Commit 4b31fd4d authored by Rafal Rogalski's avatar Rafal Rogalski Committed by Jakub Kicinski
Browse files

ice: Fix RDMA VSI removal during queue rebuild 



During qdisc create/delete, it is necessary to rebuild the queue
of VSIs. An error occurred because the VSIs created by RDMA were
still active.

Added check if RDMA is active. If yes, it disallows qdisc changes
and writes a message in the system logs.

Fixes: 348048e7 ("ice: Implement iidc operations")
Signed-off-by: default avatarRafal Rogalski <rafalx.rogalski@intel.com>
Signed-off-by: default avatarMateusz Palczewski <mateusz.palczewski@intel.com>
Signed-off-by: default avatarKamil Maziarz <kamil.maziarz@intel.com>
Tested-by: default avatarBharathi Sreenivas <bharathi.sreenivas@intel.com>
Signed-off-by: default avatarTony Nguyen <anthony.l.nguyen@intel.com>
Reviewed-by: default avatarLeon Romanovsky <leonro@nvidia.com>
Link: https://lore.kernel.org/r/20230728171243.2446101-1-anthony.l.nguyen@intel.com


Signed-off-by: default avatarJakub Kicinski <kuba@kernel.org>
parent 55c1528f

drivers/net/ethernet/intel/ice/ice_main.c

+18 −0
Original line number Diff line number Diff line
@@ -8813,6 +8813,7 @@ ice_setup_tc(struct net_device *netdev, enum tc_setup_type type, 
{

	struct ice_netdev_priv *np = netdev_priv(netdev);

	struct ice_pf *pf = np->vsi->back;

	bool locked = false;

	int err;



	switch (type) {
@@ -8822,10 +8823,27 @@ ice_setup_tc(struct net_device *netdev, enum tc_setup_type type, 
						  ice_setup_tc_block_cb,

						  np, np, true);

	case TC_SETUP_QDISC_MQPRIO:

		if (pf->adev) {

			mutex_lock(&pf->adev_mutex);

			device_lock(&pf->adev->dev);

			locked = true;

			if (pf->adev->dev.driver) {

				netdev_err(netdev, "Cannot change qdisc when RDMA is active\n");

				err = -EBUSY;

				goto adev_unlock;

			}

		}



		/* setup traffic classifier for receive side */

		mutex_lock(&pf->tc_mutex);

		err = ice_setup_tc_mqprio_qdisc(netdev, type_data);

		mutex_unlock(&pf->tc_mutex);



adev_unlock:

		if (locked) {

			device_unlock(&pf->adev->dev);

			mutex_unlock(&pf->adev_mutex);

		}

		return err;

	default:

		return -EOPNOTSUPP;