IB/ipoib: Avoid flushing the workqueue from worker context

The error flow of ipoib_ib_dev_open() invokes ipoib_ib_dev_stop() with
workqueue flushing enabled, which deadlocks if the open procedure
itself was called by a worker thread.

Fix this by adding a flush enabled flag to ipoib_ib_dev_open() and set
it accordingly from the locations where such a call is made.

The call trace was the following:

 [<ffffffff81095bc4>] ? flush_workqueue+0x54/0x80
 [<ffffffffa056c657>] ? ipoib_ib_dev_stop+0x447/0x650 [ib_ipoib]
 [<ffffffffa056cc34>] ? ipoib_ib_dev_open+0x284/0x430 [ib_ipoib]
 [<ffffffffa05674a8>] ? ipoib_open+0x78/0x1d0 [ib_ipoib]
 [<ffffffffa05697b8>] ? ipoib_pkey_open+0x38/0x40 [ib_ipoib]
 [<ffffffffa056cf3c>] ? __ipoib_ib_dev_flush+0x15c/0x2c0 [ib_ipoib]
 [<ffffffffa056ce56>] ? __ipoib_ib_dev_flush+0x76/0x2c0 [ib_ipoib]
 [<ffffffffa056d0a0>] ? ipoib_ib_dev_flush_heavy+0x0/0x20 [ib_ipoib]
 [<ffffffffa056d0ba>] ? ipoib_ib_dev_flush_heavy+0x1a/0x20 [ib_ipoib]
 [<ffffffff81094d20>] ? worker_thread+0x170/0x2a0
 [<ffffffff8109b2a0>] ? autoremove_wake_function+0x0/0x40

Signed-off-by: Erez Shitrit <erezsh@mellanox.com>
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Acked-by: Alex Estrin <alex.estrin@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>

diff --git a/drivers/infiniband/ulp/ipoib/ipoib.h b/drivers/infiniband/ulp/ipoib/ipoib.h
index 683d23a..3edce61 100644 (file)
--- a/drivers/infiniband/ulp/ipoib/ipoib.h
+++ b/drivers/infiniband/ulp/ipoib/ipoib.h
@@ -471,7 +471,7 @@ void ipoib_ib_dev_flush_heavy(struct work_struct *work);
 void ipoib_pkey_event(struct work_struct *work);
 void ipoib_ib_dev_cleanup(struct net_device *dev);
 
-int ipoib_ib_dev_open(struct net_device *dev);
+int ipoib_ib_dev_open(struct net_device *dev, int flush);
 int ipoib_ib_dev_up(struct net_device *dev);
 int ipoib_ib_dev_down(struct net_device *dev, int flush);
 int ipoib_ib_dev_stop(struct net_device *dev, int flush);

diff --git a/drivers/infiniband/ulp/ipoib/ipoib_ib.c b/drivers/infiniband/ulp/ipoib/ipoib_ib.c
index be8f971..9dcb2c9 100644 (file)
--- a/drivers/infiniband/ulp/ipoib/ipoib_ib.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_ib.c
@@ -664,7 +664,7 @@ static void ipoib_ib_tx_timer_func(unsigned long ctx)
        drain_tx_cq((struct net_device *)ctx);
 }
 
-int ipoib_ib_dev_open(struct net_device *dev)
+int ipoib_ib_dev_open(struct net_device *dev, int flush)
 {
        struct ipoib_dev_priv *priv = netdev_priv(dev);
        int ret;
@@ -705,7 +705,7 @@ int ipoib_ib_dev_open(struct net_device *dev)
 dev_stop:
        if (!test_and_set_bit(IPOIB_FLAG_INITIALIZED, &priv->flags))
                napi_enable(&priv->napi);
-       ipoib_ib_dev_stop(dev, 1);
+       ipoib_ib_dev_stop(dev, flush);
        return -1;
 }
 
@@ -916,7 +916,7 @@ int ipoib_ib_dev_init(struct net_device *dev, struct ib_device *ca, int port)
                    (unsigned long) dev);
 
        if (dev->flags & IFF_UP) {
-               if (ipoib_ib_dev_open(dev)) {
+               if (ipoib_ib_dev_open(dev, 1)) {
                        ipoib_transport_dev_cleanup(dev);
                        return -ENODEV;
                }
@@ -1033,7 +1033,7 @@ static void __ipoib_ib_dev_flush(struct ipoib_dev_priv *priv,
 
        if (level == IPOIB_FLUSH_HEAVY) {
                ipoib_ib_dev_stop(dev, 0);
-               ipoib_ib_dev_open(dev);
+               ipoib_ib_dev_open(dev, 0);
        }
 
        /*

diff --git a/drivers/infiniband/ulp/ipoib/ipoib_main.c b/drivers/infiniband/ulp/ipoib/ipoib_main.c
index 35acbd4..1bf994a 100644 (file)
--- a/drivers/infiniband/ulp/ipoib/ipoib_main.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c
@@ -114,7 +114,7 @@ int ipoib_open(struct net_device *dev)
        if (!test_bit(IPOIB_PKEY_ASSIGNED, &priv->flags))
                return 0;
 
-       if (ipoib_ib_dev_open(dev))
+       if (ipoib_ib_dev_open(dev, 1))
                goto err_disable;
 
        if (ipoib_ib_dev_up(dev))