[patch net-next 0/5] mlxsw: Introduce support for offload forward mark

[patch net-next 0/5] mlxsw: Introduce support for offload forward mark

[Jiri Pirko]

From: Jiri Pirko <jiri@mellanox.com>

Ido says:
This patchset enables the forwarding of certain control packets by the
device instead of relying on the CPU to do the forwarding.

The first two patches simplify the current switchdev offload forward
infrastructure and make it usable for stacked devices. This is done by
moving the packet and port marking to the bridge driver instead of the
switch driver.

Patches 3-5 add the mlxsw specific bits to support the forward mark.

Ido Schimmel (5):
  switchdev: Support parent ID comparison for stacked devices
  bridge: switchdev: Add forward mark support for stacked devices
  mlxsw: spectrum: Simplify traps definition
  mlxsw: spectrum: Allow different traps to have different actions
  mlxsw: spectrum: Mirror certain packets to CPU

 Documentation/networking/switchdev.txt         |  13 +--
 drivers/net/ethernet/mellanox/mlxsw/core.h     |   1 +
 drivers/net/ethernet/mellanox/mlxsw/spectrum.c | 156 +++++++------------------
 drivers/net/ethernet/rocker/rocker_main.c      |   2 +-
 drivers/net/ethernet/rocker/rocker_ofdpa.c     |   4 -
 include/linux/netdevice.h                      |   5 -
 include/linux/skbuff.h                         |  13 +--
 include/net/switchdev.h                        |   6 -
 net/bridge/Makefile                            |   2 +
 net/bridge/br_forward.c                        |   3 +-
 net/bridge/br_if.c                             |  10 +-
 net/bridge/br_input.c                          |   2 +
 net/bridge/br_private.h                        |  37 ++++++
 net/bridge/br_switchdev.c                      |  57 +++++++++
 net/core/dev.c                                 |  10 --
 net/switchdev/switchdev.c                      |  87 --------------
 16 files changed, 158 insertions(+), 250 deletions(-)
 create mode 100644 net/bridge/br_switchdev.c

-- 
2.5.5

[patch net-next 1/5] switchdev: Support parent ID comparison for stacked devices

[Jiri Pirko]

From: Ido Schimmel <idosch@mellanox.com>

switchdev_port_same_parent_id() currently expects port netdevs, but we
need it to support stacked devices in the next patch, so drop the
NO_RECURSE flag.

Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
---
 net/switchdev/switchdev.c | 2 --
 1 file changed, 2 deletions(-)

diff --git a/net/switchdev/switchdev.c b/net/switchdev/switchdev.c
index 9e90129..2c683f2 100644
--- a/net/switchdev/switchdev.c
+++ b/net/switchdev/switchdev.c
@@ -1292,12 +1292,10 @@ bool switchdev_port_same_parent_id(struct net_device *a,
 	struct switchdev_attr a_attr = {
 		.orig_dev = a,
 		.id = SWITCHDEV_ATTR_ID_PORT_PARENT_ID,
-		.flags = SWITCHDEV_F_NO_RECURSE,
 	};
 	struct switchdev_attr b_attr = {
 		.orig_dev = b,
 		.id = SWITCHDEV_ATTR_ID_PORT_PARENT_ID,
-		.flags = SWITCHDEV_F_NO_RECURSE,
 	};
 
 	if (switchdev_port_attr_get(a, &a_attr) ||
-- 
2.5.5

[patch net-next 2/5] bridge: switchdev: Add forward mark support for stacked devices

[Jiri Pirko]

From: Ido Schimmel <idosch@mellanox.com>

switchdev_port_fwd_mark_set() is used to set the 'offload_fwd_mark' of
port netdevs so that packets being flooded by the device won't be
flooded twice.

It works by assigning a unique identifier (the ifindex of the first
bridge port) to bridge ports sharing the same parent ID. This prevents
packets from being flooded twice by the same switch, but will flood
packets through bridge ports belonging to a different switch.

This method is problematic when stacked devices are taken into account,
such as VLANs. In such cases, a physical port netdev can have upper
devices being members in two different bridges, thus requiring two
different 'offload_fwd_mark's to be configured on the port netdev, which
is impossible.

The main problem is that packet and netdev marking is performed at the
physical netdev level, whereas flooding occurs between bridge ports,
which are not necessarily port netdevs.

Instead, packet and netdev marking should really be done in the bridge
driver with the switch driver only telling it which packets it already
forwarded. The bridge driver will mark such packets using the mark
assigned to the ingress bridge port and will prevent the packet from
being forwarded through any bridge port sharing the same mark (i.e.
having the same parent ID).

Remove the current switchdev 'offload_fwd_mark' implementation and
instead implement the proposed method. In addition, make rocker - the
sole user of the mark - use the proposed method.

Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
---
 Documentation/networking/switchdev.txt     | 13 ++---
 drivers/net/ethernet/rocker/rocker_main.c  |  2 +-
 drivers/net/ethernet/rocker/rocker_ofdpa.c |  4 --
 include/linux/netdevice.h                  |  5 --
 include/linux/skbuff.h                     | 13 ++---
 include/net/switchdev.h                    |  6 ---
 net/bridge/Makefile                        |  2 +
 net/bridge/br_forward.c                    |  3 +-
 net/bridge/br_if.c                         | 10 ++--
 net/bridge/br_input.c                      |  2 +
 net/bridge/br_private.h                    | 37 +++++++++++++
 net/bridge/br_switchdev.c                  | 57 ++++++++++++++++++++
 net/core/dev.c                             | 10 ----
 net/switchdev/switchdev.c                  | 85 ------------------------------
 14 files changed, 117 insertions(+), 132 deletions(-)
 create mode 100644 net/bridge/br_switchdev.c

diff --git a/Documentation/networking/switchdev.txt b/Documentation/networking/switchdev.txt
index 31c3911..44235e8 100644
--- a/Documentation/networking/switchdev.txt
+++ b/Documentation/networking/switchdev.txt
@@ -283,15 +283,10 @@ be sent to the port netdev for processing by the bridge driver.  The
 bridge should not reflood the packet to the same ports the device flooded,
 otherwise there will be duplicate packets on the wire.
 
-To avoid duplicate packets, the device/driver should mark a packet as already
-forwarded using skb->offload_fwd_mark.  The same mark is set on the device
-ports in the domain using dev->offload_fwd_mark.  If the skb->offload_fwd_mark
-is non-zero and matches the forwarding egress port's dev->skb_mark, the kernel
-will drop the skb right before transmit on the egress port, with the
-understanding that the device already forwarded the packet on same egress port.
-The driver can use switchdev_port_fwd_mark_set() to set a globally unique mark
-for port's dev->offload_fwd_mark, based on the port's parent ID (switch ID) and
-a group ifindex.
+To avoid duplicate packets, the switch driver should mark a packet as already
+forwarded by setting the skb->offload_fwd_mark bit. The bridge driver will mark
+the skb using the ingress bridge port's mark and prevent it from being forwarded
+through any bridge port with the same mark.
 
 It is possible for the switch device to not handle flooding and push the
 packets up to the bridge driver for flooding.  This is not ideal as the number
diff --git a/drivers/net/ethernet/rocker/rocker_main.c b/drivers/net/ethernet/rocker/rocker_main.c
index f0b09b0..1f0c086 100644
--- a/drivers/net/ethernet/rocker/rocker_main.c
+++ b/drivers/net/ethernet/rocker/rocker_main.c
@@ -2412,7 +2412,7 @@ static int rocker_port_rx_proc(const struct rocker *rocker,
 	skb->protocol = eth_type_trans(skb, rocker_port->dev);
 
 	if (rx_flags & ROCKER_RX_FLAGS_FWD_OFFLOAD)
-		skb->offload_fwd_mark = rocker_port->dev->offload_fwd_mark;
+		skb->offload_fwd_mark = 1;
 
 	rocker_port->dev->stats.rx_packets++;
 	rocker_port->dev->stats.rx_bytes += skb->len;
diff --git a/drivers/net/ethernet/rocker/rocker_ofdpa.c b/drivers/net/ethernet/rocker/rocker_ofdpa.c
index 1ca7963..fcad907 100644
--- a/drivers/net/ethernet/rocker/rocker_ofdpa.c
+++ b/drivers/net/ethernet/rocker/rocker_ofdpa.c
@@ -2558,7 +2558,6 @@ static int ofdpa_port_init(struct rocker_port *rocker_port)
 	struct ofdpa_port *ofdpa_port = rocker_port->wpriv;
 	int err;
 
-	switchdev_port_fwd_mark_set(ofdpa_port->dev, NULL, false);
 	rocker_port_set_learning(rocker_port,
 				 !!(ofdpa_port->brport_flags & BR_LEARNING));
 
@@ -2817,7 +2816,6 @@ static int ofdpa_port_bridge_join(struct ofdpa_port *ofdpa_port,
 		ofdpa_port_internal_vlan_id_get(ofdpa_port, bridge->ifindex);
 
 	ofdpa_port->bridge_dev = bridge;
-	switchdev_port_fwd_mark_set(ofdpa_port->dev, bridge, true);
 
 	return ofdpa_port_vlan_add(ofdpa_port, NULL, OFDPA_UNTAGGED_VID, 0);
 }
@@ -2836,8 +2834,6 @@ static int ofdpa_port_bridge_leave(struct ofdpa_port *ofdpa_port)
 		ofdpa_port_internal_vlan_id_get(ofdpa_port,
 						ofdpa_port->dev->ifindex);
 
-	switchdev_port_fwd_mark_set(ofdpa_port->dev, ofdpa_port->bridge_dev,
-				    false);
 	ofdpa_port->bridge_dev = NULL;
 
 	err = ofdpa_port_vlan_add(ofdpa_port, NULL, OFDPA_UNTAGGED_VID, 0);
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index 794bb07..d122be9 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -1562,8 +1562,6 @@ enum netdev_priv_flags {
  *
  *	@xps_maps:	XXX: need comments on this one
  *
- *	@offload_fwd_mark:	Offload device fwding mark
- *
  *	@watchdog_timeo:	Represents the timeout that is used by
  *				the watchdog (see dev_watchdog())
  *	@watchdog_timer:	List of timers
@@ -1814,9 +1812,6 @@ struct net_device {
 #ifdef CONFIG_NET_CLS_ACT
 	struct tcf_proto __rcu  *egress_cl_list;
 #endif
-#ifdef CONFIG_NET_SWITCHDEV
-	u32			offload_fwd_mark;
-#endif
 
 	/* These may be needed for future network-power-down code. */
 	struct timer_list	watchdog_timer;
diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index 7047448..cfb7219 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -612,7 +612,6 @@ static inline bool skb_mstamp_after(const struct skb_mstamp *t1,
  *	@no_fcs:  Request NIC to treat last 4 bytes as Ethernet FCS
   *	@napi_id: id of the NAPI struct this skb came from
  *	@secmark: security marking
- *	@offload_fwd_mark: fwding offload mark
  *	@mark: Generic packet mark
  *	@vlan_proto: vlan encapsulation protocol
  *	@vlan_tci: vlan tag control information
@@ -730,7 +729,10 @@ struct sk_buff {
 	__u8			ipvs_property:1;
 	__u8			inner_protocol_type:1;
 	__u8			remcsum_offload:1;
-	/* 3 or 5 bit hole */
+#ifdef CONFIG_NET_SWITCHDEV
+	__u8			offload_fwd_mark:1;
+#endif
+	/* 2, 4 or 5 bit hole */
 
 #ifdef CONFIG_NET_SCHED
 	__u16			tc_index;	/* traffic control index */
@@ -757,14 +759,9 @@ struct sk_buff {
 		unsigned int	sender_cpu;
 	};
 #endif
-	union {
 #ifdef CONFIG_NETWORK_SECMARK
-		__u32		secmark;
+	__u32		secmark;
 #endif
-#ifdef CONFIG_NET_SWITCHDEV
-		__u32		offload_fwd_mark;
-#endif
-	};
 
 	union {
 		__u32		mark;
diff --git a/include/net/switchdev.h b/include/net/switchdev.h
index 62f6a96..82f5e04 100644
--- a/include/net/switchdev.h
+++ b/include/net/switchdev.h
@@ -347,12 +347,6 @@ static inline int switchdev_port_fdb_dump(struct sk_buff *skb,
        return idx;
 }
 
-static inline void switchdev_port_fwd_mark_set(struct net_device *dev,
-					       struct net_device *group_dev,
-					       bool joining)
-{
-}
-
 static inline bool switchdev_port_same_parent_id(struct net_device *a,
 						 struct net_device *b)
 {
diff --git a/net/bridge/Makefile b/net/bridge/Makefile
index a1cda5d..0aefc01 100644
--- a/net/bridge/Makefile
+++ b/net/bridge/Makefile
@@ -20,4 +20,6 @@ bridge-$(CONFIG_BRIDGE_IGMP_SNOOPING) += br_multicast.o br_mdb.o
 
 bridge-$(CONFIG_BRIDGE_VLAN_FILTERING) += br_vlan.o
 
+bridge-$(CONFIG_NET_SWITCHDEV) += br_switchdev.o
+
 obj-$(CONFIG_NETFILTER) += netfilter/
diff --git a/net/bridge/br_forward.c b/net/bridge/br_forward.c
index 63a83d8..32a02de 100644
--- a/net/bridge/br_forward.c
+++ b/net/bridge/br_forward.c
@@ -29,7 +29,8 @@ static inline int should_deliver(const struct net_bridge_port *p,
 
 	vg = nbp_vlan_group_rcu(p);
 	return ((p->flags & BR_HAIRPIN_MODE) || skb->dev != p->dev) &&
-		br_allowed_egress(vg, skb) && p->state == BR_STATE_FORWARDING;
+		br_allowed_egress(vg, skb) && p->state == BR_STATE_FORWARDING &&
+		nbp_switchdev_allowed_egress(p, skb);
 }
 
 int br_dev_queue_push_xmit(struct net *net, struct sock *sk, struct sk_buff *skb)
diff --git a/net/bridge/br_if.c b/net/bridge/br_if.c
index f2fede0..1da3221 100644
--- a/net/bridge/br_if.c
+++ b/net/bridge/br_if.c
@@ -545,6 +545,10 @@ int br_add_if(struct net_bridge *br, struct net_device *dev)
 	if (err)
 		goto err5;
 
+	err = nbp_switchdev_mark_set(p);
+	if (err)
+		goto err6;
+
 	dev_disable_lro(dev);
 
 	list_add_rcu(&p->list, &br->port_list);
@@ -566,7 +570,7 @@ int br_add_if(struct net_bridge *br, struct net_device *dev)
 	err = nbp_vlan_init(p);
 	if (err) {
 		netdev_err(dev, "failed to initialize vlan filtering on this port\n");
-		goto err6;
+		goto err7;
 	}
 
 	spin_lock_bh(&br->lock);
@@ -589,12 +593,12 @@ int br_add_if(struct net_bridge *br, struct net_device *dev)
 
 	return 0;
 
-err6:
+err7:
 	list_del_rcu(&p->list);
 	br_fdb_delete_by_port(br, p, 0, 1);
 	nbp_update_port_count(br);
+err6:
 	netdev_upper_dev_unlink(dev, br->dev);
-
 err5:
 	dev->priv_flags &= ~IFF_BRIDGE_PORT;
 	netdev_rx_handler_unregister(dev);
diff --git a/net/bridge/br_input.c b/net/bridge/br_input.c
index 8e48620..3132cfc 100644
--- a/net/bridge/br_input.c
+++ b/net/bridge/br_input.c
@@ -145,6 +145,8 @@ int br_handle_frame_finish(struct net *net, struct sock *sk, struct sk_buff *skb
 	if (!br_allowed_ingress(p->br, nbp_vlan_group_rcu(p), skb, &vid))
 		goto out;
 
+	nbp_switchdev_frame_mark(p, skb);
+
 	/* insert into forwarding database after filtering to avoid spoofing */
 	br = p->br;
 	if (p->flags & BR_LEARNING)
diff --git a/net/bridge/br_private.h b/net/bridge/br_private.h
index aac2a6e..2379b2b 100644
--- a/net/bridge/br_private.h
+++ b/net/bridge/br_private.h
@@ -251,6 +251,9 @@ struct net_bridge_port
 #ifdef CONFIG_BRIDGE_VLAN_FILTERING
 	struct net_bridge_vlan_group	__rcu *vlgrp;
 #endif
+#ifdef CONFIG_NET_SWITCHDEV
+	int				offload_fwd_mark;
+#endif
 };
 
 #define br_auto_port(p) ((p)->flags & BR_AUTO_MASK)
@@ -359,6 +362,11 @@ struct net_bridge
 	struct timer_list		gc_timer;
 	struct kobject			*ifobj;
 	u32				auto_cnt;
+
+#ifdef CONFIG_NET_SWITCHDEV
+	int offload_fwd_mark;
+#endif
+
 #ifdef CONFIG_BRIDGE_VLAN_FILTERING
 	struct net_bridge_vlan_group	__rcu *vlgrp;
 	u8				vlan_enabled;
@@ -381,6 +389,10 @@ struct br_input_skb_cb {
 #ifdef CONFIG_BRIDGE_VLAN_FILTERING
 	bool vlan_filtered;
 #endif
+
+#ifdef CONFIG_NET_SWITCHDEV
+	int offload_fwd_mark;
+#endif
 };
 
 #define BR_INPUT_SKB_CB(__skb)	((struct br_input_skb_cb *)(__skb)->cb)
@@ -1034,4 +1046,29 @@ static inline int br_sysfs_addbr(struct net_device *dev) { return 0; }
 static inline void br_sysfs_delbr(struct net_device *dev) { return; }
 #endif /* CONFIG_SYSFS */
 
+/* br_switchdev.c */
+#ifdef CONFIG_NET_SWITCHDEV
+int nbp_switchdev_mark_set(struct net_bridge_port *p);
+void nbp_switchdev_frame_mark(const struct net_bridge_port *p,
+			      struct sk_buff *skb);
+bool nbp_switchdev_allowed_egress(const struct net_bridge_port *p,
+				  const struct sk_buff *skb);
+#else
+static inline int nbp_switchdev_mark_set(struct net_bridge_port *p)
+{
+	return 0;
+}
+
+static inline void nbp_switchdev_frame_mark(const struct net_bridge_port *p,
+					    struct sk_buff *skb)
+{
+}
+
+static inline bool nbp_switchdev_allowed_egress(const struct net_bridge_port *p,
+						const struct sk_buff *skb)
+{
+	return true;
+}
+#endif /* CONFIG_NET_SWITCHDEV */
+
 #endif
diff --git a/net/bridge/br_switchdev.c b/net/bridge/br_switchdev.c
new file mode 100644
index 0000000..f4097b9
--- /dev/null
+++ b/net/bridge/br_switchdev.c
@@ -0,0 +1,57 @@
+#include <linux/kernel.h>
+#include <linux/list.h>
+#include <linux/netdevice.h>
+#include <linux/rtnetlink.h>
+#include <linux/skbuff.h>
+#include <net/switchdev.h>
+
+#include "br_private.h"
+
+static int br_switchdev_mark_get(struct net_bridge *br, struct net_device *dev)
+{
+	struct net_bridge_port *p;
+
+	/* dev is yet to be added to the port list. */
+	list_for_each_entry(p, &br->port_list, list) {
+		if (switchdev_port_same_parent_id(dev, p->dev))
+			return p->offload_fwd_mark;
+	}
+
+	return ++br->offload_fwd_mark;
+}
+
+int nbp_switchdev_mark_set(struct net_bridge_port *p)
+{
+	struct switchdev_attr attr = {
+		.orig_dev = p->dev,
+		.id = SWITCHDEV_ATTR_ID_PORT_PARENT_ID,
+	};
+	int err;
+
+	ASSERT_RTNL();
+
+	err = switchdev_port_attr_get(p->dev, &attr);
+	if (err) {
+		if (err == -EOPNOTSUPP)
+			return 0;
+		return err;
+	}
+
+	p->offload_fwd_mark = br_switchdev_mark_get(p->br, p->dev);
+
+	return 0;
+}
+
+void nbp_switchdev_frame_mark(const struct net_bridge_port *p,
+			      struct sk_buff *skb)
+{
+	if (skb->offload_fwd_mark && !WARN_ON_ONCE(!p->offload_fwd_mark))
+		BR_INPUT_SKB_CB(skb)->offload_fwd_mark = p->offload_fwd_mark;
+}
+
+bool nbp_switchdev_allowed_egress(const struct net_bridge_port *p,
+				  const struct sk_buff *skb)
+{
+	return !skb->offload_fwd_mark ||
+	       BR_INPUT_SKB_CB(skb)->offload_fwd_mark != p->offload_fwd_mark;
+}
diff --git a/net/core/dev.c b/net/core/dev.c
index a75df86..6ff6316 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -3355,16 +3355,6 @@ static int __dev_queue_xmit(struct sk_buff *skb, void *accel_priv)
 	else
 		skb_dst_force(skb);
 
-#ifdef CONFIG_NET_SWITCHDEV
-	/* Don't forward if offload device already forwarded */
-	if (skb->offload_fwd_mark &&
-	    skb->offload_fwd_mark == dev->offload_fwd_mark) {
-		consume_skb(skb);
-		rc = NET_XMIT_SUCCESS;
-		goto out;
-	}
-#endif
-
 	txq = netdev_pick_tx(dev, skb, accel_priv);
 	q = rcu_dereference_bh(txq->qdisc);
 
diff --git a/net/switchdev/switchdev.c b/net/switchdev/switchdev.c
index 2c683f2..1031a03 100644
--- a/net/switchdev/switchdev.c
+++ b/net/switchdev/switchdev.c
@@ -1305,88 +1305,3 @@ bool switchdev_port_same_parent_id(struct net_device *a,
 	return netdev_phys_item_id_same(&a_attr.u.ppid, &b_attr.u.ppid);
 }
 EXPORT_SYMBOL_GPL(switchdev_port_same_parent_id);
-
-static u32 switchdev_port_fwd_mark_get(struct net_device *dev,
-				       struct net_device *group_dev)
-{
-	struct net_device *lower_dev;
-	struct list_head *iter;
-
-	netdev_for_each_lower_dev(group_dev, lower_dev, iter) {
-		if (lower_dev == dev)
-			continue;
-		if (switchdev_port_same_parent_id(dev, lower_dev))
-			return lower_dev->offload_fwd_mark;
-		return switchdev_port_fwd_mark_get(dev, lower_dev);
-	}
-
-	return dev->ifindex;
-}
-
-static void switchdev_port_fwd_mark_reset(struct net_device *group_dev,
-					  u32 old_mark, u32 *reset_mark)
-{
-	struct net_device *lower_dev;
-	struct list_head *iter;
-
-	netdev_for_each_lower_dev(group_dev, lower_dev, iter) {
-		if (lower_dev->offload_fwd_mark == old_mark) {
-			if (!*reset_mark)
-				*reset_mark = lower_dev->ifindex;
-			lower_dev->offload_fwd_mark = *reset_mark;
-		}
-		switchdev_port_fwd_mark_reset(lower_dev, old_mark, reset_mark);
-	}
-}
-
-/**
- *	switchdev_port_fwd_mark_set - Set port offload forwarding mark
- *
- *	@dev: port device
- *	@group_dev: containing device
- *	@joining: true if dev is joining group; false if leaving group
- *
- *	An ungrouped port's offload mark is just its ifindex.  A grouped
- *	port's (member of a bridge, for example) offload mark is the ifindex
- *	of one of the ports in the group with the same parent (switch) ID.
- *	Ports on the same device in the same group will have the same mark.
- *
- *	Example:
- *
- *		br0		ifindex=9
- *		  sw1p1		ifindex=2	mark=2
- *		  sw1p2		ifindex=3	mark=2
- *		  sw2p1		ifindex=4	mark=5
- *		  sw2p2		ifindex=5	mark=5
- *
- *	If sw2p2 leaves the bridge, we'll have:
- *
- *		br0		ifindex=9
- *		  sw1p1		ifindex=2	mark=2
- *		  sw1p2		ifindex=3	mark=2
- *		  sw2p1		ifindex=4	mark=4
- *		sw2p2		ifindex=5	mark=5
- */
-void switchdev_port_fwd_mark_set(struct net_device *dev,
-				 struct net_device *group_dev,
-				 bool joining)
-{
-	u32 mark = dev->ifindex;
-	u32 reset_mark = 0;
-
-	if (group_dev) {
-		ASSERT_RTNL();
-		if (joining)
-			mark = switchdev_port_fwd_mark_get(dev, group_dev);
-		else if (dev->offload_fwd_mark == mark)
-			/* Ohoh, this port was the mark reference port,
-			 * but it's leaving the group, so reset the
-			 * mark for the remaining ports in the group.
-			 */
-			switchdev_port_fwd_mark_reset(group_dev, mark,
-						      &reset_mark);
-	}
-
-	dev->offload_fwd_mark = mark;
-}
-EXPORT_SYMBOL_GPL(switchdev_port_fwd_mark_set);
-- 
2.5.5

[patch net-next 3/5] mlxsw: spectrum: Simplify traps definition

[Jiri Pirko]

From: Ido Schimmel <idosch@mellanox.com>

Instead of copying & pasting the same struct initialization for every
Rx listener, just use a macro.

Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlxsw/spectrum.c | 146 ++++++-------------------
 1 file changed, 31 insertions(+), 115 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum.c
index 2713a64..cc83cd2 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum.c
@@ -2570,123 +2570,39 @@ static void mlxsw_sp_rx_listener_func(struct sk_buff *skb, u8 local_port,
 	netif_receive_skb(skb);
 }
 
+#define MLXSW_SP_RXL(_func, _trap_id)			\
+	{						\
+		.func = _func,				\
+		.local_port = MLXSW_PORT_DONT_CARE,	\
+		.trap_id = MLXSW_TRAP_ID_##_trap_id,	\
+	}
+
 static const struct mlxsw_rx_listener mlxsw_sp_rx_listener[] = {
-	{
-		.func = mlxsw_sp_rx_listener_func,
-		.local_port = MLXSW_PORT_DONT_CARE,
-		.trap_id = MLXSW_TRAP_ID_FDB_MC,
-	},
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, FDB_MC),
 	/* Traps for specific L2 packet types, not trapped as FDB MC */
-	{
-		.func = mlxsw_sp_rx_listener_func,
-		.local_port = MLXSW_PORT_DONT_CARE,
-		.trap_id = MLXSW_TRAP_ID_STP,
-	},
-	{
-		.func = mlxsw_sp_rx_listener_func,
-		.local_port = MLXSW_PORT_DONT_CARE,
-		.trap_id = MLXSW_TRAP_ID_LACP,
-	},
-	{
-		.func = mlxsw_sp_rx_listener_func,
-		.local_port = MLXSW_PORT_DONT_CARE,
-		.trap_id = MLXSW_TRAP_ID_EAPOL,
-	},
-	{
-		.func = mlxsw_sp_rx_listener_func,
-		.local_port = MLXSW_PORT_DONT_CARE,
-		.trap_id = MLXSW_TRAP_ID_LLDP,
-	},
-	{
-		.func = mlxsw_sp_rx_listener_func,
-		.local_port = MLXSW_PORT_DONT_CARE,
-		.trap_id = MLXSW_TRAP_ID_MMRP,
-	},
-	{
-		.func = mlxsw_sp_rx_listener_func,
-		.local_port = MLXSW_PORT_DONT_CARE,
-		.trap_id = MLXSW_TRAP_ID_MVRP,
-	},
-	{
-		.func = mlxsw_sp_rx_listener_func,
-		.local_port = MLXSW_PORT_DONT_CARE,
-		.trap_id = MLXSW_TRAP_ID_RPVST,
-	},
-	{
-		.func = mlxsw_sp_rx_listener_func,
-		.local_port = MLXSW_PORT_DONT_CARE,
-		.trap_id = MLXSW_TRAP_ID_DHCP,
-	},
-	{
-		.func = mlxsw_sp_rx_listener_func,
-		.local_port = MLXSW_PORT_DONT_CARE,
-		.trap_id = MLXSW_TRAP_ID_IGMP_QUERY,
-	},
-	{
-		.func = mlxsw_sp_rx_listener_func,
-		.local_port = MLXSW_PORT_DONT_CARE,
-		.trap_id = MLXSW_TRAP_ID_IGMP_V1_REPORT,
-	},
-	{
-		.func = mlxsw_sp_rx_listener_func,
-		.local_port = MLXSW_PORT_DONT_CARE,
-		.trap_id = MLXSW_TRAP_ID_IGMP_V2_REPORT,
-	},
-	{
-		.func = mlxsw_sp_rx_listener_func,
-		.local_port = MLXSW_PORT_DONT_CARE,
-		.trap_id = MLXSW_TRAP_ID_IGMP_V2_LEAVE,
-	},
-	{
-		.func = mlxsw_sp_rx_listener_func,
-		.local_port = MLXSW_PORT_DONT_CARE,
-		.trap_id = MLXSW_TRAP_ID_IGMP_V3_REPORT,
-	},
-	{
-		.func = mlxsw_sp_rx_listener_func,
-		.local_port = MLXSW_PORT_DONT_CARE,
-		.trap_id = MLXSW_TRAP_ID_ARPBC,
-	},
-	{
-		.func = mlxsw_sp_rx_listener_func,
-		.local_port = MLXSW_PORT_DONT_CARE,
-		.trap_id = MLXSW_TRAP_ID_ARPUC,
-	},
-	{
-		.func = mlxsw_sp_rx_listener_func,
-		.local_port = MLXSW_PORT_DONT_CARE,
-		.trap_id = MLXSW_TRAP_ID_MTUERROR,
-	},
-	{
-		.func = mlxsw_sp_rx_listener_func,
-		.local_port = MLXSW_PORT_DONT_CARE,
-		.trap_id = MLXSW_TRAP_ID_TTLERROR,
-	},
-	{
-		.func = mlxsw_sp_rx_listener_func,
-		.local_port = MLXSW_PORT_DONT_CARE,
-		.trap_id = MLXSW_TRAP_ID_LBERROR,
-	},
-	{
-		.func = mlxsw_sp_rx_listener_func,
-		.local_port = MLXSW_PORT_DONT_CARE,
-		.trap_id = MLXSW_TRAP_ID_OSPF,
-	},
-	{
-		.func = mlxsw_sp_rx_listener_func,
-		.local_port = MLXSW_PORT_DONT_CARE,
-		.trap_id = MLXSW_TRAP_ID_IP2ME,
-	},
-	{
-		.func = mlxsw_sp_rx_listener_func,
-		.local_port = MLXSW_PORT_DONT_CARE,
-		.trap_id = MLXSW_TRAP_ID_RTR_INGRESS0,
-	},
-	{
-		.func = mlxsw_sp_rx_listener_func,
-		.local_port = MLXSW_PORT_DONT_CARE,
-		.trap_id = MLXSW_TRAP_ID_HOST_MISS_IPV4,
-	},
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, STP),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, LACP),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, EAPOL),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, LLDP),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, MMRP),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, MVRP),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, RPVST),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, DHCP),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, IGMP_QUERY),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, IGMP_V1_REPORT),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, IGMP_V2_REPORT),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, IGMP_V2_LEAVE),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, IGMP_V3_REPORT),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, ARPBC),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, ARPUC),
+	/* L3 traps */
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, MTUERROR),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, TTLERROR),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, LBERROR),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, OSPF),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, IP2ME),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, RTR_INGRESS0),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, HOST_MISS_IPV4),
 };
 
 static int mlxsw_sp_traps_init(struct mlxsw_sp *mlxsw_sp)
-- 
2.5.5

[patch net-next 4/5] mlxsw: spectrum: Allow different traps to have different actions

[Jiri Pirko]

From: Ido Schimmel <idosch@mellanox.com>

Up until now we only trapped packets to CPU, but we are going to allow
some packets to be mirrored (trap & forward) to CPU.

Extend the Rx listener with 'action' member.

Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlxsw/core.h     |  1 +
 drivers/net/ethernet/mellanox/mlxsw/spectrum.c | 59 +++++++++++++-------------
 2 files changed, 31 insertions(+), 29 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlxsw/core.h b/drivers/net/ethernet/mellanox/mlxsw/core.h
index d3476ea..d2e3297 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/core.h
+++ b/drivers/net/ethernet/mellanox/mlxsw/core.h
@@ -87,6 +87,7 @@ struct mlxsw_rx_listener {
 	void (*func)(struct sk_buff *skb, u8 local_port, void *priv);
 	u8 local_port;
 	u16 trap_id;
+	enum mlxsw_reg_hpkt_action action;
 };
 
 struct mlxsw_event_listener {
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum.c
index cc83cd2..ea4086a 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum.c
@@ -2570,39 +2570,40 @@ static void mlxsw_sp_rx_listener_func(struct sk_buff *skb, u8 local_port,
 	netif_receive_skb(skb);
 }
 
-#define MLXSW_SP_RXL(_func, _trap_id)			\
-	{						\
-		.func = _func,				\
-		.local_port = MLXSW_PORT_DONT_CARE,	\
-		.trap_id = MLXSW_TRAP_ID_##_trap_id,	\
+#define MLXSW_SP_RXL(_func, _trap_id, _action)			\
+	{							\
+		.func = _func,					\
+		.local_port = MLXSW_PORT_DONT_CARE,		\
+		.trap_id = MLXSW_TRAP_ID_##_trap_id,		\
+		.action = MLXSW_REG_HPKT_ACTION_##_action,	\
 	}
 
 static const struct mlxsw_rx_listener mlxsw_sp_rx_listener[] = {
-	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, FDB_MC),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, FDB_MC, TRAP_TO_CPU),
 	/* Traps for specific L2 packet types, not trapped as FDB MC */
-	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, STP),
-	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, LACP),
-	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, EAPOL),
-	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, LLDP),
-	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, MMRP),
-	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, MVRP),
-	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, RPVST),
-	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, DHCP),
-	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, IGMP_QUERY),
-	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, IGMP_V1_REPORT),
-	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, IGMP_V2_REPORT),
-	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, IGMP_V2_LEAVE),
-	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, IGMP_V3_REPORT),
-	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, ARPBC),
-	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, ARPUC),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, STP, TRAP_TO_CPU),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, LACP, TRAP_TO_CPU),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, EAPOL, TRAP_TO_CPU),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, LLDP, TRAP_TO_CPU),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, MMRP, TRAP_TO_CPU),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, MVRP, TRAP_TO_CPU),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, RPVST, TRAP_TO_CPU),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, DHCP, TRAP_TO_CPU),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, IGMP_QUERY, TRAP_TO_CPU),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, IGMP_V1_REPORT, TRAP_TO_CPU),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, IGMP_V2_REPORT, TRAP_TO_CPU),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, IGMP_V2_LEAVE, TRAP_TO_CPU),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, IGMP_V3_REPORT, TRAP_TO_CPU),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, ARPBC, TRAP_TO_CPU),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, ARPUC, TRAP_TO_CPU),
 	/* L3 traps */
-	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, MTUERROR),
-	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, TTLERROR),
-	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, LBERROR),
-	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, OSPF),
-	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, IP2ME),
-	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, RTR_INGRESS0),
-	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, HOST_MISS_IPV4),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, MTUERROR, TRAP_TO_CPU),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, TTLERROR, TRAP_TO_CPU),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, LBERROR, TRAP_TO_CPU),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, OSPF, TRAP_TO_CPU),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, IP2ME, TRAP_TO_CPU),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, RTR_INGRESS0, TRAP_TO_CPU),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, HOST_MISS_IPV4, TRAP_TO_CPU),
 };
 
 static int mlxsw_sp_traps_init(struct mlxsw_sp *mlxsw_sp)
@@ -2629,7 +2630,7 @@ static int mlxsw_sp_traps_init(struct mlxsw_sp *mlxsw_sp)
 		if (err)
 			goto err_rx_listener_register;
 
-		mlxsw_reg_hpkt_pack(hpkt_pl, MLXSW_REG_HPKT_ACTION_TRAP_TO_CPU,
+		mlxsw_reg_hpkt_pack(hpkt_pl, mlxsw_sp_rx_listener[i].action,
 				    mlxsw_sp_rx_listener[i].trap_id);
 		err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(hpkt), hpkt_pl);
 		if (err)
-- 
2.5.5

[patch net-next 5/5] mlxsw: spectrum: Mirror certain packets to CPU

[Jiri Pirko]

From: Ido Schimmel <idosch@mellanox.com>

Instead of trapping certain packets to the CPU and then relying on it to
flood them we can instead make the device mirror them.

The following packet types are mirrored:

* DHCP: Broadcast packets that should be flooded by the device, but also
trapped in case CPU is running the DHCP server.

* IGMP query: Multicast packets that need to be forwarded to other
bridge ports, but also trapped so that receiving netdev will be marked
as a router port by the bridge driver.

* ARP request: Broadcast packets that should be forwarded to other
bridge ports, but also trapped in case requested IP is of the local
machine.

* ARP response: Unicast packets that should be forwarded by the bridge
but also trapped in case response is directed at us.

Set the trap action of such packets to mirror and mark them using
'offload_fwd_mark' to prevent the bridge driver from forwarding them
itself.

Note that OSPF packets are also marked despite their action being trap.
The reason for this is that the device traps such packets in the
pipeline after they were already flooded.

Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlxsw/spectrum.c | 17 ++++++++++++-----
 1 file changed, 12 insertions(+), 5 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum.c
index ea4086a..8c9c124 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum.c
@@ -2570,6 +2570,13 @@ static void mlxsw_sp_rx_listener_func(struct sk_buff *skb, u8 local_port,
 	netif_receive_skb(skb);
 }
 
+static void mlxsw_sp_rx_listener_mark_func(struct sk_buff *skb, u8 local_port,
+					   void *priv)
+{
+	skb->offload_fwd_mark = 1;
+	return mlxsw_sp_rx_listener_func(skb, local_port, priv);
+}
+
 #define MLXSW_SP_RXL(_func, _trap_id, _action)			\
 	{							\
 		.func = _func,					\
@@ -2588,19 +2595,19 @@ static const struct mlxsw_rx_listener mlxsw_sp_rx_listener[] = {
 	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, MMRP, TRAP_TO_CPU),
 	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, MVRP, TRAP_TO_CPU),
 	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, RPVST, TRAP_TO_CPU),
-	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, DHCP, TRAP_TO_CPU),
-	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, IGMP_QUERY, TRAP_TO_CPU),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_mark_func, DHCP, MIRROR_TO_CPU),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_mark_func, IGMP_QUERY, MIRROR_TO_CPU),
 	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, IGMP_V1_REPORT, TRAP_TO_CPU),
 	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, IGMP_V2_REPORT, TRAP_TO_CPU),
 	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, IGMP_V2_LEAVE, TRAP_TO_CPU),
 	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, IGMP_V3_REPORT, TRAP_TO_CPU),
-	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, ARPBC, TRAP_TO_CPU),
-	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, ARPUC, TRAP_TO_CPU),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_mark_func, ARPBC, MIRROR_TO_CPU),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_mark_func, ARPUC, MIRROR_TO_CPU),
 	/* L3 traps */
 	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, MTUERROR, TRAP_TO_CPU),
 	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, TTLERROR, TRAP_TO_CPU),
 	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, LBERROR, TRAP_TO_CPU),
-	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, OSPF, TRAP_TO_CPU),
+	MLXSW_SP_RXL(mlxsw_sp_rx_listener_mark_func, OSPF, TRAP_TO_CPU),
 	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, IP2ME, TRAP_TO_CPU),
 	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, RTR_INGRESS0, TRAP_TO_CPU),
 	MLXSW_SP_RXL(mlxsw_sp_rx_listener_func, HOST_MISS_IPV4, TRAP_TO_CPU),
-- 
2.5.5

Re: [patch net-next 0/5] mlxsw: Introduce support for offload forward mark

[David Miller]

From: Jiri Pirko <jiri@resnulli.us>
Date: Thu, 25 Aug 2016 18:42:35 +0200

> Ido says:
> This patchset enables the forwarding of certain control packets by the
> device instead of relying on the CPU to do the forwarding.
> 
> The first two patches simplify the current switchdev offload forward
> infrastructure and make it usable for stacked devices. This is done by
> moving the packet and port marking to the bridge driver instead of the
> switch driver.
> 
> Patches 3-5 add the mlxsw specific bits to support the forward mark.

Series applied, thanks.