Date:   Thu, 30 Jun 2022 14:01:33 -0400
From:   Sean Anderson <sean.anderson@...o.com>
To:     Rob Herring <robh@...nel.org>
Cc:     "David S . Miller" <davem@...emloft.net>,
        Jakub Kicinski <kuba@...nel.org>,
        Madalin Bucur <madalin.bucur@....com>, netdev@...r.kernel.org,
        Russell King <linux@...linux.org.uk>,
        Paolo Abeni <pabeni@...hat.com>,
        linux-arm-kernel@...ts.infradead.org,
        Eric Dumazet <edumazet@...gle.com>,
        linux-kernel@...r.kernel.org,
        Kishon Vijay Abraham I <kishon@...com>,
        Krzysztof Kozlowski <krzysztof.kozlowski+dt@...aro.org>,
        Vinod Koul <vkoul@...nel.org>, devicetree@...r.kernel.org,
        linux-phy@...ts.infradead.org
Subject: Re: [PATCH net-next v2 01/35] dt-bindings: phy: Add QorIQ SerDes
 binding

Hi Rob,

On 6/30/22 1:27 PM, Rob Herring wrote:
> On Tue, Jun 28, 2022 at 06:13:30PM -0400, Sean Anderson wrote:
>> This adds a binding for the SerDes module found on QorIQ processors. The
>> phy reference has two cells, one for the first lane and one for the
>> last. This should allow for good support of multi-lane protocols when
>> (if) they are added. There is no protocol option, because the driver is
>> designed to be able to completely reconfigure lanes at runtime.
>> Generally, the phy consumer can select the appropriate protocol using
>> set_mode. For the most part there is only one protocol controller
>> (consumer) per lane/protocol combination. The exception to this is the
>> B4860 processor, which has some lanes which can be connected to
>> multiple MACs. For that processor, I anticipate the easiest way to
>> resolve this will be to add an additional cell with a "protocol
>> controller instance" property.
>> 
>> Each serdes has a unique set of supported protocols (and lanes). The
>> support matrix is stored in the driver and is selected based on the
>> compatible string. It is anticipated that a new compatible string will
>> need to be added for each serdes on each SoC that drivers support is
>> added for. There is no "generic" compatible string for this reason.
>> 
>> There are two PLLs, each of which can be used as the master clock for
>> each lane. Each PLL has its own reference. For the moment they are
>> required, because it simplifies the driver implementation. Absent
>> reference clocks can be modeled by a fixed-clock with a rate of 0.
>> 
>> Signed-off-by: Sean Anderson <sean.anderson@...o.com>
>> ---
>> 
>> Changes in v2:
>> - Add #clock-cells. This will allow using assigned-clocks* to configure
>>   the PLLs.
>> - Allow a value of 1 for phy-cells. This allows for compatibility with
>>   the similar (but according to Ioana Ciornei different enough) lynx-28g
>>   binding.
>> - Document phy cells in the description
>> - Document the structure of the compatible strings
>> - Fix example binding having too many cells in regs
>> - Move compatible first
>> - Refer to the device in the documentation, rather than the binding
>> - Remove minItems
>> - Rename to fsl,lynx-10g.yaml
>> - Use list for clock-names
>> 
>>  .../devicetree/bindings/phy/fsl,lynx-10g.yaml | 93 +++++++++++++++++++
>>  1 file changed, 93 insertions(+)
>>  create mode 100644 Documentation/devicetree/bindings/phy/fsl,lynx-10g.yaml
>> 
>> diff --git a/Documentation/devicetree/bindings/phy/fsl,lynx-10g.yaml b/Documentation/devicetree/bindings/phy/fsl,lynx-10g.yaml
>> new file mode 100644
>> index 000000000000..b5a6f631df9f
>> --- /dev/null
>> +++ b/Documentation/devicetree/bindings/phy/fsl,lynx-10g.yaml
>> @@ -0,0 +1,93 @@
>> +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
>> +%YAML 1.2
>> +---
>> +$id: http://devicetree.org/schemas/phy/fsl,lynx-10g.yaml#
>> +$schema: http://devicetree.org/meta-schemas/core.yaml#
>> +
>> +title: NXP Lynx 10G SerDes
>> +
>> +maintainers:
>> +  - Sean Anderson <sean.anderson@...o.com>
>> +
>> +description: |
>> +  These Lynx "SerDes" devices are found in NXP's QorIQ line of processors. The
>> +  SerDes provides up to eight lanes. Each lane may be configured individually,
>> +  or may be combined with adjacent lanes for a multi-lane protocol. The SerDes
>> +  supports a variety of protocols, including up to 10G Ethernet, PCIe, SATA, and
>> +  others. The specific protocols supported for each lane depend on the
>> +  particular SoC.
>> +
>> +properties:
>> +  compatible:
>> +    description: |
>> +      Each compatible is of the form "fsl,<soc-name>-serdes-<instance>".
>> +      Although many registers are compatible between different SoCs, the
>> +      supported protocols and lane assignments tend to be unique to each SerDes.
>> +      Additionally, the method of activating protocols may also be unique.
> 
> We typically have properties for handling these variables. Numbering 
> instances is something we avoid.

On v1, Krzysztof said that this was a better route...

>> +      Because of this, each SerDes instance will need its own compatible string.
> 
>> +      In cases where two SoCs share the same SerDes implementation (such as the
>> +      LS1046A and LS1026A), both SoCs should share the same compatible strings.
> 
> No, not unless the 2 parts are just fuse or package pinout differences. 
> Otherwise, there's always the chance they are not 'the same' and have 
> different bugs.
>
> You could have "fsl,ls1046a-serdes", "fsl,ls1026a-serdes" (whichever SoC 
> is older last) if you think they are 'the same'.

Fine by me.

>> +    enum:
>> +      - fsl,ls1046a-serdes-1
>> +      - fsl,ls1046a-serdes-2
>> +
>> +  "#clock-cells":
>> +    const: 1
>> +    description: |
>> +      The cell contains the index of the PLL, starting from 0. Note that when
>> +      assigning a rate to a PLL, the PLLs' rates are divided by 1000 to avoid
>> +      overflow. A rate of 5000000 corresponds to 5GHz.
>> +
>> +  "#phy-cells":
>> +    minimum: 1
>> +    maximum: 2
>> +    description: |
>> +      The cells contain the following arguments:
>> +      - The first lane in the group. Lanes are numbered based on the register
>> +        offsets, not the I/O ports. This corresponds to the letter-based ("Lane
>> +        A") naming scheme, and not the number-based ("Lane 0") naming scheme. On
>> +        most SoCs, "Lane A" is "Lane 0", but not always.
>> +      - Last lane. For single-lane protocols, this should be the same as the
>> +        first lane.
>> +      If no lanes in a SerDes can be grouped, then #phy-cells may be 1, and the
>> +      first cell will specify the only lane in the group.
> 
> Usually when we have per lane phys, the consumer side will have a 'phys' 
> entry per lane.

The problem is that it is hard to coordinate multiple lanes coming up at once. There
are particular ordering requirements when multiple lanes are grouped together:

> Note that if the lanes being reconfigured are grouped for multi-lane or synchronous
> mode, then the master source clock lane for the group must have TRST_B set to 1 
> after all the other lanes in the group. The master source clock lane of the group
> is indicated by LNmGCR0[1STLANE]=1. All lanes p<m (if LNmGCR1[TRSTDIR]=0) or p> (if LNmGCR1[TRSTDIR]=1) of the master source clock lane until the next lane with
> LNmGCR0[1STLANE]=1, or the end of the SerDes, are grouped with the master source
> clock lane. All grouped lanes must have the same setting of TRSTDIR.

This is trivial to do if we enable things as a "group" but not so easy if each lane
is a separate phy. In particular, we have to know up front whether lanes are being
grouped together in order to program 1STLANE correctly. I think the enable order
corresponds to the order in the dtb, but AFAIK that is not guaranteed by the phy
subsystem.

> Having a variable number of cells isn't great either. We usually only do 
> that when we have to extend an existing binding.

This is mainly to align closer to the lynx-28g binding. It can always be restricted.

>> +
>> +  clocks:
>> +    maxItems: 2
>> +    description: |
>> +      Clock for each PLL reference clock input.
>> +
>> +  clock-names:
>> +    items:
>> +      - enum: &clocks
>> +          - ref0
>> +          - ref1
>> +      - enum: *clocks
> 
> Whoa, there's a first. Don't use YAML anchors and references. We only 
> support JSON subset of YAML.

How else should this be expressed? I would like to allow both

clock-names = "ref0", "ref1";

and

clock-names = "ref1", "ref0";

ideally without repeating myself.

--Sean

>> +
>> +  reg:
>> +    maxItems: 1
>> +
>> +required:
>> +  - "#clock-cells"
>> +  - "#phy-cells"
>> +  - compatible
>> +  - clocks
>> +  - clock-names
>> +  - reg
>> +
>> +additionalProperties: false
>> +
>> +examples:
>> +  - |
>> +    serdes1: phy@...0000 {
>> +      #clock-cells = <1>;
>> +      #phy-cells = <2>;
>> +      compatible = "fsl,ls1046a-serdes-1";
>> +      reg = <0x1ea0000 0x2000>;
>> +      clocks = <&clk_100mhz>, <&clk_156mhz>;
>> +      clock-names = "ref0", "ref1";
>> +      assigned-clocks = <&serdes1 0>;
>> +      assigned-clock-rates = <5000000>;
>> +    };
>> +
>> +...
>> -- 
>> 2.35.1.1320.gc452695387.dirty
>> 
>> 
> 

Powered by blists - more mailing lists