Thread (2 messages) 2 messages, 2 authors, 2021-07-27

Re: [RFC v3] dt-binding: media: document ON Semi AR0521 sensor bindings

From: Sakari Ailus <sakari.ailus@iki.fi>
Date: 2021-07-27 10:59:21
Also in: linux-devicetree, lkml

Hi Krzysztof,

On Tue, Jul 27, 2021 at 12:36:20PM +0200, Krzysztof Hałasa wrote:
Hi Laurent,

Laurent Pinchart [off-list ref] writes:
quoted
quoted
+        properties:
+          data-lanes:
+            anyOf:
+              - items:
+                  - const: 1
+              - items:
+                  - const: 1
+                  - const: 2
+              - items:
+                  - const: 1
+                  - const: 2
+                  - const: 3
+                  - const: 4
As the sensor also supports an HiSPi output, I would add the bus-type
property:

          data-lanes:
	    const: 4
Is there any example of this? I'm not sure how should it it look like.
Something like the following?

        properties:
         data-lanes:
            anyOf:
              - items:
                  - const: 1
              - items:
                  - const: 1
                  - const: 2
              - items:
                  - const: 1
                  - const: 2
                  - const: 3
                  - const: 4
          bus-type:
            data-lanes:
              const: 4
I think Laurent meant:

	    bus-type:
	      const: 4

This way the bindings can be later amended with HiSPi support without
relying on defaults. Albeit the other busses in practice almost never end
up being used even if supported, apart from the standard BT.601, BT.656 and
CSI-2.

Either way is fine IMO.
And... HiSPi would need additional code in the driver. And preferably
some testing. I think I'd prefer to have DT and the driver staying in
some sort of sync. Also, I'm uncertain about the syntax and the meaning
of such, apparently redundant, construct. Nor about its relation to
HiSPi. An example would be welcome.
No need to add support for the driver.

-- 
Sakari Ailus
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help