~drizzle-trunk/drizzle/development

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
/*****************************************************************************

Copyright (c) 1995, 2009, Innobase Oy. All Rights Reserved.

This program is free software; you can redistribute it and/or modify it under
the terms of the GNU General Public License as published by the Free Software
Foundation; version 2 of the License.

This program is distributed in the hope that it will be useful, but WITHOUT
ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

You should have received a copy of the GNU General Public License along with
this program; if not, write to the Free Software Foundation, Inc., 59 Temple
Place, Suite 330, Boston, MA 02111-1307 USA

*****************************************************************************/

/**************************************************//**
@file os/os0sync.c
The interface to the operating system
synchronization primitives.

Created 9/6/1995 Heikki Tuuri
*******************************************************/

#include "os0sync.h"
#ifdef UNIV_NONINL
#include "os0sync.ic"
#endif

#ifdef __WIN__
#include <windows.h>
#endif

#include "ut0mem.h"
#include "srv0start.h"
#include "ha_prototypes.h"

/* Type definition for an operating system mutex struct */
struct os_mutex_struct{
	os_event_t	event;	/*!< Used by sync0arr.c for queing threads */
	void*		handle;	/*!< OS handle to mutex */
	ulint		count;	/*!< we use this counter to check
				that the same thread does not
				recursively lock the mutex: we
				do not assume that the OS mutex
				supports recursive locking, though
				NT seems to do that */
	UT_LIST_NODE_T(os_mutex_str_t) os_mutex_list;
				/* list of all 'slow' OS mutexes created */
};

/** Mutex protecting counts and the lists of OS mutexes and events */
UNIV_INTERN os_mutex_t	os_sync_mutex;
/** TRUE if os_sync_mutex has been initialized */
static ibool		os_sync_mutex_inited	= FALSE;
/** TRUE when os_sync_free() is being executed */
static ibool		os_sync_free_called	= FALSE;

/** This is incremented by 1 in os_thread_create and decremented by 1 in
os_thread_exit */
UNIV_INTERN ulint	os_thread_count		= 0;

/** The list of all events created */
static UT_LIST_BASE_NODE_T(os_event_struct_t)	os_event_list;

/** The list of all OS 'slow' mutexes */
static UT_LIST_BASE_NODE_T(os_mutex_str_t)	os_mutex_list;

UNIV_INTERN ulint	os_event_count		= 0;
UNIV_INTERN ulint	os_mutex_count		= 0;
UNIV_INTERN ulint	os_fast_mutex_count	= 0;

/* Because a mutex is embedded inside an event and there is an
event embedded inside a mutex, on free, this generates a recursive call.
This version of the free event function doesn't acquire the global lock */
static void os_event_free_internal(os_event_t	event);

/*********************************************************//**
Initializes global event and OS 'slow' mutex lists. */
UNIV_INTERN
void
os_sync_init(void)
/*==============*/
{
	UT_LIST_INIT(os_event_list);
	UT_LIST_INIT(os_mutex_list);

	os_sync_mutex = os_mutex_create(NULL);

	os_sync_mutex_inited = TRUE;
}

/*********************************************************//**
Frees created events and OS 'slow' mutexes. */
UNIV_INTERN
void
os_sync_free(void)
/*==============*/
{
	os_event_t	event;
	os_mutex_t	mutex;

	os_sync_free_called = TRUE;
	event = UT_LIST_GET_FIRST(os_event_list);

	while (event) {

		os_event_free(event);

		event = UT_LIST_GET_FIRST(os_event_list);
	}

	mutex = UT_LIST_GET_FIRST(os_mutex_list);

	while (mutex) {
		if (mutex == os_sync_mutex) {
			/* Set the flag to FALSE so that we do not try to
			reserve os_sync_mutex any more in remaining freeing
			operations in shutdown */
			os_sync_mutex_inited = FALSE;
		}

		os_mutex_free(mutex);

		mutex = UT_LIST_GET_FIRST(os_mutex_list);
	}
	os_sync_free_called = FALSE;
}

/*********************************************************//**
Creates an event semaphore, i.e., a semaphore which may just have two
states: signaled and nonsignaled. The created event is manual reset: it
must be reset explicitly by calling sync_os_reset_event.
@return	the event handle */
UNIV_INTERN
os_event_t
os_event_create(
/*============*/
	const char*	name)	/*!< in: the name of the event, if NULL
				the event is created without a name */
{
#ifdef __WIN__
	os_event_t event;

	event = ut_malloc(sizeof(struct os_event_struct));

	event->handle = CreateEvent(NULL, /* No security attributes */
				    TRUE, /* Manual reset */
				    FALSE, /* Initial state nonsignaled */
				    (LPCTSTR) name);
	if (!event->handle) {
		fprintf(stderr,
			"InnoDB: Could not create a Windows event semaphore;"
			" Windows error %lu\n",
			(ulong) GetLastError());
	}
#else /* Unix */
	os_event_t	event;

	UT_NOT_USED(name);

	event = ut_malloc(sizeof(struct os_event_struct));

	os_fast_mutex_init(&(event->os_mutex));

	ut_a(0 == pthread_cond_init(&(event->cond_var), NULL));

	event->is_set = FALSE;

	/* We return this value in os_event_reset(), which can then be
	be used to pass to the os_event_wait_low(). The value of zero
	is reserved in os_event_wait_low() for the case when the
	caller does not want to pass any signal_count value. To
	distinguish between the two cases we initialize signal_count
	to 1 here. */
	event->signal_count = 1;
#endif /* __WIN__ */

	/* The os_sync_mutex can be NULL because during startup an event
	can be created [ because it's embedded in the mutex/rwlock ] before
	this module has been initialized */
	if (os_sync_mutex != NULL) {
		os_mutex_enter(os_sync_mutex);
	}

	/* Put to the list of events */
	UT_LIST_ADD_FIRST(os_event_list, os_event_list, event);

	os_event_count++;

	if (os_sync_mutex != NULL) {
		os_mutex_exit(os_sync_mutex);
	}

	return(event);
}

/**********************************************************//**
Sets an event semaphore to the signaled state: lets waiting threads
proceed. */
UNIV_INTERN
void
os_event_set(
/*=========*/
	os_event_t	event)	/*!< in: event to set */
{
#ifdef __WIN__
	ut_a(event);
	ut_a(SetEvent(event->handle));
#else
	ut_a(event);

	os_fast_mutex_lock(&(event->os_mutex));

	if (event->is_set) {
		/* Do nothing */
	} else {
		event->is_set = TRUE;
		event->signal_count += 1;
		ut_a(0 == pthread_cond_broadcast(&(event->cond_var)));
	}

	os_fast_mutex_unlock(&(event->os_mutex));
#endif
}

/**********************************************************//**
Resets an event semaphore to the nonsignaled state. Waiting threads will
stop to wait for the event.
The return value should be passed to os_even_wait_low() if it is desired
that this thread should not wait in case of an intervening call to
os_event_set() between this os_event_reset() and the
os_event_wait_low() call. See comments for os_event_wait_low().
@return	current signal_count. */
UNIV_INTERN
ib_int64_t
os_event_reset(
/*===========*/
	os_event_t	event)	/*!< in: event to reset */
{
	ib_int64_t	ret = 0;

#ifdef __WIN__
	ut_a(event);

	ut_a(ResetEvent(event->handle));
#else
	ut_a(event);

	os_fast_mutex_lock(&(event->os_mutex));

	if (!event->is_set) {
		/* Do nothing */
	} else {
		event->is_set = FALSE;
	}
	ret = event->signal_count;

	os_fast_mutex_unlock(&(event->os_mutex));
#endif
	return(ret);
}

/**********************************************************//**
Frees an event object, without acquiring the global lock. */
static
void
os_event_free_internal(
/*===================*/
	os_event_t	event)	/*!< in: event to free */
{
#ifdef __WIN__
	ut_a(event);

	ut_a(CloseHandle(event->handle));
#else
	ut_a(event);

	/* This is to avoid freeing the mutex twice */
	os_fast_mutex_free(&(event->os_mutex));

	ut_a(0 == pthread_cond_destroy(&(event->cond_var)));
#endif
	/* Remove from the list of events */

	UT_LIST_REMOVE(os_event_list, os_event_list, event);

	os_event_count--;

	ut_free(event);
}

/**********************************************************//**
Frees an event object. */
UNIV_INTERN
void
os_event_free(
/*==========*/
	os_event_t	event)	/*!< in: event to free */

{
#ifdef __WIN__
	ut_a(event);

	ut_a(CloseHandle(event->handle));
#else
	ut_a(event);

	os_fast_mutex_free(&(event->os_mutex));
	ut_a(0 == pthread_cond_destroy(&(event->cond_var)));
#endif
	/* Remove from the list of events */

	os_mutex_enter(os_sync_mutex);

	UT_LIST_REMOVE(os_event_list, os_event_list, event);

	os_event_count--;

	os_mutex_exit(os_sync_mutex);

	ut_free(event);
}

/**********************************************************//**
Waits for an event object until it is in the signaled state. If
srv_shutdown_state == SRV_SHUTDOWN_EXIT_THREADS this also exits the
waiting thread when the event becomes signaled (or immediately if the
event is already in the signaled state).

Typically, if the event has been signalled after the os_event_reset()
we'll return immediately because event->is_set == TRUE.
There are, however, situations (e.g.: sync_array code) where we may
lose this information. For example:

thread A calls os_event_reset()
thread B calls os_event_set()   [event->is_set == TRUE]
thread C calls os_event_reset() [event->is_set == FALSE]
thread A calls os_event_wait()  [infinite wait!]
thread C calls os_event_wait()  [infinite wait!]

Where such a scenario is possible, to avoid infinite wait, the
value returned by os_event_reset() should be passed in as
reset_sig_count. */
UNIV_INTERN
void
os_event_wait_low(
/*==============*/
	os_event_t	event,		/*!< in: event to wait */
	ib_int64_t	reset_sig_count)/*!< in: zero or the value
					returned by previous call of
					os_event_reset(). */
{
#ifdef __WIN__
	DWORD	err;

	ut_a(event);

	UT_NOT_USED(reset_sig_count);

	/* Specify an infinite time limit for waiting */
	err = WaitForSingleObject(event->handle, INFINITE);

	ut_a(err == WAIT_OBJECT_0);

	if (srv_shutdown_state == SRV_SHUTDOWN_EXIT_THREADS) {
		os_thread_exit(NULL);
	}
#else
	ib_int64_t	old_signal_count;

	os_fast_mutex_lock(&(event->os_mutex));

	if (reset_sig_count) {
		old_signal_count = reset_sig_count;
	} else {
		old_signal_count = event->signal_count;
	}

	for (;;) {
		if (event->is_set == TRUE
		    || event->signal_count != old_signal_count) {

			os_fast_mutex_unlock(&(event->os_mutex));

			if (srv_shutdown_state == SRV_SHUTDOWN_EXIT_THREADS) {

				os_thread_exit(NULL);
			}
			/* Ok, we may return */

			return;
		}

		pthread_cond_wait(&(event->cond_var), &(event->os_mutex));

		/* Solaris manual said that spurious wakeups may occur: we
		have to check if the event really has been signaled after
		we came here to wait */
	}
#endif
}

/**********************************************************//**
Waits for an event object until it is in the signaled state or
a timeout is exceeded. In Unix the timeout is always infinite.
@return	0 if success, OS_SYNC_TIME_EXCEEDED if timeout was exceeded */
UNIV_INTERN
ulint
os_event_wait_time(
/*===============*/
	os_event_t	event,	/*!< in: event to wait */
	ulint		time)	/*!< in: timeout in microseconds, or
				OS_SYNC_INFINITE_TIME */
{
#ifdef __WIN__
	DWORD	err;

	ut_a(event);

	if (time != OS_SYNC_INFINITE_TIME) {
		err = WaitForSingleObject(event->handle, (DWORD) time / 1000);
	} else {
		err = WaitForSingleObject(event->handle, INFINITE);
	}

	if (err == WAIT_OBJECT_0) {

		return(0);
	} else if (err == WAIT_TIMEOUT) {

		return(OS_SYNC_TIME_EXCEEDED);
	} else {
		ut_error;
		return(1000000); /* dummy value to eliminate compiler warn. */
	}
#else
	UT_NOT_USED(time);

	/* In Posix this is just an ordinary, infinite wait */

	os_event_wait(event);

	return(0);
#endif
}

#ifdef __WIN__
/**********************************************************//**
Waits for any event in an OS native event array. Returns if even a single
one is signaled or becomes signaled.
@return	index of the event which was signaled */
UNIV_INTERN
ulint
os_event_wait_multiple(
/*===================*/
	ulint			n,	/*!< in: number of events in the
					array */
	os_native_event_t*	native_event_array)
					/*!< in: pointer to an array of event
					handles */
{
	DWORD	index;

	ut_a(native_event_array);
	ut_a(n > 0);

	index = WaitForMultipleObjects((DWORD) n, native_event_array,
				       FALSE,	   /* Wait for any 1 event */
				       INFINITE); /* Infinite wait time
						  limit */
	ut_a(index >= WAIT_OBJECT_0);	/* NOTE: Pointless comparison */
	ut_a(index < WAIT_OBJECT_0 + n);

	if (srv_shutdown_state == SRV_SHUTDOWN_EXIT_THREADS) {
		os_thread_exit(NULL);
	}

	return(index - WAIT_OBJECT_0);
}
#endif

/*********************************************************//**
Creates an operating system mutex semaphore. Because these are slow, the
mutex semaphore of InnoDB itself (mutex_t) should be used where possible.
@return	the mutex handle */
UNIV_INTERN
os_mutex_t
os_mutex_create(
/*============*/
	const char*	name)	/*!< in: the name of the mutex, if NULL
				the mutex is created without a name */
{
#ifdef __WIN__
	HANDLE		mutex;
	os_mutex_t	mutex_str;

	mutex = CreateMutex(NULL,	/* No security attributes */
			    FALSE,		/* Initial state: no owner */
			    (LPCTSTR) name);
	ut_a(mutex);
#else
	os_fast_mutex_t*	mutex;
	os_mutex_t		mutex_str;

	UT_NOT_USED(name);

	mutex = ut_malloc(sizeof(os_fast_mutex_t));

	os_fast_mutex_init(mutex);
#endif
	mutex_str = ut_malloc(sizeof(os_mutex_str_t));

	mutex_str->handle = mutex;
	mutex_str->count = 0;
	mutex_str->event = os_event_create(NULL);

	if (UNIV_LIKELY(os_sync_mutex_inited)) {
		/* When creating os_sync_mutex itself we cannot reserve it */
		os_mutex_enter(os_sync_mutex);
	}

	UT_LIST_ADD_FIRST(os_mutex_list, os_mutex_list, mutex_str);

	os_mutex_count++;

	if (UNIV_LIKELY(os_sync_mutex_inited)) {
		os_mutex_exit(os_sync_mutex);
	}

	return(mutex_str);
}

/**********************************************************//**
Acquires ownership of a mutex semaphore. */
UNIV_INTERN
void
os_mutex_enter(
/*===========*/
	os_mutex_t	mutex)	/*!< in: mutex to acquire */
{
#ifdef __WIN__
	DWORD	err;

	ut_a(mutex);

	/* Specify infinite time limit for waiting */
	err = WaitForSingleObject(mutex->handle, INFINITE);

	ut_a(err == WAIT_OBJECT_0);

	(mutex->count)++;
	ut_a(mutex->count == 1);
#else
	os_fast_mutex_lock(mutex->handle);

	(mutex->count)++;

	ut_a(mutex->count == 1);
#endif
}

/**********************************************************//**
Releases ownership of a mutex. */
UNIV_INTERN
void
os_mutex_exit(
/*==========*/
	os_mutex_t	mutex)	/*!< in: mutex to release */
{
	ut_a(mutex);

	ut_a(mutex->count == 1);

	(mutex->count)--;
#ifdef __WIN__
	ut_a(ReleaseMutex(mutex->handle));
#else
	os_fast_mutex_unlock(mutex->handle);
#endif
}

/**********************************************************//**
Frees a mutex object. */
UNIV_INTERN
void
os_mutex_free(
/*==========*/
	os_mutex_t	mutex)	/*!< in: mutex to free */
{
	ut_a(mutex);

	if (UNIV_LIKELY(!os_sync_free_called)) {
		os_event_free_internal(mutex->event);
	}

	if (UNIV_LIKELY(os_sync_mutex_inited)) {
		os_mutex_enter(os_sync_mutex);
	}

	UT_LIST_REMOVE(os_mutex_list, os_mutex_list, mutex);

	os_mutex_count--;

	if (UNIV_LIKELY(os_sync_mutex_inited)) {
		os_mutex_exit(os_sync_mutex);
	}

#ifdef __WIN__
	ut_a(CloseHandle(mutex->handle));

	ut_free(mutex);
#else
	os_fast_mutex_free(mutex->handle);
	ut_free(mutex->handle);
	ut_free(mutex);
#endif
}

/*********************************************************//**
Initializes an operating system fast mutex semaphore. */
UNIV_INTERN
void
os_fast_mutex_init(
/*===============*/
	os_fast_mutex_t*	fast_mutex)	/*!< in: fast mutex */
{
#ifdef __WIN__
	ut_a(fast_mutex);

	InitializeCriticalSection((LPCRITICAL_SECTION) fast_mutex);
#else
	ut_a(0 == innobase_fast_mutex_init(fast_mutex));
#endif
	if (UNIV_LIKELY(os_sync_mutex_inited)) {
		/* When creating os_sync_mutex itself (in Unix) we cannot
		reserve it */

		os_mutex_enter(os_sync_mutex);
	}

	os_fast_mutex_count++;

	if (UNIV_LIKELY(os_sync_mutex_inited)) {
		os_mutex_exit(os_sync_mutex);
	}
}

/**********************************************************//**
Acquires ownership of a fast mutex. */
UNIV_INTERN
void
os_fast_mutex_lock(
/*===============*/
	os_fast_mutex_t*	fast_mutex)	/*!< in: mutex to acquire */
{
#ifdef __WIN__
	EnterCriticalSection((LPCRITICAL_SECTION) fast_mutex);
#else
	pthread_mutex_lock(fast_mutex);
#endif
}

/**********************************************************//**
Releases ownership of a fast mutex. */
UNIV_INTERN
void
os_fast_mutex_unlock(
/*=================*/
	os_fast_mutex_t*	fast_mutex)	/*!< in: mutex to release */
{
#ifdef __WIN__
	LeaveCriticalSection(fast_mutex);
#else
	pthread_mutex_unlock(fast_mutex);
#endif
}

/**********************************************************//**
Frees a mutex object. */
UNIV_INTERN
void
os_fast_mutex_free(
/*===============*/
	os_fast_mutex_t*	fast_mutex)	/*!< in: mutex to free */
{
#ifdef __WIN__
	ut_a(fast_mutex);

	DeleteCriticalSection((LPCRITICAL_SECTION) fast_mutex);
#else
	int	ret;

	ret = pthread_mutex_destroy(fast_mutex);

	if (UNIV_UNLIKELY(ret != 0)) {
		ut_print_timestamp(stderr);
		fprintf(stderr,
			"  InnoDB: error: return value %lu when calling\n"
			"InnoDB: pthread_mutex_destroy().\n", (ulint)ret);
		fprintf(stderr,
			"InnoDB: Byte contents of the pthread mutex at %p:\n",
			(void*) fast_mutex);
		ut_print_buf(stderr, fast_mutex, sizeof(os_fast_mutex_t));
		putc('\n', stderr);
	}
#endif
	if (UNIV_LIKELY(os_sync_mutex_inited)) {
		/* When freeing the last mutexes, we have
		already freed os_sync_mutex */

		os_mutex_enter(os_sync_mutex);
	}

	os_fast_mutex_count--;

	if (UNIV_LIKELY(os_sync_mutex_inited)) {
		os_mutex_exit(os_sync_mutex);
	}
}