[PATCH 6/8] dt-bindings: media: nxp,imx-mipi-csi2: Add fsl,num-channels property
Rob Herring
robh at kernel.org
Wed Jun 25 12:27:28 PDT 2025
On Fri, Jun 20, 2025 at 12:02:37AM +0300, Laurent Pinchart wrote:
> On Tue, Jun 10, 2025 at 11:18:29AM +0300, Laurent Pinchart wrote:
> > On Mon, Jun 09, 2025 at 03:08:39PM -0400, Frank Li wrote:
> > > On Mon, Jun 09, 2025 at 09:20:33PM +0300, Laurent Pinchart wrote:
> > > > On Mon, Jun 09, 2025 at 12:53:48PM -0500, Adam Ford wrote:
> > > > > On Mon, Jun 9, 2025 at 10:32 AM Frank Li wrote:
> > > > > > On Mon, Jun 09, 2025 at 02:58:38AM +0300, Laurent Pinchart wrote:
> > > > > > > The CSI-2 receiver can be instantiated with up to four output channels.
> > > > > > > This is an integration-specific property, specify the number of
> > > > > > > instantiated channels through a new fsl,num-channels property. The
> > > > > > > property is optional, and defaults to 1 as only one channel is currently
> > > > > > > supported by drivers.
> > > > > > >
> > > > > > > The only known SoC to have more than one channel is the i.MX8MP. As the
> > > > > > > binding examples do not cover that SoC, don't update them.
> > > > > > >
> > > > > > > Signed-off-by: Laurent Pinchart <laurent.pinchart at ideasonboard.com>
> > > > > > > ---
> > > > > > > .../devicetree/bindings/media/nxp,imx-mipi-csi2.yaml | 7 +++++++
> > > > > > > 1 file changed, 7 insertions(+)
> > > > > > >
> > > > > > > diff --git a/Documentation/devicetree/bindings/media/nxp,imx-mipi-csi2.yaml b/Documentation/devicetree/bindings/media/nxp,imx-mipi-csi2.yaml
> > > > > > > index db4889bf881e..41ad5b84eaeb 100644
> > > > > > > --- a/Documentation/devicetree/bindings/media/nxp,imx-mipi-csi2.yaml
> > > > > > > +++ b/Documentation/devicetree/bindings/media/nxp,imx-mipi-csi2.yaml
> > > > > > > @@ -68,6 +68,13 @@ properties:
> > > > > > > default: 166000000
> > > > > > > deprecated: true
> > > > > > >
> > > > > > > + fsl,num-channels:
> > > > > > > + $ref: /schemas/types.yaml#/definitions/uint32
> > > > > > > + description: Number of output channels
> > > > > > > + minimum: 1
> > > > > > > + maximum: 4
> > > > > > > + default: 1
> > > > > > > +
> > > > > >
> > > > > > Look like it is fixed value for each compabiable string, So it is not
> > > > > > suitable for adding new property. It should be in driver data of each
> > > > > > compatible strings.
> > > > > >
> > > > > > I met similar case before. DT team generally don't agree on add such
> > > > > > property, unless there are two instances in the same chip, which have
> > > > > > difference channel number.
> > > > >
> > > > > If the DT changes are rejected, can the number of channels be added to
> > > > > the data structure inside mipi_csis_of_match? We have compatibles for
> > > > > 8mm and imx7. If we add an imx8mp compatible we could add a reference
> > > > > to the number of channels.
> > > >
> > > > I thought about it, and decided to add a new property because the number
> > > > of channels is really a synthesis time configuration parameter, and
> > > > could differ between different CSIS instances in the same SoC.
> > >
> > > Need add such information at binding doc's commit message,
> >
> > I'll update the commit message.
>
> The commit message in v2 will state
>
> dt-bindings: media: nxp,imx-mipi-csi2: Add fsl,num-channels property
>
> The CSI-2 receiver can be instantiated with up to four output channels.
> This is an integration-specific property, specify the number of
> instantiated channels through a new fsl,num-channels property. The
> property is optional, and defaults to 1 as only one channel is currently
> supported by drivers.
>
> Using the compatible string to infer the number of channels has been
> considered, but multiple instances of the same CSIS in the same SoC
> could conceptually be synthesized with a different number of channels.
> An explicit property is therefore more appropriate.
>
> The only known SoC to have more than one channel is the i.MX8MP. As the
> binding examples do not cover that SoC, don't update them.
So how many channels does i.MX8MP have in a DT without this property?
It's either 1 or the driver overrides it to 2.
> Rob, Krzysztof, Conor, are you fine with adding this property ?
Yes, but seems a little late.
Rob
More information about the linux-arm-kernel
mailing list