malloc.c source code [glibc_src_2.25/malloc/malloc.c]

1	/ Malloc implementation for multiple threads without lock contention.*
2	Copyright (C) 1996-2017 Free Software Foundation, Inc.
3	This file is part of the GNU C Library.
4	Contributed by Wolfram Gloger <wg@malloc.de>
5	and Doug Lea <dl@cs.oswego.edu>, 2001.
6
7	The GNU C Library is free software; you can redistribute it and/or
8	modify it under the terms of the GNU Lesser General Public License as
9	published by the Free Software Foundation; either version 2.1 of the
10	License, or (at your option) any later version.
11
12	The GNU C Library is distributed in the hope that it will be useful,
13	but WITHOUT ANY WARRANTY; without even the implied warranty of
14	MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
15	Lesser General Public License for more details.
16
17	You should have received a copy of the GNU Lesser General Public
18	License along with the GNU C Library; see the file COPYING.LIB. If
19	not, see <http://www.gnu.org/licenses/>. /*
20
21	/*
22	This is a version (aka ptmalloc2) of malloc/free/realloc written by
23	Doug Lea and adapted to multiple threads/arenas by Wolfram Gloger.
24
25	There have been substantial changes made after the integration into
26	glibc in all parts of the code. Do not look for much commonality
27	with the ptmalloc2 version.
28
29	* Version ptmalloc2-20011215
30	based on:
31	VERSION 2.7.0 Sun Mar 11 14:14:06 2001 Doug Lea (dl at gee)
32
33	* Quickstart
34
35	In order to compile this implementation, a Makefile is provided with
36	the ptmalloc2 distribution, which has pre-defined targets for some
37	popular systems (e.g. "make posix" for Posix threads). All that is
38	typically required with regard to compiler flags is the selection of
39	the thread package via defining one out of USE_PTHREADS, USE_THR or
40	USE_SPROC. Check the thread-m.h file for what effects this has.
41	Many/most systems will additionally require USE_TSD_DATA_HACK to be
42	defined, so this is the default for "make posix".
43
44	* Why use this malloc?
45
46	This is not the fastest, most space-conserving, most portable, or
47	most tunable malloc ever written. However it is among the fastest
48	while also being among the most space-conserving, portable and tunable.
49	Consistent balance across these factors results in a good general-purpose
50	allocator for malloc-intensive programs.
51
52	The main properties of the algorithms are:
53	* For large (>= 512 bytes) requests, it is a pure best-fit allocator,
54	with ties normally decided via FIFO (i.e. least recently used).
55	* For small (<= 64 bytes by default) requests, it is a caching
56	allocator, that maintains pools of quickly recycled chunks.
57	* In between, and for combinations of large and small requests, it does
58	the best it can trying to meet both goals at once.
59	* For very large requests (>= 128KB by default), it relies on system
60	memory mapping facilities, if supported.
61
62	For a longer but slightly out of date high-level description, see
63	http://gee.cs.oswego.edu/dl/html/malloc.html
64
65	You may already by default be using a C library containing a malloc
66	that is based on some version of this malloc (for example in
67	linux). You might still want to use the one in this file in order to
68	customize settings or to avoid overheads associated with library
69	versions.
70
71	* Contents, described in more detail in "description of public routines" below.
72
73	Standard (ANSI/SVID/...) functions:
74	malloc(size_t n);
75	calloc(size_t n_elements, size_t element_size);
76	free(void p);*
77	realloc(void p, size_t n);*
78	memalign(size_t alignment, size_t n);
79	valloc(size_t n);
80	mallinfo()
81	mallopt(int parameter_number, int parameter_value)
82
83	Additional functions:
84	independent_calloc(size_t n_elements, size_t size, void chunks[]);*
85	independent_comalloc(size_t n_elements, size_t sizes[], void chunks[]);*
86	pvalloc(size_t n);
87	cfree(void p);*
88	malloc_trim(size_t pad);
89	malloc_usable_size(void p);*
90	malloc_stats();
91
92	* Vital statistics:
93
94	Supported pointer representation: 4 or 8 bytes
95	Supported size_t representation: 4 or 8 bytes
96	Note that size_t is allowed to be 4 bytes even if pointers are 8.
97	You can adjust this by defining INTERNAL_SIZE_T
98
99	Alignment: 2 sizeof(size_t) (default)*
100	(i.e., 8 byte alignment with 4byte size_t). This suffices for
101	nearly all current machines and C compilers. However, you can
102	define MALLOC_ALIGNMENT to be wider than this if necessary.
103
104	Minimum overhead per allocated chunk: 4 or 8 bytes
105	Each malloced chunk has a hidden word of overhead holding size
106	and status information.
107
108	Minimum allocated size: 4-byte ptrs: 16 bytes (including 4 overhead)
109	8-byte ptrs: 24/32 bytes (including, 4/8 overhead)
110
111	When a chunk is freed, 12 (for 4byte ptrs) or 20 (for 8 byte
112	ptrs but 4 byte size) or 24 (for 8/8) additional bytes are
113	needed; 4 (8) for a trailing size field and 8 (16) bytes for
114	free list pointers. Thus, the minimum allocatable size is
115	16/24/32 bytes.
116
117	Even a request for zero bytes (i.e., malloc(0)) returns a
118	pointer to something of the minimum allocatable size.
119
120	The maximum overhead wastage (i.e., number of extra bytes
121	allocated than were requested in malloc) is less than or equal
122	to the minimum size, except for requests >= mmap_threshold that
123	are serviced via mmap(), where the worst case wastage is 2 *
124	sizeof(size_t) bytes plus the remainder from a system page (the
125	minimal mmap unit); typically 4096 or 8192 bytes.
126
127	Maximum allocated size: 4-byte size_t: 2^32 minus about two pages
128	8-byte size_t: 2^64 minus about two pages
129
130	It is assumed that (possibly signed) size_t values suffice to
131	represent chunk sizes. `Possibly signed' is due to the fact
132	that `size_t' may be defined on a system as either a signed or
133	an unsigned type. The ISO C standard says that it must be
134	unsigned, but a few systems are known not to adhere to this.
135	Additionally, even when size_t is unsigned, sbrk (which is by
136	default used to obtain memory from system) accepts signed
137	arguments, and may not be able to handle size_t-wide arguments
138	with negative sign bit. Generally, values that would
139	appear as negative after accounting for overhead and alignment
140	are supported only via mmap(), which does not have this
141	limitation.
142
143	Requests for sizes outside the allowed range will perform an optional
144	failure action and then return null. (Requests may also
145	also fail because a system is out of memory.)
146
147	Thread-safety: thread-safe
148
149	Compliance: I believe it is compliant with the 1997 Single Unix Specification
150	Also SVID/XPG, ANSI C, and probably others as well.
151
152	* Synopsis of compile-time options:
153
154	People have reported using previous versions of this malloc on all
155	versions of Unix, sometimes by tweaking some of the defines
156	below. It has been tested most extensively on Solaris and Linux.
157	People also report using it in stand-alone embedded systems.
158
159	The implementation is in straight, hand-tuned ANSI C. It is not
160	at all modular. (Sorry!) It uses a lot of macros. To be at all
161	usable, this code should be compiled using an optimizing compiler
162	(for example gcc -O3) that can simplify expressions and control
163	paths. (FAQ: some macros import variables as arguments rather than
164	declare locals because people reported that some debuggers
165	otherwise get confused.)
166
167	OPTION DEFAULT VALUE
168
169	Compilation Environment options:
170
171	HAVE_MREMAP 0
172
173	Changing default word sizes:
174
175	INTERNAL_SIZE_T size_t
176
177	Configuration and functionality options:
178
179	USE_PUBLIC_MALLOC_WRAPPERS NOT defined
180	USE_MALLOC_LOCK NOT defined
181	MALLOC_DEBUG NOT defined
182	REALLOC_ZERO_BYTES_FREES 1
183	TRIM_FASTBINS 0
184
185	Options for customizing MORECORE:
186
187	MORECORE sbrk
188	MORECORE_FAILURE -1
189	MORECORE_CONTIGUOUS 1
190	MORECORE_CANNOT_TRIM NOT defined
191	MORECORE_CLEARS 1
192	MMAP_AS_MORECORE_SIZE (1024 1024)*
193
194	Tuning options that are also dynamically changeable via mallopt:
195
196	DEFAULT_MXFAST 64 (for 32bit), 128 (for 64bit)
197	DEFAULT_TRIM_THRESHOLD 128 1024*
198	DEFAULT_TOP_PAD 0
199	DEFAULT_MMAP_THRESHOLD 128 1024*
200	DEFAULT_MMAP_MAX 65536
201
202	There are several other #defined constants and macros that you
203	probably don't want to touch unless you are extending or adapting malloc. /*
204
205	/*
206	void is the pointer type that malloc should say it returns*
207	*/
208
209	#ifndef void
210	#define void void
211	#endif /void/
212
213	#include <stddef.h> /* for size_t */
214	#include <stdlib.h> /* for getenv(), abort() */
215	#include <unistd.h> /* for __libc_enable_secure */
216
217	#include <atomic.h>
218	#include <_itoa.h>
219	#include <bits/wordsize.h>
220	#include <sys/sysinfo.h>
221
222	#include <ldsodefs.h>
223
224	#include <unistd.h>
225	#include <stdio.h> /* needed for malloc_stats */
226	#include <errno.h>
227
228	#include <shlib-compat.h>
229
230	/ For uintptr_t. /
231	#include <stdint.h>
232
233	/ For va_arg, va_start, va_end. /
234	#include <stdarg.h>
235
236	/ For MIN, MAX, powerof2. /
237	#include <sys/param.h>
238
239	/ For ALIGN_UP et. al. /
240	#include <libc-internal.h>
241
242	#include <malloc/malloc-internal.h>
243
244	/*
245	Debugging:
246
247	Because freed chunks may be overwritten with bookkeeping fields, this
248	malloc will often die when freed memory is overwritten by user
249	programs. This can be very effective (albeit in an annoying way)
250	in helping track down dangling pointers.
251
252	If you compile with -DMALLOC_DEBUG, a number of assertion checks are
253	enabled that will catch more memory errors. You probably won't be
254	able to make much sense of the actual assertion errors, but they
255	should help you locate incorrectly overwritten memory. The checking
256	is fairly extensive, and will slow down execution
257	noticeably. Calling malloc_stats or mallinfo with MALLOC_DEBUG set
258	will attempt to check every non-mmapped allocated and free chunk in
259	the course of computing the summmaries. (By nature, mmapped regions
260	cannot be checked very much automatically.)
261
262	Setting MALLOC_DEBUG may also be helpful if you are trying to modify
263	this code. The assertions in the check routines spell out in more
264	detail the assumptions and invariants underlying the algorithms.
265
266	Setting MALLOC_DEBUG does NOT provide an automated mechanism for
267	checking that all accesses to malloced memory stay within their
268	bounds. However, there are several add-ons and adaptations of this
269	or other mallocs available that do this.
270	*/
271
272	#ifndef MALLOC_DEBUG
273	#define MALLOC_DEBUG 0
274	#endif
275
276	#ifdef NDEBUG
277	# define assert(expr) ((void) 0)
278	#else
279	# define assert(expr) \
280	((expr) \
281	? ((void) 0) \
282	: __malloc_assert (#expr, __FILE__, __LINE__, __func__))
283
284	extern const char *__progname;
285
286	static void
287	__malloc_assert (const char assertion, const* char file, unsigned* int line,
288	const char *function)
289	{
290	(void) __fxprintf (NULL, "%s%s%s:%u: %s%sAssertion `%s' failed.\n",
291	__progname, __progname[`0`] ? ": " : "",
292	file, line,
293	function ? function : "", function ? ": " : "",
294	assertion);
295	fflush (stderr);
296	abort ();
297	}
298	#endif
299
300
301	/*
302	REALLOC_ZERO_BYTES_FREES should be set if a call to
303	realloc with zero bytes should be the same as a call to free.
304	This is required by the C standard. Otherwise, since this malloc
305	returns a unique pointer for malloc(0), so does realloc(p, 0).
306	*/
307
308	#ifndef REALLOC_ZERO_BYTES_FREES
309	#define REALLOC_ZERO_BYTES_FREES 1
310	#endif
311
312	/*
313	TRIM_FASTBINS controls whether free() of a very small chunk can
314	immediately lead to trimming. Setting to true (1) can reduce memory
315	footprint, but will almost always slow down programs that use a lot
316	of small chunks.
317
318	Define this only if you are willing to give up some speed to more
319	aggressively reduce system-level memory footprint when releasing
320	memory in programs that use many small chunks. You can get
321	essentially the same effect by setting MXFAST to 0, but this can
322	lead to even greater slowdowns in programs using many small chunks.
323	TRIM_FASTBINS is an in-between compile-time option, that disables
324	only those chunks bordering topmost memory from being placed in
325	fastbins.
326	*/
327
328	#ifndef TRIM_FASTBINS
329	#define TRIM_FASTBINS 0
330	#endif
331
332
333	/ Definition for getting more memory from the OS. /
334	#define MORECORE (*__morecore)
335	#define MORECORE_FAILURE 0
336	void * __default_morecore (ptrdiff_t);
337	void (__morecore)(ptrdiff_t) = __default_morecore;
338
339
340	#include <string.h>
341
342	/*
343	MORECORE-related declarations. By default, rely on sbrk
344	*/
345
346
347	/*
348	MORECORE is the name of the routine to call to obtain more memory
349	from the system. See below for general guidance on writing
350	alternative MORECORE functions, as well as a version for WIN32 and a
351	sample version for pre-OSX macos.
352	*/
353
354	#ifndef MORECORE
355	#define MORECORE sbrk
356	#endif
357
358	/*
359	MORECORE_FAILURE is the value returned upon failure of MORECORE
360	as well as mmap. Since it cannot be an otherwise valid memory address,
361	and must reflect values of standard sys calls, you probably ought not
362	try to redefine it.
363	*/
364
365	#ifndef MORECORE_FAILURE
366	#define MORECORE_FAILURE (-1)
367	#endif
368
369	/*
370	If MORECORE_CONTIGUOUS is true, take advantage of fact that
371	consecutive calls to MORECORE with positive arguments always return
372	contiguous increasing addresses. This is true of unix sbrk. Even
373	if not defined, when regions happen to be contiguous, malloc will
374	permit allocations spanning regions obtained from different
375	calls. But defining this when applicable enables some stronger
376	consistency checks and space efficiencies.
377	*/
378
379	#ifndef MORECORE_CONTIGUOUS
380	#define MORECORE_CONTIGUOUS 1
381	#endif
382
383	/*
384	Define MORECORE_CANNOT_TRIM if your version of MORECORE
385	cannot release space back to the system when given negative
386	arguments. This is generally necessary only if you are using
387	a hand-crafted MORECORE function that cannot handle negative arguments.
388	*/
389
390	/ #define MORECORE_CANNOT_TRIM /
391
392	/ MORECORE_CLEARS (default 1)*
393	The degree to which the routine mapped to MORECORE zeroes out
394	memory: never (0), only for newly allocated space (1) or always
395	(2). The distinction between (1) and (2) is necessary because on
396	some systems, if the application first decrements and then
397	increments the break value, the contents of the reallocated space
398	are unspecified.
399	*/
400
401	#ifndef MORECORE_CLEARS
402	# define MORECORE_CLEARS 1
403	#endif
404
405
406	/*
407	MMAP_AS_MORECORE_SIZE is the minimum mmap size argument to use if
408	sbrk fails, and mmap is used as a backup. The value must be a
409	multiple of page size. This backup strategy generally applies only
410	when systems have "holes" in address space, so sbrk cannot perform
411	contiguous expansion, but there is still space available on system.
412	On systems for which this is known to be useful (i.e. most linux
413	kernels), this occurs only when programs allocate huge amounts of
414	memory. Between this, and the fact that mmap regions tend to be
415	limited, the size should be large, to avoid too many mmap calls and
416	thus avoid running out of kernel resources. /*
417
418	#ifndef MMAP_AS_MORECORE_SIZE
419	#define MMAP_AS_MORECORE_SIZE (1024 * 1024)
420	#endif
421
422	/*
423	Define HAVE_MREMAP to make realloc() use mremap() to re-allocate
424	large blocks.
425	*/
426
427	#ifndef HAVE_MREMAP
428	#define HAVE_MREMAP 0
429	#endif
430
431	/ We may need to support __malloc_initialize_hook for backwards*
432	compatibility. /*
433
434	#if SHLIB_COMPAT (libc, GLIBC_2_0, GLIBC_2_24)
435	# define HAVE_MALLOC_INIT_HOOK 1
436	#else
437	# define HAVE_MALLOC_INIT_HOOK 0
438	#endif
439
440
441	/*
442	This version of malloc supports the standard SVID/XPG mallinfo
443	routine that returns a struct containing usage properties and
444	statistics. It should work on any SVID/XPG compliant system that has
445	a /usr/include/malloc.h defining struct mallinfo. (If you'd like to
446	install such a thing yourself, cut out the preliminary declarations
447	as described above and below and save them in a malloc.h file. But
448	there's no compelling reason to bother to do this.)
449
450	The main declaration needed is the mallinfo struct that is returned
451	(by-copy) by mallinfo(). The SVID/XPG malloinfo struct contains a
452	bunch of fields that are not even meaningful in this version of
453	malloc. These fields are are instead filled by mallinfo() with
454	other numbers that might be of interest.
455	*/
456
457
458	/ ---------- description of public routines ------------ /
459
460	/*
461	malloc(size_t n)
462	Returns a pointer to a newly allocated chunk of at least n bytes, or null
463	if no space is available. Additionally, on failure, errno is
464	set to ENOMEM on ANSI C systems.
465
466	If n is zero, malloc returns a minumum-sized chunk. (The minimum
467	size is 16 bytes on most 32bit systems, and 24 or 32 bytes on 64bit
468	systems.) On most systems, size_t is an unsigned type, so calls
469	with negative arguments are interpreted as requests for huge amounts
470	of space, which will often fail. The maximum supported value of n
471	differs across systems, but is in all cases less than the maximum
472	representable value of a size_t.
473	*/
474	void* __libc_malloc(size_t);
475	libc_hidden_proto (__libc_malloc)
476
477	/*
478	free(void p)*
479	Releases the chunk of memory pointed to by p, that had been previously
480	allocated using malloc or a related routine such as realloc.
481	It has no effect if p is null. It can have arbitrary (i.e., bad!)
482	effects if p has already been freed.
483
484	Unless disabled (using mallopt), freeing very large spaces will
485	when possible, automatically trigger operations that give
486	back unused memory to the system, thus reducing program footprint.
487	*/
488	void __libc_free(void*);
489	libc_hidden_proto (__libc_free)
490
491	/*
492	calloc(size_t n_elements, size_t element_size);
493	Returns a pointer to n_elements element_size bytes, with all locations*
494	set to zero.
495	*/
496	void* __libc_calloc(size_t, size_t);
497
498	/*
499	realloc(void p, size_t n)*
500	Returns a pointer to a chunk of size n that contains the same data
501	as does chunk p up to the minimum of (n, p's size) bytes, or null
502	if no space is available.
503
504	The returned pointer may or may not be the same as p. The algorithm
505	prefers extending p when possible, otherwise it employs the
506	equivalent of a malloc-copy-free sequence.
507
508	If p is null, realloc is equivalent to malloc.
509
510	If space is not available, realloc returns null, errno is set (if on
511	ANSI) and p is NOT freed.
512
513	if n is for fewer bytes than already held by p, the newly unused
514	space is lopped off and freed if possible. Unless the #define
515	REALLOC_ZERO_BYTES_FREES is set, realloc with a size argument of
516	zero (re)allocates a minimum-sized chunk.
517
518	Large chunks that were internally obtained via mmap will always
519	be reallocated using malloc-copy-free sequences unless
520	the system supports MREMAP (currently only linux).
521
522	The old unix realloc convention of allowing the last-free'd chunk
523	to be used as an argument to realloc is not supported.
524	*/
525	void* __libc_realloc(void*, size_t);
526	libc_hidden_proto (__libc_realloc)
527
528	/*
529	memalign(size_t alignment, size_t n);
530	Returns a pointer to a newly allocated chunk of n bytes, aligned
531	in accord with the alignment argument.
532
533	The alignment argument should be a power of two. If the argument is
534	not a power of two, the nearest greater power is used.
535	8-byte alignment is guaranteed by normal malloc calls, so don't
536	bother calling memalign with an argument of 8 or less.
537
538	Overreliance on memalign is a sure way to fragment space.
539	*/
540	void* __libc_memalign(size_t, size_t);
541	libc_hidden_proto (__libc_memalign)
542
543	/*
544	valloc(size_t n);
545	Equivalent to memalign(pagesize, n), where pagesize is the page
546	size of the system. If the pagesize is unknown, 4096 is used.
547	*/
548	void* __libc_valloc(size_t);
549
550
551
552	/*
553	mallopt(int parameter_number, int parameter_value)
554	Sets tunable parameters The format is to provide a
555	(parameter-number, parameter-value) pair. mallopt then sets the
556	corresponding parameter to the argument value if it can (i.e., so
557	long as the value is meaningful), and returns 1 if successful else
558	0. SVID/XPG/ANSI defines four standard param numbers for mallopt,
559	normally defined in malloc.h. Only one of these (M_MXFAST) is used
560	in this malloc. The others (M_NLBLKS, M_GRAIN, M_KEEP) don't apply,
561	so setting them has no effect. But this malloc also supports four
562	other options in mallopt. See below for details. Briefly, supported
563	parameters are as follows (listed defaults are for "typical"
564	configurations).
565
566	Symbol param # default allowed param values
567	M_MXFAST 1 64 0-80 (0 disables fastbins)
568	M_TRIM_THRESHOLD -1 1281024 any (-1U disables trimming)*
569	M_TOP_PAD -2 0 any
570	M_MMAP_THRESHOLD -3 1281024 any (or 0 if no MMAP support)*
571	M_MMAP_MAX -4 65536 any (0 disables use of mmap)
572	*/
573	int __libc_mallopt(int, int);
574	libc_hidden_proto (__libc_mallopt)
575
576
577	/*
578	mallinfo()
579	Returns (by copy) a struct containing various summary statistics:
580
581	arena: current total non-mmapped bytes allocated from system
582	ordblks: the number of free chunks
583	smblks: the number of fastbin blocks (i.e., small chunks that
584	have been freed but not use resused or consolidated)
585	hblks: current number of mmapped regions
586	hblkhd: total bytes held in mmapped regions
587	usmblks: always 0
588	fsmblks: total bytes held in fastbin blocks
589	uordblks: current total allocated space (normal or mmapped)
590	fordblks: total free space
591	keepcost: the maximum number of bytes that could ideally be released
592	back to system via malloc_trim. ("ideally" means that
593	it ignores page restrictions etc.)
594
595	Because these fields are ints, but internal bookkeeping may
596	be kept as longs, the reported values may wrap around zero and
597	thus be inaccurate.
598	*/
599	struct mallinfo __libc_mallinfo(void);
600
601
602	/*
603	pvalloc(size_t n);
604	Equivalent to valloc(minimum-page-that-holds(n)), that is,
605	round up n to nearest pagesize.
606	*/
607	void* __libc_pvalloc(size_t);
608
609	/*
610	malloc_trim(size_t pad);
611
612	If possible, gives memory back to the system (via negative
613	arguments to sbrk) if there is unused memory at the `high' end of
614	the malloc pool. You can call this after freeing large blocks of
615	memory to potentially reduce the system-level memory requirements
616	of a program. However, it cannot guarantee to reduce memory. Under
617	some allocation patterns, some large free blocks of memory will be
618	locked between two used chunks, so they cannot be given back to
619	the system.
620
621	The `pad' argument to malloc_trim represents the amount of free
622	trailing space to leave untrimmed. If this argument is zero,
623	only the minimum amount of memory to maintain internal data
624	structures will be left (one page or less). Non-zero arguments
625	can be supplied to maintain enough trailing space to service
626	future expected allocations without having to re-obtain memory
627	from the system.
628
629	Malloc_trim returns 1 if it actually released any memory, else 0.
630	On systems that do not support "negative sbrks", it will always
631	return 0.
632	*/
633	int __malloc_trim(size_t);
634
635	/*
636	malloc_usable_size(void p);*
637
638	Returns the number of bytes you can actually use in
639	an allocated chunk, which may be more than you requested (although
640	often not) due to alignment and minimum size constraints.
641	You can use this many bytes without worrying about
642	overwriting other allocated objects. This is not a particularly great
643	programming practice. malloc_usable_size can be more useful in
644	debugging and assertions, for example:
645
646	p = malloc(n);
647	assert(malloc_usable_size(p) >= 256);
648
649	*/
650	size_t __malloc_usable_size(void*);
651
652	/*
653	malloc_stats();
654	Prints on stderr the amount of space obtained from the system (both
655	via sbrk and mmap), the maximum amount (which may be more than
656	current if malloc_trim and/or munmap got called), and the current
657	number of bytes allocated via malloc (or realloc, etc) but not yet
658	freed. Note that this is the number of bytes allocated, not the
659	number requested. It will be larger than the number requested
660	because of alignment and bookkeeping overhead. Because it includes
661	alignment wastage as being in use, this figure may be greater than
662	zero even when no user-level chunks are allocated.
663
664	The reported current and maximum system memory can be inaccurate if
665	a program makes other calls to system memory allocation functions
666	(normally sbrk) outside of malloc.
667
668	malloc_stats prints only the most commonly interesting statistics.
669	More information can be obtained by calling mallinfo.
670
671	*/
672	void __malloc_stats(void);
673
674	/*
675	malloc_get_state(void);
676
677	Returns the state of all malloc variables in an opaque data
678	structure.
679	*/
680	void* __malloc_get_state(void);
681
682	/*
683	malloc_set_state(void state);*
684
685	Restore the state of all malloc variables from data obtained with
686	malloc_get_state().
687	*/
688	int __malloc_set_state(void*);
689
690	/*
691	posix_memalign(void memptr, size_t alignment, size_t size);
692
693	POSIX wrapper like memalign(), checking for validity of size.
694	*/
695	int __posix_memalign(void **, size_t, size_t);
696
697	/ mallopt tuning options /
698
699	/*
700	M_MXFAST is the maximum request size used for "fastbins", special bins
701	that hold returned chunks without consolidating their spaces. This
702	enables future requests for chunks of the same size to be handled
703	very quickly, but can increase fragmentation, and thus increase the
704	overall memory footprint of a program.
705
706	This malloc manages fastbins very conservatively yet still
707	efficiently, so fragmentation is rarely a problem for values less
708	than or equal to the default. The maximum supported value of MXFAST
709	is 80. You wouldn't want it any higher than this anyway. Fastbins
710	are designed especially for use with many small structs, objects or
711	strings -- the default handles structs/objects/arrays with sizes up
712	to 8 4byte fields, or small strings representing words, tokens,
713	etc. Using fastbins for larger objects normally worsens
714	fragmentation without improving speed.
715
716	M_MXFAST is set in REQUEST size units. It is internally used in
717	chunksize units, which adds padding and alignment. You can reduce
718	M_MXFAST to 0 to disable all use of fastbins. This causes the malloc
719	algorithm to be a closer approximation of fifo-best-fit in all cases,
720	not just for larger requests, but will generally cause it to be
721	slower.
722	*/
723
724
725	/ M_MXFAST is a standard SVID/XPG tuning option, usually listed in malloc.h /
726	#ifndef M_MXFAST
727	#define M_MXFAST 1
728	#endif
729
730	#ifndef DEFAULT_MXFAST
731	#define DEFAULT_MXFAST (64 * SIZE_SZ / 4)
732	#endif
733
734
735	/*
736	M_TRIM_THRESHOLD is the maximum amount of unused top-most memory
737	to keep before releasing via malloc_trim in free().
738
739	Automatic trimming is mainly useful in long-lived programs.
740	Because trimming via sbrk can be slow on some systems, and can
741	sometimes be wasteful (in cases where programs immediately
742	afterward allocate more large chunks) the value should be high
743	enough so that your overall system performance would improve by
744	releasing this much memory.
745
746	The trim threshold and the mmap control parameters (see below)
747	can be traded off with one another. Trimming and mmapping are
748	two different ways of releasing unused memory back to the
749	system. Between these two, it is often possible to keep
750	system-level demands of a long-lived program down to a bare
751	minimum. For example, in one test suite of sessions measuring
752	the XF86 X server on Linux, using a trim threshold of 128K and a
753	mmap threshold of 192K led to near-minimal long term resource
754	consumption.
755
756	If you are using this malloc in a long-lived program, it should
757	pay to experiment with these values. As a rough guide, you
758	might set to a value close to the average size of a process
759	(program) running on your system. Releasing this much memory
760	would allow such a process to run in memory. Generally, it's
761	worth it to tune for trimming rather tham memory mapping when a
762	program undergoes phases where several large chunks are
763	allocated and released in ways that can reuse each other's
764	storage, perhaps mixed with phases where there are no such
765	chunks at all. And in well-behaved long-lived programs,
766	controlling release of large blocks via trimming versus mapping
767	is usually faster.
768
769	However, in most programs, these parameters serve mainly as
770	protection against the system-level effects of carrying around
771	massive amounts of unneeded memory. Since frequent calls to
772	sbrk, mmap, and munmap otherwise degrade performance, the default
773	parameters are set to relatively high values that serve only as
774	safeguards.
775
776	The trim value It must be greater than page size to have any useful
777	effect. To disable trimming completely, you can set to
778	(unsigned long)(-1)
779
780	Trim settings interact with fastbin (MXFAST) settings: Unless
781	TRIM_FASTBINS is defined, automatic trimming never takes place upon
782	freeing a chunk with size less than or equal to MXFAST. Trimming is
783	instead delayed until subsequent freeing of larger chunks. However,
784	you can still force an attempted trim by calling malloc_trim.
785
786	Also, trimming is not generally possible in cases where
787	the main arena is obtained via mmap.
788
789	Note that the trick some people use of mallocing a huge space and
790	then freeing it at program startup, in an attempt to reserve system
791	memory, doesn't have the intended effect under automatic trimming,
792	since that memory will immediately be returned to the system.
793	*/
794
795	#define M_TRIM_THRESHOLD -1
796
797	#ifndef DEFAULT_TRIM_THRESHOLD
798	#define DEFAULT_TRIM_THRESHOLD (128 * 1024)
799	#endif
800
801	/*
802	M_TOP_PAD is the amount of extra `padding' space to allocate or
803	retain whenever sbrk is called. It is used in two ways internally:
804
805	* When sbrk is called to extend the top of the arena to satisfy
806	a new malloc request, this much padding is added to the sbrk
807	request.
808
809	* When malloc_trim is called automatically from free(),
810	it is used as the `pad' argument.
811
812	In both cases, the actual amount of padding is rounded
813	so that the end of the arena is always a system page boundary.
814
815	The main reason for using padding is to avoid calling sbrk so
816	often. Having even a small pad greatly reduces the likelihood
817	that nearly every malloc request during program start-up (or
818	after trimming) will invoke sbrk, which needlessly wastes
819	time.
820
821	Automatic rounding-up to page-size units is normally sufficient
822	to avoid measurable overhead, so the default is 0. However, in
823	systems where sbrk is relatively slow, it can pay to increase
824	this value, at the expense of carrying around more memory than
825	the program needs.
826	*/
827
828	#define M_TOP_PAD -2
829
830	#ifndef DEFAULT_TOP_PAD
831	#define DEFAULT_TOP_PAD (0)
832	#endif
833
834	/*
835	MMAP_THRESHOLD_MAX and _MIN are the bounds on the dynamically
836	adjusted MMAP_THRESHOLD.
837	*/
838
839	#ifndef DEFAULT_MMAP_THRESHOLD_MIN
840	#define DEFAULT_MMAP_THRESHOLD_MIN (128 * 1024)
841	#endif
842
843	#ifndef DEFAULT_MMAP_THRESHOLD_MAX
844	/ For 32-bit platforms we cannot increase the maximum mmap*
845	threshold much because it is also the minimum value for the
846	maximum heap size and its alignment. Going above 512k (i.e., 1M
847	for new heaps) wastes too much address space. /*
848	# if __WORDSIZE == 32
849	# define DEFAULT_MMAP_THRESHOLD_MAX (512 * 1024)
850	# else
851	# define DEFAULT_MMAP_THRESHOLD_MAX (4 * 1024 * 1024 * sizeof(long))
852	# endif
853	#endif
854
855	/*
856	M_MMAP_THRESHOLD is the request size threshold for using mmap()
857	to service a request. Requests of at least this size that cannot
858	be allocated using already-existing space will be serviced via mmap.
859	(If enough normal freed space already exists it is used instead.)
860
861	Using mmap segregates relatively large chunks of memory so that
862	they can be individually obtained and released from the host
863	system. A request serviced through mmap is never reused by any
864	other request (at least not directly; the system may just so
865	happen to remap successive requests to the same locations).
866
867	Segregating space in this way has the benefits that:
868
869	1. Mmapped space can ALWAYS be individually released back
870	to the system, which helps keep the system level memory
871	demands of a long-lived program low.
872	2. Mapped memory can never become `locked' between
873	other chunks, as can happen with normally allocated chunks, which
874	means that even trimming via malloc_trim would not release them.
875	3. On some systems with "holes" in address spaces, mmap can obtain
876	memory that sbrk cannot.
877
878	However, it has the disadvantages that:
879
880	1. The space cannot be reclaimed, consolidated, and then
881	used to service later requests, as happens with normal chunks.
882	2. It can lead to more wastage because of mmap page alignment
883	requirements
884	3. It causes malloc performance to be more dependent on host
885	system memory management support routines which may vary in
886	implementation quality and may impose arbitrary
887	limitations. Generally, servicing a request via normal
888	malloc steps is faster than going through a system's mmap.
889
890	The advantages of mmap nearly always outweigh disadvantages for
891	"large" chunks, but the value of "large" varies across systems. The
892	default is an empirically derived value that works well in most
893	systems.
894
895
896	Update in 2006:
897	The above was written in 2001. Since then the world has changed a lot.
898	Memory got bigger. Applications got bigger. The virtual address space
899	layout in 32 bit linux changed.
900
901	In the new situation, brk() and mmap space is shared and there are no
902	artificial limits on brk size imposed by the kernel. What is more,
903	applications have started using transient allocations larger than the
904	128Kb as was imagined in 2001.
905
906	The price for mmap is also high now; each time glibc mmaps from the
907	kernel, the kernel is forced to zero out the memory it gives to the
908	application. Zeroing memory is expensive and eats a lot of cache and
909	memory bandwidth. This has nothing to do with the efficiency of the
910	virtual memory system, by doing mmap the kernel just has no choice but
911	to zero.
912
913	In 2001, the kernel had a maximum size for brk() which was about 800
914	megabytes on 32 bit x86, at that point brk() would hit the first
915	mmaped shared libaries and couldn't expand anymore. With current 2.6
916	kernels, the VA space layout is different and brk() and mmap
917	both can span the entire heap at will.
918
919	Rather than using a static threshold for the brk/mmap tradeoff,
920	we are now using a simple dynamic one. The goal is still to avoid
921	fragmentation. The old goals we kept are
922	1) try to get the long lived large allocations to use mmap()
923	2) really large allocations should always use mmap()
924	and we're adding now:
925	3) transient allocations should use brk() to avoid forcing the kernel
926	having to zero memory over and over again
927
928	The implementation works with a sliding threshold, which is by default
929	limited to go between 128Kb and 32Mb (64Mb for 64 bitmachines) and starts
930	out at 128Kb as per the 2001 default.
931
932	This allows us to satisfy requirement 1) under the assumption that long
933	lived allocations are made early in the process' lifespan, before it has
934	started doing dynamic allocations of the same size (which will
935	increase the threshold).
936
937	The upperbound on the threshold satisfies requirement 2)
938
939	The threshold goes up in value when the application frees memory that was
940	allocated with the mmap allocator. The idea is that once the application
941	starts freeing memory of a certain size, it's highly probable that this is
942	a size the application uses for transient allocations. This estimator
943	is there to satisfy the new third requirement.
944
945	*/
946
947	#define M_MMAP_THRESHOLD -3
948
949	#ifndef DEFAULT_MMAP_THRESHOLD
950	#define DEFAULT_MMAP_THRESHOLD DEFAULT_MMAP_THRESHOLD_MIN
951	#endif
952
953	/*
954	M_MMAP_MAX is the maximum number of requests to simultaneously
955	service using mmap. This parameter exists because
956	some systems have a limited number of internal tables for
957	use by mmap, and using more than a few of them may degrade
958	performance.
959
960	The default is set to a value that serves only as a safeguard.
961	Setting to 0 disables use of mmap for servicing large requests.
962	*/
963
964	#define M_MMAP_MAX -4
965
966	#ifndef DEFAULT_MMAP_MAX
967	#define DEFAULT_MMAP_MAX (65536)
968	#endif
969
970	#include <malloc.h>
971
972	#ifndef RETURN_ADDRESS
973	#define RETURN_ADDRESS(X_) (NULL)
974	#endif
975
976	/ On some platforms we can compile internal, not exported functions better.*
977	Let the environment provide a macro and define it to be empty if it
978	is not available. /*
979	#ifndef internal_function
980	# define internal_function
981	#endif
982
983	/ Forward declarations. /
984	struct malloc_chunk;
985	typedef struct malloc_chunk* mchunkptr;
986
987	/ Internal routines. /
988
989	static void* _int_malloc(mstate, size_t);
990	static void _int_free(mstate, mchunkptr, int);
991	static void* _int_realloc(mstate, mchunkptr, INTERNAL_SIZE_T,
992	INTERNAL_SIZE_T);
993	static void* _int_memalign(mstate, size_t, size_t);
994	static void* _mid_memalign(size_t, size_t, void *);
995
996	static void malloc_printerr(int action, const char str, void* *ptr, mstate av);
997
998	static void* internal_function mem2mem_check(void *p, size_t sz);
999	static int internal_function top_check(void);
1000	static void internal_function munmap_chunk(mchunkptr p);
1001	#if HAVE_MREMAP
1002	static mchunkptr internal_function mremap_chunk(mchunkptr p, size_t new_size);
1003	#endif
1004
1005	static void* malloc_check(size_t sz, const void *caller);
1006	static void free_check(void* mem, const void *caller);
1007	static void* realloc_check(void* oldmem, size_t bytes,
1008	const void *caller);
1009	static void* memalign_check(size_t alignment, size_t bytes,
1010	const void *caller);
1011
1012	/ ------------------ MMAP support ------------------ /
1013
1014
1015	#include <fcntl.h>
1016	#include <sys/mman.h>
1017
1018	#if !defined(MAP_ANONYMOUS) && defined(MAP_ANON)
1019	# define MAP_ANONYMOUS MAP_ANON
1020	#endif
1021
1022	#ifndef MAP_NORESERVE
1023	# define MAP_NORESERVE 0
1024	#endif
1025
1026	#define MMAP(addr, size, prot, flags) \
1027	__mmap((addr), (size), (prot), (flags)\|MAP_ANONYMOUS\|MAP_PRIVATE, -1, 0)
1028
1029
1030	/*
1031	----------------------- Chunk representations -----------------------
1032	*/
1033
1034
1035	/*
1036	This struct declaration is misleading (but accurate and necessary).
1037	It declares a "view" into memory allowing access to necessary
1038	fields at known offsets from a given base. See explanation below.
1039	*/
1040
1041	struct malloc_chunk {
1042
1043	INTERNAL_SIZE_T mchunk_prev_size; / Size of previous chunk (if free). /
1044	INTERNAL_SIZE_T mchunk_size; / Size in bytes, including overhead. /
1045
1046	struct malloc_chunk* fd; / double links -- used only if free. /
1047	struct malloc_chunk* bk;
1048
1049	/ Only used for large blocks: pointer to next larger size. /
1050	struct malloc_chunk* fd_nextsize; / double links -- used only if free. /
1051	struct malloc_chunk* bk_nextsize;
1052	};
1053
1054
1055	/*
1056	malloc_chunk details:
1057
1058	(The following includes lightly edited explanations by Colin Plumb.)
1059
1060	Chunks of memory are maintained using a `boundary tag' method as
1061	described in e.g., Knuth or Standish. (See the paper by Paul
1062	Wilson ftp://ftp.cs.utexas.edu/pub/garbage/allocsrv.ps for a
1063	survey of such techniques.) Sizes of free chunks are stored both
1064	in the front of each chunk and at the end. This makes
1065	consolidating fragmented chunks into bigger chunks very fast. The
1066	size fields also hold bits representing whether chunks are free or
1067	in use.
1068
1069	An allocated chunk looks like this:
1070
1071
1072	chunk-> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1073	\| Size of previous chunk, if unallocated (P clear) \|
1074	+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1075	\| Size of chunk, in bytes \|A\|M\|P\|
1076	mem-> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1077	\| User data starts here... .
1078	. .
1079	. (malloc_usable_size() bytes) .
1080	. \|
1081	nextchunk-> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1082	\| (size of chunk, but used for application data) \|
1083	+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1084	\| Size of next chunk, in bytes \|A\|0\|1\|
1085	+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1086
1087	Where "chunk" is the front of the chunk for the purpose of most of
1088	the malloc code, but "mem" is the pointer that is returned to the
1089	user. "Nextchunk" is the beginning of the next contiguous chunk.
1090
1091	Chunks always begin on even word boundaries, so the mem portion
1092	(which is returned to the user) is also on an even word boundary, and
1093	thus at least double-word aligned.
1094
1095	Free chunks are stored in circular doubly-linked lists, and look like this:
1096
1097	chunk-> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1098	\| Size of previous chunk, if unallocated (P clear) \|
1099	+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1100	`head:' \| Size of chunk, in bytes \|A\|0\|P\|
1101	mem-> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1102	\| Forward pointer to next chunk in list \|
1103	+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1104	\| Back pointer to previous chunk in list \|
1105	+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1106	\| Unused space (may be 0 bytes long) .
1107	. .
1108	. \|
1109	nextchunk-> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1110	`foot:' \| Size of chunk, in bytes \|
1111	+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1112	\| Size of next chunk, in bytes \|A\|0\|0\|
1113	+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1114
1115	The P (PREV_INUSE) bit, stored in the unused low-order bit of the
1116	chunk size (which is always a multiple of two words), is an in-use
1117	bit for the previous* chunk. If that bit is clear, then the*
1118	word before the current chunk size contains the previous chunk
1119	size, and can be used to find the front of the previous chunk.
1120	The very first chunk allocated always has this bit set,
1121	preventing access to non-existent (or non-owned) memory. If
1122	prev_inuse is set for any given chunk, then you CANNOT determine
1123	the size of the previous chunk, and might even get a memory
1124	addressing fault when trying to do so.
1125
1126	The A (NON_MAIN_ARENA) bit is cleared for chunks on the initial,
1127	main arena, described by the main_arena variable. When additional
1128	threads are spawned, each thread receives its own arena (up to a
1129	configurable limit, after which arenas are reused for multiple
1130	threads), and the chunks in these arenas have the A bit set. To
1131	find the arena for a chunk on such a non-main arena, heap_for_ptr
1132	performs a bit mask operation and indirection through the ar_ptr
1133	member of the per-heap header heap_info (see arena.c).
1134
1135	Note that the `foot' of the current chunk is actually represented
1136	as the prev_size of the NEXT chunk. This makes it easier to
1137	deal with alignments etc but can be very confusing when trying
1138	to extend or adapt this code.
1139
1140	The three exceptions to all this are:
1141
1142	1. The special chunk `top' doesn't bother using the
1143	trailing size field since there is no next contiguous chunk
1144	that would have to index off it. After initialization, `top'
1145	is forced to always exist. If it would become less than
1146	MINSIZE bytes long, it is replenished.
1147
1148	2. Chunks allocated via mmap, which have the second-lowest-order
1149	bit M (IS_MMAPPED) set in their size fields. Because they are
1150	allocated one-by-one, each must contain its own trailing size
1151	field. If the M bit is set, the other bits are ignored
1152	(because mmapped chunks are neither in an arena, nor adjacent
1153	to a freed chunk). The M bit is also used for chunks which
1154	originally came from a dumped heap via malloc_set_state in
1155	hooks.c.
1156
1157	3. Chunks in fastbins are treated as allocated chunks from the
1158	point of view of the chunk allocator. They are consolidated
1159	with their neighbors only in bulk, in malloc_consolidate.
1160	*/
1161
1162	/*
1163	---------- Size and alignment checks and conversions ----------
1164	*/
1165
1166	/ conversion from malloc headers to user pointers, and back /
1167
1168	#define chunk2mem(p) ((void)((char)(p) + 2*SIZE_SZ))
1169	#define mem2chunk(mem) ((mchunkptr)((char)(mem) - 2SIZE_SZ))
1170
1171	/ The smallest possible chunk /
1172	#define MIN_CHUNK_SIZE (offsetof(struct malloc_chunk, fd_nextsize))
1173
1174	/ The smallest size we can malloc is an aligned minimal chunk /
1175
1176	#define MINSIZE \
1177	(unsigned long)(((MIN_CHUNK_SIZE+MALLOC_ALIGN_MASK) & ~MALLOC_ALIGN_MASK))
1178
1179	/ Check if m has acceptable alignment /
1180
1181	#define aligned_OK(m) (((unsigned long)(m) & MALLOC_ALIGN_MASK) == 0)
1182
1183	#define misaligned_chunk(p) \
1184	((uintptr_t)(MALLOC_ALIGNMENT == 2 * SIZE_SZ ? (p) : chunk2mem (p)) \
1185	& MALLOC_ALIGN_MASK)
1186
1187
1188	/*
1189	Check if a request is so large that it would wrap around zero when
1190	padded and aligned. To simplify some other code, the bound is made
1191	low enough so that adding MINSIZE will also not wrap around zero.
1192	*/
1193
1194	#define REQUEST_OUT_OF_RANGE(req) \
1195	((unsigned long) (req) >= \
1196	(unsigned long) (INTERNAL_SIZE_T) (-2 * MINSIZE))
1197
1198	/ pad request bytes into a usable size -- internal version /
1199
1200	#define request2size(req) \
1201	(((req) + SIZE_SZ + MALLOC_ALIGN_MASK < MINSIZE) ? \
1202	MINSIZE : \
1203	((req) + SIZE_SZ + MALLOC_ALIGN_MASK) & ~MALLOC_ALIGN_MASK)
1204
1205	/ Same, except also perform an argument and result check. First, we check*
1206	that the padding done by request2size didn't result in an integer
1207	overflow. Then we check (using REQUEST_OUT_OF_RANGE) that the resulting
1208	size isn't so large that a later alignment would lead to another integer
1209	overflow. /*
1210	#define checked_request2size(req, sz) \
1211	({ \
1212	(sz) = request2size (req); \
1213	if (((sz) < (req)) \
1214	\|\| REQUEST_OUT_OF_RANGE (sz)) \
1215	{ \
1216	__set_errno (ENOMEM); \
1217	return 0; \
1218	} \
1219	})
1220
1221	/*
1222	--------------- Physical chunk operations ---------------
1223	*/
1224
1225
1226	/ size field is or'ed with PREV_INUSE when previous adjacent chunk in use /
1227	#define PREV_INUSE 0x1
1228
1229	/ extract inuse bit of previous chunk /
1230	#define prev_inuse(p) ((p)->mchunk_size & PREV_INUSE)
1231
1232
1233	/ size field is or'ed with IS_MMAPPED if the chunk was obtained with mmap() /
1234	#define IS_MMAPPED 0x2
1235
1236	/ check for mmap()'ed chunk /
1237	#define chunk_is_mmapped(p) ((p)->mchunk_size & IS_MMAPPED)
1238
1239
1240	/ size field is or'ed with NON_MAIN_ARENA if the chunk was obtained*
1241	from a non-main arena. This is only set immediately before handing
1242	the chunk to the user, if necessary. /*
1243	#define NON_MAIN_ARENA 0x4
1244
1245	/ Check for chunk from main arena. /
1246	#define chunk_main_arena(p) (((p)->mchunk_size & NON_MAIN_ARENA) == 0)
1247
1248	/ Mark a chunk as not being on the main arena. /
1249	#define set_non_main_arena(p) ((p)->mchunk_size \|= NON_MAIN_ARENA)
1250
1251
1252	/*
1253	Bits to mask off when extracting size
1254
1255	Note: IS_MMAPPED is intentionally not masked off from size field in
1256	macros for which mmapped chunks should never be seen. This should
1257	cause helpful core dumps to occur if it is tried by accident by
1258	people extending or adapting this malloc.
1259	*/
1260	#define SIZE_BITS (PREV_INUSE \| IS_MMAPPED \| NON_MAIN_ARENA)
1261
1262	/ Get size, ignoring use bits /
1263	#define chunksize(p) (chunksize_nomask (p) & ~(SIZE_BITS))
1264
1265	/ Like chunksize, but do not mask SIZE_BITS. /
1266	#define chunksize_nomask(p) ((p)->mchunk_size)
1267
1268	/ Ptr to next physical malloc_chunk. /
1269	#define next_chunk(p) ((mchunkptr) (((char *) (p)) + chunksize (p)))
1270
1271	/ Size of the chunk below P. Only valid if prev_inuse (P). /
1272	#define prev_size(p) ((p)->mchunk_prev_size)
1273
1274	/ Set the size of the chunk below P. Only valid if prev_inuse (P). /
1275	#define set_prev_size(p, sz) ((p)->mchunk_prev_size = (sz))
1276
1277	/ Ptr to previous physical malloc_chunk. Only valid if prev_inuse (P). /
1278	#define prev_chunk(p) ((mchunkptr) (((char *) (p)) - prev_size (p)))
1279
1280	/ Treat space at ptr + offset as a chunk /
1281	#define chunk_at_offset(p, s) ((mchunkptr) (((char *) (p)) + (s)))
1282
1283	/ extract p's inuse bit /
1284	#define inuse(p) \
1285	((((mchunkptr) (((char *) (p)) + chunksize (p)))->mchunk_size) & PREV_INUSE)
1286
1287	/ set/clear chunk as being inuse without otherwise disturbing /
1288	#define set_inuse(p) \
1289	((mchunkptr) (((char *) (p)) + chunksize (p)))->mchunk_size \|= PREV_INUSE
1290
1291	#define clear_inuse(p) \
1292	((mchunkptr) (((char *) (p)) + chunksize (p)))->mchunk_size &= ~(PREV_INUSE)
1293
1294
1295	/ check/set/clear inuse bits in known places /
1296	#define inuse_bit_at_offset(p, s) \
1297	(((mchunkptr) (((char *) (p)) + (s)))->mchunk_size & PREV_INUSE)
1298
1299	#define set_inuse_bit_at_offset(p, s) \
1300	(((mchunkptr) (((char *) (p)) + (s)))->mchunk_size \|= PREV_INUSE)
1301
1302	#define clear_inuse_bit_at_offset(p, s) \
1303	(((mchunkptr) (((char *) (p)) + (s)))->mchunk_size &= ~(PREV_INUSE))
1304
1305
1306	/ Set size at head, without disturbing its use bit /
1307	#define set_head_size(p, s) ((p)->mchunk_size = (((p)->mchunk_size & SIZE_BITS) \| (s)))
1308
1309	/ Set size/use field /
1310	#define set_head(p, s) ((p)->mchunk_size = (s))
1311
1312	/ Set size at footer (only when chunk is not in use) /
1313	#define set_foot(p, s) (((mchunkptr) ((char *) (p) + (s)))->mchunk_prev_size = (s))
1314
1315
1316	#pragma GCC poison mchunk_size
1317	#pragma GCC poison mchunk_prev_size
1318
1319	/*
1320	-------------------- Internal data structures --------------------
1321
1322	All internal state is held in an instance of malloc_state defined
1323	below. There are no other static variables, except in two optional
1324	cases:
1325	* If USE_MALLOC_LOCK is defined, the mALLOC_MUTEx declared above.
1326	* If mmap doesn't support MAP_ANONYMOUS, a dummy file descriptor
1327	for mmap.
1328
1329	Beware of lots of tricks that minimize the total bookkeeping space
1330	requirements. The result is a little over 1K bytes (for 4byte
1331	pointers and size_t.)
1332	*/
1333
1334	/*
1335	Bins
1336
1337	An array of bin headers for free chunks. Each bin is doubly
1338	linked. The bins are approximately proportionally (log) spaced.
1339	There are a lot of these bins (128). This may look excessive, but
1340	works very well in practice. Most bins hold sizes that are
1341	unusual as malloc request sizes, but are more usual for fragments
1342	and consolidated sets of chunks, which is what these bins hold, so
1343	they can be found quickly. All procedures maintain the invariant
1344	that no consolidated chunk physically borders another one, so each
1345	chunk in a list is known to be preceeded and followed by either
1346	inuse chunks or the ends of memory.
1347
1348	Chunks in bins are kept in size order, with ties going to the
1349	approximately least recently used chunk. Ordering isn't needed
1350	for the small bins, which all contain the same-sized chunks, but
1351	facilitates best-fit allocation for larger chunks. These lists
1352	are just sequential. Keeping them in order almost never requires
1353	enough traversal to warrant using fancier ordered data
1354	structures.
1355
1356	Chunks of the same size are linked with the most
1357	recently freed at the front, and allocations are taken from the
1358	back. This results in LRU (FIFO) allocation order, which tends
1359	to give each chunk an equal opportunity to be consolidated with
1360	adjacent freed chunks, resulting in larger free chunks and less
1361	fragmentation.
1362
1363	To simplify use in double-linked lists, each bin header acts
1364	as a malloc_chunk. This avoids special-casing for headers.
1365	But to conserve space and improve locality, we allocate
1366	only the fd/bk pointers of bins, and then use repositioning tricks
1367	to treat these as the fields of a malloc_chunk.*
1368	*/
1369
1370	typedef struct malloc_chunk *mbinptr;
1371
1372	/ addressing -- note that bin_at(0) does not exist /
1373	#define bin_at(m, i) \
1374	(mbinptr) (((char ) &((m)->bins[((i) - 1) 2])) \
1375	- offsetof (struct malloc_chunk, fd))
1376
1377	/ analog of ++bin /
1378	#define next_bin(b) ((mbinptr) ((char *) (b) + (sizeof (mchunkptr) << 1)))
1379
1380	/ Reminders about list directionality within bins /
1381	#define first(b) ((b)->fd)
1382	#define last(b) ((b)->bk)
1383
1384	/ Take a chunk off a bin list /
1385	#define unlink(AV, P, BK, FD) { \
1386	FD = P->fd; \
1387	BK = P->bk; \
1388	if (__builtin_expect (FD->bk != P \|\| BK->fd != P, 0)) \
1389	malloc_printerr (check_action, "corrupted double-linked list", P, AV); \
1390	else { \
1391	FD->bk = BK; \
1392	BK->fd = FD; \
1393	if (!in_smallbin_range (chunksize_nomask (P)) \
1394	&& __builtin_expect (P->fd_nextsize != NULL, 0)) { \
1395	if (__builtin_expect (P->fd_nextsize->bk_nextsize != P, 0) \
1396	\|\| __builtin_expect (P->bk_nextsize->fd_nextsize != P, 0)) \
1397	malloc_printerr (check_action, \
1398	"corrupted double-linked list (not small)", \
1399	P, AV); \
1400	if (FD->fd_nextsize == NULL) { \
1401	if (P->fd_nextsize == P) \
1402	FD->fd_nextsize = FD->bk_nextsize = FD; \
1403	else { \
1404	FD->fd_nextsize = P->fd_nextsize; \
1405	FD->bk_nextsize = P->bk_nextsize; \
1406	P->fd_nextsize->bk_nextsize = FD; \
1407	P->bk_nextsize->fd_nextsize = FD; \
1408	} \
1409	} else { \
1410	P->fd_nextsize->bk_nextsize = P->bk_nextsize; \
1411	P->bk_nextsize->fd_nextsize = P->fd_nextsize; \
1412	} \
1413	} \
1414	} \
1415	}
1416
1417	/*
1418	Indexing
1419
1420	Bins for sizes < 512 bytes contain chunks of all the same size, spaced
1421	8 bytes apart. Larger bins are approximately logarithmically spaced:
1422
1423	64 bins of size 8
1424	32 bins of size 64
1425	16 bins of size 512
1426	8 bins of size 4096
1427	4 bins of size 32768
1428	2 bins of size 262144
1429	1 bin of size what's left
1430
1431	There is actually a little bit of slop in the numbers in bin_index
1432	for the sake of speed. This makes no difference elsewhere.
1433
1434	The bins top out around 1MB because we expect to service large
1435	requests via mmap.
1436
1437	Bin 0 does not exist. Bin 1 is the unordered list; if that would be
1438	a valid chunk size the small bins are bumped up one.
1439	*/
1440
1441	#define NBINS 128
1442	#define NSMALLBINS 64
1443	#define SMALLBIN_WIDTH MALLOC_ALIGNMENT
1444	#define SMALLBIN_CORRECTION (MALLOC_ALIGNMENT > 2 * SIZE_SZ)
1445	#define MIN_LARGE_SIZE ((NSMALLBINS - SMALLBIN_CORRECTION) * SMALLBIN_WIDTH)
1446
1447	#define in_smallbin_range(sz) \
1448	((unsigned long) (sz) < (unsigned long) MIN_LARGE_SIZE)
1449
1450	#define smallbin_index(sz) \
1451	((SMALLBIN_WIDTH == 16 ? (((unsigned) (sz)) >> 4) : (((unsigned) (sz)) >> 3))\
1452	+ SMALLBIN_CORRECTION)
1453
1454	#define largebin_index_32(sz) \
1455	(((((unsigned long) (sz)) >> 6) <= 38) ? 56 + (((unsigned long) (sz)) >> 6) :\
1456	((((unsigned long) (sz)) >> 9) <= 20) ? 91 + (((unsigned long) (sz)) >> 9) :\
1457	((((unsigned long) (sz)) >> 12) <= 10) ? 110 + (((unsigned long) (sz)) >> 12) :\
1458	((((unsigned long) (sz)) >> 15) <= 4) ? 119 + (((unsigned long) (sz)) >> 15) :\
1459	((((unsigned long) (sz)) >> 18) <= 2) ? 124 + (((unsigned long) (sz)) >> 18) :\
1460	126)
1461
1462	#define largebin_index_32_big(sz) \
1463	(((((unsigned long) (sz)) >> 6) <= 45) ? 49 + (((unsigned long) (sz)) >> 6) :\
1464	((((unsigned long) (sz)) >> 9) <= 20) ? 91 + (((unsigned long) (sz)) >> 9) :\
1465	((((unsigned long) (sz)) >> 12) <= 10) ? 110 + (((unsigned long) (sz)) >> 12) :\
1466	((((unsigned long) (sz)) >> 15) <= 4) ? 119 + (((unsigned long) (sz)) >> 15) :\
1467	((((unsigned long) (sz)) >> 18) <= 2) ? 124 + (((unsigned long) (sz)) >> 18) :\
1468	126)
1469
1470	// XXX It remains to be seen whether it is good to keep the widths of
1471	// XXX the buckets the same or whether it should be scaled by a factor
1472	// XXX of two as well.
1473	#define largebin_index_64(sz) \
1474	(((((unsigned long) (sz)) >> 6) <= 48) ? 48 + (((unsigned long) (sz)) >> 6) :\
1475	((((unsigned long) (sz)) >> 9) <= 20) ? 91 + (((unsigned long) (sz)) >> 9) :\
1476	((((unsigned long) (sz)) >> 12) <= 10) ? 110 + (((unsigned long) (sz)) >> 12) :\
1477	((((unsigned long) (sz)) >> 15) <= 4) ? 119 + (((unsigned long) (sz)) >> 15) :\
1478	((((unsigned long) (sz)) >> 18) <= 2) ? 124 + (((unsigned long) (sz)) >> 18) :\
1479	126)
1480
1481	#define largebin_index(sz) \
1482	(SIZE_SZ == 8 ? largebin_index_64 (sz) \
1483	: MALLOC_ALIGNMENT == 16 ? largebin_index_32_big (sz) \
1484	: largebin_index_32 (sz))
1485
1486	#define bin_index(sz) \
1487	((in_smallbin_range (sz)) ? smallbin_index (sz) : largebin_index (sz))
1488
1489
1490	/*
1491	Unsorted chunks
1492
1493	All remainders from chunk splits, as well as all returned chunks,
1494	are first placed in the "unsorted" bin. They are then placed
1495	in regular bins after malloc gives them ONE chance to be used before
1496	binning. So, basically, the unsorted_chunks list acts as a queue,
1497	with chunks being placed on it in free (and malloc_consolidate),
1498	and taken off (to be either used or placed in bins) in malloc.
1499
1500	The NON_MAIN_ARENA flag is never set for unsorted chunks, so it
1501	does not have to be taken into account in size comparisons.
1502	*/
1503
1504	/ The otherwise unindexable 1-bin is used to hold unsorted chunks. /
1505	#define unsorted_chunks(M) (bin_at (M, 1))
1506
1507	/*
1508	Top
1509
1510	The top-most available chunk (i.e., the one bordering the end of
1511	available memory) is treated specially. It is never included in
1512	any bin, is used only if no other chunk is available, and is
1513	released back to the system if it is very large (see
1514	M_TRIM_THRESHOLD). Because top initially
1515	points to its own bin with initial zero size, thus forcing
1516	extension on the first malloc request, we avoid having any special
1517	code in malloc to check whether it even exists yet. But we still
1518	need to do so when getting memory from system, so we make
1519	initial_top treat the bin as a legal but unusable chunk during the
1520	interval between initialization and the first call to
1521	sysmalloc. (This is somewhat delicate, since it relies on
1522	the 2 preceding words to be zero during this interval as well.)
1523	*/
1524
1525	/ Conveniently, the unsorted bin can be used as dummy top on first call /
1526	#define initial_top(M) (unsorted_chunks (M))
1527
1528	/*
1529	Binmap
1530
1531	To help compensate for the large number of bins, a one-level index
1532	structure is used for bin-by-bin searching. `binmap' is a
1533	bitvector recording whether bins are definitely empty so they can
1534	be skipped over during during traversals. The bits are NOT always
1535	cleared as soon as bins are empty, but instead only
1536	when they are noticed to be empty during traversal in malloc.
1537	*/
1538
1539	/ Conservatively use 32 bits per map word, even if on 64bit system /
1540	#define BINMAPSHIFT 5
1541	#define BITSPERMAP (1U << BINMAPSHIFT)
1542	#define BINMAPSIZE (NBINS / BITSPERMAP)
1543
1544	#define idx2block(i) ((i) >> BINMAPSHIFT)
1545	#define idx2bit(i) ((1U << ((i) & ((1U << BINMAPSHIFT) - 1))))
1546
1547	#define mark_bin(m, i) ((m)->binmap[idx2block (i)] \|= idx2bit (i))
1548	#define unmark_bin(m, i) ((m)->binmap[idx2block (i)] &= ~(idx2bit (i)))
1549	#define get_binmap(m, i) ((m)->binmap[idx2block (i)] & idx2bit (i))
1550
1551	/*
1552	Fastbins
1553
1554	An array of lists holding recently freed small chunks. Fastbins
1555	are not doubly linked. It is faster to single-link them, and
1556	since chunks are never removed from the middles of these lists,
1557	double linking is not necessary. Also, unlike regular bins, they
1558	are not even processed in FIFO order (they use faster LIFO) since
1559	ordering doesn't much matter in the transient contexts in which
1560	fastbins are normally used.
1561
1562	Chunks in fastbins keep their inuse bit set, so they cannot
1563	be consolidated with other free chunks. malloc_consolidate
1564	releases all chunks in fastbins and consolidates them with
1565	other free chunks.
1566	*/
1567
1568	typedef struct malloc_chunk *mfastbinptr;
1569	#define fastbin(ar_ptr, idx) ((ar_ptr)->fastbinsY[idx])
1570
1571	/ offset 2 to use otherwise unindexable first 2 bins /
1572	#define fastbin_index(sz) \
1573	((((unsigned int) (sz)) >> (SIZE_SZ == 8 ? 4 : 3)) - 2)
1574
1575
1576	/ The maximum fastbin request size we support /
1577	#define MAX_FAST_SIZE (80 * SIZE_SZ / 4)
1578
1579	#define NFASTBINS (fastbin_index (request2size (MAX_FAST_SIZE)) + 1)
1580
1581	/*
1582	FASTBIN_CONSOLIDATION_THRESHOLD is the size of a chunk in free()
1583	that triggers automatic consolidation of possibly-surrounding
1584	fastbin chunks. This is a heuristic, so the exact value should not
1585	matter too much. It is defined at half the default trim threshold as a
1586	compromise heuristic to only attempt consolidation if it is likely
1587	to lead to trimming. However, it is not dynamically tunable, since
1588	consolidation reduces fragmentation surrounding large chunks even
1589	if trimming is not used.
1590	*/
1591
1592	#define FASTBIN_CONSOLIDATION_THRESHOLD (65536UL)
1593
1594	/*
1595	Since the lowest 2 bits in max_fast don't matter in size comparisons,
1596	they are used as flags.
1597	*/
1598
1599	/*
1600	FASTCHUNKS_BIT held in max_fast indicates that there are probably
1601	some fastbin chunks. It is set true on entering a chunk into any
1602	fastbin, and cleared only in malloc_consolidate.
1603
1604	The truth value is inverted so that have_fastchunks will be true
1605	upon startup (since statics are zero-filled), simplifying
1606	initialization checks.
1607	*/
1608
1609	#define FASTCHUNKS_BIT (1U)
1610
1611	#define have_fastchunks(M) (((M)->flags & FASTCHUNKS_BIT) == 0)
1612	#define clear_fastchunks(M) catomic_or (&(M)->flags, FASTCHUNKS_BIT)
1613	#define set_fastchunks(M) catomic_and (&(M)->flags, ~FASTCHUNKS_BIT)
1614
1615	/*
1616	NONCONTIGUOUS_BIT indicates that MORECORE does not return contiguous
1617	regions. Otherwise, contiguity is exploited in merging together,
1618	when possible, results from consecutive MORECORE calls.
1619
1620	The initial value comes from MORECORE_CONTIGUOUS, but is
1621	changed dynamically if mmap is ever used as an sbrk substitute.
1622	*/
1623
1624	#define NONCONTIGUOUS_BIT (2U)
1625
1626	#define contiguous(M) (((M)->flags & NONCONTIGUOUS_BIT) == 0)
1627	#define noncontiguous(M) (((M)->flags & NONCONTIGUOUS_BIT) != 0)
1628	#define set_noncontiguous(M) ((M)->flags \|= NONCONTIGUOUS_BIT)
1629	#define set_contiguous(M) ((M)->flags &= ~NONCONTIGUOUS_BIT)
1630
1631	/ ARENA_CORRUPTION_BIT is set if a memory corruption was detected on the*
1632	arena. Such an arena is no longer used to allocate chunks. Chunks
1633	allocated in that arena before detecting corruption are not freed. /*
1634
1635	#define ARENA_CORRUPTION_BIT (4U)
1636
1637	#define arena_is_corrupt(A) (((A)->flags & ARENA_CORRUPTION_BIT))
1638	#define set_arena_corrupt(A) ((A)->flags \|= ARENA_CORRUPTION_BIT)
1639
1640	/*
1641	Set value of max_fast.
1642	Use impossibly small value if 0.
1643	Precondition: there are no existing fastbin chunks.
1644	Setting the value clears fastchunk bit but preserves noncontiguous bit.
1645	*/
1646
1647	#define set_max_fast(s) \
1648	global_max_fast = (((s) == 0) \
1649	? SMALLBIN_WIDTH : ((s + SIZE_SZ) & ~MALLOC_ALIGN_MASK))
1650	#define get_max_fast() global_max_fast
1651
1652
1653	/*
1654	----------- Internal state representation and initialization -----------
1655	*/
1656
1657	struct malloc_state
1658	{
1659	/ Serialize access. /
1660	__libc_lock_define (, mutex);
1661
1662	/ Flags (formerly in max_fast). /
1663	int flags;
1664
1665	/ Fastbins /
1666	mfastbinptr fastbinsY[NFASTBINS];
1667
1668	/ Base of the topmost chunk -- not otherwise kept in a bin /
1669	mchunkptr top;
1670
1671	/ The remainder from the most recent split of a small request /
1672	mchunkptr last_remainder;
1673
1674	/ Normal bins packed as described above /
1675	mchunkptr bins[NBINS * `2` - `2`];
1676
1677	/ Bitmap of bins /
1678	unsigned int binmap[BINMAPSIZE];
1679
1680	/ Linked list /
1681	struct malloc_state *next;
1682
1683	/ Linked list for free arenas. Access to this field is serialized*
1684	by free_list_lock in arena.c. /*
1685	struct malloc_state *next_free;
1686
1687	/ Number of threads attached to this arena. 0 if the arena is on*
1688	the free list. Access to this field is serialized by
1689	free_list_lock in arena.c. /*
1690	INTERNAL_SIZE_T attached_threads;
1691
1692	/ Memory allocated from the system in this arena. /
1693	INTERNAL_SIZE_T system_mem;
1694	INTERNAL_SIZE_T max_system_mem;
1695	};
1696
1697	struct malloc_par
1698	{
1699	/ Tunable parameters /
1700	unsigned long trim_threshold;
1701	INTERNAL_SIZE_T top_pad;
1702	INTERNAL_SIZE_T mmap_threshold;
1703	INTERNAL_SIZE_T arena_test;
1704	INTERNAL_SIZE_T arena_max;
1705
1706	/ Memory map support /
1707	int n_mmaps;
1708	int n_mmaps_max;
1709	int max_n_mmaps;
1710	/ the mmap_threshold is dynamic, until the user sets*
1711	it manually, at which point we need to disable any
1712	dynamic behavior. /*
1713	int no_dyn_threshold;
1714
1715	/ Statistics /
1716	INTERNAL_SIZE_T mmapped_mem;
1717	INTERNAL_SIZE_T max_mmapped_mem;
1718
1719	/ First address handed out by MORECORE/sbrk. /
1720	char *sbrk_base;
1721	};
1722
1723	/ There are several instances of this struct ("arenas") in this*
1724	malloc. If you are adapting this malloc in a way that does NOT use
1725	a static or mmapped malloc_state, you MUST explicitly zero-fill it
1726	before using. This malloc relies on the property that malloc_state
1727	is initialized to all zeroes (as is true of C statics). /*
1728
1729	static struct malloc_state main_arena =
1730	{
1731	.mutex = _LIBC_LOCK_INITIALIZER,
1732	.next = &main_arena,
1733	.attached_threads = `1`
1734	};
1735
1736	/ These variables are used for undumping support. Chunked are marked*
1737	as using mmap, but we leave them alone if they fall into this
1738	range. NB: The chunk size for these chunks only includes the
1739	initial size field (of SIZE_SZ bytes), there is no trailing size
1740	field (unlike with regular mmapped chunks). /*
1741	static mchunkptr dumped_main_arena_start; / Inclusive. /
1742	static mchunkptr dumped_main_arena_end; / Exclusive. /
1743
1744	/ True if the pointer falls into the dumped arena. Use this after*
1745	chunk_is_mmapped indicates a chunk is mmapped. /*
1746	#define DUMPED_MAIN_ARENA_CHUNK(p) \
1747	((p) >= dumped_main_arena_start && (p) < dumped_main_arena_end)
1748
1749	/ There is only one instance of the malloc parameters. /
1750
1751	static struct malloc_par mp_ =
1752	{
1753	.top_pad = DEFAULT_TOP_PAD,
1754	.n_mmaps_max = DEFAULT_MMAP_MAX,
1755	.mmap_threshold = DEFAULT_MMAP_THRESHOLD,
1756	.trim_threshold = DEFAULT_TRIM_THRESHOLD,
1757	#define NARENAS_FROM_NCORES(n) ((n) * (sizeof (long) == 4 ? 2 : 8))
1758	.arena_test = NARENAS_FROM_NCORES (`1`)
1759	};
1760
1761	/ Maximum size of memory handled in fastbins. /
1762	static INTERNAL_SIZE_T global_max_fast;
1763
1764	/*
1765	Initialize a malloc_state struct.
1766
1767	This is called only from within malloc_consolidate, which needs
1768	be called in the same contexts anyway. It is never called directly
1769	outside of malloc_consolidate because some optimizing compilers try
1770	to inline it at all call points, which turns out not to be an
1771	optimization at all. (Inlining it in malloc_consolidate is fine though.)
1772	*/
1773
1774	static void
1775	malloc_init_state (mstate av)
1776	{
1777	int i;
1778	mbinptr bin;
1779
1780	/ Establish circular links for normal bins /
1781	for (i = `1`; i < NBINS; ++i)
1782	{
1783	bin = bin_at (av, i);
1784	bin->fd = bin->bk = bin;
1785	}
1786
1787	#if MORECORE_CONTIGUOUS
1788	if (av != &main_arena)
1789	#endif
1790	set_noncontiguous (av);
1791	if (av == &main_arena)
1792	set_max_fast (DEFAULT_MXFAST);
1793	av->flags \|= FASTCHUNKS_BIT;
1794
1795	av->top = initial_top (av);
1796	}
1797
1798	/*
1799	Other internal utilities operating on mstates
1800	*/
1801
1802	static void *sysmalloc (INTERNAL_SIZE_T, mstate);
1803	static int systrim (size_t, mstate);
1804	static void malloc_consolidate (mstate);
1805
1806
1807	/ -------------- Early definitions for debugging hooks ---------------- /
1808
1809	/ Define and initialize the hook variables. These weak definitions must*
1810	appear before any use of the variables in a function (arena.c uses one). /*
1811	#ifndef weak_variable
1812	/ In GNU libc we want the hook variables to be weak definitions to*
1813	avoid a problem with Emacs. /*
1814	# define weak_variable weak_function
1815	#endif
1816
1817	/ Forward declarations. /
1818	static void *malloc_hook_ini (size_t sz,
1819	const void *caller) __THROW;
1820	static void realloc_hook_ini (void* *ptr, size_t sz,
1821	const void *caller) __THROW;
1822	static void *memalign_hook_ini (size_t alignment, size_t sz,
1823	const void *caller) __THROW;
1824
1825	#if HAVE_MALLOC_INIT_HOOK
1826	void weak_variable (__malloc_initialize_hook) (void*) = NULL;
1827	compat_symbol (libc, __malloc_initialize_hook,
1828	__malloc_initialize_hook, GLIBC_2_0);
1829	#endif
1830
1831	void weak_variable (__free_hook) (void* *__ptr,
1832	const void *) = NULL;
1833	void weak_variable (__malloc_hook)
1834	(size_t __size, const void *) = malloc_hook_ini;
1835	void weak_variable (__realloc_hook)
1836	(void __ptr, size_t __size, const* void *)
1837	= realloc_hook_ini;
1838	void weak_variable (__memalign_hook)
1839	(size_t __alignment, size_t __size, const void *)
1840	= memalign_hook_ini;
1841	void weak_variable (__after_morecore_hook) (void*) = NULL;
1842
1843
1844	/ ---------------- Error behavior ------------------------------------ /
1845
1846	#ifndef DEFAULT_CHECK_ACTION
1847	# define DEFAULT_CHECK_ACTION 3
1848	#endif
1849
1850	static int check_action = DEFAULT_CHECK_ACTION;
1851
1852
1853	/ ------------------ Testing support ----------------------------------/
1854
1855	static int perturb_byte;
1856
1857	static void
1858	alloc_perturb (char *p, size_t n)
1859	{
1860	if (__glibc_unlikely (perturb_byte))
1861	memset (p, perturb_byte ^ `0xff`, n);
1862	}
1863
1864	static void
1865	free_perturb (char *p, size_t n)
1866	{
1867	if (__glibc_unlikely (perturb_byte))
1868	memset (p, perturb_byte, n);
1869	}
1870
1871
1872
1873	#include <stap-probe.h>
1874
1875	/ ------------------- Support for multiple arenas -------------------- /
1876	#include "arena.c"
1877
1878	/*
1879	Debugging support
1880
1881	These routines make a number of assertions about the states
1882	of data structures that should be true at all times. If any
1883	are not true, it's very likely that a user program has somehow
1884	trashed memory. (It's also possible that there is a coding error
1885	in malloc. In which case, please report it!)
1886	*/
1887
1888	#if !MALLOC_DEBUG
1889
1890	# define check_chunk(A, P)
1891	# define check_free_chunk(A, P)
1892	# define check_inuse_chunk(A, P)
1893	# define check_remalloced_chunk(A, P, N)
1894	# define check_malloced_chunk(A, P, N)
1895	# define check_malloc_state(A)
1896
1897	#else
1898
1899	# define check_chunk(A, P) do_check_chunk (A, P)
1900	# define check_free_chunk(A, P) do_check_free_chunk (A, P)
1901	# define check_inuse_chunk(A, P) do_check_inuse_chunk (A, P)
1902	# define check_remalloced_chunk(A, P, N) do_check_remalloced_chunk (A, P, N)
1903	# define check_malloced_chunk(A, P, N) do_check_malloced_chunk (A, P, N)
1904	# define check_malloc_state(A) do_check_malloc_state (A)
1905
1906	/*
1907	Properties of all chunks
1908	*/
1909
1910	static void
1911	do_check_chunk (mstate av, mchunkptr p)
1912	{
1913	unsigned long sz = chunksize (p);
1914	/ min and max possible addresses assuming contiguous allocation /
1915	char max_address = (char* *) (av->top) + chunksize (av->top);
1916	char *min_address = max_address - av->system_mem;
1917
1918	if (!chunk_is_mmapped (p))
1919	{
1920	/ Has legal address ... /
1921	if (p != av->top)
1922	{
1923	if (contiguous (av))
1924	{
1925	assert (((char *) p) >= min_address);
1926	assert (((char ) p + sz) <= ((char* *) (av->top)));
1927	}
1928	}
1929	else
1930	{
1931	/ top size is always at least MINSIZE /
1932	assert ((unsigned long) (sz) >= MINSIZE);
1933	/ top predecessor always marked inuse /
1934	assert (prev_inuse (p));
1935	}
1936	}
1937	else if (!DUMPED_MAIN_ARENA_CHUNK (p))
1938	{
1939	/ address is outside main heap /
1940	if (contiguous (av) && av->top != initial_top (av))
1941	{
1942	assert (((char ) p) < min_address \|\| ((char* *) p) >= max_address);
1943	}
1944	/ chunk is page-aligned /
1945	assert (((prev_size (p) + sz) & (GLRO (dl_pagesize) - `1`)) == `0`);
1946	/ mem is aligned /
1947	assert (aligned_OK (chunk2mem (p)));
1948	}
1949	}
1950
1951	/*
1952	Properties of free chunks
1953	*/
1954
1955	static void
1956	do_check_free_chunk (mstate av, mchunkptr p)
1957	{
1958	INTERNAL_SIZE_T sz = p->size & ~(PREV_INUSE \| NON_MAIN_ARENA);
1959	mchunkptr next = chunk_at_offset (p, sz);
1960
1961	do_check_chunk (av, p);
1962
1963	/ Chunk must claim to be free ... /
1964	assert (!inuse (p));
1965	assert (!chunk_is_mmapped (p));
1966
1967	/ Unless a special marker, must have OK fields /
1968	if ((unsigned long) (sz) >= MINSIZE)
1969	{
1970	assert ((sz & MALLOC_ALIGN_MASK) == `0`);
1971	assert (aligned_OK (chunk2mem (p)));
1972	/ ... matching footer field /
1973	assert (prev_size (p) == sz);
1974	/ ... and is fully consolidated /
1975	assert (prev_inuse (p));
1976	assert (next == av->top \|\| inuse (next));
1977
1978	/ ... and has minimally sane links /
1979	assert (p->fd->bk == p);
1980	assert (p->bk->fd == p);
1981	}
1982	else / markers are always of size SIZE_SZ /
1983	assert (sz == SIZE_SZ);
1984	}
1985
1986	/*
1987	Properties of inuse chunks
1988	*/
1989
1990	static void
1991	do_check_inuse_chunk (mstate av, mchunkptr p)
1992	{
1993	mchunkptr next;
1994
1995	do_check_chunk (av, p);
1996
1997	if (chunk_is_mmapped (p))
1998	return; / mmapped chunks have no next/prev /
1999
2000	/ Check whether it claims to be in use ... /
2001	assert (inuse (p));
2002
2003	next = next_chunk (p);
2004
2005	/ ... and is surrounded by OK chunks.*
2006	Since more things can be checked with free chunks than inuse ones,
2007	if an inuse chunk borders them and debug is on, it's worth doing them.
2008	*/
2009	if (!prev_inuse (p))
2010	{
2011	/ Note that we cannot even look at prev unless it is not inuse /
2012	mchunkptr prv = prev_chunk (p);
2013	assert (next_chunk (prv) == p);
2014	do_check_free_chunk (av, prv);
2015	}
2016
2017	if (next == av->top)
2018	{
2019	assert (prev_inuse (next));
2020	assert (chunksize (next) >= MINSIZE);
2021	}
2022	else if (!inuse (next))
2023	do_check_free_chunk (av, next);
2024	}
2025
2026	/*
2027	Properties of chunks recycled from fastbins
2028	*/
2029
2030	static void
2031	do_check_remalloced_chunk (mstate av, mchunkptr p, INTERNAL_SIZE_T s)
2032	{
2033	INTERNAL_SIZE_T sz = p->size & ~(PREV_INUSE \| NON_MAIN_ARENA);
2034
2035	if (!chunk_is_mmapped (p))
2036	{
2037	assert (av == arena_for_chunk (p));
2038	if (chunk_main_arena (p))
2039	assert (av == &main_arena);
2040	else
2041	assert (av != &main_arena);
2042	}
2043
2044	do_check_inuse_chunk (av, p);
2045
2046	/ Legal size ... /
2047	assert ((sz & MALLOC_ALIGN_MASK) == `0`);
2048	assert ((unsigned long) (sz) >= MINSIZE);
2049	/ ... and alignment /
2050	assert (aligned_OK (chunk2mem (p)));
2051	/ chunk is less than MINSIZE more than request /
2052	assert ((long) (sz) - (long) (s) >= `0`);
2053	assert ((long) (sz) - (long) (s + MINSIZE) < `0`);
2054	}
2055
2056	/*
2057	Properties of nonrecycled chunks at the point they are malloced
2058	*/
2059
2060	static void
2061	do_check_malloced_chunk (mstate av, mchunkptr p, INTERNAL_SIZE_T s)
2062	{
2063	/ same as recycled case ... /
2064	do_check_remalloced_chunk (av, p, s);
2065
2066	/*
2067	... plus, must obey implementation invariant that prev_inuse is
2068	always true of any allocated chunk; i.e., that each allocated
2069	chunk borders either a previously allocated and still in-use
2070	chunk, or the base of its memory arena. This is ensured
2071	by making all allocations from the `lowest' part of any found
2072	chunk. This does not necessarily hold however for chunks
2073	recycled via fastbins.
2074	*/
2075
2076	assert (prev_inuse (p));
2077	}
2078
2079
2080	/*
2081	Properties of malloc_state.
2082
2083	This may be useful for debugging malloc, as well as detecting user
2084	programmer errors that somehow write into malloc_state.
2085
2086	If you are extending or experimenting with this malloc, you can
2087	probably figure out how to hack this routine to print out or
2088	display chunk addresses, sizes, bins, and other instrumentation.
2089	*/
2090
2091	static void
2092	do_check_malloc_state (mstate av)
2093	{
2094	int i;
2095	mchunkptr p;
2096	mchunkptr q;
2097	mbinptr b;
2098	unsigned int idx;
2099	INTERNAL_SIZE_T size;
2100	unsigned long total = `0`;
2101	int max_fast_bin;
2102
2103	/ internal size_t must be no wider than pointer type /
2104	assert (sizeof (INTERNAL_SIZE_T) <= sizeof (char *));
2105
2106	/ alignment is a power of 2 /
2107	assert ((MALLOC_ALIGNMENT & (MALLOC_ALIGNMENT - `1`)) == `0`);
2108
2109	/ cannot run remaining checks until fully initialized /
2110	if (av->top == `0` \|\| av->top == initial_top (av))
2111	return;
2112
2113	/ pagesize is a power of 2 /
2114	assert (powerof2(GLRO (dl_pagesize)));
2115
2116	/ A contiguous main_arena is consistent with sbrk_base. /
2117	if (av == &main_arena && contiguous (av))
2118	assert ((char *) mp_.sbrk_base + av->system_mem ==
2119	(char *) av->top + chunksize (av->top));
2120
2121	/ properties of fastbins /
2122
2123	/ max_fast is in allowed range /
2124	assert ((get_max_fast () & ~`1`) <= request2size (MAX_FAST_SIZE));
2125
2126	max_fast_bin = fastbin_index (get_max_fast ());
2127
2128	for (i = `0`; i < NFASTBINS; ++i)
2129	{
2130	p = fastbin (av, i);
2131
2132	/ The following test can only be performed for the main arena.*
2133	While mallopt calls malloc_consolidate to get rid of all fast
2134	bins (especially those larger than the new maximum) this does
2135	only happen for the main arena. Trying to do this for any
2136	other arena would mean those arenas have to be locked and
2137	malloc_consolidate be called for them. This is excessive. And
2138	even if this is acceptable to somebody it still cannot solve
2139	the problem completely since if the arena is locked a
2140	concurrent malloc call might create a new arena which then
2141	could use the newly invalid fast bins. /*
2142
2143	/ all bins past max_fast are empty /
2144	if (av == &main_arena && i > max_fast_bin)
2145	assert (p == `0`);
2146
2147	while (p != `0`)
2148	{
2149	/ each chunk claims to be inuse /
2150	do_check_inuse_chunk (av, p);
2151	total += chunksize (p);
2152	/ chunk belongs in this bin /
2153	assert (fastbin_index (chunksize (p)) == i);
2154	p = p->fd;
2155	}
2156	}
2157
2158	if (total != `0`)
2159	assert (have_fastchunks (av));
2160	else if (!have_fastchunks (av))
2161	assert (total == `0`);
2162
2163	/ check normal bins /
2164	for (i = `1`; i < NBINS; ++i)
2165	{
2166	b = bin_at (av, i);
2167
2168	/ binmap is accurate (except for bin 1 == unsorted_chunks) /
2169	if (i >= `2`)
2170	{
2171	unsigned int binbit = get_binmap (av, i);
2172	int empty = last (b) == b;
2173	if (!binbit)
2174	assert (empty);
2175	else if (!empty)
2176	assert (binbit);
2177	}
2178
2179	for (p = last (b); p != b; p = p->bk)
2180	{
2181	/ each chunk claims to be free /
2182	do_check_free_chunk (av, p);
2183	size = chunksize (p);
2184	total += size;
2185	if (i >= `2`)
2186	{
2187	/ chunk belongs in bin /
2188	idx = bin_index (size);
2189	assert (idx == i);
2190	/ lists are sorted /
2191	assert (p->bk == b \|\|
2192	(unsigned long) chunksize (p->bk) >= (unsigned long) chunksize (p));
2193
2194	if (!in_smallbin_range (size))
2195	{
2196	if (p->fd_nextsize != NULL)
2197	{
2198	if (p->fd_nextsize == p)
2199	assert (p->bk_nextsize == p);
2200	else
2201	{
2202	if (p->fd_nextsize == first (b))
2203	assert (chunksize (p) < chunksize (p->fd_nextsize));
2204	else
2205	assert (chunksize (p) > chunksize (p->fd_nextsize));
2206
2207	if (p == first (b))
2208	assert (chunksize (p) > chunksize (p->bk_nextsize));
2209	else
2210	assert (chunksize (p) < chunksize (p->bk_nextsize));
2211	}
2212	}
2213	else
2214	assert (p->bk_nextsize == NULL);
2215	}
2216	}
2217	else if (!in_smallbin_range (size))
2218	assert (p->fd_nextsize == NULL && p->bk_nextsize == NULL);
2219	/ chunk is followed by a legal chain of inuse chunks /
2220	for (q = next_chunk (p);
2221	(q != av->top && inuse (q) &&
2222	(unsigned long) (chunksize (q)) >= MINSIZE);
2223	q = next_chunk (q))
2224	do_check_inuse_chunk (av, q);
2225	}
2226	}
2227
2228	/ top chunk is OK /
2229	check_chunk (av, av->top);
2230	}
2231	#endif
2232
2233
2234	/ ----------------- Support for debugging hooks -------------------- /
2235	#include "hooks.c"
2236
2237
2238	/ ----------- Routines dealing with system allocation -------------- /
2239
2240	/*
2241	sysmalloc handles malloc cases requiring more memory from the system.
2242	On entry, it is assumed that av->top does not have enough
2243	space to service request for nb bytes, thus requiring that av->top
2244	be extended or replaced.
2245	*/
2246
2247	static void *
2248	sysmalloc (INTERNAL_SIZE_T nb, mstate av)
2249	{
2250	mchunkptr old_top; / incoming value of av->top /
2251	INTERNAL_SIZE_T old_size; / its size /
2252	char old_end; /* its end address /
2253
2254	long size; / arg to first MORECORE or mmap call /
2255	char brk; /* return value from MORECORE /
2256
2257	long correction; / arg to 2nd MORECORE call /
2258	char snd_brk; /* 2nd return val /
2259
2260	INTERNAL_SIZE_T front_misalign; / unusable bytes at front of new space /
2261	INTERNAL_SIZE_T end_misalign; / partial page left at end of new space /
2262	char aligned_brk; /* aligned offset into brk /
2263
2264	mchunkptr p; / the allocated/returned chunk /
2265	mchunkptr remainder; / remainder from allocation /
2266	unsigned long remainder_size; / its size /
2267
2268
2269	size_t pagesize = GLRO (dl_pagesize);
2270	bool tried_mmap = false;
2271
2272
2273	/*
2274	If have mmap, and the request size meets the mmap threshold, and
2275	the system supports mmap, and there are few enough currently
2276	allocated mmapped regions, try to directly map this request
2277	rather than expanding top.
2278	*/
2279
2280	if (av == NULL
2281	\|\| ((unsigned long) (nb) >= (unsigned long) (mp_.mmap_threshold)
2282	&& (mp_.n_mmaps < mp_.n_mmaps_max)))
2283	{
2284	char mm; /* return value from mmap call/
2285
2286	try_mmap:
2287	/*
2288	Round up size to nearest page. For mmapped chunks, the overhead
2289	is one SIZE_SZ unit larger than for normal chunks, because there
2290	is no following chunk whose prev_size field could be used.
2291
2292	See the front_misalign handling below, for glibc there is no
2293	need for further alignments unless we have have high alignment.
2294	*/
2295	if (MALLOC_ALIGNMENT == `2` * SIZE_SZ)
2296	size = ALIGN_UP (nb + SIZE_SZ, pagesize);
2297	else
2298	size = ALIGN_UP (nb + SIZE_SZ + MALLOC_ALIGN_MASK, pagesize);
2299	tried_mmap = true;
2300
2301	/ Don't try if size wraps around 0 /
2302	if ((unsigned long) (size) > (unsigned long) (nb))
2303	{
2304	mm = (char *) (MMAP (`0`, size, PROT_READ \| PROT_WRITE, `0`));
2305
2306	if (mm != MAP_FAILED)
2307	{
2308	/*
2309	The offset to the start of the mmapped region is stored
2310	in the prev_size field of the chunk. This allows us to adjust
2311	returned start address to meet alignment requirements here
2312	and in memalign(), and still be able to compute proper
2313	address argument for later munmap in free() and realloc().
2314	*/
2315
2316	if (MALLOC_ALIGNMENT == `2` * SIZE_SZ)
2317	{
2318	/ For glibc, chunk2mem increases the address by 2SIZE_SZ and
2319	MALLOC_ALIGN_MASK is 2SIZE_SZ-1. Each mmap'ed area is page*
2320	aligned and therefore definitely MALLOC_ALIGN_MASK-aligned. /*
2321	assert (((INTERNAL_SIZE_T) chunk2mem (mm) & MALLOC_ALIGN_MASK) == `0`);
2322	front_misalign = `0`;
2323	}
2324	else
2325	front_misalign = (INTERNAL_SIZE_T) chunk2mem (mm) & MALLOC_ALIGN_MASK;
2326	if (front_misalign > `0`)
2327	{
2328	correction = MALLOC_ALIGNMENT - front_misalign;
2329	p = (mchunkptr) (mm + correction);
2330	set_prev_size (p, correction);
2331	set_head (p, (size - correction) \| IS_MMAPPED);
2332	}
2333	else
2334	{
2335	p = (mchunkptr) mm;
2336	set_prev_size (p, `0`);
2337	set_head (p, size \| IS_MMAPPED);
2338	}
2339
2340	/ update statistics /
2341
2342	int new = atomic_exchange_and_add (&mp_.n_mmaps, `1`) + `1`;
2343	atomic_max (&mp_.max_n_mmaps, new);
2344
2345	unsigned long sum;
2346	sum = atomic_exchange_and_add (&mp_.mmapped_mem, size) + size;
2347	atomic_max (&mp_.max_mmapped_mem, sum);
2348
2349	check_chunk (av, p);
2350
2351	return chunk2mem (p);
2352	}
2353	}
2354	}
2355
2356	/ There are no usable arenas and mmap also failed. /
2357	if (av == NULL)
2358	return `0`;
2359
2360	/ Record incoming configuration of top /
2361
2362	old_top = av->top;
2363	old_size = chunksize (old_top);
2364	old_end = (char *) (chunk_at_offset (old_top, old_size));
2365
2366	brk = snd_brk = (char *) (MORECORE_FAILURE);
2367
2368	/*
2369	If not the first time through, we require old_size to be
2370	at least MINSIZE and to have prev_inuse set.
2371	*/
2372
2373	assert ((old_top == initial_top (av) && old_size == `0`) \|\|
2374	((unsigned long) (old_size) >= MINSIZE &&
2375	prev_inuse (old_top) &&
2376	((unsigned long) old_end & (pagesize - `1`)) == `0`));
2377
2378	/ Precondition: not enough current space to satisfy nb request /
2379	assert ((unsigned long) (old_size) < (unsigned long) (nb + MINSIZE));
2380
2381
2382	if (av != &main_arena)
2383	{
2384	heap_info old_heap, heap;
2385	size_t old_heap_size;
2386
2387	/ First try to extend the current heap. /
2388	old_heap = heap_for_ptr (old_top);
2389	old_heap_size = old_heap->size;
2390	if ((long) (MINSIZE + nb - old_size) > `0`
2391	&& grow_heap (old_heap, MINSIZE + nb - old_size) == `0`)
2392	{
2393	av->system_mem += old_heap->size - old_heap_size;
2394	set_head (old_top, (((char ) old_heap + old_heap->size) - (char* *) old_top)
2395	\| PREV_INUSE);
2396	}
2397	else if ((heap = new_heap (nb + (MINSIZE + sizeof (*heap)), mp_.top_pad)))
2398	{
2399	/ Use a newly allocated heap. /
2400	heap->ar_ptr = av;
2401	heap->prev = old_heap;
2402	av->system_mem += heap->size;
2403	/ Set up the new top. /
2404	top (av) = chunk_at_offset (heap, sizeof (*heap));
2405	set_head (top (av), (heap->size - sizeof (*heap)) \| PREV_INUSE);
2406
2407	/ Setup fencepost and free the old top chunk with a multiple of*
2408	MALLOC_ALIGNMENT in size. /*
2409	/ The fencepost takes at least MINSIZE bytes, because it might*
2410	become the top chunk again later. Note that a footer is set
2411	up, too, although the chunk is marked in use. /*
2412	old_size = (old_size - MINSIZE) & ~MALLOC_ALIGN_MASK;
2413	set_head (chunk_at_offset (old_top, old_size + `2` * SIZE_SZ), `0` \| PREV_INUSE);
2414	if (old_size >= MINSIZE)
2415	{
2416	set_head (chunk_at_offset (old_top, old_size), (`2` * SIZE_SZ) \| PREV_INUSE);
2417	set_foot (chunk_at_offset (old_top, old_size), (`2` * SIZE_SZ));
2418	set_head (old_top, old_size \| PREV_INUSE \| NON_MAIN_ARENA);
2419	_int_free (av, old_top, `1`);
2420	}
2421	else
2422	{
2423	set_head (old_top, (old_size + `2` * SIZE_SZ) \| PREV_INUSE);
2424	set_foot (old_top, (old_size + `2` * SIZE_SZ));
2425	}
2426	}
2427	else if (!tried_mmap)
2428	/ We can at least try to use to mmap memory. /
2429	goto try_mmap;
2430	}
2431	else / av == main_arena /
2432
2433
2434	{ / Request enough space for nb + pad + overhead /
2435	size = nb + mp_.top_pad + MINSIZE;
2436
2437	/*
2438	If contiguous, we can subtract out existing space that we hope to
2439	combine with new space. We add it back later only if
2440	we don't actually get contiguous space.
2441	*/
2442
2443	if (contiguous (av))
2444	size -= old_size;
2445
2446	/*
2447	Round to a multiple of page size.
2448	If MORECORE is not contiguous, this ensures that we only call it
2449	with whole-page arguments. And if MORECORE is contiguous and
2450	this is not first time through, this preserves page-alignment of
2451	previous calls. Otherwise, we correct to page-align below.
2452	*/
2453
2454	size = ALIGN_UP (size, pagesize);
2455
2456	/*
2457	Don't try to call MORECORE if argument is so big as to appear
2458	negative. Note that since mmap takes size_t arg, it may succeed
2459	below even if we cannot call MORECORE.
2460	*/
2461
2462	if (size > `0`)
2463	{
2464	brk = (char *) (MORECORE (size));
2465	LIBC_PROBE (memory_sbrk_more, `2`, brk, size);
2466	}
2467
2468	if (brk != (char *) (MORECORE_FAILURE))
2469	{
2470	/ Call the `morecore' hook if necessary. /
2471	void (hook) (void*) = atomic_forced_read (__after_morecore_hook);
2472	if (__builtin_expect (hook != NULL, `0`))
2473	(*hook)();
2474	}
2475	else
2476	{
2477	/*
2478	If have mmap, try using it as a backup when MORECORE fails or
2479	cannot be used. This is worth doing on systems that have "holes" in
2480	address space, so sbrk cannot extend to give contiguous space, but
2481	space is available elsewhere. Note that we ignore mmap max count
2482	and threshold limits, since the space will not be used as a
2483	segregated mmap region.
2484	*/
2485
2486	/ Cannot merge with old top, so add its size back in /
2487	if (contiguous (av))
2488	size = ALIGN_UP (size + old_size, pagesize);
2489
2490	/ If we are relying on mmap as backup, then use larger units /
2491	if ((unsigned long) (size) < (unsigned long) (MMAP_AS_MORECORE_SIZE))
2492	size = MMAP_AS_MORECORE_SIZE;
2493
2494	/ Don't try if size wraps around 0 /
2495	if ((unsigned long) (size) > (unsigned long) (nb))
2496	{
2497	char mbrk = (char* *) (MMAP (`0`, size, PROT_READ \| PROT_WRITE, `0`));
2498
2499	if (mbrk != MAP_FAILED)
2500	{
2501	/ We do not need, and cannot use, another sbrk call to find end /
2502	brk = mbrk;
2503	snd_brk = brk + size;
2504
2505	/*
2506	Record that we no longer have a contiguous sbrk region.
2507	After the first time mmap is used as backup, we do not
2508	ever rely on contiguous space since this could incorrectly
2509	bridge regions.
2510	*/
2511	set_noncontiguous (av);
2512	}
2513	}
2514	}
2515
2516	if (brk != (char *) (MORECORE_FAILURE))
2517	{
2518	if (mp_.sbrk_base == `0`)
2519	mp_.sbrk_base = brk;
2520	av->system_mem += size;
2521
2522	/*
2523	If MORECORE extends previous space, we can likewise extend top size.
2524	*/
2525
2526	if (brk == old_end && snd_brk == (char *) (MORECORE_FAILURE))
2527	set_head (old_top, (size + old_size) \| PREV_INUSE);
2528
2529	else if (contiguous (av) && old_size && brk < old_end)
2530	{
2531	/ Oops! Someone else killed our space.. Can't touch anything. /
2532	malloc_printerr (`3`, "break adjusted to free malloc space", brk,
2533	av);
2534	}
2535
2536	/*
2537	Otherwise, make adjustments:
2538
2539	* If the first time through or noncontiguous, we need to call sbrk
2540	just to find out where the end of memory lies.
2541
2542	* We need to ensure that all returned chunks from malloc will meet
2543	MALLOC_ALIGNMENT
2544
2545	* If there was an intervening foreign sbrk, we need to adjust sbrk
2546	request size to account for fact that we will not be able to
2547	combine new space with existing space in old_top.
2548
2549	* Almost all systems internally allocate whole pages at a time, in
2550	which case we might as well use the whole last page of request.
2551	So we allocate enough more memory to hit a page boundary now,
2552	which in turn causes future contiguous calls to page-align.
2553	*/
2554
2555	else
2556	{
2557	front_misalign = `0`;
2558	end_misalign = `0`;
2559	correction = `0`;
2560	aligned_brk = brk;
2561
2562	/ handle contiguous cases /
2563	if (contiguous (av))
2564	{
2565	/ Count foreign sbrk as system_mem. /
2566	if (old_size)
2567	av->system_mem += brk - old_end;
2568
2569	/ Guarantee alignment of first new chunk made from this space /
2570
2571	front_misalign = (INTERNAL_SIZE_T) chunk2mem (brk) & MALLOC_ALIGN_MASK;
2572	if (front_misalign > `0`)
2573	{
2574	/*
2575	Skip over some bytes to arrive at an aligned position.
2576	We don't need to specially mark these wasted front bytes.
2577	They will never be accessed anyway because
2578	prev_inuse of av->top (and any chunk created from its start)
2579	is always true after initialization.
2580	*/
2581
2582	correction = MALLOC_ALIGNMENT - front_misalign;
2583	aligned_brk += correction;
2584	}
2585
2586	/*
2587	If this isn't adjacent to existing space, then we will not
2588	be able to merge with old_top space, so must add to 2nd request.
2589	*/
2590
2591	correction += old_size;
2592
2593	/ Extend the end address to hit a page boundary /
2594	end_misalign = (INTERNAL_SIZE_T) (brk + size + correction);
2595	correction += (ALIGN_UP (end_misalign, pagesize)) - end_misalign;
2596
2597	assert (correction >= `0`);
2598	snd_brk = (char *) (MORECORE (correction));
2599
2600	/*
2601	If can't allocate correction, try to at least find out current
2602	brk. It might be enough to proceed without failing.
2603
2604	Note that if second sbrk did NOT fail, we assume that space
2605	is contiguous with first sbrk. This is a safe assumption unless
2606	program is multithreaded but doesn't use locks and a foreign sbrk
2607	occurred between our first and second calls.
2608	*/
2609
2610	if (snd_brk == (char *) (MORECORE_FAILURE))
2611	{
2612	correction = `0`;
2613	snd_brk = (char *) (MORECORE (`0`));
2614	}
2615	else
2616	{
2617	/ Call the `morecore' hook if necessary. /
2618	void (hook) (void*) = atomic_forced_read (__after_morecore_hook);
2619	if (__builtin_expect (hook != NULL, `0`))
2620	(*hook)();
2621	}
2622	}
2623
2624	/ handle non-contiguous cases /
2625	else
2626	{
2627	if (MALLOC_ALIGNMENT == `2` * SIZE_SZ)
2628	/ MORECORE/mmap must correctly align /
2629	assert (((unsigned long) chunk2mem (brk) & MALLOC_ALIGN_MASK) == `0`);
2630	else
2631	{
2632	front_misalign = (INTERNAL_SIZE_T) chunk2mem (brk) & MALLOC_ALIGN_MASK;
2633	if (front_misalign > `0`)
2634	{
2635	/*
2636	Skip over some bytes to arrive at an aligned position.
2637	We don't need to specially mark these wasted front bytes.
2638	They will never be accessed anyway because
2639	prev_inuse of av->top (and any chunk created from its start)
2640	is always true after initialization.
2641	*/
2642
2643	aligned_brk += MALLOC_ALIGNMENT - front_misalign;
2644	}
2645	}
2646
2647	/ Find out current end of memory /
2648	if (snd_brk == (char *) (MORECORE_FAILURE))
2649	{
2650	snd_brk = (char *) (MORECORE (`0`));
2651	}
2652	}
2653
2654	/ Adjust top based on results of second sbrk /
2655	if (snd_brk != (char *) (MORECORE_FAILURE))
2656	{
2657	av->top = (mchunkptr) aligned_brk;
2658	set_head (av->top, (snd_brk - aligned_brk + correction) \| PREV_INUSE);
2659	av->system_mem += correction;
2660
2661	/*
2662	If not the first time through, we either have a
2663	gap due to foreign sbrk or a non-contiguous region. Insert a
2664	double fencepost at old_top to prevent consolidation with space
2665	we don't own. These fenceposts are artificial chunks that are
2666	marked as inuse and are in any case too small to use. We need
2667	two to make sizes and alignments work out.
2668	*/
2669
2670	if (old_size != `0`)
2671	{
2672	/*
2673	Shrink old_top to insert fenceposts, keeping size a
2674	multiple of MALLOC_ALIGNMENT. We know there is at least
2675	enough space in old_top to do this.
2676	*/
2677	old_size = (old_size - `4` * SIZE_SZ) & ~MALLOC_ALIGN_MASK;
2678	set_head (old_top, old_size \| PREV_INUSE);
2679
2680	/*
2681	Note that the following assignments completely overwrite
2682	old_top when old_size was previously MINSIZE. This is
2683	intentional. We need the fencepost, even if old_top otherwise gets
2684	lost.
2685	*/
2686	set_head (chunk_at_offset (old_top, old_size),
2687	(`2` * SIZE_SZ) \| PREV_INUSE);
2688	set_head (chunk_at_offset (old_top, old_size + `2` * SIZE_SZ),
2689	(`2` * SIZE_SZ) \| PREV_INUSE);
2690
2691	/ If possible, release the rest. /
2692	if (old_size >= MINSIZE)
2693	{
2694	_int_free (av, old_top, `1`);
2695	}
2696	}
2697	}
2698	}
2699	}
2700	} / if (av != &main_arena) /
2701
2702	if ((unsigned long) av->system_mem > (unsigned long) (av->max_system_mem))
2703	av->max_system_mem = av->system_mem;
2704	check_malloc_state (av);
2705
2706	/ finally, do the allocation /
2707	p = av->top;
2708	size = chunksize (p);
2709
2710	/ check that one of the above allocation paths succeeded /
2711	if ((unsigned long) (size) >= (unsigned long) (nb + MINSIZE))
2712	{
2713	remainder_size = size - nb;
2714	remainder = chunk_at_offset (p, nb);
2715	av->top = remainder;
2716	set_head (p, nb \| PREV_INUSE \| (av != &main_arena ? NON_MAIN_ARENA : `0`));
2717	set_head (remainder, remainder_size \| PREV_INUSE);
2718	check_malloced_chunk (av, p, nb);
2719	return chunk2mem (p);
2720	}
2721
2722	/ catch all failure paths /
2723	__set_errno (ENOMEM);
2724	return `0`;
2725	}
2726
2727
2728	/*
2729	systrim is an inverse of sorts to sysmalloc. It gives memory back
2730	to the system (via negative arguments to sbrk) if there is unused
2731	memory at the `high' end of the malloc pool. It is called
2732	automatically by free() when top space exceeds the trim
2733	threshold. It is also called by the public malloc_trim routine. It
2734	returns 1 if it actually released any memory, else 0.
2735	*/
2736
2737	static int
2738	systrim (size_t pad, mstate av)
2739	{
2740	long top_size; / Amount of top-most memory /
2741	long extra; / Amount to release /
2742	long released; / Amount actually released /
2743	char current_brk; /* address returned by pre-check sbrk call /
2744	char new_brk; /* address returned by post-check sbrk call /
2745	size_t pagesize;
2746	long top_area;
2747
2748	pagesize = GLRO (dl_pagesize);
2749	top_size = chunksize (av->top);
2750
2751	top_area = top_size - MINSIZE - `1`;
2752	if (top_area <= pad)
2753	return `0`;
2754
2755	/ Release in pagesize units and round down to the nearest page. /
2756	extra = ALIGN_DOWN(top_area - pad, pagesize);
2757
2758	if (extra == `0`)
2759	return `0`;
2760
2761	/*
2762	Only proceed if end of memory is where we last set it.
2763	This avoids problems if there were foreign sbrk calls.
2764	*/
2765	current_brk = (char *) (MORECORE (`0`));
2766	if (current_brk == (char *) (av->top) + top_size)
2767	{
2768	/*
2769	Attempt to release memory. We ignore MORECORE return value,
2770	and instead call again to find out where new end of memory is.
2771	This avoids problems if first call releases less than we asked,
2772	of if failure somehow altered brk value. (We could still
2773	encounter problems if it altered brk in some very bad way,
2774	but the only thing we can do is adjust anyway, which will cause
2775	some downstream failure.)
2776	*/
2777
2778	MORECORE (-extra);
2779	/ Call the `morecore' hook if necessary. /
2780	void (hook) (void*) = atomic_forced_read (__after_morecore_hook);
2781	if (__builtin_expect (hook != NULL, `0`))
2782	(*hook)();
2783	new_brk = (char *) (MORECORE (`0`));
2784
2785	LIBC_PROBE (memory_sbrk_less, `2`, new_brk, extra);
2786
2787	if (new_brk != (char *) MORECORE_FAILURE)
2788	{
2789	released = (long) (current_brk - new_brk);
2790
2791	if (released != `0`)
2792	{
2793	/ Success. Adjust top. /
2794	av->system_mem -= released;
2795	set_head (av->top, (top_size - released) \| PREV_INUSE);
2796	check_malloc_state (av);
2797	return `1`;
2798	}
2799	}
2800	}
2801	return `0`;
2802	}
2803
2804	static void
2805	internal_function
2806	munmap_chunk (mchunkptr p)
2807	{
2808	INTERNAL_SIZE_T size = chunksize (p);
2809
2810	assert (chunk_is_mmapped (p));
2811
2812	/ Do nothing if the chunk is a faked mmapped chunk in the dumped*
2813	main arena. We never free this memory. /*
2814	if (DUMPED_MAIN_ARENA_CHUNK (p))
2815	return;
2816
2817	uintptr_t block = (uintptr_t) p - prev_size (p);
2818	size_t total_size = prev_size (p) + size;
2819	/ Unfortunately we have to do the compilers job by hand here. Normally*
2820	we would test BLOCK and TOTAL-SIZE separately for compliance with the
2821	page size. But gcc does not recognize the optimization possibility
2822	(in the moment at least) so we combine the two values into one before
2823	the bit test. /*
2824	if (__builtin_expect (((block \| total_size) & (GLRO (dl_pagesize) - `1`)) != `0`, `0`))
2825	{
2826	malloc_printerr (check_action, "munmap_chunk(): invalid pointer",
2827	chunk2mem (p), NULL);
2828	return;
2829	}
2830
2831	atomic_decrement (&mp_.n_mmaps);
2832	atomic_add (&mp_.mmapped_mem, -total_size);
2833
2834	/ If munmap failed the process virtual memory address space is in a*
2835	bad shape. Just leave the block hanging around, the process will
2836	terminate shortly anyway since not much can be done. /*
2837	__munmap ((char *) block, total_size);
2838	}
2839
2840	#if HAVE_MREMAP
2841
2842	static mchunkptr
2843	internal_function
2844	mremap_chunk (mchunkptr p, size_t new_size)
2845	{
2846	size_t pagesize = GLRO (dl_pagesize);
2847	INTERNAL_SIZE_T offset = prev_size (p);
2848	INTERNAL_SIZE_T size = chunksize (p);
2849	char *cp;
2850
2851	assert (chunk_is_mmapped (p));
2852	assert (((size + offset) & (GLRO (dl_pagesize) - `1`)) == `0`);
2853
2854	/ Note the extra SIZE_SZ overhead as in mmap_chunk(). /
2855	new_size = ALIGN_UP (new_size + offset + SIZE_SZ, pagesize);
2856
2857	/ No need to remap if the number of pages does not change. /
2858	if (size + offset == new_size)
2859	return p;
2860
2861	cp = (char ) __mremap ((char* *) p - offset, size + offset, new_size,
2862	MREMAP_MAYMOVE);
2863
2864	if (cp == MAP_FAILED)
2865	return `0`;
2866
2867	p = (mchunkptr) (cp + offset);
2868
2869	assert (aligned_OK (chunk2mem (p)));
2870
2871	assert (prev_size (p) == offset);
2872	set_head (p, (new_size - offset) \| IS_MMAPPED);
2873
2874	INTERNAL_SIZE_T new;
2875	new = atomic_exchange_and_add (&mp_.mmapped_mem, new_size - size - offset)
2876	+ new_size - size - offset;
2877	atomic_max (&mp_.max_mmapped_mem, new);
2878	return p;
2879	}
2880	#endif /* HAVE_MREMAP */
2881
2882	/------------------------ Public wrappers. --------------------------------/
2883
2884	void *
2885	__libc_malloc (size_t bytes)
2886	{
2887	mstate ar_ptr;
2888	void *victim;
2889
2890	void (hook) (size_t, const void *)
2891	= atomic_forced_read (__malloc_hook);
2892	if (__builtin_expect (hook != NULL, `0`))
2893	return (*hook)(bytes, RETURN_ADDRESS (`0`));
2894
2895	arena_get (ar_ptr, bytes);
2896
2897	victim = _int_malloc (ar_ptr, bytes);
2898	/ Retry with another arena only if we were able to find a usable arena*
2899	before. /*
2900	if (!victim && ar_ptr != NULL)
2901	{
2902	LIBC_PROBE (memory_malloc_retry, `1`, bytes);
2903	ar_ptr = arena_get_retry (ar_ptr, bytes);
2904	victim = _int_malloc (ar_ptr, bytes);
2905	}
2906
2907	if (ar_ptr != NULL)
2908	__libc_lock_unlock (ar_ptr->mutex);
2909
2910	assert (!victim \|\| chunk_is_mmapped (mem2chunk (victim)) \|\|
2911	ar_ptr == arena_for_chunk (mem2chunk (victim)));
2912	return victim;
2913	}
2914	libc_hidden_def (__libc_malloc)
2915
2916	void
2917	__libc_free (void *mem)
2918	{
2919	mstate ar_ptr;
2920	mchunkptr p; / chunk corresponding to mem /
2921
2922	void (hook) (void* , const* void *)
2923	= atomic_forced_read (__free_hook);
2924	if (__builtin_expect (hook != NULL, `0`))
2925	{
2926	(*hook)(mem, RETURN_ADDRESS (`0`));
2927	return;
2928	}
2929
2930	if (mem == `0`) / free(0) has no effect /
2931	return;
2932
2933	p = mem2chunk (mem);
2934
2935	if (chunk_is_mmapped (p)) / release mmapped memory. /
2936	{
2937	/ See if the dynamic brk/mmap threshold needs adjusting.*
2938	Dumped fake mmapped chunks do not affect the threshold. /*
2939	if (!mp_.no_dyn_threshold
2940	&& chunksize_nomask (p) > mp_.mmap_threshold
2941	&& chunksize_nomask (p) <= DEFAULT_MMAP_THRESHOLD_MAX
2942	&& !DUMPED_MAIN_ARENA_CHUNK (p))
2943	{
2944	mp_.mmap_threshold = chunksize (p);
2945	mp_.trim_threshold = `2` * mp_.mmap_threshold;
2946	LIBC_PROBE (memory_mallopt_free_dyn_thresholds, `2`,
2947	mp_.mmap_threshold, mp_.trim_threshold);
2948	}
2949	munmap_chunk (p);
2950	return;
2951	}
2952
2953	ar_ptr = arena_for_chunk (p);
2954	_int_free (ar_ptr, p, `0`);
2955	}
2956	libc_hidden_def (__libc_free)
2957
2958	void *
2959	__libc_realloc (void *oldmem, size_t bytes)
2960	{
2961	mstate ar_ptr;
2962	INTERNAL_SIZE_T nb; / padded request size /
2963
2964	void newp; /* chunk to return /
2965
2966	void (hook) (void , size_t, const* void *) =
2967	atomic_forced_read (__realloc_hook);
2968	if (__builtin_expect (hook != NULL, `0`))
2969	return (*hook)(oldmem, bytes, RETURN_ADDRESS (`0`));
2970
2971	#if REALLOC_ZERO_BYTES_FREES
2972	if (bytes == `0` && oldmem != NULL)
2973	{
2974	__libc_free (oldmem); return `0`;
2975	}
2976	#endif
2977
2978	/ realloc of null is supposed to be same as malloc /
2979	if (oldmem == `0`)
2980	return __libc_malloc (bytes);
2981
2982	/ chunk corresponding to oldmem /
2983	const mchunkptr oldp = mem2chunk (oldmem);
2984	/ its size /
2985	const INTERNAL_SIZE_T oldsize = chunksize (oldp);
2986
2987	if (chunk_is_mmapped (oldp))
2988	ar_ptr = NULL;
2989	else
2990	ar_ptr = arena_for_chunk (oldp);
2991
2992	/ Little security check which won't hurt performance: the allocator*
2993	never wrapps around at the end of the address space. Therefore
2994	we can exclude some size values which might appear here by
2995	accident or by "design" from some intruder. We need to bypass
2996	this check for dumped fake mmap chunks from the old main arena
2997	because the new malloc may provide additional alignment. /*
2998	if ((__builtin_expect ((uintptr_t) oldp > (uintptr_t) -oldsize, `0`)
2999	\|\| __builtin_expect (misaligned_chunk (oldp), `0`))
3000	&& !DUMPED_MAIN_ARENA_CHUNK (oldp))
3001	{
3002	malloc_printerr (check_action, "realloc(): invalid pointer", oldmem,
3003	ar_ptr);
3004	return NULL;
3005	}
3006
3007	checked_request2size (bytes, nb);
3008
3009	if (chunk_is_mmapped (oldp))
3010	{
3011	/ If this is a faked mmapped chunk from the dumped main arena,*
3012	always make a copy (and do not free the old chunk). /*
3013	if (DUMPED_MAIN_ARENA_CHUNK (oldp))
3014	{
3015	/ Must alloc, copy, free. /
3016	void *newmem = __libc_malloc (bytes);
3017	if (newmem == `0`)
3018	return NULL;
3019	/ Copy as many bytes as are available from the old chunk*
3020	and fit into the new size. NB: The overhead for faked
3021	mmapped chunks is only SIZE_SZ, not 2 SIZE_SZ as for*
3022	regular mmapped chunks. /*
3023	if (bytes > oldsize - SIZE_SZ)
3024	bytes = oldsize - SIZE_SZ;
3025	memcpy (newmem, oldmem, bytes);
3026	return newmem;
3027	}
3028
3029	void *newmem;
3030
3031	#if HAVE_MREMAP
3032	newp = mremap_chunk (oldp, nb);
3033	if (newp)
3034	return chunk2mem (newp);
3035	#endif
3036	/ Note the extra SIZE_SZ overhead. /
3037	if (oldsize - SIZE_SZ >= nb)
3038	return oldmem; / do nothing /
3039
3040	/ Must alloc, copy, free. /
3041	newmem = __libc_malloc (bytes);
3042	if (newmem == `0`)
3043	return `0`; / propagate failure /
3044
3045	memcpy (newmem, oldmem, oldsize - `2` * SIZE_SZ);
3046	munmap_chunk (oldp);
3047	return newmem;
3048	}
3049
3050	__libc_lock_lock (ar_ptr->mutex);
3051
3052	newp = _int_realloc (ar_ptr, oldp, oldsize, nb);
3053
3054	__libc_lock_unlock (ar_ptr->mutex);
3055	assert (!newp \|\| chunk_is_mmapped (mem2chunk (newp)) \|\|
3056	ar_ptr == arena_for_chunk (mem2chunk (newp)));
3057
3058	if (newp == NULL)
3059	{
3060	/ Try harder to allocate memory in other arenas. /
3061	LIBC_PROBE (memory_realloc_retry, `2`, bytes, oldmem);
3062	newp = __libc_malloc (bytes);
3063	if (newp != NULL)
3064	{
3065	memcpy (newp, oldmem, oldsize - SIZE_SZ);
3066	_int_free (ar_ptr, oldp, `0`);
3067	}
3068	}
3069
3070	return newp;
3071	}
3072	libc_hidden_def (__libc_realloc)
3073
3074	void *
3075	__libc_memalign (size_t alignment, size_t bytes)
3076	{
3077	void *address = RETURN_ADDRESS (`0`);
3078	return _mid_memalign (alignment, bytes, address);
3079	}
3080
3081	static void *
3082	_mid_memalign (size_t alignment, size_t bytes, void *address)
3083	{
3084	mstate ar_ptr;
3085	void *p;
3086
3087	void (hook) (size_t, size_t, const void *) =
3088	atomic_forced_read (__memalign_hook);
3089	if (__builtin_expect (hook != NULL, `0`))
3090	return (*hook)(alignment, bytes, address);
3091
3092	/ If we need less alignment than we give anyway, just relay to malloc. /
3093	if (alignment <= MALLOC_ALIGNMENT)
3094	return __libc_malloc (bytes);
3095
3096	/ Otherwise, ensure that it is at least a minimum chunk size /
3097	if (alignment < MINSIZE)
3098	alignment = MINSIZE;
3099
3100	/ If the alignment is greater than SIZE_MAX / 2 + 1 it cannot be a*
3101	power of 2 and will cause overflow in the check below. /*
3102	if (alignment > SIZE_MAX / `2` + `1`)
3103	{
3104	__set_errno (EINVAL);
3105	return `0`;
3106	}
3107
3108	/ Check for overflow. /
3109	if (bytes > SIZE_MAX - alignment - MINSIZE)
3110	{
3111	__set_errno (ENOMEM);
3112	return `0`;
3113	}
3114
3115
3116	/ Make sure alignment is power of 2. /
3117	if (!powerof2 (alignment))
3118	{
3119	size_t a = MALLOC_ALIGNMENT * `2`;
3120	while (a < alignment)
3121	a <<= `1`;
3122	alignment = a;
3123	}
3124
3125	arena_get (ar_ptr, bytes + alignment + MINSIZE);
3126
3127	p = _int_memalign (ar_ptr, alignment, bytes);
3128	if (!p && ar_ptr != NULL)
3129	{
3130	LIBC_PROBE (memory_memalign_retry, `2`, bytes, alignment);
3131	ar_ptr = arena_get_retry (ar_ptr, bytes);
3132	p = _int_memalign (ar_ptr, alignment, bytes);
3133	}
3134
3135	if (ar_ptr != NULL)
3136	__libc_lock_unlock (ar_ptr->mutex);
3137
3138	assert (!p \|\| chunk_is_mmapped (mem2chunk (p)) \|\|
3139	ar_ptr == arena_for_chunk (mem2chunk (p)));
3140	return p;
3141	}
3142	/ For ISO C11. /
3143	weak_alias (__libc_memalign, aligned_alloc)
3144	libc_hidden_def (__libc_memalign)
3145
3146	void *
3147	__libc_valloc (size_t bytes)
3148	{
3149	if (__malloc_initialized < `0`)
3150	ptmalloc_init ();
3151
3152	void *address = RETURN_ADDRESS (`0`);
3153	size_t pagesize = GLRO (dl_pagesize);
3154	return _mid_memalign (pagesize, bytes, address);
3155	}
3156
3157	void *
3158	__libc_pvalloc (size_t bytes)
3159	{
3160	if (__malloc_initialized < `0`)
3161	ptmalloc_init ();
3162
3163	void *address = RETURN_ADDRESS (`0`);
3164	size_t pagesize = GLRO (dl_pagesize);
3165	size_t rounded_bytes = ALIGN_UP (bytes, pagesize);
3166
3167	/ Check for overflow. /
3168	if (bytes > SIZE_MAX - `2` * pagesize - MINSIZE)
3169	{
3170	__set_errno (ENOMEM);
3171	return `0`;
3172	}
3173
3174	return _mid_memalign (pagesize, rounded_bytes, address);
3175	}
3176
3177	void *
3178	__libc_calloc (size_t n, size_t elem_size)
3179	{
3180	mstate av;
3181	mchunkptr oldtop, p;
3182	INTERNAL_SIZE_T bytes, sz, csz, oldtopsize;
3183	void *mem;
3184	unsigned long clearsize;
3185	unsigned long nclears;
3186	INTERNAL_SIZE_T *d;
3187
3188	/ size_t is unsigned so the behavior on overflow is defined. /
3189	bytes = n * elem_size;
3190	#define HALF_INTERNAL_SIZE_T \
3191	(((INTERNAL_SIZE_T) 1) << (8 * sizeof (INTERNAL_SIZE_T) / 2))
3192	if (__builtin_expect ((n \| elem_size) >= HALF_INTERNAL_SIZE_T, `0`))
3193	{
3194	if (elem_size != `0` && bytes / elem_size != n)
3195	{
3196	__set_errno (ENOMEM);
3197	return `0`;
3198	}
3199	}
3200
3201	void (hook) (size_t, const void *) =
3202	atomic_forced_read (__malloc_hook);
3203	if (__builtin_expect (hook != NULL, `0`))
3204	{
3205	sz = bytes;
3206	mem = (*hook)(sz, RETURN_ADDRESS (`0`));
3207	if (mem == `0`)
3208	return `0`;
3209
3210	return memset (mem, `0`, sz);
3211	}
3212
3213	sz = bytes;
3214
3215	arena_get (av, sz);
3216	if (av)
3217	{
3218	/ Check if we hand out the top chunk, in which case there may be no*
3219	need to clear. /*
3220	#if MORECORE_CLEARS
3221	oldtop = top (av);
3222	oldtopsize = chunksize (top (av));
3223	# if MORECORE_CLEARS < 2
3224	/ Only newly allocated memory is guaranteed to be cleared. /
3225	if (av == &main_arena &&
3226	oldtopsize < mp_.sbrk_base + av->max_system_mem - (char *) oldtop)
3227	oldtopsize = (mp_.sbrk_base + av->max_system_mem - (char *) oldtop);
3228	# endif
3229	if (av != &main_arena)
3230	{
3231	heap_info *heap = heap_for_ptr (oldtop);
3232	if (oldtopsize < (char ) heap + heap->mprotect_size - (char* *) oldtop)
3233	oldtopsize = (char ) heap + heap->mprotect_size - (char* *) oldtop;
3234	}
3235	#endif
3236	}
3237	else
3238	{
3239	/ No usable arenas. /
3240	oldtop = `0`;
3241	oldtopsize = `0`;
3242	}
3243	mem = _int_malloc (av, sz);
3244
3245
3246	assert (!mem \|\| chunk_is_mmapped (mem2chunk (mem)) \|\|
3247	av == arena_for_chunk (mem2chunk (mem)));
3248
3249	if (mem == `0` && av != NULL)
3250	{
3251	LIBC_PROBE (memory_calloc_retry, `1`, sz);
3252	av = arena_get_retry (av, sz);
3253	mem = _int_malloc (av, sz);
3254	}
3255
3256	if (av != NULL)
3257	__libc_lock_unlock (av->mutex);
3258
3259	/ Allocation failed even after a retry. /
3260	if (mem == `0`)
3261	return `0`;
3262
3263	p = mem2chunk (mem);
3264
3265	/ Two optional cases in which clearing not necessary /
3266	if (chunk_is_mmapped (p))
3267	{
3268	if (__builtin_expect (perturb_byte, `0`))
3269	return memset (mem, `0`, sz);
3270
3271	return mem;
3272	}
3273
3274	csz = chunksize (p);
3275
3276	#if MORECORE_CLEARS
3277	if (perturb_byte == `0` && (p == oldtop && csz > oldtopsize))
3278	{
3279	/ clear only the bytes from non-freshly-sbrked memory /
3280	csz = oldtopsize;
3281	}
3282	#endif
3283
3284	/ Unroll clear of <= 36 bytes (72 if 8byte sizes). We know that*
3285	contents have an odd number of INTERNAL_SIZE_T-sized words;
3286	minimally 3. /*
3287	d = (INTERNAL_SIZE_T *) mem;
3288	clearsize = csz - SIZE_SZ;
3289	nclears = clearsize / sizeof (INTERNAL_SIZE_T);
3290	assert (nclears >= `3`);
3291
3292	if (nclears > `9`)
3293	return memset (d, `0`, clearsize);
3294
3295	else
3296	{
3297	*(d + `0`) = `0`;
3298	*(d + `1`) = `0`;
3299	*(d + `2`) = `0`;
3300	if (nclears > `4`)
3301	{
3302	*(d + `3`) = `0`;
3303	*(d + `4`) = `0`;
3304	if (nclears > `6`)
3305	{
3306	*(d + `5`) = `0`;
3307	*(d + `6`) = `0`;
3308	if (nclears > `8`)
3309	{
3310	*(d + `7`) = `0`;
3311	*(d + `8`) = `0`;
3312	}
3313	}
3314	}
3315	}
3316
3317	return mem;
3318	}
3319
3320	/*
3321	------------------------------ malloc ------------------------------
3322	*/
3323
3324	static void *
3325	_int_malloc (mstate av, size_t bytes)
3326	{
3327	INTERNAL_SIZE_T nb; / normalized request size /
3328	unsigned int idx; / associated bin index /
3329	mbinptr bin; / associated bin /
3330
3331	mchunkptr victim; / inspected/selected chunk /
3332	INTERNAL_SIZE_T size; / its size /
3333	int victim_index; / its bin index /
3334
3335	mchunkptr remainder; / remainder from a split /
3336	unsigned long remainder_size; / its size /
3337
3338	unsigned int block; / bit map traverser /
3339	unsigned int bit; / bit map traverser /
3340	unsigned int map; / current word of binmap /
3341
3342	mchunkptr fwd; / misc temp for linking /
3343	mchunkptr bck; / misc temp for linking /
3344
3345	const char *errstr = NULL;
3346
3347	/*
3348	Convert request size to internal form by adding SIZE_SZ bytes
3349	overhead plus possibly more to obtain necessary alignment and/or
3350	to obtain a size of at least MINSIZE, the smallest allocatable
3351	size. Also, checked_request2size traps (returning 0) request sizes
3352	that are so large that they wrap around zero when padded and
3353	aligned.
3354	*/
3355
3356	checked_request2size (bytes, nb);
3357
3358	/ There are no usable arenas. Fall back to sysmalloc to get a chunk from*
3359	mmap. /*
3360	if (__glibc_unlikely (av == NULL))
3361	{
3362	void *p = sysmalloc (nb, av);
3363	if (p != NULL)
3364	alloc_perturb (p, bytes);
3365	return p;
3366	}
3367
3368	/*
3369	If the size qualifies as a fastbin, first check corresponding bin.
3370	This code is safe to execute even if av is not yet initialized, so we
3371	can try it without checking, which saves some time on this fast path.
3372	*/
3373
3374	if ((unsigned long) (nb) <= (unsigned long) (get_max_fast ()))
3375	{
3376	idx = fastbin_index (nb);
3377	mfastbinptr *fb = &fastbin (av, idx);
3378	mchunkptr pp = *fb;
3379	do
3380	{
3381	victim = pp;
3382	if (victim == NULL)
3383	break;
3384	}
3385	while ((pp = catomic_compare_and_exchange_val_acq (fb, victim->fd, victim))
3386	!= victim);
3387	if (victim != `0`)
3388	{
3389	if (__builtin_expect (fastbin_index (chunksize (victim)) != idx, `0`))
3390	{
3391	errstr = "malloc(): memory corruption (fast)";
3392	errout:
3393	malloc_printerr (check_action, errstr, chunk2mem (victim), av);
3394	return NULL;
3395	}
3396	check_remalloced_chunk (av, victim, nb);
3397	void *p = chunk2mem (victim);
3398	alloc_perturb (p, bytes);
3399	return p;
3400	}
3401	}
3402
3403	/*
3404	If a small request, check regular bin. Since these "smallbins"
3405	hold one size each, no searching within bins is necessary.
3406	(For a large request, we need to wait until unsorted chunks are
3407	processed to find best fit. But for small ones, fits are exact
3408	anyway, so we can check now, which is faster.)
3409	*/
3410
3411	if (in_smallbin_range (nb))
3412	{
3413	idx = smallbin_index (nb);
3414	bin = bin_at (av, idx);
3415
3416	if ((victim = last (bin)) != bin)
3417	{
3418	if (victim == `0`) / initialization check /
3419	malloc_consolidate (av);
3420	else
3421	{
3422	bck = victim->bk;
3423	if (__glibc_unlikely (bck->fd != victim))
3424	{
3425	errstr = "malloc(): smallbin double linked list corrupted";
3426	goto errout;
3427	}
3428	set_inuse_bit_at_offset (victim, nb);
3429	bin->bk = bck;
3430	bck->fd = bin;
3431
3432	if (av != &main_arena)
3433	set_non_main_arena (victim);
3434	check_malloced_chunk (av, victim, nb);
3435	void *p = chunk2mem (victim);
3436	alloc_perturb (p, bytes);
3437	return p;
3438	}
3439	}
3440	}
3441
3442	/*
3443	If this is a large request, consolidate fastbins before continuing.
3444	While it might look excessive to kill all fastbins before
3445	even seeing if there is space available, this avoids
3446	fragmentation problems normally associated with fastbins.
3447	Also, in practice, programs tend to have runs of either small or
3448	large requests, but less often mixtures, so consolidation is not
3449	invoked all that often in most programs. And the programs that
3450	it is called frequently in otherwise tend to fragment.
3451	*/
3452
3453	else
3454	{
3455	idx = largebin_index (nb);
3456	if (have_fastchunks (av))
3457	malloc_consolidate (av);
3458	}
3459
3460	/*
3461	Process recently freed or remaindered chunks, taking one only if
3462	it is exact fit, or, if this a small request, the chunk is remainder from
3463	the most recent non-exact fit. Place other traversed chunks in
3464	bins. Note that this step is the only place in any routine where
3465	chunks are placed in bins.
3466
3467	The outer loop here is needed because we might not realize until
3468	near the end of malloc that we should have consolidated, so must
3469	do so and retry. This happens at most once, and only when we would
3470	otherwise need to expand memory to service a "small" request.
3471	*/
3472
3473	for (;; )
3474	{
3475	int iters = `0`;
3476	while ((victim = unsorted_chunks (av)->bk) != unsorted_chunks (av))
3477	{
3478	bck = victim->bk;
3479	if (__builtin_expect (chunksize_nomask (victim) <= `2` * SIZE_SZ, `0`)
3480	\|\| __builtin_expect (chunksize_nomask (victim)
3481	> av->system_mem, `0`))
3482	malloc_printerr (check_action, "malloc(): memory corruption",
3483	chunk2mem (victim), av);
3484	size = chunksize (victim);
3485
3486	/*
3487	If a small request, try to use last remainder if it is the
3488	only chunk in unsorted bin. This helps promote locality for
3489	runs of consecutive small requests. This is the only
3490	exception to best-fit, and applies only when there is
3491	no exact fit for a small chunk.
3492	*/
3493
3494	if (in_smallbin_range (nb) &&
3495	bck == unsorted_chunks (av) &&
3496	victim == av->last_remainder &&
3497	(unsigned long) (size) > (unsigned long) (nb + MINSIZE))
3498	{
3499	/ split and reattach remainder /
3500	remainder_size = size - nb;
3501	remainder = chunk_at_offset (victim, nb);
3502	unsorted_chunks (av)->bk = unsorted_chunks (av)->fd = remainder;
3503	av->last_remainder = remainder;
3504	remainder->bk = remainder->fd = unsorted_chunks (av);
3505	if (!in_smallbin_range (remainder_size))
3506	{
3507	remainder->fd_nextsize = NULL;
3508	remainder->bk_nextsize = NULL;
3509	}
3510
3511	set_head (victim, nb \| PREV_INUSE \|
3512	(av != &main_arena ? NON_MAIN_ARENA : `0`));
3513	set_head (remainder, remainder_size \| PREV_INUSE);
3514	set_foot (remainder, remainder_size);
3515
3516	check_malloced_chunk (av, victim, nb);
3517	void *p = chunk2mem (victim);
3518	alloc_perturb (p, bytes);
3519	return p;
3520	}
3521
3522	/ remove from unsorted list /
3523	unsorted_chunks (av)->bk = bck;
3524	bck->fd = unsorted_chunks (av);
3525
3526	/ Take now instead of binning if exact fit /
3527
3528	if (size == nb)
3529	{
3530	set_inuse_bit_at_offset (victim, size);
3531	if (av != &main_arena)
3532	set_non_main_arena (victim);
3533	check_malloced_chunk (av, victim, nb);
3534	void *p = chunk2mem (victim);
3535	alloc_perturb (p, bytes);
3536	return p;
3537	}
3538
3539	/ place chunk in bin /
3540
3541	if (in_smallbin_range (size))
3542	{
3543	victim_index = smallbin_index (size);
3544	bck = bin_at (av, victim_index);
3545	fwd = bck->fd;
3546	}
3547	else
3548	{
3549	victim_index = largebin_index (size);
3550	bck = bin_at (av, victim_index);
3551	fwd = bck->fd;
3552
3553	/ maintain large bins in sorted order /
3554	if (fwd != bck)
3555	{
3556	/ Or with inuse bit to speed comparisons /
3557	size \|= PREV_INUSE;
3558	/ if smaller than smallest, bypass loop below /
3559	assert (chunk_main_arena (bck->bk));
3560	if ((unsigned long) (size)
3561	< (unsigned long) chunksize_nomask (bck->bk))
3562	{
3563	fwd = bck;
3564	bck = bck->bk;
3565
3566	victim->fd_nextsize = fwd->fd;
3567	victim->bk_nextsize = fwd->fd->bk_nextsize;
3568	fwd->fd->bk_nextsize = victim->bk_nextsize->fd_nextsize = victim;
3569	}
3570	else
3571	{
3572	assert (chunk_main_arena (fwd));
3573	while ((unsigned long) size < chunksize_nomask (fwd))
3574	{
3575	fwd = fwd->fd_nextsize;
3576	assert (chunk_main_arena (fwd));
3577	}
3578
3579	if ((unsigned long) size
3580	== (unsigned long) chunksize_nomask (fwd))
3581	/ Always insert in the second position. /
3582	fwd = fwd->fd;
3583	else
3584	{
3585	victim->fd_nextsize = fwd;
3586	victim->bk_nextsize = fwd->bk_nextsize;
3587	fwd->bk_nextsize = victim;
3588	victim->bk_nextsize->fd_nextsize = victim;
3589	}
3590	bck = fwd->bk;
3591	}
3592	}
3593	else
3594	victim->fd_nextsize = victim->bk_nextsize = victim;
3595	}
3596
3597	mark_bin (av, victim_index);
3598	victim->bk = bck;
3599	victim->fd = fwd;
3600	fwd->bk = victim;
3601	bck->fd = victim;
3602
3603	#define MAX_ITERS 10000
3604	if (++iters >= MAX_ITERS)
3605	break;
3606	}
3607
3608	/*
3609	If a large request, scan through the chunks of current bin in
3610	sorted order to find smallest that fits. Use the skip list for this.
3611	*/
3612
3613	if (!in_smallbin_range (nb))
3614	{
3615	bin = bin_at (av, idx);
3616
3617	/ skip scan if empty or largest chunk is too small /
3618	if ((victim = first (bin)) != bin
3619	&& (unsigned long) chunksize_nomask (victim)
3620	>= (unsigned long) (nb))
3621	{
3622	victim = victim->bk_nextsize;
3623	while (((unsigned long) (size = chunksize (victim)) <
3624	(unsigned long) (nb)))
3625	victim = victim->bk_nextsize;
3626
3627	/ Avoid removing the first entry for a size so that the skip*
3628	list does not have to be rerouted. /*
3629	if (victim != last (bin)
3630	&& chunksize_nomask (victim)
3631	== chunksize_nomask (victim->fd))
3632	victim = victim->fd;
3633
3634	remainder_size = size - nb;
3635	unlink (av, victim, bck, fwd);
3636
3637	/ Exhaust /
3638	if (remainder_size < MINSIZE)
3639	{
3640	set_inuse_bit_at_offset (victim, size);
3641	if (av != &main_arena)
3642	set_non_main_arena (victim);
3643	}
3644	/ Split /
3645	else
3646	{
3647	remainder = chunk_at_offset (victim, nb);
3648	/ We cannot assume the unsorted list is empty and therefore*
3649	have to perform a complete insert here. /*
3650	bck = unsorted_chunks (av);
3651	fwd = bck->fd;
3652	if (__glibc_unlikely (fwd->bk != bck))
3653	{
3654	errstr = "malloc(): corrupted unsorted chunks";
3655	goto errout;
3656	}
3657	remainder->bk = bck;
3658	remainder->fd = fwd;
3659	bck->fd = remainder;
3660	fwd->bk = remainder;
3661	if (!in_smallbin_range (remainder_size))
3662	{
3663	remainder->fd_nextsize = NULL;
3664	remainder->bk_nextsize = NULL;
3665	}
3666	set_head (victim, nb \| PREV_INUSE \|
3667	(av != &main_arena ? NON_MAIN_ARENA : `0`));
3668	set_head (remainder, remainder_size \| PREV_INUSE);
3669	set_foot (remainder, remainder_size);
3670	}
3671	check_malloced_chunk (av, victim, nb);
3672	void *p = chunk2mem (victim);
3673	alloc_perturb (p, bytes);
3674	return p;
3675	}
3676	}
3677
3678	/*
3679	Search for a chunk by scanning bins, starting with next largest
3680	bin. This search is strictly by best-fit; i.e., the smallest
3681	(with ties going to approximately the least recently used) chunk
3682	that fits is selected.
3683
3684	The bitmap avoids needing to check that most blocks are nonempty.
3685	The particular case of skipping all bins during warm-up phases
3686	when no chunks have been returned yet is faster than it might look.
3687	*/
3688
3689	++idx;
3690	bin = bin_at (av, idx);
3691	block = idx2block (idx);
3692	map = av->binmap[block];
3693	bit = idx2bit (idx);
3694
3695	for (;; )
3696	{
3697	/ Skip rest of block if there are no more set bits in this block. /
3698	if (bit > map \|\| bit == `0`)
3699	{
3700	do
3701	{
3702	if (++block >= BINMAPSIZE) / out of bins /
3703	goto use_top;
3704	}
3705	while ((map = av->binmap[block]) == `0`);
3706
3707	bin = bin_at (av, (block << BINMAPSHIFT));
3708	bit = `1`;
3709	}
3710
3711	/ Advance to bin with set bit. There must be one. /
3712	while ((bit & map) == `0`)
3713	{
3714	bin = next_bin (bin);
3715	bit <<= `1`;
3716	assert (bit != `0`);
3717	}
3718
3719	/ Inspect the bin. It is likely to be non-empty /
3720	victim = last (bin);
3721
3722	/ If a false alarm (empty bin), clear the bit. /
3723	if (victim == bin)
3724	{
3725	av->binmap[block] = map &= ~bit; / Write through /
3726	bin = next_bin (bin);
3727	bit <<= `1`;
3728	}
3729
3730	else
3731	{
3732	size = chunksize (victim);
3733
3734	/ We know the first chunk in this bin is big enough to use. /
3735	assert ((unsigned long) (size) >= (unsigned long) (nb));
3736
3737	remainder_size = size - nb;
3738
3739	/ unlink /
3740	unlink (av, victim, bck, fwd);
3741
3742	/ Exhaust /
3743	if (remainder_size < MINSIZE)
3744	{
3745	set_inuse_bit_at_offset (victim, size);
3746	if (av != &main_arena)
3747	set_non_main_arena (victim);
3748	}
3749
3750	/ Split /
3751	else
3752	{
3753	remainder = chunk_at_offset (victim, nb);
3754
3755	/ We cannot assume the unsorted list is empty and therefore*
3756	have to perform a complete insert here. /*
3757	bck = unsorted_chunks (av);
3758	fwd = bck->fd;
3759	if (__glibc_unlikely (fwd->bk != bck))
3760	{
3761	errstr = "malloc(): corrupted unsorted chunks 2";
3762	goto errout;
3763	}
3764	remainder->bk = bck;
3765	remainder->fd = fwd;
3766	bck->fd = remainder;
3767	fwd->bk = remainder;
3768
3769	/ advertise as last remainder /
3770	if (in_smallbin_range (nb))
3771	av->last_remainder = remainder;
3772	if (!in_smallbin_range (remainder_size))
3773	{
3774	remainder->fd_nextsize = NULL;
3775	remainder->bk_nextsize = NULL;
3776	}
3777	set_head (victim, nb \| PREV_INUSE \|
3778	(av != &main_arena ? NON_MAIN_ARENA : `0`));
3779	set_head (remainder, remainder_size \| PREV_INUSE);
3780	set_foot (remainder, remainder_size);
3781	}
3782	check_malloced_chunk (av, victim, nb);
3783	void *p = chunk2mem (victim);
3784	alloc_perturb (p, bytes);
3785	return p;
3786	}
3787	}
3788
3789	use_top:
3790	/*
3791	If large enough, split off the chunk bordering the end of memory
3792	(held in av->top). Note that this is in accord with the best-fit
3793	search rule. In effect, av->top is treated as larger (and thus
3794	less well fitting) than any other available chunk since it can
3795	be extended to be as large as necessary (up to system
3796	limitations).
3797
3798	We require that av->top always exists (i.e., has size >=
3799	MINSIZE) after initialization, so if it would otherwise be
3800	exhausted by current request, it is replenished. (The main
3801	reason for ensuring it exists is that we may need MINSIZE space
3802	to put in fenceposts in sysmalloc.)
3803	*/
3804
3805	victim = av->top;
3806	size = chunksize (victim);
3807
3808	if ((unsigned long) (size) >= (unsigned long) (nb + MINSIZE))
3809	{
3810	remainder_size = size - nb;
3811	remainder = chunk_at_offset (victim, nb);
3812	av->top = remainder;
3813	set_head (victim, nb \| PREV_INUSE \|
3814	(av != &main_arena ? NON_MAIN_ARENA : `0`));
3815	set_head (remainder, remainder_size \| PREV_INUSE);
3816
3817	check_malloced_chunk (av, victim, nb);
3818	void *p = chunk2mem (victim);
3819	alloc_perturb (p, bytes);
3820	return p;
3821	}
3822
3823	/ When we are using atomic ops to free fast chunks we can get*
3824	here for all block sizes. /*
3825	else if (have_fastchunks (av))
3826	{
3827	malloc_consolidate (av);
3828	/ restore original bin index /
3829	if (in_smallbin_range (nb))
3830	idx = smallbin_index (nb);
3831	else
3832	idx = largebin_index (nb);
3833	}
3834
3835	/*
3836	Otherwise, relay to handle system-dependent cases
3837	*/
3838	else
3839	{
3840	void *p = sysmalloc (nb, av);
3841	if (p != NULL)
3842	alloc_perturb (p, bytes);
3843	return p;
3844	}
3845	}
3846	}
3847
3848	/*
3849	------------------------------ free ------------------------------
3850	*/
3851
3852	static void
3853	_int_free (mstate av, mchunkptr p, int have_lock)
3854	{
3855	INTERNAL_SIZE_T size; / its size /
3856	mfastbinptr fb; /* associated fastbin /
3857	mchunkptr nextchunk; / next contiguous chunk /
3858	INTERNAL_SIZE_T nextsize; / its size /
3859	int nextinuse; / true if nextchunk is used /
3860	INTERNAL_SIZE_T prevsize; / size of previous contiguous chunk /
3861	mchunkptr bck; / misc temp for linking /
3862	mchunkptr fwd; / misc temp for linking /
3863
3864	const char *errstr = NULL;
3865	int locked = `0`;
3866
3867	size = chunksize (p);
3868
3869	/ Little security check which won't hurt performance: the*
3870	allocator never wrapps around at the end of the address space.
3871	Therefore we can exclude some size values which might appear
3872	here by accident or by "design" from some intruder. /*
3873	if (__builtin_expect ((uintptr_t) p > (uintptr_t) -size, `0`)
3874	\|\| __builtin_expect (misaligned_chunk (p), `0`))
3875	{
3876	errstr = "free(): invalid pointer";
3877	errout:
3878	if (!have_lock && locked)
3879	__libc_lock_unlock (av->mutex);
3880	malloc_printerr (check_action, errstr, chunk2mem (p), av);
3881	return;
3882	}
3883	/ We know that each chunk is at least MINSIZE bytes in size or a*
3884	multiple of MALLOC_ALIGNMENT. /*
3885	if (__glibc_unlikely (size < MINSIZE \|\| !aligned_OK (size)))
3886	{
3887	errstr = "free(): invalid size";
3888	goto errout;
3889	}
3890
3891	check_inuse_chunk(av, p);
3892
3893	/*
3894	If eligible, place chunk on a fastbin so it can be found
3895	and used quickly in malloc.
3896	*/
3897
3898	if ((unsigned long)(size) <= (unsigned long)(get_max_fast ())
3899
3900	#if TRIM_FASTBINS
3901	/*
3902	If TRIM_FASTBINS set, don't place chunks
3903	bordering top into fastbins
3904	*/
3905	&& (chunk_at_offset(p, size) != av->top)
3906	#endif
3907	) {
3908
3909	if (__builtin_expect (chunksize_nomask (chunk_at_offset (p, size))
3910	<= `2` * SIZE_SZ, `0`)
3911	\|\| __builtin_expect (chunksize (chunk_at_offset (p, size))
3912	>= av->system_mem, `0`))
3913	{
3914	/ We might not have a lock at this point and concurrent modifications*
3915	of system_mem might have let to a false positive. Redo the test
3916	after getting the lock. /*
3917	if (have_lock
3918	\|\| ({ assert (locked == `0`);
3919	__libc_lock_lock (av->mutex);
3920	locked = `1`;
3921	chunksize_nomask (chunk_at_offset (p, size)) <= `2` * SIZE_SZ
3922	\|\| chunksize (chunk_at_offset (p, size)) >= av->system_mem;
3923	}))
3924	{
3925	errstr = "free(): invalid next size (fast)";
3926	goto errout;
3927	}
3928	if (! have_lock)
3929	{
3930	__libc_lock_unlock (av->mutex);
3931	locked = `0`;
3932	}
3933	}
3934
3935	free_perturb (chunk2mem(p), size - `2` * SIZE_SZ);
3936
3937	set_fastchunks(av);
3938	unsigned int idx = fastbin_index(size);
3939	fb = &fastbin (av, idx);
3940
3941	/ Atomically link P to its fastbin: P->FD = FB; FB = P; /
3942	mchunkptr old = *fb, old2;
3943	unsigned int old_idx = ~`0u`;
3944	do
3945	{
3946	/ Check that the top of the bin is not the record we are going to add*
3947	(i.e., double free). /*
3948	if (__builtin_expect (old == p, `0`))
3949	{
3950	errstr = "double free or corruption (fasttop)";
3951	goto errout;
3952	}
3953	/ Check that size of fastbin chunk at the top is the same as*
3954	size of the chunk that we are adding. We can dereference OLD
3955	only if we have the lock, otherwise it might have already been
3956	deallocated. See use of OLD_IDX below for the actual check. /*
3957	if (have_lock && old != NULL)
3958	old_idx = fastbin_index(chunksize(old));
3959	p->fd = old2 = old;
3960	}
3961	while ((old = catomic_compare_and_exchange_val_rel (fb, p, old2)) != old2);
3962
3963	if (have_lock && old != NULL && __builtin_expect (old_idx != idx, `0`))
3964	{
3965	errstr = "invalid fastbin entry (free)";
3966	goto errout;
3967	}
3968	}
3969
3970	/*
3971	Consolidate other non-mmapped chunks as they arrive.
3972	*/
3973
3974	else if (!chunk_is_mmapped(p)) {
3975	if (! have_lock) {
3976	__libc_lock_lock (av->mutex);
3977	locked = `1`;
3978	}
3979
3980	nextchunk = chunk_at_offset(p, size);
3981
3982	/ Lightweight tests: check whether the block is already the*
3983	top block. /*
3984	if (__glibc_unlikely (p == av->top))
3985	{
3986	errstr = "double free or corruption (top)";
3987	goto errout;
3988	}
3989	/ Or whether the next chunk is beyond the boundaries of the arena. /
3990	if (__builtin_expect (contiguous (av)
3991	&& (char *) nextchunk
3992	>= ((char *) av->top + chunksize(av->top)), `0`))
3993	{
3994	errstr = "double free or corruption (out)";
3995	goto errout;
3996	}
3997	/ Or whether the block is actually not marked used. /
3998	if (__glibc_unlikely (!prev_inuse(nextchunk)))
3999	{
4000	errstr = "double free or corruption (!prev)";
4001	goto errout;
4002	}
4003
4004	nextsize = chunksize(nextchunk);
4005	if (__builtin_expect (chunksize_nomask (nextchunk) <= `2` * SIZE_SZ, `0`)
4006	\|\| __builtin_expect (nextsize >= av->system_mem, `0`))
4007	{
4008	errstr = "free(): invalid next size (normal)";
4009	goto errout;
4010	}
4011
4012	free_perturb (chunk2mem(p), size - `2` * SIZE_SZ);
4013
4014	/ consolidate backward /
4015	if (!prev_inuse(p)) {
4016	prevsize = prev_size (p);
4017	size += prevsize;
4018	p = chunk_at_offset(p, -((long) prevsize));
4019	unlink(av, p, bck, fwd);
4020	}
4021
4022	if (nextchunk != av->top) {
4023	/ get and clear inuse bit /
4024	nextinuse = inuse_bit_at_offset(nextchunk, nextsize);
4025
4026	/ consolidate forward /
4027	if (!nextinuse) {
4028	unlink(av, nextchunk, bck, fwd);
4029	size += nextsize;
4030	} else
4031	clear_inuse_bit_at_offset(nextchunk, `0`);
4032
4033	/*
4034	Place the chunk in unsorted chunk list. Chunks are
4035	not placed into regular bins until after they have
4036	been given one chance to be used in malloc.
4037	*/
4038
4039	bck = unsorted_chunks(av);
4040	fwd = bck->fd;
4041	if (__glibc_unlikely (fwd->bk != bck))
4042	{
4043	errstr = "free(): corrupted unsorted chunks";
4044	goto errout;
4045	}
4046	p->fd = fwd;
4047	p->bk = bck;
4048	if (!in_smallbin_range(size))
4049	{
4050	p->fd_nextsize = NULL;
4051	p->bk_nextsize = NULL;
4052	}
4053	bck->fd = p;
4054	fwd->bk = p;
4055
4056	set_head(p, size \| PREV_INUSE);
4057	set_foot(p, size);
4058
4059	check_free_chunk(av, p);
4060	}
4061
4062	/*
4063	If the chunk borders the current high end of memory,
4064	consolidate into top
4065	*/
4066
4067	else {
4068	size += nextsize;
4069	set_head(p, size \| PREV_INUSE);
4070	av->top = p;
4071	check_chunk(av, p);
4072	}
4073
4074	/*
4075	If freeing a large space, consolidate possibly-surrounding
4076	chunks. Then, if the total unused topmost memory exceeds trim
4077	threshold, ask malloc_trim to reduce top.
4078
4079	Unless max_fast is 0, we don't know if there are fastbins
4080	bordering top, so we cannot tell for sure whether threshold
4081	has been reached unless fastbins are consolidated. But we
4082	don't want to consolidate on each free. As a compromise,
4083	consolidation is performed if FASTBIN_CONSOLIDATION_THRESHOLD
4084	is reached.
4085	*/
4086
4087	if ((unsigned long)(size) >= FASTBIN_CONSOLIDATION_THRESHOLD) {
4088	if (have_fastchunks(av))
4089	malloc_consolidate(av);
4090
4091	if (av == &main_arena) {
4092	#ifndef MORECORE_CANNOT_TRIM
4093	if ((unsigned long)(chunksize(av->top)) >=
4094	(unsigned long)(mp_.trim_threshold))
4095	systrim(mp_.top_pad, av);
4096	#endif
4097	} else {
4098	/ Always try heap_trim(), even if the top chunk is not*
4099	large, because the corresponding heap might go away. /*
4100	heap_info *heap = heap_for_ptr(top(av));
4101
4102	assert(heap->ar_ptr == av);
4103	heap_trim(heap, mp_.top_pad);
4104	}
4105	}
4106
4107	if (! have_lock) {
4108	assert (locked);
4109	__libc_lock_unlock (av->mutex);
4110	}
4111	}
4112	/*
4113	If the chunk was allocated via mmap, release via munmap().
4114	*/
4115
4116	else {
4117	munmap_chunk (p);
4118	}
4119	}
4120
4121	/*
4122	------------------------- malloc_consolidate -------------------------
4123
4124	malloc_consolidate is a specialized version of free() that tears
4125	down chunks held in fastbins. Free itself cannot be used for this
4126	purpose since, among other things, it might place chunks back onto
4127	fastbins. So, instead, we need to use a minor variant of the same
4128	code.
4129
4130	Also, because this routine needs to be called the first time through
4131	malloc anyway, it turns out to be the perfect place to trigger
4132	initialization code.
4133	*/
4134
4135	static void malloc_consolidate(mstate av)
4136	{
4137	mfastbinptr* fb; / current fastbin being consolidated /
4138	mfastbinptr* maxfb; / last fastbin (for loop control) /
4139	mchunkptr p; / current chunk being consolidated /
4140	mchunkptr nextp; / next chunk to consolidate /
4141	mchunkptr unsorted_bin; / bin header /
4142	mchunkptr first_unsorted; / chunk to link to /
4143
4144	/ These have same use as in free() /
4145	mchunkptr nextchunk;
4146	INTERNAL_SIZE_T size;
4147	INTERNAL_SIZE_T nextsize;
4148	INTERNAL_SIZE_T prevsize;
4149	int nextinuse;
4150	mchunkptr bck;
4151	mchunkptr fwd;
4152
4153	/*
4154	If max_fast is 0, we know that av hasn't
4155	yet been initialized, in which case do so below
4156	*/
4157
4158	if (get_max_fast () != `0`) {
4159	clear_fastchunks(av);
4160
4161	unsorted_bin = unsorted_chunks(av);
4162
4163	/*
4164	Remove each chunk from fast bin and consolidate it, placing it
4165	then in unsorted bin. Among other reasons for doing this,
4166	placing in unsorted bin avoids needing to calculate actual bins
4167	until malloc is sure that chunks aren't immediately going to be
4168	reused anyway.
4169	*/
4170
4171	maxfb = &fastbin (av, NFASTBINS - `1`);
4172	fb = &fastbin (av, `0`);
4173	do {
4174	p = atomic_exchange_acq (fb, NULL);
4175	if (p != `0`) {
4176	do {
4177	check_inuse_chunk(av, p);
4178	nextp = p->fd;
4179
4180	/ Slightly streamlined version of consolidation code in free() /
4181	size = chunksize (p);
4182	nextchunk = chunk_at_offset(p, size);
4183	nextsize = chunksize(nextchunk);
4184
4185	if (!prev_inuse(p)) {
4186	prevsize = prev_size (p);
4187	size += prevsize;
4188	p = chunk_at_offset(p, -((long) prevsize));
4189	unlink(av, p, bck, fwd);
4190	}
4191
4192	if (nextchunk != av->top) {
4193	nextinuse = inuse_bit_at_offset(nextchunk, nextsize);
4194
4195	if (!nextinuse) {
4196	size += nextsize;
4197	unlink(av, nextchunk, bck, fwd);
4198	} else
4199	clear_inuse_bit_at_offset(nextchunk, `0`);
4200
4201	first_unsorted = unsorted_bin->fd;
4202	unsorted_bin->fd = p;
4203	first_unsorted->bk = p;
4204
4205	if (!in_smallbin_range (size)) {
4206	p->fd_nextsize = NULL;
4207	p->bk_nextsize = NULL;
4208	}
4209
4210	set_head(p, size \| PREV_INUSE);
4211	p->bk = unsorted_bin;
4212	p->fd = first_unsorted;
4213	set_foot(p, size);
4214	}
4215
4216	else {
4217	size += nextsize;
4218	set_head(p, size \| PREV_INUSE);
4219	av->top = p;
4220	}
4221
4222	} while ( (p = nextp) != `0`);
4223
4224	}
4225	} while (fb++ != maxfb);
4226	}
4227	else {
4228	malloc_init_state(av);
4229	check_malloc_state(av);
4230	}
4231	}
4232
4233	/*
4234	------------------------------ realloc ------------------------------
4235	*/
4236
4237	void*
4238	_int_realloc(mstate av, mchunkptr oldp, INTERNAL_SIZE_T oldsize,
4239	INTERNAL_SIZE_T nb)
4240	{
4241	mchunkptr newp; / chunk to return /
4242	INTERNAL_SIZE_T newsize; / its size /
4243	void* newmem; / corresponding user mem /
4244
4245	mchunkptr next; / next contiguous chunk after oldp /
4246
4247	mchunkptr remainder; / extra space at end of newp /
4248	unsigned long remainder_size; / its size /
4249
4250	mchunkptr bck; / misc temp for linking /
4251	mchunkptr fwd; / misc temp for linking /
4252
4253	const char *errstr = NULL;
4254
4255	/ oldmem size /
4256	if (__builtin_expect (chunksize_nomask (oldp) <= `2` * SIZE_SZ, `0`)
4257	\|\| __builtin_expect (oldsize >= av->system_mem, `0`))
4258	{
4259	errstr = "realloc(): invalid old size";
4260	errout:
4261	malloc_printerr (check_action, errstr, chunk2mem (oldp), av);
4262	return NULL;
4263	}
4264
4265	check_inuse_chunk (av, oldp);
4266
4267	/ All callers already filter out mmap'ed chunks. /
4268	assert (!chunk_is_mmapped (oldp));
4269
4270	next = chunk_at_offset (oldp, oldsize);
4271	INTERNAL_SIZE_T nextsize = chunksize (next);
4272	if (__builtin_expect (chunksize_nomask (next) <= `2` * SIZE_SZ, `0`)
4273	\|\| __builtin_expect (nextsize >= av->system_mem, `0`))
4274	{
4275	errstr = "realloc(): invalid next size";
4276	goto errout;
4277	}
4278
4279	if ((unsigned long) (oldsize) >= (unsigned long) (nb))
4280	{
4281	/ already big enough; split below /
4282	newp = oldp;
4283	newsize = oldsize;
4284	}
4285
4286	else
4287	{
4288	/ Try to expand forward into top /
4289	if (next == av->top &&
4290	(unsigned long) (newsize = oldsize + nextsize) >=
4291	(unsigned long) (nb + MINSIZE))
4292	{
4293	set_head_size (oldp, nb \| (av != &main_arena ? NON_MAIN_ARENA : `0`));
4294	av->top = chunk_at_offset (oldp, nb);
4295	set_head (av->top, (newsize - nb) \| PREV_INUSE);
4296	check_inuse_chunk (av, oldp);
4297	return chunk2mem (oldp);
4298	}
4299
4300	/ Try to expand forward into next chunk; split off remainder below /
4301	else if (next != av->top &&
4302	!inuse (next) &&
4303	(unsigned long) (newsize = oldsize + nextsize) >=
4304	(unsigned long) (nb))
4305	{
4306	newp = oldp;
4307	unlink (av, next, bck, fwd);
4308	}
4309
4310	/ allocate, copy, free /
4311	else
4312	{
4313	newmem = _int_malloc (av, nb - MALLOC_ALIGN_MASK);
4314	if (newmem == `0`)
4315	return `0`; / propagate failure /
4316
4317	newp = mem2chunk (newmem);
4318	newsize = chunksize (newp);
4319
4320	/*
4321	Avoid copy if newp is next chunk after oldp.
4322	*/
4323	if (newp == next)
4324	{
4325	newsize += oldsize;
4326	newp = oldp;
4327	}
4328	else
4329	{
4330	memcpy (newmem, chunk2mem (oldp), oldsize - SIZE_SZ);
4331	_int_free (av, oldp, `1`);
4332	check_inuse_chunk (av, newp);
4333	return chunk2mem (newp);
4334	}
4335	}
4336	}
4337
4338	/ If possible, free extra space in old or extended chunk /
4339
4340	assert ((unsigned long) (newsize) >= (unsigned long) (nb));
4341
4342	remainder_size = newsize - nb;
4343
4344	if (remainder_size < MINSIZE) / not enough extra to split off /
4345	{
4346	set_head_size (newp, newsize \| (av != &main_arena ? NON_MAIN_ARENA : `0`));
4347	set_inuse_bit_at_offset (newp, newsize);
4348	}
4349	else / split remainder /
4350	{
4351	remainder = chunk_at_offset (newp, nb);
4352	set_head_size (newp, nb \| (av != &main_arena ? NON_MAIN_ARENA : `0`));
4353	set_head (remainder, remainder_size \| PREV_INUSE \|
4354	(av != &main_arena ? NON_MAIN_ARENA : `0`));
4355	/ Mark remainder as inuse so free() won't complain /
4356	set_inuse_bit_at_offset (remainder, remainder_size);
4357	_int_free (av, remainder, `1`);
4358	}
4359
4360	check_inuse_chunk (av, newp);
4361	return chunk2mem (newp);
4362	}
4363
4364	/*
4365	------------------------------ memalign ------------------------------
4366	*/
4367
4368	static void *
4369	_int_memalign (mstate av, size_t alignment, size_t bytes)
4370	{
4371	INTERNAL_SIZE_T nb; / padded request size /
4372	char m; /* memory returned by malloc call /
4373	mchunkptr p; / corresponding chunk /
4374	char brk; /* alignment point within p /
4375	mchunkptr newp; / chunk to return /
4376	INTERNAL_SIZE_T newsize; / its size /
4377	INTERNAL_SIZE_T leadsize; / leading space before alignment point /
4378	mchunkptr remainder; / spare room at end to split off /
4379	unsigned long remainder_size; / its size /
4380	INTERNAL_SIZE_T size;
4381
4382
4383
4384	checked_request2size (bytes, nb);
4385
4386	/*
4387	Strategy: find a spot within that chunk that meets the alignment
4388	request, and then possibly free the leading and trailing space.
4389	*/
4390
4391
4392	/ Check for overflow. /
4393	if (nb > SIZE_MAX - alignment - MINSIZE)
4394	{
4395	__set_errno (ENOMEM);
4396	return `0`;
4397	}
4398
4399	/ Call malloc with worst case padding to hit alignment. /
4400
4401	m = (char *) (_int_malloc (av, nb + alignment + MINSIZE));
4402
4403	if (m == `0`)
4404	return `0`; / propagate failure /
4405
4406	p = mem2chunk (m);
4407
4408	if ((((unsigned long) (m)) % alignment) != `0`) / misaligned /
4409
4410	{ /*
4411	Find an aligned spot inside chunk. Since we need to give back
4412	leading space in a chunk of at least MINSIZE, if the first
4413	calculation places us at a spot with less than MINSIZE leader,
4414	we can move to the next aligned spot -- we've allocated enough
4415	total room so that this is always possible.
4416	*/
4417	brk = (char ) mem2chunk (((unsigned* long) (m + alignment - `1`)) &
4418	- ((signed long) alignment));
4419	if ((unsigned long) (brk - (char *) (p)) < MINSIZE)
4420	brk += alignment;
4421
4422	newp = (mchunkptr) brk;
4423	leadsize = brk - (char *) (p);
4424	newsize = chunksize (p) - leadsize;
4425
4426	/ For mmapped chunks, just adjust offset /
4427	if (chunk_is_mmapped (p))
4428	{
4429	set_prev_size (newp, prev_size (p) + leadsize);
4430	set_head (newp, newsize \| IS_MMAPPED);
4431	return chunk2mem (newp);
4432	}
4433
4434	/ Otherwise, give back leader, use the rest /
4435	set_head (newp, newsize \| PREV_INUSE \|
4436	(av != &main_arena ? NON_MAIN_ARENA : `0`));
4437	set_inuse_bit_at_offset (newp, newsize);
4438	set_head_size (p, leadsize \| (av != &main_arena ? NON_MAIN_ARENA : `0`));
4439	_int_free (av, p, `1`);
4440	p = newp;
4441
4442	assert (newsize >= nb &&
4443	(((unsigned long) (chunk2mem (p))) % alignment) == `0`);
4444	}
4445
4446	/ Also give back spare room at the end /
4447	if (!chunk_is_mmapped (p))
4448	{
4449	size = chunksize (p);
4450	if ((unsigned long) (size) > (unsigned long) (nb + MINSIZE))
4451	{
4452	remainder_size = size - nb;
4453	remainder = chunk_at_offset (p, nb);
4454	set_head (remainder, remainder_size \| PREV_INUSE \|
4455	(av != &main_arena ? NON_MAIN_ARENA : `0`));
4456	set_head_size (p, nb);
4457	_int_free (av, remainder, `1`);
4458	}
4459	}
4460
4461	check_inuse_chunk (av, p);
4462	return chunk2mem (p);
4463	}
4464
4465
4466	/*
4467	------------------------------ malloc_trim ------------------------------
4468	*/
4469
4470	static int
4471	mtrim (mstate av, size_t pad)
4472	{
4473	/ Don't touch corrupt arenas. /
4474	if (arena_is_corrupt (av))
4475	return `0`;
4476
4477	/ Ensure initialization/consolidation /
4478	malloc_consolidate (av);
4479
4480	const size_t ps = GLRO (dl_pagesize);
4481	int psindex = bin_index (ps);
4482	const size_t psm1 = ps - `1`;
4483
4484	int result = `0`;
4485	for (int i = `1`; i < NBINS; ++i)
4486	if (i == `1` \|\| i >= psindex)
4487	{
4488	mbinptr bin = bin_at (av, i);
4489
4490	for (mchunkptr p = last (bin); p != bin; p = p->bk)
4491	{
4492	INTERNAL_SIZE_T size = chunksize (p);
4493
4494	if (size > psm1 + sizeof (struct malloc_chunk))
4495	{
4496	/ See whether the chunk contains at least one unused page. /
4497	char paligned_mem = (char* *) (((uintptr_t) p
4498	+ sizeof (struct malloc_chunk)
4499	+ psm1) & ~psm1);
4500
4501	assert ((char ) chunk2mem (p) + `4` SIZE_SZ <= paligned_mem);
4502	assert ((char *) p + size > paligned_mem);
4503
4504	/ This is the size we could potentially free. /
4505	size -= paligned_mem - (char *) p;
4506
4507	if (size > psm1)
4508	{
4509	#if MALLOC_DEBUG
4510	/ When debugging we simulate destroying the memory*
4511	content. /*
4512	memset (paligned_mem, `0x89`, size & ~psm1);
4513	#endif
4514	__madvise (paligned_mem, size & ~psm1, MADV_DONTNEED);
4515
4516	result = `1`;
4517	}
4518	}
4519	}
4520	}
4521
4522	#ifndef MORECORE_CANNOT_TRIM
4523	return result \| (av == &main_arena ? systrim (pad, av) : `0`);
4524
4525	#else
4526	return result;
4527	#endif
4528	}
4529
4530
4531	int
4532	__malloc_trim (size_t s)
4533	{
4534	int result = `0`;
4535
4536	if (__malloc_initialized < `0`)
4537	ptmalloc_init ();
4538
4539	mstate ar_ptr = &main_arena;
4540	do
4541	{
4542	__libc_lock_lock (ar_ptr->mutex);
4543	result \|= mtrim (ar_ptr, s);
4544	__libc_lock_unlock (ar_ptr->mutex);
4545
4546	ar_ptr = ar_ptr->next;
4547	}
4548	while (ar_ptr != &main_arena);
4549
4550	return result;
4551	}
4552
4553
4554	/*
4555	------------------------- malloc_usable_size -------------------------
4556	*/
4557
4558	static size_t
4559	musable (void *mem)
4560	{
4561	mchunkptr p;
4562	if (mem != `0`)
4563	{
4564	p = mem2chunk (mem);
4565
4566	if (__builtin_expect (using_malloc_checking == `1`, `0`))
4567	return malloc_check_get_size (p);
4568
4569	if (chunk_is_mmapped (p))
4570	{
4571	if (DUMPED_MAIN_ARENA_CHUNK (p))
4572	return chunksize (p) - SIZE_SZ;
4573	else
4574	return chunksize (p) - `2` * SIZE_SZ;
4575	}
4576	else if (inuse (p))
4577	return chunksize (p) - SIZE_SZ;
4578	}
4579	return `0`;
4580	}
4581
4582
4583	size_t
4584	__malloc_usable_size (void *m)
4585	{
4586	size_t result;
4587
4588	result = musable (m);
4589	return result;
4590	}
4591
4592	/*
4593	------------------------------ mallinfo ------------------------------
4594	Accumulate malloc statistics for arena AV into M.
4595	*/
4596
4597	static void
4598	int_mallinfo (mstate av, struct mallinfo *m)
4599	{
4600	size_t i;
4601	mbinptr b;
4602	mchunkptr p;
4603	INTERNAL_SIZE_T avail;
4604	INTERNAL_SIZE_T fastavail;
4605	int nblocks;
4606	int nfastblocks;
4607
4608	/ Ensure initialization /
4609	if (av->top == `0`)
4610	malloc_consolidate (av);
4611
4612	check_malloc_state (av);
4613
4614	/ Account for top /
4615	avail = chunksize (av->top);
4616	nblocks = `1`; / top always exists /
4617
4618	/ traverse fastbins /
4619	nfastblocks = `0`;
4620	fastavail = `0`;
4621
4622	for (i = `0`; i < NFASTBINS; ++i)
4623	{
4624	for (p = fastbin (av, i); p != `0`; p = p->fd)
4625	{
4626	++nfastblocks;
4627	fastavail += chunksize (p);
4628	}
4629	}
4630
4631	avail += fastavail;
4632
4633	/ traverse regular bins /
4634	for (i = `1`; i < NBINS; ++i)
4635	{
4636	b = bin_at (av, i);
4637	for (p = last (b); p != b; p = p->bk)
4638	{
4639	++nblocks;
4640	avail += chunksize (p);
4641	}
4642	}
4643
4644	m->smblks += nfastblocks;
4645	m->ordblks += nblocks;
4646	m->fordblks += avail;
4647	m->uordblks += av->system_mem - avail;
4648	m->arena += av->system_mem;
4649	m->fsmblks += fastavail;
4650	if (av == &main_arena)
4651	{
4652	m->hblks = mp_.n_mmaps;
4653	m->hblkhd = mp_.mmapped_mem;
4654	m->usmblks = `0`;
4655	m->keepcost = chunksize (av->top);
4656	}
4657	}
4658
4659
4660	struct mallinfo
4661	__libc_mallinfo (void)
4662	{
4663	struct mallinfo m;
4664	mstate ar_ptr;
4665
4666	if (__malloc_initialized < `0`)
4667	ptmalloc_init ();
4668
4669	memset (&m, `0`, sizeof (m));
4670	ar_ptr = &main_arena;
4671	do
4672	{
4673	__libc_lock_lock (ar_ptr->mutex);
4674	int_mallinfo (ar_ptr, &m);
4675	__libc_lock_unlock (ar_ptr->mutex);
4676
4677	ar_ptr = ar_ptr->next;
4678	}
4679	while (ar_ptr != &main_arena);
4680
4681	return m;
4682	}
4683
4684	/*
4685	------------------------------ malloc_stats ------------------------------
4686	*/
4687
4688	void
4689	__malloc_stats (void)
4690	{
4691	int i;
4692	mstate ar_ptr;
4693	unsigned int in_use_b = mp_.mmapped_mem, system_b = in_use_b;
4694
4695	if (__malloc_initialized < `0`)
4696	ptmalloc_init ();
4697	_IO_flockfile (stderr);
4698	int old_flags2 = ((_IO_FILE *) stderr)->_flags2;
4699	((_IO_FILE *) stderr)->_flags2 \|= _IO_FLAGS2_NOTCANCEL;
4700	for (i = `0`, ar_ptr = &main_arena;; i++)
4701	{
4702	struct mallinfo mi;
4703
4704	memset (&mi, `0`, sizeof (mi));
4705	__libc_lock_lock (ar_ptr->mutex);
4706	int_mallinfo (ar_ptr, &mi);
4707	fprintf (stderr, "Arena %d:\n", i);
4708	fprintf (stderr, "system bytes = %10u\n", (unsigned int) mi.arena);
4709	fprintf (stderr, "in use bytes = %10u\n", (unsigned int) mi.uordblks);
4710	#if MALLOC_DEBUG > 1
4711	if (i > `0`)
4712	dump_heap (heap_for_ptr (top (ar_ptr)));
4713	#endif
4714	system_b += mi.arena;
4715	in_use_b += mi.uordblks;
4716	__libc_lock_unlock (ar_ptr->mutex);
4717	ar_ptr = ar_ptr->next;
4718	if (ar_ptr == &main_arena)
4719	break;
4720	}
4721	fprintf (stderr, "Total (incl. mmap):\n");
4722	fprintf (stderr, "system bytes = %10u\n", system_b);
4723	fprintf (stderr, "in use bytes = %10u\n", in_use_b);
4724	fprintf (stderr, "max mmap regions = %10u\n", (unsigned int) mp_.max_n_mmaps);
4725	fprintf (stderr, "max mmap bytes = %10lu\n",
4726	(unsigned long) mp_.max_mmapped_mem);
4727	((_IO_FILE *) stderr)->_flags2 \|= old_flags2;
4728	_IO_funlockfile (stderr);
4729	}
4730
4731
4732	/*
4733	------------------------------ mallopt ------------------------------
4734	*/
4735	static inline int
4736	__always_inline
4737	do_set_trim_threshold (size_t value)
4738	{
4739	LIBC_PROBE (memory_mallopt_trim_threshold, `3`, value, mp_.trim_threshold,
4740	mp_.no_dyn_threshold);
4741	mp_.trim_threshold = value;
4742	mp_.no_dyn_threshold = `1`;
4743	return `1`;
4744	}
4745
4746	static inline int
4747	__always_inline
4748	do_set_top_pad (size_t value)
4749	{
4750	LIBC_PROBE (memory_mallopt_top_pad, `3`, value, mp_.top_pad,
4751	mp_.no_dyn_threshold);
4752	mp_.top_pad = value;
4753	mp_.no_dyn_threshold = `1`;
4754	return `1`;
4755	}
4756
4757	static inline int
4758	__always_inline
4759	do_set_mmap_threshold (size_t value)
4760	{
4761	/ Forbid setting the threshold too high. /
4762	if (value <= HEAP_MAX_SIZE / `2`)
4763	{
4764	LIBC_PROBE (memory_mallopt_mmap_threshold, `3`, value, mp_.mmap_threshold,
4765	mp_.no_dyn_threshold);
4766	mp_.mmap_threshold = value;
4767	mp_.no_dyn_threshold = `1`;
4768	return `1`;
4769	}
4770	return `0`;
4771	}
4772
4773	static inline int
4774	__always_inline
4775	do_set_mmaps_max (int32_t value)
4776	{
4777	LIBC_PROBE (memory_mallopt_mmap_max, `3`, value, mp_.n_mmaps_max,
4778	mp_.no_dyn_threshold);
4779	mp_.n_mmaps_max = value;
4780	mp_.no_dyn_threshold = `1`;
4781	return `1`;
4782	}
4783
4784	static inline int
4785	__always_inline
4786	do_set_mallopt_check (int32_t value)
4787	{
4788	LIBC_PROBE (memory_mallopt_check_action, `2`, value, check_action);
4789	check_action = value;
4790	return `1`;
4791	}
4792
4793	static inline int
4794	__always_inline
4795	do_set_perturb_byte (int32_t value)
4796	{
4797	LIBC_PROBE (memory_mallopt_perturb, `2`, value, perturb_byte);
4798	perturb_byte = value;
4799	return `1`;
4800	}
4801
4802	static inline int
4803	__always_inline
4804	do_set_arena_test (size_t value)
4805	{
4806	LIBC_PROBE (memory_mallopt_arena_test, `2`, value, mp_.arena_test);
4807	mp_.arena_test = value;
4808	return `1`;
4809	}
4810
4811	static inline int
4812	__always_inline
4813	do_set_arena_max (size_t value)
4814	{
4815	LIBC_PROBE (memory_mallopt_arena_max, `2`, value, mp_.arena_max);
4816	mp_.arena_max = value;
4817	return `1`;
4818	}
4819
4820
4821	int
4822	__libc_mallopt (int param_number, int value)
4823	{
4824	mstate av = &main_arena;
4825	int res = `1`;
4826
4827	if (__malloc_initialized < `0`)
4828	ptmalloc_init ();
4829	__libc_lock_lock (av->mutex);
4830	/ Ensure initialization/consolidation /
4831	malloc_consolidate (av);
4832
4833	LIBC_PROBE (memory_mallopt, `2`, param_number, value);
4834
4835	switch (param_number)
4836	{
4837	case M_MXFAST:
4838	if (value >= `0` && value <= MAX_FAST_SIZE)
4839	{
4840	LIBC_PROBE (memory_mallopt_mxfast, `2`, value, get_max_fast ());
4841	set_max_fast (value);
4842	}
4843	else
4844	res = `0`;
4845	break;
4846
4847	case M_TRIM_THRESHOLD:
4848	do_set_trim_threshold (value);
4849	break;
4850
4851	case M_TOP_PAD:
4852	do_set_top_pad (value);
4853	break;
4854
4855	case M_MMAP_THRESHOLD:
4856	res = do_set_mmap_threshold (value);
4857	break;
4858
4859	case M_MMAP_MAX:
4860	do_set_mmaps_max (value);
4861	break;
4862
4863	case M_CHECK_ACTION:
4864	do_set_mallopt_check (value);
4865	break;
4866
4867	case M_PERTURB:
4868	do_set_perturb_byte (value);
4869	break;
4870
4871	case M_ARENA_TEST:
4872	if (value > `0`)
4873	do_set_arena_test (value);
4874	break;
4875
4876	case M_ARENA_MAX:
4877	if (value > `0`)
4878	do_set_arena_max (value);
4879	break;
4880	}
4881	__libc_lock_unlock (av->mutex);
4882	return res;
4883	}
4884	libc_hidden_def (__libc_mallopt)
4885
4886
4887	/*
4888	-------------------- Alternative MORECORE functions --------------------
4889	*/
4890
4891
4892	/*
4893	General Requirements for MORECORE.
4894
4895	The MORECORE function must have the following properties:
4896
4897	If MORECORE_CONTIGUOUS is false:
4898
4899	* MORECORE must allocate in multiples of pagesize. It will
4900	only be called with arguments that are multiples of pagesize.
4901
4902	* MORECORE(0) must return an address that is at least
4903	MALLOC_ALIGNMENT aligned. (Page-aligning always suffices.)
4904
4905	else (i.e. If MORECORE_CONTIGUOUS is true):
4906
4907	* Consecutive calls to MORECORE with positive arguments
4908	return increasing addresses, indicating that space has been
4909	contiguously extended.
4910
4911	* MORECORE need not allocate in multiples of pagesize.
4912	Calls to MORECORE need not have args of multiples of pagesize.
4913
4914	* MORECORE need not page-align.
4915
4916	In either case:
4917
4918	* MORECORE may allocate more memory than requested. (Or even less,
4919	but this will generally result in a malloc failure.)
4920
4921	* MORECORE must not allocate memory when given argument zero, but
4922	instead return one past the end address of memory from previous
4923	nonzero call. This malloc does NOT call MORECORE(0)
4924	until at least one call with positive arguments is made, so
4925	the initial value returned is not important.
4926
4927	* Even though consecutive calls to MORECORE need not return contiguous
4928	addresses, it must be OK for malloc'ed chunks to span multiple
4929	regions in those cases where they do happen to be contiguous.
4930
4931	* MORECORE need not handle negative arguments -- it may instead
4932	just return MORECORE_FAILURE when given negative arguments.
4933	Negative arguments are always multiples of pagesize. MORECORE
4934	must not misinterpret negative args as large positive unsigned
4935	args. You can suppress all such calls from even occurring by defining
4936	MORECORE_CANNOT_TRIM,
4937
4938	There is some variation across systems about the type of the
4939	argument to sbrk/MORECORE. If size_t is unsigned, then it cannot
4940	actually be size_t, because sbrk supports negative args, so it is
4941	normally the signed type of the same width as size_t (sometimes
4942	declared as "intptr_t", and sometimes "ptrdiff_t"). It doesn't much
4943	matter though. Internally, we use "long" as arguments, which should
4944	work across all reasonable possibilities.
4945
4946	Additionally, if MORECORE ever returns failure for a positive
4947	request, then mmap is used as a noncontiguous system allocator. This
4948	is a useful backup strategy for systems with holes in address spaces
4949	-- in this case sbrk cannot contiguously expand the heap, but mmap
4950	may be able to map noncontiguous space.
4951
4952	If you'd like mmap to ALWAYS be used, you can define MORECORE to be
4953	a function that always returns MORECORE_FAILURE.
4954
4955	If you are using this malloc with something other than sbrk (or its
4956	emulation) to supply memory regions, you probably want to set
4957	MORECORE_CONTIGUOUS as false. As an example, here is a custom
4958	allocator kindly contributed for pre-OSX macOS. It uses virtually
4959	but not necessarily physically contiguous non-paged memory (locked
4960	in, present and won't get swapped out). You can use it by
4961	uncommenting this section, adding some #includes, and setting up the
4962	appropriate defines above:
4963
4964	*#define MORECORE osMoreCore
4965	*#define MORECORE_CONTIGUOUS 0
4966
4967	There is also a shutdown routine that should somehow be called for
4968	cleanup upon program exit.
4969
4970	*#define MAX_POOL_ENTRIES 100
4971	#define MINIMUM_MORECORE_SIZE (64 1024)
4972	static int next_os_pool;
4973	void our_os_pools[MAX_POOL_ENTRIES];*
4974
4975	void osMoreCore(int size)*
4976	{
4977	void ptr = 0;*
4978	static void sbrk_top = 0;*
4979
4980	if (size > 0)
4981	{
4982	if (size < MINIMUM_MORECORE_SIZE)
4983	size = MINIMUM_MORECORE_SIZE;
4984	if (CurrentExecutionLevel() == kTaskLevel)
4985	ptr = PoolAllocateResident(size + RM_PAGE_SIZE, 0);
4986	if (ptr == 0)
4987	{
4988	return (void ) MORECORE_FAILURE;*
4989	}
4990	// save ptrs so they can be freed during cleanup
4991	our_os_pools[next_os_pool] = ptr;
4992	next_os_pool++;
4993	ptr = (void ) ((((unsigned long) ptr) + RM_PAGE_MASK) & ~RM_PAGE_MASK);*
4994	sbrk_top = (char ) ptr + size;*
4995	return ptr;
4996	}
4997	else if (size < 0)
4998	{
4999	// we don't currently support shrink behavior
5000	return (void ) MORECORE_FAILURE;*
5001	}
5002	else
5003	{
5004	return sbrk_top;
5005	}
5006	}
5007
5008	// cleanup any allocated memory pools
5009	// called as last thing before shutting down driver
5010
5011	void osCleanupMem(void)
5012	{
5013	void ptr;
5014
5015	for (ptr = our_os_pools; ptr < &our_os_pools[MAX_POOL_ENTRIES]; ptr++)
5016	if (ptr)*
5017	{
5018	PoolDeallocate(ptr);*
5019	* ptr = 0;
5020	}
5021	}
5022
5023	*/
5024
5025
5026	/ Helper code. /
5027
5028	extern char **__libc_argv attribute_hidden;
5029
5030	static void
5031	malloc_printerr (int action, const char str, void* *ptr, mstate ar_ptr)
5032	{
5033	/ Avoid using this arena in future. We do not attempt to synchronize this*
5034	with anything else because we minimally want to ensure that __libc_message
5035	gets its resources safely without stumbling on the current corruption. /*
5036	if (ar_ptr)
5037	set_arena_corrupt (ar_ptr);
5038
5039	if ((action & `5`) == `5`)
5040	__libc_message (action & `2`, "%s\n", str);
5041	else if (action & `1`)
5042	{
5043	char buf[`2` * sizeof (uintptr_t) + `1`];
5044
5045	buf[sizeof (buf) - `1`] = `'\0'`;
5046	char cp = _itoa_word ((uintptr_t) ptr, &buf[sizeof* (buf) - `1`], `16`, `0`);
5047	while (cp > buf)
5048	*--cp = `'0'`;
5049
5050	__libc_message (action & `2`, "* Error in `%s': %s: 0x%s *\n",
5051	__libc_argv[`0`] ? : "<unknown>", str, cp);
5052	}
5053	else if (action & `2`)
5054	abort ();
5055	}
5056
5057	/ We need a wrapper function for one of the additions of POSIX. /
5058	int
5059	__posix_memalign (void **memptr, size_t alignment, size_t size)
5060	{
5061	void *mem;
5062
5063	/ Test whether the SIZE argument is valid. It must be a power of*
5064	two multiple of sizeof (void ). /
5065	if (alignment % sizeof (void *) != `0`
5066	\|\| !powerof2 (alignment / sizeof (void *))
5067	\|\| alignment == `0`)
5068	return EINVAL;
5069
5070
5071	void *address = RETURN_ADDRESS (`0`);
5072	mem = _mid_memalign (alignment, size, address);
5073
5074	if (mem != NULL)
5075	{
5076	*memptr = mem;
5077	return `0`;
5078	}
5079
5080	return ENOMEM;
5081	}
5082	weak_alias (__posix_memalign, posix_memalign)
5083
5084
5085	int
5086	__malloc_info (int options, FILE *fp)
5087	{
5088	/ For now, at least. /
5089	if (options != `0`)
5090	return EINVAL;
5091
5092	int n = `0`;
5093	size_t total_nblocks = `0`;
5094	size_t total_nfastblocks = `0`;
5095	size_t total_avail = `0`;
5096	size_t total_fastavail = `0`;
5097	size_t total_system = `0`;
5098	size_t total_max_system = `0`;
5099	size_t total_aspace = `0`;
5100	size_t total_aspace_mprotect = `0`;
5101
5102
5103
5104	if (__malloc_initialized < `0`)
5105	ptmalloc_init ();
5106
5107	fputs ("<malloc version=\"1\">\n", fp);
5108
5109	/ Iterate over all arenas currently in use. /
5110	mstate ar_ptr = &main_arena;
5111	do
5112	{
5113	fprintf (fp, "<heap nr=\"%d\">\n<sizes>\n", n++);
5114
5115	size_t nblocks = `0`;
5116	size_t nfastblocks = `0`;
5117	size_t avail = `0`;
5118	size_t fastavail = `0`;
5119	struct
5120	{
5121	size_t from;
5122	size_t to;
5123	size_t total;
5124	size_t count;
5125	} sizes[NFASTBINS + NBINS - `1`];
5126	#define nsizes (sizeof (sizes) / sizeof (sizes[0]))
5127
5128	__libc_lock_lock (ar_ptr->mutex);
5129
5130	for (size_t i = `0`; i < NFASTBINS; ++i)
5131	{
5132	mchunkptr p = fastbin (ar_ptr, i);
5133	if (p != NULL)
5134	{
5135	size_t nthissize = `0`;
5136	size_t thissize = chunksize (p);
5137
5138	while (p != NULL)
5139	{
5140	++nthissize;
5141	p = p->fd;
5142	}
5143
5144	fastavail += nthissize * thissize;
5145	nfastblocks += nthissize;
5146	sizes[i].from = thissize - (MALLOC_ALIGNMENT - `1`);
5147	sizes[i].to = thissize;
5148	sizes[i].count = nthissize;
5149	}
5150	else
5151	sizes[i].from = sizes[i].to = sizes[i].count = `0`;
5152
5153	sizes[i].total = sizes[i].count * sizes[i].to;
5154	}
5155
5156
5157	mbinptr bin;
5158	struct malloc_chunk *r;
5159
5160	for (size_t i = `1`; i < NBINS; ++i)
5161	{
5162	bin = bin_at (ar_ptr, i);
5163	r = bin->fd;
5164	sizes[NFASTBINS - `1` + i].from = ~((size_t) `0`);
5165	sizes[NFASTBINS - `1` + i].to = sizes[NFASTBINS - `1` + i].total
5166	= sizes[NFASTBINS - `1` + i].count = `0`;
5167
5168	if (r != NULL)
5169	while (r != bin)
5170	{
5171	size_t r_size = chunksize_nomask (r);
5172	++sizes[NFASTBINS - `1` + i].count;
5173	sizes[NFASTBINS - `1` + i].total += r_size;
5174	sizes[NFASTBINS - `1` + i].from
5175	= MIN (sizes[NFASTBINS - `1` + i].from, r_size);
5176	sizes[NFASTBINS - `1` + i].to = MAX (sizes[NFASTBINS - `1` + i].to,
5177	r_size);
5178
5179	r = r->fd;
5180	}
5181
5182	if (sizes[NFASTBINS - `1` + i].count == `0`)
5183	sizes[NFASTBINS - `1` + i].from = `0`;
5184	nblocks += sizes[NFASTBINS - `1` + i].count;
5185	avail += sizes[NFASTBINS - `1` + i].total;
5186	}
5187
5188	__libc_lock_unlock (ar_ptr->mutex);
5189
5190	total_nfastblocks += nfastblocks;
5191	total_fastavail += fastavail;
5192
5193	total_nblocks += nblocks;
5194	total_avail += avail;
5195
5196	for (size_t i = `0`; i < nsizes; ++i)
5197	if (sizes[i].count != `0` && i != NFASTBINS)
5198	fprintf (fp, " \
5199	<size from=\"%zu\" to=\"%zu\" total=\"%zu\" count=\"%zu\"/>\n",
5200	sizes[i].from, sizes[i].to, sizes[i].total, sizes[i].count);
5201
5202	if (sizes[NFASTBINS].count != `0`)
5203	fprintf (fp, "\
5204	<unsorted from=\"%zu\" to=\"%zu\" total=\"%zu\" count=\"%zu\"/>\n",
5205	sizes[NFASTBINS].from, sizes[NFASTBINS].to,
5206	sizes[NFASTBINS].total, sizes[NFASTBINS].count);
5207
5208	total_system += ar_ptr->system_mem;
5209	total_max_system += ar_ptr->max_system_mem;
5210
5211	fprintf (fp,
5212	"</sizes>\n<total type=\"fast\" count=\"%zu\" size=\"%zu\"/>\n"
5213	"<total type=\"rest\" count=\"%zu\" size=\"%zu\"/>\n"
5214	"<system type=\"current\" size=\"%zu\"/>\n"
5215	"<system type=\"max\" size=\"%zu\"/>\n",
5216	nfastblocks, fastavail, nblocks, avail,
5217	ar_ptr->system_mem, ar_ptr->max_system_mem);
5218
5219	if (ar_ptr != &main_arena)
5220	{
5221	heap_info *heap = heap_for_ptr (top (ar_ptr));
5222	fprintf (fp,
5223	"<aspace type=\"total\" size=\"%zu\"/>\n"
5224	"<aspace type=\"mprotect\" size=\"%zu\"/>\n",
5225	heap->size, heap->mprotect_size);
5226	total_aspace += heap->size;
5227	total_aspace_mprotect += heap->mprotect_size;
5228	}
5229	else
5230	{
5231	fprintf (fp,
5232	"<aspace type=\"total\" size=\"%zu\"/>\n"
5233	"<aspace type=\"mprotect\" size=\"%zu\"/>\n",
5234	ar_ptr->system_mem, ar_ptr->system_mem);
5235	total_aspace += ar_ptr->system_mem;
5236	total_aspace_mprotect += ar_ptr->system_mem;
5237	}
5238
5239	fputs ("</heap>\n", fp);
5240	ar_ptr = ar_ptr->next;
5241	}
5242	while (ar_ptr != &main_arena);
5243
5244	fprintf (fp,
5245	"<total type=\"fast\" count=\"%zu\" size=\"%zu\"/>\n"
5246	"<total type=\"rest\" count=\"%zu\" size=\"%zu\"/>\n"
5247	"<total type=\"mmap\" count=\"%d\" size=\"%zu\"/>\n"
5248	"<system type=\"current\" size=\"%zu\"/>\n"
5249	"<system type=\"max\" size=\"%zu\"/>\n"
5250	"<aspace type=\"total\" size=\"%zu\"/>\n"
5251	"<aspace type=\"mprotect\" size=\"%zu\"/>\n"
5252	"</malloc>\n",
5253	total_nfastblocks, total_fastavail, total_nblocks, total_avail,
5254	mp_.n_mmaps, mp_.mmapped_mem,
5255	total_system, total_max_system,
5256	total_aspace, total_aspace_mprotect);
5257
5258	return `0`;
5259	}
5260	weak_alias (__malloc_info, malloc_info)
5261
5262
5263	strong_alias (__libc_calloc, __calloc) weak_alias (__libc_calloc, calloc)
5264	strong_alias (__libc_free, __cfree) weak_alias (__libc_free, cfree)
5265	strong_alias (__libc_free, __free) strong_alias (__libc_free, free)
5266	strong_alias (__libc_malloc, __malloc) strong_alias (__libc_malloc, malloc)
5267	strong_alias (__libc_memalign, __memalign)
5268	weak_alias (__libc_memalign, memalign)
5269	strong_alias (__libc_realloc, __realloc) strong_alias (__libc_realloc, realloc)
5270	strong_alias (__libc_valloc, __valloc) weak_alias (__libc_valloc, valloc)
5271	strong_alias (__libc_pvalloc, __pvalloc) weak_alias (__libc_pvalloc, pvalloc)
5272	strong_alias (__libc_mallinfo, __mallinfo)
5273	weak_alias (__libc_mallinfo, mallinfo)
5274	strong_alias (__libc_mallopt, __mallopt) weak_alias (__libc_mallopt, mallopt)
5275
5276	weak_alias (__malloc_stats, malloc_stats)
5277	weak_alias (__malloc_usable_size, malloc_usable_size)
5278	weak_alias (__malloc_trim, malloc_trim)
5279
5280
5281	/ ------------------------------------------------------------*
5282	History:
5283
5284	[see ftp://g.oswego.edu/pub/misc/malloc.c for the history of dlmalloc]
5285
5286	*/
5287	/*
5288	* Local variables:
5289	* c-basic-offset: 2
5290	* End:
5291	*/
5292

Browse the source code of glibc_src_2.25/malloc/malloc.c