[U-Boot] [RFC PATCH 1/2] net: phy: ti: Allow the driver to be more configurable

[Dan Murphy <dmurphy@ti.com>]

Tom

On 03/31/2016 10:42 AM, Tom Rini wrote:
> On Thu, Mar 31, 2016 at 10:27:31AM -0500, Dan Murphy wrote:
>> Tom
>>
>> On 03/31/2016 09:11 AM, Tom Rini wrote:
>>> On Thu, Mar 31, 2016 at 07:42:39AM -0500, Dan Murphy wrote:
>>>
>>>> Not all devices use the same internal delay or fifo depth.
>>>> Add the ability to set the internal delay for rx or tx and the
>>>> fifo depth via the config file.  If the value is not set in the
>>>> config file then set the delay to the default.
>>>>
>>>> Signed-off-by: Dan Murphy <dmurphy@ti.com>
>>>> ---
>>>>  drivers/net/phy/ti.c                 | 71 ++++++++++++++++++++++++++++++++----
>>>>  include/dt-bindings/net/ti-dp83867.h | 35 ++++++++++++++++++
>>> I don't think this is taking things down the right path.  If it's a DT
>>> binding, it comes from the device tree (which is fine and good!) but
>>> that means the binding needs to meet the usual reviews and not just come
>>> in via U-Boot like this.  We really don't want to add a DT binding that
>>> gets values from the config.h file.
>>>
>> This binding file I created that is already part of the Mainline kernel.
>> I just brought it in to use #defines and once the drivers are ported to use
>> DT then the bindings will already be available.
>>
>> And the config.h should be getting its values from the dt-binding.
> OK.  Then lets bring in the whole binding as its own patch.  And
> Mugunthan made cpsw do DM_ETH I see drivers/net/phy/micrel.c talks
> DM_ETH and DT so lets get this PHY driver updated and then get this
> additional bit from the DT.  Thanks!
>
Thanks for the DT pointer I will look at it.  What about platforms that do not use DT yet?
Well maybe the more appropriate question is are there any platforms that still use the config file?

Dan

-- 
------------------
Dan Murphy