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, 855 insertions, 0 deletions

diff --git a/drivers/usb/gadget/udc/aspeed-vhub/epn.c b/drivers/usb/gadget/udc/aspeed-vhub/epn.c
new file mode 100644
index 000000000..56e55472d
--- /dev/null
+++ b/drivers/usb/gadget/udc/aspeed-vhub/epn.c
@@ -0,0 +1,855 @@
+// SPDX-License-Identifier: GPL-2.0+
+/*
+ * aspeed-vhub -- Driver for Aspeed SoC "vHub" USB gadget
+ *
+ * epn.c - Generic endpoints management
+ *
+ * Copyright 2017 IBM Corporation
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/delay.h>
+#include <linux/ioport.h>
+#include <linux/slab.h>
+#include <linux/errno.h>
+#include <linux/list.h>
+#include <linux/interrupt.h>
+#include <linux/proc_fs.h>
+#include <linux/prefetch.h>
+#include <linux/clk.h>
+#include <linux/usb/gadget.h>
+#include <linux/of.h>
+#include <linux/of_gpio.h>
+#include <linux/regmap.h>
+#include <linux/dma-mapping.h>
+
+#include "vhub.h"
+
+#define EXTRA_CHECKS
+
+#ifdef EXTRA_CHECKS
+#define CHECK(ep, expr, fmt...)					\
+	do {							\
+		if (!(expr)) EPDBG(ep, "CHECK:" fmt);		\
+	} while(0)
+#else
+#define CHECK(ep, expr, fmt...)	do { } while(0)
+#endif
+
+static void ast_vhub_epn_kick(struct ast_vhub_ep *ep, struct ast_vhub_req *req)
+{
+	unsigned int act = req->req.actual;
+	unsigned int len = req->req.length;
+	unsigned int chunk;
+
+	/* There should be no DMA ongoing */
+	WARN_ON(req->active);
+
+	/* Calculate next chunk size */
+	chunk = len - act;
+	if (chunk > ep->ep.maxpacket)
+		chunk = ep->ep.maxpacket;
+	else if ((chunk < ep->ep.maxpacket) || !req->req.zero)
+		req->last_desc = 1;
+
+	EPVDBG(ep, "kick req %p act=%d/%d chunk=%d last=%d\n",
+	       req, act, len, chunk, req->last_desc);
+
+	/* If DMA unavailable, using staging EP buffer */
+	if (!req->req.dma) {
+
+		/* For IN transfers, copy data over first */
+		if (ep->epn.is_in) {
+			memcpy(ep->buf, req->req.buf + act, chunk);
+			vhub_dma_workaround(ep->buf);
+		}
+		writel(ep->buf_dma, ep->epn.regs + AST_VHUB_EP_DESC_BASE);
+	} else {
+		if (ep->epn.is_in)
+			vhub_dma_workaround(req->req.buf);
+		writel(req->req.dma + act, ep->epn.regs + AST_VHUB_EP_DESC_BASE);
+	}
+
+	/* Start DMA */
+	req->active = true;
+	writel(VHUB_EP_DMA_SET_TX_SIZE(chunk),
+	       ep->epn.regs + AST_VHUB_EP_DESC_STATUS);
+	writel(VHUB_EP_DMA_SET_TX_SIZE(chunk) | VHUB_EP_DMA_SINGLE_KICK,
+	       ep->epn.regs + AST_VHUB_EP_DESC_STATUS);
+}
+
+static void ast_vhub_epn_handle_ack(struct ast_vhub_ep *ep)
+{
+	struct ast_vhub_req *req;
+	unsigned int len;
+	int status = 0;
+	u32 stat;
+
+	/* Read EP status */
+	stat = readl(ep->epn.regs + AST_VHUB_EP_DESC_STATUS);
+
+	/* Grab current request if any */
+	req = list_first_entry_or_null(&ep->queue, struct ast_vhub_req, queue);
+
+	EPVDBG(ep, "ACK status=%08x is_in=%d, req=%p (active=%d)\n",
+	       stat, ep->epn.is_in, req, req ? req->active : 0);
+
+	/* In absence of a request, bail out, must have been dequeued */
+	if (!req)
+		return;
+
+	/*
+	 * Request not active, move on to processing queue, active request
+	 * was probably dequeued
+	 */
+	if (!req->active)
+		goto next_chunk;
+
+	/* Check if HW has moved on */
+	if (VHUB_EP_DMA_RPTR(stat) != 0) {
+		EPDBG(ep, "DMA read pointer not 0 !\n");
+		return;
+	}
+
+	/* No current DMA ongoing */
+	req->active = false;
+
+	/* Grab length out of HW */
+	len = VHUB_EP_DMA_TX_SIZE(stat);
+
+	/* If not using DMA, copy data out if needed */
+	if (!req->req.dma && !ep->epn.is_in && len) {
+		if (req->req.actual + len > req->req.length) {
+			req->last_desc = 1;
+			status = -EOVERFLOW;
+			goto done;
+		} else {
+			memcpy(req->req.buf + req->req.actual, ep->buf, len);
+		}
+	}
+	/* Adjust size */
+	req->req.actual += len;
+
+	/* Check for short packet */
+	if (len < ep->ep.maxpacket)
+		req->last_desc = 1;
+
+done:
+	/* That's it ? complete the request and pick a new one */
+	if (req->last_desc >= 0) {
+		ast_vhub_done(ep, req, status);
+		req = list_first_entry_or_null(&ep->queue, struct ast_vhub_req,
+					       queue);
+
+		/*
+		 * Due to lock dropping inside "done" the next request could
+		 * already be active, so check for that and bail if needed.
+		 */
+		if (!req || req->active)
+			return;
+	}
+
+ next_chunk:
+	ast_vhub_epn_kick(ep, req);
+}
+
+static inline unsigned int ast_vhub_count_free_descs(struct ast_vhub_ep *ep)
+{
+	/*
+	 * d_next == d_last means descriptor list empty to HW,
+	 * thus we can only have AST_VHUB_DESCS_COUNT-1 descriptors
+	 * in the list
+	 */
+	return (ep->epn.d_last + AST_VHUB_DESCS_COUNT - ep->epn.d_next - 1) &
+		(AST_VHUB_DESCS_COUNT - 1);
+}
+
+static void ast_vhub_epn_kick_desc(struct ast_vhub_ep *ep,
+				   struct ast_vhub_req *req)
+{
+	struct ast_vhub_desc *desc = NULL;
+	unsigned int act = req->act_count;
+	unsigned int len = req->req.length;
+	unsigned int chunk;
+
+	/* Mark request active if not already */
+	req->active = true;
+
+	/* If the request was already completely written, do nothing */
+	if (req->last_desc >= 0)
+		return;
+
+	EPVDBG(ep, "kick act=%d/%d chunk_max=%d free_descs=%d\n",
+	       act, len, ep->epn.chunk_max, ast_vhub_count_free_descs(ep));
+
+	/* While we can create descriptors */
+	while (ast_vhub_count_free_descs(ep) && req->last_desc < 0) {
+		unsigned int d_num;
+
+		/* Grab next free descriptor */
+		d_num = ep->epn.d_next;
+		desc = &ep->epn.descs[d_num];
+		ep->epn.d_next = (d_num + 1) & (AST_VHUB_DESCS_COUNT - 1);
+
+		/* Calculate next chunk size */
+		chunk = len - act;
+		if (chunk <= ep->epn.chunk_max) {
+			/*
+			 * Is this the last packet ? Because of having up to 8
+			 * packets in a descriptor we can't just compare "chunk"
+			 * with ep.maxpacket. We have to see if it's a multiple
+			 * of it to know if we have to send a zero packet.
+			 * Sadly that involves a modulo which is a bit expensive
+			 * but probably still better than not doing it.
+			 */
+			if (!chunk || !req->req.zero || (chunk % ep->ep.maxpacket) != 0)
+				req->last_desc = d_num;
+		} else {
+			chunk = ep->epn.chunk_max;
+		}
+
+		EPVDBG(ep, " chunk: act=%d/%d chunk=%d last=%d desc=%d free=%d\n",
+		       act, len, chunk, req->last_desc, d_num,
+		       ast_vhub_count_free_descs(ep));
+
+		/* Populate descriptor */
+		desc->w0 = cpu_to_le32(req->req.dma + act);
+
+		/* Interrupt if end of request or no more descriptors */
+
+		/*
+		 * TODO: Be smarter about it, if we don't have enough
+		 * descriptors request an interrupt before queue empty
+		 * or so in order to be able to populate more before
+		 * the HW runs out. This isn't a problem at the moment
+		 * as we use 256 descriptors and only put at most one
+		 * request in the ring.
+		 */
+		desc->w1 = cpu_to_le32(VHUB_DSC1_IN_SET_LEN(chunk));
+		if (req->last_desc >= 0 || !ast_vhub_count_free_descs(ep))
+			desc->w1 |= cpu_to_le32(VHUB_DSC1_IN_INTERRUPT);
+
+		/* Account packet */
+		req->act_count = act = act + chunk;
+	}
+
+	if (likely(desc))
+		vhub_dma_workaround(desc);
+
+	/* Tell HW about new descriptors */
+	writel(VHUB_EP_DMA_SET_CPU_WPTR(ep->epn.d_next),
+	       ep->epn.regs + AST_VHUB_EP_DESC_STATUS);
+
+	EPVDBG(ep, "HW kicked, d_next=%d dstat=%08x\n",
+	       ep->epn.d_next, readl(ep->epn.regs + AST_VHUB_EP_DESC_STATUS));
+}
+
+static void ast_vhub_epn_handle_ack_desc(struct ast_vhub_ep *ep)
+{
+	struct ast_vhub_req *req;
+	unsigned int len, d_last;
+	u32 stat, stat1;
+
+	/* Read EP status, workaround HW race */
+	do {
+		stat = readl(ep->epn.regs + AST_VHUB_EP_DESC_STATUS);
+		stat1 = readl(ep->epn.regs + AST_VHUB_EP_DESC_STATUS);
+	} while(stat != stat1);
+
+	/* Extract RPTR */
+	d_last = VHUB_EP_DMA_RPTR(stat);
+
+	/* Grab current request if any */
+	req = list_first_entry_or_null(&ep->queue, struct ast_vhub_req, queue);
+
+	EPVDBG(ep, "ACK status=%08x is_in=%d ep->d_last=%d..%d\n",
+	       stat, ep->epn.is_in, ep->epn.d_last, d_last);
+
+	/* Check all completed descriptors */
+	while (ep->epn.d_last != d_last) {
+		struct ast_vhub_desc *desc;
+		unsigned int d_num;
+		bool is_last_desc;
+
+		/* Grab next completed descriptor */
+		d_num = ep->epn.d_last;
+		desc = &ep->epn.descs[d_num];
+		ep->epn.d_last = (d_num + 1) & (AST_VHUB_DESCS_COUNT - 1);
+
+		/* Grab len out of descriptor */
+		len = VHUB_DSC1_IN_LEN(le32_to_cpu(desc->w1));
+
+		EPVDBG(ep, " desc %d len=%d req=%p (act=%d)\n",
+		       d_num, len, req, req ? req->active : 0);
+
+		/* If no active request pending, move on */
+		if (!req || !req->active)
+			continue;
+
+		/* Adjust size */
+		req->req.actual += len;
+
+		/* Is that the last chunk ? */
+		is_last_desc = req->last_desc == d_num;
+		CHECK(ep, is_last_desc == (len < ep->ep.maxpacket ||
+					   (req->req.actual >= req->req.length &&
+					    !req->req.zero)),
+		      "Last packet discrepancy: last_desc=%d len=%d r.act=%d "
+		      "r.len=%d r.zero=%d mp=%d\n",
+		      is_last_desc, len, req->req.actual, req->req.length,
+		      req->req.zero, ep->ep.maxpacket);
+
+		if (is_last_desc) {
+			/*
+			 * Because we can only have one request at a time
+			 * in our descriptor list in this implementation,
+			 * d_last and ep->d_last should now be equal
+			 */
+			CHECK(ep, d_last == ep->epn.d_last,
+			      "DMA read ptr mismatch %d vs %d\n",
+			      d_last, ep->epn.d_last);
+
+			/* Note: done will drop and re-acquire the lock */
+			ast_vhub_done(ep, req, 0);
+			req = list_first_entry_or_null(&ep->queue,
+						       struct ast_vhub_req,
+						       queue);
+			break;
+		}
+	}
+
+	/* More work ? */
+	if (req)
+		ast_vhub_epn_kick_desc(ep, req);
+}
+
+void ast_vhub_epn_ack_irq(struct ast_vhub_ep *ep)
+{
+	if (ep->epn.desc_mode)
+		ast_vhub_epn_handle_ack_desc(ep);
+	else
+		ast_vhub_epn_handle_ack(ep);
+}
+
+static int ast_vhub_epn_queue(struct usb_ep* u_ep, struct usb_request *u_req,
+			      gfp_t gfp_flags)
+{
+	struct ast_vhub_req *req = to_ast_req(u_req);
+	struct ast_vhub_ep *ep = to_ast_ep(u_ep);
+	struct ast_vhub *vhub = ep->vhub;
+	unsigned long flags;
+	bool empty;
+	int rc;
+
+	/* Paranoid checks */
+	if (!u_req || !u_req->complete || !u_req->buf) {
+		dev_warn(&vhub->pdev->dev, "Bogus EPn request ! u_req=%p\n", u_req);
+		if (u_req) {
+			dev_warn(&vhub->pdev->dev, "complete=%p internal=%d\n",
+				 u_req->complete, req->internal);
+		}
+		return -EINVAL;
+	}
+
+	/* Endpoint enabled ? */
+	if (!ep->epn.enabled || !u_ep->desc || !ep->dev || !ep->d_idx ||
+	    !ep->dev->enabled) {
+		EPDBG(ep, "Enqueuing request on wrong or disabled EP\n");
+		return -ESHUTDOWN;
+	}
+
+	/* Map request for DMA if possible. For now, the rule for DMA is
+	 * that:
+	 *
+	 *  * For single stage mode (no descriptors):
+	 *
+	 *   - The buffer is aligned to a 8 bytes boundary (HW requirement)
+	 *   - For a OUT endpoint, the request size is a multiple of the EP
+	 *     packet size (otherwise the controller will DMA past the end
+	 *     of the buffer if the host is sending a too long packet).
+	 *
+	 *  * For descriptor mode (tx only for now), always.
+	 *
+	 * We could relax the latter by making the decision to use the bounce
+	 * buffer based on the size of a given *segment* of the request rather
+	 * than the whole request.
+	 */
+	if (ep->epn.desc_mode ||
+	    ((((unsigned long)u_req->buf & 7) == 0) &&
+	     (ep->epn.is_in || !(u_req->length & (u_ep->maxpacket - 1))))) {
+		rc = usb_gadget_map_request_by_dev(&vhub->pdev->dev, u_req,
+					    ep->epn.is_in);
+		if (rc) {
+			dev_warn(&vhub->pdev->dev,
+				 "Request mapping failure %d\n", rc);
+			return rc;
+		}
+	} else
+		u_req->dma = 0;
+
+	EPVDBG(ep, "enqueue req @%p\n", req);
+	EPVDBG(ep, " l=%d dma=0x%x zero=%d noshort=%d noirq=%d is_in=%d\n",
+	       u_req->length, (u32)u_req->dma, u_req->zero,
+	       u_req->short_not_ok, u_req->no_interrupt,
+	       ep->epn.is_in);
+
+	/* Initialize request progress fields */
+	u_req->status = -EINPROGRESS;
+	u_req->actual = 0;
+	req->act_count = 0;
+	req->active = false;
+	req->last_desc = -1;
+	spin_lock_irqsave(&vhub->lock, flags);
+	empty = list_empty(&ep->queue);
+
+	/* Add request to list and kick processing if empty */
+	list_add_tail(&req->queue, &ep->queue);
+	if (empty) {
+		if (ep->epn.desc_mode)
+			ast_vhub_epn_kick_desc(ep, req);
+		else
+			ast_vhub_epn_kick(ep, req);
+	}
+	spin_unlock_irqrestore(&vhub->lock, flags);
+
+	return 0;
+}
+
+static void ast_vhub_stop_active_req(struct ast_vhub_ep *ep,
+				     bool restart_ep)
+{
+	u32 state, reg, loops;
+
+	/* Stop DMA activity */
+	if (ep->epn.desc_mode)
+		writel(VHUB_EP_DMA_CTRL_RESET, ep->epn.regs + AST_VHUB_EP_DMA_CTLSTAT);
+	else
+		writel(0, ep->epn.regs + AST_VHUB_EP_DMA_CTLSTAT);
+
+	/* Wait for it to complete */
+	for (loops = 0; loops < 1000; loops++) {
+		state = readl(ep->epn.regs + AST_VHUB_EP_DMA_CTLSTAT);
+		state = VHUB_EP_DMA_PROC_STATUS(state);
+		if (state == EP_DMA_PROC_RX_IDLE ||
+		    state == EP_DMA_PROC_TX_IDLE)
+			break;
+		udelay(1);
+	}
+	if (loops >= 1000)
+		dev_warn(&ep->vhub->pdev->dev, "Timeout waiting for DMA\n");
+
+	/* If we don't have to restart the endpoint, that's it */
+	if (!restart_ep)
+		return;
+
+	/* Restart the endpoint */
+	if (ep->epn.desc_mode) {
+		/*
+		 * Take out descriptors by resetting the DMA read
+		 * pointer to be equal to the CPU write pointer.
+		 *
+		 * Note: If we ever support creating descriptors for
+		 * requests that aren't the head of the queue, we
+		 * may have to do something more complex here,
+		 * especially if the request being taken out is
+		 * not the current head descriptors.
+		 */
+		reg = VHUB_EP_DMA_SET_RPTR(ep->epn.d_next) |
+			VHUB_EP_DMA_SET_CPU_WPTR(ep->epn.d_next);
+		writel(reg, ep->epn.regs + AST_VHUB_EP_DESC_STATUS);
+
+		/* Then turn it back on */
+		writel(ep->epn.dma_conf,
+		       ep->epn.regs + AST_VHUB_EP_DMA_CTLSTAT);
+	} else {
+		/* Single mode: just turn it back on */
+		writel(ep->epn.dma_conf,
+		       ep->epn.regs + AST_VHUB_EP_DMA_CTLSTAT);
+	}
+}
+
+static int ast_vhub_epn_dequeue(struct usb_ep* u_ep, struct usb_request *u_req)
+{
+	struct ast_vhub_ep *ep = to_ast_ep(u_ep);
+	struct ast_vhub *vhub = ep->vhub;
+	struct ast_vhub_req *req = NULL, *iter;
+	unsigned long flags;
+	int rc = -EINVAL;
+
+	spin_lock_irqsave(&vhub->lock, flags);
+
+	/* Make sure it's actually queued on this endpoint */
+	list_for_each_entry(iter, &ep->queue, queue) {
+		if (&iter->req != u_req)
+			continue;
+		req = iter;
+		break;
+	}
+
+	if (req) {
+		EPVDBG(ep, "dequeue req @%p active=%d\n",
+		       req, req->active);
+		if (req->active)
+			ast_vhub_stop_active_req(ep, true);
+		ast_vhub_done(ep, req, -ECONNRESET);
+		rc = 0;
+	}
+
+	spin_unlock_irqrestore(&vhub->lock, flags);
+	return rc;
+}
+
+void ast_vhub_update_epn_stall(struct ast_vhub_ep *ep)
+{
+	u32 reg;
+
+	if (WARN_ON(ep->d_idx == 0))
+		return;
+	reg = readl(ep->epn.regs + AST_VHUB_EP_CONFIG);
+	if (ep->epn.stalled || ep->epn.wedged)
+		reg |= VHUB_EP_CFG_STALL_CTRL;
+	else
+		reg &= ~VHUB_EP_CFG_STALL_CTRL;
+	writel(reg, ep->epn.regs + AST_VHUB_EP_CONFIG);
+
+	if (!ep->epn.stalled && !ep->epn.wedged)
+		writel(VHUB_EP_TOGGLE_SET_EPNUM(ep->epn.g_idx),
+		       ep->vhub->regs + AST_VHUB_EP_TOGGLE);
+}
+
+static int ast_vhub_set_halt_and_wedge(struct usb_ep* u_ep, bool halt,
+				      bool wedge)
+{
+	struct ast_vhub_ep *ep = to_ast_ep(u_ep);
+	struct ast_vhub *vhub = ep->vhub;
+	unsigned long flags;
+
+	EPDBG(ep, "Set halt (%d) & wedge (%d)\n", halt, wedge);
+
+	if (!u_ep || !u_ep->desc)
+		return -EINVAL;
+	if (ep->d_idx == 0)
+		return 0;
+	if (ep->epn.is_iso)
+		return -EOPNOTSUPP;
+
+	spin_lock_irqsave(&vhub->lock, flags);
+
+	/* Fail with still-busy IN endpoints */
+	if (halt && ep->epn.is_in && !list_empty(&ep->queue)) {
+		spin_unlock_irqrestore(&vhub->lock, flags);
+		return -EAGAIN;
+	}
+	ep->epn.stalled = halt;
+	ep->epn.wedged = wedge;
+	ast_vhub_update_epn_stall(ep);
+
+	spin_unlock_irqrestore(&vhub->lock, flags);
+
+	return 0;
+}
+
+static int ast_vhub_epn_set_halt(struct usb_ep *u_ep, int value)
+{
+	return ast_vhub_set_halt_and_wedge(u_ep, value != 0, false);
+}
+
+static int ast_vhub_epn_set_wedge(struct usb_ep *u_ep)
+{
+	return ast_vhub_set_halt_and_wedge(u_ep, true, true);
+}
+
+static int ast_vhub_epn_disable(struct usb_ep* u_ep)
+{
+	struct ast_vhub_ep *ep = to_ast_ep(u_ep);
+	struct ast_vhub *vhub = ep->vhub;
+	unsigned long flags;
+	u32 imask, ep_ier;
+
+	EPDBG(ep, "Disabling !\n");
+
+	spin_lock_irqsave(&vhub->lock, flags);
+
+	ep->epn.enabled = false;
+
+	/* Stop active DMA if any */
+	ast_vhub_stop_active_req(ep, false);
+
+	/* Disable endpoint */
+	writel(0, ep->epn.regs + AST_VHUB_EP_CONFIG);
+
+	/* Disable ACK interrupt */
+	imask = VHUB_EP_IRQ(ep->epn.g_idx);
+	ep_ier = readl(vhub->regs + AST_VHUB_EP_ACK_IER);
+	ep_ier &= ~imask;
+	writel(ep_ier, vhub->regs + AST_VHUB_EP_ACK_IER);
+	writel(imask, vhub->regs + AST_VHUB_EP_ACK_ISR);
+
+	/* Nuke all pending requests */
+	ast_vhub_nuke(ep, -ESHUTDOWN);
+
+	/* No more descriptor associated with request */
+	ep->ep.desc = NULL;
+
+	spin_unlock_irqrestore(&vhub->lock, flags);
+
+	return 0;
+}
+
+static int ast_vhub_epn_enable(struct usb_ep* u_ep,
+			       const struct usb_endpoint_descriptor *desc)
+{
+	struct ast_vhub_ep *ep = to_ast_ep(u_ep);
+	struct ast_vhub_dev *dev;
+	struct ast_vhub *vhub;
+	u16 maxpacket, type;
+	unsigned long flags;
+	u32 ep_conf, ep_ier, imask;
+
+	/* Check arguments */
+	if (!u_ep || !desc)
+		return -EINVAL;
+
+	maxpacket = usb_endpoint_maxp(desc);
+	if (!ep->d_idx || !ep->dev ||
+	    desc->bDescriptorType != USB_DT_ENDPOINT ||
+	    maxpacket == 0 || maxpacket > ep->ep.maxpacket) {
+		EPDBG(ep, "Invalid EP enable,d_idx=%d,dev=%p,type=%d,mp=%d/%d\n",
+		      ep->d_idx, ep->dev, desc->bDescriptorType,
+		      maxpacket, ep->ep.maxpacket);
+		return -EINVAL;
+	}
+	if (ep->d_idx != usb_endpoint_num(desc)) {
+		EPDBG(ep, "EP number mismatch !\n");
+		return -EINVAL;
+	}
+
+	if (ep->epn.enabled) {
+		EPDBG(ep, "Already enabled\n");
+		return -EBUSY;
+	}
+	dev = ep->dev;
+	vhub = ep->vhub;
+
+	/* Check device state */
+	if (!dev->driver) {
+		EPDBG(ep, "Bogus device state: driver=%p speed=%d\n",
+		       dev->driver, dev->gadget.speed);
+		return -ESHUTDOWN;
+	}
+
+	/* Grab some info from the descriptor */
+	ep->epn.is_in = usb_endpoint_dir_in(desc);
+	ep->ep.maxpacket = maxpacket;
+	type = usb_endpoint_type(desc);
+	ep->epn.d_next = ep->epn.d_last = 0;
+	ep->epn.is_iso = false;
+	ep->epn.stalled = false;
+	ep->epn.wedged = false;
+
+	EPDBG(ep, "Enabling [%s] %s num %d maxpacket=%d\n",
+	      ep->epn.is_in ? "in" : "out", usb_ep_type_string(type),
+	      usb_endpoint_num(desc), maxpacket);
+
+	/* Can we use DMA descriptor mode ? */
+	ep->epn.desc_mode = ep->epn.descs && ep->epn.is_in;
+	if (ep->epn.desc_mode)
+		memset(ep->epn.descs, 0, 8 * AST_VHUB_DESCS_COUNT);
+
+	/*
+	 * Large send function can send up to 8 packets from
+	 * one descriptor with a limit of 4095 bytes.
+	 */
+	ep->epn.chunk_max = ep->ep.maxpacket;
+	if (ep->epn.is_in) {
+		ep->epn.chunk_max <<= 3;
+		while (ep->epn.chunk_max > 4095)
+			ep->epn.chunk_max -= ep->ep.maxpacket;
+	}
+
+	switch(type) {
+	case USB_ENDPOINT_XFER_CONTROL:
+		EPDBG(ep, "Only one control endpoint\n");
+		return -EINVAL;
+	case USB_ENDPOINT_XFER_INT:
+		ep_conf = VHUB_EP_CFG_SET_TYPE(EP_TYPE_INT);
+		break;
+	case USB_ENDPOINT_XFER_BULK:
+		ep_conf = VHUB_EP_CFG_SET_TYPE(EP_TYPE_BULK);
+		break;
+	case USB_ENDPOINT_XFER_ISOC:
+		ep_conf = VHUB_EP_CFG_SET_TYPE(EP_TYPE_ISO);
+		ep->epn.is_iso = true;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	/* Encode the rest of the EP config register */
+	if (maxpacket < 1024)
+		ep_conf |= VHUB_EP_CFG_SET_MAX_PKT(maxpacket);
+	if (!ep->epn.is_in)
+		ep_conf |= VHUB_EP_CFG_DIR_OUT;
+	ep_conf |= VHUB_EP_CFG_SET_EP_NUM(usb_endpoint_num(desc));
+	ep_conf |= VHUB_EP_CFG_ENABLE;
+	ep_conf |= VHUB_EP_CFG_SET_DEV(dev->index + 1);
+	EPVDBG(ep, "config=%08x\n", ep_conf);
+
+	spin_lock_irqsave(&vhub->lock, flags);
+
+	/* Disable HW and reset DMA */
+	writel(0, ep->epn.regs + AST_VHUB_EP_CONFIG);
+	writel(VHUB_EP_DMA_CTRL_RESET,
+	       ep->epn.regs + AST_VHUB_EP_DMA_CTLSTAT);
+
+	/* Configure and enable */
+	writel(ep_conf, ep->epn.regs + AST_VHUB_EP_CONFIG);
+
+	if (ep->epn.desc_mode) {
+		/* Clear DMA status, including the DMA read ptr */
+		writel(0, ep->epn.regs + AST_VHUB_EP_DESC_STATUS);
+
+		/* Set descriptor base */
+		writel(ep->epn.descs_dma,
+		       ep->epn.regs + AST_VHUB_EP_DESC_BASE);
+
+		/* Set base DMA config value */
+		ep->epn.dma_conf = VHUB_EP_DMA_DESC_MODE;
+		if (ep->epn.is_in)
+			ep->epn.dma_conf |= VHUB_EP_DMA_IN_LONG_MODE;
+
+		/* First reset and disable all operations */
+		writel(ep->epn.dma_conf | VHUB_EP_DMA_CTRL_RESET,
+		       ep->epn.regs + AST_VHUB_EP_DMA_CTLSTAT);
+
+		/* Enable descriptor mode */
+		writel(ep->epn.dma_conf,
+		       ep->epn.regs + AST_VHUB_EP_DMA_CTLSTAT);
+	} else {
+		/* Set base DMA config value */
+		ep->epn.dma_conf = VHUB_EP_DMA_SINGLE_STAGE;
+
+		/* Reset and switch to single stage mode */
+		writel(ep->epn.dma_conf | VHUB_EP_DMA_CTRL_RESET,
+		       ep->epn.regs + AST_VHUB_EP_DMA_CTLSTAT);
+		writel(ep->epn.dma_conf,
+		       ep->epn.regs + AST_VHUB_EP_DMA_CTLSTAT);
+		writel(0, ep->epn.regs + AST_VHUB_EP_DESC_STATUS);
+	}
+
+	/* Cleanup data toggle just in case */
+	writel(VHUB_EP_TOGGLE_SET_EPNUM(ep->epn.g_idx),
+	       vhub->regs + AST_VHUB_EP_TOGGLE);
+
+	/* Cleanup and enable ACK interrupt */
+	imask = VHUB_EP_IRQ(ep->epn.g_idx);
+	writel(imask, vhub->regs + AST_VHUB_EP_ACK_ISR);
+	ep_ier = readl(vhub->regs + AST_VHUB_EP_ACK_IER);
+	ep_ier |= imask;
+	writel(ep_ier, vhub->regs + AST_VHUB_EP_ACK_IER);
+
+	/* Woot, we are online ! */
+	ep->epn.enabled = true;
+
+	spin_unlock_irqrestore(&vhub->lock, flags);
+
+	return 0;
+}
+
+static void ast_vhub_epn_dispose(struct usb_ep *u_ep)
+{
+	struct ast_vhub_ep *ep = to_ast_ep(u_ep);
+
+	if (WARN_ON(!ep->dev || !ep->d_idx))
+		return;
+
+	EPDBG(ep, "Releasing endpoint\n");
+
+	/* Take it out of the EP list */
+	list_del_init(&ep->ep.ep_list);
+
+	/* Mark the address free in the device */
+	ep->dev->epns[ep->d_idx - 1] = NULL;
+
+	/* Free name & DMA buffers */
+	kfree(ep->ep.name);
+	ep->ep.name = NULL;
+	dma_free_coherent(&ep->vhub->pdev->dev,
+			  AST_VHUB_EPn_MAX_PACKET +
+			  8 * AST_VHUB_DESCS_COUNT,
+			  ep->buf, ep->buf_dma);
+	ep->buf = NULL;
+	ep->epn.descs = NULL;
+
+	/* Mark free */
+	ep->dev = NULL;
+}
+
+static const struct usb_ep_ops ast_vhub_epn_ops = {
+	.enable		= ast_vhub_epn_enable,
+	.disable	= ast_vhub_epn_disable,
+	.dispose	= ast_vhub_epn_dispose,
+	.queue		= ast_vhub_epn_queue,
+	.dequeue	= ast_vhub_epn_dequeue,
+	.set_halt	= ast_vhub_epn_set_halt,
+	.set_wedge	= ast_vhub_epn_set_wedge,
+	.alloc_request	= ast_vhub_alloc_request,
+	.free_request	= ast_vhub_free_request,
+};
+
+struct ast_vhub_ep *ast_vhub_alloc_epn(struct ast_vhub_dev *d, u8 addr)
+{
+	struct ast_vhub *vhub = d->vhub;
+	struct ast_vhub_ep *ep;
+	unsigned long flags;
+	int i;
+
+	/* Find a free one (no device) */
+	spin_lock_irqsave(&vhub->lock, flags);
+	for (i = 0; i < vhub->max_epns; i++)
+		if (vhub->epns[i].dev == NULL)
+			break;
+	if (i >= vhub->max_epns) {
+		spin_unlock_irqrestore(&vhub->lock, flags);
+		return NULL;
+	}
+
+	/* Set it up */
+	ep = &vhub->epns[i];
+	ep->dev = d;
+	spin_unlock_irqrestore(&vhub->lock, flags);
+
+	DDBG(d, "Allocating gen EP %d for addr %d\n", i, addr);
+	INIT_LIST_HEAD(&ep->queue);
+	ep->d_idx = addr;
+	ep->vhub = vhub;
+	ep->ep.ops = &ast_vhub_epn_ops;
+	ep->ep.name = kasprintf(GFP_KERNEL, "ep%d", addr);
+	d->epns[addr-1] = ep;
+	ep->epn.g_idx = i;
+	ep->epn.regs = vhub->regs + 0x200 + (i * 0x10);
+
+	ep->buf = dma_alloc_coherent(&vhub->pdev->dev,
+				     AST_VHUB_EPn_MAX_PACKET +
+				     8 * AST_VHUB_DESCS_COUNT,
+				     &ep->buf_dma, GFP_KERNEL);
+	if (!ep->buf) {
+		kfree(ep->ep.name);
+		ep->ep.name = NULL;
+		return NULL;
+	}
+	ep->epn.descs = ep->buf + AST_VHUB_EPn_MAX_PACKET;
+	ep->epn.descs_dma = ep->buf_dma + AST_VHUB_EPn_MAX_PACKET;
+
+	usb_ep_set_maxpacket_limit(&ep->ep, AST_VHUB_EPn_MAX_PACKET);
+	list_add_tail(&ep->ep.ep_list, &d->gadget.ep_list);
+	ep->ep.caps.type_iso = true;
+	ep->ep.caps.type_bulk = true;
+	ep->ep.caps.type_int = true;
+	ep->ep.caps.dir_in = true;
+	ep->ep.caps.dir_out = true;
+
+	return ep;
+}