The legacy hwmon_device_register() function is deprecated and triggers
warnings in dmesg. Convert the driver to the modern
hwmon_device_register_with_info() API.
This conversion does the following:
- Dynamically allocates standard HWMON temp, fan, and pwm channels.
- Configures HWMON ops callbacks (.is_visible, .read, .read_string, .write).
- Standardizes attribute naming to match the HWMON ABI:
- fanX_output -> fanX_target (HWMON_F_TARGET)
- fanX_manual -> pwmX_enable (HWMON_PWM_ENABLE)
- Dynamically registers non-standard fanX_safe attributes under the HWMON
class directory via extra_groups.
- Cleans up legacy sysfs nodes, groups, and unused show/store static functions
to avoid unused symbol compiler warnings.
- Avoids recursive mutex deadlocks when writing to pwmX_enable by locklessly
resolving the entry and invoking the underlying raw SMC read/write calls.
- Avoids UAF race condition on module exit by using unmanaged registration and
explicitly calling hwmon_device_unregister() as the first step of applesmc_exit(),
guaranteeing that HWMON nodes are destroyed before static structures are freed.
Signed-off-by: Shih-Yuan Lee <fourdollars@debian.org>
Link: https://lore.kernel.org/r/20260711093323.14529-4-fourdollars@debian.org
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
In applesmc_get_entry_by_index(), the cache->valid flag is checked
locklessly, but setting it to true lacks memory barriers. This can lead to
a data race (TOCTOU) where another thread sees cache->valid as true
before the actual cache contents (cache->key, cache->len, cache->type, etc.)
are fully committed and visible to that CPU, potentially causing it to read
uninitialized data and send incorrect keys to the Apple SMC hardware.
Introduce memory barriers (smp_load_acquire and smp_store_release) with
explanatory comments to ensure cache synchronization is thread-safe and
fully visible across all CPUs.
Signed-off-by: Shih-Yuan Lee <fourdollars@debian.org>
Link: https://lore.kernel.org/r/20260711093323.14529-3-fourdollars@debian.org
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
To support the read_string callback for fan labels in the modern HWMON
API, load and cache the fan position names in smcreg.fan_positions
during register initialization.
Pre-pad fallback labels with four spaces to match the "+ 4" pointer
arithmetic offset used by all fan labels in the read_string callback.
Signed-off-by: Shih-Yuan Lee <fourdollars@debian.org>
Link: https://lore.kernel.org/r/20260711093323.14529-2-fourdollars@debian.org
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
The driver has a match table for the i2c bus wired into its driver
structure, but the table is not exported with MODULE_DEVICE_TABLE().
Add the missing MODULE_DEVICE_TABLE() entry so module alias information
is generated for automatic module loading.
This is a source-level fix. It does not claim dynamic hardware
reproduction; the evidence is the driver-owned match table, its use by
the driver registration structure, and the missing module alias
publication.
Signed-off-by: Pengpeng Hou <pengpeng@iscas.ac.cn>
Link: https://lore.kernel.org/r/20260704151543.35069-1-pengpeng@iscas.ac.cn
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
scripts/kernel-doc -Wall -none reports:
Warning: drivers/hwmon/peci/common.h:39 function parameter 'state' not
described in 'peci_sensor_need_update'
Warning: drivers/hwmon/peci/common.h:39 Excess function parameter
'sensor' description in 'peci_sensor_need_update'
Warning: drivers/hwmon/peci/common.h:49 function parameter 'state' not
described in 'peci_sensor_mark_updated'
Warning: drivers/hwmon/peci/common.h:49 Excess function parameter
'sensor' description in 'peci_sensor_mark_updated'
Both peci_sensor_need_update() and peci_sensor_mark_updated() take a
pointer to struct peci_sensor_state named 'state', but their kernel-doc
comments document a parameter named 'sensor' and describe it as a
pointer to the sensor data struct. The comments have been out of sync
with the code since the file was added by commit bf3608f338 ("hwmon:
peci: Add cputemp driver").
Update the kernel-doc comments to document the 'state' parameter.
No functional change.
Assisted-by: Claude:claude-sonnet-5
Signed-off-by: Babanpreet Singh <bbnpreetsingh@gmail.com>
Link: https://lore.kernel.org/r/20260712081258.7-1-bbnpreetsingh@gmail.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
The GXCAS GXHT30 is a humidity and temperature sensor that is a
drop-in replacement for and fully software compatible with the
Sensirion SHT30.
It is handled by the "sensirion,sht30" fallback compatible without
any driver changes; manual instantiation works with the existing
sht3x device name. Document the chip as supported.
Signed-off-by: Zaixiang Xu <zaixiang.xu.dev@gmail.com>
Link: https://lore.kernel.org/r/20260713074559.12196-5-zaixiang.xu.dev@gmail.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Add an of_device_id table to support devicetree based instantiation
of the supported sensors.
The devicetree binding models all higher accuracy parts with a
fallback to the base part of their group, so only the two base
compatibles need to be listed. The match data distinguishes the
humidity and temperature parts (SHT3x) from the temperature-only
parts (STS3x), which require different configuration.
Start enum sht3x_chips at 1: with sht3x equal to 0, the OF match
data for the SHT devices would be NULL, causing i2c_get_match_data()
to fall back to i2c_device_id name matching, which fails for
devicetree names. The devices would then only work because the
resulting chip id 0 happens to equal sht3x. Non-zero match data
avoids relying on that coincidence.
Signed-off-by: Zaixiang Xu <zaixiang.xu.dev@gmail.com>
Link: https://lore.kernel.org/r/20260713074559.12196-4-zaixiang.xu.dev@gmail.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Add a YAML devicetree binding for the Sensirion SHT3x/STS3x sensor
family and the compatible GXCAS GXHT30, with the ALERT interrupt,
nRESET GPIO and VDD supply as optional properties.
The SHT30, SHT31, SHT35 and SHT85 measure humidity and temperature
and differ only in accuracy and packaging, while the STS30, STS31,
STS32 and STS35 are temperature-only parts. Within each group the
parts share the same software interface, so model them with a
fallback to the base part of their group: "sensirion,sht30" for the
SHT parts and "sensirion,sts30" for the STS parts.
The GXCAS GXHT30 is a drop-in replacement for the SHT30 and uses it
as fallback as well.
Signed-off-by: Zaixiang Xu <zaixiang.xu.dev@gmail.com>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@oss.qualcomm.com>
Link: https://lore.kernel.org/r/20260713074559.12196-3-zaixiang.xu.dev@gmail.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Enable T_Sensor for ROG STRIX X870E-E GAMING WIFI.
Board specs list the sensor, reading at the standard for this family
address without the thermistor connected to the motherboard, returns
-62, which is one of the known blank values (the other one in -40)
readings for temperature sensors in the ASUS EC when there is no
thermistor connected to the motherboard socket. With that evidence we
conclude that the T_Sensor can be enabled in the driver for this board.
Signed-off-by: Vishaal Saraiya <vishaal.saraiya@gmail.com>
Signed-off-by: Eugene Shalygin <eugene.shalygin@gmail.com>
Link: https://lore.kernel.org/r/20260711073319.553163-2-eugene.shalygin@gmail.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
tmp401 is missing HWMON_C_REGISTER_TZ. So a devicetree thermal
zone referencing a such sensor ('thermal-sensors = <&tmp411 N>')
fails to register it. Then its cooling-maps is not set, which means
that the fans are not managed.
Same as lm90 or jc42, set HWMON_C_REGISTER_TZ.
Boards without the property into the DT keep the same behavior.
Signed-off-by: Vincent Jardin <vjardin@free.fr>
Link: https://lore.kernel.org/r/20260713-for-upstream-hwmon-tmp401-register-tz-v1-1-47315d8617bc@free.fr
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Add the Lenovo LOQ 15IAX9 product family to the yogafan dmi_system_id
table. Unlike the Legion series, ACPI table extraction reveals this LOQ
utilizes an 8-bit EC architecture with a 100x multiplier, but maintains
a dual-fan setup via the \_SB.PC00.LPCB.EC0.FA1S and FA2S paths.
A new loq_15iax9_8bit_dual_cfg structure was added to properly map these
hardware registers and calculate RPMs correctly while avoiding path
conflicts with AMD-based LOQ variants.
Signed-off-by: Sarbajit Sarkar <sarbajitsarkar16108@gmail.com>
Link: https://lore.kernel.org/r/20260709183500.13761-1-sarbajitsarkar16108@gmail.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
The driver populates acpi_device_name() and acpi_device_class() that are
never read afterward, so make it stop doing that and drop the symbols
defined for this purpose.
Since ACPI_POWER_METER_CLASS is also used for generating ACPI netlink
events, use the "pwr_meter_resource" string literal directly instead of
it for that.
No intentional functional impact.
This will facilitate the removal of device_name and device_class from
struct acpi_device_pnp in the future.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Link: https://lore.kernel.org/r/2297496.irdbgypaU6@rafael.j.wysocki
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
The ChromeOS EC APIs use integers representing degrees kelvin for
temperatures. The default conversions from linux/units.h will then
always convert these integer degrees celsius with a 150 millidegree
offset. This is a bit confusing, as it also differs from other CrOS EC
tooling. Internally the EC uses a kelvin to celsius offset of a round
273, so the current conversion is also not entirely accurate.
Implement a custom conversion which preserves round values.
Signed-off-by: Thomas Weißschuh <linux@weissschuh.net>
Link: https://lore.kernel.org/r/20260630-cros_ec-hwmon-overflow-v1-1-3d2ecd3eb0f2@weissschuh.net
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Add a driver for the temperature and voltage sensors on PolarFire SoC.
The temperature reports how hot the die is, and the voltages are the
SoC's 1.05, 1.8 and 2.5 volt rails respectively.
The hardware supports alarms in theory, but there is an erratum that
prevents clearing them once triggered, so no support is added for them.
The hardware measures voltage with 16 bits, of which 1 is a sign bit and
the remainder holds the voltage as a fixed point integer value. It's
improbable that the hardware will work if the voltages are negative, so
the driver ignores the sign bits.
There's no dt support etc here because this is the child of a simple-mfd
syscon.
Signed-off-by: Lars Randers <lranders@mail.dk>
Co-developed-by: Conor Dooley <conor.dooley@microchip.com>
Signed-off-by: Conor Dooley <conor.dooley@microchip.com>
Link: https://lore.kernel.org/r/20260629-wriggle-headscarf-c85a4070dddb@spud
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Add support for ROHM BD12790 hot-swap controller which is largely
similar to Analog Devices adm1272.
The BD12790 uses the same selectable 60V/100V voltage ranges and
15mV/30mV current-sense ranges as the ADM1272, and the same VRANGE
(bit 5) and IRANGE (bit 0) layout in PMON_CONFIG. It therefore uses
a dedicated coefficient table that mirrors adm1272_coefficients, with
the following differences derived from BD12790 datasheet Table 1 (p.18):
- power 60V/30mV: m=17560 (vs. 17561)
- power 100V/30mV: m=10536 (vs. 10535)
- temperature: b=31880 (vs. 31871, reflecting T[11:0] = 4.2*T + 3188)
Signed-off-by: Matti Vaittinen <mazziesaccount@gmail.com>
Assisted-by: GitHub Copilot:claude-sonnet-4.6
Link: https://lore.kernel.org/r/b209c1b47712b69f17b52cfd7a7a38ed76024ca7.1782458224.git.mazziesaccount@gmail.com
[groeck: Fixed comment describing BD12790 coefficients]
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
ROHM BD12780 and BD12780A are hot-swap controllers. They are largely
similar to Analog Devices ADM1278. Besides the ID registers and some
added functionality, the BD12780 and BD12780A mark PMON_CONFIG bits
[15:14] as reserved. Hence TSFILT setting must be omitted on these ICs.
The BD12780 has 3 pins usable for configuring the I2C address. The
BD12780A lists the ADDR3-pin as "not connect".
Support ROHM BD12780 and BD12780A controllers.
Signed-off-by: Matti Vaittinen <mazziesaccount@gmail.com>
Link: https://lore.kernel.org/r/e1e6e6218c08b562311356ef9c57378d32c26b08.1782458224.git.mazziesaccount@gmail.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Populating the i2c_device_id -table is not enough to make the
driver module automatically load when device-tree node for the
power-monitor is parsed at boot.
Adding the of_device_id tables causes the driver module to be
automatically load at boot. Testing has been done with rather old Debian
system.
When inspecting the generated module-aliases with the insmod, following
entries seem to be the difference:
alias: of:N*T*Cadi,adm1075C*
alias: of:N*T*Cadi,adm1075
I suspect these are required for the module loading to work.
Signed-off-by: Matti Vaittinen <mazziesaccount@gmail.com>
Link: https://lore.kernel.org/r/9eca6831f9fe2d781bb88337397c39b10e36f5c6.1782458224.git.mazziesaccount@gmail.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Add support for the 'ti,current-range' devicetree property to configure
the current limit via the DEVICE_SETUP (0xD9) register, overriding the
physical CL pin setting.
This configuration is supported on all chips in this driver (LM25066,
LM5064, LM5066, LM5066i) except LM25056.
The property values "low" and "high" map to:
- LM25066: low = 25 mV, high = 46 mV
- LM5064, LM5066, LM5066i: low = 26 mV, high = 50 mV
The Bit 4 mapping to High/Low current limit is handled dynamically on
probe because it is swapped for LM25066 compared to the other supported
chips.
Signed-off-by: Potin Lai <potin.lai.pt@gmail.com>
Link: https://lore.kernel.org/r/20260615-lm25066-cl-config-v3-2-decb4f5b0b77@gmail.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Add a 'ti,current-range' string property to configure the device's Current
Limit (CL) behavior to "high" or "low" via the register, overriding the
physical CL pin setting.
This configuration is supported on LM25066, LM5064, LM5066, and LM5066i.
LM25056 is excluded because it does not support configuring the current
limit via the DEVICE_SETUP register (bit 2 of DEVICE_SETUP is reserved).
The values "low" and "high" map to the respective low/high threshold
voltages of the chips:
- LM25066: low = 25 mV, high = 46 mV
- LM5064, LM5066, LM5066i: low = 26 mV, high = 50 mV
Signed-off-by: Potin Lai <potin.lai.pt@gmail.com>
Reviewed-by: Conor Dooley <conor.dooley@microchip.com>
Link: https://lore.kernel.org/r/20260615-lm25066-cl-config-v3-1-decb4f5b0b77@gmail.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
