Message-ID: <17b8de50-426a-2543-a79a-aab44c9d52cf@cogentembedded.com>
Date:   Thu, 13 Jan 2022 00:10:35 +0300
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:91.0) Gecko/20100101
 Thunderbird/91.4.1
Subject: Re: [PATCH 2/3] arm64: dts: renesas: r8a77961: Add lvds0 device node
Content-Language: en-US
To:     Laurent Pinchart <laurent.pinchart@ideasonboard.com>,
        Geert Uytterhoeven <geert+renesas@glider.be>
Cc:     David Airlie <airlied@linux.ie>, Rob Herring <robh+dt@kernel.org>,
        linux-renesas-soc@vger.kernel.org, devicetree@vger.kernel.org,
        linux-kernel@vger.kernel.org
References: <20211224052309.1997096-1-nikita.yoush@cogentembedded.com>
 <20211224052309.1997096-3-nikita.yoush@cogentembedded.com>
 <YcyTV4fJqMHIeyYB@pendragon.ideasonboard.com>
 <87626d61-ada0-c220-bea5-5330f5256629@cogentembedded.com>
 <YcyXQxW3kRqQ2Yv0@pendragon.ideasonboard.com>
From:   Nikita Yushchenko <nikita.yoush@cogentembedded.com>
In-Reply-To: <YcyXQxW3kRqQ2Yv0@pendragon.ideasonboard.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Precedence: bulk

>>>> +				port@1 {
>>>> +					reg = <1>;
>>>> +					lvds0_out: endpoint {
>>>>    					};
>>>
>>> Endpoints must have a remote-endpoint property. Let's drop the endpoint
>>> here and keep the port only, the endpoint can be declared in board
>>> files.
>>>
>>> If you're fine with this change I can make it when applying the patch.
>>
>> This empty endpoint is currently defined in dtsi files for other r-car
>> gen3 SoCs.
>>
>> Goal here is to define lvds0_out label that is then used in extension
>> board dtsi files to link to the port.
>>
>> In this patch I just used the same approach as in files laying nearby.
>>
>> If this approach is not appropriate, then perhaps need to fix it in
>> files for all SoCs, to make it possible for extension board dtsi to be
>> compatible with all of them.
> 
> I'm writing a patch to drop those right now :-) I'll CC you.

This is not the only place where rcag-gen3 dtsi files are using empty-endpoint pattern.

du rgb port is defined in the same way.

And, I've submitted a patch some weeks ago [1] that hooked into that.

[1] https://lore.kernel.org/lkml/20211225115308.2152364-1-nikita.yoush@cogentembedded.com/

Since there was no reply, I am about to resubmit it.
But, perhaps need to do something with empty-endpoint pattern first?

Nikita