On Wed, Jun 14, 2017 at 09:59:06AM +0300, Leon Romanovsky wrote:
From: Alex Vesker <valex@xxxxxxxxxxxx>
Limit calls to free_rdma_netdev() for capable devices only.
Fixes: cd565b4b51e5 ('IB/IPoIB: Support acceleration options callbacks')
Signed-off-by: Alex Vesker <valex@xxxxxxxxxxxx>
Signed-off-by: Leon Romanovsky <leon@xxxxxxxxxx>
---
drivers/infiniband/hw/mlx5/main.c | 6 ++++--
drivers/infiniband/ulp/ipoib/ipoib_main.c | 5 ++++-
2 files changed, 8 insertions(+), 3 deletions(-)
diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index 0c79983c8b1a..9ecc089d4529 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -3692,8 +3692,10 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev)
dev->ib_dev.check_mr_status = mlx5_ib_check_mr_status;
dev->ib_dev.get_port_immutable = mlx5_port_immutable;
dev->ib_dev.get_dev_fw_str = get_dev_fw_str;
- dev->ib_dev.alloc_rdma_netdev = mlx5_ib_alloc_rdma_netdev;
- dev->ib_dev.free_rdma_netdev = mlx5_ib_free_rdma_netdev;
+ if (MLX5_CAP_GEN(mdev, ipoib_enhanced_offloads)) {
+ dev->ib_dev.alloc_rdma_netdev = mlx5_ib_alloc_rdma_netdev;
+ dev->ib_dev.free_rdma_netdev = mlx5_ib_free_rdma_netdev;
+ }
if (mlx5_core_is_pf(mdev)) {
dev->ib_dev.get_vf_config = mlx5_ib_get_vf_config;
dev->ib_dev.set_vf_link_state = mlx5_ib_set_vf_link_state;
diff --git a/drivers/infiniband/ulp/ipoib/ipoib_main.c b/drivers/infiniband/ulp/ipoib/ipoib_main.c
index 0ddd9709e1df..91fae34bdd4f 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib_main.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c
@@ -2301,7 +2301,10 @@ static void ipoib_remove_one(struct ib_device *device, void *client_data)
flush_workqueue(priv->wq);
unregister_netdev(priv->dev);
- free_netdev(priv->dev);
+ if (device->free_rdma_netdev)
+ device->free_rdma_netdev(priv->dev);
+ else
+ free_netdev(priv->dev);
This is causing an crash in hfi1 driver.
If the call to alloc_rdma_netdev() has returned -EOPNOTSUPP, we shouldn't be
calling free_rdma_netdev() here (as device driver doesn't own that netdev).
I will send out a patch to hfi1 driver to guard check for the type in its
free_rdma_netdev() call to handle such issues.
But we need to properly free the netdev here, probably by something like this.
if ((priv->dev->netdev_ops != ipoib_netdev_default_pf) &&
device->free_rdma_netdev)
device->free_rdma_netdev(priv->dev);
else
free_netdev(priv->dev);
Niranjana
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
