Re: [PATCH v2 1/3] dt-bindings: net: Add T-HEAD dwmac support

[Drew Fustini <dfustini@tenstorrent.com>]

On Fri, Sep 27, 2024 at 11:34:48AM +0200, Krzysztof Kozlowski wrote:
> On Thu, Sep 26, 2024 at 11:15:50AM -0700, Drew Fustini wrote:
> > From: Jisheng Zhang <jszhang@kernel.org>
> > 
> > Add documentation to describe T-HEAD dwmac.
> > 
> > Signed-off-by: Jisheng Zhang <jszhang@kernel.org>
> > Signed-off-by: Emil Renner Berthing <emil.renner.berthing@canonical.com>
> > [drew: change apb registers from syscon to second reg of gmac node]
> > [drew: rename compatible, add thead rx/tx internal delay properties]
> > Signed-off-by: Drew Fustini <dfustini@tenstorrent.com>
> > ---
> >  .../devicetree/bindings/net/snps,dwmac.yaml        |   1 +
> >  .../devicetree/bindings/net/thead,th1520-gmac.yaml | 109 +++++++++++++++++++++
> >  MAINTAINERS                                        |   1 +
> >  3 files changed, 111 insertions(+)
> > 
> > diff --git a/Documentation/devicetree/bindings/net/snps,dwmac.yaml b/Documentation/devicetree/bindings/net/snps,dwmac.yaml
> > index 4e2ba1bf788c..474ade185033 100644
> > --- a/Documentation/devicetree/bindings/net/snps,dwmac.yaml
> > +++ b/Documentation/devicetree/bindings/net/snps,dwmac.yaml
> > @@ -99,6 +99,7 @@ properties:
> >          - snps,dwxgmac-2.10
> >          - starfive,jh7100-dwmac
> >          - starfive,jh7110-dwmac
> > +        - thead,th1520-gmac
> >  
> >    reg:
> >      minItems: 1
> > diff --git a/Documentation/devicetree/bindings/net/thead,th1520-gmac.yaml b/Documentation/devicetree/bindings/net/thead,th1520-gmac.yaml
> > new file mode 100644
> > index 000000000000..1070e891c025
> > --- /dev/null
> > +++ b/Documentation/devicetree/bindings/net/thead,th1520-gmac.yaml
> > @@ -0,0 +1,109 @@
> > +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
> > +%YAML 1.2
> > +---
> > +$id: http://devicetree.org/schemas/net/thead,th1520-gmac.yaml#
> > +$schema: http://devicetree.org/meta-schemas/core.yaml#
> > +
> > +title: T-HEAD TH1520 GMAC Ethernet controller
> > +
> > +maintainers:
> > +  - Drew Fustini <dfustini@tenstorrent.com>
> > +
> > +description: |
> > +  The TH1520 GMAC is described in the TH1520 Peripheral Interface User Manual
> > +  https://git.beagleboard.org/beaglev-ahead/beaglev-ahead/-/tree/main/docs
> > +
> > +  Features include
> > +    - Compliant with IEEE802.3 Specification
> > +    - IEEE 1588-2008 standard for precision networked clock synchronization
> > +    - Supports 10/100/1000Mbps data transfer rate
> > +    - Supports RGMII/MII interface
> > +    - Preamble and start of frame data (SFD) insertion in Transmit path
> > +    - Preamble and SFD deletion in the Receive path
> > +    - Automatic CRC and pad generation options for receive frames
> > +    - MDIO master interface for PHY device configuration and management
> > +
> > +  The GMAC Registers consists of two parts
> > +    - APB registers are used to configure clock frequency/clock enable/clock
> > +      direction/PHY interface type.
> > +    - AHB registers are use to configure GMAC core (DesignWare Core part).
> > +      GMAC core register consists of DMA registers and GMAC registers.
> > +
> > +select:
> > +  properties:
> > +    compatible:
> > +      contains:
> > +        enum:
> > +          - thead,th1520-gmac
> > +  required:
> > +    - compatible
> > +
> > +allOf:
> > +  - $ref: snps,dwmac.yaml#
> > +
> > +properties:
> > +  compatible:
> > +    items:
> > +      - enum:
> > +          - thead,th1520-gmac
> > +      - const: snps,dwmac-3.70a
> > +
> > +  reg:
> > +    items:
> > +      - description: DesignWare GMAC IP core registers
> > +      - description: GMAC APB registers
> > +
> > +  reg-names:
> > +    items:
> > +      - const: dwmac
> > +      - const: apb
> > +
> > +  thead,rx-internal-delay:
> > +    $ref: /schemas/types.yaml#/definitions/uint32
> > +    description: |
> > +      RGMII receive clock delay. The value is used for the delay_ctrl
> > +      field in GMAC_RXCLK_DELAY_CTRL. Units are not specified.
> 
> What do you mean by "unspecified units"? They are always specified,
> hardware does not work randomly, e.g. once uses clock cycles, but next
> time you run it will use picoseconds.
> 
> You also miss default (property is not required) and some sort of constraints.

I should have stated that I don't know the units for delay_ctrl. The
5-bit field has a max value of 31 which seems far too small for
picoseconds. Unfortunately, the documentation from the SoC vendor does
not give anymore details about what the value represents.

Andrew Lunn replied [1] to my cover letter that it is best to hard code
the field to 0 (which is the hardware reset value) if I don't know what
the units are for delay_ctrl. The hardware that I have works okay with
delay_ctrl of 0, so it seems these new vendor properties are not needed.

Thanks,
Drew

[1] https://lore.kernel.org/lkml/5e379911-e3de-478c-b785-61dbcc9627b1@lunn.ch/