Merge tag 'net-next-6.16' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next

Pull networking updates from Paolo Abeni:
 "Core:

   - Implement the Device Memory TCP transmit path, allowing zero-copy
     data transmission on top of TCP from e.g. GPU memory to the wire.

   - Move all the IPv6 routing tables management outside the RTNL scope,
     under its own lock and RCU. The route control path is now 3x times
     faster.

   - Convert queue related netlink ops to instance lock, reducing again
     the scope of the RTNL lock. This improves the control plane
     scalability.

   - Refactor the software crc32c implementation, removing unneeded
     abstraction layers and improving significantly the related
     micro-benchmarks.

   - Optimize the GRO engine for UDP-tunneled traffic, for a 10%
     performance improvement in related stream tests.

   - Cover more per-CPU storage with local nested BH locking; this is a
     prep work to remove the current per-CPU lock in local_bh_disable()
     on PREMPT_RT.

   - Introduce and use nlmsg_payload helper, combining buffer bounds
     verification with accessing payload carried by netlink messages.

  Netfilter:

   - Rewrite the procfs conntrack table implementation, improving
     considerably the dump performance. A lot of user-space tools still
     use this interface.

   - Implement support for wildcard netdevice in netdev basechain and
     flowtables.

   - Integrate conntrack information into nft trace infrastructure.

   - Export set count and backend name to userspace, for better
     introspection.

  BPF:

   - BPF qdisc support: BPF-qdisc can be implemented with BPF struct_ops
     programs and can be controlled in similar way to traditional qdiscs
     using the "tc qdisc" command.

   - Refactor the UDP socket iterator, addressing long standing issues
     WRT duplicate hits or missed sockets.

  Protocols:

   - Improve TCP receive buffer auto-tuning and increase the default
     upper bound for the receive buffer; overall this improves the
     single flow maximum thoughput on 200Gbs link by over 60%.

   - Add AFS GSSAPI security class to AF_RXRPC; it provides transport
     security for connections to the AFS fileserver and VL server.

   - Improve TCP multipath routing, so that the sources address always
     matches the nexthop device.

   - Introduce SO_PASSRIGHTS for AF_UNIX, to allow disabling SCM_RIGHTS,
     and thus preventing DoS caused by passing around problematic FDs.

   - Retire DCCP socket. DCCP only receives updates for bugs, and major
     distros disable it by default. Its removal allows for better
     organisation of TCP fields to reduce the number of cache lines hit
     in the fast path.

   - Extend TCP drop-reason support to cover PAWS checks.

  Driver API:

   - Reorganize PTP ioctl flag support to require an explicit opt-in for
     the drivers, avoiding the problem of drivers not rejecting new
     unsupported flags.

   - Converted several device drivers to timestamping APIs.

   - Introduce per-PHY ethtool dump helpers, improving the support for
     dump operations targeting PHYs.

  Tests and tooling:

   - Add support for classic netlink in user space C codegen, so that
     ynl-c can now read, create and modify links, routes addresses and
     qdisc layer configuration.

   - Add ynl sub-types for binary attributes, allowing ynl-c to output
     known struct instead of raw binary data, clarifying the classic
     netlink output.

   - Extend MPTCP selftests to improve the code-coverage.

   - Add tests for XDP tail adjustment in AF_XDP.

  New hardware / drivers:

   - OpenVPN virtual driver: offload OpenVPN data channels processing to
     the kernel-space, increasing the data transfer throughput WRT the
     user-space implementation.

   - Renesas glue driver for the gigabit ethernet RZ/V2H(P) SoC.

   - Broadcom asp-v3.0 ethernet driver.

   - AMD Renoir ethernet device.

   - ReakTek MT9888 2.5G ethernet PHY driver.

   - Aeonsemi 10G C45 PHYs driver.

  Drivers:

   - Ethernet high-speed NICs:
       - nVidia/Mellanox (mlx5):
           - refactor the steering table handling to significantly
             reduce the amount of memory used
           - add support for complex matches in H/W flow steering
           - improve flow streeing error handling
           - convert to netdev instance locking
       - Intel (100G, ice, igb, ixgbe, idpf):
           - ice: add switchdev support for LLDP traffic over VF
           - ixgbe: add firmware manipulation and regions devlink support
           - igb: introduce support for frame transmission premption
           - igb: adds persistent NAPI configuration
           - idpf: introduce RDMA support
           - idpf: add initial PTP support
       - Meta (fbnic):
           - extend hardware stats coverage
           - add devlink dev flash support
       - Broadcom (bnxt):
           - add support for RX-side device memory TCP
       - Wangxun (txgbe):
           - implement support for udp tunnel offload
           - complete PTP and SRIOV support for AML 25G/10G devices

   - Ethernet NICs embedded and virtual:
       - Google (gve):
           - add device memory TCP TX support
       - Amazon (ena):
           - support persistent per-NAPI config
       - Airoha:
           - add H/W support for L2 traffic offload
           - add per flow stats for flow offloading
       - RealTek (rtl8211): add support for WoL magic packet
       - Synopsys (stmmac):
           - dwmac-socfpga 1000BaseX support
           - add Loongson-2K3000 support
           - introduce support for hardware-accelerated VLAN stripping
       - Broadcom (bcmgenet):
           - expose more H/W stats
       - Freescale (enetc, dpaa2-eth):
           - enetc: add MAC filter, VLAN filter RSS and loopback support
           - dpaa2-eth: convert to H/W timestamping APIs
       - vxlan: convert FDB table to rhashtable, for better scalabilty
       - veth: apply qdisc backpressure on full ring to reduce TX drops

   - Ethernet switches:
       - Microchip (kzZ88x3): add ETS scheduler support

   - Ethernet PHYs:
       - RealTek (rtl8211):
           - add support for WoL magic packet
           - add support for PHY LEDs

   - CAN:
       - Adds RZ/G3E CANFD support to the rcar_canfd driver.
       - Preparatory work for CAN-XL support.
       - Add self-tests framework with support for CAN physical interfaces.

   - WiFi:
       - mac80211:
           - scan improvements with multi-link operation (MLO)
       - Qualcomm (ath12k):
           - enable AHB support for IPQ5332
           - add monitor interface support to QCN9274
           - add multi-link operation support to WCN7850
           - add 802.11d scan offload support to WCN7850
           - monitor mode for WCN7850, better 6 GHz regulatory
       - Qualcomm (ath11k):
           - restore hibernation support
       - MediaTek (mt76):
           - WiFi-7 improvements
           - implement support for mt7990
       - Intel (iwlwifi):
           - enhanced multi-link single-radio (EMLSR) support on 5 GHz links
           - rework device configuration
       - RealTek (rtw88):
           - improve throughput for RTL8814AU
       - RealTek (rtw89):
           - add multi-link operation support
           - STA/P2P concurrency improvements
           - support different SAR configs by antenna

   - Bluetooth:
       - introduce HCI Driver protocol
       - btintel_pcie: do not generate coredump for diagnostic events
       - btusb: add HCI Drv commands for configuring altsetting
       - btusb: add RTL8851BE device 0x0bda:0xb850
       - btusb: add new VID/PID 13d3/3584 for MT7922
       - btusb: add new VID/PID 13d3/3630 and 13d3/3613 for MT7925
       - btnxpuart: implement host-wakeup feature"

* tag 'net-next-6.16' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next: (1611 commits)
  selftests/bpf: Fix bpf selftest build warning
  selftests: netfilter: Fix skip of wildcard interface test
  net: phy: mscc: Stop clearing the the UDPv4 checksum for L2 frames
  net: openvswitch: Fix the dead loop of MPLS parse
  calipso: Don't call calipso functions for AF_INET sk.
  selftests/tc-testing: Add a test for HFSC eltree double add with reentrant enqueue behaviour on netem
  net_sched: hfsc: Address reentrant enqueue adding class to eltree twice
  octeontx2-pf: QOS: Refactor TC_HTB_LEAF_DEL_LAST callback
  octeontx2-pf: QOS: Perform cache sync on send queue teardown
  net: mana: Add support for Multi Vports on Bare metal
  net: devmem: ncdevmem: remove unused variable
  net: devmem: ksft: upgrade rx test to send 1K data
  net: devmem: ksft: add 5 tuple FS support
  net: devmem: ksft: add exit_wait to make rx test pass
  net: devmem: ksft: add ipv4 support
  net: devmem: preserve sockc_err
  page_pool: fix ugly page_pool formatting
  net: devmem: move list_add to net_devmem_bind_dmabuf.
  selftests: netfilter: nft_queue.sh: include file transfer duration in log message
  net: phy: mscc: Fix memory leak when using one step timestamping
  ...

1 files changed, 421 insertions, 71 deletions

diff --git a/drivers/net/ethernet/airoha/airoha_eth.c b/drivers/net/ethernet/airoha/airoha_eth.c
index 1e9ab65218ff..d1d3b854361e 100644
--- a/drivers/net/ethernet/airoha/airoha_eth.c
+++ b/drivers/net/ethernet/airoha/airoha_eth.c
@@ -5,6 +5,7 @@
  */
 #include <linux/of.h>
 #include <linux/of_net.h>
+#include <linux/of_reserved_mem.h>
 #include <linux/platform_device.h>
 #include <linux/tcp.h>
 #include <linux/u64_stats_sync.h>
@@ -34,46 +35,40 @@ u32 airoha_rmw(void __iomem *base, u32 offset, u32 mask, u32 val)
 	return val;
 }
 
-static void airoha_qdma_set_irqmask(struct airoha_qdma *qdma, int index,
-				    u32 clear, u32 set)
+static void airoha_qdma_set_irqmask(struct airoha_irq_bank *irq_bank,
+				    int index, u32 clear, u32 set)
 {
+	struct airoha_qdma *qdma = irq_bank->qdma;
+	int bank = irq_bank - &qdma->irq_banks[0];
 	unsigned long flags;
 
-	if (WARN_ON_ONCE(index >= ARRAY_SIZE(qdma->irqmask)))
+	if (WARN_ON_ONCE(index >= ARRAY_SIZE(irq_bank->irqmask)))
 		return;
 
-	spin_lock_irqsave(&qdma->irq_lock, flags);
+	spin_lock_irqsave(&irq_bank->irq_lock, flags);
 
-	qdma->irqmask[index] &= ~clear;
-	qdma->irqmask[index] |= set;
-	airoha_qdma_wr(qdma, REG_INT_ENABLE(index), qdma->irqmask[index]);
+	irq_bank->irqmask[index] &= ~clear;
+	irq_bank->irqmask[index] |= set;
+	airoha_qdma_wr(qdma, REG_INT_ENABLE(bank, index),
+		       irq_bank->irqmask[index]);
 	/* Read irq_enable register in order to guarantee the update above
 	 * completes in the spinlock critical section.
 	 */
-	airoha_qdma_rr(qdma, REG_INT_ENABLE(index));
+	airoha_qdma_rr(qdma, REG_INT_ENABLE(bank, index));
 
-	spin_unlock_irqrestore(&qdma->irq_lock, flags);
+	spin_unlock_irqrestore(&irq_bank->irq_lock, flags);
 }
 
-static void airoha_qdma_irq_enable(struct airoha_qdma *qdma, int index,
-				   u32 mask)
+static void airoha_qdma_irq_enable(struct airoha_irq_bank *irq_bank,
+				   int index, u32 mask)
 {
-	airoha_qdma_set_irqmask(qdma, index, 0, mask);
+	airoha_qdma_set_irqmask(irq_bank, index, 0, mask);
 }
 
-static void airoha_qdma_irq_disable(struct airoha_qdma *qdma, int index,
-				    u32 mask)
+static void airoha_qdma_irq_disable(struct airoha_irq_bank *irq_bank,
+				    int index, u32 mask)
 {
-	airoha_qdma_set_irqmask(qdma, index, mask, 0);
-}
-
-static bool airhoa_is_lan_gdm_port(struct airoha_gdm_port *port)
-{
-	/* GDM1 port on EN7581 SoC is connected to the lan dsa switch.
-	 * GDM{2,3,4} can be used as wan port connected to an external
-	 * phy module.
-	 */
-	return port->id == 1;
+	airoha_qdma_set_irqmask(irq_bank, index, mask, 0);
 }
 
 static void airoha_set_macaddr(struct airoha_gdm_port *port, const u8 *addr)
@@ -527,6 +522,25 @@ static int airoha_fe_init(struct airoha_eth *eth)
 	/* disable IFC by default */
 	airoha_fe_clear(eth, REG_FE_CSR_IFC_CFG, FE_IFC_EN_MASK);
 
+	airoha_fe_wr(eth, REG_PPE_DFT_CPORT0(0),
+		     FIELD_PREP(DFT_CPORT_MASK(7), FE_PSE_PORT_CDM1) |
+		     FIELD_PREP(DFT_CPORT_MASK(6), FE_PSE_PORT_CDM1) |
+		     FIELD_PREP(DFT_CPORT_MASK(5), FE_PSE_PORT_CDM1) |
+		     FIELD_PREP(DFT_CPORT_MASK(4), FE_PSE_PORT_CDM1) |
+		     FIELD_PREP(DFT_CPORT_MASK(3), FE_PSE_PORT_CDM1) |
+		     FIELD_PREP(DFT_CPORT_MASK(2), FE_PSE_PORT_CDM1) |
+		     FIELD_PREP(DFT_CPORT_MASK(1), FE_PSE_PORT_CDM1) |
+		     FIELD_PREP(DFT_CPORT_MASK(0), FE_PSE_PORT_CDM1));
+	airoha_fe_wr(eth, REG_PPE_DFT_CPORT0(1),
+		     FIELD_PREP(DFT_CPORT_MASK(7), FE_PSE_PORT_CDM2) |
+		     FIELD_PREP(DFT_CPORT_MASK(6), FE_PSE_PORT_CDM2) |
+		     FIELD_PREP(DFT_CPORT_MASK(5), FE_PSE_PORT_CDM2) |
+		     FIELD_PREP(DFT_CPORT_MASK(4), FE_PSE_PORT_CDM2) |
+		     FIELD_PREP(DFT_CPORT_MASK(3), FE_PSE_PORT_CDM2) |
+		     FIELD_PREP(DFT_CPORT_MASK(2), FE_PSE_PORT_CDM2) |
+		     FIELD_PREP(DFT_CPORT_MASK(1), FE_PSE_PORT_CDM2) |
+		     FIELD_PREP(DFT_CPORT_MASK(0), FE_PSE_PORT_CDM2));
+
 	/* enable 1:N vlan action, init vlan table */
 	airoha_fe_set(eth, REG_MC_VLAN_EN, MC_VLAN_EN_MASK);
 
@@ -687,7 +701,7 @@ static int airoha_qdma_rx_process(struct airoha_queue *q, int budget)
 
 		reason = FIELD_GET(AIROHA_RXD4_PPE_CPU_REASON, msg1);
 		if (reason == PPE_CPU_REASON_HIT_UNBIND_RATE_REACHED)
-			airoha_ppe_check_skb(eth->ppe, hash);
+			airoha_ppe_check_skb(eth->ppe, q->skb, hash);
 
 		done++;
 		napi_gro_receive(&q->napi, q->skb);
@@ -716,9 +730,20 @@ static int airoha_qdma_rx_napi_poll(struct napi_struct *napi, int budget)
 		done += cur;
 	} while (cur && done < budget);
 
-	if (done < budget && napi_complete(napi))
-		airoha_qdma_irq_enable(q->qdma, QDMA_INT_REG_IDX1,
-				       RX_DONE_INT_MASK);
+	if (done < budget && napi_complete(napi)) {
+		struct airoha_qdma *qdma = q->qdma;
+		int i, qid = q - &qdma->q_rx[0];
+		int intr_reg = qid < RX_DONE_HIGH_OFFSET ? QDMA_INT_REG_IDX1
+							 : QDMA_INT_REG_IDX2;
+
+		for (i = 0; i < ARRAY_SIZE(qdma->irq_banks); i++) {
+			if (!(BIT(qid) & RX_IRQ_BANK_PIN_MASK(i)))
+				continue;
+
+			airoha_qdma_irq_enable(&qdma->irq_banks[i], intr_reg,
+					       BIT(qid % RX_DONE_HIGH_OFFSET));
+		}
+	}
 
 	return done;
 }
@@ -921,7 +946,7 @@ unlock:
 	}
 
 	if (done < budget && napi_complete(napi))
-		airoha_qdma_irq_enable(qdma, QDMA_INT_REG_IDX0,
+		airoha_qdma_irq_enable(&qdma->irq_banks[0], QDMA_INT_REG_IDX0,
 				       TX_DONE_INT_MASK(id));
 
 	return done;
@@ -1039,24 +1064,46 @@ static void airoha_qdma_cleanup_tx_queue(struct airoha_queue *q)
 static int airoha_qdma_init_hfwd_queues(struct airoha_qdma *qdma)
 {
 	struct airoha_eth *eth = qdma->eth;
+	int id = qdma - &eth->qdma[0];
 	dma_addr_t dma_addr;
+	const char *name;
+	int size, index;
 	u32 status;
-	int size;
 
 	size = HW_DSCP_NUM * sizeof(struct airoha_qdma_fwd_desc);
-	qdma->hfwd.desc = dmam_alloc_coherent(eth->dev, size, &dma_addr,
-					      GFP_KERNEL);
-	if (!qdma->hfwd.desc)
+	if (!dmam_alloc_coherent(eth->dev, size, &dma_addr, GFP_KERNEL))
 		return -ENOMEM;
 
 	airoha_qdma_wr(qdma, REG_FWD_DSCP_BASE, dma_addr);
 
-	size = AIROHA_MAX_PACKET_SIZE * HW_DSCP_NUM;
-	qdma->hfwd.q = dmam_alloc_coherent(eth->dev, size, &dma_addr,
-					   GFP_KERNEL);
-	if (!qdma->hfwd.q)
+	name = devm_kasprintf(eth->dev, GFP_KERNEL, "qdma%d-buf", id);
+	if (!name)
 		return -ENOMEM;
 
+	index = of_property_match_string(eth->dev->of_node,
+					 "memory-region-names", name);
+	if (index >= 0) {
+		struct reserved_mem *rmem;
+		struct device_node *np;
+
+		/* Consume reserved memory for hw forwarding buffers queue if
+		 * available in the DTS
+		 */
+		np = of_parse_phandle(eth->dev->of_node, "memory-region",
+				      index);
+		if (!np)
+			return -ENODEV;
+
+		rmem = of_reserved_mem_lookup(np);
+		of_node_put(np);
+		dma_addr = rmem->base;
+	} else {
+		size = AIROHA_MAX_PACKET_SIZE * HW_DSCP_NUM;
+		if (!dmam_alloc_coherent(eth->dev, size, &dma_addr,
+					 GFP_KERNEL))
+			return -ENOMEM;
+	}
+
 	airoha_qdma_wr(qdma, REG_FWD_BUF_BASE, dma_addr);
 
 	airoha_qdma_rmw(qdma, REG_HW_FWD_DSCP_CFG,
@@ -1068,7 +1115,7 @@ static int airoha_qdma_init_hfwd_queues(struct airoha_qdma *qdma)
 			LMGR_INIT_START | LMGR_SRAM_MODE_MASK |
 			HW_FWD_DESC_NUM_MASK,
 			FIELD_PREP(HW_FWD_DESC_NUM_MASK, HW_DSCP_NUM) |
-			LMGR_INIT_START);
+			LMGR_INIT_START | LMGR_SRAM_MODE_MASK);
 
 	return read_poll_timeout(airoha_qdma_rr, status,
 				 !(status & LMGR_INIT_START), USEC_PER_MSEC,
@@ -1151,14 +1198,24 @@ static int airoha_qdma_hw_init(struct airoha_qdma *qdma)
 {
 	int i;
 
-	/* clear pending irqs */
-	for (i = 0; i < ARRAY_SIZE(qdma->irqmask); i++)
+	for (i = 0; i < ARRAY_SIZE(qdma->irq_banks); i++) {
+		/* clear pending irqs */
 		airoha_qdma_wr(qdma, REG_INT_STATUS(i), 0xffffffff);
-
-	/* setup irqs */
-	airoha_qdma_irq_enable(qdma, QDMA_INT_REG_IDX0, INT_IDX0_MASK);
-	airoha_qdma_irq_enable(qdma, QDMA_INT_REG_IDX1, INT_IDX1_MASK);
-	airoha_qdma_irq_enable(qdma, QDMA_INT_REG_IDX4, INT_IDX4_MASK);
+		/* setup rx irqs */
+		airoha_qdma_irq_enable(&qdma->irq_banks[i], QDMA_INT_REG_IDX0,
+				       INT_RX0_MASK(RX_IRQ_BANK_PIN_MASK(i)));
+		airoha_qdma_irq_enable(&qdma->irq_banks[i], QDMA_INT_REG_IDX1,
+				       INT_RX1_MASK(RX_IRQ_BANK_PIN_MASK(i)));
+		airoha_qdma_irq_enable(&qdma->irq_banks[i], QDMA_INT_REG_IDX2,
+				       INT_RX2_MASK(RX_IRQ_BANK_PIN_MASK(i)));
+		airoha_qdma_irq_enable(&qdma->irq_banks[i], QDMA_INT_REG_IDX3,
+				       INT_RX3_MASK(RX_IRQ_BANK_PIN_MASK(i)));
+	}
+	/* setup tx irqs */
+	airoha_qdma_irq_enable(&qdma->irq_banks[0], QDMA_INT_REG_IDX0,
+			       TX_COHERENT_LOW_INT_MASK | INT_TX_MASK);
+	airoha_qdma_irq_enable(&qdma->irq_banks[0], QDMA_INT_REG_IDX4,
+			       TX_COHERENT_HIGH_INT_MASK);
 
 	/* setup irq binding */
 	for (i = 0; i < ARRAY_SIZE(qdma->q_tx); i++) {
@@ -1203,30 +1260,39 @@ static int airoha_qdma_hw_init(struct airoha_qdma *qdma)
 
 static irqreturn_t airoha_irq_handler(int irq, void *dev_instance)
 {
-	struct airoha_qdma *qdma = dev_instance;
-	u32 intr[ARRAY_SIZE(qdma->irqmask)];
+	struct airoha_irq_bank *irq_bank = dev_instance;
+	struct airoha_qdma *qdma = irq_bank->qdma;
+	u32 rx_intr_mask = 0, rx_intr1, rx_intr2;
+	u32 intr[ARRAY_SIZE(irq_bank->irqmask)];
 	int i;
 
-	for (i = 0; i < ARRAY_SIZE(qdma->irqmask); i++) {
+	for (i = 0; i < ARRAY_SIZE(intr); i++) {
 		intr[i] = airoha_qdma_rr(qdma, REG_INT_STATUS(i));
-		intr[i] &= qdma->irqmask[i];
+		intr[i] &= irq_bank->irqmask[i];
 		airoha_qdma_wr(qdma, REG_INT_STATUS(i), intr[i]);
 	}
 
 	if (!test_bit(DEV_STATE_INITIALIZED, &qdma->eth->state))
 		return IRQ_NONE;
 
-	if (intr[1] & RX_DONE_INT_MASK) {
-		airoha_qdma_irq_disable(qdma, QDMA_INT_REG_IDX1,
-					RX_DONE_INT_MASK);
+	rx_intr1 = intr[1] & RX_DONE_LOW_INT_MASK;
+	if (rx_intr1) {
+		airoha_qdma_irq_disable(irq_bank, QDMA_INT_REG_IDX1, rx_intr1);
+		rx_intr_mask |= rx_intr1;
+	}
+
+	rx_intr2 = intr[2] & RX_DONE_HIGH_INT_MASK;
+	if (rx_intr2) {
+		airoha_qdma_irq_disable(irq_bank, QDMA_INT_REG_IDX2, rx_intr2);
+		rx_intr_mask |= (rx_intr2 << 16);
+	}
 
-		for (i = 0; i < ARRAY_SIZE(qdma->q_rx); i++) {
-			if (!qdma->q_rx[i].ndesc)
-				continue;
+	for (i = 0; rx_intr_mask && i < ARRAY_SIZE(qdma->q_rx); i++) {
+		if (!qdma->q_rx[i].ndesc)
+			continue;
 
-			if (intr[1] & BIT(i))
-				napi_schedule(&qdma->q_rx[i].napi);
-		}
+		if (rx_intr_mask & BIT(i))
+			napi_schedule(&qdma->q_rx[i].napi);
 	}
 
 	if (intr[0] & INT_TX_MASK) {
@@ -1234,7 +1300,7 @@ static irqreturn_t airoha_irq_handler(int irq, void *dev_instance)
 			if (!(intr[0] & TX_DONE_INT_MASK(i)))
 				continue;
 
-			airoha_qdma_irq_disable(qdma, QDMA_INT_REG_IDX0,
+			airoha_qdma_irq_disable(irq_bank, QDMA_INT_REG_IDX0,
 						TX_DONE_INT_MASK(i));
 			napi_schedule(&qdma->q_tx_irq[i].napi);
 		}
@@ -1243,6 +1309,39 @@ static irqreturn_t airoha_irq_handler(int irq, void *dev_instance)
 	return IRQ_HANDLED;
 }
 
+static int airoha_qdma_init_irq_banks(struct platform_device *pdev,
+				      struct airoha_qdma *qdma)
+{
+	struct airoha_eth *eth = qdma->eth;
+	int i, id = qdma - &eth->qdma[0];
+
+	for (i = 0; i < ARRAY_SIZE(qdma->irq_banks); i++) {
+		struct airoha_irq_bank *irq_bank = &qdma->irq_banks[i];
+		int err, irq_index = 4 * id + i;
+		const char *name;
+
+		spin_lock_init(&irq_bank->irq_lock);
+		irq_bank->qdma = qdma;
+
+		irq_bank->irq = platform_get_irq(pdev, irq_index);
+		if (irq_bank->irq < 0)
+			return irq_bank->irq;
+
+		name = devm_kasprintf(eth->dev, GFP_KERNEL,
+				      KBUILD_MODNAME ".%d", irq_index);
+		if (!name)
+			return -ENOMEM;
+
+		err = devm_request_irq(eth->dev, irq_bank->irq,
+				       airoha_irq_handler, IRQF_SHARED, name,
+				       irq_bank);
+		if (err)
+			return err;
+	}
+
+	return 0;
+}
+
 static int airoha_qdma_init(struct platform_device *pdev,
 			    struct airoha_eth *eth,
 			    struct airoha_qdma *qdma)
@@ -1250,9 +1349,7 @@ static int airoha_qdma_init(struct platform_device *pdev,
 	int err, id = qdma - &eth->qdma[0];
 	const char *res;
 
-	spin_lock_init(&qdma->irq_lock);
 	qdma->eth = eth;
-
 	res = devm_kasprintf(eth->dev, GFP_KERNEL, "qdma%d", id);
 	if (!res)
 		return -ENOMEM;
@@ -1262,12 +1359,7 @@ static int airoha_qdma_init(struct platform_device *pdev,
 		return dev_err_probe(eth->dev, PTR_ERR(qdma->regs),
 				     "failed to iomap qdma%d regs\n", id);
 
-	qdma->irq = platform_get_irq(pdev, 4 * id);
-	if (qdma->irq < 0)
-		return qdma->irq;
-
-	err = devm_request_irq(eth->dev, qdma->irq, airoha_irq_handler,
-			       IRQF_SHARED, KBUILD_MODNAME, qdma);
+	err = airoha_qdma_init_irq_banks(pdev, qdma);
 	if (err)
 		return err;
 
@@ -1627,7 +1719,6 @@ static void airhoha_set_gdm2_loopback(struct airoha_gdm_port *port)
 
 	if (port->id == 3) {
 		/* FIXME: handle XSI_PCE1_PORT */
-		airoha_fe_wr(eth, REG_PPE_DFT_CPORT0(0),  0x5500);
 		airoha_fe_rmw(eth, REG_FE_WAN_PORT,
 			      WAN1_EN_MASK | WAN1_MASK | WAN0_MASK,
 			      FIELD_PREP(WAN0_MASK, HSGMII_LAN_PCIE0_SRCPORT));
@@ -2105,6 +2196,125 @@ static int airoha_tc_setup_qdisc_ets(struct airoha_gdm_port *port,
 	}
 }
 
+static int airoha_qdma_get_rl_param(struct airoha_qdma *qdma, int queue_id,
+				    u32 addr, enum trtcm_param_type param,
+				    u32 *val_low, u32 *val_high)
+{
+	u32 idx = QDMA_METER_IDX(queue_id), group = QDMA_METER_GROUP(queue_id);
+	u32 val, config = FIELD_PREP(RATE_LIMIT_PARAM_TYPE_MASK, param) |
+			  FIELD_PREP(RATE_LIMIT_METER_GROUP_MASK, group) |
+			  FIELD_PREP(RATE_LIMIT_PARAM_INDEX_MASK, idx);
+
+	airoha_qdma_wr(qdma, REG_TRTCM_CFG_PARAM(addr), config);
+	if (read_poll_timeout(airoha_qdma_rr, val,
+			      val & RATE_LIMIT_PARAM_RW_DONE_MASK,
+			      USEC_PER_MSEC, 10 * USEC_PER_MSEC, true, qdma,
+			      REG_TRTCM_CFG_PARAM(addr)))
+		return -ETIMEDOUT;
+
+	*val_low = airoha_qdma_rr(qdma, REG_TRTCM_DATA_LOW(addr));
+	if (val_high)
+		*val_high = airoha_qdma_rr(qdma, REG_TRTCM_DATA_HIGH(addr));
+
+	return 0;
+}
+
+static int airoha_qdma_set_rl_param(struct airoha_qdma *qdma, int queue_id,
+				    u32 addr, enum trtcm_param_type param,
+				    u32 val)
+{
+	u32 idx = QDMA_METER_IDX(queue_id), group = QDMA_METER_GROUP(queue_id);
+	u32 config = RATE_LIMIT_PARAM_RW_MASK |
+		     FIELD_PREP(RATE_LIMIT_PARAM_TYPE_MASK, param) |
+		     FIELD_PREP(RATE_LIMIT_METER_GROUP_MASK, group) |
+		     FIELD_PREP(RATE_LIMIT_PARAM_INDEX_MASK, idx);
+
+	airoha_qdma_wr(qdma, REG_TRTCM_DATA_LOW(addr), val);
+	airoha_qdma_wr(qdma, REG_TRTCM_CFG_PARAM(addr), config);
+
+	return read_poll_timeout(airoha_qdma_rr, val,
+				 val & RATE_LIMIT_PARAM_RW_DONE_MASK,
+				 USEC_PER_MSEC, 10 * USEC_PER_MSEC, true,
+				 qdma, REG_TRTCM_CFG_PARAM(addr));
+}
+
+static int airoha_qdma_set_rl_config(struct airoha_qdma *qdma, int queue_id,
+				     u32 addr, bool enable, u32 enable_mask)
+{
+	u32 val;
+	int err;
+
+	err = airoha_qdma_get_rl_param(qdma, queue_id, addr, TRTCM_MISC_MODE,
+				       &val, NULL);
+	if (err)
+		return err;
+
+	val = enable ? val | enable_mask : val & ~enable_mask;
+
+	return airoha_qdma_set_rl_param(qdma, queue_id, addr, TRTCM_MISC_MODE,
+					val);
+}
+
+static int airoha_qdma_set_rl_token_bucket(struct airoha_qdma *qdma,
+					   int queue_id, u32 rate_val,
+					   u32 bucket_size)
+{
+	u32 val, config, tick, unit, rate, rate_frac;
+	int err;
+
+	err = airoha_qdma_get_rl_param(qdma, queue_id, REG_INGRESS_TRTCM_CFG,
+				       TRTCM_MISC_MODE, &config, NULL);
+	if (err)
+		return err;
+
+	val = airoha_qdma_rr(qdma, REG_INGRESS_TRTCM_CFG);
+	tick = FIELD_GET(INGRESS_FAST_TICK_MASK, val);
+	if (config & TRTCM_TICK_SEL)
+		tick *= FIELD_GET(INGRESS_SLOW_TICK_RATIO_MASK, val);
+	if (!tick)
+		return -EINVAL;
+
+	unit = (config & TRTCM_PKT_MODE) ? 1000000 / tick : 8000 / tick;
+	if (!unit)
+		return -EINVAL;
+
+	rate = rate_val / unit;
+	rate_frac = rate_val % unit;
+	rate_frac = FIELD_PREP(TRTCM_TOKEN_RATE_MASK, rate_frac) / unit;
+	rate = FIELD_PREP(TRTCM_TOKEN_RATE_MASK, rate) |
+	       FIELD_PREP(TRTCM_TOKEN_RATE_FRACTION_MASK, rate_frac);
+
+	err = airoha_qdma_set_rl_param(qdma, queue_id, REG_INGRESS_TRTCM_CFG,
+				       TRTCM_TOKEN_RATE_MODE, rate);
+	if (err)
+		return err;
+
+	val = bucket_size;
+	if (!(config & TRTCM_PKT_MODE))
+		val = max_t(u32, val, MIN_TOKEN_SIZE);
+	val = min_t(u32, __fls(val), MAX_TOKEN_SIZE_OFFSET);
+
+	return airoha_qdma_set_rl_param(qdma, queue_id, REG_INGRESS_TRTCM_CFG,
+					TRTCM_BUCKETSIZE_SHIFT_MODE, val);
+}
+
+static int airoha_qdma_init_rl_config(struct airoha_qdma *qdma, int queue_id,
+				      bool enable, enum trtcm_unit_type unit)
+{
+	bool tick_sel = queue_id == 0 || queue_id == 2 || queue_id == 8;
+	enum trtcm_param mode = TRTCM_METER_MODE;
+	int err;
+
+	mode |= unit == TRTCM_PACKET_UNIT ? TRTCM_PKT_MODE : 0;
+	err = airoha_qdma_set_rl_config(qdma, queue_id, REG_INGRESS_TRTCM_CFG,
+					enable, mode);
+	if (err)
+		return err;
+
+	return airoha_qdma_set_rl_config(qdma, queue_id, REG_INGRESS_TRTCM_CFG,
+					 tick_sel, TRTCM_TICK_SEL);
+}
+
 static int airoha_qdma_get_trtcm_param(struct airoha_qdma *qdma, int channel,
 				       u32 addr, enum trtcm_param_type param,
 				       enum trtcm_mode_type mode,
@@ -2269,10 +2479,142 @@ static int airoha_tc_htb_alloc_leaf_queue(struct airoha_gdm_port *port,
 	return 0;
 }
 
+static int airoha_qdma_set_rx_meter(struct airoha_gdm_port *port,
+				    u32 rate, u32 bucket_size,
+				    enum trtcm_unit_type unit_type)
+{
+	struct airoha_qdma *qdma = port->qdma;
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(qdma->q_rx); i++) {
+		int err;
+
+		if (!qdma->q_rx[i].ndesc)
+			continue;
+
+		err = airoha_qdma_init_rl_config(qdma, i, !!rate, unit_type);
+		if (err)
+			return err;
+
+		err = airoha_qdma_set_rl_token_bucket(qdma, i, rate,
+						      bucket_size);
+		if (err)
+			return err;
+	}
+
+	return 0;
+}
+
+static int airoha_tc_matchall_act_validate(struct tc_cls_matchall_offload *f)
+{
+	const struct flow_action *actions = &f->rule->action;
+	const struct flow_action_entry *act;
+
+	if (!flow_action_has_entries(actions)) {
+		NL_SET_ERR_MSG_MOD(f->common.extack,
+				   "filter run with no actions");
+		return -EINVAL;
+	}
+
+	if (!flow_offload_has_one_action(actions)) {
+		NL_SET_ERR_MSG_MOD(f->common.extack,
+				   "only once action per filter is supported");
+		return -EOPNOTSUPP;
+	}
+
+	act = &actions->entries[0];
+	if (act->id != FLOW_ACTION_POLICE) {
+		NL_SET_ERR_MSG_MOD(f->common.extack, "unsupported action");
+		return -EOPNOTSUPP;
+	}
+
+	if (act->police.exceed.act_id != FLOW_ACTION_DROP) {
+		NL_SET_ERR_MSG_MOD(f->common.extack,
+				   "invalid exceed action id");
+		return -EOPNOTSUPP;
+	}
+
+	if (act->police.notexceed.act_id != FLOW_ACTION_ACCEPT) {
+		NL_SET_ERR_MSG_MOD(f->common.extack,
+				   "invalid notexceed action id");
+		return -EOPNOTSUPP;
+	}
+
+	if (act->police.notexceed.act_id == FLOW_ACTION_ACCEPT &&
+	    !flow_action_is_last_entry(actions, act)) {
+		NL_SET_ERR_MSG_MOD(f->common.extack,
+				   "action accept must be last");
+		return -EOPNOTSUPP;
+	}
+
+	if (act->police.peakrate_bytes_ps || act->police.avrate ||
+	    act->police.overhead || act->police.mtu) {
+		NL_SET_ERR_MSG_MOD(f->common.extack,
+				   "peakrate/avrate/overhead/mtu unsupported");
+		return -EOPNOTSUPP;
+	}
+
+	return 0;
+}
+
+static int airoha_dev_tc_matchall(struct net_device *dev,
+				  struct tc_cls_matchall_offload *f)
+{
+	enum trtcm_unit_type unit_type = TRTCM_BYTE_UNIT;
+	struct airoha_gdm_port *port = netdev_priv(dev);
+	u32 rate = 0, bucket_size = 0;
+
+	switch (f->command) {
+	case TC_CLSMATCHALL_REPLACE: {
+		const struct flow_action_entry *act;
+		int err;
+
+		err = airoha_tc_matchall_act_validate(f);
+		if (err)
+			return err;
+
+		act = &f->rule->action.entries[0];
+		if (act->police.rate_pkt_ps) {
+			rate = act->police.rate_pkt_ps;
+			bucket_size = act->police.burst_pkt;
+			unit_type = TRTCM_PACKET_UNIT;
+		} else {
+			rate = div_u64(act->police.rate_bytes_ps, 1000);
+			rate = rate << 3; /* Kbps */
+			bucket_size = act->police.burst;
+		}
+		fallthrough;
+	}
+	case TC_CLSMATCHALL_DESTROY:
+		return airoha_qdma_set_rx_meter(port, rate, bucket_size,
+						unit_type);
+	default:
+		return -EOPNOTSUPP;
+	}
+}
+
+static int airoha_dev_setup_tc_block_cb(enum tc_setup_type type,
+					void *type_data, void *cb_priv)
+{
+	struct net_device *dev = cb_priv;
+
+	if (!tc_can_offload(dev))
+		return -EOPNOTSUPP;
+
+	switch (type) {
+	case TC_SETUP_CLSFLOWER:
+		return airoha_ppe_setup_tc_block_cb(dev, type_data);
+	case TC_SETUP_CLSMATCHALL:
+		return airoha_dev_tc_matchall(dev, type_data);
+	default:
+		return -EOPNOTSUPP;
+	}
+}
+
 static int airoha_dev_setup_tc_block(struct airoha_gdm_port *port,
 				     struct flow_block_offload *f)
 {
-	flow_setup_cb_t *cb = airoha_ppe_setup_tc_block_cb;
+	flow_setup_cb_t *cb = airoha_dev_setup_tc_block_cb;
 	static LIST_HEAD(block_cb_list);
 	struct flow_block_cb *block_cb;
 
@@ -2511,7 +2853,7 @@ static int airoha_alloc_gdm_port(struct airoha_eth *eth,
 	dev->features |= dev->hw_features;
 	dev->vlan_features = dev->hw_features;
 	dev->dev.of_node = np;
-	dev->irq = qdma->irq;
+	dev->irq = qdma->irq_banks[0].irq;
 	SET_NETDEV_DEV(dev, eth->dev);
 
 	/* reserve hw queues for HTB offloading */
@@ -2541,7 +2883,15 @@ static int airoha_alloc_gdm_port(struct airoha_eth *eth,
 	if (err)
 		return err;
 
-	return register_netdev(dev);
+	err = register_netdev(dev);
+	if (err)
+		goto free_metadata_dst;
+
+	return 0;
+
+free_metadata_dst:
+	airoha_metadata_dst_free(port);
+	return err;
 }
 
 static int airoha_probe(struct platform_device *pdev)