The gpiod functions include variants for managed gpiod resources. Use it
to simplify the remove function.

As the driver handles a device node without a specification of dvs gpios
just fine, additionally use the variant of gpiod_get exactly for this
use case. This makes error checking more strict.

As a third benefit this patch makes the driver use the flags parameter
of gpiod_get* which will not be optional any more after 4.2 and so
prevents a build failure when the respective gpiod commit is merged.
Signed-off-by: NUwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: NMark Brown <broonie@kernel.org>

This fixes a build problem on mips found by the kbuild test robot:

drivers/regulator/rk808-regulator.c: In function 'rk808_buck1_2_get_voltage_sel_regmap':
drivers/regulator/rk808-regulator.c:97:2: error: implicit declaration of function 'gpiod_get_value' [-Werror=implicit-function-declaration]
  if (IS_ERR(gpio) || gpiod_get_value(gpio) == 0)
  ^

Fixes: bad47ad2 ("regulator: rk808: fixed the overshoot when adjust voltage")
Signed-off-by: NUwe Kleine-König <uwe@kleine-koenig.org>
Signed-off-by: NMark Brown <broonie@kernel.org>

There is a overshoot in DCDC1/DCDC2, we have 2 method to workaround:
1st is use dvs pin to switch the voltage between value in BUCKn_ON_VSEL
and BUCKn_DVS_VSEL. If DVS pin is inactive, the voltage of DCDC1/DCDC2
are controlled by BUCKn_ON_VSEL, when we pull dvs1/dvs2 pin to active,
they would be controlled by BUCKn_DVS_VSEL. In this case, the ramp rate
is same as the value programmed in BUCKn_RATE, and the fastest rate is
10mv/us.
2nd method is gradual adjustment, adjust the voltage to a target value
step by step via i2c, each step is set to 100 mA. If you write the
voltage directly using an i2c write the rk808 will always ramp as fast
as it possibly can, about 100mv/us.
Signed-off-by: NChris Zhong <zyw@rock-chips.com>
Reviewed-by: NDoug Anderson <dianders@chromium.org>
Signed-off-by: NMark Brown <broonie@kernel.org>

The LDOs are documented in the rk808 datasheet to have a soft start
time of 400us.  Add that to the driver.  If this time takes longer on
a certain board the device tree should be able to override with
"regulator-enable-ramp-delay".

This fixes some dw_mmc probing problems (together with other patches
posted to the mmc maiing lists) on rk3288.
Signed-off-by: NDoug Anderson <dianders@chromium.org>
Signed-off-by: NMark Brown <broonie@kernel.org>
Cc: stable@vger.kernel.org

Fixes the following sparse warnings:

drivers/regulator/rk808-regulator.c:100:5: warning:
 symbol 'rk808_set_suspend_voltage' was not declared. Should it be static?
drivers/regulator/rk808-regulator.c:115:5: warning:
 symbol 'rk808_set_suspend_enable' was not declared. Should it be static?
drivers/regulator/rk808-regulator.c:126:5: warning:
 symbol 'rk808_set_suspend_disable' was not declared. Should it be static?
Signed-off-by: NWei Yongjun <yongjun_wei@trendmicro.com.cn>
Signed-off-by: NMark Brown <broonie@kernel.org>

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>

support setting suspend voltage and disable regulator in suspend.
Signed-off-by: NChris Zhong <zyw@rock-chips.com>
Reviewed-by: NDoug Anderson <dianders@chromium.org>
Signed-off-by: NMark Brown <broonie@kernel.org>

The min_uv in DCDC1 & DCDC2 should be 712.5mv
Signed-off-by: NChris Zhong <zyw@rock-chips.com>
Reviwed-by: NDoug Anderson <dianders@chromium.org>
Signed-off-by: NMark Brown <broonie@kernel.org>

On rk808 buck1 and buck2 have programmable ramp delays.  Let's add a
function to allow a client of rk808 to set them.
Signed-off-by: NDoug Anderson <dianders@chromium.org>
Signed-off-by: NMark Brown <broonie@kernel.org>

1. Pass &pdev->dev rather than &client->dev to of_regulator_match, the *dev
   argument is used for devres to ensure devm_of_regulator_put_matches() will
   be called when unload the module.

2. of_get_child_by_name() returns a node pointer with refcount incremented.
   Thus add missing of_node_put(reg_np).
Signed-off-by: NAxel Lin <axel.lin@ingics.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Also remove non-informative comment.
Signed-off-by: NAxel Lin <axel.lin@ingics.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Signed-off-by: NChris Zhong <zyw@rock-chips.com>
Tested-by: NHeiko Stuebner <heiko@sntech.de>
Reviewed-by: NDoug Anderson <dianders@chromium.org>
Tested-by: NDoug Anderson <dianders@chromium.org>
Signed-off-by: NMark Brown <broonie@kernel.org>

The original RK808 regulator driver didn't setup input supplies
properly.  Add them.
Signed-off-by: NDoug Anderson <dianders@chromium.org>
Signed-off-by: NMark Brown <broonie@kernel.org>

The min_sel is 0, max_sel is 15, so n_voltages should be 16.
Signed-off-by: NAxel Lin <axel.lin@ingics.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

The memory allocated in rk808_regulator_probe() needs to be freed when the
module is unloaded. Thus pass &pdev->dev rather than &client->dev to
devm_kzalloc.
Signed-off-by: NAxel Lin <axel.lin@ingics.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

remove the redundant code, since pdata has been removed from stuct rk808
Signed-off-by: NChris Zhong <zyw@rock-chips.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

The "dev" has been deleted from "struct rk808" in rk808 mfd driver
so rk808->dev should be replaced by &client->dev here.
Signed-off-by: NChris Zhong <zyw@rock-chips.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

The RK808 regulator driver was putting its config on the stack but not
initting it.  That means that you got a semi-random config.  Fix this.
Signed-off-by: NDoug Anderson <dianders@chromium.org>
Signed-off-by: NMark Brown <broonie@linaro.org>

The regulator module consists of 4 DCDCs, 8 LDOs and 2 switches.
The output voltages are configurable and are meant to supply power
to the main processor and other components
Signed-off-by: NChris Zhong <zyw@rock-chips.com>
Signed-off-by: NMark Brown <broonie@linaro.org>