Re: [PATCH 10/13] dt-bindings: spi: add bindings for microchip mpfs spi

On 09/11/2021 14:20, Conor.Dooley@microchip.com wrote:

On 09/11/2021 13:04, Krzysztof Kozlowski wrote:

quoted

EXTERNAL EMAIL: Do not click links or open attachments unless you know the content is safe

On 09/11/2021 13:58, Conor.Dooley@microchip.com wrote:

quoted

On 09/11/2021 12:53, Krzysztof Kozlowski wrote:

quoted

EXTERNAL EMAIL: Do not click links or open attachments unless you know the content is safe

On 09/11/2021 13:16, Conor.Dooley@microchip.com wrote:

quoted

On 09/11/2021 04:06, Rob Herring wrote:

quoted

EXTERNAL EMAIL: Do not click links or open attachments unless you know the content is safe

On Mon, 08 Nov 2021 15:05:51 +0000, conor.dooley@microchip.com wrote:

quoted

From: Conor Dooley <conor.dooley@microchip.com>

Add device tree bindings for the {q,}spi controller on
the Microchip PolarFire SoC.

Signed-off-by: Conor Dooley <conor.dooley@microchip.com>
---
    .../bindings/spi/microchip,mpfs-spi.yaml      | 72 +++++++++++++++++++
    1 file changed, 72 insertions(+)
    create mode 100644 Documentation/devicetree/bindings/spi/microchip,mpfs-spi.yaml

My bot found errors running 'make DT_CHECKER_FLAGS=-m dt_binding_check'
on your patch (DT_CHECKER_FLAGS is new in v5.13):

yamllint warnings/errors:

dtschema/dtc warnings/errors:
Documentation/devicetree/bindings/spi/microchip,mpfs-spi.example.dts:19:18: fatal error: dt-bindings/clock/microchip,mpfs-clock.h: No such file or directory
      19 |         #include "dt-bindings/clock/microchip,mpfs-clock.h"

Rob,
Should I drop the header from the example or is there a way for me
specify the dependent patch to pass this check?

The error has to be fixed, although not necessarily by dropping the
header, but by posting it. How this can pass on your system? There is no
such file added in this patchset.

I linked the patch adding the clock as a dependency in the cover letter
[1], which is why I was wondering if there was a better way to do so
that would get picked up by the checker bot.

It's not only about the bot, but dependency when applied. If you did not
warn clk maintainer that clock bindings should go via Rob's tree or
should be provided as a tag, the patches here cannot be applied in this
cycle.

It was not my (our) intention to send the clock patches via rob's tree.
And since this is my first time trying to upstream wholescale changes to 
a device tree I honestly didn't expect this series to get accepted in 
this cycle anyway.

OK :)

Assuming your new bindings pass db_binding_check with
DT_CHECKER_FLAGS=-m (on top of clock patch), I propose to keep the
header here.

Another idea would be to submit without the header and use raw IDs
(numbers) and convert it later. I prefer the first- base on clock patches.


Best regards,
Krzysztof