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>
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
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>
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>
