Merge tag 'net-next-6.3' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-nextgrafted

Pull networking updates from Jakub Kicinski:
 "Core:

   - Add dedicated kmem_cache for typical/small skb->head, avoid having
     to access struct page at kfree time, and improve memory use.

   - Introduce sysctl to set default RPS configuration for new netdevs.

   - Define Netlink protocol specification format which can be used to
     describe messages used by each family and auto-generate parsers.
     Add tools for generating kernel data structures and uAPI headers.

   - Expose all net/core sysctls inside netns.

   - Remove 4s sleep in netpoll if carrier is instantly detected on
     boot.

   - Add configurable limit of MDB entries per port, and port-vlan.

   - Continue populating drop reasons throughout the stack.

   - Retire a handful of legacy Qdiscs and classifiers.

  Protocols:

   - Support IPv4 big TCP (TSO frames larger than 64kB).

   - Add IP_LOCAL_PORT_RANGE socket option, to control local port range
     on socket by socket basis.

   - Track and report in procfs number of MPTCP sockets used.

   - Support mixing IPv4 and IPv6 flows in the in-kernel MPTCP path
     manager.

   - IPv6: don't check net.ipv6.route.max_size and rely on garbage
     collection to free memory (similarly to IPv4).

   - Support Penultimate Segment Pop (PSP) flavor in SRv6 (RFC8986).

   - ICMP: add per-rate limit counters.

   - Add support for user scanning requests in ieee802154.

   - Remove static WEP support.

   - Support minimal Wi-Fi 7 Extremely High Throughput (EHT) rate
     reporting.

   - WiFi 7 EHT channel puncturing support (client & AP).

  BPF:

   - Add a rbtree data structure following the "next-gen data structure"
     precedent set by recently added linked list, that is, by using
     kfunc + kptr instead of adding a new BPF map type.

   - Expose XDP hints via kfuncs with initial support for RX hash and
     timestamp metadata.

   - Add BPF_F_NO_TUNNEL_KEY extension to bpf_skb_set_tunnel_key to
     better support decap on GRE tunnel devices not operating in collect
     metadata.

   - Improve x86 JIT's codegen for PROBE_MEM runtime error checks.

   - Remove the need for trace_printk_lock for bpf_trace_printk and
     bpf_trace_vprintk helpers.

   - Extend libbpf's bpf_tracing.h support for tracing arguments of
     kprobes/uprobes and syscall as a special case.

   - Significantly reduce the search time for module symbols by
     livepatch and BPF.

   - Enable cpumasks to be used as kptrs, which is useful for tracing
     programs tracking which tasks end up running on which CPUs in
     different time intervals.

   - Add support for BPF trampoline on s390x and riscv64.

   - Add capability to export the XDP features supported by the NIC.

   - Add __bpf_kfunc tag for marking kernel functions as kfuncs.

   - Add cgroup.memory=nobpf kernel parameter option to disable BPF
     memory accounting for container environments.

  Netfilter:

   - Remove the CLUSTERIP target. It has been marked as obsolete for
     years, and we still have WARN splats wrt races of the out-of-band
     /proc interface installed by this target.

   - Add 'destroy' commands to nf_tables. They are identical to the
     existing 'delete' commands, but do not return an error if the
     referenced object (set, chain, rule...) did not exist.

  Driver API:

   - Improve cpumask_local_spread() locality to help NICs set the right
     IRQ affinity on AMD platforms.

   - Separate C22 and C45 MDIO bus transactions more clearly.

   - Introduce new DCB table to control DSCP rewrite on egress.

   - Support configuration of Physical Layer Collision Avoidance (PLCA)
     Reconciliation Sublayer (RS) (802.3cg-2019). Modern version of
     shared medium Ethernet.

   - Support for MAC Merge layer (IEEE 802.3-2018 clause 99). Allowing
     preemption of low priority frames by high priority frames.

   - Add support for controlling MACSec offload using netlink SET.

   - Rework devlink instance refcounts to allow registration and
     de-registration under the instance lock. Split the code into
     multiple files, drop some of the unnecessarily granular locks and
     factor out common parts of netlink operation handling.

   - Add TX frame aggregation parameters (for USB drivers).

   - Add a new attr TCA_EXT_WARN_MSG to report TC (offload) warning
     messages with notifications for debug.

   - Allow offloading of UDP NEW connections via act_ct.

   - Add support for per action HW stats in TC.

   - Support hardware miss to TC action (continue processing in SW from
     a specific point in the action chain).

   - Warn if old Wireless Extension user space interface is used with
     modern cfg80211/mac80211 drivers. Do not support Wireless
     Extensions for Wi-Fi 7 devices at all. Everyone should switch to
     using nl80211 interface instead.

   - Improve the CAN bit timing configuration. Use extack to return
     error messages directly to user space, update the SJW handling,
     including the definition of a new default value that will benefit
     CAN-FD controllers, by increasing their oscillator tolerance.

  New hardware / drivers:

   - Ethernet:
      - nVidia BlueField-3 support (control traffic driver)
      - Ethernet support for imx93 SoCs
      - Motorcomm yt8531 gigabit Ethernet PHY
      - onsemi NCN26000 10BASE-T1S PHY (with support for PLCA)
      - Microchip LAN8841 PHY (incl. cable diagnostics and PTP)
      - Amlogic gxl MDIO mux

   - WiFi:
      - RealTek RTL8188EU (rtl8xxxu)
      - Qualcomm Wi-Fi 7 devices (ath12k)

   - CAN:
      - Renesas R-Car V4H

  Drivers:

   - Bluetooth:
      - Set Per Platform Antenna Gain (PPAG) for Intel controllers.

   - Ethernet NICs:
      - Intel (1G, igc):
         - support TSN / Qbv / packet scheduling features of i226 model
      - Intel (100G, ice):
         - use GNSS subsystem instead of TTY
         - multi-buffer XDP support
         - extend support for GPIO pins to E823 devices
      - nVidia/Mellanox:
         - update the shared buffer configuration on PFC commands
         - implement PTP adjphase function for HW offset control
         - TC support for Geneve and GRE with VF tunnel offload
         - more efficient crypto key management method
         - multi-port eswitch support
      - Netronome/Corigine:
         - add DCB IEEE support
         - support IPsec offloading for NFP3800
      - Freescale/NXP (enetc):
         - support XDP_REDIRECT for XDP non-linear buffers
         - improve reconfig, avoid link flap and waiting for idle
         - support MAC Merge layer
      - Other NICs:
         - sfc/ef100: add basic devlink support for ef100
         - ionic: rx_push mode operation (writing descriptors via MMIO)
         - bnxt: use the auxiliary bus abstraction for RDMA
         - r8169: disable ASPM and reset bus in case of tx timeout
         - cpsw: support QSGMII mode for J721e CPSW9G
         - cpts: support pulse-per-second output
         - ngbe: add an mdio bus driver
         - usbnet: optimize usbnet_bh() by avoiding unnecessary queuing
         - r8152: handle devices with FW with NCM support
         - amd-xgbe: support 10Mbps, 2.5GbE speeds and rx-adaptation
         - virtio-net: support multi buffer XDP
         - virtio/vsock: replace virtio_vsock_pkt with sk_buff
         - tsnep: XDP support

   - Ethernet high-speed switches:
      - nVidia/Mellanox (mlxsw):
         - add support for latency TLV (in FW control messages)
      - Microchip (sparx5):
         - separate explicit and implicit traffic forwarding rules, make
           the implicit rules always active
         - add support for egress DSCP rewrite
         - IS0 VCAP support (Ingress Classification)
         - IS2 VCAP filters (protos, L3 addrs, L4 ports, flags, ToS
           etc.)
         - ES2 VCAP support (Egress Access Control)
         - support for Per-Stream Filtering and Policing (802.1Q,
           8.6.5.1)

   - Ethernet embedded switches:
      - Marvell (mv88e6xxx):
         - add MAB (port auth) offload support
         - enable PTP receive for mv88e6390
      - NXP (ocelot):
         - support MAC Merge layer
         - support for the the vsc7512 internal copper phys
      - Microchip:
         - lan9303: convert to PHYLINK
         - lan966x: support TC flower filter statistics
         - lan937x: PTP support for KSZ9563/KSZ8563 and LAN937x
         - lan937x: support Credit Based Shaper configuration
         - ksz9477: support Energy Efficient Ethernet
      - other:
         - qca8k: convert to regmap read/write API, use bulk operations
         - rswitch: Improve TX timestamp accuracy

   - Intel WiFi (iwlwifi):
      - EHT (Wi-Fi 7) rate reporting
      - STEP equalizer support: transfer some STEP (connection to radio
        on platforms with integrated wifi) related parameters from the
        BIOS to the firmware.

   - Qualcomm 802.11ax WiFi (ath11k):
      - IPQ5018 support
      - Fine Timing Measurement (FTM) responder role support
      - channel 177 support

   - MediaTek WiFi (mt76):
      - per-PHY LED support
      - mt7996: EHT (Wi-Fi 7) support
      - Wireless Ethernet Dispatch (WED) reset support
      - switch to using page pool allocator

   - RealTek WiFi (rtw89):
      - support new version of Bluetooth co-existance

   - Mobile:
      - rmnet: support TX aggregation"

* tag 'net-next-6.3' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next: (1872 commits)
  page_pool: add a comment explaining the fragment counter usage
  net: ethtool: fix __ethtool_dev_mm_supported() implementation
  ethtool: pse-pd: Fix double word in comments
  xsk: add linux/vmalloc.h to xsk.c
  sefltests: netdevsim: wait for devlink instance after netns removal
  selftest: fib_tests: Always cleanup before exit
  net/mlx5e: Align IPsec ASO result memory to be as required by hardware
  net/mlx5e: TC, Set CT miss to the specific ct action instance
  net/mlx5e: Rename CHAIN_TO_REG to MAPPED_OBJ_TO_REG
  net/mlx5: Refactor tc miss handling to a single function
  net/mlx5: Kconfig: Make tc offload depend on tc skb extension
  net/sched: flower: Support hardware miss to tc action
  net/sched: flower: Move filter handle initialization earlier
  net/sched: cls_api: Support hardware miss to tc action
  net/sched: Rename user cookie and act cookie
  sfc: fix builds without CONFIG_RTC_LIB
  sfc: clean up some inconsistent indentings
  net/mlx4_en: Introduce flexible array to silence overflow warning
  net: lan966x: Fix possible deadlock inside PTP
  net/ulp: Remove redundant ->clone() test in inet_clone_ulp().
  ...

1 files changed, 1137 insertions, 0 deletions

diff --git a/drivers/mailbox/arm_mhuv2.c b/drivers/mailbox/arm_mhuv2.c
new file mode 100644
index 000000000..c6d4957c4
--- /dev/null
+++ b/drivers/mailbox/arm_mhuv2.c
@@ -0,0 +1,1137 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * ARM Message Handling Unit Version 2 (MHUv2) driver.
+ *
+ * Copyright (C) 2020 ARM Ltd.
+ * Copyright (C) 2020 Linaro Ltd.
+ *
+ * An MHUv2 mailbox controller can provide up to 124 channel windows (each 32
+ * bit long) and the driver allows any combination of both the transport
+ * protocol modes: data-transfer and doorbell, to be used on those channel
+ * windows.
+ *
+ * The transport protocols should be specified in the device tree entry for the
+ * device. The transport protocols determine how the underlying hardware
+ * resources of the device are utilized when transmitting data. Refer to the
+ * device tree bindings of the ARM MHUv2 controller for more details.
+ *
+ * The number of registered mailbox channels is dependent on both the underlying
+ * hardware - mainly the number of channel windows implemented by the platform,
+ * as well as the selected transport protocols.
+ *
+ * The MHUv2 controller can work both as a sender and receiver, but the driver
+ * and the DT bindings support unidirectional transfers for better allocation of
+ * the channels. That is, this driver will be probed for two separate devices
+ * for each mailbox controller, a sender device and a receiver device.
+ */
+
+#include <linux/amba/bus.h>
+#include <linux/interrupt.h>
+#include <linux/mailbox_controller.h>
+#include <linux/mailbox/arm_mhuv2_message.h>
+#include <linux/module.h>
+#include <linux/of_address.h>
+#include <linux/spinlock.h>
+
+/* ====== MHUv2 Registers ====== */
+
+/* Maximum number of channel windows */
+#define MHUV2_CH_WN_MAX			124
+/* Number of combined interrupt status registers */
+#define MHUV2_CMB_INT_ST_REG_CNT	4
+#define MHUV2_STAT_BYTES		(sizeof(u32))
+#define MHUV2_STAT_BITS			(MHUV2_STAT_BYTES * __CHAR_BIT__)
+
+#define LSB_MASK(n)			((1 << (n * __CHAR_BIT__)) - 1)
+#define MHUV2_PROTOCOL_PROP		"arm,mhuv2-protocols"
+
+/* Register Message Handling Unit Configuration fields */
+struct mhu_cfg_t {
+	u32 num_ch : 7;
+	u32 pad : 25;
+} __packed;
+
+/* register Interrupt Status fields */
+struct int_st_t {
+	u32 nr2r : 1;
+	u32 r2nr : 1;
+	u32 pad : 30;
+} __packed;
+
+/* Register Interrupt Clear fields */
+struct int_clr_t {
+	u32 nr2r : 1;
+	u32 r2nr : 1;
+	u32 pad : 30;
+} __packed;
+
+/* Register Interrupt Enable fields */
+struct int_en_t {
+	u32 r2nr : 1;
+	u32 nr2r : 1;
+	u32 chcomb : 1;
+	u32 pad : 29;
+} __packed;
+
+/* Register Implementer Identification fields */
+struct iidr_t {
+	u32 implementer : 12;
+	u32 revision : 4;
+	u32 variant : 4;
+	u32 product_id : 12;
+} __packed;
+
+/* Register Architecture Identification Register fields */
+struct aidr_t {
+	u32 arch_minor_rev : 4;
+	u32 arch_major_rev : 4;
+	u32 pad : 24;
+} __packed;
+
+/* Sender Channel Window fields */
+struct mhu2_send_ch_wn_reg {
+	u32 stat;
+	u8 pad1[0x0C - 0x04];
+	u32 stat_set;
+	u32 int_st;
+	u32 int_clr;
+	u32 int_en;
+	u8 pad2[0x20 - 0x1C];
+} __packed;
+
+/* Sender frame register fields */
+struct mhu2_send_frame_reg {
+	struct mhu2_send_ch_wn_reg ch_wn[MHUV2_CH_WN_MAX];
+	struct mhu_cfg_t mhu_cfg;
+	u32 resp_cfg;
+	u32 access_request;
+	u32 access_ready;
+	struct int_st_t int_st;
+	struct int_clr_t int_clr;
+	struct int_en_t int_en;
+	u32 reserved0;
+	u32 chcomb_int_st[MHUV2_CMB_INT_ST_REG_CNT];
+	u8 pad[0xFC8 - 0xFB0];
+	struct iidr_t iidr;
+	struct aidr_t aidr;
+} __packed;
+
+/* Receiver Channel Window fields */
+struct mhu2_recv_ch_wn_reg {
+	u32 stat;
+	u32 stat_masked;
+	u32 stat_clear;
+	u8 reserved0[0x10 - 0x0C];
+	u32 mask;
+	u32 mask_set;
+	u32 mask_clear;
+	u8 pad[0x20 - 0x1C];
+} __packed;
+
+/* Receiver frame register fields */
+struct mhu2_recv_frame_reg {
+	struct mhu2_recv_ch_wn_reg ch_wn[MHUV2_CH_WN_MAX];
+	struct mhu_cfg_t mhu_cfg;
+	u8 reserved0[0xF90 - 0xF84];
+	struct int_st_t int_st;
+	struct int_clr_t int_clr;
+	struct int_en_t int_en;
+	u32 pad;
+	u32 chcomb_int_st[MHUV2_CMB_INT_ST_REG_CNT];
+	u8 reserved2[0xFC8 - 0xFB0];
+	struct iidr_t iidr;
+	struct aidr_t aidr;
+} __packed;
+
+
+/* ====== MHUv2 data structures ====== */
+
+enum mhuv2_transport_protocol {
+	DOORBELL = 0,
+	DATA_TRANSFER = 1
+};
+
+enum mhuv2_frame {
+	RECEIVER_FRAME,
+	SENDER_FRAME
+};
+
+/**
+ * struct mhuv2 - MHUv2 mailbox controller data
+ *
+ * @mbox:	Mailbox controller belonging to the MHU frame.
+ * @send:	Base address of the register mapping region.
+ * @recv:	Base address of the register mapping region.
+ * @frame:	Frame type: RECEIVER_FRAME or SENDER_FRAME.
+ * @irq:	Interrupt.
+ * @windows:	Channel windows implemented by the platform.
+ * @minor:	Minor version of the controller.
+ * @length:	Length of the protocols array in bytes.
+ * @protocols:	Raw protocol information, derived from device tree.
+ * @doorbell_pending_lock: spinlock required for correct operation of Tx
+ *		interrupt for doorbells.
+ */
+struct mhuv2 {
+	struct mbox_controller mbox;
+	union {
+		struct mhu2_send_frame_reg __iomem *send;
+		struct mhu2_recv_frame_reg __iomem *recv;
+	};
+	enum mhuv2_frame frame;
+	unsigned int irq;
+	unsigned int windows;
+	unsigned int minor;
+	unsigned int length;
+	u32 *protocols;
+
+	spinlock_t doorbell_pending_lock;
+};
+
+#define mhu_from_mbox(_mbox) container_of(_mbox, struct mhuv2, mbox)
+
+/**
+ * struct mhuv2_protocol_ops - MHUv2 operations
+ *
+ * Each transport protocol must provide an implementation of the operations
+ * provided here.
+ *
+ * @rx_startup: Startup callback for receiver.
+ * @rx_shutdown: Shutdown callback for receiver.
+ * @read_data: Reads and clears newly available data.
+ * @tx_startup: Startup callback for receiver.
+ * @tx_shutdown: Shutdown callback for receiver.
+ * @last_tx_done: Report back if the last tx is completed or not.
+ * @send_data: Send data to the receiver.
+ */
+struct mhuv2_protocol_ops {
+	int (*rx_startup)(struct mhuv2 *mhu, struct mbox_chan *chan);
+	void (*rx_shutdown)(struct mhuv2 *mhu, struct mbox_chan *chan);
+	void *(*read_data)(struct mhuv2 *mhu, struct mbox_chan *chan);
+
+	void (*tx_startup)(struct mhuv2 *mhu, struct mbox_chan *chan);
+	void (*tx_shutdown)(struct mhuv2 *mhu, struct mbox_chan *chan);
+	int (*last_tx_done)(struct mhuv2 *mhu, struct mbox_chan *chan);
+	int (*send_data)(struct mhuv2 *mhu, struct mbox_chan *chan, void *arg);
+};
+
+/*
+ * MHUv2 mailbox channel's private information
+ *
+ * @ops:	protocol specific ops for the channel.
+ * @ch_wn_idx:	Channel window index allocated to the channel.
+ * @windows:	Total number of windows consumed by the channel, only relevant
+ *		in DATA_TRANSFER protocol.
+ * @doorbell:	Doorbell bit number within the ch_wn_idx window, only relevant
+ *		in DOORBELL protocol.
+ * @pending:	Flag indicating pending doorbell interrupt, only relevant in
+ *		DOORBELL protocol.
+ */
+struct mhuv2_mbox_chan_priv {
+	const struct mhuv2_protocol_ops *ops;
+	u32 ch_wn_idx;
+	union {
+		u32 windows;
+		struct {
+			u32 doorbell;
+			u32 pending;
+		};
+	};
+};
+
+/* Macro for reading a bitfield within a physically mapped packed struct */
+#define readl_relaxed_bitfield(_regptr, _type, _field)			\
+	({								\
+		u32 _regval;						\
+		_regval = readl_relaxed((_regptr));			\
+		(*(_type *)(&_regval))._field;				\
+	})
+
+/* Macro for writing a bitfield within a physically mapped packed struct */
+#define writel_relaxed_bitfield(_value, _regptr, _type, _field)		\
+	({								\
+		u32 _regval;						\
+		_regval = readl_relaxed(_regptr);			\
+		(*(_type *)(&_regval))._field = _value;			\
+		writel_relaxed(_regval, _regptr);			\
+	})
+
+
+/* =================== Doorbell transport protocol operations =============== */
+
+static int mhuv2_doorbell_rx_startup(struct mhuv2 *mhu, struct mbox_chan *chan)
+{
+	struct mhuv2_mbox_chan_priv *priv = chan->con_priv;
+
+	writel_relaxed(BIT(priv->doorbell),
+		       &mhu->recv->ch_wn[priv->ch_wn_idx].mask_clear);
+	return 0;
+}
+
+static void mhuv2_doorbell_rx_shutdown(struct mhuv2 *mhu,
+				       struct mbox_chan *chan)
+{
+	struct mhuv2_mbox_chan_priv *priv = chan->con_priv;
+
+	writel_relaxed(BIT(priv->doorbell),
+		       &mhu->recv->ch_wn[priv->ch_wn_idx].mask_set);
+}
+
+static void *mhuv2_doorbell_read_data(struct mhuv2 *mhu, struct mbox_chan *chan)
+{
+	struct mhuv2_mbox_chan_priv *priv = chan->con_priv;
+
+	writel_relaxed(BIT(priv->doorbell),
+		       &mhu->recv->ch_wn[priv->ch_wn_idx].stat_clear);
+	return NULL;
+}
+
+static int mhuv2_doorbell_last_tx_done(struct mhuv2 *mhu,
+				       struct mbox_chan *chan)
+{
+	struct mhuv2_mbox_chan_priv *priv = chan->con_priv;
+
+	return !(readl_relaxed(&mhu->send->ch_wn[priv->ch_wn_idx].stat) &
+		 BIT(priv->doorbell));
+}
+
+static int mhuv2_doorbell_send_data(struct mhuv2 *mhu, struct mbox_chan *chan,
+				    void *arg)
+{
+	struct mhuv2_mbox_chan_priv *priv = chan->con_priv;
+	unsigned long flags;
+
+	spin_lock_irqsave(&mhu->doorbell_pending_lock, flags);
+
+	priv->pending = 1;
+	writel_relaxed(BIT(priv->doorbell),
+		       &mhu->send->ch_wn[priv->ch_wn_idx].stat_set);
+
+	spin_unlock_irqrestore(&mhu->doorbell_pending_lock, flags);
+
+	return 0;
+}
+
+static const struct mhuv2_protocol_ops mhuv2_doorbell_ops = {
+	.rx_startup = mhuv2_doorbell_rx_startup,
+	.rx_shutdown = mhuv2_doorbell_rx_shutdown,
+	.read_data = mhuv2_doorbell_read_data,
+	.last_tx_done = mhuv2_doorbell_last_tx_done,
+	.send_data = mhuv2_doorbell_send_data,
+};
+#define IS_PROTOCOL_DOORBELL(_priv) (_priv->ops == &mhuv2_doorbell_ops)
+
+/* ============= Data transfer transport protocol operations ================ */
+
+static int mhuv2_data_transfer_rx_startup(struct mhuv2 *mhu,
+					  struct mbox_chan *chan)
+{
+	struct mhuv2_mbox_chan_priv *priv = chan->con_priv;
+	int i = priv->ch_wn_idx + priv->windows - 1;
+
+	/*
+	 * The protocol mandates that all but the last status register must be
+	 * masked.
+	 */
+	writel_relaxed(0xFFFFFFFF, &mhu->recv->ch_wn[i].mask_clear);
+	return 0;
+}
+
+static void mhuv2_data_transfer_rx_shutdown(struct mhuv2 *mhu,
+					    struct mbox_chan *chan)
+{
+	struct mhuv2_mbox_chan_priv *priv = chan->con_priv;
+	int i = priv->ch_wn_idx + priv->windows - 1;
+
+	writel_relaxed(0xFFFFFFFF, &mhu->recv->ch_wn[i].mask_set);
+}
+
+static void *mhuv2_data_transfer_read_data(struct mhuv2 *mhu,
+					   struct mbox_chan *chan)
+{
+	struct mhuv2_mbox_chan_priv *priv = chan->con_priv;
+	const int windows = priv->windows;
+	struct arm_mhuv2_mbox_msg *msg;
+	u32 *data;
+	int i, idx;
+
+	msg = kzalloc(sizeof(*msg) + windows * MHUV2_STAT_BYTES, GFP_KERNEL);
+	if (!msg)
+		return ERR_PTR(-ENOMEM);
+
+	data = msg->data = msg + 1;
+	msg->len = windows * MHUV2_STAT_BYTES;
+
+	/*
+	 * Messages are expected in order of most significant word to least
+	 * significant word. Refer mhuv2_data_transfer_send_data() for more
+	 * details.
+	 *
+	 * We also need to read the stat register instead of stat_masked, as we
+	 * masked all but the last window.
+	 *
+	 * Last channel window must be cleared as the final operation. Upon
+	 * clearing the last channel window register, which is unmasked in
+	 * data-transfer protocol, the interrupt is de-asserted.
+	 */
+	for (i = 0; i < windows; i++) {
+		idx = priv->ch_wn_idx + i;
+		data[windows - 1 - i] = readl_relaxed(&mhu->recv->ch_wn[idx].stat);
+		writel_relaxed(0xFFFFFFFF, &mhu->recv->ch_wn[idx].stat_clear);
+	}
+
+	return msg;
+}
+
+static void mhuv2_data_transfer_tx_startup(struct mhuv2 *mhu,
+					   struct mbox_chan *chan)
+{
+	struct mhuv2_mbox_chan_priv *priv = chan->con_priv;
+	int i = priv->ch_wn_idx + priv->windows - 1;
+
+	/* Enable interrupts only for the last window */
+	if (mhu->minor) {
+		writel_relaxed(0x1, &mhu->send->ch_wn[i].int_clr);
+		writel_relaxed(0x1, &mhu->send->ch_wn[i].int_en);
+	}
+}
+
+static void mhuv2_data_transfer_tx_shutdown(struct mhuv2 *mhu,
+					    struct mbox_chan *chan)
+{
+	struct mhuv2_mbox_chan_priv *priv = chan->con_priv;
+	int i = priv->ch_wn_idx + priv->windows - 1;
+
+	if (mhu->minor)
+		writel_relaxed(0x0, &mhu->send->ch_wn[i].int_en);
+}
+
+static int mhuv2_data_transfer_last_tx_done(struct mhuv2 *mhu,
+					    struct mbox_chan *chan)
+{
+	struct mhuv2_mbox_chan_priv *priv = chan->con_priv;
+	int i = priv->ch_wn_idx + priv->windows - 1;
+
+	/* Just checking the last channel window should be enough */
+	return !readl_relaxed(&mhu->send->ch_wn[i].stat);
+}
+
+/*
+ * Message will be transmitted from most significant to least significant word.
+ * This is to allow for messages shorter than channel windows to still trigger
+ * the receiver interrupt which gets activated when the last stat register is
+ * written. As an example, a 6-word message is to be written on a 4-channel MHU
+ * connection: Registers marked with '*' are masked, and will not generate an
+ * interrupt on the receiver side once written.
+ *
+ * u32 *data =	[0x00000001], [0x00000002], [0x00000003], [0x00000004],
+ *		[0x00000005], [0x00000006]
+ *
+ * ROUND 1:
+ * stat reg		To write	Write sequence
+ * [ stat 3 ]	<-	[0x00000001]	4 <- triggers interrupt on receiver
+ * [ stat 2 ]	<-	[0x00000002]	3
+ * [ stat 1 ]	<-	[0x00000003]	2
+ * [ stat 0 ]	<-	[0x00000004]	1
+ *
+ * data += 4 // Increment data pointer by number of stat regs
+ *
+ * ROUND 2:
+ * stat reg		To write	Write sequence
+ * [ stat 3 ]	<-	[0x00000005]	2 <- triggers interrupt on receiver
+ * [ stat 2 ]	<-	[0x00000006]	1
+ * [ stat 1 ]	<-	[0x00000000]
+ * [ stat 0 ]	<-	[0x00000000]
+ */
+static int mhuv2_data_transfer_send_data(struct mhuv2 *mhu,
+					 struct mbox_chan *chan, void *arg)
+{
+	const struct arm_mhuv2_mbox_msg *msg = arg;
+	int bytes_left = msg->len, bytes_to_send, bytes_in_round, i;
+	struct mhuv2_mbox_chan_priv *priv = chan->con_priv;
+	int windows = priv->windows;
+	u32 *data = msg->data, word;
+
+	while (bytes_left) {
+		if (!data[0]) {
+			dev_err(mhu->mbox.dev, "Data aligned at first window can't be zero to guarantee interrupt generation at receiver");
+			return -EINVAL;
+		}
+
+		while(!mhuv2_data_transfer_last_tx_done(mhu, chan))
+			continue;
+
+		bytes_in_round = min(bytes_left, (int)(windows * MHUV2_STAT_BYTES));
+
+		for (i = windows - 1; i >= 0; i--) {
+			/* Data less than windows can transfer ? */
+			if (unlikely(bytes_in_round <= i * MHUV2_STAT_BYTES))
+				continue;
+
+			word = data[i];
+			bytes_to_send = bytes_in_round & (MHUV2_STAT_BYTES - 1);
+			if (unlikely(bytes_to_send))
+				word &= LSB_MASK(bytes_to_send);
+			else
+				bytes_to_send = MHUV2_STAT_BYTES;
+
+			writel_relaxed(word, &mhu->send->ch_wn[priv->ch_wn_idx + windows - 1 - i].stat_set);
+			bytes_left -= bytes_to_send;
+			bytes_in_round -= bytes_to_send;
+		}
+
+		data += windows;
+	}
+
+	return 0;
+}
+
+static const struct mhuv2_protocol_ops mhuv2_data_transfer_ops = {
+	.rx_startup = mhuv2_data_transfer_rx_startup,
+	.rx_shutdown = mhuv2_data_transfer_rx_shutdown,
+	.read_data = mhuv2_data_transfer_read_data,
+	.tx_startup = mhuv2_data_transfer_tx_startup,
+	.tx_shutdown = mhuv2_data_transfer_tx_shutdown,
+	.last_tx_done = mhuv2_data_transfer_last_tx_done,
+	.send_data = mhuv2_data_transfer_send_data,
+};
+
+/* Interrupt handlers */
+
+static struct mbox_chan *get_irq_chan_comb(struct mhuv2 *mhu, u32 __iomem *reg)
+{
+	struct mbox_chan *chans = mhu->mbox.chans;
+	int channel = 0, i, offset = 0, windows, protocol, ch_wn;
+	u32 stat;
+
+	for (i = 0; i < MHUV2_CMB_INT_ST_REG_CNT; i++) {
+		stat = readl_relaxed(reg + i);
+		if (!stat)
+			continue;
+
+		ch_wn = i * MHUV2_STAT_BITS + __builtin_ctz(stat);
+
+		for (i = 0; i < mhu->length; i += 2) {
+			protocol = mhu->protocols[i];
+			windows = mhu->protocols[i + 1];
+
+			if (ch_wn >= offset + windows) {
+				if (protocol == DOORBELL)
+					channel += MHUV2_STAT_BITS * windows;
+				else
+					channel++;
+
+				offset += windows;
+				continue;
+			}
+
+			/* Return first chan of the window in doorbell mode */
+			if (protocol == DOORBELL)
+				channel += MHUV2_STAT_BITS * (ch_wn - offset);
+
+			return &chans[channel];
+		}
+	}
+
+	return ERR_PTR(-EIO);
+}
+
+static irqreturn_t mhuv2_sender_interrupt(int irq, void *data)
+{
+	struct mhuv2 *mhu = data;
+	struct device *dev = mhu->mbox.dev;
+	struct mhuv2_mbox_chan_priv *priv;
+	struct mbox_chan *chan;
+	unsigned long flags;
+	int i, found = 0;
+	u32 stat;
+
+	chan = get_irq_chan_comb(mhu, mhu->send->chcomb_int_st);
+	if (IS_ERR(chan)) {
+		dev_warn(dev, "Failed to find channel for the Tx interrupt\n");
+		return IRQ_NONE;
+	}
+	priv = chan->con_priv;
+
+	if (!IS_PROTOCOL_DOORBELL(priv)) {
+		writel_relaxed(1, &mhu->send->ch_wn[priv->ch_wn_idx + priv->windows - 1].int_clr);
+
+		if (chan->cl) {
+			mbox_chan_txdone(chan, 0);
+			return IRQ_HANDLED;
+		}
+
+		dev_warn(dev, "Tx interrupt Received on channel (%u) not currently attached to a mailbox client\n",
+			 priv->ch_wn_idx);
+		return IRQ_NONE;
+	}
+
+	/* Clear the interrupt first, so we don't miss any doorbell later */
+	writel_relaxed(1, &mhu->send->ch_wn[priv->ch_wn_idx].int_clr);
+
+	/*
+	 * In Doorbell mode, make sure no new transitions happen while the
+	 * interrupt handler is trying to find the finished doorbell tx
+	 * operations, else we may think few of the transfers were complete
+	 * before they actually were.
+	 */
+	spin_lock_irqsave(&mhu->doorbell_pending_lock, flags);
+
+	/*
+	 * In case of doorbell mode, the first channel of the window is returned
+	 * by get_irq_chan_comb(). Find all the pending channels here.
+	 */
+	stat = readl_relaxed(&mhu->send->ch_wn[priv->ch_wn_idx].stat);
+
+	for (i = 0; i < MHUV2_STAT_BITS; i++) {
+		priv = chan[i].con_priv;
+
+		/* Find cases where pending was 1, but stat's bit is cleared */
+		if (priv->pending ^ ((stat >> i) & 0x1)) {
+			BUG_ON(!priv->pending);
+
+			if (!chan->cl) {
+				dev_warn(dev, "Tx interrupt received on doorbell (%u : %u) channel not currently attached to a mailbox client\n",
+					 priv->ch_wn_idx, i);
+				continue;
+			}
+
+			mbox_chan_txdone(&chan[i], 0);
+			priv->pending = 0;
+			found++;
+		}
+	}
+
+	spin_unlock_irqrestore(&mhu->doorbell_pending_lock, flags);
+
+	if (!found) {
+		/*
+		 * We may have already processed the doorbell in the previous
+		 * iteration if the interrupt came right after we cleared it but
+		 * before we read the stat register.
+		 */
+		dev_dbg(dev, "Couldn't find the doorbell (%u) for the Tx interrupt interrupt\n",
+			priv->ch_wn_idx);
+		return IRQ_NONE;
+	}
+
+	return IRQ_HANDLED;
+}
+
+static struct mbox_chan *get_irq_chan_comb_rx(struct mhuv2 *mhu)
+{
+	struct mhuv2_mbox_chan_priv *priv;
+	struct mbox_chan *chan;
+	u32 stat;
+
+	chan = get_irq_chan_comb(mhu, mhu->recv->chcomb_int_st);
+	if (IS_ERR(chan))
+		return chan;
+
+	priv = chan->con_priv;
+	if (!IS_PROTOCOL_DOORBELL(priv))
+		return chan;
+
+	/*
+	 * In case of doorbell mode, the first channel of the window is returned
+	 * by the routine. Find the exact channel here.
+	 */
+	stat = readl_relaxed(&mhu->recv->ch_wn[priv->ch_wn_idx].stat_masked);
+	BUG_ON(!stat);
+
+	return chan + __builtin_ctz(stat);
+}
+
+static struct mbox_chan *get_irq_chan_stat_rx(struct mhuv2 *mhu)
+{
+	struct mbox_chan *chans = mhu->mbox.chans;
+	struct mhuv2_mbox_chan_priv *priv;
+	u32 stat;
+	int i = 0;
+
+	while (i < mhu->mbox.num_chans) {
+		priv = chans[i].con_priv;
+		stat = readl_relaxed(&mhu->recv->ch_wn[priv->ch_wn_idx].stat_masked);
+
+		if (stat) {
+			if (IS_PROTOCOL_DOORBELL(priv))
+				i += __builtin_ctz(stat);
+			return &chans[i];
+		}
+
+		i += IS_PROTOCOL_DOORBELL(priv) ? MHUV2_STAT_BITS : 1;
+	}
+
+	return ERR_PTR(-EIO);
+}
+
+static struct mbox_chan *get_irq_chan_rx(struct mhuv2 *mhu)
+{
+	if (!mhu->minor)
+		return get_irq_chan_stat_rx(mhu);
+
+	return get_irq_chan_comb_rx(mhu);
+}
+
+static irqreturn_t mhuv2_receiver_interrupt(int irq, void *arg)
+{
+	struct mhuv2 *mhu = arg;
+	struct mbox_chan *chan = get_irq_chan_rx(mhu);
+	struct device *dev = mhu->mbox.dev;
+	struct mhuv2_mbox_chan_priv *priv;
+	int ret = IRQ_NONE;
+	void *data;
+
+	if (IS_ERR(chan)) {
+		dev_warn(dev, "Failed to find channel for the rx interrupt\n");
+		return IRQ_NONE;
+	}
+	priv = chan->con_priv;
+
+	/* Read and clear the data first */
+	data = priv->ops->read_data(mhu, chan);
+
+	if (!chan->cl) {
+		dev_warn(dev, "Received data on channel (%u) not currently attached to a mailbox client\n",
+			 priv->ch_wn_idx);
+	} else if (IS_ERR(data)) {
+		dev_err(dev, "Failed to read data: %lu\n", PTR_ERR(data));
+	} else {
+		mbox_chan_received_data(chan, data);
+		ret = IRQ_HANDLED;
+	}
+
+	if (!IS_ERR(data))
+		kfree(data);
+
+	return ret;
+}
+
+/* Sender and receiver ops */
+static bool mhuv2_sender_last_tx_done(struct mbox_chan *chan)
+{
+	struct mhuv2 *mhu = mhu_from_mbox(chan->mbox);
+	struct mhuv2_mbox_chan_priv *priv = chan->con_priv;
+
+	return priv->ops->last_tx_done(mhu, chan);
+}
+
+static int mhuv2_sender_send_data(struct mbox_chan *chan, void *data)
+{
+	struct mhuv2 *mhu = mhu_from_mbox(chan->mbox);
+	struct mhuv2_mbox_chan_priv *priv = chan->con_priv;
+
+	if (!priv->ops->last_tx_done(mhu, chan))
+		return -EBUSY;
+
+	return priv->ops->send_data(mhu, chan, data);
+}
+
+static int mhuv2_sender_startup(struct mbox_chan *chan)
+{
+	struct mhuv2 *mhu = mhu_from_mbox(chan->mbox);
+	struct mhuv2_mbox_chan_priv *priv = chan->con_priv;
+
+	if (priv->ops->tx_startup)
+		priv->ops->tx_startup(mhu, chan);
+	return 0;
+}
+
+static void mhuv2_sender_shutdown(struct mbox_chan *chan)
+{
+	struct mhuv2 *mhu = mhu_from_mbox(chan->mbox);
+	struct mhuv2_mbox_chan_priv *priv = chan->con_priv;
+
+	if (priv->ops->tx_shutdown)
+		priv->ops->tx_shutdown(mhu, chan);
+}
+
+static const struct mbox_chan_ops mhuv2_sender_ops = {
+	.send_data = mhuv2_sender_send_data,
+	.startup = mhuv2_sender_startup,
+	.shutdown = mhuv2_sender_shutdown,
+	.last_tx_done = mhuv2_sender_last_tx_done,
+};
+
+static int mhuv2_receiver_startup(struct mbox_chan *chan)
+{
+	struct mhuv2 *mhu = mhu_from_mbox(chan->mbox);
+	struct mhuv2_mbox_chan_priv *priv = chan->con_priv;
+
+	return priv->ops->rx_startup(mhu, chan);
+}
+
+static void mhuv2_receiver_shutdown(struct mbox_chan *chan)
+{
+	struct mhuv2 *mhu = mhu_from_mbox(chan->mbox);
+	struct mhuv2_mbox_chan_priv *priv = chan->con_priv;
+
+	priv->ops->rx_shutdown(mhu, chan);
+}
+
+static int mhuv2_receiver_send_data(struct mbox_chan *chan, void *data)
+{
+	dev_err(chan->mbox->dev,
+		"Trying to transmit on a receiver MHU frame\n");
+	return -EIO;
+}
+
+static bool mhuv2_receiver_last_tx_done(struct mbox_chan *chan)
+{
+	dev_err(chan->mbox->dev, "Trying to Tx poll on a receiver MHU frame\n");
+	return true;
+}
+
+static const struct mbox_chan_ops mhuv2_receiver_ops = {
+	.send_data = mhuv2_receiver_send_data,
+	.startup = mhuv2_receiver_startup,
+	.shutdown = mhuv2_receiver_shutdown,
+	.last_tx_done = mhuv2_receiver_last_tx_done,
+};
+
+static struct mbox_chan *mhuv2_mbox_of_xlate(struct mbox_controller *mbox,
+					     const struct of_phandle_args *pa)
+{
+	struct mhuv2 *mhu = mhu_from_mbox(mbox);
+	struct mbox_chan *chans = mbox->chans;
+	int channel = 0, i, offset, doorbell, protocol, windows;
+
+	if (pa->args_count != 2)
+		return ERR_PTR(-EINVAL);
+
+	offset = pa->args[0];
+	doorbell = pa->args[1];
+	if (doorbell >= MHUV2_STAT_BITS)
+		goto out;
+
+	for (i = 0; i < mhu->length; i += 2) {
+		protocol = mhu->protocols[i];
+		windows = mhu->protocols[i + 1];
+
+		if (protocol == DOORBELL) {
+			if (offset < windows)
+				return &chans[channel + MHUV2_STAT_BITS * offset + doorbell];
+
+			channel += MHUV2_STAT_BITS * windows;
+			offset -= windows;
+		} else {
+			if (offset == 0) {
+				if (doorbell)
+					goto out;
+
+				return &chans[channel];
+			}
+
+			channel++;
+			offset--;
+		}
+	}
+
+out:
+	dev_err(mbox->dev, "Couldn't xlate to a valid channel (%d: %d)\n",
+		pa->args[0], doorbell);
+	return ERR_PTR(-ENODEV);
+}
+
+static int mhuv2_verify_protocol(struct mhuv2 *mhu)
+{
+	struct device *dev = mhu->mbox.dev;
+	int protocol, windows, channels = 0, total_windows = 0, i;
+
+	for (i = 0; i < mhu->length; i += 2) {
+		protocol = mhu->protocols[i];
+		windows = mhu->protocols[i + 1];
+
+		if (!windows) {
+			dev_err(dev, "Window size can't be zero (%d)\n", i);
+			return -EINVAL;
+		}
+		total_windows += windows;
+
+		if (protocol == DOORBELL) {
+			channels += MHUV2_STAT_BITS * windows;
+		} else if (protocol == DATA_TRANSFER) {
+			channels++;
+		} else {
+			dev_err(dev, "Invalid protocol (%d) present in %s property at index %d\n",
+				protocol, MHUV2_PROTOCOL_PROP, i);
+			return -EINVAL;
+		}
+	}
+
+	if (total_windows > mhu->windows) {
+		dev_err(dev, "Channel windows can't be more than what's implemented by the hardware ( %d: %d)\n",
+			total_windows, mhu->windows);
+		return -EINVAL;
+	}
+
+	mhu->mbox.num_chans = channels;
+	return 0;
+}
+
+static int mhuv2_allocate_channels(struct mhuv2 *mhu)
+{
+	struct mbox_controller *mbox = &mhu->mbox;
+	struct mhuv2_mbox_chan_priv *priv;
+	struct device *dev = mbox->dev;
+	struct mbox_chan *chans;
+	int protocol, windows = 0, next_window = 0, i, j, k;
+
+	chans = devm_kcalloc(dev, mbox->num_chans, sizeof(*chans), GFP_KERNEL);
+	if (!chans)
+		return -ENOMEM;
+
+	mbox->chans = chans;
+
+	for (i = 0; i < mhu->length; i += 2) {
+		next_window += windows;
+
+		protocol = mhu->protocols[i];
+		windows = mhu->protocols[i + 1];
+
+		if (protocol == DATA_TRANSFER) {
+			priv = devm_kmalloc(dev, sizeof(*priv), GFP_KERNEL);
+			if (!priv)
+				return -ENOMEM;
+
+			priv->ch_wn_idx = next_window;
+			priv->ops = &mhuv2_data_transfer_ops;
+			priv->windows = windows;
+			chans++->con_priv = priv;
+			continue;
+		}
+
+		for (j = 0; j < windows; j++) {
+			for (k = 0; k < MHUV2_STAT_BITS; k++) {
+				priv = devm_kmalloc(dev, sizeof(*priv), GFP_KERNEL);
+				if (!priv)
+					return -ENOMEM;
+
+				priv->ch_wn_idx = next_window + j;
+				priv->ops = &mhuv2_doorbell_ops;
+				priv->doorbell = k;
+				chans++->con_priv = priv;
+			}
+
+			/*
+			 * Permanently enable interrupt as we can't
+			 * control it per doorbell.
+			 */
+			if (mhu->frame == SENDER_FRAME && mhu->minor)
+				writel_relaxed(0x1, &mhu->send->ch_wn[priv->ch_wn_idx].int_en);
+		}
+	}
+
+	/* Make sure we have initialized all channels */
+	BUG_ON(chans - mbox->chans != mbox->num_chans);
+
+	return 0;
+}
+
+static int mhuv2_parse_channels(struct mhuv2 *mhu)
+{
+	struct device *dev = mhu->mbox.dev;
+	const struct device_node *np = dev->of_node;
+	int ret, count;
+	u32 *protocols;
+
+	count = of_property_count_u32_elems(np, MHUV2_PROTOCOL_PROP);
+	if (count <= 0 || count % 2) {
+		dev_err(dev, "Invalid %s property (%d)\n", MHUV2_PROTOCOL_PROP,
+			count);
+		return -EINVAL;
+	}
+
+	protocols = devm_kmalloc_array(dev, count, sizeof(*protocols), GFP_KERNEL);
+	if (!protocols)
+		return -ENOMEM;
+
+	ret = of_property_read_u32_array(np, MHUV2_PROTOCOL_PROP, protocols, count);
+	if (ret) {
+		dev_err(dev, "Failed to read %s property: %d\n",
+			MHUV2_PROTOCOL_PROP, ret);
+		return ret;
+	}
+
+	mhu->protocols = protocols;
+	mhu->length = count;
+
+	ret = mhuv2_verify_protocol(mhu);
+	if (ret)
+		return ret;
+
+	return mhuv2_allocate_channels(mhu);
+}
+
+static int mhuv2_tx_init(struct amba_device *adev, struct mhuv2 *mhu,
+			 void __iomem *reg)
+{
+	struct device *dev = mhu->mbox.dev;
+	int ret, i;
+
+	mhu->frame = SENDER_FRAME;
+	mhu->mbox.ops = &mhuv2_sender_ops;
+	mhu->send = reg;
+
+	mhu->windows = readl_relaxed_bitfield(&mhu->send->mhu_cfg, struct mhu_cfg_t, num_ch);
+	mhu->minor = readl_relaxed_bitfield(&mhu->send->aidr, struct aidr_t, arch_minor_rev);
+
+	spin_lock_init(&mhu->doorbell_pending_lock);
+
+	/*
+	 * For minor version 1 and forward, tx interrupt is provided by
+	 * the controller.
+	 */
+	if (mhu->minor && adev->irq[0]) {
+		ret = devm_request_threaded_irq(dev, adev->irq[0], NULL,
+						mhuv2_sender_interrupt,
+						IRQF_ONESHOT, "mhuv2-tx", mhu);
+		if (ret) {
+			dev_err(dev, "Failed to request tx IRQ, fallback to polling mode: %d\n",
+				ret);
+		} else {
+			mhu->mbox.txdone_irq = true;
+			mhu->mbox.txdone_poll = false;
+			mhu->irq = adev->irq[0];
+
+			writel_relaxed_bitfield(1, &mhu->send->int_en, struct int_en_t, chcomb);
+
+			/* Disable all channel interrupts */
+			for (i = 0; i < mhu->windows; i++)
+				writel_relaxed(0x0, &mhu->send->ch_wn[i].int_en);
+
+			goto out;
+		}
+	}
+
+	mhu->mbox.txdone_irq = false;
+	mhu->mbox.txdone_poll = true;
+	mhu->mbox.txpoll_period = 1;
+
+out:
+	/* Wait for receiver to be ready */
+	writel_relaxed(0x1, &mhu->send->access_request);
+	while (!readl_relaxed(&mhu->send->access_ready))
+		continue;
+
+	return 0;
+}
+
+static int mhuv2_rx_init(struct amba_device *adev, struct mhuv2 *mhu,
+			 void __iomem *reg)
+{
+	struct device *dev = mhu->mbox.dev;
+	int ret, i;
+
+	mhu->frame = RECEIVER_FRAME;
+	mhu->mbox.ops = &mhuv2_receiver_ops;
+	mhu->recv = reg;
+
+	mhu->windows = readl_relaxed_bitfield(&mhu->recv->mhu_cfg, struct mhu_cfg_t, num_ch);
+	mhu->minor = readl_relaxed_bitfield(&mhu->recv->aidr, struct aidr_t, arch_minor_rev);
+
+	mhu->irq = adev->irq[0];
+	if (!mhu->irq) {
+		dev_err(dev, "Missing receiver IRQ\n");
+		return -EINVAL;
+	}
+
+	ret = devm_request_threaded_irq(dev, mhu->irq, NULL,
+					mhuv2_receiver_interrupt, IRQF_ONESHOT,
+					"mhuv2-rx", mhu);
+	if (ret) {
+		dev_err(dev, "Failed to request rx IRQ\n");
+		return ret;
+	}
+
+	/* Mask all the channel windows */
+	for (i = 0; i < mhu->windows; i++)
+		writel_relaxed(0xFFFFFFFF, &mhu->recv->ch_wn[i].mask_set);
+
+	if (mhu->minor)
+		writel_relaxed_bitfield(1, &mhu->recv->int_en, struct int_en_t, chcomb);
+
+	return 0;
+}
+
+static int mhuv2_probe(struct amba_device *adev, const struct amba_id *id)
+{
+	struct device *dev = &adev->dev;
+	const struct device_node *np = dev->of_node;
+	struct mhuv2 *mhu;
+	void __iomem *reg;
+	int ret = -EINVAL;
+
+	reg = devm_of_iomap(dev, dev->of_node, 0, NULL);
+	if (IS_ERR(reg))
+		return PTR_ERR(reg);
+
+	mhu = devm_kzalloc(dev, sizeof(*mhu), GFP_KERNEL);
+	if (!mhu)
+		return -ENOMEM;
+
+	mhu->mbox.dev = dev;
+	mhu->mbox.of_xlate = mhuv2_mbox_of_xlate;
+
+	if (of_device_is_compatible(np, "arm,mhuv2-tx"))
+		ret = mhuv2_tx_init(adev, mhu, reg);
+	else if (of_device_is_compatible(np, "arm,mhuv2-rx"))
+		ret = mhuv2_rx_init(adev, mhu, reg);
+	else
+		dev_err(dev, "Invalid compatible property\n");
+
+	if (ret)
+		return ret;
+
+	/* Channel windows can't be 0 */
+	BUG_ON(!mhu->windows);
+
+	ret = mhuv2_parse_channels(mhu);
+	if (ret)
+		return ret;
+
+	amba_set_drvdata(adev, mhu);
+
+	ret = devm_mbox_controller_register(dev, &mhu->mbox);
+	if (ret)
+		dev_err(dev, "failed to register ARM MHUv2 driver %d\n", ret);
+
+	return ret;
+}
+
+static void mhuv2_remove(struct amba_device *adev)
+{
+	struct mhuv2 *mhu = amba_get_drvdata(adev);
+
+	if (mhu->frame == SENDER_FRAME)
+		writel_relaxed(0x0, &mhu->send->access_request);
+}
+
+static struct amba_id mhuv2_ids[] = {
+	{
+		/* 2.0 */
+		.id = 0xbb0d1,
+		.mask = 0xfffff,
+	},
+	{
+		/* 2.1 */
+		.id = 0xbb076,
+		.mask = 0xfffff,
+	},
+	{ 0, 0 },
+};
+MODULE_DEVICE_TABLE(amba, mhuv2_ids);
+
+static struct amba_driver mhuv2_driver = {
+	.drv = {
+		.name	= "arm-mhuv2",
+	},
+	.id_table	= mhuv2_ids,
+	.probe		= mhuv2_probe,
+	.remove		= mhuv2_remove,
+};
+module_amba_driver(mhuv2_driver);
+
+MODULE_LICENSE("GPL v2");
+MODULE_DESCRIPTION("ARM MHUv2 Driver");
+MODULE_AUTHOR("Viresh Kumar <viresh.kumar@linaro.org>");
+MODULE_AUTHOR("Tushar Khandelwal <tushar.khandelwal@arm.com>");