X-Git-Url: https://err.no/cgi-bin/gitweb.cgi?a=blobdiff_plain;ds=sidebyside;f=net%2Fsched%2Fsch_generic.c;h=deafeb9cb70ef24f1579198e69e4c3f29db58fad;hb=9f9afec48221fe4a19f84a9341f5b304bf7d7783;hp=d7aca8ef524ab8f9ea9d01b986d0ea7a88c72c3a;hpb=af57d238aa2107e1b45d8dacad6e50db938f0567;p=linux-2.6 diff --git a/net/sched/sch_generic.c b/net/sched/sch_generic.c index d7aca8ef52..deafeb9cb7 100644 --- a/net/sched/sch_generic.c +++ b/net/sched/sch_generic.c @@ -14,7 +14,6 @@ #include #include #include -#include #include #include #include @@ -46,11 +45,10 @@ The idea is the following: - enqueue, dequeue are serialized via top level device spinlock dev->queue_lock. - - tree walking is protected by read_lock_bh(qdisc_tree_lock) + - tree walking is protected by read_lock(qdisc_tree_lock) and this lock is used only in process context. - - updates to tree are made under rtnl semaphore or - from softirq context (__qdisc_destroy rcu-callback) - hence this lock needs local bh disabling. + - updates to tree are made only under rtnl semaphore, + hence this lock may be made without local bh disabling. qdisc_tree_lock must be grabbed BEFORE dev->queue_lock! */ @@ -58,14 +56,14 @@ DEFINE_RWLOCK(qdisc_tree_lock); void qdisc_lock_tree(struct net_device *dev) { - write_lock_bh(&qdisc_tree_lock); + write_lock(&qdisc_tree_lock); spin_lock_bh(&dev->queue_lock); } void qdisc_unlock_tree(struct net_device *dev) { spin_unlock_bh(&dev->queue_lock); - write_unlock_bh(&qdisc_tree_lock); + write_unlock(&qdisc_tree_lock); } /* @@ -96,8 +94,11 @@ static inline int qdisc_restart(struct net_device *dev) struct sk_buff *skb; /* Dequeue packet */ - if ((skb = q->dequeue(q)) != NULL) { + if (((skb = dev->gso_skb)) || ((skb = q->dequeue(q)))) { unsigned nolock = (dev->features & NETIF_F_LLTX); + + dev->gso_skb = NULL; + /* * When the driver has LLTX set it does its own locking * in start_xmit. No need to add additional overhead by @@ -134,10 +135,8 @@ static inline int qdisc_restart(struct net_device *dev) if (!netif_queue_stopped(dev)) { int ret; - if (netdev_nit) - dev_queue_xmit_nit(skb, dev); - ret = dev->hard_start_xmit(skb, dev); + ret = dev_hard_start_xmit(skb, dev); if (ret == NETDEV_TX_OK) { if (!nolock) { netif_tx_unlock(dev); @@ -171,7 +170,10 @@ static inline int qdisc_restart(struct net_device *dev) */ requeue: - q->ops->requeue(skb, q); + if (skb->next) + dev->gso_skb = skb; + else + q->ops->requeue(skb, q); netif_schedule(dev); return 1; } @@ -181,9 +183,13 @@ requeue: void __qdisc_run(struct net_device *dev) { + if (unlikely(dev->qdisc == &noop_qdisc)) + goto out; + while (qdisc_restart(dev) < 0 && !netif_queue_stopped(dev)) /* NOTHING */; +out: clear_bit(__LINK_STATE_QDISC_RUNNING, &dev->state); } @@ -231,9 +237,7 @@ void __netdev_watchdog_up(struct net_device *dev) static void dev_watchdog_up(struct net_device *dev) { - netif_tx_lock_bh(dev); __netdev_watchdog_up(dev); - netif_tx_unlock_bh(dev); } static void dev_watchdog_down(struct net_device *dev) @@ -425,10 +429,9 @@ struct Qdisc *qdisc_alloc(struct net_device *dev, struct Qdisc_ops *ops) size = QDISC_ALIGN(sizeof(*sch)); size += ops->priv_size + (QDISC_ALIGNTO - 1); - p = kmalloc(size, GFP_KERNEL); + p = kzalloc(size, GFP_KERNEL); if (!p) goto errout; - memset(p, 0, size); sch = (struct Qdisc *) QDISC_ALIGN((unsigned long) p); sch->padded = (char *) sch - (char *) p; @@ -447,13 +450,15 @@ errout: return ERR_PTR(-err); } -struct Qdisc * qdisc_create_dflt(struct net_device *dev, struct Qdisc_ops *ops) +struct Qdisc * qdisc_create_dflt(struct net_device *dev, struct Qdisc_ops *ops, + unsigned int parentid) { struct Qdisc *sch; sch = qdisc_alloc(dev, ops); if (IS_ERR(sch)) goto errout; + sch->parent = parentid; if (!ops->init || ops->init(sch, NULL) == 0) return sch; @@ -479,20 +484,6 @@ void qdisc_reset(struct Qdisc *qdisc) static void __qdisc_destroy(struct rcu_head *head) { struct Qdisc *qdisc = container_of(head, struct Qdisc, q_rcu); - struct Qdisc_ops *ops = qdisc->ops; - -#ifdef CONFIG_NET_ESTIMATOR - gen_kill_estimator(&qdisc->bstats, &qdisc->rate_est); -#endif - write_lock(&qdisc_tree_lock); - if (ops->reset) - ops->reset(qdisc); - if (ops->destroy) - ops->destroy(qdisc); - write_unlock(&qdisc_tree_lock); - module_put(ops->owner); - - dev_put(qdisc->dev); kfree((char *) qdisc - qdisc->padded); } @@ -500,32 +491,23 @@ static void __qdisc_destroy(struct rcu_head *head) void qdisc_destroy(struct Qdisc *qdisc) { - struct list_head cql = LIST_HEAD_INIT(cql); - struct Qdisc *cq, *q, *n; + struct Qdisc_ops *ops = qdisc->ops; if (qdisc->flags & TCQ_F_BUILTIN || - !atomic_dec_and_test(&qdisc->refcnt)) + !atomic_dec_and_test(&qdisc->refcnt)) return; - if (!list_empty(&qdisc->list)) { - if (qdisc->ops->cl_ops == NULL) - list_del(&qdisc->list); - else - list_move(&qdisc->list, &cql); - } - - /* unlink inner qdiscs from dev->qdisc_list immediately */ - list_for_each_entry(cq, &cql, list) - list_for_each_entry_safe(q, n, &qdisc->dev->qdisc_list, list) - if (TC_H_MAJ(q->parent) == TC_H_MAJ(cq->handle)) { - if (q->ops->cl_ops == NULL) - list_del_init(&q->list); - else - list_move_tail(&q->list, &cql); - } - list_for_each_entry_safe(cq, n, &cql, list) - list_del_init(&cq->list); + list_del(&qdisc->list); +#ifdef CONFIG_NET_ESTIMATOR + gen_kill_estimator(&qdisc->bstats, &qdisc->rate_est); +#endif + if (ops->reset) + ops->reset(qdisc); + if (ops->destroy) + ops->destroy(qdisc); + module_put(ops->owner); + dev_put(qdisc->dev); call_rcu(&qdisc->q_rcu, __qdisc_destroy); } @@ -540,20 +522,21 @@ void dev_activate(struct net_device *dev) if (dev->qdisc_sleeping == &noop_qdisc) { struct Qdisc *qdisc; if (dev->tx_queue_len) { - qdisc = qdisc_create_dflt(dev, &pfifo_fast_ops); + qdisc = qdisc_create_dflt(dev, &pfifo_fast_ops, + TC_H_ROOT); if (qdisc == NULL) { printk(KERN_INFO "%s: activation failed\n", dev->name); return; } - write_lock_bh(&qdisc_tree_lock); + write_lock(&qdisc_tree_lock); list_add_tail(&qdisc->list, &dev->qdisc_list); - write_unlock_bh(&qdisc_tree_lock); + write_unlock(&qdisc_tree_lock); } else { qdisc = &noqueue_qdisc; } - write_lock_bh(&qdisc_tree_lock); + write_lock(&qdisc_tree_lock); dev->qdisc_sleeping = qdisc; - write_unlock_bh(&qdisc_tree_lock); + write_unlock(&qdisc_tree_lock); } if (!netif_carrier_ok(dev)) @@ -583,10 +566,17 @@ void dev_deactivate(struct net_device *dev) dev_watchdog_down(dev); - while (test_bit(__LINK_STATE_SCHED, &dev->state)) + /* Wait for outstanding dev_queue_xmit calls. */ + synchronize_rcu(); + + /* Wait for outstanding qdisc_run calls. */ + while (test_bit(__LINK_STATE_QDISC_RUNNING, &dev->state)) yield(); - spin_unlock_wait(&dev->_xmit_lock); + if (dev->gso_skb) { + kfree_skb(dev->gso_skb); + dev->gso_skb = NULL; + } } void dev_init_scheduler(struct net_device *dev)