1 /* 2 * Network-device interface management. 3 * 4 * Copyright (c) 2004-2005, Keir Fraser 5 * 6 * This program is free software; you can redistribute it and/or 7 * modify it under the terms of the GNU General Public License version 2 8 * as published by the Free Software Foundation; or, when distributed 9 * separately from the Linux kernel or incorporated into other 10 * software packages, subject to the following license: 11 * 12 * Permission is hereby granted, free of charge, to any person obtaining a copy 13 * of this source file (the "Software"), to deal in the Software without 14 * restriction, including without limitation the rights to use, copy, modify, 15 * merge, publish, distribute, sublicense, and/or sell copies of the Software, 16 * and to permit persons to whom the Software is furnished to do so, subject to 17 * the following conditions: 18 * 19 * The above copyright notice and this permission notice shall be included in 20 * all copies or substantial portions of the Software. 21 * 22 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR 23 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, 24 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE 25 * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER 26 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING 27 * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS 28 * IN THE SOFTWARE. 29 */ 30 31 #include "common.h" 32 33 #include <linux/kthread.h> 34 #include <linux/sched/task.h> 35 #include <linux/ethtool.h> 36 #include <linux/rtnetlink.h> 37 #include <linux/if_vlan.h> 38 #include <linux/vmalloc.h> 39 40 #include <xen/events.h> 41 #include <asm/xen/hypercall.h> 42 #include <xen/balloon.h> 43 44 /* Number of bytes allowed on the internal guest Rx queue. */ 45 #define XENVIF_RX_QUEUE_BYTES (XEN_NETIF_RX_RING_SIZE/2 * PAGE_SIZE) 46 47 /* This function is used to set SKBFL_ZEROCOPY_ENABLE as well as 48 * increasing the inflight counter. We need to increase the inflight 49 * counter because core driver calls into xenvif_zerocopy_callback 50 * which calls xenvif_skb_zerocopy_complete. 51 */ 52 void xenvif_skb_zerocopy_prepare(struct xenvif_queue *queue, 53 struct sk_buff *skb) 54 { 55 skb_shinfo(skb)->flags |= SKBFL_ZEROCOPY_ENABLE; 56 atomic_inc(&queue->inflight_packets); 57 } 58 59 void xenvif_skb_zerocopy_complete(struct xenvif_queue *queue) 60 { 61 atomic_dec(&queue->inflight_packets); 62 63 /* Wake the dealloc thread _after_ decrementing inflight_packets so 64 * that if kthread_stop() has already been called, the dealloc thread 65 * does not wait forever with nothing to wake it. 66 */ 67 wake_up(&queue->dealloc_wq); 68 } 69 70 static int xenvif_schedulable(struct xenvif *vif) 71 { 72 return netif_running(vif->dev) && 73 test_bit(VIF_STATUS_CONNECTED, &vif->status) && 74 !vif->disabled; 75 } 76 77 static bool xenvif_handle_tx_interrupt(struct xenvif_queue *queue) 78 { 79 bool rc; 80 81 rc = RING_HAS_UNCONSUMED_REQUESTS(&queue->tx); 82 if (rc) 83 napi_schedule(&queue->napi); 84 return rc; 85 } 86 87 static irqreturn_t xenvif_tx_interrupt(int irq, void *dev_id) 88 { 89 struct xenvif_queue *queue = dev_id; 90 int old; 91 92 old = atomic_fetch_or(NETBK_TX_EOI, &queue->eoi_pending); 93 WARN(old & NETBK_TX_EOI, "Interrupt while EOI pending\n"); 94 95 if (!xenvif_handle_tx_interrupt(queue)) { 96 atomic_andnot(NETBK_TX_EOI, &queue->eoi_pending); 97 xen_irq_lateeoi(irq, XEN_EOI_FLAG_SPURIOUS); 98 } 99 100 return IRQ_HANDLED; 101 } 102 103 static int xenvif_poll(struct napi_struct *napi, int budget) 104 { 105 struct xenvif_queue *queue = 106 container_of(napi, struct xenvif_queue, napi); 107 int work_done; 108 109 /* This vif is rogue, we pretend we've there is nothing to do 110 * for this vif to deschedule it from NAPI. But this interface 111 * will be turned off in thread context later. 112 */ 113 if (unlikely(queue->vif->disabled)) { 114 napi_complete(napi); 115 return 0; 116 } 117 118 work_done = xenvif_tx_action(queue, budget); 119 120 if (work_done < budget) { 121 napi_complete_done(napi, work_done); 122 /* If the queue is rate-limited, it shall be 123 * rescheduled in the timer callback. 124 */ 125 if (likely(!queue->rate_limited)) 126 xenvif_napi_schedule_or_enable_events(queue); 127 } 128 129 return work_done; 130 } 131 132 static bool xenvif_handle_rx_interrupt(struct xenvif_queue *queue) 133 { 134 bool rc; 135 136 rc = xenvif_have_rx_work(queue, false); 137 if (rc) 138 xenvif_kick_thread(queue); 139 return rc; 140 } 141 142 static irqreturn_t xenvif_rx_interrupt(int irq, void *dev_id) 143 { 144 struct xenvif_queue *queue = dev_id; 145 int old; 146 147 old = atomic_fetch_or(NETBK_RX_EOI, &queue->eoi_pending); 148 WARN(old & NETBK_RX_EOI, "Interrupt while EOI pending\n"); 149 150 if (!xenvif_handle_rx_interrupt(queue)) { 151 atomic_andnot(NETBK_RX_EOI, &queue->eoi_pending); 152 xen_irq_lateeoi(irq, XEN_EOI_FLAG_SPURIOUS); 153 } 154 155 return IRQ_HANDLED; 156 } 157 158 irqreturn_t xenvif_interrupt(int irq, void *dev_id) 159 { 160 struct xenvif_queue *queue = dev_id; 161 int old; 162 bool has_rx, has_tx; 163 164 old = atomic_fetch_or(NETBK_COMMON_EOI, &queue->eoi_pending); 165 WARN(old, "Interrupt while EOI pending\n"); 166 167 has_tx = xenvif_handle_tx_interrupt(queue); 168 has_rx = xenvif_handle_rx_interrupt(queue); 169 170 if (!has_rx && !has_tx) { 171 atomic_andnot(NETBK_COMMON_EOI, &queue->eoi_pending); 172 xen_irq_lateeoi(irq, XEN_EOI_FLAG_SPURIOUS); 173 } 174 175 return IRQ_HANDLED; 176 } 177 178 static u16 xenvif_select_queue(struct net_device *dev, struct sk_buff *skb, 179 struct net_device *sb_dev) 180 { 181 struct xenvif *vif = netdev_priv(dev); 182 unsigned int size = vif->hash.size; 183 unsigned int num_queues; 184 185 /* If queues are not set up internally - always return 0 186 * as the packet going to be dropped anyway */ 187 num_queues = READ_ONCE(vif->num_queues); 188 if (num_queues < 1) 189 return 0; 190 191 if (vif->hash.alg == XEN_NETIF_CTRL_HASH_ALGORITHM_NONE) 192 return netdev_pick_tx(dev, skb, NULL) % 193 dev->real_num_tx_queues; 194 195 xenvif_set_skb_hash(vif, skb); 196 197 if (size == 0) 198 return skb_get_hash_raw(skb) % dev->real_num_tx_queues; 199 200 return vif->hash.mapping[vif->hash.mapping_sel] 201 [skb_get_hash_raw(skb) % size]; 202 } 203 204 static netdev_tx_t 205 xenvif_start_xmit(struct sk_buff *skb, struct net_device *dev) 206 { 207 struct xenvif *vif = netdev_priv(dev); 208 struct xenvif_queue *queue = NULL; 209 unsigned int num_queues; 210 u16 index; 211 struct xenvif_rx_cb *cb; 212 213 BUG_ON(skb->dev != dev); 214 215 /* Drop the packet if queues are not set up. 216 * This handler should be called inside an RCU read section 217 * so we don't need to enter it here explicitly. 218 */ 219 num_queues = READ_ONCE(vif->num_queues); 220 if (num_queues < 1) 221 goto drop; 222 223 /* Obtain the queue to be used to transmit this packet */ 224 index = skb_get_queue_mapping(skb); 225 if (index >= num_queues) { 226 pr_warn_ratelimited("Invalid queue %hu for packet on interface %s\n", 227 index, vif->dev->name); 228 index %= num_queues; 229 } 230 queue = &vif->queues[index]; 231 232 /* Drop the packet if queue is not ready */ 233 if (queue->task == NULL || 234 queue->dealloc_task == NULL || 235 !xenvif_schedulable(vif)) 236 goto drop; 237 238 if (vif->multicast_control && skb->pkt_type == PACKET_MULTICAST) { 239 struct ethhdr *eth = (struct ethhdr *)skb->data; 240 241 if (!xenvif_mcast_match(vif, eth->h_dest)) 242 goto drop; 243 } 244 245 cb = XENVIF_RX_CB(skb); 246 cb->expires = jiffies + vif->drain_timeout; 247 248 /* If there is no hash algorithm configured then make sure there 249 * is no hash information in the socket buffer otherwise it 250 * would be incorrectly forwarded to the frontend. 251 */ 252 if (vif->hash.alg == XEN_NETIF_CTRL_HASH_ALGORITHM_NONE) 253 skb_clear_hash(skb); 254 255 if (!xenvif_rx_queue_tail(queue, skb)) 256 goto drop; 257 258 xenvif_kick_thread(queue); 259 260 return NETDEV_TX_OK; 261 262 drop: 263 vif->dev->stats.tx_dropped++; 264 dev_kfree_skb_any(skb); 265 return NETDEV_TX_OK; 266 } 267 268 static struct net_device_stats *xenvif_get_stats(struct net_device *dev) 269 { 270 struct xenvif *vif = netdev_priv(dev); 271 struct xenvif_queue *queue = NULL; 272 unsigned int num_queues; 273 u64 rx_bytes = 0; 274 u64 rx_packets = 0; 275 u64 tx_bytes = 0; 276 u64 tx_packets = 0; 277 unsigned int index; 278 279 rcu_read_lock(); 280 num_queues = READ_ONCE(vif->num_queues); 281 282 /* Aggregate tx and rx stats from each queue */ 283 for (index = 0; index < num_queues; ++index) { 284 queue = &vif->queues[index]; 285 rx_bytes += queue->stats.rx_bytes; 286 rx_packets += queue->stats.rx_packets; 287 tx_bytes += queue->stats.tx_bytes; 288 tx_packets += queue->stats.tx_packets; 289 } 290 291 rcu_read_unlock(); 292 293 vif->dev->stats.rx_bytes = rx_bytes; 294 vif->dev->stats.rx_packets = rx_packets; 295 vif->dev->stats.tx_bytes = tx_bytes; 296 vif->dev->stats.tx_packets = tx_packets; 297 298 return &vif->dev->stats; 299 } 300 301 static void xenvif_up(struct xenvif *vif) 302 { 303 struct xenvif_queue *queue = NULL; 304 unsigned int num_queues = vif->num_queues; 305 unsigned int queue_index; 306 307 for (queue_index = 0; queue_index < num_queues; ++queue_index) { 308 queue = &vif->queues[queue_index]; 309 napi_enable(&queue->napi); 310 enable_irq(queue->tx_irq); 311 if (queue->tx_irq != queue->rx_irq) 312 enable_irq(queue->rx_irq); 313 xenvif_napi_schedule_or_enable_events(queue); 314 } 315 } 316 317 static void xenvif_down(struct xenvif *vif) 318 { 319 struct xenvif_queue *queue = NULL; 320 unsigned int num_queues = vif->num_queues; 321 unsigned int queue_index; 322 323 for (queue_index = 0; queue_index < num_queues; ++queue_index) { 324 queue = &vif->queues[queue_index]; 325 disable_irq(queue->tx_irq); 326 if (queue->tx_irq != queue->rx_irq) 327 disable_irq(queue->rx_irq); 328 napi_disable(&queue->napi); 329 del_timer_sync(&queue->credit_timeout); 330 } 331 } 332 333 static int xenvif_open(struct net_device *dev) 334 { 335 struct xenvif *vif = netdev_priv(dev); 336 if (test_bit(VIF_STATUS_CONNECTED, &vif->status)) 337 xenvif_up(vif); 338 netif_tx_start_all_queues(dev); 339 return 0; 340 } 341 342 static int xenvif_close(struct net_device *dev) 343 { 344 struct xenvif *vif = netdev_priv(dev); 345 if (test_bit(VIF_STATUS_CONNECTED, &vif->status)) 346 xenvif_down(vif); 347 netif_tx_stop_all_queues(dev); 348 return 0; 349 } 350 351 static int xenvif_change_mtu(struct net_device *dev, int mtu) 352 { 353 struct xenvif *vif = netdev_priv(dev); 354 int max = vif->can_sg ? ETH_MAX_MTU - VLAN_ETH_HLEN : ETH_DATA_LEN; 355 356 if (mtu > max) 357 return -EINVAL; 358 dev->mtu = mtu; 359 return 0; 360 } 361 362 static netdev_features_t xenvif_fix_features(struct net_device *dev, 363 netdev_features_t features) 364 { 365 struct xenvif *vif = netdev_priv(dev); 366 367 if (!vif->can_sg) 368 features &= ~NETIF_F_SG; 369 if (~(vif->gso_mask) & GSO_BIT(TCPV4)) 370 features &= ~NETIF_F_TSO; 371 if (~(vif->gso_mask) & GSO_BIT(TCPV6)) 372 features &= ~NETIF_F_TSO6; 373 if (!vif->ip_csum) 374 features &= ~NETIF_F_IP_CSUM; 375 if (!vif->ipv6_csum) 376 features &= ~NETIF_F_IPV6_CSUM; 377 378 return features; 379 } 380 381 static const struct xenvif_stat { 382 char name[ETH_GSTRING_LEN]; 383 u16 offset; 384 } xenvif_stats[] = { 385 { 386 "rx_gso_checksum_fixup", 387 offsetof(struct xenvif_stats, rx_gso_checksum_fixup) 388 }, 389 /* If (sent != success + fail), there are probably packets never 390 * freed up properly! 391 */ 392 { 393 "tx_zerocopy_sent", 394 offsetof(struct xenvif_stats, tx_zerocopy_sent), 395 }, 396 { 397 "tx_zerocopy_success", 398 offsetof(struct xenvif_stats, tx_zerocopy_success), 399 }, 400 { 401 "tx_zerocopy_fail", 402 offsetof(struct xenvif_stats, tx_zerocopy_fail) 403 }, 404 /* Number of packets exceeding MAX_SKB_FRAG slots. You should use 405 * a guest with the same MAX_SKB_FRAG 406 */ 407 { 408 "tx_frag_overflow", 409 offsetof(struct xenvif_stats, tx_frag_overflow) 410 }, 411 }; 412 413 static int xenvif_get_sset_count(struct net_device *dev, int string_set) 414 { 415 switch (string_set) { 416 case ETH_SS_STATS: 417 return ARRAY_SIZE(xenvif_stats); 418 default: 419 return -EINVAL; 420 } 421 } 422 423 static void xenvif_get_ethtool_stats(struct net_device *dev, 424 struct ethtool_stats *stats, u64 * data) 425 { 426 struct xenvif *vif = netdev_priv(dev); 427 unsigned int num_queues; 428 int i; 429 unsigned int queue_index; 430 431 rcu_read_lock(); 432 num_queues = READ_ONCE(vif->num_queues); 433 434 for (i = 0; i < ARRAY_SIZE(xenvif_stats); i++) { 435 unsigned long accum = 0; 436 for (queue_index = 0; queue_index < num_queues; ++queue_index) { 437 void *vif_stats = &vif->queues[queue_index].stats; 438 accum += *(unsigned long *)(vif_stats + xenvif_stats[i].offset); 439 } 440 data[i] = accum; 441 } 442 443 rcu_read_unlock(); 444 } 445 446 static void xenvif_get_strings(struct net_device *dev, u32 stringset, u8 * data) 447 { 448 int i; 449 450 switch (stringset) { 451 case ETH_SS_STATS: 452 for (i = 0; i < ARRAY_SIZE(xenvif_stats); i++) 453 memcpy(data + i * ETH_GSTRING_LEN, 454 xenvif_stats[i].name, ETH_GSTRING_LEN); 455 break; 456 } 457 } 458 459 static const struct ethtool_ops xenvif_ethtool_ops = { 460 .get_link = ethtool_op_get_link, 461 462 .get_sset_count = xenvif_get_sset_count, 463 .get_ethtool_stats = xenvif_get_ethtool_stats, 464 .get_strings = xenvif_get_strings, 465 }; 466 467 static const struct net_device_ops xenvif_netdev_ops = { 468 .ndo_select_queue = xenvif_select_queue, 469 .ndo_start_xmit = xenvif_start_xmit, 470 .ndo_get_stats = xenvif_get_stats, 471 .ndo_open = xenvif_open, 472 .ndo_stop = xenvif_close, 473 .ndo_change_mtu = xenvif_change_mtu, 474 .ndo_fix_features = xenvif_fix_features, 475 .ndo_set_mac_address = eth_mac_addr, 476 .ndo_validate_addr = eth_validate_addr, 477 }; 478 479 struct xenvif *xenvif_alloc(struct device *parent, domid_t domid, 480 unsigned int handle) 481 { 482 static const u8 dummy_addr[ETH_ALEN] = { 483 0xfe, 0xff, 0xff, 0xff, 0xff, 0xff, 484 }; 485 int err; 486 struct net_device *dev; 487 struct xenvif *vif; 488 char name[IFNAMSIZ] = {}; 489 490 snprintf(name, IFNAMSIZ - 1, "vif%u.%u", domid, handle); 491 /* Allocate a netdev with the max. supported number of queues. 492 * When the guest selects the desired number, it will be updated 493 * via netif_set_real_num_*_queues(). 494 */ 495 dev = alloc_netdev_mq(sizeof(struct xenvif), name, NET_NAME_UNKNOWN, 496 ether_setup, xenvif_max_queues); 497 if (dev == NULL) { 498 pr_warn("Could not allocate netdev for %s\n", name); 499 return ERR_PTR(-ENOMEM); 500 } 501 502 SET_NETDEV_DEV(dev, parent); 503 504 vif = netdev_priv(dev); 505 506 vif->domid = domid; 507 vif->handle = handle; 508 vif->can_sg = 1; 509 vif->ip_csum = 1; 510 vif->dev = dev; 511 vif->disabled = false; 512 vif->drain_timeout = msecs_to_jiffies(rx_drain_timeout_msecs); 513 vif->stall_timeout = msecs_to_jiffies(rx_stall_timeout_msecs); 514 515 /* Start out with no queues. */ 516 vif->queues = NULL; 517 vif->num_queues = 0; 518 519 vif->xdp_headroom = 0; 520 521 spin_lock_init(&vif->lock); 522 INIT_LIST_HEAD(&vif->fe_mcast_addr); 523 524 dev->netdev_ops = &xenvif_netdev_ops; 525 dev->hw_features = NETIF_F_SG | 526 NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM | 527 NETIF_F_TSO | NETIF_F_TSO6 | NETIF_F_FRAGLIST; 528 dev->features = dev->hw_features | NETIF_F_RXCSUM; 529 dev->ethtool_ops = &xenvif_ethtool_ops; 530 531 dev->min_mtu = ETH_MIN_MTU; 532 dev->max_mtu = ETH_MAX_MTU - VLAN_ETH_HLEN; 533 534 /* 535 * Initialise a dummy MAC address. We choose the numerically 536 * largest non-broadcast address to prevent the address getting 537 * stolen by an Ethernet bridge for STP purposes. 538 * (FE:FF:FF:FF:FF:FF) 539 */ 540 eth_hw_addr_set(dev, dummy_addr); 541 542 netif_carrier_off(dev); 543 544 err = register_netdev(dev); 545 if (err) { 546 netdev_warn(dev, "Could not register device: err=%d\n", err); 547 free_netdev(dev); 548 return ERR_PTR(err); 549 } 550 551 netdev_dbg(dev, "Successfully created xenvif\n"); 552 553 __module_get(THIS_MODULE); 554 555 return vif; 556 } 557 558 int xenvif_init_queue(struct xenvif_queue *queue) 559 { 560 int err, i; 561 562 queue->credit_bytes = queue->remaining_credit = ~0UL; 563 queue->credit_usec = 0UL; 564 timer_setup(&queue->credit_timeout, xenvif_tx_credit_callback, 0); 565 queue->credit_window_start = get_jiffies_64(); 566 567 queue->rx_queue_max = XENVIF_RX_QUEUE_BYTES; 568 569 skb_queue_head_init(&queue->rx_queue); 570 skb_queue_head_init(&queue->tx_queue); 571 572 queue->pending_cons = 0; 573 queue->pending_prod = MAX_PENDING_REQS; 574 for (i = 0; i < MAX_PENDING_REQS; ++i) 575 queue->pending_ring[i] = i; 576 577 spin_lock_init(&queue->callback_lock); 578 spin_lock_init(&queue->response_lock); 579 580 /* If ballooning is disabled, this will consume real memory, so you 581 * better enable it. The long term solution would be to use just a 582 * bunch of valid page descriptors, without dependency on ballooning 583 */ 584 err = gnttab_alloc_pages(MAX_PENDING_REQS, 585 queue->mmap_pages); 586 if (err) { 587 netdev_err(queue->vif->dev, "Could not reserve mmap_pages\n"); 588 return -ENOMEM; 589 } 590 591 for (i = 0; i < MAX_PENDING_REQS; i++) { 592 queue->pending_tx_info[i].callback_struct = (struct ubuf_info_msgzc) 593 { { .callback = xenvif_zerocopy_callback }, 594 { { .ctx = NULL, 595 .desc = i } } }; 596 queue->grant_tx_handle[i] = NETBACK_INVALID_HANDLE; 597 } 598 599 return 0; 600 } 601 602 void xenvif_carrier_on(struct xenvif *vif) 603 { 604 rtnl_lock(); 605 if (!vif->can_sg && vif->dev->mtu > ETH_DATA_LEN) 606 dev_set_mtu(vif->dev, ETH_DATA_LEN); 607 netdev_update_features(vif->dev); 608 set_bit(VIF_STATUS_CONNECTED, &vif->status); 609 if (netif_running(vif->dev)) 610 xenvif_up(vif); 611 rtnl_unlock(); 612 } 613 614 int xenvif_connect_ctrl(struct xenvif *vif, grant_ref_t ring_ref, 615 unsigned int evtchn) 616 { 617 struct net_device *dev = vif->dev; 618 struct xenbus_device *xendev = xenvif_to_xenbus_device(vif); 619 void *addr; 620 struct xen_netif_ctrl_sring *shared; 621 RING_IDX rsp_prod, req_prod; 622 int err; 623 624 err = xenbus_map_ring_valloc(xendev, &ring_ref, 1, &addr); 625 if (err) 626 goto err; 627 628 shared = (struct xen_netif_ctrl_sring *)addr; 629 rsp_prod = READ_ONCE(shared->rsp_prod); 630 req_prod = READ_ONCE(shared->req_prod); 631 632 BACK_RING_ATTACH(&vif->ctrl, shared, rsp_prod, XEN_PAGE_SIZE); 633 634 err = -EIO; 635 if (req_prod - rsp_prod > RING_SIZE(&vif->ctrl)) 636 goto err_unmap; 637 638 err = bind_interdomain_evtchn_to_irq_lateeoi(xendev, evtchn); 639 if (err < 0) 640 goto err_unmap; 641 642 vif->ctrl_irq = err; 643 644 xenvif_init_hash(vif); 645 646 err = request_threaded_irq(vif->ctrl_irq, NULL, xenvif_ctrl_irq_fn, 647 IRQF_ONESHOT, "xen-netback-ctrl", vif); 648 if (err) { 649 pr_warn("Could not setup irq handler for %s\n", dev->name); 650 goto err_deinit; 651 } 652 653 return 0; 654 655 err_deinit: 656 xenvif_deinit_hash(vif); 657 unbind_from_irqhandler(vif->ctrl_irq, vif); 658 vif->ctrl_irq = 0; 659 660 err_unmap: 661 xenbus_unmap_ring_vfree(xendev, vif->ctrl.sring); 662 vif->ctrl.sring = NULL; 663 664 err: 665 return err; 666 } 667 668 static void xenvif_disconnect_queue(struct xenvif_queue *queue) 669 { 670 if (queue->task) { 671 kthread_stop_put(queue->task); 672 queue->task = NULL; 673 } 674 675 if (queue->dealloc_task) { 676 kthread_stop(queue->dealloc_task); 677 queue->dealloc_task = NULL; 678 } 679 680 if (queue->napi.poll) { 681 netif_napi_del(&queue->napi); 682 queue->napi.poll = NULL; 683 } 684 685 if (queue->tx_irq) { 686 unbind_from_irqhandler(queue->tx_irq, queue); 687 if (queue->tx_irq == queue->rx_irq) 688 queue->rx_irq = 0; 689 queue->tx_irq = 0; 690 } 691 692 if (queue->rx_irq) { 693 unbind_from_irqhandler(queue->rx_irq, queue); 694 queue->rx_irq = 0; 695 } 696 697 xenvif_unmap_frontend_data_rings(queue); 698 } 699 700 int xenvif_connect_data(struct xenvif_queue *queue, 701 unsigned long tx_ring_ref, 702 unsigned long rx_ring_ref, 703 unsigned int tx_evtchn, 704 unsigned int rx_evtchn) 705 { 706 struct xenbus_device *dev = xenvif_to_xenbus_device(queue->vif); 707 struct task_struct *task; 708 int err; 709 710 BUG_ON(queue->tx_irq); 711 BUG_ON(queue->task); 712 BUG_ON(queue->dealloc_task); 713 714 err = xenvif_map_frontend_data_rings(queue, tx_ring_ref, 715 rx_ring_ref); 716 if (err < 0) 717 goto err; 718 719 init_waitqueue_head(&queue->wq); 720 init_waitqueue_head(&queue->dealloc_wq); 721 atomic_set(&queue->inflight_packets, 0); 722 723 netif_napi_add(queue->vif->dev, &queue->napi, xenvif_poll); 724 725 queue->stalled = true; 726 727 task = kthread_run(xenvif_kthread_guest_rx, queue, 728 "%s-guest-rx", queue->name); 729 if (IS_ERR(task)) 730 goto kthread_err; 731 queue->task = task; 732 /* 733 * Take a reference to the task in order to prevent it from being freed 734 * if the thread function returns before kthread_stop is called. 735 */ 736 get_task_struct(task); 737 738 task = kthread_run(xenvif_dealloc_kthread, queue, 739 "%s-dealloc", queue->name); 740 if (IS_ERR(task)) 741 goto kthread_err; 742 queue->dealloc_task = task; 743 744 if (tx_evtchn == rx_evtchn) { 745 /* feature-split-event-channels == 0 */ 746 err = bind_interdomain_evtchn_to_irqhandler_lateeoi( 747 dev, tx_evtchn, xenvif_interrupt, 0, 748 queue->name, queue); 749 if (err < 0) 750 goto err; 751 queue->tx_irq = queue->rx_irq = err; 752 disable_irq(queue->tx_irq); 753 } else { 754 /* feature-split-event-channels == 1 */ 755 snprintf(queue->tx_irq_name, sizeof(queue->tx_irq_name), 756 "%s-tx", queue->name); 757 err = bind_interdomain_evtchn_to_irqhandler_lateeoi( 758 dev, tx_evtchn, xenvif_tx_interrupt, 0, 759 queue->tx_irq_name, queue); 760 if (err < 0) 761 goto err; 762 queue->tx_irq = err; 763 disable_irq(queue->tx_irq); 764 765 snprintf(queue->rx_irq_name, sizeof(queue->rx_irq_name), 766 "%s-rx", queue->name); 767 err = bind_interdomain_evtchn_to_irqhandler_lateeoi( 768 dev, rx_evtchn, xenvif_rx_interrupt, 0, 769 queue->rx_irq_name, queue); 770 if (err < 0) 771 goto err; 772 queue->rx_irq = err; 773 disable_irq(queue->rx_irq); 774 } 775 776 return 0; 777 778 kthread_err: 779 pr_warn("Could not allocate kthread for %s\n", queue->name); 780 err = PTR_ERR(task); 781 err: 782 xenvif_disconnect_queue(queue); 783 return err; 784 } 785 786 void xenvif_carrier_off(struct xenvif *vif) 787 { 788 struct net_device *dev = vif->dev; 789 790 rtnl_lock(); 791 if (test_and_clear_bit(VIF_STATUS_CONNECTED, &vif->status)) { 792 netif_carrier_off(dev); /* discard queued packets */ 793 if (netif_running(dev)) 794 xenvif_down(vif); 795 } 796 rtnl_unlock(); 797 } 798 799 void xenvif_disconnect_data(struct xenvif *vif) 800 { 801 struct xenvif_queue *queue = NULL; 802 unsigned int num_queues = vif->num_queues; 803 unsigned int queue_index; 804 805 xenvif_carrier_off(vif); 806 807 for (queue_index = 0; queue_index < num_queues; ++queue_index) { 808 queue = &vif->queues[queue_index]; 809 810 xenvif_disconnect_queue(queue); 811 } 812 813 xenvif_mcast_addr_list_free(vif); 814 } 815 816 void xenvif_disconnect_ctrl(struct xenvif *vif) 817 { 818 if (vif->ctrl_irq) { 819 xenvif_deinit_hash(vif); 820 unbind_from_irqhandler(vif->ctrl_irq, vif); 821 vif->ctrl_irq = 0; 822 } 823 824 if (vif->ctrl.sring) { 825 xenbus_unmap_ring_vfree(xenvif_to_xenbus_device(vif), 826 vif->ctrl.sring); 827 vif->ctrl.sring = NULL; 828 } 829 } 830 831 /* Reverse the relevant parts of xenvif_init_queue(). 832 * Used for queue teardown from xenvif_free(), and on the 833 * error handling paths in xenbus.c:connect(). 834 */ 835 void xenvif_deinit_queue(struct xenvif_queue *queue) 836 { 837 gnttab_free_pages(MAX_PENDING_REQS, queue->mmap_pages); 838 } 839 840 void xenvif_free(struct xenvif *vif) 841 { 842 struct xenvif_queue *queues = vif->queues; 843 unsigned int num_queues = vif->num_queues; 844 unsigned int queue_index; 845 846 unregister_netdev(vif->dev); 847 free_netdev(vif->dev); 848 849 for (queue_index = 0; queue_index < num_queues; ++queue_index) 850 xenvif_deinit_queue(&queues[queue_index]); 851 vfree(queues); 852 853 module_put(THIS_MODULE); 854 } 855