This patch is a follow-up to yesterday's "part 1" which is needed before
applying this patch. You can also find these patches at
!!!Please note: Do neither call pr_perf_get_p_state nor pr_perf_set_p_state
yet. Both perform an acpi-low-level write to an unitialized = random(?) =
Only two things are missing now to complete P-States support: reading the
_PSS and _PCT packages. Is bm_evaluate_object the right call for that?