diff options
Diffstat (limited to 'drivers/net/ethernet/mellanox')
40 files changed, 2343 insertions, 800 deletions
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_cq.c b/drivers/net/ethernet/mellanox/mlx4/en_cq.c index 85fe17e4dcfb..87d1f4d2a77b 100644 --- a/drivers/net/ethernet/mellanox/mlx4/en_cq.c +++ b/drivers/net/ethernet/mellanox/mlx4/en_cq.c @@ -208,12 +208,10 @@ int mlx4_en_set_cq_moder(struct mlx4_en_priv *priv, struct mlx4_en_cq *cq) cq->moder_cnt, cq->moder_time); } -int mlx4_en_arm_cq(struct mlx4_en_priv *priv, struct mlx4_en_cq *cq) +void mlx4_en_arm_cq(struct mlx4_en_priv *priv, struct mlx4_en_cq *cq) { mlx4_cq_arm(&cq->mcq, MLX4_CQ_DB_REQ_NOT, priv->mdev->uar_map, &priv->mdev->uar_lock); - - return 0; } diff --git a/drivers/net/ethernet/mellanox/mlx4/en_main.c b/drivers/net/ethernet/mellanox/mlx4/en_main.c index 2b0cbca4beb5..686e18de9a97 100644 --- a/drivers/net/ethernet/mellanox/mlx4/en_main.c +++ b/drivers/net/ethernet/mellanox/mlx4/en_main.c @@ -147,7 +147,7 @@ void mlx4_en_update_loopback_state(struct net_device *dev, mutex_unlock(&priv->mdev->state_lock); } -static int mlx4_en_get_profile(struct mlx4_en_dev *mdev) +static void mlx4_en_get_profile(struct mlx4_en_dev *mdev) { struct mlx4_en_profile *params = &mdev->profile; int i; @@ -176,8 +176,6 @@ static int mlx4_en_get_profile(struct mlx4_en_dev *mdev) params->prof[i].rss_rings = 0; params->prof[i].inline_thold = inline_thold; } - - return 0; } static void *mlx4_en_get_netdev(struct mlx4_dev *dev, void *ctx, u8 port) @@ -309,10 +307,7 @@ static void *mlx4_en_add(struct mlx4_dev *dev) } /* Build device profile according to supplied module parameters */ - if (mlx4_en_get_profile(mdev)) { - mlx4_err(mdev, "Bad module parameters, aborting\n"); - goto err_mr; - } + mlx4_en_get_profile(mdev); /* Configure which ports to start according to module parameters */ mdev->port_cnt = 0; diff --git a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c index 3a291fc1780a..6e67ca7aa7f5 100644 --- a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c +++ b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c @@ -130,19 +130,20 @@ out: return err; } -static int __mlx4_en_setup_tc(struct net_device *dev, u32 handle, - u32 chain_index, __be16 proto, - struct tc_to_netdev *tc) +static int __mlx4_en_setup_tc(struct net_device *dev, enum tc_setup_type type, + void *type_data) { - if (tc->type != TC_SETUP_MQPRIO) - return -EINVAL; + struct tc_mqprio_qopt *mqprio = type_data; + + if (type != TC_SETUP_MQPRIO) + return -EOPNOTSUPP; - if (tc->mqprio->num_tc && tc->mqprio->num_tc != MLX4_EN_NUM_UP_HIGH) + if (mqprio->num_tc && mqprio->num_tc != MLX4_EN_NUM_UP_HIGH) return -EINVAL; - tc->mqprio->hw = TC_MQPRIO_HW_OFFLOAD_TCS; + mqprio->hw = TC_MQPRIO_HW_OFFLOAD_TCS; - return mlx4_en_alloc_tx_queue_per_tc(dev, tc->mqprio->num_tc); + return mlx4_en_alloc_tx_queue_per_tc(dev, mqprio->num_tc); } #ifdef CONFIG_RFS_ACCEL diff --git a/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h b/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h index d350b2158104..fdb3ad0cbe54 100644 --- a/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h +++ b/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h @@ -685,7 +685,7 @@ int mlx4_en_activate_cq(struct mlx4_en_priv *priv, struct mlx4_en_cq *cq, int cq_idx); void mlx4_en_deactivate_cq(struct mlx4_en_priv *priv, struct mlx4_en_cq *cq); int mlx4_en_set_cq_moder(struct mlx4_en_priv *priv, struct mlx4_en_cq *cq); -int mlx4_en_arm_cq(struct mlx4_en_priv *priv, struct mlx4_en_cq *cq); +void mlx4_en_arm_cq(struct mlx4_en_priv *priv, struct mlx4_en_cq *cq); void mlx4_en_tx_irq(struct mlx4_cq *mcq); u16 mlx4_en_select_queue(struct net_device *dev, struct sk_buff *skb, diff --git a/drivers/net/ethernet/mellanox/mlx5/core/Kconfig b/drivers/net/ethernet/mellanox/mlx5/core/Kconfig index 5aee05992f27..fdaef00465d7 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/Kconfig +++ b/drivers/net/ethernet/mellanox/mlx5/core/Kconfig @@ -34,6 +34,27 @@ config MLX5_CORE_EN ---help--- Ethernet support in Mellanox Technologies ConnectX-4 NIC. +config MLX5_MPFS + bool "Mellanox Technologies MLX5 MPFS support" + depends on MLX5_CORE_EN + default y + ---help--- + Mellanox Technologies Ethernet Multi-Physical Function Switch (MPFS) + support in ConnectX NIC. MPFs is required for when multi-PF configuration + is enabled to allow passing user configured unicast MAC addresses to the + requesting PF. + +config MLX5_ESWITCH + bool "Mellanox Technologies MLX5 SRIOV E-Switch support" + depends on MLX5_CORE_EN + default y + ---help--- + Mellanox Technologies Ethernet SRIOV E-Switch support in ConnectX NIC. + E-Switch provides internal SRIOV packet steering and switching for the + enabled VFs and PF in two available modes: + Legacy SRIOV mode (L2 mac vlan steering based). + Switchdev mode (eswitch offloads). + config MLX5_CORE_EN_DCB bool "Data Center Bridging (DCB) Support" default y diff --git a/drivers/net/ethernet/mellanox/mlx5/core/Makefile b/drivers/net/ethernet/mellanox/mlx5/core/Makefile index 9d17e4e76d3a..22ed657d263a 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/Makefile +++ b/drivers/net/ethernet/mellanox/mlx5/core/Makefile @@ -11,10 +11,13 @@ mlx5_core-$(CONFIG_MLX5_ACCEL) += accel/ipsec.o mlx5_core-$(CONFIG_MLX5_FPGA) += fpga/cmd.o fpga/core.o fpga/conn.o fpga/sdk.o \ fpga/ipsec.o -mlx5_core-$(CONFIG_MLX5_CORE_EN) += eswitch.o eswitch_offloads.o \ - en_main.o en_common.o en_fs.o en_ethtool.o en_tx.o \ - en_rx.o en_rx_am.o en_txrx.o en_clock.o vxlan.o \ - en_tc.o en_arfs.o en_rep.o en_fs_ethtool.o en_selftest.o +mlx5_core-$(CONFIG_MLX5_CORE_EN) += en_main.o en_common.o en_fs.o en_ethtool.o \ + en_tx.o en_rx.o en_rx_am.o en_txrx.o en_clock.o vxlan.o \ + en_arfs.o en_fs_ethtool.o en_selftest.o + +mlx5_core-$(CONFIG_MLX5_MPFS) += lib/mpfs.o + +mlx5_core-$(CONFIG_MLX5_ESWITCH) += eswitch.o eswitch_offloads.o en_rep.o en_tc.o mlx5_core-$(CONFIG_MLX5_CORE_EN_DCB) += en_dcbnl.o diff --git a/drivers/net/ethernet/mellanox/mlx5/core/dev.c b/drivers/net/ethernet/mellanox/mlx5/core/dev.c index a62f4b6a21a5..ff60cf7342ca 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/dev.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/dev.c @@ -45,11 +45,70 @@ struct mlx5_device_context { unsigned long state; }; +struct mlx5_delayed_event { + struct list_head list; + struct mlx5_core_dev *dev; + enum mlx5_dev_event event; + unsigned long param; +}; + enum { MLX5_INTERFACE_ADDED, MLX5_INTERFACE_ATTACHED, }; +static void add_delayed_event(struct mlx5_priv *priv, + struct mlx5_core_dev *dev, + enum mlx5_dev_event event, + unsigned long param) +{ + struct mlx5_delayed_event *delayed_event; + + delayed_event = kzalloc(sizeof(*delayed_event), GFP_ATOMIC); + if (!delayed_event) { + mlx5_core_err(dev, "event %d is missed\n", event); + return; + } + + mlx5_core_dbg(dev, "Accumulating event %d\n", event); + delayed_event->dev = dev; + delayed_event->event = event; + delayed_event->param = param; + list_add_tail(&delayed_event->list, &priv->waiting_events_list); +} + +static void fire_delayed_event_locked(struct mlx5_device_context *dev_ctx, + struct mlx5_core_dev *dev, + struct mlx5_priv *priv) +{ + struct mlx5_delayed_event *de; + struct mlx5_delayed_event *n; + + /* stop delaying events */ + priv->is_accum_events = false; + + /* fire all accumulated events before new event comes */ + list_for_each_entry_safe(de, n, &priv->waiting_events_list, list) { + dev_ctx->intf->event(dev, dev_ctx->context, de->event, de->param); + list_del(&de->list); + kfree(de); + } +} + +static void cleanup_delayed_evets(struct mlx5_priv *priv) +{ + struct mlx5_delayed_event *de; + struct mlx5_delayed_event *n; + + spin_lock_irq(&priv->ctx_lock); + priv->is_accum_events = false; + list_for_each_entry_safe(de, n, &priv->waiting_events_list, list) { + list_del(&de->list); + kfree(de); + } + spin_unlock_irq(&priv->ctx_lock); +} + void mlx5_add_device(struct mlx5_interface *intf, struct mlx5_priv *priv) { struct mlx5_device_context *dev_ctx; @@ -63,6 +122,12 @@ void mlx5_add_device(struct mlx5_interface *intf, struct mlx5_priv *priv) return; dev_ctx->intf = intf; + /* accumulating events that can come after mlx5_ib calls to + * ib_register_device, till adding that interface to the events list. + */ + + priv->is_accum_events = true; + dev_ctx->context = intf->add(dev); set_bit(MLX5_INTERFACE_ADDED, &dev_ctx->state); if (intf->attach) @@ -71,6 +136,9 @@ void mlx5_add_device(struct mlx5_interface *intf, struct mlx5_priv *priv) if (dev_ctx->context) { spin_lock_irq(&priv->ctx_lock); list_add_tail(&dev_ctx->list, &priv->ctx_list); + + fire_delayed_event_locked(dev_ctx, dev, priv); + #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING if (dev_ctx->intf->pfault) { if (priv->pfault) { @@ -84,6 +152,8 @@ void mlx5_add_device(struct mlx5_interface *intf, struct mlx5_priv *priv) spin_unlock_irq(&priv->ctx_lock); } else { kfree(dev_ctx); + /* delete all accumulated events */ + cleanup_delayed_evets(priv); } } @@ -341,6 +411,9 @@ void mlx5_core_event(struct mlx5_core_dev *dev, enum mlx5_dev_event event, spin_lock_irqsave(&priv->ctx_lock, flags); + if (priv->is_accum_events) + add_delayed_event(priv, dev, event, param); + list_for_each_entry(dev_ctx, &priv->ctx_list, list) if (dev_ctx->intf->event) dev_ctx->intf->event(dev, dev_ctx->context, event, param); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c b/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c index dfccb5305e9c..eecbc6d4f51f 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c @@ -36,6 +36,7 @@ #include <linux/tcp.h> #include <linux/mlx5/fs.h> #include "en.h" +#include "lib/mpfs.h" static int mlx5e_add_l2_flow_rule(struct mlx5e_priv *priv, struct mlx5e_l2_rule *ai, int type); @@ -65,6 +66,7 @@ struct mlx5e_l2_hash_node { struct hlist_node hlist; u8 action; struct mlx5e_l2_rule ai; + bool mpfs; }; static inline int mlx5e_hash_l2(u8 *addr) @@ -362,17 +364,30 @@ static void mlx5e_del_vlan_rules(struct mlx5e_priv *priv) static void mlx5e_execute_l2_action(struct mlx5e_priv *priv, struct mlx5e_l2_hash_node *hn) { - switch (hn->action) { + u8 action = hn->action; + int l2_err = 0; + + switch (action) { case MLX5E_ACTION_ADD: mlx5e_add_l2_flow_rule(priv, &hn->ai, MLX5E_FULLMATCH); + if (!is_multicast_ether_addr(hn->ai.addr)) { + l2_err = mlx5_mpfs_add_mac(priv->mdev, hn->ai.addr); + hn->mpfs = !l2_err; + } hn->action = MLX5E_ACTION_NONE; break; case MLX5E_ACTION_DEL: + if (!is_multicast_ether_addr(hn->ai.addr) && hn->mpfs) + l2_err = mlx5_mpfs_del_mac(priv->mdev, hn->ai.addr); mlx5e_del_l2_flow_rule(priv, &hn->ai); mlx5e_del_l2_from_hash(hn); break; } + + if (l2_err) + netdev_warn(priv->netdev, "MPFS, failed to %s mac %pM, err(%d)\n", + action == MLX5E_ACTION_ADD ? "add" : "del", hn->ai.addr, l2_err); } static void mlx5e_sync_netdev_addr(struct mlx5e_priv *priv) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c index 57f31fa478ce..981f8415b546 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c @@ -2581,12 +2581,6 @@ static void mlx5e_build_channels_tx_maps(struct mlx5e_priv *priv) } } -static bool mlx5e_is_eswitch_vport_mngr(struct mlx5_core_dev *mdev) -{ - return (MLX5_CAP_GEN(mdev, vport_group_manager) && - MLX5_CAP_GEN(mdev, port_type) == MLX5_CAP_PORT_TYPE_ETH); -} - void mlx5e_activate_priv_channels(struct mlx5e_priv *priv) { int num_txqs = priv->channels.num * priv->channels.params.num_tc; @@ -2600,7 +2594,7 @@ void mlx5e_activate_priv_channels(struct mlx5e_priv *priv) mlx5e_activate_channels(&priv->channels); netif_tx_start_all_queues(priv->netdev); - if (mlx5e_is_eswitch_vport_mngr(priv->mdev)) + if (MLX5_VPORT_MANAGER(priv->mdev)) mlx5e_add_sqs_fwd_rules(priv); mlx5e_wait_channels_min_rx_wqes(&priv->channels); @@ -2611,7 +2605,7 @@ void mlx5e_deactivate_priv_channels(struct mlx5e_priv *priv) { mlx5e_redirect_rqts_to_drop(priv); - if (mlx5e_is_eswitch_vport_mngr(priv->mdev)) + if (MLX5_VPORT_MANAGER(priv->mdev)) mlx5e_remove_sqs_fwd_rules(priv); /* FIXME: This is a W/A only for tx timeout watch dog false alarm when @@ -2998,12 +2992,16 @@ static int mlx5e_modify_channels_vsd(struct mlx5e_channels *chs, bool vsd) return 0; } -static int mlx5e_setup_tc(struct net_device *netdev, u8 tc) +static int mlx5e_setup_tc_mqprio(struct net_device *netdev, + struct tc_mqprio_qopt *mqprio) { struct mlx5e_priv *priv = netdev_priv(netdev); struct mlx5e_channels new_channels = {}; + u8 tc = mqprio->num_tc; int err = 0; + mqprio->hw = TC_MQPRIO_HW_OFFLOAD_TCS; + if (tc && tc != MLX5E_MAX_NUM_TC) return -EINVAL; @@ -3027,39 +3025,42 @@ out: return err; } -static int mlx5e_ndo_setup_tc(struct net_device *dev, u32 handle, - u32 chain_index, __be16 proto, - struct tc_to_netdev *tc) +#ifdef CONFIG_MLX5_ESWITCH +static int mlx5e_setup_tc_cls_flower(struct net_device *dev, + struct tc_cls_flower_offload *cls_flower) { struct mlx5e_priv *priv = netdev_priv(dev); - if (TC_H_MAJ(handle) != TC_H_MAJ(TC_H_INGRESS)) - goto mqprio; + if (TC_H_MAJ(cls_flower->common.handle) != TC_H_MAJ(TC_H_INGRESS) || + cls_flower->common.chain_index) + return -EOPNOTSUPP; - if (chain_index) + switch (cls_flower->command) { + case TC_CLSFLOWER_REPLACE: + return mlx5e_configure_flower(priv, cls_flower); + case TC_CLSFLOWER_DESTROY: + return mlx5e_delete_flower(priv, cls_flower); + case TC_CLSFLOWER_STATS: + return mlx5e_stats_flower(priv, cls_flower); + default: return -EOPNOTSUPP; + } +} +#endif - switch (tc->type) { +static int mlx5e_setup_tc(struct net_device *dev, enum tc_setup_type type, + void *type_data) +{ + switch (type) { +#ifdef CONFIG_MLX5_ESWITCH case TC_SETUP_CLSFLOWER: - switch (tc->cls_flower->command) { - case TC_CLSFLOWER_REPLACE: - return mlx5e_configure_flower(priv, proto, tc->cls_flower); - case TC_CLSFLOWER_DESTROY: - return mlx5e_delete_flower(priv, tc->cls_flower); - case TC_CLSFLOWER_STATS: - return mlx5e_stats_flower(priv, tc->cls_flower); - } + return mlx5e_setup_tc_cls_flower(dev, type_data); +#endif + case TC_SETUP_MQPRIO: + return mlx5e_setup_tc_mqprio(dev, type_data); default: return -EOPNOTSUPP; } - -mqprio: - if (tc->type != TC_SETUP_MQPRIO) - return -EINVAL; - - tc->mqprio->hw = TC_MQPRIO_HW_OFFLOAD_TCS; - - return mlx5e_setup_tc(dev, tc->mqprio->num_tc); } static void @@ -3356,6 +3357,7 @@ static int mlx5e_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd) } } +#ifdef CONFIG_MLX5_ESWITCH static int mlx5e_set_vf_mac(struct net_device *dev, int vf, u8 *mac) { struct mlx5e_priv *priv = netdev_priv(dev); @@ -3458,6 +3460,7 @@ static int mlx5e_get_vf_stats(struct net_device *dev, return mlx5_eswitch_get_vport_stats(mdev->priv.eswitch, vf + 1, vf_stats); } +#endif static void mlx5e_add_vxlan_port(struct net_device *netdev, struct udp_tunnel_info *ti) @@ -3691,11 +3694,11 @@ static void mlx5e_netpoll(struct net_device *dev) } #endif -static const struct net_device_ops mlx5e_netdev_ops_basic = { +static const struct net_device_ops mlx5e_netdev_ops = { .ndo_open = mlx5e_open, .ndo_stop = mlx5e_close, .ndo_start_xmit = mlx5e_xmit, - .ndo_setup_tc = mlx5e_ndo_setup_tc, + .ndo_setup_tc = mlx5e_setup_tc, .ndo_select_queue = mlx5e_select_queue, .ndo_get_stats64 = mlx5e_get_stats, .ndo_set_rx_mode = mlx5e_set_rx_mode, @@ -3706,6 +3709,9 @@ static const struct net_device_ops mlx5e_netdev_ops_basic = { .ndo_change_mtu = mlx5e_change_mtu, .ndo_do_ioctl = mlx5e_ioctl, .ndo_set_tx_maxrate = mlx5e_set_tx_maxrate, + .ndo_udp_tunnel_add = mlx5e_add_vxlan_port, + .ndo_udp_tunnel_del = mlx5e_del_vxlan_port, + .ndo_features_check = mlx5e_features_check, #ifdef CONFIG_RFS_ACCEL .ndo_rx_flow_steer = mlx5e_rx_flow_steer, #endif @@ -3714,29 +3720,8 @@ static const struct net_device_ops mlx5e_netdev_ops_basic = { #ifdef CONFIG_NET_POLL_CONTROLLER .ndo_poll_controller = mlx5e_netpoll, #endif -}; - -static const struct net_device_ops mlx5e_netdev_ops_sriov = { - .ndo_open = mlx5e_open, - .ndo_stop = mlx5e_close, - .ndo_start_xmit = mlx5e_xmit, - .ndo_setup_tc = mlx5e_ndo_setup_tc, - .ndo_select_queue = mlx5e_select_queue, - .ndo_get_stats64 = mlx5e_get_stats, - .ndo_set_rx_mode = mlx5e_set_rx_mode, - .ndo_set_mac_address = mlx5e_set_mac, - .ndo_vlan_rx_add_vid = mlx5e_vlan_rx_add_vid, - .ndo_vlan_rx_kill_vid = mlx5e_vlan_rx_kill_vid, - .ndo_set_features = mlx5e_set_features, - .ndo_change_mtu = mlx5e_change_mtu, - .ndo_do_ioctl = mlx5e_ioctl, - .ndo_udp_tunnel_add = mlx5e_add_vxlan_port, - .ndo_udp_tunnel_del = mlx5e_del_vxlan_port, - .ndo_set_tx_maxrate = mlx5e_set_tx_maxrate, - .ndo_features_check = mlx5e_features_check, -#ifdef CONFIG_RFS_ACCEL - .ndo_rx_flow_steer = mlx5e_rx_flow_steer, -#endif +#ifdef CONFIG_MLX5_ESWITCH + /* SRIOV E-Switch NDOs */ .ndo_set_vf_mac = mlx5e_set_vf_mac, .ndo_set_vf_vlan = mlx5e_set_vf_vlan, .ndo_set_vf_spoofchk = mlx5e_set_vf_spoofchk, @@ -3745,13 +3730,9 @@ static const struct net_device_ops mlx5e_netdev_ops_sriov = { .ndo_get_vf_config = mlx5e_get_vf_config, .ndo_set_vf_link_state = mlx5e_set_vf_link_state, .ndo_get_vf_stats = mlx5e_get_vf_stats, - .ndo_tx_timeout = mlx5e_tx_timeout, - .ndo_xdp = mlx5e_xdp, -#ifdef CONFIG_NET_POLL_CONTROLLER - .ndo_poll_controller = mlx5e_netpoll, -#endif .ndo_has_offload_stats = mlx5e_has_offload_stats, .ndo_get_offload_stats = mlx5e_get_offload_stats, +#endif }; static int mlx5e_check_required_hca_cap(struct mlx5_core_dev *mdev) @@ -3981,9 +3962,11 @@ static void mlx5e_set_netdev_dev_addr(struct net_device *netdev) } } +#if IS_ENABLED(CONFIG_NET_SWITCHDEV) && IS_ENABLED(CONFIG_MLX5_ESWITCH) static const struct switchdev_ops mlx5e_switchdev_ops = { .switchdev_port_attr_get = mlx5e_attr_get, }; +#endif static void mlx5e_build_nic_netdev(struct net_device *netdev) { @@ -3994,15 +3977,12 @@ static void mlx5e_build_nic_netdev(struct net_device *netdev) SET_NETDEV_DEV(netdev, &mdev->pdev->dev); - if (MLX5_CAP_GEN(mdev, vport_group_manager)) { - netdev->netdev_ops = &mlx5e_netdev_ops_sriov; + netdev->netdev_ops = &mlx5e_netdev_ops; + #ifdef CONFIG_MLX5_CORE_EN_DCB - if (MLX5_CAP_GEN(mdev, qos)) - netdev->dcbnl_ops = &mlx5e_dcbnl_ops; + if (MLX5_CAP_GEN(mdev, vport_group_manager) && MLX5_CAP_GEN(mdev, qos)) + netdev->dcbnl_ops = &mlx5e_dcbnl_ops; #endif - } else { - netdev->netdev_ops = &mlx5e_netdev_ops_basic; - } netdev->watchdog_timeo = 15 * HZ; @@ -4074,8 +4054,8 @@ static void mlx5e_build_nic_netdev(struct net_device *netdev) mlx5e_set_netdev_dev_addr(netdev); -#ifdef CONFIG_NET_SWITCHDEV - if (MLX5_CAP_GEN(mdev, vport_group_manager)) +#if IS_ENABLED(CONFIG_NET_SWITCHDEV) && IS_ENABLED(CONFIG_MLX5_ESWITCH) + if (MLX5_VPORT_MANAGER(mdev)) netdev->switchdev_ops = &mlx5e_switchdev_ops; #endif @@ -4217,7 +4197,7 @@ static void mlx5e_nic_enable(struct mlx5e_priv *priv) mlx5e_enable_async_events(priv); - if (MLX5_CAP_GEN(mdev, vport_group_manager)) + if (MLX5_VPORT_MANAGER(priv->mdev)) mlx5e_register_vport_reps(priv); if (netdev->reg_state != NETREG_REGISTERED) @@ -4251,7 +4231,7 @@ static void mlx5e_nic_disable(struct mlx5e_priv *priv) queue_work(priv->wq, &priv->set_rx_mode_work); - if (MLX5_CAP_GEN(mdev, vport_group_manager)) + if (MLX5_VPORT_MANAGER(priv->mdev)) mlx5e_unregister_vport_reps(priv); mlx5e_disable_async_events(priv); @@ -4424,32 +4404,29 @@ static void mlx5e_detach(struct mlx5_core_dev *mdev, void *vpriv) static void *mlx5e_add(struct mlx5_core_dev *mdev) { - struct mlx5_eswitch *esw = mdev->priv.eswitch; - int total_vfs = MLX5_TOTAL_VPORTS(mdev); - struct mlx5e_rep_priv *rpriv = NULL; + struct net_device *netdev; + void *rpriv = NULL; void *priv; - int vport; int err; - struct net_device *netdev; err = mlx5e_check_required_hca_cap(mdev); if (err) return NULL; - if (MLX5_CAP_GEN(mdev, vport_group_manager)) { - rpriv = kzalloc(sizeof(*rpriv), GFP_KERNEL); +#ifdef CONFIG_MLX5_ESWITCH + if (MLX5_VPORT_MANAGER(mdev)) { + rpriv = mlx5e_alloc_nic_rep_priv(mdev); if (!rpriv) { - mlx5_core_warn(mdev, - "Not creating net device, Failed to alloc rep priv data\n"); + mlx5_core_warn(mdev, "Failed to alloc NIC rep priv data\n"); return NULL; } - rpriv->rep = &esw->offloads.vport_reps[0]; } +#endif netdev = mlx5e_create_netdev(mdev, &mlx5e_nic_profile, rpriv); if (!netdev) { mlx5_core_err(mdev, "mlx5e_create_netdev failed\n"); - goto err_unregister_reps; + goto err_free_rpriv; } priv = netdev_priv(netdev); @@ -4470,14 +4447,9 @@ static void *mlx5e_add(struct mlx5_core_dev *mdev) err_detach: mlx5e_detach(mdev, priv); - err_destroy_netdev: mlx5e_destroy_netdev(priv); - -err_unregister_reps: - for (vport = 1; vport < total_vfs; vport++) - mlx5_eswitch_unregister_vport_rep(esw, vport); - +err_free_rpriv: kfree(rpriv); return NULL; } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c index 45e60be9c277..f3c494a4ecdf 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c @@ -651,37 +651,42 @@ static int mlx5e_rep_get_phys_port_name(struct net_device *dev, return 0; } -static int mlx5e_rep_ndo_setup_tc(struct net_device *dev, u32 handle, - u32 chain_index, __be16 proto, - struct tc_to_netdev *tc) +static int +mlx5e_rep_setup_tc_cls_flower(struct net_device *dev, + struct tc_cls_flower_offload *cls_flower) { struct mlx5e_priv *priv = netdev_priv(dev); - if (TC_H_MAJ(handle) != TC_H_MAJ(TC_H_INGRESS)) + if (TC_H_MAJ(cls_flower->common.handle) != TC_H_MAJ(TC_H_INGRESS) || + cls_flower->common.chain_index) return -EOPNOTSUPP; - if (tc->egress_dev) { + if (cls_flower->egress_dev) { struct mlx5_eswitch *esw = priv->mdev->priv.eswitch; - struct net_device *uplink_dev = mlx5_eswitch_get_uplink_netdev(esw); - return uplink_dev->netdev_ops->ndo_setup_tc(uplink_dev, handle, - chain_index, - proto, tc); + dev = mlx5_eswitch_get_uplink_netdev(esw); + return dev->netdev_ops->ndo_setup_tc(dev, TC_SETUP_CLSFLOWER, + cls_flower); } - if (chain_index) + switch (cls_flower->command) { + case TC_CLSFLOWER_REPLACE: + return mlx5e_configure_flower(priv, cls_flower); + case TC_CLSFLOWER_DESTROY: + return mlx5e_delete_flower(priv, cls_flower); + case TC_CLSFLOWER_STATS: + return mlx5e_stats_flower(priv, cls_flower); + default: return -EOPNOTSUPP; + } +} - switch (tc->type) { +static int mlx5e_rep_setup_tc(struct net_device *dev, enum tc_setup_type type, + void *type_data) +{ + switch (type) { case TC_SETUP_CLSFLOWER: - switch (tc->cls_flower->command) { - case TC_CLSFLOWER_REPLACE: - return mlx5e_configure_flower(priv, proto, tc->cls_flower); - case TC_CLSFLOWER_DESTROY: - return mlx5e_delete_flower(priv, tc->cls_flower); - case TC_CLSFLOWER_STATS: - return mlx5e_stats_flower(priv, tc->cls_flower); - } + return mlx5e_rep_setup_tc_cls_flower(dev, type_data); default: return -EOPNOTSUPP; } @@ -773,7 +778,7 @@ static const struct net_device_ops mlx5e_netdev_ops_rep = { .ndo_stop = mlx5e_rep_close, .ndo_start_xmit = mlx5e_xmit, .ndo_get_phys_port_name = mlx5e_rep_get_phys_port_name, - .ndo_setup_tc = mlx5e_rep_ndo_setup_tc, + .ndo_setup_tc = mlx5e_rep_setup_tc, .ndo_get_stats64 = mlx5e_rep_get_stats, .ndo_has_offload_stats = mlx5e_has_offload_stats, .ndo_get_offload_stats = mlx5e_get_offload_stats, @@ -1099,3 +1104,16 @@ void mlx5e_unregister_vport_reps(struct mlx5e_priv *priv) mlx5e_rep_unregister_vf_vports(priv); /* VFs vports */ mlx5_eswitch_unregister_vport_rep(esw, 0); /* UPLINK PF*/ } + +void *mlx5e_alloc_nic_rep_priv(struct mlx5_core_dev *mdev) +{ + struct mlx5_eswitch *esw = mdev->priv.eswitch; + struct mlx5e_rep_priv *rpriv; + + rpriv = kzalloc(sizeof(*rpriv), GFP_KERNEL); + if (!rpriv) + return NULL; + + rpriv->rep = &esw->offloads.vport_reps[0]; + return rpriv; +} diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h index a0a1a7a1d6c0..5659ed9f51e6 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h @@ -38,6 +38,7 @@ #include "eswitch.h" #include "en.h" +#ifdef CONFIG_MLX5_ESWITCH struct mlx5e_neigh_update_table { struct rhashtable neigh_ht; /* Save the neigh hash entries in a list in addition to the hash table @@ -123,6 +124,7 @@ struct mlx5e_encap_entry { int encap_size; }; +void *mlx5e_alloc_nic_rep_priv(struct mlx5_core_dev *mdev); void mlx5e_register_vport_reps(struct mlx5e_priv *priv); void mlx5e_unregister_vport_reps(struct mlx5e_priv *priv); bool mlx5e_is_uplink_rep(struct mlx5e_priv *priv); @@ -141,5 +143,12 @@ void mlx5e_rep_encap_entry_detach(struct mlx5e_priv *priv, struct mlx5e_encap_entry *e); void mlx5e_rep_queue_neigh_stats_work(struct mlx5e_priv *priv); +#else /* CONFIG_MLX5_ESWITCH */ +static inline void mlx5e_register_vport_reps(struct mlx5e_priv *priv) {} +static inline void mlx5e_unregister_vport_reps(struct mlx5e_priv *priv) {} +static inline bool mlx5e_is_uplink_rep(struct mlx5e_priv *priv) { return false; } +static inline int mlx5e_add_sqs_fwd_rules(struct mlx5e_priv *priv) { return 0; } +static inline void mlx5e_remove_sqs_fwd_rules(struct mlx5e_priv *priv) {} +#endif #endif /* __MLX5E_REP_H__ */ diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c index 325b2c8c1c6d..8e224bcbc6a6 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c @@ -857,6 +857,7 @@ wq_ll_pop: &wqe->next.next_wqe_index); } +#ifdef CONFIG_MLX5_ESWITCH void mlx5e_handle_rx_cqe_rep(struct mlx5e_rq *rq, struct mlx5_cqe64 *cqe) { struct net_device *netdev = rq->netdev; @@ -901,6 +902,7 @@ wq_ll_pop: mlx5_wq_ll_pop(&rq->wq, wqe_counter_be, &wqe->next.next_wqe_index); } +#endif static inline void mlx5e_mpwqe_fill_rx_skb(struct mlx5e_rq *rq, struct mlx5_cqe64 *cqe, diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c index 3c536f560dd2..3b10d3df7627 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c @@ -1326,7 +1326,7 @@ static int parse_tc_nic_actions(struct mlx5e_priv *priv, struct tcf_exts *exts, LIST_HEAD(actions); int err; - if (tc_no_actions(exts)) + if (!tcf_exts_has_actions(exts)) return -EINVAL; attr->flow_tag = MLX5_FS_DEFAULT_FLOW_TAG; @@ -1839,7 +1839,7 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv, struct tcf_exts *exts, bool encap = false; int err = 0; - if (tc_no_actions(exts)) + if (!tcf_exts_has_actions(exts)) return -EINVAL; memset(attr, 0, sizeof(*attr)); @@ -1939,7 +1939,7 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv, struct tcf_exts *exts, return err; } -int mlx5e_configure_flower(struct mlx5e_priv *priv, __be16 protocol, +int mlx5e_configure_flower(struct mlx5e_priv *priv, struct tc_cls_flower_offload *f) { struct mlx5_eswitch *esw = priv->mdev->priv.eswitch; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.h b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.h index ecbe30d808ae..c14c263a739b 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.h @@ -33,12 +33,15 @@ #ifndef __MLX5_EN_TC_H__ #define __MLX5_EN_TC_H__ +#include <net/pkt_cls.h> + #define MLX5E_TC_FLOW_ID_MASK 0x0000ffff +#ifdef CONFIG_MLX5_ESWITCH int mlx5e_tc_init(struct mlx5e_priv *priv); void mlx5e_tc_cleanup(struct mlx5e_priv *priv); -int mlx5e_configure_flower(struct mlx5e_priv *priv, __be16 protocol, +int mlx5e_configure_flower(struct mlx5e_priv *priv, struct tc_cls_flower_offload *f); int mlx5e_delete_flower(struct mlx5e_priv *priv, struct tc_cls_flower_offload *f); @@ -60,4 +63,10 @@ static inline int mlx5e_tc_num_filters(struct mlx5e_priv *priv) return atomic_read(&priv->fs.tc.ht.nelems); } +#else /* CONFIG_MLX5_ESWITCH */ +static inline int mlx5e_tc_init(struct mlx5e_priv *priv) { return 0; } +static inline void mlx5e_tc_cleanup(struct mlx5e_priv *priv) {} +static inline int mlx5e_tc_num_filters(struct mlx5e_priv *priv) { return 0; } +#endif + #endif /* __MLX5_EN_TC_H__ */ diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eq.c b/drivers/net/ethernet/mellanox/mlx5/core/eq.c index 52b9a64cd3a2..de704ff5619a 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eq.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eq.c @@ -36,9 +36,7 @@ #include <linux/mlx5/cmd.h> #include "mlx5_core.h" #include "fpga/core.h" -#ifdef CONFIG_MLX5_CORE_EN #include "eswitch.h" -#endif enum { MLX5_EQE_SIZE = sizeof(struct mlx5_eqe), @@ -467,11 +465,9 @@ static irqreturn_t mlx5_eq_int(int irq, void *eq_ptr) } break; -#ifdef CONFIG_MLX5_CORE_EN case MLX5_EVENT_TYPE_NIC_VPORT_CHANGE: mlx5_eswitch_vport_event(dev->priv.eswitch, eqe); break; -#endif case MLX5_EVENT_TYPE_PORT_MODULE_EVENT: mlx5_port_module_event(dev, eqe); @@ -688,9 +684,7 @@ int mlx5_start_eqs(struct mlx5_core_dev *dev) u64 async_event_mask = MLX5_ASYNC_EVENT_MASK; int err; - if (MLX5_CAP_GEN(dev, port_type) == MLX5_CAP_PORT_TYPE_ETH && - MLX5_CAP_GEN(dev, vport_group_manager) && - mlx5_core_is_pf(dev)) + if (MLX5_VPORT_MANAGER(dev)) async_event_mask |= (1ull << MLX5_EVENT_TYPE_NIC_VPORT_CHANGE); if (MLX5_CAP_GEN(dev, port_module_event)) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c index 8b18cc9ec026..fd51f0ea8df9 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c @@ -46,19 +46,13 @@ enum { MLX5_ACTION_DEL = 2, }; -/* E-Switch UC L2 table hash node */ -struct esw_uc_addr { - struct l2addr_node node; - u32 table_index; - u32 vport; -}; - /* Vport UC/MC hash node */ struct vport_addr { struct l2addr_node node; u8 action; u32 vport; - struct mlx5_flow_handle *flow_rule; /* SRIOV only */ + struct mlx5_flow_handle *flow_rule; + bool mpfs; /* UC MAC was added to MPFs */ /* A flag indicating that mac was added due to mc promiscuous vport */ bool mc_promisc; }; @@ -154,81 +148,6 @@ static int modify_esw_vport_cvlan(struct mlx5_core_dev *dev, u32 vport, return modify_esw_vport_context_cmd(dev, vport, in, sizeof(in)); } -/* HW L2 Table (MPFS) management */ -static int set_l2_table_entry_cmd(struct mlx5_core_dev *dev, u32 index, - u8 *mac, u8 vlan_valid, u16 vlan) -{ - u32 in[MLX5_ST_SZ_DW(set_l2_table_entry_in)] = {0}; - u32 out[MLX5_ST_SZ_DW(set_l2_table_entry_out)] = {0}; - u8 *in_mac_addr; - - MLX5_SET(set_l2_table_entry_in, in, opcode, - MLX5_CMD_OP_SET_L2_TABLE_ENTRY); - MLX5_SET(set_l2_table_entry_in, in, table_index, index); - MLX5_SET(set_l2_table_entry_in, in, vlan_valid, vlan_valid); - MLX5_SET(set_l2_table_entry_in, in, vlan, vlan); - - in_mac_addr = MLX5_ADDR_OF(set_l2_table_entry_in, in, mac_address); - ether_addr_copy(&in_mac_addr[2], mac); - - return mlx5_cmd_exec(dev, in, sizeof(in), out, sizeof(out)); -} - -static int del_l2_table_entry_cmd(struct mlx5_core_dev *dev, u32 index) -{ - u32 in[MLX5_ST_SZ_DW(delete_l2_table_entry_in)] = {0}; - u32 out[MLX5_ST_SZ_DW(delete_l2_table_entry_out)] = {0}; - - MLX5_SET(delete_l2_table_entry_in, in, opcode, - MLX5_CMD_OP_DELETE_L2_TABLE_ENTRY); - MLX5_SET(delete_l2_table_entry_in, in, table_index, index); - return mlx5_cmd_exec(dev, in, sizeof(in), out, sizeof(out)); -} - -static int alloc_l2_table_index(struct mlx5_l2_table *l2_table, u32 *ix) -{ - int err = 0; - - *ix = find_first_zero_bit(l2_table->bitmap, l2_table->size); - if (*ix >= l2_table->size) - err = -ENOSPC; - else - __set_bit(*ix, l2_table->bitmap); - - return err; -} - -static void free_l2_table_index(struct mlx5_l2_table *l2_table, u32 ix) -{ - __clear_bit(ix, l2_table->bitmap); -} - -static int set_l2_table_entry(struct mlx5_core_dev *dev, u8 *mac, - u8 vlan_valid, u16 vlan, - u32 *index) -{ - struct mlx5_l2_table *l2_table = &dev->priv.eswitch->l2_table; - int err; - - err = alloc_l2_table_index(l2_table, index); - if (err) - return err; - - err = set_l2_table_entry_cmd(dev, *index, mac, vlan_valid, vlan); - if (err) - free_l2_table_index(l2_table, *index); - - return err; -} - -static void del_l2_table_entry(struct mlx5_core_dev *dev, u32 index) -{ - struct mlx5_l2_table *l2_table = &dev->priv.eswitch->l2_table; - - del_l2_table_entry_cmd(dev, index); - free_l2_table_index(l2_table, index); -} - /* E-Switch FDB */ static struct mlx5_flow_handle * __esw_fdb_set_vport_rule(struct mlx5_eswitch *esw, u32 vport, bool rx_rule, @@ -455,65 +374,60 @@ typedef int (*vport_addr_action)(struct mlx5_eswitch *esw, static int esw_add_uc_addr(struct mlx5_eswitch *esw, struct vport_addr *vaddr) { - struct hlist_head *hash = esw->l2_table.l2_hash; - struct esw_uc_addr *esw_uc; u8 *mac = vaddr->node.addr; u32 vport = vaddr->vport; int err; - esw_uc = l2addr_hash_find(hash, mac, struct esw_uc_addr); - if (esw_uc) { + /* Skip mlx5_mpfs_add_mac for PFs, + * it is already done by the PF netdev in mlx5e_execute_l2_action + */ + if (!vport) + goto fdb_add; + + err = mlx5_mpfs_add_mac(esw->dev, mac); + if (err) { esw_warn(esw->dev, - "Failed to set L2 mac(%pM) for vport(%d), mac is already in use by vport(%d)\n", - mac, vport, esw_uc->vport); - return -EEXIST; + "Failed to add L2 table mac(%pM) for vport(%d), err(%d)\n", + mac, vport, err); + return err; } + vaddr->mpfs = true; - esw_uc = l2addr_hash_add(hash, mac, struct esw_uc_addr, GFP_KERNEL); - if (!esw_uc) - return -ENOMEM; - esw_uc->vport = vport; - - err = set_l2_table_entry(esw->dev, mac, 0, 0, &esw_uc->table_index); - if (err) - goto abort; - +fdb_add: /* SRIOV is enabled: Forward UC MAC to vport */ if (esw->fdb_table.fdb && esw->mode == SRIOV_LEGACY) vaddr->flow_rule = esw_fdb_set_vport_rule(esw, mac, vport); - esw_debug(esw->dev, "\tADDED UC MAC: vport[%d] %pM index:%d fr(%p)\n", - vport, mac, esw_uc->table_index, vaddr->flow_rule); - return err; -abort: - l2addr_hash_del(esw_uc); + esw_debug(esw->dev, "\tADDED UC MAC: vport[%d] %pM fr(%p)\n", + vport, mac, vaddr->flow_rule); + return err; } static int esw_del_uc_addr(struct mlx5_eswitch *esw, struct vport_addr *vaddr) { - struct hlist_head *hash = esw->l2_table.l2_hash; - struct esw_uc_addr *esw_uc; u8 *mac = vaddr->node.addr; u32 vport = vaddr->vport; + int err = 0; - esw_uc = l2addr_hash_find(hash, mac, struct esw_uc_addr); - if (!esw_uc || esw_uc->vport != vport) { - esw_debug(esw->dev, - "MAC(%pM) doesn't belong to vport (%d)\n", - mac, vport); - return -EINVAL; - } - esw_debug(esw->dev, "\tDELETE UC MAC: vport[%d] %pM index:%d fr(%p)\n", - vport, mac, esw_uc->table_index, vaddr->flow_rule); + /* Skip mlx5_mpfs_del_mac for PFs, + * it is already done by the PF netdev in mlx5e_execute_l2_action + */ + if (!vport || !vaddr->mpfs) + goto fdb_del; - del_l2_table_entry(esw->dev, esw_uc->table_index); + err = mlx5_mpfs_del_mac(esw->dev, mac); + if (err) + esw_warn(esw->dev, + "Failed to del L2 table mac(%pM) for vport(%d), err(%d)\n", + mac, vport, err); + vaddr->mpfs = false; +fdb_del: if (vaddr->flow_rule) mlx5_del_flow_rules(vaddr->flow_rule); vaddr->flow_rule = NULL; - l2addr_hash_del(esw_uc); return 0; } @@ -1611,13 +1525,14 @@ static void esw_disable_vport(struct mlx5_eswitch *esw, int vport_num) } /* Public E-Switch API */ +#define ESW_ALLOWED(esw) ((esw) && MLX5_VPORT_MANAGER((esw)->dev)) + int mlx5_eswitch_enable_sriov(struct mlx5_eswitch *esw, int nvfs, int mode) { int err; int i, enabled_events; - if (!esw || !MLX5_CAP_GEN(esw->dev, vport_group_manager) || - MLX5_CAP_GEN(esw->dev, port_type) != MLX5_CAP_PORT_TYPE_ETH) + if (!ESW_ALLOWED(esw)) return 0; if (!MLX5_CAP_GEN(esw->dev, eswitch_flow_table) || @@ -1634,7 +1549,6 @@ int mlx5_eswitch_enable_sriov(struct mlx5_eswitch *esw, int nvfs, int mode) esw_info(esw->dev, "E-Switch enable SRIOV: nvfs(%d) mode (%d)\n", nvfs, mode); esw->mode = mode; - esw_disable_vport(esw, 0); if (mode == SRIOV_LEGACY) err = esw_create_legacy_fdb_table(esw, nvfs + 1); @@ -1647,7 +1561,11 @@ int mlx5_eswitch_enable_sriov(struct mlx5_eswitch *esw, int nvfs, int mode) if (err) esw_warn(esw->dev, "Failed to create eswitch TSAR"); - enabled_events = (mode == SRIOV_LEGACY) ? SRIOV_VPORT_EVENTS : UC_ADDR_CHANGE; + /* Don't enable vport events when in SRIOV_OFFLOADS mode, since: + * 1. L2 table (MPFS) is programmed by PF/VF representors netdevs set_rx_mode + * 2. FDB/Eswitch is programmed by user space tools + */ + enabled_events = (mode == SRIOV_LEGACY) ? SRIOV_VPORT_EVENTS : 0; for (i = 0; i <= nvfs; i++) esw_enable_vport(esw, i, enabled_events); @@ -1656,7 +1574,6 @@ int mlx5_eswitch_enable_sriov(struct mlx5_eswitch *esw, int nvfs, int mode) return 0; abort: - esw_enable_vport(esw, 0, UC_ADDR_CHANGE); esw->mode = SRIOV_NONE; return err; } @@ -1667,9 +1584,7 @@ void mlx5_eswitch_disable_sriov(struct mlx5_eswitch *esw) int nvports; int i; - if (!esw || !MLX5_CAP_GEN(esw->dev, vport_group_manager) || - MLX5_CAP_GEN(esw->dev, port_type) != MLX5_CAP_PORT_TYPE_ETH || - esw->mode == SRIOV_NONE) + if (!ESW_ALLOWED(esw) || esw->mode == SRIOV_NONE) return; esw_info(esw->dev, "disable SRIOV: active vports(%d) mode(%d)\n", @@ -1692,44 +1607,21 @@ void mlx5_eswitch_disable_sriov(struct mlx5_eswitch *esw) esw_offloads_cleanup(esw, nvports); esw->mode = SRIOV_NONE; - /* VPORT 0 (PF) must be enabled back with non-sriov configuration */ - esw_enable_vport(esw, 0, UC_ADDR_CHANGE); -} - -void mlx5_eswitch_attach(struct mlx5_eswitch *esw) -{ - if (!esw || !MLX5_CAP_GEN(esw->dev, vport_group_manager) || - MLX5_CAP_GEN(esw->dev, port_type) != MLX5_CAP_PORT_TYPE_ETH) - return; - - esw_enable_vport(esw, 0, UC_ADDR_CHANGE); - /* VF Vports will be enabled when SRIOV is enabled */ -} - -void mlx5_eswitch_detach(struct mlx5_eswitch *esw) -{ - if (!esw || !MLX5_CAP_GEN(esw->dev, vport_group_manager) || - MLX5_CAP_GEN(esw->dev, port_type) != MLX5_CAP_PORT_TYPE_ETH) - return; - - esw_disable_vport(esw, 0); } int mlx5_eswitch_init(struct mlx5_core_dev *dev) { - int l2_table_size = 1 << MLX5_CAP_GEN(dev, log_max_l2_table); int total_vports = MLX5_TOTAL_VPORTS(dev); struct mlx5_eswitch *esw; int vport_num; int err; - if (!MLX5_CAP_GEN(dev, vport_group_manager) || - MLX5_CAP_GEN(dev, port_type) != MLX5_CAP_PORT_TYPE_ETH) + if (!MLX5_VPORT_MANAGER(dev)) return 0; esw_info(dev, - "Total vports %d, l2 table size(%d), per vport: max uc(%d) max mc(%d)\n", - total_vports, l2_table_size, + "Total vports %d, per vport: max uc(%d) max mc(%d)\n", + total_vports, MLX5_MAX_UC_PER_VPORT(dev), MLX5_MAX_MC_PER_VPORT(dev)); @@ -1739,14 +1631,6 @@ int mlx5_eswitch_init(struct mlx5_core_dev *dev) esw->dev = dev; - esw->l2_table.bitmap = kcalloc(BITS_TO_LONGS(l2_table_size), - sizeof(uintptr_t), GFP_KERNEL); - if (!esw->l2_table.bitmap) { - err = -ENOMEM; - goto abort; - } - esw->l2_table.size = l2_table_size; - esw->work_queue = create_singlethread_workqueue("mlx5_esw_wq"); if (!esw->work_queue) { err = -ENOMEM; @@ -1797,7 +1681,6 @@ int mlx5_eswitch_init(struct mlx5_core_dev *dev) abort: if (esw->work_queue) destroy_workqueue(esw->work_queue); - kfree(esw->l2_table.bitmap); kfree(esw->vports); kfree(esw->offloads.vport_reps); kfree(esw); @@ -1806,15 +1689,13 @@ abort: void mlx5_eswitch_cleanup(struct mlx5_eswitch *esw) { - if (!esw || !MLX5_CAP_GEN(esw->dev, vport_group_manager) || - MLX5_CAP_GEN(esw->dev, port_type) != MLX5_CAP_PORT_TYPE_ETH) + if (!esw || !MLX5_VPORT_MANAGER(esw->dev)) return; esw_info(esw->dev, "cleanup\n"); esw->dev->priv.eswitch = NULL; destroy_workqueue(esw->work_queue); - kfree(esw->l2_table.bitmap); kfree(esw->offloads.vport_reps); kfree(esw->vports); kfree(esw); @@ -1838,8 +1719,6 @@ void mlx5_eswitch_vport_event(struct mlx5_eswitch *esw, struct mlx5_eqe *eqe) } /* Vport Administration */ -#define ESW_ALLOWED(esw) \ - (esw && MLX5_CAP_GEN(esw->dev, vport_group_manager) && mlx5_core_is_pf(esw->dev)) #define LEGAL_VPORT(esw, vport) (vport >= 0 && vport < esw->total_vports) int mlx5_eswitch_set_vport_mac(struct mlx5_eswitch *esw, diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h index 834a33050969..565c8b7a399a 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h @@ -37,6 +37,15 @@ #include <linux/if_link.h> #include <net/devlink.h> #include <linux/mlx5/device.h> +#include "lib/mpfs.h" + +enum { + SRIOV_NONE, + SRIOV_LEGACY, + SRIOV_OFFLOADS +}; + +#ifdef CONFIG_MLX5_ESWITCH #define MLX5_MAX_UC_PER_VPORT(dev) \ (1 << MLX5_CAP_GEN(dev, log_max_current_uc_list)) @@ -44,9 +53,6 @@ #define MLX5_MAX_MC_PER_VPORT(dev) \ (1 << MLX5_CAP_GEN(dev, log_max_current_mc_list)) -#define MLX5_L2_ADDR_HASH_SIZE (BIT(BITS_PER_BYTE)) -#define MLX5_L2_ADDR_HASH(addr) (addr[5]) - #define FDB_UPLINK_VPORT 0xffff #define MLX5_MIN_BW_SHARE 1 @@ -54,48 +60,6 @@ #define MLX5_RATE_TO_BW_SHARE(rate, divider, limit) \ min_t(u32, max_t(u32, (rate) / (divider), MLX5_MIN_BW_SHARE), limit) -/* L2 -mac address based- hash helpers */ -struct l2addr_node { - struct hlist_node hlist; - u8 addr[ETH_ALEN]; -}; - -#define for_each_l2hash_node(hn, tmp, hash, i) \ - for (i = 0; i < MLX5_L2_ADDR_HASH_SIZE; i++) \ - hlist_for_each_entry_safe(hn, tmp, &hash[i], hlist) - -#define l2addr_hash_find(hash, mac, type) ({ \ - int ix = MLX5_L2_ADDR_HASH(mac); \ - bool found = false; \ - type *ptr = NULL; \ - \ - hlist_for_each_entry(ptr, &hash[ix], node.hlist) \ - if (ether_addr_equal(ptr->node.addr, mac)) {\ - found = true; \ - break; \ - } \ - if (!found) \ - ptr = NULL; \ - ptr; \ -}) - -#define l2addr_hash_add(hash, mac, type, gfp) ({ \ - int ix = MLX5_L2_ADDR_HASH(mac); \ - type *ptr = NULL; \ - \ - ptr = kzalloc(sizeof(type), gfp); \ - if (ptr) { \ - ether_addr_copy(ptr->node.addr, mac); \ - hlist_add_head(&ptr->node.hlist, &hash[ix]);\ - } \ - ptr; \ -}) - -#define l2addr_hash_del(ptr) ({ \ - hlist_del(&ptr->node.hlist); \ - kfree(ptr); \ -}) - struct vport_ingress { struct mlx5_flow_table *acl; struct mlx5_flow_group *allow_untagged_spoofchk_grp; @@ -150,12 +114,6 @@ struct mlx5_vport { u16 enabled_events; }; -struct mlx5_l2_table { - struct hlist_head l2_hash[MLX5_L2_ADDR_HASH_SIZE]; - u32 size; - unsigned long *bitmap; -}; - struct mlx5_eswitch_fdb { void *fdb; union { @@ -175,12 +133,6 @@ struct mlx5_eswitch_fdb { }; }; -enum { - SRIOV_NONE, - SRIOV_LEGACY, - SRIOV_OFFLOADS -}; - struct mlx5_esw_sq { struct mlx5_flow_handle *send_to_vport_rule; struct list_head list; @@ -222,7 +174,6 @@ struct esw_mc_addr { /* SRIOV only */ struct mlx5_eswitch { struct mlx5_core_dev *dev; - struct mlx5_l2_table l2_table; struct mlx5_eswitch_fdb fdb_table; struct hlist_head mc_table[MLX5_L2_ADDR_HASH_SIZE]; struct workqueue_struct *work_queue; @@ -250,8 +201,6 @@ int esw_offloads_init(struct mlx5_eswitch *esw, int nvports); /* E-Switch API */ int mlx5_eswitch_init(struct mlx5_core_dev *dev); void mlx5_eswitch_cleanup(struct mlx5_eswitch *esw); -void mlx5_eswitch_attach(struct mlx5_eswitch *esw); -void mlx5_eswitch_detach(struct mlx5_eswitch *esw); void mlx5_eswitch_vport_event(struct mlx5_eswitch *esw, struct mlx5_eqe *eqe); int mlx5_eswitch_enable_sriov(struct mlx5_eswitch *esw, int nvfs, int mode); void mlx5_eswitch_disable_sriov(struct mlx5_eswitch *esw); @@ -345,4 +294,13 @@ int __mlx5_eswitch_set_vport_vlan(struct mlx5_eswitch *esw, #define esw_debug(dev, format, ...) \ mlx5_core_dbg_mask(dev, MLX5_DEBUG_ESWITCH_MASK, format, ##__VA_ARGS__) +#else /* CONFIG_MLX5_ESWITCH */ +/* eswitch API stubs */ +static inline int mlx5_eswitch_init(struct mlx5_core_dev *dev) { return 0; } +static inline void mlx5_eswitch_cleanup(struct mlx5_eswitch *esw) {} +static inline void mlx5_eswitch_vport_event(struct mlx5_eswitch *esw, struct mlx5_eqe *eqe) {} +static inline int mlx5_eswitch_enable_sriov(struct mlx5_eswitch *esw, int nvfs, int mode) { return 0; } +static inline void mlx5_eswitch_disable_sriov(struct mlx5_eswitch *esw) {} +#endif /* CONFIG_MLX5_ESWITCH */ + #endif /* __MLX5_ESWITCH_H__ */ diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c index 95b64025ce36..e7c186b58579 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c @@ -433,6 +433,8 @@ static int esw_create_offloads_fast_fdb_table(struct mlx5_eswitch *esw) struct mlx5_flow_table *fdb = NULL; int esw_size, err = 0; u32 flags = 0; + u32 max_flow_counter = (MLX5_CAP_GEN(dev, max_flow_counter_31_16) << 16) | + MLX5_CAP_GEN(dev, max_flow_counter_15_0); root_ns = mlx5_get_flow_namespace(dev, MLX5_FLOW_NAMESPACE_FDB); if (!root_ns) { @@ -443,9 +445,9 @@ static int esw_create_offloads_fast_fdb_table(struct mlx5_eswitch *esw) esw_debug(dev, "Create offloads FDB table, min (max esw size(2^%d), max counters(%d)*groups(%d))\n", MLX5_CAP_ESW_FLOWTABLE_FDB(dev, log_max_ft_size), - MLX5_CAP_GEN(dev, max_flow_counter), ESW_OFFLOADS_NUM_GROUPS); + max_flow_counter, ESW_OFFLOADS_NUM_GROUPS); - esw_size = min_t(int, MLX5_CAP_GEN(dev, max_flow_counter) * ESW_OFFLOADS_NUM_GROUPS, + esw_size = min_t(int, max_flow_counter * ESW_OFFLOADS_NUM_GROUPS, 1 << MLX5_CAP_ESW_FLOWTABLE_FDB(dev, log_max_ft_size)); if (esw->offloads.encap != DEVLINK_ESWITCH_ENCAP_MODE_NONE) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c index e750f07793b8..16b32f31d691 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c @@ -359,7 +359,7 @@ int mlx5_cmd_delete_fte(struct mlx5_core_dev *dev, return mlx5_cmd_exec(dev, in, sizeof(in), out, sizeof(out)); } -int mlx5_cmd_fc_alloc(struct mlx5_core_dev *dev, u16 *id) +int mlx5_cmd_fc_alloc(struct mlx5_core_dev *dev, u32 *id) { u32 in[MLX5_ST_SZ_DW(alloc_flow_counter_in)] = {0}; u32 out[MLX5_ST_SZ_DW(alloc_flow_counter_out)] = {0}; @@ -374,7 +374,7 @@ int mlx5_cmd_fc_alloc(struct mlx5_core_dev *dev, u16 *id) return err; } -int mlx5_cmd_fc_free(struct mlx5_core_dev *dev, u16 id) +int mlx5_cmd_fc_free(struct mlx5_core_dev *dev, u32 id) { u32 in[MLX5_ST_SZ_DW(dealloc_flow_counter_in)] = {0}; u32 out[MLX5_ST_SZ_DW(dealloc_flow_counter_out)] = {0}; @@ -385,7 +385,7 @@ int mlx5_cmd_fc_free(struct mlx5_core_dev *dev, u16 id) return mlx5_cmd_exec(dev, in, sizeof(in), out, sizeof(out)); } -int mlx5_cmd_fc_query(struct mlx5_core_dev *dev, u16 id, +int mlx5_cmd_fc_query(struct mlx5_core_dev *dev, u32 id, u64 *packets, u64 *bytes) { u32 out[MLX5_ST_SZ_BYTES(query_flow_counter_out) + @@ -409,14 +409,14 @@ int mlx5_cmd_fc_query(struct mlx5_core_dev *dev, u16 id, } struct mlx5_cmd_fc_bulk { - u16 id; + u32 id; int num; int outlen; u32 out[0]; }; struct mlx5_cmd_fc_bulk * -mlx5_cmd_fc_bulk_alloc(struct mlx5_core_dev *dev, u16 id, int num) +mlx5_cmd_fc_bulk_alloc(struct mlx5_core_dev *dev, u32 id, int num) { struct mlx5_cmd_fc_bulk *b; int outlen = @@ -453,7 +453,7 @@ mlx5_cmd_fc_bulk_query(struct mlx5_core_dev *dev, struct mlx5_cmd_fc_bulk *b) } void mlx5_cmd_fc_bulk_get(struct mlx5_core_dev *dev, - struct mlx5_cmd_fc_bulk *b, u16 id, + struct mlx5_cmd_fc_bulk *b, u32 id, u64 *packets, u64 *bytes) { int index = id - b->id; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.h b/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.h index 0f98a7cf4877..c6d7bdf255b6 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.h @@ -74,20 +74,20 @@ int mlx5_cmd_update_root_ft(struct mlx5_core_dev *dev, struct mlx5_flow_table *ft, u32 underlay_qpn); -int mlx5_cmd_fc_alloc(struct mlx5_core_dev *dev, u16 *id); -int mlx5_cmd_fc_free(struct mlx5_core_dev *dev, u16 id); -int mlx5_cmd_fc_query(struct mlx5_core_dev *dev, u16 id, +int mlx5_cmd_fc_alloc(struct mlx5_core_dev *dev, u32 *id); +int mlx5_cmd_fc_free(struct mlx5_core_dev *dev, u32 id); +int mlx5_cmd_fc_query(struct mlx5_core_dev *dev, u32 id, u64 *packets, u64 *bytes); struct mlx5_cmd_fc_bulk; struct mlx5_cmd_fc_bulk * -mlx5_cmd_fc_bulk_alloc(struct mlx5_core_dev *dev, u16 id, int num); +mlx5_cmd_fc_bulk_alloc(struct mlx5_core_dev *dev, u32 id, int num); void mlx5_cmd_fc_bulk_free(struct mlx5_cmd_fc_bulk *b); int mlx5_cmd_fc_bulk_query(struct mlx5_core_dev *dev, struct mlx5_cmd_fc_bulk *b); void mlx5_cmd_fc_bulk_get(struct mlx5_core_dev *dev, - struct mlx5_cmd_fc_bulk *b, u16 id, + struct mlx5_cmd_fc_bulk *b, u32 id, u64 *packets, u64 *bytes); #endif diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h index 990acee6fb09..9fb5a333df52 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h @@ -136,7 +136,7 @@ struct mlx5_fc { u64 lastpackets; u64 lastbytes; - u16 id; + u32 id; bool deleted; bool aging; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_counters.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_counters.c index 6507d8acc54d..89d1f8650033 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/fs_counters.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_counters.c @@ -38,6 +38,8 @@ #include "fs_cmd.h" #define MLX5_FC_STATS_PERIOD msecs_to_jiffies(1000) +/* Max number of counters to query in bulk read is 32K */ +#define MLX5_SW_MAX_COUNTERS_BULK BIT(15) /* locking scheme: * @@ -90,16 +92,21 @@ static void mlx5_fc_stats_insert(struct rb_root *root, struct mlx5_fc *counter) rb_insert_color(&counter->node, root); } +/* The function returns the last node that was queried so the caller + * function can continue calling it till all counters are queried. + */ static struct rb_node *mlx5_fc_stats_query(struct mlx5_core_dev *dev, struct mlx5_fc *first, - u16 last_id) + u32 last_id) { struct mlx5_cmd_fc_bulk *b; struct rb_node *node = NULL; - u16 afirst_id; + u32 afirst_id; int num; int err; - int max_bulk = 1 << MLX5_CAP_GEN(dev, log_max_flow_counter_bulk); + + int max_bulk = min_t(int, MLX5_SW_MAX_COUNTERS_BULK, + (1 << MLX5_CAP_GEN(dev, log_max_flow_counter_bulk))); /* first id must be aligned to 4 when using bulk query */ afirst_id = first->id & ~0x3; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lib/mpfs.c b/drivers/net/ethernet/mellanox/mlx5/core/lib/mpfs.c new file mode 100644 index 000000000000..7cb67122e8b5 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlx5/core/lib/mpfs.c @@ -0,0 +1,201 @@ +/* + * Copyright (c) 2017, Mellanox Technologies. All rights reserved. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * OpenIB.org BSD license below: + * + * Redistribution and use in source and binary forms, with or + * without modification, are permitted provided that the following + * conditions are met: + * + * - Redistributions of source code must retain the above + * copyright notice, this list of conditions and the following + * disclaimer. + * + * - Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials + * provided with the distribution. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + */ + +#include <linux/etherdevice.h> +#include <linux/mlx5/driver.h> +#include <linux/mlx5/mlx5_ifc.h> +#include "mlx5_core.h" +#include "lib/mpfs.h" + +/* HW L2 Table (MPFS) management */ +static int set_l2table_entry_cmd(struct mlx5_core_dev *dev, u32 index, u8 *mac) +{ + u32 in[MLX5_ST_SZ_DW(set_l2_table_entry_in)] = {0}; + u32 out[MLX5_ST_SZ_DW(set_l2_table_entry_out)] = {0}; + u8 *in_mac_addr; + + MLX5_SET(set_l2_table_entry_in, in, opcode, MLX5_CMD_OP_SET_L2_TABLE_ENTRY); + MLX5_SET(set_l2_table_entry_in, in, table_index, index); + + in_mac_addr = MLX5_ADDR_OF(set_l2_table_entry_in, in, mac_address); + ether_addr_copy(&in_mac_addr[2], mac); + + return mlx5_cmd_exec(dev, in, sizeof(in), out, sizeof(out)); +} + +static int del_l2table_entry_cmd(struct mlx5_core_dev *dev, u32 index) +{ + u32 in[MLX5_ST_SZ_DW(delete_l2_table_entry_in)] = {0}; + u32 out[MLX5_ST_SZ_DW(delete_l2_table_entry_out)] = {0}; + + MLX5_SET(delete_l2_table_entry_in, in, opcode, MLX5_CMD_OP_DELETE_L2_TABLE_ENTRY); + MLX5_SET(delete_l2_table_entry_in, in, table_index, index); + return mlx5_cmd_exec(dev, in, sizeof(in), out, sizeof(out)); +} + +/* UC L2 table hash node */ +struct l2table_node { + struct l2addr_node node; + u32 index; /* index in HW l2 table */ +}; + +struct mlx5_mpfs { + struct hlist_head hash[MLX5_L2_ADDR_HASH_SIZE]; + struct mutex lock; /* Synchronize l2 table access */ + u32 size; + unsigned long *bitmap; +}; + +static int alloc_l2table_index(struct mlx5_mpfs *l2table, u32 *ix) +{ + int err = 0; + + *ix = find_first_zero_bit(l2table->bitmap, l2table->size); + if (*ix >= l2table->size) + err = -ENOSPC; + else + __set_bit(*ix, l2table->bitmap); + + return err; +} + +static void free_l2table_index(struct mlx5_mpfs *l2table, u32 ix) +{ + __clear_bit(ix, l2table->bitmap); +} + +int mlx5_mpfs_init(struct mlx5_core_dev *dev) +{ + int l2table_size = 1 << MLX5_CAP_GEN(dev, log_max_l2_table); + struct mlx5_mpfs *mpfs; + + if (!MLX5_VPORT_MANAGER(dev)) + return 0; + + mpfs = kzalloc(sizeof(*mpfs), GFP_KERNEL); + if (!mpfs) + return -ENOMEM; + + mutex_init(&mpfs->lock); + mpfs->size = l2table_size; + mpfs->bitmap = kcalloc(BITS_TO_LONGS(l2table_size), + sizeof(uintptr_t), GFP_KERNEL); + if (!mpfs->bitmap) { + kfree(mpfs); + return -ENOMEM; + } + + dev->priv.mpfs = mpfs; + return 0; +} + +void mlx5_mpfs_cleanup(struct mlx5_core_dev *dev) +{ + struct mlx5_mpfs *mpfs = dev->priv.mpfs; + + if (!MLX5_VPORT_MANAGER(dev)) + return; + + WARN_ON(!hlist_empty(mpfs->hash)); + kfree(mpfs->bitmap); + kfree(mpfs); +} + +int mlx5_mpfs_add_mac(struct mlx5_core_dev *dev, u8 *mac) +{ + struct mlx5_mpfs *mpfs = dev->priv.mpfs; + struct l2table_node *l2addr; + u32 index; + int err; + + if (!MLX5_VPORT_MANAGER(dev)) + return 0; + + mutex_lock(&mpfs->lock); + + l2addr = l2addr_hash_find(mpfs->hash, mac, struct l2table_node); + if (l2addr) { + err = -EEXIST; + goto abort; + } + + err = alloc_l2table_index(mpfs, &index); + if (err) + goto abort; + + l2addr = l2addr_hash_add(mpfs->hash, mac, struct l2table_node, GFP_KERNEL); + if (!l2addr) { + free_l2table_index(mpfs, index); + err = -ENOMEM; + goto abort; + } + + l2addr->index = index; + err = set_l2table_entry_cmd(dev, index, mac); + if (err) { + l2addr_hash_del(l2addr); + free_l2table_index(mpfs, index); + } + + mlx5_core_dbg(dev, "MPFS mac added %pM, index (%d)\n", mac, index); +abort: + mutex_unlock(&mpfs->lock); + return err; +} + +int mlx5_mpfs_del_mac(struct mlx5_core_dev *dev, u8 *mac) +{ + struct mlx5_mpfs *mpfs = dev->priv.mpfs; + struct l2table_node *l2addr; + int err = 0; + u32 index; + + if (!MLX5_VPORT_MANAGER(dev)) + return 0; + + mutex_lock(&mpfs->lock); + + l2addr = l2addr_hash_find(mpfs->hash, mac, struct l2table_node); + if (!l2addr) { + err = -ENOENT; + goto unlock; + } + + index = l2addr->index; + del_l2table_entry_cmd(dev, index); + l2addr_hash_del(l2addr); + free_l2table_index(mpfs, index); + mlx5_core_dbg(dev, "MPFS mac deleted %pM, index (%d)\n", mac, index); +unlock: + mutex_unlock(&mpfs->lock); + return err; +} diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lib/mpfs.h b/drivers/net/ethernet/mellanox/mlx5/core/lib/mpfs.h new file mode 100644 index 000000000000..4a7b2c3203a7 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlx5/core/lib/mpfs.h @@ -0,0 +1,95 @@ +/* + * Copyright (c) 2017, Mellanox Technologies, Ltd. All rights reserved. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * OpenIB.org BSD license below: + * + * Redistribution and use in source and binary forms, with or + * without modification, are permitted provided that the following + * conditions are met: + * + * - Redistributions of source code must retain the above + * copyright notice, this list of conditions and the following + * disclaimer. + * + * - Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials + * provided with the distribution. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + */ + +#ifndef __MLX5_MPFS_H__ +#define __MLX5_MPFS_H__ + +#include <linux/if_ether.h> +#include <linux/mlx5/device.h> + +/* L2 -mac address based- hash helpers */ +#define MLX5_L2_ADDR_HASH_SIZE (BIT(BITS_PER_BYTE)) +#define MLX5_L2_ADDR_HASH(addr) (addr[5]) + +struct l2addr_node { + struct hlist_node hlist; + u8 addr[ETH_ALEN]; +}; + +#define for_each_l2hash_node(hn, tmp, hash, i) \ + for (i = 0; i < MLX5_L2_ADDR_HASH_SIZE; i++) \ + hlist_for_each_entry_safe(hn, tmp, &(hash)[i], hlist) + +#define l2addr_hash_find(hash, mac, type) ({ \ + int ix = MLX5_L2_ADDR_HASH(mac); \ + bool found = false; \ + type *ptr = NULL; \ + \ + hlist_for_each_entry(ptr, &(hash)[ix], node.hlist) \ + if (ether_addr_equal(ptr->node.addr, mac)) {\ + found = true; \ + break; \ + } \ + if (!found) \ + ptr = NULL; \ + ptr; \ +}) + +#define l2addr_hash_add(hash, mac, type, gfp) ({ \ + int ix = MLX5_L2_ADDR_HASH(mac); \ + type *ptr = NULL; \ + \ + ptr = kzalloc(sizeof(type), gfp); \ + if (ptr) { \ + ether_addr_copy(ptr->node.addr, mac); \ + hlist_add_head(&ptr->node.hlist, &(hash)[ix]);\ + } \ + ptr; \ +}) + +#define l2addr_hash_del(ptr) ({ \ + hlist_del(&(ptr)->node.hlist); \ + kfree(ptr); \ +}) + +#ifdef CONFIG_MLX5_MPFS +int mlx5_mpfs_init(struct mlx5_core_dev *dev); +void mlx5_mpfs_cleanup(struct mlx5_core_dev *dev); +int mlx5_mpfs_add_mac(struct mlx5_core_dev *dev, u8 *mac); +int mlx5_mpfs_del_mac(struct mlx5_core_dev *dev, u8 *mac); +#else /* #ifndef CONFIG_MLX5_MPFS */ +static inline int mlx5_mpfs_init(struct mlx5_core_dev *dev) { return 0; } +static inline void mlx5_mpfs_cleanup(struct mlx5_core_dev *dev) {} +static inline int mlx5_mpfs_add_mac(struct mlx5_core_dev *dev, u8 *mac) { return 0; } +static inline int mlx5_mpfs_del_mac(struct mlx5_core_dev *dev, u8 *mac) { return 0; } +#endif +#endif diff --git a/drivers/net/ethernet/mellanox/mlx5/core/main.c b/drivers/net/ethernet/mellanox/mlx5/core/main.c index c065132b956d..6dbd637b4e66 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/main.c @@ -53,9 +53,8 @@ #include <net/devlink.h> #include "mlx5_core.h" #include "fs_core.h" -#ifdef CONFIG_MLX5_CORE_EN +#include "lib/mpfs.h" #include "eswitch.h" -#endif #include "lib/mlx5.h" #include "fpga/core.h" #include "accel/ipsec.h" @@ -946,13 +945,17 @@ static int mlx5_init_once(struct mlx5_core_dev *dev, struct mlx5_priv *priv) goto err_tables_cleanup; } -#ifdef CONFIG_MLX5_CORE_EN + err = mlx5_mpfs_init(dev); + if (err) { + dev_err(&pdev->dev, "Failed to init l2 table %d\n", err); + goto err_rl_cleanup; + } + err = mlx5_eswitch_init(dev); if (err) { dev_err(&pdev->dev, "Failed to init eswitch %d\n", err); - goto err_rl_cleanup; + goto err_mpfs_cleanup; } -#endif err = mlx5_sriov_init(dev); if (err) { @@ -971,13 +974,11 @@ static int mlx5_init_once(struct mlx5_core_dev *dev, struct mlx5_priv *priv) err_sriov_cleanup: mlx5_sriov_cleanup(dev); err_eswitch_cleanup: -#ifdef CONFIG_MLX5_CORE_EN mlx5_eswitch_cleanup(dev->priv.eswitch); - +err_mpfs_cleanup: + mlx5_mpfs_cleanup(dev); err_rl_cleanup: -#endif mlx5_cleanup_rl_table(dev); - err_tables_cleanup: mlx5_cleanup_mkey_table(dev); mlx5_cleanup_srq_table(dev); @@ -995,9 +996,8 @@ static void mlx5_cleanup_once(struct mlx5_core_dev *dev) { mlx5_fpga_cleanup(dev); mlx5_sriov_cleanup(dev); -#ifdef CONFIG_MLX5_CORE_EN mlx5_eswitch_cleanup(dev->priv.eswitch); -#endif + mlx5_mpfs_cleanup(dev); mlx5_cleanup_rl_table(dev); mlx5_cleanup_reserved_gids(dev); mlx5_cleanup_mkey_table(dev); @@ -1155,10 +1155,6 @@ static int mlx5_load_one(struct mlx5_core_dev *dev, struct mlx5_priv *priv, goto err_fs; } -#ifdef CONFIG_MLX5_CORE_EN - mlx5_eswitch_attach(dev->priv.eswitch); -#endif - err = mlx5_sriov_attach(dev); if (err) { dev_err(&pdev->dev, "sriov init failed %d\n", err); @@ -1202,9 +1198,6 @@ err_fpga_start: mlx5_sriov_detach(dev); err_sriov: -#ifdef CONFIG_MLX5_CORE_EN - mlx5_eswitch_detach(dev->priv.eswitch); -#endif mlx5_cleanup_fs(dev); err_fs: @@ -1279,9 +1272,6 @@ static int mlx5_unload_one(struct mlx5_core_dev *dev, struct mlx5_priv *priv, mlx5_fpga_device_stop(dev); mlx5_sriov_detach(dev); -#ifdef CONFIG_MLX5_CORE_EN - mlx5_eswitch_detach(dev->priv.eswitch); -#endif mlx5_cleanup_fs(dev); mlx5_irq_clear_affinity_hints(dev); free_comp_eqs(dev); @@ -1313,7 +1303,7 @@ struct mlx5_core_event_handler { }; static const struct devlink_ops mlx5_devlink_ops = { -#ifdef CONFIG_MLX5_CORE_EN +#ifdef CONFIG_MLX5_ESWITCH .eswitch_mode_set = mlx5_devlink_eswitch_mode_set, .eswitch_mode_get = mlx5_devlink_eswitch_mode_get, .eswitch_inline_mode_set = mlx5_devlink_eswitch_inline_mode_set, @@ -1353,6 +1343,9 @@ static int init_one(struct pci_dev *pdev, mutex_init(&dev->pci_status_mutex); mutex_init(&dev->intf_state_mutex); + INIT_LIST_HEAD(&priv->waiting_events_list); + priv->is_accum_events = false; + #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING err = init_srcu_struct(&priv->pfault_srcu); if (err) { diff --git a/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h b/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h index 6a263e8d883a..d8da9240a00b 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h @@ -43,6 +43,10 @@ #define DRIVER_VERSION "5.0-0" #define MLX5_TOTAL_VPORTS(mdev) (1 + pci_sriov_get_totalvfs(mdev->pdev)) +#define MLX5_VPORT_MANAGER(mdev) \ + (MLX5_CAP_GEN(mdev, vport_group_manager) && \ + (MLX5_CAP_GEN(mdev, port_type) == MLX5_CAP_PORT_TYPE_ETH) && \ + mlx5_core_is_pf(mdev)) extern uint mlx5_core_debug_mask; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/sriov.c b/drivers/net/ethernet/mellanox/mlx5/core/sriov.c index bf99d40e30b4..5e7ffc9fad78 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/sriov.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/sriov.c @@ -33,9 +33,7 @@ #include <linux/pci.h> #include <linux/mlx5/driver.h> #include "mlx5_core.h" -#ifdef CONFIG_MLX5_CORE_EN #include "eswitch.h" -#endif bool mlx5_sriov_is_enabled(struct mlx5_core_dev *dev) { @@ -57,14 +55,12 @@ static int mlx5_device_enable_sriov(struct mlx5_core_dev *dev, int num_vfs) return -EBUSY; } -#ifdef CONFIG_MLX5_CORE_EN err = mlx5_eswitch_enable_sriov(dev->priv.eswitch, num_vfs, SRIOV_LEGACY); if (err) { mlx5_core_warn(dev, "failed to enable eswitch SRIOV (%d)\n", err); return err; } -#endif for (vf = 0; vf < num_vfs; vf++) { err = mlx5_core_enable_hca(dev, vf + 1); @@ -88,11 +84,7 @@ static void mlx5_device_disable_sriov(struct mlx5_core_dev *dev) int vf; if (!sriov->enabled_vfs) -#ifdef CONFIG_MLX5_CORE_EN - goto disable_sriov_resources; -#else - return; -#endif + goto out; for (vf = 0; vf < sriov->num_vfs; vf++) { if (!sriov->vfs_ctx[vf].enabled) @@ -106,10 +98,8 @@ static void mlx5_device_disable_sriov(struct mlx5_core_dev *dev) sriov->enabled_vfs--; } -#ifdef CONFIG_MLX5_CORE_EN -disable_sriov_resources: +out: mlx5_eswitch_disable_sriov(dev->priv.eswitch); -#endif if (mlx5_wait_for_vf_pages(dev)) mlx5_core_warn(dev, "timeout reclaiming VFs pages\n"); diff --git a/drivers/net/ethernet/mellanox/mlxsw/Kconfig b/drivers/net/ethernet/mellanox/mlxsw/Kconfig index 695adff89d71..d56eea310509 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/Kconfig +++ b/drivers/net/ethernet/mellanox/mlxsw/Kconfig @@ -75,6 +75,7 @@ config MLXSW_SPECTRUM depends on MLXSW_CORE && MLXSW_PCI && NET_SWITCHDEV && VLAN_8021Q depends on PSAMPLE || PSAMPLE=n depends on BRIDGE || BRIDGE=n + depends on IPV6 || IPV6=n select PARMAN select MLXFW default m diff --git a/drivers/net/ethernet/mellanox/mlxsw/core.c b/drivers/net/ethernet/mellanox/mlxsw/core.c index affe84eb4bff..9d5e7cf288be 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/core.c +++ b/drivers/net/ethernet/mellanox/mlxsw/core.c @@ -667,7 +667,7 @@ static int mlxsw_emad_reg_access(struct mlxsw_core *mlxsw_core, int err; dev_dbg(mlxsw_core->bus_info->dev, "EMAD reg access (tid=%llx,reg_id=%x(%s),type=%s)\n", - trans->tid, reg->id, mlxsw_reg_id_str(reg->id), + tid, reg->id, mlxsw_reg_id_str(reg->id), mlxsw_core_reg_access_type_str(type)); skb = mlxsw_emad_alloc(mlxsw_core, reg->len); diff --git a/drivers/net/ethernet/mellanox/mlxsw/core_acl_flex_keys.h b/drivers/net/ethernet/mellanox/mlxsw/core_acl_flex_keys.h index 9807ef814e42..f6963b0b4a55 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/core_acl_flex_keys.h +++ b/drivers/net/ethernet/mellanox/mlxsw/core_acl_flex_keys.h @@ -57,6 +57,9 @@ enum mlxsw_afk_element { MLXSW_AFK_ELEMENT_VID, MLXSW_AFK_ELEMENT_PCP, MLXSW_AFK_ELEMENT_TCP_FLAGS, + MLXSW_AFK_ELEMENT_IP_TTL_, + MLXSW_AFK_ELEMENT_IP_ECN, + MLXSW_AFK_ELEMENT_IP_DSCP, MLXSW_AFK_ELEMENT_MAX, }; @@ -104,6 +107,9 @@ static const struct mlxsw_afk_element_info mlxsw_afk_element_infos[] = { MLXSW_AFK_ELEMENT_INFO_U32(VID, 0x10, 8, 12), MLXSW_AFK_ELEMENT_INFO_U32(PCP, 0x10, 20, 3), MLXSW_AFK_ELEMENT_INFO_U32(TCP_FLAGS, 0x10, 23, 9), + MLXSW_AFK_ELEMENT_INFO_U32(IP_TTL_, 0x14, 0, 8), + MLXSW_AFK_ELEMENT_INFO_U32(IP_ECN, 0x14, 9, 2), + MLXSW_AFK_ELEMENT_INFO_U32(IP_DSCP, 0x14, 11, 6), MLXSW_AFK_ELEMENT_INFO_U32(SRC_IP4, 0x18, 0, 32), MLXSW_AFK_ELEMENT_INFO_U32(DST_IP4, 0x1C, 0, 32), MLXSW_AFK_ELEMENT_INFO_BUF(SRC_IP6_HI, 0x18, 8), diff --git a/drivers/net/ethernet/mellanox/mlxsw/reg.h b/drivers/net/ethernet/mellanox/mlxsw/reg.h index 1bd34d9a7b9e..7e8ba546c3a4 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/reg.h +++ b/drivers/net/ethernet/mellanox/mlxsw/reg.h @@ -3679,15 +3679,17 @@ enum mlxsw_reg_htgt_trap_group { MLXSW_REG_HTGT_TRAP_GROUP_SP_LACP, MLXSW_REG_HTGT_TRAP_GROUP_SP_LLDP, MLXSW_REG_HTGT_TRAP_GROUP_SP_IGMP, - MLXSW_REG_HTGT_TRAP_GROUP_SP_BGP_IPV4, + MLXSW_REG_HTGT_TRAP_GROUP_SP_BGP, MLXSW_REG_HTGT_TRAP_GROUP_SP_OSPF, MLXSW_REG_HTGT_TRAP_GROUP_SP_ARP, - MLXSW_REG_HTGT_TRAP_GROUP_SP_ARP_MISS, + MLXSW_REG_HTGT_TRAP_GROUP_SP_HOST_MISS, MLXSW_REG_HTGT_TRAP_GROUP_SP_ROUTER_EXP, MLXSW_REG_HTGT_TRAP_GROUP_SP_REMOTE_ROUTE, MLXSW_REG_HTGT_TRAP_GROUP_SP_IP2ME, MLXSW_REG_HTGT_TRAP_GROUP_SP_DHCP, MLXSW_REG_HTGT_TRAP_GROUP_SP_EVENT, + MLXSW_REG_HTGT_TRAP_GROUP_SP_IPV6_MLD, + MLXSW_REG_HTGT_TRAP_GROUP_SP_IPV6_ND, }; /* reg_htgt_trap_group @@ -3952,10 +3954,12 @@ MLXSW_ITEM32(reg, rgcr, pcp_rw, 0x18, 16, 2); */ MLXSW_ITEM32(reg, rgcr, activity_dis, 0x20, 0, 8); -static inline void mlxsw_reg_rgcr_pack(char *payload, bool ipv4_en) +static inline void mlxsw_reg_rgcr_pack(char *payload, bool ipv4_en, + bool ipv6_en) { MLXSW_REG_ZERO(rgcr, payload); mlxsw_reg_rgcr_ipv4_en_set(payload, ipv4_en); + mlxsw_reg_rgcr_ipv6_en_set(payload, ipv6_en); } /* RITR - Router Interface Table Register @@ -3988,16 +3992,16 @@ MLXSW_ITEM32(reg, ritr, ipv4, 0x00, 29, 1); MLXSW_ITEM32(reg, ritr, ipv6, 0x00, 28, 1); enum mlxsw_reg_ritr_if_type { + /* VLAN interface. */ MLXSW_REG_RITR_VLAN_IF, + /* FID interface. */ MLXSW_REG_RITR_FID_IF, + /* Sub-port interface. */ MLXSW_REG_RITR_SP_IF, }; /* reg_ritr_type - * Router interface type. - * 0 - VLAN interface. - * 1 - FID interface. - * 2 - Sub-port interface. + * Router interface type as per enum mlxsw_reg_ritr_if_type. * Access: RW */ MLXSW_ITEM32(reg, ritr, type, 0x00, 23, 3); @@ -4203,10 +4207,12 @@ static inline void mlxsw_reg_ritr_pack(char *payload, bool enable, MLXSW_REG_ZERO(ritr, payload); mlxsw_reg_ritr_enable_set(payload, enable); mlxsw_reg_ritr_ipv4_set(payload, 1); + mlxsw_reg_ritr_ipv6_set(payload, 1); mlxsw_reg_ritr_type_set(payload, type); mlxsw_reg_ritr_op_set(payload, op); mlxsw_reg_ritr_rif_set(payload, rif); mlxsw_reg_ritr_ipv4_fe_set(payload, 1); + mlxsw_reg_ritr_ipv6_fe_set(payload, 1); mlxsw_reg_ritr_lb_en_set(payload, 1); mlxsw_reg_ritr_virtual_router_set(payload, vr_id); mlxsw_reg_ritr_mtu_set(payload, mtu); @@ -4712,12 +4718,13 @@ MLXSW_ITEM32(reg, ralue, prefix_len, 0x08, 0, 8); /* reg_ralue_dip* * The prefix of the route or of the marker that the object of the LPM * is compared with. The most significant bits of the dip are the prefix. - * The list significant bits must be '0' if the prefix_len is smaller + * The least significant bits must be '0' if the prefix_len is smaller * than 128 for IPv6 or smaller than 32 for IPv4. * IPv4 address uses bits dip[31:0] and bits dip[127:32] are reserved. * Access: Index */ MLXSW_ITEM32(reg, ralue, dip4, 0x18, 0, 32); +MLXSW_ITEM_BUF(reg, ralue, dip6, 0x0C, 16); enum mlxsw_reg_ralue_entry_type { MLXSW_REG_RALUE_ENTRY_TYPE_MARKER_ENTRY = 1, @@ -4806,7 +4813,7 @@ MLXSW_ITEM32(reg, ralue, ecmp_size, 0x28, 0, 13); */ MLXSW_ITEM32(reg, ralue, local_erif, 0x24, 0, 16); -/* reg_ralue_v +/* reg_ralue_ip2me_v * Valid bit for the tunnel_ptr field. * If valid = 0 then trap to CPU as IP2ME trap ID. * If valid = 1 and the packet format allows NVE or IPinIP tunnel @@ -4816,15 +4823,15 @@ MLXSW_ITEM32(reg, ralue, local_erif, 0x24, 0, 16); * Only relevant in case of IP2ME action. * Access: RW */ -MLXSW_ITEM32(reg, ralue, v, 0x24, 31, 1); +MLXSW_ITEM32(reg, ralue, ip2me_v, 0x24, 31, 1); -/* reg_ralue_tunnel_ptr +/* reg_ralue_ip2me_tunnel_ptr * Tunnel Pointer for NVE or IPinIP tunnel decapsulation. * For Spectrum, pointer to KVD Linear. * Only relevant in case of IP2ME action. * Access: RW */ -MLXSW_ITEM32(reg, ralue, tunnel_ptr, 0x24, 0, 24); +MLXSW_ITEM32(reg, ralue, ip2me_tunnel_ptr, 0x24, 0, 24); static inline void mlxsw_reg_ralue_pack(char *payload, enum mlxsw_reg_ralxx_protocol protocol, @@ -4851,6 +4858,16 @@ static inline void mlxsw_reg_ralue_pack4(char *payload, mlxsw_reg_ralue_dip4_set(payload, dip); } +static inline void mlxsw_reg_ralue_pack6(char *payload, + enum mlxsw_reg_ralxx_protocol protocol, + enum mlxsw_reg_ralue_op op, + u16 virtual_router, u8 prefix_len, + const void *dip) +{ + mlxsw_reg_ralue_pack(payload, protocol, op, virtual_router, prefix_len); + mlxsw_reg_ralue_dip6_memcpy_to(payload, dip); +} + static inline void mlxsw_reg_ralue_act_remote_pack(char *payload, enum mlxsw_reg_ralue_trap_action trap_action, @@ -4954,6 +4971,7 @@ MLXSW_ITEM32(reg, rauht, rif, 0x00, 0, 16); * Access: Index */ MLXSW_ITEM32(reg, rauht, dip4, 0x1C, 0x0, 32); +MLXSW_ITEM_BUF(reg, rauht, dip6, 0x10, 16); enum mlxsw_reg_rauht_trap_action { MLXSW_REG_RAUHT_TRAP_ACTION_NOP, @@ -5018,6 +5036,15 @@ static inline void mlxsw_reg_rauht_pack4(char *payload, mlxsw_reg_rauht_dip4_set(payload, dip); } +static inline void mlxsw_reg_rauht_pack6(char *payload, + enum mlxsw_reg_rauht_op op, u16 rif, + const char *mac, const char *dip) +{ + mlxsw_reg_rauht_pack(payload, op, rif, mac); + mlxsw_reg_rauht_type_set(payload, MLXSW_REG_RAUHT_TYPE_IPV6); + mlxsw_reg_rauht_dip6_memcpy_to(payload, dip); +} + /* RALEU - Router Algorithmic LPM ECMP Update Register * --------------------------------------------------- * The register enables updating the ECMP section in the action for multiple @@ -5216,6 +5243,30 @@ MLXSW_ITEM32_INDEXED(reg, rauhtd, ipv4_ent_rif, MLXSW_REG_RAUHTD_BASE_LEN, 0, MLXSW_ITEM32_INDEXED(reg, rauhtd, ipv4_ent_dip, MLXSW_REG_RAUHTD_BASE_LEN, 0, 32, MLXSW_REG_RAUHTD_IPV4_ENT_LEN, 0x04, false); +#define MLXSW_REG_RAUHTD_IPV6_ENT_LEN 0x20 + +/* reg_rauhtd_ipv6_ent_a + * Activity. Set for new entries. Set if a packet lookup has hit on the + * specific entry. + * Access: RO + */ +MLXSW_ITEM32_INDEXED(reg, rauhtd, ipv6_ent_a, MLXSW_REG_RAUHTD_BASE_LEN, 16, 1, + MLXSW_REG_RAUHTD_IPV6_ENT_LEN, 0x00, false); + +/* reg_rauhtd_ipv6_ent_rif + * Router interface. + * Access: RO + */ +MLXSW_ITEM32_INDEXED(reg, rauhtd, ipv6_ent_rif, MLXSW_REG_RAUHTD_BASE_LEN, 0, + 16, MLXSW_REG_RAUHTD_IPV6_ENT_LEN, 0x00, false); + +/* reg_rauhtd_ipv6_ent_dip + * Destination IPv6 address. + * Access: RO + */ +MLXSW_ITEM_BUF_INDEXED(reg, rauhtd, ipv6_ent_dip, MLXSW_REG_RAUHTD_BASE_LEN, + 16, MLXSW_REG_RAUHTD_IPV6_ENT_LEN, 0x10); + static inline void mlxsw_reg_rauhtd_ent_ipv4_unpack(char *payload, int ent_index, u16 *p_rif, u32 *p_dip) @@ -5224,6 +5275,14 @@ static inline void mlxsw_reg_rauhtd_ent_ipv4_unpack(char *payload, *p_dip = mlxsw_reg_rauhtd_ipv4_ent_dip_get(payload, ent_index); } +static inline void mlxsw_reg_rauhtd_ent_ipv6_unpack(char *payload, + int rec_index, u16 *p_rif, + char *p_dip) +{ + *p_rif = mlxsw_reg_rauhtd_ipv6_ent_rif_get(payload, rec_index); + mlxsw_reg_rauhtd_ipv6_ent_dip_memcpy_from(payload, rec_index, p_dip); +} + /* MFCR - Management Fan Control Register * -------------------------------------- * This register controls the settings of the Fan Speed PWM mechanism. diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum.c index 60bf8f27cc00..eb7c4549f464 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum.c @@ -58,6 +58,7 @@ #include <net/tc_act/tc_mirred.h> #include <net/netevent.h> #include <net/tc_act/tc_sample.h> +#include <net/addrconf.h> #include "spectrum.h" #include "pci.h" @@ -1616,16 +1617,16 @@ mlxsw_sp_port_del_cls_matchall_sample(struct mlxsw_sp_port *mlxsw_sp_port) } static int mlxsw_sp_port_add_cls_matchall(struct mlxsw_sp_port *mlxsw_sp_port, - __be16 protocol, - struct tc_cls_matchall_offload *cls, + struct tc_cls_matchall_offload *f, bool ingress) { struct mlxsw_sp_port_mall_tc_entry *mall_tc_entry; + __be16 protocol = f->common.protocol; const struct tc_action *a; LIST_HEAD(actions); int err; - if (!tc_single_action(cls->exts)) { + if (!tcf_exts_has_one_action(f->exts)) { netdev_err(mlxsw_sp_port->dev, "only singular actions are supported\n"); return -EOPNOTSUPP; } @@ -1633,9 +1634,9 @@ static int mlxsw_sp_port_add_cls_matchall(struct mlxsw_sp_port *mlxsw_sp_port, mall_tc_entry = kzalloc(sizeof(*mall_tc_entry), GFP_KERNEL); if (!mall_tc_entry) return -ENOMEM; - mall_tc_entry->cookie = cls->cookie; + mall_tc_entry->cookie = f->cookie; - tcf_exts_to_list(cls->exts, &actions); + tcf_exts_to_list(f->exts, &actions); a = list_first_entry(&actions, struct tc_action, list); if (is_tcf_mirred_egress_mirror(a) && protocol == htons(ETH_P_ALL)) { @@ -1647,7 +1648,7 @@ static int mlxsw_sp_port_add_cls_matchall(struct mlxsw_sp_port *mlxsw_sp_port, mirror, a, ingress); } else if (is_tcf_sample(a) && protocol == htons(ETH_P_ALL)) { mall_tc_entry->type = MLXSW_SP_PORT_MALL_SAMPLE; - err = mlxsw_sp_port_add_cls_matchall_sample(mlxsw_sp_port, cls, + err = mlxsw_sp_port_add_cls_matchall_sample(mlxsw_sp_port, f, a, ingress); } else { err = -EOPNOTSUPP; @@ -1665,12 +1666,12 @@ err_add_action: } static void mlxsw_sp_port_del_cls_matchall(struct mlxsw_sp_port *mlxsw_sp_port, - struct tc_cls_matchall_offload *cls) + struct tc_cls_matchall_offload *f) { struct mlxsw_sp_port_mall_tc_entry *mall_tc_entry; mall_tc_entry = mlxsw_sp_port_mall_tc_entry_find(mlxsw_sp_port, - cls->cookie); + f->cookie); if (!mall_tc_entry) { netdev_dbg(mlxsw_sp_port->dev, "tc entry not found on port\n"); return; @@ -1692,49 +1693,61 @@ static void mlxsw_sp_port_del_cls_matchall(struct mlxsw_sp_port *mlxsw_sp_port, kfree(mall_tc_entry); } -static int mlxsw_sp_setup_tc(struct net_device *dev, u32 handle, - u32 chain_index, __be16 proto, - struct tc_to_netdev *tc) +static int mlxsw_sp_setup_tc_cls_matchall(struct mlxsw_sp_port *mlxsw_sp_port, + struct tc_cls_matchall_offload *f) { - struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); - bool ingress = TC_H_MAJ(handle) == TC_H_MAJ(TC_H_INGRESS); + bool ingress = TC_H_MAJ(f->common.handle) == TC_H_MAJ(TC_H_INGRESS); - if (chain_index) + if (f->common.chain_index) return -EOPNOTSUPP; - switch (tc->type) { - case TC_SETUP_MATCHALL: - switch (tc->cls_mall->command) { - case TC_CLSMATCHALL_REPLACE: - return mlxsw_sp_port_add_cls_matchall(mlxsw_sp_port, - proto, - tc->cls_mall, - ingress); - case TC_CLSMATCHALL_DESTROY: - mlxsw_sp_port_del_cls_matchall(mlxsw_sp_port, - tc->cls_mall); - return 0; - default: - return -EOPNOTSUPP; - } - case TC_SETUP_CLSFLOWER: - switch (tc->cls_flower->command) { - case TC_CLSFLOWER_REPLACE: - return mlxsw_sp_flower_replace(mlxsw_sp_port, ingress, - proto, tc->cls_flower); - case TC_CLSFLOWER_DESTROY: - mlxsw_sp_flower_destroy(mlxsw_sp_port, ingress, - tc->cls_flower); - return 0; - case TC_CLSFLOWER_STATS: - return mlxsw_sp_flower_stats(mlxsw_sp_port, ingress, - tc->cls_flower); - default: - return -EOPNOTSUPP; - } + switch (f->command) { + case TC_CLSMATCHALL_REPLACE: + return mlxsw_sp_port_add_cls_matchall(mlxsw_sp_port, f, + ingress); + case TC_CLSMATCHALL_DESTROY: + mlxsw_sp_port_del_cls_matchall(mlxsw_sp_port, f); + return 0; + default: + return -EOPNOTSUPP; + } +} + +static int +mlxsw_sp_setup_tc_cls_flower(struct mlxsw_sp_port *mlxsw_sp_port, + struct tc_cls_flower_offload *f) +{ + bool ingress = TC_H_MAJ(f->common.handle) == TC_H_MAJ(TC_H_INGRESS); + + if (f->common.chain_index) + return -EOPNOTSUPP; + + switch (f->command) { + case TC_CLSFLOWER_REPLACE: + return mlxsw_sp_flower_replace(mlxsw_sp_port, ingress, f); + case TC_CLSFLOWER_DESTROY: + mlxsw_sp_flower_destroy(mlxsw_sp_port, ingress, f); + return 0; + case TC_CLSFLOWER_STATS: + return mlxsw_sp_flower_stats(mlxsw_sp_port, ingress, f); + default: + return -EOPNOTSUPP; } +} + +static int mlxsw_sp_setup_tc(struct net_device *dev, enum tc_setup_type type, + void *type_data) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); - return -EOPNOTSUPP; + switch (type) { + case TC_SETUP_CLSMATCHALL: + return mlxsw_sp_setup_tc_cls_matchall(mlxsw_sp_port, type_data); + case TC_SETUP_CLSFLOWER: + return mlxsw_sp_setup_tc_cls_flower(mlxsw_sp_port, type_data); + default: + return -EOPNOTSUPP; + } } static const struct net_device_ops mlxsw_sp_port_netdev_ops = { @@ -3333,15 +3346,47 @@ static const struct mlxsw_listener mlxsw_sp_listener[] = { MLXSW_SP_RXL_MARK(ARPBC, MIRROR_TO_CPU, ARP, false), MLXSW_SP_RXL_MARK(ARPUC, MIRROR_TO_CPU, ARP, false), MLXSW_SP_RXL_NO_MARK(FID_MISS, TRAP_TO_CPU, IP2ME, false), + MLXSW_SP_RXL_MARK(IPV6_MLDV12_LISTENER_QUERY, MIRROR_TO_CPU, IPV6_MLD, + false), + MLXSW_SP_RXL_NO_MARK(IPV6_MLDV1_LISTENER_REPORT, TRAP_TO_CPU, IPV6_MLD, + false), + MLXSW_SP_RXL_NO_MARK(IPV6_MLDV1_LISTENER_DONE, TRAP_TO_CPU, IPV6_MLD, + false), + MLXSW_SP_RXL_NO_MARK(IPV6_MLDV2_LISTENER_REPORT, TRAP_TO_CPU, IPV6_MLD, + false), /* L3 traps */ - MLXSW_SP_RXL_NO_MARK(MTUERROR, TRAP_TO_CPU, ROUTER_EXP, false), - MLXSW_SP_RXL_NO_MARK(TTLERROR, TRAP_TO_CPU, ROUTER_EXP, false), - MLXSW_SP_RXL_NO_MARK(LBERROR, TRAP_TO_CPU, ROUTER_EXP, false), - MLXSW_SP_RXL_MARK(OSPF, TRAP_TO_CPU, OSPF, false), - MLXSW_SP_RXL_NO_MARK(IP2ME, TRAP_TO_CPU, IP2ME, false), - MLXSW_SP_RXL_NO_MARK(RTR_INGRESS0, TRAP_TO_CPU, REMOTE_ROUTE, false), - MLXSW_SP_RXL_NO_MARK(HOST_MISS_IPV4, TRAP_TO_CPU, ARP_MISS, false), - MLXSW_SP_RXL_NO_MARK(BGP_IPV4, TRAP_TO_CPU, BGP_IPV4, false), + MLXSW_SP_RXL_MARK(MTUERROR, TRAP_TO_CPU, ROUTER_EXP, false), + MLXSW_SP_RXL_MARK(TTLERROR, TRAP_TO_CPU, ROUTER_EXP, false), + MLXSW_SP_RXL_MARK(LBERROR, TRAP_TO_CPU, ROUTER_EXP, false), + MLXSW_SP_RXL_MARK(IP2ME, TRAP_TO_CPU, IP2ME, false), + MLXSW_SP_RXL_MARK(IPV6_UNSPECIFIED_ADDRESS, TRAP_TO_CPU, ROUTER_EXP, + false), + MLXSW_SP_RXL_MARK(IPV6_LINK_LOCAL_DEST, TRAP_TO_CPU, ROUTER_EXP, false), + MLXSW_SP_RXL_MARK(IPV6_LINK_LOCAL_SRC, TRAP_TO_CPU, ROUTER_EXP, false), + MLXSW_SP_RXL_MARK(IPV6_ALL_NODES_LINK, TRAP_TO_CPU, ROUTER_EXP, false), + MLXSW_SP_RXL_MARK(IPV6_ALL_ROUTERS_LINK, TRAP_TO_CPU, ROUTER_EXP, + false), + MLXSW_SP_RXL_MARK(IPV4_OSPF, TRAP_TO_CPU, OSPF, false), + MLXSW_SP_RXL_MARK(IPV6_OSPF, TRAP_TO_CPU, OSPF, false), + MLXSW_SP_RXL_MARK(IPV6_DHCP, TRAP_TO_CPU, DHCP, false), + MLXSW_SP_RXL_MARK(RTR_INGRESS0, TRAP_TO_CPU, REMOTE_ROUTE, false), + MLXSW_SP_RXL_MARK(IPV4_BGP, TRAP_TO_CPU, BGP, false), + MLXSW_SP_RXL_MARK(IPV6_BGP, TRAP_TO_CPU, BGP, false), + MLXSW_SP_RXL_MARK(L3_IPV6_ROUTER_SOLICITATION, TRAP_TO_CPU, IPV6_ND, + false), + MLXSW_SP_RXL_MARK(L3_IPV6_ROUTER_ADVERTISMENT, TRAP_TO_CPU, IPV6_ND, + false), + MLXSW_SP_RXL_MARK(L3_IPV6_NEIGHBOR_SOLICITATION, TRAP_TO_CPU, IPV6_ND, + false), + MLXSW_SP_RXL_MARK(L3_IPV6_NEIGHBOR_ADVERTISMENT, TRAP_TO_CPU, IPV6_ND, + false), + MLXSW_SP_RXL_MARK(L3_IPV6_REDIRECTION, TRAP_TO_CPU, IPV6_ND, false), + MLXSW_SP_RXL_MARK(IPV6_MC_LINK_LOCAL_DEST, TRAP_TO_CPU, ROUTER_EXP, + false), + MLXSW_SP_RXL_MARK(HOST_MISS_IPV4, TRAP_TO_CPU, HOST_MISS, false), + MLXSW_SP_RXL_MARK(HOST_MISS_IPV6, TRAP_TO_CPU, HOST_MISS, false), + MLXSW_SP_RXL_MARK(ROUTER_ALERT_IPV4, TRAP_TO_CPU, ROUTER_EXP, false), + MLXSW_SP_RXL_MARK(ROUTER_ALERT_IPV6, TRAP_TO_CPU, ROUTER_EXP, false), /* PKT Sample trap */ MLXSW_RXL(mlxsw_sp_rx_listener_sample_func, PKT_SAMPLE, MIRROR_TO_CPU, false, SP_IP2ME, DISCARD), @@ -3376,15 +3421,17 @@ static int mlxsw_sp_cpu_policers_set(struct mlxsw_core *mlxsw_core) burst_size = 7; break; case MLXSW_REG_HTGT_TRAP_GROUP_SP_IGMP: + case MLXSW_REG_HTGT_TRAP_GROUP_SP_IPV6_MLD: rate = 16 * 1024; burst_size = 10; break; - case MLXSW_REG_HTGT_TRAP_GROUP_SP_BGP_IPV4: + case MLXSW_REG_HTGT_TRAP_GROUP_SP_BGP: case MLXSW_REG_HTGT_TRAP_GROUP_SP_ARP: case MLXSW_REG_HTGT_TRAP_GROUP_SP_DHCP: - case MLXSW_REG_HTGT_TRAP_GROUP_SP_ARP_MISS: + case MLXSW_REG_HTGT_TRAP_GROUP_SP_HOST_MISS: case MLXSW_REG_HTGT_TRAP_GROUP_SP_ROUTER_EXP: case MLXSW_REG_HTGT_TRAP_GROUP_SP_REMOTE_ROUTE: + case MLXSW_REG_HTGT_TRAP_GROUP_SP_IPV6_ND: rate = 1024; burst_size = 7; break; @@ -3433,21 +3480,23 @@ static int mlxsw_sp_trap_groups_set(struct mlxsw_core *mlxsw_core) priority = 5; tc = 5; break; - case MLXSW_REG_HTGT_TRAP_GROUP_SP_BGP_IPV4: + case MLXSW_REG_HTGT_TRAP_GROUP_SP_BGP: case MLXSW_REG_HTGT_TRAP_GROUP_SP_DHCP: priority = 4; tc = 4; break; case MLXSW_REG_HTGT_TRAP_GROUP_SP_IGMP: case MLXSW_REG_HTGT_TRAP_GROUP_SP_IP2ME: + case MLXSW_REG_HTGT_TRAP_GROUP_SP_IPV6_MLD: priority = 3; tc = 3; break; case MLXSW_REG_HTGT_TRAP_GROUP_SP_ARP: + case MLXSW_REG_HTGT_TRAP_GROUP_SP_IPV6_ND: priority = 2; tc = 2; break; - case MLXSW_REG_HTGT_TRAP_GROUP_SP_ARP_MISS: + case MLXSW_REG_HTGT_TRAP_GROUP_SP_HOST_MISS: case MLXSW_REG_HTGT_TRAP_GROUP_SP_ROUTER_EXP: case MLXSW_REG_HTGT_TRAP_GROUP_SP_REMOTE_ROUTE: priority = 1; @@ -4357,6 +4406,10 @@ static struct notifier_block mlxsw_sp_inetaddr_nb __read_mostly = { .priority = 10, /* Must be called before FIB notifier block */ }; +static struct notifier_block mlxsw_sp_inet6addr_nb __read_mostly = { + .notifier_call = mlxsw_sp_inet6addr_event, +}; + static struct notifier_block mlxsw_sp_router_netevent_nb __read_mostly = { .notifier_call = mlxsw_sp_router_netevent_event, }; @@ -4377,6 +4430,7 @@ static int __init mlxsw_sp_module_init(void) register_netdevice_notifier(&mlxsw_sp_netdevice_nb); register_inetaddr_notifier(&mlxsw_sp_inetaddr_nb); + register_inet6addr_notifier(&mlxsw_sp_inet6addr_nb); register_netevent_notifier(&mlxsw_sp_router_netevent_nb); err = mlxsw_core_driver_register(&mlxsw_sp_driver); @@ -4393,6 +4447,7 @@ err_pci_driver_register: mlxsw_core_driver_unregister(&mlxsw_sp_driver); err_core_driver_register: unregister_netevent_notifier(&mlxsw_sp_router_netevent_nb); + unregister_inet6addr_notifier(&mlxsw_sp_inet6addr_nb); unregister_inetaddr_notifier(&mlxsw_sp_inetaddr_nb); unregister_netdevice_notifier(&mlxsw_sp_netdevice_nb); return err; @@ -4403,6 +4458,7 @@ static void __exit mlxsw_sp_module_exit(void) mlxsw_pci_driver_unregister(&mlxsw_sp_pci_driver); mlxsw_core_driver_unregister(&mlxsw_sp_driver); unregister_netevent_notifier(&mlxsw_sp_router_netevent_nb); + unregister_inet6addr_notifier(&mlxsw_sp_inet6addr_nb); unregister_inetaddr_notifier(&mlxsw_sp_inetaddr_nb); unregister_netdevice_notifier(&mlxsw_sp_netdevice_nb); } diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum.h index 5ef98d4d0ab6..8452d1db2f3f 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum.h +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum.h @@ -384,6 +384,8 @@ int mlxsw_sp_router_netevent_event(struct notifier_block *unused, int mlxsw_sp_netdevice_router_port_event(struct net_device *dev); int mlxsw_sp_inetaddr_event(struct notifier_block *unused, unsigned long event, void *ptr); +int mlxsw_sp_inet6addr_event(struct notifier_block *unused, + unsigned long event, void *ptr); int mlxsw_sp_netdevice_vrf_event(struct net_device *l3_dev, unsigned long event, struct netdev_notifier_changeupper_info *info); void @@ -506,7 +508,7 @@ extern const struct mlxsw_sp_acl_ops mlxsw_sp_acl_tcam_ops; /* spectrum_flower.c */ int mlxsw_sp_flower_replace(struct mlxsw_sp_port *mlxsw_sp_port, bool ingress, - __be16 protocol, struct tc_cls_flower_offload *f); + struct tc_cls_flower_offload *f); void mlxsw_sp_flower_destroy(struct mlxsw_sp_port *mlxsw_sp_port, bool ingress, struct tc_cls_flower_offload *f); int mlxsw_sp_flower_stats(struct mlxsw_sp_port *mlxsw_sp_port, bool ingress, diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl.c index 01a1501b56ca..508b5fcacd77 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl.c @@ -369,7 +369,7 @@ int mlxsw_sp_acl_rulei_act_fwd(struct mlxsw_sp *mlxsw_sp, local_port = mlxsw_sp_port->local_port; in_port = false; } else { - /* If out_dev is NULL, the called wants to + /* If out_dev is NULL, the caller wants to * set forward to ingress port. */ local_port = 0; diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_flex_keys.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_flex_keys.h index 85d5001a5818..fb8031828454 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_flex_keys.h +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_flex_keys.h @@ -70,6 +70,9 @@ static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_ipv4_dip[] = { static struct mlxsw_afk_element_inst mlxsw_sp_afk_element_info_ipv4[] = { MLXSW_AFK_ELEMENT_INST_U32(SRC_IP4, 0x00, 0, 32), + MLXSW_AFK_ELEMENT_INST_U32(IP_ECN, 0x04, 4, 2), + MLXSW_AFK_ELEMENT_INST_U32(IP_TTL_, 0x04, 24, 8), + MLXSW_AFK_ELEMENT_INST_U32(IP_DSCP, 0x08, 0, 6), MLXSW_AFK_ELEMENT_INST_U32(TCP_FLAGS, 0x08, 8, 9), /* TCP_CONTROL+TCP_ECN */ }; diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c index 61a10f166f97..bc5173f1b5c1 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c @@ -984,6 +984,9 @@ static const enum mlxsw_afk_element mlxsw_sp_acl_tcam_pattern_ipv4[] = { MLXSW_AFK_ELEMENT_VID, MLXSW_AFK_ELEMENT_PCP, MLXSW_AFK_ELEMENT_TCP_FLAGS, + MLXSW_AFK_ELEMENT_IP_TTL_, + MLXSW_AFK_ELEMENT_IP_ECN, + MLXSW_AFK_ELEMENT_IP_DSCP, }; static const enum mlxsw_afk_element mlxsw_sp_acl_tcam_pattern_ipv6[] = { diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_fid.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_fid.c index 6afbe9ec64e2..bbd238e50f05 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_fid.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_fid.c @@ -109,7 +109,6 @@ static const int mlxsw_sp_sfgc_uc_packet_types[MLXSW_REG_SFGC_TYPE_MAX] = { static const int mlxsw_sp_sfgc_bc_packet_types[MLXSW_REG_SFGC_TYPE_MAX] = { [MLXSW_REG_SFGC_TYPE_BROADCAST] = 1, - [MLXSW_REG_SFGC_TYPE_UNREGISTERED_MULTICAST_IPV6] = 1, [MLXSW_REG_SFGC_TYPE_UNREGISTERED_MULTICAST_NON_IP] = 1, [MLXSW_REG_SFGC_TYPE_IPV4_LINK_LOCAL] = 1, [MLXSW_REG_SFGC_TYPE_IPV6_ALL_HOST] = 1, @@ -117,6 +116,7 @@ static const int mlxsw_sp_sfgc_bc_packet_types[MLXSW_REG_SFGC_TYPE_MAX] = { static const int mlxsw_sp_sfgc_mc_packet_types[MLXSW_REG_SFGC_TYPE_MAX] = { [MLXSW_REG_SFGC_TYPE_UNREGISTERED_MULTICAST_IPV4] = 1, + [MLXSW_REG_SFGC_TYPE_UNREGISTERED_MULTICAST_IPV6] = 1, }; static const int *mlxsw_sp_packet_type_sfgc_types[] = { diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_flower.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_flower.c index 21bb2bf62d3e..95428b41c50f 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_flower.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_flower.c @@ -53,7 +53,7 @@ static int mlxsw_sp_flower_parse_actions(struct mlxsw_sp *mlxsw_sp, LIST_HEAD(actions); int err; - if (tc_no_actions(exts)) + if (!tcf_exts_has_actions(exts)) return 0; /* Count action is inserted first */ @@ -212,11 +212,46 @@ static int mlxsw_sp_flower_parse_tcp(struct mlxsw_sp *mlxsw_sp, return 0; } +static int mlxsw_sp_flower_parse_ip(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule_info *rulei, + struct tc_cls_flower_offload *f, + u16 n_proto) +{ + struct flow_dissector_key_ip *key, *mask; + + if (!dissector_uses_key(f->dissector, FLOW_DISSECTOR_KEY_IP)) + return 0; + + if (n_proto != ETH_P_IP && n_proto != ETH_P_IPV6) { + dev_err(mlxsw_sp->bus_info->dev, "IP keys supported only for IPv4/6\n"); + return -EINVAL; + } + + key = skb_flow_dissector_target(f->dissector, + FLOW_DISSECTOR_KEY_IP, + f->key); + mask = skb_flow_dissector_target(f->dissector, + FLOW_DISSECTOR_KEY_IP, + f->mask); + mlxsw_sp_acl_rulei_keymask_u32(rulei, MLXSW_AFK_ELEMENT_IP_TTL_, + key->ttl, mask->ttl); + + mlxsw_sp_acl_rulei_keymask_u32(rulei, MLXSW_AFK_ELEMENT_IP_ECN, + key->tos & 0x3, mask->tos & 0x3); + + mlxsw_sp_acl_rulei_keymask_u32(rulei, MLXSW_AFK_ELEMENT_IP_DSCP, + key->tos >> 6, mask->tos >> 6); + + return 0; +} + static int mlxsw_sp_flower_parse(struct mlxsw_sp *mlxsw_sp, struct net_device *dev, struct mlxsw_sp_acl_rule_info *rulei, struct tc_cls_flower_offload *f) { + u16 n_proto_mask = 0; + u16 n_proto_key = 0; u16 addr_type = 0; u8 ip_proto = 0; int err; @@ -229,12 +264,13 @@ static int mlxsw_sp_flower_parse(struct mlxsw_sp *mlxsw_sp, BIT(FLOW_DISSECTOR_KEY_IPV6_ADDRS) | BIT(FLOW_DISSECTOR_KEY_PORTS) | BIT(FLOW_DISSECTOR_KEY_TCP) | + BIT(FLOW_DISSECTOR_KEY_IP) | BIT(FLOW_DISSECTOR_KEY_VLAN))) { dev_err(mlxsw_sp->bus_info->dev, "Unsupported key\n"); return -EOPNOTSUPP; } - mlxsw_sp_acl_rulei_priority(rulei, f->prio); + mlxsw_sp_acl_rulei_priority(rulei, f->common.prio); if (dissector_uses_key(f->dissector, FLOW_DISSECTOR_KEY_CONTROL)) { struct flow_dissector_key_control *key = @@ -253,8 +289,8 @@ static int mlxsw_sp_flower_parse(struct mlxsw_sp *mlxsw_sp, skb_flow_dissector_target(f->dissector, FLOW_DISSECTOR_KEY_BASIC, f->mask); - u16 n_proto_key = ntohs(key->n_proto); - u16 n_proto_mask = ntohs(mask->n_proto); + n_proto_key = ntohs(key->n_proto); + n_proto_mask = ntohs(mask->n_proto); if (n_proto_key == ETH_P_ALL) { n_proto_key = 0; @@ -324,11 +360,15 @@ static int mlxsw_sp_flower_parse(struct mlxsw_sp *mlxsw_sp, if (err) return err; + err = mlxsw_sp_flower_parse_ip(mlxsw_sp, rulei, f, n_proto_key & n_proto_mask); + if (err) + return err; + return mlxsw_sp_flower_parse_actions(mlxsw_sp, dev, rulei, f->exts); } int mlxsw_sp_flower_replace(struct mlxsw_sp_port *mlxsw_sp_port, bool ingress, - __be16 protocol, struct tc_cls_flower_offload *f) + struct tc_cls_flower_offload *f) { struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; struct net_device *dev = mlxsw_sp_port->dev; diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c index 4b2e0fd7d51e..93b6da88e79c 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c @@ -43,12 +43,19 @@ #include <linux/inetdevice.h> #include <linux/netdevice.h> #include <linux/if_bridge.h> +#include <linux/socket.h> +#include <linux/route.h> #include <net/netevent.h> #include <net/neighbour.h> #include <net/arp.h> #include <net/ip_fib.h> +#include <net/ip6_fib.h> #include <net/fib_rules.h> #include <net/l3mdev.h> +#include <net/addrconf.h> +#include <net/ndisc.h> +#include <net/ipv6.h> +#include <net/fib_notifier.h> #include "spectrum.h" #include "core.h" @@ -304,7 +311,7 @@ static struct mlxsw_sp_rif * mlxsw_sp_rif_find_by_dev(const struct mlxsw_sp *mlxsw_sp, const struct net_device *dev); -#define MLXSW_SP_PREFIX_COUNT (sizeof(struct in6_addr) * BITS_PER_BYTE) +#define MLXSW_SP_PREFIX_COUNT (sizeof(struct in6_addr) * BITS_PER_BYTE + 1) struct mlxsw_sp_prefix_usage { DECLARE_BITMAP(b, MLXSW_SP_PREFIX_COUNT); @@ -384,21 +391,31 @@ struct mlxsw_sp_fib_node { struct mlxsw_sp_fib_key key; }; -struct mlxsw_sp_fib_entry_params { +struct mlxsw_sp_fib_entry { + struct list_head list; + struct mlxsw_sp_fib_node *fib_node; + enum mlxsw_sp_fib_entry_type type; + struct list_head nexthop_group_node; + struct mlxsw_sp_nexthop_group *nh_group; +}; + +struct mlxsw_sp_fib4_entry { + struct mlxsw_sp_fib_entry common; u32 tb_id; u32 prio; u8 tos; u8 type; }; -struct mlxsw_sp_fib_entry { +struct mlxsw_sp_fib6_entry { + struct mlxsw_sp_fib_entry common; + struct list_head rt6_list; + unsigned int nrt6; +}; + +struct mlxsw_sp_rt6 { struct list_head list; - struct mlxsw_sp_fib_node *fib_node; - enum mlxsw_sp_fib_entry_type type; - struct list_head nexthop_group_node; - struct mlxsw_sp_nexthop_group *nh_group; - struct mlxsw_sp_fib_entry_params params; - bool offloaded; + struct rt6_info *rt; }; enum mlxsw_sp_l3proto { @@ -428,6 +445,7 @@ struct mlxsw_sp_vr { u32 tb_id; /* kernel fib table id */ unsigned int rif_count; struct mlxsw_sp_fib *fib4; + struct mlxsw_sp_fib *fib6; }; static const struct rhashtable_params mlxsw_sp_fib_ht_params; @@ -625,7 +643,7 @@ static void mlxsw_sp_lpm_fini(struct mlxsw_sp *mlxsw_sp) static bool mlxsw_sp_vr_is_used(const struct mlxsw_sp_vr *vr) { - return !!vr->fib4; + return !!vr->fib4 || !!vr->fib6; } static struct mlxsw_sp_vr *mlxsw_sp_vr_find_unused(struct mlxsw_sp *mlxsw_sp) @@ -694,7 +712,7 @@ static struct mlxsw_sp_fib *mlxsw_sp_vr_fib(const struct mlxsw_sp_vr *vr, case MLXSW_SP_L3_PROTO_IPV4: return vr->fib4; case MLXSW_SP_L3_PROTO_IPV6: - BUG_ON(1); + return vr->fib6; } return NULL; } @@ -703,6 +721,7 @@ static struct mlxsw_sp_vr *mlxsw_sp_vr_create(struct mlxsw_sp *mlxsw_sp, u32 tb_id) { struct mlxsw_sp_vr *vr; + int err; vr = mlxsw_sp_vr_find_unused(mlxsw_sp); if (!vr) @@ -710,12 +729,24 @@ static struct mlxsw_sp_vr *mlxsw_sp_vr_create(struct mlxsw_sp *mlxsw_sp, vr->fib4 = mlxsw_sp_fib_create(vr, MLXSW_SP_L3_PROTO_IPV4); if (IS_ERR(vr->fib4)) return ERR_CAST(vr->fib4); + vr->fib6 = mlxsw_sp_fib_create(vr, MLXSW_SP_L3_PROTO_IPV6); + if (IS_ERR(vr->fib6)) { + err = PTR_ERR(vr->fib6); + goto err_fib6_create; + } vr->tb_id = tb_id; return vr; + +err_fib6_create: + mlxsw_sp_fib_destroy(vr->fib4); + vr->fib4 = NULL; + return ERR_PTR(err); } static void mlxsw_sp_vr_destroy(struct mlxsw_sp_vr *vr) { + mlxsw_sp_fib_destroy(vr->fib6); + vr->fib6 = NULL; mlxsw_sp_fib_destroy(vr->fib4); vr->fib4 = NULL; } @@ -773,7 +804,8 @@ static struct mlxsw_sp_vr *mlxsw_sp_vr_get(struct mlxsw_sp *mlxsw_sp, u32 tb_id) static void mlxsw_sp_vr_put(struct mlxsw_sp_vr *vr) { - if (!vr->rif_count && list_empty(&vr->fib4->node_list)) + if (!vr->rif_count && list_empty(&vr->fib4->node_list) && + list_empty(&vr->fib6->node_list)) mlxsw_sp_vr_destroy(vr); } @@ -929,8 +961,15 @@ mlxsw_sp_neigh_entry_lookup(struct mlxsw_sp *mlxsw_sp, struct neighbour *n) static void mlxsw_sp_router_neighs_update_interval_init(struct mlxsw_sp *mlxsw_sp) { - unsigned long interval = NEIGH_VAR(&arp_tbl.parms, DELAY_PROBE_TIME); + unsigned long interval; +#if IS_ENABLED(CONFIG_IPV6) + interval = min_t(unsigned long, + NEIGH_VAR(&arp_tbl.parms, DELAY_PROBE_TIME), + NEIGH_VAR(&nd_tbl.parms, DELAY_PROBE_TIME)); +#else + interval = NEIGH_VAR(&arp_tbl.parms, DELAY_PROBE_TIME); +#endif mlxsw_sp->router->neighs_update.interval = jiffies_to_msecs(interval); } @@ -965,6 +1004,44 @@ static void mlxsw_sp_router_neigh_ent_ipv4_process(struct mlxsw_sp *mlxsw_sp, neigh_release(n); } +#if IS_ENABLED(IPV6) +static void mlxsw_sp_router_neigh_ent_ipv6_process(struct mlxsw_sp *mlxsw_sp, + char *rauhtd_pl, + int rec_index) +{ + struct net_device *dev; + struct neighbour *n; + struct in6_addr dip; + u16 rif; + + mlxsw_reg_rauhtd_ent_ipv6_unpack(rauhtd_pl, rec_index, &rif, + (char *) &dip); + + if (!mlxsw_sp->router->rifs[rif]) { + dev_err_ratelimited(mlxsw_sp->bus_info->dev, "Incorrect RIF in neighbour entry\n"); + return; + } + + dev = mlxsw_sp->router->rifs[rif]->dev; + n = neigh_lookup(&nd_tbl, &dip, dev); + if (!n) { + netdev_err(dev, "Failed to find matching neighbour for IP=%pI6c\n", + &dip); + return; + } + + netdev_dbg(dev, "Updating neighbour with IP=%pI6c\n", &dip); + neigh_event_send(n, NULL); + neigh_release(n); +} +#else +static void mlxsw_sp_router_neigh_ent_ipv6_process(struct mlxsw_sp *mlxsw_sp, + char *rauhtd_pl, + int rec_index) +{ +} +#endif + static void mlxsw_sp_router_neigh_rec_ipv4_process(struct mlxsw_sp *mlxsw_sp, char *rauhtd_pl, int rec_index) @@ -988,6 +1065,15 @@ static void mlxsw_sp_router_neigh_rec_ipv4_process(struct mlxsw_sp *mlxsw_sp, } +static void mlxsw_sp_router_neigh_rec_ipv6_process(struct mlxsw_sp *mlxsw_sp, + char *rauhtd_pl, + int rec_index) +{ + /* One record contains one entry. */ + mlxsw_sp_router_neigh_ent_ipv6_process(mlxsw_sp, rauhtd_pl, + rec_index); +} + static void mlxsw_sp_router_neigh_rec_process(struct mlxsw_sp *mlxsw_sp, char *rauhtd_pl, int rec_index) { @@ -997,7 +1083,8 @@ static void mlxsw_sp_router_neigh_rec_process(struct mlxsw_sp *mlxsw_sp, rec_index); break; case MLXSW_REG_RAUHTD_TYPE_IPV6: - WARN_ON_ONCE(1); + mlxsw_sp_router_neigh_rec_ipv6_process(mlxsw_sp, rauhtd_pl, + rec_index); break; } } @@ -1022,22 +1109,20 @@ static bool mlxsw_sp_router_rauhtd_is_full(char *rauhtd_pl) return false; } -static int mlxsw_sp_router_neighs_update_rauhtd(struct mlxsw_sp *mlxsw_sp) +static int +__mlxsw_sp_router_neighs_update_rauhtd(struct mlxsw_sp *mlxsw_sp, + char *rauhtd_pl, + enum mlxsw_reg_rauhtd_type type) { - char *rauhtd_pl; - u8 num_rec; - int i, err; - - rauhtd_pl = kmalloc(MLXSW_REG_RAUHTD_LEN, GFP_KERNEL); - if (!rauhtd_pl) - return -ENOMEM; + int i, num_rec; + int err; /* Make sure the neighbour's netdev isn't removed in the * process. */ rtnl_lock(); do { - mlxsw_reg_rauhtd_pack(rauhtd_pl, MLXSW_REG_RAUHTD_TYPE_IPV4); + mlxsw_reg_rauhtd_pack(rauhtd_pl, type); err = mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(rauhtd), rauhtd_pl); if (err) { @@ -1051,6 +1136,27 @@ static int mlxsw_sp_router_neighs_update_rauhtd(struct mlxsw_sp *mlxsw_sp) } while (mlxsw_sp_router_rauhtd_is_full(rauhtd_pl)); rtnl_unlock(); + return err; +} + +static int mlxsw_sp_router_neighs_update_rauhtd(struct mlxsw_sp *mlxsw_sp) +{ + enum mlxsw_reg_rauhtd_type type; + char *rauhtd_pl; + int err; + + rauhtd_pl = kmalloc(MLXSW_REG_RAUHTD_LEN, GFP_KERNEL); + if (!rauhtd_pl) + return -ENOMEM; + + type = MLXSW_REG_RAUHTD_TYPE_IPV4; + err = __mlxsw_sp_router_neighs_update_rauhtd(mlxsw_sp, rauhtd_pl, type); + if (err) + goto out; + + type = MLXSW_REG_RAUHTD_TYPE_IPV6; + err = __mlxsw_sp_router_neighs_update_rauhtd(mlxsw_sp, rauhtd_pl, type); +out: kfree(rauhtd_pl); return err; } @@ -1147,6 +1253,32 @@ mlxsw_sp_router_neigh_entry_op4(struct mlxsw_sp *mlxsw_sp, } static void +mlxsw_sp_router_neigh_entry_op6(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_neigh_entry *neigh_entry, + enum mlxsw_reg_rauht_op op) +{ + struct neighbour *n = neigh_entry->key.n; + char rauht_pl[MLXSW_REG_RAUHT_LEN]; + const char *dip = n->primary_key; + + mlxsw_reg_rauht_pack6(rauht_pl, op, neigh_entry->rif, neigh_entry->ha, + dip); + mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(rauht), rauht_pl); +} + +static bool mlxsw_sp_neigh_ipv6_ignore(struct neighbour *n) +{ + /* Packets with a link-local destination address are trapped + * after LPM lookup and never reach the neighbour table, so + * there is no need to program such neighbours to the device. + */ + if (ipv6_addr_type((struct in6_addr *) &n->primary_key) & + IPV6_ADDR_LINKLOCAL) + return true; + return false; +} + +static void mlxsw_sp_neigh_entry_update(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_neigh_entry *neigh_entry, bool adding) @@ -1154,11 +1286,17 @@ mlxsw_sp_neigh_entry_update(struct mlxsw_sp *mlxsw_sp, if (!adding && !neigh_entry->connected) return; neigh_entry->connected = adding; - if (neigh_entry->key.n->tbl == &arp_tbl) + if (neigh_entry->key.n->tbl->family == AF_INET) { mlxsw_sp_router_neigh_entry_op4(mlxsw_sp, neigh_entry, mlxsw_sp_rauht_op(adding)); - else + } else if (neigh_entry->key.n->tbl->family == AF_INET6) { + if (mlxsw_sp_neigh_ipv6_ignore(neigh_entry->key.n)) + return; + mlxsw_sp_router_neigh_entry_op6(mlxsw_sp, neigh_entry, + mlxsw_sp_rauht_op(adding)); + } else { WARN_ON_ONCE(1); + } } struct mlxsw_sp_neigh_event_work { @@ -1227,7 +1365,8 @@ int mlxsw_sp_router_netevent_event(struct notifier_block *unused, p = ptr; /* We don't care about changes in the default table. */ - if (!p->dev || p->tbl != &arp_tbl) + if (!p->dev || (p->tbl->family != AF_INET && + p->tbl->family != AF_INET6)) return NOTIFY_DONE; /* We are in atomic context and can't take RTNL mutex, @@ -1246,7 +1385,7 @@ int mlxsw_sp_router_netevent_event(struct notifier_block *unused, case NETEVENT_NEIGH_UPDATE: n = ptr; - if (n->tbl != &arp_tbl) + if (n->tbl->family != AF_INET && n->tbl->family != AF_INET6) return NOTIFY_DONE; mlxsw_sp_port = mlxsw_sp_port_lower_dev_hold(n->dev); @@ -1307,25 +1446,16 @@ static void mlxsw_sp_neigh_fini(struct mlxsw_sp *mlxsw_sp) rhashtable_destroy(&mlxsw_sp->router->neigh_ht); } -static int mlxsw_sp_neigh_rif_flush(struct mlxsw_sp *mlxsw_sp, - const struct mlxsw_sp_rif *rif) -{ - char rauht_pl[MLXSW_REG_RAUHT_LEN]; - - mlxsw_reg_rauht_pack(rauht_pl, MLXSW_REG_RAUHT_OP_WRITE_DELETE_ALL, - rif->rif_index, rif->addr); - return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(rauht), rauht_pl); -} - static void mlxsw_sp_neigh_rif_gone_sync(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_rif *rif) { struct mlxsw_sp_neigh_entry *neigh_entry, *tmp; - mlxsw_sp_neigh_rif_flush(mlxsw_sp, rif); list_for_each_entry_safe(neigh_entry, tmp, &rif->neigh_list, - rif_list_node) + rif_list_node) { + mlxsw_sp_neigh_entry_update(mlxsw_sp, neigh_entry, false); mlxsw_sp_neigh_entry_destroy(mlxsw_sp, neigh_entry); + } } struct mlxsw_sp_nexthop_key { @@ -1340,6 +1470,7 @@ struct mlxsw_sp_nexthop { */ struct rhash_head ht_node; struct mlxsw_sp_nexthop_key key; + unsigned char gw_addr[sizeof(struct in6_addr)]; struct mlxsw_sp_rif *rif; u8 should_offload:1, /* set indicates this neigh is connected and * should be put to KVD linear area of this group. @@ -1360,6 +1491,7 @@ struct mlxsw_sp_nexthop_group_key { struct mlxsw_sp_nexthop_group { struct rhash_head ht_node; struct list_head fib_list; /* list of fib entries that use this group */ + struct neigh_table *neigh_tbl; struct mlxsw_sp_nexthop_group_key key; u8 adj_index_valid:1, gateway:1; /* routes using the group use a gateway */ @@ -1535,6 +1667,24 @@ mlxsw_sp_nexthop_fib_entries_update(struct mlxsw_sp *mlxsw_sp, } static void +mlxsw_sp_fib_entry_offload_refresh(struct mlxsw_sp_fib_entry *fib_entry, + enum mlxsw_reg_ralue_op op, int err); + +static void +mlxsw_sp_nexthop_fib_entries_refresh(struct mlxsw_sp_nexthop_group *nh_grp) +{ + enum mlxsw_reg_ralue_op op = MLXSW_REG_RALUE_OP_WRITE_WRITE; + struct mlxsw_sp_fib_entry *fib_entry; + + list_for_each_entry(fib_entry, &nh_grp->fib_list, nexthop_group_node) { + if (!mlxsw_sp_fib_node_entry_is_first(fib_entry->fib_node, + fib_entry)) + continue; + mlxsw_sp_fib_entry_offload_refresh(fib_entry, op, 0); + } +} + +static void mlxsw_sp_nexthop_group_refresh(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_nexthop_group *nh_grp) { @@ -1556,7 +1706,7 @@ mlxsw_sp_nexthop_group_refresh(struct mlxsw_sp *mlxsw_sp, for (i = 0; i < nh_grp->count; i++) { nh = &nh_grp->nexthops[i]; - if (nh->should_offload ^ nh->offloaded) { + if (nh->should_offload != nh->offloaded) { offload_change = true; if (nh->should_offload) nh->update = 1; @@ -1621,6 +1771,10 @@ mlxsw_sp_nexthop_group_refresh(struct mlxsw_sp *mlxsw_sp, dev_warn(mlxsw_sp->bus_info->dev, "Failed to mass-update adjacency index for nexthop group.\n"); goto set_trap; } + + /* Offload state within the group changed, so update the flags. */ + mlxsw_sp_nexthop_fib_entries_refresh(nh_grp); + return; set_trap: @@ -1640,9 +1794,9 @@ set_trap: static void __mlxsw_sp_nexthop_neigh_update(struct mlxsw_sp_nexthop *nh, bool removing) { - if (!removing && !nh->should_offload) + if (!removing) nh->should_offload = 1; - else if (removing && nh->offloaded) + else if (nh->offloaded) nh->should_offload = 0; nh->update = 1; } @@ -1684,7 +1838,6 @@ static int mlxsw_sp_nexthop_neigh_init(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_nexthop *nh) { struct mlxsw_sp_neigh_entry *neigh_entry; - struct fib_nh *fib_nh = nh->key.fib_nh; struct neighbour *n; u8 nud_state, dead; int err; @@ -1693,13 +1846,14 @@ static int mlxsw_sp_nexthop_neigh_init(struct mlxsw_sp *mlxsw_sp, return 0; /* Take a reference of neigh here ensuring that neigh would - * not be detructed before the nexthop entry is finished. + * not be destructed before the nexthop entry is finished. * The reference is taken either in neigh_lookup() or * in neigh_create() in case n is not found. */ - n = neigh_lookup(&arp_tbl, &fib_nh->nh_gw, fib_nh->nh_dev); + n = neigh_lookup(nh->nh_grp->neigh_tbl, &nh->gw_addr, nh->rif->dev); if (!n) { - n = neigh_create(&arp_tbl, &fib_nh->nh_gw, fib_nh->nh_dev); + n = neigh_create(nh->nh_grp->neigh_tbl, &nh->gw_addr, + nh->rif->dev); if (IS_ERR(n)) return PTR_ERR(n); neigh_event_send(n, NULL); @@ -1761,10 +1915,10 @@ static void mlxsw_sp_nexthop_neigh_fini(struct mlxsw_sp *mlxsw_sp, neigh_release(n); } -static int mlxsw_sp_nexthop_init(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_nexthop_group *nh_grp, - struct mlxsw_sp_nexthop *nh, - struct fib_nh *fib_nh) +static int mlxsw_sp_nexthop4_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group *nh_grp, + struct mlxsw_sp_nexthop *nh, + struct fib_nh *fib_nh) { struct net_device *dev = fib_nh->nh_dev; struct in_device *in_dev; @@ -1773,6 +1927,7 @@ static int mlxsw_sp_nexthop_init(struct mlxsw_sp *mlxsw_sp, nh->nh_grp = nh_grp; nh->key.fib_nh = fib_nh; + memcpy(&nh->gw_addr, &fib_nh->nh_gw, sizeof(fib_nh->nh_gw)); err = mlxsw_sp_nexthop_insert(mlxsw_sp, nh); if (err) return err; @@ -1802,16 +1957,16 @@ err_nexthop_neigh_init: return err; } -static void mlxsw_sp_nexthop_fini(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_nexthop *nh) +static void mlxsw_sp_nexthop4_fini(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop *nh) { mlxsw_sp_nexthop_neigh_fini(mlxsw_sp, nh); mlxsw_sp_nexthop_rif_fini(nh); mlxsw_sp_nexthop_remove(mlxsw_sp, nh); } -static void mlxsw_sp_nexthop_event(struct mlxsw_sp *mlxsw_sp, - unsigned long event, struct fib_nh *fib_nh) +static void mlxsw_sp_nexthop4_event(struct mlxsw_sp *mlxsw_sp, + unsigned long event, struct fib_nh *fib_nh) { struct mlxsw_sp_nexthop_key key; struct mlxsw_sp_nexthop *nh; @@ -1856,7 +2011,7 @@ static void mlxsw_sp_nexthop_rif_gone_sync(struct mlxsw_sp *mlxsw_sp, } static struct mlxsw_sp_nexthop_group * -mlxsw_sp_nexthop_group_create(struct mlxsw_sp *mlxsw_sp, struct fib_info *fi) +mlxsw_sp_nexthop4_group_create(struct mlxsw_sp *mlxsw_sp, struct fib_info *fi) { struct mlxsw_sp_nexthop_group *nh_grp; struct mlxsw_sp_nexthop *nh; @@ -1871,6 +2026,8 @@ mlxsw_sp_nexthop_group_create(struct mlxsw_sp *mlxsw_sp, struct fib_info *fi) if (!nh_grp) return ERR_PTR(-ENOMEM); INIT_LIST_HEAD(&nh_grp->fib_list); + nh_grp->neigh_tbl = &arp_tbl; + nh_grp->gateway = fi->fib_nh->nh_scope == RT_SCOPE_LINK; nh_grp->count = fi->fib_nhs; nh_grp->key.fi = fi; @@ -1878,9 +2035,9 @@ mlxsw_sp_nexthop_group_create(struct mlxsw_sp *mlxsw_sp, struct fib_info *fi) for (i = 0; i < nh_grp->count; i++) { nh = &nh_grp->nexthops[i]; fib_nh = &fi->fib_nh[i]; - err = mlxsw_sp_nexthop_init(mlxsw_sp, nh_grp, nh, fib_nh); + err = mlxsw_sp_nexthop4_init(mlxsw_sp, nh_grp, nh, fib_nh); if (err) - goto err_nexthop_init; + goto err_nexthop4_init; } err = mlxsw_sp_nexthop_group_insert(mlxsw_sp, nh_grp); if (err) @@ -1889,10 +2046,10 @@ mlxsw_sp_nexthop_group_create(struct mlxsw_sp *mlxsw_sp, struct fib_info *fi) return nh_grp; err_nexthop_group_insert: -err_nexthop_init: +err_nexthop4_init: for (i--; i >= 0; i--) { nh = &nh_grp->nexthops[i]; - mlxsw_sp_nexthop_fini(mlxsw_sp, nh); + mlxsw_sp_nexthop4_fini(mlxsw_sp, nh); } fib_info_put(nh_grp->key.fi); kfree(nh_grp); @@ -1900,8 +2057,8 @@ err_nexthop_init: } static void -mlxsw_sp_nexthop_group_destroy(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_nexthop_group *nh_grp) +mlxsw_sp_nexthop4_group_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group *nh_grp) { struct mlxsw_sp_nexthop *nh; int i; @@ -1909,7 +2066,7 @@ mlxsw_sp_nexthop_group_destroy(struct mlxsw_sp *mlxsw_sp, mlxsw_sp_nexthop_group_remove(mlxsw_sp, nh_grp); for (i = 0; i < nh_grp->count; i++) { nh = &nh_grp->nexthops[i]; - mlxsw_sp_nexthop_fini(mlxsw_sp, nh); + mlxsw_sp_nexthop4_fini(mlxsw_sp, nh); } mlxsw_sp_nexthop_group_refresh(mlxsw_sp, nh_grp); WARN_ON_ONCE(nh_grp->adj_index_valid); @@ -1917,9 +2074,9 @@ mlxsw_sp_nexthop_group_destroy(struct mlxsw_sp *mlxsw_sp, kfree(nh_grp); } -static int mlxsw_sp_nexthop_group_get(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_fib_entry *fib_entry, - struct fib_info *fi) +static int mlxsw_sp_nexthop4_group_get(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry, + struct fib_info *fi) { struct mlxsw_sp_nexthop_group_key key; struct mlxsw_sp_nexthop_group *nh_grp; @@ -1927,7 +2084,7 @@ static int mlxsw_sp_nexthop_group_get(struct mlxsw_sp *mlxsw_sp, key.fi = fi; nh_grp = mlxsw_sp_nexthop_group_lookup(mlxsw_sp, key); if (!nh_grp) { - nh_grp = mlxsw_sp_nexthop_group_create(mlxsw_sp, fi); + nh_grp = mlxsw_sp_nexthop4_group_create(mlxsw_sp, fi); if (IS_ERR(nh_grp)) return PTR_ERR(nh_grp); } @@ -1936,15 +2093,25 @@ static int mlxsw_sp_nexthop_group_get(struct mlxsw_sp *mlxsw_sp, return 0; } -static void mlxsw_sp_nexthop_group_put(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_fib_entry *fib_entry) +static void mlxsw_sp_nexthop4_group_put(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) { struct mlxsw_sp_nexthop_group *nh_grp = fib_entry->nh_group; list_del(&fib_entry->nexthop_group_node); if (!list_empty(&nh_grp->fib_list)) return; - mlxsw_sp_nexthop_group_destroy(mlxsw_sp, nh_grp); + mlxsw_sp_nexthop4_group_destroy(mlxsw_sp, nh_grp); +} + +static bool +mlxsw_sp_fib4_entry_should_offload(const struct mlxsw_sp_fib_entry *fib_entry) +{ + struct mlxsw_sp_fib4_entry *fib4_entry; + + fib4_entry = container_of(fib_entry, struct mlxsw_sp_fib4_entry, + common); + return !fib4_entry->tos; } static bool @@ -1952,8 +2119,14 @@ mlxsw_sp_fib_entry_should_offload(const struct mlxsw_sp_fib_entry *fib_entry) { struct mlxsw_sp_nexthop_group *nh_group = fib_entry->nh_group; - if (fib_entry->params.tos) - return false; + switch (fib_entry->fib_node->fib->proto) { + case MLXSW_SP_L3_PROTO_IPV4: + if (!mlxsw_sp_fib4_entry_should_offload(fib_entry)) + return false; + break; + case MLXSW_SP_L3_PROTO_IPV6: + break; + } switch (fib_entry->type) { case MLXSW_SP_FIB_ENTRY_TYPE_REMOTE: @@ -1965,16 +2138,111 @@ mlxsw_sp_fib_entry_should_offload(const struct mlxsw_sp_fib_entry *fib_entry) } } -static void mlxsw_sp_fib_entry_offload_set(struct mlxsw_sp_fib_entry *fib_entry) +static struct mlxsw_sp_nexthop * +mlxsw_sp_rt6_nexthop(struct mlxsw_sp_nexthop_group *nh_grp, + const struct mlxsw_sp_rt6 *mlxsw_sp_rt6) +{ + int i; + + for (i = 0; i < nh_grp->count; i++) { + struct mlxsw_sp_nexthop *nh = &nh_grp->nexthops[i]; + struct rt6_info *rt = mlxsw_sp_rt6->rt; + + if (nh->rif && nh->rif->dev == rt->dst.dev && + ipv6_addr_equal((const struct in6_addr *) &nh->gw_addr, + &rt->rt6i_gateway)) + return nh; + continue; + } + + return NULL; +} + +static void +mlxsw_sp_fib4_entry_offload_set(struct mlxsw_sp_fib_entry *fib_entry) +{ + struct mlxsw_sp_nexthop_group *nh_grp = fib_entry->nh_group; + int i; + + if (fib_entry->type == MLXSW_SP_FIB_ENTRY_TYPE_LOCAL) { + nh_grp->nexthops->key.fib_nh->nh_flags |= RTNH_F_OFFLOAD; + return; + } + + for (i = 0; i < nh_grp->count; i++) { + struct mlxsw_sp_nexthop *nh = &nh_grp->nexthops[i]; + + if (nh->offloaded) + nh->key.fib_nh->nh_flags |= RTNH_F_OFFLOAD; + else + nh->key.fib_nh->nh_flags &= ~RTNH_F_OFFLOAD; + } +} + +static void +mlxsw_sp_fib4_entry_offload_unset(struct mlxsw_sp_fib_entry *fib_entry) +{ + struct mlxsw_sp_nexthop_group *nh_grp = fib_entry->nh_group; + int i; + + for (i = 0; i < nh_grp->count; i++) { + struct mlxsw_sp_nexthop *nh = &nh_grp->nexthops[i]; + + nh->key.fib_nh->nh_flags &= ~RTNH_F_OFFLOAD; + } +} + +static void +mlxsw_sp_fib6_entry_offload_set(struct mlxsw_sp_fib_entry *fib_entry) { - fib_entry->offloaded = true; + struct mlxsw_sp_fib6_entry *fib6_entry; + struct mlxsw_sp_rt6 *mlxsw_sp_rt6; + + fib6_entry = container_of(fib_entry, struct mlxsw_sp_fib6_entry, + common); + + if (fib_entry->type == MLXSW_SP_FIB_ENTRY_TYPE_LOCAL) { + list_first_entry(&fib6_entry->rt6_list, struct mlxsw_sp_rt6, + list)->rt->rt6i_flags |= RTF_OFFLOAD; + return; + } + + list_for_each_entry(mlxsw_sp_rt6, &fib6_entry->rt6_list, list) { + struct mlxsw_sp_nexthop_group *nh_grp = fib_entry->nh_group; + struct mlxsw_sp_nexthop *nh; + + nh = mlxsw_sp_rt6_nexthop(nh_grp, mlxsw_sp_rt6); + if (nh && nh->offloaded) + mlxsw_sp_rt6->rt->rt6i_flags |= RTF_OFFLOAD; + else + mlxsw_sp_rt6->rt->rt6i_flags &= ~RTF_OFFLOAD; + } +} +static void +mlxsw_sp_fib6_entry_offload_unset(struct mlxsw_sp_fib_entry *fib_entry) +{ + struct mlxsw_sp_fib6_entry *fib6_entry; + struct mlxsw_sp_rt6 *mlxsw_sp_rt6; + + fib6_entry = container_of(fib_entry, struct mlxsw_sp_fib6_entry, + common); + list_for_each_entry(mlxsw_sp_rt6, &fib6_entry->rt6_list, list) { + struct rt6_info *rt = mlxsw_sp_rt6->rt; + + rt->rt6i_flags &= ~RTF_OFFLOAD; + } +} + +static void mlxsw_sp_fib_entry_offload_set(struct mlxsw_sp_fib_entry *fib_entry) +{ switch (fib_entry->fib_node->fib->proto) { case MLXSW_SP_L3_PROTO_IPV4: - fib_info_offload_inc(fib_entry->nh_group->key.fi); + mlxsw_sp_fib4_entry_offload_set(fib_entry); break; case MLXSW_SP_L3_PROTO_IPV6: - WARN_ON_ONCE(1); + mlxsw_sp_fib6_entry_offload_set(fib_entry); + break; } } @@ -1983,13 +2251,12 @@ mlxsw_sp_fib_entry_offload_unset(struct mlxsw_sp_fib_entry *fib_entry) { switch (fib_entry->fib_node->fib->proto) { case MLXSW_SP_L3_PROTO_IPV4: - fib_info_offload_dec(fib_entry->nh_group->key.fi); + mlxsw_sp_fib4_entry_offload_unset(fib_entry); break; case MLXSW_SP_L3_PROTO_IPV6: - WARN_ON_ONCE(1); + mlxsw_sp_fib6_entry_offload_unset(fib_entry); + break; } - - fib_entry->offloaded = false; } static void @@ -1998,17 +2265,13 @@ mlxsw_sp_fib_entry_offload_refresh(struct mlxsw_sp_fib_entry *fib_entry, { switch (op) { case MLXSW_REG_RALUE_OP_WRITE_DELETE: - if (!fib_entry->offloaded) - return; return mlxsw_sp_fib_entry_offload_unset(fib_entry); case MLXSW_REG_RALUE_OP_WRITE_WRITE: if (err) return; - if (mlxsw_sp_fib_entry_should_offload(fib_entry) && - !fib_entry->offloaded) + if (mlxsw_sp_fib_entry_should_offload(fib_entry)) mlxsw_sp_fib_entry_offload_set(fib_entry); - else if (!mlxsw_sp_fib_entry_should_offload(fib_entry) && - fib_entry->offloaded) + else if (!mlxsw_sp_fib_entry_should_offload(fib_entry)) mlxsw_sp_fib_entry_offload_unset(fib_entry); return; default: @@ -2016,13 +2279,37 @@ mlxsw_sp_fib_entry_offload_refresh(struct mlxsw_sp_fib_entry *fib_entry, } } -static int mlxsw_sp_fib_entry_op4_remote(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_fib_entry *fib_entry, - enum mlxsw_reg_ralue_op op) +static void +mlxsw_sp_fib_entry_ralue_pack(char *ralue_pl, + const struct mlxsw_sp_fib_entry *fib_entry, + enum mlxsw_reg_ralue_op op) { - char ralue_pl[MLXSW_REG_RALUE_LEN]; struct mlxsw_sp_fib *fib = fib_entry->fib_node->fib; - u32 *p_dip = (u32 *) fib_entry->fib_node->key.addr; + enum mlxsw_reg_ralxx_protocol proto; + u32 *p_dip; + + proto = (enum mlxsw_reg_ralxx_protocol) fib->proto; + + switch (fib->proto) { + case MLXSW_SP_L3_PROTO_IPV4: + p_dip = (u32 *) fib_entry->fib_node->key.addr; + mlxsw_reg_ralue_pack4(ralue_pl, proto, op, fib->vr->id, + fib_entry->fib_node->key.prefix_len, + *p_dip); + break; + case MLXSW_SP_L3_PROTO_IPV6: + mlxsw_reg_ralue_pack6(ralue_pl, proto, op, fib->vr->id, + fib_entry->fib_node->key.prefix_len, + fib_entry->fib_node->key.addr); + break; + } +} + +static int mlxsw_sp_fib_entry_op_remote(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry, + enum mlxsw_reg_ralue_op op) +{ + char ralue_pl[MLXSW_REG_RALUE_LEN]; enum mlxsw_reg_ralue_trap_action trap_action; u16 trap_id = 0; u32 adjacency_index = 0; @@ -2041,24 +2328,19 @@ static int mlxsw_sp_fib_entry_op4_remote(struct mlxsw_sp *mlxsw_sp, trap_id = MLXSW_TRAP_ID_RTR_INGRESS0; } - mlxsw_reg_ralue_pack4(ralue_pl, - (enum mlxsw_reg_ralxx_protocol) fib->proto, op, - fib->vr->id, fib_entry->fib_node->key.prefix_len, - *p_dip); + mlxsw_sp_fib_entry_ralue_pack(ralue_pl, fib_entry, op); mlxsw_reg_ralue_act_remote_pack(ralue_pl, trap_action, trap_id, adjacency_index, ecmp_size); return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ralue), ralue_pl); } -static int mlxsw_sp_fib_entry_op4_local(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_fib_entry *fib_entry, - enum mlxsw_reg_ralue_op op) +static int mlxsw_sp_fib_entry_op_local(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry, + enum mlxsw_reg_ralue_op op) { struct mlxsw_sp_rif *rif = fib_entry->nh_group->nh_rif; - struct mlxsw_sp_fib *fib = fib_entry->fib_node->fib; enum mlxsw_reg_ralue_trap_action trap_action; char ralue_pl[MLXSW_REG_RALUE_LEN]; - u32 *p_dip = (u32 *) fib_entry->fib_node->key.addr; u16 trap_id = 0; u16 rif_index = 0; @@ -2070,42 +2352,34 @@ static int mlxsw_sp_fib_entry_op4_local(struct mlxsw_sp *mlxsw_sp, trap_id = MLXSW_TRAP_ID_RTR_INGRESS0; } - mlxsw_reg_ralue_pack4(ralue_pl, - (enum mlxsw_reg_ralxx_protocol) fib->proto, op, - fib->vr->id, fib_entry->fib_node->key.prefix_len, - *p_dip); + mlxsw_sp_fib_entry_ralue_pack(ralue_pl, fib_entry, op); mlxsw_reg_ralue_act_local_pack(ralue_pl, trap_action, trap_id, rif_index); return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ralue), ralue_pl); } -static int mlxsw_sp_fib_entry_op4_trap(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_fib_entry *fib_entry, - enum mlxsw_reg_ralue_op op) +static int mlxsw_sp_fib_entry_op_trap(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry, + enum mlxsw_reg_ralue_op op) { - struct mlxsw_sp_fib *fib = fib_entry->fib_node->fib; char ralue_pl[MLXSW_REG_RALUE_LEN]; - u32 *p_dip = (u32 *) fib_entry->fib_node->key.addr; - mlxsw_reg_ralue_pack4(ralue_pl, - (enum mlxsw_reg_ralxx_protocol) fib->proto, op, - fib->vr->id, fib_entry->fib_node->key.prefix_len, - *p_dip); + mlxsw_sp_fib_entry_ralue_pack(ralue_pl, fib_entry, op); mlxsw_reg_ralue_act_ip2me_pack(ralue_pl); return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ralue), ralue_pl); } -static int mlxsw_sp_fib_entry_op4(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_fib_entry *fib_entry, - enum mlxsw_reg_ralue_op op) +static int __mlxsw_sp_fib_entry_op(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry, + enum mlxsw_reg_ralue_op op) { switch (fib_entry->type) { case MLXSW_SP_FIB_ENTRY_TYPE_REMOTE: - return mlxsw_sp_fib_entry_op4_remote(mlxsw_sp, fib_entry, op); + return mlxsw_sp_fib_entry_op_remote(mlxsw_sp, fib_entry, op); case MLXSW_SP_FIB_ENTRY_TYPE_LOCAL: - return mlxsw_sp_fib_entry_op4_local(mlxsw_sp, fib_entry, op); + return mlxsw_sp_fib_entry_op_local(mlxsw_sp, fib_entry, op); case MLXSW_SP_FIB_ENTRY_TYPE_TRAP: - return mlxsw_sp_fib_entry_op4_trap(mlxsw_sp, fib_entry, op); + return mlxsw_sp_fib_entry_op_trap(mlxsw_sp, fib_entry, op); } return -EINVAL; } @@ -2114,16 +2388,10 @@ static int mlxsw_sp_fib_entry_op(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_fib_entry *fib_entry, enum mlxsw_reg_ralue_op op) { - int err = -EINVAL; + int err = __mlxsw_sp_fib_entry_op(mlxsw_sp, fib_entry, op); - switch (fib_entry->fib_node->fib->proto) { - case MLXSW_SP_L3_PROTO_IPV4: - err = mlxsw_sp_fib_entry_op4(mlxsw_sp, fib_entry, op); - break; - case MLXSW_SP_L3_PROTO_IPV6: - return err; - } mlxsw_sp_fib_entry_offload_refresh(fib_entry, op, err); + return err; } @@ -2173,72 +2441,80 @@ mlxsw_sp_fib4_entry_type_set(struct mlxsw_sp *mlxsw_sp, } } -static struct mlxsw_sp_fib_entry * +static struct mlxsw_sp_fib4_entry * mlxsw_sp_fib4_entry_create(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_fib_node *fib_node, const struct fib_entry_notifier_info *fen_info) { + struct mlxsw_sp_fib4_entry *fib4_entry; struct mlxsw_sp_fib_entry *fib_entry; int err; - fib_entry = kzalloc(sizeof(*fib_entry), GFP_KERNEL); - if (!fib_entry) { - err = -ENOMEM; - goto err_fib_entry_alloc; - } + fib4_entry = kzalloc(sizeof(*fib4_entry), GFP_KERNEL); + if (!fib4_entry) + return ERR_PTR(-ENOMEM); + fib_entry = &fib4_entry->common; err = mlxsw_sp_fib4_entry_type_set(mlxsw_sp, fen_info, fib_entry); if (err) goto err_fib4_entry_type_set; - err = mlxsw_sp_nexthop_group_get(mlxsw_sp, fib_entry, fen_info->fi); + err = mlxsw_sp_nexthop4_group_get(mlxsw_sp, fib_entry, fen_info->fi); if (err) - goto err_nexthop_group_get; + goto err_nexthop4_group_get; - fib_entry->params.prio = fen_info->fi->fib_priority; - fib_entry->params.tb_id = fen_info->tb_id; - fib_entry->params.type = fen_info->type; - fib_entry->params.tos = fen_info->tos; + fib4_entry->prio = fen_info->fi->fib_priority; + fib4_entry->tb_id = fen_info->tb_id; + fib4_entry->type = fen_info->type; + fib4_entry->tos = fen_info->tos; fib_entry->fib_node = fib_node; - return fib_entry; + return fib4_entry; -err_nexthop_group_get: +err_nexthop4_group_get: err_fib4_entry_type_set: - kfree(fib_entry); -err_fib_entry_alloc: + kfree(fib4_entry); return ERR_PTR(err); } static void mlxsw_sp_fib4_entry_destroy(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_fib_entry *fib_entry) + struct mlxsw_sp_fib4_entry *fib4_entry) { - mlxsw_sp_nexthop_group_put(mlxsw_sp, fib_entry); - kfree(fib_entry); + mlxsw_sp_nexthop4_group_put(mlxsw_sp, &fib4_entry->common); + kfree(fib4_entry); } static struct mlxsw_sp_fib_node * -mlxsw_sp_fib4_node_get(struct mlxsw_sp *mlxsw_sp, - const struct fib_entry_notifier_info *fen_info); +mlxsw_sp_fib_node_lookup(struct mlxsw_sp_fib *fib, const void *addr, + size_t addr_len, unsigned char prefix_len); -static struct mlxsw_sp_fib_entry * +static struct mlxsw_sp_fib4_entry * mlxsw_sp_fib4_entry_lookup(struct mlxsw_sp *mlxsw_sp, const struct fib_entry_notifier_info *fen_info) { - struct mlxsw_sp_fib_entry *fib_entry; + struct mlxsw_sp_fib4_entry *fib4_entry; struct mlxsw_sp_fib_node *fib_node; + struct mlxsw_sp_fib *fib; + struct mlxsw_sp_vr *vr; - fib_node = mlxsw_sp_fib4_node_get(mlxsw_sp, fen_info); - if (IS_ERR(fib_node)) + vr = mlxsw_sp_vr_find(mlxsw_sp, fen_info->tb_id); + if (!vr) + return NULL; + fib = mlxsw_sp_vr_fib(vr, MLXSW_SP_L3_PROTO_IPV4); + + fib_node = mlxsw_sp_fib_node_lookup(fib, &fen_info->dst, + sizeof(fen_info->dst), + fen_info->dst_len); + if (!fib_node) return NULL; - list_for_each_entry(fib_entry, &fib_node->entry_list, list) { - if (fib_entry->params.tb_id == fen_info->tb_id && - fib_entry->params.tos == fen_info->tos && - fib_entry->params.type == fen_info->type && - fib_entry->nh_group->key.fi == fen_info->fi) { - return fib_entry; + list_for_each_entry(fib4_entry, &fib_node->entry_list, common.list) { + if (fib4_entry->tb_id == fen_info->tb_id && + fib4_entry->tos == fen_info->tos && + fib4_entry->type == fen_info->type && + fib4_entry->common.nh_group->key.fi == fen_info->fi) { + return fib4_entry; } } @@ -2395,28 +2671,25 @@ static void mlxsw_sp_fib_node_fini(struct mlxsw_sp *mlxsw_sp, } static struct mlxsw_sp_fib_node * -mlxsw_sp_fib4_node_get(struct mlxsw_sp *mlxsw_sp, - const struct fib_entry_notifier_info *fen_info) +mlxsw_sp_fib_node_get(struct mlxsw_sp *mlxsw_sp, u32 tb_id, const void *addr, + size_t addr_len, unsigned char prefix_len, + enum mlxsw_sp_l3proto proto) { struct mlxsw_sp_fib_node *fib_node; struct mlxsw_sp_fib *fib; struct mlxsw_sp_vr *vr; int err; - vr = mlxsw_sp_vr_get(mlxsw_sp, fen_info->tb_id); + vr = mlxsw_sp_vr_get(mlxsw_sp, tb_id); if (IS_ERR(vr)) return ERR_CAST(vr); - fib = mlxsw_sp_vr_fib(vr, MLXSW_SP_L3_PROTO_IPV4); + fib = mlxsw_sp_vr_fib(vr, proto); - fib_node = mlxsw_sp_fib_node_lookup(fib, &fen_info->dst, - sizeof(fen_info->dst), - fen_info->dst_len); + fib_node = mlxsw_sp_fib_node_lookup(fib, addr, addr_len, prefix_len); if (fib_node) return fib_node; - fib_node = mlxsw_sp_fib_node_create(fib, &fen_info->dst, - sizeof(fen_info->dst), - fen_info->dst_len); + fib_node = mlxsw_sp_fib_node_create(fib, addr, addr_len, prefix_len); if (!fib_node) { err = -ENOMEM; goto err_fib_node_create; @@ -2435,8 +2708,8 @@ err_fib_node_create: return ERR_PTR(err); } -static void mlxsw_sp_fib4_node_put(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_fib_node *fib_node) +static void mlxsw_sp_fib_node_put(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_node *fib_node) { struct mlxsw_sp_vr *vr = fib_node->fib->vr; @@ -2447,95 +2720,100 @@ static void mlxsw_sp_fib4_node_put(struct mlxsw_sp *mlxsw_sp, mlxsw_sp_vr_put(vr); } -static struct mlxsw_sp_fib_entry * +static struct mlxsw_sp_fib4_entry * mlxsw_sp_fib4_node_entry_find(const struct mlxsw_sp_fib_node *fib_node, - const struct mlxsw_sp_fib_entry_params *params) + const struct mlxsw_sp_fib4_entry *new4_entry) { - struct mlxsw_sp_fib_entry *fib_entry; + struct mlxsw_sp_fib4_entry *fib4_entry; - list_for_each_entry(fib_entry, &fib_node->entry_list, list) { - if (fib_entry->params.tb_id > params->tb_id) + list_for_each_entry(fib4_entry, &fib_node->entry_list, common.list) { + if (fib4_entry->tb_id > new4_entry->tb_id) continue; - if (fib_entry->params.tb_id != params->tb_id) + if (fib4_entry->tb_id != new4_entry->tb_id) break; - if (fib_entry->params.tos > params->tos) + if (fib4_entry->tos > new4_entry->tos) continue; - if (fib_entry->params.prio >= params->prio || - fib_entry->params.tos < params->tos) - return fib_entry; + if (fib4_entry->prio >= new4_entry->prio || + fib4_entry->tos < new4_entry->tos) + return fib4_entry; } return NULL; } -static int mlxsw_sp_fib4_node_list_append(struct mlxsw_sp_fib_entry *fib_entry, - struct mlxsw_sp_fib_entry *new_entry) +static int +mlxsw_sp_fib4_node_list_append(struct mlxsw_sp_fib4_entry *fib4_entry, + struct mlxsw_sp_fib4_entry *new4_entry) { struct mlxsw_sp_fib_node *fib_node; - if (WARN_ON(!fib_entry)) + if (WARN_ON(!fib4_entry)) return -EINVAL; - fib_node = fib_entry->fib_node; - list_for_each_entry_from(fib_entry, &fib_node->entry_list, list) { - if (fib_entry->params.tb_id != new_entry->params.tb_id || - fib_entry->params.tos != new_entry->params.tos || - fib_entry->params.prio != new_entry->params.prio) + fib_node = fib4_entry->common.fib_node; + list_for_each_entry_from(fib4_entry, &fib_node->entry_list, + common.list) { + if (fib4_entry->tb_id != new4_entry->tb_id || + fib4_entry->tos != new4_entry->tos || + fib4_entry->prio != new4_entry->prio) break; } - list_add_tail(&new_entry->list, &fib_entry->list); + list_add_tail(&new4_entry->common.list, &fib4_entry->common.list); return 0; } static int -mlxsw_sp_fib4_node_list_insert(struct mlxsw_sp_fib_node *fib_node, - struct mlxsw_sp_fib_entry *new_entry, +mlxsw_sp_fib4_node_list_insert(struct mlxsw_sp_fib4_entry *new4_entry, bool replace, bool append) { - struct mlxsw_sp_fib_entry *fib_entry; + struct mlxsw_sp_fib_node *fib_node = new4_entry->common.fib_node; + struct mlxsw_sp_fib4_entry *fib4_entry; - fib_entry = mlxsw_sp_fib4_node_entry_find(fib_node, &new_entry->params); + fib4_entry = mlxsw_sp_fib4_node_entry_find(fib_node, new4_entry); if (append) - return mlxsw_sp_fib4_node_list_append(fib_entry, new_entry); - if (replace && WARN_ON(!fib_entry)) + return mlxsw_sp_fib4_node_list_append(fib4_entry, new4_entry); + if (replace && WARN_ON(!fib4_entry)) return -EINVAL; /* Insert new entry before replaced one, so that we can later * remove the second. */ - if (fib_entry) { - list_add_tail(&new_entry->list, &fib_entry->list); + if (fib4_entry) { + list_add_tail(&new4_entry->common.list, + &fib4_entry->common.list); } else { - struct mlxsw_sp_fib_entry *last; + struct mlxsw_sp_fib4_entry *last; - list_for_each_entry(last, &fib_node->entry_list, list) { - if (new_entry->params.tb_id > last->params.tb_id) + list_for_each_entry(last, &fib_node->entry_list, common.list) { + if (new4_entry->tb_id > last->tb_id) break; - fib_entry = last; + fib4_entry = last; } - if (fib_entry) - list_add(&new_entry->list, &fib_entry->list); + if (fib4_entry) + list_add(&new4_entry->common.list, + &fib4_entry->common.list); else - list_add(&new_entry->list, &fib_node->entry_list); + list_add(&new4_entry->common.list, + &fib_node->entry_list); } return 0; } static void -mlxsw_sp_fib4_node_list_remove(struct mlxsw_sp_fib_entry *fib_entry) +mlxsw_sp_fib4_node_list_remove(struct mlxsw_sp_fib4_entry *fib4_entry) { - list_del(&fib_entry->list); + list_del(&fib4_entry->common.list); } -static int -mlxsw_sp_fib4_node_entry_add(struct mlxsw_sp *mlxsw_sp, - const struct mlxsw_sp_fib_node *fib_node, - struct mlxsw_sp_fib_entry *fib_entry) +static int mlxsw_sp_fib_node_entry_add(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) { + struct mlxsw_sp_fib_node *fib_node = fib_entry->fib_node; + if (!mlxsw_sp_fib_node_entry_is_first(fib_node, fib_entry)) return 0; @@ -2552,11 +2830,11 @@ mlxsw_sp_fib4_node_entry_add(struct mlxsw_sp *mlxsw_sp, return mlxsw_sp_fib_entry_update(mlxsw_sp, fib_entry); } -static void -mlxsw_sp_fib4_node_entry_del(struct mlxsw_sp *mlxsw_sp, - const struct mlxsw_sp_fib_node *fib_node, - struct mlxsw_sp_fib_entry *fib_entry) +static void mlxsw_sp_fib_node_entry_del(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) { + struct mlxsw_sp_fib_node *fib_node = fib_entry->fib_node; + if (!mlxsw_sp_fib_node_entry_is_first(fib_node, fib_entry)) return; @@ -2574,54 +2852,50 @@ mlxsw_sp_fib4_node_entry_del(struct mlxsw_sp *mlxsw_sp, } static int mlxsw_sp_fib4_node_entry_link(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_fib_entry *fib_entry, + struct mlxsw_sp_fib4_entry *fib4_entry, bool replace, bool append) { - struct mlxsw_sp_fib_node *fib_node = fib_entry->fib_node; int err; - err = mlxsw_sp_fib4_node_list_insert(fib_node, fib_entry, replace, - append); + err = mlxsw_sp_fib4_node_list_insert(fib4_entry, replace, append); if (err) return err; - err = mlxsw_sp_fib4_node_entry_add(mlxsw_sp, fib_node, fib_entry); + err = mlxsw_sp_fib_node_entry_add(mlxsw_sp, &fib4_entry->common); if (err) - goto err_fib4_node_entry_add; + goto err_fib_node_entry_add; return 0; -err_fib4_node_entry_add: - mlxsw_sp_fib4_node_list_remove(fib_entry); +err_fib_node_entry_add: + mlxsw_sp_fib4_node_list_remove(fib4_entry); return err; } static void mlxsw_sp_fib4_node_entry_unlink(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_fib_entry *fib_entry) + struct mlxsw_sp_fib4_entry *fib4_entry) { - struct mlxsw_sp_fib_node *fib_node = fib_entry->fib_node; - - mlxsw_sp_fib4_node_entry_del(mlxsw_sp, fib_node, fib_entry); - mlxsw_sp_fib4_node_list_remove(fib_entry); + mlxsw_sp_fib_node_entry_del(mlxsw_sp, &fib4_entry->common); + mlxsw_sp_fib4_node_list_remove(fib4_entry); } static void mlxsw_sp_fib4_entry_replace(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_fib_entry *fib_entry, + struct mlxsw_sp_fib4_entry *fib4_entry, bool replace) { - struct mlxsw_sp_fib_node *fib_node = fib_entry->fib_node; - struct mlxsw_sp_fib_entry *replaced; + struct mlxsw_sp_fib_node *fib_node = fib4_entry->common.fib_node; + struct mlxsw_sp_fib4_entry *replaced; if (!replace) return; /* We inserted the new entry before replaced one */ - replaced = list_next_entry(fib_entry, list); + replaced = list_next_entry(fib4_entry, common.list); mlxsw_sp_fib4_node_entry_unlink(mlxsw_sp, replaced); mlxsw_sp_fib4_entry_destroy(mlxsw_sp, replaced); - mlxsw_sp_fib4_node_put(mlxsw_sp, fib_node); + mlxsw_sp_fib_node_put(mlxsw_sp, fib_node); } static int @@ -2629,76 +2903,723 @@ mlxsw_sp_router_fib4_add(struct mlxsw_sp *mlxsw_sp, const struct fib_entry_notifier_info *fen_info, bool replace, bool append) { - struct mlxsw_sp_fib_entry *fib_entry; + struct mlxsw_sp_fib4_entry *fib4_entry; struct mlxsw_sp_fib_node *fib_node; int err; if (mlxsw_sp->router->aborted) return 0; - fib_node = mlxsw_sp_fib4_node_get(mlxsw_sp, fen_info); + fib_node = mlxsw_sp_fib_node_get(mlxsw_sp, fen_info->tb_id, + &fen_info->dst, sizeof(fen_info->dst), + fen_info->dst_len, + MLXSW_SP_L3_PROTO_IPV4); if (IS_ERR(fib_node)) { dev_warn(mlxsw_sp->bus_info->dev, "Failed to get FIB node\n"); return PTR_ERR(fib_node); } - fib_entry = mlxsw_sp_fib4_entry_create(mlxsw_sp, fib_node, fen_info); - if (IS_ERR(fib_entry)) { + fib4_entry = mlxsw_sp_fib4_entry_create(mlxsw_sp, fib_node, fen_info); + if (IS_ERR(fib4_entry)) { dev_warn(mlxsw_sp->bus_info->dev, "Failed to create FIB entry\n"); - err = PTR_ERR(fib_entry); + err = PTR_ERR(fib4_entry); goto err_fib4_entry_create; } - err = mlxsw_sp_fib4_node_entry_link(mlxsw_sp, fib_entry, replace, + err = mlxsw_sp_fib4_node_entry_link(mlxsw_sp, fib4_entry, replace, append); if (err) { dev_warn(mlxsw_sp->bus_info->dev, "Failed to link FIB entry to node\n"); goto err_fib4_node_entry_link; } - mlxsw_sp_fib4_entry_replace(mlxsw_sp, fib_entry, replace); + mlxsw_sp_fib4_entry_replace(mlxsw_sp, fib4_entry, replace); return 0; err_fib4_node_entry_link: - mlxsw_sp_fib4_entry_destroy(mlxsw_sp, fib_entry); + mlxsw_sp_fib4_entry_destroy(mlxsw_sp, fib4_entry); err_fib4_entry_create: - mlxsw_sp_fib4_node_put(mlxsw_sp, fib_node); + mlxsw_sp_fib_node_put(mlxsw_sp, fib_node); return err; } static void mlxsw_sp_router_fib4_del(struct mlxsw_sp *mlxsw_sp, struct fib_entry_notifier_info *fen_info) { + struct mlxsw_sp_fib4_entry *fib4_entry; + struct mlxsw_sp_fib_node *fib_node; + + if (mlxsw_sp->router->aborted) + return; + + fib4_entry = mlxsw_sp_fib4_entry_lookup(mlxsw_sp, fen_info); + if (WARN_ON(!fib4_entry)) + return; + fib_node = fib4_entry->common.fib_node; + + mlxsw_sp_fib4_node_entry_unlink(mlxsw_sp, fib4_entry); + mlxsw_sp_fib4_entry_destroy(mlxsw_sp, fib4_entry); + mlxsw_sp_fib_node_put(mlxsw_sp, fib_node); +} + +static bool mlxsw_sp_fib6_rt_should_ignore(const struct rt6_info *rt) +{ + /* Packets with link-local destination IP arriving to the router + * are trapped to the CPU, so no need to program specific routes + * for them. + */ + if (ipv6_addr_type(&rt->rt6i_dst.addr) & IPV6_ADDR_LINKLOCAL) + return true; + + /* Multicast routes aren't supported, so ignore them. Neighbour + * Discovery packets are specifically trapped. + */ + if (ipv6_addr_type(&rt->rt6i_dst.addr) & IPV6_ADDR_MULTICAST) + return true; + + /* Cloned routes are irrelevant in the forwarding path. */ + if (rt->rt6i_flags & RTF_CACHE) + return true; + + return false; +} + +static struct mlxsw_sp_rt6 *mlxsw_sp_rt6_create(struct rt6_info *rt) +{ + struct mlxsw_sp_rt6 *mlxsw_sp_rt6; + + mlxsw_sp_rt6 = kzalloc(sizeof(*mlxsw_sp_rt6), GFP_KERNEL); + if (!mlxsw_sp_rt6) + return ERR_PTR(-ENOMEM); + + /* In case of route replace, replaced route is deleted with + * no notification. Take reference to prevent accessing freed + * memory. + */ + mlxsw_sp_rt6->rt = rt; + rt6_hold(rt); + + return mlxsw_sp_rt6; +} + +#if IS_ENABLED(CONFIG_IPV6) +static void mlxsw_sp_rt6_release(struct rt6_info *rt) +{ + rt6_release(rt); +} +#else +static void mlxsw_sp_rt6_release(struct rt6_info *rt) +{ +} +#endif + +static void mlxsw_sp_rt6_destroy(struct mlxsw_sp_rt6 *mlxsw_sp_rt6) +{ + mlxsw_sp_rt6_release(mlxsw_sp_rt6->rt); + kfree(mlxsw_sp_rt6); +} + +static bool mlxsw_sp_fib6_rt_can_mp(const struct rt6_info *rt) +{ + /* RTF_CACHE routes are ignored */ + return (rt->rt6i_flags & (RTF_GATEWAY | RTF_ADDRCONF)) == RTF_GATEWAY; +} + +static struct rt6_info * +mlxsw_sp_fib6_entry_rt(const struct mlxsw_sp_fib6_entry *fib6_entry) +{ + return list_first_entry(&fib6_entry->rt6_list, struct mlxsw_sp_rt6, + list)->rt; +} + +static struct mlxsw_sp_fib6_entry * +mlxsw_sp_fib6_node_mp_entry_find(const struct mlxsw_sp_fib_node *fib_node, + const struct rt6_info *nrt, bool replace) +{ + struct mlxsw_sp_fib6_entry *fib6_entry; + + if (!mlxsw_sp_fib6_rt_can_mp(nrt) || replace) + return NULL; + + list_for_each_entry(fib6_entry, &fib_node->entry_list, common.list) { + struct rt6_info *rt = mlxsw_sp_fib6_entry_rt(fib6_entry); + + /* RT6_TABLE_LOCAL and RT6_TABLE_MAIN share the same + * virtual router. + */ + if (rt->rt6i_table->tb6_id > nrt->rt6i_table->tb6_id) + continue; + if (rt->rt6i_table->tb6_id != nrt->rt6i_table->tb6_id) + break; + if (rt->rt6i_metric < nrt->rt6i_metric) + continue; + if (rt->rt6i_metric == nrt->rt6i_metric && + mlxsw_sp_fib6_rt_can_mp(rt)) + return fib6_entry; + if (rt->rt6i_metric > nrt->rt6i_metric) + break; + } + + return NULL; +} + +static struct mlxsw_sp_rt6 * +mlxsw_sp_fib6_entry_rt_find(const struct mlxsw_sp_fib6_entry *fib6_entry, + const struct rt6_info *rt) +{ + struct mlxsw_sp_rt6 *mlxsw_sp_rt6; + + list_for_each_entry(mlxsw_sp_rt6, &fib6_entry->rt6_list, list) { + if (mlxsw_sp_rt6->rt == rt) + return mlxsw_sp_rt6; + } + + return NULL; +} + +static int mlxsw_sp_nexthop6_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group *nh_grp, + struct mlxsw_sp_nexthop *nh, + const struct rt6_info *rt) +{ + struct net_device *dev = rt->dst.dev; + struct mlxsw_sp_rif *rif; + int err; + + nh->nh_grp = nh_grp; + memcpy(&nh->gw_addr, &rt->rt6i_gateway, sizeof(nh->gw_addr)); + + if (!dev) + return 0; + + rif = mlxsw_sp_rif_find_by_dev(mlxsw_sp, dev); + if (!rif) + return 0; + mlxsw_sp_nexthop_rif_init(nh, rif); + + err = mlxsw_sp_nexthop_neigh_init(mlxsw_sp, nh); + if (err) + goto err_nexthop_neigh_init; + + return 0; + +err_nexthop_neigh_init: + mlxsw_sp_nexthop_rif_fini(nh); + return err; +} + +static void mlxsw_sp_nexthop6_fini(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop *nh) +{ + mlxsw_sp_nexthop_neigh_fini(mlxsw_sp, nh); + mlxsw_sp_nexthop_rif_fini(nh); +} + +static struct mlxsw_sp_nexthop_group * +mlxsw_sp_nexthop6_group_create(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib6_entry *fib6_entry) +{ + struct mlxsw_sp_nexthop_group *nh_grp; + struct mlxsw_sp_rt6 *mlxsw_sp_rt6; + struct mlxsw_sp_nexthop *nh; + size_t alloc_size; + int i = 0; + int err; + + alloc_size = sizeof(*nh_grp) + + fib6_entry->nrt6 * sizeof(struct mlxsw_sp_nexthop); + nh_grp = kzalloc(alloc_size, GFP_KERNEL); + if (!nh_grp) + return ERR_PTR(-ENOMEM); + INIT_LIST_HEAD(&nh_grp->fib_list); +#if IS_ENABLED(CONFIG_IPV6) + nh_grp->neigh_tbl = &nd_tbl; +#endif + mlxsw_sp_rt6 = list_first_entry(&fib6_entry->rt6_list, + struct mlxsw_sp_rt6, list); + nh_grp->gateway = !!(mlxsw_sp_rt6->rt->rt6i_flags & RTF_GATEWAY); + nh_grp->count = fib6_entry->nrt6; + for (i = 0; i < nh_grp->count; i++) { + struct rt6_info *rt = mlxsw_sp_rt6->rt; + + nh = &nh_grp->nexthops[i]; + err = mlxsw_sp_nexthop6_init(mlxsw_sp, nh_grp, nh, rt); + if (err) + goto err_nexthop6_init; + mlxsw_sp_rt6 = list_next_entry(mlxsw_sp_rt6, list); + } + mlxsw_sp_nexthop_group_refresh(mlxsw_sp, nh_grp); + return nh_grp; + +err_nexthop6_init: + for (i--; i >= 0; i--) { + nh = &nh_grp->nexthops[i]; + mlxsw_sp_nexthop6_fini(mlxsw_sp, nh); + } + kfree(nh_grp); + return ERR_PTR(err); +} + +static void +mlxsw_sp_nexthop6_group_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group *nh_grp) +{ + struct mlxsw_sp_nexthop *nh; + int i = nh_grp->count; + + for (i--; i >= 0; i--) { + nh = &nh_grp->nexthops[i]; + mlxsw_sp_nexthop6_fini(mlxsw_sp, nh); + } + mlxsw_sp_nexthop_group_refresh(mlxsw_sp, nh_grp); + WARN_ON(nh_grp->adj_index_valid); + kfree(nh_grp); +} + +static int mlxsw_sp_nexthop6_group_get(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib6_entry *fib6_entry) +{ + struct mlxsw_sp_nexthop_group *nh_grp; + + /* For now, don't consolidate nexthop groups */ + nh_grp = mlxsw_sp_nexthop6_group_create(mlxsw_sp, fib6_entry); + if (IS_ERR(nh_grp)) + return PTR_ERR(nh_grp); + + list_add_tail(&fib6_entry->common.nexthop_group_node, + &nh_grp->fib_list); + fib6_entry->common.nh_group = nh_grp; + + return 0; +} + +static void mlxsw_sp_nexthop6_group_put(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) +{ + struct mlxsw_sp_nexthop_group *nh_grp = fib_entry->nh_group; + + list_del(&fib_entry->nexthop_group_node); + if (!list_empty(&nh_grp->fib_list)) + return; + mlxsw_sp_nexthop6_group_destroy(mlxsw_sp, nh_grp); +} + +static int +mlxsw_sp_nexthop6_group_update(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib6_entry *fib6_entry) +{ + struct mlxsw_sp_nexthop_group *old_nh_grp = fib6_entry->common.nh_group; + int err; + + fib6_entry->common.nh_group = NULL; + list_del(&fib6_entry->common.nexthop_group_node); + + err = mlxsw_sp_nexthop6_group_get(mlxsw_sp, fib6_entry); + if (err) + goto err_nexthop6_group_get; + + /* In case this entry is offloaded, then the adjacency index + * currently associated with it in the device's table is that + * of the old group. Start using the new one instead. + */ + err = mlxsw_sp_fib_node_entry_add(mlxsw_sp, &fib6_entry->common); + if (err) + goto err_fib_node_entry_add; + + if (list_empty(&old_nh_grp->fib_list)) + mlxsw_sp_nexthop6_group_destroy(mlxsw_sp, old_nh_grp); + + return 0; + +err_fib_node_entry_add: + mlxsw_sp_nexthop6_group_put(mlxsw_sp, &fib6_entry->common); +err_nexthop6_group_get: + list_add_tail(&fib6_entry->common.nexthop_group_node, + &old_nh_grp->fib_list); + fib6_entry->common.nh_group = old_nh_grp; + return err; +} + +static int +mlxsw_sp_fib6_entry_nexthop_add(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib6_entry *fib6_entry, + struct rt6_info *rt) +{ + struct mlxsw_sp_rt6 *mlxsw_sp_rt6; + int err; + + mlxsw_sp_rt6 = mlxsw_sp_rt6_create(rt); + if (IS_ERR(mlxsw_sp_rt6)) + return PTR_ERR(mlxsw_sp_rt6); + + list_add_tail(&mlxsw_sp_rt6->list, &fib6_entry->rt6_list); + fib6_entry->nrt6++; + + err = mlxsw_sp_nexthop6_group_update(mlxsw_sp, fib6_entry); + if (err) + goto err_nexthop6_group_update; + + return 0; + +err_nexthop6_group_update: + fib6_entry->nrt6--; + list_del(&mlxsw_sp_rt6->list); + mlxsw_sp_rt6_destroy(mlxsw_sp_rt6); + return err; +} + +static void +mlxsw_sp_fib6_entry_nexthop_del(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib6_entry *fib6_entry, + struct rt6_info *rt) +{ + struct mlxsw_sp_rt6 *mlxsw_sp_rt6; + + mlxsw_sp_rt6 = mlxsw_sp_fib6_entry_rt_find(fib6_entry, rt); + if (WARN_ON(!mlxsw_sp_rt6)) + return; + + fib6_entry->nrt6--; + list_del(&mlxsw_sp_rt6->list); + mlxsw_sp_nexthop6_group_update(mlxsw_sp, fib6_entry); + mlxsw_sp_rt6_destroy(mlxsw_sp_rt6); +} + +static void mlxsw_sp_fib6_entry_type_set(struct mlxsw_sp_fib_entry *fib_entry, + const struct rt6_info *rt) +{ + /* Packets hitting RTF_REJECT routes need to be discarded by the + * stack. We can rely on their destination device not having a + * RIF (it's the loopback device) and can thus use action type + * local, which will cause them to be trapped with a lower + * priority than packets that need to be locally received. + */ + if (rt->rt6i_flags & RTF_LOCAL) + fib_entry->type = MLXSW_SP_FIB_ENTRY_TYPE_TRAP; + else if (rt->rt6i_flags & RTF_REJECT) + fib_entry->type = MLXSW_SP_FIB_ENTRY_TYPE_LOCAL; + else if (rt->rt6i_flags & RTF_GATEWAY) + fib_entry->type = MLXSW_SP_FIB_ENTRY_TYPE_REMOTE; + else + fib_entry->type = MLXSW_SP_FIB_ENTRY_TYPE_LOCAL; +} + +static void +mlxsw_sp_fib6_entry_rt_destroy_all(struct mlxsw_sp_fib6_entry *fib6_entry) +{ + struct mlxsw_sp_rt6 *mlxsw_sp_rt6, *tmp; + + list_for_each_entry_safe(mlxsw_sp_rt6, tmp, &fib6_entry->rt6_list, + list) { + fib6_entry->nrt6--; + list_del(&mlxsw_sp_rt6->list); + mlxsw_sp_rt6_destroy(mlxsw_sp_rt6); + } +} + +static struct mlxsw_sp_fib6_entry * +mlxsw_sp_fib6_entry_create(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_node *fib_node, + struct rt6_info *rt) +{ + struct mlxsw_sp_fib6_entry *fib6_entry; struct mlxsw_sp_fib_entry *fib_entry; + struct mlxsw_sp_rt6 *mlxsw_sp_rt6; + int err; + + fib6_entry = kzalloc(sizeof(*fib6_entry), GFP_KERNEL); + if (!fib6_entry) + return ERR_PTR(-ENOMEM); + fib_entry = &fib6_entry->common; + + mlxsw_sp_rt6 = mlxsw_sp_rt6_create(rt); + if (IS_ERR(mlxsw_sp_rt6)) { + err = PTR_ERR(mlxsw_sp_rt6); + goto err_rt6_create; + } + + mlxsw_sp_fib6_entry_type_set(fib_entry, mlxsw_sp_rt6->rt); + + INIT_LIST_HEAD(&fib6_entry->rt6_list); + list_add_tail(&mlxsw_sp_rt6->list, &fib6_entry->rt6_list); + fib6_entry->nrt6 = 1; + err = mlxsw_sp_nexthop6_group_get(mlxsw_sp, fib6_entry); + if (err) + goto err_nexthop6_group_get; + + fib_entry->fib_node = fib_node; + + return fib6_entry; + +err_nexthop6_group_get: + list_del(&mlxsw_sp_rt6->list); + mlxsw_sp_rt6_destroy(mlxsw_sp_rt6); +err_rt6_create: + kfree(fib6_entry); + return ERR_PTR(err); +} + +static void mlxsw_sp_fib6_entry_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib6_entry *fib6_entry) +{ + mlxsw_sp_nexthop6_group_put(mlxsw_sp, &fib6_entry->common); + mlxsw_sp_fib6_entry_rt_destroy_all(fib6_entry); + WARN_ON(fib6_entry->nrt6); + kfree(fib6_entry); +} + +static struct mlxsw_sp_fib6_entry * +mlxsw_sp_fib6_node_entry_find(const struct mlxsw_sp_fib_node *fib_node, + const struct rt6_info *nrt, bool replace) +{ + struct mlxsw_sp_fib6_entry *fib6_entry, *fallback = NULL; + + list_for_each_entry(fib6_entry, &fib_node->entry_list, common.list) { + struct rt6_info *rt = mlxsw_sp_fib6_entry_rt(fib6_entry); + + if (rt->rt6i_table->tb6_id > nrt->rt6i_table->tb6_id) + continue; + if (rt->rt6i_table->tb6_id != nrt->rt6i_table->tb6_id) + break; + if (replace && rt->rt6i_metric == nrt->rt6i_metric) { + if (mlxsw_sp_fib6_rt_can_mp(rt) == + mlxsw_sp_fib6_rt_can_mp(nrt)) + return fib6_entry; + if (mlxsw_sp_fib6_rt_can_mp(nrt)) + fallback = fallback ?: fib6_entry; + } + if (rt->rt6i_metric > nrt->rt6i_metric) + return fallback ?: fib6_entry; + } + + return fallback; +} + +static int +mlxsw_sp_fib6_node_list_insert(struct mlxsw_sp_fib6_entry *new6_entry, + bool replace) +{ + struct mlxsw_sp_fib_node *fib_node = new6_entry->common.fib_node; + struct rt6_info *nrt = mlxsw_sp_fib6_entry_rt(new6_entry); + struct mlxsw_sp_fib6_entry *fib6_entry; + + fib6_entry = mlxsw_sp_fib6_node_entry_find(fib_node, nrt, replace); + + if (replace && WARN_ON(!fib6_entry)) + return -EINVAL; + + if (fib6_entry) { + list_add_tail(&new6_entry->common.list, + &fib6_entry->common.list); + } else { + struct mlxsw_sp_fib6_entry *last; + + list_for_each_entry(last, &fib_node->entry_list, common.list) { + struct rt6_info *rt = mlxsw_sp_fib6_entry_rt(last); + + if (nrt->rt6i_table->tb6_id > rt->rt6i_table->tb6_id) + break; + fib6_entry = last; + } + + if (fib6_entry) + list_add(&new6_entry->common.list, + &fib6_entry->common.list); + else + list_add(&new6_entry->common.list, + &fib_node->entry_list); + } + + return 0; +} + +static void +mlxsw_sp_fib6_node_list_remove(struct mlxsw_sp_fib6_entry *fib6_entry) +{ + list_del(&fib6_entry->common.list); +} + +static int mlxsw_sp_fib6_node_entry_link(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib6_entry *fib6_entry, + bool replace) +{ + int err; + + err = mlxsw_sp_fib6_node_list_insert(fib6_entry, replace); + if (err) + return err; + + err = mlxsw_sp_fib_node_entry_add(mlxsw_sp, &fib6_entry->common); + if (err) + goto err_fib_node_entry_add; + + return 0; + +err_fib_node_entry_add: + mlxsw_sp_fib6_node_list_remove(fib6_entry); + return err; +} + +static void +mlxsw_sp_fib6_node_entry_unlink(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib6_entry *fib6_entry) +{ + mlxsw_sp_fib_node_entry_del(mlxsw_sp, &fib6_entry->common); + mlxsw_sp_fib6_node_list_remove(fib6_entry); +} + +static struct mlxsw_sp_fib6_entry * +mlxsw_sp_fib6_entry_lookup(struct mlxsw_sp *mlxsw_sp, + const struct rt6_info *rt) +{ + struct mlxsw_sp_fib6_entry *fib6_entry; + struct mlxsw_sp_fib_node *fib_node; + struct mlxsw_sp_fib *fib; + struct mlxsw_sp_vr *vr; + + vr = mlxsw_sp_vr_find(mlxsw_sp, rt->rt6i_table->tb6_id); + if (!vr) + return NULL; + fib = mlxsw_sp_vr_fib(vr, MLXSW_SP_L3_PROTO_IPV6); + + fib_node = mlxsw_sp_fib_node_lookup(fib, &rt->rt6i_dst.addr, + sizeof(rt->rt6i_dst.addr), + rt->rt6i_dst.plen); + if (!fib_node) + return NULL; + + list_for_each_entry(fib6_entry, &fib_node->entry_list, common.list) { + struct rt6_info *iter_rt = mlxsw_sp_fib6_entry_rt(fib6_entry); + + if (rt->rt6i_table->tb6_id == iter_rt->rt6i_table->tb6_id && + rt->rt6i_metric == iter_rt->rt6i_metric && + mlxsw_sp_fib6_entry_rt_find(fib6_entry, rt)) + return fib6_entry; + } + + return NULL; +} + +static void mlxsw_sp_fib6_entry_replace(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib6_entry *fib6_entry, + bool replace) +{ + struct mlxsw_sp_fib_node *fib_node = fib6_entry->common.fib_node; + struct mlxsw_sp_fib6_entry *replaced; + + if (!replace) + return; + + replaced = list_next_entry(fib6_entry, common.list); + + mlxsw_sp_fib6_node_entry_unlink(mlxsw_sp, replaced); + mlxsw_sp_fib6_entry_destroy(mlxsw_sp, replaced); + mlxsw_sp_fib_node_put(mlxsw_sp, fib_node); +} + +static int mlxsw_sp_router_fib6_add(struct mlxsw_sp *mlxsw_sp, + struct rt6_info *rt, bool replace) +{ + struct mlxsw_sp_fib6_entry *fib6_entry; + struct mlxsw_sp_fib_node *fib_node; + int err; + + if (mlxsw_sp->router->aborted) + return 0; + + if (rt->rt6i_src.plen) + return -EINVAL; + + if (mlxsw_sp_fib6_rt_should_ignore(rt)) + return 0; + + fib_node = mlxsw_sp_fib_node_get(mlxsw_sp, rt->rt6i_table->tb6_id, + &rt->rt6i_dst.addr, + sizeof(rt->rt6i_dst.addr), + rt->rt6i_dst.plen, + MLXSW_SP_L3_PROTO_IPV6); + if (IS_ERR(fib_node)) + return PTR_ERR(fib_node); + + /* Before creating a new entry, try to append route to an existing + * multipath entry. + */ + fib6_entry = mlxsw_sp_fib6_node_mp_entry_find(fib_node, rt, replace); + if (fib6_entry) { + err = mlxsw_sp_fib6_entry_nexthop_add(mlxsw_sp, fib6_entry, rt); + if (err) + goto err_fib6_entry_nexthop_add; + return 0; + } + + fib6_entry = mlxsw_sp_fib6_entry_create(mlxsw_sp, fib_node, rt); + if (IS_ERR(fib6_entry)) { + err = PTR_ERR(fib6_entry); + goto err_fib6_entry_create; + } + + err = mlxsw_sp_fib6_node_entry_link(mlxsw_sp, fib6_entry, replace); + if (err) + goto err_fib6_node_entry_link; + + mlxsw_sp_fib6_entry_replace(mlxsw_sp, fib6_entry, replace); + + return 0; + +err_fib6_node_entry_link: + mlxsw_sp_fib6_entry_destroy(mlxsw_sp, fib6_entry); +err_fib6_entry_create: +err_fib6_entry_nexthop_add: + mlxsw_sp_fib_node_put(mlxsw_sp, fib_node); + return err; +} + +static void mlxsw_sp_router_fib6_del(struct mlxsw_sp *mlxsw_sp, + struct rt6_info *rt) +{ + struct mlxsw_sp_fib6_entry *fib6_entry; struct mlxsw_sp_fib_node *fib_node; if (mlxsw_sp->router->aborted) return; - fib_entry = mlxsw_sp_fib4_entry_lookup(mlxsw_sp, fen_info); - if (WARN_ON(!fib_entry)) + if (mlxsw_sp_fib6_rt_should_ignore(rt)) + return; + + fib6_entry = mlxsw_sp_fib6_entry_lookup(mlxsw_sp, rt); + if (WARN_ON(!fib6_entry)) return; - fib_node = fib_entry->fib_node; - mlxsw_sp_fib4_node_entry_unlink(mlxsw_sp, fib_entry); - mlxsw_sp_fib4_entry_destroy(mlxsw_sp, fib_entry); - mlxsw_sp_fib4_node_put(mlxsw_sp, fib_node); + /* If route is part of a multipath entry, but not the last one + * removed, then only reduce its nexthop group. + */ + if (!list_is_singular(&fib6_entry->rt6_list)) { + mlxsw_sp_fib6_entry_nexthop_del(mlxsw_sp, fib6_entry, rt); + return; + } + + fib_node = fib6_entry->common.fib_node; + + mlxsw_sp_fib6_node_entry_unlink(mlxsw_sp, fib6_entry); + mlxsw_sp_fib6_entry_destroy(mlxsw_sp, fib6_entry); + mlxsw_sp_fib_node_put(mlxsw_sp, fib_node); } -static int mlxsw_sp_router_set_abort_trap(struct mlxsw_sp *mlxsw_sp) +static int __mlxsw_sp_router_set_abort_trap(struct mlxsw_sp *mlxsw_sp, + enum mlxsw_reg_ralxx_protocol proto, + u8 tree_id) { char ralta_pl[MLXSW_REG_RALTA_LEN]; char ralst_pl[MLXSW_REG_RALST_LEN]; int i, err; - mlxsw_reg_ralta_pack(ralta_pl, true, MLXSW_REG_RALXX_PROTOCOL_IPV4, - MLXSW_SP_LPM_TREE_MIN); + mlxsw_reg_ralta_pack(ralta_pl, true, proto, tree_id); err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ralta), ralta_pl); if (err) return err; - mlxsw_reg_ralst_pack(ralst_pl, 0xff, MLXSW_SP_LPM_TREE_MIN); + mlxsw_reg_ralst_pack(ralst_pl, 0xff, tree_id); err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ralst), ralst_pl); if (err) return err; @@ -2711,17 +3632,14 @@ static int mlxsw_sp_router_set_abort_trap(struct mlxsw_sp *mlxsw_sp) if (!mlxsw_sp_vr_is_used(vr)) continue; - mlxsw_reg_raltb_pack(raltb_pl, vr->id, - MLXSW_REG_RALXX_PROTOCOL_IPV4, - MLXSW_SP_LPM_TREE_MIN); + mlxsw_reg_raltb_pack(raltb_pl, vr->id, proto, tree_id); err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(raltb), raltb_pl); if (err) return err; - mlxsw_reg_ralue_pack4(ralue_pl, MLXSW_SP_L3_PROTO_IPV4, - MLXSW_REG_RALUE_OP_WRITE_WRITE, vr->id, 0, - 0); + mlxsw_reg_ralue_pack(ralue_pl, proto, + MLXSW_REG_RALUE_OP_WRITE_WRITE, vr->id, 0); mlxsw_reg_ralue_act_ip2me_pack(ralue_pl); err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ralue), ralue_pl); @@ -2732,17 +3650,33 @@ static int mlxsw_sp_router_set_abort_trap(struct mlxsw_sp *mlxsw_sp) return 0; } +static int mlxsw_sp_router_set_abort_trap(struct mlxsw_sp *mlxsw_sp) +{ + enum mlxsw_reg_ralxx_protocol proto = MLXSW_REG_RALXX_PROTOCOL_IPV4; + int err; + + err = __mlxsw_sp_router_set_abort_trap(mlxsw_sp, proto, + MLXSW_SP_LPM_TREE_MIN); + if (err) + return err; + + proto = MLXSW_REG_RALXX_PROTOCOL_IPV6; + return __mlxsw_sp_router_set_abort_trap(mlxsw_sp, proto, + MLXSW_SP_LPM_TREE_MIN + 1); +} + static void mlxsw_sp_fib4_node_flush(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_fib_node *fib_node) { - struct mlxsw_sp_fib_entry *fib_entry, *tmp; + struct mlxsw_sp_fib4_entry *fib4_entry, *tmp; - list_for_each_entry_safe(fib_entry, tmp, &fib_node->entry_list, list) { - bool do_break = &tmp->list == &fib_node->entry_list; + list_for_each_entry_safe(fib4_entry, tmp, &fib_node->entry_list, + common.list) { + bool do_break = &tmp->common.list == &fib_node->entry_list; - mlxsw_sp_fib4_node_entry_unlink(mlxsw_sp, fib_entry); - mlxsw_sp_fib4_entry_destroy(mlxsw_sp, fib_entry); - mlxsw_sp_fib4_node_put(mlxsw_sp, fib_node); + mlxsw_sp_fib4_node_entry_unlink(mlxsw_sp, fib4_entry); + mlxsw_sp_fib4_entry_destroy(mlxsw_sp, fib4_entry); + mlxsw_sp_fib_node_put(mlxsw_sp, fib_node); /* Break when entry list is empty and node was freed. * Otherwise, we'll access freed memory in the next * iteration. @@ -2752,6 +3686,23 @@ static void mlxsw_sp_fib4_node_flush(struct mlxsw_sp *mlxsw_sp, } } +static void mlxsw_sp_fib6_node_flush(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_node *fib_node) +{ + struct mlxsw_sp_fib6_entry *fib6_entry, *tmp; + + list_for_each_entry_safe(fib6_entry, tmp, &fib_node->entry_list, + common.list) { + bool do_break = &tmp->common.list == &fib_node->entry_list; + + mlxsw_sp_fib6_node_entry_unlink(mlxsw_sp, fib6_entry); + mlxsw_sp_fib6_entry_destroy(mlxsw_sp, fib6_entry); + mlxsw_sp_fib_node_put(mlxsw_sp, fib_node); + if (do_break) + break; + } +} + static void mlxsw_sp_fib_node_flush(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_fib_node *fib_node) { @@ -2760,7 +3711,7 @@ static void mlxsw_sp_fib_node_flush(struct mlxsw_sp *mlxsw_sp, mlxsw_sp_fib4_node_flush(mlxsw_sp, fib_node); break; case MLXSW_SP_L3_PROTO_IPV6: - WARN_ON_ONCE(1); + mlxsw_sp_fib6_node_flush(mlxsw_sp, fib_node); break; } } @@ -2791,10 +3742,17 @@ static void mlxsw_sp_router_fib_flush(struct mlxsw_sp *mlxsw_sp) if (!mlxsw_sp_vr_is_used(vr)) continue; mlxsw_sp_vr_fib_flush(mlxsw_sp, vr, MLXSW_SP_L3_PROTO_IPV4); + + /* If virtual router was only used for IPv4, then it's no + * longer used. + */ + if (!mlxsw_sp_vr_is_used(vr)) + continue; + mlxsw_sp_vr_fib_flush(mlxsw_sp, vr, MLXSW_SP_L3_PROTO_IPV6); } } -static void mlxsw_sp_router_fib4_abort(struct mlxsw_sp *mlxsw_sp) +static void mlxsw_sp_router_fib_abort(struct mlxsw_sp *mlxsw_sp) { int err; @@ -2811,6 +3769,7 @@ static void mlxsw_sp_router_fib4_abort(struct mlxsw_sp *mlxsw_sp) struct mlxsw_sp_fib_event_work { struct work_struct work; union { + struct fib6_entry_notifier_info fen6_info; struct fib_entry_notifier_info fen_info; struct fib_rule_notifier_info fr_info; struct fib_nh_notifier_info fnh_info; @@ -2819,7 +3778,7 @@ struct mlxsw_sp_fib_event_work { unsigned long event; }; -static void mlxsw_sp_router_fib_event_work(struct work_struct *work) +static void mlxsw_sp_router_fib4_event_work(struct work_struct *work) { struct mlxsw_sp_fib_event_work *fib_work = container_of(work, struct mlxsw_sp_fib_event_work, work); @@ -2839,7 +3798,7 @@ static void mlxsw_sp_router_fib_event_work(struct work_struct *work) err = mlxsw_sp_router_fib4_add(mlxsw_sp, &fib_work->fen_info, replace, append); if (err) - mlxsw_sp_router_fib4_abort(mlxsw_sp); + mlxsw_sp_router_fib_abort(mlxsw_sp); fib_info_put(fib_work->fen_info.fi); break; case FIB_EVENT_ENTRY_DEL: @@ -2850,13 +3809,13 @@ static void mlxsw_sp_router_fib_event_work(struct work_struct *work) case FIB_EVENT_RULE_DEL: rule = fib_work->fr_info.rule; if (!fib4_rule_default(rule) && !rule->l3mdev) - mlxsw_sp_router_fib4_abort(mlxsw_sp); + mlxsw_sp_router_fib_abort(mlxsw_sp); fib_rule_put(rule); break; case FIB_EVENT_NH_ADD: /* fall through */ case FIB_EVENT_NH_DEL: - mlxsw_sp_nexthop_event(mlxsw_sp, fib_work->event, - fib_work->fnh_info.fib_nh); + mlxsw_sp_nexthop4_event(mlxsw_sp, fib_work->event, + fib_work->fnh_info.fib_nh); fib_info_put(fib_work->fnh_info.fib_nh->nh_parent); break; } @@ -2864,6 +3823,87 @@ static void mlxsw_sp_router_fib_event_work(struct work_struct *work) kfree(fib_work); } +static void mlxsw_sp_router_fib6_event_work(struct work_struct *work) +{ + struct mlxsw_sp_fib_event_work *fib_work = + container_of(work, struct mlxsw_sp_fib_event_work, work); + struct mlxsw_sp *mlxsw_sp = fib_work->mlxsw_sp; + struct fib_rule *rule; + bool replace; + int err; + + rtnl_lock(); + switch (fib_work->event) { + case FIB_EVENT_ENTRY_REPLACE: /* fall through */ + case FIB_EVENT_ENTRY_ADD: + replace = fib_work->event == FIB_EVENT_ENTRY_REPLACE; + err = mlxsw_sp_router_fib6_add(mlxsw_sp, + fib_work->fen6_info.rt, replace); + if (err) + mlxsw_sp_router_fib_abort(mlxsw_sp); + mlxsw_sp_rt6_release(fib_work->fen6_info.rt); + break; + case FIB_EVENT_ENTRY_DEL: + mlxsw_sp_router_fib6_del(mlxsw_sp, fib_work->fen6_info.rt); + mlxsw_sp_rt6_release(fib_work->fen6_info.rt); + break; + case FIB_EVENT_RULE_ADD: /* fall through */ + case FIB_EVENT_RULE_DEL: + rule = fib_work->fr_info.rule; + if (!fib6_rule_default(rule) && !rule->l3mdev) + mlxsw_sp_router_fib_abort(mlxsw_sp); + fib_rule_put(rule); + break; + } + rtnl_unlock(); + kfree(fib_work); +} + +static void mlxsw_sp_router_fib4_event(struct mlxsw_sp_fib_event_work *fib_work, + struct fib_notifier_info *info) +{ + switch (fib_work->event) { + case FIB_EVENT_ENTRY_REPLACE: /* fall through */ + case FIB_EVENT_ENTRY_APPEND: /* fall through */ + case FIB_EVENT_ENTRY_ADD: /* fall through */ + case FIB_EVENT_ENTRY_DEL: + memcpy(&fib_work->fen_info, info, sizeof(fib_work->fen_info)); + /* Take referece on fib_info to prevent it from being + * freed while work is queued. Release it afterwards. + */ + fib_info_hold(fib_work->fen_info.fi); + break; + case FIB_EVENT_RULE_ADD: /* fall through */ + case FIB_EVENT_RULE_DEL: + memcpy(&fib_work->fr_info, info, sizeof(fib_work->fr_info)); + fib_rule_get(fib_work->fr_info.rule); + break; + case FIB_EVENT_NH_ADD: /* fall through */ + case FIB_EVENT_NH_DEL: + memcpy(&fib_work->fnh_info, info, sizeof(fib_work->fnh_info)); + fib_info_hold(fib_work->fnh_info.fib_nh->nh_parent); + break; + } +} + +static void mlxsw_sp_router_fib6_event(struct mlxsw_sp_fib_event_work *fib_work, + struct fib_notifier_info *info) +{ + switch (fib_work->event) { + case FIB_EVENT_ENTRY_REPLACE: /* fall through */ + case FIB_EVENT_ENTRY_ADD: /* fall through */ + case FIB_EVENT_ENTRY_DEL: + memcpy(&fib_work->fen6_info, info, sizeof(fib_work->fen6_info)); + rt6_hold(fib_work->fen6_info.rt); + break; + case FIB_EVENT_RULE_ADD: /* fall through */ + case FIB_EVENT_RULE_DEL: + memcpy(&fib_work->fr_info, info, sizeof(fib_work->fr_info)); + fib_rule_get(fib_work->fr_info.rule); + break; + } +} + /* Called with rcu_read_lock() */ static int mlxsw_sp_router_fib_event(struct notifier_block *nb, unsigned long event, void *ptr) @@ -2879,31 +3919,18 @@ static int mlxsw_sp_router_fib_event(struct notifier_block *nb, if (WARN_ON(!fib_work)) return NOTIFY_BAD; - INIT_WORK(&fib_work->work, mlxsw_sp_router_fib_event_work); router = container_of(nb, struct mlxsw_sp_router, fib_nb); fib_work->mlxsw_sp = router->mlxsw_sp; fib_work->event = event; - switch (event) { - case FIB_EVENT_ENTRY_REPLACE: /* fall through */ - case FIB_EVENT_ENTRY_APPEND: /* fall through */ - case FIB_EVENT_ENTRY_ADD: /* fall through */ - case FIB_EVENT_ENTRY_DEL: - memcpy(&fib_work->fen_info, ptr, sizeof(fib_work->fen_info)); - /* Take referece on fib_info to prevent it from being - * freed while work is queued. Release it afterwards. - */ - fib_info_hold(fib_work->fen_info.fi); - break; - case FIB_EVENT_RULE_ADD: /* fall through */ - case FIB_EVENT_RULE_DEL: - memcpy(&fib_work->fr_info, ptr, sizeof(fib_work->fr_info)); - fib_rule_get(fib_work->fr_info.rule); + switch (info->family) { + case AF_INET: + INIT_WORK(&fib_work->work, mlxsw_sp_router_fib4_event_work); + mlxsw_sp_router_fib4_event(fib_work, info); break; - case FIB_EVENT_NH_ADD: /* fall through */ - case FIB_EVENT_NH_DEL: - memcpy(&fib_work->fnh_info, ptr, sizeof(fib_work->fnh_info)); - fib_info_hold(fib_work->fnh_info.fib_nh->nh_parent); + case AF_INET6: + INIT_WORK(&fib_work->work, mlxsw_sp_router_fib6_event_work); + mlxsw_sp_router_fib6_event(fib_work, info); break; } @@ -2948,17 +3975,28 @@ static void mlxsw_sp_router_rif_gone_sync(struct mlxsw_sp *mlxsw_sp, mlxsw_sp_neigh_rif_gone_sync(mlxsw_sp, rif); } -static bool mlxsw_sp_rif_should_config(struct mlxsw_sp_rif *rif, - const struct in_device *in_dev, - unsigned long event) +static bool +mlxsw_sp_rif_should_config(struct mlxsw_sp_rif *rif, struct net_device *dev, + unsigned long event) { + struct inet6_dev *inet6_dev; + bool addr_list_empty = true; + struct in_device *idev; + switch (event) { case NETDEV_UP: - if (!rif) - return true; - return false; + return rif == NULL; case NETDEV_DOWN: - if (rif && !in_dev->ifa_list && + idev = __in_dev_get_rtnl(dev); + if (idev && idev->ifa_list) + addr_list_empty = false; + + inet6_dev = __in6_dev_get(dev); + if (addr_list_empty && inet6_dev && + !list_empty(&inet6_dev->addr_list)) + addr_list_empty = false; + + if (rif && addr_list_empty && !netif_is_l3_slave(rif->dev)) return true; /* It is possible we already removed the RIF ourselves @@ -3356,7 +4394,7 @@ int mlxsw_sp_inetaddr_event(struct notifier_block *unused, goto out; rif = mlxsw_sp_rif_find_by_dev(mlxsw_sp, dev); - if (!mlxsw_sp_rif_should_config(rif, ifa->ifa_dev, event)) + if (!mlxsw_sp_rif_should_config(rif, dev, event)) goto out; err = __mlxsw_sp_inetaddr_event(dev, event); @@ -3364,6 +4402,61 @@ out: return notifier_from_errno(err); } +struct mlxsw_sp_inet6addr_event_work { + struct work_struct work; + struct net_device *dev; + unsigned long event; +}; + +static void mlxsw_sp_inet6addr_event_work(struct work_struct *work) +{ + struct mlxsw_sp_inet6addr_event_work *inet6addr_work = + container_of(work, struct mlxsw_sp_inet6addr_event_work, work); + struct net_device *dev = inet6addr_work->dev; + unsigned long event = inet6addr_work->event; + struct mlxsw_sp *mlxsw_sp; + struct mlxsw_sp_rif *rif; + + rtnl_lock(); + mlxsw_sp = mlxsw_sp_lower_get(dev); + if (!mlxsw_sp) + goto out; + + rif = mlxsw_sp_rif_find_by_dev(mlxsw_sp, dev); + if (!mlxsw_sp_rif_should_config(rif, dev, event)) + goto out; + + __mlxsw_sp_inetaddr_event(dev, event); +out: + rtnl_unlock(); + dev_put(dev); + kfree(inet6addr_work); +} + +/* Called with rcu_read_lock() */ +int mlxsw_sp_inet6addr_event(struct notifier_block *unused, + unsigned long event, void *ptr) +{ + struct inet6_ifaddr *if6 = (struct inet6_ifaddr *) ptr; + struct mlxsw_sp_inet6addr_event_work *inet6addr_work; + struct net_device *dev = if6->idev->dev; + + if (!mlxsw_sp_port_dev_lower_find_rcu(dev)) + return NOTIFY_DONE; + + inet6addr_work = kzalloc(sizeof(*inet6addr_work), GFP_ATOMIC); + if (!inet6addr_work) + return NOTIFY_BAD; + + INIT_WORK(&inet6addr_work->work, mlxsw_sp_inet6addr_event_work); + inet6addr_work->dev = dev; + inet6addr_work->event = event; + dev_hold(dev); + mlxsw_core_schedule_work(&inet6addr_work->work); + + return NOTIFY_DONE; +} + static int mlxsw_sp_rif_edit(struct mlxsw_sp *mlxsw_sp, u16 rif_index, const char *mac, int mtu) { @@ -3565,6 +4658,11 @@ static int mlxsw_sp_rif_vlan_configure(struct mlxsw_sp_rif *rif) if (err) return err; + err = mlxsw_sp_fid_flood_set(rif->fid, MLXSW_SP_FLOOD_TYPE_MC, + mlxsw_sp_router_port(mlxsw_sp), true); + if (err) + goto err_fid_mc_flood_set; + err = mlxsw_sp_fid_flood_set(rif->fid, MLXSW_SP_FLOOD_TYPE_BC, mlxsw_sp_router_port(mlxsw_sp), true); if (err) @@ -3573,6 +4671,9 @@ static int mlxsw_sp_rif_vlan_configure(struct mlxsw_sp_rif *rif) return 0; err_fid_bc_flood_set: + mlxsw_sp_fid_flood_set(rif->fid, MLXSW_SP_FLOOD_TYPE_MC, + mlxsw_sp_router_port(mlxsw_sp), false); +err_fid_mc_flood_set: mlxsw_sp_rif_vlan_fid_op(rif, MLXSW_REG_RITR_VLAN_IF, vid, false); return err; } @@ -3584,6 +4685,8 @@ static void mlxsw_sp_rif_vlan_deconfigure(struct mlxsw_sp_rif *rif) mlxsw_sp_fid_flood_set(rif->fid, MLXSW_SP_FLOOD_TYPE_BC, mlxsw_sp_router_port(mlxsw_sp), false); + mlxsw_sp_fid_flood_set(rif->fid, MLXSW_SP_FLOOD_TYPE_MC, + mlxsw_sp_router_port(mlxsw_sp), false); mlxsw_sp_rif_vlan_fid_op(rif, MLXSW_REG_RITR_VLAN_IF, vid, false); } @@ -3614,6 +4717,11 @@ static int mlxsw_sp_rif_fid_configure(struct mlxsw_sp_rif *rif) if (err) return err; + err = mlxsw_sp_fid_flood_set(rif->fid, MLXSW_SP_FLOOD_TYPE_MC, + mlxsw_sp_router_port(mlxsw_sp), true); + if (err) + goto err_fid_mc_flood_set; + err = mlxsw_sp_fid_flood_set(rif->fid, MLXSW_SP_FLOOD_TYPE_BC, mlxsw_sp_router_port(mlxsw_sp), true); if (err) @@ -3622,6 +4730,9 @@ static int mlxsw_sp_rif_fid_configure(struct mlxsw_sp_rif *rif) return 0; err_fid_bc_flood_set: + mlxsw_sp_fid_flood_set(rif->fid, MLXSW_SP_FLOOD_TYPE_MC, + mlxsw_sp_router_port(mlxsw_sp), false); +err_fid_mc_flood_set: mlxsw_sp_rif_vlan_fid_op(rif, MLXSW_REG_RITR_FID_IF, fid_index, false); return err; } @@ -3633,6 +4744,8 @@ static void mlxsw_sp_rif_fid_deconfigure(struct mlxsw_sp_rif *rif) mlxsw_sp_fid_flood_set(rif->fid, MLXSW_SP_FLOOD_TYPE_BC, mlxsw_sp_router_port(mlxsw_sp), false); + mlxsw_sp_fid_flood_set(rif->fid, MLXSW_SP_FLOOD_TYPE_MC, + mlxsw_sp_router_port(mlxsw_sp), false); mlxsw_sp_rif_vlan_fid_op(rif, MLXSW_REG_RITR_FID_IF, fid_index, false); } @@ -3704,7 +4817,7 @@ static int __mlxsw_sp_router_init(struct mlxsw_sp *mlxsw_sp) return -EIO; max_rifs = MLXSW_CORE_RES_GET(mlxsw_sp->core, MAX_RIFS); - mlxsw_reg_rgcr_pack(rgcr_pl, true); + mlxsw_reg_rgcr_pack(rgcr_pl, true, true); mlxsw_reg_rgcr_max_router_interfaces_set(rgcr_pl, max_rifs); err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(rgcr), rgcr_pl); if (err) @@ -3716,7 +4829,7 @@ static void __mlxsw_sp_router_fini(struct mlxsw_sp *mlxsw_sp) { char rgcr_pl[MLXSW_REG_RGCR_LEN]; - mlxsw_reg_rgcr_pack(rgcr_pl, false); + mlxsw_reg_rgcr_pack(rgcr_pl, false, false); mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(rgcr), rgcr_pl); } diff --git a/drivers/net/ethernet/mellanox/mlxsw/trap.h b/drivers/net/ethernet/mellanox/mlxsw/trap.h index 12b5ed58f3eb..61652396bf75 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/trap.h +++ b/drivers/net/ethernet/mellanox/mlxsw/trap.h @@ -61,11 +61,32 @@ enum { MLXSW_TRAP_ID_MTUERROR = 0x52, MLXSW_TRAP_ID_TTLERROR = 0x53, MLXSW_TRAP_ID_LBERROR = 0x54, - MLXSW_TRAP_ID_OSPF = 0x55, + MLXSW_TRAP_ID_IPV4_OSPF = 0x55, MLXSW_TRAP_ID_IP2ME = 0x5F, + MLXSW_TRAP_ID_IPV6_UNSPECIFIED_ADDRESS = 0x60, + MLXSW_TRAP_ID_IPV6_LINK_LOCAL_DEST = 0x61, + MLXSW_TRAP_ID_IPV6_LINK_LOCAL_SRC = 0x62, + MLXSW_TRAP_ID_IPV6_ALL_NODES_LINK = 0x63, + MLXSW_TRAP_ID_IPV6_OSPF = 0x64, + MLXSW_TRAP_ID_IPV6_MLDV12_LISTENER_QUERY = 0x65, + MLXSW_TRAP_ID_IPV6_MLDV1_LISTENER_REPORT = 0x66, + MLXSW_TRAP_ID_IPV6_MLDV1_LISTENER_DONE = 0x67, + MLXSW_TRAP_ID_IPV6_MLDV2_LISTENER_REPORT = 0x68, + MLXSW_TRAP_ID_IPV6_DHCP = 0x69, + MLXSW_TRAP_ID_IPV6_ALL_ROUTERS_LINK = 0x6F, MLXSW_TRAP_ID_RTR_INGRESS0 = 0x70, - MLXSW_TRAP_ID_BGP_IPV4 = 0x88, + MLXSW_TRAP_ID_IPV4_BGP = 0x88, + MLXSW_TRAP_ID_IPV6_BGP = 0x89, + MLXSW_TRAP_ID_L3_IPV6_ROUTER_SOLICITATION = 0x8A, + MLXSW_TRAP_ID_L3_IPV6_ROUTER_ADVERTISMENT = 0x8B, + MLXSW_TRAP_ID_L3_IPV6_NEIGHBOR_SOLICITATION = 0x8C, + MLXSW_TRAP_ID_L3_IPV6_NEIGHBOR_ADVERTISMENT = 0x8D, + MLXSW_TRAP_ID_L3_IPV6_REDIRECTION = 0x8E, MLXSW_TRAP_ID_HOST_MISS_IPV4 = 0x90, + MLXSW_TRAP_ID_IPV6_MC_LINK_LOCAL_DEST = 0x91, + MLXSW_TRAP_ID_HOST_MISS_IPV6 = 0x92, + MLXSW_TRAP_ID_ROUTER_ALERT_IPV4 = 0xD6, + MLXSW_TRAP_ID_ROUTER_ALERT_IPV6 = 0xD7, MLXSW_TRAP_ID_ACL0 = 0x1C0, MLXSW_TRAP_ID_MAX = 0x1FF |