Re: [PATCH] dt-bindings: mmc: describe the Allwinner A733 eMMC controller

[Krzysztof Kozlowski]

On Fri, Oct 30, 2026 at 10:26:47PM -0400, Enzo Adriano wrote:
> SMHC3 uses a different DMA descriptor format from SMHC0/2: the length
> field is 12 bits and the ownership/control word carries different
> control and high-address fields. An older A100/D1 fallback would claim
> unsupported programming compatibility with the existing sunxi-mmc
> descriptor builder.
> 
> Give SMHC3 a distinct compatible and describe its register range,
> interrupt, bus/module/storage-MBUS/MSI-Lite1 clocks and bus reset. This
> binding permits an accurate disabled hardware description; no
> operational driver match or eMMC functionality is claimed here.
> 
> AI assistance: Codex:gpt-6-astra.

Drop, we do not advertise slop companies anymore.

> 
> Assisted-by: LLM
Drop

> Signed-off-by: Enzo Adriano <enzo.adriano.code@xxxxxxxxx>
> ---
> This binding is a separate prerequisite for the disabled MMC3 node in

How this can be a separate prerequisite? Binding without user is not a
prerequisite.

> the A733 DT v2. It does not add a sunxi-mmc driver match or claim an
> operational eMMC path.

Hm? Why?

> 
> The original A733 manual and public vendor implementation distinguish
> SMHC3's descriptor control/high-address fields and 12-bit buffer
> length from SMHC0-2. Its resources are the 0x04023000/0x1000 register
> window, SPI 164, AHB/module/storage-MBUS/MSI-Lite1 clocks and AHB reset.
> The generic A733/D1 MMC0-2 compatible already in the base is unchanged.
> 
> The example uses the public A733 CCU header IDs. The tested standalone
> base is official commit d89e1d186ec9271bf78e62771779e9fc03bf0fc3 plus
> the clock helpers and CCU v6:
> https://lore.kernel.org/r/20260906-sunxi-clk-no-p-v1-0-cbde21c9fe69@xxxxxxxxxxxxx
> https://lore.kernel.org/r/20261009-a733-clk-v6-0-ae8178fce7e8@xxxxxxxxxxxxx

...

> 
> base-commit: d89e1d186ec9271bf78e62771779e9fc03bf0fc3
> prerequisite-patch-id: 0df2ae41836b86fd868dfa01ddfde57f888f761a
> prerequisite-patch-id: 88f3a431e7c551a9fa732f062188f10fdd362eeb
> prerequisite-patch-id: 3e88bc873c909059d1199f82cda4013765e7a5e5
> prerequisite-patch-id: 0dbc27cd7417352319a0705a051dff35085e8922
> prerequisite-patch-id: 0ed6b92774b5b794761c7ce34db653a5435dc833
> prerequisite-patch-id: 27ed3814dbd0c7ad830532ff125625917c873b4b
> prerequisite-patch-id: 74d03898c5ce14763f4e89fb73e021de59f0fb9f
> prerequisite-patch-id: 9225cb094cf6cbb338b14ac63711d68bed12d24d
> prerequisite-patch-id: 568dae1c91a0d7318d6eb14a18054343b84b603a
> prerequisite-patch-id: ff3806332e741f0e2c8dac1c11f4c71449c16c8c
> prerequisite-patch-id: 39461e0543aefb5c40184664625cbab19ac8408a
> prerequisite-patch-id: 6fc03bfbd343042dd3b660938fcc09c5e7d48314

So this cannot really be tested.

Best regards,
Krzysztof