1 /* 2 * Vhost User Bridge 3 * 4 * Copyright (c) 2015 Red Hat, Inc. 5 * 6 * Authors: 7 * Victor Kaplansky <victork@redhat.com> 8 * 9 * This work is licensed under the terms of the GNU GPL, version 2 or 10 * later. See the COPYING file in the top-level directory. 11 */ 12 13 /* 14 * TODO: 15 * - main should get parameters from the command line. 16 * - implement all request handlers. Still not implemented: 17 * vubr_get_queue_num_exec() 18 * vubr_send_rarp_exec() 19 * - test for broken requests and virtqueue. 20 * - implement features defined by Virtio 1.0 spec. 21 * - support mergeable buffers and indirect descriptors. 22 * - implement clean shutdown. 23 * - implement non-blocking writes to UDP backend. 24 * - implement polling strategy. 25 * - implement clean starting/stopping of vq processing 26 * - implement clean starting/stopping of used and buffers 27 * dirty page logging. 28 */ 29 30 #define _FILE_OFFSET_BITS 64 31 32 #include "qemu/osdep.h" 33 #include "qemu/iov.h" 34 #include "standard-headers/linux/virtio_net.h" 35 #include "contrib/libvhost-user/libvhost-user.h" 36 37 #define VHOST_USER_BRIDGE_DEBUG 1 38 39 #define DPRINT(...) \ 40 do { \ 41 if (VHOST_USER_BRIDGE_DEBUG) { \ 42 printf(__VA_ARGS__); \ 43 } \ 44 } while (0) 45 46 typedef void (*CallbackFunc)(int sock, void *ctx); 47 48 typedef struct Event { 49 void *ctx; 50 CallbackFunc callback; 51 } Event; 52 53 typedef struct Dispatcher { 54 int max_sock; 55 fd_set fdset; 56 Event events[FD_SETSIZE]; 57 } Dispatcher; 58 59 typedef struct VubrDev { 60 VuDev vudev; 61 Dispatcher dispatcher; 62 int backend_udp_sock; 63 struct sockaddr_in backend_udp_dest; 64 int hdrlen; 65 int sock; 66 int ready; 67 int quit; 68 } VubrDev; 69 70 static void 71 vubr_die(const char *s) 72 { 73 perror(s); 74 exit(1); 75 } 76 77 static int 78 dispatcher_init(Dispatcher *dispr) 79 { 80 FD_ZERO(&dispr->fdset); 81 dispr->max_sock = -1; 82 return 0; 83 } 84 85 static int 86 dispatcher_add(Dispatcher *dispr, int sock, void *ctx, CallbackFunc cb) 87 { 88 if (sock >= FD_SETSIZE) { 89 fprintf(stderr, 90 "Error: Failed to add new event. sock %d should be less than %d\n", 91 sock, FD_SETSIZE); 92 return -1; 93 } 94 95 dispr->events[sock].ctx = ctx; 96 dispr->events[sock].callback = cb; 97 98 FD_SET(sock, &dispr->fdset); 99 if (sock > dispr->max_sock) { 100 dispr->max_sock = sock; 101 } 102 DPRINT("Added sock %d for watching. max_sock: %d\n", 103 sock, dispr->max_sock); 104 return 0; 105 } 106 107 static int 108 dispatcher_remove(Dispatcher *dispr, int sock) 109 { 110 if (sock >= FD_SETSIZE) { 111 fprintf(stderr, 112 "Error: Failed to remove event. sock %d should be less than %d\n", 113 sock, FD_SETSIZE); 114 return -1; 115 } 116 117 FD_CLR(sock, &dispr->fdset); 118 DPRINT("Sock %d removed from dispatcher watch.\n", sock); 119 return 0; 120 } 121 122 /* timeout in us */ 123 static int 124 dispatcher_wait(Dispatcher *dispr, uint32_t timeout) 125 { 126 struct timeval tv; 127 tv.tv_sec = timeout / 1000000; 128 tv.tv_usec = timeout % 1000000; 129 130 fd_set fdset = dispr->fdset; 131 132 /* wait until some of sockets become readable. */ 133 int rc = select(dispr->max_sock + 1, &fdset, 0, 0, &tv); 134 135 if (rc == -1) { 136 vubr_die("select"); 137 } 138 139 /* Timeout */ 140 if (rc == 0) { 141 return 0; 142 } 143 144 /* Now call callback for every ready socket. */ 145 146 int sock; 147 for (sock = 0; sock < dispr->max_sock + 1; sock++) { 148 /* The callback on a socket can remove other sockets from the 149 * dispatcher, thus we have to check that the socket is 150 * still not removed from dispatcher's list 151 */ 152 if (FD_ISSET(sock, &fdset) && FD_ISSET(sock, &dispr->fdset)) { 153 Event *e = &dispr->events[sock]; 154 e->callback(sock, e->ctx); 155 } 156 } 157 158 return 0; 159 } 160 161 static void 162 vubr_handle_tx(VuDev *dev, int qidx) 163 { 164 VuVirtq *vq = vu_get_queue(dev, qidx); 165 VubrDev *vubr = container_of(dev, VubrDev, vudev); 166 int hdrlen = vubr->hdrlen; 167 VuVirtqElement *elem = NULL; 168 169 assert(qidx % 2); 170 171 for (;;) { 172 ssize_t ret; 173 unsigned int out_num; 174 struct iovec sg[VIRTQUEUE_MAX_SIZE], *out_sg; 175 176 elem = vu_queue_pop(dev, vq, sizeof(VuVirtqElement)); 177 if (!elem) { 178 break; 179 } 180 181 out_num = elem->out_num; 182 out_sg = elem->out_sg; 183 if (out_num < 1) { 184 fprintf(stderr, "virtio-net header not in first element\n"); 185 break; 186 } 187 if (VHOST_USER_BRIDGE_DEBUG) { 188 iov_hexdump(out_sg, out_num, stderr, "TX:", 1024); 189 } 190 191 if (hdrlen) { 192 unsigned sg_num = iov_copy(sg, ARRAY_SIZE(sg), 193 out_sg, out_num, 194 hdrlen, -1); 195 out_num = sg_num; 196 out_sg = sg; 197 } 198 199 struct msghdr msg = { 200 .msg_name = (struct sockaddr *) &vubr->backend_udp_dest, 201 .msg_namelen = sizeof(struct sockaddr_in), 202 .msg_iov = out_sg, 203 .msg_iovlen = out_num, 204 }; 205 do { 206 ret = sendmsg(vubr->backend_udp_sock, &msg, 0); 207 } while (ret == -1 && (errno == EAGAIN || errno == EINTR)); 208 209 if (ret == -1) { 210 vubr_die("sendmsg()"); 211 } 212 213 vu_queue_push(dev, vq, elem, 0); 214 vu_queue_notify(dev, vq); 215 216 free(elem); 217 elem = NULL; 218 } 219 220 free(elem); 221 } 222 223 static void 224 iov_restore_front(struct iovec *front, struct iovec *iov, size_t bytes) 225 { 226 struct iovec *cur; 227 228 for (cur = front; front != iov; cur++) { 229 bytes -= cur->iov_len; 230 } 231 232 cur->iov_base -= bytes; 233 cur->iov_len += bytes; 234 } 235 236 static void 237 iov_truncate(struct iovec *iov, unsigned iovc, size_t bytes) 238 { 239 unsigned i; 240 241 for (i = 0; i < iovc; i++, iov++) { 242 if (bytes < iov->iov_len) { 243 iov->iov_len = bytes; 244 return; 245 } 246 247 bytes -= iov->iov_len; 248 } 249 250 assert(!"couldn't truncate iov"); 251 } 252 253 static void 254 vubr_backend_recv_cb(int sock, void *ctx) 255 { 256 VubrDev *vubr = (VubrDev *) ctx; 257 VuDev *dev = &vubr->vudev; 258 VuVirtq *vq = vu_get_queue(dev, 0); 259 VuVirtqElement *elem = NULL; 260 struct iovec mhdr_sg[VIRTQUEUE_MAX_SIZE]; 261 struct virtio_net_hdr_mrg_rxbuf mhdr; 262 unsigned mhdr_cnt = 0; 263 int hdrlen = vubr->hdrlen; 264 int i = 0; 265 struct virtio_net_hdr hdr = { 266 .flags = 0, 267 .gso_type = VIRTIO_NET_HDR_GSO_NONE 268 }; 269 270 DPRINT("\n\n *** IN UDP RECEIVE CALLBACK ***\n\n"); 271 DPRINT(" hdrlen = %d\n", hdrlen); 272 273 if (!vu_queue_enabled(dev, vq) || 274 !vu_queue_avail_bytes(dev, vq, hdrlen, 0)) { 275 DPRINT("Got UDP packet, but no available descriptors on RX virtq.\n"); 276 return; 277 } 278 279 do { 280 struct iovec *sg; 281 ssize_t ret, total = 0; 282 unsigned int num; 283 284 elem = vu_queue_pop(dev, vq, sizeof(VuVirtqElement)); 285 if (!elem) { 286 break; 287 } 288 289 if (elem->in_num < 1) { 290 fprintf(stderr, "virtio-net contains no in buffers\n"); 291 break; 292 } 293 294 sg = elem->in_sg; 295 num = elem->in_num; 296 if (i == 0) { 297 if (hdrlen == 12) { 298 mhdr_cnt = iov_copy(mhdr_sg, ARRAY_SIZE(mhdr_sg), 299 sg, elem->in_num, 300 offsetof(typeof(mhdr), num_buffers), 301 sizeof(mhdr.num_buffers)); 302 } 303 iov_from_buf(sg, elem->in_num, 0, &hdr, sizeof hdr); 304 total += hdrlen; 305 assert(iov_discard_front(&sg, &num, hdrlen) == hdrlen); 306 } 307 308 struct msghdr msg = { 309 .msg_name = (struct sockaddr *) &vubr->backend_udp_dest, 310 .msg_namelen = sizeof(struct sockaddr_in), 311 .msg_iov = sg, 312 .msg_iovlen = elem->in_num, 313 .msg_flags = MSG_DONTWAIT, 314 }; 315 do { 316 ret = recvmsg(vubr->backend_udp_sock, &msg, 0); 317 } while (ret == -1 && (errno == EINTR)); 318 319 if (i == 0) { 320 iov_restore_front(elem->in_sg, sg, hdrlen); 321 } 322 323 if (ret == -1) { 324 if (errno == EWOULDBLOCK) { 325 vu_queue_rewind(dev, vq, 1); 326 break; 327 } 328 329 vubr_die("recvmsg()"); 330 } 331 332 total += ret; 333 iov_truncate(elem->in_sg, elem->in_num, total); 334 vu_queue_fill(dev, vq, elem, total, i++); 335 336 free(elem); 337 elem = NULL; 338 } while (false); /* could loop if DONTWAIT worked? */ 339 340 if (mhdr_cnt) { 341 mhdr.num_buffers = i; 342 iov_from_buf(mhdr_sg, mhdr_cnt, 343 0, 344 &mhdr.num_buffers, sizeof mhdr.num_buffers); 345 } 346 347 vu_queue_flush(dev, vq, i); 348 vu_queue_notify(dev, vq); 349 350 free(elem); 351 } 352 353 static void 354 vubr_receive_cb(int sock, void *ctx) 355 { 356 VubrDev *vubr = (VubrDev *)ctx; 357 358 if (!vu_dispatch(&vubr->vudev)) { 359 fprintf(stderr, "Error while dispatching\n"); 360 } 361 } 362 363 typedef struct WatchData { 364 VuDev *dev; 365 vu_watch_cb cb; 366 void *data; 367 } WatchData; 368 369 static void 370 watch_cb(int sock, void *ctx) 371 { 372 struct WatchData *wd = ctx; 373 374 wd->cb(wd->dev, VU_WATCH_IN, wd->data); 375 } 376 377 static void 378 vubr_set_watch(VuDev *dev, int fd, int condition, 379 vu_watch_cb cb, void *data) 380 { 381 VubrDev *vubr = container_of(dev, VubrDev, vudev); 382 static WatchData watches[FD_SETSIZE]; 383 struct WatchData *wd = &watches[fd]; 384 385 wd->cb = cb; 386 wd->data = data; 387 wd->dev = dev; 388 dispatcher_add(&vubr->dispatcher, fd, wd, watch_cb); 389 } 390 391 static void 392 vubr_remove_watch(VuDev *dev, int fd) 393 { 394 VubrDev *vubr = container_of(dev, VubrDev, vudev); 395 396 dispatcher_remove(&vubr->dispatcher, fd); 397 } 398 399 static int 400 vubr_send_rarp_exec(VuDev *dev, VhostUserMsg *vmsg) 401 { 402 DPRINT("Function %s() not implemented yet.\n", __func__); 403 return 0; 404 } 405 406 static int 407 vubr_process_msg(VuDev *dev, VhostUserMsg *vmsg, int *do_reply) 408 { 409 switch (vmsg->request) { 410 case VHOST_USER_SEND_RARP: 411 *do_reply = vubr_send_rarp_exec(dev, vmsg); 412 return 1; 413 default: 414 /* let the library handle the rest */ 415 return 0; 416 } 417 418 return 0; 419 } 420 421 static void 422 vubr_set_features(VuDev *dev, uint64_t features) 423 { 424 VubrDev *vubr = container_of(dev, VubrDev, vudev); 425 426 if ((features & (1ULL << VIRTIO_F_VERSION_1)) || 427 (features & (1ULL << VIRTIO_NET_F_MRG_RXBUF))) { 428 vubr->hdrlen = 12; 429 } else { 430 vubr->hdrlen = 10; 431 } 432 } 433 434 static uint64_t 435 vubr_get_features(VuDev *dev) 436 { 437 return 1ULL << VIRTIO_NET_F_GUEST_ANNOUNCE | 438 1ULL << VIRTIO_NET_F_MRG_RXBUF; 439 } 440 441 static void 442 vubr_queue_set_started(VuDev *dev, int qidx, bool started) 443 { 444 VuVirtq *vq = vu_get_queue(dev, qidx); 445 446 if (qidx % 2 == 1) { 447 vu_set_queue_handler(dev, vq, started ? vubr_handle_tx : NULL); 448 } 449 } 450 451 static void 452 vubr_panic(VuDev *dev, const char *msg) 453 { 454 VubrDev *vubr = container_of(dev, VubrDev, vudev); 455 456 fprintf(stderr, "PANIC: %s\n", msg); 457 458 dispatcher_remove(&vubr->dispatcher, dev->sock); 459 vubr->quit = 1; 460 } 461 462 static const VuDevIface vuiface = { 463 .get_features = vubr_get_features, 464 .set_features = vubr_set_features, 465 .process_msg = vubr_process_msg, 466 .queue_set_started = vubr_queue_set_started, 467 }; 468 469 static void 470 vubr_accept_cb(int sock, void *ctx) 471 { 472 VubrDev *dev = (VubrDev *)ctx; 473 int conn_fd; 474 struct sockaddr_un un; 475 socklen_t len = sizeof(un); 476 477 conn_fd = accept(sock, (struct sockaddr *) &un, &len); 478 if (conn_fd == -1) { 479 vubr_die("accept()"); 480 } 481 DPRINT("Got connection from remote peer on sock %d\n", conn_fd); 482 483 vu_init(&dev->vudev, 484 conn_fd, 485 vubr_panic, 486 vubr_set_watch, 487 vubr_remove_watch, 488 &vuiface); 489 490 dispatcher_add(&dev->dispatcher, conn_fd, ctx, vubr_receive_cb); 491 dispatcher_remove(&dev->dispatcher, sock); 492 } 493 494 static VubrDev * 495 vubr_new(const char *path, bool client) 496 { 497 VubrDev *dev = (VubrDev *) calloc(1, sizeof(VubrDev)); 498 struct sockaddr_un un; 499 CallbackFunc cb; 500 size_t len; 501 502 /* Get a UNIX socket. */ 503 dev->sock = socket(AF_UNIX, SOCK_STREAM, 0); 504 if (dev->sock == -1) { 505 vubr_die("socket"); 506 } 507 508 un.sun_family = AF_UNIX; 509 strcpy(un.sun_path, path); 510 len = sizeof(un.sun_family) + strlen(path); 511 512 if (!client) { 513 unlink(path); 514 515 if (bind(dev->sock, (struct sockaddr *) &un, len) == -1) { 516 vubr_die("bind"); 517 } 518 519 if (listen(dev->sock, 1) == -1) { 520 vubr_die("listen"); 521 } 522 cb = vubr_accept_cb; 523 524 DPRINT("Waiting for connections on UNIX socket %s ...\n", path); 525 } else { 526 if (connect(dev->sock, (struct sockaddr *)&un, len) == -1) { 527 vubr_die("connect"); 528 } 529 vu_init(&dev->vudev, 530 dev->sock, 531 vubr_panic, 532 vubr_set_watch, 533 vubr_remove_watch, 534 &vuiface); 535 cb = vubr_receive_cb; 536 } 537 538 dispatcher_init(&dev->dispatcher); 539 540 dispatcher_add(&dev->dispatcher, dev->sock, (void *)dev, cb); 541 542 return dev; 543 } 544 545 static void 546 vubr_set_host(struct sockaddr_in *saddr, const char *host) 547 { 548 if (isdigit(host[0])) { 549 if (!inet_aton(host, &saddr->sin_addr)) { 550 fprintf(stderr, "inet_aton() failed.\n"); 551 exit(1); 552 } 553 } else { 554 struct hostent *he = gethostbyname(host); 555 556 if (!he) { 557 fprintf(stderr, "gethostbyname() failed.\n"); 558 exit(1); 559 } 560 saddr->sin_addr = *(struct in_addr *)he->h_addr; 561 } 562 } 563 564 static void 565 vubr_backend_udp_setup(VubrDev *dev, 566 const char *local_host, 567 const char *local_port, 568 const char *remote_host, 569 const char *remote_port) 570 { 571 int sock; 572 const char *r; 573 574 int lport, rport; 575 576 lport = strtol(local_port, (char **)&r, 0); 577 if (r == local_port) { 578 fprintf(stderr, "lport parsing failed.\n"); 579 exit(1); 580 } 581 582 rport = strtol(remote_port, (char **)&r, 0); 583 if (r == remote_port) { 584 fprintf(stderr, "rport parsing failed.\n"); 585 exit(1); 586 } 587 588 struct sockaddr_in si_local = { 589 .sin_family = AF_INET, 590 .sin_port = htons(lport), 591 }; 592 593 vubr_set_host(&si_local, local_host); 594 595 /* setup destination for sends */ 596 dev->backend_udp_dest = (struct sockaddr_in) { 597 .sin_family = AF_INET, 598 .sin_port = htons(rport), 599 }; 600 vubr_set_host(&dev->backend_udp_dest, remote_host); 601 602 sock = socket(AF_INET, SOCK_DGRAM, IPPROTO_UDP); 603 if (sock == -1) { 604 vubr_die("socket"); 605 } 606 607 if (bind(sock, (struct sockaddr *)&si_local, sizeof(si_local)) == -1) { 608 vubr_die("bind"); 609 } 610 611 dev->backend_udp_sock = sock; 612 dispatcher_add(&dev->dispatcher, sock, dev, vubr_backend_recv_cb); 613 DPRINT("Waiting for data from udp backend on %s:%d...\n", 614 local_host, lport); 615 } 616 617 static void 618 vubr_run(VubrDev *dev) 619 { 620 while (!dev->quit) { 621 /* timeout 200ms */ 622 dispatcher_wait(&dev->dispatcher, 200000); 623 /* Here one can try polling strategy. */ 624 } 625 } 626 627 static int 628 vubr_parse_host_port(const char **host, const char **port, const char *buf) 629 { 630 char *p = strchr(buf, ':'); 631 632 if (!p) { 633 return -1; 634 } 635 *p = '\0'; 636 *host = strdup(buf); 637 *port = strdup(p + 1); 638 return 0; 639 } 640 641 #define DEFAULT_UD_SOCKET "/tmp/vubr.sock" 642 #define DEFAULT_LHOST "127.0.0.1" 643 #define DEFAULT_LPORT "4444" 644 #define DEFAULT_RHOST "127.0.0.1" 645 #define DEFAULT_RPORT "5555" 646 647 static const char *ud_socket_path = DEFAULT_UD_SOCKET; 648 static const char *lhost = DEFAULT_LHOST; 649 static const char *lport = DEFAULT_LPORT; 650 static const char *rhost = DEFAULT_RHOST; 651 static const char *rport = DEFAULT_RPORT; 652 653 int 654 main(int argc, char *argv[]) 655 { 656 VubrDev *dev; 657 int opt; 658 bool client = false; 659 660 while ((opt = getopt(argc, argv, "l:r:u:c")) != -1) { 661 662 switch (opt) { 663 case 'l': 664 if (vubr_parse_host_port(&lhost, &lport, optarg) < 0) { 665 goto out; 666 } 667 break; 668 case 'r': 669 if (vubr_parse_host_port(&rhost, &rport, optarg) < 0) { 670 goto out; 671 } 672 break; 673 case 'u': 674 ud_socket_path = strdup(optarg); 675 break; 676 case 'c': 677 client = true; 678 break; 679 default: 680 goto out; 681 } 682 } 683 684 DPRINT("ud socket: %s (%s)\n", ud_socket_path, 685 client ? "client" : "server"); 686 DPRINT("local: %s:%s\n", lhost, lport); 687 DPRINT("remote: %s:%s\n", rhost, rport); 688 689 dev = vubr_new(ud_socket_path, client); 690 if (!dev) { 691 return 1; 692 } 693 694 vubr_backend_udp_setup(dev, lhost, lport, rhost, rport); 695 vubr_run(dev); 696 697 vu_deinit(&dev->vudev); 698 699 return 0; 700 701 out: 702 fprintf(stderr, "Usage: %s ", argv[0]); 703 fprintf(stderr, "[-c] [-u ud_socket_path] [-l lhost:lport] [-r rhost:rport]\n"); 704 fprintf(stderr, "\t-u path to unix doman socket. default: %s\n", 705 DEFAULT_UD_SOCKET); 706 fprintf(stderr, "\t-l local host and port. default: %s:%s\n", 707 DEFAULT_LHOST, DEFAULT_LPORT); 708 fprintf(stderr, "\t-r remote host and port. default: %s:%s\n", 709 DEFAULT_RHOST, DEFAULT_RPORT); 710 fprintf(stderr, "\t-c client mode\n"); 711 712 return 1; 713 } 714