drivers/net/xen-netback/netback.c

   1 /*
   2  * Back-end of the driver for virtual network devices. This portion of the
   3  * driver exports a 'unified' network-device interface that can be accessed
   4  * by any operating system that implements a compatible front end. A
   5  * reference front-end implementation can be found in:
   6  *  drivers/net/xen-netfront.c
   7  *
   8  * Copyright (c) 2002-2005, K A Fraser
   9  *
  10  * This program is free software; you can redistribute it and/or
  11  * modify it under the terms of the GNU General Public License version 2
  12  * as published by the Free Software Foundation; or, when distributed
  13  * separately from the Linux kernel or incorporated into other
  14  * software packages, subject to the following license:
  15  *
  16  * Permission is hereby granted, free of charge, to any person obtaining a copy
  17  * of this source file (the "Software"), to deal in the Software without
  18  * restriction, including without limitation the rights to use, copy, modify,
  19  * merge, publish, distribute, sublicense, and/or sell copies of the Software,
  20  * and to permit persons to whom the Software is furnished to do so, subject to
  21  * the following conditions:
  22  *
  23  * The above copyright notice and this permission notice shall be included in
  24  * all copies or substantial portions of the Software.
  25  *
  26  * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
  27  * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
  28  * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
  29  * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
  30  * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
  31  * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
  32  * IN THE SOFTWARE.
  33  */
  34
  35 #include "common.h"
  36
  37 #include <linux/kthread.h>
  38 #include <linux/if_vlan.h>
  39 #include <linux/udp.h>
  40 #include <linux/highmem.h>
  41
  42 #include <net/tcp.h>
  43
  44 #include <xen/xen.h>
  45 #include <xen/events.h>
  46 #include <xen/interface/memory.h>
  47 #include <xen/page.h>
  48
  49 #include <asm/xen/hypercall.h>
  50
  51 /* Provide an option to disable split event channels at load time as
  52  * event channels are limited resource. Split event channels are
  53  * enabled by default.
  54  */
  55 bool separate_tx_rx_irq = true;
  56 module_param(separate_tx_rx_irq, bool, 0644);
  57
  58 /* The time that packets can stay on the guest Rx internal queue
  59  * before they are dropped.
  60  */
  61 unsigned int rx_drain_timeout_msecs = 10000;
  62 module_param(rx_drain_timeout_msecs, uint, 0444);
  63
  64 /* The length of time before the frontend is considered unresponsive
  65  * because it isn't providing Rx slots.
  66  */
  67 unsigned int rx_stall_timeout_msecs = 60000;
  68 module_param(rx_stall_timeout_msecs, uint, 0444);
  69
  70 #define MAX_QUEUES_DEFAULT 8
  71 unsigned int xenvif_max_queues;
  72 module_param_named(max_queues, xenvif_max_queues, uint, 0644);
  73 MODULE_PARM_DESC(max_queues,
  74                  "Maximum number of queues per virtual interface");
  75
  76 /*
  77  * This is the maximum slots a skb can have. If a guest sends a skb
  78  * which exceeds this limit it is considered malicious.
  79  */
  80 #define FATAL_SKB_SLOTS_DEFAULT 20
  81 static unsigned int fatal_skb_slots = FATAL_SKB_SLOTS_DEFAULT;
  82 module_param(fatal_skb_slots, uint, 0444);
  83
  84 /* The amount to copy out of the first guest Tx slot into the skb's
  85  * linear area.  If the first slot has more data, it will be mapped
  86  * and put into the first frag.
  87  *
  88  * This is sized to avoid pulling headers from the frags for most
  89  * TCP/IP packets.
  90  */
  91 #define XEN_NETBACK_TX_COPY_LEN 128
  92
  93 /* This is the maximum number of flows in the hash cache. */
  94 #define XENVIF_HASH_CACHE_SIZE_DEFAULT 64
  95 unsigned int xenvif_hash_cache_size = XENVIF_HASH_CACHE_SIZE_DEFAULT;
  96 module_param_named(hash_cache_size, xenvif_hash_cache_size, uint, 0644);
  97 MODULE_PARM_DESC(hash_cache_size, "Number of flows in the hash cache");
  98
  99 /* The module parameter tells that we have to put data
 100  * for xen-netfront with the XDP_PACKET_HEADROOM offset
 101  * needed for XDP processing
 102  */
 103 bool provides_xdp_headroom = true;
 104 module_param(provides_xdp_headroom, bool, 0644);
 105
 106 static void xenvif_idx_release(struct xenvif_queue *queue, u16 pending_idx,
 107                                u8 status);
 108
 109 static void make_tx_response(struct xenvif_queue *queue,
 110                              struct xen_netif_tx_request *txp,
 111                              unsigned int extra_count,
 112                              s8       st);
 113 static void push_tx_responses(struct xenvif_queue *queue);
 114
 115 static void xenvif_idx_unmap(struct xenvif_queue *queue, u16 pending_idx);
 116
 117 static inline int tx_work_todo(struct xenvif_queue *queue);
 118
 119 static inline unsigned long idx_to_pfn(struct xenvif_queue *queue,
 120                                        u16 idx)
 121 {
 122         return page_to_pfn(queue->mmap_pages[idx]);
 123 }
 124
 125 static inline unsigned long idx_to_kaddr(struct xenvif_queue *queue,
 126                                          u16 idx)
 127 {
 128         return (unsigned long)pfn_to_kaddr(idx_to_pfn(queue, idx));
 129 }
 130
 131 #define callback_param(vif, pending_idx) \
 132         (vif->pending_tx_info[pending_idx].callback_struct)
 133
 134 /* Find the containing VIF's structure from a pointer in pending_tx_info array
 135  */
 136 static inline struct xenvif_queue *ubuf_to_queue(const struct ubuf_info *ubuf)
 137 {
 138         u16 pending_idx = ubuf->desc;
 139         struct pending_tx_info *temp =
 140                 container_of(ubuf, struct pending_tx_info, callback_struct);
 141         return container_of(temp - pending_idx,
 142                             struct xenvif_queue,
 143                             pending_tx_info[0]);
 144 }
 145
 146 static u16 frag_get_pending_idx(skb_frag_t *frag)
 147 {
 148         return (u16)skb_frag_off(frag);
 149 }
 150
 151 static void frag_set_pending_idx(skb_frag_t *frag, u16 pending_idx)
 152 {
 153         skb_frag_off_set(frag, pending_idx);
 154 }
 155
 156 static inline pending_ring_idx_t pending_index(unsigned i)
 157 {
 158         return i & (MAX_PENDING_REQS-1);
 159 }
 160
 161 void xenvif_kick_thread(struct xenvif_queue *queue)
 162 {
 163         wake_up(&queue->wq);
 164 }
 165
 166 void xenvif_napi_schedule_or_enable_events(struct xenvif_queue *queue)
 167 {
 168         int more_to_do;
 169
 170         RING_FINAL_CHECK_FOR_REQUESTS(&queue->tx, more_to_do);
 171
 172         if (more_to_do)
 173                 napi_schedule(&queue->napi);
 174         else if (atomic_fetch_andnot(NETBK_TX_EOI | NETBK_COMMON_EOI,
 175                                      &queue->eoi_pending) &
 176                  (NETBK_TX_EOI | NETBK_COMMON_EOI))
 177                 xen_irq_lateeoi(queue->tx_irq, 0);
 178 }
 179
 180 static void tx_add_credit(struct xenvif_queue *queue)
 181 {
 182         unsigned long max_burst, max_credit;
 183
 184         /*
 185          * Allow a burst big enough to transmit a jumbo packet of up to 128kB.
 186          * Otherwise the interface can seize up due to insufficient credit.
 187          */
 188         max_burst = max(131072UL, queue->credit_bytes);
 189
 190         /* Take care that adding a new chunk of credit doesn't wrap to zero. */
 191         max_credit = queue->remaining_credit + queue->credit_bytes;
 192         if (max_credit < queue->remaining_credit)
 193                 max_credit = ULONG_MAX; /* wrapped: clamp to ULONG_MAX */
 194
 195         queue->remaining_credit = min(max_credit, max_burst);
 196         queue->rate_limited = false;
 197 }
 198
 199 void xenvif_tx_credit_callback(struct timer_list *t)
 200 {
 201         struct xenvif_queue *queue = from_timer(queue, t, credit_timeout);
 202         tx_add_credit(queue);
 203         xenvif_napi_schedule_or_enable_events(queue);
 204 }
 205
 206 static void xenvif_tx_err(struct xenvif_queue *queue,
 207                           struct xen_netif_tx_request *txp,
 208                           unsigned int extra_count, RING_IDX end)
 209 {
 210         RING_IDX cons = queue->tx.req_cons;
 211         unsigned long flags;
 212
 213         do {
 214                 spin_lock_irqsave(&queue->response_lock, flags);
 215                 make_tx_response(queue, txp, extra_count, XEN_NETIF_RSP_ERROR);
 216                 push_tx_responses(queue);
 217                 spin_unlock_irqrestore(&queue->response_lock, flags);
 218                 if (cons == end)
 219                         break;
 220                 RING_COPY_REQUEST(&queue->tx, cons++, txp);
 221                 extra_count = 0; /* only the first frag can have extras */
 222         } while (1);
 223         queue->tx.req_cons = cons;
 224 }
 225
 226 static void xenvif_fatal_tx_err(struct xenvif *vif)
 227 {
 228         netdev_err(vif->dev, "fatal error; disabling device\n");
 229         vif->disabled = true;
 230         /* Disable the vif from queue 0's kthread */
 231         if (vif->num_queues)
 232                 xenvif_kick_thread(&vif->queues[0]);
 233 }
 234
 235 static int xenvif_count_requests(struct xenvif_queue *queue,
 236                                  struct xen_netif_tx_request *first,
 237                                  unsigned int extra_count,
 238                                  struct xen_netif_tx_request *txp,
 239                                  int work_to_do)
 240 {
 241         RING_IDX cons = queue->tx.req_cons;
 242         int slots = 0;
 243         int drop_err = 0;
 244         int more_data;
 245
 246         if (!(first->flags & XEN_NETTXF_more_data))
 247                 return 0;
 248
 249         do {
 250                 struct xen_netif_tx_request dropped_tx = { 0 };
 251
 252                 if (slots >= work_to_do) {
 253                         netdev_err(queue->vif->dev,
 254                                    "Asked for %d slots but exceeds this limit\n",
 255                                    work_to_do);
 256                         xenvif_fatal_tx_err(queue->vif);
 257                         return -ENODATA;
 258                 }
 259
 260                 /* This guest is really using too many slots and
 261                  * considered malicious.
 262                  */
 263                 if (unlikely(slots >= fatal_skb_slots)) {
 264                         netdev_err(queue->vif->dev,
 265                                    "Malicious frontend using %d slots, threshold %u\n",
 266                                    slots, fatal_skb_slots);
 267                         xenvif_fatal_tx_err(queue->vif);
 268                         return -E2BIG;
 269                 }
 270
 271                 /* Xen network protocol had implicit dependency on
 272                  * MAX_SKB_FRAGS. XEN_NETBK_LEGACY_SLOTS_MAX is set to
 273                  * the historical MAX_SKB_FRAGS value 18 to honor the
 274                  * same behavior as before. Any packet using more than
 275                  * 18 slots but less than fatal_skb_slots slots is
 276                  * dropped
 277                  */
 278                 if (!drop_err && slots >= XEN_NETBK_LEGACY_SLOTS_MAX) {
 279                         if (net_ratelimit())
 280                                 netdev_dbg(queue->vif->dev,
 281                                            "Too many slots (%d) exceeding limit (%d), dropping packet\n",
 282                                            slots, XEN_NETBK_LEGACY_SLOTS_MAX);
 283                         drop_err = -E2BIG;
 284                 }
 285
 286                 if (drop_err)
 287                         txp = &dropped_tx;
 288
 289                 RING_COPY_REQUEST(&queue->tx, cons + slots, txp);
 290
 291                 /* If the guest submitted a frame >= 64 KiB then
 292                  * first->size overflowed and following slots will
 293                  * appear to be larger than the frame.
 294                  *
 295                  * This cannot be fatal error as there are buggy
 296                  * frontends that do this.
 297                  *
 298                  * Consume all slots and drop the packet.
 299                  */
 300                 if (!drop_err && txp->size > first->size) {
 301                         if (net_ratelimit())
 302                                 netdev_dbg(queue->vif->dev,
 303                                            "Invalid tx request, slot size %u > remaining size %u\n",
 304                                            txp->size, first->size);
 305                         drop_err = -EIO;
 306                 }
 307
 308                 first->size -= txp->size;
 309                 slots++;
 310
 311                 if (unlikely((txp->offset + txp->size) > XEN_PAGE_SIZE)) {
 312                         netdev_err(queue->vif->dev, "Cross page boundary, txp->offset: %u, size: %u\n",
 313                                  txp->offset, txp->size);
 314                         xenvif_fatal_tx_err(queue->vif);
 315                         return -EINVAL;
 316                 }
 317
 318                 more_data = txp->flags & XEN_NETTXF_more_data;
 319
 320                 if (!drop_err)
 321                         txp++;
 322
 323         } while (more_data);
 324
 325         if (drop_err) {
 326                 xenvif_tx_err(queue, first, extra_count, cons + slots);
 327                 return drop_err;
 328         }
 329
 330         return slots;
 331 }
 332
 333
 334 struct xenvif_tx_cb {
 335         u16 pending_idx;
 336 };
 337
 338 #define XENVIF_TX_CB(skb) ((struct xenvif_tx_cb *)(skb)->cb)
 339
 340 static inline void xenvif_tx_create_map_op(struct xenvif_queue *queue,
 341                                            u16 pending_idx,
 342                                            struct xen_netif_tx_request *txp,
 343                                            unsigned int extra_count,
 344                                            struct gnttab_map_grant_ref *mop)
 345 {
 346         queue->pages_to_map[mop-queue->tx_map_ops] = queue->mmap_pages[pending_idx];
 347         gnttab_set_map_op(mop, idx_to_kaddr(queue, pending_idx),
 348                           GNTMAP_host_map | GNTMAP_readonly,
 349                           txp->gref, queue->vif->domid);
 350
 351         memcpy(&queue->pending_tx_info[pending_idx].req, txp,
 352                sizeof(*txp));
 353         queue->pending_tx_info[pending_idx].extra_count = extra_count;
 354 }
 355
 356 static inline struct sk_buff *xenvif_alloc_skb(unsigned int size)
 357 {
 358         struct sk_buff *skb =
 359                 alloc_skb(size + NET_SKB_PAD + NET_IP_ALIGN,
 360                           GFP_ATOMIC | __GFP_NOWARN);
 361         if (unlikely(skb == NULL))
 362                 return NULL;
 363
 364         /* Packets passed to netif_rx() must have some headroom. */
 365         skb_reserve(skb, NET_SKB_PAD + NET_IP_ALIGN);
 366
 367         /* Initialize it here to avoid later surprises */
 368         skb_shinfo(skb)->destructor_arg = NULL;
 369
 370         return skb;
 371 }
 372
 373 static struct gnttab_map_grant_ref *xenvif_get_requests(struct xenvif_queue *queue,
 374                                                         struct sk_buff *skb,
 375                                                         struct xen_netif_tx_request *txp,
 376                                                         struct gnttab_map_grant_ref *gop,
 377                                                         unsigned int frag_overflow,
 378                                                         struct sk_buff *nskb)
 379 {
 380         struct skb_shared_info *shinfo = skb_shinfo(skb);
 381         skb_frag_t *frags = shinfo->frags;
 382         u16 pending_idx = XENVIF_TX_CB(skb)->pending_idx;
 383         int start;
 384         pending_ring_idx_t index;
 385         unsigned int nr_slots;
 386
 387         nr_slots = shinfo->nr_frags;
 388
 389         /* Skip first skb fragment if it is on same page as header fragment. */
 390         start = (frag_get_pending_idx(&shinfo->frags[0]) == pending_idx);
 391
 392         for (shinfo->nr_frags = start; shinfo->nr_frags < nr_slots;
 393              shinfo->nr_frags++, txp++, gop++) {
 394                 index = pending_index(queue->pending_cons++);
 395                 pending_idx = queue->pending_ring[index];
 396                 xenvif_tx_create_map_op(queue, pending_idx, txp, 0, gop);
 397                 frag_set_pending_idx(&frags[shinfo->nr_frags], pending_idx);
 398         }
 399
 400         if (frag_overflow) {
 401
 402                 shinfo = skb_shinfo(nskb);
 403                 frags = shinfo->frags;
 404
 405                 for (shinfo->nr_frags = 0; shinfo->nr_frags < frag_overflow;
 406                      shinfo->nr_frags++, txp++, gop++) {
 407                         index = pending_index(queue->pending_cons++);
 408                         pending_idx = queue->pending_ring[index];
 409                         xenvif_tx_create_map_op(queue, pending_idx, txp, 0,
 410                                                 gop);
 411                         frag_set_pending_idx(&frags[shinfo->nr_frags],
 412                                              pending_idx);
 413                 }
 414
 415                 skb_shinfo(skb)->frag_list = nskb;
 416         }
 417
 418         return gop;
 419 }
 420
 421 static inline void xenvif_grant_handle_set(struct xenvif_queue *queue,
 422                                            u16 pending_idx,
 423                                            grant_handle_t handle)
 424 {
 425         if (unlikely(queue->grant_tx_handle[pending_idx] !=
 426                      NETBACK_INVALID_HANDLE)) {
 427                 netdev_err(queue->vif->dev,
 428                            "Trying to overwrite active handle! pending_idx: 0x%x\n",
 429                            pending_idx);
 430                 BUG();
 431         }
 432         queue->grant_tx_handle[pending_idx] = handle;
 433 }
 434
 435 static inline void xenvif_grant_handle_reset(struct xenvif_queue *queue,
 436                                              u16 pending_idx)
 437 {
 438         if (unlikely(queue->grant_tx_handle[pending_idx] ==
 439                      NETBACK_INVALID_HANDLE)) {
 440                 netdev_err(queue->vif->dev,
 441                            "Trying to unmap invalid handle! pending_idx: 0x%x\n",
 442                            pending_idx);
 443                 BUG();
 444         }
 445         queue->grant_tx_handle[pending_idx] = NETBACK_INVALID_HANDLE;
 446 }
 447
 448 static int xenvif_tx_check_gop(struct xenvif_queue *queue,
 449                                struct sk_buff *skb,
 450                                struct gnttab_map_grant_ref **gopp_map,
 451                                struct gnttab_copy **gopp_copy)
 452 {
 453         struct gnttab_map_grant_ref *gop_map = *gopp_map;
 454         u16 pending_idx = XENVIF_TX_CB(skb)->pending_idx;
 455         /* This always points to the shinfo of the skb being checked, which
 456          * could be either the first or the one on the frag_list
 457          */
 458         struct skb_shared_info *shinfo = skb_shinfo(skb);
 459         /* If this is non-NULL, we are currently checking the frag_list skb, and
 460          * this points to the shinfo of the first one
 461          */
 462         struct skb_shared_info *first_shinfo = NULL;
 463         int nr_frags = shinfo->nr_frags;
 464         const bool sharedslot = nr_frags &&
 465                                 frag_get_pending_idx(&shinfo->frags[0]) == pending_idx;
 466         int i, err;
 467
 468         /* Check status of header. */
 469         err = (*gopp_copy)->status;
 470         if (unlikely(err)) {
 471                 if (net_ratelimit())
 472                         netdev_dbg(queue->vif->dev,
 473                                    "Grant copy of header failed! status: %d pending_idx: %u ref: %u\n",
 474                                    (*gopp_copy)->status,
 475                                    pending_idx,
 476                                    (*gopp_copy)->source.u.ref);
 477                 /* The first frag might still have this slot mapped */
 478                 if (!sharedslot)
 479                         xenvif_idx_release(queue, pending_idx,
 480                                            XEN_NETIF_RSP_ERROR);
 481         }
 482         (*gopp_copy)++;
 483
 484 check_frags:
 485         for (i = 0; i < nr_frags; i++, gop_map++) {
 486                 int j, newerr;
 487
 488                 pending_idx = frag_get_pending_idx(&shinfo->frags[i]);
 489
 490                 /* Check error status: if okay then remember grant handle. */
 491                 newerr = gop_map->status;
 492
 493                 if (likely(!newerr)) {
 494                         xenvif_grant_handle_set(queue,
 495                                                 pending_idx,
 496                                                 gop_map->handle);
 497                         /* Had a previous error? Invalidate this fragment. */
 498                         if (unlikely(err)) {
 499                                 xenvif_idx_unmap(queue, pending_idx);
 500                                 /* If the mapping of the first frag was OK, but
 501                                  * the header's copy failed, and they are
 502                                  * sharing a slot, send an error
 503                                  */
 504                                 if (i == 0 && !first_shinfo && sharedslot)
 505                                         xenvif_idx_release(queue, pending_idx,
 506                                                            XEN_NETIF_RSP_ERROR);
 507                                 else
 508                                         xenvif_idx_release(queue, pending_idx,
 509                                                            XEN_NETIF_RSP_OKAY);
 510                         }
 511                         continue;
 512                 }
 513
 514                 /* Error on this fragment: respond to client with an error. */
 515                 if (net_ratelimit())
 516                         netdev_dbg(queue->vif->dev,
 517                                    "Grant map of %d. frag failed! status: %d pending_idx: %u ref: %u\n",
 518                                    i,
 519                                    gop_map->status,
 520                                    pending_idx,
 521                                    gop_map->ref);
 522
 523                 xenvif_idx_release(queue, pending_idx, XEN_NETIF_RSP_ERROR);
 524
 525                 /* Not the first error? Preceding frags already invalidated. */
 526                 if (err)
 527                         continue;
 528
 529                 /* First error: if the header haven't shared a slot with the
 530                  * first frag, release it as well.
 531                  */
 532                 if (!sharedslot)
 533                         xenvif_idx_release(queue,
 534                                            XENVIF_TX_CB(skb)->pending_idx,
 535                                            XEN_NETIF_RSP_OKAY);
 536
 537                 /* Invalidate preceding fragments of this skb. */
 538                 for (j = 0; j < i; j++) {
 539                         pending_idx = frag_get_pending_idx(&shinfo->frags[j]);
 540                         xenvif_idx_unmap(queue, pending_idx);
 541                         xenvif_idx_release(queue, pending_idx,
 542                                            XEN_NETIF_RSP_OKAY);
 543                 }
 544
 545                 /* And if we found the error while checking the frag_list, unmap
 546                  * the first skb's frags
 547                  */
 548                 if (first_shinfo) {
 549                         for (j = 0; j < first_shinfo->nr_frags; j++) {
 550                                 pending_idx = frag_get_pending_idx(&first_shinfo->frags[j]);
 551                                 xenvif_idx_unmap(queue, pending_idx);
 552                                 xenvif_idx_release(queue, pending_idx,
 553                                                    XEN_NETIF_RSP_OKAY);
 554                         }
 555                 }
 556
 557                 /* Remember the error: invalidate all subsequent fragments. */
 558                 err = newerr;
 559         }
 560
 561         if (skb_has_frag_list(skb) && !first_shinfo) {
 562                 first_shinfo = shinfo;
 563                 shinfo = skb_shinfo(shinfo->frag_list);
 564                 nr_frags = shinfo->nr_frags;
 565
 566                 goto check_frags;
 567         }
 568
 569         *gopp_map = gop_map;
 570         return err;
 571 }
 572
 573 static void xenvif_fill_frags(struct xenvif_queue *queue, struct sk_buff *skb)
 574 {
 575         struct skb_shared_info *shinfo = skb_shinfo(skb);
 576         int nr_frags = shinfo->nr_frags;
 577         int i;
 578         u16 prev_pending_idx = INVALID_PENDING_IDX;
 579
 580         for (i = 0; i < nr_frags; i++) {
 581                 skb_frag_t *frag = shinfo->frags + i;
 582                 struct xen_netif_tx_request *txp;
 583                 struct page *page;
 584                 u16 pending_idx;
 585
 586                 pending_idx = frag_get_pending_idx(frag);
 587
 588                 /* If this is not the first frag, chain it to the previous*/
 589                 if (prev_pending_idx == INVALID_PENDING_IDX)
 590                         skb_shinfo(skb)->destructor_arg =
 591                                 &callback_param(queue, pending_idx);
 592                 else
 593                         callback_param(queue, prev_pending_idx).ctx =
 594                                 &callback_param(queue, pending_idx);
 595
 596                 callback_param(queue, pending_idx).ctx = NULL;
 597                 prev_pending_idx = pending_idx;
 598
 599                 txp = &queue->pending_tx_info[pending_idx].req;
 600                 page = virt_to_page(idx_to_kaddr(queue, pending_idx));
 601                 __skb_fill_page_desc(skb, i, page, txp->offset, txp->size);
 602                 skb->len += txp->size;
 603                 skb->data_len += txp->size;
 604                 skb->truesize += txp->size;
 605
 606                 /* Take an extra reference to offset network stack's put_page */
 607                 get_page(queue->mmap_pages[pending_idx]);
 608         }
 609 }
 610
 611 static int xenvif_get_extras(struct xenvif_queue *queue,
 612                              struct xen_netif_extra_info *extras,
 613                              unsigned int *extra_count,
 614                              int work_to_do)
 615 {
 616         struct xen_netif_extra_info extra;
 617         RING_IDX cons = queue->tx.req_cons;
 618
 619         do {
 620                 if (unlikely(work_to_do-- <= 0)) {
 621                         netdev_err(queue->vif->dev, "Missing extra info\n");
 622                         xenvif_fatal_tx_err(queue->vif);
 623                         return -EBADR;
 624                 }
 625
 626                 RING_COPY_REQUEST(&queue->tx, cons, &extra);
 627
 628                 queue->tx.req_cons = ++cons;
 629                 (*extra_count)++;
 630
 631                 if (unlikely(!extra.type ||
 632                              extra.type >= XEN_NETIF_EXTRA_TYPE_MAX)) {
 633                         netdev_err(queue->vif->dev,
 634                                    "Invalid extra type: %d\n", extra.type);
 635                         xenvif_fatal_tx_err(queue->vif);
 636                         return -EINVAL;
 637                 }
 638
 639                 memcpy(&extras[extra.type - 1], &extra, sizeof(extra));
 640         } while (extra.flags & XEN_NETIF_EXTRA_FLAG_MORE);
 641
 642         return work_to_do;
 643 }
 644
 645 static int xenvif_set_skb_gso(struct xenvif *vif,
 646                               struct sk_buff *skb,
 647                               struct xen_netif_extra_info *gso)
 648 {
 649         if (!gso->u.gso.size) {
 650                 netdev_err(vif->dev, "GSO size must not be zero.\n");
 651                 xenvif_fatal_tx_err(vif);
 652                 return -EINVAL;
 653         }
 654
 655         switch (gso->u.gso.type) {
 656         case XEN_NETIF_GSO_TYPE_TCPV4:
 657                 skb_shinfo(skb)->gso_type = SKB_GSO_TCPV4;
 658                 break;
 659         case XEN_NETIF_GSO_TYPE_TCPV6:
 660                 skb_shinfo(skb)->gso_type = SKB_GSO_TCPV6;
 661                 break;
 662         default:
 663                 netdev_err(vif->dev, "Bad GSO type %d.\n", gso->u.gso.type);
 664                 xenvif_fatal_tx_err(vif);
 665                 return -EINVAL;
 666         }
 667
 668         skb_shinfo(skb)->gso_size = gso->u.gso.size;
 669         /* gso_segs will be calculated later */
 670
 671         return 0;
 672 }
 673
 674 static int checksum_setup(struct xenvif_queue *queue, struct sk_buff *skb)
 675 {
 676         bool recalculate_partial_csum = false;
 677
 678         /* A GSO SKB must be CHECKSUM_PARTIAL. However some buggy
 679          * peers can fail to set NETRXF_csum_blank when sending a GSO
 680          * frame. In this case force the SKB to CHECKSUM_PARTIAL and
 681          * recalculate the partial checksum.
 682          */
 683         if (skb->ip_summed != CHECKSUM_PARTIAL && skb_is_gso(skb)) {
 684                 queue->stats.rx_gso_checksum_fixup++;
 685                 skb->ip_summed = CHECKSUM_PARTIAL;
 686                 recalculate_partial_csum = true;
 687         }
 688
 689         /* A non-CHECKSUM_PARTIAL SKB does not require setup. */
 690         if (skb->ip_summed != CHECKSUM_PARTIAL)
 691                 return 0;
 692
 693         return skb_checksum_setup(skb, recalculate_partial_csum);
 694 }
 695
 696 static bool tx_credit_exceeded(struct xenvif_queue *queue, unsigned size)
 697 {
 698         u64 now = get_jiffies_64();
 699         u64 next_credit = queue->credit_window_start +
 700                 msecs_to_jiffies(queue->credit_usec / 1000);
 701
 702         /* Timer could already be pending in rare cases. */
 703         if (timer_pending(&queue->credit_timeout)) {
 704                 queue->rate_limited = true;
 705                 return true;
 706         }
 707
 708         /* Passed the point where we can replenish credit? */
 709         if (time_after_eq64(now, next_credit)) {
 710                 queue->credit_window_start = now;
 711                 tx_add_credit(queue);
 712         }
 713
 714         /* Still too big to send right now? Set a callback. */
 715         if (size > queue->remaining_credit) {
 716                 mod_timer(&queue->credit_timeout,
 717                           next_credit);
 718                 queue->credit_window_start = next_credit;
 719                 queue->rate_limited = true;
 720
 721                 return true;
 722         }
 723
 724         return false;
 725 }
 726
 727 /* No locking is required in xenvif_mcast_add/del() as they are
 728  * only ever invoked from NAPI poll. An RCU list is used because
 729  * xenvif_mcast_match() is called asynchronously, during start_xmit.
 730  */
 731
 732 static int xenvif_mcast_add(struct xenvif *vif, const u8 *addr)
 733 {
 734         struct xenvif_mcast_addr *mcast;
 735
 736         if (vif->fe_mcast_count == XEN_NETBK_MCAST_MAX) {
 737                 if (net_ratelimit())
 738                         netdev_err(vif->dev,
 739                                    "Too many multicast addresses\n");
 740                 return -ENOSPC;
 741         }
 742
 743         mcast = kzalloc(sizeof(*mcast), GFP_ATOMIC);
 744         if (!mcast)
 745                 return -ENOMEM;
 746
 747         ether_addr_copy(mcast->addr, addr);
 748         list_add_tail_rcu(&mcast->entry, &vif->fe_mcast_addr);
 749         vif->fe_mcast_count++;
 750
 751         return 0;
 752 }
 753
 754 static void xenvif_mcast_del(struct xenvif *vif, const u8 *addr)
 755 {
 756         struct xenvif_mcast_addr *mcast;
 757
 758         list_for_each_entry_rcu(mcast, &vif->fe_mcast_addr, entry) {
 759                 if (ether_addr_equal(addr, mcast->addr)) {
 760                         --vif->fe_mcast_count;
 761                         list_del_rcu(&mcast->entry);
 762                         kfree_rcu(mcast, rcu);
 763                         break;
 764                 }
 765         }
 766 }
 767
 768 bool xenvif_mcast_match(struct xenvif *vif, const u8 *addr)
 769 {
 770         struct xenvif_mcast_addr *mcast;
 771
 772         rcu_read_lock();
 773         list_for_each_entry_rcu(mcast, &vif->fe_mcast_addr, entry) {
 774                 if (ether_addr_equal(addr, mcast->addr)) {
 775                         rcu_read_unlock();
 776                         return true;
 777                 }
 778         }
 779         rcu_read_unlock();
 780
 781         return false;
 782 }
 783
 784 void xenvif_mcast_addr_list_free(struct xenvif *vif)
 785 {
 786         /* No need for locking or RCU here. NAPI poll and TX queue
 787          * are stopped.
 788          */
 789         while (!list_empty(&vif->fe_mcast_addr)) {
 790                 struct xenvif_mcast_addr *mcast;
 791
 792                 mcast = list_first_entry(&vif->fe_mcast_addr,
 793                                          struct xenvif_mcast_addr,
 794                                          entry);
 795                 --vif->fe_mcast_count;
 796                 list_del(&mcast->entry);
 797                 kfree(mcast);
 798         }
 799 }
 800
 801 static void xenvif_tx_build_gops(struct xenvif_queue *queue,
 802                                      int budget,
 803                                      unsigned *copy_ops,
 804                                      unsigned *map_ops)
 805 {
 806         struct gnttab_map_grant_ref *gop = queue->tx_map_ops;
 807         struct sk_buff *skb, *nskb;
 808         int ret;
 809         unsigned int frag_overflow;
 810
 811         while (skb_queue_len(&queue->tx_queue) < budget) {
 812                 struct xen_netif_tx_request txreq;
 813                 struct xen_netif_tx_request txfrags[XEN_NETBK_LEGACY_SLOTS_MAX];
 814                 struct xen_netif_extra_info extras[XEN_NETIF_EXTRA_TYPE_MAX-1];
 815                 unsigned int extra_count;
 816                 u16 pending_idx;
 817                 RING_IDX idx;
 818                 int work_to_do;
 819                 unsigned int data_len;
 820                 pending_ring_idx_t index;
 821
 822                 if (queue->tx.sring->req_prod - queue->tx.req_cons >
 823                     XEN_NETIF_TX_RING_SIZE) {
 824                         netdev_err(queue->vif->dev,
 825                                    "Impossible number of requests. "
 826                                    "req_prod %d, req_cons %d, size %ld\n",
 827                                    queue->tx.sring->req_prod, queue->tx.req_cons,
 828                                    XEN_NETIF_TX_RING_SIZE);
 829                         xenvif_fatal_tx_err(queue->vif);
 830                         break;
 831                 }
 832
 833                 work_to_do = XEN_RING_NR_UNCONSUMED_REQUESTS(&queue->tx);
 834                 if (!work_to_do)
 835                         break;
 836
 837                 idx = queue->tx.req_cons;
 838                 rmb(); /* Ensure that we see the request before we copy it. */
 839                 RING_COPY_REQUEST(&queue->tx, idx, &txreq);
 840
 841                 /* Credit-based scheduling. */
 842                 if (txreq.size > queue->remaining_credit &&
 843                     tx_credit_exceeded(queue, txreq.size))
 844                         break;
 845
 846                 queue->remaining_credit -= txreq.size;
 847
 848                 work_to_do--;
 849                 queue->tx.req_cons = ++idx;
 850
 851                 memset(extras, 0, sizeof(extras));
 852                 extra_count = 0;
 853                 if (txreq.flags & XEN_NETTXF_extra_info) {
 854                         work_to_do = xenvif_get_extras(queue, extras,
 855                                                        &extra_count,
 856                                                        work_to_do);
 857                         idx = queue->tx.req_cons;
 858                         if (unlikely(work_to_do < 0))
 859                                 break;
 860                 }
 861
 862                 if (extras[XEN_NETIF_EXTRA_TYPE_MCAST_ADD - 1].type) {
 863                         struct xen_netif_extra_info *extra;
 864
 865                         extra = &extras[XEN_NETIF_EXTRA_TYPE_MCAST_ADD - 1];
 866                         ret = xenvif_mcast_add(queue->vif, extra->u.mcast.addr);
 867
 868                         make_tx_response(queue, &txreq, extra_count,
 869                                          (ret == 0) ?
 870                                          XEN_NETIF_RSP_OKAY :
 871                                          XEN_NETIF_RSP_ERROR);
 872                         push_tx_responses(queue);
 873                         continue;
 874                 }
 875
 876                 if (extras[XEN_NETIF_EXTRA_TYPE_MCAST_DEL - 1].type) {
 877                         struct xen_netif_extra_info *extra;
 878
 879                         extra = &extras[XEN_NETIF_EXTRA_TYPE_MCAST_DEL - 1];
 880                         xenvif_mcast_del(queue->vif, extra->u.mcast.addr);
 881
 882                         make_tx_response(queue, &txreq, extra_count,
 883                                          XEN_NETIF_RSP_OKAY);
 884                         push_tx_responses(queue);
 885                         continue;
 886                 }
 887
 888                 ret = xenvif_count_requests(queue, &txreq, extra_count,
 889                                             txfrags, work_to_do);
 890                 if (unlikely(ret < 0))
 891                         break;
 892
 893                 idx += ret;
 894
 895                 if (unlikely(txreq.size < ETH_HLEN)) {
 896                         netdev_dbg(queue->vif->dev,
 897                                    "Bad packet size: %d\n", txreq.size);
 898                         xenvif_tx_err(queue, &txreq, extra_count, idx);
 899                         break;
 900                 }
 901
 902                 /* No crossing a page as the payload mustn't fragment. */
 903                 if (unlikely((txreq.offset + txreq.size) > XEN_PAGE_SIZE)) {
 904                         netdev_err(queue->vif->dev,
 905                                    "txreq.offset: %u, size: %u, end: %lu\n",
 906                                    txreq.offset, txreq.size,
 907                                    (unsigned long)(txreq.offset&~XEN_PAGE_MASK) + txreq.size);
 908                         xenvif_fatal_tx_err(queue->vif);
 909                         break;
 910                 }
 911
 912                 index = pending_index(queue->pending_cons);
 913                 pending_idx = queue->pending_ring[index];
 914
 915                 data_len = (txreq.size > XEN_NETBACK_TX_COPY_LEN &&
 916                             ret < XEN_NETBK_LEGACY_SLOTS_MAX) ?
 917                         XEN_NETBACK_TX_COPY_LEN : txreq.size;
 918
 919                 skb = xenvif_alloc_skb(data_len);
 920                 if (unlikely(skb == NULL)) {
 921                         netdev_dbg(queue->vif->dev,
 922                                    "Can't allocate a skb in start_xmit.\n");
 923                         xenvif_tx_err(queue, &txreq, extra_count, idx);
 924                         break;
 925                 }
 926
 927                 skb_shinfo(skb)->nr_frags = ret;
 928                 if (data_len < txreq.size)
 929                         skb_shinfo(skb)->nr_frags++;
 930                 /* At this point shinfo->nr_frags is in fact the number of
 931                  * slots, which can be as large as XEN_NETBK_LEGACY_SLOTS_MAX.
 932                  */
 933                 frag_overflow = 0;
 934                 nskb = NULL;
 935                 if (skb_shinfo(skb)->nr_frags > MAX_SKB_FRAGS) {
 936                         frag_overflow = skb_shinfo(skb)->nr_frags - MAX_SKB_FRAGS;
 937                         BUG_ON(frag_overflow > MAX_SKB_FRAGS);
 938                         skb_shinfo(skb)->nr_frags = MAX_SKB_FRAGS;
 939                         nskb = xenvif_alloc_skb(0);
 940                         if (unlikely(nskb == NULL)) {
 941                                 skb_shinfo(skb)->nr_frags = 0;
 942                                 kfree_skb(skb);
 943                                 xenvif_tx_err(queue, &txreq, extra_count, idx);
 944                                 if (net_ratelimit())
 945                                         netdev_err(queue->vif->dev,
 946                                                    "Can't allocate the frag_list skb.\n");
 947                                 break;
 948                         }
 949                 }
 950
 951                 if (extras[XEN_NETIF_EXTRA_TYPE_GSO - 1].type) {
 952                         struct xen_netif_extra_info *gso;
 953                         gso = &extras[XEN_NETIF_EXTRA_TYPE_GSO - 1];
 954
 955                         if (xenvif_set_skb_gso(queue->vif, skb, gso)) {
 956                                 /* Failure in xenvif_set_skb_gso is fatal. */
 957                                 skb_shinfo(skb)->nr_frags = 0;
 958                                 kfree_skb(skb);
 959                                 kfree_skb(nskb);
 960                                 break;
 961                         }
 962                 }
 963
 964                 if (extras[XEN_NETIF_EXTRA_TYPE_HASH - 1].type) {
 965                         struct xen_netif_extra_info *extra;
 966                         enum pkt_hash_types type = PKT_HASH_TYPE_NONE;
 967
 968                         extra = &extras[XEN_NETIF_EXTRA_TYPE_HASH - 1];
 969
 970                         switch (extra->u.hash.type) {
 971                         case _XEN_NETIF_CTRL_HASH_TYPE_IPV4:
 972                         case _XEN_NETIF_CTRL_HASH_TYPE_IPV6:
 973                                 type = PKT_HASH_TYPE_L3;
 974                                 break;
 975
 976                         case _XEN_NETIF_CTRL_HASH_TYPE_IPV4_TCP:
 977                         case _XEN_NETIF_CTRL_HASH_TYPE_IPV6_TCP:
 978                                 type = PKT_HASH_TYPE_L4;
 979                                 break;
 980
 981                         default:
 982                                 break;
 983                         }
 984
 985                         if (type != PKT_HASH_TYPE_NONE)
 986                                 skb_set_hash(skb,
 987                                              *(u32 *)extra->u.hash.value,
 988                                              type);
 989                 }
 990
 991                 XENVIF_TX_CB(skb)->pending_idx = pending_idx;
 992
 993                 __skb_put(skb, data_len);
 994                 queue->tx_copy_ops[*copy_ops].source.u.ref = txreq.gref;
 995                 queue->tx_copy_ops[*copy_ops].source.domid = queue->vif->domid;
 996                 queue->tx_copy_ops[*copy_ops].source.offset = txreq.offset;
 997
 998                 queue->tx_copy_ops[*copy_ops].dest.u.gmfn =
 999                         virt_to_gfn(skb->data);
1000                 queue->tx_copy_ops[*copy_ops].dest.domid = DOMID_SELF;
1001                 queue->tx_copy_ops[*copy_ops].dest.offset =
1002                         offset_in_page(skb->data) & ~XEN_PAGE_MASK;
1003
1004                 queue->tx_copy_ops[*copy_ops].len = data_len;
1005                 queue->tx_copy_ops[*copy_ops].flags = GNTCOPY_source_gref;
1006
1007                 (*copy_ops)++;
1008
1009                 if (data_len < txreq.size) {
1010                         frag_set_pending_idx(&skb_shinfo(skb)->frags[0],
1011                                              pending_idx);
1012                         xenvif_tx_create_map_op(queue, pending_idx, &txreq,
1013                                                 extra_count, gop);
1014                         gop++;
1015                 } else {
1016                         frag_set_pending_idx(&skb_shinfo(skb)->frags[0],
1017                                              INVALID_PENDING_IDX);
1018                         memcpy(&queue->pending_tx_info[pending_idx].req,
1019                                &txreq, sizeof(txreq));
1020                         queue->pending_tx_info[pending_idx].extra_count =
1021                                 extra_count;
1022                 }
1023
1024                 queue->pending_cons++;
1025
1026                 gop = xenvif_get_requests(queue, skb, txfrags, gop,
1027                                           frag_overflow, nskb);
1028
1029                 __skb_queue_tail(&queue->tx_queue, skb);
1030
1031                 queue->tx.req_cons = idx;
1032
1033                 if (((gop-queue->tx_map_ops) >= ARRAY_SIZE(queue->tx_map_ops)) ||
1034                     (*copy_ops >= ARRAY_SIZE(queue->tx_copy_ops)))
1035                         break;
1036         }
1037
1038         (*map_ops) = gop - queue->tx_map_ops;
1039         return;
1040 }
1041
1042 /* Consolidate skb with a frag_list into a brand new one with local pages on
1043  * frags. Returns 0 or -ENOMEM if can't allocate new pages.
1044  */
1045 static int xenvif_handle_frag_list(struct xenvif_queue *queue, struct sk_buff *skb)
1046 {
1047         unsigned int offset = skb_headlen(skb);
1048         skb_frag_t frags[MAX_SKB_FRAGS];
1049         int i, f;
1050         struct ubuf_info *uarg;
1051         struct sk_buff *nskb = skb_shinfo(skb)->frag_list;
1052
1053         queue->stats.tx_zerocopy_sent += 2;
1054         queue->stats.tx_frag_overflow++;
1055
1056         xenvif_fill_frags(queue, nskb);
1057         /* Subtract frags size, we will correct it later */
1058         skb->truesize -= skb->data_len;
1059         skb->len += nskb->len;
1060         skb->data_len += nskb->len;
1061
1062         /* create a brand new frags array and coalesce there */
1063         for (i = 0; offset < skb->len; i++) {
1064                 struct page *page;
1065                 unsigned int len;
1066
1067                 BUG_ON(i >= MAX_SKB_FRAGS);
1068                 page = alloc_page(GFP_ATOMIC);
1069                 if (!page) {
1070                         int j;
1071                         skb->truesize += skb->data_len;
1072                         for (j = 0; j < i; j++)
1073                                 put_page(skb_frag_page(&frags[j]));
1074                         return -ENOMEM;
1075                 }
1076
1077                 if (offset + PAGE_SIZE < skb->len)
1078                         len = PAGE_SIZE;
1079                 else
1080                         len = skb->len - offset;
1081                 if (skb_copy_bits(skb, offset, page_address(page), len))
1082                         BUG();
1083
1084                 offset += len;
1085                 __skb_frag_set_page(&frags[i], page);
1086                 skb_frag_off_set(&frags[i], 0);
1087                 skb_frag_size_set(&frags[i], len);
1088         }
1089
1090         /* Release all the original (foreign) frags. */
1091         for (f = 0; f < skb_shinfo(skb)->nr_frags; f++)
1092                 skb_frag_unref(skb, f);
1093         uarg = skb_shinfo(skb)->destructor_arg;
1094         /* increase inflight counter to offset decrement in callback */
1095         atomic_inc(&queue->inflight_packets);
1096         uarg->callback(NULL, uarg, true);
1097         skb_shinfo(skb)->destructor_arg = NULL;
1098
1099         /* Fill the skb with the new (local) frags. */
1100         memcpy(skb_shinfo(skb)->frags, frags, i * sizeof(skb_frag_t));
1101         skb_shinfo(skb)->nr_frags = i;
1102         skb->truesize += i * PAGE_SIZE;
1103
1104         return 0;
1105 }
1106
1107 static int xenvif_tx_submit(struct xenvif_queue *queue)
1108 {
1109         struct gnttab_map_grant_ref *gop_map = queue->tx_map_ops;
1110         struct gnttab_copy *gop_copy = queue->tx_copy_ops;
1111         struct sk_buff *skb;
1112         int work_done = 0;
1113
1114         while ((skb = __skb_dequeue(&queue->tx_queue)) != NULL) {
1115                 struct xen_netif_tx_request *txp;
1116                 u16 pending_idx;
1117                 unsigned data_len;
1118
1119                 pending_idx = XENVIF_TX_CB(skb)->pending_idx;
1120                 txp = &queue->pending_tx_info[pending_idx].req;
1121
1122                 /* Check the remap error code. */
1123                 if (unlikely(xenvif_tx_check_gop(queue, skb, &gop_map, &gop_copy))) {
1124                         /* If there was an error, xenvif_tx_check_gop is
1125                          * expected to release all the frags which were mapped,
1126                          * so kfree_skb shouldn't do it again
1127                          */
1128                         skb_shinfo(skb)->nr_frags = 0;
1129                         if (skb_has_frag_list(skb)) {
1130                                 struct sk_buff *nskb =
1131                                                 skb_shinfo(skb)->frag_list;
1132                                 skb_shinfo(nskb)->nr_frags = 0;
1133                         }
1134                         kfree_skb(skb);
1135                         continue;
1136                 }
1137
1138                 data_len = skb->len;
1139                 callback_param(queue, pending_idx).ctx = NULL;
1140                 if (data_len < txp->size) {
1141                         /* Append the packet payload as a fragment. */
1142                         txp->offset += data_len;
1143                         txp->size -= data_len;
1144                 } else {
1145                         /* Schedule a response immediately. */
1146                         xenvif_idx_release(queue, pending_idx,
1147                                            XEN_NETIF_RSP_OKAY);
1148                 }
1149
1150                 if (txp->flags & XEN_NETTXF_csum_blank)
1151                         skb->ip_summed = CHECKSUM_PARTIAL;
1152                 else if (txp->flags & XEN_NETTXF_data_validated)
1153                         skb->ip_summed = CHECKSUM_UNNECESSARY;
1154
1155                 xenvif_fill_frags(queue, skb);
1156
1157                 if (unlikely(skb_has_frag_list(skb))) {
1158                         struct sk_buff *nskb = skb_shinfo(skb)->frag_list;
1159                         xenvif_skb_zerocopy_prepare(queue, nskb);
1160                         if (xenvif_handle_frag_list(queue, skb)) {
1161                                 if (net_ratelimit())
1162                                         netdev_err(queue->vif->dev,
1163                                                    "Not enough memory to consolidate frag_list!\n");
1164                                 xenvif_skb_zerocopy_prepare(queue, skb);
1165                                 kfree_skb(skb);
1166                                 continue;
1167                         }
1168                         /* Copied all the bits from the frag list -- free it. */
1169                         skb_frag_list_init(skb);
1170                         kfree_skb(nskb);
1171                 }
1172
1173                 skb->dev      = queue->vif->dev;
1174                 skb->protocol = eth_type_trans(skb, skb->dev);
1175                 skb_reset_network_header(skb);
1176
1177                 if (checksum_setup(queue, skb)) {
1178                         netdev_dbg(queue->vif->dev,
1179                                    "Can't setup checksum in net_tx_action\n");
1180                         /* We have to set this flag to trigger the callback */
1181                         if (skb_shinfo(skb)->destructor_arg)
1182                                 xenvif_skb_zerocopy_prepare(queue, skb);
1183                         kfree_skb(skb);
1184                         continue;
1185                 }
1186
1187                 skb_probe_transport_header(skb);
1188
1189                 /* If the packet is GSO then we will have just set up the
1190                  * transport header offset in checksum_setup so it's now
1191                  * straightforward to calculate gso_segs.
1192                  */
1193                 if (skb_is_gso(skb)) {
1194                         int mss, hdrlen;
1195
1196                         /* GSO implies having the L4 header. */
1197                         WARN_ON_ONCE(!skb_transport_header_was_set(skb));
1198                         if (unlikely(!skb_transport_header_was_set(skb))) {
1199                                 kfree_skb(skb);
1200                                 continue;
1201                         }
1202
1203                         mss = skb_shinfo(skb)->gso_size;
1204                         hdrlen = skb_tcp_all_headers(skb);
1205
1206                         skb_shinfo(skb)->gso_segs =
1207                                 DIV_ROUND_UP(skb->len - hdrlen, mss);
1208                 }
1209
1210                 queue->stats.rx_bytes += skb->len;
1211                 queue->stats.rx_packets++;
1212
1213                 work_done++;
1214
1215                 /* Set this flag right before netif_receive_skb, otherwise
1216                  * someone might think this packet already left netback, and
1217                  * do a skb_copy_ubufs while we are still in control of the
1218                  * skb. E.g. the __pskb_pull_tail earlier can do such thing.
1219                  */
1220                 if (skb_shinfo(skb)->destructor_arg) {
1221                         xenvif_skb_zerocopy_prepare(queue, skb);
1222                         queue->stats.tx_zerocopy_sent++;
1223                 }
1224
1225                 netif_receive_skb(skb);
1226         }
1227
1228         return work_done;
1229 }
1230
1231 void xenvif_zerocopy_callback(struct sk_buff *skb, struct ubuf_info *ubuf,
1232                               bool zerocopy_success)
1233 {
1234         unsigned long flags;
1235         pending_ring_idx_t index;
1236         struct xenvif_queue *queue = ubuf_to_queue(ubuf);
1237
1238         /* This is the only place where we grab this lock, to protect callbacks
1239          * from each other.
1240          */
1241         spin_lock_irqsave(&queue->callback_lock, flags);
1242         do {
1243                 u16 pending_idx = ubuf->desc;
1244                 ubuf = (struct ubuf_info *) ubuf->ctx;
1245                 BUG_ON(queue->dealloc_prod - queue->dealloc_cons >=
1246                         MAX_PENDING_REQS);
1247                 index = pending_index(queue->dealloc_prod);
1248                 queue->dealloc_ring[index] = pending_idx;
1249                 /* Sync with xenvif_tx_dealloc_action:
1250                  * insert idx then incr producer.
1251                  */
1252                 smp_wmb();
1253                 queue->dealloc_prod++;
1254         } while (ubuf);
1255         spin_unlock_irqrestore(&queue->callback_lock, flags);
1256
1257         if (likely(zerocopy_success))
1258                 queue->stats.tx_zerocopy_success++;
1259         else
1260                 queue->stats.tx_zerocopy_fail++;
1261         xenvif_skb_zerocopy_complete(queue);
1262 }
1263
1264 static inline void xenvif_tx_dealloc_action(struct xenvif_queue *queue)
1265 {
1266         struct gnttab_unmap_grant_ref *gop;
1267         pending_ring_idx_t dc, dp;
1268         u16 pending_idx, pending_idx_release[MAX_PENDING_REQS];
1269         unsigned int i = 0;
1270
1271         dc = queue->dealloc_cons;
1272         gop = queue->tx_unmap_ops;
1273
1274         /* Free up any grants we have finished using */
1275         do {
1276                 dp = queue->dealloc_prod;
1277
1278                 /* Ensure we see all indices enqueued by all
1279                  * xenvif_zerocopy_callback().
1280                  */
1281                 smp_rmb();
1282
1283                 while (dc != dp) {
1284                         BUG_ON(gop - queue->tx_unmap_ops >= MAX_PENDING_REQS);
1285                         pending_idx =
1286                                 queue->dealloc_ring[pending_index(dc++)];
1287
1288                         pending_idx_release[gop - queue->tx_unmap_ops] =
1289                                 pending_idx;
1290                         queue->pages_to_unmap[gop - queue->tx_unmap_ops] =
1291                                 queue->mmap_pages[pending_idx];
1292                         gnttab_set_unmap_op(gop,
1293                                             idx_to_kaddr(queue, pending_idx),
1294                                             GNTMAP_host_map,
1295                                             queue->grant_tx_handle[pending_idx]);
1296                         xenvif_grant_handle_reset(queue, pending_idx);
1297                         ++gop;
1298                 }
1299
1300         } while (dp != queue->dealloc_prod);
1301
1302         queue->dealloc_cons = dc;
1303
1304         if (gop - queue->tx_unmap_ops > 0) {
1305                 int ret;
1306                 ret = gnttab_unmap_refs(queue->tx_unmap_ops,
1307                                         NULL,
1308                                         queue->pages_to_unmap,
1309                                         gop - queue->tx_unmap_ops);
1310                 if (ret) {
1311                         netdev_err(queue->vif->dev, "Unmap fail: nr_ops %tu ret %d\n",
1312                                    gop - queue->tx_unmap_ops, ret);
1313                         for (i = 0; i < gop - queue->tx_unmap_ops; ++i) {
1314                                 if (gop[i].status != GNTST_okay)
1315                                         netdev_err(queue->vif->dev,
1316                                                    " host_addr: 0x%llx handle: 0x%x status: %d\n",
1317                                                    gop[i].host_addr,
1318                                                    gop[i].handle,
1319                                                    gop[i].status);
1320                         }
1321                         BUG();
1322                 }
1323         }
1324
1325         for (i = 0; i < gop - queue->tx_unmap_ops; ++i)
1326                 xenvif_idx_release(queue, pending_idx_release[i],
1327                                    XEN_NETIF_RSP_OKAY);
1328 }
1329
1330
1331 /* Called after netfront has transmitted */
1332 int xenvif_tx_action(struct xenvif_queue *queue, int budget)
1333 {
1334         unsigned nr_mops, nr_cops = 0;
1335         int work_done, ret;
1336
1337         if (unlikely(!tx_work_todo(queue)))
1338                 return 0;
1339
1340         xenvif_tx_build_gops(queue, budget, &nr_cops, &nr_mops);
1341
1342         if (nr_cops == 0)
1343                 return 0;
1344
1345         gnttab_batch_copy(queue->tx_copy_ops, nr_cops);
1346         if (nr_mops != 0) {
1347                 ret = gnttab_map_refs(queue->tx_map_ops,
1348                                       NULL,
1349                                       queue->pages_to_map,
1350                                       nr_mops);
1351                 if (ret) {
1352                         unsigned int i;
1353
1354                         netdev_err(queue->vif->dev, "Map fail: nr %u ret %d\n",
1355                                    nr_mops, ret);
1356                         for (i = 0; i < nr_mops; ++i)
1357                                 WARN_ON_ONCE(queue->tx_map_ops[i].status ==
1358                                              GNTST_okay);
1359                 }
1360         }
1361
1362         work_done = xenvif_tx_submit(queue);
1363
1364         return work_done;
1365 }
1366
1367 static void xenvif_idx_release(struct xenvif_queue *queue, u16 pending_idx,
1368                                u8 status)
1369 {
1370         struct pending_tx_info *pending_tx_info;
1371         pending_ring_idx_t index;
1372         unsigned long flags;
1373
1374         pending_tx_info = &queue->pending_tx_info[pending_idx];
1375
1376         spin_lock_irqsave(&queue->response_lock, flags);
1377
1378         make_tx_response(queue, &pending_tx_info->req,
1379                          pending_tx_info->extra_count, status);
1380
1381         /* Release the pending index before pusing the Tx response so
1382          * its available before a new Tx request is pushed by the
1383          * frontend.
1384          */
1385         index = pending_index(queue->pending_prod++);
1386         queue->pending_ring[index] = pending_idx;
1387
1388         push_tx_responses(queue);
1389
1390         spin_unlock_irqrestore(&queue->response_lock, flags);
1391 }
1392
1393
1394 static void make_tx_response(struct xenvif_queue *queue,
1395                              struct xen_netif_tx_request *txp,
1396                              unsigned int extra_count,
1397                              s8       st)
1398 {
1399         RING_IDX i = queue->tx.rsp_prod_pvt;
1400         struct xen_netif_tx_response *resp;
1401
1402         resp = RING_GET_RESPONSE(&queue->tx, i);
1403         resp->id     = txp->id;
1404         resp->status = st;
1405
1406         while (extra_count-- != 0)
1407                 RING_GET_RESPONSE(&queue->tx, ++i)->status = XEN_NETIF_RSP_NULL;
1408
1409         queue->tx.rsp_prod_pvt = ++i;
1410 }
1411
1412 static void push_tx_responses(struct xenvif_queue *queue)
1413 {
1414         int notify;
1415
1416         RING_PUSH_RESPONSES_AND_CHECK_NOTIFY(&queue->tx, notify);
1417         if (notify)
1418                 notify_remote_via_irq(queue->tx_irq);
1419 }
1420
1421 static void xenvif_idx_unmap(struct xenvif_queue *queue, u16 pending_idx)
1422 {
1423         int ret;
1424         struct gnttab_unmap_grant_ref tx_unmap_op;
1425
1426         gnttab_set_unmap_op(&tx_unmap_op,
1427                             idx_to_kaddr(queue, pending_idx),
1428                             GNTMAP_host_map,
1429                             queue->grant_tx_handle[pending_idx]);
1430         xenvif_grant_handle_reset(queue, pending_idx);
1431
1432         ret = gnttab_unmap_refs(&tx_unmap_op, NULL,
1433                                 &queue->mmap_pages[pending_idx], 1);
1434         if (ret) {
1435                 netdev_err(queue->vif->dev,
1436                            "Unmap fail: ret: %d pending_idx: %d host_addr: %llx handle: 0x%x status: %d\n",
1437                            ret,
1438                            pending_idx,
1439                            tx_unmap_op.host_addr,
1440                            tx_unmap_op.handle,
1441                            tx_unmap_op.status);
1442                 BUG();
1443         }
1444 }
1445
1446 static inline int tx_work_todo(struct xenvif_queue *queue)
1447 {
1448         if (likely(RING_HAS_UNCONSUMED_REQUESTS(&queue->tx)))
1449                 return 1;
1450
1451         return 0;
1452 }
1453
1454 static inline bool tx_dealloc_work_todo(struct xenvif_queue *queue)
1455 {
1456         return queue->dealloc_cons != queue->dealloc_prod;
1457 }
1458
1459 void xenvif_unmap_frontend_data_rings(struct xenvif_queue *queue)
1460 {
1461         if (queue->tx.sring)
1462                 xenbus_unmap_ring_vfree(xenvif_to_xenbus_device(queue->vif),
1463                                         queue->tx.sring);
1464         if (queue->rx.sring)
1465                 xenbus_unmap_ring_vfree(xenvif_to_xenbus_device(queue->vif),
1466                                         queue->rx.sring);
1467 }
1468
1469 int xenvif_map_frontend_data_rings(struct xenvif_queue *queue,
1470                                    grant_ref_t tx_ring_ref,
1471                                    grant_ref_t rx_ring_ref)
1472 {
1473         void *addr;
1474         struct xen_netif_tx_sring *txs;
1475         struct xen_netif_rx_sring *rxs;
1476         RING_IDX rsp_prod, req_prod;
1477         int err;
1478
1479         err = xenbus_map_ring_valloc(xenvif_to_xenbus_device(queue->vif),
1480                                      &tx_ring_ref, 1, &addr);
1481         if (err)
1482                 goto err;
1483
1484         txs = (struct xen_netif_tx_sring *)addr;
1485         rsp_prod = READ_ONCE(txs->rsp_prod);
1486         req_prod = READ_ONCE(txs->req_prod);
1487
1488         BACK_RING_ATTACH(&queue->tx, txs, rsp_prod, XEN_PAGE_SIZE);
1489
1490         err = -EIO;
1491         if (req_prod - rsp_prod > RING_SIZE(&queue->tx))
1492                 goto err;
1493
1494         err = xenbus_map_ring_valloc(xenvif_to_xenbus_device(queue->vif),
1495                                      &rx_ring_ref, 1, &addr);
1496         if (err)
1497                 goto err;
1498
1499         rxs = (struct xen_netif_rx_sring *)addr;
1500         rsp_prod = READ_ONCE(rxs->rsp_prod);
1501         req_prod = READ_ONCE(rxs->req_prod);
1502
1503         BACK_RING_ATTACH(&queue->rx, rxs, rsp_prod, XEN_PAGE_SIZE);
1504
1505         err = -EIO;
1506         if (req_prod - rsp_prod > RING_SIZE(&queue->rx))
1507                 goto err;
1508
1509         return 0;
1510
1511 err:
1512         xenvif_unmap_frontend_data_rings(queue);
1513         return err;
1514 }
1515
1516 static bool xenvif_dealloc_kthread_should_stop(struct xenvif_queue *queue)
1517 {
1518         /* Dealloc thread must remain running until all inflight
1519          * packets complete.
1520          */
1521         return kthread_should_stop() &&
1522                 !atomic_read(&queue->inflight_packets);
1523 }
1524
1525 int xenvif_dealloc_kthread(void *data)
1526 {
1527         struct xenvif_queue *queue = data;
1528
1529         for (;;) {
1530                 wait_event_interruptible(queue->dealloc_wq,
1531                                          tx_dealloc_work_todo(queue) ||
1532                                          xenvif_dealloc_kthread_should_stop(queue));
1533                 if (xenvif_dealloc_kthread_should_stop(queue))
1534                         break;
1535
1536                 xenvif_tx_dealloc_action(queue);
1537                 cond_resched();
1538         }
1539
1540         /* Unmap anything remaining*/
1541         if (tx_dealloc_work_todo(queue))
1542                 xenvif_tx_dealloc_action(queue);
1543
1544         return 0;
1545 }
1546
1547 static void make_ctrl_response(struct xenvif *vif,
1548                                const struct xen_netif_ctrl_request *req,
1549                                u32 status, u32 data)
1550 {
1551         RING_IDX idx = vif->ctrl.rsp_prod_pvt;
1552         struct xen_netif_ctrl_response rsp = {
1553                 .id = req->id,
1554                 .type = req->type,
1555                 .status = status,
1556                 .data = data,
1557         };
1558
1559         *RING_GET_RESPONSE(&vif->ctrl, idx) = rsp;
1560         vif->ctrl.rsp_prod_pvt = ++idx;
1561 }
1562
1563 static void push_ctrl_response(struct xenvif *vif)
1564 {
1565         int notify;
1566
1567         RING_PUSH_RESPONSES_AND_CHECK_NOTIFY(&vif->ctrl, notify);
1568         if (notify)
1569                 notify_remote_via_irq(vif->ctrl_irq);
1570 }
1571
1572 static void process_ctrl_request(struct xenvif *vif,
1573                                  const struct xen_netif_ctrl_request *req)
1574 {
1575         u32 status = XEN_NETIF_CTRL_STATUS_NOT_SUPPORTED;
1576         u32 data = 0;
1577
1578         switch (req->type) {
1579         case XEN_NETIF_CTRL_TYPE_SET_HASH_ALGORITHM:
1580                 status = xenvif_set_hash_alg(vif, req->data[0]);
1581                 break;
1582
1583         case XEN_NETIF_CTRL_TYPE_GET_HASH_FLAGS:
1584                 status = xenvif_get_hash_flags(vif, &data);
1585                 break;
1586
1587         case XEN_NETIF_CTRL_TYPE_SET_HASH_FLAGS:
1588                 status = xenvif_set_hash_flags(vif, req->data[0]);
1589                 break;
1590
1591         case XEN_NETIF_CTRL_TYPE_SET_HASH_KEY:
1592                 status = xenvif_set_hash_key(vif, req->data[0],
1593                                              req->data[1]);
1594                 break;
1595
1596         case XEN_NETIF_CTRL_TYPE_GET_HASH_MAPPING_SIZE:
1597                 status = XEN_NETIF_CTRL_STATUS_SUCCESS;
1598                 data = XEN_NETBK_MAX_HASH_MAPPING_SIZE;
1599                 break;
1600
1601         case XEN_NETIF_CTRL_TYPE_SET_HASH_MAPPING_SIZE:
1602                 status = xenvif_set_hash_mapping_size(vif,
1603                                                       req->data[0]);
1604                 break;
1605
1606         case XEN_NETIF_CTRL_TYPE_SET_HASH_MAPPING:
1607                 status = xenvif_set_hash_mapping(vif, req->data[0],
1608                                                  req->data[1],
1609                                                  req->data[2]);
1610                 break;
1611
1612         default:
1613                 break;
1614         }
1615
1616         make_ctrl_response(vif, req, status, data);
1617         push_ctrl_response(vif);
1618 }
1619
1620 static void xenvif_ctrl_action(struct xenvif *vif)
1621 {
1622         for (;;) {
1623                 RING_IDX req_prod, req_cons;
1624
1625                 req_prod = vif->ctrl.sring->req_prod;
1626                 req_cons = vif->ctrl.req_cons;
1627
1628                 /* Make sure we can see requests before we process them. */
1629                 rmb();
1630
1631                 if (req_cons == req_prod)
1632                         break;
1633
1634                 while (req_cons != req_prod) {
1635                         struct xen_netif_ctrl_request req;
1636
1637                         RING_COPY_REQUEST(&vif->ctrl, req_cons, &req);
1638                         req_cons++;
1639
1640                         process_ctrl_request(vif, &req);
1641                 }
1642
1643                 vif->ctrl.req_cons = req_cons;
1644                 vif->ctrl.sring->req_event = req_cons + 1;
1645         }
1646 }
1647
1648 static bool xenvif_ctrl_work_todo(struct xenvif *vif)
1649 {
1650         if (likely(RING_HAS_UNCONSUMED_REQUESTS(&vif->ctrl)))
1651                 return true;
1652
1653         return false;
1654 }
1655
1656 irqreturn_t xenvif_ctrl_irq_fn(int irq, void *data)
1657 {
1658         struct xenvif *vif = data;
1659         unsigned int eoi_flag = XEN_EOI_FLAG_SPURIOUS;
1660
1661         while (xenvif_ctrl_work_todo(vif)) {
1662                 xenvif_ctrl_action(vif);
1663                 eoi_flag = 0;
1664         }
1665
1666         xen_irq_lateeoi(irq, eoi_flag);
1667
1668         return IRQ_HANDLED;
1669 }
1670
1671 static int __init netback_init(void)
1672 {
1673         int rc = 0;
1674
1675         if (!xen_domain())
1676                 return -ENODEV;
1677
1678         /* Allow as many queues as there are CPUs but max. 8 if user has not
1679          * specified a value.
1680          */
1681         if (xenvif_max_queues == 0)
1682                 xenvif_max_queues = min_t(unsigned int, MAX_QUEUES_DEFAULT,
1683                                           num_online_cpus());
1684
1685         if (fatal_skb_slots < XEN_NETBK_LEGACY_SLOTS_MAX) {
1686                 pr_info("fatal_skb_slots too small (%d), bump it to XEN_NETBK_LEGACY_SLOTS_MAX (%d)\n",
1687                         fatal_skb_slots, XEN_NETBK_LEGACY_SLOTS_MAX);
1688                 fatal_skb_slots = XEN_NETBK_LEGACY_SLOTS_MAX;
1689         }
1690
1691         rc = xenvif_xenbus_init();
1692         if (rc)
1693                 goto failed_init;
1694
1695 #ifdef CONFIG_DEBUG_FS
1696         xen_netback_dbg_root = debugfs_create_dir("xen-netback", NULL);
1697 #endif /* CONFIG_DEBUG_FS */
1698
1699         return 0;
1700
1701 failed_init:
1702         return rc;
1703 }
1704
1705 module_init(netback_init);
1706
1707 static void __exit netback_fini(void)
1708 {
1709 #ifdef CONFIG_DEBUG_FS
1710         debugfs_remove_recursive(xen_netback_dbg_root);
1711 #endif /* CONFIG_DEBUG_FS */
1712         xenvif_xenbus_fini();
1713 }
1714 module_exit(netback_fini);
1715
1716 MODULE_LICENSE("Dual BSD/GPL");
1717 MODULE_ALIAS("xen-backend:vif");