net: bcmgenet: fix off-by-one in incrementing read pointer
authorFlorian Fainelli <f.fainelli@gmail.com>
Fri, 10 Oct 2014 17:51:52 +0000 (10:51 -0700)
committerDavid S. Miller <davem@davemloft.net>
Fri, 10 Oct 2014 19:39:15 +0000 (15:39 -0400)
Commit b629be5c8399d7c423b92135eb43a86c924d1cbc ("net: bcmgenet: check
harder for out of memory conditions") moved the increment of the local
read pointer *before* reading from the hardware descriptor using
dmadesc_get_length_status(), which creates an off-by-one situation.

Fix this by moving again the read_ptr increment after we have read the
hardware descriptor to get both the control block and the read pointer
back in sync.

Fixes: b629be5c8399 ("net: bcmgenet: check harder for out of memory conditions")
Signed-off-by: Jaedon Shin <jaedon.shin@gmail.com>
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Acked-by: Petri Gynther <pgynther@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
drivers/net/ethernet/broadcom/genet/bcmgenet.c

index fff2634b6f3414258655d2174e7988a1be3f843d..fdc9ec09e453510636460780226e0349cc51b9a5 100644 (file)
@@ -1285,11 +1285,6 @@ static unsigned int bcmgenet_desc_rx(struct bcmgenet_priv *priv,
                cb = &priv->rx_cbs[priv->rx_read_ptr];
                skb = cb->skb;
 
-               rxpktprocessed++;
-
-               priv->rx_read_ptr++;
-               priv->rx_read_ptr &= (priv->num_rx_bds - 1);
-
                /* We do not have a backing SKB, so we do not have a
                 * corresponding DMA mapping for this incoming packet since
                 * bcmgenet_rx_refill always either has both skb and mapping or
@@ -1404,6 +1399,10 @@ refill:
                err = bcmgenet_rx_refill(priv, cb);
                if (err)
                        netif_err(priv, rx_err, dev, "Rx refill failed\n");
+
+               rxpktprocessed++;
+               priv->rx_read_ptr++;
+               priv->rx_read_ptr &= (priv->num_rx_bds - 1);
        }
 
        return rxpktprocessed;