Re: [PATCH v0 03/14] drm/gma500,drm/i915: Make I2C terminology more inclusive

[Jani Nikula]

On Tue, 02 Apr 2024, Easwar Hariharan <eahariha@xxxxxxxxxxxxxxxxxxx> wrote:
> On 4/2/2024 7:32 AM, Jani Nikula wrote:
>> On Tue, 02 Apr 2024, Easwar Hariharan <eahariha@xxxxxxxxxxxxxxxxxxx> wrote:
>>> On 4/2/2024 12:48 AM, Jani Nikula wrote:
>>>> On Fri, 29 Mar 2024, Easwar Hariharan <eahariha@xxxxxxxxxxxxxxxxxxx> wrote:
>>>>> I2C v7, SMBus 3.2, and I3C specifications have replaced "master/slave"
>>>>> with more appropriate terms. Inspired by and following on to Wolfram's
>>>>> series to fix drivers/i2c/[1], fix the terminology for users of
>>>>> I2C_ALGOBIT bitbanging interface, now that the approved verbiage exists
>>>>> in the specification.
>>>>
>>>> gma500 and i915 changes should be split. See MAINTAINERS.
>>>>
>>>> Might also split the i915 changes to smaller pieces, it's kind of
>>>> random. And the changes here are not strictly related to I2C AFAICT, so
>>>> the commit message should be updated.
>>>>
>>>> BR,
>>>> Jani.
>>>>
>>>>
>>>
>>> <snip>
>>>
>>> I will split gma500 and i915 into their respective patches if possible in v2.
>>>
>>> Can you say more about the changes being "not strictly related to I2C"? My
>>> heuristic was to grep for master/slave, and look in the surrounding context for
>>> i2c-related terminology (i2c_pin, 7-bit address, struct i2c_adapter, i2c_bus, etc)
>>> to confirm that they are i2c-related, then following the references around to
>>> make the compiler happy. For e.g., I did not change the many references to bigjoiner
>>> master and slave because I understood from context they were not i2c references.
>>>
>>> A couple examples would help me restrict the changes to I2C, since as mentioned in the
>>> discussion on Wolfram's thread, there are places where migrating away from master/slave
>>> terms in the code would conflict with the original technical manuals and reduce correlation
>>> and understanding of the code.
>> 
>> I guess I was looking at the VBT changes in intel_bios.c. Granted, they
>> do end up being used as i2c addresses. No big deal.
>> 
>> I think I'd expect the treewide i2c adapter changes to land first, via
>> i2c, and subsequent cleanups to happen next, via individual driver
>> trees. There's quite a bit of conflict potential merging this outside of
>> drm-intel-next, and there's really no need for that.
>> 
>> BR,
>> Jani.
>> 
>
> Great! Just so I'm clear, do you still want the i915 changes split up more, along with them being
> split off from gma500?

If we can merge the i915 changes via drm-intel-next, it's probably fine
as a big i915 patch. Just the gma500 separated. (The struct
i2c_algorithm change etc. necessarily has to go via I2C tree of course.)

BR,
Jani.



>
> Thanks,
> Easwar

-- 
Jani Nikula, Intel