On 22/11/2023 19:46, Krzysztof Kozlowski wrote:
> On 22/11/2023 19:39, William Zhang wrote:
>> Hi,
>>
>> On 11/22/2023 07:52 AM, Rafał Miłecki wrote:
>>> On 22.11.2023 16:50, Krzysztof Kozlowski wrote:
>>>> On 22/11/2023 16:49, Rafał Miłecki wrote:
>>>>>>> For example a year ago I added binding for BCMBCA SoC timer without
>>>>>>> actual driver, see e112f2de151b ("dt-bindings: timer: Add Broadcom's
>>>>>>> BCMBCA timers").
>>>>>>>
>>>>>>> I'm not sure if we're going to agree on this, but personally I like
>>>>>>> describing hardware as much as I can. So it's well documented /
>>>>>>> understood and people may eventually write drivers for it. Maybe it's
>>>>>>> partially because I come from Broadcom's world that isn't well known
>>>>>>> for upstream efforts in general.
>>>>>>
>>>>>> The problem is that "brcm,bcmbca-hs-uart" is not describing
>>>>>> hardware. It
>>>>>> is saying that all these devices have similar (compatible) programming
>>>>>> model, so the OS can use just one compatible. This goes away from pure
>>>>>> hardware description into interpretation.
>>>>>>
>> It is the same hardware IP block used in bcmbca SoCs. To me, it
>> perfectly describe the hardware IP block and it does not need fallback
>> because there is no fallback. We did that for SPI controller although
>> it has two revisions of that IP block so we have brcm,bcmbca-hsspi-v1.0
>> and 1.1
>>
>>>>>> Rob already commented on such non-SoC compatibles multiple times. I do
>>>>>> not see any reason here to not use specific compatible as fallback.
>>>>>
>> Sorry I missed Rob's comments. If we have any new rule or notes about
>> this, I would like to check it out.
>>
>>>>> Do I get it right we should rather have some base specific compatible
>>>>> like: "brcm,bcm63138-hs-uart" and then if anything use fallback to it
>>>>> like: "brcm,bcm4908-hs-uart", "brcm,bcm63138-hs-uart"; ?
>>>>
>>>> Yes, or the other way around, depends which is probably the oldest.
>> If we absolutely can not use bcmbca-hs-uart, I would suggest to use
>
> We can, but I am surprised that you want without any driver. What's the
> point of generic compatible?
>
>> bcm63xx-hs-uart to be more soc specific and in fact the oldest SoC have
>
> What is xx? Wildcard? I mean... ehhh...
OK, it's not worth my time. Neither Rafał's.
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@xxxxxxxxxx>
I can go to Embedded OSS every year and give the same speech every year
and still people will on:
1. insist on generic fallback compatible,
2. wildcards
3. families
I will keep this email and use it to justify the same, third speech next
year. Which won't be listened to, so I will go in 2025 fourth time. :)
Best regards,
Krzysztof
