• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1 // SPDX-License-Identifier: GPL-2.0-or-later
2 /* Client connection-specific management code.
3  *
4  * Copyright (C) 2016, 2020 Red Hat, Inc. All Rights Reserved.
5  * Written by David Howells (dhowells@redhat.com)
6  *
7  * Client connections need to be cached for a little while after they've made a
8  * call so as to handle retransmitted DATA packets in case the server didn't
9  * receive the final ACK or terminating ABORT we sent it.
10  *
11  * There are flags of relevance to the cache:
12  *
13  *  (2) DONT_REUSE - The connection should be discarded as soon as possible and
14  *      should not be reused.  This is set when an exclusive connection is used
15  *      or a call ID counter overflows.
16  *
17  * The caching state may only be changed if the cache lock is held.
18  *
19  * There are two idle client connection expiry durations.  If the total number
20  * of connections is below the reap threshold, we use the normal duration; if
21  * it's above, we use the fast duration.
22  */
23 
24 #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
25 
26 #include <linux/slab.h>
27 #include <linux/idr.h>
28 #include <linux/timer.h>
29 #include <linux/sched/signal.h>
30 
31 #include "ar-internal.h"
32 
33 __read_mostly unsigned int rxrpc_reap_client_connections = 900;
34 __read_mostly unsigned long rxrpc_conn_idle_client_expiry = 2 * 60 * HZ;
35 __read_mostly unsigned long rxrpc_conn_idle_client_fast_expiry = 2 * HZ;
36 
37 /*
38  * We use machine-unique IDs for our client connections.
39  */
40 DEFINE_IDR(rxrpc_client_conn_ids);
41 static DEFINE_SPINLOCK(rxrpc_conn_id_lock);
42 
43 static void rxrpc_deactivate_bundle(struct rxrpc_bundle *bundle);
44 
45 /*
46  * Get a connection ID and epoch for a client connection from the global pool.
47  * The connection struct pointer is then recorded in the idr radix tree.  The
48  * epoch doesn't change until the client is rebooted (or, at least, unless the
49  * module is unloaded).
50  */
rxrpc_get_client_connection_id(struct rxrpc_connection * conn,gfp_t gfp)51 static int rxrpc_get_client_connection_id(struct rxrpc_connection *conn,
52 					  gfp_t gfp)
53 {
54 	struct rxrpc_net *rxnet = conn->params.local->rxnet;
55 	int id;
56 
57 	_enter("");
58 
59 	idr_preload(gfp);
60 	spin_lock(&rxrpc_conn_id_lock);
61 
62 	id = idr_alloc_cyclic(&rxrpc_client_conn_ids, conn,
63 			      1, 0x40000000, GFP_NOWAIT);
64 	if (id < 0)
65 		goto error;
66 
67 	spin_unlock(&rxrpc_conn_id_lock);
68 	idr_preload_end();
69 
70 	conn->proto.epoch = rxnet->epoch;
71 	conn->proto.cid = id << RXRPC_CIDSHIFT;
72 	set_bit(RXRPC_CONN_HAS_IDR, &conn->flags);
73 	_leave(" [CID %x]", conn->proto.cid);
74 	return 0;
75 
76 error:
77 	spin_unlock(&rxrpc_conn_id_lock);
78 	idr_preload_end();
79 	_leave(" = %d", id);
80 	return id;
81 }
82 
83 /*
84  * Release a connection ID for a client connection from the global pool.
85  */
rxrpc_put_client_connection_id(struct rxrpc_connection * conn)86 static void rxrpc_put_client_connection_id(struct rxrpc_connection *conn)
87 {
88 	if (test_bit(RXRPC_CONN_HAS_IDR, &conn->flags)) {
89 		spin_lock(&rxrpc_conn_id_lock);
90 		idr_remove(&rxrpc_client_conn_ids,
91 			   conn->proto.cid >> RXRPC_CIDSHIFT);
92 		spin_unlock(&rxrpc_conn_id_lock);
93 	}
94 }
95 
96 /*
97  * Destroy the client connection ID tree.
98  */
rxrpc_destroy_client_conn_ids(void)99 void rxrpc_destroy_client_conn_ids(void)
100 {
101 	struct rxrpc_connection *conn;
102 	int id;
103 
104 	if (!idr_is_empty(&rxrpc_client_conn_ids)) {
105 		idr_for_each_entry(&rxrpc_client_conn_ids, conn, id) {
106 			pr_err("AF_RXRPC: Leaked client conn %p {%d}\n",
107 			       conn, refcount_read(&conn->ref));
108 		}
109 		BUG();
110 	}
111 
112 	idr_destroy(&rxrpc_client_conn_ids);
113 }
114 
115 /*
116  * Allocate a connection bundle.
117  */
rxrpc_alloc_bundle(struct rxrpc_conn_parameters * cp,gfp_t gfp)118 static struct rxrpc_bundle *rxrpc_alloc_bundle(struct rxrpc_conn_parameters *cp,
119 					       gfp_t gfp)
120 {
121 	struct rxrpc_bundle *bundle;
122 
123 	bundle = kzalloc(sizeof(*bundle), gfp);
124 	if (bundle) {
125 		bundle->params = *cp;
126 		rxrpc_get_peer(bundle->params.peer);
127 		refcount_set(&bundle->ref, 1);
128 		atomic_set(&bundle->active, 1);
129 		spin_lock_init(&bundle->channel_lock);
130 		INIT_LIST_HEAD(&bundle->waiting_calls);
131 	}
132 	return bundle;
133 }
134 
rxrpc_get_bundle(struct rxrpc_bundle * bundle)135 struct rxrpc_bundle *rxrpc_get_bundle(struct rxrpc_bundle *bundle)
136 {
137 	refcount_inc(&bundle->ref);
138 	return bundle;
139 }
140 
rxrpc_free_bundle(struct rxrpc_bundle * bundle)141 static void rxrpc_free_bundle(struct rxrpc_bundle *bundle)
142 {
143 	rxrpc_put_peer(bundle->params.peer);
144 	kfree(bundle);
145 }
146 
rxrpc_put_bundle(struct rxrpc_bundle * bundle)147 void rxrpc_put_bundle(struct rxrpc_bundle *bundle)
148 {
149 	unsigned int d = bundle->debug_id;
150 	bool dead;
151 	int r;
152 
153 	dead = __refcount_dec_and_test(&bundle->ref, &r);
154 
155 	_debug("PUT B=%x %d", d, r - 1);
156 	if (dead)
157 		rxrpc_free_bundle(bundle);
158 }
159 
160 /*
161  * Allocate a client connection.
162  */
163 static struct rxrpc_connection *
rxrpc_alloc_client_connection(struct rxrpc_bundle * bundle,gfp_t gfp)164 rxrpc_alloc_client_connection(struct rxrpc_bundle *bundle, gfp_t gfp)
165 {
166 	struct rxrpc_connection *conn;
167 	struct rxrpc_net *rxnet = bundle->params.local->rxnet;
168 	int ret;
169 
170 	_enter("");
171 
172 	conn = rxrpc_alloc_connection(gfp);
173 	if (!conn) {
174 		_leave(" = -ENOMEM");
175 		return ERR_PTR(-ENOMEM);
176 	}
177 
178 	refcount_set(&conn->ref, 1);
179 	conn->bundle		= bundle;
180 	conn->params		= bundle->params;
181 	conn->out_clientflag	= RXRPC_CLIENT_INITIATED;
182 	conn->state		= RXRPC_CONN_CLIENT;
183 	conn->service_id	= conn->params.service_id;
184 
185 	ret = rxrpc_get_client_connection_id(conn, gfp);
186 	if (ret < 0)
187 		goto error_0;
188 
189 	ret = rxrpc_init_client_conn_security(conn);
190 	if (ret < 0)
191 		goto error_1;
192 
193 	ret = conn->security->prime_packet_security(conn);
194 	if (ret < 0)
195 		goto error_2;
196 
197 	atomic_inc(&rxnet->nr_conns);
198 	write_lock(&rxnet->conn_lock);
199 	list_add_tail(&conn->proc_link, &rxnet->conn_proc_list);
200 	write_unlock(&rxnet->conn_lock);
201 
202 	rxrpc_get_bundle(bundle);
203 	rxrpc_get_peer(conn->params.peer);
204 	rxrpc_get_local(conn->params.local);
205 	key_get(conn->params.key);
206 
207 	trace_rxrpc_conn(conn->debug_id, rxrpc_conn_new_client,
208 			 refcount_read(&conn->ref),
209 			 __builtin_return_address(0));
210 
211 	atomic_inc(&rxnet->nr_client_conns);
212 	trace_rxrpc_client(conn, -1, rxrpc_client_alloc);
213 	_leave(" = %p", conn);
214 	return conn;
215 
216 error_2:
217 	conn->security->clear(conn);
218 error_1:
219 	rxrpc_put_client_connection_id(conn);
220 error_0:
221 	kfree(conn);
222 	_leave(" = %d", ret);
223 	return ERR_PTR(ret);
224 }
225 
226 /*
227  * Determine if a connection may be reused.
228  */
rxrpc_may_reuse_conn(struct rxrpc_connection * conn)229 static bool rxrpc_may_reuse_conn(struct rxrpc_connection *conn)
230 {
231 	struct rxrpc_net *rxnet;
232 	int id_cursor, id, distance, limit;
233 
234 	if (!conn)
235 		goto dont_reuse;
236 
237 	rxnet = conn->params.local->rxnet;
238 	if (test_bit(RXRPC_CONN_DONT_REUSE, &conn->flags))
239 		goto dont_reuse;
240 
241 	if (conn->state != RXRPC_CONN_CLIENT ||
242 	    conn->proto.epoch != rxnet->epoch)
243 		goto mark_dont_reuse;
244 
245 	/* The IDR tree gets very expensive on memory if the connection IDs are
246 	 * widely scattered throughout the number space, so we shall want to
247 	 * kill off connections that, say, have an ID more than about four
248 	 * times the maximum number of client conns away from the current
249 	 * allocation point to try and keep the IDs concentrated.
250 	 */
251 	id_cursor = idr_get_cursor(&rxrpc_client_conn_ids);
252 	id = conn->proto.cid >> RXRPC_CIDSHIFT;
253 	distance = id - id_cursor;
254 	if (distance < 0)
255 		distance = -distance;
256 	limit = max_t(unsigned long, atomic_read(&rxnet->nr_conns) * 4, 1024);
257 	if (distance > limit)
258 		goto mark_dont_reuse;
259 
260 	return true;
261 
262 mark_dont_reuse:
263 	set_bit(RXRPC_CONN_DONT_REUSE, &conn->flags);
264 dont_reuse:
265 	return false;
266 }
267 
268 /*
269  * Look up the conn bundle that matches the connection parameters, adding it if
270  * it doesn't yet exist.
271  */
rxrpc_look_up_bundle(struct rxrpc_conn_parameters * cp,gfp_t gfp)272 static struct rxrpc_bundle *rxrpc_look_up_bundle(struct rxrpc_conn_parameters *cp,
273 						 gfp_t gfp)
274 {
275 	static atomic_t rxrpc_bundle_id;
276 	struct rxrpc_bundle *bundle, *candidate;
277 	struct rxrpc_local *local = cp->local;
278 	struct rb_node *p, **pp, *parent;
279 	long diff;
280 
281 	_enter("{%px,%x,%u,%u}",
282 	       cp->peer, key_serial(cp->key), cp->security_level, cp->upgrade);
283 
284 	if (cp->exclusive)
285 		return rxrpc_alloc_bundle(cp, gfp);
286 
287 	/* First, see if the bundle is already there. */
288 	_debug("search 1");
289 	spin_lock(&local->client_bundles_lock);
290 	p = local->client_bundles.rb_node;
291 	while (p) {
292 		bundle = rb_entry(p, struct rxrpc_bundle, local_node);
293 
294 #define cmp(X) ((long)bundle->params.X - (long)cp->X)
295 		diff = (cmp(peer) ?:
296 			cmp(key) ?:
297 			cmp(security_level) ?:
298 			cmp(upgrade));
299 #undef cmp
300 		if (diff < 0)
301 			p = p->rb_left;
302 		else if (diff > 0)
303 			p = p->rb_right;
304 		else
305 			goto found_bundle;
306 	}
307 	spin_unlock(&local->client_bundles_lock);
308 	_debug("not found");
309 
310 	/* It wasn't.  We need to add one. */
311 	candidate = rxrpc_alloc_bundle(cp, gfp);
312 	if (!candidate)
313 		return NULL;
314 
315 	_debug("search 2");
316 	spin_lock(&local->client_bundles_lock);
317 	pp = &local->client_bundles.rb_node;
318 	parent = NULL;
319 	while (*pp) {
320 		parent = *pp;
321 		bundle = rb_entry(parent, struct rxrpc_bundle, local_node);
322 
323 #define cmp(X) ((long)bundle->params.X - (long)cp->X)
324 		diff = (cmp(peer) ?:
325 			cmp(key) ?:
326 			cmp(security_level) ?:
327 			cmp(upgrade));
328 #undef cmp
329 		if (diff < 0)
330 			pp = &(*pp)->rb_left;
331 		else if (diff > 0)
332 			pp = &(*pp)->rb_right;
333 		else
334 			goto found_bundle_free;
335 	}
336 
337 	_debug("new bundle");
338 	candidate->debug_id = atomic_inc_return(&rxrpc_bundle_id);
339 	rb_link_node(&candidate->local_node, parent, pp);
340 	rb_insert_color(&candidate->local_node, &local->client_bundles);
341 	rxrpc_get_bundle(candidate);
342 	spin_unlock(&local->client_bundles_lock);
343 	_leave(" = %u [new]", candidate->debug_id);
344 	return candidate;
345 
346 found_bundle_free:
347 	rxrpc_free_bundle(candidate);
348 found_bundle:
349 	rxrpc_get_bundle(bundle);
350 	atomic_inc(&bundle->active);
351 	spin_unlock(&local->client_bundles_lock);
352 	_leave(" = %u [found]", bundle->debug_id);
353 	return bundle;
354 }
355 
356 /*
357  * Create or find a client bundle to use for a call.
358  *
359  * If we return with a connection, the call will be on its waiting list.  It's
360  * left to the caller to assign a channel and wake up the call.
361  */
rxrpc_prep_call(struct rxrpc_sock * rx,struct rxrpc_call * call,struct rxrpc_conn_parameters * cp,struct sockaddr_rxrpc * srx,gfp_t gfp)362 static struct rxrpc_bundle *rxrpc_prep_call(struct rxrpc_sock *rx,
363 					    struct rxrpc_call *call,
364 					    struct rxrpc_conn_parameters *cp,
365 					    struct sockaddr_rxrpc *srx,
366 					    gfp_t gfp)
367 {
368 	struct rxrpc_bundle *bundle;
369 
370 	_enter("{%d,%lx},", call->debug_id, call->user_call_ID);
371 
372 	cp->peer = rxrpc_lookup_peer(rx, cp->local, srx, gfp);
373 	if (!cp->peer)
374 		goto error;
375 
376 	call->cong_cwnd = cp->peer->cong_cwnd;
377 	if (call->cong_cwnd >= call->cong_ssthresh)
378 		call->cong_mode = RXRPC_CALL_CONGEST_AVOIDANCE;
379 	else
380 		call->cong_mode = RXRPC_CALL_SLOW_START;
381 	if (cp->upgrade)
382 		__set_bit(RXRPC_CALL_UPGRADE, &call->flags);
383 
384 	/* Find the client connection bundle. */
385 	bundle = rxrpc_look_up_bundle(cp, gfp);
386 	if (!bundle)
387 		goto error;
388 
389 	/* Get this call queued.  Someone else may activate it whilst we're
390 	 * lining up a new connection, but that's fine.
391 	 */
392 	spin_lock(&bundle->channel_lock);
393 	list_add_tail(&call->chan_wait_link, &bundle->waiting_calls);
394 	spin_unlock(&bundle->channel_lock);
395 
396 	_leave(" = [B=%x]", bundle->debug_id);
397 	return bundle;
398 
399 error:
400 	_leave(" = -ENOMEM");
401 	return ERR_PTR(-ENOMEM);
402 }
403 
404 /*
405  * Allocate a new connection and add it into a bundle.
406  */
rxrpc_add_conn_to_bundle(struct rxrpc_bundle * bundle,gfp_t gfp)407 static void rxrpc_add_conn_to_bundle(struct rxrpc_bundle *bundle, gfp_t gfp)
408 	__releases(bundle->channel_lock)
409 {
410 	struct rxrpc_connection *candidate = NULL, *old = NULL;
411 	bool conflict;
412 	int i;
413 
414 	_enter("");
415 
416 	conflict = bundle->alloc_conn;
417 	if (!conflict)
418 		bundle->alloc_conn = true;
419 	spin_unlock(&bundle->channel_lock);
420 	if (conflict) {
421 		_leave(" [conf]");
422 		return;
423 	}
424 
425 	candidate = rxrpc_alloc_client_connection(bundle, gfp);
426 
427 	spin_lock(&bundle->channel_lock);
428 	bundle->alloc_conn = false;
429 
430 	if (IS_ERR(candidate)) {
431 		bundle->alloc_error = PTR_ERR(candidate);
432 		spin_unlock(&bundle->channel_lock);
433 		_leave(" [err %ld]", PTR_ERR(candidate));
434 		return;
435 	}
436 
437 	bundle->alloc_error = 0;
438 
439 	for (i = 0; i < ARRAY_SIZE(bundle->conns); i++) {
440 		unsigned int shift = i * RXRPC_MAXCALLS;
441 		int j;
442 
443 		old = bundle->conns[i];
444 		if (!rxrpc_may_reuse_conn(old)) {
445 			if (old)
446 				trace_rxrpc_client(old, -1, rxrpc_client_replace);
447 			candidate->bundle_shift = shift;
448 			atomic_inc(&bundle->active);
449 			bundle->conns[i] = candidate;
450 			for (j = 0; j < RXRPC_MAXCALLS; j++)
451 				set_bit(shift + j, &bundle->avail_chans);
452 			candidate = NULL;
453 			break;
454 		}
455 
456 		old = NULL;
457 	}
458 
459 	spin_unlock(&bundle->channel_lock);
460 
461 	if (candidate) {
462 		_debug("discard C=%x", candidate->debug_id);
463 		trace_rxrpc_client(candidate, -1, rxrpc_client_duplicate);
464 		rxrpc_put_connection(candidate);
465 	}
466 
467 	rxrpc_put_connection(old);
468 	_leave("");
469 }
470 
471 /*
472  * Add a connection to a bundle if there are no usable connections or we have
473  * connections waiting for extra capacity.
474  */
rxrpc_maybe_add_conn(struct rxrpc_bundle * bundle,gfp_t gfp)475 static void rxrpc_maybe_add_conn(struct rxrpc_bundle *bundle, gfp_t gfp)
476 {
477 	struct rxrpc_call *call;
478 	int i, usable;
479 
480 	_enter("");
481 
482 	spin_lock(&bundle->channel_lock);
483 
484 	/* See if there are any usable connections. */
485 	usable = 0;
486 	for (i = 0; i < ARRAY_SIZE(bundle->conns); i++)
487 		if (rxrpc_may_reuse_conn(bundle->conns[i]))
488 			usable++;
489 
490 	if (!usable && !list_empty(&bundle->waiting_calls)) {
491 		call = list_first_entry(&bundle->waiting_calls,
492 					struct rxrpc_call, chan_wait_link);
493 		if (test_bit(RXRPC_CALL_UPGRADE, &call->flags))
494 			bundle->try_upgrade = true;
495 	}
496 
497 	if (!usable)
498 		goto alloc_conn;
499 
500 	if (!bundle->avail_chans &&
501 	    !bundle->try_upgrade &&
502 	    !list_empty(&bundle->waiting_calls) &&
503 	    usable < ARRAY_SIZE(bundle->conns))
504 		goto alloc_conn;
505 
506 	spin_unlock(&bundle->channel_lock);
507 	_leave("");
508 	return;
509 
510 alloc_conn:
511 	return rxrpc_add_conn_to_bundle(bundle, gfp);
512 }
513 
514 /*
515  * Assign a channel to the call at the front of the queue and wake the call up.
516  * We don't increment the callNumber counter until this number has been exposed
517  * to the world.
518  */
rxrpc_activate_one_channel(struct rxrpc_connection * conn,unsigned int channel)519 static void rxrpc_activate_one_channel(struct rxrpc_connection *conn,
520 				       unsigned int channel)
521 {
522 	struct rxrpc_channel *chan = &conn->channels[channel];
523 	struct rxrpc_bundle *bundle = conn->bundle;
524 	struct rxrpc_call *call = list_entry(bundle->waiting_calls.next,
525 					     struct rxrpc_call, chan_wait_link);
526 	u32 call_id = chan->call_counter + 1;
527 
528 	_enter("C=%x,%u", conn->debug_id, channel);
529 
530 	trace_rxrpc_client(conn, channel, rxrpc_client_chan_activate);
531 
532 	/* Cancel the final ACK on the previous call if it hasn't been sent yet
533 	 * as the DATA packet will implicitly ACK it.
534 	 */
535 	clear_bit(RXRPC_CONN_FINAL_ACK_0 + channel, &conn->flags);
536 	clear_bit(conn->bundle_shift + channel, &bundle->avail_chans);
537 
538 	rxrpc_see_call(call);
539 	list_del_init(&call->chan_wait_link);
540 	call->peer	= rxrpc_get_peer(conn->params.peer);
541 	call->conn	= rxrpc_get_connection(conn);
542 	call->cid	= conn->proto.cid | channel;
543 	call->call_id	= call_id;
544 	call->security	= conn->security;
545 	call->security_ix = conn->security_ix;
546 	call->service_id = conn->service_id;
547 
548 	trace_rxrpc_connect_call(call);
549 	_net("CONNECT call %08x:%08x as call %d on conn %d",
550 	     call->cid, call->call_id, call->debug_id, conn->debug_id);
551 
552 	write_lock_bh(&call->state_lock);
553 	call->state = RXRPC_CALL_CLIENT_SEND_REQUEST;
554 	write_unlock_bh(&call->state_lock);
555 
556 	/* Paired with the read barrier in rxrpc_connect_call().  This orders
557 	 * cid and epoch in the connection wrt to call_id without the need to
558 	 * take the channel_lock.
559 	 *
560 	 * We provisionally assign a callNumber at this point, but we don't
561 	 * confirm it until the call is about to be exposed.
562 	 *
563 	 * TODO: Pair with a barrier in the data_ready handler when that looks
564 	 * at the call ID through a connection channel.
565 	 */
566 	smp_wmb();
567 
568 	chan->call_id		= call_id;
569 	chan->call_debug_id	= call->debug_id;
570 	rcu_assign_pointer(chan->call, call);
571 	wake_up(&call->waitq);
572 }
573 
574 /*
575  * Remove a connection from the idle list if it's on it.
576  */
rxrpc_unidle_conn(struct rxrpc_bundle * bundle,struct rxrpc_connection * conn)577 static void rxrpc_unidle_conn(struct rxrpc_bundle *bundle, struct rxrpc_connection *conn)
578 {
579 	struct rxrpc_net *rxnet = bundle->params.local->rxnet;
580 	bool drop_ref;
581 
582 	if (!list_empty(&conn->cache_link)) {
583 		drop_ref = false;
584 		spin_lock(&rxnet->client_conn_cache_lock);
585 		if (!list_empty(&conn->cache_link)) {
586 			list_del_init(&conn->cache_link);
587 			drop_ref = true;
588 		}
589 		spin_unlock(&rxnet->client_conn_cache_lock);
590 		if (drop_ref)
591 			rxrpc_put_connection(conn);
592 	}
593 }
594 
595 /*
596  * Assign channels and callNumbers to waiting calls with channel_lock
597  * held by caller.
598  */
rxrpc_activate_channels_locked(struct rxrpc_bundle * bundle)599 static void rxrpc_activate_channels_locked(struct rxrpc_bundle *bundle)
600 {
601 	struct rxrpc_connection *conn;
602 	unsigned long avail, mask;
603 	unsigned int channel, slot;
604 
605 	if (bundle->try_upgrade)
606 		mask = 1;
607 	else
608 		mask = ULONG_MAX;
609 
610 	while (!list_empty(&bundle->waiting_calls)) {
611 		avail = bundle->avail_chans & mask;
612 		if (!avail)
613 			break;
614 		channel = __ffs(avail);
615 		clear_bit(channel, &bundle->avail_chans);
616 
617 		slot = channel / RXRPC_MAXCALLS;
618 		conn = bundle->conns[slot];
619 		if (!conn)
620 			break;
621 
622 		if (bundle->try_upgrade)
623 			set_bit(RXRPC_CONN_PROBING_FOR_UPGRADE, &conn->flags);
624 		rxrpc_unidle_conn(bundle, conn);
625 
626 		channel &= (RXRPC_MAXCALLS - 1);
627 		conn->act_chans	|= 1 << channel;
628 		rxrpc_activate_one_channel(conn, channel);
629 	}
630 }
631 
632 /*
633  * Assign channels and callNumbers to waiting calls.
634  */
rxrpc_activate_channels(struct rxrpc_bundle * bundle)635 static void rxrpc_activate_channels(struct rxrpc_bundle *bundle)
636 {
637 	_enter("B=%x", bundle->debug_id);
638 
639 	trace_rxrpc_client(NULL, -1, rxrpc_client_activate_chans);
640 
641 	if (!bundle->avail_chans)
642 		return;
643 
644 	spin_lock(&bundle->channel_lock);
645 	rxrpc_activate_channels_locked(bundle);
646 	spin_unlock(&bundle->channel_lock);
647 	_leave("");
648 }
649 
650 /*
651  * Wait for a callNumber and a channel to be granted to a call.
652  */
rxrpc_wait_for_channel(struct rxrpc_bundle * bundle,struct rxrpc_call * call,gfp_t gfp)653 static int rxrpc_wait_for_channel(struct rxrpc_bundle *bundle,
654 				  struct rxrpc_call *call, gfp_t gfp)
655 {
656 	DECLARE_WAITQUEUE(myself, current);
657 	int ret = 0;
658 
659 	_enter("%d", call->debug_id);
660 
661 	if (!gfpflags_allow_blocking(gfp)) {
662 		rxrpc_maybe_add_conn(bundle, gfp);
663 		rxrpc_activate_channels(bundle);
664 		ret = bundle->alloc_error ?: -EAGAIN;
665 		goto out;
666 	}
667 
668 	add_wait_queue_exclusive(&call->waitq, &myself);
669 	for (;;) {
670 		rxrpc_maybe_add_conn(bundle, gfp);
671 		rxrpc_activate_channels(bundle);
672 		ret = bundle->alloc_error;
673 		if (ret < 0)
674 			break;
675 
676 		switch (call->interruptibility) {
677 		case RXRPC_INTERRUPTIBLE:
678 		case RXRPC_PREINTERRUPTIBLE:
679 			set_current_state(TASK_INTERRUPTIBLE);
680 			break;
681 		case RXRPC_UNINTERRUPTIBLE:
682 		default:
683 			set_current_state(TASK_UNINTERRUPTIBLE);
684 			break;
685 		}
686 		if (READ_ONCE(call->state) != RXRPC_CALL_CLIENT_AWAIT_CONN)
687 			break;
688 		if ((call->interruptibility == RXRPC_INTERRUPTIBLE ||
689 		     call->interruptibility == RXRPC_PREINTERRUPTIBLE) &&
690 		    signal_pending(current)) {
691 			ret = -ERESTARTSYS;
692 			break;
693 		}
694 		schedule();
695 	}
696 	remove_wait_queue(&call->waitq, &myself);
697 	__set_current_state(TASK_RUNNING);
698 
699 out:
700 	_leave(" = %d", ret);
701 	return ret;
702 }
703 
704 /*
705  * find a connection for a call
706  * - called in process context with IRQs enabled
707  */
rxrpc_connect_call(struct rxrpc_sock * rx,struct rxrpc_call * call,struct rxrpc_conn_parameters * cp,struct sockaddr_rxrpc * srx,gfp_t gfp)708 int rxrpc_connect_call(struct rxrpc_sock *rx,
709 		       struct rxrpc_call *call,
710 		       struct rxrpc_conn_parameters *cp,
711 		       struct sockaddr_rxrpc *srx,
712 		       gfp_t gfp)
713 {
714 	struct rxrpc_bundle *bundle;
715 	struct rxrpc_net *rxnet = cp->local->rxnet;
716 	int ret = 0;
717 
718 	_enter("{%d,%lx},", call->debug_id, call->user_call_ID);
719 
720 	rxrpc_discard_expired_client_conns(&rxnet->client_conn_reaper);
721 
722 	bundle = rxrpc_prep_call(rx, call, cp, srx, gfp);
723 	if (IS_ERR(bundle)) {
724 		ret = PTR_ERR(bundle);
725 		goto out;
726 	}
727 
728 	if (call->state == RXRPC_CALL_CLIENT_AWAIT_CONN) {
729 		ret = rxrpc_wait_for_channel(bundle, call, gfp);
730 		if (ret < 0)
731 			goto wait_failed;
732 	}
733 
734 granted_channel:
735 	/* Paired with the write barrier in rxrpc_activate_one_channel(). */
736 	smp_rmb();
737 
738 out_put_bundle:
739 	rxrpc_deactivate_bundle(bundle);
740 	rxrpc_put_bundle(bundle);
741 out:
742 	_leave(" = %d", ret);
743 	return ret;
744 
745 wait_failed:
746 	spin_lock(&bundle->channel_lock);
747 	list_del_init(&call->chan_wait_link);
748 	spin_unlock(&bundle->channel_lock);
749 
750 	if (call->state != RXRPC_CALL_CLIENT_AWAIT_CONN) {
751 		ret = 0;
752 		goto granted_channel;
753 	}
754 
755 	trace_rxrpc_client(call->conn, ret, rxrpc_client_chan_wait_failed);
756 	rxrpc_set_call_completion(call, RXRPC_CALL_LOCAL_ERROR, 0, ret);
757 	rxrpc_disconnect_client_call(bundle, call);
758 	goto out_put_bundle;
759 }
760 
761 /*
762  * Note that a call, and thus a connection, is about to be exposed to the
763  * world.
764  */
rxrpc_expose_client_call(struct rxrpc_call * call)765 void rxrpc_expose_client_call(struct rxrpc_call *call)
766 {
767 	unsigned int channel = call->cid & RXRPC_CHANNELMASK;
768 	struct rxrpc_connection *conn = call->conn;
769 	struct rxrpc_channel *chan = &conn->channels[channel];
770 
771 	if (!test_and_set_bit(RXRPC_CALL_EXPOSED, &call->flags)) {
772 		/* Mark the call ID as being used.  If the callNumber counter
773 		 * exceeds ~2 billion, we kill the connection after its
774 		 * outstanding calls have finished so that the counter doesn't
775 		 * wrap.
776 		 */
777 		chan->call_counter++;
778 		if (chan->call_counter >= INT_MAX)
779 			set_bit(RXRPC_CONN_DONT_REUSE, &conn->flags);
780 		trace_rxrpc_client(conn, channel, rxrpc_client_exposed);
781 	}
782 }
783 
784 /*
785  * Set the reap timer.
786  */
rxrpc_set_client_reap_timer(struct rxrpc_net * rxnet)787 static void rxrpc_set_client_reap_timer(struct rxrpc_net *rxnet)
788 {
789 	if (!rxnet->kill_all_client_conns) {
790 		unsigned long now = jiffies;
791 		unsigned long reap_at = now + rxrpc_conn_idle_client_expiry;
792 
793 		if (rxnet->live)
794 			timer_reduce(&rxnet->client_conn_reap_timer, reap_at);
795 	}
796 }
797 
798 /*
799  * Disconnect a client call.
800  */
rxrpc_disconnect_client_call(struct rxrpc_bundle * bundle,struct rxrpc_call * call)801 void rxrpc_disconnect_client_call(struct rxrpc_bundle *bundle, struct rxrpc_call *call)
802 {
803 	struct rxrpc_connection *conn;
804 	struct rxrpc_channel *chan = NULL;
805 	struct rxrpc_net *rxnet = bundle->params.local->rxnet;
806 	unsigned int channel;
807 	bool may_reuse;
808 	u32 cid;
809 
810 	_enter("c=%x", call->debug_id);
811 
812 	spin_lock(&bundle->channel_lock);
813 	set_bit(RXRPC_CALL_DISCONNECTED, &call->flags);
814 
815 	/* Calls that have never actually been assigned a channel can simply be
816 	 * discarded.
817 	 */
818 	conn = call->conn;
819 	if (!conn) {
820 		_debug("call is waiting");
821 		ASSERTCMP(call->call_id, ==, 0);
822 		ASSERT(!test_bit(RXRPC_CALL_EXPOSED, &call->flags));
823 		list_del_init(&call->chan_wait_link);
824 		goto out;
825 	}
826 
827 	cid = call->cid;
828 	channel = cid & RXRPC_CHANNELMASK;
829 	chan = &conn->channels[channel];
830 	trace_rxrpc_client(conn, channel, rxrpc_client_chan_disconnect);
831 
832 	if (rcu_access_pointer(chan->call) != call) {
833 		spin_unlock(&bundle->channel_lock);
834 		BUG();
835 	}
836 
837 	may_reuse = rxrpc_may_reuse_conn(conn);
838 
839 	/* If a client call was exposed to the world, we save the result for
840 	 * retransmission.
841 	 *
842 	 * We use a barrier here so that the call number and abort code can be
843 	 * read without needing to take a lock.
844 	 *
845 	 * TODO: Make the incoming packet handler check this and handle
846 	 * terminal retransmission without requiring access to the call.
847 	 */
848 	if (test_bit(RXRPC_CALL_EXPOSED, &call->flags)) {
849 		_debug("exposed %u,%u", call->call_id, call->abort_code);
850 		__rxrpc_disconnect_call(conn, call);
851 
852 		if (test_and_clear_bit(RXRPC_CONN_PROBING_FOR_UPGRADE, &conn->flags)) {
853 			trace_rxrpc_client(conn, channel, rxrpc_client_to_active);
854 			bundle->try_upgrade = false;
855 			if (may_reuse)
856 				rxrpc_activate_channels_locked(bundle);
857 		}
858 
859 	}
860 
861 	/* See if we can pass the channel directly to another call. */
862 	if (may_reuse && !list_empty(&bundle->waiting_calls)) {
863 		trace_rxrpc_client(conn, channel, rxrpc_client_chan_pass);
864 		rxrpc_activate_one_channel(conn, channel);
865 		goto out;
866 	}
867 
868 	/* Schedule the final ACK to be transmitted in a short while so that it
869 	 * can be skipped if we find a follow-on call.  The first DATA packet
870 	 * of the follow on call will implicitly ACK this call.
871 	 */
872 	if (call->completion == RXRPC_CALL_SUCCEEDED &&
873 	    test_bit(RXRPC_CALL_EXPOSED, &call->flags)) {
874 		unsigned long final_ack_at = jiffies + 2;
875 
876 		WRITE_ONCE(chan->final_ack_at, final_ack_at);
877 		smp_wmb(); /* vs rxrpc_process_delayed_final_acks() */
878 		set_bit(RXRPC_CONN_FINAL_ACK_0 + channel, &conn->flags);
879 		rxrpc_reduce_conn_timer(conn, final_ack_at);
880 	}
881 
882 	/* Deactivate the channel. */
883 	rcu_assign_pointer(chan->call, NULL);
884 	set_bit(conn->bundle_shift + channel, &conn->bundle->avail_chans);
885 	conn->act_chans	&= ~(1 << channel);
886 
887 	/* If no channels remain active, then put the connection on the idle
888 	 * list for a short while.  Give it a ref to stop it going away if it
889 	 * becomes unbundled.
890 	 */
891 	if (!conn->act_chans) {
892 		trace_rxrpc_client(conn, channel, rxrpc_client_to_idle);
893 		conn->idle_timestamp = jiffies;
894 
895 		rxrpc_get_connection(conn);
896 		spin_lock(&rxnet->client_conn_cache_lock);
897 		list_move_tail(&conn->cache_link, &rxnet->idle_client_conns);
898 		spin_unlock(&rxnet->client_conn_cache_lock);
899 
900 		rxrpc_set_client_reap_timer(rxnet);
901 	}
902 
903 out:
904 	spin_unlock(&bundle->channel_lock);
905 	_leave("");
906 	return;
907 }
908 
909 /*
910  * Remove a connection from a bundle.
911  */
rxrpc_unbundle_conn(struct rxrpc_connection * conn)912 static void rxrpc_unbundle_conn(struct rxrpc_connection *conn)
913 {
914 	struct rxrpc_bundle *bundle = conn->bundle;
915 	unsigned int bindex;
916 	bool need_drop = false;
917 	int i;
918 
919 	_enter("C=%x", conn->debug_id);
920 
921 	if (conn->flags & RXRPC_CONN_FINAL_ACK_MASK)
922 		rxrpc_process_delayed_final_acks(conn, true);
923 
924 	spin_lock(&bundle->channel_lock);
925 	bindex = conn->bundle_shift / RXRPC_MAXCALLS;
926 	if (bundle->conns[bindex] == conn) {
927 		_debug("clear slot %u", bindex);
928 		bundle->conns[bindex] = NULL;
929 		for (i = 0; i < RXRPC_MAXCALLS; i++)
930 			clear_bit(conn->bundle_shift + i, &bundle->avail_chans);
931 		need_drop = true;
932 	}
933 	spin_unlock(&bundle->channel_lock);
934 
935 	if (need_drop) {
936 		rxrpc_deactivate_bundle(bundle);
937 		rxrpc_put_connection(conn);
938 	}
939 }
940 
941 /*
942  * Drop the active count on a bundle.
943  */
rxrpc_deactivate_bundle(struct rxrpc_bundle * bundle)944 static void rxrpc_deactivate_bundle(struct rxrpc_bundle *bundle)
945 {
946 	struct rxrpc_local *local = bundle->params.local;
947 	bool need_put = false;
948 
949 	if (atomic_dec_and_lock(&bundle->active, &local->client_bundles_lock)) {
950 		if (!bundle->params.exclusive) {
951 			_debug("erase bundle");
952 			rb_erase(&bundle->local_node, &local->client_bundles);
953 			need_put = true;
954 		}
955 
956 		spin_unlock(&local->client_bundles_lock);
957 		if (need_put)
958 			rxrpc_put_bundle(bundle);
959 	}
960 }
961 
962 /*
963  * Clean up a dead client connection.
964  */
rxrpc_kill_client_conn(struct rxrpc_connection * conn)965 static void rxrpc_kill_client_conn(struct rxrpc_connection *conn)
966 {
967 	struct rxrpc_local *local = conn->params.local;
968 	struct rxrpc_net *rxnet = local->rxnet;
969 
970 	_enter("C=%x", conn->debug_id);
971 
972 	trace_rxrpc_client(conn, -1, rxrpc_client_cleanup);
973 	atomic_dec(&rxnet->nr_client_conns);
974 
975 	rxrpc_put_client_connection_id(conn);
976 	rxrpc_kill_connection(conn);
977 }
978 
979 /*
980  * Clean up a dead client connections.
981  */
rxrpc_put_client_conn(struct rxrpc_connection * conn)982 void rxrpc_put_client_conn(struct rxrpc_connection *conn)
983 {
984 	const void *here = __builtin_return_address(0);
985 	unsigned int debug_id = conn->debug_id;
986 	bool dead;
987 	int r;
988 
989 	dead = __refcount_dec_and_test(&conn->ref, &r);
990 	trace_rxrpc_conn(debug_id, rxrpc_conn_put_client, r - 1, here);
991 	if (dead)
992 		rxrpc_kill_client_conn(conn);
993 }
994 
995 /*
996  * Discard expired client connections from the idle list.  Each conn in the
997  * idle list has been exposed and holds an extra ref because of that.
998  *
999  * This may be called from conn setup or from a work item so cannot be
1000  * considered non-reentrant.
1001  */
rxrpc_discard_expired_client_conns(struct work_struct * work)1002 void rxrpc_discard_expired_client_conns(struct work_struct *work)
1003 {
1004 	struct rxrpc_connection *conn;
1005 	struct rxrpc_net *rxnet =
1006 		container_of(work, struct rxrpc_net, client_conn_reaper);
1007 	unsigned long expiry, conn_expires_at, now;
1008 	unsigned int nr_conns;
1009 
1010 	_enter("");
1011 
1012 	if (list_empty(&rxnet->idle_client_conns)) {
1013 		_leave(" [empty]");
1014 		return;
1015 	}
1016 
1017 	/* Don't double up on the discarding */
1018 	if (!spin_trylock(&rxnet->client_conn_discard_lock)) {
1019 		_leave(" [already]");
1020 		return;
1021 	}
1022 
1023 	/* We keep an estimate of what the number of conns ought to be after
1024 	 * we've discarded some so that we don't overdo the discarding.
1025 	 */
1026 	nr_conns = atomic_read(&rxnet->nr_client_conns);
1027 
1028 next:
1029 	spin_lock(&rxnet->client_conn_cache_lock);
1030 
1031 	if (list_empty(&rxnet->idle_client_conns))
1032 		goto out;
1033 
1034 	conn = list_entry(rxnet->idle_client_conns.next,
1035 			  struct rxrpc_connection, cache_link);
1036 
1037 	if (!rxnet->kill_all_client_conns) {
1038 		/* If the number of connections is over the reap limit, we
1039 		 * expedite discard by reducing the expiry timeout.  We must,
1040 		 * however, have at least a short grace period to be able to do
1041 		 * final-ACK or ABORT retransmission.
1042 		 */
1043 		expiry = rxrpc_conn_idle_client_expiry;
1044 		if (nr_conns > rxrpc_reap_client_connections)
1045 			expiry = rxrpc_conn_idle_client_fast_expiry;
1046 		if (conn->params.local->service_closed)
1047 			expiry = rxrpc_closed_conn_expiry * HZ;
1048 
1049 		conn_expires_at = conn->idle_timestamp + expiry;
1050 
1051 		now = READ_ONCE(jiffies);
1052 		if (time_after(conn_expires_at, now))
1053 			goto not_yet_expired;
1054 	}
1055 
1056 	trace_rxrpc_client(conn, -1, rxrpc_client_discard);
1057 	list_del_init(&conn->cache_link);
1058 
1059 	spin_unlock(&rxnet->client_conn_cache_lock);
1060 
1061 	rxrpc_unbundle_conn(conn);
1062 	rxrpc_put_connection(conn); /* Drop the ->cache_link ref */
1063 
1064 	nr_conns--;
1065 	goto next;
1066 
1067 not_yet_expired:
1068 	/* The connection at the front of the queue hasn't yet expired, so
1069 	 * schedule the work item for that point if we discarded something.
1070 	 *
1071 	 * We don't worry if the work item is already scheduled - it can look
1072 	 * after rescheduling itself at a later time.  We could cancel it, but
1073 	 * then things get messier.
1074 	 */
1075 	_debug("not yet");
1076 	if (!rxnet->kill_all_client_conns)
1077 		timer_reduce(&rxnet->client_conn_reap_timer, conn_expires_at);
1078 
1079 out:
1080 	spin_unlock(&rxnet->client_conn_cache_lock);
1081 	spin_unlock(&rxnet->client_conn_discard_lock);
1082 	_leave("");
1083 }
1084 
1085 /*
1086  * Preemptively destroy all the client connection records rather than waiting
1087  * for them to time out
1088  */
rxrpc_destroy_all_client_connections(struct rxrpc_net * rxnet)1089 void rxrpc_destroy_all_client_connections(struct rxrpc_net *rxnet)
1090 {
1091 	_enter("");
1092 
1093 	spin_lock(&rxnet->client_conn_cache_lock);
1094 	rxnet->kill_all_client_conns = true;
1095 	spin_unlock(&rxnet->client_conn_cache_lock);
1096 
1097 	del_timer_sync(&rxnet->client_conn_reap_timer);
1098 
1099 	if (!rxrpc_queue_work(&rxnet->client_conn_reaper))
1100 		_debug("destroy: queue failed");
1101 
1102 	_leave("");
1103 }
1104 
1105 /*
1106  * Clean up the client connections on a local endpoint.
1107  */
rxrpc_clean_up_local_conns(struct rxrpc_local * local)1108 void rxrpc_clean_up_local_conns(struct rxrpc_local *local)
1109 {
1110 	struct rxrpc_connection *conn, *tmp;
1111 	struct rxrpc_net *rxnet = local->rxnet;
1112 	LIST_HEAD(graveyard);
1113 
1114 	_enter("");
1115 
1116 	spin_lock(&rxnet->client_conn_cache_lock);
1117 
1118 	list_for_each_entry_safe(conn, tmp, &rxnet->idle_client_conns,
1119 				 cache_link) {
1120 		if (conn->params.local == local) {
1121 			trace_rxrpc_client(conn, -1, rxrpc_client_discard);
1122 			list_move(&conn->cache_link, &graveyard);
1123 		}
1124 	}
1125 
1126 	spin_unlock(&rxnet->client_conn_cache_lock);
1127 
1128 	while (!list_empty(&graveyard)) {
1129 		conn = list_entry(graveyard.next,
1130 				  struct rxrpc_connection, cache_link);
1131 		list_del_init(&conn->cache_link);
1132 		rxrpc_unbundle_conn(conn);
1133 		rxrpc_put_connection(conn);
1134 	}
1135 
1136 	_leave(" [culled]");
1137 }
1138