net_sched: transform qdisc running bit into a seqcount Instead of using a single bit (__QDISC___STATE_RUNNING) in sch->__state, use a seqcount. This adds lockdep support, but more importantly it will allow us to sample qdisc/class statistics without having to grab qdisc root lock. Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Cong Wang <xiyou.wangcong@gmail.com> Cc: Jamal Hadi Salim <jhs@mojatatu.com> Signed-off-by: David S. Miller <davem@davemloft.net>
diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c index 941ec99..681af31a 100644 --- a/drivers/net/bonding/bond_main.c +++ b/drivers/net/bonding/bond_main.c
@@ -4610,6 +4610,7 @@ static struct lock_class_key bonding_netdev_xmit_lock_key; static struct lock_class_key bonding_netdev_addr_lock_key; static struct lock_class_key bonding_tx_busylock_key; +static struct lock_class_key bonding_qdisc_running_key; static void bond_set_lockdep_class_one(struct net_device *dev, struct netdev_queue *txq, @@ -4625,6 +4626,7 @@ &bonding_netdev_addr_lock_key); netdev_for_each_tx_queue(dev, bond_set_lockdep_class_one, NULL); dev->qdisc_tx_busylock = &bonding_tx_busylock_key; + dev->qdisc_running_key = &bonding_qdisc_running_key; } /* Called from registration process */
diff --git a/drivers/net/ppp/ppp_generic.c b/drivers/net/ppp/ppp_generic.c index 8dedafa..aeabaa4 100644 --- a/drivers/net/ppp/ppp_generic.c +++ b/drivers/net/ppp/ppp_generic.c
@@ -1313,9 +1313,12 @@ } static struct lock_class_key ppp_tx_busylock; +static struct lock_class_key ppp_qdisc_running_key; + static int ppp_dev_init(struct net_device *dev) { dev->qdisc_tx_busylock = &ppp_tx_busylock; + dev->qdisc_running_key = &ppp_qdisc_running_key; return 0; }
diff --git a/drivers/net/team/team.c b/drivers/net/team/team.c index 2ace126..00eb389 100644 --- a/drivers/net/team/team.c +++ b/drivers/net/team/team.c
@@ -1577,6 +1577,7 @@ static struct lock_class_key team_netdev_xmit_lock_key; static struct lock_class_key team_netdev_addr_lock_key; static struct lock_class_key team_tx_busylock_key; +static struct lock_class_key team_qdisc_running_key; static void team_set_lockdep_class_one(struct net_device *dev, struct netdev_queue *txq, @@ -1590,6 +1591,7 @@ lockdep_set_class(&dev->addr_list_lock, &team_netdev_addr_lock_key); netdev_for_each_tx_queue(dev, team_set_lockdep_class_one, NULL); dev->qdisc_tx_busylock = &team_tx_busylock_key; + dev->qdisc_running_key = &team_qdisc_running_key; } static int team_init(struct net_device *dev)
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index fa6df26..59d7e06 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h
@@ -1862,6 +1862,7 @@ #endif struct phy_device *phydev; struct lock_class_key *qdisc_tx_busylock; + struct lock_class_key *qdisc_running_key; bool proto_down; }; #define to_net_dev(d) container_of(d, struct net_device, dev)
diff --git a/include/net/sch_generic.h b/include/net/sch_generic.h index a1fd76c..bff8d89 100644 --- a/include/net/sch_generic.h +++ b/include/net/sch_generic.h
@@ -29,13 +29,6 @@ __QDISC_STATE_THROTTLED, }; -/* - * following bits are only changed while qdisc lock is held - */ -enum qdisc___state_t { - __QDISC___STATE_RUNNING = 1, -}; - struct qdisc_size_table { struct rcu_head rcu; struct list_head list; @@ -93,7 +86,7 @@ unsigned long state; struct sk_buff_head q; struct gnet_stats_basic_packed bstats; - unsigned int __state; + seqcount_t running; struct gnet_stats_queue qstats; struct rcu_head rcu_head; int padded; @@ -104,20 +97,20 @@ static inline bool qdisc_is_running(const struct Qdisc *qdisc) { - return (qdisc->__state & __QDISC___STATE_RUNNING) ? true : false; + return (raw_read_seqcount(&qdisc->running) & 1) ? true : false; } static inline bool qdisc_run_begin(struct Qdisc *qdisc) { if (qdisc_is_running(qdisc)) return false; - qdisc->__state |= __QDISC___STATE_RUNNING; + write_seqcount_begin(&qdisc->running); return true; } static inline void qdisc_run_end(struct Qdisc *qdisc) { - qdisc->__state &= ~__QDISC___STATE_RUNNING; + write_seqcount_end(&qdisc->running); } static inline bool qdisc_may_bulk(const struct Qdisc *qdisc)
diff --git a/net/bluetooth/6lowpan.c b/net/bluetooth/6lowpan.c index 780089d..977a11e 100644 --- a/net/bluetooth/6lowpan.c +++ b/net/bluetooth/6lowpan.c
@@ -629,6 +629,7 @@ static struct lock_class_key bt_tx_busylock; static struct lock_class_key bt_netdev_xmit_lock_key; +static struct lock_class_key bt_qdisc_running_key; static void bt_set_lockdep_class_one(struct net_device *dev, struct netdev_queue *txq, @@ -641,6 +642,7 @@ { netdev_for_each_tx_queue(dev, bt_set_lockdep_class_one, NULL); dev->qdisc_tx_busylock = &bt_tx_busylock; + dev->qdisc_running_key = &bt_qdisc_running_key; return 0; }
diff --git a/net/core/dev.c b/net/core/dev.c index 896b686..e0bcc39 100644 --- a/net/core/dev.c +++ b/net/core/dev.c
@@ -3075,7 +3075,7 @@ /* * Heuristic to force contended enqueues to serialize on a * separate lock before trying to get qdisc main lock. - * This permits __QDISC___STATE_RUNNING owner to get the lock more + * This permits qdisc->running owner to get the lock more * often and dequeue packets faster. */ contended = qdisc_is_running(q);
diff --git a/net/ieee802154/6lowpan/core.c b/net/ieee802154/6lowpan/core.c index dd085db..14aa5ef 100644 --- a/net/ieee802154/6lowpan/core.c +++ b/net/ieee802154/6lowpan/core.c
@@ -60,6 +60,7 @@ static struct lock_class_key lowpan_tx_busylock; static struct lock_class_key lowpan_netdev_xmit_lock_key; +static struct lock_class_key lowpan_qdisc_running_key; static void lowpan_set_lockdep_class_one(struct net_device *ldev, struct netdev_queue *txq, @@ -73,6 +74,8 @@ { netdev_for_each_tx_queue(ldev, lowpan_set_lockdep_class_one, NULL); ldev->qdisc_tx_busylock = &lowpan_tx_busylock; + ldev->qdisc_running_key = &lowpan_qdisc_running_key; + return 0; }
diff --git a/net/l2tp/l2tp_eth.c b/net/l2tp/l2tp_eth.c index e253c26..c00d72d 100644 --- a/net/l2tp/l2tp_eth.c +++ b/net/l2tp/l2tp_eth.c
@@ -68,6 +68,8 @@ } static struct lock_class_key l2tp_eth_tx_busylock; +static struct lock_class_key l2tp_qdisc_running_key; + static int l2tp_eth_dev_init(struct net_device *dev) { struct l2tp_eth *priv = netdev_priv(dev); @@ -76,6 +78,8 @@ eth_hw_addr_random(dev); eth_broadcast_addr(dev->broadcast); dev->qdisc_tx_busylock = &l2tp_eth_tx_busylock; + dev->qdisc_running_key = &l2tp_qdisc_running_key; + return 0; }
diff --git a/net/sched/sch_generic.c b/net/sched/sch_generic.c index 269dd71..cebea73 100644 --- a/net/sched/sch_generic.c +++ b/net/sched/sch_generic.c
@@ -110,7 +110,7 @@ /* * Transmit possibly several skbs, and handle the return status as - * required. Holding the __QDISC___STATE_RUNNING bit guarantees that + * required. Owning running seqcount bit guarantees that * only one CPU can execute this function. * * Returns to the caller: @@ -137,10 +137,10 @@ HARD_TX_UNLOCK(dev, txq); } else { - spin_lock(root_lock); + spin_lock_nested(root_lock, SINGLE_DEPTH_NESTING); return qdisc_qlen(q); } - spin_lock(root_lock); + spin_lock_nested(root_lock, SINGLE_DEPTH_NESTING); if (dev_xmit_complete(ret)) { /* Driver sent out skb successfully or skb was consumed */ @@ -163,7 +163,7 @@ /* * NOTE: Called under qdisc_lock(q) with locally disabled BH. * - * __QDISC___STATE_RUNNING guarantees only one CPU can process + * running seqcount guarantees only one CPU can process * this qdisc at a time. qdisc_lock(q) serializes queue accesses for * this queue. * @@ -379,6 +379,7 @@ .list = LIST_HEAD_INIT(noop_qdisc.list), .q.lock = __SPIN_LOCK_UNLOCKED(noop_qdisc.q.lock), .dev_queue = &noop_netdev_queue, + .running = SEQCNT_ZERO(noop_qdisc.running), .busylock = __SPIN_LOCK_UNLOCKED(noop_qdisc.busylock), }; EXPORT_SYMBOL(noop_qdisc); @@ -537,6 +538,7 @@ EXPORT_SYMBOL(pfifo_fast_ops); static struct lock_class_key qdisc_tx_busylock; +static struct lock_class_key qdisc_running_key; struct Qdisc *qdisc_alloc(struct netdev_queue *dev_queue, const struct Qdisc_ops *ops) @@ -570,6 +572,10 @@ lockdep_set_class(&sch->busylock, dev->qdisc_tx_busylock ?: &qdisc_tx_busylock); + seqcount_init(&sch->running); + lockdep_set_class(&sch->running, + dev->qdisc_running_key ?: &qdisc_running_key); + sch->ops = ops; sch->enqueue = ops->enqueue; sch->dequeue = ops->dequeue;