Re: [PATCH 3/5] drm/i2c: adv7511: Refactor encoder slave functions

[Archit Taneja <architt@codeaurora.org>]

Hi Laurent,

On 12/3/2015 9:41 PM, Archit Taneja wrote:
>
>
> On 12/3/2015 9:25 PM, Rob Clark wrote:
>> On Thu, Dec 3, 2015 at 10:28 AM, Laurent Pinchart
>> <laurent.pinchart@ideasonboard.com> wrote:
>>> On Thursday 03 December 2015 10:02:02 Rob Clark wrote:
>>>> On Mon, Jul 27, 2015 at 4:59 AM, Laurent Pinchart wrote:
>>>>> On Monday 27 July 2015 11:46:57 Archit Taneja wrote:
>>>>>> ADV7511 is represented as an i2c drm slave encoder device.
>>>>>> ADV7533, on
>>>>>> the other hand, is going be a normal i2c client device creating
>>>>>> bridge
>>>>>> and connector entities.
<snip>

>
> Laurent, as we'd discussed in the past, I did an initial study of how
> we could map bridge ops to the encoder slave ops, and change it for the
> rcar kms driver (and others) that might use adv7511.
>
> The drm_encoder_slave_funcs ops are a superset of bridge ops. The
> rcar-du driver uses a subset of these ops (encoder-like ops) to
> implement the hdmi encoder (rcar_du_hdmienc.c) and the other
> (connector-like ops)for implementing the hdmi connector
> (rcar_du_hdmicon.c)
>
> We have two options:
>
> 1. If we want to use drm_bridge as it is, the rcar driver shouldn't
> create a hdmi connector anymore, and rely on the adv7511 driver to make
> a connector for it. This is the easiest way, but it will break the nice
> representation of the hardware in DT.
>
> 2. We add more drm_bridge ops (the ones that implement the connector
> ops), and make the hdmi connector created by rcar-du use these bridge
> ops.
>
> Both these options have issues. The 2) one is the probably the better of
> the two, but I don't know if adding more ops to the bridge is the right
> way to go.
>
> Therefore, I don't know how we can solve this straight away. If there is
> more debate needed on this, then it is probably easier to get adv7533
> support in, and then figure out how two merge the two.
>

I gave approach (1) a try. I modified the rcar-du driver to use bridge
instead of i2c slave encoder and posted a RFC [1]. The hdmi connector
DT nodes (like in r8a7790-lager.dts) aren't really used by the rcar-du
driver. Now, with the adv7511 driver creating its own connector, they
mean even much less.

The adv7511 refactor from i2c slave encoder to bridge is done in [2].

[1] http://marc.info/?l=dri-devel&m=145235835902378&w=1
[2] http://marc.info/?l=dri-devel&m=145235823602353&w=1

Could you please review and test?

Thanks,
Archit

>
>>
>> BR,
>> -R
>>
>>> --
>>> Regards,
>>>
>>> Laurent Pinchart
>>>
>

-- 
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
a Linux Foundation Collaborative Project