[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <47ccfe2d-5b96-2a76-90b7-6c3679dcec27@linaro.org>
Date: Sun, 24 Apr 2022 16:34:46 +0200
From: Krzysztof Kozlowski <krzysztof.kozlowski@...aro.org>
To: Ashish Mhetre <amhetre@...dia.com>, Rob Herring <robh@...nel.org>
Cc: dmitry.osipenko@...labora.com, digetx@...il.com,
thierry.reding@...il.com, jonathanh@...dia.com,
linux-kernel@...r.kernel.org, devicetree@...r.kernel.org,
linux-tegra@...r.kernel.org, krzysztof.kozlowski+dt@...aro.org,
vdumpa@...dia.com, Snikam@...dia.com
Subject: Re: [Patch v7 3/4] dt-bindings: memory: Update reg/reg-names
validation
On 24/04/2022 07:20, Ashish Mhetre wrote:
>>>
>>> New, added properties cannot be required. That's an ABI break.
>>>
>> This is handled in driver code to make sure driver works with old dts
>> as well. So is this bindings change fine or shall I change it such that
>> dt bindings check shall pass with older dts as well?
>> Or as mentioned by Dmitry, I can update the commit message to reflect
>> that ABI change is intended and driver is compatible with older DTBs as
>> well.
>>
> Hi Rob,
> Can you please confirm how shall I go in next version?
> Is it fine for dt bindings check to fail if driver is compatible with
> old as well as new dts? Or dt bindings check shall pass with old as
> well as new dts?
The driver works fine without reg-names and accepts old DTB, right? In
such case, just mention this in commit msg, that the bindings require
reg-names but backwards compatibility will be preserved in the driver. I
think it's fine to alter bindings such way.
Best regards,
Krzysztof
Powered by blists - more mailing lists