mlx4_core: INIT/CLOSE port logic for IB ports in SR-IOV mode

Normally, INIT_PORT and CLOSE_PORT are invoked when special QP0
transitions to RTR, or transitions to ERR/RESET respectively.

In SR-IOV mode, however, the master is also paravirtualized.  This in
turn requires that we not do INIT_PORT until the entire QP0 path (real
QP0 and proxy QP0) is ready to receive.  When the real QP0 goes down,
we should indicate that the port is not active.

Signed-off-by: Jack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: Roland Dreier <roland@purestorage.com>

2 files changed, 92 insertions, 17 deletions

diff --git a/drivers/net/ethernet/mellanox/mlx4/fw.c b/drivers/net/ethernet/mellanox/mlx4/fw.c
index b788039..bf159c0 100644
--- a/drivers/net/ethernet/mellanox/mlx4/fw.c
+++ b/drivers/net/ethernet/mellanox/mlx4/fw.c
@@ -1365,6 +1365,19 @@ out:
 	return err;
 }
 
+/* for IB-type ports only in SRIOV mode. Checks that both proxy QP0
+ * and real QP0 are active, so that the paravirtualized QP0 is ready
+ * to operate */
+static int check_qp0_state(struct mlx4_dev *dev, int function, int port)
+{
+	struct mlx4_priv *priv = mlx4_priv(dev);
+	/* irrelevant if not infiniband */
+	if (priv->mfunc.master.qp0_state[port].proxy_qp0_active &&
+	    priv->mfunc.master.qp0_state[port].qp0_active)
+		return 1;
+	return 0;
+}
+
 int mlx4_INIT_PORT_wrapper(struct mlx4_dev *dev, int slave,
 			   struct mlx4_vhcr *vhcr,
 			   struct mlx4_cmd_mailbox *inbox,
@@ -1381,14 +1394,29 @@ int mlx4_INIT_PORT_wrapper(struct mlx4_dev *dev, int slave,
 	if (dev->caps.port_mask[port] == MLX4_PORT_TYPE_IB)
 		return -ENODEV;
 
-	/* Enable port only if it was previously disabled */
-	if (!priv->mfunc.master.init_port_ref[port]) {
-		err = mlx4_cmd(dev, 0, port, 0, MLX4_CMD_INIT_PORT,
-			       MLX4_CMD_TIME_CLASS_A, MLX4_CMD_NATIVE);
-		if (err)
-			return err;
+	if (dev->caps.port_mask[port] != MLX4_PORT_TYPE_IB) {
+		/* Enable port only if it was previously disabled */
+		if (!priv->mfunc.master.init_port_ref[port]) {
+			err = mlx4_cmd(dev, 0, port, 0, MLX4_CMD_INIT_PORT,
+				       MLX4_CMD_TIME_CLASS_A, MLX4_CMD_NATIVE);
+			if (err)
+				return err;
+		}
+		priv->mfunc.master.slave_state[slave].init_port_mask |= (1 << port);
+	} else {
+		if (slave == mlx4_master_func_num(dev)) {
+			if (check_qp0_state(dev, slave, port) &&
+			    !priv->mfunc.master.qp0_state[port].port_active) {
+				err = mlx4_cmd(dev, 0, port, 0, MLX4_CMD_INIT_PORT,
+					       MLX4_CMD_TIME_CLASS_A, MLX4_CMD_NATIVE);
+				if (err)
+					return err;
+				priv->mfunc.master.qp0_state[port].port_active = 1;
+				priv->mfunc.master.slave_state[slave].init_port_mask |= (1 << port);
+			}
+		} else
+			priv->mfunc.master.slave_state[slave].init_port_mask |= (1 << port);
 	}
-	priv->mfunc.master.slave_state[slave].init_port_mask |= (1 << port);
 	++priv->mfunc.master.init_port_ref[port];
 	return 0;
 }
@@ -1463,13 +1491,30 @@ int mlx4_CLOSE_PORT_wrapper(struct mlx4_dev *dev, int slave,
 
 	if (dev->caps.port_mask[port] == MLX4_PORT_TYPE_IB)
 		return -ENODEV;
-	if (priv->mfunc.master.init_port_ref[port] == 1) {
-		err = mlx4_cmd(dev, 0, port, 0, MLX4_CMD_CLOSE_PORT, 1000,
-			       MLX4_CMD_NATIVE);
-		if (err)
-			return err;
+
+	if (dev->caps.port_mask[port] != MLX4_PORT_TYPE_IB) {
+		if (priv->mfunc.master.init_port_ref[port] == 1) {
+			err = mlx4_cmd(dev, 0, port, 0, MLX4_CMD_CLOSE_PORT,
+				       1000, MLX4_CMD_NATIVE);
+			if (err)
+				return err;
+		}
+		priv->mfunc.master.slave_state[slave].init_port_mask &= ~(1 << port);
+	} else {
+		/* infiniband port */
+		if (slave == mlx4_master_func_num(dev)) {
+			if (!priv->mfunc.master.qp0_state[port].qp0_active &&
+			    priv->mfunc.master.qp0_state[port].port_active) {
+				err = mlx4_cmd(dev, 0, port, 0, MLX4_CMD_CLOSE_PORT,
+					       1000, MLX4_CMD_NATIVE);
+				if (err)
+					return err;
+				priv->mfunc.master.slave_state[slave].init_port_mask &= ~(1 << port);
+				priv->mfunc.master.qp0_state[port].port_active = 0;
+			}
+		} else
+			priv->mfunc.master.slave_state[slave].init_port_mask &= ~(1 << port);
 	}
-	priv->mfunc.master.slave_state[slave].init_port_mask &= ~(1 << port);
 	--priv->mfunc.master.init_port_ref[port];
 	return 0;
 }
diff --git a/drivers/net/ethernet/mellanox/mlx4/qp.c b/drivers/net/ethernet/mellanox/mlx4/qp.c
index b8da72b..436ef6c 100644
--- a/drivers/net/ethernet/mellanox/mlx4/qp.c
+++ b/drivers/net/ethernet/mellanox/mlx4/qp.c
@@ -67,10 +67,18 @@ void mlx4_qp_event(struct mlx4_dev *dev, u32 qpn, int event_type)
 		complete(&qp->free);
 }
 
-static int is_qp0(struct mlx4_dev *dev, struct mlx4_qp *qp)
+/* used for INIT/CLOSE port logic */
+static int is_qp0(struct mlx4_dev *dev, struct mlx4_qp *qp, int *real_qp0, int *proxy_qp0)
 {
-	return qp->qpn >= dev->caps.sqp_start &&
+	/* qp0 is either the proxy qp0, or the real qp0 */
+	*proxy_qp0 = qp->qpn >= dev->caps.sqp_start &&
 		qp->qpn <= dev->caps.sqp_start + 1;
+
+	*real_qp0 = mlx4_is_master(dev) &&
+		qp->qpn >= dev->caps.base_sqpn &&
+		qp->qpn <= dev->caps.base_sqpn + 1;
+
+	return *real_qp0 || *proxy_qp0;
 }
 
 static int __mlx4_qp_modify(struct mlx4_dev *dev, struct mlx4_mtt *mtt,
@@ -122,6 +130,8 @@ static int __mlx4_qp_modify(struct mlx4_dev *dev, struct mlx4_mtt *mtt,
 	struct mlx4_priv *priv = mlx4_priv(dev);
 	struct mlx4_cmd_mailbox *mailbox;
 	int ret = 0;
+	int real_qp0 = 0;
+	int proxy_qp0 = 0;
 	u8 port;
 
 	if (cur_state >= MLX4_QP_NUM_STATE || new_state >= MLX4_QP_NUM_STATE ||
@@ -133,9 +143,12 @@ static int __mlx4_qp_modify(struct mlx4_dev *dev, struct mlx4_mtt *mtt,
 			MLX4_CMD_2RST_QP, MLX4_CMD_TIME_CLASS_A, native);
 		if (mlx4_is_master(dev) && cur_state != MLX4_QP_STATE_ERR &&
 		    cur_state != MLX4_QP_STATE_RST &&
-		    is_qp0(dev, qp)) {
+		    is_qp0(dev, qp, &real_qp0, &proxy_qp0)) {
 			port = (qp->qpn & 1) + 1;
-			priv->mfunc.master.qp0_state[port].qp0_active = 0;
+			if (proxy_qp0)
+				priv->mfunc.master.qp0_state[port].proxy_qp0_active = 0;
+			else
+				priv->mfunc.master.qp0_state[port].qp0_active = 0;
 		}
 		return ret;
 	}
@@ -162,6 +175,23 @@ static int __mlx4_qp_modify(struct mlx4_dev *dev, struct mlx4_mtt *mtt,
 		       new_state == MLX4_QP_STATE_RST ? 2 : 0,
 		       op[cur_state][new_state], MLX4_CMD_TIME_CLASS_C, native);
 
+	if (mlx4_is_master(dev) && is_qp0(dev, qp, &real_qp0, &proxy_qp0)) {
+		port = (qp->qpn & 1) + 1;
+		if (cur_state != MLX4_QP_STATE_ERR &&
+		    cur_state != MLX4_QP_STATE_RST &&
+		    new_state == MLX4_QP_STATE_ERR) {
+			if (proxy_qp0)
+				priv->mfunc.master.qp0_state[port].proxy_qp0_active = 0;
+			else
+				priv->mfunc.master.qp0_state[port].qp0_active = 0;
+		} else if (new_state == MLX4_QP_STATE_RTR) {
+			if (proxy_qp0)
+				priv->mfunc.master.qp0_state[port].proxy_qp0_active = 1;
+			else
+				priv->mfunc.master.qp0_state[port].qp0_active = 1;
+		}
+	}
+
 	mlx4_free_cmd_mailbox(dev, mailbox);
 	return ret;
 }