cpufreq: powernv: Migrate to ->exit() callback instead of ->stop_cpu()
authorViresh Kumar <viresh.kumar@linaro.org>
Wed, 23 Jun 2021 06:01:14 +0000 (11:31 +0530)
committerRafael J. Wysocki <rafael.j.wysocki@intel.com>
Wed, 30 Jun 2021 16:51:43 +0000 (18:51 +0200)
commit952da0c9ab5b047665442dc239cee36d5c9edb98
tree20be4be8eac91f1ef2c2a3e3459d4d693bbcf66a
parent9357a380f90a89a168d505561d11f68272e0e768
cpufreq: powernv: Migrate to ->exit() callback instead of ->stop_cpu()

Commit 367dc4aa932b ("cpufreq: Add stop CPU callback to cpufreq_driver
interface") added the ->stop_cpu() callback to allow the drivers to do
clean up before the CPU is completely down and its state can't be
modified.

At that time the CPU hotplug framework used to call the cpufreq core's
registered notifier for different events like CPU_DOWN_PREPARE and
CPU_POST_DEAD. The ->stop_cpu() callback was called during the
CPU_DOWN_PREPARE event.

This is no longer the case, cpuhp_cpufreq_offline() is called only
once by the CPU hotplug core now and we don't really need two
separate callbacks for cpufreq drivers, i.e. ->stop_cpu() and
->exit(), as everything can be done from the ->exit() callback
itself.

Migrate to using the ->exit() callback instead of ->stop_cpu().

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
[ rjw: Minor changelog edits ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
drivers/cpufreq/powernv-cpufreq.c