X-Git-Url: http://git.onelab.eu/?a=blobdiff_plain;f=drivers%2Finfiniband%2Fulp%2Fipoib%2Fipoib_multicast.c;h=1dae4b238252d3b204d68570bc4fa73f9e73959c;hb=43bc926fffd92024b46cafaf7350d669ba9ca884;hp=70208c3d21e28d415a8f9a82f886e62b64c4590d;hpb=cee37fe97739d85991964371c1f3a745c00dd236;p=linux-2.6.git diff --git a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c index 70208c3d2..1dae4b238 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c @@ -1,5 +1,7 @@ /* * Copyright (c) 2004, 2005 Topspin Communications. All rights reserved. + * Copyright (c) 2005 Sun Microsystems, Inc. All rights reserved. + * Copyright (c) 2004 Voltaire, Inc. All rights reserved. * * This software is available to you under a choice of one of two * licenses. You may choose to be licensed under the terms of the GNU @@ -41,6 +43,8 @@ #include #include +#include + #include "ipoib.h" #ifdef CONFIG_INFINIBAND_IPOIB_DEBUG @@ -51,7 +55,7 @@ MODULE_PARM_DESC(mcast_debug_level, "Enable multicast debug tracing if > 0"); #endif -static DECLARE_MUTEX(mcast_mutex); +static DEFINE_MUTEX(mcast_mutex); /* Used for all multicast joins (broadcast, IPv4 mcast and IPv6 mcast) */ struct ipoib_mcast { @@ -93,8 +97,7 @@ static void ipoib_mcast_free(struct ipoib_mcast *mcast) struct ipoib_dev_priv *priv = netdev_priv(dev); struct ipoib_neigh *neigh, *tmp; unsigned long flags; - LIST_HEAD(ah_list); - struct ipoib_ah *ah, *tah; + int tx_dropped = 0; ipoib_dbg_mcast(netdev_priv(dev), "deleting multicast group " IPOIB_GID_FMT "\n", @@ -103,28 +106,31 @@ static void ipoib_mcast_free(struct ipoib_mcast *mcast) spin_lock_irqsave(&priv->lock, flags); list_for_each_entry_safe(neigh, tmp, &mcast->neigh_list, list) { + /* + * It's safe to call ipoib_put_ah() inside priv->lock + * here, because we know that mcast->ah will always + * hold one more reference, so ipoib_put_ah() will + * never do more than decrement the ref count. + */ if (neigh->ah) - list_add_tail(&neigh->ah->list, &ah_list); - *to_ipoib_neigh(neigh->neighbour) = NULL; - neigh->neighbour->ops->destructor = NULL; - kfree(neigh); + ipoib_put_ah(neigh->ah); + ipoib_neigh_free(neigh); } spin_unlock_irqrestore(&priv->lock, flags); - list_for_each_entry_safe(ah, tah, &ah_list, list) - ipoib_put_ah(ah); - if (mcast->ah) ipoib_put_ah(mcast->ah); while (!skb_queue_empty(&mcast->pkt_queue)) { - struct sk_buff *skb = skb_dequeue(&mcast->pkt_queue); - - skb->dev = dev; - dev_kfree_skb_any(skb); + ++tx_dropped; + dev_kfree_skb_any(skb_dequeue(&mcast->pkt_queue)); } + spin_lock_irqsave(&priv->tx_lock, flags); + priv->stats.tx_dropped += tx_dropped; + spin_unlock_irqrestore(&priv->tx_lock, flags); + kfree(mcast); } @@ -133,26 +139,18 @@ static struct ipoib_mcast *ipoib_mcast_alloc(struct net_device *dev, { struct ipoib_mcast *mcast; - mcast = kmalloc(sizeof (*mcast), can_sleep ? GFP_KERNEL : GFP_ATOMIC); + mcast = kzalloc(sizeof *mcast, can_sleep ? GFP_KERNEL : GFP_ATOMIC); if (!mcast) return NULL; - memset(mcast, 0, sizeof (*mcast)); - - init_completion(&mcast->done); - mcast->dev = dev; mcast->created = jiffies; - mcast->backoff = HZ; - mcast->logcount = 0; + mcast->backoff = 1; INIT_LIST_HEAD(&mcast->list); INIT_LIST_HEAD(&mcast->neigh_list); skb_queue_head_init(&mcast->pkt_queue); - mcast->ah = NULL; - mcast->query = NULL; - return mcast; } @@ -213,6 +211,7 @@ static int ipoib_mcast_join_finish(struct ipoib_mcast *mcast, { struct net_device *dev = mcast->dev; struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_ah *ah; int ret; mcast->mcmember = *mcmember; @@ -251,6 +250,7 @@ static int ipoib_mcast_join_finish(struct ipoib_mcast *mcast, .port_num = priv->port, .sl = mcast->mcmember.sl, .ah_flags = IB_AH_GRH, + .static_rate = mcast->mcmember.rate, .grh = { .flow_label = be32_to_cpu(mcast->mcmember.flow_label), .hop_limit = mcast->mcmember.hop_limit, @@ -258,19 +258,10 @@ static int ipoib_mcast_join_finish(struct ipoib_mcast *mcast, .traffic_class = mcast->mcmember.traffic_class } }; - int path_rate = ib_sa_rate_enum_to_int(mcast->mcmember.rate); - av.grh.dgid = mcast->mcmember.mgid; - if (path_rate > 0 && priv->local_rate > path_rate) - av.static_rate = (priv->local_rate - 1) / path_rate; - - ipoib_dbg_mcast(priv, "static_rate %d for local port %dX, mcmember %dX\n", - av.static_rate, priv->local_rate, - ib_sa_rate_enum_to_int(mcast->mcmember.rate)); - - mcast->ah = ipoib_create_ah(dev, priv->pd, &av); - if (!mcast->ah) { + ah = ipoib_create_ah(dev, priv->pd, &av); + if (!ah) { ipoib_warn(priv, "ib_address_create failed\n"); } else { ipoib_dbg_mcast(priv, "MGID " IPOIB_GID_FMT @@ -280,11 +271,17 @@ static int ipoib_mcast_join_finish(struct ipoib_mcast *mcast, be16_to_cpu(mcast->mcmember.mlid), mcast->mcmember.sl); } + + spin_lock_irq(&priv->lock); + mcast->ah = ah; + spin_unlock_irq(&priv->lock); } /* actually send any queued packets */ + spin_lock_irq(&priv->tx_lock); while (!skb_queue_empty(&mcast->pkt_queue)) { struct sk_buff *skb = skb_dequeue(&mcast->pkt_queue); + spin_unlock_irq(&priv->tx_lock); skb->dev = dev; @@ -295,7 +292,9 @@ static int ipoib_mcast_join_finish(struct ipoib_mcast *mcast, if (dev_queue_xmit(skb)) ipoib_warn(priv, "dev_queue_xmit failed to requeue packet\n"); + spin_lock_irq(&priv->tx_lock); } + spin_unlock_irq(&priv->tx_lock); return 0; } @@ -307,6 +306,7 @@ ipoib_mcast_sendonly_join_complete(int status, { struct ipoib_mcast *mcast = mcast_ptr; struct net_device *dev = mcast->dev; + struct ipoib_dev_priv *priv = netdev_priv(dev); if (!status) ipoib_mcast_join_finish(mcast, mcmember); @@ -317,13 +317,12 @@ ipoib_mcast_sendonly_join_complete(int status, IPOIB_GID_ARG(mcast->mcmember.mgid), status); /* Flush out any queued packets */ + spin_lock_irq(&priv->tx_lock); while (!skb_queue_empty(&mcast->pkt_queue)) { - struct sk_buff *skb = skb_dequeue(&mcast->pkt_queue); - - skb->dev = dev; - - dev_kfree_skb_any(skb); + ++priv->stats.tx_dropped; + dev_kfree_skb_any(skb_dequeue(&mcast->pkt_queue)); } + spin_unlock_irq(&priv->tx_lock); /* Clear the busy flag so we try again */ clear_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags); @@ -357,7 +356,9 @@ static int ipoib_mcast_sendonly_join(struct ipoib_mcast *mcast) rec.mgid = mcast->mcmember.mgid; rec.port_gid = priv->local_gid; - rec.pkey = be16_to_cpu(priv->pkey); + rec.pkey = cpu_to_be16(priv->pkey); + + init_completion(&mcast->done); ret = ib_sa_mcmember_rec_set(priv->ca, priv->port, &rec, IB_SA_MCMEMBER_REC_MGID | @@ -394,11 +395,11 @@ static void ipoib_mcast_join_complete(int status, IPOIB_GID_ARG(mcast->mcmember.mgid), status); if (!status && !ipoib_mcast_join_finish(mcast, mcmember)) { - mcast->backoff = HZ; - down(&mcast_mutex); + mcast->backoff = 1; + mutex_lock(&mcast_mutex); if (test_bit(IPOIB_MCAST_RUN, &priv->flags)) queue_work(ipoib_workqueue, &priv->mcast_task); - up(&mcast_mutex); + mutex_unlock(&mcast_mutex); complete(&mcast->done); return; } @@ -426,9 +427,11 @@ static void ipoib_mcast_join_complete(int status, if (mcast->backoff > IPOIB_MAX_BACKOFF_SECONDS) mcast->backoff = IPOIB_MAX_BACKOFF_SECONDS; + mutex_lock(&mcast_mutex); + + spin_lock_irq(&priv->lock); mcast->query = NULL; - down(&mcast_mutex); if (test_bit(IPOIB_MCAST_RUN, &priv->flags)) { if (status == -ETIMEDOUT) queue_work(ipoib_workqueue, &priv->mcast_task); @@ -437,7 +440,8 @@ static void ipoib_mcast_join_complete(int status, mcast->backoff * HZ); } else complete(&mcast->done); - up(&mcast_mutex); + spin_unlock_irq(&priv->lock); + mutex_unlock(&mcast_mutex); return; } @@ -457,7 +461,7 @@ static void ipoib_mcast_join(struct net_device *dev, struct ipoib_mcast *mcast, rec.mgid = mcast->mcmember.mgid; rec.port_gid = priv->local_gid; - rec.pkey = be16_to_cpu(priv->pkey); + rec.pkey = cpu_to_be16(priv->pkey); comp_mask = IB_SA_MCMEMBER_REC_MGID | @@ -478,6 +482,8 @@ static void ipoib_mcast_join(struct net_device *dev, struct ipoib_mcast *mcast, rec.traffic_class = priv->broadcast->mcmember.traffic_class; } + init_completion(&mcast->done); + ret = ib_sa_mcmember_rec_set(priv->ca, priv->port, &rec, comp_mask, mcast->backoff * 1000, GFP_ATOMIC, ipoib_mcast_join_complete, @@ -490,12 +496,12 @@ static void ipoib_mcast_join(struct net_device *dev, struct ipoib_mcast *mcast, if (mcast->backoff > IPOIB_MAX_BACKOFF_SECONDS) mcast->backoff = IPOIB_MAX_BACKOFF_SECONDS; - down(&mcast_mutex); + mutex_lock(&mcast_mutex); if (test_bit(IPOIB_MCAST_RUN, &priv->flags)) queue_delayed_work(ipoib_workqueue, &priv->mcast_task, - mcast->backoff); - up(&mcast_mutex); + mcast->backoff * HZ); + mutex_unlock(&mcast_mutex); } else mcast->query_id = ret; } @@ -525,21 +531,24 @@ void ipoib_mcast_join_task(void *dev_ptr) } if (!priv->broadcast) { - priv->broadcast = ipoib_mcast_alloc(dev, 1); - if (!priv->broadcast) { + struct ipoib_mcast *broadcast; + + broadcast = ipoib_mcast_alloc(dev, 1); + if (!broadcast) { ipoib_warn(priv, "failed to allocate broadcast group\n"); - down(&mcast_mutex); + mutex_lock(&mcast_mutex); if (test_bit(IPOIB_MCAST_RUN, &priv->flags)) queue_delayed_work(ipoib_workqueue, &priv->mcast_task, HZ); - up(&mcast_mutex); + mutex_unlock(&mcast_mutex); return; } - memcpy(priv->broadcast->mcmember.mgid.raw, priv->dev->broadcast + 4, + spin_lock_irq(&priv->lock); + memcpy(broadcast->mcmember.mgid.raw, priv->dev->broadcast + 4, sizeof (union ib_gid)); + priv->broadcast = broadcast; - spin_lock_irq(&priv->lock); __ipoib_mcast_add(dev, priv->broadcast); spin_unlock_irq(&priv->lock); } @@ -588,44 +597,57 @@ int ipoib_mcast_start_thread(struct net_device *dev) ipoib_dbg_mcast(priv, "starting multicast thread\n"); - down(&mcast_mutex); + mutex_lock(&mcast_mutex); if (!test_and_set_bit(IPOIB_MCAST_RUN, &priv->flags)) queue_work(ipoib_workqueue, &priv->mcast_task); - up(&mcast_mutex); + mutex_unlock(&mcast_mutex); + + spin_lock_irq(&priv->lock); + set_bit(IPOIB_MCAST_STARTED, &priv->flags); + spin_unlock_irq(&priv->lock); return 0; } -int ipoib_mcast_stop_thread(struct net_device *dev) +static void wait_for_mcast_join(struct ipoib_dev_priv *priv, + struct ipoib_mcast *mcast) +{ + spin_lock_irq(&priv->lock); + if (mcast && mcast->query) { + ib_sa_cancel_query(mcast->query_id, mcast->query); + mcast->query = NULL; + spin_unlock_irq(&priv->lock); + ipoib_dbg_mcast(priv, "waiting for MGID " IPOIB_GID_FMT "\n", + IPOIB_GID_ARG(mcast->mcmember.mgid)); + wait_for_completion(&mcast->done); + } + else + spin_unlock_irq(&priv->lock); +} + +int ipoib_mcast_stop_thread(struct net_device *dev, int flush) { struct ipoib_dev_priv *priv = netdev_priv(dev); struct ipoib_mcast *mcast; ipoib_dbg_mcast(priv, "stopping multicast thread\n"); - down(&mcast_mutex); + spin_lock_irq(&priv->lock); + clear_bit(IPOIB_MCAST_STARTED, &priv->flags); + spin_unlock_irq(&priv->lock); + + mutex_lock(&mcast_mutex); clear_bit(IPOIB_MCAST_RUN, &priv->flags); cancel_delayed_work(&priv->mcast_task); - up(&mcast_mutex); + mutex_unlock(&mcast_mutex); - flush_workqueue(ipoib_workqueue); + if (flush) + flush_workqueue(ipoib_workqueue); - if (priv->broadcast && priv->broadcast->query) { - ib_sa_cancel_query(priv->broadcast->query_id, priv->broadcast->query); - priv->broadcast->query = NULL; - ipoib_dbg_mcast(priv, "waiting for bcast\n"); - wait_for_completion(&priv->broadcast->done); - } + wait_for_mcast_join(priv, priv->broadcast); - list_for_each_entry(mcast, &priv->multicast_list, list) { - if (mcast->query) { - ib_sa_cancel_query(mcast->query_id, mcast->query); - mcast->query = NULL; - ipoib_dbg_mcast(priv, "waiting for MGID " IPOIB_GID_FMT "\n", - IPOIB_GID_ARG(mcast->mcmember.mgid)); - wait_for_completion(&mcast->done); - } - } + list_for_each_entry(mcast, &priv->multicast_list, list) + wait_for_mcast_join(priv, mcast); return 0; } @@ -646,7 +668,7 @@ static int ipoib_mcast_leave(struct net_device *dev, struct ipoib_mcast *mcast) rec.mgid = mcast->mcmember.mgid; rec.port_gid = priv->local_gid; - rec.pkey = be16_to_cpu(priv->pkey); + rec.pkey = cpu_to_be16(priv->pkey); /* Remove ourselves from the multicast group */ ret = ipoib_mcast_detach(dev, be16_to_cpu(mcast->mcmember.mlid), @@ -684,6 +706,14 @@ void ipoib_mcast_send(struct net_device *dev, union ib_gid *mgid, */ spin_lock(&priv->lock); + if (!test_bit(IPOIB_MCAST_STARTED, &priv->flags) || + !priv->broadcast || + !test_bit(IPOIB_MCAST_FLAG_ATTACHED, &priv->broadcast->flags)) { + ++priv->stats.tx_dropped; + dev_kfree_skb_any(skb); + goto unlock; + } + mcast = __ipoib_mcast_find(dev, mgid); if (!mcast) { /* Let's create a new send only group now */ @@ -694,6 +724,7 @@ void ipoib_mcast_send(struct net_device *dev, union ib_gid *mgid, if (!mcast) { ipoib_warn(priv, "unable to allocate memory for " "multicast structure\n"); + ++priv->stats.tx_dropped; dev_kfree_skb_any(skb); goto out; } @@ -707,8 +738,10 @@ void ipoib_mcast_send(struct net_device *dev, union ib_gid *mgid, if (!mcast->ah) { if (skb_queue_len(&mcast->pkt_queue) < IPOIB_MAX_MCAST_QUEUE) skb_queue_tail(&mcast->pkt_queue, skb); - else + else { + ++priv->stats.tx_dropped; dev_kfree_skb_any(skb); + } if (mcast->query) ipoib_dbg_mcast(priv, "no address vector, " @@ -728,13 +761,11 @@ out: if (skb->dst && skb->dst->neighbour && !*to_ipoib_neigh(skb->dst->neighbour)) { - struct ipoib_neigh *neigh = kmalloc(sizeof *neigh, GFP_ATOMIC); + struct ipoib_neigh *neigh = ipoib_neigh_alloc(skb->dst->neighbour); if (neigh) { kref_get(&mcast->ah->ref); neigh->ah = mcast->ah; - neigh->neighbour = skb->dst->neighbour; - *to_ipoib_neigh(skb->dst->neighbour) = neigh; list_add_tail(&neigh->list, &mcast->neigh_list); } } @@ -742,6 +773,7 @@ out: ipoib_send(dev, skb, mcast->ah, IB_MULTICAST_QPN); } +unlock: spin_unlock(&priv->lock); } @@ -749,48 +781,23 @@ void ipoib_mcast_dev_flush(struct net_device *dev) { struct ipoib_dev_priv *priv = netdev_priv(dev); LIST_HEAD(remove_list); - struct ipoib_mcast *mcast, *tmcast, *nmcast; + struct ipoib_mcast *mcast, *tmcast; unsigned long flags; ipoib_dbg_mcast(priv, "flushing multicast list\n"); spin_lock_irqsave(&priv->lock, flags); - list_for_each_entry_safe(mcast, tmcast, &priv->multicast_list, list) { - nmcast = ipoib_mcast_alloc(dev, 0); - if (nmcast) { - nmcast->flags = - mcast->flags & (1 << IPOIB_MCAST_FLAG_SENDONLY); - - nmcast->mcmember.mgid = mcast->mcmember.mgid; - - /* Add the new group in before the to-be-destroyed group */ - list_add_tail(&nmcast->list, &mcast->list); - list_del_init(&mcast->list); - rb_replace_node(&mcast->rb_node, &nmcast->rb_node, - &priv->multicast_tree); - - list_add_tail(&mcast->list, &remove_list); - } else { - ipoib_warn(priv, "could not reallocate multicast group " - IPOIB_GID_FMT "\n", - IPOIB_GID_ARG(mcast->mcmember.mgid)); - } + list_for_each_entry_safe(mcast, tmcast, &priv->multicast_list, list) { + list_del(&mcast->list); + rb_erase(&mcast->rb_node, &priv->multicast_tree); + list_add_tail(&mcast->list, &remove_list); } if (priv->broadcast) { - nmcast = ipoib_mcast_alloc(dev, 0); - if (nmcast) { - nmcast->mcmember.mgid = priv->broadcast->mcmember.mgid; - - rb_replace_node(&priv->broadcast->rb_node, - &nmcast->rb_node, - &priv->multicast_tree); - - list_add_tail(&priv->broadcast->list, &remove_list); - } - - priv->broadcast = nmcast; + rb_erase(&priv->broadcast->rb_node, &priv->multicast_tree); + list_add_tail(&priv->broadcast->list, &remove_list); + priv->broadcast = NULL; } spin_unlock_irqrestore(&priv->lock, flags); @@ -801,24 +808,6 @@ void ipoib_mcast_dev_flush(struct net_device *dev) } } -void ipoib_mcast_dev_down(struct net_device *dev) -{ - struct ipoib_dev_priv *priv = netdev_priv(dev); - unsigned long flags; - - /* Delete broadcast since it will be recreated */ - if (priv->broadcast) { - ipoib_dbg_mcast(priv, "deleting broadcast group\n"); - - spin_lock_irqsave(&priv->lock, flags); - rb_erase(&priv->broadcast->rb_node, &priv->multicast_tree); - spin_unlock_irqrestore(&priv->lock, flags); - ipoib_mcast_leave(dev, priv->broadcast); - ipoib_mcast_free(priv->broadcast); - priv->broadcast = NULL; - } -} - void ipoib_mcast_restart_task(void *dev_ptr) { struct net_device *dev = dev_ptr; @@ -830,9 +819,10 @@ void ipoib_mcast_restart_task(void *dev_ptr) ipoib_dbg_mcast(priv, "restarting multicast task\n"); - ipoib_mcast_stop_thread(dev); + ipoib_mcast_stop_thread(dev, 0); - spin_lock_irqsave(&priv->lock, flags); + spin_lock_irqsave(&dev->xmit_lock, flags); + spin_lock(&priv->lock); /* * Unfortunately, the networking core only gives us a list of all of @@ -904,10 +894,13 @@ void ipoib_mcast_restart_task(void *dev_ptr) list_add_tail(&mcast->list, &remove_list); } } - spin_unlock_irqrestore(&priv->lock, flags); + + spin_unlock(&priv->lock); + spin_unlock_irqrestore(&dev->xmit_lock, flags); /* We have to cancel outside of the spinlock */ list_for_each_entry_safe(mcast, tmcast, &remove_list, list) { + wait_for_mcast_join(priv, mcast); ipoib_mcast_leave(mcast->dev, mcast); ipoib_mcast_free(mcast); } @@ -916,6 +909,8 @@ void ipoib_mcast_restart_task(void *dev_ptr) ipoib_mcast_start_thread(dev); } +#ifdef CONFIG_INFINIBAND_IPOIB_DEBUG + struct ipoib_mcast_iter *ipoib_mcast_iter_init(struct net_device *dev) { struct ipoib_mcast_iter *iter; @@ -925,21 +920,16 @@ struct ipoib_mcast_iter *ipoib_mcast_iter_init(struct net_device *dev) return NULL; iter->dev = dev; - memset(iter->mgid.raw, 0, sizeof iter->mgid); + memset(iter->mgid.raw, 0, 16); if (ipoib_mcast_iter_next(iter)) { - ipoib_mcast_iter_free(iter); + kfree(iter); return NULL; } return iter; } -void ipoib_mcast_iter_free(struct ipoib_mcast_iter *iter) -{ - kfree(iter); -} - int ipoib_mcast_iter_next(struct ipoib_mcast_iter *iter) { struct ipoib_dev_priv *priv = netdev_priv(iter->dev); @@ -988,3 +978,5 @@ void ipoib_mcast_iter_read(struct ipoib_mcast_iter *iter, *complete = iter->complete; *send_only = iter->send_only; } + +#endif /* CONFIG_INFINIBAND_IPOIB_DEBUG */