xref: /freebsd/sys/rpc/svc_vc.c (revision f56f82e0)
1 /*	$NetBSD: svc_vc.c,v 1.7 2000/08/03 00:01:53 fvdl Exp $	*/
2 
3 /*-
4  * Copyright (c) 2009, Sun Microsystems, Inc.
5  * All rights reserved.
6  *
7  * Redistribution and use in source and binary forms, with or without
8  * modification, are permitted provided that the following conditions are met:
9  * - Redistributions of source code must retain the above copyright notice,
10  *   this list of conditions and the following disclaimer.
11  * - Redistributions in binary form must reproduce the above copyright notice,
12  *   this list of conditions and the following disclaimer in the documentation
13  *   and/or other materials provided with the distribution.
14  * - Neither the name of Sun Microsystems, Inc. nor the names of its
15  *   contributors may be used to endorse or promote products derived
16  *   from this software without specific prior written permission.
17  *
18  * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
19  * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
20  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
21  * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE
22  * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
23  * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
24  * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
25  * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
26  * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
27  * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
28  * POSSIBILITY OF SUCH DAMAGE.
29  */
30 
31 #if defined(LIBC_SCCS) && !defined(lint)
32 static char *sccsid2 = "@(#)svc_tcp.c 1.21 87/08/11 Copyr 1984 Sun Micro";
33 static char *sccsid = "@(#)svc_tcp.c	2.2 88/08/01 4.0 RPCSRC";
34 #endif
35 #include <sys/cdefs.h>
36 __FBSDID("$FreeBSD$");
37 
38 /*
39  * svc_vc.c, Server side for Connection Oriented based RPC.
40  *
41  * Actually implements two flavors of transporter -
42  * a tcp rendezvouser (a listner and connection establisher)
43  * and a record/tcp stream.
44  */
45 
46 #include <sys/param.h>
47 #include <sys/lock.h>
48 #include <sys/kernel.h>
49 #include <sys/malloc.h>
50 #include <sys/mbuf.h>
51 #include <sys/mutex.h>
52 #include <sys/proc.h>
53 #include <sys/protosw.h>
54 #include <sys/queue.h>
55 #include <sys/socket.h>
56 #include <sys/socketvar.h>
57 #include <sys/sx.h>
58 #include <sys/systm.h>
59 #include <sys/uio.h>
60 
61 #include <net/vnet.h>
62 
63 #include <netinet/tcp.h>
64 
65 #include <rpc/rpc.h>
66 
67 #include <rpc/krpc.h>
68 #include <rpc/rpc_com.h>
69 
70 #include <security/mac/mac_framework.h>
71 
72 static bool_t svc_vc_rendezvous_recv(SVCXPRT *, struct rpc_msg *,
73     struct sockaddr **, struct mbuf **);
74 static enum xprt_stat svc_vc_rendezvous_stat(SVCXPRT *);
75 static void svc_vc_rendezvous_destroy(SVCXPRT *);
76 static bool_t svc_vc_null(void);
77 static void svc_vc_destroy(SVCXPRT *);
78 static enum xprt_stat svc_vc_stat(SVCXPRT *);
79 static bool_t svc_vc_ack(SVCXPRT *, uint32_t *);
80 static bool_t svc_vc_recv(SVCXPRT *, struct rpc_msg *,
81     struct sockaddr **, struct mbuf **);
82 static bool_t svc_vc_reply(SVCXPRT *, struct rpc_msg *,
83     struct sockaddr *, struct mbuf *, uint32_t *seq);
84 static bool_t svc_vc_control(SVCXPRT *xprt, const u_int rq, void *in);
85 static bool_t svc_vc_rendezvous_control (SVCXPRT *xprt, const u_int rq,
86     void *in);
87 static void svc_vc_backchannel_destroy(SVCXPRT *);
88 static enum xprt_stat svc_vc_backchannel_stat(SVCXPRT *);
89 static bool_t svc_vc_backchannel_recv(SVCXPRT *, struct rpc_msg *,
90     struct sockaddr **, struct mbuf **);
91 static bool_t svc_vc_backchannel_reply(SVCXPRT *, struct rpc_msg *,
92     struct sockaddr *, struct mbuf *, uint32_t *);
93 static bool_t svc_vc_backchannel_control(SVCXPRT *xprt, const u_int rq,
94     void *in);
95 static SVCXPRT *svc_vc_create_conn(SVCPOOL *pool, struct socket *so,
96     struct sockaddr *raddr);
97 static int svc_vc_accept(struct socket *head, struct socket **sop);
98 static int svc_vc_soupcall(struct socket *so, void *arg, int waitflag);
99 static int svc_vc_rendezvous_soupcall(struct socket *, void *, int);
100 
101 static struct xp_ops svc_vc_rendezvous_ops = {
102 	.xp_recv =	svc_vc_rendezvous_recv,
103 	.xp_stat =	svc_vc_rendezvous_stat,
104 	.xp_reply =	(bool_t (*)(SVCXPRT *, struct rpc_msg *,
105 		struct sockaddr *, struct mbuf *, uint32_t *))svc_vc_null,
106 	.xp_destroy =	svc_vc_rendezvous_destroy,
107 	.xp_control =	svc_vc_rendezvous_control
108 };
109 
110 static struct xp_ops svc_vc_ops = {
111 	.xp_recv =	svc_vc_recv,
112 	.xp_stat =	svc_vc_stat,
113 	.xp_ack =	svc_vc_ack,
114 	.xp_reply =	svc_vc_reply,
115 	.xp_destroy =	svc_vc_destroy,
116 	.xp_control =	svc_vc_control
117 };
118 
119 static struct xp_ops svc_vc_backchannel_ops = {
120 	.xp_recv =	svc_vc_backchannel_recv,
121 	.xp_stat =	svc_vc_backchannel_stat,
122 	.xp_reply =	svc_vc_backchannel_reply,
123 	.xp_destroy =	svc_vc_backchannel_destroy,
124 	.xp_control =	svc_vc_backchannel_control
125 };
126 
127 /*
128  * Usage:
129  *	xprt = svc_vc_create(sock, send_buf_size, recv_buf_size);
130  *
131  * Creates, registers, and returns a (rpc) tcp based transporter.
132  * Once *xprt is initialized, it is registered as a transporter
133  * see (svc.h, xprt_register).  This routine returns
134  * a NULL if a problem occurred.
135  *
136  * The filedescriptor passed in is expected to refer to a bound, but
137  * not yet connected socket.
138  *
139  * Since streams do buffered io similar to stdio, the caller can specify
140  * how big the send and receive buffers are via the second and third parms;
141  * 0 => use the system default.
142  */
143 SVCXPRT *
144 svc_vc_create(SVCPOOL *pool, struct socket *so, size_t sendsize,
145     size_t recvsize)
146 {
147 	SVCXPRT *xprt;
148 	struct sockaddr* sa;
149 	int error;
150 
151 	SOCK_LOCK(so);
152 	if (so->so_state & (SS_ISCONNECTED|SS_ISDISCONNECTED)) {
153 		SOCK_UNLOCK(so);
154 		CURVNET_SET(so->so_vnet);
155 		error = so->so_proto->pr_usrreqs->pru_peeraddr(so, &sa);
156 		CURVNET_RESTORE();
157 		if (error)
158 			return (NULL);
159 		xprt = svc_vc_create_conn(pool, so, sa);
160 		free(sa, M_SONAME);
161 		return (xprt);
162 	}
163 	SOCK_UNLOCK(so);
164 
165 	xprt = svc_xprt_alloc();
166 	sx_init(&xprt->xp_lock, "xprt->xp_lock");
167 	xprt->xp_pool = pool;
168 	xprt->xp_socket = so;
169 	xprt->xp_p1 = NULL;
170 	xprt->xp_p2 = NULL;
171 	xprt->xp_ops = &svc_vc_rendezvous_ops;
172 
173 	CURVNET_SET(so->so_vnet);
174 	error = so->so_proto->pr_usrreqs->pru_sockaddr(so, &sa);
175 	CURVNET_RESTORE();
176 	if (error) {
177 		goto cleanup_svc_vc_create;
178 	}
179 
180 	memcpy(&xprt->xp_ltaddr, sa, sa->sa_len);
181 	free(sa, M_SONAME);
182 
183 	xprt_register(xprt);
184 
185 	solisten(so, -1, curthread);
186 
187 	SOLISTEN_LOCK(so);
188 	xprt->xp_upcallset = 1;
189 	solisten_upcall_set(so, svc_vc_rendezvous_soupcall, xprt);
190 	SOLISTEN_UNLOCK(so);
191 
192 	return (xprt);
193 
194 cleanup_svc_vc_create:
195 	sx_destroy(&xprt->xp_lock);
196 	svc_xprt_free(xprt);
197 
198 	return (NULL);
199 }
200 
201 /*
202  * Create a new transport for a socket optained via soaccept().
203  */
204 SVCXPRT *
205 svc_vc_create_conn(SVCPOOL *pool, struct socket *so, struct sockaddr *raddr)
206 {
207 	SVCXPRT *xprt;
208 	struct cf_conn *cd;
209 	struct sockaddr* sa = NULL;
210 	struct sockopt opt;
211 	int one = 1;
212 	int error;
213 
214 	bzero(&opt, sizeof(struct sockopt));
215 	opt.sopt_dir = SOPT_SET;
216 	opt.sopt_level = SOL_SOCKET;
217 	opt.sopt_name = SO_KEEPALIVE;
218 	opt.sopt_val = &one;
219 	opt.sopt_valsize = sizeof(one);
220 	error = sosetopt(so, &opt);
221 	if (error) {
222 		return (NULL);
223 	}
224 
225 	if (so->so_proto->pr_protocol == IPPROTO_TCP) {
226 		bzero(&opt, sizeof(struct sockopt));
227 		opt.sopt_dir = SOPT_SET;
228 		opt.sopt_level = IPPROTO_TCP;
229 		opt.sopt_name = TCP_NODELAY;
230 		opt.sopt_val = &one;
231 		opt.sopt_valsize = sizeof(one);
232 		error = sosetopt(so, &opt);
233 		if (error) {
234 			return (NULL);
235 		}
236 	}
237 
238 	cd = mem_alloc(sizeof(*cd));
239 	cd->strm_stat = XPRT_IDLE;
240 
241 	xprt = svc_xprt_alloc();
242 	sx_init(&xprt->xp_lock, "xprt->xp_lock");
243 	xprt->xp_pool = pool;
244 	xprt->xp_socket = so;
245 	xprt->xp_p1 = cd;
246 	xprt->xp_p2 = NULL;
247 	xprt->xp_ops = &svc_vc_ops;
248 
249 	/*
250 	 * See http://www.connectathon.org/talks96/nfstcp.pdf - client
251 	 * has a 5 minute timer, server has a 6 minute timer.
252 	 */
253 	xprt->xp_idletimeout = 6 * 60;
254 
255 	memcpy(&xprt->xp_rtaddr, raddr, raddr->sa_len);
256 
257 	CURVNET_SET(so->so_vnet);
258 	error = so->so_proto->pr_usrreqs->pru_sockaddr(so, &sa);
259 	CURVNET_RESTORE();
260 	if (error)
261 		goto cleanup_svc_vc_create;
262 
263 	memcpy(&xprt->xp_ltaddr, sa, sa->sa_len);
264 	free(sa, M_SONAME);
265 
266 	xprt_register(xprt);
267 
268 	SOCKBUF_LOCK(&so->so_rcv);
269 	xprt->xp_upcallset = 1;
270 	soupcall_set(so, SO_RCV, svc_vc_soupcall, xprt);
271 	SOCKBUF_UNLOCK(&so->so_rcv);
272 
273 	/*
274 	 * Throw the transport into the active list in case it already
275 	 * has some data buffered.
276 	 */
277 	sx_xlock(&xprt->xp_lock);
278 	xprt_active(xprt);
279 	sx_xunlock(&xprt->xp_lock);
280 
281 	return (xprt);
282 cleanup_svc_vc_create:
283 	sx_destroy(&xprt->xp_lock);
284 	svc_xprt_free(xprt);
285 	mem_free(cd, sizeof(*cd));
286 
287 	return (NULL);
288 }
289 
290 /*
291  * Create a new transport for a backchannel on a clnt_vc socket.
292  */
293 SVCXPRT *
294 svc_vc_create_backchannel(SVCPOOL *pool)
295 {
296 	SVCXPRT *xprt = NULL;
297 	struct cf_conn *cd = NULL;
298 
299 	cd = mem_alloc(sizeof(*cd));
300 	cd->strm_stat = XPRT_IDLE;
301 
302 	xprt = svc_xprt_alloc();
303 	sx_init(&xprt->xp_lock, "xprt->xp_lock");
304 	xprt->xp_pool = pool;
305 	xprt->xp_socket = NULL;
306 	xprt->xp_p1 = cd;
307 	xprt->xp_p2 = NULL;
308 	xprt->xp_ops = &svc_vc_backchannel_ops;
309 	return (xprt);
310 }
311 
312 /*
313  * This does all of the accept except the final call to soaccept. The
314  * caller will call soaccept after dropping its locks (soaccept may
315  * call malloc).
316  */
317 int
318 svc_vc_accept(struct socket *head, struct socket **sop)
319 {
320 	struct socket *so;
321 	int error = 0;
322 	short nbio;
323 
324 	/* XXXGL: shouldn't that be an assertion? */
325 	if ((head->so_options & SO_ACCEPTCONN) == 0) {
326 		error = EINVAL;
327 		goto done;
328 	}
329 #ifdef MAC
330 	error = mac_socket_check_accept(curthread->td_ucred, head);
331 	if (error != 0)
332 		goto done;
333 #endif
334 	/*
335 	 * XXXGL: we want non-blocking semantics.  The socket could be a
336 	 * socket created by kernel as well as socket shared with userland,
337 	 * so we can't be sure about presense of SS_NBIO.  We also shall not
338 	 * toggle it on the socket, since that may surprise userland.  So we
339 	 * set SS_NBIO only temporarily.
340 	 */
341 	SOLISTEN_LOCK(head);
342 	nbio = head->so_state & SS_NBIO;
343 	head->so_state |= SS_NBIO;
344 	error = solisten_dequeue(head, &so, 0);
345 	head->so_state &= (nbio & ~SS_NBIO);
346 	if (error)
347 		goto done;
348 
349 	so->so_state |= nbio;
350 	*sop = so;
351 
352 	/* connection has been removed from the listen queue */
353 	KNOTE_UNLOCKED(&head->so_rdsel.si_note, 0);
354 done:
355 	return (error);
356 }
357 
358 /*ARGSUSED*/
359 static bool_t
360 svc_vc_rendezvous_recv(SVCXPRT *xprt, struct rpc_msg *msg,
361     struct sockaddr **addrp, struct mbuf **mp)
362 {
363 	struct socket *so = NULL;
364 	struct sockaddr *sa = NULL;
365 	int error;
366 	SVCXPRT *new_xprt;
367 
368 	/*
369 	 * The socket upcall calls xprt_active() which will eventually
370 	 * cause the server to call us here. We attempt to accept a
371 	 * connection from the socket and turn it into a new
372 	 * transport. If the accept fails, we have drained all pending
373 	 * connections so we call xprt_inactive().
374 	 */
375 	sx_xlock(&xprt->xp_lock);
376 
377 	error = svc_vc_accept(xprt->xp_socket, &so);
378 
379 	if (error == EWOULDBLOCK) {
380 		/*
381 		 * We must re-test for new connections after taking
382 		 * the lock to protect us in the case where a new
383 		 * connection arrives after our call to accept fails
384 		 * with EWOULDBLOCK.
385 		 */
386 		SOLISTEN_LOCK(xprt->xp_socket);
387 		if (TAILQ_EMPTY(&xprt->xp_socket->sol_comp))
388 			xprt_inactive_self(xprt);
389 		SOLISTEN_UNLOCK(xprt->xp_socket);
390 		sx_xunlock(&xprt->xp_lock);
391 		return (FALSE);
392 	}
393 
394 	if (error) {
395 		SOLISTEN_LOCK(xprt->xp_socket);
396 		if (xprt->xp_upcallset) {
397 			xprt->xp_upcallset = 0;
398 			soupcall_clear(xprt->xp_socket, SO_RCV);
399 		}
400 		SOLISTEN_UNLOCK(xprt->xp_socket);
401 		xprt_inactive_self(xprt);
402 		sx_xunlock(&xprt->xp_lock);
403 		return (FALSE);
404 	}
405 
406 	sx_xunlock(&xprt->xp_lock);
407 
408 	sa = NULL;
409 	error = soaccept(so, &sa);
410 
411 	if (error) {
412 		/*
413 		 * XXX not sure if I need to call sofree or soclose here.
414 		 */
415 		if (sa)
416 			free(sa, M_SONAME);
417 		return (FALSE);
418 	}
419 
420 	/*
421 	 * svc_vc_create_conn will call xprt_register - we don't need
422 	 * to do anything with the new connection except derefence it.
423 	 */
424 	new_xprt = svc_vc_create_conn(xprt->xp_pool, so, sa);
425 	if (!new_xprt) {
426 		soclose(so);
427 	} else {
428 		SVC_RELEASE(new_xprt);
429 	}
430 
431 	free(sa, M_SONAME);
432 
433 	return (FALSE); /* there is never an rpc msg to be processed */
434 }
435 
436 /*ARGSUSED*/
437 static enum xprt_stat
438 svc_vc_rendezvous_stat(SVCXPRT *xprt)
439 {
440 
441 	return (XPRT_IDLE);
442 }
443 
444 static void
445 svc_vc_destroy_common(SVCXPRT *xprt)
446 {
447 
448 	if (xprt->xp_socket)
449 		(void)soclose(xprt->xp_socket);
450 
451 	if (xprt->xp_netid)
452 		(void) mem_free(xprt->xp_netid, strlen(xprt->xp_netid) + 1);
453 	svc_xprt_free(xprt);
454 }
455 
456 static void
457 svc_vc_rendezvous_destroy(SVCXPRT *xprt)
458 {
459 
460 	SOLISTEN_LOCK(xprt->xp_socket);
461 	if (xprt->xp_upcallset) {
462 		xprt->xp_upcallset = 0;
463 		solisten_upcall_set(xprt->xp_socket, NULL, NULL);
464 	}
465 	SOLISTEN_UNLOCK(xprt->xp_socket);
466 
467 	svc_vc_destroy_common(xprt);
468 }
469 
470 static void
471 svc_vc_destroy(SVCXPRT *xprt)
472 {
473 	struct cf_conn *cd = (struct cf_conn *)xprt->xp_p1;
474 
475 	SOCKBUF_LOCK(&xprt->xp_socket->so_rcv);
476 	if (xprt->xp_upcallset) {
477 		xprt->xp_upcallset = 0;
478 		soupcall_clear(xprt->xp_socket, SO_RCV);
479 	}
480 	SOCKBUF_UNLOCK(&xprt->xp_socket->so_rcv);
481 
482 	svc_vc_destroy_common(xprt);
483 
484 	if (cd->mreq)
485 		m_freem(cd->mreq);
486 	if (cd->mpending)
487 		m_freem(cd->mpending);
488 	mem_free(cd, sizeof(*cd));
489 }
490 
491 static void
492 svc_vc_backchannel_destroy(SVCXPRT *xprt)
493 {
494 	struct cf_conn *cd = (struct cf_conn *)xprt->xp_p1;
495 	struct mbuf *m, *m2;
496 
497 	svc_xprt_free(xprt);
498 	m = cd->mreq;
499 	while (m != NULL) {
500 		m2 = m;
501 		m = m->m_nextpkt;
502 		m_freem(m2);
503 	}
504 	mem_free(cd, sizeof(*cd));
505 }
506 
507 /*ARGSUSED*/
508 static bool_t
509 svc_vc_control(SVCXPRT *xprt, const u_int rq, void *in)
510 {
511 	return (FALSE);
512 }
513 
514 static bool_t
515 svc_vc_rendezvous_control(SVCXPRT *xprt, const u_int rq, void *in)
516 {
517 
518 	return (FALSE);
519 }
520 
521 static bool_t
522 svc_vc_backchannel_control(SVCXPRT *xprt, const u_int rq, void *in)
523 {
524 
525 	return (FALSE);
526 }
527 
528 static enum xprt_stat
529 svc_vc_stat(SVCXPRT *xprt)
530 {
531 	struct cf_conn *cd;
532 
533 	cd = (struct cf_conn *)(xprt->xp_p1);
534 
535 	if (cd->strm_stat == XPRT_DIED)
536 		return (XPRT_DIED);
537 
538 	if (cd->mreq != NULL && cd->resid == 0 && cd->eor)
539 		return (XPRT_MOREREQS);
540 
541 	if (soreadable(xprt->xp_socket))
542 		return (XPRT_MOREREQS);
543 
544 	return (XPRT_IDLE);
545 }
546 
547 static bool_t
548 svc_vc_ack(SVCXPRT *xprt, uint32_t *ack)
549 {
550 
551 	*ack = atomic_load_acq_32(&xprt->xp_snt_cnt);
552 	*ack -= sbused(&xprt->xp_socket->so_snd);
553 	return (TRUE);
554 }
555 
556 static enum xprt_stat
557 svc_vc_backchannel_stat(SVCXPRT *xprt)
558 {
559 	struct cf_conn *cd;
560 
561 	cd = (struct cf_conn *)(xprt->xp_p1);
562 
563 	if (cd->mreq != NULL)
564 		return (XPRT_MOREREQS);
565 
566 	return (XPRT_IDLE);
567 }
568 
569 /*
570  * If we have an mbuf chain in cd->mpending, try to parse a record from it,
571  * leaving the result in cd->mreq. If we don't have a complete record, leave
572  * the partial result in cd->mreq and try to read more from the socket.
573  */
574 static int
575 svc_vc_process_pending(SVCXPRT *xprt)
576 {
577 	struct cf_conn *cd = (struct cf_conn *) xprt->xp_p1;
578 	struct socket *so = xprt->xp_socket;
579 	struct mbuf *m;
580 
581 	/*
582 	 * If cd->resid is non-zero, we have part of the
583 	 * record already, otherwise we are expecting a record
584 	 * marker.
585 	 */
586 	if (!cd->resid && cd->mpending) {
587 		/*
588 		 * See if there is enough data buffered to
589 		 * make up a record marker. Make sure we can
590 		 * handle the case where the record marker is
591 		 * split across more than one mbuf.
592 		 */
593 		size_t n = 0;
594 		uint32_t header;
595 
596 		m = cd->mpending;
597 		while (n < sizeof(uint32_t) && m) {
598 			n += m->m_len;
599 			m = m->m_next;
600 		}
601 		if (n < sizeof(uint32_t)) {
602 			so->so_rcv.sb_lowat = sizeof(uint32_t) - n;
603 			return (FALSE);
604 		}
605 		m_copydata(cd->mpending, 0, sizeof(header),
606 		    (char *)&header);
607 		header = ntohl(header);
608 		cd->eor = (header & 0x80000000) != 0;
609 		cd->resid = header & 0x7fffffff;
610 		m_adj(cd->mpending, sizeof(uint32_t));
611 	}
612 
613 	/*
614 	 * Start pulling off mbufs from cd->mpending
615 	 * until we either have a complete record or
616 	 * we run out of data. We use m_split to pull
617 	 * data - it will pull as much as possible and
618 	 * split the last mbuf if necessary.
619 	 */
620 	while (cd->mpending && cd->resid) {
621 		m = cd->mpending;
622 		if (cd->mpending->m_next
623 		    || cd->mpending->m_len > cd->resid)
624 			cd->mpending = m_split(cd->mpending,
625 			    cd->resid, M_WAITOK);
626 		else
627 			cd->mpending = NULL;
628 		if (cd->mreq)
629 			m_last(cd->mreq)->m_next = m;
630 		else
631 			cd->mreq = m;
632 		while (m) {
633 			cd->resid -= m->m_len;
634 			m = m->m_next;
635 		}
636 	}
637 
638 	/*
639 	 * Block receive upcalls if we have more data pending,
640 	 * otherwise report our need.
641 	 */
642 	if (cd->mpending)
643 		so->so_rcv.sb_lowat = INT_MAX;
644 	else
645 		so->so_rcv.sb_lowat =
646 		    imax(1, imin(cd->resid, so->so_rcv.sb_hiwat / 2));
647 	return (TRUE);
648 }
649 
650 static bool_t
651 svc_vc_recv(SVCXPRT *xprt, struct rpc_msg *msg,
652     struct sockaddr **addrp, struct mbuf **mp)
653 {
654 	struct cf_conn *cd = (struct cf_conn *) xprt->xp_p1;
655 	struct uio uio;
656 	struct mbuf *m;
657 	struct socket* so = xprt->xp_socket;
658 	XDR xdrs;
659 	int error, rcvflag;
660 	uint32_t xid_plus_direction[2];
661 
662 	/*
663 	 * Serialise access to the socket and our own record parsing
664 	 * state.
665 	 */
666 	sx_xlock(&xprt->xp_lock);
667 
668 	for (;;) {
669 		/* If we have no request ready, check pending queue. */
670 		while (cd->mpending &&
671 		    (cd->mreq == NULL || cd->resid != 0 || !cd->eor)) {
672 			if (!svc_vc_process_pending(xprt))
673 				break;
674 		}
675 
676 		/* Process and return complete request in cd->mreq. */
677 		if (cd->mreq != NULL && cd->resid == 0 && cd->eor) {
678 
679 			/*
680 			 * Now, check for a backchannel reply.
681 			 * The XID is in the first uint32_t of the reply
682 			 * and the message direction is the second one.
683 			 */
684 			if ((cd->mreq->m_len >= sizeof(xid_plus_direction) ||
685 			    m_length(cd->mreq, NULL) >=
686 			    sizeof(xid_plus_direction)) &&
687 			    xprt->xp_p2 != NULL) {
688 				m_copydata(cd->mreq, 0,
689 				    sizeof(xid_plus_direction),
690 				    (char *)xid_plus_direction);
691 				xid_plus_direction[0] =
692 				    ntohl(xid_plus_direction[0]);
693 				xid_plus_direction[1] =
694 				    ntohl(xid_plus_direction[1]);
695 				/* Check message direction. */
696 				if (xid_plus_direction[1] == REPLY) {
697 					clnt_bck_svccall(xprt->xp_p2,
698 					    cd->mreq,
699 					    xid_plus_direction[0]);
700 					cd->mreq = NULL;
701 					continue;
702 				}
703 			}
704 
705 			xdrmbuf_create(&xdrs, cd->mreq, XDR_DECODE);
706 			cd->mreq = NULL;
707 
708 			/* Check for next request in a pending queue. */
709 			svc_vc_process_pending(xprt);
710 			if (cd->mreq == NULL || cd->resid != 0) {
711 				SOCKBUF_LOCK(&so->so_rcv);
712 				if (!soreadable(so))
713 					xprt_inactive_self(xprt);
714 				SOCKBUF_UNLOCK(&so->so_rcv);
715 			}
716 
717 			sx_xunlock(&xprt->xp_lock);
718 
719 			if (! xdr_callmsg(&xdrs, msg)) {
720 				XDR_DESTROY(&xdrs);
721 				return (FALSE);
722 			}
723 
724 			*addrp = NULL;
725 			*mp = xdrmbuf_getall(&xdrs);
726 			XDR_DESTROY(&xdrs);
727 
728 			return (TRUE);
729 		}
730 
731 		/*
732 		 * The socket upcall calls xprt_active() which will eventually
733 		 * cause the server to call us here. We attempt to
734 		 * read as much as possible from the socket and put
735 		 * the result in cd->mpending. If the read fails,
736 		 * we have drained both cd->mpending and the socket so
737 		 * we can call xprt_inactive().
738 		 */
739 		uio.uio_resid = 1000000000;
740 		uio.uio_td = curthread;
741 		m = NULL;
742 		rcvflag = MSG_DONTWAIT;
743 		error = soreceive(so, NULL, &uio, &m, NULL, &rcvflag);
744 
745 		if (error == EWOULDBLOCK) {
746 			/*
747 			 * We must re-test for readability after
748 			 * taking the lock to protect us in the case
749 			 * where a new packet arrives on the socket
750 			 * after our call to soreceive fails with
751 			 * EWOULDBLOCK.
752 			 */
753 			SOCKBUF_LOCK(&so->so_rcv);
754 			if (!soreadable(so))
755 				xprt_inactive_self(xprt);
756 			SOCKBUF_UNLOCK(&so->so_rcv);
757 			sx_xunlock(&xprt->xp_lock);
758 			return (FALSE);
759 		}
760 
761 		if (error) {
762 			SOCKBUF_LOCK(&so->so_rcv);
763 			if (xprt->xp_upcallset) {
764 				xprt->xp_upcallset = 0;
765 				soupcall_clear(so, SO_RCV);
766 			}
767 			SOCKBUF_UNLOCK(&so->so_rcv);
768 			xprt_inactive_self(xprt);
769 			cd->strm_stat = XPRT_DIED;
770 			sx_xunlock(&xprt->xp_lock);
771 			return (FALSE);
772 		}
773 
774 		if (!m) {
775 			/*
776 			 * EOF - the other end has closed the socket.
777 			 */
778 			xprt_inactive_self(xprt);
779 			cd->strm_stat = XPRT_DIED;
780 			sx_xunlock(&xprt->xp_lock);
781 			return (FALSE);
782 		}
783 
784 		if (cd->mpending)
785 			m_last(cd->mpending)->m_next = m;
786 		else
787 			cd->mpending = m;
788 	}
789 }
790 
791 static bool_t
792 svc_vc_backchannel_recv(SVCXPRT *xprt, struct rpc_msg *msg,
793     struct sockaddr **addrp, struct mbuf **mp)
794 {
795 	struct cf_conn *cd = (struct cf_conn *) xprt->xp_p1;
796 	struct ct_data *ct;
797 	struct mbuf *m;
798 	XDR xdrs;
799 
800 	sx_xlock(&xprt->xp_lock);
801 	ct = (struct ct_data *)xprt->xp_p2;
802 	if (ct == NULL) {
803 		sx_xunlock(&xprt->xp_lock);
804 		return (FALSE);
805 	}
806 	mtx_lock(&ct->ct_lock);
807 	m = cd->mreq;
808 	if (m == NULL) {
809 		xprt_inactive_self(xprt);
810 		mtx_unlock(&ct->ct_lock);
811 		sx_xunlock(&xprt->xp_lock);
812 		return (FALSE);
813 	}
814 	cd->mreq = m->m_nextpkt;
815 	mtx_unlock(&ct->ct_lock);
816 	sx_xunlock(&xprt->xp_lock);
817 
818 	xdrmbuf_create(&xdrs, m, XDR_DECODE);
819 	if (! xdr_callmsg(&xdrs, msg)) {
820 		XDR_DESTROY(&xdrs);
821 		return (FALSE);
822 	}
823 	*addrp = NULL;
824 	*mp = xdrmbuf_getall(&xdrs);
825 	XDR_DESTROY(&xdrs);
826 	return (TRUE);
827 }
828 
829 static bool_t
830 svc_vc_reply(SVCXPRT *xprt, struct rpc_msg *msg,
831     struct sockaddr *addr, struct mbuf *m, uint32_t *seq)
832 {
833 	XDR xdrs;
834 	struct mbuf *mrep;
835 	bool_t stat = TRUE;
836 	int error, len;
837 
838 	/*
839 	 * Leave space for record mark.
840 	 */
841 	mrep = m_gethdr(M_WAITOK, MT_DATA);
842 	mrep->m_data += sizeof(uint32_t);
843 
844 	xdrmbuf_create(&xdrs, mrep, XDR_ENCODE);
845 
846 	if (msg->rm_reply.rp_stat == MSG_ACCEPTED &&
847 	    msg->rm_reply.rp_acpt.ar_stat == SUCCESS) {
848 		if (!xdr_replymsg(&xdrs, msg))
849 			stat = FALSE;
850 		else
851 			xdrmbuf_append(&xdrs, m);
852 	} else {
853 		stat = xdr_replymsg(&xdrs, msg);
854 	}
855 
856 	if (stat) {
857 		m_fixhdr(mrep);
858 
859 		/*
860 		 * Prepend a record marker containing the reply length.
861 		 */
862 		M_PREPEND(mrep, sizeof(uint32_t), M_WAITOK);
863 		len = mrep->m_pkthdr.len;
864 		*mtod(mrep, uint32_t *) =
865 			htonl(0x80000000 | (len - sizeof(uint32_t)));
866 		atomic_add_32(&xprt->xp_snd_cnt, len);
867 		error = sosend(xprt->xp_socket, NULL, NULL, mrep, NULL,
868 		    0, curthread);
869 		if (!error) {
870 			atomic_add_rel_32(&xprt->xp_snt_cnt, len);
871 			if (seq)
872 				*seq = xprt->xp_snd_cnt;
873 			stat = TRUE;
874 		} else
875 			atomic_subtract_32(&xprt->xp_snd_cnt, len);
876 	} else {
877 		m_freem(mrep);
878 	}
879 
880 	XDR_DESTROY(&xdrs);
881 
882 	return (stat);
883 }
884 
885 static bool_t
886 svc_vc_backchannel_reply(SVCXPRT *xprt, struct rpc_msg *msg,
887     struct sockaddr *addr, struct mbuf *m, uint32_t *seq)
888 {
889 	struct ct_data *ct;
890 	XDR xdrs;
891 	struct mbuf *mrep;
892 	bool_t stat = TRUE;
893 	int error;
894 
895 	/*
896 	 * Leave space for record mark.
897 	 */
898 	mrep = m_gethdr(M_WAITOK, MT_DATA);
899 	mrep->m_data += sizeof(uint32_t);
900 
901 	xdrmbuf_create(&xdrs, mrep, XDR_ENCODE);
902 
903 	if (msg->rm_reply.rp_stat == MSG_ACCEPTED &&
904 	    msg->rm_reply.rp_acpt.ar_stat == SUCCESS) {
905 		if (!xdr_replymsg(&xdrs, msg))
906 			stat = FALSE;
907 		else
908 			xdrmbuf_append(&xdrs, m);
909 	} else {
910 		stat = xdr_replymsg(&xdrs, msg);
911 	}
912 
913 	if (stat) {
914 		m_fixhdr(mrep);
915 
916 		/*
917 		 * Prepend a record marker containing the reply length.
918 		 */
919 		M_PREPEND(mrep, sizeof(uint32_t), M_WAITOK);
920 		*mtod(mrep, uint32_t *) =
921 			htonl(0x80000000 | (mrep->m_pkthdr.len
922 				- sizeof(uint32_t)));
923 		sx_xlock(&xprt->xp_lock);
924 		ct = (struct ct_data *)xprt->xp_p2;
925 		if (ct != NULL)
926 			error = sosend(ct->ct_socket, NULL, NULL, mrep, NULL,
927 			    0, curthread);
928 		else
929 			error = EPIPE;
930 		sx_xunlock(&xprt->xp_lock);
931 		if (!error) {
932 			stat = TRUE;
933 		}
934 	} else {
935 		m_freem(mrep);
936 	}
937 
938 	XDR_DESTROY(&xdrs);
939 
940 	return (stat);
941 }
942 
943 static bool_t
944 svc_vc_null()
945 {
946 
947 	return (FALSE);
948 }
949 
950 static int
951 svc_vc_soupcall(struct socket *so, void *arg, int waitflag)
952 {
953 	SVCXPRT *xprt = (SVCXPRT *) arg;
954 
955 	if (soreadable(xprt->xp_socket))
956 		xprt_active(xprt);
957 	return (SU_OK);
958 }
959 
960 static int
961 svc_vc_rendezvous_soupcall(struct socket *head, void *arg, int waitflag)
962 {
963 	SVCXPRT *xprt = (SVCXPRT *) arg;
964 
965 	if (!TAILQ_EMPTY(&head->sol_comp))
966 		xprt_active(xprt);
967 	return (SU_OK);
968 }
969 
970 #if 0
971 /*
972  * Get the effective UID of the sending process. Used by rpcbind, keyserv
973  * and rpc.yppasswdd on AF_LOCAL.
974  */
975 int
976 __rpc_get_local_uid(SVCXPRT *transp, uid_t *uid) {
977 	int sock, ret;
978 	gid_t egid;
979 	uid_t euid;
980 	struct sockaddr *sa;
981 
982 	sock = transp->xp_fd;
983 	sa = (struct sockaddr *)transp->xp_rtaddr;
984 	if (sa->sa_family == AF_LOCAL) {
985 		ret = getpeereid(sock, &euid, &egid);
986 		if (ret == 0)
987 			*uid = euid;
988 		return (ret);
989 	} else
990 		return (-1);
991 }
992 #endif
993