[mirror_ubuntu-jammy-kernel.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, see <http://www.gnu.org/licenses/>.
 *
 * 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/if_vlan.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;
	struct work_struct work;
};

#define RING_SIZE_MIN 64
static int ring_size = 128;
module_param(ring_size, int, S_IRUGO);
MODULE_PARM_DESC(ring_size, "Ring buffer size (# of pages)");

static void do_set_multicast(struct work_struct *w)
{
	struct net_device_context *ndevctx =
		container_of(w, struct net_device_context, work);
	struct netvsc_device *nvdev;
	struct rndis_device *rdev;

	nvdev = hv_get_drvdata(ndevctx->device_ctx);
	if (nvdev == NULL || nvdev->ndev == NULL)
		return;

	rdev = nvdev->extension;
	if (rdev == NULL)
		return;

	if (nvdev->ndev->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);
}

static void netvsc_set_multicast_list(struct net_device *net)
{
	struct net_device_context *net_device_ctx = netdev_priv(net);

	schedule_work(&net_device_ctx->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);

	/* Make sure netvsc_set_multicast_list doesn't re-enable filter! */
	cancel_work_sync(&net_device_ctx->work);
	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) +
			 NDIS_VLAN_PPI_SIZE, 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_OK;
	}

	packet->vlan_tci = skb->vlan_tci;

	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 {
		kfree(packet);
		if (ret != -EAGAIN) {
			dev_kfree_skb_any(skb);
			net->stats.tx_dropped++;
		}
	}

	return (ret == -EAGAIN) ? 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);
		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);
	}
}

/*
 * 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 || net->reg_state != NETREG_REGISTERED) {
		packet->status = NVSP_STAT_FAIL;
		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;
		packet->status = NVSP_STAT_FAIL;
		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;
	if (packet->vlan_tci & VLAN_TAG_PRESENT)
		__vlan_hwaccel_put_tag(skb, htons(ETH_P_8021Q),
				       packet->vlan_tci);

	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)
{
	strlcpy(info->driver, KBUILD_MODNAME, sizeof(info->driver));
	strlcpy(info->fw_version, "N/A", sizeof(info->fw_version));
}

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_work_sync(&ndevctx->work);
	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 int netvsc_set_mac_addr(struct net_device *ndev, void *p)
{
	struct net_device_context *ndevctx = netdev_priv(ndev);
	struct hv_device *hdev =  ndevctx->device_ctx;
	struct sockaddr *addr = p;
	char save_adr[ETH_ALEN];
	unsigned char save_aatype;
	int err;

	memcpy(save_adr, ndev->dev_addr, ETH_ALEN);
	save_aatype = ndev->addr_assign_type;

	err = eth_mac_addr(ndev, p);
	if (err != 0)
		return err;

	err = rndis_filter_set_device_mac(hdev, addr->sa_data);
	if (err != 0) {
		/* roll back to saved MAC */
		memcpy(ndev->dev_addr, save_adr, ETH_ALEN);
		ndev->addr_assign_type = save_aatype;
	}

	return err;
}


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 =		netvsc_set_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;
	netdev_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);
	INIT_WORK(&net_device_ctx->work, do_set_multicast);

	net->netdev_ops = &device_ops;

	/* TODO: Add GSO and Checksum offload */
	net->hw_features = 0;
	net->features = NETIF_F_HW_VLAN_CTAG_TX;

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

	/* 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);
		free_netdev(net);
		hv_set_drvdata(dev, NULL);
		return ret;
	}
	memcpy(net->dev_addr, device_info.mac_adr, ETH_ALEN);

	netif_carrier_on(net);

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

	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);
	cancel_work_sync(&ndev_ctx->work);

	/* 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 */
	{ HV_NIC_GUID, },
	{ },
};

MODULE_DEVICE_TABLE(vmbus, id_table);

/* The one and only one */
static struct  hv_driver netvsc_drv = {
	.name = KBUILD_MODNAME,
	.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)
{
	if (ring_size < RING_SIZE_MIN) {
		ring_size = RING_SIZE_MIN;
		pr_info("Increased ring_size to %d (min allowed)\n",
			ring_size);
	}
	return vmbus_driver_register(&netvsc_drv);
}

MODULE_LICENSE("GPL");
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
adf8d3ff	14	* this program; if not, see <http://www.gnu.org/licenses/>.
fceaf24a HJ	15	*
fceaf24a HJ	16	* Authors:
d0e94d17	17	* Haiyang Zhang <haiyangz@microsoft.com>
fceaf24a	18	* Hank Janssen <hjanssen@microsoft.com>
fceaf24a	19	*/
eb335bc4 HJ	20	#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
eb335bc4 HJ	21
fceaf24a	22	#include <linux/init.h>
9079ce69	23	#include <linux/atomic.h>
fceaf24a HJ	24	#include <linux/module.h>
	25	#include <linux/highmem.h>
	26	#include <linux/device.h>
fceaf24a	27	#include <linux/io.h>
fceaf24a HJ	28	#include <linux/delay.h>
	29	#include <linux/netdevice.h>
	30	#include <linux/inetdevice.h>
	31	#include <linux/etherdevice.h>
	32	#include <linux/skbuff.h>
c802db11	33	#include <linux/if_vlan.h>
fceaf24a	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;
792df872	47	struct work_struct work;
fceaf24a HJ	48	};
fceaf24a HJ	49
fa85a6c2	50	#define RING_SIZE_MIN 64
99c8da0f	51	static int ring_size = 128;
450d7a4b SH	52	module_param(ring_size, int, S_IRUGO);
450d7a4b SH	53	MODULE_PARM_DESC(ring_size, "Ring buffer size (# of pages)");
fceaf24a	54
d426b2e3 HZ	55	static void do_set_multicast(struct work_struct *w)
d426b2e3 HZ	56	{
792df872 WM	57	struct net_device_context *ndevctx =
792df872 WM	58	container_of(w, struct net_device_context, work);
d426b2e3 HZ	59	struct netvsc_device *nvdev;
	60	struct rndis_device *rdev;
	61
	62	nvdev = hv_get_drvdata(ndevctx->device_ctx);
792df872 WM	63	if (nvdev == NULL \|\| nvdev->ndev == NULL)
792df872 WM	64	return;
d426b2e3 HZ	65
	66	rdev = nvdev->extension;
	67	if (rdev == NULL)
792df872	68	return;
d426b2e3	69
792df872	70	if (nvdev->ndev->flags & IFF_PROMISC)
d426b2e3 HZ	71	rndis_filter_set_packet_filter(rdev,
	72	NDIS_PACKET_TYPE_PROMISCUOUS);
	73	else
	74	rndis_filter_set_packet_filter(rdev,
	75	NDIS_PACKET_TYPE_BROADCAST \|
	76	NDIS_PACKET_TYPE_ALL_MULTICAST \|
	77	NDIS_PACKET_TYPE_DIRECTED);
d426b2e3 HZ	78	}
d426b2e3 HZ	79
4e9bfefa	80	static void netvsc_set_multicast_list(struct net_device *net)
fceaf24a	81	{
792df872	82	struct net_device_context *net_device_ctx = netdev_priv(net);
d426b2e3	83
792df872	84	schedule_work(&net_device_ctx->work);
fceaf24a HJ	85	}
fceaf24a HJ	86
fceaf24a HJ	87	static int netvsc_open(struct net_device *net)
fceaf24a HJ	88	{
fceaf24a	89	struct net_device_context *net_device_ctx = netdev_priv(net);
6bad88da	90	struct hv_device *device_obj = net_device_ctx->device_ctx;
02fafbc6	91	int ret = 0;
fceaf24a	92
d515d0ff HZ	93	/* Open up the device */
	94	ret = rndis_filter_open(device_obj);
	95	if (ret != 0) {
	96	netdev_err(net, "unable to open device (ret %d).\n", ret);
	97	return ret;
fceaf24a HJ	98	}
fceaf24a HJ	99
d515d0ff HZ	100	netif_start_queue(net);
d515d0ff HZ	101
fceaf24a HJ	102	return ret;
	103	}
	104
fceaf24a HJ	105	static int netvsc_close(struct net_device *net)
fceaf24a HJ	106	{
fceaf24a	107	struct net_device_context *net_device_ctx = netdev_priv(net);
6bad88da	108	struct hv_device *device_obj = net_device_ctx->device_ctx;
02fafbc6	109	int ret;
fceaf24a	110
0a282538	111	netif_tx_disable(net);
fceaf24a	112
792df872 WM	113	/* Make sure netvsc_set_multicast_list doesn't re-enable filter! */
792df872 WM	114	cancel_work_sync(&net_device_ctx->work);
9c26aa0d	115	ret = rndis_filter_close(device_obj);
fceaf24a	116	if (ret != 0)
eb335bc4	117	netdev_err(net, "unable to close device (ret %d).\n", ret);
fceaf24a	118
fceaf24a HJ	119	return ret;
	120	}
	121
fceaf24a HJ	122	static void netvsc_xmit_completion(void *context)
fceaf24a HJ	123	{
4193d4f4	124	struct hv_netvsc_packet packet = (struct hv_netvsc_packet )context;
02fafbc6	125	struct sk_buff skb = (struct sk_buff )
72a2f5bd	126	(unsigned long)packet->completion.send.send_completion_tid;
fceaf24a	127
fceaf24a HJ	128	kfree(packet);
fceaf24a HJ	129
1d06825b	130	if (skb)
b220f5f9	131	dev_kfree_skb_any(skb);
fceaf24a HJ	132	}
fceaf24a HJ	133
02fafbc6	134	static int netvsc_start_xmit(struct sk_buff skb, struct net_device net)
fceaf24a	135	{
fceaf24a	136	struct net_device_context *net_device_ctx = netdev_priv(net);
4193d4f4	137	struct hv_netvsc_packet *packet;
02fafbc6	138	int ret;
4d447c9a	139	unsigned int i, num_pages, npg_data;
fceaf24a	140
c31c151b	141	/* Add multipages for skb->data and additional 2 for RNDIS */
4d447c9a HZ	142	npg_data = (((unsigned long)skb->data + skb_headlen(skb) - 1)
4d447c9a HZ	143	>> PAGE_SHIFT) - ((unsigned long)skb->data >> PAGE_SHIFT) + 1;
c31c151b	144	num_pages = skb_shinfo(skb)->nr_frags + npg_data + 2;
fceaf24a	145
454f18a9	146	/* Allocate a netvsc packet based on # of frags. */
02fafbc6	147	packet = kzalloc(sizeof(struct hv_netvsc_packet) +
6048718d	148	(num_pages * sizeof(struct hv_page_buffer)) +
1f5f3a75 HZ	149	sizeof(struct rndis_filter_packet) +
1f5f3a75 HZ	150	NDIS_VLAN_PPI_SIZE, GFP_ATOMIC);
02fafbc6	151	if (!packet) {
bf769375	152	/* out of memory, drop packet */
eb335bc4	153	netdev_err(net, "unable to allocate hv_netvsc_packet\n");
b220f5f9 SH	154
	155	dev_kfree_skb(skb);
	156	net->stats.tx_dropped++;
bb6d5e76	157	return NETDEV_TX_OK;
fceaf24a HJ	158	}
fceaf24a HJ	159
1f5f3a75 HZ	160	packet->vlan_tci = skb->vlan_tci;
1f5f3a75 HZ	161
72a2f5bd	162	packet->extension = (void *)(unsigned long)packet +
02fafbc6	163	sizeof(struct hv_netvsc_packet) +
6048718d	164	(num_pages * sizeof(struct hv_page_buffer));
fceaf24a	165
c31c151b HZ	166	/* If the rndis msg goes beyond 1 page, we will add 1 later */
c31c151b HZ	167	packet->page_buf_cnt = num_pages - 1;
fceaf24a	168
454f18a9	169	/* Initialize it from the skb */
4d447c9a	170	packet->total_data_buflen = skb->len;
fceaf24a	171
6048718d	172	/* Start filling in the page buffers starting after RNDIS buffer. */
ca623ad3 HZ	173	packet->page_buf[1].pfn = virt_to_phys(skb->data) >> PAGE_SHIFT;
ca623ad3 HZ	174	packet->page_buf[1].offset
6048718d	175	= (unsigned long)skb->data & (PAGE_SIZE - 1);
4d447c9a HZ	176	if (npg_data == 1)
	177	packet->page_buf[1].len = skb_headlen(skb);
	178	else
	179	packet->page_buf[1].len = PAGE_SIZE
	180	- packet->page_buf[1].offset;
	181
	182	for (i = 2; i <= npg_data; i++) {
	183	packet->page_buf[i].pfn = virt_to_phys(skb->data
	184	+ PAGE_SIZE * (i-1)) >> PAGE_SHIFT;
	185	packet->page_buf[i].offset = 0;
	186	packet->page_buf[i].len = PAGE_SIZE;
	187	}
	188	if (npg_data > 1)
	189	packet->page_buf[npg_data].len = (((unsigned long)skb->data
	190	+ skb_headlen(skb) - 1) & (PAGE_SIZE - 1)) + 1;
6048718d SH	191
	192	/* Additional fragments are after SKB data */
	193	for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
9e903e08	194	const skb_frag_t *f = &skb_shinfo(skb)->frags[i];
6048718d	195
4d447c9a HZ	196	packet->page_buf[i+npg_data+1].pfn =
	197	page_to_pfn(skb_frag_page(f));
	198	packet->page_buf[i+npg_data+1].offset = f->page_offset;
	199	packet->page_buf[i+npg_data+1].len = skb_frag_size(f);
fceaf24a HJ	200	}
fceaf24a HJ	201
454f18a9	202	/* Set the completion routine */
72a2f5bd HZ	203	packet->completion.send.send_completion = netvsc_xmit_completion;
	204	packet->completion.send.send_completion_ctx = packet;
	205	packet->completion.send.send_completion_tid = (unsigned long)skb;
fceaf24a	206
55acb696	207	ret = rndis_filter_send(net_device_ctx->device_ctx,
02fafbc6	208	packet);
02fafbc6	209	if (ret == 0) {
b852fdce SH	210	net->stats.tx_bytes += skb->len;
b852fdce SH	211	net->stats.tx_packets++;
b220f5f9	212	} else {
8a5f9edc	213	kfree(packet);
33be96e4 HZ	214	if (ret != -EAGAIN) {
	215	dev_kfree_skb_any(skb);
	216	net->stats.tx_dropped++;
	217	}
fceaf24a HJ	218	}
fceaf24a HJ	219
33be96e4	220	return (ret == -EAGAIN) ? NETDEV_TX_BUSY : NETDEV_TX_OK;
fceaf24a HJ	221	}
fceaf24a HJ	222
3e189519	223	/*
02fafbc6 GKH	224	* netvsc_linkstatus_callback - Link up/down notification
02fafbc6 GKH	225	*/
90ef117a	226	void netvsc_linkstatus_callback(struct hv_device *device_obj,
02fafbc6	227	unsigned int status)
fceaf24a	228	{
2ddd5e5f	229	struct net_device *net;
c996edcf	230	struct net_device_context *ndev_ctx;
2ddd5e5f S	231	struct netvsc_device *net_device;
	232
	233	net_device = hv_get_drvdata(device_obj);
	234	net = net_device->ndev;
fceaf24a	235
02fafbc6	236	if (!net) {
eb335bc4 HJ	237	netdev_err(net, "got link status but net device "
eb335bc4 HJ	238	"not initialized yet\n");
fceaf24a HJ	239	return;
	240	}
	241
02fafbc6	242	if (status == 1) {
fceaf24a	243	netif_carrier_on(net);
c996edcf	244	ndev_ctx = netdev_priv(net);
c4b6a2ea	245	schedule_delayed_work(&ndev_ctx->dwork, 0);
122a5f64	246	schedule_delayed_work(&ndev_ctx->dwork, msecs_to_jiffies(20));
02fafbc6	247	} else {
fceaf24a	248	netif_carrier_off(net);
fceaf24a	249	}
fceaf24a HJ	250	}
fceaf24a HJ	251
3e189519 HJ	252	/*
	253	* netvsc_recv_callback - Callback when we receive a packet from the
	254	* "wire" on the specified device.
02fafbc6	255	*/
f79adf8f	256	int netvsc_recv_callback(struct hv_device *device_obj,
02fafbc6	257	struct hv_netvsc_packet *packet)
fceaf24a	258	{
6f4c4446	259	struct net_device *net;
fceaf24a	260	struct sk_buff *skb;
fceaf24a	261
6f4c4446	262	net = ((struct netvsc_device *)hv_get_drvdata(device_obj))->ndev;
a68f9614	263	if (!net \|\| net->reg_state != NETREG_REGISTERED) {
63f6921d	264	packet->status = NVSP_STAT_FAIL;
fceaf24a HJ	265	return 0;
	266	}
	267
9495c282	268	/* Allocate a skb - TODO direct I/O to pages? */
72a2f5bd	269	skb = netdev_alloc_skb_ip_align(net, packet->total_data_buflen);
9495c282 SH	270	if (unlikely(!skb)) {
9495c282 SH	271	++net->stats.rx_dropped;
63f6921d	272	packet->status = NVSP_STAT_FAIL;
9495c282 SH	273	return 0;
9495c282 SH	274	}
fceaf24a	275
02fafbc6 GKH	276	/*
	277	* Copy to skb. This copy is needed here since the memory pointed by
	278	* hv_netvsc_packet cannot be deallocated
	279	*/
45326342 HZ	280	memcpy(skb_put(skb, packet->total_data_buflen), packet->data,
45326342 HZ	281	packet->total_data_buflen);
fceaf24a HJ	282
fceaf24a HJ	283	skb->protocol = eth_type_trans(skb, net);
fceaf24a	284	skb->ip_summed = CHECKSUM_NONE;
93725cbd HZ	285	if (packet->vlan_tci & VLAN_TAG_PRESENT)
	286	__vlan_hwaccel_put_tag(skb, htons(ETH_P_8021Q),
	287	packet->vlan_tci);
fceaf24a	288
9495c282	289	net->stats.rx_packets++;
48c38839	290	net->stats.rx_bytes += packet->total_data_buflen;
9495c282	291
02fafbc6 GKH	292	/*
02fafbc6 GKH	293	* Pass the skb back up. Network stack will deallocate the skb when it
9495c282 SH	294	* is done.
9495c282 SH	295	* TODO - use NAPI?
02fafbc6	296	*/
9495c282	297	netif_rx(skb);
fceaf24a	298
fceaf24a HJ	299	return 0;
	300	}
	301
f82f4ad7 SH	302	static void netvsc_get_drvinfo(struct net_device *net,
	303	struct ethtool_drvinfo *info)
	304	{
7826d43f	305	strlcpy(info->driver, KBUILD_MODNAME, sizeof(info->driver));
7826d43f	306	strlcpy(info->fw_version, "N/A", sizeof(info->fw_version));
f82f4ad7 SH	307	}
f82f4ad7 SH	308
4d447c9a HZ	309	static int netvsc_change_mtu(struct net_device *ndev, int mtu)
	310	{
	311	struct net_device_context *ndevctx = netdev_priv(ndev);
	312	struct hv_device *hdev = ndevctx->device_ctx;
	313	struct netvsc_device *nvdev = hv_get_drvdata(hdev);
	314	struct netvsc_device_info device_info;
	315	int limit = ETH_DATA_LEN;
	316
	317	if (nvdev == NULL \|\| nvdev->destroy)
	318	return -ENODEV;
	319
	320	if (nvdev->nvsp_version == NVSP_PROTOCOL_VERSION_2)
	321	limit = NETVSC_MTU;
	322
	323	if (mtu < 68 \|\| mtu > limit)
	324	return -EINVAL;
	325
	326	nvdev->start_remove = true;
792df872	327	cancel_work_sync(&ndevctx->work);
0a282538	328	netif_tx_disable(ndev);
4d447c9a HZ	329	rndis_filter_device_remove(hdev);
	330
	331	ndev->mtu = mtu;
	332
	333	ndevctx->device_ctx = hdev;
	334	hv_set_drvdata(hdev, ndev);
	335	device_info.ring_size = ring_size;
	336	rndis_filter_device_add(hdev, &device_info);
	337	netif_wake_queue(ndev);
	338
	339	return 0;
	340	}
	341
1ce09e89 HZ	342
	343	static int netvsc_set_mac_addr(struct net_device ndev, void p)
	344	{
	345	struct net_device_context *ndevctx = netdev_priv(ndev);
	346	struct hv_device *hdev = ndevctx->device_ctx;
	347	struct sockaddr *addr = p;
9a4c831e	348	char save_adr[ETH_ALEN];
1ce09e89 HZ	349	unsigned char save_aatype;
	350	int err;
	351
	352	memcpy(save_adr, ndev->dev_addr, ETH_ALEN);
	353	save_aatype = ndev->addr_assign_type;
	354
	355	err = eth_mac_addr(ndev, p);
	356	if (err != 0)
	357	return err;
	358
	359	err = rndis_filter_set_device_mac(hdev, addr->sa_data);
	360	if (err != 0) {
	361	/* roll back to saved MAC */
	362	memcpy(ndev->dev_addr, save_adr, ETH_ALEN);
	363	ndev->addr_assign_type = save_aatype;
	364	}
	365
	366	return err;
	367	}
	368
	369
f82f4ad7 SH	370	static const struct ethtool_ops ethtool_ops = {
f82f4ad7 SH	371	.get_drvinfo = netvsc_get_drvinfo,
f82f4ad7 SH	372	.get_link = ethtool_op_get_link,
	373	};
	374
df2fff28 GKH	375	static const struct net_device_ops device_ops = {
	376	.ndo_open = netvsc_open,
	377	.ndo_stop = netvsc_close,
	378	.ndo_start_xmit = netvsc_start_xmit,
afc4b13d	379	.ndo_set_rx_mode = netvsc_set_multicast_list,
4d447c9a	380	.ndo_change_mtu = netvsc_change_mtu,
b681b588	381	.ndo_validate_addr = eth_validate_addr,
1ce09e89	382	.ndo_set_mac_address = netvsc_set_mac_addr,
df2fff28 GKH	383	};
df2fff28 GKH	384
c996edcf HZ	385	/*
	386	* Send GARP packet to network peers after migrations.
	387	* After Quick Migration, the network is not immediately operational in the
	388	* current context when receiving RNDIS_STATUS_MEDIA_CONNECT event. So, add
122a5f64	389	* another netif_notify_peers() into a delayed work, otherwise GARP packet
c996edcf HZ	390	* will not be sent after quick migration, and cause network disconnection.
	391	*/
	392	static void netvsc_send_garp(struct work_struct *w)
	393	{
	394	struct net_device_context *ndev_ctx;
	395	struct net_device *net;
2ddd5e5f	396	struct netvsc_device *net_device;
c996edcf	397
122a5f64	398	ndev_ctx = container_of(w, struct net_device_context, dwork.work);
2ddd5e5f S	399	net_device = hv_get_drvdata(ndev_ctx->device_ctx);
2ddd5e5f S	400	net = net_device->ndev;
ee89bab1	401	netdev_notify_peers(net);
c996edcf HZ	402	}
	403
	404
84946899 S	405	static int netvsc_probe(struct hv_device *dev,
84946899 S	406	const struct hv_vmbus_device_id *dev_id)
df2fff28	407	{
df2fff28 GKH	408	struct net_device *net = NULL;
	409	struct net_device_context *net_device_ctx;
	410	struct netvsc_device_info device_info;
	411	int ret;
	412
546d9e10	413	net = alloc_etherdev(sizeof(struct net_device_context));
df2fff28	414	if (!net)
51a805d0	415	return -ENOMEM;
df2fff28 GKH	416
	417	/* Set initial state */
	418	netif_carrier_off(net);
df2fff28 GKH	419
df2fff28 GKH	420	net_device_ctx = netdev_priv(net);
9efd21e1	421	net_device_ctx->device_ctx = dev;
2ddd5e5f	422	hv_set_drvdata(dev, net);
122a5f64	423	INIT_DELAYED_WORK(&net_device_ctx->dwork, netvsc_send_garp);
792df872	424	INIT_WORK(&net_device_ctx->work, do_set_multicast);
df2fff28	425
df2fff28 GKH	426	net->netdev_ops = &device_ops;
df2fff28 GKH	427
6048718d	428	/* TODO: Add GSO and Checksum offload */
f4570820 HZ	429	net->hw_features = 0;
f4570820 HZ	430	net->features = NETIF_F_HW_VLAN_CTAG_TX;
6048718d	431
f82f4ad7	432	SET_ETHTOOL_OPS(net, &ethtool_ops);
9efd21e1	433	SET_NETDEV_DEV(net, &dev->device);
df2fff28	434
692e084e HZ	435	/* Notify the netvsc driver of the new device */
	436	device_info.ring_size = ring_size;
	437	ret = rndis_filter_device_add(dev, &device_info);
	438	if (ret != 0) {
	439	netdev_err(net, "unable to add netvsc device (ret %d)\n", ret);
df2fff28	440	free_netdev(net);
2ddd5e5f	441	hv_set_drvdata(dev, NULL);
692e084e	442	return ret;
df2fff28	443	}
692e084e HZ	444	memcpy(net->dev_addr, device_info.mac_adr, ETH_ALEN);
	445
	446	netif_carrier_on(net);
df2fff28	447
a68f9614 HZ	448	ret = register_netdev(net);
	449	if (ret != 0) {
	450	pr_err("Unable to register netdev.\n");
	451	rndis_filter_device_remove(dev);
	452	free_netdev(net);
	453	}
	454
df2fff28 GKH	455	return ret;
	456	}
	457
415b023a	458	static int netvsc_remove(struct hv_device *dev)
df2fff28	459	{
2ddd5e5f	460	struct net_device *net;
122a5f64	461	struct net_device_context *ndev_ctx;
2ddd5e5f S	462	struct netvsc_device *net_device;
	463
	464	net_device = hv_get_drvdata(dev);
	465	net = net_device->ndev;
df2fff28	466
df2fff28	467	if (net == NULL) {
415b023a	468	dev_err(&dev->device, "No net device to remove\n");
df2fff28 GKH	469	return 0;
	470	}
	471
4d447c9a HZ	472	net_device->start_remove = true;
4d447c9a HZ	473
122a5f64 HZ	474	ndev_ctx = netdev_priv(net);
122a5f64 HZ	475	cancel_delayed_work_sync(&ndev_ctx->dwork);
792df872	476	cancel_work_sync(&ndev_ctx->work);
122a5f64	477
df2fff28	478	/* Stop outbound asap */
0a282538	479	netif_tx_disable(net);
df2fff28 GKH	480
	481	unregister_netdev(net);
	482
	483	/*
	484	* Call to the vsc driver to let it know that the device is being
	485	* removed
	486	*/
df06bcff	487	rndis_filter_device_remove(dev);
df2fff28 GKH	488
df2fff28 GKH	489	free_netdev(net);
df06bcff	490	return 0;
df2fff28 GKH	491	}
df2fff28 GKH	492
345c4cc3	493	static const struct hv_vmbus_device_id id_table[] = {
c45cf2d4	494	/* Network guid */
8f505944	495	{ HV_NIC_GUID, },
c45cf2d4	496	{ },
345c4cc3 S	497	};
	498
	499	MODULE_DEVICE_TABLE(vmbus, id_table);
	500
f1542a66	501	/* The one and only one */
fde0ef9b	502	static struct hv_driver netvsc_drv = {
d31b20fc	503	.name = KBUILD_MODNAME,
345c4cc3	504	.id_table = id_table,
fde0ef9b S	505	.probe = netvsc_probe,
fde0ef9b S	506	.remove = netvsc_remove,
d4890970	507	};
f1542a66	508
a9869c94	509	static void __exit netvsc_drv_exit(void)
fceaf24a	510	{
768fa219	511	vmbus_driver_unregister(&netvsc_drv);
fceaf24a HJ	512	}
fceaf24a HJ	513
1fde28cf	514	static int __init netvsc_drv_init(void)
df2fff28	515	{
fa85a6c2 HZ	516	if (ring_size < RING_SIZE_MIN) {
	517	ring_size = RING_SIZE_MIN;
	518	pr_info("Increased ring_size to %d (min allowed)\n",
	519	ring_size);
	520	}
768fa219	521	return vmbus_driver_register(&netvsc_drv);
df2fff28 GKH	522	}
df2fff28 GKH	523
26c14cc1	524	MODULE_LICENSE("GPL");
7880fc54	525	MODULE_DESCRIPTION("Microsoft Hyper-V network driver");
fceaf24a	526
1fde28cf	527	module_init(netvsc_drv_init);
a9869c94	528	module_exit(netvsc_drv_exit);