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 malloc_trim(size_t pad);
88 malloc_usable_size(void* p);
89 malloc_stats();
90
91* Vital statistics:
92
93 Supported pointer representation: 4 or 8 bytes
94 Supported size_t representation: 4 or 8 bytes
95 Note that size_t is allowed to be 4 bytes even if pointers are 8.
96 You can adjust this by defining INTERNAL_SIZE_T
97
98 Alignment: 2 * sizeof(size_t) (default)
99 (i.e., 8 byte alignment with 4byte size_t). This suffices for
100 nearly all current machines and C compilers. However, you can
101 define MALLOC_ALIGNMENT to be wider than this if necessary.
102
103 Minimum overhead per allocated chunk: 4 or 8 bytes
104 Each malloced chunk has a hidden word of overhead holding size
105 and status information.
106
107 Minimum allocated size: 4-byte ptrs: 16 bytes (including 4 overhead)
108 8-byte ptrs: 24/32 bytes (including, 4/8 overhead)
109
110 When a chunk is freed, 12 (for 4byte ptrs) or 20 (for 8 byte
111 ptrs but 4 byte size) or 24 (for 8/8) additional bytes are
112 needed; 4 (8) for a trailing size field and 8 (16) bytes for
113 free list pointers. Thus, the minimum allocatable size is
114 16/24/32 bytes.
115
116 Even a request for zero bytes (i.e., malloc(0)) returns a
117 pointer to something of the minimum allocatable size.
118
119 The maximum overhead wastage (i.e., number of extra bytes
120 allocated than were requested in malloc) is less than or equal
121 to the minimum size, except for requests >= mmap_threshold that
122 are serviced via mmap(), where the worst case wastage is 2 *
123 sizeof(size_t) bytes plus the remainder from a system page (the
124 minimal mmap unit); typically 4096 or 8192 bytes.
125
126 Maximum allocated size: 4-byte size_t: 2^32 minus about two pages
127 8-byte size_t: 2^64 minus about two pages
128
129 It is assumed that (possibly signed) size_t values suffice to
130 represent chunk sizes. `Possibly signed' is due to the fact
131 that `size_t' may be defined on a system as either a signed or
132 an unsigned type. The ISO C standard says that it must be
133 unsigned, but a few systems are known not to adhere to this.
134 Additionally, even when size_t is unsigned, sbrk (which is by
135 default used to obtain memory from system) accepts signed
136 arguments, and may not be able to handle size_t-wide arguments
137 with negative sign bit. Generally, values that would
138 appear as negative after accounting for overhead and alignment
139 are supported only via mmap(), which does not have this
140 limitation.
141
142 Requests for sizes outside the allowed range will perform an optional
143 failure action and then return null. (Requests may also
144 also fail because a system is out of memory.)
145
146 Thread-safety: thread-safe
147
148 Compliance: I believe it is compliant with the 1997 Single Unix Specification
149 Also SVID/XPG, ANSI C, and probably others as well.
150
151* Synopsis of compile-time options:
152
153 People have reported using previous versions of this malloc on all
154 versions of Unix, sometimes by tweaking some of the defines
155 below. It has been tested most extensively on Solaris and Linux.
156 People also report using it in stand-alone embedded systems.
157
158 The implementation is in straight, hand-tuned ANSI C. It is not
159 at all modular. (Sorry!) It uses a lot of macros. To be at all
160 usable, this code should be compiled using an optimizing compiler
161 (for example gcc -O3) that can simplify expressions and control
162 paths. (FAQ: some macros import variables as arguments rather than
163 declare locals because people reported that some debuggers
164 otherwise get confused.)
165
166 OPTION DEFAULT VALUE
167
168 Compilation Environment options:
169
170 HAVE_MREMAP 0
171
172 Changing default word sizes:
173
174 INTERNAL_SIZE_T size_t
175
176 Configuration and functionality options:
177
178 USE_PUBLIC_MALLOC_WRAPPERS NOT defined
179 USE_MALLOC_LOCK NOT defined
180 MALLOC_DEBUG NOT defined
181 REALLOC_ZERO_BYTES_FREES 1
182 TRIM_FASTBINS 0
183
184 Options for customizing MORECORE:
185
186 MORECORE sbrk
187 MORECORE_FAILURE -1
188 MORECORE_CONTIGUOUS 1
189 MORECORE_CANNOT_TRIM NOT defined
190 MORECORE_CLEARS 1
191 MMAP_AS_MORECORE_SIZE (1024 * 1024)
192
193 Tuning options that are also dynamically changeable via mallopt:
194
195 DEFAULT_MXFAST 64 (for 32bit), 128 (for 64bit)
196 DEFAULT_TRIM_THRESHOLD 128 * 1024
197 DEFAULT_TOP_PAD 0
198 DEFAULT_MMAP_THRESHOLD 128 * 1024
199 DEFAULT_MMAP_MAX 65536
200
201 There are several other #defined constants and macros that you
202 probably don't want to touch unless you are extending or adapting malloc. */
203
204/*
205 void* is the pointer type that malloc should say it returns
206*/
207
208#ifndef void
209#define void void
210#endif /*void*/
211
212#include <stddef.h> /* for size_t */
213#include <stdlib.h> /* for getenv(), abort() */
214#include <unistd.h> /* for __libc_enable_secure */
215
216#include <atomic.h>
217#include <_itoa.h>
218#include <bits/wordsize.h>
219#include <sys/sysinfo.h>
220
221#include <ldsodefs.h>
222
223#include <unistd.h>
224#include <stdio.h> /* needed for malloc_stats */
225#include <errno.h>
226
227#include <shlib-compat.h>
228
229/* For uintptr_t. */
230#include <stdint.h>
231
232/* For va_arg, va_start, va_end. */
233#include <stdarg.h>
234
235/* For MIN, MAX, powerof2. */
236#include <sys/param.h>
237
238/* For ALIGN_UP et. al. */
239#include <libc-pointer-arith.h>
240
241/* For DIAG_PUSH/POP_NEEDS_COMMENT et al. */
242#include <libc-diag.h>
243
244#include <malloc/malloc-internal.h>
245
246/* For SINGLE_THREAD_P. */
247#include <sysdep-cancel.h>
248
249/*
250 Debugging:
251
252 Because freed chunks may be overwritten with bookkeeping fields, this
253 malloc will often die when freed memory is overwritten by user
254 programs. This can be very effective (albeit in an annoying way)
255 in helping track down dangling pointers.
256
257 If you compile with -DMALLOC_DEBUG, a number of assertion checks are
258 enabled that will catch more memory errors. You probably won't be
259 able to make much sense of the actual assertion errors, but they
260 should help you locate incorrectly overwritten memory. The checking
261 is fairly extensive, and will slow down execution
262 noticeably. Calling malloc_stats or mallinfo with MALLOC_DEBUG set
263 will attempt to check every non-mmapped allocated and free chunk in
264 the course of computing the summmaries. (By nature, mmapped regions
265 cannot be checked very much automatically.)
266
267 Setting MALLOC_DEBUG may also be helpful if you are trying to modify
268 this code. The assertions in the check routines spell out in more
269 detail the assumptions and invariants underlying the algorithms.
270
271 Setting MALLOC_DEBUG does NOT provide an automated mechanism for
272 checking that all accesses to malloced memory stay within their
273 bounds. However, there are several add-ons and adaptations of this
274 or other mallocs available that do this.
275*/
276
277#ifndef MALLOC_DEBUG
278#define MALLOC_DEBUG 0
279#endif
280
281#ifdef NDEBUG
282# define assert(expr) ((void) 0)
283#else
284# define assert(expr) \
285 ((expr) \
286 ? ((void) 0) \
287 : __malloc_assert (#expr, __FILE__, __LINE__, __func__))
288
289extern const char *__progname;
290
291static void
292__malloc_assert (const char *assertion, const char *file, unsigned int line,
293 const char *function)
294{
295 (void) __fxprintf (NULL, "%s%s%s:%u: %s%sAssertion `%s' failed.\n",
296 __progname, __progname[0] ? ": " : "",
297 file, line,
298 function ? function : "", function ? ": " : "",
299 assertion);
300 fflush (stderr);
301 abort ();
302}
303#endif
304
305#if USE_TCACHE
306/* We want 64 entries. This is an arbitrary limit, which tunables can reduce. */
307# define TCACHE_MAX_BINS 64
308# define MAX_TCACHE_SIZE tidx2usize (TCACHE_MAX_BINS-1)
309
310/* Only used to pre-fill the tunables. */
311# define tidx2usize(idx) (((size_t) idx) * MALLOC_ALIGNMENT + MINSIZE - SIZE_SZ)
312
313/* When "x" is from chunksize(). */
314# define csize2tidx(x) (((x) - MINSIZE + MALLOC_ALIGNMENT - 1) / MALLOC_ALIGNMENT)
315/* When "x" is a user-provided size. */
316# define usize2tidx(x) csize2tidx (request2size (x))
317
318/* With rounding and alignment, the bins are...
319 idx 0 bytes 0..24 (64-bit) or 0..12 (32-bit)
320 idx 1 bytes 25..40 or 13..20
321 idx 2 bytes 41..56 or 21..28
322 etc. */
323
324/* This is another arbitrary limit, which tunables can change. Each
325 tcache bin will hold at most this number of chunks. */
326# define TCACHE_FILL_COUNT 7
327#endif
328
329
330/*
331 REALLOC_ZERO_BYTES_FREES should be set if a call to
332 realloc with zero bytes should be the same as a call to free.
333 This is required by the C standard. Otherwise, since this malloc
334 returns a unique pointer for malloc(0), so does realloc(p, 0).
335*/
336
337#ifndef REALLOC_ZERO_BYTES_FREES
338#define REALLOC_ZERO_BYTES_FREES 1
339#endif
340
341/*
342 TRIM_FASTBINS controls whether free() of a very small chunk can
343 immediately lead to trimming. Setting to true (1) can reduce memory
344 footprint, but will almost always slow down programs that use a lot
345 of small chunks.
346
347 Define this only if you are willing to give up some speed to more
348 aggressively reduce system-level memory footprint when releasing
349 memory in programs that use many small chunks. You can get
350 essentially the same effect by setting MXFAST to 0, but this can
351 lead to even greater slowdowns in programs using many small chunks.
352 TRIM_FASTBINS is an in-between compile-time option, that disables
353 only those chunks bordering topmost memory from being placed in
354 fastbins.
355*/
356
357#ifndef TRIM_FASTBINS
358#define TRIM_FASTBINS 0
359#endif
360
361
362/* Definition for getting more memory from the OS. */
363#define MORECORE (*__morecore)
364#define MORECORE_FAILURE 0
365void * __default_morecore (ptrdiff_t);
366void *(*__morecore)(ptrdiff_t) = __default_morecore;
367
368
369#include <string.h>
370
371/*
372 MORECORE-related declarations. By default, rely on sbrk
373*/
374
375
376/*
377 MORECORE is the name of the routine to call to obtain more memory
378 from the system. See below for general guidance on writing
379 alternative MORECORE functions, as well as a version for WIN32 and a
380 sample version for pre-OSX macos.
381*/
382
383#ifndef MORECORE
384#define MORECORE sbrk
385#endif
386
387/*
388 MORECORE_FAILURE is the value returned upon failure of MORECORE
389 as well as mmap. Since it cannot be an otherwise valid memory address,
390 and must reflect values of standard sys calls, you probably ought not
391 try to redefine it.
392*/
393
394#ifndef MORECORE_FAILURE
395#define MORECORE_FAILURE (-1)
396#endif
397
398/*
399 If MORECORE_CONTIGUOUS is true, take advantage of fact that
400 consecutive calls to MORECORE with positive arguments always return
401 contiguous increasing addresses. This is true of unix sbrk. Even
402 if not defined, when regions happen to be contiguous, malloc will
403 permit allocations spanning regions obtained from different
404 calls. But defining this when applicable enables some stronger
405 consistency checks and space efficiencies.
406*/
407
408#ifndef MORECORE_CONTIGUOUS
409#define MORECORE_CONTIGUOUS 1
410#endif
411
412/*
413 Define MORECORE_CANNOT_TRIM if your version of MORECORE
414 cannot release space back to the system when given negative
415 arguments. This is generally necessary only if you are using
416 a hand-crafted MORECORE function that cannot handle negative arguments.
417*/
418
419/* #define MORECORE_CANNOT_TRIM */
420
421/* MORECORE_CLEARS (default 1)
422 The degree to which the routine mapped to MORECORE zeroes out
423 memory: never (0), only for newly allocated space (1) or always
424 (2). The distinction between (1) and (2) is necessary because on
425 some systems, if the application first decrements and then
426 increments the break value, the contents of the reallocated space
427 are unspecified.
428 */
429
430#ifndef MORECORE_CLEARS
431# define MORECORE_CLEARS 1
432#endif
433
434
435/*
436 MMAP_AS_MORECORE_SIZE is the minimum mmap size argument to use if
437 sbrk fails, and mmap is used as a backup. The value must be a
438 multiple of page size. This backup strategy generally applies only
439 when systems have "holes" in address space, so sbrk cannot perform
440 contiguous expansion, but there is still space available on system.
441 On systems for which this is known to be useful (i.e. most linux
442 kernels), this occurs only when programs allocate huge amounts of
443 memory. Between this, and the fact that mmap regions tend to be
444 limited, the size should be large, to avoid too many mmap calls and
445 thus avoid running out of kernel resources. */
446
447#ifndef MMAP_AS_MORECORE_SIZE
448#define MMAP_AS_MORECORE_SIZE (1024 * 1024)
449#endif
450
451/*
452 Define HAVE_MREMAP to make realloc() use mremap() to re-allocate
453 large blocks.
454*/
455
456#ifndef HAVE_MREMAP
457#define HAVE_MREMAP 0
458#endif
459
460/* We may need to support __malloc_initialize_hook for backwards
461 compatibility. */
462
463#if SHLIB_COMPAT (libc, GLIBC_2_0, GLIBC_2_24)
464# define HAVE_MALLOC_INIT_HOOK 1
465#else
466# define HAVE_MALLOC_INIT_HOOK 0
467#endif
468
469
470/*
471 This version of malloc supports the standard SVID/XPG mallinfo
472 routine that returns a struct containing usage properties and
473 statistics. It should work on any SVID/XPG compliant system that has
474 a /usr/include/malloc.h defining struct mallinfo. (If you'd like to
475 install such a thing yourself, cut out the preliminary declarations
476 as described above and below and save them in a malloc.h file. But
477 there's no compelling reason to bother to do this.)
478
479 The main declaration needed is the mallinfo struct that is returned
480 (by-copy) by mallinfo(). The SVID/XPG malloinfo struct contains a
481 bunch of fields that are not even meaningful in this version of
482 malloc. These fields are are instead filled by mallinfo() with
483 other numbers that might be of interest.
484*/
485
486
487/* ---------- description of public routines ------------ */
488
489/*
490 malloc(size_t n)
491 Returns a pointer to a newly allocated chunk of at least n bytes, or null
492 if no space is available. Additionally, on failure, errno is
493 set to ENOMEM on ANSI C systems.
494
495 If n is zero, malloc returns a minumum-sized chunk. (The minimum
496 size is 16 bytes on most 32bit systems, and 24 or 32 bytes on 64bit
497 systems.) On most systems, size_t is an unsigned type, so calls
498 with negative arguments are interpreted as requests for huge amounts
499 of space, which will often fail. The maximum supported value of n
500 differs across systems, but is in all cases less than the maximum
501 representable value of a size_t.
502*/
503void* __libc_malloc(size_t);
504libc_hidden_proto (__libc_malloc)
505
506/*
507 free(void* p)
508 Releases the chunk of memory pointed to by p, that had been previously
509 allocated using malloc or a related routine such as realloc.
510 It has no effect if p is null. It can have arbitrary (i.e., bad!)
511 effects if p has already been freed.
512
513 Unless disabled (using mallopt), freeing very large spaces will
514 when possible, automatically trigger operations that give
515 back unused memory to the system, thus reducing program footprint.
516*/
517void __libc_free(void*);
518libc_hidden_proto (__libc_free)
519
520/*
521 calloc(size_t n_elements, size_t element_size);
522 Returns a pointer to n_elements * element_size bytes, with all locations
523 set to zero.
524*/
525void* __libc_calloc(size_t, size_t);
526
527/*
528 realloc(void* p, size_t n)
529 Returns a pointer to a chunk of size n that contains the same data
530 as does chunk p up to the minimum of (n, p's size) bytes, or null
531 if no space is available.
532
533 The returned pointer may or may not be the same as p. The algorithm
534 prefers extending p when possible, otherwise it employs the
535 equivalent of a malloc-copy-free sequence.
536
537 If p is null, realloc is equivalent to malloc.
538
539 If space is not available, realloc returns null, errno is set (if on
540 ANSI) and p is NOT freed.
541
542 if n is for fewer bytes than already held by p, the newly unused
543 space is lopped off and freed if possible. Unless the #define
544 REALLOC_ZERO_BYTES_FREES is set, realloc with a size argument of
545 zero (re)allocates a minimum-sized chunk.
546
547 Large chunks that were internally obtained via mmap will always be
548 grown using malloc-copy-free sequences unless the system supports
549 MREMAP (currently only linux).
550
551 The old unix realloc convention of allowing the last-free'd chunk
552 to be used as an argument to realloc is not supported.
553*/
554void* __libc_realloc(void*, size_t);
555libc_hidden_proto (__libc_realloc)
556
557/*
558 memalign(size_t alignment, size_t n);
559 Returns a pointer to a newly allocated chunk of n bytes, aligned
560 in accord with the alignment argument.
561
562 The alignment argument should be a power of two. If the argument is
563 not a power of two, the nearest greater power is used.
564 8-byte alignment is guaranteed by normal malloc calls, so don't
565 bother calling memalign with an argument of 8 or less.
566
567 Overreliance on memalign is a sure way to fragment space.
568*/
569void* __libc_memalign(size_t, size_t);
570libc_hidden_proto (__libc_memalign)
571
572/*
573 valloc(size_t n);
574 Equivalent to memalign(pagesize, n), where pagesize is the page
575 size of the system. If the pagesize is unknown, 4096 is used.
576*/
577void* __libc_valloc(size_t);
578
579
580
581/*
582 mallopt(int parameter_number, int parameter_value)
583 Sets tunable parameters The format is to provide a
584 (parameter-number, parameter-value) pair. mallopt then sets the
585 corresponding parameter to the argument value if it can (i.e., so
586 long as the value is meaningful), and returns 1 if successful else
587 0. SVID/XPG/ANSI defines four standard param numbers for mallopt,
588 normally defined in malloc.h. Only one of these (M_MXFAST) is used
589 in this malloc. The others (M_NLBLKS, M_GRAIN, M_KEEP) don't apply,
590 so setting them has no effect. But this malloc also supports four
591 other options in mallopt. See below for details. Briefly, supported
592 parameters are as follows (listed defaults are for "typical"
593 configurations).
594
595 Symbol param # default allowed param values
596 M_MXFAST 1 64 0-80 (0 disables fastbins)
597 M_TRIM_THRESHOLD -1 128*1024 any (-1U disables trimming)
598 M_TOP_PAD -2 0 any
599 M_MMAP_THRESHOLD -3 128*1024 any (or 0 if no MMAP support)
600 M_MMAP_MAX -4 65536 any (0 disables use of mmap)
601*/
602int __libc_mallopt(int, int);
603libc_hidden_proto (__libc_mallopt)
604
605
606/*
607 mallinfo()
608 Returns (by copy) a struct containing various summary statistics:
609
610 arena: current total non-mmapped bytes allocated from system
611 ordblks: the number of free chunks
612 smblks: the number of fastbin blocks (i.e., small chunks that
613 have been freed but not use resused or consolidated)
614 hblks: current number of mmapped regions
615 hblkhd: total bytes held in mmapped regions
616 usmblks: always 0
617 fsmblks: total bytes held in fastbin blocks
618 uordblks: current total allocated space (normal or mmapped)
619 fordblks: total free space
620 keepcost: the maximum number of bytes that could ideally be released
621 back to system via malloc_trim. ("ideally" means that
622 it ignores page restrictions etc.)
623
624 Because these fields are ints, but internal bookkeeping may
625 be kept as longs, the reported values may wrap around zero and
626 thus be inaccurate.
627*/
628struct mallinfo __libc_mallinfo(void);
629
630
631/*
632 pvalloc(size_t n);
633 Equivalent to valloc(minimum-page-that-holds(n)), that is,
634 round up n to nearest pagesize.
635 */
636void* __libc_pvalloc(size_t);
637
638/*
639 malloc_trim(size_t pad);
640
641 If possible, gives memory back to the system (via negative
642 arguments to sbrk) if there is unused memory at the `high' end of
643 the malloc pool. You can call this after freeing large blocks of
644 memory to potentially reduce the system-level memory requirements
645 of a program. However, it cannot guarantee to reduce memory. Under
646 some allocation patterns, some large free blocks of memory will be
647 locked between two used chunks, so they cannot be given back to
648 the system.
649
650 The `pad' argument to malloc_trim represents the amount of free
651 trailing space to leave untrimmed. If this argument is zero,
652 only the minimum amount of memory to maintain internal data
653 structures will be left (one page or less). Non-zero arguments
654 can be supplied to maintain enough trailing space to service
655 future expected allocations without having to re-obtain memory
656 from the system.
657
658 Malloc_trim returns 1 if it actually released any memory, else 0.
659 On systems that do not support "negative sbrks", it will always
660 return 0.
661*/
662int __malloc_trim(size_t);
663
664/*
665 malloc_usable_size(void* p);
666
667 Returns the number of bytes you can actually use in
668 an allocated chunk, which may be more than you requested (although
669 often not) due to alignment and minimum size constraints.
670 You can use this many bytes without worrying about
671 overwriting other allocated objects. This is not a particularly great
672 programming practice. malloc_usable_size can be more useful in
673 debugging and assertions, for example:
674
675 p = malloc(n);
676 assert(malloc_usable_size(p) >= 256);
677
678*/
679size_t __malloc_usable_size(void*);
680
681/*
682 malloc_stats();
683 Prints on stderr the amount of space obtained from the system (both
684 via sbrk and mmap), the maximum amount (which may be more than
685 current if malloc_trim and/or munmap got called), and the current
686 number of bytes allocated via malloc (or realloc, etc) but not yet
687 freed. Note that this is the number of bytes allocated, not the
688 number requested. It will be larger than the number requested
689 because of alignment and bookkeeping overhead. Because it includes
690 alignment wastage as being in use, this figure may be greater than
691 zero even when no user-level chunks are allocated.
692
693 The reported current and maximum system memory can be inaccurate if
694 a program makes other calls to system memory allocation functions
695 (normally sbrk) outside of malloc.
696
697 malloc_stats prints only the most commonly interesting statistics.
698 More information can be obtained by calling mallinfo.
699
700*/
701void __malloc_stats(void);
702
703/*
704 malloc_get_state(void);
705
706 Returns the state of all malloc variables in an opaque data
707 structure.
708*/
709void* __malloc_get_state(void);
710
711/*
712 malloc_set_state(void* state);
713
714 Restore the state of all malloc variables from data obtained with
715 malloc_get_state().
716*/
717int __malloc_set_state(void*);
718
719/*
720 posix_memalign(void **memptr, size_t alignment, size_t size);
721
722 POSIX wrapper like memalign(), checking for validity of size.
723*/
724int __posix_memalign(void **, size_t, size_t);
725
726/* mallopt tuning options */
727
728/*
729 M_MXFAST is the maximum request size used for "fastbins", special bins
730 that hold returned chunks without consolidating their spaces. This
731 enables future requests for chunks of the same size to be handled
732 very quickly, but can increase fragmentation, and thus increase the
733 overall memory footprint of a program.
734
735 This malloc manages fastbins very conservatively yet still
736 efficiently, so fragmentation is rarely a problem for values less
737 than or equal to the default. The maximum supported value of MXFAST
738 is 80. You wouldn't want it any higher than this anyway. Fastbins
739 are designed especially for use with many small structs, objects or
740 strings -- the default handles structs/objects/arrays with sizes up
741 to 8 4byte fields, or small strings representing words, tokens,
742 etc. Using fastbins for larger objects normally worsens
743 fragmentation without improving speed.
744
745 M_MXFAST is set in REQUEST size units. It is internally used in
746 chunksize units, which adds padding and alignment. You can reduce
747 M_MXFAST to 0 to disable all use of fastbins. This causes the malloc
748 algorithm to be a closer approximation of fifo-best-fit in all cases,
749 not just for larger requests, but will generally cause it to be
750 slower.
751*/
752
753
754/* M_MXFAST is a standard SVID/XPG tuning option, usually listed in malloc.h */
755#ifndef M_MXFAST
756#define M_MXFAST 1
757#endif
758
759#ifndef DEFAULT_MXFAST
760#define DEFAULT_MXFAST (64 * SIZE_SZ / 4)
761#endif
762
763
764/*
765 M_TRIM_THRESHOLD is the maximum amount of unused top-most memory
766 to keep before releasing via malloc_trim in free().
767
768 Automatic trimming is mainly useful in long-lived programs.
769 Because trimming via sbrk can be slow on some systems, and can
770 sometimes be wasteful (in cases where programs immediately
771 afterward allocate more large chunks) the value should be high
772 enough so that your overall system performance would improve by
773 releasing this much memory.
774
775 The trim threshold and the mmap control parameters (see below)
776 can be traded off with one another. Trimming and mmapping are
777 two different ways of releasing unused memory back to the
778 system. Between these two, it is often possible to keep
779 system-level demands of a long-lived program down to a bare
780 minimum. For example, in one test suite of sessions measuring
781 the XF86 X server on Linux, using a trim threshold of 128K and a
782 mmap threshold of 192K led to near-minimal long term resource
783 consumption.
784
785 If you are using this malloc in a long-lived program, it should
786 pay to experiment with these values. As a rough guide, you
787 might set to a value close to the average size of a process
788 (program) running on your system. Releasing this much memory
789 would allow such a process to run in memory. Generally, it's
790 worth it to tune for trimming rather tham memory mapping when a
791 program undergoes phases where several large chunks are
792 allocated and released in ways that can reuse each other's
793 storage, perhaps mixed with phases where there are no such
794 chunks at all. And in well-behaved long-lived programs,
795 controlling release of large blocks via trimming versus mapping
796 is usually faster.
797
798 However, in most programs, these parameters serve mainly as
799 protection against the system-level effects of carrying around
800 massive amounts of unneeded memory. Since frequent calls to
801 sbrk, mmap, and munmap otherwise degrade performance, the default
802 parameters are set to relatively high values that serve only as
803 safeguards.
804
805 The trim value It must be greater than page size to have any useful
806 effect. To disable trimming completely, you can set to
807 (unsigned long)(-1)
808
809 Trim settings interact with fastbin (MXFAST) settings: Unless
810 TRIM_FASTBINS is defined, automatic trimming never takes place upon
811 freeing a chunk with size less than or equal to MXFAST. Trimming is
812 instead delayed until subsequent freeing of larger chunks. However,
813 you can still force an attempted trim by calling malloc_trim.
814
815 Also, trimming is not generally possible in cases where
816 the main arena is obtained via mmap.
817
818 Note that the trick some people use of mallocing a huge space and
819 then freeing it at program startup, in an attempt to reserve system
820 memory, doesn't have the intended effect under automatic trimming,
821 since that memory will immediately be returned to the system.
822*/
823
824#define M_TRIM_THRESHOLD -1
825
826#ifndef DEFAULT_TRIM_THRESHOLD
827#define DEFAULT_TRIM_THRESHOLD (128 * 1024)
828#endif
829
830/*
831 M_TOP_PAD is the amount of extra `padding' space to allocate or
832 retain whenever sbrk is called. It is used in two ways internally:
833
834 * When sbrk is called to extend the top of the arena to satisfy
835 a new malloc request, this much padding is added to the sbrk
836 request.
837
838 * When malloc_trim is called automatically from free(),
839 it is used as the `pad' argument.
840
841 In both cases, the actual amount of padding is rounded
842 so that the end of the arena is always a system page boundary.
843
844 The main reason for using padding is to avoid calling sbrk so
845 often. Having even a small pad greatly reduces the likelihood
846 that nearly every malloc request during program start-up (or
847 after trimming) will invoke sbrk, which needlessly wastes
848 time.
849
850 Automatic rounding-up to page-size units is normally sufficient
851 to avoid measurable overhead, so the default is 0. However, in
852 systems where sbrk is relatively slow, it can pay to increase
853 this value, at the expense of carrying around more memory than
854 the program needs.
855*/
856
857#define M_TOP_PAD -2
858
859#ifndef DEFAULT_TOP_PAD
860#define DEFAULT_TOP_PAD (0)
861#endif
862
863/*
864 MMAP_THRESHOLD_MAX and _MIN are the bounds on the dynamically
865 adjusted MMAP_THRESHOLD.
866*/
867
868#ifndef DEFAULT_MMAP_THRESHOLD_MIN
869#define DEFAULT_MMAP_THRESHOLD_MIN (128 * 1024)
870#endif
871
872#ifndef DEFAULT_MMAP_THRESHOLD_MAX
873 /* For 32-bit platforms we cannot increase the maximum mmap
874 threshold much because it is also the minimum value for the
875 maximum heap size and its alignment. Going above 512k (i.e., 1M
876 for new heaps) wastes too much address space. */
877# if __WORDSIZE == 32
878# define DEFAULT_MMAP_THRESHOLD_MAX (512 * 1024)
879# else
880# define DEFAULT_MMAP_THRESHOLD_MAX (4 * 1024 * 1024 * sizeof(long))
881# endif
882#endif
883
884/*
885 M_MMAP_THRESHOLD is the request size threshold for using mmap()
886 to service a request. Requests of at least this size that cannot
887 be allocated using already-existing space will be serviced via mmap.
888 (If enough normal freed space already exists it is used instead.)
889
890 Using mmap segregates relatively large chunks of memory so that
891 they can be individually obtained and released from the host
892 system. A request serviced through mmap is never reused by any
893 other request (at least not directly; the system may just so
894 happen to remap successive requests to the same locations).
895
896 Segregating space in this way has the benefits that:
897
898 1. Mmapped space can ALWAYS be individually released back
899 to the system, which helps keep the system level memory
900 demands of a long-lived program low.
901 2. Mapped memory can never become `locked' between
902 other chunks, as can happen with normally allocated chunks, which
903 means that even trimming via malloc_trim would not release them.
904 3. On some systems with "holes" in address spaces, mmap can obtain
905 memory that sbrk cannot.
906
907 However, it has the disadvantages that:
908
909 1. The space cannot be reclaimed, consolidated, and then
910 used to service later requests, as happens with normal chunks.
911 2. It can lead to more wastage because of mmap page alignment
912 requirements
913 3. It causes malloc performance to be more dependent on host
914 system memory management support routines which may vary in
915 implementation quality and may impose arbitrary
916 limitations. Generally, servicing a request via normal
917 malloc steps is faster than going through a system's mmap.
918
919 The advantages of mmap nearly always outweigh disadvantages for
920 "large" chunks, but the value of "large" varies across systems. The
921 default is an empirically derived value that works well in most
922 systems.
923
924
925 Update in 2006:
926 The above was written in 2001. Since then the world has changed a lot.
927 Memory got bigger. Applications got bigger. The virtual address space
928 layout in 32 bit linux changed.
929
930 In the new situation, brk() and mmap space is shared and there are no
931 artificial limits on brk size imposed by the kernel. What is more,
932 applications have started using transient allocations larger than the
933 128Kb as was imagined in 2001.
934
935 The price for mmap is also high now; each time glibc mmaps from the
936 kernel, the kernel is forced to zero out the memory it gives to the
937 application. Zeroing memory is expensive and eats a lot of cache and
938 memory bandwidth. This has nothing to do with the efficiency of the
939 virtual memory system, by doing mmap the kernel just has no choice but
940 to zero.
941
942 In 2001, the kernel had a maximum size for brk() which was about 800
943 megabytes on 32 bit x86, at that point brk() would hit the first
944 mmaped shared libaries and couldn't expand anymore. With current 2.6
945 kernels, the VA space layout is different and brk() and mmap
946 both can span the entire heap at will.
947
948 Rather than using a static threshold for the brk/mmap tradeoff,
949 we are now using a simple dynamic one. The goal is still to avoid
950 fragmentation. The old goals we kept are
951 1) try to get the long lived large allocations to use mmap()
952 2) really large allocations should always use mmap()
953 and we're adding now:
954 3) transient allocations should use brk() to avoid forcing the kernel
955 having to zero memory over and over again
956
957 The implementation works with a sliding threshold, which is by default
958 limited to go between 128Kb and 32Mb (64Mb for 64 bitmachines) and starts
959 out at 128Kb as per the 2001 default.
960
961 This allows us to satisfy requirement 1) under the assumption that long
962 lived allocations are made early in the process' lifespan, before it has
963 started doing dynamic allocations of the same size (which will
964 increase the threshold).
965
966 The upperbound on the threshold satisfies requirement 2)
967
968 The threshold goes up in value when the application frees memory that was
969 allocated with the mmap allocator. The idea is that once the application
970 starts freeing memory of a certain size, it's highly probable that this is
971 a size the application uses for transient allocations. This estimator
972 is there to satisfy the new third requirement.
973
974*/
975
976#define M_MMAP_THRESHOLD -3
977
978#ifndef DEFAULT_MMAP_THRESHOLD
979#define DEFAULT_MMAP_THRESHOLD DEFAULT_MMAP_THRESHOLD_MIN
980#endif
981
982/*
983 M_MMAP_MAX is the maximum number of requests to simultaneously
984 service using mmap. This parameter exists because
985 some systems have a limited number of internal tables for
986 use by mmap, and using more than a few of them may degrade
987 performance.
988
989 The default is set to a value that serves only as a safeguard.
990 Setting to 0 disables use of mmap for servicing large requests.
991*/
992
993#define M_MMAP_MAX -4
994
995#ifndef DEFAULT_MMAP_MAX
996#define DEFAULT_MMAP_MAX (65536)
997#endif
998
999#include <malloc.h>
1000
1001#ifndef RETURN_ADDRESS
1002#define RETURN_ADDRESS(X_) (NULL)
1003#endif
1004
1005/* Forward declarations. */
1006struct malloc_chunk;
1007typedef struct malloc_chunk* mchunkptr;
1008
1009/* Internal routines. */
1010
1011static void* _int_malloc(mstate, size_t);
1012static void _int_free(mstate, mchunkptr, int);
1013static void* _int_realloc(mstate, mchunkptr, INTERNAL_SIZE_T,
1014 INTERNAL_SIZE_T);
1015static void* _int_memalign(mstate, size_t, size_t);
1016static void* _mid_memalign(size_t, size_t, void *);
1017
1018static void malloc_printerr(const char *str) __attribute__ ((noreturn));
1019
1020static void* mem2mem_check(void *p, size_t sz);
1021static void top_check(void);
1022static void munmap_chunk(mchunkptr p);
1023#if HAVE_MREMAP
1024static mchunkptr mremap_chunk(mchunkptr p, size_t new_size);
1025#endif
1026
1027static void* malloc_check(size_t sz, const void *caller);
1028static void free_check(void* mem, const void *caller);
1029static void* realloc_check(void* oldmem, size_t bytes,
1030 const void *caller);
1031static void* memalign_check(size_t alignment, size_t bytes,
1032 const void *caller);
1033
1034/* ------------------ MMAP support ------------------ */
1035
1036
1037#include <fcntl.h>
1038#include <sys/mman.h>
1039
1040#if !defined(MAP_ANONYMOUS) && defined(MAP_ANON)
1041# define MAP_ANONYMOUS MAP_ANON
1042#endif
1043
1044#ifndef MAP_NORESERVE
1045# define MAP_NORESERVE 0
1046#endif
1047
1048#define MMAP(addr, size, prot, flags) \
1049 __mmap((addr), (size), (prot), (flags)|MAP_ANONYMOUS|MAP_PRIVATE, -1, 0)
1050
1051
1052/*
1053 ----------------------- Chunk representations -----------------------
1054*/
1055
1056
1057/*
1058 This struct declaration is misleading (but accurate and necessary).
1059 It declares a "view" into memory allowing access to necessary
1060 fields at known offsets from a given base. See explanation below.
1061*/
1062
1063struct malloc_chunk {
1064
1065 INTERNAL_SIZE_T mchunk_prev_size; /* Size of previous chunk (if free). */
1066 INTERNAL_SIZE_T mchunk_size; /* Size in bytes, including overhead. */
1067
1068 struct malloc_chunk* fd; /* double links -- used only if free. */
1069 struct malloc_chunk* bk;
1070
1071 /* Only used for large blocks: pointer to next larger size. */
1072 struct malloc_chunk* fd_nextsize; /* double links -- used only if free. */
1073 struct malloc_chunk* bk_nextsize;
1074};
1075
1076
1077/*
1078 malloc_chunk details:
1079
1080 (The following includes lightly edited explanations by Colin Plumb.)
1081
1082 Chunks of memory are maintained using a `boundary tag' method as
1083 described in e.g., Knuth or Standish. (See the paper by Paul
1084 Wilson ftp://ftp.cs.utexas.edu/pub/garbage/allocsrv.ps for a
1085 survey of such techniques.) Sizes of free chunks are stored both
1086 in the front of each chunk and at the end. This makes
1087 consolidating fragmented chunks into bigger chunks very fast. The
1088 size fields also hold bits representing whether chunks are free or
1089 in use.
1090
1091 An allocated chunk looks like this:
1092
1093
1094 chunk-> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1095 | Size of previous chunk, if unallocated (P clear) |
1096 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1097 | Size of chunk, in bytes |A|M|P|
1098 mem-> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1099 | User data starts here... .
1100 . .
1101 . (malloc_usable_size() bytes) .
1102 . |
1103nextchunk-> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1104 | (size of chunk, but used for application data) |
1105 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1106 | Size of next chunk, in bytes |A|0|1|
1107 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1108
1109 Where "chunk" is the front of the chunk for the purpose of most of
1110 the malloc code, but "mem" is the pointer that is returned to the
1111 user. "Nextchunk" is the beginning of the next contiguous chunk.
1112
1113 Chunks always begin on even word boundaries, so the mem portion
1114 (which is returned to the user) is also on an even word boundary, and
1115 thus at least double-word aligned.
1116
1117 Free chunks are stored in circular doubly-linked lists, and look like this:
1118
1119 chunk-> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1120 | Size of previous chunk, if unallocated (P clear) |
1121 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1122 `head:' | Size of chunk, in bytes |A|0|P|
1123 mem-> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1124 | Forward pointer to next chunk in list |
1125 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1126 | Back pointer to previous chunk in list |
1127 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1128 | Unused space (may be 0 bytes long) .
1129 . .
1130 . |
1131nextchunk-> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1132 `foot:' | Size of chunk, in bytes |
1133 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1134 | Size of next chunk, in bytes |A|0|0|
1135 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1136
1137 The P (PREV_INUSE) bit, stored in the unused low-order bit of the
1138 chunk size (which is always a multiple of two words), is an in-use
1139 bit for the *previous* chunk. If that bit is *clear*, then the
1140 word before the current chunk size contains the previous chunk
1141 size, and can be used to find the front of the previous chunk.
1142 The very first chunk allocated always has this bit set,
1143 preventing access to non-existent (or non-owned) memory. If
1144 prev_inuse is set for any given chunk, then you CANNOT determine
1145 the size of the previous chunk, and might even get a memory
1146 addressing fault when trying to do so.
1147
1148 The A (NON_MAIN_ARENA) bit is cleared for chunks on the initial,
1149 main arena, described by the main_arena variable. When additional
1150 threads are spawned, each thread receives its own arena (up to a
1151 configurable limit, after which arenas are reused for multiple
1152 threads), and the chunks in these arenas have the A bit set. To
1153 find the arena for a chunk on such a non-main arena, heap_for_ptr
1154 performs a bit mask operation and indirection through the ar_ptr
1155 member of the per-heap header heap_info (see arena.c).
1156
1157 Note that the `foot' of the current chunk is actually represented
1158 as the prev_size of the NEXT chunk. This makes it easier to
1159 deal with alignments etc but can be very confusing when trying
1160 to extend or adapt this code.
1161
1162 The three exceptions to all this are:
1163
1164 1. The special chunk `top' doesn't bother using the
1165 trailing size field since there is no next contiguous chunk
1166 that would have to index off it. After initialization, `top'
1167 is forced to always exist. If it would become less than
1168 MINSIZE bytes long, it is replenished.
1169
1170 2. Chunks allocated via mmap, which have the second-lowest-order
1171 bit M (IS_MMAPPED) set in their size fields. Because they are
1172 allocated one-by-one, each must contain its own trailing size
1173 field. If the M bit is set, the other bits are ignored
1174 (because mmapped chunks are neither in an arena, nor adjacent
1175 to a freed chunk). The M bit is also used for chunks which
1176 originally came from a dumped heap via malloc_set_state in
1177 hooks.c.
1178
1179 3. Chunks in fastbins are treated as allocated chunks from the
1180 point of view of the chunk allocator. They are consolidated
1181 with their neighbors only in bulk, in malloc_consolidate.
1182*/
1183
1184/*
1185 ---------- Size and alignment checks and conversions ----------
1186*/
1187
1188/* conversion from malloc headers to user pointers, and back */
1189
1190#define chunk2mem(p) ((void*)((char*)(p) + 2*SIZE_SZ))
1191#define mem2chunk(mem) ((mchunkptr)((char*)(mem) - 2*SIZE_SZ))
1192
1193/* The smallest possible chunk */
1194#define MIN_CHUNK_SIZE (offsetof(struct malloc_chunk, fd_nextsize))
1195
1196/* The smallest size we can malloc is an aligned minimal chunk */
1197
1198#define MINSIZE \
1199 (unsigned long)(((MIN_CHUNK_SIZE+MALLOC_ALIGN_MASK) & ~MALLOC_ALIGN_MASK))
1200
1201/* Check if m has acceptable alignment */
1202
1203#define aligned_OK(m) (((unsigned long)(m) & MALLOC_ALIGN_MASK) == 0)
1204
1205#define misaligned_chunk(p) \
1206 ((uintptr_t)(MALLOC_ALIGNMENT == 2 * SIZE_SZ ? (p) : chunk2mem (p)) \
1207 & MALLOC_ALIGN_MASK)
1208
1209
1210/*
1211 Check if a request is so large that it would wrap around zero when
1212 padded and aligned. To simplify some other code, the bound is made
1213 low enough so that adding MINSIZE will also not wrap around zero.
1214 */
1215
1216#define REQUEST_OUT_OF_RANGE(req) \
1217 ((unsigned long) (req) >= \
1218 (unsigned long) (INTERNAL_SIZE_T) (-2 * MINSIZE))
1219
1220/* pad request bytes into a usable size -- internal version */
1221
1222#define request2size(req) \
1223 (((req) + SIZE_SZ + MALLOC_ALIGN_MASK < MINSIZE) ? \
1224 MINSIZE : \
1225 ((req) + SIZE_SZ + MALLOC_ALIGN_MASK) & ~MALLOC_ALIGN_MASK)
1226
1227/* Same, except also perform argument check */
1228
1229#define checked_request2size(req, sz) \
1230 if (REQUEST_OUT_OF_RANGE (req)) { \
1231 __set_errno (ENOMEM); \
1232 return 0; \
1233 } \
1234 (sz) = request2size (req);
1235
1236/*
1237 --------------- Physical chunk operations ---------------
1238 */
1239
1240
1241/* size field is or'ed with PREV_INUSE when previous adjacent chunk in use */
1242#define PREV_INUSE 0x1
1243
1244/* extract inuse bit of previous chunk */
1245#define prev_inuse(p) ((p)->mchunk_size & PREV_INUSE)
1246
1247
1248/* size field is or'ed with IS_MMAPPED if the chunk was obtained with mmap() */
1249#define IS_MMAPPED 0x2
1250
1251/* check for mmap()'ed chunk */
1252#define chunk_is_mmapped(p) ((p)->mchunk_size & IS_MMAPPED)
1253
1254
1255/* size field is or'ed with NON_MAIN_ARENA if the chunk was obtained
1256 from a non-main arena. This is only set immediately before handing
1257 the chunk to the user, if necessary. */
1258#define NON_MAIN_ARENA 0x4
1259
1260/* Check for chunk from main arena. */
1261#define chunk_main_arena(p) (((p)->mchunk_size & NON_MAIN_ARENA) == 0)
1262
1263/* Mark a chunk as not being on the main arena. */
1264#define set_non_main_arena(p) ((p)->mchunk_size |= NON_MAIN_ARENA)
1265
1266
1267/*
1268 Bits to mask off when extracting size
1269
1270 Note: IS_MMAPPED is intentionally not masked off from size field in
1271 macros for which mmapped chunks should never be seen. This should
1272 cause helpful core dumps to occur if it is tried by accident by
1273 people extending or adapting this malloc.
1274 */
1275#define SIZE_BITS (PREV_INUSE | IS_MMAPPED | NON_MAIN_ARENA)
1276
1277/* Get size, ignoring use bits */
1278#define chunksize(p) (chunksize_nomask (p) & ~(SIZE_BITS))
1279
1280/* Like chunksize, but do not mask SIZE_BITS. */
1281#define chunksize_nomask(p) ((p)->mchunk_size)
1282
1283/* Ptr to next physical malloc_chunk. */
1284#define next_chunk(p) ((mchunkptr) (((char *) (p)) + chunksize (p)))
1285
1286/* Size of the chunk below P. Only valid if prev_inuse (P). */
1287#define prev_size(p) ((p)->mchunk_prev_size)
1288
1289/* Set the size of the chunk below P. Only valid if prev_inuse (P). */
1290#define set_prev_size(p, sz) ((p)->mchunk_prev_size = (sz))
1291
1292/* Ptr to previous physical malloc_chunk. Only valid if prev_inuse (P). */
1293#define prev_chunk(p) ((mchunkptr) (((char *) (p)) - prev_size (p)))
1294
1295/* Treat space at ptr + offset as a chunk */
1296#define chunk_at_offset(p, s) ((mchunkptr) (((char *) (p)) + (s)))
1297
1298/* extract p's inuse bit */
1299#define inuse(p) \
1300 ((((mchunkptr) (((char *) (p)) + chunksize (p)))->mchunk_size) & PREV_INUSE)
1301
1302/* set/clear chunk as being inuse without otherwise disturbing */
1303#define set_inuse(p) \
1304 ((mchunkptr) (((char *) (p)) + chunksize (p)))->mchunk_size |= PREV_INUSE
1305
1306#define clear_inuse(p) \
1307 ((mchunkptr) (((char *) (p)) + chunksize (p)))->mchunk_size &= ~(PREV_INUSE)
1308
1309
1310/* check/set/clear inuse bits in known places */
1311#define inuse_bit_at_offset(p, s) \
1312 (((mchunkptr) (((char *) (p)) + (s)))->mchunk_size & PREV_INUSE)
1313
1314#define set_inuse_bit_at_offset(p, s) \
1315 (((mchunkptr) (((char *) (p)) + (s)))->mchunk_size |= PREV_INUSE)
1316
1317#define clear_inuse_bit_at_offset(p, s) \
1318 (((mchunkptr) (((char *) (p)) + (s)))->mchunk_size &= ~(PREV_INUSE))
1319
1320
1321/* Set size at head, without disturbing its use bit */
1322#define set_head_size(p, s) ((p)->mchunk_size = (((p)->mchunk_size & SIZE_BITS) | (s)))
1323
1324/* Set size/use field */
1325#define set_head(p, s) ((p)->mchunk_size = (s))
1326
1327/* Set size at footer (only when chunk is not in use) */
1328#define set_foot(p, s) (((mchunkptr) ((char *) (p) + (s)))->mchunk_prev_size = (s))
1329
1330
1331#pragma GCC poison mchunk_size
1332#pragma GCC poison mchunk_prev_size
1333
1334/*
1335 -------------------- Internal data structures --------------------
1336
1337 All internal state is held in an instance of malloc_state defined
1338 below. There are no other static variables, except in two optional
1339 cases:
1340 * If USE_MALLOC_LOCK is defined, the mALLOC_MUTEx declared above.
1341 * If mmap doesn't support MAP_ANONYMOUS, a dummy file descriptor
1342 for mmap.
1343
1344 Beware of lots of tricks that minimize the total bookkeeping space
1345 requirements. The result is a little over 1K bytes (for 4byte
1346 pointers and size_t.)
1347 */
1348
1349/*
1350 Bins
1351
1352 An array of bin headers for free chunks. Each bin is doubly
1353 linked. The bins are approximately proportionally (log) spaced.
1354 There are a lot of these bins (128). This may look excessive, but
1355 works very well in practice. Most bins hold sizes that are
1356 unusual as malloc request sizes, but are more usual for fragments
1357 and consolidated sets of chunks, which is what these bins hold, so
1358 they can be found quickly. All procedures maintain the invariant
1359 that no consolidated chunk physically borders another one, so each
1360 chunk in a list is known to be preceeded and followed by either
1361 inuse chunks or the ends of memory.
1362
1363 Chunks in bins are kept in size order, with ties going to the
1364 approximately least recently used chunk. Ordering isn't needed
1365 for the small bins, which all contain the same-sized chunks, but
1366 facilitates best-fit allocation for larger chunks. These lists
1367 are just sequential. Keeping them in order almost never requires
1368 enough traversal to warrant using fancier ordered data
1369 structures.
1370
1371 Chunks of the same size are linked with the most
1372 recently freed at the front, and allocations are taken from the
1373 back. This results in LRU (FIFO) allocation order, which tends
1374 to give each chunk an equal opportunity to be consolidated with
1375 adjacent freed chunks, resulting in larger free chunks and less
1376 fragmentation.
1377
1378 To simplify use in double-linked lists, each bin header acts
1379 as a malloc_chunk. This avoids special-casing for headers.
1380 But to conserve space and improve locality, we allocate
1381 only the fd/bk pointers of bins, and then use repositioning tricks
1382 to treat these as the fields of a malloc_chunk*.
1383 */
1384
1385typedef struct malloc_chunk *mbinptr;
1386
1387/* addressing -- note that bin_at(0) does not exist */
1388#define bin_at(m, i) \
1389 (mbinptr) (((char *) &((m)->bins[((i) - 1) * 2])) \
1390 - offsetof (struct malloc_chunk, fd))
1391
1392/* analog of ++bin */
1393#define next_bin(b) ((mbinptr) ((char *) (b) + (sizeof (mchunkptr) << 1)))
1394
1395/* Reminders about list directionality within bins */
1396#define first(b) ((b)->fd)
1397#define last(b) ((b)->bk)
1398
1399/* Take a chunk off a bin list */
1400#define unlink(AV, P, BK, FD) { \
1401 if (__builtin_expect (chunksize(P) != prev_size (next_chunk(P)), 0)) \
1402 malloc_printerr ("corrupted size vs. prev_size"); \
1403 FD = P->fd; \
1404 BK = P->bk; \
1405 if (__builtin_expect (FD->bk != P || BK->fd != P, 0)) \
1406 malloc_printerr ("corrupted double-linked list"); \
1407 else { \
1408 FD->bk = BK; \
1409 BK->fd = FD; \
1410 if (!in_smallbin_range (chunksize_nomask (P)) \
1411 && __builtin_expect (P->fd_nextsize != NULL, 0)) { \
1412 if (__builtin_expect (P->fd_nextsize->bk_nextsize != P, 0) \
1413 || __builtin_expect (P->bk_nextsize->fd_nextsize != P, 0)) \
1414 malloc_printerr ("corrupted double-linked list (not small)"); \
1415 if (FD->fd_nextsize == NULL) { \
1416 if (P->fd_nextsize == P) \
1417 FD->fd_nextsize = FD->bk_nextsize = FD; \
1418 else { \
1419 FD->fd_nextsize = P->fd_nextsize; \
1420 FD->bk_nextsize = P->bk_nextsize; \
1421 P->fd_nextsize->bk_nextsize = FD; \
1422 P->bk_nextsize->fd_nextsize = FD; \
1423 } \
1424 } else { \
1425 P->fd_nextsize->bk_nextsize = P->bk_nextsize; \
1426 P->bk_nextsize->fd_nextsize = P->fd_nextsize; \
1427 } \
1428 } \
1429 } \
1430}
1431
1432/*
1433 Indexing
1434
1435 Bins for sizes < 512 bytes contain chunks of all the same size, spaced
1436 8 bytes apart. Larger bins are approximately logarithmically spaced:
1437
1438 64 bins of size 8
1439 32 bins of size 64
1440 16 bins of size 512
1441 8 bins of size 4096
1442 4 bins of size 32768
1443 2 bins of size 262144
1444 1 bin of size what's left
1445
1446 There is actually a little bit of slop in the numbers in bin_index
1447 for the sake of speed. This makes no difference elsewhere.
1448
1449 The bins top out around 1MB because we expect to service large
1450 requests via mmap.
1451
1452 Bin 0 does not exist. Bin 1 is the unordered list; if that would be
1453 a valid chunk size the small bins are bumped up one.
1454 */
1455
1456#define NBINS 128
1457#define NSMALLBINS 64
1458#define SMALLBIN_WIDTH MALLOC_ALIGNMENT
1459#define SMALLBIN_CORRECTION (MALLOC_ALIGNMENT > 2 * SIZE_SZ)
1460#define MIN_LARGE_SIZE ((NSMALLBINS - SMALLBIN_CORRECTION) * SMALLBIN_WIDTH)
1461
1462#define in_smallbin_range(sz) \
1463 ((unsigned long) (sz) < (unsigned long) MIN_LARGE_SIZE)
1464
1465#define smallbin_index(sz) \
1466 ((SMALLBIN_WIDTH == 16 ? (((unsigned) (sz)) >> 4) : (((unsigned) (sz)) >> 3))\
1467 + SMALLBIN_CORRECTION)
1468
1469#define largebin_index_32(sz) \
1470 (((((unsigned long) (sz)) >> 6) <= 38) ? 56 + (((unsigned long) (sz)) >> 6) :\
1471 ((((unsigned long) (sz)) >> 9) <= 20) ? 91 + (((unsigned long) (sz)) >> 9) :\
1472 ((((unsigned long) (sz)) >> 12) <= 10) ? 110 + (((unsigned long) (sz)) >> 12) :\
1473 ((((unsigned long) (sz)) >> 15) <= 4) ? 119 + (((unsigned long) (sz)) >> 15) :\
1474 ((((unsigned long) (sz)) >> 18) <= 2) ? 124 + (((unsigned long) (sz)) >> 18) :\
1475 126)
1476
1477#define largebin_index_32_big(sz) \
1478 (((((unsigned long) (sz)) >> 6) <= 45) ? 49 + (((unsigned long) (sz)) >> 6) :\
1479 ((((unsigned long) (sz)) >> 9) <= 20) ? 91 + (((unsigned long) (sz)) >> 9) :\
1480 ((((unsigned long) (sz)) >> 12) <= 10) ? 110 + (((unsigned long) (sz)) >> 12) :\
1481 ((((unsigned long) (sz)) >> 15) <= 4) ? 119 + (((unsigned long) (sz)) >> 15) :\
1482 ((((unsigned long) (sz)) >> 18) <= 2) ? 124 + (((unsigned long) (sz)) >> 18) :\
1483 126)
1484
1485// XXX It remains to be seen whether it is good to keep the widths of
1486// XXX the buckets the same or whether it should be scaled by a factor
1487// XXX of two as well.
1488#define largebin_index_64(sz) \
1489 (((((unsigned long) (sz)) >> 6) <= 48) ? 48 + (((unsigned long) (sz)) >> 6) :\
1490 ((((unsigned long) (sz)) >> 9) <= 20) ? 91 + (((unsigned long) (sz)) >> 9) :\
1491 ((((unsigned long) (sz)) >> 12) <= 10) ? 110 + (((unsigned long) (sz)) >> 12) :\
1492 ((((unsigned long) (sz)) >> 15) <= 4) ? 119 + (((unsigned long) (sz)) >> 15) :\
1493 ((((unsigned long) (sz)) >> 18) <= 2) ? 124 + (((unsigned long) (sz)) >> 18) :\
1494 126)
1495
1496#define largebin_index(sz) \
1497 (SIZE_SZ == 8 ? largebin_index_64 (sz) \
1498 : MALLOC_ALIGNMENT == 16 ? largebin_index_32_big (sz) \
1499 : largebin_index_32 (sz))
1500
1501#define bin_index(sz) \
1502 ((in_smallbin_range (sz)) ? smallbin_index (sz) : largebin_index (sz))
1503
1504
1505/*
1506 Unsorted chunks
1507
1508 All remainders from chunk splits, as well as all returned chunks,
1509 are first placed in the "unsorted" bin. They are then placed
1510 in regular bins after malloc gives them ONE chance to be used before
1511 binning. So, basically, the unsorted_chunks list acts as a queue,
1512 with chunks being placed on it in free (and malloc_consolidate),
1513 and taken off (to be either used or placed in bins) in malloc.
1514
1515 The NON_MAIN_ARENA flag is never set for unsorted chunks, so it
1516 does not have to be taken into account in size comparisons.
1517 */
1518
1519/* The otherwise unindexable 1-bin is used to hold unsorted chunks. */
1520#define unsorted_chunks(M) (bin_at (M, 1))
1521
1522/*
1523 Top
1524
1525 The top-most available chunk (i.e., the one bordering the end of
1526 available memory) is treated specially. It is never included in
1527 any bin, is used only if no other chunk is available, and is
1528 released back to the system if it is very large (see
1529 M_TRIM_THRESHOLD). Because top initially
1530 points to its own bin with initial zero size, thus forcing
1531 extension on the first malloc request, we avoid having any special
1532 code in malloc to check whether it even exists yet. But we still
1533 need to do so when getting memory from system, so we make
1534 initial_top treat the bin as a legal but unusable chunk during the
1535 interval between initialization and the first call to
1536 sysmalloc. (This is somewhat delicate, since it relies on
1537 the 2 preceding words to be zero during this interval as well.)
1538 */
1539
1540/* Conveniently, the unsorted bin can be used as dummy top on first call */
1541#define initial_top(M) (unsorted_chunks (M))
1542
1543/*
1544 Binmap
1545
1546 To help compensate for the large number of bins, a one-level index
1547 structure is used for bin-by-bin searching. `binmap' is a
1548 bitvector recording whether bins are definitely empty so they can
1549 be skipped over during during traversals. The bits are NOT always
1550 cleared as soon as bins are empty, but instead only
1551 when they are noticed to be empty during traversal in malloc.
1552 */
1553
1554/* Conservatively use 32 bits per map word, even if on 64bit system */
1555#define BINMAPSHIFT 5
1556#define BITSPERMAP (1U << BINMAPSHIFT)
1557#define BINMAPSIZE (NBINS / BITSPERMAP)
1558
1559#define idx2block(i) ((i) >> BINMAPSHIFT)
1560#define idx2bit(i) ((1U << ((i) & ((1U << BINMAPSHIFT) - 1))))
1561
1562#define mark_bin(m, i) ((m)->binmap[idx2block (i)] |= idx2bit (i))
1563#define unmark_bin(m, i) ((m)->binmap[idx2block (i)] &= ~(idx2bit (i)))
1564#define get_binmap(m, i) ((m)->binmap[idx2block (i)] & idx2bit (i))
1565
1566/*
1567 Fastbins
1568
1569 An array of lists holding recently freed small chunks. Fastbins
1570 are not doubly linked. It is faster to single-link them, and
1571 since chunks are never removed from the middles of these lists,
1572 double linking is not necessary. Also, unlike regular bins, they
1573 are not even processed in FIFO order (they use faster LIFO) since
1574 ordering doesn't much matter in the transient contexts in which
1575 fastbins are normally used.
1576
1577 Chunks in fastbins keep their inuse bit set, so they cannot
1578 be consolidated with other free chunks. malloc_consolidate
1579 releases all chunks in fastbins and consolidates them with
1580 other free chunks.
1581 */
1582
1583typedef struct malloc_chunk *mfastbinptr;
1584#define fastbin(ar_ptr, idx) ((ar_ptr)->fastbinsY[idx])
1585
1586/* offset 2 to use otherwise unindexable first 2 bins */
1587#define fastbin_index(sz) \
1588 ((((unsigned int) (sz)) >> (SIZE_SZ == 8 ? 4 : 3)) - 2)
1589
1590
1591/* The maximum fastbin request size we support */
1592#define MAX_FAST_SIZE (80 * SIZE_SZ / 4)
1593
1594#define NFASTBINS (fastbin_index (request2size (MAX_FAST_SIZE)) + 1)
1595
1596/*
1597 FASTBIN_CONSOLIDATION_THRESHOLD is the size of a chunk in free()
1598 that triggers automatic consolidation of possibly-surrounding
1599 fastbin chunks. This is a heuristic, so the exact value should not
1600 matter too much. It is defined at half the default trim threshold as a
1601 compromise heuristic to only attempt consolidation if it is likely
1602 to lead to trimming. However, it is not dynamically tunable, since
1603 consolidation reduces fragmentation surrounding large chunks even
1604 if trimming is not used.
1605 */
1606
1607#define FASTBIN_CONSOLIDATION_THRESHOLD (65536UL)
1608
1609/*
1610 NONCONTIGUOUS_BIT indicates that MORECORE does not return contiguous
1611 regions. Otherwise, contiguity is exploited in merging together,
1612 when possible, results from consecutive MORECORE calls.
1613
1614 The initial value comes from MORECORE_CONTIGUOUS, but is
1615 changed dynamically if mmap is ever used as an sbrk substitute.
1616 */
1617
1618#define NONCONTIGUOUS_BIT (2U)
1619
1620#define contiguous(M) (((M)->flags & NONCONTIGUOUS_BIT) == 0)
1621#define noncontiguous(M) (((M)->flags & NONCONTIGUOUS_BIT) != 0)
1622#define set_noncontiguous(M) ((M)->flags |= NONCONTIGUOUS_BIT)
1623#define set_contiguous(M) ((M)->flags &= ~NONCONTIGUOUS_BIT)
1624
1625/* Maximum size of memory handled in fastbins. */
1626static INTERNAL_SIZE_T global_max_fast;
1627
1628/*
1629 Set value of max_fast.
1630 Use impossibly small value if 0.
1631 Precondition: there are no existing fastbin chunks in the main arena.
1632 Since do_check_malloc_state () checks this, we call malloc_consolidate ()
1633 before changing max_fast. Note other arenas will leak their fast bin
1634 entries if max_fast is reduced.
1635 */
1636
1637#define set_max_fast(s) \
1638 global_max_fast = (((s) == 0) \
1639 ? SMALLBIN_WIDTH : ((s + SIZE_SZ) & ~MALLOC_ALIGN_MASK))
1640
1641static inline INTERNAL_SIZE_T
1642get_max_fast (void)
1643{
1644 /* Tell the GCC optimizers that global_max_fast is never larger
1645 than MAX_FAST_SIZE. This avoids out-of-bounds array accesses in
1646 _int_malloc after constant propagation of the size parameter.
1647 (The code never executes because malloc preserves the
1648 global_max_fast invariant, but the optimizers may not recognize
1649 this.) */
1650 if (global_max_fast > MAX_FAST_SIZE)
1651 __builtin_unreachable ();
1652 return global_max_fast;
1653}
1654
1655/*
1656 ----------- Internal state representation and initialization -----------
1657 */
1658
1659/*
1660 have_fastchunks indicates that there are probably some fastbin chunks.
1661 It is set true on entering a chunk into any fastbin, and cleared early in
1662 malloc_consolidate. The value is approximate since it may be set when there
1663 are no fastbin chunks, or it may be clear even if there are fastbin chunks
1664 available. Given it's sole purpose is to reduce number of redundant calls to
1665 malloc_consolidate, it does not affect correctness. As a result we can safely
1666 use relaxed atomic accesses.
1667 */
1668
1669
1670struct malloc_state
1671{
1672 /* Serialize access. */
1673 __libc_lock_define (, mutex);
1674
1675 /* Flags (formerly in max_fast). */
1676 int flags;
1677
1678 /* Set if the fastbin chunks contain recently inserted free blocks. */
1679 /* Note this is a bool but not all targets support atomics on booleans. */
1680 int have_fastchunks;
1681
1682 /* Fastbins */
1683 mfastbinptr fastbinsY[NFASTBINS];
1684
1685 /* Base of the topmost chunk -- not otherwise kept in a bin */
1686 mchunkptr top;
1687
1688 /* The remainder from the most recent split of a small request */
1689 mchunkptr last_remainder;
1690
1691 /* Normal bins packed as described above */
1692 mchunkptr bins[NBINS * 2 - 2];
1693
1694 /* Bitmap of bins */
1695 unsigned int binmap[BINMAPSIZE];
1696
1697 /* Linked list */
1698 struct malloc_state *next;
1699
1700 /* Linked list for free arenas. Access to this field is serialized
1701 by free_list_lock in arena.c. */
1702 struct malloc_state *next_free;
1703
1704 /* Number of threads attached to this arena. 0 if the arena is on
1705 the free list. Access to this field is serialized by
1706 free_list_lock in arena.c. */
1707 INTERNAL_SIZE_T attached_threads;
1708
1709 /* Memory allocated from the system in this arena. */
1710 INTERNAL_SIZE_T system_mem;
1711 INTERNAL_SIZE_T max_system_mem;
1712};
1713
1714struct malloc_par
1715{
1716 /* Tunable parameters */
1717 unsigned long trim_threshold;
1718 INTERNAL_SIZE_T top_pad;
1719 INTERNAL_SIZE_T mmap_threshold;
1720 INTERNAL_SIZE_T arena_test;
1721 INTERNAL_SIZE_T arena_max;
1722
1723 /* Memory map support */
1724 int n_mmaps;
1725 int n_mmaps_max;
1726 int max_n_mmaps;
1727 /* the mmap_threshold is dynamic, until the user sets
1728 it manually, at which point we need to disable any
1729 dynamic behavior. */
1730 int no_dyn_threshold;
1731
1732 /* Statistics */
1733 INTERNAL_SIZE_T mmapped_mem;
1734 INTERNAL_SIZE_T max_mmapped_mem;
1735
1736 /* First address handed out by MORECORE/sbrk. */
1737 char *sbrk_base;
1738
1739#if USE_TCACHE
1740 /* Maximum number of buckets to use. */
1741 size_t tcache_bins;
1742 size_t tcache_max_bytes;
1743 /* Maximum number of chunks in each bucket. */
1744 size_t tcache_count;
1745 /* Maximum number of chunks to remove from the unsorted list, which
1746 aren't used to prefill the cache. */
1747 size_t tcache_unsorted_limit;
1748#endif
1749};
1750
1751/* There are several instances of this struct ("arenas") in this
1752 malloc. If you are adapting this malloc in a way that does NOT use
1753 a static or mmapped malloc_state, you MUST explicitly zero-fill it
1754 before using. This malloc relies on the property that malloc_state
1755 is initialized to all zeroes (as is true of C statics). */
1756
1757static struct malloc_state main_arena =
1758{
1759 .mutex = _LIBC_LOCK_INITIALIZER,
1760 .next = &main_arena,
1761 .attached_threads = 1
1762};
1763
1764/* These variables are used for undumping support. Chunked are marked
1765 as using mmap, but we leave them alone if they fall into this
1766 range. NB: The chunk size for these chunks only includes the
1767 initial size field (of SIZE_SZ bytes), there is no trailing size
1768 field (unlike with regular mmapped chunks). */
1769static mchunkptr dumped_main_arena_start; /* Inclusive. */
1770static mchunkptr dumped_main_arena_end; /* Exclusive. */
1771
1772/* True if the pointer falls into the dumped arena. Use this after
1773 chunk_is_mmapped indicates a chunk is mmapped. */
1774#define DUMPED_MAIN_ARENA_CHUNK(p) \
1775 ((p) >= dumped_main_arena_start && (p) < dumped_main_arena_end)
1776
1777/* There is only one instance of the malloc parameters. */
1778
1779static struct malloc_par mp_ =
1780{
1781 .top_pad = DEFAULT_TOP_PAD,
1782 .n_mmaps_max = DEFAULT_MMAP_MAX,
1783 .mmap_threshold = DEFAULT_MMAP_THRESHOLD,
1784 .trim_threshold = DEFAULT_TRIM_THRESHOLD,
1785#define NARENAS_FROM_NCORES(n) ((n) * (sizeof (long) == 4 ? 2 : 8))
1786 .arena_test = NARENAS_FROM_NCORES (1)
1787#if USE_TCACHE
1788 ,
1789 .tcache_count = TCACHE_FILL_COUNT,
1790 .tcache_bins = TCACHE_MAX_BINS,
1791 .tcache_max_bytes = tidx2usize (TCACHE_MAX_BINS-1),
1792 .tcache_unsorted_limit = 0 /* No limit. */
1793#endif
1794};
1795
1796/*
1797 Initialize a malloc_state struct.
1798
1799 This is called from ptmalloc_init () or from _int_new_arena ()
1800 when creating a new arena.
1801 */
1802
1803static void
1804malloc_init_state (mstate av)
1805{
1806 int i;
1807 mbinptr bin;
1808
1809 /* Establish circular links for normal bins */
1810 for (i = 1; i < NBINS; ++i)
1811 {
1812 bin = bin_at (av, i);
1813 bin->fd = bin->bk = bin;
1814 }
1815
1816#if MORECORE_CONTIGUOUS
1817 if (av != &main_arena)
1818#endif
1819 set_noncontiguous (av);
1820 if (av == &main_arena)
1821 set_max_fast (DEFAULT_MXFAST);
1822 atomic_store_relaxed (&av->have_fastchunks, false);
1823
1824 av->top = initial_top (av);
1825}
1826
1827/*
1828 Other internal utilities operating on mstates
1829 */
1830
1831static void *sysmalloc (INTERNAL_SIZE_T, mstate);
1832static int systrim (size_t, mstate);
1833static void malloc_consolidate (mstate);
1834
1835
1836/* -------------- Early definitions for debugging hooks ---------------- */
1837
1838/* Define and initialize the hook variables. These weak definitions must
1839 appear before any use of the variables in a function (arena.c uses one). */
1840#ifndef weak_variable
1841/* In GNU libc we want the hook variables to be weak definitions to
1842 avoid a problem with Emacs. */
1843# define weak_variable weak_function
1844#endif
1845
1846/* Forward declarations. */
1847static void *malloc_hook_ini (size_t sz,
1848 const void *caller) __THROW;
1849static void *realloc_hook_ini (void *ptr, size_t sz,
1850 const void *caller) __THROW;
1851static void *memalign_hook_ini (size_t alignment, size_t sz,
1852 const void *caller) __THROW;
1853
1854#if HAVE_MALLOC_INIT_HOOK
1855void weak_variable (*__malloc_initialize_hook) (void) = NULL;
1856compat_symbol (libc, __malloc_initialize_hook,
1857 __malloc_initialize_hook, GLIBC_2_0);
1858#endif
1859
1860void weak_variable (*__free_hook) (void *__ptr,
1861 const void *) = NULL;
1862void *weak_variable (*__malloc_hook)
1863 (size_t __size, const void *) = malloc_hook_ini;
1864void *weak_variable (*__realloc_hook)
1865 (void *__ptr, size_t __size, const void *)
1866 = realloc_hook_ini;
1867void *weak_variable (*__memalign_hook)
1868 (size_t __alignment, size_t __size, const void *)
1869 = memalign_hook_ini;
1870void weak_variable (*__after_morecore_hook) (void) = NULL;
1871
1872
1873/* ------------------ Testing support ----------------------------------*/
1874
1875static int perturb_byte;
1876
1877static void
1878alloc_perturb (char *p, size_t n)
1879{
1880 if (__glibc_unlikely (perturb_byte))
1881 memset (p, perturb_byte ^ 0xff, n);
1882}
1883
1884static void
1885free_perturb (char *p, size_t n)
1886{
1887 if (__glibc_unlikely (perturb_byte))
1888 memset (p, perturb_byte, n);
1889}
1890
1891
1892
1893#include <stap-probe.h>
1894
1895/* ------------------- Support for multiple arenas -------------------- */
1896#include "arena.c"
1897
1898/*
1899 Debugging support
1900
1901 These routines make a number of assertions about the states
1902 of data structures that should be true at all times. If any
1903 are not true, it's very likely that a user program has somehow
1904 trashed memory. (It's also possible that there is a coding error
1905 in malloc. In which case, please report it!)
1906 */
1907
1908#if !MALLOC_DEBUG
1909
1910# define check_chunk(A, P)
1911# define check_free_chunk(A, P)
1912# define check_inuse_chunk(A, P)
1913# define check_remalloced_chunk(A, P, N)
1914# define check_malloced_chunk(A, P, N)
1915# define check_malloc_state(A)
1916
1917#else
1918
1919# define check_chunk(A, P) do_check_chunk (A, P)
1920# define check_free_chunk(A, P) do_check_free_chunk (A, P)
1921# define check_inuse_chunk(A, P) do_check_inuse_chunk (A, P)
1922# define check_remalloced_chunk(A, P, N) do_check_remalloced_chunk (A, P, N)
1923# define check_malloced_chunk(A, P, N) do_check_malloced_chunk (A, P, N)
1924# define check_malloc_state(A) do_check_malloc_state (A)
1925
1926/*
1927 Properties of all chunks
1928 */
1929
1930static void
1931do_check_chunk (mstate av, mchunkptr p)
1932{
1933 unsigned long sz = chunksize (p);
1934 /* min and max possible addresses assuming contiguous allocation */
1935 char *max_address = (char *) (av->top) + chunksize (av->top);
1936 char *min_address = max_address - av->system_mem;
1937
1938 if (!chunk_is_mmapped (p))
1939 {
1940 /* Has legal address ... */
1941 if (p != av->top)
1942 {
1943 if (contiguous (av))
1944 {
1945 assert (((char *) p) >= min_address);
1946 assert (((char *) p + sz) <= ((char *) (av->top)));
1947 }
1948 }
1949 else
1950 {
1951 /* top size is always at least MINSIZE */
1952 assert ((unsigned long) (sz) >= MINSIZE);
1953 /* top predecessor always marked inuse */
1954 assert (prev_inuse (p));
1955 }
1956 }
1957 else if (!DUMPED_MAIN_ARENA_CHUNK (p))
1958 {
1959 /* address is outside main heap */
1960 if (contiguous (av) && av->top != initial_top (av))
1961 {
1962 assert (((char *) p) < min_address || ((char *) p) >= max_address);
1963 }
1964 /* chunk is page-aligned */
1965 assert (((prev_size (p) + sz) & (GLRO (dl_pagesize) - 1)) == 0);
1966 /* mem is aligned */
1967 assert (aligned_OK (chunk2mem (p)));
1968 }
1969}
1970
1971/*
1972 Properties of free chunks
1973 */
1974
1975static void
1976do_check_free_chunk (mstate av, mchunkptr p)
1977{
1978 INTERNAL_SIZE_T sz = chunksize_nomask (p) & ~(PREV_INUSE | NON_MAIN_ARENA);
1979 mchunkptr next = chunk_at_offset (p, sz);
1980
1981 do_check_chunk (av, p);
1982
1983 /* Chunk must claim to be free ... */
1984 assert (!inuse (p));
1985 assert (!chunk_is_mmapped (p));
1986
1987 /* Unless a special marker, must have OK fields */
1988 if ((unsigned long) (sz) >= MINSIZE)
1989 {
1990 assert ((sz & MALLOC_ALIGN_MASK) == 0);
1991 assert (aligned_OK (chunk2mem (p)));
1992 /* ... matching footer field */
1993 assert (prev_size (next_chunk (p)) == sz);
1994 /* ... and is fully consolidated */
1995 assert (prev_inuse (p));
1996 assert (next == av->top || inuse (next));
1997
1998 /* ... and has minimally sane links */
1999 assert (p->fd->bk == p);
2000 assert (p->bk->fd == p);
2001 }
2002 else /* markers are always of size SIZE_SZ */
2003 assert (sz == SIZE_SZ);
2004}
2005
2006/*
2007 Properties of inuse chunks
2008 */
2009
2010static void
2011do_check_inuse_chunk (mstate av, mchunkptr p)
2012{
2013 mchunkptr next;
2014
2015 do_check_chunk (av, p);
2016
2017 if (chunk_is_mmapped (p))
2018 return; /* mmapped chunks have no next/prev */
2019
2020 /* Check whether it claims to be in use ... */
2021 assert (inuse (p));
2022
2023 next = next_chunk (p);
2024
2025 /* ... and is surrounded by OK chunks.
2026 Since more things can be checked with free chunks than inuse ones,
2027 if an inuse chunk borders them and debug is on, it's worth doing them.
2028 */
2029 if (!prev_inuse (p))
2030 {
2031 /* Note that we cannot even look at prev unless it is not inuse */
2032 mchunkptr prv = prev_chunk (p);
2033 assert (next_chunk (prv) == p);
2034 do_check_free_chunk (av, prv);
2035 }
2036
2037 if (next == av->top)
2038 {
2039 assert (prev_inuse (next));
2040 assert (chunksize (next) >= MINSIZE);
2041 }
2042 else if (!inuse (next))
2043 do_check_free_chunk (av, next);
2044}
2045
2046/*
2047 Properties of chunks recycled from fastbins
2048 */
2049
2050static void
2051do_check_remalloced_chunk (mstate av, mchunkptr p, INTERNAL_SIZE_T s)
2052{
2053 INTERNAL_SIZE_T sz = chunksize_nomask (p) & ~(PREV_INUSE | NON_MAIN_ARENA);
2054
2055 if (!chunk_is_mmapped (p))
2056 {
2057 assert (av == arena_for_chunk (p));
2058 if (chunk_main_arena (p))
2059 assert (av == &main_arena);
2060 else
2061 assert (av != &main_arena);
2062 }
2063
2064 do_check_inuse_chunk (av, p);
2065
2066 /* Legal size ... */
2067 assert ((sz & MALLOC_ALIGN_MASK) == 0);
2068 assert ((unsigned long) (sz) >= MINSIZE);
2069 /* ... and alignment */
2070 assert (aligned_OK (chunk2mem (p)));
2071 /* chunk is less than MINSIZE more than request */
2072 assert ((long) (sz) - (long) (s) >= 0);
2073 assert ((long) (sz) - (long) (s + MINSIZE) < 0);
2074}
2075
2076/*
2077 Properties of nonrecycled chunks at the point they are malloced
2078 */
2079
2080static void
2081do_check_malloced_chunk (mstate av, mchunkptr p, INTERNAL_SIZE_T s)
2082{
2083 /* same as recycled case ... */
2084 do_check_remalloced_chunk (av, p, s);
2085
2086 /*
2087 ... plus, must obey implementation invariant that prev_inuse is
2088 always true of any allocated chunk; i.e., that each allocated
2089 chunk borders either a previously allocated and still in-use
2090 chunk, or the base of its memory arena. This is ensured
2091 by making all allocations from the `lowest' part of any found
2092 chunk. This does not necessarily hold however for chunks
2093 recycled via fastbins.
2094 */
2095
2096 assert (prev_inuse (p));
2097}
2098
2099
2100/*
2101 Properties of malloc_state.
2102
2103 This may be useful for debugging malloc, as well as detecting user
2104 programmer errors that somehow write into malloc_state.
2105
2106 If you are extending or experimenting with this malloc, you can
2107 probably figure out how to hack this routine to print out or
2108 display chunk addresses, sizes, bins, and other instrumentation.
2109 */
2110
2111static void
2112do_check_malloc_state (mstate av)
2113{
2114 int i;
2115 mchunkptr p;
2116 mchunkptr q;
2117 mbinptr b;
2118 unsigned int idx;
2119 INTERNAL_SIZE_T size;
2120 unsigned long total = 0;
2121 int max_fast_bin;
2122
2123 /* internal size_t must be no wider than pointer type */
2124 assert (sizeof (INTERNAL_SIZE_T) <= sizeof (char *));
2125
2126 /* alignment is a power of 2 */
2127 assert ((MALLOC_ALIGNMENT & (MALLOC_ALIGNMENT - 1)) == 0);
2128
2129 /* Check the arena is initialized. */
2130 assert (av->top != 0);
2131
2132 /* No memory has been allocated yet, so doing more tests is not possible. */
2133 if (av->top == initial_top (av))
2134 return;
2135
2136 /* pagesize is a power of 2 */
2137 assert (powerof2(GLRO (dl_pagesize)));
2138
2139 /* A contiguous main_arena is consistent with sbrk_base. */
2140 if (av == &main_arena && contiguous (av))
2141 assert ((char *) mp_.sbrk_base + av->system_mem ==
2142 (char *) av->top + chunksize (av->top));
2143
2144 /* properties of fastbins */
2145
2146 /* max_fast is in allowed range */
2147 assert ((get_max_fast () & ~1) <= request2size (MAX_FAST_SIZE));
2148
2149 max_fast_bin = fastbin_index (get_max_fast ());
2150
2151 for (i = 0; i < NFASTBINS; ++i)
2152 {
2153 p = fastbin (av, i);
2154
2155 /* The following test can only be performed for the main arena.
2156 While mallopt calls malloc_consolidate to get rid of all fast
2157 bins (especially those larger than the new maximum) this does
2158 only happen for the main arena. Trying to do this for any
2159 other arena would mean those arenas have to be locked and
2160 malloc_consolidate be called for them. This is excessive. And
2161 even if this is acceptable to somebody it still cannot solve
2162 the problem completely since if the arena is locked a
2163 concurrent malloc call might create a new arena which then
2164 could use the newly invalid fast bins. */
2165
2166 /* all bins past max_fast are empty */
2167 if (av == &main_arena && i > max_fast_bin)
2168 assert (p == 0);
2169
2170 while (p != 0)
2171 {
2172 /* each chunk claims to be inuse */
2173 do_check_inuse_chunk (av, p);
2174 total += chunksize (p);
2175 /* chunk belongs in this bin */
2176 assert (fastbin_index (chunksize (p)) == i);
2177 p = p->fd;
2178 }
2179 }
2180
2181 /* check normal bins */
2182 for (i = 1; i < NBINS; ++i)
2183 {
2184 b = bin_at (av, i);
2185
2186 /* binmap is accurate (except for bin 1 == unsorted_chunks) */
2187 if (i >= 2)
2188 {
2189 unsigned int binbit = get_binmap (av, i);
2190 int empty = last (b) == b;
2191 if (!binbit)
2192 assert (empty);
2193 else if (!empty)
2194 assert (binbit);
2195 }
2196
2197 for (p = last (b); p != b; p = p->bk)
2198 {
2199 /* each chunk claims to be free */
2200 do_check_free_chunk (av, p);
2201 size = chunksize (p);
2202 total += size;
2203 if (i >= 2)
2204 {
2205 /* chunk belongs in bin */
2206 idx = bin_index (size);
2207 assert (idx == i);
2208 /* lists are sorted */
2209 assert (p->bk == b ||
2210 (unsigned long) chunksize (p->bk) >= (unsigned long) chunksize (p));
2211
2212 if (!in_smallbin_range (size))
2213 {
2214 if (p->fd_nextsize != NULL)
2215 {
2216 if (p->fd_nextsize == p)
2217 assert (p->bk_nextsize == p);
2218 else
2219 {
2220 if (p->fd_nextsize == first (b))
2221 assert (chunksize (p) < chunksize (p->fd_nextsize));
2222 else
2223 assert (chunksize (p) > chunksize (p->fd_nextsize));
2224
2225 if (p == first (b))
2226 assert (chunksize (p) > chunksize (p->bk_nextsize));
2227 else
2228 assert (chunksize (p) < chunksize (p->bk_nextsize));
2229 }
2230 }
2231 else
2232 assert (p->bk_nextsize == NULL);
2233 }
2234 }
2235 else if (!in_smallbin_range (size))
2236 assert (p->fd_nextsize == NULL && p->bk_nextsize == NULL);
2237 /* chunk is followed by a legal chain of inuse chunks */
2238 for (q = next_chunk (p);
2239 (q != av->top && inuse (q) &&
2240 (unsigned long) (chunksize (q)) >= MINSIZE);
2241 q = next_chunk (q))
2242 do_check_inuse_chunk (av, q);
2243 }
2244 }
2245
2246 /* top chunk is OK */
2247 check_chunk (av, av->top);
2248}
2249#endif
2250
2251
2252/* ----------------- Support for debugging hooks -------------------- */
2253#include "hooks.c"
2254
2255
2256/* ----------- Routines dealing with system allocation -------------- */
2257
2258/*
2259 sysmalloc handles malloc cases requiring more memory from the system.
2260 On entry, it is assumed that av->top does not have enough
2261 space to service request for nb bytes, thus requiring that av->top
2262 be extended or replaced.
2263 */
2264
2265static void *
2266sysmalloc (INTERNAL_SIZE_T nb, mstate av)
2267{
2268 mchunkptr old_top; /* incoming value of av->top */
2269 INTERNAL_SIZE_T old_size; /* its size */
2270 char *old_end; /* its end address */
2271
2272 long size; /* arg to first MORECORE or mmap call */
2273 char *brk; /* return value from MORECORE */
2274
2275 long correction; /* arg to 2nd MORECORE call */
2276 char *snd_brk; /* 2nd return val */
2277
2278 INTERNAL_SIZE_T front_misalign; /* unusable bytes at front of new space */
2279 INTERNAL_SIZE_T end_misalign; /* partial page left at end of new space */
2280 char *aligned_brk; /* aligned offset into brk */
2281
2282 mchunkptr p; /* the allocated/returned chunk */
2283 mchunkptr remainder; /* remainder from allocation */
2284 unsigned long remainder_size; /* its size */
2285
2286
2287 size_t pagesize = GLRO (dl_pagesize);
2288 bool tried_mmap = false;
2289
2290
2291 /*
2292 If have mmap, and the request size meets the mmap threshold, and
2293 the system supports mmap, and there are few enough currently
2294 allocated mmapped regions, try to directly map this request
2295 rather than expanding top.
2296 */
2297
2298 if (av == NULL
2299 || ((unsigned long) (nb) >= (unsigned long) (mp_.mmap_threshold)
2300 && (mp_.n_mmaps < mp_.n_mmaps_max)))
2301 {
2302 char *mm; /* return value from mmap call*/
2303
2304 try_mmap:
2305 /*
2306 Round up size to nearest page. For mmapped chunks, the overhead
2307 is one SIZE_SZ unit larger than for normal chunks, because there
2308 is no following chunk whose prev_size field could be used.
2309
2310 See the front_misalign handling below, for glibc there is no
2311 need for further alignments unless we have have high alignment.
2312 */
2313 if (MALLOC_ALIGNMENT == 2 * SIZE_SZ)
2314 size = ALIGN_UP (nb + SIZE_SZ, pagesize);
2315 else
2316 size = ALIGN_UP (nb + SIZE_SZ + MALLOC_ALIGN_MASK, pagesize);
2317 tried_mmap = true;
2318
2319 /* Don't try if size wraps around 0 */
2320 if ((unsigned long) (size) > (unsigned long) (nb))
2321 {
2322 mm = (char *) (MMAP (0, size, PROT_READ | PROT_WRITE, 0));
2323
2324 if (mm != MAP_FAILED)
2325 {
2326 /*
2327 The offset to the start of the mmapped region is stored
2328 in the prev_size field of the chunk. This allows us to adjust
2329 returned start address to meet alignment requirements here
2330 and in memalign(), and still be able to compute proper
2331 address argument for later munmap in free() and realloc().
2332 */
2333
2334 if (MALLOC_ALIGNMENT == 2 * SIZE_SZ)
2335 {
2336 /* For glibc, chunk2mem increases the address by 2*SIZE_SZ and
2337 MALLOC_ALIGN_MASK is 2*SIZE_SZ-1. Each mmap'ed area is page
2338 aligned and therefore definitely MALLOC_ALIGN_MASK-aligned. */
2339 assert (((INTERNAL_SIZE_T) chunk2mem (mm) & MALLOC_ALIGN_MASK) == 0);
2340 front_misalign = 0;
2341 }
2342 else
2343 front_misalign = (INTERNAL_SIZE_T) chunk2mem (mm) & MALLOC_ALIGN_MASK;
2344 if (front_misalign > 0)
2345 {
2346 correction = MALLOC_ALIGNMENT - front_misalign;
2347 p = (mchunkptr) (mm + correction);
2348 set_prev_size (p, correction);
2349 set_head (p, (size - correction) | IS_MMAPPED);
2350 }
2351 else
2352 {
2353 p = (mchunkptr) mm;
2354 set_prev_size (p, 0);
2355 set_head (p, size | IS_MMAPPED);
2356 }
2357
2358 /* update statistics */
2359
2360 int new = atomic_exchange_and_add (&mp_.n_mmaps, 1) + 1;
2361 atomic_max (&mp_.max_n_mmaps, new);
2362
2363 unsigned long sum;
2364 sum = atomic_exchange_and_add (&mp_.mmapped_mem, size) + size;
2365 atomic_max (&mp_.max_mmapped_mem, sum);
2366
2367 check_chunk (av, p);
2368
2369 return chunk2mem (p);
2370 }
2371 }
2372 }
2373
2374 /* There are no usable arenas and mmap also failed. */
2375 if (av == NULL)
2376 return 0;
2377
2378 /* Record incoming configuration of top */
2379
2380 old_top = av->top;
2381 old_size = chunksize (old_top);
2382 old_end = (char *) (chunk_at_offset (old_top, old_size));
2383
2384 brk = snd_brk = (char *) (MORECORE_FAILURE);
2385
2386 /*
2387 If not the first time through, we require old_size to be
2388 at least MINSIZE and to have prev_inuse set.
2389 */
2390
2391 assert ((old_top == initial_top (av) && old_size == 0) ||
2392 ((unsigned long) (old_size) >= MINSIZE &&
2393 prev_inuse (old_top) &&
2394 ((unsigned long) old_end & (pagesize - 1)) == 0));
2395
2396 /* Precondition: not enough current space to satisfy nb request */
2397 assert ((unsigned long) (old_size) < (unsigned long) (nb + MINSIZE));
2398
2399
2400 if (av != &main_arena)
2401 {
2402 heap_info *old_heap, *heap;
2403 size_t old_heap_size;
2404
2405 /* First try to extend the current heap. */
2406 old_heap = heap_for_ptr (old_top);
2407 old_heap_size = old_heap->size;
2408 if ((long) (MINSIZE + nb - old_size) > 0
2409 && grow_heap (old_heap, MINSIZE + nb - old_size) == 0)
2410 {
2411 av->system_mem += old_heap->size - old_heap_size;
2412 set_head (old_top, (((char *) old_heap + old_heap->size) - (char *) old_top)
2413 | PREV_INUSE);
2414 }
2415 else if ((heap = new_heap (nb + (MINSIZE + sizeof (*heap)), mp_.top_pad)))
2416 {
2417 /* Use a newly allocated heap. */
2418 heap->ar_ptr = av;
2419 heap->prev = old_heap;
2420 av->system_mem += heap->size;
2421 /* Set up the new top. */
2422 top (av) = chunk_at_offset (heap, sizeof (*heap));
2423 set_head (top (av), (heap->size - sizeof (*heap)) | PREV_INUSE);
2424
2425 /* Setup fencepost and free the old top chunk with a multiple of
2426 MALLOC_ALIGNMENT in size. */
2427 /* The fencepost takes at least MINSIZE bytes, because it might
2428 become the top chunk again later. Note that a footer is set
2429 up, too, although the chunk is marked in use. */
2430 old_size = (old_size - MINSIZE) & ~MALLOC_ALIGN_MASK;
2431 set_head (chunk_at_offset (old_top, old_size + 2 * SIZE_SZ), 0 | PREV_INUSE);
2432 if (old_size >= MINSIZE)
2433 {
2434 set_head (chunk_at_offset (old_top, old_size), (2 * SIZE_SZ) | PREV_INUSE);
2435 set_foot (chunk_at_offset (old_top, old_size), (2 * SIZE_SZ));
2436 set_head (old_top, old_size | PREV_INUSE | NON_MAIN_ARENA);
2437 _int_free (av, old_top, 1);
2438 }
2439 else
2440 {
2441 set_head (old_top, (old_size + 2 * SIZE_SZ) | PREV_INUSE);
2442 set_foot (old_top, (old_size + 2 * SIZE_SZ));
2443 }
2444 }
2445 else if (!tried_mmap)
2446 /* We can at least try to use to mmap memory. */
2447 goto try_mmap;
2448 }
2449 else /* av == main_arena */
2450
2451
2452 { /* Request enough space for nb + pad + overhead */
2453 size = nb + mp_.top_pad + MINSIZE;
2454
2455 /*
2456 If contiguous, we can subtract out existing space that we hope to
2457 combine with new space. We add it back later only if
2458 we don't actually get contiguous space.
2459 */
2460
2461 if (contiguous (av))
2462 size -= old_size;
2463
2464 /*
2465 Round to a multiple of page size.
2466 If MORECORE is not contiguous, this ensures that we only call it
2467 with whole-page arguments. And if MORECORE is contiguous and
2468 this is not first time through, this preserves page-alignment of
2469 previous calls. Otherwise, we correct to page-align below.
2470 */
2471
2472 size = ALIGN_UP (size, pagesize);
2473
2474 /*
2475 Don't try to call MORECORE if argument is so big as to appear
2476 negative. Note that since mmap takes size_t arg, it may succeed
2477 below even if we cannot call MORECORE.
2478 */
2479
2480 if (size > 0)
2481 {
2482 brk = (char *) (MORECORE (size));
2483 LIBC_PROBE (memory_sbrk_more, 2, brk, size);
2484 }
2485
2486 if (brk != (char *) (MORECORE_FAILURE))
2487 {
2488 /* Call the `morecore' hook if necessary. */
2489 void (*hook) (void) = atomic_forced_read (__after_morecore_hook);
2490 if (__builtin_expect (hook != NULL, 0))
2491 (*hook)();
2492 }
2493 else
2494 {
2495 /*
2496 If have mmap, try using it as a backup when MORECORE fails or
2497 cannot be used. This is worth doing on systems that have "holes" in
2498 address space, so sbrk cannot extend to give contiguous space, but
2499 space is available elsewhere. Note that we ignore mmap max count
2500 and threshold limits, since the space will not be used as a
2501 segregated mmap region.
2502 */
2503
2504 /* Cannot merge with old top, so add its size back in */
2505 if (contiguous (av))
2506 size = ALIGN_UP (size + old_size, pagesize);
2507
2508 /* If we are relying on mmap as backup, then use larger units */
2509 if ((unsigned long) (size) < (unsigned long) (MMAP_AS_MORECORE_SIZE))
2510 size = MMAP_AS_MORECORE_SIZE;
2511
2512 /* Don't try if size wraps around 0 */
2513 if ((unsigned long) (size) > (unsigned long) (nb))
2514 {
2515 char *mbrk = (char *) (MMAP (0, size, PROT_READ | PROT_WRITE, 0));
2516
2517 if (mbrk != MAP_FAILED)
2518 {
2519 /* We do not need, and cannot use, another sbrk call to find end */
2520 brk = mbrk;
2521 snd_brk = brk + size;
2522
2523 /*
2524 Record that we no longer have a contiguous sbrk region.
2525 After the first time mmap is used as backup, we do not
2526 ever rely on contiguous space since this could incorrectly
2527 bridge regions.
2528 */
2529 set_noncontiguous (av);
2530 }
2531 }
2532 }
2533
2534 if (brk != (char *) (MORECORE_FAILURE))
2535 {
2536 if (mp_.sbrk_base == 0)
2537 mp_.sbrk_base = brk;
2538 av->system_mem += size;
2539
2540 /*
2541 If MORECORE extends previous space, we can likewise extend top size.
2542 */
2543
2544 if (brk == old_end && snd_brk == (char *) (MORECORE_FAILURE))
2545 set_head (old_top, (size + old_size) | PREV_INUSE);
2546
2547 else if (contiguous (av) && old_size && brk < old_end)
2548 /* Oops! Someone else killed our space.. Can't touch anything. */
2549 malloc_printerr ("break adjusted to free malloc space");
2550
2551 /*
2552 Otherwise, make adjustments:
2553
2554 * If the first time through or noncontiguous, we need to call sbrk
2555 just to find out where the end of memory lies.
2556
2557 * We need to ensure that all returned chunks from malloc will meet
2558 MALLOC_ALIGNMENT
2559
2560 * If there was an intervening foreign sbrk, we need to adjust sbrk
2561 request size to account for fact that we will not be able to
2562 combine new space with existing space in old_top.
2563
2564 * Almost all systems internally allocate whole pages at a time, in
2565 which case we might as well use the whole last page of request.
2566 So we allocate enough more memory to hit a page boundary now,
2567 which in turn causes future contiguous calls to page-align.
2568 */
2569
2570 else
2571 {
2572 front_misalign = 0;
2573 end_misalign = 0;
2574 correction = 0;
2575 aligned_brk = brk;
2576
2577 /* handle contiguous cases */
2578 if (contiguous (av))
2579 {
2580 /* Count foreign sbrk as system_mem. */
2581 if (old_size)
2582 av->system_mem += brk - old_end;
2583
2584 /* Guarantee alignment of first new chunk made from this space */
2585
2586 front_misalign = (INTERNAL_SIZE_T) chunk2mem (brk) & MALLOC_ALIGN_MASK;
2587 if (front_misalign > 0)
2588 {
2589 /*
2590 Skip over some bytes to arrive at an aligned position.
2591 We don't need to specially mark these wasted front bytes.
2592 They will never be accessed anyway because
2593 prev_inuse of av->top (and any chunk created from its start)
2594 is always true after initialization.
2595 */
2596
2597 correction = MALLOC_ALIGNMENT - front_misalign;
2598 aligned_brk += correction;
2599 }
2600
2601 /*
2602 If this isn't adjacent to existing space, then we will not
2603 be able to merge with old_top space, so must add to 2nd request.
2604 */
2605
2606 correction += old_size;
2607
2608 /* Extend the end address to hit a page boundary */
2609 end_misalign = (INTERNAL_SIZE_T) (brk + size + correction);
2610 correction += (ALIGN_UP (end_misalign, pagesize)) - end_misalign;
2611
2612 assert (correction >= 0);
2613 snd_brk = (char *) (MORECORE (correction));
2614
2615 /*
2616 If can't allocate correction, try to at least find out current
2617 brk. It might be enough to proceed without failing.
2618
2619 Note that if second sbrk did NOT fail, we assume that space
2620 is contiguous with first sbrk. This is a safe assumption unless
2621 program is multithreaded but doesn't use locks and a foreign sbrk
2622 occurred between our first and second calls.
2623 */
2624
2625 if (snd_brk == (char *) (MORECORE_FAILURE))
2626 {
2627 correction = 0;
2628 snd_brk = (char *) (MORECORE (0));
2629 }
2630 else
2631 {
2632 /* Call the `morecore' hook if necessary. */
2633 void (*hook) (void) = atomic_forced_read (__after_morecore_hook);
2634 if (__builtin_expect (hook != NULL, 0))
2635 (*hook)();
2636 }
2637 }
2638
2639 /* handle non-contiguous cases */
2640 else
2641 {
2642 if (MALLOC_ALIGNMENT == 2 * SIZE_SZ)
2643 /* MORECORE/mmap must correctly align */
2644 assert (((unsigned long) chunk2mem (brk) & MALLOC_ALIGN_MASK) == 0);
2645 else
2646 {
2647 front_misalign = (INTERNAL_SIZE_T) chunk2mem (brk) & MALLOC_ALIGN_MASK;
2648 if (front_misalign > 0)
2649 {
2650 /*
2651 Skip over some bytes to arrive at an aligned position.
2652 We don't need to specially mark these wasted front bytes.
2653 They will never be accessed anyway because
2654 prev_inuse of av->top (and any chunk created from its start)
2655 is always true after initialization.
2656 */
2657
2658 aligned_brk += MALLOC_ALIGNMENT - front_misalign;
2659 }
2660 }
2661
2662 /* Find out current end of memory */
2663 if (snd_brk == (char *) (MORECORE_FAILURE))
2664 {
2665 snd_brk = (char *) (MORECORE (0));
2666 }
2667 }
2668
2669 /* Adjust top based on results of second sbrk */
2670 if (snd_brk != (char *) (MORECORE_FAILURE))
2671 {
2672 av->top = (mchunkptr) aligned_brk;
2673 set_head (av->top, (snd_brk - aligned_brk + correction) | PREV_INUSE);
2674 av->system_mem += correction;
2675
2676 /*
2677 If not the first time through, we either have a
2678 gap due to foreign sbrk or a non-contiguous region. Insert a
2679 double fencepost at old_top to prevent consolidation with space
2680 we don't own. These fenceposts are artificial chunks that are
2681 marked as inuse and are in any case too small to use. We need
2682 two to make sizes and alignments work out.
2683 */
2684
2685 if (old_size != 0)
2686 {
2687 /*
2688 Shrink old_top to insert fenceposts, keeping size a
2689 multiple of MALLOC_ALIGNMENT. We know there is at least
2690 enough space in old_top to do this.
2691 */
2692 old_size = (old_size - 4 * SIZE_SZ) & ~MALLOC_ALIGN_MASK;
2693 set_head (old_top, old_size | PREV_INUSE);
2694
2695 /*
2696 Note that the following assignments completely overwrite
2697 old_top when old_size was previously MINSIZE. This is
2698 intentional. We need the fencepost, even if old_top otherwise gets
2699 lost.
2700 */
2701 set_head (chunk_at_offset (old_top, old_size),
2702 (2 * SIZE_SZ) | PREV_INUSE);
2703 set_head (chunk_at_offset (old_top, old_size + 2 * SIZE_SZ),
2704 (2 * SIZE_SZ) | PREV_INUSE);
2705
2706 /* If possible, release the rest. */
2707 if (old_size >= MINSIZE)
2708 {
2709 _int_free (av, old_top, 1);
2710 }
2711 }
2712 }
2713 }
2714 }
2715 } /* if (av != &main_arena) */
2716
2717 if ((unsigned long) av->system_mem > (unsigned long) (av->max_system_mem))
2718 av->max_system_mem = av->system_mem;
2719 check_malloc_state (av);
2720
2721 /* finally, do the allocation */
2722 p = av->top;
2723 size = chunksize (p);
2724
2725 /* check that one of the above allocation paths succeeded */
2726 if ((unsigned long) (size) >= (unsigned long) (nb + MINSIZE))
2727 {
2728 remainder_size = size - nb;
2729 remainder = chunk_at_offset (p, nb);
2730 av->top = remainder;
2731 set_head (p, nb | PREV_INUSE | (av != &main_arena ? NON_MAIN_ARENA : 0));
2732 set_head (remainder, remainder_size | PREV_INUSE);
2733 check_malloced_chunk (av, p, nb);
2734 return chunk2mem (p);
2735 }
2736
2737 /* catch all failure paths */
2738 __set_errno (ENOMEM);
2739 return 0;
2740}
2741
2742
2743/*
2744 systrim is an inverse of sorts to sysmalloc. It gives memory back
2745 to the system (via negative arguments to sbrk) if there is unused
2746 memory at the `high' end of the malloc pool. It is called
2747 automatically by free() when top space exceeds the trim
2748 threshold. It is also called by the public malloc_trim routine. It
2749 returns 1 if it actually released any memory, else 0.
2750 */
2751
2752static int
2753systrim (size_t pad, mstate av)
2754{
2755 long top_size; /* Amount of top-most memory */
2756 long extra; /* Amount to release */
2757 long released; /* Amount actually released */
2758 char *current_brk; /* address returned by pre-check sbrk call */
2759 char *new_brk; /* address returned by post-check sbrk call */
2760 size_t pagesize;
2761 long top_area;
2762
2763 pagesize = GLRO (dl_pagesize);
2764 top_size = chunksize (av->top);
2765
2766 top_area = top_size - MINSIZE - 1;
2767 if (top_area <= pad)
2768 return 0;
2769
2770 /* Release in pagesize units and round down to the nearest page. */
2771 extra = ALIGN_DOWN(top_area - pad, pagesize);
2772
2773 if (extra == 0)
2774 return 0;
2775
2776 /*
2777 Only proceed if end of memory is where we last set it.
2778 This avoids problems if there were foreign sbrk calls.
2779 */
2780 current_brk = (char *) (MORECORE (0));
2781 if (current_brk == (char *) (av->top) + top_size)
2782 {
2783 /*
2784 Attempt to release memory. We ignore MORECORE return value,
2785 and instead call again to find out where new end of memory is.
2786 This avoids problems if first call releases less than we asked,
2787 of if failure somehow altered brk value. (We could still
2788 encounter problems if it altered brk in some very bad way,
2789 but the only thing we can do is adjust anyway, which will cause
2790 some downstream failure.)
2791 */
2792
2793 MORECORE (-extra);
2794 /* Call the `morecore' hook if necessary. */
2795 void (*hook) (void) = atomic_forced_read (__after_morecore_hook);
2796 if (__builtin_expect (hook != NULL, 0))
2797 (*hook)();
2798 new_brk = (char *) (MORECORE (0));
2799
2800 LIBC_PROBE (memory_sbrk_less, 2, new_brk, extra);
2801
2802 if (new_brk != (char *) MORECORE_FAILURE)
2803 {
2804 released = (long) (current_brk - new_brk);
2805
2806 if (released != 0)
2807 {
2808 /* Success. Adjust top. */
2809 av->system_mem -= released;
2810 set_head (av->top, (top_size - released) | PREV_INUSE);
2811 check_malloc_state (av);
2812 return 1;
2813 }
2814 }
2815 }
2816 return 0;
2817}
2818
2819static void
2820munmap_chunk (mchunkptr p)
2821{
2822 INTERNAL_SIZE_T size = chunksize (p);
2823
2824 assert (chunk_is_mmapped (p));
2825
2826 /* Do nothing if the chunk is a faked mmapped chunk in the dumped
2827 main arena. We never free this memory. */
2828 if (DUMPED_MAIN_ARENA_CHUNK (p))
2829 return;
2830
2831 uintptr_t block = (uintptr_t) p - prev_size (p);
2832 size_t total_size = prev_size (p) + size;
2833 /* Unfortunately we have to do the compilers job by hand here. Normally
2834 we would test BLOCK and TOTAL-SIZE separately for compliance with the
2835 page size. But gcc does not recognize the optimization possibility
2836 (in the moment at least) so we combine the two values into one before
2837 the bit test. */
2838 if (__builtin_expect (((block | total_size) & (GLRO (dl_pagesize) - 1)) != 0, 0))
2839 malloc_printerr ("munmap_chunk(): invalid pointer");
2840
2841 atomic_decrement (&mp_.n_mmaps);
2842 atomic_add (&mp_.mmapped_mem, -total_size);
2843
2844 /* If munmap failed the process virtual memory address space is in a
2845 bad shape. Just leave the block hanging around, the process will
2846 terminate shortly anyway since not much can be done. */
2847 __munmap ((char *) block, total_size);
2848}
2849
2850#if HAVE_MREMAP
2851
2852static mchunkptr
2853mremap_chunk (mchunkptr p, size_t new_size)
2854{
2855 size_t pagesize = GLRO (dl_pagesize);
2856 INTERNAL_SIZE_T offset = prev_size (p);
2857 INTERNAL_SIZE_T size = chunksize (p);
2858 char *cp;
2859
2860 assert (chunk_is_mmapped (p));
2861 assert (((size + offset) & (GLRO (dl_pagesize) - 1)) == 0);
2862
2863 /* Note the extra SIZE_SZ overhead as in mmap_chunk(). */
2864 new_size = ALIGN_UP (new_size + offset + SIZE_SZ, pagesize);
2865
2866 /* No need to remap if the number of pages does not change. */
2867 if (size + offset == new_size)
2868 return p;
2869
2870 cp = (char *) __mremap ((char *) p - offset, size + offset, new_size,
2871 MREMAP_MAYMOVE);
2872
2873 if (cp == MAP_FAILED)
2874 return 0;
2875
2876 p = (mchunkptr) (cp + offset);
2877
2878 assert (aligned_OK (chunk2mem (p)));
2879
2880 assert (prev_size (p) == offset);
2881 set_head (p, (new_size - offset) | IS_MMAPPED);
2882
2883 INTERNAL_SIZE_T new;
2884 new = atomic_exchange_and_add (&mp_.mmapped_mem, new_size - size - offset)
2885 + new_size - size - offset;
2886 atomic_max (&mp_.max_mmapped_mem, new);
2887 return p;
2888}
2889#endif /* HAVE_MREMAP */
2890
2891/*------------------------ Public wrappers. --------------------------------*/
2892
2893#if USE_TCACHE
2894
2895/* We overlay this structure on the user-data portion of a chunk when
2896 the chunk is stored in the per-thread cache. */
2897typedef struct tcache_entry
2898{
2899 struct tcache_entry *next;
2900} tcache_entry;
2901
2902/* There is one of these for each thread, which contains the
2903 per-thread cache (hence "tcache_perthread_struct"). Keeping
2904 overall size low is mildly important. Note that COUNTS and ENTRIES
2905 are redundant (we could have just counted the linked list each
2906 time), this is for performance reasons. */
2907typedef struct tcache_perthread_struct
2908{
2909 char counts[TCACHE_MAX_BINS];
2910 tcache_entry *entries[TCACHE_MAX_BINS];
2911} tcache_perthread_struct;
2912
2913static __thread bool tcache_shutting_down = false;
2914static __thread tcache_perthread_struct *tcache = NULL;
2915
2916/* Caller must ensure that we know tc_idx is valid and there's room
2917 for more chunks. */
2918static __always_inline void
2919tcache_put (mchunkptr chunk, size_t tc_idx)
2920{
2921 tcache_entry *e = (tcache_entry *) chunk2mem (chunk);
2922 assert (tc_idx < TCACHE_MAX_BINS);
2923 e->next = tcache->entries[tc_idx];
2924 tcache->entries[tc_idx] = e;
2925 ++(tcache->counts[tc_idx]);
2926}
2927
2928/* Caller must ensure that we know tc_idx is valid and there's
2929 available chunks to remove. */
2930static __always_inline void *
2931tcache_get (size_t tc_idx)
2932{
2933 tcache_entry *e = tcache->entries[tc_idx];
2934 assert (tc_idx < TCACHE_MAX_BINS);
2935 assert (tcache->entries[tc_idx] > 0);
2936 tcache->entries[tc_idx] = e->next;
2937 --(tcache->counts[tc_idx]);
2938 return (void *) e;
2939}
2940
2941static void __attribute__ ((section ("__libc_thread_freeres_fn")))
2942tcache_thread_freeres (void)
2943{
2944 int i;
2945 tcache_perthread_struct *tcache_tmp = tcache;
2946
2947 if (!tcache)
2948 return;
2949
2950 /* Disable the tcache and prevent it from being reinitialized. */
2951 tcache = NULL;
2952 tcache_shutting_down = true;
2953
2954 /* Free all of the entries and the tcache itself back to the arena
2955 heap for coalescing. */
2956 for (i = 0; i < TCACHE_MAX_BINS; ++i)
2957 {
2958 while (tcache_tmp->entries[i])
2959 {
2960 tcache_entry *e = tcache_tmp->entries[i];
2961 tcache_tmp->entries[i] = e->next;
2962 __libc_free (e);
2963 }
2964 }
2965
2966 __libc_free (tcache_tmp);
2967}
2968text_set_element (__libc_thread_subfreeres, tcache_thread_freeres);
2969
2970static void
2971tcache_init(void)
2972{
2973 mstate ar_ptr;
2974 void *victim = 0;
2975 const size_t bytes = sizeof (tcache_perthread_struct);
2976
2977 if (tcache_shutting_down)
2978 return;
2979
2980 arena_get (ar_ptr, bytes);
2981 victim = _int_malloc (ar_ptr, bytes);
2982 if (!victim && ar_ptr != NULL)
2983 {
2984 ar_ptr = arena_get_retry (ar_ptr, bytes);
2985 victim = _int_malloc (ar_ptr, bytes);
2986 }
2987
2988
2989 if (ar_ptr != NULL)
2990 __libc_lock_unlock (ar_ptr->mutex);
2991
2992 /* In a low memory situation, we may not be able to allocate memory
2993 - in which case, we just keep trying later. However, we
2994 typically do this very early, so either there is sufficient
2995 memory, or there isn't enough memory to do non-trivial
2996 allocations anyway. */
2997 if (victim)
2998 {
2999 tcache = (tcache_perthread_struct *) victim;
3000 memset (tcache, 0, sizeof (tcache_perthread_struct));
3001 }
3002
3003}
3004
3005#define MAYBE_INIT_TCACHE() \
3006 if (__glibc_unlikely (tcache == NULL)) \
3007 tcache_init();
3008
3009#else
3010#define MAYBE_INIT_TCACHE()
3011#endif
3012
3013void *
3014__libc_malloc (size_t bytes)
3015{
3016 mstate ar_ptr;
3017 void *victim;
3018
3019 void *(*hook) (size_t, const void *)
3020 = atomic_forced_read (__malloc_hook);
3021 if (__builtin_expect (hook != NULL, 0))
3022 return (*hook)(bytes, RETURN_ADDRESS (0));
3023#if USE_TCACHE
3024 /* int_free also calls request2size, be careful to not pad twice. */
3025 size_t tbytes = request2size (bytes);
3026 size_t tc_idx = csize2tidx (tbytes);
3027
3028 MAYBE_INIT_TCACHE ();
3029
3030 DIAG_PUSH_NEEDS_COMMENT;
3031 if (tc_idx < mp_.tcache_bins
3032 /*&& tc_idx < TCACHE_MAX_BINS*/ /* to appease gcc */
3033 && tcache
3034 && tcache->entries[tc_idx] != NULL)
3035 {
3036 return tcache_get (tc_idx);
3037 }
3038 DIAG_POP_NEEDS_COMMENT;
3039#endif
3040
3041 if (SINGLE_THREAD_P)
3042 {
3043 victim = _int_malloc (&main_arena, bytes);
3044 assert (!victim || chunk_is_mmapped (mem2chunk (victim)) ||
3045 &main_arena == arena_for_chunk (mem2chunk (victim)));
3046 return victim;
3047 }
3048
3049 arena_get (ar_ptr, bytes);
3050
3051 victim = _int_malloc (ar_ptr, bytes);
3052 /* Retry with another arena only if we were able to find a usable arena
3053 before. */
3054 if (!victim && ar_ptr != NULL)
3055 {
3056 LIBC_PROBE (memory_malloc_retry, 1, bytes);
3057 ar_ptr = arena_get_retry (ar_ptr, bytes);
3058 victim = _int_malloc (ar_ptr, bytes);
3059 }
3060
3061 if (ar_ptr != NULL)
3062 __libc_lock_unlock (ar_ptr->mutex);
3063
3064 assert (!victim || chunk_is_mmapped (mem2chunk (victim)) ||
3065 ar_ptr == arena_for_chunk (mem2chunk (victim)));
3066 return victim;
3067}
3068libc_hidden_def (__libc_malloc)
3069
3070void
3071__libc_free (void *mem)
3072{
3073 mstate ar_ptr;
3074 mchunkptr p; /* chunk corresponding to mem */
3075
3076 void (*hook) (void *, const void *)
3077 = atomic_forced_read (__free_hook);
3078 if (__builtin_expect (hook != NULL, 0))
3079 {
3080 (*hook)(mem, RETURN_ADDRESS (0));
3081 return;
3082 }
3083
3084 if (mem == 0) /* free(0) has no effect */
3085 return;
3086
3087 p = mem2chunk (mem);
3088
3089 if (chunk_is_mmapped (p)) /* release mmapped memory. */
3090 {
3091 /* See if the dynamic brk/mmap threshold needs adjusting.
3092 Dumped fake mmapped chunks do not affect the threshold. */
3093 if (!mp_.no_dyn_threshold
3094 && chunksize_nomask (p) > mp_.mmap_threshold
3095 && chunksize_nomask (p) <= DEFAULT_MMAP_THRESHOLD_MAX
3096 && !DUMPED_MAIN_ARENA_CHUNK (p))
3097 {
3098 mp_.mmap_threshold = chunksize (p);
3099 mp_.trim_threshold = 2 * mp_.mmap_threshold;
3100 LIBC_PROBE (memory_mallopt_free_dyn_thresholds, 2,
3101 mp_.mmap_threshold, mp_.trim_threshold);
3102 }
3103 munmap_chunk (p);
3104 return;
3105 }
3106
3107 MAYBE_INIT_TCACHE ();
3108
3109 ar_ptr = arena_for_chunk (p);
3110 _int_free (ar_ptr, p, 0);
3111}
3112libc_hidden_def (__libc_free)
3113
3114void *
3115__libc_realloc (void *oldmem, size_t bytes)
3116{
3117 mstate ar_ptr;
3118 INTERNAL_SIZE_T nb; /* padded request size */
3119
3120 void *newp; /* chunk to return */
3121
3122 void *(*hook) (void *, size_t, const void *) =
3123 atomic_forced_read (__realloc_hook);
3124 if (__builtin_expect (hook != NULL, 0))
3125 return (*hook)(oldmem, bytes, RETURN_ADDRESS (0));
3126
3127#if REALLOC_ZERO_BYTES_FREES
3128 if (bytes == 0 && oldmem != NULL)
3129 {
3130 __libc_free (oldmem); return 0;
3131 }
3132#endif
3133
3134 /* realloc of null is supposed to be same as malloc */
3135 if (oldmem == 0)
3136 return __libc_malloc (bytes);
3137
3138 /* chunk corresponding to oldmem */
3139 const mchunkptr oldp = mem2chunk (oldmem);
3140 /* its size */
3141 const INTERNAL_SIZE_T oldsize = chunksize (oldp);
3142
3143 if (chunk_is_mmapped (oldp))
3144 ar_ptr = NULL;
3145 else
3146 {
3147 MAYBE_INIT_TCACHE ();
3148 ar_ptr = arena_for_chunk (oldp);
3149 }
3150
3151 /* Little security check which won't hurt performance: the allocator
3152 never wrapps around at the end of the address space. Therefore
3153 we can exclude some size values which might appear here by
3154 accident or by "design" from some intruder. We need to bypass
3155 this check for dumped fake mmap chunks from the old main arena
3156 because the new malloc may provide additional alignment. */
3157 if ((__builtin_expect ((uintptr_t) oldp > (uintptr_t) -oldsize, 0)
3158 || __builtin_expect (misaligned_chunk (oldp), 0))
3159 && !DUMPED_MAIN_ARENA_CHUNK (oldp))
3160 malloc_printerr ("realloc(): invalid pointer");
3161
3162 checked_request2size (bytes, nb);
3163
3164 if (chunk_is_mmapped (oldp))
3165 {
3166 /* If this is a faked mmapped chunk from the dumped main arena,
3167 always make a copy (and do not free the old chunk). */
3168 if (DUMPED_MAIN_ARENA_CHUNK (oldp))
3169 {
3170 /* Must alloc, copy, free. */
3171 void *newmem = __libc_malloc (bytes);
3172 if (newmem == 0)
3173 return NULL;
3174 /* Copy as many bytes as are available from the old chunk
3175 and fit into the new size. NB: The overhead for faked
3176 mmapped chunks is only SIZE_SZ, not 2 * SIZE_SZ as for
3177 regular mmapped chunks. */
3178 if (bytes > oldsize - SIZE_SZ)
3179 bytes = oldsize - SIZE_SZ;
3180 memcpy (newmem, oldmem, bytes);
3181 return newmem;
3182 }
3183
3184 void *newmem;
3185
3186#if HAVE_MREMAP
3187 newp = mremap_chunk (oldp, nb);
3188 if (newp)
3189 return chunk2mem (newp);
3190#endif
3191 /* Note the extra SIZE_SZ overhead. */
3192 if (oldsize - SIZE_SZ >= nb)
3193 return oldmem; /* do nothing */
3194
3195 /* Must alloc, copy, free. */
3196 newmem = __libc_malloc (bytes);
3197 if (newmem == 0)
3198 return 0; /* propagate failure */
3199
3200 memcpy (newmem, oldmem, oldsize - 2 * SIZE_SZ);
3201 munmap_chunk (oldp);
3202 return newmem;
3203 }
3204
3205 if (SINGLE_THREAD_P)
3206 {
3207 newp = _int_realloc (ar_ptr, oldp, oldsize, nb);
3208 assert (!newp || chunk_is_mmapped (mem2chunk (newp)) ||
3209 ar_ptr == arena_for_chunk (mem2chunk (newp)));
3210
3211 return newp;
3212 }
3213
3214 __libc_lock_lock (ar_ptr->mutex);
3215
3216 newp = _int_realloc (ar_ptr, oldp, oldsize, nb);
3217
3218 __libc_lock_unlock (ar_ptr->mutex);
3219 assert (!newp || chunk_is_mmapped (mem2chunk (newp)) ||
3220 ar_ptr == arena_for_chunk (mem2chunk (newp)));
3221
3222 if (newp == NULL)
3223 {
3224 /* Try harder to allocate memory in other arenas. */
3225 LIBC_PROBE (memory_realloc_retry, 2, bytes, oldmem);
3226 newp = __libc_malloc (bytes);
3227 if (newp != NULL)
3228 {
3229 memcpy (newp, oldmem, oldsize - SIZE_SZ);
3230 _int_free (ar_ptr, oldp, 0);
3231 }
3232 }
3233
3234 return newp;
3235}
3236libc_hidden_def (__libc_realloc)
3237
3238void *
3239__libc_memalign (size_t alignment, size_t bytes)
3240{
3241 void *address = RETURN_ADDRESS (0);
3242 return _mid_memalign (alignment, bytes, address);
3243}
3244
3245static void *
3246_mid_memalign (size_t alignment, size_t bytes, void *address)
3247{
3248 mstate ar_ptr;
3249 void *p;
3250
3251 void *(*hook) (size_t, size_t, const void *) =
3252 atomic_forced_read (__memalign_hook);
3253 if (__builtin_expect (hook != NULL, 0))
3254 return (*hook)(alignment, bytes, address);
3255
3256 /* If we need less alignment than we give anyway, just relay to malloc. */
3257 if (alignment <= MALLOC_ALIGNMENT)
3258 return __libc_malloc (bytes);
3259
3260 /* Otherwise, ensure that it is at least a minimum chunk size */
3261 if (alignment < MINSIZE)
3262 alignment = MINSIZE;
3263
3264 /* If the alignment is greater than SIZE_MAX / 2 + 1 it cannot be a
3265 power of 2 and will cause overflow in the check below. */
3266 if (alignment > SIZE_MAX / 2 + 1)
3267 {
3268 __set_errno (EINVAL);
3269 return 0;
3270 }
3271
3272 /* Check for overflow. */
3273 if (bytes > SIZE_MAX - alignment - MINSIZE)
3274 {
3275 __set_errno (ENOMEM);
3276 return 0;
3277 }
3278
3279
3280 /* Make sure alignment is power of 2. */
3281 if (!powerof2 (alignment))
3282 {
3283 size_t a = MALLOC_ALIGNMENT * 2;
3284 while (a < alignment)
3285 a <<= 1;
3286 alignment = a;
3287 }
3288
3289 if (SINGLE_THREAD_P)
3290 {
3291 p = _int_memalign (&main_arena, alignment, bytes);
3292 assert (!p || chunk_is_mmapped (mem2chunk (p)) ||
3293 &main_arena == arena_for_chunk (mem2chunk (p)));
3294
3295 return p;
3296 }
3297
3298 arena_get (ar_ptr, bytes + alignment + MINSIZE);
3299
3300 p = _int_memalign (ar_ptr, alignment, bytes);
3301 if (!p && ar_ptr != NULL)
3302 {
3303 LIBC_PROBE (memory_memalign_retry, 2, bytes, alignment);
3304 ar_ptr = arena_get_retry (ar_ptr, bytes);
3305 p = _int_memalign (ar_ptr, alignment, bytes);
3306 }
3307
3308 if (ar_ptr != NULL)
3309 __libc_lock_unlock (ar_ptr->mutex);
3310
3311 assert (!p || chunk_is_mmapped (mem2chunk (p)) ||
3312 ar_ptr == arena_for_chunk (mem2chunk (p)));
3313 return p;
3314}
3315/* For ISO C11. */
3316weak_alias (__libc_memalign, aligned_alloc)
3317libc_hidden_def (__libc_memalign)
3318
3319void *
3320__libc_valloc (size_t bytes)
3321{
3322 if (__malloc_initialized < 0)
3323 ptmalloc_init ();
3324
3325 void *address = RETURN_ADDRESS (0);
3326 size_t pagesize = GLRO (dl_pagesize);
3327 return _mid_memalign (pagesize, bytes, address);
3328}
3329
3330void *
3331__libc_pvalloc (size_t bytes)
3332{
3333 if (__malloc_initialized < 0)
3334 ptmalloc_init ();
3335
3336 void *address = RETURN_ADDRESS (0);
3337 size_t pagesize = GLRO (dl_pagesize);
3338 size_t rounded_bytes = ALIGN_UP (bytes, pagesize);
3339
3340 /* Check for overflow. */
3341 if (bytes > SIZE_MAX - 2 * pagesize - MINSIZE)
3342 {
3343 __set_errno (ENOMEM);
3344 return 0;
3345 }
3346
3347 return _mid_memalign (pagesize, rounded_bytes, address);
3348}
3349
3350void *
3351__libc_calloc (size_t n, size_t elem_size)
3352{
3353 mstate av;
3354 mchunkptr oldtop, p;
3355 INTERNAL_SIZE_T bytes, sz, csz, oldtopsize;
3356 void *mem;
3357 unsigned long clearsize;
3358 unsigned long nclears;
3359 INTERNAL_SIZE_T *d;
3360
3361 /* size_t is unsigned so the behavior on overflow is defined. */
3362 bytes = n * elem_size;
3363#define HALF_INTERNAL_SIZE_T \
3364 (((INTERNAL_SIZE_T) 1) << (8 * sizeof (INTERNAL_SIZE_T) / 2))
3365 if (__builtin_expect ((n | elem_size) >= HALF_INTERNAL_SIZE_T, 0))
3366 {
3367 if (elem_size != 0 && bytes / elem_size != n)
3368 {
3369 __set_errno (ENOMEM);
3370 return 0;
3371 }
3372 }
3373
3374 void *(*hook) (size_t, const void *) =
3375 atomic_forced_read (__malloc_hook);
3376 if (__builtin_expect (hook != NULL, 0))
3377 {
3378 sz = bytes;
3379 mem = (*hook)(sz, RETURN_ADDRESS (0));
3380 if (mem == 0)
3381 return 0;
3382
3383 return memset (mem, 0, sz);
3384 }
3385
3386 sz = bytes;
3387
3388 MAYBE_INIT_TCACHE ();
3389
3390 if (SINGLE_THREAD_P)
3391 av = &main_arena;
3392 else
3393 arena_get (av, sz);
3394
3395 if (av)
3396 {
3397 /* Check if we hand out the top chunk, in which case there may be no
3398 need to clear. */
3399#if MORECORE_CLEARS
3400 oldtop = top (av);
3401 oldtopsize = chunksize (top (av));
3402# if MORECORE_CLEARS < 2
3403 /* Only newly allocated memory is guaranteed to be cleared. */
3404 if (av == &main_arena &&
3405 oldtopsize < mp_.sbrk_base + av->max_system_mem - (char *) oldtop)
3406 oldtopsize = (mp_.sbrk_base + av->max_system_mem - (char *) oldtop);
3407# endif
3408 if (av != &main_arena)
3409 {
3410 heap_info *heap = heap_for_ptr (oldtop);
3411 if (oldtopsize < (char *) heap + heap->mprotect_size - (char *) oldtop)
3412 oldtopsize = (char *) heap + heap->mprotect_size - (char *) oldtop;
3413 }
3414#endif
3415 }
3416 else
3417 {
3418 /* No usable arenas. */
3419 oldtop = 0;
3420 oldtopsize = 0;
3421 }
3422 mem = _int_malloc (av, sz);
3423
3424 assert (!mem || chunk_is_mmapped (mem2chunk (mem)) ||
3425 av == arena_for_chunk (mem2chunk (mem)));
3426
3427 if (!SINGLE_THREAD_P)
3428 {
3429 if (mem == 0 && av != NULL)
3430 {
3431 LIBC_PROBE (memory_calloc_retry, 1, sz);
3432 av = arena_get_retry (av, sz);
3433 mem = _int_malloc (av, sz);
3434 }
3435
3436 if (av != NULL)
3437 __libc_lock_unlock (av->mutex);
3438 }
3439
3440 /* Allocation failed even after a retry. */
3441 if (mem == 0)
3442 return 0;
3443
3444 p = mem2chunk (mem);
3445
3446 /* Two optional cases in which clearing not necessary */
3447 if (chunk_is_mmapped (p))
3448 {
3449 if (__builtin_expect (perturb_byte, 0))
3450 return memset (mem, 0, sz);
3451
3452 return mem;
3453 }
3454
3455 csz = chunksize (p);
3456
3457#if MORECORE_CLEARS
3458 if (perturb_byte == 0 && (p == oldtop && csz > oldtopsize))
3459 {
3460 /* clear only the bytes from non-freshly-sbrked memory */
3461 csz = oldtopsize;
3462 }
3463#endif
3464
3465 /* Unroll clear of <= 36 bytes (72 if 8byte sizes). We know that
3466 contents have an odd number of INTERNAL_SIZE_T-sized words;
3467 minimally 3. */
3468 d = (INTERNAL_SIZE_T *) mem;
3469 clearsize = csz - SIZE_SZ;
3470 nclears = clearsize / sizeof (INTERNAL_SIZE_T);
3471 assert (nclears >= 3);
3472
3473 if (nclears > 9)
3474 return memset (d, 0, clearsize);
3475
3476 else
3477 {
3478 *(d + 0) = 0;
3479 *(d + 1) = 0;
3480 *(d + 2) = 0;
3481 if (nclears > 4)
3482 {
3483 *(d + 3) = 0;
3484 *(d + 4) = 0;
3485 if (nclears > 6)
3486 {
3487 *(d + 5) = 0;
3488 *(d + 6) = 0;
3489 if (nclears > 8)
3490 {
3491 *(d + 7) = 0;
3492 *(d + 8) = 0;
3493 }
3494 }
3495 }
3496 }
3497
3498 return mem;
3499}
3500
3501/*
3502 ------------------------------ malloc ------------------------------
3503 */
3504
3505static void *
3506_int_malloc (mstate av, size_t bytes)
3507{
3508 INTERNAL_SIZE_T nb; /* normalized request size */
3509 unsigned int idx; /* associated bin index */
3510 mbinptr bin; /* associated bin */
3511
3512 mchunkptr victim; /* inspected/selected chunk */
3513 INTERNAL_SIZE_T size; /* its size */
3514 int victim_index; /* its bin index */
3515
3516 mchunkptr remainder; /* remainder from a split */
3517 unsigned long remainder_size; /* its size */
3518
3519 unsigned int block; /* bit map traverser */
3520 unsigned int bit; /* bit map traverser */
3521 unsigned int map; /* current word of binmap */
3522
3523 mchunkptr fwd; /* misc temp for linking */
3524 mchunkptr bck; /* misc temp for linking */
3525
3526#if USE_TCACHE
3527 size_t tcache_unsorted_count; /* count of unsorted chunks processed */
3528#endif
3529
3530 /*
3531 Convert request size to internal form by adding SIZE_SZ bytes
3532 overhead plus possibly more to obtain necessary alignment and/or
3533 to obtain a size of at least MINSIZE, the smallest allocatable
3534 size. Also, checked_request2size traps (returning 0) request sizes
3535 that are so large that they wrap around zero when padded and
3536 aligned.
3537 */
3538
3539 checked_request2size (bytes, nb);
3540
3541 /* There are no usable arenas. Fall back to sysmalloc to get a chunk from
3542 mmap. */
3543 if (__glibc_unlikely (av == NULL))
3544 {
3545 void *p = sysmalloc (nb, av);
3546 if (p != NULL)
3547 alloc_perturb (p, bytes);
3548 return p;
3549 }
3550
3551 /*
3552 If the size qualifies as a fastbin, first check corresponding bin.
3553 This code is safe to execute even if av is not yet initialized, so we
3554 can try it without checking, which saves some time on this fast path.
3555 */
3556
3557#define REMOVE_FB(fb, victim, pp) \
3558 do \
3559 { \
3560 victim = pp; \
3561 if (victim == NULL) \
3562 break; \
3563 } \
3564 while ((pp = catomic_compare_and_exchange_val_acq (fb, victim->fd, victim)) \
3565 != victim); \
3566
3567 if ((unsigned long) (nb) <= (unsigned long) (get_max_fast ()))
3568 {
3569 idx = fastbin_index (nb);
3570 mfastbinptr *fb = &fastbin (av, idx);
3571 mchunkptr pp;
3572 victim = *fb;
3573
3574 if (victim != NULL)
3575 {
3576 if (SINGLE_THREAD_P)
3577 *fb = victim->fd;
3578 else
3579 REMOVE_FB (fb, pp, victim);
3580 if (__glibc_likely (victim != NULL))
3581 {
3582 size_t victim_idx = fastbin_index (chunksize (victim));
3583 if (__builtin_expect (victim_idx != idx, 0))
3584 malloc_printerr ("malloc(): memory corruption (fast)");
3585 check_remalloced_chunk (av, victim, nb);
3586#if USE_TCACHE
3587 /* While we're here, if we see other chunks of the same size,
3588 stash them in the tcache. */
3589 size_t tc_idx = csize2tidx (nb);
3590 if (tcache && tc_idx < mp_.tcache_bins)
3591 {
3592 mchunkptr tc_victim;
3593
3594 /* While bin not empty and tcache not full, copy chunks. */
3595 while (tcache->counts[tc_idx] < mp_.tcache_count
3596 && (tc_victim = *fb) != NULL)
3597 {
3598 if (SINGLE_THREAD_P)
3599 *fb = tc_victim->fd;
3600 else
3601 {
3602 REMOVE_FB (fb, pp, tc_victim);
3603 if (__glibc_unlikely (tc_victim == NULL))
3604 break;
3605 }
3606 tcache_put (tc_victim, tc_idx);
3607 }
3608 }
3609#endif
3610 void *p = chunk2mem (victim);
3611 alloc_perturb (p, bytes);
3612 return p;
3613 }
3614 }
3615 }
3616
3617 /*
3618 If a small request, check regular bin. Since these "smallbins"
3619 hold one size each, no searching within bins is necessary.
3620 (For a large request, we need to wait until unsorted chunks are
3621 processed to find best fit. But for small ones, fits are exact
3622 anyway, so we can check now, which is faster.)
3623 */
3624
3625 if (in_smallbin_range (nb))
3626 {
3627 idx = smallbin_index (nb);
3628 bin = bin_at (av, idx);
3629
3630 if ((victim = last (bin)) != bin)
3631 {
3632 bck = victim->bk;
3633 if (__glibc_unlikely (bck->fd != victim))
3634 malloc_printerr ("malloc(): smallbin double linked list corrupted");
3635 set_inuse_bit_at_offset (victim, nb);
3636 bin->bk = bck;
3637 bck->fd = bin;
3638
3639 if (av != &main_arena)
3640 set_non_main_arena (victim);
3641 check_malloced_chunk (av, victim, nb);
3642#if USE_TCACHE
3643 /* While we're here, if we see other chunks of the same size,
3644 stash them in the tcache. */
3645 size_t tc_idx = csize2tidx (nb);
3646 if (tcache && tc_idx < mp_.tcache_bins)
3647 {
3648 mchunkptr tc_victim;
3649
3650 /* While bin not empty and tcache not full, copy chunks over. */
3651 while (tcache->counts[tc_idx] < mp_.tcache_count
3652 && (tc_victim = last (bin)) != bin)
3653 {
3654 if (tc_victim != 0)
3655 {
3656 bck = tc_victim->bk;
3657 set_inuse_bit_at_offset (tc_victim, nb);
3658 if (av != &main_arena)
3659 set_non_main_arena (tc_victim);
3660 bin->bk = bck;
3661 bck->fd = bin;
3662
3663 tcache_put (tc_victim, tc_idx);
3664 }
3665 }
3666 }
3667#endif
3668 void *p = chunk2mem (victim);
3669 alloc_perturb (p, bytes);
3670 return p;
3671 }
3672 }
3673
3674 /*
3675 If this is a large request, consolidate fastbins before continuing.
3676 While it might look excessive to kill all fastbins before
3677 even seeing if there is space available, this avoids
3678 fragmentation problems normally associated with fastbins.
3679 Also, in practice, programs tend to have runs of either small or
3680 large requests, but less often mixtures, so consolidation is not
3681 invoked all that often in most programs. And the programs that
3682 it is called frequently in otherwise tend to fragment.
3683 */
3684
3685 else
3686 {
3687 idx = largebin_index (nb);
3688 if (atomic_load_relaxed (&av->have_fastchunks))
3689 malloc_consolidate (av);
3690 }
3691
3692 /*
3693 Process recently freed or remaindered chunks, taking one only if
3694 it is exact fit, or, if this a small request, the chunk is remainder from
3695 the most recent non-exact fit. Place other traversed chunks in
3696 bins. Note that this step is the only place in any routine where
3697 chunks are placed in bins.
3698
3699 The outer loop here is needed because we might not realize until
3700 near the end of malloc that we should have consolidated, so must
3701 do so and retry. This happens at most once, and only when we would
3702 otherwise need to expand memory to service a "small" request.
3703 */
3704
3705#if USE_TCACHE
3706 INTERNAL_SIZE_T tcache_nb = 0;
3707 size_t tc_idx = csize2tidx (nb);
3708 if (tcache && tc_idx < mp_.tcache_bins)
3709 tcache_nb = nb;
3710 int return_cached = 0;
3711
3712 tcache_unsorted_count = 0;
3713#endif
3714
3715 for (;; )
3716 {
3717 int iters = 0;
3718 while ((victim = unsorted_chunks (av)->bk) != unsorted_chunks (av))
3719 {
3720 bck = victim->bk;
3721 if (__builtin_expect (chunksize_nomask (victim) <= 2 * SIZE_SZ, 0)
3722 || __builtin_expect (chunksize_nomask (victim)
3723 > av->system_mem, 0))
3724 malloc_printerr ("malloc(): memory corruption");
3725 size = chunksize (victim);
3726
3727 /*
3728 If a small request, try to use last remainder if it is the
3729 only chunk in unsorted bin. This helps promote locality for
3730 runs of consecutive small requests. This is the only
3731 exception to best-fit, and applies only when there is
3732 no exact fit for a small chunk.
3733 */
3734
3735 if (in_smallbin_range (nb) &&
3736 bck == unsorted_chunks (av) &&
3737 victim == av->last_remainder &&
3738 (unsigned long) (size) > (unsigned long) (nb + MINSIZE))
3739 {
3740 /* split and reattach remainder */
3741 remainder_size = size - nb;
3742 remainder = chunk_at_offset (victim, nb);
3743 unsorted_chunks (av)->bk = unsorted_chunks (av)->fd = remainder;
3744 av->last_remainder = remainder;
3745 remainder->bk = remainder->fd = unsorted_chunks (av);
3746 if (!in_smallbin_range (remainder_size))
3747 {
3748 remainder->fd_nextsize = NULL;
3749 remainder->bk_nextsize = NULL;
3750 }
3751
3752 set_head (victim, nb | PREV_INUSE |
3753 (av != &main_arena ? NON_MAIN_ARENA : 0));
3754 set_head (remainder, remainder_size | PREV_INUSE);
3755 set_foot (remainder, remainder_size);
3756
3757 check_malloced_chunk (av, victim, nb);
3758 void *p = chunk2mem (victim);
3759 alloc_perturb (p, bytes);
3760 return p;
3761 }
3762
3763 /* remove from unsorted list */
3764 unsorted_chunks (av)->bk = bck;
3765 bck->fd = unsorted_chunks (av);
3766
3767 /* Take now instead of binning if exact fit */
3768
3769 if (size == nb)
3770 {
3771 set_inuse_bit_at_offset (victim, size);
3772 if (av != &main_arena)
3773 set_non_main_arena (victim);
3774#if USE_TCACHE
3775 /* Fill cache first, return to user only if cache fills.
3776 We may return one of these chunks later. */
3777 if (tcache_nb
3778 && tcache->counts[tc_idx] < mp_.tcache_count)
3779 {
3780 tcache_put (victim, tc_idx);
3781 return_cached = 1;
3782 continue;
3783 }
3784 else
3785 {
3786#endif
3787 check_malloced_chunk (av, victim, nb);
3788 void *p = chunk2mem (victim);
3789 alloc_perturb (p, bytes);
3790 return p;
3791#if USE_TCACHE
3792 }
3793#endif
3794 }
3795
3796 /* place chunk in bin */
3797
3798 if (in_smallbin_range (size))
3799 {
3800 victim_index = smallbin_index (size);
3801 bck = bin_at (av, victim_index);
3802 fwd = bck->fd;
3803 }
3804 else
3805 {
3806 victim_index = largebin_index (size);
3807 bck = bin_at (av, victim_index);
3808 fwd = bck->fd;
3809
3810 /* maintain large bins in sorted order */
3811 if (fwd != bck)
3812 {
3813 /* Or with inuse bit to speed comparisons */
3814 size |= PREV_INUSE;
3815 /* if smaller than smallest, bypass loop below */
3816 assert (chunk_main_arena (bck->bk));
3817 if ((unsigned long) (size)
3818 < (unsigned long) chunksize_nomask (bck->bk))
3819 {
3820 fwd = bck;
3821 bck = bck->bk;
3822
3823 victim->fd_nextsize = fwd->fd;
3824 victim->bk_nextsize = fwd->fd->bk_nextsize;
3825 fwd->fd->bk_nextsize = victim->bk_nextsize->fd_nextsize = victim;
3826 }
3827 else
3828 {
3829 assert (chunk_main_arena (fwd));
3830 while ((unsigned long) size < chunksize_nomask (fwd))
3831 {
3832 fwd = fwd->fd_nextsize;
3833 assert (chunk_main_arena (fwd));
3834 }
3835
3836 if ((unsigned long) size
3837 == (unsigned long) chunksize_nomask (fwd))
3838 /* Always insert in the second position. */
3839 fwd = fwd->fd;
3840 else
3841 {
3842 victim->fd_nextsize = fwd;
3843 victim->bk_nextsize = fwd->bk_nextsize;
3844 fwd->bk_nextsize = victim;
3845 victim->bk_nextsize->fd_nextsize = victim;
3846 }
3847 bck = fwd->bk;
3848 }
3849 }
3850 else
3851 victim->fd_nextsize = victim->bk_nextsize = victim;
3852 }
3853
3854 mark_bin (av, victim_index);
3855 victim->bk = bck;
3856 victim->fd = fwd;
3857 fwd->bk = victim;
3858 bck->fd = victim;
3859
3860#if USE_TCACHE
3861 /* If we've processed as many chunks as we're allowed while
3862 filling the cache, return one of the cached ones. */
3863 ++tcache_unsorted_count;
3864 if (return_cached
3865 && mp_.tcache_unsorted_limit > 0
3866 && tcache_unsorted_count > mp_.tcache_unsorted_limit)
3867 {
3868 return tcache_get (tc_idx);
3869 }
3870#endif
3871
3872#define MAX_ITERS 10000
3873 if (++iters >= MAX_ITERS)
3874 break;
3875 }
3876
3877#if USE_TCACHE
3878 /* If all the small chunks we found ended up cached, return one now. */
3879 if (return_cached)
3880 {
3881 return tcache_get (tc_idx);
3882 }
3883#endif
3884
3885 /*
3886 If a large request, scan through the chunks of current bin in
3887 sorted order to find smallest that fits. Use the skip list for this.
3888 */
3889
3890 if (!in_smallbin_range (nb))
3891 {
3892 bin = bin_at (av, idx);
3893
3894 /* skip scan if empty or largest chunk is too small */
3895 if ((victim = first (bin)) != bin
3896 && (unsigned long) chunksize_nomask (victim)
3897 >= (unsigned long) (nb))
3898 {
3899 victim = victim->bk_nextsize;
3900 while (((unsigned long) (size = chunksize (victim)) <
3901 (unsigned long) (nb)))
3902 victim = victim->bk_nextsize;
3903
3904 /* Avoid removing the first entry for a size so that the skip
3905 list does not have to be rerouted. */
3906 if (victim != last (bin)
3907 && chunksize_nomask (victim)
3908 == chunksize_nomask (victim->fd))
3909 victim = victim->fd;
3910
3911 remainder_size = size - nb;
3912 unlink (av, victim, bck, fwd);
3913
3914 /* Exhaust */
3915 if (remainder_size < MINSIZE)
3916 {
3917 set_inuse_bit_at_offset (victim, size);
3918 if (av != &main_arena)
3919 set_non_main_arena (victim);
3920 }
3921 /* Split */
3922 else
3923 {
3924 remainder = chunk_at_offset (victim, nb);
3925 /* We cannot assume the unsorted list is empty and therefore
3926 have to perform a complete insert here. */
3927 bck = unsorted_chunks (av);
3928 fwd = bck->fd;
3929 if (__glibc_unlikely (fwd->bk != bck))
3930 malloc_printerr ("malloc(): corrupted unsorted chunks");
3931 remainder->bk = bck;
3932 remainder->fd = fwd;
3933 bck->fd = remainder;
3934 fwd->bk = remainder;
3935 if (!in_smallbin_range (remainder_size))
3936 {
3937 remainder->fd_nextsize = NULL;
3938 remainder->bk_nextsize = NULL;
3939 }
3940 set_head (victim, nb | PREV_INUSE |
3941 (av != &main_arena ? NON_MAIN_ARENA : 0));
3942 set_head (remainder, remainder_size | PREV_INUSE);
3943 set_foot (remainder, remainder_size);
3944 }
3945 check_malloced_chunk (av, victim, nb);
3946 void *p = chunk2mem (victim);
3947 alloc_perturb (p, bytes);
3948 return p;
3949 }
3950 }
3951
3952 /*
3953 Search for a chunk by scanning bins, starting with next largest
3954 bin. This search is strictly by best-fit; i.e., the smallest
3955 (with ties going to approximately the least recently used) chunk
3956 that fits is selected.
3957
3958 The bitmap avoids needing to check that most blocks are nonempty.
3959 The particular case of skipping all bins during warm-up phases
3960 when no chunks have been returned yet is faster than it might look.
3961 */
3962
3963 ++idx;
3964 bin = bin_at (av, idx);
3965 block = idx2block (idx);
3966 map = av->binmap[block];
3967 bit = idx2bit (idx);
3968
3969 for (;; )
3970 {
3971 /* Skip rest of block if there are no more set bits in this block. */
3972 if (bit > map || bit == 0)
3973 {
3974 do
3975 {
3976 if (++block >= BINMAPSIZE) /* out of bins */
3977 goto use_top;
3978 }
3979 while ((map = av->binmap[block]) == 0);
3980
3981 bin = bin_at (av, (block << BINMAPSHIFT));
3982 bit = 1;
3983 }
3984
3985 /* Advance to bin with set bit. There must be one. */
3986 while ((bit & map) == 0)
3987 {
3988 bin = next_bin (bin);
3989 bit <<= 1;
3990 assert (bit != 0);
3991 }
3992
3993 /* Inspect the bin. It is likely to be non-empty */
3994 victim = last (bin);
3995
3996 /* If a false alarm (empty bin), clear the bit. */
3997 if (victim == bin)
3998 {
3999 av->binmap[block] = map &= ~bit; /* Write through */
4000 bin = next_bin (bin);
4001 bit <<= 1;
4002 }
4003
4004 else
4005 {
4006 size = chunksize (victim);
4007
4008 /* We know the first chunk in this bin is big enough to use. */
4009 assert ((unsigned long) (size) >= (unsigned long) (nb));
4010
4011 remainder_size = size - nb;
4012
4013 /* unlink */
4014 unlink (av, victim, bck, fwd);
4015
4016 /* Exhaust */
4017 if (remainder_size < MINSIZE)
4018 {
4019 set_inuse_bit_at_offset (victim, size);
4020 if (av != &main_arena)
4021 set_non_main_arena (victim);
4022 }
4023
4024 /* Split */
4025 else
4026 {
4027 remainder = chunk_at_offset (victim, nb);
4028
4029 /* We cannot assume the unsorted list is empty and therefore
4030 have to perform a complete insert here. */
4031 bck = unsorted_chunks (av);
4032 fwd = bck->fd;
4033 if (__glibc_unlikely (fwd->bk != bck))
4034 malloc_printerr ("malloc(): corrupted unsorted chunks 2");
4035 remainder->bk = bck;
4036 remainder->fd = fwd;
4037 bck->fd = remainder;
4038 fwd->bk = remainder;
4039
4040 /* advertise as last remainder */
4041 if (in_smallbin_range (nb))
4042 av->last_remainder = remainder;
4043 if (!in_smallbin_range (remainder_size))
4044 {
4045 remainder->fd_nextsize = NULL;
4046 remainder->bk_nextsize = NULL;
4047 }
4048 set_head (victim, nb | PREV_INUSE |
4049 (av != &main_arena ? NON_MAIN_ARENA : 0));
4050 set_head (remainder, remainder_size | PREV_INUSE);
4051 set_foot (remainder, remainder_size);
4052 }
4053 check_malloced_chunk (av, victim, nb);
4054 void *p = chunk2mem (victim);
4055 alloc_perturb (p, bytes);
4056 return p;
4057 }
4058 }
4059
4060 use_top:
4061 /*
4062 If large enough, split off the chunk bordering the end of memory
4063 (held in av->top). Note that this is in accord with the best-fit
4064 search rule. In effect, av->top is treated as larger (and thus
4065 less well fitting) than any other available chunk since it can
4066 be extended to be as large as necessary (up to system
4067 limitations).
4068
4069 We require that av->top always exists (i.e., has size >=
4070 MINSIZE) after initialization, so if it would otherwise be
4071 exhausted by current request, it is replenished. (The main
4072 reason for ensuring it exists is that we may need MINSIZE space
4073 to put in fenceposts in sysmalloc.)
4074 */
4075
4076 victim = av->top;
4077 size = chunksize (victim);
4078
4079 if ((unsigned long) (size) >= (unsigned long) (nb + MINSIZE))
4080 {
4081 remainder_size = size - nb;
4082 remainder = chunk_at_offset (victim, nb);
4083 av->top = remainder;
4084 set_head (victim, nb | PREV_INUSE |
4085 (av != &main_arena ? NON_MAIN_ARENA : 0));
4086 set_head (remainder, remainder_size | PREV_INUSE);
4087
4088 check_malloced_chunk (av, victim, nb);
4089 void *p = chunk2mem (victim);
4090 alloc_perturb (p, bytes);
4091 return p;
4092 }
4093
4094 /* When we are using atomic ops to free fast chunks we can get
4095 here for all block sizes. */
4096 else if (atomic_load_relaxed (&av->have_fastchunks))
4097 {
4098 malloc_consolidate (av);
4099 /* restore original bin index */
4100 if (in_smallbin_range (nb))
4101 idx = smallbin_index (nb);
4102 else
4103 idx = largebin_index (nb);
4104 }
4105
4106 /*
4107 Otherwise, relay to handle system-dependent cases
4108 */
4109 else
4110 {
4111 void *p = sysmalloc (nb, av);
4112 if (p != NULL)
4113 alloc_perturb (p, bytes);
4114 return p;
4115 }
4116 }
4117}
4118
4119/*
4120 ------------------------------ free ------------------------------
4121 */
4122
4123static void
4124_int_free (mstate av, mchunkptr p, int have_lock)
4125{
4126 INTERNAL_SIZE_T size; /* its size */
4127 mfastbinptr *fb; /* associated fastbin */
4128 mchunkptr nextchunk; /* next contiguous chunk */
4129 INTERNAL_SIZE_T nextsize; /* its size */
4130 int nextinuse; /* true if nextchunk is used */
4131 INTERNAL_SIZE_T prevsize; /* size of previous contiguous chunk */
4132 mchunkptr bck; /* misc temp for linking */
4133 mchunkptr fwd; /* misc temp for linking */
4134
4135 size = chunksize (p);
4136
4137 /* Little security check which won't hurt performance: the
4138 allocator never wrapps around at the end of the address space.
4139 Therefore we can exclude some size values which might appear
4140 here by accident or by "design" from some intruder. */
4141 if (__builtin_expect ((uintptr_t) p > (uintptr_t) -size, 0)
4142 || __builtin_expect (misaligned_chunk (p), 0))
4143 malloc_printerr ("free(): invalid pointer");
4144 /* We know that each chunk is at least MINSIZE bytes in size or a
4145 multiple of MALLOC_ALIGNMENT. */
4146 if (__glibc_unlikely (size < MINSIZE || !aligned_OK (size)))
4147 malloc_printerr ("free(): invalid size");
4148
4149 check_inuse_chunk(av, p);
4150
4151#if USE_TCACHE
4152 {
4153 size_t tc_idx = csize2tidx (size);
4154
4155 if (tcache
4156 && tc_idx < mp_.tcache_bins
4157 && tcache->counts[tc_idx] < mp_.tcache_count)
4158 {
4159 tcache_put (p, tc_idx);
4160 return;
4161 }
4162 }
4163#endif
4164
4165 /*
4166 If eligible, place chunk on a fastbin so it can be found
4167 and used quickly in malloc.
4168 */
4169
4170 if ((unsigned long)(size) <= (unsigned long)(get_max_fast ())
4171
4172#if TRIM_FASTBINS
4173 /*
4174 If TRIM_FASTBINS set, don't place chunks
4175 bordering top into fastbins
4176 */
4177 && (chunk_at_offset(p, size) != av->top)
4178#endif
4179 ) {
4180
4181 if (__builtin_expect (chunksize_nomask (chunk_at_offset (p, size))
4182 <= 2 * SIZE_SZ, 0)
4183 || __builtin_expect (chunksize (chunk_at_offset (p, size))
4184 >= av->system_mem, 0))
4185 {
4186 bool fail = true;
4187 /* We might not have a lock at this point and concurrent modifications
4188 of system_mem might result in a false positive. Redo the test after
4189 getting the lock. */
4190 if (!have_lock)
4191 {
4192 __libc_lock_lock (av->mutex);
4193 fail = (chunksize_nomask (chunk_at_offset (p, size)) <= 2 * SIZE_SZ
4194 || chunksize (chunk_at_offset (p, size)) >= av->system_mem);
4195 __libc_lock_unlock (av->mutex);
4196 }
4197
4198 if (fail)
4199 malloc_printerr ("free(): invalid next size (fast)");
4200 }
4201
4202 free_perturb (chunk2mem(p), size - 2 * SIZE_SZ);
4203
4204 atomic_store_relaxed (&av->have_fastchunks, true);
4205 unsigned int idx = fastbin_index(size);
4206 fb = &fastbin (av, idx);
4207
4208 /* Atomically link P to its fastbin: P->FD = *FB; *FB = P; */
4209 mchunkptr old = *fb, old2;
4210
4211 if (SINGLE_THREAD_P)
4212 {
4213 /* Check that the top of the bin is not the record we are going to
4214 add (i.e., double free). */
4215 if (__builtin_expect (old == p, 0))
4216 malloc_printerr ("double free or corruption (fasttop)");
4217 p->fd = old;
4218 *fb = p;
4219 }
4220 else
4221 do
4222 {
4223 /* Check that the top of the bin is not the record we are going to
4224 add (i.e., double free). */
4225 if (__builtin_expect (old == p, 0))
4226 malloc_printerr ("double free or corruption (fasttop)");
4227 p->fd = old2 = old;
4228 }
4229 while ((old = catomic_compare_and_exchange_val_rel (fb, p, old2))
4230 != old2);
4231
4232 /* Check that size of fastbin chunk at the top is the same as
4233 size of the chunk that we are adding. We can dereference OLD
4234 only if we have the lock, otherwise it might have already been
4235 allocated again. */
4236 if (have_lock && old != NULL
4237 && __builtin_expect (fastbin_index (chunksize (old)) != idx, 0))
4238 malloc_printerr ("invalid fastbin entry (free)");
4239 }
4240
4241 /*
4242 Consolidate other non-mmapped chunks as they arrive.
4243 */
4244
4245 else if (!chunk_is_mmapped(p)) {
4246
4247 /* If we're single-threaded, don't lock the arena. */
4248 if (SINGLE_THREAD_P)
4249 have_lock = true;
4250
4251 if (!have_lock)
4252 __libc_lock_lock (av->mutex);
4253
4254 nextchunk = chunk_at_offset(p, size);
4255
4256 /* Lightweight tests: check whether the block is already the
4257 top block. */
4258 if (__glibc_unlikely (p == av->top))
4259 malloc_printerr ("double free or corruption (top)");
4260 /* Or whether the next chunk is beyond the boundaries of the arena. */
4261 if (__builtin_expect (contiguous (av)
4262 && (char *) nextchunk
4263 >= ((char *) av->top + chunksize(av->top)), 0))
4264 malloc_printerr ("double free or corruption (out)");
4265 /* Or whether the block is actually not marked used. */
4266 if (__glibc_unlikely (!prev_inuse(nextchunk)))
4267 malloc_printerr ("double free or corruption (!prev)");
4268
4269 nextsize = chunksize(nextchunk);
4270 if (__builtin_expect (chunksize_nomask (nextchunk) <= 2 * SIZE_SZ, 0)
4271 || __builtin_expect (nextsize >= av->system_mem, 0))
4272 malloc_printerr ("free(): invalid next size (normal)");
4273
4274 free_perturb (chunk2mem(p), size - 2 * SIZE_SZ);
4275
4276 /* consolidate backward */
4277 if (!prev_inuse(p)) {
4278 prevsize = prev_size (p);
4279 size += prevsize;
4280 p = chunk_at_offset(p, -((long) prevsize));
4281 unlink(av, p, bck, fwd);
4282 }
4283
4284 if (nextchunk != av->top) {
4285 /* get and clear inuse bit */
4286 nextinuse = inuse_bit_at_offset(nextchunk, nextsize);
4287
4288 /* consolidate forward */
4289 if (!nextinuse) {
4290 unlink(av, nextchunk, bck, fwd);
4291 size += nextsize;
4292 } else
4293 clear_inuse_bit_at_offset(nextchunk, 0);
4294
4295 /*
4296 Place the chunk in unsorted chunk list. Chunks are
4297 not placed into regular bins until after they have
4298 been given one chance to be used in malloc.
4299 */
4300
4301 bck = unsorted_chunks(av);
4302 fwd = bck->fd;
4303 if (__glibc_unlikely (fwd->bk != bck))
4304 malloc_printerr ("free(): corrupted unsorted chunks");
4305 p->fd = fwd;
4306 p->bk = bck;
4307 if (!in_smallbin_range(size))
4308 {
4309 p->fd_nextsize = NULL;
4310 p->bk_nextsize = NULL;
4311 }
4312 bck->fd = p;
4313 fwd->bk = p;
4314
4315 set_head(p, size | PREV_INUSE);
4316 set_foot(p, size);
4317
4318 check_free_chunk(av, p);
4319 }
4320
4321 /*
4322 If the chunk borders the current high end of memory,
4323 consolidate into top
4324 */
4325
4326 else {
4327 size += nextsize;
4328 set_head(p, size | PREV_INUSE);
4329 av->top = p;
4330 check_chunk(av, p);
4331 }
4332
4333 /*
4334 If freeing a large space, consolidate possibly-surrounding
4335 chunks. Then, if the total unused topmost memory exceeds trim
4336 threshold, ask malloc_trim to reduce top.
4337
4338 Unless max_fast is 0, we don't know if there are fastbins
4339 bordering top, so we cannot tell for sure whether threshold
4340 has been reached unless fastbins are consolidated. But we
4341 don't want to consolidate on each free. As a compromise,
4342 consolidation is performed if FASTBIN_CONSOLIDATION_THRESHOLD
4343 is reached.
4344 */
4345
4346 if ((unsigned long)(size) >= FASTBIN_CONSOLIDATION_THRESHOLD) {
4347 if (atomic_load_relaxed (&av->have_fastchunks))
4348 malloc_consolidate(av);
4349
4350 if (av == &main_arena) {
4351#ifndef MORECORE_CANNOT_TRIM
4352 if ((unsigned long)(chunksize(av->top)) >=
4353 (unsigned long)(mp_.trim_threshold))
4354 systrim(mp_.top_pad, av);
4355#endif
4356 } else {
4357 /* Always try heap_trim(), even if the top chunk is not
4358 large, because the corresponding heap might go away. */
4359 heap_info *heap = heap_for_ptr(top(av));
4360
4361 assert(heap->ar_ptr == av);
4362 heap_trim(heap, mp_.top_pad);
4363 }
4364 }
4365
4366 if (!have_lock)
4367 __libc_lock_unlock (av->mutex);
4368 }
4369 /*
4370 If the chunk was allocated via mmap, release via munmap().
4371 */
4372
4373 else {
4374 munmap_chunk (p);
4375 }
4376}
4377
4378/*
4379 ------------------------- malloc_consolidate -------------------------
4380
4381 malloc_consolidate is a specialized version of free() that tears
4382 down chunks held in fastbins. Free itself cannot be used for this
4383 purpose since, among other things, it might place chunks back onto
4384 fastbins. So, instead, we need to use a minor variant of the same
4385 code.
4386*/
4387
4388static void malloc_consolidate(mstate av)
4389{
4390 mfastbinptr* fb; /* current fastbin being consolidated */
4391 mfastbinptr* maxfb; /* last fastbin (for loop control) */
4392 mchunkptr p; /* current chunk being consolidated */
4393 mchunkptr nextp; /* next chunk to consolidate */
4394 mchunkptr unsorted_bin; /* bin header */
4395 mchunkptr first_unsorted; /* chunk to link to */
4396
4397 /* These have same use as in free() */
4398 mchunkptr nextchunk;
4399 INTERNAL_SIZE_T size;
4400 INTERNAL_SIZE_T nextsize;
4401 INTERNAL_SIZE_T prevsize;
4402 int nextinuse;
4403 mchunkptr bck;
4404 mchunkptr fwd;
4405
4406 atomic_store_relaxed (&av->have_fastchunks, false);
4407
4408 unsorted_bin = unsorted_chunks(av);
4409
4410 /*
4411 Remove each chunk from fast bin and consolidate it, placing it
4412 then in unsorted bin. Among other reasons for doing this,
4413 placing in unsorted bin avoids needing to calculate actual bins
4414 until malloc is sure that chunks aren't immediately going to be
4415 reused anyway.
4416 */
4417
4418 maxfb = &fastbin (av, NFASTBINS - 1);
4419 fb = &fastbin (av, 0);
4420 do {
4421 p = atomic_exchange_acq (fb, NULL);
4422 if (p != 0) {
4423 do {
4424 check_inuse_chunk(av, p);
4425 nextp = p->fd;
4426
4427 /* Slightly streamlined version of consolidation code in free() */
4428 size = chunksize (p);
4429 nextchunk = chunk_at_offset(p, size);
4430 nextsize = chunksize(nextchunk);
4431
4432 if (!prev_inuse(p)) {
4433 prevsize = prev_size (p);
4434 size += prevsize;
4435 p = chunk_at_offset(p, -((long) prevsize));
4436 unlink(av, p, bck, fwd);
4437 }
4438
4439 if (nextchunk != av->top) {
4440 nextinuse = inuse_bit_at_offset(nextchunk, nextsize);
4441
4442 if (!nextinuse) {
4443 size += nextsize;
4444 unlink(av, nextchunk, bck, fwd);
4445 } else
4446 clear_inuse_bit_at_offset(nextchunk, 0);
4447
4448 first_unsorted = unsorted_bin->fd;
4449 unsorted_bin->fd = p;
4450 first_unsorted->bk = p;
4451
4452 if (!in_smallbin_range (size)) {
4453 p->fd_nextsize = NULL;
4454 p->bk_nextsize = NULL;
4455 }
4456
4457 set_head(p, size | PREV_INUSE);
4458 p->bk = unsorted_bin;
4459 p->fd = first_unsorted;
4460 set_foot(p, size);
4461 }
4462
4463 else {
4464 size += nextsize;
4465 set_head(p, size | PREV_INUSE);
4466 av->top = p;
4467 }
4468
4469 } while ( (p = nextp) != 0);
4470
4471 }
4472 } while (fb++ != maxfb);
4473}
4474
4475/*
4476 ------------------------------ realloc ------------------------------
4477*/
4478
4479void*
4480_int_realloc(mstate av, mchunkptr oldp, INTERNAL_SIZE_T oldsize,
4481 INTERNAL_SIZE_T nb)
4482{
4483 mchunkptr newp; /* chunk to return */
4484 INTERNAL_SIZE_T newsize; /* its size */
4485 void* newmem; /* corresponding user mem */
4486
4487 mchunkptr next; /* next contiguous chunk after oldp */
4488
4489 mchunkptr remainder; /* extra space at end of newp */
4490 unsigned long remainder_size; /* its size */
4491
4492 mchunkptr bck; /* misc temp for linking */
4493 mchunkptr fwd; /* misc temp for linking */
4494
4495 unsigned long copysize; /* bytes to copy */
4496 unsigned int ncopies; /* INTERNAL_SIZE_T words to copy */
4497 INTERNAL_SIZE_T* s; /* copy source */
4498 INTERNAL_SIZE_T* d; /* copy destination */
4499
4500 /* oldmem size */
4501 if (__builtin_expect (chunksize_nomask (oldp) <= 2 * SIZE_SZ, 0)
4502 || __builtin_expect (oldsize >= av->system_mem, 0))
4503 malloc_printerr ("realloc(): invalid old size");
4504
4505 check_inuse_chunk (av, oldp);
4506
4507 /* All callers already filter out mmap'ed chunks. */
4508 assert (!chunk_is_mmapped (oldp));
4509
4510 next = chunk_at_offset (oldp, oldsize);
4511 INTERNAL_SIZE_T nextsize = chunksize (next);
4512 if (__builtin_expect (chunksize_nomask (next) <= 2 * SIZE_SZ, 0)
4513 || __builtin_expect (nextsize >= av->system_mem, 0))
4514 malloc_printerr ("realloc(): invalid next size");
4515
4516 if ((unsigned long) (oldsize) >= (unsigned long) (nb))
4517 {
4518 /* already big enough; split below */
4519 newp = oldp;
4520 newsize = oldsize;
4521 }
4522
4523 else
4524 {
4525 /* Try to expand forward into top */
4526 if (next == av->top &&
4527 (unsigned long) (newsize = oldsize + nextsize) >=
4528 (unsigned long) (nb + MINSIZE))
4529 {
4530 set_head_size (oldp, nb | (av != &main_arena ? NON_MAIN_ARENA : 0));
4531 av->top = chunk_at_offset (oldp, nb);
4532 set_head (av->top, (newsize - nb) | PREV_INUSE);
4533 check_inuse_chunk (av, oldp);
4534 return chunk2mem (oldp);
4535 }
4536
4537 /* Try to expand forward into next chunk; split off remainder below */
4538 else if (next != av->top &&
4539 !inuse (next) &&
4540 (unsigned long) (newsize = oldsize + nextsize) >=
4541 (unsigned long) (nb))
4542 {
4543 newp = oldp;
4544 unlink (av, next, bck, fwd);
4545 }
4546
4547 /* allocate, copy, free */
4548 else
4549 {
4550 newmem = _int_malloc (av, nb - MALLOC_ALIGN_MASK);
4551 if (newmem == 0)
4552 return 0; /* propagate failure */
4553
4554 newp = mem2chunk (newmem);
4555 newsize = chunksize (newp);
4556
4557 /*
4558 Avoid copy if newp is next chunk after oldp.
4559 */
4560 if (newp == next)
4561 {
4562 newsize += oldsize;
4563 newp = oldp;
4564 }
4565 else
4566 {
4567 /*
4568 Unroll copy of <= 36 bytes (72 if 8byte sizes)
4569 We know that contents have an odd number of
4570 INTERNAL_SIZE_T-sized words; minimally 3.
4571 */
4572
4573 copysize = oldsize - SIZE_SZ;
4574 s = (INTERNAL_SIZE_T *) (chunk2mem (oldp));
4575 d = (INTERNAL_SIZE_T *) (newmem);
4576 ncopies = copysize / sizeof (INTERNAL_SIZE_T);
4577 assert (ncopies >= 3);
4578
4579 if (ncopies > 9)
4580 memcpy (d, s, copysize);
4581
4582 else
4583 {
4584 *(d + 0) = *(s + 0);
4585 *(d + 1) = *(s + 1);
4586 *(d + 2) = *(s + 2);
4587 if (ncopies > 4)
4588 {
4589 *(d + 3) = *(s + 3);
4590 *(d + 4) = *(s + 4);
4591 if (ncopies > 6)
4592 {
4593 *(d + 5) = *(s + 5);
4594 *(d + 6) = *(s + 6);
4595 if (ncopies > 8)
4596 {
4597 *(d + 7) = *(s + 7);
4598 *(d + 8) = *(s + 8);
4599 }
4600 }
4601 }
4602 }
4603
4604 _int_free (av, oldp, 1);
4605 check_inuse_chunk (av, newp);
4606 return chunk2mem (newp);
4607 }
4608 }
4609 }
4610
4611 /* If possible, free extra space in old or extended chunk */
4612
4613 assert ((unsigned long) (newsize) >= (unsigned long) (nb));
4614
4615 remainder_size = newsize - nb;
4616
4617 if (remainder_size < MINSIZE) /* not enough extra to split off */
4618 {
4619 set_head_size (newp, newsize | (av != &main_arena ? NON_MAIN_ARENA : 0));
4620 set_inuse_bit_at_offset (newp, newsize);
4621 }
4622 else /* split remainder */
4623 {
4624 remainder = chunk_at_offset (newp, nb);
4625 set_head_size (newp, nb | (av != &main_arena ? NON_MAIN_ARENA : 0));
4626 set_head (remainder, remainder_size | PREV_INUSE |
4627 (av != &main_arena ? NON_MAIN_ARENA : 0));
4628 /* Mark remainder as inuse so free() won't complain */
4629 set_inuse_bit_at_offset (remainder, remainder_size);
4630 _int_free (av, remainder, 1);
4631 }
4632
4633 check_inuse_chunk (av, newp);
4634 return chunk2mem (newp);
4635}
4636
4637/*
4638 ------------------------------ memalign ------------------------------
4639 */
4640
4641static void *
4642_int_memalign (mstate av, size_t alignment, size_t bytes)
4643{
4644 INTERNAL_SIZE_T nb; /* padded request size */
4645 char *m; /* memory returned by malloc call */
4646 mchunkptr p; /* corresponding chunk */
4647 char *brk; /* alignment point within p */
4648 mchunkptr newp; /* chunk to return */
4649 INTERNAL_SIZE_T newsize; /* its size */
4650 INTERNAL_SIZE_T leadsize; /* leading space before alignment point */
4651 mchunkptr remainder; /* spare room at end to split off */
4652 unsigned long remainder_size; /* its size */
4653 INTERNAL_SIZE_T size;
4654
4655
4656
4657 checked_request2size (bytes, nb);
4658
4659 /*
4660 Strategy: find a spot within that chunk that meets the alignment
4661 request, and then possibly free the leading and trailing space.
4662 */
4663
4664
4665 /* Call malloc with worst case padding to hit alignment. */
4666
4667 m = (char *) (_int_malloc (av, nb + alignment + MINSIZE));
4668
4669 if (m == 0)
4670 return 0; /* propagate failure */
4671
4672 p = mem2chunk (m);
4673
4674 if ((((unsigned long) (m)) % alignment) != 0) /* misaligned */
4675
4676 { /*
4677 Find an aligned spot inside chunk. Since we need to give back
4678 leading space in a chunk of at least MINSIZE, if the first
4679 calculation places us at a spot with less than MINSIZE leader,
4680 we can move to the next aligned spot -- we've allocated enough
4681 total room so that this is always possible.
4682 */
4683 brk = (char *) mem2chunk (((unsigned long) (m + alignment - 1)) &
4684 - ((signed long) alignment));
4685 if ((unsigned long) (brk - (char *) (p)) < MINSIZE)
4686 brk += alignment;
4687
4688 newp = (mchunkptr) brk;
4689 leadsize = brk - (char *) (p);
4690 newsize = chunksize (p) - leadsize;
4691
4692 /* For mmapped chunks, just adjust offset */
4693 if (chunk_is_mmapped (p))
4694 {
4695 set_prev_size (newp, prev_size (p) + leadsize);
4696 set_head (newp, newsize | IS_MMAPPED);
4697 return chunk2mem (newp);
4698 }
4699
4700 /* Otherwise, give back leader, use the rest */
4701 set_head (newp, newsize | PREV_INUSE |
4702 (av != &main_arena ? NON_MAIN_ARENA : 0));
4703 set_inuse_bit_at_offset (newp, newsize);
4704 set_head_size (p, leadsize | (av != &main_arena ? NON_MAIN_ARENA : 0));
4705 _int_free (av, p, 1);
4706 p = newp;
4707
4708 assert (newsize >= nb &&
4709 (((unsigned long) (chunk2mem (p))) % alignment) == 0);
4710 }
4711
4712 /* Also give back spare room at the end */
4713 if (!chunk_is_mmapped (p))
4714 {
4715 size = chunksize (p);
4716 if ((unsigned long) (size) > (unsigned long) (nb + MINSIZE))
4717 {
4718 remainder_size = size - nb;
4719 remainder = chunk_at_offset (p, nb);
4720 set_head (remainder, remainder_size | PREV_INUSE |
4721 (av != &main_arena ? NON_MAIN_ARENA : 0));
4722 set_head_size (p, nb);
4723 _int_free (av, remainder, 1);
4724 }
4725 }
4726
4727 check_inuse_chunk (av, p);
4728 return chunk2mem (p);
4729}
4730
4731
4732/*
4733 ------------------------------ malloc_trim ------------------------------
4734 */
4735
4736static int
4737mtrim (mstate av, size_t pad)
4738{
4739 /* Ensure all blocks are consolidated. */
4740 malloc_consolidate (av);
4741
4742 const size_t ps = GLRO (dl_pagesize);
4743 int psindex = bin_index (ps);
4744 const size_t psm1 = ps - 1;
4745
4746 int result = 0;
4747 for (int i = 1; i < NBINS; ++i)
4748 if (i == 1 || i >= psindex)
4749 {
4750 mbinptr bin = bin_at (av, i);
4751
4752 for (mchunkptr p = last (bin); p != bin; p = p->bk)
4753 {
4754 INTERNAL_SIZE_T size = chunksize (p);
4755
4756 if (size > psm1 + sizeof (struct malloc_chunk))
4757 {
4758 /* See whether the chunk contains at least one unused page. */
4759 char *paligned_mem = (char *) (((uintptr_t) p
4760 + sizeof (struct malloc_chunk)
4761 + psm1) & ~psm1);
4762
4763 assert ((char *) chunk2mem (p) + 4 * SIZE_SZ <= paligned_mem);
4764 assert ((char *) p + size > paligned_mem);
4765
4766 /* This is the size we could potentially free. */
4767 size -= paligned_mem - (char *) p;
4768
4769 if (size > psm1)
4770 {
4771#if MALLOC_DEBUG
4772 /* When debugging we simulate destroying the memory
4773 content. */
4774 memset (paligned_mem, 0x89, size & ~psm1);
4775#endif
4776 __madvise (paligned_mem, size & ~psm1, MADV_DONTNEED);
4777
4778 result = 1;
4779 }
4780 }
4781 }
4782 }
4783
4784#ifndef MORECORE_CANNOT_TRIM
4785 return result | (av == &main_arena ? systrim (pad, av) : 0);
4786
4787#else
4788 return result;
4789#endif
4790}
4791
4792
4793int
4794__malloc_trim (size_t s)
4795{
4796 int result = 0;
4797
4798 if (__malloc_initialized < 0)
4799 ptmalloc_init ();
4800
4801 mstate ar_ptr = &main_arena;
4802 do
4803 {
4804 __libc_lock_lock (ar_ptr->mutex);
4805 result |= mtrim (ar_ptr, s);
4806 __libc_lock_unlock (ar_ptr->mutex);
4807
4808 ar_ptr = ar_ptr->next;
4809 }
4810 while (ar_ptr != &main_arena);
4811
4812 return result;
4813}
4814
4815
4816/*
4817 ------------------------- malloc_usable_size -------------------------
4818 */
4819
4820static size_t
4821musable (void *mem)
4822{
4823 mchunkptr p;
4824 if (mem != 0)
4825 {
4826 p = mem2chunk (mem);
4827
4828 if (__builtin_expect (using_malloc_checking == 1, 0))
4829 return malloc_check_get_size (p);
4830
4831 if (chunk_is_mmapped (p))
4832 {
4833 if (DUMPED_MAIN_ARENA_CHUNK (p))
4834 return chunksize (p) - SIZE_SZ;
4835 else
4836 return chunksize (p) - 2 * SIZE_SZ;
4837 }
4838 else if (inuse (p))
4839 return chunksize (p) - SIZE_SZ;
4840 }
4841 return 0;
4842}
4843
4844
4845size_t
4846__malloc_usable_size (void *m)
4847{
4848 size_t result;
4849
4850 result = musable (m);
4851 return result;
4852}
4853
4854/*
4855 ------------------------------ mallinfo ------------------------------
4856 Accumulate malloc statistics for arena AV into M.
4857 */
4858
4859static void
4860int_mallinfo (mstate av, struct mallinfo *m)
4861{
4862 size_t i;
4863 mbinptr b;
4864 mchunkptr p;
4865 INTERNAL_SIZE_T avail;
4866 INTERNAL_SIZE_T fastavail;
4867 int nblocks;
4868 int nfastblocks;
4869
4870 check_malloc_state (av);
4871
4872 /* Account for top */
4873 avail = chunksize (av->top);
4874 nblocks = 1; /* top always exists */
4875
4876 /* traverse fastbins */
4877 nfastblocks = 0;
4878 fastavail = 0;
4879
4880 for (i = 0; i < NFASTBINS; ++i)
4881 {
4882 for (p = fastbin (av, i); p != 0; p = p->fd)
4883 {
4884 ++nfastblocks;
4885 fastavail += chunksize (p);
4886 }
4887 }
4888
4889 avail += fastavail;
4890
4891 /* traverse regular bins */
4892 for (i = 1; i < NBINS; ++i)
4893 {
4894 b = bin_at (av, i);
4895 for (p = last (b); p != b; p = p->bk)
4896 {
4897 ++nblocks;
4898 avail += chunksize (p);
4899 }
4900 }
4901
4902 m->smblks += nfastblocks;
4903 m->ordblks += nblocks;
4904 m->fordblks += avail;
4905 m->uordblks += av->system_mem - avail;
4906 m->arena += av->system_mem;
4907 m->fsmblks += fastavail;
4908 if (av == &main_arena)
4909 {
4910 m->hblks = mp_.n_mmaps;
4911 m->hblkhd = mp_.mmapped_mem;
4912 m->usmblks = 0;
4913 m->keepcost = chunksize (av->top);
4914 }
4915}
4916
4917
4918struct mallinfo
4919__libc_mallinfo (void)
4920{
4921 struct mallinfo m;
4922 mstate ar_ptr;
4923
4924 if (__malloc_initialized < 0)
4925 ptmalloc_init ();
4926
4927 memset (&m, 0, sizeof (m));
4928 ar_ptr = &main_arena;
4929 do
4930 {
4931 __libc_lock_lock (ar_ptr->mutex);
4932 int_mallinfo (ar_ptr, &m);
4933 __libc_lock_unlock (ar_ptr->mutex);
4934
4935 ar_ptr = ar_ptr->next;
4936 }
4937 while (ar_ptr != &main_arena);
4938
4939 return m;
4940}
4941
4942/*
4943 ------------------------------ malloc_stats ------------------------------
4944 */
4945
4946void
4947__malloc_stats (void)
4948{
4949 int i;
4950 mstate ar_ptr;
4951 unsigned int in_use_b = mp_.mmapped_mem, system_b = in_use_b;
4952
4953 if (__malloc_initialized < 0)
4954 ptmalloc_init ();
4955 _IO_flockfile (stderr);
4956 int old_flags2 = ((_IO_FILE *) stderr)->_flags2;
4957 ((_IO_FILE *) stderr)->_flags2 |= _IO_FLAGS2_NOTCANCEL;
4958 for (i = 0, ar_ptr = &main_arena;; i++)
4959 {
4960 struct mallinfo mi;
4961
4962 memset (&mi, 0, sizeof (mi));
4963 __libc_lock_lock (ar_ptr->mutex);
4964 int_mallinfo (ar_ptr, &mi);
4965 fprintf (stderr, "Arena %d:\n", i);
4966 fprintf (stderr, "system bytes = %10u\n", (unsigned int) mi.arena);
4967 fprintf (stderr, "in use bytes = %10u\n", (unsigned int) mi.uordblks);
4968#if MALLOC_DEBUG > 1
4969 if (i > 0)
4970 dump_heap (heap_for_ptr (top (ar_ptr)));
4971#endif
4972 system_b += mi.arena;
4973 in_use_b += mi.uordblks;
4974 __libc_lock_unlock (ar_ptr->mutex);
4975 ar_ptr = ar_ptr->next;
4976 if (ar_ptr == &main_arena)
4977 break;
4978 }
4979 fprintf (stderr, "Total (incl. mmap):\n");
4980 fprintf (stderr, "system bytes = %10u\n", system_b);
4981 fprintf (stderr, "in use bytes = %10u\n", in_use_b);
4982 fprintf (stderr, "max mmap regions = %10u\n", (unsigned int) mp_.max_n_mmaps);
4983 fprintf (stderr, "max mmap bytes = %10lu\n",
4984 (unsigned long) mp_.max_mmapped_mem);
4985 ((_IO_FILE *) stderr)->_flags2 |= old_flags2;
4986 _IO_funlockfile (stderr);
4987}
4988
4989
4990/*
4991 ------------------------------ mallopt ------------------------------
4992 */
4993static inline int
4994__always_inline
4995do_set_trim_threshold (size_t value)
4996{
4997 LIBC_PROBE (memory_mallopt_trim_threshold, 3, value, mp_.trim_threshold,
4998 mp_.no_dyn_threshold);
4999 mp_.trim_threshold = value;
5000 mp_.no_dyn_threshold = 1;
5001 return 1;
5002}
5003
5004static inline int
5005__always_inline
5006do_set_top_pad (size_t value)
5007{
5008 LIBC_PROBE (memory_mallopt_top_pad, 3, value, mp_.top_pad,
5009 mp_.no_dyn_threshold);
5010 mp_.top_pad = value;
5011 mp_.no_dyn_threshold = 1;
5012 return 1;
5013}
5014
5015static inline int
5016__always_inline
5017do_set_mmap_threshold (size_t value)
5018{
5019 /* Forbid setting the threshold too high. */
5020 if (value <= HEAP_MAX_SIZE / 2)
5021 {
5022 LIBC_PROBE (memory_mallopt_mmap_threshold, 3, value, mp_.mmap_threshold,
5023 mp_.no_dyn_threshold);
5024 mp_.mmap_threshold = value;
5025 mp_.no_dyn_threshold = 1;
5026 return 1;
5027 }
5028 return 0;
5029}
5030
5031static inline int
5032__always_inline
5033do_set_mmaps_max (int32_t value)
5034{
5035 LIBC_PROBE (memory_mallopt_mmap_max, 3, value, mp_.n_mmaps_max,
5036 mp_.no_dyn_threshold);
5037 mp_.n_mmaps_max = value;
5038 mp_.no_dyn_threshold = 1;
5039 return 1;
5040}
5041
5042static inline int
5043__always_inline
5044do_set_mallopt_check (int32_t value)
5045{
5046 return 1;
5047}
5048
5049static inline int
5050__always_inline
5051do_set_perturb_byte (int32_t value)
5052{
5053 LIBC_PROBE (memory_mallopt_perturb, 2, value, perturb_byte);
5054 perturb_byte = value;
5055 return 1;
5056}
5057
5058static inline int
5059__always_inline
5060do_set_arena_test (size_t value)
5061{
5062 LIBC_PROBE (memory_mallopt_arena_test, 2, value, mp_.arena_test);
5063 mp_.arena_test = value;
5064 return 1;
5065}
5066
5067static inline int
5068__always_inline
5069do_set_arena_max (size_t value)
5070{
5071 LIBC_PROBE (memory_mallopt_arena_max, 2, value, mp_.arena_max);
5072 mp_.arena_max = value;
5073 return 1;
5074}
5075
5076#if USE_TCACHE
5077static inline int
5078__always_inline
5079do_set_tcache_max (size_t value)
5080{
5081 if (value >= 0 && value <= MAX_TCACHE_SIZE)
5082 {
5083 LIBC_PROBE (memory_tunable_tcache_max_bytes, 2, value, mp_.tcache_max_bytes);
5084 mp_.tcache_max_bytes = value;
5085 mp_.tcache_bins = csize2tidx (request2size(value)) + 1;
5086 }
5087 return 1;
5088}
5089
5090static inline int
5091__always_inline
5092do_set_tcache_count (size_t value)
5093{
5094 LIBC_PROBE (memory_tunable_tcache_count, 2, value, mp_.tcache_count);
5095 mp_.tcache_count = value;
5096 return 1;
5097}
5098
5099static inline int
5100__always_inline
5101do_set_tcache_unsorted_limit (size_t value)
5102{
5103 LIBC_PROBE (memory_tunable_tcache_unsorted_limit, 2, value, mp_.tcache_unsorted_limit);
5104 mp_.tcache_unsorted_limit = value;
5105 return 1;
5106}
5107#endif
5108
5109int
5110__libc_mallopt (int param_number, int value)
5111{
5112 mstate av = &main_arena;
5113 int res = 1;
5114
5115 if (__malloc_initialized < 0)
5116 ptmalloc_init ();
5117 __libc_lock_lock (av->mutex);
5118
5119 LIBC_PROBE (memory_mallopt, 2, param_number, value);
5120
5121 /* We must consolidate main arena before changing max_fast
5122 (see definition of set_max_fast). */
5123 malloc_consolidate (av);
5124
5125 switch (param_number)
5126 {
5127 case M_MXFAST:
5128 if (value >= 0 && value <= MAX_FAST_SIZE)
5129 {
5130 LIBC_PROBE (memory_mallopt_mxfast, 2, value, get_max_fast ());
5131 set_max_fast (value);
5132 }
5133 else
5134 res = 0;
5135 break;
5136
5137 case M_TRIM_THRESHOLD:
5138 do_set_trim_threshold (value);
5139 break;
5140
5141 case M_TOP_PAD:
5142 do_set_top_pad (value);
5143 break;
5144
5145 case M_MMAP_THRESHOLD:
5146 res = do_set_mmap_threshold (value);
5147 break;
5148
5149 case M_MMAP_MAX:
5150 do_set_mmaps_max (value);
5151 break;
5152
5153 case M_CHECK_ACTION:
5154 do_set_mallopt_check (value);
5155 break;
5156
5157 case M_PERTURB:
5158 do_set_perturb_byte (value);
5159 break;
5160
5161 case M_ARENA_TEST:
5162 if (value > 0)
5163 do_set_arena_test (value);
5164 break;
5165
5166 case M_ARENA_MAX:
5167 if (value > 0)
5168 do_set_arena_max (value);
5169 break;
5170 }
5171 __libc_lock_unlock (av->mutex);
5172 return res;
5173}
5174libc_hidden_def (__libc_mallopt)
5175
5176
5177/*
5178 -------------------- Alternative MORECORE functions --------------------
5179 */
5180
5181
5182/*
5183 General Requirements for MORECORE.
5184
5185 The MORECORE function must have the following properties:
5186
5187 If MORECORE_CONTIGUOUS is false:
5188
5189 * MORECORE must allocate in multiples of pagesize. It will
5190 only be called with arguments that are multiples of pagesize.
5191
5192 * MORECORE(0) must return an address that is at least
5193 MALLOC_ALIGNMENT aligned. (Page-aligning always suffices.)
5194
5195 else (i.e. If MORECORE_CONTIGUOUS is true):
5196
5197 * Consecutive calls to MORECORE with positive arguments
5198 return increasing addresses, indicating that space has been
5199 contiguously extended.
5200
5201 * MORECORE need not allocate in multiples of pagesize.
5202 Calls to MORECORE need not have args of multiples of pagesize.
5203
5204 * MORECORE need not page-align.
5205
5206 In either case:
5207
5208 * MORECORE may allocate more memory than requested. (Or even less,
5209 but this will generally result in a malloc failure.)
5210
5211 * MORECORE must not allocate memory when given argument zero, but
5212 instead return one past the end address of memory from previous
5213 nonzero call. This malloc does NOT call MORECORE(0)
5214 until at least one call with positive arguments is made, so
5215 the initial value returned is not important.
5216
5217 * Even though consecutive calls to MORECORE need not return contiguous
5218 addresses, it must be OK for malloc'ed chunks to span multiple
5219 regions in those cases where they do happen to be contiguous.
5220
5221 * MORECORE need not handle negative arguments -- it may instead
5222 just return MORECORE_FAILURE when given negative arguments.
5223 Negative arguments are always multiples of pagesize. MORECORE
5224 must not misinterpret negative args as large positive unsigned
5225 args. You can suppress all such calls from even occurring by defining
5226 MORECORE_CANNOT_TRIM,
5227
5228 There is some variation across systems about the type of the
5229 argument to sbrk/MORECORE. If size_t is unsigned, then it cannot
5230 actually be size_t, because sbrk supports negative args, so it is
5231 normally the signed type of the same width as size_t (sometimes
5232 declared as "intptr_t", and sometimes "ptrdiff_t"). It doesn't much
5233 matter though. Internally, we use "long" as arguments, which should
5234 work across all reasonable possibilities.
5235
5236 Additionally, if MORECORE ever returns failure for a positive
5237 request, then mmap is used as a noncontiguous system allocator. This
5238 is a useful backup strategy for systems with holes in address spaces
5239 -- in this case sbrk cannot contiguously expand the heap, but mmap
5240 may be able to map noncontiguous space.
5241
5242 If you'd like mmap to ALWAYS be used, you can define MORECORE to be
5243 a function that always returns MORECORE_FAILURE.
5244
5245 If you are using this malloc with something other than sbrk (or its
5246 emulation) to supply memory regions, you probably want to set
5247 MORECORE_CONTIGUOUS as false. As an example, here is a custom
5248 allocator kindly contributed for pre-OSX macOS. It uses virtually
5249 but not necessarily physically contiguous non-paged memory (locked
5250 in, present and won't get swapped out). You can use it by
5251 uncommenting this section, adding some #includes, and setting up the
5252 appropriate defines above:
5253
5254 *#define MORECORE osMoreCore
5255 *#define MORECORE_CONTIGUOUS 0
5256
5257 There is also a shutdown routine that should somehow be called for
5258 cleanup upon program exit.
5259
5260 *#define MAX_POOL_ENTRIES 100
5261 *#define MINIMUM_MORECORE_SIZE (64 * 1024)
5262 static int next_os_pool;
5263 void *our_os_pools[MAX_POOL_ENTRIES];
5264
5265 void *osMoreCore(int size)
5266 {
5267 void *ptr = 0;
5268 static void *sbrk_top = 0;
5269
5270 if (size > 0)
5271 {
5272 if (size < MINIMUM_MORECORE_SIZE)
5273 size = MINIMUM_MORECORE_SIZE;
5274 if (CurrentExecutionLevel() == kTaskLevel)
5275 ptr = PoolAllocateResident(size + RM_PAGE_SIZE, 0);
5276 if (ptr == 0)
5277 {
5278 return (void *) MORECORE_FAILURE;
5279 }
5280 // save ptrs so they can be freed during cleanup
5281 our_os_pools[next_os_pool] = ptr;
5282 next_os_pool++;
5283 ptr = (void *) ((((unsigned long) ptr) + RM_PAGE_MASK) & ~RM_PAGE_MASK);
5284 sbrk_top = (char *) ptr + size;
5285 return ptr;
5286 }
5287 else if (size < 0)
5288 {
5289 // we don't currently support shrink behavior
5290 return (void *) MORECORE_FAILURE;
5291 }
5292 else
5293 {
5294 return sbrk_top;
5295 }
5296 }
5297
5298 // cleanup any allocated memory pools
5299 // called as last thing before shutting down driver
5300
5301 void osCleanupMem(void)
5302 {
5303 void **ptr;
5304
5305 for (ptr = our_os_pools; ptr < &our_os_pools[MAX_POOL_ENTRIES]; ptr++)
5306 if (*ptr)
5307 {
5308 PoolDeallocate(*ptr);
5309 * ptr = 0;
5310 }
5311 }
5312
5313 */
5314
5315
5316/* Helper code. */
5317
5318extern char **__libc_argv attribute_hidden;
5319
5320static void
5321malloc_printerr (const char *str)
5322{
5323 __libc_message (do_abort, "%s\n", str);
5324 __builtin_unreachable ();
5325}
5326
5327/* We need a wrapper function for one of the additions of POSIX. */
5328int
5329__posix_memalign (void **memptr, size_t alignment, size_t size)
5330{
5331 void *mem;
5332
5333 /* Test whether the SIZE argument is valid. It must be a power of
5334 two multiple of sizeof (void *). */
5335 if (alignment % sizeof (void *) != 0
5336 || !powerof2 (alignment / sizeof (void *))
5337 || alignment == 0)
5338 return EINVAL;
5339
5340
5341 void *address = RETURN_ADDRESS (0);
5342 mem = _mid_memalign (alignment, size, address);
5343
5344 if (mem != NULL)
5345 {
5346 *memptr = mem;
5347 return 0;
5348 }
5349
5350 return ENOMEM;
5351}
5352weak_alias (__posix_memalign, posix_memalign)
5353
5354
5355int
5356__malloc_info (int options, FILE *fp)
5357{
5358 /* For now, at least. */
5359 if (options != 0)
5360 return EINVAL;
5361
5362 int n = 0;
5363 size_t total_nblocks = 0;
5364 size_t total_nfastblocks = 0;
5365 size_t total_avail = 0;
5366 size_t total_fastavail = 0;
5367 size_t total_system = 0;
5368 size_t total_max_system = 0;
5369 size_t total_aspace = 0;
5370 size_t total_aspace_mprotect = 0;
5371
5372
5373
5374 if (__malloc_initialized < 0)
5375 ptmalloc_init ();
5376
5377 fputs ("<malloc version=\"1\">\n", fp);
5378
5379 /* Iterate over all arenas currently in use. */
5380 mstate ar_ptr = &main_arena;
5381 do
5382 {
5383 fprintf (fp, "<heap nr=\"%d\">\n<sizes>\n", n++);
5384
5385 size_t nblocks = 0;
5386 size_t nfastblocks = 0;
5387 size_t avail = 0;
5388 size_t fastavail = 0;
5389 struct
5390 {
5391 size_t from;
5392 size_t to;
5393 size_t total;
5394 size_t count;
5395 } sizes[NFASTBINS + NBINS - 1];
5396#define nsizes (sizeof (sizes) / sizeof (sizes[0]))
5397
5398 __libc_lock_lock (ar_ptr->mutex);
5399
5400 for (size_t i = 0; i < NFASTBINS; ++i)
5401 {
5402 mchunkptr p = fastbin (ar_ptr, i);
5403 if (p != NULL)
5404 {
5405 size_t nthissize = 0;
5406 size_t thissize = chunksize (p);
5407
5408 while (p != NULL)
5409 {
5410 ++nthissize;
5411 p = p->fd;
5412 }
5413
5414 fastavail += nthissize * thissize;
5415 nfastblocks += nthissize;
5416 sizes[i].from = thissize - (MALLOC_ALIGNMENT - 1);
5417 sizes[i].to = thissize;
5418 sizes[i].count = nthissize;
5419 }
5420 else
5421 sizes[i].from = sizes[i].to = sizes[i].count = 0;
5422
5423 sizes[i].total = sizes[i].count * sizes[i].to;
5424 }
5425
5426
5427 mbinptr bin;
5428 struct malloc_chunk *r;
5429
5430 for (size_t i = 1; i < NBINS; ++i)
5431 {
5432 bin = bin_at (ar_ptr, i);
5433 r = bin->fd;
5434 sizes[NFASTBINS - 1 + i].from = ~((size_t) 0);
5435 sizes[NFASTBINS - 1 + i].to = sizes[NFASTBINS - 1 + i].total
5436 = sizes[NFASTBINS - 1 + i].count = 0;
5437
5438 if (r != NULL)
5439 while (r != bin)
5440 {
5441 size_t r_size = chunksize_nomask (r);
5442 ++sizes[NFASTBINS - 1 + i].count;
5443 sizes[NFASTBINS - 1 + i].total += r_size;
5444 sizes[NFASTBINS - 1 + i].from
5445 = MIN (sizes[NFASTBINS - 1 + i].from, r_size);
5446 sizes[NFASTBINS - 1 + i].to = MAX (sizes[NFASTBINS - 1 + i].to,
5447 r_size);
5448
5449 r = r->fd;
5450 }
5451
5452 if (sizes[NFASTBINS - 1 + i].count == 0)
5453 sizes[NFASTBINS - 1 + i].from = 0;
5454 nblocks += sizes[NFASTBINS - 1 + i].count;
5455 avail += sizes[NFASTBINS - 1 + i].total;
5456 }
5457
5458 size_t heap_size = 0;
5459 size_t heap_mprotect_size = 0;
5460 size_t heap_count = 0;
5461 if (ar_ptr != &main_arena)
5462 {
5463 /* Iterate over the arena heaps from back to front. */
5464 heap_info *heap = heap_for_ptr (top (ar_ptr));
5465 do
5466 {
5467 heap_size += heap->size;
5468 heap_mprotect_size += heap->mprotect_size;
5469 heap = heap->prev;
5470 ++heap_count;
5471 }
5472 while (heap != NULL);
5473 }
5474
5475 __libc_lock_unlock (ar_ptr->mutex);
5476
5477 total_nfastblocks += nfastblocks;
5478 total_fastavail += fastavail;
5479
5480 total_nblocks += nblocks;
5481 total_avail += avail;
5482
5483 for (size_t i = 0; i < nsizes; ++i)
5484 if (sizes[i].count != 0 && i != NFASTBINS)
5485 fprintf (fp, " \
5486 <size from=\"%zu\" to=\"%zu\" total=\"%zu\" count=\"%zu\"/>\n",
5487 sizes[i].from, sizes[i].to, sizes[i].total, sizes[i].count);
5488
5489 if (sizes[NFASTBINS].count != 0)
5490 fprintf (fp, "\
5491 <unsorted from=\"%zu\" to=\"%zu\" total=\"%zu\" count=\"%zu\"/>\n",
5492 sizes[NFASTBINS].from, sizes[NFASTBINS].to,
5493 sizes[NFASTBINS].total, sizes[NFASTBINS].count);
5494
5495 total_system += ar_ptr->system_mem;
5496 total_max_system += ar_ptr->max_system_mem;
5497
5498 fprintf (fp,
5499 "</sizes>\n<total type=\"fast\" count=\"%zu\" size=\"%zu\"/>\n"
5500 "<total type=\"rest\" count=\"%zu\" size=\"%zu\"/>\n"
5501 "<system type=\"current\" size=\"%zu\"/>\n"
5502 "<system type=\"max\" size=\"%zu\"/>\n",
5503 nfastblocks, fastavail, nblocks, avail,
5504 ar_ptr->system_mem, ar_ptr->max_system_mem);
5505
5506 if (ar_ptr != &main_arena)
5507 {
5508 fprintf (fp,
5509 "<aspace type=\"total\" size=\"%zu\"/>\n"
5510 "<aspace type=\"mprotect\" size=\"%zu\"/>\n"
5511 "<aspace type=\"subheaps\" size=\"%zu\"/>\n",
5512 heap_size, heap_mprotect_size, heap_count);
5513 total_aspace += heap_size;
5514 total_aspace_mprotect += heap_mprotect_size;
5515 }
5516 else
5517 {
5518 fprintf (fp,
5519 "<aspace type=\"total\" size=\"%zu\"/>\n"
5520 "<aspace type=\"mprotect\" size=\"%zu\"/>\n",
5521 ar_ptr->system_mem, ar_ptr->system_mem);
5522 total_aspace += ar_ptr->system_mem;
5523 total_aspace_mprotect += ar_ptr->system_mem;
5524 }
5525
5526 fputs ("</heap>\n", fp);
5527 ar_ptr = ar_ptr->next;
5528 }
5529 while (ar_ptr != &main_arena);
5530
5531 fprintf (fp,
5532 "<total type=\"fast\" count=\"%zu\" size=\"%zu\"/>\n"
5533 "<total type=\"rest\" count=\"%zu\" size=\"%zu\"/>\n"
5534 "<total type=\"mmap\" count=\"%d\" size=\"%zu\"/>\n"
5535 "<system type=\"current\" size=\"%zu\"/>\n"
5536 "<system type=\"max\" size=\"%zu\"/>\n"
5537 "<aspace type=\"total\" size=\"%zu\"/>\n"
5538 "<aspace type=\"mprotect\" size=\"%zu\"/>\n"
5539 "</malloc>\n",
5540 total_nfastblocks, total_fastavail, total_nblocks, total_avail,
5541 mp_.n_mmaps, mp_.mmapped_mem,
5542 total_system, total_max_system,
5543 total_aspace, total_aspace_mprotect);
5544
5545 return 0;
5546}
5547weak_alias (__malloc_info, malloc_info)
5548
5549
5550strong_alias (__libc_calloc, __calloc) weak_alias (__libc_calloc, calloc)
5551strong_alias (__libc_free, __free) strong_alias (__libc_free, free)
5552strong_alias (__libc_malloc, __malloc) strong_alias (__libc_malloc, malloc)
5553strong_alias (__libc_memalign, __memalign)
5554weak_alias (__libc_memalign, memalign)
5555strong_alias (__libc_realloc, __realloc) strong_alias (__libc_realloc, realloc)
5556strong_alias (__libc_valloc, __valloc) weak_alias (__libc_valloc, valloc)
5557strong_alias (__libc_pvalloc, __pvalloc) weak_alias (__libc_pvalloc, pvalloc)
5558strong_alias (__libc_mallinfo, __mallinfo)
5559weak_alias (__libc_mallinfo, mallinfo)
5560strong_alias (__libc_mallopt, __mallopt<