1 /* A simple event-driven programming library. Originally I wrote this code
2 * for the Jim's event-loop (Jim is a Tcl interpreter) but later translated
3 * it in form of a library for easy reuse.
4 *
5 * Copyright (c) 2006-2010, Salvatore Sanfilippo <antirez at gmail dot com>
6 * All rights reserved.
7 *
8 * Redistribution and use in source and binary forms, with or without
9 * modification, are permitted provided that the following conditions are met:
10 *
11 * * Redistributions of source code must retain the above copyright notice,
12 * this list of conditions and the following disclaimer.
13 * * Redistributions in binary form must reproduce the above copyright
14 * notice, this list of conditions and the following disclaimer in the
15 * documentation and/or other materials provided with the distribution.
16 * * Neither the name of Redis nor the names of its contributors may be used
17 * to endorse or promote products derived from this software without
18 * specific prior written permission.
19 *
20 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
21 * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
22 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
23 * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
24 * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
25 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
26 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
27 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
28 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
29 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
30 * POSSIBILITY OF SUCH DAMAGE.
31 */
32
33 #include <stdio.h>
34 #include <sys/time.h>
35 #include <sys/types.h>
36 #include <unistd.h>
37 #include <stdlib.h>
38 #include <poll.h>
39 #include <string.h>
40 #include <time.h>
41 #include <errno.h>
42
43 #include "ae.h"
44 #include "zmalloc.h"
45 #include "config.h"
46
47 /* Include the best multiplexing layer supported by this system.
48 * The following should be ordered by performances, descending. */
49 #ifdef HAVE_EVPORT
50 #include "ae_evport.c"
51 #else
52 #ifdef HAVE_EPOLL
53 #include "ae_epoll.c"
54 #else
55 #ifdef HAVE_KQUEUE
56 #include "ae_kqueue.c"
57 #else
58 #include "ae_select.c"
59 #endif
60 #endif
61 #endif
62
aeCreateEventLoop(int setsize)63 aeEventLoop *aeCreateEventLoop(int setsize) {
64 aeEventLoop *eventLoop;
65 int i;
66
67 if ((eventLoop = zmalloc(sizeof(*eventLoop))) == NULL) goto err;
68 eventLoop->events = zmalloc(sizeof(aeFileEvent)*setsize);
69 eventLoop->fired = zmalloc(sizeof(aeFiredEvent)*setsize);
70 if (eventLoop->events == NULL || eventLoop->fired == NULL) goto err;
71 eventLoop->setsize = setsize;
72 eventLoop->lastTime = time(NULL);
73 eventLoop->timeEventHead = NULL;
74 eventLoop->timeEventNextId = 0;
75 eventLoop->stop = 0;
76 eventLoop->maxfd = -1;
77 eventLoop->beforesleep = NULL;
78 eventLoop->aftersleep = NULL;
79 if (aeApiCreate(eventLoop) == -1) goto err;
80 /* Events with mask == AE_NONE are not set. So let's initialize the
81 * vector with it. */
82 for (i = 0; i < setsize; i++)
83 eventLoop->events[i].mask = AE_NONE;
84 return eventLoop;
85
86 err:
87 if (eventLoop) {
88 zfree(eventLoop->events);
89 zfree(eventLoop->fired);
90 zfree(eventLoop);
91 }
92 return NULL;
93 }
94
95 /* Return the current set size. */
aeGetSetSize(aeEventLoop * eventLoop)96 int aeGetSetSize(aeEventLoop *eventLoop) {
97 return eventLoop->setsize;
98 }
99
100 /* Resize the maximum set size of the event loop.
101 * If the requested set size is smaller than the current set size, but
102 * there is already a file descriptor in use that is >= the requested
103 * set size minus one, AE_ERR is returned and the operation is not
104 * performed at all.
105 *
106 * Otherwise AE_OK is returned and the operation is successful. */
aeResizeSetSize(aeEventLoop * eventLoop,int setsize)107 int aeResizeSetSize(aeEventLoop *eventLoop, int setsize) {
108 int i;
109
110 if (setsize == eventLoop->setsize) return AE_OK;
111 if (eventLoop->maxfd >= setsize) return AE_ERR;
112 if (aeApiResize(eventLoop,setsize) == -1) return AE_ERR;
113
114 eventLoop->events = zrealloc(eventLoop->events,sizeof(aeFileEvent)*setsize);
115 eventLoop->fired = zrealloc(eventLoop->fired,sizeof(aeFiredEvent)*setsize);
116 eventLoop->setsize = setsize;
117
118 /* Make sure that if we created new slots, they are initialized with
119 * an AE_NONE mask. */
120 for (i = eventLoop->maxfd+1; i < setsize; i++)
121 eventLoop->events[i].mask = AE_NONE;
122 return AE_OK;
123 }
124
aeDeleteEventLoop(aeEventLoop * eventLoop)125 void aeDeleteEventLoop(aeEventLoop *eventLoop) {
126 aeApiFree(eventLoop);
127 zfree(eventLoop->events);
128 zfree(eventLoop->fired);
129 zfree(eventLoop);
130 }
131
aeStop(aeEventLoop * eventLoop)132 void aeStop(aeEventLoop *eventLoop) {
133 eventLoop->stop = 1;
134 }
135
aeCreateFileEvent(aeEventLoop * eventLoop,int fd,int mask,aeFileProc * proc,void * clientData)136 int aeCreateFileEvent(aeEventLoop *eventLoop, int fd, int mask,
137 aeFileProc *proc, void *clientData)
138 {
139 if (fd >= eventLoop->setsize) {
140 errno = ERANGE;
141 return AE_ERR;
142 }
143 aeFileEvent *fe = &eventLoop->events[fd];
144
145 if (aeApiAddEvent(eventLoop, fd, mask) == -1)
146 return AE_ERR;
147 fe->mask |= mask;
148 if (mask & AE_READABLE) fe->rfileProc = proc;
149 if (mask & AE_WRITABLE) fe->wfileProc = proc;
150 fe->clientData = clientData;
151 if (fd > eventLoop->maxfd)
152 eventLoop->maxfd = fd;
153 return AE_OK;
154 }
155
aeDeleteFileEvent(aeEventLoop * eventLoop,int fd,int mask)156 void aeDeleteFileEvent(aeEventLoop *eventLoop, int fd, int mask)
157 {
158 if (fd >= eventLoop->setsize) return;
159 aeFileEvent *fe = &eventLoop->events[fd];
160 if (fe->mask == AE_NONE) return;
161
162 /* We want to always remove AE_BARRIER if set when AE_WRITABLE
163 * is removed. */
164 if (mask & AE_WRITABLE) mask |= AE_BARRIER;
165
166 aeApiDelEvent(eventLoop, fd, mask);
167 fe->mask = fe->mask & (~mask);
168 if (fd == eventLoop->maxfd && fe->mask == AE_NONE) {
169 /* Update the max fd */
170 int j;
171
172 for (j = eventLoop->maxfd-1; j >= 0; j--)
173 if (eventLoop->events[j].mask != AE_NONE) break;
174 eventLoop->maxfd = j;
175 }
176 }
177
aeGetFileEvents(aeEventLoop * eventLoop,int fd)178 int aeGetFileEvents(aeEventLoop *eventLoop, int fd) {
179 if (fd >= eventLoop->setsize) return 0;
180 aeFileEvent *fe = &eventLoop->events[fd];
181
182 return fe->mask;
183 }
184
aeGetTime(long * seconds,long * milliseconds)185 static void aeGetTime(long *seconds, long *milliseconds)
186 {
187 struct timeval tv;
188
189 gettimeofday(&tv, NULL);
190 *seconds = tv.tv_sec;
191 *milliseconds = tv.tv_usec/1000;
192 }
193
aeAddMillisecondsToNow(long long milliseconds,long * sec,long * ms)194 static void aeAddMillisecondsToNow(long long milliseconds, long *sec, long *ms) {
195 long cur_sec, cur_ms, when_sec, when_ms;
196
197 aeGetTime(&cur_sec, &cur_ms);
198 when_sec = cur_sec + milliseconds/1000;
199 when_ms = cur_ms + milliseconds%1000;
200 if (when_ms >= 1000) {
201 when_sec ++;
202 when_ms -= 1000;
203 }
204 *sec = when_sec;
205 *ms = when_ms;
206 }
207
aeCreateTimeEvent(aeEventLoop * eventLoop,long long milliseconds,aeTimeProc * proc,void * clientData,aeEventFinalizerProc * finalizerProc)208 long long aeCreateTimeEvent(aeEventLoop *eventLoop, long long milliseconds,
209 aeTimeProc *proc, void *clientData,
210 aeEventFinalizerProc *finalizerProc)
211 {
212 long long id = eventLoop->timeEventNextId++;
213 aeTimeEvent *te;
214
215 te = zmalloc(sizeof(*te));
216 if (te == NULL) return AE_ERR;
217 te->id = id;
218 aeAddMillisecondsToNow(milliseconds,&te->when_sec,&te->when_ms);
219 te->timeProc = proc;
220 te->finalizerProc = finalizerProc;
221 te->clientData = clientData;
222 te->prev = NULL;
223 te->next = eventLoop->timeEventHead;
224 if (te->next)
225 te->next->prev = te;
226 eventLoop->timeEventHead = te;
227 return id;
228 }
229
aeDeleteTimeEvent(aeEventLoop * eventLoop,long long id)230 int aeDeleteTimeEvent(aeEventLoop *eventLoop, long long id)
231 {
232 aeTimeEvent *te = eventLoop->timeEventHead;
233 while(te) {
234 if (te->id == id) {
235 te->id = AE_DELETED_EVENT_ID;
236 return AE_OK;
237 }
238 te = te->next;
239 }
240 return AE_ERR; /* NO event with the specified ID found */
241 }
242
243 /* Search the first timer to fire.
244 * This operation is useful to know how many time the select can be
245 * put in sleep without to delay any event.
246 * If there are no timers NULL is returned.
247 *
248 * Note that's O(N) since time events are unsorted.
249 * Possible optimizations (not needed by Redis so far, but...):
250 * 1) Insert the event in order, so that the nearest is just the head.
251 * Much better but still insertion or deletion of timers is O(N).
252 * 2) Use a skiplist to have this operation as O(1) and insertion as O(log(N)).
253 */
aeSearchNearestTimer(aeEventLoop * eventLoop)254 static aeTimeEvent *aeSearchNearestTimer(aeEventLoop *eventLoop)
255 {
256 aeTimeEvent *te = eventLoop->timeEventHead;
257 aeTimeEvent *nearest = NULL;
258
259 while(te) {
260 if (!nearest || te->when_sec < nearest->when_sec ||
261 (te->when_sec == nearest->when_sec &&
262 te->when_ms < nearest->when_ms))
263 nearest = te;
264 te = te->next;
265 }
266 return nearest;
267 }
268
269 /* Process time events */
processTimeEvents(aeEventLoop * eventLoop)270 static int processTimeEvents(aeEventLoop *eventLoop) {
271 int processed = 0;
272 aeTimeEvent *te;
273 long long maxId;
274 time_t now = time(NULL);
275
276 /* If the system clock is moved to the future, and then set back to the
277 * right value, time events may be delayed in a random way. Often this
278 * means that scheduled operations will not be performed soon enough.
279 *
280 * Here we try to detect system clock skews, and force all the time
281 * events to be processed ASAP when this happens: the idea is that
282 * processing events earlier is less dangerous than delaying them
283 * indefinitely, and practice suggests it is. */
284 if (now < eventLoop->lastTime) {
285 te = eventLoop->timeEventHead;
286 while(te) {
287 te->when_sec = 0;
288 te = te->next;
289 }
290 }
291 eventLoop->lastTime = now;
292
293 te = eventLoop->timeEventHead;
294 maxId = eventLoop->timeEventNextId-1;
295 while(te) {
296 long now_sec, now_ms;
297 long long id;
298
299 /* Remove events scheduled for deletion. */
300 if (te->id == AE_DELETED_EVENT_ID) {
301 aeTimeEvent *next = te->next;
302 if (te->prev)
303 te->prev->next = te->next;
304 else
305 eventLoop->timeEventHead = te->next;
306 if (te->next)
307 te->next->prev = te->prev;
308 if (te->finalizerProc)
309 te->finalizerProc(eventLoop, te->clientData);
310 zfree(te);
311 te = next;
312 continue;
313 }
314
315 /* Make sure we don't process time events created by time events in
316 * this iteration. Note that this check is currently useless: we always
317 * add new timers on the head, however if we change the implementation
318 * detail, this check may be useful again: we keep it here for future
319 * defense. */
320 if (te->id > maxId) {
321 te = te->next;
322 continue;
323 }
324 aeGetTime(&now_sec, &now_ms);
325 if (now_sec > te->when_sec ||
326 (now_sec == te->when_sec && now_ms >= te->when_ms))
327 {
328 int retval;
329
330 id = te->id;
331 retval = te->timeProc(eventLoop, id, te->clientData);
332 processed++;
333 if (retval != AE_NOMORE) {
334 aeAddMillisecondsToNow(retval,&te->when_sec,&te->when_ms);
335 } else {
336 te->id = AE_DELETED_EVENT_ID;
337 }
338 }
339 te = te->next;
340 }
341 return processed;
342 }
343
344 /* Process every pending time event, then every pending file event
345 * (that may be registered by time event callbacks just processed).
346 * Without special flags the function sleeps until some file event
347 * fires, or when the next time event occurs (if any).
348 *
349 * If flags is 0, the function does nothing and returns.
350 * if flags has AE_ALL_EVENTS set, all the kind of events are processed.
351 * if flags has AE_FILE_EVENTS set, file events are processed.
352 * if flags has AE_TIME_EVENTS set, time events are processed.
353 * if flags has AE_DONT_WAIT set the function returns ASAP until all
354 * if flags has AE_CALL_AFTER_SLEEP set, the aftersleep callback is called.
355 * the events that's possible to process without to wait are processed.
356 *
357 * The function returns the number of events processed. */
aeProcessEvents(aeEventLoop * eventLoop,int flags)358 int aeProcessEvents(aeEventLoop *eventLoop, int flags)
359 {
360 int processed = 0, numevents;
361
362 /* Nothing to do? return ASAP */
363 if (!(flags & AE_TIME_EVENTS) && !(flags & AE_FILE_EVENTS)) return 0;
364
365 /* Note that we want call select() even if there are no
366 * file events to process as long as we want to process time
367 * events, in order to sleep until the next time event is ready
368 * to fire. */
369 if (eventLoop->maxfd != -1 ||
370 ((flags & AE_TIME_EVENTS) && !(flags & AE_DONT_WAIT))) {
371 int j;
372 aeTimeEvent *shortest = NULL;
373 struct timeval tv, *tvp;
374
375 if (flags & AE_TIME_EVENTS && !(flags & AE_DONT_WAIT))
376 shortest = aeSearchNearestTimer(eventLoop);
377 if (shortest) {
378 long now_sec, now_ms;
379
380 aeGetTime(&now_sec, &now_ms);
381 tvp = &tv;
382
383 /* How many milliseconds we need to wait for the next
384 * time event to fire? */
385 long long ms =
386 (shortest->when_sec - now_sec)*1000 +
387 shortest->when_ms - now_ms;
388
389 if (ms > 0) {
390 tvp->tv_sec = ms/1000;
391 tvp->tv_usec = (ms % 1000)*1000;
392 } else {
393 tvp->tv_sec = 0;
394 tvp->tv_usec = 0;
395 }
396 } else {
397 /* If we have to check for events but need to return
398 * ASAP because of AE_DONT_WAIT we need to set the timeout
399 * to zero */
400 if (flags & AE_DONT_WAIT) {
401 tv.tv_sec = tv.tv_usec = 0;
402 tvp = &tv;
403 } else {
404 /* Otherwise we can block */
405 tvp = NULL; /* wait forever */
406 }
407 }
408
409 /* Call the multiplexing API, will return only on timeout or when
410 * some event fires. */
411 numevents = aeApiPoll(eventLoop, tvp);
412
413 /* After sleep callback. */
414 if (eventLoop->aftersleep != NULL && flags & AE_CALL_AFTER_SLEEP)
415 eventLoop->aftersleep(eventLoop);
416
417 for (j = 0; j < numevents; j++) {
418 aeFileEvent *fe = &eventLoop->events[eventLoop->fired[j].fd];
419 int mask = eventLoop->fired[j].mask;
420 int fd = eventLoop->fired[j].fd;
421 int fired = 0; /* Number of events fired for current fd. */
422
423 /* Normally we execute the readable event first, and the writable
424 * event laster. This is useful as sometimes we may be able
425 * to serve the reply of a query immediately after processing the
426 * query.
427 *
428 * However if AE_BARRIER is set in the mask, our application is
429 * asking us to do the reverse: never fire the writable event
430 * after the readable. In such a case, we invert the calls.
431 * This is useful when, for instance, we want to do things
432 * in the beforeSleep() hook, like fsynching a file to disk,
433 * before replying to a client. */
434 int invert = fe->mask & AE_BARRIER;
435
436 /* Note the "fe->mask & mask & ..." code: maybe an already
437 * processed event removed an element that fired and we still
438 * didn't processed, so we check if the event is still valid.
439 *
440 * Fire the readable event if the call sequence is not
441 * inverted. */
442 if (!invert && fe->mask & mask & AE_READABLE) {
443 fe->rfileProc(eventLoop,fd,fe->clientData,mask);
444 fired++;
445 }
446
447 /* Fire the writable event. */
448 if (fe->mask & mask & AE_WRITABLE) {
449 if (!fired || fe->wfileProc != fe->rfileProc) {
450 fe->wfileProc(eventLoop,fd,fe->clientData,mask);
451 fired++;
452 }
453 }
454
455 /* If we have to invert the call, fire the readable event now
456 * after the writable one. */
457 if (invert && fe->mask & mask & AE_READABLE) {
458 if (!fired || fe->wfileProc != fe->rfileProc) {
459 fe->rfileProc(eventLoop,fd,fe->clientData,mask);
460 fired++;
461 }
462 }
463
464 processed++;
465 }
466 }
467 /* Check time events */
468 if (flags & AE_TIME_EVENTS)
469 processed += processTimeEvents(eventLoop);
470
471 return processed; /* return the number of processed file/time events */
472 }
473
474 /* Wait for milliseconds until the given file descriptor becomes
475 * writable/readable/exception */
aeWait(int fd,int mask,long long milliseconds)476 int aeWait(int fd, int mask, long long milliseconds) {
477 struct pollfd pfd;
478 int retmask = 0, retval;
479
480 memset(&pfd, 0, sizeof(pfd));
481 pfd.fd = fd;
482 if (mask & AE_READABLE) pfd.events |= POLLIN;
483 if (mask & AE_WRITABLE) pfd.events |= POLLOUT;
484
485 if ((retval = poll(&pfd, 1, milliseconds))== 1) {
486 if (pfd.revents & POLLIN) retmask |= AE_READABLE;
487 if (pfd.revents & POLLOUT) retmask |= AE_WRITABLE;
488 if (pfd.revents & POLLERR) retmask |= AE_WRITABLE;
489 if (pfd.revents & POLLHUP) retmask |= AE_WRITABLE;
490 return retmask;
491 } else {
492 return retval;
493 }
494 }
495
aeMain(aeEventLoop * eventLoop)496 void aeMain(aeEventLoop *eventLoop) {
497 eventLoop->stop = 0;
498 while (!eventLoop->stop) {
499 if (eventLoop->beforesleep != NULL)
500 eventLoop->beforesleep(eventLoop);
501 aeProcessEvents(eventLoop, AE_ALL_EVENTS|AE_CALL_AFTER_SLEEP);
502 }
503 }
504
aeGetApiName(void)505 char *aeGetApiName(void) {
506 return aeApiName();
507 }
508
aeSetBeforeSleepProc(aeEventLoop * eventLoop,aeBeforeSleepProc * beforesleep)509 void aeSetBeforeSleepProc(aeEventLoop *eventLoop, aeBeforeSleepProc *beforesleep) {
510 eventLoop->beforesleep = beforesleep;
511 }
512
aeSetAfterSleepProc(aeEventLoop * eventLoop,aeBeforeSleepProc * aftersleep)513 void aeSetAfterSleepProc(aeEventLoop *eventLoop, aeBeforeSleepProc *aftersleep) {
514 eventLoop->aftersleep = aftersleep;
515 }
516