net/mlx5: E-Switch, Add SR-IOV (FDB) support

Enabling E-Switch SRIOV for nvfs+1 vports.

Create E-Switch FDB for L2 UC/MC mac steering between VFs/PF and
external vport (Uplink).

FDB contains forwarding rules such as:
	UC MAC0 -> vport0(PF).
	UC MAC1 -> vport1.
	UC MAC2 -> vport2.
	MC MACX -> vport0, vport2, Uplink.
	MC MACY -> vport1, Uplink.

For unmatched traffic FDB has the following default rules:
	Unmached Traffic (src vport != Uplink) -> Uplink.
	Unmached Traffic (src vport == Uplink) -> vport0(PF).

FDB rules population:
Each NIC vport (VF) will notify E-Switch manager of its UC/MC vport
context changes via modify vport context command, which will be
translated to an event that will be handled by E-Switch manager (PF)
which will update FDB table accordingly.
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>
Signed-off-by: NOr Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

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

@@ -86,10 +86,25 @@ struct l2addr_node {
 	kfree(ptr);                                         \
 })
 
+struct mlx5_flow_rule {
+	void             *ft;
+	u32              fi;
+	u8               match_criteria_enable;
+	u32              *match_criteria;
+	u32              *match_value;
+	u32              action;
+	u32              flow_tag;
+	bool             valid;
+	atomic_t         refcount;
+	struct mutex     mutex; /* protect flow rule updates */
+	struct list_head dest_list;
+};
+
 struct mlx5_vport {
 	struct mlx5_core_dev    *dev;
 	int                     vport;
 	struct hlist_head       uc_list[MLX5_L2_ADDR_HASH_SIZE];
+	struct hlist_head       mc_list[MLX5_L2_ADDR_HASH_SIZE];
 	struct work_struct      vport_change_handler;
 
 	/* This spinlock protects access to vport data, between
@@ -98,6 +113,7 @@ struct mlx5_vport {
 	 */
 	spinlock_t              lock; /* vport events sync */
 	bool                    enabled;
+	u16                     enabled_events;
 };
 
 struct mlx5_l2_table {
@@ -106,17 +122,26 @@ struct mlx5_l2_table {
 	unsigned long        *bitmap;
 };
 
+struct mlx5_eswitch_fdb {
+	void *fdb;
+};
+
 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;
 	struct mlx5_vport       *vports;
 	int                     total_vports;
+	int                     enabled_vports;
 };
 
 /* E-Switch API */
 int mlx5_eswitch_init(struct mlx5_core_dev *dev);
 void mlx5_eswitch_cleanup(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);
+void mlx5_eswitch_disable_sriov(struct mlx5_eswitch *esw);
 
 #endif /* __MLX5_ESWITCH_H__ */

drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h

@@ -36,6 +36,7 @@
 #include <linux/types.h>
 #include <linux/kernel.h>
 #include <linux/sched.h>
+#include <linux/if_link.h>
 
 #define DRIVER_NAME "mlx5_core"
 #define DRIVER_VERSION "3.0-1"

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

@@ -33,6 +33,9 @@
 #include <linux/pci.h>
 #include <linux/mlx5/driver.h>
 #include "mlx5_core.h"
+#ifdef CONFIG_MLX5_CORE_EN
+#include "eswitch.h"
+#endif
 
 static void enable_vfs(struct mlx5_core_dev *dev, int num_vfs)
 {
@@ -144,13 +147,15 @@ int mlx5_core_sriov_configure(struct pci_dev *pdev, int num_vfs)
 	mlx5_core_cleanup_vfs(dev);
 
 	if (!num_vfs) {
+#ifdef CONFIG_MLX5_CORE_EN
+		mlx5_eswitch_disable_sriov(dev->priv.eswitch);
+#endif
 		kfree(sriov->vfs_ctx);
 		sriov->vfs_ctx = NULL;
 		if (!pci_vfs_assigned(pdev))
 			pci_disable_sriov(pdev);
 		else
 			pr_info("unloading PF driver while leaving orphan VFs\n");
-
 		return 0;
 	}
 
@@ -161,6 +166,9 @@ int mlx5_core_sriov_configure(struct pci_dev *pdev, int num_vfs)
 	}
 
 	mlx5_core_init_vfs(dev, num_vfs);
+#ifdef CONFIG_MLX5_CORE_EN
+	mlx5_eswitch_enable_sriov(dev->priv.eswitch, num_vfs);
+#endif
 
 	return num_vfs;
 }
@@ -199,6 +207,10 @@ int mlx5_sriov_init(struct mlx5_core_dev *dev)
 	sriov->enabled_vfs = cur_vfs;
 
 	mlx5_core_init_vfs(dev, cur_vfs);
+#ifdef CONFIG_MLX5_CORE_EN
+	if (cur_vfs)
+		mlx5_eswitch_enable_sriov(dev->priv.eswitch, cur_vfs);
+#endif
 
 	enable_vfs(dev, cur_vfs);
 

include/linux/mlx5/device.h

@@ -1074,6 +1074,12 @@ enum {
 	VPORT_STATE_UP			= 0x1,
 };
 
+enum {
+	MLX5_ESW_VPORT_ADMIN_STATE_DOWN  = 0x0,
+	MLX5_ESW_VPORT_ADMIN_STATE_UP    = 0x1,
+	MLX5_ESW_VPORT_ADMIN_STATE_AUTO  = 0x2,
+};
+
 enum {
 	MLX5_L3_PROT_TYPE_IPV4		= 0,
 	MLX5_L3_PROT_TYPE_IPV6		= 1,

include/linux/mlx5/flow_table.h

@@ -41,6 +41,15 @@ struct mlx5_flow_table_group {
 	u32	match_criteria[MLX5_ST_SZ_DW(fte_match_param)];
 };
 
+struct mlx5_flow_destination {
+	enum mlx5_flow_destination_type	type;
+	union {
+		u32			tir_num;
+		void			*ft;
+		u32			vport_num;
+	};
+};
+
 void *mlx5_create_flow_table(struct mlx5_core_dev *dev, u8 level, u8 table_type,
 			     u16 num_groups,
 			     struct mlx5_flow_table_group *group);

include/linux/mlx5/mlx5_ifc.h

@@ -827,9 +827,10 @@ struct mlx5_ifc_cmd_hca_cap_bits {
 	u8         reserved_69[0x220];
 };
 
-enum {
-	MLX5_DEST_FORMAT_STRUCT_DESTINATION_TYPE_FLOW_TABLE_  = 0x1,
-	MLX5_DEST_FORMAT_STRUCT_DESTINATION_TYPE_TIR          = 0x2,
+enum mlx5_flow_destination_type {
+	MLX5_FLOW_DESTINATION_TYPE_VPORT        = 0x0,
+	MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE   = 0x1,
+	MLX5_FLOW_DESTINATION_TYPE_TIR          = 0x2,
 };
 
 struct mlx5_ifc_dest_format_struct_bits {