Re: [PATCH v2 1/4] dt-bindings: net: ethernet-phy: move clocks property to invidivual PHY bindings

On 6/5/26 01:39, Rob Herring wrote:

On Tue, Jun 02, 2026 at 10:50:37AM +0400, George Moussalem wrote:

quoted

Move the clock property and restriction from the ethernet-phy.yaml file
to the individual PHY binding files. This allows each PHY to manage its
own clock requirements.

Signed-off-by: George Moussalem <redacted>
---
Commit 350b7a258f20 introduced the clocks property with a restriction to
maximum 1 to the main ethernet-phy.yaml binding for Realtek to add an
optional external clock source. This is restrictive to all PHY bindings,
as some PHYs may require more than 1 clock such as the IPQ5018 PHY which
requires 2 clocks (for RX and TX).

There are three other PHY drivers that require clock management:
- Micrel: requires 1 optional clock and the micrel.yaml file already
accomodates for the clock property.

It does? Where?

apologies, I see it's clock-names only:
https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git/tree/Documentation/devicetree/bindings/net/micrel.yaml#n103

quoted

- SMSC: requires an optional clock and the legacy bindings file
(smsc-lan87xx.txt) already accomodates for the clock property.
- BCM7xxx: requires an optional clock. I could not find a bindings file
for this PHY family.

Because it only uses what is defined in ethernet-phy.yaml. The only way 
to enforce 1 clock in this case is split ethernet-phy.yaml into common 
properties (removing the 'select') and a 'generic' phy schema that 
selects nodes with name 'ethernet-phy' and also have no 'compatible' 
property. Not sure if that's really worth doing. We generally try to 
require compatible for any new phys, but that's only if they try to add 
something to ethernet-phy.yaml so we see it.

The simple solution is you need to keep 'clocks' and make it 1-2 
clocks. It's been how many years until we needed 2 clocks? So probably 
some time before needing more if ever.

I'll update in next version, thanks.

Rob

thanks,
George