Merge branch 'mlx5-next' of git://git.kernel.org/pub/scm/linux/kernel/git/mellanox/linux

Saeed Mahameed says:

====================
mlx5-next branch 2021-12-15

Hi Dave, Jakub, Jason

This pulls mlx5-next branch into net-next and rdma branches.
All patches already reviewed on both rdma and netdev mailing lists.

Please pull and let me know if there's any problem.

1) Add multiple FDB steering priorities [1]
2) Introduce HW bits needed to configure MAC list size of VF/SF.
   Required for ("net/mlx5: Memory optimizations") upcoming series [2].

[1] https://lore.kernel.org/netdev/20211201193621.9129-1-saeed@kernel.org/
[2] https://lore.kernel.org/lkml/20211208141722.13646-1-shayd@nvidia.com/
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

6 files changed, 76 insertions, 28 deletions

diff --git a/drivers/infiniband/hw/mlx5/fs.c b/drivers/infiniband/hw/mlx5/fs.c
index b780185d9dc6..661ed2b44508 100644
--- a/drivers/infiniband/hw/mlx5/fs.c
+++ b/drivers/infiniband/hw/mlx5/fs.c
@@ -1508,7 +1508,7 @@ _get_flow_table(struct mlx5_ib_dev *dev,
 		    !esw_encap)
 			flags |= MLX5_FLOW_TABLE_TUNNEL_EN_REFORMAT;
 		break;
-	case MLX5_FLOW_NAMESPACE_FDB:
+	case MLX5_FLOW_NAMESPACE_FDB_BYPASS:
 		max_table_size = BIT(
 			MLX5_CAP_ESW_FLOWTABLE_FDB(dev->mdev, log_max_ft_size));
 		if (MLX5_CAP_ESW_FLOWTABLE_FDB(dev->mdev, decap) && esw_encap)
@@ -1517,7 +1517,7 @@ _get_flow_table(struct mlx5_ib_dev *dev,
 					       reformat_l3_tunnel_to_l2) &&
 		    esw_encap)
 			flags |= MLX5_FLOW_TABLE_TUNNEL_EN_REFORMAT;
-		priority = FDB_BYPASS_PATH;
+		priority = fs_matcher->priority;
 		break;
 	case MLX5_FLOW_NAMESPACE_RDMA_RX:
 		max_table_size = BIT(
@@ -1546,8 +1546,8 @@ _get_flow_table(struct mlx5_ib_dev *dev,
 	case MLX5_FLOW_NAMESPACE_EGRESS:
 		prio = &dev->flow_db->egress_prios[priority];
 		break;
-	case MLX5_FLOW_NAMESPACE_FDB:
-		prio = &dev->flow_db->fdb;
+	case MLX5_FLOW_NAMESPACE_FDB_BYPASS:
+		prio = &dev->flow_db->fdb[priority];
 		break;
 	case MLX5_FLOW_NAMESPACE_RDMA_RX:
 		prio = &dev->flow_db->rdma_rx[priority];
@@ -1937,7 +1937,7 @@ mlx5_ib_ft_type_to_namespace(enum mlx5_ib_uapi_flow_table_type table_type,
 		*namespace = MLX5_FLOW_NAMESPACE_EGRESS;
 		break;
 	case MLX5_IB_UAPI_FLOW_TABLE_TYPE_FDB:
-		*namespace = MLX5_FLOW_NAMESPACE_FDB;
+		*namespace = MLX5_FLOW_NAMESPACE_FDB_BYPASS;
 		break;
 	case MLX5_IB_UAPI_FLOW_TABLE_TYPE_RDMA_RX:
 		*namespace = MLX5_FLOW_NAMESPACE_RDMA_RX;
@@ -2029,8 +2029,8 @@ static int get_dests(struct uverbs_attr_bundle *attrs,
 	}
 
 	/* Allow only DEVX object, drop as dest for FDB */
-	if (fs_matcher->ns_type == MLX5_FLOW_NAMESPACE_FDB && !(dest_devx ||
-	     (*flags & MLX5_IB_ATTR_CREATE_FLOW_FLAGS_DROP)))
+	if (fs_matcher->ns_type == MLX5_FLOW_NAMESPACE_FDB_BYPASS &&
+	    !(dest_devx || (*flags & MLX5_IB_ATTR_CREATE_FLOW_FLAGS_DROP)))
 		return -EINVAL;
 
 	/* Allow only DEVX object or QP as dest when inserting to RDMA_RX */
@@ -2050,7 +2050,7 @@ static int get_dests(struct uverbs_attr_bundle *attrs,
 		if (!is_flow_dest(devx_obj, dest_id, dest_type))
 			return -EINVAL;
 		/* Allow only flow table as dest when inserting to FDB or RDMA_RX */
-		if ((fs_matcher->ns_type == MLX5_FLOW_NAMESPACE_FDB ||
+		if ((fs_matcher->ns_type == MLX5_FLOW_NAMESPACE_FDB_BYPASS ||
 		     fs_matcher->ns_type == MLX5_FLOW_NAMESPACE_RDMA_RX) &&
 		    *dest_type != MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE)
 			return -EINVAL;
@@ -2320,7 +2320,7 @@ static int UVERBS_HANDLER(MLX5_IB_METHOD_FLOW_MATCHER_CREATE)(
 	if (err)
 		goto end;
 
-	if (obj->ns_type == MLX5_FLOW_NAMESPACE_FDB &&
+	if (obj->ns_type == MLX5_FLOW_NAMESPACE_FDB_BYPASS &&
 	    mlx5_eswitch_mode(dev->mdev) != MLX5_ESWITCH_OFFLOADS) {
 		err = -EINVAL;
 		goto end;
diff --git a/drivers/infiniband/hw/mlx5/mlx5_ib.h b/drivers/infiniband/hw/mlx5/mlx5_ib.h
index e636e954f6bf..e3c33be9c5a0 100644
--- a/drivers/infiniband/hw/mlx5/mlx5_ib.h
+++ b/drivers/infiniband/hw/mlx5/mlx5_ib.h
@@ -232,6 +232,7 @@ enum {
 #define MLX5_IB_NUM_FLOW_FT		(MLX5_IB_FLOW_LEFTOVERS_PRIO + 1)
 #define MLX5_IB_NUM_SNIFFER_FTS		2
 #define MLX5_IB_NUM_EGRESS_FTS		1
+#define MLX5_IB_NUM_FDB_FTS		MLX5_BY_PASS_NUM_REGULAR_PRIOS
 struct mlx5_ib_flow_prio {
 	struct mlx5_flow_table		*flow_table;
 	unsigned int			refcount;
@@ -276,7 +277,7 @@ struct mlx5_ib_flow_db {
 	struct mlx5_ib_flow_prio	egress_prios[MLX5_IB_NUM_FLOW_FT];
 	struct mlx5_ib_flow_prio	sniffer[MLX5_IB_NUM_SNIFFER_FTS];
 	struct mlx5_ib_flow_prio	egress[MLX5_IB_NUM_EGRESS_FTS];
-	struct mlx5_ib_flow_prio	fdb;
+	struct mlx5_ib_flow_prio	fdb[MLX5_IB_NUM_FDB_FTS];
 	struct mlx5_ib_flow_prio	rdma_rx[MLX5_IB_NUM_FLOW_FT];
 	struct mlx5_ib_flow_prio	rdma_tx[MLX5_IB_NUM_FLOW_FT];
 	struct mlx5_ib_flow_prio	opfcs[MLX5_IB_OPCOUNTER_MAX];
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c
index 750b21124a1a..762b9730a897 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c
@@ -788,7 +788,8 @@ static int mlx5_cmd_packet_reformat_alloc(struct mlx5_flow_root_namespace *ns,
 	int err;
 	u32 *in;
 
-	if (namespace == MLX5_FLOW_NAMESPACE_FDB)
+	if (namespace == MLX5_FLOW_NAMESPACE_FDB ||
+	    namespace == MLX5_FLOW_NAMESPACE_FDB_BYPASS)
 		max_encap_size = MLX5_CAP_ESW(dev, max_encap_header_size);
 	else
 		max_encap_size = MLX5_CAP_FLOWTABLE(dev, max_encap_header_size);
@@ -860,6 +861,7 @@ static int mlx5_cmd_modify_header_alloc(struct mlx5_flow_root_namespace *ns,
 
 	switch (namespace) {
 	case MLX5_FLOW_NAMESPACE_FDB:
+	case MLX5_FLOW_NAMESPACE_FDB_BYPASS:
 		max_actions = MLX5_CAP_ESW_FLOWTABLE_FDB(dev, max_modify_header_actions);
 		table_type = FS_FT_FDB;
 		break;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
index 386ab9a2d490..a8671d7b7ca8 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
@@ -2206,6 +2206,22 @@ struct mlx5_flow_namespace *mlx5_get_fdb_sub_ns(struct mlx5_core_dev *dev,
 }
 EXPORT_SYMBOL(mlx5_get_fdb_sub_ns);
 
+static bool is_nic_rx_ns(enum mlx5_flow_namespace_type type)
+{
+	switch (type) {
+	case MLX5_FLOW_NAMESPACE_BYPASS:
+	case MLX5_FLOW_NAMESPACE_LAG:
+	case MLX5_FLOW_NAMESPACE_OFFLOADS:
+	case MLX5_FLOW_NAMESPACE_ETHTOOL:
+	case MLX5_FLOW_NAMESPACE_KERNEL:
+	case MLX5_FLOW_NAMESPACE_LEFTOVERS:
+	case MLX5_FLOW_NAMESPACE_ANCHOR:
+		return true;
+	default:
+		return false;
+	}
+}
+
 struct mlx5_flow_namespace *mlx5_get_flow_namespace(struct mlx5_core_dev *dev,
 						    enum mlx5_flow_namespace_type type)
 {
@@ -2235,31 +2251,39 @@ struct mlx5_flow_namespace *mlx5_get_flow_namespace(struct mlx5_core_dev *dev,
 		if (steering->sniffer_tx_root_ns)
 			return &steering->sniffer_tx_root_ns->ns;
 		return NULL;
-	default:
+	case MLX5_FLOW_NAMESPACE_FDB_BYPASS:
+		root_ns = steering->fdb_root_ns;
+		prio =  FDB_BYPASS_PATH;
 		break;
-	}
-
-	if (type == MLX5_FLOW_NAMESPACE_EGRESS ||
-	    type == MLX5_FLOW_NAMESPACE_EGRESS_KERNEL) {
+	case MLX5_FLOW_NAMESPACE_EGRESS:
+	case MLX5_FLOW_NAMESPACE_EGRESS_KERNEL:
 		root_ns = steering->egress_root_ns;
 		prio = type - MLX5_FLOW_NAMESPACE_EGRESS;
-	} else if (type == MLX5_FLOW_NAMESPACE_RDMA_RX) {
+		break;
+	case MLX5_FLOW_NAMESPACE_RDMA_RX:
 		root_ns = steering->rdma_rx_root_ns;
 		prio = RDMA_RX_BYPASS_PRIO;
-	} else if (type == MLX5_FLOW_NAMESPACE_RDMA_RX_KERNEL) {
+		break;
+	case MLX5_FLOW_NAMESPACE_RDMA_RX_KERNEL:
 		root_ns = steering->rdma_rx_root_ns;
 		prio = RDMA_RX_KERNEL_PRIO;
-	} else if (type == MLX5_FLOW_NAMESPACE_RDMA_TX) {
+		break;
+	case MLX5_FLOW_NAMESPACE_RDMA_TX:
 		root_ns = steering->rdma_tx_root_ns;
-	} else if (type == MLX5_FLOW_NAMESPACE_RDMA_RX_COUNTERS) {
+		break;
+	case MLX5_FLOW_NAMESPACE_RDMA_RX_COUNTERS:
 		root_ns = steering->rdma_rx_root_ns;
 		prio = RDMA_RX_COUNTERS_PRIO;
-	} else if (type == MLX5_FLOW_NAMESPACE_RDMA_TX_COUNTERS) {
+		break;
+	case MLX5_FLOW_NAMESPACE_RDMA_TX_COUNTERS:
 		root_ns = steering->rdma_tx_root_ns;
 		prio = RDMA_TX_COUNTERS_PRIO;
-	} else { /* Must be NIC RX */
+		break;
+	default: /* Must be NIC RX */
+		WARN_ON(!is_nic_rx_ns(type));
 		root_ns = steering->root_ns;
 		prio = type;
+		break;
 	}
 
 	if (!root_ns)
@@ -2822,6 +2846,28 @@ static int create_fdb_fast_path(struct mlx5_flow_steering *steering)
 	return 0;
 }
 
+static int create_fdb_bypass(struct mlx5_flow_steering *steering)
+{
+	struct mlx5_flow_namespace *ns;
+	struct fs_prio *prio;
+	int i;
+
+	prio = fs_create_prio(&steering->fdb_root_ns->ns, FDB_BYPASS_PATH, 0);
+	if (IS_ERR(prio))
+		return PTR_ERR(prio);
+
+	ns = fs_create_namespace(prio, MLX5_FLOW_TABLE_MISS_ACTION_DEF);
+	if (IS_ERR(ns))
+		return PTR_ERR(ns);
+
+	for (i = 0; i < MLX5_BY_PASS_NUM_REGULAR_PRIOS; i++) {
+		prio = fs_create_prio(ns, i, 1);
+		if (IS_ERR(prio))
+			return PTR_ERR(prio);
+	}
+	return 0;
+}
+
 static int init_fdb_root_ns(struct mlx5_flow_steering *steering)
 {
 	struct fs_prio *maj_prio;
@@ -2831,12 +2877,10 @@ static int init_fdb_root_ns(struct mlx5_flow_steering *steering)
 	if (!steering->fdb_root_ns)
 		return -ENOMEM;
 
-	maj_prio = fs_create_prio(&steering->fdb_root_ns->ns, FDB_BYPASS_PATH,
-				  1);
-	if (IS_ERR(maj_prio)) {
-		err = PTR_ERR(maj_prio);
+	err = create_fdb_bypass(steering);
+	if (err)
 		goto out_err;
-	}
+
 	err = create_fdb_fast_path(steering);
 	if (err)
 		goto out_err;
diff --git a/include/linux/mlx5/fs.h b/include/linux/mlx5/fs.h
index cd2d4c572367..b1aad14689e3 100644
--- a/include/linux/mlx5/fs.h
+++ b/include/linux/mlx5/fs.h
@@ -73,6 +73,7 @@ enum mlx5_flow_namespace_type {
 	MLX5_FLOW_NAMESPACE_KERNEL,
 	MLX5_FLOW_NAMESPACE_LEFTOVERS,
 	MLX5_FLOW_NAMESPACE_ANCHOR,
+	MLX5_FLOW_NAMESPACE_FDB_BYPASS,
 	MLX5_FLOW_NAMESPACE_FDB,
 	MLX5_FLOW_NAMESPACE_ESW_EGRESS,
 	MLX5_FLOW_NAMESPACE_ESW_INGRESS,
diff --git a/include/linux/mlx5/mlx5_ifc.h b/include/linux/mlx5/mlx5_ifc.h
index fbaab440a484..e9db12aae8f9 100644
--- a/include/linux/mlx5/mlx5_ifc.h
+++ b/include/linux/mlx5/mlx5_ifc.h
@@ -1621,7 +1621,7 @@ struct mlx5_ifc_cmd_hca_cap_bits {
 
 	u8         ext_stride_num_range[0x1];
 	u8         roce_rw_supported[0x1];
-	u8         reserved_at_3a2[0x1];
+	u8         log_max_current_uc_list_wr_supported[0x1];
 	u8         log_max_stride_sz_rq[0x5];
 	u8         reserved_at_3a8[0x3];
 	u8         log_min_stride_sz_rq[0x5];