Merge branch 'net-dsa-microchip-add-support-for-credit-based-shaper'

Arun Ramadoss says:

====================
net: dsa: microchip: add support for credit based shaper

LAN937x switch family, KSZ9477, KSZ9567, KSZ9563 and KSZ8563 supports
the credit based shaper. But there were few difference between LAN937x and KSZ
switch like
- number of queues for LAN937x is 8 and for others it is 4.
- size of credit increment register for LAN937x is 24 and for other is 16-bit.
This patch series add the credit based shaper with common implementation for
LAN937x and KSZ swithes.
====================

Link: https://lore.kernel.org/r/20230120052135.32120-1-arun.ramadoss@microchip.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

9 files changed, 213 insertions, 26 deletions

diff --git a/drivers/net/dsa/microchip/ksz9477.c b/drivers/net/dsa/microchip/ksz9477.c
index 6178a96e389f..bf13d47c26cf 100644
--- a/drivers/net/dsa/microchip/ksz9477.c
+++ b/drivers/net/dsa/microchip/ksz9477.c
@@ -980,6 +980,22 @@ int ksz9477_set_ageing_time(struct ksz_device *dev, unsigned int msecs)
 	return ksz_write8(dev, REG_SW_LUE_CTRL_0, value);
 }
 
+void ksz9477_port_queue_split(struct ksz_device *dev, int port)
+{
+	u8 data;
+
+	if (dev->info->num_tx_queues == 8)
+		data = PORT_EIGHT_QUEUE;
+	else if (dev->info->num_tx_queues == 4)
+		data = PORT_FOUR_QUEUE;
+	else if (dev->info->num_tx_queues == 2)
+		data = PORT_TWO_QUEUE;
+	else
+		data = PORT_SINGLE_QUEUE;
+
+	ksz_prmw8(dev, port, REG_PORT_CTRL_0, PORT_QUEUE_SPLIT_MASK, data);
+}
+
 void ksz9477_port_setup(struct ksz_device *dev, int port, bool cpu_port)
 {
 	struct dsa_switch *ds = dev->ds;
@@ -991,6 +1007,8 @@ void ksz9477_port_setup(struct ksz_device *dev, int port, bool cpu_port)
 		ksz_port_cfg(dev, port, REG_PORT_CTRL_0, PORT_TAIL_TAG_ENABLE,
 			     true);
 
+	ksz9477_port_queue_split(dev, port);
+
 	ksz_port_cfg(dev, port, REG_PORT_CTRL_0, PORT_MAC_LOOPBACK, false);
 
 	/* set back pressure */
@@ -1166,6 +1184,13 @@ u32 ksz9477_get_port_addr(int port, int offset)
 	return PORT_CTRL_ADDR(port, offset);
 }
 
+int ksz9477_tc_cbs_set_cinc(struct ksz_device *dev, int port, u32 val)
+{
+	val = val >> 8;
+
+	return ksz_pwrite16(dev, port, REG_PORT_MTI_CREDIT_INCREMENT, val);
+}
+
 int ksz9477_switch_init(struct ksz_device *dev)
 {
 	u8 data8;
diff --git a/drivers/net/dsa/microchip/ksz9477.h b/drivers/net/dsa/microchip/ksz9477.h
index 7c5bb3032772..b6f7e3c46e3f 100644
--- a/drivers/net/dsa/microchip/ksz9477.h
+++ b/drivers/net/dsa/microchip/ksz9477.h
@@ -51,10 +51,12 @@ int ksz9477_mdb_del(struct ksz_device *dev, int port,
 		    const struct switchdev_obj_port_mdb *mdb, struct dsa_db db);
 int ksz9477_change_mtu(struct ksz_device *dev, int port, int mtu);
 void ksz9477_config_cpu_port(struct dsa_switch *ds);
+int ksz9477_tc_cbs_set_cinc(struct ksz_device *dev, int port, u32 val);
 int ksz9477_enable_stp_addr(struct ksz_device *dev);
 int ksz9477_reset_switch(struct ksz_device *dev);
 int ksz9477_dsa_init(struct ksz_device *dev);
 int ksz9477_switch_init(struct ksz_device *dev);
 void ksz9477_switch_exit(struct ksz_device *dev);
+void ksz9477_port_queue_split(struct ksz_device *dev, int port);
 
 #endif
diff --git a/drivers/net/dsa/microchip/ksz9477_reg.h b/drivers/net/dsa/microchip/ksz9477_reg.h
index cc457fa64939..cba3dba58bc3 100644
--- a/drivers/net/dsa/microchip/ksz9477_reg.h
+++ b/drivers/net/dsa/microchip/ksz9477_reg.h
@@ -850,7 +850,11 @@
 #define PORT_FORCE_TX_FLOW_CTRL		BIT(4)
 #define PORT_FORCE_RX_FLOW_CTRL		BIT(3)
 #define PORT_TAIL_TAG_ENABLE		BIT(2)
-#define PORT_QUEUE_SPLIT_ENABLE		0x3
+#define PORT_QUEUE_SPLIT_MASK		GENMASK(1, 0)
+#define PORT_EIGHT_QUEUE		0x3
+#define PORT_FOUR_QUEUE			0x2
+#define PORT_TWO_QUEUE			0x1
+#define PORT_SINGLE_QUEUE		0x0
 
 #define REG_PORT_CTRL_1			0x0021
 
@@ -1480,33 +1484,10 @@
 
 /* 9 - Shaping */
 
-#define REG_PORT_MTI_QUEUE_INDEX__4	0x0900
+#define REG_PORT_MTI_QUEUE_CTRL_0__4   0x0904
 
-#define REG_PORT_MTI_QUEUE_CTRL_0__4	0x0904
+#define MTI_PVID_REPLACE               BIT(0)
 
-#define MTI_PVID_REPLACE		BIT(0)
-
-#define REG_PORT_MTI_QUEUE_CTRL_0	0x0914
-
-#define MTI_SCHEDULE_MODE_M		0x3
-#define MTI_SCHEDULE_MODE_S		6
-#define MTI_SCHEDULE_STRICT_PRIO	0
-#define MTI_SCHEDULE_WRR		2
-#define MTI_SHAPING_M			0x3
-#define MTI_SHAPING_S			4
-#define MTI_SHAPING_OFF			0
-#define MTI_SHAPING_SRP			1
-#define MTI_SHAPING_TIME_AWARE		2
-
-#define REG_PORT_MTI_QUEUE_CTRL_1	0x0915
-
-#define MTI_TX_RATIO_M			(BIT(7) - 1)
-
-#define REG_PORT_MTI_QUEUE_CTRL_2__2	0x0916
-#define REG_PORT_MTI_HI_WATER_MARK	0x0916
-#define REG_PORT_MTI_QUEUE_CTRL_3__2	0x0918
-#define REG_PORT_MTI_LO_WATER_MARK	0x0918
-#define REG_PORT_MTI_QUEUE_CTRL_4__2	0x091A
 #define REG_PORT_MTI_CREDIT_INCREMENT	0x091A
 
 /* A - QM */
diff --git a/drivers/net/dsa/microchip/ksz_common.c b/drivers/net/dsa/microchip/ksz_common.c
index 28d26e80e256..46becc0382d6 100644
--- a/drivers/net/dsa/microchip/ksz_common.c
+++ b/drivers/net/dsa/microchip/ksz_common.c
@@ -23,6 +23,7 @@
 #include <linux/of_net.h>
 #include <linux/micrel_phy.h>
 #include <net/dsa.h>
+#include <net/pkt_cls.h>
 #include <net/switchdev.h>
 
 #include "ksz_common.h"
@@ -31,6 +32,10 @@
 #include "ksz9477.h"
 #include "lan937x.h"
 
+#define KSZ_CBS_ENABLE ((MTI_SCHEDULE_STRICT_PRIO << MTI_SCHEDULE_MODE_S) | \
+			(MTI_SHAPING_SRP << MTI_SHAPING_S))
+#define KSZ_CBS_DISABLE ((MTI_SCHEDULE_WRR << MTI_SCHEDULE_MODE_S) |\
+			 (MTI_SHAPING_OFF << MTI_SHAPING_S))
 #define MIB_COUNTER_NUM 0x20
 
 struct ksz_stats_raw {
@@ -250,6 +255,7 @@ static const struct ksz_dev_ops ksz9477_dev_ops = {
 	.change_mtu = ksz9477_change_mtu,
 	.phylink_mac_link_up = ksz9477_phylink_mac_link_up,
 	.config_cpu_port = ksz9477_config_cpu_port,
+	.tc_cbs_set_cinc = ksz9477_tc_cbs_set_cinc,
 	.enable_stp_addr = ksz9477_enable_stp_addr,
 	.reset = ksz9477_reset_switch,
 	.init = ksz9477_switch_init,
@@ -286,6 +292,7 @@ static const struct ksz_dev_ops lan937x_dev_ops = {
 	.change_mtu = lan937x_change_mtu,
 	.phylink_mac_link_up = ksz9477_phylink_mac_link_up,
 	.config_cpu_port = lan937x_config_cpu_port,
+	.tc_cbs_set_cinc = lan937x_tc_cbs_set_cinc,
 	.enable_stp_addr = ksz9477_enable_stp_addr,
 	.reset = lan937x_reset_switch,
 	.init = lan937x_switch_init,
@@ -1080,6 +1087,8 @@ const struct ksz_chip_data ksz_switch_chips[] = {
 		.cpu_ports = 0x07,	/* can be configured as cpu port */
 		.port_cnt = 3,		/* total port count */
 		.port_nirqs = 3,
+		.num_tx_queues = 4,
+		.tc_cbs_supported = true,
 		.ops = &ksz9477_dev_ops,
 		.mib_names = ksz9477_mib_names,
 		.mib_cnt = ARRAY_SIZE(ksz9477_mib_names),
@@ -1106,6 +1115,7 @@ const struct ksz_chip_data ksz_switch_chips[] = {
 		.num_statics = 8,
 		.cpu_ports = 0x10,	/* can be configured as cpu port */
 		.port_cnt = 5,		/* total cpu and user ports */
+		.num_tx_queues = 4,
 		.ops = &ksz8_dev_ops,
 		.ksz87xx_eee_link_erratum = true,
 		.mib_names = ksz9477_mib_names,
@@ -1144,6 +1154,7 @@ const struct ksz_chip_data ksz_switch_chips[] = {
 		.num_statics = 8,
 		.cpu_ports = 0x10,	/* can be configured as cpu port */
 		.port_cnt = 5,		/* total cpu and user ports */
+		.num_tx_queues = 4,
 		.ops = &ksz8_dev_ops,
 		.ksz87xx_eee_link_erratum = true,
 		.mib_names = ksz9477_mib_names,
@@ -1168,6 +1179,7 @@ const struct ksz_chip_data ksz_switch_chips[] = {
 		.num_statics = 8,
 		.cpu_ports = 0x10,	/* can be configured as cpu port */
 		.port_cnt = 5,		/* total cpu and user ports */
+		.num_tx_queues = 4,
 		.ops = &ksz8_dev_ops,
 		.ksz87xx_eee_link_erratum = true,
 		.mib_names = ksz9477_mib_names,
@@ -1192,6 +1204,7 @@ const struct ksz_chip_data ksz_switch_chips[] = {
 		.num_statics = 8,
 		.cpu_ports = 0x4,	/* can be configured as cpu port */
 		.port_cnt = 3,
+		.num_tx_queues = 4,
 		.ops = &ksz8_dev_ops,
 		.mib_names = ksz88xx_mib_names,
 		.mib_cnt = ARRAY_SIZE(ksz88xx_mib_names),
@@ -1213,6 +1226,8 @@ const struct ksz_chip_data ksz_switch_chips[] = {
 		.cpu_ports = 0x7F,	/* can be configured as cpu port */
 		.port_cnt = 7,		/* total physical port count */
 		.port_nirqs = 4,
+		.num_tx_queues = 4,
+		.tc_cbs_supported = true,
 		.ops = &ksz9477_dev_ops,
 		.phy_errata_9477 = true,
 		.mib_names = ksz9477_mib_names,
@@ -1245,6 +1260,7 @@ const struct ksz_chip_data ksz_switch_chips[] = {
 		.cpu_ports = 0x3F,	/* can be configured as cpu port */
 		.port_cnt = 6,		/* total physical port count */
 		.port_nirqs = 2,
+		.num_tx_queues = 4,
 		.ops = &ksz9477_dev_ops,
 		.phy_errata_9477 = true,
 		.mib_names = ksz9477_mib_names,
@@ -1277,6 +1293,7 @@ const struct ksz_chip_data ksz_switch_chips[] = {
 		.cpu_ports = 0x7F,	/* can be configured as cpu port */
 		.port_cnt = 7,		/* total physical port count */
 		.port_nirqs = 2,
+		.num_tx_queues = 4,
 		.ops = &ksz9477_dev_ops,
 		.phy_errata_9477 = true,
 		.mib_names = ksz9477_mib_names,
@@ -1307,6 +1324,7 @@ const struct ksz_chip_data ksz_switch_chips[] = {
 		.cpu_ports = 0x07,	/* can be configured as cpu port */
 		.port_cnt = 3,		/* total port count */
 		.port_nirqs = 2,
+		.num_tx_queues = 4,
 		.ops = &ksz9477_dev_ops,
 		.mib_names = ksz9477_mib_names,
 		.mib_cnt = ARRAY_SIZE(ksz9477_mib_names),
@@ -1332,6 +1350,8 @@ const struct ksz_chip_data ksz_switch_chips[] = {
 		.cpu_ports = 0x07,	/* can be configured as cpu port */
 		.port_cnt = 3,		/* total port count */
 		.port_nirqs = 3,
+		.num_tx_queues = 4,
+		.tc_cbs_supported = true,
 		.ops = &ksz9477_dev_ops,
 		.mib_names = ksz9477_mib_names,
 		.mib_cnt = ARRAY_SIZE(ksz9477_mib_names),
@@ -1357,6 +1377,8 @@ const struct ksz_chip_data ksz_switch_chips[] = {
 		.cpu_ports = 0x7F,	/* can be configured as cpu port */
 		.port_cnt = 7,		/* total physical port count */
 		.port_nirqs = 3,
+		.num_tx_queues = 4,
+		.tc_cbs_supported = true,
 		.ops = &ksz9477_dev_ops,
 		.phy_errata_9477 = true,
 		.mib_names = ksz9477_mib_names,
@@ -1387,6 +1409,8 @@ const struct ksz_chip_data ksz_switch_chips[] = {
 		.cpu_ports = 0x10,	/* can be configured as cpu port */
 		.port_cnt = 5,		/* total physical port count */
 		.port_nirqs = 6,
+		.num_tx_queues = 8,
+		.tc_cbs_supported = true,
 		.ops = &lan937x_dev_ops,
 		.mib_names = ksz9477_mib_names,
 		.mib_cnt = ARRAY_SIZE(ksz9477_mib_names),
@@ -1411,6 +1435,8 @@ const struct ksz_chip_data ksz_switch_chips[] = {
 		.cpu_ports = 0x30,	/* can be configured as cpu port */
 		.port_cnt = 6,		/* total physical port count */
 		.port_nirqs = 6,
+		.num_tx_queues = 8,
+		.tc_cbs_supported = true,
 		.ops = &lan937x_dev_ops,
 		.mib_names = ksz9477_mib_names,
 		.mib_cnt = ARRAY_SIZE(ksz9477_mib_names),
@@ -1435,6 +1461,8 @@ const struct ksz_chip_data ksz_switch_chips[] = {
 		.cpu_ports = 0x30,	/* can be configured as cpu port */
 		.port_cnt = 8,		/* total physical port count */
 		.port_nirqs = 6,
+		.num_tx_queues = 8,
+		.tc_cbs_supported = true,
 		.ops = &lan937x_dev_ops,
 		.mib_names = ksz9477_mib_names,
 		.mib_cnt = ARRAY_SIZE(ksz9477_mib_names),
@@ -1463,6 +1491,8 @@ const struct ksz_chip_data ksz_switch_chips[] = {
 		.cpu_ports = 0x38,	/* can be configured as cpu port */
 		.port_cnt = 5,		/* total physical port count */
 		.port_nirqs = 6,
+		.num_tx_queues = 8,
+		.tc_cbs_supported = true,
 		.ops = &lan937x_dev_ops,
 		.mib_names = ksz9477_mib_names,
 		.mib_cnt = ARRAY_SIZE(ksz9477_mib_names),
@@ -1491,6 +1521,8 @@ const struct ksz_chip_data ksz_switch_chips[] = {
 		.cpu_ports = 0x30,	/* can be configured as cpu port */
 		.port_cnt = 8,		/* total physical port count */
 		.port_nirqs = 6,
+		.num_tx_queues = 8,
+		.tc_cbs_supported = true,
 		.ops = &lan937x_dev_ops,
 		.mib_names = ksz9477_mib_names,
 		.mib_cnt = ARRAY_SIZE(ksz9477_mib_names),
@@ -2065,6 +2097,8 @@ static int ksz_setup(struct dsa_switch *ds)
 
 	dev->dev_ops->enable_stp_addr(dev);
 
+	ds->num_tx_queues = dev->info->num_tx_queues;
+
 	regmap_update_bits(dev->regmap[0], regs[S_MULTICAST_CTRL],
 			   MULTICAST_STORM_DISABLE, MULTICAST_STORM_DISABLE);
 
@@ -2958,6 +2992,101 @@ static int ksz_switch_detect(struct ksz_device *dev)
 	return 0;
 }
 
+/* Bandwidth is calculated by idle slope/transmission speed. Then the Bandwidth
+ * is converted to Hex-decimal using the successive multiplication method. On
+ * every step, integer part is taken and decimal part is carry forwarded.
+ */
+static int cinc_cal(s32 idle_slope, s32 send_slope, u32 *bw)
+{
+	u32 cinc = 0;
+	u32 txrate;
+	u32 rate;
+	u8 temp;
+	u8 i;
+
+	txrate = idle_slope - send_slope;
+
+	if (!txrate)
+		return -EINVAL;
+
+	rate = idle_slope;
+
+	/* 24 bit register */
+	for (i = 0; i < 6; i++) {
+		rate = rate * 16;
+
+		temp = rate / txrate;
+
+		rate %= txrate;
+
+		cinc = ((cinc << 4) | temp);
+	}
+
+	*bw = cinc;
+
+	return 0;
+}
+
+static int ksz_setup_tc_cbs(struct dsa_switch *ds, int port,
+			    struct tc_cbs_qopt_offload *qopt)
+{
+	struct ksz_device *dev = ds->priv;
+	int ret;
+	u32 bw;
+
+	if (!dev->info->tc_cbs_supported)
+		return -EOPNOTSUPP;
+
+	if (qopt->queue > dev->info->num_tx_queues)
+		return -EINVAL;
+
+	/* Queue Selection */
+	ret = ksz_pwrite32(dev, port, REG_PORT_MTI_QUEUE_INDEX__4, qopt->queue);
+	if (ret)
+		return ret;
+
+	if (!qopt->enable)
+		return ksz_pwrite8(dev, port, REG_PORT_MTI_QUEUE_CTRL_0,
+				   KSZ_CBS_DISABLE);
+
+	/* High Credit */
+	ret = ksz_pwrite16(dev, port, REG_PORT_MTI_HI_WATER_MARK,
+			   qopt->hicredit);
+	if (ret)
+		return ret;
+
+	/* Low Credit */
+	ret = ksz_pwrite16(dev, port, REG_PORT_MTI_LO_WATER_MARK,
+			   qopt->locredit);
+	if (ret)
+		return ret;
+
+	/* Credit Increment Register */
+	ret = cinc_cal(qopt->idleslope, qopt->sendslope, &bw);
+	if (ret)
+		return ret;
+
+	if (dev->dev_ops->tc_cbs_set_cinc) {
+		ret = dev->dev_ops->tc_cbs_set_cinc(dev, port, bw);
+		if (ret)
+			return ret;
+	}
+
+	return ksz_pwrite8(dev, port, REG_PORT_MTI_QUEUE_CTRL_0,
+			   KSZ_CBS_ENABLE);
+}
+
+static int ksz_setup_tc(struct dsa_switch *ds, int port,
+			enum tc_setup_type type, void *type_data)
+{
+	switch (type) {
+	case TC_SETUP_QDISC_CBS:
+		return ksz_setup_tc_cbs(ds, port, type_data);
+	default:
+		return -EOPNOTSUPP;
+	}
+}
+
 static const struct dsa_switch_ops ksz_switch_ops = {
 	.get_tag_protocol	= ksz_get_tag_protocol,
 	.connect_tag_protocol   = ksz_connect_tag_protocol,
@@ -3000,6 +3129,7 @@ static const struct dsa_switch_ops ksz_switch_ops = {
 	.port_hwtstamp_set	= ksz_hwtstamp_set,
 	.port_txtstamp		= ksz_port_txtstamp,
 	.port_rxtstamp		= ksz_port_rxtstamp,
+	.port_setup_tc		= ksz_setup_tc,
 };
 
 struct ksz_device *ksz_switch_alloc(struct device *base, void *priv)
diff --git a/drivers/net/dsa/microchip/ksz_common.h b/drivers/net/dsa/microchip/ksz_common.h
index 7260528e5c57..d2d5761d58e9 100644
--- a/drivers/net/dsa/microchip/ksz_common.h
+++ b/drivers/net/dsa/microchip/ksz_common.h
@@ -49,6 +49,8 @@ struct ksz_chip_data {
 	int cpu_ports;
 	int port_cnt;
 	u8 port_nirqs;
+	u8 num_tx_queues;
+	bool tc_cbs_supported;
 	const struct ksz_dev_ops *ops;
 	bool phy_errata_9477;
 	bool ksz87xx_eee_link_erratum;
@@ -353,6 +355,7 @@ struct ksz_dev_ops {
 				    struct phy_device *phydev, int speed,
 				    int duplex, bool tx_pause, bool rx_pause);
 	void (*setup_rgmii_delay)(struct ksz_device *dev, int port);
+	int (*tc_cbs_set_cinc)(struct ksz_device *dev, int port, u32 val);
 	void (*config_cpu_port)(struct dsa_switch *ds);
 	int (*enable_stp_addr)(struct ksz_device *dev);
 	int (*reset)(struct ksz_device *dev);
@@ -646,6 +649,24 @@ static inline int is_lan937x(struct ksz_device *dev)
 #define KSZ8_LEGAL_PACKET_SIZE		1518
 #define KSZ9477_MAX_FRAME_SIZE		9000
 
+/* CBS related registers */
+#define REG_PORT_MTI_QUEUE_INDEX__4	0x0900
+
+#define REG_PORT_MTI_QUEUE_CTRL_0	0x0914
+
+#define MTI_SCHEDULE_MODE_M		0x3
+#define MTI_SCHEDULE_MODE_S		6
+#define MTI_SCHEDULE_STRICT_PRIO	0
+#define MTI_SCHEDULE_WRR		2
+#define MTI_SHAPING_M			0x3
+#define MTI_SHAPING_S			4
+#define MTI_SHAPING_OFF			0
+#define MTI_SHAPING_SRP			1
+#define MTI_SHAPING_TIME_AWARE		2
+
+#define REG_PORT_MTI_HI_WATER_MARK	0x0916
+#define REG_PORT_MTI_LO_WATER_MARK	0x0918
+
 /* Regmap tables generation */
 #define KSZ_SPI_OP_RD		3
 #define KSZ_SPI_OP_WR		2
diff --git a/drivers/net/dsa/microchip/lan937x.h b/drivers/net/dsa/microchip/lan937x.h
index 8e9e66d6728d..3388d91dbc44 100644
--- a/drivers/net/dsa/microchip/lan937x.h
+++ b/drivers/net/dsa/microchip/lan937x.h
@@ -20,4 +20,5 @@ void lan937x_phylink_get_caps(struct ksz_device *dev, int port,
 			      struct phylink_config *config);
 void lan937x_setup_rgmii_delay(struct ksz_device *dev, int port);
 int lan937x_set_ageing_time(struct ksz_device *dev, unsigned int msecs);
+int lan937x_tc_cbs_set_cinc(struct ksz_device *dev, int port, u32 val);
 #endif
diff --git a/drivers/net/dsa/microchip/lan937x_main.c b/drivers/net/dsa/microchip/lan937x_main.c
index 06d3d0308cba..399a3905e6ca 100644
--- a/drivers/net/dsa/microchip/lan937x_main.c
+++ b/drivers/net/dsa/microchip/lan937x_main.c
@@ -15,6 +15,7 @@
 
 #include "lan937x_reg.h"
 #include "ksz_common.h"
+#include "ksz9477.h"
 #include "lan937x.h"
 
 static int lan937x_cfg(struct ksz_device *dev, u32 addr, u8 bits, bool set)
@@ -180,6 +181,9 @@ void lan937x_port_setup(struct ksz_device *dev, int port, bool cpu_port)
 		lan937x_port_cfg(dev, port, REG_PORT_CTRL_0,
 				 PORT_TAIL_TAG_ENABLE, true);
 
+	/* Enable the Port Queue split */
+	ksz9477_port_queue_split(dev, port);
+
 	/* set back pressure for half duplex */
 	lan937x_port_cfg(dev, port, REG_PORT_MAC_CTRL_1, PORT_BACK_PRESSURE,
 			 true);
@@ -336,6 +340,11 @@ void lan937x_setup_rgmii_delay(struct ksz_device *dev, int port)
 	}
 }
 
+int lan937x_tc_cbs_set_cinc(struct ksz_device *dev, int port, u32 val)
+{
+	return ksz_pwrite32(dev, port, REG_PORT_MTI_CREDIT_INCREMENT, val);
+}
+
 int lan937x_switch_init(struct ksz_device *dev)
 {
 	dev->port_mask = (1 << dev->info->port_cnt) - 1;
diff --git a/drivers/net/dsa/microchip/lan937x_reg.h b/drivers/net/dsa/microchip/lan937x_reg.h
index 5bc16a4c4441..45b606b6429f 100644
--- a/drivers/net/dsa/microchip/lan937x_reg.h
+++ b/drivers/net/dsa/microchip/lan937x_reg.h
@@ -185,6 +185,9 @@
 
 #define P_PRIO_CTRL			REG_PORT_MRI_PRIO_CTRL
 
+/* 9 - Shaping */
+#define REG_PORT_MTI_CREDIT_INCREMENT	0x091C
+
 /* The port number as per the datasheet */
 #define RGMII_2_PORT_NUM		5
 #define RGMII_1_PORT_NUM		6
diff --git a/net/dsa/tag_ksz.c b/net/dsa/tag_ksz.c
index 694478fe07d6..0eb1c7784c3d 100644
--- a/net/dsa/tag_ksz.c
+++ b/net/dsa/tag_ksz.c
@@ -178,6 +178,7 @@ MODULE_ALIAS_DSA_TAG_DRIVER(DSA_TAG_PROTO_KSZ8795, KSZ8795_NAME);
 #define KSZ9477_PTP_TAG_LEN		4
 #define KSZ9477_PTP_TAG_INDICATION	0x80
 
+#define KSZ9477_TAIL_TAG_PRIO		GENMASK(8, 7)
 #define KSZ9477_TAIL_TAG_OVERRIDE	BIT(9)
 #define KSZ9477_TAIL_TAG_LOOKUP		BIT(10)
 
@@ -269,6 +270,8 @@ static struct sk_buff *ksz_defer_xmit(struct dsa_port *dp, struct sk_buff *skb)
 static struct sk_buff *ksz9477_xmit(struct sk_buff *skb,
 				    struct net_device *dev)
 {
+	u16 queue_mapping = skb_get_queue_mapping(skb);
+	u8 prio = netdev_txq_to_tc(dev, queue_mapping);
 	struct dsa_port *dp = dsa_slave_to_port(dev);
 	__be16 *tag;
 	u8 *addr;
@@ -285,6 +288,8 @@ static struct sk_buff *ksz9477_xmit(struct sk_buff *skb,
 
 	val = BIT(dp->index);
 
+	val |= FIELD_PREP(KSZ9477_TAIL_TAG_PRIO, prio);
+
 	if (is_link_local_ether_addr(addr))
 		val |= KSZ9477_TAIL_TAG_OVERRIDE;
 
@@ -322,12 +327,15 @@ static const struct dsa_device_ops ksz9477_netdev_ops = {
 DSA_TAG_DRIVER(ksz9477_netdev_ops);
 MODULE_ALIAS_DSA_TAG_DRIVER(DSA_TAG_PROTO_KSZ9477, KSZ9477_NAME);
 
+#define KSZ9893_TAIL_TAG_PRIO		GENMASK(4, 3)
 #define KSZ9893_TAIL_TAG_OVERRIDE	BIT(5)
 #define KSZ9893_TAIL_TAG_LOOKUP		BIT(6)
 
 static struct sk_buff *ksz9893_xmit(struct sk_buff *skb,
 				    struct net_device *dev)
 {
+	u16 queue_mapping = skb_get_queue_mapping(skb);
+	u8 prio = netdev_txq_to_tc(dev, queue_mapping);
 	struct dsa_port *dp = dsa_slave_to_port(dev);
 	u8 *addr;
 	u8 *tag;
@@ -343,6 +351,8 @@ static struct sk_buff *ksz9893_xmit(struct sk_buff *skb,
 
 	*tag = BIT(dp->index);
 
+	*tag |= FIELD_PREP(KSZ9893_TAIL_TAG_PRIO, prio);
+
 	if (is_link_local_ether_addr(addr))
 		*tag |= KSZ9893_TAIL_TAG_OVERRIDE;
 
@@ -384,11 +394,14 @@ MODULE_ALIAS_DSA_TAG_DRIVER(DSA_TAG_PROTO_KSZ9893, KSZ9893_NAME);
 #define LAN937X_TAIL_TAG_BLOCKING_OVERRIDE	BIT(11)
 #define LAN937X_TAIL_TAG_LOOKUP			BIT(12)
 #define LAN937X_TAIL_TAG_VALID			BIT(13)
+#define LAN937X_TAIL_TAG_PRIO			GENMASK(10, 8)
 #define LAN937X_TAIL_TAG_PORT_MASK		7
 
 static struct sk_buff *lan937x_xmit(struct sk_buff *skb,
 				    struct net_device *dev)
 {
+	u16 queue_mapping = skb_get_queue_mapping(skb);
+	u8 prio = netdev_txq_to_tc(dev, queue_mapping);
 	struct dsa_port *dp = dsa_slave_to_port(dev);
 	const struct ethhdr *hdr = eth_hdr(skb);
 	__be16 *tag;
@@ -403,6 +416,8 @@ static struct sk_buff *lan937x_xmit(struct sk_buff *skb,
 
 	val = BIT(dp->index);
 
+	val |= FIELD_PREP(LAN937X_TAIL_TAG_PRIO, prio);
+
 	if (is_link_local_ether_addr(hdr->h_dest))
 		val |= LAN937X_TAIL_TAG_BLOCKING_OVERRIDE;