cpufreq: intel_pstate: Set P-state upfront in performance mode
authorRafael J. Wysocki <rafael.j.wysocki@intel.com>
Wed, 19 Oct 2016 00:57:22 +0000 (02:57 +0200)
committerRafael J. Wysocki <rafael.j.wysocki@intel.com>
Fri, 21 Oct 2016 20:18:22 +0000 (22:18 +0200)
commita6c6ead14183ea4ec8ce7551e1f3451024b9c4db
tree4d83d3c8a987c477ecae017915c7aee58afceafc
parentc6fe46a79ecd79606bb96fada4515f6b23f87b62
cpufreq: intel_pstate: Set P-state upfront in performance mode

After commit a4675fbc4a7a (cpufreq: intel_pstate: Replace timers with
utilization update callbacks) the cpufreq governor callbacks may not
be invoked on NOHZ_FULL CPUs and, in particular, switching to the
"performance" policy via sysfs may not have any effect on them.  That
is a problem, because it usually is desirable to squeeze the last
bit of performance out of those CPUs, so work around it by setting
the maximum P-state (within the limits) in intel_pstate_set_policy()
upfront when the policy is CPUFREQ_POLICY_PERFORMANCE.

Fixes: a4675fbc4a7a (cpufreq: intel_pstate: Replace timers with utilization update callbacks)
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
drivers/cpufreq/intel_pstate.c