hi i tried the example given on the examples page to duplicate selected traffic like tc qdisc add dev eth0 root handle 1: prio tc qdisc add dev eth0 parent 1:3 handle 3: netem duplicate 40% tc filter add dev eth0 protocol ip parent 1:0 prio 3 u32 match ip dst 11.0.2.2 flowid 1:3 when i ping from 11.0.2.2 to this interface my machine hangs. the same thing works for drop or delay. i would appreciate if someone can tell me what i am doing wrong. thanks. __________________________________ Do you Yahoo!? Yahoo! Small Business - Try our new resources site! http://smallbusiness.yahoo.com/resources/
js si wrote:> hi > > i tried the example given on the examples page to > duplicate selected traffic like > > tc qdisc add dev eth0 root handle 1: prio > tc qdisc add dev eth0 parent 1:3 handle 3: netem > duplicate 40% > tc filter add dev eth0 protocol ip parent 1:0 prio 3 > u32 match ip dst 11.0.2.2 flowid 1:3 > > when i ping from 11.0.2.2 to this interface my machine > hangs. the same thing works for drop or delay. > > i would appreciate if someone can tell me what i am > doing wrong.Using netem with an upper qdisc screws up the upper qdisc''s q.qlen counter and qdisc_run() tries to dequeue it indefinitely. I don''t know whether netem is intended to be useable only as top-level qdisc, if so it should make sure that it really is used this way to avoid hanging the system. Regards Patrick
On Wed, 30 Mar 2005 10:05:04 -0800 (PST) js si <nistnet_user@yahoo.com> wrote:> hi > > i tried the example given on the examples page to > duplicate selected traffic like > > tc qdisc add dev eth0 root handle 1: prio > tc qdisc add dev eth0 parent 1:3 handle 3: netem > duplicate 40% > tc filter add dev eth0 protocol ip parent 1:0 prio 3 > u32 match ip dst 11.0.2.2 flowid 1:3 >Try this. qdisc_restart() has a bug. It will spin until packet becomes available, which is wrong. This effects both netem and tbf, maybe other qdisc as well. --- linux-2.6/net/sched/sch_generic.c 2005-03-14 14:30:52.000000000 -0800 +++ tcp-2.6/net/sched/sch_generic.c 2005-04-04 10:28:31.000000000 -0700 @@ -176,10 +176,10 @@ int qdisc_restart(struct net_device *dev requeue: q->ops->requeue(skb, q); - netif_schedule(dev); - return 1; + netif_schedule(dev); + return q->q.qlen; } - return q->q.qlen; + return 0; } static void dev_watchdog(unsigned long arg)
Stephen Hemminger
2005-Apr-04 21:04 UTC
Re: [Netem] Re: netem with prio hangs on duplicate
On Mon, 4 Apr 2005 11:00:36 -0700 Stephen Hemminger <shemminger@osdl.org> wrote:> On Wed, 30 Mar 2005 10:05:04 -0800 (PST) > js si <nistnet_user@yahoo.com> wrote: > > > hi > > > > i tried the example given on the examples page to > > duplicate selected traffic like > > > > tc qdisc add dev eth0 root handle 1: prio > > tc qdisc add dev eth0 parent 1:3 handle 3: netem > > duplicate 40% > > tc filter add dev eth0 protocol ip parent 1:0 prio 3 > > u32 match ip dst 11.0.2.2 flowid 1:3 > > > > Try this. qdisc_restart() has a bug. It will spin until packet becomes available, which > is wrong. This effects both netem and tbf, maybe other qdisc as well. >Never mind, that patch is crap...
On Wed, 30 Mar 2005 10:05:04 -0800 (PST) js si <nistnet_user@yahoo.com> wrote:> hi > > i tried the example given on the examples page to > duplicate selected traffic like > > tc qdisc add dev eth0 root handle 1: prio > tc qdisc add dev eth0 parent 1:3 handle 3: netem > duplicate 40% > tc filter add dev eth0 protocol ip parent 1:0 prio 3 > u32 match ip dst 11.0.2.2 flowid 1:3 > > when i ping from 11.0.2.2 to this interface my machine > hangs. the same thing works for drop or delay. > > i would appreciate if someone can tell me what i am > doing wrong.I can''t reproduce this on my SMP box, will try UP. Could you get a backtrace with sysrq-P and sysrq-T?
Try this alternative, it changes where netem does the delaying and
doesn''t
do queuing in the timer routine. It is stable for the basic tests, but I still
consider
it experimental and needs more testing.
Patch against the version of sch_netem.c in 2.6.12-rc2
--- linux-2.6.12-rc2/net/sched/sch_netem.c 2005-04-04 09:39:41.000000000 -0700
+++ netem-2.6.12-rc2/net/sched/sch_netem.c 2005-04-06 15:39:16.000000000 -0700
@@ -138,38 +138,78 @@ static long tabledist(unsigned long mu,
}
/* Put skb in the private delayed queue. */
-static int delay_skb(struct Qdisc *sch, struct sk_buff *skb)
+static int netem_delay(struct Qdisc *sch, struct sk_buff *skb)
{
struct netem_sched_data *q = qdisc_priv(sch);
- struct netem_skb_cb *cb = (struct netem_skb_cb *)skb->cb;
psched_tdiff_t td;
psched_time_t now;
PSCHED_GET_TIME(now);
td = tabledist(q->latency, q->jitter, &q->delay_cor,
q->delay_dist);
- PSCHED_TADD2(now, td, cb->time_to_send);
/* Always queue at tail to keep packets in order */
if (likely(q->delayed.qlen < q->limit)) {
+ struct netem_skb_cb *cb = (struct netem_skb_cb *)skb->cb;
+
+ PSCHED_TADD2(now, td, cb->time_to_send);
+
+ pr_debug("netem_delay: skb=%p now=%llu tosend=%llu\n", skb,
+ now, cb->time_to_send);
+
__skb_queue_tail(&q->delayed, skb);
- if (!timer_pending(&q->timer)) {
- q->timer.expires = jiffies + PSCHED_US2JIFFIE(td);
- add_timer(&q->timer);
- }
return NET_XMIT_SUCCESS;
}
+ pr_debug("netem_delay: queue over limit %d\n", q->limit);
+ sch->qstats.overlimits++;
kfree_skb(skb);
return NET_XMIT_DROP;
}
+/*
+ * Move a packet that is ready to send from the delay holding
+ * list to the underlying qdisc.
+ */
+static int netem_run(struct Qdisc *sch)
+{
+ struct netem_sched_data *q = qdisc_priv(sch);
+ struct sk_buff *skb;
+ psched_time_t now;
+
+ PSCHED_GET_TIME(now);
+
+ skb = skb_peek(&q->delayed);
+ if (skb) {
+ const struct netem_skb_cb *cb
+ = (const struct netem_skb_cb *)skb->cb;
+ long delay
+ = PSCHED_US2JIFFIE(PSCHED_TDIFF(cb->time_to_send, now));
+ pr_debug("netem_run: skb=%p delay=%ld\n", skb, delay);
+
+ /* if more time remaining? */
+ if (delay > 0) {
+ mod_timer(&q->timer, jiffies + delay);
+ return 1;
+ }
+
+ __skb_unlink(skb, &q->delayed);
+
+ if (q->qdisc->enqueue(skb, q->qdisc)) {
+ sch->q.qlen--;
+ sch->qstats.drops++;
+ }
+ }
+
+ return 0;
+}
+
static int netem_enqueue(struct sk_buff *skb, struct Qdisc *sch)
{
struct netem_sched_data *q = qdisc_priv(sch);
struct sk_buff *skb2;
int ret;
- pr_debug("netem_enqueue skb=%p @%lu\n", skb, jiffies);
+ pr_debug("netem_enqueue skb=%p\n", skb);
/* Random packet drop 0 => none, ~0 => all */
if (q->loss && q->loss >= get_crandom(&q->loss_cor)) {
@@ -184,7 +224,7 @@ static int netem_enqueue(struct sk_buff
&& (skb2 = skb_clone(skb, GFP_ATOMIC)) != NULL) {
pr_debug("netem_enqueue: dup %p\n", skb2);
- if (delay_skb(sch, skb2)) {
+ if (netem_delay(sch, skb2)) {
sch->q.qlen++;
sch->bstats.bytes += skb2->len;
sch->bstats.packets++;
@@ -202,7 +242,8 @@ static int netem_enqueue(struct sk_buff
ret = q->qdisc->enqueue(skb, q->qdisc);
} else {
q->counter = 0;
- ret = delay_skb(sch, skb);
+ ret = netem_delay(sch, skb);
+ netem_run(sch);
}
if (likely(ret == NET_XMIT_SUCCESS)) {
@@ -241,56 +282,35 @@ static unsigned int netem_drop(struct Qd
return len;
}
-/* Dequeue packet.
- * Move all packets that are ready to send from the delay holding
- * list to the underlying qdisc, then just call dequeue
- */
static struct sk_buff *netem_dequeue(struct Qdisc *sch)
{
struct netem_sched_data *q = qdisc_priv(sch);
struct sk_buff *skb;
+ int pending;
+
+ pending = netem_run(sch);
skb = q->qdisc->dequeue(q->qdisc);
- if (skb)
+ if (skb) {
+ pr_debug("netem_dequeue: return skb=%p\n", skb);
sch->q.qlen--;
+ sch->flags &= ~TCQ_F_THROTTLED;
+ }
+ else if (pending) {
+ pr_debug("netem_dequeue: throttling\n");
+ sch->flags |= TCQ_F_THROTTLED;
+ }
+
return skb;
}
static void netem_watchdog(unsigned long arg)
{
struct Qdisc *sch = (struct Qdisc *)arg;
- struct netem_sched_data *q = qdisc_priv(sch);
- struct net_device *dev = sch->dev;
- struct sk_buff *skb;
- psched_time_t now;
- pr_debug("netem_watchdog: fired @%lu\n", jiffies);
-
- spin_lock_bh(&dev->queue_lock);
- PSCHED_GET_TIME(now);
-
- while ((skb = skb_peek(&q->delayed)) != NULL) {
- const struct netem_skb_cb *cb
- = (const struct netem_skb_cb *)skb->cb;
- long delay
- = PSCHED_US2JIFFIE(PSCHED_TDIFF(cb->time_to_send, now));
- pr_debug("netem_watchdog: skb %p@%lu %ld\n",
- skb, jiffies, delay);
-
- /* if more time remaining? */
- if (delay > 0) {
- mod_timer(&q->timer, jiffies + delay);
- break;
- }
- __skb_unlink(skb, &q->delayed);
-
- if (q->qdisc->enqueue(skb, q->qdisc)) {
- sch->q.qlen--;
- sch->qstats.drops++;
- }
- }
- qdisc_run(dev);
- spin_unlock_bh(&dev->queue_lock);
+ pr_debug("netem_watchdog qlen=%d\n", sch->q.qlen);
+ sch->flags &= ~TCQ_F_THROTTLED;
+ netif_schedule(sch->dev);
}
static void netem_reset(struct Qdisc *sch)
@@ -301,6 +321,7 @@ static void netem_reset(struct Qdisc *sc
skb_queue_purge(&q->delayed);
sch->q.qlen = 0;
+ sch->flags &= ~TCQ_F_THROTTLED;
del_timer_sync(&q->timer);
}