Merge branch 'net-stmmac-fpe-via-ethtool-tc'

Furong Xu says:

====================
net: stmmac: FPE via ethtool + tc

Move the Frame Preemption(FPE) over to the new standard API which uses
ethtool-mm/tc-mqprio/tc-taprio.
====================

Link: https://patch.msgid.link/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>

Author: kuba-moo

drivers/net/ethernet/stmicro/stmmac/dwmac4_core.c

@@ -58,10 +58,6 @@ static void dwmac4_core_init(struct mac_device_info *hw,
 	if (hw->pcs)
 		value |= GMAC_PCS_IRQ_DEFAULT;
 
-	/* Enable FPE interrupt */
-	if ((GMAC_HW_FEAT_FPESEL & readl(ioaddr + GMAC_HW_FEATURE3)) >> 26)
-		value |= GMAC_INT_FPE_EN;
-
 	writel(value, ioaddr + GMAC_INT_EN);
 
 	if (GMAC_INT_DEFAULT_ENABLE & GMAC_INT_TSIE)
@@ -1268,6 +1264,9 @@ const struct stmmac_ops dwmac410_ops = {
 	.fpe_configure = dwmac5_fpe_configure,
 	.fpe_send_mpacket = dwmac5_fpe_send_mpacket,
 	.fpe_irq_status = dwmac5_fpe_irq_status,
+	.fpe_get_add_frag_size = dwmac5_fpe_get_add_frag_size,
+	.fpe_set_add_frag_size = dwmac5_fpe_set_add_frag_size,
+	.fpe_map_preemption_class = dwmac5_fpe_map_preemption_class,
 	.add_hw_vlan_rx_fltr = dwmac4_add_hw_vlan_rx_fltr,
 	.del_hw_vlan_rx_fltr = dwmac4_del_hw_vlan_rx_fltr,
 	.restore_hw_vlan_rx_fltr = dwmac4_restore_hw_vlan_rx_fltr,
@@ -1320,6 +1319,9 @@ const struct stmmac_ops dwmac510_ops = {
 	.fpe_configure = dwmac5_fpe_configure,
 	.fpe_send_mpacket = dwmac5_fpe_send_mpacket,
 	.fpe_irq_status = dwmac5_fpe_irq_status,
+	.fpe_get_add_frag_size = dwmac5_fpe_get_add_frag_size,
+	.fpe_set_add_frag_size = dwmac5_fpe_set_add_frag_size,
+	.fpe_map_preemption_class = dwmac5_fpe_map_preemption_class,
 	.add_hw_vlan_rx_fltr = dwmac4_add_hw_vlan_rx_fltr,
 	.del_hw_vlan_rx_fltr = dwmac4_del_hw_vlan_rx_fltr,
 	.restore_hw_vlan_rx_fltr = dwmac4_restore_hw_vlan_rx_fltr,

drivers/net/ethernet/stmicro/stmmac/dwmac5.c

@@ -575,11 +575,11 @@ int dwmac5_flex_pps_config(void __iomem *ioaddr, int index,
 
 void dwmac5_fpe_configure(void __iomem *ioaddr, struct stmmac_fpe_cfg *cfg,
 			  u32 num_txq, u32 num_rxq,
-			  bool enable)
+			  bool tx_enable, bool pmac_enable)
 {
 	u32 value;
 
-	if (enable) {
+	if (tx_enable) {
 		cfg->fpe_csr = EFPE;
 		value = readl(ioaddr + GMAC_RXQ_CTRL1);
 		value &= ~GMAC_RXQCTRL_FPRQ;
@@ -589,6 +589,21 @@ void dwmac5_fpe_configure(void __iomem *ioaddr, struct stmmac_fpe_cfg *cfg,
 		cfg->fpe_csr = 0;
 	}
 	writel(cfg->fpe_csr, ioaddr + MAC_FPE_CTRL_STS);
+
+	value = readl(ioaddr + GMAC_INT_EN);
+
+	if (pmac_enable) {
+		if (!(value & GMAC_INT_FPE_EN)) {
+			/* Dummy read to clear any pending masked interrupts */
+			readl(ioaddr + MAC_FPE_CTRL_STS);
+
+			value |= GMAC_INT_FPE_EN;
+		}
+	} else {
+		value &= ~GMAC_INT_FPE_EN;
+	}
+
+	writel(value, ioaddr + GMAC_INT_EN);
 }
 
 int dwmac5_fpe_irq_status(void __iomem *ioaddr, struct net_device *dev)
@@ -605,22 +620,22 @@ int dwmac5_fpe_irq_status(void __iomem *ioaddr, struct net_device *dev)
 
 	if (value & TRSP) {
 		status |= FPE_EVENT_TRSP;
-		netdev_info(dev, "FPE: Respond mPacket is transmitted\n");
+		netdev_dbg(dev, "FPE: Respond mPacket is transmitted\n");
 	}
 
 	if (value & TVER) {
 		status |= FPE_EVENT_TVER;
-		netdev_info(dev, "FPE: Verify mPacket is transmitted\n");
+		netdev_dbg(dev, "FPE: Verify mPacket is transmitted\n");
 	}
 
 	if (value & RRSP) {
 		status |= FPE_EVENT_RRSP;
-		netdev_info(dev, "FPE: Respond mPacket is received\n");
+		netdev_dbg(dev, "FPE: Respond mPacket is received\n");
 	}
 
 	if (value & RVER) {
 		status |= FPE_EVENT_RVER;
-		netdev_info(dev, "FPE: Verify mPacket is received\n");
+		netdev_dbg(dev, "FPE: Verify mPacket is received\n");
 	}
 
 	return status;
@@ -638,3 +653,72 @@ void dwmac5_fpe_send_mpacket(void __iomem *ioaddr, struct stmmac_fpe_cfg *cfg,
 
 	writel(value, ioaddr + MAC_FPE_CTRL_STS);
 }
+
+int dwmac5_fpe_get_add_frag_size(const void __iomem *ioaddr)
+{
+	return FIELD_GET(DWMAC5_ADD_FRAG_SZ, readl(ioaddr + MTL_FPE_CTRL_STS));
+}
+
+void dwmac5_fpe_set_add_frag_size(void __iomem *ioaddr, u32 add_frag_size)
+{
+	u32 value;
+
+	value = readl(ioaddr + MTL_FPE_CTRL_STS);
+	writel(u32_replace_bits(value, add_frag_size, DWMAC5_ADD_FRAG_SZ),
+	       ioaddr + MTL_FPE_CTRL_STS);
+}
+
+#define ALG_ERR_MSG "TX algorithm SP is not suitable for one-to-many mapping"
+#define WEIGHT_ERR_MSG "TXQ weight %u differs across other TXQs in TC: [%u]"
+
+int dwmac5_fpe_map_preemption_class(struct net_device *ndev,
+				    struct netlink_ext_ack *extack, u32 pclass)
+{
+	u32 val, offset, count, queue_weight, preemptible_txqs = 0;
+	struct stmmac_priv *priv = netdev_priv(ndev);
+	u32 num_tc = ndev->num_tc;
+
+	if (!pclass)
+		goto update_mapping;
+
+	/* DWMAC CORE4+ can not program TC:TXQ mapping to hardware.
+	 *
+	 * Synopsys Databook:
+	 * "The number of Tx DMA channels is equal to the number of Tx queues,
+	 * and is direct one-to-one mapping."
+	 */
+	for (u32 tc = 0; tc < num_tc; tc++) {
+		count = ndev->tc_to_txq[tc].count;
+		offset = ndev->tc_to_txq[tc].offset;
+
+		if (pclass & BIT(tc))
+			preemptible_txqs |= GENMASK(offset + count - 1, offset);
+
+		/* This is 1:1 mapping, go to next TC */
+		if (count == 1)
+			continue;
+
+		if (priv->plat->tx_sched_algorithm == MTL_TX_ALGORITHM_SP) {
+			NL_SET_ERR_MSG_MOD(extack, ALG_ERR_MSG);
+			return -EINVAL;
+		}
+
+		queue_weight = priv->plat->tx_queues_cfg[offset].weight;
+
+		for (u32 i = 1; i < count; i++) {
+			if (priv->plat->tx_queues_cfg[offset + i].weight !=
+			    queue_weight) {
+				NL_SET_ERR_MSG_FMT_MOD(extack, WEIGHT_ERR_MSG,
+						       queue_weight, tc);
+				return -EINVAL;
+			}
+		}
+	}
+
+update_mapping:
+	val = readl(priv->ioaddr + MTL_FPE_CTRL_STS);
+	writel(u32_replace_bits(val, preemptible_txqs, DWMAC5_PREEMPTION_CLASS),
+	       priv->ioaddr + MTL_FPE_CTRL_STS);
+
+	return 0;
+}

drivers/net/ethernet/stmicro/stmmac/dwmac5.h

@@ -39,6 +39,12 @@
 #define MAC_PPSx_INTERVAL(x)		(0x00000b88 + ((x) * 0x10))
 #define MAC_PPSx_WIDTH(x)		(0x00000b8c + ((x) * 0x10))
 
+#define MTL_FPE_CTRL_STS		0x00000c90
+/* Preemption Classification */
+#define DWMAC5_PREEMPTION_CLASS		GENMASK(15, 8)
+/* Additional Fragment Size of preempted frames */
+#define DWMAC5_ADD_FRAG_SZ		GENMASK(1, 0)
+
 #define MTL_RXP_CONTROL_STATUS		0x00000ca0
 #define RXPI				BIT(31)
 #define NPE				GENMASK(23, 16)
@@ -104,10 +110,14 @@ int dwmac5_flex_pps_config(void __iomem *ioaddr, int index,
 			   u32 sub_second_inc, u32 systime_flags);
 void dwmac5_fpe_configure(void __iomem *ioaddr, struct stmmac_fpe_cfg *cfg,
 			  u32 num_txq, u32 num_rxq,
-			  bool enable);
+			  bool tx_enable, bool pmac_enable);
 void dwmac5_fpe_send_mpacket(void __iomem *ioaddr,
 			     struct stmmac_fpe_cfg *cfg,
 			     enum stmmac_mpacket_type type);
 int dwmac5_fpe_irq_status(void __iomem *ioaddr, struct net_device *dev);
+int dwmac5_fpe_get_add_frag_size(const void __iomem *ioaddr);
+void dwmac5_fpe_set_add_frag_size(void __iomem *ioaddr, u32 add_frag_size);
+int dwmac5_fpe_map_preemption_class(struct net_device *ndev,
+				    struct netlink_ext_ack *extack, u32 pclass);
 
 #endif /* __DWMAC5_H__ */

drivers/net/ethernet/stmicro/stmmac/dwxgmac2_core.c

@@ -1504,13 +1504,14 @@ static void dwxgmac2_set_arp_offload(struct mac_device_info *hw, bool en,
 	writel(value, ioaddr + XGMAC_RX_CONFIG);
 }
 
-static void dwxgmac3_fpe_configure(void __iomem *ioaddr, struct stmmac_fpe_cfg *cfg,
-				   u32 num_txq,
-				   u32 num_rxq, bool enable)
+static void dwxgmac3_fpe_configure(void __iomem *ioaddr,
+				   struct stmmac_fpe_cfg *cfg,
+				   u32 num_txq, u32 num_rxq,
+				   bool tx_enable, bool pmac_enable)
 {
 	u32 value;
 
-	if (!enable) {
+	if (!tx_enable) {
 		value = readl(ioaddr + XGMAC_FPE_CTRL_STS);
 
 		value &= ~XGMAC_EFPE;

drivers/net/ethernet/stmicro/stmmac/hwif.c

@@ -171,7 +171,7 @@ static const struct stmmac_hwif_entry {
 		.mac = &dwmac4_ops,
 		.hwtimestamp = &stmmac_ptp,
 		.mode = NULL,
-		.tc = &dwmac510_tc_ops,
+		.tc = &dwmac4_tc_ops,
 		.mmc = &dwmac_mmc_ops,
 		.est = &dwmac510_est_ops,
 		.setup = dwmac4_setup,
@@ -252,7 +252,7 @@ static const struct stmmac_hwif_entry {
 		.mac = &dwxgmac210_ops,
 		.hwtimestamp = &stmmac_ptp,
 		.mode = NULL,
-		.tc = &dwmac510_tc_ops,
+		.tc = &dwxgmac_tc_ops,
 		.mmc = &dwxgmac_mmc_ops,
 		.est = &dwmac510_est_ops,
 		.setup = dwxgmac2_setup,
@@ -273,7 +273,7 @@ static const struct stmmac_hwif_entry {
 		.mac = &dwxlgmac2_ops,
 		.hwtimestamp = &stmmac_ptp,
 		.mode = NULL,
-		.tc = &dwmac510_tc_ops,
+		.tc = &dwxgmac_tc_ops,
 		.mmc = &dwxgmac_mmc_ops,
 		.est = &dwmac510_est_ops,
 		.setup = dwxlgmac2_setup,

drivers/net/ethernet/stmicro/stmmac/hwif.h

@@ -7,6 +7,7 @@
 
 #include <linux/netdevice.h>
 #include <linux/stmmac.h>
+#include <net/pkt_cls.h>
 
 #define stmmac_do_void_callback(__priv, __module, __cname,  __arg0, __args...) \
 ({ \
@@ -28,6 +29,8 @@
 struct stmmac_extra_stats;
 struct stmmac_priv;
 struct stmmac_safety_stats;
+struct stmmac_fpe_cfg;
+enum stmmac_mpacket_type;
 struct dma_desc;
 struct dma_extended_desc;
 struct dma_edesc;
@@ -419,11 +422,16 @@ struct stmmac_ops {
 	void (*set_arp_offload)(struct mac_device_info *hw, bool en, u32 addr);
 	void (*fpe_configure)(void __iomem *ioaddr, struct stmmac_fpe_cfg *cfg,
 			      u32 num_txq, u32 num_rxq,
-			      bool enable);
+			      bool tx_enable, bool pmac_enable);
 	void (*fpe_send_mpacket)(void __iomem *ioaddr,
 				 struct stmmac_fpe_cfg *cfg,
 				 enum stmmac_mpacket_type type);
 	int (*fpe_irq_status)(void __iomem *ioaddr, struct net_device *dev);
+	int (*fpe_get_add_frag_size)(const void __iomem *ioaddr);
+	void (*fpe_set_add_frag_size)(void __iomem *ioaddr, u32 add_frag_size);
+	int (*fpe_map_preemption_class)(struct net_device *ndev,
+					struct netlink_ext_ack *extack,
+					u32 pclass);
 };
 
 #define stmmac_core_init(__priv, __args...) \
@@ -528,6 +536,12 @@ struct stmmac_ops {
 	stmmac_do_void_callback(__priv, mac, fpe_send_mpacket, __args)
 #define stmmac_fpe_irq_status(__priv, __args...) \
 	stmmac_do_callback(__priv, mac, fpe_irq_status, __args)
+#define stmmac_fpe_get_add_frag_size(__priv, __args...) \
+	stmmac_do_callback(__priv, mac, fpe_get_add_frag_size, __args)
+#define stmmac_fpe_set_add_frag_size(__priv, __args...) \
+	stmmac_do_void_callback(__priv, mac, fpe_set_add_frag_size, __args)
+#define stmmac_fpe_map_preemption_class(__priv, __args...) \
+	stmmac_do_void_callback(__priv, mac, fpe_map_preemption_class, __args)
 
 /* PTP and HW Timer helpers */
 struct stmmac_hwtimestamp {
@@ -615,6 +629,8 @@ struct stmmac_tc_ops {
 			 struct tc_etf_qopt_offload *qopt);
 	int (*query_caps)(struct stmmac_priv *priv,
 			  struct tc_query_caps_base *base);
+	int (*setup_mqprio)(struct stmmac_priv *priv,
+			    struct tc_mqprio_qopt_offload *qopt);
 };
 
 #define stmmac_tc_init(__priv, __args...) \
@@ -631,6 +647,8 @@ struct stmmac_tc_ops {
 	stmmac_do_callback(__priv, tc, setup_etf, __args)
 #define stmmac_tc_query_caps(__priv, __args...) \
 	stmmac_do_callback(__priv, tc, query_caps, __args)
+#define stmmac_tc_setup_mqprio(__priv, __args...) \
+	stmmac_do_callback(__priv, tc, setup_mqprio, __args)
 
 struct stmmac_counters;
 
@@ -674,7 +692,9 @@ extern const struct stmmac_dma_ops dwmac4_dma_ops;
 extern const struct stmmac_ops dwmac410_ops;
 extern const struct stmmac_dma_ops dwmac410_dma_ops;
 extern const struct stmmac_ops dwmac510_ops;
+extern const struct stmmac_tc_ops dwmac4_tc_ops;
 extern const struct stmmac_tc_ops dwmac510_tc_ops;
+extern const struct stmmac_tc_ops dwxgmac_tc_ops;
 extern const struct stmmac_ops dwxgmac210_ops;
 extern const struct stmmac_ops dwxlgmac2_ops;
 extern const struct stmmac_dma_ops dwxgmac210_dma_ops;

drivers/net/ethernet/stmicro/stmmac/stmmac.h

@@ -146,6 +146,32 @@ struct stmmac_channel {
 	u32 index;
 };
 
+/* FPE link-partner hand-shaking mPacket type */
+enum stmmac_mpacket_type {
+	MPACKET_VERIFY = 0,
+	MPACKET_RESPONSE = 1,
+};
+
+#define STMMAC_FPE_MM_MAX_VERIFY_RETRIES	3
+#define STMMAC_FPE_MM_MAX_VERIFY_TIME_MS	128
+
+struct stmmac_fpe_cfg {
+	/* Serialize access to MAC Merge state between ethtool requests
+	 * and link state updates.
+	 */
+	spinlock_t lock;
+
+	u32 fpe_csr;				/* MAC_FPE_CTRL_STS reg cache */
+
+	enum ethtool_mm_verify_status status;
+	struct timer_list verify_timer;
+	bool verify_enabled;
+	int verify_retries;
+	bool pmac_enabled;
+	u32 verify_time;
+	bool tx_enabled;
+};
+
 struct stmmac_tc_entry {
 	bool in_use;
 	bool in_hw;
@@ -339,11 +365,8 @@ struct stmmac_priv {
 	struct workqueue_struct *wq;
 	struct work_struct service_task;
 
-	/* Workqueue for handling FPE hand-shaking */
-	unsigned long fpe_task_state;
-	struct workqueue_struct *fpe_wq;
-	struct work_struct fpe_task;
-	char wq_name[IFNAMSIZ + 4];
+	/* Frame Preemption feature (FPE) */
+	struct stmmac_fpe_cfg fpe_cfg;
 
 	/* TC Handling */
 	unsigned int tc_entries_max;
@@ -397,7 +420,7 @@ bool stmmac_eee_init(struct stmmac_priv *priv);
 int stmmac_reinit_queues(struct net_device *dev, u32 rx_cnt, u32 tx_cnt);
 int stmmac_reinit_ringparam(struct net_device *dev, u32 rx_size, u32 tx_size);
 int stmmac_bus_clks_config(struct stmmac_priv *priv, bool enabled);
-void stmmac_fpe_handshake(struct stmmac_priv *priv, bool enable);
+void stmmac_fpe_apply(struct stmmac_priv *priv);
 
 static inline bool stmmac_xdp_is_enabled(struct stmmac_priv *priv)
 {