linux/drivers/regulator, branch v6.13

regulator: rename regulator-uv-survival-time-ms according to DT binding

2024-12-19T11:15:24Z

The regulator bindings don't document regulator-uv-survival-time-ms, but the more descriptive regulator-uv-less-critical-window-ms instead. Looking back at v3[1] and v4[2] of the series adding the support, the property was indeed renamed between these patch series, but unfortunately the rename only made it into the DT bindings with the driver code still using the old name. Let's therefore rename the property in the driver code to follow suit. This will break backwards compatibility, but there are no upstream device trees using the property and we never documented the old name of the property anyway. ¯\_(ツ)_/¯" [1]: https://lore.kernel.org/all/20231025084614.3092295-7-o.rempel@pengutronix.de/ [2]: https://lore.kernel.org/all/20231026144824.4065145-5-o.rempel@pengutronix.de/ Signed-off-by: Ahmad Fatoum Link: https://patch.msgid.link/20241218-regulator-uv-survival-time-ms-rename-v1-1-6cac9c3c75da@pengutronix.de Signed-off-by: Mark Brown

regulator: axp20x: AXP717: set ramp_delay

2024-12-09T13:06:07Z

AXP717 datasheet says that regulator ramp delay is 15.625 us/step, which is 10mV in our case. Add a AXP_DESC_RANGES_DELAY macro and update AXP_DESC_RANGES macro to expand to AXP_DESC_RANGES_DELAY with ramp_delay = 0 For DCDC4, steps is 100mv Add a AXP_DESC_DELAY macro and update AXP_DESC macro to expand to AXP_DESC_DELAY with ramp_delay = 0 This patch fix crashes when using CPU DVFS. Signed-off-by: Philippe Simons Tested-by: Hironori KIKUCHI Tested-by: Chris Morgan Reviewed-by: Chen-Yu Tsai Fixes: d2ac3df75c3a ("regulator: axp20x: add support for the AXP717") Link: https://patch.msgid.link/20241208124308.5630-1-simons.philippe@gmail.com Signed-off-by: Mark Brown

Merge tag 'regulator-fix-v6.13-merge-window' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regulator

2024-11-28T17:40:53Z

Pull regulator fixes from Mark Brown: "A couple of fixes that came in during the merge window, plus documetation of a new device ID for the Qualcomm LABIBB driver. There's a core fix for the rarely used current constraints and a fix for the Qualcomm RPMH driver which had described only one of the two voltage ranges that the hardware could control, creating a potential incompatibility with the configuration left by firmware" * tag 'regulator-fix-v6.13-merge-window' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regulator: regulator: core: Ignore unset max_uA constraints in current limit check dt-bindings: regulator: qcom-labibb-regulator: document the pmi8950 labibb regulator regulator: qcom-rpmh: Update ranges for FTSMPS525

Merge tag 'mfd-next-6.13' of git://git.kernel.org/pub/scm/linux/kernel/git/lee/mfd

2024-11-23T00:19:47Z

Pull MFD updates from Lee Jones: - Several drivers, including atmel-flexcom/rk8xx-core, palmas, and tps65010, have undergone minor code improvements to enhance consistency and fix race conditions. - The syscon driver now utilizes the regmap max_register_is_0 capability for consistent register map configuration across syscons of all sizes. - New device support has been added for QCS8300, qcs615, SA8255p, and samsung,s2dos05, expanding the range of compatible hardware. - The cros_ec driver now supports loading cros_ec_ucsi on supported ECs and avoids loading the charger with UCSI, streamlining functionality. - The bd96801 driver now utilizes the more modern maple tree register cache, improving performance. - The da9052-spi driver has undergone a fix to change the read-mask to write-mask, preventing potential issues. - Unused declarations in max77693 have been removed, and support for samsung,s2dos05 has been added, enhancing code clarity and device compatibility. - Error handling in cs42l43 has been fixed to avoid unbalanced regulator put and ensure proper synchronization during driver removal. - The wcd934x driver now uses MODULE_DEVICE_TABLE() instead of MODULE_ALIAS(), improving code consistency. - Documentation for qcom,tcsr, syscon, and atmel-smc has been updated and reorganized for better clarity and maintainability. - The intel_soc_pmic_bxtwc driver has undergone significant improvements, including the use of IRQ domains for various devices, fixing IRQ domain names duplication, and code refactoring for better consistency and maintainability. - The ipaq-micro driver has received a fix for a missing break statement in the default case, enhancing code robustness. - Support for the AXP323 PMIC has been added to the axp20x driver, along with ensuring a clear relationship between IDs and model names, and allowing multiple regulators, broadening hardware compatibility. - The cs42l43 driver now disables IRQs during suspend for improved power management. - The adp5585 driver has reduced its dependencies by dropping the obsolete dependency on COMPILE_TEST. - Initial support for the MT6328 PMIC has been added to the mt6397 driver, expanding the range of supported hardware. - The rtc-bd70528 driver has been simplified by dropping the IC name from IRQ, improving code readability. - Documentation for qcom,spmi-pmic, ti,twl, and zii,rave-sp has been updated to enhance clarity and incorporate new features. - The rt5033 driver has received a fix for a missing regmap_del_irq_chip() in the error handling path. - New device support has been added for MSM8917, and the intel_soc_pmic_crc driver now supports non-ACPI instantiated i2c_client. - The 88pm886 driver has added support for the RTC cell, and the tqmx86 driver has improved its GPIO IRQ setup and added I2C IRQ support, increasing functionality. - The sprd,sc2731 DT schema has been updated and converted to YAML format for better readability and maintainability. * tag 'mfd-next-6.13' of git://git.kernel.org/pub/scm/linux/kernel/git/lee/mfd: (62 commits) dt-bindings: mfd: bd71828: Use charger resistor in mOhm instead of MOhm dt-bindings: mfd: sprd,sc2731: Convert to YAML mfd: tqmx86: Add I2C IRQ support mfd: tqmx86: Make IRQ setup errors non-fatal mfd: tqmx86: Refactor GPIO IRQ setup mfd: tqmx86: Improve gpio_irq module parameter description mfd: tqmx86: Add board definitions for TQMx120UC, TQMx130UC and TQMxE41S mfd: 88pm886: Add the RTC cell dt-bindings: mfd: Add Realtek RTL9300 switch peripherals mfd: intel_soc_pmic_crc: Add support for non ACPI instantiated i2c_client mfd: intel_soc_pmic_*: Consistently use filename as driver name dt-bindings: mfd: qcom,tcsr: Add compatible for MSM8917 mfd: rt5033: Fix missing regmap_del_irq_chip() mfd: cgbc-core: Fix error handling paths in cgbc_init_device() dt-bindings: mfd: aspeed: Support for AST2700 mfd: Switch back to struct platform_driver::remove() dt-bindings: mfd: qcom,spmi-pmic: Document PMICs added in SM8750 mfd: rtc: bd7xxxx Drop IC name from IRQ mfd: mt6397: Add initial support for MT6328 mfd: adp5585: Drop obsolete dependency on COMPILE_TEST ...

regulator: core: Ignore unset max_uA constraints in current limit check

2024-11-21T14:48:25Z

We should only consider max_uA constraints if they are explicitly defined. In cases where it is not set, we should assume the regulator has no current limit. Signed-off-by: Kory Maincent Link: https://patch.msgid.link/20241121-feature_poe_port_prio-v3-2-83299fa6967c@bootlin.com Signed-off-by: Mark Brown

Merge tag 'regulator-v6.13' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regulator

2024-11-20T20:18:50Z

Pull regulator updates from Mark Brown: "This was a quite quiet release for regulators on the drivers front, but we do have two small but useful core improvements: - Improve handling of cases where DT platforms specify both DT and init_data based configuration for a single regulator. - A helper of_regulator_get_optional() to simplify writing generic bindings for regulator consumers in subsystems. There's also some YAML conversions for the DT bindings which are a big part of the diffstat" * tag 'regulator-v6.13' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regulator: regulator: dt-bindings: qcom,rpmh: Correct PM8550VE supplies regulator: Switch back to struct platform_driver::remove() regulator: doc: remove documentation comment for regulator_init regulator: doc: add missing documentation for init_cb regulator: rk808: Restrict DVS GPIOs to the RK808 variant only regulator: rk808: Use dev_err_probe() in the probe path regulator: rk808: Perform trivial code cleanups regulator: dt-bindings: qcom,qca6390-pmu: add more properties for wcn6855 regulator: dt-bindings: lltc,ltc3676: convert to YAML regulator: core: Use fsleep() to get best sleep mechanism regulator: core: remove machine init callback from config regulator: core: add callback to perform runtime init regulator: core: do not silently ignore provided init_data regulator: max5970: Drop unused structs regulator: dt-bindings: vctrl-regulator: convert to YAML regulator: qcom-smd: make smd_vreg_rpm static regulator: Call of_node_put() only once in rzg2l_usb_vbus_regulator_probe() regulator: isl6271a: Drop explicit initialization of struct i2c_device_id::driver_data to 0 regulator: Add devres version of of_regulator_get_optional() regulator: Add of_regulator_get_optional() for pure DT regulator lookup

regulator: qcom-rpmh: Update ranges for FTSMPS525

2024-11-18T15:44:30Z

All FTSMPS525 regulators support LV and MV ranges; however, the boot loader firmware will determine which range to use as the device boots. Nonetheless, the driver cannot determine which range was selected, so hardcoding the ranges as either LV or MV will not cover all cases as it's possible for the firmware to select a range not supported by the driver's current hardcoded values. To this end, combine the ranges for the FTSMPS525s into one struct and point all regulators to the updated combined struct. This should work on all boards regardless of which range is selected by the firmware and more accurately caputres the capability of this regulator on a hardware level. Signed-off-by: Melody Olvera Reviewed-by: Dmitry Baryshkov Link: https://patch.msgid.link/20241112002645.2803506-1-quic_molvera@quicinc.com Signed-off-by: Mark Brown

regulator: Switch back to struct platform_driver::remove()

2024-11-12T13:08:35Z

After commit 0edb555a65d1 ("platform: Make platform_driver::remove() return void") .remove() is (again) the right callback to implement for platform drivers. Convert all platform drivers below drivers/regulator to use .remove(), with the eventual goal to drop struct platform_driver::remove_new(). As .remove() and .remove_new() have the same prototypes, conversion is done by just changing the structure member name in the driver initializer. A few whitespace changes are done en passant to make indention consistent. Signed-off-by: Uwe Kleine-König Link: https://patch.msgid.link/ab85510f83fa901e44d5d563fe6e768054229bfe.1731398433.git.u.kleine-koenig@baylibre.com Signed-off-by: Mark Brown

regulator: rk808: Add apply_bit for BUCK3 on RK809

2024-11-01T14:47:08Z

Currently, RK809's BUCK3 regulator is modelled in the driver as a configurable regulator with 0.5-2.4V voltage range. But the voltage setting is not actually applied, because when bit 6 of PMIC_POWER_CONFIG register is set to 0 (default), BUCK3 output voltage is determined by the external feedback resistor. Fix this, by setting bit 6 when voltage selection is set. Existing users which do not specify voltage constraints in their device trees will not be affected by this change, since no voltage setting is applied in those cases, and bit 6 is not enabled. Signed-off-by: Mikhail Rudenko Link: https://patch.msgid.link/20241017-rk809-dcdc3-v1-1-e3c3de92f39c@gmail.com Signed-off-by: Mark Brown

regulator: axp20x: add support for the AXP323

2024-10-31T15:51:31Z

The X-Powers AXP323 is a very close sibling of the AXP313A. The only difference seems to be the ability to dual-phase the first two DC/DC converters. Place the new AXP323 ID next to the existing AXP313A checks, to let them share most code. The only difference is the poly-phase detection code, which gets extended to check the respective bit in a newly used register. Signed-off-by: Andre Przywara Reviewed-by: Chen-Yu Tsai Reviewed-by: Mark Brown Link: https://lore.kernel.org/r/20241007001408.27249-6-andre.przywara@arm.com Signed-off-by: Lee Jones