[openib-general] Fwd: issues in ipoib

[Michael S. Tsirkin]

Quoting r. Roland Dreier <rdreier at cisco.com>:
> Subject: Re: Fwd: issues in ipoib
> 
>  > 1. pkey cache issues
>  > http://thread.gmane.org/gmane.linux.drivers.openib/26684/focus=26692
> 
> I thought we fixed the P_Key cache issues by correcting the oversight
> in retrying the P_Key query?
> 
>  > 3. ipoib race reported after code review by Eitan Rabin
>  > http://openib.org/pipermail/openib-general/2006-June/022916.html
> 
> Yeah, might be a problem I guess.  Does it work to do
> netif_stop_queue() in ipoib_ib_dev_down()?

Hmm. Since we are lockless, could ipoib_start_xmit run even after we call
netif_stop_queue?  Since interrupts are disabled anyway, can we just just take
tx_lock?  How does the following look?

---

Prevent flush task from freeing the ipoib_neigh pointer,
while ipoib_start_xmit is accessing the ipoib_neigh through
the pointer is has loaded from the hardware address.

Signed-off-by: Michael S. Tsirkin <mst at mellanox.co.il>

diff --git a/drivers/infiniband/ulp/ipoib/ipoib_main.c b/drivers/infiniband/ulp/ipoib/ipoib_main.c
index cf71d2a..31c4b05 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib_main.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c
@@ -336,7 +336,8 @@ void ipoib_flush_paths(struct net_device
 	struct ipoib_path *path, *tp;
 	LIST_HEAD(remove_list);
 
-	spin_lock_irq(&priv->lock);
+	spin_lock_irq(&priv->tx_lock);
+	spin_lock(&priv->lock);
 
 	list_splice(&priv->path_list, &remove_list);
 	INIT_LIST_HEAD(&priv->path_list);
@@ -352,7 +353,8 @@ void ipoib_flush_paths(struct net_device
 		path_free(dev, path);
 		spin_lock_irq(&priv->lock);
 	}
-	spin_unlock_irq(&priv->lock);
+	spin_unlock(&priv->lock);
+	spin_unlock_irq(&priv->tx_lock);
 }
 
 static void path_rec_completion(int status,


-- 
MST