Re: [PATCH v1 1/2] dt-bindings: display: bridge: sn65dsi83: Add DSI video mode

[Marek Vasut <marex@denx.de>]

On 4/3/23 23:15, Francesco Dolcini wrote:
> On Mon, Apr 03, 2023 at 04:06:22PM -0500, Rob Herring wrote:
>> On Thu, Mar 30, 2023 at 12:17:51PM +0200, Francesco Dolcini wrote:
>>> From: Francesco Dolcini <francesco.dolcini@toradex.com>
>>>
>>> SN65DSI8[34] device supports burst video mode and non-burst video mode
>>> with sync events or with sync pulses packet transmission as described in
>>> the DSI specification.
>>>
>>> Add property to select the expected mode, this allows for example to
>>> select a mode that is compatible with the DSI host interface.
>>
>> Why does this need to be in DT?
> 
>> The source and sink drivers should know what their capabilities are
>> and pick the best common one.
> 
> Is there a best mode?

I think yes:
Burst (is better than) Sync Events (is better than) Sync Pulses

Burst is most energy efficient, Sync-Pulses is the simplest and least 
energy efficient and with most constraints.

> Isn't this a decision how do we want the 2 peers
> to communicate?

I don't think so, I believe the Host and nearest bridge should be able 
to negotiate their capabilities (mode, link rate, etc.) within the DRM 
subsystem.

> For the MIPI-DSI Linux/DRM experts: am I missing something? Is there
> another way to have a DSI video sink to ask for a specific mode?

I'm afraid this is not implemented yet, so ... plumbing needed.

[...]