public inbox for dmaengine@vger.kernel.org
 help / color / mirror / Atom feed
From: "Joseph, Abin" <Abin.Joseph@amd.com>
To: Krzysztof Kozlowski <krzk@kernel.org>
Cc: vkoul@kernel.org, Frank.Li@kernel.org, robh@kernel.org,
	krzk+dt@kernel.org, conor+dt@kernel.org, michal.simek@amd.com,
	radhey.shyam.pandey@amd.com, git@amd.com,
	dmaengine@vger.kernel.org, devicetree@vger.kernel.org,
	linux-arm-kernel@lists.infradead.org,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH v2] dt-bindings: dmaengine: xlnx,axi-dma: Convert bindings into yaml
Date: Tue, 3 Mar 2026 12:16:27 +0530	[thread overview]
Message-ID: <98bc3a1e-27dc-4b05-a949-d8830cc552d1@amd.com> (raw)
In-Reply-To: <20260225-cerise-oryx-of-pluck-8e2f30@quoll>


Hi Krzysztof,


On 2/25/2026 3:52 PM, Krzysztof Kozlowski wrote:
> Caution: This message originated from an External Source. Use proper caution when opening attachments, clicking links, or responding.
> 
> 
> On Wed, Feb 25, 2026 at 10:35:21AM +0530, Abin Joseph wrote:
>> Convert the bindings document for Xilinx DMA from txt to yaml.
>> No changes to existing binding description.
>>
>> Signed-off-by: Abin Joseph <abin.joseph@amd.com>
>> ---
>>
>> v2:
>> -> Add examples for each compatible
> 
> No, why? We did not ask for that and (see writing schema) we ask for one
> example, more only if they are different.
> 

Sorry about that, my mistake.

Based on Frank Li’s comment,
[the YAML schema example should use only one of the compatible strings.]

I misunderstood this and thought we needed
to add examples for each compatible.

I’ll fix this and update it correctly in the next revision.


> 
> ...
> 
> 
>> +properties:
>> +  compatible:
>> +    enum:
>> +      - xlnx,axi-cdma-1.00.a
>> +      - xlnx,axi-dma-1.00.a
>> +      - xlnx,axi-mcdma-1.00.a
>> +      - xlnx,axi-vdma-1.00.a
>> +
>> +  reg:
>> +    maxItems: 1
>> +
>> +  "#dma-cells":
>> +    const: 1
>> +
>> +  "#address-cells":
>> +    const: 1
>> +
>> +  "#size-cells":
>> +    const: 1
>> +
>> +  interrupts:
>> +    items:
>> +      - description: Interrupt for single channel (MM2S or S2MM)
>> +      - description: Interrupt for dual channel configuration
>> +    minItems: 1
>> +    description:
>> +      Interrupt lines for the DMA controller. Only used when
>> +      xlnx,axistream-connected is present (DMA connected to AXI Stream
>> +      IP). When child dma-channel nodes are present, interrupts are
>> +      specified in the child nodes instead.
>> +
>> +  clocks:
>> +    minItems: 1
>> +    maxItems: 5
>> +
>> +  clock-names:
>> +    minItems: 1
>> +    maxItems: 5
>> +
>> +  dma-ranges: true
>> +
>> +  xlnx,addrwidth:
>> +    $ref: /schemas/types.yaml#/definitions/uint32
>> +    enum: [32, 64]
>> +    description: The DMA addressing size in bits.
>> +
>> +  xlnx,num-fstores:
>> +    $ref: /schemas/types.yaml#/definitions/uint32
>> +    minimum: 1
>> +    maximum: 32
>> +    description: Should be the number of framebuffers as configured in h/w.
>> +
>> +  xlnx,flush-fsync:
>> +    type: boolean
>> +    description: Tells which channel to Flush on Frame sync.
>> +
>> +  xlnx,sg-length-width:
>> +    $ref: /schemas/types.yaml#/definitions/uint32
>> +    minimum: 8
>> +    maximum: 26
>> +    default: 23
>> +    description:
>> +      Width in bits of the length register as configured in hardware.
>> +
>> +  xlnx,irq-delay:
>> +    $ref: /schemas/types.yaml#/definitions/uint32
>> +    minimum: 0
>> +    maximum: 255
>> +    description:
>> +      Tells the interrupt delay timeout value. Valid range is from 0-255.
>> +      Setting this value to zero disables the delay timer interrupt.
>> +      1 timeout interval = 125 * clock period of SG clock.
>> +
>> +  xlnx,axistream-connected:
>> +    type: boolean
>> +    description: Tells whether DMA is connected to AXI stream IP.
>> +
>> +patternProperties:
>> +  "^dma-channel(-mm2s|-s2mm)?$":
>> +    type: object
>> +    description:
>> +      Should have at least one channel and can have up to two channels per
>> +      device. This node specifies the properties of each DMA channel.
>> +
>> +    properties:
>> +      compatible:
>> +        enum:
>> +          - xlnx,axi-vdma-mm2s-channel
>> +          - xlnx,axi-vdma-s2mm-channel
>> +          - xlnx,axi-cdma-channel
>> +          - xlnx,axi-dma-mm2s-channel
>> +          - xlnx,axi-dma-s2mm-channel
>> +
>> +      interrupts:
>> +        maxItems: 1
>> +
>> +      xlnx,datawidth:
>> +        $ref: /schemas/types.yaml#/definitions/uint32
>> +        enum: [32, 64, 128, 256, 512, 1024]
>> +        description: Should contain the stream data width, take values {32,64...1024}.
>> +
>> +      xlnx,include-dre:
>> +        type: boolean
>> +        description: Tells hardware is configured for Data Realignment Engine.
>> +
>> +      xlnx,genlock-mode:
>> +        type: boolean
>> +        description: Tells Genlock synchronization is enabled/disabled in hardware.
>> +
>> +      xlnx,enable-vert-flip:
>> +        type: boolean
>> +        description:
>> +          Tells vertical flip is enabled/disabled in hardware(S2MM path).
>> +
>> +      dma-channels:
>> +        $ref: /schemas/types.yaml#/definitions/uint32
>> +        description: Number of dma channels in child node.
>> +
>> +    required:
>> +      - compatible
>> +      - interrupts
>> +      - xlnx,datawidth
>> +
>> +    additionalProperties: false
>> +
>> +allOf:
>> +  - $ref: ../dma-controller.yaml#
>> +
>> +  - if:
>> +      properties:
>> +        compatible:
>> +          contains:
>> +            const: xlnx,axi-vdma-1.00.a
>> +    then:
>> +      properties:
>> +        clock-names:
>> +          contains:
>> +            const: s_axi_lite_aclk
> 
> This is REALLY unexpected syntax. You are supposed to have strictly
> ordered list - why do you need need it to be so flexible? And if element
> is required it should be the first item. There is no single DTS even
> mentioning it!


I will fix this syntax

> 
>> +          items:
>> +            enum:
>> +              - s_axi_lite_aclk
>> +              - m_axi_mm2s_aclk
>> +              - m_axi_s2mm_aclk
>> +              - m_axis_mm2s_aclk
>> +              - s_axis_s2mm_aclk
>> +          minItems: 1
>> +          maxItems: 5
>> +      patternProperties:
>> +        "^dma-channel(-mm2s|-s2mm)?$":
>> +          properties:
>> +            compatible:
>> +              enum:
>> +                - xlnx,axi-vdma-mm2s-channel
>> +                - xlnx,axi-vdma-s2mm-channel
>> +      required:
>> +        - xlnx,num-fstores
>> +
>> +  - if:
>> +      properties:
>> +        compatible:
>> +          contains:
>> +            const: xlnx,axi-cdma-1.00.a
>> +    then:
>> +      properties:
>> +        clock-names:
>> +          items:
>> +            - const: s_axi_lite_aclk
>> +            - const: m_axi_aclk
>> +      patternProperties:
>> +        "^dma-channel(-mm2s|-s2mm)?$":
>> +          properties:
>> +            compatible:
>> +              enum:
>> +                - xlnx,axi-cdma-channel
>> +
>> +  - if:
>> +      properties:
>> +        compatible:
>> +          contains:
>> +            enum:
>> +              - xlnx,axi-dma-1.00.a
>> +              - xlnx,axi-mcdma-1.00.a
>> +    then:
>> +      properties:
>> +        clock-names:
>> +          contains:
>> +            const: s_axi_lite_aclk
> 
> Why do you need this?
> 
>> +          items:
>> +            enum:
>> +              - s_axi_lite_aclk
>> +              - m_axi_mm2s_aclk
>> +              - m_axi_s2mm_aclk
>> +              - m_axi_sg_aclk
> 
> Why this cannot be ordered list like we expect (see writing bindings)?
> 

sure will do it.

> 
>> +          minItems: 1
>> +          maxItems: 4
>> +      patternProperties:
>> +        "^dma-channel(-mm2s|-s2mm)?(@[0-9a-f]+)?$":
>> +          properties:
>> +            compatible:
>> +              enum:
>> +                - xlnx,axi-dma-mm2s-channel
>> +                - xlnx,axi-dma-s2mm-channel
>> +
>> +  - if:
>> +      properties:
>> +        compatible:
>> +          contains:
>> +            enum:
>> +              - xlnx,axi-cdma-1.00.a
>> +              - xlnx,axi-mcdma-1.00.a
>> +              - xlnx,axi-dma-1.00.a
>> +    then:
>> +      properties:
>> +        interrupts: false
> 
> Why interrupts are flexible in other case?
> 
> This should be probably squashed in each of previous if:then:.

I will squash the interrupt constraints into each if:then: block as 
suggested.

For VDMA and CDMA, interrupts are always in child nodes regardless of
axistream-connected, so I'll add "interrupts: false" to their respective 
blocks.

For AXI DMA and MCDMA:
- When xlnx,axistream-connected is present: interrupts are at parent level
- When not connected to AXI stream: interrupts are in child nodes

Since the current schema already allows both parent and child node 
interrupts
for DMA/MCDMA (flexible behavior), I won't add "interrupts: false" to their
block, which correctly permits both configurations.

I will fix this in the next version.


> 
> 
>> +
>> +required:
>> +  - "#dma-cells"
>> +  - reg
>> +  - xlnx,addrwidth
>> +  - dma-ranges
>> +  - clocks
>> +  - clock-names
>> +
>> +additionalProperties: false
>> +
>> +examples:
>> +  - |
>> +    #include <dt-bindings/interrupt-controller/arm-gic.h>
>> +
>> +    dma-controller@40030000 {
>> +        compatible = "xlnx,axi-vdma-1.00.a";
>> +        #dma-cells = <1>;
>> +        #address-cells = <1>;
>> +        #size-cells = <1>;
>> +        reg = <0x40030000 0x10000>;
>> +        dma-ranges = <0x0 0x0 0x40000000>;
>> +        xlnx,num-fstores = <8>;
>> +        xlnx,flush-fsync;
>> +        xlnx,addrwidth = <32>;
>> +        clocks = <&clk 0>, <&clk 1>, <&clk 2>, <&clk 3>, <&clk 4>;
>> +        clock-names = "s_axi_lite_aclk", "m_axi_mm2s_aclk",
>> +                      "m_axi_s2mm_aclk", "m_axis_mm2s_aclk",
>> +                      "s_axis_s2mm_aclk";
>> +
>> +        dma-channel-mm2s {
>> +            compatible = "xlnx,axi-vdma-mm2s-channel";
>> +            interrupts = <GIC_SPI 54 IRQ_TYPE_LEVEL_HIGH>;
>> +            xlnx,datawidth = <64>;
>> +        };
>> +
>> +        dma-channel-s2mm {
>> +            compatible = "xlnx,axi-vdma-s2mm-channel";
>> +            interrupts = <GIC_SPI 53 IRQ_TYPE_LEVEL_HIGH>;
>> +            xlnx,datawidth = <64>;
>> +        };
>> +    };
>> +
>> +  - |
>> +    #include <dt-bindings/interrupt-controller/arm-gic.h>
>> +
>> +    dma-controller@a4030000 {
>> +        compatible = "xlnx,axi-dma-1.00.a";
>> +        #dma-cells = <1>;
>> +        #address-cells = <1>;
>> +        #size-cells = <1>;
>> +        reg = <0xa4030000 0x10000>;
>> +        dma-ranges = <0x0 0x0 0x40000000>;
>> +        xlnx,addrwidth = <32>;
>> +        xlnx,sg-length-width = <14>;
>> +        clocks = <&clk 0>, <&clk 1>, <&clk 2>, <&clk 3>;
>> +        clock-names = "s_axi_lite_aclk", "m_axi_mm2s_aclk",
>> +                      "m_axi_s2mm_aclk", "m_axi_sg_aclk";
>> +
>> +        dma-channel-mm2s {
>> +            compatible = "xlnx,axi-dma-mm2s-channel";
>> +            interrupts = <GIC_SPI 86 IRQ_TYPE_LEVEL_HIGH>;
>> +            xlnx,datawidth = <64>;
>> +            xlnx,include-dre;
>> +        };
>> +
>> +        dma-channel-s2mm {
>> +            compatible = "xlnx,axi-dma-s2mm-channel";
>> +            interrupts = <GIC_SPI 87 IRQ_TYPE_LEVEL_HIGH>;
>> +            xlnx,datawidth = <64>;
>> +            xlnx,include-dre;
>> +        };
>> +    };
>> +
> 
> Drop all examples past this point.

sure, thanks

Abin Joseph

> 
> Best regards,
> Krzysztof
> 


  reply	other threads:[~2026-03-03  6:46 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-02-25  5:05 [PATCH v2] dt-bindings: dmaengine: xlnx,axi-dma: Convert bindings into yaml Abin Joseph
2026-02-25 10:22 ` Krzysztof Kozlowski
2026-03-03  6:46   ` Joseph, Abin [this message]
2026-02-25 10:23 ` Krzysztof Kozlowski

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=98bc3a1e-27dc-4b05-a949-d8830cc552d1@amd.com \
    --to=abin.joseph@amd.com \
    --cc=Frank.Li@kernel.org \
    --cc=conor+dt@kernel.org \
    --cc=devicetree@vger.kernel.org \
    --cc=dmaengine@vger.kernel.org \
    --cc=git@amd.com \
    --cc=krzk+dt@kernel.org \
    --cc=krzk@kernel.org \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=michal.simek@amd.com \
    --cc=radhey.shyam.pandey@amd.com \
    --cc=robh@kernel.org \
    --cc=vkoul@kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox