Merge branch 'Generic-adjustment-for-flow-dissector-in-DSA'

Vladimir Oltean says:

====================
Generic adjustment for flow dissector in DSA

This is the v2 of a series initially submitted in May:
https://www.spinics.net/lists/netdev/msg651866.html

The end goal is to get rid of the unintuitive code for the flow
dissector that currently exists in the taggers. It can all be replaced
by a single, common function.

Some background work needs to be done for that. Especially the ocelot
driver poses some problems, since it has a different tag length between
RX and TX, and I didn't want to make DSA aware of that, since I could
instead make the tag lengths equal.

Changes in v3:
- Added an optimization (08/15) that makes the generic case not need to
  call the .flow_dissect function pointer. Basically .flow_dissect now
  currently only exists for sja1105.
- Moved the .promisc_on_master property to the tagger structure.
- Added the .tail_tag property to the tagger structure.
- Disabled "suppresscc = all" from my .gitconfig.
====================

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

19 files changed, 158 insertions, 135 deletions

diff --git a/drivers/net/dsa/ocelot/felix.c b/drivers/net/dsa/ocelot/felix.c
index a56fc50f5be4..ab3ee5c3fd02 100644
--- a/drivers/net/dsa/ocelot/felix.c
+++ b/drivers/net/dsa/ocelot/felix.c
@@ -439,6 +439,8 @@ static int felix_init_structs(struct felix *felix, int num_phys_ports)
 	ocelot->vcap_is2_actions= felix->info->vcap_is2_actions;
 	ocelot->vcap		= felix->info->vcap;
 	ocelot->ops		= felix->info->ops;
+	ocelot->inj_prefix	= OCELOT_TAG_PREFIX_SHORT;
+	ocelot->xtr_prefix	= OCELOT_TAG_PREFIX_SHORT;
 
 	port_phy_modes = kcalloc(num_phys_ports, sizeof(phy_interface_t),
 				 GFP_KERNEL);
@@ -509,7 +511,7 @@ static int felix_init_structs(struct felix *felix, int num_phys_ports)
 			return PTR_ERR(target);
 		}
 
-		template = devm_kzalloc(ocelot->dev, OCELOT_TAG_LEN,
+		template = devm_kzalloc(ocelot->dev, OCELOT_TOTAL_TAG_LEN,
 					GFP_KERNEL);
 		if (!template) {
 			dev_err(ocelot->dev,
@@ -538,6 +540,28 @@ static int felix_init_structs(struct felix *felix, int num_phys_ports)
 	return 0;
 }
 
+/* The CPU port module is connected to the Node Processor Interface (NPI). This
+ * is the mode through which frames can be injected from and extracted to an
+ * external CPU, over Ethernet.
+ */
+static void felix_npi_port_init(struct ocelot *ocelot, int port)
+{
+	ocelot->npi = port;
+
+	ocelot_write(ocelot, QSYS_EXT_CPU_CFG_EXT_CPUQ_MSK_M |
+		     QSYS_EXT_CPU_CFG_EXT_CPU_PORT(port),
+		     QSYS_EXT_CPU_CFG);
+
+	/* NPI port Injection/Extraction configuration */
+	ocelot_fields_write(ocelot, port, SYS_PORT_MODE_INCL_XTR_HDR,
+			    ocelot->xtr_prefix);
+	ocelot_fields_write(ocelot, port, SYS_PORT_MODE_INCL_INJ_HDR,
+			    ocelot->inj_prefix);
+
+	/* Disable transmission of pause frames */
+	ocelot_fields_write(ocelot, port, SYS_PAUSE_CFG_PAUSE_ENA, 0);
+}
+
 /* Hardware initialization done here so that we can allocate structures with
  * devm without fear of dsa_register_switch returning -EPROBE_DEFER and causing
  * us to allocate structures twice (leak memory) and map PCI memory twice
@@ -570,11 +594,8 @@ static int felix_setup(struct dsa_switch *ds)
 	for (port = 0; port < ds->num_ports; port++) {
 		ocelot_init_port(ocelot, port);
 
-		/* Bring up the CPU port module and configure the NPI port */
 		if (dsa_is_cpu_port(ds, port))
-			ocelot_configure_cpu(ocelot, port,
-					     OCELOT_TAG_PREFIX_NONE,
-					     OCELOT_TAG_PREFIX_LONG);
+			felix_npi_port_init(ocelot, port);
 
 		/* Set the default QoS Classification based on PCP and DEI
 		 * bits of vlan tag.
diff --git a/drivers/net/dsa/ocelot/felix_vsc9959.c b/drivers/net/dsa/ocelot/felix_vsc9959.c
index 3ab6d6847c5b..2e9270c00096 100644
--- a/drivers/net/dsa/ocelot/felix_vsc9959.c
+++ b/drivers/net/dsa/ocelot/felix_vsc9959.c
@@ -1155,6 +1155,8 @@ static void vsc9959_xmit_template_populate(struct ocelot *ocelot, int port)
 	struct ocelot_port *ocelot_port = ocelot->ports[port];
 	u8 *template = ocelot_port->xmit_template;
 	u64 bypass, dest, src;
+	__be32 *prefix;
+	u8 *injection;
 
 	/* Set the source port as the CPU port module and not the
 	 * NPI port
@@ -1163,9 +1165,14 @@ static void vsc9959_xmit_template_populate(struct ocelot *ocelot, int port)
 	dest = BIT(port);
 	bypass = true;
 
-	packing(template, &bypass, 127, 127, OCELOT_TAG_LEN, PACK, 0);
-	packing(template, &dest,    68,  56, OCELOT_TAG_LEN, PACK, 0);
-	packing(template, &src,     46,  43, OCELOT_TAG_LEN, PACK, 0);
+	injection = template + OCELOT_SHORT_PREFIX_LEN;
+	prefix = (__be32 *)template;
+
+	packing(injection, &bypass, 127, 127, OCELOT_TAG_LEN, PACK, 0);
+	packing(injection, &dest,    68,  56, OCELOT_TAG_LEN, PACK, 0);
+	packing(injection, &src,     46,  43, OCELOT_TAG_LEN, PACK, 0);
+
+	*prefix = cpu_to_be32(0x8880000a);
 }
 
 static const struct felix_info felix_info_vsc9959 = {
diff --git a/drivers/net/dsa/ocelot/seville_vsc9953.c b/drivers/net/dsa/ocelot/seville_vsc9953.c
index b0ff90c0ae16..e28dd600f464 100644
--- a/drivers/net/dsa/ocelot/seville_vsc9953.c
+++ b/drivers/net/dsa/ocelot/seville_vsc9953.c
@@ -1003,6 +1003,8 @@ static void vsc9953_xmit_template_populate(struct ocelot *ocelot, int port)
 	struct ocelot_port *ocelot_port = ocelot->ports[port];
 	u8 *template = ocelot_port->xmit_template;
 	u64 bypass, dest, src;
+	__be32 *prefix;
+	u8 *injection;
 
 	/* Set the source port as the CPU port module and not the
 	 * NPI port
@@ -1011,9 +1013,14 @@ static void vsc9953_xmit_template_populate(struct ocelot *ocelot, int port)
 	dest = BIT(port);
 	bypass = true;
 
-	packing(template, &bypass, 127, 127, OCELOT_TAG_LEN, PACK, 0);
-	packing(template, &dest,    67,  57, OCELOT_TAG_LEN, PACK, 0);
-	packing(template, &src,     46,  43, OCELOT_TAG_LEN, PACK, 0);
+	injection = template + OCELOT_SHORT_PREFIX_LEN;
+	prefix = (__be32 *)template;
+
+	packing(injection, &bypass, 127, 127, OCELOT_TAG_LEN, PACK, 0);
+	packing(injection, &dest,    67,  57, OCELOT_TAG_LEN, PACK, 0);
+	packing(injection, &src,     46,  43, OCELOT_TAG_LEN, PACK, 0);
+
+	*prefix = cpu_to_be32(0x88800005);
 }
 
 static const struct felix_info seville_info_vsc9953 = {
diff --git a/drivers/net/ethernet/mscc/ocelot.c b/drivers/net/ethernet/mscc/ocelot.c
index 0445c5ee5551..b9375d96cdbc 100644
--- a/drivers/net/ethernet/mscc/ocelot.c
+++ b/drivers/net/ethernet/mscc/ocelot.c
@@ -1346,22 +1346,14 @@ void ocelot_init_port(struct ocelot *ocelot, int port)
 }
 EXPORT_SYMBOL(ocelot_init_port);
 
-/* Configure and enable the CPU port module, which is a set of queues.
- * If @npi contains a valid port index, the CPU port module is connected
- * to the Node Processor Interface (NPI). This is the mode through which
- * frames can be injected from and extracted to an external CPU,
- * over Ethernet.
+/* Configure and enable the CPU port module, which is a set of queues
+ * accessible through register MMIO, frame DMA or Ethernet (in case
+ * NPI mode is used).
  */
-void ocelot_configure_cpu(struct ocelot *ocelot, int npi,
-			  enum ocelot_tag_prefix injection,
-			  enum ocelot_tag_prefix extraction)
+static void ocelot_cpu_port_init(struct ocelot *ocelot)
 {
 	int cpu = ocelot->num_phys_ports;
 
-	ocelot->npi = npi;
-	ocelot->inj_prefix = injection;
-	ocelot->xtr_prefix = extraction;
-
 	/* The unicast destination PGID for the CPU port module is unused */
 	ocelot_write_rix(ocelot, 0, ANA_PGID_PGID, cpu);
 	/* Instead set up a multicast destination PGID for traffic copied to
@@ -1373,31 +1365,13 @@ void ocelot_configure_cpu(struct ocelot *ocelot, int npi,
 			 ANA_PORT_PORT_CFG_PORTID_VAL(cpu),
 			 ANA_PORT_PORT_CFG, cpu);
 
-	if (npi >= 0 && npi < ocelot->num_phys_ports) {
-		ocelot_write(ocelot, QSYS_EXT_CPU_CFG_EXT_CPUQ_MSK_M |
-			     QSYS_EXT_CPU_CFG_EXT_CPU_PORT(npi),
-			     QSYS_EXT_CPU_CFG);
-
-		/* Enable NPI port */
-		ocelot_fields_write(ocelot, npi,
-				    QSYS_SWITCH_PORT_MODE_PORT_ENA, 1);
-		/* NPI port Injection/Extraction configuration */
-		ocelot_fields_write(ocelot, npi, SYS_PORT_MODE_INCL_XTR_HDR,
-				    extraction);
-		ocelot_fields_write(ocelot, npi, SYS_PORT_MODE_INCL_INJ_HDR,
-				    injection);
-
-		/* Disable transmission of pause frames */
-		ocelot_fields_write(ocelot, npi, SYS_PAUSE_CFG_PAUSE_ENA, 0);
-	}
-
 	/* Enable CPU port module */
 	ocelot_fields_write(ocelot, cpu, QSYS_SWITCH_PORT_MODE_PORT_ENA, 1);
 	/* CPU port Injection/Extraction configuration */
 	ocelot_fields_write(ocelot, cpu, SYS_PORT_MODE_INCL_XTR_HDR,
-			    extraction);
+			    ocelot->xtr_prefix);
 	ocelot_fields_write(ocelot, cpu, SYS_PORT_MODE_INCL_INJ_HDR,
-			    injection);
+			    ocelot->inj_prefix);
 
 	/* Configure the CPU port to be VLAN aware */
 	ocelot_write_gix(ocelot, ANA_PORT_VLAN_CFG_VLAN_VID(0) |
@@ -1405,7 +1379,6 @@ void ocelot_configure_cpu(struct ocelot *ocelot, int npi,
 				 ANA_PORT_VLAN_CFG_VLAN_POP_CNT(1),
 			 ANA_PORT_VLAN_CFG, cpu);
 }
-EXPORT_SYMBOL(ocelot_configure_cpu);
 
 int ocelot_init(struct ocelot *ocelot)
 {
@@ -1445,6 +1418,7 @@ int ocelot_init(struct ocelot *ocelot)
 	ocelot_mact_init(ocelot);
 	ocelot_vlan_init(ocelot);
 	ocelot_vcap_init(ocelot);
+	ocelot_cpu_port_init(ocelot);
 
 	for (port = 0; port < ocelot->num_phys_ports; port++) {
 		/* Clear all counters (5 groups) */
diff --git a/drivers/net/ethernet/mscc/ocelot_vsc7514.c b/drivers/net/ethernet/mscc/ocelot_vsc7514.c
index dfb1535f26f2..d7aef2fb9848 100644
--- a/drivers/net/ethernet/mscc/ocelot_vsc7514.c
+++ b/drivers/net/ethernet/mscc/ocelot_vsc7514.c
@@ -930,10 +930,6 @@ static int mscc_ocelot_init_ports(struct platform_device *pdev,
 	if (!ocelot->ports)
 		return -ENOMEM;
 
-	/* No NPI port */
-	ocelot_configure_cpu(ocelot, -1, OCELOT_TAG_PREFIX_NONE,
-			     OCELOT_TAG_PREFIX_NONE);
-
 	for_each_available_child_of_node(ports, portnp) {
 		struct ocelot_port_private *priv;
 		struct ocelot_port *ocelot_port;
@@ -1120,6 +1116,9 @@ static int mscc_ocelot_probe(struct platform_device *pdev)
 	ocelot->vcap_is2_keys = vsc7514_vcap_is2_keys;
 	ocelot->vcap_is2_actions = vsc7514_vcap_is2_actions;
 	ocelot->vcap = vsc7514_vcap_props;
+	ocelot->inj_prefix = OCELOT_TAG_PREFIX_NONE;
+	ocelot->xtr_prefix = OCELOT_TAG_PREFIX_NONE;
+	ocelot->npi = -1;
 
 	err = ocelot_init(ocelot);
 	if (err)
diff --git a/include/net/dsa.h b/include/net/dsa.h
index d16057c5987a..b502a63d196e 100644
--- a/include/net/dsa.h
+++ b/include/net/dsa.h
@@ -74,8 +74,8 @@ struct dsa_device_ops {
 	struct sk_buff *(*xmit)(struct sk_buff *skb, struct net_device *dev);
 	struct sk_buff *(*rcv)(struct sk_buff *skb, struct net_device *dev,
 			       struct packet_type *pt);
-	int (*flow_dissect)(const struct sk_buff *skb, __be16 *proto,
-			    int *offset);
+	void (*flow_dissect)(const struct sk_buff *skb, __be16 *proto,
+			     int *offset);
 	/* Used to determine which traffic should match the DSA filter in
 	 * eth_type_trans, and which, if any, should bypass it and be processed
 	 * as regular on the master net device.
@@ -84,6 +84,13 @@ struct dsa_device_ops {
 	unsigned int overhead;
 	const char *name;
 	enum dsa_tag_protocol proto;
+	/* Some tagging protocols either mangle or shift the destination MAC
+	 * address, in which case the DSA master would drop packets on ingress
+	 * if what it understands out of the destination MAC address is not in
+	 * its RX filter.
+	 */
+	bool promisc_on_master;
+	bool tail_tag;
 };
 
 /* This structure defines the control interfaces that are overlayed by the
@@ -705,6 +712,32 @@ static inline bool dsa_can_decode(const struct sk_buff *skb,
 	return false;
 }
 
+/* All DSA tags that push the EtherType to the right (basically all except tail
+ * tags, which don't break dissection) can be treated the same from the
+ * perspective of the flow dissector.
+ *
+ * We need to return:
+ *  - offset: the (B - A) difference between:
+ *    A. the position of the real EtherType and
+ *    B. the current skb->data (aka ETH_HLEN bytes into the frame, aka 2 bytes
+ *       after the normal EtherType was supposed to be)
+ *    The offset in bytes is exactly equal to the tagger overhead (and half of
+ *    that, in __be16 shorts).
+ *
+ *  - proto: the value of the real EtherType.
+ */
+static inline void dsa_tag_generic_flow_dissect(const struct sk_buff *skb,
+						__be16 *proto, int *offset)
+{
+#if IS_ENABLED(CONFIG_NET_DSA)
+	const struct dsa_device_ops *ops = skb->dev->dsa_ptr->tag_ops;
+	int tag_len = ops->overhead;
+
+	*offset = tag_len;
+	*proto = ((__be16 *)skb->data)[(tag_len / 2) - 1];
+#endif
+}
+
 #if IS_ENABLED(CONFIG_NET_DSA)
 static inline int __dsa_netdevice_ops_check(struct net_device *dev)
 {
diff --git a/include/soc/mscc/ocelot.h b/include/soc/mscc/ocelot.h
index 3105bbb6cdcf..3093385f6147 100644
--- a/include/soc/mscc/ocelot.h
+++ b/include/soc/mscc/ocelot.h
@@ -101,6 +101,7 @@
 #define OCELOT_TAG_LEN			16
 #define OCELOT_SHORT_PREFIX_LEN		4
 #define OCELOT_LONG_PREFIX_LEN		16
+#define OCELOT_TOTAL_TAG_LEN	(OCELOT_SHORT_PREFIX_LEN + OCELOT_TAG_LEN)
 
 #define OCELOT_SPEED_2500		0
 #define OCELOT_SPEED_1000		1
@@ -672,9 +673,6 @@ void __ocelot_rmw_ix(struct ocelot *ocelot, u32 val, u32 mask, u32 reg,
 int ocelot_regfields_init(struct ocelot *ocelot,
 			  const struct reg_field *const regfields);
 struct regmap *ocelot_regmap_init(struct ocelot *ocelot, struct resource *res);
-void ocelot_configure_cpu(struct ocelot *ocelot, int npi,
-			  enum ocelot_tag_prefix injection,
-			  enum ocelot_tag_prefix extraction);
 int ocelot_init(struct ocelot *ocelot);
 void ocelot_deinit(struct ocelot *ocelot);
 void ocelot_init_port(struct ocelot *ocelot, int port);
diff --git a/net/core/flow_dissector.c b/net/core/flow_dissector.c
index 29806eb765cf..e21950a2c897 100644
--- a/net/core/flow_dissector.c
+++ b/net/core/flow_dissector.c
@@ -932,8 +932,14 @@ bool __skb_flow_dissect(const struct net *net,
 			int offset = 0;
 
 			ops = skb->dev->dsa_ptr->tag_ops;
-			if (ops->flow_dissect &&
-			    !ops->flow_dissect(skb, &proto, &offset)) {
+			/* Tail taggers don't break flow dissection */
+			if (!ops->tail_tag) {
+				if (ops->flow_dissect)
+					ops->flow_dissect(skb, &proto, &offset);
+				else
+					dsa_tag_generic_flow_dissect(skb,
+								     &proto,
+								     &offset);
 				hlen -= offset;
 				nhoff += offset;
 			}
diff --git a/net/dsa/master.c b/net/dsa/master.c
index 61615ebc70e9..c91de041a91d 100644
--- a/net/dsa/master.c
+++ b/net/dsa/master.c
@@ -259,6 +259,18 @@ static void dsa_netdev_ops_set(struct net_device *dev,
 	dev->dsa_ptr->netdev_ops = ops;
 }
 
+static void dsa_master_set_promiscuity(struct net_device *dev, int inc)
+{
+	const struct dsa_device_ops *ops = dev->dsa_ptr->tag_ops;
+
+	if (!ops->promisc_on_master)
+		return;
+
+	rtnl_lock();
+	dev_set_promiscuity(dev, inc);
+	rtnl_unlock();
+}
+
 static ssize_t tagging_show(struct device *d, struct device_attribute *attr,
 			    char *buf)
 {
@@ -314,9 +326,12 @@ int dsa_master_setup(struct net_device *dev, struct dsa_port *cpu_dp)
 	dev->dsa_ptr = cpu_dp;
 	lockdep_set_class(&dev->addr_list_lock,
 			  &dsa_master_addr_list_lock_key);
+
+	dsa_master_set_promiscuity(dev, 1);
+
 	ret = dsa_master_ethtool_setup(dev);
 	if (ret)
-		return ret;
+		goto out_err_reset_promisc;
 
 	dsa_netdev_ops_set(dev, &dsa_netdev_ops);
 
@@ -329,6 +344,8 @@ int dsa_master_setup(struct net_device *dev, struct dsa_port *cpu_dp)
 out_err_ndo_teardown:
 	dsa_netdev_ops_set(dev, NULL);
 	dsa_master_ethtool_teardown(dev);
+out_err_reset_promisc:
+	dsa_master_set_promiscuity(dev, -1);
 	return ret;
 }
 
@@ -338,6 +355,7 @@ void dsa_master_teardown(struct net_device *dev)
 	dsa_netdev_ops_set(dev, NULL);
 	dsa_master_ethtool_teardown(dev);
 	dsa_master_reset_mtu(dev);
+	dsa_master_set_promiscuity(dev, -1);
 
 	dev->dsa_ptr = NULL;
 
diff --git a/net/dsa/tag_brcm.c b/net/dsa/tag_brcm.c
index 1dab212a294f..69d6b8c597a9 100644
--- a/net/dsa/tag_brcm.c
+++ b/net/dsa/tag_brcm.c
@@ -107,6 +107,18 @@ static struct sk_buff *brcm_tag_xmit_ll(struct sk_buff *skb,
 	return skb;
 }
 
+/* Frames with this tag have one of these two layouts:
+ * -----------------------------------
+ * | MAC DA | MAC SA | 4b tag | Type | DSA_TAG_PROTO_BRCM
+ * -----------------------------------
+ * -----------------------------------
+ * | 4b tag | MAC DA | MAC SA | Type | DSA_TAG_PROTO_BRCM_PREPEND
+ * -----------------------------------
+ * In both cases, at receive time, skb->data points 2 bytes before the actual
+ * Ethernet type field and we have an offset of 4bytes between where skb->data
+ * and where the payload starts. So the same low-level receive function can be
+ * used.
+ */
 static struct sk_buff *brcm_tag_rcv_ll(struct sk_buff *skb,
 				       struct net_device *dev,
 				       struct packet_type *pt,
@@ -149,27 +161,6 @@ static struct sk_buff *brcm_tag_rcv_ll(struct sk_buff *skb,
 
 	return skb;
 }
-
-static int brcm_tag_flow_dissect(const struct sk_buff *skb, __be16 *proto,
-				 int *offset)
-{
-	/* We have been called on the DSA master network device after
-	 * eth_type_trans() which pulled the Ethernet header already.
-	 * Frames have one of these two layouts:
-	 * -----------------------------------
-	 * | MAC DA | MAC SA | 4b tag | Type | DSA_TAG_PROTO_BRCM
-	 * -----------------------------------
-	 * -----------------------------------
-	 * | 4b tag | MAC DA | MAC SA | Type | DSA_TAG_PROTO_BRCM_PREPEND
-	 * -----------------------------------
-	 * skb->data points 2 bytes before the actual Ethernet type field and
-	 * we have an offset of 4bytes between where skb->data and where the
-	 * payload starts.
-	 */
-	*offset = BRCM_TAG_LEN;
-	*proto = ((__be16 *)skb->data)[1];
-	return 0;
-}
 #endif
 
 #if IS_ENABLED(CONFIG_NET_DSA_TAG_BRCM)
@@ -205,7 +196,6 @@ static const struct dsa_device_ops brcm_netdev_ops = {
 	.xmit	= brcm_tag_xmit,
 	.rcv	= brcm_tag_rcv,
 	.overhead = BRCM_TAG_LEN,
-	.flow_dissect = brcm_tag_flow_dissect,
 };
 
 DSA_TAG_DRIVER(brcm_netdev_ops);
@@ -240,7 +230,6 @@ static const struct dsa_device_ops brcm_prepend_netdev_ops = {
 	.xmit	= brcm_tag_xmit_prepend,
 	.rcv	= brcm_tag_rcv_prepend,
 	.overhead = BRCM_TAG_LEN,
-	.flow_dissect = brcm_tag_flow_dissect,
 };
 
 DSA_TAG_DRIVER(brcm_prepend_netdev_ops);
diff --git a/net/dsa/tag_dsa.c b/net/dsa/tag_dsa.c
index 7ddec9794477..0b756fae68a5 100644
--- a/net/dsa/tag_dsa.c
+++ b/net/dsa/tag_dsa.c
@@ -142,20 +142,11 @@ static struct sk_buff *dsa_rcv(struct sk_buff *skb, struct net_device *dev,
 	return skb;
 }
 
-static int dsa_tag_flow_dissect(const struct sk_buff *skb, __be16 *proto,
-				int *offset)
-{
-	*offset = 4;
-	*proto = ((__be16 *)skb->data)[1];
-	return 0;
-}
-
 static const struct dsa_device_ops dsa_netdev_ops = {
 	.name	= "dsa",
 	.proto	= DSA_TAG_PROTO_DSA,
 	.xmit	= dsa_xmit,
 	.rcv	= dsa_rcv,
-	.flow_dissect   = dsa_tag_flow_dissect,
 	.overhead = DSA_HLEN,
 };
 
diff --git a/net/dsa/tag_edsa.c b/net/dsa/tag_edsa.c
index d6200ff98200..120614240319 100644
--- a/net/dsa/tag_edsa.c
+++ b/net/dsa/tag_edsa.c
@@ -192,20 +192,11 @@ static struct sk_buff *edsa_rcv(struct sk_buff *skb, struct net_device *dev,
 	return skb;
 }
 
-static int edsa_tag_flow_dissect(const struct sk_buff *skb, __be16 *proto,
-				 int *offset)
-{
-	*offset = 8;
-	*proto = ((__be16 *)skb->data)[3];
-	return 0;
-}
-
 static const struct dsa_device_ops edsa_netdev_ops = {
 	.name	= "edsa",
 	.proto	= DSA_TAG_PROTO_EDSA,
 	.xmit	= edsa_xmit,
 	.rcv	= edsa_rcv,
-	.flow_dissect   = edsa_tag_flow_dissect,
 	.overhead = EDSA_HLEN,
 };
 
diff --git a/net/dsa/tag_ksz.c b/net/dsa/tag_ksz.c
index bd1a3158d79a..945a9bd5ba35 100644
--- a/net/dsa/tag_ksz.c
+++ b/net/dsa/tag_ksz.c
@@ -237,6 +237,7 @@ static const struct dsa_device_ops ksz9893_netdev_ops = {
 	.xmit	= ksz9893_xmit,
 	.rcv	= ksz9477_rcv,
 	.overhead = KSZ_INGRESS_TAG_LEN,
+	.tail_tag = true,
 };
 
 DSA_TAG_DRIVER(ksz9893_netdev_ops);
diff --git a/net/dsa/tag_mtk.c b/net/dsa/tag_mtk.c
index f602fc758d68..4cdd9cf428fb 100644
--- a/net/dsa/tag_mtk.c
+++ b/net/dsa/tag_mtk.c
@@ -105,21 +105,11 @@ static struct sk_buff *mtk_tag_rcv(struct sk_buff *skb, struct net_device *dev,
 	return skb;
 }
 
-static int mtk_tag_flow_dissect(const struct sk_buff *skb, __be16 *proto,
-				int *offset)
-{
-	*offset = 4;
-	*proto = ((__be16 *)skb->data)[1];
-
-	return 0;
-}
-
 static const struct dsa_device_ops mtk_netdev_ops = {
 	.name		= "mtk",
 	.proto		= DSA_TAG_PROTO_MTK,
 	.xmit		= mtk_tag_xmit,
 	.rcv		= mtk_tag_rcv,
-	.flow_dissect	= mtk_tag_flow_dissect,
 	.overhead	= MTK_HDR_LEN,
 };
 
diff --git a/net/dsa/tag_ocelot.c b/net/dsa/tag_ocelot.c
index d1a7e224adff..ec16badb7812 100644
--- a/net/dsa/tag_ocelot.c
+++ b/net/dsa/tag_ocelot.c
@@ -141,10 +141,12 @@ static struct sk_buff *ocelot_xmit(struct sk_buff *skb,
 	struct dsa_switch *ds = dp->ds;
 	struct ocelot *ocelot = ds->priv;
 	struct ocelot_port *ocelot_port;
+	u8 *prefix, *injection;
 	u64 qos_class, rew_op;
-	u8 *injection;
+	int err;
 
-	if (unlikely(skb_cow_head(skb, OCELOT_TAG_LEN) < 0)) {
+	err = skb_cow_head(skb, OCELOT_TOTAL_TAG_LEN);
+	if (unlikely(err < 0)) {
 		netdev_err(netdev, "Cannot make room for tag.\n");
 		return NULL;
 	}
@@ -153,7 +155,10 @@ static struct sk_buff *ocelot_xmit(struct sk_buff *skb,
 
 	injection = skb_push(skb, OCELOT_TAG_LEN);
 
-	memcpy(injection, ocelot_port->xmit_template, OCELOT_TAG_LEN);
+	prefix = skb_push(skb, OCELOT_SHORT_PREFIX_LEN);
+
+	memcpy(prefix, ocelot_port->xmit_template, OCELOT_TOTAL_TAG_LEN);
+
 	/* Fix up the fields which are not statically determined
 	 * in the template
 	 */
@@ -187,11 +192,11 @@ static struct sk_buff *ocelot_rcv(struct sk_buff *skb,
 	 * so it points to the beginning of the frame.
 	 */
 	skb_push(skb, ETH_HLEN);
-	/* We don't care about the long prefix, it is just for easy entrance
+	/* We don't care about the short prefix, it is just for easy entrance
 	 * into the DSA master's RX filter. Discard it now by moving it into
 	 * the headroom.
 	 */
-	skb_pull(skb, OCELOT_LONG_PREFIX_LEN);
+	skb_pull(skb, OCELOT_SHORT_PREFIX_LEN);
 	/* And skb->data now points to the extraction frame header.
 	 * Keep a pointer to it.
 	 */
@@ -205,7 +210,7 @@ static struct sk_buff *ocelot_rcv(struct sk_buff *skb,
 	skb_pull(skb, ETH_HLEN);
 
 	/* Remove from inet csum the extraction header */
-	skb_postpull_rcsum(skb, start, OCELOT_LONG_PREFIX_LEN + OCELOT_TAG_LEN);
+	skb_postpull_rcsum(skb, start, OCELOT_TOTAL_TAG_LEN);
 
 	packing(extraction, &src_port,  46, 43, OCELOT_TAG_LEN, UNPACK, 0);
 	packing(extraction, &qos_class, 19, 17, OCELOT_TAG_LEN, UNPACK, 0);
@@ -231,7 +236,8 @@ static const struct dsa_device_ops ocelot_netdev_ops = {
 	.proto			= DSA_TAG_PROTO_OCELOT,
 	.xmit			= ocelot_xmit,
 	.rcv			= ocelot_rcv,
-	.overhead		= OCELOT_TAG_LEN + OCELOT_LONG_PREFIX_LEN,
+	.overhead		= OCELOT_TOTAL_TAG_LEN,
+	.promisc_on_master	= true,
 };
 
 MODULE_LICENSE("GPL v2");
diff --git a/net/dsa/tag_qca.c b/net/dsa/tag_qca.c
index 7066f5e697d7..1b9e8507112b 100644
--- a/net/dsa/tag_qca.c
+++ b/net/dsa/tag_qca.c
@@ -89,21 +89,11 @@ static struct sk_buff *qca_tag_rcv(struct sk_buff *skb, struct net_device *dev,
 	return skb;
 }
 
-static int qca_tag_flow_dissect(const struct sk_buff *skb, __be16 *proto,
-                                int *offset)
-{
-	*offset = QCA_HDR_LEN;
-	*proto = ((__be16 *)skb->data)[0];
-
-	return 0;
-}
-
 static const struct dsa_device_ops qca_netdev_ops = {
 	.name	= "qca",
 	.proto	= DSA_TAG_PROTO_QCA,
 	.xmit	= qca_tag_xmit,
 	.rcv	= qca_tag_rcv,
-	.flow_dissect = qca_tag_flow_dissect,
 	.overhead = QCA_HDR_LEN,
 };
 
diff --git a/net/dsa/tag_rtl4_a.c b/net/dsa/tag_rtl4_a.c
index 7b63010fa87b..2646abe5a69e 100644
--- a/net/dsa/tag_rtl4_a.c
+++ b/net/dsa/tag_rtl4_a.c
@@ -106,22 +106,11 @@ static struct sk_buff *rtl4a_tag_rcv(struct sk_buff *skb,
 	return skb;
 }
 
-static int rtl4a_tag_flow_dissect(const struct sk_buff *skb, __be16 *proto,
-				  int *offset)
-{
-	*offset = RTL4_A_HDR_LEN;
-	/* Skip past the tag and fetch the encapsulated Ethertype */
-	*proto = ((__be16 *)skb->data)[1];
-
-	return 0;
-}
-
 static const struct dsa_device_ops rtl4a_netdev_ops = {
 	.name	= "rtl4a",
 	.proto	= DSA_TAG_PROTO_RTL4_A,
 	.xmit	= rtl4a_tag_xmit,
 	.rcv	= rtl4a_tag_rcv,
-	.flow_dissect = rtl4a_tag_flow_dissect,
 	.overhead = RTL4_A_HDR_LEN,
 };
 module_dsa_tag_driver(rtl4a_netdev_ops);
diff --git a/net/dsa/tag_sja1105.c b/net/dsa/tag_sja1105.c
index 3710f9daa46d..50496013cdb7 100644
--- a/net/dsa/tag_sja1105.c
+++ b/net/dsa/tag_sja1105.c
@@ -346,6 +346,16 @@ static struct sk_buff *sja1105_rcv(struct sk_buff *skb,
 					      is_meta);
 }
 
+static void sja1105_flow_dissect(const struct sk_buff *skb, __be16 *proto,
+				 int *offset)
+{
+	/* No tag added for management frames, all ok */
+	if (unlikely(sja1105_is_link_local(skb)))
+		return;
+
+	dsa_tag_generic_flow_dissect(skb, proto, offset);
+}
+
 static const struct dsa_device_ops sja1105_netdev_ops = {
 	.name = "sja1105",
 	.proto = DSA_TAG_PROTO_SJA1105,
@@ -353,6 +363,8 @@ static const struct dsa_device_ops sja1105_netdev_ops = {
 	.rcv = sja1105_rcv,
 	.filter = sja1105_filter,
 	.overhead = VLAN_HLEN,
+	.flow_dissect = sja1105_flow_dissect,
+	.promisc_on_master = true,
 };
 
 MODULE_LICENSE("GPL v2");
diff --git a/net/dsa/tag_trailer.c b/net/dsa/tag_trailer.c
index 4f8ab62f0208..3a1cc24a4f0a 100644
--- a/net/dsa/tag_trailer.c
+++ b/net/dsa/tag_trailer.c
@@ -83,6 +83,7 @@ static const struct dsa_device_ops trailer_netdev_ops = {
 	.xmit	= trailer_xmit,
 	.rcv	= trailer_rcv,
 	.overhead = 4,
+	.tail_tag = true,
 };
 
 MODULE_LICENSE("GPL");