Re: [PATCH] vmpressure: fix divide-by-0 in vmpressure_work_fn

[Anton Vorontsov <anton@enomsg.org>]

On Fri, Sep 06, 2013 at 10:59:16PM -0700, Hugh Dickins wrote:
> Hit divide-by-0 in vmpressure_work_fn(): checking vmpr->scanned before
> taking the lock is not enough, we must check scanned afterwards too.
> 
> Signed-off-by: Hugh Dickins <hughd@google.com>
> Cc: stable@vger.kernel.org

Hm... Just trying to understand this one. I don't see how this can happen,
considering that only one instance of vmpressure_work_fn() supposed to be
running (unlike vmpressure()), and the only place where we zero
vmpr->scanned is vmpressure_work_fn() itself?

> ---
> 
>  mm/vmpressure.c |    3 +++
>  1 file changed, 3 insertions(+)
> 
> --- 3.11/mm/vmpressure.c	2013-09-02 13:46:10.000000000 -0700
> +++ linux/mm/vmpressure.c	2013-09-06 22:43:03.596003080 -0700
> @@ -187,6 +187,9 @@ static void vmpressure_work_fn(struct wo
>  	vmpr->reclaimed = 0;
>  	spin_unlock(&vmpr->sr_lock);
>  
> +	if (!scanned)
> +		return;
> +
>  	do {
>  		if (vmpressure_event(vmpr, scanned, reclaimed))
>  			break;

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>