Re: [PATCH v2 4/6] clk: qcom: rpmh: Add support for Kaanapali rpmh clocks

[Taniya Das <taniya.das@oss.qualcomm.com>]

On 11/18/2025 10:04 AM, Dmitry Baryshkov wrote:
> On Mon, Nov 17, 2025 at 01:28:46PM +0530, Taniya Das wrote:
>>
>>
>> On 11/14/2025 4:38 PM, Dmitry Baryshkov wrote:
>>> On Fri, Nov 14, 2025 at 02:13:49PM +0530, Taniya Das wrote:
>>>>
>>>>
>>>> On 11/11/2025 4:16 PM, Dmitry Baryshkov wrote:
>>>>> On Thu, Oct 30, 2025 at 04:39:07PM +0530, Taniya Das wrote:
>>>>>> Add the RPMH clocks present in Kaanapali SoC.
>>>>>>
>>>>>> Signed-off-by: Jingyi Wang <jingyi.wang@oss.qualcomm.com>
>>>>>> Signed-off-by: Taniya Das <taniya.das@oss.qualcomm.com>
>>>>>> ---
>>>>>>  drivers/clk/qcom/clk-rpmh.c | 42 ++++++++++++++++++++++++++++++++++++++++++
>>>>>>  1 file changed, 42 insertions(+)
>>>>>>
>>>>>> diff --git a/drivers/clk/qcom/clk-rpmh.c b/drivers/clk/qcom/clk-rpmh.c
>>>>>> index 1a98b3a0c528c24b600326e6b951b2edb6dcadd7..fd0fe312a7f2830a27e6effc0c0bd905d9d5ebed 100644
>>>>>> --- a/drivers/clk/qcom/clk-rpmh.c
>>>>>> +++ b/drivers/clk/qcom/clk-rpmh.c
>>>>>> @@ -395,6 +395,19 @@ DEFINE_CLK_RPMH_VRM(clk4, _a, "C4A_E0", 1);
>>>>>>  DEFINE_CLK_RPMH_VRM(clk5, _a, "C5A_E0", 1);
>>>>>>  DEFINE_CLK_RPMH_VRM(clk8, _a, "C8A_E0", 1);
>>>>>>  
>>>>>> +DEFINE_CLK_RPMH_VRM(ln_bb_clk1, _a2_e0, "C6A_E0", 2);
>>>>>> +DEFINE_CLK_RPMH_VRM(ln_bb_clk2, _a2_e0, "C7A_E0", 2);
>>>>>> +DEFINE_CLK_RPMH_VRM(ln_bb_clk3, _a2_e0, "C8A_E0", 2);
>>>>>> +
>>>>>> +DEFINE_CLK_RPMH_VRM(rf_clk1, _a_e0, "C1A_E0", 1);
>>>>>> +DEFINE_CLK_RPMH_VRM(rf_clk2, _a_e0, "C2A_E0", 1);
>>>>>
>>>>> What is the difference between these clocks and clk[3458] defined few
>>>>> lines above? Why are they named differently? If the other name is
>>>>> incorrect, please fix it.
>>>>>
>>>>
>>>> Dmitry, my intention was to make a clear distinction between the ‘rf’
>>>> clocks and the ‘ln’ clocks. Since there could be overlap in the
>>>> numbering, I added prefixes for clarity. I should have applied the same
>>>> approach to clk[3458] as well. I will add the fix-up for the same.
>>>
>>> Why do we need to distinguish between them here? The resources in CMD-DB
>>> don't have such a difference. You'll select whether the clock is RF or
>>> LN when describing the platform data.
>>>
>>
>> It is more for readibility and maintain a direct mapping with the PMIC
>> clock grid. This way we can immediately identify the clock type without
>> cross-referencing desc as the clock mapping here would indicate the type
>> of clock. Yes, the CMD-DB name does not reflect anything with the names
>> here. Please do let me know your suggestion.
> 
> I'd prefer if variables reflect CMD-DB resource names rather than the
> actual clock usage. In the end, platform data does exactly that - it
> maps usage to CMD-DB resources.
> 

Actually the grid does not reflect the cmd-db resource name at all, so I
do not think that is the right approach. Let me move to completely
remove the 'ln' and 'rf' references and refer them to 'clkN'.

> Also, if we follow your proposal, we will end up with ln_bb_clk and
> rf_clk referencing exactly the same resource, causing possible
> confusion.
> 

Sure, Dmitry, I will move to use clkN and remove the suffixes.

-- 
Thanks,
Taniya Das