mbox series

[net-next,v2,00/10] arm64: dts: layerscape: update MAC nodes with PHY information

Message ID 20201002210737.27645-1-ioana.ciornei@nxp.com
Headers show
Series arm64: dts: layerscape: update MAC nodes with PHY information | expand

Message

Ioana Ciornei Oct. 2, 2020, 9:07 p.m. UTC
This patch set aims to add the necessary DTS nodes to complete the
MAC/PCS/PHY representation on DPAA2 devices. The external MDIO bus nodes
and the PHYs found on them are added, along with the PCS MDIO internal
buses and their PCS PHYs. Also, links to these PHYs are added from the
DPMAC node.

I am resending these via netdev because I am not really sure if Shawn is
still able to take them in time for 5.10 since his last activity on the
tree has been some time ago.
I tested them on linux-next and there are no conflicts.

Changes in v2:
 - documented the dpmac node into a new yaml entry
 - dropped the '0x' from some unit addresses

Ioana Ciornei (10):
  dt-bindings: net: add the dpaa2-mac DTS definition
  arm64: dts: ls1088a: add external MDIO device nodes
  arm64: dts: ls1088ardb: add QSGMII PHY nodes
  arm64: dts: ls1088ardb: add necessary DTS nodes for DPMAC2
  arm64: dts: ls208xa: add the external MDIO nodes
  arm64: dts: ls2088ardb: add PHY nodes for the CS4340 PHYs
  arm64: dts: ls2088ardb: add PHY nodes for the AQR405 PHYs
  arm64: dts: ls208xa: add PCS MDIO and PCS PHY nodes
  arm64: dts: lx2160a: add PCS MDIO and PCS PHY nodes
  arm64: dts: lx2160ardb: add nodes for the AQR107 PHYs

 .../devicetree/bindings/net/dpaa2-mac.yaml    |  55 ++++
 .../boot/dts/freescale/fsl-ls1088a-rdb.dts    | 119 +++++++++
 .../arm64/boot/dts/freescale/fsl-ls1088a.dtsi |  81 ++++++
 .../boot/dts/freescale/fsl-ls2088a-rdb.dts    | 118 ++++++++
 .../arm64/boot/dts/freescale/fsl-ls208xa.dtsi | 242 +++++++++++++++++
 .../boot/dts/freescale/fsl-lx2160a-rdb.dts    |  32 +++
 .../arm64/boot/dts/freescale/fsl-lx2160a.dtsi | 252 ++++++++++++++++++
 7 files changed, 899 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/net/dpaa2-mac.yaml

Comments

David Miller Oct. 2, 2020, 11:13 p.m. UTC | #1
From: Ioana Ciornei <ioana.ciornei@nxp.com>
Date: Sat,  3 Oct 2020 00:07:27 +0300

> This patch set aims to add the necessary DTS nodes to complete the
> MAC/PCS/PHY representation on DPAA2 devices. The external MDIO bus nodes
> and the PHYs found on them are added, along with the PCS MDIO internal
> buses and their PCS PHYs. Also, links to these PHYs are added from the
> DPMAC node.
> 
> I am resending these via netdev because I am not really sure if Shawn is
> still able to take them in time for 5.10 since his last activity on the
> tree has been some time ago.
> I tested them on linux-next and there are no conflicts.
> 
> Changes in v2:
>  - documented the dpmac node into a new yaml entry
>  - dropped the '0x' from some unit addresses

I don't feel comfortable taking such a sizable set of DT changes into
the networking tree rather than the devicetree or ARM tree(s).

I know we're fast and more responsive than the other subsystems (by
several orders of magnitude) but that isn't a reason to bypass the
correct tree for these changes.

Thank you.
Ioana Ciornei Oct. 3, 2020, 4:35 a.m. UTC | #2
On Fri, Oct 02, 2020 at 04:13:18PM -0700, David Miller wrote:
> From: Ioana Ciornei <ioana.ciornei@nxp.com>
> Date: Sat,  3 Oct 2020 00:07:27 +0300
> 
> > This patch set aims to add the necessary DTS nodes to complete the
> > MAC/PCS/PHY representation on DPAA2 devices. The external MDIO bus nodes
> > and the PHYs found on them are added, along with the PCS MDIO internal
> > buses and their PCS PHYs. Also, links to these PHYs are added from the
> > DPMAC node.
> > 
> > I am resending these via netdev because I am not really sure if Shawn is
> > still able to take them in time for 5.10 since his last activity on the
> > tree has been some time ago.
> > I tested them on linux-next and there are no conflicts.
> > 
> > Changes in v2:
> >  - documented the dpmac node into a new yaml entry
> >  - dropped the '0x' from some unit addresses
> 
> I don't feel comfortable taking such a sizable set of DT changes into
> the networking tree rather than the devicetree or ARM tree(s).
> 
> I know we're fast and more responsive than the other subsystems (by
> several orders of magnitude) but that isn't a reason to bypass the
> correct tree for these changes.
> 
> Thank you.

No problem. At least I cleaned-up the patch set and now it's more or
less ready to go for next time.
Thanks for the feedback!

Ioana
Rob Herring Oct. 6, 2020, 8:56 p.m. UTC | #3
On Sat, Oct 03, 2020 at 12:07:28AM +0300, Ioana Ciornei wrote:
> Add a documentation entry for the DTS bindings needed and supported by
> the dpaa2-mac driver.
> 
> Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
> ---
> Changes in v2:
>  - new patch
> 
>  .../devicetree/bindings/net/dpaa2-mac.yaml    | 55 +++++++++++++++++++

Use the compatible string for the filename.

>  1 file changed, 55 insertions(+)
>  create mode 100644 Documentation/devicetree/bindings/net/dpaa2-mac.yaml
> 
> diff --git a/Documentation/devicetree/bindings/net/dpaa2-mac.yaml b/Documentation/devicetree/bindings/net/dpaa2-mac.yaml
> new file mode 100644
> index 000000000000..744b0590278d
> --- /dev/null
> +++ b/Documentation/devicetree/bindings/net/dpaa2-mac.yaml
> @@ -0,0 +1,55 @@
> +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
> +%YAML 1.2
> +---
> +$id: http://devicetree.org/schemas/net/dpaa2-mac.yaml#
> +$schema: http://devicetree.org/meta-schemas/core.yaml#
> +
> +title: DPAA2 MAC bindings
> +
> +maintainers:
> +  - Ioana Ciornei <ioana.ciornei@nxp.com>
> +
> +description:
> +  This binding represents the DPAA2 MAC objects found on the fsl-mc bus and
> +  located under the 'dpmacs' node for the fsl-mc bus DTS node.

Need $ref to ethernet-controller.yaml

> +
> +properties:
> +  compatible:
> +    const: "fsl,qoriq-mc-dpmac"

Don't need quotes.

> +
> +  reg:
> +    maxItems: 1
> +    description: The DPMAC number
> +
> +  phy-handle: true
> +
> +  phy-connection-type: true
> +
> +  phy-mode: true
> +
> +  pcs-handle:
> +    $ref: /schemas/types.yaml#definitions/phandle
> +    description:
> +      A reference to a node representing a PCS PHY device found on
> +      the internal MDIO bus.

Perhaps use the 'phys' binding? (Too many PHYs with ethernet...)

This would be the on-chip XAUI/SerDes phy? That's typically 'phys' where 
as 'phy-handle' is ethernet PHY. 

> +
> +  managed: true
> +
> +required:
> +  - reg

addtionalProperties: false

> +
> +examples:
> +  - |
> +    dpmacs {
> +      #address-cells = <1>;
> +      #size-cells = <0>;
> +
> +      dpmac@4 {

ethernet@4

> +        compatible = "fsl,qoriq-mc-dpmac";
> +        reg = <0x4>;
> +        phy-handle = <&mdio1_phy6>;
> +        phy-connection-type = "qsgmii";
> +        managed = "in-band-status";
> +        pcs-handle = <&pcs3_1>;
> +      };
> +    };
> -- 
> 2.28.0
>
Ioana Ciornei Oct. 7, 2020, 10:51 a.m. UTC | #4
On Tue, Oct 06, 2020 at 03:56:35PM -0500, Rob Herring wrote:
> On Sat, Oct 03, 2020 at 12:07:28AM +0300, Ioana Ciornei wrote:
> > Add a documentation entry for the DTS bindings needed and supported by
> > the dpaa2-mac driver.
> > 
> > Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
> > ---
> > Changes in v2:
> >  - new patch
> > 
> >  .../devicetree/bindings/net/dpaa2-mac.yaml    | 55 +++++++++++++++++++
> 
> Use the compatible string for the filename.

Sure.

> > +title: DPAA2 MAC bindings
> > +
> > +maintainers:
> > +  - Ioana Ciornei <ioana.ciornei@nxp.com>
> > +
> > +description:
> > +  This binding represents the DPAA2 MAC objects found on the fsl-mc bus and
> > +  located under the 'dpmacs' node for the fsl-mc bus DTS node.
> 
> Need $ref to ethernet-controller.yaml
> 
> > +
> > +properties:
> > +  compatible:
> > +    const: "fsl,qoriq-mc-dpmac"
> 
> Don't need quotes.

Got it.

> 
> > +
> > +  reg:
> > +    maxItems: 1
> > +    description: The DPMAC number
> > +
> > +  phy-handle: true
> > +
> > +  phy-connection-type: true
> > +
> > +  phy-mode: true
> > +
> > +  pcs-handle:
> > +    $ref: /schemas/types.yaml#definitions/phandle
> > +    description:
> > +      A reference to a node representing a PCS PHY device found on
> > +      the internal MDIO bus.
> 
> Perhaps use the 'phys' binding? (Too many PHYs with ethernet...)
> 
> This would be the on-chip XAUI/SerDes phy? That's typically 'phys' where 
> as 'phy-handle' is ethernet PHY. 
> 

The PCS deals with proper coding (8b/10b, 64b/66b) and auto-negotiation
between the MAC and whatever is connected to it. It is, logically
speaking, above the SERDES which implements the physical layer
(PMA/PMD). We are not describing or configuring electrical parameters of
SERDES lanes and such here, we are just referencing the PCS found on the
internal MDIO bus of the MAC.

> > +
> > +  managed: true
> > +
> > +required:
> > +  - reg
> 
> addtionalProperties: false

Ok, I'll add it.
I didn't find a reference to it in the writing-schema.rst. I think that
would have helped.

> 
> > +
> > +examples:
> > +  - |
> > +    dpmacs {
> > +      #address-cells = <1>;
> > +      #size-cells = <0>;
> > +
> > +      dpmac@4 {
> 
> ethernet@4

Ok.

> 
> > +        compatible = "fsl,qoriq-mc-dpmac";
> > +        reg = <0x4>;
> > +        phy-handle = <&mdio1_phy6>;
> > +        phy-connection-type = "qsgmii";
> > +        managed = "in-band-status";
> > +        pcs-handle = <&pcs3_1>;
> > +      };
> > +    };
> > -- 
> > 2.28.0
> >