xref: /freebsd/sys/rpc/svc_vc.c (revision 0b3105a37d7adcadcb720112fed4dc4e8040be99)
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 
100 static struct xp_ops svc_vc_rendezvous_ops = {
101 	.xp_recv =	svc_vc_rendezvous_recv,
102 	.xp_stat =	svc_vc_rendezvous_stat,
103 	.xp_reply =	(bool_t (*)(SVCXPRT *, struct rpc_msg *,
104 		struct sockaddr *, struct mbuf *, uint32_t *))svc_vc_null,
105 	.xp_destroy =	svc_vc_rendezvous_destroy,
106 	.xp_control =	svc_vc_rendezvous_control
107 };
108 
109 static struct xp_ops svc_vc_ops = {
110 	.xp_recv =	svc_vc_recv,
111 	.xp_stat =	svc_vc_stat,
112 	.xp_ack =	svc_vc_ack,
113 	.xp_reply =	svc_vc_reply,
114 	.xp_destroy =	svc_vc_destroy,
115 	.xp_control =	svc_vc_control
116 };
117 
118 static struct xp_ops svc_vc_backchannel_ops = {
119 	.xp_recv =	svc_vc_backchannel_recv,
120 	.xp_stat =	svc_vc_backchannel_stat,
121 	.xp_reply =	svc_vc_backchannel_reply,
122 	.xp_destroy =	svc_vc_backchannel_destroy,
123 	.xp_control =	svc_vc_backchannel_control
124 };
125 
126 /*
127  * Usage:
128  *	xprt = svc_vc_create(sock, send_buf_size, recv_buf_size);
129  *
130  * Creates, registers, and returns a (rpc) tcp based transporter.
131  * Once *xprt is initialized, it is registered as a transporter
132  * see (svc.h, xprt_register).  This routine returns
133  * a NULL if a problem occurred.
134  *
135  * The filedescriptor passed in is expected to refer to a bound, but
136  * not yet connected socket.
137  *
138  * Since streams do buffered io similar to stdio, the caller can specify
139  * how big the send and receive buffers are via the second and third parms;
140  * 0 => use the system default.
141  */
142 SVCXPRT *
143 svc_vc_create(SVCPOOL *pool, struct socket *so, size_t sendsize,
144     size_t recvsize)
145 {
146 	SVCXPRT *xprt = NULL;
147 	struct sockaddr* sa;
148 	int error;
149 
150 	SOCK_LOCK(so);
151 	if (so->so_state & (SS_ISCONNECTED|SS_ISDISCONNECTED)) {
152 		SOCK_UNLOCK(so);
153 		CURVNET_SET(so->so_vnet);
154 		error = so->so_proto->pr_usrreqs->pru_peeraddr(so, &sa);
155 		CURVNET_RESTORE();
156 		if (error)
157 			return (NULL);
158 		xprt = svc_vc_create_conn(pool, so, sa);
159 		free(sa, M_SONAME);
160 		return (xprt);
161 	}
162 	SOCK_UNLOCK(so);
163 
164 	xprt = svc_xprt_alloc();
165 	sx_init(&xprt->xp_lock, "xprt->xp_lock");
166 	xprt->xp_pool = pool;
167 	xprt->xp_socket = so;
168 	xprt->xp_p1 = NULL;
169 	xprt->xp_p2 = NULL;
170 	xprt->xp_ops = &svc_vc_rendezvous_ops;
171 
172 	CURVNET_SET(so->so_vnet);
173 	error = so->so_proto->pr_usrreqs->pru_sockaddr(so, &sa);
174 	CURVNET_RESTORE();
175 	if (error) {
176 		goto cleanup_svc_vc_create;
177 	}
178 
179 	memcpy(&xprt->xp_ltaddr, sa, sa->sa_len);
180 	free(sa, M_SONAME);
181 
182 	xprt_register(xprt);
183 
184 	solisten(so, -1, curthread);
185 
186 	SOCKBUF_LOCK(&so->so_rcv);
187 	xprt->xp_upcallset = 1;
188 	soupcall_set(so, SO_RCV, svc_vc_soupcall, xprt);
189 	SOCKBUF_UNLOCK(&so->so_rcv);
190 
191 	return (xprt);
192 cleanup_svc_vc_create:
193 	if (xprt) {
194 		sx_destroy(&xprt->xp_lock);
195 		svc_xprt_free(xprt);
196 	}
197 	return (NULL);
198 }
199 
200 /*
201  * Create a new transport for a socket optained via soaccept().
202  */
203 SVCXPRT *
204 svc_vc_create_conn(SVCPOOL *pool, struct socket *so, struct sockaddr *raddr)
205 {
206 	SVCXPRT *xprt = NULL;
207 	struct cf_conn *cd = NULL;
208 	struct sockaddr* sa = NULL;
209 	struct sockopt opt;
210 	int one = 1;
211 	int error;
212 
213 	bzero(&opt, sizeof(struct sockopt));
214 	opt.sopt_dir = SOPT_SET;
215 	opt.sopt_level = SOL_SOCKET;
216 	opt.sopt_name = SO_KEEPALIVE;
217 	opt.sopt_val = &one;
218 	opt.sopt_valsize = sizeof(one);
219 	error = sosetopt(so, &opt);
220 	if (error) {
221 		return (NULL);
222 	}
223 
224 	if (so->so_proto->pr_protocol == IPPROTO_TCP) {
225 		bzero(&opt, sizeof(struct sockopt));
226 		opt.sopt_dir = SOPT_SET;
227 		opt.sopt_level = IPPROTO_TCP;
228 		opt.sopt_name = TCP_NODELAY;
229 		opt.sopt_val = &one;
230 		opt.sopt_valsize = sizeof(one);
231 		error = sosetopt(so, &opt);
232 		if (error) {
233 			return (NULL);
234 		}
235 	}
236 
237 	cd = mem_alloc(sizeof(*cd));
238 	cd->strm_stat = XPRT_IDLE;
239 
240 	xprt = svc_xprt_alloc();
241 	sx_init(&xprt->xp_lock, "xprt->xp_lock");
242 	xprt->xp_pool = pool;
243 	xprt->xp_socket = so;
244 	xprt->xp_p1 = cd;
245 	xprt->xp_p2 = NULL;
246 	xprt->xp_ops = &svc_vc_ops;
247 
248 	/*
249 	 * See http://www.connectathon.org/talks96/nfstcp.pdf - client
250 	 * has a 5 minute timer, server has a 6 minute timer.
251 	 */
252 	xprt->xp_idletimeout = 6 * 60;
253 
254 	memcpy(&xprt->xp_rtaddr, raddr, raddr->sa_len);
255 
256 	CURVNET_SET(so->so_vnet);
257 	error = so->so_proto->pr_usrreqs->pru_sockaddr(so, &sa);
258 	CURVNET_RESTORE();
259 	if (error)
260 		goto cleanup_svc_vc_create;
261 
262 	memcpy(&xprt->xp_ltaddr, sa, sa->sa_len);
263 	free(sa, M_SONAME);
264 
265 	xprt_register(xprt);
266 
267 	SOCKBUF_LOCK(&so->so_rcv);
268 	xprt->xp_upcallset = 1;
269 	soupcall_set(so, SO_RCV, svc_vc_soupcall, xprt);
270 	SOCKBUF_UNLOCK(&so->so_rcv);
271 
272 	/*
273 	 * Throw the transport into the active list in case it already
274 	 * has some data buffered.
275 	 */
276 	sx_xlock(&xprt->xp_lock);
277 	xprt_active(xprt);
278 	sx_xunlock(&xprt->xp_lock);
279 
280 	return (xprt);
281 cleanup_svc_vc_create:
282 	if (xprt) {
283 		sx_destroy(&xprt->xp_lock);
284 		svc_xprt_free(xprt);
285 	}
286 	if (cd)
287 		mem_free(cd, sizeof(*cd));
288 	return (NULL);
289 }
290 
291 /*
292  * Create a new transport for a backchannel on a clnt_vc socket.
293  */
294 SVCXPRT *
295 svc_vc_create_backchannel(SVCPOOL *pool)
296 {
297 	SVCXPRT *xprt = NULL;
298 	struct cf_conn *cd = NULL;
299 
300 	cd = mem_alloc(sizeof(*cd));
301 	cd->strm_stat = XPRT_IDLE;
302 
303 	xprt = svc_xprt_alloc();
304 	sx_init(&xprt->xp_lock, "xprt->xp_lock");
305 	xprt->xp_pool = pool;
306 	xprt->xp_socket = NULL;
307 	xprt->xp_p1 = cd;
308 	xprt->xp_p2 = NULL;
309 	xprt->xp_ops = &svc_vc_backchannel_ops;
310 	return (xprt);
311 }
312 
313 /*
314  * This does all of the accept except the final call to soaccept. The
315  * caller will call soaccept after dropping its locks (soaccept may
316  * call malloc).
317  */
318 int
319 svc_vc_accept(struct socket *head, struct socket **sop)
320 {
321 	int error = 0;
322 	struct socket *so;
323 
324 	if ((head->so_options & SO_ACCEPTCONN) == 0) {
325 		error = EINVAL;
326 		goto done;
327 	}
328 #ifdef MAC
329 	error = mac_socket_check_accept(curthread->td_ucred, head);
330 	if (error != 0)
331 		goto done;
332 #endif
333 	ACCEPT_LOCK();
334 	if (TAILQ_EMPTY(&head->so_comp)) {
335 		ACCEPT_UNLOCK();
336 		error = EWOULDBLOCK;
337 		goto done;
338 	}
339 	so = TAILQ_FIRST(&head->so_comp);
340 	KASSERT(!(so->so_qstate & SQ_INCOMP), ("svc_vc_accept: so SQ_INCOMP"));
341 	KASSERT(so->so_qstate & SQ_COMP, ("svc_vc_accept: so not SQ_COMP"));
342 
343 	/*
344 	 * Before changing the flags on the socket, we have to bump the
345 	 * reference count.  Otherwise, if the protocol calls sofree(),
346 	 * the socket will be released due to a zero refcount.
347 	 * XXX might not need soref() since this is simpler than kern_accept.
348 	 */
349 	SOCK_LOCK(so);			/* soref() and so_state update */
350 	soref(so);			/* file descriptor reference */
351 
352 	TAILQ_REMOVE(&head->so_comp, so, so_list);
353 	head->so_qlen--;
354 	so->so_state |= (head->so_state & SS_NBIO);
355 	so->so_qstate &= ~SQ_COMP;
356 	so->so_head = NULL;
357 
358 	SOCK_UNLOCK(so);
359 	ACCEPT_UNLOCK();
360 
361 	*sop = so;
362 
363 	/* connection has been removed from the listen queue */
364 	KNOTE_UNLOCKED(&head->so_rcv.sb_sel.si_note, 0);
365 done:
366 	return (error);
367 }
368 
369 /*ARGSUSED*/
370 static bool_t
371 svc_vc_rendezvous_recv(SVCXPRT *xprt, struct rpc_msg *msg,
372     struct sockaddr **addrp, struct mbuf **mp)
373 {
374 	struct socket *so = NULL;
375 	struct sockaddr *sa = NULL;
376 	int error;
377 	SVCXPRT *new_xprt;
378 
379 	/*
380 	 * The socket upcall calls xprt_active() which will eventually
381 	 * cause the server to call us here. We attempt to accept a
382 	 * connection from the socket and turn it into a new
383 	 * transport. If the accept fails, we have drained all pending
384 	 * connections so we call xprt_inactive().
385 	 */
386 	sx_xlock(&xprt->xp_lock);
387 
388 	error = svc_vc_accept(xprt->xp_socket, &so);
389 
390 	if (error == EWOULDBLOCK) {
391 		/*
392 		 * We must re-test for new connections after taking
393 		 * the lock to protect us in the case where a new
394 		 * connection arrives after our call to accept fails
395 		 * with EWOULDBLOCK.
396 		 */
397 		ACCEPT_LOCK();
398 		if (TAILQ_EMPTY(&xprt->xp_socket->so_comp))
399 			xprt_inactive_self(xprt);
400 		ACCEPT_UNLOCK();
401 		sx_xunlock(&xprt->xp_lock);
402 		return (FALSE);
403 	}
404 
405 	if (error) {
406 		SOCKBUF_LOCK(&xprt->xp_socket->so_rcv);
407 		if (xprt->xp_upcallset) {
408 			xprt->xp_upcallset = 0;
409 			soupcall_clear(xprt->xp_socket, SO_RCV);
410 		}
411 		SOCKBUF_UNLOCK(&xprt->xp_socket->so_rcv);
412 		xprt_inactive_self(xprt);
413 		sx_xunlock(&xprt->xp_lock);
414 		return (FALSE);
415 	}
416 
417 	sx_xunlock(&xprt->xp_lock);
418 
419 	sa = 0;
420 	error = soaccept(so, &sa);
421 
422 	if (error) {
423 		/*
424 		 * XXX not sure if I need to call sofree or soclose here.
425 		 */
426 		if (sa)
427 			free(sa, M_SONAME);
428 		return (FALSE);
429 	}
430 
431 	/*
432 	 * svc_vc_create_conn will call xprt_register - we don't need
433 	 * to do anything with the new connection except derefence it.
434 	 */
435 	new_xprt = svc_vc_create_conn(xprt->xp_pool, so, sa);
436 	if (!new_xprt) {
437 		soclose(so);
438 	} else {
439 		SVC_RELEASE(new_xprt);
440 	}
441 
442 	free(sa, M_SONAME);
443 
444 	return (FALSE); /* there is never an rpc msg to be processed */
445 }
446 
447 /*ARGSUSED*/
448 static enum xprt_stat
449 svc_vc_rendezvous_stat(SVCXPRT *xprt)
450 {
451 
452 	return (XPRT_IDLE);
453 }
454 
455 static void
456 svc_vc_destroy_common(SVCXPRT *xprt)
457 {
458 	SOCKBUF_LOCK(&xprt->xp_socket->so_rcv);
459 	if (xprt->xp_upcallset) {
460 		xprt->xp_upcallset = 0;
461 		soupcall_clear(xprt->xp_socket, SO_RCV);
462 	}
463 	SOCKBUF_UNLOCK(&xprt->xp_socket->so_rcv);
464 
465 	if (xprt->xp_socket)
466 		(void)soclose(xprt->xp_socket);
467 
468 	if (xprt->xp_netid)
469 		(void) mem_free(xprt->xp_netid, strlen(xprt->xp_netid) + 1);
470 	svc_xprt_free(xprt);
471 }
472 
473 static void
474 svc_vc_rendezvous_destroy(SVCXPRT *xprt)
475 {
476 
477 	svc_vc_destroy_common(xprt);
478 }
479 
480 static void
481 svc_vc_destroy(SVCXPRT *xprt)
482 {
483 	struct cf_conn *cd = (struct cf_conn *)xprt->xp_p1;
484 
485 	svc_vc_destroy_common(xprt);
486 
487 	if (cd->mreq)
488 		m_freem(cd->mreq);
489 	if (cd->mpending)
490 		m_freem(cd->mpending);
491 	mem_free(cd, sizeof(*cd));
492 }
493 
494 static void
495 svc_vc_backchannel_destroy(SVCXPRT *xprt)
496 {
497 	struct cf_conn *cd = (struct cf_conn *)xprt->xp_p1;
498 	struct mbuf *m, *m2;
499 
500 	svc_xprt_free(xprt);
501 	m = cd->mreq;
502 	while (m != NULL) {
503 		m2 = m;
504 		m = m->m_nextpkt;
505 		m_freem(m2);
506 	}
507 	mem_free(cd, sizeof(*cd));
508 }
509 
510 /*ARGSUSED*/
511 static bool_t
512 svc_vc_control(SVCXPRT *xprt, const u_int rq, void *in)
513 {
514 	return (FALSE);
515 }
516 
517 static bool_t
518 svc_vc_rendezvous_control(SVCXPRT *xprt, const u_int rq, void *in)
519 {
520 
521 	return (FALSE);
522 }
523 
524 static bool_t
525 svc_vc_backchannel_control(SVCXPRT *xprt, const u_int rq, void *in)
526 {
527 
528 	return (FALSE);
529 }
530 
531 static enum xprt_stat
532 svc_vc_stat(SVCXPRT *xprt)
533 {
534 	struct cf_conn *cd;
535 
536 	cd = (struct cf_conn *)(xprt->xp_p1);
537 
538 	if (cd->strm_stat == XPRT_DIED)
539 		return (XPRT_DIED);
540 
541 	if (cd->mreq != NULL && cd->resid == 0 && cd->eor)
542 		return (XPRT_MOREREQS);
543 
544 	if (soreadable(xprt->xp_socket))
545 		return (XPRT_MOREREQS);
546 
547 	return (XPRT_IDLE);
548 }
549 
550 static bool_t
551 svc_vc_ack(SVCXPRT *xprt, uint32_t *ack)
552 {
553 
554 	*ack = atomic_load_acq_32(&xprt->xp_snt_cnt);
555 	*ack -= sbused(&xprt->xp_socket->so_snd);
556 	return (TRUE);
557 }
558 
559 static enum xprt_stat
560 svc_vc_backchannel_stat(SVCXPRT *xprt)
561 {
562 	struct cf_conn *cd;
563 
564 	cd = (struct cf_conn *)(xprt->xp_p1);
565 
566 	if (cd->mreq != NULL)
567 		return (XPRT_MOREREQS);
568 
569 	return (XPRT_IDLE);
570 }
571 
572 /*
573  * If we have an mbuf chain in cd->mpending, try to parse a record from it,
574  * leaving the result in cd->mreq. If we don't have a complete record, leave
575  * the partial result in cd->mreq and try to read more from the socket.
576  */
577 static int
578 svc_vc_process_pending(SVCXPRT *xprt)
579 {
580 	struct cf_conn *cd = (struct cf_conn *) xprt->xp_p1;
581 	struct socket *so = xprt->xp_socket;
582 	struct mbuf *m;
583 
584 	/*
585 	 * If cd->resid is non-zero, we have part of the
586 	 * record already, otherwise we are expecting a record
587 	 * marker.
588 	 */
589 	if (!cd->resid && cd->mpending) {
590 		/*
591 		 * See if there is enough data buffered to
592 		 * make up a record marker. Make sure we can
593 		 * handle the case where the record marker is
594 		 * split across more than one mbuf.
595 		 */
596 		size_t n = 0;
597 		uint32_t header;
598 
599 		m = cd->mpending;
600 		while (n < sizeof(uint32_t) && m) {
601 			n += m->m_len;
602 			m = m->m_next;
603 		}
604 		if (n < sizeof(uint32_t)) {
605 			so->so_rcv.sb_lowat = sizeof(uint32_t) - n;
606 			return (FALSE);
607 		}
608 		m_copydata(cd->mpending, 0, sizeof(header),
609 		    (char *)&header);
610 		header = ntohl(header);
611 		cd->eor = (header & 0x80000000) != 0;
612 		cd->resid = header & 0x7fffffff;
613 		m_adj(cd->mpending, sizeof(uint32_t));
614 	}
615 
616 	/*
617 	 * Start pulling off mbufs from cd->mpending
618 	 * until we either have a complete record or
619 	 * we run out of data. We use m_split to pull
620 	 * data - it will pull as much as possible and
621 	 * split the last mbuf if necessary.
622 	 */
623 	while (cd->mpending && cd->resid) {
624 		m = cd->mpending;
625 		if (cd->mpending->m_next
626 		    || cd->mpending->m_len > cd->resid)
627 			cd->mpending = m_split(cd->mpending,
628 			    cd->resid, M_WAITOK);
629 		else
630 			cd->mpending = NULL;
631 		if (cd->mreq)
632 			m_last(cd->mreq)->m_next = m;
633 		else
634 			cd->mreq = m;
635 		while (m) {
636 			cd->resid -= m->m_len;
637 			m = m->m_next;
638 		}
639 	}
640 
641 	/*
642 	 * Block receive upcalls if we have more data pending,
643 	 * otherwise report our need.
644 	 */
645 	if (cd->mpending)
646 		so->so_rcv.sb_lowat = INT_MAX;
647 	else
648 		so->so_rcv.sb_lowat =
649 		    imax(1, imin(cd->resid, so->so_rcv.sb_hiwat / 2));
650 	return (TRUE);
651 }
652 
653 static bool_t
654 svc_vc_recv(SVCXPRT *xprt, struct rpc_msg *msg,
655     struct sockaddr **addrp, struct mbuf **mp)
656 {
657 	struct cf_conn *cd = (struct cf_conn *) xprt->xp_p1;
658 	struct uio uio;
659 	struct mbuf *m;
660 	struct socket* so = xprt->xp_socket;
661 	XDR xdrs;
662 	int error, rcvflag;
663 	uint32_t xid_plus_direction[2];
664 
665 	/*
666 	 * Serialise access to the socket and our own record parsing
667 	 * state.
668 	 */
669 	sx_xlock(&xprt->xp_lock);
670 
671 	for (;;) {
672 		/* If we have no request ready, check pending queue. */
673 		while (cd->mpending &&
674 		    (cd->mreq == NULL || cd->resid != 0 || !cd->eor)) {
675 			if (!svc_vc_process_pending(xprt))
676 				break;
677 		}
678 
679 		/* Process and return complete request in cd->mreq. */
680 		if (cd->mreq != NULL && cd->resid == 0 && cd->eor) {
681 
682 			/*
683 			 * Now, check for a backchannel reply.
684 			 * The XID is in the first uint32_t of the reply
685 			 * and the message direction is the second one.
686 			 */
687 			if ((cd->mreq->m_len >= sizeof(xid_plus_direction) ||
688 			    m_length(cd->mreq, NULL) >=
689 			    sizeof(xid_plus_direction)) &&
690 			    xprt->xp_p2 != NULL) {
691 				m_copydata(cd->mreq, 0,
692 				    sizeof(xid_plus_direction),
693 				    (char *)xid_plus_direction);
694 				xid_plus_direction[0] =
695 				    ntohl(xid_plus_direction[0]);
696 				xid_plus_direction[1] =
697 				    ntohl(xid_plus_direction[1]);
698 				/* Check message direction. */
699 				if (xid_plus_direction[1] == REPLY) {
700 					clnt_bck_svccall(xprt->xp_p2,
701 					    cd->mreq,
702 					    xid_plus_direction[0]);
703 					cd->mreq = NULL;
704 					continue;
705 				}
706 			}
707 
708 			xdrmbuf_create(&xdrs, cd->mreq, XDR_DECODE);
709 			cd->mreq = NULL;
710 
711 			/* Check for next request in a pending queue. */
712 			svc_vc_process_pending(xprt);
713 			if (cd->mreq == NULL || cd->resid != 0) {
714 				SOCKBUF_LOCK(&so->so_rcv);
715 				if (!soreadable(so))
716 					xprt_inactive_self(xprt);
717 				SOCKBUF_UNLOCK(&so->so_rcv);
718 			}
719 
720 			sx_xunlock(&xprt->xp_lock);
721 
722 			if (! xdr_callmsg(&xdrs, msg)) {
723 				XDR_DESTROY(&xdrs);
724 				return (FALSE);
725 			}
726 
727 			*addrp = NULL;
728 			*mp = xdrmbuf_getall(&xdrs);
729 			XDR_DESTROY(&xdrs);
730 
731 			return (TRUE);
732 		}
733 
734 		/*
735 		 * The socket upcall calls xprt_active() which will eventually
736 		 * cause the server to call us here. We attempt to
737 		 * read as much as possible from the socket and put
738 		 * the result in cd->mpending. If the read fails,
739 		 * we have drained both cd->mpending and the socket so
740 		 * we can call xprt_inactive().
741 		 */
742 		uio.uio_resid = 1000000000;
743 		uio.uio_td = curthread;
744 		m = NULL;
745 		rcvflag = MSG_DONTWAIT;
746 		error = soreceive(so, NULL, &uio, &m, NULL, &rcvflag);
747 
748 		if (error == EWOULDBLOCK) {
749 			/*
750 			 * We must re-test for readability after
751 			 * taking the lock to protect us in the case
752 			 * where a new packet arrives on the socket
753 			 * after our call to soreceive fails with
754 			 * EWOULDBLOCK.
755 			 */
756 			SOCKBUF_LOCK(&so->so_rcv);
757 			if (!soreadable(so))
758 				xprt_inactive_self(xprt);
759 			SOCKBUF_UNLOCK(&so->so_rcv);
760 			sx_xunlock(&xprt->xp_lock);
761 			return (FALSE);
762 		}
763 
764 		if (error) {
765 			SOCKBUF_LOCK(&so->so_rcv);
766 			if (xprt->xp_upcallset) {
767 				xprt->xp_upcallset = 0;
768 				soupcall_clear(so, SO_RCV);
769 			}
770 			SOCKBUF_UNLOCK(&so->so_rcv);
771 			xprt_inactive_self(xprt);
772 			cd->strm_stat = XPRT_DIED;
773 			sx_xunlock(&xprt->xp_lock);
774 			return (FALSE);
775 		}
776 
777 		if (!m) {
778 			/*
779 			 * EOF - the other end has closed the socket.
780 			 */
781 			xprt_inactive_self(xprt);
782 			cd->strm_stat = XPRT_DIED;
783 			sx_xunlock(&xprt->xp_lock);
784 			return (FALSE);
785 		}
786 
787 		if (cd->mpending)
788 			m_last(cd->mpending)->m_next = m;
789 		else
790 			cd->mpending = m;
791 	}
792 }
793 
794 static bool_t
795 svc_vc_backchannel_recv(SVCXPRT *xprt, struct rpc_msg *msg,
796     struct sockaddr **addrp, struct mbuf **mp)
797 {
798 	struct cf_conn *cd = (struct cf_conn *) xprt->xp_p1;
799 	struct ct_data *ct;
800 	struct mbuf *m;
801 	XDR xdrs;
802 
803 	sx_xlock(&xprt->xp_lock);
804 	ct = (struct ct_data *)xprt->xp_p2;
805 	if (ct == NULL) {
806 		sx_xunlock(&xprt->xp_lock);
807 		return (FALSE);
808 	}
809 	mtx_lock(&ct->ct_lock);
810 	m = cd->mreq;
811 	if (m == NULL) {
812 		xprt_inactive_self(xprt);
813 		mtx_unlock(&ct->ct_lock);
814 		sx_xunlock(&xprt->xp_lock);
815 		return (FALSE);
816 	}
817 	cd->mreq = m->m_nextpkt;
818 	mtx_unlock(&ct->ct_lock);
819 	sx_xunlock(&xprt->xp_lock);
820 
821 	xdrmbuf_create(&xdrs, m, XDR_DECODE);
822 	if (! xdr_callmsg(&xdrs, msg)) {
823 		XDR_DESTROY(&xdrs);
824 		return (FALSE);
825 	}
826 	*addrp = NULL;
827 	*mp = xdrmbuf_getall(&xdrs);
828 	XDR_DESTROY(&xdrs);
829 	return (TRUE);
830 }
831 
832 static bool_t
833 svc_vc_reply(SVCXPRT *xprt, struct rpc_msg *msg,
834     struct sockaddr *addr, struct mbuf *m, uint32_t *seq)
835 {
836 	XDR xdrs;
837 	struct mbuf *mrep;
838 	bool_t stat = TRUE;
839 	int error, len;
840 
841 	/*
842 	 * Leave space for record mark.
843 	 */
844 	mrep = m_gethdr(M_WAITOK, MT_DATA);
845 	mrep->m_data += sizeof(uint32_t);
846 
847 	xdrmbuf_create(&xdrs, mrep, XDR_ENCODE);
848 
849 	if (msg->rm_reply.rp_stat == MSG_ACCEPTED &&
850 	    msg->rm_reply.rp_acpt.ar_stat == SUCCESS) {
851 		if (!xdr_replymsg(&xdrs, msg))
852 			stat = FALSE;
853 		else
854 			xdrmbuf_append(&xdrs, m);
855 	} else {
856 		stat = xdr_replymsg(&xdrs, msg);
857 	}
858 
859 	if (stat) {
860 		m_fixhdr(mrep);
861 
862 		/*
863 		 * Prepend a record marker containing the reply length.
864 		 */
865 		M_PREPEND(mrep, sizeof(uint32_t), M_WAITOK);
866 		len = mrep->m_pkthdr.len;
867 		*mtod(mrep, uint32_t *) =
868 			htonl(0x80000000 | (len - sizeof(uint32_t)));
869 		atomic_add_32(&xprt->xp_snd_cnt, len);
870 		error = sosend(xprt->xp_socket, NULL, NULL, mrep, NULL,
871 		    0, curthread);
872 		if (!error) {
873 			atomic_add_rel_32(&xprt->xp_snt_cnt, len);
874 			if (seq)
875 				*seq = xprt->xp_snd_cnt;
876 			stat = TRUE;
877 		} else
878 			atomic_subtract_32(&xprt->xp_snd_cnt, len);
879 	} else {
880 		m_freem(mrep);
881 	}
882 
883 	XDR_DESTROY(&xdrs);
884 
885 	return (stat);
886 }
887 
888 static bool_t
889 svc_vc_backchannel_reply(SVCXPRT *xprt, struct rpc_msg *msg,
890     struct sockaddr *addr, struct mbuf *m, uint32_t *seq)
891 {
892 	struct ct_data *ct;
893 	XDR xdrs;
894 	struct mbuf *mrep;
895 	bool_t stat = TRUE;
896 	int error;
897 
898 	/*
899 	 * Leave space for record mark.
900 	 */
901 	mrep = m_gethdr(M_WAITOK, MT_DATA);
902 	mrep->m_data += sizeof(uint32_t);
903 
904 	xdrmbuf_create(&xdrs, mrep, XDR_ENCODE);
905 
906 	if (msg->rm_reply.rp_stat == MSG_ACCEPTED &&
907 	    msg->rm_reply.rp_acpt.ar_stat == SUCCESS) {
908 		if (!xdr_replymsg(&xdrs, msg))
909 			stat = FALSE;
910 		else
911 			xdrmbuf_append(&xdrs, m);
912 	} else {
913 		stat = xdr_replymsg(&xdrs, msg);
914 	}
915 
916 	if (stat) {
917 		m_fixhdr(mrep);
918 
919 		/*
920 		 * Prepend a record marker containing the reply length.
921 		 */
922 		M_PREPEND(mrep, sizeof(uint32_t), M_WAITOK);
923 		*mtod(mrep, uint32_t *) =
924 			htonl(0x80000000 | (mrep->m_pkthdr.len
925 				- sizeof(uint32_t)));
926 		sx_xlock(&xprt->xp_lock);
927 		ct = (struct ct_data *)xprt->xp_p2;
928 		if (ct != NULL)
929 			error = sosend(ct->ct_socket, NULL, NULL, mrep, NULL,
930 			    0, curthread);
931 		else
932 			error = EPIPE;
933 		sx_xunlock(&xprt->xp_lock);
934 		if (!error) {
935 			stat = TRUE;
936 		}
937 	} else {
938 		m_freem(mrep);
939 	}
940 
941 	XDR_DESTROY(&xdrs);
942 
943 	return (stat);
944 }
945 
946 static bool_t
947 svc_vc_null()
948 {
949 
950 	return (FALSE);
951 }
952 
953 static int
954 svc_vc_soupcall(struct socket *so, void *arg, int waitflag)
955 {
956 	SVCXPRT *xprt = (SVCXPRT *) arg;
957 
958 	if (soreadable(xprt->xp_socket))
959 		xprt_active(xprt);
960 	return (SU_OK);
961 }
962 
963 #if 0
964 /*
965  * Get the effective UID of the sending process. Used by rpcbind, keyserv
966  * and rpc.yppasswdd on AF_LOCAL.
967  */
968 int
969 __rpc_get_local_uid(SVCXPRT *transp, uid_t *uid) {
970 	int sock, ret;
971 	gid_t egid;
972 	uid_t euid;
973 	struct sockaddr *sa;
974 
975 	sock = transp->xp_fd;
976 	sa = (struct sockaddr *)transp->xp_rtaddr;
977 	if (sa->sa_family == AF_LOCAL) {
978 		ret = getpeereid(sock, &euid, &egid);
979 		if (ret == 0)
980 			*uid = euid;
981 		return (ret);
982 	} else
983 		return (-1);
984 }
985 #endif
986