Merge tag 'mlx5-updates-2023-06-21' of git://git.kernel.org/pub/scm/l…

…inux/kernel/git/saeed/linux

Saeed Mahameed says:

====================
mlx5-updates-2023-06-21

mlx5 driver minor cleanup and fixes to net-next

* tag 'mlx5-updates-2023-06-21' of git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linux:
  net/mlx5: Remove pointless vport lookup from mlx5_esw_check_port_type()
  net/mlx5: Remove redundant check from mlx5_esw_query_vport_vhca_id()
  net/mlx5: Remove redundant is_mdev_switchdev_mode() check from is_ib_rep_supported()
  net/mlx5: Remove redundant MLX5_ESWITCH_MANAGER() check from is_ib_rep_supported()
  net/mlx5e: E-Switch, Fix shared fdb error flow
  net/mlx5e: Remove redundant comment
  net/mlx5e: E-Switch, Pass other_vport flag if vport is not 0
  net/mlx5e: E-Switch, Use xarray for devcom paired device index
  net/mlx5e: E-Switch, Add peer fdb miss rules for vport manager or ecpf
  net/mlx5e: Use vhca_id for device index in vport rx rules
  net/mlx5: Lag, Remove duplicate code checking lag is supported
  net/mlx5: Fix error code in mlx5_is_reset_now_capable()
  net/mlx5: Fix reserved at offset in hca_cap register
  net/mlx5: Fix SFs kernel documentation error
  net/mlx5: Fix UAF in mlx5_eswitch_cleanup()
====================

Link: https://lore.kernel.org/r/20230623192907.39033-1-saeed@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Documentation/networking/device_drivers/ethernet/mellanox/mlx5/switchdev.rst

@@ -51,19 +51,21 @@ This will allow user to configure the SF before the SF have been fully probed,
 which will save time.
 
 Usage example:
-Create SF:
-$ devlink port add pci/0000:08:00.0 flavour pcisf pfnum 0 sfnum 11
-$ devlink port function set pci/0000:08:00.0/32768 \
-               hw_addr 00:00:00:00:00:11 state active
 
-Enable ETH auxiliary device:
-$ devlink dev param set auxiliary/mlx5_core.sf.1 \
-              name enable_eth value true cmode driverinit
+- Create SF::
 
-Now, in order to fully probe the SF, use devlink reload:
-$ devlink dev reload auxiliary/mlx5_core.sf.1
+    $ devlink port add pci/0000:08:00.0 flavour pcisf pfnum 0 sfnum 11
+    $ devlink port function set pci/0000:08:00.0/32768 hw_addr 00:00:00:00:00:11 state active
 
-mlx5 supports ETH,rdma and vdpa (vnet) auxiliary devices devlink params (see :ref:`Documentation/networking/devlink/devlink-params.rst`)
+- Enable ETH auxiliary device::
+
+    $ devlink dev param set auxiliary/mlx5_core.sf.1 name enable_eth value true cmode driverinit
+
+- Now, in order to fully probe the SF, use devlink reload::
+
+    $ devlink dev reload auxiliary/mlx5_core.sf.1
+
+mlx5 supports ETH,rdma and vdpa (vnet) auxiliary devices devlink params (see :ref:`Documentation/networking/devlink/devlink-params.rst <devlink_params_generic>`).
 
 mlx5 supports subfunction management using devlink port (see :ref:`Documentation/networking/devlink/devlink-port.rst <devlink_port>`) interface.
 

drivers/net/ethernet/mellanox/mlx5/core/dev.c

@@ -151,12 +151,6 @@ static bool is_ib_rep_supported(struct mlx5_core_dev *dev)
 	if (!is_eth_rep_supported(dev))
 		return false;
 
-	if (!MLX5_ESWITCH_MANAGER(dev))
-		return false;
-
-	if (!is_mdev_switchdev_mode(dev))
-		return false;
-
 	if (mlx5_core_mp_enabled(dev))
 		return false;
 

drivers/net/ethernet/mellanox/mlx5/core/en_rep.c

@@ -408,7 +408,7 @@ static int mlx5e_sqs2vport_add_peers_rules(struct mlx5_eswitch *esw, struct mlx5
 
 	mlx5_devcom_for_each_peer_entry(devcom, MLX5_DEVCOM_ESW_OFFLOADS,
 					peer_esw, tmp) {
-		int peer_rule_idx = mlx5_get_dev_index(peer_esw->dev);
+		u16 peer_rule_idx = MLX5_CAP_GEN(peer_esw->dev, vhca_id);
 		struct mlx5e_rep_sq_peer *sq_peer;
 		int err;
 
@@ -1581,7 +1581,7 @@ static void *mlx5e_vport_rep_get_proto_dev(struct mlx5_eswitch_rep *rep)
 static void mlx5e_vport_rep_event_unpair(struct mlx5_eswitch_rep *rep,
 					 struct mlx5_eswitch *peer_esw)
 {
-	int i = mlx5_get_dev_index(peer_esw->dev);
+	u16 i = MLX5_CAP_GEN(peer_esw->dev, vhca_id);
 	struct mlx5e_rep_priv *rpriv;
 	struct mlx5e_rep_sq *rep_sq;
 
@@ -1603,7 +1603,7 @@ static int mlx5e_vport_rep_event_pair(struct mlx5_eswitch *esw,
 				      struct mlx5_eswitch_rep *rep,
 				      struct mlx5_eswitch *peer_esw)
 {
-	int i = mlx5_get_dev_index(peer_esw->dev);
+	u16 i = MLX5_CAP_GEN(peer_esw->dev, vhca_id);
 	struct mlx5_flow_handle *flow_rule;
 	struct mlx5e_rep_sq_peer *sq_peer;
 	struct mlx5e_rep_priv *rpriv;

drivers/net/ethernet/mellanox/mlx5/core/eswitch.c

@@ -1751,16 +1751,14 @@ int mlx5_eswitch_init(struct mlx5_core_dev *dev)
 	if (!MLX5_VPORT_MANAGER(dev) && !MLX5_ESWITCH_MANAGER(dev))
 		return 0;
 
+	esw = kzalloc(sizeof(*esw), GFP_KERNEL);
+	if (!esw)
+		return -ENOMEM;
+
 	err = devl_params_register(priv_to_devlink(dev), mlx5_eswitch_params,
 				   ARRAY_SIZE(mlx5_eswitch_params));
 	if (err)
-		return err;
-
-	esw = kzalloc(sizeof(*esw), GFP_KERNEL);
-	if (!esw) {
-		err = -ENOMEM;
-		goto unregister_param;
-	}
+		goto free_esw;
 
 	esw->dev = dev;
 	esw->manager_vport = mlx5_eswitch_manager_vport(dev);
@@ -1821,10 +1819,10 @@ int mlx5_eswitch_init(struct mlx5_core_dev *dev)
 	if (esw->work_queue)
 		destroy_workqueue(esw->work_queue);
 	debugfs_remove_recursive(esw->debugfs_root);
-	kfree(esw);
-unregister_param:
 	devl_params_unregister(priv_to_devlink(dev), mlx5_eswitch_params,
 			       ARRAY_SIZE(mlx5_eswitch_params));
+free_esw:
+	kfree(esw);
 	return err;
 }
 
@@ -1848,9 +1846,9 @@ void mlx5_eswitch_cleanup(struct mlx5_eswitch *esw)
 	esw_offloads_cleanup(esw);
 	mlx5_esw_vports_cleanup(esw);
 	debugfs_remove_recursive(esw->debugfs_root);
-	kfree(esw);
 	devl_params_unregister(priv_to_devlink(esw->dev), mlx5_eswitch_params,
 			       ARRAY_SIZE(mlx5_eswitch_params));
+	kfree(esw);
 }
 
 /* Vport Administration */
@@ -1910,12 +1908,6 @@ int mlx5_eswitch_set_vport_mac(struct mlx5_eswitch *esw,
 
 static bool mlx5_esw_check_port_type(struct mlx5_eswitch *esw, u16 vport_num, xa_mark_t mark)
 {
-	struct mlx5_vport *vport;
-
-	vport = mlx5_eswitch_get_vport(esw, vport_num);
-	if (IS_ERR(vport))
-		return false;
-
 	return xa_get_mark(&esw->vports, vport_num, mark);
 }
 

drivers/net/ethernet/mellanox/mlx5/core/eswitch.h

@@ -353,7 +353,7 @@ struct mlx5_eswitch {
 		u32             large_group_num;
 	}  params;
 	struct blocking_notifier_head n_head;
-	bool paired[MLX5_MAX_PORTS];
+	struct xarray paired;
 };
 
 void esw_offloads_disable(struct mlx5_eswitch *esw);

drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c

@@ -1069,6 +1069,9 @@ static int esw_add_fdb_peer_miss_rules(struct mlx5_eswitch *esw,
 	void *misc;
 	int err;
 
+	if (!MLX5_VPORT_MANAGER(esw->dev) && !mlx5_core_is_ecpf_esw_manager(esw->dev))
+		return 0;
+
 	spec = kvzalloc(sizeof(*spec), GFP_KERNEL);
 	if (!spec)
 		return -ENOMEM;
@@ -1177,11 +1180,14 @@ static int esw_add_fdb_peer_miss_rules(struct mlx5_eswitch *esw,
 static void esw_del_fdb_peer_miss_rules(struct mlx5_eswitch *esw,
 					struct mlx5_core_dev *peer_dev)
 {
+	u16 peer_index = mlx5_get_dev_index(peer_dev);
 	struct mlx5_flow_handle **flows;
 	struct mlx5_vport *vport;
 	unsigned long i;
 
-	flows = esw->fdb_table.offloads.peer_miss_rules[mlx5_get_dev_index(peer_dev)];
+	flows = esw->fdb_table.offloads.peer_miss_rules[peer_index];
+	if (!flows)
+		return;
 
 	if (mlx5_core_ec_sriov_enabled(esw->dev)) {
 		mlx5_esw_for_each_ec_vf_vport(esw, i, vport, mlx5_core_max_ec_vfs(esw->dev)) {
@@ -1206,7 +1212,9 @@ static void esw_del_fdb_peer_miss_rules(struct mlx5_eswitch *esw,
 		vport = mlx5_eswitch_get_vport(esw, MLX5_VPORT_PF);
 		mlx5_del_flow_rules(flows[vport->index]);
 	}
+
 	kvfree(flows);
+	esw->fdb_table.offloads.peer_miss_rules[peer_index] = NULL;
 }
 
 static int esw_add_fdb_miss_rule(struct mlx5_eswitch *esw)
@@ -1896,7 +1904,6 @@ static int esw_create_vport_rx_group(struct mlx5_eswitch *esw)
 	if (!flow_group_in)
 		return -ENOMEM;
 
-	/* create vport rx group */
 	mlx5_esw_set_flow_group_source_port(esw, flow_group_in, 0);
 
 	MLX5_SET(create_flow_group_in, flow_group_in, start_flow_index, 0);
@@ -2543,13 +2550,13 @@ static int __esw_set_master_egress_rule(struct mlx5_core_dev *master,
 	return err;
 }
 
-static int esw_master_egress_create_resources(struct mlx5_flow_namespace *egress_ns,
+static int esw_master_egress_create_resources(struct mlx5_eswitch *esw,
+					      struct mlx5_flow_namespace *egress_ns,
 					      struct mlx5_vport *vport, size_t count)
 {
 	int inlen = MLX5_ST_SZ_BYTES(create_flow_group_in);
 	struct mlx5_flow_table_attr ft_attr = {
 		.max_fte = count, .prio = 0, .level = 0,
-		.flags = MLX5_FLOW_TABLE_OTHER_VPORT,
 	};
 	struct mlx5_flow_table *acl;
 	struct mlx5_flow_group *g;
@@ -2564,6 +2571,9 @@ static int esw_master_egress_create_resources(struct mlx5_flow_namespace *egress
 	if (!flow_group_in)
 		return -ENOMEM;
 
+	if (vport->vport || mlx5_core_is_ecpf(esw->dev))
+		ft_attr.flags = MLX5_FLOW_TABLE_OTHER_VPORT;
+
 	acl = mlx5_create_vport_flow_table(egress_ns, &ft_attr, vport->vport);
 	if (IS_ERR(acl)) {
 		err = PTR_ERR(acl);
@@ -2608,8 +2618,12 @@ static int esw_master_egress_create_resources(struct mlx5_flow_namespace *egress
 
 static void esw_master_egress_destroy_resources(struct mlx5_vport *vport)
 {
+	if (!xa_empty(&vport->egress.offloads.bounce_rules))
+		return;
 	mlx5_destroy_flow_group(vport->egress.offloads.bounce_grp);
+	vport->egress.offloads.bounce_grp = NULL;
 	mlx5_destroy_flow_table(vport->egress.acl);
+	vport->egress.acl = NULL;
 }
 
 static int esw_set_master_egress_rule(struct mlx5_core_dev *master,
@@ -2634,7 +2648,7 @@ static int esw_set_master_egress_rule(struct mlx5_core_dev *master,
 	if (vport->egress.acl && vport->egress.type != VPORT_EGRESS_ACL_TYPE_SHARED_FDB)
 		return 0;
 
-	err = esw_master_egress_create_resources(egress_ns, vport, count);
+	err = esw_master_egress_create_resources(esw, egress_ns, vport, count);
 	if (err)
 		return err;
 
@@ -2799,15 +2813,21 @@ static int mlx5_esw_offloads_devcom_event(int event,
 	struct mlx5_eswitch *esw = my_data;
 	struct mlx5_devcom *devcom = esw->dev->priv.devcom;
 	struct mlx5_eswitch *peer_esw = event_data;
+	u16 esw_i, peer_esw_i;
+	bool esw_paired;
 	int err;
 
+	peer_esw_i = MLX5_CAP_GEN(peer_esw->dev, vhca_id);
+	esw_i = MLX5_CAP_GEN(esw->dev, vhca_id);
+	esw_paired = !!xa_load(&esw->paired, peer_esw_i);
+
 	switch (event) {
 	case ESW_OFFLOADS_DEVCOM_PAIR:
 		if (mlx5_eswitch_vport_match_metadata_enabled(esw) !=
 		    mlx5_eswitch_vport_match_metadata_enabled(peer_esw))
 			break;
 
-		if (esw->paired[mlx5_get_dev_index(peer_esw->dev)])
+		if (esw_paired)
 			break;
 
 		err = mlx5_esw_offloads_set_ns_peer(esw, peer_esw, true);
@@ -2821,23 +2841,29 @@ static int mlx5_esw_offloads_devcom_event(int event,
 		if (err)
 			goto err_pair;
 
-		esw->paired[mlx5_get_dev_index(peer_esw->dev)] = true;
-		peer_esw->paired[mlx5_get_dev_index(esw->dev)] = true;
+		err = xa_insert(&esw->paired, peer_esw_i, peer_esw, GFP_KERNEL);
+		if (err)
+			goto err_xa;
+
+		err = xa_insert(&peer_esw->paired, esw_i, esw, GFP_KERNEL);
+		if (err)
+			goto err_peer_xa;
+
 		esw->num_peers++;
 		peer_esw->num_peers++;
 		mlx5_devcom_comp_set_ready(devcom, MLX5_DEVCOM_ESW_OFFLOADS, true);
 		break;
 
 	case ESW_OFFLOADS_DEVCOM_UNPAIR:
-		if (!esw->paired[mlx5_get_dev_index(peer_esw->dev)])
+		if (!esw_paired)
 			break;
 
 		peer_esw->num_peers--;
 		esw->num_peers--;
 		if (!esw->num_peers && !peer_esw->num_peers)
 			mlx5_devcom_comp_set_ready(devcom, MLX5_DEVCOM_ESW_OFFLOADS, false);
-		esw->paired[mlx5_get_dev_index(peer_esw->dev)] = false;
-		peer_esw->paired[mlx5_get_dev_index(esw->dev)] = false;
+		xa_erase(&peer_esw->paired, esw_i);
+		xa_erase(&esw->paired, peer_esw_i);
 		mlx5_esw_offloads_unpair(peer_esw, esw);
 		mlx5_esw_offloads_unpair(esw, peer_esw);
 		mlx5_esw_offloads_set_ns_peer(esw, peer_esw, false);
@@ -2846,6 +2872,10 @@ static int mlx5_esw_offloads_devcom_event(int event,
 
 	return 0;
 
+err_peer_xa:
+	xa_erase(&esw->paired, peer_esw_i);
+err_xa:
+	mlx5_esw_offloads_unpair(peer_esw, esw);
 err_pair:
 	mlx5_esw_offloads_unpair(esw, peer_esw);
 err_peer:
@@ -2868,9 +2898,10 @@ void mlx5_esw_offloads_devcom_init(struct mlx5_eswitch *esw)
 	if (!MLX5_CAP_ESW(esw->dev, merged_eswitch))
 		return;
 
-	if (!mlx5_is_lag_supported(esw->dev))
+	if (!mlx5_lag_is_supported(esw->dev))
 		return;
 
+	xa_init(&esw->paired);
 	mlx5_devcom_register_component(devcom,
 				       MLX5_DEVCOM_ESW_OFFLOADS,
 				       mlx5_esw_offloads_devcom_event,
@@ -2890,14 +2921,15 @@ void mlx5_esw_offloads_devcom_cleanup(struct mlx5_eswitch *esw)
 	if (!MLX5_CAP_ESW(esw->dev, merged_eswitch))
 		return;
 
-	if (!mlx5_is_lag_supported(esw->dev))
+	if (!mlx5_lag_is_supported(esw->dev))
 		return;
 
 	mlx5_devcom_send_event(devcom, MLX5_DEVCOM_ESW_OFFLOADS,
 			       ESW_OFFLOADS_DEVCOM_UNPAIR,
 			       ESW_OFFLOADS_DEVCOM_UNPAIR, esw);
 
 	mlx5_devcom_unregister_component(devcom, MLX5_DEVCOM_ESW_OFFLOADS);
+	xa_destroy(&esw->paired);
 }
 
 bool mlx5_esw_vport_match_metadata_supported(const struct mlx5_eswitch *esw)
@@ -3929,9 +3961,6 @@ static int mlx5_esw_query_vport_vhca_id(struct mlx5_eswitch *esw, u16 vport_num,
 	int err;
 
 	*vhca_id = 0;
-	if (mlx5_esw_is_manager_vport(esw, vport_num) ||
-	    !MLX5_CAP_GEN(esw->dev, vhca_resource_manager))
-		return -EPERM;
 
 	query_ctx = kzalloc(query_out_sz, GFP_KERNEL);
 	if (!query_ctx)

drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c

@@ -327,7 +327,7 @@ static bool mlx5_is_reset_now_capable(struct mlx5_core_dev *dev)
 
 	if (!MLX5_CAP_GEN(dev, fast_teardown)) {
 		mlx5_core_warn(dev, "fast teardown is not supported by firmware\n");
-		return -EOPNOTSUPP;
+		return false;
 	}
 
 	err = pci_read_config_word(dev->pdev, PCI_DEVICE_ID, &dev_id);

drivers/net/ethernet/mellanox/mlx5/core/lag/lag.c

@@ -1268,14 +1268,21 @@ void mlx5_lag_remove_mdev(struct mlx5_core_dev *dev)
 	mlx5_ldev_put(ldev);
 }
 
+bool mlx5_lag_is_supported(struct mlx5_core_dev *dev)
+{
+	if (!MLX5_CAP_GEN(dev, vport_group_manager) ||
+	    !MLX5_CAP_GEN(dev, lag_master) ||
+	    MLX5_CAP_GEN(dev, num_lag_ports) < 2 ||
+	    MLX5_CAP_GEN(dev, num_lag_ports) > MLX5_MAX_PORTS)
+		return false;
+	return true;
+}
+
 void mlx5_lag_add_mdev(struct mlx5_core_dev *dev)
 {
 	int err;
 
-	if (!MLX5_CAP_GEN(dev, vport_group_manager) ||
-	    !MLX5_CAP_GEN(dev, lag_master) ||
-	    (MLX5_CAP_GEN(dev, num_lag_ports) > MLX5_MAX_PORTS ||
-	     MLX5_CAP_GEN(dev, num_lag_ports) <= 1))
+	if (!mlx5_lag_is_supported(dev))
 		return;
 
 recheck:

drivers/net/ethernet/mellanox/mlx5/core/lag/lag.h

@@ -74,15 +74,7 @@ struct mlx5_lag {
 	struct lag_mpesw	  lag_mpesw;
 };
 
-static inline bool mlx5_is_lag_supported(struct mlx5_core_dev *dev)
-{
-	if (!MLX5_CAP_GEN(dev, vport_group_manager) ||
-	    !MLX5_CAP_GEN(dev, lag_master) ||
-	    MLX5_CAP_GEN(dev, num_lag_ports) < 2 ||
-	    MLX5_CAP_GEN(dev, num_lag_ports) > MLX5_MAX_PORTS)
-		return false;
-	return true;
-}
+bool mlx5_lag_is_supported(struct mlx5_core_dev *dev);
 
 static inline struct mlx5_lag *
 mlx5_lag_dev(struct mlx5_core_dev *dev)