/* vi: set sw=4 ts=4: */ /* * Simple telnet server * Bjorn Wesen, Axis Communications AB (bjornw@axis.com) * * Licensed under GPLv2 or later, see file LICENSE in this source tree. * * --------------------------------------------------------------------------- * (C) Copyright 2000, Axis Communications AB, LUND, SWEDEN **************************************************************************** * * The telnetd manpage says it all: * * Telnetd operates by allocating a pseudo-terminal device (see pty(4)) for * a client, then creating a login process which has the slave side of the * pseudo-terminal as stdin, stdout, and stderr. Telnetd manipulates the * master side of the pseudo-terminal, implementing the telnet protocol and * passing characters between the remote client and the login process. * * Vladimir Oleynik 2001 * Set process group corrections, initial busybox port */ //config:config TELNETD //config: bool "telnetd (12 kb)" //config: default y //config: select FEATURE_SYSLOG //config: help //config: A daemon for the TELNET protocol, allowing you to log onto the host //config: running the daemon. Please keep in mind that the TELNET protocol //config: sends passwords in plain text. If you can't afford the space for an //config: SSH daemon and you trust your network, you may say 'y' here. As a //config: more secure alternative, you should seriously consider installing the //config: very small Dropbear SSH daemon instead: //config: http://matt.ucc.asn.au/dropbear/dropbear.html //config: //config: Note that for busybox telnetd to work you need several things: //config: First of all, your kernel needs: //config: CONFIG_UNIX98_PTYS=y //config: //config: Next, you need a /dev/pts directory on your root filesystem: //config: //config: $ ls -ld /dev/pts //config: drwxr-xr-x 2 root root 0 Sep 23 13:21 /dev/pts/ //config: //config: Next you need the pseudo terminal master multiplexer /dev/ptmx: //config: //config: $ ls -la /dev/ptmx //config: crw-rw-rw- 1 root tty 5, 2 Sep 23 13:55 /dev/ptmx //config: //config: Any /dev/ttyp[0-9]* files you may have can be removed. //config: Next, you need to mount the devpts filesystem on /dev/pts using: //config: //config: mount -t devpts devpts /dev/pts //config: //config: You need to be sure that busybox has LOGIN and //config: FEATURE_SUID enabled. And finally, you should make //config: certain that busybox has been installed setuid root: //config: //config: chown root.root /bin/busybox //config: chmod 4755 /bin/busybox //config: //config: with all that done, telnetd _should_ work.... //config: //config:config FEATURE_TELNETD_STANDALONE //config: bool "Support standalone telnetd (not inetd only)" //config: default y //config: depends on TELNETD //config: help //config: Selecting this will make telnetd able to run standalone. //config: //config:config FEATURE_TELNETD_PORT_DEFAULT //config: int "Default port" //config: default 23 //config: range 1 65535 //config: depends on FEATURE_TELNETD_STANDALONE //config: //config:config FEATURE_TELNETD_INETD_WAIT //config: bool "Support -w SEC option (inetd wait mode)" //config: default y //config: depends on FEATURE_TELNETD_STANDALONE //config: help //config: This option allows you to run telnetd in "inet wait" mode. //config: Example inetd.conf line (note "wait", not usual "nowait"): //config: //config: telnet stream tcp wait root /bin/telnetd telnetd -w10 //config: //config: In this example, inetd passes _listening_ socket_ as fd 0 //config: to telnetd when connection appears. //config: telnetd will wait for connections until all existing //config: connections are closed, and no new connections //config: appear during 10 seconds. Then it exits, and inetd continues //config: to listen for new connections. //config: //config: This option is rarely used. "tcp nowait" is much more usual //config: way of running tcp services, including telnetd. //config: You most probably want to say N here. //applet:IF_TELNETD(APPLET(telnetd, BB_DIR_USR_SBIN, BB_SUID_DROP)) //kbuild:lib-$(CONFIG_TELNETD) += telnetd.o //usage:#define telnetd_trivial_usage //usage: "[OPTIONS]" //usage:#define telnetd_full_usage "\n\n" //usage: "Handle incoming telnet connections" //usage: IF_NOT_FEATURE_TELNETD_STANDALONE(" via inetd") "\n" //usage: "\n -l LOGIN Exec LOGIN on connect" //usage: "\n -f ISSUE_FILE Display ISSUE_FILE instead of /etc/issue" //usage: "\n -K Close connection as soon as login exits" //usage: "\n (normally wait until all programs close slave pty)" //usage: IF_FEATURE_TELNETD_STANDALONE( //usage: "\n -p PORT Port to listen on. Default "STR(CONFIG_FEATURE_TELNETD_PORT_DEFAULT) //usage: "\n -b ADDR[:PORT] Address to bind to" //usage: "\n -F Run in foreground" //usage: "\n -i Inetd mode" //usage: IF_FEATURE_TELNETD_INETD_WAIT( //usage: "\n -w SEC Inetd 'wait' mode, linger time SEC" //usage: "\n inetd.conf line: 23 stream tcp wait root telnetd telnetd -w10" //usage: "\n -S Log to syslog (implied by -i or without -F and -w)" //usage: ) //usage: ) #define DEBUG 0 #include "libbb.h" #include "common_bufsiz.h" #include #if DEBUG # define TELCMDS # define TELOPTS #endif #include struct tsession { struct tsession *next; pid_t shell_pid; int sockfd_read; int sockfd_write; int ptyfd; smallint buffered_IAC_for_pty; /* two circular buffers */ /*char *buf1, *buf2;*/ /*#define TS_BUF1(ts) ts->buf1*/ /*#define TS_BUF2(ts) TS_BUF2(ts)*/ #define TS_BUF1(ts) ((unsigned char*)(ts + 1)) #define TS_BUF2(ts) (((unsigned char*)(ts + 1)) + BUFSIZE) int rdidx1, wridx1, size1; int rdidx2, wridx2, size2; }; /* Two buffers are directly after tsession in malloced memory. * Make whole thing fit in 4k */ enum { BUFSIZE = (4 * 1024 - sizeof(struct tsession)) / 2 }; /* Globals */ struct globals { struct tsession *sessions; const char *loginpath; const char *issuefile; int maxfd; } FIX_ALIASING; #define G (*(struct globals*)bb_common_bufsiz1) #define INIT_G() do { \ setup_common_bufsiz(); \ G.loginpath = "/bin/login"; \ G.issuefile = "/etc/issue.net"; \ } while (0) /* Write some buf1 data to pty, processing IACs. * Update wridx1 and size1. Return < 0 on error. * Buggy if IAC is present but incomplete: skips them. */ static ssize_t safe_write_to_pty_decode_iac(struct tsession *ts) { unsigned wr; ssize_t rc; unsigned char *buf; unsigned char *found; buf = TS_BUF1(ts) + ts->wridx1; wr = MIN(BUFSIZE - ts->wridx1, ts->size1); /* wr is at least 1 here */ if (ts->buffered_IAC_for_pty) { /* Last time we stopped on a "dangling" IAC byte. * We removed it from the buffer back then. * Now pretend it's still there, and jump to IAC processing. */ ts->buffered_IAC_for_pty = 0; wr++; ts->size1++; buf--; /* Yes, this can point before the buffer. It's ok */ ts->wridx1--; goto handle_iac; } found = memchr(buf, IAC, wr); if (found != buf) { /* There is a "prefix" of non-IAC chars. * Write only them, and return. */ if (found) wr = found - buf; /* We map \r\n ==> \r for pragmatic reasons: * many client implementations send \r\n when * the user hits the CarriageReturn key. * See RFC 1123 3.3.1 Telnet End-of-Line Convention. */ rc = wr; found = memchr(buf, '\r', wr); if (found) rc = found - buf + 1; rc = safe_write(ts->ptyfd, buf, rc); if (rc <= 0) return rc; if (rc < wr /* don't look past available data */ && buf[rc-1] == '\r' /* need this: imagine that write was _short_ */ && (buf[rc] == '\n' || buf[rc] == '\0') ) { rc++; } goto update_and_return; } /* buf starts with IAC char. Process that sequence. * Example: we get this from our own (bbox) telnet client: * read(5, "\377\374\1""\377\373\37""\377\372\37\0\262\0@\377\360""\377\375\1""\377\375\3"): * IAC WONT ECHO, IAC WILL NAWS, IAC SB NAWS IAC SE, IAC DO SGA * Another example (telnet-0.17 from old-netkit): * read(4, "\377\375\3""\377\373\30""\377\373\37""\377\373 ""\377\373!""\377\373\"""\377\373'" * "\377\375\5""\377\373#""\377\374\1""\377\372\37\0\257\0I\377\360""\377\375\1"): * IAC DO SGA, IAC WILL TTYPE, IAC WILL NAWS, IAC WILL TSPEED, IAC WILL LFLOW, IAC WILL LINEMODE, IAC WILL NEW_ENVIRON, * IAC DO STATUS, IAC WILL XDISPLOC, IAC WONT ECHO, IAC SB NAWS IAC SE, IAC DO ECHO */ if (wr <= 1) { /* Only the single IAC byte is in the buffer, eat it * and set a flag "process the rest of the sequence * next time we are here". */ //bb_error_msg("dangling IAC!"); ts->buffered_IAC_for_pty = 1; rc = 1; goto update_and_return; } handle_iac: /* 2-byte commands (240..250 and 255): * IAC IAC (255) Literal 255. Supported. * IAC SE (240) End of subnegotiation. Treated as NOP. * IAC NOP (241) NOP. Supported. * IAC BRK (243) Break. Like serial line break. TODO via tcsendbreak()? * IAC AYT (246) Are you there. * These don't look useful: * IAC DM (242) Data mark. What is this? * IAC IP (244) Suspend, interrupt or abort the process. (Ancient cousin of ^C). * IAC AO (245) Abort output. "You can continue running, but do not send me the output". * IAC EC (247) Erase character. The receiver should delete the last received char. * IAC EL (248) Erase line. The receiver should delete everything up tp last newline. * IAC GA (249) Go ahead. For half-duplex lines: "now you talk". * Implemented only as part of NAWS: * IAC SB (250) Subnegotiation of an option follows. */ if (buf[1] == IAC) { /* Literal 255 (emacs M-DEL) */ //bb_error_msg("255!"); rc = safe_write(ts->ptyfd, &buf[1], 1); /* * If we went through buffered_IAC_for_pty==1 path, * bailing out on error like below messes up the buffer. * EAGAIN is highly unlikely here, other errors will be * repeated on next write, let's just skip error check. */ #if 0 if (rc <= 0) return rc; #endif rc = 2; goto update_and_return; } if (buf[1] == AYT) { if (ts->size2 == 0) { /* if nothing buffered yet... */ /* Send back evidence that AYT was seen */ unsigned char *buf2 = TS_BUF2(ts); buf2[0] = IAC; buf2[1] = NOP; ts->wridx2 = 0; ts->rdidx2 = ts->size2 = 2; } rc = 2; goto update_and_return; } if (buf[1] >= 240 && buf[1] <= 249) { /* NOP (241). Ignore (putty keepalive, etc) */ /* All other 2-byte commands also treated as NOPs here */ rc = 2; goto update_and_return; } if (wr <= 2) { /* BUG: only 2 bytes of the IAC is in the buffer, we just eat them. * This is not a practical problem since >2 byte IACs are seen only * in initial negotiation, when buffer is empty */ rc = 2; goto update_and_return; } if (buf[1] == SB) { if (buf[2] == TELOPT_NAWS) { /* IAC SB, TELOPT_NAWS, 4-byte, IAC SE */ struct winsize ws; if (wr <= 6) { /* BUG: incomplete, can't process */ rc = wr; goto update_and_return; } memset(&ws, 0, sizeof(ws)); /* pixel sizes are set to 0 */ ws.ws_col = (buf[3] << 8) | buf[4]; ws.ws_row = (buf[5] << 8) | buf[6]; ioctl(ts->ptyfd, TIOCSWINSZ, (char *)&ws); rc = 7; /* trailing IAC SE will be eaten separately, as 2-byte NOP */ goto update_and_return; } /* else: other subnegs not supported yet */ } /* Assume it is a 3-byte WILL/WONT/DO/DONT 251..254 command and skip it */ #if DEBUG fprintf(stderr, "Ignoring IAC %s,%s\n", TELCMD(buf[1]), TELOPT(buf[2])); #endif rc = 3; update_and_return: ts->wridx1 += rc; if (ts->wridx1 >= BUFSIZE) /* actually == BUFSIZE */ ts->wridx1 = 0; ts->size1 -= rc; /* * Hack. We cannot process IACs which wrap around buffer's end. * Since properly fixing it requires writing bigger code, * we rely instead on this code making it virtually impossible * to have wrapped IAC (people don't type at 2k/second). * It also allows for bigger reads in common case. */ if (ts->size1 == 0) { /* very typical */ //bb_error_msg("zero size1"); ts->rdidx1 = 0; ts->wridx1 = 0; return rc; } wr = ts->wridx1; if (wr != 0 && wr < ts->rdidx1) { /* Buffer is not wrapped yet. * We can easily move it to the beginning. */ //bb_error_msg("moved %d", wr); memmove(TS_BUF1(ts), TS_BUF1(ts) + wr, ts->size1); ts->rdidx1 -= wr; ts->wridx1 = 0; } return rc; } /* * Converting single IAC into double on output */ static size_t safe_write_double_iac(int fd, const char *buf, size_t count) { const char *IACptr; size_t wr, rc, total; total = 0; while (1) { if (count == 0) return total; if (*buf == (char)IAC) { static const char IACIAC[] ALIGN1 = { IAC, IAC }; rc = safe_write(fd, IACIAC, 2); /* BUG: if partial write was only 1 byte long, we end up emitting just one IAC */ if (rc != 2) break; buf++; total++; count--; continue; } /* count != 0, *buf != IAC */ IACptr = memchr(buf, IAC, count); wr = count; if (IACptr) wr = IACptr - buf; rc = safe_write(fd, buf, wr); if (rc != wr) break; buf += rc; total += rc; count -= rc; } /* here: rc - result of last short write */ if ((ssize_t)rc < 0) { /* error? */ if (total == 0) return rc; rc = 0; } return total + rc; } /* Must match getopt32 string */ enum { OPT_WATCHCHILD = (1 << 2), /* -K */ OPT_INETD = (1 << 3) * ENABLE_FEATURE_TELNETD_STANDALONE, /* -i */ OPT_PORT = (1 << 4) * ENABLE_FEATURE_TELNETD_STANDALONE, /* -p PORT */ OPT_FOREGROUND = (1 << 6) * ENABLE_FEATURE_TELNETD_STANDALONE, /* -F */ OPT_SYSLOG = (1 << 7) * ENABLE_FEATURE_TELNETD_INETD_WAIT, /* -S */ OPT_WAIT = (1 << 8) * ENABLE_FEATURE_TELNETD_INETD_WAIT, /* -w SEC */ }; static struct tsession * make_new_session( IF_FEATURE_TELNETD_STANDALONE(int sock) IF_NOT_FEATURE_TELNETD_STANDALONE(void) ) { #if !ENABLE_FEATURE_TELNETD_STANDALONE enum { sock = 0 }; #endif const char *login_argv[2]; struct termios termbuf; int fd, pid; char tty_name[GETPTY_BUFSIZE]; struct tsession *ts = xzalloc(sizeof(struct tsession) + BUFSIZE * 2); /*ts->buf1 = (char *)(ts + 1);*/ /*ts->buf2 = ts->buf1 + BUFSIZE;*/ /* Got a new connection, set up a tty */ fd = xgetpty(tty_name); if (fd > G.maxfd) G.maxfd = fd; ts->ptyfd = fd; ndelay_on(fd); close_on_exec_on(fd); /* SO_KEEPALIVE by popular demand */ setsockopt_keepalive(sock); #if ENABLE_FEATURE_TELNETD_STANDALONE ts->sockfd_read = sock; ndelay_on(sock); if (sock == 0) { /* We are called with fd 0 - we are in inetd mode */ sock++; /* so use fd 1 for output */ ndelay_on(sock); } ts->sockfd_write = sock; if (sock > G.maxfd) G.maxfd = sock; #else /* ts->sockfd_read = 0; - done by xzalloc */ ts->sockfd_write = 1; ndelay_on(0); ndelay_on(1); #endif /* Make the telnet client understand we will echo characters so it * should not do it locally. We don't tell the client to run linemode, * because we want to handle line editing and tab completion and other * stuff that requires char-by-char support. */ { static const char iacs_to_send[] ALIGN1 = { IAC, DO, TELOPT_ECHO, IAC, DO, TELOPT_NAWS, /* This requires telnetd.ctrlSQ.patch (incomplete) */ /*IAC, DO, TELOPT_LFLOW,*/ IAC, WILL, TELOPT_ECHO, IAC, WILL, TELOPT_SGA }; /* This confuses safe_write_double_iac(), it will try to duplicate * each IAC... */ //memcpy(TS_BUF2(ts), iacs_to_send, sizeof(iacs_to_send)); //ts->rdidx2 = sizeof(iacs_to_send); //ts->size2 = sizeof(iacs_to_send); /* So just stuff it into TCP stream! (no error check...) */ #if ENABLE_FEATURE_TELNETD_STANDALONE safe_write(sock, iacs_to_send, sizeof(iacs_to_send)); #else safe_write(1, iacs_to_send, sizeof(iacs_to_send)); #endif /*ts->rdidx2 = 0; - xzalloc did it */ /*ts->size2 = 0;*/ } fflush_all(); pid = vfork(); /* NOMMU-friendly */ if (pid < 0) { free(ts); close(fd); /* sock will be closed by caller */ bb_simple_perror_msg("vfork"); return NULL; } if (pid > 0) { /* Parent */ ts->shell_pid = pid; return ts; } /* Child */ /* Careful - we are after vfork! */ /* Restore default signal handling ASAP */ bb_signals((1 << SIGCHLD) + (1 << SIGPIPE), SIG_DFL); pid = getpid(); if (ENABLE_FEATURE_UTMP) { len_and_sockaddr *lsa = get_peer_lsa(sock); char *hostname = NULL; if (lsa) { hostname = xmalloc_sockaddr2dotted(&lsa->u.sa); free(lsa); } write_new_utmp(pid, LOGIN_PROCESS, tty_name, /*username:*/ "LOGIN", hostname); free(hostname); } /* Make new session and process group */ setsid(); /* Open the child's side of the tty */ /* NB: setsid() disconnects from any previous ctty's. Therefore * we must open child's side of the tty AFTER setsid! */ close(0); xopen(tty_name, O_RDWR); /* becomes our ctty */ xdup2(0, 1); xdup2(0, 2); tcsetpgrp(0, pid); /* switch this tty's process group to us */ /* The pseudo-terminal allocated to the client is configured to operate * in cooked mode, and with XTABS CRMOD enabled (see tty(4)) */ tcgetattr(0, &termbuf); termbuf.c_lflag |= ECHO; /* if we use readline we dont want this */ termbuf.c_oflag |= ONLCR | XTABS; termbuf.c_iflag |= ICRNL; termbuf.c_iflag &= ~IXOFF; /*termbuf.c_lflag &= ~ICANON;*/ tcsetattr_stdin_TCSANOW(&termbuf); /* Uses FILE-based I/O to stdout, but does fflush_all(), * so should be safe with vfork. * I fear, though, that some users will have ridiculously big * issue files, and they may block writing to fd 1, * (parent is supposed to read it, but parent waits * for vforked child to exec!) */ print_login_issue(G.issuefile, tty_name); /* Exec shell / login / whatever */ login_argv[0] = G.loginpath; login_argv[1] = NULL; /* exec busybox applet (if PREFER_APPLETS=y), if that fails, * exec external program. * NB: sock is either 0 or has CLOEXEC set on it. * fd has CLOEXEC set on it too. These two fds will be closed here. */ BB_EXECVP(G.loginpath, (char **)login_argv); /* _exit is safer with vfork, and we shouldn't send message * to remote clients anyway */ _exit_FAILURE(); /*bb_perror_msg_and_die("execv %s", G.loginpath);*/ } #if ENABLE_FEATURE_TELNETD_STANDALONE static void free_session(struct tsession *ts) { struct tsession *t; if (option_mask32 & OPT_INETD) exit_SUCCESS(); /* Unlink this telnet session from the session list */ t = G.sessions; if (t == ts) G.sessions = ts->next; else { while (t->next != ts) t = t->next; t->next = ts->next; } #if 0 /* It was said that "normal" telnetd just closes ptyfd, * doesn't send SIGKILL. When we close ptyfd, * kernel sends SIGHUP to processes having slave side opened. */ kill(ts->shell_pid, SIGKILL); waitpid(ts->shell_pid, NULL, 0); #endif close(ts->ptyfd); close(ts->sockfd_read); /* We do not need to close(ts->sockfd_write), it's the same * as sockfd_read unless we are in inetd mode. But in inetd mode * we do not reach this */ free(ts); /* Scan all sessions and find new maxfd */ G.maxfd = 0; ts = G.sessions; while (ts) { if (G.maxfd < ts->ptyfd) G.maxfd = ts->ptyfd; if (G.maxfd < ts->sockfd_read) G.maxfd = ts->sockfd_read; #if 0 /* Again, sockfd_write == sockfd_read here */ if (G.maxfd < ts->sockfd_write) G.maxfd = ts->sockfd_write; #endif ts = ts->next; } } #else /* !FEATURE_TELNETD_STANDALONE */ /* Used in main() only, thus "return 0" actually is exit(EXIT_SUCCESS). */ #define free_session(ts) return 0 #endif static void handle_sigchld(int sig UNUSED_PARAM) { pid_t pid; struct tsession *ts; int save_errno = errno; /* Looping: more than one child may have exited */ while (1) { pid = wait_any_nohang(NULL); if (pid <= 0) break; ts = G.sessions; while (ts) { if (ts->shell_pid == pid) { ts->shell_pid = -1; update_utmp_DEAD_PROCESS(pid); break; } ts = ts->next; } } errno = save_errno; } int telnetd_main(int argc, char **argv) MAIN_EXTERNALLY_VISIBLE; int telnetd_main(int argc UNUSED_PARAM, char **argv) { fd_set rdfdset, wrfdset; unsigned opt; int count; struct tsession *ts; #if ENABLE_FEATURE_TELNETD_STANDALONE #define IS_INETD (opt & OPT_INETD) int master_fd = master_fd; /* for compiler */ int sec_linger = sec_linger; char *opt_bindaddr = NULL; char *opt_portnbr; #else enum { IS_INETD = 1, master_fd = -1, }; #endif INIT_G(); /* Even if !STANDALONE, we accept (and ignore) -i, thus people * don't need to guess whether it's ok to pass -i to us */ opt = getopt32(argv, "^" "f:l:Ki" IF_FEATURE_TELNETD_STANDALONE("p:b:F") IF_FEATURE_TELNETD_INETD_WAIT("Sw:+") /* -w NUM */ "\0" /* -w implies -F. -w and -i don't mix */ IF_FEATURE_TELNETD_INETD_WAIT("wF:i--w:w--i"), &G.issuefile, &G.loginpath IF_FEATURE_TELNETD_STANDALONE(, &opt_portnbr, &opt_bindaddr) IF_FEATURE_TELNETD_INETD_WAIT(, &sec_linger) ); if (!IS_INETD /*&& !re_execed*/) { /* inform that we start in standalone mode? * May be useful when people forget to give -i */ /*bb_error_msg("listening for connections");*/ if (!(opt & OPT_FOREGROUND)) { /* DAEMON_CHDIR_ROOT was giving inconsistent * behavior with/without -F, -i */ bb_daemonize_or_rexec(0 /*was DAEMON_CHDIR_ROOT*/, argv); } } /* Redirect log to syslog early, if needed */ if (IS_INETD || (opt & OPT_SYSLOG) || !(opt & OPT_FOREGROUND)) { openlog(applet_name, LOG_PID, LOG_DAEMON); logmode = LOGMODE_SYSLOG; } #if ENABLE_FEATURE_TELNETD_STANDALONE if (IS_INETD) { G.sessions = make_new_session(0); if (!G.sessions) /* pty opening or vfork problem, exit */ return 1; /* make_new_session printed error message */ } else { master_fd = 0; if (!(opt & OPT_WAIT)) { unsigned portnbr = CONFIG_FEATURE_TELNETD_PORT_DEFAULT; if (opt & OPT_PORT) portnbr = xatou16(opt_portnbr); master_fd = create_and_bind_stream_or_die(opt_bindaddr, portnbr); xlisten(master_fd, 1); } close_on_exec_on(master_fd); } #else G.sessions = make_new_session(); if (!G.sessions) /* pty opening or vfork problem, exit */ return 1; /* make_new_session printed error message */ #endif /* We don't want to die if just one session is broken */ signal(SIGPIPE, SIG_IGN); if (opt & OPT_WATCHCHILD) signal(SIGCHLD, handle_sigchld); else /* prevent dead children from becoming zombies */ signal(SIGCHLD, SIG_IGN); /* This is how the buffers are used. The arrows indicate data flow. +-------+ wridx1++ +------+ rdidx1++ +----------+ | | <-------------- | buf1 | <-------------- | | | | size1-- +------+ size1++ | | | pty | | socket | | | rdidx2++ +------+ wridx2++ | | | | --------------> | buf2 | --------------> | | +-------+ size2++ +------+ size2-- +----------+ size1: "how many bytes are buffered for pty between rdidx1 and wridx1?" size2: "how many bytes are buffered for socket between rdidx2 and wridx2?" Each session has got two buffers. Buffers are circular. If sizeN == 0, buffer is empty. If sizeN == BUFSIZE, buffer is full. In both these cases rdidxN == wridxN. */ again: FD_ZERO(&rdfdset); FD_ZERO(&wrfdset); /* Select on the master socket, all telnet sockets and their * ptys if there is room in their session buffers. * NB: scalability problem: we recalculate entire bitmap * before each select. Can be a problem with 500+ connections. */ ts = G.sessions; while (ts) { struct tsession *next = ts->next; /* in case we free ts */ if (ts->shell_pid == -1) { /* Child died and we detected that */ free_session(ts); } else { if (ts->size1 > 0) /* can write to pty */ FD_SET(ts->ptyfd, &wrfdset); if (ts->size1 < BUFSIZE) /* can read from socket */ FD_SET(ts->sockfd_read, &rdfdset); if (ts->size2 > 0) /* can write to socket */ FD_SET(ts->sockfd_write, &wrfdset); if (ts->size2 < BUFSIZE) /* can read from pty */ FD_SET(ts->ptyfd, &rdfdset); } ts = next; } if (!IS_INETD) { FD_SET(master_fd, &rdfdset); /* This is needed because free_session() does not * take master_fd into account when it finds new * maxfd among remaining fd's */ if (master_fd > G.maxfd) G.maxfd = master_fd; } { struct timeval *tv_ptr = NULL; #if ENABLE_FEATURE_TELNETD_INETD_WAIT struct timeval tv; if ((opt & OPT_WAIT) && !G.sessions) { tv.tv_sec = sec_linger; tv.tv_usec = 0; tv_ptr = &tv; } #endif count = select(G.maxfd + 1, &rdfdset, &wrfdset, NULL, tv_ptr); } if (count == 0) /* "telnetd -w SEC" timed out */ return 0; if (count < 0) goto again; /* EINTR or ENOMEM */ #if ENABLE_FEATURE_TELNETD_STANDALONE /* Check for and accept new sessions */ if (!IS_INETD && FD_ISSET(master_fd, &rdfdset)) { int fd; struct tsession *new_ts; fd = accept(master_fd, NULL, NULL); if (fd < 0) goto again; close_on_exec_on(fd); /* Create a new session and link it into active list */ new_ts = make_new_session(fd); if (new_ts) { new_ts->next = G.sessions; G.sessions = new_ts; } else { close(fd); } } #endif /* Then check for data tunneling */ ts = G.sessions; while (ts) { /* For all sessions... */ struct tsession *next = ts->next; /* in case we free ts */ if (/*ts->size1 &&*/ FD_ISSET(ts->ptyfd, &wrfdset)) { /* Write to pty from buffer 1 */ count = safe_write_to_pty_decode_iac(ts); if (count < 0) { if (errno == EAGAIN) goto skip1; goto kill_session; } } skip1: if (/*ts->size2 &&*/ FD_ISSET(ts->sockfd_write, &wrfdset)) { /* Write to socket from buffer 2 */ count = MIN(BUFSIZE - ts->wridx2, ts->size2); count = safe_write_double_iac(ts->sockfd_write, (void*)(TS_BUF2(ts) + ts->wridx2), count); if (count < 0) { if (errno == EAGAIN) goto skip2; goto kill_session; } ts->wridx2 += count; if (ts->wridx2 >= BUFSIZE) /* actually == BUFSIZE */ ts->wridx2 = 0; ts->size2 -= count; if (ts->size2 == 0) { ts->rdidx2 = 0; ts->wridx2 = 0; } } skip2: if (/*ts->size1 < BUFSIZE &&*/ FD_ISSET(ts->sockfd_read, &rdfdset)) { /* Read from socket to buffer 1 */ count = MIN(BUFSIZE - ts->rdidx1, BUFSIZE - ts->size1); count = safe_read(ts->sockfd_read, TS_BUF1(ts) + ts->rdidx1, count); if (count <= 0) { if (count < 0 && errno == EAGAIN) goto skip3; goto kill_session; } /* Ignore trailing NUL if it is there */ if (!TS_BUF1(ts)[ts->rdidx1 + count - 1]) { --count; } ts->size1 += count; ts->rdidx1 += count; if (ts->rdidx1 >= BUFSIZE) /* actually == BUFSIZE */ ts->rdidx1 = 0; } skip3: if (/*ts->size2 < BUFSIZE &&*/ FD_ISSET(ts->ptyfd, &rdfdset)) { /* Read from pty to buffer 2 */ int eio = 0; read_pty: count = MIN(BUFSIZE - ts->rdidx2, BUFSIZE - ts->size2); count = safe_read(ts->ptyfd, TS_BUF2(ts) + ts->rdidx2, count); if (count <= 0) { if (count < 0) { if (errno == EAGAIN) goto skip4; /* login process might call vhangup(), * which causes intermittent EIOs on read above * (observed on kernel 4.12.0). Try up to 10 ms. */ if (errno == EIO && eio < 10) { eio++; //bb_error_msg("EIO pty %u", eio); usleep(1000); goto read_pty; } } goto kill_session; } ts->size2 += count; ts->rdidx2 += count; if (ts->rdidx2 >= BUFSIZE) /* actually == BUFSIZE */ ts->rdidx2 = 0; } skip4: ts = next; continue; kill_session: if (ts->shell_pid > 0) update_utmp_DEAD_PROCESS(ts->shell_pid); free_session(ts); ts = next; } goto again; }