File: thread.c

package info (click to toggle)
memcached 1.5.6-1.1
  • links: PTS, VCS
  • area: main
  • in suites: bullseye, buster, sid
  • size: 2,696 kB
  • sloc: ansic: 16,931; perl: 5,523; sh: 4,302; xml: 1,638; python: 393; makefile: 170
file content (819 lines) | stat: -rw-r--r-- 22,202 bytes parent folder | download
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
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
/* -*- Mode: C; tab-width: 4; c-basic-offset: 4; indent-tabs-mode: nil -*- */
/*
 * Thread management for memcached.
 */
#include "memcached.h"
#ifdef EXTSTORE
#include "storage.h"
#endif
#include <assert.h>
#include <stdio.h>
#include <errno.h>
#include <stdlib.h>
#include <string.h>
#include <pthread.h>

#ifdef __sun
#include <atomic.h>
#endif

#define ITEMS_PER_ALLOC 64

/* An item in the connection queue. */
enum conn_queue_item_modes {
    queue_new_conn,   /* brand new connection. */
    queue_redispatch, /* redispatching from side thread */
};
typedef struct conn_queue_item CQ_ITEM;
struct conn_queue_item {
    int               sfd;
    enum conn_states  init_state;
    int               event_flags;
    int               read_buffer_size;
    enum network_transport     transport;
    enum conn_queue_item_modes mode;
    conn *c;
    CQ_ITEM          *next;
};

/* A connection queue. */
typedef struct conn_queue CQ;
struct conn_queue {
    CQ_ITEM *head;
    CQ_ITEM *tail;
    pthread_mutex_t lock;
};

/* Locks for cache LRU operations */
pthread_mutex_t lru_locks[POWER_LARGEST];

/* Connection lock around accepting new connections */
pthread_mutex_t conn_lock = PTHREAD_MUTEX_INITIALIZER;

#if !defined(HAVE_GCC_ATOMICS) && !defined(__sun)
pthread_mutex_t atomics_mutex = PTHREAD_MUTEX_INITIALIZER;
#endif

/* Lock for global stats */
static pthread_mutex_t stats_lock = PTHREAD_MUTEX_INITIALIZER;

/* Lock to cause worker threads to hang up after being woken */
static pthread_mutex_t worker_hang_lock;

/* Free list of CQ_ITEM structs */
static CQ_ITEM *cqi_freelist;
static pthread_mutex_t cqi_freelist_lock;

static pthread_mutex_t *item_locks;
/* size of the item lock hash table */
static uint32_t item_lock_count;
unsigned int item_lock_hashpower;
#define hashsize(n) ((unsigned long int)1<<(n))
#define hashmask(n) (hashsize(n)-1)

/*
 * Each libevent instance has a wakeup pipe, which other threads
 * can use to signal that they've put a new connection on its queue.
 */
static LIBEVENT_THREAD *threads;

/*
 * Number of worker threads that have finished setting themselves up.
 */
static int init_count = 0;
static pthread_mutex_t init_lock;
static pthread_cond_t init_cond;


static void thread_libevent_process(int fd, short which, void *arg);

/* item_lock() must be held for an item before any modifications to either its
 * associated hash bucket, or the structure itself.
 * LRU modifications must hold the item lock, and the LRU lock.
 * LRU's accessing items must item_trylock() before modifying an item.
 * Items accessible from an LRU must not be freed or modified
 * without first locking and removing from the LRU.
 */

void item_lock(uint32_t hv) {
    mutex_lock(&item_locks[hv & hashmask(item_lock_hashpower)]);
}

void *item_trylock(uint32_t hv) {
    pthread_mutex_t *lock = &item_locks[hv & hashmask(item_lock_hashpower)];
    if (pthread_mutex_trylock(lock) == 0) {
        return lock;
    }
    return NULL;
}

void item_trylock_unlock(void *lock) {
    mutex_unlock((pthread_mutex_t *) lock);
}

void item_unlock(uint32_t hv) {
    mutex_unlock(&item_locks[hv & hashmask(item_lock_hashpower)]);
}

static void wait_for_thread_registration(int nthreads) {
    while (init_count < nthreads) {
        pthread_cond_wait(&init_cond, &init_lock);
    }
}

static void register_thread_initialized(void) {
    pthread_mutex_lock(&init_lock);
    init_count++;
    pthread_cond_signal(&init_cond);
    pthread_mutex_unlock(&init_lock);
    /* Force worker threads to pile up if someone wants us to */
    pthread_mutex_lock(&worker_hang_lock);
    pthread_mutex_unlock(&worker_hang_lock);
}

/* Must not be called with any deeper locks held */
void pause_threads(enum pause_thread_types type) {
    char buf[1];
    int i;

    buf[0] = 0;
    switch (type) {
        case PAUSE_ALL_THREADS:
            lru_maintainer_pause();
            slabs_rebalancer_pause();
            lru_crawler_pause();
#ifdef EXTSTORE
            storage_compact_pause();
#endif
        case PAUSE_WORKER_THREADS:
            buf[0] = 'p';
            pthread_mutex_lock(&worker_hang_lock);
            break;
        case RESUME_ALL_THREADS:
            lru_maintainer_resume();
            slabs_rebalancer_resume();
            lru_crawler_resume();
#ifdef EXTSTORE
            storage_compact_resume();
#endif
        case RESUME_WORKER_THREADS:
            pthread_mutex_unlock(&worker_hang_lock);
            break;
        default:
            fprintf(stderr, "Unknown lock type: %d\n", type);
            assert(1 == 0);
            break;
    }

    /* Only send a message if we have one. */
    if (buf[0] == 0) {
        return;
    }

    pthread_mutex_lock(&init_lock);
    init_count = 0;
    for (i = 0; i < settings.num_threads; i++) {
        if (write(threads[i].notify_send_fd, buf, 1) != 1) {
            perror("Failed writing to notify pipe");
            /* TODO: This is a fatal problem. Can it ever happen temporarily? */
        }
    }
    wait_for_thread_registration(settings.num_threads);
    pthread_mutex_unlock(&init_lock);
}

/*
 * Initializes a connection queue.
 */
static void cq_init(CQ *cq) {
    pthread_mutex_init(&cq->lock, NULL);
    cq->head = NULL;
    cq->tail = NULL;
}

/*
 * Looks for an item on a connection queue, but doesn't block if there isn't
 * one.
 * Returns the item, or NULL if no item is available
 */
static CQ_ITEM *cq_pop(CQ *cq) {
    CQ_ITEM *item;

    pthread_mutex_lock(&cq->lock);
    item = cq->head;
    if (NULL != item) {
        cq->head = item->next;
        if (NULL == cq->head)
            cq->tail = NULL;
    }
    pthread_mutex_unlock(&cq->lock);

    return item;
}

/*
 * Adds an item to a connection queue.
 */
static void cq_push(CQ *cq, CQ_ITEM *item) {
    item->next = NULL;

    pthread_mutex_lock(&cq->lock);
    if (NULL == cq->tail)
        cq->head = item;
    else
        cq->tail->next = item;
    cq->tail = item;
    pthread_mutex_unlock(&cq->lock);
}

/*
 * Returns a fresh connection queue item.
 */
static CQ_ITEM *cqi_new(void) {
    CQ_ITEM *item = NULL;
    pthread_mutex_lock(&cqi_freelist_lock);
    if (cqi_freelist) {
        item = cqi_freelist;
        cqi_freelist = item->next;
    }
    pthread_mutex_unlock(&cqi_freelist_lock);

    if (NULL == item) {
        int i;

        /* Allocate a bunch of items at once to reduce fragmentation */
        item = malloc(sizeof(CQ_ITEM) * ITEMS_PER_ALLOC);
        if (NULL == item) {
            STATS_LOCK();
            stats.malloc_fails++;
            STATS_UNLOCK();
            return NULL;
        }

        /*
         * Link together all the new items except the first one
         * (which we'll return to the caller) for placement on
         * the freelist.
         */
        for (i = 2; i < ITEMS_PER_ALLOC; i++)
            item[i - 1].next = &item[i];

        pthread_mutex_lock(&cqi_freelist_lock);
        item[ITEMS_PER_ALLOC - 1].next = cqi_freelist;
        cqi_freelist = &item[1];
        pthread_mutex_unlock(&cqi_freelist_lock);
    }

    return item;
}


/*
 * Frees a connection queue item (adds it to the freelist.)
 */
static void cqi_free(CQ_ITEM *item) {
    pthread_mutex_lock(&cqi_freelist_lock);
    item->next = cqi_freelist;
    cqi_freelist = item;
    pthread_mutex_unlock(&cqi_freelist_lock);
}


/*
 * Creates a worker thread.
 */
static void create_worker(void *(*func)(void *), void *arg) {
    pthread_attr_t  attr;
    int             ret;

    pthread_attr_init(&attr);

    if ((ret = pthread_create(&((LIBEVENT_THREAD*)arg)->thread_id, &attr, func, arg)) != 0) {
        fprintf(stderr, "Can't create thread: %s\n",
                strerror(ret));
        exit(1);
    }
}

/*
 * Sets whether or not we accept new connections.
 */
void accept_new_conns(const bool do_accept) {
    pthread_mutex_lock(&conn_lock);
    do_accept_new_conns(do_accept);
    pthread_mutex_unlock(&conn_lock);
}
/****************************** LIBEVENT THREADS *****************************/

/*
 * Set up a thread's information.
 */
static void setup_thread(LIBEVENT_THREAD *me) {
#if defined(LIBEVENT_VERSION_NUMBER) && LIBEVENT_VERSION_NUMBER >= 0x02000101
    struct event_config *ev_config;
    ev_config = event_config_new();
    event_config_set_flag(ev_config, EVENT_BASE_FLAG_NOLOCK);
    me->base = event_base_new_with_config(ev_config);
    event_config_free(ev_config);
#else
    me->base = event_init();
#endif

    if (! me->base) {
        fprintf(stderr, "Can't allocate event base\n");
        exit(1);
    }

    /* Listen for notifications from other threads */
    event_set(&me->notify_event, me->notify_receive_fd,
              EV_READ | EV_PERSIST, thread_libevent_process, me);
    event_base_set(me->base, &me->notify_event);

    if (event_add(&me->notify_event, 0) == -1) {
        fprintf(stderr, "Can't monitor libevent notify pipe\n");
        exit(1);
    }

    me->new_conn_queue = malloc(sizeof(struct conn_queue));
    if (me->new_conn_queue == NULL) {
        perror("Failed to allocate memory for connection queue");
        exit(EXIT_FAILURE);
    }
    cq_init(me->new_conn_queue);

    if (pthread_mutex_init(&me->stats.mutex, NULL) != 0) {
        perror("Failed to initialize mutex");
        exit(EXIT_FAILURE);
    }

    me->suffix_cache = cache_create("suffix", SUFFIX_SIZE, sizeof(char*),
                                    NULL, NULL);
    if (me->suffix_cache == NULL) {
        fprintf(stderr, "Failed to create suffix cache\n");
        exit(EXIT_FAILURE);
    }
#ifdef EXTSTORE
    me->io_cache = cache_create("io", sizeof(io_wrap), sizeof(char*), NULL, NULL);
    if (me->io_cache == NULL) {
        fprintf(stderr, "Failed to create IO object cache\n");
        exit(EXIT_FAILURE);
    }
#endif
}

/*
 * Worker thread: main event loop
 */
static void *worker_libevent(void *arg) {
    LIBEVENT_THREAD *me = arg;

    /* Any per-thread setup can happen here; memcached_thread_init() will block until
     * all threads have finished initializing.
     */
    me->l = logger_create();
    me->lru_bump_buf = item_lru_bump_buf_create();
    if (me->l == NULL || me->lru_bump_buf == NULL) {
        abort();
    }

    if (settings.drop_privileges) {
        drop_worker_privileges();
    }

    register_thread_initialized();

    event_base_loop(me->base, 0);

    event_base_free(me->base);
    return NULL;
}


/*
 * Processes an incoming "handle a new connection" item. This is called when
 * input arrives on the libevent wakeup pipe.
 */
static void thread_libevent_process(int fd, short which, void *arg) {
    LIBEVENT_THREAD *me = arg;
    CQ_ITEM *item;
    char buf[1];
    conn *c;
    unsigned int timeout_fd;

    if (read(fd, buf, 1) != 1) {
        if (settings.verbose > 0)
            fprintf(stderr, "Can't read from libevent pipe\n");
        return;
    }

    switch (buf[0]) {
    case 'c':
        item = cq_pop(me->new_conn_queue);

        if (NULL == item) {
            break;
        }
        switch (item->mode) {
            case queue_new_conn:
                c = conn_new(item->sfd, item->init_state, item->event_flags,
                                   item->read_buffer_size, item->transport,
                                   me->base);
                if (c == NULL) {
                    if (IS_UDP(item->transport)) {
                        fprintf(stderr, "Can't listen for events on UDP socket\n");
                        exit(1);
                    } else {
                        if (settings.verbose > 0) {
                            fprintf(stderr, "Can't listen for events on fd %d\n",
                                item->sfd);
                        }
                        close(item->sfd);
                    }
                } else {
                    c->thread = me;
                }
                break;

            case queue_redispatch:
                conn_worker_readd(item->c);
                break;
        }
        cqi_free(item);
        break;
    /* we were told to pause and report in */
    case 'p':
        register_thread_initialized();
        break;
    /* a client socket timed out */
    case 't':
        if (read(fd, &timeout_fd, sizeof(timeout_fd)) != sizeof(timeout_fd)) {
            if (settings.verbose > 0)
                fprintf(stderr, "Can't read timeout fd from libevent pipe\n");
            return;
        }
        conn_close_idle(conns[timeout_fd]);
        break;
    }
}

/* Which thread we assigned a connection to most recently. */
static int last_thread = -1;

/*
 * Dispatches a new connection to another thread. This is only ever called
 * from the main thread, either during initialization (for UDP) or because
 * of an incoming connection.
 */
void dispatch_conn_new(int sfd, enum conn_states init_state, int event_flags,
                       int read_buffer_size, enum network_transport transport) {
    CQ_ITEM *item = cqi_new();
    char buf[1];
    if (item == NULL) {
        close(sfd);
        /* given that malloc failed this may also fail, but let's try */
        fprintf(stderr, "Failed to allocate memory for connection object\n");
        return ;
    }

    int tid = (last_thread + 1) % settings.num_threads;

    LIBEVENT_THREAD *thread = threads + tid;

    last_thread = tid;

    item->sfd = sfd;
    item->init_state = init_state;
    item->event_flags = event_flags;
    item->read_buffer_size = read_buffer_size;
    item->transport = transport;
    item->mode = queue_new_conn;

    cq_push(thread->new_conn_queue, item);

    MEMCACHED_CONN_DISPATCH(sfd, thread->thread_id);
    buf[0] = 'c';
    if (write(thread->notify_send_fd, buf, 1) != 1) {
        perror("Writing to thread notify pipe");
    }
}

/*
 * Re-dispatches a connection back to the original thread. Can be called from
 * any side thread borrowing a connection.
 */
void redispatch_conn(conn *c) {
    CQ_ITEM *item = cqi_new();
    char buf[1];
    if (item == NULL) {
        /* Can't cleanly redispatch connection. close it forcefully. */
        c->state = conn_closed;
        close(c->sfd);
        return;
    }
    LIBEVENT_THREAD *thread = c->thread;
    item->sfd = c->sfd;
    item->init_state = conn_new_cmd;
    item->c = c;
    item->mode = queue_redispatch;

    cq_push(thread->new_conn_queue, item);

    buf[0] = 'c';
    if (write(thread->notify_send_fd, buf, 1) != 1) {
        perror("Writing to thread notify pipe");
    }
}

/* This misses the allow_new_conns flag :( */
void sidethread_conn_close(conn *c) {
    c->state = conn_closed;
    if (settings.verbose > 1)
        fprintf(stderr, "<%d connection closed from side thread.\n", c->sfd);
    close(c->sfd);

    STATS_LOCK();
    stats_state.curr_conns--;
    STATS_UNLOCK();

    return;
}

/********************************* ITEM ACCESS *******************************/

/*
 * Allocates a new item.
 */
item *item_alloc(char *key, size_t nkey, int flags, rel_time_t exptime, int nbytes) {
    item *it;
    /* do_item_alloc handles its own locks */
    it = do_item_alloc(key, nkey, flags, exptime, nbytes);
    return it;
}

/*
 * Returns an item if it hasn't been marked as expired,
 * lazy-expiring as needed.
 */
item *item_get(const char *key, const size_t nkey, conn *c, const bool do_update) {
    item *it;
    uint32_t hv;
    hv = hash(key, nkey);
    item_lock(hv);
    it = do_item_get(key, nkey, hv, c, do_update);
    item_unlock(hv);
    return it;
}

item *item_touch(const char *key, size_t nkey, uint32_t exptime, conn *c) {
    item *it;
    uint32_t hv;
    hv = hash(key, nkey);
    item_lock(hv);
    it = do_item_touch(key, nkey, exptime, hv, c);
    item_unlock(hv);
    return it;
}

/*
 * Links an item into the LRU and hashtable.
 */
int item_link(item *item) {
    int ret;
    uint32_t hv;

    hv = hash(ITEM_key(item), item->nkey);
    item_lock(hv);
    ret = do_item_link(item, hv);
    item_unlock(hv);
    return ret;
}

/*
 * Decrements the reference count on an item and adds it to the freelist if
 * needed.
 */
void item_remove(item *item) {
    uint32_t hv;
    hv = hash(ITEM_key(item), item->nkey);

    item_lock(hv);
    do_item_remove(item);
    item_unlock(hv);
}

/*
 * Replaces one item with another in the hashtable.
 * Unprotected by a mutex lock since the core server does not require
 * it to be thread-safe.
 */
int item_replace(item *old_it, item *new_it, const uint32_t hv) {
    return do_item_replace(old_it, new_it, hv);
}

/*
 * Unlinks an item from the LRU and hashtable.
 */
void item_unlink(item *item) {
    uint32_t hv;
    hv = hash(ITEM_key(item), item->nkey);
    item_lock(hv);
    do_item_unlink(item, hv);
    item_unlock(hv);
}

/*
 * Does arithmetic on a numeric item value.
 */
enum delta_result_type add_delta(conn *c, const char *key,
                                 const size_t nkey, int incr,
                                 const int64_t delta, char *buf,
                                 uint64_t *cas) {
    enum delta_result_type ret;
    uint32_t hv;

    hv = hash(key, nkey);
    item_lock(hv);
    ret = do_add_delta(c, key, nkey, incr, delta, buf, cas, hv);
    item_unlock(hv);
    return ret;
}

/*
 * Stores an item in the cache (high level, obeys set/add/replace semantics)
 */
enum store_item_type store_item(item *item, int comm, conn* c) {
    enum store_item_type ret;
    uint32_t hv;

    hv = hash(ITEM_key(item), item->nkey);
    item_lock(hv);
    ret = do_store_item(item, comm, c, hv);
    item_unlock(hv);
    return ret;
}

/******************************* GLOBAL STATS ******************************/

void STATS_LOCK() {
    pthread_mutex_lock(&stats_lock);
}

void STATS_UNLOCK() {
    pthread_mutex_unlock(&stats_lock);
}

void threadlocal_stats_reset(void) {
    int ii;
    for (ii = 0; ii < settings.num_threads; ++ii) {
        pthread_mutex_lock(&threads[ii].stats.mutex);
#define X(name) threads[ii].stats.name = 0;
        THREAD_STATS_FIELDS
#ifdef EXTSTORE
        EXTSTORE_THREAD_STATS_FIELDS
#endif
#undef X

        memset(&threads[ii].stats.slab_stats, 0,
                sizeof(threads[ii].stats.slab_stats));
        memset(&threads[ii].stats.lru_hits, 0,
                sizeof(uint64_t) * POWER_LARGEST);

        pthread_mutex_unlock(&threads[ii].stats.mutex);
    }
}

void threadlocal_stats_aggregate(struct thread_stats *stats) {
    int ii, sid;

    /* The struct has a mutex, but we can safely set the whole thing
     * to zero since it is unused when aggregating. */
    memset(stats, 0, sizeof(*stats));

    for (ii = 0; ii < settings.num_threads; ++ii) {
        pthread_mutex_lock(&threads[ii].stats.mutex);
#define X(name) stats->name += threads[ii].stats.name;
        THREAD_STATS_FIELDS
#ifdef EXTSTORE
        EXTSTORE_THREAD_STATS_FIELDS
#endif
#undef X

        for (sid = 0; sid < MAX_NUMBER_OF_SLAB_CLASSES; sid++) {
#define X(name) stats->slab_stats[sid].name += \
            threads[ii].stats.slab_stats[sid].name;
            SLAB_STATS_FIELDS
#undef X
        }

        for (sid = 0; sid < POWER_LARGEST; sid++) {
            stats->lru_hits[sid] +=
                threads[ii].stats.lru_hits[sid];
            stats->slab_stats[CLEAR_LRU(sid)].get_hits +=
                threads[ii].stats.lru_hits[sid];
        }

        pthread_mutex_unlock(&threads[ii].stats.mutex);
    }
}

void slab_stats_aggregate(struct thread_stats *stats, struct slab_stats *out) {
    int sid;

    memset(out, 0, sizeof(*out));

    for (sid = 0; sid < MAX_NUMBER_OF_SLAB_CLASSES; sid++) {
#define X(name) out->name += stats->slab_stats[sid].name;
        SLAB_STATS_FIELDS
#undef X
    }
}

/*
 * Initializes the thread subsystem, creating various worker threads.
 *
 * nthreads  Number of worker event handler threads to spawn
 */
void memcached_thread_init(int nthreads, void *arg) {
    int         i;
    int         power;

    for (i = 0; i < POWER_LARGEST; i++) {
        pthread_mutex_init(&lru_locks[i], NULL);
    }
    pthread_mutex_init(&worker_hang_lock, NULL);

    pthread_mutex_init(&init_lock, NULL);
    pthread_cond_init(&init_cond, NULL);

    pthread_mutex_init(&cqi_freelist_lock, NULL);
    cqi_freelist = NULL;

    /* Want a wide lock table, but don't waste memory */
    if (nthreads < 3) {
        power = 10;
    } else if (nthreads < 4) {
        power = 11;
    } else if (nthreads < 5) {
        power = 12;
    } else if (nthreads <= 10) {
        power = 13;
    } else if (nthreads <= 20) {
        power = 14;
    } else {
        /* 32k buckets. just under the hashpower default. */
        power = 15;
    }

    if (power >= hashpower) {
        fprintf(stderr, "Hash table power size (%d) cannot be equal to or less than item lock table (%d)\n", hashpower, power);
        fprintf(stderr, "Item lock table grows with `-t N` (worker threadcount)\n");
        fprintf(stderr, "Hash table grows with `-o hashpower=N` \n");
        exit(1);
    }

    item_lock_count = hashsize(power);
    item_lock_hashpower = power;

    item_locks = calloc(item_lock_count, sizeof(pthread_mutex_t));
    if (! item_locks) {
        perror("Can't allocate item locks");
        exit(1);
    }
    for (i = 0; i < item_lock_count; i++) {
        pthread_mutex_init(&item_locks[i], NULL);
    }

    threads = calloc(nthreads, sizeof(LIBEVENT_THREAD));
    if (! threads) {
        perror("Can't allocate thread descriptors");
        exit(1);
    }

    for (i = 0; i < nthreads; i++) {
        int fds[2];
        if (pipe(fds)) {
            perror("Can't create notify pipe");
            exit(1);
        }

        threads[i].notify_receive_fd = fds[0];
        threads[i].notify_send_fd = fds[1];
#ifdef EXTSTORE
        threads[i].storage = arg;
#endif
        setup_thread(&threads[i]);
        /* Reserve three fds for the libevent base, and two for the pipe */
        stats_state.reserved_fds += 5;
    }

    /* Create threads after we've done all the libevent setup. */
    for (i = 0; i < nthreads; i++) {
        create_worker(worker_libevent, &threads[i]);
    }

    /* Wait for all the threads to set themselves up before returning. */
    pthread_mutex_lock(&init_lock);
    wait_for_thread_registration(nthreads);
    pthread_mutex_unlock(&init_lock);
}