1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041
|
/* Copyright 2008 Bernhard R. Fischer, Daniel Haslinger.
*
* This file is part of OnionCat.
*
* OnionCat is free software: you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation, version 3 of the License.
*
* OnionCat is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with OnionCat. If not, see <http://www.gnu.org/licenses/>.
*/
/*! ocatroute.c
* Contains functions for managing both kind of TCP peers.
* Those are active SOCKS4A and passive TCP-LISTEN.
*
* @author Bernhard Fischer <rahra _at_ cypherpunk at>
* @version 2008/02/03-01
*/
#include "ocat.h"
#ifdef HAVE_STRUCT_IPHDR
#define IPPKTLEN(x) ntohs(((struct iphdr*) (x))->tot_len)
#define IPHDLEN sizeof(struct iphdr)
#else
#define IPPKTLEN(x) ntohs(((struct ip*) (x))->ip_len)
#define IPHDLEN sizeof(struct ip)
#endif
// file descriptors of socket_receiver pipe
// used for internal communication
static int lpfd_[2];
#ifdef PACKET_QUEUE
// packet queue pointer
static PacketQueue_t *queue_ = NULL;
// mutex and condition variable for packet queue
static pthread_mutex_t queue_mutex_ = PTHREAD_MUTEX_INITIALIZER;
static pthread_cond_t queue_cond_ = PTHREAD_COND_INITIALIZER;
#endif
int forward_packet(const struct in6_addr *addr, const char *buf, int buflen)
{
OcatPeer_t *peer;
int len;
lock_peers();
if ((peer = search_peer(addr)))
lock_peer(peer);
unlock_peers();
if (!peer)
{
log_debug("no peer for forwarding");
return E_FWD_NOPEER;
}
log_debug("forwarding %d bytes to TCP fd %d", buflen, peer->tcpfd);
if ((len = send(peer->tcpfd, buf, buflen, MSG_DONTWAIT)) == -1)
{
log_msg(LOG_ERR, "could not write %d bytes to peer %d: \"%s\", dropping", buflen, peer->tcpfd, strerror(errno));
}
else
{
if (len != buflen)
{
// FIXME: there should be sender frag handling!
log_msg(LOG_ERR, "could not write %d bytes to peer %d, %d bytes written", buflen, peer->tcpfd, len);
}
peer->time = time(NULL);
peer->out += len;
}
unlock_peer(peer);
return 0;
}
#ifdef PACKET_QUEUE
void queue_packet(const struct in6_addr *addr, const char *buf, int buflen)
{
PacketQueue_t *queue;
log_debug("copying packet to heap for queue");
if (!(queue = malloc(sizeof(PacketQueue_t) + buflen)))
{
log_msg(LOG_ERR, "%s for packet to queue", strerror(errno));
return;
}
//memcpy(&queue->addr, addr, sizeof(struct in6_addr));
IN6_ADDR_COPY(&queue->addr, addr);
queue->psize = buflen;
queue->data = ((char*)queue) + sizeof(PacketQueue_t);
memcpy(queue->data, buf, buflen);
queue->time = time(NULL);
log_debug("queuing packet");
pthread_mutex_lock(&queue_mutex_);
queue->next = queue_;
queue_ = queue;
log_debug("waking up dequeuer");
pthread_cond_signal(&queue_cond_);
pthread_mutex_unlock(&queue_mutex_);
}
void *packet_dequeuer(void *p)
{
PacketQueue_t **queue, *fqueue;
struct timespec ts;
struct timeval tv;
int rc, timed = 0;
time_t delay;
for (;;)
{
pthread_mutex_lock(&queue_mutex_);
if (timed)
{
// replaced clock_gettime() due to portability issues
if (gettimeofday(&tv, NULL) == -1)
{
log_msg(LOG_ERR, "couldn't gettime: \"%s\"", strerror(errno));
memset(&tv, 0, sizeof(tv));
}
else
{
ts.tv_sec = tv.tv_sec;
ts.tv_nsec = tv.tv_usec * 1000;
}
ts.tv_sec += DEQUEUER_WAKEUP;
log_debug("timed conditional wait...");
rc = pthread_cond_timedwait(&queue_cond_, &queue_mutex_, &ts);
}
else
{
log_debug("conditional wait...");
rc = pthread_cond_wait(&queue_cond_, &queue_mutex_);
}
if (rc)
log_msg(LOG_EMERG, "woke up: \"%s\"", strerror(rc));
log_debug("starting dequeuing");
for (queue = &queue_; *queue; /*queue = &(*queue)->next*/)
{
rc = forward_packet(&(*queue)->addr, (*queue)->data, (*queue)->psize);
// delete packet from queue if it was sent or is too old
delay = time(NULL) - (*queue)->time;
if (!rc || (delay > MAX_QUEUE_DELAY))
{
fqueue = *queue;
*queue = (*queue)->next;
free(fqueue);
log_debug("packet dequeued, delay = %d", delay);
continue;
}
queue = &(*queue)->next;
}
timed = queue_ != NULL;
pthread_mutex_unlock(&queue_mutex_);
}
}
#endif
/*! Check if source and destination address has
* the TOR IPv6 prefix.
* @return 0 on error or packet length else. */
int check_tor_prefix(const struct ip6_hdr *ihd)
{
char buf[INET6_ADDRSTRLEN];
if (!has_tor_prefix(&ihd->ip6_dst))
{
log_msg(LOG_ERR, "destination %s unreachable", inet_ntop(AF_INET6, &ihd->ip6_dst, buf, INET6_ADDRSTRLEN));
return 0;
}
if (!has_tor_prefix(&ihd->ip6_src))
{
log_msg(LOG_ERR, "source address invalid. Remote ocat could not reply");
return 0;
}
return ntohs(ihd->ip6_plen);
}
void cleanup_socket(int fd, OcatPeer_t *peer)
{
log_msg(LOG_INFO | LOG_FCONN, "fd %d reached EOF, closing.", fd);
oe_close(fd);
lock_peers();
delete_peer(peer);
unlock_peers();
}
#ifdef HANDLE_HTTP
#define BSTRLEN 1024
int handle_http(const OcatPeer_t *peer)
{
time_t t;
char response[BSTRLEN], timestr[BSTRLEN];
struct tm tm;
// simple check if packet could be an HTTP request
if (strncmp(peer->fragbuf, "GET ", 4))
return 0;
t = time(NULL);
(void) localtime_r(&t, &tm);
strftime(timestr, BSTRLEN, "%a, %d %b %Y %H:%M:%S %z", &tm);
snprintf(response, BSTRLEN,
"HTTP/1.0 301 HTTP not possible\r\nLocation: %s\r\nDate: %s\r\nContent-Type: text/html; charset=iso-8859-1\r\n\r\n"
"<html><body><h1>HTTP not possible!<br>OnionCat is running on this port at \"%s.onion\"</h1></body></html>\r\n",
OCAT_URL, timestr, CNF(onion_url)
);
log_msg(LOG_NOTICE, "request seems to be HTTP");
if (send(peer->tcpfd, response, strlen(response), MSG_DONTWAIT) == -1)
log_msg(LOG_ERR, "could not send html response");
return 1;
}
#endif
int set_peer_dest(struct in6_addr *dest, const struct in6_addr *addr)
{
if (!has_tor_prefix(addr))
{
log_debug("remote address does not have OC prefix");
return -1;
}
if (IN6_ARE_ADDR_EQUAL(addr, &CNF(ocat_addr)))
{
log_debug("source address is local address");
return -1;
}
*dest = *addr;
return 0;
}
/*! Set select timeout a little bit "random" to diverse wakeup periods. */
void set_select_timeout(struct timeval *tv)
{
tv->tv_usec = rand() % 1000000;
tv->tv_sec = SELECT_TIMEOUT + (tv->tv_usec & 1);
log_debug("timeout %d.%06d", tv->tv_sec, tv->tv_usec);
}
void set_tunheader(char *buf, uint32_t tunhdr)
{
uint32_t *ibuf = (uint32_t*) buf;
*ibuf = tunhdr;
}
uint32_t get_tunheader(char *buf)
{
uint32_t *ibuf = (uint32_t*) buf;
return *ibuf;
}
void *socket_receiver(void *p)
{
int maxfd, len;
char buf[FRAME_SIZE];
char addr[INET6_ADDRSTRLEN];
fd_set rset;
OcatPeer_t *peer;
struct in6_addr *in6;
int drop = 0;
struct ether_header *eh = (struct ether_header*) (buf + 4);
struct timeval tv;
if (pipe(lpfd_) < 0)
log_msg(LOG_EMERG, "could not create pipe for socket_receiver: \"%s\"", strerror(errno)), exit(1);
for (;;)
{
// check for termination request
if (term_req())
break;
FD_ZERO(&rset);
FD_SET(lpfd_[0], &rset);
maxfd = lpfd_[0];
// create set for all available peers to read
lock_peers();
for (peer = get_first_peer(); peer; peer = peer->next)
{
lock_peer(peer);
// only select active peers
if (peer->state != PEER_ACTIVE)
{
unlock_peer(peer);
continue;
}
if (peer->tcpfd >= FD_SETSIZE)
log_msg(LOG_EMERG, "%d >= FD_SETIZE(%d)", peer->tcpfd, FD_SETSIZE), exit(1);
FD_SET(peer->tcpfd, &rset);
if (peer->tcpfd > maxfd)
maxfd = peer->tcpfd;
unlock_peer(peer);
}
unlock_peers();
set_select_timeout(&tv);
log_debug("selecting (maxfd = %d)", maxfd);
if ((maxfd = select(maxfd + 1, &rset, NULL, NULL, &tv)) == -1)
{
log_msg(LOG_ERR, "select encountered error: \"%s\", restarting", strerror(errno));
continue;
}
// thread woke up because of internal pipe read => restart selection
if (FD_ISSET(lpfd_[0], &rset))
{
read(lpfd_[0], buf, FRAME_SIZE - 4);
maxfd--;
}
peer = NULL;
while (maxfd)
{
// the following 10 locs look somehow strange and someone may tend
// to write this as a for loop but it's necessary for thread locking!
lock_peers();
if (!peer)
peer = get_first_peer();
else if (!(peer = peer->next))
{
log_msg(LOG_EMERG, "fd %d ready but no peer found");
unlock_peers();
break;
}
lock_peer(peer);
unlock_peers();
if (peer->state != PEER_ACTIVE)
{
unlock_peer(peer);
continue;
}
if (!FD_ISSET(peer->tcpfd, &rset))
{
unlock_peer(peer);
continue;
}
maxfd--;
log_debug("reading from %d", peer->tcpfd);
// read/append data to peer's fragment buffer
if ((len = read(peer->tcpfd, peer->fragbuf + peer->fraglen, FRAME_SIZE - 4 - peer->fraglen)) == -1)
{
// this might happen on linux, see SELECT(2)
log_debug("spurious wakup of %d: \"%s\"", peer->tcpfd, strerror(errno));
unlock_peer(peer);
continue;
}
log_debug("received %d bytes on %d", len, peer->tcpfd);
// if len == 0 EOF reached => close session
if (!len)
{
log_msg(LOG_INFO | LOG_FCONN, "fd %d reached EOF, closing.", peer->tcpfd);
oe_close(peer->tcpfd);
// restart connection of permanent peers
if (peer->perm)
{
log_debug("reconnection permanent peer");
socks_queue(peer->addr, 1);
}
unlock_peer(peer);
// deleting peer
// FIXME: there might be a race-condition with restarted permanent peers
lock_peers();
delete_peer(peer);
unlock_peers();
continue;
}
peer->fraglen += len;
// update timestamp
peer->time = time(NULL);
peer->in += len;
while (peer->fraglen)
{
// incoming packet seems to be IPv6
if ((peer->fragbuf[0] & 0xf0) == 0x60)
{
log_debug("identified IPv6 packet");
if ((peer->fraglen < IP6HLEN) || (peer->fraglen < ntohs(((struct ip6_hdr*) peer->fragbuf)->ip6_plen) + IP6HLEN))
{
log_debug("keeping %d bytes frag", peer->fraglen);
break;
}
len = ntohs(((struct ip6_hdr*)peer->fragbuf)->ip6_plen) + IP6HLEN;
*peer->tunhdr = CNF(fhd_key[IPV6_KEY]);
}
// incoming packet seems to be IPv4
else if ((peer->fragbuf[0] & 0xf0) == 0x40)
{
if ((peer->fragbuf[0] & 0x0f) < 5)
{
log_debug("dropping packet, not IPv4 - resetting fragment buffer");
peer->fraglen = 0;
break;
}
#ifdef HANDLE_HTTP
if (handle_http(peer))
{
log_msg(LOG_INFO | LOG_FCONN, "closing %d due to HTTP", peer->tcpfd);
oe_close(peer->tcpfd);
unlock_peer(peer);
lock_peers();
delete_peer(peer);
unlock_peers();
}
#endif
log_debug("identified IPv4 packet");
if ((peer->fraglen < IPHDLEN) || (peer->fraglen < IPPKTLEN(peer->fragbuf)))
{
log_debug("keeping %d bytes frag", peer->fraglen);
break;
}
len = IPPKTLEN(peer->fragbuf);
*peer->tunhdr = CNF(fhd_key[IPV4_KEY]);
}
else
{
/*
log_debug("fragment buffer reset");
peer->fraglen = 0;
*/
log_debug("fragment buffer resynchronization");
len = 1;
drop = 1;
break;
}
// set IP address if it is not set yet and frame is valid
if (!drop && IN6_IS_ADDR_UNSPECIFIED(&peer->addr))
{
if (*peer->tunhdr == CNF(fhd_key[IPV6_KEY]))
{
//memcpy(&peer->addr, &((struct ip6_hdr*)peer->fragbuf)->ip6_src, sizeof(struct in6_addr));
if (set_peer_dest(&peer->addr, &((struct ip6_hdr*)peer->fragbuf)->ip6_src))
drop = 1;
}
else if (*peer->tunhdr == CNF(fhd_key[IPV4_KEY]))
{
// check if there is a route back
#ifdef HAVE_STRUCT_IPHDR
if (!(in6 = ipv4_lookup_route(ntohl(((struct iphdr*) peer->fragbuf)->saddr))))
#else
if (!(in6 = ipv4_lookup_route(ntohl(((struct ip*) peer->fragbuf)->ip_src.s_addr))))
#endif
{
drop = 1;
log_debug("no route back");
}
else
{
//memcpy(&peer->addr, in6, sizeof(struct in6_addr));
if (set_peer_dest(&peer->addr, in6))
drop = 1;
}
}
if (!drop)
log_msg(LOG_INFO | LOG_FCONN, "incoming connection on %d from %s is now identified", peer->tcpfd,
inet_ntop(AF_INET6, &peer->addr, addr, INET6_ADDRSTRLEN));
}
if (!drop)
{
// write directly on TUN device
if (!CNF(use_tap))
{
log_debug("writing to tun %d framesize %d + %d", CNF(tunfd[1]), len, 4 - BUF_OFF);
if (tun_write(CNF(tunfd[1]), ((char*) peer->tunhdr) + BUF_OFF, len + 4 - BUF_OFF) != (len + 4 - BUF_OFF))
log_msg(LOG_ERR, "could not write %d bytes to tunnel %d", len + 4 - BUF_OFF, CNF(tunfd[1]));
}
// create ethernet header and handle MAC on TAP device
else if (*peer->tunhdr == CNF(fhd_key[IPV6_KEY]))
{
log_debug("creating ethernet header");
// FIXME: should differentiate between IPv6 and IP!!
memset(eh->ether_dst, 0, ETHER_ADDR_LEN);
if (mac_set(&((struct ip6_hdr*)peer->fragbuf)->ip6_dst, eh->ether_dst) == -1)
{
log_debug("dest MAC unknown, resolving");
ndp_solicit(&((struct ip6_hdr*)peer->fragbuf)->ip6_src, &((struct ip6_hdr*)peer->fragbuf)->ip6_dst);
}
else
{
set_tunheader(buf, *peer->tunhdr);
memcpy(buf + 4 + sizeof(struct ether_header), peer->fragbuf, len);
memcpy(eh->ether_src, CNF(ocat_hwaddr), ETHER_ADDR_LEN);
if (*peer->tunhdr == CNF(fhd_key[IPV6_KEY]))
eh->ether_type = htons(ETHERTYPE_IPV6);
else if (*peer->tunhdr == CNF(fhd_key[IPV4_KEY]))
eh->ether_type = htons(ETHERTYPE_IP);
if (tun_write(CNF(tunfd[1]), buf + BUF_OFF, len + 4 + sizeof(struct ether_header) - BUF_OFF) != (len + 4 + sizeof(struct ether_header) - BUF_OFF))
log_msg(LOG_ERR, "could not write %d bytes to tunnel %d", len + 4 + sizeof(struct ether_header) - BUF_OFF, CNF(tunfd[1]));
}
}
else
{
log_debug("protocol %x not implemented on TAP device", ntohs(*peer->tunhdr));
}
}
else
{
log_msg(LOG_ERR, "dropping packet with %d bytes", len);
drop = 0;
}
peer->fraglen -= len;
if (peer->fraglen)
{
log_debug("moving fragment. fragsize %d", peer->fraglen);
memmove(peer->fragbuf, peer->fragbuf + len, FRAME_SIZE - 4 - len);
}
else
log_debug("fragbuf empty");
} // while (peer->fraglen)
unlock_peer(peer);
} // while (maxfd)
} // for (;;)
// closing pipe
oe_close(lpfd_[0]);
oe_close(lpfd_[1]);
return NULL;
}
void set_nonblock(int fd)
{
long flags;
if ((flags = fcntl(fd, F_GETFL, 0)) == -1)
{
log_msg(LOG_ERR, "could not get socket flags for %d: \"%s\"", fd, strerror(errno));
flags = 0;
}
log_debug("O_NONBLOCK currently is %x", flags & O_NONBLOCK);
if ((fcntl(fd, F_SETFL, flags | O_NONBLOCK)) == -1)
log_msg(LOG_ERR, "could not set O_NONBLOCK for %d: \"%s\"", fd, strerror(errno));
}
int insert_peer(int fd, const SocksQueue_t *sq, /*const struct in6_addr *addr,*/ time_t dly)
{
OcatPeer_t *peer;
log_msg(LOG_INFO | LOG_FCONN, "inserting peer fd %d to active peer list", fd);
set_nonblock(fd);
lock_peers();
if (!(peer = get_empty_peer()))
{
unlock_peers();
log_msg(LOG_ERR, "could not get new empty peer");
return 0;
}
lock_peer(peer);
unlock_peers();
peer->tcpfd = fd;
peer->state = PEER_ACTIVE;
peer->otime = peer->time = time(NULL);
peer->sdelay = dly;
if (sq)
{
//memcpy(&peer->addr, &sq->addr, sizeof(struct in6_addr));
IN6_ADDR_COPY(&peer->addr, &sq->addr);
peer->dir = PEER_OUTGOING;
peer->perm = sq->perm;
}
else
peer->dir = PEER_INCOMING;
unlock_peer(peer);
// wake up socket_receiver
log_debug("waking up socket_receiver");
if (write(lpfd_[1], &fd, 1) != 1)
log_msg(LOG_EMERG, "couldn't write to socket_receiver pipe: \"%s\"", strerror(errno));
return 1;
}
int insert_anon_peer(int fd)
{
return insert_peer(fd, NULL, 0);
}
int create_listener(struct sockaddr *addr, int sock_len)
{
int family;
int fd, so;
switch (addr->sa_family)
{
case AF_INET:
family = PF_INET;
break;
case AF_INET6:
family = PF_INET6;
break;
default:
log_msg(LOG_EMERG, "unknown address family %d", addr->sa_family);
return -1;
}
if ((fd = socket(family, SOCK_STREAM, 0)) < 0)
{
log_msg(LOG_EMERG, "could not create listener socker: \"%s\"", strerror(errno));
return -1;
}
so = 1;
if (setsockopt(fd, SOL_SOCKET, SO_REUSEADDR, &so, sizeof(so)) == -1)
log_msg(LOG_WARNING, "could not set socket %d to SO_REUSEADDR: \"%s\"", fd, strerror(errno));
if (bind(fd, addr, sock_len) == -1)
{
log_msg(LOG_EMERG, "could not bind listener %d: \"%s\"", fd, strerror(errno));
oe_close(fd);
return -1;
}
if (listen(fd, 32) < 0)
{
log_msg(LOG_EMERG, "could not bring listener %d to listening state: \"%s\"", fd, strerror(errno));
oe_close(fd);
return -1;
}
log_debug("created listener, fd = %d", fd);
return fd;
}
/*! run_listeners(...) is a generic socket acceptor for TCP ports. It listens
* on a given list of sockets. Every time a connection comes in the function
* action_accept is called with the incoming file descriptor as parameter.
*
* @param addr Double pointer to sockaddr structs. It MUST be terminated by a
* null pointer.
* @param sockfd Points to an int array. The array must contain at least as
* much entries as the sockaddr pointer has entries.
* @param action_accept Function pointer to function that should be called if a
* connection arrives.
* @return Always 0.
*/
int run_listeners(struct sockaddr **addr, int *sockfd, int cnt, int (action_accept)(int))
{
int fd;
struct sockaddr_in6 in6;
fd_set rset;
int maxfd, i;
socklen_t alen;
char iabuf[INET6_ADDRSTRLEN];
struct timeval tv;
for (i = 0; i < cnt; i++)
{
log_debug("create listener");
if ((sockfd[i] = create_listener(addr[i], SOCKADDR_SIZE(addr[i]))) == -1)
log_msg(LOG_EMERG, "exiting"), exit(1);
}
for (;;)
{
// check for termination request
if (term_req())
break;
log_debug("setting up fd_set");
FD_ZERO(&rset);
maxfd = -1;
for (i = 0; i < cnt; i++)
{
if (sockfd[i] == -1)
continue;
FD_SET(sockfd[i], &rset);
if (sockfd[i] > maxfd)
maxfd = sockfd[i];
}
if (maxfd == -1)
{
log_debug("no active listener fds available");
break;
}
set_select_timeout(&tv);
log_debug("selecting (maxfd = %d)", maxfd);
if ((maxfd = select(maxfd + 1, &rset, NULL, NULL, &tv)) == -1)
{
log_debug("select returned: \"%s\"", strerror(errno));
continue;
}
log_debug("select returned %d fds ready", maxfd);
for (i = 0; maxfd && (i < cnt); i++)
{
log_debug("checking fd %d (maxfd = %d, i = %d)", sockfd[i], maxfd, i);
if (!FD_ISSET(sockfd[i], &rset))
continue;
maxfd--;
alen = sizeof(in6);
log_debug("accepting connection on %d", sockfd[i]);
if ((fd = accept(sockfd[i], (struct sockaddr*) &in6, &alen)) < 0)
{
log_msg(LOG_ERR, "error accepting connection on %d: \"%s\"", sockfd[i], strerror(errno));
// FIXME: there should be additional error handling!
continue;
}
inet_ntop(in6.sin6_family,
in6.sin6_family == AF_INET6 ? &in6.sin6_addr :
(void*) &((struct sockaddr_in*) &in6)->sin_addr,
iabuf, INET6_ADDRSTRLEN);
log_msg(LOG_INFO | LOG_FCONN, "connection %d [%d] accepted on listener %d from %s port %d", fd, i, sockfd[i], iabuf, ntohs(in6.sin6_port));
(void) action_accept(fd);
} // for
}
// closing listeners
for (i = 0; i < cnt; i++)
oe_close(sockfd[i]);
log_debug("run_listeners returns");
return 0;
}
void *socket_acceptor(void *p)
{
run_listeners(CNF(oc_listen), CNF(oc_listen_fd), CNF(oc_listen_cnt), insert_anon_peer);
return NULL;
}
void packet_forwarder(void)
{
char buf[FRAME_SIZE];
int rlen;
struct in6_addr *dest;
struct in_addr in;
struct ether_header *eh = (struct ether_header*) &buf[4];
#ifdef PACKET_LOG
int pktlog;
log_debug("opening packetlog");
if ((pktlog = open("pkt_log", O_WRONLY | O_CREAT | O_TRUNC, S_IRUSR | S_IWUSR)) == -1)
log_debug("could not open packet log: %s", strerror(errno));
#endif
for (;;)
{
// check for termination request
if (term_req())
break;
#ifdef __OpenBSD__
// workaround for OpenBSD userland threads
fcntl(CNF(tunfd[0]), F_SETFL, fcntl(CNF(tunfd[0]), F_GETFL) & ~O_NONBLOCK);
#endif
log_debug("reading from tunfd[0] = %d", CNF(tunfd[0]));
if ((rlen = tun_read(CNF(tunfd[0]), buf + BUF_OFF, FRAME_SIZE - BUF_OFF)) == -1)
{
rlen = errno;
log_debug("read from tun %d returned on error: \"%s\"", CNF(tunfd[0]), strerror(rlen));
if (rlen == EINTR)
{
log_debug("signal caught");
if (CNF(sig_term))
{
log_msg(LOG_NOTICE, "caught termination request");
// set global termination flag
set_term_req();
}
if (CNF(sig_usr1))
{
lock_setup();
CNF(clear_stats) = 1;
unlock_setup();
log_msg(LOG_NOTICE, "stats will be cleared after next stats output");
}
}
log_debug("restarting");
continue;
}
rlen += BUF_OFF;
log_debug("received on tunfd %d, framesize %d + %d", CNF(tunfd[0]), rlen - 4, 4 - BUF_OFF);
#ifdef PACKET_LOG
if ((pktlog != -1) && (write(pktlog, buf, rlen) == -1))
log_debug("could not write frame to packet log: %s", strerror(errno));
#endif
// just to be on the safe side but this should never happen
if ((!CNF(use_tap) && (rlen < 4)) || (CNF(use_tap) && (rlen < 4 + sizeof(struct ether_header))))
{
log_msg(LOG_ERR, "frame effectively too short (rlen = %d)", rlen);
continue;
}
// in case of TAP device handle ethernet header
if (CNF(use_tap))
{
if (eth_check(buf, rlen))
continue;
// removing ethernet header
// FIXME: it would be better to adjust pointers instead of moving data
rlen -= sizeof(struct ether_header);
memmove(eh, eh + 1, rlen - 4);
}
#if defined(__sun__) || defined(__CYGWIN__)
// Solaris tunnel driver does not send tunnel
// header thus we guess and set it manually
if ((buf[BUF_OFF] & 0xf0) == 0x60)
set_tunheader(buf, CNF(fhd_key[IPV6_KEY]));
else if ((buf[BUF_OFF] & 0xf0) == 0x40)
set_tunheader(buf, CNF(fhd_key[IPV4_KEY]));
else
set_tunheader(buf, -1);
#endif
if (get_tunheader(buf) == CNF(fhd_key[IPV6_KEY]))
{
if (((rlen - 4) < IP6HLEN))
{
log_debug("IPv6 packet too short (%d bytes). dropping", rlen - 4);
continue;
}
#ifndef CHECK_IPSRC
if (!check_tor_prefix((struct ip6_hdr*) &buf[4]))
{
log_msg(LOG_ERR, "dropping frame");
continue;
}
#endif
if (!(dest = ipv6_lookup_route(&((struct ip6_hdr*) &buf[4])->ip6_dst)))
dest = &((struct ip6_hdr*) &buf[4])->ip6_dst;
}
else if (get_tunheader(buf) == CNF(fhd_key[IPV4_KEY]))
{
if (((rlen - 4) < IPHDLEN))
{
log_debug("IPv4 packet too short (%d bytes). dropping", rlen - 4);
continue;
}
#ifdef HAVE_STRUCT_IPHDR
in.s_addr = ((struct iphdr*) &buf[4])->daddr;
#else
in.s_addr = ((struct ip*) &buf[4])->ip_dst.s_addr;
#endif
if (!(dest = ipv4_lookup_route(ntohl(in.s_addr))))
{
log_msg(LOG_ERR, "no route to destination %s, dropping frame.", inet_ntoa(in));
continue;
}
}
else
{
log_msg(LOG_ERR, "protocol 0x%08x not supported. dropping frame.", ntohl(get_tunheader(buf)));
continue;
}
// now forward either directly or to the queue
if (forward_packet(dest, buf + 4, rlen - 4) == E_FWD_NOPEER)
{
log_debug("adding destination to SOCKS queue");
socks_queue(*dest, 0);
#ifdef PACKET_QUEUE
log_debug("queuing packet");
queue_packet(dest, buf + 4, rlen - 4);
#endif
}
}
}
int send_keepalive(OcatPeer_t *peer)
{
struct ip6_hdr hdr;
int len;
memset(&hdr, 0, sizeof(hdr));
//memcpy(&hdr.ip6_dst, &peer->addr, sizeof(struct in6_addr));
IN6_ADDR_COPY(&hdr.ip6_dst, &peer->addr);
//memcpy(&hdr.ip6_src, &CNF(ocat_addr), sizeof(struct in6_addr));
IN6_ADDR_COPY(&hdr.ip6_src, &CNF(ocat_addr));
hdr.ip6_vfc = 0x60;
hdr.ip6_nxt = IPPROTO_NONE;
hdr.ip6_hops = 1;
log_debug("sending %d bytes keepalive to fd %d", sizeof(hdr), peer->tcpfd);
if ((len = send(peer->tcpfd, &hdr, sizeof(hdr), MSG_DONTWAIT)) == -1)
{
log_msg(LOG_ERR, "could not send keepalive: %s", strerror(errno));
return -1;
}
peer->out += len;
if (len != sizeof(hdr))
{
log_msg(LOG_ERR, "sending of %d bytes keepalive truncated to %d", sizeof(hdr), len);
return -1;
}
return 0;
}
void *socket_cleaner(void *ptr)
{
OcatPeer_t *peer, **p;
int stat_wup = 0;
time_t act_time;
for (;;)
{
// check for termination request
if (term_req())
break;
sleep(CLEANER_WAKEUP);
log_debug("wakeup");
act_time = time(NULL);
// stats output
if (act_time - stat_wup >= STAT_WAKEUP)
{
stat_wup = act_time;
log_msg(LOG_INFO, "stats: ... (not implemented yet)");
lock_setup();
if (CNF(clear_stats))
{
CNF(clear_stats) = 0;
// FIXME: implement stats clearing here
log_debug("stats cleared");
}
unlock_setup();
}
// cleanup MAC table
mac_cleanup();
// cleanup peers
lock_peers();
for (p = get_first_peer_ptr(); *p; p = &(*p)->next)
{
lock_peer(*p);
// handle permanent connections
if ((*p)->perm)
{
// sending keepalive
if (act_time - (*p)->time >= KEEPALIVE_TIME)
{
send_keepalive(*p);
(*p)->time = act_time;
}
unlock_peer(*p);
}
// handle temporary connections
else if ((*p)->state && act_time - (*p)->time >= MAX_IDLE_TIME)
{
peer = *p;
*p = peer->next;
log_msg(LOG_INFO | LOG_FCONN, "peer %d timed out, closing.", peer->tcpfd);
oe_close(peer->tcpfd);
unlock_peer(peer);
delete_peer(peer);
if (!(*p))
{
log_debug("last peer in list deleted, breaking loop");
break;
}
}
else
unlock_peer(*p);
}
unlock_peers();
}
return NULL;
}
|