2 * linux/net/sunrpc/xprt.c
4 * This is a generic RPC call interface supporting congestion avoidance,
5 * and asynchronous calls.
7 * The interface works like this:
9 * - When a process places a call, it allocates a request slot if
10 * one is available. Otherwise, it sleeps on the backlog queue
12 * - Next, the caller puts together the RPC message, stuffs it into
13 * the request struct, and calls xprt_call().
14 * - xprt_call transmits the message and installs the caller on the
15 * socket's wait list. At the same time, it installs a timer that
16 * is run after the packet's timeout has expired.
17 * - When a packet arrives, the data_ready handler walks the list of
18 * pending requests for that socket. If a matching XID is found, the
19 * caller is woken up, and the timer removed.
20 * - When no reply arrives within the timeout interval, the timer is
21 * fired by the kernel and runs xprt_timer(). It either adjusts the
22 * timeout values (minor timeout) or wakes up the caller with a status
24 * - When the caller receives a notification from RPC that a reply arrived,
25 * it should release the RPC slot, and process the reply.
26 * If the call timed out, it may choose to retry the operation by
27 * adjusting the initial timeout value, and simply calling rpc_call
30 * Support for async RPC is done through a set of RPC-specific scheduling
31 * primitives that `transparently' work for processes as well as async
32 * tasks that rely on callbacks.
34 * Copyright (C) 1995-1997, Olaf Kirch <okir@monad.swb.de>
36 * TCP callback races fixes (C) 1998 Red Hat Software <alan@redhat.com>
37 * TCP send fixes (C) 1998 Red Hat Software <alan@redhat.com>
38 * TCP NFS related read + write fixes
39 * (C) 1999 Dave Airlie, University of Limerick, Ireland <airlied@linux.ie>
41 * Rewrite of larges part of the code in order to stabilize TCP stuff.
42 * Fix behaviour when socket buffer is full.
43 * (C) 1999 Trond Myklebust <trond.myklebust@fys.uio.no>
46 #include <linux/types.h>
47 #include <linux/slab.h>
48 #include <linux/capability.h>
49 #include <linux/sched.h>
50 #include <linux/errno.h>
51 #include <linux/socket.h>
53 #include <linux/net.h>
55 #include <linux/udp.h>
56 #include <linux/tcp.h>
57 #include <linux/sunrpc/clnt.h>
58 #include <linux/file.h>
59 #include <linux/workqueue.h>
60 #include <linux/random.h>
63 #include <net/checksum.h>
72 # undef RPC_DEBUG_DATA
73 # define RPCDBG_FACILITY RPCDBG_XPRT
76 #define XPRT_MAX_BACKOFF (8)
77 #define XPRT_IDLE_TIMEOUT (5*60*HZ)
78 #define XPRT_MAX_RESVPORT (800)
83 static void xprt_request_init(struct rpc_task *, struct rpc_xprt *);
84 static inline void do_xprt_reserve(struct rpc_task *);
85 static void xprt_disconnect(struct rpc_xprt *);
86 static void xprt_connect_status(struct rpc_task *task);
87 static struct rpc_xprt * xprt_setup(int proto, struct sockaddr_in *ap,
88 struct rpc_timeout *to);
89 static struct socket *xprt_create_socket(struct rpc_xprt *, int, int);
90 static void xprt_bind_socket(struct rpc_xprt *, struct socket *);
91 static int __xprt_get_cong(struct rpc_xprt *, struct rpc_task *);
93 static int xprt_clear_backlog(struct rpc_xprt *xprt);
97 * Print the buffer contents (first 128 bytes only--just enough for
101 xprt_pktdump(char *msg, u32 *packet, unsigned int count)
103 u8 *buf = (u8 *) packet;
106 dprintk("RPC: %s\n", msg);
107 for (j = 0; j < count && j < 128; j += 4) {
111 dprintk("0x%04x ", j);
113 dprintk("%02x%02x%02x%02x ",
114 buf[j], buf[j+1], buf[j+2], buf[j+3]);
120 xprt_pktdump(char *msg, u32 *packet, unsigned int count)
127 * Look up RPC transport given an INET socket
129 static inline struct rpc_xprt *
130 xprt_from_sock(struct sock *sk)
132 return (struct rpc_xprt *) sk->sk_user_data;
136 * Serialize write access to sockets, in order to prevent different
137 * requests from interfering with each other.
138 * Also prevents TCP socket connects from colliding with writes.
141 __xprt_lock_write(struct rpc_xprt *xprt, struct rpc_task *task)
143 struct rpc_rqst *req = task->tk_rqstp;
145 if (test_and_set_bit(XPRT_LOCKED, &xprt->sockstate)) {
146 if (task == xprt->snd_task)
150 if (xprt->nocong || __xprt_get_cong(xprt, task)) {
151 xprt->snd_task = task;
153 req->rq_bytes_sent = 0;
158 smp_mb__before_clear_bit();
159 clear_bit(XPRT_LOCKED, &xprt->sockstate);
160 smp_mb__after_clear_bit();
162 dprintk("RPC: %4d failed to lock socket %p\n", task->tk_pid, xprt);
163 task->tk_timeout = 0;
164 task->tk_status = -EAGAIN;
165 if (req && req->rq_ntrans)
166 rpc_sleep_on(&xprt->resend, task, NULL, NULL);
168 rpc_sleep_on(&xprt->sending, task, NULL, NULL);
173 xprt_lock_write(struct rpc_xprt *xprt, struct rpc_task *task)
177 spin_lock_bh(&xprt->sock_lock);
178 retval = __xprt_lock_write(xprt, task);
179 spin_unlock_bh(&xprt->sock_lock);
185 __xprt_lock_write_next(struct rpc_xprt *xprt)
187 struct rpc_task *task;
189 if (test_and_set_bit(XPRT_LOCKED, &xprt->sockstate))
191 if (!xprt->nocong && RPCXPRT_CONGESTED(xprt))
193 task = rpc_wake_up_next(&xprt->resend);
195 task = rpc_wake_up_next(&xprt->sending);
199 if (xprt->nocong || __xprt_get_cong(xprt, task)) {
200 struct rpc_rqst *req = task->tk_rqstp;
201 xprt->snd_task = task;
203 req->rq_bytes_sent = 0;
209 smp_mb__before_clear_bit();
210 clear_bit(XPRT_LOCKED, &xprt->sockstate);
211 smp_mb__after_clear_bit();
215 * Releases the socket for use by other requests.
218 __xprt_release_write(struct rpc_xprt *xprt, struct rpc_task *task)
220 if (xprt->snd_task == task) {
221 xprt->snd_task = NULL;
222 smp_mb__before_clear_bit();
223 clear_bit(XPRT_LOCKED, &xprt->sockstate);
224 smp_mb__after_clear_bit();
225 __xprt_lock_write_next(xprt);
230 xprt_release_write(struct rpc_xprt *xprt, struct rpc_task *task)
232 spin_lock_bh(&xprt->sock_lock);
233 __xprt_release_write(xprt, task);
234 spin_unlock_bh(&xprt->sock_lock);
238 * Write data to socket.
241 xprt_sendmsg(struct rpc_xprt *xprt, struct rpc_rqst *req)
243 struct socket *sock = xprt->sock;
244 struct xdr_buf *xdr = &req->rq_snd_buf;
245 struct sockaddr *addr = NULL;
253 xprt_pktdump("packet data:",
254 req->rq_svec->iov_base,
255 req->rq_svec->iov_len);
257 /* For UDP, we need to provide an address */
259 addr = (struct sockaddr *) &xprt->addr;
260 addrlen = sizeof(xprt->addr);
262 /* Dont repeat bytes */
263 skip = req->rq_bytes_sent;
265 clear_bit(SOCK_ASYNC_NOSPACE, &sock->flags);
266 result = xdr_sendpages(sock, addr, addrlen, xdr, skip, MSG_DONTWAIT);
268 dprintk("RPC: xprt_sendmsg(%d) = %d\n", xdr->len - skip, result);
275 /* When the server has died, an ICMP port unreachable message
276 * prompts ECONNREFUSED.
283 /* connection broken */
288 printk(KERN_NOTICE "RPC: sendmsg returned error %d\n", -result);
294 * Van Jacobson congestion avoidance. Check if the congestion window
295 * overflowed. Put the task to sleep if this is the case.
298 __xprt_get_cong(struct rpc_xprt *xprt, struct rpc_task *task)
300 struct rpc_rqst *req = task->tk_rqstp;
304 dprintk("RPC: %4d xprt_cwnd_limited cong = %ld cwnd = %ld\n",
305 task->tk_pid, xprt->cong, xprt->cwnd);
306 if (RPCXPRT_CONGESTED(xprt))
309 xprt->cong += RPC_CWNDSCALE;
314 * Adjust the congestion window, and wake up the next task
315 * that has been sleeping due to congestion
318 __xprt_put_cong(struct rpc_xprt *xprt, struct rpc_rqst *req)
323 xprt->cong -= RPC_CWNDSCALE;
324 __xprt_lock_write_next(xprt);
328 * Adjust RPC congestion window
329 * We use a time-smoothed congestion estimator to avoid heavy oscillation.
332 xprt_adjust_cwnd(struct rpc_xprt *xprt, int result)
337 if (result >= 0 && cwnd <= xprt->cong) {
338 /* The (cwnd >> 1) term makes sure
339 * the result gets rounded properly. */
340 cwnd += (RPC_CWNDSCALE * RPC_CWNDSCALE + (cwnd >> 1)) / cwnd;
341 if (cwnd > RPC_MAXCWND(xprt))
342 cwnd = RPC_MAXCWND(xprt);
343 __xprt_lock_write_next(xprt);
344 } else if (result == -ETIMEDOUT) {
346 if (cwnd < RPC_CWNDSCALE)
347 cwnd = RPC_CWNDSCALE;
349 dprintk("RPC: cong %ld, cwnd was %ld, now %ld\n",
350 xprt->cong, xprt->cwnd, cwnd);
355 * Reset the major timeout value
357 static void xprt_reset_majortimeo(struct rpc_rqst *req)
359 struct rpc_timeout *to = &req->rq_xprt->timeout;
361 req->rq_majortimeo = req->rq_timeout;
362 if (to->to_exponential)
363 req->rq_majortimeo <<= to->to_retries;
365 req->rq_majortimeo += to->to_increment * to->to_retries;
366 if (req->rq_majortimeo > to->to_maxval || req->rq_majortimeo == 0)
367 req->rq_majortimeo = to->to_maxval;
368 req->rq_majortimeo += jiffies;
372 * Adjust timeout values etc for next retransmit
374 int xprt_adjust_timeout(struct rpc_rqst *req)
376 struct rpc_xprt *xprt = req->rq_xprt;
377 struct rpc_timeout *to = &xprt->timeout;
380 if (time_before(jiffies, req->rq_majortimeo)) {
381 if (to->to_exponential)
382 req->rq_timeout <<= 1;
384 req->rq_timeout += to->to_increment;
385 if (to->to_maxval && req->rq_timeout >= to->to_maxval)
386 req->rq_timeout = to->to_maxval;
388 pprintk("RPC: %lu retrans\n", jiffies);
390 req->rq_timeout = to->to_initval;
392 xprt_reset_majortimeo(req);
393 /* Reset the RTT counters == "slow start" */
394 spin_lock_bh(&xprt->sock_lock);
395 rpc_init_rtt(req->rq_task->tk_client->cl_rtt, to->to_initval);
396 spin_unlock_bh(&xprt->sock_lock);
397 pprintk("RPC: %lu timeout\n", jiffies);
401 if (req->rq_timeout == 0) {
402 printk(KERN_WARNING "xprt_adjust_timeout: rq_timeout = 0!\n");
403 req->rq_timeout = 5 * HZ;
409 * Close down a transport socket
412 xprt_close(struct rpc_xprt *xprt)
414 struct socket *sock = xprt->sock;
415 struct sock *sk = xprt->inet;
420 write_lock_bh(&sk->sk_callback_lock);
424 sk->sk_user_data = NULL;
425 sk->sk_data_ready = xprt->old_data_ready;
426 sk->sk_state_change = xprt->old_state_change;
427 sk->sk_write_space = xprt->old_write_space;
428 write_unlock_bh(&sk->sk_callback_lock);
436 xprt_socket_autoclose(void *args)
438 struct rpc_xprt *xprt = (struct rpc_xprt *)args;
440 xprt_disconnect(xprt);
442 xprt_release_write(xprt, NULL);
446 * Mark a transport as disconnected
449 xprt_disconnect(struct rpc_xprt *xprt)
451 dprintk("RPC: disconnected transport %p\n", xprt);
452 spin_lock_bh(&xprt->sock_lock);
453 xprt_clear_connected(xprt);
454 rpc_wake_up_status(&xprt->pending, -ENOTCONN);
455 spin_unlock_bh(&xprt->sock_lock);
459 * Used to allow disconnection when we've been idle
462 xprt_init_autodisconnect(unsigned long data)
464 struct rpc_xprt *xprt = (struct rpc_xprt *)data;
466 spin_lock(&xprt->sock_lock);
467 if (!list_empty(&xprt->recv) || xprt->shutdown)
469 if (test_and_set_bit(XPRT_LOCKED, &xprt->sockstate))
471 spin_unlock(&xprt->sock_lock);
472 /* Let keventd close the socket */
473 if (test_bit(XPRT_CONNECTING, &xprt->sockstate) != 0)
474 xprt_release_write(xprt, NULL);
476 schedule_work(&xprt->task_cleanup);
479 spin_unlock(&xprt->sock_lock);
482 static void xprt_socket_connect(void *args)
484 struct rpc_xprt *xprt = (struct rpc_xprt *)args;
485 struct socket *sock = xprt->sock;
488 if (xprt->shutdown || xprt->addr.sin_port == 0)
492 * Start by resetting any existing state
495 sock = xprt_create_socket(xprt, xprt->prot, xprt->resvport);
497 /* couldn't create socket or bind to reserved port;
498 * this is likely a permanent error, so cause an abort */
501 xprt_bind_socket(xprt, sock);
502 xprt_sock_setbufsize(xprt);
509 * Tell the socket layer to start connecting...
511 status = sock->ops->connect(sock, (struct sockaddr *) &xprt->addr,
512 sizeof(xprt->addr), O_NONBLOCK);
513 dprintk("RPC: %p connect status %d connected %d sock state %d\n",
514 xprt, -status, xprt_connected(xprt), sock->sk->sk_state);
524 rpc_wake_up_status(&xprt->pending, status);
526 rpc_wake_up(&xprt->pending);
528 smp_mb__before_clear_bit();
529 clear_bit(XPRT_CONNECTING, &xprt->sockstate);
530 smp_mb__after_clear_bit();
534 * Attempt to connect a TCP socket.
537 void xprt_connect(struct rpc_task *task)
539 struct rpc_xprt *xprt = task->tk_xprt;
541 dprintk("RPC: %4d xprt_connect xprt %p %s connected\n", task->tk_pid,
542 xprt, (xprt_connected(xprt) ? "is" : "is not"));
544 if (xprt->shutdown) {
545 task->tk_status = -EIO;
548 if (!xprt->addr.sin_port) {
549 task->tk_status = -EIO;
552 if (!xprt_lock_write(xprt, task))
554 if (xprt_connected(xprt))
558 task->tk_rqstp->rq_bytes_sent = 0;
560 task->tk_timeout = RPC_CONNECT_TIMEOUT;
561 rpc_sleep_on(&xprt->pending, task, xprt_connect_status, NULL);
562 if (!test_and_set_bit(XPRT_CONNECTING, &xprt->sockstate)) {
563 /* Note: if we are here due to a dropped connection
564 * we delay reconnecting by RPC_REESTABLISH_TIMEOUT/HZ
567 if (xprt->sock != NULL)
568 schedule_delayed_work(&xprt->sock_connect,
569 RPC_REESTABLISH_TIMEOUT);
571 schedule_work(&xprt->sock_connect);
572 if (!RPC_IS_ASYNC(task))
573 flush_scheduled_work();
578 xprt_release_write(xprt, task);
582 * We arrive here when awoken from waiting on connection establishment.
585 xprt_connect_status(struct rpc_task *task)
587 struct rpc_xprt *xprt = task->tk_xprt;
589 if (task->tk_status >= 0) {
590 dprintk("RPC: %4d xprt_connect_status: connection established\n",
595 /* if soft mounted, just cause this RPC to fail */
596 if (RPC_IS_SOFT(task))
597 task->tk_status = -EIO;
599 switch (task->tk_status) {
605 dprintk("RPC: %4d xprt_connect_status: timed out\n",
609 printk(KERN_ERR "RPC: error %d connecting to server %s\n",
610 -task->tk_status, task->tk_client->cl_server);
612 xprt_release_write(xprt, task);
616 * Look up the RPC request corresponding to a reply, and then lock it.
618 static inline struct rpc_rqst *
619 xprt_lookup_rqst(struct rpc_xprt *xprt, u32 xid)
621 struct list_head *pos;
622 struct rpc_rqst *req = NULL;
624 list_for_each(pos, &xprt->recv) {
625 struct rpc_rqst *entry = list_entry(pos, struct rpc_rqst, rq_list);
626 if (entry->rq_xid == xid) {
635 * Complete reply received.
636 * The TCP code relies on us to remove the request from xprt->pending.
639 xprt_complete_rqst(struct rpc_xprt *xprt, struct rpc_rqst *req, int copied)
641 struct rpc_task *task = req->rq_task;
642 struct rpc_clnt *clnt = task->tk_client;
644 /* Adjust congestion window */
646 unsigned timer = task->tk_msg.rpc_proc->p_timer;
647 xprt_adjust_cwnd(xprt, copied);
648 __xprt_put_cong(xprt, req);
650 if (req->rq_ntrans == 1)
651 rpc_update_rtt(clnt->cl_rtt, timer,
652 (long)jiffies - req->rq_xtime);
653 rpc_set_timeo(clnt->cl_rtt, timer, req->rq_ntrans - 1);
658 /* Profile only reads for now */
660 static unsigned long nextstat;
661 static unsigned long pkt_rtt, pkt_len, pkt_cnt;
664 pkt_len += req->rq_slen + copied;
665 pkt_rtt += jiffies - req->rq_xtime;
666 if (time_before(nextstat, jiffies)) {
667 printk("RPC: %lu %ld cwnd\n", jiffies, xprt->cwnd);
668 printk("RPC: %ld %ld %ld %ld stat\n",
669 jiffies, pkt_cnt, pkt_len, pkt_rtt);
670 pkt_rtt = pkt_len = pkt_cnt = 0;
671 nextstat = jiffies + 5 * HZ;
676 dprintk("RPC: %4d has input (%d bytes)\n", task->tk_pid, copied);
677 list_del_init(&req->rq_list);
678 req->rq_received = req->rq_private_buf.len = copied;
680 /* ... and wake up the process. */
681 rpc_wake_up_task(task);
686 skb_read_bits(skb_reader_t *desc, void *to, size_t len)
688 if (len > desc->count)
690 if (skb_copy_bits(desc->skb, desc->offset, to, len))
698 skb_read_and_csum_bits(skb_reader_t *desc, void *to, size_t len)
700 unsigned int csum2, pos;
702 if (len > desc->count)
705 csum2 = skb_copy_and_csum_bits(desc->skb, pos, to, len, 0);
706 desc->csum = csum_block_add(desc->csum, csum2, pos);
713 * We have set things up such that we perform the checksum of the UDP
714 * packet in parallel with the copies into the RPC client iovec. -DaveM
717 csum_partial_copy_to_xdr(struct xdr_buf *xdr, struct sk_buff *skb)
722 desc.offset = sizeof(struct udphdr);
723 desc.count = skb->len - desc.offset;
725 if (skb->ip_summed == CHECKSUM_UNNECESSARY)
728 desc.csum = csum_partial(skb->data, desc.offset, skb->csum);
729 if (xdr_partial_copy_from_skb(xdr, 0, &desc, skb_read_and_csum_bits) < 0)
731 if (desc.offset != skb->len) {
733 csum2 = skb_checksum(skb, desc.offset, skb->len - desc.offset, 0);
734 desc.csum = csum_block_add(desc.csum, csum2, desc.offset);
738 if ((unsigned short)csum_fold(desc.csum))
742 if (xdr_partial_copy_from_skb(xdr, 0, &desc, skb_read_bits) < 0)
750 * Input handler for RPC replies. Called from a bottom half and hence
754 udp_data_ready(struct sock *sk, int len)
756 struct rpc_task *task;
757 struct rpc_xprt *xprt;
758 struct rpc_rqst *rovr;
760 int err, repsize, copied;
763 read_lock(&sk->sk_callback_lock);
764 dprintk("RPC: udp_data_ready...\n");
765 if (!(xprt = xprt_from_sock(sk))) {
766 printk("RPC: udp_data_ready request not found!\n");
770 dprintk("RPC: udp_data_ready client %p\n", xprt);
772 if ((skb = skb_recv_datagram(sk, 0, 1, &err)) == NULL)
778 repsize = skb->len - sizeof(struct udphdr);
780 printk("RPC: impossible RPC reply size %d!\n", repsize);
784 /* Copy the XID from the skb... */
785 xp = skb_header_pointer(skb, sizeof(struct udphdr),
786 sizeof(_xid), &_xid);
790 /* Look up and lock the request corresponding to the given XID */
791 spin_lock(&xprt->sock_lock);
792 rovr = xprt_lookup_rqst(xprt, *xp);
795 task = rovr->rq_task;
797 dprintk("RPC: %4d received reply\n", task->tk_pid);
799 if ((copied = rovr->rq_private_buf.buflen) > repsize)
802 /* Suck it into the iovec, verify checksum if not done by hw. */
803 if (csum_partial_copy_to_xdr(&rovr->rq_private_buf, skb))
806 /* Something worked... */
807 dst_confirm(skb->dst);
809 xprt_complete_rqst(xprt, rovr, copied);
812 spin_unlock(&xprt->sock_lock);
814 skb_free_datagram(sk, skb);
816 read_unlock(&sk->sk_callback_lock);
820 * Copy from an skb into memory and shrink the skb.
823 tcp_copy_data(skb_reader_t *desc, void *p, size_t len)
825 if (len > desc->count)
827 if (skb_copy_bits(desc->skb, desc->offset, p, len)) {
828 dprintk("RPC: failed to copy %zu bytes from skb. %zu bytes remain\n",
834 dprintk("RPC: copied %zu bytes from skb. %zu bytes remain\n",
840 * TCP read fragment marker
843 tcp_read_fraghdr(struct rpc_xprt *xprt, skb_reader_t *desc)
848 p = ((char *) &xprt->tcp_recm) + xprt->tcp_offset;
849 len = sizeof(xprt->tcp_recm) - xprt->tcp_offset;
850 used = tcp_copy_data(desc, p, len);
851 xprt->tcp_offset += used;
854 xprt->tcp_reclen = ntohl(xprt->tcp_recm);
855 if (xprt->tcp_reclen & 0x80000000)
856 xprt->tcp_flags |= XPRT_LAST_FRAG;
858 xprt->tcp_flags &= ~XPRT_LAST_FRAG;
859 xprt->tcp_reclen &= 0x7fffffff;
860 xprt->tcp_flags &= ~XPRT_COPY_RECM;
861 xprt->tcp_offset = 0;
862 /* Sanity check of the record length */
863 if (xprt->tcp_reclen < 4) {
864 printk(KERN_ERR "RPC: Invalid TCP record fragment length\n");
865 xprt_disconnect(xprt);
867 dprintk("RPC: reading TCP record fragment of length %d\n",
872 tcp_check_recm(struct rpc_xprt *xprt)
874 dprintk("RPC: xprt = %p, tcp_copied = %lu, tcp_offset = %u, tcp_reclen = %u, tcp_flags = %lx\n",
875 xprt, xprt->tcp_copied, xprt->tcp_offset, xprt->tcp_reclen, xprt->tcp_flags);
876 if (xprt->tcp_offset == xprt->tcp_reclen) {
877 xprt->tcp_flags |= XPRT_COPY_RECM;
878 xprt->tcp_offset = 0;
879 if (xprt->tcp_flags & XPRT_LAST_FRAG) {
880 xprt->tcp_flags &= ~XPRT_COPY_DATA;
881 xprt->tcp_flags |= XPRT_COPY_XID;
882 xprt->tcp_copied = 0;
891 tcp_read_xid(struct rpc_xprt *xprt, skb_reader_t *desc)
896 len = sizeof(xprt->tcp_xid) - xprt->tcp_offset;
897 dprintk("RPC: reading XID (%Zu bytes)\n", len);
898 p = ((char *) &xprt->tcp_xid) + xprt->tcp_offset;
899 used = tcp_copy_data(desc, p, len);
900 xprt->tcp_offset += used;
903 xprt->tcp_flags &= ~XPRT_COPY_XID;
904 xprt->tcp_flags |= XPRT_COPY_DATA;
905 xprt->tcp_copied = 4;
906 dprintk("RPC: reading reply for XID %08x\n",
907 ntohl(xprt->tcp_xid));
908 tcp_check_recm(xprt);
912 * TCP read and complete request
915 tcp_read_request(struct rpc_xprt *xprt, skb_reader_t *desc)
917 struct rpc_rqst *req;
918 struct xdr_buf *rcvbuf;
922 /* Find and lock the request corresponding to this xid */
923 spin_lock(&xprt->sock_lock);
924 req = xprt_lookup_rqst(xprt, xprt->tcp_xid);
926 xprt->tcp_flags &= ~XPRT_COPY_DATA;
927 dprintk("RPC: XID %08x request not found!\n",
928 ntohl(xprt->tcp_xid));
929 spin_unlock(&xprt->sock_lock);
933 rcvbuf = &req->rq_private_buf;
935 if (len > xprt->tcp_reclen - xprt->tcp_offset) {
936 skb_reader_t my_desc;
938 len = xprt->tcp_reclen - xprt->tcp_offset;
939 memcpy(&my_desc, desc, sizeof(my_desc));
941 r = xdr_partial_copy_from_skb(rcvbuf, xprt->tcp_copied,
942 &my_desc, tcp_copy_data);
946 r = xdr_partial_copy_from_skb(rcvbuf, xprt->tcp_copied,
947 desc, tcp_copy_data);
950 xprt->tcp_copied += r;
951 xprt->tcp_offset += r;
954 /* Error when copying to the receive buffer,
955 * usually because we weren't able to allocate
956 * additional buffer pages. All we can do now
957 * is turn off XPRT_COPY_DATA, so the request
958 * will not receive any additional updates,
960 * Any remaining data from this record will
963 xprt->tcp_flags &= ~XPRT_COPY_DATA;
964 dprintk("RPC: XID %08x truncated request\n",
965 ntohl(xprt->tcp_xid));
966 dprintk("RPC: xprt = %p, tcp_copied = %lu, tcp_offset = %u, tcp_reclen = %u\n",
967 xprt, xprt->tcp_copied, xprt->tcp_offset, xprt->tcp_reclen);
971 dprintk("RPC: XID %08x read %Zd bytes\n",
972 ntohl(xprt->tcp_xid), r);
973 dprintk("RPC: xprt = %p, tcp_copied = %lu, tcp_offset = %u, tcp_reclen = %u\n",
974 xprt, xprt->tcp_copied, xprt->tcp_offset, xprt->tcp_reclen);
976 if (xprt->tcp_copied == req->rq_private_buf.buflen)
977 xprt->tcp_flags &= ~XPRT_COPY_DATA;
978 else if (xprt->tcp_offset == xprt->tcp_reclen) {
979 if (xprt->tcp_flags & XPRT_LAST_FRAG)
980 xprt->tcp_flags &= ~XPRT_COPY_DATA;
984 if (!(xprt->tcp_flags & XPRT_COPY_DATA)) {
985 dprintk("RPC: %4d received reply complete\n",
986 req->rq_task->tk_pid);
987 xprt_complete_rqst(xprt, req, xprt->tcp_copied);
989 spin_unlock(&xprt->sock_lock);
990 tcp_check_recm(xprt);
994 * TCP discard extra bytes from a short read
997 tcp_read_discard(struct rpc_xprt *xprt, skb_reader_t *desc)
1001 len = xprt->tcp_reclen - xprt->tcp_offset;
1002 if (len > desc->count)
1005 desc->offset += len;
1006 xprt->tcp_offset += len;
1007 dprintk("RPC: discarded %Zu bytes\n", len);
1008 tcp_check_recm(xprt);
1012 * TCP record receive routine
1013 * We first have to grab the record marker, then the XID, then the data.
1016 tcp_data_recv(read_descriptor_t *rd_desc, struct sk_buff *skb,
1017 unsigned int offset, size_t len)
1019 struct rpc_xprt *xprt = rd_desc->arg.data;
1020 skb_reader_t desc = {
1027 dprintk("RPC: tcp_data_recv\n");
1029 /* Read in a new fragment marker if necessary */
1030 /* Can we ever really expect to get completely empty fragments? */
1031 if (xprt->tcp_flags & XPRT_COPY_RECM) {
1032 tcp_read_fraghdr(xprt, &desc);
1035 /* Read in the xid if necessary */
1036 if (xprt->tcp_flags & XPRT_COPY_XID) {
1037 tcp_read_xid(xprt, &desc);
1040 /* Read in the request data */
1041 if (xprt->tcp_flags & XPRT_COPY_DATA) {
1042 tcp_read_request(xprt, &desc);
1045 /* Skip over any trailing bytes on short reads */
1046 tcp_read_discard(xprt, &desc);
1047 } while (desc.count);
1048 dprintk("RPC: tcp_data_recv done\n");
1049 return len - desc.count;
1052 static void tcp_data_ready(struct sock *sk, int bytes)
1054 struct rpc_xprt *xprt;
1055 read_descriptor_t rd_desc;
1057 read_lock(&sk->sk_callback_lock);
1058 dprintk("RPC: tcp_data_ready...\n");
1059 if (!(xprt = xprt_from_sock(sk))) {
1060 printk("RPC: tcp_data_ready socket info not found!\n");
1066 /* We use rd_desc to pass struct xprt to tcp_data_recv */
1067 rd_desc.arg.data = xprt;
1068 rd_desc.count = 65536;
1069 tcp_read_sock(sk, &rd_desc, tcp_data_recv);
1071 read_unlock(&sk->sk_callback_lock);
1075 tcp_state_change(struct sock *sk)
1077 struct rpc_xprt *xprt;
1079 read_lock(&sk->sk_callback_lock);
1080 if (!(xprt = xprt_from_sock(sk)))
1082 dprintk("RPC: tcp_state_change client %p...\n", xprt);
1083 dprintk("RPC: state %x conn %d dead %d zapped %d\n",
1084 sk->sk_state, xprt_connected(xprt),
1085 sock_flag(sk, SOCK_DEAD),
1086 sock_flag(sk, SOCK_ZAPPED));
1088 switch (sk->sk_state) {
1089 case TCP_ESTABLISHED:
1090 spin_lock_bh(&xprt->sock_lock);
1091 if (!xprt_test_and_set_connected(xprt)) {
1092 /* Reset TCP record info */
1093 xprt->tcp_offset = 0;
1094 xprt->tcp_reclen = 0;
1095 xprt->tcp_copied = 0;
1096 xprt->tcp_flags = XPRT_COPY_RECM | XPRT_COPY_XID;
1097 rpc_wake_up(&xprt->pending);
1099 spin_unlock_bh(&xprt->sock_lock);
1105 xprt_disconnect(xprt);
1109 read_unlock(&sk->sk_callback_lock);
1113 * Called when more output buffer space is available for this socket.
1114 * We try not to wake our writers until they can make "significant"
1115 * progress, otherwise we'll waste resources thrashing sock_sendmsg
1116 * with a bunch of small requests.
1119 xprt_write_space(struct sock *sk)
1121 struct rpc_xprt *xprt;
1122 struct socket *sock;
1124 read_lock(&sk->sk_callback_lock);
1125 if (!(xprt = xprt_from_sock(sk)) || !(sock = sk->sk_socket))
1130 /* Wait until we have enough socket memory */
1132 /* from net/core/stream.c:sk_stream_write_space */
1133 if (sk_stream_wspace(sk) < sk_stream_min_wspace(sk))
1136 /* from net/core/sock.c:sock_def_write_space */
1137 if (!sock_writeable(sk))
1141 if (!test_and_clear_bit(SOCK_NOSPACE, &sock->flags))
1144 spin_lock_bh(&xprt->sock_lock);
1146 rpc_wake_up_task(xprt->snd_task);
1147 spin_unlock_bh(&xprt->sock_lock);
1149 read_unlock(&sk->sk_callback_lock);
1153 * RPC receive timeout handler.
1156 xprt_timer(struct rpc_task *task)
1158 struct rpc_rqst *req = task->tk_rqstp;
1159 struct rpc_xprt *xprt = req->rq_xprt;
1161 spin_lock(&xprt->sock_lock);
1162 if (req->rq_received)
1165 xprt_adjust_cwnd(req->rq_xprt, -ETIMEDOUT);
1166 __xprt_put_cong(xprt, req);
1168 dprintk("RPC: %4d xprt_timer (%s request)\n",
1169 task->tk_pid, req ? "pending" : "backlogged");
1171 task->tk_status = -ETIMEDOUT;
1173 task->tk_timeout = 0;
1174 rpc_wake_up_task(task);
1175 spin_unlock(&xprt->sock_lock);
1179 * Place the actual RPC call.
1180 * We have to copy the iovec because sendmsg fiddles with its contents.
1183 xprt_prepare_transmit(struct rpc_task *task)
1185 struct rpc_rqst *req = task->tk_rqstp;
1186 struct rpc_xprt *xprt = req->rq_xprt;
1189 dprintk("RPC: %4d xprt_prepare_transmit\n", task->tk_pid);
1194 spin_lock_bh(&xprt->sock_lock);
1195 if (req->rq_received && !req->rq_bytes_sent) {
1196 err = req->rq_received;
1199 if (!__xprt_lock_write(xprt, task)) {
1204 if (!xprt_connected(xprt)) {
1209 spin_unlock_bh(&xprt->sock_lock);
1214 xprt_transmit(struct rpc_task *task)
1216 struct rpc_clnt *clnt = task->tk_client;
1217 struct rpc_rqst *req = task->tk_rqstp;
1218 struct rpc_xprt *xprt = req->rq_xprt;
1219 int status, retry = 0;
1222 dprintk("RPC: %4d xprt_transmit(%u)\n", task->tk_pid, req->rq_slen);
1224 /* set up everything as needed. */
1225 /* Write the record marker */
1227 u32 *marker = req->rq_svec[0].iov_base;
1229 *marker = htonl(0x80000000|(req->rq_slen-sizeof(*marker)));
1233 if (!req->rq_received) {
1234 if (list_empty(&req->rq_list)) {
1235 spin_lock_bh(&xprt->sock_lock);
1236 /* Update the softirq receive buffer */
1237 memcpy(&req->rq_private_buf, &req->rq_rcv_buf,
1238 sizeof(req->rq_private_buf));
1239 /* Add request to the receive list */
1240 list_add_tail(&req->rq_list, &xprt->recv);
1241 spin_unlock_bh(&xprt->sock_lock);
1242 xprt_reset_majortimeo(req);
1243 /* Turn off autodisconnect */
1244 del_singleshot_timer_sync(&xprt->timer);
1246 } else if (!req->rq_bytes_sent)
1249 /* Continue transmitting the packet/record. We must be careful
1250 * to cope with writespace callbacks arriving _after_ we have
1251 * called xprt_sendmsg().
1254 req->rq_xtime = jiffies;
1255 status = xprt_sendmsg(xprt, req);
1261 req->rq_bytes_sent += status;
1263 /* If we've sent the entire packet, immediately
1264 * reset the count of bytes sent. */
1265 if (req->rq_bytes_sent >= req->rq_slen) {
1266 req->rq_bytes_sent = 0;
1270 if (status >= req->rq_slen)
1276 dprintk("RPC: %4d xmit incomplete (%d left of %d)\n",
1277 task->tk_pid, req->rq_slen - req->rq_bytes_sent,
1285 /* Note: at this point, task->tk_sleeping has not yet been set,
1286 * hence there is no danger of the waking up task being put on
1287 * schedq, and being picked up by a parallel run of rpciod().
1289 task->tk_status = status;
1293 if (test_bit(SOCK_ASYNC_NOSPACE, &xprt->sock->flags)) {
1294 /* Protect against races with xprt_write_space */
1295 spin_lock_bh(&xprt->sock_lock);
1296 /* Don't race with disconnect */
1297 if (!xprt_connected(xprt))
1298 task->tk_status = -ENOTCONN;
1299 else if (test_bit(SOCK_NOSPACE, &xprt->sock->flags)) {
1300 task->tk_timeout = req->rq_timeout;
1301 rpc_sleep_on(&xprt->pending, task, NULL, NULL);
1303 spin_unlock_bh(&xprt->sock_lock);
1306 /* Keep holding the socket if it is blocked */
1307 rpc_delay(task, HZ>>4);
1310 task->tk_timeout = RPC_REESTABLISH_TIMEOUT;
1311 rpc_sleep_on(&xprt->sending, task, NULL, NULL);
1316 xprt_disconnect(xprt);
1318 xprt_release_write(xprt, task);
1321 dprintk("RPC: %4d xmit complete\n", task->tk_pid);
1322 /* Set the task's receive timeout value */
1323 spin_lock_bh(&xprt->sock_lock);
1324 if (!xprt->nocong) {
1325 int timer = task->tk_msg.rpc_proc->p_timer;
1326 task->tk_timeout = rpc_calc_rto(clnt->cl_rtt, timer);
1327 task->tk_timeout <<= rpc_ntimeo(clnt->cl_rtt, timer) + req->rq_retries;
1328 if (task->tk_timeout > xprt->timeout.to_maxval || task->tk_timeout == 0)
1329 task->tk_timeout = xprt->timeout.to_maxval;
1331 task->tk_timeout = req->rq_timeout;
1332 /* Don't race with disconnect */
1333 if (!xprt_connected(xprt))
1334 task->tk_status = -ENOTCONN;
1335 else if (!req->rq_received)
1336 rpc_sleep_on(&xprt->pending, task, NULL, xprt_timer);
1337 __xprt_release_write(xprt, task);
1338 spin_unlock_bh(&xprt->sock_lock);
1342 * Reserve an RPC call slot.
1345 do_xprt_reserve(struct rpc_task *task)
1347 struct rpc_xprt *xprt = task->tk_xprt;
1349 task->tk_status = 0;
1352 if (!list_empty(&xprt->free)) {
1353 struct rpc_rqst *req = list_entry(xprt->free.next, struct rpc_rqst, rq_list);
1354 list_del_init(&req->rq_list);
1355 task->tk_rqstp = req;
1356 xprt_request_init(task, xprt);
1359 dprintk("RPC: waiting for request slot\n");
1360 task->tk_status = -EAGAIN;
1361 task->tk_timeout = 0;
1362 rpc_sleep_on(&xprt->backlog, task, NULL, NULL);
1366 xprt_reserve(struct rpc_task *task)
1368 struct rpc_xprt *xprt = task->tk_xprt;
1370 task->tk_status = -EIO;
1371 if (!xprt->shutdown) {
1372 spin_lock(&xprt->xprt_lock);
1373 do_xprt_reserve(task);
1374 spin_unlock(&xprt->xprt_lock);
1379 * Allocate a 'unique' XID
1381 static inline u32 xprt_alloc_xid(struct rpc_xprt *xprt)
1386 static inline void xprt_init_xid(struct rpc_xprt *xprt)
1388 get_random_bytes(&xprt->xid, sizeof(xprt->xid));
1392 * Initialize RPC request
1395 xprt_request_init(struct rpc_task *task, struct rpc_xprt *xprt)
1397 struct rpc_rqst *req = task->tk_rqstp;
1399 req->rq_timeout = xprt->timeout.to_initval;
1400 req->rq_task = task;
1401 req->rq_xprt = xprt;
1402 req->rq_xid = xprt_alloc_xid(xprt);
1403 dprintk("RPC: %4d reserved req %p xid %08x\n", task->tk_pid,
1404 req, ntohl(req->rq_xid));
1408 * Release an RPC call slot
1411 xprt_release(struct rpc_task *task)
1413 struct rpc_xprt *xprt = task->tk_xprt;
1414 struct rpc_rqst *req;
1416 if (!(req = task->tk_rqstp))
1418 spin_lock_bh(&xprt->sock_lock);
1419 __xprt_release_write(xprt, task);
1420 __xprt_put_cong(xprt, req);
1421 if (!list_empty(&req->rq_list))
1422 list_del(&req->rq_list);
1423 xprt->last_used = jiffies;
1424 if (list_empty(&xprt->recv) && !xprt->shutdown)
1425 mod_timer(&xprt->timer, xprt->last_used + XPRT_IDLE_TIMEOUT);
1426 spin_unlock_bh(&xprt->sock_lock);
1427 task->tk_rqstp = NULL;
1428 memset(req, 0, sizeof(*req)); /* mark unused */
1430 dprintk("RPC: %4d release request %p\n", task->tk_pid, req);
1432 spin_lock(&xprt->xprt_lock);
1433 list_add(&req->rq_list, &xprt->free);
1434 xprt_clear_backlog(xprt);
1435 spin_unlock(&xprt->xprt_lock);
1439 * Set default timeout parameters
1442 xprt_default_timeout(struct rpc_timeout *to, int proto)
1444 if (proto == IPPROTO_UDP)
1445 xprt_set_timeout(to, 5, 5 * HZ);
1447 xprt_set_timeout(to, 5, 60 * HZ);
1451 * Set constant timeout
1454 xprt_set_timeout(struct rpc_timeout *to, unsigned int retr, unsigned long incr)
1457 to->to_increment = incr;
1458 to->to_maxval = incr * retr;
1459 to->to_retries = retr;
1460 to->to_exponential = 0;
1463 unsigned int xprt_udp_slot_table_entries = RPC_DEF_SLOT_TABLE;
1464 unsigned int xprt_tcp_slot_table_entries = RPC_DEF_SLOT_TABLE;
1467 * Initialize an RPC client
1469 static struct rpc_xprt *
1470 xprt_setup(int proto, struct sockaddr_in *ap, struct rpc_timeout *to)
1472 struct rpc_xprt *xprt;
1473 unsigned int entries;
1474 size_t slot_table_size;
1475 struct rpc_rqst *req;
1477 dprintk("RPC: setting up %s transport...\n",
1478 proto == IPPROTO_UDP? "UDP" : "TCP");
1480 entries = (proto == IPPROTO_TCP)?
1481 xprt_tcp_slot_table_entries : xprt_udp_slot_table_entries;
1483 if ((xprt = kmalloc(sizeof(struct rpc_xprt), GFP_KERNEL)) == NULL)
1484 return ERR_PTR(-ENOMEM);
1485 memset(xprt, 0, sizeof(*xprt)); /* Nnnngh! */
1486 xprt->max_reqs = entries;
1487 slot_table_size = entries * sizeof(xprt->slot[0]);
1488 xprt->slot = kmalloc(slot_table_size, GFP_KERNEL);
1489 if (xprt->slot == NULL) {
1491 return ERR_PTR(-ENOMEM);
1493 memset(xprt->slot, 0, slot_table_size);
1497 xprt->stream = (proto == IPPROTO_TCP)? 1 : 0;
1499 xprt->cwnd = RPC_MAXCWND(xprt);
1501 xprt->max_payload = (1U << 31) - 1;
1503 xprt->cwnd = RPC_INITCWND;
1504 xprt->max_payload = (1U << 16) - (MAX_HEADER << 3);
1506 spin_lock_init(&xprt->sock_lock);
1507 spin_lock_init(&xprt->xprt_lock);
1508 init_waitqueue_head(&xprt->cong_wait);
1510 INIT_LIST_HEAD(&xprt->free);
1511 INIT_LIST_HEAD(&xprt->recv);
1512 INIT_WORK(&xprt->sock_connect, xprt_socket_connect, xprt);
1513 INIT_WORK(&xprt->task_cleanup, xprt_socket_autoclose, xprt);
1514 init_timer(&xprt->timer);
1515 xprt->timer.function = xprt_init_autodisconnect;
1516 xprt->timer.data = (unsigned long) xprt;
1517 xprt->last_used = jiffies;
1518 xprt->port = XPRT_MAX_RESVPORT;
1520 /* Set timeout parameters */
1522 xprt->timeout = *to;
1524 xprt_default_timeout(&xprt->timeout, xprt->prot);
1526 rpc_init_wait_queue(&xprt->pending, "xprt_pending");
1527 rpc_init_wait_queue(&xprt->sending, "xprt_sending");
1528 rpc_init_wait_queue(&xprt->resend, "xprt_resend");
1529 rpc_init_priority_wait_queue(&xprt->backlog, "xprt_backlog");
1531 /* initialize free list */
1532 for (req = &xprt->slot[entries-1]; req >= &xprt->slot[0]; req--)
1533 list_add(&req->rq_list, &xprt->free);
1535 xprt_init_xid(xprt);
1537 /* Check whether we want to use a reserved port */
1538 xprt->resvport = capable(CAP_NET_BIND_SERVICE) ? 1 : 0;
1540 dprintk("RPC: created transport %p with %u slots\n", xprt,
1547 * Bind to a reserved port
1549 static inline int xprt_bindresvport(struct rpc_xprt *xprt, struct socket *sock)
1551 struct sockaddr_in myaddr = {
1552 .sin_family = AF_INET,
1556 /* Were we already bound to a given port? Try to reuse it */
1559 myaddr.sin_port = htons(port);
1560 err = sock->ops->bind(sock, (struct sockaddr *) &myaddr,
1567 port = XPRT_MAX_RESVPORT;
1568 } while (err == -EADDRINUSE && port != xprt->port);
1570 printk("RPC: Can't bind to reserved port (%d).\n", -err);
1575 xprt_bind_socket(struct rpc_xprt *xprt, struct socket *sock)
1577 struct sock *sk = sock->sk;
1582 write_lock_bh(&sk->sk_callback_lock);
1583 sk->sk_user_data = xprt;
1584 xprt->old_data_ready = sk->sk_data_ready;
1585 xprt->old_state_change = sk->sk_state_change;
1586 xprt->old_write_space = sk->sk_write_space;
1587 if (xprt->prot == IPPROTO_UDP) {
1588 sk->sk_data_ready = udp_data_ready;
1589 sk->sk_no_check = UDP_CSUM_NORCV;
1590 xprt_set_connected(xprt);
1592 tcp_sk(sk)->nonagle = 1; /* disable Nagle's algorithm */
1593 sk->sk_data_ready = tcp_data_ready;
1594 sk->sk_state_change = tcp_state_change;
1595 xprt_clear_connected(xprt);
1597 sk->sk_write_space = xprt_write_space;
1599 /* Reset to new socket */
1602 write_unlock_bh(&sk->sk_callback_lock);
1608 * Set socket buffer length
1611 xprt_sock_setbufsize(struct rpc_xprt *xprt)
1613 struct sock *sk = xprt->inet;
1617 if (xprt->rcvsize) {
1618 sk->sk_userlocks |= SOCK_RCVBUF_LOCK;
1619 sk->sk_rcvbuf = xprt->rcvsize * xprt->max_reqs * 2;
1621 if (xprt->sndsize) {
1622 sk->sk_userlocks |= SOCK_SNDBUF_LOCK;
1623 sk->sk_sndbuf = xprt->sndsize * xprt->max_reqs * 2;
1624 sk->sk_write_space(sk);
1629 * Datastream sockets are created here, but xprt_connect will create
1630 * and connect stream sockets.
1632 static struct socket * xprt_create_socket(struct rpc_xprt *xprt, int proto, int resvport)
1634 struct socket *sock;
1637 dprintk("RPC: xprt_create_socket(%s %d)\n",
1638 (proto == IPPROTO_UDP)? "udp" : "tcp", proto);
1640 type = (proto == IPPROTO_UDP)? SOCK_DGRAM : SOCK_STREAM;
1642 if ((err = sock_create_kern(PF_INET, type, proto, &sock)) < 0) {
1643 printk("RPC: can't create socket (%d).\n", -err);
1647 /* If the caller has the capability, bind to a reserved port */
1648 if (resvport && xprt_bindresvport(xprt, sock) < 0) {
1649 printk("RPC: can't bind to reserved port.\n");
1661 * Create an RPC client transport given the protocol and peer address.
1664 xprt_create_proto(int proto, struct sockaddr_in *sap, struct rpc_timeout *to)
1666 struct rpc_xprt *xprt;
1668 xprt = xprt_setup(proto, sap, to);
1670 dprintk("RPC: xprt_create_proto failed\n");
1672 dprintk("RPC: xprt_create_proto created xprt %p\n", xprt);
1677 * Prepare for transport shutdown.
1680 xprt_shutdown(struct rpc_xprt *xprt)
1683 rpc_wake_up(&xprt->sending);
1684 rpc_wake_up(&xprt->resend);
1685 rpc_wake_up(&xprt->pending);
1686 rpc_wake_up(&xprt->backlog);
1687 wake_up(&xprt->cong_wait);
1688 del_timer_sync(&xprt->timer);
1690 /* synchronously wait for connect worker to finish */
1691 cancel_delayed_work(&xprt->sock_connect);
1692 flush_scheduled_work();
1696 * Clear the xprt backlog queue
1699 xprt_clear_backlog(struct rpc_xprt *xprt) {
1700 rpc_wake_up_next(&xprt->backlog);
1701 wake_up(&xprt->cong_wait);
1706 * Destroy an RPC transport, killing off all requests.
1709 xprt_destroy(struct rpc_xprt *xprt)
1711 dprintk("RPC: destroying transport %p\n", xprt);
1712 xprt_shutdown(xprt);
1713 xprt_disconnect(xprt);