gpio: grgpio: Make irq_chip immutable

Kernel warns about mutable irq_chips:
"not an immutable chip, please consider fixing!"

Constify grgpio_irq_chip, flag the irq_chip as IRQCHIP_IMMUTABLE, add the
new helper functions, and call the appropriate gpiolib functions.

Signed-off-by: Peng Fan <peng.fan@nxp.com>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20250509-gpio-v1-2-639377c98288@nxp.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: bcm-kona: make irq_chip immutable

The gpiolib is moving to make irq_chip immutable, otherwise
there is warning: "not an immutable chip, please consider fixing it!"

The bcm_gpio_irq_chip already has irq hooks configured correctly,
bcm_kona_gpio_irq_mask/bcm_kona_gpio_irq_unmask calls gpiochip_disable_irq/
gpiochip_enable_irq, and bcm_kona_gpio_irq_reqres/irq_release_resources
calls gpiochip_reqres_irq/gpiochip_relres_irq.

So just need to  flag it as IRQCHIP_IMMUTABLE.

Signed-off-by: Peng Fan <peng.fan@nxp.com>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20250509-gpio-v1-1-639377c98288@nxp.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: TODO: add item about GPIO drivers reading struct gpio_chip::base

drivers/pinctrl/pinctrl-at91.c uses struct gpio_chip::base to find out
which bit to set in a register:

  dev_dbg(npct->dev, "enable pin %u as GPIO\n", offset);

  mask = 1 << (offset - chip->base);

This adds a non-obvious dependency on the global numberspace from the
GPIO driver itself, even if all consumers use the descriptor API.

More such instances may exist and will need to be fixed in the quest for
removal of the global numberspace, so add a reminder about it to the
TODO list.

Link: https://lore.kernel.org/all/1d00c056-3d61-4c22-bedd-3bae0bf1ddc4@pengutronix.de/
Suggested-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Signed-off-by: Ahmad Fatoum <a.fatoum@pengutronix.de>
Link: https://lore.kernel.org/r/20250507-gpio-chip-base-readback-v1-1-ade56e38480b@pengutronix.de
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: mxc: configure dynamic GPIO base for CONFIG_GPIO_SYSFS=n

i.MX GPIO numbering has been deterministic since commit 7e6086d9e54a
("gpio/mxc: specify gpio base for device tree probe"), a year after
device tree support was first added back in 2011.

Reverting to dynamically allocated GPIO base now would break most
systems making use of the sysfs API. These systems will be eventually
broken by the removal of the sysfs API, but that would result in GPIO
scripts not working instead of essentially toggling at random according
to probe order, which would happen if we unconditionally set base to -1.

Yet, the warning is annoying and has resulted in many rejected attempts
to remove it over the years[1][2][3].

As the i.MX GPIO driver is device tree only, GPIO_SYSFS is the only
consumer of the deterministic GPIO numbering. Let's therefore restrict the
static base and the warning that comes with it to configurations
with CONFIG_GPIO_SYSFS enabled.

[1]: https://lore.kernel.org/all/20230226205319.1013332-1-dario.binacchi@amarulasolutions.com/
[2]: https://lore.kernel.org/all/20230506085928.933737-2-haibo.chen@nxp.com/
[3]: https://lore.kernel.org/all/20241121145515.3087855-1-catalin.popescu@leica-geosystems.com/

Signed-off-by: Ahmad Fatoum <a.fatoum@pengutronix.de>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20250507-b4-imx-gpio-base-warning-v2-1-d43d09e2c6bf@pengutronix.de
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

dt-bindings: gpio: tegra186: Add gpio-ranges

Add optional gpio-ranges property.

Signed-off-by: Aaron Kling <webgeek1234@gmail.com>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Acked-by: Rob Herring (Arm) <robh@kernel.org>
Link: https://lore.kernel.org/r/20250429-tegra186-pinctrl-v1-2-722c7c42394e@gmail.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: ml-ioh: use new GPIO line value setter callbacks

struct gpio_chip now has callbacks for setting line values that return
an integer, allowing to indicate failures. Convert the driver to using
them.

Link: https://lore.kernel.org/r/20250506-gpiochip-set-rv-gpio-part3-v1-12-0fbdea5a9667@linaro.org
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: mc33880: use new GPIO line value setter callbacks

struct gpio_chip now has callbacks for setting line values that return
an integer, allowing to indicate failures. Convert the driver to using
them.

Link: https://lore.kernel.org/r/20250506-gpiochip-set-rv-gpio-part3-v1-11-0fbdea5a9667@linaro.org
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: mb86s7x: use new GPIO line value setter callbacks

struct gpio_chip now has callbacks for setting line values that return
an integer, allowing to indicate failures. Convert the driver to using
them.

Link: https://lore.kernel.org/r/20250506-gpiochip-set-rv-gpio-part3-v1-10-0fbdea5a9667@linaro.org
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: max77620: use new GPIO line value setter callbacks

struct gpio_chip now has callbacks for setting line values that return
an integer, allowing to indicate failures. Convert the driver to using
them.

Link: https://lore.kernel.org/r/20250506-gpiochip-set-rv-gpio-part3-v1-9-0fbdea5a9667@linaro.org
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: max732x: use new GPIO line value setter callbacks

struct gpio_chip now has callbacks for setting line values that return
an integer, allowing to indicate failures. Convert the driver to using
them.

Link: https://lore.kernel.org/r/20250506-gpiochip-set-rv-gpio-part3-v1-8-0fbdea5a9667@linaro.org
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: max730x: use new GPIO line value setter callbacks

struct gpio_chip now has callbacks for setting line values that return
an integer, allowing to indicate failures. Convert the driver to using
them.

Link: https://lore.kernel.org/r/20250506-gpiochip-set-rv-gpio-part3-v1-7-0fbdea5a9667@linaro.org
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: max3191x: remove unused callbacks

This driver is input-only and as such doesn't need to define empty set()
and direction_output() callbacks. Remove them.

Link: https://lore.kernel.org/r/20250506-gpiochip-set-rv-gpio-part3-v1-6-0fbdea5a9667@linaro.org
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: madera: use new GPIO line value setter callbacks

struct gpio_chip now has callbacks for setting line values that return
an integer, allowing to indicate failures. Convert the driver to using
them.

Reviewed-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20250506-gpiochip-set-rv-gpio-part3-v1-5-0fbdea5a9667@linaro.org
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: lpc32xx: use new GPIO line value setter callbacks

struct gpio_chip now has callbacks for setting line values that return
an integer, allowing to indicate failures. Convert the driver to using
them.

Link: https://lore.kernel.org/r/20250506-gpiochip-set-rv-gpio-part3-v1-4-0fbdea5a9667@linaro.org
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: lpc18xx: use new GPIO line value setter callbacks

struct gpio_chip now has callbacks for setting line values that return
an integer, allowing to indicate failures. Convert the driver to using
them.

Link: https://lore.kernel.org/r/20250506-gpiochip-set-rv-gpio-part3-v1-3-0fbdea5a9667@linaro.org
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: lp87565: use new GPIO line value setter callbacks

struct gpio_chip now has callbacks for setting line values that return
an integer, allowing to indicate failures. Convert the driver to using
them.

Link: https://lore.kernel.org/r/20250506-gpiochip-set-rv-gpio-part3-v1-2-0fbdea5a9667@linaro.org
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: lp873x: use new GPIO line value setter callbacks

struct gpio_chip now has callbacks for setting line values that return
an integer, allowing to indicate failures. Convert the driver to using
them.

Link: https://lore.kernel.org/r/20250506-gpiochip-set-rv-gpio-part3-v1-1-0fbdea5a9667@linaro.org
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

Merge tag 'ib-mfd-gpio-nvmem-v6.16' of git://git./linux/kernel/git/lee/mfd into gpio/for-next

Immutable branch between MFD, GPIO and NVMEM due for the v6.16 merge window

nvmem: max77759: Add Maxim MAX77759 NVMEM driver

The Maxim MAX77759 is a companion PMIC for USB Type-C applications and
includes Battery Charger, Fuel Gauge, temperature sensors, USB Type-C
Port Controller (TCPC), NVMEM, and a GPIO expander.

This driver exposes the non volatile memory using the platform device
registered by the core MFD driver.

Signed-off-by: André Draszik <andre.draszik@linaro.org>
Acked-by: Srinivas Kandagatla <srini@kernel.org>
Acked-by: Peter Griffin <peter.griffin@linaro.org>
Link: https://lore.kernel.org/r/20250509-max77759-mfd-v10-3-962ac15ee3ef@linaro.org
Signed-off-by: Lee Jones <lee@kernel.org>

gpio: max77759: Add Maxim MAX77759 gpio driver

The Maxim MAX77759 is a companion PMIC for USB Type-C applications and
includes Battery Charger, Fuel Gauge, temperature sensors, USB Type-C
Port Controller (TCPC), NVMEM, and a GPIO expander.

This driver supports the GPIO functions using the platform device
registered by the core MFD driver.

Signed-off-by: André Draszik <andre.draszik@linaro.org>
Acked-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Acked-by: Peter Griffin <peter.griffin@linaro.org>
Link: https://lore.kernel.org/r/20250509-max77759-mfd-v10-2-962ac15ee3ef@linaro.org
Signed-off-by: Lee Jones <lee@kernel.org>

mfd: max77759: Add Maxim MAX77759 core driver

The Maxim MAX77759 is a companion PMIC for USB Type-C applications and
includes Battery Charger, Fuel Gauge, temperature sensors, USB Type-C
Port Controller (TCPC), NVMEM, and a GPIO expander.

Fuel Gauge and TCPC have separate and independent I2C addresses,
register maps, and interrupt lines and are therefore excluded from the
MFD core device driver here.

The GPIO and NVMEM interfaces are accessed via specific commands to the
built-in microprocessor. This driver implements an API that client
drivers can use for accessing those.

Signed-off-by: André Draszik <andre.draszik@linaro.org>
Acked-by: Peter Griffin <peter.griffin@linaro.org>
Link: https://lore.kernel.org/r/20250509-max77759-mfd-v10-1-962ac15ee3ef@linaro.org
Signed-off-by: Lee Jones <lee@kernel.org>

dt-bindings: mfd: Add max77759 binding

The Maxim MAX77759 is a companion PMIC for USB Type-C applications and
includes Battery Charger, Fuel Gauge, temperature sensors, USB Type-C
Port Controller (TCPC), NVMEM, and a GPIO expander.

This describes the top-level device.

Signed-off-by: André Draszik <andre.draszik@linaro.org>
Reviewed-by: "Rob Herring (Arm)" <robh@kernel.org>
Link: https://lore.kernel.org/r/20250325-max77759-mfd-v6-3-c0870ca662ba@linaro.org
Signed-off-by: Lee Jones <lee@kernel.org>

dt-bindings: nvmem: Add max77759 binding

The Maxim MAX77759 is a companion PMIC for USB Type-C applications and
includes Battery Charger, Fuel Gauge, temperature sensors, USB Type-C
Port Controller (TCPC), NVMEM, and a GPIO expander.

This describes its storage module (NVMEM).

Signed-off-by: André Draszik <andre.draszik@linaro.org>
Reviewed-by: "Rob Herring (Arm)" <robh@kernel.org>
Link: https://lore.kernel.org/r/20250325-max77759-mfd-v6-2-c0870ca662ba@linaro.org
Signed-off-by: Lee Jones <lee@kernel.org>

dt-bindings: gpio: Add max77759 binding

The Maxim MAX77759 is a companion PMIC for USB Type-C applications and
includes Battery Charger, Fuel Gauge, temperature sensors, USB Type-C
Port Controller (TCPC), NVMEM, and a GPIO expander.

This describes its GPIO module.

Signed-off-by: André Draszik <andre.draszik@linaro.org>
Reviewed-by: "Rob Herring (Arm)" <robh@kernel.org>
Acked-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Link: https://lore.kernel.org/r/20250325-max77759-mfd-v6-1-c0870ca662ba@linaro.org
Signed-off-by: Lee Jones <lee@kernel.org>

gpio: blzp1600: remove incorrect pf_match_ptr()

The blzp1600_gpio_of_match is defined unconditionally and can always be referenced
by the driver, so don't use of_match_ptr() and avoid this warning:

drivers/gpio/gpio-blzp1600.c:263:34: error: 'blzp1600_gpio_of_match' defined but not used [-Werror=unused-const-variable=]
  263 | static const struct of_device_id blzp1600_gpio_of_match[] = {

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Link: https://lore.kernel.org/r/20250502203721.2117186-1-arnd@kernel.org
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: lp3943: use new line value setter callbacks

struct gpio_chip now has callbacks for setting line values that return
an integer, allowing to indicate failures. Convert the driver to using
them.

Reviewed-by: Peng Fan <peng.fan@nxp.com>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20250423-gpiochip-set-rv-gpio-part2-v1-12-b22245cde81a@linaro.org
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: loongson: use new line value setter callbacks

struct gpio_chip now has callbacks for setting line values that return
an integer, allowing to indicate failures. Convert the driver to using
them.

Reviewed-by: Peng Fan <peng.fan@nxp.com>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20250423-gpiochip-set-rv-gpio-part2-v1-11-b22245cde81a@linaro.org
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: loongson-64bit: use new line value setter callbacks

struct gpio_chip now has callbacks for setting line values that return
an integer, allowing to indicate failures. Convert the driver to using
them.

Reviewed-by: Peng Fan <peng.fan@nxp.com>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20250423-gpiochip-set-rv-gpio-part2-v1-10-b22245cde81a@linaro.org
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: logicvc: use new line value setter callbacks

struct gpio_chip now has callbacks for setting line values that return
an integer, allowing to indicate failures. Convert the driver to using
them.

Reviewed-by: Peng Fan <peng.fan@nxp.com>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20250423-gpiochip-set-rv-gpio-part2-v1-9-b22245cde81a@linaro.org
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: ljca: use new line value setter callbacks

struct gpio_chip now has callbacks for setting line values that return
an integer, allowing to indicate failures. Convert the driver to using
them.

Reviewed-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Reviewed-by: Peng Fan <peng.fan@nxp.com>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20250423-gpiochip-set-rv-gpio-part2-v1-8-b22245cde81a@linaro.org
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: kempld: use new line value setter callbacks

struct gpio_chip now has callbacks for setting line values that return
an integer, allowing to indicate failures. Convert the driver to using
them.

Reviewed-by: Peng Fan <peng.fan@nxp.com>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20250423-gpiochip-set-rv-gpio-part2-v1-7-b22245cde81a@linaro.org
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: janz-ttl: use new line value setter callbacks

struct gpio_chip now has callbacks for setting line values that return
an integer, allowing to indicate failures. Convert the driver to using
them.

Reviewed-by: Peng Fan <peng.fan@nxp.com>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20250423-gpiochip-set-rv-gpio-part2-v1-6-b22245cde81a@linaro.org
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: it87: use new line value setter callbacks

struct gpio_chip now has callbacks for setting line values that return
an integer, allowing to indicate failures. Convert the driver to using
them.

Reviewed-by: Peng Fan <peng.fan@nxp.com>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20250423-gpiochip-set-rv-gpio-part2-v1-5-b22245cde81a@linaro.org
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: imx-scu: use new line value setter callbacks

struct gpio_chip now has callbacks for setting line values that return
an integer, allowing to indicate failures. Convert the driver to using
them.

Reviewed-by: Peng Fan <peng.fan@nxp.com>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20250423-gpiochip-set-rv-gpio-part2-v1-4-b22245cde81a@linaro.org
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: imx-scu: destroy the mutex in detach path

Use devm_mutex_init() in order to clean up after the mutex debug data in
detach and error path.

Reviewed-by: Peng Fan <peng.fan@nxp.com>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20250423-gpiochip-set-rv-gpio-part2-v1-3-b22245cde81a@linaro.org
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: imx-scu: use lock guards

Simplify the code by using lock guards from cleanup.h with the driver's
mutex.

Reviewed-by: Peng Fan <peng.fan@nxp.com>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20250423-gpiochip-set-rv-gpio-part2-v1-2-b22245cde81a@linaro.org
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: imx-scu: don't check the GPIO range

Core GPIO code already does this for drivers, no need to duplicate the
check.

Reviewed-by: Peng Fan <peng.fan@nxp.com>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20250423-gpiochip-set-rv-gpio-part2-v1-1-b22245cde81a@linaro.org
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpiolib: devres: Finish the conversion to use devm_add_action()

With a recently added helper we can complete the conversion of
the GPIOLIB code to use devm_add_action() in all suitable cases.
So do this.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Link: https://lore.kernel.org/r/20250220162238.2738038-4-andriy.shevchenko@linux.intel.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

Merge tag 'gpiod-devm-is-action-added-for-v6.16-rc1' into gpio/for-next

Immutable tag for the driver core tree to pull from

devres: Move devm_*_action*() APIs to devres.h
devres: Add devm_is_action_added() helper

devres: Add devm_is_action_added() helper

In some code we would like to know if the action in device managed resources
was added by devm_add_action() family of calls. Introduce a helper for that.

Reviewed-by: Raag Jadav <raag.jadav@intel.com>
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: Zijun Hu <quic_zijuhu@quicinc.com>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Link: https://lore.kernel.org/r/20250220162238.2738038-3-andriy.shevchenko@linux.intel.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

devres: Move devm_*_action*() APIs to devres.h

We have a newly created header linux/device/devres.h that gathers
device managed APIs, so users won't need to include entire device.h
for only these ones. Move devm_*_action*() APIs to devres.h as well.

Reviewed-by: Raag Jadav <raag.jadav@intel.com>
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: Zijun Hu <quic_zijuhu@quicinc.com>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Link: https://lore.kernel.org/r/20250220162238.2738038-2-andriy.shevchenko@linux.intel.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: pca953x: Add support for level-triggered interrupts

Adds support for level-triggered interrupts in the PCA953x GPIO
expander driver. Previously, the driver only supported edge-triggered
interrupts, which could lead to missed events in scenarios where an
interrupt condition persists until it is explicitly cleared.

By enabling level-triggered interrupts, the driver can now detect and
respond to sustained interrupt conditions more reliably.

Signed-off-by: Potin Lai <potin.lai.pt@gmail.com>
Link: https://lore.kernel.org/r/20250409-gpio-pca953x-level-triggered-irq-v3-1-7f184d814934@gmail.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: vf610: Allow building as a module

Support for building it as a module for use on the Android platform,
because the Android kernel(GKI) doesn't contain board-specific drivers,
it requires that these drivers be built as a module and loaded into the
system.

Signed-off-by: Jindong Yue <jindong.yue@nxp.com>
Link: https://lore.kernel.org/r/20250416062605.1948856-1-jindong.yue@nxp.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: Restrict GPIO_ICH to compile testing with HAS_IOPORT

When building gpio-ich.c for a platform that does not have IO port
access like s390, there are several errors due to the use of inl() and
outl() throughout the driver:

  drivers/gpio/gpio-ich.c: In function 'ichx_read_bit':
  include/asm-generic/io.h:578:14: error: call to '_inl' declared with attribute error: inl() requires CONFIG_HAS_IOPORT

Only allow compile testing when the targeted platform selects HAS_IOPORT
so that there are no compile time errors.

Fixes: f3592d252f77 ("gpio: ich: enable building with COMPILE_TEST=y")
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202504171941.g1yXja0j-lkp@intel.com/
Signed-off-by: Nathan Chancellor <nathan@kernel.org>
Link: https://lore.kernel.org/r/20250418-gpio-ich-fix-build-without-ioport-v1-1-83fc753438ec@kernel.org
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpiolib: Remove redundant assignment of return variable

In some functions the returned variable is assigned to 0 and then
reassigned to the actual value. Remove redundant assignments.

In one case make it more clear that the assignment is not needed.

Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Link: https://lore.kernel.org/r/20250416095645.2027695-9-andriy.shevchenko@linux.intel.com
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

gpiolib: Reuse return variable in gpiod_to_irq()

There are two variables for the same used in the gpiod_to_irq().
Replace the second by reusing the function top-level one.
While at it, refactor the branch to have less lines of code.

Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Link: https://lore.kernel.org/r/20250416095645.2027695-8-andriy.shevchenko@linux.intel.com
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

gpiolib: Call validate_desc() when VALIDATE_DESC() can't be used

Call validate_desc() directly when VALIDATE_DESC() can't be used.
It will print additional information useful for debugging.

Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Link: https://lore.kernel.org/r/20250416095645.2027695-7-andriy.shevchenko@linux.intel.com
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

gpiolib: Move validate_desc() and Co upper in the code

Move validate_desc() and Co upper in the code to be able to use
in the further changes.

Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Link: https://lore.kernel.org/r/20250416095645.2027695-6-andriy.shevchenko@linux.intel.com
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

gpiolib: Revert "Don't WARN on gpiod_put() for optional GPIO"

No need to double check the pointer for NULL since gpiod_free()
is using VALIDATE_DESC_VOID() which simply returns in that case.

This reverts commit 1d7765ba15aca68f3bc52f59434c1c34855bbb54.

Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Link: https://lore.kernel.org/r/20250416095645.2027695-5-andriy.shevchenko@linux.intel.com
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

gpiolib: Print actual error when descriptor contains an error pointer

Print the actual error when descriptor contains an error pointer.
This might help debugging those rare cases.

Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Link: https://lore.kernel.org/r/20250416095645.2027695-4-andriy.shevchenko@linux.intel.com
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

gpiolib: Convert to use guard()() for gpio_machine_hogs_mutex

The driver uses guard()()/scoped_guard() for the rest of the synchronisation
calls. Convert to use the same for gpio_machine_hogs_mutex.

Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Link: https://lore.kernel.org/r/20250416095645.2027695-3-andriy.shevchenko@linux.intel.com
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

gpiolib: Make taking gpio_lookup_lock consistent

There are two ways to take a lock: plain call to the mutex_lock()
or using guard()() / scoped_guard(). The driver inconsistently uses
both. Make taking gpio_lookup_lock consistent.

Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Link: https://lore.kernel.org/r/20250416095645.2027695-2-andriy.shevchenko@linux.intel.com
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

gpio: spacemit: add support for K1 SoC

Implement GPIO functionality which capable of setting pin as
input, output. Also, each pin can be used as interrupt which
support rising, falling, or both edge type trigger.

Reviewed-by: Alex Elder <elder@riscstar.com>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Yixun Lan <dlan@gentoo.org>
Link: https://lore.kernel.org/r/20250412-03-k1-gpio-v8-2-1c6862d272ec@gentoo.org
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

dt-bindings: gpio: spacemit: add support for K1 SoC

The GPIO controller of K1 support basic functions as input/output,
all pins can be used as interrupt which route to one IRQ line,
trigger type can be select between rising edge, falling edge, or both.
There are four GPIO banks, each consisting of 32 pins.

Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Reviewed-by: Rob Herring (Arm) <robh@kernel.org>
Signed-off-by: Yixun Lan <dlan@gentoo.org>
Link: https://lore.kernel.org/r/20250412-03-k1-gpio-v8-1-1c6862d272ec@gentoo.org
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

gpio: tn48m: enable building the module with COMPILE_TEST=y

Extend the build coverage by allowing to build the module with
COMPILE_TEST enabled.

Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20250408-gpio-compile-test-v1-6-140e108e9392@linaro.org
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

gpio: rtd: enable building the module with COMPILE_TEST=y

Extend the build coverage by allowing to build the module with
COMPILE_TEST enabled.

Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20250408-gpio-compile-test-v1-4-140e108e9392@linaro.org
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

gpio: pl061: enable building the module with COMPILE_TEST=y

Extend the build coverage by allowing to build the module with
COMPILE_TEST enabled.

Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20250408-gpio-compile-test-v1-3-140e108e9392@linaro.org
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

gpio: msc313: enable building the module with COMPILE_TEST=y

Extend the build coverage by allowing to build the module with
COMPILE_TEST enabled.

Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20250408-gpio-compile-test-v1-2-140e108e9392@linaro.org
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

gpio: zynq: enable building the modules with COMPILE_TEST=y

Extend the build coverage by allowing to build the modules with
COMPILE_TEST enabled.

Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20250408-gpio-compile-test-v1-1-140e108e9392@linaro.org
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

gpio: blzp1600: drop dependency on OF headers

Use the generic boolean device property getter instead of the
OF-specific variant. This allows us to stop pulling in linux/of.h. While
at it: drop the of_irq.h inclusion as none of its symbols are used in
this driver.

Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Reviewed-by: Nikolaos Pasaloukos <nikolaos.pasaloukos@blaize.com>
Link: https://lore.kernel.org/r/20250408143629.125576-1-brgl@bgdev.pl
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

gpio: ds4520: don't check the 'ngpios' property in the driver

This is already done by the gpio-regmap abstraction by calling
gpiochip_get_ngpios(). We don't need to do this at the driver level.

Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Link: https://lore.kernel.org/r/20250408-gpio-regmap-ngpios-v2-1-26f254014f57@linaro.org
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

gpio: ich: use new line value setter callbacks

struct gpio_chip now has callbacks for setting line values that return
an integer, allowing to indicate failures. Convert the driver to using
them.

Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Link: https://lore.kernel.org/r/20250407-gpiochip-set-rv-gpio-part1-v1-12-78399683ca38@linaro.org
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

gpio: ich: enable building with COMPILE_TEST=y

Extend the build coverage by allowing to build the module with
COMPILE_TEST enabled.

Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20250407-gpiochip-set-rv-gpio-part1-v1-11-78399683ca38@linaro.org
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

gpio: htc-egpio: use new line value setter callbacks

struct gpio_chip now has callbacks for setting line values that return
an integer, allowing to indicate failures. Convert the driver to using
them.

Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Link: https://lore.kernel.org/r/20250407-gpiochip-set-rv-gpio-part1-v1-10-78399683ca38@linaro.org
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

gpio: htc-egpio: enable building with COMPILE_TEST=y

Extend the build coverage by allowing to build the module with
COMPILE_TEST enabled.

Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20250407-gpiochip-set-rv-gpio-part1-v1-9-78399683ca38@linaro.org
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

gpio: gw-pld: use new line value setter callbacks

struct gpio_chip now has callbacks for setting line values that return
an integer, allowing to indicate failures. Convert the driver to using
them.

Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Link: https://lore.kernel.org/r/20250407-gpiochip-set-rv-gpio-part1-v1-8-78399683ca38@linaro.org
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

gpio: graniterapids: use new line value setter callbacks

struct gpio_chip now has callbacks for setting line values that return
an integer, allowing to indicate failures. Convert the driver to using
them.

Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Link: https://lore.kernel.org/r/20250407-gpiochip-set-rv-gpio-part1-v1-7-78399683ca38@linaro.org
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

gpio: f7188: use new line value setter callbacks

struct gpio_chip now has callbacks for setting line values that return
an integer, allowing to indicate failures. Convert the driver to using
them.

Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Link: https://lore.kernel.org/r/20250407-gpiochip-set-rv-gpio-part1-v1-6-78399683ca38@linaro.org
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

gpio: exar: use new line value setter callbacks

struct gpio_chip now has callbacks for setting line values that return
an integer, allowing to indicate failures. Convert the driver to using
them.

Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Link: https://lore.kernel.org/r/20250407-gpiochip-set-rv-gpio-part1-v1-4-78399683ca38@linaro.org
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

gpio: em: use new line value setter callbacks

struct gpio_chip now has callbacks for setting line values that return
an integer, allowing to indicate failures. Convert the driver to using
them.

Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Link: https://lore.kernel.org/r/20250407-gpiochip-set-rv-gpio-part1-v1-3-78399683ca38@linaro.org
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

gpio: eic-sprd: use new line value setter callbacks

struct gpio_chip now has callbacks for setting line values that return
an integer, allowing to indicate failures. Convert the driver to using
them.

Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Link: https://lore.kernel.org/r/20250407-gpiochip-set-rv-gpio-part1-v1-2-78399683ca38@linaro.org
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

gpio: dln2: use new line value setter callbacks

struct gpio_chip now has callbacks for setting line values that return
an integer, allowing to indicate failures. Convert the driver to using
them.

Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Link: https://lore.kernel.org/r/20250407-gpiochip-set-rv-gpio-part1-v1-1-78399683ca38@linaro.org
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

gpio: GPIO_BLZP1600 should depend on ARCH_BLAIZE

The Blaize BLZP1600 GPIO controller is only present on BLaize BLZP1600
SoCs.  Hence add a dependency on ARCH_BLAIZE, to prevent asking the user
about this driver when configuring a kernel without Blaize platform
support.

Fixes: 602ae04f56542d8e ("gpio: Enable Blaize BLZP1600 GPIO support")
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: Nikolaos Pasaloukos <nikolaos.pasaloukos@blaize.com>
Link: https://lore.kernel.org/r/a75af4b623579a193b4a59350c063969aabb5b71.1744190524.git.geert+renesas@glider.be
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

selftests: gpio: gpio-aggregator: add a test case for _sysfs prefix reservation

The kernel doc for gpio-aggregator configfs interface, which was recently
added, states that users should not be able to create an aggregator with a
name prefixed by "_sysfs" via configfs. However, it was found that this
guard does not function as expected (thanks to Dan Carpenter for
identifying and fixing the issue).

Add a test case to verify the guard.

Signed-off-by: Koichiro Den <koichiro.den@canonical.com>
Link: https://lore.kernel.org/r/20250412150119.1461023-1-koichiro.den@canonical.com
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

gpio: aggregator: Fix leak in gpio_aggregator_parse()

Call gpio_aggregator_free_lines() before returning on this error path.

Fixes: 83c8e3df642f ("gpio: aggregator: expose aggregator created via legacy sysfs to configfs")
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Acked-by: Koichiro Den <koichiro.den@canonical.com>
Link: https://lore.kernel.org/r/e023bfe52509ce1bef6209ec7c47e99279c551dd.1744452787.git.dan.carpenter@linaro.org
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

gpio: aggregator: Fix error code in gpio_aggregator_activate()

Propagate the error code if gpio_aggregator_make_device_sw_node() fails.
Don't return success.

Fixes: 86f162e73d2d ("gpio: aggregator: introduce basic configfs interface")
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Acked-by: Koichiro Den <koichiro.den@canonical.com>
Link: https://lore.kernel.org/r/79b804a0769a434698616bebedacc0e5d5605fdc.1744452787.git.dan.carpenter@linaro.org
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

gpio: aggregator: Return an error if there are no GPIOs in gpio_aggregator_parse()

The error handling in gpio_aggregator_parse() was re-written.  It now
returns success if there are no GPIOs.  Restore the previous behavior
and return -EINVAL instead.

Fixes: 83c8e3df642f ("gpio: aggregator: expose aggregator created via legacy sysfs to configfs")
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Acked-by: Koichiro Den <koichiro.den@canonical.com>
Link: https://lore.kernel.org/r/9dcd5fda7a3819e896d9eee4156e7c46c9a64595.1744452787.git.dan.carpenter@linaro.org
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

gpio: aggregator: Fix gpio_aggregator_line_alloc() checking

The gpio_aggregator_line_alloc() function returns error pointers, but
the callers check for NULL.  Update the error checking in the callers.

Fixes: 83c8e3df642f ("gpio: aggregator: expose aggregator created via legacy sysfs to configfs")
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Acked-by: Koichiro Den <koichiro.den@canonical.com>
Link: https://lore.kernel.org/r/cc71d8cf6e9bb4bb8cd9ae5050100081891d9345.1744452787.git.dan.carpenter@linaro.org
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

gpio: aggregator: fix "_sysfs" prefix check in gpio_aggregator_make_group()

This code is intended to reject strings that start with "_sysfs" but the
strcmp() limit is wrong so checks the whole string instead of the prefix.

Fixes: 83c8e3df642f ("gpio: aggregator: expose aggregator created via legacy sysfs to configfs")
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Acked-by: Koichiro Den <koichiro.den@canonical.com>
Link: https://lore.kernel.org/r/30210ed77b40b4b6629de659cb56b9ec7832c447.1744452787.git.dan.carpenter@linaro.org
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>

selftests: gpio: add test cases for gpio-aggregator

Add a set of tests for gpio-aggregator module. This test covers both
pre-existing new_device/delete_device interface and new configfs-based
interface.

Signed-off-by: Koichiro Den <koichiro.den@canonical.com>
Link: https://lore.kernel.org/r/20250407043019.4105613-10-koichiro.den@canonical.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

Documentation: gpio: document configfs interface for gpio-aggregator

Add documentation for the newly added configfs-based interface for GPIO
aggregator.

Signed-off-by: Koichiro Den <koichiro.den@canonical.com>
Link: https://lore.kernel.org/r/20250407043019.4105613-9-koichiro.den@canonical.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: aggregator: cancel deferred probe for devices created via configfs

For aggregators initialized via configfs, write 1 to 'live' waits for
probe completion and returns an error if the probe fails, unlike the
legacy sysfs interface, which is asynchronous.

Since users control the liveness of the aggregator device and might be
editing configurations while 'live' is 0, deferred probing is both
unnatural and unsafe.

Cancel deferred probe for purely configfs-based aggregators when probe
fails.

Signed-off-by: Koichiro Den <koichiro.den@canonical.com>
Link: https://lore.kernel.org/r/20250407043019.4105613-8-koichiro.den@canonical.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: aggregator: expose aggregator created via legacy sysfs to configfs

Expose settings for aggregators created using the sysfs 'new_device'
interface to configfs. Once written to 'new_device', an "_sysfs.<N>" path
appears in the configfs regardless of whether the probe succeeds.
Consequently, users can no longer use that prefix for custom GPIO
aggregator names. The 'live' attribute changes to 1 when the probe
succeeds and the GPIO forwarder is instantiated.

Note that the aggregator device created via sysfs is asynchronous, i.e.
writing into 'new_device' returns without waiting for probe completion,
and the probe may succeed, fail, or eventually succeed via deferred
probe. Thus, the 'live' attribute may change from 0 to 1 asynchronously
without notice. So, editing key/offset/name while it's waiting for
deferred probe is prohibited.

The configfs auto-generation relies on create_default_group(), which
inherently prohibits rmdir(2). To align with the limitation, this commit
also prohibits mkdir(2) for them. When users want to change the number
of lines for an aggregator initialized via 'new_device', they need to
tear down the device using 'delete_device' and reconfigure it from
scratch. This does not break previous behavior; users of legacy sysfs
interface simply gain additional almost read-only configfs exposure.

Still, users can write to the 'live' attribute to toggle the device
unless it's waiting for deferred probe. So once probe succeeds, they can
deactivate it in the same manner as the devices initialized via
configfs.

Signed-off-by: Koichiro Den <koichiro.den@canonical.com>
Link: https://lore.kernel.org/r/20250407043019.4105613-7-koichiro.den@canonical.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: aggregator: rename 'name' to 'key' in gpio_aggregator_parse()

Rename the local variable 'name' in gpio_aggregator_parse() to 'key'
because struct gpio_aggregator_line now uses the 'name' field for the
custom line name and the local variable actually represents a 'key'.
This change prepares for the next but one commit.

No functional change.

Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Koichiro Den <koichiro.den@canonical.com>
Link: https://lore.kernel.org/r/20250407043019.4105613-6-koichiro.den@canonical.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: aggregator: introduce basic configfs interface

The existing sysfs 'new_device' interface has several limitations:
* No way to determine when GPIO aggregator creation is complete.
* No way to retrieve errors when creating a GPIO aggregator.
* No way to trace a GPIO line of an aggregator back to its
  corresponding physical device.
* The 'new_device' echo does not indicate which virtual gpiochip<N>
  was created.
* No way to assign names to GPIO lines exported through an aggregator.

Introduce the new configfs interface for gpio-aggregator to address
these limitations. It provides a more streamlined, modern, and
extensible configuration method. For backward compatibility, the
'new_device' interface and its behavior is retained for now.

This commit implements basic functionalities:

  /config/gpio-aggregator/<name-of-your-choice>/
  /config/gpio-aggregator/<name-of-your-choice>/live
  /config/gpio-aggregator/<name-of-your-choice>/dev_name
  /config/gpio-aggregator/<name-of-your-choice>/<lineY>/
  /config/gpio-aggregator/<name-of-your-choice>/<lineY>/key
  /config/gpio-aggregator/<name-of-your-choice>/<lineY>/offset
  /config/gpio-aggregator/<name-of-your-choice>/<lineY>/name

Basic setup flow is:
1. Create a directory for a GPIO aggregator.
2. Create subdirectories for each line you want to instantiate.
3. In each line directory, configure the key and offset.
   The key/offset semantics are as follows:
   * If offset is >= 0:
     - key specifies the name of the chip this GPIO belongs to
     - offset specifies the line offset within that chip.
   * If offset is <0:
     - key needs to specify the GPIO line name.
4. Return to the aggregator's root directory and write '1' to the live
   attribute.

For example, the command in the existing kernel doc:

  echo 'e6052000.gpio 19 e6050000.gpio 20-21' > new_device

is equivalent to:

  mkdir /sys/kernel/config/gpio-aggregator/<custom-name>
  # Change <custom-name> to name of your choice (e.g. "aggr0")
  cd /sys/kernel/config/gpio-aggregator/<custom-name>
  mkdir line0 line1 line2  # Only "line<Y>" naming allowed.
  echo e6052000.gpio > line0/key
  echo 19            > line0/offset
  echo e6050000.gpio > line1/key
  echo 20            > line1/offset
  echo e6050000.gpio > line2/key
  echo 21            > line2/offset
  echo 1             > live

The corresponding gpio_device id can be identified as follows:

  cd /sys/kernel/config/gpio-aggregator/<custom-name>
  ls -d /sys/devices/platform/`cat dev_name`/gpiochip*

Also, via configfs, custom GPIO line name can be set like this:

  cd /sys/kernel/config/gpio-aggregator/<custom-name>
  echo "abc"         > line1/name

Signed-off-by: Koichiro Den <koichiro.den@canonical.com>
Link: https://lore.kernel.org/r/20250407043019.4105613-5-koichiro.den@canonical.com
[Bartosz: remove stray newlines]
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: aggregator: add gpio_aggregator_{alloc,free}()

Prepare for the upcoming configfs interface. These functions will be
used by both the existing sysfs interface and the new configfs
interface, reducing code duplication.

No functional change.

Signed-off-by: Koichiro Den <koichiro.den@canonical.com>
Link: https://lore.kernel.org/r/20250407043019.4105613-4-koichiro.den@canonical.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: aggregator: unify function naming

Unify function names to use gpio_aggregator_ prefix (except GPIO
forwarder implementations, which remain unchanged in subsequent
commits). While at it, rename the pre-existing gpio_aggregator_free() to
gpio_aggregator_destory(), since that name will be used by new
alloc/free functions introduced in the next commit, for which the name
is more appropriate.

No functional change.

Signed-off-by: Koichiro Den <koichiro.den@canonical.com>
Link: https://lore.kernel.org/r/20250407043019.4105613-3-koichiro.den@canonical.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpio: aggregator: reorder functions to prepare for configfs introduction

Reorder functions in drivers/gpio/gpio-aggregator.c to prepare for the
configfs-based interface additions in subsequent commits. Arrange the
code so that the configfs implementations will appear above the existing
sysfs-specific code, since the latter will partly depend on the configfs
interface implementations when it starts to expose the settings to
configfs.

The order in drivers/gpio/gpio-aggregator.c will be as follows:

* Basic gpio_aggregator/gpio_aggregator_line representations
* Common utility functions
* GPIO Forwarder implementations
* Configfs interface implementations
* Sysfs interface implementations
* Platform device implementations
* Module init/exit implementations

This separate commit ensures a clean diff for the subsequent commits.

No functional change.

Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Koichiro Den <koichiro.den@canonical.com>
Link: https://lore.kernel.org/r/20250407043019.4105613-2-koichiro.den@canonical.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

Merge tag 'gpiod-is-equal-for-v6.16-rc1' of gitolite.pub/scm/linux/kernel/git/brgl/linux into gpio/for-next

Immutable tag for the regulator tree to pull from

gpio: provide gpiod_is_equal()

gpio: provide gpiod_is_equal()

There are users in the kernel that directly compare raw GPIO descriptor
pointers in order to determine whether they refer to the same physical
GPIO pin. This accidentally works like this but is not guaranteed by any
API contract. Let's provide a comparator function that hides the actual
logic.

Reviewed-by: Mark Brown <broonie@kernel.org>
Link: https://lore.kernel.org/r/20250407-gpiod-is-equal-v1-1-7d85f568ae6e@linaro.org
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpiolib: support parsing gpio three-cell interrupts scheme

gpio irq which using three-cell scheme should always call
instance_match() function to find the correct irqdomain.

The select() function will be called with !DOMAIN_BUS_ANY,
so for specific gpio irq driver, it need to set bus token
explicitly, something like:
  irq_domain_update_bus_token(girq->domain, DOMAIN_BUS_WIRED);

Signed-off-by: Yixun Lan <dlan@gentoo.org>
Link: https://lore.kernel.org/r/20250408-04-gpio-irq-threecell-v4-1-fd170d5e2d2b@gentoo.org
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

Merge tag 'irqdomain-04-08-25' of git://git./linux/kernel/git/tip/tip into gpio/for-next

Stable tag for the GPIO tree to pull from

gpio: Enable Blaize BLZP1600 GPIO support

Blaize BLZP1600 GPIO controller is provided by VeriSilicon
Microelectronics based on the GPIO APB v0.2 design. It has 32
input/output ports which can be configured as edge or level
triggered interrupts. It also provides a de-bounce feature.
This controller is used on the Blaize BLZP1600 SoC.

Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Nikolaos Pasaloukos <nikolaos.pasaloukos@blaize.com>
Link: https://lore.kernel.org/r/20250327-kernel-upstreaming-add_gpio_support-v2-2-bbe51f8d66da@blaize.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

dt-bindings: Document Blaize BLZP1600 GPIO driver

This is a custom silicon GPIO driver provided by VeriSilicon
Microelectronics. It has 32 input/output ports which can be
configured as edge or level triggered interrupts. It also provides
a de-bounce feature.
This controller is used on the Blaize BLZP1600 SoC.

Signed-off-by: Nikolaos Pasaloukos <nikolaos.pasaloukos@blaize.com>
Reviewed-by: Conor Dooley <conor.dooley@microchip.com>
Link: https://lore.kernel.org/r/20250327-kernel-upstreaming-add_gpio_support-v2-1-bbe51f8d66da@blaize.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

dt-bindings: gpio: pca95xx: add Toradex ecgpiol16

The Toradex ecgpiol16 is a 16-bit I2C I/O expander implemented using a
small MCU.
Its register interface and behavior are compatible with the PCAL6416.

Signed-off-by: Emanuele Ghidoli <emanuele.ghidoli@toradex.com>
Signed-off-by: Francesco Dolcini <francesco.dolcini@toradex.com>
Acked-by: Conor Dooley <conor.dooley@microchip.com>
Link: https://lore.kernel.org/r/20250331072644.17921-1-francesco@dolcini.it
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

dt-bindings: gpio: Correct indentation and style in DTS example

DTS example in the bindings should be indented with 2- or 4-spaces and
aligned with opening '- |', so correct any differences like 3-spaces or
mixtures 2- and 4-spaces in one binding.  While re-indenting, drop
unused labels.

No functional changes here, but saves some comments during reviews of
new patches built on existing code.

Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Reviewed-by: Sander Vanheule <sander@svanheule.net>
Reviewed-by: Manikandan Muralidharan <manikandan.m@microchip.com>
Reviewed-by: Rob Herring (Arm) <robh@kernel.org>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Link: https://lore.kernel.org/r/20250324125326.82270-1-krzysztof.kozlowski@linaro.org
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

Merge tag 'gpio-s5m8767-polarity-quirk-for-v6.16-rc1' into gpio/for-next

gpiolib: of: Add polarity quirk for s5m8767

irqdomain: Support three-cell scheme interrupts

Add new function *_twothreecell() to extend support to parse three-cell
interrupts which encoded as <instance hwirq irqflag>, the translate
function will retrieve irq number and flag from last two cells.

This API will be used in gpio irq driver which need to work with
two or three cells cases.

Signed-off-by: Yixun Lan <dlan@gentoo.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/all/20250326-04-gpio-irq-threecell-v3-1-aab006ab0e00@gentoo.org

gpiolib: of: Add polarity quirk for s5m8767

This is prepare patch for switching s5m8767 regulator driver to
use GPIO descriptor. DTS for exynos5250 spring incorrectly specifies
"active low" polarity for the DVS and DS line. But per datasheet,
they are actually active high. So add polarity quirk for it.

Signed-off-by: Peng Fan <peng.fan@nxp.com>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20250327004945.563765-1-peng.fan@oss.nxp.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

gpiolib: don't allow setting values on input lines

Some drivers as well as the character device and sysfs code check
whether the line actually is in output mode before allowing the user to
set a value.

However, GPIO value setters now return integer values and can indicate
failures. This allows us to move these checks into the core code.

Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20250311-gpio-set-check-output-v1-1-d971bca9e6fa@linaro.org
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>