[net,v5,3/3] net: sched: fix tx action reschedule issue with stopped queue

Message ID 1620266264-48109-4-git-send-email-linyunsheng@huawei.com
State Superseded
Headers show
Series
  • fix packet stuck problem for lockless qdisc
Related show

Commit Message

Yunsheng Lin May 6, 2021, 1:57 a.m.
The netdev qeueue might be stopped when byte queue limit has
reached or tx hw ring is full, net_tx_action() may still be
rescheduled endlessly if STATE_MISSED is set, which consumes
a lot of cpu without dequeuing and transmiting any skb because
the netdev queue is stopped, see qdisc_run_end().

This patch fixes it by checking the netdev queue state before
calling qdisc_run() and clearing STATE_MISSED if netdev queue is
stopped during qdisc_run(), the net_tx_action() is recheduled
again when netdev qeueue is restarted, see netif_tx_wake_queue().

As q->enqueue() may return NET_XMIT_DROP when there is no enough
space, running qdisc_run() will likely consume unnecessary cpu, so
avoid calling qdisc_run() when q->enqueue() returns NET_XMIT_DROP
too.

Fixes: 6b3ba9146fe6 ("net: sched: allow qdiscs to handle locking")
Reported-by: Michal Kubecek <mkubecek@suse.cz>
Signed-off-by: Yunsheng Lin <linyunsheng@huawei.com>
---
 net/core/dev.c          | 4 +++-
 net/sched/sch_generic.c | 8 +++++++-
 2 files changed, 10 insertions(+), 2 deletions(-)

Patch

diff --git a/net/core/dev.c b/net/core/dev.c
index d596cd7..005bc3e 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -3853,7 +3853,9 @@  static inline int __dev_xmit_skb(struct sk_buff *skb, struct Qdisc *q,
 
 	if (q->flags & TCQ_F_NOLOCK) {
 		rc = q->enqueue(skb, q, &to_free) & NET_XMIT_MASK;
-		qdisc_run(q);
+		if (likely(rc != NET_XMIT_DROP &&
+			   !netif_xmit_frozen_or_stopped(txq)))
+			qdisc_run(q);
 
 		if (unlikely(to_free))
 			kfree_skb_list(to_free);
diff --git a/net/sched/sch_generic.c b/net/sched/sch_generic.c
index c32ac5b..2bb829ea 100644
--- a/net/sched/sch_generic.c
+++ b/net/sched/sch_generic.c
@@ -74,6 +74,7 @@  static inline struct sk_buff *__skb_dequeue_bad_txq(struct Qdisc *q)
 			}
 		} else {
 			skb = SKB_XOFF_MAGIC;
+			clear_bit(__QDISC_STATE_MISSED, &q->state);
 		}
 	}
 
@@ -242,6 +243,7 @@  static struct sk_buff *dequeue_skb(struct Qdisc *q, bool *validate,
 			}
 		} else {
 			skb = NULL;
+			clear_bit(__QDISC_STATE_MISSED, &q->state);
 		}
 		if (lock)
 			spin_unlock(lock);
@@ -251,8 +253,10 @@  static struct sk_buff *dequeue_skb(struct Qdisc *q, bool *validate,
 	*validate = true;
 
 	if ((q->flags & TCQ_F_ONETXQUEUE) &&
-	    netif_xmit_frozen_or_stopped(txq))
+	    netif_xmit_frozen_or_stopped(txq)) {
+		clear_bit(__QDISC_STATE_MISSED, &q->state);
 		return skb;
+	}
 
 	skb = qdisc_dequeue_skb_bad_txq(q);
 	if (unlikely(skb)) {
@@ -311,6 +315,8 @@  bool sch_direct_xmit(struct sk_buff *skb, struct Qdisc *q,
 		HARD_TX_LOCK(dev, txq, smp_processor_id());
 		if (!netif_xmit_frozen_or_stopped(txq))
 			skb = dev_hard_start_xmit(skb, dev, txq, &ret);
+		else
+			clear_bit(__QDISC_STATE_MISSED, &q->state);
 
 		HARD_TX_UNLOCK(dev, txq);
 	} else {