diff options
| author | Eric Dumazet <eric.dumazet@gmail.com> | 2010-05-10 14:01:31 +0200 |
|---|---|---|
| committer | David S. Miller <davem@davemloft.net> | 2010-05-10 14:01:31 +0200 |
| commit | 1ae5dc342ac78d7a42965fd1f323815f6f5ef2c1 (patch) | |
| tree | d1955a7639e99832590df26466a34d5786a880ae /drivers/net/eexpress.c | |
| parent | Bluetooth: Fix issues where sk_sleep() helper is needed now (diff) | |
| download | linux-1ae5dc342ac78d7a42965fd1f323815f6f5ef2c1.tar.xz linux-1ae5dc342ac78d7a42965fd1f323815f6f5ef2c1.zip | |
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>
Diffstat (limited to 'drivers/net/eexpress.c')
| -rw-r--r-- | drivers/net/eexpress.c | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/drivers/net/eexpress.c b/drivers/net/eexpress.c index c31dd0685553..43c9c9c5cf4c 100644 --- a/drivers/net/eexpress.c +++ b/drivers/net/eexpress.c @@ -543,7 +543,7 @@ static void unstick_cu(struct net_device *dev) if (lp->started) { - if (time_after(jiffies, dev->trans_start + 50)) + if (time_after(jiffies, dev_trans_start(dev) + HZ/2)) { if (lp->tx_link==lp->last_tx_restart) { @@ -1018,7 +1018,7 @@ static void eexp_hw_tx_pio(struct net_device *dev, unsigned short *buf, outw(lp->tx_head+0x16, ioaddr + DATAPORT); outw(0, ioaddr + DATAPORT); - outsw(ioaddr + DATAPORT, buf, (len+1)>>1); + outsw(ioaddr + DATAPORT, buf, (len+1)>>1); outw(lp->tx_tail+0xc, ioaddr + WRITE_PTR); outw(lp->tx_head, ioaddr + DATAPORT); |