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, 789 insertions, 0 deletions

diff --git a/drivers/net/wireless/ath/ath12k/htc.c b/drivers/net/wireless/ath/ath12k/htc.c
new file mode 100644
index 000000000..23f7428ab
--- /dev/null
+++ b/drivers/net/wireless/ath/ath12k/htc.c
@@ -0,0 +1,789 @@
+// SPDX-License-Identifier: BSD-3-Clause-Clear
+/*
+ * Copyright (c) 2018-2021 The Linux Foundation. All rights reserved.
+ * Copyright (c) 2021-2022 Qualcomm Innovation Center, Inc. All rights reserved.
+ */
+#include <linux/skbuff.h>
+#include <linux/ctype.h>
+
+#include "debug.h"
+#include "hif.h"
+
+struct sk_buff *ath12k_htc_alloc_skb(struct ath12k_base *ab, int size)
+{
+	struct sk_buff *skb;
+
+	skb = dev_alloc_skb(size + sizeof(struct ath12k_htc_hdr));
+	if (!skb)
+		return NULL;
+
+	skb_reserve(skb, sizeof(struct ath12k_htc_hdr));
+
+	/* FW/HTC requires 4-byte aligned streams */
+	if (!IS_ALIGNED((unsigned long)skb->data, 4))
+		ath12k_warn(ab, "Unaligned HTC tx skb\n");
+
+	return skb;
+}
+
+static void ath12k_htc_control_tx_complete(struct ath12k_base *ab,
+					   struct sk_buff *skb)
+{
+	kfree_skb(skb);
+}
+
+static struct sk_buff *ath12k_htc_build_tx_ctrl_skb(void)
+{
+	struct sk_buff *skb;
+	struct ath12k_skb_cb *skb_cb;
+
+	skb = dev_alloc_skb(ATH12K_HTC_CONTROL_BUFFER_SIZE);
+	if (!skb)
+		return NULL;
+
+	skb_reserve(skb, sizeof(struct ath12k_htc_hdr));
+	WARN_ON_ONCE(!IS_ALIGNED((unsigned long)skb->data, 4));
+
+	skb_cb = ATH12K_SKB_CB(skb);
+	memset(skb_cb, 0, sizeof(*skb_cb));
+
+	return skb;
+}
+
+static void ath12k_htc_prepare_tx_skb(struct ath12k_htc_ep *ep,
+				      struct sk_buff *skb)
+{
+	struct ath12k_htc_hdr *hdr;
+
+	hdr = (struct ath12k_htc_hdr *)skb->data;
+
+	memset(hdr, 0, sizeof(*hdr));
+	hdr->htc_info = le32_encode_bits(ep->eid, HTC_HDR_ENDPOINTID) |
+			le32_encode_bits((skb->len - sizeof(*hdr)),
+					 HTC_HDR_PAYLOADLEN);
+
+	if (ep->tx_credit_flow_enabled)
+		hdr->htc_info |= le32_encode_bits(ATH12K_HTC_FLAG_NEED_CREDIT_UPDATE,
+						  HTC_HDR_FLAGS);
+
+	spin_lock_bh(&ep->htc->tx_lock);
+	hdr->ctrl_info = le32_encode_bits(ep->seq_no++, HTC_HDR_CONTROLBYTES1);
+	spin_unlock_bh(&ep->htc->tx_lock);
+}
+
+int ath12k_htc_send(struct ath12k_htc *htc,
+		    enum ath12k_htc_ep_id eid,
+		    struct sk_buff *skb)
+{
+	struct ath12k_htc_ep *ep = &htc->endpoint[eid];
+	struct ath12k_skb_cb *skb_cb = ATH12K_SKB_CB(skb);
+	struct device *dev = htc->ab->dev;
+	struct ath12k_base *ab = htc->ab;
+	int credits = 0;
+	int ret;
+
+	if (eid >= ATH12K_HTC_EP_COUNT) {
+		ath12k_warn(ab, "Invalid endpoint id: %d\n", eid);
+		return -ENOENT;
+	}
+
+	skb_push(skb, sizeof(struct ath12k_htc_hdr));
+
+	if (ep->tx_credit_flow_enabled) {
+		credits = DIV_ROUND_UP(skb->len, htc->target_credit_size);
+		spin_lock_bh(&htc->tx_lock);
+		if (ep->tx_credits < credits) {
+			ath12k_dbg(ab, ATH12K_DBG_HTC,
+				   "htc insufficient credits ep %d required %d available %d\n",
+				   eid, credits, ep->tx_credits);
+			spin_unlock_bh(&htc->tx_lock);
+			ret = -EAGAIN;
+			goto err_pull;
+		}
+		ep->tx_credits -= credits;
+		ath12k_dbg(ab, ATH12K_DBG_HTC,
+			   "htc ep %d consumed %d credits (total %d)\n",
+			   eid, credits, ep->tx_credits);
+		spin_unlock_bh(&htc->tx_lock);
+	}
+
+	ath12k_htc_prepare_tx_skb(ep, skb);
+
+	skb_cb->paddr = dma_map_single(dev, skb->data, skb->len, DMA_TO_DEVICE);
+	ret = dma_mapping_error(dev, skb_cb->paddr);
+	if (ret) {
+		ret = -EIO;
+		goto err_credits;
+	}
+
+	ret = ath12k_ce_send(htc->ab, skb, ep->ul_pipe_id, ep->eid);
+	if (ret)
+		goto err_unmap;
+
+	return 0;
+
+err_unmap:
+	dma_unmap_single(dev, skb_cb->paddr, skb->len, DMA_TO_DEVICE);
+err_credits:
+	if (ep->tx_credit_flow_enabled) {
+		spin_lock_bh(&htc->tx_lock);
+		ep->tx_credits += credits;
+		ath12k_dbg(ab, ATH12K_DBG_HTC,
+			   "htc ep %d reverted %d credits back (total %d)\n",
+			   eid, credits, ep->tx_credits);
+		spin_unlock_bh(&htc->tx_lock);
+
+		if (ep->ep_ops.ep_tx_credits)
+			ep->ep_ops.ep_tx_credits(htc->ab);
+	}
+err_pull:
+	skb_pull(skb, sizeof(struct ath12k_htc_hdr));
+	return ret;
+}
+
+static void
+ath12k_htc_process_credit_report(struct ath12k_htc *htc,
+				 const struct ath12k_htc_credit_report *report,
+				 int len,
+				 enum ath12k_htc_ep_id eid)
+{
+	struct ath12k_base *ab = htc->ab;
+	struct ath12k_htc_ep *ep;
+	int i, n_reports;
+
+	if (len % sizeof(*report))
+		ath12k_warn(ab, "Uneven credit report len %d", len);
+
+	n_reports = len / sizeof(*report);
+
+	spin_lock_bh(&htc->tx_lock);
+	for (i = 0; i < n_reports; i++, report++) {
+		if (report->eid >= ATH12K_HTC_EP_COUNT)
+			break;
+
+		ep = &htc->endpoint[report->eid];
+		ep->tx_credits += report->credits;
+
+		ath12k_dbg(ab, ATH12K_DBG_HTC, "htc ep %d got %d credits (total %d)\n",
+			   report->eid, report->credits, ep->tx_credits);
+
+		if (ep->ep_ops.ep_tx_credits) {
+			spin_unlock_bh(&htc->tx_lock);
+			ep->ep_ops.ep_tx_credits(htc->ab);
+			spin_lock_bh(&htc->tx_lock);
+		}
+	}
+	spin_unlock_bh(&htc->tx_lock);
+}
+
+static int ath12k_htc_process_trailer(struct ath12k_htc *htc,
+				      u8 *buffer,
+				      int length,
+				      enum ath12k_htc_ep_id src_eid)
+{
+	struct ath12k_base *ab = htc->ab;
+	int status = 0;
+	struct ath12k_htc_record *record;
+	size_t len;
+
+	while (length > 0) {
+		record = (struct ath12k_htc_record *)buffer;
+
+		if (length < sizeof(record->hdr)) {
+			status = -EINVAL;
+			break;
+		}
+
+		if (record->hdr.len > length) {
+			/* no room left in buffer for record */
+			ath12k_warn(ab, "Invalid record length: %d\n",
+				    record->hdr.len);
+			status = -EINVAL;
+			break;
+		}
+
+		switch (record->hdr.id) {
+		case ATH12K_HTC_RECORD_CREDITS:
+			len = sizeof(struct ath12k_htc_credit_report);
+			if (record->hdr.len < len) {
+				ath12k_warn(ab, "Credit report too long\n");
+				status = -EINVAL;
+				break;
+			}
+			ath12k_htc_process_credit_report(htc,
+							 record->credit_report,
+							 record->hdr.len,
+							 src_eid);
+			break;
+		default:
+			ath12k_warn(ab, "Unhandled record: id:%d length:%d\n",
+				    record->hdr.id, record->hdr.len);
+			break;
+		}
+
+		if (status)
+			break;
+
+		/* multiple records may be present in a trailer */
+		buffer += sizeof(record->hdr) + record->hdr.len;
+		length -= sizeof(record->hdr) + record->hdr.len;
+	}
+
+	return status;
+}
+
+static void ath12k_htc_suspend_complete(struct ath12k_base *ab, bool ack)
+{
+	ath12k_dbg(ab, ATH12K_DBG_BOOT, "boot suspend complete %d\n", ack);
+
+	if (ack)
+		set_bit(ATH12K_FLAG_HTC_SUSPEND_COMPLETE, &ab->dev_flags);
+	else
+		clear_bit(ATH12K_FLAG_HTC_SUSPEND_COMPLETE, &ab->dev_flags);
+
+	complete(&ab->htc_suspend);
+}
+
+void ath12k_htc_rx_completion_handler(struct ath12k_base *ab,
+				      struct sk_buff *skb)
+{
+	int status = 0;
+	struct ath12k_htc *htc = &ab->htc;
+	struct ath12k_htc_hdr *hdr;
+	struct ath12k_htc_ep *ep;
+	u16 payload_len;
+	u32 trailer_len = 0;
+	size_t min_len;
+	u8 eid;
+	bool trailer_present;
+
+	hdr = (struct ath12k_htc_hdr *)skb->data;
+	skb_pull(skb, sizeof(*hdr));
+
+	eid = le32_get_bits(hdr->htc_info, HTC_HDR_ENDPOINTID);
+
+	if (eid >= ATH12K_HTC_EP_COUNT) {
+		ath12k_warn(ab, "HTC Rx: invalid eid %d\n", eid);
+		goto out;
+	}
+
+	ep = &htc->endpoint[eid];
+
+	payload_len = le32_get_bits(hdr->htc_info, HTC_HDR_PAYLOADLEN);
+
+	if (payload_len + sizeof(*hdr) > ATH12K_HTC_MAX_LEN) {
+		ath12k_warn(ab, "HTC rx frame too long, len: %zu\n",
+			    payload_len + sizeof(*hdr));
+		goto out;
+	}
+
+	if (skb->len < payload_len) {
+		ath12k_warn(ab, "HTC Rx: insufficient length, got %d, expected %d\n",
+			    skb->len, payload_len);
+		goto out;
+	}
+
+	/* get flags to check for trailer */
+	trailer_present = le32_get_bits(hdr->htc_info, HTC_HDR_FLAGS) &
+			  ATH12K_HTC_FLAG_TRAILER_PRESENT;
+
+	if (trailer_present) {
+		u8 *trailer;
+
+		trailer_len = le32_get_bits(hdr->ctrl_info,
+					    HTC_HDR_CONTROLBYTES0);
+		min_len = sizeof(struct ath12k_htc_record_hdr);
+
+		if ((trailer_len < min_len) ||
+		    (trailer_len > payload_len)) {
+			ath12k_warn(ab, "Invalid trailer length: %d\n",
+				    trailer_len);
+			goto out;
+		}
+
+		trailer = (u8 *)hdr;
+		trailer += sizeof(*hdr);
+		trailer += payload_len;
+		trailer -= trailer_len;
+		status = ath12k_htc_process_trailer(htc, trailer,
+						    trailer_len, eid);
+		if (status)
+			goto out;
+
+		skb_trim(skb, skb->len - trailer_len);
+	}
+
+	if (trailer_len >= payload_len)
+		/* zero length packet with trailer data, just drop these */
+		goto out;
+
+	if (eid == ATH12K_HTC_EP_0) {
+		struct ath12k_htc_msg *msg = (struct ath12k_htc_msg *)skb->data;
+
+		switch (le32_get_bits(msg->msg_svc_id, HTC_MSG_MESSAGEID)) {
+		case ATH12K_HTC_MSG_READY_ID:
+		case ATH12K_HTC_MSG_CONNECT_SERVICE_RESP_ID:
+			/* handle HTC control message */
+			if (completion_done(&htc->ctl_resp)) {
+				/* this is a fatal error, target should not be
+				 * sending unsolicited messages on the ep 0
+				 */
+				ath12k_warn(ab, "HTC rx ctrl still processing\n");
+				complete(&htc->ctl_resp);
+				goto out;
+			}
+
+			htc->control_resp_len =
+				min_t(int, skb->len,
+				      ATH12K_HTC_MAX_CTRL_MSG_LEN);
+
+			memcpy(htc->control_resp_buffer, skb->data,
+			       htc->control_resp_len);
+
+			complete(&htc->ctl_resp);
+			break;
+		case ATH12K_HTC_MSG_SEND_SUSPEND_COMPLETE:
+			ath12k_htc_suspend_complete(ab, true);
+			break;
+		case ATH12K_HTC_MSG_NACK_SUSPEND:
+			ath12k_htc_suspend_complete(ab, false);
+			break;
+		case ATH12K_HTC_MSG_WAKEUP_FROM_SUSPEND_ID:
+			break;
+		default:
+			ath12k_warn(ab, "ignoring unsolicited htc ep0 event %u\n",
+				    le32_get_bits(msg->msg_svc_id, HTC_MSG_MESSAGEID));
+			break;
+		}
+		goto out;
+	}
+
+	ath12k_dbg(ab, ATH12K_DBG_HTC, "htc rx completion ep %d skb %pK\n",
+		   eid, skb);
+	ep->ep_ops.ep_rx_complete(ab, skb);
+
+	/* poll tx completion for interrupt disabled CE's */
+	ath12k_ce_poll_send_completed(ab, ep->ul_pipe_id);
+
+	/* skb is now owned by the rx completion handler */
+	skb = NULL;
+out:
+	kfree_skb(skb);
+}
+
+static void ath12k_htc_control_rx_complete(struct ath12k_base *ab,
+					   struct sk_buff *skb)
+{
+	/* This is unexpected. FW is not supposed to send regular rx on this
+	 * endpoint.
+	 */
+	ath12k_warn(ab, "unexpected htc rx\n");
+	kfree_skb(skb);
+}
+
+static const char *htc_service_name(enum ath12k_htc_svc_id id)
+{
+	switch (id) {
+	case ATH12K_HTC_SVC_ID_RESERVED:
+		return "Reserved";
+	case ATH12K_HTC_SVC_ID_RSVD_CTRL:
+		return "Control";
+	case ATH12K_HTC_SVC_ID_WMI_CONTROL:
+		return "WMI";
+	case ATH12K_HTC_SVC_ID_WMI_DATA_BE:
+		return "DATA BE";
+	case ATH12K_HTC_SVC_ID_WMI_DATA_BK:
+		return "DATA BK";
+	case ATH12K_HTC_SVC_ID_WMI_DATA_VI:
+		return "DATA VI";
+	case ATH12K_HTC_SVC_ID_WMI_DATA_VO:
+		return "DATA VO";
+	case ATH12K_HTC_SVC_ID_WMI_CONTROL_MAC1:
+		return "WMI MAC1";
+	case ATH12K_HTC_SVC_ID_WMI_CONTROL_MAC2:
+		return "WMI MAC2";
+	case ATH12K_HTC_SVC_ID_NMI_CONTROL:
+		return "NMI Control";
+	case ATH12K_HTC_SVC_ID_NMI_DATA:
+		return "NMI Data";
+	case ATH12K_HTC_SVC_ID_HTT_DATA_MSG:
+		return "HTT Data";
+	case ATH12K_HTC_SVC_ID_TEST_RAW_STREAMS:
+		return "RAW";
+	case ATH12K_HTC_SVC_ID_IPA_TX:
+		return "IPA TX";
+	case ATH12K_HTC_SVC_ID_PKT_LOG:
+		return "PKT LOG";
+	case ATH12K_HTC_SVC_ID_WMI_CONTROL_DIAG:
+		return "WMI DIAG";
+	}
+
+	return "Unknown";
+}
+
+static void ath12k_htc_reset_endpoint_states(struct ath12k_htc *htc)
+{
+	struct ath12k_htc_ep *ep;
+	int i;
+
+	for (i = ATH12K_HTC_EP_0; i < ATH12K_HTC_EP_COUNT; i++) {
+		ep = &htc->endpoint[i];
+		ep->service_id = ATH12K_HTC_SVC_ID_UNUSED;
+		ep->max_ep_message_len = 0;
+		ep->max_tx_queue_depth = 0;
+		ep->eid = i;
+		ep->htc = htc;
+		ep->tx_credit_flow_enabled = true;
+	}
+}
+
+static u8 ath12k_htc_get_credit_allocation(struct ath12k_htc *htc,
+					   u16 service_id)
+{
+	struct ath12k_htc_svc_tx_credits *serv_entry;
+	u8 i, allocation = 0;
+
+	serv_entry = htc->service_alloc_table;
+
+	for (i = 0; i < ATH12K_HTC_MAX_SERVICE_ALLOC_ENTRIES; i++) {
+		if (serv_entry[i].service_id == service_id) {
+			allocation = serv_entry[i].credit_allocation;
+			break;
+		}
+	}
+
+	return allocation;
+}
+
+static int ath12k_htc_setup_target_buffer_assignments(struct ath12k_htc *htc)
+{
+	struct ath12k_htc_svc_tx_credits *serv_entry;
+	static const u32 svc_id[] = {
+					ATH12K_HTC_SVC_ID_WMI_CONTROL,
+					ATH12K_HTC_SVC_ID_WMI_CONTROL_MAC1,
+					ATH12K_HTC_SVC_ID_WMI_CONTROL_MAC2,
+				};
+	int i, credits;
+
+	credits = htc->total_transmit_credits;
+	serv_entry = htc->service_alloc_table;
+
+	if ((htc->wmi_ep_count == 0) ||
+	    (htc->wmi_ep_count > ARRAY_SIZE(svc_id)))
+		return -EINVAL;
+
+	/* Divide credits among number of endpoints for WMI */
+	credits = credits / htc->wmi_ep_count;
+	for (i = 0; i < htc->wmi_ep_count; i++) {
+		serv_entry[i].service_id = svc_id[i];
+		serv_entry[i].credit_allocation = credits;
+	}
+
+	return 0;
+}
+
+int ath12k_htc_wait_target(struct ath12k_htc *htc)
+{
+	int i, status = 0;
+	struct ath12k_base *ab = htc->ab;
+	unsigned long time_left;
+	struct ath12k_htc_ready *ready;
+	u16 message_id;
+	u16 credit_count;
+	u16 credit_size;
+
+	time_left = wait_for_completion_timeout(&htc->ctl_resp,
+						ATH12K_HTC_WAIT_TIMEOUT_HZ);
+	if (!time_left) {
+		ath12k_warn(ab, "failed to receive control response completion, polling..\n");
+
+		for (i = 0; i < ab->hw_params->ce_count; i++)
+			ath12k_ce_per_engine_service(htc->ab, i);
+
+		time_left =
+			wait_for_completion_timeout(&htc->ctl_resp,
+						    ATH12K_HTC_WAIT_TIMEOUT_HZ);
+
+		if (!time_left)
+			status = -ETIMEDOUT;
+	}
+
+	if (status < 0) {
+		ath12k_warn(ab, "ctl_resp never came in (%d)\n", status);
+		return status;
+	}
+
+	if (htc->control_resp_len < sizeof(*ready)) {
+		ath12k_warn(ab, "Invalid HTC ready msg len:%d\n",
+			    htc->control_resp_len);
+		return -ECOMM;
+	}
+
+	ready = (struct ath12k_htc_ready *)htc->control_resp_buffer;
+	message_id = le32_get_bits(ready->id_credit_count, HTC_MSG_MESSAGEID);
+	credit_count = le32_get_bits(ready->id_credit_count,
+				     HTC_READY_MSG_CREDITCOUNT);
+	credit_size = le32_get_bits(ready->size_ep, HTC_READY_MSG_CREDITSIZE);
+
+	if (message_id != ATH12K_HTC_MSG_READY_ID) {
+		ath12k_warn(ab, "Invalid HTC ready msg: 0x%x\n", message_id);
+		return -ECOMM;
+	}
+
+	htc->total_transmit_credits = credit_count;
+	htc->target_credit_size = credit_size;
+
+	ath12k_dbg(ab, ATH12K_DBG_HTC,
+		   "Target ready! transmit resources: %d size:%d\n",
+		   htc->total_transmit_credits, htc->target_credit_size);
+
+	if ((htc->total_transmit_credits == 0) ||
+	    (htc->target_credit_size == 0)) {
+		ath12k_warn(ab, "Invalid credit size received\n");
+		return -ECOMM;
+	}
+
+	ath12k_htc_setup_target_buffer_assignments(htc);
+
+	return 0;
+}
+
+int ath12k_htc_connect_service(struct ath12k_htc *htc,
+			       struct ath12k_htc_svc_conn_req *conn_req,
+			       struct ath12k_htc_svc_conn_resp *conn_resp)
+{
+	struct ath12k_base *ab = htc->ab;
+	struct ath12k_htc_conn_svc *req_msg;
+	struct ath12k_htc_conn_svc_resp resp_msg_dummy;
+	struct ath12k_htc_conn_svc_resp *resp_msg = &resp_msg_dummy;
+	enum ath12k_htc_ep_id assigned_eid = ATH12K_HTC_EP_COUNT;
+	struct ath12k_htc_ep *ep;
+	struct sk_buff *skb;
+	unsigned int max_msg_size = 0;
+	int length, status;
+	unsigned long time_left;
+	bool disable_credit_flow_ctrl = false;
+	u16 message_id, service_id, flags = 0;
+	u8 tx_alloc = 0;
+
+	/* special case for HTC pseudo control service */
+	if (conn_req->service_id == ATH12K_HTC_SVC_ID_RSVD_CTRL) {
+		disable_credit_flow_ctrl = true;
+		assigned_eid = ATH12K_HTC_EP_0;
+		max_msg_size = ATH12K_HTC_MAX_CTRL_MSG_LEN;
+		memset(&resp_msg_dummy, 0, sizeof(resp_msg_dummy));
+		goto setup;
+	}
+
+	tx_alloc = ath12k_htc_get_credit_allocation(htc,
+						    conn_req->service_id);
+	if (!tx_alloc)
+		ath12k_dbg(ab, ATH12K_DBG_BOOT,
+			   "boot htc service %s does not allocate target credits\n",
+			   htc_service_name(conn_req->service_id));
+
+	skb = ath12k_htc_build_tx_ctrl_skb();
+	if (!skb) {
+		ath12k_warn(ab, "Failed to allocate HTC packet\n");
+		return -ENOMEM;
+	}
+
+	length = sizeof(*req_msg);
+	skb_put(skb, length);
+	memset(skb->data, 0, length);
+
+	req_msg = (struct ath12k_htc_conn_svc *)skb->data;
+	req_msg->msg_svc_id = le32_encode_bits(ATH12K_HTC_MSG_CONNECT_SERVICE_ID,
+					       HTC_MSG_MESSAGEID);
+
+	flags |= u32_encode_bits(tx_alloc, ATH12K_HTC_CONN_FLAGS_RECV_ALLOC);
+
+	/* Only enable credit flow control for WMI ctrl service */
+	if (!(conn_req->service_id == ATH12K_HTC_SVC_ID_WMI_CONTROL ||
+	      conn_req->service_id == ATH12K_HTC_SVC_ID_WMI_CONTROL_MAC1 ||
+	      conn_req->service_id == ATH12K_HTC_SVC_ID_WMI_CONTROL_MAC2)) {
+		flags |= ATH12K_HTC_CONN_FLAGS_DISABLE_CREDIT_FLOW_CTRL;
+		disable_credit_flow_ctrl = true;
+	}
+
+	req_msg->flags_len = le32_encode_bits(flags, HTC_SVC_MSG_CONNECTIONFLAGS);
+	req_msg->msg_svc_id |= le32_encode_bits(conn_req->service_id,
+						HTC_SVC_MSG_SERVICE_ID);
+
+	reinit_completion(&htc->ctl_resp);
+
+	status = ath12k_htc_send(htc, ATH12K_HTC_EP_0, skb);
+	if (status) {
+		kfree_skb(skb);
+		return status;
+	}
+
+	/* wait for response */
+	time_left = wait_for_completion_timeout(&htc->ctl_resp,
+						ATH12K_HTC_CONN_SVC_TIMEOUT_HZ);
+	if (!time_left) {
+		ath12k_err(ab, "Service connect timeout\n");
+		return -ETIMEDOUT;
+	}
+
+	/* we controlled the buffer creation, it's aligned */
+	resp_msg = (struct ath12k_htc_conn_svc_resp *)htc->control_resp_buffer;
+	message_id = le32_get_bits(resp_msg->msg_svc_id, HTC_MSG_MESSAGEID);
+	service_id = le32_get_bits(resp_msg->msg_svc_id,
+				   HTC_SVC_RESP_MSG_SERVICEID);
+
+	if ((message_id != ATH12K_HTC_MSG_CONNECT_SERVICE_RESP_ID) ||
+	    (htc->control_resp_len < sizeof(*resp_msg))) {
+		ath12k_err(ab, "Invalid resp message ID 0x%x", message_id);
+		return -EPROTO;
+	}
+
+	ath12k_dbg(ab, ATH12K_DBG_HTC,
+		   "HTC Service %s connect response: status: %u, assigned ep: %u\n",
+		   htc_service_name(service_id),
+		   le32_get_bits(resp_msg->flags_len, HTC_SVC_RESP_MSG_STATUS),
+		   le32_get_bits(resp_msg->flags_len, HTC_SVC_RESP_MSG_ENDPOINTID));
+
+	conn_resp->connect_resp_code = le32_get_bits(resp_msg->flags_len,
+						     HTC_SVC_RESP_MSG_STATUS);
+
+	/* check response status */
+	if (conn_resp->connect_resp_code != ATH12K_HTC_CONN_SVC_STATUS_SUCCESS) {
+		ath12k_err(ab, "HTC Service %s connect request failed: 0x%x)\n",
+			   htc_service_name(service_id),
+		       conn_resp->connect_resp_code);
+		return -EPROTO;
+	}
+
+	assigned_eid = le32_get_bits(resp_msg->flags_len,
+				     HTC_SVC_RESP_MSG_ENDPOINTID);
+
+	max_msg_size = le32_get_bits(resp_msg->flags_len,
+				     HTC_SVC_RESP_MSG_MAXMSGSIZE);
+
+setup:
+
+	if (assigned_eid >= ATH12K_HTC_EP_COUNT)
+		return -EPROTO;
+
+	if (max_msg_size == 0)
+		return -EPROTO;
+
+	ep = &htc->endpoint[assigned_eid];
+	ep->eid = assigned_eid;
+
+	if (ep->service_id != ATH12K_HTC_SVC_ID_UNUSED)
+		return -EPROTO;
+
+	/* return assigned endpoint to caller */
+	conn_resp->eid = assigned_eid;
+	conn_resp->max_msg_len = le32_get_bits(resp_msg->flags_len,
+					       HTC_SVC_RESP_MSG_MAXMSGSIZE);
+
+	/* setup the endpoint */
+	ep->service_id = conn_req->service_id;
+	ep->max_tx_queue_depth = conn_req->max_send_queue_depth;
+	ep->max_ep_message_len = le32_get_bits(resp_msg->flags_len,
+					       HTC_SVC_RESP_MSG_MAXMSGSIZE);
+	ep->tx_credits = tx_alloc;
+
+	/* copy all the callbacks */
+	ep->ep_ops = conn_req->ep_ops;
+
+	status = ath12k_hif_map_service_to_pipe(htc->ab,
+						ep->service_id,
+						&ep->ul_pipe_id,
+						&ep->dl_pipe_id);
+	if (status)
+		return status;
+
+	ath12k_dbg(ab, ATH12K_DBG_BOOT,
+		   "boot htc service '%s' ul pipe %d dl pipe %d eid %d ready\n",
+		   htc_service_name(ep->service_id), ep->ul_pipe_id,
+		   ep->dl_pipe_id, ep->eid);
+
+	if (disable_credit_flow_ctrl && ep->tx_credit_flow_enabled) {
+		ep->tx_credit_flow_enabled = false;
+		ath12k_dbg(ab, ATH12K_DBG_BOOT,
+			   "boot htc service '%s' eid %d TX flow control disabled\n",
+			   htc_service_name(ep->service_id), assigned_eid);
+	}
+
+	return status;
+}
+
+int ath12k_htc_start(struct ath12k_htc *htc)
+{
+	struct sk_buff *skb;
+	int status;
+	struct ath12k_base *ab = htc->ab;
+	struct ath12k_htc_setup_complete_extended *msg;
+
+	skb = ath12k_htc_build_tx_ctrl_skb();
+	if (!skb)
+		return -ENOMEM;
+
+	skb_put(skb, sizeof(*msg));
+	memset(skb->data, 0, skb->len);
+
+	msg = (struct ath12k_htc_setup_complete_extended *)skb->data;
+	msg->msg_id = le32_encode_bits(ATH12K_HTC_MSG_SETUP_COMPLETE_EX_ID,
+				       HTC_MSG_MESSAGEID);
+
+	ath12k_dbg(ab, ATH12K_DBG_HTC, "HTC is using TX credit flow control\n");
+
+	status = ath12k_htc_send(htc, ATH12K_HTC_EP_0, skb);
+	if (status) {
+		kfree_skb(skb);
+		return status;
+	}
+
+	return 0;
+}
+
+int ath12k_htc_init(struct ath12k_base *ab)
+{
+	struct ath12k_htc *htc = &ab->htc;
+	struct ath12k_htc_svc_conn_req conn_req = { };
+	struct ath12k_htc_svc_conn_resp conn_resp = { };
+	int ret;
+
+	spin_lock_init(&htc->tx_lock);
+
+	ath12k_htc_reset_endpoint_states(htc);
+
+	htc->ab = ab;
+
+	switch (ab->wmi_ab.preferred_hw_mode) {
+	case WMI_HOST_HW_MODE_SINGLE:
+		htc->wmi_ep_count = 1;
+		break;
+	case WMI_HOST_HW_MODE_DBS:
+	case WMI_HOST_HW_MODE_DBS_OR_SBS:
+		htc->wmi_ep_count = 2;
+		break;
+	case WMI_HOST_HW_MODE_DBS_SBS:
+		htc->wmi_ep_count = 3;
+		break;
+	default:
+		htc->wmi_ep_count = ab->hw_params->max_radios;
+		break;
+	}
+
+	/* setup our pseudo HTC control endpoint connection */
+	conn_req.ep_ops.ep_tx_complete = ath12k_htc_control_tx_complete;
+	conn_req.ep_ops.ep_rx_complete = ath12k_htc_control_rx_complete;
+	conn_req.max_send_queue_depth = ATH12K_NUM_CONTROL_TX_BUFFERS;
+	conn_req.service_id = ATH12K_HTC_SVC_ID_RSVD_CTRL;
+
+	/* connect fake service */
+	ret = ath12k_htc_connect_service(htc, &conn_req, &conn_resp);
+	if (ret) {
+		ath12k_err(ab, "could not connect to htc service (%d)\n", ret);
+		return ret;
+	}
+
+	init_completion(&htc->ctl_resp);
+
+	return 0;
+}