[PATCH v6 1/2] dt-bindings: net: Add FSD EQoS device tree bindings

[Swathi K S]

> -----Original Message-----
> From: Krzysztof Kozlowski <krzk at kernel.org>
> Sent: 14 February 2025 13:02
> To: Swathi K S <swathi.ks at samsung.com>
> Cc: krzk+dt at kernel.org; andrew+netdev at lunn.ch; davem at davemloft.net;
> edumazet at google.com; kuba at kernel.org; pabeni at redhat.com;
> robh at kernel.org; conor+dt at kernel.org; richardcochran at gmail.com;
> mcoquelin.stm32 at gmail.com; alexandre.torgue at foss.st.com;
> rmk+kernel at armlinux.org.uk; netdev at vger.kernel.org;
> devicetree at vger.kernel.org; linux-stm32 at st-md-mailman.stormreply.com;
> linux-arm-kernel at lists.infradead.org; linux-kernel at vger.kernel.org
> Subject: Re: [PATCH v6 1/2] dt-bindings: net: Add FSD EQoS device tree
> bindings
> 
> On 14/02/2025 05:53, Swathi K S wrote:
> >
> >
> >> -----Original Message-----
> >> From: Krzysztof Kozlowski <krzk at kernel.org>
> >> Sent: 13 February 2025 17:31
> >> To: Swathi K S <swathi.ks at samsung.com>
> >> Cc: krzk+dt at kernel.org; andrew+netdev at lunn.ch;
> davem at davemloft.net;
> >> edumazet at google.com; kuba at kernel.org; pabeni at redhat.com;
> >> robh at kernel.org; conor+dt at kernel.org; richardcochran at gmail.com;
> >> mcoquelin.stm32 at gmail.com; alexandre.torgue at foss.st.com;
> >> rmk+kernel at armlinux.org.uk; netdev at vger.kernel.org;
> >> devicetree at vger.kernel.org; linux-stm32 at st-md-
> mailman.stormreply.com;
> >> linux-arm-kernel at lists.infradead.org; linux-kernel at vger.kernel.org
> >> Subject: Re: [PATCH v6 1/2] dt-bindings: net: Add FSD EQoS device
> >> tree bindings
> >>
> >> On 13/02/2025 12:04, Swathi K S wrote:
> >>>
> >>>
> >>>> -----Original Message-----
> >>>> From: Krzysztof Kozlowski <krzk at kernel.org>
> >>>> Sent: 13 February 2025 13:24
> >>>> To: Swathi K S <swathi.ks at samsung.com>
> >>>> Cc: krzk+dt at kernel.org; andrew+netdev at lunn.ch;
> >> davem at davemloft.net;
> >>>> edumazet at google.com; kuba at kernel.org; pabeni at redhat.com;
> >>>> robh at kernel.org; conor+dt at kernel.org; richardcochran at gmail.com;
> >>>> mcoquelin.stm32 at gmail.com; alexandre.torgue at foss.st.com;
> >>>> rmk+kernel at armlinux.org.uk; netdev at vger.kernel.org;
> >>>> devicetree at vger.kernel.org; linux-stm32 at st-md-
> >> mailman.stormreply.com;
> >>>> linux-arm-kernel at lists.infradead.org; linux-kernel at vger.kernel.org
> >>>> Subject: Re: [PATCH v6 1/2] dt-bindings: net: Add FSD EQoS device
> >>>> tree bindings
> >>>>
> >>>> On Thu, Feb 13, 2025 at 10:16:23AM +0530, Swathi K S wrote:
> >>>>> +  clock-names:
> >>>>> +    minItems: 5
> >>>>> +    maxItems: 10
> >>>>> +    contains:
> >>>>> +      enum:
> >>>>> +        - ptp_ref
> >>>>> +        - master_bus
> >>>>> +        - slave_bus
> >>>>> +        - tx
> >>>>> +        - rx
> >>>>> +        - master2_bus
> >>>>> +        - slave2_bus
> >>>>> +        - eqos_rxclk_mux
> >>>>> +        - eqos_phyrxclk
> >>>>> +        - dout_peric_rgmii_clk
> >>>>
> >>>> This does not match the previous entry. It should be strictly
> >>>> ordered with
> >>>> minItems: 5.
> >>>
> >>> Hi Krzysztof,
> >>> Thanks for reviewing.
> >>> In FSD SoC, we have 2 instances of ethernet in two blocks.
> >>> One instance needs 5 clocks and the other needs 10 clocks.
> >>
> >> I understand and I do not think this is contradictory to what I asked.
> >> If it is, then why/how?
> >>
> >>>
> >>> I tried to understand this by looking at some other dt-binding files
> >>> as given below, but looks like they follow similar approach
> >>> Documentation/devicetree/bindings/net/stm32-dwmac.yaml
> >>> Documentation/devicetree/bindings/net/rockchip-dwmac.yaml
> >>>
> >>> Could you please guide me on how to implement this?
> >>> Also, please help me understand what is meant by 'strictly ordered'
> >>
> >> Every other 99% of bindings. Just like your clocks property.
> >
> > Hi Krzysztof,
> > Thanks for your feedback.
> > I want to make sure I fully understand your comment.
> > I can see we have added clocks and clock names in the same order.
> 
> No, you did not. You can see syntax is very different - one uses items, other
> uses contains-enum. And now test it, change the order in DTS and see if you
> see any warning.
> 
> > Could you please help in detail what specifically needs to be modified
> regarding the ordering and minItems/maxItems usage?
> You did not try hard enough.
> 
> Open other bindings and look how they list clocks. For example any Samsung
> clock controller bindings. Or any qcom bindings.

Thanks for your insight. I tried understanding the dt-binding implementations from Samsung/ qcom as you suggested.
I am thinking of making the following modification in clock-names:

clock-names:
     minItems: 5
     items:
         - const: ptp_ref
         - const: master_bus
         - const: slave_bus
         - const: tx
         - const: rx
         - const: master2_bus
         - const: slave2_bus
         - const: eqos_rxclk_mux
         - const: eqos_phyrxclk
         - const: dout_peric_rgmii_clk
		 
Does this align with your feedback or do you suggest any further changes?

-Swathi

> 
> Best regards,
> Krzysztof