[PATCH linux dev-4.10 v2 1/3] drivers/hwmon: Document optional dev tree property 'fault-max-fan'

[Joel Stanley]

On Fri, Jun 16, 2017 at 5:30 AM, Christopher Bostic
<cbostic at linux.vnet.ibm.com> wrote:
> Document new optional device tree property 'fault-max-fan'.  This
> property indicates that the max31785 should be configured to drive
> fan with 100% PWM duty cycle on fault condition.
>
> Signed-off-by: Christopher Bostic <cbostic at linux.vnet.ibm.com>
> ---
>  Documentation/devicetree/bindings/hwmon/max31785.txt | 20 ++++++++++++++++++++
>  1 file changed, 20 insertions(+)
>  create mode 100644 Documentation/devicetree/bindings/hwmon/max31785.txt
>
> diff --git a/Documentation/devicetree/bindings/hwmon/max31785.txt b/Documentation/devicetree/bindings/hwmon/max31785.txt
> new file mode 100644
> index 0000000..2b0190d
> --- /dev/null
> +++ b/Documentation/devicetree/bindings/hwmon/max31785.txt
> @@ -0,0 +1,20 @@
> +Bindings for MAX31785 I2C fan controller
> +
> +Reference:
> +[1]    https://datasheets.maximintegrated.com/en/ds/MAX31785.pdf
> +
> +Required properties:
> +- compatible : "maxim,max31785"
> +- reg        : I2C address, 0x52.
> +
> +Optional properties, default is to retain the chip's current setting:
> +- fault-max-fan : When present this property indicates to set fan to 100% PWM
> +                  duty cycle on FAULT condition.

The hardware has 6 pages that can be configured to assert this on
fault. We should describe that in the binding.

Should the binding describe the four pages as separate properties, or
just the one?

One option is to say: we are describing the presence of the wire that
connects the fault pin to something, so the presence/absence of the
hardware is all that will live in the binding?

Cheers,

Joel

>
> +Example:
> +
> +        max31785 at 52 {
> +                compatible = "maxim,max31785";
> +                reg = <0x52>;
> +                fault-max-fan;
> +        };
> --
> 1.8.2.2
>