e1000e: stop cleaning when we reach tx_ring->next_to_use
authorTerry Loftin <terry.loftin@hp.com>
Fri, 9 Apr 2010 10:29:49 +0000 (10:29 +0000)
committerDavid S. Miller <davem@davemloft.net>
Tue, 13 Apr 2010 10:03:13 +0000 (03:03 -0700)
Tx ring buffers after tx_ring->next_to_use are volatile and could
change, possibly causing a crash.  Stop cleaning when we hit
tx_ring->next_to_use.

Signed-off-by: Terry Loftin <terry.loftin@hp.com>
Acked-by: Bruce Allan <bruce.w.allan@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
drivers/net/e1000e/netdev.c

index cfd09ce..73d43c5 100644 (file)
@@ -661,6 +661,8 @@ static bool e1000_clean_tx_irq(struct e1000_adapter *adapter)
                                i = 0;
                }
 
+               if (i == tx_ring->next_to_use)
+                       break;
                eop = tx_ring->buffer_info[i].next_to_watch;
                eop_desc = E1000_TX_DESC(*tx_ring, eop);
        }