Re: [PATCH v3 6/6] mfd: max77620: Provide system power-off functionality

[Thierry Reding <thierry.reding@gmail.com>]

[-- Attachment #1: Type: text/plain, Size: 2801 bytes --]

On Thu, Apr 25, 2019 at 01:49:00AM +0300, Dmitry Osipenko wrote:
> Provide system power-off functionality that allows to turn off machine
> gracefully.
> 
> Signed-off-by: Dmitry Osipenko <digetx@gmail.com>
> ---
>  drivers/mfd/max77620.c | 19 +++++++++++++++++++
>  1 file changed, 19 insertions(+)
> 
> diff --git a/drivers/mfd/max77620.c b/drivers/mfd/max77620.c
> index 9b0009c29610..e56223bde568 100644
> --- a/drivers/mfd/max77620.c
> +++ b/drivers/mfd/max77620.c
> @@ -37,6 +37,8 @@
>  #include <linux/regmap.h>
>  #include <linux/slab.h>
>  
> +static struct max77620_chip *max77620_scratch;
> +
>  static const struct resource gpio_resources[] = {
>  	DEFINE_RES_IRQ(MAX77620_IRQ_TOP_GPIO),
>  };
> @@ -481,6 +483,15 @@ static int max77620_read_es_version(struct max77620_chip *chip)
>  	return ret;
>  }
>  
> +static void max77620_pm_power_off(void)
> +{
> +	struct max77620_chip *chip = max77620_scratch;
> +
> +	regmap_update_bits(chip->rmap, MAX77620_REG_ONOFFCNFG1,
> +			   MAX77620_ONOFFCNFG1_SFT_RST,
> +			   MAX77620_ONOFFCNFG1_SFT_RST);
> +}

I think this is only partially correct. See here for a driver that I had
proposed a while back:

	https://github.com/thierryreding/linux/commit/d0eaa77b402f62bd236d76e3edeb3ccf296cbe81

Note that that driver is part of a larger series to move away from all
the pm_power_off hackery. There was a fair bit of discussion back when I
proposed the original power off driver for max77620:

	https://lkml.org/lkml/2017/1/12/470

I think I may have a more up-to-date local branch of the system-power
branch from my github repository if you're interested in looking at some
of that code. There wasn't a whole lot of feedback on the patches, but
the feedback I did get was generally positive. However, since it didn't
gain any traction I eventually abandoned that effort. It might be worth
picking it up again, since, as far as I can tell, the situation around
power off and restart hasn't changed in the meantime.

Thierry

> +
>  static int max77620_probe(struct i2c_client *client,
>  			  const struct i2c_device_id *id)
>  {
> @@ -488,6 +499,7 @@ static int max77620_probe(struct i2c_client *client,
>  	struct max77620_chip *chip;
>  	const struct mfd_cell *mfd_cells;
>  	int n_mfd_cells;
> +	bool pm_off;
>  	int ret;
>  
>  	chip = devm_kzalloc(&client->dev, sizeof(*chip), GFP_KERNEL);
> @@ -554,6 +566,13 @@ static int max77620_probe(struct i2c_client *client,
>  		return ret;
>  	}
>  
> +	pm_off = of_property_read_bool(client->dev.of_node,
> +				       "maxim,system-power-controller");
> +	if (pm_off && !pm_power_off) {
> +		max77620_scratch = chip;
> +		pm_power_off = max77620_pm_power_off;
> +	}
> +
>  	return 0;
>  }
>  
> -- 
> 2.21.0
> 

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]