Linux Ethernet Bridging - Mar 2022 - [Bridge] [PATCH v2 net-next 04/10] net: bridge: mst: Notify switchdev drivers of VLAN MSTI migrations

The bridge has had per-VLAN STP support for a while now, since:

https://lore.kernel.org/netdev/20200124114022.10883-1-nikolay at
cumulusnetworks.com/

The current implementation has some problems:

- The mapping from VLAN to STP state is fixed as 1:1, i.e. each VLAN
  is managed independently. This is awkward from an MSTP (802.1Q-2018,
  Clause 13.5) point of view, where the model is that multiple VLANs
  are grouped into MST instances.

  Because of the way that the standard is written, presumably, this is
  also reflected in hardware implementations. It is not uncommon for a
  switch to support the full 4k range of VIDs, but that the pool of
  MST instances is much smaller. Some examples:

  Marvell LinkStreet (mv88e6xxx): 4k VLANs, but only 64 MSTIs
  Marvell Prestera: 4k VLANs, but only 128 MSTIs
  Microchip SparX-5i: 4k VLANs, but only 128 MSTIs

- By default, the feature is enabled, and there is no way to disable
  it. This makes it hard to add offloading in a backwards compatible
  way, since any underlying switchdevs have no way to refuse the
  function if the hardware does not support it

- The port-global STP state has precedence over per-VLAN states. In
  MSTP, as far as I understand it, all VLANs will use the common
  spanning tree (CST) by default - through traffic engineering you can
  then optimize your network to group subsets of VLANs to use
  different trees (MSTI). To my understanding, the way this is
  typically managed in silicon is roughly:

  Incoming packet:
  .----.----.--------------.----.-------------
  | DA | SA | 802.1Q VID=X | ET | Payload ...
  '----'----'--------------'----'-------------
                        |
                        '->|\     .----------------------------.
                           | +--> | VID | Members | ... | MSTI |
                   PVID -->|/     |-----|---------|-----|------|
                                  |   1 | 0001001 | ... |    0 |
                                  |   2 | 0001010 | ... |   10 |
                                  |   3 | 0001100 | ... |   10 |
                                  '----------------------------'
                                                             |
                               .-----------------------------'
                               |  .------------------------.
                               '->| MSTI | Fwding | Lrning |
                                  |------|--------|--------|
                                  |    0 | 111110 | 111110 |
                                  |   10 | 110111 | 110111 |
                                  '------------------------'

  What this is trying to show is that the STP state (whether MSTP is
  used, or ye olde STP) is always accessed via the VLAN table. If STP
  is running, all MSTI pointers in that table will reference the same
  index in the STP stable - if MSTP is running, some VLANs may point
  to other trees (like in this example).

  The fact that in the Linux bridge, the global state (think: index 0
  in most hardware implementations) is supposed to override the
  per-VLAN state, is very awkward to offload. In effect, this means
  that when the global state changes to blocking, drivers will have to
  iterate over all MSTIs in use, and alter them all to match. This
  also means that you have to cache whether the hardware state is
  currently tracking the global state or the per-VLAN state. In the
  first case, you also have to cache the per-VLAN state so that you
  can restore it if the global state transitions back to forwarding.

This series adds a new mst_enable bridge setting (as suggested by Nik)
that can only be changed when no VLANs are configured on the
bridge. Enabling this mode has the following effect:

- The port-global STP state is used to represent the CST (Common
  Spanning Tree) (1/10)

- Ingress STP filtering is deferred until the frame's VLAN has been
  resolved (1/10)

- The preexisting per-VLAN states can no longer be controlled directly
  (1/10). They are instead placed under the MST module's control,
  which is managed using a new netlink interface (described in 3/10)

- VLANs can br mapped to MSTIs in an arbitrary M:N fashion, using a
  new global VLAN option (2/10)

4-5/10 adds switchdev notifications so that a driver can track VID to
MSTI mappings and MST port states.

An offloading implementation is this provided for mv88e6xxx.

A proposal for the corresponding iproute2 interface is available here:

https://github.com/wkz/iproute2/tree/mst

Tobias Waldekranz (10):
  net: bridge: mst: Multiple Spanning Tree (MST) mode
  net: bridge: mst: Allow changing a VLAN's MSTI
  net: bridge: mst: Support setting and reporting MST port states
  net: bridge: mst: Notify switchdev drivers of VLAN MSTI migrations
  net: bridge: mst: Notify switchdev drivers of MST state changes
  net: dsa: Pass VLAN MSTI migration notifications to driver
  net: dsa: Pass MST state changes to driver
  net: dsa: mv88e6xxx: Disentangle STU from VTU
  net: dsa: mv88e6xxx: Export STU as devlink region
  net: dsa: mv88e6xxx: MST Offloading

 drivers/net/dsa/mv88e6xxx/chip.c        | 232 ++++++++++++++
 drivers/net/dsa/mv88e6xxx/chip.h        |  38 +++
 drivers/net/dsa/mv88e6xxx/devlink.c     |  94 ++++++
 drivers/net/dsa/mv88e6xxx/global1.h     |  10 +
 drivers/net/dsa/mv88e6xxx/global1_vtu.c | 311 ++++++++++--------
 include/net/dsa.h                       |   5 +
 include/net/switchdev.h                 |  17 +
 include/uapi/linux/if_bridge.h          |  17 +
 include/uapi/linux/if_link.h            |   1 +
 include/uapi/linux/rtnetlink.h          |   5 +
 net/bridge/Makefile                     |   2 +-
 net/bridge/br_input.c                   |  17 +-
 net/bridge/br_mst.c                     | 402 ++++++++++++++++++++++++
 net/bridge/br_netlink.c                 |  17 +-
 net/bridge/br_private.h                 |  31 ++
 net/bridge/br_stp.c                     |   3 +
 net/bridge/br_switchdev.c               |  57 ++++
 net/bridge/br_vlan.c                    |  20 +-
 net/bridge/br_vlan_options.c            |  24 +-
 net/dsa/dsa_priv.h                      |   3 +
 net/dsa/port.c                          |  40 +++
 net/dsa/slave.c                         |  12 +
 22 files changed, 1216 insertions(+), 142 deletions(-)
 create mode 100644 net/bridge/br_mst.c

-- 
2.25.1

Allow the user to switch from the current per-VLAN STP mode to an MST
mode.

Up to this point, per-VLAN STP states where always isolated from each
other. This is in contrast to the MSTP standard (802.1Q-2018, Clause
13.5), where VLANs are grouped into MST instances (MSTIs), and the
state is managed on a per-MSTI level, rather that at the per-VLAN
level.

Perhaps due to the prevalence of the standard, many switching ASICs
are built after the same model. Therefore, add a corresponding MST
mode to the bridge, which we can later add offloading support for in a
straight-forward way.

For now, all VLANs are fixed to MSTI 0, also called the Common
Spanning Tree (CST). That is, all VLANs will follow the port-global
state.

Upcoming changes will make this actually useful by allowing VLANs to
be mapped to arbitrary MSTIs and allow individual MSTI states to be
changed.

Signed-off-by: Tobias Waldekranz <tobias at waldekranz.com>
---
 include/uapi/linux/if_link.h |  1 +
 net/bridge/Makefile          |  2 +-
 net/bridge/br_input.c        | 17 +++++++-
 net/bridge/br_mst.c          | 83 ++++++++++++++++++++++++++++++++++++
 net/bridge/br_netlink.c      | 14 +++++-
 net/bridge/br_private.h      | 26 +++++++++++
 net/bridge/br_stp.c          |  3 ++
 net/bridge/br_vlan.c         | 20 ++++++++-
 net/bridge/br_vlan_options.c |  5 +++
 9 files changed, 166 insertions(+), 5 deletions(-)
 create mode 100644 net/bridge/br_mst.c

diff --git a/include/uapi/linux/if_link.h b/include/uapi/linux/if_link.h
index e315e53125f4..7e0a653aafa3 100644
--- a/include/uapi/linux/if_link.h
+++ b/include/uapi/linux/if_link.h
@@ -482,6 +482,7 @@ enum {
 	IFLA_BR_VLAN_STATS_PER_PORT,
 	IFLA_BR_MULTI_BOOLOPT,
 	IFLA_BR_MCAST_QUERIER_STATE,
+	IFLA_BR_MST_ENABLED,
 	__IFLA_BR_MAX,
 };
 
diff --git a/net/bridge/Makefile b/net/bridge/Makefile
index 7fb9a021873b..24bd1c0a9a5a 100644
--- a/net/bridge/Makefile
+++ b/net/bridge/Makefile
@@ -20,7 +20,7 @@ obj-$(CONFIG_BRIDGE_NETFILTER) += br_netfilter.o
 
 bridge-$(CONFIG_BRIDGE_IGMP_SNOOPING) += br_multicast.o br_mdb.o
br_multicast_eht.o
 
-bridge-$(CONFIG_BRIDGE_VLAN_FILTERING) += br_vlan.o br_vlan_tunnel.o
br_vlan_options.o
+bridge-$(CONFIG_BRIDGE_VLAN_FILTERING) += br_vlan.o br_vlan_tunnel.o
br_vlan_options.o br_mst.o
 
 bridge-$(CONFIG_NET_SWITCHDEV) += br_switchdev.o
 
diff --git a/net/bridge/br_input.c b/net/bridge/br_input.c
index e0c13fcc50ed..196417859c4a 100644
--- a/net/bridge/br_input.c
+++ b/net/bridge/br_input.c
@@ -78,13 +78,22 @@ int br_handle_frame_finish(struct net *net, struct sock *sk,
struct sk_buff *skb
 	u16 vid = 0;
 	u8 state;
 
-	if (!p || p->state == BR_STATE_DISABLED)
+	if (!p)
 		goto drop;
 
 	br = p->br;
+
+	if (br_mst_is_enabled(br)) {
+		state = BR_STATE_FORWARDING;
+	} else {
+		if (p->state == BR_STATE_DISABLED)
+			goto drop;
+
+		state = p->state;
+	}
+
 	brmctx = &p->br->multicast_ctx;
 	pmctx = &p->multicast_ctx;
-	state = p->state;
 	if (!br_allowed_ingress(p->br, nbp_vlan_group_rcu(p), skb, &vid,
 				&state, &vlan))
 		goto out;
@@ -370,9 +379,13 @@ static rx_handler_result_t br_handle_frame(struct sk_buff
**pskb)
 		return RX_HANDLER_PASS;
 
 forward:
+	if (br_mst_is_enabled(p->br))
+		goto defer_stp_filtering;
+
 	switch (p->state) {
 	case BR_STATE_FORWARDING:
 	case BR_STATE_LEARNING:
+defer_stp_filtering:
 		if (ether_addr_equal(p->br->dev->dev_addr, dest))
 			skb->pkt_type = PACKET_HOST;
 
diff --git a/net/bridge/br_mst.c b/net/bridge/br_mst.c
new file mode 100644
index 000000000000..ad6e91670fa8
--- /dev/null
+++ b/net/bridge/br_mst.c
@@ -0,0 +1,83 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ *	Bridge Multiple Spanning Tree Support
+ *
+ *	Authors:
+ *	Tobias Waldekranz		<tobias at waldekranz.com>
+ */
+
+#include <linux/kernel.h>
+
+#include "br_private.h"
+
+DEFINE_STATIC_KEY_FALSE(br_mst_used);
+
+void br_mst_vlan_set_state(struct net_bridge_port *p, struct net_bridge_vlan
*v,
+			   u8 state)
+{
+	struct net_bridge_vlan_group *vg = nbp_vlan_group(p);
+
+	if (v->state == state)
+		return;
+
+	br_vlan_set_state(v, state);
+
+	if (v->vid == vg->pvid)
+		br_vlan_set_pvid_state(vg, state);
+}
+
+void br_mst_set_state(struct net_bridge_port *p, u16 msti, u8 state)
+{
+	struct net_bridge_vlan_group *vg;
+	struct net_bridge_vlan *v;
+
+	vg = nbp_vlan_group(p);
+	if (!vg)
+		return;
+
+	list_for_each_entry(v, &vg->vlan_list, vlist) {
+		if (v->brvlan->msti != msti)
+			continue;
+
+		br_mst_vlan_set_state(p, v, state);
+	}
+}
+
+void br_mst_vlan_init_state(struct net_bridge_vlan *v)
+{
+	/* VLANs always start out in MSTI 0 (CST) */
+	v->msti = 0;
+
+	if (br_vlan_is_master(v))
+		v->state = BR_STATE_FORWARDING;
+	else
+		v->state = v->port->state;
+}
+
+int br_mst_set_enabled(struct net_bridge *br, unsigned long val)
+{
+	struct net_bridge_vlan_group *vg;
+	struct net_bridge_port *p;
+
+	/* Mode may only be changed when there are no port VLANs. */
+	list_for_each_entry(p, &br->port_list, list) {
+		vg = nbp_vlan_group(p);
+
+		if (vg->num_vlans)
+			return -EBUSY;
+	}
+
+	if (val > 1)
+		return -EINVAL;
+
+	if (!!val == br_opt_get(br, BROPT_MST_ENABLED))
+		return 0;
+
+	if (val)
+		static_branch_enable(&br_mst_used);
+	else
+		static_branch_disable(&br_mst_used);
+
+	br_opt_toggle(br, BROPT_MST_ENABLED, !!val);
+	return 0;
+}
diff --git a/net/bridge/br_netlink.c b/net/bridge/br_netlink.c
index 7d4432ca9a20..a17a0fe25a58 100644
--- a/net/bridge/br_netlink.c
+++ b/net/bridge/br_netlink.c
@@ -1163,6 +1163,7 @@ static const struct nla_policy br_policy[IFLA_BR_MAX + 1]
= {
 	[IFLA_BR_MCAST_IGMP_VERSION] = { .type = NLA_U8 },
 	[IFLA_BR_MCAST_MLD_VERSION] = { .type = NLA_U8 },
 	[IFLA_BR_VLAN_STATS_PER_PORT] = { .type = NLA_U8 },
+	[IFLA_BR_MST_ENABLED] = { .type = NLA_U8 },
 	[IFLA_BR_MULTI_BOOLOPT]  		NLA_POLICY_EXACT_LEN(sizeof(struct
br_boolopt_multi)),
 };
@@ -1255,6 +1256,14 @@ static int br_changelink(struct net_device *brdev, struct
nlattr *tb[],
 		if (err)
 			return err;
 	}
+
+	if (data[IFLA_BR_MST_ENABLED]) {
+		__u8 mst = nla_get_u8(data[IFLA_BR_MST_ENABLED]);
+
+		err = br_mst_set_enabled(br, mst);
+		if (err)
+			return err;
+	}
 #endif
 
 	if (data[IFLA_BR_GROUP_FWD_MASK]) {
@@ -1475,6 +1484,7 @@ static size_t br_get_size(const struct net_device *brdev)
 	       nla_total_size(sizeof(u16)) +    /* IFLA_BR_VLAN_DEFAULT_PVID */
 	       nla_total_size(sizeof(u8)) +     /* IFLA_BR_VLAN_STATS_ENABLED */
 	       nla_total_size(sizeof(u8)) +	/* IFLA_BR_VLAN_STATS_PER_PORT */
+	       nla_total_size(sizeof(u8)) +	/* IFLA_BR_MST_ENABLED */
 #endif
 	       nla_total_size(sizeof(u16)) +    /* IFLA_BR_GROUP_FWD_MASK */
 	       nla_total_size(sizeof(struct ifla_bridge_id)) +   /* IFLA_BR_ROOT_ID */
@@ -1572,7 +1582,9 @@ static int br_fill_info(struct sk_buff *skb, const struct
net_device *brdev)
 	    nla_put_u8(skb, IFLA_BR_VLAN_STATS_ENABLED,
 		       br_opt_get(br, BROPT_VLAN_STATS_ENABLED)) ||
 	    nla_put_u8(skb, IFLA_BR_VLAN_STATS_PER_PORT,
-		       br_opt_get(br, BROPT_VLAN_STATS_PER_PORT)))
+		       br_opt_get(br, BROPT_VLAN_STATS_PER_PORT)) ||
+	    nla_put_u8(skb, IFLA_BR_MST_ENABLED,
+		       br_opt_get(br, BROPT_MST_ENABLED)))
 		return -EMSGSIZE;
 #endif
 #ifdef CONFIG_BRIDGE_IGMP_SNOOPING
diff --git a/net/bridge/br_private.h b/net/bridge/br_private.h
index 48bc61ebc211..af50ad036b06 100644
--- a/net/bridge/br_private.h
+++ b/net/bridge/br_private.h
@@ -178,6 +178,7 @@ enum {
  * @br_mcast_ctx: if MASTER flag set, this is the global vlan multicast context
  * @port_mcast_ctx: if MASTER flag unset, this is the per-port/vlan multicast
  *                  context
+ * @msti: if MASTER flag set, this holds the VLANs MST instance
  * @vlist: sorted list of VLAN entries
  * @rcu: used for entry destruction
  *
@@ -210,6 +211,8 @@ struct net_bridge_vlan {
 		struct net_bridge_mcast_port	port_mcast_ctx;
 	};
 
+	u16				msti;
+
 	struct list_head		vlist;
 
 	struct rcu_head			rcu;
@@ -445,6 +448,7 @@ enum net_bridge_opts {
 	BROPT_NO_LL_LEARN,
 	BROPT_VLAN_BRIDGE_BINDING,
 	BROPT_MCAST_VLAN_SNOOPING_ENABLED,
+	BROPT_MST_ENABLED,
 };
 
 struct net_bridge {
@@ -1765,6 +1769,28 @@ static inline bool br_vlan_state_allowed(u8 state, bool
learn_allow)
 }
 #endif
 
+/* br_mst.c */
+#ifdef CONFIG_BRIDGE_VLAN_FILTERING
+DECLARE_STATIC_KEY_FALSE(br_mst_used);
+static inline bool br_mst_is_enabled(struct net_bridge *br)
+{
+	return static_branch_unlikely(&br_mst_used) &&
+		br_opt_get(br, BROPT_MST_ENABLED);
+}
+
+void br_mst_set_state(struct net_bridge_port *p, u16 msti, u8 state);
+void br_mst_vlan_init_state(struct net_bridge_vlan *v);
+int br_mst_set_enabled(struct net_bridge *br, unsigned long val);
+#else
+static inline bool br_mst_is_enabled(struct net_bridge *br)
+{
+	return false;
+}
+
+static inline void br_mst_set_state(struct net_bridge_port *p,
+				    u16 msti, u8 state) {}
+#endif
+
 struct nf_br_ops {
 	int (*br_dev_xmit_hook)(struct sk_buff *skb);
 };
diff --git a/net/bridge/br_stp.c b/net/bridge/br_stp.c
index 1d80f34a139c..82a97a021a57 100644
--- a/net/bridge/br_stp.c
+++ b/net/bridge/br_stp.c
@@ -43,6 +43,9 @@ void br_set_state(struct net_bridge_port *p, unsigned int
state)
 		return;
 
 	p->state = state;
+	if (br_opt_get(p->br, BROPT_MST_ENABLED))
+		br_mst_set_state(p, 0, state);
+
 	err = switchdev_port_attr_set(p->dev, &attr, NULL);
 	if (err && err != -EOPNOTSUPP)
 		br_warn(p->br, "error setting offload STP state on port
%u(%s)\n",
diff --git a/net/bridge/br_vlan.c b/net/bridge/br_vlan.c
index 7557e90b60e1..0f5e75ccac79 100644
--- a/net/bridge/br_vlan.c
+++ b/net/bridge/br_vlan.c
@@ -226,6 +226,24 @@ static void nbp_vlan_rcu_free(struct rcu_head *rcu)
 	kfree(v);
 }
 
+static void br_vlan_init_state(struct net_bridge_vlan *v)
+{
+	struct net_bridge *br;
+
+	if (br_vlan_is_master(v))
+		br = v->br;
+	else
+		br = v->port->br;
+
+	if (br_opt_get(br, BROPT_MST_ENABLED)) {
+		br_mst_vlan_init_state(v);
+		return;
+	}
+
+	v->state = BR_STATE_FORWARDING;
+	v->msti = 0;
+}
+
 /* This is the shared VLAN add function which works for both ports and bridge
  * devices. There are four possible calls to this function in terms of the
  * vlan entry type:
@@ -322,7 +340,7 @@ static int __vlan_add(struct net_bridge_vlan *v, u16 flags,
 	}
 
 	/* set the state before publishing */
-	v->state = BR_STATE_FORWARDING;
+	br_vlan_init_state(v);
 
 	err = rhashtable_lookup_insert_fast(&vg->vlan_hash, &v->vnode,
 					    br_vlan_rht_params);
diff --git a/net/bridge/br_vlan_options.c b/net/bridge/br_vlan_options.c
index a6382973b3e7..09112b56e79c 100644
--- a/net/bridge/br_vlan_options.c
+++ b/net/bridge/br_vlan_options.c
@@ -99,6 +99,11 @@ static int br_vlan_modify_state(struct net_bridge_vlan_group
*vg,
 		return -EBUSY;
 	}
 
+	if (br_opt_get(br, BROPT_MST_ENABLED)) {
+		NL_SET_ERR_MSG_MOD(extack, "Can't modify vlan state directly when
MST is enabled");
+		return -EBUSY;
+	}
+
 	if (v->state == state)
 		return 0;
 
-- 
2.25.1

Allow a VLAN to move out of the CST (MSTI 0), to an independent tree.

The user manages the VID to MSTI mappings via a global VLAN
setting. The proposed iproute2 interface would be:

    bridge vlan global set dev br0 vid <VID> msti <MSTI>

Changing the state in non-zero MSTIs is still not supported, but will
be addressed in upcoming changes.

Signed-off-by: Tobias Waldekranz <tobias at waldekranz.com>
---
 include/uapi/linux/if_bridge.h |  1 +
 net/bridge/br_mst.c            | 39 ++++++++++++++++++++++++++++++++++
 net/bridge/br_private.h        |  1 +
 net/bridge/br_vlan_options.c   | 19 +++++++++++++++--
 4 files changed, 58 insertions(+), 2 deletions(-)

diff --git a/include/uapi/linux/if_bridge.h b/include/uapi/linux/if_bridge.h
index 2711c3522010..b68016f625b7 100644
--- a/include/uapi/linux/if_bridge.h
+++ b/include/uapi/linux/if_bridge.h
@@ -564,6 +564,7 @@ enum {
 	BRIDGE_VLANDB_GOPTS_MCAST_QUERIER,
 	BRIDGE_VLANDB_GOPTS_MCAST_ROUTER_PORTS,
 	BRIDGE_VLANDB_GOPTS_MCAST_QUERIER_STATE,
+	BRIDGE_VLANDB_GOPTS_MSTI,
 	__BRIDGE_VLANDB_GOPTS_MAX
 };
 #define BRIDGE_VLANDB_GOPTS_MAX (__BRIDGE_VLANDB_GOPTS_MAX - 1)
diff --git a/net/bridge/br_mst.c b/net/bridge/br_mst.c
index ad6e91670fa8..f3b8e279b85c 100644
--- a/net/bridge/br_mst.c
+++ b/net/bridge/br_mst.c
@@ -43,6 +43,45 @@ void br_mst_set_state(struct net_bridge_port *p, u16 msti, u8
state)
 	}
 }
 
+static void br_mst_vlan_sync_state(struct net_bridge_vlan *pv, u16 msti)
+{
+	struct net_bridge_vlan_group *vg = nbp_vlan_group(pv->port);
+	struct net_bridge_vlan *v;
+
+	list_for_each_entry(v, &vg->vlan_list, vlist) {
+		/* If this port already has a defined state in this
+		 * MSTI (through some other VLAN membership), inherit
+		 * it.
+		 */
+		if (v != pv && v->brvlan->msti == msti) {
+			br_mst_vlan_set_state(pv->port, pv, v->state);
+			return;
+		}
+	}
+
+	/* Otherwise, start out in a new MSTI with all ports disabled. */
+	return br_mst_vlan_set_state(pv->port, pv, BR_STATE_DISABLED);
+}
+
+int br_mst_vlan_set_msti(struct net_bridge_vlan *mv, u16 msti)
+{
+	struct net_bridge_vlan_group *vg;
+	struct net_bridge_vlan *pv;
+	struct net_bridge_port *p;
+
+	mv->msti = msti;
+
+	list_for_each_entry(p, &mv->br->port_list, list) {
+		vg = nbp_vlan_group(p);
+
+		pv = br_vlan_find(vg, mv->vid);
+		if (pv)
+			br_mst_vlan_sync_state(pv, msti);
+	}
+
+	return 0;
+}
+
 void br_mst_vlan_init_state(struct net_bridge_vlan *v)
 {
 	/* VLANs always start out in MSTI 0 (CST) */
diff --git a/net/bridge/br_private.h b/net/bridge/br_private.h
index af50ad036b06..63601043abca 100644
--- a/net/bridge/br_private.h
+++ b/net/bridge/br_private.h
@@ -1779,6 +1779,7 @@ static inline bool br_mst_is_enabled(struct net_bridge
*br)
 }
 
 void br_mst_set_state(struct net_bridge_port *p, u16 msti, u8 state);
+int br_mst_vlan_set_msti(struct net_bridge_vlan *v, u16 msti);
 void br_mst_vlan_init_state(struct net_bridge_vlan *v);
 int br_mst_set_enabled(struct net_bridge *br, unsigned long val);
 #else
diff --git a/net/bridge/br_vlan_options.c b/net/bridge/br_vlan_options.c
index 09112b56e79c..87a9c80d26d3 100644
--- a/net/bridge/br_vlan_options.c
+++ b/net/bridge/br_vlan_options.c
@@ -44,8 +44,8 @@ bool br_vlan_opts_eq_range(const struct net_bridge_vlan
*v_curr,
 	u8 curr_mc_rtr = br_vlan_multicast_router(v_curr);
 
 	return v_curr->state == range_end->state &&
-	       __vlan_tun_can_enter_range(v_curr, range_end) &&
-	       curr_mc_rtr == range_mc_rtr;
+		__vlan_tun_can_enter_range(v_curr, range_end) &&
+		curr_mc_rtr == range_mc_rtr;
 }
 
 bool br_vlan_opts_fill(struct sk_buff *skb, const struct net_bridge_vlan *v)
@@ -296,6 +296,7 @@ bool br_vlan_global_opts_can_enter_range(const struct
net_bridge_vlan *v_curr,
 					 const struct net_bridge_vlan *r_end)
 {
 	return v_curr->vid - r_end->vid == 1 &&
+		v_curr->msti == r_end->msti &&
 	       ((v_curr->priv_flags ^ r_end->priv_flags) &
 		BR_VLFLAG_GLOBAL_MCAST_ENABLED) == 0 &&
 		br_multicast_ctx_options_equal(&v_curr->br_mcast_ctx,
@@ -384,6 +385,9 @@ bool br_vlan_global_opts_fill(struct sk_buff *skb, u16 vid,
u16 vid_range,
 #endif
 #endif
 
+	if (nla_put_u16(skb, BRIDGE_VLANDB_GOPTS_MSTI, v_opts->msti))
+		goto out_err;
+
 	nla_nest_end(skb, nest);
 
 	return true;
@@ -415,6 +419,7 @@ static size_t rtnl_vlan_global_opts_nlmsg_size(const struct
net_bridge_vlan *v)
 		+ nla_total_size(0) /* BRIDGE_VLANDB_GOPTS_MCAST_ROUTER_PORTS */
 		+ br_rports_size(&v->br_mcast_ctx) /*
BRIDGE_VLANDB_GOPTS_MCAST_ROUTER_PORTS */
 #endif
+		+ nla_total_size(sizeof(u16)) /* BRIDGE_VLANDB_GOPTS_MSTI */
 		+ nla_total_size(sizeof(u16)); /* BRIDGE_VLANDB_GOPTS_RANGE */
 }
 
@@ -564,6 +569,15 @@ static int br_vlan_process_global_one_opts(const struct
net_bridge *br,
 	}
 #endif
 #endif
+	if (tb[BRIDGE_VLANDB_GOPTS_MSTI]) {
+		u16 msti;
+
+		msti = nla_get_u16(tb[BRIDGE_VLANDB_GOPTS_MSTI]);
+		err = br_mst_vlan_set_msti(v, msti);
+		if (err)
+			return err;
+		*changed = true;
+	}
 
 	return 0;
 }
@@ -583,6 +597,7 @@ static const struct nla_policy
br_vlan_db_gpol[BRIDGE_VLANDB_GOPTS_MAX + 1] = {
 	[BRIDGE_VLANDB_GOPTS_MCAST_QUERIER_INTVL]	= { .type = NLA_U64 },
 	[BRIDGE_VLANDB_GOPTS_MCAST_STARTUP_QUERY_INTVL]	= { .type = NLA_U64 },
 	[BRIDGE_VLANDB_GOPTS_MCAST_QUERY_RESPONSE_INTVL] = { .type = NLA_U64 },
+	[BRIDGE_VLANDB_GOPTS_MSTI] = NLA_POLICY_MAX(NLA_U16, VLAN_N_VID - 1),
 };
 
 int br_vlan_rtm_process_global_options(struct net_device *dev,
-- 
2.25.1

Make it possible to change the port state in a given MSTI. This is
done through a new netlink interface, since the MSTIs are objects in
their own right. The proposed iproute2 interface would be:

    bridge mst set dev <PORT> msti <MSTI> state <STATE>

Current states in all applicable MSTIs can also be dumped. The
proposed iproute interface looks like this:

$ bridge mst
port              msti
vb1               0
		    state forwarding
		  100
		    state disabled
vb2               0
		    state forwarding
		  100
		    state forwarding

The preexisting per-VLAN states are still valid in the MST
mode (although they are read-only), and can be queried as usual if one
is interested in knowing a particular VLAN's state without having to
care about the VID to MSTI mapping (in this example VLAN 20 and 30 are
bound to MSTI 100):

$ bridge -d vlan
port              vlan-id
vb1               10
		    state forwarding mcast_router 1
		  20
		    state disabled mcast_router 1
		  30
		    state disabled mcast_router 1
		  40
		    state forwarding mcast_router 1
vb2               10
		    state forwarding mcast_router 1
		  20
		    state forwarding mcast_router 1
		  30
		    state forwarding mcast_router 1
		  40
		    state forwarding mcast_router 1

Signed-off-by: Tobias Waldekranz <tobias at waldekranz.com>
---
 include/uapi/linux/if_bridge.h |  16 +++
 include/uapi/linux/rtnetlink.h |   5 +
 net/bridge/br_mst.c            | 244 +++++++++++++++++++++++++++++++++
 net/bridge/br_netlink.c        |   3 +
 net/bridge/br_private.h        |   4 +
 5 files changed, 272 insertions(+)

diff --git a/include/uapi/linux/if_bridge.h b/include/uapi/linux/if_bridge.h
index b68016f625b7..784482527861 100644
--- a/include/uapi/linux/if_bridge.h
+++ b/include/uapi/linux/if_bridge.h
@@ -785,4 +785,20 @@ enum {
 	__BRIDGE_QUERIER_MAX
 };
 #define BRIDGE_QUERIER_MAX (__BRIDGE_QUERIER_MAX - 1)
+
+enum {
+	BRIDGE_MST_UNSPEC,
+	BRIDGE_MST_ENTRY,
+	__BRIDGE_MST_MAX,
+};
+#define BRIDGE_MST_MAX (__BRIDGE_MST_MAX - 1)
+
+enum {
+	BRIDGE_MST_ENTRY_UNSPEC,
+	BRIDGE_MST_ENTRY_MSTI,
+	BRIDGE_MST_ENTRY_STATE,
+	__BRIDGE_MST_ENTRY_MAX,
+};
+#define BRIDGE_MST_ENTRY_MAX (__BRIDGE_MST_ENTRY_MAX - 1)
+
 #endif /* _UAPI_LINUX_IF_BRIDGE_H */
diff --git a/include/uapi/linux/rtnetlink.h b/include/uapi/linux/rtnetlink.h
index 0970cb4b1b88..4a48f3ce862c 100644
--- a/include/uapi/linux/rtnetlink.h
+++ b/include/uapi/linux/rtnetlink.h
@@ -192,6 +192,11 @@ enum {
 	RTM_GETTUNNEL,
 #define RTM_GETTUNNEL	RTM_GETTUNNEL
 
+	RTM_GETMST = 124 + 2,
+#define RTM_GETMST	RTM_GETMST
+	RTM_SETMST,
+#define RTM_SETMST	RTM_SETMST
+
 	__RTM_MAX,
 #define RTM_MAX		(((__RTM_MAX + 3) & ~3) - 1)
 };
diff --git a/net/bridge/br_mst.c b/net/bridge/br_mst.c
index f3b8e279b85c..8dea8e7257fd 100644
--- a/net/bridge/br_mst.c
+++ b/net/bridge/br_mst.c
@@ -120,3 +120,247 @@ int br_mst_set_enabled(struct net_bridge *br, unsigned
long val)
 	br_opt_toggle(br, BROPT_MST_ENABLED, !!val);
 	return 0;
 }
+
+static int br_mst_nl_get_one(struct net_bridge_port *p, struct sk_buff *skb,
+			     struct netlink_callback *cb)
+{
+	struct net_bridge_vlan_group *vg = nbp_vlan_group(p);
+	int err = 0, idx = 0, s_idx = cb->args[1];
+	struct net_bridge_vlan *v;
+	struct br_port_msg *bpm;
+	struct nlmsghdr *nlh;
+	struct nlattr *nest;
+	unsigned long *seen;
+
+	nlh = nlmsg_put(skb, NETLINK_CB(cb->skb).portid, cb->nlh->nlmsg_seq,
+			RTM_GETMST, sizeof(*bpm), NLM_F_MULTI);
+	if (!nlh)
+		return -EMSGSIZE;
+
+	bpm = nlmsg_data(nlh);
+	memset(bpm, 0, sizeof(*bpm));
+	bpm->ifindex = p->dev->ifindex;
+
+	seen = bitmap_zalloc(VLAN_N_VID, 0);
+	if (!seen)
+		return -ENOMEM;
+
+	list_for_each_entry(v, &vg->vlan_list, vlist) {
+		if (test_bit(v->brvlan->msti, seen))
+			continue;
+
+		if (idx < s_idx)
+			goto skip;
+
+		nest = nla_nest_start_noflag(skb, BRIDGE_MST_ENTRY);
+		if (!nest ||
+		    nla_put_u16(skb, BRIDGE_MST_ENTRY_MSTI, v->brvlan->msti) ||
+		    nla_put_u8(skb, BRIDGE_MST_ENTRY_STATE, v->state)) {
+			err = -EMSGSIZE;
+			break;
+		}
+		nla_nest_end(skb, nest);
+
+		set_bit(v->brvlan->msti, seen);
+skip:
+		idx++;
+	}
+
+	kfree(seen);
+	nlmsg_end(skb, nlh);
+	return err;
+}
+
+static struct net_bridge_port *br_mst_nl_get_parse(struct net *net,
+						   struct netlink_callback *cb)
+{
+	struct netlink_ext_ack *extack = cb->extack;
+	const struct nlmsghdr *nlh = cb->nlh;
+	struct net_bridge_port *p;
+	struct br_port_msg *bpm;
+	struct net_device *dev;
+
+	if (nlh->nlmsg_len < nlmsg_msg_size(sizeof(*bpm))) {
+		NL_SET_ERR_MSG_MOD(extack, "Invalid header for mst get request");
+		return ERR_PTR(-EINVAL);
+	}
+
+	if (nlmsg_attrlen(nlh, sizeof(*bpm))) {
+		NL_SET_ERR_MSG(extack, "Invalid data after header in mst get
request");
+		return ERR_PTR(-EINVAL);
+	}
+
+	bpm = nlmsg_data(nlh);
+	if (!bpm->ifindex)
+		return NULL;
+
+	dev = __dev_get_by_index(net, bpm->ifindex);
+	if (!dev)
+		return ERR_PTR(-ENODEV);
+
+	if (!netif_is_bridge_port(dev)) {
+		NL_SET_ERR_MSG_MOD(extack, "The device is not a valid bridge
port");
+		return ERR_PTR(-EINVAL);
+	}
+
+	p = br_port_get_rtnl(dev);
+	if (WARN_ON(!p))
+		return ERR_PTR(-ENODEV);
+
+	if (!br_opt_get(p->br, BROPT_MST_ENABLED)) {
+		NL_SET_ERR_MSG_MOD(extack, "Can't query MST state when MST is
disabled");
+		return ERR_PTR(-EINVAL);
+	}
+
+	return p;
+}
+
+static int br_mst_nl_get(struct sk_buff *skb, struct netlink_callback *cb)
+{
+	int err = 0, idx = 0, s_idx = cb->args[0];
+	struct net *net = sock_net(skb->sk);
+	struct net_bridge_port *p;
+	struct net_device *dev;
+
+	p = br_mst_nl_get_parse(net, cb);
+	if (IS_ERR(p))
+		return PTR_ERR(p);
+
+	if (p) {
+		err = br_mst_nl_get_one(p, skb, cb);
+		if (err != -EMSGSIZE)
+			return err;
+	} else {
+		for_each_netdev(net, dev) {
+			if (!netif_is_bridge_port(dev))
+				continue;
+
+			if (idx < s_idx)
+				goto skip;
+
+			p = br_port_get_rtnl(dev);
+			if (WARN_ON(!p))
+				return -ENODEV;
+
+			err = br_mst_nl_get_one(p, skb, cb);
+			if (err == -EMSGSIZE)
+				break;
+skip:
+			idx++;
+		}
+	}
+
+	cb->args[0] = idx;
+	return skb->len;
+}
+
+static const struct nla_policy br_mst_nl_policy[BRIDGE_MST_ENTRY_MAX + 1] = {
+	[BRIDGE_MST_ENTRY_MSTI] = NLA_POLICY_RANGE(NLA_U16,
+						   1, /* 0 reserved for CST */
+						   VLAN_N_VID - 1),
+	[BRIDGE_MST_ENTRY_STATE] = NLA_POLICY_RANGE(NLA_U8,
+						    BR_STATE_DISABLED,
+						    BR_STATE_BLOCKING),
+};
+
+static int br_mst_nl_set_one(struct net_bridge_port *p,
+			     const struct nlattr *attr,
+			     struct netlink_ext_ack *extack)
+{
+	struct nlattr *tb[BRIDGE_MST_ENTRY_MAX + 1];
+	u16 msti;
+	u8 state;
+	int err;
+
+	err = nla_parse_nested(tb, BRIDGE_MST_ENTRY_MAX, attr,
+			       br_mst_nl_policy, extack);
+	if (err)
+		return err;
+
+	if (!tb[BRIDGE_MST_ENTRY_MSTI]) {
+		NL_SET_ERR_MSG_MOD(extack, "MSTI not specified");
+		return -EINVAL;
+	}
+
+	if (!tb[BRIDGE_MST_ENTRY_STATE]) {
+		NL_SET_ERR_MSG_MOD(extack, "State not specified");
+		return -EINVAL;
+	}
+
+	msti = nla_get_u16(tb[BRIDGE_MST_ENTRY_MSTI]);
+	state = nla_get_u8(tb[BRIDGE_MST_ENTRY_STATE]);
+
+	br_mst_set_state(p, msti, state);
+	return 0;
+}
+
+static int br_mst_nl_set(struct sk_buff *skb, struct nlmsghdr *nlh,
+			 struct netlink_ext_ack *extack)
+{
+	struct net *net = sock_net(skb->sk);
+	struct net_bridge_port *p;
+	struct br_port_msg *bpm;
+	struct net_device *dev;
+	struct nlattr *attr;
+	int err, msts = 0;
+	int rem;
+
+	err = nlmsg_parse(nlh, sizeof(*bpm), NULL, BRIDGE_MST_MAX, NULL,
+			  extack);
+	if (err < 0)
+		return err;
+
+	bpm = nlmsg_data(nlh);
+	dev = __dev_get_by_index(net, bpm->ifindex);
+	if (!dev)
+		return -ENODEV;
+
+	if (!netif_is_bridge_port(dev)) {
+		NL_SET_ERR_MSG_MOD(extack, "The device is not a valid bridge
port");
+		return -EINVAL;
+	}
+
+	p = br_port_get_rtnl(dev);
+	if (WARN_ON(!p))
+		return -ENODEV;
+
+	if (!br_opt_get(p->br, BROPT_MST_ENABLED)) {
+		NL_SET_ERR_MSG_MOD(extack, "Can't modify MST state when MST is
disabled");
+		return -EBUSY;
+	}
+
+	nlmsg_for_each_attr(attr, nlh, sizeof(*bpm), rem) {
+		switch (nla_type(attr)) {
+		case BRIDGE_MST_ENTRY:
+			err = br_mst_nl_set_one(p, attr, extack);
+			break;
+		default:
+			continue;
+		}
+
+		msts++;
+		if (err)
+			break;
+	}
+
+	if (!msts) {
+		NL_SET_ERR_MSG_MOD(extack, "Found no MST entries to process");
+		err = -EINVAL;
+	}
+
+	return err;
+}
+
+void br_mst_rtnl_init(void)
+{
+	rtnl_register_module(THIS_MODULE, PF_BRIDGE, RTM_GETMST, NULL,
+			     br_mst_nl_get, 0);
+	rtnl_register_module(THIS_MODULE, PF_BRIDGE, RTM_SETMST,
+			     br_mst_nl_set, NULL, 0);
+}
+
+void br_mst_rtnl_uninit(void)
+{
+	rtnl_unregister(PF_BRIDGE, RTM_SETMST);
+	rtnl_unregister(PF_BRIDGE, RTM_GETMST);
+}
diff --git a/net/bridge/br_netlink.c b/net/bridge/br_netlink.c
index a17a0fe25a58..6d70d6f9cf17 100644
--- a/net/bridge/br_netlink.c
+++ b/net/bridge/br_netlink.c
@@ -1813,6 +1813,7 @@ int __init br_netlink_init(void)
 
 	br_mdb_init();
 	br_vlan_rtnl_init();
+	br_mst_rtnl_init();
 	rtnl_af_register(&br_af_ops);
 
 	err = rtnl_link_register(&br_link_ops);
@@ -1824,6 +1825,7 @@ int __init br_netlink_init(void)
 out_af:
 	rtnl_af_unregister(&br_af_ops);
 	br_mdb_uninit();
+	br_mst_rtnl_uninit();
 	return err;
 }
 
@@ -1831,6 +1833,7 @@ void br_netlink_fini(void)
 {
 	br_mdb_uninit();
 	br_vlan_rtnl_uninit();
+	br_mst_rtnl_uninit();
 	rtnl_af_unregister(&br_af_ops);
 	rtnl_link_unregister(&br_link_ops);
 }
diff --git a/net/bridge/br_private.h b/net/bridge/br_private.h
index 63601043abca..7882a65ffb43 100644
--- a/net/bridge/br_private.h
+++ b/net/bridge/br_private.h
@@ -1782,6 +1782,8 @@ void br_mst_set_state(struct net_bridge_port *p, u16 msti,
u8 state);
 int br_mst_vlan_set_msti(struct net_bridge_vlan *v, u16 msti);
 void br_mst_vlan_init_state(struct net_bridge_vlan *v);
 int br_mst_set_enabled(struct net_bridge *br, unsigned long val);
+void br_mst_rtnl_init(void);
+void br_mst_rtnl_uninit(void);
 #else
 static inline bool br_mst_is_enabled(struct net_bridge *br)
 {
@@ -1790,6 +1792,8 @@ static inline bool br_mst_is_enabled(struct net_bridge
*br)
 
 static inline void br_mst_set_state(struct net_bridge_port *p,
 				    u16 msti, u8 state) {}
+static inline void br_mst_rtnl_init(void) {}
+static inline void br_mst_rtnl_uninit(void) {}
 #endif
 
 struct nf_br_ops {
-- 
2.25.1

Whenever a VLAN moves to a new MSTI, send a switchdev notification so
that switchdevs can...

...either refuse the migration if the hardware does not support
offloading of MST...

..or track a bridge's VID to MSTI mapping when offloading is
supported.

Signed-off-by: Tobias Waldekranz <tobias at waldekranz.com>
---
 include/net/switchdev.h   | 10 +++++++
 net/bridge/br_mst.c       | 15 +++++++++++
 net/bridge/br_switchdev.c | 57 +++++++++++++++++++++++++++++++++++++++
 3 files changed, 82 insertions(+)

diff --git a/include/net/switchdev.h b/include/net/switchdev.h
index 3e424d40fae3..39e57aa5005a 100644
--- a/include/net/switchdev.h
+++ b/include/net/switchdev.h
@@ -28,6 +28,7 @@ enum switchdev_attr_id {
 	SWITCHDEV_ATTR_ID_BRIDGE_MC_DISABLED,
 	SWITCHDEV_ATTR_ID_BRIDGE_MROUTER,
 	SWITCHDEV_ATTR_ID_MRP_PORT_ROLE,
+	SWITCHDEV_ATTR_ID_VLAN_MSTI,
 };
 
 struct switchdev_brport_flags {
@@ -35,6 +36,14 @@ struct switchdev_brport_flags {
 	unsigned long mask;
 };
 
+struct switchdev_vlan_attr {
+	u16 vid;
+
+	union {
+		u16 msti;
+	};
+};
+
 struct switchdev_attr {
 	struct net_device *orig_dev;
 	enum switchdev_attr_id id;
@@ -50,6 +59,7 @@ struct switchdev_attr {
 		u16 vlan_protocol;			/* BRIDGE_VLAN_PROTOCOL */
 		bool mc_disabled;			/* MC_DISABLED */
 		u8 mrp_port_role;			/* MRP_PORT_ROLE */
+		struct switchdev_vlan_attr vlan_attr;	/* VLAN_* */
 	} u;
 };
 
diff --git a/net/bridge/br_mst.c b/net/bridge/br_mst.c
index 8dea8e7257fd..aba603675165 100644
--- a/net/bridge/br_mst.c
+++ b/net/bridge/br_mst.c
@@ -7,6 +7,7 @@
  */
 
 #include <linux/kernel.h>
+#include <net/switchdev.h>
 
 #include "br_private.h"
 
@@ -65,9 +66,23 @@ static void br_mst_vlan_sync_state(struct net_bridge_vlan
*pv, u16 msti)
 
 int br_mst_vlan_set_msti(struct net_bridge_vlan *mv, u16 msti)
 {
+	struct switchdev_attr attr = {
+		.id = SWITCHDEV_ATTR_ID_VLAN_MSTI,
+		.flags = SWITCHDEV_F_DEFER,
+		.orig_dev = mv->br->dev,
+		.u.vlan_attr = {
+			.vid = mv->vid,
+			.msti = msti,
+		},
+	};
 	struct net_bridge_vlan_group *vg;
 	struct net_bridge_vlan *pv;
 	struct net_bridge_port *p;
+	int err;
+
+	err = switchdev_port_attr_set(mv->br->dev, &attr, NULL);
+	if (err && err != -EOPNOTSUPP)
+		return err;
 
 	mv->msti = msti;
 
diff --git a/net/bridge/br_switchdev.c b/net/bridge/br_switchdev.c
index 6f6a70121a5e..160d7659f88a 100644
--- a/net/bridge/br_switchdev.c
+++ b/net/bridge/br_switchdev.c
@@ -428,6 +428,57 @@ static int br_switchdev_vlan_replay(struct net_device
*br_dev,
 	return 0;
 }
 
+static int br_switchdev_mst_replay(struct net_device *br_dev,
+				   const void *ctx, bool adding,
+				   struct notifier_block *nb,
+				   struct netlink_ext_ack *extack)
+{
+	struct switchdev_notifier_port_attr_info attr_info = {
+		.info = {
+			.dev = br_dev,
+			.extack = extack,
+			.ctx = ctx,
+		},
+	};
+	struct net_bridge *br = netdev_priv(br_dev);
+	struct net_bridge_vlan_group *vg;
+	struct net_bridge_vlan *v;
+	int err;
+
+	ASSERT_RTNL();
+
+	if (!nb)
+		return 0;
+
+	if (!netif_is_bridge_master(br_dev))
+		return -EINVAL;
+
+	vg = br_vlan_group(br);
+
+	list_for_each_entry(v, &vg->vlan_list, vlist) {
+		struct switchdev_attr attr = {
+			.id = SWITCHDEV_ATTR_ID_VLAN_MSTI,
+			.flags = SWITCHDEV_F_DEFER,
+			.orig_dev = br_dev,
+			.u.vlan_attr = {
+				.vid = v->vid,
+				.msti = v->msti,
+			}
+		};
+
+		if (!v->msti)
+			continue;
+
+		attr_info.attr = &attr;
+		err = nb->notifier_call(nb, SWITCHDEV_PORT_ATTR_SET, &attr_info);
+		err = notifier_to_errno(err);
+		if (err)
+			return err;
+	}
+
+	return 0;
+}
+
 #ifdef CONFIG_BRIDGE_IGMP_SNOOPING
 struct br_switchdev_mdb_complete_info {
 	struct net_bridge_port *port;
@@ -695,6 +746,10 @@ static int nbp_switchdev_sync_objs(struct net_bridge_port
*p, const void *ctx,
 	if (err && err != -EOPNOTSUPP)
 		return err;
 
+	err = br_switchdev_mst_replay(br_dev, ctx, true, blocking_nb, extack);
+	if (err && err != -EOPNOTSUPP)
+		return err;
+
 	err = br_switchdev_mdb_replay(br_dev, dev, ctx, true, blocking_nb,
 				      extack);
 	if (err && err != -EOPNOTSUPP)
@@ -719,6 +774,8 @@ static void nbp_switchdev_unsync_objs(struct net_bridge_port
*p,
 
 	br_switchdev_mdb_replay(br_dev, dev, ctx, false, blocking_nb, NULL);
 
+	br_switchdev_mst_replay(br_dev, ctx, false, blocking_nb, NULL);
+
 	br_switchdev_vlan_replay(br_dev, ctx, false, blocking_nb, NULL);
 }
 
-- 
2.25.1

Generate a switchdev notification whenever an MST state changes. This
notification is keyed by the VLANs MSTI rather than the VID, since
multiple VLANs may share the same MST instance.

Signed-off-by: Tobias Waldekranz <tobias at waldekranz.com>
---
 include/net/switchdev.h |  7 +++++++
 net/bridge/br_mst.c     | 21 +++++++++++++++++++++
 2 files changed, 28 insertions(+)

diff --git a/include/net/switchdev.h b/include/net/switchdev.h
index 39e57aa5005a..441beeb2fda5 100644
--- a/include/net/switchdev.h
+++ b/include/net/switchdev.h
@@ -19,6 +19,7 @@
 enum switchdev_attr_id {
 	SWITCHDEV_ATTR_ID_UNDEFINED,
 	SWITCHDEV_ATTR_ID_PORT_STP_STATE,
+	SWITCHDEV_ATTR_ID_PORT_MST_STATE,
 	SWITCHDEV_ATTR_ID_PORT_BRIDGE_FLAGS,
 	SWITCHDEV_ATTR_ID_PORT_PRE_BRIDGE_FLAGS,
 	SWITCHDEV_ATTR_ID_PORT_MROUTER,
@@ -31,6 +32,11 @@ enum switchdev_attr_id {
 	SWITCHDEV_ATTR_ID_VLAN_MSTI,
 };
 
+struct switchdev_mst_state {
+	u16 msti;
+	u8 state;
+};
+
 struct switchdev_brport_flags {
 	unsigned long val;
 	unsigned long mask;
@@ -52,6 +58,7 @@ struct switchdev_attr {
 	void (*complete)(struct net_device *dev, int err, void *priv);
 	union {
 		u8 stp_state;				/* PORT_STP_STATE */
+		struct switchdev_mst_state mst_state;	/* PORT_MST_STATE */
 		struct switchdev_brport_flags brport_flags; /* PORT_BRIDGE_FLAGS */
 		bool mrouter;				/* PORT_MROUTER */
 		clock_t ageing_time;			/* BRIDGE_AGEING_TIME */
diff --git a/net/bridge/br_mst.c b/net/bridge/br_mst.c
index aba603675165..9cdd7d9e07c6 100644
--- a/net/bridge/br_mst.c
+++ b/net/bridge/br_mst.c
@@ -29,8 +29,18 @@ void br_mst_vlan_set_state(struct net_bridge_port *p, struct
net_bridge_vlan *v,
 
 void br_mst_set_state(struct net_bridge_port *p, u16 msti, u8 state)
 {
+	struct switchdev_attr attr = {
+		.id = SWITCHDEV_ATTR_ID_PORT_MST_STATE,
+		.flags = SWITCHDEV_F_DEFER,
+		.orig_dev = p->dev,
+		.u.mst_state = {
+			.msti = msti,
+			.state = state,
+		},
+	};
 	struct net_bridge_vlan_group *vg;
 	struct net_bridge_vlan *v;
+	int err;
 
 	vg = nbp_vlan_group(p);
 	if (!vg)
@@ -42,6 +52,17 @@ void br_mst_set_state(struct net_bridge_port *p, u16 msti, u8
state)
 
 		br_mst_vlan_set_state(p, v, state);
 	}
+
+	if (!msti)
+		/* MSTI 0 (CST) state changes are notified via the
+		 * regular SWITCHDEV_ATTR_ID_PORT_STP_STATE.
+		 */
+		return;
+
+	err = switchdev_port_attr_set(p->dev, &attr, NULL);
+	if (err && err != -EOPNOTSUPP)
+		br_warn(p->br, "unable to offload MST state on %s in MSTI %u",
+			netdev_name(p->dev), msti);
 }
 
 static void br_mst_vlan_sync_state(struct net_bridge_vlan *pv, u16 msti)
-- 
2.25.1

Add the usual trampoline functionality from the generic DSA layer down
to the drivers for VLAN MSTI migrations.

Signed-off-by: Tobias Waldekranz <tobias at waldekranz.com>
---
 include/net/dsa.h  |  3 +++
 net/dsa/dsa_priv.h |  1 +
 net/dsa/port.c     | 10 ++++++++++
 net/dsa/slave.c    |  6 ++++++
 4 files changed, 20 insertions(+)

diff --git a/include/net/dsa.h b/include/net/dsa.h
index cfedcfb86350..cc8acb01bd9b 100644
--- a/include/net/dsa.h
+++ b/include/net/dsa.h
@@ -962,6 +962,9 @@ struct dsa_switch_ops {
 				 struct netlink_ext_ack *extack);
 	int	(*port_vlan_del)(struct dsa_switch *ds, int port,
 				 const struct switchdev_obj_port_vlan *vlan);
+	int	(*vlan_msti_set)(struct dsa_switch *ds,
+				 const struct switchdev_attr *attr);
+
 	/*
 	 * Forwarding database
 	 */
diff --git a/net/dsa/dsa_priv.h b/net/dsa/dsa_priv.h
index 07c0ad52395a..87ec0697e92e 100644
--- a/net/dsa/dsa_priv.h
+++ b/net/dsa/dsa_priv.h
@@ -217,6 +217,7 @@ int dsa_port_vlan_filtering(struct dsa_port *dp, bool
vlan_filtering,
 			    struct netlink_ext_ack *extack);
 bool dsa_port_skip_vlan_configuration(struct dsa_port *dp);
 int dsa_port_ageing_time(struct dsa_port *dp, clock_t ageing_clock);
+int dsa_port_vlan_msti(struct dsa_port *dp, const struct switchdev_attr *attr);
 int dsa_port_mtu_change(struct dsa_port *dp, int new_mtu,
 			bool targeted_match);
 int dsa_port_fdb_add(struct dsa_port *dp, const unsigned char *addr,
diff --git a/net/dsa/port.c b/net/dsa/port.c
index d9da425a17fb..5f45cb7d70ba 100644
--- a/net/dsa/port.c
+++ b/net/dsa/port.c
@@ -778,6 +778,16 @@ int dsa_port_bridge_flags(struct dsa_port *dp,
 	return 0;
 }
 
+int dsa_port_vlan_msti(struct dsa_port *dp, const struct switchdev_attr *attr)
+{
+	struct dsa_switch *ds = dp->ds;
+
+	if (!ds->ops->vlan_msti_set)
+		return -EOPNOTSUPP;
+
+	return ds->ops->vlan_msti_set(ds, attr);
+}
+
 int dsa_port_mtu_change(struct dsa_port *dp, int new_mtu,
 			bool targeted_match)
 {
diff --git a/net/dsa/slave.c b/net/dsa/slave.c
index 089616206b11..c6ffcd782b5a 100644
--- a/net/dsa/slave.c
+++ b/net/dsa/slave.c
@@ -314,6 +314,12 @@ static int dsa_slave_port_attr_set(struct net_device *dev,
const void *ctx,
 
 		ret = dsa_port_bridge_flags(dp, attr->u.brport_flags, extack);
 		break;
+	case SWITCHDEV_ATTR_ID_VLAN_MSTI:
+		if (!dsa_port_offloads_bridge_dev(dp, attr->orig_dev))
+			return -EOPNOTSUPP;
+
+		ret = dsa_port_vlan_msti(dp, attr);
+		break;
 	default:
 		ret = -EOPNOTSUPP;
 		break;
-- 
2.25.1

Add the usual trampoline functionality from the generic DSA layer down
to the drivers for MST state changes.

Signed-off-by: Tobias Waldekranz <tobias at waldekranz.com>
---
 include/net/dsa.h  |  2 ++
 net/dsa/dsa_priv.h |  2 ++
 net/dsa/port.c     | 30 ++++++++++++++++++++++++++++++
 net/dsa/slave.c    |  6 ++++++
 4 files changed, 40 insertions(+)

diff --git a/include/net/dsa.h b/include/net/dsa.h
index cc8acb01bd9b..096e6e3a8e1e 100644
--- a/include/net/dsa.h
+++ b/include/net/dsa.h
@@ -943,6 +943,8 @@ struct dsa_switch_ops {
 				     struct dsa_bridge bridge);
 	void	(*port_stp_state_set)(struct dsa_switch *ds, int port,
 				      u8 state);
+	int	(*port_mst_state_set)(struct dsa_switch *ds, int port,
+				      const struct switchdev_mst_state *state);
 	void	(*port_fast_age)(struct dsa_switch *ds, int port);
 	int	(*port_pre_bridge_flags)(struct dsa_switch *ds, int port,
 					 struct switchdev_brport_flags flags,
diff --git a/net/dsa/dsa_priv.h b/net/dsa/dsa_priv.h
index 87ec0697e92e..a620e079ebc5 100644
--- a/net/dsa/dsa_priv.h
+++ b/net/dsa/dsa_priv.h
@@ -198,6 +198,8 @@ static inline struct net_device
*dsa_master_find_slave(struct net_device *dev,
 void dsa_port_set_tag_protocol(struct dsa_port *cpu_dp,
 			       const struct dsa_device_ops *tag_ops);
 int dsa_port_set_state(struct dsa_port *dp, u8 state, bool do_fast_age);
+int dsa_port_set_mst_state(struct dsa_port *dp,
+			   const struct switchdev_mst_state *state);
 int dsa_port_enable_rt(struct dsa_port *dp, struct phy_device *phy);
 int dsa_port_enable(struct dsa_port *dp, struct phy_device *phy);
 void dsa_port_disable_rt(struct dsa_port *dp);
diff --git a/net/dsa/port.c b/net/dsa/port.c
index 5f45cb7d70ba..26cfbc8ab499 100644
--- a/net/dsa/port.c
+++ b/net/dsa/port.c
@@ -108,6 +108,36 @@ int dsa_port_set_state(struct dsa_port *dp, u8 state, bool
do_fast_age)
 	return 0;
 }
 
+int dsa_port_set_mst_state(struct dsa_port *dp,
+			   const struct switchdev_mst_state *state)
+{
+	struct dsa_switch *ds = dp->ds;
+	int err, port = dp->index;
+
+	if (!ds->ops->port_mst_state_set)
+		return -EOPNOTSUPP;
+
+	err = ds->ops->port_mst_state_set(ds, port, state);
+	if (err)
+		return err;
+
+	if (!dsa_port_can_configure_learning(dp) || dp->learning) {
+		switch (state->state) {
+		case BR_STATE_DISABLED:
+		case BR_STATE_BLOCKING:
+		case BR_STATE_LISTENING:
+			/* Ideally we would only fast age entries
+			 * belonging to VLANs controlled by this
+			 * MST.
+			 */
+			dsa_port_fast_age(dp);
+			break;
+		}
+	}
+
+	return 0;
+}
+
 static void dsa_port_set_state_now(struct dsa_port *dp, u8 state,
 				   bool do_fast_age)
 {
diff --git a/net/dsa/slave.c b/net/dsa/slave.c
index c6ffcd782b5a..32b006a5b778 100644
--- a/net/dsa/slave.c
+++ b/net/dsa/slave.c
@@ -288,6 +288,12 @@ static int dsa_slave_port_attr_set(struct net_device *dev,
const void *ctx,
 
 		ret = dsa_port_set_state(dp, attr->u.stp_state, true);
 		break;
+	case SWITCHDEV_ATTR_ID_PORT_MST_STATE:
+		if (!dsa_port_offloads_bridge_port(dp, attr->orig_dev))
+			return -EOPNOTSUPP;
+
+		ret = dsa_port_set_mst_state(dp, &attr->u.mst_state);
+		break;
 	case SWITCHDEV_ATTR_ID_BRIDGE_VLAN_FILTERING:
 		if (!dsa_port_offloads_bridge_dev(dp, attr->orig_dev))
 			return -EOPNOTSUPP;
-- 
2.25.1

In early LinkStreet silicon (e.g. 6095/6185), the per-VLAN STP states
were kept in the VTU - there was no concept of a SID. Later, the
information was split into two tables, where the VTU only tracked
memberships and deferred the STP state tracking to the STU via a
pointer (SID). This meant that a group of VLANs could share the same
STU entry. Most likely, this was done to align with MSTP (802.1Q-2018,
Clause 13), which is built on this principle.

While the VTU is still 4k lines on most devices, the STU is capped at
64 entries. This means that the current stategy, updating STU info
whenever a VTU entry is updated, can not easily support MSTP because:

- The maximum number of VIDs would also be capped at 64, as we would
  have to allocate one SID for every VTU entry - even if many VLANs
  would effectively share the same MST.

- MSTP updates would be unnecessarily slow as you would have to
  iterate over all VLANs that share the same MST.

In order to support MSTP offloading in the future, manage the STU as a
separate entity from the VTU.

Only add support for newer hardware with separate VTU and
STU. VTU-only devices can also be supported, but essentially this
requires a software implementation of an STU (fanning out state
changed to all VLANs tied to the same MST).

Signed-off-by: Tobias Waldekranz <tobias at waldekranz.com>
---
 drivers/net/dsa/mv88e6xxx/chip.c        |  54 ++++
 drivers/net/dsa/mv88e6xxx/chip.h        |  24 ++
 drivers/net/dsa/mv88e6xxx/global1.h     |  10 +
 drivers/net/dsa/mv88e6xxx/global1_vtu.c | 311 ++++++++++++++----------
 4 files changed, 264 insertions(+), 135 deletions(-)

diff --git a/drivers/net/dsa/mv88e6xxx/chip.c b/drivers/net/dsa/mv88e6xxx/chip.c
index 84b90fc36c58..c14a62aa6a6c 100644
--- a/drivers/net/dsa/mv88e6xxx/chip.c
+++ b/drivers/net/dsa/mv88e6xxx/chip.c
@@ -1791,6 +1791,33 @@ static int mv88e6xxx_atu_new(struct mv88e6xxx_chip *chip,
u16 *fid)
 	return mv88e6xxx_g1_atu_flush(chip, *fid, true);
 }
 
+static int mv88e6xxx_stu_loadpurge(struct mv88e6xxx_chip *chip,
+				   struct mv88e6xxx_stu_entry *entry)
+{
+	if (!chip->info->ops->stu_loadpurge)
+		return -EOPNOTSUPP;
+
+	return chip->info->ops->stu_loadpurge(chip, entry);
+}
+
+static int mv88e6xxx_stu_setup(struct mv88e6xxx_chip *chip)
+{
+	struct mv88e6xxx_stu_entry stu = {
+		.valid = true,
+		.sid = 0
+	};
+
+	if (!mv88e6xxx_has_stu(chip))
+		return 0;
+
+	/* Make sure that SID 0 is always valid. This is used by VTU
+	 * entries that do not make use of the STU, e.g. when creating
+	 * a VLAN upper on a port that is also part of a VLAN
+	 * filtering bridge.
+	 */
+	return mv88e6xxx_stu_loadpurge(chip, &stu);
+}
+
 static int mv88e6xxx_port_check_hw_vlan(struct dsa_switch *ds, int port,
 					u16 vid)
 {
@@ -3427,6 +3454,13 @@ static int mv88e6xxx_setup(struct dsa_switch *ds)
 	if (err)
 		goto unlock;
 
+	/* Must be called after mv88e6xxx_vtu_setup (which flushes the
+	 * VTU, thereby also flushing the STU).
+	 */
+	err = mv88e6xxx_stu_setup(chip);
+	if (err)
+		goto unlock;
+
 	/* Setup Switch Port Registers */
 	for (i = 0; i < mv88e6xxx_num_ports(chip); i++) {
 		if (dsa_is_unused_port(ds, i))
@@ -3882,6 +3916,8 @@ static const struct mv88e6xxx_ops mv88e6097_ops = {
 	.vtu_getnext = mv88e6352_g1_vtu_getnext,
 	.vtu_loadpurge = mv88e6352_g1_vtu_loadpurge,
 	.phylink_get_caps = mv88e6095_phylink_get_caps,
+	.stu_getnext = mv88e6352_g1_stu_getnext,
+	.stu_loadpurge = mv88e6352_g1_stu_loadpurge,
 	.set_max_frame_size = mv88e6185_g1_set_max_frame_size,
 };
 
@@ -4968,6 +5004,8 @@ static const struct mv88e6xxx_ops mv88e6352_ops = {
 	.atu_set_hash = mv88e6165_g1_atu_set_hash,
 	.vtu_getnext = mv88e6352_g1_vtu_getnext,
 	.vtu_loadpurge = mv88e6352_g1_vtu_loadpurge,
+	.stu_getnext = mv88e6352_g1_stu_getnext,
+	.stu_loadpurge = mv88e6352_g1_stu_loadpurge,
 	.serdes_get_lane = mv88e6352_serdes_get_lane,
 	.serdes_pcs_get_state = mv88e6352_serdes_pcs_get_state,
 	.serdes_pcs_config = mv88e6352_serdes_pcs_config,
@@ -5033,6 +5071,8 @@ static const struct mv88e6xxx_ops mv88e6390_ops = {
 	.atu_set_hash = mv88e6165_g1_atu_set_hash,
 	.vtu_getnext = mv88e6390_g1_vtu_getnext,
 	.vtu_loadpurge = mv88e6390_g1_vtu_loadpurge,
+	.stu_getnext = mv88e6390_g1_stu_getnext,
+	.stu_loadpurge = mv88e6390_g1_stu_loadpurge,
 	.serdes_power = mv88e6390_serdes_power,
 	.serdes_get_lane = mv88e6390_serdes_get_lane,
 	/* Check status register pause & lpa register */
@@ -5098,6 +5138,8 @@ static const struct mv88e6xxx_ops mv88e6390x_ops = {
 	.atu_set_hash = mv88e6165_g1_atu_set_hash,
 	.vtu_getnext = mv88e6390_g1_vtu_getnext,
 	.vtu_loadpurge = mv88e6390_g1_vtu_loadpurge,
+	.stu_getnext = mv88e6390_g1_stu_getnext,
+	.stu_loadpurge = mv88e6390_g1_stu_loadpurge,
 	.serdes_power = mv88e6390_serdes_power,
 	.serdes_get_lane = mv88e6390x_serdes_get_lane,
 	.serdes_pcs_get_state = mv88e6390_serdes_pcs_get_state,
@@ -5166,6 +5208,8 @@ static const struct mv88e6xxx_ops mv88e6393x_ops = {
 	.atu_set_hash = mv88e6165_g1_atu_set_hash,
 	.vtu_getnext = mv88e6390_g1_vtu_getnext,
 	.vtu_loadpurge = mv88e6390_g1_vtu_loadpurge,
+	.stu_getnext = mv88e6390_g1_stu_getnext,
+	.stu_loadpurge = mv88e6390_g1_stu_loadpurge,
 	.serdes_power = mv88e6393x_serdes_power,
 	.serdes_get_lane = mv88e6393x_serdes_get_lane,
 	.serdes_pcs_get_state = mv88e6393x_serdes_pcs_get_state,
@@ -5234,6 +5278,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_ports = 11,
 		.num_internal_phys = 8,
 		.max_vid = 4095,
+		.max_sid = 63,
 		.port_base_addr = 0x10,
 		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
@@ -5487,6 +5532,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_internal_phys = 9,
 		.num_gpio = 16,
 		.max_vid = 8191,
+		.max_sid = 63,
 		.port_base_addr = 0x0,
 		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
@@ -5510,6 +5556,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_internal_phys = 9,
 		.num_gpio = 16,
 		.max_vid = 8191,
+		.max_sid = 63,
 		.port_base_addr = 0x0,
 		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
@@ -5532,6 +5579,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_ports = 11,	/* 10 + Z80 */
 		.num_internal_phys = 9,
 		.max_vid = 8191,
+		.max_sid = 63,
 		.port_base_addr = 0x0,
 		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
@@ -5554,6 +5602,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_ports = 11,	/* 10 + Z80 */
 		.num_internal_phys = 9,
 		.max_vid = 8191,
+		.max_sid = 63,
 		.port_base_addr = 0x0,
 		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
@@ -5576,6 +5625,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_ports = 11,	/* 10 + Z80 */
 		.num_internal_phys = 9,
 		.max_vid = 8191,
+		.max_sid = 63,
 		.port_base_addr = 0x0,
 		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
@@ -5815,6 +5865,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_internal_phys = 5,
 		.num_gpio = 15,
 		.max_vid = 4095,
+		.max_sid = 63,
 		.port_base_addr = 0x10,
 		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
@@ -5839,6 +5890,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_internal_phys = 9,
 		.num_gpio = 16,
 		.max_vid = 8191,
+		.max_sid = 63,
 		.port_base_addr = 0x0,
 		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
@@ -5863,6 +5915,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_internal_phys = 9,
 		.num_gpio = 16,
 		.max_vid = 8191,
+		.max_sid = 63,
 		.port_base_addr = 0x0,
 		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
@@ -5886,6 +5939,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_ports = 11,	/* 10 + Z80 */
 		.num_internal_phys = 9,
 		.max_vid = 8191,
+		.max_sid = 63,
 		.port_base_addr = 0x0,
 		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
diff --git a/drivers/net/dsa/mv88e6xxx/chip.h b/drivers/net/dsa/mv88e6xxx/chip.h
index 30b92a265613..be654be69982 100644
--- a/drivers/net/dsa/mv88e6xxx/chip.h
+++ b/drivers/net/dsa/mv88e6xxx/chip.h
@@ -20,6 +20,7 @@
 
 #define EDSA_HLEN		8
 #define MV88E6XXX_N_FID		4096
+#define MV88E6XXX_N_SID		64
 
 #define MV88E6XXX_FID_STANDALONE	0
 #define MV88E6XXX_FID_BRIDGED		1
@@ -130,6 +131,7 @@ struct mv88e6xxx_info {
 	unsigned int num_internal_phys;
 	unsigned int num_gpio;
 	unsigned int max_vid;
+	unsigned int max_sid;
 	unsigned int port_base_addr;
 	unsigned int phy_base_addr;
 	unsigned int global1_addr;
@@ -181,6 +183,12 @@ struct mv88e6xxx_vtu_entry {
 	bool	valid;
 	bool	policy;
 	u8	member[DSA_MAX_PORTS];
+	u8	state[DSA_MAX_PORTS];	/* Older silicon has no STU */
+};
+
+struct mv88e6xxx_stu_entry {
+	u8	sid;
+	bool	valid;
 	u8	state[DSA_MAX_PORTS];
 };
 
@@ -602,6 +610,12 @@ struct mv88e6xxx_ops {
 	int (*vtu_loadpurge)(struct mv88e6xxx_chip *chip,
 			     struct mv88e6xxx_vtu_entry *entry);
 
+	/* Spanning Tree Unit operations */
+	int (*stu_getnext)(struct mv88e6xxx_chip *chip,
+			   struct mv88e6xxx_stu_entry *entry);
+	int (*stu_loadpurge)(struct mv88e6xxx_chip *chip,
+			     struct mv88e6xxx_stu_entry *entry);
+
 	/* GPIO operations */
 	const struct mv88e6xxx_gpio_ops *gpio_ops;
 
@@ -700,6 +714,11 @@ struct mv88e6xxx_hw_stat {
 	int type;
 };
 
+static inline bool mv88e6xxx_has_stu(struct mv88e6xxx_chip *chip)
+{
+	return chip->info->max_sid > 0;
+}
+
 static inline bool mv88e6xxx_has_pvt(struct mv88e6xxx_chip *chip)
 {
 	return chip->info->pvt;
@@ -730,6 +749,11 @@ static inline unsigned int mv88e6xxx_max_vid(struct
mv88e6xxx_chip *chip)
 	return chip->info->max_vid;
 }
 
+static inline unsigned int mv88e6xxx_max_sid(struct mv88e6xxx_chip *chip)
+{
+	return chip->info->max_sid;
+}
+
 static inline u16 mv88e6xxx_port_mask(struct mv88e6xxx_chip *chip)
 {
 	return GENMASK((s32)mv88e6xxx_num_ports(chip) - 1, 0);
diff --git a/drivers/net/dsa/mv88e6xxx/global1.h
b/drivers/net/dsa/mv88e6xxx/global1.h
index 2c1607c858a1..65958b2a0d3a 100644
--- a/drivers/net/dsa/mv88e6xxx/global1.h
+++ b/drivers/net/dsa/mv88e6xxx/global1.h
@@ -348,6 +348,16 @@ int mv88e6390_g1_vtu_getnext(struct mv88e6xxx_chip *chip,
 int mv88e6390_g1_vtu_loadpurge(struct mv88e6xxx_chip *chip,
 			       struct mv88e6xxx_vtu_entry *entry);
 int mv88e6xxx_g1_vtu_flush(struct mv88e6xxx_chip *chip);
+int mv88e6xxx_g1_stu_getnext(struct mv88e6xxx_chip *chip,
+			     struct mv88e6xxx_stu_entry *entry);
+int mv88e6352_g1_stu_getnext(struct mv88e6xxx_chip *chip,
+			     struct mv88e6xxx_stu_entry *entry);
+int mv88e6352_g1_stu_loadpurge(struct mv88e6xxx_chip *chip,
+			       struct mv88e6xxx_stu_entry *entry);
+int mv88e6390_g1_stu_getnext(struct mv88e6xxx_chip *chip,
+			     struct mv88e6xxx_stu_entry *entry);
+int mv88e6390_g1_stu_loadpurge(struct mv88e6xxx_chip *chip,
+			       struct mv88e6xxx_stu_entry *entry);
 int mv88e6xxx_g1_vtu_prob_irq_setup(struct mv88e6xxx_chip *chip);
 void mv88e6xxx_g1_vtu_prob_irq_free(struct mv88e6xxx_chip *chip);
 int mv88e6xxx_g1_atu_get_next(struct mv88e6xxx_chip *chip, u16 fid);
diff --git a/drivers/net/dsa/mv88e6xxx/global1_vtu.c
b/drivers/net/dsa/mv88e6xxx/global1_vtu.c
index b1bd9274a562..38e18f5811bf 100644
--- a/drivers/net/dsa/mv88e6xxx/global1_vtu.c
+++ b/drivers/net/dsa/mv88e6xxx/global1_vtu.c
@@ -44,8 +44,7 @@ static int mv88e6xxx_g1_vtu_fid_write(struct mv88e6xxx_chip
*chip,
 
 /* Offset 0x03: VTU SID Register */
 
-static int mv88e6xxx_g1_vtu_sid_read(struct mv88e6xxx_chip *chip,
-				     struct mv88e6xxx_vtu_entry *entry)
+static int mv88e6xxx_g1_vtu_sid_read(struct mv88e6xxx_chip *chip, u8 *sid)
 {
 	u16 val;
 	int err;
@@ -54,15 +53,14 @@ static int mv88e6xxx_g1_vtu_sid_read(struct mv88e6xxx_chip
*chip,
 	if (err)
 		return err;
 
-	entry->sid = val & MV88E6352_G1_VTU_SID_MASK;
+	*sid = val & MV88E6352_G1_VTU_SID_MASK;
 
 	return 0;
 }
 
-static int mv88e6xxx_g1_vtu_sid_write(struct mv88e6xxx_chip *chip,
-				      struct mv88e6xxx_vtu_entry *entry)
+static int mv88e6xxx_g1_vtu_sid_write(struct mv88e6xxx_chip *chip, u8 sid)
 {
-	u16 val = entry->sid & MV88E6352_G1_VTU_SID_MASK;
+	u16 val = sid & MV88E6352_G1_VTU_SID_MASK;
 
 	return mv88e6xxx_g1_write(chip, MV88E6352_G1_VTU_SID, val);
 }
@@ -91,7 +89,7 @@ static int mv88e6xxx_g1_vtu_op(struct mv88e6xxx_chip *chip,
u16 op)
 /* Offset 0x06: VTU VID Register */
 
 static int mv88e6xxx_g1_vtu_vid_read(struct mv88e6xxx_chip *chip,
-				     struct mv88e6xxx_vtu_entry *entry)
+				     bool *valid, u16 *vid)
 {
 	u16 val;
 	int err;
@@ -100,25 +98,28 @@ static int mv88e6xxx_g1_vtu_vid_read(struct mv88e6xxx_chip
*chip,
 	if (err)
 		return err;
 
-	entry->vid = val & 0xfff;
+	if (vid) {
+		*vid = val & 0xfff;
 
-	if (val & MV88E6390_G1_VTU_VID_PAGE)
-		entry->vid |= 0x1000;
+		if (val & MV88E6390_G1_VTU_VID_PAGE)
+			*vid |= 0x1000;
+	}
 
-	entry->valid = !!(val & MV88E6XXX_G1_VTU_VID_VALID);
+	if (valid)
+		*valid = !!(val & MV88E6XXX_G1_VTU_VID_VALID);
 
 	return 0;
 }
 
 static int mv88e6xxx_g1_vtu_vid_write(struct mv88e6xxx_chip *chip,
-				      struct mv88e6xxx_vtu_entry *entry)
+				      bool valid, u16 vid)
 {
-	u16 val = entry->vid & 0xfff;
+	u16 val = vid & 0xfff;
 
-	if (entry->vid & 0x1000)
+	if (vid & 0x1000)
 		val |= MV88E6390_G1_VTU_VID_PAGE;
 
-	if (entry->valid)
+	if (valid)
 		val |= MV88E6XXX_G1_VTU_VID_VALID;
 
 	return mv88e6xxx_g1_write(chip, MV88E6XXX_G1_VTU_VID, val);
@@ -147,7 +148,7 @@ static int mv88e6185_g1_vtu_stu_data_read(struct
mv88e6xxx_chip *chip,
 }
 
 static int mv88e6185_g1_vtu_data_read(struct mv88e6xxx_chip *chip,
-				      struct mv88e6xxx_vtu_entry *entry)
+				      u8 *member, u8 *state)
 {
 	u16 regs[3];
 	int err;
@@ -160,36 +161,20 @@ static int mv88e6185_g1_vtu_data_read(struct
mv88e6xxx_chip *chip,
 	/* Extract MemberTag data */
 	for (i = 0; i < mv88e6xxx_num_ports(chip); ++i) {
 		unsigned int member_offset = (i % 4) * 4;
+		unsigned int state_offset = member_offset + 2;
 
-		entry->member[i] = (regs[i / 4] >> member_offset) & 0x3;
-	}
-
-	return 0;
-}
-
-static int mv88e6185_g1_stu_data_read(struct mv88e6xxx_chip *chip,
-				      struct mv88e6xxx_vtu_entry *entry)
-{
-	u16 regs[3];
-	int err;
-	int i;
-
-	err = mv88e6185_g1_vtu_stu_data_read(chip, regs);
-	if (err)
-		return err;
+		if (member)
+			member[i] = (regs[i / 4] >> member_offset) & 0x3;
 
-	/* Extract PortState data */
-	for (i = 0; i < mv88e6xxx_num_ports(chip); ++i) {
-		unsigned int state_offset = (i % 4) * 4 + 2;
-
-		entry->state[i] = (regs[i / 4] >> state_offset) & 0x3;
+		if (state)
+			state[i] = (regs[i / 4] >> state_offset) & 0x3;
 	}
 
 	return 0;
 }
 
 static int mv88e6185_g1_vtu_data_write(struct mv88e6xxx_chip *chip,
-				       struct mv88e6xxx_vtu_entry *entry)
+				       u8 *member, u8 *state)
 {
 	u16 regs[3] = { 0 };
 	int i;
@@ -199,8 +184,11 @@ static int mv88e6185_g1_vtu_data_write(struct
mv88e6xxx_chip *chip,
 		unsigned int member_offset = (i % 4) * 4;
 		unsigned int state_offset = member_offset + 2;
 
-		regs[i / 4] |= (entry->member[i] & 0x3) << member_offset;
-		regs[i / 4] |= (entry->state[i] & 0x3) << state_offset;
+		if (member)
+			regs[i / 4] |= (member[i] & 0x3) << member_offset;
+
+		if (state)
+			regs[i / 4] |= (state[i] & 0x3) << state_offset;
 	}
 
 	/* Write all 3 VTU/STU Data registers */
@@ -268,48 +256,6 @@ static int mv88e6390_g1_vtu_data_write(struct
mv88e6xxx_chip *chip, u8 *data)
 
 /* VLAN Translation Unit Operations */
 
-static int mv88e6xxx_g1_vtu_stu_getnext(struct mv88e6xxx_chip *chip,
-					struct mv88e6xxx_vtu_entry *entry)
-{
-	int err;
-
-	err = mv88e6xxx_g1_vtu_sid_write(chip, entry);
-	if (err)
-		return err;
-
-	err = mv88e6xxx_g1_vtu_op(chip, MV88E6XXX_G1_VTU_OP_STU_GET_NEXT);
-	if (err)
-		return err;
-
-	err = mv88e6xxx_g1_vtu_sid_read(chip, entry);
-	if (err)
-		return err;
-
-	return mv88e6xxx_g1_vtu_vid_read(chip, entry);
-}
-
-static int mv88e6xxx_g1_vtu_stu_get(struct mv88e6xxx_chip *chip,
-				    struct mv88e6xxx_vtu_entry *vtu)
-{
-	struct mv88e6xxx_vtu_entry stu;
-	int err;
-
-	err = mv88e6xxx_g1_vtu_sid_read(chip, vtu);
-	if (err)
-		return err;
-
-	stu.sid = vtu->sid - 1;
-
-	err = mv88e6xxx_g1_vtu_stu_getnext(chip, &stu);
-	if (err)
-		return err;
-
-	if (stu.sid != vtu->sid || !stu.valid)
-		return -EINVAL;
-
-	return 0;
-}
-
 int mv88e6xxx_g1_vtu_getnext(struct mv88e6xxx_chip *chip,
 			     struct mv88e6xxx_vtu_entry *entry)
 {
@@ -327,7 +273,7 @@ int mv88e6xxx_g1_vtu_getnext(struct mv88e6xxx_chip *chip,
 	 * write the VID only once, when the entry is given as invalid.
 	 */
 	if (!entry->valid) {
-		err = mv88e6xxx_g1_vtu_vid_write(chip, entry);
+		err = mv88e6xxx_g1_vtu_vid_write(chip, false, entry->vid);
 		if (err)
 			return err;
 	}
@@ -336,7 +282,7 @@ int mv88e6xxx_g1_vtu_getnext(struct mv88e6xxx_chip *chip,
 	if (err)
 		return err;
 
-	return mv88e6xxx_g1_vtu_vid_read(chip, entry);
+	return mv88e6xxx_g1_vtu_vid_read(chip, &entry->valid,
&entry->vid);
 }
 
 int mv88e6185_g1_vtu_getnext(struct mv88e6xxx_chip *chip,
@@ -350,11 +296,7 @@ int mv88e6185_g1_vtu_getnext(struct mv88e6xxx_chip *chip,
 		return err;
 
 	if (entry->valid) {
-		err = mv88e6185_g1_vtu_data_read(chip, entry);
-		if (err)
-			return err;
-
-		err = mv88e6185_g1_stu_data_read(chip, entry);
+		err = mv88e6185_g1_vtu_data_read(chip, entry->member, entry->state);
 		if (err)
 			return err;
 
@@ -384,7 +326,7 @@ int mv88e6352_g1_vtu_getnext(struct mv88e6xxx_chip *chip,
 		return err;
 
 	if (entry->valid) {
-		err = mv88e6185_g1_vtu_data_read(chip, entry);
+		err = mv88e6185_g1_vtu_data_read(chip, entry->member, NULL);
 		if (err)
 			return err;
 
@@ -392,12 +334,7 @@ int mv88e6352_g1_vtu_getnext(struct mv88e6xxx_chip *chip,
 		if (err)
 			return err;
 
-		/* Fetch VLAN PortState data from the STU */
-		err = mv88e6xxx_g1_vtu_stu_get(chip, entry);
-		if (err)
-			return err;
-
-		err = mv88e6185_g1_stu_data_read(chip, entry);
+		err = mv88e6xxx_g1_vtu_sid_read(chip, &entry->sid);
 		if (err)
 			return err;
 	}
@@ -420,16 +357,11 @@ int mv88e6390_g1_vtu_getnext(struct mv88e6xxx_chip *chip,
 		if (err)
 			return err;
 
-		/* Fetch VLAN PortState data from the STU */
-		err = mv88e6xxx_g1_vtu_stu_get(chip, entry);
-		if (err)
-			return err;
-
-		err = mv88e6390_g1_vtu_data_read(chip, entry->state);
+		err = mv88e6xxx_g1_vtu_fid_read(chip, entry);
 		if (err)
 			return err;
 
-		err = mv88e6xxx_g1_vtu_fid_read(chip, entry);
+		err = mv88e6xxx_g1_vtu_sid_read(chip, &entry->sid);
 		if (err)
 			return err;
 	}
@@ -447,12 +379,12 @@ int mv88e6185_g1_vtu_loadpurge(struct mv88e6xxx_chip
*chip,
 	if (err)
 		return err;
 
-	err = mv88e6xxx_g1_vtu_vid_write(chip, entry);
+	err = mv88e6xxx_g1_vtu_vid_write(chip, entry->valid, entry->vid);
 	if (err)
 		return err;
 
 	if (entry->valid) {
-		err = mv88e6185_g1_vtu_data_write(chip, entry);
+		err = mv88e6185_g1_vtu_data_write(chip, entry->member, entry->state);
 		if (err)
 			return err;
 
@@ -479,27 +411,21 @@ int mv88e6352_g1_vtu_loadpurge(struct mv88e6xxx_chip
*chip,
 	if (err)
 		return err;
 
-	err = mv88e6xxx_g1_vtu_vid_write(chip, entry);
+	err = mv88e6xxx_g1_vtu_vid_write(chip, entry->valid, entry->vid);
 	if (err)
 		return err;
 
 	if (entry->valid) {
-		/* Write MemberTag and PortState data */
-		err = mv88e6185_g1_vtu_data_write(chip, entry);
-		if (err)
-			return err;
-
-		err = mv88e6xxx_g1_vtu_sid_write(chip, entry);
+		/* Write MemberTag data */
+		err = mv88e6185_g1_vtu_data_write(chip, entry->member, NULL);
 		if (err)
 			return err;
 
-		/* Load STU entry */
-		err = mv88e6xxx_g1_vtu_op(chip,
-					  MV88E6XXX_G1_VTU_OP_STU_LOAD_PURGE);
+		err = mv88e6xxx_g1_vtu_fid_write(chip, entry);
 		if (err)
 			return err;
 
-		err = mv88e6xxx_g1_vtu_fid_write(chip, entry);
+		err = mv88e6xxx_g1_vtu_sid_write(chip, entry->sid);
 		if (err)
 			return err;
 	}
@@ -517,41 +443,113 @@ int mv88e6390_g1_vtu_loadpurge(struct mv88e6xxx_chip
*chip,
 	if (err)
 		return err;
 
-	err = mv88e6xxx_g1_vtu_vid_write(chip, entry);
+	err = mv88e6xxx_g1_vtu_vid_write(chip, entry->valid, entry->vid);
 	if (err)
 		return err;
 
 	if (entry->valid) {
-		/* Write PortState data */
-		err = mv88e6390_g1_vtu_data_write(chip, entry->state);
+		/* Write MemberTag data */
+		err = mv88e6390_g1_vtu_data_write(chip, entry->member);
 		if (err)
 			return err;
 
-		err = mv88e6xxx_g1_vtu_sid_write(chip, entry);
+		err = mv88e6xxx_g1_vtu_fid_write(chip, entry);
 		if (err)
 			return err;
 
-		/* Load STU entry */
-		err = mv88e6xxx_g1_vtu_op(chip,
-					  MV88E6XXX_G1_VTU_OP_STU_LOAD_PURGE);
+		err = mv88e6xxx_g1_vtu_sid_write(chip, entry->sid);
 		if (err)
 			return err;
+	}
 
-		/* Write MemberTag data */
-		err = mv88e6390_g1_vtu_data_write(chip, entry->member);
+	/* Load/Purge VTU entry */
+	return mv88e6xxx_g1_vtu_op(chip, MV88E6XXX_G1_VTU_OP_VTU_LOAD_PURGE);
+}
+
+int mv88e6xxx_g1_vtu_flush(struct mv88e6xxx_chip *chip)
+{
+	int err;
+
+	err = mv88e6xxx_g1_vtu_op_wait(chip);
+	if (err)
+		return err;
+
+	return mv88e6xxx_g1_vtu_op(chip, MV88E6XXX_G1_VTU_OP_FLUSH_ALL);
+}
+
+/* Spanning Tree Unit Operations */
+
+int mv88e6xxx_g1_stu_getnext(struct mv88e6xxx_chip *chip,
+			     struct mv88e6xxx_stu_entry *entry)
+{
+	int err;
+
+	err = mv88e6xxx_g1_vtu_op_wait(chip);
+	if (err)
+		return err;
+
+	/* To get the next higher active SID, the STU GetNext operation can be
+	 * started again without setting the SID registers since it already
+	 * contains the last SID.
+	 *
+	 * To save a few hardware accesses and abstract this to the caller,
+	 * write the SID only once, when the entry is given as invalid.
+	 */
+	if (!entry->valid) {
+		err = mv88e6xxx_g1_vtu_sid_write(chip, entry->sid);
 		if (err)
 			return err;
+	}
 
-		err = mv88e6xxx_g1_vtu_fid_write(chip, entry);
+	err = mv88e6xxx_g1_vtu_op(chip, MV88E6XXX_G1_VTU_OP_STU_GET_NEXT);
+	if (err)
+		return err;
+
+	err = mv88e6xxx_g1_vtu_vid_read(chip, &entry->valid, NULL);
+	if (err)
+		return err;
+
+	if (entry->valid) {
+		err = mv88e6xxx_g1_vtu_sid_read(chip, &entry->sid);
 		if (err)
 			return err;
 	}
 
-	/* Load/Purge VTU entry */
-	return mv88e6xxx_g1_vtu_op(chip, MV88E6XXX_G1_VTU_OP_VTU_LOAD_PURGE);
+	return 0;
 }
 
-int mv88e6xxx_g1_vtu_flush(struct mv88e6xxx_chip *chip)
+int mv88e6352_g1_stu_getnext(struct mv88e6xxx_chip *chip,
+			     struct mv88e6xxx_stu_entry *entry)
+{
+	int err;
+
+	err = mv88e6xxx_g1_stu_getnext(chip, entry);
+	if (err)
+		return err;
+
+	if (!entry->valid)
+		return 0;
+
+	return mv88e6185_g1_vtu_data_read(chip, NULL, entry->state);
+}
+
+int mv88e6390_g1_stu_getnext(struct mv88e6xxx_chip *chip,
+			     struct mv88e6xxx_stu_entry *entry)
+{
+	int err;
+
+	err = mv88e6xxx_g1_stu_getnext(chip, entry);
+	if (err)
+		return err;
+
+	if (!entry->valid)
+		return 0;
+
+	return mv88e6390_g1_vtu_data_read(chip, entry->state);
+}
+
+int mv88e6352_g1_stu_loadpurge(struct mv88e6xxx_chip *chip,
+			       struct mv88e6xxx_stu_entry *entry)
 {
 	int err;
 
@@ -559,16 +557,59 @@ int mv88e6xxx_g1_vtu_flush(struct mv88e6xxx_chip *chip)
 	if (err)
 		return err;
 
-	return mv88e6xxx_g1_vtu_op(chip, MV88E6XXX_G1_VTU_OP_FLUSH_ALL);
+	err = mv88e6xxx_g1_vtu_vid_write(chip, entry->valid, 0);
+	if (err)
+		return err;
+
+	err = mv88e6xxx_g1_vtu_sid_write(chip, entry->sid);
+	if (err)
+		return err;
+
+	if (entry->valid) {
+		err = mv88e6185_g1_vtu_data_write(chip, NULL, entry->state);
+		if (err)
+			return err;
+	}
+
+	/* Load/Purge STU entry */
+	return mv88e6xxx_g1_vtu_op(chip, MV88E6XXX_G1_VTU_OP_STU_LOAD_PURGE);
+}
+
+int mv88e6390_g1_stu_loadpurge(struct mv88e6xxx_chip *chip,
+			       struct mv88e6xxx_stu_entry *entry)
+{
+	int err;
+
+	err = mv88e6xxx_g1_vtu_op_wait(chip);
+	if (err)
+		return err;
+
+	err = mv88e6xxx_g1_vtu_vid_write(chip, entry->valid, 0);
+	if (err)
+		return err;
+
+	err = mv88e6xxx_g1_vtu_sid_write(chip, entry->sid);
+	if (err)
+		return err;
+
+	if (entry->valid) {
+		err = mv88e6390_g1_vtu_data_write(chip, entry->state);
+		if (err)
+			return err;
+	}
+
+	/* Load/Purge STU entry */
+	return mv88e6xxx_g1_vtu_op(chip, MV88E6XXX_G1_VTU_OP_STU_LOAD_PURGE);
 }
 
+/* VTU Violation Management */
+
 static irqreturn_t mv88e6xxx_g1_vtu_prob_irq_thread_fn(int irq, void *dev_id)
 {
 	struct mv88e6xxx_chip *chip = dev_id;
-	struct mv88e6xxx_vtu_entry entry;
+	u16 val, vid;
 	int spid;
 	int err;
-	u16 val;
 
 	mv88e6xxx_reg_lock(chip);
 
@@ -580,7 +621,7 @@ static irqreturn_t mv88e6xxx_g1_vtu_prob_irq_thread_fn(int
irq, void *dev_id)
 	if (err)
 		goto out;
 
-	err = mv88e6xxx_g1_vtu_vid_read(chip, &entry);
+	err = mv88e6xxx_g1_vtu_vid_read(chip, NULL, &vid);
 	if (err)
 		goto out;
 
@@ -588,13 +629,13 @@ static irqreturn_t mv88e6xxx_g1_vtu_prob_irq_thread_fn(int
irq, void *dev_id)
 
 	if (val & MV88E6XXX_G1_VTU_OP_MEMBER_VIOLATION) {
 		dev_err_ratelimited(chip->dev, "VTU member violation for vid %d,
source port %d\n",
-				    entry.vid, spid);
+				    vid, spid);
 		chip->ports[spid].vtu_member_violation++;
 	}
 
 	if (val & MV88E6XXX_G1_VTU_OP_MISS_VIOLATION) {
 		dev_dbg_ratelimited(chip->dev, "VTU miss violation for vid %d, source
port %d\n",
-				    entry.vid, spid);
+				    vid, spid);
 		chip->ports[spid].vtu_miss_violation++;
 	}
 
-- 
2.25.1

Export the raw STU data in a devlink region so that it can be
inspected from userspace and compared to the current bridge
configuration.

Signed-off-by: Tobias Waldekranz <tobias at waldekranz.com>
---
 drivers/net/dsa/mv88e6xxx/chip.h    |  1 +
 drivers/net/dsa/mv88e6xxx/devlink.c | 94 +++++++++++++++++++++++++++++
 2 files changed, 95 insertions(+)

diff --git a/drivers/net/dsa/mv88e6xxx/chip.h b/drivers/net/dsa/mv88e6xxx/chip.h
index be654be69982..6d4daa24d3e5 100644
--- a/drivers/net/dsa/mv88e6xxx/chip.h
+++ b/drivers/net/dsa/mv88e6xxx/chip.h
@@ -287,6 +287,7 @@ enum mv88e6xxx_region_id {
 	MV88E6XXX_REGION_GLOBAL2,
 	MV88E6XXX_REGION_ATU,
 	MV88E6XXX_REGION_VTU,
+	MV88E6XXX_REGION_STU,
 	MV88E6XXX_REGION_PVT,
 
 	_MV88E6XXX_REGION_MAX,
diff --git a/drivers/net/dsa/mv88e6xxx/devlink.c
b/drivers/net/dsa/mv88e6xxx/devlink.c
index 381068395c63..1266eabee086 100644
--- a/drivers/net/dsa/mv88e6xxx/devlink.c
+++ b/drivers/net/dsa/mv88e6xxx/devlink.c
@@ -503,6 +503,85 @@ static int mv88e6xxx_region_vtu_snapshot(struct devlink
*dl,
 	return 0;
 }
 
+/**
+ * struct mv88e6xxx_devlink_stu_entry - Devlink STU entry
+ * @sid:   Global1/3:   SID, unknown filters and learning.
+ * @vid:   Global1/6:   Valid bit.
+ * @data:  Global1/7-9: Membership data and priority override.
+ * @resvd: Reserved. In case we forgot something.
+ *
+ * The STU entry format varies between chipset generations. Peridot
+ * and Amethyst packs the STU data into Global1/7-8. Older silicon
+ * spreads the information across all three VTU data registers -
+ * inheriting the layout of even older hardware that had no STU at
+ * all. Since this is a low-level debug interface, copy all data
+ * verbatim and defer parsing to the consumer.
+ */
+struct mv88e6xxx_devlink_stu_entry {
+	u16 sid;
+	u16 vid;
+	u16 data[3];
+	u16 resvd;
+};
+
+static int mv88e6xxx_region_stu_snapshot(struct devlink *dl,
+					 const struct devlink_region_ops *ops,
+					 struct netlink_ext_ack *extack,
+					 u8 **data)
+{
+	struct mv88e6xxx_devlink_stu_entry *table, *entry;
+	struct dsa_switch *ds = dsa_devlink_to_ds(dl);
+	struct mv88e6xxx_chip *chip = ds->priv;
+	struct mv88e6xxx_stu_entry stu;
+	int err;
+
+	table = kcalloc(mv88e6xxx_max_sid(chip) + 1,
+			sizeof(struct mv88e6xxx_devlink_stu_entry),
+			GFP_KERNEL);
+	if (!table)
+		return -ENOMEM;
+
+	entry = table;
+	stu.sid = mv88e6xxx_max_sid(chip);
+	stu.valid = false;
+
+	mv88e6xxx_reg_lock(chip);
+
+	do {
+		err = mv88e6xxx_g1_stu_getnext(chip, &stu);
+		if (err)
+			break;
+
+		if (!stu.valid)
+			break;
+
+		err = err ? : mv88e6xxx_g1_read(chip, MV88E6352_G1_VTU_SID,
+						&entry->sid);
+		err = err ? : mv88e6xxx_g1_read(chip, MV88E6XXX_G1_VTU_VID,
+						&entry->vid);
+		err = err ? : mv88e6xxx_g1_read(chip, MV88E6XXX_G1_VTU_DATA1,
+						&entry->data[0]);
+		err = err ? : mv88e6xxx_g1_read(chip, MV88E6XXX_G1_VTU_DATA2,
+						&entry->data[1]);
+		err = err ? : mv88e6xxx_g1_read(chip, MV88E6XXX_G1_VTU_DATA3,
+						&entry->data[2]);
+		if (err)
+			break;
+
+		entry++;
+	} while (stu.sid < mv88e6xxx_max_sid(chip));
+
+	mv88e6xxx_reg_unlock(chip);
+
+	if (err) {
+		kfree(table);
+		return err;
+	}
+
+	*data = (u8 *)table;
+	return 0;
+}
+
 static int mv88e6xxx_region_pvt_snapshot(struct devlink *dl,
 					 const struct devlink_region_ops *ops,
 					 struct netlink_ext_ack *extack,
@@ -605,6 +684,12 @@ static struct devlink_region_ops mv88e6xxx_region_vtu_ops =
{
 	.destructor = kfree,
 };
 
+static struct devlink_region_ops mv88e6xxx_region_stu_ops = {
+	.name = "stu",
+	.snapshot = mv88e6xxx_region_stu_snapshot,
+	.destructor = kfree,
+};
+
 static struct devlink_region_ops mv88e6xxx_region_pvt_ops = {
 	.name = "pvt",
 	.snapshot = mv88e6xxx_region_pvt_snapshot,
@@ -640,6 +725,11 @@ static struct mv88e6xxx_region mv88e6xxx_regions[] = {
 		.ops = &mv88e6xxx_region_vtu_ops
 	  /* calculated at runtime */
 	},
+	[MV88E6XXX_REGION_STU] = {
+		.ops = &mv88e6xxx_region_stu_ops,
+		.cond = mv88e6xxx_has_stu,
+	  /* calculated at runtime */
+	},
 	[MV88E6XXX_REGION_PVT] = {
 		.ops = &mv88e6xxx_region_pvt_ops,
 		.size = MV88E6XXX_MAX_PVT_ENTRIES * sizeof(u16),
@@ -706,6 +796,10 @@ int mv88e6xxx_setup_devlink_regions_global(struct
dsa_switch *ds)
 			size = (mv88e6xxx_max_vid(chip) + 1) *
 				sizeof(struct mv88e6xxx_devlink_vtu_entry);
 			break;
+		case MV88E6XXX_REGION_STU:
+			size = (mv88e6xxx_max_sid(chip) + 1) *
+				sizeof(struct mv88e6xxx_devlink_stu_entry);
+			break;
 		}
 
 		region = dsa_devlink_region_create(ds, ops, 1, size);
-- 
2.25.1

Allocate a SID in the STU for each MSTID in use by a bridge and handle
the mapping of MSTIDs to VLANs using the SID field of each VTU entry.

Signed-off-by: Tobias Waldekranz <tobias at waldekranz.com>
---
 drivers/net/dsa/mv88e6xxx/chip.c | 178 +++++++++++++++++++++++++++++++
 drivers/net/dsa/mv88e6xxx/chip.h |  13 +++
 2 files changed, 191 insertions(+)

diff --git a/drivers/net/dsa/mv88e6xxx/chip.c b/drivers/net/dsa/mv88e6xxx/chip.c
index c14a62aa6a6c..4fb4ec1dff79 100644
--- a/drivers/net/dsa/mv88e6xxx/chip.c
+++ b/drivers/net/dsa/mv88e6xxx/chip.c
@@ -1818,6 +1818,137 @@ static int mv88e6xxx_stu_setup(struct mv88e6xxx_chip
*chip)
 	return mv88e6xxx_stu_loadpurge(chip, &stu);
 }
 
+static int mv88e6xxx_sid_new(struct mv88e6xxx_chip *chip, u8 *sid)
+{
+	DECLARE_BITMAP(busy, MV88E6XXX_N_SID) = { 0 };
+	struct mv88e6xxx_mst *mst;
+
+	set_bit(0, busy);
+
+	list_for_each_entry(mst, &chip->msts, node) {
+		set_bit(mst->stu.sid, busy);
+	}
+
+	*sid = find_first_zero_bit(busy, MV88E6XXX_N_SID);
+
+	return (*sid >= mv88e6xxx_max_sid(chip)) ? -ENOSPC : 0;
+}
+
+static int mv88e6xxx_sid_put(struct mv88e6xxx_chip *chip, u8 sid)
+{
+	struct mv88e6xxx_mst *mst, *tmp;
+	int err = 0;
+
+	list_for_each_entry_safe(mst, tmp, &chip->msts, node) {
+		if (mst->stu.sid == sid) {
+			if (refcount_dec_and_test(&mst->refcnt)) {
+				mst->stu.valid = false;
+				err = mv88e6xxx_stu_loadpurge(chip, &mst->stu);
+				list_del(&mst->node);
+				kfree(mst);
+			}
+
+			return err;
+		}
+	}
+
+	return -ENOENT;
+}
+
+static int mv88e6xxx_sid_get(struct mv88e6xxx_chip *chip, struct net_device
*br,
+			     u16 msti, u8 *sid)
+{
+	struct mv88e6xxx_mst *mst;
+	int err, i;
+
+	if (!br)
+		return 0;
+
+	if (!mv88e6xxx_has_stu(chip))
+		return -EOPNOTSUPP;
+
+	list_for_each_entry(mst, &chip->msts, node) {
+		if (mst->br == br && mst->msti == msti) {
+			refcount_inc(&mst->refcnt);
+			*sid = mst->stu.sid;
+			return 0;
+		}
+	}
+
+	err = mv88e6xxx_sid_new(chip, sid);
+	if (err)
+		return err;
+
+	mst = kzalloc(sizeof(*mst), GFP_KERNEL);
+	if (!mst)
+		return -ENOMEM;
+
+	INIT_LIST_HEAD(&mst->node);
+	refcount_set(&mst->refcnt, 1);
+	mst->br = br;
+	mst->msti = msti;
+	mst->stu.valid = true;
+	mst->stu.sid = *sid;
+
+	/* The bridge starts out all ports in the disabled state. But
+	 * a STU state of disabled means to go by the port-global
+	 * state. So we set all user port's initial state to blocking,
+	 * to match the bridge's behavior.
+	 */
+	for (i = 0; i < mv88e6xxx_num_ports(chip); i++)
+		mst->stu.state[i] = dsa_is_user_port(chip->ds, i) ?
+			MV88E6XXX_PORT_CTL0_STATE_BLOCKING :
+			MV88E6XXX_PORT_CTL0_STATE_DISABLED;
+
+	list_add_tail(&mst->node, &chip->msts);
+	return mv88e6xxx_stu_loadpurge(chip, &mst->stu);
+}
+
+static int mv88e6xxx_port_mst_state_set(struct dsa_switch *ds, int port,
+					const struct switchdev_mst_state *st)
+{
+	struct dsa_port *dp = dsa_to_port(ds, port);
+	struct mv88e6xxx_chip *chip = ds->priv;
+	struct mv88e6xxx_mst *mst;
+	u8 state;
+	int err;
+
+	if (!mv88e6xxx_has_stu(chip))
+		return -EOPNOTSUPP;
+
+	switch (st->state) {
+	case BR_STATE_DISABLED:
+	case BR_STATE_BLOCKING:
+	case BR_STATE_LISTENING:
+		state = MV88E6XXX_PORT_CTL0_STATE_BLOCKING;
+		break;
+	case BR_STATE_LEARNING:
+		state = MV88E6XXX_PORT_CTL0_STATE_LEARNING;
+		break;
+	case BR_STATE_FORWARDING:
+		state = MV88E6XXX_PORT_CTL0_STATE_FORWARDING;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	list_for_each_entry(mst, &chip->msts, node) {
+		if (mst->br == dsa_port_bridge_dev_get(dp) &&
+		    mst->msti == st->msti) {
+			if (mst->stu.state[port] == state)
+				return 0;
+
+			mst->stu.state[port] = state;
+			mv88e6xxx_reg_lock(chip);
+			err = mv88e6xxx_stu_loadpurge(chip, &mst->stu);
+			mv88e6xxx_reg_unlock(chip);
+			return err;
+		}
+	}
+
+	return -ENOENT;
+}
+
 static int mv88e6xxx_port_check_hw_vlan(struct dsa_switch *ds, int port,
 					u16 vid)
 {
@@ -2437,6 +2568,12 @@ static int mv88e6xxx_port_vlan_leave(struct
mv88e6xxx_chip *chip,
 	if (err)
 		return err;
 
+	if (!vlan.valid && vlan.sid) {
+		err = mv88e6xxx_sid_put(chip, vlan.sid);
+		if (err)
+			return err;
+	}
+
 	return mv88e6xxx_g1_atu_remove(chip, vlan.fid, port, false);
 }
 
@@ -2482,6 +2619,44 @@ static int mv88e6xxx_port_vlan_del(struct dsa_switch *ds,
int port,
 	return err;
 }
 
+static int mv88e6xxx_vlan_msti_set(struct dsa_switch *ds,
+				   const struct switchdev_attr *attr)
+{
+	const struct switchdev_vlan_attr *vattr = &attr->u.vlan_attr;
+	struct mv88e6xxx_chip *chip = ds->priv;
+	struct mv88e6xxx_vtu_entry vlan;
+	u8 new_sid;
+	int err;
+
+	mv88e6xxx_reg_lock(chip);
+
+	err = mv88e6xxx_vtu_get(chip, vattr->vid, &vlan);
+	if (err)
+		goto unlock;
+
+	if (!vlan.valid) {
+		err = -EINVAL;
+		goto unlock;
+	}
+
+	err = mv88e6xxx_sid_get(chip, attr->orig_dev, vattr->msti,
&new_sid);
+	if (err)
+		goto unlock;
+
+	if (vlan.sid) {
+		err = mv88e6xxx_sid_put(chip, vlan.sid);
+		if (err)
+			goto unlock;
+	}
+
+	vlan.sid = new_sid;
+	err = mv88e6xxx_vtu_loadpurge(chip, &vlan);
+
+unlock:
+	mv88e6xxx_reg_unlock(chip);
+	return err;
+}
+
 static int mv88e6xxx_port_fdb_add(struct dsa_switch *ds, int port,
 				  const unsigned char *addr, u16 vid,
 				  struct dsa_db db)
@@ -6008,6 +6183,7 @@ static struct mv88e6xxx_chip *mv88e6xxx_alloc_chip(struct
device *dev)
 	mutex_init(&chip->reg_lock);
 	INIT_LIST_HEAD(&chip->mdios);
 	idr_init(&chip->policies);
+	INIT_LIST_HEAD(&chip->msts);
 
 	return chip;
 }
@@ -6540,10 +6716,12 @@ static const struct dsa_switch_ops mv88e6xxx_switch_ops
= {
 	.port_pre_bridge_flags	= mv88e6xxx_port_pre_bridge_flags,
 	.port_bridge_flags	= mv88e6xxx_port_bridge_flags,
 	.port_stp_state_set	= mv88e6xxx_port_stp_state_set,
+	.port_mst_state_set	= mv88e6xxx_port_mst_state_set,
 	.port_fast_age		= mv88e6xxx_port_fast_age,
 	.port_vlan_filtering	= mv88e6xxx_port_vlan_filtering,
 	.port_vlan_add		= mv88e6xxx_port_vlan_add,
 	.port_vlan_del		= mv88e6xxx_port_vlan_del,
+	.vlan_msti_set		= mv88e6xxx_vlan_msti_set,
 	.port_fdb_add           = mv88e6xxx_port_fdb_add,
 	.port_fdb_del           = mv88e6xxx_port_fdb_del,
 	.port_fdb_dump          = mv88e6xxx_port_fdb_dump,
diff --git a/drivers/net/dsa/mv88e6xxx/chip.h b/drivers/net/dsa/mv88e6xxx/chip.h
index 6d4daa24d3e5..6a0b66354e1d 100644
--- a/drivers/net/dsa/mv88e6xxx/chip.h
+++ b/drivers/net/dsa/mv88e6xxx/chip.h
@@ -297,6 +297,16 @@ struct mv88e6xxx_region_priv {
 	enum mv88e6xxx_region_id id;
 };
 
+struct mv88e6xxx_mst {
+	struct list_head node;
+
+	refcount_t refcnt;
+	struct net_device *br;
+	u16 msti;
+
+	struct mv88e6xxx_stu_entry stu;
+};
+
 struct mv88e6xxx_chip {
 	const struct mv88e6xxx_info *info;
 
@@ -397,6 +407,9 @@ struct mv88e6xxx_chip {
 
 	/* devlink regions */
 	struct devlink_region *regions[_MV88E6XXX_REGION_MAX];
+
+	/* Bridge MST to SID mappings */
+	struct list_head msts;
 };
 
 struct mv88e6xxx_bus_ops {
-- 
2.25.1

Hi Tobias,

On Tue, Mar 01, 2022 at 11:03:11AM +0100, Tobias Waldekranz
wrote:
> A proposal for the corresponding iproute2 interface is available here:
> 
> https://github.com/wkz/iproute2/tree/mst
Please pardon my ignorance. Is there a user-mode STP protocol application
that supports MSTP, and that you've tested these patches with?
I'd like to give it a try.