From: Martin Blumenstingl <[email protected]>
The CPU port has to specify a phy-mode and either a phy or a fixed-link.
Since GSWIP is connected using a SoC internal protocol there's no PHY
involved. Add phy-mode = "internal" and a fixed-link to describe the
communication between the PMAC (Ethernet controller) and GSWIP switch.
Signed-off-by: Martin Blumenstingl <[email protected]>
---
Documentation/devicetree/bindings/net/dsa/lantiq-gswip.txt | 6 ++++++
1 file changed, 6 insertions(+)
diff --git a/Documentation/devicetree/bindings/net/dsa/lantiq-gswip.txt b/Documentation/devicetree/bindings/net/dsa/lantiq-gswip.txt
index 8bb1eff21cb1..e81ba0e0da0f 100644
--- a/Documentation/devicetree/bindings/net/dsa/lantiq-gswip.txt
+++ b/Documentation/devicetree/bindings/net/dsa/lantiq-gswip.txt
@@ -96,7 +96,13 @@ switch@e108000 {
port@6 {
reg = <0x6>;
+ phy-mode = "internal";
ethernet = <ð0>;
+
+ fixed-link {
+ speed = <1000>;
+ full-duplex;
+ };
};
};
--
2.39.2
On Thu, Jun 06, 2024 at 10:52:22AM +0200, Martin Schiller wrote:
> From: Martin Blumenstingl <[email protected]>
>
> The CPU port has to specify a phy-mode and either a phy or a fixed-link.
> Since GSWIP is connected using a SoC internal protocol there's no PHY
> involved. Add phy-mode = "internal" and a fixed-link to describe the
> communication between the PMAC (Ethernet controller) and GSWIP switch.
>
> Signed-off-by: Martin Blumenstingl <[email protected]>
> ---
Reviewed-by: Vladimir Oltean <[email protected]>
Long-term it would be good to also see a dt-schema conversion.
On 06/06/2024 10:52, Martin Schiller wrote:
> From: Martin Blumenstingl <[email protected]>
>
> The CPU port has to specify a phy-mode and either a phy or a fixed-link.
> Since GSWIP is connected using a SoC internal protocol there's no PHY
> involved. Add phy-mode = "internal" and a fixed-link to describe the
> communication between the PMAC (Ethernet controller) and GSWIP switch.
You did nothing in the binding to describe them. You only extended
example, which does not really matter if there is DTS with it.
Best regards,
Krzysztof
On 2024-06-10 10:55, Krzysztof Kozlowski wrote:
> On 06/06/2024 10:52, Martin Schiller wrote:
>> From: Martin Blumenstingl <[email protected]>
>>
>> The CPU port has to specify a phy-mode and either a phy or a
>> fixed-link.
>> Since GSWIP is connected using a SoC internal protocol there's no PHY
>> involved. Add phy-mode = "internal" and a fixed-link to describe the
>> communication between the PMAC (Ethernet controller) and GSWIP switch.
>
> You did nothing in the binding to describe them. You only extended
> example, which does not really matter if there is DTS with it.
>
> Best regards,
> Krzysztof
OK, so I'll update subject and commit message to signal that we only
update the example code.
On Mon, Jun 10, 2024 at 11:07:15AM +0200, Martin Schiller wrote:
> On 2024-06-10 10:55, Krzysztof Kozlowski wrote:
> > On 06/06/2024 10:52, Martin Schiller wrote:
> > > From: Martin Blumenstingl <[email protected]>
> > >
> > > The CPU port has to specify a phy-mode and either a phy or a
> > > fixed-link.
> > > Since GSWIP is connected using a SoC internal protocol there's no PHY
> > > involved. Add phy-mode = "internal" and a fixed-link to describe the
> > > communication between the PMAC (Ethernet controller) and GSWIP switch.
> >
> > You did nothing in the binding to describe them. You only extended
> > example, which does not really matter if there is DTS with it.
> >
> > Best regards,
> > Krzysztof
>
> OK, so I'll update subject and commit message to signal that we only
> update the example code.
Either convert it or leave it alone. If you are worried about users' DTs
being wrong due to copying a bad example, then you should care enough to
do the conversion. Given the errors we find in examples, it's likely
not the only problem.
Rob
On 2024-06-11 00:05, Rob Herring wrote:
> On Mon, Jun 10, 2024 at 11:07:15AM +0200, Martin Schiller wrote:
>> On 2024-06-10 10:55, Krzysztof Kozlowski wrote:
>> > On 06/06/2024 10:52, Martin Schiller wrote:
>> > > From: Martin Blumenstingl <[email protected]>
>> > >
>> > > The CPU port has to specify a phy-mode and either a phy or a
>> > > fixed-link.
>> > > Since GSWIP is connected using a SoC internal protocol there's no PHY
>> > > involved. Add phy-mode = "internal" and a fixed-link to describe the
>> > > communication between the PMAC (Ethernet controller) and GSWIP switch.
>> >
>> > You did nothing in the binding to describe them. You only extended
>> > example, which does not really matter if there is DTS with it.
>> >
>> > Best regards,
>> > Krzysztof
>>
>> OK, so I'll update subject and commit message to signal that we only
>> update the example code.
>
> Either convert it or leave it alone. If you are worried about users'
> DTs
> being wrong due to copying a bad example, then you should care enough
> to
> do the conversion. Given the errors we find in examples, it's likely
> not the only problem.
>
> Rob
Then I will convert the bindings to the new YAML format and send another
update of this patch-set.
Thanks,
Martin