (4.4)
On bionic, unlike xenial, even with the BIOS set to "BIOS controlled dynamic" mode, the intel_pstate driver is loaded instead of pcc-cpufreq
Pull power management fix from Rafael Wysocki:
"Fix a relatively old initialization issue in intel_pstate causing the
pcc-cpufreq driver to be used instead of it on some HP Proliant
systems.
This turned into a functional regression during the 4.17 cycle,
because pcc-cpufreq is a scalability disaster and that was amplified
by the idle loop rework done at that time (Rafael Wysocki).
This suggests there has definitely been some related change in this area that sound very much similar to this which is worth further research.
On the previously mentioned HP server today I was able to get closer to reproducing the situation by testing with bionic (4.15.0-47-generic) instead of xenial (4.4)
On bionic, unlike xenial, even with the BIOS set to "BIOS controlled dynamic" mode, the intel_pstate driver is loaded instead of pcc-cpufreq
Found this kernel commit: https:/ /github. com/torvalds/ linux/commit/ bfa54a3a00e2f7f f051a50f3957e4f ca3d73f6e7
Pull power management fix from Rafael Wysocki:
"Fix a relatively old initialization issue in intel_pstate causing the
pcc-cpufreq driver to be used instead of it on some HP Proliant
systems.
This turned into a functional regression during the 4.17 cycle,
because pcc-cpufreq is a scalability disaster and that was amplified
by the idle loop rework done at that time (Rafael Wysocki).
This suggests there has definitely been some related change in this area that sound very much similar to this which is worth further research.