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, 601 insertions, 0 deletions

diff --git a/drivers/net/ethernet/fungible/funcore/fun_queue.c b/drivers/net/ethernet/fungible/funcore/fun_queue.c
new file mode 100644
index 000000000..8ab9f6843
--- /dev/null
+++ b/drivers/net/ethernet/fungible/funcore/fun_queue.c
@@ -0,0 +1,601 @@
+// SPDX-License-Identifier: (GPL-2.0-only OR BSD-3-Clause)
+
+#include <linux/dma-mapping.h>
+#include <linux/interrupt.h>
+#include <linux/log2.h>
+#include <linux/mm.h>
+#include <linux/netdevice.h>
+#include <linux/pci.h>
+#include <linux/slab.h>
+
+#include "fun_dev.h"
+#include "fun_queue.h"
+
+/* Allocate memory for a queue. This includes the memory for the HW descriptor
+ * ring, an optional 64b HW write-back area, and an optional SW state ring.
+ * Returns the virtual and DMA addresses of the HW ring, the VA of the SW ring,
+ * and the VA of the write-back area.
+ */
+void *fun_alloc_ring_mem(struct device *dma_dev, size_t depth,
+			 size_t hw_desc_sz, size_t sw_desc_sz, bool wb,
+			 int numa_node, dma_addr_t *dma_addr, void **sw_va,
+			 volatile __be64 **wb_va)
+{
+	int dev_node = dev_to_node(dma_dev);
+	size_t dma_sz;
+	void *va;
+
+	if (numa_node == NUMA_NO_NODE)
+		numa_node = dev_node;
+
+	/* Place optional write-back area at end of descriptor ring. */
+	dma_sz = hw_desc_sz * depth;
+	if (wb)
+		dma_sz += sizeof(u64);
+
+	set_dev_node(dma_dev, numa_node);
+	va = dma_alloc_coherent(dma_dev, dma_sz, dma_addr, GFP_KERNEL);
+	set_dev_node(dma_dev, dev_node);
+	if (!va)
+		return NULL;
+
+	if (sw_desc_sz) {
+		*sw_va = kvzalloc_node(sw_desc_sz * depth, GFP_KERNEL,
+				       numa_node);
+		if (!*sw_va) {
+			dma_free_coherent(dma_dev, dma_sz, va, *dma_addr);
+			return NULL;
+		}
+	}
+
+	if (wb)
+		*wb_va = va + dma_sz - sizeof(u64);
+	return va;
+}
+EXPORT_SYMBOL_GPL(fun_alloc_ring_mem);
+
+void fun_free_ring_mem(struct device *dma_dev, size_t depth, size_t hw_desc_sz,
+		       bool wb, void *hw_va, dma_addr_t dma_addr, void *sw_va)
+{
+	if (hw_va) {
+		size_t sz = depth * hw_desc_sz;
+
+		if (wb)
+			sz += sizeof(u64);
+		dma_free_coherent(dma_dev, sz, hw_va, dma_addr);
+	}
+	kvfree(sw_va);
+}
+EXPORT_SYMBOL_GPL(fun_free_ring_mem);
+
+/* Prepare and issue an admin command to create an SQ on the device with the
+ * provided parameters. If the queue ID is auto-allocated by the device it is
+ * returned in *sqidp.
+ */
+int fun_sq_create(struct fun_dev *fdev, u16 flags, u32 sqid, u32 cqid,
+		  u8 sqe_size_log2, u32 sq_depth, dma_addr_t dma_addr,
+		  u8 coal_nentries, u8 coal_usec, u32 irq_num,
+		  u32 scan_start_id, u32 scan_end_id,
+		  u32 rq_buf_size_log2, u32 *sqidp, u32 __iomem **dbp)
+{
+	union {
+		struct fun_admin_epsq_req req;
+		struct fun_admin_generic_create_rsp rsp;
+	} cmd;
+	dma_addr_t wb_addr;
+	u32 hw_qid;
+	int rc;
+
+	if (sq_depth > fdev->q_depth)
+		return -EINVAL;
+	if (flags & FUN_ADMIN_EPSQ_CREATE_FLAG_RQ)
+		sqe_size_log2 = ilog2(sizeof(struct fun_eprq_rqbuf));
+
+	wb_addr = dma_addr + (sq_depth << sqe_size_log2);
+
+	cmd.req.common = FUN_ADMIN_REQ_COMMON_INIT2(FUN_ADMIN_OP_EPSQ,
+						    sizeof(cmd.req));
+	cmd.req.u.create =
+		FUN_ADMIN_EPSQ_CREATE_REQ_INIT(FUN_ADMIN_SUBOP_CREATE, flags,
+					       sqid, cqid, sqe_size_log2,
+					       sq_depth - 1, dma_addr, 0,
+					       coal_nentries, coal_usec,
+					       irq_num, scan_start_id,
+					       scan_end_id, 0,
+					       rq_buf_size_log2,
+					       ilog2(sizeof(u64)), wb_addr);
+
+	rc = fun_submit_admin_sync_cmd(fdev, &cmd.req.common,
+				       &cmd.rsp, sizeof(cmd.rsp), 0);
+	if (rc)
+		return rc;
+
+	hw_qid = be32_to_cpu(cmd.rsp.id);
+	*dbp = fun_sq_db_addr(fdev, hw_qid);
+	if (flags & FUN_ADMIN_RES_CREATE_FLAG_ALLOCATOR)
+		*sqidp = hw_qid;
+	return rc;
+}
+EXPORT_SYMBOL_GPL(fun_sq_create);
+
+/* Prepare and issue an admin command to create a CQ on the device with the
+ * provided parameters. If the queue ID is auto-allocated by the device it is
+ * returned in *cqidp.
+ */
+int fun_cq_create(struct fun_dev *fdev, u16 flags, u32 cqid, u32 rqid,
+		  u8 cqe_size_log2, u32 cq_depth, dma_addr_t dma_addr,
+		  u16 headroom, u16 tailroom, u8 coal_nentries, u8 coal_usec,
+		  u32 irq_num, u32 scan_start_id, u32 scan_end_id, u32 *cqidp,
+		  u32 __iomem **dbp)
+{
+	union {
+		struct fun_admin_epcq_req req;
+		struct fun_admin_generic_create_rsp rsp;
+	} cmd;
+	u32 hw_qid;
+	int rc;
+
+	if (cq_depth > fdev->q_depth)
+		return -EINVAL;
+
+	cmd.req.common = FUN_ADMIN_REQ_COMMON_INIT2(FUN_ADMIN_OP_EPCQ,
+						    sizeof(cmd.req));
+	cmd.req.u.create =
+		FUN_ADMIN_EPCQ_CREATE_REQ_INIT(FUN_ADMIN_SUBOP_CREATE, flags,
+					       cqid, rqid, cqe_size_log2,
+					       cq_depth - 1, dma_addr, tailroom,
+					       headroom / 2, 0, coal_nentries,
+					       coal_usec, irq_num,
+					       scan_start_id, scan_end_id, 0);
+
+	rc = fun_submit_admin_sync_cmd(fdev, &cmd.req.common,
+				       &cmd.rsp, sizeof(cmd.rsp), 0);
+	if (rc)
+		return rc;
+
+	hw_qid = be32_to_cpu(cmd.rsp.id);
+	*dbp = fun_cq_db_addr(fdev, hw_qid);
+	if (flags & FUN_ADMIN_RES_CREATE_FLAG_ALLOCATOR)
+		*cqidp = hw_qid;
+	return rc;
+}
+EXPORT_SYMBOL_GPL(fun_cq_create);
+
+static bool fun_sq_is_head_wb(const struct fun_queue *funq)
+{
+	return funq->sq_flags & FUN_ADMIN_EPSQ_CREATE_FLAG_HEAD_WB_ADDRESS;
+}
+
+static void fun_clean_rq(struct fun_queue *funq)
+{
+	struct fun_dev *fdev = funq->fdev;
+	struct fun_rq_info *rqinfo;
+	unsigned int i;
+
+	for (i = 0; i < funq->rq_depth; i++) {
+		rqinfo = &funq->rq_info[i];
+		if (rqinfo->page) {
+			dma_unmap_page(fdev->dev, rqinfo->dma, PAGE_SIZE,
+				       DMA_FROM_DEVICE);
+			put_page(rqinfo->page);
+			rqinfo->page = NULL;
+		}
+	}
+}
+
+static int fun_fill_rq(struct fun_queue *funq)
+{
+	struct device *dev = funq->fdev->dev;
+	int i, node = dev_to_node(dev);
+	struct fun_rq_info *rqinfo;
+
+	for (i = 0; i < funq->rq_depth; i++) {
+		rqinfo = &funq->rq_info[i];
+		rqinfo->page = alloc_pages_node(node, GFP_KERNEL, 0);
+		if (unlikely(!rqinfo->page))
+			return -ENOMEM;
+
+		rqinfo->dma = dma_map_page(dev, rqinfo->page, 0,
+					   PAGE_SIZE, DMA_FROM_DEVICE);
+		if (unlikely(dma_mapping_error(dev, rqinfo->dma))) {
+			put_page(rqinfo->page);
+			rqinfo->page = NULL;
+			return -ENOMEM;
+		}
+
+		funq->rqes[i] = FUN_EPRQ_RQBUF_INIT(rqinfo->dma);
+	}
+
+	funq->rq_tail = funq->rq_depth - 1;
+	return 0;
+}
+
+static void fun_rq_update_pos(struct fun_queue *funq, int buf_offset)
+{
+	if (buf_offset <= funq->rq_buf_offset) {
+		struct fun_rq_info *rqinfo = &funq->rq_info[funq->rq_buf_idx];
+		struct device *dev = funq->fdev->dev;
+
+		dma_sync_single_for_device(dev, rqinfo->dma, PAGE_SIZE,
+					   DMA_FROM_DEVICE);
+		funq->num_rqe_to_fill++;
+		if (++funq->rq_buf_idx == funq->rq_depth)
+			funq->rq_buf_idx = 0;
+	}
+	funq->rq_buf_offset = buf_offset;
+}
+
+/* Given a command response with data scattered across >= 1 RQ buffers return
+ * a pointer to a contiguous buffer containing all the data. If the data is in
+ * one RQ buffer the start address within that buffer is returned, otherwise a
+ * new buffer is allocated and the data is gathered into it.
+ */
+static void *fun_data_from_rq(struct fun_queue *funq,
+			      const struct fun_rsp_common *rsp, bool *need_free)
+{
+	u32 bufoff, total_len, remaining, fragsize, dataoff;
+	struct device *dma_dev = funq->fdev->dev;
+	const struct fun_dataop_rqbuf *databuf;
+	const struct fun_dataop_hdr *dataop;
+	const struct fun_rq_info *rqinfo;
+	void *data;
+
+	dataop = (void *)rsp + rsp->suboff8 * 8;
+	total_len = be32_to_cpu(dataop->total_len);
+
+	if (likely(dataop->nsgl == 1)) {
+		databuf = (struct fun_dataop_rqbuf *)dataop->imm;
+		bufoff = be32_to_cpu(databuf->bufoff);
+		fun_rq_update_pos(funq, bufoff);
+		rqinfo = &funq->rq_info[funq->rq_buf_idx];
+		dma_sync_single_for_cpu(dma_dev, rqinfo->dma + bufoff,
+					total_len, DMA_FROM_DEVICE);
+		*need_free = false;
+		return page_address(rqinfo->page) + bufoff;
+	}
+
+	/* For scattered completions gather the fragments into one buffer. */
+
+	data = kmalloc(total_len, GFP_ATOMIC);
+	/* NULL is OK here. In case of failure we still need to consume the data
+	 * for proper buffer accounting but indicate an error in the response.
+	 */
+	if (likely(data))
+		*need_free = true;
+
+	dataoff = 0;
+	for (remaining = total_len; remaining; remaining -= fragsize) {
+		fun_rq_update_pos(funq, 0);
+		fragsize = min_t(unsigned int, PAGE_SIZE, remaining);
+		if (data) {
+			rqinfo = &funq->rq_info[funq->rq_buf_idx];
+			dma_sync_single_for_cpu(dma_dev, rqinfo->dma, fragsize,
+						DMA_FROM_DEVICE);
+			memcpy(data + dataoff, page_address(rqinfo->page),
+			       fragsize);
+			dataoff += fragsize;
+		}
+	}
+	return data;
+}
+
+unsigned int __fun_process_cq(struct fun_queue *funq, unsigned int max)
+{
+	const struct fun_cqe_info *info;
+	struct fun_rsp_common *rsp;
+	unsigned int new_cqes;
+	u16 sf_p, flags;
+	bool need_free;
+	void *cqe;
+
+	if (!max)
+		max = funq->cq_depth - 1;
+
+	for (new_cqes = 0; new_cqes < max; new_cqes++) {
+		cqe = funq->cqes + (funq->cq_head << funq->cqe_size_log2);
+		info = funq_cqe_info(funq, cqe);
+		sf_p = be16_to_cpu(info->sf_p);
+
+		if ((sf_p & 1) != funq->cq_phase)
+			break;
+
+		/* ensure the phase tag is read before other CQE fields */
+		dma_rmb();
+
+		if (++funq->cq_head == funq->cq_depth) {
+			funq->cq_head = 0;
+			funq->cq_phase = !funq->cq_phase;
+		}
+
+		rsp = cqe;
+		flags = be16_to_cpu(rsp->flags);
+
+		need_free = false;
+		if (unlikely(flags & FUN_REQ_COMMON_FLAG_CQE_IN_RQBUF)) {
+			rsp = fun_data_from_rq(funq, rsp, &need_free);
+			if (!rsp) {
+				rsp = cqe;
+				rsp->len8 = 1;
+				if (rsp->ret == 0)
+					rsp->ret = ENOMEM;
+			}
+		}
+
+		if (funq->cq_cb)
+			funq->cq_cb(funq, funq->cb_data, rsp, info);
+		if (need_free)
+			kfree(rsp);
+	}
+
+	dev_dbg(funq->fdev->dev, "CQ %u, new CQEs %u/%u, head %u, phase %u\n",
+		funq->cqid, new_cqes, max, funq->cq_head, funq->cq_phase);
+	return new_cqes;
+}
+
+unsigned int fun_process_cq(struct fun_queue *funq, unsigned int max)
+{
+	unsigned int processed;
+	u32 db;
+
+	processed = __fun_process_cq(funq, max);
+
+	if (funq->num_rqe_to_fill) {
+		funq->rq_tail = (funq->rq_tail + funq->num_rqe_to_fill) %
+				funq->rq_depth;
+		funq->num_rqe_to_fill = 0;
+		writel(funq->rq_tail, funq->rq_db);
+	}
+
+	db = funq->cq_head | FUN_DB_IRQ_ARM_F;
+	writel(db, funq->cq_db);
+	return processed;
+}
+
+static int fun_alloc_sqes(struct fun_queue *funq)
+{
+	funq->sq_cmds = fun_alloc_ring_mem(funq->fdev->dev, funq->sq_depth,
+					   1 << funq->sqe_size_log2, 0,
+					   fun_sq_is_head_wb(funq),
+					   NUMA_NO_NODE, &funq->sq_dma_addr,
+					   NULL, &funq->sq_head);
+	return funq->sq_cmds ? 0 : -ENOMEM;
+}
+
+static int fun_alloc_cqes(struct fun_queue *funq)
+{
+	funq->cqes = fun_alloc_ring_mem(funq->fdev->dev, funq->cq_depth,
+					1 << funq->cqe_size_log2, 0, false,
+					NUMA_NO_NODE, &funq->cq_dma_addr, NULL,
+					NULL);
+	return funq->cqes ? 0 : -ENOMEM;
+}
+
+static int fun_alloc_rqes(struct fun_queue *funq)
+{
+	funq->rqes = fun_alloc_ring_mem(funq->fdev->dev, funq->rq_depth,
+					sizeof(*funq->rqes),
+					sizeof(*funq->rq_info), false,
+					NUMA_NO_NODE, &funq->rq_dma_addr,
+					(void **)&funq->rq_info, NULL);
+	return funq->rqes ? 0 : -ENOMEM;
+}
+
+/* Free a queue's structures. */
+void fun_free_queue(struct fun_queue *funq)
+{
+	struct device *dev = funq->fdev->dev;
+
+	fun_free_ring_mem(dev, funq->cq_depth, 1 << funq->cqe_size_log2, false,
+			  funq->cqes, funq->cq_dma_addr, NULL);
+	fun_free_ring_mem(dev, funq->sq_depth, 1 << funq->sqe_size_log2,
+			  fun_sq_is_head_wb(funq), funq->sq_cmds,
+			  funq->sq_dma_addr, NULL);
+
+	if (funq->rqes) {
+		fun_clean_rq(funq);
+		fun_free_ring_mem(dev, funq->rq_depth, sizeof(*funq->rqes),
+				  false, funq->rqes, funq->rq_dma_addr,
+				  funq->rq_info);
+	}
+
+	kfree(funq);
+}
+
+/* Allocate and initialize a funq's structures. */
+struct fun_queue *fun_alloc_queue(struct fun_dev *fdev, int qid,
+				  const struct fun_queue_alloc_req *req)
+{
+	struct fun_queue *funq = kzalloc(sizeof(*funq), GFP_KERNEL);
+
+	if (!funq)
+		return NULL;
+
+	funq->fdev = fdev;
+	spin_lock_init(&funq->sq_lock);
+
+	funq->qid = qid;
+
+	/* Initial CQ/SQ/RQ ids */
+	if (req->rq_depth) {
+		funq->cqid = 2 * qid;
+		if (funq->qid) {
+			/* I/O Q: use rqid = cqid, sqid = +1 */
+			funq->rqid = funq->cqid;
+			funq->sqid = funq->rqid + 1;
+		} else {
+			/* Admin Q: sqid is always 0, use ID 1 for RQ */
+			funq->sqid = 0;
+			funq->rqid = 1;
+		}
+	} else {
+		funq->cqid = qid;
+		funq->sqid = qid;
+	}
+
+	funq->cq_flags = req->cq_flags;
+	funq->sq_flags = req->sq_flags;
+
+	funq->cqe_size_log2 = req->cqe_size_log2;
+	funq->sqe_size_log2 = req->sqe_size_log2;
+
+	funq->cq_depth = req->cq_depth;
+	funq->sq_depth = req->sq_depth;
+
+	funq->cq_intcoal_nentries = req->cq_intcoal_nentries;
+	funq->cq_intcoal_usec = req->cq_intcoal_usec;
+
+	funq->sq_intcoal_nentries = req->sq_intcoal_nentries;
+	funq->sq_intcoal_usec = req->sq_intcoal_usec;
+
+	if (fun_alloc_cqes(funq))
+		goto free_funq;
+
+	funq->cq_phase = 1;
+
+	if (fun_alloc_sqes(funq))
+		goto free_funq;
+
+	if (req->rq_depth) {
+		funq->rq_flags = req->rq_flags | FUN_ADMIN_EPSQ_CREATE_FLAG_RQ;
+		funq->rq_depth = req->rq_depth;
+		funq->rq_buf_offset = -1;
+
+		if (fun_alloc_rqes(funq) || fun_fill_rq(funq))
+			goto free_funq;
+	}
+
+	funq->cq_vector = -1;
+	funq->cqe_info_offset = (1 << funq->cqe_size_log2) - sizeof(struct fun_cqe_info);
+
+	/* SQ/CQ 0 are implicitly created, assign their doorbells now.
+	 * Other queues are assigned doorbells at their explicit creation.
+	 */
+	if (funq->sqid == 0)
+		funq->sq_db = fun_sq_db_addr(fdev, 0);
+	if (funq->cqid == 0)
+		funq->cq_db = fun_cq_db_addr(fdev, 0);
+
+	return funq;
+
+free_funq:
+	fun_free_queue(funq);
+	return NULL;
+}
+
+/* Create a funq's CQ on the device. */
+static int fun_create_cq(struct fun_queue *funq)
+{
+	struct fun_dev *fdev = funq->fdev;
+	unsigned int rqid;
+	int rc;
+
+	rqid = funq->cq_flags & FUN_ADMIN_EPCQ_CREATE_FLAG_RQ ?
+		funq->rqid : FUN_HCI_ID_INVALID;
+	rc = fun_cq_create(fdev, funq->cq_flags, funq->cqid, rqid,
+			   funq->cqe_size_log2, funq->cq_depth,
+			   funq->cq_dma_addr, 0, 0, funq->cq_intcoal_nentries,
+			   funq->cq_intcoal_usec, funq->cq_vector, 0, 0,
+			   &funq->cqid, &funq->cq_db);
+	if (!rc)
+		dev_dbg(fdev->dev, "created CQ %u\n", funq->cqid);
+
+	return rc;
+}
+
+/* Create a funq's SQ on the device. */
+static int fun_create_sq(struct fun_queue *funq)
+{
+	struct fun_dev *fdev = funq->fdev;
+	int rc;
+
+	rc = fun_sq_create(fdev, funq->sq_flags, funq->sqid, funq->cqid,
+			   funq->sqe_size_log2, funq->sq_depth,
+			   funq->sq_dma_addr, funq->sq_intcoal_nentries,
+			   funq->sq_intcoal_usec, funq->cq_vector, 0, 0,
+			   0, &funq->sqid, &funq->sq_db);
+	if (!rc)
+		dev_dbg(fdev->dev, "created SQ %u\n", funq->sqid);
+
+	return rc;
+}
+
+/* Create a funq's RQ on the device. */
+int fun_create_rq(struct fun_queue *funq)
+{
+	struct fun_dev *fdev = funq->fdev;
+	int rc;
+
+	rc = fun_sq_create(fdev, funq->rq_flags, funq->rqid, funq->cqid, 0,
+			   funq->rq_depth, funq->rq_dma_addr, 0, 0,
+			   funq->cq_vector, 0, 0, PAGE_SHIFT, &funq->rqid,
+			   &funq->rq_db);
+	if (!rc)
+		dev_dbg(fdev->dev, "created RQ %u\n", funq->rqid);
+
+	return rc;
+}
+
+static unsigned int funq_irq(struct fun_queue *funq)
+{
+	return pci_irq_vector(to_pci_dev(funq->fdev->dev), funq->cq_vector);
+}
+
+int fun_request_irq(struct fun_queue *funq, const char *devname,
+		    irq_handler_t handler, void *data)
+{
+	int rc;
+
+	if (funq->cq_vector < 0)
+		return -EINVAL;
+
+	funq->irq_handler = handler;
+	funq->irq_data = data;
+
+	snprintf(funq->irqname, sizeof(funq->irqname),
+		 funq->qid ? "%s-q[%d]" : "%s-adminq", devname, funq->qid);
+
+	rc = request_irq(funq_irq(funq), handler, 0, funq->irqname, data);
+	if (rc)
+		funq->irq_handler = NULL;
+
+	return rc;
+}
+
+/* Create all component queues of a funq  on the device. */
+int fun_create_queue(struct fun_queue *funq)
+{
+	int rc;
+
+	rc = fun_create_cq(funq);
+	if (rc)
+		return rc;
+
+	if (funq->rq_depth) {
+		rc = fun_create_rq(funq);
+		if (rc)
+			goto release_cq;
+	}
+
+	rc = fun_create_sq(funq);
+	if (rc)
+		goto release_rq;
+
+	return 0;
+
+release_rq:
+	fun_destroy_sq(funq->fdev, funq->rqid);
+release_cq:
+	fun_destroy_cq(funq->fdev, funq->cqid);
+	return rc;
+}
+
+void fun_free_irq(struct fun_queue *funq)
+{
+	if (funq->irq_handler) {
+		unsigned int vector = funq_irq(funq);
+
+		free_irq(vector, funq->irq_data);
+		funq->irq_handler = NULL;
+		funq->irq_data = NULL;
+	}
+}