Re: [PATCH v4 3/3] input: misc: Add Qualcomm SPMI PMIC haptics driver

[Fenglin Wu]

On 7/18/2026 12:51 AM, Dmitry Torokhov wrote:
>> +
>> +static void qcom_haptics_remove(struct platform_device *pdev)
>> +{
>> +	struct qcom_haptics *h = platform_get_drvdata(pdev);
>> +
>> +	/*
>> +	 * Unregister the input device explicitly at the beginning
>> +	 * to avoid the input device being used after the resources
>> +	 * are freed.
>> +	 */
>> +	input_unregister_device(h->input);
>> +	disable_delayed_work_sync(&h->play_work);
>> +	scoped_guard(mutex, &h->play_lock) {
>> +		haptics_queue_flush(h);
>> +		haptics_stop_locked(h, true);
>> +	}
>> +
>> +	haptics_enable_module(h, false);
>> +
>> +	scoped_guard(mutex, &h->fifo_lock) {
>> +		for (int i = 0; i < HAPTICS_MAX_EFFECTS; i++) {
>> +			kvfree(h->effects[i].fifo_data);
>> +			h->effects[i].fifo_data = NULL;
>> +		}
>> +	}
> 
> Most of this is better suited to close() method of input device.

Hi Dmitry,

I tried moving this into the close() function of the input device, but
I’m running into an issue: 'disable_delayed_work_sync(&h->play_work)'
may be called multiple times when the input device is opened and closed,
and the delayed work can sometimes end up being disabled unexpectedly.
That would mean I need another flag to track whether the delayed work is
enabled. Do you see any concerns with that, or should I still keep it in
remove()?

Thanks
Fenglin
> 
>> +}

>