diff options
author | Denys Vlasenko | 2009-12-02 01:51:24 +0100 |
---|---|---|
committer | Denys Vlasenko | 2009-12-02 01:51:24 +0100 |
commit | 386960a38e5aee4f56a04214f86709c71dff120d (patch) | |
tree | 416cf9f19ee69749fbeb33931bc464edafe74321 | |
parent | fae9f499b2ab3becb8a672d982c50f318114dee9 (diff) | |
download | busybox-386960a38e5aee4f56a04214f86709c71dff120d.zip busybox-386960a38e5aee4f56a04214f86709c71dff120d.tar.gz |
htpd: better logging; removed some unused code and data. -212 bytes
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
-rw-r--r-- | networking/ntpd.c | 443 |
1 files changed, 222 insertions, 221 deletions
diff --git a/networking/ntpd.c b/networking/ntpd.c index f10a81c..3b9e7b7 100644 --- a/networking/ntpd.c +++ b/networking/ntpd.c @@ -67,27 +67,21 @@ enum { }; typedef struct { - uint8_t status; /* status of local clock and leap info */ - uint8_t stratum; /* stratum level */ - uint8_t ppoll; /* poll value */ - int8_t precision; - s_fixedpt_t rootdelay; - s_fixedpt_t dispersion; - uint32_t refid; - l_fixedpt_t reftime; - l_fixedpt_t orgtime; - l_fixedpt_t rectime; - l_fixedpt_t xmttime; - uint32_t keyid; - uint8_t digest[NTP_DIGESTSIZE]; + uint8_t m_status; /* status of local clock and leap info */ + uint8_t m_stratum; /* stratum level */ + uint8_t m_ppoll; /* poll value */ + int8_t m_precision; + s_fixedpt_t m_rootdelay; + s_fixedpt_t m_dispersion; + uint32_t m_refid; + l_fixedpt_t m_reftime; + l_fixedpt_t m_orgtime; + l_fixedpt_t m_rectime; + l_fixedpt_t m_xmttime; + uint32_t m_keyid; + uint8_t m_digest[NTP_DIGESTSIZE]; } ntp_msg_t; -typedef struct { - int fd; - ntp_msg_t msg; - double xmttime; -} ntp_query_t; - enum { NTP_VERSION = 4, NTP_MAXSTRATUM = 15, @@ -116,48 +110,38 @@ enum { #define OFFSET_1900_1970 2208988800UL /* 1970 - 1900 in seconds */ -enum client_state { - STATE_NONE, - STATE_QUERY_SENT, - STATE_REPLY_RECEIVED, -}; - typedef struct { - double rootdelay; - double rootdispersion; - double reftime; - uint32_t refid; - uint32_t refid4; - uint8_t synced; - uint8_t leap; - int8_t precision; - uint8_t poll; - uint8_t stratum; -} ntp_status_t; - -typedef struct { - ntp_status_t status; - double offset; - double delay; - double error; - time_t rcvd; - uint8_t good; + double o_offset; + double o_delay; + //UNUSED: double o_error; + time_t o_rcvd; + uint32_t o_refid4; + uint8_t o_leap; + uint8_t o_stratum; + uint8_t o_good; } ntp_offset_t; typedef struct { -//TODO: -// (1) store dotted addr str, to avoid constant translations -// (2) periodically re-resolve DNS names +//TODO: periodically re-resolve DNS names? len_and_sockaddr *lsa; - ntp_query_t query; - ntp_offset_t reply[OFFSET_ARRAY_SIZE]; - ntp_offset_t update; - enum client_state state; + char *dotted; + double xmttime; time_t next; time_t deadline; + int fd; + uint8_t state; uint8_t shift; uint8_t trustlevel; + ntp_msg_t msg; + ntp_offset_t update; + ntp_offset_t reply[OFFSET_ARRAY_SIZE]; } ntp_peer_t; +/* for ntp_peer_t::state */ +enum { + STATE_NONE, + STATE_QUERY_SENT, + STATE_REPLY_RECEIVED, +}; enum { OPT_n = (1 << 0), @@ -172,14 +156,21 @@ enum { struct globals { - unsigned verbose; + double rootdelay; + double reftime; + llist_t *ntp_peers; #if ENABLE_FEATURE_NTPD_SERVER int listen_fd; #endif + unsigned verbose; unsigned peer_cnt; - llist_t *ntp_peers; - ntp_status_t status; + uint32_t refid; + uint32_t refid4; uint32_t scale; + uint8_t synced; + uint8_t leap; + int8_t precision; + uint8_t stratum; uint8_t time_is_set; uint8_t first_adj_done; }; @@ -190,7 +181,7 @@ static const int const_IPTOS_LOWDELAY = IPTOS_LOWDELAY; static void -set_next(ntp_peer_t *p, time_t t) +set_next(ntp_peer_t *p, unsigned t) { p->next = time(NULL) + t; p->deadline = 0; @@ -204,12 +195,12 @@ add_peers(const char *s) p = xzalloc(sizeof(*p)); //TODO: big ntpd uses all IPs, not just 1st, do we need to mimic that? p->lsa = xhost2sockaddr(s, 123); - p->query.fd = -1; - p->query.msg.status = MODE_CLIENT | (NTP_VERSION << 3); + p->dotted = xmalloc_sockaddr2dotted_noport(&p->lsa->u.sa); + p->fd = -1; + p->msg.m_status = MODE_CLIENT | (NTP_VERSION << 3); if (STATE_NONE != 0) p->state = STATE_NONE; p->trustlevel = TRUSTLEVEL_PATHETIC; - p->query.fd = -1; set_next(p, 0); llist_add_to(&G.ntp_peers, p); @@ -241,6 +232,7 @@ lfp_to_d(l_fixedpt_t lfp) return ret; } +#if 0 //UNUSED static double sfp_to_d(s_fixedpt_t sfp) { @@ -250,6 +242,7 @@ sfp_to_d(s_fixedpt_t sfp) ret = (double)sfp.int_parts + ((double)sfp.fractions / USHRT_MAX); return ret; } +#endif #if ENABLE_FEATURE_NTPD_SERVER static l_fixedpt_t @@ -282,17 +275,17 @@ set_deadline(ntp_peer_t *p, time_t t) p->next = 0; } -static time_t +static unsigned error_interval(void) { - time_t interval, r; + unsigned interval, r; interval = INTERVAL_QUERY_PATHETIC * QSCALE_OFF_MAX / QSCALE_OFF_MIN; - r = (unsigned)random() % (unsigned long)(interval / 10); + r = (unsigned)random() % (unsigned)(interval / 10); return (interval + r); } static int -sendmsg_wrap(int fd, +do_sendto(int fd, const struct sockaddr *from, const struct sockaddr *to, socklen_t addrlen, ntp_msg_t *msg, ssize_t len) { @@ -331,16 +324,16 @@ send_query_to_peer(ntp_peer_t *p) // Uncomment this and use strace to see it in action: #define PROBE_LOCAL_ADDR // { len_and_sockaddr lsa; lsa.len = LSA_SIZEOF_SA; getsockname(p->query.fd, &lsa.u.sa, &lsa.len); } - if (p->query.fd == -1) { + if (p->fd == -1) { int fd, family; len_and_sockaddr *local_lsa; family = p->lsa->u.sa.sa_family; - //was: p->query.fd = xsocket(family, SOCK_DGRAM, 0); - p->query.fd = fd = xsocket_type(&local_lsa, family, SOCK_DGRAM); + //was: p->fd = xsocket(family, SOCK_DGRAM, 0); + p->fd = fd = xsocket_type(&local_lsa, family, SOCK_DGRAM); /* local_lsa has "null" address and port 0 now. * bind() ensures we have a *particular port* selected by kernel - * and remembered in p->query.fd, thus later recv(p->query.fd) + * and remembered in p->fd, thus later recv(p->fd) * receives only packets sent to this port. */ PROBE_LOCAL_ADDR @@ -367,16 +360,19 @@ send_query_to_peer(ntp_peer_t *p) * Save the real transmit timestamp locally. */ - p->query.msg.xmttime.int_partl = random(); - p->query.msg.xmttime.fractionl = random(); - p->query.xmttime = gettime1900fp(); + p->msg.m_xmttime.int_partl = random(); + p->msg.m_xmttime.fractionl = random(); + p->xmttime = gettime1900fp(); - if (sendmsg_wrap(p->query.fd, /*from:*/ NULL, /*to:*/ &p->lsa->u.sa, /*addrlen:*/ p->lsa->len, - &p->query.msg, NTP_MSGSIZE_NOAUTH) == -1) { + if (do_sendto(p->fd, /*from:*/ NULL, /*to:*/ &p->lsa->u.sa, /*addrlen:*/ p->lsa->len, + &p->msg, NTP_MSGSIZE_NOAUTH) == -1 + ) { set_next(p, INTERVAL_QUERY_PATHETIC); return -1; } + if (G.verbose) + bb_error_msg("sent request to %s", p->dotted); p->state = STATE_QUERY_SENT; set_deadline(p, QUERYTIME_MAX); @@ -384,13 +380,13 @@ send_query_to_peer(ntp_peer_t *p) } static int -offset_compare(const void *aa, const void *bb) +compare_offsets(const void *aa, const void *bb) { const ntp_peer_t *const *a = aa; const ntp_peer_t *const *b = bb; - if ((*a)->update.offset < (*b)->update.offset) + if ((*a)->update.o_offset < (*b)->update.o_offset) return -1; - return ((*a)->update.offset > (*b)->update.offset); + return ((*a)->update.o_offset > (*b)->update.o_offset); } static uint32_t @@ -406,24 +402,20 @@ updated_scale(double offset) } static void -adjtime_wrap(void) +slew_time(void) { ntp_peer_t *p; + llist_t *item; unsigned offset_cnt; - unsigned middle; - int i = 0; - ntp_peer_t **peers; double offset_median; - llist_t *item; - len_and_sockaddr *lsa; - struct timeval tv, olddelta; + struct timeval tv; offset_cnt = 0; for (item = G.ntp_peers; item != NULL; item = item->link) { p = (ntp_peer_t *) item->data; if (p->trustlevel < TRUSTLEVEL_BADPEER) continue; - if (!p->update.good) + if (!p->update.o_good) return; offset_cnt++; } @@ -431,68 +423,74 @@ adjtime_wrap(void) if (offset_cnt == 0) goto clear_good; - peers = xzalloc(sizeof(peers[0]) * offset_cnt); - for (item = G.ntp_peers; item != NULL; item = item->link) { - p = (ntp_peer_t *) item->data; - if (p->trustlevel < TRUSTLEVEL_BADPEER) - continue; - peers[i++] = p; - } - - qsort(peers, offset_cnt, sizeof(peers[0]), offset_compare); + { + len_and_sockaddr *lsa; + unsigned middle; + unsigned i = 0; + ntp_peer_t **peers = xzalloc(sizeof(peers[0]) * offset_cnt); + + for (item = G.ntp_peers; item != NULL; item = item->link) { + p = (ntp_peer_t *) item->data; + if (p->trustlevel < TRUSTLEVEL_BADPEER) + continue; + peers[i++] = p; + } - middle = offset_cnt / 2; - if (middle != 0 && (offset_cnt & 1) == 0) { - offset_median = (peers[middle-1]->update.offset + peers[middle]->update.offset) / 2; - G.status.rootdelay = (peers[middle-1]->update.delay + peers[middle]->update.delay) / 2; - G.status.stratum = MAX(peers[middle-1]->update.status.stratum, peers[middle]->update.status.stratum); - } else { - offset_median = peers[middle]->update.offset; - G.status.rootdelay = peers[middle]->update.delay; - G.status.stratum = peers[middle]->update.status.stratum; + qsort(peers, offset_cnt, sizeof(peers[0]), compare_offsets); + + middle = offset_cnt / 2; + if (middle != 0 && (offset_cnt & 1) == 0) { + offset_median = (peers[middle-1]->update.o_offset + peers[middle]->update.o_offset) / 2; + G.rootdelay = (peers[middle-1]->update.o_delay + peers[middle]->update.o_delay) / 2; + G.stratum = 1 + MAX(peers[middle-1]->update.o_stratum, peers[middle]->update.o_stratum); + } else { + offset_median = peers[middle]->update.o_offset; + G.rootdelay = peers[middle]->update.o_delay; + G.stratum = 1 + peers[middle]->update.o_stratum; + } + G.leap = peers[middle]->update.o_leap; + G.refid4 = peers[middle]->update.o_refid4; + lsa = peers[middle]->lsa; + G.refid = +#if ENABLE_FEATURE_IPV6 + lsa->u.sa.sa_family != AF_INET ? + G.refid4 : +#endif + lsa->u.sin.sin_addr.s_addr; + free(peers); } - G.status.leap = peers[middle]->update.status.leap; - bb_info_msg("adjusting local clock by %fs", offset_median); + bb_error_msg("adjusting clock by %fs, our stratum is %u", offset_median, G.stratum); + errno = 0; d_to_tv(offset_median, &tv); - if (adjtime(&tv, &olddelta) == -1) { + if (adjtime(&tv, &tv) == -1) { bb_perror_msg("adjtime failed"); //TODO: maybe _and_die? - } else - if (G.first_adj_done - && olddelta.tv_sec == 0 - && olddelta.tv_usec == 0 - && !G.status.synced - ) { - bb_info_msg("clock synced"); - G.status.synced = 1; - } else - if (G.status.synced) { - bb_info_msg("clock unsynced"); - G.status.synced = 0; + } else { + if (G.verbose >= 2) + bb_error_msg("old adjust: %d.%06u", (int)tv.tv_sec, (unsigned)tv.tv_usec); + if (G.first_adj_done + && tv.tv_sec == 0 + && tv.tv_usec == 0 // TODO: allow for tiny values? + && !G.synced + ) { + G.synced = 1; + bb_error_msg("clock %ssynced", ""); + } else + if (G.synced) { + G.synced = 0; + bb_error_msg("clock %ssynced", "un"); + } } G.first_adj_done = 1; - G.status.reftime = gettime1900fp(); - G.status.stratum++; /* one more than selected peer */ + G.reftime = gettime1900fp(); G.scale = updated_scale(offset_median); - G.status.refid4 = peers[middle]->update.status.refid4; - - lsa = peers[middle]->lsa; - G.status.refid = -#if ENABLE_FEATURE_IPV6 - lsa->u.sa.sa_family != AF_INET ? - G.status.refid4 : -#endif - lsa->u.sin.sin_addr.s_addr; - - free(peers); - clear_good: for (item = G.ntp_peers; item != NULL; item = item->link) { p = (ntp_peer_t *) item->data; - p->update.good = 0; + p->update.o_good = 0; } } @@ -530,7 +528,7 @@ step_time_once(double offset) strftime(buf, sizeof(buf), "%a %b %e %H:%M:%S %Z %Y", localtime(&tval)); /* Do we want to print message below to system log when daemonized? */ - bb_info_msg("set local clock to %s (offset %fs)", buf, offset); + bb_error_msg("set local clock to %s (offset %fs)", buf, offset); for (item = G.ntp_peers; item != NULL; item = item->link) { p = (ntp_peer_t *) item->data; @@ -548,69 +546,68 @@ step_time_once(double offset) static void update_peer_data(ntp_peer_t *p) { - int i, best = 0, good = 0; - - /* - * clock filter - * find the offset which arrived with the lowest delay - * use that as the peer update - * invalidate it and all older ones + /* Clock filter. + * Find the offset which arrived with the lowest delay. + * Use that as the peer update. + * Invalidate it and all older ones. */ + int i; + int best = best; /* for compiler */ + int good; - for (i = 0; good == 0 && i < OFFSET_ARRAY_SIZE; i++) { - if (p->reply[i].good) { + good = 0; + for (i = 0; i < OFFSET_ARRAY_SIZE; i++) { + if (p->reply[i].o_good) { good++; - best = i; + best = i++; + break; } } for (; i < OFFSET_ARRAY_SIZE; i++) { - if (p->reply[i].good) { + if (p->reply[i].o_good) { good++; - if (p->reply[i].delay < p->reply[best].delay) + if (p->reply[i].o_delay < p->reply[best].o_delay) best = i; } } - if (good < 8) + if (good < 8) //FIXME: was it meant to be OFFSET_ARRAY_SIZE, not 8? return; memcpy(&p->update, &p->reply[best], sizeof(p->update)); - adjtime_wrap(); + slew_time(); for (i = 0; i < OFFSET_ARRAY_SIZE; i++) - if (p->reply[i].rcvd <= p->reply[best].rcvd) - p->reply[i].good = 0; + if (p->reply[i].o_rcvd <= p->reply[best].o_rcvd) + p->reply[i].o_good = 0; } -static time_t -scale_interval(time_t requested) +static unsigned +scale_interval(unsigned requested) { - time_t interval, r; + unsigned interval, r; interval = requested * G.scale; - r = (unsigned)random() % (unsigned long)(MAX(5, interval / 10)); + r = (unsigned)random() % (unsigned)(MAX(5, interval / 10)); return (interval + r); } static void recv_and_process_peer_pkt(ntp_peer_t *p) { - char *addr; ssize_t size; ntp_msg_t msg; double T1, T2, T3, T4; - time_t interval; + unsigned interval; ntp_offset_t *offset; - addr = xmalloc_sockaddr2dotted_noport(&p->lsa->u.sa); - /* We can recvfrom here and check from.IP, but some multihomed * ntp servers reply from their *other IP*. * TODO: maybe we should check at least what we can: from.port == 123? */ - size = recv(p->query.fd, &msg, sizeof(msg), MSG_DONTWAIT); + size = recv(p->fd, &msg, sizeof(msg), MSG_DONTWAIT); if (size == -1) { - bb_perror_msg("recv(%s) error", addr); + bb_perror_msg("recv(%s) error", p->dotted); if (errno == EHOSTUNREACH || errno == EHOSTDOWN || errno == ENETUNREACH || errno == ENETDOWN || errno == ECONNREFUSED || errno == EADDRNOTAVAIL @@ -618,7 +615,7 @@ recv_and_process_peer_pkt(ntp_peer_t *p) ) { //TODO: always do this? set_next(p, error_interval()); - goto bail; + goto close_sock; } xfunc_die(); } @@ -626,23 +623,23 @@ recv_and_process_peer_pkt(ntp_peer_t *p) T4 = gettime1900fp(); if (size != NTP_MSGSIZE_NOAUTH && size != NTP_MSGSIZE) { - bb_error_msg("malformed packet received from %s", addr); + bb_error_msg("malformed packet received from %s", p->dotted); goto bail; } - if (msg.orgtime.int_partl != p->query.msg.xmttime.int_partl - || msg.orgtime.fractionl != p->query.msg.xmttime.fractionl + if (msg.m_orgtime.int_partl != p->msg.m_xmttime.int_partl + || msg.m_orgtime.fractionl != p->msg.m_xmttime.fractionl ) { goto bail; } - if ((msg.status & LI_ALARM) == LI_ALARM - || msg.stratum == 0 - || msg.stratum > NTP_MAXSTRATUM + if ((msg.m_status & LI_ALARM) == LI_ALARM + || msg.m_stratum == 0 + || msg.m_stratum > NTP_MAXSTRATUM ) { interval = error_interval(); - bb_info_msg("reply from %s: not synced, next query %ds", addr, (int) interval); - goto bail; + bb_error_msg("reply from %s: not synced, next query %us", p->dotted, interval); + goto close_sock; } /* @@ -660,34 +657,34 @@ recv_and_process_peer_pkt(ntp_peer_t *p) * d = (T4 - T1) - (T3 - T2) t = ((T2 - T1) + (T3 - T4)) / 2. */ - T1 = p->query.xmttime; - T2 = lfp_to_d(msg.rectime); - T3 = lfp_to_d(msg.xmttime); + T1 = p->xmttime; + T2 = lfp_to_d(msg.m_rectime); + T3 = lfp_to_d(msg.m_xmttime); offset = &p->reply[p->shift]; - offset->offset = ((T2 - T1) + (T3 - T4)) / 2; - offset->delay = (T4 - T1) - (T3 - T2); - if (offset->delay < 0) { + offset->o_offset = ((T2 - T1) + (T3 - T4)) / 2; + offset->o_delay = (T4 - T1) - (T3 - T2); + if (offset->o_delay < 0) { interval = error_interval(); set_next(p, interval); - bb_info_msg("reply from %s: negative delay %f", addr, p->reply[p->shift].delay); - goto bail; + bb_error_msg("reply from %s: negative delay %f", p->dotted, p->reply[p->shift].o_delay); + goto close_sock; } - offset->error = (T2 - T1) - (T3 - T4); + //UNUSED: offset->o_error = (T2 - T1) - (T3 - T4); // Can we use (T4 - OFFSET_1900_1970) instead of time(NULL)? - offset->rcvd = time(NULL); - offset->good = 1; - - offset->status.leap = (msg.status & LI_MASK); - offset->status.precision = msg.precision; - offset->status.rootdelay = sfp_to_d(msg.rootdelay); - offset->status.rootdispersion = sfp_to_d(msg.dispersion); - offset->status.refid = ntohl(msg.refid); - offset->status.refid4 = msg.xmttime.fractionl; - offset->status.reftime = lfp_to_d(msg.reftime); - offset->status.poll = msg.ppoll; - offset->status.stratum = msg.stratum; + offset->o_rcvd = time(NULL); + offset->o_good = 1; + + offset->o_leap = (msg.m_status & LI_MASK); + //UNUSED: offset->o_precision = msg.m_precision; + //UNUSED: offset->o_rootdelay = sfp_to_d(msg.m_rootdelay); + //UNUSED: offset->o_rootdispersion = sfp_to_d(msg.m_dispersion); + //UNUSED: offset->o_refid = ntohl(msg.m_refid); + offset->o_refid4 = msg.m_xmttime.fractionl; + //UNUSED: offset->o_reftime = lfp_to_d(msg.m_reftime); + //UNUSED: offset->o_poll = msg.m_ppoll; + offset->o_stratum = msg.m_stratum; if (p->trustlevel < TRUSTLEVEL_PATHETIC) interval = scale_interval(INTERVAL_QUERY_PATHETIC); @@ -703,21 +700,29 @@ recv_and_process_peer_pkt(ntp_peer_t *p) if (p->trustlevel < TRUSTLEVEL_MAX) { p->trustlevel++; if (p->trustlevel == TRUSTLEVEL_BADPEER) - bb_info_msg("peer %s now valid", addr); + bb_error_msg("peer %s now valid", p->dotted); } - bb_info_msg("reply from %s: offset %f delay %f, next query %ds", addr, - offset->offset, offset->delay, (int) interval); + if (G.verbose) + bb_error_msg("reply from %s: offset %f delay %f, next query %us", p->dotted, + offset->o_offset, offset->o_delay, interval); update_peer_data(p); - step_time_once(offset->offset); + step_time_once(offset->o_offset); p->shift++; if (p->shift >= OFFSET_ARRAY_SIZE) p->shift = 0; + close_sock: + /* We do not expect any more packets for now. + * Closing the socket informs kernel about it. + * We open a new socket when we send a new query. + */ + close(p->fd); + p->fd = -1; bail: - free(addr); + return; } #if ENABLE_FEATURE_NTPD_SERVER @@ -751,31 +756,31 @@ recv_and_process_client_pkt(void /*int fd*/) goto bail; } - query_status = msg.status; - query_ppoll = msg.ppoll; - query_xmttime = msg.xmttime; + query_status = msg.m_status; + query_ppoll = msg.m_ppoll; + query_xmttime = msg.m_xmttime; /* Build a reply packet */ memset(&msg, 0, sizeof(msg)); - msg.status = G.status.synced ? G.status.leap : LI_ALARM; - msg.status |= (query_status & VERSION_MASK); - msg.status |= ((query_status & MODE_MASK) == MODE_CLIENT) ? + msg.m_status = G.synced ? G.leap : LI_ALARM; + msg.m_status |= (query_status & VERSION_MASK); + msg.m_status |= ((query_status & MODE_MASK) == MODE_CLIENT) ? MODE_SERVER : MODE_SYM_PAS; - msg.stratum = G.status.stratum; - msg.ppoll = query_ppoll; - msg.precision = G.status.precision; + msg.m_stratum = G.stratum; + msg.m_ppoll = query_ppoll; + msg.m_precision = G.precision; rectime = gettime1900fp(); - msg.xmttime = msg.rectime = d_to_lfp(rectime); - msg.reftime = d_to_lfp(G.status.reftime); - //msg.xmttime = d_to_lfp(gettime1900fp()); // = msg.rectime - msg.orgtime = query_xmttime; - msg.rootdelay = d_to_sfp(G.status.rootdelay); + msg.m_xmttime = msg.m_rectime = d_to_lfp(rectime); + msg.m_reftime = d_to_lfp(G.reftime); + //msg.m_xmttime = d_to_lfp(gettime1900fp()); // = msg.m_rectime + msg.m_orgtime = query_xmttime; + msg.m_rootdelay = d_to_sfp(G.rootdelay); version = (query_status & VERSION_MASK); /* ... >> VERSION_SHIFT - done below instead */ - msg.refid = (version > (3 << VERSION_SHIFT)) ? G.status.refid4 : G.status.refid; + msg.m_refid = (version > (3 << VERSION_SHIFT)) ? G.refid4 : G.refid; /* We reply from the local address packet was sent to, * this makes to/from look swapped here: */ - sendmsg_wrap(G.listen_fd, + do_sendto(G.listen_fd, /*from:*/ &to->u.sa, /*to:*/ from, /*addrlen:*/ to->len, &msg, size); @@ -931,7 +936,7 @@ static NOINLINE void ntp_init(char **argv) #endif while (b > 1) prec--, b >>= 1; - G.status.precision = prec; + G.precision = prec; } G.scale = 1; @@ -991,18 +996,14 @@ int ntpd_main(int argc UNUSED_PARAM, char **argv) } if (p->deadline != 0 && p->deadline <= cur_time) { /* Timed out waiting for reply */ - char *addr = xmalloc_sockaddr2dotted_noport(&p->lsa->u.sa); - timeout = error_interval(); - bb_info_msg("no reply from %s received in time, " - "next query %ds", addr, timeout); + bb_error_msg("timed out waiting for %s, " + "next query %us", p->dotted, timeout); if (p->trustlevel >= TRUSTLEVEL_BADPEER) { p->trustlevel /= 2; if (p->trustlevel < TRUSTLEVEL_BADPEER) - bb_info_msg("peer %s now invalid", addr); + bb_error_msg("peer %s now invalid", p->dotted); } - free(addr); - set_next(p, timeout); } @@ -1013,7 +1014,7 @@ int ntpd_main(int argc UNUSED_PARAM, char **argv) if (p->state == STATE_QUERY_SENT) { /* Wait for reply from this peer */ - pfd[i].fd = p->query.fd; + pfd[i].fd = p->fd; pfd[i].events = POLLIN; idx2peer[i] = p; i++; @@ -1024,12 +1025,12 @@ int ntpd_main(int argc UNUSED_PARAM, char **argv) step_time_once(0); /* no good peers, don't wait */ timeout = nextaction - cur_time; - if (timeout < 0) - timeout = 0; + if (timeout < 1) + timeout = 1; /* Here we may block */ - if (g.verbose) - bb_error_msg("entering poll %u secs", timeout); + if (g.verbose >= 2) + bb_error_msg("poll %u sec, waiting on %u sockets", timeout, i); nfds = poll(pfd, i, timeout * 1000); if (nfds <= 0) continue; |