Re: [PATCH v1 1/3] pinctrl: qcom: lpass-lpi: Switch to PM clock framework for runtime PM

[Ajay Kumar Nandam <ajay.nandam@oss.qualcomm.com>]

On 4/14/2026 2:07 PM, Konrad Dybcio wrote:
> On 4/13/26 2:22 PM, Ajay Kumar Nandam wrote:
>> Convert the LPASS LPI pinctrl driver to use the PM clock framework for
>> runtime power management.
>>
>> This allows the LPASS LPI pinctrl driver to drop clock votes when idle,
>> improves power efficiency on platforms using LPASS LPI island mode, and
>> aligns the driver with common runtime PM patterns used across Qualcomm
>> LPASS subsystems.
>>
>> Signed-off-by: Ajay Kumar Nandam <ajay.nandam@oss.qualcomm.com>
>> ---
>>   drivers/pinctrl/qcom/pinctrl-lpass-lpi.c      | 36 +++++++++++++------
>>   drivers/pinctrl/qcom/pinctrl-lpass-lpi.h      |  2 ++
>>   .../pinctrl/qcom/pinctrl-sc7280-lpass-lpi.c   |  5 +++
>>   3 files changed, 32 insertions(+), 11 deletions(-)
>>
>> diff --git a/drivers/pinctrl/qcom/pinctrl-lpass-lpi.c b/drivers/pinctrl/qcom/pinctrl-lpass-lpi.c
>> index 76aed3296..6d50e06ef 100644
>> --- a/drivers/pinctrl/qcom/pinctrl-lpass-lpi.c
>> +++ b/drivers/pinctrl/qcom/pinctrl-lpass-lpi.c
>> @@ -14,15 +14,16 @@
>>   
>>   #include <linux/pinctrl/pinconf-generic.h>
>>   #include <linux/pinctrl/pinconf.h>
>> +#include <linux/pm_runtime.h>
>>   #include <linux/pinctrl/pinmux.h>
>>   
>>   #include "../pinctrl-utils.h"
>>   
>>   #include "pinctrl-lpass-lpi.h"
>> +#include <linux/pm_clock.h>
> 
> Please move it up, together with other non-local includes
> 
> [...]

ACK, will correct in the next version

Thanks
Ajay Kumar Nandam

> 
>> +	pm_runtime_set_autosuspend_delay(dev, 100);
>> +	pm_runtime_use_autosuspend(dev);
>> +	pm_runtime_enable(dev);
> 
> devm_pm_runtime_enable() will let you drop the manual disablement below and
> in .remove()

ACK, will switch in the next version.

Thanks
Ajay Kumar Nandam

> 
>>   
>>   	pctrl->desc.pctlops = &lpi_gpio_pinctrl_ops;
>>   	pctrl->desc.pmxops = &lpi_gpio_pinmux_ops;
>> @@ -539,20 +540,33 @@ int lpi_pinctrl_probe(struct platform_device *pdev)
>>   	return 0;
>>   
>>   err_pinctrl:
>> +	pm_runtime_disable(dev);
>>   	mutex_destroy(&pctrl->lock);
>> -	clk_bulk_disable_unprepare(MAX_LPI_NUM_CLKS, pctrl->clks);
>>   
>>   	return ret;
>>   }
> 
> [...]
> 
>> --- a/drivers/pinctrl/qcom/pinctrl-sc7280-lpass-lpi.c
>> +++ b/drivers/pinctrl/qcom/pinctrl-sc7280-lpass-lpi.c
>> @@ -139,10 +139,15 @@ static const struct of_device_id lpi_pinctrl_of_match[] = {
>>   };
>>   MODULE_DEVICE_TABLE(of, lpi_pinctrl_of_match);
>>   
>> +static const struct dev_pm_ops lpi_pinctrl_pm_ops = {
>> +	RUNTIME_PM_OPS(lpi_pinctrl_runtime_suspend, lpi_pinctrl_runtime_resume, NULL)
>> +};
>> +
>>   static struct platform_driver lpi_pinctrl_driver = {
>>   	.driver = {
>>   		   .name = "qcom-sc7280-lpass-lpi-pinctrl",
>>   		   .of_match_table = lpi_pinctrl_of_match,
>> +		   .pm = pm_ptr(&lpi_pinctrl_pm_ops),
> 
> I believe SoCs other than kodiak also require this change to avoid
> regressions.

This patch is currently limited to SC7280, where the runtime PM behavior
has been validated.

Based on this feedback, I’ll revisit the series and extend the same
runtime PM support to the other LPASS‑LPI SoC drivers that share this
block, so the behavior remains consistent across platforms and avoids
potential regressions.

Thanks
Ajay Kumar Nandam



> 
> Konrad