Use the static SRCU initializer for `cpufreq_transition_notifier_list'.
This avoids the init_cpufreq_transition_notifier_list() initcall. Its
only purpose is to initialize the SRCU notifier once during boot and set
another variable which is used as an indicator whether the init was
perfromed before cpufreq_register_notifier() was used.
Signed-off-by: NSebastian Andrzej Siewior <bigeasy@linutronix.de>
Acked-by: NViresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

If the policy limits are updated via cpufreq_update_policy() and
subsequently via sysfs, the limits stored in user_policy may be
set incorrectly.

For example, if both min and max are set via sysfs to the maximum
available frequency, user_policy.min and user_policy.max will also
be the maximum.  If a policy notifier triggered by
cpufreq_update_policy() lowers both the min and the max at this
point, that change is not reflected by the user_policy limits, so
if the max is updated again via sysfs to the same lower value,
then user_policy.max will be lower than user_policy.min which
shouldn't happen.  In particular, if one of the policy CPUs is
then taken offline and back online, cpufreq_set_policy() will
fail for it due to a failing limits check.

To prevent that from happening, initialize the min and max fields
of the new_policy object to the ones stored in user_policy that
were previously set via sysfs.
Signed-off-by: NKevin Wangtao <kevin.wangtao@hisilicon.com>
Acked-by: NViresh Kumar <viresh.kumar@linaro.org>
[ rjw: Subject & changelog ]
Cc: All applicable <stable@vger.kernel.org>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

cpufreq_notify_transition() calls __cpufreq_notify_transition() for each
CPU of a policy. There is a lot of code in __cpufreq_notify_transition()
though which isn't required to be executed for each CPU, like checking
about disabled cpufreq or irqs, adjusting jiffies, updating cpufreq
stats and some debug print messages.

This commit merges __cpufreq_notify_transition() into
cpufreq_notify_transition() and modifies cpufreq_notify_transition() to
execute minimum amount of code for each CPU.

Also fix the kerneldoc for cpufreq_notify_transition() while at it.
Signed-off-by: NViresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

The cpufreq table is already validated by the cpufreq core and none of
the users of cpufreq_generic_init() have any dependency on it to
validate the table as well.

Don't validate the cpufreq table anymore from cpufreq_generic_init().
Signed-off-by: NViresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

By design, cpufreq drivers are responsible for calling
cpufreq_frequency_table_cpuinfo() from their ->init()
callbacks to validate the frequency table.

However, if a cpufreq driver is buggy and fails to do so properly, it
lead to unexpected behavior of the driver or the cpufreq core at a
later point in time.  It would be better if the core could
validate the frequency table during driver initialization.

To that end, introduce cpufreq_table_validate_and_sort() and make
the cpufreq core call it right after invoking the ->init() callback
of the driver and destroy the cpufreq policy if the table is invalid.

For the time being the validation of the table happens twice, once
from the driver and then from the core.  The individual drivers will
be updated separately to drop table validation if they don't need it
for other reasons.

The frequency table is marked "sorted" or "unsorted" by the new helper
now instead of in cpufreq_table_validate_and_show(), as it should only
be done after validating the table (which the drivers won't do going
forward).
Signed-off-by: NViresh Kumar <viresh.kumar@linaro.org>
[ rjw: Subject/changelog ]
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

Ideally the de-allocation of resources should happen in the exact
opposite order in which they were allocated. It helps maintain the code
in long term, even if nothing really breaks with incorrect ordering.

That wasn't followed in cpufreq_online() and it has some
inconsistencies.  For example, the symlinks were created from within
the locked region while they are removed only after putting the locks.
Also ->exit() should have been called only after the symlinks are
removed and the lock is dropped, as that was the case when ->init()
was first called.
Signed-off-by: NViresh Kumar <viresh.kumar@linaro.org>
[ rjw: Subject ]
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

cpufreq_resume can be called even without preceding cpufreq_suspend.
This can happen in following scenario:

    suspend_devices_and_enter
       --> dpm_suspend_start
          --> dpm_prepare
              --> device_prepare : this function errors out
          --> dpm_suspend: this is skipped due to dpm_prepare failure
                           this means cpufreq_suspend is skipped over
       --> goto Recover_platform, due to previous error
       --> goto Resume_devices
       --> dpm_resume_end
           --> dpm_resume
               --> cpufreq_resume

In case schedutil is used as frequency governor, cpufreq_resume will
eventually call sugov_start, which does following:

    memset(sg_cpu, 0, sizeof(*sg_cpu));
    ....

This effectively erases function pointer for frequency update, causing
crash later on. The function pointer would have been set correctly if
subsequent cpufreq_add_update_util_hook runs successfully, but that
function returns earlier because cpufreq_suspend was not called:

    if (WARN_ON(per_cpu(cpufreq_update_util_data, cpu)))
		return;

The fix is to check cpufreq_suspended first, if it's false, that means
cpufreq_suspend was not called in the first place, so do not resume
cpufreq.
Signed-off-by: NBo Yan <byan@nvidia.com>
Acked-by: NViresh Kumar <viresh.kumar@linaro.org>
[ rjw: Dropped printing a message ]
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

It is possible to remove a cpufreq governor module after
cpufreq_parse_governor() has returned success in
store_scaling_governor() and before cpufreq_set_policy()
acquires a reference to it, because the governor list is
not protected during that period and nothing prevents the
governor from being unregistered then.

Prevent that from happening by acquiring an extra reference
to the governor module temporarily in cpufreq_parse_governor(),
under cpufreq_governor_mutex, and dropping it in
store_scaling_governor(), when cpufreq_set_policy() returns.

Note that the second cpufreq_parse_governor() call site is fine,
because it only cares about the policy member of new_policy.
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: NViresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

Drop a pointless return statement from cpufreq_unregister_governor().
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: NViresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

Pass policy pointer to cpufreq_parse_governor() instead of passing
pointers to two members of it so as to make the code slightly more
straightforward.
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: NViresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

Drop an unnecessary local variable from cpufreq_parse_governor()
and rearrange the code in there to make it easier to follow.
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: NViresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

Frequency-invariant accounting support based on the ratio of current
frequency and maximum supported frequency is an optional feature an arch
can implement.

Since there are cpufreq drivers (e.g. cpufreq-dt) which can be build for
different arch's a default implementation of the frequency-invariance
setter function arch_set_freq_scale() is needed.

This default implementation is an empty weak function which will be
overwritten by a strong function in case the arch provides one.

The setter function passes the cpumask of related (to the frequency
change) cpus (online and offline cpus), the (new) current frequency and
the maximum supported frequency.
Signed-off-by: NDietmar Eggemann <dietmar.eggemann@arm.com>
Acked-by: NViresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

If transition_delay_us isn't defined by the cpufreq driver, the default
value of transition delay (time after which the cpufreq governor will
try updating the frequency again) is currently calculated by multiplying
transition_latency (nsec) with LATENCY_MULTIPLIER (1000) and then
converting this time to usec. That gives the exact same value as
transition_latency, just that the time unit is usec instead of nsec.

With acpi-cpufreq for example, transition_latency is set to around 10
usec and we get transition delay as 10 ms. Which seems to be a
reasonable amount of time to reevaluate the frequency again.

But for platforms where frequency switching isn't that fast (like ARM),
the transition_latency varies from 500 usec to 3 ms, and the transition
delay becomes 500 ms to 3 seconds. Of course, that is a pretty bad
default value to start with.

We can try to come across a better formula (instead of multiplying with
LATENCY_MULTIPLIER) to solve this problem, but will that be worth it ?

This patch tries a simple approach and caps the maximum value of default
transition delay to 10 ms. Of course, userspace can still come in and
change this value anytime or individual drivers can rather provide
transition_delay_us instead.
Signed-off-by: NViresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

CPUFREQ_ENTRY_INVALID is a special symbol which is used to specify that
an entry in the cpufreq table is invalid. But using it outside of the
scope of the cpufreq table looks a bit incorrect.

We can represent an invalid frequency by writing it as 0 instead if we
need. Note that it is already done that way for the return value of the
->get() callback.

Lets do the same for ->fast_switch() and not use CPUFREQ_ENTRY_INVALID
outside of the scope of cpufreq table.

Also update the comment over cpufreq_driver_fast_switch() to clearly
mention what this returns.

None of the drivers return CPUFREQ_ENTRY_INVALID as of now from
->fast_switch() callback and so we don't need to update any of those.
Signed-off-by: NViresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

With the recent updates, CPUFREQ_ETERNAL is only used by the drivers
which don't know their transition latency but want to use dynamic
switching.

Anyway, the routine cpufreq_policy_transition_delay_us() caps the value
of transition latency to 10 ms now and that can be used safely with such
platforms.

Remove the check from cpufreq_init_governor() and allow dynamic
switching for such configurations as well.
Signed-off-by: NViresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

The policy->transition_latency field is used for multiple purposes
today and its not straight forward at all. This is how it is used:

A. Set the correct transition_latency value.

B. Set it to CPUFREQ_ETERNAL because:
   1. We don't want automatic dynamic switching (with
      ondemand/conservative) to happen at all.
   2. We don't know the transition latency.

This patch handles the B.1. case in a more readable way. A new flag for
the cpufreq drivers is added to disallow use of cpufreq governors which
have dynamic_switching flag set.

All the current cpufreq drivers which are setting transition_latency
unconditionally to CPUFREQ_ETERNAL are updated to use it. They don't
need to set transition_latency anymore.

There shouldn't be any functional change after this patch.
Signed-off-by: NViresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: NDominik Brodowski <linux@dominikbrodowski.net>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

There is no limitation in the ondemand or conservative governors which
disallow the transition_latency to be greater than 10 ms.

The max_transition_latency field is rather used to disallow automatic
dynamic frequency switching for platforms which didn't wanted these
governors to run.

Replace max_transition_latency with a boolean (dynamic_switching) and
check for transition_latency == CPUFREQ_ETERNAL along with that. This
makes it pretty straight forward to read/understand now.
Signed-off-by: NViresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

The policy->transition_delay_us field is used only by the schedutil
governor currently, and this field describes how fast the driver wants
the cpufreq governor to change CPUs frequency. It should rather be a
common thing across all governors, as it doesn't have any schedutil
dependency here.

Create a new helper cpufreq_policy_transition_delay_us() to get the
transition delay across all governors.
Signed-off-by: NViresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

The goal of this change is to give users a uniform and meaningful
result when they read /sys/...cpufreq/scaling_cur_freq
on modern x86 hardware, as compared to what they get today.

Modern x86 processors include the hardware needed
to accurately calculate frequency over an interval --
APERF, MPERF, and the TSC.

Here we provide an x86 routine to make this calculation
on supported hardware, and use it in preference to any
driver driver-specific cpufreq_driver.get() routine.

MHz is computed like so:

MHz = base_MHz * delta_APERF / delta_MPERF

MHz is the average frequency of the busy processor
over a measurement interval.  The interval is
defined to be the time between successive invocations
of aperfmperf_khz_on_cpu(), which are expected to to
happen on-demand when users read sysfs attribute
cpufreq/scaling_cur_freq.

As with previous methods of calculating MHz,
idle time is excluded.

base_MHz above is from TSC calibration global "cpu_khz".

This x86 native method to calculate MHz returns a meaningful result
no matter if P-states are controlled by hardware or firmware
and/or if the Linux cpufreq sub-system is or is-not installed.

When this routine is invoked more frequently, the measurement
interval becomes shorter.  However, the code limits re-computation
to 10ms intervals so that average frequency remains meaningful.

Discerning users are encouraged to take advantage of
the turbostat(8) utility, which can gracefully handle
concurrent measurement intervals of arbitrary length.
Signed-off-by: NLen Brown <len.brown@intel.com>
Reviewed-by: NThomas Gleixner <tglx@linutronix.de>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

For a driver that does not set the CPUFREQ_STICKY flag, if all of the
->init() calls fail, cpufreq_register_driver() should return an error.
This will prevent the driver from loading.

Fixes: ce1bcfe9 (cpufreq: check cpufreq_policy_list instead of scanning policies for all CPUs)
Cc: 4.0+ <stable@vger.kernel.org> # 4.0+
Signed-off-by: NDavid Arcari <darcari@redhat.com>
Acked-by: NViresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

cpufreq holds get_online_cpus() while invoking cpuhp_setup_state_nocalls()
to make subsys_interface_register() and the registration of hotplug calls
atomic versus cpu hotplug.

cpuhp_setup_state_nocalls() invokes get_online_cpus() as well. This is
correct, but prevents the conversion of the hotplug locking to a percpu
rwsem.

Use cpuhp_setup/remove_state_nocalls_cpuslocked() to avoid the nested
call. Convert *_online_cpus() to the new interfaces while at it.
Signed-off-by: NSebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Tested-by: NPaul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: NIngo Molnar <mingo@kernel.org>
Acked-by: N"Rafael J. Wysocki" <rafael.j.wysocki@intel.com>
Acked-by: NViresh Kumar <viresh.kumar@linaro.org>
Cc: linux-pm@vger.kernel.org
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Link: http://lkml.kernel.org/r/20170524081547.731628408@linutronix.de

There is a report that after commit 27622b06 ("cpufreq: Convert
to hotplug state machine"), the normal CPU offline/online cycle
fails on some platforms.

According to the ftrace result, this problem was triggered on
platforms using acpi-cpufreq as the default cpufreq driver,
and due to the lack of some ACPI freq method (eg. _PCT),
cpufreq_online() failed and returned a negative value, so the CPU
hotplug state machine rolled back the CPU online process.  Actually,
from the user's perspective, the failure of cpufreq_online() should
not prevent that CPU from being brought up, although cpufreq might
not work on that CPU.

BTW, during system startup cpufreq_online() is not invoked via CPU
online but by the cpufreq device creation process, so the APs can be
brought up even though cpufreq_online() fails in that stage.

This patch ignores the return value of cpufreq_online/offline() and
lets the cpufreq framework deal with the failure.  cpufreq_online()
itself will do a proper rollback in that case and if _PCT is missing,
the ACPI cpufreq driver will print a warning if the corresponding
debug options have been enabled.

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=194581
Fixes: 27622b06 ("cpufreq: Convert to hotplug state machine")
Reported-and-tested-by: NTomasz Maciej Nowak <tmn505@gmail.com>
Signed-off-by: NChen Yu <yu.c.chen@intel.com>
Acked-by: NViresh Kumar <viresh.kumar@linaro.org>
Cc: 4.9+ <stable@vger.kernel.org> # 4.9+
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

The cpufreq core only tries to create symbolic links from CPU
directories in sysfs to policy directories in cpufreq_add_dev(),
either when a given CPU is registered or when the cpufreq driver
is registered, whichever happens first.  That is not sufficient,
however, because cpufreq_add_dev() may be called for an offline CPU
whose policy object has not been created yet and, quite obviously,
the symbolic cannot be added in that case.

Fix that by making cpufreq_online() attempt to add symbolic links to
policy objects for the CPUs in the related_cpus mask of every new
policy object created by it.

The cpufreq_driver_lock locking around the for_each_cpu() loop
in cpufreq_online() is dropped, because it is not necessary and the
code is somewhat simpler without it.  Moreover, failures to create
a symbolic link will not be regarded as hard errors any more and
the CPUs without those links will not be taken offline automatically,
but that should not be problematic in practice.
Reported-and-tested-by: NPrashanth Prakash <pprakash@codeaurora.org>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Cc: 4.9+ <stable@vger.kernel.org> # 4.9+

On CPU online the cpufreq core restores the previous governor (or
the previous "policy" setting for ->setpolicy drivers), but it does
not restore the min/max limits at the same time, which is confusing,
inconsistent and real pain for users who set the limits and then
suspend/resume the system (using full suspend), in which case the
limits are reset on all CPUs except for the boot one.

Fix this by making cpufreq_online() restore the limits when an inactive
policy is brought online.

The commit log and patch are inspired from Rafael's earlier work.
Reported-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: NViresh Kumar <viresh.kumar@linaro.org>
Cc: 4.3+ <stable@vger.kernel.org> # 4.3+
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

There is a missing newline in show_cpuinfo_cur_freq(), so add it,
but while at it clean that function up somewhat too.
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: NViresh Kumar <viresh.kumar@linaro.org>
Cc: All applicable <stable@vger.kernel.org>

Add the "cpufreq.off=1" cmdline option.

At boot-time, this allows a user to request CONFIG_CPU_FREQ=n
behavior from a kernel built with CONFIG_CPU_FREQ=y.

This is analogous to the existing "cpuidle.off=1" option
and CONFIG_CPU_IDLE=y

This capability is valuable when we need to debug end-user
issues in the BIOS or in Linux.  It is also convenient
for enabling comparisons, which may otherwise require a new kernel,
or help from BIOS SETUP, which may be buggy or unavailable.
Signed-off-by: NLen Brown <len.brown@intel.com>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

If new_policy is set in cpufreq_online(), the policy object has just
been created and its real_cpus mask has been zeroed on allocation,
and the driver's ->init() callback should not touch it.

It doesn't need to be cleared again, so don't do that.
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: NViresh Kumar <viresh.kumar@linaro.org>

Its not used anymore, remove it.
Signed-off-by: NViresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

Those were added by:

commit fcd7af91 ("cpufreq: stats: handle cpufreq_unregister_driver()
and suspend/resume properly")

but aren't used anymore since:

commit 1aefc75b ("cpufreq: stats: Make the stats code non-modular").

Remove them. Also remove the redundant parameter to the respective
routines.
Signed-off-by: NViresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

Kernel CPU stats are stored in cputime_t which is an architecture
defined type, and hence a bit opaque and requiring accessors and mutators
for any operation.

Converting them to nsecs simplifies the code and is one step toward
the removal of cputime_t in the core code.
Signed-off-by: NFrederic Weisbecker <fweisbec@gmail.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Link: http://lkml.kernel.org/r/1485832191-26889-4-git-send-email-fweisbec@gmail.comSigned-off-by: NIngo Molnar <mingo@kernel.org>

The return value of cpufreq_update_policy() is never used, so make
it void.
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: NViresh Kumar <viresh.kumar@linaro.org>

There are two places in the cpufreq core in which low-level driver
callbacks may be invoked for an inactive cpufreq policy, which isn't
guaranteed to work in general.  Both are due to possible races with
CPU offline.

First, in cpufreq_get(), the policy may become inactive after
the check against policy->cpus in cpufreq_cpu_get() and before
policy->rwsem is acquired, in which case using it going forward may
not be correct.

Second, an analogous situation is possible in cpufreq_update_policy().

Avoid using inactive policies by adding policy_is_inactive() checks
to the code in the above places.
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: NViresh Kumar <viresh.kumar@linaro.org>

The function cpufreq_register_driver() returns zero on success and since
commit 27622b06 ("cpufreq: Convert to hotplug state machine")
erroneously a positive number. Due to the "if (x) assume_error" construct
all callers assumed an error and as a consequence the cpu freq kworker
crashes with a NULL pointer dereference.

Reset the return value back to zero in the success case.

Fixes: 27622b06 ("cpufreq: Convert to hotplug state machine")
Reported-by: NBorislav Petkov <bp@alien8.de>
Reported-and-tested-by: NIngo Molnar <mingo@kernel.org>
Signed-off-by: NSebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: peterz@infradead.org
Cc: rjw@rjwysocki.net
Link: http://lkml.kernel.org/r/20160920145628.lp2bmq72ip3oiash@linutronix.deSigned-off-by: NThomas Gleixner <tglx@linutronix.de>

Install the callbacks via the state machine.
Signed-off-by: NSebastian Andrzej Siewior <bigeasy@linutronix.de>
Acked-by: N"Rafael J. Wysocki" <rjw@rjwysocki.net>
Cc: linux-pm@vger.kernel.org
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Viresh Kumar <viresh.kumar@linaro.or
Cc: rt@linutronix.de
Link: http://lkml.kernel.org/r/20160906170457.32393-13-bigeasy@linutronix.deSigned-off-by: NThomas Gleixner <tglx@linutronix.de>

If a cpufreq driver is registered very early in the boot stage (e.g.
registered from postcore_initcall()), then cpufreq core may generate
kernel warnings for it.

In this case, the CPUs are brought online, then the cpufreq driver is
registered, and then the CPU topology devices are registered. However,
by the time cpufreq_add_dev() gets called, the cpu device isn't stored
in the per-cpu variable (cpu_sys_devices,) which is read by
get_cpu_device().

So the cpufreq core fails to get device for the CPU, for which
cpufreq_add_dev() was called in the first place and we will hit a
WARN_ON(!cpu_dev).

Even if we reuse the 'dev' parameter passed to cpufreq_add_dev() to
avoid that warning, there might be other CPUs online that share the
policy with the cpu for which cpufreq_add_dev() is called. Eventually
get_cpu_device() will return NULL for them as well, and we will hit the
same WARN_ON() again.

In order to fix these issues, change cpufreq core to create links to the
policy for a cpu only when cpufreq_add_dev() is called for that CPU.

Reuse the 'real_cpus' mask to track that as well.

Note that cpufreq_remove_dev() already handles removal of the links for
individual CPUs and cpufreq_add_dev() has aligned with that now.
Reported-by: NRussell King <rmk+kernel@arm.linux.org.uk>
Tested-by: NRussell King <rmk+kernel@arm.linux.org.uk>
Signed-off-by: NViresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

Since cpufreq_policy_alloc() doesn't use its dev variable for
anything useful, drop that variable from there along with the
NULL check against it.
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: NViresh Kumar <viresh.kumar@linaro.org>

Export cpufreq_driver_resolve_freq() since governors may be compiled as
modules.
Reported-by: NStephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: NSteve Muckle <smuckle@linaro.org>
Acked-by: NViresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

The handlers provided by cpufreq core are sufficient for resolving the
frequency for drivers providing ->target_index(), as the core already
has the frequency table and so ->resolve_freq() isn't required for such
platforms.

This patch disallows drivers with ->target_index() callback to use the
->resolve_freq() callback.

Also, it fixes a potential kernel crash for drivers providing ->target()
but no ->resolve_freq().

Fixes: e3c06236 "cpufreq: add cpufreq_driver_resolve_freq()"
Signed-off-by: NViresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

Cpufreq governors may need to know what a particular target frequency
maps to in the driver without necessarily wanting to set the frequency.
Support this operation via a new cpufreq API,
cpufreq_driver_resolve_freq(). This API returns the lowest driver
frequency equal or greater than the target frequency
(CPUFREQ_RELATION_L), subject to any policy (min/max) or driver
limitations. The mapping is also cached in the policy so that a
subsequent fast_switch operation can avoid repeating the same lookup.

The API will call a new cpufreq driver callback, resolve_freq(), if it
has been registered by the driver. Otherwise the frequency is resolved
via cpufreq_frequency_table_target(). Rather than require ->target()
style drivers to provide a resolve_freq() callback it is left to the
caller to ensure that the driver implements this callback if necessary
to use cpufreq_driver_resolve_freq().
Suggested-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: NSteve Muckle <smuckle@linaro.org>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

Both callers of cpufreq_update_current_freq(), cpufreq_update_policy()
and cpufreq_start_governor(), check cpufreq_suspended before calling
that function, so drop the redundant cpufreq_suspended check from it.
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: NViresh Kumar <viresh.kumar@linaro.org>