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 <asm/xen/page.h> 49 #include <xen/xen.h> 50 #include <xen/xenbus.h> 51 #include <xen/events.h> 52 #include <xen/page.h> 53 #include <xen/platform_pci.h> 54 #include <xen/grant_table.h> 55 56 #include <xen/interface/io/netif.h> 57 #include <xen/interface/memory.h> 58 #include <xen/interface/grant_table.h> 59 60 /* Module parameters */ 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, PAGE_SIZE) 79 #define NET_RX_RING_SIZE __CONST_RING_SIZE(xen_netif_rx, 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 struct netfront_stats { 91 u64 packets; 92 u64 bytes; 93 struct u64_stats_sync syncp; 94 }; 95 96 struct netfront_info; 97 98 struct netfront_queue { 99 unsigned int id; /* Queue ID, 0-based */ 100 char name[QUEUE_NAME_SIZE]; /* DEVNAME-qN */ 101 struct netfront_info *info; 102 103 struct napi_struct napi; 104 105 /* Split event channels support, tx_* == rx_* when using 106 * single event channel. 107 */ 108 unsigned int tx_evtchn, rx_evtchn; 109 unsigned int tx_irq, rx_irq; 110 /* Only used when split event channels support is enabled */ 111 char tx_irq_name[IRQ_NAME_SIZE]; /* DEVNAME-qN-tx */ 112 char rx_irq_name[IRQ_NAME_SIZE]; /* DEVNAME-qN-rx */ 113 114 spinlock_t tx_lock; 115 struct xen_netif_tx_front_ring tx; 116 int tx_ring_ref; 117 118 /* 119 * {tx,rx}_skbs store outstanding skbuffs. Free tx_skb entries 120 * are linked from tx_skb_freelist through skb_entry.link. 121 * 122 * NB. Freelist index entries are always going to be less than 123 * PAGE_OFFSET, whereas pointers to skbs will always be equal or 124 * greater than PAGE_OFFSET: we use this property to distinguish 125 * them. 126 */ 127 union skb_entry { 128 struct sk_buff *skb; 129 unsigned long link; 130 } tx_skbs[NET_TX_RING_SIZE]; 131 grant_ref_t gref_tx_head; 132 grant_ref_t grant_tx_ref[NET_TX_RING_SIZE]; 133 struct page *grant_tx_page[NET_TX_RING_SIZE]; 134 unsigned tx_skb_freelist; 135 136 spinlock_t rx_lock ____cacheline_aligned_in_smp; 137 struct xen_netif_rx_front_ring rx; 138 int rx_ring_ref; 139 140 struct timer_list rx_refill_timer; 141 142 struct sk_buff *rx_skbs[NET_RX_RING_SIZE]; 143 grant_ref_t gref_rx_head; 144 grant_ref_t grant_rx_ref[NET_RX_RING_SIZE]; 145 146 unsigned long rx_pfn_array[NET_RX_RING_SIZE]; 147 struct multicall_entry rx_mcl[NET_RX_RING_SIZE+1]; 148 struct mmu_update rx_mmu[NET_RX_RING_SIZE]; 149 }; 150 151 struct netfront_info { 152 struct list_head list; 153 struct net_device *netdev; 154 155 struct xenbus_device *xbdev; 156 157 /* Multi-queue support */ 158 struct netfront_queue *queues; 159 160 /* Statistics */ 161 struct netfront_stats __percpu *rx_stats; 162 struct netfront_stats __percpu *tx_stats; 163 164 atomic_t rx_gso_checksum_fixup; 165 }; 166 167 struct netfront_rx_info { 168 struct xen_netif_rx_response rx; 169 struct xen_netif_extra_info extras[XEN_NETIF_EXTRA_TYPE_MAX - 1]; 170 }; 171 172 static void skb_entry_set_link(union skb_entry *list, unsigned short id) 173 { 174 list->link = id; 175 } 176 177 static int skb_entry_is_link(const union skb_entry *list) 178 { 179 BUILD_BUG_ON(sizeof(list->skb) != sizeof(list->link)); 180 return (unsigned long)list->skb < PAGE_OFFSET; 181 } 182 183 /* 184 * Access macros for acquiring freeing slots in tx_skbs[]. 185 */ 186 187 static void add_id_to_freelist(unsigned *head, union skb_entry *list, 188 unsigned short id) 189 { 190 skb_entry_set_link(&list[id], *head); 191 *head = id; 192 } 193 194 static unsigned short get_id_from_freelist(unsigned *head, 195 union skb_entry *list) 196 { 197 unsigned int id = *head; 198 *head = list[id].link; 199 return id; 200 } 201 202 static int xennet_rxidx(RING_IDX idx) 203 { 204 return idx & (NET_RX_RING_SIZE - 1); 205 } 206 207 static struct sk_buff *xennet_get_rx_skb(struct netfront_queue *queue, 208 RING_IDX ri) 209 { 210 int i = xennet_rxidx(ri); 211 struct sk_buff *skb = queue->rx_skbs[i]; 212 queue->rx_skbs[i] = NULL; 213 return skb; 214 } 215 216 static grant_ref_t xennet_get_rx_ref(struct netfront_queue *queue, 217 RING_IDX ri) 218 { 219 int i = xennet_rxidx(ri); 220 grant_ref_t ref = queue->grant_rx_ref[i]; 221 queue->grant_rx_ref[i] = GRANT_INVALID_REF; 222 return ref; 223 } 224 225 #ifdef CONFIG_SYSFS 226 static int xennet_sysfs_addif(struct net_device *netdev); 227 static void xennet_sysfs_delif(struct net_device *netdev); 228 #else /* !CONFIG_SYSFS */ 229 #define xennet_sysfs_addif(dev) (0) 230 #define xennet_sysfs_delif(dev) do { } while (0) 231 #endif 232 233 static bool xennet_can_sg(struct net_device *dev) 234 { 235 return dev->features & NETIF_F_SG; 236 } 237 238 239 static void rx_refill_timeout(unsigned long data) 240 { 241 struct netfront_queue *queue = (struct netfront_queue *)data; 242 napi_schedule(&queue->napi); 243 } 244 245 static int netfront_tx_slot_available(struct netfront_queue *queue) 246 { 247 return (queue->tx.req_prod_pvt - queue->tx.rsp_cons) < 248 (NET_TX_RING_SIZE - MAX_SKB_FRAGS - 2); 249 } 250 251 static void xennet_maybe_wake_tx(struct netfront_queue *queue) 252 { 253 struct net_device *dev = queue->info->netdev; 254 struct netdev_queue *dev_queue = netdev_get_tx_queue(dev, queue->id); 255 256 if (unlikely(netif_tx_queue_stopped(dev_queue)) && 257 netfront_tx_slot_available(queue) && 258 likely(netif_running(dev))) 259 netif_tx_wake_queue(netdev_get_tx_queue(dev, queue->id)); 260 } 261 262 263 static struct sk_buff *xennet_alloc_one_rx_buffer(struct netfront_queue *queue) 264 { 265 struct sk_buff *skb; 266 struct page *page; 267 268 skb = __netdev_alloc_skb(queue->info->netdev, 269 RX_COPY_THRESHOLD + NET_IP_ALIGN, 270 GFP_ATOMIC | __GFP_NOWARN); 271 if (unlikely(!skb)) 272 return NULL; 273 274 page = alloc_page(GFP_ATOMIC | __GFP_NOWARN); 275 if (!page) { 276 kfree_skb(skb); 277 return NULL; 278 } 279 skb_add_rx_frag(skb, 0, page, 0, 0, PAGE_SIZE); 280 281 /* Align ip header to a 16 bytes boundary */ 282 skb_reserve(skb, NET_IP_ALIGN); 283 skb->dev = queue->info->netdev; 284 285 return skb; 286 } 287 288 289 static void xennet_alloc_rx_buffers(struct netfront_queue *queue) 290 { 291 RING_IDX req_prod = queue->rx.req_prod_pvt; 292 int notify; 293 294 if (unlikely(!netif_carrier_ok(queue->info->netdev))) 295 return; 296 297 for (req_prod = queue->rx.req_prod_pvt; 298 req_prod - queue->rx.rsp_cons < NET_RX_RING_SIZE; 299 req_prod++) { 300 struct sk_buff *skb; 301 unsigned short id; 302 grant_ref_t ref; 303 unsigned long pfn; 304 struct xen_netif_rx_request *req; 305 306 skb = xennet_alloc_one_rx_buffer(queue); 307 if (!skb) 308 break; 309 310 id = xennet_rxidx(req_prod); 311 312 BUG_ON(queue->rx_skbs[id]); 313 queue->rx_skbs[id] = skb; 314 315 ref = gnttab_claim_grant_reference(&queue->gref_rx_head); 316 BUG_ON((signed short)ref < 0); 317 queue->grant_rx_ref[id] = ref; 318 319 pfn = page_to_pfn(skb_frag_page(&skb_shinfo(skb)->frags[0])); 320 321 req = RING_GET_REQUEST(&queue->rx, req_prod); 322 gnttab_grant_foreign_access_ref(ref, 323 queue->info->xbdev->otherend_id, 324 pfn_to_mfn(pfn), 325 0); 326 327 req->id = id; 328 req->gref = ref; 329 } 330 331 queue->rx.req_prod_pvt = req_prod; 332 333 /* Not enough requests? Try again later. */ 334 if (req_prod - queue->rx.rsp_cons < NET_RX_SLOTS_MIN) { 335 mod_timer(&queue->rx_refill_timer, jiffies + (HZ/10)); 336 return; 337 } 338 339 wmb(); /* barrier so backend seens requests */ 340 341 RING_PUSH_REQUESTS_AND_CHECK_NOTIFY(&queue->rx, notify); 342 if (notify) 343 notify_remote_via_irq(queue->rx_irq); 344 } 345 346 static int xennet_open(struct net_device *dev) 347 { 348 struct netfront_info *np = netdev_priv(dev); 349 unsigned int num_queues = dev->real_num_tx_queues; 350 unsigned int i = 0; 351 struct netfront_queue *queue = NULL; 352 353 for (i = 0; i < num_queues; ++i) { 354 queue = &np->queues[i]; 355 napi_enable(&queue->napi); 356 357 spin_lock_bh(&queue->rx_lock); 358 if (netif_carrier_ok(dev)) { 359 xennet_alloc_rx_buffers(queue); 360 queue->rx.sring->rsp_event = queue->rx.rsp_cons + 1; 361 if (RING_HAS_UNCONSUMED_RESPONSES(&queue->rx)) 362 napi_schedule(&queue->napi); 363 } 364 spin_unlock_bh(&queue->rx_lock); 365 } 366 367 netif_tx_start_all_queues(dev); 368 369 return 0; 370 } 371 372 static void xennet_tx_buf_gc(struct netfront_queue *queue) 373 { 374 RING_IDX cons, prod; 375 unsigned short id; 376 struct sk_buff *skb; 377 378 BUG_ON(!netif_carrier_ok(queue->info->netdev)); 379 380 do { 381 prod = queue->tx.sring->rsp_prod; 382 rmb(); /* Ensure we see responses up to 'rp'. */ 383 384 for (cons = queue->tx.rsp_cons; cons != prod; cons++) { 385 struct xen_netif_tx_response *txrsp; 386 387 txrsp = RING_GET_RESPONSE(&queue->tx, cons); 388 if (txrsp->status == XEN_NETIF_RSP_NULL) 389 continue; 390 391 id = txrsp->id; 392 skb = queue->tx_skbs[id].skb; 393 if (unlikely(gnttab_query_foreign_access( 394 queue->grant_tx_ref[id]) != 0)) { 395 pr_alert("%s: warning -- grant still in use by backend domain\n", 396 __func__); 397 BUG(); 398 } 399 gnttab_end_foreign_access_ref( 400 queue->grant_tx_ref[id], GNTMAP_readonly); 401 gnttab_release_grant_reference( 402 &queue->gref_tx_head, queue->grant_tx_ref[id]); 403 queue->grant_tx_ref[id] = GRANT_INVALID_REF; 404 queue->grant_tx_page[id] = NULL; 405 add_id_to_freelist(&queue->tx_skb_freelist, queue->tx_skbs, id); 406 dev_kfree_skb_irq(skb); 407 } 408 409 queue->tx.rsp_cons = prod; 410 411 /* 412 * Set a new event, then check for race with update of tx_cons. 413 * Note that it is essential to schedule a callback, no matter 414 * how few buffers are pending. Even if there is space in the 415 * transmit ring, higher layers may be blocked because too much 416 * data is outstanding: in such cases notification from Xen is 417 * likely to be the only kick that we'll get. 418 */ 419 queue->tx.sring->rsp_event = 420 prod + ((queue->tx.sring->req_prod - prod) >> 1) + 1; 421 mb(); /* update shared area */ 422 } while ((cons == prod) && (prod != queue->tx.sring->rsp_prod)); 423 424 xennet_maybe_wake_tx(queue); 425 } 426 427 static void xennet_make_frags(struct sk_buff *skb, struct netfront_queue *queue, 428 struct xen_netif_tx_request *tx) 429 { 430 char *data = skb->data; 431 unsigned long mfn; 432 RING_IDX prod = queue->tx.req_prod_pvt; 433 int frags = skb_shinfo(skb)->nr_frags; 434 unsigned int offset = offset_in_page(data); 435 unsigned int len = skb_headlen(skb); 436 unsigned int id; 437 grant_ref_t ref; 438 int i; 439 440 /* While the header overlaps a page boundary (including being 441 larger than a page), split it it into page-sized chunks. */ 442 while (len > PAGE_SIZE - offset) { 443 tx->size = PAGE_SIZE - offset; 444 tx->flags |= XEN_NETTXF_more_data; 445 len -= tx->size; 446 data += tx->size; 447 offset = 0; 448 449 id = get_id_from_freelist(&queue->tx_skb_freelist, queue->tx_skbs); 450 queue->tx_skbs[id].skb = skb_get(skb); 451 tx = RING_GET_REQUEST(&queue->tx, prod++); 452 tx->id = id; 453 ref = gnttab_claim_grant_reference(&queue->gref_tx_head); 454 BUG_ON((signed short)ref < 0); 455 456 mfn = virt_to_mfn(data); 457 gnttab_grant_foreign_access_ref(ref, queue->info->xbdev->otherend_id, 458 mfn, GNTMAP_readonly); 459 460 queue->grant_tx_page[id] = virt_to_page(data); 461 tx->gref = queue->grant_tx_ref[id] = ref; 462 tx->offset = offset; 463 tx->size = len; 464 tx->flags = 0; 465 } 466 467 /* Grant backend access to each skb fragment page. */ 468 for (i = 0; i < frags; i++) { 469 skb_frag_t *frag = skb_shinfo(skb)->frags + i; 470 struct page *page = skb_frag_page(frag); 471 472 len = skb_frag_size(frag); 473 offset = frag->page_offset; 474 475 /* Skip unused frames from start of page */ 476 page += offset >> PAGE_SHIFT; 477 offset &= ~PAGE_MASK; 478 479 while (len > 0) { 480 unsigned long bytes; 481 482 bytes = PAGE_SIZE - offset; 483 if (bytes > len) 484 bytes = len; 485 486 tx->flags |= XEN_NETTXF_more_data; 487 488 id = get_id_from_freelist(&queue->tx_skb_freelist, 489 queue->tx_skbs); 490 queue->tx_skbs[id].skb = skb_get(skb); 491 tx = RING_GET_REQUEST(&queue->tx, prod++); 492 tx->id = id; 493 ref = gnttab_claim_grant_reference(&queue->gref_tx_head); 494 BUG_ON((signed short)ref < 0); 495 496 mfn = pfn_to_mfn(page_to_pfn(page)); 497 gnttab_grant_foreign_access_ref(ref, 498 queue->info->xbdev->otherend_id, 499 mfn, GNTMAP_readonly); 500 501 queue->grant_tx_page[id] = page; 502 tx->gref = queue->grant_tx_ref[id] = ref; 503 tx->offset = offset; 504 tx->size = bytes; 505 tx->flags = 0; 506 507 offset += bytes; 508 len -= bytes; 509 510 /* Next frame */ 511 if (offset == PAGE_SIZE && len) { 512 BUG_ON(!PageCompound(page)); 513 page++; 514 offset = 0; 515 } 516 } 517 } 518 519 queue->tx.req_prod_pvt = prod; 520 } 521 522 /* 523 * Count how many ring slots are required to send the frags of this 524 * skb. Each frag might be a compound page. 525 */ 526 static int xennet_count_skb_frag_slots(struct sk_buff *skb) 527 { 528 int i, frags = skb_shinfo(skb)->nr_frags; 529 int pages = 0; 530 531 for (i = 0; i < frags; i++) { 532 skb_frag_t *frag = skb_shinfo(skb)->frags + i; 533 unsigned long size = skb_frag_size(frag); 534 unsigned long offset = frag->page_offset; 535 536 /* Skip unused frames from start of page */ 537 offset &= ~PAGE_MASK; 538 539 pages += PFN_UP(offset + size); 540 } 541 542 return pages; 543 } 544 545 static u16 xennet_select_queue(struct net_device *dev, struct sk_buff *skb, 546 void *accel_priv, select_queue_fallback_t fallback) 547 { 548 unsigned int num_queues = dev->real_num_tx_queues; 549 u32 hash; 550 u16 queue_idx; 551 552 /* First, check if there is only one queue */ 553 if (num_queues == 1) { 554 queue_idx = 0; 555 } else { 556 hash = skb_get_hash(skb); 557 queue_idx = hash % num_queues; 558 } 559 560 return queue_idx; 561 } 562 563 static int xennet_start_xmit(struct sk_buff *skb, struct net_device *dev) 564 { 565 unsigned short id; 566 struct netfront_info *np = netdev_priv(dev); 567 struct netfront_stats *tx_stats = this_cpu_ptr(np->tx_stats); 568 struct xen_netif_tx_request *tx; 569 char *data = skb->data; 570 RING_IDX i; 571 grant_ref_t ref; 572 unsigned long mfn; 573 int notify; 574 int slots; 575 unsigned int offset = offset_in_page(data); 576 unsigned int len = skb_headlen(skb); 577 unsigned long flags; 578 struct netfront_queue *queue = NULL; 579 unsigned int num_queues = dev->real_num_tx_queues; 580 u16 queue_index; 581 582 /* Drop the packet if no queues are set up */ 583 if (num_queues < 1) 584 goto drop; 585 /* Determine which queue to transmit this SKB on */ 586 queue_index = skb_get_queue_mapping(skb); 587 queue = &np->queues[queue_index]; 588 589 /* If skb->len is too big for wire format, drop skb and alert 590 * user about misconfiguration. 591 */ 592 if (unlikely(skb->len > XEN_NETIF_MAX_TX_SIZE)) { 593 net_alert_ratelimited( 594 "xennet: skb->len = %u, too big for wire format\n", 595 skb->len); 596 goto drop; 597 } 598 599 slots = DIV_ROUND_UP(offset + len, PAGE_SIZE) + 600 xennet_count_skb_frag_slots(skb); 601 if (unlikely(slots > MAX_SKB_FRAGS + 1)) { 602 net_dbg_ratelimited("xennet: skb rides the rocket: %d slots, %d bytes\n", 603 slots, skb->len); 604 if (skb_linearize(skb)) 605 goto drop; 606 data = skb->data; 607 offset = offset_in_page(data); 608 len = skb_headlen(skb); 609 } 610 611 spin_lock_irqsave(&queue->tx_lock, flags); 612 613 if (unlikely(!netif_carrier_ok(dev) || 614 (slots > 1 && !xennet_can_sg(dev)) || 615 netif_needs_gso(dev, skb, netif_skb_features(skb)))) { 616 spin_unlock_irqrestore(&queue->tx_lock, flags); 617 goto drop; 618 } 619 620 i = queue->tx.req_prod_pvt; 621 622 id = get_id_from_freelist(&queue->tx_skb_freelist, queue->tx_skbs); 623 queue->tx_skbs[id].skb = skb; 624 625 tx = RING_GET_REQUEST(&queue->tx, i); 626 627 tx->id = id; 628 ref = gnttab_claim_grant_reference(&queue->gref_tx_head); 629 BUG_ON((signed short)ref < 0); 630 mfn = virt_to_mfn(data); 631 gnttab_grant_foreign_access_ref( 632 ref, queue->info->xbdev->otherend_id, mfn, GNTMAP_readonly); 633 queue->grant_tx_page[id] = virt_to_page(data); 634 tx->gref = queue->grant_tx_ref[id] = ref; 635 tx->offset = offset; 636 tx->size = len; 637 638 tx->flags = 0; 639 if (skb->ip_summed == CHECKSUM_PARTIAL) 640 /* local packet? */ 641 tx->flags |= XEN_NETTXF_csum_blank | XEN_NETTXF_data_validated; 642 else if (skb->ip_summed == CHECKSUM_UNNECESSARY) 643 /* remote but checksummed. */ 644 tx->flags |= XEN_NETTXF_data_validated; 645 646 if (skb_shinfo(skb)->gso_size) { 647 struct xen_netif_extra_info *gso; 648 649 gso = (struct xen_netif_extra_info *) 650 RING_GET_REQUEST(&queue->tx, ++i); 651 652 tx->flags |= XEN_NETTXF_extra_info; 653 654 gso->u.gso.size = skb_shinfo(skb)->gso_size; 655 gso->u.gso.type = (skb_shinfo(skb)->gso_type & SKB_GSO_TCPV6) ? 656 XEN_NETIF_GSO_TYPE_TCPV6 : 657 XEN_NETIF_GSO_TYPE_TCPV4; 658 gso->u.gso.pad = 0; 659 gso->u.gso.features = 0; 660 661 gso->type = XEN_NETIF_EXTRA_TYPE_GSO; 662 gso->flags = 0; 663 } 664 665 queue->tx.req_prod_pvt = i + 1; 666 667 xennet_make_frags(skb, queue, tx); 668 tx->size = skb->len; 669 670 RING_PUSH_REQUESTS_AND_CHECK_NOTIFY(&queue->tx, notify); 671 if (notify) 672 notify_remote_via_irq(queue->tx_irq); 673 674 u64_stats_update_begin(&tx_stats->syncp); 675 tx_stats->bytes += skb->len; 676 tx_stats->packets++; 677 u64_stats_update_end(&tx_stats->syncp); 678 679 /* Note: It is not safe to access skb after xennet_tx_buf_gc()! */ 680 xennet_tx_buf_gc(queue); 681 682 if (!netfront_tx_slot_available(queue)) 683 netif_tx_stop_queue(netdev_get_tx_queue(dev, queue->id)); 684 685 spin_unlock_irqrestore(&queue->tx_lock, flags); 686 687 return NETDEV_TX_OK; 688 689 drop: 690 dev->stats.tx_dropped++; 691 dev_kfree_skb_any(skb); 692 return NETDEV_TX_OK; 693 } 694 695 static int xennet_close(struct net_device *dev) 696 { 697 struct netfront_info *np = netdev_priv(dev); 698 unsigned int num_queues = dev->real_num_tx_queues; 699 unsigned int i; 700 struct netfront_queue *queue; 701 netif_tx_stop_all_queues(np->netdev); 702 for (i = 0; i < num_queues; ++i) { 703 queue = &np->queues[i]; 704 napi_disable(&queue->napi); 705 } 706 return 0; 707 } 708 709 static void xennet_move_rx_slot(struct netfront_queue *queue, struct sk_buff *skb, 710 grant_ref_t ref) 711 { 712 int new = xennet_rxidx(queue->rx.req_prod_pvt); 713 714 BUG_ON(queue->rx_skbs[new]); 715 queue->rx_skbs[new] = skb; 716 queue->grant_rx_ref[new] = ref; 717 RING_GET_REQUEST(&queue->rx, queue->rx.req_prod_pvt)->id = new; 718 RING_GET_REQUEST(&queue->rx, queue->rx.req_prod_pvt)->gref = ref; 719 queue->rx.req_prod_pvt++; 720 } 721 722 static int xennet_get_extras(struct netfront_queue *queue, 723 struct xen_netif_extra_info *extras, 724 RING_IDX rp) 725 726 { 727 struct xen_netif_extra_info *extra; 728 struct device *dev = &queue->info->netdev->dev; 729 RING_IDX cons = queue->rx.rsp_cons; 730 int err = 0; 731 732 do { 733 struct sk_buff *skb; 734 grant_ref_t ref; 735 736 if (unlikely(cons + 1 == rp)) { 737 if (net_ratelimit()) 738 dev_warn(dev, "Missing extra info\n"); 739 err = -EBADR; 740 break; 741 } 742 743 extra = (struct xen_netif_extra_info *) 744 RING_GET_RESPONSE(&queue->rx, ++cons); 745 746 if (unlikely(!extra->type || 747 extra->type >= XEN_NETIF_EXTRA_TYPE_MAX)) { 748 if (net_ratelimit()) 749 dev_warn(dev, "Invalid extra type: %d\n", 750 extra->type); 751 err = -EINVAL; 752 } else { 753 memcpy(&extras[extra->type - 1], extra, 754 sizeof(*extra)); 755 } 756 757 skb = xennet_get_rx_skb(queue, cons); 758 ref = xennet_get_rx_ref(queue, cons); 759 xennet_move_rx_slot(queue, skb, ref); 760 } while (extra->flags & XEN_NETIF_EXTRA_FLAG_MORE); 761 762 queue->rx.rsp_cons = cons; 763 return err; 764 } 765 766 static int xennet_get_responses(struct netfront_queue *queue, 767 struct netfront_rx_info *rinfo, RING_IDX rp, 768 struct sk_buff_head *list) 769 { 770 struct xen_netif_rx_response *rx = &rinfo->rx; 771 struct xen_netif_extra_info *extras = rinfo->extras; 772 struct device *dev = &queue->info->netdev->dev; 773 RING_IDX cons = queue->rx.rsp_cons; 774 struct sk_buff *skb = xennet_get_rx_skb(queue, cons); 775 grant_ref_t ref = xennet_get_rx_ref(queue, cons); 776 int max = MAX_SKB_FRAGS + (rx->status <= RX_COPY_THRESHOLD); 777 int slots = 1; 778 int err = 0; 779 unsigned long ret; 780 781 if (rx->flags & XEN_NETRXF_extra_info) { 782 err = xennet_get_extras(queue, extras, rp); 783 cons = queue->rx.rsp_cons; 784 } 785 786 for (;;) { 787 if (unlikely(rx->status < 0 || 788 rx->offset + rx->status > PAGE_SIZE)) { 789 if (net_ratelimit()) 790 dev_warn(dev, "rx->offset: %x, size: %u\n", 791 rx->offset, rx->status); 792 xennet_move_rx_slot(queue, skb, ref); 793 err = -EINVAL; 794 goto next; 795 } 796 797 /* 798 * This definitely indicates a bug, either in this driver or in 799 * the backend driver. In future this should flag the bad 800 * situation to the system controller to reboot the backend. 801 */ 802 if (ref == GRANT_INVALID_REF) { 803 if (net_ratelimit()) 804 dev_warn(dev, "Bad rx response id %d.\n", 805 rx->id); 806 err = -EINVAL; 807 goto next; 808 } 809 810 ret = gnttab_end_foreign_access_ref(ref, 0); 811 BUG_ON(!ret); 812 813 gnttab_release_grant_reference(&queue->gref_rx_head, ref); 814 815 __skb_queue_tail(list, skb); 816 817 next: 818 if (!(rx->flags & XEN_NETRXF_more_data)) 819 break; 820 821 if (cons + slots == rp) { 822 if (net_ratelimit()) 823 dev_warn(dev, "Need more slots\n"); 824 err = -ENOENT; 825 break; 826 } 827 828 rx = RING_GET_RESPONSE(&queue->rx, cons + slots); 829 skb = xennet_get_rx_skb(queue, cons + slots); 830 ref = xennet_get_rx_ref(queue, cons + slots); 831 slots++; 832 } 833 834 if (unlikely(slots > max)) { 835 if (net_ratelimit()) 836 dev_warn(dev, "Too many slots\n"); 837 err = -E2BIG; 838 } 839 840 if (unlikely(err)) 841 queue->rx.rsp_cons = cons + slots; 842 843 return err; 844 } 845 846 static int xennet_set_skb_gso(struct sk_buff *skb, 847 struct xen_netif_extra_info *gso) 848 { 849 if (!gso->u.gso.size) { 850 if (net_ratelimit()) 851 pr_warn("GSO size must not be zero\n"); 852 return -EINVAL; 853 } 854 855 if (gso->u.gso.type != XEN_NETIF_GSO_TYPE_TCPV4 && 856 gso->u.gso.type != XEN_NETIF_GSO_TYPE_TCPV6) { 857 if (net_ratelimit()) 858 pr_warn("Bad GSO type %d\n", gso->u.gso.type); 859 return -EINVAL; 860 } 861 862 skb_shinfo(skb)->gso_size = gso->u.gso.size; 863 skb_shinfo(skb)->gso_type = 864 (gso->u.gso.type == XEN_NETIF_GSO_TYPE_TCPV4) ? 865 SKB_GSO_TCPV4 : 866 SKB_GSO_TCPV6; 867 868 /* Header must be checked, and gso_segs computed. */ 869 skb_shinfo(skb)->gso_type |= SKB_GSO_DODGY; 870 skb_shinfo(skb)->gso_segs = 0; 871 872 return 0; 873 } 874 875 static RING_IDX xennet_fill_frags(struct netfront_queue *queue, 876 struct sk_buff *skb, 877 struct sk_buff_head *list) 878 { 879 struct skb_shared_info *shinfo = skb_shinfo(skb); 880 RING_IDX cons = queue->rx.rsp_cons; 881 struct sk_buff *nskb; 882 883 while ((nskb = __skb_dequeue(list))) { 884 struct xen_netif_rx_response *rx = 885 RING_GET_RESPONSE(&queue->rx, ++cons); 886 skb_frag_t *nfrag = &skb_shinfo(nskb)->frags[0]; 887 888 if (shinfo->nr_frags == MAX_SKB_FRAGS) { 889 unsigned int pull_to = NETFRONT_SKB_CB(skb)->pull_to; 890 891 BUG_ON(pull_to <= skb_headlen(skb)); 892 __pskb_pull_tail(skb, pull_to - skb_headlen(skb)); 893 } 894 BUG_ON(shinfo->nr_frags >= MAX_SKB_FRAGS); 895 896 skb_add_rx_frag(skb, shinfo->nr_frags, skb_frag_page(nfrag), 897 rx->offset, rx->status, PAGE_SIZE); 898 899 skb_shinfo(nskb)->nr_frags = 0; 900 kfree_skb(nskb); 901 } 902 903 return cons; 904 } 905 906 static int checksum_setup(struct net_device *dev, struct sk_buff *skb) 907 { 908 bool recalculate_partial_csum = false; 909 910 /* 911 * A GSO SKB must be CHECKSUM_PARTIAL. However some buggy 912 * peers can fail to set NETRXF_csum_blank when sending a GSO 913 * frame. In this case force the SKB to CHECKSUM_PARTIAL and 914 * recalculate the partial checksum. 915 */ 916 if (skb->ip_summed != CHECKSUM_PARTIAL && skb_is_gso(skb)) { 917 struct netfront_info *np = netdev_priv(dev); 918 atomic_inc(&np->rx_gso_checksum_fixup); 919 skb->ip_summed = CHECKSUM_PARTIAL; 920 recalculate_partial_csum = true; 921 } 922 923 /* A non-CHECKSUM_PARTIAL SKB does not require setup. */ 924 if (skb->ip_summed != CHECKSUM_PARTIAL) 925 return 0; 926 927 return skb_checksum_setup(skb, recalculate_partial_csum); 928 } 929 930 static int handle_incoming_queue(struct netfront_queue *queue, 931 struct sk_buff_head *rxq) 932 { 933 struct netfront_stats *rx_stats = this_cpu_ptr(queue->info->rx_stats); 934 int packets_dropped = 0; 935 struct sk_buff *skb; 936 937 while ((skb = __skb_dequeue(rxq)) != NULL) { 938 int pull_to = NETFRONT_SKB_CB(skb)->pull_to; 939 940 if (pull_to > skb_headlen(skb)) 941 __pskb_pull_tail(skb, pull_to - skb_headlen(skb)); 942 943 /* Ethernet work: Delayed to here as it peeks the header. */ 944 skb->protocol = eth_type_trans(skb, queue->info->netdev); 945 skb_reset_network_header(skb); 946 947 if (checksum_setup(queue->info->netdev, skb)) { 948 kfree_skb(skb); 949 packets_dropped++; 950 queue->info->netdev->stats.rx_errors++; 951 continue; 952 } 953 954 u64_stats_update_begin(&rx_stats->syncp); 955 rx_stats->packets++; 956 rx_stats->bytes += skb->len; 957 u64_stats_update_end(&rx_stats->syncp); 958 959 /* Pass it up. */ 960 napi_gro_receive(&queue->napi, skb); 961 } 962 963 return packets_dropped; 964 } 965 966 static int xennet_poll(struct napi_struct *napi, int budget) 967 { 968 struct netfront_queue *queue = container_of(napi, struct netfront_queue, napi); 969 struct net_device *dev = queue->info->netdev; 970 struct sk_buff *skb; 971 struct netfront_rx_info rinfo; 972 struct xen_netif_rx_response *rx = &rinfo.rx; 973 struct xen_netif_extra_info *extras = rinfo.extras; 974 RING_IDX i, rp; 975 int work_done; 976 struct sk_buff_head rxq; 977 struct sk_buff_head errq; 978 struct sk_buff_head tmpq; 979 int err; 980 981 spin_lock(&queue->rx_lock); 982 983 skb_queue_head_init(&rxq); 984 skb_queue_head_init(&errq); 985 skb_queue_head_init(&tmpq); 986 987 rp = queue->rx.sring->rsp_prod; 988 rmb(); /* Ensure we see queued responses up to 'rp'. */ 989 990 i = queue->rx.rsp_cons; 991 work_done = 0; 992 while ((i != rp) && (work_done < budget)) { 993 memcpy(rx, RING_GET_RESPONSE(&queue->rx, i), sizeof(*rx)); 994 memset(extras, 0, sizeof(rinfo.extras)); 995 996 err = xennet_get_responses(queue, &rinfo, rp, &tmpq); 997 998 if (unlikely(err)) { 999 err: 1000 while ((skb = __skb_dequeue(&tmpq))) 1001 __skb_queue_tail(&errq, skb); 1002 dev->stats.rx_errors++; 1003 i = queue->rx.rsp_cons; 1004 continue; 1005 } 1006 1007 skb = __skb_dequeue(&tmpq); 1008 1009 if (extras[XEN_NETIF_EXTRA_TYPE_GSO - 1].type) { 1010 struct xen_netif_extra_info *gso; 1011 gso = &extras[XEN_NETIF_EXTRA_TYPE_GSO - 1]; 1012 1013 if (unlikely(xennet_set_skb_gso(skb, gso))) { 1014 __skb_queue_head(&tmpq, skb); 1015 queue->rx.rsp_cons += skb_queue_len(&tmpq); 1016 goto err; 1017 } 1018 } 1019 1020 NETFRONT_SKB_CB(skb)->pull_to = rx->status; 1021 if (NETFRONT_SKB_CB(skb)->pull_to > RX_COPY_THRESHOLD) 1022 NETFRONT_SKB_CB(skb)->pull_to = RX_COPY_THRESHOLD; 1023 1024 skb_shinfo(skb)->frags[0].page_offset = rx->offset; 1025 skb_frag_size_set(&skb_shinfo(skb)->frags[0], rx->status); 1026 skb->data_len = rx->status; 1027 skb->len += rx->status; 1028 1029 i = xennet_fill_frags(queue, skb, &tmpq); 1030 1031 if (rx->flags & XEN_NETRXF_csum_blank) 1032 skb->ip_summed = CHECKSUM_PARTIAL; 1033 else if (rx->flags & XEN_NETRXF_data_validated) 1034 skb->ip_summed = CHECKSUM_UNNECESSARY; 1035 1036 __skb_queue_tail(&rxq, skb); 1037 1038 queue->rx.rsp_cons = ++i; 1039 work_done++; 1040 } 1041 1042 __skb_queue_purge(&errq); 1043 1044 work_done -= handle_incoming_queue(queue, &rxq); 1045 1046 xennet_alloc_rx_buffers(queue); 1047 1048 if (work_done < budget) { 1049 int more_to_do = 0; 1050 1051 napi_complete(napi); 1052 1053 RING_FINAL_CHECK_FOR_RESPONSES(&queue->rx, more_to_do); 1054 if (more_to_do) 1055 napi_schedule(napi); 1056 } 1057 1058 spin_unlock(&queue->rx_lock); 1059 1060 return work_done; 1061 } 1062 1063 static int xennet_change_mtu(struct net_device *dev, int mtu) 1064 { 1065 int max = xennet_can_sg(dev) ? 1066 XEN_NETIF_MAX_TX_SIZE - MAX_TCP_HEADER : ETH_DATA_LEN; 1067 1068 if (mtu > max) 1069 return -EINVAL; 1070 dev->mtu = mtu; 1071 return 0; 1072 } 1073 1074 static struct rtnl_link_stats64 *xennet_get_stats64(struct net_device *dev, 1075 struct rtnl_link_stats64 *tot) 1076 { 1077 struct netfront_info *np = netdev_priv(dev); 1078 int cpu; 1079 1080 for_each_possible_cpu(cpu) { 1081 struct netfront_stats *rx_stats = per_cpu_ptr(np->rx_stats, cpu); 1082 struct netfront_stats *tx_stats = per_cpu_ptr(np->tx_stats, cpu); 1083 u64 rx_packets, rx_bytes, tx_packets, tx_bytes; 1084 unsigned int start; 1085 1086 do { 1087 start = u64_stats_fetch_begin_irq(&tx_stats->syncp); 1088 tx_packets = tx_stats->packets; 1089 tx_bytes = tx_stats->bytes; 1090 } while (u64_stats_fetch_retry_irq(&tx_stats->syncp, start)); 1091 1092 do { 1093 start = u64_stats_fetch_begin_irq(&rx_stats->syncp); 1094 rx_packets = rx_stats->packets; 1095 rx_bytes = rx_stats->bytes; 1096 } while (u64_stats_fetch_retry_irq(&rx_stats->syncp, start)); 1097 1098 tot->rx_packets += rx_packets; 1099 tot->tx_packets += tx_packets; 1100 tot->rx_bytes += rx_bytes; 1101 tot->tx_bytes += tx_bytes; 1102 } 1103 1104 tot->rx_errors = dev->stats.rx_errors; 1105 tot->tx_dropped = dev->stats.tx_dropped; 1106 1107 return tot; 1108 } 1109 1110 static void xennet_release_tx_bufs(struct netfront_queue *queue) 1111 { 1112 struct sk_buff *skb; 1113 int i; 1114 1115 for (i = 0; i < NET_TX_RING_SIZE; i++) { 1116 /* Skip over entries which are actually freelist references */ 1117 if (skb_entry_is_link(&queue->tx_skbs[i])) 1118 continue; 1119 1120 skb = queue->tx_skbs[i].skb; 1121 get_page(queue->grant_tx_page[i]); 1122 gnttab_end_foreign_access(queue->grant_tx_ref[i], 1123 GNTMAP_readonly, 1124 (unsigned long)page_address(queue->grant_tx_page[i])); 1125 queue->grant_tx_page[i] = NULL; 1126 queue->grant_tx_ref[i] = GRANT_INVALID_REF; 1127 add_id_to_freelist(&queue->tx_skb_freelist, queue->tx_skbs, i); 1128 dev_kfree_skb_irq(skb); 1129 } 1130 } 1131 1132 static void xennet_release_rx_bufs(struct netfront_queue *queue) 1133 { 1134 int id, ref; 1135 1136 spin_lock_bh(&queue->rx_lock); 1137 1138 for (id = 0; id < NET_RX_RING_SIZE; id++) { 1139 struct sk_buff *skb; 1140 struct page *page; 1141 1142 skb = queue->rx_skbs[id]; 1143 if (!skb) 1144 continue; 1145 1146 ref = queue->grant_rx_ref[id]; 1147 if (ref == GRANT_INVALID_REF) 1148 continue; 1149 1150 page = skb_frag_page(&skb_shinfo(skb)->frags[0]); 1151 1152 /* gnttab_end_foreign_access() needs a page ref until 1153 * foreign access is ended (which may be deferred). 1154 */ 1155 get_page(page); 1156 gnttab_end_foreign_access(ref, 0, 1157 (unsigned long)page_address(page)); 1158 queue->grant_rx_ref[id] = GRANT_INVALID_REF; 1159 1160 kfree_skb(skb); 1161 } 1162 1163 spin_unlock_bh(&queue->rx_lock); 1164 } 1165 1166 static netdev_features_t xennet_fix_features(struct net_device *dev, 1167 netdev_features_t features) 1168 { 1169 struct netfront_info *np = netdev_priv(dev); 1170 int val; 1171 1172 if (features & NETIF_F_SG) { 1173 if (xenbus_scanf(XBT_NIL, np->xbdev->otherend, "feature-sg", 1174 "%d", &val) < 0) 1175 val = 0; 1176 1177 if (!val) 1178 features &= ~NETIF_F_SG; 1179 } 1180 1181 if (features & NETIF_F_IPV6_CSUM) { 1182 if (xenbus_scanf(XBT_NIL, np->xbdev->otherend, 1183 "feature-ipv6-csum-offload", "%d", &val) < 0) 1184 val = 0; 1185 1186 if (!val) 1187 features &= ~NETIF_F_IPV6_CSUM; 1188 } 1189 1190 if (features & NETIF_F_TSO) { 1191 if (xenbus_scanf(XBT_NIL, np->xbdev->otherend, 1192 "feature-gso-tcpv4", "%d", &val) < 0) 1193 val = 0; 1194 1195 if (!val) 1196 features &= ~NETIF_F_TSO; 1197 } 1198 1199 if (features & NETIF_F_TSO6) { 1200 if (xenbus_scanf(XBT_NIL, np->xbdev->otherend, 1201 "feature-gso-tcpv6", "%d", &val) < 0) 1202 val = 0; 1203 1204 if (!val) 1205 features &= ~NETIF_F_TSO6; 1206 } 1207 1208 return features; 1209 } 1210 1211 static int xennet_set_features(struct net_device *dev, 1212 netdev_features_t features) 1213 { 1214 if (!(features & NETIF_F_SG) && dev->mtu > ETH_DATA_LEN) { 1215 netdev_info(dev, "Reducing MTU because no SG offload"); 1216 dev->mtu = ETH_DATA_LEN; 1217 } 1218 1219 return 0; 1220 } 1221 1222 static irqreturn_t xennet_tx_interrupt(int irq, void *dev_id) 1223 { 1224 struct netfront_queue *queue = dev_id; 1225 unsigned long flags; 1226 1227 spin_lock_irqsave(&queue->tx_lock, flags); 1228 xennet_tx_buf_gc(queue); 1229 spin_unlock_irqrestore(&queue->tx_lock, flags); 1230 1231 return IRQ_HANDLED; 1232 } 1233 1234 static irqreturn_t xennet_rx_interrupt(int irq, void *dev_id) 1235 { 1236 struct netfront_queue *queue = dev_id; 1237 struct net_device *dev = queue->info->netdev; 1238 1239 if (likely(netif_carrier_ok(dev) && 1240 RING_HAS_UNCONSUMED_RESPONSES(&queue->rx))) 1241 napi_schedule(&queue->napi); 1242 1243 return IRQ_HANDLED; 1244 } 1245 1246 static irqreturn_t xennet_interrupt(int irq, void *dev_id) 1247 { 1248 xennet_tx_interrupt(irq, dev_id); 1249 xennet_rx_interrupt(irq, dev_id); 1250 return IRQ_HANDLED; 1251 } 1252 1253 #ifdef CONFIG_NET_POLL_CONTROLLER 1254 static void xennet_poll_controller(struct net_device *dev) 1255 { 1256 /* Poll each queue */ 1257 struct netfront_info *info = netdev_priv(dev); 1258 unsigned int num_queues = dev->real_num_tx_queues; 1259 unsigned int i; 1260 for (i = 0; i < num_queues; ++i) 1261 xennet_interrupt(0, &info->queues[i]); 1262 } 1263 #endif 1264 1265 static const struct net_device_ops xennet_netdev_ops = { 1266 .ndo_open = xennet_open, 1267 .ndo_stop = xennet_close, 1268 .ndo_start_xmit = xennet_start_xmit, 1269 .ndo_change_mtu = xennet_change_mtu, 1270 .ndo_get_stats64 = xennet_get_stats64, 1271 .ndo_set_mac_address = eth_mac_addr, 1272 .ndo_validate_addr = eth_validate_addr, 1273 .ndo_fix_features = xennet_fix_features, 1274 .ndo_set_features = xennet_set_features, 1275 .ndo_select_queue = xennet_select_queue, 1276 #ifdef CONFIG_NET_POLL_CONTROLLER 1277 .ndo_poll_controller = xennet_poll_controller, 1278 #endif 1279 }; 1280 1281 static void xennet_free_netdev(struct net_device *netdev) 1282 { 1283 struct netfront_info *np = netdev_priv(netdev); 1284 1285 free_percpu(np->rx_stats); 1286 free_percpu(np->tx_stats); 1287 free_netdev(netdev); 1288 } 1289 1290 static struct net_device *xennet_create_dev(struct xenbus_device *dev) 1291 { 1292 int err; 1293 struct net_device *netdev; 1294 struct netfront_info *np; 1295 1296 netdev = alloc_etherdev_mq(sizeof(struct netfront_info), xennet_max_queues); 1297 if (!netdev) 1298 return ERR_PTR(-ENOMEM); 1299 1300 np = netdev_priv(netdev); 1301 np->xbdev = dev; 1302 1303 /* No need to use rtnl_lock() before the call below as it 1304 * happens before register_netdev(). 1305 */ 1306 netif_set_real_num_tx_queues(netdev, 0); 1307 np->queues = NULL; 1308 1309 err = -ENOMEM; 1310 np->rx_stats = netdev_alloc_pcpu_stats(struct netfront_stats); 1311 if (np->rx_stats == NULL) 1312 goto exit; 1313 np->tx_stats = netdev_alloc_pcpu_stats(struct netfront_stats); 1314 if (np->tx_stats == NULL) 1315 goto exit; 1316 1317 netdev->netdev_ops = &xennet_netdev_ops; 1318 1319 netdev->features = NETIF_F_IP_CSUM | NETIF_F_RXCSUM | 1320 NETIF_F_GSO_ROBUST; 1321 netdev->hw_features = NETIF_F_SG | 1322 NETIF_F_IPV6_CSUM | 1323 NETIF_F_TSO | NETIF_F_TSO6; 1324 1325 /* 1326 * Assume that all hw features are available for now. This set 1327 * will be adjusted by the call to netdev_update_features() in 1328 * xennet_connect() which is the earliest point where we can 1329 * negotiate with the backend regarding supported features. 1330 */ 1331 netdev->features |= netdev->hw_features; 1332 1333 netdev->ethtool_ops = &xennet_ethtool_ops; 1334 SET_NETDEV_DEV(netdev, &dev->dev); 1335 1336 netif_set_gso_max_size(netdev, XEN_NETIF_MAX_TX_SIZE - MAX_TCP_HEADER); 1337 1338 np->netdev = netdev; 1339 1340 netif_carrier_off(netdev); 1341 1342 return netdev; 1343 1344 exit: 1345 xennet_free_netdev(netdev); 1346 return ERR_PTR(err); 1347 } 1348 1349 /** 1350 * Entry point to this code when a new device is created. Allocate the basic 1351 * structures and the ring buffers for communication with the backend, and 1352 * inform the backend of the appropriate details for those. 1353 */ 1354 static int netfront_probe(struct xenbus_device *dev, 1355 const struct xenbus_device_id *id) 1356 { 1357 int err; 1358 struct net_device *netdev; 1359 struct netfront_info *info; 1360 1361 netdev = xennet_create_dev(dev); 1362 if (IS_ERR(netdev)) { 1363 err = PTR_ERR(netdev); 1364 xenbus_dev_fatal(dev, err, "creating netdev"); 1365 return err; 1366 } 1367 1368 info = netdev_priv(netdev); 1369 dev_set_drvdata(&dev->dev, info); 1370 1371 err = register_netdev(info->netdev); 1372 if (err) { 1373 pr_warn("%s: register_netdev err=%d\n", __func__, err); 1374 goto fail; 1375 } 1376 1377 err = xennet_sysfs_addif(info->netdev); 1378 if (err) { 1379 unregister_netdev(info->netdev); 1380 pr_warn("%s: add sysfs failed err=%d\n", __func__, err); 1381 goto fail; 1382 } 1383 1384 return 0; 1385 1386 fail: 1387 xennet_free_netdev(netdev); 1388 dev_set_drvdata(&dev->dev, NULL); 1389 return err; 1390 } 1391 1392 static void xennet_end_access(int ref, void *page) 1393 { 1394 /* This frees the page as a side-effect */ 1395 if (ref != GRANT_INVALID_REF) 1396 gnttab_end_foreign_access(ref, 0, (unsigned long)page); 1397 } 1398 1399 static void xennet_disconnect_backend(struct netfront_info *info) 1400 { 1401 unsigned int i = 0; 1402 unsigned int num_queues = info->netdev->real_num_tx_queues; 1403 1404 netif_carrier_off(info->netdev); 1405 1406 for (i = 0; i < num_queues; ++i) { 1407 struct netfront_queue *queue = &info->queues[i]; 1408 1409 if (queue->tx_irq && (queue->tx_irq == queue->rx_irq)) 1410 unbind_from_irqhandler(queue->tx_irq, queue); 1411 if (queue->tx_irq && (queue->tx_irq != queue->rx_irq)) { 1412 unbind_from_irqhandler(queue->tx_irq, queue); 1413 unbind_from_irqhandler(queue->rx_irq, queue); 1414 } 1415 queue->tx_evtchn = queue->rx_evtchn = 0; 1416 queue->tx_irq = queue->rx_irq = 0; 1417 1418 napi_synchronize(&queue->napi); 1419 1420 xennet_release_tx_bufs(queue); 1421 xennet_release_rx_bufs(queue); 1422 gnttab_free_grant_references(queue->gref_tx_head); 1423 gnttab_free_grant_references(queue->gref_rx_head); 1424 1425 /* End access and free the pages */ 1426 xennet_end_access(queue->tx_ring_ref, queue->tx.sring); 1427 xennet_end_access(queue->rx_ring_ref, queue->rx.sring); 1428 1429 queue->tx_ring_ref = GRANT_INVALID_REF; 1430 queue->rx_ring_ref = GRANT_INVALID_REF; 1431 queue->tx.sring = NULL; 1432 queue->rx.sring = NULL; 1433 } 1434 } 1435 1436 /** 1437 * We are reconnecting to the backend, due to a suspend/resume, or a backend 1438 * driver restart. We tear down our netif structure and recreate it, but 1439 * leave the device-layer structures intact so that this is transparent to the 1440 * rest of the kernel. 1441 */ 1442 static int netfront_resume(struct xenbus_device *dev) 1443 { 1444 struct netfront_info *info = dev_get_drvdata(&dev->dev); 1445 1446 dev_dbg(&dev->dev, "%s\n", dev->nodename); 1447 1448 xennet_disconnect_backend(info); 1449 return 0; 1450 } 1451 1452 static int xen_net_read_mac(struct xenbus_device *dev, u8 mac[]) 1453 { 1454 char *s, *e, *macstr; 1455 int i; 1456 1457 macstr = s = xenbus_read(XBT_NIL, dev->nodename, "mac", NULL); 1458 if (IS_ERR(macstr)) 1459 return PTR_ERR(macstr); 1460 1461 for (i = 0; i < ETH_ALEN; i++) { 1462 mac[i] = simple_strtoul(s, &e, 16); 1463 if ((s == e) || (*e != ((i == ETH_ALEN-1) ? '\0' : ':'))) { 1464 kfree(macstr); 1465 return -ENOENT; 1466 } 1467 s = e+1; 1468 } 1469 1470 kfree(macstr); 1471 return 0; 1472 } 1473 1474 static int setup_netfront_single(struct netfront_queue *queue) 1475 { 1476 int err; 1477 1478 err = xenbus_alloc_evtchn(queue->info->xbdev, &queue->tx_evtchn); 1479 if (err < 0) 1480 goto fail; 1481 1482 err = bind_evtchn_to_irqhandler(queue->tx_evtchn, 1483 xennet_interrupt, 1484 0, queue->info->netdev->name, queue); 1485 if (err < 0) 1486 goto bind_fail; 1487 queue->rx_evtchn = queue->tx_evtchn; 1488 queue->rx_irq = queue->tx_irq = err; 1489 1490 return 0; 1491 1492 bind_fail: 1493 xenbus_free_evtchn(queue->info->xbdev, queue->tx_evtchn); 1494 queue->tx_evtchn = 0; 1495 fail: 1496 return err; 1497 } 1498 1499 static int setup_netfront_split(struct netfront_queue *queue) 1500 { 1501 int err; 1502 1503 err = xenbus_alloc_evtchn(queue->info->xbdev, &queue->tx_evtchn); 1504 if (err < 0) 1505 goto fail; 1506 err = xenbus_alloc_evtchn(queue->info->xbdev, &queue->rx_evtchn); 1507 if (err < 0) 1508 goto alloc_rx_evtchn_fail; 1509 1510 snprintf(queue->tx_irq_name, sizeof(queue->tx_irq_name), 1511 "%s-tx", queue->name); 1512 err = bind_evtchn_to_irqhandler(queue->tx_evtchn, 1513 xennet_tx_interrupt, 1514 0, queue->tx_irq_name, queue); 1515 if (err < 0) 1516 goto bind_tx_fail; 1517 queue->tx_irq = err; 1518 1519 snprintf(queue->rx_irq_name, sizeof(queue->rx_irq_name), 1520 "%s-rx", queue->name); 1521 err = bind_evtchn_to_irqhandler(queue->rx_evtchn, 1522 xennet_rx_interrupt, 1523 0, queue->rx_irq_name, queue); 1524 if (err < 0) 1525 goto bind_rx_fail; 1526 queue->rx_irq = err; 1527 1528 return 0; 1529 1530 bind_rx_fail: 1531 unbind_from_irqhandler(queue->tx_irq, queue); 1532 queue->tx_irq = 0; 1533 bind_tx_fail: 1534 xenbus_free_evtchn(queue->info->xbdev, queue->rx_evtchn); 1535 queue->rx_evtchn = 0; 1536 alloc_rx_evtchn_fail: 1537 xenbus_free_evtchn(queue->info->xbdev, queue->tx_evtchn); 1538 queue->tx_evtchn = 0; 1539 fail: 1540 return err; 1541 } 1542 1543 static int setup_netfront(struct xenbus_device *dev, 1544 struct netfront_queue *queue, unsigned int feature_split_evtchn) 1545 { 1546 struct xen_netif_tx_sring *txs; 1547 struct xen_netif_rx_sring *rxs; 1548 int err; 1549 1550 queue->tx_ring_ref = GRANT_INVALID_REF; 1551 queue->rx_ring_ref = GRANT_INVALID_REF; 1552 queue->rx.sring = NULL; 1553 queue->tx.sring = NULL; 1554 1555 txs = (struct xen_netif_tx_sring *)get_zeroed_page(GFP_NOIO | __GFP_HIGH); 1556 if (!txs) { 1557 err = -ENOMEM; 1558 xenbus_dev_fatal(dev, err, "allocating tx ring page"); 1559 goto fail; 1560 } 1561 SHARED_RING_INIT(txs); 1562 FRONT_RING_INIT(&queue->tx, txs, PAGE_SIZE); 1563 1564 err = xenbus_grant_ring(dev, virt_to_mfn(txs)); 1565 if (err < 0) 1566 goto grant_tx_ring_fail; 1567 queue->tx_ring_ref = err; 1568 1569 rxs = (struct xen_netif_rx_sring *)get_zeroed_page(GFP_NOIO | __GFP_HIGH); 1570 if (!rxs) { 1571 err = -ENOMEM; 1572 xenbus_dev_fatal(dev, err, "allocating rx ring page"); 1573 goto alloc_rx_ring_fail; 1574 } 1575 SHARED_RING_INIT(rxs); 1576 FRONT_RING_INIT(&queue->rx, rxs, PAGE_SIZE); 1577 1578 err = xenbus_grant_ring(dev, virt_to_mfn(rxs)); 1579 if (err < 0) 1580 goto grant_rx_ring_fail; 1581 queue->rx_ring_ref = err; 1582 1583 if (feature_split_evtchn) 1584 err = setup_netfront_split(queue); 1585 /* setup single event channel if 1586 * a) feature-split-event-channels == 0 1587 * b) feature-split-event-channels == 1 but failed to setup 1588 */ 1589 if (!feature_split_evtchn || (feature_split_evtchn && err)) 1590 err = setup_netfront_single(queue); 1591 1592 if (err) 1593 goto alloc_evtchn_fail; 1594 1595 return 0; 1596 1597 /* If we fail to setup netfront, it is safe to just revoke access to 1598 * granted pages because backend is not accessing it at this point. 1599 */ 1600 alloc_evtchn_fail: 1601 gnttab_end_foreign_access_ref(queue->rx_ring_ref, 0); 1602 grant_rx_ring_fail: 1603 free_page((unsigned long)rxs); 1604 alloc_rx_ring_fail: 1605 gnttab_end_foreign_access_ref(queue->tx_ring_ref, 0); 1606 grant_tx_ring_fail: 1607 free_page((unsigned long)txs); 1608 fail: 1609 return err; 1610 } 1611 1612 /* Queue-specific initialisation 1613 * This used to be done in xennet_create_dev() but must now 1614 * be run per-queue. 1615 */ 1616 static int xennet_init_queue(struct netfront_queue *queue) 1617 { 1618 unsigned short i; 1619 int err = 0; 1620 1621 spin_lock_init(&queue->tx_lock); 1622 spin_lock_init(&queue->rx_lock); 1623 1624 init_timer(&queue->rx_refill_timer); 1625 queue->rx_refill_timer.data = (unsigned long)queue; 1626 queue->rx_refill_timer.function = rx_refill_timeout; 1627 1628 snprintf(queue->name, sizeof(queue->name), "%s-q%u", 1629 queue->info->netdev->name, queue->id); 1630 1631 /* Initialise tx_skbs as a free chain containing every entry. */ 1632 queue->tx_skb_freelist = 0; 1633 for (i = 0; i < NET_TX_RING_SIZE; i++) { 1634 skb_entry_set_link(&queue->tx_skbs[i], i+1); 1635 queue->grant_tx_ref[i] = GRANT_INVALID_REF; 1636 queue->grant_tx_page[i] = NULL; 1637 } 1638 1639 /* Clear out rx_skbs */ 1640 for (i = 0; i < NET_RX_RING_SIZE; i++) { 1641 queue->rx_skbs[i] = NULL; 1642 queue->grant_rx_ref[i] = GRANT_INVALID_REF; 1643 } 1644 1645 /* A grant for every tx ring slot */ 1646 if (gnttab_alloc_grant_references(NET_TX_RING_SIZE, 1647 &queue->gref_tx_head) < 0) { 1648 pr_alert("can't alloc tx grant refs\n"); 1649 err = -ENOMEM; 1650 goto exit; 1651 } 1652 1653 /* A grant for every rx ring slot */ 1654 if (gnttab_alloc_grant_references(NET_RX_RING_SIZE, 1655 &queue->gref_rx_head) < 0) { 1656 pr_alert("can't alloc rx grant refs\n"); 1657 err = -ENOMEM; 1658 goto exit_free_tx; 1659 } 1660 1661 return 0; 1662 1663 exit_free_tx: 1664 gnttab_free_grant_references(queue->gref_tx_head); 1665 exit: 1666 return err; 1667 } 1668 1669 static int write_queue_xenstore_keys(struct netfront_queue *queue, 1670 struct xenbus_transaction *xbt, int write_hierarchical) 1671 { 1672 /* Write the queue-specific keys into XenStore in the traditional 1673 * way for a single queue, or in a queue subkeys for multiple 1674 * queues. 1675 */ 1676 struct xenbus_device *dev = queue->info->xbdev; 1677 int err; 1678 const char *message; 1679 char *path; 1680 size_t pathsize; 1681 1682 /* Choose the correct place to write the keys */ 1683 if (write_hierarchical) { 1684 pathsize = strlen(dev->nodename) + 10; 1685 path = kzalloc(pathsize, GFP_KERNEL); 1686 if (!path) { 1687 err = -ENOMEM; 1688 message = "out of memory while writing ring references"; 1689 goto error; 1690 } 1691 snprintf(path, pathsize, "%s/queue-%u", 1692 dev->nodename, queue->id); 1693 } else { 1694 path = (char *)dev->nodename; 1695 } 1696 1697 /* Write ring references */ 1698 err = xenbus_printf(*xbt, path, "tx-ring-ref", "%u", 1699 queue->tx_ring_ref); 1700 if (err) { 1701 message = "writing tx-ring-ref"; 1702 goto error; 1703 } 1704 1705 err = xenbus_printf(*xbt, path, "rx-ring-ref", "%u", 1706 queue->rx_ring_ref); 1707 if (err) { 1708 message = "writing rx-ring-ref"; 1709 goto error; 1710 } 1711 1712 /* Write event channels; taking into account both shared 1713 * and split event channel scenarios. 1714 */ 1715 if (queue->tx_evtchn == queue->rx_evtchn) { 1716 /* Shared event channel */ 1717 err = xenbus_printf(*xbt, path, 1718 "event-channel", "%u", queue->tx_evtchn); 1719 if (err) { 1720 message = "writing event-channel"; 1721 goto error; 1722 } 1723 } else { 1724 /* Split event channels */ 1725 err = xenbus_printf(*xbt, path, 1726 "event-channel-tx", "%u", queue->tx_evtchn); 1727 if (err) { 1728 message = "writing event-channel-tx"; 1729 goto error; 1730 } 1731 1732 err = xenbus_printf(*xbt, path, 1733 "event-channel-rx", "%u", queue->rx_evtchn); 1734 if (err) { 1735 message = "writing event-channel-rx"; 1736 goto error; 1737 } 1738 } 1739 1740 if (write_hierarchical) 1741 kfree(path); 1742 return 0; 1743 1744 error: 1745 if (write_hierarchical) 1746 kfree(path); 1747 xenbus_dev_fatal(dev, err, "%s", message); 1748 return err; 1749 } 1750 1751 static void xennet_destroy_queues(struct netfront_info *info) 1752 { 1753 unsigned int i; 1754 1755 rtnl_lock(); 1756 1757 for (i = 0; i < info->netdev->real_num_tx_queues; i++) { 1758 struct netfront_queue *queue = &info->queues[i]; 1759 1760 if (netif_running(info->netdev)) 1761 napi_disable(&queue->napi); 1762 netif_napi_del(&queue->napi); 1763 } 1764 1765 rtnl_unlock(); 1766 1767 kfree(info->queues); 1768 info->queues = NULL; 1769 } 1770 1771 static int xennet_create_queues(struct netfront_info *info, 1772 unsigned int num_queues) 1773 { 1774 unsigned int i; 1775 int ret; 1776 1777 info->queues = kcalloc(num_queues, sizeof(struct netfront_queue), 1778 GFP_KERNEL); 1779 if (!info->queues) 1780 return -ENOMEM; 1781 1782 rtnl_lock(); 1783 1784 for (i = 0; i < num_queues; i++) { 1785 struct netfront_queue *queue = &info->queues[i]; 1786 1787 queue->id = i; 1788 queue->info = info; 1789 1790 ret = xennet_init_queue(queue); 1791 if (ret < 0) { 1792 dev_warn(&info->netdev->dev, 1793 "only created %d queues\n", i); 1794 num_queues = i; 1795 break; 1796 } 1797 1798 netif_napi_add(queue->info->netdev, &queue->napi, 1799 xennet_poll, 64); 1800 if (netif_running(info->netdev)) 1801 napi_enable(&queue->napi); 1802 } 1803 1804 netif_set_real_num_tx_queues(info->netdev, num_queues); 1805 1806 rtnl_unlock(); 1807 1808 if (num_queues == 0) { 1809 dev_err(&info->netdev->dev, "no queues\n"); 1810 return -EINVAL; 1811 } 1812 return 0; 1813 } 1814 1815 /* Common code used when first setting up, and when resuming. */ 1816 static int talk_to_netback(struct xenbus_device *dev, 1817 struct netfront_info *info) 1818 { 1819 const char *message; 1820 struct xenbus_transaction xbt; 1821 int err; 1822 unsigned int feature_split_evtchn; 1823 unsigned int i = 0; 1824 unsigned int max_queues = 0; 1825 struct netfront_queue *queue = NULL; 1826 unsigned int num_queues = 1; 1827 1828 info->netdev->irq = 0; 1829 1830 /* Check if backend supports multiple queues */ 1831 err = xenbus_scanf(XBT_NIL, info->xbdev->otherend, 1832 "multi-queue-max-queues", "%u", &max_queues); 1833 if (err < 0) 1834 max_queues = 1; 1835 num_queues = min(max_queues, xennet_max_queues); 1836 1837 /* Check feature-split-event-channels */ 1838 err = xenbus_scanf(XBT_NIL, info->xbdev->otherend, 1839 "feature-split-event-channels", "%u", 1840 &feature_split_evtchn); 1841 if (err < 0) 1842 feature_split_evtchn = 0; 1843 1844 /* Read mac addr. */ 1845 err = xen_net_read_mac(dev, info->netdev->dev_addr); 1846 if (err) { 1847 xenbus_dev_fatal(dev, err, "parsing %s/mac", dev->nodename); 1848 goto out; 1849 } 1850 1851 if (info->queues) 1852 xennet_destroy_queues(info); 1853 1854 err = xennet_create_queues(info, num_queues); 1855 if (err < 0) 1856 goto destroy_ring; 1857 1858 /* Create shared ring, alloc event channel -- for each queue */ 1859 for (i = 0; i < num_queues; ++i) { 1860 queue = &info->queues[i]; 1861 err = setup_netfront(dev, queue, feature_split_evtchn); 1862 if (err) { 1863 /* setup_netfront() will tidy up the current 1864 * queue on error, but we need to clean up 1865 * those already allocated. 1866 */ 1867 if (i > 0) { 1868 rtnl_lock(); 1869 netif_set_real_num_tx_queues(info->netdev, i); 1870 rtnl_unlock(); 1871 goto destroy_ring; 1872 } else { 1873 goto out; 1874 } 1875 } 1876 } 1877 1878 again: 1879 err = xenbus_transaction_start(&xbt); 1880 if (err) { 1881 xenbus_dev_fatal(dev, err, "starting transaction"); 1882 goto destroy_ring; 1883 } 1884 1885 if (num_queues == 1) { 1886 err = write_queue_xenstore_keys(&info->queues[0], &xbt, 0); /* flat */ 1887 if (err) 1888 goto abort_transaction_no_dev_fatal; 1889 } else { 1890 /* Write the number of queues */ 1891 err = xenbus_printf(xbt, dev->nodename, "multi-queue-num-queues", 1892 "%u", num_queues); 1893 if (err) { 1894 message = "writing multi-queue-num-queues"; 1895 goto abort_transaction_no_dev_fatal; 1896 } 1897 1898 /* Write the keys for each queue */ 1899 for (i = 0; i < num_queues; ++i) { 1900 queue = &info->queues[i]; 1901 err = write_queue_xenstore_keys(queue, &xbt, 1); /* hierarchical */ 1902 if (err) 1903 goto abort_transaction_no_dev_fatal; 1904 } 1905 } 1906 1907 /* The remaining keys are not queue-specific */ 1908 err = xenbus_printf(xbt, dev->nodename, "request-rx-copy", "%u", 1909 1); 1910 if (err) { 1911 message = "writing request-rx-copy"; 1912 goto abort_transaction; 1913 } 1914 1915 err = xenbus_printf(xbt, dev->nodename, "feature-rx-notify", "%d", 1); 1916 if (err) { 1917 message = "writing feature-rx-notify"; 1918 goto abort_transaction; 1919 } 1920 1921 err = xenbus_printf(xbt, dev->nodename, "feature-sg", "%d", 1); 1922 if (err) { 1923 message = "writing feature-sg"; 1924 goto abort_transaction; 1925 } 1926 1927 err = xenbus_printf(xbt, dev->nodename, "feature-gso-tcpv4", "%d", 1); 1928 if (err) { 1929 message = "writing feature-gso-tcpv4"; 1930 goto abort_transaction; 1931 } 1932 1933 err = xenbus_write(xbt, dev->nodename, "feature-gso-tcpv6", "1"); 1934 if (err) { 1935 message = "writing feature-gso-tcpv6"; 1936 goto abort_transaction; 1937 } 1938 1939 err = xenbus_write(xbt, dev->nodename, "feature-ipv6-csum-offload", 1940 "1"); 1941 if (err) { 1942 message = "writing feature-ipv6-csum-offload"; 1943 goto abort_transaction; 1944 } 1945 1946 err = xenbus_transaction_end(xbt, 0); 1947 if (err) { 1948 if (err == -EAGAIN) 1949 goto again; 1950 xenbus_dev_fatal(dev, err, "completing transaction"); 1951 goto destroy_ring; 1952 } 1953 1954 return 0; 1955 1956 abort_transaction: 1957 xenbus_dev_fatal(dev, err, "%s", message); 1958 abort_transaction_no_dev_fatal: 1959 xenbus_transaction_end(xbt, 1); 1960 destroy_ring: 1961 xennet_disconnect_backend(info); 1962 kfree(info->queues); 1963 info->queues = NULL; 1964 rtnl_lock(); 1965 netif_set_real_num_tx_queues(info->netdev, 0); 1966 rtnl_unlock(); 1967 out: 1968 return err; 1969 } 1970 1971 static int xennet_connect(struct net_device *dev) 1972 { 1973 struct netfront_info *np = netdev_priv(dev); 1974 unsigned int num_queues = 0; 1975 int err; 1976 unsigned int feature_rx_copy; 1977 unsigned int j = 0; 1978 struct netfront_queue *queue = NULL; 1979 1980 err = xenbus_scanf(XBT_NIL, np->xbdev->otherend, 1981 "feature-rx-copy", "%u", &feature_rx_copy); 1982 if (err != 1) 1983 feature_rx_copy = 0; 1984 1985 if (!feature_rx_copy) { 1986 dev_info(&dev->dev, 1987 "backend does not support copying receive path\n"); 1988 return -ENODEV; 1989 } 1990 1991 err = talk_to_netback(np->xbdev, np); 1992 if (err) 1993 return err; 1994 1995 /* talk_to_netback() sets the correct number of queues */ 1996 num_queues = dev->real_num_tx_queues; 1997 1998 rtnl_lock(); 1999 netdev_update_features(dev); 2000 rtnl_unlock(); 2001 2002 /* 2003 * All public and private state should now be sane. Get 2004 * ready to start sending and receiving packets and give the driver 2005 * domain a kick because we've probably just requeued some 2006 * packets. 2007 */ 2008 netif_carrier_on(np->netdev); 2009 for (j = 0; j < num_queues; ++j) { 2010 queue = &np->queues[j]; 2011 2012 notify_remote_via_irq(queue->tx_irq); 2013 if (queue->tx_irq != queue->rx_irq) 2014 notify_remote_via_irq(queue->rx_irq); 2015 2016 spin_lock_irq(&queue->tx_lock); 2017 xennet_tx_buf_gc(queue); 2018 spin_unlock_irq(&queue->tx_lock); 2019 2020 spin_lock_bh(&queue->rx_lock); 2021 xennet_alloc_rx_buffers(queue); 2022 spin_unlock_bh(&queue->rx_lock); 2023 } 2024 2025 return 0; 2026 } 2027 2028 /** 2029 * Callback received when the backend's state changes. 2030 */ 2031 static void netback_changed(struct xenbus_device *dev, 2032 enum xenbus_state backend_state) 2033 { 2034 struct netfront_info *np = dev_get_drvdata(&dev->dev); 2035 struct net_device *netdev = np->netdev; 2036 2037 dev_dbg(&dev->dev, "%s\n", xenbus_strstate(backend_state)); 2038 2039 switch (backend_state) { 2040 case XenbusStateInitialising: 2041 case XenbusStateInitialised: 2042 case XenbusStateReconfiguring: 2043 case XenbusStateReconfigured: 2044 case XenbusStateUnknown: 2045 break; 2046 2047 case XenbusStateInitWait: 2048 if (dev->state != XenbusStateInitialising) 2049 break; 2050 if (xennet_connect(netdev) != 0) 2051 break; 2052 xenbus_switch_state(dev, XenbusStateConnected); 2053 break; 2054 2055 case XenbusStateConnected: 2056 netdev_notify_peers(netdev); 2057 break; 2058 2059 case XenbusStateClosed: 2060 if (dev->state == XenbusStateClosed) 2061 break; 2062 /* Missed the backend's CLOSING state -- fallthrough */ 2063 case XenbusStateClosing: 2064 xenbus_frontend_closed(dev); 2065 break; 2066 } 2067 } 2068 2069 static const struct xennet_stat { 2070 char name[ETH_GSTRING_LEN]; 2071 u16 offset; 2072 } xennet_stats[] = { 2073 { 2074 "rx_gso_checksum_fixup", 2075 offsetof(struct netfront_info, rx_gso_checksum_fixup) 2076 }, 2077 }; 2078 2079 static int xennet_get_sset_count(struct net_device *dev, int string_set) 2080 { 2081 switch (string_set) { 2082 case ETH_SS_STATS: 2083 return ARRAY_SIZE(xennet_stats); 2084 default: 2085 return -EINVAL; 2086 } 2087 } 2088 2089 static void xennet_get_ethtool_stats(struct net_device *dev, 2090 struct ethtool_stats *stats, u64 * data) 2091 { 2092 void *np = netdev_priv(dev); 2093 int i; 2094 2095 for (i = 0; i < ARRAY_SIZE(xennet_stats); i++) 2096 data[i] = atomic_read((atomic_t *)(np + xennet_stats[i].offset)); 2097 } 2098 2099 static void xennet_get_strings(struct net_device *dev, u32 stringset, u8 * data) 2100 { 2101 int i; 2102 2103 switch (stringset) { 2104 case ETH_SS_STATS: 2105 for (i = 0; i < ARRAY_SIZE(xennet_stats); i++) 2106 memcpy(data + i * ETH_GSTRING_LEN, 2107 xennet_stats[i].name, ETH_GSTRING_LEN); 2108 break; 2109 } 2110 } 2111 2112 static const struct ethtool_ops xennet_ethtool_ops = 2113 { 2114 .get_link = ethtool_op_get_link, 2115 2116 .get_sset_count = xennet_get_sset_count, 2117 .get_ethtool_stats = xennet_get_ethtool_stats, 2118 .get_strings = xennet_get_strings, 2119 }; 2120 2121 #ifdef CONFIG_SYSFS 2122 static ssize_t show_rxbuf(struct device *dev, 2123 struct device_attribute *attr, char *buf) 2124 { 2125 return sprintf(buf, "%lu\n", NET_RX_RING_SIZE); 2126 } 2127 2128 static ssize_t store_rxbuf(struct device *dev, 2129 struct device_attribute *attr, 2130 const char *buf, size_t len) 2131 { 2132 char *endp; 2133 unsigned long target; 2134 2135 if (!capable(CAP_NET_ADMIN)) 2136 return -EPERM; 2137 2138 target = simple_strtoul(buf, &endp, 0); 2139 if (endp == buf) 2140 return -EBADMSG; 2141 2142 /* rxbuf_min and rxbuf_max are no longer configurable. */ 2143 2144 return len; 2145 } 2146 2147 static struct device_attribute xennet_attrs[] = { 2148 __ATTR(rxbuf_min, S_IRUGO|S_IWUSR, show_rxbuf, store_rxbuf), 2149 __ATTR(rxbuf_max, S_IRUGO|S_IWUSR, show_rxbuf, store_rxbuf), 2150 __ATTR(rxbuf_cur, S_IRUGO, show_rxbuf, NULL), 2151 }; 2152 2153 static int xennet_sysfs_addif(struct net_device *netdev) 2154 { 2155 int i; 2156 int err; 2157 2158 for (i = 0; i < ARRAY_SIZE(xennet_attrs); i++) { 2159 err = device_create_file(&netdev->dev, 2160 &xennet_attrs[i]); 2161 if (err) 2162 goto fail; 2163 } 2164 return 0; 2165 2166 fail: 2167 while (--i >= 0) 2168 device_remove_file(&netdev->dev, &xennet_attrs[i]); 2169 return err; 2170 } 2171 2172 static void xennet_sysfs_delif(struct net_device *netdev) 2173 { 2174 int i; 2175 2176 for (i = 0; i < ARRAY_SIZE(xennet_attrs); i++) 2177 device_remove_file(&netdev->dev, &xennet_attrs[i]); 2178 } 2179 2180 #endif /* CONFIG_SYSFS */ 2181 2182 static int xennet_remove(struct xenbus_device *dev) 2183 { 2184 struct netfront_info *info = dev_get_drvdata(&dev->dev); 2185 unsigned int num_queues = info->netdev->real_num_tx_queues; 2186 struct netfront_queue *queue = NULL; 2187 unsigned int i = 0; 2188 2189 dev_dbg(&dev->dev, "%s\n", dev->nodename); 2190 2191 xennet_disconnect_backend(info); 2192 2193 xennet_sysfs_delif(info->netdev); 2194 2195 unregister_netdev(info->netdev); 2196 2197 for (i = 0; i < num_queues; ++i) { 2198 queue = &info->queues[i]; 2199 del_timer_sync(&queue->rx_refill_timer); 2200 } 2201 2202 if (num_queues) { 2203 kfree(info->queues); 2204 info->queues = NULL; 2205 } 2206 2207 xennet_free_netdev(info->netdev); 2208 2209 return 0; 2210 } 2211 2212 static const struct xenbus_device_id netfront_ids[] = { 2213 { "vif" }, 2214 { "" } 2215 }; 2216 2217 static struct xenbus_driver netfront_driver = { 2218 .ids = netfront_ids, 2219 .probe = netfront_probe, 2220 .remove = xennet_remove, 2221 .resume = netfront_resume, 2222 .otherend_changed = netback_changed, 2223 }; 2224 2225 static int __init netif_init(void) 2226 { 2227 if (!xen_domain()) 2228 return -ENODEV; 2229 2230 if (!xen_has_pv_nic_devices()) 2231 return -ENODEV; 2232 2233 pr_info("Initialising Xen virtual ethernet driver\n"); 2234 2235 /* Allow as many queues as there are CPUs, by default */ 2236 xennet_max_queues = num_online_cpus(); 2237 2238 return xenbus_register_frontend(&netfront_driver); 2239 } 2240 module_init(netif_init); 2241 2242 2243 static void __exit netif_exit(void) 2244 { 2245 xenbus_unregister_driver(&netfront_driver); 2246 } 2247 module_exit(netif_exit); 2248 2249 MODULE_DESCRIPTION("Xen virtual network device frontend"); 2250 MODULE_LICENSE("GPL"); 2251 MODULE_ALIAS("xen:vif"); 2252 MODULE_ALIAS("xennet"); 2253