Re: [PATCH v11 1/2] dt-bindings: media: Add Allwinner V3s Camera Sensor Interface (CSI)

[Chen-Yu Tsai]

Hi,

On Fri, Sep 28, 2018 at 8:56 PM Sakari Ailus
<sakari.ailus@xxxxxxxxxxxxxxx> wrote:
>
> Hi Chen-Yu,
>
> On Fri, Sep 28, 2018 at 07:10:58PM +0800, Chen-Yu Tsai wrote:
> > On Fri, Sep 28, 2018 at 6:23 PM Sakari Ailus
> > <sakari.ailus@xxxxxxxxxxxxxxx> wrote:
> > >
> > > Hi Laurent,
> > >
> > > On Fri, Sep 28, 2018 at 12:45:12PM +0300, Laurent Pinchart wrote:
> > > > Hi Sakari,
> > > >
> > > > On Friday, 28 September 2018 12:38:33 EEST Sakari Ailus wrote:
> > > > > On Wed, Sep 26, 2018 at 04:40:04PM +0800, Yong Deng wrote:
> > > > > > Add binding documentation for Allwinner V3s CSI.
> > > > > >
> > > > > > Acked-by: Maxime Ripard <maxime.ripard@xxxxxxxxxxx>
> > > > > > Acked-by: Sakari Ailus <sakari.ailus@xxxxxxxxxxxxxxx>
> > > > >
> > > > > I know... but I have a few more comments.
> > > > >
> > > > > > Reviewed-by: Rob Herring <robh@xxxxxxxxxx>
> > > > > > Signed-off-by: Yong Deng <yong.deng@xxxxxxxxxxxx>
> > > > > > ---
> > > > > >
> > > > > >  .../devicetree/bindings/media/sun6i-csi.txt        | 59 +++++++++++++++++
> > > > > >  1 file changed, 59 insertions(+)
> > > > > >  create mode 100644 Documentation/devicetree/bindings/media/sun6i-csi.txt
> > > > > >
> > > > > > diff --git a/Documentation/devicetree/bindings/media/sun6i-csi.txt
> > > > > > b/Documentation/devicetree/bindings/media/sun6i-csi.txt new file mode
> > > > > > 100644
> > > > > > index 000000000000..2ff47a9507a6
> > > > > > --- /dev/null
> > > > > > +++ b/Documentation/devicetree/bindings/media/sun6i-csi.txt
> > > > > > @@ -0,0 +1,59 @@
> > > > > > +Allwinner V3s Camera Sensor Interface
> > > > > > +-------------------------------------
> > > > > > +
> > > > > > +Allwinner V3s SoC features two CSI module. CSI0 is used for MIPI CSI-2
> > > > > > +interface and CSI1 is used for parallel interface.
> > > > > > +
> > > > > > +Required properties:
> > > > > > +  - compatible: value must be "allwinner,sun8i-v3s-csi"
> > > > > > +  - reg: base address and size of the memory-mapped region.
> > > > > > +  - interrupts: interrupt associated to this IP
> > > > > > +  - clocks: phandles to the clocks feeding the CSI
> > > > > > +    * bus: the CSI interface clock
> > > > > > +    * mod: the CSI module clock
> > > > > > +    * ram: the CSI DRAM clock
> > > > > > +  - clock-names: the clock names mentioned above
> > > > > > +  - resets: phandles to the reset line driving the CSI
> > > > > > +
> > > > > > +Each CSI node should contain one 'port' child node with one child
> > > > > > 'endpoint' +node, according to the bindings defined in
> > > > > > +Documentation/devicetree/bindings/media/video-interfaces.txt. As
> > > > > > mentioned
> > > > > > +above, the endpoint's bus type should be MIPI CSI-2 for CSI0 and parallel
> > > > > > or +Bt656 for CSI1.
> > > > >
> > > > > Which port represents CSI0 and which one is CSI1? That needs to be
> > > > > documented.
> > > >
> > > > There are two CSI devices, named CSI0 and CSI1, with one port each. The CSI0
> > > > device supports CSI-2 only, and the CSI1 device parallel (BT.601 or BT.656)
> > > > only.
> > > >
> > > > > > +
> > > > > > +Endpoint node properties for CSI1
> > > > >
> > > > > How about CSI0? I'd expect at least data-lanes, and clock-lanes as well if
> > > > > the hardware supports lane mapping.
> > > >
> > > > I enquired about that too. As far as I understand, CSI0 isn't supported yet in
> > > > the driver due to lack of documentation and lack of open-source vendor-
> > > > provided source code. While DT bindings are not tied to driver
> > > > implementations, it's not the best idea to design DT bindings without at least
> > > > one working implementation to test them. I thus proposed just listing CSI0 as
> > > > being unsupported for now.
> > >
> > > Ack.
> > >
> > > We should still define which receiver corresponds to a given port. Probably
> > > 1 for CSI1 would make sense, in order to avoid changing the order the
> > > hardware already uses. 0 doesn't need to be documented no IMO.
> > >
> > > What do you think?
> >
> > AFAICT it would be a completely seperate node, since they have different address
> > spaces, clocks and reset controls. So there's no possibility of confusion.
> >
> > According to Yong, CSI0 is tied internally to some unknown MIPI CSI2-receiver,
> > which is the undocumented part. CSI1 has its parallel data pins exposed to the
> > outside.
>
> Thanks for clearing up the confusion. If these are truly different kinds of
> devices, then don't they also deserve different compatible strings? And
> possibly also different DT binding documentation in a separate file.

Allwinner's documents aren't particularly clear about this. AFAICT it does not
say anything about them being different. The CSI section gives a register table
with two base addresses. The V3S SDK listed on linux-sunxi wiki also doesn't
differentiate between the two. What's more is there's actually code and register
addresses for the MIPI stuff [1], though the license is questionable,
as with most
BSPs. It indeed seems like a CSI controller, which basically just takes whatever
input it is configured for and sends it to either a downstream ISP or DRAM.
There are MIPI-CSI receiver and DPHY blocks in addresses following CSI0, but we
can treat them as separate hardware blocks.

ChenYu

[1] https://github.com/yeashen/v3s-linux-sdk/blob/master/linux-3.4/drivers/media/video/sunxi-vfe/platform/sun8iw8p1_vfe_cfg.h#L27