xref: /netbsd-src/lib/libc/stdlib/jemalloc.c (revision 466a16a118933bd295a8a104f095714fadf9cf68)
1 /*	$NetBSD: jemalloc.c,v 1.19 2008/06/23 10:46:25 ad Exp $	*/
2 
3 /*-
4  * Copyright (C) 2006,2007 Jason Evans <jasone@FreeBSD.org>.
5  * All rights reserved.
6  *
7  * Redistribution and use in source and binary forms, with or without
8  * modification, are permitted provided that the following conditions
9  * are met:
10  * 1. Redistributions of source code must retain the above copyright
11  *    notice(s), this list of conditions and the following disclaimer as
12  *    the first lines of this file unmodified other than the possible
13  *    addition of one or more copyright notices.
14  * 2. Redistributions in binary form must reproduce the above copyright
15  *    notice(s), this list of conditions and the following disclaimer in
16  *    the documentation and/or other materials provided with the
17  *    distribution.
18  *
19  * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDER(S) ``AS IS'' AND ANY
20  * EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
21  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
22  * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT HOLDER(S) BE
23  * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
24  * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
25  * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR
26  * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
27  * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE
28  * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE,
29  * EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
30  *
31  *******************************************************************************
32  *
33  * This allocator implementation is designed to provide scalable performance
34  * for multi-threaded programs on multi-processor systems.  The following
35  * features are included for this purpose:
36  *
37  *   + Multiple arenas are used if there are multiple CPUs, which reduces lock
38  *     contention and cache sloshing.
39  *
40  *   + Cache line sharing between arenas is avoided for internal data
41  *     structures.
42  *
43  *   + Memory is managed in chunks and runs (chunks can be split into runs),
44  *     rather than as individual pages.  This provides a constant-time
45  *     mechanism for associating allocations with particular arenas.
46  *
47  * Allocation requests are rounded up to the nearest size class, and no record
48  * of the original request size is maintained.  Allocations are broken into
49  * categories according to size class.  Assuming runtime defaults, 4 kB pages
50  * and a 16 byte quantum, the size classes in each category are as follows:
51  *
52  *   |=====================================|
53  *   | Category | Subcategory    |    Size |
54  *   |=====================================|
55  *   | Small    | Tiny           |       2 |
56  *   |          |                |       4 |
57  *   |          |                |       8 |
58  *   |          |----------------+---------|
59  *   |          | Quantum-spaced |      16 |
60  *   |          |                |      32 |
61  *   |          |                |      48 |
62  *   |          |                |     ... |
63  *   |          |                |     480 |
64  *   |          |                |     496 |
65  *   |          |                |     512 |
66  *   |          |----------------+---------|
67  *   |          | Sub-page       |    1 kB |
68  *   |          |                |    2 kB |
69  *   |=====================================|
70  *   | Large                     |    4 kB |
71  *   |                           |    8 kB |
72  *   |                           |   12 kB |
73  *   |                           |     ... |
74  *   |                           | 1012 kB |
75  *   |                           | 1016 kB |
76  *   |                           | 1020 kB |
77  *   |=====================================|
78  *   | Huge                      |    1 MB |
79  *   |                           |    2 MB |
80  *   |                           |    3 MB |
81  *   |                           |     ... |
82  *   |=====================================|
83  *
84  * A different mechanism is used for each category:
85  *
86  *   Small : Each size class is segregated into its own set of runs.  Each run
87  *           maintains a bitmap of which regions are free/allocated.
88  *
89  *   Large : Each allocation is backed by a dedicated run.  Metadata are stored
90  *           in the associated arena chunk header maps.
91  *
92  *   Huge : Each allocation is backed by a dedicated contiguous set of chunks.
93  *          Metadata are stored in a separate red-black tree.
94  *
95  *******************************************************************************
96  */
97 
98 /* LINTLIBRARY */
99 
100 #ifdef __NetBSD__
101 #  define xutrace(a, b)		utrace("malloc", (a), (b))
102 #  define __DECONST(x, y)	((x)__UNCONST(y))
103 #  define NO_TLS
104 #else
105 #  define xutrace(a, b)		utrace((a), (b))
106 #endif	/* __NetBSD__ */
107 
108 /*
109  * MALLOC_PRODUCTION disables assertions and statistics gathering.  It also
110  * defaults the A and J runtime options to off.  These settings are appropriate
111  * for production systems.
112  */
113 #define MALLOC_PRODUCTION
114 
115 #ifndef MALLOC_PRODUCTION
116 #  define MALLOC_DEBUG
117 #endif
118 
119 #include <sys/cdefs.h>
120 /* __FBSDID("$FreeBSD: src/lib/libc/stdlib/malloc.c,v 1.147 2007/06/15 22:00:16 jasone Exp $"); */
121 __RCSID("$NetBSD: jemalloc.c,v 1.19 2008/06/23 10:46:25 ad Exp $");
122 
123 #ifdef __FreeBSD__
124 #include "libc_private.h"
125 #ifdef MALLOC_DEBUG
126 #  define _LOCK_DEBUG
127 #endif
128 #include "spinlock.h"
129 #endif
130 #include "namespace.h"
131 #include <sys/mman.h>
132 #include <sys/param.h>
133 #ifdef __FreeBSD__
134 #include <sys/stddef.h>
135 #endif
136 #include <sys/time.h>
137 #include <sys/types.h>
138 #include <sys/sysctl.h>
139 #include <sys/tree.h>
140 #include <sys/uio.h>
141 #include <sys/ktrace.h> /* Must come after several other sys/ includes. */
142 
143 #ifdef __FreeBSD__
144 #include <machine/atomic.h>
145 #include <machine/cpufunc.h>
146 #endif
147 #include <machine/vmparam.h>
148 
149 #include <errno.h>
150 #include <limits.h>
151 #include <pthread.h>
152 #include <sched.h>
153 #include <stdarg.h>
154 #include <stdbool.h>
155 #include <stdio.h>
156 #include <stdint.h>
157 #include <stdlib.h>
158 #include <string.h>
159 #include <strings.h>
160 #include <unistd.h>
161 
162 #ifdef __NetBSD__
163 #  include <reentrant.h>
164 #  include "extern.h"
165 
166 #define STRERROR_R(a, b, c)	__strerror_r(a, b, c);
167 /*
168  * A non localized version of strerror, that avoids bringing in
169  * stdio and the locale code. All the malloc messages are in English
170  * so why bother?
171  */
172 static int
173 __strerror_r(int e, char *s, size_t l)
174 {
175 	int rval;
176 	size_t slen;
177 
178 	if (e >= 0 && e < sys_nerr) {
179 		slen = strlcpy(s, sys_errlist[e], l);
180 		rval = 0;
181 	} else {
182 		slen = snprintf_ss(s, l, "Unknown error %u", e);
183 		rval = EINVAL;
184 	}
185 	return slen >= l ? ERANGE : rval;
186 }
187 #endif
188 
189 #ifdef __FreeBSD__
190 #define STRERROR_R(a, b, c)	strerror_r(a, b, c);
191 #include "un-namespace.h"
192 #endif
193 
194 /* MALLOC_STATS enables statistics calculation. */
195 #ifndef MALLOC_PRODUCTION
196 #  define MALLOC_STATS
197 #endif
198 
199 #ifdef MALLOC_DEBUG
200 #  ifdef NDEBUG
201 #    undef NDEBUG
202 #  endif
203 #else
204 #  ifndef NDEBUG
205 #    define NDEBUG
206 #  endif
207 #endif
208 #include <assert.h>
209 
210 #ifdef MALLOC_DEBUG
211    /* Disable inlining to make debugging easier. */
212 #  define inline
213 #endif
214 
215 /* Size of stack-allocated buffer passed to strerror_r(). */
216 #define	STRERROR_BUF		64
217 
218 /* Minimum alignment of allocations is 2^QUANTUM_2POW_MIN bytes. */
219 #ifdef __i386__
220 #  define QUANTUM_2POW_MIN	4
221 #  define SIZEOF_PTR_2POW	2
222 #  define USE_BRK
223 #endif
224 #ifdef __ia64__
225 #  define QUANTUM_2POW_MIN	4
226 #  define SIZEOF_PTR_2POW	3
227 #endif
228 #ifdef __alpha__
229 #  define QUANTUM_2POW_MIN	4
230 #  define SIZEOF_PTR_2POW	3
231 #  define NO_TLS
232 #endif
233 #ifdef __sparc64__
234 #  define QUANTUM_2POW_MIN	4
235 #  define SIZEOF_PTR_2POW	3
236 #  define NO_TLS
237 #endif
238 #ifdef __amd64__
239 #  define QUANTUM_2POW_MIN	4
240 #  define SIZEOF_PTR_2POW	3
241 #endif
242 #ifdef __arm__
243 #  define QUANTUM_2POW_MIN	3
244 #  define SIZEOF_PTR_2POW	2
245 #  define USE_BRK
246 #  define NO_TLS
247 #endif
248 #ifdef __powerpc__
249 #  define QUANTUM_2POW_MIN	4
250 #  define SIZEOF_PTR_2POW	2
251 #  define USE_BRK
252 #endif
253 #if defined(__sparc__) && !defined(__sparc64__)
254 #  define QUANTUM_2POW_MIN	4
255 #  define SIZEOF_PTR_2POW	2
256 #  define USE_BRK
257 #endif
258 #ifdef __vax__
259 #  define QUANTUM_2POW_MIN	4
260 #  define SIZEOF_PTR_2POW	2
261 #  define USE_BRK
262 #endif
263 #ifdef __sh__
264 #  define QUANTUM_2POW_MIN	4
265 #  define SIZEOF_PTR_2POW	2
266 #  define USE_BRK
267 #endif
268 #ifdef __m68k__
269 #  define QUANTUM_2POW_MIN	4
270 #  define SIZEOF_PTR_2POW	2
271 #  define USE_BRK
272 #endif
273 #ifdef __mips__
274 #  define QUANTUM_2POW_MIN	4
275 #  define SIZEOF_PTR_2POW	2
276 #  define USE_BRK
277 #endif
278 #ifdef __hppa__
279 #  define QUANTUM_2POW_MIN     4
280 #  define SIZEOF_PTR_2POW      2
281 #  define USE_BRK
282 #endif
283 
284 #define	SIZEOF_PTR		(1 << SIZEOF_PTR_2POW)
285 
286 /* sizeof(int) == (1 << SIZEOF_INT_2POW). */
287 #ifndef SIZEOF_INT_2POW
288 #  define SIZEOF_INT_2POW	2
289 #endif
290 
291 /* We can't use TLS in non-PIC programs, since TLS relies on loader magic. */
292 #if (!defined(PIC) && !defined(NO_TLS))
293 #  define NO_TLS
294 #endif
295 
296 /*
297  * Size and alignment of memory chunks that are allocated by the OS's virtual
298  * memory system.
299  */
300 #define	CHUNK_2POW_DEFAULT	20
301 
302 /*
303  * Maximum size of L1 cache line.  This is used to avoid cache line aliasing,
304  * so over-estimates are okay (up to a point), but under-estimates will
305  * negatively affect performance.
306  */
307 #define	CACHELINE_2POW		6
308 #define	CACHELINE		((size_t)(1 << CACHELINE_2POW))
309 
310 /* Smallest size class to support. */
311 #define	TINY_MIN_2POW		1
312 
313 /*
314  * Maximum size class that is a multiple of the quantum, but not (necessarily)
315  * a power of 2.  Above this size, allocations are rounded up to the nearest
316  * power of 2.
317  */
318 #define	SMALL_MAX_2POW_DEFAULT	9
319 #define	SMALL_MAX_DEFAULT	(1 << SMALL_MAX_2POW_DEFAULT)
320 
321 /*
322  * Maximum desired run header overhead.  Runs are sized as small as possible
323  * such that this setting is still honored, without violating other constraints.
324  * The goal is to make runs as small as possible without exceeding a per run
325  * external fragmentation threshold.
326  *
327  * Note that it is possible to set this low enough that it cannot be honored
328  * for some/all object sizes, since there is one bit of header overhead per
329  * object (plus a constant).  In such cases, this constraint is relaxed.
330  *
331  * RUN_MAX_OVRHD_RELAX specifies the maximum number of bits per region of
332  * overhead for which RUN_MAX_OVRHD is relaxed.
333  */
334 #define RUN_MAX_OVRHD		0.015
335 #define RUN_MAX_OVRHD_RELAX	1.5
336 
337 /* Put a cap on small object run size.  This overrides RUN_MAX_OVRHD. */
338 #define RUN_MAX_SMALL_2POW	15
339 #define RUN_MAX_SMALL		(1 << RUN_MAX_SMALL_2POW)
340 
341 /******************************************************************************/
342 
343 #ifdef __FreeBSD__
344 /*
345  * Mutexes based on spinlocks.  We can't use normal pthread mutexes, because
346  * they require malloc()ed memory.
347  */
348 typedef struct {
349 	spinlock_t	lock;
350 } malloc_mutex_t;
351 
352 /* Set to true once the allocator has been initialized. */
353 static bool malloc_initialized = false;
354 
355 /* Used to avoid initialization races. */
356 static malloc_mutex_t init_lock = {_SPINLOCK_INITIALIZER};
357 #else
358 #define	malloc_mutex_t	mutex_t
359 
360 /* Set to true once the allocator has been initialized. */
361 static bool malloc_initialized = false;
362 
363 /* Used to avoid initialization races. */
364 static mutex_t init_lock = MUTEX_INITIALIZER;
365 #endif
366 
367 /******************************************************************************/
368 /*
369  * Statistics data structures.
370  */
371 
372 #ifdef MALLOC_STATS
373 
374 typedef struct malloc_bin_stats_s malloc_bin_stats_t;
375 struct malloc_bin_stats_s {
376 	/*
377 	 * Number of allocation requests that corresponded to the size of this
378 	 * bin.
379 	 */
380 	uint64_t	nrequests;
381 
382 	/* Total number of runs created for this bin's size class. */
383 	uint64_t	nruns;
384 
385 	/*
386 	 * Total number of runs reused by extracting them from the runs tree for
387 	 * this bin's size class.
388 	 */
389 	uint64_t	reruns;
390 
391 	/* High-water mark for this bin. */
392 	unsigned long	highruns;
393 
394 	/* Current number of runs in this bin. */
395 	unsigned long	curruns;
396 };
397 
398 typedef struct arena_stats_s arena_stats_t;
399 struct arena_stats_s {
400 	/* Number of bytes currently mapped. */
401 	size_t		mapped;
402 
403 	/* Per-size-category statistics. */
404 	size_t		allocated_small;
405 	uint64_t	nmalloc_small;
406 	uint64_t	ndalloc_small;
407 
408 	size_t		allocated_large;
409 	uint64_t	nmalloc_large;
410 	uint64_t	ndalloc_large;
411 };
412 
413 typedef struct chunk_stats_s chunk_stats_t;
414 struct chunk_stats_s {
415 	/* Number of chunks that were allocated. */
416 	uint64_t	nchunks;
417 
418 	/* High-water mark for number of chunks allocated. */
419 	unsigned long	highchunks;
420 
421 	/*
422 	 * Current number of chunks allocated.  This value isn't maintained for
423 	 * any other purpose, so keep track of it in order to be able to set
424 	 * highchunks.
425 	 */
426 	unsigned long	curchunks;
427 };
428 
429 #endif /* #ifdef MALLOC_STATS */
430 
431 /******************************************************************************/
432 /*
433  * Chunk data structures.
434  */
435 
436 /* Tree of chunks. */
437 typedef struct chunk_node_s chunk_node_t;
438 struct chunk_node_s {
439 	/* Linkage for the chunk tree. */
440 	RB_ENTRY(chunk_node_s) link;
441 
442 	/*
443 	 * Pointer to the chunk that this tree node is responsible for.  In some
444 	 * (but certainly not all) cases, this data structure is placed at the
445 	 * beginning of the corresponding chunk, so this field may point to this
446 	 * node.
447 	 */
448 	void	*chunk;
449 
450 	/* Total chunk size. */
451 	size_t	size;
452 };
453 typedef struct chunk_tree_s chunk_tree_t;
454 RB_HEAD(chunk_tree_s, chunk_node_s);
455 
456 /******************************************************************************/
457 /*
458  * Arena data structures.
459  */
460 
461 typedef struct arena_s arena_t;
462 typedef struct arena_bin_s arena_bin_t;
463 
464 typedef struct arena_chunk_map_s arena_chunk_map_t;
465 struct arena_chunk_map_s {
466 	/* Number of pages in run. */
467 	uint32_t	npages;
468 	/*
469 	 * Position within run.  For a free run, this is POS_FREE for the first
470 	 * and last pages.  The POS_FREE special value makes it possible to
471 	 * quickly coalesce free runs.
472 	 *
473 	 * This is the limiting factor for chunksize; there can be at most 2^31
474 	 * pages in a run.
475 	 */
476 #define POS_FREE ((uint32_t)0xffffffffU)
477 	uint32_t	pos;
478 };
479 
480 /* Arena chunk header. */
481 typedef struct arena_chunk_s arena_chunk_t;
482 struct arena_chunk_s {
483 	/* Arena that owns the chunk. */
484 	arena_t *arena;
485 
486 	/* Linkage for the arena's chunk tree. */
487 	RB_ENTRY(arena_chunk_s) link;
488 
489 	/*
490 	 * Number of pages in use.  This is maintained in order to make
491 	 * detection of empty chunks fast.
492 	 */
493 	uint32_t pages_used;
494 
495 	/*
496 	 * Every time a free run larger than this value is created/coalesced,
497 	 * this value is increased.  The only way that the value decreases is if
498 	 * arena_run_alloc() fails to find a free run as large as advertised by
499 	 * this value.
500 	 */
501 	uint32_t max_frun_npages;
502 
503 	/*
504 	 * Every time a free run that starts at an earlier page than this value
505 	 * is created/coalesced, this value is decreased.  It is reset in a
506 	 * similar fashion to max_frun_npages.
507 	 */
508 	uint32_t min_frun_ind;
509 
510 	/*
511 	 * Map of pages within chunk that keeps track of free/large/small.  For
512 	 * free runs, only the map entries for the first and last pages are
513 	 * kept up to date, so that free runs can be quickly coalesced.
514 	 */
515 	arena_chunk_map_t map[1]; /* Dynamically sized. */
516 };
517 typedef struct arena_chunk_tree_s arena_chunk_tree_t;
518 RB_HEAD(arena_chunk_tree_s, arena_chunk_s);
519 
520 typedef struct arena_run_s arena_run_t;
521 struct arena_run_s {
522 	/* Linkage for run trees. */
523 	RB_ENTRY(arena_run_s) link;
524 
525 #ifdef MALLOC_DEBUG
526 	uint32_t	magic;
527 #  define ARENA_RUN_MAGIC 0x384adf93
528 #endif
529 
530 	/* Bin this run is associated with. */
531 	arena_bin_t	*bin;
532 
533 	/* Index of first element that might have a free region. */
534 	unsigned	regs_minelm;
535 
536 	/* Number of free regions in run. */
537 	unsigned	nfree;
538 
539 	/* Bitmask of in-use regions (0: in use, 1: free). */
540 	unsigned	regs_mask[1]; /* Dynamically sized. */
541 };
542 typedef struct arena_run_tree_s arena_run_tree_t;
543 RB_HEAD(arena_run_tree_s, arena_run_s);
544 
545 struct arena_bin_s {
546 	/*
547 	 * Current run being used to service allocations of this bin's size
548 	 * class.
549 	 */
550 	arena_run_t	*runcur;
551 
552 	/*
553 	 * Tree of non-full runs.  This tree is used when looking for an
554 	 * existing run when runcur is no longer usable.  We choose the
555 	 * non-full run that is lowest in memory; this policy tends to keep
556 	 * objects packed well, and it can also help reduce the number of
557 	 * almost-empty chunks.
558 	 */
559 	arena_run_tree_t runs;
560 
561 	/* Size of regions in a run for this bin's size class. */
562 	size_t		reg_size;
563 
564 	/* Total size of a run for this bin's size class. */
565 	size_t		run_size;
566 
567 	/* Total number of regions in a run for this bin's size class. */
568 	uint32_t	nregs;
569 
570 	/* Number of elements in a run's regs_mask for this bin's size class. */
571 	uint32_t	regs_mask_nelms;
572 
573 	/* Offset of first region in a run for this bin's size class. */
574 	uint32_t	reg0_offset;
575 
576 #ifdef MALLOC_STATS
577 	/* Bin statistics. */
578 	malloc_bin_stats_t stats;
579 #endif
580 };
581 
582 struct arena_s {
583 #ifdef MALLOC_DEBUG
584 	uint32_t		magic;
585 #  define ARENA_MAGIC 0x947d3d24
586 #endif
587 
588 	/* All operations on this arena require that mtx be locked. */
589 	malloc_mutex_t		mtx;
590 
591 #ifdef MALLOC_STATS
592 	arena_stats_t		stats;
593 #endif
594 
595 	/*
596 	 * Tree of chunks this arena manages.
597 	 */
598 	arena_chunk_tree_t	chunks;
599 
600 	/*
601 	 * In order to avoid rapid chunk allocation/deallocation when an arena
602 	 * oscillates right on the cusp of needing a new chunk, cache the most
603 	 * recently freed chunk.  This caching is disabled by opt_hint.
604 	 *
605 	 * There is one spare chunk per arena, rather than one spare total, in
606 	 * order to avoid interactions between multiple threads that could make
607 	 * a single spare inadequate.
608 	 */
609 	arena_chunk_t *spare;
610 
611 	/*
612 	 * bins is used to store rings of free regions of the following sizes,
613 	 * assuming a 16-byte quantum, 4kB pagesize, and default MALLOC_OPTIONS.
614 	 *
615 	 *   bins[i] | size |
616 	 *   --------+------+
617 	 *        0  |    2 |
618 	 *        1  |    4 |
619 	 *        2  |    8 |
620 	 *   --------+------+
621 	 *        3  |   16 |
622 	 *        4  |   32 |
623 	 *        5  |   48 |
624 	 *        6  |   64 |
625 	 *           :      :
626 	 *           :      :
627 	 *       33  |  496 |
628 	 *       34  |  512 |
629 	 *   --------+------+
630 	 *       35  | 1024 |
631 	 *       36  | 2048 |
632 	 *   --------+------+
633 	 */
634 	arena_bin_t		bins[1]; /* Dynamically sized. */
635 };
636 
637 /******************************************************************************/
638 /*
639  * Data.
640  */
641 
642 /* Number of CPUs. */
643 static unsigned		ncpus;
644 
645 /* VM page size. */
646 static size_t		pagesize;
647 static size_t		pagesize_mask;
648 static int		pagesize_2pow;
649 
650 /* Various bin-related settings. */
651 static size_t		bin_maxclass; /* Max size class for bins. */
652 static unsigned		ntbins; /* Number of (2^n)-spaced tiny bins. */
653 static unsigned		nqbins; /* Number of quantum-spaced bins. */
654 static unsigned		nsbins; /* Number of (2^n)-spaced sub-page bins. */
655 static size_t		small_min;
656 static size_t		small_max;
657 
658 /* Various quantum-related settings. */
659 static size_t		quantum;
660 static size_t		quantum_mask; /* (quantum - 1). */
661 
662 /* Various chunk-related settings. */
663 static size_t		chunksize;
664 static size_t		chunksize_mask; /* (chunksize - 1). */
665 static int		chunksize_2pow;
666 static unsigned		chunk_npages;
667 static unsigned		arena_chunk_header_npages;
668 static size_t		arena_maxclass; /* Max size class for arenas. */
669 
670 /********/
671 /*
672  * Chunks.
673  */
674 
675 /* Protects chunk-related data structures. */
676 static malloc_mutex_t	chunks_mtx;
677 
678 /* Tree of chunks that are stand-alone huge allocations. */
679 static chunk_tree_t	huge;
680 
681 #ifdef USE_BRK
682 /*
683  * Try to use brk for chunk-size allocations, due to address space constraints.
684  */
685 /*
686  * Protects sbrk() calls.  This must be separate from chunks_mtx, since
687  * base_pages_alloc() also uses sbrk(), but cannot lock chunks_mtx (doing so
688  * could cause recursive lock acquisition).
689  */
690 static malloc_mutex_t	brk_mtx;
691 /* Result of first sbrk(0) call. */
692 static void		*brk_base;
693 /* Current end of brk, or ((void *)-1) if brk is exhausted. */
694 static void		*brk_prev;
695 /* Current upper limit on brk addresses. */
696 static void		*brk_max;
697 #endif
698 
699 #ifdef MALLOC_STATS
700 /* Huge allocation statistics. */
701 static uint64_t		huge_nmalloc;
702 static uint64_t		huge_ndalloc;
703 static uint64_t		huge_nralloc;
704 static size_t		huge_allocated;
705 #endif
706 
707 /*
708  * Tree of chunks that were previously allocated.  This is used when allocating
709  * chunks, in an attempt to re-use address space.
710  */
711 static chunk_tree_t	old_chunks;
712 
713 /****************************/
714 /*
715  * base (internal allocation).
716  */
717 
718 /*
719  * Current pages that are being used for internal memory allocations.  These
720  * pages are carved up in cacheline-size quanta, so that there is no chance of
721  * false cache line sharing.
722  */
723 static void		*base_pages;
724 static void		*base_next_addr;
725 static void		*base_past_addr; /* Addr immediately past base_pages. */
726 static chunk_node_t	*base_chunk_nodes; /* LIFO cache of chunk nodes. */
727 static malloc_mutex_t	base_mtx;
728 #ifdef MALLOC_STATS
729 static size_t		base_mapped;
730 #endif
731 
732 /********/
733 /*
734  * Arenas.
735  */
736 
737 /*
738  * Arenas that are used to service external requests.  Not all elements of the
739  * arenas array are necessarily used; arenas are created lazily as needed.
740  */
741 static arena_t		**arenas;
742 static unsigned		narenas;
743 static unsigned		next_arena;
744 static malloc_mutex_t	arenas_mtx; /* Protects arenas initialization. */
745 
746 #ifndef NO_TLS
747 /*
748  * Map of pthread_self() --> arenas[???], used for selecting an arena to use
749  * for allocations.
750  */
751 static __thread arena_t	*arenas_map;
752 #define	get_arenas_map()	(arenas_map)
753 #define	set_arenas_map(x)	(arenas_map = x)
754 #else
755 static thread_key_t arenas_map_key;
756 #define	get_arenas_map()	thr_getspecific(arenas_map_key)
757 #define	set_arenas_map(x)	thr_setspecific(arenas_map_key, x)
758 #endif
759 
760 #ifdef MALLOC_STATS
761 /* Chunk statistics. */
762 static chunk_stats_t	stats_chunks;
763 #endif
764 
765 /*******************************/
766 /*
767  * Runtime configuration options.
768  */
769 const char	*_malloc_options;
770 
771 #ifndef MALLOC_PRODUCTION
772 static bool	opt_abort = true;
773 static bool	opt_junk = true;
774 #else
775 static bool	opt_abort = false;
776 static bool	opt_junk = false;
777 #endif
778 static bool	opt_hint = false;
779 static bool	opt_print_stats = false;
780 static int	opt_quantum_2pow = QUANTUM_2POW_MIN;
781 static int	opt_small_max_2pow = SMALL_MAX_2POW_DEFAULT;
782 static int	opt_chunk_2pow = CHUNK_2POW_DEFAULT;
783 static bool	opt_utrace = false;
784 static bool	opt_sysv = false;
785 static bool	opt_xmalloc = false;
786 static bool	opt_zero = false;
787 static int32_t	opt_narenas_lshift = 0;
788 
789 typedef struct {
790 	void	*p;
791 	size_t	s;
792 	void	*r;
793 } malloc_utrace_t;
794 
795 #define	UTRACE(a, b, c)							\
796 	if (opt_utrace) {						\
797 		malloc_utrace_t ut;					\
798 		ut.p = a;						\
799 		ut.s = b;						\
800 		ut.r = c;						\
801 		xutrace(&ut, sizeof(ut));				\
802 	}
803 
804 /******************************************************************************/
805 /*
806  * Begin function prototypes for non-inline static functions.
807  */
808 
809 static void	wrtmessage(const char *p1, const char *p2, const char *p3,
810 		const char *p4);
811 #ifdef MALLOC_STATS
812 static void	malloc_printf(const char *format, ...);
813 #endif
814 static char	*umax2s(uintmax_t x, char *s);
815 static bool	base_pages_alloc(size_t minsize);
816 static void	*base_alloc(size_t size);
817 static chunk_node_t *base_chunk_node_alloc(void);
818 static void	base_chunk_node_dealloc(chunk_node_t *node);
819 #ifdef MALLOC_STATS
820 static void	stats_print(arena_t *arena);
821 #endif
822 static void	*pages_map(void *addr, size_t size);
823 static void	*pages_map_align(void *addr, size_t size, int align);
824 static void	pages_unmap(void *addr, size_t size);
825 static void	*chunk_alloc(size_t size);
826 static void	chunk_dealloc(void *chunk, size_t size);
827 static void	arena_run_split(arena_t *arena, arena_run_t *run, size_t size);
828 static arena_chunk_t *arena_chunk_alloc(arena_t *arena);
829 static void	arena_chunk_dealloc(arena_t *arena, arena_chunk_t *chunk);
830 static arena_run_t *arena_run_alloc(arena_t *arena, size_t size);
831 static void	arena_run_dalloc(arena_t *arena, arena_run_t *run, size_t size);
832 static arena_run_t *arena_bin_nonfull_run_get(arena_t *arena, arena_bin_t *bin);
833 static void *arena_bin_malloc_hard(arena_t *arena, arena_bin_t *bin);
834 static size_t arena_bin_run_size_calc(arena_bin_t *bin, size_t min_run_size);
835 static void	*arena_malloc(arena_t *arena, size_t size);
836 static void	*arena_palloc(arena_t *arena, size_t alignment, size_t size,
837     size_t alloc_size);
838 static size_t	arena_salloc(const void *ptr);
839 static void	*arena_ralloc(void *ptr, size_t size, size_t oldsize);
840 static void	arena_dalloc(arena_t *arena, arena_chunk_t *chunk, void *ptr);
841 static bool	arena_new(arena_t *arena);
842 static arena_t	*arenas_extend(unsigned ind);
843 static void	*huge_malloc(size_t size);
844 static void	*huge_palloc(size_t alignment, size_t size);
845 static void	*huge_ralloc(void *ptr, size_t size, size_t oldsize);
846 static void	huge_dalloc(void *ptr);
847 static void	*imalloc(size_t size);
848 static void	*ipalloc(size_t alignment, size_t size);
849 static void	*icalloc(size_t size);
850 static size_t	isalloc(const void *ptr);
851 static void	*iralloc(void *ptr, size_t size);
852 static void	idalloc(void *ptr);
853 static void	malloc_print_stats(void);
854 static bool	malloc_init_hard(void);
855 
856 /*
857  * End function prototypes.
858  */
859 /******************************************************************************/
860 /*
861  * Begin mutex.
862  */
863 
864 #ifdef __NetBSD__
865 #define	malloc_mutex_init(m)	mutex_init(m, NULL)
866 #define	malloc_mutex_lock(m)	mutex_lock(m)
867 #define	malloc_mutex_unlock(m)	mutex_unlock(m)
868 #else	/* __NetBSD__ */
869 static inline void
870 malloc_mutex_init(malloc_mutex_t *a_mutex)
871 {
872 	static const spinlock_t lock = _SPINLOCK_INITIALIZER;
873 
874 	a_mutex->lock = lock;
875 }
876 
877 static inline void
878 malloc_mutex_lock(malloc_mutex_t *a_mutex)
879 {
880 
881 	if (__isthreaded)
882 		_SPINLOCK(&a_mutex->lock);
883 }
884 
885 static inline void
886 malloc_mutex_unlock(malloc_mutex_t *a_mutex)
887 {
888 
889 	if (__isthreaded)
890 		_SPINUNLOCK(&a_mutex->lock);
891 }
892 #endif	/* __NetBSD__ */
893 
894 /*
895  * End mutex.
896  */
897 /******************************************************************************/
898 /*
899  * Begin Utility functions/macros.
900  */
901 
902 /* Return the chunk address for allocation address a. */
903 #define	CHUNK_ADDR2BASE(a)						\
904 	((void *)((uintptr_t)(a) & ~chunksize_mask))
905 
906 /* Return the chunk offset of address a. */
907 #define	CHUNK_ADDR2OFFSET(a)						\
908 	((size_t)((uintptr_t)(a) & chunksize_mask))
909 
910 /* Return the smallest chunk multiple that is >= s. */
911 #define	CHUNK_CEILING(s)						\
912 	(((s) + chunksize_mask) & ~chunksize_mask)
913 
914 /* Return the smallest cacheline multiple that is >= s. */
915 #define	CACHELINE_CEILING(s)						\
916 	(((s) + (CACHELINE - 1)) & ~(CACHELINE - 1))
917 
918 /* Return the smallest quantum multiple that is >= a. */
919 #define	QUANTUM_CEILING(a)						\
920 	(((a) + quantum_mask) & ~quantum_mask)
921 
922 /* Return the smallest pagesize multiple that is >= s. */
923 #define	PAGE_CEILING(s)							\
924 	(((s) + pagesize_mask) & ~pagesize_mask)
925 
926 /* Compute the smallest power of 2 that is >= x. */
927 static inline size_t
928 pow2_ceil(size_t x)
929 {
930 
931 	x--;
932 	x |= x >> 1;
933 	x |= x >> 2;
934 	x |= x >> 4;
935 	x |= x >> 8;
936 	x |= x >> 16;
937 #if (SIZEOF_PTR == 8)
938 	x |= x >> 32;
939 #endif
940 	x++;
941 	return (x);
942 }
943 
944 static void
945 wrtmessage(const char *p1, const char *p2, const char *p3, const char *p4)
946 {
947 
948 	write(STDERR_FILENO, p1, strlen(p1));
949 	write(STDERR_FILENO, p2, strlen(p2));
950 	write(STDERR_FILENO, p3, strlen(p3));
951 	write(STDERR_FILENO, p4, strlen(p4));
952 }
953 
954 void	(*_malloc_message)(const char *p1, const char *p2, const char *p3,
955 	    const char *p4) = wrtmessage;
956 
957 #ifdef MALLOC_STATS
958 /*
959  * Print to stderr in such a way as to (hopefully) avoid memory allocation.
960  */
961 static void
962 malloc_printf(const char *format, ...)
963 {
964 	char buf[4096];
965 	va_list ap;
966 
967 	va_start(ap, format);
968 	vsnprintf(buf, sizeof(buf), format, ap);
969 	va_end(ap);
970 	_malloc_message(buf, "", "", "");
971 }
972 #endif
973 
974 /*
975  * We don't want to depend on vsnprintf() for production builds, since that can
976  * cause unnecessary bloat for static binaries.  umax2s() provides minimal
977  * integer printing functionality, so that malloc_printf() use can be limited to
978  * MALLOC_STATS code.
979  */
980 #define UMAX2S_BUFSIZE	21
981 static char *
982 umax2s(uintmax_t x, char *s)
983 {
984 	unsigned i;
985 
986 	/* Make sure UMAX2S_BUFSIZE is large enough. */
987 	/* LINTED */
988 	assert(sizeof(uintmax_t) <= 8);
989 
990 	i = UMAX2S_BUFSIZE - 1;
991 	s[i] = '\0';
992 	do {
993 		i--;
994 		s[i] = "0123456789"[(int)x % 10];
995 		x /= (uintmax_t)10LL;
996 	} while (x > 0);
997 
998 	return (&s[i]);
999 }
1000 
1001 /******************************************************************************/
1002 
1003 static bool
1004 base_pages_alloc(size_t minsize)
1005 {
1006 	size_t csize = 0;
1007 
1008 #ifdef USE_BRK
1009 	/*
1010 	 * Do special brk allocation here, since base allocations don't need to
1011 	 * be chunk-aligned.
1012 	 */
1013 	if (brk_prev != (void *)-1) {
1014 		void *brk_cur;
1015 		intptr_t incr;
1016 
1017 		if (minsize != 0)
1018 			csize = CHUNK_CEILING(minsize);
1019 
1020 		malloc_mutex_lock(&brk_mtx);
1021 		do {
1022 			/* Get the current end of brk. */
1023 			brk_cur = sbrk(0);
1024 
1025 			/*
1026 			 * Calculate how much padding is necessary to
1027 			 * chunk-align the end of brk.  Don't worry about
1028 			 * brk_cur not being chunk-aligned though.
1029 			 */
1030 			incr = (intptr_t)chunksize
1031 			    - (intptr_t)CHUNK_ADDR2OFFSET(brk_cur);
1032 			if (incr < minsize)
1033 				incr += csize;
1034 
1035 			brk_prev = sbrk(incr);
1036 			if (brk_prev == brk_cur) {
1037 				/* Success. */
1038 				malloc_mutex_unlock(&brk_mtx);
1039 				base_pages = brk_cur;
1040 				base_next_addr = base_pages;
1041 				base_past_addr = (void *)((uintptr_t)base_pages
1042 				    + incr);
1043 #ifdef MALLOC_STATS
1044 				base_mapped += incr;
1045 #endif
1046 				return (false);
1047 			}
1048 		} while (brk_prev != (void *)-1);
1049 		malloc_mutex_unlock(&brk_mtx);
1050 	}
1051 	if (minsize == 0) {
1052 		/*
1053 		 * Failure during initialization doesn't matter, so avoid
1054 		 * falling through to the mmap-based page mapping code.
1055 		 */
1056 		return (true);
1057 	}
1058 #endif
1059 	assert(minsize != 0);
1060 	csize = PAGE_CEILING(minsize);
1061 	base_pages = pages_map(NULL, csize);
1062 	if (base_pages == NULL)
1063 		return (true);
1064 	base_next_addr = base_pages;
1065 	base_past_addr = (void *)((uintptr_t)base_pages + csize);
1066 #ifdef MALLOC_STATS
1067 	base_mapped += csize;
1068 #endif
1069 	return (false);
1070 }
1071 
1072 static void *
1073 base_alloc(size_t size)
1074 {
1075 	void *ret;
1076 	size_t csize;
1077 
1078 	/* Round size up to nearest multiple of the cacheline size. */
1079 	csize = CACHELINE_CEILING(size);
1080 
1081 	malloc_mutex_lock(&base_mtx);
1082 
1083 	/* Make sure there's enough space for the allocation. */
1084 	if ((uintptr_t)base_next_addr + csize > (uintptr_t)base_past_addr) {
1085 		if (base_pages_alloc(csize)) {
1086 			ret = NULL;
1087 			goto RETURN;
1088 		}
1089 	}
1090 
1091 	/* Allocate. */
1092 	ret = base_next_addr;
1093 	base_next_addr = (void *)((uintptr_t)base_next_addr + csize);
1094 
1095 RETURN:
1096 	malloc_mutex_unlock(&base_mtx);
1097 	return (ret);
1098 }
1099 
1100 static chunk_node_t *
1101 base_chunk_node_alloc(void)
1102 {
1103 	chunk_node_t *ret;
1104 
1105 	malloc_mutex_lock(&base_mtx);
1106 	if (base_chunk_nodes != NULL) {
1107 		ret = base_chunk_nodes;
1108 		/* LINTED */
1109 		base_chunk_nodes = *(chunk_node_t **)ret;
1110 		malloc_mutex_unlock(&base_mtx);
1111 	} else {
1112 		malloc_mutex_unlock(&base_mtx);
1113 		ret = (chunk_node_t *)base_alloc(sizeof(chunk_node_t));
1114 	}
1115 
1116 	return (ret);
1117 }
1118 
1119 static void
1120 base_chunk_node_dealloc(chunk_node_t *node)
1121 {
1122 
1123 	malloc_mutex_lock(&base_mtx);
1124 	/* LINTED */
1125 	*(chunk_node_t **)node = base_chunk_nodes;
1126 	base_chunk_nodes = node;
1127 	malloc_mutex_unlock(&base_mtx);
1128 }
1129 
1130 /******************************************************************************/
1131 
1132 #ifdef MALLOC_STATS
1133 static void
1134 stats_print(arena_t *arena)
1135 {
1136 	unsigned i;
1137 	int gap_start;
1138 
1139 	malloc_printf(
1140 	    "          allocated/mapped            nmalloc      ndalloc\n");
1141 
1142 	malloc_printf("small: %12zu %-12s %12llu %12llu\n",
1143 	    arena->stats.allocated_small, "", arena->stats.nmalloc_small,
1144 	    arena->stats.ndalloc_small);
1145 	malloc_printf("large: %12zu %-12s %12llu %12llu\n",
1146 	    arena->stats.allocated_large, "", arena->stats.nmalloc_large,
1147 	    arena->stats.ndalloc_large);
1148 	malloc_printf("total: %12zu/%-12zu %12llu %12llu\n",
1149 	    arena->stats.allocated_small + arena->stats.allocated_large,
1150 	    arena->stats.mapped,
1151 	    arena->stats.nmalloc_small + arena->stats.nmalloc_large,
1152 	    arena->stats.ndalloc_small + arena->stats.ndalloc_large);
1153 
1154 	malloc_printf("bins:     bin   size regs pgs  requests   newruns"
1155 	    "    reruns maxruns curruns\n");
1156 	for (i = 0, gap_start = -1; i < ntbins + nqbins + nsbins; i++) {
1157 		if (arena->bins[i].stats.nrequests == 0) {
1158 			if (gap_start == -1)
1159 				gap_start = i;
1160 		} else {
1161 			if (gap_start != -1) {
1162 				if (i > gap_start + 1) {
1163 					/* Gap of more than one size class. */
1164 					malloc_printf("[%u..%u]\n",
1165 					    gap_start, i - 1);
1166 				} else {
1167 					/* Gap of one size class. */
1168 					malloc_printf("[%u]\n", gap_start);
1169 				}
1170 				gap_start = -1;
1171 			}
1172 			malloc_printf(
1173 			    "%13u %1s %4u %4u %3u %9llu %9llu"
1174 			    " %9llu %7lu %7lu\n",
1175 			    i,
1176 			    i < ntbins ? "T" : i < ntbins + nqbins ? "Q" : "S",
1177 			    arena->bins[i].reg_size,
1178 			    arena->bins[i].nregs,
1179 			    arena->bins[i].run_size >> pagesize_2pow,
1180 			    arena->bins[i].stats.nrequests,
1181 			    arena->bins[i].stats.nruns,
1182 			    arena->bins[i].stats.reruns,
1183 			    arena->bins[i].stats.highruns,
1184 			    arena->bins[i].stats.curruns);
1185 		}
1186 	}
1187 	if (gap_start != -1) {
1188 		if (i > gap_start + 1) {
1189 			/* Gap of more than one size class. */
1190 			malloc_printf("[%u..%u]\n", gap_start, i - 1);
1191 		} else {
1192 			/* Gap of one size class. */
1193 			malloc_printf("[%u]\n", gap_start);
1194 		}
1195 	}
1196 }
1197 #endif
1198 
1199 /*
1200  * End Utility functions/macros.
1201  */
1202 /******************************************************************************/
1203 /*
1204  * Begin chunk management functions.
1205  */
1206 
1207 #ifndef lint
1208 static inline int
1209 chunk_comp(chunk_node_t *a, chunk_node_t *b)
1210 {
1211 
1212 	assert(a != NULL);
1213 	assert(b != NULL);
1214 
1215 	if ((uintptr_t)a->chunk < (uintptr_t)b->chunk)
1216 		return (-1);
1217 	else if (a->chunk == b->chunk)
1218 		return (0);
1219 	else
1220 		return (1);
1221 }
1222 
1223 /* Generate red-black tree code for chunks. */
1224 RB_GENERATE_STATIC(chunk_tree_s, chunk_node_s, link, chunk_comp);
1225 #endif
1226 
1227 static void *
1228 pages_map_align(void *addr, size_t size, int align)
1229 {
1230 	void *ret;
1231 
1232 	/*
1233 	 * We don't use MAP_FIXED here, because it can cause the *replacement*
1234 	 * of existing mappings, and we only want to create new mappings.
1235 	 */
1236 	ret = mmap(addr, size, PROT_READ | PROT_WRITE,
1237 	    MAP_PRIVATE | MAP_ANON | MAP_ALIGNED(align), -1, 0);
1238 	assert(ret != NULL);
1239 
1240 	if (ret == MAP_FAILED)
1241 		ret = NULL;
1242 	else if (addr != NULL && ret != addr) {
1243 		/*
1244 		 * We succeeded in mapping memory, but not in the right place.
1245 		 */
1246 		if (munmap(ret, size) == -1) {
1247 			char buf[STRERROR_BUF];
1248 
1249 			STRERROR_R(errno, buf, sizeof(buf));
1250 			_malloc_message(getprogname(),
1251 			    ": (malloc) Error in munmap(): ", buf, "\n");
1252 			if (opt_abort)
1253 				abort();
1254 		}
1255 		ret = NULL;
1256 	}
1257 
1258 	assert(ret == NULL || (addr == NULL && ret != addr)
1259 	    || (addr != NULL && ret == addr));
1260 	return (ret);
1261 }
1262 
1263 static void *
1264 pages_map(void *addr, size_t size)
1265 {
1266 
1267 	return pages_map_align(addr, size, 0);
1268 }
1269 
1270 static void
1271 pages_unmap(void *addr, size_t size)
1272 {
1273 
1274 	if (munmap(addr, size) == -1) {
1275 		char buf[STRERROR_BUF];
1276 
1277 		STRERROR_R(errno, buf, sizeof(buf));
1278 		_malloc_message(getprogname(),
1279 		    ": (malloc) Error in munmap(): ", buf, "\n");
1280 		if (opt_abort)
1281 			abort();
1282 	}
1283 }
1284 
1285 static void *
1286 chunk_alloc(size_t size)
1287 {
1288 	void *ret, *chunk;
1289 	chunk_node_t *tchunk, *delchunk;
1290 
1291 	assert(size != 0);
1292 	assert((size & chunksize_mask) == 0);
1293 
1294 	malloc_mutex_lock(&chunks_mtx);
1295 
1296 	if (size == chunksize) {
1297 		/*
1298 		 * Check for address ranges that were previously chunks and try
1299 		 * to use them.
1300 		 */
1301 
1302 		/* LINTED */
1303 		tchunk = RB_MIN(chunk_tree_s, &old_chunks);
1304 		while (tchunk != NULL) {
1305 			/* Found an address range.  Try to recycle it. */
1306 
1307 			chunk = tchunk->chunk;
1308 			delchunk = tchunk;
1309 			/* LINTED */
1310 			tchunk = RB_NEXT(chunk_tree_s, &old_chunks, delchunk);
1311 
1312 			/* Remove delchunk from the tree. */
1313 			/* LINTED */
1314 			RB_REMOVE(chunk_tree_s, &old_chunks, delchunk);
1315 			base_chunk_node_dealloc(delchunk);
1316 
1317 #ifdef USE_BRK
1318 			if ((uintptr_t)chunk >= (uintptr_t)brk_base
1319 			    && (uintptr_t)chunk < (uintptr_t)brk_max) {
1320 				/* Re-use a previously freed brk chunk. */
1321 				ret = chunk;
1322 				goto RETURN;
1323 			}
1324 #endif
1325 			if ((ret = pages_map(chunk, size)) != NULL) {
1326 				/* Success. */
1327 				goto RETURN;
1328 			}
1329 		}
1330 	}
1331 
1332 	/*
1333 	 * Try to over-allocate, but allow the OS to place the allocation
1334 	 * anywhere.  Beware of size_t wrap-around.
1335 	 */
1336 	if (size + chunksize > size) {
1337 		if ((ret = pages_map_align(NULL, size, chunksize_2pow))
1338 		    != NULL) {
1339 			goto RETURN;
1340 		}
1341 	}
1342 
1343 #ifdef USE_BRK
1344 	/*
1345 	 * Try to create allocations in brk, in order to make full use of
1346 	 * limited address space.
1347 	 */
1348 	if (brk_prev != (void *)-1) {
1349 		void *brk_cur;
1350 		intptr_t incr;
1351 
1352 		/*
1353 		 * The loop is necessary to recover from races with other
1354 		 * threads that are using brk for something other than malloc.
1355 		 */
1356 		malloc_mutex_lock(&brk_mtx);
1357 		do {
1358 			/* Get the current end of brk. */
1359 			brk_cur = sbrk(0);
1360 
1361 			/*
1362 			 * Calculate how much padding is necessary to
1363 			 * chunk-align the end of brk.
1364 			 */
1365 			incr = (intptr_t)size
1366 			    - (intptr_t)CHUNK_ADDR2OFFSET(brk_cur);
1367 			if (incr == size) {
1368 				ret = brk_cur;
1369 			} else {
1370 				ret = (void *)((intptr_t)brk_cur + incr);
1371 				incr += size;
1372 			}
1373 
1374 			brk_prev = sbrk(incr);
1375 			if (brk_prev == brk_cur) {
1376 				/* Success. */
1377 				malloc_mutex_unlock(&brk_mtx);
1378 				brk_max = (void *)((intptr_t)ret + size);
1379 				goto RETURN;
1380 			}
1381 		} while (brk_prev != (void *)-1);
1382 		malloc_mutex_unlock(&brk_mtx);
1383 	}
1384 #endif
1385 
1386 	/* All strategies for allocation failed. */
1387 	ret = NULL;
1388 RETURN:
1389 	if (ret != NULL) {
1390 		chunk_node_t key;
1391 		/*
1392 		 * Clean out any entries in old_chunks that overlap with the
1393 		 * memory we just allocated.
1394 		 */
1395 		key.chunk = ret;
1396 		/* LINTED */
1397 		tchunk = RB_NFIND(chunk_tree_s, &old_chunks, &key);
1398 		while (tchunk != NULL
1399 		    && (uintptr_t)tchunk->chunk >= (uintptr_t)ret
1400 		    && (uintptr_t)tchunk->chunk < (uintptr_t)ret + size) {
1401 			delchunk = tchunk;
1402 			/* LINTED */
1403 			tchunk = RB_NEXT(chunk_tree_s, &old_chunks, delchunk);
1404 			/* LINTED */
1405 			RB_REMOVE(chunk_tree_s, &old_chunks, delchunk);
1406 			base_chunk_node_dealloc(delchunk);
1407 		}
1408 
1409 	}
1410 #ifdef MALLOC_STATS
1411 	if (ret != NULL) {
1412 		stats_chunks.nchunks += (size / chunksize);
1413 		stats_chunks.curchunks += (size / chunksize);
1414 	}
1415 	if (stats_chunks.curchunks > stats_chunks.highchunks)
1416 		stats_chunks.highchunks = stats_chunks.curchunks;
1417 #endif
1418 	malloc_mutex_unlock(&chunks_mtx);
1419 
1420 	assert(CHUNK_ADDR2BASE(ret) == ret);
1421 	return (ret);
1422 }
1423 
1424 static void
1425 chunk_dealloc(void *chunk, size_t size)
1426 {
1427 	chunk_node_t *node;
1428 
1429 	assert(chunk != NULL);
1430 	assert(CHUNK_ADDR2BASE(chunk) == chunk);
1431 	assert(size != 0);
1432 	assert((size & chunksize_mask) == 0);
1433 
1434 	malloc_mutex_lock(&chunks_mtx);
1435 
1436 #ifdef USE_BRK
1437 	if ((uintptr_t)chunk >= (uintptr_t)brk_base
1438 	    && (uintptr_t)chunk < (uintptr_t)brk_max) {
1439 		void *brk_cur;
1440 
1441 		malloc_mutex_lock(&brk_mtx);
1442 		/* Get the current end of brk. */
1443 		brk_cur = sbrk(0);
1444 
1445 		/*
1446 		 * Try to shrink the data segment if this chunk is at the end
1447 		 * of the data segment.  The sbrk() call here is subject to a
1448 		 * race condition with threads that use brk(2) or sbrk(2)
1449 		 * directly, but the alternative would be to leak memory for
1450 		 * the sake of poorly designed multi-threaded programs.
1451 		 */
1452 		if (brk_cur == brk_max
1453 		    && (void *)((uintptr_t)chunk + size) == brk_max
1454 		    && sbrk(-(intptr_t)size) == brk_max) {
1455 			malloc_mutex_unlock(&brk_mtx);
1456 			if (brk_prev == brk_max) {
1457 				/* Success. */
1458 				brk_prev = (void *)((intptr_t)brk_max
1459 				    - (intptr_t)size);
1460 				brk_max = brk_prev;
1461 			}
1462 		} else {
1463 			size_t offset;
1464 
1465 			malloc_mutex_unlock(&brk_mtx);
1466 			madvise(chunk, size, MADV_FREE);
1467 
1468 			/*
1469 			 * Iteratively create records of each chunk-sized
1470 			 * memory region that 'chunk' is comprised of, so that
1471 			 * the address range can be recycled if memory usage
1472 			 * increases later on.
1473 			 */
1474 			for (offset = 0; offset < size; offset += chunksize) {
1475 				node = base_chunk_node_alloc();
1476 				if (node == NULL)
1477 					break;
1478 
1479 				node->chunk = (void *)((uintptr_t)chunk
1480 				    + (uintptr_t)offset);
1481 				node->size = chunksize;
1482 				/* LINTED */
1483 				RB_INSERT(chunk_tree_s, &old_chunks, node);
1484 			}
1485 		}
1486 	} else {
1487 #endif
1488 		pages_unmap(chunk, size);
1489 
1490 		/*
1491 		 * Make a record of the chunk's address, so that the address
1492 		 * range can be recycled if memory usage increases later on.
1493 		 * Don't bother to create entries if (size > chunksize), since
1494 		 * doing so could cause scalability issues for truly gargantuan
1495 		 * objects (many gigabytes or larger).
1496 		 */
1497 		if (size == chunksize) {
1498 			node = base_chunk_node_alloc();
1499 			if (node != NULL) {
1500 				node->chunk = (void *)(uintptr_t)chunk;
1501 				node->size = chunksize;
1502 				/* LINTED */
1503 				RB_INSERT(chunk_tree_s, &old_chunks, node);
1504 			}
1505 		}
1506 #ifdef USE_BRK
1507 	}
1508 #endif
1509 
1510 #ifdef MALLOC_STATS
1511 	stats_chunks.curchunks -= (size / chunksize);
1512 #endif
1513 	malloc_mutex_unlock(&chunks_mtx);
1514 }
1515 
1516 /*
1517  * End chunk management functions.
1518  */
1519 /******************************************************************************/
1520 /*
1521  * Begin arena.
1522  */
1523 
1524 /*
1525  * Choose an arena based on a per-thread and (optimistically) per-CPU value.
1526  *
1527  * We maintain at least one block of arenas.  Usually there are more.
1528  * The blocks are $ncpu arenas in size.  Whole blocks are 'hashed'
1529  * amongst threads.  To accomplish this, next_arena advances only in
1530  * ncpu steps.
1531  */
1532 static __noinline arena_t *
1533 choose_arena_hard(void)
1534 {
1535 	unsigned i, curcpu;
1536 	arena_t **map;
1537 
1538 	/* Initialize the current block of arenas and advance to next. */
1539 	malloc_mutex_lock(&arenas_mtx);
1540 	assert(next_arena % ncpus == 0);
1541 	assert(narenas % ncpus == 0);
1542 	map = &arenas[next_arena];
1543 	set_arenas_map(map);
1544 	for (i = 0; i < ncpus; i++) {
1545 		if (arenas[next_arena] == NULL)
1546 			arenas_extend(next_arena);
1547 		next_arena = (next_arena + 1) % narenas;
1548 	}
1549 	malloc_mutex_unlock(&arenas_mtx);
1550 
1551 	/*
1552 	 * If we were unable to allocate an arena above, then default to
1553 	 * the first arena, which is always present.
1554 	 */
1555 	curcpu = thr_curcpu();
1556 	if (map[curcpu] != NULL)
1557 		return map[curcpu];
1558 	return arenas[0];
1559 }
1560 
1561 static inline arena_t *
1562 choose_arena(void)
1563 {
1564 	unsigned curcpu;
1565 	arena_t **map;
1566 
1567 	map = get_arenas_map();
1568 	curcpu = thr_curcpu();
1569 	if (__predict_true(map != NULL && map[curcpu] != NULL))
1570 		return map[curcpu];
1571 
1572         return choose_arena_hard();
1573 }
1574 
1575 #ifndef lint
1576 static inline int
1577 arena_chunk_comp(arena_chunk_t *a, arena_chunk_t *b)
1578 {
1579 
1580 	assert(a != NULL);
1581 	assert(b != NULL);
1582 
1583 	if ((uintptr_t)a < (uintptr_t)b)
1584 		return (-1);
1585 	else if (a == b)
1586 		return (0);
1587 	else
1588 		return (1);
1589 }
1590 
1591 /* Generate red-black tree code for arena chunks. */
1592 RB_GENERATE_STATIC(arena_chunk_tree_s, arena_chunk_s, link, arena_chunk_comp);
1593 #endif
1594 
1595 #ifndef lint
1596 static inline int
1597 arena_run_comp(arena_run_t *a, arena_run_t *b)
1598 {
1599 
1600 	assert(a != NULL);
1601 	assert(b != NULL);
1602 
1603 	if ((uintptr_t)a < (uintptr_t)b)
1604 		return (-1);
1605 	else if (a == b)
1606 		return (0);
1607 	else
1608 		return (1);
1609 }
1610 
1611 /* Generate red-black tree code for arena runs. */
1612 RB_GENERATE_STATIC(arena_run_tree_s, arena_run_s, link, arena_run_comp);
1613 #endif
1614 
1615 static inline void *
1616 arena_run_reg_alloc(arena_run_t *run, arena_bin_t *bin)
1617 {
1618 	void *ret;
1619 	unsigned i, mask, bit, regind;
1620 
1621 	assert(run->magic == ARENA_RUN_MAGIC);
1622 	assert(run->regs_minelm < bin->regs_mask_nelms);
1623 
1624 	/*
1625 	 * Move the first check outside the loop, so that run->regs_minelm can
1626 	 * be updated unconditionally, without the possibility of updating it
1627 	 * multiple times.
1628 	 */
1629 	i = run->regs_minelm;
1630 	mask = run->regs_mask[i];
1631 	if (mask != 0) {
1632 		/* Usable allocation found. */
1633 		bit = ffs((int)mask) - 1;
1634 
1635 		regind = ((i << (SIZEOF_INT_2POW + 3)) + bit);
1636 		ret = (void *)(((uintptr_t)run) + bin->reg0_offset
1637 		    + (bin->reg_size * regind));
1638 
1639 		/* Clear bit. */
1640 		mask ^= (1 << bit);
1641 		run->regs_mask[i] = mask;
1642 
1643 		return (ret);
1644 	}
1645 
1646 	for (i++; i < bin->regs_mask_nelms; i++) {
1647 		mask = run->regs_mask[i];
1648 		if (mask != 0) {
1649 			/* Usable allocation found. */
1650 			bit = ffs((int)mask) - 1;
1651 
1652 			regind = ((i << (SIZEOF_INT_2POW + 3)) + bit);
1653 			ret = (void *)(((uintptr_t)run) + bin->reg0_offset
1654 			    + (bin->reg_size * regind));
1655 
1656 			/* Clear bit. */
1657 			mask ^= (1 << bit);
1658 			run->regs_mask[i] = mask;
1659 
1660 			/*
1661 			 * Make a note that nothing before this element
1662 			 * contains a free region.
1663 			 */
1664 			run->regs_minelm = i; /* Low payoff: + (mask == 0); */
1665 
1666 			return (ret);
1667 		}
1668 	}
1669 	/* Not reached. */
1670 	/* LINTED */
1671 	assert(0);
1672 	return (NULL);
1673 }
1674 
1675 static inline void
1676 arena_run_reg_dalloc(arena_run_t *run, arena_bin_t *bin, void *ptr, size_t size)
1677 {
1678 	/*
1679 	 * To divide by a number D that is not a power of two we multiply
1680 	 * by (2^21 / D) and then right shift by 21 positions.
1681 	 *
1682 	 *   X / D
1683 	 *
1684 	 * becomes
1685 	 *
1686 	 *   (X * size_invs[(D >> QUANTUM_2POW_MIN) - 3]) >> SIZE_INV_SHIFT
1687 	 */
1688 #define SIZE_INV_SHIFT 21
1689 #define SIZE_INV(s) (((1 << SIZE_INV_SHIFT) / (s << QUANTUM_2POW_MIN)) + 1)
1690 	static const unsigned size_invs[] = {
1691 	    SIZE_INV(3),
1692 	    SIZE_INV(4), SIZE_INV(5), SIZE_INV(6), SIZE_INV(7),
1693 	    SIZE_INV(8), SIZE_INV(9), SIZE_INV(10), SIZE_INV(11),
1694 	    SIZE_INV(12),SIZE_INV(13), SIZE_INV(14), SIZE_INV(15),
1695 	    SIZE_INV(16),SIZE_INV(17), SIZE_INV(18), SIZE_INV(19),
1696 	    SIZE_INV(20),SIZE_INV(21), SIZE_INV(22), SIZE_INV(23),
1697 	    SIZE_INV(24),SIZE_INV(25), SIZE_INV(26), SIZE_INV(27),
1698 	    SIZE_INV(28),SIZE_INV(29), SIZE_INV(30), SIZE_INV(31)
1699 #if (QUANTUM_2POW_MIN < 4)
1700 	    ,
1701 	    SIZE_INV(32), SIZE_INV(33), SIZE_INV(34), SIZE_INV(35),
1702 	    SIZE_INV(36), SIZE_INV(37), SIZE_INV(38), SIZE_INV(39),
1703 	    SIZE_INV(40), SIZE_INV(41), SIZE_INV(42), SIZE_INV(43),
1704 	    SIZE_INV(44), SIZE_INV(45), SIZE_INV(46), SIZE_INV(47),
1705 	    SIZE_INV(48), SIZE_INV(49), SIZE_INV(50), SIZE_INV(51),
1706 	    SIZE_INV(52), SIZE_INV(53), SIZE_INV(54), SIZE_INV(55),
1707 	    SIZE_INV(56), SIZE_INV(57), SIZE_INV(58), SIZE_INV(59),
1708 	    SIZE_INV(60), SIZE_INV(61), SIZE_INV(62), SIZE_INV(63)
1709 #endif
1710 	};
1711 	unsigned diff, regind, elm, bit;
1712 
1713 	/* LINTED */
1714 	assert(run->magic == ARENA_RUN_MAGIC);
1715 	assert(((sizeof(size_invs)) / sizeof(unsigned)) + 3
1716 	    >= (SMALL_MAX_DEFAULT >> QUANTUM_2POW_MIN));
1717 
1718 	/*
1719 	 * Avoid doing division with a variable divisor if possible.  Using
1720 	 * actual division here can reduce allocator throughput by over 20%!
1721 	 */
1722 	diff = (unsigned)((uintptr_t)ptr - (uintptr_t)run - bin->reg0_offset);
1723 	if ((size & (size - 1)) == 0) {
1724 		/*
1725 		 * log2_table allows fast division of a power of two in the
1726 		 * [1..128] range.
1727 		 *
1728 		 * (x / divisor) becomes (x >> log2_table[divisor - 1]).
1729 		 */
1730 		static const unsigned char log2_table[] = {
1731 		    0, 1, 0, 2, 0, 0, 0, 3, 0, 0, 0, 0, 0, 0, 0, 4,
1732 		    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 5,
1733 		    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
1734 		    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 6,
1735 		    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
1736 		    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
1737 		    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
1738 		    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 7
1739 		};
1740 
1741 		if (size <= 128)
1742 			regind = (diff >> log2_table[size - 1]);
1743 		else if (size <= 32768)
1744 			regind = diff >> (8 + log2_table[(size >> 8) - 1]);
1745 		else {
1746 			/*
1747 			 * The page size is too large for us to use the lookup
1748 			 * table.  Use real division.
1749 			 */
1750 			regind = (unsigned)(diff / size);
1751 		}
1752 	} else if (size <= ((sizeof(size_invs) / sizeof(unsigned))
1753 	    << QUANTUM_2POW_MIN) + 2) {
1754 		regind = size_invs[(size >> QUANTUM_2POW_MIN) - 3] * diff;
1755 		regind >>= SIZE_INV_SHIFT;
1756 	} else {
1757 		/*
1758 		 * size_invs isn't large enough to handle this size class, so
1759 		 * calculate regind using actual division.  This only happens
1760 		 * if the user increases small_max via the 'S' runtime
1761 		 * configuration option.
1762 		 */
1763 		regind = (unsigned)(diff / size);
1764 	};
1765 	assert(diff == regind * size);
1766 	assert(regind < bin->nregs);
1767 
1768 	elm = regind >> (SIZEOF_INT_2POW + 3);
1769 	if (elm < run->regs_minelm)
1770 		run->regs_minelm = elm;
1771 	bit = regind - (elm << (SIZEOF_INT_2POW + 3));
1772 	assert((run->regs_mask[elm] & (1 << bit)) == 0);
1773 	run->regs_mask[elm] |= (1 << bit);
1774 #undef SIZE_INV
1775 #undef SIZE_INV_SHIFT
1776 }
1777 
1778 static void
1779 arena_run_split(arena_t *arena, arena_run_t *run, size_t size)
1780 {
1781 	arena_chunk_t *chunk;
1782 	unsigned run_ind, map_offset, total_pages, need_pages, rem_pages;
1783 	unsigned i;
1784 
1785 	chunk = (arena_chunk_t *)CHUNK_ADDR2BASE(run);
1786 	run_ind = (unsigned)(((uintptr_t)run - (uintptr_t)chunk)
1787 	    >> pagesize_2pow);
1788 	total_pages = chunk->map[run_ind].npages;
1789 	need_pages = (unsigned)(size >> pagesize_2pow);
1790 	assert(need_pages <= total_pages);
1791 	rem_pages = total_pages - need_pages;
1792 
1793 	/* Split enough pages from the front of run to fit allocation size. */
1794 	map_offset = run_ind;
1795 	for (i = 0; i < need_pages; i++) {
1796 		chunk->map[map_offset + i].npages = need_pages;
1797 		chunk->map[map_offset + i].pos = i;
1798 	}
1799 
1800 	/* Keep track of trailing unused pages for later use. */
1801 	if (rem_pages > 0) {
1802 		/* Update map for trailing pages. */
1803 		map_offset += need_pages;
1804 		chunk->map[map_offset].npages = rem_pages;
1805 		chunk->map[map_offset].pos = POS_FREE;
1806 		chunk->map[map_offset + rem_pages - 1].npages = rem_pages;
1807 		chunk->map[map_offset + rem_pages - 1].pos = POS_FREE;
1808 	}
1809 
1810 	chunk->pages_used += need_pages;
1811 }
1812 
1813 static arena_chunk_t *
1814 arena_chunk_alloc(arena_t *arena)
1815 {
1816 	arena_chunk_t *chunk;
1817 
1818 	if (arena->spare != NULL) {
1819 		chunk = arena->spare;
1820 		arena->spare = NULL;
1821 
1822 		/* LINTED */
1823 		RB_INSERT(arena_chunk_tree_s, &arena->chunks, chunk);
1824 	} else {
1825 		chunk = (arena_chunk_t *)chunk_alloc(chunksize);
1826 		if (chunk == NULL)
1827 			return (NULL);
1828 #ifdef MALLOC_STATS
1829 		arena->stats.mapped += chunksize;
1830 #endif
1831 
1832 		chunk->arena = arena;
1833 
1834 		/* LINTED */
1835 		RB_INSERT(arena_chunk_tree_s, &arena->chunks, chunk);
1836 
1837 		/*
1838 		 * Claim that no pages are in use, since the header is merely
1839 		 * overhead.
1840 		 */
1841 		chunk->pages_used = 0;
1842 
1843 		chunk->max_frun_npages = chunk_npages -
1844 		    arena_chunk_header_npages;
1845 		chunk->min_frun_ind = arena_chunk_header_npages;
1846 
1847 		/*
1848 		 * Initialize enough of the map to support one maximal free run.
1849 		 */
1850 		chunk->map[arena_chunk_header_npages].npages = chunk_npages -
1851 		    arena_chunk_header_npages;
1852 		chunk->map[arena_chunk_header_npages].pos = POS_FREE;
1853 		chunk->map[chunk_npages - 1].npages = chunk_npages -
1854 		    arena_chunk_header_npages;
1855 		chunk->map[chunk_npages - 1].pos = POS_FREE;
1856 	}
1857 
1858 	return (chunk);
1859 }
1860 
1861 static void
1862 arena_chunk_dealloc(arena_t *arena, arena_chunk_t *chunk)
1863 {
1864 
1865 	/*
1866 	 * Remove chunk from the chunk tree, regardless of whether this chunk
1867 	 * will be cached, so that the arena does not use it.
1868 	 */
1869 	/* LINTED */
1870 	RB_REMOVE(arena_chunk_tree_s, &chunk->arena->chunks, chunk);
1871 
1872 	if (opt_hint == false) {
1873 		if (arena->spare != NULL) {
1874 			chunk_dealloc((void *)arena->spare, chunksize);
1875 #ifdef MALLOC_STATS
1876 			arena->stats.mapped -= chunksize;
1877 #endif
1878 		}
1879 		arena->spare = chunk;
1880 	} else {
1881 		assert(arena->spare == NULL);
1882 		chunk_dealloc((void *)chunk, chunksize);
1883 #ifdef MALLOC_STATS
1884 		arena->stats.mapped -= chunksize;
1885 #endif
1886 	}
1887 }
1888 
1889 static arena_run_t *
1890 arena_run_alloc(arena_t *arena, size_t size)
1891 {
1892 	arena_chunk_t *chunk;
1893 	arena_run_t *run;
1894 	unsigned need_npages, limit_pages, compl_need_npages;
1895 
1896 	assert(size <= (chunksize - (arena_chunk_header_npages <<
1897 	    pagesize_2pow)));
1898 	assert((size & pagesize_mask) == 0);
1899 
1900 	/*
1901 	 * Search through arena's chunks in address order for a free run that is
1902 	 * large enough.  Look for the first fit.
1903 	 */
1904 	need_npages = (unsigned)(size >> pagesize_2pow);
1905 	limit_pages = chunk_npages - arena_chunk_header_npages;
1906 	compl_need_npages = limit_pages - need_npages;
1907 	/* LINTED */
1908 	RB_FOREACH(chunk, arena_chunk_tree_s, &arena->chunks) {
1909 		/*
1910 		 * Avoid searching this chunk if there are not enough
1911 		 * contiguous free pages for there to possibly be a large
1912 		 * enough free run.
1913 		 */
1914 		if (chunk->pages_used <= compl_need_npages &&
1915 		    need_npages <= chunk->max_frun_npages) {
1916 			arena_chunk_map_t *mapelm;
1917 			unsigned i;
1918 			unsigned max_frun_npages = 0;
1919 			unsigned min_frun_ind = chunk_npages;
1920 
1921 			assert(chunk->min_frun_ind >=
1922 			    arena_chunk_header_npages);
1923 			for (i = chunk->min_frun_ind; i < chunk_npages;) {
1924 				mapelm = &chunk->map[i];
1925 				if (mapelm->pos == POS_FREE) {
1926 					if (mapelm->npages >= need_npages) {
1927 						run = (arena_run_t *)
1928 						    ((uintptr_t)chunk + (i <<
1929 						    pagesize_2pow));
1930 						/* Update page map. */
1931 						arena_run_split(arena, run,
1932 						    size);
1933 						return (run);
1934 					}
1935 					if (mapelm->npages >
1936 					    max_frun_npages) {
1937 						max_frun_npages =
1938 						    mapelm->npages;
1939 					}
1940 					if (i < min_frun_ind) {
1941 						min_frun_ind = i;
1942 						if (i < chunk->min_frun_ind)
1943 							chunk->min_frun_ind = i;
1944 					}
1945 				}
1946 				i += mapelm->npages;
1947 			}
1948 			/*
1949 			 * Search failure.  Reset cached chunk->max_frun_npages.
1950 			 * chunk->min_frun_ind was already reset above (if
1951 			 * necessary).
1952 			 */
1953 			chunk->max_frun_npages = max_frun_npages;
1954 		}
1955 	}
1956 
1957 	/*
1958 	 * No usable runs.  Create a new chunk from which to allocate the run.
1959 	 */
1960 	chunk = arena_chunk_alloc(arena);
1961 	if (chunk == NULL)
1962 		return (NULL);
1963 	run = (arena_run_t *)((uintptr_t)chunk + (arena_chunk_header_npages <<
1964 	    pagesize_2pow));
1965 	/* Update page map. */
1966 	arena_run_split(arena, run, size);
1967 	return (run);
1968 }
1969 
1970 static void
1971 arena_run_dalloc(arena_t *arena, arena_run_t *run, size_t size)
1972 {
1973 	arena_chunk_t *chunk;
1974 	unsigned run_ind, run_pages;
1975 
1976 	chunk = (arena_chunk_t *)CHUNK_ADDR2BASE(run);
1977 
1978 	run_ind = (unsigned)(((uintptr_t)run - (uintptr_t)chunk)
1979 	    >> pagesize_2pow);
1980 	assert(run_ind >= arena_chunk_header_npages);
1981 	assert(run_ind < (chunksize >> pagesize_2pow));
1982 	run_pages = (unsigned)(size >> pagesize_2pow);
1983 	assert(run_pages == chunk->map[run_ind].npages);
1984 
1985 	/* Subtract pages from count of pages used in chunk. */
1986 	chunk->pages_used -= run_pages;
1987 
1988 	/* Mark run as deallocated. */
1989 	assert(chunk->map[run_ind].npages == run_pages);
1990 	chunk->map[run_ind].pos = POS_FREE;
1991 	assert(chunk->map[run_ind + run_pages - 1].npages == run_pages);
1992 	chunk->map[run_ind + run_pages - 1].pos = POS_FREE;
1993 
1994 	/*
1995 	 * Tell the kernel that we don't need the data in this run, but only if
1996 	 * requested via runtime configuration.
1997 	 */
1998 	if (opt_hint)
1999 		madvise(run, size, MADV_FREE);
2000 
2001 	/* Try to coalesce with neighboring runs. */
2002 	if (run_ind > arena_chunk_header_npages &&
2003 	    chunk->map[run_ind - 1].pos == POS_FREE) {
2004 		unsigned prev_npages;
2005 
2006 		/* Coalesce with previous run. */
2007 		prev_npages = chunk->map[run_ind - 1].npages;
2008 		run_ind -= prev_npages;
2009 		assert(chunk->map[run_ind].npages == prev_npages);
2010 		assert(chunk->map[run_ind].pos == POS_FREE);
2011 		run_pages += prev_npages;
2012 
2013 		chunk->map[run_ind].npages = run_pages;
2014 		assert(chunk->map[run_ind].pos == POS_FREE);
2015 		chunk->map[run_ind + run_pages - 1].npages = run_pages;
2016 		assert(chunk->map[run_ind + run_pages - 1].pos == POS_FREE);
2017 	}
2018 
2019 	if (run_ind + run_pages < chunk_npages &&
2020 	    chunk->map[run_ind + run_pages].pos == POS_FREE) {
2021 		unsigned next_npages;
2022 
2023 		/* Coalesce with next run. */
2024 		next_npages = chunk->map[run_ind + run_pages].npages;
2025 		run_pages += next_npages;
2026 		assert(chunk->map[run_ind + run_pages - 1].npages ==
2027 		    next_npages);
2028 		assert(chunk->map[run_ind + run_pages - 1].pos == POS_FREE);
2029 
2030 		chunk->map[run_ind].npages = run_pages;
2031 		chunk->map[run_ind].pos = POS_FREE;
2032 		chunk->map[run_ind + run_pages - 1].npages = run_pages;
2033 		assert(chunk->map[run_ind + run_pages - 1].pos == POS_FREE);
2034 	}
2035 
2036 	if (chunk->map[run_ind].npages > chunk->max_frun_npages)
2037 		chunk->max_frun_npages = chunk->map[run_ind].npages;
2038 	if (run_ind < chunk->min_frun_ind)
2039 		chunk->min_frun_ind = run_ind;
2040 
2041 	/* Deallocate chunk if it is now completely unused. */
2042 	if (chunk->pages_used == 0)
2043 		arena_chunk_dealloc(arena, chunk);
2044 }
2045 
2046 static arena_run_t *
2047 arena_bin_nonfull_run_get(arena_t *arena, arena_bin_t *bin)
2048 {
2049 	arena_run_t *run;
2050 	unsigned i, remainder;
2051 
2052 	/* Look for a usable run. */
2053 	/* LINTED */
2054 	if ((run = RB_MIN(arena_run_tree_s, &bin->runs)) != NULL) {
2055 		/* run is guaranteed to have available space. */
2056 		/* LINTED */
2057 		RB_REMOVE(arena_run_tree_s, &bin->runs, run);
2058 #ifdef MALLOC_STATS
2059 		bin->stats.reruns++;
2060 #endif
2061 		return (run);
2062 	}
2063 	/* No existing runs have any space available. */
2064 
2065 	/* Allocate a new run. */
2066 	run = arena_run_alloc(arena, bin->run_size);
2067 	if (run == NULL)
2068 		return (NULL);
2069 
2070 	/* Initialize run internals. */
2071 	run->bin = bin;
2072 
2073 	for (i = 0; i < bin->regs_mask_nelms; i++)
2074 		run->regs_mask[i] = UINT_MAX;
2075 	remainder = bin->nregs & ((1 << (SIZEOF_INT_2POW + 3)) - 1);
2076 	if (remainder != 0) {
2077 		/* The last element has spare bits that need to be unset. */
2078 		run->regs_mask[i] = (UINT_MAX >> ((1 << (SIZEOF_INT_2POW + 3))
2079 		    - remainder));
2080 	}
2081 
2082 	run->regs_minelm = 0;
2083 
2084 	run->nfree = bin->nregs;
2085 #ifdef MALLOC_DEBUG
2086 	run->magic = ARENA_RUN_MAGIC;
2087 #endif
2088 
2089 #ifdef MALLOC_STATS
2090 	bin->stats.nruns++;
2091 	bin->stats.curruns++;
2092 	if (bin->stats.curruns > bin->stats.highruns)
2093 		bin->stats.highruns = bin->stats.curruns;
2094 #endif
2095 	return (run);
2096 }
2097 
2098 /* bin->runcur must have space available before this function is called. */
2099 static inline void *
2100 arena_bin_malloc_easy(arena_t *arena, arena_bin_t *bin, arena_run_t *run)
2101 {
2102 	void *ret;
2103 
2104 	assert(run->magic == ARENA_RUN_MAGIC);
2105 	assert(run->nfree > 0);
2106 
2107 	ret = arena_run_reg_alloc(run, bin);
2108 	assert(ret != NULL);
2109 	run->nfree--;
2110 
2111 	return (ret);
2112 }
2113 
2114 /* Re-fill bin->runcur, then call arena_bin_malloc_easy(). */
2115 static void *
2116 arena_bin_malloc_hard(arena_t *arena, arena_bin_t *bin)
2117 {
2118 
2119 	bin->runcur = arena_bin_nonfull_run_get(arena, bin);
2120 	if (bin->runcur == NULL)
2121 		return (NULL);
2122 	assert(bin->runcur->magic == ARENA_RUN_MAGIC);
2123 	assert(bin->runcur->nfree > 0);
2124 
2125 	return (arena_bin_malloc_easy(arena, bin, bin->runcur));
2126 }
2127 
2128 /*
2129  * Calculate bin->run_size such that it meets the following constraints:
2130  *
2131  *   *) bin->run_size >= min_run_size
2132  *   *) bin->run_size <= arena_maxclass
2133  *   *) bin->run_size <= RUN_MAX_SMALL
2134  *   *) run header overhead <= RUN_MAX_OVRHD (or header overhead relaxed).
2135  *
2136  * bin->nregs, bin->regs_mask_nelms, and bin->reg0_offset are
2137  * also calculated here, since these settings are all interdependent.
2138  */
2139 static size_t
2140 arena_bin_run_size_calc(arena_bin_t *bin, size_t min_run_size)
2141 {
2142 	size_t try_run_size, good_run_size;
2143 	unsigned good_nregs, good_mask_nelms, good_reg0_offset;
2144 	unsigned try_nregs, try_mask_nelms, try_reg0_offset;
2145 	float max_ovrhd = RUN_MAX_OVRHD;
2146 
2147 	assert(min_run_size >= pagesize);
2148 	assert(min_run_size <= arena_maxclass);
2149 	assert(min_run_size <= RUN_MAX_SMALL);
2150 
2151 	/*
2152 	 * Calculate known-valid settings before entering the run_size
2153 	 * expansion loop, so that the first part of the loop always copies
2154 	 * valid settings.
2155 	 *
2156 	 * The do..while loop iteratively reduces the number of regions until
2157 	 * the run header and the regions no longer overlap.  A closed formula
2158 	 * would be quite messy, since there is an interdependency between the
2159 	 * header's mask length and the number of regions.
2160 	 */
2161 	try_run_size = min_run_size;
2162 	try_nregs = (unsigned)(((try_run_size - sizeof(arena_run_t)) /
2163 	    bin->reg_size) + 1); /* Counter-act the first line of the loop. */
2164 	do {
2165 		try_nregs--;
2166 		try_mask_nelms = (try_nregs >> (SIZEOF_INT_2POW + 3)) +
2167 		    ((try_nregs & ((1 << (SIZEOF_INT_2POW + 3)) - 1)) ? 1 : 0);
2168 		try_reg0_offset = (unsigned)(try_run_size -
2169 		    (try_nregs * bin->reg_size));
2170 	} while (sizeof(arena_run_t) + (sizeof(unsigned) * (try_mask_nelms - 1))
2171 	    > try_reg0_offset);
2172 
2173 	/* run_size expansion loop. */
2174 	do {
2175 		/*
2176 		 * Copy valid settings before trying more aggressive settings.
2177 		 */
2178 		good_run_size = try_run_size;
2179 		good_nregs = try_nregs;
2180 		good_mask_nelms = try_mask_nelms;
2181 		good_reg0_offset = try_reg0_offset;
2182 
2183 		/* Try more aggressive settings. */
2184 		try_run_size += pagesize;
2185 		try_nregs = (unsigned)(((try_run_size - sizeof(arena_run_t)) /
2186 		    bin->reg_size) + 1); /* Counter-act try_nregs-- in loop. */
2187 		do {
2188 			try_nregs--;
2189 			try_mask_nelms = (try_nregs >> (SIZEOF_INT_2POW + 3)) +
2190 			    ((try_nregs & ((1 << (SIZEOF_INT_2POW + 3)) - 1)) ?
2191 			    1 : 0);
2192 			try_reg0_offset = (unsigned)(try_run_size - (try_nregs *
2193 			    bin->reg_size));
2194 		} while (sizeof(arena_run_t) + (sizeof(unsigned) *
2195 		    (try_mask_nelms - 1)) > try_reg0_offset);
2196 	} while (try_run_size <= arena_maxclass && try_run_size <= RUN_MAX_SMALL
2197 	    && max_ovrhd > RUN_MAX_OVRHD_RELAX / ((float)(bin->reg_size << 3))
2198 	    && ((float)(try_reg0_offset)) / ((float)(try_run_size)) >
2199 	    max_ovrhd);
2200 
2201 	assert(sizeof(arena_run_t) + (sizeof(unsigned) * (good_mask_nelms - 1))
2202 	    <= good_reg0_offset);
2203 	assert((good_mask_nelms << (SIZEOF_INT_2POW + 3)) >= good_nregs);
2204 
2205 	/* Copy final settings. */
2206 	bin->run_size = good_run_size;
2207 	bin->nregs = good_nregs;
2208 	bin->regs_mask_nelms = good_mask_nelms;
2209 	bin->reg0_offset = good_reg0_offset;
2210 
2211 	return (good_run_size);
2212 }
2213 
2214 static void *
2215 arena_malloc(arena_t *arena, size_t size)
2216 {
2217 	void *ret;
2218 
2219 	assert(arena != NULL);
2220 	assert(arena->magic == ARENA_MAGIC);
2221 	assert(size != 0);
2222 	assert(QUANTUM_CEILING(size) <= arena_maxclass);
2223 
2224 	if (size <= bin_maxclass) {
2225 		arena_bin_t *bin;
2226 		arena_run_t *run;
2227 
2228 		/* Small allocation. */
2229 
2230 		if (size < small_min) {
2231 			/* Tiny. */
2232 			size = pow2_ceil(size);
2233 			bin = &arena->bins[ffs((int)(size >> (TINY_MIN_2POW +
2234 			    1)))];
2235 #if (!defined(NDEBUG) || defined(MALLOC_STATS))
2236 			/*
2237 			 * Bin calculation is always correct, but we may need
2238 			 * to fix size for the purposes of assertions and/or
2239 			 * stats accuracy.
2240 			 */
2241 			if (size < (1 << TINY_MIN_2POW))
2242 				size = (1 << TINY_MIN_2POW);
2243 #endif
2244 		} else if (size <= small_max) {
2245 			/* Quantum-spaced. */
2246 			size = QUANTUM_CEILING(size);
2247 			bin = &arena->bins[ntbins + (size >> opt_quantum_2pow)
2248 			    - 1];
2249 		} else {
2250 			/* Sub-page. */
2251 			size = pow2_ceil(size);
2252 			bin = &arena->bins[ntbins + nqbins
2253 			    + (ffs((int)(size >> opt_small_max_2pow)) - 2)];
2254 		}
2255 		assert(size == bin->reg_size);
2256 
2257 		malloc_mutex_lock(&arena->mtx);
2258 		if ((run = bin->runcur) != NULL && run->nfree > 0)
2259 			ret = arena_bin_malloc_easy(arena, bin, run);
2260 		else
2261 			ret = arena_bin_malloc_hard(arena, bin);
2262 
2263 		if (ret == NULL) {
2264 			malloc_mutex_unlock(&arena->mtx);
2265 			return (NULL);
2266 		}
2267 
2268 #ifdef MALLOC_STATS
2269 		bin->stats.nrequests++;
2270 		arena->stats.nmalloc_small++;
2271 		arena->stats.allocated_small += size;
2272 #endif
2273 	} else {
2274 		/* Large allocation. */
2275 		size = PAGE_CEILING(size);
2276 		malloc_mutex_lock(&arena->mtx);
2277 		ret = (void *)arena_run_alloc(arena, size);
2278 		if (ret == NULL) {
2279 			malloc_mutex_unlock(&arena->mtx);
2280 			return (NULL);
2281 		}
2282 #ifdef MALLOC_STATS
2283 		arena->stats.nmalloc_large++;
2284 		arena->stats.allocated_large += size;
2285 #endif
2286 	}
2287 
2288 	malloc_mutex_unlock(&arena->mtx);
2289 
2290 	if (opt_junk)
2291 		memset(ret, 0xa5, size);
2292 	else if (opt_zero)
2293 		memset(ret, 0, size);
2294 	return (ret);
2295 }
2296 
2297 static inline void
2298 arena_palloc_trim(arena_t *arena, arena_chunk_t *chunk, unsigned pageind,
2299     unsigned npages)
2300 {
2301 	unsigned i;
2302 
2303 	assert(npages > 0);
2304 
2305 	/*
2306 	 * Modifiy the map such that arena_run_dalloc() sees the run as
2307 	 * separately allocated.
2308 	 */
2309 	for (i = 0; i < npages; i++) {
2310 		chunk->map[pageind + i].npages = npages;
2311 		chunk->map[pageind + i].pos = i;
2312 	}
2313 	arena_run_dalloc(arena, (arena_run_t *)((uintptr_t)chunk + (pageind <<
2314 	    pagesize_2pow)), npages << pagesize_2pow);
2315 }
2316 
2317 /* Only handles large allocations that require more than page alignment. */
2318 static void *
2319 arena_palloc(arena_t *arena, size_t alignment, size_t size, size_t alloc_size)
2320 {
2321 	void *ret;
2322 	size_t offset;
2323 	arena_chunk_t *chunk;
2324 	unsigned pageind, i, npages;
2325 
2326 	assert((size & pagesize_mask) == 0);
2327 	assert((alignment & pagesize_mask) == 0);
2328 
2329 	npages = (unsigned)(size >> pagesize_2pow);
2330 
2331 	malloc_mutex_lock(&arena->mtx);
2332 	ret = (void *)arena_run_alloc(arena, alloc_size);
2333 	if (ret == NULL) {
2334 		malloc_mutex_unlock(&arena->mtx);
2335 		return (NULL);
2336 	}
2337 
2338 	chunk = (arena_chunk_t *)CHUNK_ADDR2BASE(ret);
2339 
2340 	offset = (uintptr_t)ret & (alignment - 1);
2341 	assert((offset & pagesize_mask) == 0);
2342 	assert(offset < alloc_size);
2343 	if (offset == 0) {
2344 		pageind = (unsigned)(((uintptr_t)ret - (uintptr_t)chunk) >>
2345 		    pagesize_2pow);
2346 
2347 		/* Update the map for the run to be kept. */
2348 		for (i = 0; i < npages; i++) {
2349 			chunk->map[pageind + i].npages = npages;
2350 			assert(chunk->map[pageind + i].pos == i);
2351 		}
2352 
2353 		/* Trim trailing space. */
2354 		arena_palloc_trim(arena, chunk, pageind + npages,
2355 		    (unsigned)((alloc_size - size) >> pagesize_2pow));
2356 	} else {
2357 		size_t leadsize, trailsize;
2358 
2359 		leadsize = alignment - offset;
2360 		ret = (void *)((uintptr_t)ret + leadsize);
2361 		pageind = (unsigned)(((uintptr_t)ret - (uintptr_t)chunk) >>
2362 		    pagesize_2pow);
2363 
2364 		/* Update the map for the run to be kept. */
2365 		for (i = 0; i < npages; i++) {
2366 			chunk->map[pageind + i].npages = npages;
2367 			chunk->map[pageind + i].pos = i;
2368 		}
2369 
2370 		/* Trim leading space. */
2371 		arena_palloc_trim(arena, chunk,
2372 		    (unsigned)(pageind - (leadsize >> pagesize_2pow)),
2373 		    (unsigned)(leadsize >> pagesize_2pow));
2374 
2375 		trailsize = alloc_size - leadsize - size;
2376 		if (trailsize != 0) {
2377 			/* Trim trailing space. */
2378 			assert(trailsize < alloc_size);
2379 			arena_palloc_trim(arena, chunk, pageind + npages,
2380 			    (unsigned)(trailsize >> pagesize_2pow));
2381 		}
2382 	}
2383 
2384 #ifdef MALLOC_STATS
2385 	arena->stats.nmalloc_large++;
2386 	arena->stats.allocated_large += size;
2387 #endif
2388 	malloc_mutex_unlock(&arena->mtx);
2389 
2390 	if (opt_junk)
2391 		memset(ret, 0xa5, size);
2392 	else if (opt_zero)
2393 		memset(ret, 0, size);
2394 	return (ret);
2395 }
2396 
2397 /* Return the size of the allocation pointed to by ptr. */
2398 static size_t
2399 arena_salloc(const void *ptr)
2400 {
2401 	size_t ret;
2402 	arena_chunk_t *chunk;
2403 	arena_chunk_map_t *mapelm;
2404 	unsigned pageind;
2405 
2406 	assert(ptr != NULL);
2407 	assert(CHUNK_ADDR2BASE(ptr) != ptr);
2408 
2409 	/*
2410 	 * No arena data structures that we query here can change in a way that
2411 	 * affects this function, so we don't need to lock.
2412 	 */
2413 	chunk = (arena_chunk_t *)CHUNK_ADDR2BASE(ptr);
2414 	pageind = (unsigned)(((uintptr_t)ptr - (uintptr_t)chunk) >>
2415 	    pagesize_2pow);
2416 	mapelm = &chunk->map[pageind];
2417 	if (mapelm->pos != 0 || ptr != (char *)((uintptr_t)chunk) + (pageind <<
2418 	    pagesize_2pow)) {
2419 		arena_run_t *run;
2420 
2421 		pageind -= mapelm->pos;
2422 
2423 		run = (arena_run_t *)((uintptr_t)chunk + (pageind <<
2424 		    pagesize_2pow));
2425 		assert(run->magic == ARENA_RUN_MAGIC);
2426 		ret = run->bin->reg_size;
2427 	} else
2428 		ret = mapelm->npages << pagesize_2pow;
2429 
2430 	return (ret);
2431 }
2432 
2433 static void *
2434 arena_ralloc(void *ptr, size_t size, size_t oldsize)
2435 {
2436 	void *ret;
2437 
2438 	/* Avoid moving the allocation if the size class would not change. */
2439 	if (size < small_min) {
2440 		if (oldsize < small_min &&
2441 		    ffs((int)(pow2_ceil(size) >> (TINY_MIN_2POW + 1)))
2442 		    == ffs((int)(pow2_ceil(oldsize) >> (TINY_MIN_2POW + 1))))
2443 			goto IN_PLACE;
2444 	} else if (size <= small_max) {
2445 		if (oldsize >= small_min && oldsize <= small_max &&
2446 		    (QUANTUM_CEILING(size) >> opt_quantum_2pow)
2447 		    == (QUANTUM_CEILING(oldsize) >> opt_quantum_2pow))
2448 			goto IN_PLACE;
2449 	} else {
2450 		/*
2451 		 * We make no attempt to resize runs here, though it would be
2452 		 * possible to do so.
2453 		 */
2454 		if (oldsize > small_max && PAGE_CEILING(size) == oldsize)
2455 			goto IN_PLACE;
2456 	}
2457 
2458 	/*
2459 	 * If we get here, then size and oldsize are different enough that we
2460 	 * need to use a different size class.  In that case, fall back to
2461 	 * allocating new space and copying.
2462 	 */
2463 	ret = arena_malloc(choose_arena(), size);
2464 	if (ret == NULL)
2465 		return (NULL);
2466 
2467 	/* Junk/zero-filling were already done by arena_malloc(). */
2468 	if (size < oldsize)
2469 		memcpy(ret, ptr, size);
2470 	else
2471 		memcpy(ret, ptr, oldsize);
2472 	idalloc(ptr);
2473 	return (ret);
2474 IN_PLACE:
2475 	if (opt_junk && size < oldsize)
2476 		memset((void *)((uintptr_t)ptr + size), 0x5a, oldsize - size);
2477 	else if (opt_zero && size > oldsize)
2478 		memset((void *)((uintptr_t)ptr + oldsize), 0, size - oldsize);
2479 	return (ptr);
2480 }
2481 
2482 static void
2483 arena_dalloc(arena_t *arena, arena_chunk_t *chunk, void *ptr)
2484 {
2485 	unsigned pageind;
2486 	arena_chunk_map_t *mapelm;
2487 	size_t size;
2488 
2489 	assert(arena != NULL);
2490 	assert(arena->magic == ARENA_MAGIC);
2491 	assert(chunk->arena == arena);
2492 	assert(ptr != NULL);
2493 	assert(CHUNK_ADDR2BASE(ptr) != ptr);
2494 
2495 	pageind = (unsigned)(((uintptr_t)ptr - (uintptr_t)chunk) >>
2496 	    pagesize_2pow);
2497 	mapelm = &chunk->map[pageind];
2498 	if (mapelm->pos != 0 || ptr != (char *)((uintptr_t)chunk) + (pageind <<
2499 	    pagesize_2pow)) {
2500 		arena_run_t *run;
2501 		arena_bin_t *bin;
2502 
2503 		/* Small allocation. */
2504 
2505 		pageind -= mapelm->pos;
2506 
2507 		run = (arena_run_t *)((uintptr_t)chunk + (pageind <<
2508 		    pagesize_2pow));
2509 		assert(run->magic == ARENA_RUN_MAGIC);
2510 		bin = run->bin;
2511 		size = bin->reg_size;
2512 
2513 		if (opt_junk)
2514 			memset(ptr, 0x5a, size);
2515 
2516 		malloc_mutex_lock(&arena->mtx);
2517 		arena_run_reg_dalloc(run, bin, ptr, size);
2518 		run->nfree++;
2519 
2520 		if (run->nfree == bin->nregs) {
2521 			/* Deallocate run. */
2522 			if (run == bin->runcur)
2523 				bin->runcur = NULL;
2524 			else if (bin->nregs != 1) {
2525 				/*
2526 				 * This block's conditional is necessary because
2527 				 * if the run only contains one region, then it
2528 				 * never gets inserted into the non-full runs
2529 				 * tree.
2530 				 */
2531 				/* LINTED */
2532 				RB_REMOVE(arena_run_tree_s, &bin->runs, run);
2533 			}
2534 #ifdef MALLOC_DEBUG
2535 			run->magic = 0;
2536 #endif
2537 			arena_run_dalloc(arena, run, bin->run_size);
2538 #ifdef MALLOC_STATS
2539 			bin->stats.curruns--;
2540 #endif
2541 		} else if (run->nfree == 1 && run != bin->runcur) {
2542 			/*
2543 			 * Make sure that bin->runcur always refers to the
2544 			 * lowest non-full run, if one exists.
2545 			 */
2546 			if (bin->runcur == NULL)
2547 				bin->runcur = run;
2548 			else if ((uintptr_t)run < (uintptr_t)bin->runcur) {
2549 				/* Switch runcur. */
2550 				if (bin->runcur->nfree > 0) {
2551 					/* Insert runcur. */
2552 					/* LINTED */
2553 					RB_INSERT(arena_run_tree_s, &bin->runs,
2554 					    bin->runcur);
2555 				}
2556 				bin->runcur = run;
2557 			} else {
2558 				/* LINTED */
2559 				RB_INSERT(arena_run_tree_s, &bin->runs, run);
2560 			}
2561 		}
2562 #ifdef MALLOC_STATS
2563 		arena->stats.allocated_small -= size;
2564 		arena->stats.ndalloc_small++;
2565 #endif
2566 	} else {
2567 		/* Large allocation. */
2568 
2569 		size = mapelm->npages << pagesize_2pow;
2570 		assert((((uintptr_t)ptr) & pagesize_mask) == 0);
2571 
2572 		if (opt_junk)
2573 			memset(ptr, 0x5a, size);
2574 
2575 		malloc_mutex_lock(&arena->mtx);
2576 		arena_run_dalloc(arena, (arena_run_t *)ptr, size);
2577 #ifdef MALLOC_STATS
2578 		arena->stats.allocated_large -= size;
2579 		arena->stats.ndalloc_large++;
2580 #endif
2581 	}
2582 
2583 	malloc_mutex_unlock(&arena->mtx);
2584 }
2585 
2586 static bool
2587 arena_new(arena_t *arena)
2588 {
2589 	unsigned i;
2590 	arena_bin_t *bin;
2591 	size_t prev_run_size;
2592 
2593 	malloc_mutex_init(&arena->mtx);
2594 
2595 #ifdef MALLOC_STATS
2596 	memset(&arena->stats, 0, sizeof(arena_stats_t));
2597 #endif
2598 
2599 	/* Initialize chunks. */
2600 	RB_INIT(&arena->chunks);
2601 	arena->spare = NULL;
2602 
2603 	/* Initialize bins. */
2604 	prev_run_size = pagesize;
2605 
2606 	/* (2^n)-spaced tiny bins. */
2607 	for (i = 0; i < ntbins; i++) {
2608 		bin = &arena->bins[i];
2609 		bin->runcur = NULL;
2610 		RB_INIT(&bin->runs);
2611 
2612 		bin->reg_size = (1 << (TINY_MIN_2POW + i));
2613 		prev_run_size = arena_bin_run_size_calc(bin, prev_run_size);
2614 
2615 #ifdef MALLOC_STATS
2616 		memset(&bin->stats, 0, sizeof(malloc_bin_stats_t));
2617 #endif
2618 	}
2619 
2620 	/* Quantum-spaced bins. */
2621 	for (; i < ntbins + nqbins; i++) {
2622 		bin = &arena->bins[i];
2623 		bin->runcur = NULL;
2624 		RB_INIT(&bin->runs);
2625 
2626 		bin->reg_size = quantum * (i - ntbins + 1);
2627 /*
2628 		pow2_size = pow2_ceil(quantum * (i - ntbins + 1));
2629 */
2630 		prev_run_size = arena_bin_run_size_calc(bin, prev_run_size);
2631 
2632 #ifdef MALLOC_STATS
2633 		memset(&bin->stats, 0, sizeof(malloc_bin_stats_t));
2634 #endif
2635 	}
2636 
2637 	/* (2^n)-spaced sub-page bins. */
2638 	for (; i < ntbins + nqbins + nsbins; i++) {
2639 		bin = &arena->bins[i];
2640 		bin->runcur = NULL;
2641 		RB_INIT(&bin->runs);
2642 
2643 		bin->reg_size = (small_max << (i - (ntbins + nqbins) + 1));
2644 
2645 		prev_run_size = arena_bin_run_size_calc(bin, prev_run_size);
2646 
2647 #ifdef MALLOC_STATS
2648 		memset(&bin->stats, 0, sizeof(malloc_bin_stats_t));
2649 #endif
2650 	}
2651 
2652 #ifdef MALLOC_DEBUG
2653 	arena->magic = ARENA_MAGIC;
2654 #endif
2655 
2656 	return (false);
2657 }
2658 
2659 /* Create a new arena and insert it into the arenas array at index ind. */
2660 static arena_t *
2661 arenas_extend(unsigned ind)
2662 {
2663 	arena_t *ret;
2664 
2665 	/* Allocate enough space for trailing bins. */
2666 	ret = (arena_t *)base_alloc(sizeof(arena_t)
2667 	    + (sizeof(arena_bin_t) * (ntbins + nqbins + nsbins - 1)));
2668 	if (ret != NULL && arena_new(ret) == false) {
2669 		arenas[ind] = ret;
2670 		return (ret);
2671 	}
2672 	/* Only reached if there is an OOM error. */
2673 
2674 	/*
2675 	 * OOM here is quite inconvenient to propagate, since dealing with it
2676 	 * would require a check for failure in the fast path.  Instead, punt
2677 	 * by using arenas[0].  In practice, this is an extremely unlikely
2678 	 * failure.
2679 	 */
2680 	_malloc_message(getprogname(),
2681 	    ": (malloc) Error initializing arena\n", "", "");
2682 	if (opt_abort)
2683 		abort();
2684 
2685 	return (arenas[0]);
2686 }
2687 
2688 /*
2689  * End arena.
2690  */
2691 /******************************************************************************/
2692 /*
2693  * Begin general internal functions.
2694  */
2695 
2696 static void *
2697 huge_malloc(size_t size)
2698 {
2699 	void *ret;
2700 	size_t csize;
2701 	chunk_node_t *node;
2702 
2703 	/* Allocate one or more contiguous chunks for this request. */
2704 
2705 	csize = CHUNK_CEILING(size);
2706 	if (csize == 0) {
2707 		/* size is large enough to cause size_t wrap-around. */
2708 		return (NULL);
2709 	}
2710 
2711 	/* Allocate a chunk node with which to track the chunk. */
2712 	node = base_chunk_node_alloc();
2713 	if (node == NULL)
2714 		return (NULL);
2715 
2716 	ret = chunk_alloc(csize);
2717 	if (ret == NULL) {
2718 		base_chunk_node_dealloc(node);
2719 		return (NULL);
2720 	}
2721 
2722 	/* Insert node into huge. */
2723 	node->chunk = ret;
2724 	node->size = csize;
2725 
2726 	malloc_mutex_lock(&chunks_mtx);
2727 	RB_INSERT(chunk_tree_s, &huge, node);
2728 #ifdef MALLOC_STATS
2729 	huge_nmalloc++;
2730 	huge_allocated += csize;
2731 #endif
2732 	malloc_mutex_unlock(&chunks_mtx);
2733 
2734 	if (opt_junk)
2735 		memset(ret, 0xa5, csize);
2736 	else if (opt_zero)
2737 		memset(ret, 0, csize);
2738 
2739 	return (ret);
2740 }
2741 
2742 /* Only handles large allocations that require more than chunk alignment. */
2743 static void *
2744 huge_palloc(size_t alignment, size_t size)
2745 {
2746 	void *ret;
2747 	size_t alloc_size, chunk_size, offset;
2748 	chunk_node_t *node;
2749 
2750 	/*
2751 	 * This allocation requires alignment that is even larger than chunk
2752 	 * alignment.  This means that huge_malloc() isn't good enough.
2753 	 *
2754 	 * Allocate almost twice as many chunks as are demanded by the size or
2755 	 * alignment, in order to assure the alignment can be achieved, then
2756 	 * unmap leading and trailing chunks.
2757 	 */
2758 	assert(alignment >= chunksize);
2759 
2760 	chunk_size = CHUNK_CEILING(size);
2761 
2762 	if (size >= alignment)
2763 		alloc_size = chunk_size + alignment - chunksize;
2764 	else
2765 		alloc_size = (alignment << 1) - chunksize;
2766 
2767 	/* Allocate a chunk node with which to track the chunk. */
2768 	node = base_chunk_node_alloc();
2769 	if (node == NULL)
2770 		return (NULL);
2771 
2772 	ret = chunk_alloc(alloc_size);
2773 	if (ret == NULL) {
2774 		base_chunk_node_dealloc(node);
2775 		return (NULL);
2776 	}
2777 
2778 	offset = (uintptr_t)ret & (alignment - 1);
2779 	assert((offset & chunksize_mask) == 0);
2780 	assert(offset < alloc_size);
2781 	if (offset == 0) {
2782 		/* Trim trailing space. */
2783 		chunk_dealloc((void *)((uintptr_t)ret + chunk_size), alloc_size
2784 		    - chunk_size);
2785 	} else {
2786 		size_t trailsize;
2787 
2788 		/* Trim leading space. */
2789 		chunk_dealloc(ret, alignment - offset);
2790 
2791 		ret = (void *)((uintptr_t)ret + (alignment - offset));
2792 
2793 		trailsize = alloc_size - (alignment - offset) - chunk_size;
2794 		if (trailsize != 0) {
2795 		    /* Trim trailing space. */
2796 		    assert(trailsize < alloc_size);
2797 		    chunk_dealloc((void *)((uintptr_t)ret + chunk_size),
2798 			trailsize);
2799 		}
2800 	}
2801 
2802 	/* Insert node into huge. */
2803 	node->chunk = ret;
2804 	node->size = chunk_size;
2805 
2806 	malloc_mutex_lock(&chunks_mtx);
2807 	RB_INSERT(chunk_tree_s, &huge, node);
2808 #ifdef MALLOC_STATS
2809 	huge_nmalloc++;
2810 	huge_allocated += chunk_size;
2811 #endif
2812 	malloc_mutex_unlock(&chunks_mtx);
2813 
2814 	if (opt_junk)
2815 		memset(ret, 0xa5, chunk_size);
2816 	else if (opt_zero)
2817 		memset(ret, 0, chunk_size);
2818 
2819 	return (ret);
2820 }
2821 
2822 static void *
2823 huge_ralloc(void *ptr, size_t size, size_t oldsize)
2824 {
2825 	void *ret;
2826 
2827 	/* Avoid moving the allocation if the size class would not change. */
2828 	if (oldsize > arena_maxclass &&
2829 	    CHUNK_CEILING(size) == CHUNK_CEILING(oldsize)) {
2830 		if (opt_junk && size < oldsize) {
2831 			memset((void *)((uintptr_t)ptr + size), 0x5a, oldsize
2832 			    - size);
2833 		} else if (opt_zero && size > oldsize) {
2834 			memset((void *)((uintptr_t)ptr + oldsize), 0, size
2835 			    - oldsize);
2836 		}
2837 		return (ptr);
2838 	}
2839 
2840 	if (CHUNK_ADDR2BASE(ptr) == ptr
2841 #ifdef USE_BRK
2842 	    && ((uintptr_t)ptr < (uintptr_t)brk_base
2843 	    || (uintptr_t)ptr >= (uintptr_t)brk_max)
2844 #endif
2845 	    ) {
2846 		chunk_node_t *node, key;
2847 		void *newptr;
2848 		size_t oldcsize;
2849 		size_t newcsize;
2850 
2851 		newcsize = CHUNK_CEILING(size);
2852 		oldcsize = CHUNK_CEILING(oldsize);
2853 		assert(oldcsize != newcsize);
2854 		if (newcsize == 0) {
2855 			/* size_t wrap-around */
2856 			return (NULL);
2857 		}
2858 		newptr = mremap(ptr, oldcsize, NULL, newcsize,
2859 		    MAP_ALIGNED(chunksize_2pow));
2860 		if (newptr != MAP_FAILED) {
2861 			assert(CHUNK_ADDR2BASE(newptr) == newptr);
2862 
2863 			/* update tree */
2864 			malloc_mutex_lock(&chunks_mtx);
2865 			key.chunk = __DECONST(void *, ptr);
2866 			/* LINTED */
2867 			node = RB_FIND(chunk_tree_s, &huge, &key);
2868 			assert(node != NULL);
2869 			assert(node->chunk == ptr);
2870 			assert(node->size == oldcsize);
2871 			node->size = newcsize;
2872 			if (ptr != newptr) {
2873 				RB_REMOVE(chunk_tree_s, &huge, node);
2874 				node->chunk = newptr;
2875 				RB_INSERT(chunk_tree_s, &huge, node);
2876 			}
2877 #ifdef MALLOC_STATS
2878 			huge_nralloc++;
2879 			huge_allocated += newcsize - oldcsize;
2880 			if (newcsize > oldcsize) {
2881 				stats_chunks.curchunks +=
2882 				    (newcsize - oldcsize) / chunksize;
2883 				if (stats_chunks.curchunks >
2884 				    stats_chunks.highchunks)
2885 					stats_chunks.highchunks =
2886 					    stats_chunks.curchunks;
2887 			} else {
2888 				stats_chunks.curchunks -=
2889 				    (oldcsize - newcsize) / chunksize;
2890 			}
2891 #endif
2892 			malloc_mutex_unlock(&chunks_mtx);
2893 
2894 			if (opt_junk && size < oldsize) {
2895 				memset((void *)((uintptr_t)newptr + size), 0x5a,
2896 				    newcsize - size);
2897 			} else if (opt_zero && size > oldsize) {
2898 				memset((void *)((uintptr_t)newptr + oldsize), 0,
2899 				    size - oldsize);
2900 			}
2901 			return (newptr);
2902 		}
2903 	}
2904 
2905 	/*
2906 	 * If we get here, then size and oldsize are different enough that we
2907 	 * need to use a different size class.  In that case, fall back to
2908 	 * allocating new space and copying.
2909 	 */
2910 	ret = huge_malloc(size);
2911 	if (ret == NULL)
2912 		return (NULL);
2913 
2914 	if (CHUNK_ADDR2BASE(ptr) == ptr) {
2915 		/* The old allocation is a chunk. */
2916 		if (size < oldsize)
2917 			memcpy(ret, ptr, size);
2918 		else
2919 			memcpy(ret, ptr, oldsize);
2920 	} else {
2921 		/* The old allocation is a region. */
2922 		assert(oldsize < size);
2923 		memcpy(ret, ptr, oldsize);
2924 	}
2925 	idalloc(ptr);
2926 	return (ret);
2927 }
2928 
2929 static void
2930 huge_dalloc(void *ptr)
2931 {
2932 	chunk_node_t key;
2933 	chunk_node_t *node;
2934 
2935 	malloc_mutex_lock(&chunks_mtx);
2936 
2937 	/* Extract from tree of huge allocations. */
2938 	key.chunk = ptr;
2939 	/* LINTED */
2940 	node = RB_FIND(chunk_tree_s, &huge, &key);
2941 	assert(node != NULL);
2942 	assert(node->chunk == ptr);
2943 	/* LINTED */
2944 	RB_REMOVE(chunk_tree_s, &huge, node);
2945 
2946 #ifdef MALLOC_STATS
2947 	huge_ndalloc++;
2948 	huge_allocated -= node->size;
2949 #endif
2950 
2951 	malloc_mutex_unlock(&chunks_mtx);
2952 
2953 	/* Unmap chunk. */
2954 #ifdef USE_BRK
2955 	if (opt_junk)
2956 		memset(node->chunk, 0x5a, node->size);
2957 #endif
2958 	chunk_dealloc(node->chunk, node->size);
2959 
2960 	base_chunk_node_dealloc(node);
2961 }
2962 
2963 static void *
2964 imalloc(size_t size)
2965 {
2966 	void *ret;
2967 
2968 	assert(size != 0);
2969 
2970 	if (size <= arena_maxclass)
2971 		ret = arena_malloc(choose_arena(), size);
2972 	else
2973 		ret = huge_malloc(size);
2974 
2975 	return (ret);
2976 }
2977 
2978 static void *
2979 ipalloc(size_t alignment, size_t size)
2980 {
2981 	void *ret;
2982 	size_t ceil_size;
2983 
2984 	/*
2985 	 * Round size up to the nearest multiple of alignment.
2986 	 *
2987 	 * This done, we can take advantage of the fact that for each small
2988 	 * size class, every object is aligned at the smallest power of two
2989 	 * that is non-zero in the base two representation of the size.  For
2990 	 * example:
2991 	 *
2992 	 *   Size |   Base 2 | Minimum alignment
2993 	 *   -----+----------+------------------
2994 	 *     96 |  1100000 |  32
2995 	 *    144 | 10100000 |  32
2996 	 *    192 | 11000000 |  64
2997 	 *
2998 	 * Depending on runtime settings, it is possible that arena_malloc()
2999 	 * will further round up to a power of two, but that never causes
3000 	 * correctness issues.
3001 	 */
3002 	ceil_size = (size + (alignment - 1)) & (-alignment);
3003 	/*
3004 	 * (ceil_size < size) protects against the combination of maximal
3005 	 * alignment and size greater than maximal alignment.
3006 	 */
3007 	if (ceil_size < size) {
3008 		/* size_t overflow. */
3009 		return (NULL);
3010 	}
3011 
3012 	if (ceil_size <= pagesize || (alignment <= pagesize
3013 	    && ceil_size <= arena_maxclass))
3014 		ret = arena_malloc(choose_arena(), ceil_size);
3015 	else {
3016 		size_t run_size;
3017 
3018 		/*
3019 		 * We can't achieve sub-page alignment, so round up alignment
3020 		 * permanently; it makes later calculations simpler.
3021 		 */
3022 		alignment = PAGE_CEILING(alignment);
3023 		ceil_size = PAGE_CEILING(size);
3024 		/*
3025 		 * (ceil_size < size) protects against very large sizes within
3026 		 * pagesize of SIZE_T_MAX.
3027 		 *
3028 		 * (ceil_size + alignment < ceil_size) protects against the
3029 		 * combination of maximal alignment and ceil_size large enough
3030 		 * to cause overflow.  This is similar to the first overflow
3031 		 * check above, but it needs to be repeated due to the new
3032 		 * ceil_size value, which may now be *equal* to maximal
3033 		 * alignment, whereas before we only detected overflow if the
3034 		 * original size was *greater* than maximal alignment.
3035 		 */
3036 		if (ceil_size < size || ceil_size + alignment < ceil_size) {
3037 			/* size_t overflow. */
3038 			return (NULL);
3039 		}
3040 
3041 		/*
3042 		 * Calculate the size of the over-size run that arena_palloc()
3043 		 * would need to allocate in order to guarantee the alignment.
3044 		 */
3045 		if (ceil_size >= alignment)
3046 			run_size = ceil_size + alignment - pagesize;
3047 		else {
3048 			/*
3049 			 * It is possible that (alignment << 1) will cause
3050 			 * overflow, but it doesn't matter because we also
3051 			 * subtract pagesize, which in the case of overflow
3052 			 * leaves us with a very large run_size.  That causes
3053 			 * the first conditional below to fail, which means
3054 			 * that the bogus run_size value never gets used for
3055 			 * anything important.
3056 			 */
3057 			run_size = (alignment << 1) - pagesize;
3058 		}
3059 
3060 		if (run_size <= arena_maxclass) {
3061 			ret = arena_palloc(choose_arena(), alignment, ceil_size,
3062 			    run_size);
3063 		} else if (alignment <= chunksize)
3064 			ret = huge_malloc(ceil_size);
3065 		else
3066 			ret = huge_palloc(alignment, ceil_size);
3067 	}
3068 
3069 	assert(((uintptr_t)ret & (alignment - 1)) == 0);
3070 	return (ret);
3071 }
3072 
3073 static void *
3074 icalloc(size_t size)
3075 {
3076 	void *ret;
3077 
3078 	if (size <= arena_maxclass) {
3079 		ret = arena_malloc(choose_arena(), size);
3080 		if (ret == NULL)
3081 			return (NULL);
3082 		memset(ret, 0, size);
3083 	} else {
3084 		/*
3085 		 * The virtual memory system provides zero-filled pages, so
3086 		 * there is no need to do so manually, unless opt_junk is
3087 		 * enabled, in which case huge_malloc() fills huge allocations
3088 		 * with junk.
3089 		 */
3090 		ret = huge_malloc(size);
3091 		if (ret == NULL)
3092 			return (NULL);
3093 
3094 		if (opt_junk)
3095 			memset(ret, 0, size);
3096 #ifdef USE_BRK
3097 		else if ((uintptr_t)ret >= (uintptr_t)brk_base
3098 		    && (uintptr_t)ret < (uintptr_t)brk_max) {
3099 			/*
3100 			 * This may be a re-used brk chunk.  Therefore, zero
3101 			 * the memory.
3102 			 */
3103 			memset(ret, 0, size);
3104 		}
3105 #endif
3106 	}
3107 
3108 	return (ret);
3109 }
3110 
3111 static size_t
3112 isalloc(const void *ptr)
3113 {
3114 	size_t ret;
3115 	arena_chunk_t *chunk;
3116 
3117 	assert(ptr != NULL);
3118 
3119 	chunk = (arena_chunk_t *)CHUNK_ADDR2BASE(ptr);
3120 	if (chunk != ptr) {
3121 		/* Region. */
3122 		assert(chunk->arena->magic == ARENA_MAGIC);
3123 
3124 		ret = arena_salloc(ptr);
3125 	} else {
3126 		chunk_node_t *node, key;
3127 
3128 		/* Chunk (huge allocation). */
3129 
3130 		malloc_mutex_lock(&chunks_mtx);
3131 
3132 		/* Extract from tree of huge allocations. */
3133 		key.chunk = __DECONST(void *, ptr);
3134 		/* LINTED */
3135 		node = RB_FIND(chunk_tree_s, &huge, &key);
3136 		assert(node != NULL);
3137 
3138 		ret = node->size;
3139 
3140 		malloc_mutex_unlock(&chunks_mtx);
3141 	}
3142 
3143 	return (ret);
3144 }
3145 
3146 static void *
3147 iralloc(void *ptr, size_t size)
3148 {
3149 	void *ret;
3150 	size_t oldsize;
3151 
3152 	assert(ptr != NULL);
3153 	assert(size != 0);
3154 
3155 	oldsize = isalloc(ptr);
3156 
3157 	if (size <= arena_maxclass)
3158 		ret = arena_ralloc(ptr, size, oldsize);
3159 	else
3160 		ret = huge_ralloc(ptr, size, oldsize);
3161 
3162 	return (ret);
3163 }
3164 
3165 static void
3166 idalloc(void *ptr)
3167 {
3168 	arena_chunk_t *chunk;
3169 
3170 	assert(ptr != NULL);
3171 
3172 	chunk = (arena_chunk_t *)CHUNK_ADDR2BASE(ptr);
3173 	if (chunk != ptr) {
3174 		/* Region. */
3175 		arena_dalloc(chunk->arena, chunk, ptr);
3176 	} else
3177 		huge_dalloc(ptr);
3178 }
3179 
3180 static void
3181 malloc_print_stats(void)
3182 {
3183 
3184 	if (opt_print_stats) {
3185 		char s[UMAX2S_BUFSIZE];
3186 		_malloc_message("___ Begin malloc statistics ___\n", "", "",
3187 		    "");
3188 		_malloc_message("Assertions ",
3189 #ifdef NDEBUG
3190 		    "disabled",
3191 #else
3192 		    "enabled",
3193 #endif
3194 		    "\n", "");
3195 		_malloc_message("Boolean MALLOC_OPTIONS: ",
3196 		    opt_abort ? "A" : "a",
3197 		    opt_junk ? "J" : "j",
3198 		    opt_hint ? "H" : "h");
3199 		_malloc_message(opt_utrace ? "PU" : "Pu",
3200 		    opt_sysv ? "V" : "v",
3201 		    opt_xmalloc ? "X" : "x",
3202 		    opt_zero ? "Z\n" : "z\n");
3203 
3204 		_malloc_message("CPUs: ", umax2s(ncpus, s), "\n", "");
3205 		_malloc_message("Max arenas: ", umax2s(narenas, s), "\n", "");
3206 		_malloc_message("Pointer size: ", umax2s(sizeof(void *), s),
3207 		    "\n", "");
3208 		_malloc_message("Quantum size: ", umax2s(quantum, s), "\n", "");
3209 		_malloc_message("Max small size: ", umax2s(small_max, s), "\n",
3210 		    "");
3211 
3212 		_malloc_message("Chunk size: ", umax2s(chunksize, s), "", "");
3213 		_malloc_message(" (2^", umax2s(opt_chunk_2pow, s), ")\n", "");
3214 
3215 #ifdef MALLOC_STATS
3216 		{
3217 			size_t allocated, mapped;
3218 			unsigned i;
3219 			arena_t *arena;
3220 
3221 			/* Calculate and print allocated/mapped stats. */
3222 
3223 			/* arenas. */
3224 			for (i = 0, allocated = 0; i < narenas; i++) {
3225 				if (arenas[i] != NULL) {
3226 					malloc_mutex_lock(&arenas[i]->mtx);
3227 					allocated +=
3228 					    arenas[i]->stats.allocated_small;
3229 					allocated +=
3230 					    arenas[i]->stats.allocated_large;
3231 					malloc_mutex_unlock(&arenas[i]->mtx);
3232 				}
3233 			}
3234 
3235 			/* huge/base. */
3236 			malloc_mutex_lock(&chunks_mtx);
3237 			allocated += huge_allocated;
3238 			mapped = stats_chunks.curchunks * chunksize;
3239 			malloc_mutex_unlock(&chunks_mtx);
3240 
3241 			malloc_mutex_lock(&base_mtx);
3242 			mapped += base_mapped;
3243 			malloc_mutex_unlock(&base_mtx);
3244 
3245 			malloc_printf("Allocated: %zu, mapped: %zu\n",
3246 			    allocated, mapped);
3247 
3248 			/* Print chunk stats. */
3249 			{
3250 				chunk_stats_t chunks_stats;
3251 
3252 				malloc_mutex_lock(&chunks_mtx);
3253 				chunks_stats = stats_chunks;
3254 				malloc_mutex_unlock(&chunks_mtx);
3255 
3256 				malloc_printf("chunks: nchunks   "
3257 				    "highchunks    curchunks\n");
3258 				malloc_printf("  %13llu%13lu%13lu\n",
3259 				    chunks_stats.nchunks,
3260 				    chunks_stats.highchunks,
3261 				    chunks_stats.curchunks);
3262 			}
3263 
3264 			/* Print chunk stats. */
3265 			malloc_printf(
3266 			    "huge: nmalloc      ndalloc      "
3267 			    "nralloc    allocated\n");
3268 			malloc_printf(" %12llu %12llu %12llu %12zu\n",
3269 			    huge_nmalloc, huge_ndalloc, huge_nralloc,
3270 			    huge_allocated);
3271 
3272 			/* Print stats for each arena. */
3273 			for (i = 0; i < narenas; i++) {
3274 				arena = arenas[i];
3275 				if (arena != NULL) {
3276 					malloc_printf(
3277 					    "\narenas[%u] @ %p\n", i, arena);
3278 					malloc_mutex_lock(&arena->mtx);
3279 					stats_print(arena);
3280 					malloc_mutex_unlock(&arena->mtx);
3281 				}
3282 			}
3283 		}
3284 #endif /* #ifdef MALLOC_STATS */
3285 		_malloc_message("--- End malloc statistics ---\n", "", "", "");
3286 	}
3287 }
3288 
3289 /*
3290  * FreeBSD's pthreads implementation calls malloc(3), so the malloc
3291  * implementation has to take pains to avoid infinite recursion during
3292  * initialization.
3293  */
3294 static inline bool
3295 malloc_init(void)
3296 {
3297 
3298 	if (malloc_initialized == false)
3299 		return (malloc_init_hard());
3300 
3301 	return (false);
3302 }
3303 
3304 static bool
3305 malloc_init_hard(void)
3306 {
3307 	unsigned i, j;
3308 	ssize_t linklen;
3309 	char buf[PATH_MAX + 1];
3310 	const char *opts = "";
3311 
3312 	malloc_mutex_lock(&init_lock);
3313 	if (malloc_initialized) {
3314 		/*
3315 		 * Another thread initialized the allocator before this one
3316 		 * acquired init_lock.
3317 		 */
3318 		malloc_mutex_unlock(&init_lock);
3319 		return (false);
3320 	}
3321 
3322 	/* Get number of CPUs. */
3323 	{
3324 		int mib[2];
3325 		size_t len;
3326 
3327 		mib[0] = CTL_HW;
3328 		mib[1] = HW_NCPU;
3329 		len = sizeof(ncpus);
3330 		if (sysctl(mib, 2, &ncpus, &len, (void *) 0, 0) == -1) {
3331 			/* Error. */
3332 			ncpus = 1;
3333 		}
3334 	}
3335 
3336 	/* Get page size. */
3337 	{
3338 		long result;
3339 
3340 		result = sysconf(_SC_PAGESIZE);
3341 		assert(result != -1);
3342 		pagesize = (unsigned) result;
3343 
3344 		/*
3345 		 * We assume that pagesize is a power of 2 when calculating
3346 		 * pagesize_mask and pagesize_2pow.
3347 		 */
3348 		assert(((result - 1) & result) == 0);
3349 		pagesize_mask = result - 1;
3350 		pagesize_2pow = ffs((int)result) - 1;
3351 	}
3352 
3353 	for (i = 0; i < 3; i++) {
3354 		/* Get runtime configuration. */
3355 		switch (i) {
3356 		case 0:
3357 			if ((linklen = readlink("/etc/malloc.conf", buf,
3358 						sizeof(buf) - 1)) != -1) {
3359 				/*
3360 				 * Use the contents of the "/etc/malloc.conf"
3361 				 * symbolic link's name.
3362 				 */
3363 				buf[linklen] = '\0';
3364 				opts = buf;
3365 			} else {
3366 				/* No configuration specified. */
3367 				buf[0] = '\0';
3368 				opts = buf;
3369 			}
3370 			break;
3371 		case 1:
3372 			if ((opts = getenv("MALLOC_OPTIONS")) != NULL &&
3373 			    issetugid() == 0) {
3374 				/*
3375 				 * Do nothing; opts is already initialized to
3376 				 * the value of the MALLOC_OPTIONS environment
3377 				 * variable.
3378 				 */
3379 			} else {
3380 				/* No configuration specified. */
3381 				buf[0] = '\0';
3382 				opts = buf;
3383 			}
3384 			break;
3385 		case 2:
3386 			if (_malloc_options != NULL) {
3387 			    /*
3388 			     * Use options that were compiled into the program.
3389 			     */
3390 			    opts = _malloc_options;
3391 			} else {
3392 				/* No configuration specified. */
3393 				buf[0] = '\0';
3394 				opts = buf;
3395 			}
3396 			break;
3397 		default:
3398 			/* NOTREACHED */
3399 			/* LINTED */
3400 			assert(false);
3401 		}
3402 
3403 		for (j = 0; opts[j] != '\0'; j++) {
3404 			switch (opts[j]) {
3405 			case 'a':
3406 				opt_abort = false;
3407 				break;
3408 			case 'A':
3409 				opt_abort = true;
3410 				break;
3411 			case 'h':
3412 				opt_hint = false;
3413 				break;
3414 			case 'H':
3415 				opt_hint = true;
3416 				break;
3417 			case 'j':
3418 				opt_junk = false;
3419 				break;
3420 			case 'J':
3421 				opt_junk = true;
3422 				break;
3423 			case 'k':
3424 				/*
3425 				 * Chunks always require at least one header
3426 				 * page, so chunks can never be smaller than
3427 				 * two pages.
3428 				 */
3429 				if (opt_chunk_2pow > pagesize_2pow + 1)
3430 					opt_chunk_2pow--;
3431 				break;
3432 			case 'K':
3433 				/*
3434 				 * There must be fewer pages in a chunk than
3435 				 * can be recorded by the pos field of
3436 				 * arena_chunk_map_t, in order to make POS_FREE
3437 				 * special.
3438 				 */
3439 				if (opt_chunk_2pow - pagesize_2pow
3440 				    < (sizeof(uint32_t) << 3) - 1)
3441 					opt_chunk_2pow++;
3442 				break;
3443 			case 'n':
3444 				opt_narenas_lshift--;
3445 				break;
3446 			case 'N':
3447 				opt_narenas_lshift++;
3448 				break;
3449 			case 'p':
3450 				opt_print_stats = false;
3451 				break;
3452 			case 'P':
3453 				opt_print_stats = true;
3454 				break;
3455 			case 'q':
3456 				if (opt_quantum_2pow > QUANTUM_2POW_MIN)
3457 					opt_quantum_2pow--;
3458 				break;
3459 			case 'Q':
3460 				if (opt_quantum_2pow < pagesize_2pow - 1)
3461 					opt_quantum_2pow++;
3462 				break;
3463 			case 's':
3464 				if (opt_small_max_2pow > QUANTUM_2POW_MIN)
3465 					opt_small_max_2pow--;
3466 				break;
3467 			case 'S':
3468 				if (opt_small_max_2pow < pagesize_2pow - 1)
3469 					opt_small_max_2pow++;
3470 				break;
3471 			case 'u':
3472 				opt_utrace = false;
3473 				break;
3474 			case 'U':
3475 				opt_utrace = true;
3476 				break;
3477 			case 'v':
3478 				opt_sysv = false;
3479 				break;
3480 			case 'V':
3481 				opt_sysv = true;
3482 				break;
3483 			case 'x':
3484 				opt_xmalloc = false;
3485 				break;
3486 			case 'X':
3487 				opt_xmalloc = true;
3488 				break;
3489 			case 'z':
3490 				opt_zero = false;
3491 				break;
3492 			case 'Z':
3493 				opt_zero = true;
3494 				break;
3495 			default: {
3496 				char cbuf[2];
3497 
3498 				cbuf[0] = opts[j];
3499 				cbuf[1] = '\0';
3500 				_malloc_message(getprogname(),
3501 				    ": (malloc) Unsupported character in "
3502 				    "malloc options: '", cbuf, "'\n");
3503 			}
3504 			}
3505 		}
3506 	}
3507 
3508 	/* Take care to call atexit() only once. */
3509 	if (opt_print_stats) {
3510 		/* Print statistics at exit. */
3511 		atexit(malloc_print_stats);
3512 	}
3513 
3514 	/* Set variables according to the value of opt_small_max_2pow. */
3515 	if (opt_small_max_2pow < opt_quantum_2pow)
3516 		opt_small_max_2pow = opt_quantum_2pow;
3517 	small_max = (1 << opt_small_max_2pow);
3518 
3519 	/* Set bin-related variables. */
3520 	bin_maxclass = (pagesize >> 1);
3521 	assert(opt_quantum_2pow >= TINY_MIN_2POW);
3522 	ntbins = (unsigned)(opt_quantum_2pow - TINY_MIN_2POW);
3523 	assert(ntbins <= opt_quantum_2pow);
3524 	nqbins = (unsigned)(small_max >> opt_quantum_2pow);
3525 	nsbins = (unsigned)(pagesize_2pow - opt_small_max_2pow - 1);
3526 
3527 	/* Set variables according to the value of opt_quantum_2pow. */
3528 	quantum = (1 << opt_quantum_2pow);
3529 	quantum_mask = quantum - 1;
3530 	if (ntbins > 0)
3531 		small_min = (quantum >> 1) + 1;
3532 	else
3533 		small_min = 1;
3534 	assert(small_min <= quantum);
3535 
3536 	/* Set variables according to the value of opt_chunk_2pow. */
3537 	chunksize = (1LU << opt_chunk_2pow);
3538 	chunksize_mask = chunksize - 1;
3539 	chunksize_2pow = (unsigned)opt_chunk_2pow;
3540 	chunk_npages = (unsigned)(chunksize >> pagesize_2pow);
3541 	{
3542 		unsigned header_size;
3543 
3544 		header_size = (unsigned)(sizeof(arena_chunk_t) +
3545 		    (sizeof(arena_chunk_map_t) * (chunk_npages - 1)));
3546 		arena_chunk_header_npages = (header_size >> pagesize_2pow);
3547 		if ((header_size & pagesize_mask) != 0)
3548 			arena_chunk_header_npages++;
3549 	}
3550 	arena_maxclass = chunksize - (arena_chunk_header_npages <<
3551 	    pagesize_2pow);
3552 
3553 	UTRACE(0, 0, 0);
3554 
3555 #ifdef MALLOC_STATS
3556 	memset(&stats_chunks, 0, sizeof(chunk_stats_t));
3557 #endif
3558 
3559 	/* Various sanity checks that regard configuration. */
3560 	assert(quantum >= sizeof(void *));
3561 	assert(quantum <= pagesize);
3562 	assert(chunksize >= pagesize);
3563 	assert(quantum * 4 <= chunksize);
3564 
3565 	/* Initialize chunks data. */
3566 	malloc_mutex_init(&chunks_mtx);
3567 	RB_INIT(&huge);
3568 #ifdef USE_BRK
3569 	malloc_mutex_init(&brk_mtx);
3570 	brk_base = sbrk(0);
3571 	brk_prev = brk_base;
3572 	brk_max = brk_base;
3573 #endif
3574 #ifdef MALLOC_STATS
3575 	huge_nmalloc = 0;
3576 	huge_ndalloc = 0;
3577 	huge_nralloc = 0;
3578 	huge_allocated = 0;
3579 #endif
3580 	RB_INIT(&old_chunks);
3581 
3582 	/* Initialize base allocation data structures. */
3583 #ifdef MALLOC_STATS
3584 	base_mapped = 0;
3585 #endif
3586 #ifdef USE_BRK
3587 	/*
3588 	 * Allocate a base chunk here, since it doesn't actually have to be
3589 	 * chunk-aligned.  Doing this before allocating any other chunks allows
3590 	 * the use of space that would otherwise be wasted.
3591 	 */
3592 	base_pages_alloc(0);
3593 #endif
3594 	base_chunk_nodes = NULL;
3595 	malloc_mutex_init(&base_mtx);
3596 
3597 	if (ncpus > 1) {
3598 		/*
3599 		 * For SMP systems, create four times as many arenas as there
3600 		 * are CPUs by default.
3601 		 */
3602 		opt_narenas_lshift += 2;
3603 	}
3604 
3605 #ifdef NO_TLS
3606 	/* Initialize arena key. */
3607 	(void)thr_keycreate(&arenas_map_key, NULL);
3608 #endif
3609 
3610 	/* Determine how many arenas to use. */
3611 	narenas = ncpus;
3612 	if (opt_narenas_lshift > 0) {
3613 		if ((narenas << opt_narenas_lshift) > narenas)
3614 			narenas <<= opt_narenas_lshift;
3615 		/*
3616 		 * Make sure not to exceed the limits of what base_malloc()
3617 		 * can handle.
3618 		 */
3619 		if (narenas * sizeof(arena_t *) > chunksize)
3620 			narenas = (unsigned)(chunksize / sizeof(arena_t *));
3621 	} else if (opt_narenas_lshift < 0) {
3622 		if ((narenas << opt_narenas_lshift) < narenas)
3623 			narenas <<= opt_narenas_lshift;
3624 		/* Make sure there is at least one arena. */
3625 		if (narenas == 0)
3626 			narenas = 1;
3627 	}
3628 
3629 	next_arena = 0;
3630 
3631 	/* Allocate and initialize arenas. */
3632 	arenas = (arena_t **)base_alloc(sizeof(arena_t *) * narenas);
3633 	if (arenas == NULL) {
3634 		malloc_mutex_unlock(&init_lock);
3635 		return (true);
3636 	}
3637 	/*
3638 	 * Zero the array.  In practice, this should always be pre-zeroed,
3639 	 * since it was just mmap()ed, but let's be sure.
3640 	 */
3641 	memset(arenas, 0, sizeof(arena_t *) * narenas);
3642 
3643 	/*
3644 	 * Initialize one arena here.  The rest are lazily created in
3645 	 * arena_choose_hard().
3646 	 */
3647 	arenas_extend(0);
3648 	if (arenas[0] == NULL) {
3649 		malloc_mutex_unlock(&init_lock);
3650 		return (true);
3651 	}
3652 
3653 	malloc_mutex_init(&arenas_mtx);
3654 
3655 	malloc_initialized = true;
3656 	malloc_mutex_unlock(&init_lock);
3657 	return (false);
3658 }
3659 
3660 /*
3661  * End general internal functions.
3662  */
3663 /******************************************************************************/
3664 /*
3665  * Begin malloc(3)-compatible functions.
3666  */
3667 
3668 void *
3669 malloc(size_t size)
3670 {
3671 	void *ret;
3672 
3673 	if (malloc_init()) {
3674 		ret = NULL;
3675 		goto RETURN;
3676 	}
3677 
3678 	if (size == 0) {
3679 		if (opt_sysv == false)
3680 			size = 1;
3681 		else {
3682 			ret = NULL;
3683 			goto RETURN;
3684 		}
3685 	}
3686 
3687 	ret = imalloc(size);
3688 
3689 RETURN:
3690 	if (ret == NULL) {
3691 		if (opt_xmalloc) {
3692 			_malloc_message(getprogname(),
3693 			    ": (malloc) Error in malloc(): out of memory\n", "",
3694 			    "");
3695 			abort();
3696 		}
3697 		errno = ENOMEM;
3698 	}
3699 
3700 	UTRACE(0, size, ret);
3701 	return (ret);
3702 }
3703 
3704 int
3705 posix_memalign(void **memptr, size_t alignment, size_t size)
3706 {
3707 	int ret;
3708 	void *result;
3709 
3710 	if (malloc_init())
3711 		result = NULL;
3712 	else {
3713 		/* Make sure that alignment is a large enough power of 2. */
3714 		if (((alignment - 1) & alignment) != 0
3715 		    || alignment < sizeof(void *)) {
3716 			if (opt_xmalloc) {
3717 				_malloc_message(getprogname(),
3718 				    ": (malloc) Error in posix_memalign(): "
3719 				    "invalid alignment\n", "", "");
3720 				abort();
3721 			}
3722 			result = NULL;
3723 			ret = EINVAL;
3724 			goto RETURN;
3725 		}
3726 
3727 		result = ipalloc(alignment, size);
3728 	}
3729 
3730 	if (result == NULL) {
3731 		if (opt_xmalloc) {
3732 			_malloc_message(getprogname(),
3733 			": (malloc) Error in posix_memalign(): out of memory\n",
3734 			"", "");
3735 			abort();
3736 		}
3737 		ret = ENOMEM;
3738 		goto RETURN;
3739 	}
3740 
3741 	*memptr = result;
3742 	ret = 0;
3743 
3744 RETURN:
3745 	UTRACE(0, size, result);
3746 	return (ret);
3747 }
3748 
3749 void *
3750 calloc(size_t num, size_t size)
3751 {
3752 	void *ret;
3753 	size_t num_size;
3754 
3755 	if (malloc_init()) {
3756 		num_size = 0;
3757 		ret = NULL;
3758 		goto RETURN;
3759 	}
3760 
3761 	num_size = num * size;
3762 	if (num_size == 0) {
3763 		if ((opt_sysv == false) && ((num == 0) || (size == 0)))
3764 			num_size = 1;
3765 		else {
3766 			ret = NULL;
3767 			goto RETURN;
3768 		}
3769 	/*
3770 	 * Try to avoid division here.  We know that it isn't possible to
3771 	 * overflow during multiplication if neither operand uses any of the
3772 	 * most significant half of the bits in a size_t.
3773 	 */
3774 	} else if ((unsigned long long)((num | size) &
3775 	   ((unsigned long long)SIZE_T_MAX << (sizeof(size_t) << 2))) &&
3776 	   (num_size / size != num)) {
3777 		/* size_t overflow. */
3778 		ret = NULL;
3779 		goto RETURN;
3780 	}
3781 
3782 	ret = icalloc(num_size);
3783 
3784 RETURN:
3785 	if (ret == NULL) {
3786 		if (opt_xmalloc) {
3787 			_malloc_message(getprogname(),
3788 			    ": (malloc) Error in calloc(): out of memory\n", "",
3789 			    "");
3790 			abort();
3791 		}
3792 		errno = ENOMEM;
3793 	}
3794 
3795 	UTRACE(0, num_size, ret);
3796 	return (ret);
3797 }
3798 
3799 void *
3800 realloc(void *ptr, size_t size)
3801 {
3802 	void *ret;
3803 
3804 	if (size == 0) {
3805 		if (opt_sysv == false)
3806 			size = 1;
3807 		else {
3808 			if (ptr != NULL)
3809 				idalloc(ptr);
3810 			ret = NULL;
3811 			goto RETURN;
3812 		}
3813 	}
3814 
3815 	if (ptr != NULL) {
3816 		assert(malloc_initialized);
3817 
3818 		ret = iralloc(ptr, size);
3819 
3820 		if (ret == NULL) {
3821 			if (opt_xmalloc) {
3822 				_malloc_message(getprogname(),
3823 				    ": (malloc) Error in realloc(): out of "
3824 				    "memory\n", "", "");
3825 				abort();
3826 			}
3827 			errno = ENOMEM;
3828 		}
3829 	} else {
3830 		if (malloc_init())
3831 			ret = NULL;
3832 		else
3833 			ret = imalloc(size);
3834 
3835 		if (ret == NULL) {
3836 			if (opt_xmalloc) {
3837 				_malloc_message(getprogname(),
3838 				    ": (malloc) Error in realloc(): out of "
3839 				    "memory\n", "", "");
3840 				abort();
3841 			}
3842 			errno = ENOMEM;
3843 		}
3844 	}
3845 
3846 RETURN:
3847 	UTRACE(ptr, size, ret);
3848 	return (ret);
3849 }
3850 
3851 void
3852 free(void *ptr)
3853 {
3854 
3855 	UTRACE(ptr, 0, 0);
3856 	if (ptr != NULL) {
3857 		assert(malloc_initialized);
3858 
3859 		idalloc(ptr);
3860 	}
3861 }
3862 
3863 /*
3864  * End malloc(3)-compatible functions.
3865  */
3866 /******************************************************************************/
3867 /*
3868  * Begin non-standard functions.
3869  */
3870 #ifndef __NetBSD__
3871 size_t
3872 malloc_usable_size(const void *ptr)
3873 {
3874 
3875 	assert(ptr != NULL);
3876 
3877 	return (isalloc(ptr));
3878 }
3879 #endif
3880 
3881 /*
3882  * End non-standard functions.
3883  */
3884 /******************************************************************************/
3885 /*
3886  * Begin library-private functions, used by threading libraries for protection
3887  * of malloc during fork().  These functions are only called if the program is
3888  * running in threaded mode, so there is no need to check whether the program
3889  * is threaded here.
3890  */
3891 
3892 void
3893 _malloc_prefork(void)
3894 {
3895 	unsigned i;
3896 
3897 	/* Acquire all mutexes in a safe order. */
3898 
3899 	malloc_mutex_lock(&arenas_mtx);
3900 	for (i = 0; i < narenas; i++) {
3901 		if (arenas[i] != NULL)
3902 			malloc_mutex_lock(&arenas[i]->mtx);
3903 	}
3904 	malloc_mutex_unlock(&arenas_mtx);
3905 
3906 	malloc_mutex_lock(&base_mtx);
3907 
3908 	malloc_mutex_lock(&chunks_mtx);
3909 }
3910 
3911 void
3912 _malloc_postfork(void)
3913 {
3914 	unsigned i;
3915 
3916 	/* Release all mutexes, now that fork() has completed. */
3917 
3918 	malloc_mutex_unlock(&chunks_mtx);
3919 
3920 	malloc_mutex_unlock(&base_mtx);
3921 
3922 	malloc_mutex_lock(&arenas_mtx);
3923 	for (i = 0; i < narenas; i++) {
3924 		if (arenas[i] != NULL)
3925 			malloc_mutex_unlock(&arenas[i]->mtx);
3926 	}
3927 	malloc_mutex_unlock(&arenas_mtx);
3928 }
3929 
3930 /*
3931  * End library-private functions.
3932  */
3933 /******************************************************************************/
3934