Re: [bug?] r8169: hangs under heavy load

[Eric Dumazet <eric.dumazet@gmail.com>]

Le vendredi 25 novembre 2011 à 14:54 -0600, Jonathan Nieder a écrit :
> Eric Dumazet wrote:
> 
> > rtl8169_rx_interrupt(..., budget) can return budget + 1 sometimes
> > because of :
> >
> >                 /* Work around for AMD plateform. */
> >                 if ((desc->opts2 & cpu_to_le32(0xfffe000)) &&
> >                     (tp->mac_version == RTL_GIGA_MAC_VER_05)) {
> >                         desc->opts2 = 0;
> >                         cur_rx++;
> >                 }
> >
> > Sorry, I wont patch this today, its black Friday, and David said to
> > patch submitters :
> >
> > "stick to turkey and wine you're better at it" 
> 
> :)  Sensible enough.  Thanks for the quick response!

;)

You could try this non official patch, if you are in a hurry ;)

diff --git a/drivers/net/ethernet/realtek/r8169.c b/drivers/net/ethernet/realtek/r8169.c
index 6f06aa1..b8a537f 100644
--- a/drivers/net/ethernet/realtek/r8169.c
+++ b/drivers/net/ethernet/realtek/r8169.c
@@ -5710,6 +5710,7 @@ static int rtl8169_rx_interrupt(struct net_device *dev,
 {
 	unsigned int cur_rx, rx_left;
 	unsigned int count;
+	unsigned int workaround = 0;
 
 	cur_rx = tp->cur_rx;
 	rx_left = NUM_RX_DESC + tp->dirty_rx - cur_rx;
@@ -5780,6 +5781,7 @@ static int rtl8169_rx_interrupt(struct net_device *dev,
 		    (tp->mac_version == RTL_GIGA_MAC_VER_05)) {
 			desc->opts2 = 0;
 			cur_rx++;
+			workaround++;
 		}
 	}
 
@@ -5788,7 +5790,7 @@ static int rtl8169_rx_interrupt(struct net_device *dev,
 
 	tp->dirty_rx += count;
 
-	return count;
+	return count - workaround;
 }
 
 static irqreturn_t rtl8169_interrupt(int irq, void *dev_instance)