[deliverable/linux.git] / drivers / net / hyperv / netvsc_drv.c

/*
 * Copyright (c) 2009, Microsoft Corporation.
 *
 * This program is free software; you can redistribute it and/or modify it
 * under the terms and conditions of the GNU General Public License,
 * version 2, as published by the Free Software Foundation.
 *
 * This program is distributed in the hope it will be useful, but WITHOUT
 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
 * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
 * more details.
 *
 * You should have received a copy of the GNU General Public License along with
 * this program; if not, write to the Free Software Foundation, Inc., 59 Temple
 * Place - Suite 330, Boston, MA 02111-1307 USA.
 *
 * Authors:
 *   Haiyang Zhang <haiyangz@microsoft.com>
 *   Hank Janssen  <hjanssen@microsoft.com>
 */
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt

#include <linux/init.h>
#include <linux/atomic.h>
#include <linux/module.h>
#include <linux/highmem.h>
#include <linux/device.h>
#include <linux/io.h>
#include <linux/delay.h>
#include <linux/netdevice.h>
#include <linux/inetdevice.h>
#include <linux/etherdevice.h>
#include <linux/skbuff.h>
#include <linux/in.h>
#include <linux/slab.h>
#include <net/arp.h>
#include <net/route.h>
#include <net/sock.h>
#include <net/pkt_sched.h>

#include "hyperv_net.h"

struct net_device_context {
	/* point back to our device context */
	struct hv_device *device_ctx;
	struct delayed_work dwork;
};


static int ring_size = 128;
module_param(ring_size, int, S_IRUGO);
MODULE_PARM_DESC(ring_size, "Ring buffer size (# of pages)");

struct set_multicast_work {
	struct work_struct work;
	struct net_device *net;
};

static void do_set_multicast(struct work_struct *w)
{
	struct set_multicast_work *swk =
		container_of(w, struct set_multicast_work, work);
	struct net_device *net = swk->net;

	struct net_device_context *ndevctx = netdev_priv(net);
	struct netvsc_device *nvdev;
	struct rndis_device *rdev;

	nvdev = hv_get_drvdata(ndevctx->device_ctx);
	if (nvdev == NULL)
		goto out;

	rdev = nvdev->extension;
	if (rdev == NULL)
		goto out;

	if (net->flags & IFF_PROMISC)
		rndis_filter_set_packet_filter(rdev,
			NDIS_PACKET_TYPE_PROMISCUOUS);
	else
		rndis_filter_set_packet_filter(rdev,
			NDIS_PACKET_TYPE_BROADCAST |
			NDIS_PACKET_TYPE_ALL_MULTICAST |
			NDIS_PACKET_TYPE_DIRECTED);

out:
	kfree(w);
}

static void netvsc_set_multicast_list(struct net_device *net)
{
	struct set_multicast_work *swk =
		kmalloc(sizeof(struct set_multicast_work), GFP_ATOMIC);
	if (swk == NULL)
		return;

	swk->net = net;
	INIT_WORK(&swk->work, do_set_multicast);
	schedule_work(&swk->work);
}

static int netvsc_open(struct net_device *net)
{
	struct net_device_context *net_device_ctx = netdev_priv(net);
	struct hv_device *device_obj = net_device_ctx->device_ctx;
	int ret = 0;

	/* Open up the device */
	ret = rndis_filter_open(device_obj);
	if (ret != 0) {
		netdev_err(net, "unable to open device (ret %d).\n", ret);
		return ret;
	}

	netif_start_queue(net);

	return ret;
}

static int netvsc_close(struct net_device *net)
{
	struct net_device_context *net_device_ctx = netdev_priv(net);
	struct hv_device *device_obj = net_device_ctx->device_ctx;
	int ret;

	netif_tx_disable(net);

	ret = rndis_filter_close(device_obj);
	if (ret != 0)
		netdev_err(net, "unable to close device (ret %d).\n", ret);

	return ret;
}

static void netvsc_xmit_completion(void *context)
{
	struct hv_netvsc_packet *packet = (struct hv_netvsc_packet *)context;
	struct sk_buff *skb = (struct sk_buff *)
		(unsigned long)packet->completion.send.send_completion_tid;

	kfree(packet);

	if (skb)
		dev_kfree_skb_any(skb);
}

static int netvsc_start_xmit(struct sk_buff *skb, struct net_device *net)
{
	struct net_device_context *net_device_ctx = netdev_priv(net);
	struct hv_netvsc_packet *packet;
	int ret;
	unsigned int i, num_pages, npg_data;

	/* Add multipages for skb->data and additional 2 for RNDIS */
	npg_data = (((unsigned long)skb->data + skb_headlen(skb) - 1)
		>> PAGE_SHIFT) - ((unsigned long)skb->data >> PAGE_SHIFT) + 1;
	num_pages = skb_shinfo(skb)->nr_frags + npg_data + 2;

	/* Allocate a netvsc packet based on # of frags. */
	packet = kzalloc(sizeof(struct hv_netvsc_packet) +
			 (num_pages * sizeof(struct hv_page_buffer)) +
			 sizeof(struct rndis_filter_packet), GFP_ATOMIC);
	if (!packet) {
		/* out of memory, drop packet */
		netdev_err(net, "unable to allocate hv_netvsc_packet\n");

		dev_kfree_skb(skb);
		net->stats.tx_dropped++;
		return NETDEV_TX_BUSY;
	}

	packet->extension = (void *)(unsigned long)packet +
				sizeof(struct hv_netvsc_packet) +
				    (num_pages * sizeof(struct hv_page_buffer));

	/* If the rndis msg goes beyond 1 page, we will add 1 later */
	packet->page_buf_cnt = num_pages - 1;

	/* Initialize it from the skb */
	packet->total_data_buflen = skb->len;

	/* Start filling in the page buffers starting after RNDIS buffer. */
	packet->page_buf[1].pfn = virt_to_phys(skb->data) >> PAGE_SHIFT;
	packet->page_buf[1].offset
		= (unsigned long)skb->data & (PAGE_SIZE - 1);
	if (npg_data == 1)
		packet->page_buf[1].len = skb_headlen(skb);
	else
		packet->page_buf[1].len = PAGE_SIZE
			- packet->page_buf[1].offset;

	for (i = 2; i <= npg_data; i++) {
		packet->page_buf[i].pfn = virt_to_phys(skb->data
			+ PAGE_SIZE * (i-1)) >> PAGE_SHIFT;
		packet->page_buf[i].offset = 0;
		packet->page_buf[i].len = PAGE_SIZE;
	}
	if (npg_data > 1)
		packet->page_buf[npg_data].len = (((unsigned long)skb->data
			+ skb_headlen(skb) - 1) & (PAGE_SIZE - 1)) + 1;

	/* Additional fragments are after SKB data */
	for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
		const skb_frag_t *f = &skb_shinfo(skb)->frags[i];

		packet->page_buf[i+npg_data+1].pfn =
			page_to_pfn(skb_frag_page(f));
		packet->page_buf[i+npg_data+1].offset = f->page_offset;
		packet->page_buf[i+npg_data+1].len = skb_frag_size(f);
	}

	/* Set the completion routine */
	packet->completion.send.send_completion = netvsc_xmit_completion;
	packet->completion.send.send_completion_ctx = packet;
	packet->completion.send.send_completion_tid = (unsigned long)skb;

	ret = rndis_filter_send(net_device_ctx->device_ctx,
				  packet);
	if (ret == 0) {
		net->stats.tx_bytes += skb->len;
		net->stats.tx_packets++;
	} else {
		/* we are shutting down or bus overloaded, just drop packet */
		net->stats.tx_dropped++;
		kfree(packet);
		dev_kfree_skb_any(skb);
	}

	return ret ? NETDEV_TX_BUSY : NETDEV_TX_OK;
}

/*
 * netvsc_linkstatus_callback - Link up/down notification
 */
void netvsc_linkstatus_callback(struct hv_device *device_obj,
				       unsigned int status)
{
	struct net_device *net;
	struct net_device_context *ndev_ctx;
	struct netvsc_device *net_device;

	net_device = hv_get_drvdata(device_obj);
	net = net_device->ndev;

	if (!net) {
		netdev_err(net, "got link status but net device "
				"not initialized yet\n");
		return;
	}

	if (status == 1) {
		netif_carrier_on(net);
		netif_wake_queue(net);
		ndev_ctx = netdev_priv(net);
		schedule_delayed_work(&ndev_ctx->dwork, 0);
		schedule_delayed_work(&ndev_ctx->dwork, msecs_to_jiffies(20));
	} else {
		netif_carrier_off(net);
		netif_tx_disable(net);
	}
}

/*
 * netvsc_recv_callback -  Callback when we receive a packet from the
 * "wire" on the specified device.
 */
int netvsc_recv_callback(struct hv_device *device_obj,
				struct hv_netvsc_packet *packet)
{
	struct net_device *net;
	struct sk_buff *skb;

	net = ((struct netvsc_device *)hv_get_drvdata(device_obj))->ndev;
	if (!net) {
		netdev_err(net, "got receive callback but net device"
			" not initialized yet\n");
		return 0;
	}

	/* Allocate a skb - TODO direct I/O to pages? */
	skb = netdev_alloc_skb_ip_align(net, packet->total_data_buflen);
	if (unlikely(!skb)) {
		++net->stats.rx_dropped;
		return 0;
	}

	/*
	 * Copy to skb. This copy is needed here since the memory pointed by
	 * hv_netvsc_packet cannot be deallocated
	 */
	memcpy(skb_put(skb, packet->total_data_buflen), packet->data,
		packet->total_data_buflen);

	skb->protocol = eth_type_trans(skb, net);
	skb->ip_summed = CHECKSUM_NONE;

	net->stats.rx_packets++;
	net->stats.rx_bytes += packet->total_data_buflen;

	/*
	 * Pass the skb back up. Network stack will deallocate the skb when it
	 * is done.
	 * TODO - use NAPI?
	 */
	netif_rx(skb);

	return 0;
}

static void netvsc_get_drvinfo(struct net_device *net,
			       struct ethtool_drvinfo *info)
{
	strcpy(info->driver, "hv_netvsc");
	strcpy(info->version, HV_DRV_VERSION);
	strcpy(info->fw_version, "N/A");
}

static int netvsc_change_mtu(struct net_device *ndev, int mtu)
{
	struct net_device_context *ndevctx = netdev_priv(ndev);
	struct hv_device *hdev =  ndevctx->device_ctx;
	struct netvsc_device *nvdev = hv_get_drvdata(hdev);
	struct netvsc_device_info device_info;
	int limit = ETH_DATA_LEN;

	if (nvdev == NULL || nvdev->destroy)
		return -ENODEV;

	if (nvdev->nvsp_version == NVSP_PROTOCOL_VERSION_2)
		limit = NETVSC_MTU;

	if (mtu < 68 || mtu > limit)
		return -EINVAL;

	nvdev->start_remove = true;
	cancel_delayed_work_sync(&ndevctx->dwork);
	netif_tx_disable(ndev);
	rndis_filter_device_remove(hdev);

	ndev->mtu = mtu;

	ndevctx->device_ctx = hdev;
	hv_set_drvdata(hdev, ndev);
	device_info.ring_size = ring_size;
	rndis_filter_device_add(hdev, &device_info);
	netif_wake_queue(ndev);

	return 0;
}

static const struct ethtool_ops ethtool_ops = {
	.get_drvinfo	= netvsc_get_drvinfo,
	.get_link	= ethtool_op_get_link,
};

static const struct net_device_ops device_ops = {
	.ndo_open =			netvsc_open,
	.ndo_stop =			netvsc_close,
	.ndo_start_xmit =		netvsc_start_xmit,
	.ndo_set_rx_mode =		netvsc_set_multicast_list,
	.ndo_change_mtu =		netvsc_change_mtu,
	.ndo_validate_addr =		eth_validate_addr,
	.ndo_set_mac_address =		eth_mac_addr,
};

/*
 * Send GARP packet to network peers after migrations.
 * After Quick Migration, the network is not immediately operational in the
 * current context when receiving RNDIS_STATUS_MEDIA_CONNECT event. So, add
 * another netif_notify_peers() into a delayed work, otherwise GARP packet
 * will not be sent after quick migration, and cause network disconnection.
 */
static void netvsc_send_garp(struct work_struct *w)
{
	struct net_device_context *ndev_ctx;
	struct net_device *net;
	struct netvsc_device *net_device;

	ndev_ctx = container_of(w, struct net_device_context, dwork.work);
	net_device = hv_get_drvdata(ndev_ctx->device_ctx);
	net = net_device->ndev;
	netif_notify_peers(net);
}


static int netvsc_probe(struct hv_device *dev,
			const struct hv_vmbus_device_id *dev_id)
{
	struct net_device *net = NULL;
	struct net_device_context *net_device_ctx;
	struct netvsc_device_info device_info;
	int ret;

	net = alloc_etherdev(sizeof(struct net_device_context));
	if (!net)
		return -ENOMEM;

	/* Set initial state */
	netif_carrier_off(net);

	net_device_ctx = netdev_priv(net);
	net_device_ctx->device_ctx = dev;
	hv_set_drvdata(dev, net);
	INIT_DELAYED_WORK(&net_device_ctx->dwork, netvsc_send_garp);

	net->netdev_ops = &device_ops;

	/* TODO: Add GSO and Checksum offload */
	net->hw_features = NETIF_F_SG;
	net->features = NETIF_F_SG;

	SET_ETHTOOL_OPS(net, &ethtool_ops);
	SET_NETDEV_DEV(net, &dev->device);

	ret = register_netdev(net);
	if (ret != 0) {
		pr_err("Unable to register netdev.\n");
		free_netdev(net);
		goto out;
	}

	/* Notify the netvsc driver of the new device */
	device_info.ring_size = ring_size;
	ret = rndis_filter_device_add(dev, &device_info);
	if (ret != 0) {
		netdev_err(net, "unable to add netvsc device (ret %d)\n", ret);
		unregister_netdev(net);
		free_netdev(net);
		hv_set_drvdata(dev, NULL);
		return ret;
	}
	memcpy(net->dev_addr, device_info.mac_adr, ETH_ALEN);

	netif_carrier_on(net);

out:
	return ret;
}

static int netvsc_remove(struct hv_device *dev)
{
	struct net_device *net;
	struct net_device_context *ndev_ctx;
	struct netvsc_device *net_device;

	net_device = hv_get_drvdata(dev);
	net = net_device->ndev;

	if (net == NULL) {
		dev_err(&dev->device, "No net device to remove\n");
		return 0;
	}

	net_device->start_remove = true;

	ndev_ctx = netdev_priv(net);
	cancel_delayed_work_sync(&ndev_ctx->dwork);

	/* Stop outbound asap */
	netif_tx_disable(net);

	unregister_netdev(net);

	/*
	 * Call to the vsc driver to let it know that the device is being
	 * removed
	 */
	rndis_filter_device_remove(dev);

	free_netdev(net);
	return 0;
}

static const struct hv_vmbus_device_id id_table[] = {
	/* Network guid */
	{ VMBUS_DEVICE(0x63, 0x51, 0x61, 0xF8, 0x3E, 0xDF, 0xc5, 0x46,
		       0x91, 0x3F, 0xF2, 0xD2, 0xF9, 0x65, 0xED, 0x0E) },
	{ },
};

MODULE_DEVICE_TABLE(vmbus, id_table);

/* The one and only one */
static struct  hv_driver netvsc_drv = {
	.name = "netvsc",
	.id_table = id_table,
	.probe = netvsc_probe,
	.remove = netvsc_remove,
};

static void __exit netvsc_drv_exit(void)
{
	vmbus_driver_unregister(&netvsc_drv);
}

static int __init netvsc_drv_init(void)
{
	return vmbus_driver_register(&netvsc_drv);
}

MODULE_LICENSE("GPL");
MODULE_VERSION(HV_DRV_VERSION);
MODULE_DESCRIPTION("Microsoft Hyper-V network driver");

module_init(netvsc_drv_init);
module_exit(netvsc_drv_exit);
Commit	Line	Data
fceaf24a	1	/*
fceaf24a HJ	2	* Copyright (c) 2009, Microsoft Corporation.
	3	*
	4	* This program is free software; you can redistribute it and/or modify it
	5	* under the terms and conditions of the GNU General Public License,
	6	* version 2, as published by the Free Software Foundation.
	7	*
	8	* This program is distributed in the hope it will be useful, but WITHOUT
	9	* ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
	10	* FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
	11	* more details.
	12	*
	13	* You should have received a copy of the GNU General Public License along with
	14	* this program; if not, write to the Free Software Foundation, Inc., 59 Temple
	15	* Place - Suite 330, Boston, MA 02111-1307 USA.
	16	*
	17	* Authors:
d0e94d17	18	* Haiyang Zhang <haiyangz@microsoft.com>
fceaf24a	19	* Hank Janssen <hjanssen@microsoft.com>
fceaf24a	20	*/
eb335bc4 HJ	21	#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
eb335bc4 HJ	22
fceaf24a	23	#include <linux/init.h>
9079ce69	24	#include <linux/atomic.h>
fceaf24a HJ	25	#include <linux/module.h>
	26	#include <linux/highmem.h>
	27	#include <linux/device.h>
fceaf24a	28	#include <linux/io.h>
fceaf24a HJ	29	#include <linux/delay.h>
	30	#include <linux/netdevice.h>
	31	#include <linux/inetdevice.h>
	32	#include <linux/etherdevice.h>
	33	#include <linux/skbuff.h>
	34	#include <linux/in.h>
5a0e3ad6	35	#include <linux/slab.h>
fceaf24a HJ	36	#include <net/arp.h>
	37	#include <net/route.h>
	38	#include <net/sock.h>
	39	#include <net/pkt_sched.h>
3f335ea2	40
5ca7252a	41	#include "hyperv_net.h"
fceaf24a	42
fceaf24a	43	struct net_device_context {
02fafbc6	44	/* point back to our device context */
6bad88da	45	struct hv_device *device_ctx;
122a5f64	46	struct delayed_work dwork;
fceaf24a HJ	47	};
fceaf24a HJ	48
fceaf24a	49
99c8da0f	50	static int ring_size = 128;
450d7a4b SH	51	module_param(ring_size, int, S_IRUGO);
450d7a4b SH	52	MODULE_PARM_DESC(ring_size, "Ring buffer size (# of pages)");
fceaf24a	53
d426b2e3 HZ	54	struct set_multicast_work {
	55	struct work_struct work;
	56	struct net_device *net;
	57	};
	58
	59	static void do_set_multicast(struct work_struct *w)
	60	{
	61	struct set_multicast_work *swk =
	62	container_of(w, struct set_multicast_work, work);
	63	struct net_device *net = swk->net;
	64
	65	struct net_device_context *ndevctx = netdev_priv(net);
	66	struct netvsc_device *nvdev;
	67	struct rndis_device *rdev;
	68
	69	nvdev = hv_get_drvdata(ndevctx->device_ctx);
	70	if (nvdev == NULL)
c11bf1c8	71	goto out;
d426b2e3 HZ	72
	73	rdev = nvdev->extension;
	74	if (rdev == NULL)
c11bf1c8	75	goto out;
d426b2e3 HZ	76
	77	if (net->flags & IFF_PROMISC)
	78	rndis_filter_set_packet_filter(rdev,
	79	NDIS_PACKET_TYPE_PROMISCUOUS);
	80	else
	81	rndis_filter_set_packet_filter(rdev,
	82	NDIS_PACKET_TYPE_BROADCAST \|
	83	NDIS_PACKET_TYPE_ALL_MULTICAST \|
	84	NDIS_PACKET_TYPE_DIRECTED);
	85
c11bf1c8	86	out:
d426b2e3 HZ	87	kfree(w);
	88	}
	89
4e9bfefa	90	static void netvsc_set_multicast_list(struct net_device *net)
fceaf24a	91	{
d426b2e3 HZ	92	struct set_multicast_work *swk =
	93	kmalloc(sizeof(struct set_multicast_work), GFP_ATOMIC);
	94	if (swk == NULL)
	95	return;
	96
	97	swk->net = net;
	98	INIT_WORK(&swk->work, do_set_multicast);
	99	schedule_work(&swk->work);
fceaf24a HJ	100	}
fceaf24a HJ	101
fceaf24a HJ	102	static int netvsc_open(struct net_device *net)
fceaf24a HJ	103	{
fceaf24a	104	struct net_device_context *net_device_ctx = netdev_priv(net);
6bad88da	105	struct hv_device *device_obj = net_device_ctx->device_ctx;
02fafbc6	106	int ret = 0;
fceaf24a	107
d515d0ff HZ	108	/* Open up the device */
	109	ret = rndis_filter_open(device_obj);
	110	if (ret != 0) {
	111	netdev_err(net, "unable to open device (ret %d).\n", ret);
	112	return ret;
fceaf24a HJ	113	}
fceaf24a HJ	114
d515d0ff HZ	115	netif_start_queue(net);
d515d0ff HZ	116
fceaf24a HJ	117	return ret;
	118	}
	119
fceaf24a HJ	120	static int netvsc_close(struct net_device *net)
fceaf24a HJ	121	{
fceaf24a	122	struct net_device_context *net_device_ctx = netdev_priv(net);
6bad88da	123	struct hv_device *device_obj = net_device_ctx->device_ctx;
02fafbc6	124	int ret;
fceaf24a	125
0a282538	126	netif_tx_disable(net);
fceaf24a	127
9c26aa0d	128	ret = rndis_filter_close(device_obj);
fceaf24a	129	if (ret != 0)
eb335bc4	130	netdev_err(net, "unable to close device (ret %d).\n", ret);
fceaf24a	131
fceaf24a HJ	132	return ret;
	133	}
	134
fceaf24a HJ	135	static void netvsc_xmit_completion(void *context)
fceaf24a HJ	136	{
4193d4f4	137	struct hv_netvsc_packet packet = (struct hv_netvsc_packet )context;
02fafbc6	138	struct sk_buff skb = (struct sk_buff )
72a2f5bd	139	(unsigned long)packet->completion.send.send_completion_tid;
fceaf24a	140
fceaf24a HJ	141	kfree(packet);
fceaf24a HJ	142
1d06825b	143	if (skb)
b220f5f9	144	dev_kfree_skb_any(skb);
fceaf24a HJ	145	}
fceaf24a HJ	146
02fafbc6	147	static int netvsc_start_xmit(struct sk_buff skb, struct net_device net)
fceaf24a	148	{
fceaf24a	149	struct net_device_context *net_device_ctx = netdev_priv(net);
4193d4f4	150	struct hv_netvsc_packet *packet;
02fafbc6	151	int ret;
4d447c9a	152	unsigned int i, num_pages, npg_data;
fceaf24a	153
c31c151b	154	/* Add multipages for skb->data and additional 2 for RNDIS */
4d447c9a HZ	155	npg_data = (((unsigned long)skb->data + skb_headlen(skb) - 1)
4d447c9a HZ	156	>> PAGE_SHIFT) - ((unsigned long)skb->data >> PAGE_SHIFT) + 1;
c31c151b	157	num_pages = skb_shinfo(skb)->nr_frags + npg_data + 2;
fceaf24a	158
454f18a9	159	/* Allocate a netvsc packet based on # of frags. */
02fafbc6	160	packet = kzalloc(sizeof(struct hv_netvsc_packet) +
6048718d	161	(num_pages * sizeof(struct hv_page_buffer)) +
f8ba8c70	162	sizeof(struct rndis_filter_packet), GFP_ATOMIC);
02fafbc6	163	if (!packet) {
bf769375	164	/* out of memory, drop packet */
eb335bc4	165	netdev_err(net, "unable to allocate hv_netvsc_packet\n");
b220f5f9 SH	166
	167	dev_kfree_skb(skb);
	168	net->stats.tx_dropped++;
bf769375	169	return NETDEV_TX_BUSY;
fceaf24a HJ	170	}
fceaf24a HJ	171
72a2f5bd	172	packet->extension = (void *)(unsigned long)packet +
02fafbc6	173	sizeof(struct hv_netvsc_packet) +
6048718d	174	(num_pages * sizeof(struct hv_page_buffer));
fceaf24a	175
c31c151b HZ	176	/* If the rndis msg goes beyond 1 page, we will add 1 later */
c31c151b HZ	177	packet->page_buf_cnt = num_pages - 1;
fceaf24a	178
454f18a9	179	/* Initialize it from the skb */
4d447c9a	180	packet->total_data_buflen = skb->len;
fceaf24a	181
6048718d	182	/* Start filling in the page buffers starting after RNDIS buffer. */
ca623ad3 HZ	183	packet->page_buf[1].pfn = virt_to_phys(skb->data) >> PAGE_SHIFT;
ca623ad3 HZ	184	packet->page_buf[1].offset
6048718d	185	= (unsigned long)skb->data & (PAGE_SIZE - 1);
4d447c9a HZ	186	if (npg_data == 1)
	187	packet->page_buf[1].len = skb_headlen(skb);
	188	else
	189	packet->page_buf[1].len = PAGE_SIZE
	190	- packet->page_buf[1].offset;
	191
	192	for (i = 2; i <= npg_data; i++) {
	193	packet->page_buf[i].pfn = virt_to_phys(skb->data
	194	+ PAGE_SIZE * (i-1)) >> PAGE_SHIFT;
	195	packet->page_buf[i].offset = 0;
	196	packet->page_buf[i].len = PAGE_SIZE;
	197	}
	198	if (npg_data > 1)
	199	packet->page_buf[npg_data].len = (((unsigned long)skb->data
	200	+ skb_headlen(skb) - 1) & (PAGE_SIZE - 1)) + 1;
6048718d SH	201
	202	/* Additional fragments are after SKB data */
	203	for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
9e903e08	204	const skb_frag_t *f = &skb_shinfo(skb)->frags[i];
6048718d	205
4d447c9a HZ	206	packet->page_buf[i+npg_data+1].pfn =
	207	page_to_pfn(skb_frag_page(f));
	208	packet->page_buf[i+npg_data+1].offset = f->page_offset;
	209	packet->page_buf[i+npg_data+1].len = skb_frag_size(f);
fceaf24a HJ	210	}
fceaf24a HJ	211
454f18a9	212	/* Set the completion routine */
72a2f5bd HZ	213	packet->completion.send.send_completion = netvsc_xmit_completion;
	214	packet->completion.send.send_completion_ctx = packet;
	215	packet->completion.send.send_completion_tid = (unsigned long)skb;
fceaf24a	216
55acb696	217	ret = rndis_filter_send(net_device_ctx->device_ctx,
02fafbc6	218	packet);
02fafbc6	219	if (ret == 0) {
b852fdce SH	220	net->stats.tx_bytes += skb->len;
b852fdce SH	221	net->stats.tx_packets++;
b220f5f9 SH	222	} else {
b220f5f9 SH	223	/* we are shutting down or bus overloaded, just drop packet */
b852fdce	224	net->stats.tx_dropped++;
8a5f9edc HZ	225	kfree(packet);
8a5f9edc HZ	226	dev_kfree_skb_any(skb);
fceaf24a HJ	227	}
fceaf24a HJ	228
bf769375	229	return ret ? NETDEV_TX_BUSY : NETDEV_TX_OK;
fceaf24a HJ	230	}
fceaf24a HJ	231
3e189519	232	/*
02fafbc6 GKH	233	* netvsc_linkstatus_callback - Link up/down notification
02fafbc6 GKH	234	*/
90ef117a	235	void netvsc_linkstatus_callback(struct hv_device *device_obj,
02fafbc6	236	unsigned int status)
fceaf24a	237	{
2ddd5e5f	238	struct net_device *net;
c996edcf	239	struct net_device_context *ndev_ctx;
2ddd5e5f S	240	struct netvsc_device *net_device;
	241
	242	net_device = hv_get_drvdata(device_obj);
	243	net = net_device->ndev;
fceaf24a	244
02fafbc6	245	if (!net) {
eb335bc4 HJ	246	netdev_err(net, "got link status but net device "
eb335bc4 HJ	247	"not initialized yet\n");
fceaf24a HJ	248	return;
	249	}
	250
02fafbc6	251	if (status == 1) {
fceaf24a HJ	252	netif_carrier_on(net);
fceaf24a HJ	253	netif_wake_queue(net);
c996edcf	254	ndev_ctx = netdev_priv(net);
c4b6a2ea	255	schedule_delayed_work(&ndev_ctx->dwork, 0);
122a5f64	256	schedule_delayed_work(&ndev_ctx->dwork, msecs_to_jiffies(20));
02fafbc6	257	} else {
fceaf24a	258	netif_carrier_off(net);
0a282538	259	netif_tx_disable(net);
fceaf24a	260	}
fceaf24a HJ	261	}
fceaf24a HJ	262
3e189519 HJ	263	/*
	264	* netvsc_recv_callback - Callback when we receive a packet from the
	265	* "wire" on the specified device.
02fafbc6	266	*/
f79adf8f	267	int netvsc_recv_callback(struct hv_device *device_obj,
02fafbc6	268	struct hv_netvsc_packet *packet)
fceaf24a	269	{
6f4c4446	270	struct net_device *net;
fceaf24a	271	struct sk_buff *skb;
fceaf24a	272
6f4c4446	273	net = ((struct netvsc_device *)hv_get_drvdata(device_obj))->ndev;
02fafbc6	274	if (!net) {
eb335bc4 HJ	275	netdev_err(net, "got receive callback but net device"
eb335bc4 HJ	276	" not initialized yet\n");
fceaf24a HJ	277	return 0;
	278	}
	279
9495c282	280	/* Allocate a skb - TODO direct I/O to pages? */
72a2f5bd	281	skb = netdev_alloc_skb_ip_align(net, packet->total_data_buflen);
9495c282 SH	282	if (unlikely(!skb)) {
	283	++net->stats.rx_dropped;
	284	return 0;
	285	}
fceaf24a	286
02fafbc6 GKH	287	/*
	288	* Copy to skb. This copy is needed here since the memory pointed by
	289	* hv_netvsc_packet cannot be deallocated
	290	*/
45326342 HZ	291	memcpy(skb_put(skb, packet->total_data_buflen), packet->data,
45326342 HZ	292	packet->total_data_buflen);
fceaf24a HJ	293
fceaf24a HJ	294	skb->protocol = eth_type_trans(skb, net);
fceaf24a HJ	295	skb->ip_summed = CHECKSUM_NONE;
fceaf24a HJ	296
9495c282	297	net->stats.rx_packets++;
48c38839	298	net->stats.rx_bytes += packet->total_data_buflen;
9495c282	299
02fafbc6 GKH	300	/*
02fafbc6 GKH	301	* Pass the skb back up. Network stack will deallocate the skb when it
9495c282 SH	302	* is done.
9495c282 SH	303	* TODO - use NAPI?
02fafbc6	304	*/
9495c282	305	netif_rx(skb);
fceaf24a	306
fceaf24a HJ	307	return 0;
	308	}
	309
f82f4ad7 SH	310	static void netvsc_get_drvinfo(struct net_device *net,
	311	struct ethtool_drvinfo *info)
	312	{
	313	strcpy(info->driver, "hv_netvsc");
	314	strcpy(info->version, HV_DRV_VERSION);
	315	strcpy(info->fw_version, "N/A");
	316	}
	317
4d447c9a HZ	318	static int netvsc_change_mtu(struct net_device *ndev, int mtu)
	319	{
	320	struct net_device_context *ndevctx = netdev_priv(ndev);
	321	struct hv_device *hdev = ndevctx->device_ctx;
	322	struct netvsc_device *nvdev = hv_get_drvdata(hdev);
	323	struct netvsc_device_info device_info;
	324	int limit = ETH_DATA_LEN;
	325
	326	if (nvdev == NULL \|\| nvdev->destroy)
	327	return -ENODEV;
	328
	329	if (nvdev->nvsp_version == NVSP_PROTOCOL_VERSION_2)
	330	limit = NETVSC_MTU;
	331
	332	if (mtu < 68 \|\| mtu > limit)
	333	return -EINVAL;
	334
	335	nvdev->start_remove = true;
	336	cancel_delayed_work_sync(&ndevctx->dwork);
0a282538	337	netif_tx_disable(ndev);
4d447c9a HZ	338	rndis_filter_device_remove(hdev);
	339
	340	ndev->mtu = mtu;
	341
	342	ndevctx->device_ctx = hdev;
	343	hv_set_drvdata(hdev, ndev);
	344	device_info.ring_size = ring_size;
	345	rndis_filter_device_add(hdev, &device_info);
	346	netif_wake_queue(ndev);
	347
	348	return 0;
	349	}
	350
f82f4ad7 SH	351	static const struct ethtool_ops ethtool_ops = {
f82f4ad7 SH	352	.get_drvinfo = netvsc_get_drvinfo,
f82f4ad7 SH	353	.get_link = ethtool_op_get_link,
	354	};
	355
df2fff28 GKH	356	static const struct net_device_ops device_ops = {
	357	.ndo_open = netvsc_open,
	358	.ndo_stop = netvsc_close,
	359	.ndo_start_xmit = netvsc_start_xmit,
afc4b13d	360	.ndo_set_rx_mode = netvsc_set_multicast_list,
4d447c9a	361	.ndo_change_mtu = netvsc_change_mtu,
b681b588 HZ	362	.ndo_validate_addr = eth_validate_addr,
b681b588 HZ	363	.ndo_set_mac_address = eth_mac_addr,
df2fff28 GKH	364	};
df2fff28 GKH	365
c996edcf HZ	366	/*
	367	* Send GARP packet to network peers after migrations.
	368	* After Quick Migration, the network is not immediately operational in the
	369	* current context when receiving RNDIS_STATUS_MEDIA_CONNECT event. So, add
122a5f64	370	* another netif_notify_peers() into a delayed work, otherwise GARP packet
c996edcf HZ	371	* will not be sent after quick migration, and cause network disconnection.
	372	*/
	373	static void netvsc_send_garp(struct work_struct *w)
	374	{
	375	struct net_device_context *ndev_ctx;
	376	struct net_device *net;
2ddd5e5f	377	struct netvsc_device *net_device;
c996edcf	378
122a5f64	379	ndev_ctx = container_of(w, struct net_device_context, dwork.work);
2ddd5e5f S	380	net_device = hv_get_drvdata(ndev_ctx->device_ctx);
2ddd5e5f S	381	net = net_device->ndev;
c996edcf HZ	382	netif_notify_peers(net);
	383	}
	384
	385
84946899 S	386	static int netvsc_probe(struct hv_device *dev,
84946899 S	387	const struct hv_vmbus_device_id *dev_id)
df2fff28	388	{
df2fff28 GKH	389	struct net_device *net = NULL;
	390	struct net_device_context *net_device_ctx;
	391	struct netvsc_device_info device_info;
	392	int ret;
	393
546d9e10	394	net = alloc_etherdev(sizeof(struct net_device_context));
df2fff28	395	if (!net)
51a805d0	396	return -ENOMEM;
df2fff28 GKH	397
	398	/* Set initial state */
	399	netif_carrier_off(net);
df2fff28 GKH	400
df2fff28 GKH	401	net_device_ctx = netdev_priv(net);
9efd21e1	402	net_device_ctx->device_ctx = dev;
2ddd5e5f	403	hv_set_drvdata(dev, net);
122a5f64	404	INIT_DELAYED_WORK(&net_device_ctx->dwork, netvsc_send_garp);
df2fff28	405
df2fff28 GKH	406	net->netdev_ops = &device_ops;
df2fff28 GKH	407
6048718d	408	/* TODO: Add GSO and Checksum offload */
877a344b	409	net->hw_features = NETIF_F_SG;
6048718d SH	410	net->features = NETIF_F_SG;
6048718d SH	411
f82f4ad7	412	SET_ETHTOOL_OPS(net, &ethtool_ops);
9efd21e1	413	SET_NETDEV_DEV(net, &dev->device);
df2fff28 GKH	414
	415	ret = register_netdev(net);
	416	if (ret != 0) {
692e084e	417	pr_err("Unable to register netdev.\n");
df2fff28	418	free_netdev(net);
692e084e	419	goto out;
df2fff28 GKH	420	}
df2fff28 GKH	421
692e084e HZ	422	/* Notify the netvsc driver of the new device */
	423	device_info.ring_size = ring_size;
	424	ret = rndis_filter_device_add(dev, &device_info);
	425	if (ret != 0) {
	426	netdev_err(net, "unable to add netvsc device (ret %d)\n", ret);
	427	unregister_netdev(net);
df2fff28	428	free_netdev(net);
2ddd5e5f	429	hv_set_drvdata(dev, NULL);
692e084e	430	return ret;
df2fff28	431	}
692e084e HZ	432	memcpy(net->dev_addr, device_info.mac_adr, ETH_ALEN);
	433
	434	netif_carrier_on(net);
df2fff28	435
692e084e	436	out:
df2fff28 GKH	437	return ret;
	438	}
	439
415b023a	440	static int netvsc_remove(struct hv_device *dev)
df2fff28	441	{
2ddd5e5f	442	struct net_device *net;
122a5f64	443	struct net_device_context *ndev_ctx;
2ddd5e5f S	444	struct netvsc_device *net_device;
	445
	446	net_device = hv_get_drvdata(dev);
	447	net = net_device->ndev;
df2fff28	448
df2fff28	449	if (net == NULL) {
415b023a	450	dev_err(&dev->device, "No net device to remove\n");
df2fff28 GKH	451	return 0;
	452	}
	453
4d447c9a HZ	454	net_device->start_remove = true;
4d447c9a HZ	455
122a5f64 HZ	456	ndev_ctx = netdev_priv(net);
	457	cancel_delayed_work_sync(&ndev_ctx->dwork);
	458
df2fff28	459	/* Stop outbound asap */
0a282538	460	netif_tx_disable(net);
df2fff28 GKH	461
	462	unregister_netdev(net);
	463
	464	/*
	465	* Call to the vsc driver to let it know that the device is being
	466	* removed
	467	*/
df06bcff	468	rndis_filter_device_remove(dev);
df2fff28 GKH	469
df2fff28 GKH	470	free_netdev(net);
df06bcff	471	return 0;
df2fff28 GKH	472	}
df2fff28 GKH	473
345c4cc3	474	static const struct hv_vmbus_device_id id_table[] = {
c45cf2d4 GKH	475	/* Network guid */
	476	{ VMBUS_DEVICE(0x63, 0x51, 0x61, 0xF8, 0x3E, 0xDF, 0xc5, 0x46,
	477	0x91, 0x3F, 0xF2, 0xD2, 0xF9, 0x65, 0xED, 0x0E) },
	478	{ },
345c4cc3 S	479	};
	480
	481	MODULE_DEVICE_TABLE(vmbus, id_table);
	482
f1542a66	483	/* The one and only one */
fde0ef9b	484	static struct hv_driver netvsc_drv = {
768fa219	485	.name = "netvsc",
345c4cc3	486	.id_table = id_table,
fde0ef9b S	487	.probe = netvsc_probe,
fde0ef9b S	488	.remove = netvsc_remove,
d4890970	489	};
f1542a66	490
a9869c94	491	static void __exit netvsc_drv_exit(void)
fceaf24a	492	{
768fa219	493	vmbus_driver_unregister(&netvsc_drv);
fceaf24a HJ	494	}
fceaf24a HJ	495
1fde28cf	496	static int __init netvsc_drv_init(void)
df2fff28	497	{
768fa219	498	return vmbus_driver_register(&netvsc_drv);
df2fff28 GKH	499	}
df2fff28 GKH	500
26c14cc1 HJ	501	MODULE_LICENSE("GPL");
26c14cc1 HJ	502	MODULE_VERSION(HV_DRV_VERSION);
7880fc54	503	MODULE_DESCRIPTION("Microsoft Hyper-V network driver");
fceaf24a	504
1fde28cf	505	module_init(netvsc_drv_init);
a9869c94	506	module_exit(netvsc_drv_exit);