svc.c revision 193603
1/*	$NetBSD: svc.c,v 1.21 2000/07/06 03:10:35 christos Exp $	*/
2
3/*
4 * Sun RPC is a product of Sun Microsystems, Inc. and is provided for
5 * unrestricted use provided that this legend is included on all tape
6 * media and as a part of the software program in whole or part.  Users
7 * may copy or modify Sun RPC without charge, but are not authorized
8 * to license or distribute it to anyone else except as part of a product or
9 * program developed by the user.
10 *
11 * SUN RPC IS PROVIDED AS IS WITH NO WARRANTIES OF ANY KIND INCLUDING THE
12 * WARRANTIES OF DESIGN, MERCHANTIBILITY AND FITNESS FOR A PARTICULAR
13 * PURPOSE, OR ARISING FROM A COURSE OF DEALING, USAGE OR TRADE PRACTICE.
14 *
15 * Sun RPC is provided with no support and without any obligation on the
16 * part of Sun Microsystems, Inc. to assist in its use, correction,
17 * modification or enhancement.
18 *
19 * SUN MICROSYSTEMS, INC. SHALL HAVE NO LIABILITY WITH RESPECT TO THE
20 * INFRINGEMENT OF COPYRIGHTS, TRADE SECRETS OR ANY PATENTS BY SUN RPC
21 * OR ANY PART THEREOF.
22 *
23 * In no event will Sun Microsystems, Inc. be liable for any lost revenue
24 * or profits or other special, indirect and consequential damages, even if
25 * Sun has been advised of the possibility of such damages.
26 *
27 * Sun Microsystems, Inc.
28 * 2550 Garcia Avenue
29 * Mountain View, California  94043
30 */
31
32#if defined(LIBC_SCCS) && !defined(lint)
33static char *sccsid2 = "@(#)svc.c 1.44 88/02/08 Copyr 1984 Sun Micro";
34static char *sccsid = "@(#)svc.c	2.4 88/08/11 4.0 RPCSRC";
35#endif
36#include <sys/cdefs.h>
37__FBSDID("$FreeBSD: head/sys/rpc/svc.c 193603 2009-06-07 01:06:56Z rmacklem $");
38
39/*
40 * svc.c, Server-side remote procedure call interface.
41 *
42 * There are two sets of procedures here.  The xprt routines are
43 * for handling transport handles.  The svc routines handle the
44 * list of service routines.
45 *
46 * Copyright (C) 1984, Sun Microsystems, Inc.
47 */
48
49#include <sys/param.h>
50#include <sys/lock.h>
51#include <sys/kernel.h>
52#include <sys/kthread.h>
53#include <sys/malloc.h>
54#include <sys/mbuf.h>
55#include <sys/mutex.h>
56#include <sys/proc.h>
57#include <sys/queue.h>
58#include <sys/socketvar.h>
59#include <sys/systm.h>
60#include <sys/ucred.h>
61
62#include <rpc/rpc.h>
63#include <rpc/rpcb_clnt.h>
64#include <rpc/replay.h>
65
66#include <rpc/rpc_com.h>
67
68#define SVC_VERSQUIET 0x0001		/* keep quiet about vers mismatch */
69#define version_keepquiet(xp) (SVC_EXT(xp)->xp_flags & SVC_VERSQUIET)
70
71static struct svc_callout *svc_find(SVCPOOL *pool, rpcprog_t, rpcvers_t,
72    char *);
73static void svc_new_thread(SVCPOOL *pool);
74static void xprt_unregister_locked(SVCXPRT *xprt);
75
76/* ***************  SVCXPRT related stuff **************** */
77
78static int svcpool_minthread_sysctl(SYSCTL_HANDLER_ARGS);
79static int svcpool_maxthread_sysctl(SYSCTL_HANDLER_ARGS);
80
81SVCPOOL*
82svcpool_create(const char *name, struct sysctl_oid_list *sysctl_base)
83{
84	SVCPOOL *pool;
85
86	pool = malloc(sizeof(SVCPOOL), M_RPC, M_WAITOK|M_ZERO);
87
88	mtx_init(&pool->sp_lock, "sp_lock", NULL, MTX_DEF);
89	pool->sp_name = name;
90	pool->sp_state = SVCPOOL_INIT;
91	pool->sp_proc = NULL;
92	TAILQ_INIT(&pool->sp_xlist);
93	TAILQ_INIT(&pool->sp_active);
94	TAILQ_INIT(&pool->sp_callouts);
95	LIST_INIT(&pool->sp_threads);
96	LIST_INIT(&pool->sp_idlethreads);
97	pool->sp_minthreads = 1;
98	pool->sp_maxthreads = 1;
99	pool->sp_threadcount = 0;
100
101	/*
102	 * Don't use more than a quarter of mbuf clusters or more than
103	 * 45Mb buffering requests.
104	 */
105	pool->sp_space_high = nmbclusters * MCLBYTES / 4;
106	if (pool->sp_space_high > 45 << 20)
107		pool->sp_space_high = 45 << 20;
108	pool->sp_space_low = 2 * pool->sp_space_high / 3;
109
110	sysctl_ctx_init(&pool->sp_sysctl);
111	if (sysctl_base) {
112		SYSCTL_ADD_PROC(&pool->sp_sysctl, sysctl_base, OID_AUTO,
113		    "minthreads", CTLTYPE_INT | CTLFLAG_RW,
114		    pool, 0, svcpool_minthread_sysctl, "I", "");
115		SYSCTL_ADD_PROC(&pool->sp_sysctl, sysctl_base, OID_AUTO,
116		    "maxthreads", CTLTYPE_INT | CTLFLAG_RW,
117		    pool, 0, svcpool_maxthread_sysctl, "I", "");
118		SYSCTL_ADD_INT(&pool->sp_sysctl, sysctl_base, OID_AUTO,
119		    "threads", CTLFLAG_RD, &pool->sp_threadcount, 0, "");
120
121		SYSCTL_ADD_UINT(&pool->sp_sysctl, sysctl_base, OID_AUTO,
122		    "request_space_used", CTLFLAG_RD,
123		    &pool->sp_space_used, 0,
124		    "Space in parsed but not handled requests.");
125
126		SYSCTL_ADD_UINT(&pool->sp_sysctl, sysctl_base, OID_AUTO,
127		    "request_space_used_highest", CTLFLAG_RD,
128		    &pool->sp_space_used_highest, 0,
129		    "Highest space used since reboot.");
130
131		SYSCTL_ADD_UINT(&pool->sp_sysctl, sysctl_base, OID_AUTO,
132		    "request_space_high", CTLFLAG_RW,
133		    &pool->sp_space_high, 0,
134		    "Maximum space in parsed but not handled requests.");
135
136		SYSCTL_ADD_UINT(&pool->sp_sysctl, sysctl_base, OID_AUTO,
137		    "request_space_low", CTLFLAG_RW,
138		    &pool->sp_space_low, 0,
139		    "Low water mark for request space.");
140
141		SYSCTL_ADD_UINT(&pool->sp_sysctl, sysctl_base, OID_AUTO,
142		    "request_space_throttled", CTLFLAG_RD,
143		    &pool->sp_space_throttled, 0,
144		    "Whether nfs requests are currently throttled");
145
146		SYSCTL_ADD_UINT(&pool->sp_sysctl, sysctl_base, OID_AUTO,
147		    "request_space_throttle_count", CTLFLAG_RD,
148		    &pool->sp_space_throttle_count, 0,
149		    "Count of times throttling based on request space has occurred");
150	}
151
152	return pool;
153}
154
155void
156svcpool_destroy(SVCPOOL *pool)
157{
158	SVCXPRT *xprt, *nxprt;
159	struct svc_callout *s;
160	struct svcxprt_list cleanup;
161
162	TAILQ_INIT(&cleanup);
163	mtx_lock(&pool->sp_lock);
164
165	while (TAILQ_FIRST(&pool->sp_xlist)) {
166		xprt = TAILQ_FIRST(&pool->sp_xlist);
167		xprt_unregister_locked(xprt);
168		TAILQ_INSERT_TAIL(&cleanup, xprt, xp_link);
169	}
170
171	while (TAILQ_FIRST(&pool->sp_callouts)) {
172		s = TAILQ_FIRST(&pool->sp_callouts);
173		mtx_unlock(&pool->sp_lock);
174		svc_unreg(pool, s->sc_prog, s->sc_vers);
175		mtx_lock(&pool->sp_lock);
176	}
177	mtx_unlock(&pool->sp_lock);
178
179	TAILQ_FOREACH_SAFE(xprt, &cleanup, xp_link, nxprt) {
180		SVC_RELEASE(xprt);
181	}
182
183	mtx_destroy(&pool->sp_lock);
184
185	if (pool->sp_rcache)
186		replay_freecache(pool->sp_rcache);
187
188	sysctl_ctx_free(&pool->sp_sysctl);
189	free(pool, M_RPC);
190}
191
192static bool_t
193svcpool_active(SVCPOOL *pool)
194{
195	enum svcpool_state state = pool->sp_state;
196
197	if (state == SVCPOOL_INIT || state == SVCPOOL_CLOSING)
198		return (FALSE);
199	return (TRUE);
200}
201
202/*
203 * Sysctl handler to set the minimum thread count on a pool
204 */
205static int
206svcpool_minthread_sysctl(SYSCTL_HANDLER_ARGS)
207{
208	SVCPOOL *pool;
209	int newminthreads, error, n;
210
211	pool = oidp->oid_arg1;
212	newminthreads = pool->sp_minthreads;
213	error = sysctl_handle_int(oidp, &newminthreads, 0, req);
214	if (error == 0 && newminthreads != pool->sp_minthreads) {
215		if (newminthreads > pool->sp_maxthreads)
216			return (EINVAL);
217		mtx_lock(&pool->sp_lock);
218		if (newminthreads > pool->sp_minthreads
219		    && svcpool_active(pool)) {
220			/*
221			 * If the pool is running and we are
222			 * increasing, create some more threads now.
223			 */
224			n = newminthreads - pool->sp_threadcount;
225			if (n > 0) {
226				mtx_unlock(&pool->sp_lock);
227				while (n--)
228					svc_new_thread(pool);
229				mtx_lock(&pool->sp_lock);
230			}
231		}
232		pool->sp_minthreads = newminthreads;
233		mtx_unlock(&pool->sp_lock);
234	}
235	return (error);
236}
237
238/*
239 * Sysctl handler to set the maximum thread count on a pool
240 */
241static int
242svcpool_maxthread_sysctl(SYSCTL_HANDLER_ARGS)
243{
244	SVCPOOL *pool;
245	SVCTHREAD *st;
246	int newmaxthreads, error;
247
248	pool = oidp->oid_arg1;
249	newmaxthreads = pool->sp_maxthreads;
250	error = sysctl_handle_int(oidp, &newmaxthreads, 0, req);
251	if (error == 0 && newmaxthreads != pool->sp_maxthreads) {
252		if (newmaxthreads < pool->sp_minthreads)
253			return (EINVAL);
254		mtx_lock(&pool->sp_lock);
255		if (newmaxthreads < pool->sp_maxthreads
256		    && svcpool_active(pool)) {
257			/*
258			 * If the pool is running and we are
259			 * decreasing, wake up some idle threads to
260			 * encourage them to exit.
261			 */
262			LIST_FOREACH(st, &pool->sp_idlethreads, st_ilink)
263				cv_signal(&st->st_cond);
264		}
265		pool->sp_maxthreads = newmaxthreads;
266		mtx_unlock(&pool->sp_lock);
267	}
268	return (error);
269}
270
271/*
272 * Activate a transport handle.
273 */
274void
275xprt_register(SVCXPRT *xprt)
276{
277	SVCPOOL *pool = xprt->xp_pool;
278
279	mtx_lock(&pool->sp_lock);
280	xprt->xp_registered = TRUE;
281	xprt->xp_active = FALSE;
282	TAILQ_INSERT_TAIL(&pool->sp_xlist, xprt, xp_link);
283	mtx_unlock(&pool->sp_lock);
284}
285
286/*
287 * De-activate a transport handle. Note: the locked version doesn't
288 * release the transport - caller must do that after dropping the pool
289 * lock.
290 */
291static void
292xprt_unregister_locked(SVCXPRT *xprt)
293{
294	SVCPOOL *pool = xprt->xp_pool;
295
296	if (xprt->xp_active) {
297		TAILQ_REMOVE(&pool->sp_active, xprt, xp_alink);
298		xprt->xp_active = FALSE;
299	}
300	TAILQ_REMOVE(&pool->sp_xlist, xprt, xp_link);
301	xprt->xp_registered = FALSE;
302}
303
304void
305xprt_unregister(SVCXPRT *xprt)
306{
307	SVCPOOL *pool = xprt->xp_pool;
308
309	mtx_lock(&pool->sp_lock);
310	xprt_unregister_locked(xprt);
311	mtx_unlock(&pool->sp_lock);
312
313	SVC_RELEASE(xprt);
314}
315
316static void
317xprt_assignthread(SVCXPRT *xprt)
318{
319	SVCPOOL *pool = xprt->xp_pool;
320	SVCTHREAD *st;
321
322	/*
323	 * Attempt to assign a service thread to this
324	 * transport.
325	 */
326	LIST_FOREACH(st, &pool->sp_idlethreads, st_ilink) {
327		if (st->st_xprt == NULL && STAILQ_EMPTY(&st->st_reqs))
328			break;
329	}
330	if (st) {
331		SVC_ACQUIRE(xprt);
332		xprt->xp_thread = st;
333		st->st_xprt = xprt;
334		cv_signal(&st->st_cond);
335	} else {
336		/*
337		 * See if we can create a new thread. The
338		 * actual thread creation happens in
339		 * svc_run_internal because our locking state
340		 * is poorly defined (we are typically called
341		 * from a socket upcall). Don't create more
342		 * than one thread per second.
343		 */
344		if (pool->sp_state == SVCPOOL_ACTIVE
345		    && pool->sp_lastcreatetime < time_uptime
346		    && pool->sp_threadcount < pool->sp_maxthreads) {
347			pool->sp_state = SVCPOOL_THREADWANTED;
348		}
349	}
350}
351
352void
353xprt_active(SVCXPRT *xprt)
354{
355	SVCPOOL *pool = xprt->xp_pool;
356
357	mtx_lock(&pool->sp_lock);
358
359	if (!xprt->xp_registered) {
360		/*
361		 * Race with xprt_unregister - we lose.
362		 */
363		mtx_unlock(&pool->sp_lock);
364		return;
365	}
366
367	if (!xprt->xp_active) {
368		TAILQ_INSERT_TAIL(&pool->sp_active, xprt, xp_alink);
369		xprt->xp_active = TRUE;
370		xprt_assignthread(xprt);
371	}
372
373	mtx_unlock(&pool->sp_lock);
374}
375
376void
377xprt_inactive_locked(SVCXPRT *xprt)
378{
379	SVCPOOL *pool = xprt->xp_pool;
380
381	if (xprt->xp_active) {
382		TAILQ_REMOVE(&pool->sp_active, xprt, xp_alink);
383		xprt->xp_active = FALSE;
384	}
385}
386
387void
388xprt_inactive(SVCXPRT *xprt)
389{
390	SVCPOOL *pool = xprt->xp_pool;
391
392	mtx_lock(&pool->sp_lock);
393	xprt_inactive_locked(xprt);
394	mtx_unlock(&pool->sp_lock);
395}
396
397/*
398 * Add a service program to the callout list.
399 * The dispatch routine will be called when a rpc request for this
400 * program number comes in.
401 */
402bool_t
403svc_reg(SVCXPRT *xprt, const rpcprog_t prog, const rpcvers_t vers,
404    void (*dispatch)(struct svc_req *, SVCXPRT *),
405    const struct netconfig *nconf)
406{
407	SVCPOOL *pool = xprt->xp_pool;
408	struct svc_callout *s;
409	char *netid = NULL;
410	int flag = 0;
411
412/* VARIABLES PROTECTED BY svc_lock: s, svc_head */
413
414	if (xprt->xp_netid) {
415		netid = strdup(xprt->xp_netid, M_RPC);
416		flag = 1;
417	} else if (nconf && nconf->nc_netid) {
418		netid = strdup(nconf->nc_netid, M_RPC);
419		flag = 1;
420	} /* must have been created with svc_raw_create */
421	if ((netid == NULL) && (flag == 1)) {
422		return (FALSE);
423	}
424
425	mtx_lock(&pool->sp_lock);
426	if ((s = svc_find(pool, prog, vers, netid)) != NULL) {
427		if (netid)
428			free(netid, M_RPC);
429		if (s->sc_dispatch == dispatch)
430			goto rpcb_it; /* he is registering another xptr */
431		mtx_unlock(&pool->sp_lock);
432		return (FALSE);
433	}
434	s = malloc(sizeof (struct svc_callout), M_RPC, M_NOWAIT);
435	if (s == NULL) {
436		if (netid)
437			free(netid, M_RPC);
438		mtx_unlock(&pool->sp_lock);
439		return (FALSE);
440	}
441
442	s->sc_prog = prog;
443	s->sc_vers = vers;
444	s->sc_dispatch = dispatch;
445	s->sc_netid = netid;
446	TAILQ_INSERT_TAIL(&pool->sp_callouts, s, sc_link);
447
448	if ((xprt->xp_netid == NULL) && (flag == 1) && netid)
449		((SVCXPRT *) xprt)->xp_netid = strdup(netid, M_RPC);
450
451rpcb_it:
452	mtx_unlock(&pool->sp_lock);
453	/* now register the information with the local binder service */
454	if (nconf) {
455		bool_t dummy;
456		struct netconfig tnc;
457		struct netbuf nb;
458		tnc = *nconf;
459		nb.buf = &xprt->xp_ltaddr;
460		nb.len = xprt->xp_ltaddr.ss_len;
461		dummy = rpcb_set(prog, vers, &tnc, &nb);
462		return (dummy);
463	}
464	return (TRUE);
465}
466
467/*
468 * Remove a service program from the callout list.
469 */
470void
471svc_unreg(SVCPOOL *pool, const rpcprog_t prog, const rpcvers_t vers)
472{
473	struct svc_callout *s;
474
475	/* unregister the information anyway */
476	(void) rpcb_unset(prog, vers, NULL);
477	mtx_lock(&pool->sp_lock);
478	while ((s = svc_find(pool, prog, vers, NULL)) != NULL) {
479		TAILQ_REMOVE(&pool->sp_callouts, s, sc_link);
480		if (s->sc_netid)
481			mem_free(s->sc_netid, sizeof (s->sc_netid) + 1);
482		mem_free(s, sizeof (struct svc_callout));
483	}
484	mtx_unlock(&pool->sp_lock);
485}
486
487/* ********************** CALLOUT list related stuff ************* */
488
489/*
490 * Search the callout list for a program number, return the callout
491 * struct.
492 */
493static struct svc_callout *
494svc_find(SVCPOOL *pool, rpcprog_t prog, rpcvers_t vers, char *netid)
495{
496	struct svc_callout *s;
497
498	mtx_assert(&pool->sp_lock, MA_OWNED);
499	TAILQ_FOREACH(s, &pool->sp_callouts, sc_link) {
500		if (s->sc_prog == prog && s->sc_vers == vers
501		    && (netid == NULL || s->sc_netid == NULL ||
502			strcmp(netid, s->sc_netid) == 0))
503			break;
504	}
505
506	return (s);
507}
508
509/* ******************* REPLY GENERATION ROUTINES  ************ */
510
511static bool_t
512svc_sendreply_common(struct svc_req *rqstp, struct rpc_msg *rply,
513    struct mbuf *body)
514{
515	SVCXPRT *xprt = rqstp->rq_xprt;
516	bool_t ok;
517
518	if (rqstp->rq_args) {
519		m_freem(rqstp->rq_args);
520		rqstp->rq_args = NULL;
521	}
522
523	if (xprt->xp_pool->sp_rcache)
524		replay_setreply(xprt->xp_pool->sp_rcache,
525		    rply, svc_getrpccaller(rqstp), body);
526
527	if (!SVCAUTH_WRAP(&rqstp->rq_auth, &body))
528		return (FALSE);
529
530	ok = SVC_REPLY(xprt, rply, rqstp->rq_addr, body);
531	if (rqstp->rq_addr) {
532		free(rqstp->rq_addr, M_SONAME);
533		rqstp->rq_addr = NULL;
534	}
535
536	return (ok);
537}
538
539/*
540 * Send a reply to an rpc request
541 */
542bool_t
543svc_sendreply(struct svc_req *rqstp, xdrproc_t xdr_results, void * xdr_location)
544{
545	struct rpc_msg rply;
546	struct mbuf *m;
547	XDR xdrs;
548	bool_t ok;
549
550	rply.rm_xid = rqstp->rq_xid;
551	rply.rm_direction = REPLY;
552	rply.rm_reply.rp_stat = MSG_ACCEPTED;
553	rply.acpted_rply.ar_verf = rqstp->rq_verf;
554	rply.acpted_rply.ar_stat = SUCCESS;
555	rply.acpted_rply.ar_results.where = NULL;
556	rply.acpted_rply.ar_results.proc = (xdrproc_t) xdr_void;
557
558	MGET(m, M_WAIT, MT_DATA);
559	MCLGET(m, M_WAIT);
560	m->m_len = 0;
561	xdrmbuf_create(&xdrs, m, XDR_ENCODE);
562	ok = xdr_results(&xdrs, xdr_location);
563	XDR_DESTROY(&xdrs);
564
565	if (ok) {
566		return (svc_sendreply_common(rqstp, &rply, m));
567	} else {
568		m_freem(m);
569		return (FALSE);
570	}
571}
572
573bool_t
574svc_sendreply_mbuf(struct svc_req *rqstp, struct mbuf *m)
575{
576	struct rpc_msg rply;
577
578	rply.rm_xid = rqstp->rq_xid;
579	rply.rm_direction = REPLY;
580	rply.rm_reply.rp_stat = MSG_ACCEPTED;
581	rply.acpted_rply.ar_verf = rqstp->rq_verf;
582	rply.acpted_rply.ar_stat = SUCCESS;
583	rply.acpted_rply.ar_results.where = NULL;
584	rply.acpted_rply.ar_results.proc = (xdrproc_t) xdr_void;
585
586	return (svc_sendreply_common(rqstp, &rply, m));
587}
588
589/*
590 * No procedure error reply
591 */
592void
593svcerr_noproc(struct svc_req *rqstp)
594{
595	SVCXPRT *xprt = rqstp->rq_xprt;
596	struct rpc_msg rply;
597
598	rply.rm_xid = rqstp->rq_xid;
599	rply.rm_direction = REPLY;
600	rply.rm_reply.rp_stat = MSG_ACCEPTED;
601	rply.acpted_rply.ar_verf = rqstp->rq_verf;
602	rply.acpted_rply.ar_stat = PROC_UNAVAIL;
603
604	if (xprt->xp_pool->sp_rcache)
605		replay_setreply(xprt->xp_pool->sp_rcache,
606		    &rply, svc_getrpccaller(rqstp), NULL);
607
608	svc_sendreply_common(rqstp, &rply, NULL);
609}
610
611/*
612 * Can't decode args error reply
613 */
614void
615svcerr_decode(struct svc_req *rqstp)
616{
617	SVCXPRT *xprt = rqstp->rq_xprt;
618	struct rpc_msg rply;
619
620	rply.rm_xid = rqstp->rq_xid;
621	rply.rm_direction = REPLY;
622	rply.rm_reply.rp_stat = MSG_ACCEPTED;
623	rply.acpted_rply.ar_verf = rqstp->rq_verf;
624	rply.acpted_rply.ar_stat = GARBAGE_ARGS;
625
626	if (xprt->xp_pool->sp_rcache)
627		replay_setreply(xprt->xp_pool->sp_rcache,
628		    &rply, (struct sockaddr *) &xprt->xp_rtaddr, NULL);
629
630	svc_sendreply_common(rqstp, &rply, NULL);
631}
632
633/*
634 * Some system error
635 */
636void
637svcerr_systemerr(struct svc_req *rqstp)
638{
639	SVCXPRT *xprt = rqstp->rq_xprt;
640	struct rpc_msg rply;
641
642	rply.rm_xid = rqstp->rq_xid;
643	rply.rm_direction = REPLY;
644	rply.rm_reply.rp_stat = MSG_ACCEPTED;
645	rply.acpted_rply.ar_verf = rqstp->rq_verf;
646	rply.acpted_rply.ar_stat = SYSTEM_ERR;
647
648	if (xprt->xp_pool->sp_rcache)
649		replay_setreply(xprt->xp_pool->sp_rcache,
650		    &rply, svc_getrpccaller(rqstp), NULL);
651
652	svc_sendreply_common(rqstp, &rply, NULL);
653}
654
655/*
656 * Authentication error reply
657 */
658void
659svcerr_auth(struct svc_req *rqstp, enum auth_stat why)
660{
661	SVCXPRT *xprt = rqstp->rq_xprt;
662	struct rpc_msg rply;
663
664	rply.rm_xid = rqstp->rq_xid;
665	rply.rm_direction = REPLY;
666	rply.rm_reply.rp_stat = MSG_DENIED;
667	rply.rjcted_rply.rj_stat = AUTH_ERROR;
668	rply.rjcted_rply.rj_why = why;
669
670	if (xprt->xp_pool->sp_rcache)
671		replay_setreply(xprt->xp_pool->sp_rcache,
672		    &rply, svc_getrpccaller(rqstp), NULL);
673
674	svc_sendreply_common(rqstp, &rply, NULL);
675}
676
677/*
678 * Auth too weak error reply
679 */
680void
681svcerr_weakauth(struct svc_req *rqstp)
682{
683
684	svcerr_auth(rqstp, AUTH_TOOWEAK);
685}
686
687/*
688 * Program unavailable error reply
689 */
690void
691svcerr_noprog(struct svc_req *rqstp)
692{
693	SVCXPRT *xprt = rqstp->rq_xprt;
694	struct rpc_msg rply;
695
696	rply.rm_xid = rqstp->rq_xid;
697	rply.rm_direction = REPLY;
698	rply.rm_reply.rp_stat = MSG_ACCEPTED;
699	rply.acpted_rply.ar_verf = rqstp->rq_verf;
700	rply.acpted_rply.ar_stat = PROG_UNAVAIL;
701
702	if (xprt->xp_pool->sp_rcache)
703		replay_setreply(xprt->xp_pool->sp_rcache,
704		    &rply, svc_getrpccaller(rqstp), NULL);
705
706	svc_sendreply_common(rqstp, &rply, NULL);
707}
708
709/*
710 * Program version mismatch error reply
711 */
712void
713svcerr_progvers(struct svc_req *rqstp, rpcvers_t low_vers, rpcvers_t high_vers)
714{
715	SVCXPRT *xprt = rqstp->rq_xprt;
716	struct rpc_msg rply;
717
718	rply.rm_xid = rqstp->rq_xid;
719	rply.rm_direction = REPLY;
720	rply.rm_reply.rp_stat = MSG_ACCEPTED;
721	rply.acpted_rply.ar_verf = rqstp->rq_verf;
722	rply.acpted_rply.ar_stat = PROG_MISMATCH;
723	rply.acpted_rply.ar_vers.low = (uint32_t)low_vers;
724	rply.acpted_rply.ar_vers.high = (uint32_t)high_vers;
725
726	if (xprt->xp_pool->sp_rcache)
727		replay_setreply(xprt->xp_pool->sp_rcache,
728		    &rply, svc_getrpccaller(rqstp), NULL);
729
730	svc_sendreply_common(rqstp, &rply, NULL);
731}
732
733/*
734 * Allocate a new server transport structure. All fields are
735 * initialized to zero and xp_p3 is initialized to point at an
736 * extension structure to hold various flags and authentication
737 * parameters.
738 */
739SVCXPRT *
740svc_xprt_alloc()
741{
742	SVCXPRT *xprt;
743	SVCXPRT_EXT *ext;
744
745	xprt = mem_alloc(sizeof(SVCXPRT));
746	memset(xprt, 0, sizeof(SVCXPRT));
747	ext = mem_alloc(sizeof(SVCXPRT_EXT));
748	memset(ext, 0, sizeof(SVCXPRT_EXT));
749	xprt->xp_p3 = ext;
750	refcount_init(&xprt->xp_refs, 1);
751
752	return (xprt);
753}
754
755/*
756 * Free a server transport structure.
757 */
758void
759svc_xprt_free(xprt)
760	SVCXPRT *xprt;
761{
762
763	mem_free(xprt->xp_p3, sizeof(SVCXPRT_EXT));
764	mem_free(xprt, sizeof(SVCXPRT));
765}
766
767/* ******************* SERVER INPUT STUFF ******************* */
768
769/*
770 * Read RPC requests from a transport and queue them to be
771 * executed. We handle authentication and replay cache replies here.
772 * Actually dispatching the RPC is deferred till svc_executereq.
773 */
774static enum xprt_stat
775svc_getreq(SVCXPRT *xprt, struct svc_req **rqstp_ret)
776{
777	SVCPOOL *pool = xprt->xp_pool;
778	struct svc_req *r;
779	struct rpc_msg msg;
780	struct mbuf *args;
781	enum xprt_stat stat;
782
783	/* now receive msgs from xprtprt (support batch calls) */
784	r = malloc(sizeof(*r), M_RPC, M_WAITOK|M_ZERO);
785
786	msg.rm_call.cb_cred.oa_base = r->rq_credarea;
787	msg.rm_call.cb_verf.oa_base = &r->rq_credarea[MAX_AUTH_BYTES];
788	r->rq_clntcred = &r->rq_credarea[2*MAX_AUTH_BYTES];
789	if (SVC_RECV(xprt, &msg, &r->rq_addr, &args)) {
790		enum auth_stat why;
791
792		/*
793		 * Handle replays and authenticate before queuing the
794		 * request to be executed.
795		 */
796		SVC_ACQUIRE(xprt);
797		r->rq_xprt = xprt;
798		if (pool->sp_rcache) {
799			struct rpc_msg repmsg;
800			struct mbuf *repbody;
801			enum replay_state rs;
802			rs = replay_find(pool->sp_rcache, &msg,
803			    svc_getrpccaller(r), &repmsg, &repbody);
804			switch (rs) {
805			case RS_NEW:
806				break;
807			case RS_DONE:
808				SVC_REPLY(xprt, &repmsg, r->rq_addr,
809				    repbody);
810				if (r->rq_addr) {
811					free(r->rq_addr, M_SONAME);
812					r->rq_addr = NULL;
813				}
814				goto call_done;
815
816			default:
817				goto call_done;
818			}
819		}
820
821		r->rq_xid = msg.rm_xid;
822		r->rq_prog = msg.rm_call.cb_prog;
823		r->rq_vers = msg.rm_call.cb_vers;
824		r->rq_proc = msg.rm_call.cb_proc;
825		r->rq_size = sizeof(*r) + m_length(args, NULL);
826		r->rq_args = args;
827		if ((why = _authenticate(r, &msg)) != AUTH_OK) {
828			/*
829			 * RPCSEC_GSS uses this return code
830			 * for requests that form part of its
831			 * context establishment protocol and
832			 * should not be dispatched to the
833			 * application.
834			 */
835			if (why != RPCSEC_GSS_NODISPATCH)
836				svcerr_auth(r, why);
837			goto call_done;
838		}
839
840		if (!SVCAUTH_UNWRAP(&r->rq_auth, &r->rq_args)) {
841			svcerr_decode(r);
842			goto call_done;
843		}
844
845		/*
846		 * Everything checks out, return request to caller.
847		 */
848		*rqstp_ret = r;
849		r = NULL;
850	}
851call_done:
852	if (r) {
853		svc_freereq(r);
854		r = NULL;
855	}
856	if ((stat = SVC_STAT(xprt)) == XPRT_DIED) {
857		xprt_unregister(xprt);
858	}
859
860	return (stat);
861}
862
863static void
864svc_executereq(struct svc_req *rqstp)
865{
866	SVCXPRT *xprt = rqstp->rq_xprt;
867	SVCPOOL *pool = xprt->xp_pool;
868	int prog_found;
869	rpcvers_t low_vers;
870	rpcvers_t high_vers;
871	struct svc_callout *s;
872
873	/* now match message with a registered service*/
874	prog_found = FALSE;
875	low_vers = (rpcvers_t) -1L;
876	high_vers = (rpcvers_t) 0L;
877	TAILQ_FOREACH(s, &pool->sp_callouts, sc_link) {
878		if (s->sc_prog == rqstp->rq_prog) {
879			if (s->sc_vers == rqstp->rq_vers) {
880				/*
881				 * We hand ownership of r to the
882				 * dispatch method - they must call
883				 * svc_freereq.
884				 */
885				(*s->sc_dispatch)(rqstp, xprt);
886				return;
887			}  /* found correct version */
888			prog_found = TRUE;
889			if (s->sc_vers < low_vers)
890				low_vers = s->sc_vers;
891			if (s->sc_vers > high_vers)
892				high_vers = s->sc_vers;
893		}   /* found correct program */
894	}
895
896	/*
897	 * if we got here, the program or version
898	 * is not served ...
899	 */
900	if (prog_found)
901		svcerr_progvers(rqstp, low_vers, high_vers);
902	else
903		svcerr_noprog(rqstp);
904
905	svc_freereq(rqstp);
906}
907
908static void
909svc_checkidle(SVCPOOL *pool)
910{
911	SVCXPRT *xprt, *nxprt;
912	time_t timo;
913	struct svcxprt_list cleanup;
914
915	TAILQ_INIT(&cleanup);
916	TAILQ_FOREACH_SAFE(xprt, &pool->sp_xlist, xp_link, nxprt) {
917		/*
918		 * Only some transports have idle timers. Don't time
919		 * something out which is just waking up.
920		 */
921		if (!xprt->xp_idletimeout || xprt->xp_thread)
922			continue;
923
924		timo = xprt->xp_lastactive + xprt->xp_idletimeout;
925		if (time_uptime > timo) {
926			xprt_unregister_locked(xprt);
927			TAILQ_INSERT_TAIL(&cleanup, xprt, xp_link);
928		}
929	}
930
931	mtx_unlock(&pool->sp_lock);
932	TAILQ_FOREACH_SAFE(xprt, &cleanup, xp_link, nxprt) {
933		SVC_RELEASE(xprt);
934	}
935	mtx_lock(&pool->sp_lock);
936
937}
938
939static void
940svc_assign_waiting_sockets(SVCPOOL *pool)
941{
942	SVCXPRT *xprt;
943
944	TAILQ_FOREACH(xprt, &pool->sp_active, xp_alink) {
945		if (!xprt->xp_thread) {
946			xprt_assignthread(xprt);
947		}
948	}
949}
950
951static bool_t
952svc_request_space_available(SVCPOOL *pool)
953{
954
955	mtx_assert(&pool->sp_lock, MA_OWNED);
956
957	if (pool->sp_space_throttled) {
958		/*
959		 * Below the low-water yet? If so, assign any waiting sockets.
960		 */
961		if (pool->sp_space_used < pool->sp_space_low) {
962			pool->sp_space_throttled = FALSE;
963			svc_assign_waiting_sockets(pool);
964			return TRUE;
965		}
966
967		return FALSE;
968	} else {
969		if (pool->sp_space_used
970		    >= pool->sp_space_high) {
971			pool->sp_space_throttled = TRUE;
972			pool->sp_space_throttle_count++;
973			return FALSE;
974		}
975
976		return TRUE;
977	}
978}
979
980static void
981svc_run_internal(SVCPOOL *pool, bool_t ismaster)
982{
983	SVCTHREAD *st, *stpref;
984	SVCXPRT *xprt;
985	enum xprt_stat stat;
986	struct svc_req *rqstp;
987	int error;
988
989	st = mem_alloc(sizeof(*st));
990	st->st_xprt = NULL;
991	STAILQ_INIT(&st->st_reqs);
992	cv_init(&st->st_cond, "rpcsvc");
993
994	mtx_lock(&pool->sp_lock);
995	LIST_INSERT_HEAD(&pool->sp_threads, st, st_link);
996
997	/*
998	 * If we are a new thread which was spawned to cope with
999	 * increased load, set the state back to SVCPOOL_ACTIVE.
1000	 */
1001	if (pool->sp_state == SVCPOOL_THREADSTARTING)
1002		pool->sp_state = SVCPOOL_ACTIVE;
1003
1004	while (pool->sp_state != SVCPOOL_CLOSING) {
1005		/*
1006		 * Check for idle transports once per second.
1007		 */
1008		if (time_uptime > pool->sp_lastidlecheck) {
1009			pool->sp_lastidlecheck = time_uptime;
1010			svc_checkidle(pool);
1011		}
1012
1013		xprt = st->st_xprt;
1014		if (!xprt && STAILQ_EMPTY(&st->st_reqs)) {
1015			/*
1016			 * Enforce maxthreads count.
1017			 */
1018			if (pool->sp_threadcount > pool->sp_maxthreads)
1019				break;
1020
1021			/*
1022			 * Before sleeping, see if we can find an
1023			 * active transport which isn't being serviced
1024			 * by a thread.
1025			 */
1026			if (svc_request_space_available(pool)) {
1027				TAILQ_FOREACH(xprt, &pool->sp_active,
1028				    xp_alink) {
1029					if (!xprt->xp_thread) {
1030						SVC_ACQUIRE(xprt);
1031						xprt->xp_thread = st;
1032						st->st_xprt = xprt;
1033						break;
1034					}
1035				}
1036			}
1037			if (st->st_xprt)
1038				continue;
1039
1040			LIST_INSERT_HEAD(&pool->sp_idlethreads, st, st_ilink);
1041			error = cv_timedwait_sig(&st->st_cond, &pool->sp_lock,
1042				5 * hz);
1043			LIST_REMOVE(st, st_ilink);
1044
1045			/*
1046			 * Reduce worker thread count when idle.
1047			 */
1048			if (error == EWOULDBLOCK) {
1049				if (!ismaster
1050				    && (pool->sp_threadcount
1051					> pool->sp_minthreads)
1052					&& !st->st_xprt
1053					&& STAILQ_EMPTY(&st->st_reqs))
1054					break;
1055			}
1056			if (error == EWOULDBLOCK)
1057				continue;
1058			if (error) {
1059				if (pool->sp_state != SVCPOOL_CLOSING) {
1060					mtx_unlock(&pool->sp_lock);
1061					svc_exit(pool);
1062					mtx_lock(&pool->sp_lock);
1063				}
1064				break;
1065			}
1066
1067			if (pool->sp_state == SVCPOOL_THREADWANTED) {
1068				pool->sp_state = SVCPOOL_THREADSTARTING;
1069				pool->sp_lastcreatetime = time_uptime;
1070				mtx_unlock(&pool->sp_lock);
1071				svc_new_thread(pool);
1072				mtx_lock(&pool->sp_lock);
1073			}
1074			continue;
1075		}
1076
1077		if (xprt) {
1078			/*
1079			 * Drain the transport socket and queue up any
1080			 * RPCs.
1081			 */
1082			xprt->xp_lastactive = time_uptime;
1083			stat = XPRT_IDLE;
1084			do {
1085				if (!svc_request_space_available(pool))
1086					break;
1087				rqstp = NULL;
1088				mtx_unlock(&pool->sp_lock);
1089				stat = svc_getreq(xprt, &rqstp);
1090				mtx_lock(&pool->sp_lock);
1091				if (rqstp) {
1092					/*
1093					 * See if the application has
1094					 * a preference for some other
1095					 * thread.
1096					 */
1097					stpref = st;
1098					if (pool->sp_assign)
1099						stpref = pool->sp_assign(st,
1100						    rqstp);
1101
1102					pool->sp_space_used +=
1103						rqstp->rq_size;
1104					if (pool->sp_space_used
1105					    > pool->sp_space_used_highest)
1106						pool->sp_space_used_highest =
1107							pool->sp_space_used;
1108					rqstp->rq_thread = stpref;
1109					STAILQ_INSERT_TAIL(&stpref->st_reqs,
1110					    rqstp, rq_link);
1111					stpref->st_reqcount++;
1112
1113					/*
1114					 * If we assigned the request
1115					 * to another thread, make
1116					 * sure its awake and continue
1117					 * reading from the
1118					 * socket. Otherwise, try to
1119					 * find some other thread to
1120					 * read from the socket and
1121					 * execute the request
1122					 * immediately.
1123					 */
1124					if (stpref != st) {
1125						cv_signal(&stpref->st_cond);
1126						continue;
1127					} else {
1128						break;
1129					}
1130				}
1131			} while (stat == XPRT_MOREREQS
1132			    && pool->sp_state != SVCPOOL_CLOSING);
1133
1134			/*
1135			 * Move this transport to the end of the
1136			 * active list to ensure fairness when
1137			 * multiple transports are active. If this was
1138			 * the last queued request, svc_getreq will
1139			 * end up calling xprt_inactive to remove from
1140			 * the active list.
1141			 */
1142			xprt->xp_thread = NULL;
1143			st->st_xprt = NULL;
1144			if (xprt->xp_active) {
1145				xprt_assignthread(xprt);
1146				TAILQ_REMOVE(&pool->sp_active, xprt, xp_alink);
1147				TAILQ_INSERT_TAIL(&pool->sp_active, xprt,
1148				    xp_alink);
1149			}
1150			mtx_unlock(&pool->sp_lock);
1151			SVC_RELEASE(xprt);
1152			mtx_lock(&pool->sp_lock);
1153		}
1154
1155		/*
1156		 * Execute what we have queued.
1157		 */
1158		while ((rqstp = STAILQ_FIRST(&st->st_reqs)) != NULL) {
1159			size_t sz = rqstp->rq_size;
1160			mtx_unlock(&pool->sp_lock);
1161			svc_executereq(rqstp);
1162			mtx_lock(&pool->sp_lock);
1163			pool->sp_space_used -= sz;
1164		}
1165	}
1166
1167	if (st->st_xprt) {
1168		xprt = st->st_xprt;
1169		st->st_xprt = NULL;
1170		SVC_RELEASE(xprt);
1171	}
1172
1173	KASSERT(STAILQ_EMPTY(&st->st_reqs), ("stray reqs on exit"));
1174	LIST_REMOVE(st, st_link);
1175	pool->sp_threadcount--;
1176
1177	mtx_unlock(&pool->sp_lock);
1178
1179	cv_destroy(&st->st_cond);
1180	mem_free(st, sizeof(*st));
1181
1182	if (!ismaster)
1183		wakeup(pool);
1184}
1185
1186static void
1187svc_thread_start(void *arg)
1188{
1189
1190	svc_run_internal((SVCPOOL *) arg, FALSE);
1191	kthread_exit();
1192}
1193
1194static void
1195svc_new_thread(SVCPOOL *pool)
1196{
1197	struct thread *td;
1198
1199	pool->sp_threadcount++;
1200	kthread_add(svc_thread_start, pool,
1201	    pool->sp_proc, &td, 0, 0,
1202	    "%s: service", pool->sp_name);
1203}
1204
1205void
1206svc_run(SVCPOOL *pool)
1207{
1208	int i;
1209	struct proc *p;
1210	struct thread *td;
1211
1212	p = curproc;
1213	td = curthread;
1214	snprintf(td->td_name, sizeof(td->td_name),
1215	    "%s: master", pool->sp_name);
1216	pool->sp_state = SVCPOOL_ACTIVE;
1217	pool->sp_proc = p;
1218	pool->sp_lastcreatetime = time_uptime;
1219	pool->sp_threadcount = 1;
1220
1221	for (i = 1; i < pool->sp_minthreads; i++) {
1222		svc_new_thread(pool);
1223	}
1224
1225	svc_run_internal(pool, TRUE);
1226
1227	mtx_lock(&pool->sp_lock);
1228	while (pool->sp_threadcount > 0)
1229		msleep(pool, &pool->sp_lock, 0, "svcexit", 0);
1230	mtx_unlock(&pool->sp_lock);
1231}
1232
1233void
1234svc_exit(SVCPOOL *pool)
1235{
1236	SVCTHREAD *st;
1237
1238	mtx_lock(&pool->sp_lock);
1239
1240	pool->sp_state = SVCPOOL_CLOSING;
1241	LIST_FOREACH(st, &pool->sp_idlethreads, st_ilink)
1242		cv_signal(&st->st_cond);
1243
1244	mtx_unlock(&pool->sp_lock);
1245}
1246
1247bool_t
1248svc_getargs(struct svc_req *rqstp, xdrproc_t xargs, void *args)
1249{
1250	struct mbuf *m;
1251	XDR xdrs;
1252	bool_t stat;
1253
1254	m = rqstp->rq_args;
1255	rqstp->rq_args = NULL;
1256
1257	xdrmbuf_create(&xdrs, m, XDR_DECODE);
1258	stat = xargs(&xdrs, args);
1259	XDR_DESTROY(&xdrs);
1260
1261	return (stat);
1262}
1263
1264bool_t
1265svc_freeargs(struct svc_req *rqstp, xdrproc_t xargs, void *args)
1266{
1267	XDR xdrs;
1268
1269	if (rqstp->rq_addr) {
1270		free(rqstp->rq_addr, M_SONAME);
1271		rqstp->rq_addr = NULL;
1272	}
1273
1274	xdrs.x_op = XDR_FREE;
1275	return (xargs(&xdrs, args));
1276}
1277
1278void
1279svc_freereq(struct svc_req *rqstp)
1280{
1281	SVCTHREAD *st;
1282	SVCXPRT *xprt;
1283	SVCPOOL *pool;
1284
1285	st = rqstp->rq_thread;
1286	xprt = rqstp->rq_xprt;
1287	if (xprt)
1288		pool = xprt->xp_pool;
1289	else
1290		pool = NULL;
1291	if (st) {
1292		mtx_lock(&pool->sp_lock);
1293		KASSERT(rqstp == STAILQ_FIRST(&st->st_reqs),
1294		    ("Freeing request out of order"));
1295		STAILQ_REMOVE_HEAD(&st->st_reqs, rq_link);
1296		st->st_reqcount--;
1297		if (pool->sp_done)
1298			pool->sp_done(st, rqstp);
1299		mtx_unlock(&pool->sp_lock);
1300	}
1301
1302	if (rqstp->rq_auth.svc_ah_ops)
1303		SVCAUTH_RELEASE(&rqstp->rq_auth);
1304
1305	if (rqstp->rq_xprt) {
1306		SVC_RELEASE(rqstp->rq_xprt);
1307	}
1308
1309	if (rqstp->rq_addr)
1310		free(rqstp->rq_addr, M_SONAME);
1311
1312	if (rqstp->rq_args)
1313		m_freem(rqstp->rq_args);
1314
1315	free(rqstp, M_RPC);
1316}
1317