Re: [PATCH 1/5] dt-bindings: connector: add optional properties for Type-B

[Hans de Goede <hdegoede@redhat.com>]

Hi,

On 11-03-19 09:06, Hans de Goede wrote:
> Hi,
> 
> On 11-03-19 06:33, Chunfeng Yun wrote:
>> Hi,
>>
>> On Fri, 2019-03-08 at 13:07 +0100, Hans de Goede wrote:
>>> Hi,
>>>
>>> On 08-03-19 07:13, Chunfeng Yun wrote:
>>>> Add id-gpios, vbus-gpios, vbus-supply and pinctrl properties for
>>>> usb-b-connector
>>>>
>>>> Signed-off-by: Chunfeng Yun <chunfeng.yun@mediatek.com>
>>>> ---
>>>>    .../devicetree/bindings/connector/usb-connector.txt    | 10 ++++++++++
>>>>    1 file changed, 10 insertions(+)
>>>>
>>>> diff --git a/Documentation/devicetree/bindings/connector/usb-connector.txt b/Documentation/devicetree/bindings/connector/usb-connector.txt
>>>> index a9a2f2fc44f2..7a07b0f4f973 100644
>>>> --- a/Documentation/devicetree/bindings/connector/usb-connector.txt
>>>> +++ b/Documentation/devicetree/bindings/connector/usb-connector.txt
>>>> @@ -17,6 +17,16 @@ Optional properties:
>>>>    - self-powered: Set this property if the usb device that has its own power
>>>>      source.
>>>> +Optional properties for usb-b-connector:
>>>> +- id-gpios: gpio for USB ID pin.
>>>
>>> What about boards where the ID pin is *not* connected to a GPIO,
>>> but e.g. to a special pin on the PMIC which can also detect
>>> an ACA adapter ? Currently this case is handled by extcon
>>> drivers, but we have no way to set e.g. vbus-supply for the
>>> connector. Maybe in this case the usb-connector node should
>>> be a child of the PMIC node ?
>> Yes, it would be, PMIC is in charger of detecting the status of ID pin
> 
> Ok, then I think this should be documented too.
> 
>>> And in many cases there also is a mux to switch the datalines
>>> between the host and device(gadget) controllers, how should
>>> that be described in this model?  See the new usb-role-switch
>>> code under drivers/usb/roles
>>>
>>> In some cases the mux is controlled through a gpio, so we
>>> may want to add a "mux-gpios" here in which case we also
>>> need to define what 0/1 means.
>> I'm not sure, the mux seems not belong to this connector,
>> and may need another driver to register usb-role-switch,
>> similar to:
>>
>> [v2,2/2] usb: typec: add typec switch via GPIO control
>> https://patchwork.kernel.org/patch/10834327/
> 
> Right the mux/role-switch will need a driver, but the "owner"
> of the usb_connector, e.g. the PMIC or the owner of the
> id GPIO pin needs to know which device is the role-switch so
> that it can set the role correctly based on the id-pin.
> 
> Your binding already contains Vbus info, allowing the owner
> of the usb_connector to enable/disable Vbus based on the id-pin,
> but the owner will also be responsible for setting the role-switch.
> 
> Note we cannot simply assume there will be only one role-switch,
> we really need some link from the usb_connector to the role-switch
> (or if it is a GPIO driven role-switch simply a role-switch-gpios
> member in the usb_connector).

I see now in your "[PATCH v3 1/2] dt-bindings: usb: add documentation for
typec switch via GPIO" that you plan to use a
Documentation/devicetree/bindings/graph.txt style binding for this,
adding a remote-endpoint to the orientation-switch (in that patch), or
to the role-switch.

But a Type-C port can have both an orientation-switch and a role-switch
(and a mux if it supports e.g. DP-altmode), how is the driver driving
the device which has the actual usb_c_connecter child-node to which
the remote-endpoints for both the orientation- and the role-switch points
supposed to figure out which is which ?

Also it feels the wrong-way around to me to have the orientation-switch
point to the usb_c_connector, to me it would make more sense to have
the usb_c_connector point to a port on the orientation-switch.

Regards,

Hans