From: Moshe Shemesh moshe@mellanox.com
[ Upstream commit fed91613c9dd455dd154b22fa8e11b8526466082 ]
Add restarting state flag to avoid scheduling another restart task while such task is already running. Change task name from watchdog_task to restart_task to better fit the task role.
Fixes: 1e338db56e5a ("mlx4_en: Fix a race at restart task") Signed-off-by: Moshe Shemesh moshe@mellanox.com Signed-off-by: Tariq Toukan tariqt@nvidia.com Signed-off-by: David S. Miller davem@davemloft.net Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org --- drivers/net/ethernet/mellanox/mlx4/en_netdev.c | 20 +++++++++++++------- drivers/net/ethernet/mellanox/mlx4/mlx4_en.h | 7 ++++++- 2 files changed, 19 insertions(+), 8 deletions(-)
--- a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c +++ b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c @@ -1311,8 +1311,10 @@ static void mlx4_en_tx_timeout(struct ne }
priv->port_stats.tx_timeout++; - en_dbg(DRV, priv, "Scheduling watchdog\n"); - queue_work(mdev->workqueue, &priv->watchdog_task); + if (!test_and_set_bit(MLX4_EN_STATE_FLAG_RESTARTING, &priv->state)) { + en_dbg(DRV, priv, "Scheduling port restart\n"); + queue_work(mdev->workqueue, &priv->restart_task); + } }
@@ -1746,6 +1748,7 @@ int mlx4_en_start_port(struct net_device local_bh_enable(); }
+ clear_bit(MLX4_EN_STATE_FLAG_RESTARTING, &priv->state); netif_tx_start_all_queues(dev); netif_device_attach(dev);
@@ -1900,7 +1903,7 @@ void mlx4_en_stop_port(struct net_device static void mlx4_en_restart(struct work_struct *work) { struct mlx4_en_priv *priv = container_of(work, struct mlx4_en_priv, - watchdog_task); + restart_task); struct mlx4_en_dev *mdev = priv->mdev; struct net_device *dev = priv->dev;
@@ -2220,7 +2223,7 @@ static int mlx4_en_change_mtu(struct net if (netif_running(dev)) { mutex_lock(&mdev->state_lock); if (!mdev->device_up) { - /* NIC is probably restarting - let watchdog task reset + /* NIC is probably restarting - let restart task reset * the port */ en_dbg(DRV, priv, "Change MTU called with card down!?\n"); } else { @@ -2229,7 +2232,9 @@ static int mlx4_en_change_mtu(struct net if (err) { en_err(priv, "Failed restarting port:%d\n", priv->port); - queue_work(mdev->workqueue, &priv->watchdog_task); + if (!test_and_set_bit(MLX4_EN_STATE_FLAG_RESTARTING, + &priv->state)) + queue_work(mdev->workqueue, &priv->restart_task); } } mutex_unlock(&mdev->state_lock); @@ -2701,7 +2706,8 @@ static int mlx4_xdp_set(struct net_devic if (err) { en_err(priv, "Failed starting port %d for XDP change\n", priv->port); - queue_work(mdev->workqueue, &priv->watchdog_task); + if (!test_and_set_bit(MLX4_EN_STATE_FLAG_RESTARTING, &priv->state)) + queue_work(mdev->workqueue, &priv->restart_task); } }
@@ -3080,7 +3086,7 @@ int mlx4_en_init_netdev(struct mlx4_en_d priv->counter_index = MLX4_SINK_COUNTER_INDEX(mdev->dev); spin_lock_init(&priv->stats_lock); INIT_WORK(&priv->rx_mode_task, mlx4_en_do_set_rx_mode); - INIT_WORK(&priv->watchdog_task, mlx4_en_restart); + INIT_WORK(&priv->restart_task, mlx4_en_restart); INIT_WORK(&priv->linkstate_task, mlx4_en_linkstate); INIT_DELAYED_WORK(&priv->stats_task, mlx4_en_do_get_stats); INIT_DELAYED_WORK(&priv->service_task, mlx4_en_service_task); --- a/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h +++ b/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h @@ -522,6 +522,10 @@ struct mlx4_en_stats_bitmap { struct mutex mutex; /* for mutual access to stats bitmap */ };
+enum { + MLX4_EN_STATE_FLAG_RESTARTING, +}; + struct mlx4_en_priv { struct mlx4_en_dev *mdev; struct mlx4_en_port_profile *prof; @@ -586,7 +590,7 @@ struct mlx4_en_priv { struct mlx4_en_cq *rx_cq[MAX_RX_RINGS]; struct mlx4_qp drop_qp; struct work_struct rx_mode_task; - struct work_struct watchdog_task; + struct work_struct restart_task; struct work_struct linkstate_task; struct delayed_work stats_task; struct delayed_work service_task; @@ -632,6 +636,7 @@ struct mlx4_en_priv { u32 pflags; u8 rss_key[MLX4_EN_RSS_KEY_SIZE]; u8 rss_hash_fn; + unsigned long state; };
enum mlx4_en_wol {