This reverts the main change of commit bff23714 ("leds: leds-gpio:
Set of_node for created LED devices") because of_node assignment is
handled by core since commit 7ea79ae8 ("leds: gpio: use OF variant
of LED registering function").  Basically the code was overwriting the
of_node with same value.

No functional change expected.
Signed-off-by: NKrzysztof Kozlowski <krzk@kernel.org>
Tested-by: NRafał Miłecki <rafal@milecki.pl>
Signed-off-by: NJacek Anaszewski <jacek.anaszewski@gmail.com>

Commit 45d4c6de ("leds: gpio: Try to lookup gpiod from device")
removed the common code path that set the led_dat->gpiod pointer in
create_gpio_led(), but only added it back for the devices that have
a valid gpio_led_platform_data structure. Calling gpio_leds_create()
in gpio_led_probe() is not enough to get a valid gpiod pointer.

Fortunately enough, gpio_leds_create() already gets the needed pointer,
we just need to assign it to the relevant gpio_led_data structure.

Fixes: 45d4c6de ("leds: gpio: Try to lookup gpiod from device")
Signed-off-by: NLiviu Dudau <liviu@dudau.co.uk>
Reviewed-by: NLinus Walleij <linus.walleij@linaro.org>
Acked-by: NPavel Machek <pavel@ucw.cz>
Signed-off-by: NJacek Anaszewski <jacek.anaszewski@gmail.com>

This augments the GPIO lookup code in the GPIO LEDs to
attempt to look up a GPIO descriptor from the device with
index.

This makes it possible to use GPIO machine look-up tables
and stop passing global GPIO numbers through platform data.

Using this we can stepwise convert existing board files
to use machine descriptor tables and then eventually drop
the legacy GPIO support and only include
<linux/gpio/consumer.h> and use descriptors exclusively.
Signed-off-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NJacek Anaszewski <jacek.anaszewski@gmail.com>

In some systems, such as Baseboard Management Controllers (BMCs), we
want to retain the state of LEDs across a reboot of the BMC (whilst the
host remains up). Implement support for the retain-state-shutdown
devicetree property in leds-gpio.
Signed-off-by: NAndrew Jeffery <andrew@aj.id.au>
Acked-by: NPavel Machek <pavel@ucw.cz>
Tested-by: NBrandon Wyman <bjwyman@gmail.com>
Signed-off-by: NJacek Anaszewski <jacek.anaszewski@gmail.com>

In leds-gpio we support LEDs specified in DT so we should use
(devm_)of_led_classdev_register. This allows passing DT node as argument
for use by the LED subsystem.
Signed-off-by: NRafał Miłecki <rafal@milecki.pl>
Acked-by: NPavel Machek <pavel@ucw.cz>
Signed-off-by: NJacek Anaszewski <jacek.anaszewski@gmail.com>

Rename devm_get_gpiod_from_child() into
devm_fwnode_get_gpiod_from_child() to reflect the fact that this
function is operating on a fwnode object.
Signed-off-by: NBoris Brezillon <boris.brezillon@free-electrons.com>
Acked-by: NJacek Anaszewski <jacek.anaszewski@gmail.com>
Acked-by: NDmitry Torokhov <dmitry.torokhov@gmail.com>
Signed-off-by: NLinus Walleij <linus.walleij@linaro.org>

Currently all users of fwnode_get_named_gpiod() have no way to
specify a label for the GPIO. So GPIOs listed in debugfs are shown
with label "?". With this change a proper label is used.

Also adjust all users so they can pass a label, properly retrieved
from device tree properties.
Signed-off-by: NAlexander Stein <alexander.stein@systec-electronic.com>
Cc: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Acked-by: NJacek Anaszewski <jacek.anaszewski@gmail.com>
Acked-by: NBartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Acked-by: NDmitry Torokhov <dmitry.torokhov@gmail.com>
Signed-off-by: NLinus Walleij <linus.walleij@linaro.org>

Make fwnode_get_named_gpiod() consistent with the rest of
gpiod_get() like API, i.e. configure GPIO pin immediately after
request.

Besides obvious clean up it will help to configure pins based
on firmware provided resources.
Reviewed-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NLinus Walleij <linus.walleij@linaro.org>

Simplify the error handling and add a missing call to fwnode_handle_put
when checking led.name.
Signed-off-by: NHeiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: NJacek Anaszewski <j.anaszewski@samsung.com>

Using the managed version of led_classdev_register allows to
significantly simplify the code.
Signed-off-by: NHeiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: NJacek Anaszewski <j.anaszewski@samsung.com>

Checking for the presence of the property first isn't strictly needed
as we can react on the return code of fwnode_property_read_string.
Also, even if the presence of a property "label" was checked,
reading a string value for it theoretically still can fail and
this case isn't handled.
Signed-off-by: NHeiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: NJacek Anaszewski <j.anaszewski@samsung.com>

Definition of np can be moved into the loop as well to simplify
the code a little.
Signed-off-by: NHeiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: NJacek Anaszewski <j.anaszewski@samsung.com>

Add a helper for the container_of as it's used more than once.
Signed-off-by: NHeiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: NJacek Anaszewski <j.anaszewski@samsung.com>

gpiod_get_value_cansleep returns 0, 1, or an error code.
So far errors are not handled and treated the same as 1.
Change this to bail out if an error code is returned and
remove the double negation.
Signed-off-by: NHeiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: NJacek Anaszewski <j.anaszewski@samsung.com>

Introduce a typedef gpio_blink_set_t to improve readability of the code.
Signed-off-by: NHeiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: NJacek Anaszewski <j.anaszewski@samsung.com>

When working with Device Tree we iterate over children of "gpio-leds"
compatible node and create LED device for each of them. We take care of
all common DT properties (label, default trigger, state, etc.), however
some triggers may want to support more of them.

It could be useful for timer trigger to support setting delay on/off
values with Device Tree property. Or for transient trigger to support
e.g. duration property.

We obviously should handle such properties in triggers, not in generic
code. To let trigger drivers read properties from DT node we need to set
of_node to point the relevant node. This change allows using all kind of
of helpers in e.g. "activate" callbacks.
Signed-off-by: NRafał Miłecki <zajec5@gmail.com>
Signed-off-by: NJacek Anaszewski <j.anaszewski@samsung.com>

Calling a GPIO LEDs is quite likely to work even if the kernel
has paniced, so they are ideal to blink in this situation.
This commit adds support for the new "panic-indicator"
firmware property, allowing to mark a given LED to blink on
a kernel panic.
Signed-off-by: NEzequiel Garcia <ezequiel@vanguardiasur.com.ar>
Reviewed-by: NMatthias Brugger <mbrugger@suse.com>
Acked-by: NPavel Machek <pavel@ucw.cz>
Signed-off-by: NJacek Anaszewski <j.anaszewski@samsung.com>

Use the GPIO flag definition instead of a numeric literal, so the casual
reader grepping for GPIOF_ will find the GPIO flags used.
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NJacek Anaszewski <j.anaszewski@samsung.com>

In case of error, the function gpio_to_desc() returns NULL
pointer not ERR_PTR(). The IS_ERR() test in the return value
check should be replaced with NULL test.
Signed-off-by: NWei Yongjun <yongjun_wei@trendmicro.com.cn>
Signed-off-by: NJacek Anaszewski <j.anaszewski@samsung.com>

Now the core implements the work queue, remove it from the drivers,
and switch to using brightness_set_blocking op.
Signed-off-by: NJacek Anaszewski <j.anaszewski@samsung.com>
Cc: Raphael Assenat <raph@8d.com>

add a shutdown function for setting the gpio-leds
into off state when shuting down.
Signed-off-by: NHeiko Schocher <hs@denx.de>
Signed-off-by: NJacek Anaszewski <j.anaszewski@samsung.com>

Commit 8a0662d9 introduced of_node and acpi_node symbols in global namespace
but there were already ~63 of_node local variables or function parameters
(no single acpi_node though, but anyway).

After debugging undefined but used of_node local varible (which turned out
to reference static function of_node() instead) it became clear that the names
for the functions are too short and too generic for global scope.
Signed-off-by: NAlexander Sverdlin <alexander.sverdlin@gmail.com>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

Signed-off-by: NGeert Uytterhoeven <geert@linux-m68k.org>
Acked-by: NJacek Anaszewski <j.anaszewski@samsung.com>
Signed-off-by: NBryan Wu <cooloney@gmail.com>

Make sure that already registered LEDs will be torn down properly
if the name of the next LED to create is unavailable.
Signed-off-by: NJacek Anaszewski <j.anaszewski@samsung.com>
Cc: Richard Purdie <rpurdie@rpsys.net>
Signed-off-by: NBryan Wu <cooloney@gmail.com>

In gpio_leds_create(), when devm_get_gpiod_from_child() fails with
-EPROBE_DEFER on the second gpio led to be created, the first already
registered led is not torn down properly. This causes create_gpio_led()
to fail for the first led on re-probe().

Fix this misbehaviour by incrementing num_leds only if all
potentially failing calls completed successfully.
Signed-off-by: NSebastian Hesselbarth <sebastian.hesselbarth@gmail.com>
Acked-by: NJacek Anaszewski <j.anaszewski@samsung.com>
Signed-off-by: NBryan Wu <cooloney@gmail.com>

The gpio binding document says that new code should always use named
gpios. Patch 40b73183 added support to parse a list of gpios from child
nodes, but does not make it possible to use named gpios. This patch adds
the con_id property and implements it is done in gpiolib.c, where the
old-style of using unnamed gpios still works.
Signed-off-by: NOlliver Schinagl <oliver@schinagl.nl>
Acked-by: NBryan Wu <cooloney@gmail.com>
Acked-by: NDmitry Torokhov <dmitry.torokhov@gmail.com>
Reviewed-by: NAlexandre Courbot <acourbot@nvidia.com>
Signed-off-by: NLinus Walleij <linus.walleij@linaro.org>

Instead of overriding error codes, pass them on unmodified. This
way a EPROBE_DEFER is correctly passed to the driver core. This results
in the LED driver correctly requesting probe deferral in cases the GPIO
controller is not yet available.
Signed-off-by: NSoren Brinkmann <soren.brinkmann@xilinx.com>
Reported-and-tested-by: NAndreas Färber <afaerber@suse.de>
Signed-off-by: NBryan Wu <cooloney@gmail.com>

Commit a43f2cbb ("leds: leds-gpio: Make use of device property API")
caused a regression by reading the incorrect string for the "default-state"
property.

Fix the property string check to restore the original behaviour.

Fixes: a43f2cbb (leds: leds-gpio: Make use of device property API)
Reported-by: NJean-Michel Hautbois <jean-michel.hautbois@vodalys.com>
Signed-off-by: NFabio Estevam <fabio.estevam@freescale.com>
Acked-by: NBryan Wu <cooloney@gmail.com>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

Since commit a43f2cbb ("leds: leds-gpio: Make use of device property
API") it is no longer possible to register multiple gpio leds without passing
the 'label' property.

According to Documentation/devicetree/bindings/leds/common.txt:

"Optional properties for child nodes:
- label : The label for this LED.  If omitted, the label is
  taken from the node name (excluding the unit address)."

So retrieve the node name when the 'label' property is absent to keep the old
behaviour and fix this regression.

Fixes: a43f2cbb (leds: leds-gpio: Make use of device property API)
Reported-by: NJean-Michel Hautbois <jean-michel.hautbois@vodalys.com>
Signed-off-by: NFabio Estevam <fabio.estevam@freescale.com>
Acked-by: NGrant Likely <grant.likely@linaro.org>
Acked-by: NBryan Wu <cooloney@gmail.com>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

In the legacy case, led_dat->gpiod is initialized correctly, but
overwritten later by template->gpiod, which is NULL, causing leds-gpio
to fail with:

    gpiod_direction_output: invalid GPIO
    leds-gpio: probe of leds-gpio failed with error -22

Move the initialization of led_dat->gpiod from template->gpiod up, and
always use led_dat->gpiod later, to fix this.

Fixes: 5c51277a (leds: leds-gpio: Add support for GPIO descriptors)
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

Commit 21f2aae91e902aad ("leds: leds-gpio: Add support for GPIO
descriptors") already converted most of the driver to use GPIO descriptors.
What is still missing is the platform specific hook gpio_blink_set() and
board files which pass legacy GPIO numbers to this driver in platform data.

In this patch we handle the former and convert gpio_blink_set() to take
GPIO descriptor instead. In order to do this we convert the existing four
users to accept GPIO descriptor and translate it to legacy GPIO number in
the platform code. This effectively "pushes" legacy GPIO number usage from
the driver to platforms.

Also add comment to the remaining block describing that it is legacy code
path and we are getting rid of it eventually.
Suggested-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Acked-by: NAndrew Lunn <andrew@lunn.ch>
Reviewed-by: NLinus Walleij <linus.walleij@linaro.org>
Acked-by: NAlexandre Courbot <acourbot@nvidia.com>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

Make use of device property API in this driver so that both OF and ACPI
based system can use the same driver.

This change contains material from Max Eliaser and Mika Westerberg.
Signed-off-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Acked-by: NBryan Wu <cooloney@gmail.com>
Acked-by: NGrant Likely <grant.likely@linaro.org>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

GPIO descriptors are the preferred way over legacy GPIO numbers
nowadays. Convert the driver to use GPIO descriptors internally but
still allow passing legacy GPIO numbers from platform data to support
existing platforms.
Signed-off-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Acked-by: NAlexandre Courbot <acourbot@nvidia.com>
Acked-by: NBryan Wu <cooloney@gmail.com>
Acked-by: NArnd Bergmann <arnd@arndb.de>
Acked-by: NGrant Likely <grant.likely@linaro.org>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

A platform_driver does not need to set an owner, it will be populated by the
driver core.
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

Remove stray blank line and space.
Signed-off-by: NXiubo Li <Li.Xiubo@freescale.com>
Signed-off-by: NBryan Wu <cooloney@gmail.com>

If the inlcude headers aren't sorted alphabetically, then the
logical choice is to append new ones, however that creates a
lot of potential for conflicts or duplicates because every change
will then add new includes in the same location.
Signed-off-by: NXiubo Li <Li.Xiubo@freescale.com>
Signed-off-by: NBryan Wu <cooloney@gmail.com>

Enable autoloading of leds-gpio module when a corresponing DT entry is present.
Signed-off-by: NPaolo Pisati <paolo.pisati@canonical.com>
Signed-off-by: NBryan Wu <cooloney@gmail.com>

None of these files are actually using any __init type directives
and hence don't need to include <linux/init.h>.  Most are just a
left over from __devinit and __cpuinit removal, or simply due to
code getting copied from one driver to the next.

Cc: Bryan Wu <cooloney@gmail.com>
Cc: Richard Purdie <rpurdie@rpsys.net>
Cc: linux-leds@vger.kernel.org
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: NBryan Wu <cooloney@gmail.com>

Some gpio-leds need retain the state even in suspend, such as charger led.
But this property missed in devicetree, add it.

(cooloney@gmail.com: fold DT binding updates into this patch)
Signed-off-by: NRobin Gong <b38343@freescale.com>
Signed-off-by: NBryan Wu <cooloney@gmail.com>

now the leds-gpio driver will create every child led node without
checking the status is disabled or not.

for example, if we have a led node like d3, and its status is disabled:
	leds {
		d3 {
			label = "d3";
			gpios = <&pioE 24 0>;
			status = "disabled";
		};
	};

we except the d3 should not be created. And the gpios should not be
request as well.

But current driver will create d3 and request its gpio.

This patch fix this by using for_each_available_child_of_node() and
of_get_available_child_count() to enumerate all child nodes. So the
disabled node will be inavailable.
Signed-off-by: NJosh Wu <josh.wu@atmel.com>
Signed-off-by: NBryan Wu <cooloney@gmail.com>