Function sugov_eas_rebuild_sd() defined in the schedutil cpufreq governor
implements generic functionality that may be useful in other places. In
particular, there is a plan to use it in the intel_pstate driver in the
future.
For this reason, move it from schedutil to the energy model code and
rename it to em_rebuild_sched_domains().
This also helps to get rid of some #ifdeffery in schedutil which is a
plus.
No intentional functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Christian Loehle <christian.loehle@arm.com>
When I booted my RK3588 based system I noticed that cpufreq complained
about system clock:
[ +0.007211] cpufreq: cpufreq_online: CPU0: Running at unlisted initial frequency: 816000 KHz, changing to: 1008000 KHz
Then I realized that unit is displayed wrong: "KHz" instead of "kHz".
Signed-off-by: Marcin Juszkiewicz <marcin.juszkiewicz@linaro.org>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Link: https://patch.msgid.link/20240909095529.2325103-1-marcin.juszkiewicz@linaro.org
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The current LATENCY_MULTIPLIER which has been around for nearly 20 years
causes rate_limit_us to be always in ms range.
On M1 mac mini I get 50 and 56us transition latency, but due to the 1000
multiplier we end up setting rate_limit_us to 50 and 56ms, which gets
capped into 2ms and was 10ms before e13aa799c2 ("cpufreq: Change
default transition delay to 2ms")
On Intel I5 system transition latency is 20us but due to the multiplier
we end up with 20ms that again is capped to 2ms.
Given how good modern hardware and how modern workloads require systems
to be more responsive to cater for sudden changes in workload (tasks
sleeping/wakeup/migrating, uclamp causing a sudden boost or cap) and
that 2ms is quarter of the time of 120Hz refresh rate system, drop the
old logic in favour of providing 50% headroom.
rate_limit_us = 1.5 * latency.
I considered not adding any headroom which could mean that we can end up
with infinite back-to-back requests.
I also considered providing a constant headroom (e.g: 100us) assuming
that any h/w or f/w dealing with the request shouldn't require a large
headroom when transition_latency is actually high.
But for both cases I wasn't sure if h/w or f/w can end up being
overwhelmed dealing with the freq requests in a potentially busy system.
So I opted for providing 50% breathing room.
This is expected to impact schedutil only as the other user,
dbs_governor, takes the max(2*tick, transition_delay_us) and the former
was at least 2ms on 1ms TICK, which is equivalent to the max_delay_us
before applying this patch. For systems with TICK of 4ms, this value
would have almost always ended up with 8ms sampling rate.
For systems that report 0 transition latency, we still default to
returning 1ms as transition delay.
This helps in eliminating a source of latency for applying requests as
mentioned in [1]. For example if we have a 1ms tick, most systems will
miss sending an update at tick when updating the util_avg for a task/CPU
(rate_limit_us will be 2ms for most systems).
Link: https://lore.kernel.org/lkml/20240724212255.mfr2ybiv2j2uqek7@airbuntu/ # [1]
Link: https://lore.kernel.org/lkml/20240205022500.2232124-1-qyousef@layalina.io/
Signed-off-by: Qais Yousef <qyousef@layalina.io>
Link: https://patch.msgid.link/20240728192659.58115-1-qyousef@layalina.io
[ rjw: Subject edits ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The behavior introduced in commit f37a4d6b4a ("cpufreq: Fix per-policy
boost behavior on SoCs using cpufreq_boost_set_sw()") sets up the boost
policy incorrectly when boost has been enabled by the platform firmware
initially even if a driver sets the policy up.
This is because policy_has_boost_freq() assumes that there is a frequency
table set up by the driver and that the boost frequencies are advertised
in that table. This assumption doesn't work for acpi-cpufreq or
amd-pstate. Only use this check to enable boost if it's not already
enabled instead of also disabling it if alreayd enabled.
Fixes: f37a4d6b4a ("cpufreq: Fix per-policy boost behavior on SoCs using cpufreq_boost_set_sw()")
Link: https://patch.msgid.link/20240626204723.6237-1-mario.limonciello@amd.com
Reviewed-by: Sibi Sankar <quic_sibis@quicinc.com>
Reviewed-by: Dhruva Gole <d-gole@ti.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Gautham R. Shenoy <gautham.shenoy@amd.com>
Suggested-by: Viresh Kumar <viresh.kumar@linaro.org>
Suggested-by: Gautham R. Shenoy <gautham.shenoy@amd.com>
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Cc: All applicable <stable@vger.kernel.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Since this function is supposed to return boost_enabled which is anyway
a bool type make sure that it's return value is also marked as bool.
This helps maintain better consistency in data types being used.
Signed-off-by: Dhruva Gole <d-gole@ti.com>
Reviewed-by: Mario Limonciello <mario.limonciello@amd.com>
Link: https://patch.msgid.link/20240627060117.1809477-1-d-gole@ti.com
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
* Enables amd-pstate by default in "shared memory" designs without a dedicated MSR.
* Adds extra infrastructure for debugging problems.
* Bug fixes found for init/unload failure
-----BEGIN PGP SIGNATURE-----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=Eea8
-----END PGP SIGNATURE-----
Merge tag 'amd-pstate-v6.11-2024-06-24' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/superm1/linux
Merge the second round of changes for amd-pstate in 6.11 from Mario
Limonciello:
"* Enables amd-pstate by default in "shared memory" designs without
a dedicated MSR.
* Adds extra infrastructure for debugging problems.
* Bug fixes found for init/unload failure."
* tag 'amd-pstate-v6.11-2024-06-24' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/superm1/linux:
cpufreq: simplify boolean parsing with kstrtobool in store function
cpufreq: amd-pstate: Don't create attributes when registration fails
cpufreq: amd-pstate: Make amd-pstate unit tests depend on amd-pstate
cpufreq/amd-pstate: fix setting policy current frequency value
cpufreq: amd-pstate: auto-load pstate driver by default
cpufreq: amd-pstate: enable shared memory type CPPC by default
cpufreq: amd-pstate: switch boot_cpu_has() to cpu_feature_enabled()
Documentation: PM: amd-pstate: add guided mode to the Operation mode
cpufreq: amd-pstate: add debug message while CPPC is supported and disabled by SBIOS
cpufreq: amd-pstate: show CPPC debug message if CPPC is not supported
cpufreq: amd-pstate: remove unused variable nominal_freq
cpufreq: amd-pstate: optimize the initial frequency values verification
cpufreq: amd-pstate: Allow users to write 'default' EPP string
Update the cpufreq store function to use kstrtobool for parsing boolean
values. This simplifies the code and improves readability by using a
standard kernel function for boolean string conversion.
Reviewed-by: Mario Limonciello <mario.limonciello@amd.com>
Reviewed-by: Gautham R. Shenoy <gautham.shenoy@amd.com>
Signed-off-by: Perry Yuan <perry.yuan@amd.com>
Link: https://lore.kernel.org/r/d392eba3bad1231776124c321cef8c184ce1482d.1718988436.git.perry.yuan@amd.com
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Replace sprintf() and scnprintf() with sysfs_emit() and sysfs_emit_at()
in the cpufreq core.
This ensures safer buffer handling and consistency with sysfs interfaces.
Update show_scaling_available_governors() and related functions for
compliance with the new API.
Signed-off-by: Perry Yuan <perry.yuan@amd.com>
Reviewed-by: Mario Limonciello <mario.limonciello@amd.com>
Link: https://patch.msgid.link/20240619081520.259971-1-perry.yuan@amd.com
[ rjw: Changelog edits ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
- Rework the handling of disabled turbo in the intel_pstate driver and
make it update the maximum CPU frequency consistently regardless of
the reason on top of a number of cleanups (Rafael Wysocki).
- Add missing checks for NULL .exit() cpufreq driver callback to the
cpufreq core (Viresh Kumar).
- Prevent pulicy->max from going above the frequency QoS maximum value
when cpufreq_frequency_table_verify() is used (Xuewen Yan).
- Prevent a negative CPU number or frequency value from being printed
if they are really large (Joshua Yeong).
- Update MAINTAINERS entry for amd-pstate to add two new submaintainers
and a designated reviewer (Huang Rui).
- Clean up the amd-pstate driver and update its documentation (Gautham
Shenoy).
- Fix the highest frequency issue in the amd-pstate driver which limits
performance (Perry Yuan).
- Enable CPPC v2 for certain processors in the family 17H, as requested
by TR40 processor users who expect improved performance and lower
system temperature (Perry Yuan).
- Change latency and delay values to be read from platform firmware
firstly for more accurate timing (Perry Yuan).
- A new quirk is introduced for supporting amd-pstate on legacy
processors which either lack CPPC capability, or only only have CPPC
v2 capability (Perry Yuan).
- Sun50i cpufreq: Add support for opp_supported_hw, H616 platform and
general cleanups (Andre Przywara, Martin Botka, Brandon Cheo Fusi,
Dan Carpenter, Viresh Kumar).
- CPPC cpufreq: Fix possible null pointer dereference (Aleksandr
Mishin).
- Eliminate uses of of_node_put() from cpufreq (Javier Carrasco,
Shivani Gupta).
- brcmstb-avs: ISO C90 forbids mixed declarations (Portia Stephens).
- mediatek cpufreq: Add support for MT7988A (Sam Shih).
- cpufreq-qcom-hw: Add SM4450 compatibles in DT bindings (Tengfei Fan).
- Fix struct cpudata::epp_cached kernel-doc in the intel_pstate cpufreq
driver (Jeff Johnson).
- Fix kerneldoc description of ladder_do_selection() (Jeff Johnson).
- Convert the cpuidle kirkwood driver to platform remove callback
returning void (Yangtao Li).
- Replace deprecated strncpy() with strscpy() in the hibernation core
code (Justin Stitt).
- Use %ps to simplify debug output in the core system-wide suspend and
resume code (Len Brown).
- Remove unnecessary else from device_init_wakeup() and make
device_wakeup_disable() return void (Dhruva Gole).
- Enable PMU support in the Intel TPMI RAPL driver (Zhang Rui).
- Add support for ArrowLake-H platform to the Intel RAPL driver (Zhang
Rui).
- Avoid explicit cpumask allocation on stack in DTPM (Dawei Li).
- Make the Samsung exynos-asv driver update the Energy Model after
adjusting voltage on top of some preliminary changes of the OPP and
Enery Model generic code (Lukasz Luba).
- Remove a reference to a function that has been dropped from the power
management documentation (Bjorn Helgaas).
- Convert the platfrom remove callback to .remove_new for the
exyno-nocp, exynos-ppmu, mtk-cci-devfreq, sun8i-a33-mbus, and
rk3399_dmc devfreq drivers (Uwe Kleine-König).
- Use DEFINE_SIMPLE_PM_OPS for exyno-bus.c driver (Anand Moon).
-----BEGIN PGP SIGNATURE-----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=el5Y
-----END PGP SIGNATURE-----
Merge tag 'pm-6.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm
Pull power management updates from Rafael Wysocki:
"These are mostly cpufreq updates, including a significant intel-pstate
driver update and several amd-pstate improvements plus some updates of
ARM cpufreq drivers, general fixes and cleanups.
Also included are changes related to system sleep, power capping
updates adding support for a new platform and a new hardware feature
(among other things), a Samsung exynos-asv driver update allowing it
to change its Energy Model after adjusting voltage, minor cpuidle and
devfreq updates and a small documentation cleanup.
Specifics:
- Rework the handling of disabled turbo in the intel_pstate driver
and make it update the maximum CPU frequency consistently
regardless of the reason on top of a number of cleanups (Rafael
Wysocki)
- Add missing checks for NULL .exit() cpufreq driver callback to the
cpufreq core (Viresh Kumar)
- Prevent pulicy->max from going above the frequency QoS maximum
value when cpufreq_frequency_table_verify() is used (Xuewen Yan)
- Prevent a negative CPU number or frequency value from being printed
if they are really large (Joshua Yeong)
- Update MAINTAINERS entry for amd-pstate to add two new
submaintainers and a designated reviewer (Huang Rui)
- Clean up the amd-pstate driver and update its documentation
(Gautham Shenoy)
- Fix the highest frequency issue in the amd-pstate driver which
limits performance (Perry Yuan)
- Enable CPPC v2 for certain processors in the family 17H, as
requested by TR40 processor users who expect improved performance
and lower system temperature (Perry Yuan)
- Change latency and delay values to be read from platform firmware
firstly for more accurate timing (Perry Yuan)
- A new quirk is introduced for supporting amd-pstate on legacy
processors which either lack CPPC capability, or only only have
CPPC v2 capability (Perry Yuan)
- Sun50i cpufreq: Add support for opp_supported_hw, H616 platform and
general cleanups (Andre Przywara, Martin Botka, Brandon Cheo Fusi,
Dan Carpenter, Viresh Kumar)
- CPPC cpufreq: Fix possible null pointer dereference (Aleksandr
Mishin)
- Eliminate uses of of_node_put() from cpufreq (Javier Carrasco,
Shivani Gupta)
- brcmstb-avs: ISO C90 forbids mixed declarations (Portia Stephens)
- mediatek cpufreq: Add support for MT7988A (Sam Shih)
- cpufreq-qcom-hw: Add SM4450 compatibles in DT bindings (Tengfei
Fan)
- Fix struct cpudata::epp_cached kernel-doc in the intel_pstate
cpufreq driver (Jeff Johnson)
- Fix kerneldoc description of ladder_do_selection() (Jeff Johnson)
- Convert the cpuidle kirkwood driver to platform remove callback
returning void (Yangtao Li)
- Replace deprecated strncpy() with strscpy() in the hibernation core
code (Justin Stitt)
- Use %ps to simplify debug output in the core system-wide suspend
and resume code (Len Brown)
- Remove unnecessary else from device_init_wakeup() and make
device_wakeup_disable() return void (Dhruva Gole)
- Enable PMU support in the Intel TPMI RAPL driver (Zhang Rui)
- Add support for ArrowLake-H platform to the Intel RAPL driver
(Zhang Rui)
- Avoid explicit cpumask allocation on stack in DTPM (Dawei Li)
- Make the Samsung exynos-asv driver update the Energy Model after
adjusting voltage on top of some preliminary changes of the OPP and
Enery Model generic code (Lukasz Luba)
- Remove a reference to a function that has been dropped from the
power management documentation (Bjorn Helgaas)
- Convert the platfrom remove callback to .remove_new for the
exyno-nocp, exynos-ppmu, mtk-cci-devfreq, sun8i-a33-mbus, and
rk3399_dmc devfreq drivers (Uwe Kleine-König)
- Use DEFINE_SIMPLE_PM_OPS for exyno-bus.c driver (Anand Moon)"
* tag 'pm-6.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm: (68 commits)
PM / devfreq: exynos: Use DEFINE_SIMPLE_DEV_PM_OPS for PM functions
PM / devfreq: rk3399_dmc: Convert to platform remove callback returning void
PM / devfreq: sun8i-a33-mbus: Convert to platform remove callback returning void
PM / devfreq: mtk-cci: Convert to platform remove callback returning void
PM / devfreq: exynos-ppmu: Convert to platform remove callback returning void
PM / devfreq: exynos-nocp: Convert to platform remove callback returning void
cpufreq: amd-pstate: fix the highest frequency issue which limits performance
cpufreq: intel_pstate: fix struct cpudata::epp_cached kernel-doc
cpuidle: ladder: fix ladder_do_selection() kernel-doc
powercap: intel_rapl_tpmi: Enable PMU support
powercap: intel_rapl: Introduce APIs for PMU support
PM: hibernate: replace deprecated strncpy() with strscpy()
cpufreq: Fix up printing large CPU numbers and frequency values
MAINTAINERS: cpufreq: amd-pstate: Add co-maintainers and reviewer
cpufreq: amd-pstate: remove unused variable lowest_nonlinear_freq
cpufreq: amd-pstate: fix code format problems
cpufreq: amd-pstate: Add quirk for the pstate CPPC capabilities missing
cppc_acpi: print error message if CPPC is unsupported
cpufreq: amd-pstate: get transition delay and latency value from ACPI tables
cpufreq: amd-pstate: Bail out if min/max/nominal_freq is 0
...
Provide to the scheduler a feedback about the temporary max available
capacity. Unlike arch_update_thermal_pressure(), this doesn't need to be
filtered as the pressure will happen for dozens of ms or more.
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Tested-by: Lukasz Luba <lukasz.luba@arm.com>
Reviewed-by: Qais Yousef <qyousef@layalina.io>
Reviewed-by: Lukasz Luba <lukasz.luba@arm.com>
Reviewed-by: Dhruva Gole <d-gole@ti.com>
Acked-by: Rafael J. Wysocki <rafael@kernel.org>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Link: https://lore.kernel.org/r/20240326091616.3696851-2-vincent.guittot@linaro.org
The exit() callback is optional and shouldn't be called without checking
a valid pointer first.
Also, we must clear freq_table pointer even if the exit() callback isn't
present.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Fixes: 91a12e91dc ("cpufreq: Allow light-weight tear down and bring up of CPUs")
Fixes: f339f35417 ("cpufreq: Rearrange locking in cpufreq_remove_dev()")
Reported-by: Lizhe <sensor1010@163.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
In the existing code, per-policy flags don't have any impact i.e.
if cpufreq_driver boost is enabled and boost is disabled for one or
more of the policies, the cpufreq driver will behave as if boost is
enabled.
Fix this by incorporating per-policy boost flag in the policy->max
computation used in cpufreq_frequency_table_cpuinfo and setting the
default per-policy boost to mirror the cpufreq_driver boost flag.
Fixes: 218a06a79d ("cpufreq: Support per-policy performance boost")
Reported-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Dhruva Gole <d-gole@ti.com>
Signed-off-by: Sibi Sankar <quic_sibis@quicinc.com>
Tested-by:Yipeng Zou <zouyipeng@huawei.com> <mailto:zouyipeng@huawei.com>
Reviewed-by: Yipeng Zou <zouyipeng@huawei.com> <mailto:zouyipeng@huawei.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Offlining a CPU and bringing it back online is a common operation and it
happens frequently during system suspend/resume, where the non-boot CPUs
are hotplugged out during suspend and brought back at resume.
The cpufreq core already tries to make this path as fast as possible as
the changes are only temporary in nature and full cleanup of resources
isn't required in this case. For example the drivers can implement
online()/offline() callbacks to avoid a lot of tear down of resources.
On similar lines, there is no need to unregister the cpufreq cooling
device during suspend / resume, but only while the policy is getting
removed.
Moreover, unregistering the cpufreq cooling device is resulting in an
unwanted outcome, where the system suspend is eventually aborted in the
process. Currently, during system suspend the cpufreq core unregisters
the cooling device, which in turn removes a kobject using device_del()
and that generates a notification to the userspace via uevent broadcast.
This causes system suspend to abort in some setups.
This was also earlier reported (indirectly) by Roman [1]. Maybe there is
another way around to fixing that problem properly, but this change
makes sense anyways.
Move the registering and unregistering of the cooling device to policy
creation and removal times onlyy.
Closes: https://bugzilla.kernel.org/show_bug.cgi?id=218521
Reported-by: Manaf Meethalavalappu Pallikunhi <quic_manafm@quicinc.com>
Reported-by: Roman Stratiienko <r.stratiienko@gmail.com>
Link: https://patchwork.kernel.org/project/linux-pm/patch/20220710164026.541466-1-r.stratiienko@gmail.com/ [1]
Tested-by: Manaf Meethalavalappu Pallikunhi <quic_manafm@quicinc.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Dhruva Gole <d-gole@ti.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Some platforms like Arm's Juno can have a high transition latency that
can be larger than the 2ms cap introduced. If a driver reports
a transition_latency that is higher than the cap, then use it as-is.
Update comment s/10/2/ to reflect the new cap of 2ms.
Reported-by: Pierre Gondois <pierre.gondois@arm.com>
Signed-off-by: Qais Yousef <qyousef@layalina.io>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
10ms is too high for today's hardware, even low end ones. This default
end up being used a lot on Arm machines at least. Pine64, mac mini and
pixel 6 all end up with 10ms rate_limit_us when using schedutil, and
it's too high for all of them.
Change the default to 2ms which should be 'pessimistic' enough for worst
case scenario, but not too high for platforms with fast DVFS hardware.
Signed-off-by: Qais Yousef <qyousef@layalina.io>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
cpuinfo.max_freq can change at runtime because of boost as an example. This
implies that the value could be different from the frequency that has been
used to compute the capacity of a CPU.
The new arch_scale_freq_ref() returns a fixed and coherent frequency
that can be used to compute the capacity for a given frequency.
[ Also fix a arch_set_freq_scale() newline style wart in <linux/cpufreq.h>. ]
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Tested-by: Lukasz Luba <lukasz.luba@arm.com>
Reviewed-by: Lukasz Luba <lukasz.luba@arm.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Acked-by: Rafael J. Wysocki <rafael@kernel.org>
Link: https://lore.kernel.org/r/20231211104855.558096-3-vincent.guittot@linaro.org
- Add support for several Qualcomm SoC versions and other similar
changes (Christian Marangi, Dmitry Baryshkov, Luca Weiss, Neil
Armstrong, Richard Acayan, Robert Marko, Rohit Agarwal, Stephan
Gerhold and Varadarajan Narayanan).
- Clean up the tegra cpufreq driver (Sumit Gupta).
- Use of_property_read_reg() to parse "reg" in pmac32 driver (Rob
Herring).
- Add support for TI's am62p5 Soc (Bryan Brattlof).
- Make ARM_BRCMSTB_AVS_CPUFREQ depends on !ARM_SCMI_CPUFREQ (Florian
Fainelli).
- Update Kconfig to mention i.MX7 as well (Alexander Stein).
- Revise global turbo disable check in intel_pstate (Srinivas
Pandruvada).
- Carry out initialization of sg_cpu in the schedutil cpufreq governor
in one loop (Liao Chang).
- Simplify the condition for storing 'down_threshold' in the
conservative cpufreq governor (Liao Chang).
- Use fine-grained mutex in the userspace cpufreq governor (Liao
Chang).
- Move is_managed indicator in the userspace cpufreq governor into a
per-policy structure (Liao Chang).
- Rebuild sched-domains when removing cpufreq driver (Pierre Gondois).
- Fix buffer overflow detection in trans_stats() (Christian Marangi).
- Switch to dev_pm_opp_find_freq_(ceil/floor)_indexed() APIs to support
specific devices like UFS which handle multiple clocks through OPP
(Operating Performance Point) framework (Manivannan Sadhasivam).
- Add perf support to the Rockchip DFI (DDR Monitor Module) devfreq-
event driver:
* Generalize rockchip-dfi.c to support new RK3568/RK3588 using
different DDR type (Sascha Hauer).
* Convert DT binding document format to yaml (Sascha Hauer).
* Add perf support for DFI (a unit suitable for measuring DDR
utilization) to rockchip-dfi.c to extend DFI usage (Sascha Hauer).
- Add locking to the OPP handling code in the Mediatek CCI devfreq
driver, because the voltage of shared OPP might be changed by
multiple drivers (Mark Tseng, Dan Carpenter).
- Use device_get_match_data() in the Samsung Exynos PPMU devfreq-event
driver (Rob Herring).
- Extend support for the opp-level beyond required-opps (Ulf Hansson).
- Add dev_pm_opp_find_level_floor() (Krishna chaitanya chundru).
- dt-bindings: Allow opp-peak-kBpsfor kryo CPUs, support Qualcomm Krait
SoCs and document named opp-microvolt property (Bjorn Andersson,
Dmitry Baryshkov and Christian Marangi).
- Fix -Wunsequenced warning _of_add_opp_table_v1() (Nathan Chancellor).
- General cleanup of OPP code (Viresh Kumar).
- Use __get_safe_page() rather than touching the list in hibernation
snapshot code (Brian Geffon).
- Fix symbol export for _SIMPLE_ variants of _PM_OPS() (Raag Jadav).
- Clean up sync_read handling in snapshot_write_next() (Brian Geffon).
- Fix kerneldoc comments for swsusp_check() and swsusp_close() to
better match code (Christoph Hellwig).
- Downgrade BIOS locked limits pr_warn() in the Intel RAPL power
capping driver to pr_debug() (Ville Syrjälä).
- Change the minimum python version for the intel_pstate_tracer utility
from 2.7 to 3.6 (Doug Smythies).
-----BEGIN PGP SIGNATURE-----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=h8my
-----END PGP SIGNATURE-----
Merge tag 'pm-6.7-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm
Pull power management updates from Rafael Wysocki:
"These add new hardware support (new Qualcomm SoC versions in cpufreq,
RK3568/RK3588 in devfreq), extend the OPP (operating performance
points) framework, improve cpufreq governors, fix issues and clean up
code (most of the changes are in cpufreq and devfreq).
Specifics:
- Add support for several Qualcomm SoC versions and other similar
changes (Christian Marangi, Dmitry Baryshkov, Luca Weiss, Neil
Armstrong, Richard Acayan, Robert Marko, Rohit Agarwal, Stephan
Gerhold and Varadarajan Narayanan)
- Clean up the tegra cpufreq driver (Sumit Gupta)
- Use of_property_read_reg() to parse "reg" in pmac32 driver (Rob
Herring)
- Add support for TI's am62p5 Soc (Bryan Brattlof)
- Make ARM_BRCMSTB_AVS_CPUFREQ depends on !ARM_SCMI_CPUFREQ (Florian
Fainelli)
- Update Kconfig to mention i.MX7 as well (Alexander Stein)
- Revise global turbo disable check in intel_pstate (Srinivas
Pandruvada)
- Carry out initialization of sg_cpu in the schedutil cpufreq
governor in one loop (Liao Chang)
- Simplify the condition for storing 'down_threshold' in the
conservative cpufreq governor (Liao Chang)
- Use fine-grained mutex in the userspace cpufreq governor (Liao
Chang)
- Move is_managed indicator in the userspace cpufreq governor into a
per-policy structure (Liao Chang)
- Rebuild sched-domains when removing cpufreq driver (Pierre Gondois)
- Fix buffer overflow detection in trans_stats() (Christian Marangi)
- Switch to dev_pm_opp_find_freq_(ceil/floor)_indexed() APIs to
support specific devices like UFS which handle multiple clocks
through OPP (Operating Performance Point) framework (Manivannan
Sadhasivam)
- Add perf support to the Rockchip DFI (DDR Monitor Module) devfreq-
event driver:
* Generalize rockchip-dfi.c to support new RK3568/RK3588 using
different DDR type (Sascha Hauer).
* Convert DT binding document format to yaml (Sascha Hauer).
* Add perf support for DFI (a unit suitable for measuring DDR
utilization) to rockchip-dfi.c to extend DFI usage (Sascha
Hauer)
- Add locking to the OPP handling code in the Mediatek CCI devfreq
driver, because the voltage of shared OPP might be changed by
multiple drivers (Mark Tseng, Dan Carpenter)
- Use device_get_match_data() in the Samsung Exynos PPMU
devfreq-event driver (Rob Herring)
- Extend support for the opp-level beyond required-opps (Ulf Hansson)
- Add dev_pm_opp_find_level_floor() (Krishna chaitanya chundru)
- dt-bindings: Allow opp-peak-kBpsfor kryo CPUs, support Qualcomm
Krait SoCs and document named opp-microvolt property (Bjorn
Andersson, Dmitry Baryshkov and Christian Marangi)
- Fix -Wunsequenced warning _of_add_opp_table_v1() (Nathan
Chancellor)
- General cleanup of OPP code (Viresh Kumar)
- Use __get_safe_page() rather than touching the list in hibernation
snapshot code (Brian Geffon)
- Fix symbol export for _SIMPLE_ variants of _PM_OPS() (Raag Jadav)
- Clean up sync_read handling in snapshot_write_next() (Brian Geffon)
- Fix kerneldoc comments for swsusp_check() and swsusp_close() to
better match code (Christoph Hellwig)
- Downgrade BIOS locked limits pr_warn() in the Intel RAPL power
capping driver to pr_debug() (Ville Syrjälä)
- Change the minimum python version for the intel_pstate_tracer
utility from 2.7 to 3.6 (Doug Smythies)"
* tag 'pm-6.7-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm: (82 commits)
dt-bindings: cpufreq: qcom-hw: document SM8650 CPUFREQ Hardware
cpufreq: arm: Kconfig: Add i.MX7 to supported SoC for ARM_IMX_CPUFREQ_DT
cpufreq: qcom-nvmem: add support for IPQ8064
cpufreq: qcom-nvmem: also accept operating-points-v2-krait-cpu
cpufreq: qcom-nvmem: drop pvs_ver for format a fuses
dt-bindings: cpufreq: qcom-cpufreq-nvmem: Document krait-cpu
cpufreq: qcom-nvmem: add support for IPQ6018
dt-bindings: cpufreq: qcom-cpufreq-nvmem: document IPQ6018
cpufreq: qcom-nvmem: Add MSM8909
cpufreq: qcom-nvmem: Simplify driver data allocation
powercap: intel_rapl: Downgrade BIOS locked limits pr_warn() to pr_debug()
cpufreq: stats: Fix buffer overflow detection in trans_stats()
dt-bindings: devfreq: event: rockchip,dfi: Add rk3588 support
dt-bindings: devfreq: event: rockchip,dfi: Add rk3568 support
dt-bindings: devfreq: event: convert Rockchip DFI binding to yaml
PM / devfreq: rockchip-dfi: add support for RK3588
PM / devfreq: rockchip-dfi: account for multiple DDRMON_CTRL registers
PM / devfreq: rockchip-dfi: make register stride SoC specific
PM / devfreq: rockchip-dfi: Add perf support
PM / devfreq: rockchip-dfi: give variable a better name
...
The Energy Aware Scheduler (EAS) relies on the schedutil governor.
When moving to/from the schedutil governor, sched domains must be
rebuilt to allow re-evaluating the enablement conditions of EAS.
This is done through sched_cpufreq_governor_change().
Having a cpufreq governor assumes a cpufreq driver is running.
Inserting/removing a cpufreq driver should trigger a re-evaluation
of EAS enablement conditions, avoiding to see EAS enabled when
removing a running cpufreq driver.
Rebuild the sched domains in schedutil's sugov_init()/sugov_exit(),
allowing to check EAS's enablement condition whenever schedutil
governor is initialized/exited from.
Move relevant code up in schedutil.c to avoid a split and conditional
function declaration.
Rename sched_cpufreq_governor_change() to sugov_eas_rebuild_sd().
Signed-off-by: Pierre Gondois <pierre.gondois@arm.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The boost control currently applies to the whole system. However, users
may prefer to boost a subset of cores in order to provide prioritized
performance to workloads running on the boosted cores.
Enable per-policy boost by adding a 'boost' sysfs interface under each
policy path. This can be found at:
/sys/devices/system/cpu/cpufreq/policy<*>/boost
Same to the global boost switch, writing 1/0 to the per-policy 'boost'
enables/disables boost on a cpufreq policy respectively.
The user view of global and per-policy boost controls should be:
1. Enabling global boost initially enables boost on all policies, and
per-policy boost can then be enabled or disabled individually, given that
the platform does support so.
2. Disabling global boost makes the per-policy boost interface illegal.
Signed-off-by: Jie Zhan <zhanjie9@hisilicon.com>
Reviewed-by: Wei Xu <xuwei5@hisilicon.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The field 'transition_task' of policy structure is used to track the
task which is performing the frequency transition. Using this field to
print a warning once detect a case where the same task is calling
_begin() again before completing the preivous frequency transition via
the _end().
However, there is a potential race condition in _end() and _begin() APIs
while updating the field 'transition_task' of policy, the scenario is
depicted below:
Task A Task B
/* 1st freq transition */
Invoke _begin() {
...
...
}
/* 2nd freq transition */
Invoke _begin() {
... //waiting for A to
... //clear
... //transition_ongoing
... //in _end() for
... //the 1st transition
|
Change the frequency |
|
Invoke _end() { |
... |
... |
transition_ongoing = false; V
transition_ongoing = true;
transition_task = current;
transition_task = NULL;
... //A overwrites the task
... //performing the transition
... //result in error warning.
}
To fix this race condition, the transition_lock of policy structure is
now acquired before updating policy structure in _end() API. Which ensure
that only one task can update the 'transition_task' field at a time.
Link: https://lore.kernel.org/all/b3c61d8a-d52d-3136-fbf0-d1de9f1ba411@huawei.com/
Fixes: ca654dc3a9 ("cpufreq: Catch double invocations of cpufreq_freq_transition_begin/end")
Signed-off-by: Liao Chang <liaochang1@huawei.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The pointer value of policy and driver structure are currently printed
in the error messages of cpufreq_resume(), this is not recommended and
helpful.
In order to be consistent with the error message in cpufreq_suspend()
and easier to understand, print the name of driver strcture and the
manage CPU of policy structure individually in the error messages of
cpufreq_resume().
Link: https://lore.kernel.org/all/b7be717c-41d8-bbbf-3e97-3799948ab757@huawei.com
Signed-off-by: Liao Chang <liaochang1@huawei.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
When a cpufreq_policy is allocated, the cpus, related_cpus and real_cpus
of policy are still unset. Therefore, it is preferable to print the
passed 'cpu' parameter instead of a empty 'cpus' cpumask in error
message when registering MIN/MAX QoS notifier fails.
Signed-off-by: Liao Chang <liaochang1@huawei.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
If fast_switch_possible flag is set by the scaling driver, the governor
is free to select fast_switch function even if adjust_perf is set. Some
scaling drivers which use adjust_perf don't set fast_switch thinking
that the governor would never fall back to fast_switch. But the governor
can fall back to fast_switch even in runtime if frequency invariance is
disabled due to some reason. This could crash the kernel if the driver
didn't set the fast_switch function pointer.
Therefore, fail driver registration if it has adjust_perf without
fast_switch.
Suggested-by: Rafael J. Wysocki <rafael@kernel.org>
Suggested-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Wyes Karny <wyes.karny@amd.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Here is the large set of driver core changes for 6.4-rc1.
Once again, a busy development cycle, with lots of changes happening in
the driver core in the quest to be able to move "struct bus" and "struct
class" into read-only memory, a task now complete with these changes.
This will make the future rust interactions with the driver core more
"provably correct" as well as providing more obvious lifetime rules for
all busses and classes in the kernel.
The changes required for this did touch many individual classes and
busses as many callbacks were changed to take const * parameters
instead. All of these changes have been submitted to the various
subsystem maintainers, giving them plenty of time to review, and most of
them actually did so.
Other than those changes, included in here are a small set of other
things:
- kobject logging improvements
- cacheinfo improvements and updates
- obligatory fw_devlink updates and fixes
- documentation updates
- device property cleanups and const * changes
- firwmare loader dependency fixes.
All of these have been in linux-next for a while with no reported
problems.
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-----BEGIN PGP SIGNATURE-----
iG0EABECAC0WIQT0tgzFv3jCIUoxPcsxR9QN2y37KQUCZEp7Sw8cZ3JlZ0Brcm9h
aC5jb20ACgkQMUfUDdst+ykitQCfamUHpxGcKOAGuLXMotXNakTEsxgAoIquENm5
LEGadNS38k5fs+73UaxV
=7K4B
-----END PGP SIGNATURE-----
Merge tag 'driver-core-6.4-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core
Pull driver core updates from Greg KH:
"Here is the large set of driver core changes for 6.4-rc1.
Once again, a busy development cycle, with lots of changes happening
in the driver core in the quest to be able to move "struct bus" and
"struct class" into read-only memory, a task now complete with these
changes.
This will make the future rust interactions with the driver core more
"provably correct" as well as providing more obvious lifetime rules
for all busses and classes in the kernel.
The changes required for this did touch many individual classes and
busses as many callbacks were changed to take const * parameters
instead. All of these changes have been submitted to the various
subsystem maintainers, giving them plenty of time to review, and most
of them actually did so.
Other than those changes, included in here are a small set of other
things:
- kobject logging improvements
- cacheinfo improvements and updates
- obligatory fw_devlink updates and fixes
- documentation updates
- device property cleanups and const * changes
- firwmare loader dependency fixes.
All of these have been in linux-next for a while with no reported
problems"
* tag 'driver-core-6.4-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core: (120 commits)
device property: make device_property functions take const device *
driver core: update comments in device_rename()
driver core: Don't require dynamic_debug for initcall_debug probe timing
firmware_loader: rework crypto dependencies
firmware_loader: Strip off \n from customized path
zram: fix up permission for the hot_add sysfs file
cacheinfo: Add use_arch[|_cache]_info field/function
arch_topology: Remove early cacheinfo error message if -ENOENT
cacheinfo: Check cache properties are present in DT
cacheinfo: Check sib_leaf in cache_leaves_are_shared()
cacheinfo: Allow early level detection when DT/ACPI info is missing/broken
cacheinfo: Add arm64 early level initializer implementation
cacheinfo: Add arch specific early level initializer
tty: make tty_class a static const structure
driver core: class: remove struct class_interface * from callbacks
driver core: class: mark the struct class in struct class_interface constant
driver core: class: make class_register() take a const *
driver core: class: mark class_release() as taking a const *
driver core: remove incorrect comment for device_create*
MIPS: vpe-cmp: remove module owner pointer from struct class usage.
...
cpufreq_verify_current_freq checks() if the frequency returned by
the hardware has a slight delta with the valid frequency value
last set and returns "policy->cur" if the delta is within "1 MHz".
In the comparison, "policy->cur" is in "kHz" but it's compared
against HZ_PER_MHZ. So, the comparison range becomes "1 GHz".
Fix this by comparing against KHZ_PER_MHZ instead of HZ_PER_MHZ.
Fixes: f55ae08c89 ("cpufreq: Avoid unnecessary frequency updates due to mismatch")
Signed-off-by: Sanjay Chandrashekara <sanjayc@nvidia.com>
[ sumit gupta: Commit message update ]
Signed-off-by: Sumit Gupta <sumitg@nvidia.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Since the cpufreq core directly uses freq_table, for cpufreq drivers
that set their target_index() callback, make it mandatory for them to
set the same.
Since this is set per policy and normally from policy->init(), do this
from cpufreq_table_validate_and_sort() which gets called right after
->init().
Reported-by: Yajun Deng <yajun.deng@linux.dev>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
When echo an invalid val to scaling_min_freq:
> echo 123abc123 > scaling_min_freq
It looks weird to have a return val of 0:
> echo $?
> 0
Sane people won't echo strings like that into these interfaces but fuzz
tests may do. Also, maybe it's better to inform people if input is
invalid.
After this:
> echo 123abc123 > scaling_min_freq
> -bash: echo: write error: Invalid argument
Signed-off-by: qinyu <qinyu32@huawei.com>
Tested-by: zhangxiaofeng <zhangxiaofeng46@huawei.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Direct access to the struct bus_type dev_root pointer is going away soon
so replace that with a call to bus_get_dev_root() instead, which is what
it is there for.
Cc: Viresh Kumar <viresh.kumar@linaro.org>
Cc: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Cc: Len Brown <lenb@kernel.org>
Cc: linux-pm@vger.kernel.org
Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Link: https://lore.kernel.org/r/20230313182918.1312597-3-gregkh@linuxfoundation.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Since commit ee6d3dd4ed ("driver core: make kobj_type constant.")
the driver core allows the usage of const struct kobj_type.
Take advantage of this to constify the structure definition to prevent
modification at runtime.
Signed-off-by: Thomas Weißschuh <linux@weissschuh.net>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
All but a few drivers ignore the return value of
cpufreq_unregister_driver(). Those few that don't only call it after
cpufreq_register_driver() succeeded, in which case the call doesn't
fail.
Make the function return no value and add a WARN_ON for the case that
the function is called in an invalid situation (i.e. without a previous
successful call to cpufreq_register_driver()).
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Acked-by: Florian Fainelli <f.fainelli@gmail.com> # brcmstb-avs-cpufreq.c
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
In cpufreq_policy_alloc(), it will call uninitialed completion in
cpufreq_sysfs_release() when kobject_init_and_add() fails. And
that will cause a crash such as the following page fault in complete:
BUG: unable to handle page fault for address: fffffffffffffff8
[..]
RIP: 0010:complete+0x98/0x1f0
[..]
Call Trace:
kobject_put+0x1be/0x4c0
cpufreq_online.cold+0xee/0x1fd
cpufreq_add_dev+0x183/0x1e0
subsys_interface_register+0x3f5/0x4e0
cpufreq_register_driver+0x3b7/0x670
acpi_cpufreq_init+0x56c/0x1000 [acpi_cpufreq]
do_one_initcall+0x13d/0x780
do_init_module+0x1c3/0x630
load_module+0x6e67/0x73b0
__do_sys_finit_module+0x181/0x240
do_syscall_64+0x35/0x80
entry_SYSCALL_64_after_hwframe+0x63/0xcd
Fixes: 4ebe36c94a ("cpufreq: Fix kobject memleak")
Signed-off-by: Yongqiang Liu <liuyongqiang13@huawei.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Cc: 5.2+ <stable@vger.kernel.org> # 5.2+
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
There is no need to check if the cpufreq driver implements callback
cpufreq_driver::target_index. The logic in the __resolve_freq uses
the frequency table available in the policy. It doesn't matter if the
driver provides 'target_index' or 'target' callback. It just has to
populate the 'policy->freq_table'.
Thus, check only frequency table during the frequency resolving call.
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Lukasz Luba <lukasz.luba@arm.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
- Fix return error code in mtk_cpu_dvfs_info_init (Yang Yingliang).
- Minor cleanups and support for new boards for Qcom cpufreq drivers
(Bryan O'Donoghue, Konrad Dybcio, Pierre Gondois, and Yicong Yang).
- Fix sparse warnings for Tegra driver (Viresh Kumar).
-----BEGIN PGP SIGNATURE-----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=fX64
-----END PGP SIGNATURE-----
Merge tag 'cpufreq-arm-updates-5.20-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/vireshk/pm
Pull cpufreq/ARM updates for 5.20-rc1 from Viresh Kumar:
"- Fix return error code in mtk_cpu_dvfs_info_init (Yang Yingliang).
- Minor cleanups and support for new boards for Qcom cpufreq drivers
(Bryan O'Donoghue, Konrad Dybcio, Pierre Gondois, and Yicong Yang).
- Fix sparse warnings for Tegra driver (Viresh Kumar)."
* tag 'cpufreq-arm-updates-5.20-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/vireshk/pm:
cpufreq: tegra194: Staticize struct tegra_cpufreq_soc instances
dt-bindings: cpufreq: cpufreq-qcom-hw: Add SM6375 compatible
dt-bindings: opp: Add msm8939 to the compatible list
dt-bindings: opp: Add missing compat devices
dt-bindings: opp: opp-v2-kryo-cpu: Fix example binding checks
cpufreq: Change order of online() CB and policy->cpus modification
cpufreq: qcom-hw: Remove deprecated irq_set_affinity_hint() call
cpufreq: qcom-hw: Disable LMH irq when disabling policy
cpufreq: qcom-hw: Reset cancel_throttle when policy is re-enabled
cpufreq: qcom-cpufreq-hw: use HZ_PER_KHZ macro in units.h
cpufreq: mediatek: fix error return code in mtk_cpu_dvfs_info_init()
From a state where all policy->related_cpus are offline, putting one
of the policy's CPU back online re-activates the policy by:
1. Calling cpufreq_driver->online()
2. Setting the CPU in policy->cpus
qcom_cpufreq_hw_cpu_online() makes use of policy->cpus. Thus 1. and 2.
should be inverted to avoid having a policy->cpus empty. The
qcom-cpufreq-hw is the only driver affected by this.
Signed-off-by: Pierre Gondois <pierre.gondois@arm.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
With the new design in place, the show() and store() callbacks check if
the policy is active or not before proceeding any further to avoid
potential races. And in order to guarantee that cpufreq_policy_free()
must be called after clearing the policy->cpus mask, i.e. by marking the
policy inactive.
In order to avoid introducing a bug around this later, print a warning
message if we end up freeing an active policy.
Also update cpufreq_online() a bit to make sure we clear the cpus mask
for each error case before calling cpufreq_policy_free().
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
This change was introduced long back by commit 4f750c9308 ("cpufreq:
Synchronize the cpufreq store_*() routines with CPU hotplug").
Since then, both cpufreq and hotplug core have been reworked and have
much better locking in place. The race mentioned in commit 4f750c9308
isn't possible anymore.
Drop the unnecessary locking.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Instead of specially adding a space for each CPU, except the first one,
lets add space for each of them and remove it at the end.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
cpufreq_offline() calls offline() and exit() under the policy rwsem
But they are called outside the rwsem in cpufreq_online().
Make cpufreq_online() call offline() and exit() as well as online() and
init() under the policy rwsem to achieve a clear lock relationship.
All of the init() and online() implementations in the tree only
initialize the policy object without attempting to acquire the policy
rwsem and they won't call cpufreq APIs attempting to acquire it.
Signed-off-by: Schspa Shi <schspa@gmail.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
[ rjw: Changelog edits ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
If policy initialization fails after the sysfs files are created,
there is a possibility to end up running show()/store() callbacks
for half-initialized policies, which may have unpredictable
outcomes.
Abort show()/store() in such a case by making sure the policy is active.
Also dectivate the policy on such failures.
Signed-off-by: Schspa Shi <schspa@gmail.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
[ rjw: Subject and changelog edits ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Currently, cpufreq_remove_dev() invokes the ->exit() driver callback
without holding the policy rwsem which is inconsistent with what
happens if ->exit() is invoked directly from cpufreq_offline().
It also manipulates the real_cpus mask and removes the CPU device
symlink without holding the policy rwsem, but cpufreq_offline() holds
the rwsem around the modifications thereof.
For consistency, modify cpufreq_remove_dev() to hold the policy rwsem
until the ->exit() callback has been called (or it has been determined
that it is not necessary to call it).
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Split the "core" part running under the policy rwsem out of
cpufreq_offline() to allow the locking in cpufreq_remove_dev() to be
rearranged more easily.
As a side-effect this eliminates the unlock label that's not needed
any more.
No expected functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Notice that cpufreq_offline() only needs to check policy_is_inactive()
once and rearrange the code in there to make that happen.
No expected functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
add_cpu_dev_symlink() is responsible for setting the CPUs in the
real_cpus mask, the reverse of which should be done from
remove_cpu_dev_symlink() to make it look clean and avoid any breakage
later on.
Move the call to clear the mask to remove_cpu_dev_symlink().
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
This reverts commit f346e96267.
The commit tried to fix a possible real bug but it made it even worse.
The fix was simply buggy as now an error out to out_offline_policy or
out_exit_policy will try to release a semaphore which was never taken in
the first place. This works fine only if we failed late, i.e. via
out_destroy_policy.
Fixes: f346e96267 ("cpufreq: Fix possible race in cpufreq online error path")
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
For some platforms, the frequency returned by hardware may be slightly
different from what is provided in the frequency table. For example,
hardware may return 499 MHz instead of 500 MHz. In such cases it is
better to avoid getting into unnecessary frequency updates, as we may
end up switching policy->cur between the two and sending unnecessary
pre/post update notifications, etc.
This patch has chosen allows the hardware frequency and table frequency
to deviate by 1 MHz for now, we may want to increase it a bit later on
if someone still complains.
Reported-by: Rex-BC Chen <rex-bc.chen@mediatek.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Tested-by: Jia-wei Chang <jia-wei.chang@mediatek.com>
Reviewed-by: Matthias Brugger <matthias.bgg@gmail.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
When cpufreq online fails, the policy->cpus mask is not cleared and
policy->rwsem is released too early, so the driver can be invoked
via the cpuinfo_cur_freq sysfs attribute while its ->offline() or
->exit() callbacks are being run.
Take policy->clk as an example:
static int cpufreq_online(unsigned int cpu)
{
...
// policy->cpus != 0 at this time
down_write(&policy->rwsem);
ret = cpufreq_add_dev_interface(policy);
up_write(&policy->rwsem);
return 0;
out_destroy_policy:
for_each_cpu(j, policy->real_cpus)
remove_cpu_dev_symlink(policy, get_cpu_device(j));
up_write(&policy->rwsem);
...
out_exit_policy:
if (cpufreq_driver->exit)
cpufreq_driver->exit(policy);
clk_put(policy->clk);
// policy->clk is a wild pointer
...
^
|
Another process access
__cpufreq_get
cpufreq_verify_current_freq
cpufreq_generic_get
// acces wild pointer of policy->clk;
|
|
out_offline_policy: |
cpufreq_policy_free(policy); |
// deleted here, and will wait for no body reference
cpufreq_policy_put_kobj(policy);
}
Address this by modifying cpufreq_online() to release policy->rwsem
in the error path after the driver callbacks have run and to clear
policy->cpus before releasing the semaphore.
Fixes: 7106e02bae ("cpufreq: release policy->rwsem on error")
Signed-off-by: Schspa Shi <schspa@gmail.com>
[ rjw: Subject and changelog edits ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
This effectively revert '4bf8e582119e ("cpufreq: Remove ready()
callback")', in order to reintroduce the ready callback.
This is needed in order to be able to leave the thermal pressure
interrupts in the Qualcomm CPUfreq driver disabled during
initialization, so that it doesn't fire while related_cpus are still 0.
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
[ Viresh: Added the Chinese translation as well and updated commit msg ]
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
There are currently 2 ways to create a set of sysfs files for a
kobj_type, through the default_attrs field, and the default_groups
field. Move the cpufreq code to use default_groups field which has been
the preferred way since aa30f47cf6 ("kobject: Add support for default
attribute groups to kobj_type") so that we can soon get rid of the
obsolete default_attrs field.
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
