Re: [PATCH RFT 09/10] arm64: dts: microchip: sparx5_pcb134: drop duplicated NOR flash

From: Krzysztof Kozlowski
Date: Fri Apr 05 2024 - 11:45:58 EST


On 05/04/2024 14:10, Steen Hegelund wrote:
> Hi Krzysztof,
>
> On Mon, 2024-04-01 at 17:37 +0200, Krzysztof Kozlowski wrote:
>> EXTERNAL EMAIL: Do not click links or open attachments unless you know the content is safe
>>
>> Since beginning the DTS extended the SPI0 in two places adding two SPI
>> muxes, each with same SPI NOR flash.  Both used exactly the same
>> chip-selects, so this was clearly buggy code.  Without checking in
>> datasheet, assume device has only one SPI NOR flash, so code was
>> duplicated.
>>
>> Fixes dtc W=1 warnings:
>>
>>   sparx5_pcb134_board.dtsi:277.10-281.4: Warning (unique_unit_address_if_enabled):
>> /axi@600000000/spi@600104000/flash@0: duplicate unit-address (also used in node
>> /axi@600000000/spi@600104000/spi@0)
>>
>> Signed-off-by: Krzysztof Kozlowski <krzk@xxxxxxxxxx>
>>
>> ---
>>
>> Not tested on hardware
>> ---
>>  .../boot/dts/microchip/sparx5_pcb134_board.dtsi  | 16 ----------------
>>  1 file changed, 16 deletions(-)
>>
>> diff --git a/arch/arm64/boot/dts/microchip/sparx5_pcb134_board.dtsi
>> b/arch/arm64/boot/dts/microchip/sparx5_pcb134_board.dtsi
>> index f165a409bc1d..dc7b59dfcb40 100644
>> --- a/arch/arm64/boot/dts/microchip/sparx5_pcb134_board.dtsi
>> +++ b/arch/arm64/boot/dts/microchip/sparx5_pcb134_board.dtsi
>> @@ -281,22 +281,6 @@ flash@0 {
>>         };
>>  };
>>
>> -&spi0 {
>> -       status = "okay";
>> -       spi@0 {
>> -               compatible = "spi-mux";
>> -               mux-controls = <&mux>;
>> -               #address-cells = <1>;
>> -               #size-cells = <0>;
>> -               reg = <0>;      /* CS0 */
>> -               flash@9 {
>> -                       compatible = "jedec,spi-nor";
>> -                       spi-max-frequency = <8000000>;
>> -                       reg = <0x9>;    /* SPI */
>> -               };
>> -       };
>> -};
>> -
>
> When testing this on actual HW the SPI NOR is no longer accessible.
> The reason is that it sits behind a SPI-MUX and that needs to be present in the Device Tree.
>
> So if you do the "reverse" clean-up it works fine: Remove the simple spi0 node and keep the one that
> has the spi-mux reference.

Aaaa, so we need to do it that way. Thanks for testing! I'll send v2.

Best regards,
Krzysztof