Re: [PATCH v2] watchdog: avoid usage of iterator after loop

[Guenter Roeck]

On 3/13/23 06:59, Jakob Koschel wrote:
> If potentially no valid element is found, 'p' would contain an invalid
> pointer past the iterator loop. To ensure 'p' is valid under any
> circumstances, the kfree() should be within the loop body.
>
> Additionally, Linus proposed to avoid any use of the list iterator
> variable after the loop, in the attempt to move the list iterator
> variable declaration into the macro to avoid any potential misuse after
> the loop [1].
>
> Link: https://lore.kernel.org/all/CAHk-=wgRr_D8CB-D9Kg-c=EHreAsk5SqXPwr9Y7k9sA6cWXJ6w@xxxxxxxxxxxxxx/ [1]
> Signed-off-by: Jakob Koschel <jkl820.git@xxxxxxxxx>

Reviewed-by: Guenter Roeck <linux@xxxxxxxxxxxx>

> ---
> Changes in v2:
> - avoid returning within the list iterator
> - Link to v1: https://lore.kernel.org/r/20230301-watchdog-avoid-iter-after-loop-v1-1-851f583be9f7@xxxxxxxxx
> ---
>   drivers/watchdog/watchdog_pretimeout.c | 3 +--
>   1 file changed, 1 insertion(+), 2 deletions(-)
>
> diff --git a/drivers/watchdog/watchdog_pretimeout.c b/drivers/watchdog/watchdog_pretimeout.c
> index 376a495ab80c..e5295c990fa1 100644
> --- a/drivers/watchdog/watchdog_pretimeout.c
> +++ b/drivers/watchdog/watchdog_pretimeout.c
> @@ -207,10 +207,9 @@ void watchdog_unregister_pretimeout(struct watchdog_device *wdd)
>   	list_for_each_entry_safe(p, t, &pretimeout_list, entry) {
>   		if (p->wdd == wdd) {
>   			list_del(&p->entry);
> +			kfree(p);
>   			break;
>   		}
>   	}
>   	spin_unlock_irq(&pretimeout_lock);
> -
> -	kfree(p);
>   }
>
> ---
> base-commit: c0927a7a5391f7d8e593e5e50ead7505a23cadf9
> change-id: 20230301-watchdog-avoid-iter-after-loop-a197bf201435
>
> Best regards,