Re: [PATCH REVIEW 1/2] watchdog: extend the mutex-protected section in watchdog_cdev_unregister()

[Guenter Roeck]

On Thu, Aug 18, 2022 at 03:06:23AM +0100, Alexey Klimov wrote:
> watchdog_stop() should be called with wd_data->lock mutex locked.
> Updates to wdd->status also occur under this look throughout the
> watchdog_dev.c as well as functions that deal with pretimeout hrtimer
> like watchdog_hrtimer_pretimeout_stop() here.
> 

What actual problems are you fixing here ?

Guenter

> Signed-off-by: Alexey Klimov <aklimov@xxxxxxxxxx>
> ---
>  drivers/watchdog/watchdog_dev.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/watchdog/watchdog_dev.c b/drivers/watchdog/watchdog_dev.c
> index 54903f3c851e..804236a094f6 100644
> --- a/drivers/watchdog/watchdog_dev.c
> +++ b/drivers/watchdog/watchdog_dev.c
> @@ -1095,6 +1095,8 @@ static void watchdog_cdev_unregister(struct watchdog_device *wdd)
>  		old_wd_data = NULL;
>  	}
>  
> +	mutex_lock(&wd_data->lock);
> +
>  	if (watchdog_active(wdd) &&
>  	    test_bit(WDOG_STOP_ON_UNREGISTER, &wdd->status)) {
>  		watchdog_stop(wdd);
> @@ -1102,7 +1104,6 @@ static void watchdog_cdev_unregister(struct watchdog_device *wdd)
>  
>  	watchdog_hrtimer_pretimeout_stop(wdd);
>  
> -	mutex_lock(&wd_data->lock);
>  	wd_data->wdd = NULL;
>  	wdd->wd_data = NULL;
>  	mutex_unlock(&wd_data->lock);
> -- 
> 2.37.2
>