xref: /linux/net/rds/threads.c (revision bd628c1bed7902ec1f24ba0fe70758949146abbe)
1 /*
2  * Copyright (c) 2006, 2018 Oracle and/or its affiliates. All rights reserved.
3  *
4  * This software is available to you under a choice of one of two
5  * licenses.  You may choose to be licensed under the terms of the GNU
6  * General Public License (GPL) Version 2, available from the file
7  * COPYING in the main directory of this source tree, or the
8  * OpenIB.org BSD license below:
9  *
10  *     Redistribution and use in source and binary forms, with or
11  *     without modification, are permitted provided that the following
12  *     conditions are met:
13  *
14  *      - Redistributions of source code must retain the above
15  *        copyright notice, this list of conditions and the following
16  *        disclaimer.
17  *
18  *      - Redistributions in binary form must reproduce the above
19  *        copyright notice, this list of conditions and the following
20  *        disclaimer in the documentation and/or other materials
21  *        provided with the distribution.
22  *
23  * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
24  * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
25  * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
26  * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
27  * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
28  * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
29  * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
30  * SOFTWARE.
31  *
32  */
33 #include <linux/kernel.h>
34 #include <linux/random.h>
35 #include <linux/export.h>
36 
37 #include "rds.h"
38 
39 /*
40  * All of connection management is simplified by serializing it through
41  * work queues that execute in a connection managing thread.
42  *
43  * TCP wants to send acks through sendpage() in response to data_ready(),
44  * but it needs a process context to do so.
45  *
46  * The receive paths need to allocate but can't drop packets (!) so we have
47  * a thread around to block allocating if the receive fast path sees an
48  * allocation failure.
49  */
50 
51 /* Grand Unified Theory of connection life cycle:
52  * At any point in time, the connection can be in one of these states:
53  * DOWN, CONNECTING, UP, DISCONNECTING, ERROR
54  *
55  * The following transitions are possible:
56  *  ANY		  -> ERROR
57  *  UP		  -> DISCONNECTING
58  *  ERROR	  -> DISCONNECTING
59  *  DISCONNECTING -> DOWN
60  *  DOWN	  -> CONNECTING
61  *  CONNECTING	  -> UP
62  *
63  * Transition to state DISCONNECTING/DOWN:
64  *  -	Inside the shutdown worker; synchronizes with xmit path
65  *	through RDS_IN_XMIT, and with connection management callbacks
66  *	via c_cm_lock.
67  *
68  *	For receive callbacks, we rely on the underlying transport
69  *	(TCP, IB/RDMA) to provide the necessary synchronisation.
70  */
71 struct workqueue_struct *rds_wq;
72 EXPORT_SYMBOL_GPL(rds_wq);
73 
74 void rds_connect_path_complete(struct rds_conn_path *cp, int curr)
75 {
76 	if (!rds_conn_path_transition(cp, curr, RDS_CONN_UP)) {
77 		printk(KERN_WARNING "%s: Cannot transition to state UP, "
78 				"current state is %d\n",
79 				__func__,
80 				atomic_read(&cp->cp_state));
81 		rds_conn_path_drop(cp, false);
82 		return;
83 	}
84 
85 	rdsdebug("conn %p for %pI6c to %pI6c complete\n",
86 		 cp->cp_conn, &cp->cp_conn->c_laddr, &cp->cp_conn->c_faddr);
87 
88 	cp->cp_reconnect_jiffies = 0;
89 	set_bit(0, &cp->cp_conn->c_map_queued);
90 	rcu_read_lock();
91 	if (!rds_destroy_pending(cp->cp_conn)) {
92 		queue_delayed_work(rds_wq, &cp->cp_send_w, 0);
93 		queue_delayed_work(rds_wq, &cp->cp_recv_w, 0);
94 	}
95 	rcu_read_unlock();
96 }
97 EXPORT_SYMBOL_GPL(rds_connect_path_complete);
98 
99 void rds_connect_complete(struct rds_connection *conn)
100 {
101 	rds_connect_path_complete(&conn->c_path[0], RDS_CONN_CONNECTING);
102 }
103 EXPORT_SYMBOL_GPL(rds_connect_complete);
104 
105 /*
106  * This random exponential backoff is relied on to eventually resolve racing
107  * connects.
108  *
109  * If connect attempts race then both parties drop both connections and come
110  * here to wait for a random amount of time before trying again.  Eventually
111  * the backoff range will be so much greater than the time it takes to
112  * establish a connection that one of the pair will establish the connection
113  * before the other's random delay fires.
114  *
115  * Connection attempts that arrive while a connection is already established
116  * are also considered to be racing connects.  This lets a connection from
117  * a rebooted machine replace an existing stale connection before the transport
118  * notices that the connection has failed.
119  *
120  * We should *always* start with a random backoff; otherwise a broken connection
121  * will always take several iterations to be re-established.
122  */
123 void rds_queue_reconnect(struct rds_conn_path *cp)
124 {
125 	unsigned long rand;
126 	struct rds_connection *conn = cp->cp_conn;
127 
128 	rdsdebug("conn %p for %pI6c to %pI6c reconnect jiffies %lu\n",
129 		 conn, &conn->c_laddr, &conn->c_faddr,
130 		 cp->cp_reconnect_jiffies);
131 
132 	/* let peer with smaller addr initiate reconnect, to avoid duels */
133 	if (conn->c_trans->t_type == RDS_TRANS_TCP &&
134 	    rds_addr_cmp(&conn->c_laddr, &conn->c_faddr) >= 0)
135 		return;
136 
137 	set_bit(RDS_RECONNECT_PENDING, &cp->cp_flags);
138 	if (cp->cp_reconnect_jiffies == 0) {
139 		cp->cp_reconnect_jiffies = rds_sysctl_reconnect_min_jiffies;
140 		rcu_read_lock();
141 		if (!rds_destroy_pending(cp->cp_conn))
142 			queue_delayed_work(rds_wq, &cp->cp_conn_w, 0);
143 		rcu_read_unlock();
144 		return;
145 	}
146 
147 	get_random_bytes(&rand, sizeof(rand));
148 	rdsdebug("%lu delay %lu ceil conn %p for %pI6c -> %pI6c\n",
149 		 rand % cp->cp_reconnect_jiffies, cp->cp_reconnect_jiffies,
150 		 conn, &conn->c_laddr, &conn->c_faddr);
151 	rcu_read_lock();
152 	if (!rds_destroy_pending(cp->cp_conn))
153 		queue_delayed_work(rds_wq, &cp->cp_conn_w,
154 				   rand % cp->cp_reconnect_jiffies);
155 	rcu_read_unlock();
156 
157 	cp->cp_reconnect_jiffies = min(cp->cp_reconnect_jiffies * 2,
158 					rds_sysctl_reconnect_max_jiffies);
159 }
160 
161 void rds_connect_worker(struct work_struct *work)
162 {
163 	struct rds_conn_path *cp = container_of(work,
164 						struct rds_conn_path,
165 						cp_conn_w.work);
166 	struct rds_connection *conn = cp->cp_conn;
167 	int ret;
168 
169 	if (cp->cp_index > 0 &&
170 	    rds_addr_cmp(&cp->cp_conn->c_laddr, &cp->cp_conn->c_faddr) >= 0)
171 		return;
172 	clear_bit(RDS_RECONNECT_PENDING, &cp->cp_flags);
173 	ret = rds_conn_path_transition(cp, RDS_CONN_DOWN, RDS_CONN_CONNECTING);
174 	if (ret) {
175 		ret = conn->c_trans->conn_path_connect(cp);
176 		rdsdebug("conn %p for %pI6c to %pI6c dispatched, ret %d\n",
177 			 conn, &conn->c_laddr, &conn->c_faddr, ret);
178 
179 		if (ret) {
180 			if (rds_conn_path_transition(cp,
181 						     RDS_CONN_CONNECTING,
182 						     RDS_CONN_DOWN))
183 				rds_queue_reconnect(cp);
184 			else
185 				rds_conn_path_error(cp, "connect failed\n");
186 		}
187 	}
188 }
189 
190 void rds_send_worker(struct work_struct *work)
191 {
192 	struct rds_conn_path *cp = container_of(work,
193 						struct rds_conn_path,
194 						cp_send_w.work);
195 	int ret;
196 
197 	if (rds_conn_path_state(cp) == RDS_CONN_UP) {
198 		clear_bit(RDS_LL_SEND_FULL, &cp->cp_flags);
199 		ret = rds_send_xmit(cp);
200 		cond_resched();
201 		rdsdebug("conn %p ret %d\n", cp->cp_conn, ret);
202 		switch (ret) {
203 		case -EAGAIN:
204 			rds_stats_inc(s_send_immediate_retry);
205 			queue_delayed_work(rds_wq, &cp->cp_send_w, 0);
206 			break;
207 		case -ENOMEM:
208 			rds_stats_inc(s_send_delayed_retry);
209 			queue_delayed_work(rds_wq, &cp->cp_send_w, 2);
210 		default:
211 			break;
212 		}
213 	}
214 }
215 
216 void rds_recv_worker(struct work_struct *work)
217 {
218 	struct rds_conn_path *cp = container_of(work,
219 						struct rds_conn_path,
220 						cp_recv_w.work);
221 	int ret;
222 
223 	if (rds_conn_path_state(cp) == RDS_CONN_UP) {
224 		ret = cp->cp_conn->c_trans->recv_path(cp);
225 		rdsdebug("conn %p ret %d\n", cp->cp_conn, ret);
226 		switch (ret) {
227 		case -EAGAIN:
228 			rds_stats_inc(s_recv_immediate_retry);
229 			queue_delayed_work(rds_wq, &cp->cp_recv_w, 0);
230 			break;
231 		case -ENOMEM:
232 			rds_stats_inc(s_recv_delayed_retry);
233 			queue_delayed_work(rds_wq, &cp->cp_recv_w, 2);
234 		default:
235 			break;
236 		}
237 	}
238 }
239 
240 void rds_shutdown_worker(struct work_struct *work)
241 {
242 	struct rds_conn_path *cp = container_of(work,
243 						struct rds_conn_path,
244 						cp_down_w);
245 
246 	rds_conn_shutdown(cp);
247 }
248 
249 void rds_threads_exit(void)
250 {
251 	destroy_workqueue(rds_wq);
252 }
253 
254 int rds_threads_init(void)
255 {
256 	rds_wq = create_singlethread_workqueue("krdsd");
257 	if (!rds_wq)
258 		return -ENOMEM;
259 
260 	return 0;
261 }
262 
263 /* Compare two IPv6 addresses.  Return 0 if the two addresses are equal.
264  * Return 1 if the first is greater.  Return -1 if the second is greater.
265  */
266 int rds_addr_cmp(const struct in6_addr *addr1,
267 		 const struct in6_addr *addr2)
268 {
269 #if defined(CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS) && BITS_PER_LONG == 64
270 	const __be64 *a1, *a2;
271 	u64 x, y;
272 
273 	a1 = (__be64 *)addr1;
274 	a2 = (__be64 *)addr2;
275 
276 	if (*a1 != *a2) {
277 		if (be64_to_cpu(*a1) < be64_to_cpu(*a2))
278 			return -1;
279 		else
280 			return 1;
281 	} else {
282 		x = be64_to_cpu(*++a1);
283 		y = be64_to_cpu(*++a2);
284 		if (x < y)
285 			return -1;
286 		else if (x > y)
287 			return 1;
288 		else
289 			return 0;
290 	}
291 #else
292 	u32 a, b;
293 	int i;
294 
295 	for (i = 0; i < 4; i++) {
296 		if (addr1->s6_addr32[i] != addr2->s6_addr32[i]) {
297 			a = ntohl(addr1->s6_addr32[i]);
298 			b = ntohl(addr2->s6_addr32[i]);
299 			if (a < b)
300 				return -1;
301 			else if (a > b)
302 				return 1;
303 		}
304 	}
305 	return 0;
306 #endif
307 }
308 EXPORT_SYMBOL_GPL(rds_addr_cmp);
309