xref: /linux/net/rxrpc/recvmsg.c (revision 8fc4e4aa2bfca8d32e8bc2a01526ea2da450e6cb)
1 // SPDX-License-Identifier: GPL-2.0-or-later
2 /* RxRPC recvmsg() implementation
3  *
4  * Copyright (C) 2007 Red Hat, Inc. All Rights Reserved.
5  * Written by David Howells (dhowells@redhat.com)
6  */
7 
8 #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
9 
10 #include <linux/net.h>
11 #include <linux/skbuff.h>
12 #include <linux/export.h>
13 #include <linux/sched/signal.h>
14 
15 #include <net/sock.h>
16 #include <net/af_rxrpc.h>
17 #include "ar-internal.h"
18 
19 /*
20  * Post a call for attention by the socket or kernel service.  Further
21  * notifications are suppressed by putting recvmsg_link on a dummy queue.
22  */
23 void rxrpc_notify_socket(struct rxrpc_call *call)
24 {
25 	struct rxrpc_sock *rx;
26 	struct sock *sk;
27 
28 	_enter("%d", call->debug_id);
29 
30 	if (!list_empty(&call->recvmsg_link))
31 		return;
32 
33 	rcu_read_lock();
34 
35 	rx = rcu_dereference(call->socket);
36 	sk = &rx->sk;
37 	if (rx && sk->sk_state < RXRPC_CLOSE) {
38 		if (call->notify_rx) {
39 			spin_lock_bh(&call->notify_lock);
40 			call->notify_rx(sk, call, call->user_call_ID);
41 			spin_unlock_bh(&call->notify_lock);
42 		} else {
43 			write_lock_bh(&rx->recvmsg_lock);
44 			if (list_empty(&call->recvmsg_link)) {
45 				rxrpc_get_call(call, rxrpc_call_got);
46 				list_add_tail(&call->recvmsg_link, &rx->recvmsg_q);
47 			}
48 			write_unlock_bh(&rx->recvmsg_lock);
49 
50 			if (!sock_flag(sk, SOCK_DEAD)) {
51 				_debug("call %ps", sk->sk_data_ready);
52 				sk->sk_data_ready(sk);
53 			}
54 		}
55 	}
56 
57 	rcu_read_unlock();
58 	_leave("");
59 }
60 
61 /*
62  * Transition a call to the complete state.
63  */
64 bool __rxrpc_set_call_completion(struct rxrpc_call *call,
65 				 enum rxrpc_call_completion compl,
66 				 u32 abort_code,
67 				 int error)
68 {
69 	if (call->state < RXRPC_CALL_COMPLETE) {
70 		call->abort_code = abort_code;
71 		call->error = error;
72 		call->completion = compl;
73 		call->state = RXRPC_CALL_COMPLETE;
74 		trace_rxrpc_call_complete(call);
75 		wake_up(&call->waitq);
76 		rxrpc_notify_socket(call);
77 		return true;
78 	}
79 	return false;
80 }
81 
82 bool rxrpc_set_call_completion(struct rxrpc_call *call,
83 			       enum rxrpc_call_completion compl,
84 			       u32 abort_code,
85 			       int error)
86 {
87 	bool ret = false;
88 
89 	if (call->state < RXRPC_CALL_COMPLETE) {
90 		write_lock_bh(&call->state_lock);
91 		ret = __rxrpc_set_call_completion(call, compl, abort_code, error);
92 		write_unlock_bh(&call->state_lock);
93 	}
94 	return ret;
95 }
96 
97 /*
98  * Record that a call successfully completed.
99  */
100 bool __rxrpc_call_completed(struct rxrpc_call *call)
101 {
102 	return __rxrpc_set_call_completion(call, RXRPC_CALL_SUCCEEDED, 0, 0);
103 }
104 
105 bool rxrpc_call_completed(struct rxrpc_call *call)
106 {
107 	bool ret = false;
108 
109 	if (call->state < RXRPC_CALL_COMPLETE) {
110 		write_lock_bh(&call->state_lock);
111 		ret = __rxrpc_call_completed(call);
112 		write_unlock_bh(&call->state_lock);
113 	}
114 	return ret;
115 }
116 
117 /*
118  * Record that a call is locally aborted.
119  */
120 bool __rxrpc_abort_call(const char *why, struct rxrpc_call *call,
121 			rxrpc_seq_t seq, u32 abort_code, int error)
122 {
123 	trace_rxrpc_abort(call->debug_id, why, call->cid, call->call_id, seq,
124 			  abort_code, error);
125 	return __rxrpc_set_call_completion(call, RXRPC_CALL_LOCALLY_ABORTED,
126 					   abort_code, error);
127 }
128 
129 bool rxrpc_abort_call(const char *why, struct rxrpc_call *call,
130 		      rxrpc_seq_t seq, u32 abort_code, int error)
131 {
132 	bool ret;
133 
134 	write_lock_bh(&call->state_lock);
135 	ret = __rxrpc_abort_call(why, call, seq, abort_code, error);
136 	write_unlock_bh(&call->state_lock);
137 	return ret;
138 }
139 
140 /*
141  * Pass a call terminating message to userspace.
142  */
143 static int rxrpc_recvmsg_term(struct rxrpc_call *call, struct msghdr *msg)
144 {
145 	u32 tmp = 0;
146 	int ret;
147 
148 	switch (call->completion) {
149 	case RXRPC_CALL_SUCCEEDED:
150 		ret = 0;
151 		if (rxrpc_is_service_call(call))
152 			ret = put_cmsg(msg, SOL_RXRPC, RXRPC_ACK, 0, &tmp);
153 		break;
154 	case RXRPC_CALL_REMOTELY_ABORTED:
155 		tmp = call->abort_code;
156 		ret = put_cmsg(msg, SOL_RXRPC, RXRPC_ABORT, 4, &tmp);
157 		break;
158 	case RXRPC_CALL_LOCALLY_ABORTED:
159 		tmp = call->abort_code;
160 		ret = put_cmsg(msg, SOL_RXRPC, RXRPC_ABORT, 4, &tmp);
161 		break;
162 	case RXRPC_CALL_NETWORK_ERROR:
163 		tmp = -call->error;
164 		ret = put_cmsg(msg, SOL_RXRPC, RXRPC_NET_ERROR, 4, &tmp);
165 		break;
166 	case RXRPC_CALL_LOCAL_ERROR:
167 		tmp = -call->error;
168 		ret = put_cmsg(msg, SOL_RXRPC, RXRPC_LOCAL_ERROR, 4, &tmp);
169 		break;
170 	default:
171 		pr_err("Invalid terminal call state %u\n", call->state);
172 		BUG();
173 		break;
174 	}
175 
176 	trace_rxrpc_recvmsg(call, rxrpc_recvmsg_terminal, call->rx_hard_ack,
177 			    call->rx_pkt_offset, call->rx_pkt_len, ret);
178 	return ret;
179 }
180 
181 /*
182  * End the packet reception phase.
183  */
184 static void rxrpc_end_rx_phase(struct rxrpc_call *call, rxrpc_serial_t serial)
185 {
186 	_enter("%d,%s", call->debug_id, rxrpc_call_states[call->state]);
187 
188 	trace_rxrpc_receive(call, rxrpc_receive_end, 0, call->rx_top);
189 	ASSERTCMP(call->rx_hard_ack, ==, call->rx_top);
190 
191 	if (call->state == RXRPC_CALL_CLIENT_RECV_REPLY) {
192 		rxrpc_propose_ACK(call, RXRPC_ACK_IDLE, serial, false, true,
193 				  rxrpc_propose_ack_terminal_ack);
194 		//rxrpc_send_ack_packet(call, false, NULL);
195 	}
196 
197 	write_lock_bh(&call->state_lock);
198 
199 	switch (call->state) {
200 	case RXRPC_CALL_CLIENT_RECV_REPLY:
201 		__rxrpc_call_completed(call);
202 		write_unlock_bh(&call->state_lock);
203 		break;
204 
205 	case RXRPC_CALL_SERVER_RECV_REQUEST:
206 		call->tx_phase = true;
207 		call->state = RXRPC_CALL_SERVER_ACK_REQUEST;
208 		call->expect_req_by = jiffies + MAX_JIFFY_OFFSET;
209 		write_unlock_bh(&call->state_lock);
210 		rxrpc_propose_ACK(call, RXRPC_ACK_DELAY, serial, false, true,
211 				  rxrpc_propose_ack_processing_op);
212 		break;
213 	default:
214 		write_unlock_bh(&call->state_lock);
215 		break;
216 	}
217 }
218 
219 /*
220  * Discard a packet we've used up and advance the Rx window by one.
221  */
222 static void rxrpc_rotate_rx_window(struct rxrpc_call *call)
223 {
224 	struct rxrpc_skb_priv *sp;
225 	struct sk_buff *skb;
226 	rxrpc_serial_t serial;
227 	rxrpc_seq_t hard_ack, top;
228 	bool last = false;
229 	u8 subpacket;
230 	int ix;
231 
232 	_enter("%d", call->debug_id);
233 
234 	hard_ack = call->rx_hard_ack;
235 	top = smp_load_acquire(&call->rx_top);
236 	ASSERT(before(hard_ack, top));
237 
238 	hard_ack++;
239 	ix = hard_ack & RXRPC_RXTX_BUFF_MASK;
240 	skb = call->rxtx_buffer[ix];
241 	rxrpc_see_skb(skb, rxrpc_skb_rotated);
242 	sp = rxrpc_skb(skb);
243 
244 	subpacket = call->rxtx_annotations[ix] & RXRPC_RX_ANNO_SUBPACKET;
245 	serial = sp->hdr.serial + subpacket;
246 
247 	if (subpacket == sp->nr_subpackets - 1 &&
248 	    sp->rx_flags & RXRPC_SKB_INCL_LAST)
249 		last = true;
250 
251 	call->rxtx_buffer[ix] = NULL;
252 	call->rxtx_annotations[ix] = 0;
253 	/* Barrier against rxrpc_input_data(). */
254 	smp_store_release(&call->rx_hard_ack, hard_ack);
255 
256 	rxrpc_free_skb(skb, rxrpc_skb_freed);
257 
258 	trace_rxrpc_receive(call, rxrpc_receive_rotate, serial, hard_ack);
259 	if (last) {
260 		rxrpc_end_rx_phase(call, serial);
261 	} else {
262 		/* Check to see if there's an ACK that needs sending. */
263 		if (after_eq(hard_ack, call->ackr_consumed + 2) ||
264 		    after_eq(top, call->ackr_seen + 2) ||
265 		    (hard_ack == top && after(hard_ack, call->ackr_consumed)))
266 			rxrpc_propose_ACK(call, RXRPC_ACK_DELAY, serial,
267 					  true, true,
268 					  rxrpc_propose_ack_rotate_rx);
269 		if (call->ackr_reason && call->ackr_reason != RXRPC_ACK_DELAY)
270 			rxrpc_send_ack_packet(call, false, NULL);
271 	}
272 }
273 
274 /*
275  * Decrypt and verify a (sub)packet.  The packet's length may be changed due to
276  * padding, but if this is the case, the packet length will be resident in the
277  * socket buffer.  Note that we can't modify the master skb info as the skb may
278  * be the home to multiple subpackets.
279  */
280 static int rxrpc_verify_packet(struct rxrpc_call *call, struct sk_buff *skb,
281 			       u8 annotation,
282 			       unsigned int offset, unsigned int len)
283 {
284 	struct rxrpc_skb_priv *sp = rxrpc_skb(skb);
285 	rxrpc_seq_t seq = sp->hdr.seq;
286 	u16 cksum = sp->hdr.cksum;
287 	u8 subpacket = annotation & RXRPC_RX_ANNO_SUBPACKET;
288 
289 	_enter("");
290 
291 	/* For all but the head jumbo subpacket, the security checksum is in a
292 	 * jumbo header immediately prior to the data.
293 	 */
294 	if (subpacket > 0) {
295 		__be16 tmp;
296 		if (skb_copy_bits(skb, offset - 2, &tmp, 2) < 0)
297 			BUG();
298 		cksum = ntohs(tmp);
299 		seq += subpacket;
300 	}
301 
302 	return call->security->verify_packet(call, skb, offset, len,
303 					     seq, cksum);
304 }
305 
306 /*
307  * Locate the data within a packet.  This is complicated by:
308  *
309  * (1) An skb may contain a jumbo packet - so we have to find the appropriate
310  *     subpacket.
311  *
312  * (2) The (sub)packets may be encrypted and, if so, the encrypted portion
313  *     contains an extra header which includes the true length of the data,
314  *     excluding any encrypted padding.
315  */
316 static int rxrpc_locate_data(struct rxrpc_call *call, struct sk_buff *skb,
317 			     u8 *_annotation,
318 			     unsigned int *_offset, unsigned int *_len,
319 			     bool *_last)
320 {
321 	struct rxrpc_skb_priv *sp = rxrpc_skb(skb);
322 	unsigned int offset = sizeof(struct rxrpc_wire_header);
323 	unsigned int len;
324 	bool last = false;
325 	int ret;
326 	u8 annotation = *_annotation;
327 	u8 subpacket = annotation & RXRPC_RX_ANNO_SUBPACKET;
328 
329 	/* Locate the subpacket */
330 	offset += subpacket * RXRPC_JUMBO_SUBPKTLEN;
331 	len = skb->len - offset;
332 	if (subpacket < sp->nr_subpackets - 1)
333 		len = RXRPC_JUMBO_DATALEN;
334 	else if (sp->rx_flags & RXRPC_SKB_INCL_LAST)
335 		last = true;
336 
337 	if (!(annotation & RXRPC_RX_ANNO_VERIFIED)) {
338 		ret = rxrpc_verify_packet(call, skb, annotation, offset, len);
339 		if (ret < 0)
340 			return ret;
341 		*_annotation |= RXRPC_RX_ANNO_VERIFIED;
342 	}
343 
344 	*_offset = offset;
345 	*_len = len;
346 	*_last = last;
347 	call->security->locate_data(call, skb, _offset, _len);
348 	return 0;
349 }
350 
351 /*
352  * Deliver messages to a call.  This keeps processing packets until the buffer
353  * is filled and we find either more DATA (returns 0) or the end of the DATA
354  * (returns 1).  If more packets are required, it returns -EAGAIN.
355  */
356 static int rxrpc_recvmsg_data(struct socket *sock, struct rxrpc_call *call,
357 			      struct msghdr *msg, struct iov_iter *iter,
358 			      size_t len, int flags, size_t *_offset)
359 {
360 	struct rxrpc_skb_priv *sp;
361 	struct sk_buff *skb;
362 	rxrpc_serial_t serial;
363 	rxrpc_seq_t hard_ack, top, seq;
364 	size_t remain;
365 	bool rx_pkt_last;
366 	unsigned int rx_pkt_offset, rx_pkt_len;
367 	int ix, copy, ret = -EAGAIN, ret2;
368 
369 	if (test_and_clear_bit(RXRPC_CALL_RX_UNDERRUN, &call->flags) &&
370 	    call->ackr_reason)
371 		rxrpc_send_ack_packet(call, false, NULL);
372 
373 	rx_pkt_offset = call->rx_pkt_offset;
374 	rx_pkt_len = call->rx_pkt_len;
375 	rx_pkt_last = call->rx_pkt_last;
376 
377 	if (call->state >= RXRPC_CALL_SERVER_ACK_REQUEST) {
378 		seq = call->rx_hard_ack;
379 		ret = 1;
380 		goto done;
381 	}
382 
383 	/* Barriers against rxrpc_input_data(). */
384 	hard_ack = call->rx_hard_ack;
385 	seq = hard_ack + 1;
386 
387 	while (top = smp_load_acquire(&call->rx_top),
388 	       before_eq(seq, top)
389 	       ) {
390 		ix = seq & RXRPC_RXTX_BUFF_MASK;
391 		skb = call->rxtx_buffer[ix];
392 		if (!skb) {
393 			trace_rxrpc_recvmsg(call, rxrpc_recvmsg_hole, seq,
394 					    rx_pkt_offset, rx_pkt_len, 0);
395 			break;
396 		}
397 		smp_rmb();
398 		rxrpc_see_skb(skb, rxrpc_skb_seen);
399 		sp = rxrpc_skb(skb);
400 
401 		if (!(flags & MSG_PEEK)) {
402 			serial = sp->hdr.serial;
403 			serial += call->rxtx_annotations[ix] & RXRPC_RX_ANNO_SUBPACKET;
404 			trace_rxrpc_receive(call, rxrpc_receive_front,
405 					    serial, seq);
406 		}
407 
408 		if (msg)
409 			sock_recv_timestamp(msg, sock->sk, skb);
410 
411 		if (rx_pkt_offset == 0) {
412 			ret2 = rxrpc_locate_data(call, skb,
413 						 &call->rxtx_annotations[ix],
414 						 &rx_pkt_offset, &rx_pkt_len,
415 						 &rx_pkt_last);
416 			trace_rxrpc_recvmsg(call, rxrpc_recvmsg_next, seq,
417 					    rx_pkt_offset, rx_pkt_len, ret2);
418 			if (ret2 < 0) {
419 				ret = ret2;
420 				goto out;
421 			}
422 		} else {
423 			trace_rxrpc_recvmsg(call, rxrpc_recvmsg_cont, seq,
424 					    rx_pkt_offset, rx_pkt_len, 0);
425 		}
426 
427 		/* We have to handle short, empty and used-up DATA packets. */
428 		remain = len - *_offset;
429 		copy = rx_pkt_len;
430 		if (copy > remain)
431 			copy = remain;
432 		if (copy > 0) {
433 			ret2 = skb_copy_datagram_iter(skb, rx_pkt_offset, iter,
434 						      copy);
435 			if (ret2 < 0) {
436 				ret = ret2;
437 				goto out;
438 			}
439 
440 			/* handle piecemeal consumption of data packets */
441 			rx_pkt_offset += copy;
442 			rx_pkt_len -= copy;
443 			*_offset += copy;
444 		}
445 
446 		if (rx_pkt_len > 0) {
447 			trace_rxrpc_recvmsg(call, rxrpc_recvmsg_full, seq,
448 					    rx_pkt_offset, rx_pkt_len, 0);
449 			ASSERTCMP(*_offset, ==, len);
450 			ret = 0;
451 			break;
452 		}
453 
454 		/* The whole packet has been transferred. */
455 		if (!(flags & MSG_PEEK))
456 			rxrpc_rotate_rx_window(call);
457 		rx_pkt_offset = 0;
458 		rx_pkt_len = 0;
459 
460 		if (rx_pkt_last) {
461 			ASSERTCMP(seq, ==, READ_ONCE(call->rx_top));
462 			ret = 1;
463 			goto out;
464 		}
465 
466 		seq++;
467 	}
468 
469 out:
470 	if (!(flags & MSG_PEEK)) {
471 		call->rx_pkt_offset = rx_pkt_offset;
472 		call->rx_pkt_len = rx_pkt_len;
473 		call->rx_pkt_last = rx_pkt_last;
474 	}
475 done:
476 	trace_rxrpc_recvmsg(call, rxrpc_recvmsg_data_return, seq,
477 			    rx_pkt_offset, rx_pkt_len, ret);
478 	if (ret == -EAGAIN)
479 		set_bit(RXRPC_CALL_RX_UNDERRUN, &call->flags);
480 	return ret;
481 }
482 
483 /*
484  * Receive a message from an RxRPC socket
485  * - we need to be careful about two or more threads calling recvmsg
486  *   simultaneously
487  */
488 int rxrpc_recvmsg(struct socket *sock, struct msghdr *msg, size_t len,
489 		  int flags)
490 {
491 	struct rxrpc_call *call;
492 	struct rxrpc_sock *rx = rxrpc_sk(sock->sk);
493 	struct list_head *l;
494 	size_t copied = 0;
495 	long timeo;
496 	int ret;
497 
498 	DEFINE_WAIT(wait);
499 
500 	trace_rxrpc_recvmsg(NULL, rxrpc_recvmsg_enter, 0, 0, 0, 0);
501 
502 	if (flags & (MSG_OOB | MSG_TRUNC))
503 		return -EOPNOTSUPP;
504 
505 	timeo = sock_rcvtimeo(&rx->sk, flags & MSG_DONTWAIT);
506 
507 try_again:
508 	lock_sock(&rx->sk);
509 
510 	/* Return immediately if a client socket has no outstanding calls */
511 	if (RB_EMPTY_ROOT(&rx->calls) &&
512 	    list_empty(&rx->recvmsg_q) &&
513 	    rx->sk.sk_state != RXRPC_SERVER_LISTENING) {
514 		release_sock(&rx->sk);
515 		return -EAGAIN;
516 	}
517 
518 	if (list_empty(&rx->recvmsg_q)) {
519 		ret = -EWOULDBLOCK;
520 		if (timeo == 0) {
521 			call = NULL;
522 			goto error_no_call;
523 		}
524 
525 		release_sock(&rx->sk);
526 
527 		/* Wait for something to happen */
528 		prepare_to_wait_exclusive(sk_sleep(&rx->sk), &wait,
529 					  TASK_INTERRUPTIBLE);
530 		ret = sock_error(&rx->sk);
531 		if (ret)
532 			goto wait_error;
533 
534 		if (list_empty(&rx->recvmsg_q)) {
535 			if (signal_pending(current))
536 				goto wait_interrupted;
537 			trace_rxrpc_recvmsg(NULL, rxrpc_recvmsg_wait,
538 					    0, 0, 0, 0);
539 			timeo = schedule_timeout(timeo);
540 		}
541 		finish_wait(sk_sleep(&rx->sk), &wait);
542 		goto try_again;
543 	}
544 
545 	/* Find the next call and dequeue it if we're not just peeking.  If we
546 	 * do dequeue it, that comes with a ref that we will need to release.
547 	 */
548 	write_lock_bh(&rx->recvmsg_lock);
549 	l = rx->recvmsg_q.next;
550 	call = list_entry(l, struct rxrpc_call, recvmsg_link);
551 	if (!(flags & MSG_PEEK))
552 		list_del_init(&call->recvmsg_link);
553 	else
554 		rxrpc_get_call(call, rxrpc_call_got);
555 	write_unlock_bh(&rx->recvmsg_lock);
556 
557 	trace_rxrpc_recvmsg(call, rxrpc_recvmsg_dequeue, 0, 0, 0, 0);
558 
559 	/* We're going to drop the socket lock, so we need to lock the call
560 	 * against interference by sendmsg.
561 	 */
562 	if (!mutex_trylock(&call->user_mutex)) {
563 		ret = -EWOULDBLOCK;
564 		if (flags & MSG_DONTWAIT)
565 			goto error_requeue_call;
566 		ret = -ERESTARTSYS;
567 		if (mutex_lock_interruptible(&call->user_mutex) < 0)
568 			goto error_requeue_call;
569 	}
570 
571 	release_sock(&rx->sk);
572 
573 	if (test_bit(RXRPC_CALL_RELEASED, &call->flags))
574 		BUG();
575 
576 	if (test_bit(RXRPC_CALL_HAS_USERID, &call->flags)) {
577 		if (flags & MSG_CMSG_COMPAT) {
578 			unsigned int id32 = call->user_call_ID;
579 
580 			ret = put_cmsg(msg, SOL_RXRPC, RXRPC_USER_CALL_ID,
581 				       sizeof(unsigned int), &id32);
582 		} else {
583 			unsigned long idl = call->user_call_ID;
584 
585 			ret = put_cmsg(msg, SOL_RXRPC, RXRPC_USER_CALL_ID,
586 				       sizeof(unsigned long), &idl);
587 		}
588 		if (ret < 0)
589 			goto error_unlock_call;
590 	}
591 
592 	if (msg->msg_name && call->peer) {
593 		struct sockaddr_rxrpc *srx = msg->msg_name;
594 		size_t len = sizeof(call->peer->srx);
595 
596 		memcpy(msg->msg_name, &call->peer->srx, len);
597 		srx->srx_service = call->service_id;
598 		msg->msg_namelen = len;
599 	}
600 
601 	switch (READ_ONCE(call->state)) {
602 	case RXRPC_CALL_CLIENT_RECV_REPLY:
603 	case RXRPC_CALL_SERVER_RECV_REQUEST:
604 	case RXRPC_CALL_SERVER_ACK_REQUEST:
605 		ret = rxrpc_recvmsg_data(sock, call, msg, &msg->msg_iter, len,
606 					 flags, &copied);
607 		if (ret == -EAGAIN)
608 			ret = 0;
609 
610 		if (after(call->rx_top, call->rx_hard_ack) &&
611 		    call->rxtx_buffer[(call->rx_hard_ack + 1) & RXRPC_RXTX_BUFF_MASK])
612 			rxrpc_notify_socket(call);
613 		break;
614 	default:
615 		ret = 0;
616 		break;
617 	}
618 
619 	if (ret < 0)
620 		goto error_unlock_call;
621 
622 	if (call->state == RXRPC_CALL_COMPLETE) {
623 		ret = rxrpc_recvmsg_term(call, msg);
624 		if (ret < 0)
625 			goto error_unlock_call;
626 		if (!(flags & MSG_PEEK))
627 			rxrpc_release_call(rx, call);
628 		msg->msg_flags |= MSG_EOR;
629 		ret = 1;
630 	}
631 
632 	if (ret == 0)
633 		msg->msg_flags |= MSG_MORE;
634 	else
635 		msg->msg_flags &= ~MSG_MORE;
636 	ret = copied;
637 
638 error_unlock_call:
639 	mutex_unlock(&call->user_mutex);
640 	rxrpc_put_call(call, rxrpc_call_put);
641 	trace_rxrpc_recvmsg(call, rxrpc_recvmsg_return, 0, 0, 0, ret);
642 	return ret;
643 
644 error_requeue_call:
645 	if (!(flags & MSG_PEEK)) {
646 		write_lock_bh(&rx->recvmsg_lock);
647 		list_add(&call->recvmsg_link, &rx->recvmsg_q);
648 		write_unlock_bh(&rx->recvmsg_lock);
649 		trace_rxrpc_recvmsg(call, rxrpc_recvmsg_requeue, 0, 0, 0, 0);
650 	} else {
651 		rxrpc_put_call(call, rxrpc_call_put);
652 	}
653 error_no_call:
654 	release_sock(&rx->sk);
655 error_trace:
656 	trace_rxrpc_recvmsg(call, rxrpc_recvmsg_return, 0, 0, 0, ret);
657 	return ret;
658 
659 wait_interrupted:
660 	ret = sock_intr_errno(timeo);
661 wait_error:
662 	finish_wait(sk_sleep(&rx->sk), &wait);
663 	call = NULL;
664 	goto error_trace;
665 }
666 
667 /**
668  * rxrpc_kernel_recv_data - Allow a kernel service to receive data/info
669  * @sock: The socket that the call exists on
670  * @call: The call to send data through
671  * @iter: The buffer to receive into
672  * @_len: The amount of data we want to receive (decreased on return)
673  * @want_more: True if more data is expected to be read
674  * @_abort: Where the abort code is stored if -ECONNABORTED is returned
675  * @_service: Where to store the actual service ID (may be upgraded)
676  *
677  * Allow a kernel service to receive data and pick up information about the
678  * state of a call.  Returns 0 if got what was asked for and there's more
679  * available, 1 if we got what was asked for and we're at the end of the data
680  * and -EAGAIN if we need more data.
681  *
682  * Note that we may return -EAGAIN to drain empty packets at the end of the
683  * data, even if we've already copied over the requested data.
684  *
685  * *_abort should also be initialised to 0.
686  */
687 int rxrpc_kernel_recv_data(struct socket *sock, struct rxrpc_call *call,
688 			   struct iov_iter *iter, size_t *_len,
689 			   bool want_more, u32 *_abort, u16 *_service)
690 {
691 	size_t offset = 0;
692 	int ret;
693 
694 	_enter("{%d,%s},%zu,%d",
695 	       call->debug_id, rxrpc_call_states[call->state],
696 	       *_len, want_more);
697 
698 	ASSERTCMP(call->state, !=, RXRPC_CALL_SERVER_SECURING);
699 
700 	mutex_lock(&call->user_mutex);
701 
702 	switch (READ_ONCE(call->state)) {
703 	case RXRPC_CALL_CLIENT_RECV_REPLY:
704 	case RXRPC_CALL_SERVER_RECV_REQUEST:
705 	case RXRPC_CALL_SERVER_ACK_REQUEST:
706 		ret = rxrpc_recvmsg_data(sock, call, NULL, iter,
707 					 *_len, 0, &offset);
708 		*_len -= offset;
709 		if (ret < 0)
710 			goto out;
711 
712 		/* We can only reach here with a partially full buffer if we
713 		 * have reached the end of the data.  We must otherwise have a
714 		 * full buffer or have been given -EAGAIN.
715 		 */
716 		if (ret == 1) {
717 			if (iov_iter_count(iter) > 0)
718 				goto short_data;
719 			if (!want_more)
720 				goto read_phase_complete;
721 			ret = 0;
722 			goto out;
723 		}
724 
725 		if (!want_more)
726 			goto excess_data;
727 		goto out;
728 
729 	case RXRPC_CALL_COMPLETE:
730 		goto call_complete;
731 
732 	default:
733 		ret = -EINPROGRESS;
734 		goto out;
735 	}
736 
737 read_phase_complete:
738 	ret = 1;
739 out:
740 	switch (call->ackr_reason) {
741 	case RXRPC_ACK_IDLE:
742 		break;
743 	case RXRPC_ACK_DELAY:
744 		if (ret != -EAGAIN)
745 			break;
746 		fallthrough;
747 	default:
748 		rxrpc_send_ack_packet(call, false, NULL);
749 	}
750 
751 	if (_service)
752 		*_service = call->service_id;
753 	mutex_unlock(&call->user_mutex);
754 	_leave(" = %d [%zu,%d]", ret, iov_iter_count(iter), *_abort);
755 	return ret;
756 
757 short_data:
758 	trace_rxrpc_rx_eproto(call, 0, tracepoint_string("short_data"));
759 	ret = -EBADMSG;
760 	goto out;
761 excess_data:
762 	trace_rxrpc_rx_eproto(call, 0, tracepoint_string("excess_data"));
763 	ret = -EMSGSIZE;
764 	goto out;
765 call_complete:
766 	*_abort = call->abort_code;
767 	ret = call->error;
768 	if (call->completion == RXRPC_CALL_SUCCEEDED) {
769 		ret = 1;
770 		if (iov_iter_count(iter) > 0)
771 			ret = -ECONNRESET;
772 	}
773 	goto out;
774 }
775 EXPORT_SYMBOL(rxrpc_kernel_recv_data);
776 
777 /**
778  * rxrpc_kernel_get_reply_time - Get timestamp on first reply packet
779  * @sock: The socket that the call exists on
780  * @call: The call to query
781  * @_ts: Where to put the timestamp
782  *
783  * Retrieve the timestamp from the first DATA packet of the reply if it is
784  * in the ring.  Returns true if successful, false if not.
785  */
786 bool rxrpc_kernel_get_reply_time(struct socket *sock, struct rxrpc_call *call,
787 				 ktime_t *_ts)
788 {
789 	struct sk_buff *skb;
790 	rxrpc_seq_t hard_ack, top, seq;
791 	bool success = false;
792 
793 	mutex_lock(&call->user_mutex);
794 
795 	if (READ_ONCE(call->state) != RXRPC_CALL_CLIENT_RECV_REPLY)
796 		goto out;
797 
798 	hard_ack = call->rx_hard_ack;
799 	if (hard_ack != 0)
800 		goto out;
801 
802 	seq = hard_ack + 1;
803 	top = smp_load_acquire(&call->rx_top);
804 	if (after(seq, top))
805 		goto out;
806 
807 	skb = call->rxtx_buffer[seq & RXRPC_RXTX_BUFF_MASK];
808 	if (!skb)
809 		goto out;
810 
811 	*_ts = skb_get_ktime(skb);
812 	success = true;
813 
814 out:
815 	mutex_unlock(&call->user_mutex);
816 	return success;
817 }
818 EXPORT_SYMBOL(rxrpc_kernel_get_reply_time);
819