Re: [RFC PATCH v1 1/4] media: dt-bindings: max9286: add device tree binding

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hi Geert,

On Tue, Jul 17, 2018 at 02:23:16PM +0200, Geert Uytterhoeven wrote:
> Hi Jacopo,
>
> On Tue, Jul 17, 2018 at 2:13 PM jacopo mondi <jacopo@xxxxxxxxxx> wrote:
> >    I'm replying here, even if a new version of the bindings for this
> > chip has been posted[1], as they have the same ports layout.
> >
> > [1] https://www.spinics.net/lists/linux-renesas-soc/msg29307.html
> >
> > On Wed, Jun 06, 2018 at 08:34:41AM +0200, Geert Uytterhoeven wrote:
> > > Hi Kieran,
> > >
> > > On Wed, Jun 6, 2018 at 1:34 AM, Kieran Bingham
> > > <kieran.bingham+renesas@xxxxxxxxxxxxxxxx> wrote:
> > > > Provide device tree binding documentation for the MAX9286 Quad GMSL
> > > > deserialiser.
> > > >
> > > > Signed-off-by: Kieran Bingham <kieran.bingham+renesas@xxxxxxxxxxxxxxxx>
> > >
> > > Thanks for your patch!
> > >
> > > > --- /dev/null
> > > > +++ b/Documentation/devicetree/bindings/media/i2c/max9286.txt
> > > > @@ -0,0 +1,75 @@
> > > > +* Maxim Integrated MAX9286 GMSL Quad 1.5Gbps GMSL Deserializer
> > > > +
> > > > +Required Properties:
> > > > + - compatible: Shall be "maxim,max9286"
> > > > +
> > > > +The following required properties are defined externally in
> > > > +Documentation/devicetree/bindings/i2c/i2c-mux.txt:
> > > > + - Standard I2C mux properties.
> > > > + - I2C child bus nodes.
> > > > +
> > > > +A maximum of 4 I2C child nodes can be specified on the MAX9286, to
> > > > +correspond with a maximum of 4 input devices.
> > > > +
> > > > +The device node must contain one 'port' child node per device input and output
> > > > +port, in accordance with the video interface bindings defined in
> > > > +Documentation/devicetree/bindings/media/video-interfaces.txt. The port nodes
> > > > +are numbered as follows.
> > > > +
> > > > +      Port        Type
> > > > +    ----------------------
> > > > +       0          sink
> > > > +       1          sink
> > > > +       2          sink
> > > > +       3          sink
> > > > +       4          source
> > >
> > > I assume the source and at least one sink are thus mandatory?
> > >
> > > Would it make sense to use port 0 for the source?
> > > This would simplify extending the binding to devices with more input
> > > ports later.
> >
> > I see your point, but as someone that has no idea how future chips could look
> > like, I wonder why having multiple outputs it's more un-likely to
> > happen than having more inputs added.
>
> I also don't know.
> I was just thinking "What if another chip has less or more sinks?".
>
> > Do you have any suggestion on how we can handle both cases?
>
> Instead of having a single "ports" subnode, you could split it in two subnodes,
> "sinks" and "sources"? I don't know if that's feasible.
>
Maybe I didn't get you here, and sorry to repeat something obvious for
you but, that would be against basic assumptions both in fwnode/of framework
and in media framework too. See the graph.txt documentation and
video_interfaces.txt, the layout with a single 'ports' node with
multiple 'port' sub-nodes is not avoidable, afaik.

What is not -theoretically- forbidden is to have port subnodes with
multiple endpoints, which is also used as an example in multiple
places in the documentation files I mentioned above. Unfortunately, as
we experienced with the ADV748x and the R-Car CSI-2 receiver
upstreaming effort, the v4l2-async framework relies on matching on the
remote endpoint's parent device node, and Kieran and Niklas had to use
a custom endpoint matching logic for the R-Car CSI-2 and adv748x
combo, something I'm sure nobody wants to repeat here.

We said that multiple times that we would like to tackle this
limitation (if that's limitation) in the v4l2_async framework, maybe
if GMSL-like devices with multiple input/output ports comes out, we
might have enough motivation to do that ;) ?

Thanks
   j


> Gr{oetje,eeting}s,
>
>                         Geert
>
> --
> Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@xxxxxxxxxxxxxx
>
> In personal conversations with technical people, I call myself a hacker. But
> when I'm talking to journalists I just say "programmer" or something like that.
>                                 -- Linus Torvalds

Attachment: signature.asc
Description: PGP signature


[Index of Archives]     [Linux Samsung SOC]     [Linux Wireless]     [Linux Kernel]     [ATH6KL]     [Linux Bluetooth]     [Linux Netdev]     [Kernel Newbies]     [IDE]     [Security]     [Git]     [Netfilter]     [Bugtraq]     [Yosemite News]     [MIPS Linux]     [ARM Linux]     [Linux Security]     [Linux RAID]     [Linux ATA RAID]     [Samba]     [Device Mapper]

  Powered by Linux