Re: [PATCH] Input: davinci_keyscan: Free memory occupied by davinci key scan device

[arvindY <arvind.yadav.cs@gmail.com>]

On Tuesday 14 November 2017 11:44 PM, Dmitry Torokhov wrote:
> On Tue, Nov 14, 2017 at 07:15:14PM +0530, Arvind Yadav wrote:
>> Here, key_dev is NULL in input_free_device() and It will not free
>> memory occupied by davinci key scan device. Move "key_dev = NULL"
>> after input_free_device().
> No, this is absolutely wrong. It is disallowed to call
> input_free_device() after calling input_unregister_device(), because
> input core will free the device once the last reference is dropped. That
> is why we set "key_dev" to NULL to turn input_free_device() into a noop.
Yes, You are correct. Once device is unregistered it will get freed.

Thanks, :)
>> Signed-off-by: Arvind Yadav <arvind.yadav.cs@gmail.com>
>> ---
>>   drivers/input/keyboard/davinci_keyscan.c | 2 +-
>>   1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/drivers/input/keyboard/davinci_keyscan.c b/drivers/input/keyboard/davinci_keyscan.c
>> index b20a5d0..22cb8d2 100644
>> --- a/drivers/input/keyboard/davinci_keyscan.c
>> +++ b/drivers/input/keyboard/davinci_keyscan.c
>> @@ -290,13 +290,13 @@ static int __init davinci_ks_probe(struct platform_device *pdev)
>>   	free_irq(davinci_ks->irq, davinci_ks);
>>   fail5:
>>   	input_unregister_device(davinci_ks->input);
>> -	key_dev = NULL;
>>   fail4:
>>   	iounmap(davinci_ks->base);
>>   fail3:
>>   	release_mem_region(davinci_ks->pbase, davinci_ks->base_size);
>>   fail2:
>>   	input_free_device(key_dev);
>> +	key_dev = NULL;
>>   fail1:
>>   	kfree(davinci_ks);
>>   
>> -- 
>> 1.9.1
>>
> Thanks.
>