Re: [PATCH] dt-bindings: Another round of adding missing 'additionalProperties'

[Jonathan Cameron]

On Fri, 2 Oct 2020 18:41:43 -0500
Rob Herring <robh@xxxxxxxxxx> wrote:

> Another round of wack-a-mole. The json-schema default is additional
> unknown properties are allowed, but for DT all properties should be
> defined.
> 
> Cc: Thierry Reding <thierry.reding@xxxxxxxxx>
> Cc: Linus Walleij <linus.walleij@xxxxxxxxxx>
> Cc: Stephen Boyd <sboyd@xxxxxxxxxx>
> Cc: Shawn Guo <shawnguo@xxxxxxxxxx>
> Cc: Bjorn Andersson <bjorn.andersson@xxxxxxxxxx>
> Cc: Baolin Wang <baolin.wang7@xxxxxxxxx>
> Cc: Guenter Roeck <linux@xxxxxxxxxxxx>
> Cc: Jonathan Cameron <jic23@xxxxxxxxxx>
> Cc: Mauro Carvalho Chehab <mchehab@xxxxxxxxxx>
> Cc: Laurent Pinchart <laurent.pinchart+renesas@xxxxxxxxxxxxxxxx>
> Cc: Lee Jones <lee.jones@xxxxxxxxxx>
> Cc: Ulf Hansson <ulf.hansson@xxxxxxxxxx>
> Cc: "David S. Miller" <davem@xxxxxxxxxxxxx>
> Cc: Bjorn Helgaas <bhelgaas@xxxxxxxxxx>
> Cc: Vinod Koul <vkoul@xxxxxxxxxx>
> Cc: Liam Girdwood <lgirdwood@xxxxxxxxx>
> Cc: Mark Brown <broonie@xxxxxxxxxx>
> Cc: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
> Cc: Daniel Lezcano <daniel.lezcano@xxxxxxxxxx>
> Cc: linux-clk@xxxxxxxxxxxxxxx
> Cc: dri-devel@xxxxxxxxxxxxxxxxxxxxx
> Cc: linux-spi@xxxxxxxxxxxxxxx
> Cc: linux-gpio@xxxxxxxxxxxxxxx
> Cc: linux-hwmon@xxxxxxxxxxxxxxx
> Cc: linux-iio@xxxxxxxxxxxxxxx
> Cc: openipmi-developer@xxxxxxxxxxxxxxxxxxxxx
> Cc: linux-leds@xxxxxxxxxxxxxxx
> Cc: linux-media@xxxxxxxxxxxxxxx
> Cc: linux-rockchip@xxxxxxxxxxxxxxxxxxx
> Cc: linux-stm32@xxxxxxxxxxxxxxxxxxxxxxxxxxxx
> Cc: linux-mips@xxxxxxxxxxxxxxx
> Cc: linux-mmc@xxxxxxxxxxxxxxx
> Cc: netdev@xxxxxxxxxxxxxxx
> Cc: linux-pci@xxxxxxxxxxxxxxx
> Cc: linux-pm@xxxxxxxxxxxxxxx
> Cc: linux-remoteproc@xxxxxxxxxxxxxxx
> Cc: linux-serial@xxxxxxxxxxxxxxx
> Cc: alsa-devel@xxxxxxxxxxxxxxxx
> Cc: linux-usb@xxxxxxxxxxxxxxx
> Signed-off-by: Rob Herring <robh@xxxxxxxxxx>
> ---

Hi Rob,

Acked-by: Jonathan Cameron <Jonathan.Cameron@xxxxxxxxxx> # for iio


However, one of these made me wonder if the binding was simply wrong...
(definitely highlights why we should have additionalProperties: false
where ever possible).

...


> diff --git a/Documentation/devicetree/bindings/iio/imu/invensense,icm42600.yaml b/Documentation/devicetree/bindings/iio/imu/invensense,icm42600.yaml
> index abd8d25e1136..4c1c083d0e92 100644
> --- a/Documentation/devicetree/bindings/iio/imu/invensense,icm42600.yaml
> +++ b/Documentation/devicetree/bindings/iio/imu/invensense,icm42600.yaml
> @@ -47,11 +47,17 @@ properties:
>    vddio-supply:
>      description: Regulator that provides power to the bus
>  
> +  spi-max-frequency: true
> +  spi-cpha: true
> +  spi-cpol: true

It isn't completely unheard of for a device to operate in multiple SPI modes, but
it does seem to be fairly unusual.  I took a look at the datasheet and at least
from the provided timing diagrams, these are both required in SPI mode.

http://invensense.tdk.com/wp-content/uploads/2020/09/DS-000292-ICM-42605-v1.5.pdf

That doesn't make the binding wrong as such, but we could be tighter in checking this!

I'll add this to my list to take a closer look at sometime soonish.

Thanks.

Jonathan

> +
>  required:
>    - compatible
>    - reg
>    - interrupts
>  
> +additionalProperties: false
> +
>  examples:
>    - |