Re: [PATCH v5 07/13] arm64: dts: qcom: x1e80100-crd: Add ov08x40 RGB sensor on CSIPHY4
From: Bryan O'Donoghue
Date: Fri Sep 04 2026 - 12:34:32 EST
On 04/09/2026 17:03, Nihal Kumar Gupta wrote:
On 04-09-2026 16:26, Bryan O'Donoghue wrote:
+ port@3 {Hi Bryan,
+ camss_csiphy4_inep: endpoint {
+ data-lanes = <0 1 2 3>;
+ remote-endpoint = <&csiphy4_out_ep>;
+ };
camss_csiphy4_inep has data-lanes = <0 1 2 3> and csiphy4_in_ep on the
same physical link has data-lanes = <1 2 3 4>.
Yes I'm aware of that and choose not to churn this.
This is what the controller has always done - it only cares about the number of lanes in this mode not their value - in the old method it cares about the values too.
...
Sigh, I can write a patch to count the number of lanes instead of what the code currently does.
That way the old method and a new method based on +1 will not collide.
One thing to note is that if the PHY mapped lanes <2 3> the camss side would not care if we represented that as <0 1> <1 2> or <2 3> since all it cares about after this series is the _number_ of lanes.
That said it is less confusing to have a uniform numbering scheme at both ends.
---
bod