drivers/staging/most/net/net.c - linux/kernel/git/gregkh/driver-core - Git at Google

 // SPDX-License-Identifier: GPL-2.0
 /*
  * net.c - Networking component for Mostcore
  *
  * Copyright (C) 2015, Microchip Technology Germany II GmbH & Co. KG
  */

 #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt

 #include <linux/module.h>
 #include <linux/netdevice.h>
 #include <linux/etherdevice.h>
 #include <linux/slab.h>
 #include <linux/init.h>
 #include <linux/list.h>
 #include <linux/wait.h>
 #include <linux/kobject.h>
 #include <linux/most.h>

 #define MEP_HDR_LEN 8
 #define MDP_HDR_LEN 16
 #define MAMAC_DATA_LEN (1024 - MDP_HDR_LEN)

 #define PMHL 5

 #define PMS_TELID_UNSEGM_MAMAC	0x0A
 #define PMS_FIFONO_MDP		0x01
 #define PMS_FIFONO_MEP		0x04
 #define PMS_MSGTYPE_DATA	0x04
 #define PMS_DEF_PRIO		0
 #define MEP_DEF_RETRY		15

 #define PMS_FIFONO_MASK		0x07
 #define PMS_FIFONO_SHIFT	3
 #define PMS_RETRY_SHIFT		4
 #define PMS_TELID_MASK		0x0F
 #define PMS_TELID_SHIFT		4

 #define HB(value)		((u8)((u16)(value) >> 8))
 #define LB(value)		((u8)(value))

 #define EXTRACT_BIT_SET(bitset_name, value) \
 	(((value) >> bitset_name##_SHIFT) & bitset_name##_MASK)

 #define PMS_IS_MEP(buf, len) \
 	((len) > MEP_HDR_LEN && \
 	 EXTRACT_BIT_SET(PMS_FIFONO, (buf)[3]) == PMS_FIFONO_MEP)

 static inline bool pms_is_mamac(char *buf, u32 len)
 {
 	return (len > MDP_HDR_LEN &&
 		EXTRACT_BIT_SET(PMS_FIFONO, buf[3]) == PMS_FIFONO_MDP &&
 		EXTRACT_BIT_SET(PMS_TELID, buf[14]) == PMS_TELID_UNSEGM_MAMAC);
 }

 struct net_dev_channel {
 	bool linked;
 	int ch_id;
 };

 struct net_dev_context {
 	struct most_interface *iface;
 	bool is_mamac;
 	struct net_device *dev;
 	struct net_dev_channel rx;
 	struct net_dev_channel tx;
 	struct list_head list;
 };

 static LIST_HEAD(net_devices);
 static DEFINE_MUTEX(probe_disc_mt); /* ch->linked = true, most_nd_open */
 static DEFINE_SPINLOCK(list_lock); /* list_head, ch->linked = false, dev_hold */
 static struct most_component comp;

 static int skb_to_mamac(const struct sk_buff *skb, struct mbo *mbo)
 {
 	u8 *buff = mbo->virt_address;
 	static const u8 broadcast[] = { 0x03, 0xFF };
 	const u8 *dest_addr = skb->data + 4;
 	const u8 *eth_type = skb->data + 12;
 	unsigned int payload_len = skb->len - ETH_HLEN;
 	unsigned int mdp_len = payload_len + MDP_HDR_LEN;

 	if (mdp_len < skb->len) {
 		pr_err("drop: too large packet! (%u)\n", skb->len);
 		return -EINVAL;
 	}

 	if (mbo->buffer_length < mdp_len) {
 		pr_err("drop: too small buffer! (%d for %d)\n",
 		       mbo->buffer_length, mdp_len);
 		return -EINVAL;
 	}

 	if (skb->len < ETH_HLEN) {
 		pr_err("drop: too small packet! (%d)\n", skb->len);
 		return -EINVAL;
 	}

 	if (dest_addr[0] == 0xFF && dest_addr[1] == 0xFF)
 		dest_addr = broadcast;

 	*buff++ = HB(mdp_len - 2);
 	*buff++ = LB(mdp_len - 2);

 	*buff++ = PMHL;
 	*buff++ = (PMS_FIFONO_MDP << PMS_FIFONO_SHIFT) | PMS_MSGTYPE_DATA;
 	*buff++ = PMS_DEF_PRIO;
 	*buff++ = dest_addr[0];
 	*buff++ = dest_addr[1];
 	*buff++ = 0x00;

 	*buff++ = HB(payload_len + 6);
 	*buff++ = LB(payload_len + 6);

 	/* end of FPH here */

 	*buff++ = eth_type[0];
 	*buff++ = eth_type[1];
 	*buff++ = 0;
 	*buff++ = 0;

 	*buff++ = PMS_TELID_UNSEGM_MAMAC << 4 | HB(payload_len);
 	*buff++ = LB(payload_len);

 	memcpy(buff, skb->data + ETH_HLEN, payload_len);
 	mbo->buffer_length = mdp_len;
 	return 0;
 }

 static int skb_to_mep(const struct sk_buff *skb, struct mbo *mbo)
 {
 	u8 *buff = mbo->virt_address;
 	unsigned int mep_len = skb->len + MEP_HDR_LEN;

 	if (mep_len < skb->len) {
 		pr_err("drop: too large packet! (%u)\n", skb->len);
 		return -EINVAL;
 	}

 	if (mbo->buffer_length < mep_len) {
 		pr_err("drop: too small buffer! (%d for %d)\n",
 		       mbo->buffer_length, mep_len);
 		return -EINVAL;
 	}

 	*buff++ = HB(mep_len - 2);
 	*buff++ = LB(mep_len - 2);

 	*buff++ = PMHL;
 	*buff++ = (PMS_FIFONO_MEP << PMS_FIFONO_SHIFT) | PMS_MSGTYPE_DATA;
 	*buff++ = (MEP_DEF_RETRY << PMS_RETRY_SHIFT) | PMS_DEF_PRIO;
 	*buff++ = 0;
 	*buff++ = 0;
 	*buff++ = 0;

 	memcpy(buff, skb->data, skb->len);
 	mbo->buffer_length = mep_len;
 	return 0;
 }

 static int most_nd_set_mac_address(struct net_device *dev, void *p)
 {
 	struct net_dev_context *nd = netdev_priv(dev);
 	int err = eth_mac_addr(dev, p);

 	if (err)
 		return err;

 	nd->is_mamac =
 		(dev->dev_addr[0] == 0 && dev->dev_addr[1] == 0 &&
 		 dev->dev_addr[2] == 0 && dev->dev_addr[3] == 0);

 	/*
 	 * Set default MTU for the given packet type.
 	 * It is still possible to change MTU using ip tools afterwards.
 	 */
 	dev->mtu = nd->is_mamac ? MAMAC_DATA_LEN : ETH_DATA_LEN;

 	return 0;
 }

 static void on_netinfo(struct most_interface *iface,
 		       unsigned char link_stat, unsigned char *mac_addr);

 static int most_nd_open(struct net_device *dev)
 {
 	struct net_dev_context *nd = netdev_priv(dev);
 	int ret = 0;

 	mutex_lock(&probe_disc_mt);

 	if (most_start_channel(nd->iface, nd->rx.ch_id, &comp)) {
 		netdev_err(dev, "most_start_channel() failed\n");
 		ret = -EBUSY;
 		goto unlock;
 	}

 	if (most_start_channel(nd->iface, nd->tx.ch_id, &comp)) {
 		netdev_err(dev, "most_start_channel() failed\n");
 		most_stop_channel(nd->iface, nd->rx.ch_id, &comp);
 		ret = -EBUSY;
 		goto unlock;
 	}

 	netif_carrier_off(dev);
 	if (is_valid_ether_addr(dev->dev_addr))
 		netif_dormant_off(dev);
 	else
 		netif_dormant_on(dev);
 	netif_wake_queue(dev);
 	if (nd->iface->request_netinfo)
 		nd->iface->request_netinfo(nd->iface, nd->tx.ch_id, on_netinfo);

 unlock:
 	mutex_unlock(&probe_disc_mt);
 	return ret;
 }

 static int most_nd_stop(struct net_device *dev)
 {
 	struct net_dev_context *nd = netdev_priv(dev);

 	netif_stop_queue(dev);
 	if (nd->iface->request_netinfo)
 		nd->iface->request_netinfo(nd->iface, nd->tx.ch_id, NULL);
 	most_stop_channel(nd->iface, nd->rx.ch_id, &comp);
 	most_stop_channel(nd->iface, nd->tx.ch_id, &comp);

 	return 0;
 }

 static netdev_tx_t most_nd_start_xmit(struct sk_buff *skb,
 				      struct net_device *dev)
 {
 	struct net_dev_context *nd = netdev_priv(dev);
 	struct mbo *mbo;
 	int ret;

 	mbo = most_get_mbo(nd->iface, nd->tx.ch_id, &comp);

 	if (!mbo) {
 		netif_stop_queue(dev);
 		dev->stats.tx_fifo_errors++;
 		return NETDEV_TX_BUSY;
 	}

 	if (nd->is_mamac)
 		ret = skb_to_mamac(skb, mbo);
 	else
 		ret = skb_to_mep(skb, mbo);

 	if (ret) {
 		most_put_mbo(mbo);
 		dev->stats.tx_dropped++;
 		kfree_skb(skb);
 		return NETDEV_TX_OK;
 	}

 	most_submit_mbo(mbo);
 	dev->stats.tx_packets++;
 	dev->stats.tx_bytes += skb->len;
 	kfree_skb(skb);
 	return NETDEV_TX_OK;
 }

 static const struct net_device_ops most_nd_ops = {
 	.ndo_open = most_nd_open,
 	.ndo_stop = most_nd_stop,
 	.ndo_start_xmit = most_nd_start_xmit,
 	.ndo_set_mac_address = most_nd_set_mac_address,
 };

 static void most_nd_setup(struct net_device *dev)
 {
 	ether_setup(dev);
 	dev->netdev_ops = &most_nd_ops;
 }

 static struct net_dev_context *get_net_dev(struct most_interface *iface)
 {
 	struct net_dev_context *nd;

 	list_for_each_entry(nd, &net_devices, list)
 		if (nd->iface == iface)
 			return nd;
 	return NULL;
 }

 static struct net_dev_context *get_net_dev_hold(struct most_interface *iface)
 {
 	struct net_dev_context *nd;
 	unsigned long flags;

 	spin_lock_irqsave(&list_lock, flags);
 	nd = get_net_dev(iface);
 	if (nd && nd->rx.linked && nd->tx.linked)
 		dev_hold(nd->dev);
 	else
 		nd = NULL;
 	spin_unlock_irqrestore(&list_lock, flags);
 	return nd;
 }

 static int comp_probe_channel(struct most_interface *iface, int channel_idx,
 			      struct most_channel_config *ccfg, char *name,
 			      char *args)
 {
 	struct net_dev_context *nd;
 	struct net_dev_channel *ch;
 	struct net_device *dev;
 	unsigned long flags;
 	int ret = 0;

 	if (!iface)
 		return -EINVAL;

 	if (ccfg->data_type != MOST_CH_ASYNC)
 		return -EINVAL;

 	mutex_lock(&probe_disc_mt);
 	nd = get_net_dev(iface);
 	if (!nd) {
 		dev = alloc_netdev(sizeof(struct net_dev_context), "meth%d",
 				   NET_NAME_UNKNOWN, most_nd_setup);
 		if (!dev) {
 			ret = -ENOMEM;
 			goto unlock;
 		}

 		nd = netdev_priv(dev);
 		nd->iface = iface;
 		nd->dev = dev;

 		spin_lock_irqsave(&list_lock, flags);
 		list_add(&nd->list, &net_devices);
 		spin_unlock_irqrestore(&list_lock, flags);

 		ch = ccfg->direction == MOST_CH_TX ? &nd->tx : &nd->rx;
 	} else {
 		ch = ccfg->direction == MOST_CH_TX ? &nd->tx : &nd->rx;
 		if (ch->linked) {
 			pr_err("direction is allocated\n");
 			ret = -EINVAL;
 			goto unlock;
 		}

 		if (register_netdev(nd->dev)) {
 			pr_err("register_netdev() failed\n");
 			ret = -EINVAL;
 			goto unlock;
 		}
 	}
 	ch->ch_id = channel_idx;
 	ch->linked = true;

 unlock:
 	mutex_unlock(&probe_disc_mt);
 	return ret;
 }

 static int comp_disconnect_channel(struct most_interface *iface,
 				   int channel_idx)
 {
 	struct net_dev_context *nd;
 	struct net_dev_channel *ch;
 	unsigned long flags;
 	int ret = 0;

 	mutex_lock(&probe_disc_mt);
 	nd = get_net_dev(iface);
 	if (!nd) {
 		ret = -EINVAL;
 		goto unlock;
 	}

 	if (nd->rx.linked && channel_idx == nd->rx.ch_id) {
 		ch = &nd->rx;
 	} else if (nd->tx.linked && channel_idx == nd->tx.ch_id) {
 		ch = &nd->tx;
 	} else {
 		ret = -EINVAL;
 		goto unlock;
 	}

 	if (nd->rx.linked && nd->tx.linked) {
 		spin_lock_irqsave(&list_lock, flags);
 		ch->linked = false;
 		spin_unlock_irqrestore(&list_lock, flags);

 		/*
 		 * do not call most_stop_channel() here, because channels are
 		 * going to be closed in ndo_stop() after unregister_netdev()
 		 */
 		unregister_netdev(nd->dev);
 	} else {
 		spin_lock_irqsave(&list_lock, flags);
 		list_del(&nd->list);
 		spin_unlock_irqrestore(&list_lock, flags);

 		free_netdev(nd->dev);
 	}

 unlock:
 	mutex_unlock(&probe_disc_mt);
 	return ret;
 }

 static int comp_resume_tx_channel(struct most_interface *iface,
 				  int channel_idx)
 {
 	struct net_dev_context *nd;

 	nd = get_net_dev_hold(iface);
 	if (!nd)
 		return 0;

 	if (nd->tx.ch_id != channel_idx)
 		goto put_nd;

 	netif_wake_queue(nd->dev);

 put_nd:
 	dev_put(nd->dev);
 	return 0;
 }

 static int comp_rx_data(struct mbo *mbo)
 {
 	const u32 zero = 0;
 	struct net_dev_context *nd;
 	char *buf = mbo->virt_address;
 	u32 len = mbo->processed_length;
 	struct sk_buff *skb;
 	struct net_device *dev;
 	unsigned int skb_len;
 	int ret = 0;

 	nd = get_net_dev_hold(mbo->ifp);
 	if (!nd)
 		return -EIO;

 	if (nd->rx.ch_id != mbo->hdm_channel_id) {
 		ret = -EIO;
 		goto put_nd;
 	}

 	dev = nd->dev;

 	if (nd->is_mamac) {
 		if (!pms_is_mamac(buf, len)) {
 			ret = -EIO;
 			goto put_nd;
 		}

 		skb = dev_alloc_skb(len - MDP_HDR_LEN + 2 * ETH_ALEN + 2);
 	} else {
 		if (!PMS_IS_MEP(buf, len)) {
 			ret = -EIO;
 			goto put_nd;
 		}

 		skb = dev_alloc_skb(len - MEP_HDR_LEN);
 	}

 	if (!skb) {
 		dev->stats.rx_dropped++;
 		pr_err_once("drop packet: no memory for skb\n");
 		goto out;
 	}

 	skb->dev = dev;

 	if (nd->is_mamac) {
 		/* dest */
 		ether_addr_copy(skb_put(skb, ETH_ALEN), dev->dev_addr);

 		/* src */
 		skb_put_data(skb, &zero, 4);
 		skb_put_data(skb, buf + 5, 2);

 		/* eth type */
 		skb_put_data(skb, buf + 10, 2);

 		buf += MDP_HDR_LEN;
 		len -= MDP_HDR_LEN;
 	} else {
 		buf += MEP_HDR_LEN;
 		len -= MEP_HDR_LEN;
 	}

 	skb_put_data(skb, buf, len);
 	skb->protocol = eth_type_trans(skb, dev);
 	skb_len = skb->len;
 	if (netif_rx(skb) == NET_RX_SUCCESS) {
 		dev->stats.rx_packets++;
 		dev->stats.rx_bytes += skb_len;
 	} else {
 		dev->stats.rx_dropped++;
 	}

 out:
 	most_put_mbo(mbo);

 put_nd:
 	dev_put(nd->dev);
 	return ret;
 }

 static struct most_component comp = {
 	.mod = THIS_MODULE,
 	.name = "net",
 	.probe_channel = comp_probe_channel,
 	.disconnect_channel = comp_disconnect_channel,
 	.tx_completion = comp_resume_tx_channel,
 	.rx_completion = comp_rx_data,
 };

 static int __init most_net_init(void)
 {
 	int err;

 	err = most_register_component(&comp);
 	if (err)
 		return err;
 	err = most_register_configfs_subsys(&comp);
 	if (err) {
 		most_deregister_component(&comp);
 		return err;
 	}
 	return 0;
 }

 static void __exit most_net_exit(void)
 {
 	most_deregister_configfs_subsys(&comp);
 	most_deregister_component(&comp);
 }

 /**
  * on_netinfo - callback for HDM to be informed about HW's MAC
  * @iface: most interface instance
  * @link_stat: link status
  * @mac_addr: MAC address
  */
 static void on_netinfo(struct most_interface *iface,
 		       unsigned char link_stat, unsigned char *mac_addr)
 {
 	struct net_dev_context *nd;
 	struct net_device *dev;
 	const u8 *m = mac_addr;

 	nd = get_net_dev_hold(iface);
 	if (!nd)
 		return;

 	dev = nd->dev;

 	if (link_stat)
 		netif_carrier_on(dev);
 	else
 		netif_carrier_off(dev);

 	if (m && is_valid_ether_addr(m)) {
 		if (!is_valid_ether_addr(dev->dev_addr)) {
 			netdev_info(dev, "set mac %pM\n", m);
 			eth_hw_addr_set(dev, m);
 			netif_dormant_off(dev);
 		} else if (!ether_addr_equal(dev->dev_addr, m)) {
 			netdev_warn(dev, "reject mac %pM\n", m);
 		}
 	}

 	dev_put(nd->dev);
 }

 module_init(most_net_init);
 module_exit(most_net_exit);
 MODULE_LICENSE("GPL");
 MODULE_AUTHOR("Andrey Shvetsov <andrey.shvetsov@k2l.de>");
 MODULE_DESCRIPTION("Networking Component Module for Mostcore");
	// SPDX-License-Identifier: GPL-2.0
	/*
	* net.c - Networking component for Mostcore
	*
	* Copyright (C) 2015, Microchip Technology Germany II GmbH & Co. KG
	*/

	#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt

	#include <linux/module.h>
	#include <linux/netdevice.h>
	#include <linux/etherdevice.h>
	#include <linux/slab.h>
	#include <linux/init.h>
	#include <linux/list.h>
	#include <linux/wait.h>
	#include <linux/kobject.h>
	#include <linux/most.h>

	#define MEP_HDR_LEN 8
	#define MDP_HDR_LEN 16
	#define MAMAC_DATA_LEN (1024 - MDP_HDR_LEN)

	#define PMHL 5

	#define PMS_TELID_UNSEGM_MAMAC 0x0A
	#define PMS_FIFONO_MDP 0x01
	#define PMS_FIFONO_MEP 0x04
	#define PMS_MSGTYPE_DATA 0x04
	#define PMS_DEF_PRIO 0
	#define MEP_DEF_RETRY 15

	#define PMS_FIFONO_MASK 0x07
	#define PMS_FIFONO_SHIFT 3
	#define PMS_RETRY_SHIFT 4
	#define PMS_TELID_MASK 0x0F
	#define PMS_TELID_SHIFT 4

	#define HB(value) ((u8)((u16)(value) >> 8))
	#define LB(value) ((u8)(value))

	#define EXTRACT_BIT_SET(bitset_name, value) \
	(((value) >> bitset_name##_SHIFT) & bitset_name##_MASK)

	#define PMS_IS_MEP(buf, len) \
	((len) > MEP_HDR_LEN && \
	EXTRACT_BIT_SET(PMS_FIFONO, (buf)[3]) == PMS_FIFONO_MEP)

	static inline bool pms_is_mamac(char *buf, u32 len)
	{
	return (len > MDP_HDR_LEN &&
	EXTRACT_BIT_SET(PMS_FIFONO, buf[3]) == PMS_FIFONO_MDP &&
	EXTRACT_BIT_SET(PMS_TELID, buf[14]) == PMS_TELID_UNSEGM_MAMAC);
	}

	struct net_dev_channel {
	bool linked;
	int ch_id;
	};

	struct net_dev_context {
	struct most_interface *iface;
	bool is_mamac;
	struct net_device *dev;
	struct net_dev_channel rx;
	struct net_dev_channel tx;
	struct list_head list;
	};

	static LIST_HEAD(net_devices);
	static DEFINE_MUTEX(probe_disc_mt); /* ch->linked = true, most_nd_open */
	static DEFINE_SPINLOCK(list_lock); /* list_head, ch->linked = false, dev_hold */
	static struct most_component comp;

	static int skb_to_mamac(const struct sk_buff skb, struct mbo mbo)
	{
	u8 *buff = mbo->virt_address;
	static const u8 broadcast[] = { 0x03, 0xFF };
	const u8 *dest_addr = skb->data + 4;
	const u8 *eth_type = skb->data + 12;
	unsigned int payload_len = skb->len - ETH_HLEN;
	unsigned int mdp_len = payload_len + MDP_HDR_LEN;

	if (mdp_len < skb->len) {
	pr_err("drop: too large packet! (%u)\n", skb->len);
	return -EINVAL;
	}

	if (mbo->buffer_length < mdp_len) {
	pr_err("drop: too small buffer! (%d for %d)\n",
	mbo->buffer_length, mdp_len);
	return -EINVAL;
	}

	if (skb->len < ETH_HLEN) {
	pr_err("drop: too small packet! (%d)\n", skb->len);
	return -EINVAL;
	}

	if (dest_addr[0] == 0xFF && dest_addr[1] == 0xFF)
	dest_addr = broadcast;

	*buff++ = HB(mdp_len - 2);
	*buff++ = LB(mdp_len - 2);

	*buff++ = PMHL;
	*buff++ = (PMS_FIFONO_MDP << PMS_FIFONO_SHIFT) \| PMS_MSGTYPE_DATA;
	*buff++ = PMS_DEF_PRIO;
	*buff++ = dest_addr[0];
	*buff++ = dest_addr[1];
	*buff++ = 0x00;

	*buff++ = HB(payload_len + 6);
	*buff++ = LB(payload_len + 6);

	/* end of FPH here */

	*buff++ = eth_type[0];
	*buff++ = eth_type[1];
	*buff++ = 0;
	*buff++ = 0;

	*buff++ = PMS_TELID_UNSEGM_MAMAC << 4 \| HB(payload_len);
	*buff++ = LB(payload_len);

	memcpy(buff, skb->data + ETH_HLEN, payload_len);
	mbo->buffer_length = mdp_len;
	return 0;
	}

	static int skb_to_mep(const struct sk_buff skb, struct mbo mbo)
	{
	u8 *buff = mbo->virt_address;
	unsigned int mep_len = skb->len + MEP_HDR_LEN;

	if (mep_len < skb->len) {
	pr_err("drop: too large packet! (%u)\n", skb->len);
	return -EINVAL;
	}

	if (mbo->buffer_length < mep_len) {
	pr_err("drop: too small buffer! (%d for %d)\n",
	mbo->buffer_length, mep_len);
	return -EINVAL;
	}

	*buff++ = HB(mep_len - 2);
	*buff++ = LB(mep_len - 2);

	*buff++ = PMHL;
	*buff++ = (PMS_FIFONO_MEP << PMS_FIFONO_SHIFT) \| PMS_MSGTYPE_DATA;
	*buff++ = (MEP_DEF_RETRY << PMS_RETRY_SHIFT) \| PMS_DEF_PRIO;
	*buff++ = 0;
	*buff++ = 0;
	*buff++ = 0;

	memcpy(buff, skb->data, skb->len);
	mbo->buffer_length = mep_len;
	return 0;
	}

	static int most_nd_set_mac_address(struct net_device dev, void p)
	{
	struct net_dev_context *nd = netdev_priv(dev);
	int err = eth_mac_addr(dev, p);

	if (err)
	return err;

	nd->is_mamac =
	(dev->dev_addr[0] == 0 && dev->dev_addr[1] == 0 &&
	dev->dev_addr[2] == 0 && dev->dev_addr[3] == 0);

	/*
	* Set default MTU for the given packet type.
	* It is still possible to change MTU using ip tools afterwards.
	*/
	dev->mtu = nd->is_mamac ? MAMAC_DATA_LEN : ETH_DATA_LEN;

	return 0;
	}

	static void on_netinfo(struct most_interface *iface,
	unsigned char link_stat, unsigned char *mac_addr);

	static int most_nd_open(struct net_device *dev)
	{
	struct net_dev_context *nd = netdev_priv(dev);
	int ret = 0;

	mutex_lock(&probe_disc_mt);

	if (most_start_channel(nd->iface, nd->rx.ch_id, &comp)) {
	netdev_err(dev, "most_start_channel() failed\n");
	ret = -EBUSY;
	goto unlock;
	}

	if (most_start_channel(nd->iface, nd->tx.ch_id, &comp)) {
	netdev_err(dev, "most_start_channel() failed\n");
	most_stop_channel(nd->iface, nd->rx.ch_id, &comp);
	ret = -EBUSY;
	goto unlock;
	}

	netif_carrier_off(dev);
	if (is_valid_ether_addr(dev->dev_addr))
	netif_dormant_off(dev);
	else
	netif_dormant_on(dev);
	netif_wake_queue(dev);
	if (nd->iface->request_netinfo)
	nd->iface->request_netinfo(nd->iface, nd->tx.ch_id, on_netinfo);

	unlock:
	mutex_unlock(&probe_disc_mt);
	return ret;
	}

	static int most_nd_stop(struct net_device *dev)
	{
	struct net_dev_context *nd = netdev_priv(dev);

	netif_stop_queue(dev);
	if (nd->iface->request_netinfo)
	nd->iface->request_netinfo(nd->iface, nd->tx.ch_id, NULL);
	most_stop_channel(nd->iface, nd->rx.ch_id, &comp);
	most_stop_channel(nd->iface, nd->tx.ch_id, &comp);

	return 0;
	}

	static netdev_tx_t most_nd_start_xmit(struct sk_buff *skb,
	struct net_device *dev)
	{
	struct net_dev_context *nd = netdev_priv(dev);
	struct mbo *mbo;
	int ret;

	mbo = most_get_mbo(nd->iface, nd->tx.ch_id, &comp);

	if (!mbo) {
	netif_stop_queue(dev);
	dev->stats.tx_fifo_errors++;
	return NETDEV_TX_BUSY;
	}

	if (nd->is_mamac)
	ret = skb_to_mamac(skb, mbo);
	else
	ret = skb_to_mep(skb, mbo);

	if (ret) {
	most_put_mbo(mbo);
	dev->stats.tx_dropped++;
	kfree_skb(skb);
	return NETDEV_TX_OK;
	}

	most_submit_mbo(mbo);
	dev->stats.tx_packets++;
	dev->stats.tx_bytes += skb->len;
	kfree_skb(skb);
	return NETDEV_TX_OK;
	}

	static const struct net_device_ops most_nd_ops = {
	.ndo_open = most_nd_open,
	.ndo_stop = most_nd_stop,
	.ndo_start_xmit = most_nd_start_xmit,
	.ndo_set_mac_address = most_nd_set_mac_address,
	};

	static void most_nd_setup(struct net_device *dev)
	{
	ether_setup(dev);
	dev->netdev_ops = &most_nd_ops;
	}

	static struct net_dev_context get_net_dev(struct most_interface iface)
	{
	struct net_dev_context *nd;

	list_for_each_entry(nd, &net_devices, list)
	if (nd->iface == iface)
	return nd;
	return NULL;
	}

	static struct net_dev_context get_net_dev_hold(struct most_interface iface)
	{
	struct net_dev_context *nd;
	unsigned long flags;

	spin_lock_irqsave(&list_lock, flags);
	nd = get_net_dev(iface);
	if (nd && nd->rx.linked && nd->tx.linked)
	dev_hold(nd->dev);
	else
	nd = NULL;
	spin_unlock_irqrestore(&list_lock, flags);
	return nd;
	}

	static int comp_probe_channel(struct most_interface *iface, int channel_idx,
	struct most_channel_config ccfg, char name,
	char *args)
	{
	struct net_dev_context *nd;
	struct net_dev_channel *ch;
	struct net_device *dev;
	unsigned long flags;
	int ret = 0;

	if (!iface)
	return -EINVAL;

	if (ccfg->data_type != MOST_CH_ASYNC)
	return -EINVAL;

	mutex_lock(&probe_disc_mt);
	nd = get_net_dev(iface);
	if (!nd) {
	dev = alloc_netdev(sizeof(struct net_dev_context), "meth%d",
	NET_NAME_UNKNOWN, most_nd_setup);
	if (!dev) {
	ret = -ENOMEM;
	goto unlock;
	}

	nd = netdev_priv(dev);
	nd->iface = iface;
	nd->dev = dev;

	spin_lock_irqsave(&list_lock, flags);
	list_add(&nd->list, &net_devices);
	spin_unlock_irqrestore(&list_lock, flags);

	ch = ccfg->direction == MOST_CH_TX ? &nd->tx : &nd->rx;
	} else {
	ch = ccfg->direction == MOST_CH_TX ? &nd->tx : &nd->rx;
	if (ch->linked) {
	pr_err("direction is allocated\n");
	ret = -EINVAL;
	goto unlock;
	}

	if (register_netdev(nd->dev)) {
	pr_err("register_netdev() failed\n");
	ret = -EINVAL;
	goto unlock;
	}
	}
	ch->ch_id = channel_idx;
	ch->linked = true;

	unlock:
	mutex_unlock(&probe_disc_mt);
	return ret;
	}

	static int comp_disconnect_channel(struct most_interface *iface,
	int channel_idx)
	{
	struct net_dev_context *nd;
	struct net_dev_channel *ch;
	unsigned long flags;
	int ret = 0;

	mutex_lock(&probe_disc_mt);
	nd = get_net_dev(iface);
	if (!nd) {
	ret = -EINVAL;
	goto unlock;
	}

	if (nd->rx.linked && channel_idx == nd->rx.ch_id) {
	ch = &nd->rx;
	} else if (nd->tx.linked && channel_idx == nd->tx.ch_id) {
	ch = &nd->tx;
	} else {
	ret = -EINVAL;
	goto unlock;
	}

	if (nd->rx.linked && nd->tx.linked) {
	spin_lock_irqsave(&list_lock, flags);
	ch->linked = false;
	spin_unlock_irqrestore(&list_lock, flags);

	/*
	* do not call most_stop_channel() here, because channels are
	* going to be closed in ndo_stop() after unregister_netdev()
	*/
	unregister_netdev(nd->dev);
	} else {
	spin_lock_irqsave(&list_lock, flags);
	list_del(&nd->list);
	spin_unlock_irqrestore(&list_lock, flags);

	free_netdev(nd->dev);
	}

	unlock:
	mutex_unlock(&probe_disc_mt);
	return ret;
	}

	static int comp_resume_tx_channel(struct most_interface *iface,
	int channel_idx)
	{
	struct net_dev_context *nd;

	nd = get_net_dev_hold(iface);
	if (!nd)
	return 0;

	if (nd->tx.ch_id != channel_idx)
	goto put_nd;

	netif_wake_queue(nd->dev);

	put_nd:
	dev_put(nd->dev);
	return 0;
	}

	static int comp_rx_data(struct mbo *mbo)
	{
	const u32 zero = 0;
	struct net_dev_context *nd;
	char *buf = mbo->virt_address;
	u32 len = mbo->processed_length;
	struct sk_buff *skb;
	struct net_device *dev;
	unsigned int skb_len;
	int ret = 0;

	nd = get_net_dev_hold(mbo->ifp);
	if (!nd)
	return -EIO;

	if (nd->rx.ch_id != mbo->hdm_channel_id) {
	ret = -EIO;
	goto put_nd;
	}

	dev = nd->dev;

	if (nd->is_mamac) {
	if (!pms_is_mamac(buf, len)) {
	ret = -EIO;
	goto put_nd;
	}

	skb = dev_alloc_skb(len - MDP_HDR_LEN + 2 * ETH_ALEN + 2);
	} else {
	if (!PMS_IS_MEP(buf, len)) {
	ret = -EIO;
	goto put_nd;
	}

	skb = dev_alloc_skb(len - MEP_HDR_LEN);
	}

	if (!skb) {
	dev->stats.rx_dropped++;
	pr_err_once("drop packet: no memory for skb\n");
	goto out;
	}

	skb->dev = dev;

	if (nd->is_mamac) {
	/* dest */
	ether_addr_copy(skb_put(skb, ETH_ALEN), dev->dev_addr);

	/* src */
	skb_put_data(skb, &zero, 4);
	skb_put_data(skb, buf + 5, 2);

	/* eth type */
	skb_put_data(skb, buf + 10, 2);

	buf += MDP_HDR_LEN;
	len -= MDP_HDR_LEN;
	} else {
	buf += MEP_HDR_LEN;
	len -= MEP_HDR_LEN;
	}

	skb_put_data(skb, buf, len);
	skb->protocol = eth_type_trans(skb, dev);
	skb_len = skb->len;
	if (netif_rx(skb) == NET_RX_SUCCESS) {
	dev->stats.rx_packets++;
	dev->stats.rx_bytes += skb_len;
	} else {
	dev->stats.rx_dropped++;
	}

	out:
	most_put_mbo(mbo);

	put_nd:
	dev_put(nd->dev);
	return ret;
	}

	static struct most_component comp = {
	.mod = THIS_MODULE,
	.name = "net",
	.probe_channel = comp_probe_channel,
	.disconnect_channel = comp_disconnect_channel,
	.tx_completion = comp_resume_tx_channel,
	.rx_completion = comp_rx_data,
	};

	static int __init most_net_init(void)
	{
	int err;

	err = most_register_component(&comp);
	if (err)
	return err;
	err = most_register_configfs_subsys(&comp);
	if (err) {
	most_deregister_component(&comp);
	return err;
	}
	return 0;
	}

	static void __exit most_net_exit(void)
	{
	most_deregister_configfs_subsys(&comp);
	most_deregister_component(&comp);
	}

	/**
	* on_netinfo - callback for HDM to be informed about HW's MAC
	* @iface: most interface instance
	* @link_stat: link status
	* @mac_addr: MAC address
	*/
	static void on_netinfo(struct most_interface *iface,
	unsigned char link_stat, unsigned char *mac_addr)
	{
	struct net_dev_context *nd;
	struct net_device *dev;
	const u8 *m = mac_addr;

	nd = get_net_dev_hold(iface);
	if (!nd)
	return;

	dev = nd->dev;

	if (link_stat)
	netif_carrier_on(dev);
	else
	netif_carrier_off(dev);

	if (m && is_valid_ether_addr(m)) {
	if (!is_valid_ether_addr(dev->dev_addr)) {
	netdev_info(dev, "set mac %pM\n", m);
	eth_hw_addr_set(dev, m);
	netif_dormant_off(dev);
	} else if (!ether_addr_equal(dev->dev_addr, m)) {
	netdev_warn(dev, "reject mac %pM\n", m);
	}
	}

	dev_put(nd->dev);
	}

	module_init(most_net_init);
	module_exit(most_net_exit);
	MODULE_LICENSE("GPL");
	MODULE_AUTHOR("Andrey Shvetsov <andrey.shvetsov@k2l.de>");
	MODULE_DESCRIPTION("Networking Component Module for Mostcore");