freebsd-skq/usr.sbin/ypserv/yp_main.c

582 lines
13 KiB
C
Raw Normal View History

/*-
* SPDX-License-Identifier: BSD-4-Clause
*
* Copyright (c) 1995
* Bill Paul <wpaul@ctr.columbia.edu>. All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. All advertising materials mentioning features or use of this software
* must display the following acknowledgement:
* This product includes software developed by Bill Paul.
* 4. Neither the name of the author nor the names of any co-contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY Bill Paul AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL Bill Paul OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
/*
* ypserv startup function.
* We need out own main() since we have to do some additional work
* that rpcgen won't do for us. Most of this file was generated using
* rpcgen.new, and later modified.
*/
#include <sys/types.h>
#include <sys/mman.h>
#include <sys/queue.h>
#include <sys/socket.h>
#include <sys/wait.h>
#include "yp.h"
#include <err.h>
#include <errno.h>
#include <memory.h>
#include <stdio.h>
#include <signal.h>
#include <stdarg.h>
#include <stdlib.h> /* getenv, exit */
#include <string.h> /* strcmp */
#include <syslog.h>
#include <unistd.h>
#ifdef __cplusplus
#include <sysent.h> /* getdtablesize, open */
#endif /* __cplusplus */
#include <netinet/in.h>
#include <netdb.h>
#include "yp_extern.h"
#include <netconfig.h>
#include <rpc/rpc.h>
#include <rpc/rpc_com.h>
#ifndef SIG_PF
#define SIG_PF void(*)(int)
#endif
#define _RPCSVC_CLOSEDOWN 120
int _rpcpmstart; /* Started by a port monitor ? */
static int _rpcfdtype; /* Whether Stream or Datagram? */
static int _rpcaf;
static int _rpcfd;
/* States a server can be in wrt request */
#define _IDLE 0
#define _SERVED 1
#define _SERVING 2
extern void ypprog_1(struct svc_req *, SVCXPRT *);
extern void ypprog_2(struct svc_req *, SVCXPRT *);
extern int _rpc_dtablesize(void);
extern int _rpcsvcstate; /* Set when a request is serviced */
char *progname = "ypserv";
char *yp_dir = _PATH_YP;
/*int debug = 0;*/
int do_dns = 0;
Big round o changes: - yp_dblookup.c: Create non-DB specific database access functions. Using these allows access to the underlying database functions without needing explicit knowledge of Berkeley DB. (These are used only when DB_CACHE is #defined. Other programs that use the non-caching functions (yp_mkdb, ypxfr, yppush, rpc.yppasswdd) shouldn't notice the difference.) - yp_dnslookup: Implement async DNS lookups. We send our own DNS requests using UDP and put the request in a queue. When the response arrives, we use the ID in the header to find the corresponsing queue entry and then send the response to the client. We can go about our business and handle other YP requests in the meantime. This way, we can deal with time consuming DNS requests without blocking and without forking. - yp_server.c: Convert to using new non-DB-specific database access functions. This simplifies the code a bit and removes the need for this module to know anything about Berkeley DB. Also convert the ypproc_match_2_svc() function to use the async DNS lookup routines. - yp_main.c: tweak yp_svc_run() to add the resolver socket to the set of descriptors monitored in the select() loop. Also add a timeout to select(); we may get stale DNS requests stuck in the queue which we want to invalidate after a while. If the timeout hits, we decrement the ttl on all pending DNS requests and nuke those requests that aren't handled before ttl hits zero. - yp_extern.h: Add prototypes for new stuff. - yp_svc_udp.c (new file): The async resolver code needs to be able to rummage around inside the RPC UDP transport handle in order to work correcty. There's basically one transport handle, and each time a request comes in, the transaction ID in the handle is changed. This means that if we queue a DNS request, then we handle some other unrelated requests, we will be unable to send the DNS response because the transaction ID and remote address of the client that made the DNS request will have been lost. What we need to do is save the client address and transaction ID in the queue entry for the DNS request, then put the transaction ID and address back in the transport handle when we're ready to reply. (And then we have to undo the change so as not to confuse any other part of the server.) The trouble is that the transaction ID is hidden in an opaque part of the transport handle, and only the code in the svc_udp module in the RPC library knows how to handle it. This file contains a couple of functions that let us read and set the transaction ID in spite of this. This is really a dirty trick and I should be taken out and shot for even thinking about it, but there's no other way to get this stuff to work. - Makefile: add yp_svc_udp.c to SRCS.
1996-12-22 22:30:58 +00:00
int resfd;
struct socklistent {
int sle_sock;
struct sockaddr_storage sle_ss;
SLIST_ENTRY(socklistent) sle_next;
};
static SLIST_HEAD(, socklistent) sle_head =
SLIST_HEAD_INITIALIZER(sle_head);
struct bindaddrlistent {
const char *ble_hostname;
SLIST_ENTRY(bindaddrlistent) ble_next;
};
static SLIST_HEAD(, bindaddrlistent) ble_head =
SLIST_HEAD_INITIALIZER(ble_head);
static char *servname = "0";
static
void _msgout(char* msg, ...)
{
va_list ap;
va_start(ap, msg);
if (debug) {
if (_rpcpmstart)
vsyslog(LOG_ERR, msg, ap);
else
vwarnx(msg, ap);
} else
vsyslog(LOG_ERR, msg, ap);
va_end(ap);
}
pid_t yp_pid;
static void
yp_svc_run(void)
{
#ifdef FD_SETSIZE
fd_set readfds;
#else
int readfds;
#endif /* def FD_SETSIZE */
int fd_setsize = _rpc_dtablesize();
struct timeval timeout;
/* Establish the identity of the parent ypserv process. */
yp_pid = getpid();
for (;;) {
#ifdef FD_SETSIZE
readfds = svc_fdset;
#else
readfds = svc_fds;
#endif /* def FD_SETSIZE */
Big round o changes: - yp_dblookup.c: Create non-DB specific database access functions. Using these allows access to the underlying database functions without needing explicit knowledge of Berkeley DB. (These are used only when DB_CACHE is #defined. Other programs that use the non-caching functions (yp_mkdb, ypxfr, yppush, rpc.yppasswdd) shouldn't notice the difference.) - yp_dnslookup: Implement async DNS lookups. We send our own DNS requests using UDP and put the request in a queue. When the response arrives, we use the ID in the header to find the corresponsing queue entry and then send the response to the client. We can go about our business and handle other YP requests in the meantime. This way, we can deal with time consuming DNS requests without blocking and without forking. - yp_server.c: Convert to using new non-DB-specific database access functions. This simplifies the code a bit and removes the need for this module to know anything about Berkeley DB. Also convert the ypproc_match_2_svc() function to use the async DNS lookup routines. - yp_main.c: tweak yp_svc_run() to add the resolver socket to the set of descriptors monitored in the select() loop. Also add a timeout to select(); we may get stale DNS requests stuck in the queue which we want to invalidate after a while. If the timeout hits, we decrement the ttl on all pending DNS requests and nuke those requests that aren't handled before ttl hits zero. - yp_extern.h: Add prototypes for new stuff. - yp_svc_udp.c (new file): The async resolver code needs to be able to rummage around inside the RPC UDP transport handle in order to work correcty. There's basically one transport handle, and each time a request comes in, the transaction ID in the handle is changed. This means that if we queue a DNS request, then we handle some other unrelated requests, we will be unable to send the DNS response because the transaction ID and remote address of the client that made the DNS request will have been lost. What we need to do is save the client address and transaction ID in the queue entry for the DNS request, then put the transaction ID and address back in the transport handle when we're ready to reply. (And then we have to undo the change so as not to confuse any other part of the server.) The trouble is that the transaction ID is hidden in an opaque part of the transport handle, and only the code in the svc_udp module in the RPC library knows how to handle it. This file contains a couple of functions that let us read and set the transaction ID in spite of this. This is really a dirty trick and I should be taken out and shot for even thinking about it, but there's no other way to get this stuff to work. - Makefile: add yp_svc_udp.c to SRCS.
1996-12-22 22:30:58 +00:00
FD_SET(resfd, &readfds);
Big round o changes: - yp_dblookup.c: Create non-DB specific database access functions. Using these allows access to the underlying database functions without needing explicit knowledge of Berkeley DB. (These are used only when DB_CACHE is #defined. Other programs that use the non-caching functions (yp_mkdb, ypxfr, yppush, rpc.yppasswdd) shouldn't notice the difference.) - yp_dnslookup: Implement async DNS lookups. We send our own DNS requests using UDP and put the request in a queue. When the response arrives, we use the ID in the header to find the corresponsing queue entry and then send the response to the client. We can go about our business and handle other YP requests in the meantime. This way, we can deal with time consuming DNS requests without blocking and without forking. - yp_server.c: Convert to using new non-DB-specific database access functions. This simplifies the code a bit and removes the need for this module to know anything about Berkeley DB. Also convert the ypproc_match_2_svc() function to use the async DNS lookup routines. - yp_main.c: tweak yp_svc_run() to add the resolver socket to the set of descriptors monitored in the select() loop. Also add a timeout to select(); we may get stale DNS requests stuck in the queue which we want to invalidate after a while. If the timeout hits, we decrement the ttl on all pending DNS requests and nuke those requests that aren't handled before ttl hits zero. - yp_extern.h: Add prototypes for new stuff. - yp_svc_udp.c (new file): The async resolver code needs to be able to rummage around inside the RPC UDP transport handle in order to work correcty. There's basically one transport handle, and each time a request comes in, the transaction ID in the handle is changed. This means that if we queue a DNS request, then we handle some other unrelated requests, we will be unable to send the DNS response because the transaction ID and remote address of the client that made the DNS request will have been lost. What we need to do is save the client address and transaction ID in the queue entry for the DNS request, then put the transaction ID and address back in the transport handle when we're ready to reply. (And then we have to undo the change so as not to confuse any other part of the server.) The trouble is that the transaction ID is hidden in an opaque part of the transport handle, and only the code in the svc_udp module in the RPC library knows how to handle it. This file contains a couple of functions that let us read and set the transaction ID in spite of this. This is really a dirty trick and I should be taken out and shot for even thinking about it, but there's no other way to get this stuff to work. - Makefile: add yp_svc_udp.c to SRCS.
1996-12-22 22:30:58 +00:00
timeout.tv_sec = RESOLVER_TIMEOUT;
timeout.tv_usec = 0;
switch (select(fd_setsize, &readfds, NULL, NULL,
Big round o changes: - yp_dblookup.c: Create non-DB specific database access functions. Using these allows access to the underlying database functions without needing explicit knowledge of Berkeley DB. (These are used only when DB_CACHE is #defined. Other programs that use the non-caching functions (yp_mkdb, ypxfr, yppush, rpc.yppasswdd) shouldn't notice the difference.) - yp_dnslookup: Implement async DNS lookups. We send our own DNS requests using UDP and put the request in a queue. When the response arrives, we use the ID in the header to find the corresponsing queue entry and then send the response to the client. We can go about our business and handle other YP requests in the meantime. This way, we can deal with time consuming DNS requests without blocking and without forking. - yp_server.c: Convert to using new non-DB-specific database access functions. This simplifies the code a bit and removes the need for this module to know anything about Berkeley DB. Also convert the ypproc_match_2_svc() function to use the async DNS lookup routines. - yp_main.c: tweak yp_svc_run() to add the resolver socket to the set of descriptors monitored in the select() loop. Also add a timeout to select(); we may get stale DNS requests stuck in the queue which we want to invalidate after a while. If the timeout hits, we decrement the ttl on all pending DNS requests and nuke those requests that aren't handled before ttl hits zero. - yp_extern.h: Add prototypes for new stuff. - yp_svc_udp.c (new file): The async resolver code needs to be able to rummage around inside the RPC UDP transport handle in order to work correcty. There's basically one transport handle, and each time a request comes in, the transaction ID in the handle is changed. This means that if we queue a DNS request, then we handle some other unrelated requests, we will be unable to send the DNS response because the transaction ID and remote address of the client that made the DNS request will have been lost. What we need to do is save the client address and transaction ID in the queue entry for the DNS request, then put the transaction ID and address back in the transport handle when we're ready to reply. (And then we have to undo the change so as not to confuse any other part of the server.) The trouble is that the transaction ID is hidden in an opaque part of the transport handle, and only the code in the svc_udp module in the RPC library knows how to handle it. This file contains a couple of functions that let us read and set the transaction ID in spite of this. This is really a dirty trick and I should be taken out and shot for even thinking about it, but there's no other way to get this stuff to work. - Makefile: add yp_svc_udp.c to SRCS.
1996-12-22 22:30:58 +00:00
&timeout)) {
case -1:
if (errno == EINTR) {
continue;
}
warn("svc_run: - select failed");
return;
case 0:
if (getpid() == yp_pid)
yp_prune_dnsq();
Big round o changes: - yp_dblookup.c: Create non-DB specific database access functions. Using these allows access to the underlying database functions without needing explicit knowledge of Berkeley DB. (These are used only when DB_CACHE is #defined. Other programs that use the non-caching functions (yp_mkdb, ypxfr, yppush, rpc.yppasswdd) shouldn't notice the difference.) - yp_dnslookup: Implement async DNS lookups. We send our own DNS requests using UDP and put the request in a queue. When the response arrives, we use the ID in the header to find the corresponsing queue entry and then send the response to the client. We can go about our business and handle other YP requests in the meantime. This way, we can deal with time consuming DNS requests without blocking and without forking. - yp_server.c: Convert to using new non-DB-specific database access functions. This simplifies the code a bit and removes the need for this module to know anything about Berkeley DB. Also convert the ypproc_match_2_svc() function to use the async DNS lookup routines. - yp_main.c: tweak yp_svc_run() to add the resolver socket to the set of descriptors monitored in the select() loop. Also add a timeout to select(); we may get stale DNS requests stuck in the queue which we want to invalidate after a while. If the timeout hits, we decrement the ttl on all pending DNS requests and nuke those requests that aren't handled before ttl hits zero. - yp_extern.h: Add prototypes for new stuff. - yp_svc_udp.c (new file): The async resolver code needs to be able to rummage around inside the RPC UDP transport handle in order to work correcty. There's basically one transport handle, and each time a request comes in, the transaction ID in the handle is changed. This means that if we queue a DNS request, then we handle some other unrelated requests, we will be unable to send the DNS response because the transaction ID and remote address of the client that made the DNS request will have been lost. What we need to do is save the client address and transaction ID in the queue entry for the DNS request, then put the transaction ID and address back in the transport handle when we're ready to reply. (And then we have to undo the change so as not to confuse any other part of the server.) The trouble is that the transaction ID is hidden in an opaque part of the transport handle, and only the code in the svc_udp module in the RPC library knows how to handle it. This file contains a couple of functions that let us read and set the transaction ID in spite of this. This is really a dirty trick and I should be taken out and shot for even thinking about it, but there's no other way to get this stuff to work. - Makefile: add yp_svc_udp.c to SRCS.
1996-12-22 22:30:58 +00:00
break;
default:
if (getpid() == yp_pid) {
if (FD_ISSET(resfd, &readfds)) {
yp_run_dnsq();
FD_CLR(resfd, &readfds);
}
svc_getreqset(&readfds);
Big round o changes: - yp_dblookup.c: Create non-DB specific database access functions. Using these allows access to the underlying database functions without needing explicit knowledge of Berkeley DB. (These are used only when DB_CACHE is #defined. Other programs that use the non-caching functions (yp_mkdb, ypxfr, yppush, rpc.yppasswdd) shouldn't notice the difference.) - yp_dnslookup: Implement async DNS lookups. We send our own DNS requests using UDP and put the request in a queue. When the response arrives, we use the ID in the header to find the corresponsing queue entry and then send the response to the client. We can go about our business and handle other YP requests in the meantime. This way, we can deal with time consuming DNS requests without blocking and without forking. - yp_server.c: Convert to using new non-DB-specific database access functions. This simplifies the code a bit and removes the need for this module to know anything about Berkeley DB. Also convert the ypproc_match_2_svc() function to use the async DNS lookup routines. - yp_main.c: tweak yp_svc_run() to add the resolver socket to the set of descriptors monitored in the select() loop. Also add a timeout to select(); we may get stale DNS requests stuck in the queue which we want to invalidate after a while. If the timeout hits, we decrement the ttl on all pending DNS requests and nuke those requests that aren't handled before ttl hits zero. - yp_extern.h: Add prototypes for new stuff. - yp_svc_udp.c (new file): The async resolver code needs to be able to rummage around inside the RPC UDP transport handle in order to work correcty. There's basically one transport handle, and each time a request comes in, the transaction ID in the handle is changed. This means that if we queue a DNS request, then we handle some other unrelated requests, we will be unable to send the DNS response because the transaction ID and remote address of the client that made the DNS request will have been lost. What we need to do is save the client address and transaction ID in the queue entry for the DNS request, then put the transaction ID and address back in the transport handle when we're ready to reply. (And then we have to undo the change so as not to confuse any other part of the server.) The trouble is that the transaction ID is hidden in an opaque part of the transport handle, and only the code in the svc_udp module in the RPC library knows how to handle it. This file contains a couple of functions that let us read and set the transaction ID in spite of this. This is really a dirty trick and I should be taken out and shot for even thinking about it, but there's no other way to get this stuff to work. - Makefile: add yp_svc_udp.c to SRCS.
1996-12-22 22:30:58 +00:00
}
}
if (yp_pid != getpid())
_exit(0);
}
}
static void
unregister(void)
{
(void)svc_unreg(YPPROG, YPVERS);
(void)svc_unreg(YPPROG, YPOLDVERS);
}
static void
reaper(int sig)
{
int status;
int saved_errno;
saved_errno = errno;
if (sig == SIGHUP) {
load_securenets();
Performance enhancements (I hope) and new stuff: yp_dblookup.c: - Implement database handle caching. What this means is that instead of opening and closing map databases for each request, we open a database and save the handle (and, if requested, the key index) in an array. This saves a bit of overhead on things like repeated YPPROC_NEXT calls, such as you'd get from getpwent(). Normally, each YPPROC_NEXT would require open()ing the database, seeking to the location supplied by the caller (which is time consuming with hash databases as the R_CURSOR flag doesn't work), reading the data, close()ing the database and then shipping the data off to the caller. The system call overhead is prohibitive, especially with very large maps. By caching the handle to an open database, we elimitate at least the open()/close() system calls, as well as the associated DB setup and tear-down operations, for a large percentage of the time. This improves performance substantially at the cost of consuming a little more memory than before. Note that all the caching support is surrounded by #ifdef DB_CACHE so that this same source module can still be used by other programs that don't need it. - Make yp_open_db() call yp_validdomain(). Doing it here saves cycles when caching is enabled since a hit on the map cache list by definition means that the domain being referenced is valid. - Also make yp_open_db() check for exhaustion of file descriptors, just in case. yp_server.c: - Reorganize things a little to take advantage of the database handle caching. Add a call to yp_flush_all() in ypproc_clear_2_svc(). - Remove calls to yp_validdomain() from some of the service procedures. yp_validdomain() is called inside yp_open_db() now, so procedures that call into the database package don't need to use yp_validdomain() themselves. - Fix a bogosity in ypproc_maplist_2_svc(): don't summarily initiallize the result.maps pointer to NULL. This causes yp_maplist_free() to fail and leaks memory. - Make ypproc_master_2_svc() copy the string it gets from the database package into a private static buffer before trying to NUL terminate it. This is necessary with the DB handle caching: stuffing a NUL into the data returned by DB package will goof it up internally. yp_main.c: - Stuff for DB handle caching: call yp_init_dbs() to clear the handle array and add call to yp_flush_all() to the SIGHUP signal handler. Makefile.yp: - Reorganize to deal with database caching. yp_mkdb(8) can now be used to send a YPPROC_CLEAR signal to ypserv(8). Call it after each map is created to refresh ypserv's cache. - Add support for mail.alias map. Contributed by Mike Murphy (mrm@sceard.com). - Make default location for the netgroups source file be /var/yp/netgroup instead of /etc/netgroup. mkaliases: - New file: script to generate mail.alias map. Contributed by Mike Murphy (mrm@sceard.com). Makefile: - Install Makefile.yp as /var/yp/Makefile.dist and link it to /var/yp/Makefile only if /var/yp/Makefile doesn't already exist. Suggested by Peter Wemm. - Install new mkaliases script in /usr/libexec along with mknetid. - Use somewhat saner approach to generating rpcgen-dependent files as suggested by Garrett Wollman.
1996-04-28 04:38:52 +00:00
#ifdef DB_CACHE
yp_flush_all();
#endif
errno = saved_errno;
return;
}
if (sig == SIGCHLD) {
while (wait3(&status, WNOHANG, NULL) > 0)
children--;
} else {
unregister();
exit(0);
}
errno = saved_errno;
return;
}
static void
usage(void)
{
2011-06-20 15:22:47 +00:00
fprintf(stderr, "usage: ypserv [-h addr] [-d] [-n] [-p path] [-P port]\n");
exit(1);
}
static void
closedown(int sig)
{
if (_rpcsvcstate == _IDLE) {
extern fd_set svc_fdset;
static int size;
int i, openfd;
if (_rpcfdtype == SOCK_DGRAM) {
unregister();
exit(0);
}
if (size == 0) {
size = getdtablesize();
}
for (i = 0, openfd = 0; i < size && openfd < 2; i++)
if (FD_ISSET(i, &svc_fdset))
openfd++;
if (openfd <= 1) {
unregister();
exit(0);
}
}
if (_rpcsvcstate == _SERVED)
_rpcsvcstate = _IDLE;
(void) signal(SIGALRM, (SIG_PF) closedown);
(void) alarm(_RPCSVC_CLOSEDOWN/2);
}
static int
create_service(const int sock, const struct netconfig *nconf,
const struct __rpc_sockinfo *si)
{
int error;
SVCXPRT *transp;
struct addrinfo hints, *res, *res0;
struct socklistent *slep;
struct bindaddrlistent *blep;
struct netbuf svcaddr;
SLIST_INIT(&sle_head);
memset(&hints, 0, sizeof(hints));
memset(&svcaddr, 0, sizeof(svcaddr));
hints.ai_family = si->si_af;
hints.ai_socktype = si->si_socktype;
hints.ai_protocol = si->si_proto;
/*
* Build socketlist from bindaddrlist.
*/
if (sock == RPC_ANYFD) {
SLIST_FOREACH(blep, &ble_head, ble_next) {
if (blep->ble_hostname == NULL)
hints.ai_flags = AI_PASSIVE;
else
hints.ai_flags = 0;
error = getaddrinfo(blep->ble_hostname, servname,
&hints, &res0);
if (error) {
_msgout("getaddrinfo(): %s",
gai_strerror(error));
return -1;
}
for (res = res0; res; res = res->ai_next) {
int s;
s = __rpc_nconf2fd(nconf);
if (s < 0) {
if (errno == EAFNOSUPPORT)
_msgout("unsupported"
" transport: %s",
nconf->nc_netid);
else
_msgout("cannot create"
" %s socket: %s",
nconf->nc_netid,
strerror(errno));
freeaddrinfo(res0);
return -1;
}
if (bindresvport_sa(s, res->ai_addr) == -1) {
if ((errno != EPERM) ||
(bind(s, res->ai_addr,
res->ai_addrlen) == -1)) {
_msgout("cannot bind "
"%s socket: %s",
nconf->nc_netid,
strerror(errno));
freeaddrinfo(res0);
close(sock);
return -1;
}
}
if (nconf->nc_semantics != NC_TPI_CLTS)
listen(s, SOMAXCONN);
slep = malloc(sizeof(*slep));
if (slep == NULL) {
_msgout("malloc failed: %s",
strerror(errno));
freeaddrinfo(res0);
close(s);
return -1;
}
memset(slep, 0, sizeof(*slep));
memcpy(&slep->sle_ss, res->ai_addr,
res->ai_addrlen);
slep->sle_sock = s;
SLIST_INSERT_HEAD(&sle_head, slep, sle_next);
/*
* If servname == "0", redefine it by using
* the bound socket.
*/
if (strncmp("0", servname, 1) == 0) {
struct sockaddr *sap;
socklen_t slen;
char *sname;
sname = malloc(NI_MAXSERV);
if (sname == NULL) {
_msgout("malloc(): %s",
strerror(errno));
freeaddrinfo(res0);
close(s);
return -1;
}
memset(sname, 0, NI_MAXSERV);
sap = (struct sockaddr *)&slep->sle_ss;
slen = sizeof(*sap);
error = getsockname(s, sap, &slen);
if (error) {
_msgout("getsockname(): %s",
strerror(errno));
freeaddrinfo(res0);
close(s);
free(sname);
return -1;
}
error = getnameinfo(sap, slen,
NULL, 0,
sname, NI_MAXSERV,
NI_NUMERICHOST | NI_NUMERICSERV);
if (error) {
_msgout("getnameinfo(): %s",
strerror(errno));
freeaddrinfo(res0);
close(s);
free(sname);
return -1;
}
servname = sname;
}
}
freeaddrinfo(res0);
}
} else {
slep = malloc(sizeof(*slep));
if (slep == NULL) {
_msgout("malloc failed: %s", strerror(errno));
return -1;
}
memset(slep, 0, sizeof(*slep));
slep->sle_sock = sock;
SLIST_INSERT_HEAD(&sle_head, slep, sle_next);
}
/*
* Traverse socketlist and create rpc service handles for each socket.
*/
SLIST_FOREACH(slep, &sle_head, sle_next) {
if (nconf->nc_semantics == NC_TPI_CLTS)
transp = svc_dg_create(slep->sle_sock, 0, 0);
else
transp = svc_vc_create(slep->sle_sock, RPC_MAXDATASIZE,
RPC_MAXDATASIZE);
if (transp == NULL) {
_msgout("unable to create service: %s",
nconf->nc_netid);
continue;
}
if (!svc_reg(transp, YPPROG, YPOLDVERS, ypprog_1, NULL)) {
svc_destroy(transp);
close(slep->sle_sock);
_msgout("unable to register (YPPROG, YPOLDVERS, %s):"
" %s", nconf->nc_netid, strerror(errno));
continue;
}
if (!svc_reg(transp, YPPROG, YPVERS, ypprog_2, NULL)) {
svc_destroy(transp);
close(slep->sle_sock);
_msgout("unable to register (YPPROG, YPVERS, %s): %s",
nconf->nc_netid, strerror(errno));
continue;
}
}
while(!(SLIST_EMPTY(&sle_head)))
SLIST_REMOVE_HEAD(&sle_head, sle_next);
/*
* Register RPC service to rpcbind by using AI_PASSIVE address.
*/
hints.ai_flags = AI_PASSIVE;
error = getaddrinfo(NULL, servname, &hints, &res0);
if (error) {
_msgout("getaddrinfo(): %s", gai_strerror(error));
return -1;
}
svcaddr.buf = res0->ai_addr;
svcaddr.len = res0->ai_addrlen;
if (si->si_af == AF_INET) {
/* XXX: ignore error intentionally */
rpcb_set(YPPROG, YPOLDVERS, nconf, &svcaddr);
}
/* XXX: ignore error intentionally */
rpcb_set(YPPROG, YPVERS, nconf, &svcaddr);
freeaddrinfo(res0);
return 0;
}
int
main(int argc, char *argv[])
{
int ch;
int error;
int ntrans;
void *nc_handle;
struct netconfig *nconf;
struct __rpc_sockinfo si;
struct bindaddrlistent *blep;
memset(&si, 0, sizeof(si));
SLIST_INIT(&ble_head);
while ((ch = getopt(argc, argv, "dh:np:P:")) != -1) {
switch (ch) {
case 'd':
debug = ypdb_debug = 1;
break;
case 'h':
blep = malloc(sizeof(*blep));
if (blep == NULL)
err(1, "malloc() failed: -h %s", optarg);
blep->ble_hostname = optarg;
SLIST_INSERT_HEAD(&ble_head, blep, ble_next);
break;
case 'n':
do_dns = 1;
break;
case 'p':
yp_dir = optarg;
break;
case 'P':
servname = optarg;
break;
default:
usage();
}
}
/*
* Add "anyaddr" entry if no -h is specified.
*/
if (SLIST_EMPTY(&ble_head)) {
blep = malloc(sizeof(*blep));
if (blep == NULL)
err(1, "malloc() failed");
memset(blep, 0, sizeof(*blep));
SLIST_INSERT_HEAD(&ble_head, blep, ble_next);
}
load_securenets();
Big round o changes: - yp_dblookup.c: Create non-DB specific database access functions. Using these allows access to the underlying database functions without needing explicit knowledge of Berkeley DB. (These are used only when DB_CACHE is #defined. Other programs that use the non-caching functions (yp_mkdb, ypxfr, yppush, rpc.yppasswdd) shouldn't notice the difference.) - yp_dnslookup: Implement async DNS lookups. We send our own DNS requests using UDP and put the request in a queue. When the response arrives, we use the ID in the header to find the corresponsing queue entry and then send the response to the client. We can go about our business and handle other YP requests in the meantime. This way, we can deal with time consuming DNS requests without blocking and without forking. - yp_server.c: Convert to using new non-DB-specific database access functions. This simplifies the code a bit and removes the need for this module to know anything about Berkeley DB. Also convert the ypproc_match_2_svc() function to use the async DNS lookup routines. - yp_main.c: tweak yp_svc_run() to add the resolver socket to the set of descriptors monitored in the select() loop. Also add a timeout to select(); we may get stale DNS requests stuck in the queue which we want to invalidate after a while. If the timeout hits, we decrement the ttl on all pending DNS requests and nuke those requests that aren't handled before ttl hits zero. - yp_extern.h: Add prototypes for new stuff. - yp_svc_udp.c (new file): The async resolver code needs to be able to rummage around inside the RPC UDP transport handle in order to work correcty. There's basically one transport handle, and each time a request comes in, the transaction ID in the handle is changed. This means that if we queue a DNS request, then we handle some other unrelated requests, we will be unable to send the DNS response because the transaction ID and remote address of the client that made the DNS request will have been lost. What we need to do is save the client address and transaction ID in the queue entry for the DNS request, then put the transaction ID and address back in the transport handle when we're ready to reply. (And then we have to undo the change so as not to confuse any other part of the server.) The trouble is that the transaction ID is hidden in an opaque part of the transport handle, and only the code in the svc_udp module in the RPC library knows how to handle it. This file contains a couple of functions that let us read and set the transaction ID in spite of this. This is really a dirty trick and I should be taken out and shot for even thinking about it, but there's no other way to get this stuff to work. - Makefile: add yp_svc_udp.c to SRCS.
1996-12-22 22:30:58 +00:00
yp_init_resolver();
Performance enhancements (I hope) and new stuff: yp_dblookup.c: - Implement database handle caching. What this means is that instead of opening and closing map databases for each request, we open a database and save the handle (and, if requested, the key index) in an array. This saves a bit of overhead on things like repeated YPPROC_NEXT calls, such as you'd get from getpwent(). Normally, each YPPROC_NEXT would require open()ing the database, seeking to the location supplied by the caller (which is time consuming with hash databases as the R_CURSOR flag doesn't work), reading the data, close()ing the database and then shipping the data off to the caller. The system call overhead is prohibitive, especially with very large maps. By caching the handle to an open database, we elimitate at least the open()/close() system calls, as well as the associated DB setup and tear-down operations, for a large percentage of the time. This improves performance substantially at the cost of consuming a little more memory than before. Note that all the caching support is surrounded by #ifdef DB_CACHE so that this same source module can still be used by other programs that don't need it. - Make yp_open_db() call yp_validdomain(). Doing it here saves cycles when caching is enabled since a hit on the map cache list by definition means that the domain being referenced is valid. - Also make yp_open_db() check for exhaustion of file descriptors, just in case. yp_server.c: - Reorganize things a little to take advantage of the database handle caching. Add a call to yp_flush_all() in ypproc_clear_2_svc(). - Remove calls to yp_validdomain() from some of the service procedures. yp_validdomain() is called inside yp_open_db() now, so procedures that call into the database package don't need to use yp_validdomain() themselves. - Fix a bogosity in ypproc_maplist_2_svc(): don't summarily initiallize the result.maps pointer to NULL. This causes yp_maplist_free() to fail and leaks memory. - Make ypproc_master_2_svc() copy the string it gets from the database package into a private static buffer before trying to NUL terminate it. This is necessary with the DB handle caching: stuffing a NUL into the data returned by DB package will goof it up internally. yp_main.c: - Stuff for DB handle caching: call yp_init_dbs() to clear the handle array and add call to yp_flush_all() to the SIGHUP signal handler. Makefile.yp: - Reorganize to deal with database caching. yp_mkdb(8) can now be used to send a YPPROC_CLEAR signal to ypserv(8). Call it after each map is created to refresh ypserv's cache. - Add support for mail.alias map. Contributed by Mike Murphy (mrm@sceard.com). - Make default location for the netgroups source file be /var/yp/netgroup instead of /etc/netgroup. mkaliases: - New file: script to generate mail.alias map. Contributed by Mike Murphy (mrm@sceard.com). Makefile: - Install Makefile.yp as /var/yp/Makefile.dist and link it to /var/yp/Makefile only if /var/yp/Makefile doesn't already exist. Suggested by Peter Wemm. - Install new mkaliases script in /usr/libexec along with mknetid. - Use somewhat saner approach to generating rpcgen-dependent files as suggested by Garrett Wollman.
1996-04-28 04:38:52 +00:00
#ifdef DB_CACHE
yp_init_dbs();
#endif
nc_handle = setnetconfig();
if (nc_handle == NULL)
err(1, "cannot read %s", NETCONFIG);
if (__rpc_fd2sockinfo(0, &si) != 0) {
/* invoked from inetd */
_rpcpmstart = 1;
_rpcfdtype = si.si_socktype;
_rpcaf = si.si_af;
_rpcfd = 0;
openlog("ypserv", LOG_PID, LOG_DAEMON);
} else {
/* standalone mode */
if (!debug) {
if (daemon(0,0)) {
err(1,"cannot fork");
}
openlog("ypserv", LOG_PID, LOG_DAEMON);
}
_rpcpmstart = 0;
_rpcaf = AF_INET;
_rpcfd = RPC_ANYFD;
unregister();
}
if (madvise(NULL, 0, MADV_PROTECT) != 0)
_msgout("madvise(): %s", strerror(errno));
/*
* Create RPC service for each transport.
*/
ntrans = 0;
while((nconf = getnetconfig(nc_handle))) {
if ((nconf->nc_flag & NC_VISIBLE)) {
if (__rpc_nconf2sockinfo(nconf, &si) == 0) {
_msgout("cannot get information for %s. "
"Ignored.", nconf->nc_netid);
continue;
}
if (_rpcpmstart) {
if (si.si_socktype != _rpcfdtype ||
si.si_af != _rpcaf)
continue;
} else if (si.si_af != _rpcaf)
continue;
error = create_service(_rpcfd, nconf, &si);
if (error) {
endnetconfig(nc_handle);
exit(1);
}
ntrans++;
}
}
endnetconfig(nc_handle);
while(!(SLIST_EMPTY(&ble_head)))
SLIST_REMOVE_HEAD(&ble_head, ble_next);
if (ntrans == 0) {
_msgout("no transport is available. Aborted.");
exit(1);
}
if (_rpcpmstart) {
(void) signal(SIGALRM, (SIG_PF) closedown);
(void) alarm(_RPCSVC_CLOSEDOWN/2);
}
/*
* Make sure SIGPIPE doesn't blow us away while servicing TCP
* connections.
*/
(void) signal(SIGPIPE, SIG_IGN);
(void) signal(SIGCHLD, (SIG_PF) reaper);
(void) signal(SIGTERM, (SIG_PF) reaper);
(void) signal(SIGINT, (SIG_PF) reaper);
(void) signal(SIGHUP, (SIG_PF) reaper);
yp_svc_run();
_msgout("svc_run returned");
exit(1);
/* NOTREACHED */
}