The scheduler code calling cpufreq_update_util() may run during CPU
offline on the target CPU after the IRQ work lists have been flushed
for it, so the target CPU should be prevented from running code that
may queue up an IRQ work item on it at that point.
Unfortunately, that may not be the case if dvfs_possible_from_any_cpu
is set for at least one cpufreq policy in the system, because that
allows the CPU going offline to run the utilization update callback
of the cpufreq governor on behalf of another (online) CPU in some
cases.
If that happens, the cpufreq governor callback may queue up an IRQ
work on the CPU running it, which is going offline, and the IRQ work
may not be flushed after that point. Moreover, that IRQ work cannot
be flushed until the "offlining" CPU goes back online, so if any
other CPU calls irq_work_sync() to wait for the completion of that
IRQ work, it will have to wait until the "offlining" CPU is back
online and that may not happen forever. In particular, a system-wide
deadlock may occur during CPU online as a result of that.
The failing scenario is as follows. CPU0 is the boot CPU, so it
creates a cpufreq policy and becomes the "leader" of it
(policy->cpu). It cannot go offline, because it is the boot CPU.
Next, other CPUs join the cpufreq policy as they go online and they
leave it when they go offline. The last CPU to go offline, say CPU3,
may queue up an IRQ work while running the governor callback on
behalf of CPU0 after leaving the cpufreq policy because of the
dvfs_possible_from_any_cpu effect described above. Then, CPU0 is
the only online CPU in the system and the stale IRQ work is still
queued on CPU3. When, say, CPU1 goes back online, it will run
irq_work_sync() to wait for that IRQ work to complete and so it
will wait for CPU3 to go back online (which may never happen even
in principle), but (worse yet) CPU0 is waiting for CPU1 at that
point too and a system-wide deadlock occurs.
To address this problem notice that CPUs which cannot run cpufreq
utilization update code for themselves (for example, because they
have left the cpufreq policies that they belonged to), should also
be prevented from running that code on behalf of the other CPUs that
belong to a cpufreq policy with dvfs_possible_from_any_cpu set and so
in that case the cpufreq_update_util_data pointer of the CPU running
the code must not be NULL as well as for the CPU which is the target
of the cpufreq utilization update in progress.
Accordingly, change cpufreq_this_cpu_can_update() into a regular
function in kernel/sched/cpufreq.c (instead of a static inline in a
header file) and make it check the cpufreq_update_util_data pointer
of the local CPU if dvfs_possible_from_any_cpu is set for the target
cpufreq policy.
Also update the schedutil governor to do the
cpufreq_this_cpu_can_update() check in the non-fast-switch
case too to avoid the stale IRQ work issues.
Fixes: 99d14d0e16 ("cpufreq: Process remote callbacks from any CPU if the platform permits")
Link: https://lore.kernel.org/linux-pm/20191121093557.bycvdo4xyinbc5cb@vireshk-i7/
Reported-by: Anson Huang <anson.huang@nxp.com>
Tested-by: Anson Huang <anson.huang@nxp.com>
Cc: 4.14+ <stable@vger.kernel.org> # 4.14+
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Tested-by: Peng Fan <peng.fan@nxp.com> (i.MX8QXP-MEK)
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.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>
No driver makes reference to these events now, remove them and the code
related to them.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
This effectively reverts some changes made by commit f9f41e3ef9
("cpufreq: Remove policy create/remove notifiers").
We have a new use case for policy create/remove notifiers (for
allocating/freeing QoS requests per policy), so add them back.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
[ rjw: Subject & changelog ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
* pm-cpufreq:
cpufreq: Make cpufreq_generic_init() return void
cpufreq: imx-cpufreq-dt: Add i.MX8MN support
cpufreq: Add QoS requests for userspace constraints
cpufreq: intel_pstate: Reuse refresh_frequency_limits()
cpufreq: Register notifiers with the PM QoS framework
PM / QoS: Add support for MIN/MAX frequency constraints
PM / QOS: Pass request type to dev_pm_qos_read_value()
PM / QOS: Rename __dev_pm_qos_read_value() and dev_pm_qos_raw_read_value()
PM / QOS: Pass request type to dev_pm_qos_{add|remove}_notifier()
It always returns 0 (success) and its return type should really be void.
Over that, many drivers have added error handling code based on its
return value, which is not required at all.
Change its return type to void and update all the callers.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
This implements QoS requests to manage userspace configuration of min
and max frequency.
Reviewed-by: Matthias Kaehlcke <mka@chromium.org>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Tested-by: syzbot <syzbot+de771ae9390dffed7266@syzkaller.appspotmail.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The implementation of intel_pstate_update_max_freq() is quite similar to
refresh_frequency_limits(), lets reuse it.
Finding minimum of policy->user_policy.max and policy->cpuinfo.max_freq
in intel_pstate_update_max_freq() is redundant as cpufreq_set_policy()
will call the ->verify() callback of intel-pstate driver, which will do
this comparison anyway and so dropping it from
intel_pstate_update_max_freq() doesn't harm.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Register notifiers for min/max frequency constraints with the PM QoS
framework. The constraints are also taken into consideration in
cpufreq_set_policy().
This also relocates cpufreq_policy_put_kobj() as it is required to be
called from cpufreq_policy_alloc() now.
refresh_frequency_limits() is updated to avoid calling
cpufreq_set_policy() for inactive policies and handle_update() is
updated to have proper locking in place.
No constraints are added until now though.
Reviewed-by: Matthias Kaehlcke <mka@chromium.org>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Tested-by: Pavel Machek <pavel@ucw.cz>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
* pm-cpufreq:
cpufreq: Avoid calling cpufreq_verify_current_freq() from handle_update()
cpufreq: Consolidate cpufreq_update_current_freq() and __cpufreq_get()
cpufreq: Don't skip frequency validation for has_target() drivers
cpufreq: Use has_target() instead of !setpolicy
cpufreq: Remove redundant !setpolicy check
cpufreq: Move the IS_ENABLED(CPU_THERMAL) macro into a stub
cpufreq: s5pv210: Don't flood kernel log after cpufreq change
cpufreq: pcc-cpufreq: Fail initialization if driver cannot be registered
cpufreq: add driver for Raspberry Pi
cpufreq: Switch imx7d to imx-cpufreq-dt for speed grading
cpufreq: imx-cpufreq-dt: Remove global platform match list
cpufreq: brcmstb-avs-cpufreq: Fix types for voltage/frequency
cpufreq: brcmstb-avs-cpufreq: Fix initial command check
cpufreq: armada-37xx: Remove set but not used variable 'freq'
cpufreq: imx-cpufreq-dt: Fix no OPPs available on unfused parts
dt-bindings: imx-cpufreq-dt: Document opp-supported-hw usage
cpufreq: Add imx-cpufreq-dt driver
cpufreq_online() and cpufreq_offline() [un]register the driver as
a cooling device. This is done if the driver is flagged as a cooling
device in addition with an IS_ENABLED() check to compile out the branching
code.
Group this test in a stub function added in the cpufreq header instead
of having the IS_ENABLED() in the code.
Suggested-by: Rafael J. Wysocki <rafael@kernel.org>
Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Based on 2 normalized pattern(s):
this program is free software you can redistribute it and or modify
it under the terms of the gnu general public license version 2 as
published by the free software foundation
this program is free software you can redistribute it and or modify
it under the terms of the gnu general public license version 2 as
published by the free software foundation #
extracted by the scancode license scanner the SPDX license identifier
GPL-2.0-only
has been chosen to replace the boilerplate/reference in 4122 file(s).
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Enrico Weigelt <info@metux.net>
Reviewed-by: Kate Stewart <kstewart@linuxfoundation.org>
Reviewed-by: Allison Randal <allison@lohutok.net>
Cc: linux-spdx@vger.kernel.org
Link: https://lkml.kernel.org/r/20190604081206.933168790@linutronix.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Currently, the notifiers are called once for each CPU of the policy->cpus
cpumask. It would be more optimal if the notifier can be called only
once and all the relevant information be provided to it. Out of the 23
drivers that register for the transition notifiers today, only 4 of them
do per-cpu updates and the callback for the rest can be called only once
for the policy without any impact.
This would also avoid multiple function calls to the notifier callbacks
and reduce multiple iterations of notifier core's code (which does
locking as well).
This patch adds pointer to the cpufreq policy to the struct
cpufreq_freqs, so the notifier callback has all the information
available to it with a single call. The five drivers which perform
per-cpu updates are updated to use the cpufreq policy. The freqs->cpu
field is redundant now and is removed.
Acked-by: David S. Miller <davem@davemloft.net> (sparc)
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
While the cpuinfo.max_freq value doesn't really matter for
intel_pstate in the active mode, in the passive mode it is used by
governors as the maximum physical frequency of the CPU and the
results of governor computations generally depend on it. Also it
is made available to user space via sysfs and it should match the
current HW configuration.
For this reason, make intel_pstate update cpuinfo.max_freq for all
CPUs if it detects a global change of turbo frequency settings from
"disable" to "enable" or the other way associated with a _PPC change
notification from the platform firmware.
Note that policy_is_inactive(), cpufreq_cpu_acquire(),
cpufreq_cpu_release(), and cpufreq_set_policy() need to be made
available to it for this purpose.
Link: https://bugzilla.kernel.org/show_bug.cgi?id=200759
Reported-by: Gabriele Mazzotta <gabriele.mzt@gmail.com>
Tested-by: Gabriele Mazzotta <gabriele.mzt@gmail.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
In some cases, the platform firmware disables or enables turbo
frequencies for all CPUs globally before triggering a _PPC change
notification for one of them. Obviously, that global change affects
all CPUs, not just the notified one, and it needs to be acted upon by
cpufreq.
The intel_pstate driver is able to detect such global changes of
the settings, but it also needs to update policy limits for all
CPUs if that happens, in particular if turbo frequencies are
enabled globally - to allow them to be used.
For this reason, introduce a new cpufreq driver callback to be
invoked on _PPC notifications, if present, instead of simply
calling cpufreq_update_policy() for the notified CPU and make
intel_pstate use it to trigger policy updates for all CPUs
in the system if global settings change.
Link: https://bugzilla.kernel.org/show_bug.cgi?id=200759
Reported-by: Gabriele Mazzotta <gabriele.mzt@gmail.com>
Tested-by: Gabriele Mazzotta <gabriele.mzt@gmail.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
The cpufreq core doesn't remove the cpufreq policy anymore on CPU
offline operation, rather that happens when the CPU device gets
unregistered from the kernel. This allows faster recovery when the CPU
comes back online. This is also very useful during system wide
suspend/resume where we offline all non-boot CPUs during suspend and
then bring them back on resume.
This commit takes the same idea a step ahead to allow drivers to do
light weight tear-down and bring-up during CPU offline and online
operations.
A new set of callbacks is introduced, online/offline(). online() gets
called when the first CPU of an inactive policy is brought up and
offline() gets called when all the CPUs of a policy are offlined.
The existing init/exit() callback get called on policy
creation/destruction. They also get called instead of online/offline()
callbacks if the online/offline() callbacks aren't provided.
This also moves around some code to get executed only for the new-policy
case going forward.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
All cpufreq drivers do similar things to register as a cooling device.
Provide a cpufreq driver flag so drivers can just ask the cpufreq core
to register the cooling device on their behalf. This allows us to get
rid of duplicated code in the drivers.
In order to allow this, we add a struct thermal_cooling_device pointer
to struct cpufreq_policy so that drivers don't need to store it in a
private data structure.
Suggested-by: Stephen Boyd <swboyd@chromium.org>
Suggested-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Amit Kucheria <amit.kucheria@linaro.org>
Reviewed-by: Matthias Kaehlcke <mka@chromium.org>
Tested-by: Matthias Kaehlcke <mka@chromium.org>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The cpufreq_global_kobject is created using kobject_create_and_add()
helper, which assigns the kobj_type as dynamic_kobj_ktype and show/store
routines are set to kobj_attr_show() and kobj_attr_store().
These routines pass struct kobj_attribute as an argument to the
show/store callbacks. But all the cpufreq files created using the
cpufreq_global_kobject expect the argument to be of type struct
attribute. Things work fine currently as no one accesses the "attr"
argument. We may not see issues even if the argument is used, as struct
kobj_attribute has struct attribute as its first element and so they
will both get same address.
But this is logically incorrect and we should rather use struct
kobj_attribute instead of struct global_attr in the cpufreq core and
drivers and the show/store callbacks should take struct kobj_attribute
as argument instead.
This bug is caught using CFI CLANG builds in android kernel which
catches mismatch in function prototypes for such callbacks.
Reported-by: Donghee Han <dh.han@samsung.com>
Reported-by: Sangkyu Kim <skwith.kim@samsung.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Minor clean-up to use BIT() and keep checkpatch happy. Clean up the
comment formatting while we're at it to make it easier to read.
Signed-off-by: Amit Kucheria <amit.kucheria@linaro.org>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
This routine checks if the CPU running this code belongs to the policy
of the target CPU or if not, can it do remote DVFS for it remotely. But
the current name of it implies as if it is only about doing remote
updates.
Rename it to make it more relevant.
Suggested-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
This isn't used anymore. Remove the helper and update documentation
accordingly.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael 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: Viresh Kumar <viresh.kumar@linaro.org>
[ rjw: Subject/changelog ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Pointer subtraction is slow and tedious. Therefore, replace all instances
where cpufreq_for_each_{valid_,}entry loops contained such substractions
with an iteration macro providing an index to the frequency_table entry.
Suggested-by: Al Viro <viro@ZenIV.linux.org.uk>
Link: http://lkml.kernel.org/r/20180120020237.GM13338@ZenIV.linux.org.uk
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Dominik Brodowski <linux@dominikbrodowski.net>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
After commit 890da9cf09 (Revert "x86: do not use cpufreq_quick_get()
for /proc/cpuinfo "cpu MHz"") the "cpu MHz" number in /proc/cpuinfo
on x86 can be either the nominal CPU frequency (which is constant)
or the frequency most recently requested by a scaling governor in
cpufreq, depending on the cpufreq configuration. That is somewhat
inconsistent and is different from what it was before 4.13, so in
order to restore the previous behavior, make it report the current
CPU frequency like the scaling_cur_freq sysfs file in cpufreq.
To that end, modify the /proc/cpuinfo implementation on x86 to use
aperfmperf_snapshot_khz() to snapshot the APERF and MPERF feedback
registers, if available, and use their values to compute the CPU
frequency to be reported as "cpu MHz".
However, do that carefully enough to avoid accumulating delays that
lead to unacceptable access times for /proc/cpuinfo on systems with
many CPUs. Run aperfmperf_snapshot_khz() once on all CPUs
asynchronously at the /proc/cpuinfo open time, add a single delay
upfront (if necessary) at that point and simply compute the current
frequency while running show_cpuinfo() for each individual CPU.
Also, to avoid slowing down /proc/cpuinfo accesses too much, reduce
the default delay between consecutive APERF and MPERF reads to 10 ms,
which should be sufficient to get large enough numbers for the
frequency computation in all cases.
Fixes: 890da9cf09 (Revert "x86: do not use cpufreq_quick_get() for /proc/cpuinfo "cpu MHz"")
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Ingo Molnar <mingo@kernel.org>
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: Dietmar Eggemann <dietmar.eggemann@arm.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
* pm-cpufreq-sched:
cpufreq: schedutil: Always process remote callback with slow switching
cpufreq: schedutil: Don't restrict kthread to related_cpus unnecessarily
cpufreq: Return 0 from ->fast_switch() on errors
cpufreq: Simplify cpufreq_can_do_remote_dvfs()
cpufreq: Process remote callbacks from any CPU if the platform permits
sched: cpufreq: Allow remote cpufreq callbacks
cpufreq: schedutil: Use unsigned int for iowait boost
cpufreq: schedutil: Make iowait boost more energy efficient
The if () in cpufreq_can_do_remote_dvfs() is superfluous, so drop
it and simply return the value of the expression under it.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
On many platforms, CPUs can do DVFS across cpufreq policies. i.e CPU
from policy-A can change frequency of CPUs belonging to policy-B.
This is quite common in case of ARM platforms where we don't
configure any per-cpu register.
Add a flag to identify such platforms and update
cpufreq_can_do_remote_dvfs() to allow remote callbacks if this flag is
set.
Also enable the flag for cpufreq-dt driver which is used only on ARM
platforms currently.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Acked-by: Saravana Kannan <skannan@codeaurora.org>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
With Android UI and benchmarks the latency of cpufreq response to
certain scheduling events can become very critical. Currently, callbacks
into cpufreq governors are only made from the scheduler if the target
CPU of the event is the same as the current CPU. This means there are
certain situations where a target CPU may not run the cpufreq governor
for some time.
One testcase to show this behavior is where a task starts running on
CPU0, then a new task is also spawned on CPU0 by a task on CPU1. If the
system is configured such that the new tasks should receive maximum
demand initially, this should result in CPU0 increasing frequency
immediately. But because of the above mentioned limitation though, this
does not occur.
This patch updates the scheduler core to call the cpufreq callbacks for
remote CPUs as well.
The schedutil, ondemand and conservative governors are updated to
process cpufreq utilization update hooks called for remote CPUs where
the remote CPU is managed by the cpufreq policy of the local CPU.
The intel_pstate driver is updated to always reject remote callbacks.
This is tested with couple of usecases (Android: hackbench, recentfling,
galleryfling, vellamo, Ubuntu: hackbench) on ARM hikey board (64 bit
octa-core, single policy). Only galleryfling showed minor improvements,
while others didn't had much deviation.
The reason being that this patch only targets a corner case, where
following are required to be true to improve performance and that
doesn't happen too often with these tests:
- Task is migrated to another CPU.
- The task has high demand, and should take the target CPU to higher
OPPs.
- And the target CPU doesn't call into the cpufreq governor until the
next tick.
Based on initial work from Steve Muckle.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Acked-by: Saravana Kannan <skannan@codeaurora.org>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Rafael 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: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Dominik Brodowski <linux@dominikbrodowski.net>
Signed-off-by: Rafael 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: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael 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: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The cpufreq core and governors aren't supposed to set a limit on how
fast we want to try changing the frequency. This is currently done for
the legacy governors with help of min_sampling_rate.
At worst, we may end up setting the sampling rate to a value lower than
the rate at which frequency can be changed and then one of the CPUs in
the policy will be only changing frequency for ever.
But that is something for the user to decide and there is no need to
have special handling for such cases in the core. Leave it for the user
to figure out.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Pull thermal management updates from Zhang Rui:
- Improve thermal cpu_cooling interaction with cpufreq core.
The cpu_cooling driver is designed to use CPU frequency scaling to
avoid high thermal states for a platform. But it wasn't glued really
well with cpufreq core.
For example clipped-cpus is copied from the policy structure and its
much better to use the policy->cpus (or related_cpus) fields directly
as they may have got updated. Not that things were broken before this
series, but they can be optimized a bit more.
This series tries to improve interactions between cpufreq core and
cpu_cooling driver and does some fixes/cleanups to the cpu_cooling
driver. (Viresh Kumar)
- A couple of fixes and cleanups in thermal core and imx, hisilicon,
bcm_2835, int340x thermal drivers. (Arvind Yadav, Dan Carpenter,
Sumeet Pawnikar, Srinivas Pandruvada, Willy WOLFF)
* 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/rzhang/linux: (24 commits)
thermal: bcm2835: fix an error code in probe()
thermal: hisilicon: Handle return value of clk_prepare_enable
thermal: imx: Handle return value of clk_prepare_enable
thermal: int340x: check for sensor when PTYP is missing
Thermal/int340x: Fix few typos and kernel-doc style
thermal: fix source code documentation for parameters
thermal: cpu_cooling: Replace kmalloc with kmalloc_array
thermal: cpu_cooling: Rearrange struct cpufreq_cooling_device
thermal: cpu_cooling: 'freq' can't be zero in cpufreq_state2power()
thermal: cpu_cooling: don't store cpu_dev in cpufreq_cdev
thermal: cpu_cooling: get_level() can't fail
thermal: cpu_cooling: create structure for idle time stats
thermal: cpu_cooling: merge frequency and power tables
thermal: cpu_cooling: get rid of 'allowed_cpus'
thermal: cpu_cooling: OPPs are registered for all CPUs
thermal: cpu_cooling: store cpufreq policy
cpufreq: create cpufreq_table_count_valid_entries()
thermal: cpu_cooling: use cpufreq_policy to register cooling device
thermal: cpu_cooling: get rid of a variable in cpufreq_set_cur_state()
thermal: cpu_cooling: remove cpufreq_cooling_get_level()
...
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: Len Brown <len.brown@intel.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.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>
Make the schedutil governor take the initial (default) value of the
rate_limit_us sysfs attribute from the (new) transition_delay_us
policy parameter (to be set by the scaling driver).
That will allow scaling drivers to make schedutil use smaller default
values of rate_limit_us and reduce the default average time interval
between consecutive frequency changes.
Make intel_pstate set transition_delay_us to 500.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Those were added by:
commit fcd7af917a ("cpufreq: stats: handle cpufreq_unregister_driver()
and suspend/resume properly")
but aren't used anymore since:
commit 1aefc75b24 ("cpufreq: stats: Make the stats code non-modular").
Remove them. Also remove the redundant parameter to the respective
routines.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The return value of cpufreq_update_policy() is never used, so make
it void.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Allow CPUfreq statistics to be cleared by writing anything to
/sys/.../cpufreq/stats/reset.
Signed-off-by: Markus Mayer <mmayer@broadcom.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Skip invalid entries when searching the frequency. This fixes cpufreq
at least on loongson2 MIPS board.
Fixes: da0c6dc00c (cpufreq: Handle sorted frequency tables more efficiently)
Signed-off-by: Aaro Koskinen <aaro.koskinen@iki.fi>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Cc: 4.8+ <stable@vger.kernel.org> # 4.8+
Signed-off-by: Rafael 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: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Steve Muckle <smuckle@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
cpufreq drivers aren't required to provide a sorted frequency table
today, and even the ones which provide a sorted table aren't handled
efficiently by cpufreq core.
This patch adds infrastructure to verify if the freq-table provided by
the drivers is sorted or not, and use efficient helpers if they are
sorted.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
This routine can't fail unless the frequency table is invalid and
doesn't contain any valid entries.
Make it return the index and WARN() in case it is used for an invalid
table.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The policy already has this pointer set, use it instead.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@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 modularity of cpufreq_stats is quite problematic.
First off, the usage of policy notifiers for the initialization
and cleanup in the cpufreq_stats module is inherently racy with
respect to CPU offline/online and the initialization and cleanup
of the cpufreq driver.
Second, fast frequency switching (used by the schedutil governor)
cannot be enabled if any transition notifiers are registered, so
if the cpufreq_stats module (that registers a transition notifier
for updating transition statistics) is loaded, the schedutil governor
cannot use fast frequency switching.
On the other hand, allowing cpufreq_stats to be built as a module
doesn't really add much value. Arguably, there's not much reason
for that code to be modular at all.
For the above reasons, make the cpufreq stats code non-modular,
modify the core to invoke functions provided by that code directly
and drop the notifiers from it.
Make the stats sysfs attributes appear empty if fast frequency
switching is enabled as the statistics will not be updated in that
case anyway (and returning -EBUSY from those attributes breaks
powertop).
While at it, clean up Kconfig help for the CPU_FREQ_STAT and
CPU_FREQ_STAT_DETAILS options.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
The 'initialized' field in struct cpufreq_governor is only used by
the conservative governor (as a usage counter) and the way that
happens is far from straightforward and arguably incorrect.
Namely, the value of 'initialized' is checked by
cpufreq_dbs_governor_init() and cpufreq_dbs_governor_exit() and
the results of those checks are passed (as the second argument) to
the ->init() and ->exit() callbacks in struct dbs_governor. Those
callbacks are only implemented by the ondemand and conservative
governors and ondemand doesn't use their second argument at all.
In turn, the conservative governor uses it to decide whether or not
to either register or unregister a transition notifier.
That whole mechanism is not only unnecessarily convoluted, but also
racy, because the 'initialized' field of struct cpufreq_governor is
updated in cpufreq_init_governor() and cpufreq_exit_governor() under
policy->rwsem which doesn't help if one of these functions is run
twice in parallel for different policies (which isn't impossible in
principle), for example.
Instead of it, add a proper usage counter to the conservative
governor and update it from cs_init() and cs_exit() which is
guaranteed to be non-racy, as those functions are only called
under gov_dbs_data_mutex which is global.
With that in place, drop the 'initialized' field from struct
cpufreq_governor as it is not used any more.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Create a new helper to avoid code duplication across governors.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The design of the cpufreq governor API is not very straightforward,
as struct cpufreq_governor provides only one callback to be invoked
from different code paths for different purposes. The purpose it is
invoked for is determined by its second "event" argument, causing it
to act as a "callback multiplexer" of sorts.
Unfortunately, that leads to extra complexity in governors, some of
which implement the ->governor() callback as a switch statement
that simply checks the event argument and invokes a separate function
to handle that specific event.
That extra complexity can be eliminated by replacing the all-purpose
->governor() callback with a family of callbacks to carry out specific
governor operations: initialization and exit, start and stop and policy
limits updates. That also turns out to reduce the code size too, so
do it.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Due to differences in the cpufreq core's handling of runtime CPU
offline and nonboot CPUs disabling during system suspend-to-RAM,
fast frequency switching gets disabled after a suspend-to-RAM and
resume cycle on all of the nonboot CPUs.
To prevent that from happening, move the invocation of
cpufreq_disable_fast_switch() from cpufreq_exit_governor() to
sugov_exit(), as the schedutil governor is the only user of fast
frequency switching today anyway.
That simply prevents cpufreq_disable_fast_switch() from being called
without invoking the ->governor callback for the CPUFREQ_GOV_POLICY_EXIT
event (which happens during system suspend now).
Fixes: b7898fda5b (cpufreq: Support for fast frequency switching)
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Modify the ACPI cpufreq driver to provide a method for switching
CPU frequencies from interrupt context and update the cpufreq core
to support that method if available.
Introduce a new cpufreq driver callback, ->fast_switch, to be
invoked for frequency switching from interrupt context by (future)
governors supporting that feature via (new) helper function
cpufreq_driver_fast_switch().
Add two new policy flags, fast_switch_possible, to be set by the
cpufreq driver if fast frequency switching can be used for the
given policy and fast_switch_enabled, to be set by the governor
if it is going to use fast frequency switching for the given
policy. Also add a helper for setting the latter.
Since fast frequency switching is inherently incompatible with
cpufreq transition notifiers, make it possible to set the
fast_switch_enabled only if there are no transition notifiers
already registered and make the registration of new transition
notifiers fail if fast_switch_enabled is set for at least one
policy.
Implement the ->fast_switch callback in the ACPI cpufreq driver
and make it set fast_switch_possible during policy initialization
as appropriate.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Move definitions of symbols related to transition latency and
sampling rate to include/linux/cpufreq.h so they can be used by
(future) goverernors located outside of drivers/cpufreq/.
No functional changes.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Move definitions and function headers related to struct gov_attr_set
to include/linux/cpufreq.h so they can be used by (future) goverernors
located outside of drivers/cpufreq/.
No functional changes.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Create cpufreq.c under kernel/sched/ and move the cpufreq code
related to the scheduler to that file and to sched.h.
Redefine cpufreq_update_util() as a static inline function to avoid
function calls at its call sites in the scheduler code (as suggested
by Peter Zijlstra).
Also move the definition of struct update_util_data and declaration
of cpufreq_set_update_util_data() from include/linux/cpufreq.h to
include/linux/sched.h.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
The entire sequence of events (like INIT/START or STOP/EXIT) for which
cpufreq_governor() is called, is guaranteed to be protected by
policy->rwsem now.
The additional checks that were added earlier (as we were forced to drop
policy->rwsem before calling cpufreq_governor() for EXIT event), aren't
required anymore.
Over that, they weren't sufficient really. They just take care of
START/STOP events, but not INIT/EXIT and the state machine was never
maintained properly by them.
Kill the unnecessary checks and policy->governor_enabled field.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Earlier, when the struct freq-attr was used to represent governor
attributes, the standard cpufreq show/store sysfs attribute callbacks
were applied to the governor tunable attributes and they always acquire
the policy->rwsem lock before carrying out the operation. That could
have resulted in an ABBA deadlock if governor tunable attributes are
removed under policy->rwsem while one of them is being accessed
concurrently (if sysfs attributes removal wins the race, it will wait
for the access to complete with policy->rwsem held while the attribute
callback will block on policy->rwsem indefinitely).
We attempted to address this issue by dropping policy->rwsem around
governor tunable attributes removal (that is, around invocations of the
->governor callback with the event arg equal to CPUFREQ_GOV_POLICY_EXIT)
in cpufreq_set_policy(), but that opened up race conditions that had not
been possible with policy->rwsem held all the time.
The previous commit, "cpufreq: governor: New sysfs show/store callbacks
for governor tunables", fixed the original ABBA deadlock by adding new
governor specific show/store callbacks.
We don't have to drop rwsem around invocations of governor event
CPUFREQ_GOV_POLICY_EXIT anymore, and original fix can be reverted now.
Fixes: 955ef48335 (cpufreq: Drop rwsem lock around CPUFREQ_GOV_POLICY_EXIT)
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reported-by: Juri Lelli <juri.lelli@arm.com>
Tested-by: Juri Lelli <juri.lelli@arm.com>
Tested-by: Shilpasri G Bhat <shilpa.bhat@linux.vnet.ibm.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Introduce a mechanism by which parts of the cpufreq subsystem
("setpolicy" drivers or the core) can register callbacks to be
executed from cpufreq_update_util() which is invoked by the
scheduler's update_load_avg() on CPU utilization changes.
This allows the "setpolicy" drivers to dispense with their timers
and do all of the computations they need and frequency/voltage
adjustments in the update_load_avg() code path, among other things.
The update_load_avg() changes were suggested by Peter Zijlstra.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Ingo Molnar <mingo@kernel.org>
That macro uses an internal static inline function that is first
totally unnecessary and second hard to read, so simplify it and
get rid of that monster.
No functional changes.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
The preprocessor magic used for setting the default cpufreq governor
(and for using the performance governor as a fallback one for that
matter) is really nasty, so replace it with __weak functions and
overrides.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Saravana Kannan <skannan@codeaurora.org>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Notice that the boost_supported field in struct cpufreq_driver is
redundant, because the driver's ->set_boost callback may be left
unset if "boost" is not supported. Moreover, the only driver
populating the ->set_boost callback is acpi_cpufreq, so make it
avoid populating that callback if "boost" is not supported, rework
the core to check ->set_boost instead of boost_supported to
verify "boost" support and drop boost_supported which isn't
used any more.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
cpufreq_boost_supported() is not used outside of cpufreq.c, so make
it static.
While at it, refactor it as a one-liner (which it really is).
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
For cpufreq drivers which use setpolicy interface, after offline->online
the policy is set to default. This can be reproduced by setting the
default policy of intel_pstate or longrun to ondemand and then change to
"performance". After offline and online, the setpolicy will be called with
the policy=ondemand.
For drivers using governors this condition is handled by storing
last_governor, during offline and restoring during online. The same should
be done for drivers using setpolicy interface. Storing last_policy during
offline and restoring during online.
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The cpufreq sysfs interface had been a bit inconsistent as one of the
CPUs for a policy had a real directory within its sysfs 'cpuX' directory
and all other CPUs had links to it. That also made the code a bit
complex as we need to take care of moving the sysfs directory if the CPU
containing the real directory is getting physically hot-unplugged.
Solve this by creating 'policyX' directories (per-policy) in
/sys/devices/system/cpu/cpufreq/ directory, where X is the CPU for which
the policy was first created.
This also removes the need of keeping kobj_cpu and we can remove it now.
Suggested-by: Saravana Kannan <skannan@codeaurora.org>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Saravana Kannan <skannan@codeaurora.org>
Acked-by: is more of a general agreement from the person that he is
Reviewed-by: is a more strict tag and implies that the reviewer has
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
They don't do anything special now, remove the unnecessary wrapper.
Reviewed-by: Saravana Kannan <skannan@codeaurora.org>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Later patches will need to create policy specific directories in
/sys/devices/system/cpu/cpufreq/ directory and so the cpufreq directory
wouldn't be ever empty.
And so no fun creating/destroying it on need basis anymore. Create it
once on system boot.
Reviewed-by: Saravana Kannan <skannan@codeaurora.org>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
cpufreq_cpu_get() called by get_cur_freq_on_cpu() is overkill,
because the ->get() callback is always invoked in a context in
which all of the conditions checked by cpufreq_cpu_get() are
guaranteed to be satisfied.
Use cpufreq_cpu_get_raw() instead of it and drop the
corresponding cpufreq_cpu_put() from get_cur_freq_on_cpu().
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
* pm-opp:
PM / OPP: Drop unlikely before IS_ERR(_OR_NULL)
PM / OPP: Fix static checker warning (broken 64bit big endian systems)
PM / OPP: Free resources and properly return error on failure
cpufreq-dt: make scaling_boost_freqs sysfs attr available when boost is enabled
cpufreq: dt: Add support for turbo/boost mode
cpufreq: dt: Add support for operating-points-v2 bindings
cpufreq: Allow drivers to enable boost support after registering driver
cpufreq: Update boost flag while initializing freq table from OPPs
PM / OPP: add dev_pm_opp_is_turbo() helper
PM / OPP: Add helpers for initializing CPU OPPs
PM / OPP: Add support for opp-suspend
PM / OPP: Add OPP sharing information to OPP library
PM / OPP: Add clock-latency-ns support
PM / OPP: Add support to parse "operating-points-v2" bindings
PM / OPP: Break _opp_add_dynamic() into smaller functions
PM / OPP: Allocate dev_opp from _add_device_opp()
PM / OPP: Create _remove_device_opp() for freeing dev_opp
PM / OPP: Relocate few routines
PM / OPP: Create a directory for opp bindings
PM / OPP: Update bindings to make opp-hz a 64 bit value
Its all about caching min/max freq requested by userspace, and
the name 'cpufreq_real_policy' doesn't fit that well. Rename it to
cpufreq_user_policy.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Its always same as policy->policy, and there is no need to keep another
copy of it. Remove it.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Its always same as policy->governor, and there is no need to keep
another copy of it. Remove it.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
What's being done from CPUFREQ_INCOMPATIBLE, can also be done with
CPUFREQ_ADJUST. There is nothing special with CPUFREQ_INCOMPATIBLE
notifier.
Kill CPUFREQ_INCOMPATIBLE and fix its usage sites.
This also updates the numbering of notifier events to remove holes.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Make scaling_boost_freqs sysfs attribute is available when
cpufreq-dt driver is used and boost support is enabled.
Suggested-by: Viresh Kumar <viresh.kumar@linaro.org>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
In some cases it wouldn't be known at time of driver registration, if
the driver needs to support boost frequencies.
For example, while getting boost information from DT with opp-v2
bindings, we need to parse the bindings for all the CPUs to know if
turbo/boost OPPs are supported or not.
One way out to do that efficiently is to delay supporting boost mode
(i.e. creating /sys/devices/system/cpu/cpufreq/boost file), until the
time OPP bindings are parsed.
At that point, the driver can enable boost support. This can be done at
->init(), where the frequency table is created.
To do that, the driver requires few APIs from cpufreq core that let him
do this. This patch provides these APIs.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Stephen Boyd <sboyd@codeaurora.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
After commit 87549141d5 (cpufreq: Stop migrating sysfs files on
hotplug) there is a problem with CPUs that share cpufreq policy
objects with other CPUs and are initially offline.
Say CPU1 shares a policy with CPU0 which is online and is registered
first. As part of the registration process, cpufreq_add_dev() is
called for it. It creates the policy object and a symbolic link
to it from the CPU1's sysfs directory. If CPU1 is registered
subsequently and it is offline at that time, cpufreq_add_dev() will
attempt to create a symbolic link to the policy object for it, but
that link is present already, so a warning about that will be
triggered.
To avoid that warning, make cpufreq use an additional CPU mask
containing related CPUs that are actually present for each policy
object. That mask is initialized when the policy object is populated
after its creation (for the first online CPU using it) and it includes
CPUs from the "policy CPUs" mask returned by the cpufreq driver's
->init() callback that are physically present at that time. Symbolic
links to the policy are created only for the CPUs in that mask.
If cpufreq_add_dev() is invoked for an offline CPU, it checks the
new mask and only creates the symlink if the CPU was not in it (the
CPU is added to the mask at the same time).
In turn, cpufreq_remove_dev() drops the given CPU from the new mask,
removes its symlink to the policy object and returns, unless it is
the CPU owning the policy object. In that case, the policy object
is moved to a new CPU's sysfs directory or deleted if the CPU being
removed was the last user of the policy.
While at it, notice that cpufreq_remove_dev() can't fail, because
its return value is ignored, so make it ignore return values from
__cpufreq_remove_dev_prepare() and __cpufreq_remove_dev_finish()
and prevent these functions from aborting on errors returned by
__cpufreq_governor(). Also drop the now unused sif argument from
them.
Fixes: 87549141d5 (cpufreq: Stop migrating sysfs files on hotplug)
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reported-and-tested-by: Russell King <linux@arm.linux.org.uk>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
In order to prepare for the next few commits, that will stop migrating
sysfs files on cpu hotplug, this patch starts managing sysfs-cpu
separately.
The behavior is still the same as we are still migrating sysfs files on
hotplug, later commits would change that.
Signed-off-by: Saravana Kannan <skannan@codeaurora.org>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
History of which governor was used last is common to all CPUs within a
policy and maintaining it per-cpu isn't the best approach for sure.
Apart from wasting memory, this also increases the complexity of
managing this data structure as it has to be updated for all CPUs.
To make that somewhat simpler, lets store this information in a new
field 'last_governor' in struct cpufreq_policy and update it on removal
of last cpu of a policy.
As a side-effect it also solves an old problem, consider a system with
two clusters 0 & 1. And there is one policy per cluster.
Cluster 0: CPU0 and 1.
Cluster 1: CPU2 and 3.
- CPU2 is first brought online, and governor is set to performance
(default as cpufreq_cpu_governor wasn't set).
- Governor is changed to ondemand.
- CPU2 is taken offline and cpufreq_cpu_governor is updated for CPU2.
- CPU3 is brought online.
- Because cpufreq_cpu_governor wasn't set for CPU3, the default governor
performance is picked for CPU3.
This patch fixes the bug as we now have a single variable to update for
policy.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
CPUFREQ_UPDATE_POLICY_CPU notifications were used only from cpufreq-stats which
doesn't use it anymore. Remove them.
This also decrements values of other notification macros defined after
CPUFREQ_UPDATE_POLICY_CPU by 1 to remove gaps. Hopefully all users are using
macro's instead of direct numbers and so they wouldn't break as macro values are
changed now.
Reviewed-by: Prarit Bhargava <prarit@redhat.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
'last_cpu' was used only from cpufreq-stats and isn't used anymore. Get rid of
it.
Reviewed-by: Prarit Bhargava <prarit@redhat.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
All CPUs sharing a cpufreq policy share stats too. For this reason,
add a stats pointer to struct cpufreq_policy and drop per-CPU variable
cpufreq_stats_table used for accessing cpufreq stats so as to reduce
code complexity.
Reviewed-by: Prarit Bhargava <prarit@redhat.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Currently there is no callback for cpufreq drivers which is called once the
policy is ready to be used. There are some requirements where such a callback is
required.
One of them is registering a cooling device with the help of
of_cpufreq_cooling_register(). This routine tries to get 'struct cpufreq_policy'
for CPUs which isn't yet initialed at the time ->init() is called and so we face
issues while registering the cooling device.
Because we can't register cooling device from ->init(), we need a callback that
is called after the policy is ready to be used and hence we introduce ->ready()
callback.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Eduardo Valentin <edubezval@gmail.com>
Tested-by: Eduardo Valentin <edubezval@gmail.com>
Reviewed-by: Lukasz Majewski <l.majewski@samsung.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Adding any new callback to 'struct cpufreq_driver' gives following checkpatch
warning:
WARNING: Unnecessary space before function pointer arguments
+ void (*ready) (struct cpufreq_policy *policy);
This is because we have been using a tab spacing between function pointer name
and its arguments and the new one tried to follow that.
Though we normally don't try to fix every checkpatch warning, specially around
formatting issues as that creates unnecessary noise over lists. But I thought we
better fix this so that new additions don't generate these warnings plus it
looks far better/symmetric now.
So, remove these tab spacing issues in 'struct cpufreq_driver' only + fix
alignment of all members.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Eduardo Valentin <edubezval@gmail.com>
Tested-by: Eduardo Valentin <edubezval@gmail.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
This commit extends the cpufreq_driver structure with an additional
'void *driver_data' field that can be filled by the ->probe() function
of a cpufreq driver to pass additional custom information to the
driver itself.
A new function called cpufreq_get_driver_data() is added to allow a
cpufreq driver to retrieve those driver data, since they are typically
needed from a cpufreq_policy->init() callback, which does not have
access to the cpufreq_driver structure. This function call is similar
to the existing cpufreq_get_current_driver() function call.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Drivers supporting multiple clusters or multiple 'struct cpufreq_policy'
instances may need to keep per-policy data. If the core doesn't provide support
for that, they might do it in the most unoptimized way: 'per-cpu' data.
This patch adds another field in struct cpufreq_policy: 'driver_data'. It isn't
accessed by core and is for driver's internal use only.
Tested-by: Stephen Boyd <sboyd@codeaurora.org>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Introduce CPUFREQ_RELATION_C for frequency selection.
It selects the frequency with the minimum euclidean distance to target.
In case of equal distance between 2 frequencies, it will select the
greater frequency.
Signed-off-by: Stratos Karafotis <stratosk@semaphore.gr>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Commit 5eeaf1f189 (cpufreq: Fix build error on some platforms that
use cpufreq_for_each_*) moved function cpufreq_next_valid() to a public
header. Warnings are now generated when objects including that header
are built with -Wsign-compare (as an out-of-tree module might be):
.../include/linux/cpufreq.h: In function ‘cpufreq_next_valid’:
.../include/linux/cpufreq.h:519:27: warning: comparison between signed
and unsigned integer expressions [-Wsign-compare]
while ((*pos)->frequency != CPUFREQ_TABLE_END)
^
.../include/linux/cpufreq.h:520:25: warning: comparison between signed
and unsigned integer expressions [-Wsign-compare]
if ((*pos)->frequency != CPUFREQ_ENTRY_INVALID)
^
Constants CPUFREQ_ENTRY_INVALID and CPUFREQ_TABLE_END are signed, but
are used with unsigned member 'frequency' of cpufreq_frequency_table.
Update the macro definitions to be explicitly unsigned to match their
use.
This also corrects potentially wrong behavior of clk_rate_table_iter()
if unsigned long is wider than usigned int.
Fixes: 5eeaf1f189 (cpufreq: Fix build error on some platforms that use cpufreq_for_each_*)
Signed-off-by: Brian W Hart <hartb@linux.vnet.ibm.com>
Reviewed-by: Simon Horman <horms+renesas@verge.net.au>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Douglas Anderson, recently pointed out an interesting problem due to which
udelay() was expiring earlier than it should.
While transitioning between frequencies few platforms may temporarily switch to
a stable frequency, waiting for the main PLL to stabilize.
For example: When we transition between very low frequencies on exynos, like
between 200MHz and 300MHz, we may temporarily switch to a PLL running at 800MHz.
No CPUFREQ notification is sent for that. That means there's a period of time
when we're running at 800MHz but loops_per_jiffy is calibrated at between 200MHz
and 300MHz. And so udelay behaves badly.
To get this fixed in a generic way, introduce another set of callbacks
get_intermediate() and target_intermediate(), only for drivers with
target_index() and CPUFREQ_ASYNC_NOTIFICATION unset.
get_intermediate() should return a stable intermediate frequency platform wants
to switch to, and target_intermediate() should set CPU to that frequency,
before jumping to the frequency corresponding to 'index'. Core will take care of
sending notifications and driver doesn't have to handle them in
target_intermediate() or target_index().
NOTE: ->target_index() should restore to policy->restore_freq in case of
failures as core would send notifications for that.
Tested-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Doug Anderson <dianders@chromium.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
On platforms that use cpufreq_for_each_* macros, build fails if
CONFIG_CPU_FREQ=n, e.g. ARM/shmobile/koelsch/non-multiplatform:
drivers/built-in.o: In function `clk_round_parent':
clkdev.c:(.text+0xcf168): undefined reference to `cpufreq_next_valid'
drivers/built-in.o: In function `clk_rate_table_find':
clkdev.c:(.text+0xcf820): undefined reference to `cpufreq_next_valid'
make[3]: *** [vmlinux] Error 1
Fix this making cpufreq_next_valid function inline and move it to
cpufreq.h.
Fixes: 27e289dce2 (cpufreq: Introduce macros for cpufreq_frequency_table iteration)
Reported-and-tested-by: Geert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: Stratos Karafotis <stratosk@semaphore.gr>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
CPUFreq specific helper functions for OPP (Operating Performance Points)
now use generic OPP functions that allow CPUFreq to be be moved back
into CPUFreq framework. This allows for independent modifications
or future enhancements as needed isolated to just CPUFreq framework
alone.
Here, we just move relevant code and documentation to make this part of
CPUFreq infrastructure.
Cc: Kevin Hilman <khilman@deeprootsystems.com>
Signed-off-by: Nishanth Menon <nm@ti.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Some cpufreq drivers were redundantly invoking the _begin() and _end()
APIs around frequency transitions, and this double invocation (one from
the cpufreq core and the other from the cpufreq driver) used to result
in a self-deadlock, leading to system hangs during boot. (The _begin()
API makes contending callers wait until the previous invocation is
complete. Hence, the cpufreq driver would end up waiting on itself!).
Now all such drivers have been fixed, but debugging this issue was not
very straight-forward (even lockdep didn't catch this). So let us add a
debug infrastructure to the cpufreq core to catch such issues more easily
in the future.
We add a new field called 'transition_task' to the policy structure, to keep
track of the task which is performing the frequency transition. Using this
field, we make note of this task during _begin() and print a warning if we
find a case where the same task is calling _begin() again, before completing
the previous frequency transition using the corresponding _end().
We have left out ASYNC_NOTIFICATION drivers from this debug infrastructure
for 2 reasons:
1. At the moment, we have no way to avoid a particular scenario where this
debug infrastructure can emit false-positive warnings for such drivers.
The scenario is depicted below:
Task A Task B
/* 1st freq transition */
Invoke _begin() {
...
...
}
Change the frequency
/* 2nd freq transition */
Invoke _begin() {
... //waiting for B to
... //finish _end() for
... //the 1st transition
... | Got interrupt for successful
... | change of frequency (1st one).
... |
... | /* 1st freq transition */
... | Invoke _end() {
... | ...
... V }
...
...
}
This scenario is actually deadlock-free because, once Task A changes the
frequency, it is Task B's responsibility to invoke the corresponding
_end() for the 1st frequency transition. Hence it is perfectly legal for
Task A to go ahead and attempt another frequency transition in the meantime.
(Of course it won't be able to proceed until Task B finishes the 1st _end(),
but this doesn't cause a deadlock or a hang).
The debug infrastructure cannot handle this scenario and will treat it as
a deadlock and print a warning. To avoid this, we exclude such drivers
from the purview of this code.
2. Luckily, we don't _need_ this infrastructure for ASYNC_NOTIFICATION drivers
at all! The cpufreq core does not automatically invoke the _begin() and
_end() APIs during frequency transitions in such drivers. Thus, the driver
alone is responsible for invoking _begin()/_end() and hence there shouldn't
be any conflicts which lead to double invocations. So, we can skip these
drivers, since the probability that such drivers will hit this problem is
extremely low, as outlined above.
Signed-off-by: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Many cpufreq drivers need to iterate over the cpufreq_frequency_table
for various tasks.
This patch introduces two macros which can be used for iteration over
cpufreq_frequency_table keeping a common coding style across drivers:
- cpufreq_for_each_entry: iterate over each entry of the table
- cpufreq_for_each_valid_entry: iterate over each entry that contains
a valid frequency.
It should have no functional changes.
Signed-off-by: Stratos Karafotis <stratosk@semaphore.gr>
Acked-by: Lad, Prabhakar <prabhakar.csengg@gmail.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Currently cpufreq frequency table has two fields: frequency and driver_data.
driver_data is only for drivers' internal use and cpufreq core shouldn't use
it at all. But with the introduction of BOOST frequencies, this assumption
was broken and we started using it as a flag instead.
There are two problems due to this:
- It is against the description of this field, as driver's data is used by
the core now.
- if drivers fill it with -3 for any frequency, then those frequencies are
never considered by cpufreq core as it is exactly same as value of
CPUFREQ_BOOST_FREQ, i.e. ~2.
The best way to get this fixed is by creating another field flags which
will be used for such flags. This patch does that. Along with that various
drivers need modifications due to the change of struct cpufreq_frequency_table.
Reviewed-by: Gautham R Shenoy <ego@linux.vnet.ibm.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
cpufreq_notify_transition() and cpufreq_notify_post_transition() shouldn't be
called directly by cpufreq drivers anymore and so these should be marked static.
Reviewed-by: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Whenever we change the frequency of a CPU, we call the PRECHANGE and POSTCHANGE
notifiers. They must be serialized, i.e. PRECHANGE and POSTCHANGE notifiers
should strictly alternate, thereby preventing two different sets of PRECHANGE or
POSTCHANGE notifiers from interleaving arbitrarily.
The following examples illustrate why this is important:
Scenario 1:
-----------
A thread reading the value of cpuinfo_cur_freq, will call
__cpufreq_cpu_get()->cpufreq_out_of_sync()->cpufreq_notify_transition()
The ondemand governor can decide to change the frequency of the CPU at the same
time and hence it can end up sending the notifications via ->target().
If the notifiers are not serialized, the following sequence can occur:
- PRECHANGE Notification for freq A (from cpuinfo_cur_freq)
- PRECHANGE Notification for freq B (from target())
- Freq changed by target() to B
- POSTCHANGE Notification for freq B
- POSTCHANGE Notification for freq A
We can see from the above that the last POSTCHANGE Notification happens for freq
A but the hardware is set to run at freq B.
Where would we break then?: adjust_jiffies() in cpufreq.c & cpufreq_callback()
in arch/arm/kernel/smp.c (which also adjusts the jiffies). All the
loops_per_jiffy calculations will get messed up.
Scenario 2:
-----------
The governor calls __cpufreq_driver_target() to change the frequency. At the
same time, if we change scaling_{min|max}_freq from sysfs, it will end up
calling the governor's CPUFREQ_GOV_LIMITS notification, which will also call
__cpufreq_driver_target(). And hence we end up issuing concurrent calls to
->target().
Typically, platforms have the following logic in their ->target() routines:
(Eg: cpufreq-cpu0, omap, exynos, etc)
A. If new freq is more than old: Increase voltage
B. Change freq
C. If new freq is less than old: decrease voltage
Now, if the two concurrent calls to ->target() are X and Y, where X is trying to
increase the freq and Y is trying to decrease it, we get the following race
condition:
X.A: voltage gets increased for larger freq
Y.A: nothing happens
Y.B: freq gets decreased
Y.C: voltage gets decreased
X.B: freq gets increased
X.C: nothing happens
Thus we can end up setting a freq which is not supported by the voltage we have
set. That will probably make the clock to the CPU unstable and the system might
not work properly anymore.
This patch introduces a set of synchronization primitives to serialize frequency
transitions, which are to be used as shown below:
cpufreq_freq_transition_begin();
//Perform the frequency change
cpufreq_freq_transition_end();
The _begin() call sends the PRECHANGE notification whereas the _end() call sends
the POSTCHANGE notification. Also, all the necessary synchronization is handled
within these calls. In particular, even drivers which set the ASYNC_NOTIFICATION
flag can also use these APIs for performing frequency transitions (ie., you can
call _begin() from one task, and call the corresponding _end() from a different
task).
The actual synchronization underneath is not that complicated:
The key challenge is to allow drivers to begin the transition from one thread
and end it in a completely different thread (this is to enable drivers that do
asynchronous POSTCHANGE notification from bottom-halves, to also use the same
interface).
To achieve this, a 'transition_ongoing' flag, a 'transition_lock' spinlock and a
wait-queue are added per-policy. The flag and the wait-queue are used in
conjunction to create an "uninterrupted flow" from _begin() to _end(). The
spinlock is used to ensure that only one such "flow" is in flight at any given
time. Put together, this provides us all the necessary synchronization.
Signed-off-by: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
This callback allows the driver to do clean up before the CPU is
completely down and its state cannot be modified. This is used
by the intel_pstate driver to reduce the requested P state prior to
the core going away. This is required because the requested P state
of the offline core is used to select the package P state. This
effectively sets the floor package P state to the requested P state on
the offline core.
Signed-off-by: Dirk Brandewie <dirk.j.brandewie@intel.com>
[rjw: Minor modifications]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
