On Mon, Jun 5, 2023, at 14:07, Nicolas Ferre wrote:
> On 05/06/2023 at 08:42, Krzysztof Kozlowski wrote:
>>>
>>> diff --git a/drivers/net/ethernet/cadence/macb_main.c b/drivers/net/ethernet/cadence/macb_main.c
>>> index 29a1199dad14..609c8e9305ba 100644
>>> --- a/drivers/net/ethernet/cadence/macb_main.c
>>> +++ b/drivers/net/ethernet/cadence/macb_main.c
>>> @@ -4913,6 +4913,7 @@ static const struct of_device_id macb_dt_ids[] = {
>>> { .compatible = "microchip,mpfs-macb", .data = &mpfs_config },
>>> { .compatible = "microchip,sama7g5-gem", .data = &sama7g5_gem_config },
>>> { .compatible = "microchip,sama7g5-emac", .data = &sama7g5_emac_config },
>>> + { .compatible = "microchip,sam9x7-gem", .data = &sama7g5_gem_config },
>>
>> These are compatible, aren't they? Why do you need new entry?
>
> The hardware itself is different, even if the new features are not
> supported yet in the macb driver.
> The macb driver will certainly evolve in order to add these features so
> we decided to match a new compatible string all the way to the driver.
It sounds like you can still drop this patch though, and only add a
specific entry here after the .data field is actually different
when those features get added.
The important bit for now is to have the specific string in the binding
and in the dtb, along with the fallback for I assume "microchip,sama7g5-gem".
Arnd
