net: trans_start cleanups

Now that core network takes care of trans_start updates, dont do it
in drivers themselves, if possible. Drivers can avoid one cache miss
(on dev->trans_start) in their start_xmit() handler.

Exceptions are NETIF_F_LLTX drivers

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This commit is contained in:
Eric Dumazet
2010-05-10 05:01:31 -07:00
committed by David S. Miller
parent 2b0b05ddc0
commit 1ae5dc342a
168 changed files with 103 additions and 314 deletions

View File

@@ -1905,7 +1905,7 @@ static void ns_tx_timeout(struct net_device *dev)
spin_unlock_irq(&np->lock);
enable_irq(dev->irq);
dev->trans_start = jiffies;
dev->trans_start = jiffies; /* prevent tx timeout */
np->stats.tx_errors++;
netif_wake_queue(dev);
}
@@ -2119,8 +2119,6 @@ static netdev_tx_t start_tx(struct sk_buff *skb, struct net_device *dev)
}
spin_unlock_irqrestore(&np->lock, flags);
dev->trans_start = jiffies;
if (netif_msg_tx_queued(np)) {
printk(KERN_DEBUG "%s: Transmit frame #%d queued in slot %d.\n",
dev->name, np->cur_tx, entry);