Re: [PATCH] usb/usbtmc: Remove a redundant kref_put in usbtmc_disconnect

[Greg KH <gregkh@linuxfoundation.org>]

On Thu, Mar 11, 2021 at 07:03:54AM -0800, Lv Yunlong wrote:
> In the implementation of usbtmc_free_int(), it already calls
> kref_put() to free the data as shown below. So, in
> usbtmc_disconnect, call an extra kref_put() is redundant.
> 
> """
> static void usbtmc_free_int(struct usbtmc_device_data *data)
> {
> 	if (!data->iin_ep_present || !data->iin_urb)
> 		return;
> 	usb_kill_urb(data->iin_urb);
> 	kfree(data->iin_buffer);
> 	data->iin_buffer = NULL;
> 	usb_free_urb(data->iin_urb);
> 	data->iin_urb = NULL;
> 	kref_put(&data->kref, usbtmc_delete);
> }
> """
> 
> Signed-off-by: Lv Yunlong <lyl2019@mail.ustc.edu.cn>
> ---
>  drivers/usb/class/usbtmc.c | 1 -
>  1 file changed, 1 deletion(-)
> 
> diff --git a/drivers/usb/class/usbtmc.c b/drivers/usb/class/usbtmc.c
> index 74d5a9c5238a..adcdd2df1949 100644
> --- a/drivers/usb/class/usbtmc.c
> +++ b/drivers/usb/class/usbtmc.c
> @@ -2494,7 +2494,6 @@ static void usbtmc_disconnect(struct usb_interface *intf)
>  	}
>  	mutex_unlock(&data->io_mutex);
>  	usbtmc_free_int(data);
> -	kref_put(&data->kref, usbtmc_delete);

Did you test this change?

If not, please do so, I think the code is fine as-is.

thanks,

greg k-h