net_sched: Bug in netem reordering
authorHagen Paul Pfeifer <hagen@jauu.net>
Wed, 4 Jan 2012 17:35:26 +0000 (17:35 +0000)
committerDavid S. Miller <davem@davemloft.net>
Thu, 5 Jan 2012 18:27:39 +0000 (13:27 -0500)
Not now, but it looks you are correct. q->qdisc is NULL until another
additional qdisc is attached (beside tfifo). See 50612537e9ab2969312.
The following patch should work.

From: Hagen Paul Pfeifer <hagen@jauu.net>

netem: catch NULL pointer by updating the real qdisc statistic

Reported-by: Vijay Subramanian <subramanian.vijay@gmail.com>
Signed-off-by: Hagen Paul Pfeifer <hagen@jauu.net>
Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
net/sched/sch_netem.c

index 06a5cebad3426b67df67c41f39d2275ae8163102..e7e1d0b57b3d2a28d5e99276cd195f7dfde999d0 100644 (file)
@@ -458,8 +458,8 @@ static int netem_enqueue(struct sk_buff *skb, struct Qdisc *sch)
                q->counter = 0;
 
                __skb_queue_head(&sch->q, skb);
-               q->qdisc->qstats.backlog += qdisc_pkt_len(skb);
-               q->qdisc->qstats.requeues++;
+               sch->qstats.backlog += qdisc_pkt_len(skb);
+               sch->qstats.requeues++;
                ret = NET_XMIT_SUCCESS;
        }