On 30.04.2023 15:44, Daniel Golle wrote:
On Sun, Apr 30, 2023 at 03:34:43PM +0300, Arınç ÜNAL wrote:
On 30.04.2023 14:28, David Bauer wrote:
Document the ability to add nodes for the MDIO bus connecting the
switch-internal PHYs.
This is quite interesting. Currently the PHY muxing feature for the MT7530
switch looks for some fake ethernet-phy definitions on the mdio-bus where
the switch is also defined.
Looking at the binding here, there will be an mdio node under the switch
node. This could be useful to define the ethernet-phys for PHY muxing here
instead, so we don't waste the register addresses on the parent mdio-bus for
fake things. It looks like this should work right out of the box. I will do
some tests.
Are there any examples as to what to configure on the switch PHYs with this
change?
Signed-off-by: David Bauer <mail@xxxxxxxxxxxxxxx>
---
.../devicetree/bindings/net/dsa/mediatek,mt7530.yaml | 6 ++++++
1 file changed, 6 insertions(+)
diff --git a/Documentation/devicetree/bindings/net/dsa/mediatek,mt7530.yaml b/Documentation/devicetree/bindings/net/dsa/mediatek,mt7530.yaml
index e532c6b795f4..50f8f83cc440 100644
--- a/Documentation/devicetree/bindings/net/dsa/mediatek,mt7530.yaml
+++ b/Documentation/devicetree/bindings/net/dsa/mediatek,mt7530.yaml
@@ -128,6 +128,12 @@ properties:
See Documentation/devicetree/bindings/regulator/mt6323-regulator.txt for
details for the regulator setup on these boards.
+ mdio:
+ $ref: /schemas/net/mdio.yaml#
+ unevaluatedProperties: false
+ description:
+ Node for the internal MDIO bus connected to the embedded ethernet-PHYs.
Please set this property as false for mediatek,mt7988-switch as it doesn't
use MDIO.
Well, quite the opposite is true. This change is **needed** on MT7988 as
the built-in 1GE PHYs of the MT7988 are connected to the (internal) MDIO
bus of the switch. And they do need calibration data assigned as nvmem
via device tree.
tl;dr: Despite not being connected via MDIO itself also MT7988 exposes an
internal MDIO bus for the switch PHYs.
Got it, thanks.
Arınç
