summaryrefslogtreecommitdiff
path: root/src/connection.c
blob: 58d86c31b01df2a9e84a0ebf82f6323cce710572 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
/*
 * Copyright (c) 2019, Redis Labs
 * All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions are met:
 *
 *   * Redistributions of source code must retain the above copyright notice,
 *     this list of conditions and the following disclaimer.
 *   * Redistributions in binary form must reproduce the above copyright
 *     notice, this list of conditions and the following disclaimer in the
 *     documentation and/or other materials provided with the distribution.
 *   * Neither the name of Redis nor the names of its contributors may be used
 *     to endorse or promote products derived from this software without
 *     specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
 * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
 * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
 * POSSIBILITY OF SUCH DAMAGE.
 */

#include "server.h"
#include "connhelpers.h"

/* The connections module provides a lean abstraction of network connections
 * to avoid direct socket and async event management across the Redis code base.
 *
 * It does NOT provide advanced connection features commonly found in similar
 * libraries such as complete in/out buffer management, throttling, etc. These
 * functions remain in networking.c.
 *
 * The primary goal is to allow transparent handling of TCP and TLS based
 * connections. To do so, connections have the following properties:
 *
 * 1. A connection may live before its corresponding socket exists.  This
 *    allows various context and configuration setting to be handled before
 *    establishing the actual connection.
 * 2. The caller may register/unregister logical read/write handlers to be
 *    called when the connection has data to read from/can accept writes.
 *    These logical handlers may or may not correspond to actual AE events,
 *    depending on the implementation (for TCP they are; for TLS they aren't).
 */

ConnectionType CT_Socket;

/* When a connection is created we must know its type already, but the
 * underlying socket may or may not exist:
 *
 * - For accepted connections, it exists as we do not model the listen/accept
 *   part; So caller calls connCreateSocket() followed by connAccept().
 * - For outgoing connections, the socket is created by the connection module
 *   itself; So caller calls connCreateSocket() followed by connConnect(),
 *   which registers a connect callback that fires on connected/error state
 *   (and after any transport level handshake was done).
 *
 * NOTE: An earlier version relied on connections being part of other structs
 * and not independently allocated. This could lead to further optimizations
 * like using container_of(), etc.  However it was discontinued in favor of
 * this approach for these reasons:
 *
 * 1. In some cases conns are created/handled outside the context of the
 * containing struct, in which case it gets a bit awkward to copy them.
 * 2. Future implementations may wish to allocate arbitrary data for the
 * connection.
 * 3. The container_of() approach is anyway risky because connections may
 * be embedded in different structs, not just client.
 */

connection *connCreateSocket() {
    connection *conn = zcalloc(sizeof(connection));
    conn->type = &CT_Socket;
    conn->fd = -1;

    return conn;
}

/* Create a new socket-type connection that is already associated with
 * an accepted connection.
 *
 * The socket is not read for I/O until connAccept() was called and
 * invoked the connection-level accept handler.
 */
connection *connCreateAcceptedSocket(int fd) {
    connection *conn = connCreateSocket();
    conn->fd = fd;
    conn->state = CONN_STATE_ACCEPTING;
    return conn;
}

static int connSocketConnect(connection *conn, const char *addr, int port, const char *src_addr,
        ConnectionCallbackFunc connect_handler) {
    int fd = anetTcpNonBlockBestEffortBindConnect(NULL,addr,port,src_addr);
    if (fd == -1) {
        conn->state = CONN_STATE_ERROR;
        conn->last_errno = errno;
        return C_ERR;
    }

    conn->fd = fd;
    conn->state = CONN_STATE_CONNECTING;

    conn->conn_handler = connect_handler;
    aeCreateFileEvent(server.el, conn->fd, AE_WRITABLE,
            conn->type->ae_handler, conn);

    return C_OK;
}

/* Returns true if a write handler is registered */
int connHasWriteHandler(connection *conn) {
    return conn->write_handler != NULL;
}

/* Returns true if a read handler is registered */
int connHasReadHandler(connection *conn) {
    return conn->read_handler != NULL;
}

/* Associate a private data pointer with the connection */
void connSetPrivateData(connection *conn, void *data) {
    conn->private_data = data;
}

/* Get the associated private data pointer */
void *connGetPrivateData(connection *conn) {
    return conn->private_data;
}

/* ------ Pure socket connections ------- */

/* A very incomplete list of implementation-specific calls.  Much of the above shall
 * move here as we implement additional connection types.
 */

/* Close the connection and free resources. */
static void connSocketClose(connection *conn) {
    if (conn->fd != -1) {
        aeDeleteFileEvent(server.el,conn->fd,AE_READABLE);
        aeDeleteFileEvent(server.el,conn->fd,AE_WRITABLE);
        close(conn->fd);
        conn->fd = -1;
    }

    /* If called from within a handler, schedule the close but
     * keep the connection until the handler returns.
     */
    if (conn->flags & CONN_FLAG_IN_HANDLER) {
        conn->flags |= CONN_FLAG_CLOSE_SCHEDULED;
        return;
    }

    zfree(conn);
}

static int connSocketWrite(connection *conn, const void *data, size_t data_len) {
    int ret = write(conn->fd, data, data_len);
    if (ret < 0 && errno != EAGAIN) {
        conn->last_errno = errno;
        conn->state = CONN_STATE_ERROR;
    }

    return ret;
}

static int connSocketRead(connection *conn, void *buf, size_t buf_len) {
    int ret = read(conn->fd, buf, buf_len);
    if (!ret) {
        conn->state = CONN_STATE_CLOSED;
    } else if (ret < 0 && errno != EAGAIN) {
        conn->last_errno = errno;
        conn->state = CONN_STATE_ERROR;
    }

    return ret;
}

static int connSocketAccept(connection *conn, ConnectionCallbackFunc accept_handler) {
    if (conn->state != CONN_STATE_ACCEPTING) return C_ERR;
    conn->state = CONN_STATE_CONNECTED;
    if (!callHandler(conn, accept_handler)) return C_ERR;
    return C_OK;
}

/* Register a write handler, to be called when the connection is writable.
 * If NULL, the existing handler is removed.
 *
 * The barrier flag indicates a write barrier is requested, resulting with
 * CONN_FLAG_WRITE_BARRIER set. This will ensure that the write handler is
 * always called before and not after the read handler in a single event
 * loop.
 */
static int connSocketSetWriteHandler(connection *conn, ConnectionCallbackFunc func, int barrier) {
    if (func == conn->write_handler) return C_OK;

    conn->write_handler = func;
    if (barrier)
        conn->flags |= CONN_FLAG_WRITE_BARRIER;
    else
        conn->flags &= ~CONN_FLAG_WRITE_BARRIER;
    if (!conn->write_handler)
        aeDeleteFileEvent(server.el,conn->fd,AE_WRITABLE);
    else
        if (aeCreateFileEvent(server.el,conn->fd,AE_WRITABLE,
                    conn->type->ae_handler,conn) == AE_ERR) return C_ERR;
    return C_OK;
}

/* Register a read handler, to be called when the connection is readable.
 * If NULL, the existing handler is removed.
 */
static int connSocketSetReadHandler(connection *conn, ConnectionCallbackFunc func) {
    if (func == conn->read_handler) return C_OK;

    conn->read_handler = func;
    if (!conn->read_handler)
        aeDeleteFileEvent(server.el,conn->fd,AE_READABLE);
    else
        if (aeCreateFileEvent(server.el,conn->fd,
                    AE_READABLE,conn->type->ae_handler,conn) == AE_ERR) return C_ERR;
    return C_OK;
}

static const char *connSocketGetLastError(connection *conn) {
    return strerror(conn->last_errno);
}

static void connSocketEventHandler(struct aeEventLoop *el, int fd, void *clientData, int mask)
{
    UNUSED(el);
    UNUSED(fd);
    connection *conn = clientData;

    if (conn->state == CONN_STATE_CONNECTING &&
            (mask & AE_WRITABLE) && conn->conn_handler) {

        if (connGetSocketError(conn)) {
            conn->last_errno = errno;
            conn->state = CONN_STATE_ERROR;
        } else {
            conn->state = CONN_STATE_CONNECTED;
        }

        if (!conn->write_handler) aeDeleteFileEvent(server.el,conn->fd,AE_WRITABLE);

        if (!callHandler(conn, conn->conn_handler)) return;
        conn->conn_handler = NULL;
    }

    /* Normally we execute the readable event first, and the writable
     * event later. This is useful as sometimes we may be able
     * to serve the reply of a query immediately after processing the
     * query.
     *
     * However if WRITE_BARRIER is set in the mask, our application is
     * asking us to do the reverse: never fire the writable event
     * after the readable. In such a case, we invert the calls.
     * This is useful when, for instance, we want to do things
     * in the beforeSleep() hook, like fsync'ing a file to disk,
     * before replying to a client. */
    int invert = conn->flags & CONN_FLAG_WRITE_BARRIER;

    int call_write = (mask & AE_WRITABLE) && conn->write_handler;
    int call_read = (mask & AE_READABLE) && conn->read_handler;

    /* Handle normal I/O flows */
    if (!invert && call_read) {
        if (!callHandler(conn, conn->read_handler)) return;
    }
    /* Fire the writable event. */
    if (call_write) {
        if (!callHandler(conn, conn->write_handler)) return;
    }
    /* If we have to invert the call, fire the readable event now
     * after the writable one. */
    if (invert && call_read) {
        if (!callHandler(conn, conn->read_handler)) return;
    }
}

static int connSocketBlockingConnect(connection *conn, const char *addr, int port, long long timeout) {
    int fd = anetTcpNonBlockConnect(NULL,addr,port);
    if (fd == -1) {
        conn->state = CONN_STATE_ERROR;
        conn->last_errno = errno;
        return C_ERR;
    }

    if ((aeWait(fd, AE_WRITABLE, timeout) & AE_WRITABLE) == 0) {
        conn->state = CONN_STATE_ERROR;
        conn->last_errno = ETIMEDOUT;
    }

    conn->fd = fd;
    conn->state = CONN_STATE_CONNECTED;
    return C_OK;
}

/* Connection-based versions of syncio.c functions.
 * NOTE: This should ideally be refactored out in favor of pure async work.
 */

static ssize_t connSocketSyncWrite(connection *conn, char *ptr, ssize_t size, long long timeout) {
    return syncWrite(conn->fd, ptr, size, timeout);
}

static ssize_t connSocketSyncRead(connection *conn, char *ptr, ssize_t size, long long timeout) {
    return syncRead(conn->fd, ptr, size, timeout);
}

static ssize_t connSocketSyncReadLine(connection *conn, char *ptr, ssize_t size, long long timeout) {
    return syncReadLine(conn->fd, ptr, size, timeout);
}


ConnectionType CT_Socket = {
    .ae_handler = connSocketEventHandler,
    .close = connSocketClose,
    .write = connSocketWrite,
    .read = connSocketRead,
    .accept = connSocketAccept,
    .connect = connSocketConnect,
    .set_write_handler = connSocketSetWriteHandler,
    .set_read_handler = connSocketSetReadHandler,
    .get_last_error = connSocketGetLastError,
    .blocking_connect = connSocketBlockingConnect,
    .sync_write = connSocketSyncWrite,
    .sync_read = connSocketSyncRead,
    .sync_readline = connSocketSyncReadLine
};


int connGetSocketError(connection *conn) {
    int sockerr = 0;
    socklen_t errlen = sizeof(sockerr);

    if (getsockopt(conn->fd, SOL_SOCKET, SO_ERROR, &sockerr, &errlen) == -1)
        sockerr = errno;
    return sockerr;
}

int connPeerToString(connection *conn, char *ip, size_t ip_len, int *port) {
    return anetPeerToString(conn ? conn->fd : -1, ip, ip_len, port);
}

int connFormatPeer(connection *conn, char *buf, size_t buf_len) {
    return anetFormatPeer(conn ? conn->fd : -1, buf, buf_len);
}

int connSockName(connection *conn, char *ip, size_t ip_len, int *port) {
    return anetSockName(conn->fd, ip, ip_len, port);
}

int connBlock(connection *conn) {
    if (conn->fd == -1) return C_ERR;
    return anetBlock(NULL, conn->fd);
}

int connNonBlock(connection *conn) {
    if (conn->fd == -1) return C_ERR;
    return anetNonBlock(NULL, conn->fd);
}

int connEnableTcpNoDelay(connection *conn) {
    if (conn->fd == -1) return C_ERR;
    return anetEnableTcpNoDelay(NULL, conn->fd);
}

int connDisableTcpNoDelay(connection *conn) {
    if (conn->fd == -1) return C_ERR;
    return anetDisableTcpNoDelay(NULL, conn->fd);
}

int connKeepAlive(connection *conn, int interval) {
    if (conn->fd == -1) return C_ERR;
    return anetKeepAlive(NULL, conn->fd, interval);
}

int connSendTimeout(connection *conn, long long ms) {
    return anetSendTimeout(NULL, conn->fd, ms);
}

int connRecvTimeout(connection *conn, long long ms) {
    return anetRecvTimeout(NULL, conn->fd, ms);
}

int connGetState(connection *conn) {
    return conn->state;
}

/* Return a text that describes the connection, suitable for inclusion
 * in CLIENT LIST and similar outputs.
 *
 * For sockets, we always return "fd=<fdnum>" to maintain compatibility.
 */
const char *connGetInfo(connection *conn, char *buf, size_t buf_len) {
    snprintf(buf, buf_len-1, "fd=%i", conn->fd);
    return buf;
}