2 * Copyright (c) 2009, Microsoft Corporation.
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.
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
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.
18 * Haiyang Zhang <haiyangz@microsoft.com>
19 * Hank Janssen <hjanssen@microsoft.com>
21 #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
23 #include <linux/init.h>
24 #include <linux/atomic.h>
25 #include <linux/module.h>
26 #include <linux/highmem.h>
27 #include <linux/device.h>
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>
35 #include <linux/slab.h>
37 #include <net/route.h>
39 #include <net/pkt_sched.h>
41 #include "hyperv_net.h"
43 struct net_device_context {
44 /* point back to our device context */
45 struct hv_device *device_ctx;
46 struct delayed_work dwork;
50 static int ring_size = 128;
51 module_param(ring_size, int, S_IRUGO);
52 MODULE_PARM_DESC(ring_size, "Ring buffer size (# of pages)");
54 struct set_multicast_work {
55 struct work_struct work;
56 struct net_device *net;
59 static void do_set_multicast(struct work_struct *w)
61 struct set_multicast_work *swk =
62 container_of(w, struct set_multicast_work, work);
63 struct net_device *net = swk->net;
65 struct net_device_context *ndevctx = netdev_priv(net);
66 struct netvsc_device *nvdev;
67 struct rndis_device *rdev;
69 nvdev = hv_get_drvdata(ndevctx->device_ctx);
73 rdev = nvdev->extension;
77 if (net->flags & IFF_PROMISC)
78 rndis_filter_set_packet_filter(rdev,
79 NDIS_PACKET_TYPE_PROMISCUOUS);
81 rndis_filter_set_packet_filter(rdev,
82 NDIS_PACKET_TYPE_BROADCAST |
83 NDIS_PACKET_TYPE_ALL_MULTICAST |
84 NDIS_PACKET_TYPE_DIRECTED);
90 static void netvsc_set_multicast_list(struct net_device *net)
92 struct set_multicast_work *swk =
93 kmalloc(sizeof(struct set_multicast_work), GFP_ATOMIC);
98 INIT_WORK(&swk->work, do_set_multicast);
99 schedule_work(&swk->work);
102 static int netvsc_open(struct net_device *net)
104 struct net_device_context *net_device_ctx = netdev_priv(net);
105 struct hv_device *device_obj = net_device_ctx->device_ctx;
108 /* Open up the device */
109 ret = rndis_filter_open(device_obj);
111 netdev_err(net, "unable to open device (ret %d).\n", ret);
115 netif_start_queue(net);
120 static int netvsc_close(struct net_device *net)
122 struct net_device_context *net_device_ctx = netdev_priv(net);
123 struct hv_device *device_obj = net_device_ctx->device_ctx;
126 netif_tx_disable(net);
128 ret = rndis_filter_close(device_obj);
130 netdev_err(net, "unable to close device (ret %d).\n", ret);
135 static void netvsc_xmit_completion(void *context)
137 struct hv_netvsc_packet *packet = (struct hv_netvsc_packet *)context;
138 struct sk_buff *skb = (struct sk_buff *)
139 (unsigned long)packet->completion.send.send_completion_tid;
144 dev_kfree_skb_any(skb);
147 static int netvsc_start_xmit(struct sk_buff *skb, struct net_device *net)
149 struct net_device_context *net_device_ctx = netdev_priv(net);
150 struct hv_netvsc_packet *packet;
152 unsigned int i, num_pages, npg_data;
154 /* Add multipages for skb->data and additional 2 for RNDIS */
155 npg_data = (((unsigned long)skb->data + skb_headlen(skb) - 1)
156 >> PAGE_SHIFT) - ((unsigned long)skb->data >> PAGE_SHIFT) + 1;
157 num_pages = skb_shinfo(skb)->nr_frags + npg_data + 2;
159 /* Allocate a netvsc packet based on # of frags. */
160 packet = kzalloc(sizeof(struct hv_netvsc_packet) +
161 (num_pages * sizeof(struct hv_page_buffer)) +
162 sizeof(struct rndis_filter_packet) +
163 NDIS_VLAN_PPI_SIZE, GFP_ATOMIC);
165 /* out of memory, drop packet */
166 netdev_err(net, "unable to allocate hv_netvsc_packet\n");
169 net->stats.tx_dropped++;
173 packet->vlan_tci = skb->vlan_tci;
175 packet->extension = (void *)(unsigned long)packet +
176 sizeof(struct hv_netvsc_packet) +
177 (num_pages * sizeof(struct hv_page_buffer));
179 /* If the rndis msg goes beyond 1 page, we will add 1 later */
180 packet->page_buf_cnt = num_pages - 1;
182 /* Initialize it from the skb */
183 packet->total_data_buflen = skb->len;
185 /* Start filling in the page buffers starting after RNDIS buffer. */
186 packet->page_buf[1].pfn = virt_to_phys(skb->data) >> PAGE_SHIFT;
187 packet->page_buf[1].offset
188 = (unsigned long)skb->data & (PAGE_SIZE - 1);
190 packet->page_buf[1].len = skb_headlen(skb);
192 packet->page_buf[1].len = PAGE_SIZE
193 - packet->page_buf[1].offset;
195 for (i = 2; i <= npg_data; i++) {
196 packet->page_buf[i].pfn = virt_to_phys(skb->data
197 + PAGE_SIZE * (i-1)) >> PAGE_SHIFT;
198 packet->page_buf[i].offset = 0;
199 packet->page_buf[i].len = PAGE_SIZE;
202 packet->page_buf[npg_data].len = (((unsigned long)skb->data
203 + skb_headlen(skb) - 1) & (PAGE_SIZE - 1)) + 1;
205 /* Additional fragments are after SKB data */
206 for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
207 const skb_frag_t *f = &skb_shinfo(skb)->frags[i];
209 packet->page_buf[i+npg_data+1].pfn =
210 page_to_pfn(skb_frag_page(f));
211 packet->page_buf[i+npg_data+1].offset = f->page_offset;
212 packet->page_buf[i+npg_data+1].len = skb_frag_size(f);
215 /* Set the completion routine */
216 packet->completion.send.send_completion = netvsc_xmit_completion;
217 packet->completion.send.send_completion_ctx = packet;
218 packet->completion.send.send_completion_tid = (unsigned long)skb;
220 ret = rndis_filter_send(net_device_ctx->device_ctx,
223 net->stats.tx_bytes += skb->len;
224 net->stats.tx_packets++;
229 return ret ? NETDEV_TX_BUSY : NETDEV_TX_OK;
233 * netvsc_linkstatus_callback - Link up/down notification
235 void netvsc_linkstatus_callback(struct hv_device *device_obj,
238 struct net_device *net;
239 struct net_device_context *ndev_ctx;
240 struct netvsc_device *net_device;
242 net_device = hv_get_drvdata(device_obj);
243 net = net_device->ndev;
246 netdev_err(net, "got link status but net device "
247 "not initialized yet\n");
252 netif_carrier_on(net);
253 netif_wake_queue(net);
254 ndev_ctx = netdev_priv(net);
255 schedule_delayed_work(&ndev_ctx->dwork, 0);
256 schedule_delayed_work(&ndev_ctx->dwork, msecs_to_jiffies(20));
258 netif_carrier_off(net);
259 netif_tx_disable(net);
264 * netvsc_recv_callback - Callback when we receive a packet from the
265 * "wire" on the specified device.
267 int netvsc_recv_callback(struct hv_device *device_obj,
268 struct hv_netvsc_packet *packet)
270 struct net_device *net;
273 net = ((struct netvsc_device *)hv_get_drvdata(device_obj))->ndev;
275 netdev_err(net, "got receive callback but net device"
276 " not initialized yet\n");
280 /* Allocate a skb - TODO direct I/O to pages? */
281 skb = netdev_alloc_skb_ip_align(net, packet->total_data_buflen);
282 if (unlikely(!skb)) {
283 ++net->stats.rx_dropped;
288 * Copy to skb. This copy is needed here since the memory pointed by
289 * hv_netvsc_packet cannot be deallocated
291 memcpy(skb_put(skb, packet->total_data_buflen), packet->data,
292 packet->total_data_buflen);
294 skb->protocol = eth_type_trans(skb, net);
295 skb->ip_summed = CHECKSUM_NONE;
296 skb->vlan_tci = packet->vlan_tci;
298 net->stats.rx_packets++;
299 net->stats.rx_bytes += packet->total_data_buflen;
302 * Pass the skb back up. Network stack will deallocate the skb when it
311 static void netvsc_get_drvinfo(struct net_device *net,
312 struct ethtool_drvinfo *info)
314 strcpy(info->driver, KBUILD_MODNAME);
315 strcpy(info->version, HV_DRV_VERSION);
316 strcpy(info->fw_version, "N/A");
319 static int netvsc_change_mtu(struct net_device *ndev, int mtu)
321 struct net_device_context *ndevctx = netdev_priv(ndev);
322 struct hv_device *hdev = ndevctx->device_ctx;
323 struct netvsc_device *nvdev = hv_get_drvdata(hdev);
324 struct netvsc_device_info device_info;
325 int limit = ETH_DATA_LEN;
327 if (nvdev == NULL || nvdev->destroy)
330 if (nvdev->nvsp_version == NVSP_PROTOCOL_VERSION_2)
333 if (mtu < 68 || mtu > limit)
336 nvdev->start_remove = true;
337 cancel_delayed_work_sync(&ndevctx->dwork);
338 netif_tx_disable(ndev);
339 rndis_filter_device_remove(hdev);
343 ndevctx->device_ctx = hdev;
344 hv_set_drvdata(hdev, ndev);
345 device_info.ring_size = ring_size;
346 rndis_filter_device_add(hdev, &device_info);
347 netif_wake_queue(ndev);
352 static const struct ethtool_ops ethtool_ops = {
353 .get_drvinfo = netvsc_get_drvinfo,
354 .get_link = ethtool_op_get_link,
357 static const struct net_device_ops device_ops = {
358 .ndo_open = netvsc_open,
359 .ndo_stop = netvsc_close,
360 .ndo_start_xmit = netvsc_start_xmit,
361 .ndo_set_rx_mode = netvsc_set_multicast_list,
362 .ndo_change_mtu = netvsc_change_mtu,
363 .ndo_validate_addr = eth_validate_addr,
364 .ndo_set_mac_address = eth_mac_addr,
368 * Send GARP packet to network peers after migrations.
369 * After Quick Migration, the network is not immediately operational in the
370 * current context when receiving RNDIS_STATUS_MEDIA_CONNECT event. So, add
371 * another netif_notify_peers() into a delayed work, otherwise GARP packet
372 * will not be sent after quick migration, and cause network disconnection.
374 static void netvsc_send_garp(struct work_struct *w)
376 struct net_device_context *ndev_ctx;
377 struct net_device *net;
378 struct netvsc_device *net_device;
380 ndev_ctx = container_of(w, struct net_device_context, dwork.work);
381 net_device = hv_get_drvdata(ndev_ctx->device_ctx);
382 net = net_device->ndev;
383 netif_notify_peers(net);
387 static int netvsc_probe(struct hv_device *dev,
388 const struct hv_vmbus_device_id *dev_id)
390 struct net_device *net = NULL;
391 struct net_device_context *net_device_ctx;
392 struct netvsc_device_info device_info;
395 net = alloc_etherdev(sizeof(struct net_device_context));
399 /* Set initial state */
400 netif_carrier_off(net);
402 net_device_ctx = netdev_priv(net);
403 net_device_ctx->device_ctx = dev;
404 hv_set_drvdata(dev, net);
405 INIT_DELAYED_WORK(&net_device_ctx->dwork, netvsc_send_garp);
407 net->netdev_ops = &device_ops;
409 /* TODO: Add GSO and Checksum offload */
410 net->hw_features = NETIF_F_SG;
411 net->features = NETIF_F_SG | NETIF_F_HW_VLAN_TX;
413 SET_ETHTOOL_OPS(net, ðtool_ops);
414 SET_NETDEV_DEV(net, &dev->device);
416 ret = register_netdev(net);
418 pr_err("Unable to register netdev.\n");
423 /* Notify the netvsc driver of the new device */
424 device_info.ring_size = ring_size;
425 ret = rndis_filter_device_add(dev, &device_info);
427 netdev_err(net, "unable to add netvsc device (ret %d)\n", ret);
428 unregister_netdev(net);
430 hv_set_drvdata(dev, NULL);
433 memcpy(net->dev_addr, device_info.mac_adr, ETH_ALEN);
435 netif_carrier_on(net);
441 static int netvsc_remove(struct hv_device *dev)
443 struct net_device *net;
444 struct net_device_context *ndev_ctx;
445 struct netvsc_device *net_device;
447 net_device = hv_get_drvdata(dev);
448 net = net_device->ndev;
451 dev_err(&dev->device, "No net device to remove\n");
455 net_device->start_remove = true;
457 ndev_ctx = netdev_priv(net);
458 cancel_delayed_work_sync(&ndev_ctx->dwork);
460 /* Stop outbound asap */
461 netif_tx_disable(net);
463 unregister_netdev(net);
466 * Call to the vsc driver to let it know that the device is being
469 rndis_filter_device_remove(dev);
475 static const struct hv_vmbus_device_id id_table[] = {
477 { VMBUS_DEVICE(0x63, 0x51, 0x61, 0xF8, 0x3E, 0xDF, 0xc5, 0x46,
478 0x91, 0x3F, 0xF2, 0xD2, 0xF9, 0x65, 0xED, 0x0E) },
482 MODULE_DEVICE_TABLE(vmbus, id_table);
484 /* The one and only one */
485 static struct hv_driver netvsc_drv = {
486 .name = KBUILD_MODNAME,
487 .id_table = id_table,
488 .probe = netvsc_probe,
489 .remove = netvsc_remove,
492 static void __exit netvsc_drv_exit(void)
494 vmbus_driver_unregister(&netvsc_drv);
497 static int __init netvsc_drv_init(void)
499 return vmbus_driver_register(&netvsc_drv);
502 MODULE_LICENSE("GPL");
503 MODULE_VERSION(HV_DRV_VERSION);
504 MODULE_DESCRIPTION("Microsoft Hyper-V network driver");
506 module_init(netvsc_drv_init);
507 module_exit(netvsc_drv_exit);