Re: [PATCH v6 4/8] hwmon: (amc6821) add support for tsd,mule

[Quentin Schulz <quentin.schulz@cherry.de>]

Hi Krzysztof,

On 8/12/24 1:38 PM, Krzysztof Kozlowski wrote:
> [Some people who received this message don't often get email from krzk@kernel.org. Learn why this is important at https://aka.ms/LearnAboutSenderIdentification ]
> 
> On 31/07/2024 17:12, Guenter Roeck wrote:
>> On Thu, Jul 25, 2024 at 03:27:50PM +0200, Farouk Bouabid wrote:
>>> Theobroma Systems Mule is an MCU that emulates a set of I2C devices,
>>> among which is an amc6821 and other devices that are reachable through
>>> an I2C-mux.
>>>
>>> The devices on the mux can be selected by writing the appropriate device
>>> number to an I2C config register (amc6821: reg 0xff)
>>>
>>> Implement "tsd,mule" compatible to instantiate the I2C-mux platform device
>>> when probing the amc6821.
>>>
>>> Signed-off-by: Farouk Bouabid <farouk.bouabid@cherry.de>
>>> Reviewed-by: Guenter Roeck <linux@roeck-us.net>
>>
>> Applied.
> 
> Eh, there is undocumented dependency on I2C here. Next has warning
> because of this.
> 

I think you meant to comment this on 
https://lore.kernel.org/linux-i2c/20240725-dev-mule-i2c-mux-v6-0-f9f6d7b60fb2@cherry.de/T/#mdb7976f1dc16fce0b7db9abee6fd0b1fd0a2e2ba 
(patch 3 and not 4 of the series). This patch (4) is fine on its own I 
believe, no dependency on anything else. (well, except if we expect 
bindings to be absolutely merged before the drivers? I think what 
matters is the Device Tree changes making use of the new binding be 
merged after dt-binding changes?).

I agree that there's a somewhat non-obvious dependency between patch 1 
and 3 (the dt-bindings) and 5-8 with everything before, we could have 
made this more explicit.

> Farouk, please *always mention* the dependencies between patches.
> 

I wasn't aware of that rule, my apologies for not catching this before 
upstream submission.

For anyone wondering the rule is made explicit here:
https://www.kernel.org/doc/html/latest/process/submitting-patches.html#separate-your-changes

"If one patch depends on another patch in order for a change to be 
complete, that is OK. Simply note “this patch depends on patch X” in 
your patch description."

Question about b4 workflow though. I encourage using b4 to avoid as many 
mistakes as possible and make the workflow as painless as possible. I 
believe b4 doesn't allow you to have per-patch notes, only in the 
cover-letter.
a) is this dependency list in cover-letter acceptable, or
b) need to add it to the patch note (below the ---), or
c) can add it to the patch commit log

I've seen subsystem keep vX changelogs in commit logs, and some who do 
not want it, so maybe there's no one rule here?

Cheers,
Quentin