[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20101230174918.GA2018@del.dom.local>
Date: Thu, 30 Dec 2010 18:49:18 +0100
From: Jarek Poplawski <jarkao2@...il.com>
To: Eric Dumazet <eric.dumazet@...il.com>
Cc: David Miller <davem@...emloft.net>, netdev <netdev@...r.kernel.org>
Subject: Re: [PATCH net-next-2.6] sfq: fix slot_dequeue_head()
On Thu, Dec 30, 2010 at 04:02:48PM +0100, Eric Dumazet wrote:
> Le mercredi 22 décembre 2010 ?? 07:32 +0000, Jarek Poplawski a écrit :
> > > Also, slot_dequeue_tail() should make sure slot skb chain is correctly
> > > terminated, or sfq_dump_class_stats() can access freed skbs.
> >
> > ...and a good hint for code reusing ;-)
>
> Yes, and of course same fix is needed in slot_dequeue_head(), as further
> testing on my side made it pretty clear.
>
> I was adding possibility to have more packets queued in SFQ (more
> packets than max number of flows) and got unexpected crashes.
>
> Reverting to net-next-2.6, I still got crashes. Oops.
>
> [PATCH net-next-2.6] sfq: fix slot_dequeue_head()
>
> slot_dequeue_head() should make sure slot skb chain is correct in both
> ways, or we can crash if all possible flows are in use.
Nice scenario ;-) Of course, it's easy to guess I looked for something
like this after your previous fix and missed that :-| Btw, it looks
like slot_queue_init() could go back to sfq_init() now.
Jarek P.
>
> Signed-off-by: Eric Dumazet <eric.dumazet@...il.com>
> Cc: Jarek Poplawski <jarkao2@...il.com>
> ---
> net/sched/sch_sfq.c | 1 +
> 1 files changed, 1 insertion(+)
>
> diff --git a/net/sched/sch_sfq.c b/net/sched/sch_sfq.c
> index 6a2f88f..3977e56 100644
> --- a/net/sched/sch_sfq.c
> +++ b/net/sched/sch_sfq.c
> @@ -292,6 +292,7 @@ static inline struct sk_buff *slot_dequeue_head(struct sfq_slot *slot)
> struct sk_buff *skb = slot->skblist_next;
>
> slot->skblist_next = skb->next;
> + skb->next->prev = (struct sk_buff *)slot;
> skb->next = skb->prev = NULL;
> return skb;
> }
>
>
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Powered by blists - more mailing lists