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, 642 insertions, 0 deletions

diff --git a/drivers/usb/musb/tusb6010_omap.c b/drivers/usb/musb/tusb6010_omap.c
new file mode 100644
index 000000000..60a93b8bb
--- /dev/null
+++ b/drivers/usb/musb/tusb6010_omap.c
@@ -0,0 +1,642 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * TUSB6010 USB 2.0 OTG Dual Role controller OMAP DMA interface
+ *
+ * Copyright (C) 2006 Nokia Corporation
+ * Tony Lindgren <tony@atomide.com>
+ */
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/errno.h>
+#include <linux/usb.h>
+#include <linux/platform_device.h>
+#include <linux/dma-mapping.h>
+#include <linux/slab.h>
+#include <linux/dmaengine.h>
+
+#include "musb_core.h"
+#include "tusb6010.h"
+
+#define to_chdat(c)		((struct tusb_omap_dma_ch *)(c)->private_data)
+
+#define MAX_DMAREQ		5	/* REVISIT: Really 6, but req5 not OK */
+
+struct tusb_dma_data {
+	s8			dmareq;
+	struct dma_chan		*chan;
+};
+
+struct tusb_omap_dma_ch {
+	struct musb		*musb;
+	void __iomem		*tbase;
+	unsigned long		phys_offset;
+	int			epnum;
+	u8			tx;
+	struct musb_hw_ep	*hw_ep;
+
+	struct tusb_dma_data	*dma_data;
+
+	struct tusb_omap_dma	*tusb_dma;
+
+	dma_addr_t		dma_addr;
+
+	u32			len;
+	u16			packet_sz;
+	u16			transfer_packet_sz;
+	u32			transfer_len;
+	u32			completed_len;
+};
+
+struct tusb_omap_dma {
+	struct dma_controller		controller;
+	void __iomem			*tbase;
+
+	struct tusb_dma_data		dma_pool[MAX_DMAREQ];
+	unsigned			multichannel:1;
+};
+
+/*
+ * Allocate dmareq0 to the current channel unless it's already taken
+ */
+static inline int tusb_omap_use_shared_dmareq(struct tusb_omap_dma_ch *chdat)
+{
+	u32		reg = musb_readl(chdat->tbase, TUSB_DMA_EP_MAP);
+
+	if (reg != 0) {
+		dev_dbg(chdat->musb->controller, "ep%i dmareq0 is busy for ep%i\n",
+			chdat->epnum, reg & 0xf);
+		return -EAGAIN;
+	}
+
+	if (chdat->tx)
+		reg = (1 << 4) | chdat->epnum;
+	else
+		reg = chdat->epnum;
+
+	musb_writel(chdat->tbase, TUSB_DMA_EP_MAP, reg);
+
+	return 0;
+}
+
+static inline void tusb_omap_free_shared_dmareq(struct tusb_omap_dma_ch *chdat)
+{
+	u32		reg = musb_readl(chdat->tbase, TUSB_DMA_EP_MAP);
+
+	if ((reg & 0xf) != chdat->epnum) {
+		printk(KERN_ERR "ep%i trying to release dmareq0 for ep%i\n",
+			chdat->epnum, reg & 0xf);
+		return;
+	}
+	musb_writel(chdat->tbase, TUSB_DMA_EP_MAP, 0);
+}
+
+/*
+ * See also musb_dma_completion in plat_uds.c and musb_g_[tx|rx]() in
+ * musb_gadget.c.
+ */
+static void tusb_omap_dma_cb(void *data)
+{
+	struct dma_channel	*channel = (struct dma_channel *)data;
+	struct tusb_omap_dma_ch	*chdat = to_chdat(channel);
+	struct tusb_omap_dma	*tusb_dma = chdat->tusb_dma;
+	struct musb		*musb = chdat->musb;
+	struct device		*dev = musb->controller;
+	struct musb_hw_ep	*hw_ep = chdat->hw_ep;
+	void __iomem		*ep_conf = hw_ep->conf;
+	void __iomem		*mbase = musb->mregs;
+	unsigned long		remaining, flags, pio;
+
+	spin_lock_irqsave(&musb->lock, flags);
+
+	dev_dbg(musb->controller, "ep%i %s dma callback\n",
+		chdat->epnum, chdat->tx ? "tx" : "rx");
+
+	if (chdat->tx)
+		remaining = musb_readl(ep_conf, TUSB_EP_TX_OFFSET);
+	else
+		remaining = musb_readl(ep_conf, TUSB_EP_RX_OFFSET);
+
+	remaining = TUSB_EP_CONFIG_XFR_SIZE(remaining);
+
+	/* HW issue #10: XFR_SIZE may get corrupt on DMA (both async & sync) */
+	if (unlikely(remaining > chdat->transfer_len)) {
+		dev_dbg(musb->controller, "Corrupt %s XFR_SIZE: 0x%08lx\n",
+			chdat->tx ? "tx" : "rx", remaining);
+		remaining = 0;
+	}
+
+	channel->actual_len = chdat->transfer_len - remaining;
+	pio = chdat->len - channel->actual_len;
+
+	dev_dbg(musb->controller, "DMA remaining %lu/%u\n", remaining, chdat->transfer_len);
+
+	/* Transfer remaining 1 - 31 bytes */
+	if (pio > 0 && pio < 32) {
+		u8	*buf;
+
+		dev_dbg(musb->controller, "Using PIO for remaining %lu bytes\n", pio);
+		buf = phys_to_virt((u32)chdat->dma_addr) + chdat->transfer_len;
+		if (chdat->tx) {
+			dma_unmap_single(dev, chdat->dma_addr,
+						chdat->transfer_len,
+						DMA_TO_DEVICE);
+			musb_write_fifo(hw_ep, pio, buf);
+		} else {
+			dma_unmap_single(dev, chdat->dma_addr,
+						chdat->transfer_len,
+						DMA_FROM_DEVICE);
+			musb_read_fifo(hw_ep, pio, buf);
+		}
+		channel->actual_len += pio;
+	}
+
+	if (!tusb_dma->multichannel)
+		tusb_omap_free_shared_dmareq(chdat);
+
+	channel->status = MUSB_DMA_STATUS_FREE;
+
+	musb_dma_completion(musb, chdat->epnum, chdat->tx);
+
+	/* We must terminate short tx transfers manually by setting TXPKTRDY.
+	 * REVISIT: This same problem may occur with other MUSB dma as well.
+	 * Easy to test with g_ether by pinging the MUSB board with ping -s54.
+	 */
+	if ((chdat->transfer_len < chdat->packet_sz)
+			|| (chdat->transfer_len % chdat->packet_sz != 0)) {
+		u16	csr;
+
+		if (chdat->tx) {
+			dev_dbg(musb->controller, "terminating short tx packet\n");
+			musb_ep_select(mbase, chdat->epnum);
+			csr = musb_readw(hw_ep->regs, MUSB_TXCSR);
+			csr |= MUSB_TXCSR_MODE | MUSB_TXCSR_TXPKTRDY
+				| MUSB_TXCSR_P_WZC_BITS;
+			musb_writew(hw_ep->regs, MUSB_TXCSR, csr);
+		}
+	}
+
+	spin_unlock_irqrestore(&musb->lock, flags);
+}
+
+static int tusb_omap_dma_program(struct dma_channel *channel, u16 packet_sz,
+				u8 rndis_mode, dma_addr_t dma_addr, u32 len)
+{
+	struct tusb_omap_dma_ch		*chdat = to_chdat(channel);
+	struct tusb_omap_dma		*tusb_dma = chdat->tusb_dma;
+	struct musb			*musb = chdat->musb;
+	struct device			*dev = musb->controller;
+	struct musb_hw_ep		*hw_ep = chdat->hw_ep;
+	void __iomem			*mbase = musb->mregs;
+	void __iomem			*ep_conf = hw_ep->conf;
+	dma_addr_t			fifo_addr = hw_ep->fifo_sync;
+	u32				dma_remaining;
+	u16				csr;
+	u32				psize;
+	struct tusb_dma_data		*dma_data;
+	struct dma_async_tx_descriptor	*dma_desc;
+	struct dma_slave_config		dma_cfg;
+	enum dma_transfer_direction	dma_dir;
+	u32				port_window;
+	int				ret;
+
+	if (unlikely(dma_addr & 0x1) || (len < 32) || (len > packet_sz))
+		return false;
+
+	/*
+	 * HW issue #10: Async dma will eventually corrupt the XFR_SIZE
+	 * register which will cause missed DMA interrupt. We could try to
+	 * use a timer for the callback, but it is unsafe as the XFR_SIZE
+	 * register is corrupt, and we won't know if the DMA worked.
+	 */
+	if (dma_addr & 0x2)
+		return false;
+
+	/*
+	 * Because of HW issue #10, it seems like mixing sync DMA and async
+	 * PIO access can confuse the DMA. Make sure XFR_SIZE is reset before
+	 * using the channel for DMA.
+	 */
+	if (chdat->tx)
+		dma_remaining = musb_readl(ep_conf, TUSB_EP_TX_OFFSET);
+	else
+		dma_remaining = musb_readl(ep_conf, TUSB_EP_RX_OFFSET);
+
+	dma_remaining = TUSB_EP_CONFIG_XFR_SIZE(dma_remaining);
+	if (dma_remaining) {
+		dev_dbg(musb->controller, "Busy %s dma, not using: %08x\n",
+			chdat->tx ? "tx" : "rx", dma_remaining);
+		return false;
+	}
+
+	chdat->transfer_len = len & ~0x1f;
+
+	if (len < packet_sz)
+		chdat->transfer_packet_sz = chdat->transfer_len;
+	else
+		chdat->transfer_packet_sz = packet_sz;
+
+	dma_data = chdat->dma_data;
+	if (!tusb_dma->multichannel) {
+		if (tusb_omap_use_shared_dmareq(chdat) != 0) {
+			dev_dbg(musb->controller, "could not get dma for ep%i\n", chdat->epnum);
+			return false;
+		}
+		if (dma_data->dmareq < 0) {
+			/* REVISIT: This should get blocked earlier, happens
+			 * with MSC ErrorRecoveryTest
+			 */
+			WARN_ON(1);
+			return false;
+		}
+	}
+
+	chdat->packet_sz = packet_sz;
+	chdat->len = len;
+	channel->actual_len = 0;
+	chdat->dma_addr = dma_addr;
+	channel->status = MUSB_DMA_STATUS_BUSY;
+
+	/* Since we're recycling dma areas, we need to clean or invalidate */
+	if (chdat->tx) {
+		dma_dir = DMA_MEM_TO_DEV;
+		dma_map_single(dev, phys_to_virt(dma_addr), len,
+				DMA_TO_DEVICE);
+	} else {
+		dma_dir = DMA_DEV_TO_MEM;
+		dma_map_single(dev, phys_to_virt(dma_addr), len,
+				DMA_FROM_DEVICE);
+	}
+
+	memset(&dma_cfg, 0, sizeof(dma_cfg));
+
+	/* Use 16-bit transfer if dma_addr is not 32-bit aligned */
+	if ((dma_addr & 0x3) == 0) {
+		dma_cfg.src_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
+		dma_cfg.dst_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
+		port_window = 8;
+	} else {
+		dma_cfg.src_addr_width = DMA_SLAVE_BUSWIDTH_2_BYTES;
+		dma_cfg.dst_addr_width = DMA_SLAVE_BUSWIDTH_2_BYTES;
+		port_window = 16;
+
+		fifo_addr = hw_ep->fifo_async;
+	}
+
+	dev_dbg(musb->controller,
+		"ep%i %s dma: %pad len: %u(%u) packet_sz: %i(%i)\n",
+		chdat->epnum, chdat->tx ? "tx" : "rx", &dma_addr,
+		chdat->transfer_len, len, chdat->transfer_packet_sz, packet_sz);
+
+	dma_cfg.src_addr = fifo_addr;
+	dma_cfg.dst_addr = fifo_addr;
+	dma_cfg.src_port_window_size = port_window;
+	dma_cfg.src_maxburst = port_window;
+	dma_cfg.dst_port_window_size = port_window;
+	dma_cfg.dst_maxburst = port_window;
+
+	ret = dmaengine_slave_config(dma_data->chan, &dma_cfg);
+	if (ret) {
+		dev_err(musb->controller, "DMA slave config failed: %d\n", ret);
+		return false;
+	}
+
+	dma_desc = dmaengine_prep_slave_single(dma_data->chan, dma_addr,
+					chdat->transfer_len, dma_dir,
+					DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+	if (!dma_desc) {
+		dev_err(musb->controller, "DMA prep_slave_single failed\n");
+		return false;
+	}
+
+	dma_desc->callback = tusb_omap_dma_cb;
+	dma_desc->callback_param = channel;
+	dmaengine_submit(dma_desc);
+
+	dev_dbg(musb->controller,
+		"ep%i %s using %i-bit %s dma from %pad to %pad\n",
+		chdat->epnum, chdat->tx ? "tx" : "rx",
+		dma_cfg.src_addr_width * 8,
+		((dma_addr & 0x3) == 0) ? "sync" : "async",
+		(dma_dir == DMA_MEM_TO_DEV) ? &dma_addr : &fifo_addr,
+		(dma_dir == DMA_MEM_TO_DEV) ? &fifo_addr : &dma_addr);
+
+	/*
+	 * Prepare MUSB for DMA transfer
+	 */
+	musb_ep_select(mbase, chdat->epnum);
+	if (chdat->tx) {
+		csr = musb_readw(hw_ep->regs, MUSB_TXCSR);
+		csr |= (MUSB_TXCSR_AUTOSET | MUSB_TXCSR_DMAENAB
+			| MUSB_TXCSR_DMAMODE | MUSB_TXCSR_MODE);
+		csr &= ~MUSB_TXCSR_P_UNDERRUN;
+		musb_writew(hw_ep->regs, MUSB_TXCSR, csr);
+	} else {
+		csr = musb_readw(hw_ep->regs, MUSB_RXCSR);
+		csr |= MUSB_RXCSR_DMAENAB;
+		csr &= ~(MUSB_RXCSR_AUTOCLEAR | MUSB_RXCSR_DMAMODE);
+		musb_writew(hw_ep->regs, MUSB_RXCSR,
+			csr | MUSB_RXCSR_P_WZC_BITS);
+	}
+
+	/* Start DMA transfer */
+	dma_async_issue_pending(dma_data->chan);
+
+	if (chdat->tx) {
+		/* Send transfer_packet_sz packets at a time */
+		psize = musb_readl(ep_conf, TUSB_EP_MAX_PACKET_SIZE_OFFSET);
+		psize &= ~0x7ff;
+		psize |= chdat->transfer_packet_sz;
+		musb_writel(ep_conf, TUSB_EP_MAX_PACKET_SIZE_OFFSET, psize);
+
+		musb_writel(ep_conf, TUSB_EP_TX_OFFSET,
+			TUSB_EP_CONFIG_XFR_SIZE(chdat->transfer_len));
+	} else {
+		/* Receive transfer_packet_sz packets at a time */
+		psize = musb_readl(ep_conf, TUSB_EP_MAX_PACKET_SIZE_OFFSET);
+		psize &= ~(0x7ff << 16);
+		psize |= (chdat->transfer_packet_sz << 16);
+		musb_writel(ep_conf, TUSB_EP_MAX_PACKET_SIZE_OFFSET, psize);
+
+		musb_writel(ep_conf, TUSB_EP_RX_OFFSET,
+			TUSB_EP_CONFIG_XFR_SIZE(chdat->transfer_len));
+	}
+
+	return true;
+}
+
+static int tusb_omap_dma_abort(struct dma_channel *channel)
+{
+	struct tusb_omap_dma_ch	*chdat = to_chdat(channel);
+
+	if (chdat->dma_data)
+		dmaengine_terminate_all(chdat->dma_data->chan);
+
+	channel->status = MUSB_DMA_STATUS_FREE;
+
+	return 0;
+}
+
+static inline int tusb_omap_dma_allocate_dmareq(struct tusb_omap_dma_ch *chdat)
+{
+	u32		reg = musb_readl(chdat->tbase, TUSB_DMA_EP_MAP);
+	int		i, dmareq_nr = -1;
+
+	for (i = 0; i < MAX_DMAREQ; i++) {
+		int cur = (reg & (0xf << (i * 5))) >> (i * 5);
+		if (cur == 0) {
+			dmareq_nr = i;
+			break;
+		}
+	}
+
+	if (dmareq_nr == -1)
+		return -EAGAIN;
+
+	reg |= (chdat->epnum << (dmareq_nr * 5));
+	if (chdat->tx)
+		reg |= ((1 << 4) << (dmareq_nr * 5));
+	musb_writel(chdat->tbase, TUSB_DMA_EP_MAP, reg);
+
+	chdat->dma_data = &chdat->tusb_dma->dma_pool[dmareq_nr];
+
+	return 0;
+}
+
+static inline void tusb_omap_dma_free_dmareq(struct tusb_omap_dma_ch *chdat)
+{
+	u32 reg;
+
+	if (!chdat || !chdat->dma_data || chdat->dma_data->dmareq < 0)
+		return;
+
+	reg = musb_readl(chdat->tbase, TUSB_DMA_EP_MAP);
+	reg &= ~(0x1f << (chdat->dma_data->dmareq * 5));
+	musb_writel(chdat->tbase, TUSB_DMA_EP_MAP, reg);
+
+	chdat->dma_data = NULL;
+}
+
+static struct dma_channel *dma_channel_pool[MAX_DMAREQ];
+
+static struct dma_channel *
+tusb_omap_dma_allocate(struct dma_controller *c,
+		struct musb_hw_ep *hw_ep,
+		u8 tx)
+{
+	int ret, i;
+	struct tusb_omap_dma	*tusb_dma;
+	struct musb		*musb;
+	struct dma_channel	*channel = NULL;
+	struct tusb_omap_dma_ch	*chdat = NULL;
+	struct tusb_dma_data	*dma_data = NULL;
+
+	tusb_dma = container_of(c, struct tusb_omap_dma, controller);
+	musb = tusb_dma->controller.musb;
+
+	/* REVISIT: Why does dmareq5 not work? */
+	if (hw_ep->epnum == 0) {
+		dev_dbg(musb->controller, "Not allowing DMA for ep0 %s\n", tx ? "tx" : "rx");
+		return NULL;
+	}
+
+	for (i = 0; i < MAX_DMAREQ; i++) {
+		struct dma_channel *ch = dma_channel_pool[i];
+		if (ch->status == MUSB_DMA_STATUS_UNKNOWN) {
+			ch->status = MUSB_DMA_STATUS_FREE;
+			channel = ch;
+			chdat = ch->private_data;
+			break;
+		}
+	}
+
+	if (!channel)
+		return NULL;
+
+	chdat->musb = tusb_dma->controller.musb;
+	chdat->tbase = tusb_dma->tbase;
+	chdat->hw_ep = hw_ep;
+	chdat->epnum = hw_ep->epnum;
+	chdat->completed_len = 0;
+	chdat->tusb_dma = tusb_dma;
+	if (tx)
+		chdat->tx = 1;
+	else
+		chdat->tx = 0;
+
+	channel->max_len = 0x7fffffff;
+	channel->desired_mode = 0;
+	channel->actual_len = 0;
+
+	if (!chdat->dma_data) {
+		if (tusb_dma->multichannel) {
+			ret = tusb_omap_dma_allocate_dmareq(chdat);
+			if (ret != 0)
+				goto free_dmareq;
+		} else {
+			chdat->dma_data = &tusb_dma->dma_pool[0];
+		}
+	}
+
+	dma_data = chdat->dma_data;
+
+	dev_dbg(musb->controller, "ep%i %s dma: %s dmareq%i\n",
+		chdat->epnum,
+		chdat->tx ? "tx" : "rx",
+		tusb_dma->multichannel ? "shared" : "dedicated",
+		dma_data->dmareq);
+
+	return channel;
+
+free_dmareq:
+	tusb_omap_dma_free_dmareq(chdat);
+
+	dev_dbg(musb->controller, "ep%i: Could not get a DMA channel\n", chdat->epnum);
+	channel->status = MUSB_DMA_STATUS_UNKNOWN;
+
+	return NULL;
+}
+
+static void tusb_omap_dma_release(struct dma_channel *channel)
+{
+	struct tusb_omap_dma_ch	*chdat = to_chdat(channel);
+	struct musb		*musb = chdat->musb;
+
+	dev_dbg(musb->controller, "Release for ep%i\n", chdat->epnum);
+
+	channel->status = MUSB_DMA_STATUS_UNKNOWN;
+
+	dmaengine_terminate_sync(chdat->dma_data->chan);
+	tusb_omap_dma_free_dmareq(chdat);
+
+	channel = NULL;
+}
+
+void tusb_dma_controller_destroy(struct dma_controller *c)
+{
+	struct tusb_omap_dma	*tusb_dma;
+	int			i;
+
+	tusb_dma = container_of(c, struct tusb_omap_dma, controller);
+	for (i = 0; i < MAX_DMAREQ; i++) {
+		struct dma_channel *ch = dma_channel_pool[i];
+		if (ch) {
+			kfree(ch->private_data);
+			kfree(ch);
+		}
+
+		/* Free up the DMA channels */
+		if (tusb_dma && tusb_dma->dma_pool[i].chan)
+			dma_release_channel(tusb_dma->dma_pool[i].chan);
+	}
+
+	kfree(tusb_dma);
+}
+EXPORT_SYMBOL_GPL(tusb_dma_controller_destroy);
+
+static int tusb_omap_allocate_dma_pool(struct tusb_omap_dma *tusb_dma)
+{
+	struct musb *musb = tusb_dma->controller.musb;
+	int i;
+	int ret = 0;
+
+	for (i = 0; i < MAX_DMAREQ; i++) {
+		struct tusb_dma_data *dma_data = &tusb_dma->dma_pool[i];
+
+		/*
+		 * Request DMA channels:
+		 * - one channel in case of non multichannel mode
+		 * - MAX_DMAREQ number of channels in multichannel mode
+		 */
+		if (i == 0 || tusb_dma->multichannel) {
+			char ch_name[8];
+
+			sprintf(ch_name, "dmareq%d", i);
+			dma_data->chan = dma_request_chan(musb->controller,
+							  ch_name);
+			if (IS_ERR(dma_data->chan)) {
+				dev_err(musb->controller,
+					"Failed to request %s\n", ch_name);
+				ret = PTR_ERR(dma_data->chan);
+				goto dma_error;
+			}
+
+			dma_data->dmareq = i;
+		} else {
+			dma_data->dmareq = -1;
+		}
+	}
+
+	return 0;
+
+dma_error:
+	for (; i >= 0; i--) {
+		struct tusb_dma_data *dma_data = &tusb_dma->dma_pool[i];
+
+		if (dma_data->dmareq >= 0)
+			dma_release_channel(dma_data->chan);
+	}
+
+	return ret;
+}
+
+struct dma_controller *
+tusb_dma_controller_create(struct musb *musb, void __iomem *base)
+{
+	void __iomem		*tbase = musb->ctrl_base;
+	struct tusb_omap_dma	*tusb_dma;
+	int			i;
+
+	/* REVISIT: Get dmareq lines used from board-*.c */
+
+	musb_writel(musb->ctrl_base, TUSB_DMA_INT_MASK, 0x7fffffff);
+	musb_writel(musb->ctrl_base, TUSB_DMA_EP_MAP, 0);
+
+	musb_writel(tbase, TUSB_DMA_REQ_CONF,
+		TUSB_DMA_REQ_CONF_BURST_SIZE(2)
+		| TUSB_DMA_REQ_CONF_DMA_REQ_EN(0x3f)
+		| TUSB_DMA_REQ_CONF_DMA_REQ_ASSER(2));
+
+	tusb_dma = kzalloc(sizeof(struct tusb_omap_dma), GFP_KERNEL);
+	if (!tusb_dma)
+		goto out;
+
+	tusb_dma->controller.musb = musb;
+	tusb_dma->tbase = musb->ctrl_base;
+
+	tusb_dma->controller.channel_alloc = tusb_omap_dma_allocate;
+	tusb_dma->controller.channel_release = tusb_omap_dma_release;
+	tusb_dma->controller.channel_program = tusb_omap_dma_program;
+	tusb_dma->controller.channel_abort = tusb_omap_dma_abort;
+
+	if (musb->tusb_revision >= TUSB_REV_30)
+		tusb_dma->multichannel = 1;
+
+	for (i = 0; i < MAX_DMAREQ; i++) {
+		struct dma_channel	*ch;
+		struct tusb_omap_dma_ch	*chdat;
+
+		ch = kzalloc(sizeof(struct dma_channel), GFP_KERNEL);
+		if (!ch)
+			goto cleanup;
+
+		dma_channel_pool[i] = ch;
+
+		chdat = kzalloc(sizeof(struct tusb_omap_dma_ch), GFP_KERNEL);
+		if (!chdat)
+			goto cleanup;
+
+		ch->status = MUSB_DMA_STATUS_UNKNOWN;
+		ch->private_data = chdat;
+	}
+
+	if (tusb_omap_allocate_dma_pool(tusb_dma))
+		goto cleanup;
+
+	return &tusb_dma->controller;
+
+cleanup:
+	musb_dma_controller_destroy(&tusb_dma->controller);
+out:
+	return NULL;
+}
+EXPORT_SYMBOL_GPL(tusb_dma_controller_create);