Merge branch 'mlx4-core-fixes'

Tariq Toukan says:

====================
mlx4 core fixes

This patchset contains bug fixes from Jack to the mlx4 Core driver.

Patch 1 solves a race in the flow of CQ free.
Patch 2 moves some qp context flags update to the correct qp transition.
Patch 3 eliminates warnings from the path of SRQ_LIMIT that flood the message log,
and keeps them only in the path of SRQ_CATAS_ERROR.

Series generated against net commit:
1a717fcf Merge tag 'mac80211-for-davem-2017-01-13' of git://git.kernel.org/pub/scm/linux/kernel/git/jberg/mac80211
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

drivers/net/ethernet/mellanox/mlx4/cq.c

@@ -101,13 +101,19 @@ void mlx4_cq_completion(struct mlx4_dev *dev, u32 cqn)
 {
 	struct mlx4_cq *cq;
 
+	rcu_read_lock();
 	cq = radix_tree_lookup(&mlx4_priv(dev)->cq_table.tree,
 			       cqn & (dev->caps.num_cqs - 1));
+	rcu_read_unlock();
+
 	if (!cq) {
 		mlx4_dbg(dev, "Completion event for bogus CQ %08x\n", cqn);
 		return;
 	}
 
+	/* Acessing the CQ outside of rcu_read_lock is safe, because
+	 * the CQ is freed only after interrupt handling is completed.
+	 */
 	++cq->arm_sn;
 
 	cq->comp(cq);
@@ -118,23 +124,19 @@ void mlx4_cq_event(struct mlx4_dev *dev, u32 cqn, int event_type)
 	struct mlx4_cq_table *cq_table = &mlx4_priv(dev)->cq_table;
 	struct mlx4_cq *cq;
 
-	spin_lock(&cq_table->lock);
-
+	rcu_read_lock();
 	cq = radix_tree_lookup(&cq_table->tree, cqn & (dev->caps.num_cqs - 1));
-	if (cq)
-		atomic_inc(&cq->refcount);
-
-	spin_unlock(&cq_table->lock);
+	rcu_read_unlock();
 
 	if (!cq) {
-		mlx4_warn(dev, "Async event for bogus CQ %08x\n", cqn);
+		mlx4_dbg(dev, "Async event for bogus CQ %08x\n", cqn);
 		return;
 	}
 
+	/* Acessing the CQ outside of rcu_read_lock is safe, because
+	 * the CQ is freed only after interrupt handling is completed.
+	 */
 	cq->event(cq, event_type);
-
-	if (atomic_dec_and_test(&cq->refcount))
-		complete(&cq->free);
 }
 
 static int mlx4_SW2HW_CQ(struct mlx4_dev *dev, struct mlx4_cmd_mailbox *mailbox,
@@ -301,9 +303,9 @@ int mlx4_cq_alloc(struct mlx4_dev *dev, int nent,
 	if (err)
 		return err;
 
-	spin_lock_irq(&cq_table->lock);
+	spin_lock(&cq_table->lock);
 	err = radix_tree_insert(&cq_table->tree, cq->cqn, cq);
-	spin_unlock_irq(&cq_table->lock);
+	spin_unlock(&cq_table->lock);
 	if (err)
 		goto err_icm;
 
@@ -349,9 +351,9 @@ int mlx4_cq_alloc(struct mlx4_dev *dev, int nent,
 	return 0;
 
 err_radix:
-	spin_lock_irq(&cq_table->lock);
+	spin_lock(&cq_table->lock);
 	radix_tree_delete(&cq_table->tree, cq->cqn);
-	spin_unlock_irq(&cq_table->lock);
+	spin_unlock(&cq_table->lock);
 
 err_icm:
 	mlx4_cq_free_icm(dev, cq->cqn);
@@ -370,15 +372,15 @@ void mlx4_cq_free(struct mlx4_dev *dev, struct mlx4_cq *cq)
 	if (err)
 		mlx4_warn(dev, "HW2SW_CQ failed (%d) for CQN %06x\n", err, cq->cqn);
 
+	spin_lock(&cq_table->lock);
+	radix_tree_delete(&cq_table->tree, cq->cqn);
+	spin_unlock(&cq_table->lock);
+
 	synchronize_irq(priv->eq_table.eq[MLX4_CQ_TO_EQ_VECTOR(cq->vector)].irq);
 	if (priv->eq_table.eq[MLX4_CQ_TO_EQ_VECTOR(cq->vector)].irq !=
 	    priv->eq_table.eq[MLX4_EQ_ASYNC].irq)
 		synchronize_irq(priv->eq_table.eq[MLX4_EQ_ASYNC].irq);
 
-	spin_lock_irq(&cq_table->lock);
-	radix_tree_delete(&cq_table->tree, cq->cqn);
-	spin_unlock_irq(&cq_table->lock);
-
 	if (atomic_dec_and_test(&cq->refcount))
 		complete(&cq->free);
 	wait_for_completion(&cq->free);

drivers/net/ethernet/mellanox/mlx4/eq.c

@@ -554,8 +554,9 @@ static int mlx4_eq_int(struct mlx4_dev *dev, struct mlx4_eq *eq)
 			break;
 
 		case MLX4_EVENT_TYPE_SRQ_LIMIT:
-			mlx4_dbg(dev, "%s: MLX4_EVENT_TYPE_SRQ_LIMIT\n",
-				 __func__);
+			mlx4_dbg(dev, "%s: MLX4_EVENT_TYPE_SRQ_LIMIT. srq_no=0x%x, eq 0x%x\n",
+				 __func__, be32_to_cpu(eqe->event.srq.srqn),
+				 eq->eqn);
 		case MLX4_EVENT_TYPE_SRQ_CATAS_ERROR:
 			if (mlx4_is_master(dev)) {
 				/* forward only to slave owning the SRQ */
@@ -570,12 +571,16 @@ static int mlx4_eq_int(struct mlx4_dev *dev, struct mlx4_eq *eq)
 						  eq->eqn, eq->cons_index, ret);
 					break;
 				}
+				if (eqe->type ==
+				    MLX4_EVENT_TYPE_SRQ_CATAS_ERROR)
 					mlx4_warn(dev, "%s: slave:%d, srq_no:0x%x, event: %02x(%02x)\n",
 						  __func__, slave,
 						  be32_to_cpu(eqe->event.srq.srqn),
 						  eqe->type, eqe->subtype);
 
 				if (!ret && slave != dev->caps.function) {
+					if (eqe->type ==
+					    MLX4_EVENT_TYPE_SRQ_CATAS_ERROR)
 						mlx4_warn(dev, "%s: sending event %02x(%02x) to slave:%d\n",
 							  __func__, eqe->type,
 							  eqe->subtype, slave);

drivers/net/ethernet/mellanox/mlx4/resource_tracker.c

@@ -2980,6 +2980,9 @@ int mlx4_RST2INIT_QP_wrapper(struct mlx4_dev *dev, int slave,
 		put_res(dev, slave, srqn, RES_SRQ);
 		qp->srq = srq;
 	}
+
+	/* Save param3 for dynamic changes from VST back to VGT */
+	qp->param3 = qpc->param3;
 	put_res(dev, slave, rcqn, RES_CQ);
 	put_res(dev, slave, mtt_base, RES_MTT);
 	res_end_move(dev, slave, RES_QP, qpn);
@@ -3772,7 +3775,6 @@ int mlx4_INIT2RTR_QP_wrapper(struct mlx4_dev *dev, int slave,
 	int qpn = vhcr->in_modifier & 0x7fffff;
 	struct res_qp *qp;
 	u8 orig_sched_queue;
-	__be32	orig_param3 = qpc->param3;
 	u8 orig_vlan_control = qpc->pri_path.vlan_control;
 	u8 orig_fvl_rx = qpc->pri_path.fvl_rx;
 	u8 orig_pri_path_fl = qpc->pri_path.fl;
@@ -3814,7 +3816,6 @@ int mlx4_INIT2RTR_QP_wrapper(struct mlx4_dev *dev, int slave,
 	 */
 	if (!err) {
 		qp->sched_queue = orig_sched_queue;
-		qp->param3	= orig_param3;
 		qp->vlan_control = orig_vlan_control;
 		qp->fvl_rx	=  orig_fvl_rx;
 		qp->pri_path_fl = orig_pri_path_fl;