Re: [PATCH v8 09/12] dt-bindings: mux-adg792a: document devicetree bindings for ADG792A/G mux

[Jonathan Cameron <jic23@kernel.org>]

On 27/01/17 22:09, Peter Rosin wrote:
> On 2017-01-27 20:50, Rob Herring wrote:
>> On Wed, Jan 18, 2017 at 04:57:12PM +0100, Peter Rosin wrote:
>>> Analog Devices ADG792A/G is a triple 4:1 mux.
>>>
>>> Acked-by: Jonathan Cameron <jic23@kernel.org>
>>> Signed-off-by: Peter Rosin <peda@axentia.se>
>>> ---
>>>  .../devicetree/bindings/mux/mux-adg792a.txt        | 79 ++++++++++++++++++++++
>>>  1 file changed, 79 insertions(+)
>>>  create mode 100644 Documentation/devicetree/bindings/mux/mux-adg792a.txt
>>>
>>> diff --git a/Documentation/devicetree/bindings/mux/mux-adg792a.txt b/Documentation/devicetree/bindings/mux/mux-adg792a.txt
>>> new file mode 100644
>>> index 000000000000..0b26dd11f070
>>> --- /dev/null
>>> +++ b/Documentation/devicetree/bindings/mux/mux-adg792a.txt
>>> @@ -0,0 +1,79 @@
>>> +Bindings for Analog Devices ADG792A/G Triple 4:1 Multiplexers
>>> +
>>> +Required properties:
>>> +- compatible : "adi,adg792a" or "adi,adg792g"
>>> +- #mux-control-cells : <0> if parallel, or <1> if not.
>>> +* Standard mux-controller bindings as decribed in mux-controller.txt
>>> +
>>> +Optional properties for ADG792G:
>>> +- gpio-controller : if present, #gpio-cells below is required.
>>> +- #gpio-cells : should be <2>
>>> +			  - First cell is the GPO line number, i.e. 0 or 1
>>> +			  - Second cell is used to specify active high (0)
>>> +			    or active low (1)
>>> +
>>> +Optional properties:
>>> +- adi,parallel : if present, the three muxes are bound together with a single
>>> +  mux controller, controlling all three muxes in parallel.
>>
>> Can't this be implied by #mux-control-cells == 0?
> 
> Right, good point! I'll drop adi,parallel.
> 
>>> +- adi,idle-state : if present, array of 2-tuples with mux controller number
>>> +  and state that mux controllers will have when idle. States 0 through 3
>>> +  correspond to signals A through D in the datasheet.
>>> +- adi,idle-high-impedance : if present, array of mux controller numbers that
>>> +  should be in the disconnected high-impedance state when idle.
>>
>> Perhaps combine these 2 to a common idle-state with the index being the 
>> mux controller number and a value of -1 (or 4) being hi-Z.
> 
> At one point [1] I had a straight array for the mux controllers with state 4
> being hi-Z and state 5 being the default "leave mux as-is on release", but
> Jonathan didn't like magic numbers. Can you please read the follow-ups to the
> referenced mail for some background, and then pick out the most suitable color
> for me to paint with? :-)
Hohum.  Given it can be cleanly represented as two properties rather than a single
property with a single magic value alongside the obvious ones I asked Peter
to split it.

Not all that fussed though as I don't have to maintain this ;)
*ducks*

Jonathan
> 
> [1] https://lkml.org/lkml/2016/11/30/110
> 
>>> +
>>> +Mux controller states 0 through 3 correspond to signals A through D in the
>>> +datasheet. If a mux controller is mentioned in neither adi,idle-state nor
>>> +adi,idle-high-impedance it is left in its previously selected state when idle.
>>> +
>>> +Example:
>>> +
>>> +	/*
>>> +	 * Three independent mux controllers (of which one is used).
>>> +	 * Mux 0 is disconnected when idle, mux 1 idles with signal C
>>> +	 * and mux 2 idles with signal A.
>>> +	 */
>>> +	&i2c0 {
>>> +		mux: adg792a@50 {
>>
>> mux-controller@50
> 
> Yep. Hmm, where have I seen that before? :-)
> 
> Cheers,
> peda
> 
>>> +			compatible = "adi,adg792a";
>>> +			reg = <0x50>;
>>> +			#mux-control-cells = <1>;
>>> +
>>> +			adi,idle-high-impedance = <0>;
>>> +			adi,idle-state = <1 2>, <2 0>;
>>> +		};
>>> +	};
>>> +
>>> +	adc-mux {
>>> +		compatible = "io-channel-mux";
>>> +		io-channels = <&adc 0>;
>>> +		io-channel-names = "parent";
>>> +
>>> +		mux-controls = <&mux 1>;
>>> +
>>> +		channels = "sync-1", "", "out";
>>> +	};
>>> +
>>> +
>>> +	/*
>>> +	 * Three parallel muxes with one mux controller, useful e.g. if
>>> +	 * the adc is differential, thus needing two signals to be muxed
>>> +	 * simultaneously for correct operation.
>>> +	 */
>>> +	&i2c0 {
>>> +		pmux: adg792a@50 {
>>> +			compatible = "adi,adg792a";
>>> +			reg = <0x50>;
>>> +			#mux-control-cells = <0>;
>>> +			adi,parallel;
>>> +		};
>>> +	};
>>> +
>>> +	diff-adc-mux {
>>> +		compatible = "io-channel-mux";
>>> +		io-channels = <&adc 0>;
>>> +		io-channel-names = "parent";
>>> +
>>> +		mux-controls = <&pmux>;
>>> +
>>> +		channels = "sync-1", "", "out";
>>> +	};
>>> -- 
>>> 2.1.4
>>>
>>> --
>>> To unsubscribe from this list: send the line "unsubscribe devicetree" in
>>> the body of a message to majordomo@vger.kernel.org
>>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>