Messages in this thread Patch in this message | | | Date | Tue, 12 Dec 2006 11:16:56 +0100 | From | Ingo Molnar <> | Subject | [patch] netpoll: fix netpoll lockups |
| |
Subject: [patch] netpoll: fix netpoll lockups From: Ingo Molnar <mingo@elte.hu>
current -git doesnt boot on my laptop due to the following netpoll breakages:
- unlock the tx lock in the else branch too ... - use irq-safe locking instead of bh-safe locking, netpoll is often called from irq context.
with this patch -git boots fine with lockdep enabled and there are no locking complaints and everything works fine. (The netpoll_send_skb() portion of this patch was based on Andrew's bh-locking based netpoll patch in -mm.)
Signed-off-by: Ingo Molnar <mingo@elte.hu> --- net/core/netpoll.c | 39 ++++++++++++++++++++++++--------------- 1 file changed, 24 insertions(+), 15 deletions(-)
Index: linux-hres-timers.q/net/core/netpoll.c =================================================================== --- linux-hres-timers.q.orig/net/core/netpoll.c +++ linux-hres-timers.q/net/core/netpoll.c @@ -55,6 +55,7 @@ static void queue_process(struct work_st struct netpoll_info *npinfo = container_of(work, struct netpoll_info, tx_work.work); struct sk_buff *skb; + unsigned long flags; while ((skb = skb_dequeue(&npinfo->txq))) { struct net_device *dev = skb->dev; @@ -64,15 +65,19 @@ static void queue_process(struct work_st continue; } - netif_tx_lock_bh(dev); + local_irq_save(flags); + netif_tx_lock(dev); if (netif_queue_stopped(dev) || dev->hard_start_xmit(skb, dev) != NETDEV_TX_OK) { skb_queue_head(&npinfo->txq, skb); - netif_tx_unlock_bh(dev); + netif_tx_unlock(dev); + local_irq_restore(flags); schedule_delayed_work(&npinfo->tx_work, HZ/10); return; } + netif_tx_unlock(dev); + local_irq_restore(flags); } } @@ -231,7 +236,7 @@ repeat: static void netpoll_send_skb(struct netpoll *np, struct sk_buff *skb) { int status = NETDEV_TX_BUSY; - unsigned long tries; + unsigned long tries, flags; struct net_device *dev = np->dev; struct netpoll_info *npinfo = np->dev->npinfo; @@ -242,22 +247,26 @@ static void netpoll_send_skb(struct netp /* don't get messages out of order, and no recursion */ if (skb_queue_len(&npinfo->txq) == 0 && - npinfo->poll_owner != smp_processor_id() && - netif_tx_trylock(dev)) { - /* try until next clock tick */ - for (tries = jiffies_to_usecs(1)/USEC_PER_POLL; tries > 0; --tries) { - if (!netif_queue_stopped(dev)) - status = dev->hard_start_xmit(skb, dev); + npinfo->poll_owner != smp_processor_id()) { + local_irq_save(flags); /* Where's netif_tx_trylock_irqsave()? */ + if (netif_tx_trylock(dev)) { + /* try until next clock tick */ + for (tries = jiffies_to_usecs(1)/USEC_PER_POLL; + tries > 0; --tries) { + if (!netif_queue_stopped(dev)) + status = dev->hard_start_xmit(skb, dev); - if (status == NETDEV_TX_OK) - break; + if (status == NETDEV_TX_OK) + break; - /* tickle device maybe there is some cleanup */ - netpoll_poll(np); + /* tickle device maybe there is some cleanup */ + netpoll_poll(np); - udelay(USEC_PER_POLL); + udelay(USEC_PER_POLL); + } + netif_tx_unlock(dev); } - netif_tx_unlock(dev); + local_irq_restore(flags); } if (status != NETDEV_TX_OK) { - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |