From c3f26a269c2421f97f10cf8ed05d5099b573af4d Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Thu, 31 Jul 2008 16:58:50 -0700 Subject: netdev: Fix lockdep warnings in multiqueue configurations. When support for multiple TX queues were added, the netif_tx_lock() routines we converted to iterate over all TX queues and grab each queue's spinlock. This causes heartburn for lockdep and it's not a healthy thing to do with lots of TX queues anyways. So modify this to use a top-level lock and a "frozen" state for the individual TX queues. Signed-off-by: David S. Miller --- include/linux/netdevice.h | 86 ++++++++++++++++++++++++++++++----------------- 1 file changed, 55 insertions(+), 31 deletions(-) (limited to 'include/linux/netdevice.h') diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index b4d056ceab96..ee583f642a9f 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -440,6 +440,7 @@ static inline void napi_synchronize(const struct napi_struct *n) enum netdev_queue_state_t { __QUEUE_STATE_XOFF, + __QUEUE_STATE_FROZEN, }; struct netdev_queue { @@ -636,7 +637,7 @@ struct net_device unsigned int real_num_tx_queues; unsigned long tx_queue_len; /* Max frames per queue allowed */ - + spinlock_t tx_global_lock; /* * One part is mostly used on xmit path (device) */ @@ -1099,6 +1100,11 @@ static inline int netif_queue_stopped(const struct net_device *dev) return netif_tx_queue_stopped(netdev_get_tx_queue(dev, 0)); } +static inline int netif_tx_queue_frozen(const struct netdev_queue *dev_queue) +{ + return test_bit(__QUEUE_STATE_FROZEN, &dev_queue->state); +} + /** * netif_running - test if up * @dev: network device @@ -1475,6 +1481,26 @@ static inline void __netif_tx_lock_bh(struct netdev_queue *txq) txq->xmit_lock_owner = smp_processor_id(); } +static inline int __netif_tx_trylock(struct netdev_queue *txq) +{ + int ok = spin_trylock(&txq->_xmit_lock); + if (likely(ok)) + txq->xmit_lock_owner = smp_processor_id(); + return ok; +} + +static inline void __netif_tx_unlock(struct netdev_queue *txq) +{ + txq->xmit_lock_owner = -1; + spin_unlock(&txq->_xmit_lock); +} + +static inline void __netif_tx_unlock_bh(struct netdev_queue *txq) +{ + txq->xmit_lock_owner = -1; + spin_unlock_bh(&txq->_xmit_lock); +} + /** * netif_tx_lock - grab network device transmit lock * @dev: network device @@ -1484,12 +1510,23 @@ static inline void __netif_tx_lock_bh(struct netdev_queue *txq) */ static inline void netif_tx_lock(struct net_device *dev) { - int cpu = smp_processor_id(); unsigned int i; + int cpu; + spin_lock(&dev->tx_global_lock); + cpu = smp_processor_id(); for (i = 0; i < dev->num_tx_queues; i++) { struct netdev_queue *txq = netdev_get_tx_queue(dev, i); + + /* We are the only thread of execution doing a + * freeze, but we have to grab the _xmit_lock in + * order to synchronize with threads which are in + * the ->hard_start_xmit() handler and already + * checked the frozen bit. + */ __netif_tx_lock(txq, cpu); + set_bit(__QUEUE_STATE_FROZEN, &txq->state); + __netif_tx_unlock(txq); } } @@ -1499,40 +1536,22 @@ static inline void netif_tx_lock_bh(struct net_device *dev) netif_tx_lock(dev); } -static inline int __netif_tx_trylock(struct netdev_queue *txq) -{ - int ok = spin_trylock(&txq->_xmit_lock); - if (likely(ok)) - txq->xmit_lock_owner = smp_processor_id(); - return ok; -} - -static inline int netif_tx_trylock(struct net_device *dev) -{ - return __netif_tx_trylock(netdev_get_tx_queue(dev, 0)); -} - -static inline void __netif_tx_unlock(struct netdev_queue *txq) -{ - txq->xmit_lock_owner = -1; - spin_unlock(&txq->_xmit_lock); -} - -static inline void __netif_tx_unlock_bh(struct netdev_queue *txq) -{ - txq->xmit_lock_owner = -1; - spin_unlock_bh(&txq->_xmit_lock); -} - static inline void netif_tx_unlock(struct net_device *dev) { unsigned int i; for (i = 0; i < dev->num_tx_queues; i++) { struct netdev_queue *txq = netdev_get_tx_queue(dev, i); - __netif_tx_unlock(txq); - } + /* No need to grab the _xmit_lock here. If the + * queue is not stopped for another reason, we + * force a schedule. + */ + clear_bit(__QUEUE_STATE_FROZEN, &txq->state); + if (!test_bit(__QUEUE_STATE_XOFF, &txq->state)) + __netif_schedule(txq->qdisc); + } + spin_unlock(&dev->tx_global_lock); } static inline void netif_tx_unlock_bh(struct net_device *dev) @@ -1556,13 +1575,18 @@ static inline void netif_tx_unlock_bh(struct net_device *dev) static inline void netif_tx_disable(struct net_device *dev) { unsigned int i; + int cpu; - netif_tx_lock_bh(dev); + local_bh_disable(); + cpu = smp_processor_id(); for (i = 0; i < dev->num_tx_queues; i++) { struct netdev_queue *txq = netdev_get_tx_queue(dev, i); + + __netif_tx_lock(txq, cpu); netif_tx_stop_queue(txq); + __netif_tx_unlock(txq); } - netif_tx_unlock_bh(dev); + local_bh_enable(); } static inline void netif_addr_lock(struct net_device *dev) -- cgit v1.2.2 From 378a2f090f7a478704a372a4869b8a9ac206234e Mon Sep 17 00:00:00 2001 From: Jarek Poplawski Date: Mon, 4 Aug 2008 22:31:03 -0700 Subject: net_sched: Add qdisc __NET_XMIT_STOLEN flag Patrick McHardy noticed: "The other problem that affects all qdiscs supporting actions is TC_ACT_QUEUED/TC_ACT_STOLEN getting mapped to NET_XMIT_SUCCESS even though the packet is not queued, corrupting upper qdiscs' qlen counters." and later explained: "The reason why it translates it at all seems to be to not increase the drops counter. Within a single qdisc this could be avoided by other means easily, upper qdiscs would still increase the counter when we return anything besides NET_XMIT_SUCCESS though. This means we need a new NET_XMIT return value to indicate this to the upper qdiscs. So I'd suggest to introduce NET_XMIT_STOLEN, return that to upper qdiscs and translate it to NET_XMIT_SUCCESS in dev_queue_xmit, similar to NET_XMIT_BYPASS." David Miller noticed: "Maybe these NET_XMIT_* values being passed around should be a set of bits. They could be composed of base meanings, combined with specific attributes. So you could say "NET_XMIT_DROP | __NET_XMIT_NO_DROP_COUNT" The attributes get masked out by the top-level ->enqueue() caller, such that the base meanings are the only thing that make their way up into the stack. If it's only about communication within the qdisc tree, let's simply code it that way." This patch is trying to realize these ideas. Signed-off-by: Jarek Poplawski Signed-off-by: David S. Miller --- include/linux/netdevice.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include/linux/netdevice.h') diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index ee583f642a9f..abbf5d52ec86 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -64,6 +64,7 @@ struct wireless_dev; #define NET_XMIT_BYPASS 4 /* packet does not leave via dequeue; (TC use only - dev_queue_xmit returns this as NET_XMIT_SUCCESS) */ +#define NET_XMIT_MASK 0xFFFF /* qdisc flags in net/sch_generic.h */ /* Backlog congestion levels */ #define NET_RX_SUCCESS 0 /* keep 'em coming, baby */ -- cgit v1.2.2 From cc6533e98a7f3cb7fce9d740da49195c7aa523a4 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Mon, 4 Aug 2008 23:04:08 -0700 Subject: net: Kill plain NET_XMIT_BYPASS. dst_input() was doing something completely absurd, looping on skb->dst->input() if NET_XMIT_BYPASS was seen, but these functions never return such an error. And as a result plain ole' NET_XMIT_BYPASS has no more references and can be completely killed off. Signed-off-by: David S. Miller --- include/linux/netdevice.h | 3 --- 1 file changed, 3 deletions(-) (limited to 'include/linux/netdevice.h') diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index abbf5d52ec86..488c56e649b5 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -61,9 +61,6 @@ struct wireless_dev; #define NET_XMIT_DROP 1 /* skb dropped */ #define NET_XMIT_CN 2 /* congestion notification */ #define NET_XMIT_POLICED 3 /* skb is shot by police */ -#define NET_XMIT_BYPASS 4 /* packet does not leave via dequeue; - (TC use only - dev_queue_xmit - returns this as NET_XMIT_SUCCESS) */ #define NET_XMIT_MASK 0xFFFF /* qdisc flags in net/sch_generic.h */ /* Backlog congestion levels */ -- cgit v1.2.2