Skip to contentSkip to conversation

E-Switch vport sharing & delegation

PATCH Applied Saeed Mahameed 3 participants 10 messages Aug 29, 2025 – Sep 2, 2025

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

v3 cover E-Switch vport sharing & delegation

From: Saeed Mahameed <[email protected]>

v2->v3:
- fix error handling, Simon.
- Remove redundant struct field comment, Simon.
- Add Reviewed-by: Simon Horman <[email protected]>

v1->v2:
- rename goto labels after targets. Jakub.

An mlx5 E-Switch FDB table can manage vports belonging to other sibling physical functions, such as ECPF (ARM embedded cores) and Host PF (x86). This enables a single source of truth for SDN software to manage network pipelines from one host. While such functionality already exists in mlx5, it is currently limited by static vport allocation,
meaning the number of vports shared between multi-host functions must be known pre-boot.

This patchset enables delegated/external vports to be discovered dynamically when switchdev mode is enabled, leveraging new firmware capabilities for dynamic vport creation.

Adjacent functions that delegate their SR-IOV VFs to sibling PFs, can be dynamically discovered on the sibling PF's switchdev mode enabling, after sriov was enabled on the originating PF, allowing for more flexible and scalable management in multi-host and ECPF-to-host scenarios.

The patchset consists of the following changes:

- Refactoring of ACL root namespace handling: The storage of vport ACL root namespaces is converted from a linear array to an xarray, allowing dynamic creation of ACLs per individual vport.
- Improvements for vhca_id to vport mapping.
- Dynamic querying and creation of delegated functions/vports.

Adithya Jayachandran (2):
net/mlx5: E-Switch, Add support for adjacent functions vports
discovery
net/mlx5: E-switch, Set representor attributes for adjacent VFs

Saeed Mahameed (5):
net/mlx5: FS, Convert vport acls root namespaces to xarray net/mlx5: E-Switch, Move vport acls root namespaces creation to
eswitch
net/mlx5: E-Switch, Create acls root namespace for adjacent vports net/mlx5: E-Switch, Register representors for adjacent vports net/mlx5: {DR,HWS}, Use the cached vhca_id for this device

 .../net/ethernet/mellanox/mlx5/core/Makefile  |   2 +-
 .../mellanox/mlx5/core/esw/adj_vport.c        | 209 ++++++++++++++++++
 .../mellanox/mlx5/core/esw/devlink_port.c     |  11 +-
 .../net/ethernet/mellanox/mlx5/core/eswitch.c | 131 ++++++++++-
 .../net/ethernet/mellanox/mlx5/core/eswitch.h |  17 ++
 .../mellanox/mlx5/core/eswitch_offloads.c     |  37 +++-
 .../net/ethernet/mellanox/mlx5/core/fs_core.c | 183 ++++++++-------
 .../net/ethernet/mellanox/mlx5/core/fs_core.h |  18 +-
 .../mellanox/mlx5/core/steering/hws/cmd.c     |  34 +--
 .../mellanox/mlx5/core/steering/sws/dr_cmd.c  |  34 +--
 10 files changed, 500 insertions(+), 176 deletions(-)
 create mode 100644 drivers/net/ethernet/mellanox/mlx5/core/esw/adj_vport.c
Signature · 1 line
2.50.1
#1 lore raw

PATCH v3 1/7 net/mlx5: FS, Convert vport acls root namespaces to xarray

From: Saeed Mahameed <[email protected]>

Before this patch it was a linear array and could only support a certain number of vports, in the next patches, vport numbers are not bound to a well known limit, thus convert acl root name space storage to xarray.

In addition create fs_core public API to add/remove vport acl namespaces as it is the eswitch responsibility to create the vports and their root name spaces for acls, in the next patch we will move mlx5_fs_ingress_acls_{init,cleanup} to eswitch and will use the individual mlx5_fs_vport_{egress,ingresS}_acl_ns_{add,remove} APIs for dynamically create vports.

---
 .../net/ethernet/mellanox/mlx5/core/fs_core.c | 169 ++++++++++++------
 .../net/ethernet/mellanox/mlx5/core/fs_core.h |  13 +-
 2 files changed, 123 insertions(+), 59 deletions(-)
drivers/net/ethernet/mellanox/mlx5/core/fs_core.c +112 -57
index cb165085a4c1..386acf248970 100644
@@ -2793,30 +2793,32 @@ struct mlx5_flow_namespace *mlx5_get_flow_namespace(struct mlx5_core_dev *dev,
 }
 EXPORT_SYMBOL(mlx5_get_flow_namespace);
 
+struct mlx5_vport_acl_root_ns {
+	u16 vport_idx;
+	struct mlx5_flow_root_namespace *root_ns;
+};
+
 struct mlx5_flow_namespace *
 mlx5_get_flow_vport_namespace(struct mlx5_core_dev *dev,
 			      enum mlx5_flow_namespace_type type, int vport_idx)
 {
 	struct mlx5_flow_steering *steering = dev->priv.steering;
+	struct mlx5_vport_acl_root_ns *vport_ns;
 
 	if (!steering)
 		return NULL;
 
 	switch (type) {
 	case MLX5_FLOW_NAMESPACE_ESW_EGRESS:
-		if (vport_idx >= steering->esw_egress_acl_vports)
-			return NULL;
-		if (steering->esw_egress_root_ns &&
-		    steering->esw_egress_root_ns[vport_idx])
-			return &steering->esw_egress_root_ns[vport_idx]->ns;
+		vport_ns = xa_load(&steering->esw_egress_root_ns, vport_idx);
+		if (vport_ns)
+			return &vport_ns->root_ns->ns;
 		else
 			return NULL;
 	case MLX5_FLOW_NAMESPACE_ESW_INGRESS:
-		if (vport_idx >= steering->esw_ingress_acl_vports)
-			return NULL;
-		if (steering->esw_ingress_root_ns &&
-		    steering->esw_ingress_root_ns[vport_idx])
-			return &steering->esw_ingress_root_ns[vport_idx]->ns;
+		vport_ns = xa_load(&steering->esw_ingress_root_ns, vport_idx);
+		if (vport_ns)
+			return &vport_ns->root_ns->ns;
 		else
 			return NULL;
 	case MLX5_FLOW_NAMESPACE_RDMA_TRANSPORT_RX:
@@ -3575,30 +3577,102 @@ static int init_fdb_root_ns(struct mlx5_flow_steering *steering)
 	return err;
 }
 
-static int init_egress_acl_root_ns(struct mlx5_flow_steering *steering, int vport)
+static void
+mlx5_fs_remove_vport_acl_root_ns(struct xarray *esw_acl_root_ns, u16 vport_idx)
+{
+	struct mlx5_vport_acl_root_ns *vport_ns;
+
+	vport_ns = xa_erase(esw_acl_root_ns, vport_idx);
+	if (vport_ns) {
+		cleanup_root_ns(vport_ns->root_ns);
+		kfree(vport_ns);
+	}
+}
+
+static int
+mlx5_fs_add_vport_acl_root_ns(struct mlx5_flow_steering *steering,
+			      struct xarray *esw_acl_root_ns,
+			      enum fs_flow_table_type table_type,
+			      u16 vport_idx)
 {
+	struct mlx5_vport_acl_root_ns *vport_ns;
 	struct fs_prio *prio;
+	int err;
+
+	/* sanity check, intended xarrays are used */
+	if (WARN_ON(esw_acl_root_ns != &steering->esw_egress_root_ns &&
+		    esw_acl_root_ns != &steering->esw_ingress_root_ns))
+		return -EINVAL;
 
-	steering->esw_egress_root_ns[vport] = create_root_ns(steering, FS_FT_ESW_EGRESS_ACL);
-	if (!steering->esw_egress_root_ns[vport])
+	if (table_type != FS_FT_ESW_EGRESS_ACL &&
+	    table_type != FS_FT_ESW_INGRESS_ACL) {
+		mlx5_core_err(steering->dev,
+			      "Invalid table type %d for egress/ingress ACLs\n",
+			      table_type);
+		return -EINVAL;
+	}
+
+	if (xa_load(esw_acl_root_ns, vport_idx))
+		return -EEXIST;
+
+	vport_ns = kzalloc(sizeof(*vport_ns), GFP_KERNEL);
+	if (!vport_ns)
 		return -ENOMEM;
 
+	vport_ns->root_ns = create_root_ns(steering, table_type);
+	if (!vport_ns->root_ns) {
+		err = -ENOMEM;
+		goto kfree_vport_ns;
+	}
+
 	/* create 1 prio*/
-	prio = fs_create_prio(&steering->esw_egress_root_ns[vport]->ns, 0, 1);
-	return PTR_ERR_OR_ZERO(prio);
+	prio = fs_create_prio(&vport_ns->root_ns->ns, 0, 1);
+	if (IS_ERR(prio)) {
+		err = PTR_ERR(prio);
+		goto cleanup_root_ns;
+	}
+
+	vport_ns->vport_idx = vport_idx;
+	err = xa_insert(esw_acl_root_ns, vport_idx, vport_ns, GFP_KERNEL);
+	if (err)
+		goto cleanup_root_ns;
+	return 0;
+
+cleanup_root_ns:
+	cleanup_root_ns(vport_ns->root_ns);
+kfree_vport_ns:
+	kfree(vport_ns);
+	return err;
 }
 
-static int init_ingress_acl_root_ns(struct mlx5_flow_steering *steering, int vport)
+int mlx5_fs_vport_egress_acl_ns_add(struct mlx5_flow_steering *steering,
+				    u16 vport_idx)
 {
-	struct fs_prio *prio;
+	return mlx5_fs_add_vport_acl_root_ns(steering,
+					     &steering->esw_egress_root_ns,
+					     FS_FT_ESW_EGRESS_ACL, vport_idx);
+}
 
-	steering->esw_ingress_root_ns[vport] = create_root_ns(steering, FS_FT_ESW_INGRESS_ACL);
-	if (!steering->esw_ingress_root_ns[vport])
-		return -ENOMEM;
+int mlx5_fs_vport_ingress_acl_ns_add(struct mlx5_flow_steering *steering,
+				     u16 vport_idx)
+{
+	return mlx5_fs_add_vport_acl_root_ns(steering,
+					     &steering->esw_ingress_root_ns,
+					     FS_FT_ESW_INGRESS_ACL, vport_idx);
+}
 
-	/* create 1 prio*/
-	prio = fs_create_prio(&steering->esw_ingress_root_ns[vport]->ns, 0, 1);
-	return PTR_ERR_OR_ZERO(prio);
+void mlx5_fs_vport_egress_acl_ns_remove(struct mlx5_flow_steering *steering,
+					int vport_idx)
+{
+	mlx5_fs_remove_vport_acl_root_ns(&steering->esw_egress_root_ns,
+					 vport_idx);
+}
+
+void mlx5_fs_vport_ingress_acl_ns_remove(struct mlx5_flow_steering *steering,
+					 int vport_idx)
+{
+	mlx5_fs_remove_vport_acl_root_ns(&steering->esw_ingress_root_ns,
+					 vport_idx);
 }
 
 int mlx5_fs_egress_acls_init(struct mlx5_core_dev *dev, int total_vports)
@@ -3607,15 +3681,10 @@ int mlx5_fs_egress_acls_init(struct mlx5_core_dev *dev, int total_vports)
 	int err;
 	int i;
 
-	steering->esw_egress_root_ns =
-			kcalloc(total_vports,
-				sizeof(*steering->esw_egress_root_ns),
-				GFP_KERNEL);
-	if (!steering->esw_egress_root_ns)
-		return -ENOMEM;
+	xa_init(&steering->esw_egress_root_ns);
 
 	for (i = 0; i < total_vports; i++) {
-		err = init_egress_acl_root_ns(steering, i);
+		err = mlx5_fs_vport_egress_acl_ns_add(steering, i);
 		if (err)
 			goto cleanup_root_ns;
 	}
@@ -3623,10 +3692,9 @@ int mlx5_fs_egress_acls_init(struct mlx5_core_dev *dev, int total_vports)
 	return 0;
 
 cleanup_root_ns:
-	for (i--; i >= 0; i--)
-		cleanup_root_ns(steering->esw_egress_root_ns[i]);
-	kfree(steering->esw_egress_root_ns);
-	steering->esw_egress_root_ns = NULL;
+	while (i--)
+		mlx5_fs_vport_egress_acl_ns_remove(steering, i);
+	xa_destroy(&steering->esw_egress_root_ns);
 	return err;
 }
 
@@ -3635,14 +3703,10 @@ void mlx5_fs_egress_acls_cleanup(struct mlx5_core_dev *dev)
 	struct mlx5_flow_steering *steering = dev->priv.steering;
 	int i;
 
-	if (!steering->esw_egress_root_ns)
-		return;
-
 	for (i = 0; i < steering->esw_egress_acl_vports; i++)
-		cleanup_root_ns(steering->esw_egress_root_ns[i]);
+		mlx5_fs_vport_egress_acl_ns_remove(steering, i);
 
-	kfree(steering->esw_egress_root_ns);
-	steering->esw_egress_root_ns = NULL;
+	xa_destroy(&steering->esw_egress_root_ns);
 }
 
 int mlx5_fs_ingress_acls_init(struct mlx5_core_dev *dev, int total_vports)
@@ -3651,15 +3715,10 @@ int mlx5_fs_ingress_acls_init(struct mlx5_core_dev *dev, int total_vports)
 	int err;
 	int i;
 
-	steering->esw_ingress_root_ns =
-			kcalloc(total_vports,
-				sizeof(*steering->esw_ingress_root_ns),
-				GFP_KERNEL);
-	if (!steering->esw_ingress_root_ns)
-		return -ENOMEM;
+	xa_init(&steering->esw_ingress_root_ns);
 
 	for (i = 0; i < total_vports; i++) {
-		err = init_ingress_acl_root_ns(steering, i);
+		err = mlx5_fs_vport_ingress_acl_ns_add(steering, i);
 		if (err)
 			goto cleanup_root_ns;
 	}
@@ -3667,10 +3726,10 @@ int mlx5_fs_ingress_acls_init(struct mlx5_core_dev *dev, int total_vports)
 	return 0;
 
 cleanup_root_ns:
-	for (i--; i >= 0; i--)
-		cleanup_root_ns(steering->esw_ingress_root_ns[i]);
-	kfree(steering->esw_ingress_root_ns);
-	steering->esw_ingress_root_ns = NULL;
+	while (i--)
+		mlx5_fs_vport_ingress_acl_ns_remove(steering, i);
+
+	xa_destroy(&steering->esw_ingress_root_ns);
 	return err;
 }
 
@@ -3679,14 +3738,10 @@ void mlx5_fs_ingress_acls_cleanup(struct mlx5_core_dev *dev)
 	struct mlx5_flow_steering *steering = dev->priv.steering;
 	int i;
 
-	if (!steering->esw_ingress_root_ns)
-		return;
-
 	for (i = 0; i < steering->esw_ingress_acl_vports; i++)
-		cleanup_root_ns(steering->esw_ingress_root_ns[i]);
+		mlx5_fs_vport_ingress_acl_ns_remove(steering, i);
 
-	kfree(steering->esw_ingress_root_ns);
-	steering->esw_ingress_root_ns = NULL;
+	xa_destroy(&steering->esw_ingress_root_ns);
 }
 
 u32 mlx5_fs_get_capabilities(struct mlx5_core_dev *dev, enum mlx5_flow_namespace_type type)
drivers/net/ethernet/mellanox/mlx5/core/fs_core.h +11 -2
index 500826229b0b..a7642d9fc118 100644
@@ -151,8 +151,8 @@ struct mlx5_flow_steering {
 	struct mlx5_flow_root_namespace *root_ns;
 	struct mlx5_flow_root_namespace *fdb_root_ns;
 	struct mlx5_flow_namespace	**fdb_sub_ns;
-	struct mlx5_flow_root_namespace **esw_egress_root_ns;
-	struct mlx5_flow_root_namespace **esw_ingress_root_ns;
+	struct xarray			esw_egress_root_ns;
+	struct xarray			esw_ingress_root_ns;
 	struct mlx5_flow_root_namespace	*sniffer_tx_root_ns;
 	struct mlx5_flow_root_namespace	*sniffer_rx_root_ns;
 	struct mlx5_flow_root_namespace	*rdma_rx_root_ns;
@@ -383,6 +383,15 @@ void mlx5_fs_egress_acls_cleanup(struct mlx5_core_dev *dev);
 int mlx5_fs_ingress_acls_init(struct mlx5_core_dev *dev, int total_vports);
 void mlx5_fs_ingress_acls_cleanup(struct mlx5_core_dev *dev);
 
+int mlx5_fs_vport_egress_acl_ns_add(struct mlx5_flow_steering *steering,
+				    u16 vport_idx);
+int mlx5_fs_vport_ingress_acl_ns_add(struct mlx5_flow_steering *steering,
+				     u16 vport_idx);
+void mlx5_fs_vport_egress_acl_ns_remove(struct mlx5_flow_steering *steering,
+					int vport_idx);
+void mlx5_fs_vport_ingress_acl_ns_remove(struct mlx5_flow_steering *steering,
+					 int vport_idx);
+
 u32 mlx5_fs_get_capabilities(struct mlx5_core_dev *dev, enum mlx5_flow_namespace_type type);
 
 struct mlx5_flow_root_namespace *find_root(struct fs_node *node);
Signature · 1 line
2.50.1
#2 lore raw

PATCH v3 2/7 net/mlx5: E-Switch, Move vport acls root namespaces creation to eswitch

In reply to Saeed Mahameed

From: Saeed Mahameed <[email protected]>

Move the loop that creates the vports ACLs root name spaces to eswitch, since it is the eswitch responsibility to decide when and how many vports ACLs root namespaces to create, in the next patches we will use the fs_core vport ACL root namespace APIs to create/remove root ns ACLs dynamically for dynamically created vports.

---
 .../net/ethernet/mellanox/mlx5/core/eswitch.c | 73 ++++++++++++++++--
 .../net/ethernet/mellanox/mlx5/core/fs_core.c | 76 ++-----------------
 .../net/ethernet/mellanox/mlx5/core/fs_core.h |  7 --
 3 files changed, 72 insertions(+), 84 deletions(-)
drivers/net/ethernet/mellanox/mlx5/core/eswitch.c +65 -8
index 9fe5a45124fd..900650a1a66e 100644
@@ -1439,19 +1439,76 @@ static void mlx5_esw_mode_change_notify(struct mlx5_eswitch *esw, u16 mode)
 	blocking_notifier_call_chain(&esw->n_head, 0, &info);
 }
 
+static int mlx5_esw_egress_acls_init(struct mlx5_core_dev *dev)
+{
+	struct mlx5_flow_steering *steering = dev->priv.steering;
+	int total_vports = mlx5_eswitch_get_total_vports(dev);
+	int err;
+	int i;
+
+	for (i = 0; i < total_vports; i++) {
+		err = mlx5_fs_vport_egress_acl_ns_add(steering, i);
+		if (err)
+			goto acl_ns_remove;
+	}
+	return 0;
+
+acl_ns_remove:
+	while (i--)
+		mlx5_fs_vport_egress_acl_ns_remove(steering, i);
+	return err;
+}
+
+static void mlx5_esw_egress_acls_cleanup(struct mlx5_core_dev *dev)
+{
+	struct mlx5_flow_steering *steering = dev->priv.steering;
+	int total_vports = mlx5_eswitch_get_total_vports(dev);
+	int i;
+
+	for (i = total_vports - 1; i >= 0; i--)
+		mlx5_fs_vport_egress_acl_ns_remove(steering, i);
+}
+
+static int mlx5_esw_ingress_acls_init(struct mlx5_core_dev *dev)
+{
+	struct mlx5_flow_steering *steering = dev->priv.steering;
+	int total_vports = mlx5_eswitch_get_total_vports(dev);
+	int err;
+	int i;
+
+	for (i = 0; i < total_vports; i++) {
+		err = mlx5_fs_vport_ingress_acl_ns_add(steering, i);
+		if (err)
+			goto acl_ns_remove;
+	}
+	return 0;
+
+acl_ns_remove:
+	while (i--)
+		mlx5_fs_vport_ingress_acl_ns_remove(steering, i);
+	return err;
+}
+
+static void mlx5_esw_ingress_acls_cleanup(struct mlx5_core_dev *dev)
+{
+	struct mlx5_flow_steering *steering = dev->priv.steering;
+	int total_vports = mlx5_eswitch_get_total_vports(dev);
+	int i;
+
+	for (i = total_vports - 1; i >= 0; i--)
+		mlx5_fs_vport_ingress_acl_ns_remove(steering, i);
+}
+
 static int mlx5_esw_acls_ns_init(struct mlx5_eswitch *esw)
 {
 	struct mlx5_core_dev *dev = esw->dev;
-	int total_vports;
 	int err;
 
 	if (esw->flags & MLX5_ESWITCH_VPORT_ACL_NS_CREATED)
 		return 0;
 
-	total_vports = mlx5_eswitch_get_total_vports(dev);
-
 	if (MLX5_CAP_ESW_EGRESS_ACL(dev, ft_support)) {
-		err = mlx5_fs_egress_acls_init(dev, total_vports);
+		err = mlx5_esw_egress_acls_init(dev);
 		if (err)
 			return err;
 	} else {
@@ -1459,7 +1516,7 @@ static int mlx5_esw_acls_ns_init(struct mlx5_eswitch *esw)
 	}
 
 	if (MLX5_CAP_ESW_INGRESS_ACL(dev, ft_support)) {
-		err = mlx5_fs_ingress_acls_init(dev, total_vports);
+		err = mlx5_esw_ingress_acls_init(dev);
 		if (err)
 			goto err;
 	} else {
@@ -1470,7 +1527,7 @@ static int mlx5_esw_acls_ns_init(struct mlx5_eswitch *esw)
 
 err:
 	if (MLX5_CAP_ESW_EGRESS_ACL(dev, ft_support))
-		mlx5_fs_egress_acls_cleanup(dev);
+		mlx5_esw_egress_acls_cleanup(dev);
 	return err;
 }
 
@@ -1480,9 +1537,9 @@ static void mlx5_esw_acls_ns_cleanup(struct mlx5_eswitch *esw)
 
 	esw->flags &= ~MLX5_ESWITCH_VPORT_ACL_NS_CREATED;
 	if (MLX5_CAP_ESW_INGRESS_ACL(dev, ft_support))
-		mlx5_fs_ingress_acls_cleanup(dev);
+		mlx5_esw_ingress_acls_cleanup(dev);
 	if (MLX5_CAP_ESW_EGRESS_ACL(dev, ft_support))
-		mlx5_fs_egress_acls_cleanup(dev);
+		mlx5_esw_egress_acls_cleanup(dev);
 }
 
 /**
drivers/net/ethernet/mellanox/mlx5/core/fs_core.c +7 -69

94 lines not on this page · raw message

drivers/net/ethernet/mellanox/mlx5/core/fs_core.h +0 -7
index a7642d9fc118..7877d9a2118d 100644
@@ -159,8 +159,6 @@ struct mlx5_flow_steering {
 	struct mlx5_flow_root_namespace	*rdma_tx_root_ns;
 	struct mlx5_flow_root_namespace	*egress_root_ns;
 	struct mlx5_flow_root_namespace	*port_sel_root_ns;
-	int esw_egress_acl_vports;
-	int esw_ingress_acl_vports;
 	struct mlx5_flow_root_namespace **rdma_transport_rx_root_ns;
 	struct mlx5_flow_root_namespace **rdma_transport_tx_root_ns;
 	int rdma_transport_rx_vports;
@@ -378,11 +376,6 @@ void mlx5_fs_core_free(struct mlx5_core_dev *dev);
 int mlx5_fs_core_init(struct mlx5_core_dev *dev);
 void mlx5_fs_core_cleanup(struct mlx5_core_dev *dev);
 
-int mlx5_fs_egress_acls_init(struct mlx5_core_dev *dev, int total_vports);
-void mlx5_fs_egress_acls_cleanup(struct mlx5_core_dev *dev);
-int mlx5_fs_ingress_acls_init(struct mlx5_core_dev *dev, int total_vports);
-void mlx5_fs_ingress_acls_cleanup(struct mlx5_core_dev *dev);
-
 int mlx5_fs_vport_egress_acl_ns_add(struct mlx5_flow_steering *steering,
 				    u16 vport_idx);
 int mlx5_fs_vport_ingress_acl_ns_add(struct mlx5_flow_steering *steering,
Signature · 1 line
2.50.1
#3 lore raw

PATCH v3 3/7 net/mlx5: E-Switch, Add support for adjacent functions vports discovery

In reply to Saeed Mahameed

From: Adithya Jayachandran <[email protected]>

Adding driver support to query adjacent functions vports, AKA delegated vports.

Adjacent functions can delegate their sriov vfs to other sibling PF in the system, to be managed by the eswitch capable sibling PF. E.g, ECPF to Host PF, multi host PF between each other, etc.

Only supported in switchdev mode.

---
 .../net/ethernet/mellanox/mlx5/core/Makefile  |   2 +-
 .../mellanox/mlx5/core/esw/adj_vport.c        | 185 ++++++++++++++++++
 .../net/ethernet/mellanox/mlx5/core/eswitch.c |  58 +++++-
 .../net/ethernet/mellanox/mlx5/core/eswitch.h |   7 +
 .../mellanox/mlx5/core/eswitch_offloads.c     |   4 +
 5 files changed, 251 insertions(+), 5 deletions(-)
 create mode 100644 drivers/net/ethernet/mellanox/mlx5/core/esw/adj_vport.c
drivers/net/ethernet/mellanox/mlx5/core/Makefile +1 -1
index 206223ce63a8..a65ab661375a 100644
@@ -69,7 +69,7 @@ mlx5_core-$(CONFIG_MLX5_TC_SAMPLE)   += en/tc/sample.o
 # Core extra
 #
 mlx5_core-$(CONFIG_MLX5_ESWITCH)   += eswitch.o eswitch_offloads.o eswitch_offloads_termtbl.o \
-				      ecpf.o rdma.o esw/legacy.o \
+				      ecpf.o rdma.o esw/legacy.o esw/adj_vport.o \
 				      esw/devlink_port.o esw/vporttbl.o esw/qos.o esw/ipsec.o
 
 mlx5_core-$(CONFIG_MLX5_ESWITCH)   += esw/acl/helper.o \
drivers/net/ethernet/mellanox/mlx5/core/esw/adj_vport.c new file +185 -0

185 lines not on this page · raw message

drivers/net/ethernet/mellanox/mlx5/core/eswitch.c +54 -4

103 lines not on this page · raw message

drivers/net/ethernet/mellanox/mlx5/core/eswitch.h +7 -0
index cfd6b1b8c6f4..2c0e5ca73f3d 100644
@@ -216,6 +216,7 @@ struct mlx5_vport {
 	u32                     metadata;
 	int                     vhca_id;
 
+	bool adjacent; /* delegated vhca from adjacent function */
 	struct mlx5_vport_info  info;
 
 	/* Protected with the E-Switch qos domain lock. The Vport QoS can
@@ -384,6 +385,7 @@ struct mlx5_eswitch {
 
 	struct mlx5_esw_bridge_offloads *br_offloads;
 	struct mlx5_esw_offload offloads;
+	u32 last_vport_idx;
 	int                     mode;
 	u16                     manager_vport;
 	u16                     first_host_vport;
@@ -417,6 +419,8 @@ int mlx5_esw_qos_modify_vport_rate(struct mlx5_eswitch *esw, u16 vport_num, u32
 /* E-Switch API */
 int mlx5_eswitch_init(struct mlx5_core_dev *dev);
 void mlx5_eswitch_cleanup(struct mlx5_eswitch *esw);
+int mlx5_esw_vport_alloc(struct mlx5_eswitch *esw, int index, u16 vport_num);
+void mlx5_esw_vport_free(struct mlx5_eswitch *esw, struct mlx5_vport *vport);
 
 #define MLX5_ESWITCH_IGNORE_NUM_VFS (-1)
 int mlx5_eswitch_enable_locked(struct mlx5_eswitch *esw, int num_vfs);
@@ -622,6 +626,9 @@ bool mlx5_esw_multipath_prereq(struct mlx5_core_dev *dev0,
 
 const u32 *mlx5_esw_query_functions(struct mlx5_core_dev *dev);
 
+void mlx5_esw_adjacent_vhcas_setup(struct mlx5_eswitch *esw);
+void mlx5_esw_adjacent_vhcas_cleanup(struct mlx5_eswitch *esw);
+
 #define MLX5_DEBUG_ESWITCH_MASK BIT(3)
 
 #define esw_info(__dev, format, ...)			\
drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c +4 -0
index cdba7bc448ee..fb03981d5036 100644
@@ -3538,6 +3538,8 @@ int esw_offloads_enable(struct mlx5_eswitch *esw)
 	int err;
 
 	mutex_init(&esw->offloads.termtbl_mutex);
+	mlx5_esw_adjacent_vhcas_setup(esw);
+
 	err = mlx5_rdma_enable_roce(esw->dev);
 	if (err)
 		goto err_roce;
@@ -3602,6 +3604,7 @@ int esw_offloads_enable(struct mlx5_eswitch *esw)
 err_metadata:
 	mlx5_rdma_disable_roce(esw->dev);
 err_roce:
+	mlx5_esw_adjacent_vhcas_cleanup(esw);
 	mutex_destroy(&esw->offloads.termtbl_mutex);
 	return err;
 }
@@ -3635,6 +3638,7 @@ void esw_offloads_disable(struct mlx5_eswitch *esw)
 	mapping_destroy(esw->offloads.reg_c0_obj_pool);
 	esw_offloads_metadata_uninit(esw);
 	mlx5_rdma_disable_roce(esw->dev);
+	mlx5_esw_adjacent_vhcas_cleanup(esw);
 	mutex_destroy(&esw->offloads.termtbl_mutex);
 }
 
Signature · 1 line
2.50.1
#5 lore raw

PATCH v3 4/7 net/mlx5: E-Switch, Create acls root namespace for adjacent vports

In reply to Saeed Mahameed

From: Saeed Mahameed <[email protected]>

Use the new vport acl root namespace add/remove API to create the missing acl root name spaces per each new adjacent function vport.

---
 drivers/net/ethernet/mellanox/mlx5/core/esw/adj_vport.c | 8 ++++++++
 1 file changed, 8 insertions(+)
drivers/net/ethernet/mellanox/mlx5/core/esw/adj_vport.c +8 -0
index 37a06c0949d5..1d104b3fe9e0 100644
@@ -1,6 +1,7 @@
 // SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB
 // Copyright (c) 2025, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
 
+#include "fs_core.h"
 #include "eswitch.h"
 
 enum {
@@ -82,6 +83,9 @@ static int mlx5_esw_adj_vport_create(struct mlx5_eswitch *esw, u16 vhca_id)
 	vport->adjacent = true;
 	vport->vhca_id = vhca_id;
 
+	mlx5_fs_vport_egress_acl_ns_add(esw->dev->priv.steering, vport->index);
+	mlx5_fs_vport_ingress_acl_ns_add(esw->dev->priv.steering, vport->index);
+
 	mlx5_esw_adj_vport_modify(esw->dev, vport_num, MLX5_ADJ_VPORT_CONNECT);
 	return 0;
 
@@ -99,6 +103,10 @@ static void mlx5_esw_adj_vport_destroy(struct mlx5_eswitch *esw,
 		  vport_num, vport->vhca_id);
 	mlx5_esw_adj_vport_modify(esw->dev, vport_num,
 				  MLX5_ADJ_VPORT_DISCONNECT);
+	mlx5_fs_vport_egress_acl_ns_remove(esw->dev->priv.steering,
+					   vport->index);
+	mlx5_fs_vport_ingress_acl_ns_remove(esw->dev->priv.steering,
+					    vport->index);
 	mlx5_esw_vport_free(esw, vport);
 	/* Reset the vport index back so new adj vports can use this index.
 	 * When vport count can incrementally change, this needs to be modified.
Signature · 1 line
2.50.1
#6 lore raw

PATCH v3 5/7 net/mlx5: E-Switch, Register representors for adjacent vports

In reply to Saeed Mahameed

From: Saeed Mahameed <[email protected]>

Register representors for adjacent vports dynamically when they are discovered. Dynamically added representors state will now be set to 'REGISTERED' when the representor type was already registered, otherwise they won't be loaded.

---
 .../mellanox/mlx5/core/esw/adj_vport.c        | 10 ++++++
 .../net/ethernet/mellanox/mlx5/core/eswitch.h |  5 +++
 .../mellanox/mlx5/core/eswitch_offloads.c     | 33 ++++++++++++++++---
 3 files changed, 43 insertions(+), 5 deletions(-)
drivers/net/ethernet/mellanox/mlx5/core/esw/adj_vport.c +10 -0
index 1d104b3fe9e0..3380f85678bc 100644
@@ -85,10 +85,19 @@ static int mlx5_esw_adj_vport_create(struct mlx5_eswitch *esw, u16 vhca_id)
 
 	mlx5_fs_vport_egress_acl_ns_add(esw->dev->priv.steering, vport->index);
 	mlx5_fs_vport_ingress_acl_ns_add(esw->dev->priv.steering, vport->index);
+	err = mlx5_esw_offloads_rep_add(esw, vport);
+	if (err)
+		goto acl_ns_remove;
 
 	mlx5_esw_adj_vport_modify(esw->dev, vport_num, MLX5_ADJ_VPORT_CONNECT);
 	return 0;
 
+acl_ns_remove:
+	mlx5_fs_vport_ingress_acl_ns_remove(esw->dev->priv.steering,
+					    vport->index);
+	mlx5_fs_vport_egress_acl_ns_remove(esw->dev->priv.steering,
+					   vport->index);
+	mlx5_esw_vport_free(esw, vport);
 destroy_esw_vport:
 	mlx5_esw_destroy_esw_vport(esw->dev, vport_num);
 	return err;
@@ -103,6 +112,7 @@ static void mlx5_esw_adj_vport_destroy(struct mlx5_eswitch *esw,
 		  vport_num, vport->vhca_id);
 	mlx5_esw_adj_vport_modify(esw->dev, vport_num,
 				  MLX5_ADJ_VPORT_DISCONNECT);
+	mlx5_esw_offloads_rep_remove(esw, vport);
 	mlx5_fs_vport_egress_acl_ns_remove(esw->dev->priv.steering,
 					   vport->index);
 	mlx5_fs_vport_ingress_acl_ns_remove(esw->dev->priv.steering,
drivers/net/ethernet/mellanox/mlx5/core/eswitch.h +5 -0
index 2c0e5ca73f3d..6d36d8bbb979 100644
@@ -838,6 +838,11 @@ void mlx5_esw_vport_vhca_id_unmap(struct mlx5_eswitch *esw,
 int mlx5_eswitch_vhca_id_to_vport(struct mlx5_eswitch *esw, u16 vhca_id, u16 *vport_num);
 bool mlx5_esw_vport_vhca_id(struct mlx5_eswitch *esw, u16 vportn, u16 *vhca_id);
 
+void mlx5_esw_offloads_rep_remove(struct mlx5_eswitch *esw,
+				  const struct mlx5_vport *vport);
+int mlx5_esw_offloads_rep_add(struct mlx5_eswitch *esw,
+			      const struct mlx5_vport *vport);
+
 /**
  * struct mlx5_esw_event_info - Indicates eswitch mode changed/changing.
  *
drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c +28 -5
index fb03981d5036..d57f86d297ab 100644
@@ -2378,7 +2378,20 @@ static int esw_offloads_start(struct mlx5_eswitch *esw,
 	return 0;
 }
 
-static int mlx5_esw_offloads_rep_init(struct mlx5_eswitch *esw, const struct mlx5_vport *vport)
+void mlx5_esw_offloads_rep_remove(struct mlx5_eswitch *esw,
+				  const struct mlx5_vport *vport)
+{
+	struct mlx5_eswitch_rep *rep = xa_load(&esw->offloads.vport_reps,
+					       vport->vport);
+
+	if (!rep)
+		return;
+	xa_erase(&esw->offloads.vport_reps, vport->vport);
+	kfree(rep);
+}
+
+int mlx5_esw_offloads_rep_add(struct mlx5_eswitch *esw,
+			      const struct mlx5_vport *vport)
 {
 	struct mlx5_eswitch_rep *rep;
 	int rep_type;
@@ -2390,9 +2403,19 @@ static int mlx5_esw_offloads_rep_init(struct mlx5_eswitch *esw, const struct mlx
 
 	rep->vport = vport->vport;
 	rep->vport_index = vport->index;
-	for (rep_type = 0; rep_type < NUM_REP_TYPES; rep_type++)
-		atomic_set(&rep->rep_data[rep_type].state, REP_UNREGISTERED);
-
+	for (rep_type = 0; rep_type < NUM_REP_TYPES; rep_type++) {
+		if (!esw->offloads.rep_ops[rep_type]) {
+			atomic_set(&rep->rep_data[rep_type].state,
+				   REP_UNREGISTERED);
+			continue;
+		}
+		/* Dynamic/delegated vports add their representors after
+		 * mlx5_eswitch_register_vport_reps, so mark them as registered
+		 * for them to be loaded later with the others.
+		 */
+		rep->esw = esw;
+		atomic_set(&rep->rep_data[rep_type].state, REP_REGISTERED);
+	}
 	err = xa_insert(&esw->offloads.vport_reps, rep->vport, rep, GFP_KERNEL);
 	if (err)
 		goto insert_err;
@@ -2430,7 +2453,7 @@ static int esw_offloads_init_reps(struct mlx5_eswitch *esw)
 	xa_init(&esw->offloads.vport_reps);
 
 	mlx5_esw_for_each_vport(esw, i, vport) {
-		err = mlx5_esw_offloads_rep_init(esw, vport);
+		err = mlx5_esw_offloads_rep_add(esw, vport);
 		if (err)
 			goto err;
 	}
Signature · 1 line
2.50.1
#7 lore raw

PATCH v3 6/7 net/mlx5: E-switch, Set representor attributes for adjacent VFs

In reply to Saeed Mahameed

From: Adithya Jayachandran <[email protected]>

Adjacent vfs get their devlink port information from firmware, use the information (pfnum, function id) from FW when populating the devlink port attributes.

Before:

$ devlink port show
pci/0000:00:03.0/180225: type eth netdev eth0 flavour pcivf controller 0 pfnum 0 vfnum 49152 external false splittable false
  function:
    hw_addr 00:00:00:00:00:00

After:

$ devlink port show
pci/0000:00:03.0/180225: type eth netdev enp0s3npf0vf2 flavour pcivf controller 0 pfnum 0 vfnum 2 external false splittable false
  function:
    hw_addr 00:00:00:00:00:00
---
 .../ethernet/mellanox/mlx5/core/esw/adj_vport.c  | 16 +++++++++++-----
 .../mellanox/mlx5/core/esw/devlink_port.c        | 11 ++++++++++-
 .../net/ethernet/mellanox/mlx5/core/eswitch.h    |  5 +++++
 3 files changed, 26 insertions(+), 6 deletions(-)
drivers/net/ethernet/mellanox/mlx5/core/esw/adj_vport.c +11 -5
index 3380f85678bc..0091ba697bae 100644
@@ -57,7 +57,8 @@ static int mlx5_esw_create_esw_vport(struct mlx5_core_dev *dev, u16 vhca_id,
 	return err;
 }
 
-static int mlx5_esw_adj_vport_create(struct mlx5_eswitch *esw, u16 vhca_id)
+static int mlx5_esw_adj_vport_create(struct mlx5_eswitch *esw, u16 vhca_id,
+				     const void *rid_info_reg)
 {
 	struct mlx5_vport *vport;
 	u16 vport_num;
@@ -83,6 +84,12 @@ static int mlx5_esw_adj_vport_create(struct mlx5_eswitch *esw, u16 vhca_id)
 	vport->adjacent = true;
 	vport->vhca_id = vhca_id;
 
+	vport->adj_info.parent_pci_devfn =
+		MLX5_GET(function_vhca_rid_info_reg, rid_info_reg,
+			 parent_pci_device_function);
+	vport->adj_info.function_id =
+		MLX5_GET(function_vhca_rid_info_reg, rid_info_reg, function_id);
+
 	mlx5_fs_vport_egress_acl_ns_add(esw->dev->priv.steering, vport->index);
 	mlx5_fs_vport_ingress_acl_ns_add(esw->dev->priv.steering, vport->index);
 	err = mlx5_esw_offloads_rep_add(esw, vport);
@@ -176,7 +183,7 @@ void mlx5_esw_adjacent_vhcas_setup(struct mlx5_eswitch *esw)
 	esw_debug(esw->dev, "Delegated vhca functions count %d\n", count);
 
 	for (i = 0; i < count; i++) {
-		void *rid_info, *rid_info_reg;
+		const void *rid_info, *rid_info_reg;
 		u16 vhca_id;
 
 		rid_info = MLX5_ADDR_OF(query_delegated_vhca_out, out,
@@ -187,10 +194,9 @@ void mlx5_esw_adjacent_vhcas_setup(struct mlx5_eswitch *esw)
 
 		vhca_id = MLX5_GET(function_vhca_rid_info_reg, rid_info_reg,
 				   vhca_id);
-		esw_debug(esw->dev, "Delegating vhca_id 0x%x rid info:\n",
-			  vhca_id);
+		esw_debug(esw->dev, "Delegating vhca_id 0x%x\n", vhca_id);
 
-		err = mlx5_esw_adj_vport_create(esw, vhca_id);
+		err = mlx5_esw_adj_vport_create(esw, vhca_id, rid_info_reg);
 		if (err) {
 			esw_warn(esw->dev,
 				 "Failed to init adjacent vhca 0x%x, err %d\n",
drivers/net/ethernet/mellanox/mlx5/core/esw/devlink_port.c +10 -1
index c33accadae0f..cf88a106d80d 100644
@@ -27,6 +27,7 @@ static void mlx5_esw_offloads_pf_vf_devlink_port_attrs_set(struct mlx5_eswitch *
 {
 	struct mlx5_core_dev *dev = esw->dev;
 	struct netdev_phys_item_id ppid = {};
+	struct mlx5_vport *vport;
 	u32 controller_num = 0;
 	bool external;
 	u16 pfnum;
@@ -42,10 +43,18 @@ static void mlx5_esw_offloads_pf_vf_devlink_port_attrs_set(struct mlx5_eswitch *
 		dl_port->attrs.switch_id.id_len = ppid.id_len;
 		devlink_port_attrs_pci_pf_set(dl_port, controller_num, pfnum, external);
 	} else if (mlx5_eswitch_is_vf_vport(esw, vport_num)) {
+		u16 func_id = vport_num - 1;
+
+		vport = mlx5_eswitch_get_vport(esw, vport_num);
 		memcpy(dl_port->attrs.switch_id.id, ppid.id, ppid.id_len);
 		dl_port->attrs.switch_id.id_len = ppid.id_len;
+		if (vport->adjacent) {
+			func_id = vport->adj_info.function_id;
+			pfnum = vport->adj_info.parent_pci_devfn;
+		}
+
 		devlink_port_attrs_pci_vf_set(dl_port, controller_num, pfnum,
-					      vport_num - 1, external);
+					      func_id, external);
 	}  else if (mlx5_core_is_ec_vf_vport(esw->dev, vport_num)) {
 		u16 base_vport = mlx5_core_ec_vf_vport_base(dev);
 
drivers/net/ethernet/mellanox/mlx5/core/eswitch.h +5 -0
index 6d36d8bbb979..4fe285ce32aa 100644
@@ -217,6 +217,11 @@ struct mlx5_vport {
 	int                     vhca_id;
 
 	bool adjacent; /* delegated vhca from adjacent function */
+	struct {
+		u16 parent_pci_devfn; /* Adjacent parent PCI device function */
+		u16 function_id; /* Function ID of the delegated VPort */
+	} adj_info;
+
 	struct mlx5_vport_info  info;
 
 	/* Protected with the E-Switch qos domain lock. The Vport QoS can
Signature · 1 line
2.50.1
#8 lore raw

PATCH v3 7/7 net/mlx5: {DR,HWS}, Use the cached vhca_id for this device

In reply to Saeed Mahameed

From: Saeed Mahameed <[email protected]>

The mlx5 driver caches many capabilities to be used by mlx5 layers.

In SW and HW steering we can use the cached vhca_id instead of invoking FW commands.

---
 .../mellanox/mlx5/core/steering/hws/cmd.c     | 34 ++++---------------
 .../mellanox/mlx5/core/steering/sws/dr_cmd.c  | 34 ++++---------------
 2 files changed, 14 insertions(+), 54 deletions(-)
drivers/net/ethernet/mellanox/mlx5/core/steering/hws/cmd.c +7 -27
index acb0317f930b..f22eaf506d28 100644
@@ -1200,40 +1200,20 @@ int mlx5hws_cmd_query_caps(struct mlx5_core_dev *mdev,
 int mlx5hws_cmd_query_gvmi(struct mlx5_core_dev *mdev, bool other_function,
 			   u16 vport_number, u16 *gvmi)
 {
-	u32 in[MLX5_ST_SZ_DW(query_hca_cap_in)] = {};
-	int out_size;
-	void *out;
 	int err;
 
-	if (other_function) {
-		err = mlx5_vport_get_vhca_id(mdev, vport_number, gvmi);
-		if  (!err)
-			return 0;
-
-		mlx5_core_err(mdev, "Failed to get vport vhca id for vport %d\n",
-			      vport_number);
-		return err;
+	if (!other_function) {
+		/* self vhca_id */
+		*gvmi = MLX5_CAP_GEN(mdev, vhca_id);
+		return 0;
 	}
 
-	/* get vhca_id for `this` function */
-	out_size = MLX5_ST_SZ_BYTES(query_hca_cap_out);
-	out = kzalloc(out_size, GFP_KERNEL);
-	if (!out)
-		return -ENOMEM;
-
-	MLX5_SET(query_hca_cap_in, in, opcode, MLX5_CMD_OP_QUERY_HCA_CAP);
-	MLX5_SET(query_hca_cap_in, in, op_mod,
-		 MLX5_SET_HCA_CAP_OP_MOD_GENERAL_DEVICE << 1 | HCA_CAP_OPMOD_GET_CUR);
-
-	err = mlx5_cmd_exec_inout(mdev, query_hca_cap, in, out);
+	err = mlx5_vport_get_vhca_id(mdev, vport_number, gvmi);
 	if (err) {
-		kfree(out);
+		mlx5_core_err(mdev, "Failed to get vport vhca id for vport %d\n",
+			      vport_number);
 		return err;
 	}
 
-	*gvmi = MLX5_GET(query_hca_cap_out, out, capability.cmd_hca_cap.vhca_id);
-
-	kfree(out);
-
 	return 0;
 }
drivers/net/ethernet/mellanox/mlx5/core/steering/sws/dr_cmd.c +7 -27
index bf99b933fd14..1ebb2b15c080 100644
@@ -35,41 +35,21 @@ int mlx5dr_cmd_query_esw_vport_context(struct mlx5_core_dev *mdev,
 int mlx5dr_cmd_query_gvmi(struct mlx5_core_dev *mdev, bool other_vport,
 			  u16 vport_number, u16 *gvmi)
 {
-	u32 in[MLX5_ST_SZ_DW(query_hca_cap_in)] = {};
-	int out_size;
-	void *out;
 	int err;
 
-	if (other_vport) {
-		err = mlx5_vport_get_vhca_id(mdev, vport_number, gvmi);
-		if  (!err)
-			return 0;
-
-		mlx5_core_err(mdev, "Failed to get vport vhca id for vport %d\n",
-			      vport_number);
-		return err;
+	if (!other_vport) {
+		/* self vhca_id */
+		*gvmi = MLX5_CAP_GEN(mdev, vhca_id);
+		return 0;
 	}
 
-	/* get vhca_id for `this` function */
-	out_size = MLX5_ST_SZ_BYTES(query_hca_cap_out);
-	out = kzalloc(out_size, GFP_KERNEL);
-	if (!out)
-		return -ENOMEM;
-
-	MLX5_SET(query_hca_cap_in, in, opcode, MLX5_CMD_OP_QUERY_HCA_CAP);
-	MLX5_SET(query_hca_cap_in, in, op_mod,
-		 MLX5_SET_HCA_CAP_OP_MOD_GENERAL_DEVICE << 1 |
-		 HCA_CAP_OPMOD_GET_CUR);
-
-	err = mlx5_cmd_exec_inout(mdev, query_hca_cap, in, out);
+	err = mlx5_vport_get_vhca_id(mdev, vport_number, gvmi);
 	if (err) {
-		kfree(out);
+		mlx5_core_err(mdev, "Failed to get vport vhca id for vport %d\n",
+			      vport_number);
 		return err;
 	}
 
-	*gvmi = MLX5_GET(query_hca_cap_out, out, capability.cmd_hca_cap.vhca_id);
-
-	kfree(out);
 	return 0;
 }
 
Signature · 1 line
2.50.1
reviewer Reviewed-by on v3 3/7
#4 lore raw

Re: [PATCH net-next V3 3/7] net/mlx5: E-Switch, Add support for adjacent functions vports discovery

In reply to Saeed Mahameed
Quoted 5 lines · Saeed Mahameed

From: Adithya Jayachandran <[email protected]>

Adding driver support to query adjacent functions vports, AKA delegated vports.

Quoted 8 lines

Adjacent functions can delegate their sriov vfs to other sibling PF in the system, to be managed by the eswitch capable sibling PF. E.g, ECPF to Host PF, multi host PF between each other, etc.

Only supported in switchdev mode.

#9 lore raw
In reply to Saeed Mahameed
This series was applied to netdev/net-next.git (main) by Paolo Abeni <[email protected]>:

Hello:

This series was applied to netdev/net-next.git (main) by Paolo Abeni <[email protected]>:

Quoted 8 lines · On Fri, 29 Aug 2025 15:37:15 -0700 you wrote:

From: Saeed Mahameed <[email protected]>

v2->v3:
- fix error handling, Simon.
- Remove redundant struct field comment, Simon.
- Add Reviewed-by: Simon Horman <[email protected]>

[...]

Here is the summary with links:
- [net-next,V3,1/7] net/mlx5: FS, Convert vport acls root namespaces to xarray

    https://git.kernel.org/netdev/net-next/c/2e894b99c017
  - [net-next,V3,2/7] net/mlx5: E-Switch, Move vport acls root namespaces creation to eswitch
    https://git.kernel.org/netdev/net-next/c/faa6ac53cdaa
  - [net-next,V3,3/7] net/mlx5: E-Switch, Add support for adjacent functions vports discovery
    https://git.kernel.org/netdev/net-next/c/17426c5d4b1d
  - [net-next,V3,4/7] net/mlx5: E-Switch, Create acls root namespace for adjacent vports
    https://git.kernel.org/netdev/net-next/c/9984ec9f1f50
  - [net-next,V3,5/7] net/mlx5: E-Switch, Register representors for adjacent vports
    https://git.kernel.org/netdev/net-next/c/a0a7002b9439
  - [net-next,V3,6/7] net/mlx5: E-switch, Set representor attributes for adjacent VFs
    https://git.kernel.org/netdev/net-next/c/5d8ae2c2cfe8
  - [net-next,V3,7/7] net/mlx5: {DR,HWS}, Use the cached vhca_id for this device
    https://git.kernel.org/netdev/net-next/c/0c2a02f3c066

You are awesome, thank you!

Signature · 2 lines
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html