Re: [PATCH 3/4] clk: qcom: Add support for Display Clock Controller on Shikra

[Imran Shaik <imran.shaik@oss.qualcomm.com>]

On 13-05-2026 08:38 pm, Dmitry Baryshkov wrote:
> On Wed, May 13, 2026 at 04:51:03PM +0200, Konrad Dybcio wrote:
>> On 5/13/26 4:06 PM, Dmitry Baryshkov wrote:
>>> On Wed, May 13, 2026 at 05:01:16PM +0300, Dmitry Baryshkov wrote:
>>>> On Wed, May 13, 2026 at 07:10:38PM +0530, Imran Shaik wrote:
>>>>> Add a driver for the Display clock controller on Qualcomm Shikra SoC.
>>>>>
>>>>> Signed-off-by: Imran Shaik <imran.shaik@oss.qualcomm.com>
>>>>> ---
>>>>>   drivers/clk/qcom/Kconfig         |  10 +
>>>>>   drivers/clk/qcom/Makefile        |   1 +
>>>>>   drivers/clk/qcom/dispcc-shikra.c | 565 +++++++++++++++++++++++++++++++++++++++
>>>>>   3 files changed, 576 insertions(+)
>>>>>
>>>>
>>>> Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
>>>
>>> After comparing the files...
>>>
>>> Can we use dispcc-qcm2290.c instead? It uses clock-names instead of
>>> clock-indices, but I think it should be fine to use clock-names as a
>>> one-off.
>>
>> Or we can convert it to use indices, since those are stable for agatti
>> too - the names would remain in the binding, just unused by the driver
> 
> Either is fine for me.
> 

Hi,

In Agatti, apart from the clock-names difference, I see that the AHB/XO 
clocks are not handled as always-on via the probe and instead rely on 
pm_clk style handling, whereas Shikra follows the newer pattern by 
marking required CBCRs as critical during probe. I think that attempting 
to modify this approach into Agatti may introduce unnecessary complexity.

And the Agatti DISPCC doesn't have the DT_DSI1 bindings exposed, and 
updating this might break the ABI with respect to bindings, and DT.

Given these and considering that Agatti is already stable, keeping the 
Shikra as separate GPUCC/DISPCC drivers is better to avoid the risk of 
regressions and complexity.

Please let know your thoughts on this.

Thanks,
Imran