/* $Id: slirp.c 98103 2023-01-17 14:15:46Z vboxsync $ */ /** @file * NAT - slirp glue. */ /* * Copyright (C) 2006-2023 Oracle and/or its affiliates. * * This file is part of VirtualBox base platform packages, as * available from https://www.virtualbox.org. * * This program 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, in version 3 of the * License. * * This program 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 this program; if not, see . * * SPDX-License-Identifier: GPL-3.0-only */ /* * This code is based on: * * libslirp glue * * Copyright (c) 2004-2008 Fabrice Bellard * * Permission is hereby granted, free of charge, to any person obtaining a copy * of this software and associated documentation files (the "Software"), to deal * in the Software without restriction, including without limitation the rights * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell * copies of the Software, and to permit persons to whom the Software is * furnished to do so, subject to the following conditions: * * The above copyright notice and this permission notice shall be included in * all copies or substantial portions of the Software. * * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN * THE SOFTWARE. */ #include "slirp.h" #ifdef RT_OS_OS2 # include #endif #include #include #include #include #include #include #ifndef RT_OS_WINDOWS # include # include # include #else # include # define _WINSOCK2API_ # include #endif #include #ifndef RT_OS_WINDOWS /** * XXX: It shouldn't be non-Windows specific. * resolv_conf_parser.h client's structure isn't OS specific, it's just need to be generalized a * a bit to replace slirp_state.h DNS server (domain) lists with rcp_state like structure. */ # include "resolv_conf_parser.h" #endif #ifndef RT_OS_WINDOWS # define DO_ENGAGE_EVENT1(so, fdset, label) \ do { \ if ( so->so_poll_index != -1 \ && so->s == polls[so->so_poll_index].fd) \ { \ polls[so->so_poll_index].events |= N_(fdset ## _poll); \ break; \ } \ AssertRelease(poll_index < (nfds)); \ AssertRelease(poll_index >= 0 && poll_index < (nfds)); \ polls[poll_index].fd = (so)->s; \ (so)->so_poll_index = poll_index; \ polls[poll_index].events = N_(fdset ## _poll); \ polls[poll_index].revents = 0; \ poll_index++; \ } while (0) # define DO_ENGAGE_EVENT2(so, fdset1, fdset2, label) \ do { \ if ( so->so_poll_index != -1 \ && so->s == polls[so->so_poll_index].fd) \ { \ polls[so->so_poll_index].events |= \ N_(fdset1 ## _poll) | N_(fdset2 ## _poll); \ break; \ } \ AssertRelease(poll_index < (nfds)); \ polls[poll_index].fd = (so)->s; \ (so)->so_poll_index = poll_index; \ polls[poll_index].events = \ N_(fdset1 ## _poll) | N_(fdset2 ## _poll); \ poll_index++; \ } while (0) # define DO_POLL_EVENTS(rc, error, so, events, label) do {} while (0) /* * DO_CHECK_FD_SET is used in dumping events on socket, including POLLNVAL. * gcc warns about attempts to log POLLNVAL so construction in a last to lines * used to catch POLLNVAL while logging and return false in case of error while * normal usage. */ # define DO_CHECK_FD_SET(so, events, fdset) \ ( ((so)->so_poll_index != -1) \ && ((so)->so_poll_index <= ndfs) \ && ((so)->s == polls[so->so_poll_index].fd) \ && (polls[(so)->so_poll_index].revents & N_(fdset ## _poll)) \ && ( N_(fdset ## _poll) == POLLNVAL \ || !(polls[(so)->so_poll_index].revents & POLLNVAL))) /* specific for Windows Winsock API */ # define DO_WIN_CHECK_FD_SET(so, events, fdset) 0 # ifndef RT_OS_LINUX # define readfds_poll (POLLRDNORM) # define writefds_poll (POLLWRNORM) # else # define readfds_poll (POLLIN) # define writefds_poll (POLLOUT) # endif # define xfds_poll (POLLPRI) # define closefds_poll (POLLHUP) # define rderr_poll (POLLERR) # if 0 /* unused yet */ # define rdhup_poll (POLLHUP) # define nval_poll (POLLNVAL) # endif # define ICMP_ENGAGE_EVENT(so, fdset) \ do { \ if (pData->icmp_socket.s != -1) \ DO_ENGAGE_EVENT1((so), fdset, ICMP); \ } while (0) #else /* RT_OS_WINDOWS */ /* * On Windows, we will be notified by IcmpSendEcho2() when the response arrives. * So no call to WSAEventSelect necessary. */ # define ICMP_ENGAGE_EVENT(so, fdset) do {} while (0) /* * On Windows we use FD_ALL_EVENTS to ensure that we don't miss any event. */ # define DO_ENGAGE_EVENT1(so, fdset1, label) \ do { \ rc = WSAEventSelect((so)->s, VBOX_SOCKET_EVENT, FD_ALL_EVENTS); \ if (rc == SOCKET_ERROR) \ { \ /* This should not happen */ \ error = WSAGetLastError(); \ LogRel(("WSAEventSelect (" #label ") error %d (so=%x, socket=%s, event=%x)\n", \ error, (so), (so)->s, VBOX_SOCKET_EVENT)); \ } \ } while (0); \ CONTINUE(label) # define DO_ENGAGE_EVENT2(so, fdset1, fdset2, label) \ DO_ENGAGE_EVENT1((so), (fdset1), label) # define DO_POLL_EVENTS(rc, error, so, events, label) \ (rc) = WSAEnumNetworkEvents((so)->s, VBOX_SOCKET_EVENT, (events)); \ if ((rc) == SOCKET_ERROR) \ { \ (error) = WSAGetLastError(); \ LogRel(("WSAEnumNetworkEvents %R[natsock] " #label " error %d\n", (so), (error))); \ LogFunc(("WSAEnumNetworkEvents %R[natsock] " #label " error %d\n", (so), (error))); \ CONTINUE(label); \ } # define acceptds_win FD_ACCEPT # define acceptds_win_bit FD_ACCEPT_BIT # define readfds_win FD_READ # define readfds_win_bit FD_READ_BIT # define writefds_win FD_WRITE # define writefds_win_bit FD_WRITE_BIT # define xfds_win FD_OOB # define xfds_win_bit FD_OOB_BIT # define closefds_win FD_CLOSE # define closefds_win_bit FD_CLOSE_BIT # define connectfds_win FD_CONNECT # define connectfds_win_bit FD_CONNECT_BIT # define closefds_win FD_CLOSE # define closefds_win_bit FD_CLOSE_BIT # define DO_CHECK_FD_SET(so, events, fdset) \ ((events).lNetworkEvents & fdset ## _win) # define DO_WIN_CHECK_FD_SET(so, events, fdset) DO_CHECK_FD_SET((so), (events), fdset) # define DO_UNIX_CHECK_FD_SET(so, events, fdset) 1 /*specific for Unix API */ #endif /* RT_OS_WINDOWS */ #define TCP_ENGAGE_EVENT1(so, fdset) \ DO_ENGAGE_EVENT1((so), fdset, tcp) #define TCP_ENGAGE_EVENT2(so, fdset1, fdset2) \ DO_ENGAGE_EVENT2((so), fdset1, fdset2, tcp) #ifdef RT_OS_WINDOWS # define WIN_TCP_ENGAGE_EVENT2(so, fdset, fdset2) TCP_ENGAGE_EVENT2(so, fdset1, fdset2) #endif #define UDP_ENGAGE_EVENT(so, fdset) \ DO_ENGAGE_EVENT1((so), fdset, udp) #define POLL_TCP_EVENTS(rc, error, so, events) \ DO_POLL_EVENTS((rc), (error), (so), (events), tcp) #define POLL_UDP_EVENTS(rc, error, so, events) \ DO_POLL_EVENTS((rc), (error), (so), (events), udp) #define CHECK_FD_SET(so, events, set) \ (DO_CHECK_FD_SET((so), (events), set)) #define WIN_CHECK_FD_SET(so, events, set) \ (DO_WIN_CHECK_FD_SET((so), (events), set)) /* * Loging macros */ #ifdef VBOX_WITH_DEBUG_NAT_SOCKETS # if defined(RT_OS_WINDOWS) # define DO_LOG_NAT_SOCK(so, proto, winevent, r_fdset, w_fdset, x_fdset) \ do { \ LogRel((" " #proto " %R[natsock] %R[natwinnetevents]\n", (so), (winevent))); \ } while (0) # else /* !RT_OS_WINDOWS */ # define DO_LOG_NAT_SOCK(so, proto, winevent, r_fdset, w_fdset, x_fdset) \ do { \ LogRel((" " #proto " %R[natsock] %s %s %s er: %s, %s, %s\n", (so), \ CHECK_FD_SET(so, ign ,r_fdset) ? "READ":"", \ CHECK_FD_SET(so, ign, w_fdset) ? "WRITE":"", \ CHECK_FD_SET(so, ign, x_fdset) ? "OOB":"", \ CHECK_FD_SET(so, ign, rderr) ? "RDERR":"", \ CHECK_FD_SET(so, ign, rdhup) ? "RDHUP":"", \ CHECK_FD_SET(so, ign, nval) ? "RDNVAL":"")); \ } while (0) # endif /* !RT_OS_WINDOWS */ #else /* !VBOX_WITH_DEBUG_NAT_SOCKETS */ # define DO_LOG_NAT_SOCK(so, proto, winevent, r_fdset, w_fdset, x_fdset) do {} while (0) #endif /* !VBOX_WITH_DEBUG_NAT_SOCKETS */ #define LOG_NAT_SOCK(so, proto, winevent, r_fdset, w_fdset, x_fdset) \ DO_LOG_NAT_SOCK((so), proto, (winevent), r_fdset, w_fdset, x_fdset) static const uint8_t special_ethaddr[6] = { 0x52, 0x54, 0x00, 0x12, 0x35, 0x00 }; static const uint8_t broadcast_ethaddr[6] = { 0xff, 0xff, 0xff, 0xff, 0xff, 0xff }; const uint8_t zerro_ethaddr[6] = { 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 }; /** * This helper routine do the checks in descriptions to * ''fUnderPolling'' and ''fShouldBeRemoved'' flags * @returns 1 if socket removed and 0 if no changes was made. */ static int slirpVerifyAndFreeSocket(PNATState pData, struct socket *pSocket) { AssertPtrReturn(pData, 0); AssertPtrReturn(pSocket, 0); AssertReturn(pSocket->fUnderPolling, 0); if (pSocket->fShouldBeRemoved) { pSocket->fUnderPolling = 0; sofree(pData, pSocket); /* pSocket is PHANTOM, now */ return 1; } return 0; } int slirp_init(PNATState *ppData, uint32_t u32NetAddr, uint32_t u32Netmask, bool fPassDomain, bool fUseHostResolver, int i32AliasMode, int iIcmpCacheLimit, bool fLocalhostReachable, void *pvUser) { int rc; PNATState pData; if (u32Netmask & 0x1f) { /* CTL is x.x.x.15, bootp passes up to 16 IPs (15..31) */ LogRel(("NAT: The last 5 bits of the netmask (%RTnaipv4) need to be unset\n", RT_BE2H_U32(u32Netmask))); return VERR_INVALID_PARAMETER; } pData = RTMemAllocZ(RT_ALIGN_Z(sizeof(NATState), sizeof(uint64_t))); *ppData = pData; if (!pData) return VERR_NO_MEMORY; pData->fPassDomain = !fUseHostResolver ? fPassDomain : false; pData->fUseHostResolver = fUseHostResolver; pData->fUseHostResolverPermanent = fUseHostResolver; pData->fLocalhostReachable = fLocalhostReachable; pData->pvUser = pvUser; pData->netmask = u32Netmask; rc = RTCritSectRwInit(&pData->CsRwHandlerChain); if (RT_FAILURE(rc)) return rc; /* sockets & TCP defaults */ pData->socket_rcv = 64 * _1K; pData->socket_snd = 64 * _1K; tcp_sndspace = 64 * _1K; tcp_rcvspace = 64 * _1K; /* * Use the same default here as in DevNAT.cpp (SoMaxConnection CFGM value) * to avoid release log noise. */ pData->soMaxConn = 10; #ifdef RT_OS_WINDOWS { WSADATA Data; RTLDRMOD hLdrMod; WSAStartup(MAKEWORD(2, 0), &Data); rc = RTLdrLoadSystem("Iphlpapi.dll", true /*fNoUnload*/, &hLdrMod); if (RT_SUCCESS(rc)) { rc = RTLdrGetSymbol(hLdrMod, "GetAdaptersAddresses", (void **)&pData->pfnGetAdaptersAddresses); if (RT_FAILURE(rc)) LogRel(("NAT: Can't find GetAdapterAddresses in Iphlpapi.dll\n")); RTLdrClose(hLdrMod); } } pData->phEvents[VBOX_SOCKET_EVENT_INDEX] = CreateEvent(NULL, FALSE, FALSE, NULL); #endif rc = bootp_dhcp_init(pData); if (RT_FAILURE(rc)) { Log(("NAT: DHCP server initialization failed\n")); RTMemFree(pData); *ppData = NULL; return rc; } debug_init(pData); if_init(pData); ip_init(pData); icmp_init(pData, iIcmpCacheLimit); /* Initialise mbufs *after* setting the MTU */ mbuf_init(pData); pData->special_addr.s_addr = u32NetAddr; pData->slirp_ethaddr = &special_ethaddr[0]; alias_addr.s_addr = pData->special_addr.s_addr | RT_H2N_U32_C(CTL_ALIAS); /** @todo add ability to configure this staff */ /* * Some guests won't reacquire DHCP lease on link flap when VM is * restored. Instead of forcing users to explicitly set CTL_GUEST * in port-forwarding rules, provide it as initial guess here. */ slirp_update_guest_addr_guess(pData, pData->special_addr.s_addr | RT_H2N_U32_C(CTL_GUEST), "initialization"); /* set default addresses */ inet_aton("127.0.0.1", &loopback_addr); rc = slirpTftpInit(pData); AssertRCReturn(rc, rc); if (i32AliasMode & ~(PKT_ALIAS_LOG|PKT_ALIAS_SAME_PORTS|PKT_ALIAS_PROXY_ONLY)) { LogRel(("NAT: bad alias mode 0x%x ignored\n", i32AliasMode)); i32AliasMode = 0; } else if (i32AliasMode != 0) { LogRel(("NAT: alias mode 0x%x\n", i32AliasMode)); } pData->i32AliasMode = i32AliasMode; getouraddr(pData); { int flags = 0; struct in_addr proxy_addr; pData->proxy_alias = LibAliasInit(pData, NULL); if (pData->proxy_alias == NULL) { Log(("NAT: LibAlias default rule wasn't initialized\n")); AssertMsgFailed(("NAT: LibAlias default rule wasn't initialized\n")); } flags = LibAliasSetMode(pData->proxy_alias, 0, 0); #ifndef NO_FW_PUNCH flags |= PKT_ALIAS_PUNCH_FW; #endif flags |= pData->i32AliasMode; /* do transparent proxying */ flags = LibAliasSetMode(pData->proxy_alias, flags, ~0U); proxy_addr.s_addr = RT_H2N_U32(RT_N2H_U32(pData->special_addr.s_addr) | CTL_ALIAS); LibAliasSetAddress(pData->proxy_alias, proxy_addr); ftp_alias_load(pData); nbt_alias_load(pData); } #ifdef VBOX_WITH_NAT_SEND2HOME /** @todo we should know all interfaces available on host. */ pData->pInSockAddrHomeAddress = RTMemAllocZ(sizeof(struct sockaddr)); pData->cInHomeAddressSize = 1; inet_aton("192.168.1.25", &pData->pInSockAddrHomeAddress[0].sin_addr); pData->pInSockAddrHomeAddress[0].sin_family = AF_INET; # ifdef RT_OS_DARWIN pData->pInSockAddrHomeAddress[0].sin_len = sizeof(struct sockaddr_in); # endif #endif #ifdef VBOX_WITH_DNSMAPPING_IN_HOSTRESOLVER STAILQ_INIT(&pData->DNSMapNames); STAILQ_INIT(&pData->DNSMapPatterns); #endif slirp_link_up(pData); return VINF_SUCCESS; } /** * Register statistics. */ void slirp_register_statistics(PNATState pData, PPDMDRVINS pDrvIns) { #ifdef VBOX_WITH_STATISTICS # define PROFILE_COUNTER(name, dsc) REGISTER_COUNTER(name, pData, STAMTYPE_PROFILE, STAMUNIT_TICKS_PER_CALL, dsc) # define COUNTING_COUNTER(name, dsc) REGISTER_COUNTER(name, pData, STAMTYPE_COUNTER, STAMUNIT_COUNT, dsc) # include "counters.h" # undef COUNTER /** @todo register statistics for the variables dumped by: * ipstats(pData); tcpstats(pData); udpstats(pData); icmpstats(pData); * mbufstats(pData); sockstats(pData); */ #else /* VBOX_WITH_STATISTICS */ NOREF(pData); NOREF(pDrvIns); #endif /* !VBOX_WITH_STATISTICS */ } /** * Deregister statistics. */ void slirp_deregister_statistics(PNATState pData, PPDMDRVINS pDrvIns) { if (pData == NULL) return; #ifdef VBOX_WITH_STATISTICS # define PROFILE_COUNTER(name, dsc) DEREGISTER_COUNTER(name, pData) # define COUNTING_COUNTER(name, dsc) DEREGISTER_COUNTER(name, pData) # include "counters.h" #else /* VBOX_WITH_STATISTICS */ NOREF(pData); NOREF(pDrvIns); #endif /* !VBOX_WITH_STATISTICS */ } /** * Marks the link as up, making it possible to establish new connections. */ void slirp_link_up(PNATState pData) { if (link_up == 1) return; link_up = 1; if (!pData->fUseHostResolverPermanent) slirpInitializeDnsSettings(pData); } /** * Marks the link as down and cleans up the current connections. */ void slirp_link_down(PNATState pData) { if (link_up == 0) return; slirpReleaseDnsSettings(pData); link_up = 0; } /** * Terminates the slirp component. */ void slirp_term(PNATState pData) { struct socket *so; if (pData == NULL) return; icmp_finit(pData); while ((so = tcb.so_next) != &tcb) { /* Don't miss TCB releasing */ if ( !sototcpcb(so) && ( so->so_state & SS_NOFDREF || so->s == -1)) sofree(pData, so); else tcp_close(pData, sototcpcb(so)); } while ((so = udb.so_next) != &udb) udp_detach(pData, so); slirp_link_down(pData); ftp_alias_unload(pData); nbt_alias_unload(pData); #ifdef VBOX_WITH_DNSMAPPING_IN_HOSTRESOLVER { DNSMAPPINGHEAD *heads[2]; int i; heads[0] = &pData->DNSMapNames; heads[1] = &pData->DNSMapPatterns; for (i = 0; i < RT_ELEMENTS(heads); ++i) { while (!STAILQ_EMPTY(heads[i])) { PDNSMAPPINGENTRY pDnsEntry = STAILQ_FIRST(heads[i]); STAILQ_REMOVE_HEAD(heads[i], MapList); RTStrFree(pDnsEntry->pszName); RTMemFree(pDnsEntry); } } } #endif while (!LIST_EMPTY(&instancehead)) { struct libalias *la = LIST_FIRST(&instancehead); /* libalias do all clean up */ LibAliasUninit(la); } while (!LIST_EMPTY(&pData->arp_cache)) { struct arp_cache_entry *ac = LIST_FIRST(&pData->arp_cache); LIST_REMOVE(ac, list); RTMemFree(ac); } while (!LIST_EMPTY(&pData->port_forward_rule_head)) { struct port_forward_rule *rule = LIST_FIRST(&pData->port_forward_rule_head); LIST_REMOVE(rule, list); RTMemFree(rule); } slirpTftpTerm(pData); bootp_dhcp_fini(pData); m_fini(pData); #ifdef RT_OS_WINDOWS WSACleanup(); #endif if (tftp_prefix) RTStrFree((char *)tftp_prefix); #ifdef LOG_ENABLED Log(("\n" "NAT statistics\n" "--------------\n" "\n")); ipstats(pData); tcpstats(pData); udpstats(pData); icmpstats(pData); mbufstats(pData); sockstats(pData); Log(("\n" "\n" "\n")); #endif RTCritSectRwDelete(&pData->CsRwHandlerChain); RTMemFree(pData); } #define CONN_CANFSEND(so) (((so)->so_state & (SS_FCANTSENDMORE|SS_ISFCONNECTED)) == SS_ISFCONNECTED) #define CONN_CANFRCV(so) (((so)->so_state & (SS_FCANTRCVMORE|SS_ISFCONNECTED)) == SS_ISFCONNECTED) /* * curtime kept to an accuracy of 1ms */ static void updtime(PNATState pData) { #ifdef RT_OS_WINDOWS struct _timeb tb; _ftime(&tb); curtime = (u_int)tb.time * (u_int)1000; curtime += (u_int)tb.millitm; #else gettimeofday(&tt, 0); curtime = (u_int)tt.tv_sec * (u_int)1000; curtime += (u_int)tt.tv_usec / (u_int)1000; if ((tt.tv_usec % 1000) >= 500) curtime++; #endif } #ifdef RT_OS_WINDOWS void slirp_select_fill(PNATState pData, int *pnfds) #else /* RT_OS_WINDOWS */ void slirp_select_fill(PNATState pData, int *pnfds, struct pollfd *polls) #endif /* !RT_OS_WINDOWS */ { struct socket *so, *so_next; int nfds; #if defined(RT_OS_WINDOWS) int rc; int error; #else int poll_index = 0; #endif int i; STAM_PROFILE_START(&pData->StatFill, a); nfds = *pnfds; /* * First, TCP sockets */ do_slowtimo = 0; if (!link_up) goto done; /* * *_slowtimo needs calling if there are IP fragments * in the fragment queue, or there are TCP connections active */ /* XXX: * triggering of fragment expiration should be the same but use new macroses */ do_slowtimo = (tcb.so_next != &tcb); if (!do_slowtimo) { for (i = 0; i < IPREASS_NHASH; i++) { if (!TAILQ_EMPTY(&ipq[i])) { do_slowtimo = 1; break; } } } /* always add the ICMP socket */ #ifndef RT_OS_WINDOWS pData->icmp_socket.so_poll_index = -1; #endif ICMP_ENGAGE_EVENT(&pData->icmp_socket, readfds); STAM_COUNTER_RESET(&pData->StatTCP); STAM_COUNTER_RESET(&pData->StatTCPHot); QSOCKET_FOREACH(so, so_next, tcp) /* { */ Assert(so->so_type == IPPROTO_TCP); #if !defined(RT_OS_WINDOWS) so->so_poll_index = -1; #endif STAM_COUNTER_INC(&pData->StatTCP); /* * See if we need a tcp_fasttimo */ if ( time_fasttimo == 0 && so->so_tcpcb != NULL && so->so_tcpcb->t_flags & TF_DELACK) { time_fasttimo = curtime; /* Flag when we want a fasttimo */ } /* * NOFDREF can include still connecting to local-host, * newly socreated() sockets etc. Don't want to select these. */ if (so->so_state & SS_NOFDREF || so->s == -1) CONTINUE(tcp); /* * Set for reading sockets which are accepting */ if (so->so_state & SS_FACCEPTCONN) { STAM_COUNTER_INC(&pData->StatTCPHot); TCP_ENGAGE_EVENT1(so, readfds); CONTINUE(tcp); } /* * Set for writing sockets which are connecting */ if (so->so_state & SS_ISFCONNECTING) { Log2(("connecting %R[natsock] engaged\n",so)); STAM_COUNTER_INC(&pData->StatTCPHot); #ifdef RT_OS_WINDOWS WIN_TCP_ENGAGE_EVENT2(so, writefds, connectfds); #else TCP_ENGAGE_EVENT1(so, writefds); #endif } /* * Set for writing if we are connected, can send more, and * we have something to send */ if (CONN_CANFSEND(so) && SBUF_LEN(&so->so_rcv)) { STAM_COUNTER_INC(&pData->StatTCPHot); TCP_ENGAGE_EVENT1(so, writefds); } /* * Set for reading (and urgent data) if we are connected, can * receive more, and we have room for it XXX /2 ? */ /** @todo vvl - check which predicat here will be more useful here in rerm of new sbufs. */ if ( CONN_CANFRCV(so) && (SBUF_LEN(&so->so_snd) < (SBUF_SIZE(&so->so_snd)/2)) #ifdef RT_OS_WINDOWS && !(so->so_state & SS_ISFCONNECTING) #endif ) { STAM_COUNTER_INC(&pData->StatTCPHot); TCP_ENGAGE_EVENT2(so, readfds, xfds); } LOOP_LABEL(tcp, so, so_next); } /* * UDP sockets */ STAM_COUNTER_RESET(&pData->StatUDP); STAM_COUNTER_RESET(&pData->StatUDPHot); QSOCKET_FOREACH(so, so_next, udp) /* { */ Assert(so->so_type == IPPROTO_UDP); STAM_COUNTER_INC(&pData->StatUDP); #if !defined(RT_OS_WINDOWS) so->so_poll_index = -1; #endif /* * See if it's timed out */ if (so->so_expire) { if (so->so_expire <= curtime) { Log2(("NAT: %R[natsock] expired\n", so)); if (so->so_timeout != NULL) { /* so_timeout - might change the so_expire value or * drop so_timeout* from so. */ so->so_timeout(pData, so, so->so_timeout_arg); /* on 4.2 so-> */ if ( so_next->so_prev != so /* so_timeout freed the socket */ || so->so_timeout) /* so_timeout just freed so_timeout */ CONTINUE_NO_UNLOCK(udp); } UDP_DETACH(pData, so, so_next); CONTINUE_NO_UNLOCK(udp); } } /* * When UDP packets are received from over the link, they're * sendto()'d straight away, so no need for setting for writing * Limit the number of packets queued by this session to 4. * Note that even though we try and limit this to 4 packets, * the session could have more queued if the packets needed * to be fragmented. * * (XXX <= 4 ?) */ if ((so->so_state & SS_ISFCONNECTED) && so->so_queued <= 4) { STAM_COUNTER_INC(&pData->StatUDPHot); UDP_ENGAGE_EVENT(so, readfds); } LOOP_LABEL(udp, so, so_next); } done: #if defined(RT_OS_WINDOWS) *pnfds = VBOX_EVENT_COUNT; #else /* RT_OS_WINDOWS */ AssertRelease(poll_index <= *pnfds); *pnfds = poll_index; #endif /* !RT_OS_WINDOWS */ STAM_PROFILE_STOP(&pData->StatFill, a); } /** * This function do Connection or sending tcp sequence to. * @returns if true operation completed * @note: functions call tcp_input that potentially could lead to tcp_drop */ static bool slirpConnectOrWrite(PNATState pData, struct socket *so, bool fConnectOnly) { int ret; LogFlowFunc(("ENTER: so:%R[natsock], fConnectOnly:%RTbool\n", so, fConnectOnly)); /* * Check for non-blocking, still-connecting sockets */ if (so->so_state & SS_ISFCONNECTING) { Log2(("connecting %R[natsock] catched\n", so)); /* Connected */ so->so_state &= ~SS_ISFCONNECTING; /* * This should be probably guarded by PROBE_CONN too. Anyway, * we disable it on OS/2 because the below send call returns * EFAULT which causes the opened TCP socket to close right * after it has been opened and connected. */ #ifndef RT_OS_OS2 ret = send(so->s, (const char *)&ret, 0, 0); if (ret < 0) { /* XXXXX Must fix, zero bytes is a NOP */ if ( soIgnorableErrorCode(errno) || errno == ENOTCONN) { LogFlowFunc(("LEAVE: false\n")); return false; } /* else failed */ so->so_state = SS_NOFDREF; } /* else so->so_state &= ~SS_ISFCONNECTING; */ #endif /* * Continue tcp_input */ TCP_INPUT(pData, (struct mbuf *)NULL, sizeof(struct ip), so); /* continue; */ } else if (!fConnectOnly) { SOWRITE(ret, pData, so); if (RT_LIKELY(ret > 0)) { /* * Make sure we will send window update to peer. This is * a moral equivalent of calling tcp_output() for PRU_RCVD * in tcp_usrreq() of the real stack. */ struct tcpcb *tp = sototcpcb(so); if (RT_LIKELY(tp != NULL)) tp->t_flags |= TF_DELACK; } } LogFlowFunc(("LEAVE: true\n")); return true; } #if defined(RT_OS_WINDOWS) void slirp_select_poll(PNATState pData, int fTimeout) #else /* RT_OS_WINDOWS */ void slirp_select_poll(PNATState pData, struct pollfd *polls, int ndfs) #endif /* !RT_OS_WINDOWS */ { struct socket *so, *so_next; int ret; #if defined(RT_OS_WINDOWS) WSANETWORKEVENTS NetworkEvents; int rc; int error; #endif STAM_PROFILE_START(&pData->StatPoll, a); /* Update time */ updtime(pData); /* * See if anything has timed out */ if (link_up) { if (time_fasttimo && ((curtime - time_fasttimo) >= 2)) { STAM_PROFILE_START(&pData->StatFastTimer, b); tcp_fasttimo(pData); time_fasttimo = 0; STAM_PROFILE_STOP(&pData->StatFastTimer, b); } if (do_slowtimo && ((curtime - last_slowtimo) >= 499)) { STAM_PROFILE_START(&pData->StatSlowTimer, c); ip_slowtimo(pData); tcp_slowtimo(pData); last_slowtimo = curtime; STAM_PROFILE_STOP(&pData->StatSlowTimer, c); } } #if defined(RT_OS_WINDOWS) if (fTimeout) return; /* only timer update */ #endif /* * Check sockets */ if (!link_up) goto done; #if defined(RT_OS_WINDOWS) icmpwin_process(pData); #else if ( (pData->icmp_socket.s != -1) && CHECK_FD_SET(&pData->icmp_socket, ignored, readfds)) sorecvfrom(pData, &pData->icmp_socket); #endif /* * Check TCP sockets */ QSOCKET_FOREACH(so, so_next, tcp) /* { */ Assert(!so->fUnderPolling); so->fUnderPolling = 1; if (slirpVerifyAndFreeSocket(pData, so)) CONTINUE(tcp); /* * FD_ISSET is meaningless on these sockets * (and they can crash the program) */ if (so->so_state & SS_NOFDREF || so->s == -1) { so->fUnderPolling = 0; CONTINUE(tcp); } POLL_TCP_EVENTS(rc, error, so, &NetworkEvents); LOG_NAT_SOCK(so, TCP, &NetworkEvents, readfds, writefds, xfds); if (so->so_state & SS_ISFCONNECTING) { int sockerr = 0; #if !defined(RT_OS_WINDOWS) { int revents = 0; /* * Failed connect(2) is reported by poll(2) on * different OSes with different combinations of * POLLERR, POLLHUP, and POLLOUT. */ if ( CHECK_FD_SET(so, NetworkEvents, closefds) /* POLLHUP */ || CHECK_FD_SET(so, NetworkEvents, rderr)) /* POLLERR */ { revents = POLLHUP; /* squash to single "failed" flag */ } #if defined(RT_OS_SOLARIS) || defined(RT_OS_NETBSD) /* Solaris and NetBSD report plain POLLOUT even on error */ else if (CHECK_FD_SET(so, NetworkEvents, writefds)) /* POLLOUT */ { revents = POLLOUT; } #endif if (revents != 0) { socklen_t optlen = (socklen_t)sizeof(sockerr); ret = getsockopt(so->s, SOL_SOCKET, SO_ERROR, &sockerr, &optlen); if ( RT_UNLIKELY(ret < 0) || ( (revents & POLLHUP) && RT_UNLIKELY(sockerr == 0))) sockerr = ETIMEDOUT; } } #else /* RT_OS_WINDOWS */ { if (NetworkEvents.lNetworkEvents & FD_CONNECT) sockerr = NetworkEvents.iErrorCode[FD_CONNECT_BIT]; } #endif if (sockerr != 0) { tcp_fconnect_failed(pData, so, sockerr); ret = slirpVerifyAndFreeSocket(pData, so); Assert(ret == 1); /* freed */ CONTINUE(tcp); } /* * XXX: For now just fall through to the old code to * handle successful connect(2). */ } /* * Check for URG data * This will soread as well, so no need to * test for readfds below if this succeeds */ /* out-of-band data */ if ( CHECK_FD_SET(so, NetworkEvents, xfds) #ifdef RT_OS_DARWIN /* Darwin and probably BSD hosts generates POLLPRI|POLLHUP event on receiving TCP.flags.{ACK|URG|FIN} this * combination on other Unixs hosts doesn't enter to this branch */ && !CHECK_FD_SET(so, NetworkEvents, closefds) #endif #ifdef RT_OS_WINDOWS /** * In some cases FD_CLOSE comes with FD_OOB, that confuse tcp processing. */ && !WIN_CHECK_FD_SET(so, NetworkEvents, closefds) #endif ) { sorecvoob(pData, so); if (slirpVerifyAndFreeSocket(pData, so)) CONTINUE(tcp); } /* * Check sockets for reading */ else if ( CHECK_FD_SET(so, NetworkEvents, readfds) || WIN_CHECK_FD_SET(so, NetworkEvents, acceptds)) { #ifdef RT_OS_WINDOWS if (WIN_CHECK_FD_SET(so, NetworkEvents, connectfds)) { /* Finish connection first */ /* should we ignore return value? */ bool fRet = slirpConnectOrWrite(pData, so, true); LogFunc(("fRet:%RTbool\n", fRet)); NOREF(fRet); if (slirpVerifyAndFreeSocket(pData, so)) CONTINUE(tcp); } #endif /* * Check for incoming connections */ if (so->so_state & SS_FACCEPTCONN) { TCP_CONNECT(pData, so); if (slirpVerifyAndFreeSocket(pData, so)) CONTINUE(tcp); if (!CHECK_FD_SET(so, NetworkEvents, closefds)) { so->fUnderPolling = 0; CONTINUE(tcp); } } ret = soread(pData, so); if (slirpVerifyAndFreeSocket(pData, so)) CONTINUE(tcp); /* Output it if we read something */ if (RT_LIKELY(ret > 0)) TCP_OUTPUT(pData, sototcpcb(so)); if (slirpVerifyAndFreeSocket(pData, so)) CONTINUE(tcp); } /* * Check for FD_CLOSE events. * in some cases once FD_CLOSE engaged on socket it could be flashed latter (for some reasons) */ if ( CHECK_FD_SET(so, NetworkEvents, closefds) || (so->so_close == 1)) { /* * drain the socket */ for (; so_next->so_prev == so && !slirpVerifyAndFreeSocket(pData, so);) { ret = soread(pData, so); if (slirpVerifyAndFreeSocket(pData, so)) break; if (ret > 0) TCP_OUTPUT(pData, sototcpcb(so)); else if (so_next->so_prev == so) { Log2(("%R[natsock] errno %d (%s)\n", so, errno, strerror(errno))); break; } } /* if socket freed ''so'' is PHANTOM and next socket isn't points on it */ if (so_next->so_prev != so) { CONTINUE(tcp); } else { /* mark the socket for termination _after_ it was drained */ so->so_close = 1; /* No idea about Windows but on Posix, POLLHUP means that we can't send more. * Actually in the specific error scenario, POLLERR is set as well. */ #ifndef RT_OS_WINDOWS if (CHECK_FD_SET(so, NetworkEvents, rderr)) sofcantsendmore(so); #endif } } /* * Check sockets for writing */ if ( CHECK_FD_SET(so, NetworkEvents, writefds) #ifdef RT_OS_WINDOWS || WIN_CHECK_FD_SET(so, NetworkEvents, connectfds) #endif ) { int fConnectOrWriteSuccess = slirpConnectOrWrite(pData, so, false); /* slirpConnectOrWrite could return true even if tcp_input called tcp_drop, * so we should be ready to such situations. */ if (slirpVerifyAndFreeSocket(pData, so)) CONTINUE(tcp); else if (!fConnectOrWriteSuccess) { so->fUnderPolling = 0; CONTINUE(tcp); } /* slirpConnectionOrWrite succeeded and socket wasn't dropped */ } /* * Probe a still-connecting, non-blocking socket * to check if it's still alive */ #ifdef PROBE_CONN if (so->so_state & SS_ISFCONNECTING) { ret = recv(so->s, (char *)&ret, 0, 0); if (ret < 0) { /* XXX */ if ( soIgnorableErrorCode(errno) || errno == ENOTCONN) { CONTINUE(tcp); /* Still connecting, continue */ } /* else failed */ so->so_state = SS_NOFDREF; /* tcp_input will take care of it */ } else { ret = send(so->s, &ret, 0, 0); if (ret < 0) { /* XXX */ if ( soIgnorableErrorCode(errno) || errno == ENOTCONN) { CONTINUE(tcp); } /* else failed */ so->so_state = SS_NOFDREF; } else so->so_state &= ~SS_ISFCONNECTING; } TCP_INPUT((struct mbuf *)NULL, sizeof(struct ip),so); } /* SS_ISFCONNECTING */ #endif if (!slirpVerifyAndFreeSocket(pData, so)) so->fUnderPolling = 0; LOOP_LABEL(tcp, so, so_next); } /* * Now UDP sockets. * Incoming packets are sent straight away, they're not buffered. * Incoming UDP data isn't buffered either. */ QSOCKET_FOREACH(so, so_next, udp) /* { */ #if 0 so->fUnderPolling = 1; if(slirpVerifyAndFreeSocket(pData, so)); CONTINUE(udp); so->fUnderPolling = 0; #endif POLL_UDP_EVENTS(rc, error, so, &NetworkEvents); LOG_NAT_SOCK(so, UDP, &NetworkEvents, readfds, writefds, xfds); if (so->s != -1 && CHECK_FD_SET(so, NetworkEvents, readfds)) { SORECVFROM(pData, so); } LOOP_LABEL(udp, so, so_next); } done: STAM_PROFILE_STOP(&pData->StatPoll, a); } struct arphdr { unsigned short ar_hrd; /* format of hardware address */ #define ARPHRD_ETHER 1 /* ethernet hardware format */ unsigned short ar_pro; /* format of protocol address */ unsigned char ar_hln; /* length of hardware address */ unsigned char ar_pln; /* length of protocol address */ unsigned short ar_op; /* ARP opcode (command) */ #define ARPOP_REQUEST 1 /* ARP request */ #define ARPOP_REPLY 2 /* ARP reply */ /* * Ethernet looks like this : This bit is variable sized however... */ unsigned char ar_sha[ETH_ALEN]; /* sender hardware address */ unsigned char ar_sip[4]; /* sender IP address */ unsigned char ar_tha[ETH_ALEN]; /* target hardware address */ unsigned char ar_tip[4]; /* target IP address */ }; AssertCompileSize(struct arphdr, 28); static void arp_output(PNATState pData, const uint8_t *pcu8EtherSource, const struct arphdr *pcARPHeaderSource, uint32_t ip4TargetAddress) { struct ethhdr *pEtherHeaderResponse; struct arphdr *pARPHeaderResponse; uint32_t ip4TargetAddressInHostFormat; struct mbuf *pMbufResponse; Assert((pcu8EtherSource)); if (!pcu8EtherSource) return; ip4TargetAddressInHostFormat = RT_N2H_U32(ip4TargetAddress); pMbufResponse = m_getcl(pData, M_NOWAIT, MT_HEADER, M_PKTHDR); if (!pMbufResponse) return; pEtherHeaderResponse = mtod(pMbufResponse, struct ethhdr *); /* @note: if_encap will swap src and dst*/ memcpy(pEtherHeaderResponse->h_source, pcu8EtherSource, ETH_ALEN); pMbufResponse->m_data += ETH_HLEN; pARPHeaderResponse = mtod(pMbufResponse, struct arphdr *); pMbufResponse->m_len = sizeof(struct arphdr); pARPHeaderResponse->ar_hrd = RT_H2N_U16_C(1); pARPHeaderResponse->ar_pro = RT_H2N_U16_C(ETH_P_IP); pARPHeaderResponse->ar_hln = ETH_ALEN; pARPHeaderResponse->ar_pln = 4; pARPHeaderResponse->ar_op = RT_H2N_U16_C(ARPOP_REPLY); memcpy(pARPHeaderResponse->ar_sha, special_ethaddr, ETH_ALEN); if (!slirpMbufTagService(pData, pMbufResponse, (uint8_t)(ip4TargetAddressInHostFormat & ~pData->netmask))) { static bool fTagErrorReported; if (!fTagErrorReported) { LogRel(("NAT: Couldn't add the tag(PACKET_SERVICE:%d)\n", (uint8_t)(ip4TargetAddressInHostFormat & ~pData->netmask))); fTagErrorReported = true; } } pARPHeaderResponse->ar_sha[5] = (uint8_t)(ip4TargetAddressInHostFormat & ~pData->netmask); memcpy(pARPHeaderResponse->ar_sip, pcARPHeaderSource->ar_tip, 4); memcpy(pARPHeaderResponse->ar_tha, pcARPHeaderSource->ar_sha, ETH_ALEN); memcpy(pARPHeaderResponse->ar_tip, pcARPHeaderSource->ar_sip, 4); if_encap(pData, ETH_P_ARP, pMbufResponse, ETH_ENCAP_URG); } /** * @note This function will free m! */ static void arp_input(PNATState pData, struct mbuf *m) { struct ethhdr *pEtherHeader; struct arphdr *pARPHeader; int ar_op; uint32_t ip4TargetAddress; /* drivers never return runt packets, so this should never happen */ if (RT_UNLIKELY((size_t)m->m_len < sizeof(struct ethhdr) + sizeof(struct arphdr))) goto done; pEtherHeader = mtod(m, struct ethhdr *); pARPHeader = (struct arphdr *)&pEtherHeader[1]; if (RT_UNLIKELY( pARPHeader->ar_hrd != RT_H2N_U16_C(ARPHRD_ETHER) || pARPHeader->ar_pro != RT_H2N_U16_C(ETH_P_IP) || pARPHeader->ar_hln != ETH_ALEN || pARPHeader->ar_pln != sizeof(RTNETADDRIPV4))) goto done; ar_op = RT_N2H_U16(pARPHeader->ar_op); ip4TargetAddress = *(uint32_t*)pARPHeader->ar_tip; switch (ar_op) { case ARPOP_REQUEST: if ( CTL_CHECK(ip4TargetAddress, CTL_DNS) || CTL_CHECK(ip4TargetAddress, CTL_ALIAS) || CTL_CHECK(ip4TargetAddress, CTL_TFTP)) { #if 0 /* Dropping ARP requests destined for CTL_ALIAS breaks all outgoing traffic completely, so don't do that... */ /* Don't reply to ARP requests for the hosts loopback interface if it is disabled. */ if ( CTL_CHECK(ip4TargetAddress, CTL_ALIAS) && !pData->fLocalhostReachable) break; #endif slirp_update_guest_addr_guess(pData, *(uint32_t *)pARPHeader->ar_sip, "arp request"); arp_output(pData, pEtherHeader->h_source, pARPHeader, ip4TargetAddress); break; } /* Gratuitous ARP */ if ( *(uint32_t *)pARPHeader->ar_sip == *(uint32_t *)pARPHeader->ar_tip && ( memcmp(pARPHeader->ar_tha, zerro_ethaddr, ETH_ALEN) == 0 || memcmp(pARPHeader->ar_tha, broadcast_ethaddr, ETH_ALEN) == 0) && memcmp(pEtherHeader->h_dest, broadcast_ethaddr, ETH_ALEN) == 0) { LogRel2(("NAT: Gratuitous ARP from %RTnaipv4 at %RTmac\n", *(uint32_t *)pARPHeader->ar_sip, pARPHeader->ar_sha)); slirp_update_guest_addr_guess(pData, *(uint32_t *)pARPHeader->ar_sip, "gratuitous arp"); slirp_arp_cache_update_or_add(pData, *(uint32_t *)pARPHeader->ar_sip, &pARPHeader->ar_sha[0]); } break; case ARPOP_REPLY: slirp_arp_cache_update_or_add(pData, *(uint32_t *)pARPHeader->ar_sip, &pARPHeader->ar_sha[0]); break; default: break; } done: m_freem(pData, m); } /** * Feed a packet into the slirp engine. * * @param m Data buffer, m_len is not valid. * @param cbBuf The length of the data in m. */ void slirp_input(PNATState pData, struct mbuf *m, size_t cbBuf) { int proto; static bool fWarnedIpv6; struct ethhdr *eh; m->m_len = (int)cbBuf; Assert((size_t)m->m_len == cbBuf); if (cbBuf < ETH_HLEN) { Log(("NAT: packet having size %d has been ignored\n", m->m_len)); m_freem(pData, m); return; } eh = mtod(m, struct ethhdr *); proto = RT_N2H_U16(eh->h_proto); switch(proto) { case ETH_P_ARP: arp_input(pData, m); break; case ETH_P_IP: /* Update time. Important if the network is very quiet, as otherwise * the first outgoing connection gets an incorrect timestamp. */ updtime(pData); m_adj(m, ETH_HLEN); M_ASSERTPKTHDR(m); m->m_pkthdr.header = mtod(m, void *); ip_input(pData, m); break; case ETH_P_IPV6: m_freem(pData, m); if (!fWarnedIpv6) { LogRel(("NAT: IPv6 not supported\n")); fWarnedIpv6 = true; } break; default: Log(("NAT: Unsupported protocol %x\n", proto)); m_freem(pData, m); break; } } /** * Output the IP packet to the ethernet device. * * @note This function will free m! */ void if_encap(PNATState pData, uint16_t eth_proto, struct mbuf *m, int flags) { struct ethhdr *eh; uint8_t *mbuf = NULL; int mlen; STAM_PROFILE_START(&pData->StatIF_encap, a); LogFlowFunc(("ENTER: pData:%p, eth_proto:%RX16, m:%p, flags:%d\n", pData, eth_proto, m, flags)); M_ASSERTPKTHDR(m); Assert(M_LEADINGSPACE(m) >= ETH_HLEN); m->m_data -= ETH_HLEN; m->m_len += ETH_HLEN; eh = mtod(m, struct ethhdr *); mlen = m->m_len; if (memcmp(eh->h_source, special_ethaddr, ETH_ALEN) != 0) { struct m_tag *t = m_tag_first(m); uint8_t u8ServiceId = CTL_ALIAS; memcpy(eh->h_dest, eh->h_source, ETH_ALEN); memcpy(eh->h_source, special_ethaddr, ETH_ALEN); Assert(memcmp(eh->h_dest, special_ethaddr, ETH_ALEN) != 0); if (memcmp(eh->h_dest, zerro_ethaddr, ETH_ALEN) == 0) { /* don't do anything */ m_freem(pData, m); goto done; } if ( t && (t = m_tag_find(m, PACKET_SERVICE, NULL))) { Assert(t); u8ServiceId = *(uint8_t *)&t[1]; } eh->h_source[5] = u8ServiceId; } /* * we're processing the chain, that isn't not expected. */ Assert((!m->m_next)); if (m->m_next) { Log(("NAT: if_encap's recived the chain, dropping...\n")); m_freem(pData, m); goto done; } mbuf = mtod(m, uint8_t *); eh->h_proto = RT_H2N_U16(eth_proto); LogFunc(("eh(dst:%RTmac, src:%RTmac)\n", eh->h_dest, eh->h_source)); if (flags & ETH_ENCAP_URG) slirp_urg_output(pData->pvUser, m, mbuf, mlen); else slirp_output(pData->pvUser, m, mbuf, mlen); done: STAM_PROFILE_STOP(&pData->StatIF_encap, a); LogFlowFuncLeave(); } void slirp_update_guest_addr_guess(PNATState pData, uint32_t guess, const char *msg) { Assert(msg != NULL); if (pData->guest_addr_guess.s_addr == guess) { LogRel2(("NAT: Guest address guess %RTnaipv4 re-confirmed by %s\n", pData->guest_addr_guess.s_addr, msg)); return; } if (pData->guest_addr_guess.s_addr == INADDR_ANY) { pData->guest_addr_guess.s_addr = guess; LogRel(("NAT: Guest address guess set to %RTnaipv4 by %s\n", pData->guest_addr_guess.s_addr, msg)); return; } else { LogRel(("NAT: Guest address guess changed from %RTnaipv4 to %RTnaipv4 by %s\n", pData->guest_addr_guess.s_addr, guess, msg)); pData->guest_addr_guess.s_addr = guess; return; } } static struct port_forward_rule * slirp_find_redirect(PNATState pData, int is_udp, struct in_addr host_addr, int host_port, struct in_addr guest_addr, int guest_port) { struct port_forward_rule *rule; uint16_t proto = (is_udp ? IPPROTO_UDP : IPPROTO_TCP); LIST_FOREACH(rule, &pData->port_forward_rule_head, list) { if ( rule->proto == proto && rule->host_port == host_port && rule->bind_ip.s_addr == host_addr.s_addr && rule->guest_port == guest_port && rule->guest_addr.s_addr == guest_addr.s_addr) { return rule; } } return NULL; } int slirp_add_redirect(PNATState pData, int is_udp, struct in_addr host_addr, int host_port, struct in_addr guest_addr, int guest_port) { struct port_forward_rule *rule; rule = slirp_find_redirect(pData, is_udp, host_addr, host_port, guest_addr, guest_port); if (rule != NULL) /* rule has been already registered */ { /* XXX: this shouldn't happen */ return 0; } rule = RTMemAllocZ(sizeof(struct port_forward_rule)); if (rule == NULL) return 1; rule->proto = (is_udp ? IPPROTO_UDP : IPPROTO_TCP); rule->bind_ip.s_addr = host_addr.s_addr; rule->host_port = host_port; rule->guest_addr.s_addr = guest_addr.s_addr; rule->guest_port = guest_port; if (rule->proto == IPPROTO_UDP) rule->so = udp_listen(pData, rule->bind_ip.s_addr, RT_H2N_U16(rule->host_port), rule->guest_addr.s_addr, RT_H2N_U16(rule->guest_port), 0); else rule->so = solisten(pData, rule->bind_ip.s_addr, RT_H2N_U16(rule->host_port), rule->guest_addr.s_addr, RT_H2N_U16(rule->guest_port), 0); if (rule->so == NULL) { LogRel(("NAT: Failed to redirect %s %RTnaipv4:%d -> %RTnaipv4:%d (%s)\n", rule->proto == IPPROTO_UDP ? "UDP" : "TCP", rule->bind_ip.s_addr, rule->host_port, guest_addr, rule->guest_port, strerror(errno))); RTMemFree(rule); return 1; } LogRel(("NAT: Set redirect %s %RTnaipv4:%d -> %RTnaipv4:%d\n", rule->proto == IPPROTO_UDP ? "UDP" : "TCP", rule->bind_ip.s_addr, rule->host_port, guest_addr, rule->guest_port)); LIST_INSERT_HEAD(&pData->port_forward_rule_head, rule, list); return 0; } int slirp_remove_redirect(PNATState pData, int is_udp, struct in_addr host_addr, int host_port, struct in_addr guest_addr, int guest_port) { struct port_forward_rule *rule; rule = slirp_find_redirect(pData, is_udp, host_addr, host_port, guest_addr, guest_port); if (rule == NULL) { LogRel(("NAT: Unable to find redirect %s %RTnaipv4:%d -> %RTnaipv4:%d\n", is_udp ? "UDP" : "TCP", host_addr.s_addr, host_port, guest_addr.s_addr, guest_port)); return 0; } LogRel(("NAT: Remove redirect %s %RTnaipv4:%d -> %RTnaipv4:%d\n", rule->proto == IPPROTO_UDP ? "UDP" : "TCP", rule->bind_ip.s_addr, rule->host_port, guest_addr.s_addr, rule->guest_port)); if (rule->so != NULL) { if (is_udp) udp_detach(pData, rule->so); else tcp_close(pData, sototcpcb(rule->so)); } LIST_REMOVE(rule, list); RTMemFree(rule); return 0; } #if defined(RT_OS_WINDOWS) HANDLE *slirp_get_events(PNATState pData) { return pData->phEvents; } void slirp_register_external_event(PNATState pData, HANDLE hEvent, int index) { pData->phEvents[index] = hEvent; } #endif unsigned int slirp_get_timeout_ms(PNATState pData) { if (link_up) { if (time_fasttimo) return 2; if (do_slowtimo) return 500; /* see PR_SLOWHZ */ } return 3600*1000; /* one hour */ } #ifndef RT_OS_WINDOWS int slirp_get_nsock(PNATState pData) { return pData->nsock; } #endif /* * this function called from NAT thread */ void slirp_post_sent(PNATState pData, void *pvArg) { struct mbuf *m = (struct mbuf *)pvArg; m_freem(pData, m); } void slirp_set_dhcp_TFTP_prefix(PNATState pData, const char *tftpPrefix) { Log2(("tftp_prefix: %s\n", tftpPrefix)); if (tftp_prefix) RTStrFree((char *)tftp_prefix); tftp_prefix = RTPathAbsDup(tftpPrefix); } void slirp_set_dhcp_TFTP_bootfile(PNATState pData, const char *bootFile) { Log2(("bootFile: %s\n", bootFile)); bootp_filename = bootFile; } void slirp_set_dhcp_next_server(PNATState pData, const char *next_server) { Log2(("next_server: %s\n", next_server)); if (next_server == NULL) pData->tftp_server.s_addr = RT_H2N_U32(RT_N2H_U32(pData->special_addr.s_addr) | CTL_TFTP); else inet_aton(next_server, &pData->tftp_server); } int slirp_set_binding_address(PNATState pData, char *addr) { int ok; pData->bindIP.s_addr = INADDR_ANY; if (addr == NULL || *addr == '\0') return VINF_SUCCESS; ok = inet_aton(addr, &pData->bindIP); if (!ok) { LogRel(("NAT: Unable to parse binding address: %s\n", addr)); return VERR_INVALID_PARAMETER; } if (pData->bindIP.s_addr == INADDR_ANY) return VINF_SUCCESS; if ((pData->bindIP.s_addr & RT_N2H_U32_C(0xe0000000)) == RT_N2H_U32_C(0xe0000000)) { LogRel(("NAT: Ignoring multicast binding address %RTnaipv4\n", pData->bindIP.s_addr)); pData->bindIP.s_addr = INADDR_ANY; return VERR_INVALID_PARAMETER; } LogRel(("NAT: Binding address %RTnaipv4\n", pData->bindIP.s_addr)); return VINF_SUCCESS; } void slirp_set_dhcp_dns_proxy(PNATState pData, bool fDNSProxy) { if (!pData->fUseHostResolver) { Log2(("NAT: DNS proxy switched %s\n", (fDNSProxy ? "on" : "off"))); pData->fUseDnsProxy = fDNSProxy; } else if (fDNSProxy) LogRel(("NAT: Host Resolver conflicts with DNS proxy, the last one was forcely ignored\n")); } #define CHECK_ARG(name, val, lim_min, lim_max) \ do { \ if ((val) < (lim_min) || (val) > (lim_max)) \ { \ LogRel(("NAT: (" #name ":%d) has been ignored, " \ "because out of range (%d, %d)\n", (val), (lim_min), (lim_max))); \ return; \ } \ else \ LogRel(("NAT: (" #name ":%d)\n", (val))); \ } while (0) void slirp_set_somaxconn(PNATState pData, int iSoMaxConn) { LogFlowFunc(("iSoMaxConn:%d\n", iSoMaxConn)); /* Conditions */ if (iSoMaxConn > SOMAXCONN) { LogRel(("NAT: value of somaxconn(%d) bigger than SOMAXCONN(%d)\n", iSoMaxConn, SOMAXCONN)); iSoMaxConn = SOMAXCONN; } if (iSoMaxConn < 1) { LogRel(("NAT: proposed value(%d) of somaxconn is invalid, default value is used (%d)\n", iSoMaxConn, pData->soMaxConn)); LogFlowFuncLeave(); return; } /* Asignment */ if (pData->soMaxConn != iSoMaxConn) { LogRel(("NAT: value of somaxconn has been changed from %d to %d\n", pData->soMaxConn, iSoMaxConn)); pData->soMaxConn = iSoMaxConn; } LogFlowFuncLeave(); } /* don't allow user set less 8kB and more than 1M values */ #define _8K_1M_CHECK_ARG(name, val) CHECK_ARG(name, (val), 8, 1024) void slirp_set_rcvbuf(PNATState pData, int kilobytes) { _8K_1M_CHECK_ARG("SOCKET_RCVBUF", kilobytes); pData->socket_rcv = kilobytes; } void slirp_set_sndbuf(PNATState pData, int kilobytes) { _8K_1M_CHECK_ARG("SOCKET_SNDBUF", kilobytes); pData->socket_snd = kilobytes * _1K; } void slirp_set_tcp_rcvspace(PNATState pData, int kilobytes) { _8K_1M_CHECK_ARG("TCP_RCVSPACE", kilobytes); tcp_rcvspace = kilobytes * _1K; } void slirp_set_tcp_sndspace(PNATState pData, int kilobytes) { _8K_1M_CHECK_ARG("TCP_SNDSPACE", kilobytes); tcp_sndspace = kilobytes * _1K; } /* * Looking for Ether by ip in ARP-cache * Note: it´s responsible of caller to allocate buffer for result * @returns iprt status code */ int slirp_arp_lookup_ether_by_ip(PNATState pData, uint32_t ip, uint8_t *ether) { struct arp_cache_entry *ac; if (ether == NULL) return VERR_INVALID_PARAMETER; if (LIST_EMPTY(&pData->arp_cache)) return VERR_NOT_FOUND; LIST_FOREACH(ac, &pData->arp_cache, list) { if ( ac->ip == ip && memcmp(ac->ether, broadcast_ethaddr, ETH_ALEN) != 0) { memcpy(ether, ac->ether, ETH_ALEN); return VINF_SUCCESS; } } return VERR_NOT_FOUND; } /* * Looking for IP by Ether in ARP-cache * Note: it´s responsible of caller to allocate buffer for result * @returns 0 - if found, 1 - otherwise */ int slirp_arp_lookup_ip_by_ether(PNATState pData, const uint8_t *ether, uint32_t *ip) { struct arp_cache_entry *ac; *ip = INADDR_ANY; if (LIST_EMPTY(&pData->arp_cache)) return VERR_NOT_FOUND; LIST_FOREACH(ac, &pData->arp_cache, list) { if (memcmp(ether, ac->ether, ETH_ALEN) == 0) { *ip = ac->ip; return VINF_SUCCESS; } } return VERR_NOT_FOUND; } void slirp_arp_who_has(PNATState pData, uint32_t dst) { struct mbuf *m; struct ethhdr *ehdr; struct arphdr *ahdr; static bool fWarned = false; LogFlowFunc(("ENTER: %RTnaipv4\n", dst)); /* ARP request WHO HAS 0.0.0.0 is one of the signals * that something has been broken at Slirp. Investigating * pcap dumps it's easy to miss warning ARP requests being * focused on investigation of other protocols flow. */ #ifdef DEBUG_vvl Assert((dst != INADDR_ANY)); NOREF(fWarned); #else if ( dst == INADDR_ANY && !fWarned) { LogRel(("NAT: ARP: \"WHO HAS INADDR_ANY\" request has been detected\n")); fWarned = true; } #endif /* !DEBUG_vvl */ m = m_getcl(pData, M_NOWAIT, MT_HEADER, M_PKTHDR); if (m == NULL) { Log(("NAT: Can't alloc mbuf for ARP request\n")); LogFlowFuncLeave(); return; } ehdr = mtod(m, struct ethhdr *); memset(ehdr->h_source, 0xff, ETH_ALEN); ahdr = (struct arphdr *)&ehdr[1]; ahdr->ar_hrd = RT_H2N_U16_C(1); ahdr->ar_pro = RT_H2N_U16_C(ETH_P_IP); ahdr->ar_hln = ETH_ALEN; ahdr->ar_pln = 4; ahdr->ar_op = RT_H2N_U16_C(ARPOP_REQUEST); memcpy(ahdr->ar_sha, special_ethaddr, ETH_ALEN); /* we assume that this request come from gw, but not from DNS or TFTP */ ahdr->ar_sha[5] = CTL_ALIAS; *(uint32_t *)ahdr->ar_sip = RT_H2N_U32(RT_N2H_U32(pData->special_addr.s_addr) | CTL_ALIAS); memset(ahdr->ar_tha, 0xff, ETH_ALEN); /*broadcast*/ *(uint32_t *)ahdr->ar_tip = dst; /* warn!!! should falls in mbuf minimal size */ m->m_len = sizeof(struct arphdr) + ETH_HLEN; m->m_data += ETH_HLEN; m->m_len -= ETH_HLEN; if_encap(pData, ETH_P_ARP, m, ETH_ENCAP_URG); LogFlowFuncLeave(); } /* updates the arp cache * @note: this is helper function, slirp_arp_cache_update_or_add should be used. * @returns 0 - if has found and updated * 1 - if hasn't found. */ static inline int slirp_arp_cache_update(PNATState pData, uint32_t dst, const uint8_t *mac) { struct arp_cache_entry *ac; Assert(( memcmp(mac, broadcast_ethaddr, ETH_ALEN) && memcmp(mac, zerro_ethaddr, ETH_ALEN))); LIST_FOREACH(ac, &pData->arp_cache, list) { if (ac->ip == dst) { memcpy(ac->ether, mac, ETH_ALEN); return 0; } } return 1; } /** * add entry to the arp cache * @note: this is helper function, slirp_arp_cache_update_or_add should be used. */ static inline void slirp_arp_cache_add(PNATState pData, uint32_t ip, const uint8_t *ether) { struct arp_cache_entry *ac = NULL; Assert(( memcmp(ether, broadcast_ethaddr, ETH_ALEN) && memcmp(ether, zerro_ethaddr, ETH_ALEN))); ac = RTMemAllocZ(sizeof(struct arp_cache_entry)); if (ac == NULL) { Log(("NAT: Can't allocate arp cache entry\n")); return; } ac->ip = ip; memcpy(ac->ether, ether, ETH_ALEN); LIST_INSERT_HEAD(&pData->arp_cache, ac, list); } /* updates or adds entry to the arp cache * @returns 0 - if has found and updated * 1 - if hasn't found. */ int slirp_arp_cache_update_or_add(PNATState pData, uint32_t dst, const uint8_t *mac) { if ( !memcmp(mac, broadcast_ethaddr, ETH_ALEN) || !memcmp(mac, zerro_ethaddr, ETH_ALEN)) { static bool fBroadcastEtherAddReported; if (!fBroadcastEtherAddReported) { LogRel(("NAT: Attempt to add pair [%RTmac:%RTnaipv4] in ARP cache was ignored\n", mac, dst)); fBroadcastEtherAddReported = true; } return 1; } if (slirp_arp_cache_update(pData, dst, mac)) slirp_arp_cache_add(pData, dst, mac); return 0; } void slirp_set_mtu(PNATState pData, int mtu) { if (mtu < 20 || mtu >= 16000) { LogRel(("NAT: MTU(%d) is out of range (20;16000] mtu forcely assigned to 1500\n", mtu)); mtu = 1500; } /* MTU is maximum transition unit on */ if_mtu = if_mru = mtu; } /** * Info handler. */ void slirp_info(PNATState pData, const void *pvArg, const char *pszArgs) { struct socket *so, *so_next; struct arp_cache_entry *ac; struct port_forward_rule *rule; PCDBGFINFOHLP pHlp = (PCDBGFINFOHLP)pvArg; NOREF(pszArgs); pHlp->pfnPrintf(pHlp, "NAT parameters: MTU=%d\n", if_mtu); pHlp->pfnPrintf(pHlp, "NAT TCP ports:\n"); QSOCKET_FOREACH(so, so_next, tcp) /* { */ pHlp->pfnPrintf(pHlp, " %R[natsock]\n", so); } pHlp->pfnPrintf(pHlp, "NAT UDP ports:\n"); QSOCKET_FOREACH(so, so_next, udp) /* { */ pHlp->pfnPrintf(pHlp, " %R[natsock]\n", so); } pHlp->pfnPrintf(pHlp, "NAT ARP cache:\n"); LIST_FOREACH(ac, &pData->arp_cache, list) { pHlp->pfnPrintf(pHlp, " %RTnaipv4 %RTmac\n", ac->ip, &ac->ether); } pHlp->pfnPrintf(pHlp, "NAT rules:\n"); LIST_FOREACH(rule, &pData->port_forward_rule_head, list) { pHlp->pfnPrintf(pHlp, " %s %d => %RTnaipv4:%d %c\n", rule->proto == IPPROTO_UDP ? "UDP" : "TCP", rule->host_port, rule->guest_addr.s_addr, rule->guest_port, rule->activated ? ' ' : '*'); } } /** * @note: NATState::fUseHostResolver could be changed in bootp.c::dhcp_decode * @note: this function is executed on GUI/VirtualBox or main/VBoxHeadless thread. * @note: this function can potentially race with bootp.c::dhcp_decode (except Darwin) */ int slirp_host_network_configuration_change_strategy_selector(const PNATState pData) { if (pData->fUseHostResolverPermanent) return VBOX_NAT_DNS_HOSTRESOLVER; if (pData->fUseDnsProxy) { #if HAVE_NOTIFICATION_FOR_DNS_UPDATE /* XXX */ && !defined(RT_OS_WINDOWS) /* We dont conflict with bootp.c::dhcp_decode */ struct rcp_state rcp_state; int rc; rcp_state.rcps_flags = RCPSF_IGNORE_IPV6; rc = rcp_parse(&rcp_state, RESOLV_CONF_FILE); LogRelFunc(("NAT: rcp_parse:%Rrc old domain:%s new domain:%s\n", rc, LIST_EMPTY(&pData->pDomainList) ? "(null)" : LIST_FIRST(&pData->pDomainList)->dd_pszDomain, rcp_state.rcps_domain)); if ( RT_FAILURE(rc) || LIST_EMPTY(&pData->pDomainList)) return VBOX_NAT_DNS_DNSPROXY; if ( rcp_state.rcps_domain && strcmp(rcp_state.rcps_domain, LIST_FIRST(&pData->pDomainList)->dd_pszDomain) == 0) return VBOX_NAT_DNS_DNSPROXY; else return VBOX_NAT_DNS_EXTERNAL; #else /* copy domain name */ /* domain only compare with coy version */ return VBOX_NAT_DNS_DNSPROXY; #endif } return VBOX_NAT_DNS_EXTERNAL; }