Message-ID: <03e97e7c-9793-4fd4-8733-f7cc43a4cf6c@nxp.com>
Date: Mon, 14 Oct 2024 13:54:17 +0800
From: Liu Ying <victor.liu@....com>
To: Dmitry Baryshkov <dmitry.baryshkov@...aro.org>
Cc: dri-devel@...ts.freedesktop.org, devicetree@...r.kernel.org,
 linux-kernel@...r.kernel.org, imx@...ts.linux.dev,
 linux-arm-kernel@...ts.infradead.org, andrzej.hajda@...el.com,
 neil.armstrong@...aro.org, rfoss@...nel.org,
 Laurent.pinchart@...asonboard.com, jonas@...boo.se,
 jernej.skrabec@...il.com, airlied@...il.com, simona@...ll.ch,
 maarten.lankhorst@...ux.intel.com, mripard@...nel.org, tzimmermann@...e.de,
 robh@...nel.org, krzk+dt@...nel.org, conor+dt@...nel.org,
 shawnguo@...nel.org, s.hauer@...gutronix.de, kernel@...gutronix.de,
 festevam@...il.com, catalin.marinas@....com, will@...nel.org,
 quic_bjorande@...cinc.com, geert+renesas@...der.be, arnd@...db.de,
 nfraprado@...labora.com, o.rempel@...gutronix.de, y.moog@...tec.de,
 marex@...x.de, isaac.scott@...asonboard.com, biju.das.jz@...renesas.com
Subject: Re: [PATCH v2 5/9] dt-bindings: display: bridge: Add ITE IT6263 LVDS
 to HDMI converter

On 10/14/2024, Liu Ying wrote:
> On 10/14/2024, Dmitry Baryshkov wrote:
>> On Sat, Oct 12, 2024 at 05:14:13PM +0800, Liu Ying wrote:
>>> On 10/12/2024, Dmitry Baryshkov wrote:
>>>> On Sat, Oct 12, 2024 at 03:35:39PM +0800, Liu Ying wrote:
>>>>> Document ITE IT6263 LVDS to HDMI converter.
>>>>>
>>>>> Product link:
>>>>> https://www.ite.com.tw/en/product/cate1/IT6263
>>>>>
>>>>> Signed-off-by: Liu Ying <victor.liu@....com>
>>>>> ---
>>>>> v2:
>>>>> * Document number of LVDS link data lanes.  (Biju)
>>>>> * Simplify ports property by dropping "oneOf".  (Rob)
>>>>>
>>>>>  .../bindings/display/bridge/ite,it6263.yaml   | 276 ++++++++++++++++++
>>>>>  1 file changed, 276 insertions(+)
>>>>>  create mode 100644 Documentation/devicetree/bindings/display/bridge/ite,it6263.yaml
>>>>>
>>>>> diff --git a/Documentation/devicetree/bindings/display/bridge/ite,it6263.yaml b/Documentation/devicetree/bindings/display/bridge/ite,it6263.yaml
>>>>> new file mode 100644
>>>>> index 000000000000..bc2bbec07623
>>>>> --- /dev/null
>>>>> +++ b/Documentation/devicetree/bindings/display/bridge/ite,it6263.yaml
>>>>> @@ -0,0 +1,276 @@
>>>>> +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
>>>>> +%YAML 1.2
>>>>> +---
>>>>> +$id: http://devicetree.org/schemas/display/bridge/ite,it6263.yaml#
>>>>> +$schema: http://devicetree.org/meta-schemas/core.yaml#
>>>>> +
>>>>> +title: ITE IT6263 LVDS to HDMI converter
>>>>> +
>>>>> +maintainers:
>>>>> +  - Liu Ying <victor.liu@....com>
>>>>> +
>>>>> +description: |
>>>>> +  The IT6263 is a high-performance single-chip De-SSC(De-Spread Spectrum) LVDS
>>>>> +  to HDMI converter.  Combined with LVDS receiver and HDMI 1.4a transmitter,
>>>>> +  the IT6263 supports LVDS input and HDMI 1.4 output by conversion function.
>>>>> +  The built-in LVDS receiver can support single-link and dual-link LVDS inputs,
>>>>> +  and the built-in HDMI transmitter is fully compliant with HDMI 1.4a/3D, HDCP
>>>>> +  1.2 and backward compatible with DVI 1.0 specification.
>>>>> +
>>>>> +  The IT6263 also encodes and transmits up to 8 channels of I2S digital audio,
>>>>> +  with sampling rate up to 192KHz and sample size up to 24 bits. In addition,
>>>>> +  an S/PDIF input port takes in compressed audio of up to 192KHz frame rate.
>>>>> +
>>>>> +  The newly supported High-Bit Rate(HBR) audio by HDMI specifications v1.3 is
>>>>> +  provided by the IT6263 in two interfaces: the four I2S input ports or the
>>>>> +  S/PDIF input port.  With both interfaces the highest possible HBR frame rate
>>>>> +  is supported at up to 768KHz.
>>>>> +
>>>>> +properties:
>>>>
>>>> No LVDS data-mapping support?
>>>
>>> It is enough to document number of LVDS link data lanes
>>> because OS should be able to determine the data-mapping
>>> by looking at the number and the data-mapping capability
>>> of the other side of the LVDS link. 
>>
>> From what I can see, data-mapping is specified on the consumer sink side
>> of the LVDS link. This means it should go to the bridge's device node.
> 
> Then, I won't define data-lanes, because data-mapping implies it,
> e.g., jeida-24 implies data lanes 0/1/2/3, see lvds-data-mapping.yaml.
> 
> Please let me know which one you prefer.
> 
>>
>>>
>>>>
>>>>> +  compatible:
>>>>> +    const: ite,it6263
>>>>> +
>>>>> +  reg:
>>>>> +    maxItems: 1
>>>>> +
>>>>> +  clocks:
>>>>> +    maxItems: 1
>>>>> +    description: audio master clock
>>>>> +
>>>>> +  clock-names:
>>>>> +    const: mclk
>>>>> +
>>>>> +  reset-gpios:
>>>>> +    maxItems: 1
>>>>> +
>>>>> +  ivdd-supply:
>>>>> +    description: 1.8V digital logic power
>>>>> +
>>>>> +  ovdd-supply:
>>>>> +    description: 3.3V I/O pin power
>>>>> +
>>>>> +  txavcc18-supply:
>>>>> +    description: 1.8V HDMI analog frontend power
>>>>> +
>>>>> +  txavcc33-supply:
>>>>> +    description: 3.3V HDMI analog frontend power
>>>>> +
>>>>> +  pvcc1-supply:
>>>>> +    description: 1.8V HDMI frontend core PLL power
>>>>> +
>>>>> +  pvcc2-supply:
>>>>> +    description: 1.8V HDMI frontend filter PLL power
>>>>> +
>>>>> +  avcc-supply:
>>>>> +    description: 3.3V LVDS frontend power
>>>>> +
>>>>> +  anvdd-supply:
>>>>> +    description: 1.8V LVDS frontend analog power
>>>>> +
>>>>> +  apvdd-supply:
>>>>> +    description: 1.8V LVDS frontend PLL power
>>>>> +
>>>>> +  "#sound-dai-cells":
>>>>> +    const: 0
>>>>> +
>>>>> +  ite,lvds-link-num-data-lanes:
>>>>> +    $ref: /schemas/types.yaml#/definitions/uint8
>>>>> +    enum: [3, 4, 5]
>>>>> +    description: number of data lanes per LVDS link
>>>>
>>>> Please use data-lanes property inside the OF graph.
>>>
>>> In both port@0 and port@1?
>>
>> Yes
> 
> I'm assuming if data-mapping is defined, then no need to
> define data-lanes.
> 
>>
>>>
>>>>
>>>>> +
>>>>> +  ite,i2s-audio-fifo-sources:
>>>>> +    $ref: /schemas/types.yaml#/definitions/uint32-array
>>>>> +    minItems: 1
>>>>> +    maxItems: 4
>>>>> +    items:
>>>>> +      enum: [0, 1, 2, 3]
>>>>> +    description:
>>>>> +      Each array element indicates the pin number of an I2S serial data input
>>>>> +      line which is connected to an audio FIFO, from audio FIFO0 to FIFO3.
>>>>
>>>> What does that mean from the board point of view? Routed audio links for
>>>> the multichannel audio?
>>>
>>> Yes, also for single channel audio.
>>>
>>>>
>>>>> +
>>>>> +  ite,rl-channel-swap-audio-sources:
>>>>> +    $ref: /schemas/types.yaml#/definitions/uint32-array
>>>>> +    minItems: 1
>>>>> +    maxItems: 4
>>>>> +    uniqueItems: true
>>>>> +    items:
>>>>> +      enum: [0, 1, 2, 3]
>>>>> +    description:
>>>>> +      Each array element indicates an audio source whose right channel and left
>>>>> +      channel are swapped by this converter. For I2S, the element is the pin
>>>>> +      number of an I2S serial data input line. For S/PDIF, the element is always
>>>>> +      0.
>>>>
>>>> Why?
>>>
>>> Because this converter has the capability to swap right channel
>>> and left channel and S/PDIF always uses audio source0. 
>>>
>>>>
>>>>> +
>>>>> +  ports:
>>>>> +    $ref: /schemas/graph.yaml#/properties/ports
>>>>> +
>>>>> +    properties:
>>>>> +      port@0:
>>>>> +        $ref: /schemas/graph.yaml#/$defs/port-base
>>>>> +        unevaluatedProperties: false
>>>>> +        description:
>>>>> +          The first LVDS input link.
>>>>> +          In dual-link LVDS mode, this link works together with the second LVDS
>>>>> +          input link, and one link receives odd pixels while the other receives
>>>>> +          even pixels. Specify the pixels with one of the dual-lvds-odd-pixels
>>>>> +          and dual-lvds-even-pixels properties when and only when dual-link LVDS
>>>>> +          mode is used.
>>>>> +
>>>>> +        properties:
>>>>> +          dual-lvds-odd-pixels:
>>>>> +            type: boolean
>>>>> +            description: the first sink port for odd pixels
>>>>> +
>>>>> +          dual-lvds-even-pixels:
>>>>> +            type: boolean
>>>>> +            description: the first sink port for even pixels
>>>>> +
>>>>> +      port@1:
>>>>> +        $ref: /schemas/graph.yaml#/$defs/port-base
>>>>> +        unevaluatedProperties: false
>>>>> +        description: the second LVDS input link
>>>>> +
>>>>> +        properties:
>>>>> +          dual-lvds-even-pixels:
>>>>> +            type: boolean
>>>>> +            description: the second sink port for even pixels
>>>>> +
>>>>> +          dual-lvds-odd-pixels:
>>>>> +            type: boolean
>>>>> +            description: the second sink port for odd pixels
>>>>> +
>>>>> +        oneOf:
>>>>> +          - required: [dual-lvds-even-pixels]
>>>>> +          - required: [dual-lvds-odd-pixels]
>>>>
>>>> This still allows one to specify that both ports provide odd pixels. Is
>>>> that expected? Also why do you need two properties which specify the
>>>> same option.
>>>
>>> No, that is not expected. Description for port@0 already mentions
>>> one link receives odd pixels while the other receives even pixels.
>>
>> That's not expected, but permitted by the bindings.
> 
> It is not permitted by port@1. If "dual-lvds-odd-pixels;" is added
> to port@1 in the dual-link LVDS example, the below warning will be
> generated by dt_binding_check.
> 
> Documentation/devicetree/bindings/display/bridge/ite,it6263.example.dtb: hdmi@4c: ports:port@1: {'reg': [[1]], 'dual-lvds-even-pixels': True, 'dual-lvds-odd-pixels': True, 'endpoint': {'remote-endpoint': [[4294967295]]}} is valid under each of {'required': ['dual-lvds-odd-pixels']}, {'required': ['dual-lvds-even-pixels']}
> 	from schema $id: http://devicetree.org/schemas/display/bridge/ite,it6263.yaml#
> 
> However, the binding for port@0 does allow DT writers to specify
> both even and odd pixels. I raised similar concerns in v1 discussion.
> According to Rob's comment in #devicetree IRC, the ports property
> is simplified to this form and more description for port@0 is added
> to tell when to specify the even/odd pixels.  If you know any better
> way to indicate the constraints, please shout.

Dmitry, if the binding in v1 may pass dt_binding_check against
dtschema-2024.9, then it looks like all constraints are implemented.
But, it can't pass. 

> 
>>
>>>
>>> Two options are supported for dual-link LVDS, not the same option:
>>> 1) LVDS link1(port@0) gets odd pixels
>>>    and
>>>    LVDS link2(port@1) gets even pixels.
>>>
>>> 2) LVDS link1(port@0) gets even pixels
>>>    and
>>>    LVDS link2(port@1) gets odd pixels.
>>> That's the reason why each port needs two properties to define
>>> odd/even pixels.
>>>
>>>>
>>>> My suggestion would be to add a single root-level property which
>>>> specifies which port provides even pixel data.
>>>
>>> That won't work.  The LVDS source side expects the ports of
>>> the sink side specify dual-lvds-{odd,even}-pixels properties.
>>
>> I didn't notice that these properties are already defined.
>>
>> As these properties are common between several schema files, please
>> extract them to a common schema file (like lvds.yaml).
> 
> I'm not sure how to do that. Is it obvious?
> Please shed some light. 
> 
> Only two panel schema files are defining even/odd pixels now -
> advantech,idk-2121wr.yaml and panel-simple-lvds-dual-ports.yaml.
> Maybe, extract them later when more schema files(especially for
> bridges) try to define the same?  I'd like to keep a low profile
> for now.
> 
>>
>>>
>>>>
>>>>> +
>>>>> +      port@2:
>>>>> +        $ref: /schemas/graph.yaml#/properties/port
>>>>> +        description: video port for the HDMI output
>>>>> +
>>>>> +      port@3:
>>>>> +        $ref: /schemas/graph.yaml#/properties/port
>>>>> +        description: sound input port
>>>>> +
>>>>> +    required:
>>>>> +      - port@0
>>>>> +      - port@2
>>>>> +
>>>>> +required:
>>>>> +  - compatible
>>>>> +  - reg
>>>>> +  - ivdd-supply
>>>>> +  - ovdd-supply
>>>>> +  - txavcc18-supply
>>>>> +  - txavcc33-supply
>>>>> +  - pvcc1-supply
>>>>> +  - pvcc2-supply
>>>>> +  - avcc-supply
>>>>> +  - anvdd-supply
>>>>> +  - apvdd-supply
>>>>> +  - ite,lvds-link-num-data-lanes
>>>>> +  - ports
>>>>> +
>>>>> +additionalProperties: false
>>>>> +
>>>>> +examples:
>>>>> +  - |
>>>>> +    /* single-link LVDS input */
>>>>> +    #include <dt-bindings/gpio/gpio.h>
>>>>> +
>>>>> +    i2c {
>>>>> +        #address-cells = <1>;
>>>>> +        #size-cells = <0>;
>>>>> +
>>>>> +        hdmi@4c {
>>>>> +            compatible = "ite,it6263";
>>>>> +            reg = <0x4c>;
>>>>> +            reset-gpios = <&gpio1 10 GPIO_ACTIVE_LOW>;
>>>>> +            ivdd-supply = <&reg_buck5>;
>>>>> +            ovdd-supply = <&reg_vext_3v3>;
>>>>> +            txavcc18-supply = <&reg_buck5>;
>>>>> +            txavcc33-supply = <&reg_vext_3v3>;
>>>>> +            pvcc1-supply = <&reg_buck5>;
>>>>> +            pvcc2-supply = <&reg_buck5>;
>>>>> +            avcc-supply = <&reg_vext_3v3>;
>>>>> +            anvdd-supply = <&reg_buck5>;
>>>>> +            apvdd-supply = <&reg_buck5>;
>>>>> +            ite,lvds-link-num-data-lanes = /bits/ 8 <4>;
>>>>> +
>>>>> +            ports {
>>>>> +                #address-cells = <1>;
>>>>> +                #size-cells = <0>;
>>>>> +
>>>>> +                port@0 {
>>>>> +                    reg = <0>;
>>>>> +
>>>>> +                    it6263_lvds_link1: endpoint {
>>>>> +                        remote-endpoint = <&ldb_lvds_ch0>;
>>>>> +                    };
>>>>> +                };
>>>>> +
>>>>> +                port@2 {
>>>>> +                    reg = <2>;
>>>>> +
>>>>> +                    it6263_out: endpoint {
>>>>> +                        remote-endpoint = <&hdmi_in>;
>>>>> +                    };
>>>>> +                };
>>>>> +            };
>>>>> +        };
>>>>> +    };
>>>>> +
>>>>> +  - |
>>>>> +    /* dual-link LVDS input */
>>>>> +    #include <dt-bindings/gpio/gpio.h>
>>>>> +
>>>>> +    i2c {
>>>>> +        #address-cells = <1>;
>>>>> +        #size-cells = <0>;
>>>>> +
>>>>> +        hdmi@4c {
>>>>> +            compatible = "ite,it6263";
>>>>> +            reg = <0x4c>;
>>>>> +            reset-gpios = <&gpio1 10 GPIO_ACTIVE_LOW>;
>>>>> +            ivdd-supply = <&reg_buck5>;
>>>>> +            ovdd-supply = <&reg_vext_3v3>;
>>>>> +            txavcc18-supply = <&reg_buck5>;
>>>>> +            txavcc33-supply = <&reg_vext_3v3>;
>>>>> +            pvcc1-supply = <&reg_buck5>;
>>>>> +            pvcc2-supply = <&reg_buck5>;
>>>>> +            avcc-supply = <&reg_vext_3v3>;
>>>>> +            anvdd-supply = <&reg_buck5>;
>>>>> +            apvdd-supply = <&reg_buck5>;
>>>>> +            ite,lvds-link-num-data-lanes = /bits/ 8 <4>;
>>>>> +
>>>>> +            ports {
>>>>> +                #address-cells = <1>;
>>>>> +                #size-cells = <0>;
>>>>> +
>>>>> +                port@0 {
>>>>> +                    reg = <0>;
>>>>> +                    dual-lvds-odd-pixels;
>>>>> +
>>>>> +                    it6263_lvds_link1_dual: endpoint {
>>>>> +                        remote-endpoint = <&ldb_lvds_ch0>;
>>>>> +                    };
>>>>> +                };
>>>>> +
>>>>> +                port@1 {
>>>>> +                    reg = <1>;
>>>>> +                    dual-lvds-even-pixels;
>>>>> +
>>>>> +                    it6263_lvds_link2_dual: endpoint {
>>>>> +                        remote-endpoint = <&ldb_lvds_ch1>;
>>>>> +                    };
>>>>> +                };
>>>>> +
>>>>> +                port@2 {
>>>>> +                    reg = <2>;
>>>>> +
>>>>> +                    it6263_out_dual: endpoint {
>>>>> +                        remote-endpoint = <&hdmi_in>;
>>>>> +                    };
>>>>> +                };
>>>>> +            };
>>>>> +        };
>>>>> +    };
>>>>> -- 
>>>>> 2.34.1
>>>>>
>>>>
>>>
>>> -- 
>>> Regards,
>>> Liu Ying
>>>
>>
> 

-- 
Regards,
Liu Ying

Powered by blists - more mailing lists