githubmirror/sssd

src/util/server.c
794 строки23 KB
Alexey Tikhonov
DEBUG: add 'debug_backtrace_enable' getter
23 окт 2024, 15:37
23 окт 2024, 15:3706920b4
/*
   SSSD

   Servers setup routines

   Copyright (C) Andrew Tridgell        1992-2005
   Copyright (C) Martin Pool            2002
   Copyright (C) Jelmer Vernooij        2002
   Copyright (C) James J Myers          2003 <myersjj@samba.org>
   Copyright (C) Simo Sorce             2008

   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; either version 3 of the License, or
   (at your option) any later version.

   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 <http://www.gnu.org/licenses/>.
*/

#include <stdio.h>
#include <sys/types.h>
#include <sys/wait.h>
#include <sys/stat.h>
#include <fcntl.h>
#include <unistd.h>
#include <signal.h>
#include <sys/prctl.h>
#include <ldb.h>
#include "util/util.h"
#include "confdb/confdb.h"
#include "util/sss_chain_id.h"
#include "util/sss_chain_id_tevent.h"

#ifdef HAVE_PRCTL
#include <sys/prctl.h>
#endif

static TALLOC_CTX *autofree_ctx;

static void server_atexit(void)
{
    talloc_zfree(autofree_ctx);
}

/*******************************************************************
 Close the low 3 FDs and open dev/null in their place.
********************************************************************/
static void close_low_fds(void)
{
#ifndef VALGRIND
    /* try and use up these file descriptors, so silly
       library routines writing to stdout etc. won't cause havoc */
    if (freopen ("/dev/null", "r", stdin) == NULL) {
        DEBUG(SSSDBG_FATAL_FAILURE, "Can't freopen() stdin to /dev/null\n");
    }
    if (freopen ("/dev/null", "w", stdout) == NULL) {
        DEBUG(SSSDBG_FATAL_FAILURE, "Can't freopen() stdout to /dev/null\n");
    }
    if (freopen ("/dev/null", "w", stderr) == NULL) {
        DEBUG(SSSDBG_FATAL_FAILURE, "Can't freopen() stderr to /dev/null\n");
    }
#endif
}

static void daemon_parent_sigterm(int sig)
{
    _exit(0);
}

/**
 Become a daemon, discarding the controlling terminal.
**/

static void become_daemon(void)
{
    pid_t pid, cpid;
    int status;
    int ret, error;

    pid = fork();
    if (pid == -1) {
        ret = errno;
        DEBUG(SSSDBG_FATAL_FAILURE, "fork() failed: %d [%s]\n",
                                     ret, strerror(ret));
        sss_log(SSS_LOG_ERR, "can't start: fork() failed");
        _exit(1);
    }
    if (pid != 0) {
        /* Terminate parent process on demand so we can hold systemd
         * or initd from starting next service until SSSD is initialized.
         * We use signals directly here because we don't have a tevent
         * context yet. */
        CatchSignal(SIGTERM, daemon_parent_sigterm);

        /* or exit when child process (i.e. sssd monitor) is terminated
         * and return error in this case */
        ret = 1;
        do {
            error = 0;
            cpid = waitpid(pid, &status, 0);
            if (cpid == -1) {
                /* An error occurred while waiting */
                error = errno;
                if (error != EINTR) {
                    DEBUG(SSSDBG_CRIT_FAILURE,
                          "Error [%d][%s] while waiting for child\n",
                           error, strerror(error));
                    /* Forcibly kill this child */
                    kill(pid, SIGKILL);
                }
            } else {
                if (WIFEXITED(status)) {
                    /* return our exit code if available */
                    ret = WEXITSTATUS(status);
                }
            }
        } while (error == EINTR);

        _exit(ret);
    }

    /* create new session, process group and detach from the terminal */
    if (setsid() == (pid_t) -1) {
        ret = errno;
        DEBUG(SSSDBG_FATAL_FAILURE, "setsid() failed: %d [%s]\n",
                                     ret, strerror(ret));
        sss_log(SSS_LOG_ERR, "can't start: setsid() failed");
        _exit(1);
    }

    /* chdir to / to be sure we're not on a remote filesystem */
    errno = 0;
    if(chdir("/") == -1) {
        ret = errno;
        DEBUG(SSSDBG_FATAL_FAILURE, "Cannot change directory (%d [%s])\n",
                                     ret, strerror(ret));
    }

    /* Close FDs 0,1,2. Needed if started by rsh */
    close_low_fds();
}

int check_pidfile(const char *file)
{
    char pid_str[32];
    pid_t pid;
    int fd;
    int ret, err;
    ssize_t len;
    ssize_t pidlen = sizeof(pid_str) - 1;

    fd = open(file, O_RDONLY, 0644);
    err = errno;
    if (fd != -1) {
        errno = 0;
        len = sss_atomic_read_s(fd, pid_str, pidlen);
        ret = errno;
        close(fd);
        if (len == -1) {
            DEBUG(SSSDBG_CRIT_FAILURE,
                  "read failed [%d][%s].\n", ret, strerror(ret));
            return EINVAL;
        }

        /* Ensure NULL-termination */
        pid_str[len] = '\0';

        /* let's check the pid */
        pid = (pid_t)atoi(pid_str);
        if (pid != 0) {
            errno = 0;
            ret = kill(pid, 0);
            /* succeeded in signaling the process -> another sssd process */
            if (ret == 0) {
                return EEXIST;
            }
            if (ret != 0 && errno != ESRCH) {
                err = errno;
                return err;
            }
        }

        /* nothing in the file or no process */
        ret = unlink(file);
        /* non-fatal failure */
        if (ret != EOK) {
            ret = errno;
            DEBUG(SSSDBG_MINOR_FAILURE,
                  "Failed to remove file: %s - %d [%s]!\n",
                  file, ret, sss_strerror(ret));
        }
    } else {
        if (err != ENOENT) {
            return err;
        }
    }

    return 0;
}

int pidfile(const char *file)
{
    char pid_str[32];
    int fd;
    int ret, err;
    size_t size;
    ssize_t written;

    ret = check_pidfile(file);
    if (ret != EOK) {
        return ret;
    }

    fd = open(file, O_CREAT | O_WRONLY | O_EXCL, 0644);
    err = errno;
    if (fd == -1) {
        return err;
    }

    memset(pid_str, 0, sizeof(pid_str));
    snprintf(pid_str, sizeof(pid_str) -1, "%u\n", (unsigned int) getpid());
    size = strlen(pid_str);

    errno = 0;
    written = sss_atomic_write_s(fd, pid_str, size);
    err = errno;
    close(fd);
    if (written == -1) {
        DEBUG(SSSDBG_CRIT_FAILURE,
              "write failed [%d][%s]\n", err, strerror(err));
        return err;
    }

    if (written != size) {
        DEBUG(SSSDBG_CRIT_FAILURE,
              "Wrote %zd bytes expected %zu\n", written, size);
        return EIO;
    }

    return 0;
}

void orderly_shutdown(int status)
{
#if HAVE_GETPGRP
    static int sent_sigterm;
    int debug;

    if (sent_sigterm == 0 && getpgrp() == getpid()) {
        debug = is_socket_activated() ? SSSDBG_TRACE_INTERNAL
                                      : SSSDBG_IMPORTANT_INFO;
        DEBUG(debug, "SIGTERM: killing children\n");
        sent_sigterm = 1;
        kill(-getpgrp(), SIGTERM);
    }
#endif
    DEBUG(SSSDBG_IMPORTANT_INFO, "Shutting down (status = %d)\n", status);
    sss_log(SSS_LOG_INFO, "Shutting down (status = %d)", status);
    exit(status);
}

static void default_quit(struct tevent_context *ev,
                         struct tevent_signal *se,
                         int signum,
                         int count,
                         void *siginfo,
                         void *private_data)
{
    struct main_context *ctx = talloc_get_type(private_data, struct main_context);
    talloc_free(ctx);

    orderly_shutdown(0);
}

#ifndef HAVE_PRCTL
static void sig_segv_abrt(int sig)
{
    DEBUG(SSSDBG_FATAL_FAILURE,
          "Received signal %s, shutting down\n", strsignal(sig));
    orderly_shutdown(1);
}
#endif /* HAVE_PRCTL */

/*
  setup signal masks
*/
static void setup_signals(void)
{
    /* we are never interested in SIGPIPE */
    BlockSignals(true, SIGPIPE);

#if defined(SIGFPE)
    /* we are never interested in SIGFPE */
    BlockSignals(true, SIGFPE);
#endif

    /* We are no longer interested in USR1 */
    BlockSignals(true, SIGUSR1);

    /* We are no longer interested in SIGINT except for monitor */
    BlockSignals(true, SIGINT);

#if defined(SIGUSR2)
    /* We are no longer interested in USR2 */
    BlockSignals(true, SIGUSR2);
#endif

    /* POSIX demands that signals are inherited. If the invoking process has
     * these signals masked, we will have problems, as we won't receive them. */
    BlockSignals(false, SIGHUP);
    BlockSignals(false, SIGTERM);

#ifndef HAVE_PRCTL
        /* If prctl is not defined on the system, try to handle
         * some common termination signals gracefully */
    CatchSignal(SIGSEGV, sig_segv_abrt);
    CatchSignal(SIGABRT, sig_segv_abrt);
#endif

}

/*
  handle io on stdin
*/
static void server_stdin_handler(struct tevent_context *event_ctx,
                                 struct tevent_fd *fde,
                                 uint16_t flags, void *private)
{
    const char *binary_name = (const char *)private;
    uint8_t c;

    errno = 0;
    if (sss_atomic_read_s(0, &c, 1) == 0) {
    DEBUG(SSSDBG_CRIT_FAILURE, "%s: EOF on stdin - terminating\n",
                                binary_name);
#if HAVE_GETPGRP
        if (getpgrp() == getpid()) {
            kill(-getpgrp(), SIGTERM);
        }
#endif
        exit(0);
    }
}

/*
 main server helpers.
*/

int die_if_parent_died(void)
{
#ifdef HAVE_PRCTL
    int ret;

    errno = 0;
    ret = prctl(PR_SET_PDEATHSIG, SIGTERM, 0, 0, 0);
    if (ret != 0) {
        ret = errno;
        DEBUG(SSSDBG_OP_FAILURE, "prctl failed [%d]: %s\n",
                                 ret, strerror(ret));
        return ret;
    }
#endif
    return EOK;
}

struct logrotate_ctx {
    struct confdb_ctx *confdb;
    const char *confdb_path;
};

static void te_server_hup(struct tevent_context *ev,
                          struct tevent_signal *se,
                          int signum,
                          int count,
                          void *siginfo,
                          void *private_data)
{
    errno_t ret;
    struct logrotate_ctx *lctx =
            talloc_get_type(private_data, struct logrotate_ctx);

    DEBUG(SSSDBG_IMPORTANT_INFO, "Received SIGHUP. Rotating logfiles.\n");

    ret = server_common_rotate_logs(lctx->confdb, lctx->confdb_path);
    if (ret != EOK) {
        DEBUG(SSSDBG_FATAL_FAILURE, "Could not reopen log file [%s]\n",
                                     strerror(ret));
    }
}

errno_t server_common_rotate_logs(struct confdb_ctx *confdb,
                                  const char *conf_path)
{
    errno_t ret;
    int old_debug_level = debug_level;

    ret = rotate_debug_files();
    if (ret) {
        sss_log(SSS_LOG_ALERT, "Could not rotate debug files! [%d][%s]\n",
                               ret, strerror(ret));
        return ret;
    }

    /* Get new debug level from the confdb */
    ret = confdb_get_int(confdb, conf_path,
                         CONFDB_SERVICE_DEBUG_LEVEL,
                         old_debug_level,
                         &debug_level);
    if (ret != EOK) {
        DEBUG(SSSDBG_FATAL_FAILURE, "Error reading from confdb (%d) [%s]\n",
                  ret, strerror(ret));
        /* Try to proceed with the old value */
        debug_level = old_debug_level;
    }

    if (debug_level != old_debug_level) {
        DEBUG(SSSDBG_IMPORTANT_INFO,
              "Debug level changed to %#.4x\n", debug_level);
        debug_level = debug_convert_old_level(debug_level);
    }

    return EOK;
}

errno_t generic_get_debug_level(TALLOC_CTX *mem_ctx,
                                struct sbus_request *sbus_req,
                                void *pvt_data,
                                uint32_t *_debug_level)
{
    *_debug_level = debug_level;
    return EOK;
}

errno_t generic_set_debug_level(TALLOC_CTX *mem_ctx,
                                struct sbus_request *sbus_req,
                                void *pvt_data,
                                uint32_t new_debug_level)
{
    debug_level = new_debug_level;
    return EOK;
}

static const char *get_db_path(void)
{
#ifdef UNIT_TESTING
#ifdef TEST_DB_PATH
    return TEST_DB_PATH;
#else
    #error "TEST_DB_PATH must be defined when unit testing server.c!"
#endif /* TEST_DB_PATH */
#else
    return DB_PATH;
#endif /* UNIT_TESTING */
}

static const char *get_pid_path(void)
{
#ifdef UNIT_TESTING
#ifdef TEST_PID_PATH
    return TEST_PID_PATH;
#else
    #error "TEST_PID_PATH must be defined when unit testing server.c!"
#endif /* TEST_PID_PATH */
#else
    return PID_PATH;
#endif
}

int server_setup(const char *name, bool is_responder,
                 int flags,
                 uid_t uid, gid_t gid,
                 const char *conf_entry,
                 struct main_context **main_ctx,
                 bool allow_sss_loop)
{
    struct tevent_context *event_ctx;
    struct main_context *ctx;
    uint16_t stdin_event_flags;
    char *conf_db;
    int ret = EOK;
    bool dt;
    bool dm;
    bool backtrace_enabled;
    struct tevent_signal *tes;
    struct logrotate_ctx *lctx;
    char *locale;
    int watchdog_interval;
    pid_t my_pid;
    char *pidfile_name;
    int cfg_debug_level = SSSDBG_INVALID;
    bool dumpable = true;

    if (is_responder) {
        sss_chain_id_set_format(DEBUG_CHAIN_ID_FMT_CID);
    } else {
        sss_chain_id_set_format(DEBUG_CHAIN_ID_FMT_RID);
    }

    talloc_enable_null_tracking();

    autofree_ctx = talloc_named_const(NULL, 0, "autofree_context");
    if (autofree_ctx == NULL) {
        return ENOMEM;
    }

    atexit(server_atexit);

    debug_prg_name = talloc_strdup(autofree_ctx, name);
    if (!debug_prg_name) {
        return ENOMEM;
    }

    my_pid = getpid();
    ret = setpgid(my_pid, my_pid);
    if (ret != EOK) {
        ret = errno;
        DEBUG(SSSDBG_MINOR_FAILURE,
              "Failed setting process group: %s[%d]. "
              "We might leak processes in case of failure\n",
              sss_strerror(ret), ret);
    }

    if (!is_socket_activated()) {
        ret = chown_debug_file(NULL, uid, gid);
        if (ret != EOK) {
            DEBUG(SSSDBG_MINOR_FAILURE,
                  "Cannot chown the debug files, debugging might not work!\n");
        }

        ret = become_user(uid, gid);
        if (ret != EOK) {
            DEBUG(SSSDBG_FUNC_DATA,
                  "Cannot become user [%"SPRIuid"][%"SPRIgid"].\n", uid, gid);
            return ret;
        }
    }

    if (!allow_sss_loop) {
        ret = setenv("_SSS_LOOPS", "NO", 0);
        if (ret != 0) {
            DEBUG(SSSDBG_OP_FAILURE, "Failed to set _SSS_LOOPS.\n");
            return ret;
        }
    }
    /* To make sure the domain cannot be set from the environment, unset the
     * variable explicitly when setting up any server. Backends later set the
     * value after reading domain from the configuration */
    ret = unsetenv(SSS_DOM_ENV);
    if (ret != 0) {
        DEBUG(SSSDBG_MINOR_FAILURE, "Unsetting "SSS_DOM_ENV" failed, journald "
              "logging might not work as expected\n");
    }

    setup_signals();

    /* we want default permissions on created files to be very strict */
    umask(SSS_DFL_UMASK);

    if (flags & FLAGS_DAEMON) {
        DEBUG(SSSDBG_IMPORTANT_INFO, "Becoming a daemon.\n");
        become_daemon();
    }

    if (flags & FLAGS_PID_FILE) {
        pidfile_name = talloc_asprintf(NULL, "%s/%s.pid", get_pid_path(), name);
        if (!pidfile_name) {
            return ENOMEM;
        }
        ret = pidfile(pidfile_name);
        if (ret != EOK) {
            DEBUG(SSSDBG_FATAL_FAILURE, "Error creating pidfile: %s! "
                  "(%d [%s])\n", pidfile_name, ret, strerror(ret));
            talloc_free(pidfile_name);
            return ret;
        }
        talloc_free(pidfile_name);
    }

    /* Set up locale */
    locale = setlocale(LC_ALL, "");
    if (locale == NULL) {
        /* Just print debug message and continue */
        DEBUG(SSSDBG_TRACE_FUNC, "Unable to set locale\n");
    }

    bindtextdomain(PACKAGE, LOCALEDIR);
    textdomain(PACKAGE);

    /* the event context is the top level structure.
     * Everything else should hang off that */
    event_ctx = tevent_context_init(autofree_ctx);
    if (event_ctx == NULL) {
        DEBUG(SSSDBG_FATAL_FAILURE,
              "The event context initialization failed\n");
        return 1;
    }

    ctx = talloc(event_ctx, struct main_context);
    if (ctx == NULL) {
        DEBUG(SSSDBG_FATAL_FAILURE, "Out of memory, aborting!\n");
        return ENOMEM;
    }

    ctx->parent_pid = getppid();
    ctx->event_ctx = event_ctx;

    /* Set up an event handler for a SIGINT */
    tes = tevent_add_signal(event_ctx, event_ctx, SIGINT, 0,
                            default_quit, ctx);
    if (tes == NULL) {
        return EIO;
    }

    /* Set up an event handler for a SIGTERM */
    tes = tevent_add_signal(event_ctx, event_ctx, SIGTERM, 0,
                            default_quit, ctx);
    if (tes == NULL) {
        return EIO;
    }

    conf_db = talloc_asprintf(ctx, "%s/%s",
                              get_db_path(), CONFDB_FILE);
    if (conf_db == NULL) {
        DEBUG(SSSDBG_FATAL_FAILURE, "Out of memory, aborting!\n");
        return ENOMEM;
    }

    ret = confdb_init(ctx, &ctx->confdb_ctx, conf_db);
    if (ret != EOK) {
        DEBUG(SSSDBG_FATAL_FAILURE, "The confdb initialization failed\n");
        return ret;
    }

    if (debug_level == SSSDBG_UNRESOLVED) {
        /* set debug level if any in conf_entry */
        ret = confdb_get_int(ctx->confdb_ctx, conf_entry,
                             CONFDB_SERVICE_DEBUG_LEVEL,
                             SSSDBG_INVALID,
                             &cfg_debug_level);
        if (ret != EOK) {
            DEBUG(SSSDBG_FATAL_FAILURE, "Error reading from confdb (%d) "
                                         "[%s]\n", ret, strerror(ret));
            return ret;
        }

        if (cfg_debug_level == SSSDBG_INVALID) {
            /* Check for the `debug` alias */
            ret = confdb_get_int(ctx->confdb_ctx, conf_entry,
                    CONFDB_SERVICE_DEBUG_LEVEL_ALIAS,
                    SSSDBG_DEFAULT,
                    &cfg_debug_level);
            if (ret != EOK) {
                DEBUG(SSSDBG_FATAL_FAILURE, "Error reading from confdb (%d) "
                                            "[%s]\n", ret, strerror(ret));
                return ret;
            }
        }

        debug_level = debug_convert_old_level(cfg_debug_level);
    }

    /* same for debug timestamps */
    if (debug_timestamps == SSSDBG_TIMESTAMP_UNRESOLVED) {
        ret = confdb_get_bool(ctx->confdb_ctx, conf_entry,
                              CONFDB_SERVICE_DEBUG_TIMESTAMPS,
                              SSSDBG_TIMESTAMP_DEFAULT,
                              &dt);
        if (ret != EOK) {
            DEBUG(SSSDBG_FATAL_FAILURE, "Error reading from confdb (%d) "
                                         "[%s]\n", ret, strerror(ret));
            return ret;
        }
        if (dt) debug_timestamps = SSSDBG_TIMESTAMP_ENABLED;
        else debug_timestamps = SSSDBG_TIMESTAMP_DISABLED;
    }

    /* same for debug microseconds */
    if (debug_microseconds == SSSDBG_MICROSECONDS_UNRESOLVED) {
        ret = confdb_get_bool(ctx->confdb_ctx, conf_entry,
                              CONFDB_SERVICE_DEBUG_MICROSECONDS,
                              SSSDBG_MICROSECONDS_DEFAULT,
                              &dm);
        if (ret != EOK) {
            DEBUG(SSSDBG_FATAL_FAILURE, "Error reading from confdb (%d) "
                                         "[%s]\n", ret, strerror(ret));
            return ret;
        }
        if (dm) debug_microseconds = SSSDBG_MICROSECONDS_ENABLED;
        else debug_microseconds = SSSDBG_MICROSECONDS_DISABLED;
    }

    ret = confdb_get_bool(ctx->confdb_ctx, conf_entry,
                          CONFDB_SERVICE_DEBUG_BACKTRACE_ENABLED,
                          true,
                          &backtrace_enabled);
    if (ret != EOK) {
        DEBUG(SSSDBG_FATAL_FAILURE, "Error reading %s from confdb (%d) [%s]\n",
              CONFDB_SERVICE_DEBUG_BACKTRACE_ENABLED, ret, strerror(ret));
        return ret;
    }
    sss_set_debug_backtrace_enable(backtrace_enabled);

    /* before opening the log file set up log rotation */
    lctx = talloc_zero(ctx, struct logrotate_ctx);
    if (!lctx) return ENOMEM;

    lctx->confdb = ctx->confdb_ctx;
    lctx->confdb_path = conf_entry;

    tes = tevent_add_signal(ctx->event_ctx, ctx, SIGHUP, 0,
                            te_server_hup, lctx);
    if (tes == NULL) {
        return EIO;
    }

    DEBUG(SSSDBG_IMPORTANT_INFO,
          "Starting with debug level = %#.4x\n", debug_level);

    /* Setup the internal watchdog */
    ret = confdb_get_int(ctx->confdb_ctx, conf_entry,
                         CONFDB_DOMAIN_TIMEOUT,
                         0, &watchdog_interval);
    if (ret != EOK) {
        DEBUG(SSSDBG_FATAL_FAILURE, "Error reading from confdb (%d) [%s]\n",
                                     ret, strerror(ret));
        return ret;
    }

    if ((flags & FLAGS_NO_WATCHDOG) == 0) {
        ret = setup_watchdog(ctx->event_ctx, watchdog_interval);
        if (ret != EOK) {
            DEBUG(SSSDBG_CRIT_FAILURE, "Watchdog setup failed.\n");
            return ret;
        }
    }

    ret = confdb_get_bool(ctx->confdb_ctx,
                          CONFDB_MONITOR_CONF_ENTRY,
                          CONFDB_MONITOR_DUMPABLE,
                          true, /* default value */
                          &dumpable);
    if (ret != EOK) {
        DEBUG(SSSDBG_FATAL_FAILURE, "Failed to determine "CONFDB_MONITOR_DUMPABLE"\n");
        return ret;
    }
    ret = prctl(PR_SET_DUMPABLE, dumpable ? 1 : 0);
    if (ret != 0) {
        DEBUG(SSSDBG_CRIT_FAILURE, "Failed to set PR_SET_DUMPABLE\n");
        return ret;
    } else if (!dumpable) {
        DEBUG(SSSDBG_IMPORTANT_INFO, "Core dumps are disabled!\n");
    }

    sss_chain_id_setup(ctx->event_ctx);

    sss_log(SSS_LOG_INFO, "Starting up");

    DEBUG(SSSDBG_TRACE_FUNC, "CONFDB: %s\n", conf_db);

    if (flags & FLAGS_INTERACTIVE) {
        /* terminate when stdin goes away */
        stdin_event_flags = TEVENT_FD_READ;
    } else {
        /* stay alive forever */
        stdin_event_flags = 0;
    }

    /* catch EOF on stdin */
#ifdef SIGTTIN
    signal(SIGTTIN, SIG_IGN);
#endif
    tevent_add_fd(event_ctx, event_ctx, STDIN_FILENO, stdin_event_flags,
                 server_stdin_handler, discard_const(name));

    *main_ctx = ctx;
    return EOK;
}

void server_loop(struct main_context *main_ctx)
{
    /* wait for events - this is where the server sits for most of its
       life */
    tevent_loop_wait(main_ctx->event_ctx);

    /* as everything hangs off this event context, freeing it
       should initiate a clean shutdown of all services */
    talloc_free(main_ctx->event_ctx);
}