Re: [PATCH v2 2/2] Add dt-bindings for Cadence XSPI controller

From: Rob Herring
Date: Wed Feb 05 2020 - 13:31:23 EST


On Tue, Jan 28, 2020 at 01:43:04PM +0100, Konrad Kociolek wrote:
> Add dt-bindings documentation for Cadence XSPI controller to support
> SPI based flash memories.
>
> Signed-off-by: Konrad Kociolek <konrad@xxxxxxxxxxx>
> ---
> Changes between initial version and v2:
> - renamed yaml file
> - added missing include
>
> .../devicetree/bindings/spi/cdns,xspi.yaml | 166 +++++++++++++++++++++
> 1 file changed, 166 insertions(+)
> create mode 100644 Documentation/devicetree/bindings/spi/cdns,xspi.yaml
>
> diff --git a/Documentation/devicetree/bindings/spi/cdns,xspi.yaml b/Documentation/devicetree/bindings/spi/cdns,xspi.yaml
> new file mode 100644
> index 000000000000..e8c43957fd90
> --- /dev/null
> +++ b/Documentation/devicetree/bindings/spi/cdns,xspi.yaml
> @@ -0,0 +1,166 @@
> +# SPDX-License-Identifier: (GPL-2.0)

Dual license new bindings:

(GPL-2.0-only OR BSD-2-Clause)

> +# Copyright 2020 Cadence
> +%YAML 1.2
> +---
> +$id: "http://devicetree.org/schemas/spi/cdns,xspi.yaml#";
> +$schema: "http://devicetree.org/meta-schemas/core.yaml#";
> +
> +title: Cadence XSPI Controller
> +
> +maintainers:
> + - Konrad Kociolek <konrad@xxxxxxxxxxx>
> +
> +description: |
> + The XSPI controller allows SPI protocol communication in
> + single, dual, quad or octal wire transmission modes for
> + read/write access to slaves such as SPI-NOR flash.
> +

Needs a ref to spi-controller.yaml

> +properties:
> + compatible:
> + const: cdns,xspi-nor-fpga
> +
> + reg:
> + maxItems: 3
> + description: |
> + Contains three entries, each of which is a tuple consisting of a
> + physical address and length. The first entry is the address and
> + length of the controller register set. The second entry is the
> + address and length of the Slave DMA data port. The third entry is
> + the address and length of auxiliary registers.

Split into 3 descriptions:

items:
- description: ...
- description: ...
- description: ...

With that, drop 'maxItems' as it is implied.

> +
> + interrupts:
> + maxItems: 1
> +
> + cdns,dqs-last-data-drop:
> + type: boolean
> + description: |
> + This parameter should be set when the Flash Device being used
> + issues data on negative edge of Flash clock and returns them with
> + DQS and the PHY is configured to sample data in DQS mode.
> + If this param is set the controller internally requests this redundant
> + data at the end of the transfer cleaning up the PHY FIFO.
> +
> + cdns,phy-data-select-oe-start:
> + $ref: /schemas/types.yaml#/definitions/uint32
> + description: |
> + Adjusts the starting point of the DQ pad output enable window.
> + Lower numbers pull the rising edge earlier in time and larger
> + numbers cause the rising edge to be delayed. Each bit changes
> + the output enable time by a 1/2 cycle resolution.
> +
> + cdns,phy-data-select-oe-end:
> + $ref: /schemas/types.yaml#/definitions/uint32
> + description: |
> + Adjusts the ending point of the DQ pad output enable window.
> + Lower numbers pull the falling edge earlier in time and larger
> + numbers cause the falling edge to be delayed. Each bit changes
> + the output enable time by a 1/2 cycle resolution.
> +
> + cdns,phy-dqs-select-oe-start:
> + $ref: /schemas/types.yaml#/definitions/uint32
> + description: |
> + Adjusts the starting point of the DQS pad output enable window.
> + Lower numbers pull the rising edge earlier in time and larger
> + numbers cause the rising edge to be delayed. Each bit changes
> + the output enable time by a 1/2 cycle resolution.
> +
> + cdns,phy-dqs-select-oe-end:
> + $ref: /schemas/types.yaml#/definitions/uint32
> + description: |
> + Adjusts the ending point of the DQS pad output enable window.
> + Lower numbers pull the falling edge earlier in time and larger
> + numbers cause the falling edge to be delayed. Each bit changes
> + the output enable time by a 1/2 cycle resolution.
> +
> + cdns,phy-gate-cfg-close:
> + $ref: /schemas/types.yaml#/definitions/uint32
> + description: |
> + Normally the gate is closing then all bits of dfi_cebar are high
> + or when dfi_rd_pre_post_amble and rebar_dfi are high. This parameter
> + allows to extend the closing of the DQS gate. Recommended zero.
> +
> + cdns,phy-gate-cfg:
> + $ref: /schemas/types.yaml#/definitions/uint32
> + description: |
> + Coarse adjust of gate open time. This value is the number of cycles
> + to delay the dfi_rddata_en signal prior to opening the gate in
> + full cycle increments. Decreasing this value pulls the gate earlier
> + in time. This field should be programmed such that the gate signal
> + lands in the valid DQS gate window.
> +
> + cdns,phy-rd-del-select:
> + $ref: /schemas/types.yaml#/definitions/uint32
> + description: |
> + Defines the read data delay. Holds the number of cycles to delay
> + the dfi_rddata_en signal prior to enabling the read FIFO.
> + After this delay, the read pointers begin incrementing the read FIFO.
> +
> + cdns,phy-clk-wr-delay:
> + $ref: /schemas/types.yaml#/definitions/uint32
> + description: |
> + Controls the clk_wr delay line which adjusts the write DQ bit
> + timing in 1/256th steps of the clock period in normal DLL
> + locked mode. In bypass mode this field directly programs
> + the number of delay elements.

For all of these, any constraints on the values? default?

> +
> + cdns,phy-use-lpbk-dqs:
> + type: boolean
> + description: |
> + This parameter chooses lpbk_dqs to capture data for reads.
> + Instead memory DQS will be used.
> +
> + cdns,phy-use-ext-lpbk-dqs:
> + type: boolean
> + description: |
> + This parameter chooses external lpbk_dqs for data capture
> + (lpbk_dqs connected to the lpbk_dqs_IO pad). When not used
> + mem_rebar_pad is used for data read capture.
> +
> +required:
> + - compatible
> + - reg
> + - interrupts
> + - cdns,phy-data-select-oe-start
> + - cdns,phy-data-select-oe-end
> + - cdns,phy-dqs-select-oe-start
> + - cdns,phy-dqs-select-oe-end
> + - cdns,phy-gate-cfg-close
> + - cdns,phy-gate-cfg
> + - cdns,phy-rd-del-select
> + - cdns,phy-clk-wr-delay

Is there no sensible default to make these optional?

> +
> +examples:
> + - |
> + #include <dt-bindings/interrupt-controller/irq.h>
> +
> + xspi: spi@a0010000 {
> + #address-cells = <1>;
> + #size-cells = <0>;
> + compatible = "cdns,xspi-nor-fpga";
> + reg = <0x0 0xa0010000 0x0 0x10000>,
> + <0x0 0xb0000000 0x0 0x10000>,
> + <0x0 0xa0020000 0x0 0x10000>;
> + interrupts = <0 90 IRQ_TYPE_LEVEL_HIGH>;
> + interrupt-parent = <&gic>;
> + cdns,dqs-last-data-drop;
> + cdns,phy-data-select-oe-start = <0>;
> + cdns,phy-data-select-oe-end = <4>;
> + cdns,phy-dqs-select-oe-start = <0>;
> + cdns,phy-dqs-select-oe-end = <1>;
> + cdns,phy-gate-cfg-close = <3>;
> + cdns,phy-gate-cfg = <0>;
> + cdns,phy-rd-del-select = <5>;
> + cdns,phy-clk-wr-delay = <64>;
> + cdns,phy-use-lpbk-dqs;
> + cdns,phy-use-ext-lpbk-dqs;
> + mt35xu512@0 {

flash@0

> + compatible = "spi-nor", "micron,mt35xu512";

Wrong order. Most specific first.

> + spi-max-frequency = <75000000>;
> + reg = <0>;
> + };
> + mt35xu512@1 {

flash@1

> + compatible = "spi-nor", "micron,mt35xu512";
> + spi-max-frequency = <75000000>;
> + reg = <1>;
> + };
> + };
> --
> 2.15.0
>