Merge branch 'DSA-mtu'

Vladimir Oltean says:

====================
Configure the MTU on DSA switches

This series adds support for configuring the MTU on front-panel switch
ports, while seamlessly adapting the CPU port and the DSA master to the
largest value plus the tagger overhead.

It also implements bridge MTU auto-normalization within the DSA core, as
resulted after the feedback of the implementation of this feature inside
the bridge driver in v2.

Support was added for quite a number of switches, in the hope that this
series would gain some traction:
 - sja1105
 - felix
 - vsc73xx
 - b53 and rest of the platform

V3 of this series was submitted here:
https://patchwork.ozlabs.org/cover/1262394/

V2 of this series was submitted here:
https://patchwork.ozlabs.org/cover/1261471/

V1 of this series was submitted here:
https://patchwork.ozlabs.org/cover/1199868/
====================

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

drivers/net/dsa/b53/b53_common.c

@@ -699,9 +699,6 @@ int b53_configure_vlan(struct dsa_switch *ds)
 		b53_write16(dev, B53_VLAN_PAGE,
 			    B53_VLAN_PORT_DEF_TAG(i), def_vid);
 
-	if (!is5325(dev) && !is5365(dev))
-		b53_set_jumbo(dev, dev->enable_jumbo, false);
-
 	return 0;
 }
 EXPORT_SYMBOL(b53_configure_vlan);
@@ -807,8 +804,6 @@ static int b53_phy_write16(struct dsa_switch *ds, int addr, int reg, u16 val)
 static int b53_reset_switch(struct b53_device *priv)
 {
 	/* reset vlans */
-	priv->enable_jumbo = false;
-
 	memset(priv->vlans, 0, sizeof(*priv->vlans) * priv->num_vlans);
 	memset(priv->ports, 0, sizeof(*priv->ports) * priv->num_ports);
 
@@ -2065,6 +2060,26 @@ int b53_set_mac_eee(struct dsa_switch *ds, int port, struct ethtool_eee *e)
 }
 EXPORT_SYMBOL(b53_set_mac_eee);
 
+static int b53_change_mtu(struct dsa_switch *ds, int port, int mtu)
+{
+	struct b53_device *dev = ds->priv;
+	bool enable_jumbo;
+	bool allow_10_100;
+
+	if (is5325(dev) || is5365(dev))
+		return -EOPNOTSUPP;
+
+	enable_jumbo = (mtu >= JMS_MIN_SIZE);
+	allow_10_100 = (dev->chip_id == BCM583XX_DEVICE_ID);
+
+	return b53_set_jumbo(dev, enable_jumbo, allow_10_100);
+}
+
+static int b53_get_max_mtu(struct dsa_switch *ds, int port)
+{
+	return JMS_MAX_SIZE;
+}
+
 static const struct dsa_switch_ops b53_switch_ops = {
 	.get_tag_protocol	= b53_get_tag_protocol,
 	.setup			= b53_setup,
@@ -2102,6 +2117,8 @@ static const struct dsa_switch_ops b53_switch_ops = {
 	.port_mdb_prepare	= b53_mdb_prepare,
 	.port_mdb_add		= b53_mdb_add,
 	.port_mdb_del		= b53_mdb_del,
+	.port_max_mtu		= b53_get_max_mtu,
+	.port_change_mtu	= b53_change_mtu,
 };
 
 struct b53_chip_data {

drivers/net/dsa/ocelot/felix.c

@@ -532,6 +532,7 @@ static int felix_setup(struct dsa_switch *ds)
 			 ANA_PGID_PGID_PGID(GENMASK(ocelot->num_phys_ports, 0)),
 			 ANA_PGID_PGID, PGID_UC);
 
+	ds->mtu_enforcement_ingress = true;
 	/* It looks like the MAC/PCS interrupt register - PM0_IEVENT (0x8040)
 	 * isn't instantiated for the Felix PF.
 	 * In-band AN may take a few ms to complete, so we need to poll.
@@ -609,6 +610,22 @@ static bool felix_txtstamp(struct dsa_switch *ds, int port,
 	return false;
 }
 
+static int felix_change_mtu(struct dsa_switch *ds, int port, int new_mtu)
+{
+	struct ocelot *ocelot = ds->priv;
+
+	ocelot_port_set_maxlen(ocelot, port, new_mtu);
+
+	return 0;
+}
+
+static int felix_get_max_mtu(struct dsa_switch *ds, int port)
+{
+	struct ocelot *ocelot = ds->priv;
+
+	return ocelot_get_max_mtu(ocelot, port);
+}
+
 static int felix_cls_flower_add(struct dsa_switch *ds, int port,
 				struct flow_cls_offload *cls, bool ingress)
 {
@@ -664,6 +681,8 @@ static const struct dsa_switch_ops felix_switch_ops = {
 	.port_hwtstamp_set	= felix_hwtstamp_set,
 	.port_rxtstamp		= felix_rxtstamp,
 	.port_txtstamp		= felix_txtstamp,
+	.port_change_mtu	= felix_change_mtu,
+	.port_max_mtu		= felix_get_max_mtu,
 	.cls_flower_add		= felix_cls_flower_add,
 	.cls_flower_del		= felix_cls_flower_del,
 	.cls_flower_stats	= felix_cls_flower_stats,

drivers/net/dsa/sja1105/sja1105.h

@@ -126,6 +126,7 @@ enum sja1105_reset_reason {
 	SJA1105_RX_HWTSTAMPING,
 	SJA1105_AGEING_TIME,
 	SJA1105_SCHEDULING,
+	SJA1105_BEST_EFFORT_POLICING,
 };
 
 int sja1105_static_config_reload(struct sja1105_private *priv,

drivers/net/dsa/sja1105/sja1105_main.c

@@ -519,12 +519,12 @@ static int sja1105_init_avb_params(struct sja1105_private *priv)
 #define SJA1105_RATE_MBPS(speed) (((speed) * 64000) / 1000)
 
 static void sja1105_setup_policer(struct sja1105_l2_policing_entry *policing,
-				  int index)
+				  int index, int mtu)
 {
 	policing[index].sharindx = index;
 	policing[index].smax = 65535; /* Burst size in bytes */
 	policing[index].rate = SJA1105_RATE_MBPS(1000);
-	policing[index].maxlen = ETH_FRAME_LEN + VLAN_HLEN + ETH_FCS_LEN;
+	policing[index].maxlen = mtu;
 	policing[index].partition = 0;
 }
 
@@ -556,12 +556,16 @@ static int sja1105_init_l2_policing(struct sja1105_private *priv)
 	 */
 	for (i = 0, k = 0; i < SJA1105_NUM_PORTS; i++) {
 		int bcast = (SJA1105_NUM_PORTS * SJA1105_NUM_TC) + i;
+		int mtu = VLAN_ETH_FRAME_LEN + ETH_FCS_LEN;
+
+		if (dsa_is_cpu_port(priv->ds, i))
+			mtu += VLAN_HLEN;
 
 		for (j = 0; j < SJA1105_NUM_TC; j++, k++)
-			sja1105_setup_policer(policing, k);
+			sja1105_setup_policer(policing, k, mtu);
 
 		/* Set up this port's policer for broadcast traffic */
-		sja1105_setup_policer(policing, bcast);
+		sja1105_setup_policer(policing, bcast, mtu);
 	}
 	return 0;
 }
@@ -1544,6 +1548,7 @@ static const char * const sja1105_reset_reasons[] = {
 	[SJA1105_RX_HWTSTAMPING] = "RX timestamping",
 	[SJA1105_AGEING_TIME] = "Ageing time",
 	[SJA1105_SCHEDULING] = "Time-aware scheduling",
+	[SJA1105_BEST_EFFORT_POLICING] = "Best-effort policing",
 };
 
 /* For situations where we need to change a setting at runtime that is only
@@ -1952,6 +1957,8 @@ static int sja1105_setup(struct dsa_switch *ds)
 	/* Advertise the 8 egress queues */
 	ds->num_tx_queues = SJA1105_NUM_TC;
 
+	ds->mtu_enforcement_ingress = true;
+
 	/* The DSA/switchdev model brings up switch ports in standalone mode by
 	 * default, and that means vlan_filtering is 0 since they're not under
 	 * a bridge, so it's safe to set up switch tagging at this time.
@@ -2120,6 +2127,39 @@ static int sja1105_set_ageing_time(struct dsa_switch *ds,
 	return sja1105_static_config_reload(priv, SJA1105_AGEING_TIME);
 }
 
+static int sja1105_change_mtu(struct dsa_switch *ds, int port, int new_mtu)
+{
+	int bcast = (SJA1105_NUM_PORTS * SJA1105_NUM_TC) + port;
+	struct sja1105_l2_policing_entry *policing;
+	struct sja1105_private *priv = ds->priv;
+	int tc;
+
+	new_mtu += VLAN_ETH_HLEN + ETH_FCS_LEN;
+
+	if (dsa_is_cpu_port(ds, port))
+		new_mtu += VLAN_HLEN;
+
+	policing = priv->static_config.tables[BLK_IDX_L2_POLICING].entries;
+
+	/* We set all 9 port policers to the same value, so just checking the
+	 * broadcast one is fine.
+	 */
+	if (policing[bcast].maxlen == new_mtu)
+		return 0;
+
+	for (tc = 0; tc < SJA1105_NUM_TC; tc++)
+		policing[port * SJA1105_NUM_TC + tc].maxlen = new_mtu;
+
+	policing[bcast].maxlen = new_mtu;
+
+	return sja1105_static_config_reload(priv, SJA1105_BEST_EFFORT_POLICING);
+}
+
+static int sja1105_get_max_mtu(struct dsa_switch *ds, int port)
+{
+	return 2043 - VLAN_ETH_HLEN - ETH_FCS_LEN;
+}
+
 static int sja1105_port_setup_tc(struct dsa_switch *ds, int port,
 				 enum tc_setup_type type,
 				 void *type_data)
@@ -2215,6 +2255,8 @@ static const struct dsa_switch_ops sja1105_switch_ops = {
 	.setup			= sja1105_setup,
 	.teardown		= sja1105_teardown,
 	.set_ageing_time	= sja1105_set_ageing_time,
+	.port_change_mtu	= sja1105_change_mtu,
+	.port_max_mtu		= sja1105_get_max_mtu,
 	.phylink_validate	= sja1105_phylink_validate,
 	.phylink_mac_link_state	= sja1105_mac_pcs_get_state,
 	.phylink_mac_config	= sja1105_mac_config,

drivers/net/dsa/vitesse-vsc73xx-core.c

@@ -664,16 +664,6 @@ static void vsc73xx_init_port(struct vsc73xx *vsc, int port)
 		      VSC73XX_MAC_CFG_TX_EN |
 		      VSC73XX_MAC_CFG_RX_EN);
 
-	/* Max length, we can do up to 9.6 KiB, so allow that.
-	 * According to application not "VSC7398 Jumbo Frames" setting
-	 * up the MTU to 9.6 KB does not affect the performance on standard
-	 * frames, so just enable it. It is clear from the application note
-	 * that "9.6 kilobytes" == 9600 bytes.
-	 */
-	vsc73xx_write(vsc, VSC73XX_BLOCK_MAC,
-		      port,
-		      VSC73XX_MAXLEN, 9600);
-
 	/* Flow control for the CPU port:
 	 * Use a zero delay pause frame when pause condition is left
 	 * Obey pause control frames
@@ -1030,6 +1020,24 @@ static void vsc73xx_get_ethtool_stats(struct dsa_switch *ds, int port,
 	}
 }
 
+static int vsc73xx_change_mtu(struct dsa_switch *ds, int port, int new_mtu)
+{
+	struct vsc73xx *vsc = ds->priv;
+
+	return vsc73xx_write(vsc, VSC73XX_BLOCK_MAC, port,
+			     VSC73XX_MAXLEN, new_mtu);
+}
+
+/* According to application not "VSC7398 Jumbo Frames" setting
+ * up the MTU to 9.6 KB does not affect the performance on standard
+ * frames. It is clear from the application note that
+ * "9.6 kilobytes" == 9600 bytes.
+ */
+static int vsc73xx_get_max_mtu(struct dsa_switch *ds, int port)
+{
+	return 9600;
+}
+
 static const struct dsa_switch_ops vsc73xx_ds_ops = {
 	.get_tag_protocol = vsc73xx_get_tag_protocol,
 	.setup = vsc73xx_setup,
@@ -1041,6 +1049,8 @@ static const struct dsa_switch_ops vsc73xx_ds_ops = {
 	.get_sset_count = vsc73xx_get_sset_count,
 	.port_enable = vsc73xx_port_enable,
 	.port_disable = vsc73xx_port_disable,
+	.port_change_mtu = vsc73xx_change_mtu,
+	.port_max_mtu = vsc73xx_get_max_mtu,
 };
 
 static int vsc73xx_gpio_get(struct gpio_chip *chip, unsigned int offset)

drivers/net/ethernet/broadcom/bgmac.c

@@ -1248,6 +1248,14 @@ static int bgmac_set_mac_address(struct net_device *net_dev, void *addr)
 	return 0;
 }
 
+static int bgmac_change_mtu(struct net_device *net_dev, int mtu)
+{
+	struct bgmac *bgmac = netdev_priv(net_dev);
+
+	bgmac_write(bgmac, BGMAC_RXMAX_LENGTH, 32 + mtu);
+	return 0;
+}
+
 static const struct net_device_ops bgmac_netdev_ops = {
 	.ndo_open		= bgmac_open,
 	.ndo_stop		= bgmac_stop,
@@ -1256,6 +1264,7 @@ static const struct net_device_ops bgmac_netdev_ops = {
 	.ndo_set_mac_address	= bgmac_set_mac_address,
 	.ndo_validate_addr	= eth_validate_addr,
 	.ndo_do_ioctl           = phy_do_ioctl_running,
+	.ndo_change_mtu		= bgmac_change_mtu,
 };
 
 /**************************************************
@@ -1530,6 +1539,9 @@ int bgmac_enet_probe(struct bgmac *bgmac)
 	net_dev->hw_features = net_dev->features;
 	net_dev->vlan_features = net_dev->features;
 
+	/* Omit FCS from max MTU size */
+	net_dev->max_mtu = BGMAC_RX_MAX_FRAME_SIZE - ETH_FCS_LEN;
+
 	err = register_netdev(bgmac->net_dev);
 	if (err) {
 		dev_err(bgmac->dev, "Cannot register net device\n");

drivers/net/ethernet/broadcom/bgmac.h

@@ -351,7 +351,7 @@
 #define BGMAC_DESC_CTL0_IOC			0x20000000	/* IRQ on complete */
 #define BGMAC_DESC_CTL0_EOF			0x40000000	/* End of frame */
 #define BGMAC_DESC_CTL0_SOF			0x80000000	/* Start of frame */
-#define BGMAC_DESC_CTL1_LEN			0x00001FFF
+#define BGMAC_DESC_CTL1_LEN			0x00003FFF
 
 #define BGMAC_PHY_NOREGS			BRCM_PSEUDO_PHY_ADDR
 #define BGMAC_PHY_MASK				0x1F
@@ -366,7 +366,8 @@
 #define BGMAC_RX_FRAME_OFFSET			30		/* There are 2 unused bytes between header and real data */
 #define BGMAC_RX_BUF_OFFSET			(NET_SKB_PAD + NET_IP_ALIGN - \
 						 BGMAC_RX_FRAME_OFFSET)
-#define BGMAC_RX_MAX_FRAME_SIZE			1536		/* Copied from b44/tg3 */
+/* Jumbo frame size with FCS */
+#define BGMAC_RX_MAX_FRAME_SIZE			9724
 #define BGMAC_RX_BUF_SIZE			(BGMAC_RX_FRAME_OFFSET + BGMAC_RX_MAX_FRAME_SIZE)
 #define BGMAC_RX_ALLOC_SIZE			(SKB_DATA_ALIGN(BGMAC_RX_BUF_SIZE + BGMAC_RX_BUF_OFFSET) + \
 						 SKB_DATA_ALIGN(sizeof(struct skb_shared_info)))

drivers/net/ethernet/mscc/ocelot.c

@@ -2185,13 +2185,25 @@ static int ocelot_init_timestamp(struct ocelot *ocelot)
 
 /* Configure the maximum SDU (L2 payload) on RX to the value specified in @sdu.
  * The length of VLAN tags is accounted for automatically via DEV_MAC_TAGS_CFG.
+ * In the special case that it's the NPI port that we're configuring, the
+ * length of the tag and optional prefix needs to be accounted for privately,
+ * in order to be able to sustain communication at the requested @sdu.
  */
-static void ocelot_port_set_maxlen(struct ocelot *ocelot, int port, size_t sdu)
+void ocelot_port_set_maxlen(struct ocelot *ocelot, int port, size_t sdu)
 {
 	struct ocelot_port *ocelot_port = ocelot->ports[port];
 	int maxlen = sdu + ETH_HLEN + ETH_FCS_LEN;
 	int atop_wm;
 
+	if (port == ocelot->npi) {
+		maxlen += OCELOT_TAG_LEN;
+
+		if (ocelot->inj_prefix == OCELOT_TAG_PREFIX_SHORT)
+			maxlen += OCELOT_SHORT_PREFIX_LEN;
+		else if (ocelot->inj_prefix == OCELOT_TAG_PREFIX_LONG)
+			maxlen += OCELOT_LONG_PREFIX_LEN;
+	}
+
 	ocelot_port_writel(ocelot_port, maxlen, DEV_MAC_MAXLEN_CFG);
 
 	/* Set Pause WM hysteresis
@@ -2209,6 +2221,24 @@ static void ocelot_port_set_maxlen(struct ocelot *ocelot, int port, size_t sdu)
 			 SYS_ATOP, port);
 	ocelot_write(ocelot, ocelot_wm_enc(atop_wm), SYS_ATOP_TOT_CFG);
 }
+EXPORT_SYMBOL(ocelot_port_set_maxlen);
+
+int ocelot_get_max_mtu(struct ocelot *ocelot, int port)
+{
+	int max_mtu = 65535 - ETH_HLEN - ETH_FCS_LEN;
+
+	if (port == ocelot->npi) {
+		max_mtu -= OCELOT_TAG_LEN;
+
+		if (ocelot->inj_prefix == OCELOT_TAG_PREFIX_SHORT)
+			max_mtu -= OCELOT_SHORT_PREFIX_LEN;
+		else if (ocelot->inj_prefix == OCELOT_TAG_PREFIX_LONG)
+			max_mtu -= OCELOT_LONG_PREFIX_LEN;
+	}
+
+	return max_mtu;
+}
+EXPORT_SYMBOL(ocelot_get_max_mtu);
 
 void ocelot_init_port(struct ocelot *ocelot, int port)
 {
@@ -2318,6 +2348,10 @@ void ocelot_configure_cpu(struct ocelot *ocelot, int npi,
 {
 	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
@@ -2330,19 +2364,10 @@ void ocelot_configure_cpu(struct ocelot *ocelot, int npi,
 			 ANA_PORT_PORT_CFG, cpu);
 
 	if (npi >= 0 && npi < ocelot->num_phys_ports) {
-		int sdu = ETH_DATA_LEN + OCELOT_TAG_LEN;
-
 		ocelot_write(ocelot, QSYS_EXT_CPU_CFG_EXT_CPUQ_MSK_M |
 			     QSYS_EXT_CPU_CFG_EXT_CPU_PORT(npi),
 			     QSYS_EXT_CPU_CFG);
 
-		if (injection == OCELOT_TAG_PREFIX_SHORT)
-			sdu += OCELOT_SHORT_PREFIX_LEN;
-		else if (injection == OCELOT_TAG_PREFIX_LONG)
-			sdu += OCELOT_LONG_PREFIX_LEN;
-
-		ocelot_port_set_maxlen(ocelot, npi, sdu);
-
 		/* Enable NPI port */
 		ocelot_write_rix(ocelot,
 				 QSYS_SWITCH_PORT_MODE_INGRESS_DROP_MODE |

drivers/net/phy/bcm-phy-lib.c

@@ -423,6 +423,28 @@ int bcm_phy_28nm_a0b0_afe_config_init(struct phy_device *phydev)
 }
 EXPORT_SYMBOL_GPL(bcm_phy_28nm_a0b0_afe_config_init);
 
+int bcm_phy_enable_jumbo(struct phy_device *phydev)
+{
+	int ret;
+
+	ret = bcm54xx_auxctl_read(phydev, MII_BCM54XX_AUXCTL_SHDWSEL_AUXCTL);
+	if (ret < 0)
+		return ret;
+
+	/* Enable extended length packet reception */
+	ret = bcm54xx_auxctl_write(phydev, MII_BCM54XX_AUXCTL_SHDWSEL_AUXCTL,
+				   ret | MII_BCM54XX_AUXCTL_ACTL_EXT_PKT_LEN);
+	if (ret < 0)
+		return ret;
+
+	/* Enable the elastic FIFO for raising the transmission limit from
+	 * 4.5KB to 10KB, at the expense of an additional 16 ns in propagation
+	 * latency.
+	 */
+	return phy_set_bits(phydev, MII_BCM54XX_ECR, MII_BCM54XX_ECR_FIFOE);
+}
+EXPORT_SYMBOL_GPL(bcm_phy_enable_jumbo);
+
 MODULE_DESCRIPTION("Broadcom PHY Library");
 MODULE_LICENSE("GPL v2");
 MODULE_AUTHOR("Broadcom Corporation");

drivers/net/phy/bcm-phy-lib.h

@@ -65,5 +65,6 @@ void bcm_phy_get_stats(struct phy_device *phydev, u64 *shadow,
 		       struct ethtool_stats *stats, u64 *data);
 void bcm_phy_r_rc_cal_reset(struct phy_device *phydev);
 int bcm_phy_28nm_a0b0_afe_config_init(struct phy_device *phydev);
+int bcm_phy_enable_jumbo(struct phy_device *phydev);
 
 #endif /* _LINUX_BCM_PHY_LIB_H */

drivers/net/phy/bcm7xxx.c

@@ -178,6 +178,10 @@ static int bcm7xxx_28nm_config_init(struct phy_device *phydev)
 		break;
 	}
 
+	if (ret)
+		return ret;
+
+	ret =  bcm_phy_enable_jumbo(phydev);
 	if (ret)
 		return ret;
 

include/linux/brcmphy.h

@@ -79,6 +79,7 @@
 #define MII_BCM54XX_ECR		0x10	/* BCM54xx extended control register */
 #define MII_BCM54XX_ECR_IM	0x1000	/* Interrupt mask */
 #define MII_BCM54XX_ECR_IF	0x0800	/* Interrupt force */
+#define MII_BCM54XX_ECR_FIFOE	0x0001	/* FIFO elasticity */
 
 #define MII_BCM54XX_ESR		0x11	/* BCM54xx extended status register */
 #define MII_BCM54XX_ESR_IS	0x1000	/* Interrupt status */
@@ -119,6 +120,7 @@
 #define MII_BCM54XX_AUXCTL_SHDWSEL_AUXCTL	0x00
 #define MII_BCM54XX_AUXCTL_ACTL_TX_6DB		0x0400
 #define MII_BCM54XX_AUXCTL_ACTL_SMDSP_ENA	0x0800
+#define MII_BCM54XX_AUXCTL_ACTL_EXT_PKT_LEN	0x4000
 
 #define MII_BCM54XX_AUXCTL_SHDWSEL_MISC			0x07
 #define MII_BCM54XX_AUXCTL_SHDWSEL_MISC_WIRESPEED_EN	0x0010

include/net/dsa.h

@@ -284,6 +284,12 @@ struct dsa_switch {
 	 */
 	bool			pcs_poll;
 
+	/* For switches that only have the MRU configurable. To ensure the
+	 * configured MTU is not exceeded, normalization of MRU on all bridged
+	 * interfaces is needed.
+	 */
+	bool			mtu_enforcement_ingress;
+
 	size_t num_ports;
 };
 
@@ -579,6 +585,16 @@ struct dsa_switch_ops {
 				     struct devlink_param_gset_ctx *ctx);
 	int	(*devlink_param_set)(struct dsa_switch *ds, u32 id,
 				     struct devlink_param_gset_ctx *ctx);
+
+	/*
+	 * MTU change functionality. Switches can also adjust their MRU through
+	 * this method. By MTU, one understands the SDU (L2 payload) length.
+	 * If the switch needs to account for the DSA tag on the CPU port, this
+	 * method needs to to do so privately.
+	 */
+	int	(*port_change_mtu)(struct dsa_switch *ds, int port,
+				   int new_mtu);
+	int	(*port_max_mtu)(struct dsa_switch *ds, int port);
 };
 
 #define DSA_DEVLINK_PARAM_DRIVER(_id, _name, _type, _cmodes)		\

include/soc/mscc/ocelot.h

@@ -517,6 +517,11 @@ struct ocelot {
 	 */
 	u8				num_phys_ports;
 
+	int				npi;
+
+	enum ocelot_tag_prefix		inj_prefix;
+	enum ocelot_tag_prefix		xtr_prefix;
+
 	u32				*lags;
 
 	struct list_head		multicast;
@@ -611,6 +616,8 @@ int ocelot_ptp_gettime64(struct ptp_clock_info *ptp, struct timespec64 *ts);
 int ocelot_port_add_txtstamp_skb(struct ocelot_port *ocelot_port,
 				 struct sk_buff *skb);
 void ocelot_get_txtstamp(struct ocelot *ocelot);
+void ocelot_port_set_maxlen(struct ocelot *ocelot, int port, size_t sdu);
+int ocelot_get_max_mtu(struct ocelot *ocelot, int port);
 int ocelot_cls_flower_replace(struct ocelot *ocelot, int port,
 			      struct flow_cls_offload *f, bool ingress);
 int ocelot_cls_flower_destroy(struct ocelot *ocelot, int port,

net/dsa/dsa2.c

@@ -18,8 +18,8 @@
 
 #include "dsa_priv.h"
 
-static LIST_HEAD(dsa_tree_list);
 static DEFINE_MUTEX(dsa2_mutex);
+LIST_HEAD(dsa_tree_list);
 
 static const struct devlink_ops dsa_devlink_ops = {
 };

net/dsa/dsa_priv.h

@@ -22,6 +22,7 @@ enum {
 	DSA_NOTIFIER_MDB_DEL,
 	DSA_NOTIFIER_VLAN_ADD,
 	DSA_NOTIFIER_VLAN_DEL,
+	DSA_NOTIFIER_MTU,
 };
 
 /* DSA_NOTIFIER_AGEING_TIME */
@@ -61,6 +62,14 @@ struct dsa_notifier_vlan_info {
 	int port;
 };
 
+/* DSA_NOTIFIER_MTU */
+struct dsa_notifier_mtu_info {
+	bool propagate_upstream;
+	int sw_index;
+	int port;
+	int mtu;
+};
+
 struct dsa_slave_priv {
 	/* Copy of CPU port xmit for faster access in slave transmit hot path */
 	struct sk_buff *	(*xmit)(struct sk_buff *skb,
@@ -127,6 +136,8 @@ int dsa_port_vlan_filtering(struct dsa_port *dp, bool vlan_filtering,
 			    struct switchdev_trans *trans);
 int dsa_port_ageing_time(struct dsa_port *dp, clock_t ageing_clock,
 			 struct switchdev_trans *trans);
+int dsa_port_mtu_change(struct dsa_port *dp, int new_mtu,
+			bool propagate_upstream);
 int dsa_port_fdb_add(struct dsa_port *dp, const unsigned char *addr,
 		     u16 vid);
 int dsa_port_fdb_del(struct dsa_port *dp, const unsigned char *addr,
@@ -183,4 +194,8 @@ dsa_slave_to_master(const struct net_device *dev)
 /* switch.c */
 int dsa_switch_register_notifier(struct dsa_switch *ds);
 void dsa_switch_unregister_notifier(struct dsa_switch *ds);
+
+/* dsa2.c */
+extern struct list_head dsa_tree_list;
+
 #endif

net/dsa/master.c

@@ -314,20 +314,6 @@ static const struct attribute_group dsa_group = {
 	.attrs	= dsa_slave_attrs,
 };
 
-static void dsa_master_set_mtu(struct net_device *dev, struct dsa_port *cpu_dp)
-{
-	unsigned int mtu = ETH_DATA_LEN + cpu_dp->tag_ops->overhead;
-	int err;
-
-	rtnl_lock();
-	if (mtu <= dev->max_mtu) {
-		err = dev_set_mtu(dev, mtu);
-		if (err)
-			netdev_dbg(dev, "Unable to set MTU to include for DSA overheads\n");
-	}
-	rtnl_unlock();
-}
-
 static void dsa_master_reset_mtu(struct net_device *dev)
 {
 	int err;
@@ -344,7 +330,12 @@ int dsa_master_setup(struct net_device *dev, struct dsa_port *cpu_dp)
 {
 	int ret;
 
-	dsa_master_set_mtu(dev,  cpu_dp);
+	rtnl_lock();
+	ret = dev_set_mtu(dev, ETH_DATA_LEN + cpu_dp->tag_ops->overhead);
+	rtnl_unlock();
+	if (ret)
+		netdev_warn(dev, "error %d setting MTU to include DSA overhead\n",
+			    ret);
 
 	/* If we use a tagging format that doesn't have an ethertype
 	 * field, make sure that all packets from this point on get

net/dsa/port.c

@@ -297,6 +297,19 @@ int dsa_port_mrouter(struct dsa_port *dp, bool mrouter,
 	return ds->ops->port_egress_floods(ds, port, true, mrouter);
 }
 
+int dsa_port_mtu_change(struct dsa_port *dp, int new_mtu,
+			bool propagate_upstream)
+{
+	struct dsa_notifier_mtu_info info = {
+		.sw_index = dp->ds->index,
+		.propagate_upstream = propagate_upstream,
+		.port = dp->index,
+		.mtu = new_mtu,
+	};
+
+	return dsa_port_notify(dp, DSA_NOTIFIER_MTU, &info);
+}
+
 int dsa_port_fdb_add(struct dsa_port *dp, const unsigned char *addr,
 		     u16 vid)
 {

net/dsa/slave.c

@@ -1218,6 +1218,208 @@ static int dsa_slave_vlan_rx_kill_vid(struct net_device *dev, __be16 proto,
 	return dsa_port_vid_del(dp, vid);
 }
 
+struct dsa_hw_port {
+	struct list_head list;
+	struct net_device *dev;
+	int old_mtu;
+};
+
+static int dsa_hw_port_list_set_mtu(struct list_head *hw_port_list, int mtu)
+{
+	const struct dsa_hw_port *p;
+	int err;
+
+	list_for_each_entry(p, hw_port_list, list) {
+		if (p->dev->mtu == mtu)
+			continue;
+
+		err = dev_set_mtu(p->dev, mtu);
+		if (err)
+			goto rollback;
+	}
+
+	return 0;
+
+rollback:
+	list_for_each_entry_continue_reverse(p, hw_port_list, list) {
+		if (p->dev->mtu == p->old_mtu)
+			continue;
+
+		if (dev_set_mtu(p->dev, p->old_mtu))
+			netdev_err(p->dev, "Failed to restore MTU\n");
+	}
+
+	return err;
+}
+
+static void dsa_hw_port_list_free(struct list_head *hw_port_list)
+{
+	struct dsa_hw_port *p, *n;
+
+	list_for_each_entry_safe(p, n, hw_port_list, list)
+		kfree(p);
+}
+
+/* Make the hardware datapath to/from @dev limited to a common MTU */
+void dsa_bridge_mtu_normalization(struct dsa_port *dp)
+{
+	struct list_head hw_port_list;
+	struct dsa_switch_tree *dst;
+	int min_mtu = ETH_MAX_MTU;
+	struct dsa_port *other_dp;
+	int err;
+
+	if (!dp->ds->mtu_enforcement_ingress)
+		return;
+
+	if (!dp->bridge_dev)
+		return;
+
+	INIT_LIST_HEAD(&hw_port_list);
+
+	/* Populate the list of ports that are part of the same bridge
+	 * as the newly added/modified port
+	 */
+	list_for_each_entry(dst, &dsa_tree_list, list) {
+		list_for_each_entry(other_dp, &dst->ports, list) {
+			struct dsa_hw_port *hw_port;
+			struct net_device *slave;
+
+			if (other_dp->type != DSA_PORT_TYPE_USER)
+				continue;
+
+			if (other_dp->bridge_dev != dp->bridge_dev)
+				continue;
+
+			if (!other_dp->ds->mtu_enforcement_ingress)
+				continue;
+
+			slave = other_dp->slave;
+
+			if (min_mtu > slave->mtu)
+				min_mtu = slave->mtu;
+
+			hw_port = kzalloc(sizeof(*hw_port), GFP_KERNEL);
+			if (!hw_port)
+				goto out;
+
+			hw_port->dev = slave;
+			hw_port->old_mtu = slave->mtu;
+
+			list_add(&hw_port->list, &hw_port_list);
+		}
+	}
+
+	/* Attempt to configure the entire hardware bridge to the newly added
+	 * interface's MTU first, regardless of whether the intention of the
+	 * user was to raise or lower it.
+	 */
+	err = dsa_hw_port_list_set_mtu(&hw_port_list, dp->slave->mtu);
+	if (!err)
+		goto out;
+
+	/* Clearly that didn't work out so well, so just set the minimum MTU on
+	 * all hardware bridge ports now. If this fails too, then all ports will
+	 * still have their old MTU rolled back anyway.
+	 */
+	dsa_hw_port_list_set_mtu(&hw_port_list, min_mtu);
+
+out:
+	dsa_hw_port_list_free(&hw_port_list);
+}
+
+static int dsa_slave_change_mtu(struct net_device *dev, int new_mtu)
+{
+	struct net_device *master = dsa_slave_to_master(dev);
+	struct dsa_port *dp = dsa_slave_to_port(dev);
+	struct dsa_slave_priv *p = netdev_priv(dev);
+	struct dsa_switch *ds = p->dp->ds;
+	struct dsa_port *cpu_dp;
+	int port = p->dp->index;
+	int largest_mtu = 0;
+	int new_master_mtu;
+	int old_master_mtu;
+	int mtu_limit;
+	int cpu_mtu;
+	int err, i;
+
+	if (!ds->ops->port_change_mtu)
+		return -EOPNOTSUPP;
+
+	for (i = 0; i < ds->num_ports; i++) {
+		int slave_mtu;
+
+		if (!dsa_is_user_port(ds, i))
+			continue;
+
+		/* During probe, this function will be called for each slave
+		 * device, while not all of them have been allocated. That's
+		 * ok, it doesn't change what the maximum is, so ignore it.
+		 */
+		if (!dsa_to_port(ds, i)->slave)
+			continue;
+
+		/* Pretend that we already applied the setting, which we
+		 * actually haven't (still haven't done all integrity checks)
+		 */
+		if (i == port)
+			slave_mtu = new_mtu;
+		else
+			slave_mtu = dsa_to_port(ds, i)->slave->mtu;
+
+		if (largest_mtu < slave_mtu)
+			largest_mtu = slave_mtu;
+	}
+
+	cpu_dp = dsa_to_port(ds, port)->cpu_dp;
+
+	mtu_limit = min_t(int, master->max_mtu, dev->max_mtu);
+	old_master_mtu = master->mtu;
+	new_master_mtu = largest_mtu + cpu_dp->tag_ops->overhead;
+	if (new_master_mtu > mtu_limit)
+		return -ERANGE;
+
+	/* If the master MTU isn't over limit, there's no need to check the CPU
+	 * MTU, since that surely isn't either.
+	 */
+	cpu_mtu = largest_mtu;
+
+	/* Start applying stuff */
+	if (new_master_mtu != old_master_mtu) {
+		err = dev_set_mtu(master, new_master_mtu);
+		if (err < 0)
+			goto out_master_failed;
+
+		/* We only need to propagate the MTU of the CPU port to
+		 * upstream switches.
+		 */
+		err = dsa_port_mtu_change(cpu_dp, cpu_mtu, true);
+		if (err)
+			goto out_cpu_failed;
+	}
+
+	err = dsa_port_mtu_change(dp, new_mtu, false);
+	if (err)
+		goto out_port_failed;
+
+	dev->mtu = new_mtu;
+
+	dsa_bridge_mtu_normalization(dp);
+
+	return 0;
+
+out_port_failed:
+	if (new_master_mtu != old_master_mtu)
+		dsa_port_mtu_change(cpu_dp, old_master_mtu -
+				    cpu_dp->tag_ops->overhead,
+				    true);
+out_cpu_failed:
+	if (new_master_mtu != old_master_mtu)
+		dev_set_mtu(master, old_master_mtu);
+out_master_failed:
+	return err;
+}
+
 static const struct ethtool_ops dsa_slave_ethtool_ops = {
 	.get_drvinfo		= dsa_slave_get_drvinfo,
 	.get_regs_len		= dsa_slave_get_regs_len,
@@ -1295,6 +1497,7 @@ static const struct net_device_ops dsa_slave_netdev_ops = {
 	.ndo_vlan_rx_add_vid	= dsa_slave_vlan_rx_add_vid,
 	.ndo_vlan_rx_kill_vid	= dsa_slave_vlan_rx_kill_vid,
 	.ndo_get_devlink_port	= dsa_slave_get_devlink_port,
+	.ndo_change_mtu		= dsa_slave_change_mtu,
 };
 
 static struct device_type dsa_type = {
@@ -1465,6 +1668,9 @@ int dsa_slave_create(struct dsa_port *port)
 	slave_dev->priv_flags |= IFF_NO_QUEUE;
 	slave_dev->netdev_ops = &dsa_slave_netdev_ops;
 	slave_dev->min_mtu = 0;
+	if (ds->ops->port_max_mtu)
+		slave_dev->max_mtu = ds->ops->port_max_mtu(ds, port->index);
+	else
 		slave_dev->max_mtu = ETH_MAX_MTU;
 	SET_NETDEV_DEVTYPE(slave_dev, &dsa_type);
 
@@ -1483,6 +1689,15 @@ int dsa_slave_create(struct dsa_port *port)
 	p->xmit = cpu_dp->tag_ops->xmit;
 	port->slave = slave_dev;
 
+	rtnl_lock();
+	ret = dsa_slave_change_mtu(slave_dev, ETH_DATA_LEN);
+	rtnl_unlock();
+	if (ret && ret != -EOPNOTSUPP) {
+		dev_err(ds->dev, "error %d setting MTU on port %d\n",
+			ret, port->index);
+		goto out_free;
+	}
+
 	netif_carrier_off(slave_dev);
 
 	ret = dsa_slave_phy_setup(slave_dev);
@@ -1545,6 +1760,8 @@ static int dsa_slave_changeupper(struct net_device *dev,
 	if (netif_is_bridge_master(info->upper_dev)) {
 		if (info->linking) {
 			err = dsa_port_bridge_join(dp, info->upper_dev);
+			if (!err)
+				dsa_bridge_mtu_normalization(dp);
 			err = notifier_from_errno(err);
 		} else {
 			dsa_port_bridge_leave(dp, info->upper_dev);

net/dsa/switch.c

@@ -52,6 +52,40 @@ static int dsa_switch_ageing_time(struct dsa_switch *ds,
 	return 0;
 }
 
+static bool dsa_switch_mtu_match(struct dsa_switch *ds, int port,
+				 struct dsa_notifier_mtu_info *info)
+{
+	if (ds->index == info->sw_index)
+		return (port == info->port) || dsa_is_dsa_port(ds, port);
+
+	if (!info->propagate_upstream)
+		return false;
+
+	if (dsa_is_dsa_port(ds, port) || dsa_is_cpu_port(ds, port))
+		return true;
+
+	return false;
+}
+
+static int dsa_switch_mtu(struct dsa_switch *ds,
+			  struct dsa_notifier_mtu_info *info)
+{
+	int port, ret;
+
+	if (!ds->ops->port_change_mtu)
+		return -EOPNOTSUPP;
+
+	for (port = 0; port < ds->num_ports; port++) {
+		if (dsa_switch_mtu_match(ds, port, info)) {
+			ret = ds->ops->port_change_mtu(ds, port, info->mtu);
+			if (ret)
+				return ret;
+		}
+	}
+
+	return 0;
+}
+
 static int dsa_switch_bridge_join(struct dsa_switch *ds,
 				  struct dsa_notifier_bridge_info *info)
 {
@@ -328,6 +362,9 @@ static int dsa_switch_event(struct notifier_block *nb,
 	case DSA_NOTIFIER_VLAN_DEL:
 		err = dsa_switch_vlan_del(ds, info);
 		break;
+	case DSA_NOTIFIER_MTU:
+		err = dsa_switch_mtu(ds, info);
+		break;
 	default:
 		err = -EOPNOTSUPP;
 		break;