1 /* 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: 18 * Haiyang Zhang <haiyangz@microsoft.com> 19 * Hank Janssen <hjanssen@microsoft.com> 20 */ 21 #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt 22 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> 28 #include <linux/io.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> 34 #include <linux/if_vlan.h> 35 #include <linux/in.h> 36 #include <linux/slab.h> 37 #include <net/arp.h> 38 #include <net/route.h> 39 #include <net/sock.h> 40 #include <net/pkt_sched.h> 41 42 #include "hyperv_net.h" 43 44 struct net_device_context { 45 /* point back to our device context */ 46 struct hv_device *device_ctx; 47 struct delayed_work dwork; 48 struct work_struct work; 49 }; 50 51 #define RING_SIZE_MIN 64 52 static int ring_size = 128; 53 module_param(ring_size, int, S_IRUGO); 54 MODULE_PARM_DESC(ring_size, "Ring buffer size (# of pages)"); 55 56 static void do_set_multicast(struct work_struct *w) 57 { 58 struct net_device_context *ndevctx = 59 container_of(w, struct net_device_context, work); 60 struct netvsc_device *nvdev; 61 struct rndis_device *rdev; 62 63 nvdev = hv_get_drvdata(ndevctx->device_ctx); 64 if (nvdev == NULL || nvdev->ndev == NULL) 65 return; 66 67 rdev = nvdev->extension; 68 if (rdev == NULL) 69 return; 70 71 if (nvdev->ndev->flags & IFF_PROMISC) 72 rndis_filter_set_packet_filter(rdev, 73 NDIS_PACKET_TYPE_PROMISCUOUS); 74 else 75 rndis_filter_set_packet_filter(rdev, 76 NDIS_PACKET_TYPE_BROADCAST | 77 NDIS_PACKET_TYPE_ALL_MULTICAST | 78 NDIS_PACKET_TYPE_DIRECTED); 79 } 80 81 static void netvsc_set_multicast_list(struct net_device *net) 82 { 83 struct net_device_context *net_device_ctx = netdev_priv(net); 84 85 schedule_work(&net_device_ctx->work); 86 } 87 88 static int netvsc_open(struct net_device *net) 89 { 90 struct net_device_context *net_device_ctx = netdev_priv(net); 91 struct hv_device *device_obj = net_device_ctx->device_ctx; 92 int ret = 0; 93 94 /* Open up the device */ 95 ret = rndis_filter_open(device_obj); 96 if (ret != 0) { 97 netdev_err(net, "unable to open device (ret %d).\n", ret); 98 return ret; 99 } 100 101 netif_start_queue(net); 102 103 return ret; 104 } 105 106 static int netvsc_close(struct net_device *net) 107 { 108 struct net_device_context *net_device_ctx = netdev_priv(net); 109 struct hv_device *device_obj = net_device_ctx->device_ctx; 110 int ret; 111 112 netif_tx_disable(net); 113 114 /* Make sure netvsc_set_multicast_list doesn't re-enable filter! */ 115 cancel_work_sync(&net_device_ctx->work); 116 ret = rndis_filter_close(device_obj); 117 if (ret != 0) 118 netdev_err(net, "unable to close device (ret %d).\n", ret); 119 120 return ret; 121 } 122 123 static void netvsc_xmit_completion(void *context) 124 { 125 struct hv_netvsc_packet *packet = (struct hv_netvsc_packet *)context; 126 struct sk_buff *skb = (struct sk_buff *) 127 (unsigned long)packet->completion.send.send_completion_tid; 128 129 kfree(packet); 130 131 if (skb) 132 dev_kfree_skb_any(skb); 133 } 134 135 static int netvsc_start_xmit(struct sk_buff *skb, struct net_device *net) 136 { 137 struct net_device_context *net_device_ctx = netdev_priv(net); 138 struct hv_netvsc_packet *packet; 139 int ret; 140 unsigned int i, num_pages, npg_data; 141 142 /* Add multipages for skb->data and additional 2 for RNDIS */ 143 npg_data = (((unsigned long)skb->data + skb_headlen(skb) - 1) 144 >> PAGE_SHIFT) - ((unsigned long)skb->data >> PAGE_SHIFT) + 1; 145 num_pages = skb_shinfo(skb)->nr_frags + npg_data + 2; 146 147 /* Allocate a netvsc packet based on # of frags. */ 148 packet = kzalloc(sizeof(struct hv_netvsc_packet) + 149 (num_pages * sizeof(struct hv_page_buffer)) + 150 sizeof(struct rndis_filter_packet) + 151 NDIS_VLAN_PPI_SIZE, GFP_ATOMIC); 152 if (!packet) { 153 /* out of memory, drop packet */ 154 netdev_err(net, "unable to allocate hv_netvsc_packet\n"); 155 156 dev_kfree_skb(skb); 157 net->stats.tx_dropped++; 158 return NETDEV_TX_OK; 159 } 160 161 packet->vlan_tci = skb->vlan_tci; 162 163 packet->extension = (void *)(unsigned long)packet + 164 sizeof(struct hv_netvsc_packet) + 165 (num_pages * sizeof(struct hv_page_buffer)); 166 167 /* If the rndis msg goes beyond 1 page, we will add 1 later */ 168 packet->page_buf_cnt = num_pages - 1; 169 170 /* Initialize it from the skb */ 171 packet->total_data_buflen = skb->len; 172 173 /* Start filling in the page buffers starting after RNDIS buffer. */ 174 packet->page_buf[1].pfn = virt_to_phys(skb->data) >> PAGE_SHIFT; 175 packet->page_buf[1].offset 176 = (unsigned long)skb->data & (PAGE_SIZE - 1); 177 if (npg_data == 1) 178 packet->page_buf[1].len = skb_headlen(skb); 179 else 180 packet->page_buf[1].len = PAGE_SIZE 181 - packet->page_buf[1].offset; 182 183 for (i = 2; i <= npg_data; i++) { 184 packet->page_buf[i].pfn = virt_to_phys(skb->data 185 + PAGE_SIZE * (i-1)) >> PAGE_SHIFT; 186 packet->page_buf[i].offset = 0; 187 packet->page_buf[i].len = PAGE_SIZE; 188 } 189 if (npg_data > 1) 190 packet->page_buf[npg_data].len = (((unsigned long)skb->data 191 + skb_headlen(skb) - 1) & (PAGE_SIZE - 1)) + 1; 192 193 /* Additional fragments are after SKB data */ 194 for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) { 195 const skb_frag_t *f = &skb_shinfo(skb)->frags[i]; 196 197 packet->page_buf[i+npg_data+1].pfn = 198 page_to_pfn(skb_frag_page(f)); 199 packet->page_buf[i+npg_data+1].offset = f->page_offset; 200 packet->page_buf[i+npg_data+1].len = skb_frag_size(f); 201 } 202 203 /* Set the completion routine */ 204 packet->completion.send.send_completion = netvsc_xmit_completion; 205 packet->completion.send.send_completion_ctx = packet; 206 packet->completion.send.send_completion_tid = (unsigned long)skb; 207 208 ret = rndis_filter_send(net_device_ctx->device_ctx, 209 packet); 210 if (ret == 0) { 211 net->stats.tx_bytes += skb->len; 212 net->stats.tx_packets++; 213 } else { 214 kfree(packet); 215 if (ret != -EAGAIN) { 216 dev_kfree_skb_any(skb); 217 net->stats.tx_dropped++; 218 } 219 } 220 221 return (ret == -EAGAIN) ? NETDEV_TX_BUSY : NETDEV_TX_OK; 222 } 223 224 /* 225 * netvsc_linkstatus_callback - Link up/down notification 226 */ 227 void netvsc_linkstatus_callback(struct hv_device *device_obj, 228 unsigned int status) 229 { 230 struct net_device *net; 231 struct net_device_context *ndev_ctx; 232 struct netvsc_device *net_device; 233 234 net_device = hv_get_drvdata(device_obj); 235 net = net_device->ndev; 236 237 if (!net) { 238 netdev_err(net, "got link status but net device " 239 "not initialized yet\n"); 240 return; 241 } 242 243 if (status == 1) { 244 netif_carrier_on(net); 245 ndev_ctx = netdev_priv(net); 246 schedule_delayed_work(&ndev_ctx->dwork, 0); 247 schedule_delayed_work(&ndev_ctx->dwork, msecs_to_jiffies(20)); 248 } else { 249 netif_carrier_off(net); 250 } 251 } 252 253 /* 254 * netvsc_recv_callback - Callback when we receive a packet from the 255 * "wire" on the specified device. 256 */ 257 int netvsc_recv_callback(struct hv_device *device_obj, 258 struct hv_netvsc_packet *packet) 259 { 260 struct net_device *net; 261 struct sk_buff *skb; 262 263 net = ((struct netvsc_device *)hv_get_drvdata(device_obj))->ndev; 264 if (!net) { 265 netdev_err(net, "got receive callback but net device" 266 " not initialized yet\n"); 267 packet->status = NVSP_STAT_FAIL; 268 return 0; 269 } 270 271 /* Allocate a skb - TODO direct I/O to pages? */ 272 skb = netdev_alloc_skb_ip_align(net, packet->total_data_buflen); 273 if (unlikely(!skb)) { 274 ++net->stats.rx_dropped; 275 packet->status = NVSP_STAT_FAIL; 276 return 0; 277 } 278 279 /* 280 * Copy to skb. This copy is needed here since the memory pointed by 281 * hv_netvsc_packet cannot be deallocated 282 */ 283 memcpy(skb_put(skb, packet->total_data_buflen), packet->data, 284 packet->total_data_buflen); 285 286 skb->protocol = eth_type_trans(skb, net); 287 skb->ip_summed = CHECKSUM_NONE; 288 if (packet->vlan_tci & VLAN_TAG_PRESENT) 289 __vlan_hwaccel_put_tag(skb, htons(ETH_P_8021Q), 290 packet->vlan_tci); 291 292 net->stats.rx_packets++; 293 net->stats.rx_bytes += packet->total_data_buflen; 294 295 /* 296 * Pass the skb back up. Network stack will deallocate the skb when it 297 * is done. 298 * TODO - use NAPI? 299 */ 300 netif_rx(skb); 301 302 return 0; 303 } 304 305 static void netvsc_get_drvinfo(struct net_device *net, 306 struct ethtool_drvinfo *info) 307 { 308 strlcpy(info->driver, KBUILD_MODNAME, sizeof(info->driver)); 309 strlcpy(info->version, HV_DRV_VERSION, sizeof(info->version)); 310 strlcpy(info->fw_version, "N/A", sizeof(info->fw_version)); 311 } 312 313 static int netvsc_change_mtu(struct net_device *ndev, int mtu) 314 { 315 struct net_device_context *ndevctx = netdev_priv(ndev); 316 struct hv_device *hdev = ndevctx->device_ctx; 317 struct netvsc_device *nvdev = hv_get_drvdata(hdev); 318 struct netvsc_device_info device_info; 319 int limit = ETH_DATA_LEN; 320 321 if (nvdev == NULL || nvdev->destroy) 322 return -ENODEV; 323 324 if (nvdev->nvsp_version == NVSP_PROTOCOL_VERSION_2) 325 limit = NETVSC_MTU; 326 327 if (mtu < 68 || mtu > limit) 328 return -EINVAL; 329 330 nvdev->start_remove = true; 331 cancel_delayed_work_sync(&ndevctx->dwork); 332 cancel_work_sync(&ndevctx->work); 333 netif_tx_disable(ndev); 334 rndis_filter_device_remove(hdev); 335 336 ndev->mtu = mtu; 337 338 ndevctx->device_ctx = hdev; 339 hv_set_drvdata(hdev, ndev); 340 device_info.ring_size = ring_size; 341 rndis_filter_device_add(hdev, &device_info); 342 netif_wake_queue(ndev); 343 344 return 0; 345 } 346 347 348 static int netvsc_set_mac_addr(struct net_device *ndev, void *p) 349 { 350 struct net_device_context *ndevctx = netdev_priv(ndev); 351 struct hv_device *hdev = ndevctx->device_ctx; 352 struct sockaddr *addr = p; 353 char save_adr[ETH_ALEN]; 354 unsigned char save_aatype; 355 int err; 356 357 memcpy(save_adr, ndev->dev_addr, ETH_ALEN); 358 save_aatype = ndev->addr_assign_type; 359 360 err = eth_mac_addr(ndev, p); 361 if (err != 0) 362 return err; 363 364 err = rndis_filter_set_device_mac(hdev, addr->sa_data); 365 if (err != 0) { 366 /* roll back to saved MAC */ 367 memcpy(ndev->dev_addr, save_adr, ETH_ALEN); 368 ndev->addr_assign_type = save_aatype; 369 } 370 371 return err; 372 } 373 374 375 static const struct ethtool_ops ethtool_ops = { 376 .get_drvinfo = netvsc_get_drvinfo, 377 .get_link = ethtool_op_get_link, 378 }; 379 380 static const struct net_device_ops device_ops = { 381 .ndo_open = netvsc_open, 382 .ndo_stop = netvsc_close, 383 .ndo_start_xmit = netvsc_start_xmit, 384 .ndo_set_rx_mode = netvsc_set_multicast_list, 385 .ndo_change_mtu = netvsc_change_mtu, 386 .ndo_validate_addr = eth_validate_addr, 387 .ndo_set_mac_address = netvsc_set_mac_addr, 388 }; 389 390 /* 391 * Send GARP packet to network peers after migrations. 392 * After Quick Migration, the network is not immediately operational in the 393 * current context when receiving RNDIS_STATUS_MEDIA_CONNECT event. So, add 394 * another netif_notify_peers() into a delayed work, otherwise GARP packet 395 * will not be sent after quick migration, and cause network disconnection. 396 */ 397 static void netvsc_send_garp(struct work_struct *w) 398 { 399 struct net_device_context *ndev_ctx; 400 struct net_device *net; 401 struct netvsc_device *net_device; 402 403 ndev_ctx = container_of(w, struct net_device_context, dwork.work); 404 net_device = hv_get_drvdata(ndev_ctx->device_ctx); 405 net = net_device->ndev; 406 netdev_notify_peers(net); 407 } 408 409 410 static int netvsc_probe(struct hv_device *dev, 411 const struct hv_vmbus_device_id *dev_id) 412 { 413 struct net_device *net = NULL; 414 struct net_device_context *net_device_ctx; 415 struct netvsc_device_info device_info; 416 int ret; 417 418 net = alloc_etherdev(sizeof(struct net_device_context)); 419 if (!net) 420 return -ENOMEM; 421 422 /* Set initial state */ 423 netif_carrier_off(net); 424 425 net_device_ctx = netdev_priv(net); 426 net_device_ctx->device_ctx = dev; 427 hv_set_drvdata(dev, net); 428 INIT_DELAYED_WORK(&net_device_ctx->dwork, netvsc_send_garp); 429 INIT_WORK(&net_device_ctx->work, do_set_multicast); 430 431 net->netdev_ops = &device_ops; 432 433 /* TODO: Add GSO and Checksum offload */ 434 net->hw_features = 0; 435 net->features = NETIF_F_HW_VLAN_CTAG_TX; 436 437 SET_ETHTOOL_OPS(net, ðtool_ops); 438 SET_NETDEV_DEV(net, &dev->device); 439 440 ret = register_netdev(net); 441 if (ret != 0) { 442 pr_err("Unable to register netdev.\n"); 443 free_netdev(net); 444 goto out; 445 } 446 447 /* Notify the netvsc driver of the new device */ 448 device_info.ring_size = ring_size; 449 ret = rndis_filter_device_add(dev, &device_info); 450 if (ret != 0) { 451 netdev_err(net, "unable to add netvsc device (ret %d)\n", ret); 452 unregister_netdev(net); 453 free_netdev(net); 454 hv_set_drvdata(dev, NULL); 455 return ret; 456 } 457 memcpy(net->dev_addr, device_info.mac_adr, ETH_ALEN); 458 459 netif_carrier_on(net); 460 461 out: 462 return ret; 463 } 464 465 static int netvsc_remove(struct hv_device *dev) 466 { 467 struct net_device *net; 468 struct net_device_context *ndev_ctx; 469 struct netvsc_device *net_device; 470 471 net_device = hv_get_drvdata(dev); 472 net = net_device->ndev; 473 474 if (net == NULL) { 475 dev_err(&dev->device, "No net device to remove\n"); 476 return 0; 477 } 478 479 net_device->start_remove = true; 480 481 ndev_ctx = netdev_priv(net); 482 cancel_delayed_work_sync(&ndev_ctx->dwork); 483 cancel_work_sync(&ndev_ctx->work); 484 485 /* Stop outbound asap */ 486 netif_tx_disable(net); 487 488 unregister_netdev(net); 489 490 /* 491 * Call to the vsc driver to let it know that the device is being 492 * removed 493 */ 494 rndis_filter_device_remove(dev); 495 496 free_netdev(net); 497 return 0; 498 } 499 500 static const struct hv_vmbus_device_id id_table[] = { 501 /* Network guid */ 502 { HV_NIC_GUID, }, 503 { }, 504 }; 505 506 MODULE_DEVICE_TABLE(vmbus, id_table); 507 508 /* The one and only one */ 509 static struct hv_driver netvsc_drv = { 510 .name = KBUILD_MODNAME, 511 .id_table = id_table, 512 .probe = netvsc_probe, 513 .remove = netvsc_remove, 514 }; 515 516 static void __exit netvsc_drv_exit(void) 517 { 518 vmbus_driver_unregister(&netvsc_drv); 519 } 520 521 static int __init netvsc_drv_init(void) 522 { 523 if (ring_size < RING_SIZE_MIN) { 524 ring_size = RING_SIZE_MIN; 525 pr_info("Increased ring_size to %d (min allowed)\n", 526 ring_size); 527 } 528 return vmbus_driver_register(&netvsc_drv); 529 } 530 531 MODULE_LICENSE("GPL"); 532 MODULE_VERSION(HV_DRV_VERSION); 533 MODULE_DESCRIPTION("Microsoft Hyper-V network driver"); 534 535 module_init(netvsc_drv_init); 536 module_exit(netvsc_drv_exit); 537