The newly introduced Energy Model framework manages power cost tables in
a generic way. Moreover, it supports several types of models since the
tables can come from DT or firmware (through SCMI) for example. On the
other hand, the cpu_cooling subsystem manages its own power cost tables
using only DT data.
In order to avoid the duplication of data in the kernel, and in order to
enable IPA with EMs coming from more than just DT, remove the private
tables from cpu_cooling.c and migrate it to using the centralized EM
framework. Doing so should have no visible functional impact for
existing users of IPA since:
- recent extenstions to the the PM_OPP infrastructure enable the
registration of EMs in PM_EM using the DT property used by IPA;
- the existing upstream cpufreq drivers marked with the
'CPUFREQ_IS_COOLING_DEV' flag all use the aforementioned PM_OPP
infrastructure, which means they all support PM_EM. The only two
exceptions are qoriq-cpufreq which doesn't in fact use an EM and
scmi-cpufreq which doesn't use DT for power costs.
For existing users of cpu_cooling, PM_EM tables will contain the exact
same power values that IPA used to compute on its own until now. The
only new dependency for them is to compile in CONFIG_ENERGY_MODEL.
The case where the thermal subsystem is used without an Energy Model
(cpufreq_cooling_ops) is handled by looking directly at CPUFreq's
frequency table which is already a dependency for cpu_cooling.c anyway.
Since the thermal framework expects the cooling states in a particular
order, bail out whenever the CPUFreq table is unsorted, since that is
fairly uncommon in general, and there are currently no users of
cpu_cooling for this use-case.
Acked-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Quentin Perret <qperret@google.com>
Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Link: https://lore.kernel.org/r/20191030151451.7961-5-qperret@google.com
The core CPU cooling infrastructure has power-related functions
that have only one client: IPA. Since there can be no user of those
functions if IPA is not compiled in, make sure to guard them with
checks on CONFIG_THERMAL_GOV_POWER_ALLOCATOR to not waste space
unnecessarily.
Acked-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Suggested-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Signed-off-by: Quentin Perret <qperret@google.com>
Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Link: https://lore.kernel.org/r/20191030151451.7961-4-qperret@google.com
Replace the CPU device PM QoS used for the management of min and max
frequency constraints in cpufreq (and its users) with per-policy
frequency QoS to avoid problems with cpufreq policies covering
more then one CPU.
Namely, a cpufreq driver is registered with the subsys interface
which calls cpufreq_add_dev() for each CPU, starting from CPU0, so
currently the PM QoS notifiers are added to the first CPU in the
policy (i.e. CPU0 in the majority of cases).
In turn, when the cpufreq driver is unregistered, the subsys interface
doing that calls cpufreq_remove_dev() for each CPU, starting from CPU0,
and the PM QoS notifiers are only removed when cpufreq_remove_dev() is
called for the last CPU in the policy, say CPUx, which as a rule is
not CPU0 if the policy covers more than one CPU. Then, the PM QoS
notifiers cannot be removed, because CPUx does not have them, and
they are still there in the device PM QoS notifiers list of CPU0,
which prevents new PM QoS notifiers from being registered for CPU0
on the next attempt to register the cpufreq driver.
The same issue occurs when the first CPU in the policy goes offline
before unregistering the driver.
After this change it does not matter which CPU is the policy CPU at
the driver registration time and whether or not it is online all the
time, because the frequency QoS is per policy and not per CPU.
Fixes: 67d874c3b2 ("cpufreq: Register notifiers with the PM QoS framework")
Reported-by: Dmitry Osipenko <digetx@gmail.com>
Tested-by: Dmitry Osipenko <digetx@gmail.com>
Reported-by: Sudeep Holla <sudeep.holla@arm.com>
Tested-by: Sudeep Holla <sudeep.holla@arm.com>
Diagnosed-by: Viresh Kumar <viresh.kumar@linaro.org>
Link: https://lore.kernel.org/linux-pm/5ad2624194baa2f53acc1f1e627eb7684c577a19.1562210705.git.viresh.kumar@linaro.org/T/#md2d89e95906b8c91c15f582146173dce2e86e99f
Link: https://lore.kernel.org/linux-pm/20191017094612.6tbkwoq4harsjcqv@vireshk-i7/T/#m30d48cc23b9a80467fbaa16e30f90b3828a5a29b
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
The cpufreq core now takes the min/max frequency constraints via QoS
requests and the CPUFREQ_ADJUST notifier shall get removed later on.
Switch over to using the QoS request for maximum frequency constraint
for cpu_cooling driver.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
[ rjw: Subject ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The CPU load values passed to the thermal_power_cpu_get_power
tracepoint are zero for all CPUs, unless, unless the
thermal_power_cpu_limit tracepoint is enabled too:
irq/41-rockchip-98 [000] .... 290.972410: thermal_power_cpu_get_power:
cpus=0000000f freq=1800000 load={{0x0,0x0,0x0,0x0}} dynamic_power=4815
vs
irq/41-rockchip-96 [000] .... 95.773585: thermal_power_cpu_get_power:
cpus=0000000f freq=1800000 load={{0x56,0x64,0x64,0x5e}} dynamic_power=4959
irq/41-rockchip-96 [000] .... 95.773596: thermal_power_cpu_limit:
cpus=0000000f freq=408000 cdev_state=10 power=416
There seems to be no good reason for omitting the CPU load information
depending on another tracepoint. My guess is that the intention was to
check whether thermal_power_cpu_get_power is (still) enabled, however
'load_cpu != NULL' already indicates that it was at least enabled when
cpufreq_get_requested_power() was entered, there seems little gain
from omitting the assignment if the tracepoint was just disabled, so
just remove the check.
Fixes: 6828a4711f ("thermal: add trace events to the power allocator governor")
Signed-off-by: Matthias Kaehlcke <mka@chromium.org>
Reviewed-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Acked-by: Javi Merino <javi.merino@kernel.org>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Eduardo Valentin <edubezval@gmail.com>
The structure cpufreq_cooling_device provides a backpointer to the thermal
device but this one is used for a trace and to unregister. For the trace,
we don't really need this field and the unregister function as the same
pointer passed as parameter. Remove it.
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Signed-off-by: Eduardo Valentin <edubezval@gmail.com>
The copyright format does not conform to the format requested by
Linaro: https://wiki.linaro.org/Copyright
Fix it.
Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Eduardo Valentin <edubezval@gmail.com>
When the static power computation was removed, the test with the power
being negative was not removed. However, the substraction which was
responsible of the negative value was removed and the variable is now
an u32. A double reason to remove the test which does not make sense.
Fixes: 84fe2cab48 ("cpu_cooling: Drop static-power related stuff")
Cc: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Eduardo Valentin <edubezval@gmail.com>
Make it clear that it is a failure if the cpufreq driver was unable to
register as a cooling device. Makes it easier to find in logs and
grepping for words like fail, err, warn.
Signed-off-by: Amit Kucheria <amit.kucheria@linaro.org>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Eduardo Valentin <edubezval@gmail.com>
No one has used it for the last two and half years (since it was
introduced by commit c36cf07176 (thermal: cpu_cooling: implement the
power cooling device API), get rid of it.
Acked-by: Eduardo Valentin <edubezval@gmail.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
of_cpufreq_cooling_register() isn't used by anyone and so can be
removed, but then we would be left with two routines:
cpufreq_cooling_register() and of_cpufreq_power_cooling_register() that
would look odd.
Remove current implementation of of_cpufreq_cooling_register() and
rename of_cpufreq_power_cooling_register() as
of_cpufreq_cooling_register(). This simplifies lots of stuff.
Acked-by: Eduardo Valentin <edubezval@gmail.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
It isn't used by anyone, drop it.
Acked-by: Eduardo Valentin <edubezval@gmail.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
All the callers of of_cpufreq_power_cooling_register() have almost
identical code and it makes more sense to move that code into the helper
as its all about reading DT properties.
This got rid of lot of redundant code.
Acked-by: Eduardo Valentin <edubezval@gmail.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
pr_err() messages should end with a new-line to avoid other messages
being concatenated.
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Acked-by: Javi Merino <javi.merino@kernel.org>
Signed-off-by: Arvind Yadav <arvind.yadav.cs@gmail.com>
Signed-off-by: Eduardo Valentin <edubezval@gmail.com>
This shrinks the size of the structure on arm64 by 8 bytes by avoiding
padding of 4 bytes at two places.
Also add missing doc comment for freq_table
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Lukasz Luba <lukasz.luba@arm.com>
Tested-by: Lukasz Luba <lukasz.luba@arm.com>
Signed-off-by: Eduardo Valentin <edubezval@gmail.com>
The frequency table shouldn't have any zero frequency entries and so
such a check isn't required. Though it would be better to make sure
'state' is within limits.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Lukasz Luba <lukasz.luba@arm.com>
Tested-by: Lukasz Luba <lukasz.luba@arm.com>
Signed-off-by: Eduardo Valentin <edubezval@gmail.com>
'cpu_dev' is used by only one function, get_static_power(), and it
wouldn't be time consuming to get the cpu device structure within it.
This would help removing cpu_dev from struct cpufreq_cooling_device.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Lukasz Luba <lukasz.luba@arm.com>
Tested-by: Lukasz Luba <lukasz.luba@arm.com>
Signed-off-by: Eduardo Valentin <edubezval@gmail.com>
The frequency passed to get_level() is returned by cpu_power_to_freq()
and it is guaranteed that get_level() can't fail.
Get rid of error code.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Lukasz Luba <lukasz.luba@arm.com>
Tested-by: Lukasz Luba <lukasz.luba@arm.com>
Signed-off-by: Eduardo Valentin <edubezval@gmail.com>
We keep two arrays for idle time stats and allocate memory for them
separately. It would be much easier to follow if we create an array of
idle stats structure instead and allocate it once.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Lukasz Luba <lukasz.luba@arm.com>
Tested-by: Lukasz Luba <lukasz.luba@arm.com>
Signed-off-by: Eduardo Valentin <edubezval@gmail.com>
The cpu_cooling driver keeps two tables:
- freq_table: table of frequencies in descending order, built from
policy->freq_table.
- power_table: table of frequencies and power in ascending order, built
from OPP table.
If the OPPs are used for the CPU device then both these tables are
actually built using the OPP core and should have the same frequency
entries. And there is no need to keep separate tables for this.
Lets merge them both.
Note that the new table is in descending order of frequencies and so the
'for' loops were required to be fixed at few places to make it work.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Lukasz Luba <lukasz.luba@arm.com>
Tested-by: Lukasz Luba <lukasz.luba@arm.com>
Signed-off-by: Eduardo Valentin <edubezval@gmail.com>
'allowed_cpus' is a copy of policy->related_cpus and can be replaced by
it directly. At some places we are only concerned about online CPUs and
policy->cpus can be used there.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Lukasz Luba <lukasz.luba@arm.com>
Tested-by: Lukasz Luba <lukasz.luba@arm.com>
Signed-off-by: Eduardo Valentin <edubezval@gmail.com>
The OPPs are registered for all CPUs of a cpufreq policy now and we
don't need to run the loop in build_dyn_power_table(). Just check for
the policy->cpu and we should be fine.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Lukasz Luba <lukasz.luba@arm.com>
Tested-by: Lukasz Luba <lukasz.luba@arm.com>
Signed-off-by: Eduardo Valentin <edubezval@gmail.com>
The cpufreq policy can be used by the cpu_cooling driver, lets store it
in the cpufreq_cooling_device structure.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Lukasz Luba <lukasz.luba@arm.com>
Tested-by: Lukasz Luba <lukasz.luba@arm.com>
Signed-off-by: Eduardo Valentin <edubezval@gmail.com>
We need such a routine at two places already, lets create one.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Lukasz Luba <lukasz.luba@arm.com>
Tested-by: Lukasz Luba <lukasz.luba@arm.com>
Signed-off-by: Eduardo Valentin <edubezval@gmail.com>
The CPU cooling driver uses the cpufreq policy, to get clip_cpus, the
frequency table, etc. Most of the callers of CPU cooling driver's
registration routines have the cpufreq policy with them, but they only
pass the policy->related_cpus cpumask. The __cpufreq_cooling_register()
routine then gets the policy by itself and uses it.
It would be much better if the callers can pass the policy instead
directly. This also fixes a basic design flaw, where the policy can be
freed while the CPU cooling driver is still active.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Lukasz Luba <lukasz.luba@arm.com>
Tested-by: Lukasz Luba <lukasz.luba@arm.com>
Signed-off-by: Eduardo Valentin <edubezval@gmail.com>
'cpu' is used at only one place and there is no need to keep a separate
variable for it.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Lukasz Luba <lukasz.luba@arm.com>
Tested-by: Lukasz Luba <lukasz.luba@arm.com>
Signed-off-by: Eduardo Valentin <edubezval@gmail.com>
There is only one user of cpufreq_cooling_get_level() and that already
has pointer to the cpufreq_cdev structure. It can directly call
get_level() instead and we can get rid of cpufreq_cooling_get_level().
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Lukasz Luba <lukasz.luba@arm.com>
Tested-by: Lukasz Luba <lukasz.luba@arm.com>
Signed-off-by: Eduardo Valentin <edubezval@gmail.com>
Objects of "struct thermal_cooling_device" are named a bit
inconsistently. Lets use cdev everywhere.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Lukasz Luba <lukasz.luba@arm.com>
Tested-by: Lukasz Luba <lukasz.luba@arm.com>
Signed-off-by: Eduardo Valentin <edubezval@gmail.com>
Objects of "struct cpufreq_cooling_device" are named a bit
inconsistently. Lets use cpufreq_cdev everywhere. Also note that the
lists containing such devices is renamed similarly too.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Lukasz Luba <lukasz.luba@arm.com>
Tested-by: Lukasz Luba <lukasz.luba@arm.com>
Signed-off-by: Eduardo Valentin <edubezval@gmail.com>
Just to make it look better.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Lukasz Luba <lukasz.luba@arm.com>
Tested-by: Lukasz Luba <lukasz.luba@arm.com>
Signed-off-by: Eduardo Valentin <edubezval@gmail.com>
After the lock is dropped, it is possible that the cpufreq_dev gets
freed before we call get_level() and that can cause kernel to crash.
Drop the lock after we are done using the structure.
Cc: 4.2+ <stable@vger.kernel.org> # 4.2+
Fixes: 02373d7c69 ("thermal: cpu_cooling: fix lockdep problems in cpu_cooling")
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Lukasz Luba <lukasz.luba@arm.com>
Tested-by: Lukasz Luba <lukasz.luba@arm.com>
Signed-off-by: Eduardo Valentin <edubezval@gmail.com>
It is possible for dev_pm_opp_find_freq_exact() to return errors. It was
all fine earlier as dev_pm_opp_get_voltage() had a check within it to
check for invalid OPPs, but dev_pm_opp_put() doesn't have any similar
checks and the callers need to make sure OPP is valid before calling
them.
Also update the later dev_warn_ratelimited() to not print the error
message as the OPP is guaranteed to be valid now.
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
There isn't much the user can do on seeing these warnings, as the
hardware is actually okay. dev_err suits much better here.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
cooling_list_lock is covering not just cpufreq_dev_count, but also the
calls to cpufreq_register_notifier() and cpufreq_unregister_notifier().
Since cooling_list_lock is also used within cpufreq_thermal_notifier(),
lockdep reports a potential deadlock. Fix it by testing the condition
under cooling_list_lock and dropping the lock before calling
cpufreq_register_notifier(). And variable cpufreq_dev_count is removed
at the same time, because it's no longer needed after the fix.
Fixes: ae60608962 ("thermal: convert cpu_cooling to use an IDA")
Reported-and-Tested-by: Russell King <linux@armlinux.org.uk>
Signed-off-by: Matthew Wilcox <mawilcox@microsoft.com>
Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Pull thermal management updates from Zhang Rui:
- add thermal driver for R-Car Gen3 thermal sensors.
- add thermal driver for ZTE' zx2967 family thermal sensors.
- convert thermal ID allocation from IDR to IDA.
- fix a possible NULL dereference in imx thermal driver.
- fix a ti-soc-thermal driver dependency issue so that critical thermal
control is still available when CPU_THERMAL is not defined.
- update binding information for QorIQ thermal driver.
- a couple of cleanups in thermal core, intel_powerclamp, exynos,
dra752-thermal, mtk-thermal driver.
* 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/rzhang/linux:
powerpc/mpc85xx: Update TMU device tree node for T1023/T1024
powerpc/mpc85xx: Update TMU device tree node for T1040/T1042
dt-bindings: Update QorIQ TMU thermal bindings
thermal: mtk_thermal: Staticise a number of data variables
thermal: arm: dra752: Remove all TSHUT related definitions
thermal: arm: dra752: Remove TSHUT configuration
thermal: ti-soc-thermal: Remove CPU_THERMAL Dependency from TI_THERMAL
thermal: imx: Fix possible NULL dereference.
thermal: exynos: Remove parsing unused samsung,tmu_cal_mode property
thermal: zx2967: add thermal driver for ZTE's zx2967 family
thermal: use cpumask_var_t for on-stack cpu masks
dt: bindings: add documentation for zx2967 family thermal sensor
thermal/intel_powerclamp: Remove set-but-not-used variables
thermal: rcar_gen3_thermal: Add R-Car Gen3 thermal driver
thermal: rcar_gen3_thermal: Document the R-Car Gen3
thermal: convert devfreq_cooling to use an IDA
thermal: convert cpu_cooling to use an IDA
thermal: convert clock cooling to use an IDA
thermal core: convert ID allocation to IDA
Putting a bare cpumask structure on the stack produces a warning on
large SMP configurations:
drivers/thermal/cpu_cooling.c: In function 'cpufreq_state2power':
drivers/thermal/cpu_cooling.c:644:1: warning: the frame size of 1056 bytes is larger than 1024 bytes [-Wframe-larger-than=]
drivers/thermal/cpu_cooling.c: In function '__cpufreq_cooling_register':
drivers/thermal/cpu_cooling.c:898:1: warning: the frame size of 1104 bytes is larger than 1024 bytes [-Wframe-larger-than=]
The recommended workaround is to use cpumask_var_t, which behaves just like
a normal cpu mask in most cases, but turns into a dynamic allocation
when CONFIG_CPUMASK_OFFSTACK is set.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
This patch updates dev_pm_opp_find_freq_*() routines to get a reference
to the OPPs returned by them.
Also updates the users of dev_pm_opp_find_freq_*() routines to call
dev_pm_opp_put() after they are done using the OPPs.
As it is guaranteed the that OPPs wouldn't get freed while being used,
the RCU read side locking present with the users isn't required anymore.
Drop it as well.
This patch also updates all users of devfreq_recommended_opp() which was
returning an OPP received from the OPP core.
Note that some of the OPP core routines have gained
rcu_read_{lock|unlock}() calls, as those still use RCU specific APIs
within them.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Chanwoo Choi <cw00.choi@samsung.com> [Devfreq]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
thermal cpu cooling does not use the ability to look up pointers by ID,
so convert it from using an IDR to the more space-efficient IDA.
The cooling_cpufreq_lock was being used to protect cpufreq_dev_count as
well as the IDR. Rather than keep the mutex to protect a single integer,
I expanded the scope of cooling_list_lock to also cover cpufreq_dev_count.
We could also convert cpufreq_dev_count into an atomic.
Signed-off-by: Matthew Wilcox <mawilcox@microsoft.com>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
The last_load is updated not cpufreq_get_actual_power() function call
but cpufreq_get_requested_power() function call.
Signed-off-by: Inhyuk Kang <hugh.kang@lge.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Acked-by: Javi Merino <javi.merino@arm.com>
Acked-by: Javi Merino <javi.merino@arm.com>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Currently all CPU cooling devices share a
`struct thermal_cooling_device_ops` instance. The thermal core uses the
presence of functions in this struct to determine if a cooling device
has a power model (see cdev_is_power_actor). cpu_cooling.c adds the
power model functions to the shared struct when a device is registered
with a power model.
Therefore, if a CPU cooling device is registered using
[of_]cpufreq_power_cooling_register, _all_ devices will be determined to
have a power model, including any registered with
[of_]cpufreq_cooling_register. This can result in cpufreq_state2power
being called on a device where dyn_power_table is NULL.
With this commit, instead of having a shared thermal_cooling_device_ops
which is mutated, we have two versions: one with the power functions and
one without.
Signed-off-by: Brendan Jackman <brendan.jackman@arm.com>
Cc: Amit Daniel Kachhap <amit.kachhap@gmail.com>
Cc: Viresh Kumar <viresh.kumar@linaro.org>
Cc: Javi Merino <javi.merino@arm.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Acked-by: Javi Merino <javi.merino@arm.com>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Most of the callers of cpufreq_frequency_get_table() already have the
pointer to a valid 'policy' structure and they don't really need to go
through the per-cpu variable first and then a check to validate the
frequency, in order to find the freq-table for the policy.
Directly use the policy->freq_table field instead for them.
Only one user of that API is left after above changes, cpu_cooling.c and
it accesses the freq_table in a racy way as the policy can get freed in
between.
Fix it by using cpufreq_cpu_get() properly.
Since there are no more users of cpufreq_frequency_get_table() left, get
rid of it.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Acked-by: Javi Merino <javi.merino@arm.com> (cpu_cooling.c)
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The freq_table array is not populated before calling
thermal_of_cooling_register. The code which populates the freq table was
introduced in commit f6859014.
This should be done before registering new thermal cooling device.
The log shows effects of this wrong decision.
[ 2.172614] cpu cpu1: Failed to get voltage for frequency 1984518656000: -34
[ 2.220863] cpu cpu0: Failed to get voltage for frequency 1984524416000: -34
Cc: <stable@vger.kernel.org> # 3.19+
Fixes: f6859014c7 ("thermal: cpu_cooling: Store frequencies in descending order")
Signed-off-by: Lukasz Luba <lukasz.luba@arm.com>
Acked-by: Javi Merino <javi.merino@arm.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
In __cpufreq_cooling_register() we allocate the arrays for time_in_idle
and time_in_idle_timestamp to be as big as the number of cpus in this
cpufreq device. However, in get_load() we access this array using the
cpu number as index, which can result in an out of bound access.
Index time_in_idle{,_timestamp} using the index in the cpufreq_device's
allowed_cpus mask, as we do for the load_cpu array in
cpufreq_get_requested_power()
Reported-by: Nicolas Boichat <drinkcat@chromium.org>
Cc: Amit Daniel Kachhap <amit.kachhap@gmail.com>
Cc: Zhang Rui <rui.zhang@intel.com>
Cc: Eduardo Valentin <edubezval@gmail.com>
Tested-by: Nicolas Boichat <drinkcat@chromium.org>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Javi Merino <javi.merino@arm.com>
Signed-off-by: Eduardo Valentin <edubezval@gmail.com>
In the function cpufreq_get_requested_power, the memory allocated
for load_cpu is live within the function only. And after the
allocation it is immediately freed with devm_kfree. There is no
need to allocate memory for load_cpu with devm function so replace
devm_kcalloc with kcalloc and devm_kfree with kfree.
Signed-off-by: Vaishali Thakkar <vthakkar1994@gmail.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Acked-by: Javi Merino <javi.merino@arm.com>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
The power table is not being freed on error from cpufreq_cooling
register or when unregistering. Free it.
Fixes: c36cf07176 ("thermal: cpu_cooling: implement the power cooling device API")
Cc: Zhang Rui <rui.zhang@intel.com>
Cc: Eduardo Valentin <edubezval@gmail.com>
Signed-off-by: Javi Merino <javi.merino@arm.com>
Signed-off-by: Eduardo Valentin <edubezval@gmail.com>
