1 /* 2 * Virtual network driver for conversing with remote driver backends. 3 * 4 * Copyright (c) 2002-2005, K A Fraser 5 * Copyright (c) 2005, XenSource Ltd 6 * 7 * This program is free software; you can redistribute it and/or 8 * modify it under the terms of the GNU General Public License version 2 9 * as published by the Free Software Foundation; or, when distributed 10 * separately from the Linux kernel or incorporated into other 11 * software packages, subject to the following license: 12 * 13 * Permission is hereby granted, free of charge, to any person obtaining a copy 14 * of this source file (the "Software"), to deal in the Software without 15 * restriction, including without limitation the rights to use, copy, modify, 16 * merge, publish, distribute, sublicense, and/or sell copies of the Software, 17 * and to permit persons to whom the Software is furnished to do so, subject to 18 * the following conditions: 19 * 20 * The above copyright notice and this permission notice shall be included in 21 * all copies or substantial portions of the Software. 22 * 23 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR 24 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, 25 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE 26 * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER 27 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING 28 * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS 29 * IN THE SOFTWARE. 30 */ 31 32 #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt 33 34 #include <linux/module.h> 35 #include <linux/kernel.h> 36 #include <linux/netdevice.h> 37 #include <linux/etherdevice.h> 38 #include <linux/skbuff.h> 39 #include <linux/ethtool.h> 40 #include <linux/if_ether.h> 41 #include <net/tcp.h> 42 #include <linux/udp.h> 43 #include <linux/moduleparam.h> 44 #include <linux/mm.h> 45 #include <linux/slab.h> 46 #include <net/ip.h> 47 48 #include <xen/xen.h> 49 #include <xen/xenbus.h> 50 #include <xen/events.h> 51 #include <xen/page.h> 52 #include <xen/platform_pci.h> 53 #include <xen/grant_table.h> 54 55 #include <xen/interface/io/netif.h> 56 #include <xen/interface/memory.h> 57 #include <xen/interface/grant_table.h> 58 59 /* Module parameters */ 60 #define MAX_QUEUES_DEFAULT 8 61 static unsigned int xennet_max_queues; 62 module_param_named(max_queues, xennet_max_queues, uint, 0644); 63 MODULE_PARM_DESC(max_queues, 64 "Maximum number of queues per virtual interface"); 65 66 static const struct ethtool_ops xennet_ethtool_ops; 67 68 struct netfront_cb { 69 int pull_to; 70 }; 71 72 #define NETFRONT_SKB_CB(skb) ((struct netfront_cb *)((skb)->cb)) 73 74 #define RX_COPY_THRESHOLD 256 75 76 #define GRANT_INVALID_REF 0 77 78 #define NET_TX_RING_SIZE __CONST_RING_SIZE(xen_netif_tx, XEN_PAGE_SIZE) 79 #define NET_RX_RING_SIZE __CONST_RING_SIZE(xen_netif_rx, XEN_PAGE_SIZE) 80 81 /* Minimum number of Rx slots (includes slot for GSO metadata). */ 82 #define NET_RX_SLOTS_MIN (XEN_NETIF_NR_SLOTS_MIN + 1) 83 84 /* Queue name is interface name with "-qNNN" appended */ 85 #define QUEUE_NAME_SIZE (IFNAMSIZ + 6) 86 87 /* IRQ name is queue name with "-tx" or "-rx" appended */ 88 #define IRQ_NAME_SIZE (QUEUE_NAME_SIZE + 3) 89 90 static DECLARE_WAIT_QUEUE_HEAD(module_unload_q); 91 92 struct netfront_stats { 93 u64 packets; 94 u64 bytes; 95 struct u64_stats_sync syncp; 96 }; 97 98 struct netfront_info; 99 100 struct netfront_queue { 101 unsigned int id; /* Queue ID, 0-based */ 102 char name[QUEUE_NAME_SIZE]; /* DEVNAME-qN */ 103 struct netfront_info *info; 104 105 struct napi_struct napi; 106 107 /* Split event channels support, tx_* == rx_* when using 108 * single event channel. 109 */ 110 unsigned int tx_evtchn, rx_evtchn; 111 unsigned int tx_irq, rx_irq; 112 /* Only used when split event channels support is enabled */ 113 char tx_irq_name[IRQ_NAME_SIZE]; /* DEVNAME-qN-tx */ 114 char rx_irq_name[IRQ_NAME_SIZE]; /* DEVNAME-qN-rx */ 115 116 spinlock_t tx_lock; 117 struct xen_netif_tx_front_ring tx; 118 int tx_ring_ref; 119 120 /* 121 * {tx,rx}_skbs store outstanding skbuffs. Free tx_skb entries 122 * are linked from tx_skb_freelist through skb_entry.link. 123 * 124 * NB. Freelist index entries are always going to be less than 125 * PAGE_OFFSET, whereas pointers to skbs will always be equal or 126 * greater than PAGE_OFFSET: we use this property to distinguish 127 * them. 128 */ 129 union skb_entry { 130 struct sk_buff *skb; 131 unsigned long link; 132 } tx_skbs[NET_TX_RING_SIZE]; 133 grant_ref_t gref_tx_head; 134 grant_ref_t grant_tx_ref[NET_TX_RING_SIZE]; 135 struct page *grant_tx_page[NET_TX_RING_SIZE]; 136 unsigned tx_skb_freelist; 137 138 spinlock_t rx_lock ____cacheline_aligned_in_smp; 139 struct xen_netif_rx_front_ring rx; 140 int rx_ring_ref; 141 142 struct timer_list rx_refill_timer; 143 144 struct sk_buff *rx_skbs[NET_RX_RING_SIZE]; 145 grant_ref_t gref_rx_head; 146 grant_ref_t grant_rx_ref[NET_RX_RING_SIZE]; 147 }; 148 149 struct netfront_info { 150 struct list_head list; 151 struct net_device *netdev; 152 153 struct xenbus_device *xbdev; 154 155 /* Multi-queue support */ 156 struct netfront_queue *queues; 157 158 /* Statistics */ 159 struct netfront_stats __percpu *rx_stats; 160 struct netfront_stats __percpu *tx_stats; 161 162 atomic_t rx_gso_checksum_fixup; 163 }; 164 165 struct netfront_rx_info { 166 struct xen_netif_rx_response rx; 167 struct xen_netif_extra_info extras[XEN_NETIF_EXTRA_TYPE_MAX - 1]; 168 }; 169 170 static void skb_entry_set_link(union skb_entry *list, unsigned short id) 171 { 172 list->link = id; 173 } 174 175 static int skb_entry_is_link(const union skb_entry *list) 176 { 177 BUILD_BUG_ON(sizeof(list->skb) != sizeof(list->link)); 178 return (unsigned long)list->skb < PAGE_OFFSET; 179 } 180 181 /* 182 * Access macros for acquiring freeing slots in tx_skbs[]. 183 */ 184 185 static void add_id_to_freelist(unsigned *head, union skb_entry *list, 186 unsigned short id) 187 { 188 skb_entry_set_link(&list[id], *head); 189 *head = id; 190 } 191 192 static unsigned short get_id_from_freelist(unsigned *head, 193 union skb_entry *list) 194 { 195 unsigned int id = *head; 196 *head = list[id].link; 197 return id; 198 } 199 200 static int xennet_rxidx(RING_IDX idx) 201 { 202 return idx & (NET_RX_RING_SIZE - 1); 203 } 204 205 static struct sk_buff *xennet_get_rx_skb(struct netfront_queue *queue, 206 RING_IDX ri) 207 { 208 int i = xennet_rxidx(ri); 209 struct sk_buff *skb = queue->rx_skbs[i]; 210 queue->rx_skbs[i] = NULL; 211 return skb; 212 } 213 214 static grant_ref_t xennet_get_rx_ref(struct netfront_queue *queue, 215 RING_IDX ri) 216 { 217 int i = xennet_rxidx(ri); 218 grant_ref_t ref = queue->grant_rx_ref[i]; 219 queue->grant_rx_ref[i] = GRANT_INVALID_REF; 220 return ref; 221 } 222 223 #ifdef CONFIG_SYSFS 224 static const struct attribute_group xennet_dev_group; 225 #endif 226 227 static bool xennet_can_sg(struct net_device *dev) 228 { 229 return dev->features & NETIF_F_SG; 230 } 231 232 233 static void rx_refill_timeout(struct timer_list *t) 234 { 235 struct netfront_queue *queue = from_timer(queue, t, rx_refill_timer); 236 napi_schedule(&queue->napi); 237 } 238 239 static int netfront_tx_slot_available(struct netfront_queue *queue) 240 { 241 return (queue->tx.req_prod_pvt - queue->tx.rsp_cons) < 242 (NET_TX_RING_SIZE - XEN_NETIF_NR_SLOTS_MIN - 1); 243 } 244 245 static void xennet_maybe_wake_tx(struct netfront_queue *queue) 246 { 247 struct net_device *dev = queue->info->netdev; 248 struct netdev_queue *dev_queue = netdev_get_tx_queue(dev, queue->id); 249 250 if (unlikely(netif_tx_queue_stopped(dev_queue)) && 251 netfront_tx_slot_available(queue) && 252 likely(netif_running(dev))) 253 netif_tx_wake_queue(netdev_get_tx_queue(dev, queue->id)); 254 } 255 256 257 static struct sk_buff *xennet_alloc_one_rx_buffer(struct netfront_queue *queue) 258 { 259 struct sk_buff *skb; 260 struct page *page; 261 262 skb = __netdev_alloc_skb(queue->info->netdev, 263 RX_COPY_THRESHOLD + NET_IP_ALIGN, 264 GFP_ATOMIC | __GFP_NOWARN); 265 if (unlikely(!skb)) 266 return NULL; 267 268 page = alloc_page(GFP_ATOMIC | __GFP_NOWARN); 269 if (!page) { 270 kfree_skb(skb); 271 return NULL; 272 } 273 skb_add_rx_frag(skb, 0, page, 0, 0, PAGE_SIZE); 274 275 /* Align ip header to a 16 bytes boundary */ 276 skb_reserve(skb, NET_IP_ALIGN); 277 skb->dev = queue->info->netdev; 278 279 return skb; 280 } 281 282 283 static void xennet_alloc_rx_buffers(struct netfront_queue *queue) 284 { 285 RING_IDX req_prod = queue->rx.req_prod_pvt; 286 int notify; 287 int err = 0; 288 289 if (unlikely(!netif_carrier_ok(queue->info->netdev))) 290 return; 291 292 for (req_prod = queue->rx.req_prod_pvt; 293 req_prod - queue->rx.rsp_cons < NET_RX_RING_SIZE; 294 req_prod++) { 295 struct sk_buff *skb; 296 unsigned short id; 297 grant_ref_t ref; 298 struct page *page; 299 struct xen_netif_rx_request *req; 300 301 skb = xennet_alloc_one_rx_buffer(queue); 302 if (!skb) { 303 err = -ENOMEM; 304 break; 305 } 306 307 id = xennet_rxidx(req_prod); 308 309 BUG_ON(queue->rx_skbs[id]); 310 queue->rx_skbs[id] = skb; 311 312 ref = gnttab_claim_grant_reference(&queue->gref_rx_head); 313 WARN_ON_ONCE(IS_ERR_VALUE((unsigned long)(int)ref)); 314 queue->grant_rx_ref[id] = ref; 315 316 page = skb_frag_page(&skb_shinfo(skb)->frags[0]); 317 318 req = RING_GET_REQUEST(&queue->rx, req_prod); 319 gnttab_page_grant_foreign_access_ref_one(ref, 320 queue->info->xbdev->otherend_id, 321 page, 322 0); 323 req->id = id; 324 req->gref = ref; 325 } 326 327 queue->rx.req_prod_pvt = req_prod; 328 329 /* Try again later if there are not enough requests or skb allocation 330 * failed. 331 * Enough requests is quantified as the sum of newly created slots and 332 * the unconsumed slots at the backend. 333 */ 334 if (req_prod - queue->rx.rsp_cons < NET_RX_SLOTS_MIN || 335 unlikely(err)) { 336 mod_timer(&queue->rx_refill_timer, jiffies + (HZ/10)); 337 return; 338 } 339 340 wmb(); /* barrier so backend seens requests */ 341 342 RING_PUSH_REQUESTS_AND_CHECK_NOTIFY(&queue->rx, notify); 343 if (notify) 344 notify_remote_via_irq(queue->rx_irq); 345 } 346 347 static int xennet_open(struct net_device *dev) 348 { 349 struct netfront_info *np = netdev_priv(dev); 350 unsigned int num_queues = dev->real_num_tx_queues; 351 unsigned int i = 0; 352 struct netfront_queue *queue = NULL; 353 354 if (!np->queues) 355 return -ENODEV; 356 357 for (i = 0; i < num_queues; ++i) { 358 queue = &np->queues[i]; 359 napi_enable(&queue->napi); 360 361 spin_lock_bh(&queue->rx_lock); 362 if (netif_carrier_ok(dev)) { 363 xennet_alloc_rx_buffers(queue); 364 queue->rx.sring->rsp_event = queue->rx.rsp_cons + 1; 365 if (RING_HAS_UNCONSUMED_RESPONSES(&queue->rx)) 366 napi_schedule(&queue->napi); 367 } 368 spin_unlock_bh(&queue->rx_lock); 369 } 370 371 netif_tx_start_all_queues(dev); 372 373 return 0; 374 } 375 376 static void xennet_tx_buf_gc(struct netfront_queue *queue) 377 { 378 RING_IDX cons, prod; 379 unsigned short id; 380 struct sk_buff *skb; 381 bool more_to_do; 382 383 BUG_ON(!netif_carrier_ok(queue->info->netdev)); 384 385 do { 386 prod = queue->tx.sring->rsp_prod; 387 rmb(); /* Ensure we see responses up to 'rp'. */ 388 389 for (cons = queue->tx.rsp_cons; cons != prod; cons++) { 390 struct xen_netif_tx_response *txrsp; 391 392 txrsp = RING_GET_RESPONSE(&queue->tx, cons); 393 if (txrsp->status == XEN_NETIF_RSP_NULL) 394 continue; 395 396 id = txrsp->id; 397 skb = queue->tx_skbs[id].skb; 398 if (unlikely(gnttab_query_foreign_access( 399 queue->grant_tx_ref[id]) != 0)) { 400 pr_alert("%s: warning -- grant still in use by backend domain\n", 401 __func__); 402 BUG(); 403 } 404 gnttab_end_foreign_access_ref( 405 queue->grant_tx_ref[id], GNTMAP_readonly); 406 gnttab_release_grant_reference( 407 &queue->gref_tx_head, queue->grant_tx_ref[id]); 408 queue->grant_tx_ref[id] = GRANT_INVALID_REF; 409 queue->grant_tx_page[id] = NULL; 410 add_id_to_freelist(&queue->tx_skb_freelist, queue->tx_skbs, id); 411 dev_kfree_skb_irq(skb); 412 } 413 414 queue->tx.rsp_cons = prod; 415 416 RING_FINAL_CHECK_FOR_RESPONSES(&queue->tx, more_to_do); 417 } while (more_to_do); 418 419 xennet_maybe_wake_tx(queue); 420 } 421 422 struct xennet_gnttab_make_txreq { 423 struct netfront_queue *queue; 424 struct sk_buff *skb; 425 struct page *page; 426 struct xen_netif_tx_request *tx; /* Last request */ 427 unsigned int size; 428 }; 429 430 static void xennet_tx_setup_grant(unsigned long gfn, unsigned int offset, 431 unsigned int len, void *data) 432 { 433 struct xennet_gnttab_make_txreq *info = data; 434 unsigned int id; 435 struct xen_netif_tx_request *tx; 436 grant_ref_t ref; 437 /* convenient aliases */ 438 struct page *page = info->page; 439 struct netfront_queue *queue = info->queue; 440 struct sk_buff *skb = info->skb; 441 442 id = get_id_from_freelist(&queue->tx_skb_freelist, queue->tx_skbs); 443 tx = RING_GET_REQUEST(&queue->tx, queue->tx.req_prod_pvt++); 444 ref = gnttab_claim_grant_reference(&queue->gref_tx_head); 445 WARN_ON_ONCE(IS_ERR_VALUE((unsigned long)(int)ref)); 446 447 gnttab_grant_foreign_access_ref(ref, queue->info->xbdev->otherend_id, 448 gfn, GNTMAP_readonly); 449 450 queue->tx_skbs[id].skb = skb; 451 queue->grant_tx_page[id] = page; 452 queue->grant_tx_ref[id] = ref; 453 454 tx->id = id; 455 tx->gref = ref; 456 tx->offset = offset; 457 tx->size = len; 458 tx->flags = 0; 459 460 info->tx = tx; 461 info->size += tx->size; 462 } 463 464 static struct xen_netif_tx_request *xennet_make_first_txreq( 465 struct netfront_queue *queue, struct sk_buff *skb, 466 struct page *page, unsigned int offset, unsigned int len) 467 { 468 struct xennet_gnttab_make_txreq info = { 469 .queue = queue, 470 .skb = skb, 471 .page = page, 472 .size = 0, 473 }; 474 475 gnttab_for_one_grant(page, offset, len, xennet_tx_setup_grant, &info); 476 477 return info.tx; 478 } 479 480 static void xennet_make_one_txreq(unsigned long gfn, unsigned int offset, 481 unsigned int len, void *data) 482 { 483 struct xennet_gnttab_make_txreq *info = data; 484 485 info->tx->flags |= XEN_NETTXF_more_data; 486 skb_get(info->skb); 487 xennet_tx_setup_grant(gfn, offset, len, data); 488 } 489 490 static struct xen_netif_tx_request *xennet_make_txreqs( 491 struct netfront_queue *queue, struct xen_netif_tx_request *tx, 492 struct sk_buff *skb, struct page *page, 493 unsigned int offset, unsigned int len) 494 { 495 struct xennet_gnttab_make_txreq info = { 496 .queue = queue, 497 .skb = skb, 498 .tx = tx, 499 }; 500 501 /* Skip unused frames from start of page */ 502 page += offset >> PAGE_SHIFT; 503 offset &= ~PAGE_MASK; 504 505 while (len) { 506 info.page = page; 507 info.size = 0; 508 509 gnttab_foreach_grant_in_range(page, offset, len, 510 xennet_make_one_txreq, 511 &info); 512 513 page++; 514 offset = 0; 515 len -= info.size; 516 } 517 518 return info.tx; 519 } 520 521 /* 522 * Count how many ring slots are required to send this skb. Each frag 523 * might be a compound page. 524 */ 525 static int xennet_count_skb_slots(struct sk_buff *skb) 526 { 527 int i, frags = skb_shinfo(skb)->nr_frags; 528 int slots; 529 530 slots = gnttab_count_grant(offset_in_page(skb->data), 531 skb_headlen(skb)); 532 533 for (i = 0; i < frags; i++) { 534 skb_frag_t *frag = skb_shinfo(skb)->frags + i; 535 unsigned long size = skb_frag_size(frag); 536 unsigned long offset = frag->page_offset; 537 538 /* Skip unused frames from start of page */ 539 offset &= ~PAGE_MASK; 540 541 slots += gnttab_count_grant(offset, size); 542 } 543 544 return slots; 545 } 546 547 static u16 xennet_select_queue(struct net_device *dev, struct sk_buff *skb, 548 void *accel_priv, select_queue_fallback_t fallback) 549 { 550 unsigned int num_queues = dev->real_num_tx_queues; 551 u32 hash; 552 u16 queue_idx; 553 554 /* First, check if there is only one queue */ 555 if (num_queues == 1) { 556 queue_idx = 0; 557 } else { 558 hash = skb_get_hash(skb); 559 queue_idx = hash % num_queues; 560 } 561 562 return queue_idx; 563 } 564 565 #define MAX_XEN_SKB_FRAGS (65536 / XEN_PAGE_SIZE + 1) 566 567 static netdev_tx_t xennet_start_xmit(struct sk_buff *skb, struct net_device *dev) 568 { 569 struct netfront_info *np = netdev_priv(dev); 570 struct netfront_stats *tx_stats = this_cpu_ptr(np->tx_stats); 571 struct xen_netif_tx_request *tx, *first_tx; 572 unsigned int i; 573 int notify; 574 int slots; 575 struct page *page; 576 unsigned int offset; 577 unsigned int len; 578 unsigned long flags; 579 struct netfront_queue *queue = NULL; 580 unsigned int num_queues = dev->real_num_tx_queues; 581 u16 queue_index; 582 struct sk_buff *nskb; 583 584 /* Drop the packet if no queues are set up */ 585 if (num_queues < 1) 586 goto drop; 587 /* Determine which queue to transmit this SKB on */ 588 queue_index = skb_get_queue_mapping(skb); 589 queue = &np->queues[queue_index]; 590 591 /* If skb->len is too big for wire format, drop skb and alert 592 * user about misconfiguration. 593 */ 594 if (unlikely(skb->len > XEN_NETIF_MAX_TX_SIZE)) { 595 net_alert_ratelimited( 596 "xennet: skb->len = %u, too big for wire format\n", 597 skb->len); 598 goto drop; 599 } 600 601 slots = xennet_count_skb_slots(skb); 602 if (unlikely(slots > MAX_XEN_SKB_FRAGS + 1)) { 603 net_dbg_ratelimited("xennet: skb rides the rocket: %d slots, %d bytes\n", 604 slots, skb->len); 605 if (skb_linearize(skb)) 606 goto drop; 607 } 608 609 page = virt_to_page(skb->data); 610 offset = offset_in_page(skb->data); 611 612 /* The first req should be at least ETH_HLEN size or the packet will be 613 * dropped by netback. 614 */ 615 if (unlikely(PAGE_SIZE - offset < ETH_HLEN)) { 616 nskb = skb_copy(skb, GFP_ATOMIC); 617 if (!nskb) 618 goto drop; 619 dev_consume_skb_any(skb); 620 skb = nskb; 621 page = virt_to_page(skb->data); 622 offset = offset_in_page(skb->data); 623 } 624 625 len = skb_headlen(skb); 626 627 spin_lock_irqsave(&queue->tx_lock, flags); 628 629 if (unlikely(!netif_carrier_ok(dev) || 630 (slots > 1 && !xennet_can_sg(dev)) || 631 netif_needs_gso(skb, netif_skb_features(skb)))) { 632 spin_unlock_irqrestore(&queue->tx_lock, flags); 633 goto drop; 634 } 635 636 /* First request for the linear area. */ 637 first_tx = tx = xennet_make_first_txreq(queue, skb, 638 page, offset, len); 639 offset += tx->size; 640 if (offset == PAGE_SIZE) { 641 page++; 642 offset = 0; 643 } 644 len -= tx->size; 645 646 if (skb->ip_summed == CHECKSUM_PARTIAL) 647 /* local packet? */ 648 tx->flags |= XEN_NETTXF_csum_blank | XEN_NETTXF_data_validated; 649 else if (skb->ip_summed == CHECKSUM_UNNECESSARY) 650 /* remote but checksummed. */ 651 tx->flags |= XEN_NETTXF_data_validated; 652 653 /* Optional extra info after the first request. */ 654 if (skb_shinfo(skb)->gso_size) { 655 struct xen_netif_extra_info *gso; 656 657 gso = (struct xen_netif_extra_info *) 658 RING_GET_REQUEST(&queue->tx, queue->tx.req_prod_pvt++); 659 660 tx->flags |= XEN_NETTXF_extra_info; 661 662 gso->u.gso.size = skb_shinfo(skb)->gso_size; 663 gso->u.gso.type = (skb_shinfo(skb)->gso_type & SKB_GSO_TCPV6) ? 664 XEN_NETIF_GSO_TYPE_TCPV6 : 665 XEN_NETIF_GSO_TYPE_TCPV4; 666 gso->u.gso.pad = 0; 667 gso->u.gso.features = 0; 668 669 gso->type = XEN_NETIF_EXTRA_TYPE_GSO; 670 gso->flags = 0; 671 } 672 673 /* Requests for the rest of the linear area. */ 674 tx = xennet_make_txreqs(queue, tx, skb, page, offset, len); 675 676 /* Requests for all the frags. */ 677 for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) { 678 skb_frag_t *frag = &skb_shinfo(skb)->frags[i]; 679 tx = xennet_make_txreqs(queue, tx, skb, 680 skb_frag_page(frag), frag->page_offset, 681 skb_frag_size(frag)); 682 } 683 684 /* First request has the packet length. */ 685 first_tx->size = skb->len; 686 687 RING_PUSH_REQUESTS_AND_CHECK_NOTIFY(&queue->tx, notify); 688 if (notify) 689 notify_remote_via_irq(queue->tx_irq); 690 691 u64_stats_update_begin(&tx_stats->syncp); 692 tx_stats->bytes += skb->len; 693 tx_stats->packets++; 694 u64_stats_update_end(&tx_stats->syncp); 695 696 /* Note: It is not safe to access skb after xennet_tx_buf_gc()! */ 697 xennet_tx_buf_gc(queue); 698 699 if (!netfront_tx_slot_available(queue)) 700 netif_tx_stop_queue(netdev_get_tx_queue(dev, queue->id)); 701 702 spin_unlock_irqrestore(&queue->tx_lock, flags); 703 704 return NETDEV_TX_OK; 705 706 drop: 707 dev->stats.tx_dropped++; 708 dev_kfree_skb_any(skb); 709 return NETDEV_TX_OK; 710 } 711 712 static int xennet_close(struct net_device *dev) 713 { 714 struct netfront_info *np = netdev_priv(dev); 715 unsigned int num_queues = dev->real_num_tx_queues; 716 unsigned int i; 717 struct netfront_queue *queue; 718 netif_tx_stop_all_queues(np->netdev); 719 for (i = 0; i < num_queues; ++i) { 720 queue = &np->queues[i]; 721 napi_disable(&queue->napi); 722 } 723 return 0; 724 } 725 726 static void xennet_move_rx_slot(struct netfront_queue *queue, struct sk_buff *skb, 727 grant_ref_t ref) 728 { 729 int new = xennet_rxidx(queue->rx.req_prod_pvt); 730 731 BUG_ON(queue->rx_skbs[new]); 732 queue->rx_skbs[new] = skb; 733 queue->grant_rx_ref[new] = ref; 734 RING_GET_REQUEST(&queue->rx, queue->rx.req_prod_pvt)->id = new; 735 RING_GET_REQUEST(&queue->rx, queue->rx.req_prod_pvt)->gref = ref; 736 queue->rx.req_prod_pvt++; 737 } 738 739 static int xennet_get_extras(struct netfront_queue *queue, 740 struct xen_netif_extra_info *extras, 741 RING_IDX rp) 742 743 { 744 struct xen_netif_extra_info *extra; 745 struct device *dev = &queue->info->netdev->dev; 746 RING_IDX cons = queue->rx.rsp_cons; 747 int err = 0; 748 749 do { 750 struct sk_buff *skb; 751 grant_ref_t ref; 752 753 if (unlikely(cons + 1 == rp)) { 754 if (net_ratelimit()) 755 dev_warn(dev, "Missing extra info\n"); 756 err = -EBADR; 757 break; 758 } 759 760 extra = (struct xen_netif_extra_info *) 761 RING_GET_RESPONSE(&queue->rx, ++cons); 762 763 if (unlikely(!extra->type || 764 extra->type >= XEN_NETIF_EXTRA_TYPE_MAX)) { 765 if (net_ratelimit()) 766 dev_warn(dev, "Invalid extra type: %d\n", 767 extra->type); 768 err = -EINVAL; 769 } else { 770 memcpy(&extras[extra->type - 1], extra, 771 sizeof(*extra)); 772 } 773 774 skb = xennet_get_rx_skb(queue, cons); 775 ref = xennet_get_rx_ref(queue, cons); 776 xennet_move_rx_slot(queue, skb, ref); 777 } while (extra->flags & XEN_NETIF_EXTRA_FLAG_MORE); 778 779 queue->rx.rsp_cons = cons; 780 return err; 781 } 782 783 static int xennet_get_responses(struct netfront_queue *queue, 784 struct netfront_rx_info *rinfo, RING_IDX rp, 785 struct sk_buff_head *list) 786 { 787 struct xen_netif_rx_response *rx = &rinfo->rx; 788 struct xen_netif_extra_info *extras = rinfo->extras; 789 struct device *dev = &queue->info->netdev->dev; 790 RING_IDX cons = queue->rx.rsp_cons; 791 struct sk_buff *skb = xennet_get_rx_skb(queue, cons); 792 grant_ref_t ref = xennet_get_rx_ref(queue, cons); 793 int max = XEN_NETIF_NR_SLOTS_MIN + (rx->status <= RX_COPY_THRESHOLD); 794 int slots = 1; 795 int err = 0; 796 unsigned long ret; 797 798 if (rx->flags & XEN_NETRXF_extra_info) { 799 err = xennet_get_extras(queue, extras, rp); 800 cons = queue->rx.rsp_cons; 801 } 802 803 for (;;) { 804 if (unlikely(rx->status < 0 || 805 rx->offset + rx->status > XEN_PAGE_SIZE)) { 806 if (net_ratelimit()) 807 dev_warn(dev, "rx->offset: %u, size: %d\n", 808 rx->offset, rx->status); 809 xennet_move_rx_slot(queue, skb, ref); 810 err = -EINVAL; 811 goto next; 812 } 813 814 /* 815 * This definitely indicates a bug, either in this driver or in 816 * the backend driver. In future this should flag the bad 817 * situation to the system controller to reboot the backend. 818 */ 819 if (ref == GRANT_INVALID_REF) { 820 if (net_ratelimit()) 821 dev_warn(dev, "Bad rx response id %d.\n", 822 rx->id); 823 err = -EINVAL; 824 goto next; 825 } 826 827 ret = gnttab_end_foreign_access_ref(ref, 0); 828 BUG_ON(!ret); 829 830 gnttab_release_grant_reference(&queue->gref_rx_head, ref); 831 832 __skb_queue_tail(list, skb); 833 834 next: 835 if (!(rx->flags & XEN_NETRXF_more_data)) 836 break; 837 838 if (cons + slots == rp) { 839 if (net_ratelimit()) 840 dev_warn(dev, "Need more slots\n"); 841 err = -ENOENT; 842 break; 843 } 844 845 rx = RING_GET_RESPONSE(&queue->rx, cons + slots); 846 skb = xennet_get_rx_skb(queue, cons + slots); 847 ref = xennet_get_rx_ref(queue, cons + slots); 848 slots++; 849 } 850 851 if (unlikely(slots > max)) { 852 if (net_ratelimit()) 853 dev_warn(dev, "Too many slots\n"); 854 err = -E2BIG; 855 } 856 857 if (unlikely(err)) 858 queue->rx.rsp_cons = cons + slots; 859 860 return err; 861 } 862 863 static int xennet_set_skb_gso(struct sk_buff *skb, 864 struct xen_netif_extra_info *gso) 865 { 866 if (!gso->u.gso.size) { 867 if (net_ratelimit()) 868 pr_warn("GSO size must not be zero\n"); 869 return -EINVAL; 870 } 871 872 if (gso->u.gso.type != XEN_NETIF_GSO_TYPE_TCPV4 && 873 gso->u.gso.type != XEN_NETIF_GSO_TYPE_TCPV6) { 874 if (net_ratelimit()) 875 pr_warn("Bad GSO type %d\n", gso->u.gso.type); 876 return -EINVAL; 877 } 878 879 skb_shinfo(skb)->gso_size = gso->u.gso.size; 880 skb_shinfo(skb)->gso_type = 881 (gso->u.gso.type == XEN_NETIF_GSO_TYPE_TCPV4) ? 882 SKB_GSO_TCPV4 : 883 SKB_GSO_TCPV6; 884 885 /* Header must be checked, and gso_segs computed. */ 886 skb_shinfo(skb)->gso_type |= SKB_GSO_DODGY; 887 skb_shinfo(skb)->gso_segs = 0; 888 889 return 0; 890 } 891 892 static RING_IDX xennet_fill_frags(struct netfront_queue *queue, 893 struct sk_buff *skb, 894 struct sk_buff_head *list) 895 { 896 struct skb_shared_info *shinfo = skb_shinfo(skb); 897 RING_IDX cons = queue->rx.rsp_cons; 898 struct sk_buff *nskb; 899 900 while ((nskb = __skb_dequeue(list))) { 901 struct xen_netif_rx_response *rx = 902 RING_GET_RESPONSE(&queue->rx, ++cons); 903 skb_frag_t *nfrag = &skb_shinfo(nskb)->frags[0]; 904 905 if (shinfo->nr_frags == MAX_SKB_FRAGS) { 906 unsigned int pull_to = NETFRONT_SKB_CB(skb)->pull_to; 907 908 BUG_ON(pull_to <= skb_headlen(skb)); 909 __pskb_pull_tail(skb, pull_to - skb_headlen(skb)); 910 } 911 BUG_ON(shinfo->nr_frags >= MAX_SKB_FRAGS); 912 913 skb_add_rx_frag(skb, shinfo->nr_frags, skb_frag_page(nfrag), 914 rx->offset, rx->status, PAGE_SIZE); 915 916 skb_shinfo(nskb)->nr_frags = 0; 917 kfree_skb(nskb); 918 } 919 920 return cons; 921 } 922 923 static int checksum_setup(struct net_device *dev, struct sk_buff *skb) 924 { 925 bool recalculate_partial_csum = false; 926 927 /* 928 * A GSO SKB must be CHECKSUM_PARTIAL. However some buggy 929 * peers can fail to set NETRXF_csum_blank when sending a GSO 930 * frame. In this case force the SKB to CHECKSUM_PARTIAL and 931 * recalculate the partial checksum. 932 */ 933 if (skb->ip_summed != CHECKSUM_PARTIAL && skb_is_gso(skb)) { 934 struct netfront_info *np = netdev_priv(dev); 935 atomic_inc(&np->rx_gso_checksum_fixup); 936 skb->ip_summed = CHECKSUM_PARTIAL; 937 recalculate_partial_csum = true; 938 } 939 940 /* A non-CHECKSUM_PARTIAL SKB does not require setup. */ 941 if (skb->ip_summed != CHECKSUM_PARTIAL) 942 return 0; 943 944 return skb_checksum_setup(skb, recalculate_partial_csum); 945 } 946 947 static int handle_incoming_queue(struct netfront_queue *queue, 948 struct sk_buff_head *rxq) 949 { 950 struct netfront_stats *rx_stats = this_cpu_ptr(queue->info->rx_stats); 951 int packets_dropped = 0; 952 struct sk_buff *skb; 953 954 while ((skb = __skb_dequeue(rxq)) != NULL) { 955 int pull_to = NETFRONT_SKB_CB(skb)->pull_to; 956 957 if (pull_to > skb_headlen(skb)) 958 __pskb_pull_tail(skb, pull_to - skb_headlen(skb)); 959 960 /* Ethernet work: Delayed to here as it peeks the header. */ 961 skb->protocol = eth_type_trans(skb, queue->info->netdev); 962 skb_reset_network_header(skb); 963 964 if (checksum_setup(queue->info->netdev, skb)) { 965 kfree_skb(skb); 966 packets_dropped++; 967 queue->info->netdev->stats.rx_errors++; 968 continue; 969 } 970 971 u64_stats_update_begin(&rx_stats->syncp); 972 rx_stats->packets++; 973 rx_stats->bytes += skb->len; 974 u64_stats_update_end(&rx_stats->syncp); 975 976 /* Pass it up. */ 977 napi_gro_receive(&queue->napi, skb); 978 } 979 980 return packets_dropped; 981 } 982 983 static int xennet_poll(struct napi_struct *napi, int budget) 984 { 985 struct netfront_queue *queue = container_of(napi, struct netfront_queue, napi); 986 struct net_device *dev = queue->info->netdev; 987 struct sk_buff *skb; 988 struct netfront_rx_info rinfo; 989 struct xen_netif_rx_response *rx = &rinfo.rx; 990 struct xen_netif_extra_info *extras = rinfo.extras; 991 RING_IDX i, rp; 992 int work_done; 993 struct sk_buff_head rxq; 994 struct sk_buff_head errq; 995 struct sk_buff_head tmpq; 996 int err; 997 998 spin_lock(&queue->rx_lock); 999 1000 skb_queue_head_init(&rxq); 1001 skb_queue_head_init(&errq); 1002 skb_queue_head_init(&tmpq); 1003 1004 rp = queue->rx.sring->rsp_prod; 1005 rmb(); /* Ensure we see queued responses up to 'rp'. */ 1006 1007 i = queue->rx.rsp_cons; 1008 work_done = 0; 1009 while ((i != rp) && (work_done < budget)) { 1010 memcpy(rx, RING_GET_RESPONSE(&queue->rx, i), sizeof(*rx)); 1011 memset(extras, 0, sizeof(rinfo.extras)); 1012 1013 err = xennet_get_responses(queue, &rinfo, rp, &tmpq); 1014 1015 if (unlikely(err)) { 1016 err: 1017 while ((skb = __skb_dequeue(&tmpq))) 1018 __skb_queue_tail(&errq, skb); 1019 dev->stats.rx_errors++; 1020 i = queue->rx.rsp_cons; 1021 continue; 1022 } 1023 1024 skb = __skb_dequeue(&tmpq); 1025 1026 if (extras[XEN_NETIF_EXTRA_TYPE_GSO - 1].type) { 1027 struct xen_netif_extra_info *gso; 1028 gso = &extras[XEN_NETIF_EXTRA_TYPE_GSO - 1]; 1029 1030 if (unlikely(xennet_set_skb_gso(skb, gso))) { 1031 __skb_queue_head(&tmpq, skb); 1032 queue->rx.rsp_cons += skb_queue_len(&tmpq); 1033 goto err; 1034 } 1035 } 1036 1037 NETFRONT_SKB_CB(skb)->pull_to = rx->status; 1038 if (NETFRONT_SKB_CB(skb)->pull_to > RX_COPY_THRESHOLD) 1039 NETFRONT_SKB_CB(skb)->pull_to = RX_COPY_THRESHOLD; 1040 1041 skb_shinfo(skb)->frags[0].page_offset = rx->offset; 1042 skb_frag_size_set(&skb_shinfo(skb)->frags[0], rx->status); 1043 skb->data_len = rx->status; 1044 skb->len += rx->status; 1045 1046 i = xennet_fill_frags(queue, skb, &tmpq); 1047 1048 if (rx->flags & XEN_NETRXF_csum_blank) 1049 skb->ip_summed = CHECKSUM_PARTIAL; 1050 else if (rx->flags & XEN_NETRXF_data_validated) 1051 skb->ip_summed = CHECKSUM_UNNECESSARY; 1052 1053 __skb_queue_tail(&rxq, skb); 1054 1055 queue->rx.rsp_cons = ++i; 1056 work_done++; 1057 } 1058 1059 __skb_queue_purge(&errq); 1060 1061 work_done -= handle_incoming_queue(queue, &rxq); 1062 1063 xennet_alloc_rx_buffers(queue); 1064 1065 if (work_done < budget) { 1066 int more_to_do = 0; 1067 1068 napi_complete_done(napi, work_done); 1069 1070 RING_FINAL_CHECK_FOR_RESPONSES(&queue->rx, more_to_do); 1071 if (more_to_do) 1072 napi_schedule(napi); 1073 } 1074 1075 spin_unlock(&queue->rx_lock); 1076 1077 return work_done; 1078 } 1079 1080 static int xennet_change_mtu(struct net_device *dev, int mtu) 1081 { 1082 int max = xennet_can_sg(dev) ? XEN_NETIF_MAX_TX_SIZE : ETH_DATA_LEN; 1083 1084 if (mtu > max) 1085 return -EINVAL; 1086 dev->mtu = mtu; 1087 return 0; 1088 } 1089 1090 static void xennet_get_stats64(struct net_device *dev, 1091 struct rtnl_link_stats64 *tot) 1092 { 1093 struct netfront_info *np = netdev_priv(dev); 1094 int cpu; 1095 1096 for_each_possible_cpu(cpu) { 1097 struct netfront_stats *rx_stats = per_cpu_ptr(np->rx_stats, cpu); 1098 struct netfront_stats *tx_stats = per_cpu_ptr(np->tx_stats, cpu); 1099 u64 rx_packets, rx_bytes, tx_packets, tx_bytes; 1100 unsigned int start; 1101 1102 do { 1103 start = u64_stats_fetch_begin_irq(&tx_stats->syncp); 1104 tx_packets = tx_stats->packets; 1105 tx_bytes = tx_stats->bytes; 1106 } while (u64_stats_fetch_retry_irq(&tx_stats->syncp, start)); 1107 1108 do { 1109 start = u64_stats_fetch_begin_irq(&rx_stats->syncp); 1110 rx_packets = rx_stats->packets; 1111 rx_bytes = rx_stats->bytes; 1112 } while (u64_stats_fetch_retry_irq(&rx_stats->syncp, start)); 1113 1114 tot->rx_packets += rx_packets; 1115 tot->tx_packets += tx_packets; 1116 tot->rx_bytes += rx_bytes; 1117 tot->tx_bytes += tx_bytes; 1118 } 1119 1120 tot->rx_errors = dev->stats.rx_errors; 1121 tot->tx_dropped = dev->stats.tx_dropped; 1122 } 1123 1124 static void xennet_release_tx_bufs(struct netfront_queue *queue) 1125 { 1126 struct sk_buff *skb; 1127 int i; 1128 1129 for (i = 0; i < NET_TX_RING_SIZE; i++) { 1130 /* Skip over entries which are actually freelist references */ 1131 if (skb_entry_is_link(&queue->tx_skbs[i])) 1132 continue; 1133 1134 skb = queue->tx_skbs[i].skb; 1135 get_page(queue->grant_tx_page[i]); 1136 gnttab_end_foreign_access(queue->grant_tx_ref[i], 1137 GNTMAP_readonly, 1138 (unsigned long)page_address(queue->grant_tx_page[i])); 1139 queue->grant_tx_page[i] = NULL; 1140 queue->grant_tx_ref[i] = GRANT_INVALID_REF; 1141 add_id_to_freelist(&queue->tx_skb_freelist, queue->tx_skbs, i); 1142 dev_kfree_skb_irq(skb); 1143 } 1144 } 1145 1146 static void xennet_release_rx_bufs(struct netfront_queue *queue) 1147 { 1148 int id, ref; 1149 1150 spin_lock_bh(&queue->rx_lock); 1151 1152 for (id = 0; id < NET_RX_RING_SIZE; id++) { 1153 struct sk_buff *skb; 1154 struct page *page; 1155 1156 skb = queue->rx_skbs[id]; 1157 if (!skb) 1158 continue; 1159 1160 ref = queue->grant_rx_ref[id]; 1161 if (ref == GRANT_INVALID_REF) 1162 continue; 1163 1164 page = skb_frag_page(&skb_shinfo(skb)->frags[0]); 1165 1166 /* gnttab_end_foreign_access() needs a page ref until 1167 * foreign access is ended (which may be deferred). 1168 */ 1169 get_page(page); 1170 gnttab_end_foreign_access(ref, 0, 1171 (unsigned long)page_address(page)); 1172 queue->grant_rx_ref[id] = GRANT_INVALID_REF; 1173 1174 kfree_skb(skb); 1175 } 1176 1177 spin_unlock_bh(&queue->rx_lock); 1178 } 1179 1180 static netdev_features_t xennet_fix_features(struct net_device *dev, 1181 netdev_features_t features) 1182 { 1183 struct netfront_info *np = netdev_priv(dev); 1184 1185 if (features & NETIF_F_SG && 1186 !xenbus_read_unsigned(np->xbdev->otherend, "feature-sg", 0)) 1187 features &= ~NETIF_F_SG; 1188 1189 if (features & NETIF_F_IPV6_CSUM && 1190 !xenbus_read_unsigned(np->xbdev->otherend, 1191 "feature-ipv6-csum-offload", 0)) 1192 features &= ~NETIF_F_IPV6_CSUM; 1193 1194 if (features & NETIF_F_TSO && 1195 !xenbus_read_unsigned(np->xbdev->otherend, "feature-gso-tcpv4", 0)) 1196 features &= ~NETIF_F_TSO; 1197 1198 if (features & NETIF_F_TSO6 && 1199 !xenbus_read_unsigned(np->xbdev->otherend, "feature-gso-tcpv6", 0)) 1200 features &= ~NETIF_F_TSO6; 1201 1202 return features; 1203 } 1204 1205 static int xennet_set_features(struct net_device *dev, 1206 netdev_features_t features) 1207 { 1208 if (!(features & NETIF_F_SG) && dev->mtu > ETH_DATA_LEN) { 1209 netdev_info(dev, "Reducing MTU because no SG offload"); 1210 dev->mtu = ETH_DATA_LEN; 1211 } 1212 1213 return 0; 1214 } 1215 1216 static irqreturn_t xennet_tx_interrupt(int irq, void *dev_id) 1217 { 1218 struct netfront_queue *queue = dev_id; 1219 unsigned long flags; 1220 1221 spin_lock_irqsave(&queue->tx_lock, flags); 1222 xennet_tx_buf_gc(queue); 1223 spin_unlock_irqrestore(&queue->tx_lock, flags); 1224 1225 return IRQ_HANDLED; 1226 } 1227 1228 static irqreturn_t xennet_rx_interrupt(int irq, void *dev_id) 1229 { 1230 struct netfront_queue *queue = dev_id; 1231 struct net_device *dev = queue->info->netdev; 1232 1233 if (likely(netif_carrier_ok(dev) && 1234 RING_HAS_UNCONSUMED_RESPONSES(&queue->rx))) 1235 napi_schedule(&queue->napi); 1236 1237 return IRQ_HANDLED; 1238 } 1239 1240 static irqreturn_t xennet_interrupt(int irq, void *dev_id) 1241 { 1242 xennet_tx_interrupt(irq, dev_id); 1243 xennet_rx_interrupt(irq, dev_id); 1244 return IRQ_HANDLED; 1245 } 1246 1247 #ifdef CONFIG_NET_POLL_CONTROLLER 1248 static void xennet_poll_controller(struct net_device *dev) 1249 { 1250 /* Poll each queue */ 1251 struct netfront_info *info = netdev_priv(dev); 1252 unsigned int num_queues = dev->real_num_tx_queues; 1253 unsigned int i; 1254 for (i = 0; i < num_queues; ++i) 1255 xennet_interrupt(0, &info->queues[i]); 1256 } 1257 #endif 1258 1259 static const struct net_device_ops xennet_netdev_ops = { 1260 .ndo_open = xennet_open, 1261 .ndo_stop = xennet_close, 1262 .ndo_start_xmit = xennet_start_xmit, 1263 .ndo_change_mtu = xennet_change_mtu, 1264 .ndo_get_stats64 = xennet_get_stats64, 1265 .ndo_set_mac_address = eth_mac_addr, 1266 .ndo_validate_addr = eth_validate_addr, 1267 .ndo_fix_features = xennet_fix_features, 1268 .ndo_set_features = xennet_set_features, 1269 .ndo_select_queue = xennet_select_queue, 1270 #ifdef CONFIG_NET_POLL_CONTROLLER 1271 .ndo_poll_controller = xennet_poll_controller, 1272 #endif 1273 }; 1274 1275 static void xennet_free_netdev(struct net_device *netdev) 1276 { 1277 struct netfront_info *np = netdev_priv(netdev); 1278 1279 free_percpu(np->rx_stats); 1280 free_percpu(np->tx_stats); 1281 free_netdev(netdev); 1282 } 1283 1284 static struct net_device *xennet_create_dev(struct xenbus_device *dev) 1285 { 1286 int err; 1287 struct net_device *netdev; 1288 struct netfront_info *np; 1289 1290 netdev = alloc_etherdev_mq(sizeof(struct netfront_info), xennet_max_queues); 1291 if (!netdev) 1292 return ERR_PTR(-ENOMEM); 1293 1294 np = netdev_priv(netdev); 1295 np->xbdev = dev; 1296 1297 np->queues = NULL; 1298 1299 err = -ENOMEM; 1300 np->rx_stats = netdev_alloc_pcpu_stats(struct netfront_stats); 1301 if (np->rx_stats == NULL) 1302 goto exit; 1303 np->tx_stats = netdev_alloc_pcpu_stats(struct netfront_stats); 1304 if (np->tx_stats == NULL) 1305 goto exit; 1306 1307 netdev->netdev_ops = &xennet_netdev_ops; 1308 1309 netdev->features = NETIF_F_IP_CSUM | NETIF_F_RXCSUM | 1310 NETIF_F_GSO_ROBUST; 1311 netdev->hw_features = NETIF_F_SG | 1312 NETIF_F_IPV6_CSUM | 1313 NETIF_F_TSO | NETIF_F_TSO6; 1314 1315 /* 1316 * Assume that all hw features are available for now. This set 1317 * will be adjusted by the call to netdev_update_features() in 1318 * xennet_connect() which is the earliest point where we can 1319 * negotiate with the backend regarding supported features. 1320 */ 1321 netdev->features |= netdev->hw_features; 1322 1323 netdev->ethtool_ops = &xennet_ethtool_ops; 1324 netdev->min_mtu = ETH_MIN_MTU; 1325 netdev->max_mtu = XEN_NETIF_MAX_TX_SIZE; 1326 SET_NETDEV_DEV(netdev, &dev->dev); 1327 1328 np->netdev = netdev; 1329 1330 netif_carrier_off(netdev); 1331 1332 xenbus_switch_state(dev, XenbusStateInitialising); 1333 return netdev; 1334 1335 exit: 1336 xennet_free_netdev(netdev); 1337 return ERR_PTR(err); 1338 } 1339 1340 /** 1341 * Entry point to this code when a new device is created. Allocate the basic 1342 * structures and the ring buffers for communication with the backend, and 1343 * inform the backend of the appropriate details for those. 1344 */ 1345 static int netfront_probe(struct xenbus_device *dev, 1346 const struct xenbus_device_id *id) 1347 { 1348 int err; 1349 struct net_device *netdev; 1350 struct netfront_info *info; 1351 1352 netdev = xennet_create_dev(dev); 1353 if (IS_ERR(netdev)) { 1354 err = PTR_ERR(netdev); 1355 xenbus_dev_fatal(dev, err, "creating netdev"); 1356 return err; 1357 } 1358 1359 info = netdev_priv(netdev); 1360 dev_set_drvdata(&dev->dev, info); 1361 #ifdef CONFIG_SYSFS 1362 info->netdev->sysfs_groups[0] = &xennet_dev_group; 1363 #endif 1364 1365 return 0; 1366 } 1367 1368 static void xennet_end_access(int ref, void *page) 1369 { 1370 /* This frees the page as a side-effect */ 1371 if (ref != GRANT_INVALID_REF) 1372 gnttab_end_foreign_access(ref, 0, (unsigned long)page); 1373 } 1374 1375 static void xennet_disconnect_backend(struct netfront_info *info) 1376 { 1377 unsigned int i = 0; 1378 unsigned int num_queues = info->netdev->real_num_tx_queues; 1379 1380 netif_carrier_off(info->netdev); 1381 1382 for (i = 0; i < num_queues && info->queues; ++i) { 1383 struct netfront_queue *queue = &info->queues[i]; 1384 1385 del_timer_sync(&queue->rx_refill_timer); 1386 1387 if (queue->tx_irq && (queue->tx_irq == queue->rx_irq)) 1388 unbind_from_irqhandler(queue->tx_irq, queue); 1389 if (queue->tx_irq && (queue->tx_irq != queue->rx_irq)) { 1390 unbind_from_irqhandler(queue->tx_irq, queue); 1391 unbind_from_irqhandler(queue->rx_irq, queue); 1392 } 1393 queue->tx_evtchn = queue->rx_evtchn = 0; 1394 queue->tx_irq = queue->rx_irq = 0; 1395 1396 if (netif_running(info->netdev)) 1397 napi_synchronize(&queue->napi); 1398 1399 xennet_release_tx_bufs(queue); 1400 xennet_release_rx_bufs(queue); 1401 gnttab_free_grant_references(queue->gref_tx_head); 1402 gnttab_free_grant_references(queue->gref_rx_head); 1403 1404 /* End access and free the pages */ 1405 xennet_end_access(queue->tx_ring_ref, queue->tx.sring); 1406 xennet_end_access(queue->rx_ring_ref, queue->rx.sring); 1407 1408 queue->tx_ring_ref = GRANT_INVALID_REF; 1409 queue->rx_ring_ref = GRANT_INVALID_REF; 1410 queue->tx.sring = NULL; 1411 queue->rx.sring = NULL; 1412 } 1413 } 1414 1415 /** 1416 * We are reconnecting to the backend, due to a suspend/resume, or a backend 1417 * driver restart. We tear down our netif structure and recreate it, but 1418 * leave the device-layer structures intact so that this is transparent to the 1419 * rest of the kernel. 1420 */ 1421 static int netfront_resume(struct xenbus_device *dev) 1422 { 1423 struct netfront_info *info = dev_get_drvdata(&dev->dev); 1424 1425 dev_dbg(&dev->dev, "%s\n", dev->nodename); 1426 1427 xennet_disconnect_backend(info); 1428 return 0; 1429 } 1430 1431 static int xen_net_read_mac(struct xenbus_device *dev, u8 mac[]) 1432 { 1433 char *s, *e, *macstr; 1434 int i; 1435 1436 macstr = s = xenbus_read(XBT_NIL, dev->nodename, "mac", NULL); 1437 if (IS_ERR(macstr)) 1438 return PTR_ERR(macstr); 1439 1440 for (i = 0; i < ETH_ALEN; i++) { 1441 mac[i] = simple_strtoul(s, &e, 16); 1442 if ((s == e) || (*e != ((i == ETH_ALEN-1) ? '\0' : ':'))) { 1443 kfree(macstr); 1444 return -ENOENT; 1445 } 1446 s = e+1; 1447 } 1448 1449 kfree(macstr); 1450 return 0; 1451 } 1452 1453 static int setup_netfront_single(struct netfront_queue *queue) 1454 { 1455 int err; 1456 1457 err = xenbus_alloc_evtchn(queue->info->xbdev, &queue->tx_evtchn); 1458 if (err < 0) 1459 goto fail; 1460 1461 err = bind_evtchn_to_irqhandler(queue->tx_evtchn, 1462 xennet_interrupt, 1463 0, queue->info->netdev->name, queue); 1464 if (err < 0) 1465 goto bind_fail; 1466 queue->rx_evtchn = queue->tx_evtchn; 1467 queue->rx_irq = queue->tx_irq = err; 1468 1469 return 0; 1470 1471 bind_fail: 1472 xenbus_free_evtchn(queue->info->xbdev, queue->tx_evtchn); 1473 queue->tx_evtchn = 0; 1474 fail: 1475 return err; 1476 } 1477 1478 static int setup_netfront_split(struct netfront_queue *queue) 1479 { 1480 int err; 1481 1482 err = xenbus_alloc_evtchn(queue->info->xbdev, &queue->tx_evtchn); 1483 if (err < 0) 1484 goto fail; 1485 err = xenbus_alloc_evtchn(queue->info->xbdev, &queue->rx_evtchn); 1486 if (err < 0) 1487 goto alloc_rx_evtchn_fail; 1488 1489 snprintf(queue->tx_irq_name, sizeof(queue->tx_irq_name), 1490 "%s-tx", queue->name); 1491 err = bind_evtchn_to_irqhandler(queue->tx_evtchn, 1492 xennet_tx_interrupt, 1493 0, queue->tx_irq_name, queue); 1494 if (err < 0) 1495 goto bind_tx_fail; 1496 queue->tx_irq = err; 1497 1498 snprintf(queue->rx_irq_name, sizeof(queue->rx_irq_name), 1499 "%s-rx", queue->name); 1500 err = bind_evtchn_to_irqhandler(queue->rx_evtchn, 1501 xennet_rx_interrupt, 1502 0, queue->rx_irq_name, queue); 1503 if (err < 0) 1504 goto bind_rx_fail; 1505 queue->rx_irq = err; 1506 1507 return 0; 1508 1509 bind_rx_fail: 1510 unbind_from_irqhandler(queue->tx_irq, queue); 1511 queue->tx_irq = 0; 1512 bind_tx_fail: 1513 xenbus_free_evtchn(queue->info->xbdev, queue->rx_evtchn); 1514 queue->rx_evtchn = 0; 1515 alloc_rx_evtchn_fail: 1516 xenbus_free_evtchn(queue->info->xbdev, queue->tx_evtchn); 1517 queue->tx_evtchn = 0; 1518 fail: 1519 return err; 1520 } 1521 1522 static int setup_netfront(struct xenbus_device *dev, 1523 struct netfront_queue *queue, unsigned int feature_split_evtchn) 1524 { 1525 struct xen_netif_tx_sring *txs; 1526 struct xen_netif_rx_sring *rxs; 1527 grant_ref_t gref; 1528 int err; 1529 1530 queue->tx_ring_ref = GRANT_INVALID_REF; 1531 queue->rx_ring_ref = GRANT_INVALID_REF; 1532 queue->rx.sring = NULL; 1533 queue->tx.sring = NULL; 1534 1535 txs = (struct xen_netif_tx_sring *)get_zeroed_page(GFP_NOIO | __GFP_HIGH); 1536 if (!txs) { 1537 err = -ENOMEM; 1538 xenbus_dev_fatal(dev, err, "allocating tx ring page"); 1539 goto fail; 1540 } 1541 SHARED_RING_INIT(txs); 1542 FRONT_RING_INIT(&queue->tx, txs, XEN_PAGE_SIZE); 1543 1544 err = xenbus_grant_ring(dev, txs, 1, &gref); 1545 if (err < 0) 1546 goto grant_tx_ring_fail; 1547 queue->tx_ring_ref = gref; 1548 1549 rxs = (struct xen_netif_rx_sring *)get_zeroed_page(GFP_NOIO | __GFP_HIGH); 1550 if (!rxs) { 1551 err = -ENOMEM; 1552 xenbus_dev_fatal(dev, err, "allocating rx ring page"); 1553 goto alloc_rx_ring_fail; 1554 } 1555 SHARED_RING_INIT(rxs); 1556 FRONT_RING_INIT(&queue->rx, rxs, XEN_PAGE_SIZE); 1557 1558 err = xenbus_grant_ring(dev, rxs, 1, &gref); 1559 if (err < 0) 1560 goto grant_rx_ring_fail; 1561 queue->rx_ring_ref = gref; 1562 1563 if (feature_split_evtchn) 1564 err = setup_netfront_split(queue); 1565 /* setup single event channel if 1566 * a) feature-split-event-channels == 0 1567 * b) feature-split-event-channels == 1 but failed to setup 1568 */ 1569 if (!feature_split_evtchn || (feature_split_evtchn && err)) 1570 err = setup_netfront_single(queue); 1571 1572 if (err) 1573 goto alloc_evtchn_fail; 1574 1575 return 0; 1576 1577 /* If we fail to setup netfront, it is safe to just revoke access to 1578 * granted pages because backend is not accessing it at this point. 1579 */ 1580 alloc_evtchn_fail: 1581 gnttab_end_foreign_access_ref(queue->rx_ring_ref, 0); 1582 grant_rx_ring_fail: 1583 free_page((unsigned long)rxs); 1584 alloc_rx_ring_fail: 1585 gnttab_end_foreign_access_ref(queue->tx_ring_ref, 0); 1586 grant_tx_ring_fail: 1587 free_page((unsigned long)txs); 1588 fail: 1589 return err; 1590 } 1591 1592 /* Queue-specific initialisation 1593 * This used to be done in xennet_create_dev() but must now 1594 * be run per-queue. 1595 */ 1596 static int xennet_init_queue(struct netfront_queue *queue) 1597 { 1598 unsigned short i; 1599 int err = 0; 1600 1601 spin_lock_init(&queue->tx_lock); 1602 spin_lock_init(&queue->rx_lock); 1603 1604 timer_setup(&queue->rx_refill_timer, rx_refill_timeout, 0); 1605 1606 snprintf(queue->name, sizeof(queue->name), "%s-q%u", 1607 queue->info->netdev->name, queue->id); 1608 1609 /* Initialise tx_skbs as a free chain containing every entry. */ 1610 queue->tx_skb_freelist = 0; 1611 for (i = 0; i < NET_TX_RING_SIZE; i++) { 1612 skb_entry_set_link(&queue->tx_skbs[i], i+1); 1613 queue->grant_tx_ref[i] = GRANT_INVALID_REF; 1614 queue->grant_tx_page[i] = NULL; 1615 } 1616 1617 /* Clear out rx_skbs */ 1618 for (i = 0; i < NET_RX_RING_SIZE; i++) { 1619 queue->rx_skbs[i] = NULL; 1620 queue->grant_rx_ref[i] = GRANT_INVALID_REF; 1621 } 1622 1623 /* A grant for every tx ring slot */ 1624 if (gnttab_alloc_grant_references(NET_TX_RING_SIZE, 1625 &queue->gref_tx_head) < 0) { 1626 pr_alert("can't alloc tx grant refs\n"); 1627 err = -ENOMEM; 1628 goto exit; 1629 } 1630 1631 /* A grant for every rx ring slot */ 1632 if (gnttab_alloc_grant_references(NET_RX_RING_SIZE, 1633 &queue->gref_rx_head) < 0) { 1634 pr_alert("can't alloc rx grant refs\n"); 1635 err = -ENOMEM; 1636 goto exit_free_tx; 1637 } 1638 1639 return 0; 1640 1641 exit_free_tx: 1642 gnttab_free_grant_references(queue->gref_tx_head); 1643 exit: 1644 return err; 1645 } 1646 1647 static int write_queue_xenstore_keys(struct netfront_queue *queue, 1648 struct xenbus_transaction *xbt, int write_hierarchical) 1649 { 1650 /* Write the queue-specific keys into XenStore in the traditional 1651 * way for a single queue, or in a queue subkeys for multiple 1652 * queues. 1653 */ 1654 struct xenbus_device *dev = queue->info->xbdev; 1655 int err; 1656 const char *message; 1657 char *path; 1658 size_t pathsize; 1659 1660 /* Choose the correct place to write the keys */ 1661 if (write_hierarchical) { 1662 pathsize = strlen(dev->nodename) + 10; 1663 path = kzalloc(pathsize, GFP_KERNEL); 1664 if (!path) { 1665 err = -ENOMEM; 1666 message = "out of memory while writing ring references"; 1667 goto error; 1668 } 1669 snprintf(path, pathsize, "%s/queue-%u", 1670 dev->nodename, queue->id); 1671 } else { 1672 path = (char *)dev->nodename; 1673 } 1674 1675 /* Write ring references */ 1676 err = xenbus_printf(*xbt, path, "tx-ring-ref", "%u", 1677 queue->tx_ring_ref); 1678 if (err) { 1679 message = "writing tx-ring-ref"; 1680 goto error; 1681 } 1682 1683 err = xenbus_printf(*xbt, path, "rx-ring-ref", "%u", 1684 queue->rx_ring_ref); 1685 if (err) { 1686 message = "writing rx-ring-ref"; 1687 goto error; 1688 } 1689 1690 /* Write event channels; taking into account both shared 1691 * and split event channel scenarios. 1692 */ 1693 if (queue->tx_evtchn == queue->rx_evtchn) { 1694 /* Shared event channel */ 1695 err = xenbus_printf(*xbt, path, 1696 "event-channel", "%u", queue->tx_evtchn); 1697 if (err) { 1698 message = "writing event-channel"; 1699 goto error; 1700 } 1701 } else { 1702 /* Split event channels */ 1703 err = xenbus_printf(*xbt, path, 1704 "event-channel-tx", "%u", queue->tx_evtchn); 1705 if (err) { 1706 message = "writing event-channel-tx"; 1707 goto error; 1708 } 1709 1710 err = xenbus_printf(*xbt, path, 1711 "event-channel-rx", "%u", queue->rx_evtchn); 1712 if (err) { 1713 message = "writing event-channel-rx"; 1714 goto error; 1715 } 1716 } 1717 1718 if (write_hierarchical) 1719 kfree(path); 1720 return 0; 1721 1722 error: 1723 if (write_hierarchical) 1724 kfree(path); 1725 xenbus_dev_fatal(dev, err, "%s", message); 1726 return err; 1727 } 1728 1729 static void xennet_destroy_queues(struct netfront_info *info) 1730 { 1731 unsigned int i; 1732 1733 for (i = 0; i < info->netdev->real_num_tx_queues; i++) { 1734 struct netfront_queue *queue = &info->queues[i]; 1735 1736 if (netif_running(info->netdev)) 1737 napi_disable(&queue->napi); 1738 netif_napi_del(&queue->napi); 1739 } 1740 1741 kfree(info->queues); 1742 info->queues = NULL; 1743 } 1744 1745 static int xennet_create_queues(struct netfront_info *info, 1746 unsigned int *num_queues) 1747 { 1748 unsigned int i; 1749 int ret; 1750 1751 info->queues = kcalloc(*num_queues, sizeof(struct netfront_queue), 1752 GFP_KERNEL); 1753 if (!info->queues) 1754 return -ENOMEM; 1755 1756 for (i = 0; i < *num_queues; i++) { 1757 struct netfront_queue *queue = &info->queues[i]; 1758 1759 queue->id = i; 1760 queue->info = info; 1761 1762 ret = xennet_init_queue(queue); 1763 if (ret < 0) { 1764 dev_warn(&info->xbdev->dev, 1765 "only created %d queues\n", i); 1766 *num_queues = i; 1767 break; 1768 } 1769 1770 netif_napi_add(queue->info->netdev, &queue->napi, 1771 xennet_poll, 64); 1772 if (netif_running(info->netdev)) 1773 napi_enable(&queue->napi); 1774 } 1775 1776 netif_set_real_num_tx_queues(info->netdev, *num_queues); 1777 1778 if (*num_queues == 0) { 1779 dev_err(&info->xbdev->dev, "no queues\n"); 1780 return -EINVAL; 1781 } 1782 return 0; 1783 } 1784 1785 /* Common code used when first setting up, and when resuming. */ 1786 static int talk_to_netback(struct xenbus_device *dev, 1787 struct netfront_info *info) 1788 { 1789 const char *message; 1790 struct xenbus_transaction xbt; 1791 int err; 1792 unsigned int feature_split_evtchn; 1793 unsigned int i = 0; 1794 unsigned int max_queues = 0; 1795 struct netfront_queue *queue = NULL; 1796 unsigned int num_queues = 1; 1797 1798 info->netdev->irq = 0; 1799 1800 /* Check if backend supports multiple queues */ 1801 max_queues = xenbus_read_unsigned(info->xbdev->otherend, 1802 "multi-queue-max-queues", 1); 1803 num_queues = min(max_queues, xennet_max_queues); 1804 1805 /* Check feature-split-event-channels */ 1806 feature_split_evtchn = xenbus_read_unsigned(info->xbdev->otherend, 1807 "feature-split-event-channels", 0); 1808 1809 /* Read mac addr. */ 1810 err = xen_net_read_mac(dev, info->netdev->dev_addr); 1811 if (err) { 1812 xenbus_dev_fatal(dev, err, "parsing %s/mac", dev->nodename); 1813 goto out_unlocked; 1814 } 1815 1816 rtnl_lock(); 1817 if (info->queues) 1818 xennet_destroy_queues(info); 1819 1820 err = xennet_create_queues(info, &num_queues); 1821 if (err < 0) { 1822 xenbus_dev_fatal(dev, err, "creating queues"); 1823 kfree(info->queues); 1824 info->queues = NULL; 1825 goto out; 1826 } 1827 rtnl_unlock(); 1828 1829 /* Create shared ring, alloc event channel -- for each queue */ 1830 for (i = 0; i < num_queues; ++i) { 1831 queue = &info->queues[i]; 1832 err = setup_netfront(dev, queue, feature_split_evtchn); 1833 if (err) 1834 goto destroy_ring; 1835 } 1836 1837 again: 1838 err = xenbus_transaction_start(&xbt); 1839 if (err) { 1840 xenbus_dev_fatal(dev, err, "starting transaction"); 1841 goto destroy_ring; 1842 } 1843 1844 if (xenbus_exists(XBT_NIL, 1845 info->xbdev->otherend, "multi-queue-max-queues")) { 1846 /* Write the number of queues */ 1847 err = xenbus_printf(xbt, dev->nodename, 1848 "multi-queue-num-queues", "%u", num_queues); 1849 if (err) { 1850 message = "writing multi-queue-num-queues"; 1851 goto abort_transaction_no_dev_fatal; 1852 } 1853 } 1854 1855 if (num_queues == 1) { 1856 err = write_queue_xenstore_keys(&info->queues[0], &xbt, 0); /* flat */ 1857 if (err) 1858 goto abort_transaction_no_dev_fatal; 1859 } else { 1860 /* Write the keys for each queue */ 1861 for (i = 0; i < num_queues; ++i) { 1862 queue = &info->queues[i]; 1863 err = write_queue_xenstore_keys(queue, &xbt, 1); /* hierarchical */ 1864 if (err) 1865 goto abort_transaction_no_dev_fatal; 1866 } 1867 } 1868 1869 /* The remaining keys are not queue-specific */ 1870 err = xenbus_printf(xbt, dev->nodename, "request-rx-copy", "%u", 1871 1); 1872 if (err) { 1873 message = "writing request-rx-copy"; 1874 goto abort_transaction; 1875 } 1876 1877 err = xenbus_printf(xbt, dev->nodename, "feature-rx-notify", "%d", 1); 1878 if (err) { 1879 message = "writing feature-rx-notify"; 1880 goto abort_transaction; 1881 } 1882 1883 err = xenbus_printf(xbt, dev->nodename, "feature-sg", "%d", 1); 1884 if (err) { 1885 message = "writing feature-sg"; 1886 goto abort_transaction; 1887 } 1888 1889 err = xenbus_printf(xbt, dev->nodename, "feature-gso-tcpv4", "%d", 1); 1890 if (err) { 1891 message = "writing feature-gso-tcpv4"; 1892 goto abort_transaction; 1893 } 1894 1895 err = xenbus_write(xbt, dev->nodename, "feature-gso-tcpv6", "1"); 1896 if (err) { 1897 message = "writing feature-gso-tcpv6"; 1898 goto abort_transaction; 1899 } 1900 1901 err = xenbus_write(xbt, dev->nodename, "feature-ipv6-csum-offload", 1902 "1"); 1903 if (err) { 1904 message = "writing feature-ipv6-csum-offload"; 1905 goto abort_transaction; 1906 } 1907 1908 err = xenbus_transaction_end(xbt, 0); 1909 if (err) { 1910 if (err == -EAGAIN) 1911 goto again; 1912 xenbus_dev_fatal(dev, err, "completing transaction"); 1913 goto destroy_ring; 1914 } 1915 1916 return 0; 1917 1918 abort_transaction: 1919 xenbus_dev_fatal(dev, err, "%s", message); 1920 abort_transaction_no_dev_fatal: 1921 xenbus_transaction_end(xbt, 1); 1922 destroy_ring: 1923 xennet_disconnect_backend(info); 1924 rtnl_lock(); 1925 xennet_destroy_queues(info); 1926 out: 1927 rtnl_unlock(); 1928 out_unlocked: 1929 device_unregister(&dev->dev); 1930 return err; 1931 } 1932 1933 static int xennet_connect(struct net_device *dev) 1934 { 1935 struct netfront_info *np = netdev_priv(dev); 1936 unsigned int num_queues = 0; 1937 int err; 1938 unsigned int j = 0; 1939 struct netfront_queue *queue = NULL; 1940 1941 if (!xenbus_read_unsigned(np->xbdev->otherend, "feature-rx-copy", 0)) { 1942 dev_info(&dev->dev, 1943 "backend does not support copying receive path\n"); 1944 return -ENODEV; 1945 } 1946 1947 err = talk_to_netback(np->xbdev, np); 1948 if (err) 1949 return err; 1950 1951 /* talk_to_netback() sets the correct number of queues */ 1952 num_queues = dev->real_num_tx_queues; 1953 1954 if (dev->reg_state == NETREG_UNINITIALIZED) { 1955 err = register_netdev(dev); 1956 if (err) { 1957 pr_warn("%s: register_netdev err=%d\n", __func__, err); 1958 device_unregister(&np->xbdev->dev); 1959 return err; 1960 } 1961 } 1962 1963 rtnl_lock(); 1964 netdev_update_features(dev); 1965 rtnl_unlock(); 1966 1967 /* 1968 * All public and private state should now be sane. Get 1969 * ready to start sending and receiving packets and give the driver 1970 * domain a kick because we've probably just requeued some 1971 * packets. 1972 */ 1973 netif_carrier_on(np->netdev); 1974 for (j = 0; j < num_queues; ++j) { 1975 queue = &np->queues[j]; 1976 1977 notify_remote_via_irq(queue->tx_irq); 1978 if (queue->tx_irq != queue->rx_irq) 1979 notify_remote_via_irq(queue->rx_irq); 1980 1981 spin_lock_irq(&queue->tx_lock); 1982 xennet_tx_buf_gc(queue); 1983 spin_unlock_irq(&queue->tx_lock); 1984 1985 spin_lock_bh(&queue->rx_lock); 1986 xennet_alloc_rx_buffers(queue); 1987 spin_unlock_bh(&queue->rx_lock); 1988 } 1989 1990 return 0; 1991 } 1992 1993 /** 1994 * Callback received when the backend's state changes. 1995 */ 1996 static void netback_changed(struct xenbus_device *dev, 1997 enum xenbus_state backend_state) 1998 { 1999 struct netfront_info *np = dev_get_drvdata(&dev->dev); 2000 struct net_device *netdev = np->netdev; 2001 2002 dev_dbg(&dev->dev, "%s\n", xenbus_strstate(backend_state)); 2003 2004 switch (backend_state) { 2005 case XenbusStateInitialising: 2006 case XenbusStateInitialised: 2007 case XenbusStateReconfiguring: 2008 case XenbusStateReconfigured: 2009 break; 2010 2011 case XenbusStateUnknown: 2012 wake_up_all(&module_unload_q); 2013 break; 2014 2015 case XenbusStateInitWait: 2016 if (dev->state != XenbusStateInitialising) 2017 break; 2018 if (xennet_connect(netdev) != 0) 2019 break; 2020 xenbus_switch_state(dev, XenbusStateConnected); 2021 break; 2022 2023 case XenbusStateConnected: 2024 netdev_notify_peers(netdev); 2025 break; 2026 2027 case XenbusStateClosed: 2028 wake_up_all(&module_unload_q); 2029 if (dev->state == XenbusStateClosed) 2030 break; 2031 /* Missed the backend's CLOSING state -- fallthrough */ 2032 case XenbusStateClosing: 2033 wake_up_all(&module_unload_q); 2034 xenbus_frontend_closed(dev); 2035 break; 2036 } 2037 } 2038 2039 static const struct xennet_stat { 2040 char name[ETH_GSTRING_LEN]; 2041 u16 offset; 2042 } xennet_stats[] = { 2043 { 2044 "rx_gso_checksum_fixup", 2045 offsetof(struct netfront_info, rx_gso_checksum_fixup) 2046 }, 2047 }; 2048 2049 static int xennet_get_sset_count(struct net_device *dev, int string_set) 2050 { 2051 switch (string_set) { 2052 case ETH_SS_STATS: 2053 return ARRAY_SIZE(xennet_stats); 2054 default: 2055 return -EINVAL; 2056 } 2057 } 2058 2059 static void xennet_get_ethtool_stats(struct net_device *dev, 2060 struct ethtool_stats *stats, u64 * data) 2061 { 2062 void *np = netdev_priv(dev); 2063 int i; 2064 2065 for (i = 0; i < ARRAY_SIZE(xennet_stats); i++) 2066 data[i] = atomic_read((atomic_t *)(np + xennet_stats[i].offset)); 2067 } 2068 2069 static void xennet_get_strings(struct net_device *dev, u32 stringset, u8 * data) 2070 { 2071 int i; 2072 2073 switch (stringset) { 2074 case ETH_SS_STATS: 2075 for (i = 0; i < ARRAY_SIZE(xennet_stats); i++) 2076 memcpy(data + i * ETH_GSTRING_LEN, 2077 xennet_stats[i].name, ETH_GSTRING_LEN); 2078 break; 2079 } 2080 } 2081 2082 static const struct ethtool_ops xennet_ethtool_ops = 2083 { 2084 .get_link = ethtool_op_get_link, 2085 2086 .get_sset_count = xennet_get_sset_count, 2087 .get_ethtool_stats = xennet_get_ethtool_stats, 2088 .get_strings = xennet_get_strings, 2089 }; 2090 2091 #ifdef CONFIG_SYSFS 2092 static ssize_t show_rxbuf(struct device *dev, 2093 struct device_attribute *attr, char *buf) 2094 { 2095 return sprintf(buf, "%lu\n", NET_RX_RING_SIZE); 2096 } 2097 2098 static ssize_t store_rxbuf(struct device *dev, 2099 struct device_attribute *attr, 2100 const char *buf, size_t len) 2101 { 2102 char *endp; 2103 unsigned long target; 2104 2105 if (!capable(CAP_NET_ADMIN)) 2106 return -EPERM; 2107 2108 target = simple_strtoul(buf, &endp, 0); 2109 if (endp == buf) 2110 return -EBADMSG; 2111 2112 /* rxbuf_min and rxbuf_max are no longer configurable. */ 2113 2114 return len; 2115 } 2116 2117 static DEVICE_ATTR(rxbuf_min, 0644, show_rxbuf, store_rxbuf); 2118 static DEVICE_ATTR(rxbuf_max, 0644, show_rxbuf, store_rxbuf); 2119 static DEVICE_ATTR(rxbuf_cur, 0444, show_rxbuf, NULL); 2120 2121 static struct attribute *xennet_dev_attrs[] = { 2122 &dev_attr_rxbuf_min.attr, 2123 &dev_attr_rxbuf_max.attr, 2124 &dev_attr_rxbuf_cur.attr, 2125 NULL 2126 }; 2127 2128 static const struct attribute_group xennet_dev_group = { 2129 .attrs = xennet_dev_attrs 2130 }; 2131 #endif /* CONFIG_SYSFS */ 2132 2133 static int xennet_remove(struct xenbus_device *dev) 2134 { 2135 struct netfront_info *info = dev_get_drvdata(&dev->dev); 2136 2137 dev_dbg(&dev->dev, "%s\n", dev->nodename); 2138 2139 if (xenbus_read_driver_state(dev->otherend) != XenbusStateClosed) { 2140 xenbus_switch_state(dev, XenbusStateClosing); 2141 wait_event(module_unload_q, 2142 xenbus_read_driver_state(dev->otherend) == 2143 XenbusStateClosing || 2144 xenbus_read_driver_state(dev->otherend) == 2145 XenbusStateUnknown); 2146 2147 xenbus_switch_state(dev, XenbusStateClosed); 2148 wait_event(module_unload_q, 2149 xenbus_read_driver_state(dev->otherend) == 2150 XenbusStateClosed || 2151 xenbus_read_driver_state(dev->otherend) == 2152 XenbusStateUnknown); 2153 } 2154 2155 xennet_disconnect_backend(info); 2156 2157 if (info->netdev->reg_state == NETREG_REGISTERED) 2158 unregister_netdev(info->netdev); 2159 2160 if (info->queues) { 2161 rtnl_lock(); 2162 xennet_destroy_queues(info); 2163 rtnl_unlock(); 2164 } 2165 xennet_free_netdev(info->netdev); 2166 2167 return 0; 2168 } 2169 2170 static const struct xenbus_device_id netfront_ids[] = { 2171 { "vif" }, 2172 { "" } 2173 }; 2174 2175 static struct xenbus_driver netfront_driver = { 2176 .ids = netfront_ids, 2177 .probe = netfront_probe, 2178 .remove = xennet_remove, 2179 .resume = netfront_resume, 2180 .otherend_changed = netback_changed, 2181 }; 2182 2183 static int __init netif_init(void) 2184 { 2185 if (!xen_domain()) 2186 return -ENODEV; 2187 2188 if (!xen_has_pv_nic_devices()) 2189 return -ENODEV; 2190 2191 pr_info("Initialising Xen virtual ethernet driver\n"); 2192 2193 /* Allow as many queues as there are CPUs inut max. 8 if user has not 2194 * specified a value. 2195 */ 2196 if (xennet_max_queues == 0) 2197 xennet_max_queues = min_t(unsigned int, MAX_QUEUES_DEFAULT, 2198 num_online_cpus()); 2199 2200 return xenbus_register_frontend(&netfront_driver); 2201 } 2202 module_init(netif_init); 2203 2204 2205 static void __exit netif_exit(void) 2206 { 2207 xenbus_unregister_driver(&netfront_driver); 2208 } 2209 module_exit(netif_exit); 2210 2211 MODULE_DESCRIPTION("Xen virtual network device frontend"); 2212 MODULE_LICENSE("GPL"); 2213 MODULE_ALIAS("xen:vif"); 2214 MODULE_ALIAS("xennet"); 2215