IPoIB: Disable NAPI while CQ is being drained

If NAPI is enabled while IPoIB's CQ is being drained, it creates a
race on priv->ibwc between ipoib_poll() and ipoib_drain_cq(), leading
to memory corruption.

The solution is to enable/disable NAPI in ipoib_ib_dev_{open/stop}()
instead of in ipoib_{open/stop}(), and sync NAPI on the INITIALIZED
flag instead on the ADMIN_UP flag. This way NAPI will be disabled when
ipoib_drain_cq() is called.

This fixes <https://bugs.openfabrics.org/show_bug.cgi?id=1587>.

Signed-off-by: Yossi Etigin <yosefe@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>

drivers/infiniband/ulp/ipoib/ipoib_ib.c

@@ -685,7 +685,8 @@ int ipoib_ib_dev_open(struct net_device *dev)
 	queue_delayed_work(ipoib_workqueue, &priv->ah_reap_task,
 			   round_jiffies_relative(HZ));
 
-	set_bit(IPOIB_FLAG_INITIALIZED, &priv->flags);
+	if (!test_and_set_bit(IPOIB_FLAG_INITIALIZED, &priv->flags))
+		napi_enable(&priv->napi);
 
 	return 0;
 }
@@ -804,7 +805,8 @@ int ipoib_ib_dev_stop(struct net_device *dev, int flush)
 	struct ipoib_tx_buf *tx_req;
 	int i;
 
-	clear_bit(IPOIB_FLAG_INITIALIZED, &priv->flags);
+	if (test_and_clear_bit(IPOIB_FLAG_INITIALIZED, &priv->flags))
+		napi_disable(&priv->napi);
 
 	ipoib_cm_dev_stop(dev);
 

drivers/infiniband/ulp/ipoib/ipoib_main.c

@@ -106,8 +106,7 @@ int ipoib_open(struct net_device *dev)
 
 	ipoib_dbg(priv, "bringing up interface\n");
 
-	if (!test_and_set_bit(IPOIB_FLAG_ADMIN_UP, &priv->flags))
-		napi_enable(&priv->napi);
+	set_bit(IPOIB_FLAG_ADMIN_UP, &priv->flags);
 
 	if (ipoib_pkey_dev_delay_open(dev))
 		return 0;
@@ -143,7 +142,6 @@ err_stop:
 	ipoib_ib_dev_stop(dev, 1);
 
 err_disable:
-	napi_disable(&priv->napi);
 	clear_bit(IPOIB_FLAG_ADMIN_UP, &priv->flags);
 
 	return -EINVAL;
@@ -156,7 +154,6 @@ static int ipoib_stop(struct net_device *dev)
 	ipoib_dbg(priv, "stopping interface\n");
 
 	clear_bit(IPOIB_FLAG_ADMIN_UP, &priv->flags);
-	napi_disable(&priv->napi);
 
 	netif_stop_queue(dev);