xref: /freebsd/sys/rpc/clnt_vc.c (revision 36712a94975f5bd0d26c85377283b49a2369c82f)
1 /*	$NetBSD: clnt_vc.c,v 1.4 2000/07/14 08:40:42 fvdl Exp $	*/
2 
3 /*-
4  * SPDX-License-Identifier: BSD-3-Clause
5  *
6  * Copyright (c) 2009, Sun Microsystems, Inc.
7  * All rights reserved.
8  *
9  * Redistribution and use in source and binary forms, with or without
10  * modification, are permitted provided that the following conditions are met:
11  * - Redistributions of source code must retain the above copyright notice,
12  *   this list of conditions and the following disclaimer.
13  * - Redistributions in binary form must reproduce the above copyright notice,
14  *   this list of conditions and the following disclaimer in the documentation
15  *   and/or other materials provided with the distribution.
16  * - Neither the name of Sun Microsystems, Inc. nor the names of its
17  *   contributors may be used to endorse or promote products derived
18  *   from this software without specific prior written permission.
19  *
20  * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
21  * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
22  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
23  * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE
24  * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
25  * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
26  * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
27  * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
28  * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
29  * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
30  * POSSIBILITY OF SUCH DAMAGE.
31  */
32 
33 #if defined(LIBC_SCCS) && !defined(lint)
34 static char *sccsid2 = "@(#)clnt_tcp.c 1.37 87/10/05 Copyr 1984 Sun Micro";
35 static char *sccsid = "@(#)clnt_tcp.c	2.2 88/08/01 4.0 RPCSRC";
36 static char sccsid3[] = "@(#)clnt_vc.c 1.19 89/03/16 Copyr 1988 Sun Micro";
37 #endif
38 #include <sys/cdefs.h>
39 __FBSDID("$FreeBSD$");
40 
41 /*
42  * clnt_tcp.c, Implements a TCP/IP based, client side RPC.
43  *
44  * Copyright (C) 1984, Sun Microsystems, Inc.
45  *
46  * TCP based RPC supports 'batched calls'.
47  * A sequence of calls may be batched-up in a send buffer.  The rpc call
48  * return immediately to the client even though the call was not necessarily
49  * sent.  The batching occurs if the results' xdr routine is NULL (0) AND
50  * the rpc timeout value is zero (see clnt.h, rpc).
51  *
52  * Clients should NOT casually batch calls that in fact return results; that is,
53  * the server side should be aware that a call is batched and not produce any
54  * return message.  Batched calls that produce many result messages can
55  * deadlock (netlock) the client and the server....
56  *
57  * Now go hang yourself.
58  */
59 
60 #include <sys/param.h>
61 #include <sys/systm.h>
62 #include <sys/kernel.h>
63 #include <sys/lock.h>
64 #include <sys/malloc.h>
65 #include <sys/mbuf.h>
66 #include <sys/mutex.h>
67 #include <sys/pcpu.h>
68 #include <sys/proc.h>
69 #include <sys/protosw.h>
70 #include <sys/socket.h>
71 #include <sys/socketvar.h>
72 #include <sys/sx.h>
73 #include <sys/syslog.h>
74 #include <sys/time.h>
75 #include <sys/uio.h>
76 
77 #include <net/vnet.h>
78 
79 #include <netinet/tcp.h>
80 
81 #include <rpc/rpc.h>
82 #include <rpc/rpc_com.h>
83 #include <rpc/krpc.h>
84 
85 struct cmessage {
86         struct cmsghdr cmsg;
87         struct cmsgcred cmcred;
88 };
89 
90 static enum clnt_stat clnt_vc_call(CLIENT *, struct rpc_callextra *,
91     rpcproc_t, struct mbuf *, struct mbuf **, struct timeval);
92 static void clnt_vc_geterr(CLIENT *, struct rpc_err *);
93 static bool_t clnt_vc_freeres(CLIENT *, xdrproc_t, void *);
94 static void clnt_vc_abort(CLIENT *);
95 static bool_t clnt_vc_control(CLIENT *, u_int, void *);
96 static void clnt_vc_close(CLIENT *);
97 static void clnt_vc_destroy(CLIENT *);
98 static bool_t time_not_ok(struct timeval *);
99 static int clnt_vc_soupcall(struct socket *so, void *arg, int waitflag);
100 
101 static struct clnt_ops clnt_vc_ops = {
102 	.cl_call =	clnt_vc_call,
103 	.cl_abort =	clnt_vc_abort,
104 	.cl_geterr =	clnt_vc_geterr,
105 	.cl_freeres =	clnt_vc_freeres,
106 	.cl_close =	clnt_vc_close,
107 	.cl_destroy =	clnt_vc_destroy,
108 	.cl_control =	clnt_vc_control
109 };
110 
111 static void clnt_vc_upcallsdone(struct ct_data *);
112 
113 static int	fake_wchan;
114 
115 /*
116  * Create a client handle for a connection.
117  * Default options are set, which the user can change using clnt_control()'s.
118  * The rpc/vc package does buffering similar to stdio, so the client
119  * must pick send and receive buffer sizes, 0 => use the default.
120  * NB: fd is copied into a private area.
121  * NB: The rpch->cl_auth is set null authentication. Caller may wish to
122  * set this something more useful.
123  *
124  * fd should be an open socket
125  */
126 CLIENT *
127 clnt_vc_create(
128 	struct socket *so,		/* open file descriptor */
129 	struct sockaddr *raddr,		/* servers address */
130 	const rpcprog_t prog,		/* program number */
131 	const rpcvers_t vers,		/* version number */
132 	size_t sendsz,			/* buffer recv size */
133 	size_t recvsz,			/* buffer send size */
134 	int intrflag)			/* interruptible */
135 {
136 	CLIENT *cl;			/* client handle */
137 	struct ct_data *ct = NULL;	/* client handle */
138 	struct timeval now;
139 	struct rpc_msg call_msg;
140 	static uint32_t disrupt;
141 	struct __rpc_sockinfo si;
142 	XDR xdrs;
143 	int error, interrupted, one = 1, sleep_flag;
144 	struct sockopt sopt;
145 
146 	if (disrupt == 0)
147 		disrupt = (uint32_t)(long)raddr;
148 
149 	cl = (CLIENT *)mem_alloc(sizeof (*cl));
150 	ct = (struct ct_data *)mem_alloc(sizeof (*ct));
151 
152 	mtx_init(&ct->ct_lock, "ct->ct_lock", NULL, MTX_DEF);
153 	ct->ct_threads = 0;
154 	ct->ct_closing = FALSE;
155 	ct->ct_closed = FALSE;
156 	ct->ct_upcallrefs = 0;
157 
158 	if ((so->so_state & (SS_ISCONNECTED|SS_ISCONFIRMING)) == 0) {
159 		error = soconnect(so, raddr, curthread);
160 		SOCK_LOCK(so);
161 		interrupted = 0;
162 		sleep_flag = PSOCK;
163 		if (intrflag != 0)
164 			sleep_flag |= PCATCH;
165 		while ((so->so_state & SS_ISCONNECTING)
166 		    && so->so_error == 0) {
167 			error = msleep(&so->so_timeo, SOCK_MTX(so),
168 			    sleep_flag, "connec", 0);
169 			if (error) {
170 				if (error == EINTR || error == ERESTART)
171 					interrupted = 1;
172 				break;
173 			}
174 		}
175 		if (error == 0) {
176 			error = so->so_error;
177 			so->so_error = 0;
178 		}
179 		SOCK_UNLOCK(so);
180 		if (error) {
181 			if (!interrupted)
182 				so->so_state &= ~SS_ISCONNECTING;
183 			rpc_createerr.cf_stat = RPC_SYSTEMERROR;
184 			rpc_createerr.cf_error.re_errno = error;
185 			goto err;
186 		}
187 	}
188 
189 	if (!__rpc_socket2sockinfo(so, &si)) {
190 		goto err;
191 	}
192 
193 	if (so->so_proto->pr_flags & PR_CONNREQUIRED) {
194 		bzero(&sopt, sizeof(sopt));
195 		sopt.sopt_dir = SOPT_SET;
196 		sopt.sopt_level = SOL_SOCKET;
197 		sopt.sopt_name = SO_KEEPALIVE;
198 		sopt.sopt_val = &one;
199 		sopt.sopt_valsize = sizeof(one);
200 		sosetopt(so, &sopt);
201 	}
202 
203 	if (so->so_proto->pr_protocol == IPPROTO_TCP) {
204 		bzero(&sopt, sizeof(sopt));
205 		sopt.sopt_dir = SOPT_SET;
206 		sopt.sopt_level = IPPROTO_TCP;
207 		sopt.sopt_name = TCP_NODELAY;
208 		sopt.sopt_val = &one;
209 		sopt.sopt_valsize = sizeof(one);
210 		sosetopt(so, &sopt);
211 	}
212 
213 	ct->ct_closeit = FALSE;
214 
215 	/*
216 	 * Set up private data struct
217 	 */
218 	ct->ct_socket = so;
219 	ct->ct_wait.tv_sec = -1;
220 	ct->ct_wait.tv_usec = -1;
221 	memcpy(&ct->ct_addr, raddr, raddr->sa_len);
222 
223 	/*
224 	 * Initialize call message
225 	 */
226 	getmicrotime(&now);
227 	ct->ct_xid = ((uint32_t)++disrupt) ^ __RPC_GETXID(&now);
228 	call_msg.rm_xid = ct->ct_xid;
229 	call_msg.rm_direction = CALL;
230 	call_msg.rm_call.cb_rpcvers = RPC_MSG_VERSION;
231 	call_msg.rm_call.cb_prog = (uint32_t)prog;
232 	call_msg.rm_call.cb_vers = (uint32_t)vers;
233 
234 	/*
235 	 * pre-serialize the static part of the call msg and stash it away
236 	 */
237 	xdrmem_create(&xdrs, ct->ct_mcallc, MCALL_MSG_SIZE,
238 	    XDR_ENCODE);
239 	if (! xdr_callhdr(&xdrs, &call_msg)) {
240 		if (ct->ct_closeit) {
241 			soclose(ct->ct_socket);
242 		}
243 		goto err;
244 	}
245 	ct->ct_mpos = XDR_GETPOS(&xdrs);
246 	XDR_DESTROY(&xdrs);
247 	ct->ct_waitchan = "rpcrecv";
248 	ct->ct_waitflag = 0;
249 
250 	/*
251 	 * Create a client handle which uses xdrrec for serialization
252 	 * and authnone for authentication.
253 	 */
254 	sendsz = __rpc_get_t_size(si.si_af, si.si_proto, (int)sendsz);
255 	recvsz = __rpc_get_t_size(si.si_af, si.si_proto, (int)recvsz);
256 	error = soreserve(ct->ct_socket, sendsz, recvsz);
257 	if (error != 0) {
258 		if (ct->ct_closeit) {
259 			soclose(ct->ct_socket);
260 		}
261 		goto err;
262 	}
263 	cl->cl_refs = 1;
264 	cl->cl_ops = &clnt_vc_ops;
265 	cl->cl_private = ct;
266 	cl->cl_auth = authnone_create();
267 
268 	SOCKBUF_LOCK(&ct->ct_socket->so_rcv);
269 	soupcall_set(ct->ct_socket, SO_RCV, clnt_vc_soupcall, ct);
270 	SOCKBUF_UNLOCK(&ct->ct_socket->so_rcv);
271 
272 	ct->ct_record = NULL;
273 	ct->ct_record_resid = 0;
274 	TAILQ_INIT(&ct->ct_pending);
275 	return (cl);
276 
277 err:
278 	mtx_destroy(&ct->ct_lock);
279 	mem_free(ct, sizeof (struct ct_data));
280 	mem_free(cl, sizeof (CLIENT));
281 
282 	return ((CLIENT *)NULL);
283 }
284 
285 static enum clnt_stat
286 clnt_vc_call(
287 	CLIENT		*cl,		/* client handle */
288 	struct rpc_callextra *ext,	/* call metadata */
289 	rpcproc_t	proc,		/* procedure number */
290 	struct mbuf	*args,		/* pointer to args */
291 	struct mbuf	**resultsp,	/* pointer to results */
292 	struct timeval	utimeout)
293 {
294 	struct ct_data *ct = (struct ct_data *) cl->cl_private;
295 	AUTH *auth;
296 	struct rpc_err *errp;
297 	enum clnt_stat stat;
298 	XDR xdrs;
299 	struct rpc_msg reply_msg;
300 	bool_t ok;
301 	int nrefreshes = 2;		/* number of times to refresh cred */
302 	struct timeval timeout;
303 	uint32_t xid;
304 	struct mbuf *mreq = NULL, *results;
305 	struct ct_request *cr;
306 	int error, trycnt;
307 
308 	cr = malloc(sizeof(struct ct_request), M_RPC, M_WAITOK);
309 
310 	mtx_lock(&ct->ct_lock);
311 
312 	if (ct->ct_closing || ct->ct_closed) {
313 		mtx_unlock(&ct->ct_lock);
314 		free(cr, M_RPC);
315 		return (RPC_CANTSEND);
316 	}
317 	ct->ct_threads++;
318 
319 	if (ext) {
320 		auth = ext->rc_auth;
321 		errp = &ext->rc_err;
322 	} else {
323 		auth = cl->cl_auth;
324 		errp = &ct->ct_error;
325 	}
326 
327 	cr->cr_mrep = NULL;
328 	cr->cr_error = 0;
329 
330 	if (ct->ct_wait.tv_usec == -1) {
331 		timeout = utimeout;	/* use supplied timeout */
332 	} else {
333 		timeout = ct->ct_wait;	/* use default timeout */
334 	}
335 
336 	/*
337 	 * After 15sec of looping, allow it to return RPC_CANTSEND, which will
338 	 * cause the clnt_reconnect layer to create a new TCP connection.
339 	 */
340 	trycnt = 15 * hz;
341 call_again:
342 	mtx_assert(&ct->ct_lock, MA_OWNED);
343 	if (ct->ct_closing || ct->ct_closed) {
344 		ct->ct_threads--;
345 		wakeup(ct);
346 		mtx_unlock(&ct->ct_lock);
347 		free(cr, M_RPC);
348 		return (RPC_CANTSEND);
349 	}
350 
351 	ct->ct_xid++;
352 	xid = ct->ct_xid;
353 
354 	mtx_unlock(&ct->ct_lock);
355 
356 	/*
357 	 * Leave space to pre-pend the record mark.
358 	 */
359 	mreq = m_gethdr(M_WAITOK, MT_DATA);
360 	mreq->m_data += sizeof(uint32_t);
361 	KASSERT(ct->ct_mpos + sizeof(uint32_t) <= MHLEN,
362 	    ("RPC header too big"));
363 	bcopy(ct->ct_mcallc, mreq->m_data, ct->ct_mpos);
364 	mreq->m_len = ct->ct_mpos;
365 
366 	/*
367 	 * The XID is the first thing in the request.
368 	 */
369 	*mtod(mreq, uint32_t *) = htonl(xid);
370 
371 	xdrmbuf_create(&xdrs, mreq, XDR_ENCODE);
372 
373 	errp->re_status = stat = RPC_SUCCESS;
374 
375 	if ((! XDR_PUTINT32(&xdrs, &proc)) ||
376 	    (! AUTH_MARSHALL(auth, xid, &xdrs,
377 		m_copym(args, 0, M_COPYALL, M_WAITOK)))) {
378 		errp->re_status = stat = RPC_CANTENCODEARGS;
379 		mtx_lock(&ct->ct_lock);
380 		goto out;
381 	}
382 	mreq->m_pkthdr.len = m_length(mreq, NULL);
383 
384 	/*
385 	 * Prepend a record marker containing the packet length.
386 	 */
387 	M_PREPEND(mreq, sizeof(uint32_t), M_WAITOK);
388 	*mtod(mreq, uint32_t *) =
389 		htonl(0x80000000 | (mreq->m_pkthdr.len - sizeof(uint32_t)));
390 
391 	cr->cr_xid = xid;
392 	mtx_lock(&ct->ct_lock);
393 	/*
394 	 * Check to see if the other end has already started to close down
395 	 * the connection. The upcall will have set ct_error.re_status
396 	 * to RPC_CANTRECV if this is the case.
397 	 * If the other end starts to close down the connection after this
398 	 * point, it will be detected later when cr_error is checked,
399 	 * since the request is in the ct_pending queue.
400 	 */
401 	if (ct->ct_error.re_status == RPC_CANTRECV) {
402 		if (errp != &ct->ct_error) {
403 			errp->re_errno = ct->ct_error.re_errno;
404 			errp->re_status = RPC_CANTRECV;
405 		}
406 		stat = RPC_CANTRECV;
407 		goto out;
408 	}
409 	TAILQ_INSERT_TAIL(&ct->ct_pending, cr, cr_link);
410 	mtx_unlock(&ct->ct_lock);
411 
412 	/*
413 	 * sosend consumes mreq.
414 	 */
415 	error = sosend(ct->ct_socket, NULL, NULL, mreq, NULL, 0, curthread);
416 	mreq = NULL;
417 	if (error == EMSGSIZE || (error == ERESTART &&
418 	    (ct->ct_waitflag & PCATCH) == 0 && trycnt-- > 0)) {
419 		SOCKBUF_LOCK(&ct->ct_socket->so_snd);
420 		sbwait(&ct->ct_socket->so_snd);
421 		SOCKBUF_UNLOCK(&ct->ct_socket->so_snd);
422 		AUTH_VALIDATE(auth, xid, NULL, NULL);
423 		mtx_lock(&ct->ct_lock);
424 		TAILQ_REMOVE(&ct->ct_pending, cr, cr_link);
425 		/* Sleep for 1 clock tick before trying the sosend() again. */
426 		msleep(&fake_wchan, &ct->ct_lock, 0, "rpclpsnd", 1);
427 		goto call_again;
428 	}
429 
430 	reply_msg.acpted_rply.ar_verf.oa_flavor = AUTH_NULL;
431 	reply_msg.acpted_rply.ar_verf.oa_base = cr->cr_verf;
432 	reply_msg.acpted_rply.ar_verf.oa_length = 0;
433 	reply_msg.acpted_rply.ar_results.where = NULL;
434 	reply_msg.acpted_rply.ar_results.proc = (xdrproc_t)xdr_void;
435 
436 	mtx_lock(&ct->ct_lock);
437 	if (error) {
438 		TAILQ_REMOVE(&ct->ct_pending, cr, cr_link);
439 		errp->re_errno = error;
440 		errp->re_status = stat = RPC_CANTSEND;
441 		goto out;
442 	}
443 
444 	/*
445 	 * Check to see if we got an upcall while waiting for the
446 	 * lock. In both these cases, the request has been removed
447 	 * from ct->ct_pending.
448 	 */
449 	if (cr->cr_error) {
450 		TAILQ_REMOVE(&ct->ct_pending, cr, cr_link);
451 		errp->re_errno = cr->cr_error;
452 		errp->re_status = stat = RPC_CANTRECV;
453 		goto out;
454 	}
455 	if (cr->cr_mrep) {
456 		TAILQ_REMOVE(&ct->ct_pending, cr, cr_link);
457 		goto got_reply;
458 	}
459 
460 	/*
461 	 * Hack to provide rpc-based message passing
462 	 */
463 	if (timeout.tv_sec == 0 && timeout.tv_usec == 0) {
464 		TAILQ_REMOVE(&ct->ct_pending, cr, cr_link);
465 		errp->re_status = stat = RPC_TIMEDOUT;
466 		goto out;
467 	}
468 
469 	error = msleep(cr, &ct->ct_lock, ct->ct_waitflag, ct->ct_waitchan,
470 	    tvtohz(&timeout));
471 
472 	TAILQ_REMOVE(&ct->ct_pending, cr, cr_link);
473 
474 	if (error) {
475 		/*
476 		 * The sleep returned an error so our request is still
477 		 * on the list. Turn the error code into an
478 		 * appropriate client status.
479 		 */
480 		errp->re_errno = error;
481 		switch (error) {
482 		case EINTR:
483 			stat = RPC_INTR;
484 			break;
485 		case EWOULDBLOCK:
486 			stat = RPC_TIMEDOUT;
487 			break;
488 		default:
489 			stat = RPC_CANTRECV;
490 		}
491 		errp->re_status = stat;
492 		goto out;
493 	} else {
494 		/*
495 		 * We were woken up by the upcall.  If the
496 		 * upcall had a receive error, report that,
497 		 * otherwise we have a reply.
498 		 */
499 		if (cr->cr_error) {
500 			errp->re_errno = cr->cr_error;
501 			errp->re_status = stat = RPC_CANTRECV;
502 			goto out;
503 		}
504 	}
505 
506 got_reply:
507 	/*
508 	 * Now decode and validate the response. We need to drop the
509 	 * lock since xdr_replymsg may end up sleeping in malloc.
510 	 */
511 	mtx_unlock(&ct->ct_lock);
512 
513 	if (ext && ext->rc_feedback)
514 		ext->rc_feedback(FEEDBACK_OK, proc, ext->rc_feedback_arg);
515 
516 	xdrmbuf_create(&xdrs, cr->cr_mrep, XDR_DECODE);
517 	ok = xdr_replymsg(&xdrs, &reply_msg);
518 	cr->cr_mrep = NULL;
519 
520 	if (ok) {
521 		if ((reply_msg.rm_reply.rp_stat == MSG_ACCEPTED) &&
522 		    (reply_msg.acpted_rply.ar_stat == SUCCESS))
523 			errp->re_status = stat = RPC_SUCCESS;
524 		else
525 			stat = _seterr_reply(&reply_msg, errp);
526 
527 		if (stat == RPC_SUCCESS) {
528 			results = xdrmbuf_getall(&xdrs);
529 			if (!AUTH_VALIDATE(auth, xid,
530 				&reply_msg.acpted_rply.ar_verf,
531 				&results)) {
532 				errp->re_status = stat = RPC_AUTHERROR;
533 				errp->re_why = AUTH_INVALIDRESP;
534 			} else {
535 				KASSERT(results,
536 				    ("auth validated but no result"));
537 				*resultsp = results;
538 			}
539 		}		/* end successful completion */
540 		/*
541 		 * If unsuccessful AND error is an authentication error
542 		 * then refresh credentials and try again, else break
543 		 */
544 		else if (stat == RPC_AUTHERROR)
545 			/* maybe our credentials need to be refreshed ... */
546 			if (nrefreshes > 0 &&
547 			    AUTH_REFRESH(auth, &reply_msg)) {
548 				nrefreshes--;
549 				XDR_DESTROY(&xdrs);
550 				mtx_lock(&ct->ct_lock);
551 				goto call_again;
552 			}
553 		/* end of unsuccessful completion */
554 	}	/* end of valid reply message */
555 	else {
556 		errp->re_status = stat = RPC_CANTDECODERES;
557 	}
558 	XDR_DESTROY(&xdrs);
559 	mtx_lock(&ct->ct_lock);
560 out:
561 	mtx_assert(&ct->ct_lock, MA_OWNED);
562 
563 	KASSERT(stat != RPC_SUCCESS || *resultsp,
564 	    ("RPC_SUCCESS without reply"));
565 
566 	if (mreq)
567 		m_freem(mreq);
568 	if (cr->cr_mrep)
569 		m_freem(cr->cr_mrep);
570 
571 	ct->ct_threads--;
572 	if (ct->ct_closing)
573 		wakeup(ct);
574 
575 	mtx_unlock(&ct->ct_lock);
576 
577 	if (auth && stat != RPC_SUCCESS)
578 		AUTH_VALIDATE(auth, xid, NULL, NULL);
579 
580 	free(cr, M_RPC);
581 
582 	return (stat);
583 }
584 
585 static void
586 clnt_vc_geterr(CLIENT *cl, struct rpc_err *errp)
587 {
588 	struct ct_data *ct = (struct ct_data *) cl->cl_private;
589 
590 	*errp = ct->ct_error;
591 }
592 
593 static bool_t
594 clnt_vc_freeres(CLIENT *cl, xdrproc_t xdr_res, void *res_ptr)
595 {
596 	XDR xdrs;
597 	bool_t dummy;
598 
599 	xdrs.x_op = XDR_FREE;
600 	dummy = (*xdr_res)(&xdrs, res_ptr);
601 
602 	return (dummy);
603 }
604 
605 /*ARGSUSED*/
606 static void
607 clnt_vc_abort(CLIENT *cl)
608 {
609 }
610 
611 static bool_t
612 clnt_vc_control(CLIENT *cl, u_int request, void *info)
613 {
614 	struct ct_data *ct = (struct ct_data *)cl->cl_private;
615 	void *infop = info;
616 	SVCXPRT *xprt;
617 
618 	mtx_lock(&ct->ct_lock);
619 
620 	switch (request) {
621 	case CLSET_FD_CLOSE:
622 		ct->ct_closeit = TRUE;
623 		mtx_unlock(&ct->ct_lock);
624 		return (TRUE);
625 	case CLSET_FD_NCLOSE:
626 		ct->ct_closeit = FALSE;
627 		mtx_unlock(&ct->ct_lock);
628 		return (TRUE);
629 	default:
630 		break;
631 	}
632 
633 	/* for other requests which use info */
634 	if (info == NULL) {
635 		mtx_unlock(&ct->ct_lock);
636 		return (FALSE);
637 	}
638 	switch (request) {
639 	case CLSET_TIMEOUT:
640 		if (time_not_ok((struct timeval *)info)) {
641 			mtx_unlock(&ct->ct_lock);
642 			return (FALSE);
643 		}
644 		ct->ct_wait = *(struct timeval *)infop;
645 		break;
646 	case CLGET_TIMEOUT:
647 		*(struct timeval *)infop = ct->ct_wait;
648 		break;
649 	case CLGET_SERVER_ADDR:
650 		(void) memcpy(info, &ct->ct_addr, (size_t)ct->ct_addr.ss_len);
651 		break;
652 	case CLGET_SVC_ADDR:
653 		/*
654 		 * Slightly different semantics to userland - we use
655 		 * sockaddr instead of netbuf.
656 		 */
657 		memcpy(info, &ct->ct_addr, ct->ct_addr.ss_len);
658 		break;
659 	case CLSET_SVC_ADDR:		/* set to new address */
660 		mtx_unlock(&ct->ct_lock);
661 		return (FALSE);
662 	case CLGET_XID:
663 		*(uint32_t *)info = ct->ct_xid;
664 		break;
665 	case CLSET_XID:
666 		/* This will set the xid of the NEXT call */
667 		/* decrement by 1 as clnt_vc_call() increments once */
668 		ct->ct_xid = *(uint32_t *)info - 1;
669 		break;
670 	case CLGET_VERS:
671 		/*
672 		 * This RELIES on the information that, in the call body,
673 		 * the version number field is the fifth field from the
674 		 * beginning of the RPC header. MUST be changed if the
675 		 * call_struct is changed
676 		 */
677 		*(uint32_t *)info =
678 		    ntohl(*(uint32_t *)(void *)(ct->ct_mcallc +
679 		    4 * BYTES_PER_XDR_UNIT));
680 		break;
681 
682 	case CLSET_VERS:
683 		*(uint32_t *)(void *)(ct->ct_mcallc +
684 		    4 * BYTES_PER_XDR_UNIT) =
685 		    htonl(*(uint32_t *)info);
686 		break;
687 
688 	case CLGET_PROG:
689 		/*
690 		 * This RELIES on the information that, in the call body,
691 		 * the program number field is the fourth field from the
692 		 * beginning of the RPC header. MUST be changed if the
693 		 * call_struct is changed
694 		 */
695 		*(uint32_t *)info =
696 		    ntohl(*(uint32_t *)(void *)(ct->ct_mcallc +
697 		    3 * BYTES_PER_XDR_UNIT));
698 		break;
699 
700 	case CLSET_PROG:
701 		*(uint32_t *)(void *)(ct->ct_mcallc +
702 		    3 * BYTES_PER_XDR_UNIT) =
703 		    htonl(*(uint32_t *)info);
704 		break;
705 
706 	case CLSET_WAITCHAN:
707 		ct->ct_waitchan = (const char *)info;
708 		break;
709 
710 	case CLGET_WAITCHAN:
711 		*(const char **) info = ct->ct_waitchan;
712 		break;
713 
714 	case CLSET_INTERRUPTIBLE:
715 		if (*(int *) info)
716 			ct->ct_waitflag = PCATCH;
717 		else
718 			ct->ct_waitflag = 0;
719 		break;
720 
721 	case CLGET_INTERRUPTIBLE:
722 		if (ct->ct_waitflag)
723 			*(int *) info = TRUE;
724 		else
725 			*(int *) info = FALSE;
726 		break;
727 
728 	case CLSET_BACKCHANNEL:
729 		xprt = (SVCXPRT *)info;
730 		if (ct->ct_backchannelxprt == NULL) {
731 			xprt->xp_p2 = ct;
732 			ct->ct_backchannelxprt = xprt;
733 		}
734 		break;
735 
736 	default:
737 		mtx_unlock(&ct->ct_lock);
738 		return (FALSE);
739 	}
740 
741 	mtx_unlock(&ct->ct_lock);
742 	return (TRUE);
743 }
744 
745 static void
746 clnt_vc_close(CLIENT *cl)
747 {
748 	struct ct_data *ct = (struct ct_data *) cl->cl_private;
749 	struct ct_request *cr;
750 
751 	mtx_lock(&ct->ct_lock);
752 
753 	if (ct->ct_closed) {
754 		mtx_unlock(&ct->ct_lock);
755 		return;
756 	}
757 
758 	if (ct->ct_closing) {
759 		while (ct->ct_closing)
760 			msleep(ct, &ct->ct_lock, 0, "rpcclose", 0);
761 		KASSERT(ct->ct_closed, ("client should be closed"));
762 		mtx_unlock(&ct->ct_lock);
763 		return;
764 	}
765 
766 	if (ct->ct_socket) {
767 		ct->ct_closing = TRUE;
768 		mtx_unlock(&ct->ct_lock);
769 
770 		SOCKBUF_LOCK(&ct->ct_socket->so_rcv);
771 		soupcall_clear(ct->ct_socket, SO_RCV);
772 		clnt_vc_upcallsdone(ct);
773 		SOCKBUF_UNLOCK(&ct->ct_socket->so_rcv);
774 
775 		/*
776 		 * Abort any pending requests and wait until everyone
777 		 * has finished with clnt_vc_call.
778 		 */
779 		mtx_lock(&ct->ct_lock);
780 		TAILQ_FOREACH(cr, &ct->ct_pending, cr_link) {
781 			cr->cr_xid = 0;
782 			cr->cr_error = ESHUTDOWN;
783 			wakeup(cr);
784 		}
785 
786 		while (ct->ct_threads)
787 			msleep(ct, &ct->ct_lock, 0, "rpcclose", 0);
788 	}
789 
790 	ct->ct_closing = FALSE;
791 	ct->ct_closed = TRUE;
792 	mtx_unlock(&ct->ct_lock);
793 	wakeup(ct);
794 }
795 
796 static void
797 clnt_vc_destroy(CLIENT *cl)
798 {
799 	struct ct_data *ct = (struct ct_data *) cl->cl_private;
800 	struct socket *so = NULL;
801 	SVCXPRT *xprt;
802 
803 	clnt_vc_close(cl);
804 
805 	mtx_lock(&ct->ct_lock);
806 	xprt = ct->ct_backchannelxprt;
807 	ct->ct_backchannelxprt = NULL;
808 	if (xprt != NULL) {
809 		mtx_unlock(&ct->ct_lock);	/* To avoid a LOR. */
810 		sx_xlock(&xprt->xp_lock);
811 		mtx_lock(&ct->ct_lock);
812 		xprt->xp_p2 = NULL;
813 		sx_xunlock(&xprt->xp_lock);
814 	}
815 
816 	if (ct->ct_socket) {
817 		if (ct->ct_closeit) {
818 			so = ct->ct_socket;
819 		}
820 	}
821 
822 	mtx_unlock(&ct->ct_lock);
823 
824 	mtx_destroy(&ct->ct_lock);
825 	if (so) {
826 		soshutdown(so, SHUT_WR);
827 		soclose(so);
828 	}
829 	mem_free(ct, sizeof(struct ct_data));
830 	if (cl->cl_netid && cl->cl_netid[0])
831 		mem_free(cl->cl_netid, strlen(cl->cl_netid) +1);
832 	if (cl->cl_tp && cl->cl_tp[0])
833 		mem_free(cl->cl_tp, strlen(cl->cl_tp) +1);
834 	mem_free(cl, sizeof(CLIENT));
835 }
836 
837 /*
838  * Make sure that the time is not garbage.   -1 value is disallowed.
839  * Note this is different from time_not_ok in clnt_dg.c
840  */
841 static bool_t
842 time_not_ok(struct timeval *t)
843 {
844 	return (t->tv_sec <= -1 || t->tv_sec > 100000000 ||
845 		t->tv_usec <= -1 || t->tv_usec > 1000000);
846 }
847 
848 int
849 clnt_vc_soupcall(struct socket *so, void *arg, int waitflag)
850 {
851 	struct ct_data *ct = (struct ct_data *) arg;
852 	struct uio uio;
853 	struct mbuf *m, *m2;
854 	struct ct_request *cr;
855 	int error, rcvflag, foundreq;
856 	uint32_t xid_plus_direction[2], header;
857 	bool_t do_read;
858 	SVCXPRT *xprt;
859 	struct cf_conn *cd;
860 
861 	CTASSERT(sizeof(xid_plus_direction) == 2 * sizeof(uint32_t));
862 	ct->ct_upcallrefs++;
863 	uio.uio_td = curthread;
864 	do {
865 		/*
866 		 * If ct_record_resid is zero, we are waiting for a
867 		 * record mark.
868 		 */
869 		if (ct->ct_record_resid == 0) {
870 
871 			/*
872 			 * Make sure there is either a whole record
873 			 * mark in the buffer or there is some other
874 			 * error condition
875 			 */
876 			do_read = FALSE;
877 			if (sbavail(&so->so_rcv) >= sizeof(uint32_t)
878 			    || (so->so_rcv.sb_state & SBS_CANTRCVMORE)
879 			    || so->so_error)
880 				do_read = TRUE;
881 
882 			if (!do_read)
883 				break;
884 
885 			SOCKBUF_UNLOCK(&so->so_rcv);
886 			uio.uio_resid = sizeof(uint32_t);
887 			m = NULL;
888 			rcvflag = MSG_DONTWAIT | MSG_SOCALLBCK;
889 			error = soreceive(so, NULL, &uio, &m, NULL, &rcvflag);
890 			SOCKBUF_LOCK(&so->so_rcv);
891 
892 			if (error == EWOULDBLOCK)
893 				break;
894 
895 			/*
896 			 * If there was an error, wake up all pending
897 			 * requests.
898 			 */
899 			if (error || uio.uio_resid > 0) {
900 			wakeup_all:
901 				mtx_lock(&ct->ct_lock);
902 				if (!error) {
903 					/*
904 					 * We must have got EOF trying
905 					 * to read from the stream.
906 					 */
907 					error = ECONNRESET;
908 				}
909 				ct->ct_error.re_status = RPC_CANTRECV;
910 				ct->ct_error.re_errno = error;
911 				TAILQ_FOREACH(cr, &ct->ct_pending, cr_link) {
912 					cr->cr_error = error;
913 					wakeup(cr);
914 				}
915 				mtx_unlock(&ct->ct_lock);
916 				break;
917 			}
918 			m_copydata(m, 0, sizeof(uint32_t), (char *)&header);
919 			header = ntohl(header);
920 			ct->ct_record = NULL;
921 			ct->ct_record_resid = header & 0x7fffffff;
922 			ct->ct_record_eor = ((header & 0x80000000) != 0);
923 			m_freem(m);
924 		} else {
925 			/*
926 			 * Wait until the socket has the whole record
927 			 * buffered.
928 			 */
929 			do_read = FALSE;
930 			if (sbavail(&so->so_rcv) >= ct->ct_record_resid
931 			    || (so->so_rcv.sb_state & SBS_CANTRCVMORE)
932 			    || so->so_error)
933 				do_read = TRUE;
934 
935 			if (!do_read)
936 				break;
937 
938 			/*
939 			 * We have the record mark. Read as much as
940 			 * the socket has buffered up to the end of
941 			 * this record.
942 			 */
943 			SOCKBUF_UNLOCK(&so->so_rcv);
944 			uio.uio_resid = ct->ct_record_resid;
945 			m = NULL;
946 			rcvflag = MSG_DONTWAIT | MSG_SOCALLBCK;
947 			error = soreceive(so, NULL, &uio, &m, NULL, &rcvflag);
948 			SOCKBUF_LOCK(&so->so_rcv);
949 
950 			if (error == EWOULDBLOCK)
951 				break;
952 
953 			if (error || uio.uio_resid == ct->ct_record_resid)
954 				goto wakeup_all;
955 
956 			/*
957 			 * If we have part of the record already,
958 			 * chain this bit onto the end.
959 			 */
960 			if (ct->ct_record)
961 				m_last(ct->ct_record)->m_next = m;
962 			else
963 				ct->ct_record = m;
964 
965 			ct->ct_record_resid = uio.uio_resid;
966 
967 			/*
968 			 * If we have the entire record, see if we can
969 			 * match it to a request.
970 			 */
971 			if (ct->ct_record_resid == 0
972 			    && ct->ct_record_eor) {
973 				/*
974 				 * The XID is in the first uint32_t of
975 				 * the reply and the message direction
976 				 * is the second one.
977 				 */
978 				if (ct->ct_record->m_len <
979 				    sizeof(xid_plus_direction) &&
980 				    m_length(ct->ct_record, NULL) <
981 				    sizeof(xid_plus_direction)) {
982 					m_freem(ct->ct_record);
983 					break;
984 				}
985 				m_copydata(ct->ct_record, 0,
986 				    sizeof(xid_plus_direction),
987 				    (char *)xid_plus_direction);
988 				xid_plus_direction[0] =
989 				    ntohl(xid_plus_direction[0]);
990 				xid_plus_direction[1] =
991 				    ntohl(xid_plus_direction[1]);
992 				/* Check message direction. */
993 				if (xid_plus_direction[1] == CALL) {
994 					/* This is a backchannel request. */
995 					mtx_lock(&ct->ct_lock);
996 					xprt = ct->ct_backchannelxprt;
997 					if (xprt == NULL) {
998 						mtx_unlock(&ct->ct_lock);
999 						/* Just throw it away. */
1000 						m_freem(ct->ct_record);
1001 						ct->ct_record = NULL;
1002 					} else {
1003 						cd = (struct cf_conn *)
1004 						    xprt->xp_p1;
1005 						m2 = cd->mreq;
1006 						/*
1007 						 * The requests are chained
1008 						 * in the m_nextpkt list.
1009 						 */
1010 						while (m2 != NULL &&
1011 						    m2->m_nextpkt != NULL)
1012 							/* Find end of list. */
1013 							m2 = m2->m_nextpkt;
1014 						if (m2 != NULL)
1015 							m2->m_nextpkt =
1016 							    ct->ct_record;
1017 						else
1018 							cd->mreq =
1019 							    ct->ct_record;
1020 						ct->ct_record->m_nextpkt =
1021 						    NULL;
1022 						ct->ct_record = NULL;
1023 						xprt_active(xprt);
1024 						mtx_unlock(&ct->ct_lock);
1025 					}
1026 				} else {
1027 					mtx_lock(&ct->ct_lock);
1028 					foundreq = 0;
1029 					TAILQ_FOREACH(cr, &ct->ct_pending,
1030 					    cr_link) {
1031 						if (cr->cr_xid ==
1032 						    xid_plus_direction[0]) {
1033 							/*
1034 							 * This one
1035 							 * matches. We leave
1036 							 * the reply mbuf in
1037 							 * cr->cr_mrep. Set
1038 							 * the XID to zero so
1039 							 * that we will ignore
1040 							 * any duplicated
1041 							 * replies.
1042 							 */
1043 							cr->cr_xid = 0;
1044 							cr->cr_mrep =
1045 							    ct->ct_record;
1046 							cr->cr_error = 0;
1047 							foundreq = 1;
1048 							wakeup(cr);
1049 							break;
1050 						}
1051 					}
1052 					mtx_unlock(&ct->ct_lock);
1053 
1054 					if (!foundreq)
1055 						m_freem(ct->ct_record);
1056 					ct->ct_record = NULL;
1057 				}
1058 			}
1059 		}
1060 	} while (m);
1061 	ct->ct_upcallrefs--;
1062 	if (ct->ct_upcallrefs < 0)
1063 		panic("rpcvc upcall refcnt");
1064 	if (ct->ct_upcallrefs == 0)
1065 		wakeup(&ct->ct_upcallrefs);
1066 	return (SU_OK);
1067 }
1068 
1069 /*
1070  * Wait for all upcalls in progress to complete.
1071  */
1072 static void
1073 clnt_vc_upcallsdone(struct ct_data *ct)
1074 {
1075 
1076 	SOCKBUF_LOCK_ASSERT(&ct->ct_socket->so_rcv);
1077 
1078 	while (ct->ct_upcallrefs > 0)
1079 		(void) msleep(&ct->ct_upcallrefs,
1080 		    SOCKBUF_MTX(&ct->ct_socket->so_rcv), 0, "rpcvcup", 0);
1081 }
1082