[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <87czy8p9cg.wl-kuninori.morimoto.gx@renesas.com>
Date: 14 Jan 2021 11:06:17 +0900
From: Kuninori Morimoto <kuninori.morimoto.gx@...esas.com>
To: Rob Herring <robh+dt@...nel.org>
Cc: Frank Rowand <frowand.list@...il.com>, devicetree@...r.kernel.org,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] of: property: add port base loop
Hi Rob
Thank you for your feedback
> > In such case, people want to have port base loop
> > instead of endpoints base loop.
> > This patch adds such functions/macros.
>
> I'm a bit hesitant on these too. A driver should generally know what
> each port # is (since the binding has to define them), and it should
> just request the port (or its connection) it wants. At least that was
> the premise behind of_graph_get_remote_node() and the cleanups (mostly
> DRM drivers) I did to use it. I'd rather see things move in that
> direction.
I'm working for ALSA SoC now.
To use ALSA SoC sound, it needs at least 3 drivers.
The purpose of "Card" here is for glueing SoC and Codec.
SoC <-> Card <-> Codec
If board/platform needs special handling,
user needs to create specific Card driver.
In such case, it needs to know detail of each ports.
But many cases, we want to use generic Card driver.
Because it is generic, it don't know how many ports are used,
and endpoint base loop is not enough for it.
This is my motivation for this patch.
> In any case, this needs a user before merging.
OK.
I will repost it again as part of above patch-set.
Thank you for your help !!
Best regards
---
Kuninori Morimoto
Powered by blists - more mailing lists