[PATCH] via-rhine: link state fix

[Roger Luethi <rl@hellgate.ch>]

Problems with link state detection have been reported several times in the
past months.

Denis Vlasenko did all the work tracking it down. Jeff Garzik suggested the
proper place for the fix.

When using the mii library, the driver needs to check mii->force_media
and set dev->state accordingly.

Roger

Signed-off-by: Roger Luethi <rl@hellgate.ch>

--- linux-2.6.15.6/drivers/net/via-rhine.c.orig	2006-03-12 13:32:16.000000000 +0100
+++ linux-2.6.15.6/drivers/net/via-rhine.c	2006-03-12 22:01:36.000000000 +0100
@@ -1085,6 +1085,25 @@
 	else
 	    iowrite8(ioread8(ioaddr + ChipCmd1) & ~Cmd1FDuplex,
 		   ioaddr + ChipCmd1);
+	if (debug > 1)
+		printk(KERN_INFO "%s: force_media %d, carrier %d\n", dev->name,
+			rp->mii_if.force_media, netif_carrier_ok(dev));
+}
+
+/* Called after status of force_media possibly changed */
+void rhine_set_carrier(struct mii_if_info *mii)
+{
+	if (mii->force_media) {
+		/* autoneg is off: Link is always assumed to be up */
+		if (!netif_carrier_ok(mii->dev))
+			netif_carrier_on(mii->dev);
+	}
+	else	/* Let MMI library update carrier status */
+		rhine_check_media(mii->dev, 0);
+	if (debug > 1)
+		printk(KERN_INFO "%s: force_media %d, carrier %d\n",
+		       mii->dev->name, mii->force_media,
+		       netif_carrier_ok(mii->dev));
 }
 
 static void rhine_check_media_task(struct net_device *dev)
@@ -1782,6 +1801,7 @@
 	spin_lock_irq(&rp->lock);
 	rc = mii_ethtool_sset(&rp->mii_if, cmd);
 	spin_unlock_irq(&rp->lock);
+	rhine_set_carrier(&rp->mii_if);
 
 	return rc;
 }
@@ -1869,6 +1889,7 @@
 	spin_lock_irq(&rp->lock);
 	rc = generic_mii_ioctl(&rp->mii_if, if_mii(rq), cmd, NULL);
 	spin_unlock_irq(&rp->lock);
+	rhine_set_carrier(&rp->mii_if);
 
 	return rc;
 }