[ Upstream commit 1fedbb5894 ]
According to the description in the intel_pstate.rst documentation,
Capacity-Aware Scheduling and Energy-Aware Scheduling are only
supported on a hybrid processor without SMT. Previously, the system
used sched_smt_active() for judgment, which is not a strict condition
because users can switch it on or off via /sys at any time.
This could lead to incorrect driver settings in certain scenarios.
For example, on a CPU that supports SMT, a user can disable SMT
via the nosmt parameter to enable asym capacity, and then re-enable
SMT via /sys. In such cases, some settings in the driver would no
longer be correct.
To address this issue, replace sched_smt_active() with cpu_smt_possible(),
and only enable asym capacity when CPU SMT is not possible.
Fixes: 929ebc93cc ("cpufreq: intel_pstate: Set asymmetric CPU capacity on hybrid systems")
Signed-off-by: Yaxiong Tian <tianyaxiong@kylinos.cn>
[ rjw: Subject and changelog edits ]
Link: https://patch.msgid.link/20260203024852.301066-1-tianyaxiong@kylinos.cn
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 0b7fbf9333 ]
When calling of_parse_phandle_with_args(), the caller is responsible
to call of_node_put() to release the reference of device node.
In scmi_cpu_domain_id(), it does not release the reference.
Fixes: e336baa419 ("cpufreq: scmi: Prepare to move OF parsing of domain-id to cpufreq")
Signed-off-by: Felix Gu <ustc.gu@gmail.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 4b747cc628 ]
Commit ac4e04d9e3 ("cpufreq: intel_pstate: Unchecked MSR aceess in
legacy mode") introduced a check for feature X86_FEATURE_IDA to verify
turbo mode support. Although this is the correct way to check for turbo
mode support, it causes issues on some platforms that disable turbo
during OS boot, but enable it later [1]. Before adding this feature
check, users were able to get turbo mode frequencies by writing 0 to
/sys/devices/system/cpu/intel_pstate/no_turbo post-boot.
To restore the old behavior on the affected systems while still
addressing the unchecked MSR issue on some Skylake-X systems, check
X86_FEATURE_IDA only immediately before updates of MSR_IA32_PERF_CTL
that may involve setting the Turbo Engage Bit (bit 32).
Fixes: ac4e04d9e3 ("cpufreq: intel_pstate: Unchecked MSR aceess in legacy mode")
Reported-by: Aaron Rainbolt <arainbolt@kfocus.org>
Closes: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/2122531 [1]
Tested-by: Aaron Rainbolt <arainbolt@kfocus.org>
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
[ rjw: Subject adjustment, changelog edits ]
Link: https://patch.msgid.link/20251111010840.141490-1-srinivas.pandruvada@linux.intel.com
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
[ richa: Backport to 6.12.y with context adjustments ]
Signed-off-by: Richa Bharti <richa.bharti@siemens.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 9600156bb9 upstream.
There are two reference count leaks in this driver:
1. In nforce2_fsb_read(): pci_get_subsys() increases the reference count
of the PCI device, but pci_dev_put() is never called to release it,
thus leaking the reference.
2. In nforce2_detect_chipset(): pci_get_subsys() gets a reference to the
nforce2_dev which is stored in a global variable, but the reference
is never released when the module is unloaded.
Fix both by:
- Adding pci_dev_put(nforce2_sub5) in nforce2_fsb_read() after reading
the configuration.
- Adding pci_dev_put(nforce2_dev) in nforce2_exit() to release the
global device reference.
Found via static analysis.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Cc: stable@vger.kernel.org
Signed-off-by: Miaoqian Lin <linmq006@gmail.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 2de5cb9606 ]
In function `s5pv210_cpu_init`, a possible refcount inconsistency has
been identified, causing a resource leak.
Why it is a bug:
1. For every clk_get, there should be a matching clk_put on every
successive error handling path.
2. After calling `clk_get(dmc1_clk)`, variable `dmc1_clk` will not be
freed even if any error happens.
How it is fixed: For every failed path, an extra goto label is added to
ensure `dmc1_clk` will be freed regardlessly.
Signed-off-by: Shuhao Fu <sfual@cse.ust.hk>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit bb31fef0d0 ]
amd_pstate_change_mode_without_dvr_change() calls cppc_set_auto_sel()
for all the present CPUs.
However, this callpath eventually calls cppc_set_reg_val() which
accesses the per-cpu cpc_desc_ptr object. This object is initialized
only for online CPUs via acpi_soft_cpu_online() -->
__acpi_processor_start() --> acpi_cppc_processor_probe().
Hence, restrict calling cppc_set_auto_sel() to only the online CPUs.
Fixes: 3ca7bc818d ("cpufreq: amd-pstate: Add guided mode control support via sysfs")
Suggested-by: Mario Limonciello (AMD) (kernel.org) <superm1@kernel.org>
Signed-off-by: Gautham R. Shenoy <gautham.shenoy@amd.com>
Signed-off-by: Mario Limonciello (AMD) <superm1@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit ba60189291 ]
During initialization, the EDVD_COREx_VOLT_FREQ registers for some cores
are still at reset values and not reflecting the actual frequency. This
causes get calls to fail. Set all cores to their respective max
frequency during probe to initialize the registers to working values.
Suggested-by: Mikko Perttunen <mperttunen@nvidia.com>
Signed-off-by: Aaron Kling <webgeek1234@gmail.com>
Reviewed-by: Mikko Perttunen <mperttunen@nvidia.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 7f3cfb7943 ]
IO time is considered busy by default for modern Intel processors. The
current check covers recent Family 6 models but excludes the brand new
Families 18 and 19.
According to Arjan van de Ven, the model check was mainly due to a lack
of testing on systems before INTEL_CORE2_MEROM. He suggests considering
all Intel processors as having an efficient idle.
Extend the IO busy classification to all Intel processors starting with
Family 6, including Family 15 (Pentium 4s) and upcoming Families 18/19.
Use an x86 VFM check and move the function to the header file to avoid
using arch-specific #ifdefs in the C file.
Signed-off-by: Sohil Mehta <sohil.mehta@intel.com>
Link: https://patch.msgid.link/20250908230655.2562440-1-sohil.mehta@intel.com
[ rjw: Added empty line after #include ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 592532a77b ]
longhaul_exit() was calling cpufreq_cpu_get(0) without checking
for a NULL policy pointer. On some systems, this could lead to a
NULL dereference and a kernel warning or panic.
This patch adds a check using unlikely() and returns early if the
policy is NULL.
Bugzilla: #219962
Signed-off-by: Dennis Beier <nanovim@gmail.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit b5af45302e ]
Add support for TI K3 AM62D2 SoC to read speed and revision values
from hardware and pass to OPP layer. AM62D shares the same configuations
as AM62A so use existing am62a7_soc_data.
Signed-off-by: Paresh Bhagat <p-bhagat@ti.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit f965d111e6 ]
If cppc_get_transition_latency() returns CPUFREQ_ETERNAL to indicate a
failure to retrieve the transition latency value from the platform
firmware, the CPPC cpufreq driver will use that value (converted to
microseconds) as the policy transition delay, but it is way too large
for any practical use.
Address this by making the driver use the cpufreq's default
transition latency value (in microseconds) as the transition delay
if CPUFREQ_ETERNAL is returned by cppc_get_transition_latency().
Fixes: d4f3388afd ("cpufreq / CPPC: Set platform specific transition_delay_us")
Cc: 5.19+ <stable@vger.kernel.org> # 5.19
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Mario Limonciello (AMD) <superm1@kernel.org>
Reviewed-by: Jie Zhan <zhanjie9@hisilicon.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Qais Yousef <qyousef@layalina.io>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit f97aef092e ]
Commit a755d0e2d4 ("cpufreq: Honour transition_latency over
transition_delay_us") caused platforms where cpuinfo.transition_latency
is CPUFREQ_ETERNAL to get a very large transition latency whereas
previously it had been capped at 10 ms (and later at 2 ms).
This led to a user-observable regression between 6.6 and 6.12 as
described by Shawn:
"The dbs sampling_rate was 10000 us on 6.6 and suddently becomes
6442450 us (4294967295 / 1000 * 1.5) on 6.12 for these platforms
because the default transition delay was dropped [...].
It slows down dbs governor's reacting to CPU loading change
dramatically. Also, as transition_delay_us is used by schedutil
governor as rate_limit_us, it shows a negative impact on device
idle power consumption, because the device gets slightly less time
in the lowest OPP."
Evidently, the expectation of the drivers using CPUFREQ_ETERNAL as
cpuinfo.transition_latency was that it would be capped by the core,
but they may as well return a default transition latency value instead
of CPUFREQ_ETERNAL and the core need not do anything with it.
Accordingly, introduce CPUFREQ_DEFAULT_TRANSITION_LATENCY_NS and make
all of the drivers in question use it instead of CPUFREQ_ETERNAL. Also
update the related Rust binding.
Fixes: a755d0e2d4 ("cpufreq: Honour transition_latency over transition_delay_us")
Closes: https://lore.kernel.org/linux-pm/20250922125929.453444-1-shawnguo2@yeah.net/
Reported-by: Shawn Guo <shawnguo@kernel.org>
Reviewed-by: Mario Limonciello (AMD) <superm1@kernel.org>
Reviewed-by: Jie Zhan <zhanjie9@hisilicon.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Cc: 6.6+ <stable@vger.kernel.org> # 6.6+
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Link: https://patch.msgid.link/2264949.irdbgypaU6@rafael.j.wysocki
[ rjw: Fix typo in new symbol name, drop redundant type cast from Rust binding ]
Tested-by: Shawn Guo <shawnguo@kernel.org> # with cpufreq-dt driver
Reviewed-by: Qais Yousef <qyousef@layalina.io>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
[ omitted Rust changes ]
Signed-off-by: Sasha Levin <sashal@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 69e5d50fcf upstream.
The cpufreq_cpu_put() call in update_qos_request() takes place too early
because the latter subsequently calls freq_qos_update_request() that
indirectly accesses the policy object in question through the QoS request
object passed to it.
Fortunately, update_qos_request() is called under intel_pstate_driver_lock,
so this issue does not matter for changing the intel_pstate operation
mode, but it theoretically can cause a crash to occur on CPU device hot
removal (which currently can only happen in virt, but it is formally
supported nevertheless).
Address this issue by modifying update_qos_request() to drop the
reference to the policy later.
Fixes: da5c504c7a ("cpufreq: intel_pstate: Implement QoS supported freq constraints")
Cc: 5.4+ <stable@vger.kernel.org> # 5.4+
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Zihuan Zhang <zhangzihuan@kylinos.cn>
Link: https://patch.msgid.link/2255671.irdbgypaU6@rafael.j.wysocki
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 0b1bb980fd ]
The original commit set all cores in a cluster to a shared policy, but
did not update set_target to apply a frequency change to all cores for
the policy. This caused most cores to remain stuck at their boot
frequency.
Fixes: be4ae8c194 ("cpufreq: tegra186: Share policy per cluster")
Signed-off-by: Aaron Kling <webgeek1234@gmail.com>
Reviewed-by: Mikko Perttunen <mperttunen@nvidia.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit cd5d4621ba ]
Broadcom STB platforms were early adopters (2017) of the SCMI framework and as
a result, not all deployed systems have a Device Tree entry where SCMI
protocol 0x13 (PERFORMANCE) is declared as a clock provider, nor are the
CPU Device Tree node(s) referencing protocol 0x13 as their clock
provider. This was clarified in commit e11c480b6d ("dt-bindings:
firmware: arm,scmi: Extend bindings for protocol@13") in 2023.
For those platforms, we allow the checks done by scmi_dev_used_by_cpus()
to continue, and in the event of not having done an early return, we key
off the documented compatible string and give them a pass to continue to
use scmi-cpufreq.
Fixes: 6c9bb86922 ("cpufreq: scmi: Skip SCMI devices that aren't used by the CPUs")
Signed-off-by: Florian Fainelli <florian.fainelli@broadcom.com>
Reviewed-by: Sudeep Holla <sudeep.holla@arm.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 8ffe28b4e8 ]
commit 2a6c727387 ("cpufreq: Initialize cpufreq-based
frequency-invariance later") postponed the frequency invariance
initialization to avoid disabling it in the error case.
This isn't locking safe, instead move the initialization up before
the subsys interface is registered (which will rebuild the
sched_domains) and add the corresponding disable on the error path.
Observed lockdep without this patch:
[ 0.989686] ======================================================
[ 0.989688] WARNING: possible circular locking dependency detected
[ 0.989690] 6.17.0-rc4-cix-build+ #31 Tainted: G S
[ 0.989691] ------------------------------------------------------
[ 0.989692] swapper/0/1 is trying to acquire lock:
[ 0.989693] ffff800082ada7f8 (sched_energy_mutex){+.+.}-{4:4}, at: rebuild_sched_domains_energy+0x30/0x58
[ 0.989705]
but task is already holding lock:
[ 0.989706] ffff000088c89bc8 (&policy->rwsem){+.+.}-{4:4}, at: cpufreq_online+0x7f8/0xbe0
[ 0.989713]
which lock already depends on the new lock.
Fixes: 2a6c727387 ("cpufreq: Initialize cpufreq-based frequency-invariance later")
Signed-off-by: Christian Loehle <christian.loehle@arm.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit 4a26df2332 upstream.
The freq_tables[] array has num_possible_cpus() elements so, to avoid an
out of bounds access, this loop should be capped at "< nb_cpus" instead
of "<= nb_cpus". The freq_tables[] array is allocated in
armada_8k_cpufreq_init().
Cc: stable@vger.kernel.org
Fixes: f525a67053 ("cpufreq: ap806: add cpufreq driver for Armada 8K")
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 0a1416a49e ]
AMU counters on certain CPPC-based platforms tend to yield inaccurate
delivered performance measurements on systems that are idle/mostly idle.
This results in an inaccurate frequency being stored by cpufreq in its
policy structure when the CPU is brought online. [1]
Consequently, if the userspace governor tries to set the frequency to a
new value, there is a possibility that it would be the erroneous value
stored earlier. In such a scenario, cpufreq would assume that the
requested frequency has already been set and return early, resulting in
the correct/new frequency request never making it to the hardware.
Since the operating frequency is liable to this sort of inconsistency,
mark the CPPC driver with CPUFREQ_NEED_UPDATE_LIMITS so that it is always
invoked when a target frequency update is requested.
Link: https://lore.kernel.org/linux-pm/20250619000925.415528-3-pmalani@google.com/ [1]
Suggested-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Prashant Malani <pmalani@google.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Link: https://patch.msgid.link/20250722055611.130574-2-pmalani@google.com
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit d1378d1d7e ]
In cpufreq_policy_put_kobj(), policy->rwsem is used. But in
cpufreq_policy_alloc(), if freq_qos_add_notifier() returns an error, error
path via err_kobj_remove or err_min_qos_notifier will be reached and
cpufreq_policy_put_kobj() will be called before policy->rwsem is
initialized. Thus, the calling of init_rwsem() should be moved to where
before these two error paths can be reached.
Fixes: 67d874c3b2 ("cpufreq: Register notifiers with the PM QoS framework")
Signed-off-by: Lifeng Zheng <zhenglifeng1@huawei.com>
Link: https://patch.msgid.link/20250709104145.2348017-3-zhenglifeng1@huawei.com
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 1cefe495ca ]
In the passive mode, intel_cpufreq_update_pstate() sets HWP_MIN_PERF in
accordance with the target frequency to ensure delivering adequate
performance, but it sets HWP_DESIRED_PERF to 0, so the processor has no
indication that the desired performance level is actually equal to the
floor one. This may cause it to choose a performance point way above
the desired level.
Moreover, this is inconsistent with intel_cpufreq_adjust_perf() which
actually sets HWP_DESIRED_PERF in accordance with the target performance
value.
Address this by adjusting intel_cpufreq_update_pstate() to pass
target_pstate as both the minimum and the desired performance levels
to intel_cpufreq_hwp_update().
Fixes: a365ab6b9d ("cpufreq: intel_pstate: Implement the ->adjust_perf() callback")
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Tested-by: Shashank Balaji <shashank.mahadasyam@sony.com>
Link: https://patch.msgid.link/6173276.lOV4Wx5bFT@rjwysocki.net
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit b1b41bc072 ]
An earlier patch marked one of the two CPU masks as 'static' to reduce stack
usage, but if CONFIG_NR_CPUS is large enough, the function still produces
a warning for compile testing:
drivers/cpufreq/armada-8k-cpufreq.c: In function 'armada_8k_cpufreq_init':
drivers/cpufreq/armada-8k-cpufreq.c:203:1: error: the frame size of 1416 bytes is larger than 1408 bytes [-Werror=frame-larger-than=]
Normally this should be done using alloc_cpumask_var(), but since the
driver already has a static mask and the probe function is not called
concurrently, use the same trick for both.
Fixes: 1ffec650d0 ("cpufreq: armada-8k: Avoid excessive stack usage")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 6c9bb86922 ]
Currently, all SCMI devices with performance domains attempt to register
a cpufreq driver, even if their performance domains aren't used to
control the CPUs. The cpufreq framework only supports registering a
single driver, so only the first device will succeed. And if that device
isn't used for the CPUs, then cpufreq will scale the wrong domains.
To avoid this, return early from scmi_cpufreq_probe() if the probing
SCMI device isn't referenced by the CPU device phandles.
This keeps the existing assumption that all CPUs are controlled by a
single SCMI device.
Signed-off-by: Mike Tipton <quic_mdtipton@quicinc.com>
Reviewed-by: Peng Fan <peng.fan@nxp.com>
Reviewed-by: Cristian Marussi <cristian.marussi@arm.com>
Reviewed-by: Sudeep Holla <sudeep.holla@arm.com>
Tested-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit cb6a85f38f upstream.
commit 0834667545 ("cpufreq: ACPI: Fix max-frequency computation")
modified get_max_boost_ratio() to return the nominal_freq advertised
in the _CPC object. This was for the purposes of computing the maximum
frequency. The frequencies advertised in _CPC objects are in
MHz. However, cpufreq expects the frequency to be in KHz. Since the
nominal_freq returned by get_max_boost_ratio() was not in KHz but
instead in MHz,the cpuinfo_max_frequency that was computed using this
nominal_freq was incorrect and an invalid value which resulted in
cpufreq reporting the P0 frequency as the cpuinfo_max_freq.
Fix this by converting the nominal_freq to KHz before returning the
same from get_max_boost_ratio().
Reported-by: Manu Bretelle <chantr4@gmail.com>
Closes: https://lore.kernel.org/lkml/aDaB63tDvbdcV0cg@HQ-GR2X1W2P57/
Fixes: 0834667545 ("cpufreq: ACPI: Fix max-frequency computation")
Signed-off-by: Gautham R. Shenoy <gautham.shenoy@amd.com>
Cc: 6.14+ <stable@vger.kernel.org> # 6.14+
Link: https://patch.msgid.link/20250529085143.709-1-gautham.shenoy@amd.com
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit db1cafc77a ]
set_boost is a per-policy function call, hence a driver wide lock is
unnecessary. Also this mutex_acquire can collide with the mutex_acquire
from the mode-switch path in status_store(), which can lead to a
deadlock. So, remove it.
Signed-off-by: Dhananjay Ugwekar <dhananjay.ugwekar@amd.com>
Acked-by: Mario Limonciello <mario.limonciello@amd.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit fc5414a477 ]
SM8650 have already been supported by qcom-cpufreq-hw driver, but
never been added to cpufreq-dt-platdev. This makes noise
[ 0.388525] cpufreq-dt cpufreq-dt: failed register driver: -17
[ 0.388537] cpufreq-dt cpufreq-dt: probe with driver cpufreq-dt failed with error -17
So adding it to the cpufreq-dt-platdev driver's blocklist to fix it.
Signed-off-by: Pengyu Luo <mitltlatltl@gmail.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit 7491cdf46b upstream.
Since cpufreq_driver_resolve_freq() can run in parallel with
cpufreq_set_policy() and there is no synchronization between them,
the former may access policy->min and policy->max while the latter
is updating them and it may see intermediate values of them due
to the way the update is carried out. Also the compiler is free
to apply any optimizations it wants both to the stores in
cpufreq_set_policy() and to the loads in cpufreq_driver_resolve_freq()
which may result in additional inconsistencies.
To address this, use WRITE_ONCE() when updating policy->min and
policy->max in cpufreq_set_policy() and use READ_ONCE() for reading
them in cpufreq_driver_resolve_freq(). Moreover, rearrange the update
in cpufreq_set_policy() to avoid storing intermediate values in
policy->min and policy->max with the help of the observation that
their new values are expected to be properly ordered upfront.
Also modify cpufreq_driver_resolve_freq() to take the possible reverse
ordering of policy->min and policy->max, which may happen depending on
the ordering of operations when this function and cpufreq_set_policy()
run concurrently, into account by always honoring the max when it
turns out to be less than the min (in case it comes from thermal
throttling or similar).
Fixes: 1517176906 ("cpufreq: Make policy min/max hard requirements")
Cc: 5.16+ <stable@vger.kernel.org> # 5.16+
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Christian Loehle <christian.loehle@arm.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Link: https://patch.msgid.link/5907080.DvuYhMxLoT@rjwysocki.net
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit ac4e04d9e3 upstream.
When turbo mode is unavailable on a Skylake-X system, executing the
command:
# echo 1 > /sys/devices/system/cpu/intel_pstate/no_turbo
results in an unchecked MSR access error:
WRMSR to 0x199 (attempted to write 0x0000000100001300).
This issue was reproduced on an OEM (Original Equipment Manufacturer)
system and is not a common problem across all Skylake-X systems.
This error occurs because the MSR 0x199 Turbo Engage Bit (bit 32) is set
when turbo mode is disabled. The issue arises when intel_pstate fails to
detect that turbo mode is disabled. Here intel_pstate relies on
MSR_IA32_MISC_ENABLE bit 38 to determine the status of turbo mode.
However, on this system, bit 38 is not set even when turbo mode is
disabled.
According to the Intel Software Developer's Manual (SDM), the BIOS sets
this bit during platform initialization to enable or disable
opportunistic processor performance operations. Logically, this bit
should be set in such cases. However, the SDM also specifies that "OS
and applications must use CPUID leaf 06H to detect processors with
opportunistic processor performance operations enabled."
Therefore, in addition to checking MSR_IA32_MISC_ENABLE bit 38, verify
that CPUID.06H:EAX[1] is 0 to accurately determine if turbo mode is
disabled.
Fixes: 4521e1a0ce ("cpufreq: intel_pstate: Reflect current no_turbo state correctly")
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Cc: All applicable <stable@vger.kernel.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit a374f28700 ]
Commit 3f66425a4f ("cpufreq: Enable COMPILE_TEST on Arm drivers")
enabled compile testing of most Arm CPUFreq drivers but left the
existing default values unchanged so that many drivers are enabled by
default whenever COMPILE_TEST is selected.
This specifically results in the S3C64XX CPUFreq driver being enabled
and initialised during boot of non-S3C64XX platforms with the following
error logged:
cpufreq: Unable to obtain ARMCLK: -2
Commit d4f610a9ba ("cpufreq: Do not enable by default during compile
testing") recently fixed most of the default values, but two entries
were missed and two could use a more specific default condition.
Fix the default values for drivers that can be compile tested and that
should be enabled by default when not compile testing.
Fixes: 3f66425a4f ("cpufreq: Enable COMPILE_TEST on Arm drivers")
Cc: Rob Herring (Arm) <robh@kernel.org>
Signed-off-by: Johan Hovold <johan+linaro@kernel.org>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 2b8e6b5888 ]
Returning a negative error code in a function with an unsigned
return type is a pretty bad idea. It is probably worse when the
justification for the change is "our static analisys tool found it".
Fixes: cf7de25878 ("cppc_cpufreq: Fix possible null pointer dereference")
Signed-off-by: Marc Zyngier <maz@kernel.org>
Cc: "Rafael J. Wysocki" <rafael@kernel.org>
Cc: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Lifeng Zheng <zhenglifeng1@huawei.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 73b24dc731 ]
cpufreq_cpu_get_raw() can return NULL when the target CPU is not present
in the policy->cpus mask. scpi_cpufreq_get_rate() does not check for
this case, which results in a NULL pointer dereference.
Fixes: 343a8d17fa ("cpufreq: scpi: remove arm_big_little dependency")
Signed-off-by: Henry Martin <bsdhenrymartin@gmail.com>
Acked-by: Sudeep Holla <sudeep.holla@arm.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 484d3f15cc ]
cpufreq_cpu_get_raw() can return NULL when the target CPU is not present
in the policy->cpus mask. scmi_cpufreq_get_rate() does not check for
this case, which results in a NULL pointer dereference.
Add NULL check after cpufreq_cpu_get_raw() to prevent this issue.
Fixes: 99d6bdf338 ("cpufreq: add support for CPU DVFS based on SCMI message protocol")
Signed-off-by: Henry Martin <bsdhenrymartin@gmail.com>
Acked-by: Sudeep Holla <sudeep.holla@arm.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 9992649f67 ]
cpufreq_cpu_get_raw() can return NULL when the target CPU is not present
in the policy->cpus mask. apple_soc_cpufreq_get_rate() does not check
for this case, which results in a NULL pointer dereference.
Fixes: 6286bbb405 ("cpufreq: apple-soc: Add new driver to control Apple SoC CPU P-states")
Signed-off-by: Henry Martin <bsdhenrymartin@gmail.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 14c8a41815 ]
A KASAN enabled kernel reports an out-of-bounds access when handling the
nvmem cell in the sun50i cpufreq driver:
==================================================================
BUG: KASAN: slab-out-of-bounds in sun50i_cpufreq_nvmem_probe+0x180/0x3d4
Read of size 4 at addr ffff000006bf31e0 by task kworker/u16:1/38
This is because the DT specifies the nvmem cell as covering only two
bytes, but we use a u32 pointer to read the value. DTs for other SoCs
indeed specify 4 bytes, so we cannot just shorten the variable to a u16.
Fortunately nvmem_cell_read() allows to return the length of the nvmem
cell, in bytes, so we can use that information to only access the valid
portion of the data.
To cover multiple cell sizes, use memcpy() to copy the information into a
zeroed u32 buffer, then also make sure we always read the data in little
endian fashion, as this is how the data is stored in the SID efuses.
Fixes: 6cc4bcceff ("cpufreq: sun50i: Refactor speed bin decoding")
Reported-by: Jernej Skrabec <jernej.skrabec@gmail.com>
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Jernej Škrabec <jernej.skrabec@gmail.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 4a1e3bf61f ]
Support was added for Tegra234 in the referenced commit, but the Kconfig
was not updated to allow building for the arch.
Fixes: 273bc890a2 ("cpufreq: tegra194: Add support for Tegra234")
Signed-off-by: Aaron Kling <webgeek1234@gmail.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 3698dd6b13 ]
We observed an issue that the CPU frequency can't raise up with a 100% CPU
load when NOHZ is off and the 'conservative' governor is selected.
'idle_time' can be negative if it's obtained from get_cpu_idle_time_jiffy()
when NOHZ is off. This was found and explained in commit 9485e4ca0b
("cpufreq: governor: Fix handling of special cases in dbs_update()").
However, commit 7592019634 ("cpufreq: governors: Fix long idle detection
logic in load calculation") introduced a comparison between 'idle_time' and
'samling_rate' to detect a long idle interval. While 'idle_time' is
converted to int before comparison, it's actually promoted to unsigned
again when compared with an unsigned 'sampling_rate'. Hence, this leads to
wrong idle interval detection when it's in fact 100% busy and sets
policy_dbs->idle_periods to a very large value. 'conservative' adjusts the
frequency to minimum because of the large 'idle_periods', such that the
frequency can't raise up. 'Ondemand' doesn't use policy_dbs->idle_periods
so it fortunately avoids the issue.
Correct negative 'idle_time' to 0 before any use of it in dbs_update().
Fixes: 7592019634 ("cpufreq: governors: Fix long idle detection logic in load calculation")
Signed-off-by: Jie Zhan <zhanjie9@hisilicon.com>
Reviewed-by: Chen Yu <yu.c.chen@intel.com>
Link: https://patch.msgid.link/20250213035510.2402076-1-zhanjie9@hisilicon.com
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 4742da9774 ]
The CPU rate from clk_get_rate() may not be divisible by 1000
(e.g., 133333333). But the rate calculated from frequency(kHz) is
always divisible by 1000 (e.g., 133333000).
Comparing the rate causes a warning during CPU scaling:
"cpufreq: __target_index: Failed to change cpu frequency: -5".
When we choose to compare kHz here, the issue does not occur.
Fixes: 343a8d17fa ("cpufreq: scpi: remove arm_big_little dependency")
Signed-off-by: zuoqian <zuoqian113@gmail.com>
Reviewed-by: Dan Carpenter <dan.carpenter@linaro.org>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
