xref: /linux/net/rds/threads.c (revision c4c11dd160a8cc98f402c4e12f94b1572e822ffd)
1 /*
2  * Copyright (c) 2006 Oracle.  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_complete(struct rds_connection *conn)
75 {
76 	if (!rds_conn_transition(conn, RDS_CONN_CONNECTING, RDS_CONN_UP)) {
77 		printk(KERN_WARNING "%s: Cannot transition to state UP, "
78 				"current state is %d\n",
79 				__func__,
80 				atomic_read(&conn->c_state));
81 		atomic_set(&conn->c_state, RDS_CONN_ERROR);
82 		queue_work(rds_wq, &conn->c_down_w);
83 		return;
84 	}
85 
86 	rdsdebug("conn %p for %pI4 to %pI4 complete\n",
87 	  conn, &conn->c_laddr, &conn->c_faddr);
88 
89 	conn->c_reconnect_jiffies = 0;
90 	set_bit(0, &conn->c_map_queued);
91 	queue_delayed_work(rds_wq, &conn->c_send_w, 0);
92 	queue_delayed_work(rds_wq, &conn->c_recv_w, 0);
93 }
94 EXPORT_SYMBOL_GPL(rds_connect_complete);
95 
96 /*
97  * This random exponential backoff is relied on to eventually resolve racing
98  * connects.
99  *
100  * If connect attempts race then both parties drop both connections and come
101  * here to wait for a random amount of time before trying again.  Eventually
102  * the backoff range will be so much greater than the time it takes to
103  * establish a connection that one of the pair will establish the connection
104  * before the other's random delay fires.
105  *
106  * Connection attempts that arrive while a connection is already established
107  * are also considered to be racing connects.  This lets a connection from
108  * a rebooted machine replace an existing stale connection before the transport
109  * notices that the connection has failed.
110  *
111  * We should *always* start with a random backoff; otherwise a broken connection
112  * will always take several iterations to be re-established.
113  */
114 void rds_queue_reconnect(struct rds_connection *conn)
115 {
116 	unsigned long rand;
117 
118 	rdsdebug("conn %p for %pI4 to %pI4 reconnect jiffies %lu\n",
119 	  conn, &conn->c_laddr, &conn->c_faddr,
120 	  conn->c_reconnect_jiffies);
121 
122 	set_bit(RDS_RECONNECT_PENDING, &conn->c_flags);
123 	if (conn->c_reconnect_jiffies == 0) {
124 		conn->c_reconnect_jiffies = rds_sysctl_reconnect_min_jiffies;
125 		queue_delayed_work(rds_wq, &conn->c_conn_w, 0);
126 		return;
127 	}
128 
129 	get_random_bytes(&rand, sizeof(rand));
130 	rdsdebug("%lu delay %lu ceil conn %p for %pI4 -> %pI4\n",
131 		 rand % conn->c_reconnect_jiffies, conn->c_reconnect_jiffies,
132 		 conn, &conn->c_laddr, &conn->c_faddr);
133 	queue_delayed_work(rds_wq, &conn->c_conn_w,
134 			   rand % conn->c_reconnect_jiffies);
135 
136 	conn->c_reconnect_jiffies = min(conn->c_reconnect_jiffies * 2,
137 					rds_sysctl_reconnect_max_jiffies);
138 }
139 
140 void rds_connect_worker(struct work_struct *work)
141 {
142 	struct rds_connection *conn = container_of(work, struct rds_connection, c_conn_w.work);
143 	int ret;
144 
145 	clear_bit(RDS_RECONNECT_PENDING, &conn->c_flags);
146 	if (rds_conn_transition(conn, RDS_CONN_DOWN, RDS_CONN_CONNECTING)) {
147 		ret = conn->c_trans->conn_connect(conn);
148 		rdsdebug("conn %p for %pI4 to %pI4 dispatched, ret %d\n",
149 			conn, &conn->c_laddr, &conn->c_faddr, ret);
150 
151 		if (ret) {
152 			if (rds_conn_transition(conn, RDS_CONN_CONNECTING, RDS_CONN_DOWN))
153 				rds_queue_reconnect(conn);
154 			else
155 				rds_conn_error(conn, "RDS: connect failed\n");
156 		}
157 	}
158 }
159 
160 void rds_send_worker(struct work_struct *work)
161 {
162 	struct rds_connection *conn = container_of(work, struct rds_connection, c_send_w.work);
163 	int ret;
164 
165 	if (rds_conn_state(conn) == RDS_CONN_UP) {
166 		ret = rds_send_xmit(conn);
167 		rdsdebug("conn %p ret %d\n", conn, ret);
168 		switch (ret) {
169 		case -EAGAIN:
170 			rds_stats_inc(s_send_immediate_retry);
171 			queue_delayed_work(rds_wq, &conn->c_send_w, 0);
172 			break;
173 		case -ENOMEM:
174 			rds_stats_inc(s_send_delayed_retry);
175 			queue_delayed_work(rds_wq, &conn->c_send_w, 2);
176 		default:
177 			break;
178 		}
179 	}
180 }
181 
182 void rds_recv_worker(struct work_struct *work)
183 {
184 	struct rds_connection *conn = container_of(work, struct rds_connection, c_recv_w.work);
185 	int ret;
186 
187 	if (rds_conn_state(conn) == RDS_CONN_UP) {
188 		ret = conn->c_trans->recv(conn);
189 		rdsdebug("conn %p ret %d\n", conn, ret);
190 		switch (ret) {
191 		case -EAGAIN:
192 			rds_stats_inc(s_recv_immediate_retry);
193 			queue_delayed_work(rds_wq, &conn->c_recv_w, 0);
194 			break;
195 		case -ENOMEM:
196 			rds_stats_inc(s_recv_delayed_retry);
197 			queue_delayed_work(rds_wq, &conn->c_recv_w, 2);
198 		default:
199 			break;
200 		}
201 	}
202 }
203 
204 void rds_shutdown_worker(struct work_struct *work)
205 {
206 	struct rds_connection *conn = container_of(work, struct rds_connection, c_down_w);
207 
208 	rds_conn_shutdown(conn);
209 }
210 
211 void rds_threads_exit(void)
212 {
213 	destroy_workqueue(rds_wq);
214 }
215 
216 int rds_threads_init(void)
217 {
218 	rds_wq = create_singlethread_workqueue("krdsd");
219 	if (!rds_wq)
220 		return -ENOMEM;
221 
222 	return 0;
223 }
224