Re: [PATCH] Documentation: devicetree: Add PHY no lane swap binding

[Florian Fainelli <f.fainelli@gmail.com>]

Le 02/04/17 à 09:23, Andrew Lunn a écrit :
> On Sat, Feb 04, 2017 at 04:47:47PM +0100, Lukasz Majewski wrote:
>> Add the documentation to avoid PHY lane swapping. This is a boolean
>> entry to notify the phy device drivers that the TX/RX lanes NO need
> 
> that the TX/RX lanes should not be swapped.
> 
>> to be swapped.
>> The use case for this binding mostly happens after wrong HW
>> configuration of PHY IC during bootstrap.
>>
>> Signed-off-by: Lukasz Majewski <lukma@denx.de>
>> ---
>>  Documentation/devicetree/bindings/net/phy.txt | 4 ++++
>>  1 file changed, 4 insertions(+)
>>
>> diff --git a/Documentation/devicetree/bindings/net/phy.txt b/Documentation/devicetree/bindings/net/phy.txt
>> index fb5056b..5e25bc9 100644
>> --- a/Documentation/devicetree/bindings/net/phy.txt
>> +++ b/Documentation/devicetree/bindings/net/phy.txt
>> @@ -39,6 +39,10 @@ Optional Properties:
>>  - enet-phy-lane-swap: If set, indicates the PHY will swap the TX/RX lanes to
>>    compensate for the board being designed with the lanes swapped.
>>  
>> +- enet-phy-lane-no-swap: If set, indicates that PHY will disable swap of the
>> +  TX/RX lanes. This binding allows the PHY to work correcly after e.g. wrong
>> +  bootstrap configuration caused by issues in PCB layout design.

s/binding/property/

>> +
> 
> We are leaving it undefined what it means if neither
> enet-phy-lane-no-swap nor enet-phy-lane-swap properties are present.
> Do we want to define this? That the swap should be left untouched by
> the driver?

Since this is a description of the hardware, absence of a properties
should mean that the driver is at freedom to either keep the hardware
defaults, or come up with its own settings that are sensible for that
particular PHY device.

What would you see clarified here?
-- 
Florian